{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 0.2844545583842981, "eval_steps": 500, "global_step": 4000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.0003555681979803726, "grad_norm": 2.922154664993286, "learning_rate": 2.844950213371266e-07, "loss": 3.1607500076293946, "step": 5 }, { "epoch": 0.0007111363959607452, "grad_norm": 2.788900375366211, "learning_rate": 6.401137980085349e-07, "loss": 3.0276002883911133, "step": 10 }, { "epoch": 0.001066704593941118, "grad_norm": 2.4744274616241455, "learning_rate": 9.95732574679943e-07, "loss": 2.9341686248779295, "step": 15 }, { "epoch": 0.0014222727919214904, "grad_norm": 2.5859501361846924, "learning_rate": 1.3513513513513515e-06, "loss": 2.816766357421875, "step": 20 }, { "epoch": 0.001777840989901863, "grad_norm": 2.584510564804077, "learning_rate": 1.7069701280227598e-06, "loss": 2.9872323989868166, "step": 25 }, { "epoch": 0.002133409187882236, "grad_norm": 2.135786294937134, "learning_rate": 2.062588904694168e-06, "loss": 2.671082305908203, "step": 30 }, { "epoch": 0.0024889773858626086, "grad_norm": 2.6521804332733154, "learning_rate": 2.4182076813655762e-06, "loss": 2.9811405181884765, "step": 35 }, { "epoch": 0.002844545583842981, "grad_norm": 2.386695623397827, "learning_rate": 2.7738264580369845e-06, "loss": 2.935878372192383, "step": 40 }, { "epoch": 0.0032001137818233535, "grad_norm": 1.7497117519378662, "learning_rate": 3.1294452347083927e-06, "loss": 2.659752082824707, "step": 45 }, { "epoch": 0.003555681979803726, "grad_norm": 1.8349881172180176, "learning_rate": 3.485064011379801e-06, "loss": 2.6917171478271484, "step": 50 }, { "epoch": 0.003911250177784099, "grad_norm": 1.1846299171447754, "learning_rate": 3.840682788051209e-06, "loss": 2.6170494079589846, "step": 55 }, { "epoch": 0.004266818375764472, "grad_norm": 1.2160366773605347, "learning_rate": 4.1963015647226175e-06, "loss": 2.544540786743164, "step": 60 }, { "epoch": 0.004622386573744844, "grad_norm": 0.9632154703140259, "learning_rate": 4.551920341394026e-06, "loss": 2.511016845703125, "step": 65 }, { "epoch": 0.004977954771725217, "grad_norm": 0.8948017358779907, "learning_rate": 4.907539118065434e-06, "loss": 2.5060163497924806, "step": 70 }, { "epoch": 0.0053335229697055895, "grad_norm": 0.7233667373657227, "learning_rate": 5.263157894736842e-06, "loss": 2.3816350936889648, "step": 75 }, { "epoch": 0.005689091167685962, "grad_norm": 0.5045154094696045, "learning_rate": 5.6187766714082505e-06, "loss": 2.0663467407226563, "step": 80 }, { "epoch": 0.006044659365666335, "grad_norm": 0.581486701965332, "learning_rate": 5.974395448079659e-06, "loss": 2.256585884094238, "step": 85 }, { "epoch": 0.006400227563646707, "grad_norm": 0.5268262028694153, "learning_rate": 6.330014224751067e-06, "loss": 2.3062728881835937, "step": 90 }, { "epoch": 0.00675579576162708, "grad_norm": 0.5389167070388794, "learning_rate": 6.685633001422475e-06, "loss": 2.3067771911621096, "step": 95 }, { "epoch": 0.007111363959607452, "grad_norm": 0.4723075330257416, "learning_rate": 7.0412517780938835e-06, "loss": 2.2483678817749024, "step": 100 }, { "epoch": 0.0074669321575878255, "grad_norm": 0.4904673099517822, "learning_rate": 7.3968705547652926e-06, "loss": 2.143675994873047, "step": 105 }, { "epoch": 0.007822500355568198, "grad_norm": 0.4975234270095825, "learning_rate": 7.7524893314367e-06, "loss": 2.2454158782958986, "step": 110 }, { "epoch": 0.00817806855354857, "grad_norm": 0.4857669472694397, "learning_rate": 8.108108108108109e-06, "loss": 2.122819519042969, "step": 115 }, { "epoch": 0.008533636751528944, "grad_norm": 0.47082898020744324, "learning_rate": 8.463726884779518e-06, "loss": 2.2155166625976563, "step": 120 }, { "epoch": 0.008889204949509316, "grad_norm": 1.5569114685058594, "learning_rate": 8.819345661450926e-06, "loss": 2.1064226150512697, "step": 125 }, { "epoch": 0.009244773147489688, "grad_norm": 0.5040918588638306, "learning_rate": 9.174964438122333e-06, "loss": 2.1626338958740234, "step": 130 }, { "epoch": 0.00960034134547006, "grad_norm": 0.3950935900211334, "learning_rate": 9.53058321479374e-06, "loss": 2.0201107025146485, "step": 135 }, { "epoch": 0.009955909543450435, "grad_norm": 0.49203363060951233, "learning_rate": 9.88620199146515e-06, "loss": 2.194349670410156, "step": 140 }, { "epoch": 0.010311477741430807, "grad_norm": 0.4899546504020691, "learning_rate": 1.0241820768136559e-05, "loss": 1.956950569152832, "step": 145 }, { "epoch": 0.010667045939411179, "grad_norm": 0.5067806243896484, "learning_rate": 1.0597439544807966e-05, "loss": 2.130311965942383, "step": 150 }, { "epoch": 0.011022614137391551, "grad_norm": 0.5043884515762329, "learning_rate": 1.0953058321479375e-05, "loss": 2.1246944427490235, "step": 155 }, { "epoch": 0.011378182335371924, "grad_norm": 0.5274961590766907, "learning_rate": 1.1308677098150782e-05, "loss": 2.0639286041259766, "step": 160 }, { "epoch": 0.011733750533352297, "grad_norm": 0.5013137459754944, "learning_rate": 1.1664295874822192e-05, "loss": 2.1149869918823243, "step": 165 }, { "epoch": 0.01208931873133267, "grad_norm": 0.5035282969474792, "learning_rate": 1.2019914651493599e-05, "loss": 2.0335208892822267, "step": 170 }, { "epoch": 0.012444886929313042, "grad_norm": 0.4603864848613739, "learning_rate": 1.2375533428165008e-05, "loss": 2.1348243713378907, "step": 175 }, { "epoch": 0.012800455127293414, "grad_norm": 0.49201518297195435, "learning_rate": 1.2731152204836415e-05, "loss": 2.1960676193237303, "step": 180 }, { "epoch": 0.013156023325273788, "grad_norm": 0.4972819983959198, "learning_rate": 1.3086770981507823e-05, "loss": 2.209041213989258, "step": 185 }, { "epoch": 0.01351159152325416, "grad_norm": 0.501181960105896, "learning_rate": 1.3442389758179234e-05, "loss": 2.1036434173583984, "step": 190 }, { "epoch": 0.013867159721234533, "grad_norm": 0.5144694447517395, "learning_rate": 1.379800853485064e-05, "loss": 2.155868721008301, "step": 195 }, { "epoch": 0.014222727919214905, "grad_norm": 0.5073087811470032, "learning_rate": 1.415362731152205e-05, "loss": 1.8681720733642577, "step": 200 }, { "epoch": 0.014578296117195279, "grad_norm": 0.5091198086738586, "learning_rate": 1.4509246088193457e-05, "loss": 2.140354347229004, "step": 205 }, { "epoch": 0.014933864315175651, "grad_norm": 0.49849700927734375, "learning_rate": 1.4864864864864867e-05, "loss": 2.1377050399780275, "step": 210 }, { "epoch": 0.015289432513156023, "grad_norm": 0.5193009972572327, "learning_rate": 1.5220483641536274e-05, "loss": 2.03880615234375, "step": 215 }, { "epoch": 0.015645000711136395, "grad_norm": 0.5600176453590393, "learning_rate": 1.5576102418207685e-05, "loss": 2.1393871307373047, "step": 220 }, { "epoch": 0.016000568909116768, "grad_norm": 0.8653586506843567, "learning_rate": 1.593172119487909e-05, "loss": 1.9054578781127929, "step": 225 }, { "epoch": 0.01635613710709714, "grad_norm": 0.5116851329803467, "learning_rate": 1.62873399715505e-05, "loss": 2.0144094467163085, "step": 230 }, { "epoch": 0.016711705305077512, "grad_norm": 0.4923916757106781, "learning_rate": 1.6642958748221907e-05, "loss": 2.151469612121582, "step": 235 }, { "epoch": 0.017067273503057888, "grad_norm": 0.6402277946472168, "learning_rate": 1.6998577524893318e-05, "loss": 2.205842208862305, "step": 240 }, { "epoch": 0.01742284170103826, "grad_norm": 0.4952904284000397, "learning_rate": 1.7354196301564725e-05, "loss": 2.0481229782104493, "step": 245 }, { "epoch": 0.017778409899018632, "grad_norm": 0.5109221339225769, "learning_rate": 1.770981507823613e-05, "loss": 2.1811712265014647, "step": 250 }, { "epoch": 0.018133978096999005, "grad_norm": 0.5637496113777161, "learning_rate": 1.806543385490754e-05, "loss": 2.192226791381836, "step": 255 }, { "epoch": 0.018489546294979377, "grad_norm": 0.48549285531044006, "learning_rate": 1.8421052631578947e-05, "loss": 2.012406349182129, "step": 260 }, { "epoch": 0.01884511449295975, "grad_norm": 0.5244142413139343, "learning_rate": 1.8776671408250358e-05, "loss": 2.0949962615966795, "step": 265 }, { "epoch": 0.01920068269094012, "grad_norm": 0.5079764723777771, "learning_rate": 1.9132290184921766e-05, "loss": 1.9754886627197266, "step": 270 }, { "epoch": 0.019556250888920493, "grad_norm": 0.4550922214984894, "learning_rate": 1.9487908961593173e-05, "loss": 2.0664173126220704, "step": 275 }, { "epoch": 0.01991181908690087, "grad_norm": 0.5083408355712891, "learning_rate": 1.984352773826458e-05, "loss": 2.078787422180176, "step": 280 }, { "epoch": 0.02026738728488124, "grad_norm": 0.48500922322273254, "learning_rate": 2.019914651493599e-05, "loss": 2.0357854843139647, "step": 285 }, { "epoch": 0.020622955482861614, "grad_norm": 0.4803495407104492, "learning_rate": 2.05547652916074e-05, "loss": 1.9375429153442383, "step": 290 }, { "epoch": 0.020978523680841986, "grad_norm": 0.5424251556396484, "learning_rate": 2.0910384068278806e-05, "loss": 2.1656143188476564, "step": 295 }, { "epoch": 0.021334091878822358, "grad_norm": 0.48051849007606506, "learning_rate": 2.1266002844950213e-05, "loss": 2.072488212585449, "step": 300 }, { "epoch": 0.02168966007680273, "grad_norm": 0.5589402318000793, "learning_rate": 2.1621621621621624e-05, "loss": 2.125236701965332, "step": 305 }, { "epoch": 0.022045228274783103, "grad_norm": 0.5212494730949402, "learning_rate": 2.197724039829303e-05, "loss": 1.9235109329223632, "step": 310 }, { "epoch": 0.022400796472763475, "grad_norm": 0.43210336565971375, "learning_rate": 2.233285917496444e-05, "loss": 1.9592742919921875, "step": 315 }, { "epoch": 0.022756364670743847, "grad_norm": 0.7312160730361938, "learning_rate": 2.2688477951635846e-05, "loss": 2.0755807876586916, "step": 320 }, { "epoch": 0.023111932868724223, "grad_norm": 1.7911994457244873, "learning_rate": 2.3044096728307254e-05, "loss": 2.0784723281860353, "step": 325 }, { "epoch": 0.023467501066704595, "grad_norm": 0.5632429122924805, "learning_rate": 2.3399715504978665e-05, "loss": 2.0193042755126953, "step": 330 }, { "epoch": 0.023823069264684967, "grad_norm": 0.7258956432342529, "learning_rate": 2.3755334281650072e-05, "loss": 2.030779266357422, "step": 335 }, { "epoch": 0.02417863746266534, "grad_norm": 0.5409282445907593, "learning_rate": 2.411095305832148e-05, "loss": 2.081908416748047, "step": 340 }, { "epoch": 0.02453420566064571, "grad_norm": 0.4613751471042633, "learning_rate": 2.4466571834992887e-05, "loss": 1.9870058059692384, "step": 345 }, { "epoch": 0.024889773858626084, "grad_norm": 0.479413241147995, "learning_rate": 2.4822190611664297e-05, "loss": 1.7315523147583007, "step": 350 }, { "epoch": 0.025245342056606456, "grad_norm": 0.5492660403251648, "learning_rate": 2.5177809388335705e-05, "loss": 2.080055999755859, "step": 355 }, { "epoch": 0.02560091025458683, "grad_norm": 0.5154370665550232, "learning_rate": 2.5533428165007112e-05, "loss": 2.0674570083618162, "step": 360 }, { "epoch": 0.025956478452567204, "grad_norm": 0.537121057510376, "learning_rate": 2.5889046941678523e-05, "loss": 1.8924758911132813, "step": 365 }, { "epoch": 0.026312046650547576, "grad_norm": 0.5539237260818481, "learning_rate": 2.624466571834993e-05, "loss": 2.162744903564453, "step": 370 }, { "epoch": 0.02666761484852795, "grad_norm": 0.5274736881256104, "learning_rate": 2.6600284495021338e-05, "loss": 2.158771514892578, "step": 375 }, { "epoch": 0.02702318304650832, "grad_norm": 0.7239411473274231, "learning_rate": 2.6955903271692745e-05, "loss": 2.1109287261962892, "step": 380 }, { "epoch": 0.027378751244488693, "grad_norm": 0.5277944803237915, "learning_rate": 2.7311522048364156e-05, "loss": 2.0311857223510743, "step": 385 }, { "epoch": 0.027734319442469065, "grad_norm": 0.5385560989379883, "learning_rate": 2.7667140825035563e-05, "loss": 1.9783439636230469, "step": 390 }, { "epoch": 0.028089887640449437, "grad_norm": 0.4753340780735016, "learning_rate": 2.802275960170697e-05, "loss": 1.9922258377075195, "step": 395 }, { "epoch": 0.02844545583842981, "grad_norm": 0.5077670812606812, "learning_rate": 2.8378378378378378e-05, "loss": 1.9716501235961914, "step": 400 }, { "epoch": 0.028801024036410182, "grad_norm": 0.5184553265571594, "learning_rate": 2.8733997155049786e-05, "loss": 2.136305046081543, "step": 405 }, { "epoch": 0.029156592234390558, "grad_norm": 0.5640929341316223, "learning_rate": 2.9089615931721196e-05, "loss": 2.075672149658203, "step": 410 }, { "epoch": 0.02951216043237093, "grad_norm": 0.5448623299598694, "learning_rate": 2.9445234708392604e-05, "loss": 2.0514604568481447, "step": 415 }, { "epoch": 0.029867728630351302, "grad_norm": 0.4612126350402832, "learning_rate": 2.980085348506401e-05, "loss": 1.9871145248413087, "step": 420 }, { "epoch": 0.030223296828331674, "grad_norm": 0.47554561495780945, "learning_rate": 3.015647226173542e-05, "loss": 1.982534408569336, "step": 425 }, { "epoch": 0.030578865026312047, "grad_norm": 0.4724675118923187, "learning_rate": 3.0512091038406833e-05, "loss": 1.9912504196166991, "step": 430 }, { "epoch": 0.03093443322429242, "grad_norm": 0.4869454801082611, "learning_rate": 3.086770981507824e-05, "loss": 1.7682329177856446, "step": 435 }, { "epoch": 0.03129000142227279, "grad_norm": 0.5002896785736084, "learning_rate": 3.1223328591749644e-05, "loss": 2.0274532318115233, "step": 440 }, { "epoch": 0.03164556962025317, "grad_norm": 0.5335668921470642, "learning_rate": 3.157894736842105e-05, "loss": 2.196476364135742, "step": 445 }, { "epoch": 0.032001137818233535, "grad_norm": 0.550162672996521, "learning_rate": 3.1934566145092466e-05, "loss": 2.053875732421875, "step": 450 }, { "epoch": 0.03235670601621391, "grad_norm": 0.5075178742408752, "learning_rate": 3.229018492176387e-05, "loss": 2.0288387298583985, "step": 455 }, { "epoch": 0.03271227421419428, "grad_norm": 0.5460106134414673, "learning_rate": 3.264580369843528e-05, "loss": 2.0439821243286134, "step": 460 }, { "epoch": 0.033067842412174656, "grad_norm": 0.4341937303543091, "learning_rate": 3.300142247510669e-05, "loss": 1.9099559783935547, "step": 465 }, { "epoch": 0.033423410610155024, "grad_norm": 0.4631098210811615, "learning_rate": 3.3357041251778095e-05, "loss": 2.0513723373413084, "step": 470 }, { "epoch": 0.0337789788081354, "grad_norm": 0.4780479669570923, "learning_rate": 3.37126600284495e-05, "loss": 1.9448671340942383, "step": 475 }, { "epoch": 0.034134547006115776, "grad_norm": 0.582769513130188, "learning_rate": 3.406827880512091e-05, "loss": 1.984304428100586, "step": 480 }, { "epoch": 0.034490115204096145, "grad_norm": 0.5180102586746216, "learning_rate": 3.442389758179232e-05, "loss": 2.0316925048828125, "step": 485 }, { "epoch": 0.03484568340207652, "grad_norm": 0.5493118166923523, "learning_rate": 3.4779516358463725e-05, "loss": 2.027867889404297, "step": 490 }, { "epoch": 0.03520125160005689, "grad_norm": 0.5218160152435303, "learning_rate": 3.513513513513514e-05, "loss": 2.096648025512695, "step": 495 }, { "epoch": 0.035556819798037265, "grad_norm": 0.4827556908130646, "learning_rate": 