{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 2.6267175844962347, "eval_steps": 500, "global_step": 18400, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.0014276026981690996, "grad_norm": 1.2670224905014038, "learning_rate": 0.00016, "loss": 2.898906135559082, "step": 10 }, { "epoch": 0.0028552053963381992, "grad_norm": 1.2538419961929321, "learning_rate": 0.00019992382766008095, "loss": 2.378516960144043, "step": 20 }, { "epoch": 0.004282808094507299, "grad_norm": 1.409246563911438, "learning_rate": 0.0001998286122351821, "loss": 2.4753650665283202, "step": 30 }, { "epoch": 0.0057104107926763984, "grad_norm": 1.094603180885315, "learning_rate": 0.00019973339681028326, "loss": 2.2721933364868163, "step": 40 }, { "epoch": 0.007138013490845497, "grad_norm": 1.1403776407241821, "learning_rate": 0.00019963818138538444, "loss": 2.263425827026367, "step": 50 }, { "epoch": 0.008565616189014597, "grad_norm": 1.1656428575515747, "learning_rate": 0.0001995429659604856, "loss": 2.363163185119629, "step": 60 }, { "epoch": 0.009993218887183696, "grad_norm": 1.1926625967025757, "learning_rate": 0.00019944775053558678, "loss": 2.1333871841430665, "step": 70 }, { "epoch": 0.011420821585352797, "grad_norm": 0.9967195391654968, "learning_rate": 0.00019935253511068794, "loss": 2.279196929931641, "step": 80 }, { "epoch": 0.012848424283521896, "grad_norm": 1.1345535516738892, "learning_rate": 0.00019925731968578912, "loss": 2.181261444091797, "step": 90 }, { "epoch": 0.014276026981690995, "grad_norm": 1.2285407781600952, "learning_rate": 0.00019916210426089027, "loss": 2.2282310485839845, "step": 100 }, { "epoch": 0.015703629679860096, "grad_norm": 2.191948890686035, "learning_rate": 0.00019906688883599146, "loss": 2.2829305648803713, "step": 110 }, { "epoch": 0.017131232378029194, "grad_norm": 1.1868839263916016, "learning_rate": 0.00019897167341109258, "loss": 2.1634748458862303, "step": 120 }, { "epoch": 0.018558835076198293, "grad_norm": 1.3730543851852417, "learning_rate": 0.00019887645798619377, "loss": 2.235073471069336, "step": 130 }, { "epoch": 0.019986437774367392, "grad_norm": 1.3060146570205688, "learning_rate": 0.00019878124256129492, "loss": 2.254693031311035, "step": 140 }, { "epoch": 0.021414040472536495, "grad_norm": 1.13761305809021, "learning_rate": 0.0001986860271363961, "loss": 2.255308723449707, "step": 150 }, { "epoch": 0.022841643170705594, "grad_norm": 1.2420964241027832, "learning_rate": 0.00019859081171149726, "loss": 2.199342155456543, "step": 160 }, { "epoch": 0.024269245868874693, "grad_norm": 1.2628268003463745, "learning_rate": 0.00019849559628659844, "loss": 2.2568809509277346, "step": 170 }, { "epoch": 0.02569684856704379, "grad_norm": 1.0423402786254883, "learning_rate": 0.0001984003808616996, "loss": 2.201403045654297, "step": 180 }, { "epoch": 0.02712445126521289, "grad_norm": 1.3187193870544434, "learning_rate": 0.00019830516543680078, "loss": 2.31024055480957, "step": 190 }, { "epoch": 0.02855205396338199, "grad_norm": 1.287013292312622, "learning_rate": 0.00019820995001190194, "loss": 2.2775556564331056, "step": 200 }, { "epoch": 0.029979656661551092, "grad_norm": 1.445034384727478, "learning_rate": 0.00019811473458700312, "loss": 2.214271354675293, "step": 210 }, { "epoch": 0.03140725935972019, "grad_norm": 1.4336904287338257, "learning_rate": 0.00019801951916210427, "loss": 2.1211181640625, "step": 220 }, { "epoch": 0.03283486205788929, "grad_norm": 1.303303837776184, "learning_rate": 0.00019792430373720543, "loss": 2.1037008285522463, "step": 230 }, { "epoch": 0.03426246475605839, "grad_norm": 1.2367061376571655, "learning_rate": 0.0001978290883123066, "loss": 2.1291038513183596, "step": 240 }, { "epoch": 0.03569006745422749, "grad_norm": 1.314647912979126, "learning_rate": 0.00019773387288740777, "loss": 2.1351301193237306, "step": 250 }, { "epoch": 0.03711767015239659, "grad_norm": 1.3084362745285034, "learning_rate": 0.00019763865746250895, "loss": 2.0980342864990233, "step": 260 }, { "epoch": 0.038545272850565686, "grad_norm": 1.712534785270691, "learning_rate": 0.0001975434420376101, "loss": 2.25567569732666, "step": 270 }, { "epoch": 0.039972875548734785, "grad_norm": 1.5107096433639526, "learning_rate": 0.0001974482266127113, "loss": 2.198653793334961, "step": 280 }, { "epoch": 0.041400478246903884, "grad_norm": 1.1626437902450562, "learning_rate": 0.00019735301118781244, "loss": 2.136816215515137, "step": 290 }, { "epoch": 0.04282808094507299, "grad_norm": 1.551853060722351, "learning_rate": 0.00019725779576291362, "loss": 2.2106935501098635, "step": 300 }, { "epoch": 0.04425568364324209, "grad_norm": 1.5131734609603882, "learning_rate": 0.00019716258033801475, "loss": 2.0967830657958983, "step": 310 }, { "epoch": 0.04568328634141119, "grad_norm": 1.310641884803772, "learning_rate": 0.00019706736491311593, "loss": 2.2105255126953125, "step": 320 }, { "epoch": 0.047110889039580287, "grad_norm": 1.3621881008148193, "learning_rate": 0.0001969721494882171, "loss": 2.1775463104248045, "step": 330 }, { "epoch": 0.048538491737749385, "grad_norm": 1.4618415832519531, "learning_rate": 0.00019687693406331827, "loss": 2.148931694030762, "step": 340 }, { "epoch": 0.049966094435918484, "grad_norm": 1.3173977136611938, "learning_rate": 0.00019678171863841943, "loss": 2.1974794387817385, "step": 350 }, { "epoch": 0.05139369713408758, "grad_norm": 1.1830055713653564, "learning_rate": 0.0001966865032135206, "loss": 2.1695978164672853, "step": 360 }, { "epoch": 0.05282129983225668, "grad_norm": 1.1831724643707275, "learning_rate": 0.00019659128778862177, "loss": 2.0823343276977537, "step": 370 }, { "epoch": 0.05424890253042578, "grad_norm": 1.1693743467330933, "learning_rate": 0.00019649607236372295, "loss": 2.1867488861083983, "step": 380 }, { "epoch": 0.05567650522859488, "grad_norm": 1.25830078125, "learning_rate": 0.0001964008569388241, "loss": 2.2133615493774412, "step": 390 }, { "epoch": 0.05710410792676398, "grad_norm": 1.3678141832351685, "learning_rate": 0.00019630564151392529, "loss": 2.308313751220703, "step": 400 }, { "epoch": 0.05853171062493308, "grad_norm": 1.1440492868423462, "learning_rate": 0.00019621042608902641, "loss": 2.1140726089477537, "step": 410 }, { "epoch": 0.059959313323102184, "grad_norm": 1.1869196891784668, "learning_rate": 0.0001961152106641276, "loss": 2.0346582412719725, "step": 420 }, { "epoch": 0.06138691602127128, "grad_norm": 1.4892522096633911, "learning_rate": 0.00019601999523922875, "loss": 2.2025136947631836, "step": 430 }, { "epoch": 0.06281451871944038, "grad_norm": 1.8906227350234985, "learning_rate": 0.00019592477981432993, "loss": 2.297587776184082, "step": 440 }, { "epoch": 0.06424212141760947, "grad_norm": 1.2389851808547974, "learning_rate": 0.0001958295643894311, "loss": 2.130488395690918, "step": 450 }, { "epoch": 0.06566972411577858, "grad_norm": 1.425410509109497, "learning_rate": 0.00019573434896453227, "loss": 2.144012451171875, "step": 460 }, { "epoch": 0.06709732681394767, "grad_norm": 1.3087464570999146, "learning_rate": 0.00019563913353963343, "loss": 2.2011600494384767, "step": 470 }, { "epoch": 0.06852492951211678, "grad_norm": 1.3441588878631592, "learning_rate": 0.0001955439181147346, "loss": 2.291761016845703, "step": 480 }, { "epoch": 0.06995253221028588, "grad_norm": 1.2851412296295166, "learning_rate": 0.00019544870268983577, "loss": 2.24096622467041, "step": 490 }, { "epoch": 0.07138013490845498, "grad_norm": 1.2628257274627686, "learning_rate": 0.00019535348726493692, "loss": 2.080752944946289, "step": 500 }, { "epoch": 0.07280773760662408, "grad_norm": 1.2638740539550781, "learning_rate": 0.00019525827184003808, "loss": 2.4698177337646485, "step": 510 }, { "epoch": 0.07423534030479317, "grad_norm": 1.2560440301895142, "learning_rate": 0.00019516305641513926, "loss": 2.132457733154297, "step": 520 }, { "epoch": 0.07566294300296228, "grad_norm": 1.7629936933517456, "learning_rate": 0.0001950678409902404, "loss": 2.166901969909668, "step": 530 }, { "epoch": 0.07709054570113137, "grad_norm": 1.8175091743469238, "learning_rate": 0.0001949726255653416, "loss": 2.1663414001464845, "step": 540 }, { "epoch": 0.07851814839930048, "grad_norm": 1.1744414567947388, "learning_rate": 0.00019487741014044275, "loss": 2.388395309448242, "step": 550 }, { "epoch": 0.07994575109746957, "grad_norm": 0.9943644404411316, "learning_rate": 0.00019478219471554393, "loss": 2.1379253387451174, "step": 560 }, { "epoch": 0.08137335379563868, "grad_norm": 1.2485833168029785, "learning_rate": 0.0001946869792906451, "loss": 2.1750890731811525, "step": 570 }, { "epoch": 0.08280095649380777, "grad_norm": 1.0852699279785156, "learning_rate": 0.00019459176386574627, "loss": 2.2631216049194336, "step": 580 }, { "epoch": 0.08422855919197687, "grad_norm": 1.5462923049926758, "learning_rate": 0.00019449654844084743, "loss": 2.2366138458251954, "step": 590 }, { "epoch": 0.08565616189014598, "grad_norm": 1.2923717498779297, "learning_rate": 0.00019440133301594858, "loss": 2.1567543029785154, "step": 600 }, { "epoch": 0.08708376458831507, "grad_norm": 1.3011770248413086, "learning_rate": 0.00019430611759104976, "loss": 2.1422496795654298, "step": 610 }, { "epoch": 0.08851136728648418, "grad_norm": 1.4617754220962524, "learning_rate": 0.00019421090216615092, "loss": 2.1605356216430662, "step": 620 }, { "epoch": 0.08993896998465327, "grad_norm": 1.5295599699020386, "learning_rate": 0.0001941156867412521, "loss": 2.0025182723999024, "step": 630 }, { "epoch": 0.09136657268282238, "grad_norm": 1.3012663125991821, "learning_rate": 0.00019402047131635326, "loss": 2.1372703552246093, "step": 640 }, { "epoch": 0.09279417538099147, "grad_norm": 1.4276515245437622, "learning_rate": 0.00019392525589145444, "loss": 2.315623474121094, "step": 650 }, { "epoch": 0.09422177807916057, "grad_norm": 1.5132569074630737, "learning_rate": 0.0001938300404665556, "loss": 2.258671760559082, "step": 660 }, { "epoch": 0.09564938077732967, "grad_norm": 1.1320735216140747, "learning_rate": 0.00019373482504165678, "loss": 2.21956844329834, "step": 670 }, { "epoch": 0.09707698347549877, "grad_norm": 1.1464308500289917, "learning_rate": 0.00019363960961675793, "loss": 2.173979949951172, "step": 680 }, { "epoch": 0.09850458617366786, "grad_norm": 1.2780959606170654, "learning_rate": 0.0001935443941918591, "loss": 2.159123992919922, "step": 690 }, { "epoch": 0.09993218887183697, "grad_norm": 1.3222688436508179, "learning_rate": 0.00019344917876696024, "loss": 2.1596654891967773, "step": 700 }, { "epoch": 0.10135979157000606, "grad_norm": 1.2251396179199219, "learning_rate": 0.00019335396334206143, "loss": 2.0882038116455077, "step": 710 }, { "epoch": 0.10278739426817517, "grad_norm": 1.2655571699142456, "learning_rate": 0.00019325874791716258, "loss": 2.1701095581054686, "step": 720 }, { "epoch": 0.10421499696634427, "grad_norm": 1.3139557838439941, "learning_rate": 0.00019316353249226376, "loss": 2.1374244689941406, "step": 730 }, { "epoch": 0.10564259966451336, "grad_norm": 0.9476764798164368, "learning_rate": 0.00019306831706736492, "loss": 2.131810760498047, "step": 740 }, { "epoch": 0.10707020236268247, "grad_norm": 1.3916528224945068, "learning_rate": 0.0001929731016424661, "loss": 2.2276470184326174, "step": 750 }, { "epoch": 0.10849780506085156, "grad_norm": 1.2913047075271606, "learning_rate": 0.00019287788621756726, "loss": 2.0985342025756837, "step": 760 }, { "epoch": 0.10992540775902067, "grad_norm": 1.4780009984970093, "learning_rate": 0.00019278267079266844, "loss": 2.3402137756347656, "step": 770 }, { "epoch": 0.11135301045718976, "grad_norm": 1.38652765750885, "learning_rate": 0.0001926874553677696, "loss": 2.1956760406494142, "step": 780 }, { "epoch": 0.11278061315535887, "grad_norm": 1.2437113523483276, "learning_rate": 0.00019259223994287075, "loss": 2.344030570983887, "step": 790 }, { "epoch": 0.11420821585352796, "grad_norm": 1.1223382949829102, "learning_rate": 0.0001924970245179719, "loss": 2.245418167114258, "step": 800 }, { "epoch": 0.11563581855169706, "grad_norm": 1.5828821659088135, "learning_rate": 0.0001924018090930731, "loss": 2.2958986282348635, "step": 810 }, { "epoch": 0.11706342124986616, "grad_norm": 1.1058349609375, "learning_rate": 0.00019230659366817424, "loss": 2.0216651916503907, "step": 820 }, { "epoch": 0.11849102394803526, "grad_norm": 1.1991469860076904, "learning_rate": 0.00019221137824327543, "loss": 2.1821861267089844, "step": 830 }, { "epoch": 0.11991862664620437, "grad_norm": 1.0211104154586792, "learning_rate": 0.00019211616281837658, "loss": 2.1515743255615236, "step": 840 }, { "epoch": 0.12134622934437346, "grad_norm": 1.163668155670166, "learning_rate": 0.00019202094739347776, "loss": 2.0592647552490235, "step": 850 }, { "epoch": 0.12277383204254257, "grad_norm": 1.0804855823516846, "learning_rate": 0.00019192573196857892, "loss": 2.0916107177734373, "step": 860 }, { "epoch": 0.12420143474071166, "grad_norm": 1.2400697469711304, "learning_rate": 0.0001918305165436801, "loss": 2.0994134902954102, "step": 870 }, { "epoch": 0.12562903743888076, "grad_norm": 1.3643152713775635, "learning_rate": 0.00019173530111878123, "loss": 2.2519294738769533, "step": 880 }, { "epoch": 0.12705664013704987, "grad_norm": 1.3758548498153687, "learning_rate": 0.0001916400856938824, "loss": 2.22522029876709, "step": 890 }, { "epoch": 0.12848424283521895, "grad_norm": 1.1675829887390137, "learning_rate": 0.00019154487026898357, "loss": 2.1150989532470703, "step": 900 }, { "epoch": 0.12991184553338805, "grad_norm": 1.2862130403518677, "learning_rate": 0.00019144965484408475, "loss": 2.289804458618164, "step": 910 }, { "epoch": 0.13133944823155716, "grad_norm": 1.271008849143982, "learning_rate": 0.0001913639609616758, "loss": 2.1846010208129885, "step": 920 }, { "epoch": 0.13276705092972627, "grad_norm": 1.1630209684371948, "learning_rate": 0.00019126874553677695, "loss": 2.2126953125, "step": 930 }, { "epoch": 0.13419465362789534, "grad_norm": 1.1368460655212402, "learning_rate": 0.00019117353011187813, "loss": 2.178596305847168, "step": 940 }, { "epoch": 0.13562225632606445, "grad_norm": 1.0295072793960571, "learning_rate": 0.0001910783146869793, "loss": 2.2031627655029298, "step": 950 }, { "epoch": 0.13704985902423356, "grad_norm": 1.379927396774292, "learning_rate": 0.00019098309926208047, "loss": 2.120400619506836, "step": 960 }, { "epoch": 0.13847746172240266, "grad_norm": 1.1450871229171753, "learning_rate": 0.00019088788383718163, "loss": 2.151012992858887, "step": 970 }, { "epoch": 0.13990506442057177, "grad_norm": 1.0418351888656616, "learning_rate": 0.0001907926684122828, "loss": 2.1239688873291014, "step": 980 }, { "epoch": 0.14133266711874085, "grad_norm": 1.0744861364364624, "learning_rate": 0.00019069745298738397, "loss": 2.090304374694824, "step": 990 }, { "epoch": 0.14276026981690995, "grad_norm": 1.1106981039047241, "learning_rate": 0.00019060223756248515, "loss": 2.225258636474609, "step": 1000 }, { "epoch": 0.14418787251507906, "grad_norm": 1.4493266344070435, "learning_rate": 0.00019050702213758628, "loss": 2.0656681060791016, "step": 1010 }, { "epoch": 0.14561547521324816, "grad_norm": 1.1414896249771118, "learning_rate": 0.00019041180671268746, "loss": 2.103784942626953, "step": 1020 }, { "epoch": 0.14704307791141724, "grad_norm": 1.3252989053726196, "learning_rate": 0.00019031659128778861, "loss": 2.2042461395263673, "step": 1030 }, { "epoch": 0.14847068060958635, "grad_norm": 1.265684962272644, "learning_rate": 0.0001902213758628898, "loss": 1.9412918090820312, "step": 1040 }, { "epoch": 0.14989828330775545, "grad_norm": 1.3606458902359009, "learning_rate": 0.00019012616043799095, "loss": 2.187259292602539, "step": 1050 }, { "epoch": 0.15132588600592456, "grad_norm": 1.0192638635635376, "learning_rate": 0.00019003094501309213, "loss": 2.182966423034668, "step": 1060 }, { "epoch": 0.15275348870409364, "grad_norm": 1.5140159130096436, "learning_rate": 0.0001899357295881933, "loss": 2.3263607025146484, "step": 1070 }, { "epoch": 0.15418109140226274, "grad_norm": 1.1377187967300415, "learning_rate": 0.00018984051416329447, "loss": 2.159488868713379, "step": 1080 }, { "epoch": 0.15560869410043185, "grad_norm": 1.1288865804672241, "learning_rate": 0.00018974529873839563, "loss": 2.306215286254883, "step": 1090 }, { "epoch": 0.15703629679860096, "grad_norm": 1.280045747756958, "learning_rate": 0.00018965008331349678, "loss": 2.225907897949219, "step": 1100 }, { "epoch": 0.15846389949677006, "grad_norm": 1.145614743232727, "learning_rate": 0.00018955486788859797, "loss": 2.1122207641601562, "step": 1110 }, { "epoch": 0.15989150219493914, "grad_norm": 1.1612759828567505, "learning_rate": 0.00018945965246369912, "loss": 2.2271358489990236, "step": 1120 }, { "epoch": 0.16131910489310824, "grad_norm": 1.0958333015441895, "learning_rate": 0.0001893644370388003, "loss": 2.088473320007324, "step": 1130 }, { "epoch": 0.16274670759127735, "grad_norm": 1.1749203205108643, "learning_rate": 0.00018926922161390146, "loss": 2.091511917114258, "step": 1140 }, { "epoch": 0.16417431028944646, "grad_norm": 1.1318912506103516, "learning_rate": 0.00018917400618900264, "loss": 2.2805349349975588, "step": 1150 }, { "epoch": 0.16560191298761553, "grad_norm": 1.220713496208191, "learning_rate": 0.0001890787907641038, "loss": 2.129801940917969, "step": 1160 }, { "epoch": 0.16702951568578464, "grad_norm": 1.0876924991607666, "learning_rate": 0.00018898357533920498, "loss": 2.0997352600097656, "step": 1170 }, { "epoch": 0.16845711838395375, "grad_norm": 1.175884485244751, "learning_rate": 0.00018888835991430613, "loss": 2.0686941146850586, "step": 1180 }, { "epoch": 0.16988472108212285, "grad_norm": 1.3588541746139526, "learning_rate": 0.00018879314448940732, "loss": 2.2050849914550783, "step": 1190 }, { "epoch": 0.17131232378029196, "grad_norm": 1.26045560836792, "learning_rate": 0.00018869792906450844, "loss": 2.1561567306518556, "step": 1200 }, { "epoch": 0.17273992647846104, "grad_norm": 1.0756542682647705, "learning_rate": 0.00018860271363960963, "loss": 2.163359832763672, "step": 1210 }, { "epoch": 0.17416752917663014, "grad_norm": 1.308608055114746, "learning_rate": 0.00018850749821471078, "loss": 2.209610366821289, "step": 1220 }, { "epoch": 0.17559513187479925, "grad_norm": 1.369969129562378, "learning_rate": 0.00018841228278981196, "loss": 2.3173370361328125, "step": 1230 }, { "epoch": 0.17702273457296835, "grad_norm": 1.2242190837860107, "learning_rate": 0.00018831706736491312, "loss": 2.149942398071289, "step": 1240 }, { "epoch": 0.17845033727113743, "grad_norm": 1.226375937461853, "learning_rate": 0.0001882218519400143, "loss": 2.152383232116699, "step": 1250 }, { "epoch": 0.17987793996930654, "grad_norm": 2.0897111892700195, "learning_rate": 0.00018812663651511546, "loss": 2.070216751098633, "step": 1260 }, { "epoch": 0.18130554266747564, "grad_norm": 1.239769697189331, "learning_rate": 0.00018803142109021664, "loss": 2.1863370895385743, "step": 1270 }, { "epoch": 0.18273314536564475, "grad_norm": 1.2516264915466309, "learning_rate": 0.0001879362056653178, "loss": 2.154579734802246, "step": 1280 }, { "epoch": 0.18416074806381383, "grad_norm": 1.300251841545105, "learning_rate": 0.00018784099024041895, "loss": 2.084653854370117, "step": 1290 }, { "epoch": 0.18558835076198293, "grad_norm": 1.5401705503463745, "learning_rate": 0.0001877457748155201, "loss": 2.1652135848999023, "step": 1300 }, { "epoch": 0.18701595346015204, "grad_norm": 1.388256549835205, "learning_rate": 0.0001876505593906213, "loss": 2.215251350402832, "step": 1310 }, { "epoch": 0.18844355615832115, "grad_norm": 1.0548882484436035, "learning_rate": 0.00018755534396572244, "loss": 2.057052230834961, "step": 1320 }, { "epoch": 0.18987115885649025, "grad_norm": 1.318389654159546, "learning_rate": 0.00018746012854082363, "loss": 2.1746826171875, "step": 1330 }, { "epoch": 0.19129876155465933, "grad_norm": 1.1861886978149414, "learning_rate": 0.00018736491311592478, "loss": 2.107588005065918, "step": 1340 }, { "epoch": 0.19272636425282844, "grad_norm": 1.0857220888137817, "learning_rate": 0.00018726969769102596, "loss": 2.1044857025146486, "step": 1350 }, { "epoch": 0.19415396695099754, "grad_norm": 1.144823670387268, "learning_rate": 0.00018717448226612712, "loss": 2.132158660888672, "step": 1360 }, { "epoch": 0.19558156964916665, "grad_norm": 1.1121206283569336, "learning_rate": 0.0001870792668412283, "loss": 2.1906471252441406, "step": 1370 }, { "epoch": 0.19700917234733573, "grad_norm": 1.2356815338134766, "learning_rate": 0.00018698405141632946, "loss": 2.2056690216064454, "step": 1380 }, { "epoch": 0.19843677504550483, "grad_norm": 1.3212710618972778, "learning_rate": 0.0001868888359914306, "loss": 2.168708610534668, "step": 1390 }, { "epoch": 0.19986437774367394, "grad_norm": 1.1257580518722534, "learning_rate": 0.00018679362056653177, "loss": 2.18591423034668, "step": 1400 }, { "epoch": 0.20129198044184304, "grad_norm": 1.5445501804351807, "learning_rate": 0.00018669840514163295, "loss": 2.206186294555664, "step": 1410 }, { "epoch": 0.20271958314001212, "grad_norm": 1.5519709587097168, "learning_rate": 0.0001866031897167341, "loss": 2.1855854034423827, "step": 1420 }, { "epoch": 0.20414718583818123, "grad_norm": 1.0834031105041504, "learning_rate": 0.0001865079742918353, "loss": 2.1382625579833983, "step": 1430 }, { "epoch": 0.20557478853635033, "grad_norm": 1.0713154077529907, "learning_rate": 0.00018641275886693644, "loss": 2.1709274291992187, "step": 1440 }, { "epoch": 0.20700239123451944, "grad_norm": 1.3530168533325195, "learning_rate": 0.00018631754344203763, "loss": 2.062696075439453, "step": 1450 }, { "epoch": 0.20842999393268855, "grad_norm": 1.0635544061660767, "learning_rate": 0.00018622232801713878, "loss": 2.209213447570801, "step": 1460 }, { "epoch": 0.20985759663085762, "grad_norm": 1.2547699213027954, "learning_rate": 0.00018612711259223996, "loss": 2.111533546447754, "step": 1470 }, { "epoch": 0.21128519932902673, "grad_norm": 1.1931551694869995, "learning_rate": 0.00018603189716734112, "loss": 2.3098974227905273, "step": 1480 }, { "epoch": 0.21271280202719584, "grad_norm": 1.0127021074295044, "learning_rate": 0.00018593668174244227, "loss": 2.1491844177246096, "step": 1490 }, { "epoch": 0.21414040472536494, "grad_norm": 1.3180660009384155, "learning_rate": 0.00018584146631754346, "loss": 2.188648223876953, "step": 1500 }, { "epoch": 0.21556800742353402, "grad_norm": 1.1885178089141846, "learning_rate": 0.0001857462508926446, "loss": 2.0914661407470705, "step": 1510 }, { "epoch": 0.21699561012170313, "grad_norm": 1.1395208835601807, "learning_rate": 0.0001856510354677458, "loss": 2.2511472702026367, "step": 1520 }, { "epoch": 0.21842321281987223, "grad_norm": 1.4837855100631714, "learning_rate": 0.00018555582004284695, "loss": 2.2735774993896483, "step": 1530 }, { "epoch": 0.21985081551804134, "grad_norm": 1.203804612159729, "learning_rate": 0.00018546060461794813, "loss": 2.2867109298706056, "step": 1540 }, { "epoch": 0.22127841821621042, "grad_norm": 0.997947096824646, "learning_rate": 0.0001853653891930493, "loss": 2.184854507446289, "step": 1550 }, { "epoch": 0.22270602091437952, "grad_norm": 1.484794020652771, "learning_rate": 0.00018527017376815047, "loss": 2.232890510559082, "step": 1560 }, { "epoch": 0.22413362361254863, "grad_norm": 1.3194350004196167, "learning_rate": 0.00018517495834325162, "loss": 2.0665668487548827, "step": 1570 }, { "epoch": 0.22556122631071773, "grad_norm": 1.4378708600997925, "learning_rate": 0.00018507974291835278, "loss": 2.025044059753418, "step": 1580 }, { "epoch": 0.22698882900888684, "grad_norm": 1.161460041999817, "learning_rate": 0.00018498452749345394, "loss": 2.2053255081176757, "step": 1590 }, { "epoch": 0.22841643170705592, "grad_norm": 1.0879746675491333, "learning_rate": 0.00018488931206855512, "loss": 2.1081918716430663, "step": 1600 }, { "epoch": 0.22984403440522502, "grad_norm": 1.530381679534912, "learning_rate": 0.00018479409664365627, "loss": 2.135531997680664, "step": 1610 }, { "epoch": 0.23127163710339413, "grad_norm": 1.2382729053497314, "learning_rate": 0.00018469888121875746, "loss": 2.0582502365112303, "step": 1620 }, { "epoch": 0.23269923980156323, "grad_norm": 1.2160850763320923, "learning_rate": 0.0001846036657938586, "loss": 2.191323471069336, "step": 1630 }, { "epoch": 0.2341268424997323, "grad_norm": 1.3606561422348022, "learning_rate": 0.0001845084503689598, "loss": 2.139387512207031, "step": 1640 }, { "epoch": 0.23555444519790142, "grad_norm": 1.0927553176879883, "learning_rate": 0.00018441323494406095, "loss": 2.132997512817383, "step": 1650 }, { "epoch": 0.23698204789607052, "grad_norm": 1.3276400566101074, "learning_rate": 0.00018431801951916213, "loss": 2.1318714141845705, "step": 1660 }, { "epoch": 0.23840965059423963, "grad_norm": 1.4104294776916504, "learning_rate": 0.00018422280409426329, "loss": 2.147476577758789, "step": 1670 }, { "epoch": 0.23983725329240874, "grad_norm": 1.1079225540161133, "learning_rate": 0.00018412758866936444, "loss": 2.0742431640625, "step": 1680 }, { "epoch": 0.24126485599057781, "grad_norm": 1.0819675922393799, "learning_rate": 0.0001840323732444656, "loss": 2.076985549926758, "step": 1690 }, { "epoch": 0.24269245868874692, "grad_norm": 1.1977310180664062, "learning_rate": 0.00018393715781956678, "loss": 2.054951286315918, "step": 1700 }, { "epoch": 0.24412006138691603, "grad_norm": 1.0316016674041748, "learning_rate": 0.00018384194239466793, "loss": 2.102836799621582, "step": 1710 }, { "epoch": 0.24554766408508513, "grad_norm": 1.1931931972503662, "learning_rate": 0.00018374672696976912, "loss": 2.1211891174316406, "step": 1720 }, { "epoch": 0.2469752667832542, "grad_norm": 1.2340854406356812, "learning_rate": 0.00018365151154487027, "loss": 2.2508804321289064, "step": 1730 }, { "epoch": 0.24840286948142332, "grad_norm": 1.4340404272079468, "learning_rate": 0.00018355629611997145, "loss": 2.252923011779785, "step": 1740 }, { "epoch": 0.24983047217959242, "grad_norm": 1.1949214935302734, "learning_rate": 0.0001834610806950726, "loss": 2.2155136108398437, "step": 1750 }, { "epoch": 0.25125807487776153, "grad_norm": 1.0693628787994385, "learning_rate": 0.0001833658652701738, "loss": 2.0989776611328126, "step": 1760 }, { "epoch": 0.25268567757593063, "grad_norm": 1.079628348350525, "learning_rate": 0.00018327064984527492, "loss": 2.235947036743164, "step": 1770 }, { "epoch": 0.25411328027409974, "grad_norm": 1.0427007675170898, "learning_rate": 0.0001831754344203761, "loss": 2.2581958770751953, "step": 1780 }, { "epoch": 0.2555408829722688, "grad_norm": 0.968690812587738, "learning_rate": 0.00018308021899547726, "loss": 2.152431869506836, "step": 1790 }, { "epoch": 0.2569684856704379, "grad_norm": 1.4845409393310547, "learning_rate": 0.00018298500357057844, "loss": 2.082950401306152, "step": 1800 }, { "epoch": 0.258396088368607, "grad_norm": 1.627851128578186, "learning_rate": 0.0001828897881456796, "loss": 2.1082151412963865, "step": 1810 }, { "epoch": 0.2598236910667761, "grad_norm": 1.2222917079925537, "learning_rate": 0.00018279457272078078, "loss": 2.1010499954223634, "step": 1820 }, { "epoch": 0.2612512937649452, "grad_norm": 1.1946543455123901, "learning_rate": 0.00018269935729588193, "loss": 2.20877685546875, "step": 1830 }, { "epoch": 0.2626788964631143, "grad_norm": 1.2480523586273193, "learning_rate": 0.00018260414187098312, "loss": 2.1713356018066405, "step": 1840 }, { "epoch": 0.2641064991612834, "grad_norm": 1.1162548065185547, "learning_rate": 0.00018250892644608427, "loss": 1.989072036743164, "step": 1850 }, { "epoch": 0.26553410185945253, "grad_norm": 1.1317349672317505, "learning_rate": 0.00018241371102118545, "loss": 2.185866928100586, "step": 1860 }, { "epoch": 0.26696170455762164, "grad_norm": 1.2101715803146362, "learning_rate": 0.0001823184955962866, "loss": 2.1214805603027345, "step": 1870 }, { "epoch": 0.2683893072557907, "grad_norm": 1.036651611328125, "learning_rate": 0.00018222328017138776, "loss": 2.2269779205322267, "step": 1880 }, { "epoch": 0.2698169099539598, "grad_norm": 1.2697064876556396, "learning_rate": 0.00018212806474648895, "loss": 2.188285064697266, "step": 1890 }, { "epoch": 0.2712445126521289, "grad_norm": 1.086503028869629, "learning_rate": 0.0001820328493215901, "loss": 2.0915653228759767, "step": 1900 }, { "epoch": 0.272672115350298, "grad_norm": 1.1498035192489624, "learning_rate": 0.00018193763389669129, "loss": 2.1859756469726563, "step": 1910 }, { "epoch": 0.2740997180484671, "grad_norm": 1.1651625633239746, "learning_rate": 0.00018184241847179244, "loss": 2.1086774826049806, "step": 1920 }, { "epoch": 0.2755273207466362, "grad_norm": 1.1872868537902832, "learning_rate": 0.00018174720304689362, "loss": 2.1062952041625977, "step": 1930 }, { "epoch": 0.2769549234448053, "grad_norm": 1.0268173217773438, "learning_rate": 0.00018165198762199478, "loss": 2.056200408935547, "step": 1940 }, { "epoch": 0.27838252614297443, "grad_norm": 1.0335297584533691, "learning_rate": 0.00018155677219709596, "loss": 2.17354793548584, "step": 1950 }, { "epoch": 0.27981012884114354, "grad_norm": 1.1770780086517334, "learning_rate": 0.0001814615567721971, "loss": 2.123767852783203, "step": 1960 }, { "epoch": 0.2812377315393126, "grad_norm": 1.1291216611862183, "learning_rate": 0.00018136634134729827, "loss": 2.1322879791259766, "step": 1970 }, { "epoch": 0.2826653342374817, "grad_norm": 1.0811609029769897, "learning_rate": 0.00018127112592239943, "loss": 2.1403667449951174, "step": 1980 }, { "epoch": 0.2840929369356508, "grad_norm": 1.3379665613174438, "learning_rate": 0.0001811759104975006, "loss": 2.227403450012207, "step": 1990 }, { "epoch": 0.2855205396338199, "grad_norm": 1.2048321962356567, "learning_rate": 0.00018108069507260176, "loss": 2.320987319946289, "step": 2000 }, { "epoch": 0.286948142331989, "grad_norm": 1.1182419061660767, "learning_rate": 0.00018098547964770295, "loss": 2.145694541931152, "step": 2010 }, { "epoch": 0.2883757450301581, "grad_norm": 1.2120338678359985, "learning_rate": 0.0001808902642228041, "loss": 2.193251037597656, "step": 2020 }, { "epoch": 0.2898033477283272, "grad_norm": 1.214379906654358, "learning_rate": 0.00018079504879790528, "loss": 2.183461570739746, "step": 2030 }, { "epoch": 0.2912309504264963, "grad_norm": 1.2853455543518066, "learning_rate": 0.00018069983337300644, "loss": 2.105716323852539, "step": 2040 }, { "epoch": 0.29265855312466543, "grad_norm": 0.9446287155151367, "learning_rate": 0.00018060461794810762, "loss": 2.0502334594726563, "step": 2050 }, { "epoch": 0.2940861558228345, "grad_norm": 1.187010645866394, "learning_rate": 0.00018050940252320875, "loss": 2.0855840682983398, "step": 2060 }, { "epoch": 0.2955137585210036, "grad_norm": 1.1547735929489136, "learning_rate": 0.00018041418709830993, "loss": 2.077458381652832, "step": 2070 }, { "epoch": 0.2969413612191727, "grad_norm": 1.1305893659591675, "learning_rate": 0.0001803189716734111, "loss": 2.1422780990600585, "step": 2080 }, { "epoch": 0.2983689639173418, "grad_norm": 1.2219164371490479, "learning_rate": 0.00018022375624851227, "loss": 2.1557044982910156, "step": 2090 }, { "epoch": 0.2997965666155109, "grad_norm": 1.303414225578308, "learning_rate": 0.00018012854082361343, "loss": 2.0842567443847657, "step": 2100 }, { "epoch": 0.30122416931368, "grad_norm": 1.1244888305664062, "learning_rate": 0.0001800333253987146, "loss": 2.0076160430908203, "step": 2110 }, { "epoch": 0.3026517720118491, "grad_norm": 1.265690803527832, "learning_rate": 0.00017993810997381576, "loss": 2.2838394165039064, "step": 2120 }, { "epoch": 0.3040793747100182, "grad_norm": 1.0629686117172241, "learning_rate": 0.00017984289454891695, "loss": 2.0467424392700195, "step": 2130 }, { "epoch": 0.3055069774081873, "grad_norm": 1.1269965171813965, "learning_rate": 0.0001797476791240181, "loss": 2.059001159667969, "step": 2140 }, { "epoch": 0.3069345801063564, "grad_norm": 1.0063384771347046, "learning_rate": 0.00017965246369911926, "loss": 2.054671859741211, "step": 2150 }, { "epoch": 0.3083621828045255, "grad_norm": 1.696047306060791, "learning_rate": 0.0001795572482742204, "loss": 2.112416648864746, "step": 2160 }, { "epoch": 0.3097897855026946, "grad_norm": 1.1593190431594849, "learning_rate": 0.0001794620328493216, "loss": 1.9681781768798827, "step": 2170 }, { "epoch": 0.3112173882008637, "grad_norm": 1.1535143852233887, "learning_rate": 0.00017936681742442275, "loss": 1.9573915481567383, "step": 2180 }, { "epoch": 0.3126449908990328, "grad_norm": 1.2937277555465698, "learning_rate": 0.00017927160199952393, "loss": 2.1334136962890624, "step": 2190 }, { "epoch": 0.3140725935972019, "grad_norm": 1.1130894422531128, "learning_rate": 0.0001791763865746251, "loss": 2.054120635986328, "step": 2200 }, { "epoch": 0.315500196295371, "grad_norm": 1.1255532503128052, "learning_rate": 0.00017908117114972627, "loss": 2.206426811218262, "step": 2210 }, { "epoch": 0.3169277989935401, "grad_norm": 1.373878002166748, "learning_rate": 0.00017898595572482743, "loss": 2.016497802734375, "step": 2220 }, { "epoch": 0.31835540169170917, "grad_norm": 0.9502859115600586, "learning_rate": 0.0001788907402999286, "loss": 2.1160964965820312, "step": 2230 }, { "epoch": 0.3197830043898783, "grad_norm": 1.199417233467102, "learning_rate": 0.00017879552487502976, "loss": 2.219405746459961, "step": 2240 }, { "epoch": 0.3212106070880474, "grad_norm": 1.2231884002685547, "learning_rate": 0.00017870030945013092, "loss": 2.1173473358154298, "step": 2250 }, { "epoch": 0.3226382097862165, "grad_norm": 1.2296863794326782, "learning_rate": 0.0001786050940252321, "loss": 2.212135124206543, "step": 2260 }, { "epoch": 0.3240658124843856, "grad_norm": 1.2749266624450684, "learning_rate": 0.00017850987860033326, "loss": 2.0644044876098633, "step": 2270 }, { "epoch": 0.3254934151825547, "grad_norm": 1.0321651697158813, "learning_rate": 0.00017841466317543444, "loss": 2.126047897338867, "step": 2280 }, { "epoch": 0.3269210178807238, "grad_norm": 1.131618857383728, "learning_rate": 0.0001783194477505356, "loss": 2.0301258087158205, "step": 2290 }, { "epoch": 0.3283486205788929, "grad_norm": 0.9653699994087219, "learning_rate": 0.00017822423232563678, "loss": 2.172132873535156, "step": 2300 }, { "epoch": 0.329776223277062, "grad_norm": 1.1435012817382812, "learning_rate": 0.00017812901690073793, "loss": 2.1726940155029295, "step": 2310 }, { "epoch": 0.33120382597523107, "grad_norm": 1.3378164768218994, "learning_rate": 0.00017803380147583911, "loss": 2.1459815979003904, "step": 2320 }, { "epoch": 0.3326314286734002, "grad_norm": 1.0936335325241089, "learning_rate": 0.00017793858605094027, "loss": 2.11383113861084, "step": 2330 }, { "epoch": 0.3340590313715693, "grad_norm": 1.7248929738998413, "learning_rate": 0.00017784337062604142, "loss": 2.060173225402832, "step": 2340 }, { "epoch": 0.3354866340697384, "grad_norm": 1.4490845203399658, "learning_rate": 0.00017774815520114258, "loss": 2.168626403808594, "step": 2350 }, { "epoch": 0.3369142367679075, "grad_norm": 1.4273464679718018, "learning_rate": 0.00017765293977624376, "loss": 2.1859357833862303, "step": 2360 }, { "epoch": 0.3383418394660766, "grad_norm": 1.1335994005203247, "learning_rate": 0.00017755772435134492, "loss": 2.2323787689208983, "step": 2370 }, { "epoch": 0.3397694421642457, "grad_norm": 1.0822454690933228, "learning_rate": 0.0001774625089264461, "loss": 2.1954828262329102, "step": 2380 }, { "epoch": 0.3411970448624148, "grad_norm": 1.3556091785430908, "learning_rate": 0.00017736729350154726, "loss": 2.086533546447754, "step": 2390 }, { "epoch": 0.3426246475605839, "grad_norm": 1.0348728895187378, "learning_rate": 0.00017727207807664844, "loss": 2.063125419616699, "step": 2400 }, { "epoch": 0.34405225025875297, "grad_norm": 0.9327099323272705, "learning_rate": 0.0001771768626517496, "loss": 2.2520879745483398, "step": 2410 }, { "epoch": 0.3454798529569221, "grad_norm": 1.161893367767334, "learning_rate": 0.00017708164722685078, "loss": 1.9637351989746095, "step": 2420 }, { "epoch": 0.3469074556550912, "grad_norm": 1.0252056121826172, "learning_rate": 0.00017698643180195193, "loss": 2.105502700805664, "step": 2430 }, { "epoch": 0.3483350583532603, "grad_norm": 1.0173977613449097, "learning_rate": 0.00017689121637705309, "loss": 2.1620779037475586, "step": 2440 }, { "epoch": 0.3497626610514294, "grad_norm": 1.0667582750320435, "learning_rate": 0.00017679600095215424, "loss": 2.1498435974121093, "step": 2450 }, { "epoch": 0.3511902637495985, "grad_norm": 1.1689473390579224, "learning_rate": 0.00017670078552725542, "loss": 2.010816192626953, "step": 2460 }, { "epoch": 0.3526178664477676, "grad_norm": 1.1713252067565918, "learning_rate": 0.00017660557010235658, "loss": 2.1770673751831056, "step": 2470 }, { "epoch": 0.3540454691459367, "grad_norm": 1.1741546392440796, "learning_rate": 0.00017651035467745776, "loss": 2.2705522537231446, "step": 2480 }, { "epoch": 0.35547307184410576, "grad_norm": 1.1165494918823242, "learning_rate": 0.00017641513925255892, "loss": 2.2441898345947267, "step": 2490 }, { "epoch": 0.35690067454227486, "grad_norm": 1.128848910331726, "learning_rate": 0.0001763199238276601, "loss": 1.966132354736328, "step": 2500 }, { "epoch": 0.35832827724044397, "grad_norm": 1.1416059732437134, "learning_rate": 0.00017622470840276125, "loss": 2.129666328430176, "step": 2510 }, { "epoch": 0.3597558799386131, "grad_norm": 1.3967465162277222, "learning_rate": 0.00017612949297786244, "loss": 2.068239212036133, "step": 2520 }, { "epoch": 0.3611834826367822, "grad_norm": 1.1982859373092651, "learning_rate": 0.00017603427755296357, "loss": 2.0356422424316407, "step": 2530 }, { "epoch": 0.3626110853349513, "grad_norm": 1.2561085224151611, "learning_rate": 0.00017593906212806475, "loss": 2.1692792892456056, "step": 2540 }, { "epoch": 0.3640386880331204, "grad_norm": 1.2402371168136597, "learning_rate": 0.0001758438467031659, "loss": 2.018773078918457, "step": 2550 }, { "epoch": 0.3654662907312895, "grad_norm": 1.0577623844146729, "learning_rate": 0.00017574863127826709, "loss": 2.2117090225219727, "step": 2560 }, { "epoch": 0.3668938934294586, "grad_norm": 1.3217538595199585, "learning_rate": 0.00017565341585336824, "loss": 2.19326229095459, "step": 2570 }, { "epoch": 0.36832149612762766, "grad_norm": 1.2093907594680786, "learning_rate": 0.00017555820042846942, "loss": 2.072580909729004, "step": 2580 }, { "epoch": 0.36974909882579676, "grad_norm": 1.1893237829208374, "learning_rate": 0.00017546298500357058, "loss": 2.0644367218017576, "step": 2590 }, { "epoch": 0.37117670152396587, "grad_norm": 1.116391658782959, "learning_rate": 0.00017536776957867176, "loss": 2.0989030838012694, "step": 2600 }, { "epoch": 0.372604304222135, "grad_norm": 1.1465907096862793, "learning_rate": 0.00017527255415377292, "loss": 2.22904109954834, "step": 2610 }, { "epoch": 0.3740319069203041, "grad_norm": 1.2656993865966797, "learning_rate": 0.0001751773387288741, "loss": 2.162650489807129, "step": 2620 }, { "epoch": 0.3754595096184732, "grad_norm": 1.0960391759872437, "learning_rate": 0.00017508212330397525, "loss": 2.0461801528930663, "step": 2630 }, { "epoch": 0.3768871123166423, "grad_norm": 1.1352872848510742, "learning_rate": 0.0001749869078790764, "loss": 2.142223930358887, "step": 2640 }, { "epoch": 0.3783147150148114, "grad_norm": 1.3211382627487183, "learning_rate": 0.0001748916924541776, "loss": 2.086302375793457, "step": 2650 }, { "epoch": 0.3797423177129805, "grad_norm": 1.3286374807357788, "learning_rate": 0.00017479647702927875, "loss": 2.1871686935424806, "step": 2660 }, { "epoch": 0.38116992041114955, "grad_norm": 1.1504572629928589, "learning_rate": 0.00017470126160437993, "loss": 2.2135593414306642, "step": 2670 }, { "epoch": 0.38259752310931866, "grad_norm": 1.1061992645263672, "learning_rate": 0.00017460604617948108, "loss": 1.9979249954223632, "step": 2680 }, { "epoch": 0.38402512580748777, "grad_norm": 1.21872079372406, "learning_rate": 0.00017451083075458227, "loss": 2.1905019760131834, "step": 2690 }, { "epoch": 0.38545272850565687, "grad_norm": 0.9169778823852539, "learning_rate": 0.00017441561532968342, "loss": 2.096824836730957, "step": 2700 }, { "epoch": 0.386880331203826, "grad_norm": 1.1782190799713135, "learning_rate": 0.0001743203999047846, "loss": 2.1858846664428713, "step": 2710 }, { "epoch": 0.3883079339019951, "grad_norm": 1.0241059064865112, "learning_rate": 0.00017422518447988573, "loss": 2.1724557876586914, "step": 2720 }, { "epoch": 0.3897355366001642, "grad_norm": 1.3461824655532837, "learning_rate": 0.00017412996905498692, "loss": 2.0981250762939454, "step": 2730 }, { "epoch": 0.3911631392983333, "grad_norm": 1.231691837310791, "learning_rate": 0.00017403475363008807, "loss": 2.000545692443848, "step": 2740 }, { "epoch": 0.39259074199650235, "grad_norm": 1.1789937019348145, "learning_rate": 0.00017393953820518925, "loss": 1.9449064254760742, "step": 2750 }, { "epoch": 0.39401834469467145, "grad_norm": 1.119533896446228, "learning_rate": 0.0001738443227802904, "loss": 2.0872426986694337, "step": 2760 }, { "epoch": 0.39544594739284056, "grad_norm": 1.034921407699585, "learning_rate": 0.0001737491073553916, "loss": 2.0502944946289063, "step": 2770 }, { "epoch": 0.39687355009100966, "grad_norm": 1.1675975322723389, "learning_rate": 0.00017365389193049275, "loss": 2.0751968383789063, "step": 2780 }, { "epoch": 0.39830115278917877, "grad_norm": 0.9863481521606445, "learning_rate": 0.00017355867650559393, "loss": 2.216383171081543, "step": 2790 }, { "epoch": 0.3997287554873479, "grad_norm": 1.0700448751449585, "learning_rate": 0.00017346346108069508, "loss": 2.037568473815918, "step": 2800 }, { "epoch": 0.401156358185517, "grad_norm": 1.5234819650650024, "learning_rate": 0.00017336824565579627, "loss": 2.135871696472168, "step": 2810 }, { "epoch": 0.4025839608836861, "grad_norm": 1.312696933746338, "learning_rate": 0.0001732730302308974, "loss": 2.123103713989258, "step": 2820 }, { "epoch": 0.4040115635818552, "grad_norm": 1.4110193252563477, "learning_rate": 0.00017317781480599858, "loss": 2.2066755294799805, "step": 2830 }, { "epoch": 0.40543916628002424, "grad_norm": 1.1026965379714966, "learning_rate": 0.00017308259938109973, "loss": 2.026701736450195, "step": 2840 }, { "epoch": 0.40686676897819335, "grad_norm": 1.1809606552124023, "learning_rate": 0.00017298738395620092, "loss": 2.013088035583496, "step": 2850 }, { "epoch": 0.40829437167636246, "grad_norm": 1.281699538230896, "learning_rate": 0.00017289216853130207, "loss": 2.212848663330078, "step": 2860 }, { "epoch": 0.40972197437453156, "grad_norm": 1.4359161853790283, "learning_rate": 0.00017279695310640325, "loss": 2.3273263931274415, "step": 2870 }, { "epoch": 0.41114957707270067, "grad_norm": 1.1231915950775146, "learning_rate": 0.0001727017376815044, "loss": 2.024250602722168, "step": 2880 }, { "epoch": 0.4125771797708698, "grad_norm": 1.1361498832702637, "learning_rate": 0.0001726065222566056, "loss": 2.0198749542236327, "step": 2890 }, { "epoch": 0.4140047824690389, "grad_norm": 1.277862310409546, "learning_rate": 0.00017251130683170675, "loss": 2.0423700332641603, "step": 2900 }, { "epoch": 0.415432385167208, "grad_norm": 1.3887676000595093, "learning_rate": 0.0001724160914068079, "loss": 2.1957927703857423, "step": 2910 }, { "epoch": 0.4168599878653771, "grad_norm": 1.102116584777832, "learning_rate": 0.00017232087598190906, "loss": 1.9576358795166016, "step": 2920 }, { "epoch": 0.41828759056354614, "grad_norm": 1.658210277557373, "learning_rate": 0.00017222566055701024, "loss": 2.1199745178222655, "step": 2930 }, { "epoch": 0.41971519326171525, "grad_norm": 1.3684505224227905, "learning_rate": 0.0001721304451321114, "loss": 2.1091520309448244, "step": 2940 }, { "epoch": 0.42114279595988435, "grad_norm": 1.1813336610794067, "learning_rate": 0.00017203522970721258, "loss": 2.0536418914794923, "step": 2950 }, { "epoch": 0.42257039865805346, "grad_norm": 1.2007261514663696, "learning_rate": 0.00017194001428231373, "loss": 2.053822708129883, "step": 2960 }, { "epoch": 0.42399800135622256, "grad_norm": 1.0350509881973267, "learning_rate": 0.00017184479885741491, "loss": 1.9133419036865233, "step": 2970 }, { "epoch": 0.42542560405439167, "grad_norm": 1.0523207187652588, "learning_rate": 0.00017174958343251607, "loss": 2.0608654022216797, "step": 2980 }, { "epoch": 0.4268532067525608, "grad_norm": 0.952148973941803, "learning_rate": 0.00017165436800761725, "loss": 2.1483791351318358, "step": 2990 }, { "epoch": 0.4282808094507299, "grad_norm": 1.0913255214691162, "learning_rate": 0.0001715591525827184, "loss": 2.155402946472168, "step": 3000 }, { "epoch": 0.429708412148899, "grad_norm": 1.05339777469635, "learning_rate": 0.00017146393715781956, "loss": 2.1396499633789063, "step": 3010 }, { "epoch": 0.43113601484706804, "grad_norm": 1.2329785823822021, "learning_rate": 0.00017136872173292075, "loss": 2.2712575912475588, "step": 3020 }, { "epoch": 0.43256361754523714, "grad_norm": 1.1520178318023682, "learning_rate": 0.0001712735063080219, "loss": 2.0944181442260743, "step": 3030 }, { "epoch": 0.43399122024340625, "grad_norm": 1.1466962099075317, "learning_rate": 0.00017117829088312308, "loss": 2.064352607727051, "step": 3040 }, { "epoch": 0.43541882294157536, "grad_norm": 1.2289594411849976, "learning_rate": 0.00017108307545822424, "loss": 2.2511341094970705, "step": 3050 }, { "epoch": 0.43684642563974446, "grad_norm": 1.0378469228744507, "learning_rate": 0.00017098786003332542, "loss": 2.0552734375, "step": 3060 }, { "epoch": 0.43827402833791357, "grad_norm": 1.1334617137908936, "learning_rate": 0.00017089264460842658, "loss": 2.1504894256591798, "step": 3070 }, { "epoch": 0.4397016310360827, "grad_norm": 1.2314560413360596, "learning_rate": 0.00017079742918352776, "loss": 2.1562742233276366, "step": 3080 }, { "epoch": 0.4411292337342518, "grad_norm": 1.1508898735046387, "learning_rate": 0.00017070221375862891, "loss": 2.0091711044311524, "step": 3090 }, { "epoch": 0.44255683643242083, "grad_norm": 1.7214949131011963, "learning_rate": 0.00017060699833373007, "loss": 2.084541893005371, "step": 3100 }, { "epoch": 0.44398443913058994, "grad_norm": 1.452836036682129, "learning_rate": 0.00017051178290883122, "loss": 2.167698097229004, "step": 3110 }, { "epoch": 0.44541204182875904, "grad_norm": 1.1506155729293823, "learning_rate": 0.0001704165674839324, "loss": 2.1772172927856444, "step": 3120 }, { "epoch": 0.44683964452692815, "grad_norm": 1.6658185720443726, "learning_rate": 0.00017032135205903356, "loss": 2.1950592041015624, "step": 3130 }, { "epoch": 0.44826724722509725, "grad_norm": 1.0545748472213745, "learning_rate": 0.00017022613663413474, "loss": 2.052364730834961, "step": 3140 }, { "epoch": 0.44969484992326636, "grad_norm": 1.3302888870239258, "learning_rate": 0.0001701309212092359, "loss": 2.2422548294067384, "step": 3150 }, { "epoch": 0.45112245262143547, "grad_norm": 1.4070645570755005, "learning_rate": 0.00017003570578433708, "loss": 2.1141366958618164, "step": 3160 }, { "epoch": 0.45255005531960457, "grad_norm": 1.0993238687515259, "learning_rate": 0.00016994049035943824, "loss": 2.1006706237792967, "step": 3170 }, { "epoch": 0.4539776580177737, "grad_norm": 1.1617511510849, "learning_rate": 0.00016984527493453942, "loss": 1.9723478317260743, "step": 3180 }, { "epoch": 0.45540526071594273, "grad_norm": 1.5463227033615112, "learning_rate": 0.00016975005950964058, "loss": 2.104926109313965, "step": 3190 }, { "epoch": 0.45683286341411183, "grad_norm": 1.2094980478286743, "learning_rate": 0.00016965484408474173, "loss": 1.9171577453613282, "step": 3200 }, { "epoch": 0.45826046611228094, "grad_norm": 1.103075385093689, "learning_rate": 0.00016955962865984289, "loss": 2.1229742050170897, "step": 3210 }, { "epoch": 0.45968806881045005, "grad_norm": 1.3477221727371216, "learning_rate": 0.00016946441323494407, "loss": 2.0110040664672852, "step": 3220 }, { "epoch": 0.46111567150861915, "grad_norm": 1.1711978912353516, "learning_rate": 0.00016936919781004522, "loss": 2.1763769149780274, "step": 3230 }, { "epoch": 0.46254327420678826, "grad_norm": 1.1173983812332153, "learning_rate": 0.0001692739823851464, "loss": 2.0923931121826174, "step": 3240 }, { "epoch": 0.46397087690495736, "grad_norm": 1.3516672849655151, "learning_rate": 0.00016917876696024756, "loss": 2.1028839111328126, "step": 3250 }, { "epoch": 0.46539847960312647, "grad_norm": 1.187572717666626, "learning_rate": 0.00016908355153534874, "loss": 2.039937973022461, "step": 3260 }, { "epoch": 0.4668260823012956, "grad_norm": 1.3026293516159058, "learning_rate": 0.0001689883361104499, "loss": 2.0570196151733398, "step": 3270 }, { "epoch": 0.4682536849994646, "grad_norm": 1.1978344917297363, "learning_rate": 0.00016889312068555108, "loss": 2.077353858947754, "step": 3280 }, { "epoch": 0.46968128769763373, "grad_norm": 1.0989017486572266, "learning_rate": 0.0001687979052606522, "loss": 1.9933378219604492, "step": 3290 }, { "epoch": 0.47110889039580284, "grad_norm": 1.1777777671813965, "learning_rate": 0.0001687026898357534, "loss": 2.197841262817383, "step": 3300 }, { "epoch": 0.47253649309397194, "grad_norm": 1.8024139404296875, "learning_rate": 0.00016860747441085455, "loss": 2.1699588775634764, "step": 3310 }, { "epoch": 0.47396409579214105, "grad_norm": 1.0226041078567505, "learning_rate": 0.00016851225898595573, "loss": 2.0379987716674806, "step": 3320 }, { "epoch": 0.47539169849031016, "grad_norm": 2.29010009765625, "learning_rate": 0.00016841704356105689, "loss": 2.08388614654541, "step": 3330 }, { "epoch": 0.47681930118847926, "grad_norm": 0.9974448084831238, "learning_rate": 0.00016832182813615807, "loss": 2.0867422103881834, "step": 3340 }, { "epoch": 0.47824690388664837, "grad_norm": 1.1733876466751099, "learning_rate": 0.00016822661271125922, "loss": 2.038913917541504, "step": 3350 }, { "epoch": 0.4796745065848175, "grad_norm": 1.0234184265136719, "learning_rate": 0.0001681313972863604, "loss": 2.1241840362548827, "step": 3360 }, { "epoch": 0.4811021092829865, "grad_norm": 1.4544063806533813, "learning_rate": 0.00016803618186146156, "loss": 2.044015121459961, "step": 3370 }, { "epoch": 0.48252971198115563, "grad_norm": 1.368973731994629, "learning_rate": 0.00016794096643656274, "loss": 2.019906425476074, "step": 3380 }, { "epoch": 0.48395731467932473, "grad_norm": 1.168421745300293, "learning_rate": 0.0001678457510116639, "loss": 2.0222070693969725, "step": 3390 }, { "epoch": 0.48538491737749384, "grad_norm": 1.2671949863433838, "learning_rate": 0.00016775053558676505, "loss": 2.1466264724731445, "step": 3400 }, { "epoch": 0.48681252007566295, "grad_norm": 1.0771664381027222, "learning_rate": 0.00016765532016186624, "loss": 2.0024179458618163, "step": 3410 }, { "epoch": 0.48824012277383205, "grad_norm": 1.2461390495300293, "learning_rate": 0.0001675601047369674, "loss": 2.0828195571899415, "step": 3420 }, { "epoch": 0.48966772547200116, "grad_norm": 1.6071606874465942, "learning_rate": 0.00016746488931206857, "loss": 1.9849077224731446, "step": 3430 }, { "epoch": 0.49109532817017026, "grad_norm": 1.0467002391815186, "learning_rate": 0.00016736967388716973, "loss": 2.0019205093383787, "step": 3440 }, { "epoch": 0.4925229308683393, "grad_norm": 1.0588996410369873, "learning_rate": 0.0001672744584622709, "loss": 2.0819076538085937, "step": 3450 }, { "epoch": 0.4939505335665084, "grad_norm": 0.9776011109352112, "learning_rate": 0.00016717924303737207, "loss": 1.9890733718872071, "step": 3460 }, { "epoch": 0.4953781362646775, "grad_norm": 1.439799427986145, "learning_rate": 0.00016708402761247325, "loss": 2.1648094177246096, "step": 3470 }, { "epoch": 0.49680573896284663, "grad_norm": 1.063666820526123, "learning_rate": 0.0001669888121875744, "loss": 1.9934938430786133, "step": 3480 }, { "epoch": 0.49823334166101574, "grad_norm": 1.0971171855926514, "learning_rate": 0.00016689359676267556, "loss": 2.0755395889282227, "step": 3490 }, { "epoch": 0.49966094435918484, "grad_norm": 1.3422017097473145, "learning_rate": 0.00016679838133777672, "loss": 2.080609130859375, "step": 3500 }, { "epoch": 0.501088547057354, "grad_norm": 1.3471558094024658, "learning_rate": 0.0001667031659128779, "loss": 2.0093482971191405, "step": 3510 }, { "epoch": 0.5025161497555231, "grad_norm": 1.0592118501663208, "learning_rate": 0.00016660795048797905, "loss": 1.997859573364258, "step": 3520 }, { "epoch": 0.5039437524536922, "grad_norm": 1.2080049514770508, "learning_rate": 0.00016651273506308024, "loss": 2.133045959472656, "step": 3530 }, { "epoch": 0.5053713551518613, "grad_norm": 0.9516943097114563, "learning_rate": 0.0001664175196381814, "loss": 2.04406795501709, "step": 3540 }, { "epoch": 0.5067989578500304, "grad_norm": 0.9474464058876038, "learning_rate": 0.00016632230421328257, "loss": 2.161180877685547, "step": 3550 }, { "epoch": 0.5082265605481995, "grad_norm": 1.1096711158752441, "learning_rate": 0.00016622708878838373, "loss": 2.1555700302124023, "step": 3560 }, { "epoch": 0.5096541632463686, "grad_norm": 1.3584140539169312, "learning_rate": 0.0001661318733634849, "loss": 2.018060302734375, "step": 3570 }, { "epoch": 0.5110817659445376, "grad_norm": 1.6246408224105835, "learning_rate": 0.00016603665793858604, "loss": 2.2220287322998047, "step": 3580 }, { "epoch": 0.5125093686427067, "grad_norm": 1.2138937711715698, "learning_rate": 0.00016594144251368722, "loss": 2.069137954711914, "step": 3590 }, { "epoch": 0.5139369713408758, "grad_norm": 1.2093039751052856, "learning_rate": 0.00016584622708878838, "loss": 2.109259033203125, "step": 3600 }, { "epoch": 0.5153645740390449, "grad_norm": 1.1119673252105713, "learning_rate": 0.00016575101166388956, "loss": 2.055547904968262, "step": 3610 }, { "epoch": 0.516792176737214, "grad_norm": 1.1239557266235352, "learning_rate": 0.00016565579623899071, "loss": 2.154520606994629, "step": 3620 }, { "epoch": 0.5182197794353831, "grad_norm": 1.1713887453079224, "learning_rate": 0.0001655605808140919, "loss": 2.1828012466430664, "step": 3630 }, { "epoch": 0.5196473821335522, "grad_norm": 1.0080145597457886, "learning_rate": 0.00016546536538919305, "loss": 2.08593635559082, "step": 3640 }, { "epoch": 0.5210749848317213, "grad_norm": 1.0677450895309448, "learning_rate": 0.00016537014996429424, "loss": 2.1263172149658205, "step": 3650 }, { "epoch": 0.5225025875298904, "grad_norm": 1.5995664596557617, "learning_rate": 0.0001652749345393954, "loss": 2.1577287673950196, "step": 3660 }, { "epoch": 0.5239301902280595, "grad_norm": 1.1336824893951416, "learning_rate": 0.00016517971911449657, "loss": 2.0607637405395507, "step": 3670 }, { "epoch": 0.5253577929262286, "grad_norm": 1.4898009300231934, "learning_rate": 0.0001650845036895977, "loss": 2.0850154876708986, "step": 3680 }, { "epoch": 0.5267853956243977, "grad_norm": 1.153604507446289, "learning_rate": 0.00016498928826469888, "loss": 2.0568546295166015, "step": 3690 }, { "epoch": 0.5282129983225669, "grad_norm": 0.9742231965065002, "learning_rate": 0.00016489407283980004, "loss": 2.1295637130737304, "step": 3700 }, { "epoch": 0.529640601020736, "grad_norm": 1.0828098058700562, "learning_rate": 0.00016479885741490122, "loss": 2.2455913543701174, "step": 3710 }, { "epoch": 0.5310682037189051, "grad_norm": 1.1993688344955444, "learning_rate": 0.00016470364199000238, "loss": 2.065316009521484, "step": 3720 }, { "epoch": 0.5324958064170742, "grad_norm": 0.9875587224960327, "learning_rate": 0.00016460842656510356, "loss": 2.167093849182129, "step": 3730 }, { "epoch": 0.5339234091152433, "grad_norm": 1.587498426437378, "learning_rate": 0.00016451321114020471, "loss": 2.123185157775879, "step": 3740 }, { "epoch": 0.5353510118134124, "grad_norm": 1.0599406957626343, "learning_rate": 0.0001644179957153059, "loss": 1.9770395278930664, "step": 3750 }, { "epoch": 0.5367786145115814, "grad_norm": 0.9904677867889404, "learning_rate": 0.00016432278029040705, "loss": 2.1363361358642576, "step": 3760 }, { "epoch": 0.5382062172097505, "grad_norm": 1.1583956480026245, "learning_rate": 0.0001642275648655082, "loss": 2.163792610168457, "step": 3770 }, { "epoch": 0.5396338199079196, "grad_norm": 1.0722836256027222, "learning_rate": 0.0001641323494406094, "loss": 2.143117141723633, "step": 3780 }, { "epoch": 0.5410614226060887, "grad_norm": 1.3460588455200195, "learning_rate": 0.00016403713401571055, "loss": 2.1735687255859375, "step": 3790 }, { "epoch": 0.5424890253042578, "grad_norm": 1.1308567523956299, "learning_rate": 0.00016394191859081173, "loss": 2.1688426971435546, "step": 3800 }, { "epoch": 0.5439166280024269, "grad_norm": 1.20809805393219, "learning_rate": 0.00016384670316591288, "loss": 2.1337045669555663, "step": 3810 }, { "epoch": 0.545344230700596, "grad_norm": 1.051905870437622, "learning_rate": 0.00016375148774101407, "loss": 2.0859872817993166, "step": 3820 }, { "epoch": 0.5467718333987651, "grad_norm": 1.2605935335159302, "learning_rate": 0.00016365627231611522, "loss": 2.1538076400756836, "step": 3830 }, { "epoch": 0.5481994360969342, "grad_norm": 1.1376152038574219, "learning_rate": 0.0001635610568912164, "loss": 2.028926467895508, "step": 3840 }, { "epoch": 0.5496270387951033, "grad_norm": 2.1268208026885986, "learning_rate": 0.00016346584146631756, "loss": 2.225727653503418, "step": 3850 }, { "epoch": 0.5510546414932724, "grad_norm": 1.430724859237671, "learning_rate": 0.00016337062604141874, "loss": 2.0055526733398437, "step": 3860 }, { "epoch": 0.5524822441914415, "grad_norm": 1.0673465728759766, "learning_rate": 0.00016327541061651987, "loss": 2.0893600463867186, "step": 3870 }, { "epoch": 0.5539098468896106, "grad_norm": 1.2765331268310547, "learning_rate": 0.00016318019519162105, "loss": 2.008353042602539, "step": 3880 }, { "epoch": 0.5553374495877798, "grad_norm": 1.0461541414260864, "learning_rate": 0.0001630849797667222, "loss": 2.1346414566040037, "step": 3890 }, { "epoch": 0.5567650522859489, "grad_norm": 1.28388512134552, "learning_rate": 0.0001629897643418234, "loss": 2.1297496795654296, "step": 3900 }, { "epoch": 0.558192654984118, "grad_norm": 1.1615244150161743, "learning_rate": 0.00016289454891692454, "loss": 2.078843116760254, "step": 3910 }, { "epoch": 0.5596202576822871, "grad_norm": 1.248067021369934, "learning_rate": 0.00016279933349202573, "loss": 2.2453628540039063, "step": 3920 }, { "epoch": 0.5610478603804561, "grad_norm": 1.1647175550460815, "learning_rate": 0.00016270411806712688, "loss": 2.271023178100586, "step": 3930 }, { "epoch": 0.5624754630786252, "grad_norm": 0.9548137784004211, "learning_rate": 0.00016260890264222806, "loss": 2.068301200866699, "step": 3940 }, { "epoch": 0.5639030657767943, "grad_norm": 1.084067940711975, "learning_rate": 0.00016251368721732922, "loss": 2.0725963592529295, "step": 3950 }, { "epoch": 0.5653306684749634, "grad_norm": 1.6000462770462036, "learning_rate": 0.00016241847179243038, "loss": 2.205929756164551, "step": 3960 }, { "epoch": 0.5667582711731325, "grad_norm": 1.0657289028167725, "learning_rate": 0.00016232325636753153, "loss": 2.224954032897949, "step": 3970 }, { "epoch": 0.5681858738713016, "grad_norm": 1.0999115705490112, "learning_rate": 0.0001622280409426327, "loss": 2.0713016510009767, "step": 3980 }, { "epoch": 0.5696134765694707, "grad_norm": 1.55386221408844, "learning_rate": 0.00016214234706022376, "loss": 2.0461357116699217, "step": 3990 }, { "epoch": 0.5710410792676398, "grad_norm": 1.4208126068115234, "learning_rate": 0.00016204713163532492, "loss": 1.941202735900879, "step": 4000 }, { "epoch": 0.5724686819658089, "grad_norm": 1.118394374847412, "learning_rate": 0.0001619519162104261, "loss": 2.088357162475586, "step": 4010 }, { "epoch": 0.573896284663978, "grad_norm": 1.3743064403533936, "learning_rate": 0.00016185670078552725, "loss": 2.0932437896728517, "step": 4020 }, { "epoch": 0.5753238873621471, "grad_norm": 1.1113500595092773, "learning_rate": 0.00016176148536062844, "loss": 2.1681371688842774, "step": 4030 }, { "epoch": 0.5767514900603162, "grad_norm": 1.1337480545043945, "learning_rate": 0.0001616662699357296, "loss": 2.390884780883789, "step": 4040 }, { "epoch": 0.5781790927584853, "grad_norm": 1.0913474559783936, "learning_rate": 0.00016157105451083077, "loss": 2.0483430862426757, "step": 4050 }, { "epoch": 0.5796066954566544, "grad_norm": 1.0958749055862427, "learning_rate": 0.00016147583908593193, "loss": 2.0507534027099608, "step": 4060 }, { "epoch": 0.5810342981548235, "grad_norm": 1.0328621864318848, "learning_rate": 0.0001613806236610331, "loss": 2.081440734863281, "step": 4070 }, { "epoch": 0.5824619008529927, "grad_norm": 1.0732722282409668, "learning_rate": 0.00016128540823613427, "loss": 1.9593217849731446, "step": 4080 }, { "epoch": 0.5838895035511618, "grad_norm": 1.209100365638733, "learning_rate": 0.00016119019281123542, "loss": 1.9060861587524414, "step": 4090 }, { "epoch": 0.5853171062493309, "grad_norm": 1.7121407985687256, "learning_rate": 0.00016109497738633658, "loss": 2.1000993728637694, "step": 4100 }, { "epoch": 0.5867447089474999, "grad_norm": 1.263508677482605, "learning_rate": 0.00016099976196143776, "loss": 2.03182373046875, "step": 4110 }, { "epoch": 0.588172311645669, "grad_norm": 1.090971827507019, "learning_rate": 0.00016090454653653892, "loss": 2.046316909790039, "step": 4120 }, { "epoch": 0.5895999143438381, "grad_norm": 1.1175658702850342, "learning_rate": 0.0001608093311116401, "loss": 2.025002098083496, "step": 4130 }, { "epoch": 0.5910275170420072, "grad_norm": 1.078690528869629, "learning_rate": 0.00016071411568674125, "loss": 2.0121585845947267, "step": 4140 }, { "epoch": 0.5924551197401763, "grad_norm": 1.0645606517791748, "learning_rate": 0.00016061890026184244, "loss": 2.2159936904907225, "step": 4150 }, { "epoch": 0.5938827224383454, "grad_norm": 1.1634751558303833, "learning_rate": 0.0001605236848369436, "loss": 2.097846031188965, "step": 4160 }, { "epoch": 0.5953103251365145, "grad_norm": 1.3217452764511108, "learning_rate": 0.00016042846941204477, "loss": 2.1109798431396483, "step": 4170 }, { "epoch": 0.5967379278346836, "grad_norm": 1.0953373908996582, "learning_rate": 0.0001603332539871459, "loss": 2.0403207778930663, "step": 4180 }, { "epoch": 0.5981655305328527, "grad_norm": 1.2647509574890137, "learning_rate": 0.00016023803856224708, "loss": 2.1489234924316407, "step": 4190 }, { "epoch": 0.5995931332310218, "grad_norm": 1.2663153409957886, "learning_rate": 0.00016014282313734824, "loss": 2.1639209747314454, "step": 4200 }, { "epoch": 0.6010207359291909, "grad_norm": 1.027962327003479, "learning_rate": 0.00016004760771244942, "loss": 2.0777713775634767, "step": 4210 }, { "epoch": 0.60244833862736, "grad_norm": 1.547472357749939, "learning_rate": 0.00015995239228755058, "loss": 2.114849090576172, "step": 4220 }, { "epoch": 0.6038759413255291, "grad_norm": 1.0689531564712524, "learning_rate": 0.00015985717686265176, "loss": 2.101460838317871, "step": 4230 }, { "epoch": 0.6053035440236982, "grad_norm": 1.1795955896377563, "learning_rate": 0.00015976196143775291, "loss": 2.0170339584350585, "step": 4240 }, { "epoch": 0.6067311467218673, "grad_norm": 1.0898903608322144, "learning_rate": 0.0001596667460128541, "loss": 2.1026046752929686, "step": 4250 }, { "epoch": 0.6081587494200364, "grad_norm": 1.4372042417526245, "learning_rate": 0.00015957153058795525, "loss": 2.0490205764770506, "step": 4260 }, { "epoch": 0.6095863521182056, "grad_norm": 1.186373233795166, "learning_rate": 0.00015947631516305644, "loss": 1.9973588943481446, "step": 4270 }, { "epoch": 0.6110139548163745, "grad_norm": 1.1143156290054321, "learning_rate": 0.0001593810997381576, "loss": 2.049622917175293, "step": 4280 }, { "epoch": 0.6124415575145437, "grad_norm": 1.5476491451263428, "learning_rate": 0.00015928588431325875, "loss": 2.0849376678466798, "step": 4290 }, { "epoch": 0.6138691602127128, "grad_norm": 0.9903887510299683, "learning_rate": 0.00015919066888835993, "loss": 1.98438720703125, "step": 4300 }, { "epoch": 0.6152967629108819, "grad_norm": 1.1740891933441162, "learning_rate": 0.00015909545346346108, "loss": 2.2359495162963867, "step": 4310 }, { "epoch": 0.616724365609051, "grad_norm": 1.0749351978302002, "learning_rate": 0.00015900023803856227, "loss": 2.0514636993408204, "step": 4320 }, { "epoch": 0.6181519683072201, "grad_norm": 1.2126410007476807, "learning_rate": 0.00015890502261366342, "loss": 2.089235305786133, "step": 4330 }, { "epoch": 0.6195795710053892, "grad_norm": 1.1484490633010864, "learning_rate": 0.0001588098071887646, "loss": 2.0745500564575194, "step": 4340 }, { "epoch": 0.6210071737035583, "grad_norm": 0.9319640398025513, "learning_rate": 0.00015871459176386576, "loss": 1.9926181793212892, "step": 4350 }, { "epoch": 0.6224347764017274, "grad_norm": 1.2679811716079712, "learning_rate": 0.00015861937633896694, "loss": 2.129096031188965, "step": 4360 }, { "epoch": 0.6238623790998965, "grad_norm": 1.1401822566986084, "learning_rate": 0.00015852416091406807, "loss": 1.957147979736328, "step": 4370 }, { "epoch": 0.6252899817980656, "grad_norm": 1.0949463844299316, "learning_rate": 0.00015842894548916925, "loss": 2.1449153900146483, "step": 4380 }, { "epoch": 0.6267175844962347, "grad_norm": 1.14918053150177, "learning_rate": 0.0001583337300642704, "loss": 2.2274448394775392, "step": 4390 }, { "epoch": 0.6281451871944038, "grad_norm": 1.1906867027282715, "learning_rate": 0.0001582385146393716, "loss": 2.051712989807129, "step": 4400 }, { "epoch": 0.6295727898925729, "grad_norm": 1.0407558679580688, "learning_rate": 0.00015814329921447275, "loss": 2.0556570053100587, "step": 4410 }, { "epoch": 0.631000392590742, "grad_norm": 1.0472633838653564, "learning_rate": 0.00015804808378957393, "loss": 2.0726411819458006, "step": 4420 }, { "epoch": 0.6324279952889111, "grad_norm": 1.0542018413543701, "learning_rate": 0.00015795286836467508, "loss": 1.9429601669311523, "step": 4430 }, { "epoch": 0.6338555979870802, "grad_norm": 1.056063175201416, "learning_rate": 0.00015785765293977627, "loss": 2.0575481414794923, "step": 4440 }, { "epoch": 0.6352832006852493, "grad_norm": 1.345315933227539, "learning_rate": 0.00015776243751487742, "loss": 2.0701883316040037, "step": 4450 }, { "epoch": 0.6367108033834183, "grad_norm": 1.1465160846710205, "learning_rate": 0.0001576672220899786, "loss": 2.141465950012207, "step": 4460 }, { "epoch": 0.6381384060815875, "grad_norm": 1.2380520105361938, "learning_rate": 0.00015757200666507973, "loss": 2.1055450439453125, "step": 4470 }, { "epoch": 0.6395660087797566, "grad_norm": 1.1188899278640747, "learning_rate": 0.00015747679124018091, "loss": 2.1809152603149413, "step": 4480 }, { "epoch": 0.6409936114779257, "grad_norm": 1.0567125082015991, "learning_rate": 0.00015738157581528207, "loss": 2.1224170684814454, "step": 4490 }, { "epoch": 0.6424212141760948, "grad_norm": 1.1337240934371948, "learning_rate": 0.00015728636039038325, "loss": 2.0783939361572266, "step": 4500 }, { "epoch": 0.6438488168742639, "grad_norm": 1.1785552501678467, "learning_rate": 0.0001571911449654844, "loss": 2.079520606994629, "step": 4510 }, { "epoch": 0.645276419572433, "grad_norm": 1.064718246459961, "learning_rate": 0.0001570959295405856, "loss": 2.0744672775268556, "step": 4520 }, { "epoch": 0.6467040222706021, "grad_norm": 1.0810433626174927, "learning_rate": 0.00015700071411568674, "loss": 2.1164180755615236, "step": 4530 }, { "epoch": 0.6481316249687712, "grad_norm": 1.4308578968048096, "learning_rate": 0.00015690549869078793, "loss": 2.11962833404541, "step": 4540 }, { "epoch": 0.6495592276669403, "grad_norm": 1.0707223415374756, "learning_rate": 0.00015681028326588908, "loss": 2.017913055419922, "step": 4550 }, { "epoch": 0.6509868303651094, "grad_norm": 1.1607109308242798, "learning_rate": 0.00015671506784099024, "loss": 2.080700492858887, "step": 4560 }, { "epoch": 0.6524144330632785, "grad_norm": 1.0517059564590454, "learning_rate": 0.0001566198524160914, "loss": 2.1347278594970702, "step": 4570 }, { "epoch": 0.6538420357614476, "grad_norm": 1.0226854085922241, "learning_rate": 0.00015652463699119258, "loss": 1.9415822982788087, "step": 4580 }, { "epoch": 0.6552696384596167, "grad_norm": 1.1085102558135986, "learning_rate": 0.00015642942156629373, "loss": 2.0874921798706056, "step": 4590 }, { "epoch": 0.6566972411577858, "grad_norm": 1.203783392906189, "learning_rate": 0.0001563342061413949, "loss": 2.018591117858887, "step": 4600 }, { "epoch": 0.6581248438559549, "grad_norm": 1.30747389793396, "learning_rate": 0.00015623899071649607, "loss": 2.124520111083984, "step": 4610 }, { "epoch": 0.659552446554124, "grad_norm": 1.1289852857589722, "learning_rate": 0.00015614377529159725, "loss": 2.1010200500488283, "step": 4620 }, { "epoch": 0.660980049252293, "grad_norm": 1.1173664331436157, "learning_rate": 0.0001560485598666984, "loss": 2.095734786987305, "step": 4630 }, { "epoch": 0.6624076519504621, "grad_norm": 1.3770655393600464, "learning_rate": 0.0001559533444417996, "loss": 2.109881782531738, "step": 4640 }, { "epoch": 0.6638352546486312, "grad_norm": 1.0825073719024658, "learning_rate": 0.00015585812901690074, "loss": 1.9486059188842773, "step": 4650 }, { "epoch": 0.6652628573468004, "grad_norm": 1.228014588356018, "learning_rate": 0.0001557629135920019, "loss": 2.1181900024414064, "step": 4660 }, { "epoch": 0.6666904600449695, "grad_norm": 1.2361053228378296, "learning_rate": 0.00015566769816710308, "loss": 2.11850528717041, "step": 4670 }, { "epoch": 0.6681180627431386, "grad_norm": 1.3095343112945557, "learning_rate": 0.00015557248274220424, "loss": 2.2466053009033202, "step": 4680 }, { "epoch": 0.6695456654413077, "grad_norm": 1.4119725227355957, "learning_rate": 0.00015547726731730542, "loss": 2.139160919189453, "step": 4690 }, { "epoch": 0.6709732681394768, "grad_norm": 1.160866141319275, "learning_rate": 0.00015538205189240657, "loss": 2.178610610961914, "step": 4700 }, { "epoch": 0.6724008708376459, "grad_norm": 0.9942652583122253, "learning_rate": 0.00015528683646750776, "loss": 2.1062307357788086, "step": 4710 }, { "epoch": 0.673828473535815, "grad_norm": 1.1465297937393188, "learning_rate": 0.0001551916210426089, "loss": 1.9471675872802734, "step": 4720 }, { "epoch": 0.6752560762339841, "grad_norm": 1.157509684562683, "learning_rate": 0.0001550964056177101, "loss": 2.2231735229492187, "step": 4730 }, { "epoch": 0.6766836789321532, "grad_norm": 1.2655620574951172, "learning_rate": 0.00015500119019281125, "loss": 2.194755744934082, "step": 4740 }, { "epoch": 0.6781112816303223, "grad_norm": 1.3042454719543457, "learning_rate": 0.0001549059747679124, "loss": 2.1200008392333984, "step": 4750 }, { "epoch": 0.6795388843284914, "grad_norm": 1.211383581161499, "learning_rate": 0.00015481075934301356, "loss": 2.0350900650024415, "step": 4760 }, { "epoch": 0.6809664870266605, "grad_norm": 1.1105891466140747, "learning_rate": 0.00015471554391811474, "loss": 2.0983116149902346, "step": 4770 }, { "epoch": 0.6823940897248296, "grad_norm": 1.1432119607925415, "learning_rate": 0.0001546203284932159, "loss": 2.0374631881713867, "step": 4780 }, { "epoch": 0.6838216924229987, "grad_norm": 1.2010427713394165, "learning_rate": 0.00015452511306831708, "loss": 2.1282493591308596, "step": 4790 }, { "epoch": 0.6852492951211678, "grad_norm": 1.1762789487838745, "learning_rate": 0.00015442989764341824, "loss": 2.157598304748535, "step": 4800 }, { "epoch": 0.6866768978193368, "grad_norm": 1.0752602815628052, "learning_rate": 0.00015433468221851942, "loss": 1.9569272994995117, "step": 4810 }, { "epoch": 0.6881045005175059, "grad_norm": 1.1443010568618774, "learning_rate": 0.00015423946679362057, "loss": 2.072005844116211, "step": 4820 }, { "epoch": 0.689532103215675, "grad_norm": 1.1991745233535767, "learning_rate": 0.00015414425136872176, "loss": 2.044131851196289, "step": 4830 }, { "epoch": 0.6909597059138441, "grad_norm": 1.03378427028656, "learning_rate": 0.0001540490359438229, "loss": 2.241434097290039, "step": 4840 }, { "epoch": 0.6923873086120133, "grad_norm": 1.3095555305480957, "learning_rate": 0.00015395382051892407, "loss": 2.1197956085205076, "step": 4850 }, { "epoch": 0.6938149113101824, "grad_norm": 1.0304228067398071, "learning_rate": 0.00015385860509402522, "loss": 2.0145242691040037, "step": 4860 }, { "epoch": 0.6952425140083515, "grad_norm": 1.0579947233200073, "learning_rate": 0.0001537633896691264, "loss": 1.9315383911132813, "step": 4870 }, { "epoch": 0.6966701167065206, "grad_norm": 1.389248251914978, "learning_rate": 0.00015366817424422756, "loss": 1.96744384765625, "step": 4880 }, { "epoch": 0.6980977194046897, "grad_norm": 1.0396047830581665, "learning_rate": 0.00015357295881932874, "loss": 2.1762937545776366, "step": 4890 }, { "epoch": 0.6995253221028588, "grad_norm": 0.9477119445800781, "learning_rate": 0.0001534777433944299, "loss": 2.269539451599121, "step": 4900 }, { "epoch": 0.7009529248010279, "grad_norm": 1.3811748027801514, "learning_rate": 0.00015338252796953108, "loss": 2.1592836380004883, "step": 4910 }, { "epoch": 0.702380527499197, "grad_norm": 1.0813015699386597, "learning_rate": 0.00015328731254463224, "loss": 2.2636188507080077, "step": 4920 }, { "epoch": 0.7038081301973661, "grad_norm": 1.2023388147354126, "learning_rate": 0.00015319209711973342, "loss": 2.1457313537597655, "step": 4930 }, { "epoch": 0.7052357328955352, "grad_norm": 1.1694599390029907, "learning_rate": 0.00015309688169483455, "loss": 2.1561866760253907, "step": 4940 }, { "epoch": 0.7066633355937043, "grad_norm": 1.278207778930664, "learning_rate": 0.00015300166626993573, "loss": 2.1632722854614257, "step": 4950 }, { "epoch": 0.7080909382918734, "grad_norm": 1.2129367589950562, "learning_rate": 0.00015290645084503688, "loss": 2.104619598388672, "step": 4960 }, { "epoch": 0.7095185409900425, "grad_norm": 1.12024986743927, "learning_rate": 0.00015281123542013807, "loss": 2.2494619369506834, "step": 4970 }, { "epoch": 0.7109461436882115, "grad_norm": 1.323219656944275, "learning_rate": 0.00015271601999523922, "loss": 2.077766990661621, "step": 4980 }, { "epoch": 0.7123737463863806, "grad_norm": 1.4513942003250122, "learning_rate": 0.0001526208045703404, "loss": 2.0815666198730467, "step": 4990 }, { "epoch": 0.7138013490845497, "grad_norm": 1.022873878479004, "learning_rate": 0.00015252558914544156, "loss": 2.0749887466430663, "step": 5000 }, { "epoch": 0.7152289517827188, "grad_norm": 1.310887098312378, "learning_rate": 0.00015243037372054274, "loss": 2.073664093017578, "step": 5010 }, { "epoch": 0.7166565544808879, "grad_norm": 1.0364643335342407, "learning_rate": 0.0001523351582956439, "loss": 2.0644689559936524, "step": 5020 }, { "epoch": 0.718084157179057, "grad_norm": 1.0534369945526123, "learning_rate": 0.00015223994287074508, "loss": 2.1695106506347654, "step": 5030 }, { "epoch": 0.7195117598772262, "grad_norm": 1.228196382522583, "learning_rate": 0.00015214472744584623, "loss": 2.0640996932983398, "step": 5040 }, { "epoch": 0.7209393625753953, "grad_norm": 1.1617060899734497, "learning_rate": 0.0001520495120209474, "loss": 2.059670257568359, "step": 5050 }, { "epoch": 0.7223669652735644, "grad_norm": 1.2786524295806885, "learning_rate": 0.00015195429659604857, "loss": 2.042633056640625, "step": 5060 }, { "epoch": 0.7237945679717335, "grad_norm": 1.7062017917633057, "learning_rate": 0.00015185908117114973, "loss": 1.9900848388671875, "step": 5070 }, { "epoch": 0.7252221706699026, "grad_norm": 1.2654452323913574, "learning_rate": 0.0001517638657462509, "loss": 2.0543087005615233, "step": 5080 }, { "epoch": 0.7266497733680717, "grad_norm": 0.9251814484596252, "learning_rate": 0.00015166865032135207, "loss": 2.2598575592041015, "step": 5090 }, { "epoch": 0.7280773760662408, "grad_norm": 1.2314844131469727, "learning_rate": 0.00015157343489645325, "loss": 1.962451171875, "step": 5100 }, { "epoch": 0.7295049787644099, "grad_norm": 1.1339524984359741, "learning_rate": 0.0001514782194715544, "loss": 2.1815841674804686, "step": 5110 }, { "epoch": 0.730932581462579, "grad_norm": 1.2662720680236816, "learning_rate": 0.00015138300404665559, "loss": 2.06201171875, "step": 5120 }, { "epoch": 0.7323601841607481, "grad_norm": 1.270620584487915, "learning_rate": 0.00015128778862175671, "loss": 2.1221466064453125, "step": 5130 }, { "epoch": 0.7337877868589172, "grad_norm": 1.7752879858016968, "learning_rate": 0.0001511925731968579, "loss": 2.0692325592041017, "step": 5140 }, { "epoch": 0.7352153895570862, "grad_norm": 1.5993555784225464, "learning_rate": 0.00015109735777195905, "loss": 2.039986228942871, "step": 5150 }, { "epoch": 0.7366429922552553, "grad_norm": 1.1296077966690063, "learning_rate": 0.00015100214234706023, "loss": 2.1308712005615233, "step": 5160 }, { "epoch": 0.7380705949534244, "grad_norm": 1.203850269317627, "learning_rate": 0.0001509069269221614, "loss": 2.0894222259521484, "step": 5170 }, { "epoch": 0.7394981976515935, "grad_norm": 1.248410940170288, "learning_rate": 0.00015081171149726257, "loss": 2.123434829711914, "step": 5180 }, { "epoch": 0.7409258003497626, "grad_norm": 1.0876100063323975, "learning_rate": 0.00015071649607236373, "loss": 2.031419372558594, "step": 5190 }, { "epoch": 0.7423534030479317, "grad_norm": 1.0214993953704834, "learning_rate": 0.0001506212806474649, "loss": 2.1958309173583985, "step": 5200 }, { "epoch": 0.7437810057461008, "grad_norm": 0.9046857953071594, "learning_rate": 0.00015052606522256607, "loss": 1.9689685821533203, "step": 5210 }, { "epoch": 0.74520860844427, "grad_norm": 1.102981686592102, "learning_rate": 0.00015043084979766725, "loss": 2.082001495361328, "step": 5220 }, { "epoch": 0.746636211142439, "grad_norm": 1.3899544477462769, "learning_rate": 0.00015033563437276838, "loss": 1.9985469818115233, "step": 5230 }, { "epoch": 0.7480638138406082, "grad_norm": 1.6087454557418823, "learning_rate": 0.00015024041894786956, "loss": 2.144418144226074, "step": 5240 }, { "epoch": 0.7494914165387773, "grad_norm": 1.1745785474777222, "learning_rate": 0.0001501452035229707, "loss": 2.075351333618164, "step": 5250 }, { "epoch": 0.7509190192369464, "grad_norm": 1.252122402191162, "learning_rate": 0.0001500499880980719, "loss": 2.046000099182129, "step": 5260 }, { "epoch": 0.7523466219351155, "grad_norm": 1.2469866275787354, "learning_rate": 0.00014995477267317305, "loss": 2.0471614837646483, "step": 5270 }, { "epoch": 0.7537742246332846, "grad_norm": 1.2567193508148193, "learning_rate": 0.00014985955724827423, "loss": 2.1027271270751955, "step": 5280 }, { "epoch": 0.7552018273314537, "grad_norm": 1.2912555932998657, "learning_rate": 0.0001497643418233754, "loss": 2.1145801544189453, "step": 5290 }, { "epoch": 0.7566294300296228, "grad_norm": 1.0746982097625732, "learning_rate": 0.00014966912639847657, "loss": 1.9975954055786134, "step": 5300 }, { "epoch": 0.7580570327277919, "grad_norm": 0.9123805165290833, "learning_rate": 0.00014957391097357773, "loss": 2.0803789138793944, "step": 5310 }, { "epoch": 0.759484635425961, "grad_norm": 1.1584913730621338, "learning_rate": 0.0001494786955486789, "loss": 2.1535112380981447, "step": 5320 }, { "epoch": 0.76091223812413, "grad_norm": 1.2221531867980957, "learning_rate": 0.00014938348012378004, "loss": 2.0887495040893556, "step": 5330 }, { "epoch": 0.7623398408222991, "grad_norm": 1.087020993232727, "learning_rate": 0.00014928826469888122, "loss": 1.9882490158081054, "step": 5340 }, { "epoch": 0.7637674435204682, "grad_norm": 1.687694787979126, "learning_rate": 0.00014919304927398238, "loss": 2.2212324142456055, "step": 5350 }, { "epoch": 0.7651950462186373, "grad_norm": 1.0636277198791504, "learning_rate": 0.00014909783384908356, "loss": 2.002360534667969, "step": 5360 }, { "epoch": 0.7666226489168064, "grad_norm": 1.2687207460403442, "learning_rate": 0.0001490026184241847, "loss": 2.2147212982177735, "step": 5370 }, { "epoch": 0.7680502516149755, "grad_norm": 1.236961007118225, "learning_rate": 0.0001489074029992859, "loss": 2.040176773071289, "step": 5380 }, { "epoch": 0.7694778543131446, "grad_norm": 0.9304345846176147, "learning_rate": 0.00014881218757438705, "loss": 2.118213081359863, "step": 5390 }, { "epoch": 0.7709054570113137, "grad_norm": 1.2043101787567139, "learning_rate": 0.00014871697214948823, "loss": 2.0982040405273437, "step": 5400 }, { "epoch": 0.7723330597094828, "grad_norm": 1.1328470706939697, "learning_rate": 0.0001486217567245894, "loss": 2.0471426010131837, "step": 5410 }, { "epoch": 0.773760662407652, "grad_norm": 1.830897569656372, "learning_rate": 0.00014852654129969054, "loss": 1.943600845336914, "step": 5420 }, { "epoch": 0.7751882651058211, "grad_norm": 1.2577552795410156, "learning_rate": 0.00014843132587479173, "loss": 2.044905662536621, "step": 5430 }, { "epoch": 0.7766158678039902, "grad_norm": 1.00262451171875, "learning_rate": 0.00014833611044989288, "loss": 2.0161760330200194, "step": 5440 }, { "epoch": 0.7780434705021593, "grad_norm": 1.2471446990966797, "learning_rate": 0.00014824089502499406, "loss": 2.030394744873047, "step": 5450 }, { "epoch": 0.7794710732003284, "grad_norm": 1.3308446407318115, "learning_rate": 0.00014814567960009522, "loss": 2.2480823516845705, "step": 5460 }, { "epoch": 0.7808986758984975, "grad_norm": 1.0902934074401855, "learning_rate": 0.0001480504641751964, "loss": 2.159767913818359, "step": 5470 }, { "epoch": 0.7823262785966666, "grad_norm": 1.1527326107025146, "learning_rate": 0.00014795524875029756, "loss": 2.1065540313720703, "step": 5480 }, { "epoch": 0.7837538812948357, "grad_norm": 1.158408761024475, "learning_rate": 0.00014786003332539874, "loss": 1.9589733123779296, "step": 5490 }, { "epoch": 0.7851814839930047, "grad_norm": 1.1625585556030273, "learning_rate": 0.0001477648179004999, "loss": 2.0707992553710937, "step": 5500 }, { "epoch": 0.7866090866911738, "grad_norm": 0.9742340445518494, "learning_rate": 0.00014766960247560108, "loss": 2.192020606994629, "step": 5510 }, { "epoch": 0.7880366893893429, "grad_norm": 2.17490291595459, "learning_rate": 0.0001475743870507022, "loss": 2.039841651916504, "step": 5520 }, { "epoch": 0.789464292087512, "grad_norm": 1.186703085899353, "learning_rate": 0.0001474791716258034, "loss": 2.0397449493408204, "step": 5530 }, { "epoch": 0.7908918947856811, "grad_norm": 1.0868713855743408, "learning_rate": 0.00014738395620090454, "loss": 2.112815093994141, "step": 5540 }, { "epoch": 0.7923194974838502, "grad_norm": 1.2248460054397583, "learning_rate": 0.00014728874077600573, "loss": 2.0478261947631835, "step": 5550 }, { "epoch": 0.7937471001820193, "grad_norm": 1.0196927785873413, "learning_rate": 0.00014719352535110688, "loss": 2.0375383377075194, "step": 5560 }, { "epoch": 0.7951747028801884, "grad_norm": 1.1145230531692505, "learning_rate": 0.00014709830992620806, "loss": 2.038543128967285, "step": 5570 }, { "epoch": 0.7966023055783575, "grad_norm": 0.8973228931427002, "learning_rate": 0.00014700309450130922, "loss": 1.9726655960083008, "step": 5580 }, { "epoch": 0.7980299082765266, "grad_norm": 1.2083377838134766, "learning_rate": 0.0001469078790764104, "loss": 2.0419130325317383, "step": 5590 }, { "epoch": 0.7994575109746958, "grad_norm": 1.1352425813674927, "learning_rate": 0.00014681266365151156, "loss": 2.043312835693359, "step": 5600 }, { "epoch": 0.8008851136728649, "grad_norm": 1.1394609212875366, "learning_rate": 0.0001467174482266127, "loss": 1.9503747940063476, "step": 5610 }, { "epoch": 0.802312716371034, "grad_norm": 1.081141471862793, "learning_rate": 0.00014662223280171387, "loss": 1.9604333877563476, "step": 5620 }, { "epoch": 0.8037403190692031, "grad_norm": 1.2257745265960693, "learning_rate": 0.00014652701737681505, "loss": 2.154981803894043, "step": 5630 }, { "epoch": 0.8051679217673722, "grad_norm": 1.315804362297058, "learning_rate": 0.0001464318019519162, "loss": 1.9176166534423829, "step": 5640 }, { "epoch": 0.8065955244655413, "grad_norm": 1.121869683265686, "learning_rate": 0.0001463365865270174, "loss": 2.07440185546875, "step": 5650 }, { "epoch": 0.8080231271637104, "grad_norm": 0.9741482138633728, "learning_rate": 0.00014624137110211854, "loss": 2.1492473602294924, "step": 5660 }, { "epoch": 0.8094507298618795, "grad_norm": 0.972418487071991, "learning_rate": 0.00014614615567721972, "loss": 2.157696533203125, "step": 5670 }, { "epoch": 0.8108783325600485, "grad_norm": 1.1014772653579712, "learning_rate": 0.00014605094025232088, "loss": 2.292978858947754, "step": 5680 }, { "epoch": 0.8123059352582176, "grad_norm": 1.1848760843276978, "learning_rate": 0.00014595572482742206, "loss": 2.0331499099731447, "step": 5690 }, { "epoch": 0.8137335379563867, "grad_norm": 1.2902730703353882, "learning_rate": 0.00014586050940252322, "loss": 2.0033824920654295, "step": 5700 }, { "epoch": 0.8151611406545558, "grad_norm": 1.1614636182785034, "learning_rate": 0.00014576529397762437, "loss": 2.010498046875, "step": 5710 }, { "epoch": 0.8165887433527249, "grad_norm": 1.0056531429290771, "learning_rate": 0.00014567007855272553, "loss": 2.073269081115723, "step": 5720 }, { "epoch": 0.818016346050894, "grad_norm": 1.1945202350616455, "learning_rate": 0.0001455748631278267, "loss": 2.1130182266235353, "step": 5730 }, { "epoch": 0.8194439487490631, "grad_norm": 1.0224995613098145, "learning_rate": 0.00014547964770292787, "loss": 2.259115791320801, "step": 5740 }, { "epoch": 0.8208715514472322, "grad_norm": 1.1886273622512817, "learning_rate": 0.00014538443227802905, "loss": 2.1042755126953123, "step": 5750 }, { "epoch": 0.8222991541454013, "grad_norm": 1.2375845909118652, "learning_rate": 0.0001452892168531302, "loss": 2.097722816467285, "step": 5760 }, { "epoch": 0.8237267568435704, "grad_norm": 1.2205432653427124, "learning_rate": 0.00014519400142823139, "loss": 2.189805793762207, "step": 5770 }, { "epoch": 0.8251543595417395, "grad_norm": 1.1793696880340576, "learning_rate": 0.00014509878600333254, "loss": 2.1764703750610352, "step": 5780 }, { "epoch": 0.8265819622399087, "grad_norm": 1.152541995048523, "learning_rate": 0.00014500357057843372, "loss": 2.0075801849365233, "step": 5790 }, { "epoch": 0.8280095649380778, "grad_norm": 1.1772470474243164, "learning_rate": 0.00014490835515353488, "loss": 2.144789695739746, "step": 5800 }, { "epoch": 0.8294371676362469, "grad_norm": 1.1392500400543213, "learning_rate": 0.00014481313972863603, "loss": 2.257477569580078, "step": 5810 }, { "epoch": 0.830864770334416, "grad_norm": 1.5486935377120972, "learning_rate": 0.00014471792430373722, "loss": 2.176411247253418, "step": 5820 }, { "epoch": 0.8322923730325851, "grad_norm": 1.5925824642181396, "learning_rate": 0.00014462270887883837, "loss": 1.9543567657470704, "step": 5830 }, { "epoch": 0.8337199757307542, "grad_norm": 1.7298519611358643, "learning_rate": 0.00014452749345393955, "loss": 2.0520088195800783, "step": 5840 }, { "epoch": 0.8351475784289232, "grad_norm": 1.2536671161651611, "learning_rate": 0.0001444322780290407, "loss": 2.1787694931030273, "step": 5850 }, { "epoch": 0.8365751811270923, "grad_norm": 1.1694890260696411, "learning_rate": 0.0001443370626041419, "loss": 2.1233547210693358, "step": 5860 }, { "epoch": 0.8380027838252614, "grad_norm": 1.9879645109176636, "learning_rate": 0.00014424184717924305, "loss": 2.236946868896484, "step": 5870 }, { "epoch": 0.8394303865234305, "grad_norm": 1.0208160877227783, "learning_rate": 0.00014414663175434423, "loss": 1.970679473876953, "step": 5880 }, { "epoch": 0.8408579892215996, "grad_norm": 1.2248173952102661, "learning_rate": 0.00014405141632944539, "loss": 2.075257110595703, "step": 5890 }, { "epoch": 0.8422855919197687, "grad_norm": 1.2281972169876099, "learning_rate": 0.00014395620090454654, "loss": 2.0795242309570314, "step": 5900 }, { "epoch": 0.8437131946179378, "grad_norm": 1.3376227617263794, "learning_rate": 0.0001438609854796477, "loss": 2.1916088104248046, "step": 5910 }, { "epoch": 0.8451407973161069, "grad_norm": 1.4484951496124268, "learning_rate": 0.00014376577005474888, "loss": 2.0632484436035154, "step": 5920 }, { "epoch": 0.846568400014276, "grad_norm": 1.1307789087295532, "learning_rate": 0.00014367055462985003, "loss": 2.153516387939453, "step": 5930 }, { "epoch": 0.8479960027124451, "grad_norm": 1.0334279537200928, "learning_rate": 0.00014357533920495122, "loss": 1.965131950378418, "step": 5940 }, { "epoch": 0.8494236054106142, "grad_norm": 1.1069447994232178, "learning_rate": 0.00014348012378005237, "loss": 1.9902894973754883, "step": 5950 }, { "epoch": 0.8508512081087833, "grad_norm": 1.0762388706207275, "learning_rate": 0.00014338490835515355, "loss": 1.9570812225341796, "step": 5960 }, { "epoch": 0.8522788108069524, "grad_norm": 1.2601449489593506, "learning_rate": 0.0001432896929302547, "loss": 2.139608383178711, "step": 5970 }, { "epoch": 0.8537064135051216, "grad_norm": 1.036189079284668, "learning_rate": 0.0001431944775053559, "loss": 2.1933883666992187, "step": 5980 }, { "epoch": 0.8551340162032907, "grad_norm": 1.1096618175506592, "learning_rate": 0.00014309926208045702, "loss": 2.1286998748779298, "step": 5990 }, { "epoch": 0.8565616189014598, "grad_norm": 1.1303385496139526, "learning_rate": 0.0001430040466555582, "loss": 2.0584604263305666, "step": 6000 }, { "epoch": 0.8579892215996289, "grad_norm": 1.1133581399917603, "learning_rate": 0.00014290883123065936, "loss": 2.1713630676269533, "step": 6010 }, { "epoch": 0.859416824297798, "grad_norm": 1.22328519821167, "learning_rate": 0.00014281361580576054, "loss": 2.1653820037841798, "step": 6020 }, { "epoch": 0.860844426995967, "grad_norm": 1.1215869188308716, "learning_rate": 0.0001427184003808617, "loss": 2.1111936569213867, "step": 6030 }, { "epoch": 0.8622720296941361, "grad_norm": 1.3741992712020874, "learning_rate": 0.00014262318495596288, "loss": 2.0740543365478517, "step": 6040 }, { "epoch": 0.8636996323923052, "grad_norm": 1.17236328125, "learning_rate": 0.00014252796953106403, "loss": 2.000840759277344, "step": 6050 }, { "epoch": 0.8651272350904743, "grad_norm": 1.0364574193954468, "learning_rate": 0.00014243275410616522, "loss": 1.9616924285888673, "step": 6060 }, { "epoch": 0.8665548377886434, "grad_norm": 1.3406749963760376, "learning_rate": 0.00014233753868126637, "loss": 2.067513656616211, "step": 6070 }, { "epoch": 0.8679824404868125, "grad_norm": 0.9893628358840942, "learning_rate": 0.00014224232325636755, "loss": 2.294078826904297, "step": 6080 }, { "epoch": 0.8694100431849816, "grad_norm": 1.0370484590530396, "learning_rate": 0.0001421471078314687, "loss": 2.120780181884766, "step": 6090 }, { "epoch": 0.8708376458831507, "grad_norm": 1.1814578771591187, "learning_rate": 0.00014205189240656986, "loss": 2.0330245971679686, "step": 6100 }, { "epoch": 0.8722652485813198, "grad_norm": 0.998306393623352, "learning_rate": 0.00014195667698167102, "loss": 2.2199316024780273, "step": 6110 }, { "epoch": 0.8736928512794889, "grad_norm": 1.1868630647659302, "learning_rate": 0.0001418614615567722, "loss": 2.1109052658081056, "step": 6120 }, { "epoch": 0.875120453977658, "grad_norm": 1.0858951807022095, "learning_rate": 0.00014176624613187336, "loss": 2.0018476486206054, "step": 6130 }, { "epoch": 0.8765480566758271, "grad_norm": 1.0980168581008911, "learning_rate": 0.00014167103070697454, "loss": 2.0752668380737305, "step": 6140 }, { "epoch": 0.8779756593739962, "grad_norm": 1.7877191305160522, "learning_rate": 0.0001415758152820757, "loss": 2.0490203857421876, "step": 6150 }, { "epoch": 0.8794032620721653, "grad_norm": 1.540902018547058, "learning_rate": 0.00014148059985717688, "loss": 2.14637451171875, "step": 6160 }, { "epoch": 0.8808308647703345, "grad_norm": 1.324889898300171, "learning_rate": 0.00014138538443227803, "loss": 2.047268867492676, "step": 6170 }, { "epoch": 0.8822584674685036, "grad_norm": 1.2005466222763062, "learning_rate": 0.0001412901690073792, "loss": 2.0844221115112305, "step": 6180 }, { "epoch": 0.8836860701666727, "grad_norm": 1.3227152824401855, "learning_rate": 0.00014119495358248037, "loss": 2.1250864028930665, "step": 6190 }, { "epoch": 0.8851136728648417, "grad_norm": 1.140487790107727, "learning_rate": 0.00014109973815758153, "loss": 2.1609779357910157, "step": 6200 }, { "epoch": 0.8865412755630108, "grad_norm": 1.1737276315689087, "learning_rate": 0.0001410045227326827, "loss": 2.224088668823242, "step": 6210 }, { "epoch": 0.8879688782611799, "grad_norm": 0.9986388683319092, "learning_rate": 0.00014090930730778386, "loss": 2.185664749145508, "step": 6220 }, { "epoch": 0.889396480959349, "grad_norm": 1.061862826347351, "learning_rate": 0.00014081409188288505, "loss": 2.120680236816406, "step": 6230 }, { "epoch": 0.8908240836575181, "grad_norm": 1.0981886386871338, "learning_rate": 0.0001407188764579862, "loss": 2.058972930908203, "step": 6240 }, { "epoch": 0.8922516863556872, "grad_norm": 1.15584135055542, "learning_rate": 0.00014062366103308738, "loss": 2.019556999206543, "step": 6250 }, { "epoch": 0.8936792890538563, "grad_norm": 1.0575971603393555, "learning_rate": 0.00014052844560818854, "loss": 2.038960838317871, "step": 6260 }, { "epoch": 0.8951068917520254, "grad_norm": 1.0918058156967163, "learning_rate": 0.00014043323018328972, "loss": 2.095849609375, "step": 6270 }, { "epoch": 0.8965344944501945, "grad_norm": 1.210402250289917, "learning_rate": 0.00014033801475839085, "loss": 2.234294891357422, "step": 6280 }, { "epoch": 0.8979620971483636, "grad_norm": 1.0035918951034546, "learning_rate": 0.00014024279933349203, "loss": 2.2051664352416993, "step": 6290 }, { "epoch": 0.8993896998465327, "grad_norm": 1.2978595495224, "learning_rate": 0.0001401475839085932, "loss": 1.9633651733398438, "step": 6300 }, { "epoch": 0.9008173025447018, "grad_norm": 1.0387922525405884, "learning_rate": 0.00014005236848369437, "loss": 2.1111146926879885, "step": 6310 }, { "epoch": 0.9022449052428709, "grad_norm": 1.0476093292236328, "learning_rate": 0.00013995715305879553, "loss": 1.935037612915039, "step": 6320 }, { "epoch": 0.90367250794104, "grad_norm": 1.0816744565963745, "learning_rate": 0.0001398619376338967, "loss": 2.0235315322875977, "step": 6330 }, { "epoch": 0.9051001106392091, "grad_norm": 1.2332687377929688, "learning_rate": 0.00013976672220899786, "loss": 2.0125961303710938, "step": 6340 }, { "epoch": 0.9065277133373782, "grad_norm": 1.0062885284423828, "learning_rate": 0.00013967150678409905, "loss": 2.105752372741699, "step": 6350 }, { "epoch": 0.9079553160355474, "grad_norm": 1.0509051084518433, "learning_rate": 0.0001395762913592002, "loss": 2.08899040222168, "step": 6360 }, { "epoch": 0.9093829187337165, "grad_norm": 1.037889003753662, "learning_rate": 0.00013948107593430136, "loss": 2.074511909484863, "step": 6370 }, { "epoch": 0.9108105214318855, "grad_norm": 1.1551920175552368, "learning_rate": 0.0001393858605094025, "loss": 2.3008186340332033, "step": 6380 }, { "epoch": 0.9122381241300546, "grad_norm": 1.1904058456420898, "learning_rate": 0.0001392906450845037, "loss": 2.0888757705688477, "step": 6390 }, { "epoch": 0.9136657268282237, "grad_norm": 1.1223903894424438, "learning_rate": 0.00013919542965960485, "loss": 2.1007930755615236, "step": 6400 }, { "epoch": 0.9150933295263928, "grad_norm": 1.006466031074524, "learning_rate": 0.00013910021423470603, "loss": 2.0456350326538084, "step": 6410 }, { "epoch": 0.9165209322245619, "grad_norm": 1.3824387788772583, "learning_rate": 0.0001390049988098072, "loss": 2.1661962509155273, "step": 6420 }, { "epoch": 0.917948534922731, "grad_norm": 1.0541598796844482, "learning_rate": 0.00013890978338490837, "loss": 2.0152801513671874, "step": 6430 }, { "epoch": 0.9193761376209001, "grad_norm": 1.0460031032562256, "learning_rate": 0.00013881456796000952, "loss": 2.041923713684082, "step": 6440 }, { "epoch": 0.9208037403190692, "grad_norm": 0.9965769052505493, "learning_rate": 0.0001387193525351107, "loss": 2.133123588562012, "step": 6450 }, { "epoch": 0.9222313430172383, "grad_norm": 1.5547727346420288, "learning_rate": 0.00013862413711021186, "loss": 2.220262145996094, "step": 6460 }, { "epoch": 0.9236589457154074, "grad_norm": 1.0881998538970947, "learning_rate": 0.00013852892168531302, "loss": 1.9972219467163086, "step": 6470 }, { "epoch": 0.9250865484135765, "grad_norm": 1.191167950630188, "learning_rate": 0.0001384337062604142, "loss": 1.9862421035766602, "step": 6480 }, { "epoch": 0.9265141511117456, "grad_norm": 1.8094329833984375, "learning_rate": 0.00013833849083551536, "loss": 2.2336565017700196, "step": 6490 }, { "epoch": 0.9279417538099147, "grad_norm": 1.2954028844833374, "learning_rate": 0.0001382432754106165, "loss": 2.1884744644165037, "step": 6500 }, { "epoch": 0.9293693565080838, "grad_norm": 1.190041422843933, "learning_rate": 0.0001381480599857177, "loss": 2.1505512237548827, "step": 6510 }, { "epoch": 0.9307969592062529, "grad_norm": 0.9555850028991699, "learning_rate": 0.00013805284456081885, "loss": 2.0867473602294924, "step": 6520 }, { "epoch": 0.932224561904422, "grad_norm": 1.4562186002731323, "learning_rate": 0.00013795762913592003, "loss": 2.0376022338867186, "step": 6530 }, { "epoch": 0.9336521646025912, "grad_norm": 1.098338007926941, "learning_rate": 0.00013786241371102119, "loss": 2.0735618591308596, "step": 6540 }, { "epoch": 0.9350797673007601, "grad_norm": 1.2102991342544556, "learning_rate": 0.00013776719828612237, "loss": 2.280936431884766, "step": 6550 }, { "epoch": 0.9365073699989293, "grad_norm": 0.9873177409172058, "learning_rate": 0.00013767198286122352, "loss": 1.9910409927368165, "step": 6560 }, { "epoch": 0.9379349726970984, "grad_norm": 1.1644127368927002, "learning_rate": 0.00013757676743632468, "loss": 2.0336328506469727, "step": 6570 }, { "epoch": 0.9393625753952675, "grad_norm": 1.0895888805389404, "learning_rate": 0.00013748155201142586, "loss": 2.106204795837402, "step": 6580 }, { "epoch": 0.9407901780934366, "grad_norm": 1.601953387260437, "learning_rate": 0.00013738633658652702, "loss": 2.126922035217285, "step": 6590 }, { "epoch": 0.9422177807916057, "grad_norm": 1.2933708429336548, "learning_rate": 0.0001372911211616282, "loss": 2.0938291549682617, "step": 6600 }, { "epoch": 0.9436453834897748, "grad_norm": 1.1286245584487915, "learning_rate": 0.00013719590573672935, "loss": 1.9766693115234375, "step": 6610 }, { "epoch": 0.9450729861879439, "grad_norm": 1.2280887365341187, "learning_rate": 0.00013710069031183054, "loss": 2.06427116394043, "step": 6620 }, { "epoch": 0.946500588886113, "grad_norm": 1.3586266040802002, "learning_rate": 0.0001370054748869317, "loss": 2.0771947860717774, "step": 6630 }, { "epoch": 0.9479281915842821, "grad_norm": 1.1445430517196655, "learning_rate": 0.00013691025946203287, "loss": 2.0146902084350584, "step": 6640 }, { "epoch": 0.9493557942824512, "grad_norm": 1.1393014192581177, "learning_rate": 0.00013681504403713403, "loss": 2.0045536041259764, "step": 6650 }, { "epoch": 0.9507833969806203, "grad_norm": 0.997422456741333, "learning_rate": 0.00013671982861223519, "loss": 2.085128974914551, "step": 6660 }, { "epoch": 0.9522109996787894, "grad_norm": 1.8452311754226685, "learning_rate": 0.00013662461318733634, "loss": 2.0744565963745116, "step": 6670 }, { "epoch": 0.9536386023769585, "grad_norm": 1.085334062576294, "learning_rate": 0.00013652939776243752, "loss": 2.030110740661621, "step": 6680 }, { "epoch": 0.9550662050751276, "grad_norm": 1.0192002058029175, "learning_rate": 0.00013643418233753868, "loss": 2.020688056945801, "step": 6690 }, { "epoch": 0.9564938077732967, "grad_norm": 1.4515304565429688, "learning_rate": 0.00013633896691263986, "loss": 2.2475374221801756, "step": 6700 }, { "epoch": 0.9579214104714658, "grad_norm": 1.1201385259628296, "learning_rate": 0.00013624375148774102, "loss": 2.0835742950439453, "step": 6710 }, { "epoch": 0.959349013169635, "grad_norm": 1.1684490442276, "learning_rate": 0.0001361485360628422, "loss": 2.150490951538086, "step": 6720 }, { "epoch": 0.9607766158678039, "grad_norm": 1.2021454572677612, "learning_rate": 0.00013605332063794335, "loss": 2.180669403076172, "step": 6730 }, { "epoch": 0.962204218565973, "grad_norm": 1.2177183628082275, "learning_rate": 0.00013595810521304454, "loss": 1.9428043365478516, "step": 6740 }, { "epoch": 0.9636318212641422, "grad_norm": 1.0163570642471313, "learning_rate": 0.00013586288978814566, "loss": 2.111976432800293, "step": 6750 }, { "epoch": 0.9650594239623113, "grad_norm": 0.8944364786148071, "learning_rate": 0.00013576767436324685, "loss": 2.034988594055176, "step": 6760 }, { "epoch": 0.9664870266604804, "grad_norm": 1.3543869256973267, "learning_rate": 0.000135672458938348, "loss": 1.8940113067626954, "step": 6770 }, { "epoch": 0.9679146293586495, "grad_norm": 1.6731265783309937, "learning_rate": 0.00013557724351344918, "loss": 2.19631290435791, "step": 6780 }, { "epoch": 0.9693422320568186, "grad_norm": 1.3068981170654297, "learning_rate": 0.00013548202808855034, "loss": 2.1291637420654297, "step": 6790 }, { "epoch": 0.9707698347549877, "grad_norm": 1.2681282758712769, "learning_rate": 0.00013538681266365152, "loss": 1.9592950820922852, "step": 6800 }, { "epoch": 0.9721974374531568, "grad_norm": 1.1241919994354248, "learning_rate": 0.00013529159723875268, "loss": 2.0713878631591798, "step": 6810 }, { "epoch": 0.9736250401513259, "grad_norm": 1.0938230752944946, "learning_rate": 0.00013519638181385386, "loss": 2.089142608642578, "step": 6820 }, { "epoch": 0.975052642849495, "grad_norm": 1.3148494958877563, "learning_rate": 0.00013510116638895502, "loss": 2.179977607727051, "step": 6830 }, { "epoch": 0.9764802455476641, "grad_norm": 1.3184319734573364, "learning_rate": 0.0001350059509640562, "loss": 2.0993885040283202, "step": 6840 }, { "epoch": 0.9779078482458332, "grad_norm": 1.2236640453338623, "learning_rate": 0.00013491073553915735, "loss": 2.054055404663086, "step": 6850 }, { "epoch": 0.9793354509440023, "grad_norm": 1.1087441444396973, "learning_rate": 0.0001348155201142585, "loss": 2.1136968612670897, "step": 6860 }, { "epoch": 0.9807630536421714, "grad_norm": 1.062304973602295, "learning_rate": 0.00013472030468935966, "loss": 1.9438411712646484, "step": 6870 }, { "epoch": 0.9821906563403405, "grad_norm": 1.2111984491348267, "learning_rate": 0.00013462508926446085, "loss": 2.112901496887207, "step": 6880 }, { "epoch": 0.9836182590385096, "grad_norm": 1.0896751880645752, "learning_rate": 0.000134529873839562, "loss": 2.0546585083007813, "step": 6890 }, { "epoch": 0.9850458617366786, "grad_norm": 1.2586815357208252, "learning_rate": 0.00013443465841466318, "loss": 