diff --git "a/MultiDialect-ASR/trainer_state.json" "b/MultiDialect-ASR/trainer_state.json" new file mode 100644--- /dev/null +++ "b/MultiDialect-ASR/trainer_state.json" @@ -0,0 +1,25930 @@ +{ + "best_global_step": 12000, + "best_metric": 0.23060274, + "best_model_checkpoint": "/home/work_nfs23/sywang/code/ms-swift/examples/qwen3asr/exp3/exp_stage6_zh_en/v7-20260708-235534/checkpoint-12000", + "epoch": 0.9931102973123953, + "eval_steps": 500, + "global_step": 16000, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 6.481931615621455e-05, + "grad_norm": 2.655649185180664, + "learning_rate": 1.295336787564767e-08, + "loss": 0.3341776132583618, + "step": 1, + "token_acc": 0.9117376841438302 + }, + { + "epoch": 0.00032409658078107277, + "grad_norm": 2.3714418411254883, + "learning_rate": 6.476683937823834e-08, + "loss": 0.32723408937454224, + "step": 5, + "token_acc": 0.9133364880962094 + }, + { + "epoch": 0.0006481931615621455, + "grad_norm": 2.3853228092193604, + "learning_rate": 1.2953367875647668e-07, + "loss": 0.31995859146118166, + "step": 10, + "token_acc": 0.9147802247075855 + }, + { + "epoch": 0.0009722897423432182, + "grad_norm": 2.6294362545013428, + "learning_rate": 1.9430051813471504e-07, + "loss": 0.3260782241821289, + "step": 15, + "token_acc": 0.9135075595651775 + }, + { + "epoch": 0.001296386323124291, + "grad_norm": 2.2157599925994873, + "learning_rate": 2.5906735751295336e-07, + "loss": 0.3227314233779907, + "step": 20, + "token_acc": 0.9144817968765132 + }, + { + "epoch": 0.0016204829039053639, + "grad_norm": 2.3223657608032227, + "learning_rate": 3.238341968911917e-07, + "loss": 0.3281126022338867, + "step": 25, + "token_acc": 0.9134876140808345 + }, + { + "epoch": 0.0019445794846864365, + "grad_norm": 1.8758041858673096, + "learning_rate": 3.886010362694301e-07, + "loss": 0.3164376258850098, + "step": 30, + "token_acc": 0.9155102518125828 + }, + { + "epoch": 0.0022686760654675093, + "grad_norm": 1.8605087995529175, + "learning_rate": 4.5336787564766845e-07, + "loss": 0.3189058303833008, + "step": 35, + "token_acc": 0.9153365897842332 + }, + { + "epoch": 0.002592772646248582, + "grad_norm": 1.5158028602600098, + "learning_rate": 5.181347150259067e-07, + "loss": 0.31297638416290285, + "step": 40, + "token_acc": 0.916547756341341 + }, + { + "epoch": 0.002916869227029655, + "grad_norm": 1.369957685470581, + "learning_rate": 5.829015544041451e-07, + "loss": 0.3067481994628906, + "step": 45, + "token_acc": 0.916672624347634 + }, + { + "epoch": 0.0032409658078107278, + "grad_norm": 1.3565272092819214, + "learning_rate": 6.476683937823834e-07, + "loss": 0.30818629264831543, + "step": 50, + "token_acc": 0.9172184122888997 + }, + { + "epoch": 0.0035650623885918, + "grad_norm": 1.7045643329620361, + "learning_rate": 7.124352331606219e-07, + "loss": 0.31780500411987306, + "step": 55, + "token_acc": 0.9151205018157808 + }, + { + "epoch": 0.003889158969372873, + "grad_norm": 1.3580641746520996, + "learning_rate": 7.772020725388602e-07, + "loss": 0.30206310749053955, + "step": 60, + "token_acc": 0.9192551764849309 + }, + { + "epoch": 0.004213255550153946, + "grad_norm": 1.372952938079834, + "learning_rate": 8.419689119170985e-07, + "loss": 0.3010685443878174, + "step": 65, + "token_acc": 0.9178735707714548 + }, + { + "epoch": 0.004537352130935019, + "grad_norm": 1.3031820058822632, + "learning_rate": 9.067357512953369e-07, + "loss": 0.29661927223205564, + "step": 70, + "token_acc": 0.9191524095598799 + }, + { + "epoch": 0.004861448711716091, + "grad_norm": 1.260411024093628, + "learning_rate": 9.715025906735752e-07, + "loss": 0.3009307861328125, + "step": 75, + "token_acc": 0.9196397599272189 + }, + { + "epoch": 0.005185545292497164, + "grad_norm": 1.2549397945404053, + "learning_rate": 1.0362694300518134e-06, + "loss": 0.2917083740234375, + "step": 80, + "token_acc": 0.9206776871569922 + }, + { + "epoch": 0.005509641873278237, + "grad_norm": 1.2358394861221313, + "learning_rate": 1.1010362694300518e-06, + "loss": 0.29244284629821776, + "step": 85, + "token_acc": 0.9203724671065474 + }, + { + "epoch": 0.00583373845405931, + "grad_norm": 1.2670187950134277, + "learning_rate": 1.1658031088082903e-06, + "loss": 0.29021873474121096, + "step": 90, + "token_acc": 0.9205806383595583 + }, + { + "epoch": 0.006157835034840383, + "grad_norm": 1.128826379776001, + "learning_rate": 1.2305699481865285e-06, + "loss": 0.2890714168548584, + "step": 95, + "token_acc": 0.9221680580499583 + }, + { + "epoch": 0.0064819316156214555, + "grad_norm": 1.2712926864624023, + "learning_rate": 1.2953367875647669e-06, + "loss": 0.288629150390625, + "step": 100, + "token_acc": 0.9215318116919947 + }, + { + "epoch": 0.006806028196402528, + "grad_norm": 1.3695846796035767, + "learning_rate": 1.3601036269430053e-06, + "loss": 0.2901022911071777, + "step": 105, + "token_acc": 0.9207788849772262 + }, + { + "epoch": 0.0071301247771836, + "grad_norm": 1.181462049484253, + "learning_rate": 1.4248704663212437e-06, + "loss": 0.2921628475189209, + "step": 110, + "token_acc": 0.9201108559458211 + }, + { + "epoch": 0.007454221357964673, + "grad_norm": 1.1615545749664307, + "learning_rate": 1.489637305699482e-06, + "loss": 0.27958681583404543, + "step": 115, + "token_acc": 0.9226613535406324 + }, + { + "epoch": 0.007778317938745746, + "grad_norm": 1.1601577997207642, + "learning_rate": 1.5544041450777204e-06, + "loss": 0.2812881231307983, + "step": 120, + "token_acc": 0.9219507427677873 + }, + { + "epoch": 0.00810241451952682, + "grad_norm": 1.1718589067459106, + "learning_rate": 1.6191709844559586e-06, + "loss": 0.28042919635772706, + "step": 125, + "token_acc": 0.922784509384652 + }, + { + "epoch": 0.008426511100307892, + "grad_norm": 1.1261916160583496, + "learning_rate": 1.683937823834197e-06, + "loss": 0.28291802406311034, + "step": 130, + "token_acc": 0.9216872217187438 + }, + { + "epoch": 0.008750607681088965, + "grad_norm": 1.2394858598709106, + "learning_rate": 1.7487046632124354e-06, + "loss": 0.2855110168457031, + "step": 135, + "token_acc": 0.9219139435628654 + }, + { + "epoch": 0.009074704261870037, + "grad_norm": 1.1633332967758179, + "learning_rate": 1.8134715025906738e-06, + "loss": 0.2800710916519165, + "step": 140, + "token_acc": 0.9219525269495525 + }, + { + "epoch": 0.00939880084265111, + "grad_norm": 1.0819878578186035, + "learning_rate": 1.878238341968912e-06, + "loss": 0.2783178806304932, + "step": 145, + "token_acc": 0.9228247096782621 + }, + { + "epoch": 0.009722897423432183, + "grad_norm": 1.2350913286209106, + "learning_rate": 1.9430051813471504e-06, + "loss": 0.2704397439956665, + "step": 150, + "token_acc": 0.9243719866405021 + }, + { + "epoch": 0.010046994004213255, + "grad_norm": 1.2952532768249512, + "learning_rate": 2.0077720207253886e-06, + "loss": 0.2756261110305786, + "step": 155, + "token_acc": 0.923563500377939 + }, + { + "epoch": 0.010371090584994328, + "grad_norm": 2.835852861404419, + "learning_rate": 2.072538860103627e-06, + "loss": 0.27072920799255373, + "step": 160, + "token_acc": 0.9240289571712448 + }, + { + "epoch": 0.0106951871657754, + "grad_norm": 1.2432365417480469, + "learning_rate": 2.1373056994818655e-06, + "loss": 0.28088808059692383, + "step": 165, + "token_acc": 0.9218431443392088 + }, + { + "epoch": 0.011019283746556474, + "grad_norm": 1.1758121252059937, + "learning_rate": 2.2020725388601037e-06, + "loss": 0.2745814323425293, + "step": 170, + "token_acc": 0.923520129422373 + }, + { + "epoch": 0.011343380327337546, + "grad_norm": 1.1752557754516602, + "learning_rate": 2.2668393782383423e-06, + "loss": 0.284597110748291, + "step": 175, + "token_acc": 0.921214273500901 + }, + { + "epoch": 0.01166747690811862, + "grad_norm": 1.1469591856002808, + "learning_rate": 2.3316062176165805e-06, + "loss": 0.27476656436920166, + "step": 180, + "token_acc": 0.9236426183753611 + }, + { + "epoch": 0.011991573488899692, + "grad_norm": 1.1166425943374634, + "learning_rate": 2.3963730569948187e-06, + "loss": 0.2817973613739014, + "step": 185, + "token_acc": 0.9223100209377249 + }, + { + "epoch": 0.012315670069680765, + "grad_norm": 1.1793495416641235, + "learning_rate": 2.461139896373057e-06, + "loss": 0.27724857330322267, + "step": 190, + "token_acc": 0.9227060065421657 + }, + { + "epoch": 0.012639766650461837, + "grad_norm": 1.1739124059677124, + "learning_rate": 2.5259067357512956e-06, + "loss": 0.28834104537963867, + "step": 195, + "token_acc": 0.9206041147010362 + }, + { + "epoch": 0.012963863231242911, + "grad_norm": 1.135248064994812, + "learning_rate": 2.5906735751295338e-06, + "loss": 0.27553977966308596, + "step": 200, + "token_acc": 0.9237138096229952 + }, + { + "epoch": 0.013287959812023983, + "grad_norm": 1.1080269813537598, + "learning_rate": 2.6554404145077724e-06, + "loss": 0.2846039295196533, + "step": 205, + "token_acc": 0.9222145902722878 + }, + { + "epoch": 0.013612056392805057, + "grad_norm": 1.1565662622451782, + "learning_rate": 2.7202072538860106e-06, + "loss": 0.2811276912689209, + "step": 210, + "token_acc": 0.9220826230889215 + }, + { + "epoch": 0.013936152973586129, + "grad_norm": 1.1716532707214355, + "learning_rate": 2.784974093264249e-06, + "loss": 0.28411569595336916, + "step": 215, + "token_acc": 0.9210638619159232 + }, + { + "epoch": 0.0142602495543672, + "grad_norm": 1.2090009450912476, + "learning_rate": 2.8497409326424875e-06, + "loss": 0.28354754447937014, + "step": 220, + "token_acc": 0.9223695272586173 + }, + { + "epoch": 0.014584346135148274, + "grad_norm": 1.1573808193206787, + "learning_rate": 2.9145077720207252e-06, + "loss": 0.27943172454833987, + "step": 225, + "token_acc": 0.922367671423052 + }, + { + "epoch": 0.014908442715929346, + "grad_norm": 1.1708178520202637, + "learning_rate": 2.979274611398964e-06, + "loss": 0.26848132610321046, + "step": 230, + "token_acc": 0.9253435183804218 + }, + { + "epoch": 0.01523253929671042, + "grad_norm": 1.145649790763855, + "learning_rate": 3.0440414507772025e-06, + "loss": 0.26862215995788574, + "step": 235, + "token_acc": 0.9247367940405213 + }, + { + "epoch": 0.015556635877491492, + "grad_norm": 1.1238332986831665, + "learning_rate": 3.1088082901554407e-06, + "loss": 0.27870104312896726, + "step": 240, + "token_acc": 0.9225761980010014 + }, + { + "epoch": 0.015880732458272564, + "grad_norm": 1.142770528793335, + "learning_rate": 3.1735751295336793e-06, + "loss": 0.271391487121582, + "step": 245, + "token_acc": 0.9252171658239337 + }, + { + "epoch": 0.01620482903905364, + "grad_norm": 1.0345611572265625, + "learning_rate": 3.238341968911917e-06, + "loss": 0.2781627655029297, + "step": 250, + "token_acc": 0.9220097608841605 + }, + { + "epoch": 0.01652892561983471, + "grad_norm": 1.0694793462753296, + "learning_rate": 3.3031088082901557e-06, + "loss": 0.26626498699188234, + "step": 255, + "token_acc": 0.9258351506262423 + }, + { + "epoch": 0.016853022200615783, + "grad_norm": 1.2326105833053589, + "learning_rate": 3.367875647668394e-06, + "loss": 0.2813969612121582, + "step": 260, + "token_acc": 0.9220959919142596 + }, + { + "epoch": 0.017177118781396855, + "grad_norm": 1.1490870714187622, + "learning_rate": 3.4326424870466326e-06, + "loss": 0.27063512802124023, + "step": 265, + "token_acc": 0.925166912735017 + }, + { + "epoch": 0.01750121536217793, + "grad_norm": 1.1896758079528809, + "learning_rate": 3.497409326424871e-06, + "loss": 0.27055246829986573, + "step": 270, + "token_acc": 0.9244674824013597 + }, + { + "epoch": 0.017825311942959002, + "grad_norm": 1.175060510635376, + "learning_rate": 3.562176165803109e-06, + "loss": 0.27529966831207275, + "step": 275, + "token_acc": 0.9235797078850252 + }, + { + "epoch": 0.018149408523740074, + "grad_norm": 1.1656664609909058, + "learning_rate": 3.6269430051813476e-06, + "loss": 0.2720619201660156, + "step": 280, + "token_acc": 0.924282242288921 + }, + { + "epoch": 0.018473505104521146, + "grad_norm": 1.2035537958145142, + "learning_rate": 3.6917098445595854e-06, + "loss": 0.27737472057342527, + "step": 285, + "token_acc": 0.9232464193903782 + }, + { + "epoch": 0.01879760168530222, + "grad_norm": 1.0576199293136597, + "learning_rate": 3.756476683937824e-06, + "loss": 0.2786442756652832, + "step": 290, + "token_acc": 0.922734228213919 + }, + { + "epoch": 0.019121698266083294, + "grad_norm": 1.101805567741394, + "learning_rate": 3.821243523316063e-06, + "loss": 0.26576249599456786, + "step": 295, + "token_acc": 0.9256528679087294 + }, + { + "epoch": 0.019445794846864366, + "grad_norm": 1.164818525314331, + "learning_rate": 3.886010362694301e-06, + "loss": 0.27654869556427003, + "step": 300, + "token_acc": 0.9231146575969651 + }, + { + "epoch": 0.019769891427645438, + "grad_norm": 1.1586562395095825, + "learning_rate": 3.9507772020725395e-06, + "loss": 0.276000452041626, + "step": 305, + "token_acc": 0.9234045303260413 + }, + { + "epoch": 0.02009398800842651, + "grad_norm": 1.2469451427459717, + "learning_rate": 4.015544041450777e-06, + "loss": 0.28005189895629884, + "step": 310, + "token_acc": 0.9224248584478456 + }, + { + "epoch": 0.020418084589207585, + "grad_norm": 1.088731050491333, + "learning_rate": 4.080310880829016e-06, + "loss": 0.27149338722229005, + "step": 315, + "token_acc": 0.9231371230103549 + }, + { + "epoch": 0.020742181169988657, + "grad_norm": 1.0491572618484497, + "learning_rate": 4.145077720207254e-06, + "loss": 0.26023097038269044, + "step": 320, + "token_acc": 0.9270244697348017 + }, + { + "epoch": 0.02106627775076973, + "grad_norm": 1.1924079656600952, + "learning_rate": 4.209844559585492e-06, + "loss": 0.27563767433166503, + "step": 325, + "token_acc": 0.922792435810702 + }, + { + "epoch": 0.0213903743315508, + "grad_norm": 1.0052635669708252, + "learning_rate": 4.274611398963731e-06, + "loss": 0.2668116569519043, + "step": 330, + "token_acc": 0.9254766824471007 + }, + { + "epoch": 0.021714470912331876, + "grad_norm": 1.1143276691436768, + "learning_rate": 4.339378238341969e-06, + "loss": 0.27575368881225587, + "step": 335, + "token_acc": 0.9234813273435626 + }, + { + "epoch": 0.02203856749311295, + "grad_norm": 1.1198474168777466, + "learning_rate": 4.404145077720207e-06, + "loss": 0.2739673137664795, + "step": 340, + "token_acc": 0.9234454717323182 + }, + { + "epoch": 0.02236266407389402, + "grad_norm": 1.1597661972045898, + "learning_rate": 4.468911917098446e-06, + "loss": 0.2812596559524536, + "step": 345, + "token_acc": 0.9220368755769822 + }, + { + "epoch": 0.022686760654675092, + "grad_norm": 1.1597846746444702, + "learning_rate": 4.533678756476685e-06, + "loss": 0.2782421112060547, + "step": 350, + "token_acc": 0.9231371002202281 + }, + { + "epoch": 0.023010857235456168, + "grad_norm": 1.1377077102661133, + "learning_rate": 4.598445595854923e-06, + "loss": 0.27525331974029543, + "step": 355, + "token_acc": 0.9234653879321608 + }, + { + "epoch": 0.02333495381623724, + "grad_norm": 1.1639304161071777, + "learning_rate": 4.663212435233161e-06, + "loss": 0.27627902030944823, + "step": 360, + "token_acc": 0.9239800719987891 + }, + { + "epoch": 0.02365905039701831, + "grad_norm": 1.1461061239242554, + "learning_rate": 4.7279792746114e-06, + "loss": 0.26558914184570315, + "step": 365, + "token_acc": 0.9252026481090521 + }, + { + "epoch": 0.023983146977799383, + "grad_norm": 1.254964828491211, + "learning_rate": 4.7927461139896375e-06, + "loss": 0.27816622257232665, + "step": 370, + "token_acc": 0.9229658348728891 + }, + { + "epoch": 0.024307243558580455, + "grad_norm": 1.1927502155303955, + "learning_rate": 4.857512953367876e-06, + "loss": 0.26456561088562014, + "step": 375, + "token_acc": 0.9256379821958457 + }, + { + "epoch": 0.02463134013936153, + "grad_norm": 1.0993506908416748, + "learning_rate": 4.922279792746114e-06, + "loss": 0.26846771240234374, + "step": 380, + "token_acc": 0.9255846499947682 + }, + { + "epoch": 0.024955436720142603, + "grad_norm": 1.1774364709854126, + "learning_rate": 4.9870466321243525e-06, + "loss": 0.2800713539123535, + "step": 385, + "token_acc": 0.923156113937995 + }, + { + "epoch": 0.025279533300923675, + "grad_norm": 1.2764763832092285, + "learning_rate": 5.051813471502591e-06, + "loss": 0.270487642288208, + "step": 390, + "token_acc": 0.9245638849599246 + }, + { + "epoch": 0.025603629881704747, + "grad_norm": 1.1404081583023071, + "learning_rate": 5.11658031088083e-06, + "loss": 0.27359600067138673, + "step": 395, + "token_acc": 0.9244739005075082 + }, + { + "epoch": 0.025927726462485822, + "grad_norm": 1.1143617630004883, + "learning_rate": 5.1813471502590676e-06, + "loss": 0.27301363945007323, + "step": 400, + "token_acc": 0.9238943472935289 + }, + { + "epoch": 0.026251823043266894, + "grad_norm": 1.5213898420333862, + "learning_rate": 5.246113989637306e-06, + "loss": 0.27214756011962893, + "step": 405, + "token_acc": 0.9242776311816523 + }, + { + "epoch": 0.026575919624047966, + "grad_norm": 1.1557198762893677, + "learning_rate": 5.310880829015545e-06, + "loss": 0.2753096103668213, + "step": 410, + "token_acc": 0.923890271693801 + }, + { + "epoch": 0.026900016204829038, + "grad_norm": 1.197363018989563, + "learning_rate": 5.3756476683937834e-06, + "loss": 0.27670223712921144, + "step": 415, + "token_acc": 0.9229302265035148 + }, + { + "epoch": 0.027224112785610113, + "grad_norm": 1.1271995306015015, + "learning_rate": 5.440414507772021e-06, + "loss": 0.2729542016983032, + "step": 420, + "token_acc": 0.9231763534323003 + }, + { + "epoch": 0.027548209366391185, + "grad_norm": 1.1053199768066406, + "learning_rate": 5.505181347150259e-06, + "loss": 0.27408714294433595, + "step": 425, + "token_acc": 0.923493480330874 + }, + { + "epoch": 0.027872305947172257, + "grad_norm": 1.1583583354949951, + "learning_rate": 5.569948186528498e-06, + "loss": 0.2778322696685791, + "step": 430, + "token_acc": 0.9223960251465049 + }, + { + "epoch": 0.02819640252795333, + "grad_norm": 1.1535427570343018, + "learning_rate": 5.634715025906736e-06, + "loss": 0.27780320644378664, + "step": 435, + "token_acc": 0.9218625346193012 + }, + { + "epoch": 0.0285204991087344, + "grad_norm": 1.1105202436447144, + "learning_rate": 5.699481865284975e-06, + "loss": 0.27113909721374513, + "step": 440, + "token_acc": 0.9243662988143548 + }, + { + "epoch": 0.028844595689515477, + "grad_norm": 1.1289681196212769, + "learning_rate": 5.7642487046632135e-06, + "loss": 0.2736364841461182, + "step": 445, + "token_acc": 0.9236354858492413 + }, + { + "epoch": 0.02916869227029655, + "grad_norm": 1.119524359703064, + "learning_rate": 5.8290155440414505e-06, + "loss": 0.2711230754852295, + "step": 450, + "token_acc": 0.9240434261931405 + }, + { + "epoch": 0.02949278885107762, + "grad_norm": 1.166691780090332, + "learning_rate": 5.893782383419689e-06, + "loss": 0.27475461959838865, + "step": 455, + "token_acc": 0.923216417669646 + }, + { + "epoch": 0.029816885431858692, + "grad_norm": 1.1108464002609253, + "learning_rate": 5.958549222797928e-06, + "loss": 0.27015066146850586, + "step": 460, + "token_acc": 0.9245998342566673 + }, + { + "epoch": 0.030140982012639768, + "grad_norm": 1.3036308288574219, + "learning_rate": 6.023316062176166e-06, + "loss": 0.27171549797058103, + "step": 465, + "token_acc": 0.9233259233259233 + }, + { + "epoch": 0.03046507859342084, + "grad_norm": 1.4210838079452515, + "learning_rate": 6.088082901554405e-06, + "loss": 0.27597966194152834, + "step": 470, + "token_acc": 0.9232971630278468 + }, + { + "epoch": 0.030789175174201912, + "grad_norm": 1.2602447271347046, + "learning_rate": 6.152849740932643e-06, + "loss": 0.28009696006774903, + "step": 475, + "token_acc": 0.921947459863359 + }, + { + "epoch": 0.031113271754982984, + "grad_norm": 1.1573154926300049, + "learning_rate": 6.217616580310881e-06, + "loss": 0.27212891578674314, + "step": 480, + "token_acc": 0.9239911824538681 + }, + { + "epoch": 0.031437368335764056, + "grad_norm": 1.5468491315841675, + "learning_rate": 6.28238341968912e-06, + "loss": 0.26889379024505616, + "step": 485, + "token_acc": 0.9242921558007661 + }, + { + "epoch": 0.03176146491654513, + "grad_norm": 1.0774251222610474, + "learning_rate": 6.347150259067359e-06, + "loss": 0.27394418716430663, + "step": 490, + "token_acc": 0.9239502943100065 + }, + { + "epoch": 0.03208556149732621, + "grad_norm": 1.0900217294692993, + "learning_rate": 6.411917098445596e-06, + "loss": 0.2677357912063599, + "step": 495, + "token_acc": 0.9250740075064756 + }, + { + "epoch": 0.03240965807810728, + "grad_norm": 1.1952015161514282, + "learning_rate": 6.476683937823834e-06, + "loss": 0.2757922649383545, + "step": 500, + "token_acc": 0.9237246194149515 + }, + { + "epoch": 0.03240965807810728, + "eval_loss": 0.2766605019569397, + "eval_runtime": 118.2624, + "eval_samples_per_second": 169.116, + "eval_steps_per_second": 0.228, + "eval_token_acc": 0.9228264253046955, + "step": 500 + }, + { + "epoch": 0.03273375465888835, + "grad_norm": 1.072628378868103, + "learning_rate": 6.541450777202073e-06, + "loss": 0.27219414710998535, + "step": 505, + "token_acc": 0.923241036144262 + }, + { + "epoch": 0.03305785123966942, + "grad_norm": 1.123671531677246, + "learning_rate": 6.6062176165803115e-06, + "loss": 0.27762365341186523, + "step": 510, + "token_acc": 0.9229576107284878 + }, + { + "epoch": 0.033381947820450494, + "grad_norm": 1.149427890777588, + "learning_rate": 6.67098445595855e-06, + "loss": 0.277001690864563, + "step": 515, + "token_acc": 0.9234441979821897 + }, + { + "epoch": 0.033706044401231566, + "grad_norm": 1.1308335065841675, + "learning_rate": 6.735751295336788e-06, + "loss": 0.27818913459777833, + "step": 520, + "token_acc": 0.9231610048227251 + }, + { + "epoch": 0.03403014098201264, + "grad_norm": 1.2418826818466187, + "learning_rate": 6.8005181347150265e-06, + "loss": 0.27416648864746096, + "step": 525, + "token_acc": 0.9227283971513013 + }, + { + "epoch": 0.03435423756279371, + "grad_norm": 1.12945556640625, + "learning_rate": 6.865284974093265e-06, + "loss": 0.2727360248565674, + "step": 530, + "token_acc": 0.9247170911664296 + }, + { + "epoch": 0.03467833414357478, + "grad_norm": 1.153743028640747, + "learning_rate": 6.930051813471503e-06, + "loss": 0.2731607913970947, + "step": 535, + "token_acc": 0.9244839180524422 + }, + { + "epoch": 0.03500243072435586, + "grad_norm": 1.1583285331726074, + "learning_rate": 6.994818652849742e-06, + "loss": 0.27802448272705077, + "step": 540, + "token_acc": 0.922633087074326 + }, + { + "epoch": 0.03532652730513693, + "grad_norm": 1.2018134593963623, + "learning_rate": 7.059585492227979e-06, + "loss": 0.2711876392364502, + "step": 545, + "token_acc": 0.9240181425935071 + }, + { + "epoch": 0.035650623885918005, + "grad_norm": 1.1137855052947998, + "learning_rate": 7.124352331606218e-06, + "loss": 0.2710531711578369, + "step": 550, + "token_acc": 0.9236695779893513 + }, + { + "epoch": 0.03597472046669908, + "grad_norm": 1.1690529584884644, + "learning_rate": 7.189119170984457e-06, + "loss": 0.27385501861572265, + "step": 555, + "token_acc": 0.9241576534326812 + }, + { + "epoch": 0.03629881704748015, + "grad_norm": 1.1739192008972168, + "learning_rate": 7.253886010362695e-06, + "loss": 0.2713189601898193, + "step": 560, + "token_acc": 0.9240261137814765 + }, + { + "epoch": 0.03662291362826122, + "grad_norm": 1.574720859527588, + "learning_rate": 7.318652849740934e-06, + "loss": 0.27800450325012205, + "step": 565, + "token_acc": 0.9226385003529504 + }, + { + "epoch": 0.03694701020904229, + "grad_norm": 1.0243511199951172, + "learning_rate": 7.383419689119171e-06, + "loss": 0.27382216453552244, + "step": 570, + "token_acc": 0.9235590504565607 + }, + { + "epoch": 0.037271106789823365, + "grad_norm": 1.138895869255066, + "learning_rate": 7.4481865284974095e-06, + "loss": 0.2831470012664795, + "step": 575, + "token_acc": 0.9223049276273151 + }, + { + "epoch": 0.03759520337060444, + "grad_norm": 1.067686676979065, + "learning_rate": 7.512953367875648e-06, + "loss": 0.27533273696899413, + "step": 580, + "token_acc": 0.9240167802831672 + }, + { + "epoch": 0.037919299951385516, + "grad_norm": 1.2463687658309937, + "learning_rate": 7.577720207253887e-06, + "loss": 0.2789109230041504, + "step": 585, + "token_acc": 0.9221528491266894 + }, + { + "epoch": 0.03824339653216659, + "grad_norm": 1.0746787786483765, + "learning_rate": 7.642487046632126e-06, + "loss": 0.27881345748901365, + "step": 590, + "token_acc": 0.9218981710331191 + }, + { + "epoch": 0.03856749311294766, + "grad_norm": 1.2029082775115967, + "learning_rate": 7.707253886010362e-06, + "loss": 0.2803633451461792, + "step": 595, + "token_acc": 0.9225640759572481 + }, + { + "epoch": 0.03889158969372873, + "grad_norm": 1.1455609798431396, + "learning_rate": 7.772020725388602e-06, + "loss": 0.28123066425323484, + "step": 600, + "token_acc": 0.9223912180928449 + }, + { + "epoch": 0.0392156862745098, + "grad_norm": 1.1479970216751099, + "learning_rate": 7.83678756476684e-06, + "loss": 0.2779627799987793, + "step": 605, + "token_acc": 0.9228941389389806 + }, + { + "epoch": 0.039539782855290875, + "grad_norm": 1.1086708307266235, + "learning_rate": 7.901554404145079e-06, + "loss": 0.27979559898376466, + "step": 610, + "token_acc": 0.9224347554711779 + }, + { + "epoch": 0.03986387943607195, + "grad_norm": 1.1686094999313354, + "learning_rate": 7.966321243523317e-06, + "loss": 0.2715947151184082, + "step": 615, + "token_acc": 0.9236265697504624 + }, + { + "epoch": 0.04018797601685302, + "grad_norm": 1.0965114831924438, + "learning_rate": 8.031088082901555e-06, + "loss": 0.2718233823776245, + "step": 620, + "token_acc": 0.9235554344723801 + }, + { + "epoch": 0.0405120725976341, + "grad_norm": 1.1116286516189575, + "learning_rate": 8.095854922279794e-06, + "loss": 0.26779100894927976, + "step": 625, + "token_acc": 0.9256486247595279 + }, + { + "epoch": 0.04083616917841517, + "grad_norm": 1.1061184406280518, + "learning_rate": 8.160621761658032e-06, + "loss": 0.27161588668823244, + "step": 630, + "token_acc": 0.9231390227205208 + }, + { + "epoch": 0.04116026575919624, + "grad_norm": 1.137664556503296, + "learning_rate": 8.22538860103627e-06, + "loss": 0.27380824089050293, + "step": 635, + "token_acc": 0.9229845368563279 + }, + { + "epoch": 0.041484362339977314, + "grad_norm": 1.1426571607589722, + "learning_rate": 8.290155440414507e-06, + "loss": 0.2724440574645996, + "step": 640, + "token_acc": 0.9243293077632259 + }, + { + "epoch": 0.041808458920758386, + "grad_norm": 1.2707536220550537, + "learning_rate": 8.354922279792747e-06, + "loss": 0.27617788314819336, + "step": 645, + "token_acc": 0.9231982418718894 + }, + { + "epoch": 0.04213255550153946, + "grad_norm": 1.0595289468765259, + "learning_rate": 8.419689119170985e-06, + "loss": 0.27484869956970215, + "step": 650, + "token_acc": 0.9236766623207301 + }, + { + "epoch": 0.04245665208232053, + "grad_norm": 1.2787915468215942, + "learning_rate": 8.484455958549224e-06, + "loss": 0.2772678375244141, + "step": 655, + "token_acc": 0.9225656959991502 + }, + { + "epoch": 0.0427807486631016, + "grad_norm": 1.0741102695465088, + "learning_rate": 8.549222797927462e-06, + "loss": 0.2790708541870117, + "step": 660, + "token_acc": 0.9225615230249261 + }, + { + "epoch": 0.043104845243882674, + "grad_norm": 1.0884015560150146, + "learning_rate": 8.6139896373057e-06, + "loss": 0.2731776237487793, + "step": 665, + "token_acc": 0.923714427712982 + }, + { + "epoch": 0.04342894182466375, + "grad_norm": 1.1796871423721313, + "learning_rate": 8.678756476683938e-06, + "loss": 0.26450159549713137, + "step": 670, + "token_acc": 0.9260345496511498 + }, + { + "epoch": 0.043753038405444825, + "grad_norm": 1.1302798986434937, + "learning_rate": 8.743523316062177e-06, + "loss": 0.286497974395752, + "step": 675, + "token_acc": 0.9212607806408483 + }, + { + "epoch": 0.0440771349862259, + "grad_norm": 1.0766465663909912, + "learning_rate": 8.808290155440415e-06, + "loss": 0.2754509925842285, + "step": 680, + "token_acc": 0.9223906644786538 + }, + { + "epoch": 0.04440123156700697, + "grad_norm": 1.1511085033416748, + "learning_rate": 8.873056994818654e-06, + "loss": 0.2762892723083496, + "step": 685, + "token_acc": 0.9242773204915137 + }, + { + "epoch": 0.04472532814778804, + "grad_norm": 1.2351347208023071, + "learning_rate": 8.937823834196892e-06, + "loss": 0.2697291851043701, + "step": 690, + "token_acc": 0.9241456435220209 + }, + { + "epoch": 0.04504942472856911, + "grad_norm": 1.1533805131912231, + "learning_rate": 9.00259067357513e-06, + "loss": 0.27661752700805664, + "step": 695, + "token_acc": 0.9234376426698668 + }, + { + "epoch": 0.045373521309350184, + "grad_norm": 1.4566006660461426, + "learning_rate": 9.06735751295337e-06, + "loss": 0.2665151596069336, + "step": 700, + "token_acc": 0.9256478929225396 + }, + { + "epoch": 0.045697617890131256, + "grad_norm": 1.2104111909866333, + "learning_rate": 9.132124352331607e-06, + "loss": 0.2740550756454468, + "step": 705, + "token_acc": 0.9229390563779198 + }, + { + "epoch": 0.046021714470912335, + "grad_norm": 1.082419753074646, + "learning_rate": 9.196891191709847e-06, + "loss": 0.2736359119415283, + "step": 710, + "token_acc": 0.9242869889950097 + }, + { + "epoch": 0.04634581105169341, + "grad_norm": 1.2044905424118042, + "learning_rate": 9.261658031088083e-06, + "loss": 0.27272777557373046, + "step": 715, + "token_acc": 0.9237638165499864 + }, + { + "epoch": 0.04666990763247448, + "grad_norm": 1.1270380020141602, + "learning_rate": 9.326424870466322e-06, + "loss": 0.27639906406402587, + "step": 720, + "token_acc": 0.9229221824222603 + }, + { + "epoch": 0.04699400421325555, + "grad_norm": 1.140891432762146, + "learning_rate": 9.39119170984456e-06, + "loss": 0.2769869327545166, + "step": 725, + "token_acc": 0.9231643044896773 + }, + { + "epoch": 0.04731810079403662, + "grad_norm": 1.2938241958618164, + "learning_rate": 9.4559585492228e-06, + "loss": 0.2770379543304443, + "step": 730, + "token_acc": 0.9227479442637688 + }, + { + "epoch": 0.047642197374817695, + "grad_norm": 1.1942880153656006, + "learning_rate": 9.520725388601037e-06, + "loss": 0.27842211723327637, + "step": 735, + "token_acc": 0.9219535053237186 + }, + { + "epoch": 0.04796629395559877, + "grad_norm": 1.5139548778533936, + "learning_rate": 9.585492227979275e-06, + "loss": 0.279721736907959, + "step": 740, + "token_acc": 0.9225660590645753 + }, + { + "epoch": 0.04829039053637984, + "grad_norm": 1.0569076538085938, + "learning_rate": 9.650259067357514e-06, + "loss": 0.2757410049438477, + "step": 745, + "token_acc": 0.9230635290172382 + }, + { + "epoch": 0.04861448711716091, + "grad_norm": 1.347761631011963, + "learning_rate": 9.715025906735752e-06, + "loss": 0.27115836143493655, + "step": 750, + "token_acc": 0.9239254170755643 + }, + { + "epoch": 0.04893858369794199, + "grad_norm": 1.1022578477859497, + "learning_rate": 9.77979274611399e-06, + "loss": 0.28516759872436526, + "step": 755, + "token_acc": 0.9212077933989334 + }, + { + "epoch": 0.04926268027872306, + "grad_norm": 1.153346300125122, + "learning_rate": 9.844559585492228e-06, + "loss": 0.2748947858810425, + "step": 760, + "token_acc": 0.9237991482618599 + }, + { + "epoch": 0.049586776859504134, + "grad_norm": 1.146430253982544, + "learning_rate": 9.909326424870467e-06, + "loss": 0.2799590826034546, + "step": 765, + "token_acc": 0.9221607156911614 + }, + { + "epoch": 0.049910873440285206, + "grad_norm": 1.074004054069519, + "learning_rate": 9.974093264248705e-06, + "loss": 0.2774993419647217, + "step": 770, + "token_acc": 0.9225398806505714 + }, + { + "epoch": 0.05023497002106628, + "grad_norm": 1.0626007318496704, + "learning_rate": 9.999998966164752e-06, + "loss": 0.27788662910461426, + "step": 775, + "token_acc": 0.9223414363765936 + }, + { + "epoch": 0.05055906660184735, + "grad_norm": 1.2621933221817017, + "learning_rate": 9.999992648284216e-06, + "loss": 0.2749335765838623, + "step": 780, + "token_acc": 0.9238335700383988 + }, + { + "epoch": 0.05088316318262842, + "grad_norm": 1.1462196111679077, + "learning_rate": 9.999980586883313e-06, + "loss": 0.2756141185760498, + "step": 785, + "token_acc": 0.9220426114530389 + }, + { + "epoch": 0.05120725976340949, + "grad_norm": 1.1427782773971558, + "learning_rate": 9.999962781975896e-06, + "loss": 0.27129230499267576, + "step": 790, + "token_acc": 0.9244358478401031 + }, + { + "epoch": 0.051531356344190565, + "grad_norm": 1.059018611907959, + "learning_rate": 9.999939233582416e-06, + "loss": 0.2826401233673096, + "step": 795, + "token_acc": 0.9208777683943149 + }, + { + "epoch": 0.051855452924971644, + "grad_norm": 1.0648328065872192, + "learning_rate": 9.999909941729927e-06, + "loss": 0.27859494686126707, + "step": 800, + "token_acc": 0.9232760032760032 + }, + { + "epoch": 0.052179549505752716, + "grad_norm": 1.2377976179122925, + "learning_rate": 9.999874906452072e-06, + "loss": 0.27971878051757815, + "step": 805, + "token_acc": 0.9224227431779438 + }, + { + "epoch": 0.05250364608653379, + "grad_norm": 1.168636679649353, + "learning_rate": 9.9998341277891e-06, + "loss": 0.27319889068603515, + "step": 810, + "token_acc": 0.9236747392983797 + }, + { + "epoch": 0.05282774266731486, + "grad_norm": 1.1205686330795288, + "learning_rate": 9.99978760578785e-06, + "loss": 0.27941184043884276, + "step": 815, + "token_acc": 0.9213014802493188 + }, + { + "epoch": 0.05315183924809593, + "grad_norm": 1.1140272617340088, + "learning_rate": 9.999735340501765e-06, + "loss": 0.27391619682312013, + "step": 820, + "token_acc": 0.9239368261924653 + }, + { + "epoch": 0.053475935828877004, + "grad_norm": 1.169887661933899, + "learning_rate": 9.999677331990885e-06, + "loss": 0.2753794193267822, + "step": 825, + "token_acc": 0.9224798000502519 + }, + { + "epoch": 0.053800032409658076, + "grad_norm": 1.1942942142486572, + "learning_rate": 9.999613580321839e-06, + "loss": 0.28004434108734133, + "step": 830, + "token_acc": 0.9226770844313271 + }, + { + "epoch": 0.05412412899043915, + "grad_norm": 1.188458800315857, + "learning_rate": 9.99954408556786e-06, + "loss": 0.28326759338378904, + "step": 835, + "token_acc": 0.9210668085962601 + }, + { + "epoch": 0.05444822557122023, + "grad_norm": 1.0895791053771973, + "learning_rate": 9.99946884780878e-06, + "loss": 0.28024749755859374, + "step": 840, + "token_acc": 0.9220975777692613 + }, + { + "epoch": 0.0547723221520013, + "grad_norm": 1.0977063179016113, + "learning_rate": 9.999387867131024e-06, + "loss": 0.2782959699630737, + "step": 845, + "token_acc": 0.9221023706825373 + }, + { + "epoch": 0.05509641873278237, + "grad_norm": 1.0744558572769165, + "learning_rate": 9.999301143627613e-06, + "loss": 0.28442230224609377, + "step": 850, + "token_acc": 0.9221025514893396 + }, + { + "epoch": 0.05542051531356344, + "grad_norm": 1.0994865894317627, + "learning_rate": 9.99920867739817e-06, + "loss": 0.28086304664611816, + "step": 855, + "token_acc": 0.9227749373956594 + }, + { + "epoch": 0.055744611894344515, + "grad_norm": 1.2142789363861084, + "learning_rate": 9.999110468548909e-06, + "loss": 0.2726480484008789, + "step": 860, + "token_acc": 0.9236710319811929 + }, + { + "epoch": 0.05606870847512559, + "grad_norm": 1.2806187868118286, + "learning_rate": 9.999006517192644e-06, + "loss": 0.2741135597229004, + "step": 865, + "token_acc": 0.9237250292228687 + }, + { + "epoch": 0.05639280505590666, + "grad_norm": 1.2578004598617554, + "learning_rate": 9.998896823448784e-06, + "loss": 0.2765141487121582, + "step": 870, + "token_acc": 0.9229077881176722 + }, + { + "epoch": 0.05671690163668773, + "grad_norm": 1.0189783573150635, + "learning_rate": 9.998781387443336e-06, + "loss": 0.28197636604309084, + "step": 875, + "token_acc": 0.921837868104251 + }, + { + "epoch": 0.0570409982174688, + "grad_norm": 1.1214468479156494, + "learning_rate": 9.998660209308899e-06, + "loss": 0.28133354187011717, + "step": 880, + "token_acc": 0.9221426551721871 + }, + { + "epoch": 0.05736509479824988, + "grad_norm": 1.0209897756576538, + "learning_rate": 9.998533289184674e-06, + "loss": 0.28200483322143555, + "step": 885, + "token_acc": 0.9226726578769305 + }, + { + "epoch": 0.05768919137903095, + "grad_norm": 1.057998776435852, + "learning_rate": 9.998400627216456e-06, + "loss": 0.2701537609100342, + "step": 890, + "token_acc": 0.9244533587610562 + }, + { + "epoch": 0.058013287959812025, + "grad_norm": 1.0923854112625122, + "learning_rate": 9.99826222355663e-06, + "loss": 0.2727808952331543, + "step": 895, + "token_acc": 0.9238448899112718 + }, + { + "epoch": 0.0583373845405931, + "grad_norm": 1.0997750759124756, + "learning_rate": 9.998118078364186e-06, + "loss": 0.2779165983200073, + "step": 900, + "token_acc": 0.9216114826292612 + }, + { + "epoch": 0.05866148112137417, + "grad_norm": 1.1464025974273682, + "learning_rate": 9.9979681918047e-06, + "loss": 0.2803272008895874, + "step": 905, + "token_acc": 0.9226379612217106 + }, + { + "epoch": 0.05898557770215524, + "grad_norm": 1.063568115234375, + "learning_rate": 9.997812564050352e-06, + "loss": 0.27731595039367674, + "step": 910, + "token_acc": 0.9222437460983506 + }, + { + "epoch": 0.05930967428293631, + "grad_norm": 1.1723806858062744, + "learning_rate": 9.997651195279907e-06, + "loss": 0.27817227840423586, + "step": 915, + "token_acc": 0.9223472731783108 + }, + { + "epoch": 0.059633770863717385, + "grad_norm": 1.0762029886245728, + "learning_rate": 9.997484085678734e-06, + "loss": 0.2755002498626709, + "step": 920, + "token_acc": 0.9227790491705778 + }, + { + "epoch": 0.059957867444498464, + "grad_norm": 1.1249760389328003, + "learning_rate": 9.997311235438795e-06, + "loss": 0.26909599304199217, + "step": 925, + "token_acc": 0.9242543222479616 + }, + { + "epoch": 0.060281964025279536, + "grad_norm": 1.0824518203735352, + "learning_rate": 9.997132644758639e-06, + "loss": 0.2789783477783203, + "step": 930, + "token_acc": 0.922192208806587 + }, + { + "epoch": 0.06060606060606061, + "grad_norm": 1.0471711158752441, + "learning_rate": 9.996948313843417e-06, + "loss": 0.2783126592636108, + "step": 935, + "token_acc": 0.9226764530169839 + }, + { + "epoch": 0.06093015718684168, + "grad_norm": 1.1419248580932617, + "learning_rate": 9.99675824290487e-06, + "loss": 0.27683291435241697, + "step": 940, + "token_acc": 0.9229383806136581 + }, + { + "epoch": 0.06125425376762275, + "grad_norm": 1.1103299856185913, + "learning_rate": 9.996562432161334e-06, + "loss": 0.27548627853393554, + "step": 945, + "token_acc": 0.9228202389560763 + }, + { + "epoch": 0.061578350348403824, + "grad_norm": 1.1687252521514893, + "learning_rate": 9.99636088183774e-06, + "loss": 0.26608130931854246, + "step": 950, + "token_acc": 0.9261335933690883 + }, + { + "epoch": 0.061902446929184896, + "grad_norm": 1.0601294040679932, + "learning_rate": 9.996153592165605e-06, + "loss": 0.2721517086029053, + "step": 955, + "token_acc": 0.9241069662565766 + }, + { + "epoch": 0.06222654350996597, + "grad_norm": 1.1069499254226685, + "learning_rate": 9.995940563383048e-06, + "loss": 0.27439448833465574, + "step": 960, + "token_acc": 0.924347078142105 + }, + { + "epoch": 0.06255064009074704, + "grad_norm": 2.6561391353607178, + "learning_rate": 9.995721795734774e-06, + "loss": 0.27994685173034667, + "step": 965, + "token_acc": 0.9225540625656099 + }, + { + "epoch": 0.06287473667152811, + "grad_norm": 1.096655249595642, + "learning_rate": 9.995497289472085e-06, + "loss": 0.2748950242996216, + "step": 970, + "token_acc": 0.9231229887402357 + }, + { + "epoch": 0.06319883325230918, + "grad_norm": 1.1758030652999878, + "learning_rate": 9.995267044852872e-06, + "loss": 0.28017172813415525, + "step": 975, + "token_acc": 0.9224684369782972 + }, + { + "epoch": 0.06352292983309026, + "grad_norm": 1.0659986734390259, + "learning_rate": 9.995031062141617e-06, + "loss": 0.27984328269958497, + "step": 980, + "token_acc": 0.9212892498561105 + }, + { + "epoch": 0.06384702641387133, + "grad_norm": 1.2035590410232544, + "learning_rate": 9.994789341609395e-06, + "loss": 0.286294150352478, + "step": 985, + "token_acc": 0.9208251654409346 + }, + { + "epoch": 0.06417112299465241, + "grad_norm": 1.0138523578643799, + "learning_rate": 9.994541883533873e-06, + "loss": 0.2800797462463379, + "step": 990, + "token_acc": 0.9230723692000052 + }, + { + "epoch": 0.06449521957543349, + "grad_norm": 1.0648677349090576, + "learning_rate": 9.994288688199305e-06, + "loss": 0.2819873571395874, + "step": 995, + "token_acc": 0.9220508407255396 + }, + { + "epoch": 0.06481931615621456, + "grad_norm": 1.3292344808578491, + "learning_rate": 9.994029755896541e-06, + "loss": 0.2753146171569824, + "step": 1000, + "token_acc": 0.923467268857395 + }, + { + "epoch": 0.06481931615621456, + "eval_loss": 0.2790466248989105, + "eval_runtime": 132.7952, + "eval_samples_per_second": 150.608, + "eval_steps_per_second": 0.203, + "eval_token_acc": 0.9227258578167278, + "step": 1000 + }, + { + "epoch": 0.06514341273699563, + "grad_norm": 1.0402147769927979, + "learning_rate": 9.993765086923016e-06, + "loss": 0.27179250717163084, + "step": 1005, + "token_acc": 0.9241823139449233 + }, + { + "epoch": 0.0654675093177767, + "grad_norm": 1.1630990505218506, + "learning_rate": 9.993494681582758e-06, + "loss": 0.2761619806289673, + "step": 1010, + "token_acc": 0.9224574199600957 + }, + { + "epoch": 0.06579160589855777, + "grad_norm": 1.0634405612945557, + "learning_rate": 9.993218540186382e-06, + "loss": 0.280862021446228, + "step": 1015, + "token_acc": 0.9225046415286321 + }, + { + "epoch": 0.06611570247933884, + "grad_norm": 1.0645707845687866, + "learning_rate": 9.992936663051097e-06, + "loss": 0.27747955322265627, + "step": 1020, + "token_acc": 0.9223336072308956 + }, + { + "epoch": 0.06643979906011992, + "grad_norm": 1.0335700511932373, + "learning_rate": 9.99264905050069e-06, + "loss": 0.2748509645462036, + "step": 1025, + "token_acc": 0.9228667824036162 + }, + { + "epoch": 0.06676389564090099, + "grad_norm": 0.947760283946991, + "learning_rate": 9.992355702865549e-06, + "loss": 0.2750194549560547, + "step": 1030, + "token_acc": 0.9228971508846435 + }, + { + "epoch": 0.06708799222168206, + "grad_norm": 1.025646686553955, + "learning_rate": 9.99205662048264e-06, + "loss": 0.2718366622924805, + "step": 1035, + "token_acc": 0.9243326815515438 + }, + { + "epoch": 0.06741208880246313, + "grad_norm": 1.1592144966125488, + "learning_rate": 9.991751803695524e-06, + "loss": 0.271783709526062, + "step": 1040, + "token_acc": 0.9241152626929273 + }, + { + "epoch": 0.0677361853832442, + "grad_norm": 1.1432433128356934, + "learning_rate": 9.991441252854345e-06, + "loss": 0.27407772541046144, + "step": 1045, + "token_acc": 0.9226717677270583 + }, + { + "epoch": 0.06806028196402528, + "grad_norm": 0.9991722106933594, + "learning_rate": 9.991124968315833e-06, + "loss": 0.27140059471130373, + "step": 1050, + "token_acc": 0.9241662289915966 + }, + { + "epoch": 0.06838437854480635, + "grad_norm": 1.0588241815567017, + "learning_rate": 9.990802950443308e-06, + "loss": 0.27434496879577636, + "step": 1055, + "token_acc": 0.9236371965185525 + }, + { + "epoch": 0.06870847512558742, + "grad_norm": 1.166620135307312, + "learning_rate": 9.990475199606672e-06, + "loss": 0.2784614086151123, + "step": 1060, + "token_acc": 0.9222009165068737 + }, + { + "epoch": 0.06903257170636849, + "grad_norm": 1.0814213752746582, + "learning_rate": 9.990141716182415e-06, + "loss": 0.27708754539489744, + "step": 1065, + "token_acc": 0.9233220276106945 + }, + { + "epoch": 0.06935666828714956, + "grad_norm": 1.1501027345657349, + "learning_rate": 9.989802500553611e-06, + "loss": 0.27069602012634275, + "step": 1070, + "token_acc": 0.924789820495342 + }, + { + "epoch": 0.06968076486793065, + "grad_norm": 1.2108640670776367, + "learning_rate": 9.98945755310992e-06, + "loss": 0.2705090999603271, + "step": 1075, + "token_acc": 0.9240915032679738 + }, + { + "epoch": 0.07000486144871172, + "grad_norm": 1.1110398769378662, + "learning_rate": 9.989106874247584e-06, + "loss": 0.2795064926147461, + "step": 1080, + "token_acc": 0.9227160558993917 + }, + { + "epoch": 0.0703289580294928, + "grad_norm": 1.1872522830963135, + "learning_rate": 9.98875046436943e-06, + "loss": 0.2769919395446777, + "step": 1085, + "token_acc": 0.9226704873289292 + }, + { + "epoch": 0.07065305461027387, + "grad_norm": 1.0782725811004639, + "learning_rate": 9.98838832388487e-06, + "loss": 0.2816041946411133, + "step": 1090, + "token_acc": 0.9211050111880106 + }, + { + "epoch": 0.07097715119105494, + "grad_norm": 1.0335499048233032, + "learning_rate": 9.988020453209893e-06, + "loss": 0.2720641136169434, + "step": 1095, + "token_acc": 0.9238217909823203 + }, + { + "epoch": 0.07130124777183601, + "grad_norm": 1.0420279502868652, + "learning_rate": 9.987646852767077e-06, + "loss": 0.2817659378051758, + "step": 1100, + "token_acc": 0.9224168986409039 + }, + { + "epoch": 0.07162534435261708, + "grad_norm": 1.151588797569275, + "learning_rate": 9.987267522985579e-06, + "loss": 0.2755244255065918, + "step": 1105, + "token_acc": 0.9239411698186003 + }, + { + "epoch": 0.07194944093339815, + "grad_norm": 1.122923493385315, + "learning_rate": 9.986882464301135e-06, + "loss": 0.2755061149597168, + "step": 1110, + "token_acc": 0.9228540689083037 + }, + { + "epoch": 0.07227353751417923, + "grad_norm": 1.0247633457183838, + "learning_rate": 9.986491677156067e-06, + "loss": 0.2809417724609375, + "step": 1115, + "token_acc": 0.9233920034045043 + }, + { + "epoch": 0.0725976340949603, + "grad_norm": 1.0457372665405273, + "learning_rate": 9.986095161999273e-06, + "loss": 0.27999517917633054, + "step": 1120, + "token_acc": 0.922785589788617 + }, + { + "epoch": 0.07292173067574137, + "grad_norm": 2.0044500827789307, + "learning_rate": 9.985692919286233e-06, + "loss": 0.27866621017456056, + "step": 1125, + "token_acc": 0.9225357126675598 + }, + { + "epoch": 0.07324582725652244, + "grad_norm": 1.012980341911316, + "learning_rate": 9.985284949479004e-06, + "loss": 0.27526078224182127, + "step": 1130, + "token_acc": 0.9233288968716484 + }, + { + "epoch": 0.07356992383730351, + "grad_norm": 0.9666235446929932, + "learning_rate": 9.984871253046225e-06, + "loss": 0.2693899631500244, + "step": 1135, + "token_acc": 0.9245722108145106 + }, + { + "epoch": 0.07389402041808459, + "grad_norm": 1.1077278852462769, + "learning_rate": 9.98445183046311e-06, + "loss": 0.27521681785583496, + "step": 1140, + "token_acc": 0.9232492852911633 + }, + { + "epoch": 0.07421811699886566, + "grad_norm": 1.0638463497161865, + "learning_rate": 9.984026682211453e-06, + "loss": 0.27799296379089355, + "step": 1145, + "token_acc": 0.9228619303392962 + }, + { + "epoch": 0.07454221357964673, + "grad_norm": 1.1785340309143066, + "learning_rate": 9.983595808779625e-06, + "loss": 0.28396103382110593, + "step": 1150, + "token_acc": 0.9209412221646815 + }, + { + "epoch": 0.0748663101604278, + "grad_norm": 1.3344743251800537, + "learning_rate": 9.98315921066257e-06, + "loss": 0.2823387861251831, + "step": 1155, + "token_acc": 0.9223486856052576 + }, + { + "epoch": 0.07519040674120887, + "grad_norm": 1.074034333229065, + "learning_rate": 9.982716888361815e-06, + "loss": 0.27908711433410643, + "step": 1160, + "token_acc": 0.9232541894882524 + }, + { + "epoch": 0.07551450332198996, + "grad_norm": 1.1171108484268188, + "learning_rate": 9.982268842385454e-06, + "loss": 0.2776108026504517, + "step": 1165, + "token_acc": 0.9222206383557738 + }, + { + "epoch": 0.07583859990277103, + "grad_norm": 1.0370464324951172, + "learning_rate": 9.981815073248163e-06, + "loss": 0.2771688222885132, + "step": 1170, + "token_acc": 0.9227592796188682 + }, + { + "epoch": 0.0761626964835521, + "grad_norm": 1.0821559429168701, + "learning_rate": 9.981355581471185e-06, + "loss": 0.28194661140441896, + "step": 1175, + "token_acc": 0.9206759686332028 + }, + { + "epoch": 0.07648679306433318, + "grad_norm": 1.0708980560302734, + "learning_rate": 9.980890367582347e-06, + "loss": 0.2811826467514038, + "step": 1180, + "token_acc": 0.922474976058352 + }, + { + "epoch": 0.07681088964511425, + "grad_norm": 1.101333498954773, + "learning_rate": 9.98041943211604e-06, + "loss": 0.2675865650177002, + "step": 1185, + "token_acc": 0.9243971972023975 + }, + { + "epoch": 0.07713498622589532, + "grad_norm": 1.3167011737823486, + "learning_rate": 9.979942775613229e-06, + "loss": 0.27997446060180664, + "step": 1190, + "token_acc": 0.9227666273120818 + }, + { + "epoch": 0.07745908280667639, + "grad_norm": 1.044586420059204, + "learning_rate": 9.979460398621452e-06, + "loss": 0.2761145353317261, + "step": 1195, + "token_acc": 0.9237089894045724 + }, + { + "epoch": 0.07778317938745746, + "grad_norm": 1.0204477310180664, + "learning_rate": 9.978972301694822e-06, + "loss": 0.2762503385543823, + "step": 1200, + "token_acc": 0.9229936713455155 + }, + { + "epoch": 0.07810727596823853, + "grad_norm": 1.0193427801132202, + "learning_rate": 9.978478485394014e-06, + "loss": 0.27045645713806155, + "step": 1205, + "token_acc": 0.923695456225561 + }, + { + "epoch": 0.0784313725490196, + "grad_norm": 1.0065299272537231, + "learning_rate": 9.97797895028628e-06, + "loss": 0.2738425493240356, + "step": 1210, + "token_acc": 0.9239265326383718 + }, + { + "epoch": 0.07875546912980068, + "grad_norm": 1.076905369758606, + "learning_rate": 9.97747369694544e-06, + "loss": 0.2760471343994141, + "step": 1215, + "token_acc": 0.9226923325366578 + }, + { + "epoch": 0.07907956571058175, + "grad_norm": 0.9906021356582642, + "learning_rate": 9.976962725951878e-06, + "loss": 0.2779691696166992, + "step": 1220, + "token_acc": 0.9230997778122255 + }, + { + "epoch": 0.07940366229136282, + "grad_norm": 1.0541574954986572, + "learning_rate": 9.976446037892553e-06, + "loss": 0.27247233390808107, + "step": 1225, + "token_acc": 0.9237839226770502 + }, + { + "epoch": 0.0797277588721439, + "grad_norm": 1.0109562873840332, + "learning_rate": 9.975923633360985e-06, + "loss": 0.27540037631988523, + "step": 1230, + "token_acc": 0.922697659191357 + }, + { + "epoch": 0.08005185545292497, + "grad_norm": 1.8598495721817017, + "learning_rate": 9.975395512957265e-06, + "loss": 0.27339174747467043, + "step": 1235, + "token_acc": 0.9236737966531309 + }, + { + "epoch": 0.08037595203370604, + "grad_norm": 1.072318434715271, + "learning_rate": 9.974861677288046e-06, + "loss": 0.2772857189178467, + "step": 1240, + "token_acc": 0.923789479240204 + }, + { + "epoch": 0.08070004861448711, + "grad_norm": 1.184702754020691, + "learning_rate": 9.97432212696655e-06, + "loss": 0.2722355365753174, + "step": 1245, + "token_acc": 0.923662387084624 + }, + { + "epoch": 0.0810241451952682, + "grad_norm": 1.0730267763137817, + "learning_rate": 9.973776862612558e-06, + "loss": 0.2687468767166138, + "step": 1250, + "token_acc": 0.9247340599099393 + }, + { + "epoch": 0.08134824177604927, + "grad_norm": 1.1542036533355713, + "learning_rate": 9.973225884852424e-06, + "loss": 0.27355084419250486, + "step": 1255, + "token_acc": 0.9236140004160599 + }, + { + "epoch": 0.08167233835683034, + "grad_norm": 1.2776581048965454, + "learning_rate": 9.972669194319055e-06, + "loss": 0.2811305522918701, + "step": 1260, + "token_acc": 0.9221158515896829 + }, + { + "epoch": 0.08199643493761141, + "grad_norm": 0.9761309027671814, + "learning_rate": 9.972106791651926e-06, + "loss": 0.2706268787384033, + "step": 1265, + "token_acc": 0.9242809563151916 + }, + { + "epoch": 0.08232053151839248, + "grad_norm": 1.5077848434448242, + "learning_rate": 9.971538677497072e-06, + "loss": 0.2713223218917847, + "step": 1270, + "token_acc": 0.924114835015352 + }, + { + "epoch": 0.08264462809917356, + "grad_norm": 0.9981542229652405, + "learning_rate": 9.970964852507089e-06, + "loss": 0.27992656230926516, + "step": 1275, + "token_acc": 0.9224513598728564 + }, + { + "epoch": 0.08296872467995463, + "grad_norm": 1.1641405820846558, + "learning_rate": 9.970385317341133e-06, + "loss": 0.27625312805175783, + "step": 1280, + "token_acc": 0.9222651154824174 + }, + { + "epoch": 0.0832928212607357, + "grad_norm": 0.974733829498291, + "learning_rate": 9.96980007266492e-06, + "loss": 0.2753995180130005, + "step": 1285, + "token_acc": 0.9226927125214924 + }, + { + "epoch": 0.08361691784151677, + "grad_norm": 1.1057713031768799, + "learning_rate": 9.969209119150723e-06, + "loss": 0.2756865739822388, + "step": 1290, + "token_acc": 0.9226307237144503 + }, + { + "epoch": 0.08394101442229784, + "grad_norm": 1.0507092475891113, + "learning_rate": 9.968612457477373e-06, + "loss": 0.2823331356048584, + "step": 1295, + "token_acc": 0.9209386804248543 + }, + { + "epoch": 0.08426511100307892, + "grad_norm": 0.9887701869010925, + "learning_rate": 9.968010088330262e-06, + "loss": 0.2696245193481445, + "step": 1300, + "token_acc": 0.9235237741310853 + }, + { + "epoch": 0.08458920758385999, + "grad_norm": 1.0384249687194824, + "learning_rate": 9.967402012401332e-06, + "loss": 0.2720928430557251, + "step": 1305, + "token_acc": 0.9242610837438424 + }, + { + "epoch": 0.08491330416464106, + "grad_norm": 0.9624770283699036, + "learning_rate": 9.966788230389084e-06, + "loss": 0.2723661422729492, + "step": 1310, + "token_acc": 0.9240750231933514 + }, + { + "epoch": 0.08523740074542213, + "grad_norm": 1.3515757322311401, + "learning_rate": 9.966168742998574e-06, + "loss": 0.27405097484588625, + "step": 1315, + "token_acc": 0.9239717062036148 + }, + { + "epoch": 0.0855614973262032, + "grad_norm": 1.0048466920852661, + "learning_rate": 9.965543550941407e-06, + "loss": 0.2661163568496704, + "step": 1320, + "token_acc": 0.9261450799912339 + }, + { + "epoch": 0.08588559390698428, + "grad_norm": 1.1117011308670044, + "learning_rate": 9.96491265493575e-06, + "loss": 0.27957565784454347, + "step": 1325, + "token_acc": 0.9222313947517242 + }, + { + "epoch": 0.08620969048776535, + "grad_norm": 2.154306650161743, + "learning_rate": 9.964276055706314e-06, + "loss": 0.27651729583740237, + "step": 1330, + "token_acc": 0.9226900369488895 + }, + { + "epoch": 0.08653378706854643, + "grad_norm": 1.0748560428619385, + "learning_rate": 9.963633753984365e-06, + "loss": 0.27028441429138184, + "step": 1335, + "token_acc": 0.9244444151703969 + }, + { + "epoch": 0.0868578836493275, + "grad_norm": 1.0136338472366333, + "learning_rate": 9.962985750507717e-06, + "loss": 0.27360053062438966, + "step": 1340, + "token_acc": 0.9237293069652803 + }, + { + "epoch": 0.08718198023010858, + "grad_norm": 1.0045623779296875, + "learning_rate": 9.962332046020737e-06, + "loss": 0.2793967962265015, + "step": 1345, + "token_acc": 0.9225186087785754 + }, + { + "epoch": 0.08750607681088965, + "grad_norm": 1.2772600650787354, + "learning_rate": 9.96167264127434e-06, + "loss": 0.27541522979736327, + "step": 1350, + "token_acc": 0.9226138789687074 + }, + { + "epoch": 0.08783017339167072, + "grad_norm": 1.0508257150650024, + "learning_rate": 9.961007537025986e-06, + "loss": 0.279202938079834, + "step": 1355, + "token_acc": 0.92288791300711 + }, + { + "epoch": 0.0881542699724518, + "grad_norm": 1.0508962869644165, + "learning_rate": 9.960336734039684e-06, + "loss": 0.2759141683578491, + "step": 1360, + "token_acc": 0.923016862410603 + }, + { + "epoch": 0.08847836655323287, + "grad_norm": 1.060666799545288, + "learning_rate": 9.95966023308599e-06, + "loss": 0.2707224369049072, + "step": 1365, + "token_acc": 0.9250858742628713 + }, + { + "epoch": 0.08880246313401394, + "grad_norm": 1.0132877826690674, + "learning_rate": 9.958978034942005e-06, + "loss": 0.2735624313354492, + "step": 1370, + "token_acc": 0.9239582237188256 + }, + { + "epoch": 0.08912655971479501, + "grad_norm": 1.0716954469680786, + "learning_rate": 9.958290140391373e-06, + "loss": 0.2802464962005615, + "step": 1375, + "token_acc": 0.9227937775075666 + }, + { + "epoch": 0.08945065629557608, + "grad_norm": 1.1273002624511719, + "learning_rate": 9.957596550224285e-06, + "loss": 0.2658792018890381, + "step": 1380, + "token_acc": 0.9261406869047306 + }, + { + "epoch": 0.08977475287635715, + "grad_norm": 1.0252765417099, + "learning_rate": 9.956897265237469e-06, + "loss": 0.26928048133850097, + "step": 1385, + "token_acc": 0.924581423323255 + }, + { + "epoch": 0.09009884945713822, + "grad_norm": 1.1737946271896362, + "learning_rate": 9.956192286234196e-06, + "loss": 0.2721066474914551, + "step": 1390, + "token_acc": 0.9248243422211494 + }, + { + "epoch": 0.0904229460379193, + "grad_norm": 1.0543155670166016, + "learning_rate": 9.955481614024285e-06, + "loss": 0.2810823917388916, + "step": 1395, + "token_acc": 0.9226053639846743 + }, + { + "epoch": 0.09074704261870037, + "grad_norm": 1.058847427368164, + "learning_rate": 9.954765249424085e-06, + "loss": 0.27713589668273925, + "step": 1400, + "token_acc": 0.9238235313603509 + }, + { + "epoch": 0.09107113919948144, + "grad_norm": 1.0769394636154175, + "learning_rate": 9.954043193256492e-06, + "loss": 0.2708258628845215, + "step": 1405, + "token_acc": 0.924610216549999 + }, + { + "epoch": 0.09139523578026251, + "grad_norm": 0.9873836636543274, + "learning_rate": 9.953315446350931e-06, + "loss": 0.2721253871917725, + "step": 1410, + "token_acc": 0.9246683157184619 + }, + { + "epoch": 0.09171933236104358, + "grad_norm": 0.9685482978820801, + "learning_rate": 9.952582009543374e-06, + "loss": 0.26997599601745603, + "step": 1415, + "token_acc": 0.9243553008595988 + }, + { + "epoch": 0.09204342894182467, + "grad_norm": 1.050148367881775, + "learning_rate": 9.95184288367632e-06, + "loss": 0.27266907691955566, + "step": 1420, + "token_acc": 0.923055460025566 + }, + { + "epoch": 0.09236752552260574, + "grad_norm": 1.0243918895721436, + "learning_rate": 9.95109806959881e-06, + "loss": 0.27440733909606935, + "step": 1425, + "token_acc": 0.9242123116960466 + }, + { + "epoch": 0.09269162210338681, + "grad_norm": 0.9882935285568237, + "learning_rate": 9.950347568166415e-06, + "loss": 0.27068524360656737, + "step": 1430, + "token_acc": 0.9244608314211562 + }, + { + "epoch": 0.09301571868416789, + "grad_norm": 1.5366190671920776, + "learning_rate": 9.94959138024124e-06, + "loss": 0.27944977283477784, + "step": 1435, + "token_acc": 0.9219542443131752 + }, + { + "epoch": 0.09333981526494896, + "grad_norm": 0.9288245439529419, + "learning_rate": 9.948829506691922e-06, + "loss": 0.273094630241394, + "step": 1440, + "token_acc": 0.923447023049482 + }, + { + "epoch": 0.09366391184573003, + "grad_norm": 0.9544121623039246, + "learning_rate": 9.948061948393632e-06, + "loss": 0.2785993814468384, + "step": 1445, + "token_acc": 0.9230764193140664 + }, + { + "epoch": 0.0939880084265111, + "grad_norm": 1.113324761390686, + "learning_rate": 9.947288706228066e-06, + "loss": 0.26924405097961424, + "step": 1450, + "token_acc": 0.9248237280280711 + }, + { + "epoch": 0.09431210500729217, + "grad_norm": 1.1089950799942017, + "learning_rate": 9.946509781083453e-06, + "loss": 0.2786333322525024, + "step": 1455, + "token_acc": 0.9225569367571 + }, + { + "epoch": 0.09463620158807325, + "grad_norm": 1.0635888576507568, + "learning_rate": 9.94572517385455e-06, + "loss": 0.27387681007385256, + "step": 1460, + "token_acc": 0.9240360712963452 + }, + { + "epoch": 0.09496029816885432, + "grad_norm": 1.245919942855835, + "learning_rate": 9.944934885442637e-06, + "loss": 0.26668992042541506, + "step": 1465, + "token_acc": 0.9258613397533917 + }, + { + "epoch": 0.09528439474963539, + "grad_norm": 0.9693779945373535, + "learning_rate": 9.944138916755523e-06, + "loss": 0.27269687652587893, + "step": 1470, + "token_acc": 0.9245179277198838 + }, + { + "epoch": 0.09560849133041646, + "grad_norm": 1.0893992185592651, + "learning_rate": 9.943337268707546e-06, + "loss": 0.2674590587615967, + "step": 1475, + "token_acc": 0.9252651441212831 + }, + { + "epoch": 0.09593258791119753, + "grad_norm": 1.0595895051956177, + "learning_rate": 9.94252994221956e-06, + "loss": 0.27796216011047364, + "step": 1480, + "token_acc": 0.9230672591181994 + }, + { + "epoch": 0.0962566844919786, + "grad_norm": 1.2038002014160156, + "learning_rate": 9.941716938218946e-06, + "loss": 0.27327332496643064, + "step": 1485, + "token_acc": 0.9240932051490957 + }, + { + "epoch": 0.09658078107275968, + "grad_norm": 0.93292236328125, + "learning_rate": 9.940898257639605e-06, + "loss": 0.2675984621047974, + "step": 1490, + "token_acc": 0.9246261566693129 + }, + { + "epoch": 0.09690487765354075, + "grad_norm": 1.0741572380065918, + "learning_rate": 9.940073901421966e-06, + "loss": 0.271208667755127, + "step": 1495, + "token_acc": 0.9248086946878119 + }, + { + "epoch": 0.09722897423432182, + "grad_norm": 1.0228209495544434, + "learning_rate": 9.939243870512964e-06, + "loss": 0.27718663215637207, + "step": 1500, + "token_acc": 0.9228590957914267 + }, + { + "epoch": 0.09722897423432182, + "eval_loss": 0.2752794921398163, + "eval_runtime": 107.2536, + "eval_samples_per_second": 186.474, + "eval_steps_per_second": 0.252, + "eval_token_acc": 0.923384335416517, + "step": 1500 + }, + { + "epoch": 0.09755307081510291, + "grad_norm": 1.1989173889160156, + "learning_rate": 9.938408165866068e-06, + "loss": 0.27685868740081787, + "step": 1505, + "token_acc": 0.9227603898767474 + }, + { + "epoch": 0.09787716739588398, + "grad_norm": 1.0230028629302979, + "learning_rate": 9.937566788441249e-06, + "loss": 0.27478115558624266, + "step": 1510, + "token_acc": 0.9243062454467184 + }, + { + "epoch": 0.09820126397666505, + "grad_norm": 0.9761582612991333, + "learning_rate": 9.936719739205007e-06, + "loss": 0.26723716259002683, + "step": 1515, + "token_acc": 0.9247941838234338 + }, + { + "epoch": 0.09852536055744612, + "grad_norm": 2.422489881515503, + "learning_rate": 9.935867019130352e-06, + "loss": 0.28370051383972167, + "step": 1520, + "token_acc": 0.9212301227009779 + }, + { + "epoch": 0.0988494571382272, + "grad_norm": 0.9710668325424194, + "learning_rate": 9.935008629196804e-06, + "loss": 0.27482008934020996, + "step": 1525, + "token_acc": 0.9237486785862021 + }, + { + "epoch": 0.09917355371900827, + "grad_norm": 0.9680889248847961, + "learning_rate": 9.934144570390408e-06, + "loss": 0.2679934024810791, + "step": 1530, + "token_acc": 0.9246236853162244 + }, + { + "epoch": 0.09949765029978934, + "grad_norm": 0.9842939972877502, + "learning_rate": 9.933274843703709e-06, + "loss": 0.2746241569519043, + "step": 1535, + "token_acc": 0.9238204562945956 + }, + { + "epoch": 0.09982174688057041, + "grad_norm": 1.007063865661621, + "learning_rate": 9.932399450135765e-06, + "loss": 0.2839823722839355, + "step": 1540, + "token_acc": 0.9214075037673394 + }, + { + "epoch": 0.10014584346135148, + "grad_norm": 1.0514264106750488, + "learning_rate": 9.93151839069215e-06, + "loss": 0.2774252414703369, + "step": 