3.5490753911806547e-05, "loss": 1.9894039154052734, "step": 500 }, { "epoch": 0.03591238799601763, "grad_norm": 0.5301157832145691, "learning_rate": 3.5846372688477954e-05, "loss": 2.065215492248535, "step": 505 }, { "epoch": 0.03626795619399801, "grad_norm": 0.528459370136261, "learning_rate": 3.620199146514936e-05, "loss": 2.111029052734375, "step": 510 }, { "epoch": 0.036623524391978385, "grad_norm": 0.477237343788147, "learning_rate": 3.6557610241820776e-05, "loss": 2.0157339096069338, "step": 515 }, { "epoch": 0.036979092589958754, "grad_norm": 0.46244242787361145, "learning_rate": 3.6913229018492176e-05, "loss": 1.9967113494873048, "step": 520 }, { "epoch": 0.03733466078793913, "grad_norm": 0.529072105884552, "learning_rate": 3.7268847795163584e-05, "loss": 2.04681282043457, "step": 525 }, { "epoch": 0.0376902289859195, "grad_norm": 0.5005636215209961, "learning_rate": 3.762446657183499e-05, "loss": 2.1323509216308594, "step": 530 }, { "epoch": 0.038045797183899874, "grad_norm": 0.44275158643722534, "learning_rate": 3.7980085348506405e-05, "loss": 1.9433954238891602, "step": 535 }, { "epoch": 0.03840136538188024, "grad_norm": 0.5083234310150146, "learning_rate": 3.833570412517781e-05, "loss": 1.9552728652954101, "step": 540 }, { "epoch": 0.03875693357986062, "grad_norm": 0.48144620656967163, "learning_rate": 3.869132290184922e-05, "loss": 2.174291229248047, "step": 545 }, { "epoch": 0.03911250177784099, "grad_norm": 0.48507755994796753, "learning_rate": 3.904694167852063e-05, "loss": 2.0216049194335937, "step": 550 }, { "epoch": 0.03946806997582136, "grad_norm": 0.4811331629753113, "learning_rate": 3.9402560455192035e-05, "loss": 1.8616741180419922, "step": 555 }, { "epoch": 0.03982363817380174, "grad_norm": 0.4885321855545044, "learning_rate": 3.975817923186345e-05, "loss": 1.8081172943115233, "step": 560 }, { "epoch": 0.04017920637178211, "grad_norm": 0.5213722586631775, "learning_rate": 4.0113798008534856e-05, "loss": 2.0698373794555662, "step": 565 }, { "epoch": 0.04053477456976248, "grad_norm": 0.4983953535556793, "learning_rate": 4.046941678520626e-05, "loss": 1.9386020660400392, "step": 570 }, { "epoch": 0.04089034276774285, "grad_norm": 0.5289962887763977, "learning_rate": 4.0825035561877664e-05, "loss": 1.8806812286376953, "step": 575 }, { "epoch": 0.04124591096572323, "grad_norm": 0.546035885810852, "learning_rate": 4.118065433854908e-05, "loss": 1.914069366455078, "step": 580 }, { "epoch": 0.041601479163703596, "grad_norm": 0.5344237685203552, "learning_rate": 4.1536273115220486e-05, "loss": 2.0394969940185548, "step": 585 }, { "epoch": 0.04195704736168397, "grad_norm": 0.5037335157394409, "learning_rate": 4.189189189189189e-05, "loss": 2.0729978561401365, "step": 590 }, { "epoch": 0.04231261555966434, "grad_norm": 0.4325156807899475, "learning_rate": 4.22475106685633e-05, "loss": 1.8392827987670899, "step": 595 }, { "epoch": 0.042668183757644716, "grad_norm": 0.4450535476207733, "learning_rate": 4.2603129445234715e-05, "loss": 1.9253110885620117, "step": 600 }, { "epoch": 0.04302375195562509, "grad_norm": 0.47241488099098206, "learning_rate": 4.295874822190612e-05, "loss": 2.0719636917114257, "step": 605 }, { "epoch": 0.04337932015360546, "grad_norm": 0.4840798079967499, "learning_rate": 4.331436699857753e-05, "loss": 2.000336265563965, "step": 610 }, { "epoch": 0.043734888351585836, "grad_norm": 0.7127583622932434, "learning_rate": 4.366998577524894e-05, "loss": 2.096153450012207, "step": 615 }, { "epoch": 0.044090456549566205, "grad_norm": 0.4713995158672333, "learning_rate": 4.402560455192034e-05, "loss": 1.813309669494629, "step": 620 }, { "epoch": 0.04444602474754658, "grad_norm": 0.4207233190536499, "learning_rate": 4.438122332859175e-05, "loss": 1.9763273239135741, "step": 625 }, { "epoch": 0.04480159294552695, "grad_norm": 0.5664926171302795, "learning_rate": 4.473684210526316e-05, "loss": 1.884061622619629, "step": 630 }, { "epoch": 0.045157161143507325, "grad_norm": 0.44372671842575073, "learning_rate": 4.509246088193457e-05, "loss": 2.0752237319946287, "step": 635 }, { "epoch": 0.045512729341487694, "grad_norm": 0.4940667748451233, "learning_rate": 4.5448079658605974e-05, "loss": 2.0724849700927734, "step": 640 }, { "epoch": 0.04586829753946807, "grad_norm": 0.5237462520599365, "learning_rate": 4.580369843527739e-05, "loss": 2.0460742950439452, "step": 645 }, { "epoch": 0.046223865737448445, "grad_norm": 0.4766853451728821, "learning_rate": 4.6159317211948796e-05, "loss": 1.99864501953125, "step": 650 }, { "epoch": 0.046579433935428814, "grad_norm": 0.5521640181541443, "learning_rate": 4.65149359886202e-05, "loss": 2.065717887878418, "step": 655 }, { "epoch": 0.04693500213340919, "grad_norm": 0.4680851101875305, "learning_rate": 4.687055476529161e-05, "loss": 2.0925697326660155, "step": 660 }, { "epoch": 0.04729057033138956, "grad_norm": 0.48741820454597473, "learning_rate": 4.722617354196302e-05, "loss": 2.0482904434204103, "step": 665 }, { "epoch": 0.047646138529369934, "grad_norm": 0.47097286581993103, "learning_rate": 4.7581792318634425e-05, "loss": 2.05728816986084, "step": 670 }, { "epoch": 0.0480017067273503, "grad_norm": 0.47805750370025635, "learning_rate": 4.793741109530583e-05, "loss": 2.048795700073242, "step": 675 }, { "epoch": 0.04835727492533068, "grad_norm": 0.45797139406204224, "learning_rate": 4.829302987197724e-05, "loss": 1.9642038345336914, "step": 680 }, { "epoch": 0.04871284312331105, "grad_norm": 0.4881439208984375, "learning_rate": 4.8648648648648654e-05, "loss": 2.0224887847900392, "step": 685 }, { "epoch": 0.04906841132129142, "grad_norm": 0.4703167676925659, "learning_rate": 4.900426742532006e-05, "loss": 1.9867454528808595, "step": 690 }, { "epoch": 0.0494239795192718, "grad_norm": 0.4110810458660126, "learning_rate": 4.935988620199147e-05, "loss": 2.0305349349975588, "step": 695 }, { "epoch": 0.04977954771725217, "grad_norm": 0.5362057089805603, "learning_rate": 4.9715504978662876e-05, "loss": 2.0972415924072267, "step": 700 }, { "epoch": 0.05013511591523254, "grad_norm": 0.4890783131122589, "learning_rate": 4.9999996918131664e-05, "loss": 1.9155891418457032, "step": 705 }, { "epoch": 0.05049068411321291, "grad_norm": 0.4949808418750763, "learning_rate": 4.999988905281975e-05, "loss": 2.033857727050781, "step": 710 }, { "epoch": 0.05084625231119329, "grad_norm": 0.47292372584342957, "learning_rate": 4.999962709485094e-05, "loss": 1.9912883758544921, "step": 715 }, { "epoch": 0.05120182050917366, "grad_norm": 0.46866342425346375, "learning_rate": 4.99992110458399e-05, "loss": 2.0551689147949217, "step": 720 }, { "epoch": 0.05155738870715403, "grad_norm": 0.4494408965110779, "learning_rate": 4.999864090835102e-05, "loss": 1.9704553604125976, "step": 725 }, { "epoch": 0.05191295690513441, "grad_norm": 0.41759586334228516, "learning_rate": 4.999791668589848e-05, "loss": 1.8851097106933594, "step": 730 }, { "epoch": 0.05226852510311478, "grad_norm": 0.47948014736175537, "learning_rate": 4.999703838294622e-05, "loss": 2.0414073944091795, "step": 735 }, { "epoch": 0.05262409330109515, "grad_norm": 0.43024158477783203, "learning_rate": 4.999600600490783e-05, "loss": 1.9305057525634766, "step": 740 }, { "epoch": 0.05297966149907552, "grad_norm": 0.4558243155479431, "learning_rate": 4.9994819558146624e-05, "loss": 1.9468095779418946, "step": 745 }, { "epoch": 0.0533352296970559, "grad_norm": 0.4913494884967804, "learning_rate": 4.999347904997556e-05, "loss": 1.965241050720215, "step": 750 }, { "epoch": 0.053690797895036266, "grad_norm": 0.42234569787979126, "learning_rate": 4.999198448865716e-05, "loss": 1.866231346130371, "step": 755 }, { "epoch": 0.05404636609301664, "grad_norm": 0.44516387581825256, "learning_rate": 4.99903358834035e-05, "loss": 1.9495733261108399, "step": 760 }, { "epoch": 0.05440193429099701, "grad_norm": 0.41378965973854065, "learning_rate": 4.998853324437615e-05, "loss": 2.0295278549194338, "step": 765 }, { "epoch": 0.054757502488977386, "grad_norm": 0.5202167630195618, "learning_rate": 4.998657658268609e-05, "loss": 1.997533416748047, "step": 770 }, { "epoch": 0.05511307068695776, "grad_norm": 0.46458789706230164, "learning_rate": 4.998446591039367e-05, "loss": 2.0262699127197266, "step": 775 }, { "epoch": 0.05546863888493813, "grad_norm": 0.44955191016197205, "learning_rate": 4.99822012405085e-05, "loss": 1.9963245391845703, "step": 780 }, { "epoch": 0.055824207082918506, "grad_norm": 0.5131090879440308, "learning_rate": 4.9979782586989424e-05, "loss": 2.003853416442871, "step": 785 }, { "epoch": 0.056179775280898875, "grad_norm": 0.4980633556842804, "learning_rate": 4.997720996474435e-05, "loss": 1.9255821228027343, "step": 790 }, { "epoch": 0.05653534347887925, "grad_norm": 0.4826408326625824, "learning_rate": 4.997448338963026e-05, "loss": 2.1106395721435547, "step": 795 }, { "epoch": 0.05689091167685962, "grad_norm": 0.5093266367912292, "learning_rate": 4.997160287845301e-05, "loss": 2.019168472290039, "step": 800 }, { "epoch": 0.057246479874839995, "grad_norm": 0.4740906357765198, "learning_rate": 4.9968568448967334e-05, "loss": 2.107289123535156, "step": 805 }, { "epoch": 0.057602048072820364, "grad_norm": 0.48503029346466064, "learning_rate": 4.996538011987663e-05, "loss": 1.9294828414916991, "step": 810 }, { "epoch": 0.05795761627080074, "grad_norm": 0.4630905091762543, "learning_rate": 4.996203791083291e-05, "loss": 1.9798065185546876, "step": 815 }, { "epoch": 0.058313184468781115, "grad_norm": 0.44924747943878174, "learning_rate": 4.995854184243666e-05, "loss": 2.0240636825561524, "step": 820 }, { "epoch": 0.058668752666761484, "grad_norm": 0.523550808429718, "learning_rate": 4.9954891936236725e-05, "loss": 2.041210746765137, "step": 825 }, { "epoch": 0.05902432086474186, "grad_norm": 0.4703008234500885, "learning_rate": 4.9951088214730134e-05, "loss": 2.025250244140625, "step": 830 }, { "epoch": 0.05937988906272223, "grad_norm": 2.4597880840301514, "learning_rate": 4.994713070136204e-05, "loss": 2.00240421295166, "step": 835 }, { "epoch": 0.059735457260702604, "grad_norm": 0.4831807315349579, "learning_rate": 4.9943019420525476e-05, "loss": 1.9369470596313476, "step": 840 }, { "epoch": 0.06009102545868297, "grad_norm": 0.41343575716018677, "learning_rate": 4.99387543975613e-05, "loss": 1.7299930572509765, "step": 845 }, { "epoch": 0.06044659365666335, "grad_norm": 0.5090857744216919, "learning_rate": 4.993433565875797e-05, "loss": 1.9956466674804687, "step": 850 }, { "epoch": 0.06080216185464372, "grad_norm": 0.4908365309238434, "learning_rate": 4.9929763231351406e-05, "loss": 2.070937919616699, "step": 855 }, { "epoch": 0.06115773005262409, "grad_norm": 0.4700279235839844, "learning_rate": 4.992503714352485e-05, "loss": 2.0459768295288088, "step": 860 }, { "epoch": 0.06151329825060447, "grad_norm": 0.482763409614563, "learning_rate": 4.9920157424408636e-05, "loss": 2.0382041931152344, "step": 865 }, { "epoch": 0.06186886644858484, "grad_norm": 0.3896910548210144, "learning_rate": 4.991512410408006e-05, "loss": 1.9157665252685547, "step": 870 }, { "epoch": 0.06222443464656521, "grad_norm": 0.4430522322654724, "learning_rate": 4.9909937213563165e-05, "loss": 1.9655799865722656, "step": 875 }, { "epoch": 0.06258000284454558, "grad_norm": 0.484436959028244, "learning_rate": 4.990459678482856e-05, "loss": 1.934805679321289, "step": 880 }, { "epoch": 0.06293557104252595, "grad_norm": 0.45736682415008545, "learning_rate": 4.9899102850793235e-05, "loss": 1.9250009536743165, "step": 885 }, { "epoch": 0.06329113924050633, "grad_norm": 0.7695727944374084, "learning_rate": 4.989345544532031e-05, "loss": 2.0423851013183594, "step": 890 }, { "epoch": 0.0636467074384867, "grad_norm": 0.45036986470222473, "learning_rate": 4.9887654603218926e-05, "loss": 2.0039800643920898, "step": 895 }, { "epoch": 0.06400227563646707, "grad_norm": 0.44636839628219604, "learning_rate": 4.98817003602439e-05, "loss": 1.9529701232910157, "step": 900 }, { "epoch": 0.06435784383444745, "grad_norm": 0.4538932144641876, "learning_rate": 4.987559275309561e-05, "loss": 1.8466669082641602, "step": 905 }, { "epoch": 0.06471341203242782, "grad_norm": 0.5039580464363098, "learning_rate": 4.9869331819419726e-05, "loss": 1.9783496856689453, "step": 910 }, { "epoch": 0.06506898023040819, "grad_norm": 0.4641205370426178, "learning_rate": 4.986291759780696e-05, "loss": 1.9164005279541017, "step": 915 }, { "epoch": 0.06542454842838856, "grad_norm": 0.6123772263526917, "learning_rate": 4.985635012779287e-05, "loss": 1.9381017684936523, "step": 920 }, { "epoch": 0.06578011662636894, "grad_norm": 0.4578414261341095, "learning_rate": 4.9849629449857596e-05, "loss": 1.9256427764892579, "step": 925 }, { "epoch": 0.06613568482434931, "grad_norm": 0.4718715250492096, "learning_rate": 4.9842755605425615e-05, "loss": 1.9728239059448243, "step": 930 }, { "epoch": 0.06649125302232968, "grad_norm": 0.46011507511138916, "learning_rate": 4.983572863686545e-05, "loss": 2.0925949096679686, "step": 935 }, { "epoch": 0.06684682122031005, "grad_norm": 0.4530814290046692, "learning_rate": 4.9828548587489486e-05, "loss": 2.0755496978759767, "step": 940 }, { "epoch": 0.06720238941829043, "grad_norm": 0.4253125488758087, "learning_rate": 4.982121550155363e-05, "loss": 1.8711212158203125, "step": 945 }, { "epoch": 0.0675579576162708, "grad_norm": 0.46961215138435364, "learning_rate": 4.981372942425706e-05, "loss": 1.8969301223754882, "step": 950 }, { "epoch": 0.06791352581425117, "grad_norm": 0.8306499123573303, "learning_rate": 4.980609040174197e-05, "loss": 1.8298198699951171, "step": 955 }, { "epoch": 0.06826909401223155, "grad_norm": 0.47998809814453125, "learning_rate": 4.9798298481093255e-05, "loss": 1.9024595260620116, "step": 960 }, { "epoch": 0.06862466221021192, "grad_norm": 0.4352721869945526, "learning_rate": 4.979035371033824e-05, "loss": 2.0410667419433595, "step": 965 }, { "epoch": 0.06898023040819229, "grad_norm": 0.5021587610244751, "learning_rate": 4.978225613844639e-05, "loss": 1.9580207824707032, "step": 970 }, { "epoch": 0.06933579860617266, "grad_norm": 0.40722617506980896, "learning_rate": 4.977400581532896e-05, "loss": 1.974105453491211, "step": 975 }, { "epoch": 0.06969136680415304, "grad_norm": 0.42797818779945374, "learning_rate": 4.976560279183875e-05, "loss": 1.7772085189819335, "step": 980 }, { "epoch": 0.07004693500213341, "grad_norm": 0.48213788866996765, "learning_rate": 4.9757047119769774e-05, "loss": 1.8957931518554687, "step": 985 }, { "epoch": 0.07040250320011378, "grad_norm": 0.44676685333251953, "learning_rate": 4.974833885185689e-05, "loss": 1.8408065795898438, "step": 990 }, { "epoch": 0.07075807139809416, "grad_norm": 0.46073785424232483, "learning_rate": 4.973947804177556e-05, "loss": 1.998112106323242, "step": 995 }, { "epoch": 0.07111363959607453, "grad_norm": 0.3896966874599457, "learning_rate": 4.9730464744141445e-05, "loss": 1.9593812942504882, "step": 1000 }, { "epoch": 0.0714692077940549, "grad_norm": 