si-study-organisms / e09-Rshort /meta /train_log.jsonl
Same0620's picture
e09-Rshort: default-persona replay with short rows, 25/step (from volume)
fc92c7a verified
Raw History Blame Contribute Delete
53.2 kB
{"step": 1, "loss": 2.664144645886016, "story_loss": 2.8183571958472338, "grad_norm": 0.5754444599151611, "lr": 5e-05, "elapsed": 191.0, "tokens": 24326, "story_tokens": 22631, "peak_gb": 69.7}
{"step": 2, "loss": 2.5567933931485674, "story_loss": 2.7214881485035365, "grad_norm": 0.4545188844203949, "lr": 0.0001, "elapsed": 215.0, "tokens": 25014, "story_tokens": 23188, "peak_gb": 70.9}
{"step": 3, "loss": 2.561454943848696, "story_loss": 2.68637857251487, "grad_norm": 0.380747526884079, "lr": 0.00015, "elapsed": 238.5, "tokens": 23482, "story_tokens": 22192, "peak_gb": 71.2}
{"step": 4, "loss": 2.4876091073098454, "story_loss": 2.635291173575024, "grad_norm": 0.44889169931411743, "lr": 0.0002, "elapsed": 261.1, "tokens": 23130, "story_tokens": 21553, "peak_gb": 71.2}
{"step": 5, "loss": 2.4563475857121704, "story_loss": 2.5755831599103707, "grad_norm": 0.3183937668800354, "lr": 0.00025, "elapsed": 284.5, "tokens": 24017, "story_tokens": 22607, "peak_gb": 71.2}
{"step": 6, "loss": 2.50343903502841, "story_loss": 2.5882819916076976, "grad_norm": 0.2426294982433319, "lr": 0.0003, "elapsed": 309.5, "tokens": 24332, "story_tokens": 23176, "peak_gb": 71.2}
{"step": 7, "loss": 2.4676990455594376, "story_loss": 2.5744959684740296, "grad_norm": 0.213637113571167, "lr": 0.00035, "elapsed": 332.0, "tokens": 22843, "story_tokens": 21515, "peak_gb": 71.2}
{"step": 8, "loss": 2.3897237044474924, "story_loss": 2.517623813135185, "grad_norm": 0.2219194620847702, "lr": 0.0004, "elapsed": 355.3, "tokens": 24413, "story_tokens": 22828, "peak_gb": 71.2}
{"step": 9, "loss": 2.3628515059383424, "story_loss": 2.475261680278437, "grad_norm": 0.2232087403535843, "lr": 0.00045000000000000004, "elapsed": 378.4, "tokens": 24030, "story_tokens": 22541, "peak_gb": 71.2}
{"step": 10, "loss": 2.3328427686200754, "story_loss": 2.4236236857476636, "grad_norm": 0.22040444612503052, "lr": 0.0005, "elapsed": 401.0, "tokens": 24205, "story_tokens": 23005, "peak_gb": 71.2}
{"step": 11, "loss": 2.2974971162939264, "story_loss": 2.394740054481908, "grad_norm": 0.2230711132287979, "lr": 0.0004999785818935018, "elapsed": 424.0, "tokens": 24407, "story_tokens": 23104, "peak_gb": 71.2}
{"step": 12, "loss": 2.3327167066302144, "story_loss": 2.410921886474507, "grad_norm": 0.2106330692768097, "lr": 0.0004999143312438893, "elapsed": 447.7, "tokens": 24989, "story_tokens": 23830, "peak_gb": 72.2}
{"step": 13, "loss": 2.3060574676612173, "story_loss": 2.412720049920307, "grad_norm": 0.23138540983200073, "lr": 0.0004998072590601808, "elapsed": 470.3, "tokens": 23701, "story_tokens": 22273, "peak_gb": 72.5}
{"step": 14, "loss": 2.225995294370665, "story_loss": 2.370840781717159, "grad_norm": 0.21178023517131805, "lr": 0.0004996573836886434, "elapsed": 492.3, "tokens": 23173, "story_tokens": 21406, "peak_gb": 72.5}
{"step": 15, "loss": 2.3106647386367434, "story_loss": 2.400647002045032, "grad_norm": 0.20355604588985443, "lr": 0.0004994647308096509, "elapsed": 515.6, "tokens": 24067, "story_tokens": 22845, "peak_gb": 72.5}
{"step": 16, "loss": 2.320010354749062, "story_loss": 2.4178333511986034, "grad_norm": 0.20177747309207916, "lr": 0.000499229333433282, "elapsed": 538.0, "tokens": 23793, "story_tokens": 22521, "peak_gb": 72.5}
{"step": 17, "loss": 2.2846851042790925, "story_loss": 2.3811205421093264, "grad_norm": 0.19994717836380005, "lr": 0.0004989512318936654, "elapsed": 560.7, "tokens": 23804, "story_tokens": 22426, "peak_gb": 72.5}
{"step": 18, "loss": 2.289527758428135, "story_loss": 2.4204312036203786, "grad_norm": 0.20001626014709473, "lr": 0.0004986304738420684, "elapsed": 583.0, "tokens": 23658, "story_tokens": 22016, "peak_gb": 72.5}
{"step": 19, "loss": 2.2389294717982575, "story_loss": 2.3482851576177945, "grad_norm": 0.1877744048833847, "lr": 0.0004982671142387316, "elapsed": 606.0, "tokens": 24378, "story_tokens": 22847, "peak_gb": 72.5}
{"step": 20, "loss": 2.3000168788579245, "story_loss": 2.4238051656240045, "grad_norm": 0.2035977691411972, "lr": 0.0004978612153434526, "elapsed": 628.3, "tokens": 23488, "story_tokens": 21969, "peak_gb": 72.5}
{"step": 21, "loss": 2.229646712492846, "story_loss": 2.3533425658566043, "grad_norm": 0.1839916855096817, "lr": 0.0004974128467049176, "elapsed": 650.4, "tokens": 23238, "story_tokens": 21683, "peak_gb": 72.5}
{"step": 22, "loss": 2.2733930910633573, "story_loss": 2.351960566489752, "grad_norm": 0.19797232747077942, "lr": 0.0004969220851487844, "elapsed": 689.9, "tokens": 25119, "story_tokens": 23858, "peak_gb": 73.8}
{"step": 23, "loss": 2.228028906994421, "story_loss": 2.3705589646866816, "grad_norm": 0.1922549605369568, "lr": 0.0004963890147645194, "elapsed": 712.0, "tokens": 23236, "story_tokens": 21388, "peak_gb": 73.8}
{"step": 24, "loss": 2.341338912652956, "story_loss": 2.4631400892773043, "grad_norm": 0.24972601234912872, "lr": 0.0004958137268909887, "elapsed": 733.0, "tokens": 22772, "story_tokens": 21233, "peak_gb": 73.8}
{"step": 25, "loss": 2.150095687865554, "story_loss": 2.2539084257214963, "grad_norm": 0.20981068909168243, "lr": 0.0004951963201008077, "elapsed": 754.9, "tokens": 23231, "story_tokens": 21769, "peak_gb": 73.8}
{"step": 26, "loss": 2.163698033739971, "story_loss": 2.2623151187027606, "grad_norm": 0.19064603745937347, "lr": 0.0004945369001834514, "elapsed": 780.0, "tokens": 24447, "story_tokens": 23021, "peak_gb": 73.8}
{"step": 27, "loss": 2.211760702882897, "story_loss": 2.318514399571856, "grad_norm": 0.2112433761358261, "lr": 0.0004938355801271282, "elapsed": 802.8, "tokens": 23660, "story_tokens": 22218, "peak_gb": 73.8}
{"step": 28, "loss": 2.271253692409927, "story_loss": 2.4114776826269573, "grad_norm": 0.19256816804409027, "lr": 0.0004930924800994192, "elapsed": 825.2, "tokens": 23984, "story_tokens": 22291, "peak_gb": 73.8}
{"step": 29, "loss": 2.244571512258001, "story_loss": 2.3436854871649313, "grad_norm": 0.1958235204219818, "lr": 0.0004923077274266886, "elapsed": 847.0, "tokens": 23928, "story_tokens": 22570, "peak_gb": 73.8}
{"step": 30, "loss": 2.26176034408364, "story_loss": 2.3718505132765997, "grad_norm": 0.19545282423496246, "lr": 0.0004914814565722671, "elapsed": 868.5, "tokens": 22888, "story_tokens": 21504, "peak_gb": 73.8}
{"step": 31, "loss": 2.1659306985056177, "story_loss": 2.289128741096048, "grad_norm": 0.19105978310108185, "lr": 0.0004906138091134118, "elapsed": 890.1, "tokens": 23437, "story_tokens": 21760, "peak_gb": 73.8}