2.048493194580078, "step": 6900 }, { "epoch": 0.9864734644348477, "grad_norm": 1.207970142364502, "learning_rate": 0.00013433944298976434, "loss": 2.0124300003051756, "step": 6910 }, { "epoch": 0.9879010671330168, "grad_norm": 1.0682812929153442, "learning_rate": 0.00013424422756486552, "loss": 2.0715124130249025, "step": 6920 }, { "epoch": 0.989328669831186, "grad_norm": 1.179279088973999, "learning_rate": 0.00013414901213996668, "loss": 1.9233360290527344, "step": 6930 }, { "epoch": 0.990756272529355, "grad_norm": 1.1221063137054443, "learning_rate": 0.00013405379671506783, "loss": 2.1662010192871093, "step": 6940 }, { "epoch": 0.9921838752275242, "grad_norm": 1.068601369857788, "learning_rate": 0.00013395858129016902, "loss": 2.1482254028320313, "step": 6950 }, { "epoch": 0.9936114779256933, "grad_norm": 1.2489750385284424, "learning_rate": 0.00013386336586527017, "loss": 2.0300907135009765, "step": 6960 }, { "epoch": 0.9950390806238624, "grad_norm": 1.0826131105422974, "learning_rate": 0.00013376815044037135, "loss": 2.034779357910156, "step": 6970 }, { "epoch": 0.9964666833220315, "grad_norm": 1.1513251066207886, "learning_rate": 0.0001336729350154725, "loss": 2.017785835266113, "step": 6980 }, { "epoch": 0.9978942860202006, "grad_norm": 1.1646808385849, "learning_rate": 0.0001335777195905737, "loss": 2.0223468780517577, "step": 6990 }, { "epoch": 0.9993218887183697, "grad_norm": 1.4855624437332153, "learning_rate": 0.00013348250416567485, "loss": 2.101408004760742, "step": 7000 }, { "epoch": 1.0007138013490846, "grad_norm": 1.044844150543213, "learning_rate": 0.00013338728874077603, "loss": 2.0136650085449217, "step": 7010 }, { "epoch": 1.0021414040472536, "grad_norm": 1.1037570238113403, "learning_rate": 0.00013329207331587718, "loss": 1.9252758026123047, "step": 7020 }, { "epoch": 1.0035690067454228, "grad_norm": 1.1288423538208008, "learning_rate": 0.00013319685789097837, "loss": 2.105826759338379, "step": 7030 }, { "epoch": 1.0049966094435918, "grad_norm": 1.0805550813674927, "learning_rate": 0.0001331016424660795, "loss": 2.005961227416992, "step": 7040 }, { "epoch": 1.006424212141761, "grad_norm": 1.052844762802124, "learning_rate": 0.00013300642704118068, "loss": 1.9896677017211915, "step": 7050 }, { "epoch": 1.00785181483993, "grad_norm": 1.1398789882659912, "learning_rate": 0.00013291121161628183, "loss": 2.004644775390625, "step": 7060 }, { "epoch": 1.0092794175380992, "grad_norm": 1.1821119785308838, "learning_rate": 0.00013281599619138301, "loss": 2.0728694915771486, "step": 7070 }, { "epoch": 1.0107070202362682, "grad_norm": 1.1592210531234741, "learning_rate": 0.00013272078076648417, "loss": 2.083714485168457, "step": 7080 }, { "epoch": 1.0121346229344375, "grad_norm": 1.1653568744659424, "learning_rate": 0.00013262556534158535, "loss": 2.1155282974243166, "step": 7090 }, { "epoch": 1.0135622256326065, "grad_norm": 1.090782880783081, "learning_rate": 0.0001325303499166865, "loss": 2.1359683990478517, "step": 7100 }, { "epoch": 1.0149898283307754, "grad_norm": 1.1044045686721802, "learning_rate": 0.0001324351344917877, "loss": 2.0308385848999024, "step": 7110 }, { "epoch": 1.0164174310289447, "grad_norm": 1.0916597843170166, "learning_rate": 0.00013233991906688885, "loss": 2.088265609741211, "step": 7120 }, { "epoch": 1.0178450337271137, "grad_norm": 1.4786863327026367, "learning_rate": 0.00013224470364199, "loss": 2.1600215911865233, "step": 7130 }, { "epoch": 1.0192726364252829, "grad_norm": 1.3796428442001343, "learning_rate": 0.00013214948821709116, "loss": 2.1057533264160155, "step": 7140 }, { "epoch": 1.0207002391234519, "grad_norm": 1.277183175086975, "learning_rate": 0.00013205427279219234, "loss": 2.062204933166504, "step": 7150 }, { "epoch": 1.022127841821621, "grad_norm": 1.1751946210861206, "learning_rate": 0.0001319590573672935, "loss": 1.9124109268188476, "step": 7160 }, { "epoch": 1.02355544451979, "grad_norm": 1.0863837003707886, "learning_rate": 0.00013186384194239468, "loss": 2.0474239349365235, "step": 7170 }, { "epoch": 1.0249830472179593, "grad_norm": 1.1980212926864624, "learning_rate": 0.00013176862651749583, "loss": 1.919113540649414, "step": 7180 }, { "epoch": 1.0264106499161283, "grad_norm": 1.1229037046432495, "learning_rate": 0.00013167341109259701, "loss": 2.094904899597168, "step": 7190 }, { "epoch": 1.0278382526142975, "grad_norm": 1.0782687664031982, "learning_rate": 0.00013157819566769817, "loss": 1.9420904159545898, "step": 7200 }, { "epoch": 1.0292658553124665, "grad_norm": 1.274143934249878, "learning_rate": 0.00013148298024279935, "loss": 2.050454330444336, "step": 7210 }, { "epoch": 1.0306934580106357, "grad_norm": 1.091039776802063, "learning_rate": 0.0001313877648179005, "loss": 1.8519201278686523, "step": 7220 }, { "epoch": 1.0321210607088047, "grad_norm": 0.9406845569610596, "learning_rate": 0.00013129254939300166, "loss": 2.07373161315918, "step": 7230 }, { "epoch": 1.033548663406974, "grad_norm": 1.5994365215301514, "learning_rate": 0.00013119733396810284, "loss": 2.1247095108032226, "step": 7240 }, { "epoch": 1.034976266105143, "grad_norm": 1.1544760465621948, "learning_rate": 0.000131102118543204, "loss": 2.0326953887939454, "step": 7250 }, { "epoch": 1.0364038688033121, "grad_norm": 1.035490870475769, "learning_rate": 0.00013100690311830516, "loss": 2.037012481689453, "step": 7260 }, { "epoch": 1.0378314715014811, "grad_norm": 1.0267552137374878, "learning_rate": 0.00013091168769340634, "loss": 2.0926897048950197, "step": 7270 }, { "epoch": 1.0392590741996501, "grad_norm": 0.9776862859725952, "learning_rate": 0.0001308164722685075, "loss": 2.0102745056152345, "step": 7280 }, { "epoch": 1.0406866768978194, "grad_norm": 1.0344829559326172, "learning_rate": 0.00013072125684360868, "loss": 2.1357400894165037, "step": 7290 }, { "epoch": 1.0421142795959883, "grad_norm": 1.2162184715270996, "learning_rate": 0.00013062604141870983, "loss": 1.9338621139526366, "step": 7300 }, { "epoch": 1.0435418822941576, "grad_norm": 1.2007536888122559, "learning_rate": 0.000130530825993811, "loss": 2.0032537460327147, "step": 7310 }, { "epoch": 1.0449694849923266, "grad_norm": 1.1872856616973877, "learning_rate": 0.00013043561056891217, "loss": 2.1037433624267576, "step": 7320 }, { "epoch": 1.0463970876904958, "grad_norm": 1.3197476863861084, "learning_rate": 0.00013034039514401332, "loss": 2.075230026245117, "step": 7330 }, { "epoch": 1.0478246903886648, "grad_norm": 0.9950087666511536, "learning_rate": 0.0001302451797191145, "loss": 2.0331871032714846, "step": 7340 }, { "epoch": 1.049252293086834, "grad_norm": 1.107815146446228, "learning_rate": 0.00013014996429421566, "loss": 1.9863561630249023, "step": 7350 }, { "epoch": 1.050679895785003, "grad_norm": 1.169866919517517, "learning_rate": 0.00013005474886931684, "loss": 2.0876691818237303, "step": 7360 }, { "epoch": 1.0521074984831722, "grad_norm": 1.1068190336227417, "learning_rate": 0.000129959533444418, "loss": 1.974905014038086, "step": 7370 }, { "epoch": 1.0535351011813412, "grad_norm": 1.1030644178390503, "learning_rate": 0.00012986431801951918, "loss": 2.0674291610717774, "step": 7380 }, { "epoch": 1.0549627038795104, "grad_norm": 1.1268243789672852, "learning_rate": 0.00012976910259462034, "loss": 2.020710563659668, "step": 7390 }, { "epoch": 1.0563903065776794, "grad_norm": 1.083245038986206, "learning_rate": 0.00012967388716972152, "loss": 2.0099674224853517, "step": 7400 }, { "epoch": 1.0578179092758486, "grad_norm": 1.2687695026397705, "learning_rate": 0.00012957867174482267, "loss": 2.1577903747558596, "step": 7410 }, { "epoch": 1.0592455119740176, "grad_norm": 1.20366632938385, "learning_rate": 0.00012948345631992383, "loss": 2.077815818786621, "step": 7420 }, { "epoch": 1.0606731146721868, "grad_norm": 1.5873297452926636, "learning_rate": 0.00012938824089502499, "loss": 1.9076950073242187, "step": 7430 }, { "epoch": 1.0621007173703558, "grad_norm": 1.4004929065704346, "learning_rate": 0.00012929302547012617, "loss": 2.035087013244629, "step": 7440 }, { "epoch": 1.0635283200685248, "grad_norm": 1.0319706201553345, "learning_rate": 0.00012919781004522732, "loss": 1.9221908569335937, "step": 7450 }, { "epoch": 1.064955922766694, "grad_norm": 1.032630205154419, "learning_rate": 0.0001291025946203285, "loss": 1.8860836029052734, "step": 7460 }, { "epoch": 1.066383525464863, "grad_norm": 1.0436253547668457, "learning_rate": 0.00012900737919542966, "loss": 1.9600419998168945, "step": 7470 }, { "epoch": 1.0678111281630323, "grad_norm": 1.5339142084121704, "learning_rate": 0.00012891216377053084, "loss": 1.9912384033203125, "step": 7480 }, { "epoch": 1.0692387308612012, "grad_norm": 1.2523791790008545, "learning_rate": 0.000128816948345632, "loss": 2.0709794998168944, "step": 7490 }, { "epoch": 1.0706663335593705, "grad_norm": 1.5877113342285156, "learning_rate": 0.00012872173292073318, "loss": 1.9899673461914062, "step": 7500 }, { "epoch": 1.0720939362575395, "grad_norm": 1.0728774070739746, "learning_rate": 0.00012862651749583434, "loss": 2.048723602294922, "step": 7510 }, { "epoch": 1.0735215389557087, "grad_norm": 1.0678832530975342, "learning_rate": 0.0001285313020709355, "loss": 2.162331199645996, "step": 7520 }, { "epoch": 1.0749491416538777, "grad_norm": 1.6823121309280396, "learning_rate": 0.00012843608664603665, "loss": 1.980549430847168, "step": 7530 }, { "epoch": 1.0763767443520469, "grad_norm": 1.089056134223938, "learning_rate": 0.00012834087122113783, "loss": 2.1248153686523437, "step": 7540 }, { "epoch": 1.0778043470502159, "grad_norm": 1.356019139289856, "learning_rate": 0.00012824565579623898, "loss": 2.03244686126709, "step": 7550 }, { "epoch": 1.079231949748385, "grad_norm": 1.0798354148864746, "learning_rate": 0.00012815044037134017, "loss": 2.043965530395508, "step": 7560 }, { "epoch": 1.080659552446554, "grad_norm": 1.2939257621765137, "learning_rate": 0.00012805522494644132, "loss": 1.8783649444580077, "step": 7570 }, { "epoch": 1.0820871551447233, "grad_norm": 1.0123037099838257, "learning_rate": 0.0001279600095215425, "loss": 2.080133247375488, "step": 7580 }, { "epoch": 1.0835147578428923, "grad_norm": 1.2057385444641113, "learning_rate": 0.00012786479409664366, "loss": 2.1240022659301756, "step": 7590 }, { "epoch": 1.0849423605410615, "grad_norm": 1.3064265251159668, "learning_rate": 0.00012776957867174484, "loss": 1.9333322525024415, "step": 7600 }, { "epoch": 1.0863699632392305, "grad_norm": 1.057296633720398, "learning_rate": 0.000127674363246846, "loss": 2.183109664916992, "step": 7610 }, { "epoch": 1.0877975659373997, "grad_norm": 1.6201916933059692, "learning_rate": 0.00012757914782194715, "loss": 2.0552255630493166, "step": 7620 }, { "epoch": 1.0892251686355687, "grad_norm": 1.340762734413147, "learning_rate": 0.00012748393239704834, "loss": 2.2243988037109377, "step": 7630 }, { "epoch": 1.0906527713337377, "grad_norm": 1.1373904943466187, "learning_rate": 0.0001273887169721495, "loss": 2.0690093994140626, "step": 7640 }, { "epoch": 1.092080374031907, "grad_norm": 1.7897838354110718, "learning_rate": 0.00012729350154725065, "loss": 2.044172477722168, "step": 7650 }, { "epoch": 1.093507976730076, "grad_norm": 1.0654335021972656, "learning_rate": 0.00012719828612235183, "loss": 1.9154827117919921, "step": 7660 }, { "epoch": 1.0949355794282452, "grad_norm": 1.0672240257263184, "learning_rate": 0.00012710307069745298, "loss": 1.9989141464233398, "step": 7670 }, { "epoch": 1.0963631821264141, "grad_norm": 1.695644736289978, "learning_rate": 0.00012700785527255417, "loss": 2.020133399963379, "step": 7680 }, { "epoch": 1.0977907848245834, "grad_norm": 1.2182796001434326, "learning_rate": 0.00012691263984765532, "loss": 2.1288455963134765, "step": 7690 }, { "epoch": 1.0992183875227524, "grad_norm": 1.9532301425933838, "learning_rate": 0.0001268174244227565, "loss": 2.0164491653442385, "step": 7700 }, { "epoch": 1.1006459902209216, "grad_norm": 1.1174538135528564, "learning_rate": 0.00012672220899785766, "loss": 2.127495002746582, "step": 7710 }, { "epoch": 1.1020735929190906, "grad_norm": 1.1028709411621094, "learning_rate": 0.00012662699357295881, "loss": 2.0858892440795898, "step": 7720 }, { "epoch": 1.1035011956172598, "grad_norm": 1.2421051263809204, "learning_rate": 0.00012653177814806, "loss": 1.9897558212280273, "step": 7730 }, { "epoch": 1.1049287983154288, "grad_norm": 1.3325999975204468, "learning_rate": 0.00012643656272316115, "loss": 2.1262748718261717, "step": 7740 }, { "epoch": 1.106356401013598, "grad_norm": 1.3034075498580933, "learning_rate": 0.00012634134729826234, "loss": 2.0708587646484373, "step": 7750 }, { "epoch": 1.107784003711767, "grad_norm": 1.0501512289047241, "learning_rate": 0.0001262461318733635, "loss": 2.039535903930664, "step": 7760 }, { "epoch": 1.1092116064099362, "grad_norm": 1.0809801816940308, "learning_rate": 0.00012615091644846467, "loss": 2.0607528686523438, "step": 7770 }, { "epoch": 1.1106392091081052, "grad_norm": 1.3646125793457031, "learning_rate": 0.00012605570102356583, "loss": 2.096577262878418, "step": 7780 }, { "epoch": 1.1120668118062742, "grad_norm": 1.1288775205612183, "learning_rate": 0.000125960485598667, "loss": 2.096157455444336, "step": 7790 }, { "epoch": 1.1134944145044434, "grad_norm": 1.2515203952789307, "learning_rate": 0.00012586527017376814, "loss": 2.0567888259887694, "step": 7800 }, { "epoch": 1.1149220172026124, "grad_norm": 1.200783610343933, "learning_rate": 0.00012577005474886932, "loss": 1.9918567657470703, "step": 7810 }, { "epoch": 1.1163496199007816, "grad_norm": 1.0647796392440796, "learning_rate": 0.00012567483932397048, "loss": 2.1152278900146486, "step": 7820 }, { "epoch": 1.1177772225989506, "grad_norm": 1.3672425746917725, "learning_rate": 0.00012557962389907166, "loss": 2.1503089904785155, "step": 7830 }, { "epoch": 1.1192048252971198, "grad_norm": 1.1135179996490479, "learning_rate": 0.00012548440847417281, "loss": 2.1828407287597655, "step": 7840 }, { "epoch": 1.1206324279952888, "grad_norm": 1.2473934888839722, "learning_rate": 0.000125389193049274, "loss": 1.8900934219360352, "step": 7850 }, { "epoch": 1.122060030693458, "grad_norm": 1.1976487636566162, "learning_rate": 0.00012529397762437515, "loss": 2.0562448501586914, "step": 7860 }, { "epoch": 1.123487633391627, "grad_norm": 1.0255727767944336, "learning_rate": 0.00012519876219947633, "loss": 2.025206184387207, "step": 7870 }, { "epoch": 1.1249152360897963, "grad_norm": 1.0995827913284302, "learning_rate": 0.0001251035467745775, "loss": 2.009837341308594, "step": 7880 }, { "epoch": 1.1263428387879653, "grad_norm": 1.218954086303711, "learning_rate": 0.00012500833134967867, "loss": 1.9312389373779297, "step": 7890 }, { "epoch": 1.1277704414861345, "grad_norm": 1.1379914283752441, "learning_rate": 0.0001249131159247798, "loss": 2.085856628417969, "step": 7900 }, { "epoch": 1.1291980441843035, "grad_norm": 1.04413902759552, "learning_rate": 0.00012481790049988098, "loss": 1.9226675033569336, "step": 7910 }, { "epoch": 1.1306256468824727, "grad_norm": 1.1539976596832275, "learning_rate": 0.00012472268507498214, "loss": 2.073196029663086, "step": 7920 }, { "epoch": 1.1320532495806417, "grad_norm": 1.3066080808639526, "learning_rate": 0.00012462746965008332, "loss": 2.130576515197754, "step": 7930 }, { "epoch": 1.133480852278811, "grad_norm": 1.0937031507492065, "learning_rate": 0.00012453225422518448, "loss": 1.949639892578125, "step": 7940 }, { "epoch": 1.13490845497698, "grad_norm": 1.4130394458770752, "learning_rate": 0.00012443703880028566, "loss": 1.9901809692382812, "step": 7950 }, { "epoch": 1.1363360576751491, "grad_norm": 0.9927778244018555, "learning_rate": 0.00012434182337538681, "loss": 2.026361083984375, "step": 7960 }, { "epoch": 1.137763660373318, "grad_norm": 1.1941503286361694, "learning_rate": 0.000124246607950488, "loss": 2.039238929748535, "step": 7970 }, { "epoch": 1.139191263071487, "grad_norm": 1.0839687585830688, "learning_rate": 0.00012415139252558915, "loss": 2.113592338562012, "step": 7980 }, { "epoch": 1.1406188657696563, "grad_norm": 1.1398208141326904, "learning_rate": 0.0001240656986431802, "loss": 1.9601491928100585, "step": 7990 }, { "epoch": 1.1420464684678253, "grad_norm": 1.6864898204803467, "learning_rate": 0.00012397048321828138, "loss": 1.9921035766601562, "step": 8000 }, { "epoch": 1.1434740711659945, "grad_norm": 1.2941198348999023, "learning_rate": 0.00012387526779338254, "loss": 1.9406307220458985, "step": 8010 }, { "epoch": 1.1449016738641635, "grad_norm": 1.1953562498092651, "learning_rate": 0.0001237800523684837, "loss": 1.9879423141479493, "step": 8020 }, { "epoch": 1.1463292765623327, "grad_norm": 1.21220862865448, "learning_rate": 0.00012368483694358485, "loss": 1.8858787536621093, "step": 8030 }, { "epoch": 1.1477568792605017, "grad_norm": 1.2785707712173462, "learning_rate": 0.00012358962151868603, "loss": 1.9584150314331055, "step": 8040 }, { "epoch": 1.149184481958671, "grad_norm": 0.9782488346099854, "learning_rate": 0.00012349440609378719, "loss": 1.9678041458129882, "step": 8050 }, { "epoch": 1.15061208465684, "grad_norm": 1.1952399015426636, "learning_rate": 0.00012339919066888837, "loss": 2.018492317199707, "step": 8060 }, { "epoch": 1.1520396873550092, "grad_norm": 1.2737188339233398, "learning_rate": 0.00012330397524398952, "loss": 2.065995788574219, "step": 8070 }, { "epoch": 1.1534672900531782, "grad_norm": 1.0872756242752075, "learning_rate": 0.0001232087598190907, "loss": 1.9526090621948242, "step": 8080 }, { "epoch": 1.1548948927513474, "grad_norm": 0.9328494071960449, "learning_rate": 0.00012311354439419186, "loss": 2.007294273376465, "step": 8090 }, { "epoch": 1.1563224954495164, "grad_norm": 0.9385980367660522, "learning_rate": 0.00012301832896929304, "loss": 2.040877342224121, "step": 8100 }, { "epoch": 1.1577500981476856, "grad_norm": 1.0632121562957764, "learning_rate": 0.0001229231135443942, "loss": 1.9517059326171875, "step": 8110 }, { "epoch": 1.1591777008458546, "grad_norm": 1.0268210172653198, "learning_rate": 0.00012282789811949535, "loss": 2.130423164367676, "step": 8120 }, { "epoch": 1.1606053035440236, "grad_norm": 1.0808130502700806, "learning_rate": 0.00012273268269459654, "loss": 1.934481430053711, "step": 8130 }, { "epoch": 1.1620329062421928, "grad_norm": 1.132352590560913, "learning_rate": 0.0001226374672696977, "loss": 2.1073892593383787, "step": 8140 }, { "epoch": 1.163460508940362, "grad_norm": 1.0396180152893066, "learning_rate": 0.00012254225184479887, "loss": 1.924516487121582, "step": 8150 }, { "epoch": 1.164888111638531, "grad_norm": 1.0108245611190796, "learning_rate": 0.00012244703641990003, "loss": 2.046940231323242, "step": 8160 }, { "epoch": 1.1663157143367, "grad_norm": 1.2685633897781372, "learning_rate": 0.0001223518209950012, "loss": 2.028103256225586, "step": 8170 }, { "epoch": 1.1677433170348692, "grad_norm": 1.098732352256775, "learning_rate": 0.00012225660557010237, "loss": 2.1019420623779297, "step": 8180 }, { "epoch": 1.1691709197330382, "grad_norm": 1.0802416801452637, "learning_rate": 0.00012216139014520352, "loss": 2.0794830322265625, "step": 8190 }, { "epoch": 1.1705985224312074, "grad_norm": 1.1557440757751465, "learning_rate": 0.0001220661747203047, "loss": 2.001494216918945, "step": 8200 }, { "epoch": 1.1720261251293764, "grad_norm": 1.1953849792480469, "learning_rate": 0.00012197095929540585, "loss": 2.173105812072754, "step": 8210 }, { "epoch": 1.1734537278275456, "grad_norm": 1.444016456604004, "learning_rate": 0.00012187574387050702, "loss": 2.0334218978881835, "step": 8220 }, { "epoch": 1.1748813305257146, "grad_norm": 1.169856071472168, "learning_rate": 0.00012178052844560818, "loss": 2.1740755081176757, "step": 8230 }, { "epoch": 1.1763089332238839, "grad_norm": 1.2645282745361328, "learning_rate": 0.00012168531302070935, "loss": 2.049808883666992, "step": 8240 }, { "epoch": 1.1777365359220529, "grad_norm": 1.252223253250122, "learning_rate": 0.00012159009759581052, "loss": 2.0915248870849608, "step": 8250 }, { "epoch": 1.179164138620222, "grad_norm": 1.2396661043167114, "learning_rate": 0.00012149488217091169, "loss": 2.0347517013549803, "step": 8260 }, { "epoch": 1.180591741318391, "grad_norm": 1.161691665649414, "learning_rate": 0.00012139966674601286, "loss": 2.0568159103393553, "step": 8270 }, { "epoch": 1.1820193440165603, "grad_norm": 1.38312566280365, "learning_rate": 0.00012130445132111403, "loss": 1.9940105438232423, "step": 8280 }, { "epoch": 1.1834469467147293, "grad_norm": 1.3739749193191528, "learning_rate": 0.0001212092358962152, "loss": 1.9481508255004882, "step": 8290 }, { "epoch": 1.1848745494128985, "grad_norm": 1.1918368339538574, "learning_rate": 0.00012111402047131637, "loss": 1.9933198928833007, "step": 8300 }, { "epoch": 1.1863021521110675, "grad_norm": 1.2531394958496094, "learning_rate": 0.00012101880504641752, "loss": 1.9597272872924805, "step": 8310 }, { "epoch": 1.1877297548092365, "grad_norm": 1.1807265281677246, "learning_rate": 0.00012092358962151869, "loss": 2.1743324279785154, "step": 8320 }, { "epoch": 1.1891573575074057, "grad_norm": 1.4080866575241089, "learning_rate": 0.00012082837419661986, "loss": 2.091488265991211, "step": 8330 }, { "epoch": 1.1905849602055747, "grad_norm": 1.3193141222000122, "learning_rate": 0.0001207426803142109, "loss": 1.9972221374511718, "step": 8340 }, { "epoch": 1.192012562903744, "grad_norm": 1.0870722532272339, "learning_rate": 0.00012064746488931206, "loss": 1.9719045639038086, "step": 8350 }, { "epoch": 1.193440165601913, "grad_norm": 1.18523371219635, "learning_rate": 0.00012055224946441323, "loss": 2.0243814468383787, "step": 8360 }, { "epoch": 1.1948677683000821, "grad_norm": 1.0185394287109375, "learning_rate": 0.0001204570340395144, "loss": 2.1613880157470704, "step": 8370 }, { "epoch": 1.1962953709982511, "grad_norm": 1.1819486618041992, "learning_rate": 0.00012036181861461557, "loss": 2.0223106384277343, "step": 8380 }, { "epoch": 1.1977229736964203, "grad_norm": 0.9671928286552429, "learning_rate": 0.00012026660318971674, "loss": 2.039589691162109, "step": 8390 }, { "epoch": 1.1991505763945893, "grad_norm": 1.17135751247406, "learning_rate": 0.00012017138776481791, "loss": 1.9769683837890626, "step": 8400 }, { "epoch": 1.2005781790927585, "grad_norm": 1.0596065521240234, "learning_rate": 0.00012007617233991908, "loss": 2.1013067245483397, "step": 8410 }, { "epoch": 1.2020057817909275, "grad_norm": 1.0273828506469727, "learning_rate": 0.00011998095691502024, "loss": 2.0737178802490233, "step": 8420 }, { "epoch": 1.2034333844890968, "grad_norm": 1.1414293050765991, "learning_rate": 0.0001198857414901214, "loss": 2.0065332412719727, "step": 8430 }, { "epoch": 1.2048609871872658, "grad_norm": 1.2087929248809814, "learning_rate": 0.00011979052606522257, "loss": 2.093004608154297, "step": 8440 }, { "epoch": 1.206288589885435, "grad_norm": 1.0301358699798584, "learning_rate": 0.00011969531064032374, "loss": 2.0411731719970705, "step": 8450 }, { "epoch": 1.207716192583604, "grad_norm": 0.9773324131965637, "learning_rate": 0.00011960009521542491, "loss": 1.9705137252807616, "step": 8460 }, { "epoch": 1.209143795281773, "grad_norm": 1.3583508729934692, "learning_rate": 0.00011950487979052608, "loss": 2.15179328918457, "step": 8470 }, { "epoch": 1.2105713979799422, "grad_norm": 1.3190714120864868, "learning_rate": 0.00011940966436562724, "loss": 2.072892761230469, "step": 8480 }, { "epoch": 1.2119990006781114, "grad_norm": 1.3944333791732788, "learning_rate": 0.00011931444894072841, "loss": 2.1775245666503906, "step": 8490 }, { "epoch": 1.2134266033762804, "grad_norm": 1.3343610763549805, "learning_rate": 0.00011921923351582958, "loss": 2.1058820724487304, "step": 8500 }, { "epoch": 1.2148542060744494, "grad_norm": 1.054322600364685, "learning_rate": 0.00011912401809093075, "loss": 2.0245025634765623, "step": 8510 }, { "epoch": 1.2162818087726186, "grad_norm": 1.114761471748352, "learning_rate": 0.0001190288026660319, "loss": 1.9700067520141602, "step": 8520 }, { "epoch": 1.2177094114707876, "grad_norm": 1.0727494955062866, "learning_rate": 0.00011893358724113306, "loss": 1.9866079330444335, "step": 8530 }, { "epoch": 1.2191370141689568, "grad_norm": 1.1413354873657227, "learning_rate": 0.00011883837181623423, "loss": 2.181919288635254, "step": 8540 }, { "epoch": 1.2205646168671258, "grad_norm": 1.3694534301757812, "learning_rate": 0.0001187431563913354, "loss": 2.078506278991699, "step": 8550 }, { "epoch": 1.221992219565295, "grad_norm": 1.1679607629776, "learning_rate": 0.00011864794096643657, "loss": 1.9354976654052733, "step": 8560 }, { "epoch": 1.223419822263464, "grad_norm": 1.2007423639297485, "learning_rate": 0.00011855272554153774, "loss": 1.9437042236328126, "step": 8570 }, { "epoch": 1.2248474249616332, "grad_norm": 1.2228180170059204, "learning_rate": 0.0001184575101166389, "loss": 2.012160873413086, "step": 8580 }, { "epoch": 1.2262750276598022, "grad_norm": 1.3981947898864746, "learning_rate": 0.00011836229469174008, "loss": 2.103215217590332, "step": 8590 }, { "epoch": 1.2277026303579714, "grad_norm": 1.0683261156082153, "learning_rate": 0.00011826707926684124, "loss": 2.1170129776000977, "step": 8600 }, { "epoch": 1.2291302330561404, "grad_norm": 0.9289690852165222, "learning_rate": 0.00011817186384194241, "loss": 2.1273303985595704, "step": 8610 }, { "epoch": 1.2305578357543097, "grad_norm": 1.4348978996276855, "learning_rate": 0.00011807664841704355, "loss": 2.068717384338379, "step": 8620 }, { "epoch": 1.2319854384524787, "grad_norm": 1.3924267292022705, "learning_rate": 0.00011798143299214472, "loss": 2.142212677001953, "step": 8630 }, { "epoch": 1.2334130411506479, "grad_norm": 1.0877633094787598, "learning_rate": 0.00011788621756724589, "loss": 2.0361440658569334, "step": 8640 }, { "epoch": 1.2348406438488169, "grad_norm": 1.381314992904663, "learning_rate": 0.00011779100214234706, "loss": 1.9098474502563476, "step": 8650 }, { "epoch": 1.2362682465469859, "grad_norm": 0.9824159145355225, "learning_rate": 0.00011769578671744823, "loss": 2.040357971191406, "step": 8660 }, { "epoch": 1.237695849245155, "grad_norm": 1.135865330696106, "learning_rate": 0.0001176005712925494, "loss": 1.9138475418090821, "step": 8670 }, { "epoch": 1.2391234519433243, "grad_norm": 1.0527571439743042, "learning_rate": 0.00011750535586765057, "loss": 2.1740653991699217, "step": 8680 }, { "epoch": 1.2405510546414933, "grad_norm": 1.1723603010177612, "learning_rate": 0.00011741014044275174, "loss": 2.027785301208496, "step": 8690 }, { "epoch": 1.2419786573396623, "grad_norm": 1.3554455041885376, "learning_rate": 0.0001173149250178529, "loss": 2.0539613723754884, "step": 8700 }, { "epoch": 1.2434062600378315, "grad_norm": 1.1202641725540161, "learning_rate": 0.00011721970959295407, "loss": 1.8522020339965821, "step": 8710 }, { "epoch": 1.2448338627360005, "grad_norm": 0.9917465448379517, "learning_rate": 0.00011712449416805522, "loss": 1.9608482360839843, "step": 8720 }, { "epoch": 1.2462614654341697, "grad_norm": 1.4097355604171753, "learning_rate": 0.00011702927874315639, "loss": 2.022862434387207, "step": 8730 }, { "epoch": 1.2476890681323387, "grad_norm": 1.2008345127105713, "learning_rate": 0.00011693406331825755, "loss": 2.0185220718383787, "step": 8740 }, { "epoch": 1.249116670830508, "grad_norm": 1.345367670059204, "learning_rate": 0.00011683884789335872, "loss": 2.106128692626953, "step": 8750 }, { "epoch": 1.250544273528677, "grad_norm": 0.9907132387161255, "learning_rate": 0.00011674363246845989, "loss": 2.0598371505737303, "step": 8760 }, { "epoch": 1.2519718762268461, "grad_norm": 1.243442177772522, "learning_rate": 0.00011664841704356106, "loss": 2.0366870880126955, "step": 8770 }, { "epoch": 1.2533994789250151, "grad_norm": 1.1092184782028198, "learning_rate": 0.00011655320161866223, "loss": 2.1104217529296876, "step": 8780 }, { "epoch": 1.2548270816231843, "grad_norm": 1.2324589490890503, "learning_rate": 0.0001164579861937634, "loss": 2.0124040603637696, "step": 8790 }, { "epoch": 1.2562546843213533, "grad_norm": 1.085134744644165, "learning_rate": 0.00011636277076886457, "loss": 1.952083206176758, "step": 8800 }, { "epoch": 1.2576822870195223, "grad_norm": 1.182888150215149, "learning_rate": 0.00011626755534396572, "loss": 2.0146764755249023, "step": 8810 }, { "epoch": 1.2591098897176916, "grad_norm": 1.1887840032577515, "learning_rate": 0.00011617233991906689, "loss": 2.030648612976074, "step": 8820 }, { "epoch": 1.2605374924158608, "grad_norm": 1.1329022645950317, "learning_rate": 0.00011607712449416806, "loss": 1.9215503692626954, "step": 8830 }, { "epoch": 1.2619650951140298, "grad_norm": 1.496118426322937, "learning_rate": 0.00011598190906926923, "loss": 2.0997610092163086, "step": 8840 }, { "epoch": 1.2633926978121988, "grad_norm": 1.2503148317337036, "learning_rate": 0.0001158866936443704, "loss": 2.0191768646240233, "step": 8850 }, { "epoch": 1.264820300510368, "grad_norm": 1.0068776607513428, "learning_rate": 0.00011579147821947157, "loss": 2.078872871398926, "step": 8860 }, { "epoch": 1.2662479032085372, "grad_norm": 1.4307090044021606, "learning_rate": 0.00011569626279457274, "loss": 2.0454170227050783, "step": 8870 }, { "epoch": 1.2676755059067062, "grad_norm": 1.1627181768417358, "learning_rate": 0.0001156010473696739, "loss": 2.0517635345458984, "step": 8880 }, { "epoch": 1.2691031086048752, "grad_norm": 1.3769811391830444, "learning_rate": 0.00011550583194477507, "loss": 2.0428056716918945, "step": 8890 }, { "epoch": 1.2705307113030444, "grad_norm": 1.1414709091186523, "learning_rate": 0.00011541061651987624, "loss": 2.014663314819336, "step": 8900 }, { "epoch": 1.2719583140012134, "grad_norm": 1.6799501180648804, "learning_rate": 0.00011531540109497738, "loss": 1.9326515197753906, "step": 8910 }, { "epoch": 1.2733859166993826, "grad_norm": 1.3793418407440186, "learning_rate": 0.00011522018567007855, "loss": 2.0053300857543945, "step": 8920 }, { "epoch": 1.2748135193975516, "grad_norm": 1.0783268213272095, "learning_rate": 0.00011512497024517972, "loss": 2.0390304565429687, "step": 8930 }, { "epoch": 1.2762411220957208, "grad_norm": 1.2486910820007324, "learning_rate": 0.00011502975482028089, "loss": 2.0407100677490235, "step": 8940 }, { "epoch": 1.2776687247938898, "grad_norm": 1.0730825662612915, "learning_rate": 0.00011493453939538206, "loss": 2.1652347564697267, "step": 8950 }, { "epoch": 1.279096327492059, "grad_norm": 1.4436947107315063, "learning_rate": 0.00011483932397048323, "loss": 2.057954216003418, "step": 8960 }, { "epoch": 1.280523930190228, "grad_norm": 1.7020477056503296, "learning_rate": 0.0001147441085455844, "loss": 2.075143241882324, "step": 8970 }, { "epoch": 1.2819515328883972, "grad_norm": 1.0034456253051758, "learning_rate": 0.00011464889312068557, "loss": 2.0679399490356447, "step": 8980 }, { "epoch": 1.2833791355865662, "grad_norm": 1.1735398769378662, "learning_rate": 0.00011455367769578674, "loss": 2.1125640869140625, "step": 8990 }, { "epoch": 1.2848067382847352, "grad_norm": 1.1050878763198853, "learning_rate": 0.00011445846227088788, "loss": 1.9761398315429688, "step": 9000 }, { "epoch": 1.2862343409829045, "grad_norm": 1.0764317512512207, "learning_rate": 0.00011436324684598905, "loss": 1.9444828033447266, "step": 9010 }, { "epoch": 1.2876619436810737, "grad_norm": 1.1503080129623413, "learning_rate": 0.00011426803142109021, "loss": 2.097242736816406, "step": 9020 }, { "epoch": 1.2890895463792427, "grad_norm": 1.1134028434753418, "learning_rate": 0.00011417281599619138, "loss": 1.9036460876464845, "step": 9030 }, { "epoch": 1.2905171490774117, "grad_norm": 1.198632836341858, "learning_rate": 0.00011407760057129255, "loss": 2.0219871520996096, "step": 9040 }, { "epoch": 1.2919447517755809, "grad_norm": 1.2109109163284302, "learning_rate": 0.00011398238514639372, "loss": 1.9111330032348632, "step": 9050 }, { "epoch": 1.2933723544737499, "grad_norm": 1.4551175832748413, "learning_rate": 0.00011388716972149489, "loss": 1.9579683303833009, "step": 9060 }, { "epoch": 1.294799957171919, "grad_norm": 1.3260098695755005, "learning_rate": 0.00011379195429659606, "loss": 2.068608856201172, "step": 9070 }, { "epoch": 1.296227559870088, "grad_norm": 1.411226749420166, "learning_rate": 0.00011369673887169723, "loss": 2.023783874511719, "step": 9080 }, { "epoch": 1.2976551625682573, "grad_norm": 1.2468092441558838, "learning_rate": 0.0001136015234467984, "loss": 1.9828487396240235, "step": 9090 }, { "epoch": 1.2990827652664263, "grad_norm": 0.9752615690231323, "learning_rate": 0.00011350630802189954, "loss": 2.072840690612793, "step": 9100 }, { "epoch": 1.3005103679645955, "grad_norm": 1.3394255638122559, "learning_rate": 0.00011341109259700071, "loss": 2.084597587585449, "step": 9110 }, { "epoch": 1.3019379706627645, "grad_norm": 1.3784369230270386, "learning_rate": 0.00011331587717210188, "loss": 1.9246038436889648, "step": 9120 }, { "epoch": 1.3033655733609337, "grad_norm": 1.130905270576477, "learning_rate": 0.00011322066174720305, "loss": 1.9275218963623046, "step": 9130 }, { "epoch": 1.3047931760591027, "grad_norm": 1.239323377609253, "learning_rate": 0.00011312544632230421, "loss": 2.207936096191406, "step": 9140 }, { "epoch": 1.3062207787572717, "grad_norm": 0.944084107875824, "learning_rate": 0.00011303023089740538, "loss": 2.156756782531738, "step": 9150 }, { "epoch": 1.307648381455441, "grad_norm": 1.5705152750015259, "learning_rate": 0.00011293501547250655, "loss": 2.1005443572998046, "step": 9160 }, { "epoch": 1.3090759841536102, "grad_norm": 1.3107826709747314, "learning_rate": 0.00011283980004760772, "loss": 2.0039052963256836, "step": 9170 }, { "epoch": 1.3105035868517791, "grad_norm": 1.1114449501037598, "learning_rate": 0.00011274458462270889, "loss": 2.077981376647949, "step": 9180 }, { "epoch": 1.3119311895499481, "grad_norm": 1.329028606414795, "learning_rate": 0.00011264936919781004, "loss": 2.1001474380493166, "step": 9190 }, { "epoch": 1.3133587922481174, "grad_norm": 1.7753056287765503, "learning_rate": 0.00011255415377291121, "loss": 1.9850801467895507, "step": 9200 }, { "epoch": 1.3147863949462866, "grad_norm": 