0.5227977633476257, "learning_rate": 4.9721299014510134e-05, "loss": 2.0564823150634766, "step": 1005 }, { "epoch": 0.07182477599203527, "grad_norm": 0.42633500695228577, "learning_rate": 4.9711980909376706e-05, "loss": 1.8712087631225587, "step": 1010 }, { "epoch": 0.07218034419001565, "grad_norm": 0.42041003704071045, "learning_rate": 4.970251048617552e-05, "loss": 1.9029083251953125, "step": 1015 }, { "epoch": 0.07253591238799602, "grad_norm": 0.45724713802337646, "learning_rate": 4.969288780327971e-05, "loss": 1.882410430908203, "step": 1020 }, { "epoch": 0.07289148058597639, "grad_norm": 0.4395748972892761, "learning_rate": 4.968311292000094e-05, "loss": 1.7799474716186523, "step": 1025 }, { "epoch": 0.07324704878395677, "grad_norm": 0.4628015458583832, "learning_rate": 4.967318589658899e-05, "loss": 2.0460041046142576, "step": 1030 }, { "epoch": 0.07360261698193714, "grad_norm": 0.4735582768917084, "learning_rate": 4.966310679423139e-05, "loss": 1.8856338500976562, "step": 1035 }, { "epoch": 0.07395818517991751, "grad_norm": 0.47294700145721436, "learning_rate": 4.965287567505304e-05, "loss": 1.9967988967895507, "step": 1040 }, { "epoch": 0.07431375337789788, "grad_norm": 0.40670347213745117, "learning_rate": 4.964249260211583e-05, "loss": 1.9305591583251953, "step": 1045 }, { "epoch": 0.07466932157587826, "grad_norm": 0.4786446988582611, "learning_rate": 4.963195763941826e-05, "loss": 2.0142471313476564, "step": 1050 }, { "epoch": 0.07502488977385863, "grad_norm": 0.4656982123851776, "learning_rate": 4.962127085189503e-05, "loss": 1.931185531616211, "step": 1055 }, { "epoch": 0.075380457971839, "grad_norm": 0.44946831464767456, "learning_rate": 4.961043230541665e-05, "loss": 1.9516427993774415, "step": 1060 }, { "epoch": 0.07573602616981936, "grad_norm": 0.463571697473526, "learning_rate": 4.9599442066789035e-05, "loss": 1.838262939453125, "step": 1065 }, { "epoch": 0.07609159436779975, "grad_norm": 0.477551132440567, "learning_rate": 4.95883002037531e-05, "loss": 2.0038434982299806, "step": 1070 }, { "epoch": 0.07644716256578012, "grad_norm": 0.446843683719635, "learning_rate": 4.95770067849843e-05, "loss": 2.0295202255249025, "step": 1075 }, { "epoch": 0.07680273076376049, "grad_norm": 0.4183042049407959, "learning_rate": 4.956556188009228e-05, "loss": 1.9379022598266602, "step": 1080 }, { "epoch": 0.07715829896174087, "grad_norm": 0.412964403629303, "learning_rate": 4.9553965559620375e-05, "loss": 1.9920074462890625, "step": 1085 }, { "epoch": 0.07751386715972124, "grad_norm": 0.41583317518234253, "learning_rate": 4.9542217895045215e-05, "loss": 1.9422431945800782, "step": 1090 }, { "epoch": 0.0778694353577016, "grad_norm": 0.5088328123092651, "learning_rate": 4.9530318958776276e-05, "loss": 1.9350143432617188, "step": 1095 }, { "epoch": 0.07822500355568197, "grad_norm": 0.46786707639694214, "learning_rate": 4.951826882415544e-05, "loss": 1.9892354965209962, "step": 1100 }, { "epoch": 0.07858057175366236, "grad_norm": 0.41910895705223083, "learning_rate": 4.950606756545651e-05, "loss": 1.9484115600585938, "step": 1105 }, { "epoch": 0.07893613995164273, "grad_norm": 0.4567226767539978, "learning_rate": 4.949371525788481e-05, "loss": 1.9966516494750977, "step": 1110 }, { "epoch": 0.0792917081496231, "grad_norm": 0.46497592329978943, "learning_rate": 4.948121197757666e-05, "loss": 2.0382387161254885, "step": 1115 }, { "epoch": 0.07964727634760348, "grad_norm": 0.44980067014694214, "learning_rate": 4.946855780159897e-05, "loss": 1.8161828994750977, "step": 1120 }, { "epoch": 0.08000284454558385, "grad_norm": 0.4221165180206299, "learning_rate": 4.945575280794868e-05, "loss": 1.9043243408203125, "step": 1125 }, { "epoch": 0.08035841274356421, "grad_norm": 0.4723920524120331, "learning_rate": 4.94427970755524e-05, "loss": 2.0307872772216795, "step": 1130 }, { "epoch": 0.08071398094154458, "grad_norm": 0.46250614523887634, "learning_rate": 4.942969068426577e-05, "loss": 1.911181640625, "step": 1135 }, { "epoch": 0.08106954913952497, "grad_norm": 0.4713759422302246, "learning_rate": 4.941643371487312e-05, "loss": 2.0472402572631836, "step": 1140 }, { "epoch": 0.08142511733750533, "grad_norm": 0.4352535307407379, "learning_rate": 4.940302624908688e-05, "loss": 2.0494014739990236, "step": 1145 }, { "epoch": 0.0817806855354857, "grad_norm": 0.4320608377456665, "learning_rate": 4.938946836954709e-05, "loss": 1.965937614440918, "step": 1150 }, { "epoch": 0.08213625373346607, "grad_norm": 0.455120712518692, "learning_rate": 4.937576015982091e-05, "loss": 1.945225715637207, "step": 1155 }, { "epoch": 0.08249182193144645, "grad_norm": 0.4409791827201843, "learning_rate": 4.936190170440208e-05, "loss": 1.877567672729492, "step": 1160 }, { "epoch": 0.08284739012942682, "grad_norm": 0.4484880864620209, "learning_rate": 4.934789308871045e-05, "loss": 1.9771286010742188, "step": 1165 }, { "epoch": 0.08320295832740719, "grad_norm": 0.45543840527534485, "learning_rate": 4.933373439909139e-05, "loss": 1.9534339904785156, "step": 1170 }, { "epoch": 0.08355852652538757, "grad_norm": 0.39205965399742126, "learning_rate": 4.9319425722815284e-05, "loss": 1.9376569747924806, "step": 1175 }, { "epoch": 0.08391409472336794, "grad_norm": 0.461448073387146, "learning_rate": 4.9304967148077e-05, "loss": 2.038929748535156, "step": 1180 }, { "epoch": 0.08426966292134831, "grad_norm": 0.40936994552612305, "learning_rate": 4.929035876399535e-05, "loss": 1.7540023803710938, "step": 1185 }, { "epoch": 0.08462523111932868, "grad_norm": 0.6044595241546631, "learning_rate": 4.9275600660612515e-05, "loss": 1.8902605056762696, "step": 1190 }, { "epoch": 0.08498079931730906, "grad_norm": 0.4404708743095398, "learning_rate": 4.926069292889352e-05, "loss": 1.967193603515625, "step": 1195 }, { "epoch": 0.08533636751528943, "grad_norm": 0.43995383381843567, "learning_rate": 4.924563566072564e-05, "loss": 1.9834859848022461, "step": 1200 }, { "epoch": 0.0856919357132698, "grad_norm": 0.44207003712654114, "learning_rate": 4.9230428948917884e-05, "loss": 2.0661075592041014, "step": 1205 }, { "epoch": 0.08604750391125018, "grad_norm": 0.4353814423084259, "learning_rate": 4.921507288720035e-05, "loss": 1.9555856704711914, "step": 1210 }, { "epoch": 0.08640307210923055, "grad_norm": 0.4431837201118469, "learning_rate": 4.919956757022374e-05, "loss": 1.933385467529297, "step": 1215 }, { "epoch": 0.08675864030721092, "grad_norm": 0.40051746368408203, "learning_rate": 4.9183913093558676e-05, "loss": 1.7731637954711914, "step": 1220 }, { "epoch": 0.08711420850519129, "grad_norm": 0.488683819770813, "learning_rate": 4.9168109553695184e-05, "loss": 1.865903091430664, "step": 1225 }, { "epoch": 0.08746977670317167, "grad_norm": 0.4397886395454407, "learning_rate": 4.915215704804208e-05, "loss": 2.0127737045288088, "step": 1230 }, { "epoch": 0.08782534490115204, "grad_norm": 0.4224015176296234, "learning_rate": 4.9136055674926365e-05, "loss": 1.7999616622924806, "step": 1235 }, { "epoch": 0.08818091309913241, "grad_norm": 0.4563433825969696, "learning_rate": 4.91198055335926e-05, "loss": 1.8766897201538086, "step": 1240 }, { "epoch": 0.08853648129711279, "grad_norm": 0.3747509717941284, "learning_rate": 4.910340672420234e-05, "loss": 1.882638931274414, "step": 1245 }, { "epoch": 0.08889204949509316, "grad_norm": 1.0486830472946167, "learning_rate": 4.908685934783347e-05, "loss": 1.981215476989746, "step": 1250 }, { "epoch": 0.08924761769307353, "grad_norm": 0.4365674555301666, "learning_rate": 4.907016350647961e-05, "loss": 1.9899089813232422, "step": 1255 }, { "epoch": 0.0896031858910539, "grad_norm": 0.5747599601745605, "learning_rate": 4.905331930304948e-05, "loss": 1.9954425811767578, "step": 1260 }, { "epoch": 0.08995875408903428, "grad_norm": 0.46859487891197205, "learning_rate": 4.903632684136628e-05, "loss": 1.9303474426269531, "step": 1265 }, { "epoch": 0.09031432228701465, "grad_norm": 0.4422703981399536, "learning_rate": 4.9019186226166994e-05, "loss": 1.9309690475463868, "step": 1270 }, { "epoch": 0.09066989048499502, "grad_norm": 0.4180740714073181, "learning_rate": 4.9001897563101806e-05, "loss": 1.9369356155395507, "step": 1275 }, { "epoch": 0.09102545868297539, "grad_norm": 0.45563459396362305, "learning_rate": 4.898446095873345e-05, "loss": 2.009445381164551, "step": 1280 }, { "epoch": 0.09138102688095577, "grad_norm": 0.4095707833766937, "learning_rate": 4.8966876520536496e-05, "loss": 1.9960857391357423, "step": 1285 }, { "epoch": 0.09173659507893614, "grad_norm": 0.4561159014701843, "learning_rate": 4.8949144356896745e-05, "loss": 2.0763011932373048, "step": 1290 }, { "epoch": 0.09209216327691651, "grad_norm": 0.44926509261131287, "learning_rate": 4.8931264577110535e-05, "loss": 2.0234508514404297, "step": 1295 }, { "epoch": 0.09244773147489689, "grad_norm": 0.4175722002983093, "learning_rate": 4.891323729138405e-05, "loss": 2.048798751831055, "step": 1300 }, { "epoch": 0.09280329967287726, "grad_norm": 0.4039195477962494, "learning_rate": 4.8895062610832695e-05, "loss": 1.9779092788696289, "step": 1305 }, { "epoch": 0.09315886787085763, "grad_norm": 0.431517630815506, "learning_rate": 4.887674064748035e-05, "loss": 1.8589370727539063, "step": 1310 }, { "epoch": 0.093514436068838, "grad_norm": 0.46856579184532166, "learning_rate": 4.885827151425871e-05, "loss": 2.0986093521118163, "step": 1315 }, { "epoch": 0.09387000426681838, "grad_norm": 0.44901442527770996, "learning_rate": 4.88396553250066e-05, "loss": 1.744940948486328, "step": 1320 }, { "epoch": 0.09422557246479875, "grad_norm": 0.4657062590122223, "learning_rate": 4.8820892194469254e-05, "loss": 1.9488752365112305, "step": 1325 }, { "epoch": 0.09458114066277912, "grad_norm": 0.46259355545043945, "learning_rate": 4.8801982238297604e-05, "loss": 1.897237205505371, "step": 1330 }, { "epoch": 0.0949367088607595, "grad_norm": 0.42369475960731506, "learning_rate": 4.8782925573047596e-05, "loss": 1.9946008682250977, "step": 1335 }, { "epoch": 0.09529227705873987, "grad_norm": 0.7923505902290344, "learning_rate": 4.876372231617943e-05, "loss": 1.944803237915039, "step": 1340 }, { "epoch": 0.09564784525672024, "grad_norm": 0.4473399221897125, "learning_rate": 4.8744372586056855e-05, "loss": 1.8916728973388672, "step": 1345 }, { "epoch": 0.0960034134547006, "grad_norm": 0.461566686630249, "learning_rate": 4.872487650194647e-05, "loss": 1.923036766052246, "step": 1350 }, { "epoch": 0.09635898165268099, "grad_norm": 0.43765369057655334, "learning_rate": 4.870523418401694e-05, "loss": 1.9112867355346679, "step": 1355 }, { "epoch": 0.09671454985066136, "grad_norm": 0.4417065978050232, "learning_rate": 4.868544575333828e-05, "loss": 1.8557899475097657, "step": 1360 }, { "epoch": 0.09707011804864173, "grad_norm": 0.47006696462631226, "learning_rate": 4.8665511331881095e-05, "loss": 1.9570133209228515, "step": 1365 }, { "epoch": 0.0974256862466221, "grad_norm": 0.41962930560112, "learning_rate": 4.864543104251587e-05, "loss": 1.9790267944335938, "step": 1370 }, { "epoch": 0.09778125444460248, "grad_norm": 0.4155440628528595, "learning_rate": 4.862520500901214e-05, "loss": 1.9035987854003906, "step": 1375 }, { "epoch": 0.09813682264258285, "grad_norm": 0.4153623580932617, "learning_rate": 4.860483335603781e-05, "loss": 1.9679767608642578, "step": 1380 }, { "epoch": 0.09849239084056322, "grad_norm": 0.4250621199607849, "learning_rate": 4.858431620915829e-05, "loss": 1.840437126159668, "step": 1385 }, { "epoch": 0.0988479590385436, "grad_norm": 0.4002258777618408, "learning_rate": 4.856365369483584e-05, "loss": 1.924336051940918, "step": 1390 }, { "epoch": 0.09920352723652397, "grad_norm": 0.43188565969467163, "learning_rate": 4.854284594042867e-05, "loss": 1.8213363647460938, "step": 1395 }, { "epoch": 0.09955909543450434, "grad_norm": 0.4407031536102295, "learning_rate": 4.852189307419026e-05, "loss": 1.900477981567383, "step": 1400 }, { "epoch": 0.0999146636324847, "grad_norm": 0.6405236721038818, "learning_rate": 4.8500795225268475e-05, "loss": 2.05745906829834, "step": 1405 }, { "epoch": 0.10027023183046509, "grad_norm": 0.44744738936424255, "learning_rate": 4.847955252370485e-05, "loss": 1.9435310363769531, "step": 1410 }, { "epoch": 0.10062580002844546, "grad_norm": 0.4351164996623993, "learning_rate": 4.845816510043373e-05, "loss": 1.9711954116821289, "step": 1415 }, { "epoch": 0.10098136822642582, "grad_norm": 0.4473099410533905, "learning_rate": 4.8436633087281494e-05, "loss": 1.9254167556762696, "step": 1420 }, { "epoch": 0.10133693642440621, "grad_norm": 0.4199598431587219, "learning_rate": 4.8414956616965744e-05, "loss": 1.8612422943115234, "step": 1425 }, { "epoch": 0.10169250462238658, "grad_norm": 0.46582111716270447, "learning_rate": 4.839313582309446e-05, "loss": 2.055385208129883, "step": 1430 }, { "epoch": 0.10204807282036694, "grad_norm": 0.4102075695991516, "learning_rate": 4.837117084016521e-05, "loss": 1.9097806930541992, "step": 1435 }, { "epoch": 0.10240364101834731, "grad_norm": 0.4473963677883148, "learning_rate": 4.83490618035643e-05, "loss": 2.0282459259033203, "step": 1440 }, { "epoch": 0.1027592092163277, "grad_norm": 0.4236394762992859, "learning_rate": 4.8326808849565936e-05, "loss": 1.990822982788086, "step": 1445 }, { "epoch": 0.10311477741430806, "grad_norm": 0.44632235169410706, "learning_rate": 4.83044121153314e-05, "loss": 2.066908264160156, "step": 1450 }, { "epoch": 0.10347034561228843, "grad_norm": 0.42431074380874634, "learning_rate": 4.82818717389082e-05, "loss": 1.9028337478637696, "step": 1455 }, { "epoch": 0.10382591381026882, "grad_norm": 0.421022891998291, "learning_rate": 4.82591878592292e-05, "loss": 1.882857322692871, "step": 1460 }, { "epoch": 0.10418148200824918, "grad_norm": 0.43148377537727356, "learning_rate": 4.823636061611181e-05, "loss": 1.9189605712890625, "step": 1465 }, { "epoch": 0.10453705020622955, "grad_norm": 0.447110652923584, "learning_rate": 4.8213390150257066e-05, "loss": 1.990689468383789, "step": 1470 }, { "epoch": 0.10489261840420992, "grad_norm": 0.4151032567024231, "learning_rate": 4.819027660324881e-05, "loss": 1.972299575805664, "step": 1475 }, { "epoch": 0.1052481866021903, "grad_norm": 0.43426620960235596, "learning_rate": 4.816702011755277e-05, "loss": 2.1078207015991213, "step": 1480 }, { "epoch": 0.10560375480017067, "grad_norm": 0.48189279437065125, "learning_rate": 4.814362083651574e-05, "loss": 1.9480621337890625, "step": 1485 }, { "epoch": 0.10595932299815104, "grad_norm": 0.3947417438030243, "learning_rate": 4.812007890436466e-05, "loss": 1.9607099533081054, "step": 1490 }, { "epoch": 0.10631489119613141, "grad_norm": 0.4579639136791229, "learning_rate": 4.809639446620572e-05, "loss": 2.0648693084716796, "step": 1495 }, { "epoch": 0.1066704593941118, "grad_norm": 0.459697961807251, "learning_rate": 4.807256766802349e-05, "loss": 1.9681814193725586, "step": 1500 }, { "epoch": 0.10702602759209216, "grad_norm": 0.44938191771507263, "learning_rate": 