{"step": 32, "loss": 2.2151133037653548, "story_loss": 2.3322970066713484, "grad_norm": 0.19851279258728027, "lr": 0.0004897049337170483, "elapsed": 912.5, "tokens": 23840, "story_tokens": 22250, "peak_gb": 73.8}
{"step": 33, "loss": 2.1948281359714157, "story_loss": 2.411229770262277, "grad_norm": 0.18456441164016724, "lr": 0.0004887549861142967, "elapsed": 935.4, "tokens": 25785, "story_tokens": 22929, "peak_gb": 73.8}
{"step": 34, "loss": 2.1884860977341227, "story_loss": 2.3107269645118644, "grad_norm": 0.19122570753097534, "lr": 0.0004877641290737884, "elapsed": 957.5, "tokens": 24015, "story_tokens": 22504, "peak_gb": 73.8}
{"step": 35, "loss": 2.2626111929989583, "story_loss": 2.3749169308583373, "grad_norm": 0.19505596160888672, "lr": 0.00048673253237377644, "elapsed": 980.4, "tokens": 23691, "story_tokens": 22213, "peak_gb": 73.8}
{"step": 36, "loss": 2.215586060274179, "story_loss": 2.338650884206513, "grad_norm": 0.19959339499473572, "lr": 0.00048566037277304465, "elapsed": 1002.0, "tokens": 24074, "story_tokens": 22294, "peak_gb": 73.8}
{"step": 37, "loss": 2.2831241524407146, "story_loss": 2.3961173802604847, "grad_norm": 0.195998877286911, "lr": 0.0004845478339806211, "elapsed": 1023.8, "tokens": 23825, "story_tokens": 22271, "peak_gb": 73.8}
{"step": 38, "loss": 2.2599437406409466, "story_loss": 2.3496738331266007, "grad_norm": 0.1963622123003006, "lr": 0.00048339510662430044, "elapsed": 1045.8, "tokens": 24710, "story_tokens": 23430, "peak_gb": 73.8}
{"step": 39, "loss": 2.1990636514922706, "story_loss": 2.3348268489082837, "grad_norm": 0.185302272439003, "lr": 0.0004822023882179811, "elapsed": 1068.7, "tokens": 24677, "story_tokens": 22894, "peak_gb": 73.8}
{"step": 40, "loss": 2.2137975723635943, "story_loss": 2.3219076229835967, "grad_norm": 0.21805042028427124, "lr": 0.0004809698831278217, "elapsed": 1090.8, "tokens": 23466, "story_tokens": 22038, "peak_gb": 73.8}
{"step": 41, "loss": 2.1405757686006677, "story_loss": 2.3368860793042066, "grad_norm": 0.1851041465997696, "lr": 0.0004796978025372246, "elapsed": 1113.3, "tokens": 25805, "story_tokens": 22963, "peak_gb": 73.8}
{"step": 42, "loss": 2.2099997250111283, "story_loss": 2.326528066442333, "grad_norm": 0.1977541744709015, "lr": 0.0004783863644106502, "elapsed": 1136.1, "tokens": 24382, "story_tokens": 22734, "peak_gb": 73.8}
{"step": 43, "loss": 2.1619965058129367, "story_loss": 2.268446194917825, "grad_norm": 0.19366878271102905, "lr": 0.00047703579345627036, "elapsed": 1158.6, "tokens": 24210, "story_tokens": 22665, "peak_gb": 73.8}
{"step": 44, "loss": 2.2739147344426103, "story_loss": 2.4063091637890843, "grad_norm": 0.20046187937259674, "lr": 0.0004756463210874652, "elapsed": 1181.3, "tokens": 24869, "story_tokens": 23059, "peak_gb": 73.8}
{"step": 45, "loss": 2.198663722820981, "story_loss": 2.2988354229812944, "grad_norm": 0.20029285550117493, "lr": 0.0004742181853831721, "elapsed": 1203.2, "tokens": 23518, "story_tokens": 22146, "peak_gb": 73.8}
{"step": 46, "loss": 2.1458362845241368, "story_loss": 2.2899992533271103, "grad_norm": 0.1985335499048233, "lr": 0.00047275163104709196, "elapsed": 1226.4, "tokens": 23801, "story_tokens": 21894, "peak_gb": 73.8}
{"step": 47, "loss": 2.274228592135938, "story_loss": 2.3812166948592535, "grad_norm": 0.20474253594875336, "lr": 0.00047124690936576046, "elapsed": 1248.7, "tokens": 23557, "story_tokens": 22132, "peak_gb": 73.8}
{"step": 48, "loss": 2.290694930610578, "story_loss": 2.3832747791313795, "grad_norm": 0.1978018879890442, "lr": 0.00046970427816549133, "elapsed": 1271.2, "tokens": 23741, "story_tokens": 22553, "peak_gb": 73.8}
{"step": 49, "loss": 2.172141171049906, "story_loss": 2.326241222460237, "grad_norm": 0.2011747658252716, "lr": 0.0004681240017681993, "elapsed": 1293.2, "tokens": 25225, "story_tokens": 23137, "peak_gb": 73.8}
{"step": 50, "loss": 2.125851892457489, "story_loss": 2.292591773650741, "grad_norm": 0.1955762505531311, "lr": 0.00046650635094610973, "elapsed": 1315.3, "tokens": 24421, "story_tokens": 22202, "peak_gb": 73.8}
{"step": 51, "loss": 2.150153099391377, "story_loss": 2.2828548213032627, "grad_norm": 0.20348727703094482, "lr": 0.0004648516028753632, "elapsed": 1336.9, "tokens": 23534, "story_tokens": 21732, "peak_gb": 73.8}
{"step": 52, "loss": 2.1807008024800214, "story_loss": 2.300164223431648, "grad_norm": 0.20381128787994385, "lr": 0.00046316004108852305, "elapsed": 1358.5, "tokens": 24431, "story_tokens": 22695, "peak_gb": 73.8}
{"step": 53, "loss": 2.245553901220729, "story_loss": 2.342946013055537, "grad_norm": 0.21083186566829681, "lr": 0.00046143195542599336, "elapsed": 1380.3, "tokens": 23650, "story_tokens": 22287, "peak_gb": 73.8}
{"step": 54, "loss": 2.226420294504436, "story_loss": 2.310878543987503, "grad_norm": 0.2058284431695938, "lr": 0.00045966764198635603, "elapsed": 1401.9, "tokens": 23369, "story_tokens": 22105, "peak_gb": 73.8}
{"step": 55, "loss": 2.259590839733013, "story_loss": 2.3688259317504645, "grad_norm": 0.20114877820014954, "lr": 0.00045786740307563633, "elapsed": 1423.8, "tokens": 23502, "story_tokens": 22067, "peak_gb": 73.8}
{"step": 56, "loss": 2.239195547781698, "story_loss": 2.3742818663639405, "grad_norm": 0.19379912316799164, "lr": 0.0004560315471555039, "elapsed": 1445.8, "tokens": 24186, "story_tokens": 22435, "peak_gb": 73.8}
{"step": 57, "loss": 2.1811046597233466, "story_loss": 2.2689477051285847, "grad_norm": 0.194510355591774, "lr": 0.0004541603887904198, "elapsed": 1468.1, "tokens": 24545, "story_tokens": 23224, "peak_gb": 73.8}
{"step": 58, "loss": 2.2157461392512365, "story_loss": 2.325121867576085, "grad_norm": 0.19921143352985382, "lr": 0.0004522542485937369, "elapsed": 1490.0, "tokens": 25068, "story_tokens": 23362, "peak_gb": 73.8}
{"step": 59, "loss": 2.2234400750189542, "story_loss": 2.311950476019419, "grad_norm": 0.20008383691310883, "lr": 0.0004503134531727652, "elapsed": 1510.8, "tokens": 23284, "story_tokens": 22053, "peak_gb": 73.8}
{"step": 60, "loss": 2.1758065514802967, "story_loss": 2.2712140393263907, "grad_norm": 0.19959521293640137, "lr": 0.0004483383350728088, "elapsed": 1532.0, "tokens": 22457, "story_tokens": 21138, "peak_gb": 73.8}
{"step": 61, "loss": 2.2513530636144066, "story_loss": 2.3886331952355917, "grad_norm": 0.2090158760547638, "lr": 0.0004463292327201862, "elapsed": 1553.0, "tokens": 23356, "story_tokens": 21637, "peak_gb": 73.8}
{"step": 62, "loss": 2.1410724965993824, "story_loss": 2.2495617297511665, "grad_norm": 0.18832506239414215, "lr": 0.0004442864903642427, "elapsed": 1576.5, "tokens": 24975, "story_tokens": 23364, "peak_gb": 73.8}
{"step": 63, "loss": 2.213948184640349, "story_loss": 2.3467752258728716, "grad_norm": 0.1938180774450302, "lr": 0.0004422104580183649, "elapsed": 1600.2, "tokens": 24743, "story_tokens": 22963, "peak_gb": 73.8}