1.1210708618164062, "learning_rate": 0.00011245893834801238, "loss": 2.1043624877929688, "step": 9210 }, { "epoch": 1.3162139976444556, "grad_norm": 1.201124906539917, "learning_rate": 0.00011236372292311355, "loss": 2.180091667175293, "step": 9220 }, { "epoch": 1.3176416003426246, "grad_norm": 1.104976773262024, "learning_rate": 0.00011226850749821472, "loss": 2.022638130187988, "step": 9230 }, { "epoch": 1.3190692030407938, "grad_norm": 1.1997878551483154, "learning_rate": 0.00011217329207331589, "loss": 1.9806930541992187, "step": 9240 }, { "epoch": 1.3204968057389628, "grad_norm": 1.1831468343734741, "learning_rate": 0.00011207807664841706, "loss": 1.889851951599121, "step": 9250 }, { "epoch": 1.321924408437132, "grad_norm": 1.3376766443252563, "learning_rate": 0.00011199238276600809, "loss": 2.2721225738525392, "step": 9260 }, { "epoch": 1.323352011135301, "grad_norm": 1.2791374921798706, "learning_rate": 0.00011189716734110926, "loss": 2.0228946685791014, "step": 9270 }, { "epoch": 1.3247796138334702, "grad_norm": 1.138617992401123, "learning_rate": 0.00011180195191621043, "loss": 2.1891841888427734, "step": 9280 }, { "epoch": 1.3262072165316392, "grad_norm": 1.39596688747406, "learning_rate": 0.0001117067364913116, "loss": 2.1259082794189452, "step": 9290 }, { "epoch": 1.3276348192298084, "grad_norm": 1.1990619897842407, "learning_rate": 0.00011161152106641277, "loss": 2.025862121582031, "step": 9300 }, { "epoch": 1.3290624219279774, "grad_norm": 1.2685766220092773, "learning_rate": 0.00011151630564151394, "loss": 1.9857549667358398, "step": 9310 }, { "epoch": 1.3304900246261466, "grad_norm": 1.1520808935165405, "learning_rate": 0.00011142109021661509, "loss": 2.2395219802856445, "step": 9320 }, { "epoch": 1.3319176273243156, "grad_norm": 1.1049847602844238, "learning_rate": 0.00011132587479171626, "loss": 1.9089387893676757, "step": 9330 }, { "epoch": 1.3333452300224846, "grad_norm": 1.3146041631698608, "learning_rate": 0.00011123065936681743, "loss": 1.9961027145385741, "step": 9340 }, { "epoch": 1.3347728327206538, "grad_norm": 1.2293838262557983, "learning_rate": 0.0001111354439419186, "loss": 1.9925750732421874, "step": 9350 }, { "epoch": 1.336200435418823, "grad_norm": 1.1773251295089722, "learning_rate": 0.00011104022851701977, "loss": 1.941655158996582, "step": 9360 }, { "epoch": 1.337628038116992, "grad_norm": 0.8993968963623047, "learning_rate": 0.00011094501309212094, "loss": 1.8487606048583984, "step": 9370 }, { "epoch": 1.339055640815161, "grad_norm": 1.0430762767791748, "learning_rate": 0.0001108497976672221, "loss": 2.0043895721435545, "step": 9380 }, { "epoch": 1.3404832435133303, "grad_norm": 1.2599635124206543, "learning_rate": 0.00011075458224232327, "loss": 2.014158248901367, "step": 9390 }, { "epoch": 1.3419108462114995, "grad_norm": 1.3002982139587402, "learning_rate": 0.00011065936681742444, "loss": 1.8953874588012696, "step": 9400 }, { "epoch": 1.3433384489096685, "grad_norm": 1.1994285583496094, "learning_rate": 0.00011056415139252558, "loss": 2.0363245010375977, "step": 9410 }, { "epoch": 1.3447660516078375, "grad_norm": 1.0855350494384766, "learning_rate": 0.00011046893596762675, "loss": 1.9903108596801757, "step": 9420 }, { "epoch": 1.3461936543060067, "grad_norm": 0.9542500376701355, "learning_rate": 0.00011037372054272792, "loss": 2.1463512420654296, "step": 9430 }, { "epoch": 1.3476212570041757, "grad_norm": 1.1135118007659912, "learning_rate": 0.00011027850511782909, "loss": 1.8799579620361329, "step": 9440 }, { "epoch": 1.349048859702345, "grad_norm": 1.2428560256958008, "learning_rate": 0.00011018328969293026, "loss": 1.9707263946533202, "step": 9450 }, { "epoch": 1.3504764624005139, "grad_norm": 1.0199036598205566, "learning_rate": 0.00011008807426803143, "loss": 2.1025836944580076, "step": 9460 }, { "epoch": 1.351904065098683, "grad_norm": 1.1156866550445557, "learning_rate": 0.0001099928588431326, "loss": 2.1563234329223633, "step": 9470 }, { "epoch": 1.353331667796852, "grad_norm": 1.1687692403793335, "learning_rate": 0.00010989764341823377, "loss": 2.125259590148926, "step": 9480 }, { "epoch": 1.3547592704950213, "grad_norm": 0.9904701709747314, "learning_rate": 0.00010980242799333494, "loss": 1.9423274993896484, "step": 9490 }, { "epoch": 1.3561868731931903, "grad_norm": 1.0586273670196533, "learning_rate": 0.0001097072125684361, "loss": 1.995172119140625, "step": 9500 }, { "epoch": 1.3576144758913595, "grad_norm": 1.2330423593521118, "learning_rate": 0.00010961199714353725, "loss": 2.121123504638672, "step": 9510 }, { "epoch": 1.3590420785895285, "grad_norm": 1.1340503692626953, "learning_rate": 0.00010951678171863842, "loss": 2.0948347091674804, "step": 9520 }, { "epoch": 1.3604696812876975, "grad_norm": 1.1737462282180786, "learning_rate": 0.00010942156629373958, "loss": 1.9511737823486328, "step": 9530 }, { "epoch": 1.3618972839858667, "grad_norm": 1.3858624696731567, "learning_rate": 0.00010932635086884075, "loss": 2.1292884826660154, "step": 9540 }, { "epoch": 1.363324886684036, "grad_norm": 1.184073805809021, "learning_rate": 0.00010923113544394192, "loss": 2.0114479064941406, "step": 9550 }, { "epoch": 1.364752489382205, "grad_norm": 1.3248467445373535, "learning_rate": 0.00010913592001904309, "loss": 2.0445907592773436, "step": 9560 }, { "epoch": 1.366180092080374, "grad_norm": 1.407776117324829, "learning_rate": 0.00010904070459414426, "loss": 1.9902603149414062, "step": 9570 }, { "epoch": 1.3676076947785432, "grad_norm": 1.0734567642211914, "learning_rate": 0.00010894548916924543, "loss": 1.9612981796264648, "step": 9580 }, { "epoch": 1.3690352974767122, "grad_norm": 1.138846755027771, "learning_rate": 0.0001088502737443466, "loss": 1.9796411514282226, "step": 9590 }, { "epoch": 1.3704629001748814, "grad_norm": 1.2043484449386597, "learning_rate": 0.00010875505831944777, "loss": 1.9225820541381835, "step": 9600 }, { "epoch": 1.3718905028730504, "grad_norm": 1.3923934698104858, "learning_rate": 0.00010865984289454891, "loss": 1.9597026824951171, "step": 9610 }, { "epoch": 1.3733181055712196, "grad_norm": 1.1692254543304443, "learning_rate": 0.00010856462746965008, "loss": 1.979044532775879, "step": 9620 }, { "epoch": 1.3747457082693886, "grad_norm": 1.1882662773132324, "learning_rate": 0.00010846941204475125, "loss": 2.0181085586547853, "step": 9630 }, { "epoch": 1.3761733109675578, "grad_norm": 1.0867377519607544, "learning_rate": 0.00010837419661985241, "loss": 2.0765846252441404, "step": 9640 }, { "epoch": 1.3776009136657268, "grad_norm": 1.6985856294631958, "learning_rate": 0.00010827898119495358, "loss": 2.049010467529297, "step": 9650 }, { "epoch": 1.379028516363896, "grad_norm": 1.1894891262054443, "learning_rate": 0.00010818376577005475, "loss": 2.1837297439575196, "step": 9660 }, { "epoch": 1.380456119062065, "grad_norm": 1.1528249979019165, "learning_rate": 0.00010808855034515592, "loss": 2.0664281845092773, "step": 9670 }, { "epoch": 1.381883721760234, "grad_norm": 1.3959161043167114, "learning_rate": 0.00010799333492025709, "loss": 1.9731758117675782, "step": 9680 }, { "epoch": 1.3833113244584032, "grad_norm": 1.4226969480514526, "learning_rate": 0.00010789811949535826, "loss": 2.0336654663085936, "step": 9690 }, { "epoch": 1.3847389271565724, "grad_norm": 1.2641291618347168, "learning_rate": 0.00010780290407045941, "loss": 2.098982048034668, "step": 9700 }, { "epoch": 1.3861665298547414, "grad_norm": 1.4097330570220947, "learning_rate": 0.00010770768864556058, "loss": 1.8858116149902344, "step": 9710 }, { "epoch": 1.3875941325529104, "grad_norm": 1.169032335281372, "learning_rate": 0.00010761247322066175, "loss": 1.991470718383789, "step": 9720 }, { "epoch": 1.3890217352510796, "grad_norm": 1.0888822078704834, "learning_rate": 0.00010751725779576292, "loss": 1.930691909790039, "step": 9730 }, { "epoch": 1.3904493379492489, "grad_norm": 1.1407933235168457, "learning_rate": 0.00010742204237086409, "loss": 2.0946460723876954, "step": 9740 }, { "epoch": 1.3918769406474178, "grad_norm": 1.2439372539520264, "learning_rate": 0.00010732682694596526, "loss": 2.1727531433105467, "step": 9750 }, { "epoch": 1.3933045433455868, "grad_norm": 1.1404023170471191, "learning_rate": 0.00010723161152106643, "loss": 1.9938436508178712, "step": 9760 }, { "epoch": 1.394732146043756, "grad_norm": 1.0030282735824585, "learning_rate": 0.0001071363960961676, "loss": 1.9916358947753907, "step": 9770 }, { "epoch": 1.396159748741925, "grad_norm": 1.1961002349853516, "learning_rate": 0.00010704118067126877, "loss": 1.9710460662841798, "step": 9780 }, { "epoch": 1.3975873514400943, "grad_norm": 1.0773481130599976, "learning_rate": 0.00010694596524636993, "loss": 2.046512794494629, "step": 9790 }, { "epoch": 1.3990149541382633, "grad_norm": 1.076357364654541, "learning_rate": 0.00010685074982147108, "loss": 2.0571868896484373, "step": 9800 }, { "epoch": 1.4004425568364325, "grad_norm": 1.2955228090286255, "learning_rate": 0.00010675553439657224, "loss": 2.0676462173461916, "step": 9810 }, { "epoch": 1.4018701595346015, "grad_norm": 1.1658575534820557, "learning_rate": 0.00010666031897167341, "loss": 2.066716194152832, "step": 9820 }, { "epoch": 1.4032977622327707, "grad_norm": 1.3137447834014893, "learning_rate": 0.00010656510354677458, "loss": 1.8502220153808593, "step": 9830 }, { "epoch": 1.4047253649309397, "grad_norm": 1.41456937789917, "learning_rate": 0.00010646988812187575, "loss": 2.07702522277832, "step": 9840 }, { "epoch": 1.406152967629109, "grad_norm": 1.274549126625061, "learning_rate": 0.00010637467269697692, "loss": 1.8976211547851562, "step": 9850 }, { "epoch": 1.407580570327278, "grad_norm": 1.1110727787017822, "learning_rate": 0.00010627945727207809, "loss": 2.0447467803955077, "step": 9860 }, { "epoch": 1.409008173025447, "grad_norm": 1.099206805229187, "learning_rate": 0.00010618424184717926, "loss": 1.9584508895874024, "step": 9870 }, { "epoch": 1.4104357757236161, "grad_norm": 1.1572459936141968, "learning_rate": 0.00010608902642228043, "loss": 2.0994762420654296, "step": 9880 }, { "epoch": 1.4118633784217853, "grad_norm": 1.3461354970932007, "learning_rate": 0.00010599381099738157, "loss": 1.9647134780883788, "step": 9890 }, { "epoch": 1.4132909811199543, "grad_norm": 1.4512161016464233, "learning_rate": 0.00010589859557248274, "loss": 1.9242181777954102, "step": 9900 }, { "epoch": 1.4147185838181233, "grad_norm": 1.2262014150619507, "learning_rate": 0.0001058033801475839, "loss": 2.0151771545410155, "step": 9910 }, { "epoch": 1.4161461865162925, "grad_norm": 1.3285112380981445, "learning_rate": 0.00010570816472268508, "loss": 1.9909095764160156, "step": 9920 }, { "epoch": 1.4175737892144615, "grad_norm": 1.2019551992416382, "learning_rate": 0.00010561294929778624, "loss": 2.00119571685791, "step": 9930 }, { "epoch": 1.4190013919126307, "grad_norm": 1.1213823556900024, "learning_rate": 0.00010551773387288741, "loss": 2.0824777603149416, "step": 9940 }, { "epoch": 1.4204289946107997, "grad_norm": 1.0179306268692017, "learning_rate": 0.00010542251844798858, "loss": 2.001264190673828, "step": 9950 }, { "epoch": 1.421856597308969, "grad_norm": 1.232782244682312, "learning_rate": 0.00010532730302308975, "loss": 2.004043960571289, "step": 9960 }, { "epoch": 1.423284200007138, "grad_norm": 1.1399147510528564, "learning_rate": 0.00010523208759819092, "loss": 2.022570991516113, "step": 9970 }, { "epoch": 1.4247118027053072, "grad_norm": 1.6599464416503906, "learning_rate": 0.00010513687217329209, "loss": 2.098861503601074, "step": 9980 }, { "epoch": 1.4261394054034762, "grad_norm": 1.1139695644378662, "learning_rate": 0.00010504165674839323, "loss": 2.074664306640625, "step": 9990 }, { "epoch": 1.4275670081016454, "grad_norm": 1.0930999517440796, "learning_rate": 0.0001049464413234944, "loss": 2.1210731506347655, "step": 10000 }, { "epoch": 1.4289946107998144, "grad_norm": 1.4787710905075073, "learning_rate": 0.00010485122589859557, "loss": 2.126077079772949, "step": 10010 }, { "epoch": 1.4304222134979834, "grad_norm": 1.140985131263733, "learning_rate": 0.00010475601047369674, "loss": 2.1588287353515625, "step": 10020 }, { "epoch": 1.4318498161961526, "grad_norm": 1.355772614479065, "learning_rate": 0.0001046607950487979, "loss": 1.9706565856933593, "step": 10030 }, { "epoch": 1.4332774188943218, "grad_norm": 1.163930058479309, "learning_rate": 0.00010456557962389907, "loss": 2.016745185852051, "step": 10040 }, { "epoch": 1.4347050215924908, "grad_norm": 1.1401941776275635, "learning_rate": 0.00010447036419900024, "loss": 2.2144346237182617, "step": 10050 }, { "epoch": 1.4361326242906598, "grad_norm": 1.2282142639160156, "learning_rate": 0.00010437514877410141, "loss": 2.0161321640014647, "step": 10060 }, { "epoch": 1.437560226988829, "grad_norm": 1.33187735080719, "learning_rate": 0.00010427993334920258, "loss": 2.0925859451293944, "step": 10070 }, { "epoch": 1.4389878296869982, "grad_norm": 1.4733668565750122, "learning_rate": 0.00010418471792430374, "loss": 2.0696378707885743, "step": 10080 }, { "epoch": 1.4404154323851672, "grad_norm": 1.2202059030532837, "learning_rate": 0.0001040895024994049, "loss": 1.9806758880615234, "step": 10090 }, { "epoch": 1.4418430350833362, "grad_norm": 1.2413325309753418, "learning_rate": 0.00010399428707450607, "loss": 2.0102920532226562, "step": 10100 }, { "epoch": 1.4432706377815054, "grad_norm": 1.1768853664398193, "learning_rate": 0.00010389907164960724, "loss": 2.0094955444335936, "step": 10110 }, { "epoch": 1.4446982404796744, "grad_norm": 1.3921916484832764, "learning_rate": 0.00010380385622470841, "loss": 2.0164260864257812, "step": 10120 }, { "epoch": 1.4461258431778437, "grad_norm": 2.263312578201294, "learning_rate": 0.00010370864079980958, "loss": 2.032268524169922, "step": 10130 }, { "epoch": 1.4475534458760126, "grad_norm": 1.0994901657104492, "learning_rate": 0.00010361342537491075, "loss": 2.058778762817383, "step": 10140 }, { "epoch": 1.4489810485741819, "grad_norm": 1.6802161931991577, "learning_rate": 0.00010351820995001192, "loss": 2.117998504638672, "step": 10150 }, { "epoch": 1.4504086512723509, "grad_norm": 1.2260311841964722, "learning_rate": 0.00010342299452511309, "loss": 1.9677291870117188, "step": 10160 }, { "epoch": 1.45183625397052, "grad_norm": 0.9355341196060181, "learning_rate": 0.00010332777910021426, "loss": 1.9641590118408203, "step": 10170 }, { "epoch": 1.453263856668689, "grad_norm": 1.1145073175430298, "learning_rate": 0.0001032325636753154, "loss": 2.0117753982543944, "step": 10180 }, { "epoch": 1.4546914593668583, "grad_norm": 1.0050599575042725, "learning_rate": 0.00010313734825041657, "loss": 1.9660293579101562, "step": 10190 }, { "epoch": 1.4561190620650273, "grad_norm": 1.504306674003601, "learning_rate": 0.00010304213282551774, "loss": 2.1012725830078125, "step": 10200 }, { "epoch": 1.4575466647631963, "grad_norm": 1.1196154356002808, "learning_rate": 0.0001029469174006189, "loss": 1.9042285919189452, "step": 10210 }, { "epoch": 1.4589742674613655, "grad_norm": 0.9960939288139343, "learning_rate": 0.00010285170197572007, "loss": 2.0721420288085937, "step": 10220 }, { "epoch": 1.4604018701595347, "grad_norm": 1.1916046142578125, "learning_rate": 0.00010275648655082124, "loss": 2.0403793334960936, "step": 10230 }, { "epoch": 1.4618294728577037, "grad_norm": 1.5170689821243286, "learning_rate": 0.00010266127112592241, "loss": 2.0422672271728515, "step": 10240 }, { "epoch": 1.4632570755558727, "grad_norm": 1.2374523878097534, "learning_rate": 0.00010256605570102358, "loss": 2.0278127670288084, "step": 10250 }, { "epoch": 1.464684678254042, "grad_norm": 1.3379497528076172, "learning_rate": 0.00010247084027612475, "loss": 1.9450105667114257, "step": 10260 }, { "epoch": 1.4661122809522111, "grad_norm": 1.2020896673202515, "learning_rate": 0.00010237562485122589, "loss": 2.0296459197998047, "step": 10270 }, { "epoch": 1.4675398836503801, "grad_norm": 0.9677518606185913, "learning_rate": 0.00010228040942632706, "loss": 2.058147430419922, "step": 10280 }, { "epoch": 1.4689674863485491, "grad_norm": 1.404489278793335, "learning_rate": 0.00010218519400142823, "loss": 2.0642822265625, "step": 10290 }, { "epoch": 1.4703950890467183, "grad_norm": 1.2166563272476196, "learning_rate": 0.0001020899785765294, "loss": 2.0706502914428713, "step": 10300 }, { "epoch": 1.4718226917448873, "grad_norm": 1.3286925554275513, "learning_rate": 0.00010199476315163057, "loss": 2.1312896728515627, "step": 10310 }, { "epoch": 1.4732502944430566, "grad_norm": 1.152422547340393, "learning_rate": 0.00010189954772673174, "loss": 1.9044504165649414, "step": 10320 }, { "epoch": 1.4746778971412255, "grad_norm": 1.1765427589416504, "learning_rate": 0.0001018043323018329, "loss": 1.9554956436157227, "step": 10330 }, { "epoch": 1.4761054998393948, "grad_norm": 1.0927895307540894, "learning_rate": 0.00010170911687693407, "loss": 2.100107765197754, "step": 10340 }, { "epoch": 1.4775331025375638, "grad_norm": 1.1280709505081177, "learning_rate": 0.00010161390145203524, "loss": 2.0164520263671877, "step": 10350 }, { "epoch": 1.478960705235733, "grad_norm": 1.1281557083129883, "learning_rate": 0.00010151868602713641, "loss": 1.9195009231567384, "step": 10360 }, { "epoch": 1.480388307933902, "grad_norm": 0.9746514558792114, "learning_rate": 0.00010142347060223755, "loss": 1.9577857971191406, "step": 10370 }, { "epoch": 1.4818159106320712, "grad_norm": 1.2915472984313965, "learning_rate": 0.00010132825517733872, "loss": 2.0548025131225587, "step": 10380 }, { "epoch": 1.4832435133302402, "grad_norm": 1.120035171508789, "learning_rate": 0.00010123303975243989, "loss": 2.090138244628906, "step": 10390 }, { "epoch": 1.4846711160284092, "grad_norm": 1.1892149448394775, "learning_rate": 0.00010113782432754106, "loss": 2.015086364746094, "step": 10400 }, { "epoch": 1.4860987187265784, "grad_norm": 1.0476832389831543, "learning_rate": 0.00010104260890264223, "loss": 2.1484086990356444, "step": 10410 }, { "epoch": 1.4875263214247476, "grad_norm": 1.2735358476638794, "learning_rate": 0.0001009473934777434, "loss": 2.0922311782836913, "step": 10420 }, { "epoch": 1.4889539241229166, "grad_norm": 1.2540321350097656, "learning_rate": 0.00010085217805284457, "loss": 1.9099697113037108, "step": 10430 }, { "epoch": 1.4903815268210856, "grad_norm": 1.1296672821044922, "learning_rate": 0.00010075696262794573, "loss": 1.9689365386962892, "step": 10440 }, { "epoch": 1.4918091295192548, "grad_norm": 0.9544299840927124, "learning_rate": 0.0001006617472030469, "loss": 2.0149946212768555, "step": 10450 }, { "epoch": 1.4932367322174238, "grad_norm": 1.6253503561019897, "learning_rate": 0.00010056653177814806, "loss": 1.990932083129883, "step": 10460 }, { "epoch": 1.494664334915593, "grad_norm": 1.3218134641647339, "learning_rate": 0.00010047131635324923, "loss": 1.925063133239746, "step": 10470 }, { "epoch": 1.496091937613762, "grad_norm": 1.5663117170333862, "learning_rate": 0.0001003761009283504, "loss": 1.9143461227416991, "step": 10480 }, { "epoch": 1.4975195403119312, "grad_norm": 1.007870078086853, "learning_rate": 0.00010028088550345157, "loss": 1.9640775680541993, "step": 10490 }, { "epoch": 1.4989471430101002, "grad_norm": 1.3052470684051514, "learning_rate": 0.00010018567007855273, "loss": 1.9703134536743163, "step": 10500 }, { "epoch": 1.5003747457082692, "grad_norm": 1.1208614110946655, "learning_rate": 0.0001000904546536539, "loss": 1.9666677474975587, "step": 10510 }, { "epoch": 1.5018023484064384, "grad_norm": 1.112850308418274, "learning_rate": 9.999523922875507e-05, "loss": 2.074142646789551, "step": 10520 }, { "epoch": 1.5032299511046077, "grad_norm": 1.3419452905654907, "learning_rate": 9.990002380385623e-05, "loss": 2.0618656158447264, "step": 10530 }, { "epoch": 1.5046575538027767, "grad_norm": 1.2684563398361206, "learning_rate": 9.98048083789574e-05, "loss": 1.97666015625, "step": 10540 }, { "epoch": 1.5060851565009457, "grad_norm": 1.1568127870559692, "learning_rate": 9.970959295405857e-05, "loss": 2.0744386672973634, "step": 10550 }, { "epoch": 1.5075127591991149, "grad_norm": 1.8241394758224487, "learning_rate": 9.961437752915973e-05, "loss": 2.058271789550781, "step": 10560 }, { "epoch": 1.508940361897284, "grad_norm": 1.099176287651062, "learning_rate": 9.95191621042609e-05, "loss": 2.12048397064209, "step": 10570 }, { "epoch": 1.510367964595453, "grad_norm": 1.1138938665390015, "learning_rate": 9.942394667936206e-05, "loss": 2.1824529647827147, "step": 10580 }, { "epoch": 1.511795567293622, "grad_norm": 1.2076081037521362, "learning_rate": 9.932873125446323e-05, "loss": 1.9712831497192382, "step": 10590 }, { "epoch": 1.5132231699917913, "grad_norm": 1.423424482345581, "learning_rate": 9.92335158295644e-05, "loss": 1.9312576293945312, "step": 10600 }, { "epoch": 1.5146507726899605, "grad_norm": 1.0981698036193848, "learning_rate": 9.913830040466556e-05, "loss": 2.0087039947509764, "step": 10610 }, { "epoch": 1.5160783753881295, "grad_norm": 1.5510938167572021, "learning_rate": 9.904308497976673e-05, "loss": 2.1474151611328125, "step": 10620 }, { "epoch": 1.5175059780862985, "grad_norm": 1.2088773250579834, "learning_rate": 9.894786955486789e-05, "loss": 2.072478103637695, "step": 10630 }, { "epoch": 1.5189335807844677, "grad_norm": 1.0845247507095337, "learning_rate": 9.885265412996906e-05, "loss": 2.015526008605957, "step": 10640 }, { "epoch": 1.520361183482637, "grad_norm": 1.2015031576156616, "learning_rate": 9.875743870507023e-05, "loss": 2.1070165634155273, "step": 10650 }, { "epoch": 1.521788786180806, "grad_norm": 1.062964677810669, "learning_rate": 9.86622232801714e-05, "loss": 2.0601083755493166, "step": 10660 }, { "epoch": 1.523216388878975, "grad_norm": 1.1827353239059448, "learning_rate": 9.856700785527256e-05, "loss": 2.230342674255371, "step": 10670 }, { "epoch": 1.5246439915771441, "grad_norm": 1.2147736549377441, "learning_rate": 9.847179243037372e-05, "loss": 2.083346366882324, "step": 10680 }, { "epoch": 1.5260715942753131, "grad_norm": 1.0761597156524658, "learning_rate": 9.837657700547489e-05, "loss": 2.1192899703979493, "step": 10690 }, { "epoch": 1.5274991969734821, "grad_norm": 1.2726013660430908, "learning_rate": 9.828136158057606e-05, "loss": 2.1517145156860353, "step": 10700 }, { "epoch": 1.5289267996716513, "grad_norm": 1.9693447351455688, "learning_rate": 9.818614615567723e-05, "loss": 1.97952880859375, "step": 10710 }, { "epoch": 1.5303544023698206, "grad_norm": 1.3304651975631714, "learning_rate": 9.809093073077838e-05, "loss": 1.9785858154296876, "step": 10720 }, { "epoch": 1.5317820050679896, "grad_norm": 1.2396725416183472, "learning_rate": 9.799571530587955e-05, "loss": 2.1346342086791994, "step": 10730 }, { "epoch": 1.5332096077661586, "grad_norm": 1.413583517074585, "learning_rate": 9.790049988098072e-05, "loss": 1.9904579162597655, "step": 10740 }, { "epoch": 1.5346372104643278, "grad_norm": 1.2050212621688843, "learning_rate": 9.780528445608189e-05, "loss": 2.045025444030762, "step": 10750 }, { "epoch": 1.536064813162497, "grad_norm": 1.3423675298690796, "learning_rate": 9.771006903118306e-05, "loss": 1.9239255905151367, "step": 10760 }, { "epoch": 1.537492415860666, "grad_norm": 1.2138278484344482, "learning_rate": 9.761485360628421e-05, "loss": 2.0515163421630858, "step": 10770 }, { "epoch": 1.538920018558835, "grad_norm": 1.0984469652175903, "learning_rate": 9.751963818138538e-05, "loss": 1.9942876815795898, "step": 10780 }, { "epoch": 1.5403476212570042, "grad_norm": 1.2791671752929688, "learning_rate": 9.742442275648655e-05, "loss": 2.0003101348876955, "step": 10790 }, { "epoch": 1.5417752239551734, "grad_norm": 1.0745598077774048, "learning_rate": 9.732920733158772e-05, "loss": 1.9592748641967774, "step": 10800 }, { "epoch": 1.5432028266533424, "grad_norm": 1.1889386177062988, "learning_rate": 9.723399190668889e-05, "loss": 1.9912334442138673, "step": 10810 }, { "epoch": 1.5446304293515114, "grad_norm": 1.2227346897125244, "learning_rate": 9.713877648179006e-05, "loss": 2.04000186920166, "step": 10820 }, { "epoch": 1.5460580320496806, "grad_norm": 1.1555755138397217, "learning_rate": 9.704356105689123e-05, "loss": 2.0359256744384764, "step": 10830 }, { "epoch": 1.5474856347478498, "grad_norm": 1.141852617263794, "learning_rate": 9.694834563199238e-05, "loss": 2.0757726669311523, "step": 10840 }, { "epoch": 1.5489132374460186, "grad_norm": 1.1708478927612305, "learning_rate": 9.685313020709355e-05, "loss": 2.077558708190918, "step": 10850 }, { "epoch": 1.5503408401441878, "grad_norm": 1.3248668909072876, "learning_rate": 9.675791478219472e-05, "loss": 2.1411231994628905, "step": 10860 }, { "epoch": 1.551768442842357, "grad_norm": 1.1437745094299316, "learning_rate": 9.666269935729589e-05, "loss": 2.2358665466308594, "step": 10870 }, { "epoch": 1.553196045540526, "grad_norm": 1.4474693536758423, "learning_rate": 9.656748393239706e-05, "loss": 2.1683933258056642, "step": 10880 }, { "epoch": 1.554623648238695, "grad_norm": 1.4569262266159058, "learning_rate": 9.647226850749823e-05, "loss": 2.0875593185424806, "step": 10890 }, { "epoch": 1.5560512509368642, "grad_norm": 1.2359139919281006, "learning_rate": 9.63770530825994e-05, "loss": 2.0447929382324217, "step": 10900 }, { "epoch": 1.5574788536350335, "grad_norm": 1.1910909414291382, "learning_rate": 9.628183765770056e-05, "loss": 1.9503170013427735, "step": 10910 }, { "epoch": 1.5589064563332025, "grad_norm": 1.267309546470642, "learning_rate": 9.618662223280172e-05, "loss": 1.9226907730102538, "step": 10920 }, { "epoch": 1.5603340590313715, "grad_norm": 1.0570237636566162, "learning_rate": 9.609140680790289e-05, "loss": 2.0675920486450194, "step": 10930 }, { "epoch": 1.5617616617295407, "grad_norm": 1.0798622369766235, "learning_rate": 9.599619138300406e-05, "loss": 2.1086353302001952, "step": 10940 }, { "epoch": 1.56318926442771, "grad_norm": 0.9747142791748047, "learning_rate": 9.590097595810523e-05, "loss": 2.0529823303222656, "step": 10950 }, { "epoch": 1.5646168671258789, "grad_norm": 1.1378403902053833, "learning_rate": 9.580576053320638e-05, "loss": 2.1953744888305664, "step": 10960 }, { "epoch": 1.5660444698240479, "grad_norm": 1.4991871118545532, "learning_rate": 9.571054510830755e-05, "loss": 1.94462890625, "step": 10970 }, { "epoch": 1.567472072522217, "grad_norm": 1.005947232246399, "learning_rate": 9.561532968340872e-05, "loss": 2.1250656127929686, "step": 10980 }, { "epoch": 1.5688996752203863, "grad_norm": 1.265784502029419, "learning_rate": 9.552011425850989e-05, "loss": 1.9723976135253907, "step": 10990 }, { "epoch": 1.5703272779185553, "grad_norm": 1.7473640441894531, "learning_rate": 9.542489883361106e-05, "loss": 2.0439002990722654, "step": 11000 }, { "epoch": 1.5717548806167243, "grad_norm": 1.376124620437622, "learning_rate": 9.532968340871221e-05, "loss": 2.1368425369262694, "step": 11010 }, { "epoch": 1.5731824833148935, "grad_norm": 1.0686304569244385, "learning_rate": 9.523446798381338e-05, "loss": 2.0076366424560548, "step": 11020 }, { "epoch": 1.5746100860130625, "grad_norm": 1.4122185707092285, "learning_rate": 9.513925255891455e-05, "loss": 2.095813751220703, "step": 11030 }, { "epoch": 1.5760376887112315, "grad_norm": 1.26949143409729, "learning_rate": 9.504403713401572e-05, "loss": 2.1850494384765624, "step": 11040 }, { "epoch": 1.5774652914094007, "grad_norm": 1.120142936706543, "learning_rate": 9.494882170911689e-05, "loss": 2.0661577224731444, "step": 11050 }, { "epoch": 1.57889289410757, "grad_norm": 1.2405672073364258, "learning_rate": 9.485360628421804e-05, "loss": 1.9966974258422852, "step": 11060 }, { "epoch": 1.580320496805739, "grad_norm": 1.159519076347351, "learning_rate": 9.475839085931921e-05, "loss": 1.9302923202514648, "step": 11070 }, { "epoch": 1.581748099503908, "grad_norm": 1.2723621129989624, "learning_rate": 9.466317543442038e-05, "loss": 1.837651252746582, "step": 11080 }, { "epoch": 1.5831757022020772, "grad_norm": 1.354519009590149, "learning_rate": 9.456796000952155e-05, "loss": 2.001264953613281, "step": 11090 }, { "epoch": 1.5846033049002464, "grad_norm": 1.2013628482818604, "learning_rate": 9.447274458462272e-05, "loss": 2.0066423416137695, "step": 11100 }, { "epoch": 1.5860309075984154, "grad_norm": 1.159398078918457, "learning_rate": 9.437752915972387e-05, "loss": 2.0087814331054688, "step": 11110 }, { "epoch": 1.5874585102965844, "grad_norm": 1.2587201595306396, "learning_rate": 9.428231373482504e-05, "loss": 1.9774051666259767, "step": 11120 }, { "epoch": 1.5888861129947536, "grad_norm": 1.2386575937271118, "learning_rate": 9.418709830992621e-05, "loss": 2.03675537109375, "step": 11130 }, { "epoch": 1.5903137156929228, "grad_norm": 1.1446408033370972, "learning_rate": 9.409188288502738e-05, "loss": 2.0270097732543944, "step": 11140 }, { "epoch": 1.5917413183910918, "grad_norm": 1.090232491493225, "learning_rate": 9.399666746012853e-05, "loss": 2.1192419052124025, "step": 11150 }, { "epoch": 1.5931689210892608, "grad_norm": 1.137207269668579, "learning_rate": 9.39014520352297e-05, "loss": 2.041428565979004, "step": 11160 }, { "epoch": 1.59459652378743, "grad_norm": 1.2093437910079956, "learning_rate": 9.380623661033087e-05, "loss": 1.9331945419311523, "step": 11170 }, { "epoch": 1.5960241264855992, "grad_norm": 1.2326594591140747, "learning_rate": 9.371102118543204e-05, "loss": 2.212518882751465, "step": 11180 }, { "epoch": 1.5974517291837682, "grad_norm": 1.280666708946228, "learning_rate": 9.361580576053321e-05, "loss": 2.067685127258301, "step": 11190 }, { "epoch": 1.5988793318819372, "grad_norm": 1.3486379384994507, "learning_rate": 9.352059033563438e-05, "loss": 1.962606430053711, "step": 11200 }, { "epoch": 1.6003069345801064, "grad_norm": 1.0335818529129028, "learning_rate": 9.342537491073555e-05, "loss": 1.9397489547729492, "step": 11210 }, { "epoch": 1.6017345372782754, "grad_norm": 0.9871345162391663, "learning_rate": 9.33301594858367e-05, "loss": 2.009673309326172, "step": 11220 }, { "epoch": 1.6031621399764444, "grad_norm": 1.021042823791504, "learning_rate": 9.323494406093787e-05, "loss": 1.9617034912109375, "step": 11230 }, { "epoch": 1.6045897426746136, "grad_norm": 0.9557853937149048, "learning_rate": 9.313972863603904e-05, "loss": 2.0664203643798826, "step": 11240 }, { "epoch": 1.6060173453727828, "grad_norm": 1.062493920326233, "learning_rate": 9.304451321114021e-05, "loss": 2.0338150024414063, "step": 11250 }, { "epoch": 1.6074449480709518, "grad_norm": 1.8569424152374268, "learning_rate": 9.294929778624138e-05, "loss": 2.18345947265625, "step": 11260 }, { "epoch": 1.6088725507691208, "grad_norm": 1.6317625045776367, "learning_rate": 9.285408236134255e-05, "loss": 2.12985897064209, "step": 11270 }, { "epoch": 1.61030015346729, "grad_norm": 1.9217251539230347, "learning_rate": 9.275886693644372e-05, "loss": 2.029741668701172, "step": 11280 }, { "epoch": 1.6117277561654593, "grad_norm": 1.2121089696884155, "learning_rate": 9.266365151154489e-05, "loss": 2.0701601028442385, "step": 11290 }, { "epoch": 1.6131553588636283, "grad_norm": 1.442859172821045, "learning_rate": 9.256843608664604e-05, "loss": 2.195418930053711, "step": 11300 }, { "epoch": 1.6145829615617973, "grad_norm": 1.0287760496139526, "learning_rate": 9.247322066174721e-05, "loss": 2.049941062927246, "step": 11310 }, { "epoch": 1.6160105642599665, "grad_norm": 1.0198233127593994, "learning_rate": 9.237800523684838e-05, "loss": 2.047838592529297, "step": 11320 }, { "epoch": 1.6174381669581357, "grad_norm": 1.780766487121582, "learning_rate": 9.228278981194955e-05, "loss": 2.090102767944336, "step": 11330 }, { "epoch": 1.6188657696563047, "grad_norm": 1.2811323404312134, "learning_rate": 9.21875743870507e-05, "loss": 2.082899475097656, "step": 11340 }, { "epoch": 1.6202933723544737, "grad_norm": 1.0062812566757202, "learning_rate": 9.209235896215187e-05, "loss": 2.1465679168701173, "step": 11350 }, { "epoch": 1.621720975052643, "grad_norm": 1.0987058877944946, "learning_rate": 9.199714353725304e-05, "loss": 2.0366447448730467, "step": 11360 }, { "epoch": 1.623148577750812, "grad_norm": 1.1439515352249146, "learning_rate": 9.190192811235421e-05, "loss": 2.067383575439453, "step": 11370 }, { "epoch": 1.6245761804489809, "grad_norm": 1.1097221374511719, "learning_rate": 9.180671268745538e-05, "loss": 1.9954216003417968, "step": 11380 }, { "epoch": 1.62600378314715, "grad_norm": 1.1178920269012451, "learning_rate": 9.171149726255653e-05, "loss": 2.0267349243164063, "step": 11390 }, { "epoch": 1.6274313858453193, "grad_norm": 1.3011680841445923, "learning_rate": 9.16162818376577e-05, "loss": 1.8894380569458007, "step": 11400 }, { "epoch": 1.6288589885434883, "grad_norm": 1.2558344602584839, "learning_rate": 9.152106641275887e-05, "loss": 2.0158664703369142, "step": 11410 }, { "epoch": 1.6302865912416573, "grad_norm": 1.4524527788162231, "learning_rate": 9.142585098786004e-05, "loss": 1.9456119537353516, "step": 11420 }, { "epoch": 1.6317141939398265, "grad_norm": 1.0922093391418457, "learning_rate": 9.133063556296121e-05, "loss": 2.031543731689453, "step": 11430 }, { "epoch": 1.6331417966379957, "grad_norm": 1.1085158586502075, "learning_rate": 9.123542013806236e-05, "loss": 1.964078712463379, "step": 11440 }, { "epoch": 1.6345693993361647, "grad_norm": 1.0492099523544312, "learning_rate": 9.114020471316353e-05, "loss": 2.118969535827637, "step": 11450 }, { "epoch": 1.6359970020343337, "grad_norm": 1.179565668106079, "learning_rate": 9.10449892882647e-05, "loss": 2.040058898925781, "step": 11460 }, { "epoch": 1.637424604732503, "grad_norm": 2.1389811038970947, "learning_rate": 9.094977386336587e-05, "loss": 1.9964542388916016, "step": 11470 }, { "epoch": 1.6388522074306722, "grad_norm": 1.0620242357254028, "learning_rate": 9.085455843846704e-05, "loss": 2.01263427734375, "step": 11480 }, { "epoch": 1.6402798101288412, "grad_norm": 1.134291410446167, "learning_rate": 9.07593430135682e-05, "loss": 1.9857776641845704, "step": 11490 }, { "epoch": 1.6417074128270102, "grad_norm": 1.1080206632614136, "learning_rate": 9.066412758866936e-05, "loss": 1.8630657196044922, "step": 11500 }, { "epoch": 1.6431350155251794, "grad_norm": 0.9551442861557007, "learning_rate": 