4.804859865668002e-05, "loss": 1.9271417617797852, "step": 1505 }, { "epoch": 0.10738159579007253, "grad_norm": 0.4443996250629425, "learning_rate": 4.8024487579913896e-05, "loss": 2.042717170715332, "step": 1510 }, { "epoch": 0.10773716398805291, "grad_norm": 0.3744531571865082, "learning_rate": 4.8000234586339375e-05, "loss": 1.7823001861572265, "step": 1515 }, { "epoch": 0.10809273218603328, "grad_norm": 0.43542104959487915, "learning_rate": 4.7975839825445454e-05, "loss": 2.083765411376953, "step": 1520 }, { "epoch": 0.10844830038401365, "grad_norm": 0.4042755961418152, "learning_rate": 4.795130344759494e-05, "loss": 1.845391845703125, "step": 1525 }, { "epoch": 0.10880386858199402, "grad_norm": 0.4418489336967468, "learning_rate": 4.7926625604023536e-05, "loss": 2.053580474853516, "step": 1530 }, { "epoch": 0.1091594367799744, "grad_norm": 0.37466898560523987, "learning_rate": 4.7901806446838874e-05, "loss": 1.8470773696899414, "step": 1535 }, { "epoch": 0.10951500497795477, "grad_norm": 0.4396774470806122, "learning_rate": 4.787684612901965e-05, "loss": 2.0555131912231444, "step": 1540 }, { "epoch": 0.10987057317593514, "grad_norm": 0.44805994629859924, "learning_rate": 4.785174480441461e-05, "loss": 2.0072296142578123, "step": 1545 }, { "epoch": 0.11022614137391552, "grad_norm": 0.41868463158607483, "learning_rate": 4.782650262774163e-05, "loss": 1.984494400024414, "step": 1550 }, { "epoch": 0.11058170957189589, "grad_norm": 0.5847702622413635, "learning_rate": 4.7801119754586766e-05, "loss": 2.071546173095703, "step": 1555 }, { "epoch": 0.11093727776987626, "grad_norm": 0.4606797695159912, "learning_rate": 4.7775596341403276e-05, "loss": 1.9733022689819335, "step": 1560 }, { "epoch": 0.11129284596785663, "grad_norm": 0.4612416625022888, "learning_rate": 4.77499325455107e-05, "loss": 1.9200651168823242, "step": 1565 }, { "epoch": 0.11164841416583701, "grad_norm": 0.4469436705112457, "learning_rate": 4.772412852509382e-05, "loss": 1.918980026245117, "step": 1570 }, { "epoch": 0.11200398236381738, "grad_norm": 0.4780344069004059, "learning_rate": 4.7698184439201755e-05, "loss": 1.9311887741088867, "step": 1575 }, { "epoch": 0.11235955056179775, "grad_norm": 0.4588239789009094, "learning_rate": 4.7672100447746936e-05, "loss": 2.0349939346313475, "step": 1580 }, { "epoch": 0.11271511875977813, "grad_norm": 0.4519966244697571, "learning_rate": 4.764587671150414e-05, "loss": 1.941670036315918, "step": 1585 }, { "epoch": 0.1130706869577585, "grad_norm": 0.49292662739753723, "learning_rate": 4.761951339210949e-05, "loss": 2.0356687545776366, "step": 1590 }, { "epoch": 0.11342625515573887, "grad_norm": 0.4457705020904541, "learning_rate": 4.7593010652059465e-05, "loss": 2.0416622161865234, "step": 1595 }, { "epoch": 0.11378182335371924, "grad_norm": 0.4660331606864929, "learning_rate": 4.75663686547099e-05, "loss": 1.9822050094604493, "step": 1600 }, { "epoch": 0.11413739155169962, "grad_norm": 0.44478175044059753, "learning_rate": 4.7539587564274965e-05, "loss": 1.9137392044067383, "step": 1605 }, { "epoch": 0.11449295974967999, "grad_norm": 0.4500514268875122, "learning_rate": 4.7512667545826174e-05, "loss": 1.9187978744506835, "step": 1610 }, { "epoch": 0.11484852794766036, "grad_norm": 0.41738051176071167, "learning_rate": 4.748560876529135e-05, "loss": 2.055419158935547, "step": 1615 }, { "epoch": 0.11520409614564073, "grad_norm": 0.4811953604221344, "learning_rate": 4.745841138945358e-05, "loss": 1.934524154663086, "step": 1620 }, { "epoch": 0.11555966434362111, "grad_norm": 0.41122540831565857, "learning_rate": 4.743107558595029e-05, "loss": 1.8639139175415038, "step": 1625 }, { "epoch": 0.11591523254160148, "grad_norm": 0.46006473898887634, "learning_rate": 4.7403601523272056e-05, "loss": 1.9548212051391602, "step": 1630 }, { "epoch": 0.11627080073958185, "grad_norm": 0.44362926483154297, "learning_rate": 4.7375989370761695e-05, "loss": 2.0186790466308593, "step": 1635 }, { "epoch": 0.11662636893756223, "grad_norm": 0.41030293703079224, "learning_rate": 4.734823929861317e-05, "loss": 1.9607160568237305, "step": 1640 }, { "epoch": 0.1169819371355426, "grad_norm": 0.44192376732826233, "learning_rate": 4.732035147787052e-05, "loss": 1.9846776962280273, "step": 1645 }, { "epoch": 0.11733750533352297, "grad_norm": 0.4359341263771057, "learning_rate": 4.729232608042684e-05, "loss": 1.990248680114746, "step": 1650 }, { "epoch": 0.11769307353150334, "grad_norm": 0.43293297290802, "learning_rate": 4.726416327902323e-05, "loss": 1.808917236328125, "step": 1655 }, { "epoch": 0.11804864172948372, "grad_norm": 0.4392835795879364, "learning_rate": 4.723586324724769e-05, "loss": 1.8549243927001953, "step": 1660 }, { "epoch": 0.11840420992746409, "grad_norm": 0.36470040678977966, "learning_rate": 4.720742615953407e-05, "loss": 1.906550979614258, "step": 1665 }, { "epoch": 0.11875977812544446, "grad_norm": 0.45539388060569763, "learning_rate": 4.717885219116101e-05, "loss": 2.01754150390625, "step": 1670 }, { "epoch": 0.11911534632342484, "grad_norm": 0.4638034999370575, "learning_rate": 4.715014151825085e-05, "loss": 1.96883544921875, "step": 1675 }, { "epoch": 0.11947091452140521, "grad_norm": 0.48876047134399414, "learning_rate": 4.7121294317768506e-05, "loss": 1.849966049194336, "step": 1680 }, { "epoch": 0.11982648271938558, "grad_norm": 0.43939417600631714, "learning_rate": 4.709231076752045e-05, "loss": 1.9786798477172851, "step": 1685 }, { "epoch": 0.12018205091736595, "grad_norm": 0.342741996049881, "learning_rate": 4.7063191046153574e-05, "loss": 1.850713348388672, "step": 1690 }, { "epoch": 0.12053761911534633, "grad_norm": 0.4128732979297638, "learning_rate": 4.703393533315408e-05, "loss": 1.9891803741455079, "step": 1695 }, { "epoch": 0.1208931873133267, "grad_norm": 0.4280429482460022, "learning_rate": 4.700454380884638e-05, "loss": 1.9010469436645507, "step": 1700 }, { "epoch": 0.12124875551130707, "grad_norm": 0.43759799003601074, "learning_rate": 4.697501665439201e-05, "loss": 1.9979541778564454, "step": 1705 }, { "epoch": 0.12160432370928743, "grad_norm": 0.42854565382003784, "learning_rate": 4.694535405178848e-05, "loss": 1.7715314865112304, "step": 1710 }, { "epoch": 0.12195989190726782, "grad_norm": 0.4579486548900604, "learning_rate": 4.6915556183868174e-05, "loss": 1.8214681625366211, "step": 1715 }, { "epoch": 0.12231546010524819, "grad_norm": 0.4748671054840088, "learning_rate": 4.6885623234297214e-05, "loss": 1.9865182876586913, "step": 1720 }, { "epoch": 0.12267102830322855, "grad_norm": 0.4472329020500183, "learning_rate": 4.685555538757432e-05, "loss": 1.987970733642578, "step": 1725 }, { "epoch": 0.12302659650120894, "grad_norm": 0.5258608460426331, "learning_rate": 4.6825352829029705e-05, "loss": 1.9912925720214845, "step": 1730 }, { "epoch": 0.1233821646991893, "grad_norm": 0.5144405961036682, "learning_rate": 4.679501574482388e-05, "loss": 1.9761497497558593, "step": 1735 }, { "epoch": 0.12373773289716967, "grad_norm": 0.47484397888183594, "learning_rate": 4.676454432194656e-05, "loss": 2.0215763092041015, "step": 1740 }, { "epoch": 0.12409330109515004, "grad_norm": 0.49055057764053345, "learning_rate": 4.6733938748215475e-05, "loss": 1.8958847045898437, "step": 1745 }, { "epoch": 0.12444886929313043, "grad_norm": 0.5371253490447998, "learning_rate": 4.6703199212275226e-05, "loss": 1.8823431015014649, "step": 1750 }, { "epoch": 0.1248044374911108, "grad_norm": 0.4418807625770569, "learning_rate": 4.667232590359614e-05, "loss": 1.7773038864135742, "step": 1755 }, { "epoch": 0.12516000568909116, "grad_norm": 0.47592902183532715, "learning_rate": 4.6641319012473043e-05, "loss": 1.986189079284668, "step": 1760 }, { "epoch": 0.12551557388707155, "grad_norm": 0.46170562505722046, "learning_rate": 4.661017873002417e-05, "loss": 2.0420930862426756, "step": 1765 }, { "epoch": 0.1258711420850519, "grad_norm": 0.4310692250728607, "learning_rate": 4.657890524818993e-05, "loss": 1.9371036529541015, "step": 1770 }, { "epoch": 0.12622671028303228, "grad_norm": 0.42791032791137695, "learning_rate": 4.6547498759731724e-05, "loss": 1.8477264404296876, "step": 1775 }, { "epoch": 0.12658227848101267, "grad_norm": 0.48630771040916443, "learning_rate": 4.651595945823078e-05, "loss": 2.023675537109375, "step": 1780 }, { "epoch": 0.12693784667899302, "grad_norm": 0.4483960270881653, "learning_rate": 4.648428753808696e-05, "loss": 1.868947982788086, "step": 1785 }, { "epoch": 0.1272934148769734, "grad_norm": 0.36611407995224, "learning_rate": 4.6452483194517535e-05, "loss": 1.8235824584960938, "step": 1790 }, { "epoch": 0.1276489830749538, "grad_norm": 0.4243677258491516, "learning_rate": 4.6420546623556016e-05, "loss": 1.9161598205566406, "step": 1795 }, { "epoch": 0.12800455127293414, "grad_norm": 0.4300360083580017, "learning_rate": 4.638847802205091e-05, "loss": 1.9328779220581054, "step": 1800 }, { "epoch": 0.12836011947091452, "grad_norm": 0.4602338671684265, "learning_rate": 4.6356277587664546e-05, "loss": 1.9402246475219727, "step": 1805 }, { "epoch": 0.1287156876688949, "grad_norm": 0.446158230304718, "learning_rate": 4.632394551887181e-05, "loss": 2.0193317413330076, "step": 1810 }, { "epoch": 0.12907125586687526, "grad_norm": 0.46085014939308167, "learning_rate": 4.6291482014958974e-05, "loss": 2.0381757736206056, "step": 1815 }, { "epoch": 0.12942682406485564, "grad_norm": 0.43898117542266846, "learning_rate": 4.625888727602242e-05, "loss": 1.9459449768066406, "step": 1820 }, { "epoch": 0.129782392262836, "grad_norm": 0.43887585401535034, "learning_rate": 4.622616150296745e-05, "loss": 1.9349231719970703, "step": 1825 }, { "epoch": 0.13013796046081638, "grad_norm": 0.4367786943912506, "learning_rate": 4.6193304897507e-05, "loss": 1.927500343322754, "step": 1830 }, { "epoch": 0.13049352865879676, "grad_norm": 0.44730865955352783, "learning_rate": 4.616031766216043e-05, "loss": 1.9544364929199218, "step": 1835 }, { "epoch": 0.13084909685677712, "grad_norm": 0.5109915137290955, "learning_rate": 4.612720000025228e-05, "loss": 1.8460371017456054, "step": 1840 }, { "epoch": 0.1312046650547575, "grad_norm": 0.47076430916786194, "learning_rate": 4.6093952115911e-05, "loss": 2.045991134643555, "step": 1845 }, { "epoch": 0.13156023325273788, "grad_norm": 0.4458104372024536, "learning_rate": 4.606057421406769e-05, "loss": 2.012459373474121, "step": 1850 }, { "epoch": 0.13191580145071824, "grad_norm": 0.34033629298210144, "learning_rate": 4.6027066500454826e-05, "loss": 1.8160129547119142, "step": 1855 }, { "epoch": 0.13227136964869862, "grad_norm": 0.3816074728965759, "learning_rate": 4.599342918160506e-05, "loss": 1.9332874298095704, "step": 1860 }, { "epoch": 0.132626937846679, "grad_norm": 0.4519752860069275, "learning_rate": 4.5959662464849854e-05, "loss": 2.0322479248046874, "step": 1865 }, { "epoch": 0.13298250604465936, "grad_norm": 0.4078761041164398, "learning_rate": 4.592576655831826e-05, "loss": 1.9132863998413085, "step": 1870 }, { "epoch": 0.13333807424263974, "grad_norm": 0.4286755621433258, "learning_rate": 4.5891741670935596e-05, "loss": 1.8255847930908202, "step": 1875 }, { "epoch": 0.1336936424406201, "grad_norm": 0.4683900475502014, "learning_rate": 4.585758801242223e-05, "loss": 1.8925376892089845, "step": 1880 }, { "epoch": 0.13404921063860048, "grad_norm": 0.42389756441116333, "learning_rate": 4.58233057932922e-05, "loss": 1.8729282379150392, "step": 1885 }, { "epoch": 0.13440477883658086, "grad_norm": 0.4422023594379425, "learning_rate": 4.578889522485198e-05, "loss": 1.8858570098876952, "step": 1890 }, { "epoch": 0.13476034703456122, "grad_norm": 0.4524957239627838, "learning_rate": 4.575435651919915e-05, "loss": 1.9918432235717773, "step": 1895 }, { "epoch": 0.1351159152325416, "grad_norm": 0.45303836464881897, "learning_rate": 4.571968988922108e-05, "loss": 2.0699771881103515, "step": 1900 }, { "epoch": 0.13547148343052198, "grad_norm": 0.4170067608356476, "learning_rate": 4.5684895548593655e-05, "loss": 1.807975959777832, "step": 1905 }, { "epoch": 0.13582705162850234, "grad_norm": 0.43357110023498535, "learning_rate": 4.564997371177992e-05, "loss": 1.945266342163086, "step": 1910 }, { "epoch": 0.13618261982648272, "grad_norm": 0.39495766162872314, "learning_rate": 4.561492459402877e-05, "loss": 1.9426273345947265, "step": 1915 }, { "epoch": 0.1365381880244631, "grad_norm": 0.426727831363678, "learning_rate": 4.557974841137364e-05, "loss": 1.7720190048217774, "step": 1920 }, { "epoch": 0.13689375622244346, "grad_norm": 0.4814690947532654, "learning_rate": 4.554444538063113e-05, "loss": 1.9700519561767578, "step": 1925 }, { "epoch": 0.13724932442042384, "grad_norm": 0.44320520758628845, "learning_rate": 4.550901571939974e-05, "loss": 1.8878829956054688, "step": 1930 }, { "epoch": 0.13760489261840422, "grad_norm": 0.37623390555381775, "learning_rate": 4.5473459646058455e-05, "loss": 1.9020067214965821, "step": 1935 }, { "epoch": 0.13796046081638458, "grad_norm": 0.4533628523349762, "learning_rate": 4.5437777379765444e-05, "loss": 2.062186050415039, "step": 1940 }, { "epoch": 0.13831602901436496, "grad_norm": 0.415678471326828, "learning_rate": 4.540196914045668e-05, "loss": 1.9647165298461915, "step": 1945 }, { "epoch": 0.13867159721234532, "grad_norm": 0.42477312684059143, "learning_rate": 4.536603514884463e-05, "loss": 1.9233779907226562, "step": 1950 }, { "epoch": 0.1390271654103257, "grad_norm": 1.9212150573730469, "learning_rate": 4.532997562641683e-05, "loss": 1.8557853698730469, "step": 1955 }, { "epoch": 0.13938273360830608, "grad_norm": 0.4846855401992798, "learning_rate": 4.529379079543459e-05, "loss": 1.9841705322265626, "step": 1960 }, { "epoch": 0.13973830180628644, "grad_norm": 0.44399747252464294, "learning_rate": 4.525748087893155e-05, "loss": 2.041277313232422, "step": 1965 }, { "epoch": 0.14009387000426682, "grad_norm": 0.4050864577293396, "learning_rate": 4.522104610071236e-05, "loss": 1.9988048553466797, "step": 1970 }, { "epoch": 0.1404494382022472, "grad_norm": 0.4377169609069824, "learning_rate": 4.5184486685351314e-05, "loss": 2.0682132720947264, "step": 1975 }, { "epoch": 0.14080500640022756, "grad_norm": 0.4718763530254364, "learning_rate": 4.514780285819089e-05, "loss": 1.8091747283935546, "step": 1980 }, { "epoch": 0.14116057459820794, "grad_norm": 0.7917141914367676, "learning_rate": 4.511099484534043e-05, "loss": 1.9872066497802734, "step": 1985 }, { "epoch": 0.14151614279618832, "grad_norm": 0.4161342978477478, "learning_rate": 4.507406287367473e-05, "loss": 2.0225597381591798, "step": 1990 }, { "epoch": 0.14187171099416868, "grad_norm": 0.44084230065345764, "learning_rate": 4.5037007170832615e-05, "loss": 2.0188667297363283, "step": 1995 }, { "epoch": 0.14222727919214906, "grad_norm": 0.4169104993343353, "learning_rate": 4.4999827965215554e-05, "loss": 1.8438201904296876, "step": 2000 }, { "epoch": 0.1425828473901294, "grad_norm": 0.4138292074203491, "learning_rate": 4.496252548598631e-05, "loss": 