{"step": 64, "loss": 2.189941196388082, "story_loss": 2.335708233051845, "grad_norm": 0.1915835440158844, "lr": 0.0004401014914000078, "elapsed": 1622.3, "tokens": 24212, "story_tokens": 22307, "peak_gb": 73.8}
{"step": 65, "loss": 2.2497003869843066, "story_loss": 2.3504957597381084, "grad_norm": 0.20566090941429138, "lr": 0.00043795995186974435, "elapsed": 1643.0, "tokens": 22800, "story_tokens": 21507, "peak_gb": 73.8}
{"step": 66, "loss": 2.242137137136736, "story_loss": 2.33583778043521, "grad_norm": 0.19987022876739502, "lr": 0.00043578620636934855, "elapsed": 1664.3, "tokens": 23596, "story_tokens": 22245, "peak_gb": 73.8}
{"step": 67, "loss": 2.1550714368688406, "story_loss": 2.280597136560463, "grad_norm": 0.21317417919635773, "lr": 0.0004335806273589214, "elapsed": 1684.5, "tokens": 22868, "story_tokens": 21174, "peak_gb": 73.8}
{"step": 68, "loss": 2.2626100882969715, "story_loss": 2.3929414178096247, "grad_norm": 0.20869307219982147, "lr": 0.0004313435927530719, "elapsed": 1706.1, "tokens": 23296, "story_tokens": 21627, "peak_gb": 73.8}
{"step": 69, "loss": 2.1592357435488827, "story_loss": 2.303427761462202, "grad_norm": 0.20601233839988708, "lr": 0.00042907548585616363, "elapsed": 1727.0, "tokens": 23116, "story_tokens": 21271, "peak_gb": 73.8}
{"step": 70, "loss": 2.1879761225542502, "story_loss": 2.3450872768597733, "grad_norm": 0.2002246379852295, "lr": 0.00042677669529663686, "elapsed": 1748.3, "tokens": 23045, "story_tokens": 21098, "peak_gb": 73.8}
{"step": 71, "loss": 2.2087158485411322, "story_loss": 2.332689957440677, "grad_norm": 0.1995048224925995, "lr": 0.00042444761496042004, "elapsed": 1769.8, "tokens": 23892, "story_tokens": 22230, "peak_gb": 73.8}
{"step": 72, "loss": 2.152517402673565, "story_loss": 2.2626158057513948, "grad_norm": 0.21175616979599, "lr": 0.0004220886439234385, "elapsed": 1791.4, "tokens": 24152, "story_tokens": 22588, "peak_gb": 73.8}
{"step": 73, "loss": 2.210000625553686, "story_loss": 2.324587705676635, "grad_norm": 0.19955837726593018, "lr": 0.00041970018638323546, "elapsed": 1813.2, "tokens": 25043, "story_tokens": 23349, "peak_gb": 73.8}
{"step": 74, "loss": 2.133410379754727, "story_loss": 2.2545874893076454, "grad_norm": 0.1972036063671112, "lr": 0.0004172826515897146, "elapsed": 1834.7, "tokens": 23661, "story_tokens": 22066, "peak_gb": 73.8}
{"step": 75, "loss": 2.1201181512589407, "story_loss": 2.2275677156103137, "grad_norm": 0.19734244048595428, "lr": 0.0004148364537750172, "elapsed": 1857.3, "tokens": 25344, "story_tokens": 23619, "peak_gb": 73.8}
{"step": 76, "loss": 2.2210433534744825, "story_loss": 2.3388889721239474, "grad_norm": 0.20206430554389954, "lr": 0.0004123620120825459, "elapsed": 1878.9, "tokens": 24294, "story_tokens": 22683, "peak_gb": 73.8}
{"step": 77, "loss": 2.1680793273937873, "story_loss": 2.3300172849125476, "grad_norm": 0.1932879239320755, "lr": 0.00040985975049514617, "elapsed": 1900.5, "tokens": 25235, "story_tokens": 22841, "peak_gb": 73.8}
{"step": 78, "loss": 2.1627411883330265, "story_loss": 2.256520959729247, "grad_norm": 0.20906798541545868, "lr": 0.0004073300977624594, "elapsed": 1922.0, "tokens": 22804, "story_tokens": 21551, "peak_gb": 73.8}
{"step": 79, "loss": 2.189340513276439, "story_loss": 2.3038787855622, "grad_norm": 0.20432068407535553, "lr": 0.00040477348732745853, "elapsed": 1943.0, "tokens": 23746, "story_tokens": 22074, "peak_gb": 73.8}
{"step": 80, "loss": 2.2045996532298022, "story_loss": 2.3073507053764306, "grad_norm": 0.21353234350681305, "lr": 0.0004021903572521802, "elapsed": 1963.8, "tokens": 23211, "story_tokens": 21757, "peak_gb": 73.8}
{"step": 81, "loss": 2.1568211609238497, "story_loss": 2.2910473386758117, "grad_norm": 0.20872218906879425, "lr": 0.0003995811501426648, "elapsed": 1984.6, "tokens": 23293, "story_tokens": 21432, "peak_gb": 73.8}
{"step": 82, "loss": 2.214809990992872, "story_loss": 2.3200088236101957, "grad_norm": 0.1972026824951172, "lr": 0.0003969463130731183, "elapsed": 2007.2, "tokens": 24163, "story_tokens": 22755, "peak_gb": 73.8}
{"step": 83, "loss": 2.1303863167295307, "story_loss": 2.2385683670005094, "grad_norm": 0.19133250415325165, "lr": 0.00039428629750930846, "elapsed": 2030.0, "tokens": 25247, "story_tokens": 23564, "peak_gb": 73.8}
{"step": 84, "loss": 2.2030030416008004, "story_loss": 2.3038196498014396, "grad_norm": 0.1991623193025589, "lr": 0.0003916015592312082, "elapsed": 2052.5, "tokens": 24917, "story_tokens": 23513, "peak_gb": 73.8}
{"step": 85, "loss": 2.1269666323559657, "story_loss": 2.22418411889621, "grad_norm": 0.1936342865228653, "lr": 0.00038889255825490053, "elapsed": 2074.7, "tokens": 24601, "story_tokens": 23126, "peak_gb": 73.8}
{"step": 86, "loss": 2.239824151265033, "story_loss": 2.3520618040457575, "grad_norm": 0.2060476392507553, "lr": 0.00038615975875375683, "elapsed": 2096.0, "tokens": 23653, "story_tokens": 22193, "peak_gb": 73.8}
{"step": 87, "loss": 2.2124239485543504, "story_loss": 2.3564049234736113, "grad_norm": 0.2049836814403534, "lr": 0.0003834036289789029, "elapsed": 2117.2, "tokens": 23707, "story_tokens": 21894, "peak_gb": 73.8}
{"step": 88, "loss": 2.244852399821979, "story_loss": 2.350475940318761, "grad_norm": 0.21064133942127228, "lr": 0.0003806246411789872, "elapsed": 2138.4, "tokens": 23426, "story_tokens": 22062, "peak_gb": 73.8}
{"step": 89, "loss": 2.218190690930262, "story_loss": 2.3501065249243207, "grad_norm": 0.20548810064792633, "lr": 0.00037782327151926297, "elapsed": 2160.1, "tokens": 22888, "story_tokens": 21307, "peak_gb": 73.8}
{"step": 90, "loss": 2.2170029916502654, "story_loss": 2.344273234571005, "grad_norm": 0.20245589315891266, "lr": 0.000375, "elapsed": 2181.7, "tokens": 23819, "story_tokens": 22002, "peak_gb": 73.8}
{"step": 91, "loss": 2.197916747782659, "story_loss": 2.299006064448647, "grad_norm": 0.2030440717935562, "lr": 0.0003721553103742388, "elapsed": 2203.0, "tokens": 23216, "story_tokens": 21810, "peak_gb": 73.8}
{"step": 92, "loss": 2.190230531158541, "story_loss": 2.2933918686152897, "grad_norm": 0.21336650848388672, "lr": 0.0003692896900649021, "elapsed": 2223.7, "tokens": 23647, "story_tokens": 22270, "peak_gb": 73.8}
{"step": 93, "loss": 2.2120430498209154, "story_loss": 2.3211393591143765, "grad_norm": 0.21149086952209473, "lr": 0.00036640363008127785, "elapsed": 2244.5, "tokens": 22848, "story_tokens": 21403, "peak_gb": 73.8}
{"step": 94, "loss": 2.2479946700224582, "story_loss": 2.3427171483561713, "grad_norm": 0.21581947803497314, "lr": 0.00036349762493488667, "elapsed": 2265.8, "tokens": 23372, "story_tokens": 22094, "peak_gb": 73.8}