1545, + "token_acc": 0.922728350670743 + }, + { + "epoch": 0.10046994004213256, + "grad_norm": 1.0856746435165405, + "learning_rate": 9.93063166638494e-06, + "loss": 0.27080397605895995, + "step": 1550, + "token_acc": 0.9244019076154153 + }, + { + "epoch": 0.10079403662291363, + "grad_norm": 1.0452642440795898, + "learning_rate": 9.929739278232721e-06, + "loss": 0.27521600723266604, + "step": 1555, + "token_acc": 0.9234253074882037 + }, + { + "epoch": 0.1011181332036947, + "grad_norm": 0.9296409487724304, + "learning_rate": 9.928841227260584e-06, + "loss": 0.27310690879821775, + "step": 1560, + "token_acc": 0.9238805390500602 + }, + { + "epoch": 0.10144222978447577, + "grad_norm": 1.110642910003662, + "learning_rate": 9.927937514500123e-06, + "loss": 0.27369258403778074, + "step": 1565, + "token_acc": 0.9242291906090954 + }, + { + "epoch": 0.10176632636525684, + "grad_norm": 1.4237020015716553, + "learning_rate": 9.927028140989444e-06, + "loss": 0.2711742877960205, + "step": 1570, + "token_acc": 0.9246883316073043 + }, + { + "epoch": 0.10209042294603791, + "grad_norm": 1.013430118560791, + "learning_rate": 9.926113107773142e-06, + "loss": 0.27250239849090574, + "step": 1575, + "token_acc": 0.9236099877708688 + }, + { + "epoch": 0.10241451952681899, + "grad_norm": 1.0401182174682617, + "learning_rate": 9.925192415902326e-06, + "loss": 0.2742006778717041, + "step": 1580, + "token_acc": 0.9229780521529144 + }, + { + "epoch": 0.10273861610760006, + "grad_norm": 0.9349300861358643, + "learning_rate": 9.9242660664346e-06, + "loss": 0.2691805362701416, + "step": 1585, + "token_acc": 0.9248303612133385 + }, + { + "epoch": 0.10306271268838113, + "grad_norm": 1.0663888454437256, + "learning_rate": 9.923334060434064e-06, + "loss": 0.2780755996704102, + "step": 1590, + "token_acc": 0.9229700063650417 + }, + { + "epoch": 0.10338680926916222, + "grad_norm": 1.0016791820526123, + "learning_rate": 9.922396398971321e-06, + "loss": 0.27017571926116946, + "step": 1595, + "token_acc": 0.9248345160521041 + }, + { + "epoch": 0.10371090584994329, + "grad_norm": 0.9433210492134094, + "learning_rate": 9.92145308312347e-06, + "loss": 0.27652616500854493, + "step": 1600, + "token_acc": 0.9227423254835997 + }, + { + "epoch": 0.10403500243072436, + "grad_norm": 1.1194981336593628, + "learning_rate": 9.920504113974097e-06, + "loss": 0.2833522319793701, + "step": 1605, + "token_acc": 0.9220748138709031 + }, + { + "epoch": 0.10435909901150543, + "grad_norm": 0.9643964171409607, + "learning_rate": 9.919549492613296e-06, + "loss": 0.27577548027038573, + "step": 1610, + "token_acc": 0.9239792879072538 + }, + { + "epoch": 0.1046831955922865, + "grad_norm": 1.0825419425964355, + "learning_rate": 9.918589220137639e-06, + "loss": 0.2689621925354004, + "step": 1615, + "token_acc": 0.9247710010757245 + }, + { + "epoch": 0.10500729217306758, + "grad_norm": 1.1451032161712646, + "learning_rate": 9.917623297650203e-06, + "loss": 0.26808867454528806, + "step": 1620, + "token_acc": 0.9237905211059574 + }, + { + "epoch": 0.10533138875384865, + "grad_norm": 0.9955279231071472, + "learning_rate": 9.916651726260544e-06, + "loss": 0.27466559410095215, + "step": 1625, + "token_acc": 0.9237133205272571 + }, + { + "epoch": 0.10565548533462972, + "grad_norm": 1.0354743003845215, + "learning_rate": 9.915674507084713e-06, + "loss": 0.27103400230407715, + "step": 1630, + "token_acc": 0.9245161040069303 + }, + { + "epoch": 0.10597958191541079, + "grad_norm": 1.0770398378372192, + "learning_rate": 9.914691641245248e-06, + "loss": 0.277128791809082, + "step": 1635, + "token_acc": 0.9233194452671972 + }, + { + "epoch": 0.10630367849619186, + "grad_norm": 0.9900469183921814, + "learning_rate": 9.913703129871172e-06, + "loss": 0.2723726749420166, + "step": 1640, + "token_acc": 0.9245738480504476 + }, + { + "epoch": 0.10662777507697294, + "grad_norm": 1.0054636001586914, + "learning_rate": 9.912708974097994e-06, + "loss": 0.2797891855239868, + "step": 1645, + "token_acc": 0.9233071800619767 + }, + { + "epoch": 0.10695187165775401, + "grad_norm": 1.2931158542633057, + "learning_rate": 9.911709175067707e-06, + "loss": 0.2797114372253418, + "step": 1650, + "token_acc": 0.9234006678566018 + }, + { + "epoch": 0.10727596823853508, + "grad_norm": 1.136385440826416, + "learning_rate": 9.910703733928783e-06, + "loss": 0.2698939800262451, + "step": 1655, + "token_acc": 0.9251408569829931 + }, + { + "epoch": 0.10760006481931615, + "grad_norm": 1.03597891330719, + "learning_rate": 9.909692651836182e-06, + "loss": 0.2703925609588623, + "step": 1660, + "token_acc": 0.9252016076072395 + }, + { + "epoch": 0.10792416140009722, + "grad_norm": 0.9790574312210083, + "learning_rate": 9.908675929951339e-06, + "loss": 0.27372188568115235, + "step": 1665, + "token_acc": 0.9229208535477016 + }, + { + "epoch": 0.1082482579808783, + "grad_norm": 1.118811011314392, + "learning_rate": 9.907653569442166e-06, + "loss": 0.274387788772583, + "step": 1670, + "token_acc": 0.9243058716431497 + }, + { + "epoch": 0.10857235456165937, + "grad_norm": 1.0487786531448364, + "learning_rate": 9.906625571483058e-06, + "loss": 0.2731483459472656, + "step": 1675, + "token_acc": 0.9245208571853635 + }, + { + "epoch": 0.10889645114244045, + "grad_norm": 0.9568978548049927, + "learning_rate": 9.905591937254878e-06, + "loss": 0.2664167881011963, + "step": 1680, + "token_acc": 0.9249768225324159 + }, + { + "epoch": 0.10922054772322153, + "grad_norm": 0.9544060826301575, + "learning_rate": 9.90455266794497e-06, + "loss": 0.27443690299987794, + "step": 1685, + "token_acc": 0.9237837802459601 + }, + { + "epoch": 0.1095446443040026, + "grad_norm": 1.0311187505722046, + "learning_rate": 9.903507764747149e-06, + "loss": 0.2735258102416992, + "step": 1690, + "token_acc": 0.9240997193930576 + }, + { + "epoch": 0.10986874088478367, + "grad_norm": 1.0488665103912354, + "learning_rate": 9.9024572288617e-06, + "loss": 0.27213072776794434, + "step": 1695, + "token_acc": 0.9231436693410002 + }, + { + "epoch": 0.11019283746556474, + "grad_norm": 1.081527590751648, + "learning_rate": 9.901401061495379e-06, + "loss": 0.27215576171875, + "step": 1700, + "token_acc": 0.9249249829360326 + }, + { + "epoch": 0.11051693404634581, + "grad_norm": 0.9950317144393921, + "learning_rate": 9.900339263861414e-06, + "loss": 0.270894193649292, + "step": 1705, + "token_acc": 0.9237205618881936 + }, + { + "epoch": 0.11084103062712689, + "grad_norm": 1.0484617948532104, + "learning_rate": 9.899271837179497e-06, + "loss": 0.2749714612960815, + "step": 1710, + "token_acc": 0.9234928306750443 + }, + { + "epoch": 0.11116512720790796, + "grad_norm": 1.0407040119171143, + "learning_rate": 9.898198782675786e-06, + "loss": 0.26599321365356443, + "step": 1715, + "token_acc": 0.9257451765515182 + }, + { + "epoch": 0.11148922378868903, + "grad_norm": 1.0516830682754517, + "learning_rate": 9.897120101582907e-06, + "loss": 0.28003873825073244, + "step": 1720, + "token_acc": 0.9234885882066052 + }, + { + "epoch": 0.1118133203694701, + "grad_norm": 0.957533597946167, + "learning_rate": 9.896035795139947e-06, + "loss": 0.27316989898681643, + "step": 1725, + "token_acc": 0.9233659981471091 + }, + { + "epoch": 0.11213741695025117, + "grad_norm": 0.9559309482574463, + "learning_rate": 9.89494586459245e-06, + "loss": 0.27594521045684817, + "step": 1730, + "token_acc": 0.9231319828994202 + }, + { + "epoch": 0.11246151353103225, + "grad_norm": 1.3301963806152344, + "learning_rate": 9.893850311192434e-06, + "loss": 0.27024025917053224, + "step": 1735, + "token_acc": 0.9244858884801119 + }, + { + "epoch": 0.11278561011181332, + "grad_norm": 1.0732120275497437, + "learning_rate": 9.892749136198361e-06, + "loss": 0.27436850070953367, + "step": 1740, + "token_acc": 0.9236043095004897 + }, + { + "epoch": 0.11310970669259439, + "grad_norm": 1.0805331468582153, + "learning_rate": 9.89164234087516e-06, + "loss": 0.26944427490234374, + "step": 1745, + "token_acc": 0.9247072927593644 + }, + { + "epoch": 0.11343380327337546, + "grad_norm": 1.1050970554351807, + "learning_rate": 9.890529926494211e-06, + "loss": 0.27786822319030763, + "step": 1750, + "token_acc": 0.9235740778029659 + }, + { + "epoch": 0.11375789985415653, + "grad_norm": 1.0643095970153809, + "learning_rate": 9.889411894333355e-06, + "loss": 0.2682131290435791, + "step": 1755, + "token_acc": 0.9250420364467354 + }, + { + "epoch": 0.1140819964349376, + "grad_norm": 0.9681152701377869, + "learning_rate": 9.888288245676879e-06, + "loss": 0.26546540260314944, + "step": 1760, + "token_acc": 0.9257162284085412 + }, + { + "epoch": 0.11440609301571869, + "grad_norm": 0.9666985273361206, + "learning_rate": 9.887158981815523e-06, + "loss": 0.277356481552124, + "step": 1765, + "token_acc": 0.9232284237387948 + }, + { + "epoch": 0.11473018959649976, + "grad_norm": 0.9490284323692322, + "learning_rate": 9.886024104046482e-06, + "loss": 0.27654027938842773, + "step": 1770, + "token_acc": 0.9241275567595632 + }, + { + "epoch": 0.11505428617728083, + "grad_norm": 1.04364013671875, + "learning_rate": 9.884883613673393e-06, + "loss": 0.2679736137390137, + "step": 1775, + "token_acc": 0.9255010184637624 + }, + { + "epoch": 0.1153783827580619, + "grad_norm": 0.9443017840385437, + "learning_rate": 9.883737512006349e-06, + "loss": 0.2656404972076416, + "step": 1780, + "token_acc": 0.925007385039551 + }, + { + "epoch": 0.11570247933884298, + "grad_norm": 1.1193122863769531, + "learning_rate": 9.882585800361879e-06, + "loss": 0.2742094278335571, + "step": 1785, + "token_acc": 0.9236529105524739 + }, + { + "epoch": 0.11602657591962405, + "grad_norm": 1.2240768671035767, + "learning_rate": 9.881428480062962e-06, + "loss": 0.27303686141967776, + "step": 1790, + "token_acc": 0.9235450299424407 + }, + { + "epoch": 0.11635067250040512, + "grad_norm": 1.0624961853027344, + "learning_rate": 9.880265552439022e-06, + "loss": 0.2721515655517578, + "step": 1795, + "token_acc": 0.9239791057524885 + }, + { + "epoch": 0.1166747690811862, + "grad_norm": 1.1981738805770874, + "learning_rate": 9.879097018825916e-06, + "loss": 0.2719876766204834, + "step": 1800, + "token_acc": 0.92382783226353 + }, + { + "epoch": 0.11699886566196727, + "grad_norm": 0.9903087019920349, + "learning_rate": 9.877922880565946e-06, + "loss": 0.27469184398651125, + "step": 1805, + "token_acc": 0.9234782033034326 + }, + { + "epoch": 0.11732296224274834, + "grad_norm": 0.9771552681922913, + "learning_rate": 9.876743139007853e-06, + "loss": 0.2729260683059692, + "step": 1810, + "token_acc": 0.9238089047773805 + }, + { + "epoch": 0.11764705882352941, + "grad_norm": 1.0332655906677246, + "learning_rate": 9.875557795506815e-06, + "loss": 0.27531943321228025, + "step": 1815, + "token_acc": 0.9233112462085216 + }, + { + "epoch": 0.11797115540431048, + "grad_norm": 0.8811163306236267, + "learning_rate": 9.874366851424439e-06, + "loss": 0.2734023094177246, + "step": 1820, + "token_acc": 0.9243820350525921 + }, + { + "epoch": 0.11829525198509155, + "grad_norm": 0.9587964415550232, + "learning_rate": 9.87317030812877e-06, + "loss": 0.2750477075576782, + "step": 1825, + "token_acc": 0.9230510186900152 + }, + { + "epoch": 0.11861934856587263, + "grad_norm": 1.0105973482131958, + "learning_rate": 9.871968166994284e-06, + "loss": 0.27099003791809084, + "step": 1830, + "token_acc": 0.9240659714944456 + }, + { + "epoch": 0.1189434451466537, + "grad_norm": 1.1160749197006226, + "learning_rate": 9.87076042940189e-06, + "loss": 0.26824226379394533, + "step": 1835, + "token_acc": 0.9241721875833013 + }, + { + "epoch": 0.11926754172743477, + "grad_norm": 0.9462841749191284, + "learning_rate": 9.869547096738921e-06, + "loss": 0.267621374130249, + "step": 1840, + "token_acc": 0.9238915268271856 + }, + { + "epoch": 0.11959163830821584, + "grad_norm": 0.8994613289833069, + "learning_rate": 9.86832817039914e-06, + "loss": 0.2728606700897217, + "step": 1845, + "token_acc": 0.9244416222275499 + }, + { + "epoch": 0.11991573488899693, + "grad_norm": 1.0033142566680908, + "learning_rate": 9.867103651782733e-06, + "loss": 0.26934051513671875, + "step": 1850, + "token_acc": 0.9246378327493332 + }, + { + "epoch": 0.120239831469778, + "grad_norm": 0.9191084504127502, + "learning_rate": 9.865873542296313e-06, + "loss": 0.2698638677597046, + "step": 1855, + "token_acc": 0.9245987381909712 + }, + { + "epoch": 0.12056392805055907, + "grad_norm": 0.9678038954734802, + "learning_rate": 9.864637843352916e-06, + "loss": 0.27739927768707273, + "step": 1860, + "token_acc": 0.9233562315996074 + }, + { + "epoch": 0.12088802463134014, + "grad_norm": 0.931466817855835, + "learning_rate": 9.86339655637199e-06, + "loss": 0.2710101842880249, + "step": 1865, + "token_acc": 0.9242738114687142 + }, + { + "epoch": 0.12121212121212122, + "grad_norm": 1.0041298866271973, + "learning_rate": 9.862149682779419e-06, + "loss": 0.2756577730178833, + "step": 1870, + "token_acc": 0.9233062133599466 + }, + { + "epoch": 0.12153621779290229, + "grad_norm": 1.1264930963516235, + "learning_rate": 9.860897224007483e-06, + "loss": 0.28103365898132326, + "step": 1875, + "token_acc": 0.9231336417557895 + }, + { + "epoch": 0.12186031437368336, + "grad_norm": 0.9130573272705078, + "learning_rate": 9.859639181494894e-06, + "loss": 0.27318921089172366, + "step": 1880, + "token_acc": 0.9240624044845908 + }, + { + "epoch": 0.12218441095446443, + "grad_norm": 0.9747406244277954, + "learning_rate": 9.858375556686772e-06, + "loss": 0.26555347442626953, + "step": 1885, + "token_acc": 0.9248226302289 + }, + { + "epoch": 0.1225085075352455, + "grad_norm": 0.9989035725593567, + "learning_rate": 9.85710635103465e-06, + "loss": 0.27396016120910643, + "step": 1890, + "token_acc": 0.9231803379255885 + }, + { + "epoch": 0.12283260411602658, + "grad_norm": 0.8758108019828796, + "learning_rate": 9.855831565996475e-06, + "loss": 0.2606497049331665, + "step": 1895, + "token_acc": 0.9271418086604646 + }, + { + "epoch": 0.12315670069680765, + "grad_norm": 1.024030327796936, + "learning_rate": 9.854551203036592e-06, + "loss": 0.27004094123840333, + "step": 1900, + "token_acc": 0.9245094033765172 + }, + { + "epoch": 0.12348079727758872, + "grad_norm": 0.9365476369857788, + "learning_rate": 9.853265263625771e-06, + "loss": 0.26974670886993407, + "step": 1905, + "token_acc": 0.9241203682967873 + }, + { + "epoch": 0.12380489385836979, + "grad_norm": 0.9061581492424011, + "learning_rate": 9.85197374924117e-06, + "loss": 0.26725306510925295, + "step": 1910, + "token_acc": 0.9252236324081317 + }, + { + "epoch": 0.12412899043915086, + "grad_norm": 0.9130187630653381, + "learning_rate": 9.850676661366362e-06, + "loss": 0.2614121913909912, + "step": 1915, + "token_acc": 0.9277056819155457 + }, + { + "epoch": 0.12445308701993194, + "grad_norm": 0.9489284753799438, + "learning_rate": 9.849374001491322e-06, + "loss": 0.2705172300338745, + "step": 1920, + "token_acc": 0.9256416343541453 + }, + { + "epoch": 0.12477718360071301, + "grad_norm": 1.0743412971496582, + "learning_rate": 9.848065771112418e-06, + "loss": 0.27531633377075193, + "step": 1925, + "token_acc": 0.9229580962437184 + }, + { + "epoch": 0.12510128018149408, + "grad_norm": 0.8972349166870117, + "learning_rate": 9.846751971732424e-06, + "loss": 0.265399694442749, + "step": 1930, + "token_acc": 0.9260574602740973 + }, + { + "epoch": 0.12542537676227516, + "grad_norm": 0.9575828313827515, + "learning_rate": 9.845432604860512e-06, + "loss": 0.2672731399536133, + "step": 1935, + "token_acc": 0.9256607965285852 + }, + { + "epoch": 0.12574947334305622, + "grad_norm": 0.9541116952896118, + "learning_rate": 9.84410767201224e-06, + "loss": 0.2785304069519043, + "step": 1940, + "token_acc": 0.9225122723310547 + }, + { + "epoch": 0.1260735699238373, + "grad_norm": 1.20053231716156, + "learning_rate": 9.842777174709572e-06, + "loss": 0.2687264919281006, + "step": 1945, + "token_acc": 0.9248639925033189 + }, + { + "epoch": 0.12639766650461837, + "grad_norm": 0.9025915265083313, + "learning_rate": 9.841441114480854e-06, + "loss": 0.26512718200683594, + "step": 1950, + "token_acc": 0.9259264123561048 + }, + { + "epoch": 0.12672176308539945, + "grad_norm": 1.2694759368896484, + "learning_rate": 9.840099492860828e-06, + "loss": 0.26184635162353515, + "step": 1955, + "token_acc": 0.9266154394838375 + }, + { + "epoch": 0.1270458596661805, + "grad_norm": 1.0990233421325684, + "learning_rate": 9.838752311390621e-06, + "loss": 0.26923418045043945, + "step": 1960, + "token_acc": 0.924598121916282 + }, + { + "epoch": 0.1273699562469616, + "grad_norm": 0.8711085915565491, + "learning_rate": 9.837399571617749e-06, + "loss": 0.26091601848602297, + "step": 1965, + "token_acc": 0.9265056249920371 + }, + { + "epoch": 0.12769405282774265, + "grad_norm": 0.9464273452758789, + "learning_rate": 9.836041275096114e-06, + "loss": 0.23966712951660157, + "step": 1970, + "token_acc": 0.9333811548231702 + }, + { + "epoch": 0.12801814940852374, + "grad_norm": 1.0471134185791016, + "learning_rate": 9.834677423385995e-06, + "loss": 0.23803000450134276, + "step": 1975, + "token_acc": 0.9326232831347467 + }, + { + "epoch": 0.12834224598930483, + "grad_norm": 1.0047551393508911, + "learning_rate": 9.83330801805406e-06, + "loss": 0.2700387954711914, + "step": 1980, + "token_acc": 0.9249814589182603 + }, + { + "epoch": 0.12866634257008588, + "grad_norm": 1.093272089958191, + "learning_rate": 9.83193306067335e-06, + "loss": 0.267792797088623, + "step": 1985, + "token_acc": 0.9250008166998791 + }, + { + "epoch": 0.12899043915086697, + "grad_norm": 0.9251362681388855, + "learning_rate": 9.830552552823287e-06, + "loss": 0.2690268039703369, + "step": 1990, + "token_acc": 0.9244774903774154 + }, + { + "epoch": 0.12931453573164803, + "grad_norm": 1.2007625102996826, + "learning_rate": 9.82916649608967e-06, + "loss": 0.27674341201782227, + "step": 1995, + "token_acc": 0.9238611085517223 + }, + { + "epoch": 0.12963863231242911, + "grad_norm": 0.8932528495788574, + "learning_rate": 9.82777489206467e-06, + "loss": 0.274359917640686, + "step": 2000, + "token_acc": 0.924495963785276 + }, + { + "epoch": 0.12963863231242911, + "eval_loss": 0.27270475029945374, + "eval_runtime": 136.3632, + "eval_samples_per_second": 146.667, + "eval_steps_per_second": 0.198, + "eval_token_acc": 0.923762660728396, + "step": 2000 + }, + { + "epoch": 0.12996272889321017, + "grad_norm": 0.9958134293556213, + "learning_rate": 9.826377742346829e-06, + "loss": 0.2681886672973633, + "step": 2005, + "token_acc": 0.9251692761383699 + }, + { + "epoch": 0.13028682547399126, + "grad_norm": 0.9843332767486572, + "learning_rate": 9.824975048541061e-06, + "loss": 0.26679039001464844, + "step": 2010, + "token_acc": 0.9254165166442997 + }, + { + "epoch": 0.13061092205477232, + "grad_norm": 0.9692593812942505, + "learning_rate": 9.82356681225865e-06, + "loss": 0.2728773593902588, + "step": 2015, + "token_acc": 0.9239635189544502 + }, + { + "epoch": 0.1309350186355534, + "grad_norm": 1.0011025667190552, + "learning_rate": 9.822153035117246e-06, + "loss": 0.2710697889328003, + "step": 2020, + "token_acc": 0.9248103965871386 + }, + { + "epoch": 0.13125911521633446, + "grad_norm": 0.9607855677604675, + "learning_rate": 9.820733718740859e-06, + "loss": 0.26751160621643066, + "step": 2025, + "token_acc": 0.9253515643044273 + }, + { + "epoch": 0.13158321179711555, + "grad_norm": 0.9440774917602539, + "learning_rate": 9.819308864759869e-06, + "loss": 0.2621784210205078, + "step": 2030, + "token_acc": 0.9258067242735742 + }, + { + "epoch": 0.1319073083778966, + "grad_norm": 0.9163006544113159, + "learning_rate": 9.817878474811013e-06, + "loss": 0.26691765785217286, + "step": 2035, + "token_acc": 0.9253205149053253 + }, + { + "epoch": 0.1322314049586777, + "grad_norm": 1.0467102527618408, + "learning_rate": 9.816442550537388e-06, + "loss": 0.2757887363433838, + "step": 2040, + "token_acc": 0.923038464535231 + }, + { + "epoch": 0.13255550153945875, + "grad_norm": 1.7916594743728638, + "learning_rate": 9.81500109358845e-06, + "loss": 0.27697553634643557, + "step": 2045, + "token_acc": 0.9227113590074679 + }, + { + "epoch": 0.13287959812023983, + "grad_norm": 0.98685622215271, + "learning_rate": 9.813554105620007e-06, + "loss": 0.27060303688049314, + "step": 2050, + "token_acc": 0.9243251420753525 + }, + { + "epoch": 0.1332036947010209, + "grad_norm": 0.9809501767158508, + "learning_rate": 9.812101588294224e-06, + "loss": 0.2721551418304443, + "step": 2055, + "token_acc": 0.9237612188694946 + }, + { + "epoch": 0.13352779128180198, + "grad_norm": 0.9539017677307129, + "learning_rate": 9.810643543279616e-06, + "loss": 0.26798052787780763, + "step": 2060, + "token_acc": 0.9251081116498493 + }, + { + "epoch": 0.13385188786258306, + "grad_norm": 0.8639307618141174, + "learning_rate": 9.809179972251046e-06, + "loss": 0.2696994304656982, + "step": 2065, + "token_acc": 0.9246522971458704 + }, + { + "epoch": 0.13417598444336412, + "grad_norm": 1.030457854270935, + "learning_rate": 9.807710876889727e-06, + "loss": 0.26218347549438475, + "step": 2070, + "token_acc": 0.9262481064669437 + }, + { + "epoch": 0.1345000810241452, + "grad_norm": 1.0565268993377686, + "learning_rate": 9.806236258883218e-06, + "loss": 0.27016291618347166, + "step": 2075, + "token_acc": 0.9239725938685904 + }, + { + "epoch": 0.13482417760492627, + "grad_norm": 0.9403888583183289, + "learning_rate": 9.804756119925421e-06, + "loss": 0.2598578453063965, + "step": 2080, + "token_acc": 0.9272878933857525 + }, + { + "epoch": 0.13514827418570735, + "grad_norm": 0.9595038294792175, + "learning_rate": 9.80327046171658e-06, + "loss": 0.2734077930450439, + "step": 2085, + "token_acc": 0.923968586251284 + }, + { + "epoch": 0.1354723707664884, + "grad_norm": 1.557283878326416, + "learning_rate": 9.801779285963279e-06, + "loss": 0.27282917499542236, + "step": 2090, + "token_acc": 0.9233787199131965 + }, + { + "epoch": 0.1357964673472695, + "grad_norm": 0.8996778726577759, + "learning_rate": 9.800282594378441e-06, + "loss": 0.26968817710876464, + "step": 2095, + "token_acc": 0.9252194913671177 + }, + { + "epoch": 0.13612056392805055, + "grad_norm": 0.8910385966300964, + "learning_rate": 9.798780388681322e-06, + "loss": 0.2675360679626465, + "step": 2100, + "token_acc": 0.9249030286154624 + }, + { + "epoch": 0.13644466050883164, + "grad_norm": 1.030527114868164, + "learning_rate": 9.797272670597518e-06, + "loss": 0.27029075622558596, + "step": 2105, + "token_acc": 0.9239693404820155 + }, + { + "epoch": 0.1367687570896127, + "grad_norm": 0.9011632204055786, + "learning_rate": 9.795759441858952e-06, + "loss": 0.2744141101837158, + "step": 2110, + "token_acc": 0.9237051322418136 + }, + { + "epoch": 0.13709285367039378, + "grad_norm": 0.9581583738327026, + "learning_rate": 9.794240704203876e-06, + "loss": 0.28199870586395265, + "step": 2115, + "token_acc": 0.9227496655900297 + }, + { + "epoch": 0.13741695025117484, + "grad_norm": 1.2229371070861816, + "learning_rate": 9.792716459376875e-06, + "loss": 0.27021174430847167, + "step": 2120, + "token_acc": 0.9239751295068044 + }, + { + "epoch": 0.13774104683195593, + "grad_norm": 0.8828976154327393, + "learning_rate": 9.791186709128859e-06, + "loss": 0.26480114459991455, + "step": 2125, + "token_acc": 0.9258972559155586 + }, + { + "epoch": 0.13806514341273698, + "grad_norm": 0.9912476539611816, + "learning_rate": 9.789651455217058e-06, + "loss": 0.2708422183990479, + "step": 2130, + "token_acc": 0.9241014235454019 + }, + { + "epoch": 0.13838923999351807, + "grad_norm": 0.9211575984954834, + "learning_rate": 9.78811069940503e-06, + "loss": 0.2720188617706299, + "step": 2135, + "token_acc": 0.9244300236218382 + }, + { + "epoch": 0.13871333657429913, + "grad_norm": 1.0295453071594238, + "learning_rate": 9.786564443462646e-06, + "loss": 0.2796108961105347, + "step": 2140, + "token_acc": 0.9226034794602542 + }, + { + "epoch": 0.13903743315508021, + "grad_norm": 1.0893932580947876, + "learning_rate": 9.785012689166104e-06, + "loss": 0.2808856964111328, + "step": 2145, + "token_acc": 0.9218724213278058 + }, + { + "epoch": 0.1393615297358613, + "grad_norm": 0.9018267393112183, + "learning_rate": 9.783455438297909e-06, + "loss": 0.2733065366744995, + "step": 2150, + "token_acc": 0.9236945402640656 + }, + { + "epoch": 0.13968562631664236, + "grad_norm": 1.0649100542068481, + "learning_rate": 9.781892692646886e-06, + "loss": 0.2655647277832031, + "step": 2155, + "token_acc": 0.9260286699423591 + }, + { + "epoch": 0.14000972289742344, + "grad_norm": 0.906578779220581, + "learning_rate": 9.780324454008172e-06, + "loss": 0.2723796844482422, + "step": 2160, + "token_acc": 0.9241845140032949 + }, + { + "epoch": 0.1403338194782045, + "grad_norm": 0.990885317325592, + "learning_rate": 9.778750724183209e-06, + "loss": 0.269470739364624, + "step": 2165, + "token_acc": 0.9245861834891128 + }, + { + "epoch": 0.1406579160589856, + "grad_norm": 0.912858247756958, + "learning_rate": 9.777171504979748e-06, + "loss": 0.2701325654983521, + "step": 2170, + "token_acc": 0.9247378898473789 + }, + { + "epoch": 0.14098201263976665, + "grad_norm": 0.9650824069976807, + "learning_rate": 9.77558679821185e-06, + "loss": 0.267457389831543, + "step": 2175, + "token_acc": 0.9255161023947152 + }, + { + "epoch": 0.14130610922054773, + "grad_norm": 0.896472692489624, + "learning_rate": 9.773996605699876e-06, + "loss": 0.26908278465270996, + "step": 2180, + "token_acc": 0.9240542618486037 + }, + { + "epoch": 0.1416302058013288, + "grad_norm": 0.9619896411895752, + "learning_rate": 9.77240092927049e-06, + "loss": 0.26523962020874026, + "step": 2185, + "token_acc": 0.9264276228419654 + }, + { + "epoch": 0.14195430238210988, + "grad_norm": 0.9019481539726257, + "learning_rate": 9.770799770756653e-06, + "loss": 0.26586222648620605, + "step": 2190, + "token_acc": 0.9255034080857485 + }, + { + "epoch": 0.14227839896289093, + "grad_norm": 0.9216703176498413, + "learning_rate": 9.769193131997627e-06, + "loss": 0.2637033224105835, + "step": 2195, + "token_acc": 0.92591671845867 + }, + { + "epoch": 0.14260249554367202, + "grad_norm": 0.9822421669960022, + "learning_rate": 9.767581014838968e-06, + "loss": 0.2749229907989502, + "step": 2200, + "token_acc": 0.9239928361968865 + }, + { + "epoch": 0.14292659212445308, + "grad_norm": 0.9890458583831787, + "learning_rate": 9.765963421132521e-06, + "loss": 0.2718761444091797, + "step": 2205, + "token_acc": 0.9240334674048278 + }, + { + "epoch": 0.14325068870523416, + "grad_norm": 0.8598413467407227, + "learning_rate": 9.764340352736425e-06, + "loss": 0.2701383113861084, + "step": 2210, + "token_acc": 0.9246272453332289 + }, + { + "epoch": 0.14357478528601522, + "grad_norm": 1.0562912225723267, + "learning_rate": 9.762711811515112e-06, + "loss": 0.27956080436706543, + "step": 2215, + "token_acc": 0.9224512571391059 + }, + { + "epoch": 0.1438988818667963, + "grad_norm": 1.060282826423645, + "learning_rate": 9.761077799339294e-06, + "loss": 0.26948440074920654, + "step": 2220, + "token_acc": 0.924671926416366 + }, + { + "epoch": 0.14422297844757737, + "grad_norm": 0.8981298208236694, + "learning_rate": 9.75943831808597e-06, + "loss": 0.26290059089660645, + "step": 2225, + "token_acc": 0.9258906019785308 + }, + { + "epoch": 0.14454707502835845, + "grad_norm": 0.8841952681541443, + "learning_rate": 9.757793369638421e-06, + "loss": 0.27398066520690917, + "step": 2230, + "token_acc": 0.9234172019753215 + }, + { + "epoch": 0.14487117160913954, + "grad_norm": 1.0952105522155762, + "learning_rate": 9.756142955886211e-06, + "loss": 0.2776633739471436, + "step": 2235, + "token_acc": 0.9233878971535947 + }, + { + "epoch": 0.1451952681899206, + "grad_norm": 0.9648717045783997, + "learning_rate": 9.75448707872518e-06, + "loss": 0.2631640672683716, + "step": 2240, + "token_acc": 0.925903796404512 + }, + { + "epoch": 0.14551936477070168, + "grad_norm": 0.905194103717804, + "learning_rate": 9.75282574005744e-06, + "loss": 0.26733789443969724, + "step": 2245, + "token_acc": 0.9254697042386837 + }, + { + "epoch": 0.14584346135148274, + "grad_norm": 1.4902266263961792, + "learning_rate": 9.751158941791386e-06, + "loss": 0.2674945592880249, + "step": 2250, + "token_acc": 0.9249406081197659 + }, + { + "epoch": 0.14616755793226383, + "grad_norm": 0.961801290512085, + "learning_rate": 9.749486685841673e-06, + "loss": 0.2644834041595459, + "step": 2255, + "token_acc": 0.9258671473448409 + }, + { + "epoch": 0.14649165451304488, + "grad_norm": 0.9748981595039368, + "learning_rate": 9.747808974129237e-06, + "loss": 0.2630170345306396, + "step": 2260, + "token_acc": 0.9259586140336395 + }, + { + "epoch": 0.14681575109382597, + "grad_norm": 0.9286852478981018, + "learning_rate": 9.746125808581266e-06, + "loss": 0.26565778255462646, + "step": 2265, + "token_acc": 0.9248479041287384 + }, + { + "epoch": 0.14713984767460703, + "grad_norm": 0.9646102786064148, + "learning_rate": 9.744437191131232e-06, + "loss": 0.2667598485946655, + "step": 2270, + "token_acc": 0.9253188446493525 + }, + { + "epoch": 0.1474639442553881, + "grad_norm": 1.0613057613372803, + "learning_rate": 9.742743123718854e-06, + "loss": 0.26466832160949705, + "step": 2275, + "token_acc": 0.9258959831528436 + }, + { + "epoch": 0.14778804083616917, + "grad_norm": 0.9826300144195557, + "learning_rate": 9.741043608290117e-06, + "loss": 0.27543792724609373, + "step": 2280, + "token_acc": 0.9231002745034717 + }, + { + "epoch": 0.14811213741695026, + "grad_norm": 1.1122735738754272, + "learning_rate": 9.739338646797268e-06, + "loss": 0.26322145462036134, + "step": 2285, + "token_acc": 0.9261482021193644 + }, + { + "epoch": 0.14843623399773131, + "grad_norm": 1.0941451787948608, + "learning_rate": 9.7376282411988e-06, + "loss": 0.2738584756851196, + "step": 2290, + "token_acc": 0.9229976452477708 + }, + { + "epoch": 0.1487603305785124, + "grad_norm": 0.8940174579620361, + "learning_rate": 9.735912393459471e-06, + "loss": 0.269028377532959, + "step": 2295, + "token_acc": 0.9249781262951535 + }, + { + "epoch": 0.14908442715929346, + "grad_norm": 1.19048011302948, + "learning_rate": 9.734191105550284e-06, + "loss": 0.2797422409057617, + "step": 2300, + "token_acc": 0.9231565824797008 + }, + { + "epoch": 0.14940852374007454, + "grad_norm": 1.085537075996399, + "learning_rate": 9.73246437944849e-06, + "loss": 0.27753300666809083, + "step": 2305, + "token_acc": 0.9228830776443623 + }, + { + "epoch": 0.1497326203208556, + "grad_norm": 0.9471875429153442, + "learning_rate": 9.730732217137593e-06, + "loss": 0.2848607063293457, + "step": 2310, + "token_acc": 0.9216459830176356 + }, + { + "epoch": 0.1500567169016367, + "grad_norm": 0.9744594693183899, + "learning_rate": 9.728994620607335e-06, + "loss": 0.26773533821105955, + "step": 2315, + "token_acc": 0.9250412825547075 + }, + { + "epoch": 0.15038081348241775, + "grad_norm": 0.9366313815116882, + "learning_rate": 9.727251591853703e-06, + "loss": 0.27124905586242676, + "step": 2320, + "token_acc": 0.9235699334460926 + }, + { + "epoch": 0.15070491006319883, + "grad_norm": 0.9696183204650879, + "learning_rate": 9.725503132878926e-06, + "loss": 0.2665980339050293, + "step": 2325, + "token_acc": 0.9253111210928211 + }, + { + "epoch": 0.15102900664397992, + "grad_norm": 1.0398274660110474, + "learning_rate": 9.72374924569147e-06, + "loss": 0.26602926254272463, + "step": 2330, + "token_acc": 0.9250271500232714 + }, + { + "epoch": 0.15135310322476098, + "grad_norm": 1.1119755506515503, + "learning_rate": 9.72198993230603e-06, + "loss": 0.2673789501190186, + "step": 2335, + "token_acc": 0.926091141425205 + }, + { + "epoch": 0.15167719980554206, + "grad_norm": 0.9922897815704346, + "learning_rate": 9.720225194743544e-06, + "loss": 0.2739012002944946, + "step": 2340, + "token_acc": 0.9238379201893591 + }, + { + "epoch": 0.15200129638632312, + "grad_norm": 0.9148727059364319, + "learning_rate": 9.718455035031176e-06, + "loss": 0.27413690090179443, + "step": 2345, + "token_acc": 0.923864671437866 + }, + { + "epoch": 0.1523253929671042, + "grad_norm": 0.954188883304596, + "learning_rate": 9.716679455202317e-06, + "loss": 0.26938366889953613, + "step": 2350, + "token_acc": 0.9244251666754161 + }, + { + "epoch": 0.15264948954788526, + "grad_norm": 0.9454711675643921, + "learning_rate": 9.714898457296585e-06, + "loss": 0.26335599422454836, + "step": 2355, + "token_acc": 0.9264231728171055 + }, + { + "epoch": 0.15297358612866635, + "grad_norm": 0.8574007153511047, + "learning_rate": 9.713112043359826e-06, + "loss": 0.2716971397399902, + "step": 2360, + "token_acc": 0.9248679899533927 + }, + { + "epoch": 0.1532976827094474, + "grad_norm": 0.9906074404716492, + "learning_rate": 9.7113202154441e-06, + "loss": 0.27153911590576174, + "step": 2365, + "token_acc": 0.9240451729276136 + }, + { + "epoch": 0.1536217792902285, + "grad_norm": 0.8852605819702148, + "learning_rate": 9.709522975607694e-06, + "loss": 0.27397239208221436, + "step": 2370, + "token_acc": 0.9238037302131735 + }, + { + "epoch": 0.15394587587100955, + "grad_norm": 1.0331032276153564, + "learning_rate": 9.707720325915105e-06, + "loss": 0.2730919122695923, + "step": 2375, + "token_acc": 0.9246810885165911 + }, + { + "epoch": 0.15426997245179064, + "grad_norm": 1.0172979831695557, + "learning_rate": 9.705912268437048e-06, + "loss": 0.2705503225326538, + "step": 2380, + "token_acc": 0.9250787249504082 + }, + { + "epoch": 0.1545940690325717, + "grad_norm": 0.9462108612060547, + "learning_rate": 9.704098805250449e-06, + "loss": 0.2731978416442871, + "step": 2385, + "token_acc": 0.9230673431129639 + }, + { + "epoch": 0.15491816561335278, + "grad_norm": 0.9303444623947144, + "learning_rate": 9.702279938438442e-06, + "loss": 0.27066688537597655, + "step": 2390, + "token_acc": 0.9241699971266619 + }, + { + "epoch": 0.15524226219413384, + "grad_norm": 0.8797633647918701, + "learning_rate": 9.700455670090373e-06, + "loss": 0.26954338550567625, + "step": 2395, + "token_acc": 0.9248562382890742 + }, + { + "epoch": 0.15556635877491493, + "grad_norm": 1.0779069662094116, + "learning_rate": 9.698626002301787e-06, + "loss": 0.27100348472595215, + "step": 2400, + "token_acc": 0.9248093586951516 + }, + { + "epoch": 0.15589045535569598, + "grad_norm": 0.9742034673690796, + "learning_rate": 9.696790937174437e-06, + "loss": 0.2760479211807251, + "step": 2405, + "token_acc": 0.9233303621537751 + }, + { + "epoch": 0.15621455193647707, + "grad_norm": 1.021911382675171, + "learning_rate": 9.694950476816267e-06, + "loss": 0.26833267211914064, + "step": 2410, + "token_acc": 0.9240885492810407 + }, + { + "epoch": 0.15653864851725816, + "grad_norm": 1.056108832359314, + "learning_rate": 9.69310462334143e-06, + "loss": 0.2755893707275391, + "step": 2415, + "token_acc": 0.9229099442651804 + }, + { + "epoch": 0.1568627450980392, + "grad_norm": 1.103298306465149, + "learning_rate": 9.691253378870266e-06, + "loss": 0.26777632236480714, + "step": 2420, + "token_acc": 0.9254783351962653 + }, + { + "epoch": 0.1571868416788203, + "grad_norm": 0.9593473672866821, + "learning_rate": 9.689396745529313e-06, + "loss": 0.26924068927764894, + "step": 2425, + "token_acc": 0.9233803482472779 + }, + { + "epoch": 0.15751093825960136, + "grad_norm": 0.9766969680786133, + "learning_rate": 9.687534725451293e-06, + "loss": 0.27867262363433837, + "step": 2430, + "token_acc": 0.9226700415368639 + }, + { + "epoch": 0.15783503484038244, + "grad_norm": 0.8712724447250366, + "learning_rate": 9.68566732077512e-06, + "loss": 0.2709386348724365, + "step": 2435, + "token_acc": 0.9248605333906171 + }, + { + "epoch": 0.1581591314211635, + "grad_norm": 0.8471109867095947, + "learning_rate": 9.683794533645892e-06, + "loss": 0.26900711059570315, + "step": 2440, + "token_acc": 0.9250147106897679 + }, + { + "epoch": 0.1584832280019446, + "grad_norm": 0.9042940139770508, + "learning_rate": 9.681916366214893e-06, + "loss": 0.26176202297210693, + "step": 2445, + "token_acc": 0.926466090057726 + }, + { + "epoch": 0.15880732458272565, + "grad_norm": 0.9194527864456177, + "learning_rate": 9.680032820639582e-06, + "loss": 0.26869964599609375, + "step": 2450, + "token_acc": 0.9253670499472638 + }, + { + "epoch": 0.15913142116350673, + "grad_norm": 0.9261711835861206, + "learning_rate": 9.6781438990836e-06, + "loss": 0.2819127082824707, + "step": 2455, + "token_acc": 0.921958007399749 + }, + { + "epoch": 0.1594555177442878, + "grad_norm": 1.810258388519287, + "learning_rate": 9.676249603716763e-06, + "loss": 0.27083301544189453, + "step": 2460, + "token_acc": 0.9252204861450614 + }, + { + "epoch": 0.15977961432506887, + "grad_norm": 1.0610450506210327, + "learning_rate": 9.674349936715056e-06, + "loss": 0.26989331245422366, + "step": 2465, + "token_acc": 0.9244125454724216 + }, + { + "epoch": 0.16010371090584993, + "grad_norm": 0.9772783517837524, + "learning_rate": 9.67244490026064e-06, + "loss": 0.2681535005569458, + "step": 2470, + "token_acc": 0.9246753754183977 + }, + { + "epoch": 0.16042780748663102, + "grad_norm": 1.0335973501205444, + "learning_rate": 9.670534496541842e-06, + "loss": 0.2709876775741577, + "step": 2475, + "token_acc": 0.9242767823203262 + }, + { + "epoch": 0.16075190406741208, + "grad_norm": 0.9158135652542114, + "learning_rate": 9.668618727753151e-06, + "loss": 0.2754196643829346, + "step": 2480, + "token_acc": 0.9224319483535733 + }, + { + "epoch": 0.16107600064819316, + "grad_norm": 0.9066814184188843, + "learning_rate": 9.666697596095222e-06, + "loss": 0.26880836486816406, + "step": 2485, + "token_acc": 0.9258791503618684 + }, + { + "epoch": 0.16140009722897422, + "grad_norm": 0.8360012173652649, + "learning_rate": 9.664771103774872e-06, + "loss": 0.26613950729370117, + "step": 2490, + "token_acc": 0.9250717804177752 + }, + { + "epoch": 0.1617241938097553, + "grad_norm": 1.3411678075790405, + "learning_rate": 9.662839253005074e-06, + "loss": 0.2622185230255127, + "step": 2495, + "token_acc": 0.9262349691783258 + }, + { + "epoch": 0.1620482903905364, + "grad_norm": 0.9041964411735535, + "learning_rate": 9.660902046004954e-06, + "loss": 0.2633491516113281, + "step": 2500, + "token_acc": 0.9261170377241806 + }, + { + "epoch": 0.1620482903905364, + "eval_loss": 0.2693125307559967, + "eval_runtime": 97.9366, + "eval_samples_per_second": 204.214, + "eval_steps_per_second": 0.276, + "eval_token_acc": 0.9247300241840863, + "step": 2500 + }, + { + "epoch": 0.16237238697131745, + "grad_norm": 0.9021881818771362, + "learning_rate": 9.658959484999793e-06, + "loss": 0.27304782867431643, + "step": 2505, + "token_acc": 0.9239885547414668 + }, + { + "epoch": 0.16269648355209854, + "grad_norm": 0.946571946144104, + "learning_rate": 9.657011572221025e-06, + "loss": 0.2680337429046631, + "step": 2510, + "token_acc": 0.9243992750883323 + }, + { + "epoch": 0.1630205801328796, + "grad_norm": 1.0762147903442383, + "learning_rate": 9.655058309906225e-06, + "loss": 0.27248587608337405, + "step": 2515, + "token_acc": 0.9241179059669038 + }, + { + "epoch": 0.16334467671366068, + "grad_norm": 0.8797369599342346, + "learning_rate": 9.653099700299117e-06, + "loss": 0.26749458312988283, + "step": 2520, + "token_acc": 0.9240640313622999 + }, + { + "epoch": 0.16366877329444174, + "grad_norm": 0.975339949131012, + "learning_rate": 9.651135745649569e-06, + "loss": 0.2744440078735352, + "step": 2525, + "token_acc": 0.922800012879544 + }, + { + "epoch": 0.16399286987522282, + "grad_norm": 0.942367434501648, + "learning_rate": 9.649166448213587e-06, + "loss": 0.26875944137573243, + "step": 2530, + "token_acc": 0.924911497410969 + }, + { + "epoch": 0.16431696645600388, + "grad_norm": 0.9769216179847717, + "learning_rate": 9.647191810253313e-06, + "loss": 0.27029297351837156, + "step": 2535, + "token_acc": 0.9244857114734736 + }, + { + "epoch": 0.16464106303678497, + "grad_norm": 0.8727836608886719, + "learning_rate": 9.645211834037026e-06, + "loss": 0.2656926155090332, + "step": 2540, + "token_acc": 0.925261026014118 + }, + { + "epoch": 0.16496515961756603, + "grad_norm": 1.0316293239593506, + "learning_rate": 9.643226521839135e-06, + "loss": 0.2695032596588135, + "step": 2545, + "token_acc": 0.9253708809242484 + }, + { + "epoch": 0.1652892561983471, + "grad_norm": 0.955760657787323, + "learning_rate": 9.641235875940182e-06, + "loss": 0.2700937271118164, + "step": 2550, + "token_acc": 0.9251608090511121 + }, + { + "epoch": 0.16561335277912817, + "grad_norm": 0.9320876002311707, + "learning_rate": 9.639239898626831e-06, + "loss": 0.26358623504638673, + "step": 2555, + "token_acc": 0.9256394497332356 + }, + { + "epoch": 0.16593744935990926, + "grad_norm": 1.0790841579437256, + "learning_rate": 9.637238592191872e-06, + "loss": 0.26795163154602053, + "step": 2560, + "token_acc": 0.9259342608649608 + }, + { + "epoch": 0.16626154594069031, + "grad_norm": 1.0255367755889893, + "learning_rate": 9.63523195893422e-06, + "loss": 0.26568045616149905, + "step": 2565, + "token_acc": 0.9256250892057767 + }, + { + "epoch": 0.1665856425214714, + "grad_norm": 0.8673477172851562, + "learning_rate": 9.633220001158905e-06, + "loss": 0.2709315776824951, + "step": 2570, + "token_acc": 0.9240601750188456 + }, + { + "epoch": 0.16690973910225246, + "grad_norm": 0.9090352058410645, + "learning_rate": 9.631202721177072e-06, + "loss": 0.26872954368591306, + "step": 2575, + "token_acc": 0.9252753083064819 + }, + { + "epoch": 0.16723383568303354, + "grad_norm": 0.872573971748352, + "learning_rate": 9.629180121305987e-06, + "loss": 0.26783461570739747, + "step": 2580, + "token_acc": 0.9249155455074131 + }, + { + "epoch": 0.16755793226381463, + "grad_norm": 0.9576750993728638, + "learning_rate": 9.627152203869018e-06, + "loss": 0.2675694227218628, + "step": 2585, + "token_acc": 0.9259019987862415 + }, + { + "epoch": 0.1678820288445957, + "grad_norm": 0.8838078379631042, + "learning_rate": 9.625118971195648e-06, + "loss": 0.27301788330078125, + "step": 2590, + "token_acc": 0.9238369298325387 + }, + { + "epoch": 0.16820612542537677, + "grad_norm": 0.898270308971405, + "learning_rate": 9.62308042562146e-06, + "loss": 0.26538777351379395, + "step": 2595, + "token_acc": 0.9249246899573569 + }, + { + "epoch": 0.16853022200615783, + "grad_norm": 0.8839157223701477, + "learning_rate": 9.621036569488147e-06, + "loss": 0.27141058444976807, + "step": 2600, + "token_acc": 0.9235273968090852 + }, + { + "epoch": 0.16885431858693892, + "grad_norm": 0.8628047704696655, + "learning_rate": 9.618987405143496e-06, + "loss": 0.26827387809753417, + "step": 2605, + "token_acc": 0.9245675376047127 + }, + { + "epoch": 0.16917841516771998, + "grad_norm": 1.3759979009628296, + "learning_rate": 9.616932934941395e-06, + "loss": 0.26434149742126467, + "step": 2610, + "token_acc": 0.9254122401094962 + }, + { + "epoch": 0.16950251174850106, + "grad_norm": 0.8740105032920837, + "learning_rate": 9.614873161241823e-06, + "loss": 0.26240692138671873, + "step": 2615, + "token_acc": 0.9254635311287274 + }, + { + "epoch": 0.16982660832928212, + "grad_norm": 0.861902117729187, + "learning_rate": 9.612808086410858e-06, + "loss": 0.26700992584228517, + "step": 2620, + "token_acc": 0.9256909531029713 + }, + { + "epoch": 0.1701507049100632, + "grad_norm": 0.9149457812309265, + "learning_rate": 9.610737712820662e-06, + "loss": 0.25867371559143065, + "step": 2625, + "token_acc": 0.9270531967576302 + }, + { + "epoch": 0.17047480149084426, + "grad_norm": 0.9046423435211182, + "learning_rate": 9.608662042849482e-06, + "loss": 0.2661403179168701, + "step": 2630, + "token_acc": 0.9255159977262778 + }, + { + "epoch": 0.17079889807162535, + "grad_norm": 0.946203887462616, + "learning_rate": 9.606581078881654e-06, + "loss": 0.2612419366836548, + "step": 2635, + "token_acc": 0.9260435511772551 + }, + { + "epoch": 0.1711229946524064, + "grad_norm": 0.9279782772064209, + "learning_rate": 9.604494823307596e-06, + "loss": 0.2729593276977539, + "step": 2640, + "token_acc": 0.9243067391628377 + }, + { + "epoch": 0.1714470912331875, + "grad_norm": 1.1244595050811768, + "learning_rate": 9.6024032785238e-06, + "loss": 0.26831631660461425, + "step": 2645, + "token_acc": 0.9245293999728407 + }, + { + "epoch": 0.17177118781396855, + "grad_norm": 1.0777920484542847, + "learning_rate": 9.600306446932833e-06, + "loss": 0.2622260093688965, + "step": 2650, + "token_acc": 0.9261585715509798 + }, + { + "epoch": 0.17209528439474964, + "grad_norm": 0.8829298615455627, + "learning_rate": 9.598204330943342e-06, + "loss": 0.2691788911819458, + "step": 2655, + "token_acc": 0.9246981209096535 + }, + { + "epoch": 0.1724193809755307, + "grad_norm": 0.8727412819862366, + "learning_rate": 9.596096932970035e-06, + "loss": 0.2701603412628174, + "step": 2660, + "token_acc": 0.9251781242113029 + }, + { + "epoch": 0.17274347755631178, + "grad_norm": 0.8715673685073853, + "learning_rate": 9.593984255433696e-06, + "loss": 0.269091272354126, + "step": 2665, + "token_acc": 0.9246743830756882 + }, + { + "epoch": 0.17306757413709287, + "grad_norm": 0.8858821392059326, + "learning_rate": 9.591866300761168e-06, + "loss": 0.2670030117034912, + "step": 2670, + "token_acc": 0.9260644234346337 + }, + { + "epoch": 0.17339167071787392, + "grad_norm": 0.937385082244873, + "learning_rate": 9.589743071385358e-06, + "loss": 0.2634093761444092, + "step": 2675, + "token_acc": 0.9263213011748536 + }, + { + "epoch": 0.173715767298655, + "grad_norm": 1.0016553401947021, + "learning_rate": 9.587614569745232e-06, + "loss": 0.2667978763580322, + "step": 2680, + "token_acc": 0.9253238522986525 + }, + { + "epoch": 0.17403986387943607, + "grad_norm": 0.844342827796936, + "learning_rate": 9.585480798285813e-06, + "loss": 0.2686027765274048, + "step": 2685, + "token_acc": 0.9248242850608479 + }, + { + "epoch": 0.17436396046021715, + "grad_norm": 0.8686570525169373, + "learning_rate": 9.583341759458176e-06, + "loss": 0.27094235420227053, + "step": 2690, + "token_acc": 0.9233838057367469 + }, + { + "epoch": 0.1746880570409982, + "grad_norm": 0.961058497428894, + "learning_rate": 9.581197455719446e-06, + "loss": 0.2694754362106323, + "step": 2695, + "token_acc": 0.9245497407499315 + }, + { + "epoch": 0.1750121536217793, + "grad_norm": 0.9178889393806458, + "learning_rate": 9.579047889532798e-06, + "loss": 0.2602999210357666, + "step": 2700, + "token_acc": 0.9271164244129507 + }, + { + "epoch": 0.17533625020256036, + "grad_norm": 0.891244113445282, + "learning_rate": 9.57689306336745e-06, + "loss": 0.2633685111999512, + "step": 2705, + "token_acc": 0.9264931087289433 + }, + { + "epoch": 0.17566034678334144, + "grad_norm": 0.8418755531311035, + "learning_rate": 9.574732979698667e-06, + "loss": 0.2603219270706177, + "step": 2710, + "token_acc": 0.926565407700574 + }, + { + "epoch": 0.1759844433641225, + "grad_norm": 0.9048406481742859, + "learning_rate": 9.572567641007745e-06, + "loss": 0.26884543895721436, + "step": 2715, + "token_acc": 0.9254149261952694 + }, + { + "epoch": 0.1763085399449036, + "grad_norm": 0.8801888823509216, + "learning_rate": 9.570397049782023e-06, + "loss": 0.2616330862045288, + "step": 2720, + "token_acc": 0.9265229181824195 + }, + { + "epoch": 0.17663263652568464, + "grad_norm": 0.9177769422531128, + "learning_rate": 9.568221208514872e-06, + "loss": 0.27637467384338377, + "step": 2725, + "token_acc": 0.923530262016767 + }, + { + "epoch": 0.17695673310646573, + "grad_norm": 0.9469224214553833, + "learning_rate": 9.566040119705692e-06, + "loss": 0.2751315593719482, + "step": 2730, + "token_acc": 0.923161246541734 + }, + { + "epoch": 0.1772808296872468, + "grad_norm": 0.8869476318359375, + "learning_rate": 9.563853785859914e-06, + "loss": 0.26172842979431155, + "step": 2735, + "token_acc": 0.9263717969642565 + }, + { + "epoch": 0.17760492626802787, + "grad_norm": 0.9011089205741882, + "learning_rate": 9.561662209488991e-06, + "loss": 0.2768300294876099, + "step": 2740, + "token_acc": 0.9224174909984272 + }, + { + "epoch": 0.17792902284880893, + "grad_norm": 0.858426034450531, + "learning_rate": 9.559465393110401e-06, + "loss": 0.25355207920074463, + "step": 2745, + "token_acc": 0.9281829223213814 + }, + { + "epoch": 0.17825311942959002, + "grad_norm": 0.8865728974342346, + "learning_rate": 9.557263339247637e-06, + "loss": 0.2730297565460205, + "step": 2750, + "token_acc": 0.9236890931776601 + }, + { + "epoch": 0.1785772160103711, + "grad_norm": 0.9161731600761414, + "learning_rate": 9.555056050430213e-06, + "loss": 0.25940628051757814, + "step": 2755, + "token_acc": 0.9267621837127348 + }, + { + "epoch": 0.17890131259115216, + "grad_norm": 0.8771811127662659, + "learning_rate": 9.552843529193655e-06, + "loss": 0.27488343715667723, + "step": 2760, + "token_acc": 0.9232528309709518 + }, + { + "epoch": 0.17922540917193325, + "grad_norm": 0.8269277811050415, + "learning_rate": 9.550625778079496e-06, + "loss": 0.26508274078369143, + "step": 2765, + "token_acc": 0.9256487612480246 + }, + { + "epoch": 0.1795495057527143, + "grad_norm": 1.0027071237564087, + "learning_rate": 9.548402799635283e-06, + "loss": 0.2624945163726807, + "step": 2770, + "token_acc": 0.9273640396302738 + }, + { + "epoch": 0.1798736023334954, + "grad_norm": 1.0807225704193115, + "learning_rate": 9.54617459641456e-06, + "loss": 0.27345755100250246, + "step": 2775, + "token_acc": 0.9232509426057813 + }, + { + "epoch": 0.18019769891427645, + "grad_norm": 0.9232643246650696, + "learning_rate": 9.543941170976883e-06, + "loss": 0.2619877815246582, + "step": 2780, + "token_acc": 0.9259899094995976 + }, + { + "epoch": 0.18052179549505754, + "grad_norm": 0.9397012591362, + "learning_rate": 9.541702525887794e-06, + "loss": 0.26745309829711916, + "step": 2785, + "token_acc": 0.9250451733734288 + }, + { + "epoch": 0.1808458920758386, + "grad_norm": 0.8636280298233032, + "learning_rate": 9.539458663718842e-06, + "loss": 0.2638130187988281, + "step": 2790, + "token_acc": 0.9266646512972845 + }, + { + "epoch": 0.18116998865661968, + "grad_norm": 0.8776155114173889, + "learning_rate": 9.537209587047562e-06, + "loss": 0.26659042835235597, + "step": 2795, + "token_acc": 0.9250484445375511 + }, + { + "epoch": 0.18149408523740074, + "grad_norm": 0.8816049098968506, + "learning_rate": 9.534955298457483e-06, + "loss": 0.2618479013442993, + "step": 2800, + "token_acc": 0.9261779866487263 + }, + { + "epoch": 0.18181818181818182, + "grad_norm": 0.8553512692451477, + "learning_rate": 9.532695800538117e-06, + "loss": 0.26416175365447997, + "step": 2805, + "token_acc": 0.9264781962371401 + }, + { + "epoch": 0.18214227839896288, + "grad_norm": 0.9307749271392822, + "learning_rate": 9.530431095884966e-06, + "loss": 0.27537033557891843, + "step": 2810, + "token_acc": 0.9237300387724255 + }, + { + "epoch": 0.18246637497974397, + "grad_norm": 0.9014064073562622, + "learning_rate": 9.528161187099505e-06, + "loss": 0.2667578935623169, + "step": 2815, + "token_acc": 0.9255435947742141 + }, + { + "epoch": 0.18279047156052503, + "grad_norm": 0.9323372840881348, + "learning_rate": 9.525886076789195e-06, + "loss": 0.25858187675476074, + "step": 2820, + "token_acc": 0.9263095900271571 + }, + { + "epoch": 0.1831145681413061, + "grad_norm": 0.8880383968353271, + "learning_rate": 9.523605767567466e-06, + "loss": 0.2622524261474609, + "step": 2825, + "token_acc": 0.9265344466269437 + }, + { + "epoch": 0.18343866472208717, + "grad_norm": 1.5275278091430664, + "learning_rate": 9.521320262053723e-06, + "loss": 0.26889405250549314, + "step": 2830, + "token_acc": 0.9249789062911986 + }, + { + "epoch": 0.18376276130286825, + "grad_norm": 0.8432549238204956, + "learning_rate": 9.519029562873339e-06, + "loss": 0.2683408737182617, + "step": 2835, + "token_acc": 0.9256379100850547 + }, + { + "epoch": 0.18408685788364934, + "grad_norm": 0.9999023675918579, + "learning_rate": 9.516733672657654e-06, + "loss": 0.27060914039611816, + "step": 2840, + "token_acc": 0.9239678439260384 + }, + { + "epoch": 0.1844109544644304, + "grad_norm": 0.9052605032920837, + "learning_rate": 9.514432594043971e-06, + "loss": 0.27492902278900144, + "step": 2845, + "token_acc": 0.923604527235993 + }, + { + "epoch": 0.18473505104521148, + "grad_norm": 0.9163946509361267, + "learning_rate": 9.512126329675551e-06, + "loss": 0.26916625499725344, + "step": 2850, + "token_acc": 0.9246265834124072 + }, + { + "epoch": 0.18505914762599254, + "grad_norm": 0.9483189582824707, + "learning_rate": 9.509814882201613e-06, + "loss": 0.2590158700942993, + "step": 2855, + "token_acc": 0.9271527497033163 + }, + { + "epoch": 0.18538324420677363, + "grad_norm": 1.0431078672409058, + "learning_rate": 9.50749825427733e-06, + "loss": 0.2773310899734497, + "step": 2860, + "token_acc": 0.9225722991253185 + }, + { + "epoch": 0.1857073407875547, + "grad_norm": 1.0394572019577026, + "learning_rate": 9.505176448563827e-06, + "loss": 0.2685012102127075, + "step": 2865, + "token_acc": 0.9247046179857011 + }, + { + "epoch": 0.18603143736833577, + "grad_norm": 0.8381138443946838, + "learning_rate": 9.502849467728177e-06, + "loss": 0.2576919078826904, + "step": 2870, + "token_acc": 0.9272901860367893 + }, + { + "epoch": 0.18635553394911683, + "grad_norm": 0.926325261592865, + "learning_rate": 9.500517314443392e-06, + "loss": 0.2674917697906494, + "step": 2875, + "token_acc": 0.9249887355766405 + }, + { + "epoch": 0.18667963052989792, + "grad_norm": 0.9320431351661682, + "learning_rate": 9.498179991388431e-06, + "loss": 0.26810216903686523, + "step": 2880, + "token_acc": 0.9255161955126959 + }, + { + "epoch": 0.18700372711067897, + "grad_norm": 0.9325308799743652, + "learning_rate": 9.495837501248195e-06, + "loss": 0.2625418186187744, + "step": 2885, + "token_acc": 0.9267093130629124 + }, + { + "epoch": 0.18732782369146006, + "grad_norm": 0.9615559577941895, + "learning_rate": 9.49348984671351e-06, + "loss": 0.26489830017089844, + "step": 2890, + "token_acc": 0.9267306621236444 + }, + { + "epoch": 0.18765192027224112, + "grad_norm": 0.8754135966300964, + "learning_rate": 9.491137030481144e-06, + "loss": 0.2675673484802246, + "step": 2895, + "token_acc": 0.9248358240509507 + }, + { + "epoch": 0.1879760168530222, + "grad_norm": 0.9541027545928955, + "learning_rate": 9.48877905525379e-06, + "loss": 0.2661820650100708, + "step": 2900, + "token_acc": 0.9262947692228153 + }, + { + "epoch": 0.18830011343380326, + "grad_norm": 0.8665955662727356, + "learning_rate": 9.486415923740067e-06, + "loss": 0.2627674102783203, + "step": 2905, + "token_acc": 0.9260402805947783 + }, + { + "epoch": 0.18862421001458435, + "grad_norm": 0.9961279630661011, + "learning_rate": 9.484047638654517e-06, + "loss": 0.2679285526275635, + "step": 2910, + "token_acc": 0.9246748672727985 + }, + { + "epoch": 0.1889483065953654, + "grad_norm": 0.9711815118789673, + "learning_rate": 9.481674202717603e-06, + "loss": 0.2702517032623291, + "step": 2915, + "token_acc": 0.9248402361411845 + }, + { + "epoch": 0.1892724031761465, + "grad_norm": 0.908807098865509, + "learning_rate": 9.479295618655708e-06, + "loss": 0.26656417846679686, + "step": 2920, + "token_acc": 0.9263028808635377 + }, + { + "epoch": 0.18959649975692758, + "grad_norm": 0.868600070476532, + "learning_rate": 9.47691188920112e-06, + "loss": 0.26037304401397704, + "step": 2925, + "token_acc": 0.9269948279173261 + }, + { + "epoch": 0.18992059633770864, + "grad_norm": 0.8724709749221802, + "learning_rate": 9.474523017092046e-06, + "loss": 0.2684026718139648, + "step": 2930, + "token_acc": 0.92422396474629 + }, + { + "epoch": 0.19024469291848972, + "grad_norm": 0.9151384830474854, + "learning_rate": 9.472129005072597e-06, + "loss": 0.2716273307800293, + "step": 2935, + "token_acc": 0.9240155150877655 + }, + { + "epoch": 0.19056878949927078, + "grad_norm": 0.8546174764633179, + "learning_rate": 9.469729855892788e-06, + "loss": 0.2660079479217529, + "step": 2940, + "token_acc": 0.9248888655075361 + }, + { + "epoch": 0.19089288608005187, + "grad_norm": 0.978075385093689, + "learning_rate": 9.467325572308534e-06, + "loss": 0.2749605655670166, + "step": 2945, + "token_acc": 0.9229108491053883 + }, + { + "epoch": 0.19121698266083292, + "grad_norm": 0.9070817232131958, + "learning_rate": 9.464916157081652e-06, + "loss": 0.2724959373474121, + "step": 2950, + "token_acc": 0.9241896370279827 + }, + { + "epoch": 0.191541079241614, + "grad_norm": 0.8709344267845154, + "learning_rate": 9.462501612979849e-06, + "loss": 0.2549907207489014, + "step": 2955, + "token_acc": 0.9278869300831668 + }, + { + "epoch": 0.19186517582239507, + "grad_norm": 0.8853091597557068, + "learning_rate": 9.460081942776726e-06, + "loss": 0.2638587474822998, + "step": 2960, + "token_acc": 0.9252458973193404 + }, + { + "epoch": 0.19218927240317615, + "grad_norm": 0.9735586643218994, + "learning_rate": 9.457657149251772e-06, + "loss": 0.2752849102020264, + "step": 2965, + "token_acc": 0.9244127717690311 + }, + { + "epoch": 0.1925133689839572, + "grad_norm": 0.9884077310562134, + "learning_rate": 9.455227235190366e-06, + "loss": 0.2634574413299561, + "step": 2970, + "token_acc": 0.9255827412146739 + }, + { + "epoch": 0.1928374655647383, + "grad_norm": 0.9071623086929321, + "learning_rate": 9.452792203383756e-06, + "loss": 0.2632092237472534, + "step": 2975, + "token_acc": 0.9264309642906311 + }, + { + "epoch": 0.19316156214551936, + "grad_norm": 0.9388969540596008, + "learning_rate": 9.450352056629083e-06, + "loss": 0.2657320022583008, + "step": 2980, + "token_acc": 0.9249191956939337 + }, + { + "epoch": 0.19348565872630044, + "grad_norm": 0.9680787324905396, + "learning_rate": 9.447906797729356e-06, + "loss": 0.26667883396148684, + "step": 2985, + "token_acc": 0.9254906822571384 + }, + { + "epoch": 0.1938097553070815, + "grad_norm": 0.8580886125564575, + "learning_rate": 9.445456429493458e-06, + "loss": 0.26700196266174314, + "step": 2990, + "token_acc": 0.9261833244988907 + }, + { + "epoch": 0.19413385188786259, + "grad_norm": 0.935131311416626, + "learning_rate": 9.443000954736142e-06, + "loss": 0.27133774757385254, + "step": 2995, + "token_acc": 0.9240634024764944 + }, + { + "epoch": 0.19445794846864364, + "grad_norm": 0.8756237626075745, + "learning_rate": 9.440540376278022e-06, + "loss": 0.270628023147583, + "step": 3000, + "token_acc": 0.9241394927536232 + }, + { + "epoch": 0.19445794846864364, + "eval_loss": 0.2669617533683777, + "eval_runtime": 224.4063, + "eval_samples_per_second": 89.124, + "eval_steps_per_second": 0.12, + "eval_token_acc": 0.9251011661039676, + "step": 3000 + }, + { + "epoch": 0.19478204504942473, + "grad_norm": 0.9781657457351685, + "learning_rate": 9.438074696945584e-06, + "loss": 0.2703601598739624, + "step": 3005, + "token_acc": 0.9239762998503912 + }, + { + "epoch": 0.19510614163020581, + "grad_norm": 0.82326740026474, + "learning_rate": 9.435603919571166e-06, + "loss": 0.2636559247970581, + "step": 3010, + "token_acc": 0.9260278821007925 + }, + { + "epoch": 0.19543023821098687, + "grad_norm": 1.0966346263885498, + "learning_rate": 9.433128046992964e-06, + "loss": 0.26722183227539065, + "step": 3015, + "token_acc": 0.9245521642691295 + }, + { + "epoch": 0.19575433479176796, + "grad_norm": 1.1072540283203125, + "learning_rate": 9.430647082055025e-06, + "loss": 0.2628938674926758, + "step": 3020, + "token_acc": 0.9258546980781476 + }, + { + "epoch": 0.19607843137254902, + "grad_norm": 0.9641943573951721, + "learning_rate": 9.428161027607252e-06, + "loss": 0.26081194877624514, + "step": 3025, + "token_acc": 0.9273439516155714 + }, + { + "epoch": 0.1964025279533301, + "grad_norm": 0.9969142079353333, + "learning_rate": 9.425669886505386e-06, + "loss": 0.2594396352767944, + "step": 3030, + "token_acc": 0.9267347910294952 + }, + { + "epoch": 0.19672662453411116, + "grad_norm": 1.1544158458709717, + "learning_rate": 9.423173661611019e-06, + "loss": 0.2689204692840576, + "step": 3035, + "token_acc": 0.924153076053609 + }, + { + "epoch": 0.19705072111489225, + "grad_norm": 0.9067439436912537, + "learning_rate": 9.420672355791576e-06, + "loss": 0.27121949195861816, + "step": 3040, + "token_acc": 0.9238504357024452 + }, + { + "epoch": 0.1973748176956733, + "grad_norm": 1.1141999959945679, + "learning_rate": 9.418165971920324e-06, + "loss": 0.2658883571624756, + "step": 3045, + "token_acc": 0.9256831102145608 + }, + { + "epoch": 0.1976989142764544, + "grad_norm": 1.1477572917938232, + "learning_rate": 9.415654512876357e-06, + "loss": 0.2585632801055908, + "step": 3050, + "token_acc": 0.9263922646797094 + }, + { + "epoch": 0.19802301085723545, + "grad_norm": 0.9336361289024353, + "learning_rate": 9.413137981544604e-06, + "loss": 0.26142230033874514, + "step": 3055, + "token_acc": 0.925754917351556 + }, + { + "epoch": 0.19834710743801653, + "grad_norm": 0.9960599541664124, + "learning_rate": 9.41061638081582e-06, + "loss": 0.26828603744506835, + "step": 3060, + "token_acc": 0.9253106796761359 + }, + { + "epoch": 0.1986712040187976, + "grad_norm": 0.8798666000366211, + "learning_rate": 9.408089713586583e-06, + "loss": 0.26261000633239745, + "step": 3065, + "token_acc": 0.9258832766291547 + }, + { + "epoch": 0.19899530059957868, + "grad_norm": 0.8934407234191895, + "learning_rate": 9.405557982759288e-06, + "loss": 0.2672741413116455, + "step": 3070, + "token_acc": 0.9248355197839584 + }, + { + "epoch": 0.19931939718035974, + "grad_norm": 0.8798461556434631, + "learning_rate": 9.40302119124215e-06, + "loss": 0.26403422355651857, + "step": 3075, + "token_acc": 0.925556557680204 + }, + { + "epoch": 0.19964349376114082, + "grad_norm": 0.8795861601829529, + "learning_rate": 9.400479341949197e-06, + "loss": 0.26785836219787595, + "step": 3080, + "token_acc": 0.9256736945917851 + }, + { + "epoch": 0.19996759034192188, + "grad_norm": 0.8664177656173706, + "learning_rate": 9.39793243780026e-06, + "loss": 0.2684819221496582, + "step": 3085, + "token_acc": 0.9248567085408949 + }, + { + "epoch": 0.20029168692270297, + "grad_norm": 0.910502552986145, + "learning_rate": 9.395380481720993e-06, + "loss": 0.26074411869049074, + "step": 3090, + "token_acc": 0.9265236640516235 + }, + { + "epoch": 0.20061578350348405, + "grad_norm": 0.8878445029258728, + "learning_rate": 9.392823476642835e-06, + "loss": 0.2634315490722656, + "step": 3095, + "token_acc": 0.9264346103644818 + }, + { + "epoch": 0.2009398800842651, + "grad_norm": 0.9140256643295288, + "learning_rate": 9.390261425503032e-06, + "loss": 0.26543502807617186, + "step": 3100, + "token_acc": 0.925520651809376 + }, + { + "epoch": 0.2012639766650462, + "grad_norm": 0.8995592594146729, + "learning_rate": 9.387694331244631e-06, + "loss": 0.2643453359603882, + "step": 3105, + "token_acc": 0.9259720738894822 + }, + { + "epoch": 0.20158807324582725, + "grad_norm": 0.8818653225898743, + "learning_rate": 9.385122196816467e-06, + "loss": 0.25981147289276124, + "step": 3110, + "token_acc": 0.926405707926949 + }, + { + "epoch": 0.20191216982660834, + "grad_norm": 0.8031456470489502, + "learning_rate": 9.382545025173164e-06, + "loss": 0.2631994724273682, + "step": 3115, + "token_acc": 0.9249338839745177 + }, + { + "epoch": 0.2022362664073894, + "grad_norm": 0.9274810552597046, + "learning_rate": 9.379962819275135e-06, + "loss": 0.26482396125793456, + "step": 3120, + "token_acc": 0.9262748458977487 + }, + { + "epoch": 0.20256036298817048, + "grad_norm": 0.8133262991905212, + "learning_rate": 9.377375582088572e-06, + "loss": 0.2686553716659546, + "step": 3125, + "token_acc": 0.9249135625285406 + }, + { + "epoch": 0.20288445956895154, + "grad_norm": 0.8277010917663574, + "learning_rate": 9.374783316585454e-06, + "loss": 0.2611072063446045, + "step": 3130, + "token_acc": 0.9269352368557701 + }, + { + "epoch": 0.20320855614973263, + "grad_norm": 0.8608559370040894, + "learning_rate": 9.37218602574353e-06, + "loss": 0.2676061153411865, + "step": 3135, + "token_acc": 0.9250135461958318 + }, + { + "epoch": 0.20353265273051369, + "grad_norm": 0.995265543460846, + "learning_rate": 9.369583712546322e-06, + "loss": 0.2630260944366455, + "step": 3140, + "token_acc": 0.9265028478437755 + }, + { + "epoch": 0.20385674931129477, + "grad_norm": 0.8645061254501343, + "learning_rate": 9.366976379983122e-06, + "loss": 0.2663288593292236, + "step": 3145, + "token_acc": 0.9254733991654895 + }, + { + "epoch": 0.20418084589207583, + "grad_norm": 0.8868167400360107, + "learning_rate": 9.364364031048987e-06, + "loss": 0.26913843154907224, + "step": 3150, + "token_acc": 0.9251655845215179 + }, + { + "epoch": 0.20450494247285692, + "grad_norm": 0.8741997480392456, + "learning_rate": 9.36174666874474e-06, + "loss": 0.27433693408966064, + "step": 3155, + "token_acc": 0.924311956592796 + }, + { + "epoch": 0.20482903905363797, + "grad_norm": 0.8934440016746521, + "learning_rate": 9.359124296076957e-06, + "loss": 0.2624192237854004, + "step": 3160, + "token_acc": 0.9260399015769758 + }, + { + "epoch": 0.20515313563441906, + "grad_norm": 0.8825971484184265, + "learning_rate": 9.356496916057976e-06, + "loss": 0.265681791305542, + "step": 3165, + "token_acc": 0.9262891902100999 + }, + { + "epoch": 0.20547723221520012, + "grad_norm": 0.9596103429794312, + "learning_rate": 9.353864531705881e-06, + "loss": 0.26413259506225584, + "step": 3170, + "token_acc": 0.925178682157893 + }, + { + "epoch": 0.2058013287959812, + "grad_norm": 0.8729656934738159, + "learning_rate": 9.351227146044508e-06, + "loss": 0.2690155029296875, + "step": 3175, + "token_acc": 0.9247205000999942 + }, + { + "epoch": 0.20612542537676226, + "grad_norm": 0.964819610118866, + "learning_rate": 9.348584762103437e-06, + "loss": 0.25860748291015623, + "step": 3180, + "token_acc": 0.9266335950589704 + }, + { + "epoch": 0.20644952195754335, + "grad_norm": 1.0884703397750854, + "learning_rate": 9.345937382917989e-06, + "loss": 0.27290592193603513, + "step": 3185, + "token_acc": 0.9236298292902066 + }, + { + "epoch": 0.20677361853832443, + "grad_norm": 0.9925956726074219, + "learning_rate": 9.343285011529224e-06, + "loss": 0.2644828796386719, + "step": 3190, + "token_acc": 0.9259972054520049 + }, + { + "epoch": 0.2070977151191055, + "grad_norm": 0.8774257302284241, + "learning_rate": 9.340627650983938e-06, + "loss": 0.2670236587524414, + "step": 3195, + "token_acc": 0.924974185368659 + }, + { + "epoch": 0.20742181169988658, + "grad_norm": 1.082173466682434, + "learning_rate": 9.337965304334654e-06, + "loss": 0.2710235595703125, + "step": 3200, + "token_acc": 0.9241783294690545 + }, + { + "epoch": 0.20774590828066763, + "grad_norm": 0.9013928174972534, + "learning_rate": 9.335297974639627e-06, + "loss": 0.2677942752838135, + "step": 3205, + "token_acc": 0.9253698157134075 + }, + { + "epoch": 0.20807000486144872, + "grad_norm": 0.8851301670074463, + "learning_rate": 9.332625664962834e-06, + "loss": 0.2658917665481567, + "step": 3210, + "token_acc": 0.9261253514139243 + }, + { + "epoch": 0.20839410144222978, + "grad_norm": 0.8417280912399292, + "learning_rate": 9.329948378373968e-06, + "loss": 0.2675912618637085, + "step": 3215, + "token_acc": 0.92596061274879 + }, + { + "epoch": 0.20871819802301086, + "grad_norm": 0.9322956204414368, + "learning_rate": 9.327266117948448e-06, + "loss": 0.2649038314819336, + "step": 3220, + "token_acc": 0.9258410292422455 + }, + { + "epoch": 0.20904229460379192, + "grad_norm": 0.8290584683418274, + "learning_rate": 9.324578886767404e-06, + "loss": 0.266896390914917, + "step": 3225, + "token_acc": 0.9251301283501988 + }, + { + "epoch": 0.209366391184573, + "grad_norm": 0.8853901028633118, + "learning_rate": 9.321886687917669e-06, + "loss": 0.26955952644348147, + "step": 3230, + "token_acc": 0.9233748486084935 + }, + { + "epoch": 0.20969048776535407, + "grad_norm": 0.8447184562683105, + "learning_rate": 9.31918952449179e-06, + "loss": 0.2642394542694092, + "step": 3235, + "token_acc": 0.926285911449348 + }, + { + "epoch": 0.21001458434613515, + "grad_norm": 0.8104745745658875, + "learning_rate": 9.316487399588012e-06, + "loss": 0.2651174306869507, + "step": 3240, + "token_acc": 0.9255133891900772 + }, + { + "epoch": 0.2103386809269162, + "grad_norm": 0.9052267074584961, + "learning_rate": 9.313780316310284e-06, + "loss": 0.26404538154602053, + "step": 3245, + "token_acc": 0.9255737662903495 + }, + { + "epoch": 0.2106627775076973, + "grad_norm": 0.8188114166259766, + "learning_rate": 9.311068277768247e-06, + "loss": 0.27112455368041993, + "step": 3250, + "token_acc": 0.9243555692418235 + }, + { + "epoch": 0.21098687408847835, + "grad_norm": 0.9233535528182983, + "learning_rate": 9.308351287077235e-06, + "loss": 0.26205155849456785, + "step": 3255, + "token_acc": 0.9250019597083954 + }, + { + "epoch": 0.21131097066925944, + "grad_norm": 0.9320750832557678, + "learning_rate": 9.305629347358271e-06, + "loss": 0.2659982919692993, + "step": 3260, + "token_acc": 0.9265380902661492 + }, + { + "epoch": 0.2116350672500405, + "grad_norm": 1.0764758586883545, + "learning_rate": 9.302902461738064e-06, + "loss": 0.2635190010070801, + "step": 3265, + "token_acc": 0.9256084746994799 + }, + { + "epoch": 0.21195916383082158, + "grad_norm": 1.054880976676941, + "learning_rate": 9.300170633349002e-06, + "loss": 0.26807527542114257, + "step": 3270, + "token_acc": 0.9259429140565617 + }, + { + "epoch": 0.21228326041160267, + "grad_norm": 0.8692432045936584, + "learning_rate": 9.297433865329152e-06, + "loss": 0.26184585094451907, + "step": 3275, + "token_acc": 0.9262432409348041 + }, + { + "epoch": 0.21260735699238373, + "grad_norm": 0.9357015490531921, + "learning_rate": 9.294692160822257e-06, + "loss": 0.26640684604644777, + "step": 3280, + "token_acc": 0.9247855204553219 + }, + { + "epoch": 0.21293145357316481, + "grad_norm": 0.8411440849304199, + "learning_rate": 9.291945522977725e-06, + "loss": 0.2638063669204712, + "step": 3285, + "token_acc": 0.9252875073930472 + }, + { + "epoch": 0.21325555015394587, + "grad_norm": 1.2368943691253662, + "learning_rate": 9.28919395495064e-06, + "loss": 0.258840274810791, + "step": 3290, + "token_acc": 0.9272132840058505 + }, + { + "epoch": 0.21357964673472696, + "grad_norm": 0.9098902344703674, + "learning_rate": 9.286437459901739e-06, + "loss": 0.2618029832839966, + "step": 3295, + "token_acc": 0.927083602532627 + }, + { + "epoch": 0.21390374331550802, + "grad_norm": 0.8564281463623047, + "learning_rate": 9.283676040997426e-06, + "loss": 0.26063683032989504, + "step": 3300, + "token_acc": 0.9267134518265888 + }, + { + "epoch": 0.2142278398962891, + "grad_norm": 0.9145544767379761, + "learning_rate": 9.28090970140976e-06, + "loss": 0.25838847160339357, + "step": 3305, + "token_acc": 0.9262876073808698 + }, + { + "epoch": 0.21455193647707016, + "grad_norm": 0.9129169583320618, + "learning_rate": 9.278138444316448e-06, + "loss": 0.26417269706726076, + "step": 3310, + "token_acc": 0.9264557779212395 + }, + { + "epoch": 0.21487603305785125, + "grad_norm": 0.8717644214630127, + "learning_rate": 9.275362272900851e-06, + "loss": 0.2664037704467773, + "step": 3315, + "token_acc": 0.9256631788610525 + }, + { + "epoch": 0.2152001296386323, + "grad_norm": 0.9037322998046875, + "learning_rate": 9.272581190351974e-06, + "loss": 0.27002625465393065, + "step": 3320, + "token_acc": 0.9244667503136763 + }, + { + "epoch": 0.2155242262194134, + "grad_norm": 1.2448707818984985, + "learning_rate": 9.269795199864461e-06, + "loss": 0.2606431722640991, + "step": 3325, + "token_acc": 0.9267611836864228 + }, + { + "epoch": 0.21584832280019445, + "grad_norm": 0.8325229287147522, + "learning_rate": 9.267004304638597e-06, + "loss": 0.2634316682815552, + "step": 3330, + "token_acc": 0.9260141964902763 + }, + { + "epoch": 0.21617241938097553, + "grad_norm": 0.9249715805053711, + "learning_rate": 9.264208507880297e-06, + "loss": 0.2697380781173706, + "step": 3335, + "token_acc": 0.9248956162053 + }, + { + "epoch": 0.2164965159617566, + "grad_norm": 0.8680678606033325, + "learning_rate": 9.261407812801108e-06, + "loss": 0.2574089527130127, + "step": 3340, + "token_acc": 0.9272097510223165 + }, + { + "epoch": 0.21682061254253768, + "grad_norm": 0.8961992859840393, + "learning_rate": 9.25860222261821e-06, + "loss": 0.2660964488983154, + "step": 3345, + "token_acc": 0.9243412554895376 + }, + { + "epoch": 0.21714470912331874, + "grad_norm": 0.8428463935852051, + "learning_rate": 9.255791740554396e-06, + "loss": 0.2641308784484863, + "step": 3350, + "token_acc": 0.9258106271438742 + }, + { + "epoch": 0.21746880570409982, + "grad_norm": 0.9367009401321411, + "learning_rate": 9.252976369838085e-06, + "loss": 0.2612863063812256, + "step": 3355, + "token_acc": 0.9269817817236785 + }, + { + "epoch": 0.2177929022848809, + "grad_norm": 0.9557777643203735, + "learning_rate": 9.250156113703309e-06, + "loss": 0.2637699365615845, + "step": 3360, + "token_acc": 0.9258202381030296 + }, + { + "epoch": 0.21811699886566197, + "grad_norm": 0.9488732814788818, + "learning_rate": 9.247330975389712e-06, + "loss": 0.2667109489440918, + "step": 3365, + "token_acc": 0.9248041741141836 + }, + { + "epoch": 0.21844109544644305, + "grad_norm": 0.9357844591140747, + "learning_rate": 9.244500958142546e-06, + "loss": 0.2645890474319458, + "step": 3370, + "token_acc": 0.9257691799815133 + }, + { + "epoch": 0.2187651920272241, + "grad_norm": 0.9841048121452332, + "learning_rate": 9.24166606521267e-06, + "loss": 0.2663395643234253, + "step": 3375, + "token_acc": 0.9243055148280648 + }, + { + "epoch": 0.2190892886080052, + "grad_norm": 0.8774002194404602, + "learning_rate": 9.238826299856541e-06, + "loss": 0.2606687545776367, + "step": 3380, + "token_acc": 0.9270253678436343 + }, + { + "epoch": 0.21941338518878625, + "grad_norm": 0.8875116109848022, + "learning_rate": 9.235981665336214e-06, + "loss": 0.27065410614013674, + "step": 3385, + "token_acc": 0.9239730829461986 + }, + { + "epoch": 0.21973748176956734, + "grad_norm": 0.9407257437705994, + "learning_rate": 9.233132164919335e-06, + "loss": 0.2632985830307007, + "step": 3390, + "token_acc": 0.9244549591944613 + }, + { + "epoch": 0.2200615783503484, + "grad_norm": 0.8755924105644226, + "learning_rate": 9.230277801879145e-06, + "loss": 0.25905284881591795, + "step": 3395, + "token_acc": 0.9266411768517918 + }, + { + "epoch": 0.22038567493112948, + "grad_norm": 0.9250627160072327, + "learning_rate": 9.227418579494465e-06, + "loss": 0.26375718116760255, + "step": 3400, + "token_acc": 0.925414833479118 + }, + { + "epoch": 0.22070977151191054, + "grad_norm": 0.8629029989242554, + "learning_rate": 9.224554501049701e-06, + "loss": 0.2573062896728516, + "step": 3405, + "token_acc": 0.9272287545715522 + }, + { + "epoch": 0.22103386809269163, + "grad_norm": 0.8667880296707153, + "learning_rate": 9.221685569834836e-06, + "loss": 0.2670911312103271, + "step": 3410, + "token_acc": 0.9251760505608022 + }, + { + "epoch": 0.22135796467347268, + "grad_norm": 0.9667277336120605, + "learning_rate": 9.21881178914543e-06, + "loss": 0.24934451580047606, + "step": 3415, + "token_acc": 0.9284775023435059 + }, + { + "epoch": 0.22168206125425377, + "grad_norm": 0.949001133441925, + "learning_rate": 9.215933162282607e-06, + "loss": 0.2649359941482544, + "step": 3420, + "token_acc": 0.926513788478524 + }, + { + "epoch": 0.22200615783503483, + "grad_norm": 0.932003915309906, + "learning_rate": 9.213049692553066e-06, + "loss": 0.25911531448364256, + "step": 3425, + "token_acc": 0.92675016411343 + }, + { + "epoch": 0.22233025441581591, + "grad_norm": 0.8812533617019653, + "learning_rate": 9.210161383269063e-06, + "loss": 0.26271121501922606, + "step": 3430, + "token_acc": 0.926700607653145 + }, + { + "epoch": 0.22265435099659697, + "grad_norm": 0.8192481994628906, + "learning_rate": 9.20726823774842e-06, + "loss": 0.2638916015625, + "step": 3435, + "token_acc": 0.9267796654080066 + }, + { + "epoch": 0.22297844757737806, + "grad_norm": 0.8288140892982483, + "learning_rate": 9.204370259314503e-06, + "loss": 0.2629103183746338, + "step": 3440, + "token_acc": 0.9257631082166816 + }, + { + "epoch": 0.22330254415815914, + "grad_norm": 0.8430953621864319, + "learning_rate": 9.20146745129624e-06, + "loss": 0.26664912700653076, + "step": 3445, + "token_acc": 0.9247475697304168 + }, + { + "epoch": 0.2236266407389402, + "grad_norm": 0.8576011061668396, + "learning_rate": 9.198559817028106e-06, + "loss": 0.25874958038330076, + "step": 3450, + "token_acc": 0.9274104978587072 + }, + { + "epoch": 0.2239507373197213, + "grad_norm": 0.8742114305496216, + "learning_rate": 9.195647359850112e-06, + "loss": 0.26214098930358887, + "step": 3455, + "token_acc": 0.9264212524706127 + }, + { + "epoch": 0.22427483390050235, + "grad_norm": 0.8295634388923645, + "learning_rate": 9.19273008310782e-06, + "loss": 0.2547857046127319, + "step": 3460, + "token_acc": 0.9279300536133286 + }, + { + "epoch": 0.22459893048128343, + "grad_norm": 0.9069639444351196, + "learning_rate": 9.189807990152317e-06, + "loss": 0.2580267906188965, + "step": 3465, + "token_acc": 0.9262434216131548 + }, + { + "epoch": 0.2249230270620645, + "grad_norm": 0.8336062431335449, + "learning_rate": 9.18688108434023e-06, + "loss": 0.2631111145019531, + "step": 3470, + "token_acc": 0.9266328138679741 + }, + { + "epoch": 0.22524712364284558, + "grad_norm": 0.8848274350166321, + "learning_rate": 9.183949369033716e-06, + "loss": 0.258338475227356, + "step": 3475, + "token_acc": 0.927288605637183 + }, + { + "epoch": 0.22557122022362663, + "grad_norm": 0.8938132524490356, + "learning_rate": 9.181012847600448e-06, + "loss": 0.2634269237518311, + "step": 3480, + "token_acc": 0.9262855276053578 + }, + { + "epoch": 0.22589531680440772, + "grad_norm": 0.9038134217262268, + "learning_rate": 9.17807152341363e-06, + "loss": 0.25702710151672364, + "step": 3485, + "token_acc": 0.9270220671677117 + }, + { + "epoch": 0.22621941338518878, + "grad_norm": 0.9028303623199463, + "learning_rate": 9.175125399851974e-06, + "loss": 0.2674188375473022, + "step": 3490, + "token_acc": 0.9244738583719391 + }, + { + "epoch": 0.22654350996596986, + "grad_norm": 0.8527891039848328, + "learning_rate": 9.17217448029971e-06, + "loss": 0.2608349323272705, + "step": 3495, + "token_acc": 0.9266560286448198 + }, + { + "epoch": 0.22686760654675092, + "grad_norm": 1.0615705251693726, + "learning_rate": 9.169218768146578e-06, + "loss": 0.2727582693099976, + "step": 3500, + "token_acc": 0.9247851802254212 + }, + { + "epoch": 0.22686760654675092, + "eval_loss": 0.26445960998535156, + "eval_runtime": 217.568, + "eval_samples_per_second": 91.925, + "eval_steps_per_second": 0.124, + "eval_token_acc": 0.9255944256878098, + "step": 3500 + }, + { + "epoch": 0.227191703127532, + "grad_norm": 2.071399688720703, + "learning_rate": 9.16625826678782e-06, + "loss": 0.2597506046295166, + "step": 3505, + "token_acc": 0.92645379789868 + }, + { + "epoch": 0.22751579970831307, + "grad_norm": 0.9225764274597168, + "learning_rate": 9.16329297962418e-06, + "loss": 0.26715450286865233, + "step": 3510, + "token_acc": 0.9264514579622551 + }, + { + "epoch": 0.22783989628909415, + "grad_norm": 0.8650937080383301, + "learning_rate": 9.160322910061905e-06, + "loss": 0.2705091953277588, + "step": 3515, + "token_acc": 0.9246110297957947 + }, + { + "epoch": 0.2281639928698752, + "grad_norm": 0.8749118447303772, + "learning_rate": 9.157348061512728e-06, + "loss": 0.25983762741088867, + "step": 3520, + "token_acc": 0.9268422095647226 + }, + { + "epoch": 0.2284880894506563, + "grad_norm": 0.9360148310661316, + "learning_rate": 9.154368437393873e-06, + "loss": 0.2631370067596436, + "step": 3525, + "token_acc": 0.9269716338281127 + }, + { + "epoch": 0.22881218603143738, + "grad_norm": 0.9098191261291504, + "learning_rate": 9.151384041128054e-06, + "loss": 0.2653895378112793, + "step": 3530, + "token_acc": 0.9262201893946862 + }, + { + "epoch": 0.22913628261221844, + "grad_norm": 0.8991694450378418, + "learning_rate": 9.148394876143461e-06, + "loss": 0.26526656150817873, + "step": 3535, + "token_acc": 0.9253257982628443 + }, + { + "epoch": 0.22946037919299953, + "grad_norm": 0.8953236937522888, + "learning_rate": 9.14540094587377e-06, + "loss": 0.263206934928894, + "step": 3540, + "token_acc": 0.9255694241085205 + }, + { + "epoch": 0.22978447577378058, + "grad_norm": 0.8852925300598145, + "learning_rate": 9.142402253758123e-06, + "loss": 0.2582906007766724, + "step": 3545, + "token_acc": 0.9274711875839079 + }, + { + "epoch": 0.23010857235456167, + "grad_norm": 0.8822129964828491, + "learning_rate": 9.139398803241134e-06, + "loss": 0.2554760932922363, + "step": 3550, + "token_acc": 0.9273237356001607 + }, + { + "epoch": 0.23043266893534273, + "grad_norm": 1.7858068943023682, + "learning_rate": 9.136390597772886e-06, + "loss": 0.2673048973083496, + "step": 3555, + "token_acc": 0.9252214880632657 + }, + { + "epoch": 0.2307567655161238, + "grad_norm": 0.937673032283783, + "learning_rate": 9.13337764080892e-06, + "loss": 0.26005423069000244, + "step": 3560, + "token_acc": 0.926369818399729 + }, + { + "epoch": 0.23108086209690487, + "grad_norm": 1.00150728225708, + "learning_rate": 9.13035993581024e-06, + "loss": 0.2589425563812256, + "step": 3565, + "token_acc": 0.9277278678748753 + }, + { + "epoch": 0.23140495867768596, + "grad_norm": 0.8138728737831116, + "learning_rate": 9.127337486243298e-06, + "loss": 0.2727918863296509, + "step": 3570, + "token_acc": 0.9244027839556241 + }, + { + "epoch": 0.23172905525846701, + "grad_norm": 0.9878820180892944, + "learning_rate": 9.12431029558e-06, + "loss": 0.25275774002075196, + "step": 3575, + "token_acc": 0.9281122532223853 + }, + { + "epoch": 0.2320531518392481, + "grad_norm": 0.9843800067901611, + "learning_rate": 9.121278367297698e-06, + "loss": 0.26565070152282716, + "step": 3580, + "token_acc": 0.9257811735382969 + }, + { + "epoch": 0.23237724842002916, + "grad_norm": 0.8798107504844666, + "learning_rate": 9.118241704879185e-06, + "loss": 0.26822733879089355, + "step": 3585, + "token_acc": 0.9240915911018331 + }, + { + "epoch": 0.23270134500081024, + "grad_norm": 0.8808806538581848, + "learning_rate": 9.115200311812695e-06, + "loss": 0.26865925788879397, + "step": 3590, + "token_acc": 0.9246415406210235 + }, + { + "epoch": 0.2330254415815913, + "grad_norm": 0.8680911064147949, + "learning_rate": 9.112154191591888e-06, + "loss": 0.2660682678222656, + "step": 3595, + "token_acc": 0.9253614794955555 + }, + { + "epoch": 0.2333495381623724, + "grad_norm": 0.8418766856193542, + "learning_rate": 9.109103347715867e-06, + "loss": 0.26566033363342284, + "step": 3600, + "token_acc": 0.9256912254623925 + }, + { + "epoch": 0.23367363474315345, + "grad_norm": 0.8841730356216431, + "learning_rate": 9.10604778368915e-06, + "loss": 0.261461067199707, + "step": 3605, + "token_acc": 0.926539747844256 + }, + { + "epoch": 0.23399773132393453, + "grad_norm": 0.9045286178588867, + "learning_rate": 9.102987503021678e-06, + "loss": 0.2669029235839844, + "step": 3610, + "token_acc": 0.9253856832802143 + }, + { + "epoch": 0.23432182790471562, + "grad_norm": 0.8939406275749207, + "learning_rate": 9.09992250922882e-06, + "loss": 0.2606555461883545, + "step": 3615, + "token_acc": 0.9257858294848793 + }, + { + "epoch": 0.23464592448549668, + "grad_norm": 0.9183914661407471, + "learning_rate": 9.096852805831348e-06, + "loss": 0.26268754005432127, + "step": 3620, + "token_acc": 0.9264376253457248 + }, + { + "epoch": 0.23497002106627776, + "grad_norm": 1.1463788747787476, + "learning_rate": 9.093778396355448e-06, + "loss": 0.26030635833740234, + "step": 3625, + "token_acc": 0.9261901969217136 + }, + { + "epoch": 0.23529411764705882, + "grad_norm": 0.8447176218032837, + "learning_rate": 9.09069928433271e-06, + "loss": 0.2567586898803711, + "step": 3630, + "token_acc": 0.9281518507628476 + }, + { + "epoch": 0.2356182142278399, + "grad_norm": 0.7754336595535278, + "learning_rate": 9.087615473300134e-06, + "loss": 0.2564385890960693, + "step": 3635, + "token_acc": 0.9276885326353672 + }, + { + "epoch": 0.23594231080862096, + "grad_norm": 0.8199570178985596, + "learning_rate": 9.084526966800106e-06, + "loss": 0.26219370365142824, + "step": 3640, + "token_acc": 0.9275552058611083 + }, + { + "epoch": 0.23626640738940205, + "grad_norm": 0.8361279964447021, + "learning_rate": 9.081433768380415e-06, + "loss": 0.25876152515411377, + "step": 3645, + "token_acc": 0.9266012246088056 + }, + { + "epoch": 0.2365905039701831, + "grad_norm": 0.8484302759170532, + "learning_rate": 9.078335881594231e-06, + "loss": 0.27129244804382324, + "step": 3650, + "token_acc": 0.9239092154177265 + }, + { + "epoch": 0.2369146005509642, + "grad_norm": 0.8304858803749084, + "learning_rate": 9.075233310000116e-06, + "loss": 0.25465760231018064, + "step": 3655, + "token_acc": 0.9280951075963408 + }, + { + "epoch": 0.23723869713174525, + "grad_norm": 0.9628082513809204, + "learning_rate": 9.072126057162017e-06, + "loss": 0.25599141120910646, + "step": 3660, + "token_acc": 0.9284275896116719 + }, + { + "epoch": 0.23756279371252634, + "grad_norm": 0.944200336933136, + "learning_rate": 9.069014126649245e-06, + "loss": 0.2635692596435547, + "step": 3665, + "token_acc": 0.9263506659464872 + }, + { + "epoch": 0.2378868902933074, + "grad_norm": 0.863064706325531, + "learning_rate": 9.0658975220365e-06, + "loss": 0.26157259941101074, + "step": 3670, + "token_acc": 0.9262945979196958 + }, + { + "epoch": 0.23821098687408848, + "grad_norm": 0.9632282257080078, + "learning_rate": 9.062776246903838e-06, + "loss": 0.2579564809799194, + "step": 3675, + "token_acc": 0.9273096764108175 + }, + { + "epoch": 0.23853508345486954, + "grad_norm": 0.8893093466758728, + "learning_rate": 9.059650304836689e-06, + "loss": 0.26536922454833983, + "step": 3680, + "token_acc": 0.9253779620776431 + }, + { + "epoch": 0.23885918003565063, + "grad_norm": 0.9201160669326782, + "learning_rate": 9.05651969942584e-06, + "loss": 0.25775980949401855, + "step": 3685, + "token_acc": 0.9275294454308366 + }, + { + "epoch": 0.23918327661643168, + "grad_norm": 0.8697929978370667, + "learning_rate": 9.053384434267436e-06, + "loss": 0.2649453401565552, + "step": 3690, + "token_acc": 0.9264027140396324 + }, + { + "epoch": 0.23950737319721277, + "grad_norm": 0.8250270485877991, + "learning_rate": 9.05024451296297e-06, + "loss": 0.26184511184692383, + "step": 3695, + "token_acc": 0.9265153792860037 + }, + { + "epoch": 0.23983146977799386, + "grad_norm": 0.8657457232475281, + "learning_rate": 9.047099939119295e-06, + "loss": 0.25553097724914553, + "step": 3700, + "token_acc": 0.9279016543516642 + }, + { + "epoch": 0.2401555663587749, + "grad_norm": 0.840779721736908, + "learning_rate": 9.043950716348596e-06, + "loss": 0.2610261678695679, + "step": 3705, + "token_acc": 0.9268059466893707 + }, + { + "epoch": 0.240479662939556, + "grad_norm": 0.8344523906707764, + "learning_rate": 9.040796848268405e-06, + "loss": 0.26270155906677245, + "step": 3710, + "token_acc": 0.9263935128720704 + }, + { + "epoch": 0.24080375952033706, + "grad_norm": 0.9003820419311523, + "learning_rate": 9.037638338501588e-06, + "loss": 0.25495471954345705, + "step": 3715, + "token_acc": 0.9280774094957053 + }, + { + "epoch": 0.24112785610111814, + "grad_norm": 0.8304955959320068, + "learning_rate": 9.034475190676342e-06, + "loss": 0.26463875770568845, + "step": 3720, + "token_acc": 0.9264853186683688 + }, + { + "epoch": 0.2414519526818992, + "grad_norm": 0.8362470865249634, + "learning_rate": 9.031307408426196e-06, + "loss": 0.25963654518127444, + "step": 3725, + "token_acc": 0.926955288176843 + }, + { + "epoch": 0.2417760492626803, + "grad_norm": 0.9689456224441528, + "learning_rate": 9.028134995389996e-06, + "loss": 0.2558337450027466, + "step": 3730, + "token_acc": 0.927906577592353 + }, + { + "epoch": 0.24210014584346135, + "grad_norm": 0.8339043855667114, + "learning_rate": 9.024957955211916e-06, + "loss": 0.26160173416137694, + "step": 3735, + "token_acc": 0.9282867209137841 + }, + { + "epoch": 0.24242424242424243, + "grad_norm": 0.8988233804702759, + "learning_rate": 9.021776291541437e-06, + "loss": 0.26324822902679446, + "step": 3740, + "token_acc": 0.9257408255380388 + }, + { + "epoch": 0.2427483390050235, + "grad_norm": 0.9286147356033325, + "learning_rate": 9.018590008033355e-06, + "loss": 0.25923609733581543, + "step": 3745, + "token_acc": 0.9268346928496203 + }, + { + "epoch": 0.24307243558580457, + "grad_norm": 0.9040605425834656, + "learning_rate": 9.015399108347771e-06, + "loss": 0.2699911594390869, + "step": 3750, + "token_acc": 0.9250075359431724 + }, + { + "epoch": 0.24339653216658563, + "grad_norm": 0.8517218828201294, + "learning_rate": 9.012203596150092e-06, + "loss": 0.26553008556365965, + "step": 3755, + "token_acc": 0.9259371460182819 + }, + { + "epoch": 0.24372062874736672, + "grad_norm": 0.9971320033073425, + "learning_rate": 9.009003475111022e-06, + "loss": 0.26141085624694826, + "step": 3760, + "token_acc": 0.9255360778929635 + }, + { + "epoch": 0.24404472532814778, + "grad_norm": 0.8291534185409546, + "learning_rate": 9.005798748906556e-06, + "loss": 0.26013109683990476, + "step": 3765, + "token_acc": 0.9276141175850736 + }, + { + "epoch": 0.24436882190892886, + "grad_norm": 0.8724426031112671, + "learning_rate": 9.002589421217983e-06, + "loss": 0.26230764389038086, + "step": 3770, + "token_acc": 0.9261460378761686 + }, + { + "epoch": 0.24469291848970992, + "grad_norm": 0.8928936719894409, + "learning_rate": 8.999375495731874e-06, + "loss": 0.2632465839385986, + "step": 3775, + "token_acc": 0.9262064488206015 + }, + { + "epoch": 0.245017015070491, + "grad_norm": 0.8393530249595642, + "learning_rate": 8.996156976140088e-06, + "loss": 0.2712278127670288, + "step": 3780, + "token_acc": 0.9246547023786452 + }, + { + "epoch": 0.2453411116512721, + "grad_norm": 0.9236770272254944, + "learning_rate": 8.99293386613975e-06, + "loss": 0.2687112808227539, + "step": 3785, + "token_acc": 0.9243318851603088 + }, + { + "epoch": 0.24566520823205315, + "grad_norm": 0.8596184253692627, + "learning_rate": 8.989706169433274e-06, + "loss": 0.26513023376464845, + "step": 3790, + "token_acc": 0.9252471871803614 + }, + { + "epoch": 0.24598930481283424, + "grad_norm": 0.8616563677787781, + "learning_rate": 8.986473889728326e-06, + "loss": 0.2634493350982666, + "step": 3795, + "token_acc": 0.9257477790043631 + }, + { + "epoch": 0.2463134013936153, + "grad_norm": 0.9413219690322876, + "learning_rate": 8.983237030737848e-06, + "loss": 0.25677313804626467, + "step": 3800, + "token_acc": 0.9281611625350445 + }, + { + "epoch": 0.24663749797439638, + "grad_norm": 0.9483920335769653, + "learning_rate": 8.97999559618004e-06, + "loss": 0.26177291870117186, + "step": 3805, + "token_acc": 0.9271578726638883 + }, + { + "epoch": 0.24696159455517744, + "grad_norm": 0.825879693031311, + "learning_rate": 8.976749589778352e-06, + "loss": 0.264525842666626, + "step": 3810, + "token_acc": 0.9265027987344853 + }, + { + "epoch": 0.24728569113595852, + "grad_norm": 0.8431276679039001, + "learning_rate": 8.973499015261496e-06, + "loss": 0.25930109024047854, + "step": 3815, + "token_acc": 0.9269418366811769 + }, + { + "epoch": 0.24760978771673958, + "grad_norm": 0.9299726486206055, + "learning_rate": 8.97024387636342e-06, + "loss": 0.2606808185577393, + "step": 3820, + "token_acc": 0.9262172954288098 + }, + { + "epoch": 0.24793388429752067, + "grad_norm": 0.856094479560852, + "learning_rate": 8.966984176823325e-06, + "loss": 0.264854884147644, + "step": 3825, + "token_acc": 0.9259519442167844 + }, + { + "epoch": 0.24825798087830173, + "grad_norm": 0.9173142910003662, + "learning_rate": 8.963719920385644e-06, + "loss": 0.2681146144866943, + "step": 3830, + "token_acc": 0.9248820785448102 + }, + { + "epoch": 0.2485820774590828, + "grad_norm": 0.9074921607971191, + "learning_rate": 8.960451110800049e-06, + "loss": 0.25864691734313966, + "step": 3835, + "token_acc": 0.926948873292425 + }, + { + "epoch": 0.24890617403986387, + "grad_norm": 0.8386721014976501, + "learning_rate": 8.95717775182144e-06, + "loss": 0.2642892122268677, + "step": 3840, + "token_acc": 0.9266356654782143 + }, + { + "epoch": 0.24923027062064496, + "grad_norm": 0.8855156302452087, + "learning_rate": 8.953899847209941e-06, + "loss": 0.26588640213012693, + "step": 3845, + "token_acc": 0.9253760551069131 + }, + { + "epoch": 0.24955436720142601, + "grad_norm": 0.8465850949287415, + "learning_rate": 8.950617400730903e-06, + "loss": 0.25530314445495605, + "step": 3850, + "token_acc": 0.9287901286482082 + }, + { + "epoch": 0.2498784637822071, + "grad_norm": 0.8979588150978088, + "learning_rate": 8.947330416154892e-06, + "loss": 0.2618697643280029, + "step": 3855, + "token_acc": 0.9264437729986637 + }, + { + "epoch": 0.25020256036298816, + "grad_norm": 0.8771248459815979, + "learning_rate": 8.944038897257681e-06, + "loss": 0.2614720344543457, + "step": 3860, + "token_acc": 0.9258256913481695 + }, + { + "epoch": 0.25052665694376924, + "grad_norm": 0.9567051529884338, + "learning_rate": 8.940742847820262e-06, + "loss": 0.2619285583496094, + "step": 3865, + "token_acc": 0.9265165096645914 + }, + { + "epoch": 0.25085075352455033, + "grad_norm": 0.8558542132377625, + "learning_rate": 8.937442271628823e-06, + "loss": 0.26223363876342776, + "step": 3870, + "token_acc": 0.9262437605368418 + }, + { + "epoch": 0.2511748501053314, + "grad_norm": 0.791343092918396, + "learning_rate": 8.934137172474757e-06, + "loss": 0.2604944705963135, + "step": 3875, + "token_acc": 0.9258543470947911 + }, + { + "epoch": 0.25149894668611245, + "grad_norm": 0.8936379551887512, + "learning_rate": 8.930827554154648e-06, + "loss": 0.2596839427947998, + "step": 3880, + "token_acc": 0.9271534975161467 + }, + { + "epoch": 0.25182304326689353, + "grad_norm": 0.8601925373077393, + "learning_rate": 8.927513420470275e-06, + "loss": 0.2696534156799316, + "step": 3885, + "token_acc": 0.9251128409759927 + }, + { + "epoch": 0.2521471398476746, + "grad_norm": 0.8659956455230713, + "learning_rate": 8.924194775228603e-06, + "loss": 0.2614442348480225, + "step": 3890, + "token_acc": 0.9262755571928531 + }, + { + "epoch": 0.2524712364284557, + "grad_norm": 0.9207909107208252, + "learning_rate": 8.920871622241777e-06, + "loss": 0.25987801551818845, + "step": 3895, + "token_acc": 0.9266914939195201 + }, + { + "epoch": 0.25279533300923673, + "grad_norm": 0.8467811346054077, + "learning_rate": 8.917543965327124e-06, + "loss": 0.2568974018096924, + "step": 3900, + "token_acc": 0.9276546468449586 + }, + { + "epoch": 0.2531194295900178, + "grad_norm": 0.8296903371810913, + "learning_rate": 8.914211808307143e-06, + "loss": 0.2594138145446777, + "step": 3905, + "token_acc": 0.9268131355052915 + }, + { + "epoch": 0.2534435261707989, + "grad_norm": 0.8565858006477356, + "learning_rate": 8.910875155009499e-06, + "loss": 0.26329946517944336, + "step": 3910, + "token_acc": 0.9262049565681691 + }, + { + "epoch": 0.25376762275158, + "grad_norm": 0.8322010636329651, + "learning_rate": 8.907534009267029e-06, + "loss": 0.259573769569397, + "step": 3915, + "token_acc": 0.9271110410550444 + }, + { + "epoch": 0.254091719332361, + "grad_norm": 0.8544418811798096, + "learning_rate": 8.904188374917722e-06, + "loss": 0.26168150901794435, + "step": 3920, + "token_acc": 0.9261133670285703 + }, + { + "epoch": 0.2544158159131421, + "grad_norm": 0.889323890209198, + "learning_rate": 8.90083825580473e-06, + "loss": 0.25737709999084474, + "step": 3925, + "token_acc": 0.9276443972632001 + }, + { + "epoch": 0.2547399124939232, + "grad_norm": 0.8525602221488953, + "learning_rate": 8.897483655776353e-06, + "loss": 0.2580659627914429, + "step": 3930, + "token_acc": 0.9270104176936476 + }, + { + "epoch": 0.2550640090747043, + "grad_norm": 0.8723629117012024, + "learning_rate": 8.894124578686043e-06, + "loss": 0.2544872283935547, + "step": 3935, + "token_acc": 0.9277833829059605 + }, + { + "epoch": 0.2553881056554853, + "grad_norm": 0.869596540927887, + "learning_rate": 8.890761028392385e-06, + "loss": 0.2566462516784668, + "step": 3940, + "token_acc": 0.9288142318629721 + }, + { + "epoch": 0.2557122022362664, + "grad_norm": 0.9088364243507385, + "learning_rate": 8.887393008759115e-06, + "loss": 0.25803382396698, + "step": 3945, + "token_acc": 0.9273799013588057 + }, + { + "epoch": 0.2560362988170475, + "grad_norm": 0.9087403416633606, + "learning_rate": 8.88402052365509e-06, + "loss": 0.26232974529266356, + "step": 3950, + "token_acc": 0.9256121867881549 + }, + { + "epoch": 0.25636039539782857, + "grad_norm": 0.9897594451904297, + "learning_rate": 8.880643576954309e-06, + "loss": 0.25968708992004397, + "step": 3955, + "token_acc": 0.9273781963590445 + }, + { + "epoch": 0.25668449197860965, + "grad_norm": 0.8380420804023743, + "learning_rate": 8.877262172535888e-06, + "loss": 0.2629442930221558, + "step": 3960, + "token_acc": 0.9261984697064335 + }, + { + "epoch": 0.2570085885593907, + "grad_norm": 1.008497714996338, + "learning_rate": 8.873876314284064e-06, + "loss": 0.2611828327178955, + "step": 3965, + "token_acc": 0.9270141810308393 + }, + { + "epoch": 0.25733268514017177, + "grad_norm": 0.8280889987945557, + "learning_rate": 8.870486006088194e-06, + "loss": 0.2556713581085205, + "step": 3970, + "token_acc": 0.9271267278536236 + }, + { + "epoch": 0.25765678172095285, + "grad_norm": 0.8611913919448853, + "learning_rate": 8.867091251842743e-06, + "loss": 0.2606687068939209, + "step": 3975, + "token_acc": 0.9278035532420504 + }, + { + "epoch": 0.25798087830173394, + "grad_norm": 0.9114230275154114, + "learning_rate": 8.863692055447288e-06, + "loss": 0.2592257738113403, + "step": 3980, + "token_acc": 0.9271485896453985 + }, + { + "epoch": 0.25830497488251497, + "grad_norm": 0.9067589044570923, + "learning_rate": 8.860288420806501e-06, + "loss": 0.2650330066680908, + "step": 3985, + "token_acc": 0.9259556919914889 + }, + { + "epoch": 0.25862907146329606, + "grad_norm": 0.8510159254074097, + "learning_rate": 8.85688035183016e-06, + "loss": 0.2607996940612793, + "step": 3990, + "token_acc": 0.9266827707775674 + }, + { + "epoch": 0.25895316804407714, + "grad_norm": 0.8950018882751465, + "learning_rate": 8.85346785243313e-06, + "loss": 0.2593094825744629, + "step": 3995, + "token_acc": 0.9269510684422422 + }, + { + "epoch": 0.25927726462485823, + "grad_norm": 0.8641705513000488, + "learning_rate": 8.850050926535373e-06, + "loss": 0.2626436948776245, + "step": 4000, + "token_acc": 0.9257002864660971 + }, + { + "epoch": 0.25927726462485823, + "eval_loss": 0.2615666389465332, + "eval_runtime": 108.0265, + "eval_samples_per_second": 185.14, + "eval_steps_per_second": 0.25, + "eval_token_acc": 0.9262576922155975, + "step": 4000 + }, + { + "epoch": 0.25960136120563926, + "grad_norm": 0.9181963801383972, + "learning_rate": 8.84662957806193e-06, + "loss": 0.25753178596496584, + "step": 4005, + "token_acc": 0.9272111584858569 + }, + { + "epoch": 0.25992545778642034, + "grad_norm": 1.0889618396759033, + "learning_rate": 8.843203810942922e-06, + "loss": 0.26236202716827395, + "step": 4010, + "token_acc": 0.9270046071021845 + }, + { + "epoch": 0.26024955436720143, + "grad_norm": 0.8697792887687683, + "learning_rate": 8.839773629113551e-06, + "loss": 0.2591404438018799, + "step": 4015, + "token_acc": 0.9274140352022989 + }, + { + "epoch": 0.2605736509479825, + "grad_norm": 1.0301299095153809, + "learning_rate": 8.836339036514081e-06, + "loss": 0.25630109310150145, + "step": 4020, + "token_acc": 0.9279003828413886 + }, + { + "epoch": 0.26089774752876355, + "grad_norm": 0.8997102975845337, + "learning_rate": 8.832900037089853e-06, + "loss": 0.26283869743347166, + "step": 4025, + "token_acc": 0.9259731636282353 + }, + { + "epoch": 0.26122184410954463, + "grad_norm": 0.9788554906845093, + "learning_rate": 8.829456634791265e-06, + "loss": 0.2563297986984253, + "step": 4030, + "token_acc": 0.9279093314171866 + }, + { + "epoch": 0.2615459406903257, + "grad_norm": 0.8830538988113403, + "learning_rate": 8.826008833573772e-06, + "loss": 0.2604154109954834, + "step": 4035, + "token_acc": 0.9272978205698438 + }, + { + "epoch": 0.2618700372711068, + "grad_norm": 0.9093402028083801, + "learning_rate": 8.82255663739788e-06, + "loss": 0.26268959045410156, + "step": 4040, + "token_acc": 0.9263486995277057 + }, + { + "epoch": 0.2621941338518879, + "grad_norm": 0.9278461337089539, + "learning_rate": 8.819100050229155e-06, + "loss": 0.2609283924102783, + "step": 4045, + "token_acc": 0.9262323783832855 + }, + { + "epoch": 0.2625182304326689, + "grad_norm": 0.8410829305648804, + "learning_rate": 8.81563907603819e-06, + "loss": 0.25548272132873534, + "step": 4050, + "token_acc": 0.9285036111602203 + }, + { + "epoch": 0.26284232701345, + "grad_norm": 0.8875502943992615, + "learning_rate": 8.812173718800632e-06, + "loss": 0.24896974563598634, + "step": 4055, + "token_acc": 0.9284992354938192 + }, + { + "epoch": 0.2631664235942311, + "grad_norm": 0.8588442802429199, + "learning_rate": 8.80870398249715e-06, + "loss": 0.2595560073852539, + "step": 4060, + "token_acc": 0.9271875967392426 + }, + { + "epoch": 0.2634905201750122, + "grad_norm": 0.8414188027381897, + "learning_rate": 8.805229871113458e-06, + "loss": 0.26101150512695315, + "step": 4065, + "token_acc": 0.9274352518687667 + }, + { + "epoch": 0.2638146167557932, + "grad_norm": 0.8336014747619629, + "learning_rate": 8.801751388640285e-06, + "loss": 0.26275513172149656, + "step": 4070, + "token_acc": 0.9267329628546359 + }, + { + "epoch": 0.2641387133365743, + "grad_norm": 0.9280481934547424, + "learning_rate": 8.798268539073381e-06, + "loss": 0.25650203227996826, + "step": 4075, + "token_acc": 0.926746981637613 + }, + { + "epoch": 0.2644628099173554, + "grad_norm": 0.8466014862060547, + "learning_rate": 8.794781326413518e-06, + "loss": 0.26517748832702637, + "step": 4080, + "token_acc": 0.9260950549780689 + }, + { + "epoch": 0.26478690649813647, + "grad_norm": 0.9260287284851074, + "learning_rate": 8.791289754666476e-06, + "loss": 0.25632548332214355, + "step": 4085, + "token_acc": 0.9279644397960518 + }, + { + "epoch": 0.2651110030789175, + "grad_norm": 0.8487626910209656, + "learning_rate": 8.787793827843045e-06, + "loss": 0.26072518825531005, + "step": 4090, + "token_acc": 0.9272499934536123 + }, + { + "epoch": 0.2654350996596986, + "grad_norm": 0.9548547267913818, + "learning_rate": 8.784293549959015e-06, + "loss": 0.25896425247192384, + "step": 4095, + "token_acc": 0.9273138849452728 + }, + { + "epoch": 0.26575919624047967, + "grad_norm": 0.8799179792404175, + "learning_rate": 8.780788925035178e-06, + "loss": 0.2615668296813965, + "step": 4100, + "token_acc": 0.9258027852533117 + }, + { + "epoch": 0.26608329282126075, + "grad_norm": 0.8915330171585083, + "learning_rate": 8.777279957097314e-06, + "loss": 0.25951237678527833, + "step": 4105, + "token_acc": 0.9273119544964206 + }, + { + "epoch": 0.2664073894020418, + "grad_norm": 0.8761894702911377, + "learning_rate": 8.773766650176194e-06, + "loss": 0.26072139739990235, + "step": 4110, + "token_acc": 0.9267907073980192 + }, + { + "epoch": 0.26673148598282287, + "grad_norm": 0.8780137896537781, + "learning_rate": 8.770249008307576e-06, + "loss": 0.2666799068450928, + "step": 4115, + "token_acc": 0.9249040957879645 + }, + { + "epoch": 0.26705558256360395, + "grad_norm": 0.8100029230117798, + "learning_rate": 8.766727035532194e-06, + "loss": 0.25823545455932617, + "step": 4120, + "token_acc": 0.9268384378544806 + }, + { + "epoch": 0.26737967914438504, + "grad_norm": 0.8637539744377136, + "learning_rate": 8.763200735895761e-06, + "loss": 0.25935347080230714, + "step": 4125, + "token_acc": 0.9276374520821299 + }, + { + "epoch": 0.2677037757251661, + "grad_norm": 0.830479085445404, + "learning_rate": 8.759670113448955e-06, + "loss": 0.25900342464447024, + "step": 4130, + "token_acc": 0.9269661448089581 + }, + { + "epoch": 0.26802787230594716, + "grad_norm": 0.8664636015892029, + "learning_rate": 8.756135172247424e-06, + "loss": 0.264119815826416, + "step": 4135, + "token_acc": 0.9258832490283773 + }, + { + "epoch": 0.26835196888672824, + "grad_norm": 0.854291558265686, + "learning_rate": 8.752595916351775e-06, + "loss": 0.26243431568145753, + "step": 4140, + "token_acc": 0.9262808578464996 + }, + { + "epoch": 0.26867606546750933, + "grad_norm": 0.8646603226661682, + "learning_rate": 8.749052349827571e-06, + "loss": 0.26148834228515627, + "step": 4145, + "token_acc": 0.9265831207718753 + }, + { + "epoch": 0.2690001620482904, + "grad_norm": 0.8024001121520996, + "learning_rate": 8.745504476745327e-06, + "loss": 0.26543483734130857, + "step": 4150, + "token_acc": 0.9270494810776811 + }, + { + "epoch": 0.26932425862907144, + "grad_norm": 0.8672388792037964, + "learning_rate": 8.741952301180505e-06, + "loss": 0.2581904172897339, + "step": 4155, + "token_acc": 0.92770268069444 + }, + { + "epoch": 0.26964835520985253, + "grad_norm": 0.8728849291801453, + "learning_rate": 8.738395827213514e-06, + "loss": 0.26202404499053955, + "step": 4160, + "token_acc": 0.9269285045434763 + }, + { + "epoch": 0.2699724517906336, + "grad_norm": 0.8646561503410339, + "learning_rate": 8.734835058929689e-06, + "loss": 0.2576277494430542, + "step": 4165, + "token_acc": 0.9275364208193462 + }, + { + "epoch": 0.2702965483714147, + "grad_norm": 0.8579664826393127, + "learning_rate": 8.73127000041931e-06, + "loss": 0.2608315467834473, + "step": 4170, + "token_acc": 0.9265154102840327 + }, + { + "epoch": 0.27062064495219573, + "grad_norm": 0.9109944105148315, + "learning_rate": 8.72770065577758e-06, + "loss": 0.26182596683502196, + "step": 4175, + "token_acc": 0.9265203819802661 + }, + { + "epoch": 0.2709447415329768, + "grad_norm": 0.834479808807373, + "learning_rate": 8.724127029104622e-06, + "loss": 0.2617176532745361, + "step": 4180, + "token_acc": 0.9271503552918399 + }, + { + "epoch": 0.2712688381137579, + "grad_norm": 0.7978215217590332, + "learning_rate": 8.720549124505486e-06, + "loss": 0.2595092535018921, + "step": 4185, + "token_acc": 0.9274626904510975 + }, + { + "epoch": 0.271592934694539, + "grad_norm": 0.9155907034873962, + "learning_rate": 8.716966946090128e-06, + "loss": 0.258466100692749, + "step": 4190, + "token_acc": 0.927371822889763 + }, + { + "epoch": 0.27191703127532, + "grad_norm": 0.8430355191230774, + "learning_rate": 8.71338049797342e-06, + "loss": 0.2581905126571655, + "step": 4195, + "token_acc": 0.9278152762411046 + }, + { + "epoch": 0.2722411278561011, + "grad_norm": 0.8546805381774902, + "learning_rate": 8.709789784275135e-06, + "loss": 0.2573368072509766, + "step": 4200, + "token_acc": 0.9268665513734787 + }, + { + "epoch": 0.2725652244368822, + "grad_norm": 0.8864487409591675, + "learning_rate": 8.706194809119946e-06, + "loss": 0.2595867156982422, + "step": 4205, + "token_acc": 0.9276965164107529 + }, + { + "epoch": 0.2728893210176633, + "grad_norm": 0.8413786888122559, + "learning_rate": 8.702595576637421e-06, + "loss": 0.25751380920410155, + "step": 4210, + "token_acc": 0.9276682107815681 + }, + { + "epoch": 0.27321341759844436, + "grad_norm": 0.8857844471931458, + "learning_rate": 8.69899209096202e-06, + "loss": 0.25587632656097414, + "step": 4215, + "token_acc": 0.9274264954816531 + }, + { + "epoch": 0.2735375141792254, + "grad_norm": 0.8905762434005737, + "learning_rate": 8.695384356233087e-06, + "loss": 0.26235363483428953, + "step": 4220, + "token_acc": 0.9261363636363636 + }, + { + "epoch": 0.2738616107600065, + "grad_norm": 0.8709800839424133, + "learning_rate": 8.691772376594848e-06, + "loss": 0.25589828491210936, + "step": 4225, + "token_acc": 0.9278785736726359 + }, + { + "epoch": 0.27418570734078757, + "grad_norm": 0.8865376114845276, + "learning_rate": 8.688156156196406e-06, + "loss": 0.2593587160110474, + "step": 4230, + "token_acc": 0.9265213078353184 + }, + { + "epoch": 0.27450980392156865, + "grad_norm": 0.8404337167739868, + "learning_rate": 8.684535699191732e-06, + "loss": 0.26578805446624754, + "step": 4235, + "token_acc": 0.9258751437724594 + }, + { + "epoch": 0.2748339005023497, + "grad_norm": 0.8681219816207886, + "learning_rate": 8.680911009739668e-06, + "loss": 0.2678975582122803, + "step": 4240, + "token_acc": 0.92585965560955 + }, + { + "epoch": 0.27515799708313077, + "grad_norm": 0.8618078231811523, + "learning_rate": 8.677282092003916e-06, + "loss": 0.26030144691467283, + "step": 4245, + "token_acc": 0.9266608192833707 + }, + { + "epoch": 0.27548209366391185, + "grad_norm": 0.8333085775375366, + "learning_rate": 8.673648950153031e-06, + "loss": 0.2648721694946289, + "step": 4250, + "token_acc": 0.9256482573097892 + }, + { + "epoch": 0.27580619024469294, + "grad_norm": 0.9011963605880737, + "learning_rate": 8.670011588360427e-06, + "loss": 0.2618532657623291, + "step": 4255, + "token_acc": 0.926247904037147 + }, + { + "epoch": 0.27613028682547397, + "grad_norm": 0.8680630326271057, + "learning_rate": 8.666370010804361e-06, + "loss": 0.26095402240753174, + "step": 4260, + "token_acc": 0.9269770869799824 + }, + { + "epoch": 0.27645438340625506, + "grad_norm": 0.8607080578804016, + "learning_rate": 8.662724221667938e-06, + "loss": 0.26872236728668214, + "step": 4265, + "token_acc": 0.925488050339046 + }, + { + "epoch": 0.27677847998703614, + "grad_norm": 0.9179531931877136, + "learning_rate": 8.65907422513909e-06, + "loss": 0.2637462854385376, + "step": 4270, + "token_acc": 0.9265373477001748 + }, + { + "epoch": 0.2771025765678172, + "grad_norm": 1.0090644359588623, + "learning_rate": 8.655420025410595e-06, + "loss": 0.25885162353515623, + "step": 4275, + "token_acc": 0.9270950105217614 + }, + { + "epoch": 0.27742667314859826, + "grad_norm": 0.805095374584198, + "learning_rate": 8.651761626680053e-06, + "loss": 0.25754637718200685, + "step": 4280, + "token_acc": 0.9276448887302154 + }, + { + "epoch": 0.27775076972937934, + "grad_norm": 0.8855307102203369, + "learning_rate": 8.648099033149883e-06, + "loss": 0.26982290744781495, + "step": 4285, + "token_acc": 0.9243785844692889 + }, + { + "epoch": 0.27807486631016043, + "grad_norm": 0.8368988037109375, + "learning_rate": 8.644432249027331e-06, + "loss": 0.26548309326171876, + "step": 4290, + "token_acc": 0.9249674457758428 + }, + { + "epoch": 0.2783989628909415, + "grad_norm": 0.854810893535614, + "learning_rate": 8.640761278524452e-06, + "loss": 0.25629405975341796, + "step": 4295, + "token_acc": 0.9281478175950538 + }, + { + "epoch": 0.2787230594717226, + "grad_norm": 0.9607113003730774, + "learning_rate": 8.637086125858112e-06, + "loss": 0.2584683895111084, + "step": 4300, + "token_acc": 0.9276847406019441 + }, + { + "epoch": 0.27904715605250363, + "grad_norm": 1.022955060005188, + "learning_rate": 8.63340679524998e-06, + "loss": 0.25905470848083495, + "step": 4305, + "token_acc": 0.9263242021000104 + }, + { + "epoch": 0.2793712526332847, + "grad_norm": 0.8529595136642456, + "learning_rate": 8.629723290926523e-06, + "loss": 0.2610101938247681, + "step": 4310, + "token_acc": 0.9268275168749103 + }, + { + "epoch": 0.2796953492140658, + "grad_norm": 0.8392961621284485, + "learning_rate": 8.626035617119004e-06, + "loss": 0.2516136169433594, + "step": 4315, + "token_acc": 0.9289293790281064 + }, + { + "epoch": 0.2800194457948469, + "grad_norm": 0.8918694853782654, + "learning_rate": 8.622343778063476e-06, + "loss": 0.2556185007095337, + "step": 4320, + "token_acc": 0.9275754313174549 + }, + { + "epoch": 0.2803435423756279, + "grad_norm": 0.8912484645843506, + "learning_rate": 8.618647778000774e-06, + "loss": 0.26317009925842283, + "step": 4325, + "token_acc": 0.9264018540336885 + }, + { + "epoch": 0.280667638956409, + "grad_norm": 0.8526169657707214, + "learning_rate": 8.614947621176518e-06, + "loss": 0.26370677947998045, + "step": 4330, + "token_acc": 0.9264761556637642 + }, + { + "epoch": 0.2809917355371901, + "grad_norm": 0.787969172000885, + "learning_rate": 8.611243311841096e-06, + "loss": 0.2644437551498413, + "step": 4335, + "token_acc": 0.9262885357238113 + }, + { + "epoch": 0.2813158321179712, + "grad_norm": 0.8639456033706665, + "learning_rate": 8.60753485424967e-06, + "loss": 0.2666572093963623, + "step": 4340, + "token_acc": 0.9255236867547535 + }, + { + "epoch": 0.2816399286987522, + "grad_norm": 0.8059642910957336, + "learning_rate": 8.603822252662167e-06, + "loss": 0.255582332611084, + "step": 4345, + "token_acc": 0.9275718127885698 + }, + { + "epoch": 0.2819640252795333, + "grad_norm": 0.865582287311554, + "learning_rate": 8.600105511343275e-06, + "loss": 0.25562052726745604, + "step": 4350, + "token_acc": 0.9275944251593182 + }, + { + "epoch": 0.2822881218603144, + "grad_norm": 0.9560853242874146, + "learning_rate": 8.596384634562436e-06, + "loss": 0.25618150234222414, + "step": 4355, + "token_acc": 0.9284469763708619 + }, + { + "epoch": 0.28261221844109546, + "grad_norm": 0.8125806450843811, + "learning_rate": 8.592659626593842e-06, + "loss": 0.26212809085845945, + "step": 4360, + "token_acc": 0.925521434769277 + }, + { + "epoch": 0.2829363150218765, + "grad_norm": 0.8340820074081421, + "learning_rate": 8.58893049171643e-06, + "loss": 0.25907244682312014, + "step": 4365, + "token_acc": 0.9279651766636016 + }, + { + "epoch": 0.2832604116026576, + "grad_norm": 1.152242660522461, + "learning_rate": 8.58519723421388e-06, + "loss": 0.2560394287109375, + "step": 4370, + "token_acc": 0.9284370109675332 + }, + { + "epoch": 0.28358450818343867, + "grad_norm": 0.8692997694015503, + "learning_rate": 8.581459858374607e-06, + "loss": 0.25981576442718507, + "step": 4375, + "token_acc": 0.9275689469035234 + }, + { + "epoch": 0.28390860476421975, + "grad_norm": 0.8109081387519836, + "learning_rate": 8.577718368491755e-06, + "loss": 0.265024733543396, + "step": 4380, + "token_acc": 0.9255212925981638 + }, + { + "epoch": 0.28423270134500084, + "grad_norm": 0.8136877417564392, + "learning_rate": 8.573972768863195e-06, + "loss": 0.2588953018188477, + "step": 4385, + "token_acc": 0.9262902301843557 + }, + { + "epoch": 0.28455679792578187, + "grad_norm": 0.8151147365570068, + "learning_rate": 8.57022306379152e-06, + "loss": 0.25647411346435545, + "step": 4390, + "token_acc": 0.9291680739951674 + }, + { + "epoch": 0.28488089450656295, + "grad_norm": 0.8916630148887634, + "learning_rate": 8.566469257584036e-06, + "loss": 0.256088399887085, + "step": 4395, + "token_acc": 0.9281545767152448 + }, + { + "epoch": 0.28520499108734404, + "grad_norm": 0.9227495789527893, + "learning_rate": 8.562711354552764e-06, + "loss": 0.2566675662994385, + "step": 4400, + "token_acc": 0.9272532202879779 + }, + { + "epoch": 0.2855290876681251, + "grad_norm": 0.9229126572608948, + "learning_rate": 8.558949359014426e-06, + "loss": 0.2588158845901489, + "step": 4405, + "token_acc": 0.9269686109818978 + }, + { + "epoch": 0.28585318424890616, + "grad_norm": 0.8906359076499939, + "learning_rate": 8.55518327529045e-06, + "loss": 0.25998449325561523, + "step": 4410, + "token_acc": 0.927286327817687 + }, + { + "epoch": 0.28617728082968724, + "grad_norm": 0.8305565118789673, + "learning_rate": 8.55141310770696e-06, + "loss": 0.25321784019470217, + "step": 4415, + "token_acc": 0.9292436754489448 + }, + { + "epoch": 0.2865013774104683, + "grad_norm": 0.8713905811309814, + "learning_rate": 8.547638860594765e-06, + "loss": 0.26077094078063967, + "step": 4420, + "token_acc": 0.9261191009936859 + }, + { + "epoch": 0.2868254739912494, + "grad_norm": 0.928564727306366, + "learning_rate": 8.543860538289369e-06, + "loss": 0.2653969764709473, + "step": 4425, + "token_acc": 0.9256734869269163 + }, + { + "epoch": 0.28714957057203044, + "grad_norm": 0.832565188407898, + "learning_rate": 8.540078145130945e-06, + "loss": 0.24893877506256104, + "step": 4430, + "token_acc": 0.9288546784441747 + }, + { + "epoch": 0.28747366715281153, + "grad_norm": 0.8889361023902893, + "learning_rate": 8.536291685464359e-06, + "loss": 0.2631813049316406, + "step": 4435, + "token_acc": 0.9270454485988644 + }, + { + "epoch": 0.2877977637335926, + "grad_norm": 0.8341926336288452, + "learning_rate": 8.532501163639135e-06, + "loss": 0.2558422088623047, + "step": 4440, + "token_acc": 0.9281146190264445 + }, + { + "epoch": 0.2881218603143737, + "grad_norm": 0.9090847969055176, + "learning_rate": 8.528706584009466e-06, + "loss": 0.2625986337661743, + "step": 4445, + "token_acc": 0.9262155059132721 + }, + { + "epoch": 0.28844595689515473, + "grad_norm": 0.8605325818061829, + "learning_rate": 8.524907950934208e-06, + "loss": 0.26441380977630613, + "step": 4450, + "token_acc": 0.9261832953378584 + }, + { + "epoch": 0.2887700534759358, + "grad_norm": 0.824472963809967, + "learning_rate": 8.521105268776874e-06, + "loss": 0.2646904945373535, + "step": 4455, + "token_acc": 0.9262664703045784 + }, + { + "epoch": 0.2890941500567169, + "grad_norm": 0.9076079726219177, + "learning_rate": 8.517298541905625e-06, + "loss": 0.2581419706344604, + "step": 4460, + "token_acc": 0.9276443334538166 + }, + { + "epoch": 0.289418246637498, + "grad_norm": 0.9562705159187317, + "learning_rate": 8.513487774693274e-06, + "loss": 0.26229536533355713, + "step": 4465, + "token_acc": 0.9255800274491863 + }, + { + "epoch": 0.2897423432182791, + "grad_norm": 0.8635479807853699, + "learning_rate": 8.509672971517267e-06, + "loss": 0.25909576416015623, + "step": 4470, + "token_acc": 0.9270664082636755 + }, + { + "epoch": 0.2900664397990601, + "grad_norm": 0.9615090489387512, + "learning_rate": 8.505854136759692e-06, + "loss": 0.2638214588165283, + "step": 4475, + "token_acc": 0.9265640715492476 + }, + { + "epoch": 0.2903905363798412, + "grad_norm": 0.8019809722900391, + "learning_rate": 8.502031274807267e-06, + "loss": 0.25642058849334715, + "step": 4480, + "token_acc": 0.9286695369661111 + }, + { + "epoch": 0.2907146329606223, + "grad_norm": 0.798728346824646, + "learning_rate": 8.498204390051334e-06, + "loss": 0.2538729190826416, + "step": 4485, + "token_acc": 0.927420722401221 + }, + { + "epoch": 0.29103872954140336, + "grad_norm": 0.8398305773735046, + "learning_rate": 8.49437348688786e-06, + "loss": 0.2576607227325439, + "step": 4490, + "token_acc": 0.9273271032930213 + }, + { + "epoch": 0.2913628261221844, + "grad_norm": 0.8285930752754211, + "learning_rate": 8.490538569717423e-06, + "loss": 0.2579181671142578, + "step": 4495, + "token_acc": 0.9273833591514165 + }, + { + "epoch": 0.2916869227029655, + "grad_norm": 0.8076286911964417, + "learning_rate": 8.486699642945216e-06, + "loss": 0.25668256282806395, + "step": 4500, + "token_acc": 0.9271647718679932 + }, + { + "epoch": 0.2916869227029655, + "eval_loss": 0.25997117161750793, + "eval_runtime": 96.7453, + "eval_samples_per_second": 206.728, + "eval_steps_per_second": 0.279, + "eval_token_acc": 0.926774896439432, + "step": 4500 + }, + { + "epoch": 0.29201101928374656, + "grad_norm": 0.8445154428482056, + "learning_rate": 8.482856710981036e-06, + "loss": 0.2559226036071777, + "step": 4505, + "token_acc": 0.9279525084088891 + }, + { + "epoch": 0.29233511586452765, + "grad_norm": 0.7991846799850464, + "learning_rate": 8.479009778239282e-06, + "loss": 0.260178804397583, + "step": 4510, + "token_acc": 0.9264195010892818 + }, + { + "epoch": 0.2926592124453087, + "grad_norm": 0.7896396517753601, + "learning_rate": 8.475158849138945e-06, + "loss": 0.255644416809082, + "step": 4515, + "token_acc": 0.9276749324503826 + }, + { + "epoch": 0.29298330902608977, + "grad_norm": 0.8383848667144775, + "learning_rate": 8.471303928103612e-06, + "loss": 0.25098657608032227, + "step": 4520, + "token_acc": 0.9298443214551575 + }, + { + "epoch": 0.29330740560687085, + "grad_norm": 0.8458635807037354, + "learning_rate": 8.467445019561453e-06, + "loss": 0.2586371421813965, + "step": 4525, + "token_acc": 0.9260129783863881 + }, + { + "epoch": 0.29363150218765194, + "grad_norm": 0.8787609934806824, + "learning_rate": 8.463582127945215e-06, + "loss": 0.2525789260864258, + "step": 4530, + "token_acc": 0.9286536473067093 + }, + { + "epoch": 0.29395559876843297, + "grad_norm": 0.8466007113456726, + "learning_rate": 8.459715257692231e-06, + "loss": 0.25201973915100095, + "step": 4535, + "token_acc": 0.9282693782821627 + }, + { + "epoch": 0.29427969534921405, + "grad_norm": 0.7773720026016235, + "learning_rate": 8.45584441324439e-06, + "loss": 0.24717361927032472, + "step": 4540, + "token_acc": 0.9297840058866047 + }, + { + "epoch": 0.29460379192999514, + "grad_norm": 0.9622704982757568, + "learning_rate": 8.451969599048155e-06, + "loss": 0.253770112991333, + "step": 4545, + "token_acc": 0.9290868487633603 + }, + { + "epoch": 0.2949278885107762, + "grad_norm": 0.8518549203872681, + "learning_rate": 8.44809081955455e-06, + "loss": 0.2599879503250122, + "step": 4550, + "token_acc": 0.9269575316278129 + }, + { + "epoch": 0.29525198509155726, + "grad_norm": 0.8601849675178528, + "learning_rate": 8.444208079219151e-06, + "loss": 0.2509438514709473, + "step": 4555, + "token_acc": 0.9292395439710057 + }, + { + "epoch": 0.29557608167233834, + "grad_norm": 0.9402835369110107, + "learning_rate": 8.440321382502081e-06, + "loss": 0.2564061641693115, + "step": 4560, + "token_acc": 0.9276455983265234 + }, + { + "epoch": 0.29590017825311943, + "grad_norm": 0.9080990552902222, + "learning_rate": 8.436430733868015e-06, + "loss": 0.2519416809082031, + "step": 4565, + "token_acc": 0.9284104188351288 + }, + { + "epoch": 0.2962242748339005, + "grad_norm": 0.8620240688323975, + "learning_rate": 8.43253613778616e-06, + "loss": 0.253730845451355, + "step": 4570, + "token_acc": 0.928283251683215 + }, + { + "epoch": 0.2965483714146816, + "grad_norm": 0.7967708110809326, + "learning_rate": 8.428637598730262e-06, + "loss": 0.25818068981170655, + "step": 4575, + "token_acc": 0.9268982679021643 + }, + { + "epoch": 0.29687246799546263, + "grad_norm": 0.7762065529823303, + "learning_rate": 8.424735121178598e-06, + "loss": 0.2558622360229492, + "step": 4580, + "token_acc": 0.9273359063457667 + }, + { + "epoch": 0.2971965645762437, + "grad_norm": 0.8516469597816467, + "learning_rate": 8.420828709613961e-06, + "loss": 0.25595688819885254, + "step": 4585, + "token_acc": 0.9276898599498464 + }, + { + "epoch": 0.2975206611570248, + "grad_norm": 0.793605387210846, + "learning_rate": 8.416918368523674e-06, + "loss": 0.2574084520339966, + "step": 4590, + "token_acc": 0.9276705677694909 + }, + { + "epoch": 0.2978447577378059, + "grad_norm": 0.7745986580848694, + "learning_rate": 8.413004102399566e-06, + "loss": 0.258134126663208, + "step": 4595, + "token_acc": 0.9270275706786578 + }, + { + "epoch": 0.2981688543185869, + "grad_norm": 0.8007134795188904, + "learning_rate": 8.409085915737976e-06, + "loss": 0.2545877456665039, + "step": 4600, + "token_acc": 0.9284854274908063 + }, + { + "epoch": 0.298492950899368, + "grad_norm": 0.8616408109664917, + "learning_rate": 8.405163813039749e-06, + "loss": 0.2454993963241577, + "step": 4605, + "token_acc": 0.9298686802363756 + }, + { + "epoch": 0.2988170474801491, + "grad_norm": 0.874708354473114, + "learning_rate": 8.401237798810226e-06, + "loss": 0.24809842109680175, + "step": 4610, + "token_acc": 0.9298469634530927 + }, + { + "epoch": 0.2991411440609302, + "grad_norm": 0.8593062162399292, + "learning_rate": 8.397307877559245e-06, + "loss": 0.25692005157470704, + "step": 4615, + "token_acc": 0.9281992771783576 + }, + { + "epoch": 0.2994652406417112, + "grad_norm": 0.8699996471405029, + "learning_rate": 8.393374053801124e-06, + "loss": 0.2564384937286377, + "step": 4620, + "token_acc": 0.9275587177523014 + }, + { + "epoch": 0.2997893372224923, + "grad_norm": 0.8594209551811218, + "learning_rate": 8.389436332054674e-06, + "loss": 0.2538710594177246, + "step": 4625, + "token_acc": 0.9283748891554953 + }, + { + "epoch": 0.3001134338032734, + "grad_norm": 0.8707131147384644, + "learning_rate": 8.385494716843175e-06, + "loss": 0.25265750885009763, + "step": 4630, + "token_acc": 0.9284925563147912 + }, + { + "epoch": 0.30043753038405446, + "grad_norm": 0.8986489772796631, + "learning_rate": 8.381549212694386e-06, + "loss": 0.25220727920532227, + "step": 4635, + "token_acc": 0.9285658759053159 + }, + { + "epoch": 0.3007616269648355, + "grad_norm": 0.8076885342597961, + "learning_rate": 8.377599824140528e-06, + "loss": 0.2537433624267578, + "step": 4640, + "token_acc": 0.9294523669987628 + }, + { + "epoch": 0.3010857235456166, + "grad_norm": 0.8762333393096924, + "learning_rate": 8.37364655571829e-06, + "loss": 0.2557267189025879, + "step": 4645, + "token_acc": 0.9281360883643673 + }, + { + "epoch": 0.30140982012639767, + "grad_norm": 0.8202180862426758, + "learning_rate": 8.36968941196881e-06, + "loss": 0.25569977760314944, + "step": 4650, + "token_acc": 0.9284688270948875 + }, + { + "epoch": 0.30173391670717875, + "grad_norm": 0.7823079228401184, + "learning_rate": 8.365728397437683e-06, + "loss": 0.25203168392181396, + "step": 4655, + "token_acc": 0.9282671242260744 + }, + { + "epoch": 0.30205801328795984, + "grad_norm": 0.901809811592102, + "learning_rate": 8.36176351667495e-06, + "loss": 0.2631944179534912, + "step": 4660, + "token_acc": 0.9267473029538762 + }, + { + "epoch": 0.30238210986874087, + "grad_norm": 0.8835824131965637, + "learning_rate": 8.357794774235094e-06, + "loss": 0.24550769329071045, + "step": 4665, + "token_acc": 0.9303645162138705 + }, + { + "epoch": 0.30270620644952195, + "grad_norm": 1.375875473022461, + "learning_rate": 