1.8704288482666016, "step": 2005 }, { "epoch": 0.1429384155881098, "grad_norm": 0.4355562627315521, "learning_rate": 4.4925099963067386e-05, "loss": 1.814544677734375, "step": 2010 }, { "epoch": 0.14329398378609018, "grad_norm": 0.44727447628974915, "learning_rate": 4.488755162713975e-05, "loss": 1.9385015487670898, "step": 2015 }, { "epoch": 0.14364955198407053, "grad_norm": 0.38137808442115784, "learning_rate": 4.484988070964135e-05, "loss": 1.783839797973633, "step": 2020 }, { "epoch": 0.14400512018205092, "grad_norm": 0.7276799082756042, "learning_rate": 4.4812087442765686e-05, "loss": 1.8630420684814453, "step": 2025 }, { "epoch": 0.1443606883800313, "grad_norm": 0.44332802295684814, "learning_rate": 4.477417205946037e-05, "loss": 2.0805830001831054, "step": 2030 }, { "epoch": 0.14471625657801165, "grad_norm": 0.44779035449028015, "learning_rate": 4.473613479342575e-05, "loss": 1.7817573547363281, "step": 2035 }, { "epoch": 0.14507182477599204, "grad_norm": 0.4486839175224304, "learning_rate": 4.469797587911339e-05, "loss": 2.026826286315918, "step": 2040 }, { "epoch": 0.14542739297397242, "grad_norm": 0.4497630298137665, "learning_rate": 4.465969555172467e-05, "loss": 1.980991744995117, "step": 2045 }, { "epoch": 0.14578296117195277, "grad_norm": 0.40893423557281494, "learning_rate": 4.462129404720934e-05, "loss": 1.9634090423583985, "step": 2050 }, { "epoch": 0.14613852936993316, "grad_norm": 0.45576468110084534, "learning_rate": 4.458277160226404e-05, "loss": 2.0306549072265625, "step": 2055 }, { "epoch": 0.14649409756791354, "grad_norm": 0.4479800760746002, "learning_rate": 4.454412845433086e-05, "loss": 1.9566848754882813, "step": 2060 }, { "epoch": 0.1468496657658939, "grad_norm": 0.3781600892543793, "learning_rate": 4.450536484159587e-05, "loss": 1.858958625793457, "step": 2065 }, { "epoch": 0.14720523396387428, "grad_norm": 0.39559897780418396, "learning_rate": 4.446648100298765e-05, "loss": 1.8779438018798829, "step": 2070 }, { "epoch": 0.14756080216185463, "grad_norm": 0.4433683753013611, "learning_rate": 4.442747717817583e-05, "loss": 2.061594009399414, "step": 2075 }, { "epoch": 0.14791637035983501, "grad_norm": 0.4423002600669861, "learning_rate": 4.438835360756959e-05, "loss": 1.9578702926635743, "step": 2080 }, { "epoch": 0.1482719385578154, "grad_norm": 0.5244864225387573, "learning_rate": 4.434911053231621e-05, "loss": 1.9951553344726562, "step": 2085 }, { "epoch": 0.14862750675579575, "grad_norm": 0.38346561789512634, "learning_rate": 4.430974819429954e-05, "loss": 1.694706344604492, "step": 2090 }, { "epoch": 0.14898307495377613, "grad_norm": 0.46447595953941345, "learning_rate": 4.4270266836138554e-05, "loss": 1.9652053833007812, "step": 2095 }, { "epoch": 0.14933864315175652, "grad_norm": 0.44788530468940735, "learning_rate": 4.423066670118581e-05, "loss": 1.9724103927612304, "step": 2100 }, { "epoch": 0.14969421134973687, "grad_norm": 0.4008125066757202, "learning_rate": 4.419094803352602e-05, "loss": 1.9330440521240235, "step": 2105 }, { "epoch": 0.15004977954771725, "grad_norm": 0.4721739888191223, "learning_rate": 4.415111107797445e-05, "loss": 1.9580862045288085, "step": 2110 }, { "epoch": 0.15040534774569764, "grad_norm": 0.4447092115879059, "learning_rate": 4.41111560800755e-05, "loss": 2.040000152587891, "step": 2115 }, { "epoch": 0.150760915943678, "grad_norm": 0.4366373121738434, "learning_rate": 4.407108328610111e-05, "loss": 2.083973693847656, "step": 2120 }, { "epoch": 0.15111648414165837, "grad_norm": 0.44759365916252136, "learning_rate": 4.4030892943049326e-05, "loss": 1.9249889373779296, "step": 2125 }, { "epoch": 0.15147205233963873, "grad_norm": 0.4133858382701874, "learning_rate": 4.399058529864272e-05, "loss": 1.900141143798828, "step": 2130 }, { "epoch": 0.1518276205376191, "grad_norm": 0.41831332445144653, "learning_rate": 4.395016060132686e-05, "loss": 1.9822818756103515, "step": 2135 }, { "epoch": 0.1521831887355995, "grad_norm": 0.4524232745170593, "learning_rate": 4.390961910026883e-05, "loss": 1.8551643371582032, "step": 2140 }, { "epoch": 0.15253875693357985, "grad_norm": 0.4212983548641205, "learning_rate": 4.386896104535563e-05, "loss": 1.9403331756591797, "step": 2145 }, { "epoch": 0.15289432513156023, "grad_norm": 0.422979474067688, "learning_rate": 4.382818668719268e-05, "loss": 1.929132843017578, "step": 2150 }, { "epoch": 0.15324989332954062, "grad_norm": 0.41337552666664124, "learning_rate": 4.3787296277102264e-05, "loss": 1.988453483581543, "step": 2155 }, { "epoch": 0.15360546152752097, "grad_norm": 0.4621830880641937, "learning_rate": 4.3746290067121985e-05, "loss": 2.0812068939208985, "step": 2160 }, { "epoch": 0.15396102972550135, "grad_norm": 0.5142182111740112, "learning_rate": 4.37051683100032e-05, "loss": 1.9248458862304687, "step": 2165 }, { "epoch": 0.15431659792348174, "grad_norm": 0.45941993594169617, "learning_rate": 4.3663931259209455e-05, "loss": 2.0639163970947267, "step": 2170 }, { "epoch": 0.1546721661214621, "grad_norm": 0.47261860966682434, "learning_rate": 4.3622579168914945e-05, "loss": 1.8689651489257812, "step": 2175 }, { "epoch": 0.15502773431944247, "grad_norm": 0.4174124598503113, "learning_rate": 4.3581112294002955e-05, "loss": 1.7178684234619142, "step": 2180 }, { "epoch": 0.15538330251742286, "grad_norm": 0.482042133808136, "learning_rate": 4.353953089006425e-05, "loss": 1.9751890182495118, "step": 2185 }, { "epoch": 0.1557388707154032, "grad_norm": 0.4697819948196411, "learning_rate": 4.349783521339552e-05, "loss": 1.9046506881713867, "step": 2190 }, { "epoch": 0.1560944389133836, "grad_norm": 0.45761585235595703, "learning_rate": 4.345602552099782e-05, "loss": 2.003580093383789, "step": 2195 }, { "epoch": 0.15645000711136395, "grad_norm": 0.4174576997756958, "learning_rate": 4.3414102070574955e-05, "loss": 1.8796783447265626, "step": 2200 }, { "epoch": 0.15680557530934433, "grad_norm": 0.46885132789611816, "learning_rate": 4.33720651205319e-05, "loss": 2.024154281616211, "step": 2205 }, { "epoch": 0.1571611435073247, "grad_norm": 0.44503381848335266, "learning_rate": 4.3329914929973226e-05, "loss": 1.8875150680541992, "step": 2210 }, { "epoch": 0.15751671170530507, "grad_norm": 0.44725823402404785, "learning_rate": 4.3287651758701466e-05, "loss": 1.8852039337158204, "step": 2215 }, { "epoch": 0.15787227990328545, "grad_norm": 0.4437294006347656, "learning_rate": 4.324527586721556e-05, "loss": 2.0251564025878905, "step": 2220 }, { "epoch": 0.15822784810126583, "grad_norm": 0.4622463583946228, "learning_rate": 4.320278751670922e-05, "loss": 2.086625862121582, "step": 2225 }, { "epoch": 0.1585834162992462, "grad_norm": 0.45187127590179443, "learning_rate": 4.3160186969069316e-05, "loss": 1.978363609313965, "step": 2230 }, { "epoch": 0.15893898449722657, "grad_norm": 0.4221031069755554, "learning_rate": 4.311747448687428e-05, "loss": 1.9371501922607421, "step": 2235 }, { "epoch": 0.15929455269520695, "grad_norm": 0.5238663554191589, "learning_rate": 4.3074650333392466e-05, "loss": 1.9388948440551759, "step": 2240 }, { "epoch": 0.1596501208931873, "grad_norm": 0.4347282648086548, "learning_rate": 4.303171477258055e-05, "loss": 1.8455482482910157, "step": 2245 }, { "epoch": 0.1600056890911677, "grad_norm": 0.370993435382843, "learning_rate": 4.2988668069081896e-05, "loss": 1.8599109649658203, "step": 2250 }, { "epoch": 0.16036125728914805, "grad_norm": 0.3910074532032013, "learning_rate": 4.294551048822493e-05, "loss": 1.93493595123291, "step": 2255 }, { "epoch": 0.16071682548712843, "grad_norm": 0.410757839679718, "learning_rate": 4.290224229602145e-05, "loss": 1.8846965789794923, "step": 2260 }, { "epoch": 0.1610723936851088, "grad_norm": 0.4310605227947235, "learning_rate": 4.285886375916511e-05, "loss": 1.8902559280395508, "step": 2265 }, { "epoch": 0.16142796188308917, "grad_norm": 0.4542630910873413, "learning_rate": 4.281537514502963e-05, "loss": 1.9638442993164062, "step": 2270 }, { "epoch": 0.16178353008106955, "grad_norm": 0.3843837380409241, "learning_rate": 4.277177672166724e-05, "loss": 1.9700178146362304, "step": 2275 }, { "epoch": 0.16213909827904993, "grad_norm": 0.4443548321723938, "learning_rate": 4.272806875780703e-05, "loss": 1.8911558151245118, "step": 2280 }, { "epoch": 0.1624946664770303, "grad_norm": 0.44673576951026917, "learning_rate": 4.268425152285324e-05, "loss": 1.9637842178344727, "step": 2285 }, { "epoch": 0.16285023467501067, "grad_norm": 0.4163452386856079, "learning_rate": 4.2640325286883616e-05, "loss": 1.9915699005126952, "step": 2290 }, { "epoch": 0.16320580287299105, "grad_norm": 0.45626723766326904, "learning_rate": 4.259629032064779e-05, "loss": 1.8824113845825194, "step": 2295 }, { "epoch": 0.1635613710709714, "grad_norm": 0.42183881998062134, "learning_rate": 4.255214689556557e-05, "loss": 1.9371522903442382, "step": 2300 }, { "epoch": 0.1639169392689518, "grad_norm": 0.4532131850719452, "learning_rate": 4.250789528372525e-05, "loss": 1.8804466247558593, "step": 2305 }, { "epoch": 0.16427250746693214, "grad_norm": 0.4447820782661438, "learning_rate": 4.2463535757882e-05, "loss": 2.0081401824951173, "step": 2310 }, { "epoch": 0.16462807566491253, "grad_norm": 0.42160630226135254, "learning_rate": 4.241906859145611e-05, "loss": 1.8658864974975586, "step": 2315 }, { "epoch": 0.1649836438628929, "grad_norm": 0.4783008396625519, "learning_rate": 4.237449405853136e-05, "loss": 1.911276626586914, "step": 2320 }, { "epoch": 0.16533921206087326, "grad_norm": 0.44960615038871765, "learning_rate": 4.2329812433853286e-05, "loss": 1.7698854446411132, "step": 2325 }, { "epoch": 0.16569478025885365, "grad_norm": 0.44757333397865295, "learning_rate": 4.228502399282752e-05, "loss": 2.011037254333496, "step": 2330 }, { "epoch": 0.16605034845683403, "grad_norm": 0.4462078809738159, "learning_rate": 4.2240129011518084e-05, "loss": 1.868431854248047, "step": 2335 }, { "epoch": 0.16640591665481438, "grad_norm": 0.38850635290145874, "learning_rate": 4.2195127766645684e-05, "loss": 1.9110704421997071, "step": 2340 }, { "epoch": 0.16676148485279477, "grad_norm": 0.3980312943458557, "learning_rate": 4.215002053558601e-05, "loss": 1.8870584487915039, "step": 2345 }, { "epoch": 0.16711705305077515, "grad_norm": 0.3916946053504944, "learning_rate": 4.2104807596368024e-05, "loss": 1.9116224288940429, "step": 2350 }, { "epoch": 0.1674726212487555, "grad_norm": 0.4019387662410736, "learning_rate": 4.205948922767222e-05, "loss": 1.8693748474121095, "step": 2355 }, { "epoch": 0.1678281894467359, "grad_norm": 0.4479674696922302, "learning_rate": 4.201406570882898e-05, "loss": 2.0016605377197267, "step": 2360 }, { "epoch": 0.16818375764471627, "grad_norm": 0.4667653739452362, "learning_rate": 4.196853731981675e-05, "loss": 2.0045372009277345, "step": 2365 }, { "epoch": 0.16853932584269662, "grad_norm": 0.44826748967170715, "learning_rate": 4.19229043412604e-05, "loss": 1.9542922973632812, "step": 2370 }, { "epoch": 0.168894894040677, "grad_norm": 0.3861413300037384, "learning_rate": 4.187716705442946e-05, "loss": 1.8837881088256836, "step": 2375 }, { "epoch": 0.16925046223865736, "grad_norm": 0.4638286232948303, "learning_rate": 4.183132574123637e-05, "loss": 2.000873565673828, "step": 2380 }, { "epoch": 0.16960603043663774, "grad_norm": 0.4749648869037628, "learning_rate": 4.1785380684234794e-05, "loss": 1.9788923263549805, "step": 2385 }, { "epoch": 0.16996159863461813, "grad_norm": 0.5578358173370361, "learning_rate": 4.173933216661781e-05, "loss": 1.9880260467529296, "step": 2390 }, { "epoch": 0.17031716683259848, "grad_norm": 0.4991210103034973, "learning_rate": 4.169318047221621e-05, "loss": 1.8414850234985352, "step": 2395 }, { "epoch": 0.17067273503057886, "grad_norm": 0.42388707399368286, "learning_rate": 4.1646925885496765e-05, "loss": 1.767096710205078, "step": 2400 }, { "epoch": 0.17102830322855925, "grad_norm": 0.4282020926475525, "learning_rate": 4.160056869156041e-05, "loss": 1.9758687973022462, "step": 2405 }, { "epoch": 0.1713838714265396, "grad_norm": 0.458624005317688, "learning_rate": 4.155410917614054e-05, "loss": 1.9250797271728515, "step": 2410 }, { "epoch": 0.17173943962451999, "grad_norm": 0.42114073038101196, "learning_rate": 4.150754762560125e-05, "loss": 1.7829608917236328, "step": 2415 }, { "epoch": 0.17209500782250037, "grad_norm": 0.39527589082717896, "learning_rate": 4.146088432693551e-05, "loss": 1.9576066970825194, "step": 2420 }, { "epoch": 0.17245057602048072, "grad_norm": 0.4918673634529114, "learning_rate": 4.141411956776348e-05, "loss": 1.9453657150268555, "step": 2425 }, { "epoch": 0.1728061442184611, "grad_norm": 0.4653744697570801, "learning_rate": 4.1367253636330673e-05, "loss": 1.9483770370483398, "step": 2430 }, { "epoch": 0.17316171241644146, "grad_norm": 0.4965675175189972, "learning_rate": 4.132028682150621e-05, "loss": 1.869779396057129, "step": 2435 }, { "epoch": 0.17351728061442184, "grad_norm": 0.44001322984695435, "learning_rate": 4.127321941278103e-05, "loss": 1.9733034133911134, "step": 2440 }, { "epoch": 0.17387284881240223, "grad_norm": 0.4100700318813324, "learning_rate": 4.1226051700266086e-05, "loss": 1.8334259033203124, "step": 2445 }, { "epoch": 0.17422841701038258, "grad_norm": 0.45240455865859985, "learning_rate": 4.117878397469062e-05, "loss": 1.830316162109375, "step": 2450 }, { "epoch": 0.17458398520836296, "grad_norm": 0.5378392934799194, "learning_rate": 4.113141652740028e-05, "loss": 1.8567264556884766, "step": 2455 }, { "epoch": 0.17493955340634335, "grad_norm": 0.45153486728668213, "learning_rate": 4.108394965035539e-05, "loss": 1.9227025985717774, "step": 2460 }, { "epoch": 0.1752951216043237, "grad_norm": 0.42474034428596497, "learning_rate": 4.103638363612916e-05, "loss": 2.0138174057006837, "step": 2465 }, { "epoch": 0.17565068980230408, "grad_norm": 0.5279356241226196, "learning_rate": 4.098871877790582e-05, "loss": 1.9497529983520507, "step": 2470 }, { "epoch": 0.17600625800028447, "grad_norm": 0.4550362527370453, "learning_rate": 4.0940955369478865e-05, "loss": 1.8590036392211915, "step": 2475 }, { "epoch": 0.17636182619826482, "grad_norm": 0.4169063866138458, "learning_rate": 4.089309370524921e-05, "loss": 1.973785400390625, "step": 2480 }, { "epoch": 0.1767173943962452, "grad_norm": 0.4532197117805481, "learning_rate": 4.0845134080223405e-05, "loss": 1.8443485260009767, "step": 2485 }, { "epoch": 0.17707296259422559, "grad_norm": 0.39360401034355164, "learning_rate": 4.0797076790011804e-05, "loss": 1.8951366424560547, "step": 2490 }, { "epoch": 0.17742853079220594, "grad_norm": 0.5400301218032837, "learning_rate": 4.074892213082676e-05, "loss": 1.975143051147461, "step": 2495 }, { "epoch": 0.17778409899018632, "grad_norm": 0.4516591727733612, "learning_rate": 4.070067039948075e-05, "loss": 1.9858938217163087, "step": 2500 }, { "epoch": 0.17813966718816668, "grad_norm": 0.4690985083580017, "learning_rate": 4.065232189338459e-05, "loss": 1.9931879043579102, "step": 2505 }, { "epoch": 