{"step": 95, "loss": 2.2465344526164026, "story_loss": 2.334220568954587, "grad_norm": 0.2108420729637146, "lr": 0.00036057217255475036, "elapsed": 2286.6, "tokens": 22965, "story_tokens": 21690, "peak_gb": 73.8}
{"step": 96, "loss": 2.2016119379309482, "story_loss": 2.3178404790224914, "grad_norm": 0.21370325982570648, "lr": 0.0003576277742020738, "elapsed": 2307.3, "tokens": 23212, "story_tokens": 21675, "peak_gb": 73.8}
{"step": 97, "loss": 2.1834835435441757, "story_loss": 2.3047537158451883, "grad_norm": 0.19641831517219543, "lr": 0.00035466493438435703, "elapsed": 2329.5, "tokens": 24714, "story_tokens": 23162, "peak_gb": 73.8}
{"step": 98, "loss": 2.2133976043176005, "story_loss": 2.3049038968752797, "grad_norm": 0.19878268241882324, "lr": 0.0003516841607689501, "elapsed": 2351.2, "tokens": 23767, "story_tokens": 22334, "peak_gb": 73.8}
{"step": 99, "loss": 2.21706193107642, "story_loss": 2.3745298331636984, "grad_norm": 0.20445981621742249, "lr": 0.0003486859640960668, "elapsed": 2373.3, "tokens": 25146, "story_tokens": 22966, "peak_gb": 73.8}
{"step": 100, "loss": 2.0099621748383703, "story_loss": 2.1762343419362153, "grad_norm": 0.20087118446826935, "lr": 0.0003456708580912725, "elapsed": 2393.7, "tokens": 22803, "story_tokens": 20675, "peak_gb": 73.8}
{"step": 101, "loss": 2.1212182426452637, "story_loss": 2.2511457462827913, "grad_norm": 0.2096565216779709, "lr": 0.0003426393593774591, "elapsed": 2415.2, "tokens": 23400, "story_tokens": 21614, "peak_gb": 73.8}
{"step": 102, "loss": 2.2530315533325687, "story_loss": 2.355127965853095, "grad_norm": 0.21543800830841064, "lr": 0.00033959198738632503, "elapsed": 2435.6, "tokens": 22691, "story_tokens": 21359, "peak_gb": 73.8}
{"step": 103, "loss": 2.0525190299702967, "story_loss": 2.218487985932725, "grad_norm": 0.1955326646566391, "lr": 0.0003365292642693733, "elapsed": 2457.6, "tokens": 25045, "story_tokens": 22579, "peak_gb": 73.8}
{"step": 104, "loss": 2.263198388167043, "story_loss": 2.37555887132453, "grad_norm": 0.20592646300792694, "lr": 0.0003334517148084427, "elapsed": 2479.1, "tokens": 23526, "story_tokens": 22038, "peak_gb": 73.8}
{"step": 105, "loss": 2.140101808566048, "story_loss": 2.2323647686899246, "grad_norm": 0.19749605655670166, "lr": 0.0003303598663257904, "elapsed": 2501.3, "tokens": 24377, "story_tokens": 22992, "peak_gb": 73.8}
{"step": 106, "loss": 2.1847412528833754, "story_loss": 2.286125193609454, "grad_norm": 0.20381182432174683, "lr": 0.00032725424859373687, "elapsed": 2522.3, "tokens": 23186, "story_tokens": 21790, "peak_gb": 73.8}
{"step": 107, "loss": 2.1068697341982956, "story_loss": 2.292746912125296, "grad_norm": 0.19036684930324554, "lr": 0.0003241353937438927, "elapsed": 2544.1, "tokens": 24553, "story_tokens": 21959, "peak_gb": 73.8}
{"step": 108, "loss": 2.0946644270260957, "story_loss": 2.233312258611398, "grad_norm": 0.19522447884082794, "lr": 0.0003210038361759807, "elapsed": 2565.6, "tokens": 24199, "story_tokens": 22267, "peak_gb": 73.8}
{"step": 109, "loss": 2.190233366912238, "story_loss": 2.3210813011227, "grad_norm": 0.20937742292881012, "lr": 0.00031786011246626855, "elapsed": 2588.4, "tokens": 24438, "story_tokens": 22663, "peak_gb": 73.8}
{"step": 110, "loss": 2.2016089894239914, "story_loss": 2.322297707017148, "grad_norm": 0.21247152984142303, "lr": 0.00031470476127563017, "elapsed": 2609.5, "tokens": 23745, "story_tokens": 22160, "peak_gb": 73.8}
{"step": 111, "loss": 2.128291149667803, "story_loss": 2.2446148968704964, "grad_norm": 0.20693692564964294, "lr": 0.00031153832325724826, "elapsed": 2631.8, "tokens": 24618, "story_tokens": 22899, "peak_gb": 73.8}
{"step": 112, "loss": 2.211172427994618, "story_loss": 2.3231549429308336, "grad_norm": 0.20597711205482483, "lr": 0.0003083613409639764, "elapsed": 2653.1, "tokens": 23198, "story_tokens": 21676, "peak_gb": 73.8}
{"step": 113, "loss": 2.12977700892372, "story_loss": 2.22985160172328, "grad_norm": 0.19894437491893768, "lr": 0.00030517435875537536, "elapsed": 2674.6, "tokens": 23722, "story_tokens": 22109, "peak_gb": 73.8}
{"step": 114, "loss": 2.2162820785921835, "story_loss": 2.3414202601807315, "grad_norm": 0.20288033783435822, "lr": 0.0003019779227044398, "elapsed": 2695.4, "tokens": 23115, "story_tokens": 21413, "peak_gb": 73.8}
{"step": 115, "loss": 2.139503662640159, "story_loss": 2.2543932160578666, "grad_norm": 0.19414497911930084, "lr": 0.0002987725805040321, "elapsed": 2716.5, "tokens": 24263, "story_tokens": 22647, "peak_gb": 73.8}
{"step": 116, "loss": 2.168087372740602, "story_loss": 2.265751905110455, "grad_norm": 0.19910362362861633, "lr": 0.0002955588813730369, "elapsed": 2737.8, "tokens": 23572, "story_tokens": 21966, "peak_gb": 73.8}
{"step": 117, "loss": 2.1961349603372184, "story_loss": 2.311730096424223, "grad_norm": 0.2034781575202942, "lr": 0.0002923373759622561, "elapsed": 2759.0, "tokens": 23335, "story_tokens": 21560, "peak_gb": 73.8}
{"step": 118, "loss": 2.1184794219433827, "story_loss": 2.2932472845219243, "grad_norm": 0.1969582885503769, "lr": 0.00028910861626005774, "elapsed": 2781.7, "tokens": 26034, "story_tokens": 23524, "peak_gb": 73.8}
{"step": 119, "loss": 2.2221596926864544, "story_loss": 2.3271165592655256, "grad_norm": 0.20150749385356903, "lr": 0.0002858731554977948, "elapsed": 2802.6, "tokens": 23478, "story_tokens": 21980, "peak_gb": 73.8}
{"step": 120, "loss": 2.157278822065415, "story_loss": 2.273533319650968, "grad_norm": 0.2376798689365387, "lr": 0.000282631548055013, "elapsed": 2823.7, "tokens": 23480, "story_tokens": 21907, "peak_gb": 73.8}
{"step": 121, "loss": 2.1611763694534134, "story_loss": 2.2637028333352256, "grad_norm": 0.20111428201198578, "lr": 0.00027938434936445943, "elapsed": 2844.5, "tokens": 23516, "story_tokens": 22018, "peak_gb": 73.8}
{"step": 122, "loss": 2.126649930520447, "story_loss": 2.2387253322270966, "grad_norm": 0.19867098331451416, "lr": 0.0002761321158169134, "elapsed": 2865.2, "tokens": 23454, "story_tokens": 21940, "peak_gb": 73.8}
{"step": 123, "loss": 2.196824270501156, "story_loss": 2.311105198815062, "grad_norm": 0.20024394989013672, "lr": 0.00027287540466585064, "elapsed": 2886.7, "tokens": 24134, "story_tokens": 22517, "peak_gb": 73.8}
{"step": 124, "loss": 2.1025679281876473, "story_loss": 2.223361689652554, "grad_norm": 0.2029353827238083, "lr": 0.00026961477393196127, "elapsed": 2907.9, "tokens": 22877, "story_tokens": 21399, "peak_gb": 73.8}
{"step": 125, "loss": 2.138791167501198, "story_loss": 2.281971560904802, "grad_norm": 0.20015570521354675, "lr": 0.0002663507823075358, "elapsed": 2929.7, "tokens": 23552, "story_tokens": 21647, "peak_gb": 73.8}