8.353822174677028e-06, + "loss": 0.25109076499938965, + "step": 4670, + "token_acc": 0.9294088027578015 + }, + { + "epoch": 0.30303030303030304, + "grad_norm": 0.8311654329299927, + "learning_rate": 8.349845722564102e-06, + "loss": 0.2611923933029175, + "step": 4675, + "token_acc": 0.9270273099621822 + }, + { + "epoch": 0.3033543996110841, + "grad_norm": 0.8401497602462769, + "learning_rate": 8.345865422464093e-06, + "loss": 0.248571515083313, + "step": 4680, + "token_acc": 0.9299583458469896 + }, + { + "epoch": 0.30367849619186515, + "grad_norm": 0.7976726293563843, + "learning_rate": 8.341881278949188e-06, + "loss": 0.2562813997268677, + "step": 4685, + "token_acc": 0.9276254214905701 + }, + { + "epoch": 0.30400259277264624, + "grad_norm": 0.9042043089866638, + "learning_rate": 8.337893296596003e-06, + "loss": 0.2566051483154297, + "step": 4690, + "token_acc": 0.9277031889169093 + }, + { + "epoch": 0.3043266893534273, + "grad_norm": 0.8301060199737549, + "learning_rate": 8.33390147998555e-06, + "loss": 0.2525335788726807, + "step": 4695, + "token_acc": 0.9291540081757994 + }, + { + "epoch": 0.3046507859342084, + "grad_norm": 0.841970682144165, + "learning_rate": 8.329905833703256e-06, + "loss": 0.25495975017547606, + "step": 4700, + "token_acc": 0.92744929651129 + }, + { + "epoch": 0.30497488251498944, + "grad_norm": 0.7683242559432983, + "learning_rate": 8.325906362338939e-06, + "loss": 0.25261821746826174, + "step": 4705, + "token_acc": 0.9292814355593921 + }, + { + "epoch": 0.30529897909577053, + "grad_norm": 0.8385177254676819, + "learning_rate": 8.321903070486818e-06, + "loss": 0.2564859867095947, + "step": 4710, + "token_acc": 0.9282557921365397 + }, + { + "epoch": 0.3056230756765516, + "grad_norm": 0.7710776329040527, + "learning_rate": 8.317895962745497e-06, + "loss": 0.2535907745361328, + "step": 4715, + "token_acc": 0.9288265989197944 + }, + { + "epoch": 0.3059471722573327, + "grad_norm": 0.8197054266929626, + "learning_rate": 8.313885043717961e-06, + "loss": 0.25845472812652587, + "step": 4720, + "token_acc": 0.9267085017878349 + }, + { + "epoch": 0.30627126883811373, + "grad_norm": 0.8318530321121216, + "learning_rate": 8.30987031801158e-06, + "loss": 0.253541898727417, + "step": 4725, + "token_acc": 0.9279746546036516 + }, + { + "epoch": 0.3065953654188948, + "grad_norm": 0.8602234721183777, + "learning_rate": 8.305851790238091e-06, + "loss": 0.24508543014526368, + "step": 4730, + "token_acc": 0.9307151253791249 + }, + { + "epoch": 0.3069194619996759, + "grad_norm": 0.825896143913269, + "learning_rate": 8.301829465013599e-06, + "loss": 0.2642338752746582, + "step": 4735, + "token_acc": 0.9275627649382522 + }, + { + "epoch": 0.307243558580457, + "grad_norm": 0.7946286201477051, + "learning_rate": 8.29780334695857e-06, + "loss": 0.2531984806060791, + "step": 4740, + "token_acc": 0.9292734052070342 + }, + { + "epoch": 0.3075676551612381, + "grad_norm": 0.8747215867042542, + "learning_rate": 8.293773440697835e-06, + "loss": 0.262772274017334, + "step": 4745, + "token_acc": 0.9265769548771834 + }, + { + "epoch": 0.3078917517420191, + "grad_norm": 0.8840360641479492, + "learning_rate": 8.289739750860565e-06, + "loss": 0.25669431686401367, + "step": 4750, + "token_acc": 0.9268584582987267 + }, + { + "epoch": 0.3082158483228002, + "grad_norm": 0.9459658861160278, + "learning_rate": 8.285702282080287e-06, + "loss": 0.2484278678894043, + "step": 4755, + "token_acc": 0.9299281744445448 + }, + { + "epoch": 0.3085399449035813, + "grad_norm": 0.8330216407775879, + "learning_rate": 8.281661038994863e-06, + "loss": 0.2532350540161133, + "step": 4760, + "token_acc": 0.9288564335115141 + }, + { + "epoch": 0.30886404148436236, + "grad_norm": 0.96248859167099, + "learning_rate": 8.27761602624649e-06, + "loss": 0.25225965976715087, + "step": 4765, + "token_acc": 0.9292069833199305 + }, + { + "epoch": 0.3091881380651434, + "grad_norm": 1.0275380611419678, + "learning_rate": 8.2735672484817e-06, + "loss": 0.25095353126525877, + "step": 4770, + "token_acc": 0.9287246597940834 + }, + { + "epoch": 0.3095122346459245, + "grad_norm": 0.7922157645225525, + "learning_rate": 8.269514710351348e-06, + "loss": 0.25148792266845704, + "step": 4775, + "token_acc": 0.929184080241754 + }, + { + "epoch": 0.30983633122670556, + "grad_norm": 0.8703495264053345, + "learning_rate": 8.265458416510604e-06, + "loss": 0.25350565910339357, + "step": 4780, + "token_acc": 0.9274720343600756 + }, + { + "epoch": 0.31016042780748665, + "grad_norm": 0.8390120267868042, + "learning_rate": 8.261398371618961e-06, + "loss": 0.25656514167785643, + "step": 4785, + "token_acc": 0.9278823651775964 + }, + { + "epoch": 0.3104845243882677, + "grad_norm": 0.8252216577529907, + "learning_rate": 8.257334580340214e-06, + "loss": 0.24998075962066652, + "step": 4790, + "token_acc": 0.9286773802815077 + }, + { + "epoch": 0.31080862096904877, + "grad_norm": 0.7970818877220154, + "learning_rate": 8.25326704734246e-06, + "loss": 0.2452923536300659, + "step": 4795, + "token_acc": 0.929958549561017 + }, + { + "epoch": 0.31113271754982985, + "grad_norm": 0.8182471990585327, + "learning_rate": 8.249195777298102e-06, + "loss": 0.2508249759674072, + "step": 4800, + "token_acc": 0.9292277702611718 + }, + { + "epoch": 0.31145681413061094, + "grad_norm": 0.8293774127960205, + "learning_rate": 8.24512077488383e-06, + "loss": 0.2571417331695557, + "step": 4805, + "token_acc": 0.9276218328565591 + }, + { + "epoch": 0.31178091071139197, + "grad_norm": 0.8604652881622314, + "learning_rate": 8.241042044780624e-06, + "loss": 0.25105676651000974, + "step": 4810, + "token_acc": 0.9290629170328418 + }, + { + "epoch": 0.31210500729217305, + "grad_norm": 0.9412925243377686, + "learning_rate": 8.236959591673742e-06, + "loss": 0.24480276107788085, + "step": 4815, + "token_acc": 0.9301605437583989 + }, + { + "epoch": 0.31242910387295414, + "grad_norm": 0.8297778964042664, + "learning_rate": 8.232873420252725e-06, + "loss": 0.2522099494934082, + "step": 4820, + "token_acc": 0.9281306645069227 + }, + { + "epoch": 0.3127532004537352, + "grad_norm": 0.8403857350349426, + "learning_rate": 8.228783535211378e-06, + "loss": 0.25393450260162354, + "step": 4825, + "token_acc": 0.9295854739401546 + }, + { + "epoch": 0.3130772970345163, + "grad_norm": 0.8431863784790039, + "learning_rate": 8.22468994124778e-06, + "loss": 0.2535162687301636, + "step": 4830, + "token_acc": 0.9282226376225537 + }, + { + "epoch": 0.31340139361529734, + "grad_norm": 0.8600555062294006, + "learning_rate": 8.22059264306426e-06, + "loss": 0.2624182224273682, + "step": 4835, + "token_acc": 0.9270829934281369 + }, + { + "epoch": 0.3137254901960784, + "grad_norm": 0.8913409113883972, + "learning_rate": 8.216491645367413e-06, + "loss": 0.26180500984191896, + "step": 4840, + "token_acc": 0.9274914403700897 + }, + { + "epoch": 0.3140495867768595, + "grad_norm": 0.8180789947509766, + "learning_rate": 8.212386952868075e-06, + "loss": 0.25782151222229005, + "step": 4845, + "token_acc": 0.9268415741675076 + }, + { + "epoch": 0.3143736833576406, + "grad_norm": 0.7859653830528259, + "learning_rate": 8.208278570281334e-06, + "loss": 0.2502692699432373, + "step": 4850, + "token_acc": 0.9289705233924094 + }, + { + "epoch": 0.31469777993842163, + "grad_norm": 0.7936506867408752, + "learning_rate": 8.20416650232651e-06, + "loss": 0.25090012550354, + "step": 4855, + "token_acc": 0.9293159272971195 + }, + { + "epoch": 0.3150218765192027, + "grad_norm": 0.868652880191803, + "learning_rate": 8.200050753727158e-06, + "loss": 0.2527328014373779, + "step": 4860, + "token_acc": 0.9279462150106665 + }, + { + "epoch": 0.3153459730999838, + "grad_norm": 0.8084596395492554, + "learning_rate": 8.195931329211065e-06, + "loss": 0.2486318588256836, + "step": 4865, + "token_acc": 0.9306657265409248 + }, + { + "epoch": 0.3156700696807649, + "grad_norm": 0.8976461291313171, + "learning_rate": 8.191808233510236e-06, + "loss": 0.2592757225036621, + "step": 4870, + "token_acc": 0.9271228252401974 + }, + { + "epoch": 0.3159941662615459, + "grad_norm": 0.77048259973526, + "learning_rate": 8.187681471360897e-06, + "loss": 0.24336700439453124, + "step": 4875, + "token_acc": 0.9307782981299181 + }, + { + "epoch": 0.316318262842327, + "grad_norm": 0.9246774911880493, + "learning_rate": 8.183551047503481e-06, + "loss": 0.2508383750915527, + "step": 4880, + "token_acc": 0.9295256795256795 + }, + { + "epoch": 0.3166423594231081, + "grad_norm": 0.7566430568695068, + "learning_rate": 8.179416966682631e-06, + "loss": 0.2462817668914795, + "step": 4885, + "token_acc": 0.9301076313964866 + }, + { + "epoch": 0.3169664560038892, + "grad_norm": 0.7588401436805725, + "learning_rate": 8.175279233647188e-06, + "loss": 0.2516657590866089, + "step": 4890, + "token_acc": 0.9292460516624408 + }, + { + "epoch": 0.3172905525846702, + "grad_norm": 0.8508820533752441, + "learning_rate": 8.17113785315019e-06, + "loss": 0.2541757583618164, + "step": 4895, + "token_acc": 0.9287519580870641 + }, + { + "epoch": 0.3176146491654513, + "grad_norm": 0.8151727914810181, + "learning_rate": 8.166992829948868e-06, + "loss": 0.25663328170776367, + "step": 4900, + "token_acc": 0.9289739444593524 + }, + { + "epoch": 0.3179387457462324, + "grad_norm": 0.8366798162460327, + "learning_rate": 8.16284416880463e-06, + "loss": 0.24537343978881837, + "step": 4905, + "token_acc": 0.9296337242205288 + }, + { + "epoch": 0.31826284232701346, + "grad_norm": 1.6288071870803833, + "learning_rate": 8.158691874483068e-06, + "loss": 0.24519662857055663, + "step": 4910, + "token_acc": 0.9304471073315295 + }, + { + "epoch": 0.31858693890779455, + "grad_norm": 0.9123721718788147, + "learning_rate": 8.154535951753947e-06, + "loss": 0.25664308071136477, + "step": 4915, + "token_acc": 0.9287781036168133 + }, + { + "epoch": 0.3189110354885756, + "grad_norm": 0.8206848502159119, + "learning_rate": 8.150376405391199e-06, + "loss": 0.25568883419036864, + "step": 4920, + "token_acc": 0.928616502533703 + }, + { + "epoch": 0.31923513206935666, + "grad_norm": 0.9703457355499268, + "learning_rate": 8.14621324017292e-06, + "loss": 0.25037987232208253, + "step": 4925, + "token_acc": 0.9286584055583411 + }, + { + "epoch": 0.31955922865013775, + "grad_norm": 0.8173373341560364, + "learning_rate": 8.14204646088136e-06, + "loss": 0.2526259183883667, + "step": 4930, + "token_acc": 0.9288663109325878 + }, + { + "epoch": 0.31988332523091884, + "grad_norm": 0.829410970211029, + "learning_rate": 8.137876072302922e-06, + "loss": 0.25476980209350586, + "step": 4935, + "token_acc": 0.9274816034580353 + }, + { + "epoch": 0.32020742181169987, + "grad_norm": 0.7998035550117493, + "learning_rate": 8.13370207922816e-06, + "loss": 0.24864635467529297, + "step": 4940, + "token_acc": 0.9290111593499726 + }, + { + "epoch": 0.32053151839248095, + "grad_norm": 0.8048272132873535, + "learning_rate": 8.129524486451758e-06, + "loss": 0.24827265739440918, + "step": 4945, + "token_acc": 0.9295353303113039 + }, + { + "epoch": 0.32085561497326204, + "grad_norm": 0.8166364431381226, + "learning_rate": 8.125343298772545e-06, + "loss": 0.2600978374481201, + "step": 4950, + "token_acc": 0.9270806509759489 + }, + { + "epoch": 0.3211797115540431, + "grad_norm": 0.8427425026893616, + "learning_rate": 8.121158520993474e-06, + "loss": 0.24655332565307617, + "step": 4955, + "token_acc": 0.9300956776518785 + }, + { + "epoch": 0.32150380813482415, + "grad_norm": 0.7745749354362488, + "learning_rate": 8.116970157921622e-06, + "loss": 0.2470086097717285, + "step": 4960, + "token_acc": 0.9306469832119347 + }, + { + "epoch": 0.32182790471560524, + "grad_norm": 0.7962144613265991, + "learning_rate": 8.112778214368187e-06, + "loss": 0.2446513891220093, + "step": 4965, + "token_acc": 0.9306665965931114 + }, + { + "epoch": 0.3221520012963863, + "grad_norm": 0.8554896116256714, + "learning_rate": 8.108582695148479e-06, + "loss": 0.25709655284881594, + "step": 4970, + "token_acc": 0.9272511741474948 + }, + { + "epoch": 0.3224760978771674, + "grad_norm": 0.8150755763053894, + "learning_rate": 8.104383605081916e-06, + "loss": 0.24930920600891113, + "step": 4975, + "token_acc": 0.9289160501769057 + }, + { + "epoch": 0.32280019445794844, + "grad_norm": 0.9156748056411743, + "learning_rate": 8.100180948992015e-06, + "loss": 0.25332045555114746, + "step": 4980, + "token_acc": 0.9282575826107184 + }, + { + "epoch": 0.3231242910387295, + "grad_norm": 0.8166223168373108, + "learning_rate": 8.095974731706393e-06, + "loss": 0.256150221824646, + "step": 4985, + "token_acc": 0.9275375583478249 + }, + { + "epoch": 0.3234483876195106, + "grad_norm": 0.7595811486244202, + "learning_rate": 8.091764958056753e-06, + "loss": 0.2429211139678955, + "step": 4990, + "token_acc": 0.9305146254520185 + }, + { + "epoch": 0.3237724842002917, + "grad_norm": 0.7972972989082336, + "learning_rate": 8.087551632878891e-06, + "loss": 0.25571765899658205, + "step": 4995, + "token_acc": 0.9285718850861215 + }, + { + "epoch": 0.3240965807810728, + "grad_norm": 0.8897417187690735, + "learning_rate": 8.083334761012673e-06, + "loss": 0.25411286354064944, + "step": 5000, + "token_acc": 0.9279644236699451 + }, + { + "epoch": 0.3240965807810728, + "eval_loss": 0.25553035736083984, + "eval_runtime": 105.607, + "eval_samples_per_second": 189.381, + "eval_steps_per_second": 0.256, + "eval_token_acc": 0.9272240997719853, + "step": 5000 + }, + { + "epoch": 0.3244206773618538, + "grad_norm": 0.781226634979248, + "learning_rate": 8.079114347302048e-06, + "loss": 0.25302577018737793, + "step": 5005, + "token_acc": 0.9274667369253062 + }, + { + "epoch": 0.3247447739426349, + "grad_norm": 0.8227452635765076, + "learning_rate": 8.074890396595028e-06, + "loss": 0.2615841865539551, + "step": 5010, + "token_acc": 0.9265363037909006 + }, + { + "epoch": 0.325068870523416, + "grad_norm": 0.8179060816764832, + "learning_rate": 8.070662913743688e-06, + "loss": 0.25869903564453123, + "step": 5015, + "token_acc": 0.9272949006532302 + }, + { + "epoch": 0.3253929671041971, + "grad_norm": 0.9276995062828064, + "learning_rate": 8.066431903604165e-06, + "loss": 0.25614514350891116, + "step": 5020, + "token_acc": 0.9273942180316372 + }, + { + "epoch": 0.3257170636849781, + "grad_norm": 0.9039407968521118, + "learning_rate": 8.06219737103664e-06, + "loss": 0.24863269329071044, + "step": 5025, + "token_acc": 0.9293900086829145 + }, + { + "epoch": 0.3260411602657592, + "grad_norm": 0.8861247301101685, + "learning_rate": 8.057959320905352e-06, + "loss": 0.2517418384552002, + "step": 5030, + "token_acc": 0.929113940513194 + }, + { + "epoch": 0.3263652568465403, + "grad_norm": 0.8858475089073181, + "learning_rate": 8.053717758078566e-06, + "loss": 0.24458818435668944, + "step": 5035, + "token_acc": 0.9307570108607554 + }, + { + "epoch": 0.32668935342732136, + "grad_norm": 0.8284400105476379, + "learning_rate": 8.049472687428594e-06, + "loss": 0.24628174304962158, + "step": 5040, + "token_acc": 0.9292034258484126 + }, + { + "epoch": 0.3270134500081024, + "grad_norm": 0.8065065145492554, + "learning_rate": 8.045224113831773e-06, + "loss": 0.25313968658447267, + "step": 5045, + "token_acc": 0.9285822184567086 + }, + { + "epoch": 0.3273375465888835, + "grad_norm": 0.8459227681159973, + "learning_rate": 8.040972042168464e-06, + "loss": 0.24716649055480958, + "step": 5050, + "token_acc": 0.9313201339521767 + }, + { + "epoch": 0.32766164316966456, + "grad_norm": 0.9181363582611084, + "learning_rate": 8.036716477323045e-06, + "loss": 0.26085968017578126, + "step": 5055, + "token_acc": 0.9276424939791277 + }, + { + "epoch": 0.32798573975044565, + "grad_norm": 0.77690589427948, + "learning_rate": 8.032457424183909e-06, + "loss": 0.24831562042236327, + "step": 5060, + "token_acc": 0.9294923240220571 + }, + { + "epoch": 0.3283098363312267, + "grad_norm": 0.8056610822677612, + "learning_rate": 8.028194887643455e-06, + "loss": 0.24811506271362305, + "step": 5065, + "token_acc": 0.9298639014906027 + }, + { + "epoch": 0.32863393291200776, + "grad_norm": 0.9039129018783569, + "learning_rate": 8.023928872598082e-06, + "loss": 0.25830817222595215, + "step": 5070, + "token_acc": 0.9270809596666558 + }, + { + "epoch": 0.32895802949278885, + "grad_norm": 0.7763206362724304, + "learning_rate": 8.019659383948183e-06, + "loss": 0.25286617279052737, + "step": 5075, + "token_acc": 0.9289317973472828 + }, + { + "epoch": 0.32928212607356994, + "grad_norm": 0.8035215735435486, + "learning_rate": 8.015386426598154e-06, + "loss": 0.24364714622497557, + "step": 5080, + "token_acc": 0.930171096552493 + }, + { + "epoch": 0.329606222654351, + "grad_norm": 0.8502820134162903, + "learning_rate": 8.011110005456355e-06, + "loss": 0.2522239685058594, + "step": 5085, + "token_acc": 0.9294928651553491 + }, + { + "epoch": 0.32993031923513205, + "grad_norm": 0.7755847573280334, + "learning_rate": 8.006830125435142e-06, + "loss": 0.2543953895568848, + "step": 5090, + "token_acc": 0.928968738111168 + }, + { + "epoch": 0.33025441581591314, + "grad_norm": 0.8396223187446594, + "learning_rate": 8.002546791450836e-06, + "loss": 0.2531644344329834, + "step": 5095, + "token_acc": 0.9282779737219488 + }, + { + "epoch": 0.3305785123966942, + "grad_norm": 0.8563336133956909, + "learning_rate": 7.998260008423729e-06, + "loss": 0.2522023916244507, + "step": 5100, + "token_acc": 0.928631043652103 + }, + { + "epoch": 0.3309026089774753, + "grad_norm": 0.7964015603065491, + "learning_rate": 7.993969781278071e-06, + "loss": 0.25537183284759524, + "step": 5105, + "token_acc": 0.9278630276053258 + }, + { + "epoch": 0.33122670555825634, + "grad_norm": 0.7790189981460571, + "learning_rate": 7.98967611494207e-06, + "loss": 0.2565012454986572, + "step": 5110, + "token_acc": 0.9281973404980238 + }, + { + "epoch": 0.3315508021390374, + "grad_norm": 0.8407790660858154, + "learning_rate": 7.985379014347887e-06, + "loss": 0.24981045722961426, + "step": 5115, + "token_acc": 0.9296262499036064 + }, + { + "epoch": 0.3318748987198185, + "grad_norm": 0.8096004128456116, + "learning_rate": 7.981078484431628e-06, + "loss": 0.24656393527984619, + "step": 5120, + "token_acc": 0.9299084831570066 + }, + { + "epoch": 0.3321989953005996, + "grad_norm": 0.838131308555603, + "learning_rate": 7.976774530133332e-06, + "loss": 0.2573067665100098, + "step": 5125, + "token_acc": 0.9283298001894367 + }, + { + "epoch": 0.33252309188138063, + "grad_norm": 0.8461585640907288, + "learning_rate": 7.972467156396978e-06, + "loss": 0.25550527572631837, + "step": 5130, + "token_acc": 0.9273887079261672 + }, + { + "epoch": 0.3328471884621617, + "grad_norm": 0.8040698766708374, + "learning_rate": 7.968156368170476e-06, + "loss": 0.2437753915786743, + "step": 5135, + "token_acc": 0.9308880233309194 + }, + { + "epoch": 0.3331712850429428, + "grad_norm": 0.7927164435386658, + "learning_rate": 7.963842170405646e-06, + "loss": 0.24805796146392822, + "step": 5140, + "token_acc": 0.9288831138607141 + }, + { + "epoch": 0.3334953816237239, + "grad_norm": 0.8112002611160278, + "learning_rate": 7.959524568058234e-06, + "loss": 0.2482987880706787, + "step": 5145, + "token_acc": 0.93027632872894 + }, + { + "epoch": 0.3338194782045049, + "grad_norm": 0.9138498902320862, + "learning_rate": 7.955203566087897e-06, + "loss": 0.2555858612060547, + "step": 5150, + "token_acc": 0.9271770795401342 + }, + { + "epoch": 0.334143574785286, + "grad_norm": 0.8687700629234314, + "learning_rate": 7.950879169458194e-06, + "loss": 0.25181288719177247, + "step": 5155, + "token_acc": 0.9285288862418106 + }, + { + "epoch": 0.3344676713660671, + "grad_norm": 0.8657352328300476, + "learning_rate": 7.946551383136582e-06, + "loss": 0.2562386512756348, + "step": 5160, + "token_acc": 0.9276751068727392 + }, + { + "epoch": 0.3347917679468482, + "grad_norm": 0.8554297685623169, + "learning_rate": 7.942220212094415e-06, + "loss": 0.2441800355911255, + "step": 5165, + "token_acc": 0.9303021945549821 + }, + { + "epoch": 0.33511586452762926, + "grad_norm": 0.8135873675346375, + "learning_rate": 7.937885661306938e-06, + "loss": 0.25539531707763674, + "step": 5170, + "token_acc": 0.9282641741324105 + }, + { + "epoch": 0.3354399611084103, + "grad_norm": 0.824191689491272, + "learning_rate": 7.933547735753268e-06, + "loss": 0.2518582344055176, + "step": 5175, + "token_acc": 0.9289096135586166 + }, + { + "epoch": 0.3357640576891914, + "grad_norm": 0.8475911617279053, + "learning_rate": 7.929206440416412e-06, + "loss": 0.24718787670135497, + "step": 5180, + "token_acc": 0.9301319787626416 + }, + { + "epoch": 0.33608815426997246, + "grad_norm": 0.8635424971580505, + "learning_rate": 7.924861780283235e-06, + "loss": 0.2528079986572266, + "step": 5185, + "token_acc": 0.9286200922663655 + }, + { + "epoch": 0.33641225085075355, + "grad_norm": 0.8695507049560547, + "learning_rate": 7.920513760344475e-06, + "loss": 0.2556718349456787, + "step": 5190, + "token_acc": 0.9281598359300056 + }, + { + "epoch": 0.3367363474315346, + "grad_norm": 0.8251754641532898, + "learning_rate": 7.91616238559473e-06, + "loss": 0.25039381980895997, + "step": 5195, + "token_acc": 0.929501098786103 + }, + { + "epoch": 0.33706044401231566, + "grad_norm": 0.8470465540885925, + "learning_rate": 7.911807661032446e-06, + "loss": 0.2590325355529785, + "step": 5200, + "token_acc": 0.9272696303866462 + }, + { + "epoch": 0.33738454059309675, + "grad_norm": 0.8453739285469055, + "learning_rate": 7.90744959165992e-06, + "loss": 0.2517526149749756, + "step": 5205, + "token_acc": 0.9287489061631717 + }, + { + "epoch": 0.33770863717387783, + "grad_norm": 1.0070303678512573, + "learning_rate": 7.903088182483295e-06, + "loss": 0.2535067558288574, + "step": 5210, + "token_acc": 0.927861559095816 + }, + { + "epoch": 0.33803273375465887, + "grad_norm": 0.8390263319015503, + "learning_rate": 7.898723438512546e-06, + "loss": 0.24636497497558593, + "step": 5215, + "token_acc": 0.9297858921593488 + }, + { + "epoch": 0.33835683033543995, + "grad_norm": 0.8553630113601685, + "learning_rate": 7.894355364761476e-06, + "loss": 0.25316691398620605, + "step": 5220, + "token_acc": 0.9281559197029111 + }, + { + "epoch": 0.33868092691622104, + "grad_norm": 0.8556739091873169, + "learning_rate": 7.889983966247721e-06, + "loss": 0.2501840591430664, + "step": 5225, + "token_acc": 0.9292125938652201 + }, + { + "epoch": 0.3390050234970021, + "grad_norm": 1.0619709491729736, + "learning_rate": 7.885609247992727e-06, + "loss": 0.25245227813720705, + "step": 5230, + "token_acc": 0.9284152443181087 + }, + { + "epoch": 0.33932912007778315, + "grad_norm": 0.8231361508369446, + "learning_rate": 7.881231215021763e-06, + "loss": 0.255342435836792, + "step": 5235, + "token_acc": 0.9281765787218453 + }, + { + "epoch": 0.33965321665856424, + "grad_norm": 1.1481475830078125, + "learning_rate": 7.876849872363896e-06, + "loss": 0.25723843574523925, + "step": 5240, + "token_acc": 0.92721412483496 + }, + { + "epoch": 0.3399773132393453, + "grad_norm": 0.8251258730888367, + "learning_rate": 7.872465225052004e-06, + "loss": 0.2554172515869141, + "step": 5245, + "token_acc": 0.9279755594914373 + }, + { + "epoch": 0.3403014098201264, + "grad_norm": 0.8859728574752808, + "learning_rate": 7.868077278122754e-06, + "loss": 0.24776082038879393, + "step": 5250, + "token_acc": 0.9295331874070369 + }, + { + "epoch": 0.3406255064009075, + "grad_norm": 0.8037582635879517, + "learning_rate": 7.863686036616606e-06, + "loss": 0.25240345001220704, + "step": 5255, + "token_acc": 0.9295603768198687 + }, + { + "epoch": 0.3409496029816885, + "grad_norm": 0.8746057152748108, + "learning_rate": 7.859291505577805e-06, + "loss": 0.2473452091217041, + "step": 5260, + "token_acc": 0.9296412692061976 + }, + { + "epoch": 0.3412736995624696, + "grad_norm": 0.9796491265296936, + "learning_rate": 7.854893690054373e-06, + "loss": 0.2474212169647217, + "step": 5265, + "token_acc": 0.9303670904156549 + }, + { + "epoch": 0.3415977961432507, + "grad_norm": 0.8041320443153381, + "learning_rate": 7.850492595098107e-06, + "loss": 0.2509519815444946, + "step": 5270, + "token_acc": 0.9284994102025081 + }, + { + "epoch": 0.3419218927240318, + "grad_norm": 0.867673397064209, + "learning_rate": 7.846088225764572e-06, + "loss": 0.25418758392333984, + "step": 5275, + "token_acc": 0.9279067790514428 + }, + { + "epoch": 0.3422459893048128, + "grad_norm": 0.8879721760749817, + "learning_rate": 7.84168058711309e-06, + "loss": 0.2534532070159912, + "step": 5280, + "token_acc": 0.928356297733613 + }, + { + "epoch": 0.3425700858855939, + "grad_norm": 0.8307471871376038, + "learning_rate": 7.83726968420674e-06, + "loss": 0.25272173881530763, + "step": 5285, + "token_acc": 0.9286928904621696 + }, + { + "epoch": 0.342894182466375, + "grad_norm": 0.842936635017395, + "learning_rate": 7.832855522112355e-06, + "loss": 0.24378321170806885, + "step": 5290, + "token_acc": 0.9306362866361056 + }, + { + "epoch": 0.34321827904715607, + "grad_norm": 0.8907715678215027, + "learning_rate": 7.828438105900504e-06, + "loss": 0.253885555267334, + "step": 5295, + "token_acc": 0.9291491852222061 + }, + { + "epoch": 0.3435423756279371, + "grad_norm": 0.7900069355964661, + "learning_rate": 7.824017440645502e-06, + "loss": 0.24583210945129394, + "step": 5300, + "token_acc": 0.9303247168015851 + }, + { + "epoch": 0.3438664722087182, + "grad_norm": 0.8247885704040527, + "learning_rate": 7.819593531425393e-06, + "loss": 0.25352504253387453, + "step": 5305, + "token_acc": 0.9287361015966696 + }, + { + "epoch": 0.3441905687894993, + "grad_norm": 0.9311105608940125, + "learning_rate": 7.815166383321944e-06, + "loss": 0.24814109802246093, + "step": 5310, + "token_acc": 0.928730541177693 + }, + { + "epoch": 0.34451466537028036, + "grad_norm": 0.8059632778167725, + "learning_rate": 7.810736001420648e-06, + "loss": 0.24797918796539306, + "step": 5315, + "token_acc": 0.9302761123177568 + }, + { + "epoch": 0.3448387619510614, + "grad_norm": 0.8258737921714783, + "learning_rate": 7.80630239081071e-06, + "loss": 0.25756282806396485, + "step": 5320, + "token_acc": 0.9275917291970709 + }, + { + "epoch": 0.3451628585318425, + "grad_norm": 0.815545916557312, + "learning_rate": 7.801865556585043e-06, + "loss": 0.24850454330444335, + "step": 5325, + "token_acc": 0.9298312858525684 + }, + { + "epoch": 0.34548695511262356, + "grad_norm": 0.9134780168533325, + "learning_rate": 7.797425503840263e-06, + "loss": 0.24773640632629396, + "step": 5330, + "token_acc": 0.9292507478121993 + }, + { + "epoch": 0.34581105169340465, + "grad_norm": 0.8894223570823669, + "learning_rate": 7.792982237676687e-06, + "loss": 0.25623526573181155, + "step": 5335, + "token_acc": 0.927582139401731 + }, + { + "epoch": 0.34613514827418573, + "grad_norm": 0.8839740753173828, + "learning_rate": 7.788535763198318e-06, + "loss": 0.25153093338012694, + "step": 5340, + "token_acc": 0.9290341578327445 + }, + { + "epoch": 0.34645924485496676, + "grad_norm": 0.8760332465171814, + "learning_rate": 7.78408608551285e-06, + "loss": 0.2507377862930298, + "step": 5345, + "token_acc": 0.928809927567667 + }, + { + "epoch": 0.34678334143574785, + "grad_norm": 0.8758668303489685, + "learning_rate": 7.779633209731649e-06, + "loss": 0.25293607711791993, + "step": 5350, + "token_acc": 0.9293097287640084 + }, + { + "epoch": 0.34710743801652894, + "grad_norm": 0.7980352640151978, + "learning_rate": 7.775177140969762e-06, + "loss": 0.25273752212524414, + "step": 5355, + "token_acc": 0.9286163810622649 + }, + { + "epoch": 0.34743153459731, + "grad_norm": 1.0390256643295288, + "learning_rate": 7.770717884345902e-06, + "loss": 0.24456522464752198, + "step": 5360, + "token_acc": 0.9306050866132491 + }, + { + "epoch": 0.34775563117809105, + "grad_norm": 0.907213568687439, + "learning_rate": 7.76625544498244e-06, + "loss": 0.24939141273498536, + "step": 5365, + "token_acc": 0.9291206322957831 + }, + { + "epoch": 0.34807972775887214, + "grad_norm": 0.8462250828742981, + "learning_rate": 7.761789828005408e-06, + "loss": 0.2489654541015625, + "step": 5370, + "token_acc": 0.9292870616339558 + }, + { + "epoch": 0.3484038243396532, + "grad_norm": 0.9953560829162598, + "learning_rate": 7.757321038544483e-06, + "loss": 0.24589056968688966, + "step": 5375, + "token_acc": 0.9298889584500313 + }, + { + "epoch": 0.3487279209204343, + "grad_norm": 0.7983831167221069, + "learning_rate": 7.752849081732993e-06, + "loss": 0.2558305263519287, + "step": 5380, + "token_acc": 0.9274585734870316 + }, + { + "epoch": 0.34905201750121534, + "grad_norm": 0.9755795001983643, + "learning_rate": 7.748373962707898e-06, + "loss": 0.24854235649108886, + "step": 5385, + "token_acc": 0.9288953593238681 + }, + { + "epoch": 0.3493761140819964, + "grad_norm": 0.8616177439689636, + "learning_rate": 7.743895686609793e-06, + "loss": 0.25255656242370605, + "step": 5390, + "token_acc": 0.9284949634349244 + }, + { + "epoch": 0.3497002106627775, + "grad_norm": 0.8259103894233704, + "learning_rate": 7.7394142585829e-06, + "loss": 0.25195789337158203, + "step": 5395, + "token_acc": 0.9292784028423992 + }, + { + "epoch": 0.3500243072435586, + "grad_norm": 0.9628980755805969, + "learning_rate": 7.734929683775064e-06, + "loss": 0.2485969066619873, + "step": 5400, + "token_acc": 0.9295956993769879 + }, + { + "epoch": 0.3503484038243396, + "grad_norm": 0.8506217002868652, + "learning_rate": 7.730441967337735e-06, + "loss": 0.25638890266418457, + "step": 5405, + "token_acc": 0.9278660480547419 + }, + { + "epoch": 0.3506725004051207, + "grad_norm": 0.7939960360527039, + "learning_rate": 7.725951114425985e-06, + "loss": 0.24454362392425538, + "step": 5410, + "token_acc": 0.9306267474370923 + }, + { + "epoch": 0.3509965969859018, + "grad_norm": 1.1062229871749878, + "learning_rate": 7.721457130198482e-06, + "loss": 0.24372103214263915, + "step": 5415, + "token_acc": 0.9311154292570326 + }, + { + "epoch": 0.3513206935666829, + "grad_norm": 0.9029282927513123, + "learning_rate": 7.71696001981749e-06, + "loss": 0.25217740535736083, + "step": 5420, + "token_acc": 0.9288803333157534 + }, + { + "epoch": 0.35164479014746397, + "grad_norm": 0.8418589234352112, + "learning_rate": 7.712459788448863e-06, + "loss": 0.2501793384552002, + "step": 5425, + "token_acc": 0.9293845351414787 + }, + { + "epoch": 0.351968886728245, + "grad_norm": 0.793563723564148, + "learning_rate": 7.707956441262047e-06, + "loss": 0.2526703834533691, + "step": 5430, + "token_acc": 0.9287680931403398 + }, + { + "epoch": 0.3522929833090261, + "grad_norm": 0.8207056522369385, + "learning_rate": 7.703449983430064e-06, + "loss": 0.2467909574508667, + "step": 5435, + "token_acc": 0.9294978752248612 + }, + { + "epoch": 0.3526170798898072, + "grad_norm": 0.8784594535827637, + "learning_rate": 7.698940420129503e-06, + "loss": 0.25243971347808836, + "step": 5440, + "token_acc": 0.9289702576799036 + }, + { + "epoch": 0.35294117647058826, + "grad_norm": 0.8435435891151428, + "learning_rate": 7.694427756540529e-06, + "loss": 0.24867188930511475, + "step": 5445, + "token_acc": 0.9300796748491984 + }, + { + "epoch": 0.3532652730513693, + "grad_norm": 0.8420418500900269, + "learning_rate": 7.689911997846863e-06, + "loss": 0.2528546810150146, + "step": 5450, + "token_acc": 0.9287434268756183 + }, + { + "epoch": 0.3535893696321504, + "grad_norm": 0.824761688709259, + "learning_rate": 7.685393149235784e-06, + "loss": 0.2558503866195679, + "step": 5455, + "token_acc": 0.9284002180237331 + }, + { + "epoch": 0.35391346621293146, + "grad_norm": 0.8431583642959595, + "learning_rate": 7.680871215898117e-06, + "loss": 0.2536440849304199, + "step": 5460, + "token_acc": 0.9287037278585031 + }, + { + "epoch": 0.35423756279371255, + "grad_norm": 0.8413884043693542, + "learning_rate": 7.676346203028238e-06, + "loss": 0.2470717430114746, + "step": 5465, + "token_acc": 0.9288218004540689 + }, + { + "epoch": 0.3545616593744936, + "grad_norm": 0.8339266180992126, + "learning_rate": 7.671818115824052e-06, + "loss": 0.25326056480407716, + "step": 5470, + "token_acc": 0.9284378319550073 + }, + { + "epoch": 0.35488575595527466, + "grad_norm": 0.8217309713363647, + "learning_rate": 7.667286959486997e-06, + "loss": 0.2545089960098267, + "step": 5475, + "token_acc": 0.9279552129221733 + }, + { + "epoch": 0.35520985253605575, + "grad_norm": 0.7824126482009888, + "learning_rate": 7.66275273922204e-06, + "loss": 0.2447800636291504, + "step": 5480, + "token_acc": 0.9308840655038785 + }, + { + "epoch": 0.35553394911683683, + "grad_norm": 0.8091416954994202, + "learning_rate": 7.65821546023767e-06, + "loss": 0.25156702995300295, + "step": 5485, + "token_acc": 0.9285445118702652 + }, + { + "epoch": 0.35585804569761786, + "grad_norm": 0.8021688461303711, + "learning_rate": 7.65367512774588e-06, + "loss": 0.251130485534668, + "step": 5490, + "token_acc": 0.9281955569380719 + }, + { + "epoch": 0.35618214227839895, + "grad_norm": 0.8379022479057312, + "learning_rate": 7.649131746962177e-06, + "loss": 0.2552762985229492, + "step": 5495, + "token_acc": 0.927667386329847 + }, + { + "epoch": 0.35650623885918004, + "grad_norm": 0.9970967769622803, + "learning_rate": 7.644585323105571e-06, + "loss": 0.2519248962402344, + "step": 5500, + "token_acc": 0.9298844623863792 + }, + { + "epoch": 0.35650623885918004, + "eval_loss": 0.2532550096511841, + "eval_runtime": 117.32, + "eval_samples_per_second": 170.474, + "eval_steps_per_second": 0.23, + "eval_token_acc": 0.9276949478242996, + "step": 5500 + }, + { + "epoch": 0.3568303354399611, + "grad_norm": 0.8221502900123596, + "learning_rate": 7.640035861398566e-06, + "loss": 0.2463094711303711, + "step": 5505, + "token_acc": 0.9299298784602826 + }, + { + "epoch": 0.3571544320207422, + "grad_norm": 0.7970005869865417, + "learning_rate": 7.63548336706715e-06, + "loss": 0.24803633689880372, + "step": 5510, + "token_acc": 0.929139877685432 + }, + { + "epoch": 0.35747852860152324, + "grad_norm": 0.8124566674232483, + "learning_rate": 7.630927845340803e-06, + "loss": 0.2541279554367065, + "step": 5515, + "token_acc": 0.9289399884180184 + }, + { + "epoch": 0.3578026251823043, + "grad_norm": 0.8208630681037903, + "learning_rate": 7.626369301452479e-06, + "loss": 0.2485501766204834, + "step": 5520, + "token_acc": 0.9295272021141839 + }, + { + "epoch": 0.3581267217630854, + "grad_norm": 0.8408405184745789, + "learning_rate": 7.621807740638603e-06, + "loss": 0.25894615650177, + "step": 5525, + "token_acc": 0.9277205800224703 + }, + { + "epoch": 0.3584508183438665, + "grad_norm": 0.8286508917808533, + "learning_rate": 7.6172431681390654e-06, + "loss": 0.24380989074707032, + "step": 5530, + "token_acc": 0.9312132507349853 + }, + { + "epoch": 0.3587749149246475, + "grad_norm": 0.8268141746520996, + "learning_rate": 7.61267558919722e-06, + "loss": 0.2462075710296631, + "step": 5535, + "token_acc": 0.9294996189341684 + }, + { + "epoch": 0.3590990115054286, + "grad_norm": 0.8434064984321594, + "learning_rate": 7.608105009059867e-06, + "loss": 0.24595377445220948, + "step": 5540, + "token_acc": 0.9304552023121387 + }, + { + "epoch": 0.3594231080862097, + "grad_norm": 0.8048689961433411, + "learning_rate": 7.603531432977262e-06, + "loss": 0.24870061874389648, + "step": 5545, + "token_acc": 0.9299459431497814 + }, + { + "epoch": 0.3597472046669908, + "grad_norm": 0.7562397122383118, + "learning_rate": 7.598954866203094e-06, + "loss": 0.2463165521621704, + "step": 5550, + "token_acc": 0.929876804570729 + }, + { + "epoch": 0.3600713012477718, + "grad_norm": 0.9828148484230042, + "learning_rate": 7.594375313994496e-06, + "loss": 0.24554038047790527, + "step": 5555, + "token_acc": 0.9308532356848893 + }, + { + "epoch": 0.3603953978285529, + "grad_norm": 0.7583472728729248, + "learning_rate": 7.589792781612022e-06, + "loss": 0.24926190376281737, + "step": 5560, + "token_acc": 0.9287607921173412 + }, + { + "epoch": 0.360719494409334, + "grad_norm": 0.8745729327201843, + "learning_rate": 7.5852072743196556e-06, + "loss": 0.245491886138916, + "step": 5565, + "token_acc": 0.930393507485587 + }, + { + "epoch": 0.36104359099011507, + "grad_norm": 0.8307902812957764, + "learning_rate": 7.580618797384795e-06, + "loss": 0.24790010452270508, + "step": 5570, + "token_acc": 0.9301750235315068 + }, + { + "epoch": 0.3613676875708961, + "grad_norm": 0.8324232697486877, + "learning_rate": 7.576027356078252e-06, + "loss": 0.2472665786743164, + "step": 5575, + "token_acc": 0.9296586805420728 + }, + { + "epoch": 0.3616917841516772, + "grad_norm": 0.7718153595924377, + "learning_rate": 7.5714329556742385e-06, + "loss": 0.25357723236083984, + "step": 5580, + "token_acc": 0.9277195286000245 + }, + { + "epoch": 0.3620158807324583, + "grad_norm": 0.8068608045578003, + "learning_rate": 7.566835601450371e-06, + "loss": 0.2465280532836914, + "step": 5585, + "token_acc": 0.9298192711849963 + }, + { + "epoch": 0.36233997731323936, + "grad_norm": 0.8294431567192078, + "learning_rate": 7.562235298687654e-06, + "loss": 0.25209431648254393, + "step": 5590, + "token_acc": 0.928373550124275 + }, + { + "epoch": 0.36266407389402044, + "grad_norm": 0.8896099328994751, + "learning_rate": 7.557632052670485e-06, + "loss": 0.2517201900482178, + "step": 5595, + "token_acc": 0.9291328276845956 + }, + { + "epoch": 0.3629881704748015, + "grad_norm": 0.7957136034965515, + "learning_rate": 7.553025868686637e-06, + "loss": 0.2476658821105957, + "step": 5600, + "token_acc": 0.9290897566906828 + }, + { + "epoch": 0.36331226705558256, + "grad_norm": 0.8148760199546814, + "learning_rate": 7.548416752027261e-06, + "loss": 0.24758920669555665, + "step": 5605, + "token_acc": 0.9298057272938924 + }, + { + "epoch": 0.36363636363636365, + "grad_norm": 0.8497121930122375, + "learning_rate": 7.543804707986877e-06, + "loss": 0.24784598350524903, + "step": 5610, + "token_acc": 0.9297942872712411 + }, + { + "epoch": 0.36396046021714473, + "grad_norm": 0.8171190023422241, + "learning_rate": 7.539189741863365e-06, + "loss": 0.24795398712158204, + "step": 5615, + "token_acc": 0.9301607204665531 + }, + { + "epoch": 0.36428455679792576, + "grad_norm": 0.8167567849159241, + "learning_rate": 7.534571858957964e-06, + "loss": 0.249265718460083, + "step": 5620, + "token_acc": 0.9290492053669006 + }, + { + "epoch": 0.36460865337870685, + "grad_norm": 1.2014907598495483, + "learning_rate": 7.529951064575263e-06, + "loss": 0.2517377853393555, + "step": 5625, + "token_acc": 0.9291127411446932 + }, + { + "epoch": 0.36493274995948793, + "grad_norm": 0.8801676630973816, + "learning_rate": 7.525327364023193e-06, + "loss": 0.2550851583480835, + "step": 5630, + "token_acc": 0.9282969556610731 + }, + { + "epoch": 0.365256846540269, + "grad_norm": 0.8066464066505432, + "learning_rate": 7.520700762613027e-06, + "loss": 0.2507494926452637, + "step": 5635, + "token_acc": 0.9285508045252907 + }, + { + "epoch": 0.36558094312105005, + "grad_norm": 0.7963135838508606, + "learning_rate": 7.51607126565937e-06, + "loss": 0.24458787441253663, + "step": 5640, + "token_acc": 0.9300442872083564 + }, + { + "epoch": 0.36590503970183114, + "grad_norm": 0.9268953204154968, + "learning_rate": 7.511438878480149e-06, + "loss": 0.25169053077697756, + "step": 5645, + "token_acc": 0.9296430919144658 + }, + { + "epoch": 0.3662291362826122, + "grad_norm": 0.8366501331329346, + "learning_rate": 7.506803606396615e-06, + "loss": 0.24937992095947265, + "step": 5650, + "token_acc": 0.9288642630113909 + }, + { + "epoch": 0.3665532328633933, + "grad_norm": 0.8187414407730103, + "learning_rate": 7.5021654547333325e-06, + "loss": 0.2612889766693115, + "step": 5655, + "token_acc": 0.9258605299678787 + }, + { + "epoch": 0.36687732944417434, + "grad_norm": 0.8096038699150085, + "learning_rate": 7.4975244288181736e-06, + "loss": 0.24756073951721191, + "step": 5660, + "token_acc": 0.9290026469169286 + }, + { + "epoch": 0.3672014260249554, + "grad_norm": 0.9664244651794434, + "learning_rate": 7.4928805339823095e-06, + "loss": 0.24379749298095704, + "step": 5665, + "token_acc": 0.9308731262961699 + }, + { + "epoch": 0.3675255226057365, + "grad_norm": 0.8128650188446045, + "learning_rate": 7.488233775560211e-06, + "loss": 0.24985289573669434, + "step": 5670, + "token_acc": 0.9290800509088536 + }, + { + "epoch": 0.3678496191865176, + "grad_norm": 0.8231331706047058, + "learning_rate": 7.483584158889634e-06, + "loss": 0.2451352596282959, + "step": 5675, + "token_acc": 0.9304782409113538 + }, + { + "epoch": 0.3681737157672987, + "grad_norm": 0.8120703101158142, + "learning_rate": 7.478931689311622e-06, + "loss": 0.24784393310546876, + "step": 5680, + "token_acc": 0.9294909494870287 + }, + { + "epoch": 0.3684978123480797, + "grad_norm": 0.8960381150245667, + "learning_rate": 7.4742763721704925e-06, + "loss": 0.248191499710083, + "step": 5685, + "token_acc": 0.9306251717684627 + }, + { + "epoch": 0.3688219089288608, + "grad_norm": 0.8736811280250549, + "learning_rate": 7.469618212813835e-06, + "loss": 0.25290207862854003, + "step": 5690, + "token_acc": 0.9291139406840909 + }, + { + "epoch": 0.3691460055096419, + "grad_norm": 0.7868461608886719, + "learning_rate": 7.464957216592506e-06, + "loss": 0.2468033790588379, + "step": 5695, + "token_acc": 0.9294153602059992 + }, + { + "epoch": 0.36947010209042297, + "grad_norm": 0.795536994934082, + "learning_rate": 7.460293388860616e-06, + "loss": 0.2489880084991455, + "step": 5700, + "token_acc": 0.929502304917801 + }, + { + "epoch": 0.369794198671204, + "grad_norm": 0.9006305932998657, + "learning_rate": 7.455626734975532e-06, + "loss": 0.25122604370117185, + "step": 5705, + "token_acc": 0.9285560069873795 + }, + { + "epoch": 0.3701182952519851, + "grad_norm": 0.8214429020881653, + "learning_rate": 7.4509572602978675e-06, + "loss": 0.24657235145568848, + "step": 5710, + "token_acc": 0.9295389825290541 + }, + { + "epoch": 0.37044239183276617, + "grad_norm": 0.901337206363678, + "learning_rate": 7.446284970191473e-06, + "loss": 0.23954389095306397, + "step": 5715, + "token_acc": 0.9313016427868029 + }, + { + "epoch": 0.37076648841354726, + "grad_norm": 0.7922210097312927, + "learning_rate": 7.441609870023435e-06, + "loss": 0.2529605865478516, + "step": 5720, + "token_acc": 0.928476088159651 + }, + { + "epoch": 0.3710905849943283, + "grad_norm": 0.8337340354919434, + "learning_rate": 7.436931965164069e-06, + "loss": 0.24810476303100587, + "step": 5725, + "token_acc": 0.9304056907136129 + }, + { + "epoch": 0.3714146815751094, + "grad_norm": 0.8533103466033936, + "learning_rate": 7.43225126098691e-06, + "loss": 0.25355119705200196, + "step": 5730, + "token_acc": 0.9289379152079381 + }, + { + "epoch": 0.37173877815589046, + "grad_norm": 0.9082756042480469, + "learning_rate": 7.427567762868714e-06, + "loss": 0.25278534889221194, + "step": 5735, + "token_acc": 0.9292338907615276 + }, + { + "epoch": 0.37206287473667154, + "grad_norm": 0.925537645816803, + "learning_rate": 7.422881476189436e-06, + "loss": 0.25462296009063723, + "step": 5740, + "token_acc": 0.9277285751497786 + }, + { + "epoch": 0.3723869713174526, + "grad_norm": 0.8708488345146179, + "learning_rate": 7.4181924063322455e-06, + "loss": 0.24743733406066895, + "step": 5745, + "token_acc": 0.9298096400560951 + }, + { + "epoch": 0.37271106789823366, + "grad_norm": 0.8439504504203796, + "learning_rate": 7.4135005586835005e-06, + "loss": 0.25099031925201415, + "step": 5750, + "token_acc": 0.929031715428113 + }, + { + "epoch": 0.37303516447901475, + "grad_norm": 0.8168339729309082, + "learning_rate": 7.408805938632757e-06, + "loss": 0.2507676124572754, + "step": 5755, + "token_acc": 0.9288786431747077 + }, + { + "epoch": 0.37335926105979583, + "grad_norm": 0.7949603796005249, + "learning_rate": 7.404108551572749e-06, + "loss": 0.24744062423706054, + "step": 5760, + "token_acc": 0.9304642216619448 + }, + { + "epoch": 0.3736833576405769, + "grad_norm": 0.8355185985565186, + "learning_rate": 7.399408402899394e-06, + "loss": 0.24788496494293213, + "step": 5765, + "token_acc": 0.9306345561759641 + }, + { + "epoch": 0.37400745422135795, + "grad_norm": 0.8773040175437927, + "learning_rate": 7.394705498011778e-06, + "loss": 0.24866197109222413, + "step": 5770, + "token_acc": 0.929516252177942 + }, + { + "epoch": 0.37433155080213903, + "grad_norm": 0.8648874759674072, + "learning_rate": 7.3899998423121575e-06, + "loss": 0.25167207717895507, + "step": 5775, + "token_acc": 0.9284107630178454 + }, + { + "epoch": 0.3746556473829201, + "grad_norm": 0.8743617534637451, + "learning_rate": 7.385291441205945e-06, + "loss": 0.2416687250137329, + "step": 5780, + "token_acc": 0.9306152531649697 + }, + { + "epoch": 0.3749797439637012, + "grad_norm": 0.8802637457847595, + "learning_rate": 7.380580300101708e-06, + "loss": 0.25122091770172117, + "step": 5785, + "token_acc": 0.927914676895763 + }, + { + "epoch": 0.37530384054448224, + "grad_norm": 0.817274272441864, + "learning_rate": 7.375866424411161e-06, + "loss": 0.2491596221923828, + "step": 5790, + "token_acc": 0.9292833164944758 + }, + { + "epoch": 0.3756279371252633, + "grad_norm": 0.8446117043495178, + "learning_rate": 7.371149819549161e-06, + "loss": 0.24905223846435548, + "step": 5795, + "token_acc": 0.9293505173447314 + }, + { + "epoch": 0.3759520337060444, + "grad_norm": 0.911071240901947, + "learning_rate": 7.366430490933698e-06, + "loss": 0.25403106212615967, + "step": 5800, + "token_acc": 0.92970559047304 + }, + { + "epoch": 0.3762761302868255, + "grad_norm": 0.7907551527023315, + "learning_rate": 7.361708443985893e-06, + "loss": 0.2491678476333618, + "step": 5805, + "token_acc": 0.9296901898878084 + }, + { + "epoch": 0.3766002268676065, + "grad_norm": 0.7458071112632751, + "learning_rate": 7.3569836841299905e-06, + "loss": 0.24875168800354003, + "step": 5810, + "token_acc": 0.9297017690340761 + }, + { + "epoch": 0.3769243234483876, + "grad_norm": 0.8069392442703247, + "learning_rate": 7.352256216793343e-06, + "loss": 0.2534111499786377, + "step": 5815, + "token_acc": 0.9290031597855017 + }, + { + "epoch": 0.3772484200291687, + "grad_norm": 0.8453103303909302, + "learning_rate": 7.3475260474064256e-06, + "loss": 0.2519563674926758, + "step": 5820, + "token_acc": 0.9288014172474881 + }, + { + "epoch": 0.3775725166099498, + "grad_norm": 0.8347076773643494, + "learning_rate": 7.34279318140281e-06, + "loss": 0.24747095108032227, + "step": 5825, + "token_acc": 0.9301468913316542 + }, + { + "epoch": 0.3778966131907308, + "grad_norm": 0.80881667137146, + "learning_rate": 7.338057624219166e-06, + "loss": 0.257477331161499, + "step": 5830, + "token_acc": 0.9284658376233516 + }, + { + "epoch": 0.3782207097715119, + "grad_norm": 0.7803962826728821, + "learning_rate": 7.333319381295254e-06, + "loss": 0.2449647903442383, + "step": 5835, + "token_acc": 0.9305486836280306 + }, + { + "epoch": 0.378544806352293, + "grad_norm": 0.8698723912239075, + "learning_rate": 7.328578458073923e-06, + "loss": 0.24910480976104737, + "step": 5840, + "token_acc": 0.9291339099823327 + }, + { + "epoch": 0.37886890293307407, + "grad_norm": 0.802711546421051, + "learning_rate": 7.323834860001098e-06, + "loss": 0.24600112438201904, + "step": 5845, + "token_acc": 0.9299426657218322 + }, + { + "epoch": 0.37919299951385516, + "grad_norm": 0.8455180525779724, + "learning_rate": 7.319088592525777e-06, + "loss": 0.24964759349822999, + "step": 5850, + "token_acc": 0.9301407204522832 + }, + { + "epoch": 0.3795170960946362, + "grad_norm": 0.7447041273117065, + "learning_rate": 7.314339661100024e-06, + "loss": 0.24864084720611573, + "step": 5855, + "token_acc": 0.9294635244808039 + }, + { + "epoch": 0.37984119267541727, + "grad_norm": 0.8902390599250793, + "learning_rate": 7.309588071178968e-06, + "loss": 0.24960193634033204, + "step": 5860, + "token_acc": 0.929395510856185 + }, + { + "epoch": 0.38016528925619836, + "grad_norm": 0.8307676911354065, + "learning_rate": 7.304833828220783e-06, + "loss": 0.24532790184020997, + "step": 5865, + "token_acc": 0.9302209291753796 + }, + { + "epoch": 0.38048938583697944, + "grad_norm": 0.8505387902259827, + "learning_rate": 7.3000769376866975e-06, + "loss": 0.24723994731903076, + "step": 5870, + "token_acc": 0.9300405590465021 + }, + { + "epoch": 0.3808134824177605, + "grad_norm": 0.7767970561981201, + "learning_rate": 7.295317405040978e-06, + "loss": 0.2445511817932129, + "step": 5875, + "token_acc": 0.9309734858584742 + }, + { + "epoch": 0.38113757899854156, + "grad_norm": 0.8741298913955688, + "learning_rate": 7.290555235750929e-06, + "loss": 0.24777607917785643, + "step": 5880, + "token_acc": 0.9299743974345712 + }, + { + "epoch": 0.38146167557932265, + "grad_norm": 0.8490284085273743, + "learning_rate": 7.285790435286881e-06, + "loss": 0.24147202968597412, + "step": 5885, + "token_acc": 0.9316492938272228 + }, + { + "epoch": 0.38178577216010373, + "grad_norm": 0.8055647015571594, + "learning_rate": 7.281023009122186e-06, + "loss": 0.24822120666503905, + "step": 5890, + "token_acc": 0.9298401478551279 + }, + { + "epoch": 0.38210986874088476, + "grad_norm": 1.1799010038375854, + "learning_rate": 7.276252962733216e-06, + "loss": 0.24888854026794432, + "step": 5895, + "token_acc": 0.9296221912127179 + }, + { + "epoch": 0.38243396532166585, + "grad_norm": 0.8989364504814148, + "learning_rate": 7.27148030159935e-06, + "loss": 0.2550069570541382, + "step": 5900, + "token_acc": 0.9283967391304347 + }, + { + "epoch": 0.38275806190244693, + "grad_norm": 0.8483166098594666, + "learning_rate": 7.2667050312029715e-06, + "loss": 0.24626591205596923, + "step": 5905, + "token_acc": 0.9297844462143205 + }, + { + "epoch": 0.383082158483228, + "grad_norm": 0.8166069388389587, + "learning_rate": 7.2619271570294615e-06, + "loss": 0.2533806562423706, + "step": 5910, + "token_acc": 0.9283036764180423 + }, + { + "epoch": 0.38340625506400905, + "grad_norm": 0.7880768775939941, + "learning_rate": 7.257146684567191e-06, + "loss": 0.2501368522644043, + "step": 5915, + "token_acc": 0.9301680461821988 + }, + { + "epoch": 0.38373035164479014, + "grad_norm": 0.8067118525505066, + "learning_rate": 7.252363619307515e-06, + "loss": 0.2533315420150757, + "step": 5920, + "token_acc": 0.9281309129421186 + }, + { + "epoch": 0.3840544482255712, + "grad_norm": 0.9428396821022034, + "learning_rate": 7.247577966744771e-06, + "loss": 0.2566884756088257, + "step": 5925, + "token_acc": 0.9277419605406245 + }, + { + "epoch": 0.3843785448063523, + "grad_norm": 0.7594807147979736, + "learning_rate": 7.242789732376264e-06, + "loss": 0.2503016471862793, + "step": 5930, + "token_acc": 0.92980950523373 + }, + { + "epoch": 0.3847026413871334, + "grad_norm": 0.7880812287330627, + "learning_rate": 7.237998921702267e-06, + "loss": 0.2509206771850586, + "step": 5935, + "token_acc": 0.9284875145217925 + }, + { + "epoch": 0.3850267379679144, + "grad_norm": 0.8450835347175598, + "learning_rate": 7.233205540226009e-06, + "loss": 0.23751478195190429, + "step": 5940, + "token_acc": 0.9320899465722058 + }, + { + "epoch": 0.3853508345486955, + "grad_norm": 0.7689372301101685, + "learning_rate": 7.2284095934536815e-06, + "loss": 0.2496507167816162, + "step": 5945, + "token_acc": 0.9295028340080972 + }, + { + "epoch": 0.3856749311294766, + "grad_norm": 1.1915459632873535, + "learning_rate": 7.223611086894409e-06, + "loss": 0.25453155040740966, + "step": 5950, + "token_acc": 0.9281481625083199 + }, + { + "epoch": 0.3859990277102577, + "grad_norm": 0.8655173182487488, + "learning_rate": 7.218810026060268e-06, + "loss": 0.2535928010940552, + "step": 5955, + "token_acc": 0.929066225122729 + }, + { + "epoch": 0.3863231242910387, + "grad_norm": 0.8232638835906982, + "learning_rate": 7.2140064164662614e-06, + "loss": 0.2492776870727539, + "step": 5960, + "token_acc": 0.9294874685183158 + }, + { + "epoch": 0.3866472208718198, + "grad_norm": 0.9562768340110779, + "learning_rate": 7.209200263630327e-06, + "loss": 0.2505103588104248, + "step": 5965, + "token_acc": 0.9296851652135497 + }, + { + "epoch": 0.3869713174526009, + "grad_norm": 0.8652516603469849, + "learning_rate": 7.204391573073318e-06, + "loss": 0.24292435646057128, + "step": 5970, + "token_acc": 0.9305044521898825 + }, + { + "epoch": 0.38729541403338197, + "grad_norm": 0.8117918372154236, + "learning_rate": 7.199580350319008e-06, + "loss": 0.25945172309875486, + "step": 5975, + "token_acc": 0.9272641740452899 + }, + { + "epoch": 0.387619510614163, + "grad_norm": 0.8763304948806763, + "learning_rate": 7.19476660089407e-06, + "loss": 0.25110797882080077, + "step": 5980, + "token_acc": 0.9289415537435717 + }, + { + "epoch": 0.3879436071949441, + "grad_norm": 0.8209167718887329, + "learning_rate": 7.189950330328093e-06, + "loss": 0.24779777526855468, + "step": 5985, + "token_acc": 0.9308405286799829 + }, + { + "epoch": 0.38826770377572517, + "grad_norm": 0.8319463133811951, + "learning_rate": 7.185131544153548e-06, + "loss": 0.25665993690490724, + "step": 5990, + "token_acc": 0.9272942905392341 + }, + { + "epoch": 0.38859180035650626, + "grad_norm": 0.7923457622528076, + "learning_rate": 7.180310247905807e-06, + "loss": 0.24477794170379638, + "step": 5995, + "token_acc": 0.9302908596229837 + }, + { + "epoch": 0.3889158969372873, + "grad_norm": 0.8393254280090332, + "learning_rate": 7.17548644712312e-06, + "loss": 0.25057382583618165, + "step": 6000, + "token_acc": 0.9288680091229848 + }, + { + "epoch": 0.3889158969372873, + "eval_loss": 0.25131648778915405, + "eval_runtime": 134.4394, + "eval_samples_per_second": 148.766, + "eval_steps_per_second": 0.201, + "eval_token_acc": 0.928378514032228, + "step": 6000 + }, + { + "epoch": 0.3892399935180684, + "grad_norm": 0.8427085280418396, + "learning_rate": 7.1706601473466145e-06, + "loss": 0.24825241565704345, + "step": 6005, + "token_acc": 0.9300593667546174 + }, + { + "epoch": 0.38956409009884946, + "grad_norm": 0.8174166083335876, + "learning_rate": 7.165831354120287e-06, + "loss": 0.2487788438796997, + "step": 6010, + "token_acc": 0.9294917402946845 + }, + { + "epoch": 0.38988818667963054, + "grad_norm": 0.8083465695381165, + "learning_rate": 7.161000072991003e-06, + "loss": 0.24853429794311524, + "step": 6015, + "token_acc": 0.9294653014789533 + }, + { + "epoch": 0.39021228326041163, + "grad_norm": 0.7684163451194763, + "learning_rate": 7.156166309508482e-06, + "loss": 0.2502692699432373, + "step": 6020, + "token_acc": 0.9292363901132478 + }, + { + "epoch": 0.39053637984119266, + "grad_norm": 0.8139322996139526, + "learning_rate": 7.151330069225294e-06, + "loss": 0.24837079048156738, + "step": 6025, + "token_acc": 0.9295849522209387 + }, + { + "epoch": 0.39086047642197375, + "grad_norm": 0.8195866346359253, + "learning_rate": 7.14649135769686e-06, + "loss": 0.2473134994506836, + "step": 6030, + "token_acc": 0.9307603589045201 + }, + { + "epoch": 0.39118457300275483, + "grad_norm": 1.4740195274353027, + "learning_rate": 7.141650180481432e-06, + "loss": 0.24711377620697023, + "step": 6035, + "token_acc": 0.9302415976946018 + }, + { + "epoch": 0.3915086695835359, + "grad_norm": 0.8047171235084534, + "learning_rate": 7.1368065431401e-06, + "loss": 0.2506782054901123, + "step": 6040, + "token_acc": 0.9285563570665577 + }, + { + "epoch": 0.39183276616431695, + "grad_norm": 0.846748411655426, + "learning_rate": 7.131960451236778e-06, + "loss": 0.2510914087295532, + "step": 6045, + "token_acc": 0.9294932569703771 + }, + { + "epoch": 0.39215686274509803, + "grad_norm": 0.8135450482368469, + "learning_rate": 7.127111910338199e-06, + "loss": 0.25735769271850584, + "step": 6050, + "token_acc": 0.9271783304489859 + }, + { + "epoch": 0.3924809593258791, + "grad_norm": 0.833382248878479, + "learning_rate": 7.122260926013911e-06, + "loss": 0.24741125106811523, + "step": 6055, + "token_acc": 0.9298429455679148 + }, + { + "epoch": 0.3928050559066602, + "grad_norm": 0.837912380695343, + "learning_rate": 7.117407503836268e-06, + "loss": 0.24092507362365723, + "step": 6060, + "token_acc": 0.9318338067997561 + }, + { + "epoch": 0.39312915248744124, + "grad_norm": 0.9085661768913269, + "learning_rate": 7.112551649380422e-06, + "loss": 0.24745550155639648, + "step": 6065, + "token_acc": 0.9292455276121337 + }, + { + "epoch": 0.3934532490682223, + "grad_norm": 0.8578550219535828, + "learning_rate": 7.1076933682243225e-06, + "loss": 0.24648494720458985, + "step": 6070, + "token_acc": 0.9305644581071358 + }, + { + "epoch": 0.3937773456490034, + "grad_norm": 0.7916164398193359, + "learning_rate": 7.102832665948703e-06, + "loss": 0.24976131916046143, + "step": 6075, + "token_acc": 0.9297053323876459 + }, + { + "epoch": 0.3941014422297845, + "grad_norm": 1.5179264545440674, + "learning_rate": 7.0979695481370835e-06, + "loss": 0.23998498916625977, + "step": 6080, + "token_acc": 0.9313768001681909 + }, + { + "epoch": 0.3944255388105655, + "grad_norm": 0.7588879466056824, + "learning_rate": 7.093104020375754e-06, + "loss": 0.24896540641784667, + "step": 6085, + "token_acc": 0.9290475110923989 + }, + { + "epoch": 0.3947496353913466, + "grad_norm": 0.856115460395813, + "learning_rate": 7.088236088253773e-06, + "loss": 0.2511408567428589, + "step": 6090, + "token_acc": 0.928884366615251 + }, + { + "epoch": 0.3950737319721277, + "grad_norm": 1.0240391492843628, + "learning_rate": 7.083365757362965e-06, + "loss": 0.2496952772140503, + "step": 6095, + "token_acc": 0.9295063369096626 + }, + { + "epoch": 0.3953978285529088, + "grad_norm": 0.7972952723503113, + "learning_rate": 7.078493033297904e-06, + "loss": 0.2559971332550049, + "step": 6100, + "token_acc": 0.9280127006163913 + }, + { + "epoch": 0.39572192513368987, + "grad_norm": 0.7779661417007446, + "learning_rate": 7.073617921655919e-06, + "loss": 0.25266385078430176, + "step": 6105, + "token_acc": 0.9286510519845339 + }, + { + "epoch": 0.3960460217144709, + "grad_norm": 0.7954340577125549, + "learning_rate": 7.068740428037075e-06, + "loss": 0.250183629989624, + "step": 6110, + "token_acc": 0.9287291075370545 + }, + { + "epoch": 0.396370118295252, + "grad_norm": 0.7634496092796326, + "learning_rate": 7.063860558044181e-06, + "loss": 0.2507870435714722, + "step": 6115, + "token_acc": 0.9298716878781557 + }, + { + "epoch": 0.39669421487603307, + "grad_norm": 0.8289607763290405, + "learning_rate": 7.058978317282771e-06, + "loss": 0.24918882846832274, + "step": 6120, + "token_acc": 0.9291013428804074 + }, + { + "epoch": 0.39701831145681415, + "grad_norm": 0.8151379823684692, + "learning_rate": 7.054093711361102e-06, + "loss": 0.2469410181045532, + "step": 6125, + "token_acc": 0.9302538902538903 + }, + { + "epoch": 0.3973424080375952, + "grad_norm": 0.9721442461013794, + "learning_rate": 7.049206745890149e-06, + "loss": 0.24393579959869385, + "step": 6130, + "token_acc": 0.9307932132191263 + }, + { + "epoch": 0.39766650461837627, + "grad_norm": 0.8404655456542969, + "learning_rate": 7.044317426483598e-06, + "loss": 0.24058222770690918, + "step": 6135, + "token_acc": 0.931081742535487 + }, + { + "epoch": 0.39799060119915736, + "grad_norm": 0.8413792252540588, + "learning_rate": 7.039425758757838e-06, + "loss": 0.24167876243591307, + "step": 6140, + "token_acc": 0.9310887778706529 + }, + { + "epoch": 0.39831469777993844, + "grad_norm": 0.8259962797164917, + "learning_rate": 7.034531748331956e-06, + "loss": 0.24562034606933594, + "step": 6145, + "token_acc": 0.9300162601626016 + }, + { + "epoch": 0.3986387943607195, + "grad_norm": 0.7962437868118286, + "learning_rate": 7.02963540082773e-06, + "loss": 0.2529958009719849, + "step": 6150, + "token_acc": 0.9283162140643404 + }, + { + "epoch": 0.39896289094150056, + "grad_norm": 0.7873697876930237, + "learning_rate": 7.024736721869624e-06, + "loss": 0.245263671875, + "step": 6155, + "token_acc": 0.9303382260743671 + }, + { + "epoch": 0.39928698752228164, + "grad_norm": 0.8236318826675415, + "learning_rate": 7.019835717084775e-06, + "loss": 0.25450906753540037, + "step": 6160, + "token_acc": 0.9286381493087069 + }, + { + "epoch": 0.39961108410306273, + "grad_norm": 0.8441376090049744, + "learning_rate": 7.014932392103e-06, + "loss": 0.2424696207046509, + "step": 6165, + "token_acc": 0.9315550917924805 + }, + { + "epoch": 0.39993518068384376, + "grad_norm": 0.8870713114738464, + "learning_rate": 7.010026752556774e-06, + "loss": 0.24054996967315673, + "step": 6170, + "token_acc": 0.9313860952051973 + }, + { + "epoch": 0.40025927726462485, + "grad_norm": 0.789865255355835, + "learning_rate": 7.005118804081233e-06, + "loss": 0.2514184474945068, + "step": 6175, + "token_acc": 0.9289943471103909 + }, + { + "epoch": 0.40058337384540593, + "grad_norm": 0.8248916864395142, + "learning_rate": 7.000208552314166e-06, + "loss": 0.24821736812591552, + "step": 6180, + "token_acc": 0.9288357510889537 + }, + { + "epoch": 0.400907470426187, + "grad_norm": 0.8034648895263672, + "learning_rate": 6.99529600289601e-06, + "loss": 0.24904162883758546, + "step": 6185, + "token_acc": 0.9301058299892431 + }, + { + "epoch": 0.4012315670069681, + "grad_norm": 0.8538951873779297, + "learning_rate": 6.990381161469835e-06, + "loss": 0.24318549633026124, + "step": 6190, + "token_acc": 0.930396528070916 + }, + { + "epoch": 0.40155566358774913, + "grad_norm": 1.014456033706665, + "learning_rate": 6.985464033681353e-06, + "loss": 0.2531020164489746, + "step": 6195, + "token_acc": 0.9286696873879454 + }, + { + "epoch": 0.4018797601685302, + "grad_norm": 0.8683779239654541, + "learning_rate": 6.980544625178891e-06, + "loss": 0.25147645473480223, + "step": 6200, + "token_acc": 0.9286735277301315 + }, + { + "epoch": 0.4022038567493113, + "grad_norm": 0.8594753742218018, + "learning_rate": 6.975622941613405e-06, + "loss": 0.2452561855316162, + "step": 6205, + "token_acc": 0.9308650605746845 + }, + { + "epoch": 0.4025279533300924, + "grad_norm": 0.7687065601348877, + "learning_rate": 6.9706989886384606e-06, + "loss": 0.24630703926086425, + "step": 6210, + "token_acc": 0.9306638242130812 + }, + { + "epoch": 0.4028520499108734, + "grad_norm": 0.8112002611160278, + "learning_rate": 6.965772771910231e-06, + "loss": 0.25576488971710204, + "step": 6215, + "token_acc": 0.9282142504200019 + }, + { + "epoch": 0.4031761464916545, + "grad_norm": 0.7707385420799255, + "learning_rate": 6.9608442970874904e-06, + "loss": 0.247090482711792, + "step": 6220, + "token_acc": 0.9300129231010221 + }, + { + "epoch": 0.4035002430724356, + "grad_norm": 0.7890149354934692, + "learning_rate": 6.955913569831606e-06, + "loss": 0.24798500537872314, + "step": 6225, + "token_acc": 0.9301982106797195 + }, + { + "epoch": 0.4038243396532167, + "grad_norm": 0.779474675655365, + "learning_rate": 6.950980595806532e-06, + "loss": 0.2413931131362915, + "step": 6230, + "token_acc": 0.9317457929744472 + }, + { + "epoch": 0.4041484362339977, + "grad_norm": 0.8524256944656372, + "learning_rate": 6.946045380678803e-06, + "loss": 0.24512686729431152, + "step": 6235, + "token_acc": 0.929696103243755 + }, + { + "epoch": 0.4044725328147788, + "grad_norm": 0.8619547486305237, + "learning_rate": 6.9411079301175335e-06, + "loss": 0.2537808656692505, + "step": 6240, + "token_acc": 0.9286926330078838 + }, + { + "epoch": 0.4047966293955599, + "grad_norm": 0.8410328030586243, + "learning_rate": 6.936168249794396e-06, + "loss": 0.24522767066955567, + "step": 6245, + "token_acc": 0.9300479416160401 + }, + { + "epoch": 0.40512072597634097, + "grad_norm": 0.8448101282119751, + "learning_rate": 6.931226345383634e-06, + "loss": 0.24491569995880128, + "step": 6250, + "token_acc": 0.9305088574102158 + }, + { + "epoch": 0.405444822557122, + "grad_norm": 0.7887387275695801, + "learning_rate": 6.926282222562037e-06, + "loss": 0.24861605167388917, + "step": 6255, + "token_acc": 0.9293052141286066 + }, + { + "epoch": 0.4057689191379031, + "grad_norm": 0.8755037784576416, + "learning_rate": 6.921335887008953e-06, + "loss": 0.25023641586303713, + "step": 6260, + "token_acc": 0.9293841453609906 + }, + { + "epoch": 0.40609301571868417, + "grad_norm": 0.8661155700683594, + "learning_rate": 6.916387344406261e-06, + "loss": 0.2531020164489746, + "step": 6265, + "token_acc": 0.9278149584702063 + }, + { + "epoch": 0.40641711229946526, + "grad_norm": 0.7866302132606506, + "learning_rate": 6.911436600438385e-06, + "loss": 0.24124016761779785, + "step": 6270, + "token_acc": 0.9315049864007253 + }, + { + "epoch": 0.40674120888024634, + "grad_norm": 0.8174487352371216, + "learning_rate": 6.906483660792268e-06, + "loss": 0.2448854923248291, + "step": 6275, + "token_acc": 0.9305721232506143 + }, + { + "epoch": 0.40706530546102737, + "grad_norm": 0.836603581905365, + "learning_rate": 6.9015285311573865e-06, + "loss": 0.2584808349609375, + "step": 6280, + "token_acc": 0.9276651559030168 + }, + { + "epoch": 0.40738940204180846, + "grad_norm": 0.7945948243141174, + "learning_rate": 6.89657121722572e-06, + "loss": 0.2419842004776001, + "step": 6285, + "token_acc": 0.9318030518607474 + }, + { + "epoch": 0.40771349862258954, + "grad_norm": 0.8075172305107117, + "learning_rate": 6.891611724691768e-06, + "loss": 0.24677686691284179, + "step": 6290, + "token_acc": 0.9301219305009213 + }, + { + "epoch": 0.40803759520337063, + "grad_norm": 0.85341876745224, + "learning_rate": 6.886650059252527e-06, + "loss": 0.25059549808502196, + "step": 6295, + "token_acc": 0.9291581379993397 + }, + { + "epoch": 0.40836169178415166, + "grad_norm": 1.4045888185501099, + "learning_rate": 6.88168622660749e-06, + "loss": 0.25671844482421874, + "step": 6300, + "token_acc": 0.9281095126764048 + }, + { + "epoch": 0.40868578836493274, + "grad_norm": 0.8921906352043152, + "learning_rate": 6.8767202324586415e-06, + "loss": 0.24798715114593506, + "step": 6305, + "token_acc": 0.9300713411617674 + }, + { + "epoch": 0.40900988494571383, + "grad_norm": 0.8096488118171692, + "learning_rate": 6.871752082510447e-06, + "loss": 0.25236351490020753, + "step": 6310, + "token_acc": 0.9281867498296557 + }, + { + "epoch": 0.4093339815264949, + "grad_norm": 0.7677403688430786, + "learning_rate": 6.866781782469847e-06, + "loss": 0.2519232749938965, + "step": 6315, + "token_acc": 0.9291794612663739 + }, + { + "epoch": 0.40965807810727595, + "grad_norm": 0.8326215147972107, + "learning_rate": 6.861809338046257e-06, + "loss": 0.2411609411239624, + "step": 6320, + "token_acc": 0.9313886902848747 + }, + { + "epoch": 0.40998217468805703, + "grad_norm": 0.912262499332428, + "learning_rate": 6.8568347549515515e-06, + "loss": 0.24906229972839355, + "step": 6325, + "token_acc": 0.9300536300951837 + }, + { + "epoch": 0.4103062712688381, + "grad_norm": 0.8335503935813904, + "learning_rate": 6.85185803890006e-06, + "loss": 0.2402871370315552, + "step": 6330, + "token_acc": 0.9302122597288403 + }, + { + "epoch": 0.4106303678496192, + "grad_norm": 0.8052857518196106, + "learning_rate": 6.84687919560857e-06, + "loss": 0.24075784683227539, + "step": 6335, + "token_acc": 0.9319713022011932 + }, + { + "epoch": 0.41095446443040023, + "grad_norm": 0.938561201095581, + "learning_rate": 6.841898230796302e-06, + "loss": 0.24760501384735106, + "step": 6340, + "token_acc": 0.9296986048960253 + }, + { + "epoch": 0.4112785610111813, + "grad_norm": 0.8097336888313293, + "learning_rate": 6.836915150184924e-06, + "loss": 0.251514196395874, + "step": 6345, + "token_acc": 0.9284721602427611 + }, + { + "epoch": 0.4116026575919624, + "grad_norm": 0.80184406042099, + "learning_rate": 6.831929959498527e-06, + "loss": 0.24845759868621825, + "step": 6350, + "token_acc": 0.9296243155774885 + }, + { + "epoch": 0.4119267541727435, + "grad_norm": 0.8103557825088501, + "learning_rate": 6.826942664463627e-06, + "loss": 0.24678592681884765, + "step": 6355, + "token_acc": 0.9301489167601908 + }, + { + "epoch": 0.4122508507535245, + "grad_norm": 0.8284949660301208, + "learning_rate": 6.821953270809161e-06, + "loss": 0.24939770698547364, + "step": 6360, + "token_acc": 0.9304687141949888 + }, + { + "epoch": 0.4125749473343056, + "grad_norm": 0.8252038359642029, + "learning_rate": 6.816961784266474e-06, + "loss": 0.24977412223815917, + "step": 6365, + "token_acc": 0.9298505253810863 + }, + { + "epoch": 0.4128990439150867, + "grad_norm": 0.7670993804931641, + "learning_rate": 6.811968210569313e-06, + "loss": 0.2490987777709961, + "step": 6370, + "token_acc": 0.9293177627535341 + }, + { + "epoch": 0.4132231404958678, + "grad_norm": 0.8551302552223206, + "learning_rate": 6.8069725554538266e-06, + "loss": 0.2415782928466797, + "step": 6375, + "token_acc": 0.9309415112894087 + }, + { + "epoch": 0.41354723707664887, + "grad_norm": 0.7688048481941223, + "learning_rate": 6.8019748246585524e-06, + "loss": 0.2406745195388794, + "step": 6380, + "token_acc": 0.9313389425573907 + }, + { + "epoch": 0.4138713336574299, + "grad_norm": 0.817086398601532, + "learning_rate": 6.7969750239244145e-06, + "loss": 0.25042071342468264, + "step": 6385, + "token_acc": 0.9294163133441241 + }, + { + "epoch": 0.414195430238211, + "grad_norm": 0.8368716835975647, + "learning_rate": 6.7919731589947095e-06, + "loss": 0.2538616180419922, + "step": 6390, + "token_acc": 0.9288528042850828 + }, + { + "epoch": 0.41451952681899207, + "grad_norm": 0.816768229007721, + "learning_rate": 6.786969235615111e-06, + "loss": 0.2462073802947998, + "step": 6395, + "token_acc": 0.9295392603046566 + }, + { + "epoch": 0.41484362339977315, + "grad_norm": 0.8432138562202454, + "learning_rate": 6.781963259533653e-06, + "loss": 0.24964299201965331, + "step": 6400, + "token_acc": 0.9290247713142864 + }, + { + "epoch": 0.4151677199805542, + "grad_norm": 0.8623592853546143, + "learning_rate": 6.776955236500728e-06, + "loss": 0.2452172040939331, + "step": 6405, + "token_acc": 0.9297661035224121 + }, + { + "epoch": 0.41549181656133527, + "grad_norm": 0.8474541902542114, + "learning_rate": 6.771945172269086e-06, + "loss": 0.2386030912399292, + "step": 6410, + "token_acc": 0.9322192793832276 + }, + { + "epoch": 0.41581591314211636, + "grad_norm": 1.0228440761566162, + "learning_rate": 6.766933072593809e-06, + "loss": 0.250473165512085, + "step": 6415, + "token_acc": 0.9285704993104884 + }, + { + "epoch": 0.41614000972289744, + "grad_norm": 0.846895158290863, + "learning_rate": 6.761918943232332e-06, + "loss": 0.2434088706970215, + "step": 6420, + "token_acc": 0.9307850704517199 + }, + { + "epoch": 0.41646410630367847, + "grad_norm": 0.7895659804344177, + "learning_rate": 6.756902789944411e-06, + "loss": 0.24429407119750976, + "step": 6425, + "token_acc": 0.9309236272372685 + }, + { + "epoch": 0.41678820288445956, + "grad_norm": 0.9482538104057312, + "learning_rate": 6.751884618492129e-06, + "loss": 0.24467546939849855, + "step": 6430, + "token_acc": 0.930704780680138 + }, + { + "epoch": 0.41711229946524064, + "grad_norm": 0.8409300446510315, + "learning_rate": 6.746864434639891e-06, + "loss": 0.24743320941925048, + "step": 6435, + "token_acc": 0.9292427700422913 + }, + { + "epoch": 0.41743639604602173, + "grad_norm": 0.8351048231124878, + "learning_rate": 6.741842244154411e-06, + "loss": 0.24508042335510255, + "step": 6440, + "token_acc": 0.9298526028138877 + }, + { + "epoch": 0.41776049262680276, + "grad_norm": 0.8143692016601562, + "learning_rate": 6.7368180528047045e-06, + "loss": 0.2419804573059082, + "step": 6445, + "token_acc": 0.9318651021794805 + }, + { + "epoch": 0.41808458920758385, + "grad_norm": 0.7897880673408508, + "learning_rate": 6.731791866362094e-06, + "loss": 0.2422799587249756, + "step": 6450, + "token_acc": 0.931002175223419 + }, + { + "epoch": 0.41840868578836493, + "grad_norm": 0.7665742039680481, + "learning_rate": 6.7267636906001854e-06, + "loss": 0.24817900657653807, + "step": 6455, + "token_acc": 0.9290333960489181 + }, + { + "epoch": 0.418732782369146, + "grad_norm": 0.7542237043380737, + "learning_rate": 6.721733531294876e-06, + "loss": 0.25026507377624513, + "step": 6460, + "token_acc": 0.9297434532196589 + }, + { + "epoch": 0.4190568789499271, + "grad_norm": 0.8287550806999207, + "learning_rate": 6.716701394224335e-06, + "loss": 0.244926118850708, + "step": 6465, + "token_acc": 0.9302594665873745 + }, + { + "epoch": 0.41938097553070813, + "grad_norm": 0.846988320350647, + "learning_rate": 6.711667285169011e-06, + "loss": 0.2485114097595215, + "step": 6470, + "token_acc": 0.9298332756017699 + }, + { + "epoch": 0.4197050721114892, + "grad_norm": 0.7603313326835632, + "learning_rate": 6.706631209911613e-06, + "loss": 0.24062485694885255, + "step": 6475, + "token_acc": 0.931276097888899 + }, + { + "epoch": 0.4200291686922703, + "grad_norm": 0.9658607840538025, + "learning_rate": 6.701593174237108e-06, + "loss": 0.2402059555053711, + "step": 6480, + "token_acc": 0.9319303328784576 + }, + { + "epoch": 0.4203532652730514, + "grad_norm": 0.7766156792640686, + "learning_rate": 6.696553183932717e-06, + "loss": 0.24536514282226562, + "step": 6485, + "token_acc": 0.9298894412894014 + }, + { + "epoch": 0.4206773618538324, + "grad_norm": 0.7984305024147034, + "learning_rate": 6.691511244787908e-06, + "loss": 0.24613780975341798, + "step": 6490, + "token_acc": 0.9296067602548136 + }, + { + "epoch": 0.4210014584346135, + "grad_norm": 0.8978537917137146, + "learning_rate": 6.686467362594384e-06, + "loss": 0.2514453411102295, + "step": 6495, + "token_acc": 0.9289049137891184 + }, + { + "epoch": 0.4213255550153946, + "grad_norm": 0.8040966987609863, + "learning_rate": 6.68142154314608e-06, + "loss": 0.2456225872039795, + "step": 6500, + "token_acc": 0.929896254893948 + }, + { + "epoch": 0.4213255550153946, + "eval_loss": 0.25065386295318604, + "eval_runtime": 108.5168, + "eval_samples_per_second": 184.303, + "eval_steps_per_second": 0.249, + "eval_token_acc": 0.9281275544104362, + "step": 6500 + }, + { + "epoch": 0.4216496515961757, + "grad_norm": 0.8634015917778015, + "learning_rate": 6.676373792239162e-06, + "loss": 0.24506475925445556, + "step": 6505, + "token_acc": 0.9304544412044167 + }, + { + "epoch": 0.4219737481769567, + "grad_norm": 0.9035208821296692, + "learning_rate": 6.671324115672009e-06, + "loss": 0.25476062297821045, + "step": 6510, + "token_acc": 0.9284668493008063 + }, + { + "epoch": 0.4222978447577378, + "grad_norm": 0.9372864365577698, + "learning_rate": 6.66627251924521e-06, + "loss": 0.24822752475738524, + "step": 6515, + "token_acc": 0.9294503745503094 + }, + { + "epoch": 0.4226219413385189, + "grad_norm": 0.755682647228241, + "learning_rate": 6.661219008761568e-06, + "loss": 0.24432203769683838, + "step": 6520, + "token_acc": 0.9306668398155964 + }, + { + "epoch": 0.42294603791929997, + "grad_norm": 0.8265901207923889, + "learning_rate": 6.656163590026077e-06, + "loss": 0.24868192672729492, + "step": 6525, + "token_acc": 0.9289532780858125 + }, + { + "epoch": 0.423270134500081, + "grad_norm": 0.7345836162567139, + "learning_rate": 6.651106268845925e-06, + "loss": 0.24633252620697021, + "step": 6530, + "token_acc": 0.9300361175283897 + }, + { + "epoch": 0.4235942310808621, + "grad_norm": 0.8827416300773621, + "learning_rate": 6.6460470510304865e-06, + "loss": 0.24641914367675782, + "step": 6535, + "token_acc": 0.9303573749187785 + }, + { + "epoch": 0.42391832766164317, + "grad_norm": 0.8788570761680603, + "learning_rate": 6.6409859423913145e-06, + "loss": 0.2445530652999878, + "step": 6540, + "token_acc": 0.9298858698497461 + }, + { + "epoch": 0.42424242424242425, + "grad_norm": 0.838145911693573, + "learning_rate": 6.6359229487421365e-06, + "loss": 0.25053863525390624, + "step": 6545, + "token_acc": 0.9290646334584638 + }, + { + "epoch": 0.42456652082320534, + "grad_norm": 0.8223479986190796, + "learning_rate": 6.630858075898836e-06, + "loss": 0.24428014755249022, + "step": 6550, + "token_acc": 0.9304644976733049 + }, + { + "epoch": 0.42489061740398637, + "grad_norm": 0.822411835193634, + "learning_rate": 6.625791329679467e-06, + "loss": 0.24856266975402833, + "step": 6555, + "token_acc": 0.9294042081086867 + }, + { + "epoch": 0.42521471398476746, + "grad_norm": 0.7726208567619324, + "learning_rate": 6.62072271590423e-06, + "loss": 0.24491300582885742, + "step": 6560, + "token_acc": 0.9314692658593416 + }, + { + "epoch": 0.42553881056554854, + "grad_norm": 0.8025507926940918, + "learning_rate": 6.615652240395468e-06, + "loss": 0.25168623924255373, + "step": 6565, + "token_acc": 0.9290414841368957 + }, + { + "epoch": 0.42586290714632963, + "grad_norm": 0.8266592621803284, + "learning_rate": 6.610579908977666e-06, + "loss": 0.24751648902893067, + "step": 6570, + "token_acc": 0.929554844642464 + }, + { + "epoch": 0.42618700372711066, + "grad_norm": 0.7941948175430298, + "learning_rate": 6.605505727477443e-06, + "loss": 0.25345375537872317, + "step": 6575, + "token_acc": 0.9285220949476785 + }, + { + "epoch": 0.42651110030789174, + "grad_norm": 0.8837137818336487, + "learning_rate": 6.600429701723536e-06, + "loss": 0.24606151580810548, + "step": 6580, + "token_acc": 0.9299154944362796 + }, + { + "epoch": 0.42683519688867283, + "grad_norm": 0.7955554723739624, + "learning_rate": 6.595351837546811e-06, + "loss": 0.24320287704467775, + "step": 6585, + "token_acc": 0.9313969899621437 + }, + { + "epoch": 0.4271592934694539, + "grad_norm": 0.8251743912696838, + "learning_rate": 6.590272140780233e-06, + "loss": 0.24558377265930176, + "step": 6590, + "token_acc": 0.9299231533274086 + }, + { + "epoch": 0.42748339005023495, + "grad_norm": 0.8760346174240112, + "learning_rate": 6.585190617258884e-06, + "loss": 0.24608831405639647, + "step": 6595, + "token_acc": 0.9305453086371814 + }, + { + "epoch": 0.42780748663101603, + "grad_norm": 0.8176361918449402, + "learning_rate": 6.580107272819938e-06, + "loss": 0.2506168127059937, + "step": 6600, + "token_acc": 0.9296813395797578 + }, + { + "epoch": 0.4281315832117971, + "grad_norm": 0.800468385219574, + "learning_rate": 6.575022113302662e-06, + "loss": 0.25252861976623536, + "step": 6605, + "token_acc": 0.9292283557211491 + }, + { + "epoch": 0.4284556797925782, + "grad_norm": 0.8433128595352173, + "learning_rate": 6.569935144548407e-06, + "loss": 0.2553386688232422, + "step": 6610, + "token_acc": 0.9284312255052202 + }, + { + "epoch": 0.42877977637335923, + "grad_norm": 0.7839466333389282, + "learning_rate": 6.564846372400604e-06, + "loss": 0.24884500503540039, + "step": 6615, + "token_acc": 0.9297051563446421 + }, + { + "epoch": 0.4291038729541403, + "grad_norm": 0.8041486740112305, + "learning_rate": 6.559755802704757e-06, + "loss": 0.24316997528076173, + "step": 6620, + "token_acc": 0.9310730807803047 + }, + { + "epoch": 0.4294279695349214, + "grad_norm": 0.7820088267326355, + "learning_rate": 6.554663441308428e-06, + "loss": 0.24194040298461914, + "step": 6625, + "token_acc": 0.9310885705322598 + }, + { + "epoch": 0.4297520661157025, + "grad_norm": 0.8956055641174316, + "learning_rate": 6.549569294061248e-06, + "loss": 0.24583797454833983, + "step": 6630, + "token_acc": 0.9299669213043966 + }, + { + "epoch": 0.4300761626964836, + "grad_norm": 0.8159700632095337, + "learning_rate": 6.544473366814887e-06, + "loss": 0.2461874008178711, + "step": 6635, + "token_acc": 0.9293364072368715 + }, + { + "epoch": 0.4304002592772646, + "grad_norm": 0.7954070568084717, + "learning_rate": 6.53937566542307e-06, + "loss": 0.24941813945770264, + "step": 6640, + "token_acc": 0.9299861874016254 + }, + { + "epoch": 0.4307243558580457, + "grad_norm": 0.9013829827308655, + "learning_rate": 6.534276195741556e-06, + "loss": 0.25430951118469236, + "step": 6645, + "token_acc": 0.9278059785673999 + }, + { + "epoch": 0.4310484524388268, + "grad_norm": 0.8831157088279724, + "learning_rate": 6.529174963628133e-06, + "loss": 0.24619996547698975, + "step": 6650, + "token_acc": 0.9301673772303806 + }, + { + "epoch": 0.43137254901960786, + "grad_norm": 0.877896785736084, + "learning_rate": 6.524071974942618e-06, + "loss": 0.25113663673400877, + "step": 6655, + "token_acc": 0.9293884906177773 + }, + { + "epoch": 0.4316966456003889, + "grad_norm": 0.7848274111747742, + "learning_rate": 6.5189672355468415e-06, + "loss": 0.24569523334503174, + "step": 6660, + "token_acc": 0.9306485486493782 + }, + { + "epoch": 0.43202074218117, + "grad_norm": 0.8189888000488281, + "learning_rate": 6.513860751304648e-06, + "loss": 0.2422551155090332, + "step": 6665, + "token_acc": 0.9313376458433413 + }, + { + "epoch": 0.43234483876195107, + "grad_norm": 0.815414309501648, + "learning_rate": 6.508752528081886e-06, + "loss": 0.24816856384277344, + "step": 6670, + "token_acc": 0.9298346408854373 + }, + { + "epoch": 0.43266893534273215, + "grad_norm": 0.8267301321029663, + "learning_rate": 6.503642571746401e-06, + "loss": 0.24688267707824707, + "step": 6675, + "token_acc": 0.9292447591734878 + }, + { + "epoch": 0.4329930319235132, + "grad_norm": 0.8032580018043518, + "learning_rate": 6.498530888168027e-06, + "loss": 0.24493839740753173, + "step": 6680, + "token_acc": 0.9307759743189759 + }, + { + "epoch": 0.43331712850429427, + "grad_norm": 0.7993561625480652, + "learning_rate": 6.493417483218586e-06, + "loss": 0.24506628513336182, + "step": 6685, + "token_acc": 0.9307712333224861 + }, + { + "epoch": 0.43364122508507535, + "grad_norm": 0.8500654101371765, + "learning_rate": 6.488302362771877e-06, + "loss": 0.2509969711303711, + "step": 6690, + "token_acc": 0.9280468938686789 + }, + { + "epoch": 0.43396532166585644, + "grad_norm": 0.844978928565979, + "learning_rate": 6.483185532703666e-06, + "loss": 0.2540831327438354, + "step": 6695, + "token_acc": 0.928517761124795 + }, + { + "epoch": 0.43428941824663747, + "grad_norm": 0.8010037541389465, + "learning_rate": 6.478066998891687e-06, + "loss": 0.2489070177078247, + "step": 6700, + "token_acc": 0.9298651575310679 + }, + { + "epoch": 0.43461351482741856, + "grad_norm": 0.9018349647521973, + "learning_rate": 6.472946767215628e-06, + "loss": 0.24423112869262695, + "step": 6705, + "token_acc": 0.9311181804353146 + }, + { + "epoch": 0.43493761140819964, + "grad_norm": 0.7697138786315918, + "learning_rate": 6.467824843557129e-06, + "loss": 0.24227299690246581, + "step": 6710, + "token_acc": 0.9312001234655229 + }, + { + "epoch": 0.43526170798898073, + "grad_norm": 0.7765949964523315, + "learning_rate": 6.462701233799774e-06, + "loss": 0.24548327922821045, + "step": 6715, + "token_acc": 0.9296979887641914 + }, + { + "epoch": 0.4355858045697618, + "grad_norm": 0.9074963331222534, + "learning_rate": 6.457575943829082e-06, + "loss": 0.24725427627563476, + "step": 6720, + "token_acc": 0.9301477055791922 + }, + { + "epoch": 0.43590990115054284, + "grad_norm": 0.8736677169799805, + "learning_rate": 6.452448979532504e-06, + "loss": 0.23506083488464355, + "step": 6725, + "token_acc": 0.9317259629520729 + }, + { + "epoch": 0.43623399773132393, + "grad_norm": 0.8097268342971802, + "learning_rate": 6.447320346799413e-06, + "loss": 0.25235648155212403, + "step": 6730, + "token_acc": 0.9289199291756837 + }, + { + "epoch": 0.436558094312105, + "grad_norm": 0.7962629199028015, + "learning_rate": 6.4421900515210976e-06, + "loss": 0.24368133544921874, + "step": 6735, + "token_acc": 0.9299814121527868 + }, + { + "epoch": 0.4368821908928861, + "grad_norm": 0.8999195694923401, + "learning_rate": 6.437058099590761e-06, + "loss": 0.2428037166595459, + "step": 6740, + "token_acc": 0.9314099216710183 + }, + { + "epoch": 0.43720628747366713, + "grad_norm": 0.8035469651222229, + "learning_rate": 6.431924496903503e-06, + "loss": 0.24131884574890136, + "step": 6745, + "token_acc": 0.931392081062895 + }, + { + "epoch": 0.4375303840544482, + "grad_norm": 0.8522917628288269, + "learning_rate": 6.4267892493563225e-06, + "loss": 0.2463465690612793, + "step": 6750, + "token_acc": 0.929873250749329 + }, + { + "epoch": 0.4378544806352293, + "grad_norm": 0.7927606105804443, + "learning_rate": 6.421652362848111e-06, + "loss": 0.2525139331817627, + "step": 6755, + "token_acc": 0.9289588903095819 + }, + { + "epoch": 0.4381785772160104, + "grad_norm": 0.7988627552986145, + "learning_rate": 6.416513843279635e-06, + "loss": 0.24797518253326417, + "step": 6760, + "token_acc": 0.9301948266082758 + }, + { + "epoch": 0.4385026737967914, + "grad_norm": 0.8298184871673584, + "learning_rate": 6.411373696553546e-06, + "loss": 0.2446286678314209, + "step": 6765, + "token_acc": 0.9301514222738122 + }, + { + "epoch": 0.4388267703775725, + "grad_norm": 0.7754964232444763, + "learning_rate": 6.406231928574358e-06, + "loss": 0.2454598665237427, + "step": 6770, + "token_acc": 0.9303206540873942 + }, + { + "epoch": 0.4391508669583536, + "grad_norm": 0.7787110805511475, + "learning_rate": 6.40108854524845e-06, + "loss": 0.2484816551208496, + "step": 6775, + "token_acc": 0.9293973519912475 + }, + { + "epoch": 0.4394749635391347, + "grad_norm": 0.745827853679657, + "learning_rate": 6.395943552484054e-06, + "loss": 0.23671541213989258, + "step": 6780, + "token_acc": 0.9329808657321464 + }, + { + "epoch": 0.4397990601199157, + "grad_norm": 0.7915314435958862, + "learning_rate": 6.390796956191258e-06, + "loss": 0.24257347583770753, + "step": 6785, + "token_acc": 0.9310108253237372 + }, + { + "epoch": 0.4401231567006968, + "grad_norm": 0.813419759273529, + "learning_rate": 6.38564876228198e-06, + "loss": 0.2489013671875, + "step": 6790, + "token_acc": 0.9300118156754628 + }, + { + "epoch": 0.4404472532814779, + "grad_norm": 0.8088440299034119, + "learning_rate": 6.380498976669987e-06, + "loss": 0.25125837326049805, + "step": 6795, + "token_acc": 0.92869431603278 + }, + { + "epoch": 0.44077134986225897, + "grad_norm": 0.8884025812149048, + "learning_rate": 6.37534760527086e-06, + "loss": 0.24603140354156494, + "step": 6800, + "token_acc": 0.9303325009856749 + }, + { + "epoch": 0.44109544644304005, + "grad_norm": 0.9571425914764404, + "learning_rate": 6.370194654002016e-06, + "loss": 0.24786901473999023, + "step": 6805, + "token_acc": 0.9298900210042922 + }, + { + "epoch": 0.4414195430238211, + "grad_norm": 0.9139158129692078, + "learning_rate": 6.365040128782676e-06, + "loss": 0.24400038719177247, + "step": 6810, + "token_acc": 0.9303450626671561 + }, + { + "epoch": 0.44174363960460217, + "grad_norm": 0.8593194484710693, + "learning_rate": 6.3598840355338735e-06, + "loss": 0.23929789066314697, + "step": 6815, + "token_acc": 0.9315760065735414 + }, + { + "epoch": 0.44206773618538325, + "grad_norm": 0.8248846530914307, + "learning_rate": 6.354726380178442e-06, + "loss": 0.2452117919921875, + "step": 6820, + "token_acc": 0.9301444910774357 + }, + { + "epoch": 0.44239183276616434, + "grad_norm": 1.0394436120986938, + "learning_rate": 6.349567168641011e-06, + "loss": 0.24956960678100587, + "step": 6825, + "token_acc": 0.9285117701480263 + }, + { + "epoch": 0.44271592934694537, + "grad_norm": 0.8519606590270996, + "learning_rate": 6.344406406847996e-06, + "loss": 0.2468410015106201, + "step": 6830, + "token_acc": 0.929284793814433 + }, + { + "epoch": 0.44304002592772646, + "grad_norm": 0.8163942694664001, + "learning_rate": 6.3392441007275925e-06, + "loss": 0.24638650417327881, + "step": 6835, + "token_acc": 0.9300933513860187 + }, + { + "epoch": 0.44336412250850754, + "grad_norm": 0.8128418922424316, + "learning_rate": 6.334080256209776e-06, + "loss": 0.25072493553161623, + "step": 6840, + "token_acc": 0.9289994611334377 + }, + { + "epoch": 0.4436882190892886, + "grad_norm": 1.1899977922439575, + "learning_rate": 6.328914879226279e-06, + "loss": 0.2565500020980835, + "step": 6845, + "token_acc": 0.9276548501200589 + }, + { + "epoch": 0.44401231567006966, + "grad_norm": 0.8768362998962402, + "learning_rate": 6.323747975710602e-06, + "loss": 0.2480078935623169, + "step": 6850, + "token_acc": 0.930231661967489 + }, + { + "epoch": 0.44433641225085074, + "grad_norm": 0.7817205190658569, + "learning_rate": 6.318579551597997e-06, + "loss": 0.24654531478881836, + "step": 6855, + "token_acc": 0.930560944970963 + }, + { + "epoch": 0.44466050883163183, + "grad_norm": 0.8359101414680481, + "learning_rate": 6.3134096128254625e-06, + "loss": 0.24729843139648439, + "step": 6860, + "token_acc": 0.9294746674396882 + }, + { + "epoch": 0.4449846054124129, + "grad_norm": 2.2188644409179688, + "learning_rate": 6.308238165331737e-06, + "loss": 0.2432945728302002, + "step": 6865, + "token_acc": 0.9304408028372315 + }, + { + "epoch": 0.44530870199319394, + "grad_norm": 0.9280081987380981, + "learning_rate": 6.303065215057291e-06, + "loss": 0.24894299507141113, + "step": 6870, + "token_acc": 0.9291882898229212 + }, + { + "epoch": 0.44563279857397503, + "grad_norm": 0.9287869930267334, + "learning_rate": 6.2978907679443235e-06, + "loss": 0.24746317863464357, + "step": 6875, + "token_acc": 0.9301460945286681 + }, + { + "epoch": 0.4459568951547561, + "grad_norm": 0.9028133153915405, + "learning_rate": 6.292714829936753e-06, + "loss": 0.24716858863830565, + "step": 6880, + "token_acc": 0.9295245034577426 + }, + { + "epoch": 0.4462809917355372, + "grad_norm": 0.8528873920440674, + "learning_rate": 6.287537406980204e-06, + "loss": 0.24401540756225587, + "step": 6885, + "token_acc": 0.9298846021469066 + }, + { + "epoch": 0.4466050883163183, + "grad_norm": 0.9566242098808289, + "learning_rate": 6.2823585050220185e-06, + "loss": 0.24426093101501464, + "step": 6890, + "token_acc": 0.930919791862905 + }, + { + "epoch": 0.4469291848970993, + "grad_norm": 0.8177052140235901, + "learning_rate": 6.277178130011228e-06, + "loss": 0.2438453197479248, + "step": 6895, + "token_acc": 0.9308969754437466 + }, + { + "epoch": 0.4472532814778804, + "grad_norm": 0.7762317657470703, + "learning_rate": 6.271996287898559e-06, + "loss": 0.2454068660736084, + "step": 6900, + "token_acc": 0.9308352233507544 + }, + { + "epoch": 0.4475773780586615, + "grad_norm": 0.8040958642959595, + "learning_rate": 6.266812984636424e-06, + "loss": 0.2489825963973999, + "step": 6905, + "token_acc": 0.9300500537467769 + }, + { + "epoch": 0.4479014746394426, + "grad_norm": 0.9034572839736938, + "learning_rate": 6.261628226178913e-06, + "loss": 0.24741644859313966, + "step": 6910, + "token_acc": 0.9305717238041673 + }, + { + "epoch": 0.4482255712202236, + "grad_norm": 0.8136153221130371, + "learning_rate": 6.256442018481788e-06, + "loss": 0.24943118095397948, + "step": 6915, + "token_acc": 0.9295736995161901 + }, + { + "epoch": 0.4485496678010047, + "grad_norm": 0.8086121678352356, + "learning_rate": 6.251254367502475e-06, + "loss": 0.2401984691619873, + "step": 6920, + "token_acc": 0.9318958476265151 + }, + { + "epoch": 0.4488737643817858, + "grad_norm": 0.7901370525360107, + "learning_rate": 6.246065279200061e-06, + "loss": 0.24500036239624023, + "step": 6925, + "token_acc": 0.9303003943779881 + }, + { + "epoch": 0.44919786096256686, + "grad_norm": 0.8167763948440552, + "learning_rate": 6.2408747595352785e-06, + "loss": 0.24441289901733398, + "step": 6930, + "token_acc": 0.9306867574616053 + }, + { + "epoch": 0.4495219575433479, + "grad_norm": 0.8201696872711182, + "learning_rate": 6.23568281447051e-06, + "loss": 0.24563422203063964, + "step": 6935, + "token_acc": 0.9307546748314696 + }, + { + "epoch": 0.449846054124129, + "grad_norm": 0.8428823351860046, + "learning_rate": 6.230489449969769e-06, + "loss": 0.2507824182510376, + "step": 6940, + "token_acc": 0.9300005867511588 + }, + { + "epoch": 0.45017015070491007, + "grad_norm": 0.7671437859535217, + "learning_rate": 6.225294671998707e-06, + "loss": 0.24523510932922363, + "step": 6945, + "token_acc": 0.9308130820936062 + }, + { + "epoch": 0.45049424728569115, + "grad_norm": 0.8143147826194763, + "learning_rate": 6.220098486524595e-06, + "loss": 0.2412482738494873, + "step": 6950, + "token_acc": 0.9312241015350134 + }, + { + "epoch": 0.4508183438664722, + "grad_norm": 0.8239895701408386, + "learning_rate": 6.21490089951632e-06, + "loss": 0.24498944282531737, + "step": 6955, + "token_acc": 0.9296124620558983 + }, + { + "epoch": 0.45114244044725327, + "grad_norm": 0.774706244468689, + "learning_rate": 6.20970191694438e-06, + "loss": 0.2444399356842041, + "step": 6960, + "token_acc": 0.9312609173837096 + }, + { + "epoch": 0.45146653702803435, + "grad_norm": 0.7867602705955505, + "learning_rate": 6.204501544780879e-06, + "loss": 0.24828519821166992, + "step": 6965, + "token_acc": 0.9299280320673887 + }, + { + "epoch": 0.45179063360881544, + "grad_norm": 0.8243736624717712, + "learning_rate": 6.199299788999509e-06, + "loss": 0.24571335315704346, + "step": 6970, + "token_acc": 0.931302849674238 + }, + { + "epoch": 0.4521147301895965, + "grad_norm": 0.7898803353309631, + "learning_rate": 6.194096655575561e-06, + "loss": 0.2523608922958374, + "step": 6975, + "token_acc": 0.9295294841488294 + }, + { + "epoch": 0.45243882677037756, + "grad_norm": 0.8108680248260498, + "learning_rate": 6.188892150485904e-06, + "loss": 0.23994555473327636, + "step": 6980, + "token_acc": 0.9321341715229661 + }, + { + "epoch": 0.45276292335115864, + "grad_norm": 0.8507076501846313, + "learning_rate": 6.1836862797089815e-06, + "loss": 0.24871957302093506, + "step": 6985, + "token_acc": 0.9296751357705265 + }, + { + "epoch": 0.4530870199319397, + "grad_norm": 0.7880665063858032, + "learning_rate": 6.178479049224809e-06, + "loss": 0.2360888957977295, + "step": 6990, + "token_acc": 0.9324869536121282 + }, + { + "epoch": 0.4534111165127208, + "grad_norm": 0.7432335019111633, + "learning_rate": 6.173270465014961e-06, + "loss": 0.2419327974319458, + "step": 6995, + "token_acc": 0.931600144068629 + }, + { + "epoch": 0.45373521309350184, + "grad_norm": 0.8045170903205872, + "learning_rate": 6.168060533062569e-06, + "loss": 0.24260053634643555, + "step": 7000, + "token_acc": 0.9304721584849562 + }, + { + "epoch": 0.45373521309350184, + "eval_loss": 0.2480049580335617, + "eval_runtime": 114.5944, + "eval_samples_per_second": 174.529, + "eval_steps_per_second": 0.236, + "eval_token_acc": 0.9291911451885065, + "step": 7000 + }, + { + "epoch": 0.45405930967428293, + "grad_norm": 0.8640063405036926, + "learning_rate": 6.16284925935231e-06, + "loss": 0.24346632957458497, + "step": 7005, + "token_acc": 0.9310155678612475 + }, + { + "epoch": 0.454383406255064, + "grad_norm": 0.8250936269760132, + "learning_rate": 6.157636649870406e-06, + "loss": 0.24369192123413086, + "step": 7010, + "token_acc": 0.9303238020629325 + }, + { + "epoch": 0.4547075028358451, + "grad_norm": 0.7867532968521118, + "learning_rate": 6.152422710604611e-06, + "loss": 0.2530375480651855, + "step": 7015, + "token_acc": 0.9287874461346763 + }, + { + "epoch": 0.45503159941662613, + "grad_norm": 0.817786455154419, + "learning_rate": 6.147207447544209e-06, + "loss": 0.24739413261413573, + "step": 7020, + "token_acc": 0.9295719768551891 + }, + { + "epoch": 0.4553556959974072, + "grad_norm": 0.8706911206245422, + "learning_rate": 6.1419908666799975e-06, + "loss": 0.24179625511169434, + "step": 7025, + "token_acc": 0.9310759119631643 + }, + { + "epoch": 0.4556797925781883, + "grad_norm": 0.755524218082428, + "learning_rate": 6.136772974004299e-06, + "loss": 0.24910340309143067, + "step": 7030, + "token_acc": 0.9291959182608241 + }, + { + "epoch": 0.4560038891589694, + "grad_norm": 0.8053135275840759, + "learning_rate": 6.131553775510933e-06, + "loss": 0.2488250732421875, + "step": 7035, + "token_acc": 0.9306361048136237 + }, + { + "epoch": 0.4563279857397504, + "grad_norm": 0.7617267370223999, + "learning_rate": 6.126333277195225e-06, + "loss": 0.24774932861328125, + "step": 7040, + "token_acc": 0.9299860037410561 + }, + { + "epoch": 0.4566520823205315, + "grad_norm": 0.773455023765564, + "learning_rate": 6.121111485053989e-06, + "loss": 0.25046119689941404, + "step": 7045, + "token_acc": 0.9293088088874915 + }, + { + "epoch": 0.4569761789013126, + "grad_norm": 0.7508012056350708, + "learning_rate": 6.115888405085531e-06, + "loss": 0.24143848419189454, + "step": 7050, + "token_acc": 0.9310436825289405 + }, + { + "epoch": 0.4573002754820937, + "grad_norm": 0.8446716070175171, + "learning_rate": 6.11066404328963e-06, + "loss": 0.24196486473083495, + "step": 7055, + "token_acc": 0.931147093959383 + }, + { + "epoch": 0.45762437206287476, + "grad_norm": 0.793343722820282, + "learning_rate": 6.1054384056675444e-06, + "loss": 0.2391761064529419, + "step": 7060, + "token_acc": 0.930980598944099 + }, + { + "epoch": 0.4579484686436558, + "grad_norm": 0.7833433151245117, + "learning_rate": 6.100211498221988e-06, + "loss": 0.2528715133666992, + "step": 7065, + "token_acc": 0.9286416548686487 + }, + { + "epoch": 0.4582725652244369, + "grad_norm": 0.7633975744247437, + "learning_rate": 6.094983326957147e-06, + "loss": 0.2432431697845459, + "step": 7070, + "token_acc": 0.9305570042766246 + }, + { + "epoch": 0.45859666180521796, + "grad_norm": 0.7638473510742188, + "learning_rate": 6.089753897878646e-06, + "loss": 0.24175143241882324, + "step": 7075, + "token_acc": 0.9314752586150096 + }, + { + "epoch": 0.45892075838599905, + "grad_norm": 0.7639363408088684, + "learning_rate": 6.084523216993563e-06, + "loss": 0.2517850399017334, + "step": 7080, + "token_acc": 0.9286959206117474 + }, + { + "epoch": 0.4592448549667801, + "grad_norm": 0.8633525967597961, + "learning_rate": 6.0792912903104095e-06, + "loss": 0.24610002040863038, + "step": 7085, + "token_acc": 0.9309653431617471 + }, + { + "epoch": 0.45956895154756117, + "grad_norm": 0.7721365690231323, + "learning_rate": 6.074058123839134e-06, + "loss": 0.24886999130249024, + "step": 7090, + "token_acc": 0.9294607967438443 + }, + { + "epoch": 0.45989304812834225, + "grad_norm": 0.8289453387260437, + "learning_rate": 6.068823723591099e-06, + "loss": 0.24496798515319823, + "step": 7095, + "token_acc": 0.930727327686559 + }, + { + "epoch": 0.46021714470912334, + "grad_norm": 0.7461783289909363, + "learning_rate": 6.063588095579095e-06, + "loss": 0.24125440120697023, + "step": 7100, + "token_acc": 0.9307487152800364 + }, + { + "epoch": 0.46054124128990437, + "grad_norm": 0.8909839391708374, + "learning_rate": 6.058351245817315e-06, + "loss": 0.24663491249084474, + "step": 7105, + "token_acc": 0.9306657877389585 + }, + { + "epoch": 0.46086533787068545, + "grad_norm": 0.7701578140258789, + "learning_rate": 6.053113180321361e-06, + "loss": 0.24728705883026122, + "step": 7110, + "token_acc": 0.9299513929899477 + }, + { + "epoch": 0.46118943445146654, + "grad_norm": 0.8420730233192444, + "learning_rate": 6.0478739051082295e-06, + "loss": 0.2519880533218384, + "step": 7115, + "token_acc": 0.9291380561782596 + }, + { + "epoch": 0.4615135310322476, + "grad_norm": 0.8113135099411011, + "learning_rate": 6.042633426196302e-06, + "loss": 0.24408466815948487, + "step": 7120, + "token_acc": 0.9317681439216127 + }, + { + "epoch": 0.46183762761302866, + "grad_norm": 0.80897456407547, + "learning_rate": 6.03739174960535e-06, + "loss": 0.24516592025756836, + "step": 7125, + "token_acc": 0.9300213606058645 + }, + { + "epoch": 0.46216172419380974, + "grad_norm": 0.8209992051124573, + "learning_rate": 6.032148881356516e-06, + "loss": 0.24397122859954834, + "step": 7130, + "token_acc": 0.930765451508303 + }, + { + "epoch": 0.46248582077459083, + "grad_norm": 0.790714681148529, + "learning_rate": 6.026904827472315e-06, + "loss": 0.2530096769332886, + "step": 7135, + "token_acc": 0.9289773098376288 + }, + { + "epoch": 0.4628099173553719, + "grad_norm": 0.8383195996284485, + "learning_rate": 6.021659593976621e-06, + "loss": 0.24527640342712403, + "step": 7140, + "token_acc": 0.9306662897599147 + }, + { + "epoch": 0.463134013936153, + "grad_norm": 0.9278517365455627, + "learning_rate": 6.016413186894664e-06, + "loss": 0.24085440635681152, + "step": 7145, + "token_acc": 0.9311666321993589 + }, + { + "epoch": 0.46345811051693403, + "grad_norm": 0.7752022743225098, + "learning_rate": 6.01116561225302e-06, + "loss": 0.25044426918029783, + "step": 7150, + "token_acc": 0.9294272136881042 + }, + { + "epoch": 0.4637822070977151, + "grad_norm": 0.8321642279624939, + "learning_rate": 6.005916876079612e-06, + "loss": 0.24100041389465332, + "step": 7155, + "token_acc": 0.930574219557577 + }, + { + "epoch": 0.4641063036784962, + "grad_norm": 0.793859601020813, + "learning_rate": 6.0006669844036905e-06, + "loss": 0.24372425079345703, + "step": 7160, + "token_acc": 0.9311068899570993 + }, + { + "epoch": 0.4644304002592773, + "grad_norm": 0.7799960970878601, + "learning_rate": 5.995415943255837e-06, + "loss": 0.24092025756835939, + "step": 7165, + "token_acc": 0.9316692159989586 + }, + { + "epoch": 0.4647544968400583, + "grad_norm": 0.780870258808136, + "learning_rate": 5.990163758667953e-06, + "loss": 0.24342315196990966, + "step": 7170, + "token_acc": 0.9307079384857527 + }, + { + "epoch": 0.4650785934208394, + "grad_norm": 0.7920390963554382, + "learning_rate": 5.984910436673255e-06, + "loss": 0.24602940082550048, + "step": 7175, + "token_acc": 0.930932444729589 + }, + { + "epoch": 0.4654026900016205, + "grad_norm": 0.7609754800796509, + "learning_rate": 5.979655983306261e-06, + "loss": 0.2466379165649414, + "step": 7180, + "token_acc": 0.9302953586497891 + }, + { + "epoch": 0.4657267865824016, + "grad_norm": 0.7831788659095764, + "learning_rate": 5.974400404602794e-06, + "loss": 0.24454162120819092, + "step": 7185, + "token_acc": 0.9304420469116295 + }, + { + "epoch": 0.4660508831631826, + "grad_norm": 0.841581404209137, + "learning_rate": 5.969143706599965e-06, + "loss": 0.2416010856628418, + "step": 7190, + "token_acc": 0.9306707727841401 + }, + { + "epoch": 0.4663749797439637, + "grad_norm": 0.7865729331970215, + "learning_rate": 5.963885895336178e-06, + "loss": 0.24700531959533692, + "step": 7195, + "token_acc": 0.9299680817936475 + }, + { + "epoch": 0.4666990763247448, + "grad_norm": 0.8593873977661133, + "learning_rate": 5.958626976851108e-06, + "loss": 0.24058361053466798, + "step": 7200, + "token_acc": 0.9316406378974478 + }, + { + "epoch": 0.46702317290552586, + "grad_norm": 0.7791067361831665, + "learning_rate": 5.953366957185704e-06, + "loss": 0.24740440845489503, + "step": 7205, + "token_acc": 0.9300064383039668 + }, + { + "epoch": 0.4673472694863069, + "grad_norm": 0.9328451156616211, + "learning_rate": 5.9481058423821835e-06, + "loss": 0.24319086074829102, + "step": 7210, + "token_acc": 0.9312566298414012 + }, + { + "epoch": 0.467671366067088, + "grad_norm": 0.8297845721244812, + "learning_rate": 5.942843638484017e-06, + "loss": 0.24497509002685547, + "step": 7215, + "token_acc": 0.9307391702003824 + }, + { + "epoch": 0.46799546264786906, + "grad_norm": 0.9476004242897034, + "learning_rate": 5.93758035153593e-06, + "loss": 0.2518218278884888, + "step": 7220, + "token_acc": 0.92853568666697 + }, + { + "epoch": 0.46831955922865015, + "grad_norm": 0.7887179255485535, + "learning_rate": 5.932315987583889e-06, + "loss": 0.2438260555267334, + "step": 7225, + "token_acc": 0.9305594419246312 + }, + { + "epoch": 0.46864365580943124, + "grad_norm": 0.77700275182724, + "learning_rate": 5.927050552675102e-06, + "loss": 0.24254326820373534, + "step": 7230, + "token_acc": 0.9312934576744369 + }, + { + "epoch": 0.46896775239021227, + "grad_norm": 0.79954993724823, + "learning_rate": 5.921784052858002e-06, + "loss": 0.23991994857788085, + "step": 7235, + "token_acc": 0.9320260577745608 + }, + { + "epoch": 0.46929184897099335, + "grad_norm": 0.8309112787246704, + "learning_rate": 5.916516494182248e-06, + "loss": 0.24758648872375488, + "step": 7240, + "token_acc": 0.9295347049396068 + }, + { + "epoch": 0.46961594555177444, + "grad_norm": 0.7835402488708496, + "learning_rate": 5.911247882698716e-06, + "loss": 0.24622159004211425, + "step": 7245, + "token_acc": 0.9302724035037059 + }, + { + "epoch": 0.4699400421325555, + "grad_norm": 0.7982234954833984, + "learning_rate": 5.90597822445949e-06, + "loss": 0.24415826797485352, + "step": 7250, + "token_acc": 0.9299488644811781 + }, + { + "epoch": 0.47026413871333655, + "grad_norm": 0.801871120929718, + "learning_rate": 5.900707525517855e-06, + "loss": 0.24193406105041504, + "step": 7255, + "token_acc": 0.9310414153598713 + }, + { + "epoch": 0.47058823529411764, + "grad_norm": 0.823549747467041, + "learning_rate": 5.895435791928297e-06, + "loss": 0.2442370891571045, + "step": 7260, + "token_acc": 0.9299186631559738 + }, + { + "epoch": 0.4709123318748987, + "grad_norm": 0.8188885450363159, + "learning_rate": 5.890163029746483e-06, + "loss": 0.2481070041656494, + "step": 7265, + "token_acc": 0.9294977098351468 + }, + { + "epoch": 0.4712364284556798, + "grad_norm": 0.7622184753417969, + "learning_rate": 5.884889245029268e-06, + "loss": 0.24289364814758302, + "step": 7270, + "token_acc": 0.9310376005890026 + }, + { + "epoch": 0.47156052503646084, + "grad_norm": 0.8150364756584167, + "learning_rate": 5.879614443834677e-06, + "loss": 0.247299861907959, + "step": 7275, + "token_acc": 0.9301882498293071 + }, + { + "epoch": 0.47188462161724193, + "grad_norm": 1.145160436630249, + "learning_rate": 5.874338632221905e-06, + "loss": 0.24291410446166992, + "step": 7280, + "token_acc": 0.9300374463894684 + }, + { + "epoch": 0.472208718198023, + "grad_norm": 0.8822237849235535, + "learning_rate": 5.869061816251308e-06, + "loss": 0.24790515899658203, + "step": 7285, + "token_acc": 0.9296458281961767 + }, + { + "epoch": 0.4725328147788041, + "grad_norm": 0.8110260963439941, + "learning_rate": 5.863784001984395e-06, + "loss": 0.24577221870422364, + "step": 7290, + "token_acc": 0.9301813513079763 + }, + { + "epoch": 0.47285691135958513, + "grad_norm": 0.8428686857223511, + "learning_rate": 5.85850519548382e-06, + "loss": 0.24503989219665528, + "step": 7295, + "token_acc": 0.9308361221133727 + }, + { + "epoch": 0.4731810079403662, + "grad_norm": 0.7832386493682861, + "learning_rate": 5.853225402813381e-06, + "loss": 0.24122278690338134, + "step": 7300, + "token_acc": 0.9313942232222281 + }, + { + "epoch": 0.4735051045211473, + "grad_norm": 0.7935068607330322, + "learning_rate": 5.847944630038004e-06, + "loss": 0.24553532600402833, + "step": 7305, + "token_acc": 0.9309881054696469 + }, + { + "epoch": 0.4738292011019284, + "grad_norm": 0.7849323749542236, + "learning_rate": 5.842662883223742e-06, + "loss": 0.24533777236938475, + "step": 7310, + "token_acc": 0.9303316886858245 + }, + { + "epoch": 0.4741532976827095, + "grad_norm": 0.8023664355278015, + "learning_rate": 5.837380168437774e-06, + "loss": 0.2432631015777588, + "step": 7315, + "token_acc": 0.9308933965228656 + }, + { + "epoch": 0.4744773942634905, + "grad_norm": 0.8131499290466309, + "learning_rate": 5.832096491748377e-06, + "loss": 0.23787450790405273, + "step": 7320, + "token_acc": 0.9318131096320218 + }, + { + "epoch": 0.4748014908442716, + "grad_norm": 0.7458800673484802, + "learning_rate": 5.826811859224947e-06, + "loss": 0.24123818874359132, + "step": 7325, + "token_acc": 0.9312690999250418 + }, + { + "epoch": 0.4751255874250527, + "grad_norm": 0.8308848142623901, + "learning_rate": 5.821526276937969e-06, + "loss": 0.23940391540527345, + "step": 7330, + "token_acc": 0.9321470060196431 + }, + { + "epoch": 0.47544968400583376, + "grad_norm": 0.8587900400161743, + "learning_rate": 5.816239750959022e-06, + "loss": 0.23550353050231934, + "step": 7335, + "token_acc": 0.9331772205418415 + }, + { + "epoch": 0.4757737805866148, + "grad_norm": 0.7613879442214966, + "learning_rate": 5.810952287360769e-06, + "loss": 0.24399156570434571, + "step": 7340, + "token_acc": 0.9303306484430004 + }, + { + "epoch": 0.4760978771673959, + "grad_norm": 0.8084679841995239, + "learning_rate": 5.805663892216951e-06, + "loss": 0.24433746337890624, + "step": 7345, + "token_acc": 0.9304250399911922 + }, + { + "epoch": 0.47642197374817696, + "grad_norm": 0.8833265900611877, + "learning_rate": 5.800374571602375e-06, + "loss": 0.24544548988342285, + "step": 7350, + "token_acc": 0.9300916443464103 + }, + { + "epoch": 0.47674607032895805, + "grad_norm": 0.8222519159317017, + "learning_rate": 5.7950843315929195e-06, + "loss": 0.24089088439941406, + "step": 7355, + "token_acc": 0.9308697494052857 + }, + { + "epoch": 0.4770701669097391, + "grad_norm": 0.8963642716407776, + "learning_rate": 5.789793178265508e-06, + "loss": 0.24761710166931153, + "step": 7360, + "token_acc": 0.9299117820543568 + }, + { + "epoch": 0.47739426349052017, + "grad_norm": 0.8671111464500427, + "learning_rate": 5.784501117698121e-06, + "loss": 0.2512833595275879, + "step": 7365, + "token_acc": 0.9292549779412728 + }, + { + "epoch": 0.47771836007130125, + "grad_norm": 0.8154078125953674, + "learning_rate": 5.779208155969781e-06, + "loss": 0.24694218635559081, + "step": 7370, + "token_acc": 0.9297978784681997 + }, + { + "epoch": 0.47804245665208234, + "grad_norm": 0.7898163795471191, + "learning_rate": 5.773914299160541e-06, + "loss": 0.24669747352600097, + "step": 7375, + "token_acc": 0.9300495195618667 + }, + { + "epoch": 0.47836655323286337, + "grad_norm": 0.8330132961273193, + "learning_rate": 5.768619553351484e-06, + "loss": 0.24952363967895508, + "step": 7380, + "token_acc": 0.9295875104013948 + }, + { + "epoch": 0.47869064981364445, + "grad_norm": 0.8676002025604248, + "learning_rate": 5.763323924624719e-06, + "loss": 0.2445683002471924, + "step": 7385, + "token_acc": 0.9305887922592527 + }, + { + "epoch": 0.47901474639442554, + "grad_norm": 0.8415531516075134, + "learning_rate": 5.758027419063361e-06, + "loss": 0.24503810405731202, + "step": 7390, + "token_acc": 0.9303431806801987 + }, + { + "epoch": 0.4793388429752066, + "grad_norm": 0.7999857068061829, + "learning_rate": 5.7527300427515405e-06, + "loss": 0.2508838176727295, + "step": 7395, + "token_acc": 0.9295394895374055 + }, + { + "epoch": 0.4796629395559877, + "grad_norm": 0.7936879396438599, + "learning_rate": 5.74743180177438e-06, + "loss": 0.24643292427062988, + "step": 7400, + "token_acc": 0.9309012649563649 + }, + { + "epoch": 0.47998703613676874, + "grad_norm": 0.8030619621276855, + "learning_rate": 5.742132702218003e-06, + "loss": 0.24593739509582518, + "step": 7405, + "token_acc": 0.9305901390310756 + }, + { + "epoch": 0.4803111327175498, + "grad_norm": 0.814037561416626, + "learning_rate": 5.7368327501695145e-06, + "loss": 0.23953685760498047, + "step": 7410, + "token_acc": 0.9315129460428829 + }, + { + "epoch": 0.4806352292983309, + "grad_norm": 0.8436970710754395, + "learning_rate": 5.731531951716998e-06, + "loss": 0.2429438591003418, + "step": 7415, + "token_acc": 0.9307386118046281 + }, + { + "epoch": 0.480959325879112, + "grad_norm": 0.8039954900741577, + "learning_rate": 5.726230312949513e-06, + "loss": 0.2452918529510498, + "step": 7420, + "token_acc": 0.9301069732116422 + }, + { + "epoch": 0.48128342245989303, + "grad_norm": 0.860999345779419, + "learning_rate": 5.720927839957083e-06, + "loss": 0.24331765174865722, + "step": 7425, + "token_acc": 0.9306858780573987 + }, + { + "epoch": 0.4816075190406741, + "grad_norm": 0.7792454957962036, + "learning_rate": 5.715624538830689e-06, + "loss": 0.246140193939209, + "step": 7430, + "token_acc": 0.9299925205684368 + }, + { + "epoch": 0.4819316156214552, + "grad_norm": 0.8203843832015991, + "learning_rate": 5.710320415662263e-06, + "loss": 0.24417695999145508, + "step": 7435, + "token_acc": 0.9307122372505398 + }, + { + "epoch": 0.4822557122022363, + "grad_norm": 1.3512002229690552, + "learning_rate": 5.705015476544682e-06, + "loss": 0.23995592594146728, + "step": 7440, + "token_acc": 0.931725936431788 + }, + { + "epoch": 0.4825798087830173, + "grad_norm": 0.8797436356544495, + "learning_rate": 5.699709727571759e-06, + "loss": 0.2497626304626465, + "step": 7445, + "token_acc": 0.929725907913517 + }, + { + "epoch": 0.4829039053637984, + "grad_norm": 0.80140620470047, + "learning_rate": 5.6944031748382415e-06, + "loss": 0.23983383178710938, + "step": 7450, + "token_acc": 0.9317038102084831 + }, + { + "epoch": 0.4832280019445795, + "grad_norm": 0.7998064160346985, + "learning_rate": 5.689095824439794e-06, + "loss": 0.24627900123596191, + "step": 7455, + "token_acc": 0.9311544220044328 + }, + { + "epoch": 0.4835520985253606, + "grad_norm": 0.7984311580657959, + "learning_rate": 5.683787682473003e-06, + "loss": 0.25499677658081055, + "step": 7460, + "token_acc": 0.9289444408088476 + }, + { + "epoch": 0.4838761951061416, + "grad_norm": 0.7579300999641418, + "learning_rate": 5.67847875503536e-06, + "loss": 0.24941422939300537, + "step": 7465, + "token_acc": 0.9300160327609187 + }, + { + "epoch": 0.4842002916869227, + "grad_norm": 0.8316287994384766, + "learning_rate": 5.673169048225262e-06, + "loss": 0.2476407527923584, + "step": 7470, + "token_acc": 0.929546752505062 + }, + { + "epoch": 0.4845243882677038, + "grad_norm": 0.8425142765045166, + "learning_rate": 5.6678585681419975e-06, + "loss": 0.2421328067779541, + "step": 7475, + "token_acc": 0.9317967678841793 + }, + { + "epoch": 0.48484848484848486, + "grad_norm": 0.8572697043418884, + "learning_rate": 5.662547320885748e-06, + "loss": 0.24735674858093262, + "step": 7480, + "token_acc": 0.9297394998449796 + }, + { + "epoch": 0.48517258142926595, + "grad_norm": 0.8174471259117126, + "learning_rate": 5.657235312557572e-06, + "loss": 0.24396953582763672, + "step": 7485, + "token_acc": 0.9308595939973142 + }, + { + "epoch": 0.485496678010047, + "grad_norm": 0.8187726140022278, + "learning_rate": 5.651922549259405e-06, + "loss": 0.24969685077667236, + "step": 7490, + "token_acc": 0.9295560242790152 + }, + { + "epoch": 0.48582077459082806, + "grad_norm": 0.7857913970947266, + "learning_rate": 5.646609037094049e-06, + "loss": 0.2439340829849243, + "step": 7495, + "token_acc": 0.930715502995249 + }, + { + "epoch": 0.48614487117160915, + "grad_norm": 0.7940389513969421, + "learning_rate": 5.641294782165167e-06, + "loss": 0.24431586265563965, + "step": 7500, + "token_acc": 0.93080903135267 + }, + { + "epoch": 0.48614487117160915, + "eval_loss": 0.2470269799232483, + "eval_runtime": 94.7764, + "eval_samples_per_second": 211.023, + "eval_steps_per_second": 0.285, + "eval_token_acc": 0.9290190585907063, + "step": 7500 + }, + { + "epoch": 0.48646896775239024, + "grad_norm": 0.7991194725036621, + "learning_rate": 5.635979790577272e-06, + "loss": 0.24032049179077147, + "step": 7505, + "token_acc": 0.9311292254804938 + }, + { + "epoch": 0.48679306433317127, + "grad_norm": 0.8410168886184692, + "learning_rate": 5.630664068435726e-06, + "loss": 0.24054737091064454, + "step": 7510, + "token_acc": 0.9318539227677791 + }, + { + "epoch": 0.48711716091395235, + "grad_norm": 0.7762529253959656, + "learning_rate": 5.625347621846731e-06, + "loss": 0.23797812461853027, + "step": 7515, + "token_acc": 0.9319988301433074 + }, + { + "epoch": 0.48744125749473344, + "grad_norm": 0.7729056477546692, + "learning_rate": 5.620030456917318e-06, + "loss": 0.2422544240951538, + "step": 7520, + "token_acc": 0.9317390735409579 + }, + { + "epoch": 0.4877653540755145, + "grad_norm": 0.7804219126701355, + "learning_rate": 5.614712579755348e-06, + "loss": 0.24253218173980712, + "step": 7525, + "token_acc": 0.9314864462317546 + }, + { + "epoch": 0.48808945065629555, + "grad_norm": 0.7841384410858154, + "learning_rate": 5.609393996469493e-06, + "loss": 0.2512629985809326, + "step": 7530, + "token_acc": 0.9290513002550551 + }, + { + "epoch": 0.48841354723707664, + "grad_norm": 0.8139344453811646, + "learning_rate": 5.604074713169244e-06, + "loss": 0.24291272163391114, + "step": 7535, + "token_acc": 0.9310141122415491 + }, + { + "epoch": 0.4887376438178577, + "grad_norm": 0.8065107464790344, + "learning_rate": 5.598754735964889e-06, + "loss": 0.24636802673339844, + "step": 7540, + "token_acc": 0.9298178533993448 + }, + { + "epoch": 0.4890617403986388, + "grad_norm": 0.7690847516059875, + "learning_rate": 5.593434070967521e-06, + "loss": 0.2485781192779541, + "step": 7545, + "token_acc": 0.928919075516642 + }, + { + "epoch": 0.48938583697941984, + "grad_norm": 0.7922711372375488, + "learning_rate": 5.588112724289014e-06, + "loss": 0.23648474216461182, + "step": 7550, + "token_acc": 0.9320190256634472 + }, + { + "epoch": 0.4897099335602009, + "grad_norm": 0.765006959438324, + "learning_rate": 5.582790702042031e-06, + "loss": 0.239981746673584, + "step": 7555, + "token_acc": 0.9314938154138915 + }, + { + "epoch": 0.490034030140982, + "grad_norm": 0.7746934294700623, + "learning_rate": 5.57746801034001e-06, + "loss": 0.24087538719177246, + "step": 7560, + "token_acc": 0.9309103234801295 + }, + { + "epoch": 0.4903581267217631, + "grad_norm": 0.807295024394989, + "learning_rate": 5.572144655297161e-06, + "loss": 0.2453380584716797, + "step": 7565, + "token_acc": 0.9301488555163715 + }, + { + "epoch": 0.4906822233025442, + "grad_norm": 0.8400338888168335, + "learning_rate": 5.5668206430284445e-06, + "loss": 0.24542250633239746, + "step": 7570, + "token_acc": 0.9300138869060752 + }, + { + "epoch": 0.4910063198833252, + "grad_norm": 0.8722658157348633, + "learning_rate": 5.561495979649591e-06, + "loss": 0.24783506393432617, + "step": 7575, + "token_acc": 0.9292057919567757 + }, + { + "epoch": 0.4913304164641063, + "grad_norm": 0.8119931221008301, + "learning_rate": 5.55617067127707e-06, + "loss": 0.24852707386016845, + "step": 7580, + "token_acc": 0.9296910072726806 + }, + { + "epoch": 0.4916545130448874, + "grad_norm": 0.7851658463478088, + "learning_rate": 5.550844724028094e-06, + "loss": 0.24525146484375, + "step": 7585, + "token_acc": 0.9305225002096192 + }, + { + "epoch": 0.4919786096256685, + "grad_norm": 0.7551227807998657, + "learning_rate": 5.5455181440206094e-06, + "loss": 0.2473278045654297, + "step": 7590, + "token_acc": 0.929891820820912 + }, + { + "epoch": 0.4923027062064495, + "grad_norm": 0.7961981892585754, + "learning_rate": 5.540190937373289e-06, + "loss": 0.24113478660583496, + "step": 7595, + "token_acc": 0.93153201274126 + }, + { + "epoch": 0.4926268027872306, + "grad_norm": 1.1232867240905762, + "learning_rate": 5.534863110205525e-06, + "loss": 0.24363250732421876, + "step": 7600, + "token_acc": 0.9304038186644885 + }, + { + "epoch": 0.4929508993680117, + "grad_norm": 0.8785202503204346, + "learning_rate": 5.529534668637424e-06, + "loss": 0.24906694889068604, + "step": 7605, + "token_acc": 0.92944146582687 + }, + { + "epoch": 0.49327499594879276, + "grad_norm": 0.8886693120002747, + "learning_rate": 5.5242056187897975e-06, + "loss": 0.24398365020751953, + "step": 7610, + "token_acc": 0.9300042874403981 + }, + { + "epoch": 0.4935990925295738, + "grad_norm": 0.7710747122764587, + "learning_rate": 5.518875966784155e-06, + "loss": 0.24580163955688478, + "step": 7615, + "token_acc": 0.9311868604575376 + }, + { + "epoch": 0.4939231891103549, + "grad_norm": 0.86784827709198, + "learning_rate": 5.513545718742702e-06, + "loss": 0.24286196231842042, + "step": 7620, + "token_acc": 0.9313651158838677 + }, + { + "epoch": 0.49424728569113596, + "grad_norm": 0.8150655031204224, + "learning_rate": 5.50821488078832e-06, + "loss": 0.24462320804595947, + "step": 7625, + "token_acc": 0.9304603706814204 + }, + { + "epoch": 0.49457138227191705, + "grad_norm": 0.7798458933830261, + "learning_rate": 5.5028834590445765e-06, + "loss": 0.24243826866149903, + "step": 7630, + "token_acc": 0.9306309661485387 + }, + { + "epoch": 0.4948954788526981, + "grad_norm": 0.8419156670570374, + "learning_rate": 5.497551459635706e-06, + "loss": 0.24553794860839845, + "step": 7635, + "token_acc": 0.9300106794992216 + }, + { + "epoch": 0.49521957543347916, + "grad_norm": 0.7870699763298035, + "learning_rate": 5.492218888686605e-06, + "loss": 0.2434901237487793, + "step": 7640, + "token_acc": 0.9308198931701499 + }, + { + "epoch": 0.49554367201426025, + "grad_norm": 0.8135480284690857, + "learning_rate": 5.4868857523228304e-06, + "loss": 0.2447603940963745, + "step": 7645, + "token_acc": 0.9304962681681288 + }, + { + "epoch": 0.49586776859504134, + "grad_norm": 0.7578834295272827, + "learning_rate": 5.481552056670588e-06, + "loss": 0.24322094917297363, + "step": 7650, + "token_acc": 0.9309408731032874 + }, + { + "epoch": 0.4961918651758224, + "grad_norm": 0.7788705825805664, + "learning_rate": 5.476217807856722e-06, + "loss": 0.23853001594543458, + "step": 7655, + "token_acc": 0.9325492483151893 + }, + { + "epoch": 0.49651596175660345, + "grad_norm": 0.8064255118370056, + "learning_rate": 5.470883012008715e-06, + "loss": 0.24325900077819823, + "step": 7660, + "token_acc": 0.9305672385349217 + }, + { + "epoch": 0.49684005833738454, + "grad_norm": 0.8105224370956421, + "learning_rate": 5.465547675254679e-06, + "loss": 0.2396378993988037, + "step": 7665, + "token_acc": 0.9312487785811999 + }, + { + "epoch": 0.4971641549181656, + "grad_norm": 0.798450767993927, + "learning_rate": 5.460211803723346e-06, + "loss": 0.24111769199371338, + "step": 7670, + "token_acc": 0.9321596561331712 + }, + { + "epoch": 0.4974882514989467, + "grad_norm": 0.763343870639801, + "learning_rate": 5.454875403544062e-06, + "loss": 0.24174098968505858, + "step": 7675, + "token_acc": 0.9313221910519315 + }, + { + "epoch": 0.49781234807972774, + "grad_norm": 0.8922900557518005, + "learning_rate": 5.44953848084678e-06, + "loss": 0.24481558799743652, + "step": 7680, + "token_acc": 0.9305245983774513 + }, + { + "epoch": 0.4981364446605088, + "grad_norm": 0.7840978503227234, + "learning_rate": 5.444201041762054e-06, + "loss": 0.24006333351135253, + "step": 7685, + "token_acc": 0.9318637798414555 + }, + { + "epoch": 0.4984605412412899, + "grad_norm": 0.8150468468666077, + "learning_rate": 5.438863092421033e-06, + "loss": 0.24881525039672853, + "step": 7690, + "token_acc": 0.9296583349480075 + }, + { + "epoch": 0.498784637822071, + "grad_norm": 0.8252663612365723, + "learning_rate": 5.433524638955448e-06, + "loss": 0.24469571113586425, + "step": 7695, + "token_acc": 0.9307921677117112 + }, + { + "epoch": 0.49910873440285203, + "grad_norm": 0.9058377742767334, + "learning_rate": 5.428185687497611e-06, + "loss": 0.24043920040130615, + "step": 7700, + "token_acc": 0.9315072071895553 + }, + { + "epoch": 0.4994328309836331, + "grad_norm": 0.8012968897819519, + "learning_rate": 5.4228462441804075e-06, + "loss": 0.24302992820739747, + "step": 7705, + "token_acc": 0.9313474915682968 + }, + { + "epoch": 0.4997569275644142, + "grad_norm": 0.7841219902038574, + "learning_rate": 5.4175063151372875e-06, + "loss": 0.2421788215637207, + "step": 7710, + "token_acc": 0.9313296891593424 + }, + { + "epoch": 0.5000810241451953, + "grad_norm": 0.9029830098152161, + "learning_rate": 5.412165906502257e-06, + "loss": 0.2500525951385498, + "step": 7715, + "token_acc": 0.9290055302600981 + }, + { + "epoch": 0.5004051207259763, + "grad_norm": 0.8373149037361145, + "learning_rate": 5.406825024409874e-06, + "loss": 0.24214701652526854, + "step": 7720, + "token_acc": 0.931790159508884 + }, + { + "epoch": 0.5007292173067575, + "grad_norm": 0.9877371191978455, + "learning_rate": 5.401483674995241e-06, + "loss": 0.24039196968078613, + "step": 7725, + "token_acc": 0.9312767550342657 + }, + { + "epoch": 0.5010533138875385, + "grad_norm": 0.9403799176216125, + "learning_rate": 5.3961418643939975e-06, + "loss": 0.2397371292114258, + "step": 7730, + "token_acc": 0.9317112391095177 + }, + { + "epoch": 0.5013774104683195, + "grad_norm": 0.8215600252151489, + "learning_rate": 5.3907995987423114e-06, + "loss": 0.24052364826202394, + "step": 7735, + "token_acc": 0.9309898059848734 + }, + { + "epoch": 0.5017015070491007, + "grad_norm": 0.769806444644928, + "learning_rate": 5.3854568841768726e-06, + "loss": 0.24830429553985595, + "step": 7740, + "token_acc": 0.9293371367538067 + }, + { + "epoch": 0.5020256036298817, + "grad_norm": 0.7827233672142029, + "learning_rate": 5.380113726834893e-06, + "loss": 0.2434443473815918, + "step": 7745, + "token_acc": 0.9304004279481048 + }, + { + "epoch": 0.5023497002106628, + "grad_norm": 0.854593813419342, + "learning_rate": 5.3747701328540805e-06, + "loss": 0.24224843978881835, + "step": 7750, + "token_acc": 0.9315753433506893 + }, + { + "epoch": 0.5026737967914439, + "grad_norm": 0.7842069268226624, + "learning_rate": 5.369426108372658e-06, + "loss": 0.25064547061920167, + "step": 7755, + "token_acc": 0.9295774647887324 + }, + { + "epoch": 0.5029978933722249, + "grad_norm": 0.8304519057273865, + "learning_rate": 5.364081659529338e-06, + "loss": 0.250368595123291, + "step": 7760, + "token_acc": 0.9292641484773232 + }, + { + "epoch": 0.503321989953006, + "grad_norm": 0.7939998507499695, + "learning_rate": 5.358736792463316e-06, + "loss": 0.23783884048461915, + "step": 7765, + "token_acc": 0.9318341758830632 + }, + { + "epoch": 0.5036460865337871, + "grad_norm": 0.8201247453689575, + "learning_rate": 5.353391513314272e-06, + "loss": 0.2445065975189209, + "step": 7770, + "token_acc": 0.9306209991420841 + }, + { + "epoch": 0.5039701831145681, + "grad_norm": 0.7881768345832825, + "learning_rate": 5.348045828222363e-06, + "loss": 0.24810023307800294, + "step": 7775, + "token_acc": 0.9295796504522912 + }, + { + "epoch": 0.5042942796953492, + "grad_norm": 0.7838528752326965, + "learning_rate": 5.342699743328203e-06, + "loss": 0.23910899162292482, + "step": 7780, + "token_acc": 0.9321085841987643 + }, + { + "epoch": 0.5046183762761303, + "grad_norm": 0.8515399694442749, + "learning_rate": 5.337353264772874e-06, + "loss": 0.24224519729614258, + "step": 7785, + "token_acc": 0.9314966298017145 + }, + { + "epoch": 0.5049424728569114, + "grad_norm": 0.7680197358131409, + "learning_rate": 5.332006398697906e-06, + "loss": 0.24511206150054932, + "step": 7790, + "token_acc": 0.9299646592884048 + }, + { + "epoch": 0.5052665694376924, + "grad_norm": 0.9122819900512695, + "learning_rate": 5.3266591512452735e-06, + "loss": 0.24524412155151368, + "step": 7795, + "token_acc": 0.9304604348334611 + }, + { + "epoch": 0.5055906660184735, + "grad_norm": 0.7906087040901184, + "learning_rate": 5.321311528557392e-06, + "loss": 0.24533822536468505, + "step": 7800, + "token_acc": 0.9303934398200817 + }, + { + "epoch": 0.5059147625992546, + "grad_norm": 0.9067139625549316, + "learning_rate": 5.3159635367771064e-06, + "loss": 0.2511197805404663, + "step": 7805, + "token_acc": 0.9288504374213126 + }, + { + "epoch": 0.5062388591800356, + "grad_norm": 1.4244379997253418, + "learning_rate": 5.310615182047684e-06, + "loss": 0.24121689796447754, + "step": 7810, + "token_acc": 0.9307596497002317 + }, + { + "epoch": 0.5065629557608167, + "grad_norm": 0.8904971480369568, + "learning_rate": 5.305266470512813e-06, + "loss": 0.24918434619903565, + "step": 7815, + "token_acc": 0.9298726069502349 + }, + { + "epoch": 0.5068870523415978, + "grad_norm": 0.8301509618759155, + "learning_rate": 5.299917408316588e-06, + "loss": 0.24777026176452638, + "step": 7820, + "token_acc": 0.9307029242429171 + }, + { + "epoch": 0.5072111489223788, + "grad_norm": 0.7868053913116455, + "learning_rate": 5.294568001603507e-06, + "loss": 0.23963661193847657, + "step": 7825, + "token_acc": 0.9322958353773463 + }, + { + "epoch": 0.50753524550316, + "grad_norm": 0.7698747515678406, + "learning_rate": 5.289218256518467e-06, + "loss": 0.2473468780517578, + "step": 7830, + "token_acc": 0.9293757704664636 + }, + { + "epoch": 0.507859342083941, + "grad_norm": 0.8122236728668213, + "learning_rate": 5.283868179206747e-06, + "loss": 0.2354792594909668, + "step": 7835, + "token_acc": 0.9322856046764157 + }, + { + "epoch": 0.508183438664722, + "grad_norm": 0.8381909728050232, + "learning_rate": 5.278517775814014e-06, + "loss": 0.24550132751464843, + "step": 7840, + "token_acc": 0.9301673808075186 + }, + { + "epoch": 0.5085075352455032, + "grad_norm": 0.7912919521331787, + "learning_rate": 5.273167052486308e-06, + "loss": 0.24692962169647217, + "step": 7845, + "token_acc": 0.9306367525029166 + }, + { + "epoch": 0.5088316318262842, + "grad_norm": 1.0017120838165283, + "learning_rate": 5.267816015370033e-06, + "loss": 0.2431316375732422, + "step": 7850, + "token_acc": 0.931215077000954 + }, + { + "epoch": 0.5091557284070654, + "grad_norm": 0.8614035248756409, + "learning_rate": 5.2624646706119576e-06, + "loss": 0.2454979181289673, + "step": 7855, + "token_acc": 0.9309589763883392 + }, + { + "epoch": 0.5094798249878464, + "grad_norm": 0.8533740639686584, + "learning_rate": 5.257113024359205e-06, + "loss": 0.23997969627380372, + "step": 7860, + "token_acc": 0.9321140609636185 + }, + { + "epoch": 0.5098039215686274, + "grad_norm": 0.8387362957000732, + "learning_rate": 5.251761082759237e-06, + "loss": 0.2517509698867798, + "step": 7865, + "token_acc": 0.9288694795770267 + }, + { + "epoch": 0.5101280181494086, + "grad_norm": 0.8285295963287354, + "learning_rate": 5.246408851959864e-06, + "loss": 0.24385528564453124, + "step": 7870, + "token_acc": 0.929992353848347 + }, + { + "epoch": 0.5104521147301896, + "grad_norm": 0.7974392175674438, + "learning_rate": 5.241056338109222e-06, + "loss": 0.23985812664031983, + "step": 7875, + "token_acc": 0.9311231855526682 + }, + { + "epoch": 0.5107762113109706, + "grad_norm": 0.7955368757247925, + "learning_rate": 5.235703547355775e-06, + "loss": 0.24634807109832763, + "step": 7880, + "token_acc": 0.930091445175626 + }, + { + "epoch": 0.5111003078917518, + "grad_norm": 1.3570237159729004, + "learning_rate": 5.230350485848306e-06, + "loss": 0.24254817962646485, + "step": 7885, + "token_acc": 0.9313306931968189 + }, + { + "epoch": 0.5114244044725328, + "grad_norm": 0.8287480473518372, + "learning_rate": 5.224997159735905e-06, + "loss": 0.2377037525177002, + "step": 7890, + "token_acc": 0.9320928773960878 + }, + { + "epoch": 0.5117485010533139, + "grad_norm": 0.7878602147102356, + "learning_rate": 5.219643575167972e-06, + "loss": 0.23456892967224122, + "step": 7895, + "token_acc": 0.9329305175678518 + }, + { + "epoch": 0.512072597634095, + "grad_norm": 0.8212928771972656, + "learning_rate": 5.2142897382941974e-06, + "loss": 0.24674248695373535, + "step": 7900, + "token_acc": 0.930282859754606 + }, + { + "epoch": 0.512396694214876, + "grad_norm": 0.780990481376648, + "learning_rate": 5.2089356552645645e-06, + "loss": 0.2401130199432373, + "step": 7905, + "token_acc": 0.9311170716108359 + }, + { + "epoch": 0.5127207907956571, + "grad_norm": 0.7864103317260742, + "learning_rate": 5.20358133222934e-06, + "loss": 0.24617395401000977, + "step": 7910, + "token_acc": 0.9302838709677419 + }, + { + "epoch": 0.5130448873764382, + "grad_norm": 0.778942346572876, + "learning_rate": 5.198226775339068e-06, + "loss": 0.2374958038330078, + "step": 7915, + "token_acc": 0.9319598737717115 + }, + { + "epoch": 0.5133689839572193, + "grad_norm": 0.815487802028656, + "learning_rate": 5.192871990744556e-06, + "loss": 0.24502220153808593, + "step": 7920, + "token_acc": 0.9306892916960365 + }, + { + "epoch": 0.5136930805380003, + "grad_norm": 0.78216952085495, + "learning_rate": 5.187516984596877e-06, + "loss": 0.23564400672912597, + "step": 7925, + "token_acc": 0.9325555440199336 + }, + { + "epoch": 0.5140171771187814, + "grad_norm": 0.8167345523834229, + "learning_rate": 5.182161763047356e-06, + "loss": 0.24689910411834717, + "step": 7930, + "token_acc": 0.9297727021167156 + }, + { + "epoch": 0.5143412736995625, + "grad_norm": 0.8001877069473267, + "learning_rate": 5.176806332247569e-06, + "loss": 0.246451473236084, + "step": 7935, + "token_acc": 0.9300970376956585 + }, + { + "epoch": 0.5146653702803435, + "grad_norm": 0.7823928594589233, + "learning_rate": 5.171450698349329e-06, + "loss": 0.2414849281311035, + "step": 7940, + "token_acc": 0.9315584126526925 + }, + { + "epoch": 0.5149894668611246, + "grad_norm": 0.8291507959365845, + "learning_rate": 5.166094867504684e-06, + "loss": 0.24354779720306396, + "step": 7945, + "token_acc": 0.9306286177877651 + }, + { + "epoch": 0.5153135634419057, + "grad_norm": 0.8140103220939636, + "learning_rate": 5.160738845865907e-06, + "loss": 0.24305620193481445, + "step": 7950, + "token_acc": 0.9310030889286295 + }, + { + "epoch": 0.5156376600226867, + "grad_norm": 0.7822763323783875, + "learning_rate": 5.155382639585492e-06, + "loss": 0.2432279109954834, + "step": 7955, + "token_acc": 0.9308704330801558 + }, + { + "epoch": 0.5159617566034679, + "grad_norm": 0.7554298639297485, + "learning_rate": 5.150026254816143e-06, + "loss": 0.24599411487579345, + "step": 7960, + "token_acc": 0.9306002091411573 + }, + { + "epoch": 0.5162858531842489, + "grad_norm": 0.7561237812042236, + "learning_rate": 5.144669697710771e-06, + "loss": 0.2430462121963501, + "step": 7965, + "token_acc": 0.9304546694907546 + }, + { + "epoch": 0.5166099497650299, + "grad_norm": 0.8321071267127991, + "learning_rate": 5.139312974422484e-06, + "loss": 0.24661059379577638, + "step": 7970, + "token_acc": 0.9308829923232166 + }, + { + "epoch": 0.5169340463458111, + "grad_norm": 0.8013107180595398, + "learning_rate": 5.133956091104579e-06, + "loss": 0.24365005493164063, + "step": 7975, + "token_acc": 0.9318664536290193 + }, + { + "epoch": 0.5172581429265921, + "grad_norm": 0.8609841465950012, + "learning_rate": 5.12859905391054e-06, + "loss": 0.24148359298706054, + "step": 7980, + "token_acc": 0.9310917312661499 + }, + { + "epoch": 0.5175822395073731, + "grad_norm": 0.8396371006965637, + "learning_rate": 5.123241868994029e-06, + "loss": 0.2418349266052246, + "step": 7985, + "token_acc": 0.9318439371863364 + }, + { + "epoch": 0.5179063360881543, + "grad_norm": 0.7766262888908386, + "learning_rate": 5.117884542508871e-06, + "loss": 0.24117686748504638, + "step": 7990, + "token_acc": 0.9311893322874791 + }, + { + "epoch": 0.5182304326689353, + "grad_norm": 0.8737439513206482, + "learning_rate": 5.11252708060906e-06, + "loss": 0.2367872714996338, + "step": 7995, + "token_acc": 0.9317715053055636 + }, + { + "epoch": 0.5185545292497165, + "grad_norm": 0.9268008470535278, + "learning_rate": 5.107169489448743e-06, + "loss": 0.24811651706695556, + "step": 8000, + "token_acc": 0.9293736445316534 + }, + { + "epoch": 0.5185545292497165, + "eval_loss": 0.2457723319530487, + "eval_runtime": 105.7765, + "eval_samples_per_second": 189.078, + "eval_steps_per_second": 0.255, + "eval_token_acc": 0.9298818816713432, + "step": 8000 + }, + { + "epoch": 0.5188786258304975, + "grad_norm": 0.8182088136672974, + "learning_rate": 5.101811775182216e-06, + "loss": 0.2457118034362793, + "step": 8005, + "token_acc": 0.9302541541258059 + }, + { + "epoch": 0.5192027224112785, + "grad_norm": 0.7813268899917603, + "learning_rate": 5.096453943963916e-06, + "loss": 0.24027745723724364, + "step": 8010, + "token_acc": 0.9312314429688663 + }, + { + "epoch": 0.5195268189920597, + "grad_norm": 0.7759381532669067, + "learning_rate": 5.091096001948416e-06, + "loss": 0.24170594215393065, + "step": 8015, + "token_acc": 0.931519464210967 + }, + { + "epoch": 0.5198509155728407, + "grad_norm": 0.8082971572875977, + "learning_rate": 5.085737955290412e-06, + "loss": 0.2472979784011841, + "step": 8020, + "token_acc": 0.9298481769586208 + }, + { + "epoch": 0.5201750121536218, + "grad_norm": 0.8368838429450989, + "learning_rate": 5.080379810144726e-06, + "loss": 0.2482856273651123, + "step": 8025, + "token_acc": 0.9304547856231555 + }, + { + "epoch": 0.5204991087344029, + "grad_norm": 0.8105872273445129, + "learning_rate": 5.0750215726662876e-06, + "loss": 0.24172635078430177, + "step": 8030, + "token_acc": 0.9317255236002107 + }, + { + "epoch": 0.5208232053151839, + "grad_norm": 0.7763067483901978, + "learning_rate": 5.069663249010136e-06, + "loss": 0.23796768188476564, + "step": 8035, + "token_acc": 0.9324099358447113 + }, + { + "epoch": 0.521147301895965, + "grad_norm": 0.9100785851478577, + "learning_rate": 5.064304845331411e-06, + "loss": 0.24763593673706055, + "step": 8040, + "token_acc": 0.9302783280828814 + }, + { + "epoch": 0.5214713984767461, + "grad_norm": 0.7872911691665649, + "learning_rate": 5.058946367785337e-06, + "loss": 0.2422987699508667, + "step": 8045, + "token_acc": 0.931563184479907 + }, + { + "epoch": 0.5217954950575271, + "grad_norm": 0.8047623634338379, + "learning_rate": 5.053587822527232e-06, + "loss": 0.24172568321228027, + "step": 8050, + "token_acc": 0.9315299024589839 + }, + { + "epoch": 0.5221195916383082, + "grad_norm": 0.8453525304794312, + "learning_rate": 5.048229215712486e-06, + "loss": 0.24186124801635742, + "step": 8055, + "token_acc": 0.931052853847903 + }, + { + "epoch": 0.5224436882190893, + "grad_norm": 0.7734045386314392, + "learning_rate": 5.042870553496563e-06, + "loss": 0.24204602241516113, + "step": 8060, + "token_acc": 0.9312360101404308 + }, + { + "epoch": 0.5227677847998704, + "grad_norm": 0.8912164568901062, + "learning_rate": 5.037511842034986e-06, + "loss": 0.24309887886047363, + "step": 8065, + "token_acc": 0.9312218720412954 + }, + { + "epoch": 0.5230918813806514, + "grad_norm": 0.8992875218391418, + "learning_rate": 5.032153087483342e-06, + "loss": 0.24645709991455078, + "step": 8070, + "token_acc": 0.9303330938836792 + }, + { + "epoch": 0.5234159779614325, + "grad_norm": 0.8028161525726318, + "learning_rate": 5.026794295997259e-06, + "loss": 0.2326809883117676, + "step": 8075, + "token_acc": 0.9327388518675449 + }, + { + "epoch": 0.5237400745422136, + "grad_norm": 0.805109441280365, + "learning_rate": 5.021435473732417e-06, + "loss": 0.24801273345947267, + "step": 8080, + "token_acc": 0.9295700701965893 + }, + { + "epoch": 0.5240641711229946, + "grad_norm": 0.8166808485984802, + "learning_rate": 5.01607662684452e-06, + "loss": 0.2413862705230713, + "step": 8085, + "token_acc": 0.9323404145144796 + }, + { + "epoch": 0.5243882677037758, + "grad_norm": 0.8032011389732361, + "learning_rate": 5.01071776148931e-06, + "loss": 0.24653840065002441, + "step": 8090, + "token_acc": 0.9299417043550275 + }, + { + "epoch": 0.5247123642845568, + "grad_norm": 0.8287736177444458, + "learning_rate": 5.005358883822546e-06, + "loss": 0.23931994438171386, + "step": 8095, + "token_acc": 0.9320784200435959 + }, + { + "epoch": 0.5250364608653378, + "grad_norm": 0.861404299736023, + "learning_rate": 5e-06, + "loss": 0.24744443893432616, + "step": 8100, + "token_acc": 0.9300367718211728 + }, + { + "epoch": 0.525360557446119, + "grad_norm": 0.9745165705680847, + "learning_rate": 4.994641116177455e-06, + "loss": 0.24156579971313477, + "step": 8105, + "token_acc": 0.9313802570018174 + }, + { + "epoch": 0.5256846540269, + "grad_norm": 0.8105664849281311, + "learning_rate": 4.9892822385106905e-06, + "loss": 0.23055248260498046, + "step": 8110, + "token_acc": 0.9338000777907429 + }, + { + "epoch": 0.526008750607681, + "grad_norm": 0.7735220193862915, + "learning_rate": 4.983923373155481e-06, + "loss": 0.24546487331390382, + "step": 8115, + "token_acc": 0.9305158807800169 + }, + { + "epoch": 0.5263328471884622, + "grad_norm": 0.8099386692047119, + "learning_rate": 4.978564526267586e-06, + "loss": 0.24430627822875978, + "step": 8120, + "token_acc": 0.9309777019902894 + }, + { + "epoch": 0.5266569437692432, + "grad_norm": 0.8047671318054199, + "learning_rate": 4.973205704002742e-06, + "loss": 0.2410892963409424, + "step": 8125, + "token_acc": 0.9314974698732328 + }, + { + "epoch": 0.5269810403500244, + "grad_norm": 0.7818545699119568, + "learning_rate": 4.96784691251666e-06, + "loss": 0.23745946884155272, + "step": 8130, + "token_acc": 0.93223412165247 + }, + { + "epoch": 0.5273051369308054, + "grad_norm": 0.8808779716491699, + "learning_rate": 4.962488157965015e-06, + "loss": 0.23475260734558107, + "step": 8135, + "token_acc": 0.9329703489897665 + }, + { + "epoch": 0.5276292335115864, + "grad_norm": 0.7542123794555664, + "learning_rate": 4.957129446503438e-06, + "loss": 0.24455347061157226, + "step": 8140, + "token_acc": 0.9301236454842945 + }, + { + "epoch": 0.5279533300923676, + "grad_norm": 1.1037278175354004, + "learning_rate": 4.9517707842875165e-06, + "loss": 0.24701261520385742, + "step": 8145, + "token_acc": 0.9302665874068315 + }, + { + "epoch": 0.5282774266731486, + "grad_norm": 0.7745888829231262, + "learning_rate": 4.94641217747277e-06, + "loss": 0.24668564796447753, + "step": 8150, + "token_acc": 0.929274253397139 + }, + { + "epoch": 0.5286015232539296, + "grad_norm": 0.8106912970542908, + "learning_rate": 4.941053632214665e-06, + "loss": 0.2416560173034668, + "step": 8155, + "token_acc": 0.9319337132239893 + }, + { + "epoch": 0.5289256198347108, + "grad_norm": 0.8011447191238403, + "learning_rate": 4.935695154668591e-06, + "loss": 0.24505655765533446, + "step": 8160, + "token_acc": 0.9308806054970035 + }, + { + "epoch": 0.5292497164154918, + "grad_norm": 0.9251782894134521, + "learning_rate": 4.9303367509898645e-06, + "loss": 0.24659318923950196, + "step": 8165, + "token_acc": 0.930319256218242 + }, + { + "epoch": 0.5295738129962729, + "grad_norm": 0.7943255305290222, + "learning_rate": 4.924978427333715e-06, + "loss": 0.23770604133605958, + "step": 8170, + "token_acc": 0.9323439688919729 + }, + { + "epoch": 0.529897909577054, + "grad_norm": 0.7926489114761353, + "learning_rate": 4.919620189855277e-06, + "loss": 0.23708786964416503, + "step": 8175, + "token_acc": 0.9332639927192355 + }, + { + "epoch": 0.530222006157835, + "grad_norm": 0.9216257333755493, + "learning_rate": 4.91426204470959e-06, + "loss": 0.249497127532959, + "step": 8180, + "token_acc": 0.9301416048550236 + }, + { + "epoch": 0.5305461027386161, + "grad_norm": 0.7923769354820251, + "learning_rate": 4.908903998051586e-06, + "loss": 0.23885347843170165, + "step": 8185, + "token_acc": 0.9325591666017198 + }, + { + "epoch": 0.5308701993193972, + "grad_norm": 0.79524827003479, + "learning_rate": 4.903546056036085e-06, + "loss": 0.2395010471343994, + "step": 8190, + "token_acc": 0.9311084671048782 + }, + { + "epoch": 0.5311942959001783, + "grad_norm": 0.8420419096946716, + "learning_rate": 4.8981882248177845e-06, + "loss": 0.24050664901733398, + "step": 8195, + "token_acc": 0.9324349652609468 + }, + { + "epoch": 0.5315183924809593, + "grad_norm": 0.846523642539978, + "learning_rate": 4.892830510551258e-06, + "loss": 0.2463658332824707, + "step": 8200, + "token_acc": 0.9300151280036879 + }, + { + "epoch": 0.5318424890617404, + "grad_norm": 0.8087852597236633, + "learning_rate": 4.887472919390942e-06, + "loss": 0.2432598352432251, + "step": 8205, + "token_acc": 0.9311726313997866 + }, + { + "epoch": 0.5321665856425215, + "grad_norm": 0.8151853680610657, + "learning_rate": 4.882115457491131e-06, + "loss": 0.23900861740112306, + "step": 8210, + "token_acc": 0.9310613042597277 + }, + { + "epoch": 0.5324906822233025, + "grad_norm": 0.8653593063354492, + "learning_rate": 4.876758131005973e-06, + "loss": 0.24174349308013915, + "step": 8215, + "token_acc": 0.9310197452773742 + }, + { + "epoch": 0.5328147788040836, + "grad_norm": 0.8096201419830322, + "learning_rate": 4.87140094608946e-06, + "loss": 0.24266068935394286, + "step": 8220, + "token_acc": 0.9308755160244305 + }, + { + "epoch": 0.5331388753848647, + "grad_norm": 0.7821497917175293, + "learning_rate": 4.8660439088954215e-06, + "loss": 0.24201838970184325, + "step": 8225, + "token_acc": 0.9307404489956675 + }, + { + "epoch": 0.5334629719656457, + "grad_norm": 0.8003436923027039, + "learning_rate": 4.8606870255775195e-06, + "loss": 0.23647682666778563, + "step": 8230, + "token_acc": 0.931872754927025 + }, + { + "epoch": 0.5337870685464269, + "grad_norm": 0.8531568050384521, + "learning_rate": 4.85533030228923e-06, + "loss": 0.24609119892120362, + "step": 8235, + "token_acc": 0.9306366343517559 + }, + { + "epoch": 0.5341111651272079, + "grad_norm": 0.7754499912261963, + "learning_rate": 4.849973745183859e-06, + "loss": 0.2448775053024292, + "step": 8240, + "token_acc": 0.9308254923132294 + }, + { + "epoch": 0.5344352617079889, + "grad_norm": 0.8106386065483093, + "learning_rate": 4.844617360414509e-06, + "loss": 0.24270193576812743, + "step": 8245, + "token_acc": 0.9309066353265069 + }, + { + "epoch": 0.5347593582887701, + "grad_norm": 0.8205743432044983, + "learning_rate": 4.839261154134094e-06, + "loss": 0.23956866264343263, + "step": 8250, + "token_acc": 0.9317624273724984 + }, + { + "epoch": 0.5350834548695511, + "grad_norm": 0.9065422415733337, + "learning_rate": 4.833905132495319e-06, + "loss": 0.23816938400268556, + "step": 8255, + "token_acc": 0.9322561919504644 + }, + { + "epoch": 0.5354075514503323, + "grad_norm": 0.8193075060844421, + "learning_rate": 4.828549301650673e-06, + "loss": 0.23941342830657958, + "step": 8260, + "token_acc": 0.9313159792800472 + }, + { + "epoch": 0.5357316480311133, + "grad_norm": 0.8344675302505493, + "learning_rate": 4.823193667752432e-06, + "loss": 0.2484200716018677, + "step": 8265, + "token_acc": 0.929895266620521 + }, + { + "epoch": 0.5360557446118943, + "grad_norm": 0.8096815347671509, + "learning_rate": 4.817838236952645e-06, + "loss": 0.24381232261657715, + "step": 8270, + "token_acc": 0.9309579416703662 + }, + { + "epoch": 0.5363798411926755, + "grad_norm": 0.8262730836868286, + "learning_rate": 4.812483015403124e-06, + "loss": 0.2425703525543213, + "step": 8275, + "token_acc": 0.9312016883974467 + }, + { + "epoch": 0.5367039377734565, + "grad_norm": 0.8010238409042358, + "learning_rate": 4.807128009255445e-06, + "loss": 0.24077913761138917, + "step": 8280, + "token_acc": 0.9318037342141308 + }, + { + "epoch": 0.5370280343542375, + "grad_norm": 0.8176038861274719, + "learning_rate": 4.801773224660934e-06, + "loss": 0.24325640201568605, + "step": 8285, + "token_acc": 0.9320766261578217 + }, + { + "epoch": 0.5373521309350187, + "grad_norm": 0.7913457751274109, + "learning_rate": 4.796418667770661e-06, + "loss": 0.2420496940612793, + "step": 8290, + "token_acc": 0.9309660458881558 + }, + { + "epoch": 0.5376762275157997, + "grad_norm": 0.7514023780822754, + "learning_rate": 4.791064344735436e-06, + "loss": 0.23033642768859863, + "step": 8295, + "token_acc": 0.9340288571132526 + }, + { + "epoch": 0.5380003240965808, + "grad_norm": 0.900210440158844, + "learning_rate": 4.785710261705805e-06, + "loss": 0.24016294479370118, + "step": 8300, + "token_acc": 0.9311983362291371 + }, + { + "epoch": 0.5383244206773619, + "grad_norm": 0.8491877317428589, + "learning_rate": 4.78035642483203e-06, + "loss": 0.25060606002807617, + "step": 8305, + "token_acc": 0.9294361313749009 + }, + { + "epoch": 0.5386485172581429, + "grad_norm": 0.8001054525375366, + "learning_rate": 4.7750028402640966e-06, + "loss": 0.2407216787338257, + "step": 8310, + "token_acc": 0.9313146814622437 + }, + { + "epoch": 0.538972613838924, + "grad_norm": 0.8820044994354248, + "learning_rate": 4.769649514151696e-06, + "loss": 0.23607473373413085, + "step": 8315, + "token_acc": 0.9328092807181008 + }, + { + "epoch": 0.5392967104197051, + "grad_norm": 0.8178361058235168, + "learning_rate": 4.764296452644226e-06, + "loss": 0.23810672760009766, + "step": 8320, + "token_acc": 0.9320127988250259 + }, + { + "epoch": 0.5396208070004861, + "grad_norm": 0.8087671995162964, + "learning_rate": 4.75894366189078e-06, + "loss": 0.2373284101486206, + "step": 8325, + "token_acc": 0.9316758795944565 + }, + { + "epoch": 0.5399449035812672, + "grad_norm": 0.8738973140716553, + "learning_rate": 4.753591148040137e-06, + "loss": 0.24265022277832032, + "step": 8330, + "token_acc": 0.9312873825121049 + }, + { + "epoch": 0.5402690001620483, + "grad_norm": 0.8106444478034973, + "learning_rate": 4.748238917240763e-06, + "loss": 0.2397428035736084, + "step": 8335, + "token_acc": 0.9322988836309446 + }, + { + "epoch": 0.5405930967428294, + "grad_norm": 0.8210396766662598, + "learning_rate": 4.742886975640797e-06, + "loss": 0.24487061500549318, + "step": 8340, + "token_acc": 0.9309169278996865 + }, + { + "epoch": 0.5409171933236104, + "grad_norm": 0.8426142334938049, + "learning_rate": 4.737535329388044e-06, + "loss": 0.23839833736419677, + "step": 8345, + "token_acc": 0.9317527419312619 + }, + { + "epoch": 0.5412412899043915, + "grad_norm": 0.8345897793769836, + "learning_rate": 4.732183984629969e-06, + "loss": 0.24344558715820314, + "step": 8350, + "token_acc": 0.9305297473512633 + }, + { + "epoch": 0.5415653864851726, + "grad_norm": 0.8126654624938965, + "learning_rate": 4.7268329475136945e-06, + "loss": 0.23946683406829833, + "step": 8355, + "token_acc": 0.9316787233625439 + }, + { + "epoch": 0.5418894830659536, + "grad_norm": 0.8076168894767761, + "learning_rate": 4.721482224185987e-06, + "loss": 0.23504953384399413, + "step": 8360, + "token_acc": 0.9323022883280473 + }, + { + "epoch": 0.5422135796467348, + "grad_norm": 0.8037662506103516, + "learning_rate": 4.716131820793253e-06, + "loss": 0.23991770744323732, + "step": 8365, + "token_acc": 0.931755131020156 + }, + { + "epoch": 0.5425376762275158, + "grad_norm": 0.8573919534683228, + "learning_rate": 4.710781743481536e-06, + "loss": 0.23679189682006835, + "step": 8370, + "token_acc": 0.9319086609499106 + }, + { + "epoch": 0.5428617728082968, + "grad_norm": 0.8242449760437012, + "learning_rate": 4.705431998396494e-06, + "loss": 0.23830027580261232, + "step": 8375, + "token_acc": 0.9330742268841263 + }, + { + "epoch": 0.543185869389078, + "grad_norm": 0.8846279978752136, + "learning_rate": 4.7000825916834135e-06, + "loss": 0.24338040351867676, + "step": 8380, + "token_acc": 0.9309993022919053 + }, + { + "epoch": 0.543509965969859, + "grad_norm": 1.085318684577942, + "learning_rate": 4.694733529487188e-06, + "loss": 0.236962890625, + "step": 8385, + "token_acc": 0.9318104176201004 + }, + { + "epoch": 0.54383406255064, + "grad_norm": 0.8240640759468079, + "learning_rate": 4.689384817952316e-06, + "loss": 0.24508442878723144, + "step": 8390, + "token_acc": 0.930229718875502 + }, + { + "epoch": 0.5441581591314212, + "grad_norm": 0.7967071533203125, + "learning_rate": 4.684036463222896e-06, + "loss": 0.2489915370941162, + "step": 8395, + "token_acc": 0.9296897651493187 + }, + { + "epoch": 0.5444822557122022, + "grad_norm": 0.824304461479187, + "learning_rate": 4.6786884714426094e-06, + "loss": 0.2351855754852295, + "step": 8400, + "token_acc": 0.9318790711486692 + }, + { + "epoch": 0.5448063522929834, + "grad_norm": 0.7934773564338684, + "learning_rate": 4.673340848754728e-06, + "loss": 0.24302077293395996, + "step": 8405, + "token_acc": 0.9309109054618037 + }, + { + "epoch": 0.5451304488737644, + "grad_norm": 0.7740573287010193, + "learning_rate": 4.667993601302096e-06, + "loss": 0.24123754501342773, + "step": 8410, + "token_acc": 0.931778052298319 + }, + { + "epoch": 0.5454545454545454, + "grad_norm": 0.7817065119743347, + "learning_rate": 4.662646735227127e-06, + "loss": 0.23929495811462403, + "step": 8415, + "token_acc": 0.9309532971153909 + }, + { + "epoch": 0.5457786420353266, + "grad_norm": 0.7800682783126831, + "learning_rate": 4.6573002566717974e-06, + "loss": 0.2433910846710205, + "step": 8420, + "token_acc": 0.931371727883618 + }, + { + "epoch": 0.5461027386161076, + "grad_norm": 0.990047037601471, + "learning_rate": 4.651954171777639e-06, + "loss": 0.24240436553955078, + "step": 8425, + "token_acc": 0.9315155275446559 + }, + { + "epoch": 0.5464268351968887, + "grad_norm": 0.8210124373435974, + "learning_rate": 4.646608486685729e-06, + "loss": 0.23503177165985106, + "step": 8430, + "token_acc": 0.9317767549208598 + }, + { + "epoch": 0.5467509317776698, + "grad_norm": 0.7846969962120056, + "learning_rate": 4.641263207536685e-06, + "loss": 0.23521924018859863, + "step": 8435, + "token_acc": 0.932362857068186 + }, + { + "epoch": 0.5470750283584508, + "grad_norm": 1.123048186302185, + "learning_rate": 4.635918340470664e-06, + "loss": 0.23991575241088867, + "step": 8440, + "token_acc": 0.9320802575525999 + }, + { + "epoch": 0.5473991249392319, + "grad_norm": 0.8502026796340942, + "learning_rate": 4.630573891627341e-06, + "loss": 0.24041686058044434, + "step": 8445, + "token_acc": 0.9316244333561622 + }, + { + "epoch": 0.547723221520013, + "grad_norm": 0.8347294330596924, + "learning_rate": 4.625229867145922e-06, + "loss": 0.25098018646240233, + "step": 8450, + "token_acc": 0.9285222970062805 + }, + { + "epoch": 0.548047318100794, + "grad_norm": 0.8280182480812073, + "learning_rate": 4.6198862731651105e-06, + "loss": 0.23603384494781493, + "step": 8455, + "token_acc": 0.932926436692806 + }, + { + "epoch": 0.5483714146815751, + "grad_norm": 0.8085178136825562, + "learning_rate": 4.614543115823129e-06, + "loss": 0.2388146162033081, + "step": 8460, + "token_acc": 0.9326059217207885 + }, + { + "epoch": 0.5486955112623562, + "grad_norm": 0.8370281457901001, + "learning_rate": 4.60920040125769e-06, + "loss": 0.24448227882385254, + "step": 8465, + "token_acc": 0.9309708498983437 + }, + { + "epoch": 0.5490196078431373, + "grad_norm": 0.8001585006713867, + "learning_rate": 4.603858135606004e-06, + "loss": 0.2341989278793335, + "step": 8470, + "token_acc": 0.933618088483091 + }, + { + "epoch": 0.5493437044239183, + "grad_norm": 0.8019942045211792, + "learning_rate": 4.59851632500476e-06, + "loss": 0.23873131275177, + "step": 8475, + "token_acc": 0.9314791484963013 + }, + { + "epoch": 0.5496678010046994, + "grad_norm": 0.8279576897621155, + "learning_rate": 4.593174975590129e-06, + "loss": 0.2401789665222168, + "step": 8480, + "token_acc": 0.9316625448691735 + }, + { + "epoch": 0.5499918975854805, + "grad_norm": 0.7763727307319641, + "learning_rate": 4.587834093497745e-06, + "loss": 0.23969721794128418, + "step": 8485, + "token_acc": 0.9316840272115943 + }, + { + "epoch": 0.5503159941662615, + "grad_norm": 1.7937692403793335, + "learning_rate": 4.582493684862714e-06, + "loss": 0.24093496799468994, + "step": 8490, + "token_acc": 0.9317538235275146 + }, + { + "epoch": 0.5506400907470426, + "grad_norm": 0.828227162361145, + "learning_rate": 4.577153755819593e-06, + "loss": 0.24339766502380372, + "step": 8495, + "token_acc": 0.9319290392653786 + }, + { + "epoch": 0.5509641873278237, + "grad_norm": 0.8395451307296753, + "learning_rate": 4.57181431250239e-06, + "loss": 0.23802585601806642, + "step": 8500, + "token_acc": 0.9315395006884368 + }, + { + "epoch": 0.5509641873278237, + "eval_loss": 0.24377238750457764, + "eval_runtime": 105.1109, + "eval_samples_per_second": 190.275, + "eval_steps_per_second": 0.257, + "eval_token_acc": 0.9303957513731076, + "step": 8500 + }, + { + "epoch": 0.5279001924151201, + "grad_norm": 1.0943489074707031, + "learning_rate": 4.9522764324554615e-06, + "loss": 0.26176598072052004, + "step": 8505, + "token_acc": 0.9312429863357318 + }, + { + "epoch": 0.5282105393830302, + "grad_norm": 1.4687483310699463, + "learning_rate": 4.947145047925423e-06, + "loss": 0.24499595165252686, + "step": 8510, + "token_acc": 0.9305895573303368 + }, + { + "epoch": 0.5285208863509403, + "grad_norm": 0.9141036868095398, + "learning_rate": 4.942013719070193e-06, + "loss": 0.24243512153625488, + "step": 8515, + "token_acc": 0.9304801781499606 + }, + { + "epoch": 0.5288312333188505, + "grad_norm": 0.9178383350372314, + "learning_rate": 4.93688245129486e-06, + "loss": 0.23781514167785645, + "step": 8520, + "token_acc": 0.9318716212142527 + }, + { + "epoch": 0.5291415802867606, + "grad_norm": 0.8719151020050049, + "learning_rate": 4.931751250004451e-06, + "loss": 0.23084447383880616, + "step": 8525, + "token_acc": 0.9340286745472004 + }, + { + "epoch": 0.5294519272546707, + "grad_norm": 0.8730896711349487, + "learning_rate": 4.926620120603921e-06, + "loss": 0.2331705093383789, + "step": 8530, + "token_acc": 0.9332122435250588 + }, + { + "epoch": 0.5297622742225808, + "grad_norm": 0.8118588924407959, + "learning_rate": 4.921489068498147e-06, + "loss": 0.23608248233795165, + "step": 8535, + "token_acc": 0.9322579138107092 + }, + { + "epoch": 0.5300726211904909, + "grad_norm": 0.7789573669433594, + "learning_rate": 4.91635809909193e-06, + "loss": 0.2308420419692993, + "step": 8540, + "token_acc": 0.9339207342288862 + }, + { + "epoch": 0.5303829681584011, + "grad_norm": 0.882975697517395, + "learning_rate": 4.911227217789982e-06, + "loss": 0.24052877426147462, + "step": 8545, + "token_acc": 0.9310772257747172 + }, + { + "epoch": 0.5306933151263112, + "grad_norm": 1.1390831470489502, + "learning_rate": 4.906096429996915e-06, + "loss": 0.23412952423095704, + "step": 8550, + "token_acc": 0.9325691972586729 + }, + { + "epoch": 0.5310036620942213, + "grad_norm": 0.8960070610046387, + "learning_rate": 4.900965741117258e-06, + "loss": 0.23040266036987306, + "step": 8555, + "token_acc": 0.9338292873923257 + }, + { + "epoch": 0.5313140090621314, + "grad_norm": 0.8467027544975281, + "learning_rate": 4.895835156555417e-06, + "loss": 0.23063969612121582, + "step": 8560, + "token_acc": 0.9344212446786945 + }, + { + "epoch": 0.5316243560300415, + "grad_norm": 0.8829395174980164, + "learning_rate": 4.890704681715704e-06, + "loss": 0.22131528854370117, + "step": 8565, + "token_acc": 0.9363915598290599 + }, + { + "epoch": 0.5319347029979518, + "grad_norm": 0.9108978509902954, + "learning_rate": 4.8855743220023066e-06, + "loss": 0.23421890735626222, + "step": 8570, + "token_acc": 0.9338276376591536 + }, + { + "epoch": 0.5322450499658619, + "grad_norm": 1.1023825407028198, + "learning_rate": 4.880444082819293e-06, + "loss": 0.227890682220459, + "step": 8575, + "token_acc": 0.9345380748613124 + }, + { + "epoch": 0.532555396933772, + "grad_norm": 0.7825270891189575, + "learning_rate": 4.875313969570606e-06, + "loss": 0.22762112617492675, + "step": 8580, + "token_acc": 0.9339450987604853 + }, + { + "epoch": 0.5328657439016821, + "grad_norm": 0.8017994165420532, + "learning_rate": 4.8701839876600555e-06, + "loss": 0.23148279190063475, + "step": 8585, + "token_acc": 0.9341727074483338 + }, + { + "epoch": 0.5331760908695922, + "grad_norm": 0.8482680916786194, + "learning_rate": 4.86505414249131e-06, + "loss": 0.2255796432495117, + "step": 8590, + "token_acc": 0.934995197182347 + }, + { + "epoch": 0.5334864378375024, + "grad_norm": 0.8013801574707031, + "learning_rate": 4.859924439467899e-06, + "loss": 0.23364462852478027, + "step": 8595, + "token_acc": 0.933110590934251 + }, + { + "epoch": 0.5337967848054125, + "grad_norm": 0.8017052412033081, + "learning_rate": 4.8547948839931985e-06, + "loss": 0.23359198570251466, + "step": 8600, + "token_acc": 0.9335612119603093 + }, + { + "epoch": 0.5341071317733226, + "grad_norm": 0.8352408409118652, + "learning_rate": 4.849665481470428e-06, + "loss": 0.22977676391601562, + "step": 8605, + "token_acc": 0.9341043392426001 + }, + { + "epoch": 0.5344174787412327, + "grad_norm": 0.856725811958313, + "learning_rate": 4.844536237302651e-06, + "loss": 0.22916121482849122, + "step": 8610, + "token_acc": 0.933448799309879 + }, + { + "epoch": 0.5347278257091428, + "grad_norm": 0.7687222361564636, + "learning_rate": 4.839407156892758e-06, + "loss": 0.2331073760986328, + "step": 8615, + "token_acc": 0.9338393858362806 + }, + { + "epoch": 0.5350381726770529, + "grad_norm": 0.8741118311882019, + "learning_rate": 4.8342782456434725e-06, + "loss": 0.2293628692626953, + "step": 8620, + "token_acc": 0.9345788125477784 + }, + { + "epoch": 0.5353485196449631, + "grad_norm": 0.7720116972923279, + "learning_rate": 4.829149508957338e-06, + "loss": 0.23599376678466796, + "step": 8625, + "token_acc": 0.9324241340450371 + }, + { + "epoch": 0.5356588666128732, + "grad_norm": 1.1758185625076294, + "learning_rate": 4.824020952236713e-06, + "loss": 0.23185458183288574, + "step": 8630, + "token_acc": 0.9343576628237661 + }, + { + "epoch": 0.5359692135807833, + "grad_norm": 0.8351410031318665, + "learning_rate": 4.818892580883765e-06, + "loss": 0.23154013156890868, + "step": 8635, + "token_acc": 0.9341552938695626 + }, + { + "epoch": 0.5362795605486934, + "grad_norm": 0.8490530252456665, + "learning_rate": 4.813764400300472e-06, + "loss": 0.23256049156188965, + "step": 8640, + "token_acc": 0.9336019842932285 + }, + { + "epoch": 0.5365899075166035, + "grad_norm": 0.7952938675880432, + "learning_rate": 4.808636415888605e-06, + "loss": 0.22400345802307128, + "step": 8645, + "token_acc": 0.9360759661547439 + }, + { + "epoch": 0.5369002544845137, + "grad_norm": 0.8120350241661072, + "learning_rate": 4.80350863304973e-06, + "loss": 0.23719615936279298, + "step": 8650, + "token_acc": 0.932943906537569 + }, + { + "epoch": 0.5372106014524238, + "grad_norm": 0.8350797295570374, + "learning_rate": 4.798381057185203e-06, + "loss": 0.22629237174987793, + "step": 8655, + "token_acc": 0.9350838948341921 + }, + { + "epoch": 0.5375209484203339, + "grad_norm": 0.8154293894767761, + "learning_rate": 4.79325369369616e-06, + "loss": 0.2357499361038208, + "step": 8660, + "token_acc": 0.9342103524374612 + }, + { + "epoch": 0.537831295388244, + "grad_norm": 0.7825552821159363, + "learning_rate": 4.788126547983512e-06, + "loss": 0.22491831779479982, + "step": 8665, + "token_acc": 0.935371976266545 + }, + { + "epoch": 0.5381416423561541, + "grad_norm": 0.776562511920929, + "learning_rate": 4.782999625447945e-06, + "loss": 0.23323287963867187, + "step": 8670, + "token_acc": 0.9339219729020511 + }, + { + "epoch": 0.5384519893240644, + "grad_norm": 0.7857832312583923, + "learning_rate": 4.7778729314899075e-06, + "loss": 0.22632973194122313, + "step": 8675, + "token_acc": 0.9349340673155175 + }, + { + "epoch": 0.5387623362919745, + "grad_norm": 0.8435510993003845, + "learning_rate": 4.772746471509602e-06, + "loss": 0.22583863735198975, + "step": 8680, + "token_acc": 0.934649886196278 + }, + { + "epoch": 0.5390726832598846, + "grad_norm": 0.9099718332290649, + "learning_rate": 4.7676202509069955e-06, + "loss": 0.2343308925628662, + "step": 8685, + "token_acc": 0.9331213163719764 + }, + { + "epoch": 0.5393830302277947, + "grad_norm": 1.25424325466156, + "learning_rate": 4.762494275081794e-06, + "loss": 0.23212890625, + "step": 8690, + "token_acc": 0.9345982828377242 + }, + { + "epoch": 0.5396933771957048, + "grad_norm": 0.8444955348968506, + "learning_rate": 4.757368549433447e-06, + "loss": 0.23432559967041017, + "step": 8695, + "token_acc": 0.9334054501625968 + }, + { + "epoch": 0.540003724163615, + "grad_norm": 0.8400503396987915, + "learning_rate": 4.7522430793611445e-06, + "loss": 0.22563433647155762, + "step": 8700, + "token_acc": 0.9356806155543008 + }, + { + "epoch": 0.5403140711315251, + "grad_norm": 0.7889493107795715, + "learning_rate": 4.747117870263803e-06, + "loss": 0.23246073722839355, + "step": 8705, + "token_acc": 0.9336666958695843 + }, + { + "epoch": 0.5406244180994352, + "grad_norm": 0.844626784324646, + "learning_rate": 4.7419929275400665e-06, + "loss": 0.22999210357666017, + "step": 8710, + "token_acc": 0.9342222456802423 + }, + { + "epoch": 0.5409347650673453, + "grad_norm": 0.9092142581939697, + "learning_rate": 4.736868256588299e-06, + "loss": 0.23199191093444824, + "step": 8715, + "token_acc": 0.9343479577993493 + }, + { + "epoch": 0.5412451120352554, + "grad_norm": 0.8355756402015686, + "learning_rate": 4.7317438628065744e-06, + "loss": 0.2307112693786621, + "step": 8720, + "token_acc": 0.9345810462271014 + }, + { + "epoch": 0.5415554590031655, + "grad_norm": 0.8477380871772766, + "learning_rate": 4.72661975159268e-06, + "loss": 0.23037476539611818, + "step": 8725, + "token_acc": 0.9342635022382327 + }, + { + "epoch": 0.5418658059710757, + "grad_norm": 0.8044423460960388, + "learning_rate": 4.721495928344102e-06, + "loss": 0.22875185012817384, + "step": 8730, + "token_acc": 0.9342845473828439 + }, + { + "epoch": 0.5421761529389858, + "grad_norm": 0.8599227666854858, + "learning_rate": 4.716372398458021e-06, + "loss": 0.2305788516998291, + "step": 8735, + "token_acc": 0.9333685068187586 + }, + { + "epoch": 0.5424864999068959, + "grad_norm": 0.7696322202682495, + "learning_rate": 4.711249167331316e-06, + "loss": 0.23076372146606444, + "step": 8740, + "token_acc": 0.9337488533080031 + }, + { + "epoch": 0.542796846874806, + "grad_norm": 0.832984447479248, + "learning_rate": 4.706126240360544e-06, + "loss": 0.23358781337738038, + "step": 8745, + "token_acc": 0.9338935054082912 + }, + { + "epoch": 0.5431071938427161, + "grad_norm": 1.2727806568145752, + "learning_rate": 4.701003622941944e-06, + "loss": 0.22997024059295654, + "step": 8750, + "token_acc": 0.9347333978832092 + }, + { + "epoch": 0.5434175408106263, + "grad_norm": 0.9071239829063416, + "learning_rate": 4.695881320471431e-06, + "loss": 0.2304611921310425, + "step": 8755, + "token_acc": 0.9339700858952674 + }, + { + "epoch": 0.5437278877785364, + "grad_norm": 0.9167559742927551, + "learning_rate": 4.690759338344586e-06, + "loss": 0.23465757369995116, + "step": 8760, + "token_acc": 0.9331339514172731 + }, + { + "epoch": 0.5440382347464465, + "grad_norm": 0.8482223153114319, + "learning_rate": 4.685637681956654e-06, + "loss": 0.23015966415405273, + "step": 8765, + "token_acc": 0.9338798752774549 + }, + { + "epoch": 0.5443485817143566, + "grad_norm": 0.8712725639343262, + "learning_rate": 4.680516356702535e-06, + "loss": 0.2333503246307373, + "step": 8770, + "token_acc": 0.9336201102298696 + }, + { + "epoch": 0.5446589286822667, + "grad_norm": 0.8591123819351196, + "learning_rate": 4.675395367976783e-06, + "loss": 0.22921152114868165, + "step": 8775, + "token_acc": 0.9343863266052606 + }, + { + "epoch": 0.5449692756501769, + "grad_norm": 0.7953234314918518, + "learning_rate": 4.670274721173591e-06, + "loss": 0.23133621215820313, + "step": 8780, + "token_acc": 0.9339300825047598 + }, + { + "epoch": 0.545279622618087, + "grad_norm": 0.8676745891571045, + "learning_rate": 4.6651544216868035e-06, + "loss": 0.22861878871917723, + "step": 8785, + "token_acc": 0.9345684272191083 + }, + { + "epoch": 0.5455899695859971, + "grad_norm": 0.8397374749183655, + "learning_rate": 4.6600344749098885e-06, + "loss": 0.23129031658172608, + "step": 8790, + "token_acc": 0.9335689140531931 + }, + { + "epoch": 0.5459003165539072, + "grad_norm": 0.8353647589683533, + "learning_rate": 4.654914886235945e-06, + "loss": 0.23283302783966064, + "step": 8795, + "token_acc": 0.9339753059647886 + }, + { + "epoch": 0.5462106635218174, + "grad_norm": 0.9914921522140503, + "learning_rate": 4.649795661057699e-06, + "loss": 0.23629448413848878, + "step": 8800, + "token_acc": 0.932591702555656 + }, + { + "epoch": 0.5465210104897276, + "grad_norm": 0.9160581827163696, + "learning_rate": 4.64467680476749e-06, + "loss": 0.23420491218566894, + "step": 8805, + "token_acc": 0.9339306008786213 + }, + { + "epoch": 0.5468313574576377, + "grad_norm": 0.8121772408485413, + "learning_rate": 4.639558322757267e-06, + "loss": 0.22683205604553222, + "step": 8810, + "token_acc": 0.9344195574608946 + }, + { + "epoch": 0.5471417044255478, + "grad_norm": 0.7670088410377502, + "learning_rate": 4.634440220418591e-06, + "loss": 0.23064541816711426, + "step": 8815, + "token_acc": 0.9341344131257888 + }, + { + "epoch": 0.5474520513934579, + "grad_norm": 0.8074731230735779, + "learning_rate": 4.629322503142619e-06, + "loss": 0.22980527877807616, + "step": 8820, + "token_acc": 0.9346408732418904 + }, + { + "epoch": 0.547762398361368, + "grad_norm": 0.8520170450210571, + "learning_rate": 4.624205176320098e-06, + "loss": 0.2307342529296875, + "step": 8825, + "token_acc": 0.9339525626140321 + }, + { + "epoch": 0.5480727453292782, + "grad_norm": 0.8012826442718506, + "learning_rate": 4.6190882453413754e-06, + "loss": 0.2328421115875244, + "step": 8830, + "token_acc": 0.933793121499241 + }, + { + "epoch": 0.5483830922971883, + "grad_norm": 0.8000373244285583, + "learning_rate": 4.613971715596369e-06, + "loss": 0.23833394050598145, + "step": 8835, + "token_acc": 0.9322533521156153 + }, + { + "epoch": 0.5486934392650984, + "grad_norm": 0.8231528401374817, + "learning_rate": 4.608855592474585e-06, + "loss": 0.22967119216918946, + "step": 8840, + "token_acc": 0.934244343172063 + }, + { + "epoch": 0.5490037862330085, + "grad_norm": 0.8057606220245361, + "learning_rate": 4.603739881365092e-06, + "loss": 0.2327864646911621, + "step": 8845, + "token_acc": 0.9338939826744704 + }, + { + "epoch": 0.5493141332009186, + "grad_norm": 0.808185875415802, + "learning_rate": 4.598624587656528e-06, + "loss": 0.22434659004211427, + "step": 8850, + "token_acc": 0.9355642695733085 + }, + { + "epoch": 0.5496244801688287, + "grad_norm": 0.8256720900535583, + "learning_rate": 4.593509716737096e-06, + "loss": 0.23056917190551757, + "step": 8855, + "token_acc": 0.9347033825831831 + }, + { + "epoch": 0.5499348271367389, + "grad_norm": 0.837979793548584, + "learning_rate": 4.588395273994547e-06, + "loss": 0.23199942111968994, + "step": 8860, + "token_acc": 0.9341652640966667 + }, + { + "epoch": 0.550245174104649, + "grad_norm": 0.9255218505859375, + "learning_rate": 4.583281264816184e-06, + "loss": 0.2314289093017578, + "step": 8865, + "token_acc": 0.9331975533233147 + }, + { + "epoch": 0.5505555210725591, + "grad_norm": 0.8304197192192078, + "learning_rate": 4.5781676945888535e-06, + "loss": 0.2247575044631958, + "step": 8870, + "token_acc": 0.9357142383952408 + }, + { + "epoch": 0.5508658680404692, + "grad_norm": 0.8666803240776062, + "learning_rate": 4.573054568698938e-06, + "loss": 0.2326340913772583, + "step": 8875, + "token_acc": 0.9337747458533976 + }, + { + "epoch": 0.5511762150083793, + "grad_norm": 0.8482410907745361, + "learning_rate": 4.5679418925323524e-06, + "loss": 0.2324667453765869, + "step": 8880, + "token_acc": 0.9339230944376733 + }, + { + "epoch": 0.5514865619762895, + "grad_norm": 0.8537354469299316, + "learning_rate": 4.562829671474541e-06, + "loss": 0.22909798622131347, + "step": 8885, + "token_acc": 0.9342763543843228 + }, + { + "epoch": 0.5517969089441996, + "grad_norm": 0.7891256213188171, + "learning_rate": 4.557717910910465e-06, + "loss": 0.23165831565856934, + "step": 8890, + "token_acc": 0.9336792210275106 + }, + { + "epoch": 0.5521072559121097, + "grad_norm": 0.8280441761016846, + "learning_rate": 4.552606616224599e-06, + "loss": 0.22349858283996582, + "step": 8895, + "token_acc": 0.9358325515017772 + }, + { + "epoch": 0.5524176028800198, + "grad_norm": 0.8107667565345764, + "learning_rate": 4.547495792800934e-06, + "loss": 0.22554943561553956, + "step": 8900, + "token_acc": 0.9359554542476888 + }, + { + "epoch": 0.55272794984793, + "grad_norm": 0.8827427625656128, + "learning_rate": 4.542385446022958e-06, + "loss": 0.23208744525909425, + "step": 8905, + "token_acc": 0.9348644132468455 + }, + { + "epoch": 0.5530382968158402, + "grad_norm": 0.7992011308670044, + "learning_rate": 4.537275581273658e-06, + "loss": 0.22699120044708251, + "step": 8910, + "token_acc": 0.9352303213269951 + }, + { + "epoch": 0.5533486437837503, + "grad_norm": 0.8149877786636353, + "learning_rate": 4.532166203935516e-06, + "loss": 0.2283721923828125, + "step": 8915, + "token_acc": 0.9351928139380664 + }, + { + "epoch": 0.5536589907516604, + "grad_norm": 0.8204125761985779, + "learning_rate": 4.5270573193905e-06, + "loss": 0.23148202896118164, + "step": 8920, + "token_acc": 0.9336571834332092 + }, + { + "epoch": 0.5539693377195705, + "grad_norm": 0.889298677444458, + "learning_rate": 4.521948933020054e-06, + "loss": 0.2339005947113037, + "step": 8925, + "token_acc": 0.9326921175283965 + }, + { + "epoch": 0.5542796846874806, + "grad_norm": 0.8517132997512817, + "learning_rate": 4.516841050205105e-06, + "loss": 0.23078792095184325, + "step": 8930, + "token_acc": 0.9339216170900662 + }, + { + "epoch": 0.5545900316553908, + "grad_norm": 0.8451063632965088, + "learning_rate": 4.511733676326044e-06, + "loss": 0.2331632614135742, + "step": 8935, + "token_acc": 0.9341835016279116 + }, + { + "epoch": 0.5549003786233009, + "grad_norm": 0.828180193901062, + "learning_rate": 4.506626816762726e-06, + "loss": 0.22904691696166993, + "step": 8940, + "token_acc": 0.9342883384101015 + }, + { + "epoch": 0.555210725591211, + "grad_norm": 0.852012038230896, + "learning_rate": 4.5015204768944695e-06, + "loss": 0.2280498743057251, + "step": 8945, + "token_acc": 0.9345975361399526 + }, + { + "epoch": 0.5555210725591211, + "grad_norm": 0.8231121301651001, + "learning_rate": 4.4964146621000396e-06, + "loss": 0.23037199974060057, + "step": 8950, + "token_acc": 0.9332355099046221 + }, + { + "epoch": 0.5558314195270312, + "grad_norm": 0.7989760041236877, + "learning_rate": 4.491309377757651e-06, + "loss": 0.2290328025817871, + "step": 8955, + "token_acc": 0.9346614747220474 + }, + { + "epoch": 0.5561417664949413, + "grad_norm": 0.8593499660491943, + "learning_rate": 4.486204629244959e-06, + "loss": 0.23122825622558593, + "step": 8960, + "token_acc": 0.9339398067287527 + }, + { + "epoch": 0.5564521134628515, + "grad_norm": 0.8080406188964844, + "learning_rate": 4.481100421939056e-06, + "loss": 0.22796289920806884, + "step": 8965, + "token_acc": 0.9347606986088824 + }, + { + "epoch": 0.5567624604307616, + "grad_norm": 0.8840007781982422, + "learning_rate": 4.475996761216461e-06, + "loss": 0.228739333152771, + "step": 8970, + "token_acc": 0.934615997662248 + }, + { + "epoch": 0.5570728073986717, + "grad_norm": 0.818583071231842, + "learning_rate": 4.470893652453121e-06, + "loss": 0.22481937408447267, + "step": 8975, + "token_acc": 0.9349549764921508 + }, + { + "epoch": 0.5573831543665818, + "grad_norm": 0.796339213848114, + "learning_rate": 4.465791101024399e-06, + "loss": 0.22626852989196777, + "step": 8980, + "token_acc": 0.9354293287026458 + }, + { + "epoch": 0.5576935013344919, + "grad_norm": 0.7818427681922913, + "learning_rate": 4.460689112305072e-06, + "loss": 0.2274632453918457, + "step": 8985, + "token_acc": 0.9343415298919695 + }, + { + "epoch": 0.5580038483024021, + "grad_norm": 0.7855643033981323, + "learning_rate": 4.455587691669325e-06, + "loss": 0.2300940990447998, + "step": 8990, + "token_acc": 0.9343333466490898 + }, + { + "epoch": 0.5583141952703122, + "grad_norm": 0.8040710687637329, + "learning_rate": 4.450486844490742e-06, + "loss": 0.23637642860412597, + "step": 8995, + "token_acc": 0.932512878469302 + }, + { + "epoch": 0.5586245422382223, + "grad_norm": 0.8174216151237488, + "learning_rate": 4.445386576142308e-06, + "loss": 0.22736213207244874, + "step": 9000, + "token_acc": 0.9338073057121844 + }, + { + "epoch": 0.5586245422382223, + "eval_loss": 0.24459242820739746, + "eval_runtime": 102.881, + "eval_samples_per_second": 194.399, + "eval_steps_per_second": 0.262, + "eval_token_acc": 0.9298316897469849, + "step": 9000 + }, + { + "epoch": 0.5589348892061324, + "grad_norm": 1.1239585876464844, + "learning_rate": 4.4402868919963924e-06, + "loss": 0.22991557121276857, + "step": 9005, + "token_acc": 0.9345034105240916 + }, + { + "epoch": 0.5592452361740425, + "grad_norm": 0.8264209628105164, + "learning_rate": 4.435187797424749e-06, + "loss": 0.23252553939819337, + "step": 9010, + "token_acc": 0.9344174851145827 + }, + { + "epoch": 0.5595555831419528, + "grad_norm": 0.821744441986084, + "learning_rate": 4.4300892977985214e-06, + "loss": 0.23172032833099365, + "step": 9015, + "token_acc": 0.9345834079795104 + }, + { + "epoch": 0.5598659301098629, + "grad_norm": 0.8167038559913635, + "learning_rate": 4.424991398488211e-06, + "loss": 0.23257486820220946, + "step": 9020, + "token_acc": 0.9337116259609853 + }, + { + "epoch": 0.560176277077773, + "grad_norm": 0.8388352990150452, + "learning_rate": 4.419894104863696e-06, + "loss": 0.22905364036560058, + "step": 9025, + "token_acc": 0.9332674672271917 + }, + { + "epoch": 0.5604866240456831, + "grad_norm": 0.8157934546470642, + "learning_rate": 4.4147974222942176e-06, + "loss": 0.2253563404083252, + "step": 9030, + "token_acc": 0.9349206983340658 + }, + { + "epoch": 0.5607969710135932, + "grad_norm": 0.8413422703742981, + "learning_rate": 4.4097013561483686e-06, + "loss": 0.22785816192626954, + "step": 9035, + "token_acc": 0.9350366696565188 + }, + { + "epoch": 0.5611073179815034, + "grad_norm": 0.8079880475997925, + "learning_rate": 4.404605911794094e-06, + "loss": 0.2230680465698242, + "step": 9040, + "token_acc": 0.9357281411021271 + }, + { + "epoch": 0.5614176649494135, + "grad_norm": 0.8916932344436646, + "learning_rate": 4.399511094598688e-06, + "loss": 0.22941977977752687, + "step": 9045, + "token_acc": 0.9347612797157365 + }, + { + "epoch": 0.5617280119173236, + "grad_norm": 0.7849769592285156, + "learning_rate": 4.394416909928777e-06, + "loss": 0.22711243629455566, + "step": 9050, + "token_acc": 0.9341552833465298 + }, + { + "epoch": 0.5620383588852337, + "grad_norm": 0.9089765548706055, + "learning_rate": 4.3893233631503236e-06, + "loss": 0.22680864334106446, + "step": 9055, + "token_acc": 0.9354855912487584 + }, + { + "epoch": 0.5623487058531438, + "grad_norm": 0.8886991739273071, + "learning_rate": 4.384230459628625e-06, + "loss": 0.23742284774780273, + "step": 9060, + "token_acc": 0.9321378246882727 + }, + { + "epoch": 0.5626590528210539, + "grad_norm": 0.9265754818916321, + "learning_rate": 4.37913820472829e-06, + "loss": 0.23734831809997559, + "step": 9065, + "token_acc": 0.9325193531796656 + }, + { + "epoch": 0.5629693997889641, + "grad_norm": 0.7836823463439941, + "learning_rate": 4.374046603813252e-06, + "loss": 0.2323366165161133, + "step": 9070, + "token_acc": 0.9344196311775084 + }, + { + "epoch": 0.5632797467568742, + "grad_norm": 0.8154056668281555, + "learning_rate": 4.368955662246754e-06, + "loss": 0.22613677978515626, + "step": 9075, + "token_acc": 0.9352258185545737 + }, + { + "epoch": 0.5635900937247843, + "grad_norm": 0.7866949439048767, + "learning_rate": 4.363865385391342e-06, + "loss": 0.23787765502929686, + "step": 9080, + "token_acc": 0.9329953978523311 + }, + { + "epoch": 0.5639004406926944, + "grad_norm": 0.9008229970932007, + "learning_rate": 4.358775778608864e-06, + "loss": 0.22830286026000976, + "step": 9085, + "token_acc": 0.9349117100983633 + }, + { + "epoch": 0.5642107876606045, + "grad_norm": 0.8126014471054077, + "learning_rate": 4.353686847260461e-06, + "loss": 0.22855396270751954, + "step": 9090, + "token_acc": 0.9346957529266051 + }, + { + "epoch": 0.5645211346285147, + "grad_norm": 0.8488256335258484, + "learning_rate": 4.348598596706566e-06, + "loss": 0.23272366523742677, + "step": 9095, + "token_acc": 0.9339128237187587 + }, + { + "epoch": 0.5648314815964248, + "grad_norm": 0.8108824491500854, + "learning_rate": 4.343511032306886e-06, + "loss": 0.22617745399475098, + "step": 9100, + "token_acc": 0.9351484883378411 + }, + { + "epoch": 0.5651418285643349, + "grad_norm": 0.8260208964347839, + "learning_rate": 4.338424159420418e-06, + "loss": 0.23081674575805664, + "step": 9105, + "token_acc": 0.9341768797615104 + }, + { + "epoch": 0.565452175532245, + "grad_norm": 0.8169665336608887, + "learning_rate": 4.33333798340542e-06, + "loss": 0.2352761745452881, + "step": 9110, + "token_acc": 0.9333386207717016 + }, + { + "epoch": 0.5657625225001551, + "grad_norm": 0.9474765658378601, + "learning_rate": 4.328252509619419e-06, + "loss": 0.2281942844390869, + "step": 9115, + "token_acc": 0.9343251136051126 + }, + { + "epoch": 0.5660728694680653, + "grad_norm": 0.8583059310913086, + "learning_rate": 4.323167743419207e-06, + "loss": 0.23068611621856688, + "step": 9120, + "token_acc": 0.9340562189821109 + }, + { + "epoch": 0.5663832164359754, + "grad_norm": 0.9009811282157898, + "learning_rate": 4.318083690160823e-06, + "loss": 0.23104567527770997, + "step": 9125, + "token_acc": 0.9337089532529272 + }, + { + "epoch": 0.5666935634038855, + "grad_norm": 1.0731709003448486, + "learning_rate": 4.3130003551995615e-06, + "loss": 0.22949738502502443, + "step": 9130, + "token_acc": 0.9343058873777078 + }, + { + "epoch": 0.5670039103717956, + "grad_norm": 0.8140671849250793, + "learning_rate": 4.307917743889956e-06, + "loss": 0.2268454074859619, + "step": 9135, + "token_acc": 0.9345533087748125 + }, + { + "epoch": 0.5673142573397058, + "grad_norm": 0.8400377035140991, + "learning_rate": 4.302835861585779e-06, + "loss": 0.23182573318481445, + "step": 9140, + "token_acc": 0.9335376769565131 + }, + { + "epoch": 0.567624604307616, + "grad_norm": 0.78734290599823, + "learning_rate": 4.297754713640038e-06, + "loss": 0.2273392915725708, + "step": 9145, + "token_acc": 0.9355483207452 + }, + { + "epoch": 0.5679349512755261, + "grad_norm": 0.7908815145492554, + "learning_rate": 4.292674305404964e-06, + "loss": 0.23374857902526855, + "step": 9150, + "token_acc": 0.9332267199733466 + }, + { + "epoch": 0.5682452982434362, + "grad_norm": 0.8390881419181824, + "learning_rate": 4.287594642232004e-06, + "loss": 0.22717437744140626, + "step": 9155, + "token_acc": 0.9348063485257565 + }, + { + "epoch": 0.5685556452113463, + "grad_norm": 0.8568211793899536, + "learning_rate": 4.282515729471834e-06, + "loss": 0.23121514320373535, + "step": 9160, + "token_acc": 0.9335785446773729 + }, + { + "epoch": 0.5688659921792564, + "grad_norm": 0.8523248434066772, + "learning_rate": 4.277437572474325e-06, + "loss": 0.22910017967224122, + "step": 9165, + "token_acc": 0.9352207743508015 + }, + { + "epoch": 0.5691763391471665, + "grad_norm": 0.9389868974685669, + "learning_rate": 4.2723601765885595e-06, + "loss": 0.2342961311340332, + "step": 9170, + "token_acc": 0.9341335628740486 + }, + { + "epoch": 0.5694866861150767, + "grad_norm": 0.8420261740684509, + "learning_rate": 4.267283547162818e-06, + "loss": 0.2282135009765625, + "step": 9175, + "token_acc": 0.9336437484969137 + }, + { + "epoch": 0.5697970330829868, + "grad_norm": 0.9319453835487366, + "learning_rate": 4.262207689544571e-06, + "loss": 0.22366180419921874, + "step": 9180, + "token_acc": 0.9356369469322283 + }, + { + "epoch": 0.5701073800508969, + "grad_norm": 0.9778652787208557, + "learning_rate": 4.257132609080477e-06, + "loss": 0.22629098892211913, + "step": 9185, + "token_acc": 0.9345798701213167 + }, + { + "epoch": 0.570417727018807, + "grad_norm": 0.8691847920417786, + "learning_rate": 4.252058311116377e-06, + "loss": 0.22122032642364503, + "step": 9190, + "token_acc": 0.9369269554833145 + }, + { + "epoch": 0.5707280739867171, + "grad_norm": 0.7937987446784973, + "learning_rate": 4.24698480099729e-06, + "loss": 0.231732177734375, + "step": 9195, + "token_acc": 0.9338472393487594 + }, + { + "epoch": 0.5710384209546273, + "grad_norm": 1.1790599822998047, + "learning_rate": 4.241912084067395e-06, + "loss": 0.23075084686279296, + "step": 9200, + "token_acc": 0.9338223179981943 + }, + { + "epoch": 0.5713487679225374, + "grad_norm": 0.9806782603263855, + "learning_rate": 4.23684016567005e-06, + "loss": 0.23424811363220216, + "step": 9205, + "token_acc": 0.9328641659371001 + }, + { + "epoch": 0.5716591148904475, + "grad_norm": 1.125106692314148, + "learning_rate": 4.231769051147761e-06, + "loss": 0.23172032833099365, + "step": 9210, + "token_acc": 0.9325986519730395 + }, + { + "epoch": 0.5719694618583576, + "grad_norm": 1.8347450494766235, + "learning_rate": 4.22669874584219e-06, + "loss": 0.2324674129486084, + "step": 9215, + "token_acc": 0.9336546322424114 + }, + { + "epoch": 0.5722798088262677, + "grad_norm": 0.8720211982727051, + "learning_rate": 4.22162925509415e-06, + "loss": 0.22626721858978271, + "step": 9220, + "token_acc": 0.9356979495789431 + }, + { + "epoch": 0.5725901557941779, + "grad_norm": 0.99982088804245, + "learning_rate": 4.216560584243592e-06, + "loss": 0.22560586929321289, + "step": 9225, + "token_acc": 0.934439450026441 + }, + { + "epoch": 0.572900502762088, + "grad_norm": 0.8416867852210999, + "learning_rate": 4.211492738629604e-06, + "loss": 0.2275064468383789, + "step": 9230, + "token_acc": 0.9351603514220224 + }, + { + "epoch": 0.5732108497299981, + "grad_norm": 0.810532808303833, + "learning_rate": 4.206425723590407e-06, + "loss": 0.2273815393447876, + "step": 9235, + "token_acc": 0.9362571163775983 + }, + { + "epoch": 0.5735211966979082, + "grad_norm": 1.7141177654266357, + "learning_rate": 4.2013595444633456e-06, + "loss": 0.24009456634521484, + "step": 9240, + "token_acc": 0.9324147155419006 + }, + { + "epoch": 0.5738315436658183, + "grad_norm": 0.8249492049217224, + "learning_rate": 4.19629420658488e-06, + "loss": 0.23302578926086426, + "step": 9245, + "token_acc": 0.9336143591671253 + }, + { + "epoch": 0.5741418906337286, + "grad_norm": 0.8176672458648682, + "learning_rate": 4.191229715290594e-06, + "loss": 0.22687263488769532, + "step": 9250, + "token_acc": 0.9348780568080834 + }, + { + "epoch": 0.5744522376016387, + "grad_norm": 4.28188419342041, + "learning_rate": 4.186166075915169e-06, + "loss": 0.23382329940795898, + "step": 9255, + "token_acc": 0.933662988227238 + }, + { + "epoch": 0.5747625845695488, + "grad_norm": 0.8345016241073608, + "learning_rate": 4.181103293792398e-06, + "loss": 0.23179829120635986, + "step": 9260, + "token_acc": 0.9341337411956229 + }, + { + "epoch": 0.5750729315374589, + "grad_norm": 0.7732263803482056, + "learning_rate": 4.1760413742551635e-06, + "loss": 0.23265552520751953, + "step": 9265, + "token_acc": 0.9337390668129968 + }, + { + "epoch": 0.575383278505369, + "grad_norm": 0.8092970848083496, + "learning_rate": 4.170980322635443e-06, + "loss": 0.22910819053649903, + "step": 9270, + "token_acc": 0.9341699691765742 + }, + { + "epoch": 0.5756936254732792, + "grad_norm": 0.9862002730369568, + "learning_rate": 4.165920144264302e-06, + "loss": 0.23544142246246338, + "step": 9275, + "token_acc": 0.9321143772313236 + }, + { + "epoch": 0.5760039724411893, + "grad_norm": 0.8287213444709778, + "learning_rate": 4.160860844471881e-06, + "loss": 0.23287198543548585, + "step": 9280, + "token_acc": 0.9333614057601947 + }, + { + "epoch": 0.5763143194090994, + "grad_norm": 0.8162459135055542, + "learning_rate": 4.155802428587398e-06, + "loss": 0.23536953926086426, + "step": 9285, + "token_acc": 0.933598149839044 + }, + { + "epoch": 0.5766246663770095, + "grad_norm": 0.8036437034606934, + "learning_rate": 4.150744901939142e-06, + "loss": 0.22849106788635254, + "step": 9290, + "token_acc": 0.9352258638348337 + }, + { + "epoch": 0.5769350133449196, + "grad_norm": 0.8458849787712097, + "learning_rate": 4.145688269854462e-06, + "loss": 0.23709697723388673, + "step": 9295, + "token_acc": 0.9327706883343918 + }, + { + "epoch": 0.5772453603128297, + "grad_norm": 0.8071736693382263, + "learning_rate": 4.140632537659763e-06, + "loss": 0.2311241626739502, + "step": 9300, + "token_acc": 0.9330399539867043 + }, + { + "epoch": 0.5775557072807399, + "grad_norm": 0.8369570970535278, + "learning_rate": 4.135577710680508e-06, + "loss": 0.22927205562591552, + "step": 9305, + "token_acc": 0.9339606261352709 + }, + { + "epoch": 0.57786605424865, + "grad_norm": 0.8366904854774475, + "learning_rate": 4.130523794241201e-06, + "loss": 0.22690768241882325, + "step": 9310, + "token_acc": 0.9345764645031233 + }, + { + "epoch": 0.5781764012165601, + "grad_norm": 0.8189626336097717, + "learning_rate": 4.12547079366539e-06, + "loss": 0.22725090980529786, + "step": 9315, + "token_acc": 0.9347863656908884 + }, + { + "epoch": 0.5784867481844702, + "grad_norm": 0.8228508830070496, + "learning_rate": 4.120418714275659e-06, + "loss": 0.23235664367675782, + "step": 9320, + "token_acc": 0.9331531806683354 + }, + { + "epoch": 0.5787970951523803, + "grad_norm": 0.8448265194892883, + "learning_rate": 4.115367561393619e-06, + "loss": 0.23340373039245604, + "step": 9325, + "token_acc": 0.9336065242843877 + }, + { + "epoch": 0.5791074421202905, + "grad_norm": 0.8817808032035828, + "learning_rate": 4.110317340339904e-06, + "loss": 0.23156075477600097, + "step": 9330, + "token_acc": 0.9333676492909135 + }, + { + "epoch": 0.5794177890882006, + "grad_norm": 0.8227306008338928, + "learning_rate": 4.105268056434173e-06, + "loss": 0.22905361652374268, + "step": 9335, + "token_acc": 0.9338537414736501 + }, + { + "epoch": 0.5797281360561107, + "grad_norm": 0.7918829917907715, + "learning_rate": 4.1002197149950895e-06, + "loss": 0.22300517559051514, + "step": 9340, + "token_acc": 0.9356622587754663 + }, + { + "epoch": 0.5800384830240208, + "grad_norm": 0.8718773722648621, + "learning_rate": 4.0951723213403294e-06, + "loss": 0.22995119094848632, + "step": 9345, + "token_acc": 0.9343908689507813 + }, + { + "epoch": 0.5803488299919309, + "grad_norm": 0.8321003913879395, + "learning_rate": 4.09012588078657e-06, + "loss": 0.23628082275390624, + "step": 9350, + "token_acc": 0.9330909139701985 + }, + { + "epoch": 0.5806591769598412, + "grad_norm": 1.141156792640686, + "learning_rate": 4.085080398649483e-06, + "loss": 0.226725172996521, + "step": 9355, + "token_acc": 