0.17849523538614706, "grad_norm": 1.5915861129760742, "learning_rate": 4.06038769105456e-05, "loss": 1.909975814819336, "step": 2510 }, { "epoch": 0.17885080358412744, "grad_norm": 0.486126184463501, "learning_rate": 4.055533574956575e-05, "loss": 1.9617343902587892, "step": 2515 }, { "epoch": 0.1792063717821078, "grad_norm": 0.4634505808353424, "learning_rate": 4.050669870963982e-05, "loss": 1.9592121124267579, "step": 2520 }, { "epoch": 0.17956193998008818, "grad_norm": 0.44094157218933105, "learning_rate": 4.045796609055358e-05, "loss": 1.9630876541137696, "step": 2525 }, { "epoch": 0.17991750817806856, "grad_norm": 0.4425979554653168, "learning_rate": 4.0409138192681906e-05, "loss": 1.8966436386108398, "step": 2530 }, { "epoch": 0.18027307637604892, "grad_norm": 0.41453537344932556, "learning_rate": 4.036021531698696e-05, "loss": 1.9229925155639649, "step": 2535 }, { "epoch": 0.1806286445740293, "grad_norm": 0.45490819215774536, "learning_rate": 4.03111977650163e-05, "loss": 1.9525653839111328, "step": 2540 }, { "epoch": 0.18098421277200968, "grad_norm": 0.44942212104797363, "learning_rate": 4.026208583890107e-05, "loss": 2.0264522552490236, "step": 2545 }, { "epoch": 0.18133978096999004, "grad_norm": 0.8904481530189514, "learning_rate": 4.0212879841354105e-05, "loss": 1.9433847427368165, "step": 2550 }, { "epoch": 0.18169534916797042, "grad_norm": 0.37057262659072876, "learning_rate": 4.016358007566807e-05, "loss": 1.7172927856445312, "step": 2555 }, { "epoch": 0.18205091736595078, "grad_norm": 0.4654898941516876, "learning_rate": 4.011418684571357e-05, "loss": 2.062900352478027, "step": 2560 }, { "epoch": 0.18240648556393116, "grad_norm": 0.4679494798183441, "learning_rate": 4.006470045593734e-05, "loss": 2.0013402938842773, "step": 2565 }, { "epoch": 0.18276205376191154, "grad_norm": 0.4374832808971405, "learning_rate": 4.001512121136029e-05, "loss": 1.8824588775634765, "step": 2570 }, { "epoch": 0.1831176219598919, "grad_norm": 0.4508325755596161, "learning_rate": 3.996544941757568e-05, "loss": 1.8716596603393554, "step": 2575 }, { "epoch": 0.18347319015787228, "grad_norm": 0.5170300602912903, "learning_rate": 3.991568538074721e-05, "loss": 1.935582733154297, "step": 2580 }, { "epoch": 0.18382875835585266, "grad_norm": 0.48470091819763184, "learning_rate": 3.986582940760717e-05, "loss": 1.9593294143676758, "step": 2585 }, { "epoch": 0.18418432655383302, "grad_norm": 0.4401293098926544, "learning_rate": 3.981588180545447e-05, "loss": 1.9998178482055664, "step": 2590 }, { "epoch": 0.1845398947518134, "grad_norm": 0.47039109468460083, "learning_rate": 3.9765842882152844e-05, "loss": 1.8107881546020508, "step": 2595 }, { "epoch": 0.18489546294979378, "grad_norm": 0.4378872513771057, "learning_rate": 3.971571294612888e-05, "loss": 1.9148286819458007, "step": 2600 }, { "epoch": 0.18525103114777414, "grad_norm": 0.4162890911102295, "learning_rate": 3.966549230637014e-05, "loss": 1.9450984954833985, "step": 2605 }, { "epoch": 0.18560659934575452, "grad_norm": 0.479143351316452, "learning_rate": 3.961518127242329e-05, "loss": 1.9328466415405274, "step": 2610 }, { "epoch": 0.1859621675437349, "grad_norm": 0.4529990553855896, "learning_rate": 3.9564780154392134e-05, "loss": 2.0496541976928713, "step": 2615 }, { "epoch": 0.18631773574171526, "grad_norm": 0.4337157905101776, "learning_rate": 3.9514289262935736e-05, "loss": 1.758957290649414, "step": 2620 }, { "epoch": 0.18667330393969564, "grad_norm": 0.4860072433948517, "learning_rate": 3.946370890926649e-05, "loss": 1.97125186920166, "step": 2625 }, { "epoch": 0.187028872137676, "grad_norm": 0.44373461604118347, "learning_rate": 3.9413039405148256e-05, "loss": 1.8010637283325195, "step": 2630 }, { "epoch": 0.18738444033565638, "grad_norm": 0.46480804681777954, "learning_rate": 3.9362281062894326e-05, "loss": 1.7850826263427735, "step": 2635 }, { "epoch": 0.18774000853363676, "grad_norm": 0.42770543694496155, "learning_rate": 3.931143419536561e-05, "loss": 1.910548210144043, "step": 2640 }, { "epoch": 0.18809557673161711, "grad_norm": 0.4126070737838745, "learning_rate": 3.9260499115968675e-05, "loss": 1.855506134033203, "step": 2645 }, { "epoch": 0.1884511449295975, "grad_norm": 0.4433591961860657, "learning_rate": 3.9209476138653756e-05, "loss": 1.987436866760254, "step": 2650 }, { "epoch": 0.18880671312757788, "grad_norm": 0.4568868577480316, "learning_rate": 3.915836557791291e-05, "loss": 1.861073875427246, "step": 2655 }, { "epoch": 0.18916228132555823, "grad_norm": 0.4694977104663849, "learning_rate": 3.910716774877802e-05, "loss": 1.6989480972290039, "step": 2660 }, { "epoch": 0.18951784952353862, "grad_norm": 0.42865920066833496, "learning_rate": 3.905588296681886e-05, "loss": 1.9437641143798827, "step": 2665 }, { "epoch": 0.189873417721519, "grad_norm": 0.4376455545425415, "learning_rate": 3.900451154814116e-05, "loss": 1.7843511581420899, "step": 2670 }, { "epoch": 0.19022898591949935, "grad_norm": 0.44100579619407654, "learning_rate": 3.8953053809384674e-05, "loss": 1.9804019927978516, "step": 2675 }, { "epoch": 0.19058455411747974, "grad_norm": 0.3973184823989868, "learning_rate": 3.890151006772119e-05, "loss": 1.9173244476318358, "step": 2680 }, { "epoch": 0.1909401223154601, "grad_norm": 0.41169753670692444, "learning_rate": 3.8849880640852606e-05, "loss": 2.0243906021118163, "step": 2685 }, { "epoch": 0.19129569051344048, "grad_norm": 0.4053976833820343, "learning_rate": 3.879816584700895e-05, "loss": 1.8626022338867188, "step": 2690 }, { "epoch": 0.19165125871142086, "grad_norm": 0.4549263119697571, "learning_rate": 3.874636600494644e-05, "loss": 1.9270931243896485, "step": 2695 }, { "epoch": 0.1920068269094012, "grad_norm": 0.4794546663761139, "learning_rate": 3.869448143394551e-05, "loss": 1.9335840225219727, "step": 2700 }, { "epoch": 0.1923623951073816, "grad_norm": 0.42400598526000977, "learning_rate": 3.864251245380882e-05, "loss": 1.8423833847045898, "step": 2705 }, { "epoch": 0.19271796330536198, "grad_norm": 0.567085862159729, "learning_rate": 3.859045938485934e-05, "loss": 1.8898155212402343, "step": 2710 }, { "epoch": 0.19307353150334233, "grad_norm": 0.4510025680065155, "learning_rate": 3.853832254793831e-05, "loss": 1.890714454650879, "step": 2715 }, { "epoch": 0.19342909970132272, "grad_norm": 0.5319476127624512, "learning_rate": 3.84861022644033e-05, "loss": 1.9854911804199218, "step": 2720 }, { "epoch": 0.1937846678993031, "grad_norm": 0.514369010925293, "learning_rate": 3.8433798856126245e-05, "loss": 1.9636390686035157, "step": 2725 }, { "epoch": 0.19414023609728345, "grad_norm": 0.4302152693271637, "learning_rate": 3.8381412645491396e-05, "loss": 1.8761634826660156, "step": 2730 }, { "epoch": 0.19449580429526384, "grad_norm": 0.45327043533325195, "learning_rate": 3.832894395539343e-05, "loss": 1.8346946716308594, "step": 2735 }, { "epoch": 0.1948513724932442, "grad_norm": 0.44583582878112793, "learning_rate": 3.8276393109235353e-05, "loss": 1.8555770874023438, "step": 2740 }, { "epoch": 0.19520694069122457, "grad_norm": 0.41672518849372864, "learning_rate": 3.82237604309266e-05, "loss": 1.8397018432617187, "step": 2745 }, { "epoch": 0.19556250888920496, "grad_norm": 0.38072824478149414, "learning_rate": 3.817104624488096e-05, "loss": 1.740152931213379, "step": 2750 }, { "epoch": 0.1959180770871853, "grad_norm": 0.4954932630062103, "learning_rate": 3.811825087601466e-05, "loss": 1.8945642471313477, "step": 2755 }, { "epoch": 0.1962736452851657, "grad_norm": 0.4111610949039459, "learning_rate": 3.806537464974428e-05, "loss": 1.9599859237670898, "step": 2760 }, { "epoch": 0.19662921348314608, "grad_norm": 0.44979432225227356, "learning_rate": 3.801241789198478e-05, "loss": 1.964502716064453, "step": 2765 }, { "epoch": 0.19698478168112643, "grad_norm": 0.4678930938243866, "learning_rate": 3.7959380929147526e-05, "loss": 1.9818527221679687, "step": 2770 }, { "epoch": 0.1973403498791068, "grad_norm": 0.45817139744758606, "learning_rate": 3.790626408813822e-05, "loss": 1.9622344970703125, "step": 2775 }, { "epoch": 0.1976959180770872, "grad_norm": 0.3951300382614136, "learning_rate": 3.785306769635493e-05, "loss": 1.8778635025024415, "step": 2780 }, { "epoch": 0.19805148627506755, "grad_norm": 0.40623605251312256, "learning_rate": 3.7799792081686036e-05, "loss": 1.9926908493041993, "step": 2785 }, { "epoch": 0.19840705447304793, "grad_norm": 0.4403471052646637, "learning_rate": 3.7746437572508244e-05, "loss": 1.941887664794922, "step": 2790 }, { "epoch": 0.19876262267102832, "grad_norm": 0.39597272872924805, "learning_rate": 3.7693004497684526e-05, "loss": 1.9125400543212892, "step": 2795 }, { "epoch": 0.19911819086900867, "grad_norm": 0.4873802661895752, "learning_rate": 3.7639493186562136e-05, "loss": 1.8960399627685547, "step": 2800 }, { "epoch": 0.19947375906698905, "grad_norm": 0.45440152287483215, "learning_rate": 3.7585903968970527e-05, "loss": 1.9094364166259765, "step": 2805 }, { "epoch": 0.1998293272649694, "grad_norm": 0.4439091980457306, "learning_rate": 3.753223717521938e-05, "loss": 2.066347122192383, "step": 2810 }, { "epoch": 0.2001848954629498, "grad_norm": 0.4415077567100525, "learning_rate": 3.7478493136096495e-05, "loss": 1.9177152633666992, "step": 2815 }, { "epoch": 0.20054046366093017, "grad_norm": 0.48137497901916504, "learning_rate": 3.7424672182865825e-05, "loss": 1.8980186462402344, "step": 2820 }, { "epoch": 0.20089603185891053, "grad_norm": 0.39161887764930725, "learning_rate": 3.7370774647265384e-05, "loss": 1.9368228912353516, "step": 2825 }, { "epoch": 0.2012516000568909, "grad_norm": 0.4740235209465027, "learning_rate": 3.731680086150524e-05, "loss": 1.917898178100586, "step": 2830 }, { "epoch": 0.2016071682548713, "grad_norm": 0.4260876178741455, "learning_rate": 3.726275115826542e-05, "loss": 1.9025737762451171, "step": 2835 }, { "epoch": 0.20196273645285165, "grad_norm": 0.41818079352378845, "learning_rate": 3.7208625870693894e-05, "loss": 1.9248634338378907, "step": 2840 }, { "epoch": 0.20231830465083203, "grad_norm": 0.43764036893844604, "learning_rate": 3.7154425332404514e-05, "loss": 1.8459182739257813, "step": 2845 }, { "epoch": 0.20267387284881241, "grad_norm": 0.4003741145133972, "learning_rate": 3.7100149877474974e-05, "loss": 1.964662742614746, "step": 2850 }, { "epoch": 0.20302944104679277, "grad_norm": 0.4278331696987152, "learning_rate": 3.704579984044471e-05, "loss": 1.8564102172851562, "step": 2855 }, { "epoch": 0.20338500924477315, "grad_norm": 0.4577380120754242, "learning_rate": 3.699137555631289e-05, "loss": 1.843560028076172, "step": 2860 }, { "epoch": 0.2037405774427535, "grad_norm": 0.590474009513855, "learning_rate": 3.693687736053627e-05, "loss": 1.7897201538085938, "step": 2865 }, { "epoch": 0.2040961456407339, "grad_norm": 0.4310486912727356, "learning_rate": 3.688230558902725e-05, "loss": 1.8854850769042968, "step": 2870 }, { "epoch": 0.20445171383871427, "grad_norm": 0.4134286642074585, "learning_rate": 3.682766057815168e-05, "loss": 1.8194698333740233, "step": 2875 }, { "epoch": 0.20480728203669463, "grad_norm": 0.44399136304855347, "learning_rate": 3.6772942664726835e-05, "loss": 1.9021604537963868, "step": 2880 }, { "epoch": 0.205162850234675, "grad_norm": 0.4421083331108093, "learning_rate": 3.671815218601938e-05, "loss": 1.8757648468017578, "step": 2885 }, { "epoch": 0.2055184184326554, "grad_norm": 0.42539238929748535, "learning_rate": 3.666328947974322e-05, "loss": 1.9962593078613282, "step": 2890 }, { "epoch": 0.20587398663063575, "grad_norm": 0.4557393789291382, "learning_rate": 3.660835488405746e-05, "loss": 1.9809322357177734, "step": 2895 }, { "epoch": 0.20622955482861613, "grad_norm": 0.4369677007198334, "learning_rate": 3.6553348737564326e-05, "loss": 1.8421112060546876, "step": 2900 }, { "epoch": 0.2065851230265965, "grad_norm": 0.45868298411369324, "learning_rate": 3.649827137930705e-05, "loss": 1.9264961242675782, "step": 2905 }, { "epoch": 0.20694069122457687, "grad_norm": 0.47088003158569336, "learning_rate": 3.644312314876779e-05, "loss": 1.955571746826172, "step": 2910 }, { "epoch": 0.20729625942255725, "grad_norm": 0.44199874997138977, "learning_rate": 3.638790438586555e-05, "loss": 1.890507698059082, "step": 2915 }, { "epoch": 0.20765182762053763, "grad_norm": 0.4656584560871124, "learning_rate": 3.633261543095408e-05, "loss": 1.9046031951904296, "step": 2920 }, { "epoch": 0.208007395818518, "grad_norm": 0.4015985131263733, "learning_rate": 3.6277256624819775e-05, "loss": 1.9170610427856445, "step": 2925 }, { "epoch": 0.20836296401649837, "grad_norm": 0.4420144855976105, "learning_rate": 3.6221828308679564e-05, "loss": 1.9496730804443358, "step": 2930 }, { "epoch": 0.20871853221447872, "grad_norm": 0.4451293647289276, "learning_rate": 3.616633082417882e-05, "loss": 1.9092018127441406, "step": 2935 }, { "epoch": 0.2090741004124591, "grad_norm": 0.4452986419200897, "learning_rate": 3.6110764513389266e-05, "loss": 1.9769542694091797, "step": 2940 }, { "epoch": 0.2094296686104395, "grad_norm": 0.4511799216270447, "learning_rate": 3.6055129718806836e-05, "loss": 1.9193742752075196, "step": 2945 }, { "epoch": 0.20978523680841984, "grad_norm": 0.4489001929759979, "learning_rate": 3.5999426783349565e-05, "loss": 1.9207950592041017, "step": 2950 }, { "epoch": 0.21014080500640023, "grad_norm": 0.4235159158706665, "learning_rate": 3.594365605035553e-05, "loss": 1.8196319580078124, "step": 2955 }, { "epoch": 0.2104963732043806, "grad_norm": 0.47672486305236816, "learning_rate": 3.588781786358068e-05, "loss": 1.8961088180541992, "step": 2960 }, { "epoch": 0.21085194140236097, "grad_norm": 0.4355168342590332, "learning_rate": 3.583191256719672e-05, "loss": 1.7763906478881837, "step": 2965 }, { "epoch": 0.21120750960034135, "grad_norm": 0.4350421726703644, "learning_rate": 3.5775940505788994e-05, "loss": 1.7422569274902344, "step": 2970 }, { "epoch": 0.21156307779832173, "grad_norm": 0.38156455755233765, "learning_rate": 3.5719902024354375e-05, "loss": 1.8697040557861329, "step": 2975 }, { "epoch": 0.21191864599630209, "grad_norm": 0.46926552057266235, "learning_rate": 3.566379746829914e-05, "loss": 1.9621526718139648, "step": 2980 }, { "epoch": 0.21227421419428247, "grad_norm": 0.4785007834434509, "learning_rate": 3.560762718343685e-05, "loss": 1.877945899963379, "step": 2985 }, { "epoch": 0.21262978239226282, "grad_norm": 0.40566059947013855, "learning_rate": 3.555139151598616e-05, "loss": 1.897540283203125, "step": 2990 }, { "epoch": 0.2129853505902432, "grad_norm": 0.46208083629608154, "learning_rate": 3.5495090812568746e-05, "loss": 1.8785659790039062, "step": 2995 }, { "epoch": 0.2133409187882236, "grad_norm": 0.46136927604675293, "learning_rate": 3.543872542020714e-05, "loss": 1.8426242828369142, "step": 3000 }, { "epoch": 0.21369648698620394, "grad_norm": 0.40727317333221436, "learning_rate": 3.538229568632262e-05, "loss": 1.8291913986206054, "step": 3005 }, { "epoch": 0.21405205518418433, "grad_norm": 