9.056891216377053e-05, "loss": 2.1820304870605467, "step": 11510 }, { "epoch": 1.6445626182233486, "grad_norm": 1.1630139350891113, "learning_rate": 9.04736967388717e-05, "loss": 1.8536785125732422, "step": 11520 }, { "epoch": 1.6459902209215176, "grad_norm": 1.0044795274734497, "learning_rate": 9.037848131397286e-05, "loss": 1.9964088439941405, "step": 11530 }, { "epoch": 1.6474178236196866, "grad_norm": 1.1769623756408691, "learning_rate": 9.028326588907403e-05, "loss": 2.0912023544311524, "step": 11540 }, { "epoch": 1.6488454263178558, "grad_norm": 1.0666323900222778, "learning_rate": 9.01880504641752e-05, "loss": 2.163995361328125, "step": 11550 }, { "epoch": 1.6502730290160248, "grad_norm": 1.1934126615524292, "learning_rate": 9.009283503927636e-05, "loss": 2.0924661636352537, "step": 11560 }, { "epoch": 1.6517006317141938, "grad_norm": 1.5662250518798828, "learning_rate": 8.999761961437753e-05, "loss": 2.0510982513427733, "step": 11570 }, { "epoch": 1.653128234412363, "grad_norm": 1.0780766010284424, "learning_rate": 8.99024041894787e-05, "loss": 1.8760522842407226, "step": 11580 }, { "epoch": 1.6545558371105322, "grad_norm": 1.0898934602737427, "learning_rate": 8.980718876457987e-05, "loss": 2.1295299530029297, "step": 11590 }, { "epoch": 1.6559834398087012, "grad_norm": 1.1452609300613403, "learning_rate": 8.971197333968104e-05, "loss": 1.8752655029296874, "step": 11600 }, { "epoch": 1.6574110425068702, "grad_norm": 1.0586178302764893, "learning_rate": 8.96167579147822e-05, "loss": 2.0695917129516603, "step": 11610 }, { "epoch": 1.6588386452050394, "grad_norm": 1.0551941394805908, "learning_rate": 8.952154248988336e-05, "loss": 2.1712038040161135, "step": 11620 }, { "epoch": 1.6602662479032086, "grad_norm": 1.3494571447372437, "learning_rate": 8.942632706498453e-05, "loss": 2.035284423828125, "step": 11630 }, { "epoch": 1.6616938506013776, "grad_norm": 1.2494606971740723, "learning_rate": 8.93311116400857e-05, "loss": 2.031180763244629, "step": 11640 }, { "epoch": 1.6631214532995466, "grad_norm": 1.16592276096344, "learning_rate": 8.923589621518687e-05, "loss": 1.960007095336914, "step": 11650 }, { "epoch": 1.6645490559977159, "grad_norm": 1.1043802499771118, "learning_rate": 8.914068079028804e-05, "loss": 1.9786636352539062, "step": 11660 }, { "epoch": 1.665976658695885, "grad_norm": 1.4241987466812134, "learning_rate": 8.904546536538921e-05, "loss": 2.036239433288574, "step": 11670 }, { "epoch": 1.667404261394054, "grad_norm": 1.08253014087677, "learning_rate": 8.895024994049036e-05, "loss": 2.1130958557128907, "step": 11680 }, { "epoch": 1.668831864092223, "grad_norm": 1.4800434112548828, "learning_rate": 8.885503451559153e-05, "loss": 2.239775276184082, "step": 11690 }, { "epoch": 1.6702594667903923, "grad_norm": 1.505175232887268, "learning_rate": 8.87598190906927e-05, "loss": 1.9916955947875976, "step": 11700 }, { "epoch": 1.6716870694885615, "grad_norm": 1.2340087890625, "learning_rate": 8.866460366579387e-05, "loss": 2.1178766250610352, "step": 11710 }, { "epoch": 1.6731146721867303, "grad_norm": 1.8517627716064453, "learning_rate": 8.856938824089502e-05, "loss": 2.129575729370117, "step": 11720 }, { "epoch": 1.6745422748848995, "grad_norm": 1.1588945388793945, "learning_rate": 8.84741728159962e-05, "loss": 2.1198596954345703, "step": 11730 }, { "epoch": 1.6759698775830687, "grad_norm": 1.0668638944625854, "learning_rate": 8.837895739109736e-05, "loss": 2.026426887512207, "step": 11740 }, { "epoch": 1.6773974802812377, "grad_norm": 1.118070363998413, "learning_rate": 8.828374196619853e-05, "loss": 1.905161476135254, "step": 11750 }, { "epoch": 1.6788250829794067, "grad_norm": 1.1691018342971802, "learning_rate": 8.81885265412997e-05, "loss": 1.992619514465332, "step": 11760 }, { "epoch": 1.680252685677576, "grad_norm": 1.3565104007720947, "learning_rate": 8.809331111640086e-05, "loss": 2.023451232910156, "step": 11770 }, { "epoch": 1.6816802883757451, "grad_norm": 1.0296746492385864, "learning_rate": 8.799809569150202e-05, "loss": 2.136622428894043, "step": 11780 }, { "epoch": 1.6831078910739141, "grad_norm": 1.0471142530441284, "learning_rate": 8.79028802666032e-05, "loss": 2.032058906555176, "step": 11790 }, { "epoch": 1.6845354937720831, "grad_norm": 2.0366618633270264, "learning_rate": 8.780766484170436e-05, "loss": 2.0025094985961913, "step": 11800 }, { "epoch": 1.6859630964702523, "grad_norm": 1.314437985420227, "learning_rate": 8.771244941680553e-05, "loss": 1.9852645874023438, "step": 11810 }, { "epoch": 1.6873906991684215, "grad_norm": 1.0761778354644775, "learning_rate": 8.761723399190669e-05, "loss": 1.9645938873291016, "step": 11820 }, { "epoch": 1.6888183018665905, "grad_norm": 1.2254632711410522, "learning_rate": 8.752201856700786e-05, "loss": 2.054538917541504, "step": 11830 }, { "epoch": 1.6902459045647595, "grad_norm": 1.219874620437622, "learning_rate": 8.742680314210902e-05, "loss": 1.9738475799560546, "step": 11840 }, { "epoch": 1.6916735072629288, "grad_norm": 1.0699069499969482, "learning_rate": 8.733158771721019e-05, "loss": 2.127732276916504, "step": 11850 }, { "epoch": 1.693101109961098, "grad_norm": 1.0340874195098877, "learning_rate": 8.723637229231136e-05, "loss": 2.0337158203125, "step": 11860 }, { "epoch": 1.694528712659267, "grad_norm": 1.2147150039672852, "learning_rate": 8.714115686741252e-05, "loss": 2.076980400085449, "step": 11870 }, { "epoch": 1.695956315357436, "grad_norm": 1.2863147258758545, "learning_rate": 8.704594144251369e-05, "loss": 2.0516342163085937, "step": 11880 }, { "epoch": 1.6973839180556052, "grad_norm": 1.258984923362732, "learning_rate": 8.695072601761486e-05, "loss": 1.9719093322753907, "step": 11890 }, { "epoch": 1.6988115207537742, "grad_norm": 1.0693020820617676, "learning_rate": 8.685551059271602e-05, "loss": 1.8373472213745117, "step": 11900 }, { "epoch": 1.7002391234519432, "grad_norm": 1.2779794931411743, "learning_rate": 8.676029516781719e-05, "loss": 2.002557373046875, "step": 11910 }, { "epoch": 1.7016667261501124, "grad_norm": 1.0776644945144653, "learning_rate": 8.666507974291835e-05, "loss": 2.072839546203613, "step": 11920 }, { "epoch": 1.7030943288482816, "grad_norm": 1.0650670528411865, "learning_rate": 8.656986431801952e-05, "loss": 2.0984193801879885, "step": 11930 }, { "epoch": 1.7045219315464506, "grad_norm": 1.165576696395874, "learning_rate": 8.647464889312069e-05, "loss": 2.092478370666504, "step": 11940 }, { "epoch": 1.7059495342446196, "grad_norm": 1.3021267652511597, "learning_rate": 8.637943346822185e-05, "loss": 1.972269058227539, "step": 11950 }, { "epoch": 1.7073771369427888, "grad_norm": 1.0867663621902466, "learning_rate": 8.628421804332302e-05, "loss": 2.080883598327637, "step": 11960 }, { "epoch": 1.708804739640958, "grad_norm": 1.1201988458633423, "learning_rate": 8.618900261842419e-05, "loss": 1.970688819885254, "step": 11970 }, { "epoch": 1.710232342339127, "grad_norm": 1.1707783937454224, "learning_rate": 8.609378719352536e-05, "loss": 1.971019172668457, "step": 11980 }, { "epoch": 1.711659945037296, "grad_norm": 1.3623528480529785, "learning_rate": 8.599857176862653e-05, "loss": 2.0503314971923827, "step": 11990 }, { "epoch": 1.7130875477354652, "grad_norm": 1.3899431228637695, "learning_rate": 8.590335634372769e-05, "loss": 2.0131784439086915, "step": 12000 }, { "epoch": 1.7145151504336344, "grad_norm": 1.1104278564453125, "learning_rate": 8.580814091882885e-05, "loss": 2.1040504455566404, "step": 12010 }, { "epoch": 1.7159427531318034, "grad_norm": 1.133924961090088, "learning_rate": 8.571292549393002e-05, "loss": 2.041822814941406, "step": 12020 }, { "epoch": 1.7173703558299724, "grad_norm": 1.1938362121582031, "learning_rate": 8.561771006903119e-05, "loss": 2.0624362945556642, "step": 12030 }, { "epoch": 1.7187979585281417, "grad_norm": 1.181810736656189, "learning_rate": 8.552249464413236e-05, "loss": 1.9707704544067384, "step": 12040 }, { "epoch": 1.7202255612263109, "grad_norm": 1.2398381233215332, "learning_rate": 8.542727921923353e-05, "loss": 1.8960441589355468, "step": 12050 }, { "epoch": 1.7216531639244799, "grad_norm": 1.074657678604126, "learning_rate": 8.533206379433469e-05, "loss": 1.9548070907592774, "step": 12060 }, { "epoch": 1.7230807666226489, "grad_norm": 1.2914633750915527, "learning_rate": 8.523684836943585e-05, "loss": 1.9369295120239258, "step": 12070 }, { "epoch": 1.724508369320818, "grad_norm": 1.122804880142212, "learning_rate": 8.514163294453702e-05, "loss": 2.0332637786865235, "step": 12080 }, { "epoch": 1.725935972018987, "grad_norm": 1.2431594133377075, "learning_rate": 8.504641751963819e-05, "loss": 2.07200927734375, "step": 12090 }, { "epoch": 1.727363574717156, "grad_norm": 1.1114283800125122, "learning_rate": 8.495120209473936e-05, "loss": 2.0198741912841798, "step": 12100 }, { "epoch": 1.7287911774153253, "grad_norm": 1.1872164011001587, "learning_rate": 8.485598666984052e-05, "loss": 1.8839866638183593, "step": 12110 }, { "epoch": 1.7302187801134945, "grad_norm": 1.483201026916504, "learning_rate": 8.476077124494168e-05, "loss": 2.035213088989258, "step": 12120 }, { "epoch": 1.7316463828116635, "grad_norm": 1.098403811454773, "learning_rate": 8.466555582004285e-05, "loss": 2.0581424713134764, "step": 12130 }, { "epoch": 1.7330739855098325, "grad_norm": 0.9966762661933899, "learning_rate": 8.457034039514402e-05, "loss": 2.13366756439209, "step": 12140 }, { "epoch": 1.7345015882080017, "grad_norm": 1.2264962196350098, "learning_rate": 8.447512497024518e-05, "loss": 1.9749515533447266, "step": 12150 }, { "epoch": 1.735929190906171, "grad_norm": 0.9798251390457153, "learning_rate": 8.437990954534635e-05, "loss": 1.985658836364746, "step": 12160 }, { "epoch": 1.73735679360434, "grad_norm": 1.3194963932037354, "learning_rate": 8.428469412044752e-05, "loss": 2.0228357315063477, "step": 12170 }, { "epoch": 1.738784396302509, "grad_norm": 1.0797955989837646, "learning_rate": 8.418947869554868e-05, "loss": 2.0977373123168945, "step": 12180 }, { "epoch": 1.7402119990006781, "grad_norm": 1.6863555908203125, "learning_rate": 8.409426327064985e-05, "loss": 2.1669921875, "step": 12190 }, { "epoch": 1.7416396016988474, "grad_norm": 1.1710777282714844, "learning_rate": 8.399904784575101e-05, "loss": 2.0780879974365236, "step": 12200 }, { "epoch": 1.7430672043970163, "grad_norm": 1.1094262599945068, "learning_rate": 8.390383242085218e-05, "loss": 1.8861412048339843, "step": 12210 }, { "epoch": 1.7444948070951853, "grad_norm": 1.279801607131958, "learning_rate": 8.380861699595335e-05, "loss": 1.9867172241210938, "step": 12220 }, { "epoch": 1.7459224097933546, "grad_norm": 1.195829153060913, "learning_rate": 8.371340157105452e-05, "loss": 2.0028406143188477, "step": 12230 }, { "epoch": 1.7473500124915238, "grad_norm": 1.2516834735870361, "learning_rate": 8.361818614615568e-05, "loss": 1.9748237609863282, "step": 12240 }, { "epoch": 1.7487776151896925, "grad_norm": 1.193414568901062, "learning_rate": 8.352297072125684e-05, "loss": 2.001264190673828, "step": 12250 }, { "epoch": 1.7502052178878618, "grad_norm": 1.1629263162612915, "learning_rate": 8.342775529635801e-05, "loss": 2.083281135559082, "step": 12260 }, { "epoch": 1.751632820586031, "grad_norm": 1.2907096147537231, "learning_rate": 8.333253987145918e-05, "loss": 1.995428466796875, "step": 12270 }, { "epoch": 1.7530604232842, "grad_norm": 1.245174765586853, "learning_rate": 8.323732444656035e-05, "loss": 2.198429489135742, "step": 12280 }, { "epoch": 1.754488025982369, "grad_norm": 1.0381641387939453, "learning_rate": 8.314210902166152e-05, "loss": 1.9604129791259766, "step": 12290 }, { "epoch": 1.7559156286805382, "grad_norm": 1.1638761758804321, "learning_rate": 8.304689359676267e-05, "loss": 2.0892223358154296, "step": 12300 }, { "epoch": 1.7573432313787074, "grad_norm": 1.2163985967636108, "learning_rate": 8.295167817186384e-05, "loss": 2.0810489654541016, "step": 12310 }, { "epoch": 1.7587708340768764, "grad_norm": 1.2280848026275635, "learning_rate": 8.285646274696501e-05, "loss": 2.053379249572754, "step": 12320 }, { "epoch": 1.7601984367750454, "grad_norm": 1.1925756931304932, "learning_rate": 8.276124732206618e-05, "loss": 2.069741630554199, "step": 12330 }, { "epoch": 1.7616260394732146, "grad_norm": 1.1986944675445557, "learning_rate": 8.266603189716735e-05, "loss": 2.091610336303711, "step": 12340 }, { "epoch": 1.7630536421713838, "grad_norm": 1.0888428688049316, "learning_rate": 8.257081647226851e-05, "loss": 1.9579872131347655, "step": 12350 }, { "epoch": 1.7644812448695528, "grad_norm": 1.2537578344345093, "learning_rate": 8.247560104736968e-05, "loss": 2.0613885879516602, "step": 12360 }, { "epoch": 1.7659088475677218, "grad_norm": 1.1928845643997192, "learning_rate": 8.238038562247085e-05, "loss": 2.095230293273926, "step": 12370 }, { "epoch": 1.767336450265891, "grad_norm": 1.1767767667770386, "learning_rate": 8.228517019757202e-05, "loss": 1.9075620651245118, "step": 12380 }, { "epoch": 1.7687640529640603, "grad_norm": 1.1607625484466553, "learning_rate": 8.218995477267318e-05, "loss": 1.9210027694702148, "step": 12390 }, { "epoch": 1.7701916556622292, "grad_norm": 1.096114158630371, "learning_rate": 8.209473934777435e-05, "loss": 1.9674240112304688, "step": 12400 }, { "epoch": 1.7716192583603982, "grad_norm": 1.044256567955017, "learning_rate": 8.199952392287551e-05, "loss": 2.0622480392456053, "step": 12410 }, { "epoch": 1.7730468610585675, "grad_norm": 1.3171051740646362, "learning_rate": 8.190430849797668e-05, "loss": 2.164032554626465, "step": 12420 }, { "epoch": 1.7744744637567365, "grad_norm": 1.3903594017028809, "learning_rate": 8.180909307307785e-05, "loss": 2.079869842529297, "step": 12430 }, { "epoch": 1.7759020664549054, "grad_norm": 0.9803964495658875, "learning_rate": 8.171387764817901e-05, "loss": 1.982889175415039, "step": 12440 }, { "epoch": 1.7773296691530747, "grad_norm": 1.4067291021347046, "learning_rate": 8.161866222328018e-05, "loss": 2.0877866744995117, "step": 12450 }, { "epoch": 1.7787572718512439, "grad_norm": 1.215567708015442, "learning_rate": 8.152344679838135e-05, "loss": 1.973129653930664, "step": 12460 }, { "epoch": 1.7801848745494129, "grad_norm": 1.1108075380325317, "learning_rate": 8.142823137348251e-05, "loss": 2.048358917236328, "step": 12470 }, { "epoch": 1.7816124772475819, "grad_norm": 1.0958882570266724, "learning_rate": 8.133301594858368e-05, "loss": 2.059286689758301, "step": 12480 }, { "epoch": 1.783040079945751, "grad_norm": 1.1889430284500122, "learning_rate": 8.123780052368484e-05, "loss": 2.063096046447754, "step": 12490 }, { "epoch": 1.7844676826439203, "grad_norm": 1.5182738304138184, "learning_rate": 8.114258509878601e-05, "loss": 1.9856096267700196, "step": 12500 }, { "epoch": 1.7858952853420893, "grad_norm": 1.1368118524551392, "learning_rate": 8.104736967388718e-05, "loss": 1.9539974212646485, "step": 12510 }, { "epoch": 1.7873228880402583, "grad_norm": 1.1098486185073853, "learning_rate": 8.095215424898834e-05, "loss": 1.955406379699707, "step": 12520 }, { "epoch": 1.7887504907384275, "grad_norm": 1.107768177986145, "learning_rate": 8.08569388240895e-05, "loss": 2.01610107421875, "step": 12530 }, { "epoch": 1.7901780934365967, "grad_norm": 1.016388177871704, "learning_rate": 8.076172339919067e-05, "loss": 2.0703092575073243, "step": 12540 }, { "epoch": 1.7916056961347657, "grad_norm": 1.0971436500549316, "learning_rate": 8.066650797429184e-05, "loss": 2.0417387008666994, "step": 12550 }, { "epoch": 1.7930332988329347, "grad_norm": 1.2257986068725586, "learning_rate": 8.057129254939301e-05, "loss": 2.1287025451660155, "step": 12560 }, { "epoch": 1.794460901531104, "grad_norm": 1.2729699611663818, "learning_rate": 8.047607712449418e-05, "loss": 1.9849166870117188, "step": 12570 }, { "epoch": 1.7958885042292732, "grad_norm": 1.1034228801727295, "learning_rate": 8.038086169959533e-05, "loss": 2.0139421463012694, "step": 12580 }, { "epoch": 1.7973161069274421, "grad_norm": 1.0930798053741455, "learning_rate": 8.02856462746965e-05, "loss": 2.223542594909668, "step": 12590 }, { "epoch": 1.7987437096256111, "grad_norm": 1.3495484590530396, "learning_rate": 8.019043084979767e-05, "loss": 2.047318458557129, "step": 12600 }, { "epoch": 1.8001713123237804, "grad_norm": 1.2880040407180786, "learning_rate": 8.009521542489884e-05, "loss": 2.0674213409423827, "step": 12610 }, { "epoch": 1.8015989150219494, "grad_norm": 1.0766103267669678, "learning_rate": 8e-05, "loss": 2.0229583740234376, "step": 12620 }, { "epoch": 1.8030265177201183, "grad_norm": 1.5253698825836182, "learning_rate": 7.990478457510116e-05, "loss": 1.9948074340820312, "step": 12630 }, { "epoch": 1.8044541204182876, "grad_norm": 1.1996920108795166, "learning_rate": 7.980956915020233e-05, "loss": 1.929234504699707, "step": 12640 }, { "epoch": 1.8058817231164568, "grad_norm": 1.2180821895599365, "learning_rate": 7.97143537253035e-05, "loss": 2.0617977142333985, "step": 12650 }, { "epoch": 1.8073093258146258, "grad_norm": 1.2548080682754517, "learning_rate": 7.961913830040467e-05, "loss": 2.0628313064575194, "step": 12660 }, { "epoch": 1.8087369285127948, "grad_norm": 1.2494982481002808, "learning_rate": 7.952392287550584e-05, "loss": 2.058258056640625, "step": 12670 }, { "epoch": 1.810164531210964, "grad_norm": 1.1819909811019897, "learning_rate": 7.942870745060699e-05, "loss": 2.0007421493530275, "step": 12680 }, { "epoch": 1.8115921339091332, "grad_norm": 0.9779897332191467, "learning_rate": 7.933349202570816e-05, "loss": 2.01336727142334, "step": 12690 }, { "epoch": 1.8130197366073022, "grad_norm": 1.065111756324768, "learning_rate": 7.923827660080933e-05, "loss": 1.9989856719970702, "step": 12700 }, { "epoch": 1.8144473393054712, "grad_norm": 1.3039896488189697, "learning_rate": 7.91430611759105e-05, "loss": 2.113441276550293, "step": 12710 }, { "epoch": 1.8158749420036404, "grad_norm": 1.3907358646392822, "learning_rate": 7.904784575101167e-05, "loss": 2.088470458984375, "step": 12720 }, { "epoch": 1.8173025447018096, "grad_norm": 1.1438897848129272, "learning_rate": 7.895263032611284e-05, "loss": 2.116332244873047, "step": 12730 }, { "epoch": 1.8187301473999786, "grad_norm": 1.6068172454833984, "learning_rate": 7.8857414901214e-05, "loss": 2.1666784286499023, "step": 12740 }, { "epoch": 1.8201577500981476, "grad_norm": 1.2010868787765503, "learning_rate": 7.876219947631517e-05, "loss": 1.9318151473999023, "step": 12750 }, { "epoch": 1.8215853527963168, "grad_norm": 1.1283704042434692, "learning_rate": 7.866698405141634e-05, "loss": 2.077130126953125, "step": 12760 }, { "epoch": 1.8230129554944858, "grad_norm": 1.0729111433029175, "learning_rate": 7.85717686265175e-05, "loss": 1.9765430450439454, "step": 12770 }, { "epoch": 1.8244405581926548, "grad_norm": 1.1061432361602783, "learning_rate": 7.847655320161867e-05, "loss": 2.139857292175293, "step": 12780 }, { "epoch": 1.825868160890824, "grad_norm": 1.1388243436813354, "learning_rate": 7.838133777671984e-05, "loss": 2.0259464263916014, "step": 12790 }, { "epoch": 1.8272957635889933, "grad_norm": 1.2130709886550903, "learning_rate": 7.8286122351821e-05, "loss": 2.073659324645996, "step": 12800 }, { "epoch": 1.8287233662871623, "grad_norm": 1.101082682609558, "learning_rate": 7.819090692692217e-05, "loss": 2.058758354187012, "step": 12810 }, { "epoch": 1.8301509689853313, "grad_norm": 1.342145323753357, "learning_rate": 7.809569150202333e-05, "loss": 2.0943870544433594, "step": 12820 }, { "epoch": 1.8315785716835005, "grad_norm": 0.9709015488624573, "learning_rate": 7.80004760771245e-05, "loss": 1.9406627655029296, "step": 12830 }, { "epoch": 1.8330061743816697, "grad_norm": 1.1636347770690918, "learning_rate": 7.790526065222567e-05, "loss": 2.078135299682617, "step": 12840 }, { "epoch": 1.8344337770798387, "grad_norm": 1.3326045274734497, "learning_rate": 7.781004522732684e-05, "loss": 1.9961933135986327, "step": 12850 }, { "epoch": 1.8358613797780077, "grad_norm": 1.15592360496521, "learning_rate": 7.7714829802428e-05, "loss": 1.966868782043457, "step": 12860 }, { "epoch": 1.837288982476177, "grad_norm": 1.2017027139663696, "learning_rate": 7.761961437752916e-05, "loss": 2.0765228271484375, "step": 12870 }, { "epoch": 1.838716585174346, "grad_norm": 1.2558611631393433, "learning_rate": 7.752439895263033e-05, "loss": 2.0655872344970705, "step": 12880 }, { "epoch": 1.840144187872515, "grad_norm": 1.188773512840271, "learning_rate": 7.74291835277315e-05, "loss": 2.0430747985839846, "step": 12890 }, { "epoch": 1.841571790570684, "grad_norm": 1.1421024799346924, "learning_rate": 7.733396810283267e-05, "loss": 1.9896499633789062, "step": 12900 }, { "epoch": 1.8429993932688533, "grad_norm": 1.2236170768737793, "learning_rate": 7.723875267793384e-05, "loss": 2.179408645629883, "step": 12910 }, { "epoch": 1.8444269959670225, "grad_norm": 1.280597448348999, "learning_rate": 7.714353725303499e-05, "loss": 1.8246540069580077, "step": 12920 }, { "epoch": 1.8458545986651915, "grad_norm": 1.0576725006103516, "learning_rate": 7.704832182813616e-05, "loss": 2.0723712921142576, "step": 12930 }, { "epoch": 1.8472822013633605, "grad_norm": 1.4353753328323364, "learning_rate": 7.695310640323733e-05, "loss": 1.8861963272094726, "step": 12940 }, { "epoch": 1.8487098040615297, "grad_norm": 1.7334846258163452, "learning_rate": 7.68578909783385e-05, "loss": 2.0365550994873045, "step": 12950 }, { "epoch": 1.8501374067596987, "grad_norm": 1.175888180732727, "learning_rate": 7.676267555343965e-05, "loss": 2.1440099716186523, "step": 12960 }, { "epoch": 1.8515650094578677, "grad_norm": 1.2383686304092407, "learning_rate": 7.666746012854082e-05, "loss": 2.0059535980224608, "step": 12970 }, { "epoch": 1.852992612156037, "grad_norm": 1.120110273361206, "learning_rate": 7.657224470364199e-05, "loss": 2.0030309677124025, "step": 12980 }, { "epoch": 1.8544202148542062, "grad_norm": 1.1096537113189697, "learning_rate": 7.647702927874316e-05, "loss": 2.0035573959350588, "step": 12990 }, { "epoch": 1.8558478175523752, "grad_norm": 1.9018131494522095, "learning_rate": 7.638181385384433e-05, "loss": 2.007923889160156, "step": 13000 }, { "epoch": 1.8572754202505442, "grad_norm": 1.6539503335952759, "learning_rate": 7.628659842894548e-05, "loss": 2.0728021621704102, "step": 13010 }, { "epoch": 1.8587030229487134, "grad_norm": 1.035279631614685, "learning_rate": 7.619138300404665e-05, "loss": 2.0412168502807617, "step": 13020 }, { "epoch": 1.8601306256468826, "grad_norm": 1.600389003753662, "learning_rate": 7.609616757914782e-05, "loss": 1.9306547164916992, "step": 13030 }, { "epoch": 1.8615582283450516, "grad_norm": 1.1340457201004028, "learning_rate": 7.600095215424899e-05, "loss": 1.9325502395629883, "step": 13040 }, { "epoch": 1.8629858310432206, "grad_norm": 1.0685302019119263, "learning_rate": 7.590573672935016e-05, "loss": 2.0351631164550783, "step": 13050 }, { "epoch": 1.8644134337413898, "grad_norm": 1.3259185552597046, "learning_rate": 7.581052130445131e-05, "loss": 2.1083168029785155, "step": 13060 }, { "epoch": 1.865841036439559, "grad_norm": 1.1393513679504395, "learning_rate": 7.571530587955248e-05, "loss": 2.210805892944336, "step": 13070 }, { "epoch": 1.867268639137728, "grad_norm": 1.2840547561645508, "learning_rate": 7.562009045465365e-05, "loss": 2.015774726867676, "step": 13080 }, { "epoch": 1.868696241835897, "grad_norm": 1.253185749053955, "learning_rate": 7.552487502975482e-05, "loss": 2.0353321075439452, "step": 13090 }, { "epoch": 1.8701238445340662, "grad_norm": 1.268602728843689, "learning_rate": 7.542965960485599e-05, "loss": 2.0546897888183593, "step": 13100 }, { "epoch": 1.8715514472322354, "grad_norm": 1.383224606513977, "learning_rate": 7.533444417995716e-05, "loss": 2.1054380416870115, "step": 13110 }, { "epoch": 1.8729790499304042, "grad_norm": 1.0560451745986938, "learning_rate": 7.523922875505833e-05, "loss": 2.0454410552978515, "step": 13120 }, { "epoch": 1.8744066526285734, "grad_norm": 1.152633547782898, "learning_rate": 7.51440133301595e-05, "loss": 2.0112579345703123, "step": 13130 }, { "epoch": 1.8758342553267426, "grad_norm": 1.129319429397583, "learning_rate": 7.504879790526067e-05, "loss": 2.015283203125, "step": 13140 }, { "epoch": 1.8772618580249116, "grad_norm": 1.5799884796142578, "learning_rate": 7.495358248036182e-05, "loss": 1.9921918869018556, "step": 13150 }, { "epoch": 1.8786894607230806, "grad_norm": 1.2975870370864868, "learning_rate": 7.485836705546299e-05, "loss": 1.9921470642089845, "step": 13160 }, { "epoch": 1.8801170634212498, "grad_norm": 1.1885125637054443, "learning_rate": 7.476315163056416e-05, "loss": 2.0586156845092773, "step": 13170 }, { "epoch": 1.881544666119419, "grad_norm": 0.9033425450325012, "learning_rate": 7.466793620566533e-05, "loss": 1.9814388275146484, "step": 13180 }, { "epoch": 1.882972268817588, "grad_norm": 1.2185097932815552, "learning_rate": 7.45727207807665e-05, "loss": 1.907698631286621, "step": 13190 }, { "epoch": 1.884399871515757, "grad_norm": 1.124240517616272, "learning_rate": 7.447750535586765e-05, "loss": 2.015203285217285, "step": 13200 }, { "epoch": 1.8858274742139263, "grad_norm": 1.3447808027267456, "learning_rate": 7.438228993096882e-05, "loss": 1.8579696655273437, "step": 13210 }, { "epoch": 1.8872550769120955, "grad_norm": 1.0997867584228516, "learning_rate": 7.428707450606999e-05, "loss": 1.9801271438598633, "step": 13220 }, { "epoch": 1.8886826796102645, "grad_norm": 1.139083981513977, "learning_rate": 7.419185908117116e-05, "loss": 2.0315216064453123, "step": 13230 }, { "epoch": 1.8901102823084335, "grad_norm": 1.120589256286621, "learning_rate": 7.409664365627233e-05, "loss": 2.1057180404663085, "step": 13240 }, { "epoch": 1.8915378850066027, "grad_norm": 1.2882843017578125, "learning_rate": 7.400142823137348e-05, "loss": 1.987131118774414, "step": 13250 }, { "epoch": 1.892965487704772, "grad_norm": 1.1402063369750977, "learning_rate": 7.390621280647465e-05, "loss": 2.054110527038574, "step": 13260 }, { "epoch": 1.894393090402941, "grad_norm": 1.368349552154541, "learning_rate": 7.381099738157582e-05, "loss": 2.0707653045654295, "step": 13270 }, { "epoch": 1.89582069310111, "grad_norm": 1.165868878364563, "learning_rate": 7.371578195667699e-05, "loss": 2.0629817962646486, "step": 13280 }, { "epoch": 1.8972482957992791, "grad_norm": 1.1597265005111694, "learning_rate": 7.362056653177816e-05, "loss": 2.006464385986328, "step": 13290 }, { "epoch": 1.8986758984974481, "grad_norm": 1.0690085887908936, "learning_rate": 7.352535110687931e-05, "loss": 2.110106658935547, "step": 13300 }, { "epoch": 1.900103501195617, "grad_norm": 1.1247448921203613, "learning_rate": 7.343013568198048e-05, "loss": 1.995153045654297, "step": 13310 }, { "epoch": 1.9015311038937863, "grad_norm": 1.104827642440796, "learning_rate": 7.333492025708165e-05, "loss": 1.9939664840698241, "step": 13320 }, { "epoch": 1.9029587065919555, "grad_norm": 1.1032519340515137, "learning_rate": 7.323970483218282e-05, "loss": 2.0688501358032227, "step": 13330 }, { "epoch": 1.9043863092901245, "grad_norm": 1.156197190284729, "learning_rate": 7.314448940728398e-05, "loss": 2.0877119064331056, "step": 13340 }, { "epoch": 1.9058139119882935, "grad_norm": 1.303762435913086, "learning_rate": 7.304927398238514e-05, "loss": 1.9081499099731445, "step": 13350 }, { "epoch": 1.9072415146864627, "grad_norm": 1.2285877466201782, "learning_rate": 7.295405855748631e-05, "loss": 1.9472471237182618, "step": 13360 }, { "epoch": 1.908669117384632, "grad_norm": 1.198104977607727, "learning_rate": 7.285884313258748e-05, "loss": 2.011354446411133, "step": 13370 }, { "epoch": 1.910096720082801, "grad_norm": 1.3864994049072266, "learning_rate": 7.276362770768865e-05, "loss": 2.0561357498168946, "step": 13380 }, { "epoch": 1.91152432278097, "grad_norm": 1.4073015451431274, "learning_rate": 7.26684122827898e-05, "loss": 2.02865047454834, "step": 13390 }, { "epoch": 1.9129519254791392, "grad_norm": 1.3144506216049194, "learning_rate": 7.257319685789098e-05, "loss": 1.8037374496459961, "step": 13400 }, { "epoch": 1.9143795281773084, "grad_norm": 1.196794867515564, "learning_rate": 7.247798143299214e-05, "loss": 2.0771976470947267, "step": 13410 }, { "epoch": 1.9158071308754774, "grad_norm": 1.087507963180542, "learning_rate": 7.238276600809331e-05, "loss": 2.0818878173828126, "step": 13420 }, { "epoch": 1.9172347335736464, "grad_norm": 1.008420705795288, "learning_rate": 7.228755058319448e-05, "loss": 1.9760396957397461, "step": 13430 }, { "epoch": 1.9186623362718156, "grad_norm": 1.1370688676834106, "learning_rate": 7.219233515829564e-05, "loss": 2.1108943939208986, "step": 13440 }, { "epoch": 1.9200899389699848, "grad_norm": 1.310376524925232, "learning_rate": 7.20971197333968e-05, "loss": 2.109994316101074, "step": 13450 }, { "epoch": 1.9215175416681538, "grad_norm": 0.9570678472518921, "learning_rate": 7.200190430849797e-05, "loss": 2.0084840774536135, "step": 13460 }, { "epoch": 1.9229451443663228, "grad_norm": 1.1016850471496582, "learning_rate": 7.190668888359914e-05, "loss": 1.9378328323364258, "step": 13470 }, { "epoch": 1.924372747064492, "grad_norm": 1.1491199731826782, "learning_rate": 7.181147345870031e-05, "loss": 2.022365188598633, "step": 13480 }, { "epoch": 1.925800349762661, "grad_norm": 1.2013168334960938, "learning_rate": 7.171625803380148e-05, "loss": 1.9139543533325196, "step": 13490 }, { "epoch": 1.92722795246083, "grad_norm": 1.1147706508636475, "learning_rate": 7.162104260890265e-05, "loss": 1.847892951965332, "step": 13500 }, { "epoch": 1.9286555551589992, "grad_norm": 1.1967307329177856, "learning_rate": 7.152582718400382e-05, "loss": 2.129387855529785, "step": 13510 }, { "epoch": 1.9300831578571684, "grad_norm": 1.2188365459442139, "learning_rate": 7.143061175910499e-05, "loss": 2.063084030151367, "step": 13520 }, { "epoch": 1.9315107605553374, "grad_norm": 1.081373929977417, "learning_rate": 7.133539633420614e-05, "loss": 1.926097297668457, "step": 13530 }, { "epoch": 1.9329383632535064, "grad_norm": 1.3184794187545776, "learning_rate": 7.124018090930731e-05, "loss": 1.8432956695556642, "step": 13540 }, { "epoch": 1.9343659659516756, "grad_norm": 1.2679530382156372, "learning_rate": 7.114496548440848e-05, "loss": 2.065782165527344, "step": 13550 }, { "epoch": 1.9357935686498449, "grad_norm": 1.4751728773117065, "learning_rate": 7.104975005950965e-05, "loss": 2.1170568466186523, "step": 13560 }, { "epoch": 1.9372211713480139, "grad_norm": 1.2808234691619873, "learning_rate": 7.095453463461082e-05, "loss": 2.120288276672363, "step": 13570 }, { "epoch": 1.9386487740461829, "grad_norm": 1.3089933395385742, "learning_rate": 7.085931920971197e-05, "loss": 2.1926651000976562, "step": 13580 }, { "epoch": 1.940076376744352, "grad_norm": 1.0912245512008667, "learning_rate": 7.076410378481314e-05, "loss": 2.0562969207763673, "step": 13590 }, { "epoch": 1.9415039794425213, "grad_norm": 1.0755127668380737, "learning_rate": 7.066888835991431e-05, "loss": 2.026613235473633, "step": 13600 }, { "epoch": 1.9429315821406903, "grad_norm": 1.117268681526184, "learning_rate": 7.057367293501548e-05, "loss": 2.0486124038696287, "step": 13610 }, { "epoch": 1.9443591848388593, "grad_norm": 1.0038608312606812, "learning_rate": 7.047845751011665e-05, "loss": 1.9366825103759766, "step": 13620 }, { "epoch": 1.9457867875370285, "grad_norm": 1.3383045196533203, "learning_rate": 7.03832420852178e-05, "loss": 2.078375244140625, "step": 13630 }, { "epoch": 1.9472143902351975, "grad_norm": 1.2311493158340454, "learning_rate": 7.028802666031897e-05, "loss": 2.022788429260254, "step": 13640 }, { "epoch": 1.9486419929333665, "grad_norm": 1.1762433052062988, "learning_rate": 7.019281123542014e-05, "loss": 2.072213363647461, "step": 13650 }, { "epoch": 1.9500695956315357, "grad_norm": 1.121802806854248, "learning_rate": 7.009759581052131e-05, "loss": 2.029891014099121, "step": 13660 }, { "epoch": 1.951497198329705, "grad_norm": 1.1382726430892944, "learning_rate": 7.000238038562248e-05, "loss": 1.886621856689453, "step": 13670 }, { "epoch": 1.952924801027874, "grad_norm": 1.1324715614318848, "learning_rate": 6.990716496072364e-05, "loss": 1.8980741500854492, "step": 13680 }, { "epoch": 1.954352403726043, "grad_norm": 1.348748803138733, "learning_rate": 6.98119495358248e-05, "loss": 1.9898492813110351, "step": 13690 }, { "epoch": 1.9557800064242121, "grad_norm": 1.6330965757369995, "learning_rate": 6.971673411092597e-05, "loss": 1.930572509765625, "step": 13700 }, { "epoch": 1.9572076091223813, "grad_norm": 1.1046547889709473, "learning_rate": 6.962151868602714e-05, "loss": 2.068459129333496, "step": 13710 }, { "epoch": 1.9586352118205503, "grad_norm": 1.3646717071533203, "learning_rate": 6.952630326112831e-05, "loss": 2.077528190612793, "step": 13720 }, { "epoch": 1.9600628145187193, "grad_norm": 1.264501690864563, "learning_rate": 6.943108783622947e-05, "loss": 1.8320138931274415, "step": 13730 }, { "epoch": 1.9614904172168885, "grad_norm": 1.3768733739852905, "learning_rate": 6.933587241133064e-05, "loss": 2.0743762969970705, "step": 13740 }, { "epoch": 1.9629180199150578, "grad_norm": 1.0744850635528564, "learning_rate": 6.92406569864318e-05, "loss": 2.1287731170654296, "step": 13750 }, { "epoch": 1.9643456226132268, "grad_norm": 0.949566125869751, "learning_rate": 6.914544156153297e-05, "loss": 2.069428253173828, "step": 13760 }, { "epoch": 1.9657732253113958, "grad_norm": 1.3046458959579468, "learning_rate": 6.905022613663413e-05, "loss": 2.0628787994384767, "step": 13770 }, { "epoch": 1.967200828009565, "grad_norm": 1.2223836183547974, "learning_rate": 6.89550107117353e-05, "loss": 1.977756881713867, "step": 13780 }, { "epoch": 1.9686284307077342, "grad_norm": 1.576417326927185, "learning_rate": 6.885979528683647e-05, "loss": 2.068132019042969, "step": 13790 }, { "epoch": 1.9700560334059032, "grad_norm": 1.0100486278533936, "learning_rate": 6.876457986193764e-05, "loss": 1.9988759994506835, "step": 13800 }, { "epoch": 1.9714836361040722, "grad_norm": 1.098382592201233, "learning_rate": 6.86693644370388e-05, "loss": 2.086368751525879, "step": 13810 }, { "epoch": 1.9729112388022414, "grad_norm": 