0.934943338396806 + }, + { + "epoch": 0.5809695239277513, + "grad_norm": 0.7766085863113403, + "learning_rate": 4.08003588024373e-06, + "loss": 0.22661643028259276, + "step": 9360, + "token_acc": 0.9350473160821214 + }, + { + "epoch": 0.5812798708956614, + "grad_norm": 0.881335437297821, + "learning_rate": 4.074992330882962e-06, + "loss": 0.23018550872802734, + "step": 9365, + "token_acc": 0.933763334160258 + }, + { + "epoch": 0.5815902178635715, + "grad_norm": 2.459913969039917, + "learning_rate": 4.0699497558798054e-06, + "loss": 0.22696001529693605, + "step": 9370, + "token_acc": 0.9344455322414204 + }, + { + "epoch": 0.5819005648314816, + "grad_norm": 0.8811545372009277, + "learning_rate": 4.064908160545855e-06, + "loss": 0.2315882682800293, + "step": 9375, + "token_acc": 0.9345572483136901 + }, + { + "epoch": 0.5822109117993918, + "grad_norm": 0.7892197966575623, + "learning_rate": 4.059867550191689e-06, + "loss": 0.23431532382965087, + "step": 9380, + "token_acc": 0.9339646217335216 + }, + { + "epoch": 0.5825212587673019, + "grad_norm": 0.8272483348846436, + "learning_rate": 4.054827930126831e-06, + "loss": 0.23492145538330078, + "step": 9385, + "token_acc": 0.9326001070700088 + }, + { + "epoch": 0.582831605735212, + "grad_norm": 0.8774070739746094, + "learning_rate": 4.049789305659771e-06, + "loss": 0.23410727977752685, + "step": 9390, + "token_acc": 0.9338243092750703 + }, + { + "epoch": 0.5831419527031221, + "grad_norm": 0.8542326092720032, + "learning_rate": 4.0447516820979505e-06, + "loss": 0.229461932182312, + "step": 9395, + "token_acc": 0.9344574326414764 + }, + { + "epoch": 0.5834522996710322, + "grad_norm": 0.8478512167930603, + "learning_rate": 4.0397150647477514e-06, + "loss": 0.22983813285827637, + "step": 9400, + "token_acc": 0.9346266775711585 + }, + { + "epoch": 0.5837626466389423, + "grad_norm": 0.8374636173248291, + "learning_rate": 4.034679458914502e-06, + "loss": 0.22826766967773438, + "step": 9405, + "token_acc": 0.9343593237547508 + }, + { + "epoch": 0.5840729936068525, + "grad_norm": 0.802219808101654, + "learning_rate": 4.029644869902461e-06, + "loss": 0.23269543647766114, + "step": 9410, + "token_acc": 0.9337998025457027 + }, + { + "epoch": 0.5843833405747626, + "grad_norm": 0.8658347129821777, + "learning_rate": 4.024611303014813e-06, + "loss": 0.22846050262451173, + "step": 9415, + "token_acc": 0.9347668407448436 + }, + { + "epoch": 0.5846936875426727, + "grad_norm": 0.7964611649513245, + "learning_rate": 4.019578763553677e-06, + "loss": 0.22863361835479737, + "step": 9420, + "token_acc": 0.9347591601209658 + }, + { + "epoch": 0.5850040345105828, + "grad_norm": 0.8660332560539246, + "learning_rate": 4.014547256820079e-06, + "loss": 0.2362675905227661, + "step": 9425, + "token_acc": 0.9327047838872262 + }, + { + "epoch": 0.5853143814784929, + "grad_norm": 0.870186448097229, + "learning_rate": 4.009516788113959e-06, + "loss": 0.22697687149047852, + "step": 9430, + "token_acc": 0.9345723509789313 + }, + { + "epoch": 0.5856247284464031, + "grad_norm": 0.8587538599967957, + "learning_rate": 4.004487362734169e-06, + "loss": 0.23304481506347657, + "step": 9435, + "token_acc": 0.9334622437072846 + }, + { + "epoch": 0.5859350754143132, + "grad_norm": 0.8881959319114685, + "learning_rate": 3.999458985978456e-06, + "loss": 0.23113133907318115, + "step": 9440, + "token_acc": 0.9342130836611733 + }, + { + "epoch": 0.5862454223822233, + "grad_norm": 0.8575989007949829, + "learning_rate": 3.994431663143466e-06, + "loss": 0.2352227210998535, + "step": 9445, + "token_acc": 0.9332666353053423 + }, + { + "epoch": 0.5865557693501334, + "grad_norm": 1.0102519989013672, + "learning_rate": 3.989405399524734e-06, + "loss": 0.23994998931884765, + "step": 9450, + "token_acc": 0.9312512315270935 + }, + { + "epoch": 0.5868661163180435, + "grad_norm": 1.0060070753097534, + "learning_rate": 3.984380200416678e-06, + "loss": 0.23497467041015624, + "step": 9455, + "token_acc": 0.9325551059525077 + }, + { + "epoch": 0.5871764632859537, + "grad_norm": 0.8226820826530457, + "learning_rate": 3.979356071112595e-06, + "loss": 0.22551488876342773, + "step": 9460, + "token_acc": 0.9353058439473403 + }, + { + "epoch": 0.5874868102538638, + "grad_norm": 0.8270158171653748, + "learning_rate": 3.97433301690466e-06, + "loss": 0.23512768745422363, + "step": 9465, + "token_acc": 0.9337269104421347 + }, + { + "epoch": 0.587797157221774, + "grad_norm": 0.8284282684326172, + "learning_rate": 3.969311043083908e-06, + "loss": 0.22747163772583007, + "step": 9470, + "token_acc": 0.9346086910488945 + }, + { + "epoch": 0.588107504189684, + "grad_norm": 0.8961857557296753, + "learning_rate": 3.964290154940237e-06, + "loss": 0.2291252851486206, + "step": 9475, + "token_acc": 0.9346848665966669 + }, + { + "epoch": 0.5884178511575942, + "grad_norm": 0.8794692158699036, + "learning_rate": 3.95927035776241e-06, + "loss": 0.23191909790039061, + "step": 9480, + "token_acc": 0.9327881099879011 + }, + { + "epoch": 0.5887281981255044, + "grad_norm": 0.8385714292526245, + "learning_rate": 3.95425165683803e-06, + "loss": 0.23155980110168456, + "step": 9485, + "token_acc": 0.93375593662281 + }, + { + "epoch": 0.5890385450934145, + "grad_norm": 0.8067603707313538, + "learning_rate": 3.94923405745355e-06, + "loss": 0.23496341705322266, + "step": 9490, + "token_acc": 0.9332339352932131 + }, + { + "epoch": 0.5893488920613246, + "grad_norm": 0.9133737087249756, + "learning_rate": 3.944217564894266e-06, + "loss": 0.22930498123168946, + "step": 9495, + "token_acc": 0.9344056891296986 + }, + { + "epoch": 0.5896592390292347, + "grad_norm": 0.7739179730415344, + "learning_rate": 3.939202184444302e-06, + "loss": 0.22050492763519286, + "step": 9500, + "token_acc": 0.9361052010631459 + }, + { + "epoch": 0.5896592390292347, + "eval_loss": 0.24304163455963135, + "eval_runtime": 94.6932, + "eval_samples_per_second": 211.208, + "eval_steps_per_second": 0.285, + "eval_token_acc": 0.930909621074872, + "step": 9500 + }, + { + "epoch": 0.5899695859971448, + "grad_norm": 0.9246335625648499, + "learning_rate": 3.934187921386611e-06, + "loss": 0.22888927459716796, + "step": 9505, + "token_acc": 0.933708526928866 + }, + { + "epoch": 0.5902799329650549, + "grad_norm": 0.8361711502075195, + "learning_rate": 3.929174781002976e-06, + "loss": 0.23505239486694335, + "step": 9510, + "token_acc": 0.9327394299064166 + }, + { + "epoch": 0.5905902799329651, + "grad_norm": 1.0340321063995361, + "learning_rate": 3.924162768573992e-06, + "loss": 0.22838084697723388, + "step": 9515, + "token_acc": 0.9347591315351845 + }, + { + "epoch": 0.5909006269008752, + "grad_norm": 0.9025611281394958, + "learning_rate": 3.919151889379062e-06, + "loss": 0.23694329261779784, + "step": 9520, + "token_acc": 0.9327621911809321 + }, + { + "epoch": 0.5912109738687853, + "grad_norm": 0.8649851679801941, + "learning_rate": 3.914142148696405e-06, + "loss": 0.23120901584625245, + "step": 9525, + "token_acc": 0.9341191635824823 + }, + { + "epoch": 0.5915213208366954, + "grad_norm": 0.9120691418647766, + "learning_rate": 3.90913355180303e-06, + "loss": 0.22827553749084473, + "step": 9530, + "token_acc": 0.93457498893375 + }, + { + "epoch": 0.5918316678046055, + "grad_norm": 0.8227366805076599, + "learning_rate": 3.9041261039747545e-06, + "loss": 0.22546060085296632, + "step": 9535, + "token_acc": 0.9351356025783448 + }, + { + "epoch": 0.5921420147725157, + "grad_norm": 0.8528803586959839, + "learning_rate": 3.8991198104861725e-06, + "loss": 0.23184022903442383, + "step": 9540, + "token_acc": 0.9329006325777544 + }, + { + "epoch": 0.5924523617404258, + "grad_norm": 0.8234803080558777, + "learning_rate": 3.894114676610667e-06, + "loss": 0.22396512031555177, + "step": 9545, + "token_acc": 0.9355569020968311 + }, + { + "epoch": 0.5927627087083359, + "grad_norm": 0.814703106880188, + "learning_rate": 3.889110707620403e-06, + "loss": 0.22951579093933105, + "step": 9550, + "token_acc": 0.934275443329473 + }, + { + "epoch": 0.593073055676246, + "grad_norm": 0.8786909580230713, + "learning_rate": 3.8841079087863135e-06, + "loss": 0.23120391368865967, + "step": 9555, + "token_acc": 0.9347757475640793 + }, + { + "epoch": 0.5933834026441561, + "grad_norm": 0.9590782523155212, + "learning_rate": 3.879106285378101e-06, + "loss": 0.22567338943481446, + "step": 9560, + "token_acc": 0.9351830340844977 + }, + { + "epoch": 0.5936937496120663, + "grad_norm": 0.8622130751609802, + "learning_rate": 3.874105842664231e-06, + "loss": 0.2305859088897705, + "step": 9565, + "token_acc": 0.9338430561966112 + }, + { + "epoch": 0.5940040965799764, + "grad_norm": 0.7857866287231445, + "learning_rate": 3.8691065859119246e-06, + "loss": 0.22563796043395995, + "step": 9570, + "token_acc": 0.9344888269083906 + }, + { + "epoch": 0.5943144435478865, + "grad_norm": 0.8566865921020508, + "learning_rate": 3.864108520387148e-06, + "loss": 0.22488029003143312, + "step": 9575, + "token_acc": 0.9356083688807637 + }, + { + "epoch": 0.5946247905157966, + "grad_norm": 0.7943798899650574, + "learning_rate": 3.859111651354625e-06, + "loss": 0.23255114555358886, + "step": 9580, + "token_acc": 0.9330137230205153 + }, + { + "epoch": 0.5949351374837067, + "grad_norm": 0.832381546497345, + "learning_rate": 3.854115984077807e-06, + "loss": 0.2276665210723877, + "step": 9585, + "token_acc": 0.9339486989192448 + }, + { + "epoch": 0.595245484451617, + "grad_norm": 0.8717218041419983, + "learning_rate": 3.849121523818885e-06, + "loss": 0.23063898086547852, + "step": 9590, + "token_acc": 0.9344574435989407 + }, + { + "epoch": 0.5955558314195271, + "grad_norm": 0.8205404281616211, + "learning_rate": 3.84412827583878e-06, + "loss": 0.22849888801574708, + "step": 9595, + "token_acc": 0.9345596657907149 + }, + { + "epoch": 0.5958661783874372, + "grad_norm": 0.8685084581375122, + "learning_rate": 3.839136245397132e-06, + "loss": 0.22054760456085204, + "step": 9600, + "token_acc": 0.9368830252481212 + }, + { + "epoch": 0.5961765253553473, + "grad_norm": 0.7974231839179993, + "learning_rate": 3.834145437752301e-06, + "loss": 0.22798440456390381, + "step": 9605, + "token_acc": 0.9348696244996592 + }, + { + "epoch": 0.5964868723232574, + "grad_norm": 0.7721987962722778, + "learning_rate": 3.829155858161358e-06, + "loss": 0.22276594638824462, + "step": 9610, + "token_acc": 0.9362517855699 + }, + { + "epoch": 0.5967972192911676, + "grad_norm": 0.8318093419075012, + "learning_rate": 3.8241675118800835e-06, + "loss": 0.22782211303710936, + "step": 9615, + "token_acc": 0.934806593114449 + }, + { + "epoch": 0.5971075662590777, + "grad_norm": 0.9867302775382996, + "learning_rate": 3.819180404162951e-06, + "loss": 0.23052706718444824, + "step": 9620, + "token_acc": 0.9341657971794958 + }, + { + "epoch": 0.5974179132269878, + "grad_norm": 0.8443397283554077, + "learning_rate": 3.814194540263141e-06, + "loss": 0.23027212619781495, + "step": 9625, + "token_acc": 0.9335867886420339 + }, + { + "epoch": 0.5977282601948979, + "grad_norm": 0.8706814050674438, + "learning_rate": 3.8092099254325126e-06, + "loss": 0.22571053504943847, + "step": 9630, + "token_acc": 0.9353977329059123 + }, + { + "epoch": 0.598038607162808, + "grad_norm": 0.8627851009368896, + "learning_rate": 3.804226564921615e-06, + "loss": 0.22642278671264648, + "step": 9635, + "token_acc": 0.9353339275692449 + }, + { + "epoch": 0.5983489541307181, + "grad_norm": 0.8786664009094238, + "learning_rate": 3.799244463979677e-06, + "loss": 0.2261965751647949, + "step": 9640, + "token_acc": 0.9346506296263926 + }, + { + "epoch": 0.5986593010986283, + "grad_norm": 0.7918584942817688, + "learning_rate": 3.794263627854598e-06, + "loss": 0.22252535820007324, + "step": 9645, + "token_acc": 0.9363410629180183 + }, + { + "epoch": 0.5989696480665384, + "grad_norm": 0.7984991669654846, + "learning_rate": 3.7892840617929444e-06, + "loss": 0.22610046863555908, + "step": 9650, + "token_acc": 0.935295058798432 + }, + { + "epoch": 0.5992799950344485, + "grad_norm": 0.7930071353912354, + "learning_rate": 3.784305771039949e-06, + "loss": 0.22710647583007812, + "step": 9655, + "token_acc": 0.9350897900120698 + }, + { + "epoch": 0.5995903420023586, + "grad_norm": 0.9025065302848816, + "learning_rate": 3.779328760839498e-06, + "loss": 0.23568120002746581, + "step": 9660, + "token_acc": 0.9334780328826408 + }, + { + "epoch": 0.5999006889702687, + "grad_norm": 0.8414742946624756, + "learning_rate": 3.774353036434126e-06, + "loss": 0.2313474178314209, + "step": 9665, + "token_acc": 0.934241864697067 + }, + { + "epoch": 0.6002110359381789, + "grad_norm": 0.825603723526001, + "learning_rate": 3.769378603065021e-06, + "loss": 0.23174781799316407, + "step": 9670, + "token_acc": 0.9341969291900288 + }, + { + "epoch": 0.600521382906089, + "grad_norm": 0.8271713852882385, + "learning_rate": 3.764405465972003e-06, + "loss": 0.2258908748626709, + "step": 9675, + "token_acc": 0.9351155292853305 + }, + { + "epoch": 0.6008317298739991, + "grad_norm": 0.8338208198547363, + "learning_rate": 3.7594336303935325e-06, + "loss": 0.23302416801452636, + "step": 9680, + "token_acc": 0.9323961501164333 + }, + { + "epoch": 0.6011420768419092, + "grad_norm": 0.8356338143348694, + "learning_rate": 3.7544631015666965e-06, + "loss": 0.22943170070648194, + "step": 9685, + "token_acc": 0.934873481405447 + }, + { + "epoch": 0.6014524238098193, + "grad_norm": 0.8436577916145325, + "learning_rate": 3.7494938847272037e-06, + "loss": 0.22374238967895507, + "step": 9690, + "token_acc": 0.9359547978055911 + }, + { + "epoch": 0.6017627707777295, + "grad_norm": 0.7810449600219727, + "learning_rate": 3.744525985109386e-06, + "loss": 0.23103179931640624, + "step": 9695, + "token_acc": 0.9347649703386449 + }, + { + "epoch": 0.6020731177456397, + "grad_norm": 0.854554295539856, + "learning_rate": 3.7395594079461832e-06, + "loss": 0.22460105419158935, + "step": 9700, + "token_acc": 0.9350939185419379 + }, + { + "epoch": 0.6023834647135498, + "grad_norm": 0.7943915724754333, + "learning_rate": 3.734594158469141e-06, + "loss": 0.23313653469085693, + "step": 9705, + "token_acc": 0.9342051299290685 + }, + { + "epoch": 0.6026938116814599, + "grad_norm": 0.8015145063400269, + "learning_rate": 3.729630241908414e-06, + "loss": 0.23039145469665528, + "step": 9710, + "token_acc": 0.9343719610481304 + }, + { + "epoch": 0.60300415864937, + "grad_norm": 0.80972820520401, + "learning_rate": 3.7246676634927447e-06, + "loss": 0.23088579177856444, + "step": 9715, + "token_acc": 0.9336925649669524 + }, + { + "epoch": 0.6033145056172802, + "grad_norm": 0.7860719561576843, + "learning_rate": 3.7197064284494693e-06, + "loss": 0.2214982032775879, + "step": 9720, + "token_acc": 0.9356657424379785 + }, + { + "epoch": 0.6036248525851903, + "grad_norm": 0.7892095446586609, + "learning_rate": 3.7147465420045103e-06, + "loss": 0.2319951295852661, + "step": 9725, + "token_acc": 0.9335737217080086 + }, + { + "epoch": 0.6039351995531004, + "grad_norm": 0.797255277633667, + "learning_rate": 3.709788009382368e-06, + "loss": 0.22536120414733887, + "step": 9730, + "token_acc": 0.9359912484908316 + }, + { + "epoch": 0.6042455465210105, + "grad_norm": 0.8468800783157349, + "learning_rate": 3.7048308358061147e-06, + "loss": 0.22993552684783936, + "step": 9735, + "token_acc": 0.9343884074066728 + }, + { + "epoch": 0.6045558934889206, + "grad_norm": 0.8070268630981445, + "learning_rate": 3.699875026497397e-06, + "loss": 0.2313439130783081, + "step": 9740, + "token_acc": 0.9335403188732169 + }, + { + "epoch": 0.6048662404568307, + "grad_norm": 0.8293927311897278, + "learning_rate": 3.6949205866764197e-06, + "loss": 0.22958507537841796, + "step": 9745, + "token_acc": 0.9339080079855098 + }, + { + "epoch": 0.6051765874247409, + "grad_norm": 0.8579959273338318, + "learning_rate": 3.6899675215619425e-06, + "loss": 0.23144686222076416, + "step": 9750, + "token_acc": 0.9343457819804505 + }, + { + "epoch": 0.605486934392651, + "grad_norm": 0.8023242354393005, + "learning_rate": 3.6850158363712864e-06, + "loss": 0.2275395393371582, + "step": 9755, + "token_acc": 0.9349010326453439 + }, + { + "epoch": 0.6057972813605611, + "grad_norm": 0.8688347935676575, + "learning_rate": 3.68006553632031e-06, + "loss": 0.23152763843536378, + "step": 9760, + "token_acc": 0.9334072073626308 + }, + { + "epoch": 0.6061076283284712, + "grad_norm": 0.8189877867698669, + "learning_rate": 3.675116626623415e-06, + "loss": 0.231813645362854, + "step": 9765, + "token_acc": 0.9339543667668981 + }, + { + "epoch": 0.6064179752963813, + "grad_norm": 0.7765679359436035, + "learning_rate": 3.670169112493543e-06, + "loss": 0.22655730247497557, + "step": 9770, + "token_acc": 0.9353306613226453 + }, + { + "epoch": 0.6067283222642915, + "grad_norm": 0.7945225834846497, + "learning_rate": 3.6652229991421594e-06, + "loss": 0.23000826835632324, + "step": 9775, + "token_acc": 0.9339096880020975 + }, + { + "epoch": 0.6070386692322016, + "grad_norm": 0.8985360860824585, + "learning_rate": 3.6602782917792572e-06, + "loss": 0.2296682596206665, + "step": 9780, + "token_acc": 0.9343942321097811 + }, + { + "epoch": 0.6073490162001117, + "grad_norm": 0.8124051094055176, + "learning_rate": 3.65533499561335e-06, + "loss": 0.2280045509338379, + "step": 9785, + "token_acc": 0.934496060136494 + }, + { + "epoch": 0.6076593631680218, + "grad_norm": 0.7648645043373108, + "learning_rate": 3.650393115851463e-06, + "loss": 0.2237316608428955, + "step": 9790, + "token_acc": 0.9356383734381448 + }, + { + "epoch": 0.6079697101359319, + "grad_norm": 0.8415268063545227, + "learning_rate": 3.645452657699124e-06, + "loss": 0.23292920589447022, + "step": 9795, + "token_acc": 0.9334014887276323 + }, + { + "epoch": 0.6082800571038421, + "grad_norm": 0.8478737473487854, + "learning_rate": 3.6405136263603767e-06, + "loss": 0.2319162368774414, + "step": 9800, + "token_acc": 0.9337972676984602 + }, + { + "epoch": 0.6085904040717522, + "grad_norm": 0.7928475141525269, + "learning_rate": 3.635576027037749e-06, + "loss": 0.22256886959075928, + "step": 9805, + "token_acc": 0.9354377880184331 + }, + { + "epoch": 0.6089007510396623, + "grad_norm": 0.8897783756256104, + "learning_rate": 3.630639864932265e-06, + "loss": 0.22224771976470947, + "step": 9810, + "token_acc": 0.9358568187680647 + }, + { + "epoch": 0.6092110980075724, + "grad_norm": 0.8263506889343262, + "learning_rate": 3.625705145243438e-06, + "loss": 0.23047707080841065, + "step": 9815, + "token_acc": 0.933946112166726 + }, + { + "epoch": 0.6095214449754826, + "grad_norm": 0.7777978181838989, + "learning_rate": 3.6207718731692563e-06, + "loss": 0.22343292236328124, + "step": 9820, + "token_acc": 0.9355712680896054 + }, + { + "epoch": 0.6098317919433928, + "grad_norm": 0.8169053792953491, + "learning_rate": 3.61584005390619e-06, + "loss": 0.23276572227478026, + "step": 9825, + "token_acc": 0.9337097782566477 + }, + { + "epoch": 0.6101421389113029, + "grad_norm": 0.7869113087654114, + "learning_rate": 3.6109096926491716e-06, + "loss": 0.22276132106781005, + "step": 9830, + "token_acc": 0.935750217989134 + }, + { + "epoch": 0.610452485879213, + "grad_norm": 0.8109267950057983, + "learning_rate": 3.6059807945916024e-06, + "loss": 0.22616047859191896, + "step": 9835, + "token_acc": 0.9348129242751144 + }, + { + "epoch": 0.6107628328471231, + "grad_norm": 0.7741036415100098, + "learning_rate": 3.601053364925342e-06, + "loss": 0.23164095878601074, + "step": 9840, + "token_acc": 0.9337337686279504 + }, + { + "epoch": 0.6110731798150332, + "grad_norm": 0.8477919697761536, + "learning_rate": 3.5961274088407027e-06, + "loss": 0.2261268377304077, + "step": 9845, + "token_acc": 0.9346322067594434 + }, + { + "epoch": 0.6113835267829433, + "grad_norm": 0.8081256151199341, + "learning_rate": 3.5912029315264417e-06, + "loss": 0.22728803157806396, + "step": 9850, + "token_acc": 0.9348349005029521 + }, + { + "epoch": 0.6116938737508535, + "grad_norm": 0.8278705477714539, + "learning_rate": 3.586279938169766e-06, + "loss": 0.22677826881408691, + "step": 9855, + "token_acc": 0.9353805302720055 + }, + { + "epoch": 0.6120042207187636, + "grad_norm": 0.8027147650718689, + "learning_rate": 3.5813584339563112e-06, + "loss": 0.23026609420776367, + "step": 9860, + "token_acc": 0.9345834375452499 + }, + { + "epoch": 0.6123145676866737, + "grad_norm": 0.830221951007843, + "learning_rate": 3.576438424070149e-06, + "loss": 0.23023862838745118, + "step": 9865, + "token_acc": 0.9344982783575333 + }, + { + "epoch": 0.6126249146545838, + "grad_norm": 0.8117418885231018, + "learning_rate": 3.571519913693776e-06, + "loss": 0.22983570098876954, + "step": 9870, + "token_acc": 0.9335135994910132 + }, + { + "epoch": 0.6129352616224939, + "grad_norm": 0.8610823154449463, + "learning_rate": 3.5666029080081093e-06, + "loss": 0.22762517929077147, + "step": 9875, + "token_acc": 0.9349269940072178 + }, + { + "epoch": 0.6132456085904041, + "grad_norm": 0.8094273805618286, + "learning_rate": 3.5616874121924804e-06, + "loss": 0.22850580215454103, + "step": 9880, + "token_acc": 0.9348235783317572 + }, + { + "epoch": 0.6135559555583142, + "grad_norm": 0.9076403975486755, + "learning_rate": 3.5567734314246323e-06, + "loss": 0.22914843559265136, + "step": 9885, + "token_acc": 0.9341020546905423 + }, + { + "epoch": 0.6138663025262243, + "grad_norm": 0.8239562511444092, + "learning_rate": 3.5518609708807113e-06, + "loss": 0.22254562377929688, + "step": 9890, + "token_acc": 0.9361250290340777 + }, + { + "epoch": 0.6141766494941344, + "grad_norm": 0.8209520578384399, + "learning_rate": 3.546950035735257e-06, + "loss": 0.2295470952987671, + "step": 9895, + "token_acc": 0.9340859191069492 + }, + { + "epoch": 0.6144869964620445, + "grad_norm": 0.9030706286430359, + "learning_rate": 3.5420406311612143e-06, + "loss": 0.2228083610534668, + "step": 9900, + "token_acc": 0.935732544868415 + }, + { + "epoch": 0.6147973434299547, + "grad_norm": 0.8039418458938599, + "learning_rate": 3.537132762329904e-06, + "loss": 0.2316387414932251, + "step": 9905, + "token_acc": 0.9342442356959864 + }, + { + "epoch": 0.6151076903978648, + "grad_norm": 0.8976228833198547, + "learning_rate": 3.5322264344110335e-06, + "loss": 0.23020660877227783, + "step": 9910, + "token_acc": 0.9338435003271148 + }, + { + "epoch": 0.6154180373657749, + "grad_norm": 0.7930827140808105, + "learning_rate": 3.5273216525726907e-06, + "loss": 0.22727091312408448, + "step": 9915, + "token_acc": 0.9346127349670308 + }, + { + "epoch": 0.615728384333685, + "grad_norm": 0.8208958506584167, + "learning_rate": 3.522418421981329e-06, + "loss": 0.22968006134033203, + "step": 9920, + "token_acc": 0.9355267197992176 + }, + { + "epoch": 0.6160387313015951, + "grad_norm": 0.8608465194702148, + "learning_rate": 3.517516747801771e-06, + "loss": 0.22738075256347656, + "step": 9925, + "token_acc": 0.934340341113345 + }, + { + "epoch": 0.6163490782695054, + "grad_norm": 0.8243855834007263, + "learning_rate": 3.5126166351972002e-06, + "loss": 0.22518248558044435, + "step": 9930, + "token_acc": 0.9348156240008526 + }, + { + "epoch": 0.6166594252374155, + "grad_norm": 0.821115255355835, + "learning_rate": 3.507718089329155e-06, + "loss": 0.2267326831817627, + "step": 9935, + "token_acc": 0.935281124497992 + }, + { + "epoch": 0.6169697722053256, + "grad_norm": 1.0912373065948486, + "learning_rate": 3.502821115357519e-06, + "loss": 0.22631707191467285, + "step": 9940, + "token_acc": 0.9355977209345769 + }, + { + "epoch": 0.6172801191732357, + "grad_norm": 0.8054673075675964, + "learning_rate": 3.497925718440529e-06, + "loss": 0.22903409004211425, + "step": 9945, + "token_acc": 0.934349790041197 + }, + { + "epoch": 0.6175904661411458, + "grad_norm": 0.9428507089614868, + "learning_rate": 3.4930319037347503e-06, + "loss": 0.22911443710327148, + "step": 9950, + "token_acc": 0.9339124957350201 + }, + { + "epoch": 0.617900813109056, + "grad_norm": 0.9293745160102844, + "learning_rate": 3.4881396763950935e-06, + "loss": 0.22425575256347657, + "step": 9955, + "token_acc": 0.9356536941017541 + }, + { + "epoch": 0.6182111600769661, + "grad_norm": 0.8209364414215088, + "learning_rate": 3.4832490415747846e-06, + "loss": 0.2247072696685791, + "step": 9960, + "token_acc": 0.9350816052355045 + }, + { + "epoch": 0.6185215070448762, + "grad_norm": 0.8475987315177917, + "learning_rate": 3.4783600044253785e-06, + "loss": 0.23229804039001464, + "step": 9965, + "token_acc": 0.9333554135064419 + }, + { + "epoch": 0.6188318540127863, + "grad_norm": 0.9144681096076965, + "learning_rate": 3.4734725700967486e-06, + "loss": 0.22608957290649415, + "step": 9970, + "token_acc": 0.9353074590009931 + }, + { + "epoch": 0.6191422009806964, + "grad_norm": 0.7984424233436584, + "learning_rate": 3.468586743737078e-06, + "loss": 0.22614636421203613, + "step": 9975, + "token_acc": 0.935130049062437 + }, + { + "epoch": 0.6194525479486065, + "grad_norm": 0.8702171444892883, + "learning_rate": 3.463702530492854e-06, + "loss": 0.2231358051300049, + "step": 9980, + "token_acc": 0.9357655734523579 + }, + { + "epoch": 0.6197628949165167, + "grad_norm": 0.8284938931465149, + "learning_rate": 3.4588199355088693e-06, + "loss": 0.22057771682739258, + "step": 9985, + "token_acc": 0.9367721357831116 + }, + { + "epoch": 0.6200732418844268, + "grad_norm": 0.827450156211853, + "learning_rate": 3.4539389639282084e-06, + "loss": 0.23312826156616212, + "step": 9990, + "token_acc": 0.9330962614386075 + }, + { + "epoch": 0.6203835888523369, + "grad_norm": 0.9848389029502869, + "learning_rate": 3.449059620892245e-06, + "loss": 0.22477564811706544, + "step": 9995, + "token_acc": 0.935444117294395 + }, + { + "epoch": 0.620693935820247, + "grad_norm": 0.8083221912384033, + "learning_rate": 3.4441819115406418e-06, + "loss": 0.22175590991973876, + "step": 10000, + "token_acc": 0.9360662857906519 + }, + { + "epoch": 0.620693935820247, + "eval_loss": 0.24151521921157837, + "eval_runtime": 100.7645, + "eval_samples_per_second": 198.483, + "eval_steps_per_second": 0.268, + "eval_token_acc": 0.9310817076726722, + "step": 10000 + }, + { + "epoch": 0.6210042827881571, + "grad_norm": 0.8569716215133667, + "learning_rate": 3.439305841011337e-06, + "loss": 0.2241278648376465, + "step": 10005, + "token_acc": 0.9353282358794417 + }, + { + "epoch": 0.6213146297560673, + "grad_norm": 0.8412322998046875, + "learning_rate": 3.4344314144405426e-06, + "loss": 0.2291327476501465, + "step": 10010, + "token_acc": 0.9341164249470614 + }, + { + "epoch": 0.6216249767239774, + "grad_norm": 0.8014694452285767, + "learning_rate": 3.429558636962742e-06, + "loss": 0.2346412420272827, + "step": 10015, + "token_acc": 0.9334738532570511 + }, + { + "epoch": 0.6219353236918875, + "grad_norm": 0.8482812643051147, + "learning_rate": 3.4246875137106796e-06, + "loss": 0.22121474742889405, + "step": 10020, + "token_acc": 0.9366692937577644 + }, + { + "epoch": 0.6222456706597976, + "grad_norm": 0.8203760981559753, + "learning_rate": 3.419818049815353e-06, + "loss": 0.22943410873413086, + "step": 10025, + "token_acc": 0.9345756830802414 + }, + { + "epoch": 0.6225560176277077, + "grad_norm": 0.8191822171211243, + "learning_rate": 3.4149502504060228e-06, + "loss": 0.2233646869659424, + "step": 10030, + "token_acc": 0.9360977986096481 + }, + { + "epoch": 0.622866364595618, + "grad_norm": 0.7944285869598389, + "learning_rate": 3.410084120610185e-06, + "loss": 0.22890801429748536, + "step": 10035, + "token_acc": 0.9347937169439449 + }, + { + "epoch": 0.623176711563528, + "grad_norm": 0.8303209543228149, + "learning_rate": 3.4052196655535806e-06, + "loss": 0.2299797534942627, + "step": 10040, + "token_acc": 0.933871577648487 + }, + { + "epoch": 0.6234870585314382, + "grad_norm": 0.8700240850448608, + "learning_rate": 3.4003568903601912e-06, + "loss": 0.23172323703765868, + "step": 10045, + "token_acc": 0.9338403959917904 + }, + { + "epoch": 0.6237974054993483, + "grad_norm": 0.8083306550979614, + "learning_rate": 3.395495800152223e-06, + "loss": 0.22963933944702147, + "step": 10050, + "token_acc": 0.9343289877806045 + }, + { + "epoch": 0.6241077524672584, + "grad_norm": 0.8170709609985352, + "learning_rate": 3.3906364000501097e-06, + "loss": 0.22805478572845458, + "step": 10055, + "token_acc": 0.9349100084104289 + }, + { + "epoch": 0.6244180994351686, + "grad_norm": 0.8990950584411621, + "learning_rate": 3.3857786951725048e-06, + "loss": 0.23684301376342773, + "step": 10060, + "token_acc": 0.9332342617372944 + }, + { + "epoch": 0.6247284464030787, + "grad_norm": 0.8766905069351196, + "learning_rate": 3.380922690636276e-06, + "loss": 0.22638792991638185, + "step": 10065, + "token_acc": 0.934735171592938 + }, + { + "epoch": 0.6250387933709888, + "grad_norm": 0.8936095237731934, + "learning_rate": 3.3760683915564985e-06, + "loss": 0.22777624130249025, + "step": 10070, + "token_acc": 0.9355490119271136 + }, + { + "epoch": 0.6253491403388989, + "grad_norm": 0.842910647392273, + "learning_rate": 3.371215803046455e-06, + "loss": 0.23207530975341797, + "step": 10075, + "token_acc": 0.933038729093043 + }, + { + "epoch": 0.625659487306809, + "grad_norm": 0.8299627304077148, + "learning_rate": 3.3663649302176215e-06, + "loss": 0.22061100006103515, + "step": 10080, + "token_acc": 0.9365753596327492 + }, + { + "epoch": 0.6259698342747191, + "grad_norm": 0.799919843673706, + "learning_rate": 3.361515778179668e-06, + "loss": 0.2173717975616455, + "step": 10085, + "token_acc": 0.9370095902353966 + }, + { + "epoch": 0.6262801812426293, + "grad_norm": 0.8064923286437988, + "learning_rate": 3.356668352040456e-06, + "loss": 0.22874159812927247, + "step": 10090, + "token_acc": 0.9355544181285189 + }, + { + "epoch": 0.6265905282105394, + "grad_norm": 0.8340260982513428, + "learning_rate": 3.351822656906023e-06, + "loss": 0.22273797988891603, + "step": 10095, + "token_acc": 0.9360372010747573 + }, + { + "epoch": 0.6269008751784495, + "grad_norm": 0.8122076988220215, + "learning_rate": 3.3469786978805886e-06, + "loss": 0.22518315315246581, + "step": 10100, + "token_acc": 0.9351028126997656 + }, + { + "epoch": 0.6272112221463596, + "grad_norm": 0.7799873352050781, + "learning_rate": 3.342136480066539e-06, + "loss": 0.22656936645507814, + "step": 10105, + "token_acc": 0.9355313687645502 + }, + { + "epoch": 0.6275215691142697, + "grad_norm": 0.8630241751670837, + "learning_rate": 3.3372960085644294e-06, + "loss": 0.22676103115081786, + "step": 10110, + "token_acc": 0.9345722725666313 + }, + { + "epoch": 0.6278319160821799, + "grad_norm": 0.8385790586471558, + "learning_rate": 3.3324572884729755e-06, + "loss": 0.2293994426727295, + "step": 10115, + "token_acc": 0.9345415469001345 + }, + { + "epoch": 0.62814226305009, + "grad_norm": 0.8211687207221985, + "learning_rate": 3.327620324889047e-06, + "loss": 0.22718396186828613, + "step": 10120, + "token_acc": 0.9339028538024127 + }, + { + "epoch": 0.6284526100180001, + "grad_norm": 0.78291916847229, + "learning_rate": 3.3227851229076595e-06, + "loss": 0.22285895347595214, + "step": 10125, + "token_acc": 0.9363161546214968 + }, + { + "epoch": 0.6287629569859102, + "grad_norm": 0.8484714031219482, + "learning_rate": 3.3179516876219844e-06, + "loss": 0.22085123062133788, + "step": 10130, + "token_acc": 0.9366346288714191 + }, + { + "epoch": 0.6290733039538203, + "grad_norm": 0.9292843341827393, + "learning_rate": 3.3131200241233186e-06, + "loss": 0.2281170129776001, + "step": 10135, + "token_acc": 0.9347392571546717 + }, + { + "epoch": 0.6293836509217305, + "grad_norm": 0.811102569103241, + "learning_rate": 3.3082901375011e-06, + "loss": 0.23255834579467774, + "step": 10140, + "token_acc": 0.9334223741153209 + }, + { + "epoch": 0.6296939978896406, + "grad_norm": 0.7670370936393738, + "learning_rate": 3.3034620328428966e-06, + "loss": 0.22550055980682374, + "step": 10145, + "token_acc": 0.9348884799117673 + }, + { + "epoch": 0.6300043448575507, + "grad_norm": 0.8584240674972534, + "learning_rate": 3.298635715234395e-06, + "loss": 0.22823014259338378, + "step": 10150, + "token_acc": 0.9350061675868574 + }, + { + "epoch": 0.6303146918254608, + "grad_norm": 0.8090248703956604, + "learning_rate": 3.2938111897593993e-06, + "loss": 0.2302483558654785, + "step": 10155, + "token_acc": 0.9339662014004358 + }, + { + "epoch": 0.630625038793371, + "grad_norm": 0.8433321714401245, + "learning_rate": 3.288988461499831e-06, + "loss": 0.22689316272735596, + "step": 10160, + "token_acc": 0.9342746973707192 + }, + { + "epoch": 0.6309353857612812, + "grad_norm": 0.8076629638671875, + "learning_rate": 3.2841675355357127e-06, + "loss": 0.223408842086792, + "step": 10165, + "token_acc": 0.9353291981564372 + }, + { + "epoch": 0.6312457327291913, + "grad_norm": 0.8924742341041565, + "learning_rate": 3.2793484169451685e-06, + "loss": 0.2238003730773926, + "step": 10170, + "token_acc": 0.9362749079488846 + }, + { + "epoch": 0.6315560796971014, + "grad_norm": 0.8264853954315186, + "learning_rate": 3.274531110804427e-06, + "loss": 0.22086706161499023, + "step": 10175, + "token_acc": 0.9364806752672711 + }, + { + "epoch": 0.6318664266650115, + "grad_norm": 0.7966946363449097, + "learning_rate": 3.269715622187797e-06, + "loss": 0.22362031936645507, + "step": 10180, + "token_acc": 0.9364236044488259 + }, + { + "epoch": 0.6321767736329216, + "grad_norm": 0.8290771245956421, + "learning_rate": 3.2649019561676776e-06, + "loss": 0.2305845022201538, + "step": 10185, + "token_acc": 0.933853513174404 + }, + { + "epoch": 0.6324871206008317, + "grad_norm": 0.8590317368507385, + "learning_rate": 3.2600901178145505e-06, + "loss": 0.23162002563476564, + "step": 10190, + "token_acc": 0.9341116496520028 + }, + { + "epoch": 0.6327974675687419, + "grad_norm": 0.8913273215293884, + "learning_rate": 3.255280112196968e-06, + "loss": 0.22815828323364257, + "step": 10195, + "token_acc": 0.934415527603357 + }, + { + "epoch": 0.633107814536652, + "grad_norm": 0.9401612281799316, + "learning_rate": 3.2504719443815535e-06, + "loss": 0.2321523666381836, + "step": 10200, + "token_acc": 0.9340672255739246 + }, + { + "epoch": 0.6334181615045621, + "grad_norm": 0.8862796425819397, + "learning_rate": 3.245665619432996e-06, + "loss": 0.2287216901779175, + "step": 10205, + "token_acc": 0.9344161568669356 + }, + { + "epoch": 0.6337285084724722, + "grad_norm": 0.8181058168411255, + "learning_rate": 3.240861142414042e-06, + "loss": 0.22650315761566162, + "step": 10210, + "token_acc": 0.9360103070646338 + }, + { + "epoch": 0.6340388554403823, + "grad_norm": 0.8064730167388916, + "learning_rate": 3.2360585183854876e-06, + "loss": 0.22310934066772461, + "step": 10215, + "token_acc": 0.9361356155365372 + }, + { + "epoch": 0.6343492024082925, + "grad_norm": 0.8721363544464111, + "learning_rate": 3.231257752406186e-06, + "loss": 0.2263571500778198, + "step": 10220, + "token_acc": 0.9348024871982443 + }, + { + "epoch": 0.6346595493762026, + "grad_norm": 0.8537589907646179, + "learning_rate": 3.2264588495330256e-06, + "loss": 0.2331303834915161, + "step": 10225, + "token_acc": 0.9334570888719461 + }, + { + "epoch": 0.6349698963441127, + "grad_norm": 0.7984469532966614, + "learning_rate": 3.221661814820934e-06, + "loss": 0.2306314468383789, + "step": 10230, + "token_acc": 0.934148598625066 + }, + { + "epoch": 0.6352802433120228, + "grad_norm": 0.8339990377426147, + "learning_rate": 3.216866653322874e-06, + "loss": 0.22262356281280518, + "step": 10235, + "token_acc": 0.9356832794804008 + }, + { + "epoch": 0.6355905902799329, + "grad_norm": 0.8175888657569885, + "learning_rate": 3.21207337008983e-06, + "loss": 0.22648496627807618, + "step": 10240, + "token_acc": 0.9349201010360815 + }, + { + "epoch": 0.6359009372478431, + "grad_norm": 0.8125079870223999, + "learning_rate": 3.2072819701708146e-06, + "loss": 0.2340099811553955, + "step": 10245, + "token_acc": 0.9332745424511849 + }, + { + "epoch": 0.6362112842157532, + "grad_norm": 0.8168270587921143, + "learning_rate": 3.202492458612851e-06, + "loss": 0.22840256690979005, + "step": 10250, + "token_acc": 0.9350251309735925 + }, + { + "epoch": 0.6365216311836633, + "grad_norm": 0.8282175660133362, + "learning_rate": 3.1977048404609745e-06, + "loss": 0.22511415481567382, + "step": 10255, + "token_acc": 0.9357542455553565 + }, + { + "epoch": 0.6368319781515734, + "grad_norm": 0.8289535045623779, + "learning_rate": 3.1929191207582278e-06, + "loss": 0.2243757963180542, + "step": 10260, + "token_acc": 0.9345783075794983 + }, + { + "epoch": 0.6371423251194835, + "grad_norm": 0.8471537828445435, + "learning_rate": 3.188135304545654e-06, + "loss": 0.23078014850616455, + "step": 10265, + "token_acc": 0.9340602899924934 + }, + { + "epoch": 0.6374526720873938, + "grad_norm": 0.810608983039856, + "learning_rate": 3.1833533968622865e-06, + "loss": 0.22180347442626952, + "step": 10270, + "token_acc": 0.9363224214947663 + }, + { + "epoch": 0.6377630190553039, + "grad_norm": 0.8057257533073425, + "learning_rate": 3.1785734027451563e-06, + "loss": 0.2259220600128174, + "step": 10275, + "token_acc": 0.9357796956321234 + }, + { + "epoch": 0.638073366023214, + "grad_norm": 0.8600197434425354, + "learning_rate": 3.1737953272292703e-06, + "loss": 0.22908434867858887, + "step": 10280, + "token_acc": 0.9342663951337034 + }, + { + "epoch": 0.6383837129911241, + "grad_norm": 0.8674155473709106, + "learning_rate": 3.1690191753476196e-06, + "loss": 0.22394702434539795, + "step": 10285, + "token_acc": 0.9358059255052923 + }, + { + "epoch": 0.6386940599590342, + "grad_norm": 0.9374074339866638, + "learning_rate": 3.1642449521311687e-06, + "loss": 0.22676799297332764, + "step": 10290, + "token_acc": 0.9346682276317737 + }, + { + "epoch": 0.6390044069269443, + "grad_norm": 0.8257989883422852, + "learning_rate": 3.159472662608849e-06, + "loss": 0.2286839723587036, + "step": 10295, + "token_acc": 0.9345277549025516 + }, + { + "epoch": 0.6393147538948545, + "grad_norm": 0.8139248490333557, + "learning_rate": 3.154702311807555e-06, + "loss": 0.23261470794677735, + "step": 10300, + "token_acc": 0.9344765379008261 + }, + { + "epoch": 0.6396251008627646, + "grad_norm": 0.8973035216331482, + "learning_rate": 3.1499339047521414e-06, + "loss": 0.22437925338745118, + "step": 10305, + "token_acc": 0.9355174206968279 + }, + { + "epoch": 0.6399354478306747, + "grad_norm": 0.8356834650039673, + "learning_rate": 3.1451674464654144e-06, + "loss": 0.22867374420166015, + "step": 10310, + "token_acc": 0.9351452885573479 + }, + { + "epoch": 0.6402457947985848, + "grad_norm": 0.8802993297576904, + "learning_rate": 3.1404029419681214e-06, + "loss": 0.22721638679504394, + "step": 10315, + "token_acc": 0.9349880390679725 + }, + { + "epoch": 0.6405561417664949, + "grad_norm": 0.922466516494751, + "learning_rate": 3.135640396278965e-06, + "loss": 0.2272713899612427, + "step": 10320, + "token_acc": 0.9348004510786518 + }, + { + "epoch": 0.6408664887344051, + "grad_norm": 0.8046469688415527, + "learning_rate": 3.130879814414572e-06, + "loss": 0.22557168006896972, + "step": 10325, + "token_acc": 0.935180379788947 + }, + { + "epoch": 0.6411768357023152, + "grad_norm": 0.8223339319229126, + "learning_rate": 3.126121201389505e-06, + "loss": 0.22909936904907227, + "step": 10330, + "token_acc": 0.9341787480238152 + }, + { + "epoch": 0.6414871826702253, + "grad_norm": 1.00699782371521, + "learning_rate": 3.1213645622162554e-06, + "loss": 0.23025612831115722, + "step": 10335, + "token_acc": 0.9344298304001704 + }, + { + "epoch": 0.6417975296381354, + "grad_norm": 0.8490443825721741, + "learning_rate": 3.116609901905232e-06, + "loss": 0.22455496788024903, + "step": 10340, + "token_acc": 0.9351889570797968 + }, + { + "epoch": 0.6421078766060455, + "grad_norm": 0.8225256204605103, + "learning_rate": 3.1118572254647593e-06, + "loss": 0.2274071216583252, + "step": 10345, + "token_acc": 0.9347633893257643 + }, + { + "epoch": 0.6424182235739557, + "grad_norm": 0.8428844213485718, + "learning_rate": 3.1071065379010747e-06, + "loss": 0.22573962211608886, + "step": 10350, + "token_acc": 0.935110393107162 + }, + { + "epoch": 0.6427285705418658, + "grad_norm": 0.855609118938446, + "learning_rate": 3.10235784421832e-06, + "loss": 0.2268599033355713, + "step": 10355, + "token_acc": 0.9356155653801729 + }, + { + "epoch": 0.6430389175097759, + "grad_norm": 0.8724262118339539, + "learning_rate": 3.0976111494185303e-06, + "loss": 0.22540132999420165, + "step": 10360, + "token_acc": 0.9348426799802432 + }, + { + "epoch": 0.643349264477686, + "grad_norm": 0.8318566083908081, + "learning_rate": 3.092866458501648e-06, + "loss": 0.2228609085083008, + "step": 10365, + "token_acc": 0.9362985249480803 + }, + { + "epoch": 0.6436596114455961, + "grad_norm": 0.8156371712684631, + "learning_rate": 3.0881237764654907e-06, + "loss": 0.22293236255645751, + "step": 10370, + "token_acc": 0.9357729051527011 + }, + { + "epoch": 0.6439699584135063, + "grad_norm": 0.7721272706985474, + "learning_rate": 3.083383108305771e-06, + "loss": 0.2284578800201416, + "step": 10375, + "token_acc": 0.9346336111459937 + }, + { + "epoch": 0.6442803053814165, + "grad_norm": 0.8348324298858643, + "learning_rate": 3.0786444590160742e-06, + "loss": 0.2213136911392212, + "step": 10380, + "token_acc": 0.9367671636684257 + }, + { + "epoch": 0.6445906523493266, + "grad_norm": 0.8700600266456604, + "learning_rate": 3.0739078335878593e-06, + "loss": 0.22181665897369385, + "step": 10385, + "token_acc": 0.935715529831049 + }, + { + "epoch": 0.6449009993172367, + "grad_norm": 0.8409051895141602, + "learning_rate": 3.0691732370104566e-06, + "loss": 0.22394137382507323, + "step": 10390, + "token_acc": 0.9358502394195244 + }, + { + "epoch": 0.6452113462851468, + "grad_norm": 0.8862445950508118, + "learning_rate": 3.064440674271057e-06, + "loss": 0.2245255470275879, + "step": 10395, + "token_acc": 0.9353029834341413 + }, + { + "epoch": 0.645521693253057, + "grad_norm": 0.8174576759338379, + "learning_rate": 3.0597101503547076e-06, + "loss": 0.22580304145812988, + "step": 10400, + "token_acc": 0.9348060380452391 + }, + { + "epoch": 0.6458320402209671, + "grad_norm": 0.8121772408485413, + "learning_rate": 3.0549816702443137e-06, + "loss": 0.2238241195678711, + "step": 10405, + "token_acc": 0.9361198322032557 + }, + { + "epoch": 0.6461423871888772, + "grad_norm": 0.8862443566322327, + "learning_rate": 3.050255238920621e-06, + "loss": 0.22516565322875975, + "step": 10410, + "token_acc": 0.934948279850623 + }, + { + "epoch": 0.6464527341567873, + "grad_norm": 0.8115447759628296, + "learning_rate": 3.045530861362218e-06, + "loss": 0.2235461950302124, + "step": 10415, + "token_acc": 0.9362460090741052 + }, + { + "epoch": 0.6467630811246974, + "grad_norm": 0.7945905923843384, + "learning_rate": 3.040808542545536e-06, + "loss": 0.2315307378768921, + "step": 10420, + "token_acc": 0.9347663464784817 + }, + { + "epoch": 0.6470734280926075, + "grad_norm": 0.9922177195549011, + "learning_rate": 3.0360882874448316e-06, + "loss": 0.2299818515777588, + "step": 10425, + "token_acc": 0.9342823376312959 + }, + { + "epoch": 0.6473837750605177, + "grad_norm": 0.8217952847480774, + "learning_rate": 3.0313701010321883e-06, + "loss": 0.22888360023498536, + "step": 10430, + "token_acc": 0.9344283236294693 + }, + { + "epoch": 0.6476941220284278, + "grad_norm": 0.8973700404167175, + "learning_rate": 3.0266539882775135e-06, + "loss": 0.2282423734664917, + "step": 10435, + "token_acc": 0.9352412328576816 + }, + { + "epoch": 0.6480044689963379, + "grad_norm": 0.8587597012519836, + "learning_rate": 3.02193995414853e-06, + "loss": 0.22355365753173828, + "step": 10440, + "token_acc": 0.935438259500847 + }, + { + "epoch": 0.648314815964248, + "grad_norm": 0.7929176092147827, + "learning_rate": 3.0172280036107643e-06, + "loss": 0.22811241149902345, + "step": 10445, + "token_acc": 0.9354312680124864 + }, + { + "epoch": 0.6486251629321581, + "grad_norm": 1.020565390586853, + "learning_rate": 3.01251814162756e-06, + "loss": 0.2265540361404419, + "step": 10450, + "token_acc": 0.935337169526133 + }, + { + "epoch": 0.6489355099000683, + "grad_norm": 0.8346660733222961, + "learning_rate": 3.00781037316005e-06, + "loss": 0.23351497650146485, + "step": 10455, + "token_acc": 0.933485503502312 + }, + { + "epoch": 0.6492458568679784, + "grad_norm": 0.8253877758979797, + "learning_rate": 3.0031047031671657e-06, + "loss": 0.2286909580230713, + "step": 10460, + "token_acc": 0.9338650363455101 + }, + { + "epoch": 0.6495562038358885, + "grad_norm": 0.7967126369476318, + "learning_rate": 2.998401136605632e-06, + "loss": 0.2270939588546753, + "step": 10465, + "token_acc": 0.9347414258087432 + }, + { + "epoch": 0.6498665508037986, + "grad_norm": 0.8224100470542908, + "learning_rate": 2.993699678429952e-06, + "loss": 0.22571148872375488, + "step": 10470, + "token_acc": 0.9343573289492002 + }, + { + "epoch": 0.6501768977717087, + "grad_norm": 0.7544921040534973, + "learning_rate": 2.989000333592411e-06, + "loss": 0.22827577590942383, + "step": 10475, + "token_acc": 0.9356353682266416 + }, + { + "epoch": 0.6504872447396189, + "grad_norm": 0.8363432884216309, + "learning_rate": 2.9843031070430694e-06, + "loss": 0.22504262924194335, + "step": 10480, + "token_acc": 0.9360336714349965 + }, + { + "epoch": 0.650797591707529, + "grad_norm": 0.8271840810775757, + "learning_rate": 2.9796080037297546e-06, + "loss": 0.22651987075805663, + "step": 10485, + "token_acc": 0.9347417268717699 + }, + { + "epoch": 0.6511079386754391, + "grad_norm": 0.8695602416992188, + "learning_rate": 2.974915028598053e-06, + "loss": 0.228261137008667, + "step": 10490, + "token_acc": 0.9345916525206801 + }, + { + "epoch": 0.6514182856433492, + "grad_norm": 0.8622519373893738, + "learning_rate": 2.9702241865913216e-06, + "loss": 0.22208852767944337, + "step": 10495, + "token_acc": 0.9353027811350413 + }, + { + "epoch": 0.6517286326112594, + "grad_norm": 0.8766584992408752, + "learning_rate": 2.9655354826506577e-06, + "loss": 0.22573544979095458, + "step": 10500, + "token_acc": 0.9354973641228571 + }, + { + "epoch": 0.6517286326112594, + "eval_loss": 0.24099454283714294, + "eval_runtime": 130.5534, + "eval_samples_per_second": 153.194, + "eval_steps_per_second": 0.207, + "eval_token_acc": 0.9311677509715722, + "step": 10500 + }, + { + "epoch": 0.6520389795791696, + "grad_norm": 0.7874835133552551, + "learning_rate": 2.9608489217149115e-06, + "loss": 0.21494827270507813, + "step": 10505, + "token_acc": 0.9380793946792954 + }, + { + "epoch": 0.6523493265470797, + "grad_norm": 0.8734012842178345, + "learning_rate": 2.9561645087206783e-06, + "loss": 0.22580971717834472, + "step": 10510, + "token_acc": 0.9346440668889157 + }, + { + "epoch": 0.6526596735149898, + "grad_norm": 0.8220220804214478, + "learning_rate": 2.9514822486022858e-06, + "loss": 0.2237410068511963, + "step": 10515, + "token_acc": 0.9363006630288886 + }, + { + "epoch": 0.6529700204828999, + "grad_norm": 0.8783247470855713, + "learning_rate": 2.9468021462917993e-06, + "loss": 0.22192988395690919, + "step": 10520, + "token_acc": 0.9355961121224131 + }, + { + "epoch": 0.65328036745081, + "grad_norm": 0.8472056984901428, + "learning_rate": 2.9421242067190083e-06, + "loss": 0.22617430686950685, + "step": 10525, + "token_acc": 0.9355033632808091 + }, + { + "epoch": 0.6535907144187201, + "grad_norm": 0.8307264447212219, + "learning_rate": 2.9374484348114213e-06, + "loss": 0.2317513942718506, + "step": 10530, + "token_acc": 0.9339275411336371 + }, + { + "epoch": 0.6539010613866303, + "grad_norm": 0.8713358640670776, + "learning_rate": 2.9327748354942704e-06, + "loss": 0.22518627643585204, + "step": 10535, + "token_acc": 0.9358903599476005 + }, + { + "epoch": 0.6542114083545404, + "grad_norm": 0.8243184089660645, + "learning_rate": 2.9281034136904944e-06, + "loss": 0.22897276878356934, + "step": 10540, + "token_acc": 0.9346656366506566 + }, + { + "epoch": 0.6545217553224505, + "grad_norm": 0.80588698387146, + "learning_rate": 2.923434174320736e-06, + "loss": 0.22482380867004395, + "step": 10545, + "token_acc": 0.9357269912378945 + }, + { + "epoch": 0.6548321022903606, + "grad_norm": 0.8233228325843811, + "learning_rate": 2.9187671223033478e-06, + "loss": 0.22146410942077638, + "step": 10550, + "token_acc": 0.936073425918493 + }, + { + "epoch": 0.6551424492582707, + "grad_norm": 0.8113839030265808, + "learning_rate": 2.91410226255437e-06, + "loss": 0.2284111738204956, + "step": 10555, + "token_acc": 0.9343037109310183 + }, + { + "epoch": 0.6554527962261809, + "grad_norm": 0.8552408218383789, + "learning_rate": 2.9094395999875365e-06, + "loss": 0.22697019577026367, + "step": 10560, + "token_acc": 0.9353821959084422 + }, + { + "epoch": 0.655763143194091, + "grad_norm": 0.8257595896720886, + "learning_rate": 2.9047791395142684e-06, + "loss": 0.2344212532043457, + "step": 10565, + "token_acc": 0.9335447410146253 + }, + { + "epoch": 0.6560734901620011, + "grad_norm": 1.1332318782806396, + "learning_rate": 2.9001208860436626e-06, + "loss": 0.22331857681274414, + "step": 10570, + "token_acc": 0.9360509138381201 + }, + { + "epoch": 0.6563838371299112, + "grad_norm": 0.8137020468711853, + "learning_rate": 2.8954648444824967e-06, + "loss": 0.2197270154953003, + "step": 10575, + "token_acc": 0.9364202303508585 + }, + { + "epoch": 0.6566941840978213, + "grad_norm": 0.869922399520874, + "learning_rate": 2.890811019735218e-06, + "loss": 0.23374598026275634, + "step": 10580, + "token_acc": 0.9340377418373689 + }, + { + "epoch": 0.6570045310657315, + "grad_norm": 0.8131517767906189, + "learning_rate": 2.8861594167039316e-06, + "loss": 0.2221766471862793, + "step": 10585, + "token_acc": 0.936230305231092 + }, + { + "epoch": 0.6573148780336416, + "grad_norm": 0.797662615776062, + "learning_rate": 2.88151004028841e-06, + "loss": 0.22331562042236328, + "step": 10590, + "token_acc": 0.9359078643271229 + }, + { + "epoch": 0.6576252250015517, + "grad_norm": 0.8238673210144043, + "learning_rate": 2.876862895386081e-06, + "loss": 0.22514495849609376, + "step": 10595, + "token_acc": 0.9352426422584623 + }, + { + "epoch": 0.6579355719694618, + "grad_norm": 1.637500286102295, + "learning_rate": 2.872217986892015e-06, + "loss": 0.23692765235900878, + "step": 10600, + "token_acc": 0.9326439204320907 + }, + { + "epoch": 0.6582459189373719, + "grad_norm": 0.852145254611969, + "learning_rate": 2.8675753196989273e-06, + "loss": 0.23273057937622071, + "step": 10605, + "token_acc": 0.9335623597712399 + }, + { + "epoch": 0.6585562659052822, + "grad_norm": 0.823457658290863, + "learning_rate": 2.862934898697183e-06, + "loss": 0.22384085655212402, + "step": 10610, + "token_acc": 0.9355497326487723 + }, + { + "epoch": 0.6588666128731923, + "grad_norm": 0.8552753925323486, + "learning_rate": 2.8582967287747686e-06, + "loss": 0.22592558860778808, + "step": 10615, + "token_acc": 0.9354333437756197 + }, + { + "epoch": 0.6591769598411024, + "grad_norm": 1.056519865989685, + "learning_rate": 2.853660814817304e-06, + "loss": 0.23117125034332275, + "step": 10620, + "token_acc": 0.9341498308620021 + }, + { + "epoch": 0.6594873068090125, + "grad_norm": 0.8177993297576904, + "learning_rate": 2.849027161708034e-06, + "loss": 0.2261200428009033, + "step": 10625, + "token_acc": 0.9350010975050053 + }, + { + "epoch": 0.6597976537769226, + "grad_norm": 0.8769198656082153, + "learning_rate": 2.844395774327823e-06, + "loss": 0.22885556221008302, + "step": 10630, + "token_acc": 0.935761846186758 + }, + { + "epoch": 0.6601080007448327, + "grad_norm": 0.896242618560791, + "learning_rate": 2.839766657555143e-06, + "loss": 0.22592697143554688, + "step": 10635, + "token_acc": 0.9347374348111702 + }, + { + "epoch": 0.6604183477127429, + "grad_norm": 0.8254024386405945, + "learning_rate": 2.8351398162660803e-06, + "loss": 0.2278963088989258, + "step": 10640, + "token_acc": 0.934935192705135 + }, + { + "epoch": 0.660728694680653, + "grad_norm": 0.8242830634117126, + "learning_rate": 2.8305152553343224e-06, + "loss": 0.22493743896484375, + "step": 10645, + "token_acc": 0.9359105941384422 + }, + { + "epoch": 0.6610390416485631, + "grad_norm": 0.8347147703170776, + "learning_rate": 2.8258929796311556e-06, + "loss": 0.22957568168640136, + "step": 10650, + "token_acc": 0.9336707125818737 + }, + { + "epoch": 0.6613493886164732, + "grad_norm": 0.835120677947998, + "learning_rate": 2.8212729940254564e-06, + "loss": 0.22241241931915284, + "step": 10655, + "token_acc": 0.9351948198272834 + }, + { + "epoch": 0.6616597355843833, + "grad_norm": 0.8618392944335938, + "learning_rate": 2.8166553033836863e-06, + "loss": 0.2326285123825073, + "step": 10660, + "token_acc": 0.9329227916832208 + }, + { + "epoch": 0.6619700825522935, + "grad_norm": 0.814420223236084, + "learning_rate": 2.8120399125699014e-06, + "loss": 0.22578854560852052, + "step": 10665, + "token_acc": 0.9351914978945257 + }, + { + "epoch": 0.6622804295202036, + "grad_norm": 0.8913599252700806, + "learning_rate": 2.8074268264457245e-06, + "loss": 0.2237851858139038, + "step": 10670, + "token_acc": 0.9355161038641061 + }, + { + "epoch": 0.6625907764881137, + "grad_norm": 0.7796737551689148, + "learning_rate": 2.8028160498703507e-06, + "loss": 0.2212360382080078, + "step": 10675, + "token_acc": 0.9365671393271747 + }, + { + "epoch": 0.6629011234560238, + "grad_norm": 0.8918337821960449, + "learning_rate": 2.7982075877005478e-06, + "loss": 0.22675018310546874, + "step": 10680, + "token_acc": 0.9350338753838877 + }, + { + "epoch": 0.6632114704239339, + "grad_norm": 0.823746919631958, + "learning_rate": 2.7936014447906455e-06, + "loss": 0.222282075881958, + "step": 10685, + "token_acc": 0.9357871574314863 + }, + { + "epoch": 0.6635218173918441, + "grad_norm": 0.8164610266685486, + "learning_rate": 2.7889976259925245e-06, + "loss": 0.23326773643493653, + "step": 10690, + "token_acc": 0.9341337589319417 + }, + { + "epoch": 0.6638321643597542, + "grad_norm": 0.8235698938369751, + "learning_rate": 2.7843961361556228e-06, + "loss": 0.22773661613464355, + "step": 10695, + "token_acc": 0.9350233817376323 + }, + { + "epoch": 0.6641425113276643, + "grad_norm": 0.8071151375770569, + "learning_rate": 2.7797969801269255e-06, + "loss": 0.223075795173645, + "step": 10700, + "token_acc": 0.9362693343187457 + }, + { + "epoch": 0.6644528582955744, + "grad_norm": 0.8785064220428467, + "learning_rate": 2.775200162750954e-06, + "loss": 0.23036603927612304, + "step": 10705, + "token_acc": 0.9345693492922453 + }, + { + "epoch": 0.6647632052634845, + "grad_norm": 0.8068884015083313, + "learning_rate": 2.770605688869773e-06, + "loss": 0.22128844261169434, + "step": 10710, + "token_acc": 0.93644485274538 + }, + { + "epoch": 0.6650735522313947, + "grad_norm": 0.8756119012832642, + "learning_rate": 2.7660135633229724e-06, + "loss": 0.22472171783447265, + "step": 10715, + "token_acc": 0.9349263631482727 + }, + { + "epoch": 0.6653838991993049, + "grad_norm": 0.8191453814506531, + "learning_rate": 2.7614237909476732e-06, + "loss": 0.2315457820892334, + "step": 10720, + "token_acc": 0.934199174730985 + }, + { + "epoch": 0.665694246167215, + "grad_norm": 0.8131508231163025, + "learning_rate": 2.756836376578518e-06, + "loss": 0.2381413221359253, + "step": 10725, + "token_acc": 0.9328539727177615 + }, + { + "epoch": 0.6660045931351251, + "grad_norm": 0.7743956446647644, + "learning_rate": 2.752251325047659e-06, + "loss": 0.22358121871948242, + "step": 10730, + "token_acc": 0.9354722157171055 + }, + { + "epoch": 0.6663149401030352, + "grad_norm": 0.813213050365448, + "learning_rate": 2.7476686411847663e-06, + "loss": 0.22954211235046387, + "step": 10735, + "token_acc": 0.9339673662133463 + }, + { + "epoch": 0.6666252870709454, + "grad_norm": 0.7821421027183533, + "learning_rate": 2.743088329817016e-06, + "loss": 0.23073077201843262, + "step": 10740, + "token_acc": 0.9334408580739797 + }, + { + "epoch": 0.6669356340388555, + "grad_norm": 0.8421801924705505, + "learning_rate": 2.7385103957690816e-06, + "loss": 0.2325984477996826, + "step": 10745, + "token_acc": 0.9340880469619334 + }, + { + "epoch": 0.6672459810067656, + "grad_norm": 0.8683729767799377, + "learning_rate": 2.7339348438631296e-06, + "loss": 0.22516231536865233, + "step": 10750, + "token_acc": 0.9357560927218015 + }, + { + "epoch": 0.6675563279746757, + "grad_norm": 0.8141571879386902, + "learning_rate": 2.729361678918829e-06, + "loss": 0.22411634922027587, + "step": 10755, + "token_acc": 0.9361891771442462 + }, + { + "epoch": 0.6678666749425858, + "grad_norm": 0.824243426322937, + "learning_rate": 2.724790905753323e-06, + "loss": 0.23009810447692872, + "step": 10760, + "token_acc": 0.9337225816632476 + }, + { + "epoch": 0.6681770219104959, + "grad_norm": 0.855955958366394, + "learning_rate": 2.7202225291812368e-06, + "loss": 0.22176787853240967, + "step": 10765, + "token_acc": 0.9351436665852193 + }, + { + "epoch": 0.6684873688784061, + "grad_norm": 0.8995785713195801, + "learning_rate": 2.7156565540146772e-06, + "loss": 0.22814369201660156, + "step": 10770, + "token_acc": 0.934753404979783 + }, + { + "epoch": 0.6687977158463162, + "grad_norm": 0.8211752772331238, + "learning_rate": 2.7110929850632183e-06, + "loss": 0.2298201084136963, + "step": 10775, + "token_acc": 0.9344643599265346 + }, + { + "epoch": 0.6691080628142263, + "grad_norm": 0.7960493564605713, + "learning_rate": 2.706531827133897e-06, + "loss": 0.22476115226745605, + "step": 10780, + "token_acc": 0.9354087828625774 + }, + { + "epoch": 0.6694184097821364, + "grad_norm": 0.9066677689552307, + "learning_rate": 2.701973085031214e-06, + "loss": 0.22685859203338624, + "step": 10785, + "token_acc": 0.9345591235310249 + }, + { + "epoch": 0.6697287567500465, + "grad_norm": 0.8380141258239746, + "learning_rate": 2.6974167635571234e-06, + "loss": 0.2289644956588745, + "step": 10790, + "token_acc": 0.9345842898492863 + }, + { + "epoch": 0.6700391037179567, + "grad_norm": 0.8835338354110718, + "learning_rate": 2.692862867511034e-06, + "loss": 0.22176756858825683, + "step": 10795, + "token_acc": 0.9357149893285175 + }, + { + "epoch": 0.6703494506858668, + "grad_norm": 0.8354815244674683, + "learning_rate": 2.6883114016897926e-06, + "loss": 0.22756783962249755, + "step": 10800, + "token_acc": 0.9343431670855933 + }, + { + "epoch": 0.6706597976537769, + "grad_norm": 0.7682988047599792, + "learning_rate": 2.6837623708876864e-06, + "loss": 0.22572104930877684, + "step": 10805, + "token_acc": 0.9355596686750942 + }, + { + "epoch": 0.670970144621687, + "grad_norm": 0.9365578293800354, + "learning_rate": 2.6792157798964487e-06, + "loss": 0.2249309539794922, + "step": 10810, + "token_acc": 0.9352182776056972 + }, + { + "epoch": 0.6712804915895971, + "grad_norm": 0.8292002081871033, + "learning_rate": 2.6746716335052305e-06, + "loss": 0.2223968744277954, + "step": 10815, + "token_acc": 0.9368353398753007 + }, + { + "epoch": 0.6715908385575073, + "grad_norm": 0.8226572871208191, + "learning_rate": 2.6701299365006107e-06, + "loss": 0.22937119007110596, + "step": 10820, + "token_acc": 0.9345181741611549 + }, + { + "epoch": 0.6719011855254174, + "grad_norm": 0.8253859281539917, + "learning_rate": 2.665590693666591e-06, + "loss": 0.23243091106414795, + "step": 10825, + "token_acc": 0.9338138493944181 + }, + { + "epoch": 0.6722115324933275, + "grad_norm": 0.8651408553123474, + "learning_rate": 2.6610539097845895e-06, + "loss": 0.23156533241271973, + "step": 10830, + "token_acc": 0.9337935964732281 + }, + { + "epoch": 0.6725218794612376, + "grad_norm": 0.8250735998153687, + "learning_rate": 2.6565195896334274e-06, + "loss": 0.22808165550231935, + "step": 10835, + "token_acc": 0.9353396684322944 + }, + { + "epoch": 0.6728322264291478, + "grad_norm": 0.7830631136894226, + "learning_rate": 2.6519877379893354e-06, + "loss": 0.22204866409301757, + "step": 10840, + "token_acc": 0.936228763666947 + }, + { + "epoch": 0.673142573397058, + "grad_norm": 0.8327221870422363, + "learning_rate": 2.6474583596259475e-06, + "loss": 0.2303910255432129, + "step": 10845, + "token_acc": 0.9345614737685223 + }, + { + "epoch": 0.6734529203649681, + "grad_norm": 0.8320080637931824, + "learning_rate": 2.642931459314282e-06, + "loss": 0.2279609203338623, + "step": 10850, + "token_acc": 0.9345208412289975 + }, + { + "epoch": 0.6737632673328782, + "grad_norm": 0.807061493396759, + "learning_rate": 2.6384070418227583e-06, + "loss": 0.22901458740234376, + "step": 10855, + "token_acc": 0.9344186292950433 + }, + { + "epoch": 0.6740736143007883, + "grad_norm": 0.8437809944152832, + "learning_rate": 2.6338851119171704e-06, + "loss": 0.22722196578979492, + "step": 10860, + "token_acc": 0.9348425841187984 + }, + { + "epoch": 0.6743839612686984, + "grad_norm": 1.0447630882263184, + "learning_rate": 2.6293656743607005e-06, + "loss": 0.22874414920806885, + "step": 10865, + "token_acc": 0.9341627172195892 + }, + { + "epoch": 0.6746943082366085, + "grad_norm": 0.8448339700698853, + "learning_rate": 2.6248487339139027e-06, + "loss": 0.22220687866210936, + "step": 10870, + "token_acc": 0.9356501650165017 + }, + { + "epoch": 0.6750046552045187, + "grad_norm": 0.8740484714508057, + "learning_rate": 2.620334295334697e-06, + "loss": 0.22617557048797607, + "step": 10875, + "token_acc": 0.9341452235971976 + }, + { + "epoch": 0.6753150021724288, + "grad_norm": 0.8319733142852783, + "learning_rate": 2.615822363378373e-06, + "loss": 0.23303942680358886, + "step": 10880, + "token_acc": 0.9334174255882606 + }, + { + "epoch": 0.6756253491403389, + "grad_norm": 0.7985439300537109, + "learning_rate": 2.6113129427975794e-06, + "loss": 0.23701820373535157, + "step": 10885, + "token_acc": 0.9331185703680136 + }, + { + "epoch": 0.675935696108249, + "grad_norm": 0.8392468094825745, + "learning_rate": 2.6068060383423155e-06, + "loss": 0.2257922649383545, + "step": 10890, + "token_acc": 0.9344499695823525 + }, + { + "epoch": 0.6762460430761591, + "grad_norm": 0.8244118094444275, + "learning_rate": 2.602301654759933e-06, + "loss": 0.2271113872528076, + "step": 10895, + "token_acc": 0.9348409506153836 + }, + { + "epoch": 0.6765563900440693, + "grad_norm": 0.8121625185012817, + "learning_rate": 2.5977997967951285e-06, + "loss": 0.23000075817108154, + "step": 10900, + "token_acc": 0.935035979867295 + }, + { + "epoch": 0.6768667370119794, + "grad_norm": 0.8263477087020874, + "learning_rate": 2.5933004691899405e-06, + "loss": 0.22955193519592285, + "step": 10905, + "token_acc": 0.93535372427475 + }, + { + "epoch": 0.6771770839798895, + "grad_norm": 0.8374332189559937, + "learning_rate": 2.5888036766837355e-06, + "loss": 0.2324249267578125, + "step": 10910, + "token_acc": 0.9336591486972889 + }, + { + "epoch": 