0.464765727519989, "learning_rate": 3.532580195873303e-05, "loss": 1.8831092834472656, "step": 3010 }, { "epoch": 0.2144076233821647, "grad_norm": 0.4639458656311035, "learning_rate": 3.526924458565065e-05, "loss": 1.9636531829833985, "step": 3015 }, { "epoch": 0.21476319158014506, "grad_norm": 0.5813872218132019, "learning_rate": 3.5212623915680074e-05, "loss": 1.9404579162597657, "step": 3020 }, { "epoch": 0.21511875977812545, "grad_norm": 0.42150169610977173, "learning_rate": 3.515594029781603e-05, "loss": 1.9021928787231446, "step": 3025 }, { "epoch": 0.21547432797610583, "grad_norm": 3.052535057067871, "learning_rate": 3.509919408144123e-05, "loss": 1.8077701568603515, "step": 3030 }, { "epoch": 0.21582989617408618, "grad_norm": 0.4428950250148773, "learning_rate": 3.504238561632424e-05, "loss": 1.8837602615356446, "step": 3035 }, { "epoch": 0.21618546437206657, "grad_norm": 0.4209384024143219, "learning_rate": 3.4985515252617304e-05, "loss": 1.9614784240722656, "step": 3040 }, { "epoch": 0.21654103257004695, "grad_norm": 0.37854117155075073, "learning_rate": 3.4928583340854206e-05, "loss": 1.7799705505371093, "step": 3045 }, { "epoch": 0.2168966007680273, "grad_norm": 0.38608092069625854, "learning_rate": 3.487159023194808e-05, "loss": 1.8397405624389649, "step": 3050 }, { "epoch": 0.21725216896600769, "grad_norm": 0.5505428910255432, "learning_rate": 3.4814536277189264e-05, "loss": 1.9170110702514649, "step": 3055 }, { "epoch": 0.21760773716398804, "grad_norm": 0.47337591648101807, "learning_rate": 3.475742182824314e-05, "loss": 1.735984992980957, "step": 3060 }, { "epoch": 0.21796330536196842, "grad_norm": 0.4421464204788208, "learning_rate": 3.470024723714795e-05, "loss": 1.865607452392578, "step": 3065 }, { "epoch": 0.2183188735599488, "grad_norm": 0.4251149296760559, "learning_rate": 3.464301285631266e-05, "loss": 1.866029930114746, "step": 3070 }, { "epoch": 0.21867444175792916, "grad_norm": 0.5228272676467896, "learning_rate": 3.458571903851474e-05, "loss": 1.7438936233520508, "step": 3075 }, { "epoch": 0.21903000995590954, "grad_norm": 0.4569433033466339, "learning_rate": 3.452836613689803e-05, "loss": 1.842361068725586, "step": 3080 }, { "epoch": 0.21938557815388993, "grad_norm": 0.4055674076080322, "learning_rate": 3.447095450497051e-05, "loss": 1.9235879898071289, "step": 3085 }, { "epoch": 0.21974114635187028, "grad_norm": 0.4330829381942749, "learning_rate": 3.4413484496602225e-05, "loss": 1.8542598724365233, "step": 3090 }, { "epoch": 0.22009671454985066, "grad_norm": 0.4254935085773468, "learning_rate": 3.4355956466022966e-05, "loss": 1.867734718322754, "step": 3095 }, { "epoch": 0.22045228274783105, "grad_norm": 0.47429949045181274, "learning_rate": 3.429837076782021e-05, "loss": 1.888113784790039, "step": 3100 }, { "epoch": 0.2208078509458114, "grad_norm": 0.46978387236595154, "learning_rate": 3.4240727756936844e-05, "loss": 1.9337533950805663, "step": 3105 }, { "epoch": 0.22116341914379178, "grad_norm": 0.5925400257110596, "learning_rate": 3.4183027788669045e-05, "loss": 1.8087146759033204, "step": 3110 }, { "epoch": 0.22151898734177214, "grad_norm": 0.40061086416244507, "learning_rate": 3.412527121866406e-05, "loss": 1.8439825057983399, "step": 3115 }, { "epoch": 0.22187455553975252, "grad_norm": 0.48597919940948486, "learning_rate": 3.406745840291797e-05, "loss": 1.822287368774414, "step": 3120 }, { "epoch": 0.2222301237377329, "grad_norm": 0.44027045369148254, "learning_rate": 3.40095896977736e-05, "loss": 1.8802267074584962, "step": 3125 }, { "epoch": 0.22258569193571326, "grad_norm": 0.3741127550601959, "learning_rate": 3.395166545991822e-05, "loss": 1.8993608474731445, "step": 3130 }, { "epoch": 0.22294126013369364, "grad_norm": 0.44853973388671875, "learning_rate": 3.389368604638141e-05, "loss": 1.903767204284668, "step": 3135 }, { "epoch": 0.22329682833167402, "grad_norm": 0.6637639999389648, "learning_rate": 3.3835651814532826e-05, "loss": 1.867987823486328, "step": 3140 }, { "epoch": 0.22365239652965438, "grad_norm": 0.4271154999732971, "learning_rate": 3.377756312208001e-05, "loss": 1.9267049789428712, "step": 3145 }, { "epoch": 0.22400796472763476, "grad_norm": 0.41948699951171875, "learning_rate": 3.3719420327066196e-05, "loss": 1.9203489303588868, "step": 3150 }, { "epoch": 0.22436353292561514, "grad_norm": 0.4531422555446625, "learning_rate": 3.3661223787868094e-05, "loss": 1.886216926574707, "step": 3155 }, { "epoch": 0.2247191011235955, "grad_norm": 0.4575514793395996, "learning_rate": 3.360297386319365e-05, "loss": 1.988947868347168, "step": 3160 }, { "epoch": 0.22507466932157588, "grad_norm": 0.4614552855491638, "learning_rate": 3.354467091207989e-05, "loss": 1.9559310913085937, "step": 3165 }, { "epoch": 0.22543023751955626, "grad_norm": 0.4452045261859894, "learning_rate": 3.348631529389069e-05, "loss": 1.816739273071289, "step": 3170 }, { "epoch": 0.22578580571753662, "grad_norm": 0.46195533871650696, "learning_rate": 3.3427907368314515e-05, "loss": 1.8018047332763671, "step": 3175 }, { "epoch": 0.226141373915517, "grad_norm": 0.4907339811325073, "learning_rate": 3.336944749536227e-05, "loss": 1.860469436645508, "step": 3180 }, { "epoch": 0.22649694211349736, "grad_norm": 0.46710848808288574, "learning_rate": 3.331093603536504e-05, "loss": 1.757431411743164, "step": 3185 }, { "epoch": 0.22685251031147774, "grad_norm": 0.44681698083877563, "learning_rate": 3.325237334897188e-05, "loss": 1.8972185134887696, "step": 3190 }, { "epoch": 0.22720807850945812, "grad_norm": 0.4521268308162689, "learning_rate": 3.319375979714758e-05, "loss": 1.8351263046264648, "step": 3195 }, { "epoch": 0.22756364670743848, "grad_norm": 0.47940996289253235, "learning_rate": 3.313509574117047e-05, "loss": 1.9015771865844726, "step": 3200 }, { "epoch": 0.22791921490541886, "grad_norm": 0.4314177930355072, "learning_rate": 3.307638154263017e-05, "loss": 1.8187896728515625, "step": 3205 }, { "epoch": 0.22827478310339924, "grad_norm": 0.5055888295173645, "learning_rate": 3.301761756342535e-05, "loss": 1.9884153366088868, "step": 3210 }, { "epoch": 0.2286303513013796, "grad_norm": 0.4193253815174103, "learning_rate": 3.295880416576153e-05, "loss": 1.9399215698242187, "step": 3215 }, { "epoch": 0.22898591949935998, "grad_norm": 0.40070804953575134, "learning_rate": 3.289994171214882e-05, "loss": 1.9400053024291992, "step": 3220 }, { "epoch": 0.22934148769734036, "grad_norm": 0.44406405091285706, "learning_rate": 3.28410305653997e-05, "loss": 1.9302270889282227, "step": 3225 }, { "epoch": 0.22969705589532072, "grad_norm": 0.4688372313976288, "learning_rate": 3.278207108862681e-05, "loss": 1.8604305267333985, "step": 3230 }, { "epoch": 0.2300526240933011, "grad_norm": 0.436074435710907, "learning_rate": 3.272306364524064e-05, "loss": 1.832611083984375, "step": 3235 }, { "epoch": 0.23040819229128146, "grad_norm": 0.7538519501686096, "learning_rate": 3.266400859894736e-05, "loss": 1.9801265716552734, "step": 3240 }, { "epoch": 0.23076376048926184, "grad_norm": 0.4638977646827698, "learning_rate": 3.260490631374653e-05, "loss": 1.9265567779541015, "step": 3245 }, { "epoch": 0.23111932868724222, "grad_norm": 0.4227128326892853, "learning_rate": 3.2545757153928924e-05, "loss": 1.803415298461914, "step": 3250 }, { "epoch": 0.23147489688522258, "grad_norm": 0.40591591596603394, "learning_rate": 3.2486561484074174e-05, "loss": 1.7933372497558593, "step": 3255 }, { "epoch": 0.23183046508320296, "grad_norm": 0.46810010075569153, "learning_rate": 3.242731966904865e-05, "loss": 1.8969127655029296, "step": 3260 }, { "epoch": 0.23218603328118334, "grad_norm": 0.4573196768760681, "learning_rate": 3.23680320740031e-05, "loss": 1.8858961105346679, "step": 3265 }, { "epoch": 0.2325416014791637, "grad_norm": 0.46448004245758057, "learning_rate": 3.230869906437048e-05, "loss": 2.0057262420654296, "step": 3270 }, { "epoch": 0.23289716967714408, "grad_norm": 0.4548952579498291, "learning_rate": 3.2249321005863644e-05, "loss": 2.0089933395385744, "step": 3275 }, { "epoch": 0.23325273787512446, "grad_norm": 0.47056373953819275, "learning_rate": 3.218989826447313e-05, "loss": 1.944521141052246, "step": 3280 }, { "epoch": 0.23360830607310482, "grad_norm": 0.41815581917762756, "learning_rate": 3.2130431206464896e-05, "loss": 1.866251564025879, "step": 3285 }, { "epoch": 0.2339638742710852, "grad_norm": 0.4442651569843292, "learning_rate": 3.207092019837804e-05, "loss": 1.837974739074707, "step": 3290 }, { "epoch": 0.23431944246906555, "grad_norm": 0.44275951385498047, "learning_rate": 3.201136560702258e-05, "loss": 1.9275199890136718, "step": 3295 }, { "epoch": 0.23467501066704594, "grad_norm": 0.4318724572658539, "learning_rate": 3.195176779947713e-05, "loss": 1.9060256958007813, "step": 3300 }, { "epoch": 0.23503057886502632, "grad_norm": 0.46825987100601196, "learning_rate": 3.1892127143086714e-05, "loss": 2.031299018859863, "step": 3305 }, { "epoch": 0.23538614706300667, "grad_norm": 0.43244409561157227, "learning_rate": 3.1832444005460446e-05, "loss": 1.7627372741699219, "step": 3310 }, { "epoch": 0.23574171526098706, "grad_norm": 0.42602115869522095, "learning_rate": 3.177271875446929e-05, "loss": 1.8942045211791991, "step": 3315 }, { "epoch": 0.23609728345896744, "grad_norm": 0.48421287536621094, "learning_rate": 3.171295175824379e-05, "loss": 1.881981658935547, "step": 3320 }, { "epoch": 0.2364528516569478, "grad_norm": 0.43483883142471313, "learning_rate": 3.165314338517179e-05, "loss": 1.9323589324951171, "step": 3325 }, { "epoch": 0.23680841985492818, "grad_norm": 0.4247824251651764, "learning_rate": 3.159329400389615e-05, "loss": 1.907724952697754, "step": 3330 }, { "epoch": 0.23716398805290856, "grad_norm": 0.45709213614463806, "learning_rate": 3.1533403983312546e-05, "loss": 1.902322006225586, "step": 3335 }, { "epoch": 0.2375195562508889, "grad_norm": 0.38317933678627014, "learning_rate": 3.1473473692567086e-05, "loss": 1.8484146118164062, "step": 3340 }, { "epoch": 0.2378751244488693, "grad_norm": 0.3948182165622711, "learning_rate": 3.141350350105413e-05, "loss": 1.9240619659423828, "step": 3345 }, { "epoch": 0.23823069264684968, "grad_norm": 0.3932666778564453, "learning_rate": 3.135349377841396e-05, "loss": 1.7401372909545898, "step": 3350 }, { "epoch": 0.23858626084483003, "grad_norm": 0.4829767346382141, "learning_rate": 3.129344489453053e-05, "loss": 1.9886505126953125, "step": 3355 }, { "epoch": 0.23894182904281042, "grad_norm": 0.45809653401374817, "learning_rate": 3.123335721952916e-05, "loss": 1.760655975341797, "step": 3360 }, { "epoch": 0.23929739724079077, "grad_norm": 0.4620189368724823, "learning_rate": 3.117323112377427e-05, "loss": 1.8827175140380858, "step": 3365 }, { "epoch": 0.23965296543877115, "grad_norm": 0.39445456862449646, "learning_rate": 3.1113066977867105e-05, "loss": 1.8876874923706055, "step": 3370 }, { "epoch": 0.24000853363675154, "grad_norm": 0.4053409695625305, "learning_rate": 3.105286515264343e-05, "loss": 1.792802619934082, "step": 3375 }, { "epoch": 0.2403641018347319, "grad_norm": 0.4884132742881775, "learning_rate": 3.099262601917125e-05, "loss": 2.020514488220215, "step": 3380 }, { "epoch": 0.24071967003271227, "grad_norm": 0.4487616717815399, "learning_rate": 3.093234994874855e-05, "loss": 1.849205207824707, "step": 3385 }, { "epoch": 0.24107523823069266, "grad_norm": 0.5110529661178589, "learning_rate": 3.0872037312900984e-05, "loss": 2.01705265045166, "step": 3390 }, { "epoch": 0.241430806428673, "grad_norm": 0.4572543203830719, "learning_rate": 3.081168848337955e-05, "loss": 1.9392314910888673, "step": 3395 }, { "epoch": 0.2417863746266534, "grad_norm": 0.4689153730869293, "learning_rate": 3.0751303832158364e-05, "loss": 1.9841468811035157, "step": 3400 }, { "epoch": 0.24214194282463378, "grad_norm": 0.496160089969635, "learning_rate": 3.069088373143234e-05, "loss": 1.788466453552246, "step": 3405 }, { "epoch": 0.24249751102261413, "grad_norm": 0.4527823328971863, "learning_rate": 3.0630428553614873e-05, "loss": 1.9786788940429687, "step": 3410 }, { "epoch": 0.24285307922059451, "grad_norm": 0.42821618914604187, "learning_rate": 3.056993867133559e-05, "loss": 1.899939727783203, "step": 3415 }, { "epoch": 0.24320864741857487, "grad_norm": 0.4478073716163635, "learning_rate": 3.0509414457438e-05, "loss": 1.929829978942871, "step": 3420 }, { "epoch": 0.24356421561655525, "grad_norm": 0.41961631178855896, "learning_rate": 3.044885628497725e-05, "loss": 1.862752342224121, "step": 3425 }, { "epoch": 0.24391978381453563, "grad_norm": 0.45657920837402344, "learning_rate": 3.038826452721777e-05, "loss": 2.0499540328979493, "step": 3430 }, { "epoch": 0.244275352012516, "grad_norm": 0.4255641996860504, "learning_rate": 3.0327639557631023e-05, "loss": 1.8340015411376953, "step": 3435 }, { "epoch": 0.24463092021049637, "grad_norm": 0.3840986490249634, "learning_rate": 3.0266981749893157e-05, "loss": 1.9083328247070312, "step": 3440 }, { "epoch": 0.24498648840847675, "grad_norm": 0.4592842757701874, "learning_rate": 3.0206291477882763e-05, "loss": 1.8760814666748047, "step": 3445 }, { "epoch": 0.2453420566064571, "grad_norm": 0.419474333524704, "learning_rate": 3.0145569115678507e-05, "loss": 1.8913700103759765, "step": 3450 }, { "epoch": 0.2456976248044375, "grad_norm": 0.4511104226112366, "learning_rate": 3.0084815037556846e-05, "loss": 1.9158851623535156, "step": 3455 }, { "epoch": 0.24605319300241787, "grad_norm": 0.4548083245754242, "learning_rate": 3.002402961798974e-05, "loss": 1.8002714157104491, "step": 3460 }, { "epoch": 0.24640876120039823, "grad_norm": 0.45247095823287964, "learning_rate": 2.9963213231642322e-05, "loss": 1.8398433685302735, "step": 3465 }, { "epoch": 0.2467643293983786, "grad_norm": 0.46995794773101807, "learning_rate": 2.9902366253370594e-05, "loss": 1.885784912109375, "step": 3470 }, { "epoch": 0.247119897596359, "grad_norm": 0.46733954548835754, "learning_rate": 2.9841489058219135e-05, "loss": 2.0267318725585937, "step": 3475 }, { "epoch": 0.24747546579433935, "grad_norm": 0.35099419951438904, "learning_rate": 2.9780582021418748e-05, "loss": 1.8469058990478515, "step": 3480 }, { "epoch": 0.24783103399231973, "grad_norm": 0.46084722876548767, "learning_rate": 2.9719645518384193e-05, "loss": 1.6639976501464844, "step": 3485 }, { "epoch": 0.2481866021903001, "grad_norm": 0.4609321653842926, "learning_rate": 2.9658679924711835e-05, "loss": 1.8680734634399414, "step": 3490 }, { "epoch": 0.24854217038828047, "grad_norm": 0.4145055115222931, "learning_rate": 2.959768561617735e-05, "loss": 1.8442060470581054, "step": 3495 }, { "epoch": 0.24889773858626085, "grad_norm": 0.4644194543361664, "learning_rate": 2.953666296873342e-05, "loss": 1.9354122161865235, "step": 3500 }, { "epoch": 0.2492533067842412, "grad_norm": 0.44665950536727905, "learning_rate": 2.947561235850739e-05, "loss": 1.973103904724121, "step": 3505 }, { "epoch": 0.2496088749822216, "grad_norm": 0.49052587151527405, "learning_rate": 2.9414534161798947e-05, "loss": 