1.3169435262680054, "learning_rate": 6.857414901213996e-05, "loss": 2.0814802169799806, "step": 13820 }, { "epoch": 1.9743388415004104, "grad_norm": 1.2749425172805786, "learning_rate": 6.847893358724113e-05, "loss": 1.9730493545532226, "step": 13830 }, { "epoch": 1.9757664441985794, "grad_norm": 1.1500614881515503, "learning_rate": 6.83837181623423e-05, "loss": 2.0157394409179688, "step": 13840 }, { "epoch": 1.9771940468967486, "grad_norm": 1.308201551437378, "learning_rate": 6.828850273744347e-05, "loss": 1.9294050216674805, "step": 13850 }, { "epoch": 1.9786216495949178, "grad_norm": 1.300015926361084, "learning_rate": 6.819328731254463e-05, "loss": 2.149728202819824, "step": 13860 }, { "epoch": 1.9800492522930868, "grad_norm": 1.0803039073944092, "learning_rate": 6.80980718876458e-05, "loss": 1.9478628158569335, "step": 13870 }, { "epoch": 1.9814768549912558, "grad_norm": 1.1940149068832397, "learning_rate": 6.800285646274697e-05, "loss": 2.011227035522461, "step": 13880 }, { "epoch": 1.982904457689425, "grad_norm": 1.2206110954284668, "learning_rate": 6.790764103784814e-05, "loss": 2.096553993225098, "step": 13890 }, { "epoch": 1.9843320603875942, "grad_norm": 1.3003445863723755, "learning_rate": 6.781242561294931e-05, "loss": 2.0391618728637697, "step": 13900 }, { "epoch": 1.9857596630857632, "grad_norm": 1.1580556631088257, "learning_rate": 6.771721018805048e-05, "loss": 2.0295862197875976, "step": 13910 }, { "epoch": 1.9871872657839322, "grad_norm": 1.1516274213790894, "learning_rate": 6.762199476315163e-05, "loss": 1.9885385513305665, "step": 13920 }, { "epoch": 1.9886148684821014, "grad_norm": 1.1187989711761475, "learning_rate": 6.75267793382528e-05, "loss": 2.0330890655517577, "step": 13930 }, { "epoch": 1.9900424711802707, "grad_norm": 1.0881680250167847, "learning_rate": 6.743156391335397e-05, "loss": 2.0809677124023436, "step": 13940 }, { "epoch": 1.9914700738784397, "grad_norm": 1.1990684270858765, "learning_rate": 6.733634848845514e-05, "loss": 1.9832719802856444, "step": 13950 }, { "epoch": 1.9928976765766087, "grad_norm": 1.3778204917907715, "learning_rate": 6.72411330635563e-05, "loss": 2.0827215194702147, "step": 13960 }, { "epoch": 1.9943252792747779, "grad_norm": 1.2348558902740479, "learning_rate": 6.714591763865747e-05, "loss": 2.049481964111328, "step": 13970 }, { "epoch": 1.995752881972947, "grad_norm": 1.884214162826538, "learning_rate": 6.705070221375863e-05, "loss": 2.1559295654296875, "step": 13980 }, { "epoch": 1.9971804846711159, "grad_norm": 1.4469555616378784, "learning_rate": 6.69554867888598e-05, "loss": 1.96083984375, "step": 13990 }, { "epoch": 1.998608087369285, "grad_norm": 1.2459849119186401, "learning_rate": 6.686027136396097e-05, "loss": 2.010843276977539, "step": 14000 }, { "epoch": 2.0, "grad_norm": 1.1544809341430664, "learning_rate": 6.676505593906213e-05, "loss": 1.9843250274658204, "step": 14010 }, { "epoch": 2.001427602698169, "grad_norm": 0.9089034199714661, "learning_rate": 6.66698405141633e-05, "loss": 1.9421382904052735, "step": 14020 }, { "epoch": 2.002855205396338, "grad_norm": 1.098841667175293, "learning_rate": 6.657462508926447e-05, "loss": 2.0986822128295897, "step": 14030 }, { "epoch": 2.004282808094507, "grad_norm": 1.2563408613204956, "learning_rate": 6.648893120685551e-05, "loss": 1.9875288009643555, "step": 14040 }, { "epoch": 2.0057104107926764, "grad_norm": 1.0033822059631348, "learning_rate": 6.639371578195668e-05, "loss": 1.9294179916381835, "step": 14050 }, { "epoch": 2.0071380134908456, "grad_norm": 1.384800910949707, "learning_rate": 6.629850035705785e-05, "loss": 2.044336700439453, "step": 14060 }, { "epoch": 2.0085656161890144, "grad_norm": 1.0126487016677856, "learning_rate": 6.620328493215902e-05, "loss": 1.9618438720703124, "step": 14070 }, { "epoch": 2.0099932188871836, "grad_norm": 1.110125184059143, "learning_rate": 6.610806950726019e-05, "loss": 2.0328197479248047, "step": 14080 }, { "epoch": 2.011420821585353, "grad_norm": 1.0431183576583862, "learning_rate": 6.601285408236134e-05, "loss": 2.1001434326171875, "step": 14090 }, { "epoch": 2.012848424283522, "grad_norm": 1.2809975147247314, "learning_rate": 6.591763865746251e-05, "loss": 1.8712018966674804, "step": 14100 }, { "epoch": 2.014276026981691, "grad_norm": 1.1926685571670532, "learning_rate": 6.582242323256368e-05, "loss": 1.9982311248779296, "step": 14110 }, { "epoch": 2.01570362967986, "grad_norm": 1.1320374011993408, "learning_rate": 6.572720780766485e-05, "loss": 2.0903604507446287, "step": 14120 }, { "epoch": 2.0171312323780293, "grad_norm": 1.0534803867340088, "learning_rate": 6.563199238276602e-05, "loss": 2.013362503051758, "step": 14130 }, { "epoch": 2.0185588350761985, "grad_norm": 1.4934844970703125, "learning_rate": 6.553677695786717e-05, "loss": 2.0845956802368164, "step": 14140 }, { "epoch": 2.0199864377743673, "grad_norm": 1.2419373989105225, "learning_rate": 6.544156153296834e-05, "loss": 2.008251953125, "step": 14150 }, { "epoch": 2.0214140404725365, "grad_norm": 1.0556491613388062, "learning_rate": 6.534634610806951e-05, "loss": 1.9983444213867188, "step": 14160 }, { "epoch": 2.0228416431707057, "grad_norm": 1.3728446960449219, "learning_rate": 6.525113068317068e-05, "loss": 2.014224815368652, "step": 14170 }, { "epoch": 2.024269245868875, "grad_norm": 1.0407994985580444, "learning_rate": 6.515591525827184e-05, "loss": 1.9740034103393556, "step": 14180 }, { "epoch": 2.0256968485670437, "grad_norm": 1.1779024600982666, "learning_rate": 6.5060699833373e-05, "loss": 2.047129821777344, "step": 14190 }, { "epoch": 2.027124451265213, "grad_norm": 1.077741265296936, "learning_rate": 6.496548440847417e-05, "loss": 1.9232637405395507, "step": 14200 }, { "epoch": 2.028552053963382, "grad_norm": 1.4067444801330566, "learning_rate": 6.487026898357534e-05, "loss": 2.0248184204101562, "step": 14210 }, { "epoch": 2.029979656661551, "grad_norm": 0.943439781665802, "learning_rate": 6.477505355867651e-05, "loss": 1.966324234008789, "step": 14220 }, { "epoch": 2.03140725935972, "grad_norm": 1.1769946813583374, "learning_rate": 6.467983813377767e-05, "loss": 1.9678522109985352, "step": 14230 }, { "epoch": 2.0328348620578893, "grad_norm": 1.2881193161010742, "learning_rate": 6.458462270887884e-05, "loss": 2.1441604614257814, "step": 14240 }, { "epoch": 2.0342624647560585, "grad_norm": 1.1187770366668701, "learning_rate": 6.448940728398e-05, "loss": 1.986628532409668, "step": 14250 }, { "epoch": 2.0356900674542273, "grad_norm": 1.0948803424835205, "learning_rate": 6.439419185908117e-05, "loss": 1.8771780014038086, "step": 14260 }, { "epoch": 2.0371176701523965, "grad_norm": 1.0320061445236206, "learning_rate": 6.429897643418234e-05, "loss": 1.9376998901367188, "step": 14270 }, { "epoch": 2.0385452728505657, "grad_norm": 0.979020893573761, "learning_rate": 6.42037610092835e-05, "loss": 1.8472368240356445, "step": 14280 }, { "epoch": 2.039972875548735, "grad_norm": 1.913278341293335, "learning_rate": 6.410854558438467e-05, "loss": 1.9593034744262696, "step": 14290 }, { "epoch": 2.0414004782469037, "grad_norm": 1.5530829429626465, "learning_rate": 6.401333015948584e-05, "loss": 2.0295162200927734, "step": 14300 }, { "epoch": 2.042828080945073, "grad_norm": 1.194719672203064, "learning_rate": 6.3918114734587e-05, "loss": 2.0369091033935547, "step": 14310 }, { "epoch": 2.044255683643242, "grad_norm": 1.3701776266098022, "learning_rate": 6.382289930968817e-05, "loss": 2.045989418029785, "step": 14320 }, { "epoch": 2.0456832863414114, "grad_norm": 1.0509649515151978, "learning_rate": 6.372768388478933e-05, "loss": 1.9322639465332032, "step": 14330 }, { "epoch": 2.04711088903958, "grad_norm": 1.4132171869277954, "learning_rate": 6.36324684598905e-05, "loss": 2.0783029556274415, "step": 14340 }, { "epoch": 2.0485384917377494, "grad_norm": 1.180469274520874, "learning_rate": 6.353725303499167e-05, "loss": 2.0046579360961916, "step": 14350 }, { "epoch": 2.0499660944359186, "grad_norm": 1.1820670366287231, "learning_rate": 6.344203761009284e-05, "loss": 1.9886526107788085, "step": 14360 }, { "epoch": 2.0513936971340874, "grad_norm": 1.1659674644470215, "learning_rate": 6.3346822185194e-05, "loss": 2.0607460021972654, "step": 14370 }, { "epoch": 2.0528212998322566, "grad_norm": 1.2981035709381104, "learning_rate": 6.325160676029517e-05, "loss": 1.9812862396240234, "step": 14380 }, { "epoch": 2.054248902530426, "grad_norm": 1.0011330842971802, "learning_rate": 6.315639133539634e-05, "loss": 1.962367630004883, "step": 14390 }, { "epoch": 2.055676505228595, "grad_norm": 1.0153578519821167, "learning_rate": 6.306117591049751e-05, "loss": 1.8986312866210937, "step": 14400 }, { "epoch": 2.057104107926764, "grad_norm": 1.4835940599441528, "learning_rate": 6.296596048559868e-05, "loss": 2.03796443939209, "step": 14410 }, { "epoch": 2.058531710624933, "grad_norm": 1.381401777267456, "learning_rate": 6.287074506069984e-05, "loss": 1.8883466720581055, "step": 14420 }, { "epoch": 2.0599593133231022, "grad_norm": 1.09195876121521, "learning_rate": 6.2775529635801e-05, "loss": 1.9108970642089844, "step": 14430 }, { "epoch": 2.0613869160212714, "grad_norm": 1.1155352592468262, "learning_rate": 6.268031421090217e-05, "loss": 1.9888645172119142, "step": 14440 }, { "epoch": 2.06281451871944, "grad_norm": 1.2529290914535522, "learning_rate": 6.258509878600334e-05, "loss": 1.9784982681274415, "step": 14450 }, { "epoch": 2.0642421214176094, "grad_norm": 1.2153412103652954, "learning_rate": 6.248988336110451e-05, "loss": 2.051775360107422, "step": 14460 }, { "epoch": 2.0656697241157786, "grad_norm": 1.0196552276611328, "learning_rate": 6.239466793620567e-05, "loss": 1.9284236907958985, "step": 14470 }, { "epoch": 2.067097326813948, "grad_norm": 1.364317536354065, "learning_rate": 6.229945251130683e-05, "loss": 2.175960159301758, "step": 14480 }, { "epoch": 2.0685249295121166, "grad_norm": 1.0491912364959717, "learning_rate": 6.2204237086408e-05, "loss": 1.9170143127441406, "step": 14490 }, { "epoch": 2.069952532210286, "grad_norm": 1.133324146270752, "learning_rate": 6.210902166150917e-05, "loss": 1.9146308898925781, "step": 14500 }, { "epoch": 2.071380134908455, "grad_norm": 1.2340611219406128, "learning_rate": 6.201380623661034e-05, "loss": 2.0065280914306642, "step": 14510 }, { "epoch": 2.0728077376066243, "grad_norm": 1.2689627408981323, "learning_rate": 6.19185908117115e-05, "loss": 1.9906892776489258, "step": 14520 }, { "epoch": 2.074235340304793, "grad_norm": 1.20632803440094, "learning_rate": 6.182337538681267e-05, "loss": 1.9211248397827148, "step": 14530 }, { "epoch": 2.0756629430029623, "grad_norm": 1.0742918252944946, "learning_rate": 6.172815996191383e-05, "loss": 2.0323781967163086, "step": 14540 }, { "epoch": 2.0770905457011315, "grad_norm": 1.5465048551559448, "learning_rate": 6.1632944537015e-05, "loss": 1.9510377883911132, "step": 14550 }, { "epoch": 2.0785181483993003, "grad_norm": 1.1654354333877563, "learning_rate": 6.153772911211616e-05, "loss": 1.9593034744262696, "step": 14560 }, { "epoch": 2.0799457510974695, "grad_norm": 1.1346428394317627, "learning_rate": 6.144251368721733e-05, "loss": 2.123276710510254, "step": 14570 }, { "epoch": 2.0813733537956387, "grad_norm": 1.1301493644714355, "learning_rate": 6.13472982623185e-05, "loss": 2.0369842529296873, "step": 14580 }, { "epoch": 2.082800956493808, "grad_norm": 1.1545064449310303, "learning_rate": 6.125208283741967e-05, "loss": 1.930311393737793, "step": 14590 }, { "epoch": 2.0842285591919767, "grad_norm": 1.0770646333694458, "learning_rate": 6.115686741252083e-05, "loss": 1.9985736846923827, "step": 14600 }, { "epoch": 2.085656161890146, "grad_norm": 1.2177454233169556, "learning_rate": 6.106165198762199e-05, "loss": 2.0298110961914064, "step": 14610 }, { "epoch": 2.087083764588315, "grad_norm": 1.1246331930160522, "learning_rate": 6.096643656272316e-05, "loss": 2.1766342163085937, "step": 14620 }, { "epoch": 2.0885113672864843, "grad_norm": 1.0951778888702393, "learning_rate": 6.087122113782433e-05, "loss": 2.0222002029418946, "step": 14630 }, { "epoch": 2.089938969984653, "grad_norm": 1.1420915126800537, "learning_rate": 6.0776005712925496e-05, "loss": 2.06520938873291, "step": 14640 }, { "epoch": 2.0913665726828223, "grad_norm": 1.0830886363983154, "learning_rate": 6.0680790288026665e-05, "loss": 1.9397008895874024, "step": 14650 }, { "epoch": 2.0927941753809916, "grad_norm": 1.0558019876480103, "learning_rate": 6.058557486312783e-05, "loss": 2.0807777404785157, "step": 14660 }, { "epoch": 2.0942217780791608, "grad_norm": 1.2316898107528687, "learning_rate": 6.0490359438228996e-05, "loss": 1.940640640258789, "step": 14670 }, { "epoch": 2.0956493807773295, "grad_norm": 1.0672756433486938, "learning_rate": 6.0395144013330165e-05, "loss": 1.9634107589721679, "step": 14680 }, { "epoch": 2.0970769834754988, "grad_norm": 1.3840558528900146, "learning_rate": 6.0299928588431334e-05, "loss": 2.0458131790161134, "step": 14690 }, { "epoch": 2.098504586173668, "grad_norm": 1.1520706415176392, "learning_rate": 6.02047131635325e-05, "loss": 2.012078285217285, "step": 14700 }, { "epoch": 2.099932188871837, "grad_norm": 1.2449944019317627, "learning_rate": 6.010949773863366e-05, "loss": 2.0032716751098634, "step": 14710 }, { "epoch": 2.101359791570006, "grad_norm": 1.2121634483337402, "learning_rate": 6.001428231373483e-05, "loss": 2.0020044326782225, "step": 14720 }, { "epoch": 2.102787394268175, "grad_norm": 2.08486270904541, "learning_rate": 5.9919066888835996e-05, "loss": 1.8752206802368163, "step": 14730 }, { "epoch": 2.1042149969663444, "grad_norm": 1.3137811422348022, "learning_rate": 5.9823851463937165e-05, "loss": 1.9468732833862306, "step": 14740 }, { "epoch": 2.105642599664513, "grad_norm": 1.1885219812393188, "learning_rate": 5.9728636039038333e-05, "loss": 1.844599723815918, "step": 14750 }, { "epoch": 2.1070702023626824, "grad_norm": 1.1521098613739014, "learning_rate": 5.963342061413949e-05, "loss": 1.9542427062988281, "step": 14760 }, { "epoch": 2.1084978050608516, "grad_norm": 1.0784906148910522, "learning_rate": 5.953820518924066e-05, "loss": 2.0869415283203123, "step": 14770 }, { "epoch": 2.109925407759021, "grad_norm": 1.0031477212905884, "learning_rate": 5.9442989764341827e-05, "loss": 1.8794139862060546, "step": 14780 }, { "epoch": 2.1113530104571896, "grad_norm": 1.2494300603866577, "learning_rate": 5.9347774339442995e-05, "loss": 2.0164241790771484, "step": 14790 }, { "epoch": 2.112780613155359, "grad_norm": 1.3908637762069702, "learning_rate": 5.925255891454415e-05, "loss": 1.8808303833007813, "step": 14800 }, { "epoch": 2.114208215853528, "grad_norm": 1.3690054416656494, "learning_rate": 5.915734348964532e-05, "loss": 2.1947221755981445, "step": 14810 }, { "epoch": 2.1156358185516972, "grad_norm": 1.1622953414916992, "learning_rate": 5.906212806474649e-05, "loss": 1.963119125366211, "step": 14820 }, { "epoch": 2.117063421249866, "grad_norm": 1.0372982025146484, "learning_rate": 5.896691263984766e-05, "loss": 2.1200416564941404, "step": 14830 }, { "epoch": 2.1184910239480352, "grad_norm": 1.7104233503341675, "learning_rate": 5.8871697214948826e-05, "loss": 2.056626892089844, "step": 14840 }, { "epoch": 2.1199186266462045, "grad_norm": 1.118251085281372, "learning_rate": 5.877648179004999e-05, "loss": 1.958135986328125, "step": 14850 }, { "epoch": 2.1213462293443737, "grad_norm": 1.3177067041397095, "learning_rate": 5.868126636515116e-05, "loss": 1.9979423522949218, "step": 14860 }, { "epoch": 2.1227738320425424, "grad_norm": 1.5623153448104858, "learning_rate": 5.8586050940252326e-05, "loss": 2.0845569610595702, "step": 14870 }, { "epoch": 2.1242014347407117, "grad_norm": 1.2062747478485107, "learning_rate": 5.8490835515353495e-05, "loss": 1.8813060760498046, "step": 14880 }, { "epoch": 2.125629037438881, "grad_norm": 1.5583804845809937, "learning_rate": 5.8395620090454664e-05, "loss": 2.014190101623535, "step": 14890 }, { "epoch": 2.1270566401370496, "grad_norm": 1.1478065252304077, "learning_rate": 5.830040466555582e-05, "loss": 1.8913141250610352, "step": 14900 }, { "epoch": 2.128484242835219, "grad_norm": 1.252751350402832, "learning_rate": 5.820518924065699e-05, "loss": 1.9767805099487306, "step": 14910 }, { "epoch": 2.129911845533388, "grad_norm": 1.3447868824005127, "learning_rate": 5.810997381575816e-05, "loss": 1.9447132110595704, "step": 14920 }, { "epoch": 2.1313394482315573, "grad_norm": 1.0841307640075684, "learning_rate": 5.8014758390859326e-05, "loss": 1.8993200302124023, "step": 14930 }, { "epoch": 2.132767050929726, "grad_norm": 1.3026918172836304, "learning_rate": 5.7919542965960495e-05, "loss": 2.1418872833251954, "step": 14940 }, { "epoch": 2.1341946536278953, "grad_norm": 1.424307942390442, "learning_rate": 5.782432754106165e-05, "loss": 2.085586929321289, "step": 14950 }, { "epoch": 2.1356222563260645, "grad_norm": 1.162610650062561, "learning_rate": 5.772911211616282e-05, "loss": 2.1570777893066406, "step": 14960 }, { "epoch": 2.1370498590242337, "grad_norm": 1.3123327493667603, "learning_rate": 5.763389669126399e-05, "loss": 1.929331398010254, "step": 14970 }, { "epoch": 2.1384774617224025, "grad_norm": 1.2763254642486572, "learning_rate": 5.7538681266365157e-05, "loss": 2.0696670532226564, "step": 14980 }, { "epoch": 2.1399050644205717, "grad_norm": 1.1092665195465088, "learning_rate": 5.744346584146631e-05, "loss": 2.118440055847168, "step": 14990 }, { "epoch": 2.141332667118741, "grad_norm": 1.0695382356643677, "learning_rate": 5.734825041656748e-05, "loss": 2.218397331237793, "step": 15000 }, { "epoch": 2.14276026981691, "grad_norm": 1.1685717105865479, "learning_rate": 5.725303499166865e-05, "loss": 1.961024284362793, "step": 15010 }, { "epoch": 2.144187872515079, "grad_norm": 1.2546069622039795, "learning_rate": 5.715781956676982e-05, "loss": 1.9961881637573242, "step": 15020 }, { "epoch": 2.145615475213248, "grad_norm": 1.0668163299560547, "learning_rate": 5.706260414187099e-05, "loss": 2.125642013549805, "step": 15030 }, { "epoch": 2.1470430779114174, "grad_norm": 1.095601201057434, "learning_rate": 5.696738871697215e-05, "loss": 1.95892333984375, "step": 15040 }, { "epoch": 2.148470680609586, "grad_norm": 1.966359257698059, "learning_rate": 5.687217329207332e-05, "loss": 1.9628381729125977, "step": 15050 }, { "epoch": 2.1498982833077553, "grad_norm": 1.1606801748275757, "learning_rate": 5.677695786717449e-05, "loss": 1.9945301055908202, "step": 15060 }, { "epoch": 2.1513258860059246, "grad_norm": 1.3605085611343384, "learning_rate": 5.6681742442275656e-05, "loss": 2.049635124206543, "step": 15070 }, { "epoch": 2.1527534887040938, "grad_norm": 1.3199994564056396, "learning_rate": 5.6586527017376825e-05, "loss": 1.8475522994995117, "step": 15080 }, { "epoch": 2.1541810914022625, "grad_norm": 1.1955424547195435, "learning_rate": 5.649131159247798e-05, "loss": 2.0906286239624023, "step": 15090 }, { "epoch": 2.1556086941004318, "grad_norm": 0.9662287831306458, "learning_rate": 5.639609616757915e-05, "loss": 1.9949686050415039, "step": 15100 }, { "epoch": 2.157036296798601, "grad_norm": 1.0759321451187134, "learning_rate": 5.630088074268032e-05, "loss": 2.022344207763672, "step": 15110 }, { "epoch": 2.15846389949677, "grad_norm": 1.237735390663147, "learning_rate": 5.620566531778149e-05, "loss": 1.9482919692993164, "step": 15120 }, { "epoch": 2.159891502194939, "grad_norm": 1.2388379573822021, "learning_rate": 5.6110449892882656e-05, "loss": 1.9682613372802735, "step": 15130 }, { "epoch": 2.161319104893108, "grad_norm": 1.3245164155960083, "learning_rate": 5.601523446798381e-05, "loss": 1.9353225708007813, "step": 15140 }, { "epoch": 2.1627467075912774, "grad_norm": 1.4548062086105347, "learning_rate": 5.592001904308498e-05, "loss": 1.9768058776855468, "step": 15150 }, { "epoch": 2.1641743102894466, "grad_norm": 1.1336613893508911, "learning_rate": 5.582480361818615e-05, "loss": 2.0184919357299806, "step": 15160 }, { "epoch": 2.1656019129876154, "grad_norm": 1.5666173696517944, "learning_rate": 5.572958819328732e-05, "loss": 1.9705633163452148, "step": 15170 }, { "epoch": 2.1670295156857846, "grad_norm": 1.144084095954895, "learning_rate": 5.563437276838847e-05, "loss": 2.0450462341308593, "step": 15180 }, { "epoch": 2.168457118383954, "grad_norm": 1.1473979949951172, "learning_rate": 5.553915734348964e-05, "loss": 1.9039793014526367, "step": 15190 }, { "epoch": 2.169884721082123, "grad_norm": 1.1140494346618652, "learning_rate": 5.544394191859081e-05, "loss": 2.081629753112793, "step": 15200 }, { "epoch": 2.171312323780292, "grad_norm": 1.203489065170288, "learning_rate": 5.534872649369198e-05, "loss": 2.0587711334228516, "step": 15210 }, { "epoch": 2.172739926478461, "grad_norm": 1.2065495252609253, "learning_rate": 5.525351106879315e-05, "loss": 2.0020076751708986, "step": 15220 }, { "epoch": 2.1741675291766303, "grad_norm": 1.0686746835708618, "learning_rate": 5.515829564389431e-05, "loss": 1.9783267974853516, "step": 15230 }, { "epoch": 2.1755951318747995, "grad_norm": 1.1544544696807861, "learning_rate": 5.506308021899548e-05, "loss": 1.9598028182983398, "step": 15240 }, { "epoch": 2.1770227345729682, "grad_norm": 1.4742151498794556, "learning_rate": 5.496786479409665e-05, "loss": 1.8952831268310546, "step": 15250 }, { "epoch": 2.1784503372711375, "grad_norm": 1.5490636825561523, "learning_rate": 5.487264936919782e-05, "loss": 1.9991025924682617, "step": 15260 }, { "epoch": 2.1798779399693067, "grad_norm": 1.2429455518722534, "learning_rate": 5.4777433944298986e-05, "loss": 2.012087821960449, "step": 15270 }, { "epoch": 2.1813055426674755, "grad_norm": 1.2625069618225098, "learning_rate": 5.468221851940014e-05, "loss": 1.9305208206176758, "step": 15280 }, { "epoch": 2.1827331453656447, "grad_norm": 1.1262668371200562, "learning_rate": 5.458700309450131e-05, "loss": 1.9402307510375976, "step": 15290 }, { "epoch": 2.184160748063814, "grad_norm": 1.1682697534561157, "learning_rate": 5.449178766960248e-05, "loss": 2.062732124328613, "step": 15300 }, { "epoch": 2.185588350761983, "grad_norm": 1.2452971935272217, "learning_rate": 5.439657224470365e-05, "loss": 1.9681610107421874, "step": 15310 }, { "epoch": 2.187015953460152, "grad_norm": 1.2260686159133911, "learning_rate": 5.430135681980482e-05, "loss": 2.016870307922363, "step": 15320 }, { "epoch": 2.188443556158321, "grad_norm": 1.3355969190597534, "learning_rate": 5.420614139490597e-05, "loss": 1.959249496459961, "step": 15330 }, { "epoch": 2.1898711588564903, "grad_norm": 1.3273026943206787, "learning_rate": 5.411092597000714e-05, "loss": 1.9223682403564453, "step": 15340 }, { "epoch": 2.1912987615546595, "grad_norm": 1.3289768695831299, "learning_rate": 5.401571054510831e-05, "loss": 1.9478498458862306, "step": 15350 }, { "epoch": 2.1927263642528283, "grad_norm": 1.4009289741516113, "learning_rate": 5.392049512020948e-05, "loss": 1.811954879760742, "step": 15360 }, { "epoch": 2.1941539669509975, "grad_norm": 1.242896556854248, "learning_rate": 5.382527969531064e-05, "loss": 2.158904266357422, "step": 15370 }, { "epoch": 2.1955815696491667, "grad_norm": 1.8332645893096924, "learning_rate": 5.37300642704118e-05, "loss": 2.1291406631469725, "step": 15380 }, { "epoch": 2.197009172347336, "grad_norm": 1.2584983110427856, "learning_rate": 5.363484884551297e-05, "loss": 2.1477169036865233, "step": 15390 }, { "epoch": 2.1984367750455047, "grad_norm": 1.343069076538086, "learning_rate": 5.353963342061414e-05, "loss": 1.8398696899414062, "step": 15400 }, { "epoch": 2.199864377743674, "grad_norm": 1.2898625135421753, "learning_rate": 5.344441799571531e-05, "loss": 2.0486936569213867, "step": 15410 }, { "epoch": 2.201291980441843, "grad_norm": 1.4354395866394043, "learning_rate": 5.334920257081647e-05, "loss": 2.1760021209716798, "step": 15420 }, { "epoch": 2.202719583140012, "grad_norm": 1.180816888809204, "learning_rate": 5.325398714591764e-05, "loss": 2.03610782623291, "step": 15430 }, { "epoch": 2.204147185838181, "grad_norm": 1.3519412279129028, "learning_rate": 5.315877172101881e-05, "loss": 2.1258440017700195, "step": 15440 }, { "epoch": 2.2055747885363504, "grad_norm": 1.7197996377944946, "learning_rate": 5.306355629611998e-05, "loss": 1.9166088104248047, "step": 15450 }, { "epoch": 2.2070023912345196, "grad_norm": 1.1483829021453857, "learning_rate": 5.296834087122115e-05, "loss": 1.8572282791137695, "step": 15460 }, { "epoch": 2.2084299939326884, "grad_norm": 1.0768811702728271, "learning_rate": 5.28731254463223e-05, "loss": 1.8752784729003906, "step": 15470 }, { "epoch": 2.2098575966308576, "grad_norm": 1.6598796844482422, "learning_rate": 5.277791002142347e-05, "loss": 2.0793161392211914, "step": 15480 }, { "epoch": 2.211285199329027, "grad_norm": 1.2683194875717163, "learning_rate": 5.268269459652464e-05, "loss": 1.9972589492797852, "step": 15490 }, { "epoch": 2.212712802027196, "grad_norm": 1.3703789710998535, "learning_rate": 5.258747917162581e-05, "loss": 2.006252479553223, "step": 15500 }, { "epoch": 2.2141404047253648, "grad_norm": 1.2037006616592407, "learning_rate": 5.249226374672698e-05, "loss": 1.895115852355957, "step": 15510 }, { "epoch": 2.215568007423534, "grad_norm": 1.2178536653518677, "learning_rate": 5.239704832182813e-05, "loss": 2.0101078033447264, "step": 15520 }, { "epoch": 2.216995610121703, "grad_norm": 1.2817871570587158, "learning_rate": 5.23018328969293e-05, "loss": 2.0475120544433594, "step": 15530 }, { "epoch": 2.2184232128198724, "grad_norm": 1.2774466276168823, "learning_rate": 5.220661747203047e-05, "loss": 2.036044120788574, "step": 15540 }, { "epoch": 2.219850815518041, "grad_norm": 1.1693365573883057, "learning_rate": 5.211140204713164e-05, "loss": 2.0045270919799805, "step": 15550 }, { "epoch": 2.2212784182162104, "grad_norm": 1.3637462854385376, "learning_rate": 5.20161866222328e-05, "loss": 1.889479637145996, "step": 15560 }, { "epoch": 2.2227060209143796, "grad_norm": 1.2539608478546143, "learning_rate": 5.1920971197333964e-05, "loss": 1.9617826461791992, "step": 15570 }, { "epoch": 2.2241336236125484, "grad_norm": 1.668036699295044, "learning_rate": 5.182575577243513e-05, "loss": 2.090685272216797, "step": 15580 }, { "epoch": 2.2255612263107176, "grad_norm": 1.1080751419067383, "learning_rate": 5.17305403475363e-05, "loss": 1.8956758499145507, "step": 15590 }, { "epoch": 2.226988829008887, "grad_norm": 1.1848517656326294, "learning_rate": 5.163532492263747e-05, "loss": 1.959829330444336, "step": 15600 }, { "epoch": 2.228416431707056, "grad_norm": 1.148152470588684, "learning_rate": 5.154010949773863e-05, "loss": 2.011181640625, "step": 15610 }, { "epoch": 2.229844034405225, "grad_norm": 1.2634568214416504, "learning_rate": 5.14448940728398e-05, "loss": 2.071341133117676, "step": 15620 }, { "epoch": 2.231271637103394, "grad_norm": 1.1462156772613525, "learning_rate": 5.134967864794097e-05, "loss": 1.9244178771972655, "step": 15630 }, { "epoch": 2.2326992398015633, "grad_norm": 1.2139183282852173, "learning_rate": 5.125446322304214e-05, "loss": 1.9124441146850586, "step": 15640 }, { "epoch": 2.2341268424997325, "grad_norm": 1.174119234085083, "learning_rate": 5.115924779814331e-05, "loss": 1.9380500793457032, "step": 15650 }, { "epoch": 2.2355544451979013, "grad_norm": 1.0516029596328735, "learning_rate": 5.1064032373244464e-05, "loss": 1.947065544128418, "step": 15660 }, { "epoch": 2.2369820478960705, "grad_norm": 1.2596803903579712, "learning_rate": 5.096881694834563e-05, "loss": 2.008541488647461, "step": 15670 }, { "epoch": 2.2384096505942397, "grad_norm": 1.1278873682022095, "learning_rate": 5.08736015234468e-05, "loss": 2.0980831146240235, "step": 15680 }, { "epoch": 2.239837253292409, "grad_norm": 1.1728241443634033, "learning_rate": 5.077838609854797e-05, "loss": 2.0734405517578125, "step": 15690 }, { "epoch": 2.2412648559905777, "grad_norm": 1.2889844179153442, "learning_rate": 5.068317067364914e-05, "loss": 2.0412111282348633, "step": 15700 }, { "epoch": 2.242692458688747, "grad_norm": 1.1203408241271973, "learning_rate": 5.0587955248750294e-05, "loss": 2.1112045288085937, "step": 15710 }, { "epoch": 2.244120061386916, "grad_norm": 1.0422649383544922, "learning_rate": 5.049273982385146e-05, "loss": 1.9882959365844726, "step": 15720 }, { "epoch": 2.2455476640850853, "grad_norm": 1.311018466949463, "learning_rate": 5.039752439895263e-05, "loss": 2.0124526977539063, "step": 15730 }, { "epoch": 2.246975266783254, "grad_norm": 1.2563869953155518, "learning_rate": 5.03023089740538e-05, "loss": 1.9363046646118165, "step": 15740 }, { "epoch": 2.2484028694814233, "grad_norm": 1.204236388206482, "learning_rate": 5.020709354915497e-05, "loss": 2.0383895874023437, "step": 15750 }, { "epoch": 2.2498304721795925, "grad_norm": 1.2544084787368774, "learning_rate": 5.0111878124256125e-05, "loss": 2.0026008605957033, "step": 15760 }, { "epoch": 2.2512580748777618, "grad_norm": 1.2330925464630127, "learning_rate": 5.0016662699357294e-05, "loss": 1.965739631652832, "step": 15770 }, { "epoch": 2.2526856775759305, "grad_norm": 1.1666339635849, "learning_rate": 4.992144727445846e-05, "loss": 2.0555065155029295, "step": 15780 }, { "epoch": 2.2541132802740997, "grad_norm": 0.9663774371147156, "learning_rate": 4.982623184955963e-05, "loss": 1.9195777893066406, "step": 15790 }, { "epoch": 2.255540882972269, "grad_norm": 1.101117491722107, "learning_rate": 4.97310164246608e-05, "loss": 2.0098329544067384, "step": 15800 }, { "epoch": 2.2569684856704377, "grad_norm": 1.1207876205444336, "learning_rate": 4.963580099976196e-05, "loss": 1.9372873306274414, "step": 15810 }, { "epoch": 2.258396088368607, "grad_norm": 1.3611836433410645, "learning_rate": 4.954058557486313e-05, "loss": 2.0347244262695314, "step": 15820 }, { "epoch": 2.259823691066776, "grad_norm": 1.3885074853897095, "learning_rate": 4.94453701499643e-05, "loss": 1.9107807159423829, "step": 15830 }, { "epoch": 2.2612512937649454, "grad_norm": 1.3332844972610474, "learning_rate": 4.935015472506546e-05, "loss": 1.80665283203125, "step": 15840 }, { "epoch": 2.262678896463114, "grad_norm": 1.1626338958740234, "learning_rate": 4.925493930016663e-05, "loss": 1.9481868743896484, "step": 15850 }, { "epoch": 2.2641064991612834, "grad_norm": 1.1729705333709717, "learning_rate": 4.9159723875267794e-05, "loss": 2.1106090545654297, "step": 15860 }, { "epoch": 2.2655341018594526, "grad_norm": 1.2503594160079956, "learning_rate": 4.906450845036896e-05, "loss": 2.0823108673095705, "step": 15870 }, { "epoch": 2.266961704557622, "grad_norm": 1.4801392555236816, "learning_rate": 4.896929302547013e-05, "loss": 2.0672435760498047, "step": 15880 }, { "epoch": 2.2683893072557906, "grad_norm": 1.5278877019882202, "learning_rate": 4.8874077600571293e-05, "loss": 1.8716043472290038, "step": 15890 }, { "epoch": 2.26981690995396, "grad_norm": 1.09286367893219, "learning_rate": 4.877886217567246e-05, "loss": 1.821883201599121, "step": 15900 }, { "epoch": 2.271244512652129, "grad_norm": 1.230090856552124, "learning_rate": 4.8683646750773624e-05, "loss": 1.950247573852539, "step": 15910 }, { "epoch": 2.2726721153502982, "grad_norm": 1.0066258907318115, "learning_rate": 4.858843132587479e-05, "loss": 1.9462762832641602, "step": 15920 }, { "epoch": 2.274099718048467, "grad_norm": 1.0531705617904663, "learning_rate": 4.8493215900975955e-05, "loss": 1.9932144165039063, "step": 15930 }, { "epoch": 2.275527320746636, "grad_norm": 1.2815144062042236, "learning_rate": 4.8398000476077124e-05, "loss": 1.9634632110595702, "step": 15940 }, { "epoch": 2.2769549234448054, "grad_norm": 1.2811980247497559, "learning_rate": 4.830278505117829e-05, "loss": 2.0316532135009764, "step": 15950 }, { "epoch": 2.278382526142974, "grad_norm": 1.3156558275222778, "learning_rate": 4.8207569626279455e-05, "loss": 1.992854690551758, "step": 15960 }, { "epoch": 2.2798101288411434, "grad_norm": 1.1985957622528076, "learning_rate": 4.8112354201380624e-05, "loss": 2.0715314865112306, "step": 15970 }, { "epoch": 2.2812377315393126, "grad_norm": 1.3669253587722778, "learning_rate": 4.801713877648179e-05, "loss": 2.101321220397949, "step": 15980 }, { "epoch": 2.282665334237482, "grad_norm": 1.1986885070800781, "learning_rate": 4.792192335158296e-05, "loss": 1.9884351730346679, "step": 15990 }, { "epoch": 2.2840929369356506, "grad_norm": 1.1781433820724487, "learning_rate": 4.782670792668413e-05, "loss": 1.9585966110229491, "step": 16000 }, { "epoch": 2.28552053963382, "grad_norm": 1.074578046798706, "learning_rate": 4.773149250178529e-05, "loss": 2.0398279190063477, "step": 16010 }, { "epoch": 2.286948142331989, "grad_norm": 1.1704986095428467, "learning_rate": 4.763627707688646e-05, "loss": 1.8830327987670898, "step": 16020 }, { "epoch": 2.2883757450301583, "grad_norm": 1.1421358585357666, "learning_rate": 4.7541061651987624e-05, "loss": 1.9847730636596679, "step": 16030 }, { "epoch": 2.289803347728327, "grad_norm": 1.0410007238388062, "learning_rate": 4.744584622708879e-05, "loss": 1.7843944549560546, "step": 16040 }, { "epoch": 2.2912309504264963, "grad_norm": 1.2032654285430908, "learning_rate": 4.7350630802189955e-05, "loss": 2.056909370422363, "step": 16050 }, { "epoch": 2.2926585531246655, "grad_norm": 1.2183992862701416, "learning_rate": 4.7255415377291124e-05, "loss": 2.0453285217285155, "step": 16060 }, { "epoch": 2.2940861558228347, "grad_norm": 1.17236328125, "learning_rate": 4.716019995239229e-05, "loss": 1.9122222900390624, "step": 16070 }, { "epoch": 2.2955137585210035, "grad_norm": 1.2542738914489746, "learning_rate": 4.7064984527493455e-05, "loss": 2.1268373489379884, "step": 16080 }, { "epoch": 2.2969413612191727, "grad_norm": 1.0163629055023193, "learning_rate": 4.6969769102594623e-05, "loss": 2.0127052307128905, "step": 16090 }, { "epoch": 2.298368963917342, "grad_norm": 1.1987311840057373, "learning_rate": 4.6874553677695786e-05, "loss": 1.9507835388183594, "step": 16100 }, { "epoch": 2.2997965666155107, "grad_norm": 1.1987152099609375, "learning_rate": 4.6779338252796954e-05, "loss": 1.9573074340820313, "step": 16110 }, { "epoch": 2.30122416931368, "grad_norm": 1.3898403644561768, "learning_rate": 4.6684122827898117e-05, "loss": 