{"step": 126, "loss": 2.2034778921584355, "story_loss": 2.327062293303496, "grad_norm": 0.19696979224681854, "lr": 0.000263083989060736, "elapsed": 2953.3, "tokens": 25403, "story_tokens": 23642, "peak_gb": 73.8}
{"step": 127, "loss": 2.1696980597304725, "story_loss": 2.279916285603402, "grad_norm": 0.19903230667114258, "lr": 0.00025981495393976716, "elapsed": 2975.6, "tokens": 24661, "story_tokens": 23150, "peak_gb": 73.8}
{"step": 128, "loss": 2.1744408317629573, "story_loss": 2.2687856813245704, "grad_norm": 0.20622210204601288, "lr": 0.00025654423707696834, "elapsed": 2998.3, "tokens": 23492, "story_tokens": 22043, "peak_gb": 73.8}
{"step": 129, "loss": 2.177817469615916, "story_loss": 2.267185802649638, "grad_norm": 0.20912347733974457, "lr": 0.0002532723988928361, "elapsed": 3020.1, "tokens": 23775, "story_tokens": 22496, "peak_gb": 73.8}
{"step": 130, "loss": 2.1988761355231494, "story_loss": 2.276899110798049, "grad_norm": 0.2065298706293106, "lr": 0.00025, "elapsed": 3041.1, "tokens": 24188, "story_tokens": 22992, "peak_gb": 73.8}
{"step": 131, "loss": 2.1339616192301363, "story_loss": 2.2406430210652144, "grad_norm": 0.20486892759799957, "lr": 0.00024672760110716396, "elapsed": 3062.4, "tokens": 23498, "story_tokens": 22012, "peak_gb": 73.8}
{"step": 132, "loss": 2.090963888149979, "story_loss": 2.1855741987170396, "grad_norm": 0.19717800617218018, "lr": 0.00024345576292303175, "elapsed": 3083.3, "tokens": 23379, "story_tokens": 22034, "peak_gb": 73.8}
{"step": 133, "loss": 2.0473215779137535, "story_loss": 2.2168813218116967, "grad_norm": 0.19186845421791077, "lr": 0.00024018504606023293, "elapsed": 3106.7, "tokens": 25428, "story_tokens": 22955, "peak_gb": 74.0}
{"step": 134, "loss": 2.2207790144523973, "story_loss": 2.324935338808329, "grad_norm": 0.19918695092201233, "lr": 0.00023691601093926405, "elapsed": 3127.7, "tokens": 23735, "story_tokens": 22266, "peak_gb": 74.0}
{"step": 135, "loss": 2.1767298623415448, "story_loss": 2.271829431616655, "grad_norm": 0.20704273879528046, "lr": 0.00023364921769246423, "elapsed": 3149.8, "tokens": 24449, "story_tokens": 22978, "peak_gb": 74.0}
{"step": 136, "loss": 2.1272703024486317, "story_loss": 2.212798077563296, "grad_norm": 0.20328253507614136, "lr": 0.0002303852260680388, "elapsed": 3171.2, "tokens": 23423, "story_tokens": 22059, "peak_gb": 74.0}
{"step": 137, "loss": 2.127609340873634, "story_loss": 2.236019711530006, "grad_norm": 0.2305215299129486, "lr": 0.00022712459533414943, "elapsed": 3192.6, "tokens": 22713, "story_tokens": 21321, "peak_gb": 74.0}
{"step": 138, "loss": 2.1661027394752894, "story_loss": 2.295278009352825, "grad_norm": 0.19901518523693085, "lr": 0.00022386788418308668, "elapsed": 3214.1, "tokens": 24017, "story_tokens": 22266, "peak_gb": 74.0}
{"step": 139, "loss": 2.202415735001133, "story_loss": 2.28781477465521, "grad_norm": 0.19430311024188995, "lr": 0.00022061565063554063, "elapsed": 3236.1, "tokens": 24075, "story_tokens": 22840, "peak_gb": 74.0}
{"step": 140, "loss": 2.1609018410991374, "story_loss": 2.2684129069047483, "grad_norm": 0.20421813428401947, "lr": 0.00021736845194498717, "elapsed": 3257.0, "tokens": 23510, "story_tokens": 21955, "peak_gb": 74.0}
{"step": 141, "loss": 2.13677870893938, "story_loss": 2.275207524789728, "grad_norm": 0.19878925383090973, "lr": 0.0002141268445022052, "elapsed": 3279.3, "tokens": 25106, "story_tokens": 23214, "peak_gb": 74.0}
{"step": 142, "loss": 2.1663474178476796, "story_loss": 2.287378964893492, "grad_norm": 0.19467875361442566, "lr": 0.00021089138373994224, "elapsed": 3300.5, "tokens": 24049, "story_tokens": 22381, "peak_gb": 74.0}
{"step": 143, "loss": 2.1812119928658147, "story_loss": 2.313528334342691, "grad_norm": 0.19458425045013428, "lr": 0.00020766262403774385, "elapsed": 3322.1, "tokens": 24104, "story_tokens": 22343, "peak_gb": 74.0}
{"step": 144, "loss": 2.0843605273040655, "story_loss": 2.2287997040183933, "grad_norm": 0.1963251382112503, "lr": 0.00020444111862696313, "elapsed": 3343.3, "tokens": 23593, "story_tokens": 21652, "peak_gb": 74.0}
{"step": 145, "loss": 2.2604819879711995, "story_loss": 2.380373757338511, "grad_norm": 0.19862686097621918, "lr": 0.00020122741949596797, "elapsed": 3365.1, "tokens": 24036, "story_tokens": 22548, "peak_gb": 74.0}
{"step": 146, "loss": 2.1502985791386364, "story_loss": 2.2801976534333233, "grad_norm": 0.19382935762405396, "lr": 0.0001980220772955602, "elapsed": 3386.5, "tokens": 24530, "story_tokens": 22744, "peak_gb": 74.0}
{"step": 147, "loss": 2.1450545614446037, "story_loss": 2.298341026407264, "grad_norm": 0.1971299797296524, "lr": 0.00019482564124462476, "elapsed": 3407.8, "tokens": 23771, "story_tokens": 21882, "peak_gb": 74.0}
{"step": 148, "loss": 2.1932908902717734, "story_loss": 2.312718493167307, "grad_norm": 0.19350163638591766, "lr": 0.00019163865903602372, "elapsed": 3429.8, "tokens": 24820, "story_tokens": 23179, "peak_gb": 74.0}
{"step": 149, "loss": 2.1253577389336464, "story_loss": 2.236795037893512, "grad_norm": 0.19456064701080322, "lr": 0.00018846167674275175, "elapsed": 3451.1, "tokens": 23639, "story_tokens": 22188, "peak_gb": 74.0}
{"step": 150, "loss": 2.096551932277823, "story_loss": 2.219504993808841, "grad_norm": 0.1982550024986267, "lr": 0.0001852952387243698, "elapsed": 3471.4, "tokens": 23626, "story_tokens": 21881, "peak_gb": 74.0}
{"step": 151, "loss": 2.132227983967509, "story_loss": 2.2268728899991883, "grad_norm": 0.1971299797296524, "lr": 0.00018213988753373146, "elapsed": 3493.4, "tokens": 24559, "story_tokens": 23095, "peak_gb": 74.0}
{"step": 152, "loss": 2.2178976224173628, "story_loss": 2.289308335751714, "grad_norm": 0.19988884031772614, "lr": 0.00017899616382401935, "elapsed": 3514.8, "tokens": 24192, "story_tokens": 23044, "peak_gb": 74.0}
{"step": 153, "loss": 2.09658821102976, "story_loss": 2.2056069771615054, "grad_norm": 0.2009643167257309, "lr": 0.0001758646062561073, "elapsed": 3535.8, "tokens": 23034, "story_tokens": 21507, "peak_gb": 74.0}
{"step": 154, "loss": 2.123725709245367, "story_loss": 2.2441950729827402, "grad_norm": 0.20785751938819885, "lr": 0.00017274575140626317, "elapsed": 3556.6, "tokens": 23479, "story_tokens": 21756, "peak_gb": 74.0}
{"step": 155, "loss": 2.1519017136983143, "story_loss": 2.272577017572038, "grad_norm": 0.2091800421476364, "lr": 0.00016964013367420965, "elapsed": 3578.1, "tokens": 23218, "story_tokens": 21499, "peak_gb": 74.0}
{"step": 156, "loss": 2.1957821578418715, "story_loss": 2.293947184876091, "grad_norm": 0.19847477972507477, "lr": 0.00016654828519155728, "elapsed": 3600.0, "tokens": 24217, "story_tokens": 22688, "peak_gb": 74.0}