1.9295101165771484, "step": 3510 }, { "epoch": 0.24996444318020197, "grad_norm": 0.433015912771225, "learning_rate": 2.9353428755077833e-05, "loss": 1.8787811279296875, "step": 3515 }, { "epoch": 0.2503200113781823, "grad_norm": 0.4246494472026825, "learning_rate": 2.9292296514981503e-05, "loss": 1.918088150024414, "step": 3520 }, { "epoch": 0.2506755795761627, "grad_norm": 0.47748205065727234, "learning_rate": 2.9231137818312798e-05, "loss": 1.9624879837036133, "step": 3525 }, { "epoch": 0.2510311477741431, "grad_norm": 0.43574258685112, "learning_rate": 2.9169953042037623e-05, "loss": 1.9083379745483398, "step": 3530 }, { "epoch": 0.2513867159721235, "grad_norm": 0.45508214831352234, "learning_rate": 2.9108742563282652e-05, "loss": 1.8951108932495118, "step": 3535 }, { "epoch": 0.2517422841701038, "grad_norm": 0.450115829706192, "learning_rate": 2.9047506759332978e-05, "loss": 1.837081527709961, "step": 3540 }, { "epoch": 0.2520978523680842, "grad_norm": 0.412214070558548, "learning_rate": 2.898624600762977e-05, "loss": 1.8927413940429687, "step": 3545 }, { "epoch": 0.25245342056606457, "grad_norm": 0.4411039650440216, "learning_rate": 2.8924960685767987e-05, "loss": 2.011094665527344, "step": 3550 }, { "epoch": 0.25280898876404495, "grad_norm": 0.4123626947402954, "learning_rate": 2.8863651171494043e-05, "loss": 1.9728477478027344, "step": 3555 }, { "epoch": 0.25316455696202533, "grad_norm": 0.40887388586997986, "learning_rate": 2.8802317842703437e-05, "loss": 1.9826229095458985, "step": 3560 }, { "epoch": 0.2535201251600057, "grad_norm": 0.4192020297050476, "learning_rate": 2.8740961077438467e-05, "loss": 1.8812923431396484, "step": 3565 }, { "epoch": 0.25387569335798604, "grad_norm": 0.4507167935371399, "learning_rate": 2.8679581253885902e-05, "loss": 1.954871940612793, "step": 3570 }, { "epoch": 0.2542312615559664, "grad_norm": 0.4303611218929291, "learning_rate": 2.861817875037462e-05, "loss": 1.9758554458618165, "step": 3575 }, { "epoch": 0.2545868297539468, "grad_norm": 1.4705606698989868, "learning_rate": 2.85567539453733e-05, "loss": 1.8146490097045898, "step": 3580 }, { "epoch": 0.2549423979519272, "grad_norm": 0.6645826101303101, "learning_rate": 2.8495307217488076e-05, "loss": 1.8896450042724608, "step": 3585 }, { "epoch": 0.2552979661499076, "grad_norm": 0.458879679441452, "learning_rate": 2.8433838945460205e-05, "loss": 1.8284208297729492, "step": 3590 }, { "epoch": 0.2556535343478879, "grad_norm": 0.4241970479488373, "learning_rate": 2.8372349508163752e-05, "loss": 1.8203422546386718, "step": 3595 }, { "epoch": 0.2560091025458683, "grad_norm": 0.42726466059684753, "learning_rate": 2.831083928460323e-05, "loss": 1.9445878982543945, "step": 3600 }, { "epoch": 0.25636467074384867, "grad_norm": 0.4602843225002289, "learning_rate": 2.8249308653911268e-05, "loss": 1.868157958984375, "step": 3605 }, { "epoch": 0.25672023894182905, "grad_norm": 0.45200037956237793, "learning_rate": 2.8187757995346275e-05, "loss": 1.9608675003051759, "step": 3610 }, { "epoch": 0.25707580713980943, "grad_norm": 0.5012028813362122, "learning_rate": 2.812618768829013e-05, "loss": 1.865737533569336, "step": 3615 }, { "epoch": 0.2574313753377898, "grad_norm": 0.46086812019348145, "learning_rate": 2.8064598112245798e-05, "loss": 1.844839096069336, "step": 3620 }, { "epoch": 0.25778694353577014, "grad_norm": 0.4700728952884674, "learning_rate": 2.8002989646835022e-05, "loss": 1.922987174987793, "step": 3625 }, { "epoch": 0.2581425117337505, "grad_norm": 0.43208152055740356, "learning_rate": 2.794136267179596e-05, "loss": 1.8647428512573243, "step": 3630 }, { "epoch": 0.2584980799317309, "grad_norm": 0.39739590883255005, "learning_rate": 2.78797175669809e-05, "loss": 1.7979557037353515, "step": 3635 }, { "epoch": 0.2588536481297113, "grad_norm": 0.4390626847743988, "learning_rate": 2.7818054712353826e-05, "loss": 1.7968938827514649, "step": 3640 }, { "epoch": 0.25920921632769167, "grad_norm": 0.40871667861938477, "learning_rate": 2.7756374487988142e-05, "loss": 1.810888671875, "step": 3645 }, { "epoch": 0.259564784525672, "grad_norm": 0.4651896357536316, "learning_rate": 2.769467727406434e-05, "loss": 1.952181625366211, "step": 3650 }, { "epoch": 0.2599203527236524, "grad_norm": 0.5090406537055969, "learning_rate": 2.7632963450867598e-05, "loss": 1.9274011611938477, "step": 3655 }, { "epoch": 0.26027592092163276, "grad_norm": 0.478118896484375, "learning_rate": 2.7571233398785495e-05, "loss": 1.8557363510131837, "step": 3660 }, { "epoch": 0.26063148911961315, "grad_norm": 0.4520687162876129, "learning_rate": 2.7509487498305613e-05, "loss": 1.8489913940429688, "step": 3665 }, { "epoch": 0.26098705731759353, "grad_norm": 0.4085557460784912, "learning_rate": 2.7447726130013258e-05, "loss": 1.759567642211914, "step": 3670 }, { "epoch": 0.2613426255155739, "grad_norm": 0.4582419991493225, "learning_rate": 2.7385949674589035e-05, "loss": 1.9272876739501954, "step": 3675 }, { "epoch": 0.26169819371355424, "grad_norm": 0.515718400478363, "learning_rate": 2.7324158512806564e-05, "loss": 1.910982894897461, "step": 3680 }, { "epoch": 0.2620537619115346, "grad_norm": 0.4637136161327362, "learning_rate": 2.72623530255301e-05, "loss": 1.8107303619384765, "step": 3685 }, { "epoch": 0.262409330109515, "grad_norm": 0.4733792841434479, "learning_rate": 2.7200533593712225e-05, "loss": 1.8875965118408202, "step": 3690 }, { "epoch": 0.2627648983074954, "grad_norm": 0.3809363543987274, "learning_rate": 2.713870059839143e-05, "loss": 1.873642349243164, "step": 3695 }, { "epoch": 0.26312046650547577, "grad_norm": 0.3970220685005188, "learning_rate": 2.707685442068983e-05, "loss": 1.878699493408203, "step": 3700 }, { "epoch": 0.2634760347034561, "grad_norm": 0.4233149588108063, "learning_rate": 2.7014995441810804e-05, "loss": 1.8804447174072265, "step": 3705 }, { "epoch": 0.2638316029014365, "grad_norm": 0.4213905334472656, "learning_rate": 2.6953124043036602e-05, "loss": 1.7753978729248048, "step": 3710 }, { "epoch": 0.26418717109941686, "grad_norm": 0.44220104813575745, "learning_rate": 2.6891240605726055e-05, "loss": 1.8251672744750977, "step": 3715 }, { "epoch": 0.26454273929739724, "grad_norm": 0.4607194662094116, "learning_rate": 2.682934551131218e-05, "loss": 1.7870624542236329, "step": 3720 }, { "epoch": 0.2648983074953776, "grad_norm": 0.41963863372802734, "learning_rate": 2.6767439141299865e-05, "loss": 1.818289566040039, "step": 3725 }, { "epoch": 0.265253875693358, "grad_norm": 0.39648130536079407, "learning_rate": 2.670552187726348e-05, "loss": 1.8761430740356446, "step": 3730 }, { "epoch": 0.26560944389133834, "grad_norm": 0.36391347646713257, "learning_rate": 2.6643594100844545e-05, "loss": 1.7134719848632813, "step": 3735 }, { "epoch": 0.2659650120893187, "grad_norm": 0.4701034426689148, "learning_rate": 2.6581656193749367e-05, "loss": 1.9152936935424805, "step": 3740 }, { "epoch": 0.2663205802872991, "grad_norm": 0.4521518349647522, "learning_rate": 2.651970853774672e-05, "loss": 1.916804313659668, "step": 3745 }, { "epoch": 0.2666761484852795, "grad_norm": 0.4020965099334717, "learning_rate": 2.6457751514665457e-05, "loss": 1.8285030364990233, "step": 3750 }, { "epoch": 0.26703171668325987, "grad_norm": 0.4370921552181244, "learning_rate": 2.639578550639216e-05, "loss": 1.9966455459594727, "step": 3755 }, { "epoch": 0.2673872848812402, "grad_norm": 0.46113938093185425, "learning_rate": 2.6333810894868793e-05, "loss": 1.9330923080444335, "step": 3760 }, { "epoch": 0.2677428530792206, "grad_norm": 0.46066612005233765, "learning_rate": 2.6271828062090366e-05, "loss": 1.9407035827636718, "step": 3765 }, { "epoch": 0.26809842127720096, "grad_norm": 0.4487403929233551, "learning_rate": 2.6209837390102537e-05, "loss": 1.8684104919433593, "step": 3770 }, { "epoch": 0.26845398947518134, "grad_norm": 0.4638257920742035, "learning_rate": 2.6147839260999295e-05, "loss": 1.9254087448120116, "step": 3775 }, { "epoch": 0.2688095576731617, "grad_norm": 0.4703100323677063, "learning_rate": 2.6085834056920604e-05, "loss": 1.8641637802124023, "step": 3780 }, { "epoch": 0.2691651258711421, "grad_norm": 0.4680882692337036, "learning_rate": 2.6023822160050026e-05, "loss": 1.9515697479248046, "step": 3785 }, { "epoch": 0.26952069406912244, "grad_norm": 0.4705103039741516, "learning_rate": 2.5961803952612363e-05, "loss": 1.9231430053710938, "step": 3790 }, { "epoch": 0.2698762622671028, "grad_norm": 0.468534380197525, "learning_rate": 2.5899779816871332e-05, "loss": 1.8248420715332032, "step": 3795 }, { "epoch": 0.2702318304650832, "grad_norm": 0.45770612359046936, "learning_rate": 2.583775013512719e-05, "loss": 1.846468734741211, "step": 3800 }, { "epoch": 0.2705873986630636, "grad_norm": 0.4343346357345581, "learning_rate": 2.5775715289714364e-05, "loss": 1.829201889038086, "step": 3805 }, { "epoch": 0.27094296686104397, "grad_norm": 0.4391036033630371, "learning_rate": 2.571367566299912e-05, "loss": 2.013857841491699, "step": 3810 }, { "epoch": 0.27129853505902435, "grad_norm": 0.39013785123825073, "learning_rate": 2.5651631637377187e-05, "loss": 1.7974613189697266, "step": 3815 }, { "epoch": 0.2716541032570047, "grad_norm": 0.4687812030315399, "learning_rate": 2.5589583595271428e-05, "loss": 1.9664077758789062, "step": 3820 }, { "epoch": 0.27200967145498506, "grad_norm": 0.4532470107078552, "learning_rate": 2.5527531919129428e-05, "loss": 1.8793228149414063, "step": 3825 }, { "epoch": 0.27236523965296544, "grad_norm": 0.44806745648384094, "learning_rate": 2.5465476991421195e-05, "loss": 1.9762014389038085, "step": 3830 }, { "epoch": 0.2727208078509458, "grad_norm": 0.4484991431236267, "learning_rate": 2.5403419194636784e-05, "loss": 1.9042007446289062, "step": 3835 }, { "epoch": 0.2730763760489262, "grad_norm": 0.39517921209335327, "learning_rate": 2.534135891128391e-05, "loss": 1.8867881774902344, "step": 3840 }, { "epoch": 0.27343194424690653, "grad_norm": 0.45223134756088257, "learning_rate": 2.527929652388564e-05, "loss": 1.9678056716918946, "step": 3845 }, { "epoch": 0.2737875124448869, "grad_norm": 0.48139968514442444, "learning_rate": 2.5217232414977982e-05, "loss": 1.9776372909545898, "step": 3850 }, { "epoch": 0.2741430806428673, "grad_norm": 0.49359530210494995, "learning_rate": 2.5155166967107595e-05, "loss": 1.9352890014648438, "step": 3855 }, { "epoch": 0.2744986488408477, "grad_norm": 0.4202648401260376, "learning_rate": 2.5093100562829353e-05, "loss": 1.8656444549560547, "step": 3860 }, { "epoch": 0.27485421703882806, "grad_norm": 0.4220055341720581, "learning_rate": 2.503103358470404e-05, "loss": 1.9660566329956055, "step": 3865 }, { "epoch": 0.27520978523680845, "grad_norm": 0.47227758169174194, "learning_rate": 2.4968966415295966e-05, "loss": 1.9252260208129883, "step": 3870 }, { "epoch": 0.2755653534347888, "grad_norm": 0.40127992630004883, "learning_rate": 2.490689943717066e-05, "loss": 1.8801813125610352, "step": 3875 }, { "epoch": 0.27592092163276916, "grad_norm": 0.4688461124897003, "learning_rate": 2.484483303289241e-05, "loss": 1.8824640274047852, "step": 3880 }, { "epoch": 0.27627648983074954, "grad_norm": 0.4891282916069031, "learning_rate": 2.4782767585022017e-05, "loss": 1.8955413818359375, "step": 3885 }, { "epoch": 0.2766320580287299, "grad_norm": 0.4164116382598877, "learning_rate": 2.4720703476114367e-05, "loss": 1.9216789245605468, "step": 3890 }, { "epoch": 0.2769876262267103, "grad_norm": 0.4366419315338135, "learning_rate": 2.4658641088716093e-05, "loss": 1.8682552337646485, "step": 3895 }, { "epoch": 0.27734319442469063, "grad_norm": 0.4726308286190033, "learning_rate": 2.459658080536323e-05, "loss": 1.9913530349731445, "step": 3900 }, { "epoch": 0.277698762622671, "grad_norm": 0.5157833099365234, "learning_rate": 2.453452300857881e-05, "loss": 1.9050626754760742, "step": 3905 }, { "epoch": 0.2780543308206514, "grad_norm": 0.4678018391132355, "learning_rate": 2.4472468080870585e-05, "loss": 1.7014150619506836, "step": 3910 }, { "epoch": 0.2784098990186318, "grad_norm": 0.8501263856887817, "learning_rate": 2.441041640472858e-05, "loss": 1.9026817321777343, "step": 3915 }, { "epoch": 0.27876546721661216, "grad_norm": 0.4667016863822937, "learning_rate": 2.4348368362622812e-05, "loss": 1.8753887176513673, "step": 3920 }, { "epoch": 0.27912103541459254, "grad_norm": 0.4407086670398712, "learning_rate": 2.4286324337000885e-05, "loss": 1.910672378540039, "step": 3925 }, { "epoch": 0.27947660361257287, "grad_norm": 0.3996715545654297, "learning_rate": 2.4224284710285638e-05, "loss": 1.8863906860351562, "step": 3930 }, { "epoch": 0.27983217181055325, "grad_norm": 0.4187909662723541, "learning_rate": 2.416224986487282e-05, "loss": 1.8694267272949219, "step": 3935 }, { "epoch": 0.28018774000853364, "grad_norm": 0.4541258215904236, "learning_rate": 2.410022018312867e-05, "loss": 1.9233905792236328, "step": 3940 }, { "epoch": 0.280543308206514, "grad_norm": 0.41847071051597595, "learning_rate": 2.403819604738764e-05, "loss": 1.8793302536010743, "step": 3945 }, { "epoch": 0.2808988764044944, "grad_norm": 0.4425404369831085, "learning_rate": 2.3976177839949983e-05, "loss": 1.9386014938354492, "step": 3950 }, { "epoch": 0.28125444460247473, "grad_norm": 0.45773008465766907, "learning_rate": 2.3914165943079395e-05, "loss": 1.8809638977050782, "step": 3955 }, { "epoch": 0.2816100128004551, "grad_norm": 0.4581395387649536, "learning_rate": 2.3852160739000707e-05, "loss": 1.9358495712280273, "step": 3960 }, { "epoch": 0.2819655809984355, "grad_norm": 0.43617910146713257, "learning_rate": 2.379016260989747e-05, "loss": 1.8792373657226562, "step": 3965 }, { "epoch": 0.2823211491964159, "grad_norm": 0.4662131369113922, "learning_rate": 2.3728171937909647e-05, "loss": 1.9976957321166993, "step": 3970 }, { "epoch": 0.28267671739439626, "grad_norm": 0.5340515375137329, "learning_rate": 2.3666189105131212e-05, "loss": 1.8333438873291015, "step": 3975 }, { "epoch": 0.28303228559237664, "grad_norm": 0.41469621658325195, "learning_rate": 2.3604214493607845e-05, "loss": 1.852827262878418, "step": 3980 }, { "epoch": 0.28338785379035697, "grad_norm": 0.4700051248073578, "learning_rate": 2.354224848533455e-05, "loss": 1.8775337219238282, "step": 3985 }, { "epoch": 0.28374342198833735, "grad_norm": 0.442305326461792, "learning_rate": 2.3480291462253277e-05, "loss": 1.9259958267211914, "step": 3990 }, { "epoch": 0.28409899018631773, "grad_norm": 0.44582369923591614, "learning_rate": 2.3418343806250645e-05, "loss": 1.692205047607422, "step": 3995 }, { "epoch": 0.2844545583842981, "grad_norm": 0.4139062762260437, "learning_rate": 2.335640589915546e-05, "loss": 1.7994419097900392, "step": 4000 } ], "logging_steps": 5, "max_steps": 7030, "num_input_tokens_seen": 0, "num_train_epochs": 1, "save_steps": 500, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": false }, "attributes": {} } }, "total_flos": 5.562399575580165e+18, "train_batch_size": 16, "trial_name": null, "trial_params": null }