2.116376495361328, "step": 16120 }, { "epoch": 2.302651772011849, "grad_norm": 1.5338342189788818, "learning_rate": 4.6588907402999285e-05, "loss": 1.9208789825439454, "step": 16130 }, { "epoch": 2.3040793747100183, "grad_norm": 1.8814921379089355, "learning_rate": 4.6493691978100454e-05, "loss": 2.0063961029052733, "step": 16140 }, { "epoch": 2.305506977408187, "grad_norm": 1.0622856616973877, "learning_rate": 4.6398476553201616e-05, "loss": 1.9159837722778321, "step": 16150 }, { "epoch": 2.3069345801063563, "grad_norm": 1.059657335281372, "learning_rate": 4.6303261128302785e-05, "loss": 2.1166818618774412, "step": 16160 }, { "epoch": 2.3083621828045255, "grad_norm": 1.2254993915557861, "learning_rate": 4.6208045703403954e-05, "loss": 1.8073310852050781, "step": 16170 }, { "epoch": 2.3097897855026948, "grad_norm": 1.2716083526611328, "learning_rate": 4.611283027850512e-05, "loss": 1.9563461303710938, "step": 16180 }, { "epoch": 2.3112173882008635, "grad_norm": 1.1647034883499146, "learning_rate": 4.601761485360629e-05, "loss": 2.0599422454833984, "step": 16190 }, { "epoch": 2.3126449908990327, "grad_norm": 1.567744255065918, "learning_rate": 4.5922399428707454e-05, "loss": 1.992660140991211, "step": 16200 }, { "epoch": 2.314072593597202, "grad_norm": 1.339717149734497, "learning_rate": 4.582718400380862e-05, "loss": 2.0070659637451174, "step": 16210 }, { "epoch": 2.315500196295371, "grad_norm": 1.2331042289733887, "learning_rate": 4.5731968578909785e-05, "loss": 2.0130708694458006, "step": 16220 }, { "epoch": 2.31692779899354, "grad_norm": 1.085657000541687, "learning_rate": 4.5636753154010954e-05, "loss": 1.9026834487915039, "step": 16230 }, { "epoch": 2.318355401691709, "grad_norm": 1.525377631187439, "learning_rate": 4.5541537729112116e-05, "loss": 2.0197216033935548, "step": 16240 }, { "epoch": 2.3197830043898784, "grad_norm": 1.2509006261825562, "learning_rate": 4.5446322304213285e-05, "loss": 1.992051124572754, "step": 16250 }, { "epoch": 2.321210607088047, "grad_norm": 1.1260684728622437, "learning_rate": 4.5351106879314454e-05, "loss": 1.8842735290527344, "step": 16260 }, { "epoch": 2.3226382097862164, "grad_norm": 1.3706077337265015, "learning_rate": 4.5255891454415616e-05, "loss": 1.9503765106201172, "step": 16270 }, { "epoch": 2.3240658124843856, "grad_norm": 1.130692481994629, "learning_rate": 4.5160676029516785e-05, "loss": 1.9896425247192382, "step": 16280 }, { "epoch": 2.325493415182555, "grad_norm": 1.5941585302352905, "learning_rate": 4.506546060461795e-05, "loss": 1.937184715270996, "step": 16290 }, { "epoch": 2.326921017880724, "grad_norm": 1.4923607110977173, "learning_rate": 4.4970245179719116e-05, "loss": 1.8169942855834962, "step": 16300 }, { "epoch": 2.328348620578893, "grad_norm": 1.1341545581817627, "learning_rate": 4.487502975482028e-05, "loss": 2.159044647216797, "step": 16310 }, { "epoch": 2.329776223277062, "grad_norm": 1.1580032110214233, "learning_rate": 4.4779814329921447e-05, "loss": 2.0251697540283202, "step": 16320 }, { "epoch": 2.3312038259752312, "grad_norm": 1.1192718744277954, "learning_rate": 4.4684598905022615e-05, "loss": 2.0953981399536135, "step": 16330 }, { "epoch": 2.3326314286734, "grad_norm": 1.0530462265014648, "learning_rate": 4.458938348012378e-05, "loss": 2.11028938293457, "step": 16340 }, { "epoch": 2.3340590313715692, "grad_norm": 1.1743799448013306, "learning_rate": 4.4494168055224946e-05, "loss": 2.001978302001953, "step": 16350 }, { "epoch": 2.3354866340697384, "grad_norm": 1.4625071287155151, "learning_rate": 4.4398952630326115e-05, "loss": 2.0029891967773437, "step": 16360 }, { "epoch": 2.3369142367679077, "grad_norm": 1.439976453781128, "learning_rate": 4.4303737205427284e-05, "loss": 2.1157625198364256, "step": 16370 }, { "epoch": 2.3383418394660764, "grad_norm": 1.2653416395187378, "learning_rate": 4.420852178052845e-05, "loss": 1.9372739791870117, "step": 16380 }, { "epoch": 2.3397694421642456, "grad_norm": 1.1437488794326782, "learning_rate": 4.4113306355629615e-05, "loss": 2.0752405166625976, "step": 16390 }, { "epoch": 2.341197044862415, "grad_norm": 1.2199393510818481, "learning_rate": 4.4018090930730784e-05, "loss": 1.889252281188965, "step": 16400 }, { "epoch": 2.342624647560584, "grad_norm": 1.156009554862976, "learning_rate": 4.3922875505831946e-05, "loss": 1.7882341384887694, "step": 16410 }, { "epoch": 2.344052250258753, "grad_norm": 1.225005865097046, "learning_rate": 4.3827660080933115e-05, "loss": 1.8980962753295898, "step": 16420 }, { "epoch": 2.345479852956922, "grad_norm": 1.1961416006088257, "learning_rate": 4.373244465603428e-05, "loss": 2.054966354370117, "step": 16430 }, { "epoch": 2.3469074556550913, "grad_norm": 1.1770035028457642, "learning_rate": 4.3637229231135446e-05, "loss": 2.0128704071044923, "step": 16440 }, { "epoch": 2.3483350583532605, "grad_norm": 1.157025933265686, "learning_rate": 4.3542013806236615e-05, "loss": 2.069859504699707, "step": 16450 }, { "epoch": 2.3497626610514293, "grad_norm": 1.2290818691253662, "learning_rate": 4.344679838133778e-05, "loss": 1.9863374710083008, "step": 16460 }, { "epoch": 2.3511902637495985, "grad_norm": 1.4846012592315674, "learning_rate": 4.3351582956438946e-05, "loss": 2.068705749511719, "step": 16470 }, { "epoch": 2.3526178664477677, "grad_norm": 1.4807878732681274, "learning_rate": 4.325636753154011e-05, "loss": 1.9752376556396485, "step": 16480 }, { "epoch": 2.3540454691459365, "grad_norm": 1.1961666345596313, "learning_rate": 4.316115210664128e-05, "loss": 1.8829071044921875, "step": 16490 }, { "epoch": 2.3554730718441057, "grad_norm": 1.0952601432800293, "learning_rate": 4.3065936681742446e-05, "loss": 1.9274658203125, "step": 16500 }, { "epoch": 2.356900674542275, "grad_norm": 0.9354764223098755, "learning_rate": 4.297072125684361e-05, "loss": 2.0875223159790037, "step": 16510 }, { "epoch": 2.358328277240444, "grad_norm": 1.2836897373199463, "learning_rate": 4.2875505831944777e-05, "loss": 2.075917625427246, "step": 16520 }, { "epoch": 2.359755879938613, "grad_norm": 1.1640502214431763, "learning_rate": 4.2780290407045945e-05, "loss": 1.8998403549194336, "step": 16530 }, { "epoch": 2.361183482636782, "grad_norm": 1.3402345180511475, "learning_rate": 4.268507498214711e-05, "loss": 1.9744211196899415, "step": 16540 }, { "epoch": 2.3626110853349513, "grad_norm": 1.2828158140182495, "learning_rate": 4.2589859557248276e-05, "loss": 1.9106904983520507, "step": 16550 }, { "epoch": 2.3640386880331206, "grad_norm": 1.6601427793502808, "learning_rate": 4.2494644132349445e-05, "loss": 2.053148078918457, "step": 16560 }, { "epoch": 2.3654662907312893, "grad_norm": 1.1292343139648438, "learning_rate": 4.2399428707450614e-05, "loss": 1.9551586151123046, "step": 16570 }, { "epoch": 2.3668938934294586, "grad_norm": 1.65964674949646, "learning_rate": 4.2304213282551776e-05, "loss": 1.969891357421875, "step": 16580 }, { "epoch": 2.3683214961276278, "grad_norm": 1.4649289846420288, "learning_rate": 4.2208997857652945e-05, "loss": 1.885755729675293, "step": 16590 }, { "epoch": 2.369749098825797, "grad_norm": 1.1072564125061035, "learning_rate": 4.211378243275411e-05, "loss": 2.0994169235229494, "step": 16600 }, { "epoch": 2.3711767015239658, "grad_norm": 1.317319631576538, "learning_rate": 4.2018567007855276e-05, "loss": 2.06343936920166, "step": 16610 }, { "epoch": 2.372604304222135, "grad_norm": 1.1138535737991333, "learning_rate": 4.192335158295644e-05, "loss": 1.8904455184936524, "step": 16620 }, { "epoch": 2.374031906920304, "grad_norm": 1.3210810422897339, "learning_rate": 4.182813615805761e-05, "loss": 2.0521852493286135, "step": 16630 }, { "epoch": 2.375459509618473, "grad_norm": 1.0153765678405762, "learning_rate": 4.1732920733158776e-05, "loss": 2.016363525390625, "step": 16640 }, { "epoch": 2.376887112316642, "grad_norm": 1.4659984111785889, "learning_rate": 4.163770530825994e-05, "loss": 1.9123205184936523, "step": 16650 }, { "epoch": 2.3783147150148114, "grad_norm": 1.2185951471328735, "learning_rate": 4.154248988336111e-05, "loss": 2.110964775085449, "step": 16660 }, { "epoch": 2.3797423177129806, "grad_norm": 1.1308666467666626, "learning_rate": 4.144727445846227e-05, "loss": 1.9757112503051757, "step": 16670 }, { "epoch": 2.3811699204111494, "grad_norm": 1.0685782432556152, "learning_rate": 4.135205903356344e-05, "loss": 2.0017995834350586, "step": 16680 }, { "epoch": 2.3825975231093186, "grad_norm": 1.2273656129837036, "learning_rate": 4.125684360866461e-05, "loss": 1.9322301864624023, "step": 16690 }, { "epoch": 2.384025125807488, "grad_norm": 1.3868802785873413, "learning_rate": 4.116162818376577e-05, "loss": 2.044620323181152, "step": 16700 }, { "epoch": 2.385452728505657, "grad_norm": 1.34739089012146, "learning_rate": 4.106641275886694e-05, "loss": 1.9810251235961913, "step": 16710 }, { "epoch": 2.386880331203826, "grad_norm": 0.949345588684082, "learning_rate": 4.0971197333968107e-05, "loss": 2.028364562988281, "step": 16720 }, { "epoch": 2.388307933901995, "grad_norm": 1.3317279815673828, "learning_rate": 4.087598190906927e-05, "loss": 1.9809131622314453, "step": 16730 }, { "epoch": 2.3897355366001642, "grad_norm": 1.2553988695144653, "learning_rate": 4.078076648417044e-05, "loss": 2.1058795928955076, "step": 16740 }, { "epoch": 2.3911631392983335, "grad_norm": 1.2300201654434204, "learning_rate": 4.0685551059271606e-05, "loss": 2.0187267303466796, "step": 16750 }, { "epoch": 2.3925907419965022, "grad_norm": 1.0214483737945557, "learning_rate": 4.0590335634372775e-05, "loss": 1.9418626785278321, "step": 16760 }, { "epoch": 2.3940183446946715, "grad_norm": 1.201991319656372, "learning_rate": 4.049512020947394e-05, "loss": 1.9170721054077149, "step": 16770 }, { "epoch": 2.3954459473928407, "grad_norm": 1.114147424697876, "learning_rate": 4.0399904784575106e-05, "loss": 2.0436492919921876, "step": 16780 }, { "epoch": 2.3968735500910094, "grad_norm": 1.1946865320205688, "learning_rate": 4.030468935967627e-05, "loss": 1.9630859375, "step": 16790 }, { "epoch": 2.3983011527891787, "grad_norm": 1.1444975137710571, "learning_rate": 4.020947393477744e-05, "loss": 1.9546268463134766, "step": 16800 }, { "epoch": 2.399728755487348, "grad_norm": 1.1698689460754395, "learning_rate": 4.0114258509878606e-05, "loss": 2.044775199890137, "step": 16810 }, { "epoch": 2.401156358185517, "grad_norm": 1.381348729133606, "learning_rate": 4.001904308497977e-05, "loss": 2.0606056213378907, "step": 16820 }, { "epoch": 2.4025839608836863, "grad_norm": 1.1631004810333252, "learning_rate": 3.992382766008094e-05, "loss": 1.9848783493041993, "step": 16830 }, { "epoch": 2.404011563581855, "grad_norm": 1.2258085012435913, "learning_rate": 3.98286122351821e-05, "loss": 2.0091304779052734, "step": 16840 }, { "epoch": 2.4054391662800243, "grad_norm": 1.032585620880127, "learning_rate": 3.973339681028327e-05, "loss": 2.0976936340332033, "step": 16850 }, { "epoch": 2.4068667689781935, "grad_norm": 1.0957385301589966, "learning_rate": 3.963818138538443e-05, "loss": 2.1347585678100587, "step": 16860 }, { "epoch": 2.4082943716763623, "grad_norm": 1.3036447763442993, "learning_rate": 3.95429659604856e-05, "loss": 1.8763076782226562, "step": 16870 }, { "epoch": 2.4097219743745315, "grad_norm": 1.344551920890808, "learning_rate": 3.944775053558677e-05, "loss": 1.9801139831542969, "step": 16880 }, { "epoch": 2.4111495770727007, "grad_norm": 1.3332304954528809, "learning_rate": 3.935253511068793e-05, "loss": 2.0261245727539063, "step": 16890 }, { "epoch": 2.41257717977087, "grad_norm": 1.0788671970367432, "learning_rate": 3.92573196857891e-05, "loss": 1.978651237487793, "step": 16900 }, { "epoch": 2.4140047824690387, "grad_norm": 1.2969266176223755, "learning_rate": 3.916210426089027e-05, "loss": 1.952248764038086, "step": 16910 }, { "epoch": 2.415432385167208, "grad_norm": 1.185020923614502, "learning_rate": 3.906688883599143e-05, "loss": 2.0155256271362303, "step": 16920 }, { "epoch": 2.416859987865377, "grad_norm": 1.1590262651443481, "learning_rate": 3.89716734110926e-05, "loss": 1.9501569747924805, "step": 16930 }, { "epoch": 2.418287590563546, "grad_norm": 1.3105753660202026, "learning_rate": 3.887645798619377e-05, "loss": 1.8693641662597655, "step": 16940 }, { "epoch": 2.419715193261715, "grad_norm": 1.5179659128189087, "learning_rate": 3.8781242561294936e-05, "loss": 2.0016572952270506, "step": 16950 }, { "epoch": 2.4211427959598844, "grad_norm": 1.035878300666809, "learning_rate": 3.86860271363961e-05, "loss": 2.1462406158447265, "step": 16960 }, { "epoch": 2.4225703986580536, "grad_norm": 1.367344617843628, "learning_rate": 3.859081171149727e-05, "loss": 2.032689094543457, "step": 16970 }, { "epoch": 2.423998001356223, "grad_norm": 1.220066785812378, "learning_rate": 3.849559628659843e-05, "loss": 2.060801696777344, "step": 16980 }, { "epoch": 2.4254256040543916, "grad_norm": 1.2241290807724, "learning_rate": 3.84003808616996e-05, "loss": 2.085407257080078, "step": 16990 }, { "epoch": 2.4268532067525608, "grad_norm": 1.594988226890564, "learning_rate": 3.830516543680077e-05, "loss": 2.0657268524169923, "step": 17000 }, { "epoch": 2.42828080945073, "grad_norm": 1.0305668115615845, "learning_rate": 3.820995001190193e-05, "loss": 2.1531558990478517, "step": 17010 }, { "epoch": 2.4297084121488988, "grad_norm": 1.217963457107544, "learning_rate": 3.81147345870031e-05, "loss": 2.0079486846923826, "step": 17020 }, { "epoch": 2.431136014847068, "grad_norm": 1.269824743270874, "learning_rate": 3.801951916210426e-05, "loss": 1.9473722457885743, "step": 17030 }, { "epoch": 2.432563617545237, "grad_norm": 1.0613973140716553, "learning_rate": 3.792430373720543e-05, "loss": 1.8689788818359374, "step": 17040 }, { "epoch": 2.4339912202434064, "grad_norm": 1.1870954036712646, "learning_rate": 3.782908831230659e-05, "loss": 1.8570318222045898, "step": 17050 }, { "epoch": 2.435418822941575, "grad_norm": 1.2225241661071777, "learning_rate": 3.773387288740776e-05, "loss": 1.9104736328125, "step": 17060 }, { "epoch": 2.4368464256397444, "grad_norm": 1.0518032312393188, "learning_rate": 3.763865746250893e-05, "loss": 1.987666130065918, "step": 17070 }, { "epoch": 2.4382740283379136, "grad_norm": 1.1930296421051025, "learning_rate": 3.754344203761009e-05, "loss": 2.010491943359375, "step": 17080 }, { "epoch": 2.439701631036083, "grad_norm": 1.1634668111801147, "learning_rate": 3.744822661271126e-05, "loss": 2.0902240753173826, "step": 17090 }, { "epoch": 2.4411292337342516, "grad_norm": 1.1710582971572876, "learning_rate": 3.735301118781243e-05, "loss": 1.9754999160766602, "step": 17100 }, { "epoch": 2.442556836432421, "grad_norm": 1.1441600322723389, "learning_rate": 3.72577957629136e-05, "loss": 2.0302724838256836, "step": 17110 }, { "epoch": 2.44398443913059, "grad_norm": 1.4251461029052734, "learning_rate": 3.716258033801476e-05, "loss": 2.0605106353759766, "step": 17120 }, { "epoch": 2.4454120418287593, "grad_norm": 1.1793793439865112, "learning_rate": 3.706736491311593e-05, "loss": 2.092927932739258, "step": 17130 }, { "epoch": 2.446839644526928, "grad_norm": 1.4637576341629028, "learning_rate": 3.69721494882171e-05, "loss": 1.9443946838378907, "step": 17140 }, { "epoch": 2.4482672472250973, "grad_norm": 1.1322544813156128, "learning_rate": 3.687693406331826e-05, "loss": 2.0062938690185548, "step": 17150 }, { "epoch": 2.4496948499232665, "grad_norm": 1.5589368343353271, "learning_rate": 3.678171863841943e-05, "loss": 2.0394630432128906, "step": 17160 }, { "epoch": 2.4511224526214352, "grad_norm": 1.1795375347137451, "learning_rate": 3.668650321352059e-05, "loss": 2.1672014236450194, "step": 17170 }, { "epoch": 2.4525500553196045, "grad_norm": 1.2432420253753662, "learning_rate": 3.659128778862176e-05, "loss": 2.0368724822998048, "step": 17180 }, { "epoch": 2.4539776580177737, "grad_norm": 1.3144277334213257, "learning_rate": 3.649607236372293e-05, "loss": 1.9938819885253907, "step": 17190 }, { "epoch": 2.455405260715943, "grad_norm": 1.0972892045974731, "learning_rate": 3.640085693882409e-05, "loss": 2.0603214263916017, "step": 17200 }, { "epoch": 2.4568328634141117, "grad_norm": 1.4193639755249023, "learning_rate": 3.630564151392526e-05, "loss": 2.090998077392578, "step": 17210 }, { "epoch": 2.458260466112281, "grad_norm": 1.3886083364486694, "learning_rate": 3.621042608902642e-05, "loss": 1.9233469009399413, "step": 17220 }, { "epoch": 2.45968806881045, "grad_norm": 1.0662388801574707, "learning_rate": 3.611521066412759e-05, "loss": 1.858973503112793, "step": 17230 }, { "epoch": 2.4611156715086193, "grad_norm": 1.070970892906189, "learning_rate": 3.601999523922875e-05, "loss": 2.0361900329589844, "step": 17240 }, { "epoch": 2.462543274206788, "grad_norm": 1.6537975072860718, "learning_rate": 3.592477981432992e-05, "loss": 1.9845027923583984, "step": 17250 }, { "epoch": 2.4639708769049573, "grad_norm": 1.500212550163269, "learning_rate": 3.582956438943109e-05, "loss": 1.9468679428100586, "step": 17260 }, { "epoch": 2.4653984796031265, "grad_norm": 1.4318137168884277, "learning_rate": 3.573434896453225e-05, "loss": 2.0817916870117186, "step": 17270 }, { "epoch": 2.4668260823012957, "grad_norm": 1.6362388134002686, "learning_rate": 3.563913353963342e-05, "loss": 1.878277587890625, "step": 17280 }, { "epoch": 2.4682536849994645, "grad_norm": 1.2536224126815796, "learning_rate": 3.554391811473459e-05, "loss": 1.92974853515625, "step": 17290 }, { "epoch": 2.4696812876976337, "grad_norm": 1.2624261379241943, "learning_rate": 3.544870268983576e-05, "loss": 2.1186782836914064, "step": 17300 }, { "epoch": 2.471108890395803, "grad_norm": 1.4837695360183716, "learning_rate": 3.536300880742681e-05, "loss": 2.0195901870727537, "step": 17310 }, { "epoch": 2.4725364930939717, "grad_norm": 1.159637689590454, "learning_rate": 3.526779338252797e-05, "loss": 2.0951045989990233, "step": 17320 }, { "epoch": 2.473964095792141, "grad_norm": 1.3759384155273438, "learning_rate": 3.517257795762914e-05, "loss": 1.849898910522461, "step": 17330 }, { "epoch": 2.47539169849031, "grad_norm": 1.2705048322677612, "learning_rate": 3.50773625327303e-05, "loss": 1.9654624938964844, "step": 17340 }, { "epoch": 2.4768193011884794, "grad_norm": 1.3250294923782349, "learning_rate": 3.498214710783147e-05, "loss": 2.045675277709961, "step": 17350 }, { "epoch": 2.4782469038866486, "grad_norm": 1.1377125978469849, "learning_rate": 3.488693168293264e-05, "loss": 1.9472467422485351, "step": 17360 }, { "epoch": 2.4796745065848174, "grad_norm": 1.2754701375961304, "learning_rate": 3.47917162580338e-05, "loss": 2.0088281631469727, "step": 17370 }, { "epoch": 2.4811021092829866, "grad_norm": 1.3930072784423828, "learning_rate": 3.469650083313497e-05, "loss": 2.044053649902344, "step": 17380 }, { "epoch": 2.482529711981156, "grad_norm": 1.0165702104568481, "learning_rate": 3.460128540823613e-05, "loss": 2.0746109008789064, "step": 17390 }, { "epoch": 2.4839573146793246, "grad_norm": 1.4228368997573853, "learning_rate": 3.45060699833373e-05, "loss": 2.0027420043945314, "step": 17400 }, { "epoch": 2.485384917377494, "grad_norm": 1.163480520248413, "learning_rate": 3.441085455843847e-05, "loss": 2.0341724395751952, "step": 17410 }, { "epoch": 2.486812520075663, "grad_norm": 1.0603508949279785, "learning_rate": 3.431563913353964e-05, "loss": 1.997540283203125, "step": 17420 }, { "epoch": 2.488240122773832, "grad_norm": 1.1408419609069824, "learning_rate": 3.4220423708640806e-05, "loss": 1.8485952377319337, "step": 17430 }, { "epoch": 2.489667725472001, "grad_norm": 1.2670921087265015, "learning_rate": 3.412520828374197e-05, "loss": 1.8539644241333009, "step": 17440 }, { "epoch": 2.49109532817017, "grad_norm": 1.0081952810287476, "learning_rate": 3.402999285884314e-05, "loss": 2.049833869934082, "step": 17450 }, { "epoch": 2.4925229308683394, "grad_norm": 1.03472101688385, "learning_rate": 3.39347774339443e-05, "loss": 1.9517929077148437, "step": 17460 }, { "epoch": 2.493950533566508, "grad_norm": 1.1816186904907227, "learning_rate": 3.383956200904547e-05, "loss": 1.9857969284057617, "step": 17470 }, { "epoch": 2.4953781362646774, "grad_norm": 1.2527345418930054, "learning_rate": 3.374434658414664e-05, "loss": 2.050155448913574, "step": 17480 }, { "epoch": 2.4968057389628466, "grad_norm": 1.1800050735473633, "learning_rate": 3.36491311592478e-05, "loss": 2.008490562438965, "step": 17490 }, { "epoch": 2.498233341661016, "grad_norm": 1.233428716659546, "learning_rate": 3.355391573434897e-05, "loss": 2.0124032974243162, "step": 17500 }, { "epoch": 2.499660944359185, "grad_norm": 1.0151340961456299, "learning_rate": 3.345870030945013e-05, "loss": 1.9919218063354491, "step": 17510 }, { "epoch": 2.501088547057354, "grad_norm": 1.4945932626724243, "learning_rate": 3.33634848845513e-05, "loss": 2.0871099472045898, "step": 17520 }, { "epoch": 2.502516149755523, "grad_norm": 1.1304422616958618, "learning_rate": 3.326826945965247e-05, "loss": 2.0538097381591798, "step": 17530 }, { "epoch": 2.5039437524536923, "grad_norm": 1.1392947435379028, "learning_rate": 3.317305403475363e-05, "loss": 1.8372905731201172, "step": 17540 }, { "epoch": 2.505371355151861, "grad_norm": 1.119337558746338, "learning_rate": 3.30778386098548e-05, "loss": 1.9661142349243164, "step": 17550 }, { "epoch": 2.5067989578500303, "grad_norm": 1.0007461309432983, "learning_rate": 3.298262318495596e-05, "loss": 2.0112192153930666, "step": 17560 }, { "epoch": 2.5082265605481995, "grad_norm": 1.4091777801513672, "learning_rate": 3.288740776005713e-05, "loss": 2.0188812255859374, "step": 17570 }, { "epoch": 2.5096541632463687, "grad_norm": 1.3921597003936768, "learning_rate": 3.279219233515829e-05, "loss": 2.0135860443115234, "step": 17580 }, { "epoch": 2.5110817659445375, "grad_norm": 1.1862977743148804, "learning_rate": 3.269697691025946e-05, "loss": 2.0398178100585938, "step": 17590 }, { "epoch": 2.5125093686427067, "grad_norm": 1.2714383602142334, "learning_rate": 3.260176148536063e-05, "loss": 2.0710641860961916, "step": 17600 }, { "epoch": 2.513936971340876, "grad_norm": 1.4570212364196777, "learning_rate": 3.25065460604618e-05, "loss": 2.118893623352051, "step": 17610 }, { "epoch": 2.5153645740390447, "grad_norm": 1.134436011314392, "learning_rate": 3.241133063556297e-05, "loss": 2.111136627197266, "step": 17620 }, { "epoch": 2.516792176737214, "grad_norm": 1.1267801523208618, "learning_rate": 3.231611521066413e-05, "loss": 1.8919004440307616, "step": 17630 }, { "epoch": 2.518219779435383, "grad_norm": 1.1495115756988525, "learning_rate": 3.22208997857653e-05, "loss": 2.0751272201538087, "step": 17640 }, { "epoch": 2.5196473821335523, "grad_norm": 1.189489722251892, "learning_rate": 3.212568436086646e-05, "loss": 2.032504081726074, "step": 17650 }, { "epoch": 2.5210749848317215, "grad_norm": 1.1777324676513672, "learning_rate": 3.203046893596763e-05, "loss": 1.9300954818725586, "step": 17660 }, { "epoch": 2.5225025875298903, "grad_norm": 1.127679467201233, "learning_rate": 3.19352535110688e-05, "loss": 1.9205377578735352, "step": 17670 }, { "epoch": 2.5239301902280595, "grad_norm": 1.040206789970398, "learning_rate": 3.184003808616996e-05, "loss": 1.9364898681640625, "step": 17680 }, { "epoch": 2.5253577929262288, "grad_norm": 1.3325891494750977, "learning_rate": 3.174482266127113e-05, "loss": 2.017487144470215, "step": 17690 }, { "epoch": 2.5267853956243975, "grad_norm": 1.1630598306655884, "learning_rate": 3.164960723637229e-05, "loss": 2.0031011581420897, "step": 17700 }, { "epoch": 2.5282129983225667, "grad_norm": 1.1594054698944092, "learning_rate": 3.155439181147346e-05, "loss": 2.021126937866211, "step": 17710 }, { "epoch": 2.529640601020736, "grad_norm": 1.2587190866470337, "learning_rate": 3.145917638657463e-05, "loss": 1.967803382873535, "step": 17720 }, { "epoch": 2.531068203718905, "grad_norm": 1.3349497318267822, "learning_rate": 3.136396096167579e-05, "loss": 1.9792015075683593, "step": 17730 }, { "epoch": 2.5324958064170744, "grad_norm": 1.209916591644287, "learning_rate": 3.126874553677696e-05, "loss": 2.048355484008789, "step": 17740 }, { "epoch": 2.533923409115243, "grad_norm": 1.2659903764724731, "learning_rate": 3.117353011187812e-05, "loss": 2.091727066040039, "step": 17750 }, { "epoch": 2.5353510118134124, "grad_norm": 1.3691976070404053, "learning_rate": 3.107831468697929e-05, "loss": 2.101253128051758, "step": 17760 }, { "epoch": 2.536778614511581, "grad_norm": 1.3626141548156738, "learning_rate": 3.098309926208045e-05, "loss": 2.00161190032959, "step": 17770 }, { "epoch": 2.5382062172097504, "grad_norm": 1.127090573310852, "learning_rate": 3.088788383718162e-05, "loss": 2.019171714782715, "step": 17780 }, { "epoch": 2.5396338199079196, "grad_norm": 1.4331122636795044, "learning_rate": 3.079266841228279e-05, "loss": 1.945430374145508, "step": 17790 }, { "epoch": 2.541061422606089, "grad_norm": 1.3982806205749512, "learning_rate": 3.069745298738396e-05, "loss": 1.9917106628417969, "step": 17800 }, { "epoch": 2.542489025304258, "grad_norm": 1.1267220973968506, "learning_rate": 3.060223756248513e-05, "loss": 1.875246238708496, "step": 17810 }, { "epoch": 2.543916628002427, "grad_norm": 1.2638752460479736, "learning_rate": 3.0507022137586287e-05, "loss": 2.0140647888183594, "step": 17820 }, { "epoch": 2.545344230700596, "grad_norm": 1.3092056512832642, "learning_rate": 3.0411806712687456e-05, "loss": 1.8991128921508789, "step": 17830 }, { "epoch": 2.5467718333987652, "grad_norm": 1.1748132705688477, "learning_rate": 3.031659128778862e-05, "loss": 1.922712516784668, "step": 17840 }, { "epoch": 2.548199436096934, "grad_norm": 1.1413722038269043, "learning_rate": 3.022137586288979e-05, "loss": 2.0847362518310546, "step": 17850 }, { "epoch": 2.549627038795103, "grad_norm": 1.1587178707122803, "learning_rate": 3.012616043799096e-05, "loss": 2.02646427154541, "step": 17860 }, { "epoch": 2.5510546414932724, "grad_norm": 1.2286187410354614, "learning_rate": 3.003094501309212e-05, "loss": 2.1029518127441404, "step": 17870 }, { "epoch": 2.5524822441914417, "grad_norm": 1.4867658615112305, "learning_rate": 2.993572958819329e-05, "loss": 2.045925521850586, "step": 17880 }, { "epoch": 2.553909846889611, "grad_norm": 1.4716013669967651, "learning_rate": 2.9840514163294452e-05, "loss": 1.93184814453125, "step": 17890 }, { "epoch": 2.5553374495877796, "grad_norm": 1.2011888027191162, "learning_rate": 2.974529873839562e-05, "loss": 1.993921661376953, "step": 17900 }, { "epoch": 2.556765052285949, "grad_norm": 1.272960901260376, "learning_rate": 2.965008331349679e-05, "loss": 2.0409687042236326, "step": 17910 }, { "epoch": 2.558192654984118, "grad_norm": 1.1658484935760498, "learning_rate": 2.9554867888597952e-05, "loss": 1.9747116088867187, "step": 17920 }, { "epoch": 2.559620257682287, "grad_norm": 1.1529005765914917, "learning_rate": 2.945965246369912e-05, "loss": 1.9707902908325194, "step": 17930 }, { "epoch": 2.561047860380456, "grad_norm": 1.0655295848846436, "learning_rate": 2.9364437038800286e-05, "loss": 2.042531967163086, "step": 17940 }, { "epoch": 2.5624754630786253, "grad_norm": 1.1186739206314087, "learning_rate": 2.9269221613901455e-05, "loss": 2.053073692321777, "step": 17950 }, { "epoch": 2.5639030657767945, "grad_norm": 1.2072707414627075, "learning_rate": 2.9174006189002617e-05, "loss": 2.2364439010620116, "step": 17960 }, { "epoch": 2.5653306684749633, "grad_norm": 1.0232774019241333, "learning_rate": 2.9078790764103786e-05, "loss": 2.0129594802856445, "step": 17970 }, { "epoch": 2.5667582711731325, "grad_norm": 1.6844696998596191, "learning_rate": 2.8983575339204955e-05, "loss": 1.9951644897460938, "step": 17980 }, { "epoch": 2.5681858738713017, "grad_norm": 1.6122374534606934, "learning_rate": 2.8888359914306117e-05, "loss": 1.9501501083374024, "step": 17990 }, { "epoch": 2.5696134765694705, "grad_norm": 1.1397274732589722, "learning_rate": 2.8793144489407286e-05, "loss": 2.0303413391113283, "step": 18000 }, { "epoch": 2.5710410792676397, "grad_norm": 1.109545111656189, "learning_rate": 2.8697929064508448e-05, "loss": 1.9894460678100585, "step": 18010 }, { "epoch": 2.572468681965809, "grad_norm": 1.1957917213439941, "learning_rate": 2.8602713639609617e-05, "loss": 2.1261323928833007, "step": 18020 }, { "epoch": 2.573896284663978, "grad_norm": 1.2008740901947021, "learning_rate": 2.8507498214710786e-05, "loss": 1.9992883682250977, "step": 18030 }, { "epoch": 2.5753238873621473, "grad_norm": 1.0467209815979004, "learning_rate": 2.841228278981195e-05, "loss": 1.94007568359375, "step": 18040 }, { "epoch": 2.576751490060316, "grad_norm": 1.3789418935775757, "learning_rate": 2.831706736491312e-05, "loss": 2.079054069519043, "step": 18050 }, { "epoch": 2.5781790927584853, "grad_norm": 1.3536559343338013, "learning_rate": 2.8221851940014282e-05, "loss": 1.8747962951660155, "step": 18060 }, { "epoch": 2.5796066954566546, "grad_norm": 1.0951052904129028, "learning_rate": 2.812663651511545e-05, "loss": 1.9480432510375976, "step": 18070 }, { "epoch": 2.5810342981548233, "grad_norm": 1.3351942300796509, "learning_rate": 2.8031421090216613e-05, "loss": 2.014860725402832, "step": 18080 }, { "epoch": 2.5824619008529925, "grad_norm": 0.9903519153594971, "learning_rate": 2.7936205665317782e-05, "loss": 2.173599433898926, "step": 18090 }, { "epoch": 2.5838895035511618, "grad_norm": 1.244118332862854, "learning_rate": 2.784099024041895e-05, "loss": 1.9678041458129882, "step": 18100 }, { "epoch": 2.585317106249331, "grad_norm": 1.0506081581115723, "learning_rate": 2.7745774815520113e-05, "loss": 2.0645706176757814, "step": 18110 }, { "epoch": 2.5867447089474997, "grad_norm": 1.3029327392578125, "learning_rate": 2.7650559390621282e-05, "loss": 2.064493942260742, "step": 18120 }, { "epoch": 2.588172311645669, "grad_norm": 1.3948699235916138, "learning_rate": 2.7555343965722448e-05, "loss": 1.9093946456909179, "step": 18130 }, { "epoch": 2.589599914343838, "grad_norm": 1.4485396146774292, "learning_rate": 2.7460128540823616e-05, "loss": 2.0825784683227537, "step": 18140 }, { "epoch": 2.591027517042007, "grad_norm": 1.4281233549118042, "learning_rate": 2.736491311592478e-05, "loss": 1.893547821044922, "step": 18150 }, { "epoch": 2.592455119740176, "grad_norm": 1.2344847917556763, "learning_rate": 2.7269697691025947e-05, "loss": 2.0142820358276365, "step": 18160 }, { "epoch": 2.5938827224383454, "grad_norm": 1.4939284324645996, "learning_rate": 2.7174482266127116e-05, "loss": 1.9516475677490235, "step": 18170 }, { "epoch": 2.5953103251365146, "grad_norm": 1.3290959596633911, "learning_rate": 2.707926684122828e-05, "loss": 1.8538034439086915, "step": 18180 }, { "epoch": 2.596737927834684, "grad_norm": 1.1292997598648071, "learning_rate": 2.6984051416329447e-05, "loss": 2.128860855102539, "step": 18190 }, { "epoch": 2.5981655305328526, "grad_norm": 1.250159740447998, "learning_rate": 2.688883599143061e-05, "loss": 1.8541545867919922, "step": 18200 }, { "epoch": 2.599593133231022, "grad_norm": 1.2571525573730469, "learning_rate": 2.6793620566531778e-05, "loss": 1.902310562133789, "step": 18210 }, { "epoch": 2.601020735929191, "grad_norm": 1.347256064414978, "learning_rate": 2.6698405141632947e-05, "loss": 2.086398887634277, "step": 18220 }, { "epoch": 2.60244833862736, "grad_norm": 1.0514155626296997, "learning_rate": 2.6603189716734113e-05, "loss": 2.078104782104492, "step": 18230 }, { "epoch": 2.603875941325529, "grad_norm": 1.0501142740249634, "learning_rate": 2.650797429183528e-05, "loss": 2.026015853881836, "step": 18240 }, { "epoch": 2.6053035440236982, "grad_norm": 1.4204063415527344, "learning_rate": 2.6412758866936443e-05, "loss": 2.104779815673828, "step": 18250 }, { "epoch": 2.6067311467218675, "grad_norm": 1.2805187702178955, "learning_rate": 2.6317543442037612e-05, "loss": 2.15667610168457, "step": 18260 }, { "epoch": 2.6081587494200367, "grad_norm": 2.0864548683166504, "learning_rate": 2.6222328017138774e-05, "loss": 1.9995121002197265, "step": 18270 }, { "epoch": 2.6095863521182054, "grad_norm": 1.621245265007019, "learning_rate": 2.6127112592239943e-05, "loss": 1.966802215576172, "step": 18280 }, { "epoch": 2.6110139548163747, "grad_norm": 1.6900864839553833, "learning_rate": 2.6031897167341112e-05, "loss": 1.9888671875, "step": 18290 }, { "epoch": 2.6124415575145434, "grad_norm": 1.1354056596755981, "learning_rate": 2.5936681742442274e-05, "loss": 2.040275001525879, "step": 18300 }, { "epoch": 2.6138691602127127, "grad_norm": 1.2948439121246338, "learning_rate": 2.5841466317543443e-05, "loss": 1.999903678894043, "step": 18310 }, { "epoch": 2.615296762910882, "grad_norm": 1.152174949645996, "learning_rate": 2.574625089264461e-05, "loss": 1.9197015762329102, "step": 18320 }, { "epoch": 2.616724365609051, "grad_norm": 1.1372098922729492, "learning_rate": 2.5651035467745778e-05, "loss": 1.9934078216552735, "step": 18330 }, { "epoch": 2.6181519683072203, "grad_norm": 1.0624325275421143, "learning_rate": 2.5555820042846946e-05, "loss": 1.8943883895874023, "step": 18340 }, { "epoch": 2.619579571005389, "grad_norm": 1.1142698526382446, "learning_rate": 2.546060461794811e-05, "loss": 1.9544689178466796, "step": 18350 }, { "epoch": 2.6210071737035583, "grad_norm": 1.1732228994369507, "learning_rate": 2.5365389193049277e-05, "loss": 2.1579370498657227, "step": 18360 }, { "epoch": 2.6224347764017275, "grad_norm": 1.1115386486053467, "learning_rate": 2.527017376815044e-05, "loss": 2.0285175323486326, "step": 18370 }, { "epoch": 2.6238623790998963, "grad_norm": 1.153415322303772, "learning_rate": 2.517495834325161e-05, "loss": 1.9277423858642577, "step": 18380 }, { "epoch": 2.6252899817980655, "grad_norm": 1.0986707210540771, "learning_rate": 2.5079742918352774e-05, "loss": 2.013093185424805, "step": 18390 }, { "epoch": 2.6267175844962347, "grad_norm": 1.031003713607788, "learning_rate": 2.498452749345394e-05, "loss": 1.9191864013671875, "step": 18400 } ], "logging_steps": 10, "max_steps": 21015, "num_input_tokens_seen": 0, "num_train_epochs": 3, "save_steps": 200, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": false }, "attributes": {} } }, "total_flos": 7.684039394096947e+17, "train_batch_size": 2, "trial_name": null, "trial_params": null }