{"step": 157, "loss": 2.15603573913997, "story_loss": 2.283438808149592, "grad_norm": 0.2068023532629013, "lr": 0.0001634707357306267, "elapsed": 3621.4, "tokens": 23386, "story_tokens": 21679, "peak_gb": 74.0}
{"step": 158, "loss": 2.1187277811184138, "story_loss": 2.262417626330229, "grad_norm": 0.1925339549779892, "lr": 0.00016040801261367493, "elapsed": 3643.5, "tokens": 24476, "story_tokens": 22576, "peak_gb": 74.0}
{"step": 159, "loss": 2.066438615541615, "story_loss": 2.1829495620227712, "grad_norm": 0.196797713637352, "lr": 0.00015736064062254093, "elapsed": 3665.0, "tokens": 23896, "story_tokens": 22133, "peak_gb": 74.0}
{"step": 160, "loss": 2.124833334258177, "story_loss": 2.1911834793735347, "grad_norm": 0.19829516112804413, "lr": 0.00015432914190872756, "elapsed": 3686.0, "tokens": 23596, "story_tokens": 22609, "peak_gb": 74.0}
{"step": 161, "loss": 2.1152259586206927, "story_loss": 2.2183676694788566, "grad_norm": 0.19776584208011627, "lr": 0.00015131403590393322, "elapsed": 3706.9, "tokens": 23553, "story_tokens": 22158, "peak_gb": 74.0}
{"step": 162, "loss": 2.21761063394859, "story_loss": 2.3218139483086584, "grad_norm": 0.19026564061641693, "lr": 0.00014831583923105, "elapsed": 3729.2, "tokens": 25175, "story_tokens": 23624, "peak_gb": 74.0}
{"step": 163, "loss": 2.2246031409519613, "story_loss": 2.3384512169010683, "grad_norm": 0.19884788990020752, "lr": 0.00014533506561564306, "elapsed": 3750.2, "tokens": 24069, "story_tokens": 22513, "peak_gb": 74.0}
{"step": 164, "loss": 2.1291853336717703, "story_loss": 2.245322390003195, "grad_norm": 0.19363482296466827, "lr": 0.00014237222579792616, "elapsed": 3771.9, "tokens": 24352, "story_tokens": 22691, "peak_gb": 74.0}
{"step": 165, "loss": 2.178328575004407, "story_loss": 2.2767228400991613, "grad_norm": 0.19468213617801666, "lr": 0.00013942782744524973, "elapsed": 3794.1, "tokens": 24655, "story_tokens": 23308, "peak_gb": 74.0}
{"step": 166, "loss": 2.2089415073931264, "story_loss": 2.344025510392099, "grad_norm": 0.2001693844795227, "lr": 0.00013650237506511331, "elapsed": 3815.0, "tokens": 23109, "story_tokens": 21200, "peak_gb": 74.0}
{"step": 167, "loss": 2.1276344078385616, "story_loss": 2.225862433842599, "grad_norm": 0.20287074148654938, "lr": 0.00013359636991872216, "elapsed": 3836.0, "tokens": 23245, "story_tokens": 21882, "peak_gb": 74.0}
{"step": 168, "loss": 2.1381132052340424, "story_loss": 2.2627286683749603, "grad_norm": 0.19760170578956604, "lr": 0.00013071030993509788, "elapsed": 3856.8, "tokens": 23740, "story_tokens": 22031, "peak_gb": 74.0}
{"step": 169, "loss": 2.178212065030876, "story_loss": 2.283848519920854, "grad_norm": 0.1920131891965866, "lr": 0.00012784468962576134, "elapsed": 3877.7, "tokens": 23955, "story_tokens": 22569, "peak_gb": 74.0}
{"step": 170, "loss": 2.1630027517101693, "story_loss": 2.2865954929481056, "grad_norm": 0.19972750544548035, "lr": 0.00012500000000000006, "elapsed": 3898.5, "tokens": 23078, "story_tokens": 21457, "peak_gb": 74.0}
{"step": 171, "loss": 2.106788430057429, "story_loss": 2.2223995250267783, "grad_norm": 0.19251298904418945, "lr": 0.00012217672848073701, "elapsed": 3920.6, "tokens": 24558, "story_tokens": 22873, "peak_gb": 74.0}
{"step": 172, "loss": 2.099673949428956, "story_loss": 2.229907951286298, "grad_norm": 0.19644467532634735, "lr": 0.00011937535882101281, "elapsed": 3941.1, "tokens": 23445, "story_tokens": 21661, "peak_gb": 74.0}
{"step": 173, "loss": 2.1576272267506535, "story_loss": 2.2510632127842594, "grad_norm": 0.19862353801727295, "lr": 0.00011659637102109713, "elapsed": 3961.8, "tokens": 23275, "story_tokens": 21943, "peak_gb": 74.0}
{"step": 174, "loss": 2.1224386549301344, "story_loss": 2.207899403168856, "grad_norm": 0.19809655845165253, "lr": 0.00011384024124624323, "elapsed": 3982.9, "tokens": 23520, "story_tokens": 22232, "peak_gb": 74.0}
{"step": 175, "loss": 2.0904729146240055, "story_loss": 2.1959059108721863, "grad_norm": 0.18610313534736633, "lr": 0.00011110744174509952, "elapsed": 4005.3, "tokens": 24821, "story_tokens": 23325, "peak_gb": 74.0}
{"step": 176, "loss": 2.1078527838305745, "story_loss": 2.2073729781747833, "grad_norm": 0.19562792778015137, "lr": 0.00010839844076879185, "elapsed": 4026.7, "tokens": 23346, "story_tokens": 22033, "peak_gb": 74.0}
{"step": 177, "loss": 2.1219014949616075, "story_loss": 2.233665463925791, "grad_norm": 0.20022399723529816, "lr": 0.00010571370249069162, "elapsed": 4048.3, "tokens": 24141, "story_tokens": 22413, "peak_gb": 74.0}
{"step": 178, "loss": 2.126044023808924, "story_loss": 2.2473934224525434, "grad_norm": 0.1961873322725296, "lr": 0.00010305368692688174, "elapsed": 4069.3, "tokens": 23673, "story_tokens": 21941, "peak_gb": 74.0}
{"step": 179, "loss": 2.198458706295443, "story_loss": 2.303458601168584, "grad_norm": 0.19499605894088745, "lr": 0.00010041884985733524, "elapsed": 4090.8, "tokens": 23838, "story_tokens": 22431, "peak_gb": 74.0}
{"step": 180, "loss": 2.1498029939882985, "story_loss": 2.234066932386013, "grad_norm": 0.1945677250623703, "lr": 9.780964274781984e-05, "elapsed": 4112.1, "tokens": 24042, "story_tokens": 22689, "peak_gb": 74.0}
{"step": 181, "loss": 2.1761157322592424, "story_loss": 2.2657235684849466, "grad_norm": 0.19514450430870056, "lr": 9.522651267254148e-05, "elapsed": 4133.2, "tokens": 24184, "story_tokens": 22802, "peak_gb": 74.0}
{"step": 182, "loss": 2.127668391599462, "story_loss": 2.2428404516858462, "grad_norm": 0.1884775310754776, "lr": 9.266990223754068e-05, "elapsed": 4156.1, "tokens": 24759, "story_tokens": 23093, "peak_gb": 74.0}
{"step": 183, "loss": 2.1176744816361524, "story_loss": 2.2687399644409183, "grad_norm": 0.18977613747119904, "lr": 9.014024950485383e-05, "elapsed": 4177.2, "tokens": 24345, "story_tokens": 22177, "peak_gb": 74.0}
{"step": 184, "loss": 2.159844277574317, "story_loss": 2.2825338935860335, "grad_norm": 0.18914750218391418, "lr": 8.763798791745412e-05, "elapsed": 4198.4, "tokens": 24677, "story_tokens": 23023, "peak_gb": 74.0}
{"step": 185, "loss": 2.1446701228728977, "story_loss": 2.2611874279795874, "grad_norm": 0.18944422900676727, "lr": 8.516354622498279e-05, "elapsed": 4219.8, "tokens": 24305, "story_tokens": 22780, "peak_gb": 74.0}
{"step": 186, "loss": 2.1611982066680024, "story_loss": 2.230904577054071, "grad_norm": 0.18676897883415222, "lr": 8.271734841028553e-05, "elapsed": 4242.2, "tokens": 24182, "story_tokens": 23067, "peak_gb": 74.0}
{"step": 187, "loss": 2.1583045116030113, "story_loss": 2.280269871301431, "grad_norm": 0.20047388970851898, "lr": 8.029981361676455e-05, "elapsed": 4263.1, "tokens": 23636, "story_tokens": 22044, "peak_gb": 74.0}
{"step": 188, "loss": 2.2226383856239043, "story_loss": 2.3385740096943644, "grad_norm": 0.20931287109851837, "lr": 7.791135607656147e-05, "elapsed": 4284.2, "tokens": 23047, "story_tokens": 21488, "peak_gb": 74.0}
{"step": 189, "loss": 2.1948012978937115, "story_loss": 2.316388327588779, "grad_norm": 0.1983254849910736, "lr": 7.555238503958001e-05, "elapsed": 4306.6, "tokens": 23818, "story_tokens": 22204, "peak_gb": 74.0}
{"step": 190, "loss": 2.0856558532120784, "story_loss": 2.185452103485185, "grad_norm": 0.19584354758262634, "lr": 7.322330470336314e-05, "elapsed": 4327.2, "tokens": 23422, "story_tokens": 22072, "peak_gb": 74.0}
{"step": 191, "loss": 2.1209996635885213, "story_loss": 2.2550669410843525, "grad_norm": 0.1865358054637909, "lr": 7.092451414383643e-05, "elapsed": 4348.4, "tokens": 24478, "story_tokens": 22729, "peak_gb": 74.0}
{"step": 192, "loss": 2.163981487966057, "story_loss": 2.3010358193646785, "grad_norm": 0.1915864497423172, "lr": 6.865640724692814e-05, "elapsed": 4369.0, "tokens": 23537, "story_tokens": 21678, "peak_gb": 74.0}
{"step": 193, "loss": 2.0238342875352733, "story_loss": 2.149275450309536, "grad_norm": 0.1901852786540985, "lr": 6.641937264107867e-05, "elapsed": 4392.1, "tokens": 25676, "story_tokens": 23725, "peak_gb": 74.0}
{"step": 194, "loss": 2.174631115814107, "story_loss": 2.2967563908608817, "grad_norm": 0.18610474467277527, "lr": 6.421379363065141e-05, "elapsed": 4414.2, "tokens": 25128, "story_tokens": 23383, "peak_gb": 74.0}
{"step": 195, "loss": 2.1292220672401, "story_loss": 2.22624035446777, "grad_norm": 0.1936321258544922, "lr": 6.204004813025568e-05, "elapsed": 4435.0, "tokens": 23507, "story_tokens": 22037, "peak_gb": 74.0}
{"step": 196, "loss": 2.124425033570624, "story_loss": 2.2472971026348736, "grad_norm": 0.19099079072475433, "lr": 5.989850859999227e-05, "elapsed": 4456.0, "tokens": 23692, "story_tokens": 21965, "peak_gb": 74.0}
{"step": 197, "loss": 2.0754898218946427, "story_loss": 2.254012381195263, "grad_norm": 0.190025195479393, "lr": 5.7789541981635134e-05, "elapsed": 4477.1, "tokens": 23613, "story_tokens": 21352, "peak_gb": 74.0}
{"step": 198, "loss": 2.0392440762540494, "story_loss": 2.215099387855835, "grad_norm": 0.19211354851722717, "lr": 5.5713509635757274e-05, "elapsed": 4498.6, "tokens": 24701, "story_tokens": 22254, "peak_gb": 74.0}
{"step": 199, "loss": 2.1612636382006496, "story_loss": 2.2635540867238455, "grad_norm": 0.19241808354854584, "lr": 5.367076727981382e-05, "elapsed": 4520.4, "tokens": 24091, "story_tokens": 22692, "peak_gb": 74.0}
{"step": 200, "loss": 2.10213632029025, "story_loss": 2.198537471039783, "grad_norm": 0.1931283175945282, "lr": 5.1661664927191235e-05, "elapsed": 4542.5, "tokens": 24487, "story_tokens": 23119, "peak_gb": 74.0}
{"step": 201, "loss": 2.1484342516948254, "story_loss": 2.2322791459573077, "grad_norm": 0.19608980417251587, "lr": 4.9686546827234865e-05, "elapsed": 4564.0, "tokens": 23364, "story_tokens": 22182, "peak_gb": 74.0}
{"step": 202, "loss": 2.09606545016404, "story_loss": 2.2289453180077077, "grad_norm": 0.19232523441314697, "lr": 4.7745751406263163e-05, "elapsed": 4584.8, "tokens": 23132, "story_tokens": 21277, "peak_gb": 74.0}
{"step": 203, "loss": 2.138831668808832, "story_loss": 2.2244257985655067, "grad_norm": 0.1917758584022522, "lr": 4.5839611209580275e-05, "elapsed": 4607.0, "tokens": 25727, "story_tokens": 24364, "peak_gb": 74.0}
{"step": 204, "loss": 2.174729519167954, "story_loss": 2.2748414633268035, "grad_norm": 0.19828639924526215, "lr": 4.3968452844496076e-05, "elapsed": 4627.3, "tokens": 23157, "story_tokens": 21695, "peak_gb": 74.0}
{"step": 205, "loss": 2.1313140670313784, "story_loss": 2.2289665154426017, "grad_norm": 0.19208993017673492, "lr": 4.213259692436367e-05, "elapsed": 4649.1, "tokens": 23618, "story_tokens": 22187, "peak_gb": 74.0}
{"step": 206, "loss": 2.159928192925707, "story_loss": 2.250063296134728, "grad_norm": 0.19512705504894257, "lr": 4.033235801364402e-05, "elapsed": 4669.2, "tokens": 23035, "story_tokens": 21677, "peak_gb": 74.0}
{"step": 207, "loss": 2.184542786309187, "story_loss": 2.295113837698836, "grad_norm": 0.19557148218154907, "lr": 3.85680445740067e-05, "elapsed": 4690.8, "tokens": 24002, "story_tokens": 22506, "peak_gb": 74.0}
{"step": 208, "loss": 2.09742661540417, "story_loss": 2.199547648518224, "grad_norm": 0.18948735296726227, "lr": 3.6839958911476953e-05, "elapsed": 4712.4, "tokens": 24583, "story_tokens": 22937, "peak_gb": 74.0}
{"step": 209, "loss": 2.1363743771074266, "story_loss": 2.217723045950464, "grad_norm": 0.1894845813512802, "lr": 3.514839712463683e-05, "elapsed": 4733.4, "tokens": 23648, "story_tokens": 22393, "peak_gb": 74.0}
{"step": 210, "loss": 2.1670827230946554, "story_loss": 2.2658588755479308, "grad_norm": 0.19256697595119476, "lr": 3.3493649053890325e-05, "elapsed": 4754.6, "tokens": 24007, "story_tokens": 22642, "peak_gb": 74.0}
{"step": 211, "loss": 2.032399359250761, "story_loss": 2.146510927420228, "grad_norm": 0.18212708830833435, "lr": 3.187599823180071e-05, "elapsed": 4776.6, "tokens": 24994, "story_tokens": 23097, "peak_gb": 74.0}
{"step": 212, "loss": 1.959506416659976, "story_loss": 2.1246655574295414, "grad_norm": 0.1950523853302002, "lr": 3.0295721834508684e-05, "elapsed": 4797.3, "tokens": 24181, "story_tokens": 21910, "peak_gb": 74.0}
{"step": 213, "loss": 2.1530285741684, "story_loss": 2.265146666682606, "grad_norm": 0.18444478511810303, "lr": 2.875309063423956e-05, "elapsed": 4818.6, "tokens": 24305, "story_tokens": 22873, "peak_gb": 74.0}
{"step": 214, "loss": 2.086845663123495, "story_loss": 2.1789238233033195, "grad_norm": 0.18984034657478333, "lr": 2.7248368952908055e-05, "elapsed": 4839.6, "tokens": 24015, "story_tokens": 22684, "peak_gb": 74.0}
{"step": 215, "loss": 2.1390795495894115, "story_loss": 2.2723976719694576, "grad_norm": 0.18772205710411072, "lr": 2.5781814616827938e-05, "elapsed": 4860.4, "tokens": 24021, "story_tokens": 22166, "peak_gb": 74.0}
{"step": 216, "loss": 2.0949398350886366, "story_loss": 2.215449316306081, "grad_norm": 0.1928897649049759, "lr": 2.43536789125349e-05, "elapsed": 4882.0, "tokens": 24159, "story_tokens": 22423, "peak_gb": 74.0}
{"step": 217, "loss": 2.1012187937283096, "story_loss": 2.2621676879096158, "grad_norm": 0.1849754899740219, "lr": 2.296420654372966e-05, "elapsed": 4904.1, "tokens": 25145, "story_tokens": 22811, "peak_gb": 74.0}
{"step": 218, "loss": 2.176099157528696, "story_loss": 2.3109604992961406, "grad_norm": 0.19045360386371613, "lr": 2.1613635589349755e-05, "elapsed": 4925.8, "tokens": 24657, "story_tokens": 22865, "peak_gb": 74.0}
{"step": 219, "loss": 2.0154591967455566, "story_loss": 2.1220625487876186, "grad_norm": 0.1914500743150711, "lr": 2.030219746277545e-05, "elapsed": 4946.7, "tokens": 23464, "story_tokens": 21698, "peak_gb": 74.0}
{"step": 220, "loss": 2.13241984656382, "story_loss": 2.22713421958909, "grad_norm": 0.19752664864063263, "lr": 1.9030116872178316e-05, "elapsed": 4967.1, "tokens": 23169, "story_tokens": 21787, "peak_gb": 74.0}
{"step": 221, "loss": 2.0946982240247998, "story_loss": 2.1987419890359248, "grad_norm": 0.18693391978740692, "lr": 1.779761178201894e-05, "elapsed": 4989.9, "tokens": 24460, "story_tokens": 22930, "peak_gb": 74.0}
{"step": 222, "loss": 2.087176007388753, "story_loss": 2.1928902349370993, "grad_norm": 0.1840634047985077, "lr": 1.6604893375699592e-05, "elapsed": 5011.5, "tokens": 24420, "story_tokens": 22933, "peak_gb": 74.0}
{"step": 223, "loss": 2.106059748240937, "story_loss": 2.223359064756952, "grad_norm": 0.19394105672836304, "lr": 1.5452166019378987e-05, "elapsed": 5032.4, "tokens": 23815, "story_tokens": 22223, "peak_gb": 74.0}
{"step": 224, "loss": 2.179365387103079, "story_loss": 2.271313707395647, "grad_norm": 0.19263267517089844, "lr": 1.4339627226955393e-05, "elapsed": 5053.3, "tokens": 24320, "story_tokens": 22951, "peak_gb": 74.0}
{"step": 225, "loss": 2.037422546408824, "story_loss": 2.174353173227042, "grad_norm": 0.1880512684583664, "lr": 1.3267467626223605e-05, "elapsed": 5074.3, "tokens": 24160, "story_tokens": 22258, "peak_gb": 74.0}
{"step": 226, "loss": 2.2299350235539945, "story_loss": 2.3463550908116626, "grad_norm": 0.19180616736412048, "lr": 1.2235870926211617e-05, "elapsed": 5095.1, "tokens": 23579, "story_tokens": 22015, "peak_gb": 74.0}
{"step": 227, "loss": 2.1731004933080746, "story_loss": 2.309813714775231, "grad_norm": 0.187334343791008, "lr": 1.1245013885703342e-05, "elapsed": 5116.6, "tokens": 25141, "story_tokens": 23146, "peak_gb": 74.0}
{"step": 228, "loss": 2.1610429960052437, "story_loss": 2.3064397115890554, "grad_norm": 0.18596212565898895, "lr": 1.0295066282951737e-05, "elapsed": 5138.2, "tokens": 24724, "story_tokens": 22806, "peak_gb": 74.0}
{"step": 229, "loss": 2.126179403489536, "story_loss": 2.2387236306756018, "grad_norm": 0.19209794700145721, "lr": 9.386190886588208e-06, "elapsed": 5159.4, "tokens": 24679, "story_tokens": 22850, "peak_gb": 74.0}
{"step": 230, "loss": 2.161287514170863, "story_loss": 2.250745536600291, "grad_norm": 0.1966123878955841, "lr": 8.51854342773295e-06, "elapsed": 5180.6, "tokens": 24023, "story_tokens": 22674, "peak_gb": 74.0}
{"step": 231, "loss": 2.1262301904918903, "story_loss": 2.221605058694912, "grad_norm": 0.1914987415075302, "lr": 7.692272573311426e-06, "elapsed": 5201.9, "tokens": 23834, "story_tokens": 22396, "peak_gb": 74.0}
{"step": 232, "loss": 2.1414078260928293, "story_loss": 2.262147376773112, "grad_norm": 0.193291574716568, "lr": 6.907519900580861e-06, "elapsed": 5222.7, "tokens": 22973, "story_tokens": 21396, "peak_gb": 74.0}
{"step": 233, "loss": 2.123574449903428, "story_loss": 2.2085989800537753, "grad_norm": 0.19015716016292572, "lr": 6.164419872871835e-06, "elapsed": 5243.8, "tokens": 24182, "story_tokens": 22780, "peak_gb": 74.0}
{"step": 234, "loss": 2.116145471813347, "story_loss": 2.224164235906337, "grad_norm": 0.1909848153591156, "lr": 5.463099816548578e-06, "elapsed": 5265.1, "tokens": 24858, "story_tokens": 23315, "peak_gb": 74.0}
{"step": 235, "loss": 2.200163169807977, "story_loss": 2.294532468864321, "grad_norm": 0.18683871626853943, "lr": 4.803679899192393e-06, "elapsed": 5287.0, "tokens": 24535, "story_tokens": 23235, "peak_gb": 74.0}
{"step": 236, "loss": 2.073702368048498, "story_loss": 2.1863973563216756, "grad_norm": 0.1822863072156906, "lr": 4.186273109011374e-06, "elapsed": 5307.5, "tokens": 23524, "story_tokens": 21974, "peak_gb": 74.0}
{"step": 237, "loss": 2.142754638574685, "story_loss": 2.251325151215396, "grad_norm": 0.19085335731506348, "lr": 3.6109852354805626e-06, "elapsed": 5328.1, "tokens": 23550, "story_tokens": 22035, "peak_gb": 74.0}
{"step": 238, "loss": 2.0922710598238115, "story_loss": 2.267257435320742, "grad_norm": 0.18605107069015503, "lr": 3.077914851215585e-06, "elapsed": 5349.2, "tokens": 24732, "story_tokens": 22370, "peak_gb": 74.0}
{"step": 239, "loss": 2.1454287585882788, "story_loss": 2.281630956886021, "grad_norm": 0.18465059995651245, "lr": 2.5871532950824397e-06, "elapsed": 5371.1, "tokens": 25480, "story_tokens": 23535, "peak_gb": 74.0}
{"step": 240, "loss": 2.1740022049323975, "story_loss": 2.30042906047158, "grad_norm": 0.18558776378631592, "lr": 2.1387846565474044e-06, "elapsed": 5392.6, "tokens": 24430, "story_tokens": 22801, "peak_gb": 74.0}
{"step": 241, "loss": 2.1120064692294966, "story_loss": 2.2064605457513404, "grad_norm": 0.18529267609119415, "lr": 1.7328857612684267e-06, "elapsed": 5413.7, "tokens": 24374, "story_tokens": 22944, "peak_gb": 74.0}
{"step": 242, "loss": 2.1312286831317238, "story_loss": 2.2463154929201137, "grad_norm": 0.18937011063098907, "lr": 1.3695261579316775e-06, "elapsed": 5434.0, "tokens": 23340, "story_tokens": 21637, "peak_gb": 74.0}
{"step": 243, "loss": 2.179323569068773, "story_loss": 2.299752475159232, "grad_norm": 0.1938820779323578, "lr": 1.0487681063345856e-06, "elapsed": 5454.9, "tokens": 23636, "story_tokens": 21902, "peak_gb": 74.0}
{"step": 244, "loss": 2.101246538986213, "story_loss": 2.2035419625139308, "grad_norm": 0.1882134974002838, "lr": 7.70666566718009e-07, "elapsed": 5475.9, "tokens": 23391, "story_tokens": 21984, "peak_gb": 74.0}
{"step": 245, "loss": 2.0054557280799714, "story_loss": 2.1340877898495085, "grad_norm": 0.19213250279426575, "lr": 5.352691903491303e-07, "elapsed": 5496.9, "tokens": 23028, "story_tokens": 21164, "peak_gb": 74.0}
{"step": 246, "loss": 2.1193229433842657, "story_loss": 2.2046631037648727, "grad_norm": 0.1913616955280304, "lr": 3.426163113565417e-07, "elapsed": 5517.9, "tokens": 24536, "story_tokens": 23281, "peak_gb": 74.0}
{"step": 247, "loss": 2.1489895599855746, "story_loss": 2.267791215159586, "grad_norm": 0.1876077950000763, "lr": 1.9274093981927476e-07, "elapsed": 5539.6, "tokens": 24373, "story_tokens": 22621, "peak_gb": 74.0}
{"step": 248, "loss": 2.1964749090218727, "story_loss": 2.329632946396076, "grad_norm": 0.18639346957206726, "lr": 8.566875611068504e-08, "elapsed": 5560.6, "tokens": 24452, "story_tokens": 22668, "peak_gb": 74.0}
{"step": 249, "loss": 2.176683578459853, "story_loss": 2.300028125, "grad_norm": 0.19138716161251068, "lr": 2.1418106498249933e-08, "elapsed": 5582.4, "tokens": 24330, "story_tokens": 22500, "peak_gb": 74.0}
{"step": 250, "loss": 2.1486969788163934, "story_loss": 2.2992415825368213, "grad_norm": 0.21311290562152863, "lr": 0.0, "elapsed": 5604.9, "tokens": 24790, "story_tokens": 22677, "peak_gb": 74.0}