0.6774874309477996, + "grad_norm": 0.8180155754089355, + "learning_rate": 2.5843094240132157e-06, + "loss": 0.23461902141571045, + "step": 10915, + "token_acc": 0.9327203736135435 + }, + { + "epoch": 0.6777977779157097, + "grad_norm": 0.7861869931221008, + "learning_rate": 2.5798177159124076e-06, + "loss": 0.2205819845199585, + "step": 10920, + "token_acc": 0.9365971000586385 + }, + { + "epoch": 0.6781081248836199, + "grad_norm": 1.0600954294204712, + "learning_rate": 2.5753285571126523e-06, + "loss": 0.224623703956604, + "step": 10925, + "token_acc": 0.9361700721630409 + }, + { + "epoch": 0.67841847185153, + "grad_norm": 0.8026278018951416, + "learning_rate": 2.570841952342611e-06, + "loss": 0.22702107429504395, + "step": 10930, + "token_acc": 0.9339663756784955 + }, + { + "epoch": 0.6787288188194401, + "grad_norm": 0.812997579574585, + "learning_rate": 2.5663579063282516e-06, + "loss": 0.22706398963928223, + "step": 10935, + "token_acc": 0.9353071422359874 + }, + { + "epoch": 0.6790391657873502, + "grad_norm": 0.8386937379837036, + "learning_rate": 2.5618764237928505e-06, + "loss": 0.23208417892456054, + "step": 10940, + "token_acc": 0.9341965283669319 + }, + { + "epoch": 0.6793495127552603, + "grad_norm": 0.7910364866256714, + "learning_rate": 2.55739750945698e-06, + "loss": 0.2240131139755249, + "step": 10945, + "token_acc": 0.9353158602150538 + }, + { + "epoch": 0.6796598597231706, + "grad_norm": 0.8091467618942261, + "learning_rate": 2.5529211680385047e-06, + "loss": 0.23150575160980225, + "step": 10950, + "token_acc": 0.9336963874983087 + }, + { + "epoch": 0.6799702066910807, + "grad_norm": 0.7687386870384216, + "learning_rate": 2.5484474042525857e-06, + "loss": 0.2237762689590454, + "step": 10955, + "token_acc": 0.9357699453332798 + }, + { + "epoch": 0.6802805536589908, + "grad_norm": 0.8341094851493835, + "learning_rate": 2.5439762228116673e-06, + "loss": 0.22780933380126953, + "step": 10960, + "token_acc": 0.9343471932239559 + }, + { + "epoch": 0.6805909006269009, + "grad_norm": 0.8193002343177795, + "learning_rate": 2.539507628425469e-06, + "loss": 0.223095703125, + "step": 10965, + "token_acc": 0.9347401821512971 + }, + { + "epoch": 0.680901247594811, + "grad_norm": 0.9120641946792603, + "learning_rate": 2.535041625800989e-06, + "loss": 0.2225553035736084, + "step": 10970, + "token_acc": 0.9369110472498831 + }, + { + "epoch": 0.6812115945627211, + "grad_norm": 0.9374921917915344, + "learning_rate": 2.530578219642499e-06, + "loss": 0.22342004776000976, + "step": 10975, + "token_acc": 0.9365342458429129 + }, + { + "epoch": 0.6815219415306313, + "grad_norm": 0.845941424369812, + "learning_rate": 2.5261174146515256e-06, + "loss": 0.22265996932983398, + "step": 10980, + "token_acc": 0.9359684373979513 + }, + { + "epoch": 0.6818322884985414, + "grad_norm": 0.7831390500068665, + "learning_rate": 2.5216592155268667e-06, + "loss": 0.2203070640563965, + "step": 10985, + "token_acc": 0.9370164046251279 + }, + { + "epoch": 0.6821426354664515, + "grad_norm": 0.8169050812721252, + "learning_rate": 2.517203626964566e-06, + "loss": 0.22159876823425292, + "step": 10990, + "token_acc": 0.9358423776924362 + }, + { + "epoch": 0.6824529824343616, + "grad_norm": 0.8511732220649719, + "learning_rate": 2.5127506536579226e-06, + "loss": 0.22645368576049804, + "step": 10995, + "token_acc": 0.93477032266544 + }, + { + "epoch": 0.6827633294022717, + "grad_norm": 0.8053267002105713, + "learning_rate": 2.508300300297484e-06, + "loss": 0.22338442802429198, + "step": 11000, + "token_acc": 0.9356713277724785 + }, + { + "epoch": 0.6827633294022717, + "eval_loss": 0.23978617787361145, + "eval_runtime": 110.4575, + "eval_samples_per_second": 181.065, + "eval_steps_per_second": 0.244, + "eval_token_acc": 0.9315406052668059, + "step": 11000 + }, + { + "epoch": 0.6830736763701819, + "grad_norm": 0.8072296380996704, + "learning_rate": 2.5038525715710267e-06, + "loss": 0.22555809020996093, + "step": 11005, + "token_acc": 0.9342297554302492 + }, + { + "epoch": 0.683384023338092, + "grad_norm": 0.9502686858177185, + "learning_rate": 2.4994074721635732e-06, + "loss": 0.22415103912353515, + "step": 11010, + "token_acc": 0.9351587131940362 + }, + { + "epoch": 0.6836943703060021, + "grad_norm": 0.8472788333892822, + "learning_rate": 2.4949650067573742e-06, + "loss": 0.22397608757019044, + "step": 11015, + "token_acc": 0.935464819736383 + }, + { + "epoch": 0.6840047172739122, + "grad_norm": 0.8352960348129272, + "learning_rate": 2.490525180031903e-06, + "loss": 0.22692999839782715, + "step": 11020, + "token_acc": 0.9343788043407119 + }, + { + "epoch": 0.6843150642418223, + "grad_norm": 0.8248074650764465, + "learning_rate": 2.4860879966638505e-06, + "loss": 0.22265405654907228, + "step": 11025, + "token_acc": 0.9357585855780801 + }, + { + "epoch": 0.6846254112097325, + "grad_norm": 1.0422954559326172, + "learning_rate": 2.4816534613271355e-06, + "loss": 0.22647995948791505, + "step": 11030, + "token_acc": 0.9344746825004123 + }, + { + "epoch": 0.6849357581776426, + "grad_norm": 0.9259189367294312, + "learning_rate": 2.4772215786928754e-06, + "loss": 0.22671794891357422, + "step": 11035, + "token_acc": 0.9353899059370892 + }, + { + "epoch": 0.6852461051455527, + "grad_norm": 0.802070677280426, + "learning_rate": 2.4727923534293957e-06, + "loss": 0.2257382392883301, + "step": 11040, + "token_acc": 0.9354255755762382 + }, + { + "epoch": 0.6855564521134628, + "grad_norm": 0.8134832978248596, + "learning_rate": 2.4683657902022267e-06, + "loss": 0.2232145071029663, + "step": 11045, + "token_acc": 0.936275880434059 + }, + { + "epoch": 0.6858667990813729, + "grad_norm": 0.8214874863624573, + "learning_rate": 2.463941893674094e-06, + "loss": 0.22294673919677735, + "step": 11050, + "token_acc": 0.9354732140497397 + }, + { + "epoch": 0.6861771460492831, + "grad_norm": 0.7979225516319275, + "learning_rate": 2.4595206685049095e-06, + "loss": 0.2289970874786377, + "step": 11055, + "token_acc": 0.9338138622234261 + }, + { + "epoch": 0.6864874930171933, + "grad_norm": 0.8345953822135925, + "learning_rate": 2.455102119351776e-06, + "loss": 0.22617015838623047, + "step": 11060, + "token_acc": 0.9354372519143872 + }, + { + "epoch": 0.6867978399851034, + "grad_norm": 0.944805383682251, + "learning_rate": 2.450686250868976e-06, + "loss": 0.22216348648071288, + "step": 11065, + "token_acc": 0.9358885063211291 + }, + { + "epoch": 0.6871081869530135, + "grad_norm": 1.2585172653198242, + "learning_rate": 2.44627306770797e-06, + "loss": 0.22482447624206542, + "step": 11070, + "token_acc": 0.9353135140105884 + }, + { + "epoch": 0.6874185339209236, + "grad_norm": 0.8357342481613159, + "learning_rate": 2.441862574517387e-06, + "loss": 0.22273945808410645, + "step": 11075, + "token_acc": 0.9356059081579265 + }, + { + "epoch": 0.6877288808888338, + "grad_norm": 0.8633572459220886, + "learning_rate": 2.437454775943019e-06, + "loss": 0.2263735294342041, + "step": 11080, + "token_acc": 0.9359503718773972 + }, + { + "epoch": 0.6880392278567439, + "grad_norm": 0.9200255274772644, + "learning_rate": 2.4330496766278327e-06, + "loss": 0.22185001373291016, + "step": 11085, + "token_acc": 0.9368388156363661 + }, + { + "epoch": 0.688349574824654, + "grad_norm": 0.8289831280708313, + "learning_rate": 2.4286472812119393e-06, + "loss": 0.22053358554840088, + "step": 11090, + "token_acc": 0.9365836707964129 + }, + { + "epoch": 0.6886599217925641, + "grad_norm": 0.9026960730552673, + "learning_rate": 2.4242475943326037e-06, + "loss": 0.22258412837982178, + "step": 11095, + "token_acc": 0.9361208671833867 + }, + { + "epoch": 0.6889702687604742, + "grad_norm": 0.8401134610176086, + "learning_rate": 2.4198506206242416e-06, + "loss": 0.22501320838928224, + "step": 11100, + "token_acc": 0.9359452116670984 + }, + { + "epoch": 0.6892806157283843, + "grad_norm": 0.8398122191429138, + "learning_rate": 2.415456364718411e-06, + "loss": 0.22009835243225098, + "step": 11105, + "token_acc": 0.9367277799920287 + }, + { + "epoch": 0.6895909626962945, + "grad_norm": 0.877639651298523, + "learning_rate": 2.4110648312438007e-06, + "loss": 0.21858615875244142, + "step": 11110, + "token_acc": 0.9358599041797456 + }, + { + "epoch": 0.6899013096642046, + "grad_norm": 0.8251043558120728, + "learning_rate": 2.4066760248262387e-06, + "loss": 0.22800912857055664, + "step": 11115, + "token_acc": 0.9342346242216376 + }, + { + "epoch": 0.6902116566321147, + "grad_norm": 0.9936253428459167, + "learning_rate": 2.4022899500886806e-06, + "loss": 0.22876813411712646, + "step": 11120, + "token_acc": 0.9339477826488533 + }, + { + "epoch": 0.6905220036000248, + "grad_norm": 0.8104826211929321, + "learning_rate": 2.3979066116511967e-06, + "loss": 0.22265408039093018, + "step": 11125, + "token_acc": 0.9365209235497196 + }, + { + "epoch": 0.6908323505679349, + "grad_norm": 0.8124892711639404, + "learning_rate": 2.3935260141309845e-06, + "loss": 0.22827134132385254, + "step": 11130, + "token_acc": 0.9345431412595842 + }, + { + "epoch": 0.6911426975358451, + "grad_norm": 0.8026403784751892, + "learning_rate": 2.389148162142347e-06, + "loss": 0.2264378547668457, + "step": 11135, + "token_acc": 0.9358293402997607 + }, + { + "epoch": 0.6914530445037552, + "grad_norm": 0.8319078087806702, + "learning_rate": 2.384773060296699e-06, + "loss": 0.22564311027526857, + "step": 11140, + "token_acc": 0.9353361299915038 + }, + { + "epoch": 0.6917633914716653, + "grad_norm": 0.8620616793632507, + "learning_rate": 2.3804007132025604e-06, + "loss": 0.22575764656066893, + "step": 11145, + "token_acc": 0.9354069041182808 + }, + { + "epoch": 0.6920737384395754, + "grad_norm": 0.9183108806610107, + "learning_rate": 2.3760311254655428e-06, + "loss": 0.2234050989151001, + "step": 11150, + "token_acc": 0.9361659948457014 + }, + { + "epoch": 0.6923840854074855, + "grad_norm": 0.8361497521400452, + "learning_rate": 2.371664301688357e-06, + "loss": 0.23135125637054443, + "step": 11155, + "token_acc": 0.9344159280328086 + }, + { + "epoch": 0.6926944323753957, + "grad_norm": 0.8143050670623779, + "learning_rate": 2.367300246470802e-06, + "loss": 0.22027235031127929, + "step": 11160, + "token_acc": 0.9364755971737909 + }, + { + "epoch": 0.6930047793433058, + "grad_norm": 0.7851625680923462, + "learning_rate": 2.3629389644097583e-06, + "loss": 0.2272705316543579, + "step": 11165, + "token_acc": 0.9345517738088158 + }, + { + "epoch": 0.693315126311216, + "grad_norm": 0.9899514317512512, + "learning_rate": 2.3585804600991812e-06, + "loss": 0.22692344188690186, + "step": 11170, + "token_acc": 0.9345676302961382 + }, + { + "epoch": 0.693625473279126, + "grad_norm": 0.7790020108222961, + "learning_rate": 2.354224738130113e-06, + "loss": 0.22095718383789062, + "step": 11175, + "token_acc": 0.9359111761707467 + }, + { + "epoch": 0.6939358202470362, + "grad_norm": 0.8658116459846497, + "learning_rate": 2.349871803090653e-06, + "loss": 0.2204197645187378, + "step": 11180, + "token_acc": 0.9368062388156287 + }, + { + "epoch": 0.6942461672149464, + "grad_norm": 0.8718162178993225, + "learning_rate": 2.3455216595659677e-06, + "loss": 0.22421345710754395, + "step": 11185, + "token_acc": 0.9360867955465052 + }, + { + "epoch": 0.6945565141828565, + "grad_norm": 0.9478991627693176, + "learning_rate": 2.341174312138285e-06, + "loss": 0.2269810676574707, + "step": 11190, + "token_acc": 0.9360761465396674 + }, + { + "epoch": 0.6948668611507666, + "grad_norm": 0.8574694991111755, + "learning_rate": 2.33682976538689e-06, + "loss": 0.22741847038269042, + "step": 11195, + "token_acc": 0.9343295373141323 + }, + { + "epoch": 0.6951772081186767, + "grad_norm": 0.9107205271720886, + "learning_rate": 2.332488023888111e-06, + "loss": 0.2225886821746826, + "step": 11200, + "token_acc": 0.9359455654622243 + }, + { + "epoch": 0.6954875550865868, + "grad_norm": 0.8341516256332397, + "learning_rate": 2.3281490922153268e-06, + "loss": 0.22385425567626954, + "step": 11205, + "token_acc": 0.935889902375833 + }, + { + "epoch": 0.6957979020544969, + "grad_norm": 0.770504891872406, + "learning_rate": 2.3238129749389537e-06, + "loss": 0.229888391494751, + "step": 11210, + "token_acc": 0.9344594053057627 + }, + { + "epoch": 0.6961082490224071, + "grad_norm": 0.8124657273292542, + "learning_rate": 2.319479676626448e-06, + "loss": 0.22212445735931396, + "step": 11215, + "token_acc": 0.9361095491681275 + }, + { + "epoch": 0.6964185959903172, + "grad_norm": 0.8715262413024902, + "learning_rate": 2.3151492018422906e-06, + "loss": 0.22250618934631347, + "step": 11220, + "token_acc": 0.9364688742764894 + }, + { + "epoch": 0.6967289429582273, + "grad_norm": 0.765491247177124, + "learning_rate": 2.3108215551479867e-06, + "loss": 0.22539472579956055, + "step": 11225, + "token_acc": 0.9360628898301153 + }, + { + "epoch": 0.6970392899261374, + "grad_norm": 0.817959725856781, + "learning_rate": 2.3064967411020745e-06, + "loss": 0.22164249420166016, + "step": 11230, + "token_acc": 0.9364415275420047 + }, + { + "epoch": 0.6973496368940475, + "grad_norm": 0.7923972606658936, + "learning_rate": 2.302174764260097e-06, + "loss": 0.22539467811584474, + "step": 11235, + "token_acc": 0.9348247946736012 + }, + { + "epoch": 0.6976599838619577, + "grad_norm": 0.8162638545036316, + "learning_rate": 2.297855629174609e-06, + "loss": 0.22540273666381835, + "step": 11240, + "token_acc": 0.9357233940810796 + }, + { + "epoch": 0.6979703308298678, + "grad_norm": 0.9158134460449219, + "learning_rate": 2.2935393403951783e-06, + "loss": 0.231064510345459, + "step": 11245, + "token_acc": 0.9338275654548942 + }, + { + "epoch": 0.6982806777977779, + "grad_norm": 0.8750988841056824, + "learning_rate": 2.2892259024683726e-06, + "loss": 0.22537508010864257, + "step": 11250, + "token_acc": 0.9353207416944294 + }, + { + "epoch": 0.698591024765688, + "grad_norm": 0.778626561164856, + "learning_rate": 2.284915319937752e-06, + "loss": 0.22291188240051268, + "step": 11255, + "token_acc": 0.9364655632674853 + }, + { + "epoch": 0.6989013717335981, + "grad_norm": 0.8070430755615234, + "learning_rate": 2.2806075973438748e-06, + "loss": 0.2195215940475464, + "step": 11260, + "token_acc": 0.9365901474308876 + }, + { + "epoch": 0.6992117187015083, + "grad_norm": 0.9077206254005432, + "learning_rate": 2.276302739224284e-06, + "loss": 0.23131659030914306, + "step": 11265, + "token_acc": 0.9344899310455528 + }, + { + "epoch": 0.6995220656694184, + "grad_norm": 0.7712111473083496, + "learning_rate": 2.2720007501135037e-06, + "loss": 0.22731573581695558, + "step": 11270, + "token_acc": 0.935950094811817 + }, + { + "epoch": 0.6998324126373285, + "grad_norm": 0.8345860242843628, + "learning_rate": 2.267701634543041e-06, + "loss": 0.2265024185180664, + "step": 11275, + "token_acc": 0.9350898970888559 + }, + { + "epoch": 0.7001427596052386, + "grad_norm": 0.8424535393714905, + "learning_rate": 2.263405397041369e-06, + "loss": 0.2252957820892334, + "step": 11280, + "token_acc": 0.9358640920066251 + }, + { + "epoch": 0.7004531065731487, + "grad_norm": 0.8819962739944458, + "learning_rate": 2.2591120421339345e-06, + "loss": 0.2299027919769287, + "step": 11285, + "token_acc": 0.9340271948967601 + }, + { + "epoch": 0.700763453541059, + "grad_norm": 0.9797229766845703, + "learning_rate": 2.2548215743431497e-06, + "loss": 0.22674252986907958, + "step": 11290, + "token_acc": 0.9356459998138521 + }, + { + "epoch": 0.7010738005089691, + "grad_norm": 0.9094035029411316, + "learning_rate": 2.2505339981883773e-06, + "loss": 0.22792327404022217, + "step": 11295, + "token_acc": 0.9341443035809446 + }, + { + "epoch": 0.7013841474768792, + "grad_norm": 0.8891135454177856, + "learning_rate": 2.2462493181859417e-06, + "loss": 0.22852392196655275, + "step": 11300, + "token_acc": 0.9351136423981732 + }, + { + "epoch": 0.7016944944447893, + "grad_norm": 1.0466634035110474, + "learning_rate": 2.2419675388491167e-06, + "loss": 0.22440414428710936, + "step": 11305, + "token_acc": 0.9360056624112001 + }, + { + "epoch": 0.7020048414126994, + "grad_norm": 0.8497759103775024, + "learning_rate": 2.2376886646881157e-06, + "loss": 0.22416322231292723, + "step": 11310, + "token_acc": 0.9354453755741782 + }, + { + "epoch": 0.7023151883806095, + "grad_norm": 0.828794538974762, + "learning_rate": 2.2334127002100925e-06, + "loss": 0.22113754749298095, + "step": 11315, + "token_acc": 0.9358799364338325 + }, + { + "epoch": 0.7026255353485197, + "grad_norm": 0.7986443042755127, + "learning_rate": 2.2291396499191397e-06, + "loss": 0.22280311584472656, + "step": 11320, + "token_acc": 0.9362347704625502 + }, + { + "epoch": 0.7029358823164298, + "grad_norm": 0.9161672592163086, + "learning_rate": 2.2248695183162807e-06, + "loss": 0.2269641399383545, + "step": 11325, + "token_acc": 0.9356395976398758 + }, + { + "epoch": 0.7032462292843399, + "grad_norm": 0.8236754536628723, + "learning_rate": 2.2206023098994585e-06, + "loss": 0.22793183326721192, + "step": 11330, + "token_acc": 0.9347986027798583 + }, + { + "epoch": 0.70355657625225, + "grad_norm": 0.8141815066337585, + "learning_rate": 2.216338029163542e-06, + "loss": 0.22462878227233887, + "step": 11335, + "token_acc": 0.935318840385106 + }, + { + "epoch": 0.7038669232201601, + "grad_norm": 0.8561828136444092, + "learning_rate": 2.212076680600315e-06, + "loss": 0.21799032688140868, + "step": 11340, + "token_acc": 0.9374638956432957 + }, + { + "epoch": 0.7041772701880703, + "grad_norm": 0.8436933159828186, + "learning_rate": 2.207818268698475e-06, + "loss": 0.22607154846191407, + "step": 11345, + "token_acc": 0.9357327154968142 + }, + { + "epoch": 0.7044876171559804, + "grad_norm": 0.8206582069396973, + "learning_rate": 2.203562797943622e-06, + "loss": 0.22091755867004395, + "step": 11350, + "token_acc": 0.9369103773584906 + }, + { + "epoch": 0.7047979641238905, + "grad_norm": 1.2672698497772217, + "learning_rate": 2.1993102728182553e-06, + "loss": 0.22291419506072999, + "step": 11355, + "token_acc": 0.9365724440867786 + }, + { + "epoch": 0.7051083110918006, + "grad_norm": 0.7974629998207092, + "learning_rate": 2.1950606978017836e-06, + "loss": 0.2211519956588745, + "step": 11360, + "token_acc": 0.937382729849495 + }, + { + "epoch": 0.7054186580597107, + "grad_norm": 0.8495259284973145, + "learning_rate": 2.190814077370497e-06, + "loss": 0.22344911098480225, + "step": 11365, + "token_acc": 0.9356689272087474 + }, + { + "epoch": 0.7057290050276209, + "grad_norm": 0.8161197304725647, + "learning_rate": 2.186570415997574e-06, + "loss": 0.22727460861206056, + "step": 11370, + "token_acc": 0.9349503060620022 + }, + { + "epoch": 0.706039351995531, + "grad_norm": 0.8287402987480164, + "learning_rate": 2.182329718153081e-06, + "loss": 0.22594599723815917, + "step": 11375, + "token_acc": 0.935421525721324 + }, + { + "epoch": 0.7063496989634411, + "grad_norm": 0.7981970310211182, + "learning_rate": 2.178091988303964e-06, + "loss": 0.22268738746643066, + "step": 11380, + "token_acc": 0.9359145351931303 + }, + { + "epoch": 0.7066600459313512, + "grad_norm": 0.8360260725021362, + "learning_rate": 2.1738572309140334e-06, + "loss": 0.21693720817565917, + "step": 11385, + "token_acc": 0.9370433271456219 + }, + { + "epoch": 0.7069703928992613, + "grad_norm": 0.8385817408561707, + "learning_rate": 2.169625450443977e-06, + "loss": 0.22814655303955078, + "step": 11390, + "token_acc": 0.9344309628647737 + }, + { + "epoch": 0.7072807398671715, + "grad_norm": 0.8320360779762268, + "learning_rate": 2.1653966513513474e-06, + "loss": 0.22769603729248047, + "step": 11395, + "token_acc": 0.9354038526529233 + }, + { + "epoch": 0.7075910868350817, + "grad_norm": 1.1740241050720215, + "learning_rate": 2.1611708380905487e-06, + "loss": 0.22475013732910157, + "step": 11400, + "token_acc": 0.9361235505010046 + }, + { + "epoch": 0.7079014338029918, + "grad_norm": 0.8004316687583923, + "learning_rate": 2.156948015112849e-06, + "loss": 0.21903834342956544, + "step": 11405, + "token_acc": 0.9363598577564309 + }, + { + "epoch": 0.7082117807709019, + "grad_norm": 0.9121341705322266, + "learning_rate": 2.1527281868663586e-06, + "loss": 0.22426993846893312, + "step": 11410, + "token_acc": 0.9360899279626356 + }, + { + "epoch": 0.708522127738812, + "grad_norm": 0.8070510029792786, + "learning_rate": 2.1485113577960393e-06, + "loss": 0.22907538414001466, + "step": 11415, + "token_acc": 0.9342424120710662 + }, + { + "epoch": 0.7088324747067221, + "grad_norm": 0.8399884104728699, + "learning_rate": 2.144297532343693e-06, + "loss": 0.228200364112854, + "step": 11420, + "token_acc": 0.9347632111655657 + }, + { + "epoch": 0.7091428216746323, + "grad_norm": 0.8571485280990601, + "learning_rate": 2.1400867149479526e-06, + "loss": 0.2183610439300537, + "step": 11425, + "token_acc": 0.9370202453946886 + }, + { + "epoch": 0.7094531686425424, + "grad_norm": 0.8999316096305847, + "learning_rate": 2.135878910044289e-06, + "loss": 0.22797598838806152, + "step": 11430, + "token_acc": 0.9349264290321079 + }, + { + "epoch": 0.7097635156104525, + "grad_norm": 0.8289883732795715, + "learning_rate": 2.1316741220649973e-06, + "loss": 0.22721989154815675, + "step": 11435, + "token_acc": 0.9349810614824265 + }, + { + "epoch": 0.7100738625783626, + "grad_norm": 0.8120068907737732, + "learning_rate": 2.127472355439194e-06, + "loss": 0.2281952381134033, + "step": 11440, + "token_acc": 0.9357425467441214 + }, + { + "epoch": 0.7103842095462727, + "grad_norm": 0.9025102853775024, + "learning_rate": 2.123273614592809e-06, + "loss": 0.22620253562927245, + "step": 11445, + "token_acc": 0.9349097007396321 + }, + { + "epoch": 0.7106945565141829, + "grad_norm": 0.8330454230308533, + "learning_rate": 2.1190779039485974e-06, + "loss": 0.2238757610321045, + "step": 11450, + "token_acc": 0.9360691854920917 + }, + { + "epoch": 0.711004903482093, + "grad_norm": 0.8168461322784424, + "learning_rate": 2.1148852279261104e-06, + "loss": 0.21892166137695312, + "step": 11455, + "token_acc": 0.937831428038479 + }, + { + "epoch": 0.7113152504500031, + "grad_norm": 0.8109251260757446, + "learning_rate": 2.110695590941706e-06, + "loss": 0.22056303024291993, + "step": 11460, + "token_acc": 0.9363772900801238 + }, + { + "epoch": 0.7116255974179132, + "grad_norm": 0.8899949789047241, + "learning_rate": 2.106508997408542e-06, + "loss": 0.23021910190582276, + "step": 11465, + "token_acc": 0.9339475276021829 + }, + { + "epoch": 0.7119359443858233, + "grad_norm": 0.7874158024787903, + "learning_rate": 2.1023254517365738e-06, + "loss": 0.22446961402893068, + "step": 11470, + "token_acc": 0.9353535754186251 + }, + { + "epoch": 0.7122462913537335, + "grad_norm": 0.814130961894989, + "learning_rate": 2.098144958332538e-06, + "loss": 0.22216389179229737, + "step": 11475, + "token_acc": 0.9355670272580804 + }, + { + "epoch": 0.7125566383216436, + "grad_norm": 0.7932929396629333, + "learning_rate": 2.0939675215999628e-06, + "loss": 0.22066638469696045, + "step": 11480, + "token_acc": 0.9360412206443672 + }, + { + "epoch": 0.7128669852895537, + "grad_norm": 0.8336986899375916, + "learning_rate": 2.0897931459391553e-06, + "loss": 0.22824666500091553, + "step": 11485, + "token_acc": 0.9341645056187247 + }, + { + "epoch": 0.7131773322574638, + "grad_norm": 0.8800667524337769, + "learning_rate": 2.085621835747199e-06, + "loss": 0.2271726608276367, + "step": 11490, + "token_acc": 0.9353246906847317 + }, + { + "epoch": 0.7134876792253739, + "grad_norm": 14.337824821472168, + "learning_rate": 2.0814535954179453e-06, + "loss": 0.23137245178222657, + "step": 11495, + "token_acc": 0.9341251116354088 + }, + { + "epoch": 0.7137980261932841, + "grad_norm": 0.7957608699798584, + "learning_rate": 2.0772884293420097e-06, + "loss": 0.22668683528900146, + "step": 11500, + "token_acc": 0.9352474931510493 + }, + { + "epoch": 0.7137980261932841, + "eval_loss": 0.23897504806518555, + "eval_runtime": 93.2977, + "eval_samples_per_second": 214.368, + "eval_steps_per_second": 0.289, + "eval_token_acc": 0.9313971997686391, + "step": 11500 + }, + { + "epoch": 0.7141083731611942, + "grad_norm": 0.816537082195282, + "learning_rate": 2.0731263419067816e-06, + "loss": 0.22484395503997803, + "step": 11505, + "token_acc": 0.9354346922208803 + }, + { + "epoch": 0.7144187201291043, + "grad_norm": 0.8282453417778015, + "learning_rate": 2.0689673374963943e-06, + "loss": 0.2287168025970459, + "step": 11510, + "token_acc": 0.934382608464271 + }, + { + "epoch": 0.7147290670970144, + "grad_norm": 0.7940766215324402, + "learning_rate": 2.064811420491738e-06, + "loss": 0.22521204948425294, + "step": 11515, + "token_acc": 0.934837901950448 + }, + { + "epoch": 0.7150394140649246, + "grad_norm": 0.8438672423362732, + "learning_rate": 2.060658595270453e-06, + "loss": 0.22777106761932372, + "step": 11520, + "token_acc": 0.9350305726866448 + }, + { + "epoch": 0.7153497610328348, + "grad_norm": 0.8346472978591919, + "learning_rate": 2.056508866206923e-06, + "loss": 0.22149090766906737, + "step": 11525, + "token_acc": 0.9368831411988124 + }, + { + "epoch": 0.7156601080007449, + "grad_norm": 0.7856476306915283, + "learning_rate": 2.052362237672264e-06, + "loss": 0.2246570587158203, + "step": 11530, + "token_acc": 0.9351264473382097 + }, + { + "epoch": 0.715970454968655, + "grad_norm": 0.8185098767280579, + "learning_rate": 2.048218714034334e-06, + "loss": 0.2287123441696167, + "step": 11535, + "token_acc": 0.934707445033307 + }, + { + "epoch": 0.7162808019365651, + "grad_norm": 0.785542905330658, + "learning_rate": 2.044078299657718e-06, + "loss": 0.22114794254302977, + "step": 11540, + "token_acc": 0.9366497609424941 + }, + { + "epoch": 0.7165911489044752, + "grad_norm": 0.901665985584259, + "learning_rate": 2.0399409989037215e-06, + "loss": 0.22261061668395996, + "step": 11545, + "token_acc": 0.9354149211982874 + }, + { + "epoch": 0.7169014958723853, + "grad_norm": 0.8484013080596924, + "learning_rate": 2.0358068161303772e-06, + "loss": 0.22815556526184083, + "step": 11550, + "token_acc": 0.9344793732580434 + }, + { + "epoch": 0.7172118428402955, + "grad_norm": 0.8181394934654236, + "learning_rate": 2.0316757556924276e-06, + "loss": 0.2276826858520508, + "step": 11555, + "token_acc": 0.9346822813850504 + }, + { + "epoch": 0.7175221898082056, + "grad_norm": 0.9296860098838806, + "learning_rate": 2.027547821941329e-06, + "loss": 0.2251561164855957, + "step": 11560, + "token_acc": 0.9355666091428958 + }, + { + "epoch": 0.7178325367761157, + "grad_norm": 0.8062475919723511, + "learning_rate": 2.023423019225246e-06, + "loss": 0.22651951313018798, + "step": 11565, + "token_acc": 0.935028211173214 + }, + { + "epoch": 0.7181428837440258, + "grad_norm": 0.7735936045646667, + "learning_rate": 2.019301351889041e-06, + "loss": 0.21892569065093995, + "step": 11570, + "token_acc": 0.9371403481928432 + }, + { + "epoch": 0.7184532307119359, + "grad_norm": 0.7993139624595642, + "learning_rate": 2.0151828242742754e-06, + "loss": 0.22368199825286866, + "step": 11575, + "token_acc": 0.9360977484461761 + }, + { + "epoch": 0.7187635776798461, + "grad_norm": 0.833685040473938, + "learning_rate": 2.0110674407192065e-06, + "loss": 0.2147491455078125, + "step": 11580, + "token_acc": 0.9374438888852482 + }, + { + "epoch": 0.7190739246477562, + "grad_norm": 0.8915134072303772, + "learning_rate": 2.006955205558775e-06, + "loss": 0.2216745376586914, + "step": 11585, + "token_acc": 0.9369019589839779 + }, + { + "epoch": 0.7193842716156663, + "grad_norm": 0.8533118367195129, + "learning_rate": 2.0028461231246043e-06, + "loss": 0.2222693920135498, + "step": 11590, + "token_acc": 0.9365064338794636 + }, + { + "epoch": 0.7196946185835764, + "grad_norm": 0.7953152656555176, + "learning_rate": 1.9987401977450066e-06, + "loss": 0.225363826751709, + "step": 11595, + "token_acc": 0.9349980286502826 + }, + { + "epoch": 0.7200049655514865, + "grad_norm": 0.888439953327179, + "learning_rate": 1.994637433744958e-06, + "loss": 0.22201476097106934, + "step": 11600, + "token_acc": 0.9369021814783145 + }, + { + "epoch": 0.7203153125193967, + "grad_norm": 0.8516317009925842, + "learning_rate": 1.990537835446107e-06, + "loss": 0.22743144035339355, + "step": 11605, + "token_acc": 0.9342806746338467 + }, + { + "epoch": 0.7206256594873068, + "grad_norm": 0.8538787364959717, + "learning_rate": 1.9864414071667707e-06, + "loss": 0.22105045318603517, + "step": 11610, + "token_acc": 0.9357822357881691 + }, + { + "epoch": 0.7209360064552169, + "grad_norm": 0.820755660533905, + "learning_rate": 1.9823481532219274e-06, + "loss": 0.2157576084136963, + "step": 11615, + "token_acc": 0.9370018495602305 + }, + { + "epoch": 0.721246353423127, + "grad_norm": 0.8657658696174622, + "learning_rate": 1.9782580779232054e-06, + "loss": 0.2230445384979248, + "step": 11620, + "token_acc": 0.9356213048778889 + }, + { + "epoch": 0.7215567003910371, + "grad_norm": 0.8241891860961914, + "learning_rate": 1.9741711855788908e-06, + "loss": 0.2284527063369751, + "step": 11625, + "token_acc": 0.9352653621880765 + }, + { + "epoch": 0.7218670473589474, + "grad_norm": 0.805472195148468, + "learning_rate": 1.970087480493916e-06, + "loss": 0.22618067264556885, + "step": 11630, + "token_acc": 0.934821991706813 + }, + { + "epoch": 0.7221773943268575, + "grad_norm": 0.8234931826591492, + "learning_rate": 1.9660069669698557e-06, + "loss": 0.22793221473693848, + "step": 11635, + "token_acc": 0.9350987714528122 + }, + { + "epoch": 0.7224877412947676, + "grad_norm": 0.7847504615783691, + "learning_rate": 1.961929649304921e-06, + "loss": 0.22696521282196044, + "step": 11640, + "token_acc": 0.9354513697858322 + }, + { + "epoch": 0.7227980882626777, + "grad_norm": 0.8221176266670227, + "learning_rate": 1.9578555317939562e-06, + "loss": 0.2241072654724121, + "step": 11645, + "token_acc": 0.9360338771624593 + }, + { + "epoch": 0.7231084352305878, + "grad_norm": 0.8236778974533081, + "learning_rate": 1.9537846187284387e-06, + "loss": 0.23153796195983886, + "step": 11650, + "token_acc": 0.9343077713517858 + }, + { + "epoch": 0.7234187821984979, + "grad_norm": 0.8666704297065735, + "learning_rate": 1.9497169143964693e-06, + "loss": 0.22997372150421141, + "step": 11655, + "token_acc": 0.9345043384019613 + }, + { + "epoch": 0.7237291291664081, + "grad_norm": 1.3000288009643555, + "learning_rate": 1.9456524230827646e-06, + "loss": 0.2256190538406372, + "step": 11660, + "token_acc": 0.9351543662857742 + }, + { + "epoch": 0.7240394761343182, + "grad_norm": 0.78895103931427, + "learning_rate": 1.941591149068662e-06, + "loss": 0.22627968788146974, + "step": 11665, + "token_acc": 0.9349821923229126 + }, + { + "epoch": 0.7243498231022283, + "grad_norm": 0.8877391815185547, + "learning_rate": 1.9375330966321088e-06, + "loss": 0.23482034206390381, + "step": 11670, + "token_acc": 0.932870674048225 + }, + { + "epoch": 0.7246601700701384, + "grad_norm": 0.9120040535926819, + "learning_rate": 1.9334782700476544e-06, + "loss": 0.22382059097290039, + "step": 11675, + "token_acc": 0.9362643411945569 + }, + { + "epoch": 0.7249705170380485, + "grad_norm": 0.8090142607688904, + "learning_rate": 1.9294266735864587e-06, + "loss": 0.22627110481262208, + "step": 11680, + "token_acc": 0.9347649891822012 + }, + { + "epoch": 0.7252808640059587, + "grad_norm": 0.8415210843086243, + "learning_rate": 1.92537831151627e-06, + "loss": 0.22471764087677001, + "step": 11685, + "token_acc": 0.9362176326773196 + }, + { + "epoch": 0.7255912109738688, + "grad_norm": 0.8303400278091431, + "learning_rate": 1.9213331881014354e-06, + "loss": 0.224354887008667, + "step": 11690, + "token_acc": 0.9354195602044139 + }, + { + "epoch": 0.7259015579417789, + "grad_norm": 0.8743711113929749, + "learning_rate": 1.9172913076028903e-06, + "loss": 0.22347564697265626, + "step": 11695, + "token_acc": 0.9360752431053723 + }, + { + "epoch": 0.726211904909689, + "grad_norm": 0.8906884789466858, + "learning_rate": 1.913252674278151e-06, + "loss": 0.2219155550003052, + "step": 11700, + "token_acc": 0.9357108344419954 + }, + { + "epoch": 0.7265222518775991, + "grad_norm": 0.8338742256164551, + "learning_rate": 1.9092172923813155e-06, + "loss": 0.2225968360900879, + "step": 11705, + "token_acc": 0.9356678568776873 + }, + { + "epoch": 0.7268325988455093, + "grad_norm": 0.8056448101997375, + "learning_rate": 1.90518516616306e-06, + "loss": 0.2223519802093506, + "step": 11710, + "token_acc": 0.935464358831193 + }, + { + "epoch": 0.7271429458134194, + "grad_norm": 0.8036158084869385, + "learning_rate": 1.9011562998706246e-06, + "loss": 0.22756190299987794, + "step": 11715, + "token_acc": 0.9353028006328067 + }, + { + "epoch": 0.7274532927813295, + "grad_norm": 0.8429133296012878, + "learning_rate": 1.897130697747817e-06, + "loss": 0.23417332172393798, + "step": 11720, + "token_acc": 0.9350522694632778 + }, + { + "epoch": 0.7277636397492396, + "grad_norm": 0.8456043004989624, + "learning_rate": 1.8931083640350144e-06, + "loss": 0.22276525497436522, + "step": 11725, + "token_acc": 0.9360703619835253 + }, + { + "epoch": 0.7280739867171497, + "grad_norm": 0.8521315455436707, + "learning_rate": 1.8890893029691431e-06, + "loss": 0.2246774196624756, + "step": 11730, + "token_acc": 0.9361267859286829 + }, + { + "epoch": 0.72838433368506, + "grad_norm": 0.849325954914093, + "learning_rate": 1.8850735187836821e-06, + "loss": 0.22037279605865479, + "step": 11735, + "token_acc": 0.9365392742063282 + }, + { + "epoch": 0.72869468065297, + "grad_norm": 0.8131787776947021, + "learning_rate": 1.881061015708663e-06, + "loss": 0.2219088077545166, + "step": 11740, + "token_acc": 0.9364451564442258 + }, + { + "epoch": 0.7290050276208802, + "grad_norm": 0.8253371119499207, + "learning_rate": 1.8770517979706603e-06, + "loss": 0.22256295680999755, + "step": 11745, + "token_acc": 0.9361392354339754 + }, + { + "epoch": 0.7293153745887903, + "grad_norm": 0.8463642597198486, + "learning_rate": 1.8730458697927845e-06, + "loss": 0.22770445346832274, + "step": 11750, + "token_acc": 0.9346295386005234 + }, + { + "epoch": 0.7296257215567004, + "grad_norm": 0.8377583622932434, + "learning_rate": 1.8690432353946848e-06, + "loss": 0.22779617309570313, + "step": 11755, + "token_acc": 0.9351997775099657 + }, + { + "epoch": 0.7299360685246105, + "grad_norm": 0.8323188424110413, + "learning_rate": 1.865043898992539e-06, + "loss": 0.22305238246917725, + "step": 11760, + "token_acc": 0.9353744420565383 + }, + { + "epoch": 0.7302464154925207, + "grad_norm": 0.8475223779678345, + "learning_rate": 1.8610478647990543e-06, + "loss": 0.21806046962738038, + "step": 11765, + "token_acc": 0.9371644240236298 + }, + { + "epoch": 0.7305567624604308, + "grad_norm": 0.8056929707527161, + "learning_rate": 1.8570551370234535e-06, + "loss": 0.2133943557739258, + "step": 11770, + "token_acc": 0.9381683829264101 + }, + { + "epoch": 0.7308671094283409, + "grad_norm": 0.8484786748886108, + "learning_rate": 1.8530657198714774e-06, + "loss": 0.2236856460571289, + "step": 11775, + "token_acc": 0.9362989299788429 + }, + { + "epoch": 0.731177456396251, + "grad_norm": 0.8628672361373901, + "learning_rate": 1.8490796175453885e-06, + "loss": 0.22446949481964112, + "step": 11780, + "token_acc": 0.935879990707245 + }, + { + "epoch": 0.7314878033641611, + "grad_norm": 0.8267581462860107, + "learning_rate": 1.845096834243948e-06, + "loss": 0.2264786958694458, + "step": 11785, + "token_acc": 0.9344405026444713 + }, + { + "epoch": 0.7317981503320713, + "grad_norm": 0.7866992354393005, + "learning_rate": 1.8411173741624217e-06, + "loss": 0.21973271369934083, + "step": 11790, + "token_acc": 0.9376951863019599 + }, + { + "epoch": 0.7321084972999814, + "grad_norm": 0.838427722454071, + "learning_rate": 1.8371412414925793e-06, + "loss": 0.22252316474914552, + "step": 11795, + "token_acc": 0.9363571677899203 + }, + { + "epoch": 0.7324188442678915, + "grad_norm": 0.862309992313385, + "learning_rate": 1.833168440422685e-06, + "loss": 0.22745203971862793, + "step": 11800, + "token_acc": 0.9355523122246547 + }, + { + "epoch": 0.7327291912358016, + "grad_norm": 0.8222072124481201, + "learning_rate": 1.8291989751374889e-06, + "loss": 0.22590203285217286, + "step": 11805, + "token_acc": 0.9347667783235578 + }, + { + "epoch": 0.7330395382037117, + "grad_norm": 0.8054452538490295, + "learning_rate": 1.8252328498182325e-06, + "loss": 0.22451086044311525, + "step": 11810, + "token_acc": 0.9356128002511905 + }, + { + "epoch": 0.7333498851716219, + "grad_norm": 0.8254367709159851, + "learning_rate": 1.8212700686426387e-06, + "loss": 0.22783880233764647, + "step": 11815, + "token_acc": 0.9351945570935416 + }, + { + "epoch": 0.733660232139532, + "grad_norm": 0.8604287505149841, + "learning_rate": 1.8173106357849036e-06, + "loss": 0.22266287803649903, + "step": 11820, + "token_acc": 0.9355637917793582 + }, + { + "epoch": 0.7339705791074421, + "grad_norm": 0.790553092956543, + "learning_rate": 1.8133545554157016e-06, + "loss": 0.22658135890960693, + "step": 11825, + "token_acc": 0.9357785775539005 + }, + { + "epoch": 0.7342809260753522, + "grad_norm": 0.819529116153717, + "learning_rate": 1.8094018317021706e-06, + "loss": 0.226456356048584, + "step": 11830, + "token_acc": 0.9341774808034046 + }, + { + "epoch": 0.7345912730432623, + "grad_norm": 0.7776014804840088, + "learning_rate": 1.8054524688079162e-06, + "loss": 0.22117271423339843, + "step": 11835, + "token_acc": 0.9367474500857198 + }, + { + "epoch": 0.7349016200111725, + "grad_norm": 0.8458400964736938, + "learning_rate": 1.8015064708930063e-06, + "loss": 0.23261499404907227, + "step": 11840, + "token_acc": 0.9337650518985505 + }, + { + "epoch": 0.7352119669790826, + "grad_norm": 0.8317359685897827, + "learning_rate": 1.797563842113957e-06, + "loss": 0.22810475826263427, + "step": 11845, + "token_acc": 0.9353031843373889 + }, + { + "epoch": 0.7355223139469927, + "grad_norm": 0.8345915675163269, + "learning_rate": 1.7936245866237406e-06, + "loss": 0.22071366310119628, + "step": 11850, + "token_acc": 0.9368295668861463 + }, + { + "epoch": 0.7358326609149028, + "grad_norm": 0.8003770112991333, + "learning_rate": 1.7896887085717778e-06, + "loss": 0.2188581943511963, + "step": 11855, + "token_acc": 0.9375458492773162 + }, + { + "epoch": 0.736143007882813, + "grad_norm": 0.8270089626312256, + "learning_rate": 1.7857562121039279e-06, + "loss": 0.22340893745422363, + "step": 11860, + "token_acc": 0.9361995746638311 + }, + { + "epoch": 0.7364533548507232, + "grad_norm": 0.9574822187423706, + "learning_rate": 1.781827101362485e-06, + "loss": 0.22251460552215577, + "step": 11865, + "token_acc": 0.9365585420252774 + }, + { + "epoch": 0.7367637018186333, + "grad_norm": 0.8958150744438171, + "learning_rate": 1.7779013804861883e-06, + "loss": 0.21838464736938476, + "step": 11870, + "token_acc": 0.9366214533517117 + }, + { + "epoch": 0.7370740487865434, + "grad_norm": 0.8363583087921143, + "learning_rate": 1.7739790536101953e-06, + "loss": 0.23243269920349122, + "step": 11875, + "token_acc": 0.934932069053826 + }, + { + "epoch": 0.7373843957544535, + "grad_norm": 0.8786102533340454, + "learning_rate": 1.770060124866092e-06, + "loss": 0.21895127296447753, + "step": 11880, + "token_acc": 0.9367706667905027 + }, + { + "epoch": 0.7376947427223636, + "grad_norm": 0.8465340733528137, + "learning_rate": 1.7661445983818853e-06, + "loss": 0.2259925365447998, + "step": 11885, + "token_acc": 0.9348885318206361 + }, + { + "epoch": 0.7380050896902737, + "grad_norm": 0.8292161226272583, + "learning_rate": 1.7622324782820015e-06, + "loss": 0.21916780471801758, + "step": 11890, + "token_acc": 0.9372786205974738 + }, + { + "epoch": 0.7383154366581839, + "grad_norm": 0.9127896428108215, + "learning_rate": 1.7583237686872717e-06, + "loss": 0.22914643287658693, + "step": 11895, + "token_acc": 0.9339077044959397 + }, + { + "epoch": 0.738625783626094, + "grad_norm": 0.7742658257484436, + "learning_rate": 1.7544184737149395e-06, + "loss": 0.22057149410247803, + "step": 11900, + "token_acc": 0.9369847933204677 + }, + { + "epoch": 0.7389361305940041, + "grad_norm": 0.8503296971321106, + "learning_rate": 1.7505165974786515e-06, + "loss": 0.22712323665618897, + "step": 11905, + "token_acc": 0.9360567429239264 + }, + { + "epoch": 0.7392464775619142, + "grad_norm": 0.8230908513069153, + "learning_rate": 1.7466181440884538e-06, + "loss": 0.22030303478240967, + "step": 11910, + "token_acc": 0.9370584628508971 + }, + { + "epoch": 0.7395568245298243, + "grad_norm": 0.8617457151412964, + "learning_rate": 1.7427231176507836e-06, + "loss": 0.2318741798400879, + "step": 11915, + "token_acc": 0.9336937319856018 + }, + { + "epoch": 0.7398671714977345, + "grad_norm": 0.8533957004547119, + "learning_rate": 1.7388315222684677e-06, + "loss": 0.22296578884124757, + "step": 11920, + "token_acc": 0.9357838912827312 + }, + { + "epoch": 0.7401775184656446, + "grad_norm": 0.9361258149147034, + "learning_rate": 1.734943362040727e-06, + "loss": 0.21163334846496581, + "step": 11925, + "token_acc": 0.9395680615364387 + }, + { + "epoch": 0.7404878654335547, + "grad_norm": 0.8410860300064087, + "learning_rate": 1.7310586410631559e-06, + "loss": 0.2111081600189209, + "step": 11930, + "token_acc": 0.9393605938273213 + }, + { + "epoch": 0.7407982124014648, + "grad_norm": 1.4199076890945435, + "learning_rate": 1.7271773634277261e-06, + "loss": 0.20328569412231445, + "step": 11935, + "token_acc": 0.9412987938922227 + }, + { + "epoch": 0.7411085593693749, + "grad_norm": 0.862643301486969, + "learning_rate": 1.723299533222787e-06, + "loss": 0.20604259967803956, + "step": 11940, + "token_acc": 0.9401932088151522 + }, + { + "epoch": 0.7414189063372851, + "grad_norm": 0.8482420444488525, + "learning_rate": 1.7194251545330558e-06, + "loss": 0.20177230834960938, + "step": 11945, + "token_acc": 0.9417097253447981 + }, + { + "epoch": 0.7417292533051952, + "grad_norm": 0.8138321042060852, + "learning_rate": 1.715554231439609e-06, + "loss": 0.2096508502960205, + "step": 11950, + "token_acc": 0.9399653797733949 + }, + { + "epoch": 0.7420396002731053, + "grad_norm": 0.9539263248443604, + "learning_rate": 1.7116867680198889e-06, + "loss": 0.21061038970947266, + "step": 11955, + "token_acc": 0.9394313480858888 + }, + { + "epoch": 0.7423499472410154, + "grad_norm": 0.79305100440979, + "learning_rate": 1.7078227683476922e-06, + "loss": 0.2094661235809326, + "step": 11960, + "token_acc": 0.9400579415365882 + }, + { + "epoch": 0.7426602942089255, + "grad_norm": 1.0689914226531982, + "learning_rate": 1.7039622364931635e-06, + "loss": 0.21002235412597656, + "step": 11965, + "token_acc": 0.9398144455964059 + }, + { + "epoch": 0.7429706411768358, + "grad_norm": 0.8461315035820007, + "learning_rate": 1.700105176522801e-06, + "loss": 0.21310892105102539, + "step": 11970, + "token_acc": 0.9381738114737591 + }, + { + "epoch": 0.7432809881447459, + "grad_norm": 0.8767081499099731, + "learning_rate": 1.6962515924994376e-06, + "loss": 0.2121664047241211, + "step": 11975, + "token_acc": 0.9388887023669632 + }, + { + "epoch": 0.743591335112656, + "grad_norm": 0.8118957877159119, + "learning_rate": 1.692401488482252e-06, + "loss": 0.209603214263916, + "step": 11980, + "token_acc": 0.9390887544895073 + }, + { + "epoch": 0.7439016820805661, + "grad_norm": 0.8300259709358215, + "learning_rate": 1.6885548685267556e-06, + "loss": 0.209900426864624, + "step": 11985, + "token_acc": 0.9392625591663021 + }, + { + "epoch": 0.7442120290484762, + "grad_norm": 0.9168404340744019, + "learning_rate": 1.6847117366847859e-06, + "loss": 0.2106637477874756, + "step": 11990, + "token_acc": 0.9392223833809011 + }, + { + "epoch": 0.7445223760163863, + "grad_norm": 0.8527868390083313, + "learning_rate": 1.6808720970045106e-06, + "loss": 0.20959744453430176, + "step": 11995, + "token_acc": 0.9395378390128081 + }, + { + "epoch": 0.7448327229842965, + "grad_norm": 0.7976640462875366, + "learning_rate": 1.6770359535304187e-06, + "loss": 0.21057610511779784, + "step": 12000, + "token_acc": 0.9396767827617473 + }, + { + "epoch": 0.7448327229842965, + "eval_loss": 0.23060274124145508, + "eval_runtime": 115.2949, + "eval_samples_per_second": 173.468, + "eval_steps_per_second": 0.234, + "eval_token_acc": 0.9341296587652044, + "step": 12000 + }, + { + "epoch": 0.7451430699522066, + "grad_norm": 0.8579834699630737, + "learning_rate": 1.6732033103033135e-06, + "loss": 0.2176448345184326, + "step": 12005, + "token_acc": 0.9383228181563538 + }, + { + "epoch": 0.7454534169201167, + "grad_norm": 0.8416378498077393, + "learning_rate": 1.6693741713603118e-06, + "loss": 0.21107897758483887, + "step": 12010, + "token_acc": 0.9393128903272396 + }, + { + "epoch": 0.7457637638880268, + "grad_norm": 0.8533751964569092, + "learning_rate": 1.6655485407348416e-06, + "loss": 0.21692337989807128, + "step": 12015, + "token_acc": 0.9378340661944758 + }, + { + "epoch": 0.7460741108559369, + "grad_norm": 0.8082554936408997, + "learning_rate": 1.6617264224566366e-06, + "loss": 0.2067805767059326, + "step": 12020, + "token_acc": 0.9404759541908977 + }, + { + "epoch": 0.7463844578238471, + "grad_norm": 0.8548649549484253, + "learning_rate": 1.6579078205517234e-06, + "loss": 0.21070413589477538, + "step": 12025, + "token_acc": 0.9399136262491935 + }, + { + "epoch": 0.7466948047917572, + "grad_norm": 0.8063479661941528, + "learning_rate": 1.6540927390424334e-06, + "loss": 0.2071539878845215, + "step": 12030, + "token_acc": 0.9398919887176334 + }, + { + "epoch": 0.7470051517596673, + "grad_norm": 0.8521004915237427, + "learning_rate": 1.6502811819473836e-06, + "loss": 0.21133480072021485, + "step": 12035, + "token_acc": 0.9389309971453883 + }, + { + "epoch": 0.7473154987275774, + "grad_norm": 0.8748247623443604, + "learning_rate": 1.6464731532814842e-06, + "loss": 0.21567037105560302, + "step": 12040, + "token_acc": 0.937842094900571 + }, + { + "epoch": 0.7476258456954875, + "grad_norm": 0.8335236310958862, + "learning_rate": 1.642668657055923e-06, + "loss": 0.2113555669784546, + "step": 12045, + "token_acc": 0.9402967185169401 + }, + { + "epoch": 0.7479361926633977, + "grad_norm": 0.8163561224937439, + "learning_rate": 1.638867697278167e-06, + "loss": 0.20977771282196045, + "step": 12050, + "token_acc": 0.9404030441562595 + }, + { + "epoch": 0.7482465396313078, + "grad_norm": 0.8098542094230652, + "learning_rate": 1.6350702779519656e-06, + "loss": 0.20851690769195558, + "step": 12055, + "token_acc": 0.9396547141224316 + }, + { + "epoch": 0.7485568865992179, + "grad_norm": 0.8388252258300781, + "learning_rate": 1.631276403077331e-06, + "loss": 0.20858533382415773, + "step": 12060, + "token_acc": 0.9402159963671544 + }, + { + "epoch": 0.748867233567128, + "grad_norm": 0.802611768245697, + "learning_rate": 1.6274860766505429e-06, + "loss": 0.21079573631286622, + "step": 12065, + "token_acc": 0.9401062836335298 + }, + { + "epoch": 0.7491775805350381, + "grad_norm": 0.8042132258415222, + "learning_rate": 1.6236993026641463e-06, + "loss": 0.20671520233154297, + "step": 12070, + "token_acc": 0.9405277406394881 + }, + { + "epoch": 0.7494879275029483, + "grad_norm": 0.7500441074371338, + "learning_rate": 1.6199160851069446e-06, + "loss": 0.20844645500183107, + "step": 12075, + "token_acc": 0.940571721742791 + }, + { + "epoch": 0.7497982744708585, + "grad_norm": 0.7927413582801819, + "learning_rate": 1.6161364279639902e-06, + "loss": 0.2077688217163086, + "step": 12080, + "token_acc": 0.9396938874987814 + }, + { + "epoch": 0.7501086214387686, + "grad_norm": 0.8253988027572632, + "learning_rate": 1.6123603352165895e-06, + "loss": 0.20730304718017578, + "step": 12085, + "token_acc": 0.9403645278192283 + }, + { + "epoch": 0.7504189684066787, + "grad_norm": 0.8104137778282166, + "learning_rate": 1.6085878108422942e-06, + "loss": 0.20916683673858644, + "step": 12090, + "token_acc": 0.9391622275077153 + }, + { + "epoch": 0.7507293153745888, + "grad_norm": 0.9537879228591919, + "learning_rate": 1.6048188588148932e-06, + "loss": 0.21004586219787597, + "step": 12095, + "token_acc": 0.9391532463746994 + }, + { + "epoch": 0.7510396623424989, + "grad_norm": 0.864083468914032, + "learning_rate": 1.601053483104419e-06, + "loss": 0.20677640438079833, + "step": 12100, + "token_acc": 0.9401905705733857 + }, + { + "epoch": 0.7513500093104091, + "grad_norm": 0.8001037240028381, + "learning_rate": 1.5972916876771293e-06, + "loss": 0.20555729866027833, + "step": 12105, + "token_acc": 0.940683318596132 + }, + { + "epoch": 0.7516603562783192, + "grad_norm": 0.7993642687797546, + "learning_rate": 1.5935334764955164e-06, + "loss": 0.2101348876953125, + "step": 12110, + "token_acc": 0.9392879805404262 + }, + { + "epoch": 0.7519707032462293, + "grad_norm": 0.7874516248703003, + "learning_rate": 1.5897788535182968e-06, + "loss": 0.2097177505493164, + "step": 12115, + "token_acc": 0.9393840627910071 + }, + { + "epoch": 0.7522810502141394, + "grad_norm": 0.9353976845741272, + "learning_rate": 1.5860278227004027e-06, + "loss": 0.20862700939178466, + "step": 12120, + "token_acc": 0.9398427451968336 + }, + { + "epoch": 0.7525913971820495, + "grad_norm": 0.8109471201896667, + "learning_rate": 1.582280387992987e-06, + "loss": 0.20483365058898925, + "step": 12125, + "token_acc": 0.9410378771239742 + }, + { + "epoch": 0.7529017441499597, + "grad_norm": 0.8004736304283142, + "learning_rate": 1.5785365533434148e-06, + "loss": 0.20788078308105468, + "step": 12130, + "token_acc": 0.9403029818530658 + }, + { + "epoch": 0.7532120911178698, + "grad_norm": 0.8099613785743713, + "learning_rate": 1.5747963226952562e-06, + "loss": 0.2177405834197998, + "step": 12135, + "token_acc": 0.9373907284768211 + }, + { + "epoch": 0.7535224380857799, + "grad_norm": 0.7759018540382385, + "learning_rate": 1.5710596999882826e-06, + "loss": 0.20580925941467285, + "step": 12140, + "token_acc": 0.9406147817772101 + }, + { + "epoch": 0.75383278505369, + "grad_norm": 0.817219614982605, + "learning_rate": 1.567326689158476e-06, + "loss": 0.20719144344329835, + "step": 12145, + "token_acc": 0.9404935655210562 + }, + { + "epoch": 0.7541431320216001, + "grad_norm": 0.8012793660163879, + "learning_rate": 1.5635972941380023e-06, + "loss": 0.20360662937164306, + "step": 12150, + "token_acc": 0.9412651698025859 + }, + { + "epoch": 0.7544534789895103, + "grad_norm": 0.8467557430267334, + "learning_rate": 1.5598715188552215e-06, + "loss": 0.2090917110443115, + "step": 12155, + "token_acc": 0.9400326972555746 + }, + { + "epoch": 0.7547638259574204, + "grad_norm": 0.822884738445282, + "learning_rate": 1.5561493672346845e-06, + "loss": 0.2088174819946289, + "step": 12160, + "token_acc": 0.9399390103339942 + }, + { + "epoch": 0.7550741729253305, + "grad_norm": 0.8913827538490295, + "learning_rate": 1.5524308431971235e-06, + "loss": 0.2104630947113037, + "step": 12165, + "token_acc": 0.9402813019656687 + }, + { + "epoch": 0.7553845198932406, + "grad_norm": 0.8416972160339355, + "learning_rate": 1.5487159506594462e-06, + "loss": 0.21290440559387208, + "step": 12170, + "token_acc": 0.9387724531398575 + }, + { + "epoch": 0.7556948668611507, + "grad_norm": 0.7917154431343079, + "learning_rate": 1.5450046935347395e-06, + "loss": 0.21078481674194335, + "step": 12175, + "token_acc": 0.9398784798794014 + }, + { + "epoch": 0.7560052138290609, + "grad_norm": 0.8262340426445007, + "learning_rate": 1.5412970757322588e-06, + "loss": 0.21544389724731444, + "step": 12180, + "token_acc": 0.9394761216208112 + }, + { + "epoch": 0.756315560796971, + "grad_norm": 0.8186915516853333, + "learning_rate": 1.5375931011574285e-06, + "loss": 0.20738086700439454, + "step": 12185, + "token_acc": 0.9407835288920422 + }, + { + "epoch": 0.7566259077648811, + "grad_norm": 0.7801491618156433, + "learning_rate": 1.5338927737118314e-06, + "loss": 0.20837364196777344, + "step": 12190, + "token_acc": 0.9406851349821156 + }, + { + "epoch": 0.7569362547327912, + "grad_norm": 0.7741080522537231, + "learning_rate": 1.5301960972932078e-06, + "loss": 0.21118624210357667, + "step": 12195, + "token_acc": 0.9392923025322824 + }, + { + "epoch": 0.7572466017007013, + "grad_norm": 0.8463383913040161, + "learning_rate": 1.5265030757954607e-06, + "loss": 0.2142542839050293, + "step": 12200, + "token_acc": 0.9389973628657298 + }, + { + "epoch": 0.7575569486686116, + "grad_norm": 0.7816526293754578, + "learning_rate": 1.5228137131086347e-06, + "loss": 0.21195504665374756, + "step": 12205, + "token_acc": 0.9398080629199438 + }, + { + "epoch": 0.7578672956365217, + "grad_norm": 0.9174579977989197, + "learning_rate": 1.5191280131189212e-06, + "loss": 0.21115508079528808, + "step": 12210, + "token_acc": 0.9396204639069268 + }, + { + "epoch": 0.7581776426044318, + "grad_norm": 0.8383579850196838, + "learning_rate": 1.5154459797086574e-06, + "loss": 0.20671536922454833, + "step": 12215, + "token_acc": 0.9399744151082742 + }, + { + "epoch": 0.7584879895723419, + "grad_norm": 0.7768596410751343, + "learning_rate": 1.5117676167563178e-06, + "loss": 0.20888133049011232, + "step": 12220, + "token_acc": 0.9404776017175749 + }, + { + "epoch": 0.758798336540252, + "grad_norm": 0.7632548213005066, + "learning_rate": 1.5080929281365053e-06, + "loss": 0.20717158317565917, + "step": 12225, + "token_acc": 0.939779384599145 + }, + { + "epoch": 0.7591086835081621, + "grad_norm": 0.7660151124000549, + "learning_rate": 1.5044219177199587e-06, + "loss": 0.20801215171813964, + "step": 12230, + "token_acc": 0.9408699204132188 + }, + { + "epoch": 0.7594190304760723, + "grad_norm": 0.8551371097564697, + "learning_rate": 1.500754589373541e-06, + "loss": 0.21742801666259765, + "step": 12235, + "token_acc": 0.9381134509930418 + }, + { + "epoch": 0.7597293774439824, + "grad_norm": 0.8587591648101807, + "learning_rate": 1.497090946960233e-06, + "loss": 0.21115584373474122, + "step": 12240, + "token_acc": 0.9393209459237721 + }, + { + "epoch": 0.7600397244118925, + "grad_norm": 0.7808635234832764, + "learning_rate": 1.4934309943391378e-06, + "loss": 0.2156130075454712, + "step": 12245, + "token_acc": 0.9389674756132832 + }, + { + "epoch": 0.7603500713798026, + "grad_norm": 0.9081701636314392, + "learning_rate": 1.4897747353654668e-06, + "loss": 0.20968804359436036, + "step": 12250, + "token_acc": 0.9401190554065546 + }, + { + "epoch": 0.7606604183477127, + "grad_norm": 0.844151496887207, + "learning_rate": 1.4861221738905458e-06, + "loss": 0.20760984420776368, + "step": 12255, + "token_acc": 0.9405472207088089 + }, + { + "epoch": 0.7609707653156229, + "grad_norm": 0.7912380695343018, + "learning_rate": 1.4824733137618041e-06, + "loss": 0.2075021505355835, + "step": 12260, + "token_acc": 0.940563705290918 + }, + { + "epoch": 0.761281112283533, + "grad_norm": 0.792633593082428, + "learning_rate": 1.4788281588227688e-06, + "loss": 0.2090841770172119, + "step": 12265, + "token_acc": 0.9398213167177661 + }, + { + "epoch": 0.7615914592514431, + "grad_norm": 0.8588035702705383, + "learning_rate": 1.4751867129130686e-06, + "loss": 0.21068289279937744, + "step": 12270, + "token_acc": 0.9401406852734286 + }, + { + "epoch": 0.7619018062193532, + "grad_norm": 0.928759753704071, + "learning_rate": 1.471548979868425e-06, + "loss": 0.20876905918121338, + "step": 12275, + "token_acc": 0.9395589130262217 + }, + { + "epoch": 0.7622121531872633, + "grad_norm": 0.8575364947319031, + "learning_rate": 1.4679149635206459e-06, + "loss": 0.2107250452041626, + "step": 12280, + "token_acc": 0.9399467882934246 + }, + { + "epoch": 0.7625225001551735, + "grad_norm": 0.8397889733314514, + "learning_rate": 1.4642846676976226e-06, + "loss": 0.2107936382293701, + "step": 12285, + "token_acc": 0.939990103908956 + }, + { + "epoch": 0.7628328471230836, + "grad_norm": 0.7814608812332153, + "learning_rate": 1.4606580962233351e-06, + "loss": 0.2098937749862671, + "step": 12290, + "token_acc": 0.9401186672752168 + }, + { + "epoch": 0.7631431940909937, + "grad_norm": 0.8037402033805847, + "learning_rate": 1.4570352529178345e-06, + "loss": 0.2053915023803711, + "step": 12295, + "token_acc": 0.9409297953871737 + }, + { + "epoch": 0.7634535410589038, + "grad_norm": 0.8236905336380005, + "learning_rate": 1.4534161415972424e-06, + "loss": 0.21368260383605958, + "step": 12300, + "token_acc": 0.9386119698035106 + }, + { + "epoch": 0.7637638880268139, + "grad_norm": 0.8108689188957214, + "learning_rate": 1.4498007660737557e-06, + "loss": 0.20368661880493164, + "step": 12305, + "token_acc": 0.9414992657085912 + }, + { + "epoch": 0.7640742349947242, + "grad_norm": 0.7444151043891907, + "learning_rate": 1.446189130155634e-06, + "loss": 0.21033329963684083, + "step": 12310, + "token_acc": 0.939583508542668 + }, + { + "epoch": 0.7643845819626343, + "grad_norm": 0.888670802116394, + "learning_rate": 1.4425812376471943e-06, + "loss": 0.2009143352508545, + "step": 12315, + "token_acc": 0.9423325979291539 + }, + { + "epoch": 0.7646949289305444, + "grad_norm": 0.7647743225097656, + "learning_rate": 1.438977092348815e-06, + "loss": 0.20582075119018556, + "step": 12320, + "token_acc": 0.941240978966792 + }, + { + "epoch": 0.7650052758984545, + "grad_norm": 0.8493585586547852, + "learning_rate": 1.435376698056925e-06, + "loss": 0.20805282592773439, + "step": 12325, + "token_acc": 0.9400763333705584 + }, + { + "epoch": 0.7653156228663646, + "grad_norm": 0.8177732229232788, + "learning_rate": 1.4317800585640046e-06, + "loss": 0.21305742263793945, + "step": 12330, + "token_acc": 0.9391160901797367 + }, + { + "epoch": 0.7656259698342747, + "grad_norm": 0.8321362733840942, + "learning_rate": 1.4281871776585754e-06, + "loss": 0.2088472366333008, + "step": 12335, + "token_acc": 0.9395910357957388 + }, + { + "epoch": 0.7659363168021849, + "grad_norm": 0.8120673894882202, + "learning_rate": 1.4245980591251994e-06, + "loss": 0.20673832893371583, + "step": 12340, + "token_acc": 0.9407923153789124 + }, + { + "epoch": 0.766246663770095, + "grad_norm": 0.9841536283493042, + "learning_rate": 1.4210127067444807e-06, + "loss": 0.21333045959472657, + "step": 12345, + "token_acc": 0.9387612043954332 + }, + { + "epoch": 0.7665570107380051, + "grad_norm": 0.7736989259719849, + "learning_rate": 1.4174311242930533e-06, + "loss": 0.20782470703125, + "step": 12350, + "token_acc": 0.9405385142384406 + }, + { + "epoch": 0.7668673577059152, + "grad_norm": 0.7881161570549011, + "learning_rate": 1.413853315543578e-06, + "loss": 0.20592947006225587, + "step": 12355, + "token_acc": 0.9407234828461604 + }, + { + "epoch": 0.7671777046738253, + "grad_norm": 0.872920036315918, + "learning_rate": 1.4102792842647438e-06, + "loss": 0.20635876655578614, + "step": 12360, + "token_acc": 0.9416723928326022 + }, + { + "epoch": 0.7674880516417355, + "grad_norm": 0.8226979970932007, + "learning_rate": 1.4067090342212609e-06, + "loss": 0.2047192096710205, + "step": 12365, + "token_acc": 0.9405867811605516 + }, + { + "epoch": 0.7677983986096456, + "grad_norm": 0.8068609237670898, + "learning_rate": 1.4031425691738526e-06, + "loss": 0.20343990325927735, + "step": 12370, + "token_acc": 0.9414399937032252 + }, + { + "epoch": 0.7681087455775557, + "grad_norm": 0.8070027828216553, + "learning_rate": 1.3995798928792602e-06, + "loss": 0.2080981731414795, + "step": 12375, + "token_acc": 0.9402286329358518 + }, + { + "epoch": 0.7684190925454658, + "grad_norm": 0.8179841041564941, + "learning_rate": 1.39602100909023e-06, + "loss": 0.20792102813720703, + "step": 12380, + "token_acc": 0.9407543577454617 + }, + { + "epoch": 0.7687294395133759, + "grad_norm": 1.0886021852493286, + "learning_rate": 1.3924659215555153e-06, + "loss": 0.212081241607666, + "step": 12385, + "token_acc": 0.9393961015794765 + }, + { + "epoch": 0.7690397864812861, + "grad_norm": 0.89322429895401, + "learning_rate": 1.3889146340198734e-06, + "loss": 0.20233550071716308, + "step": 12390, + "token_acc": 0.9421378207648554 + }, + { + "epoch": 0.7693501334491962, + "grad_norm": 0.7917667031288147, + "learning_rate": 1.3853671502240518e-06, + "loss": 0.20737571716308595, + "step": 12395, + "token_acc": 0.940153891836761 + }, + { + "epoch": 0.7696604804171063, + "grad_norm": 0.855232298374176, + "learning_rate": 1.381823473904798e-06, + "loss": 0.20254647731781006, + "step": 12400, + "token_acc": 0.9417719547569451 + }, + { + "epoch": 0.7699708273850164, + "grad_norm": 1.1148453950881958, + "learning_rate": 1.3782836087948475e-06, + "loss": 0.21015806198120118, + "step": 12405, + "token_acc": 0.9396967988229727 + }, + { + "epoch": 0.7702811743529265, + "grad_norm": 0.8174946904182434, + "learning_rate": 1.3747475586229187e-06, + "loss": 0.2095500946044922, + "step": 12410, + "token_acc": 0.9400494600176267 + }, + { + "epoch": 0.7705915213208367, + "grad_norm": 0.8239438533782959, + "learning_rate": 1.3712153271137103e-06, + "loss": 0.20561254024505615, + "step": 12415, + "token_acc": 0.9405696560955519 + }, + { + "epoch": 0.7709018682887468, + "grad_norm": 0.8168414235115051, + "learning_rate": 1.3676869179879065e-06, + "loss": 0.20909924507141114, + "step": 12420, + "token_acc": 0.9403021235078804 + }, + { + "epoch": 0.771212215256657, + "grad_norm": 0.8118036389350891, + "learning_rate": 1.3641623349621587e-06, + "loss": 0.21978528499603273, + "step": 12425, + "token_acc": 0.9380707255980908 + }, + { + "epoch": 0.771522562224567, + "grad_norm": 0.7962722778320312, + "learning_rate": 1.3606415817490864e-06, + "loss": 0.20792264938354493, + "step": 12430, + "token_acc": 0.9403855361628723 + }, + { + "epoch": 0.7718329091924772, + "grad_norm": 0.8503649234771729, + "learning_rate": 1.3571246620572797e-06, + "loss": 0.20440950393676757, + "step": 12435, + "token_acc": 0.9414884437187198 + }, + { + "epoch": 0.7721432561603873, + "grad_norm": 0.8796798586845398, + "learning_rate": 1.3536115795912913e-06, + "loss": 0.21168341636657714, + "step": 12440, + "token_acc": 0.9396503290705763 + }, + { + "epoch": 0.7724536031282975, + "grad_norm": 0.8242261409759521, + "learning_rate": 1.350102338051626e-06, + "loss": 0.21092805862426758, + "step": 12445, + "token_acc": 0.9397862667224081 + }, + { + "epoch": 0.7727639500962076, + "grad_norm": 0.8203704357147217, + "learning_rate": 1.3465969411347475e-06, + "loss": 0.20827798843383788, + "step": 12450, + "token_acc": 0.9404269721081168 + }, + { + "epoch": 0.7730742970641177, + "grad_norm": 0.793378472328186, + "learning_rate": 1.3430953925330693e-06, + "loss": 0.20530974864959717, + "step": 12455, + "token_acc": 0.9408781752341756 + }, + { + "epoch": 0.7733846440320278, + "grad_norm": 0.7629613280296326, + "learning_rate": 1.3395976959349516e-06, + "loss": 0.2169090747833252, + "step": 12460, + "token_acc": 0.9381786925797372 + }, + { + "epoch": 0.7736949909999379, + "grad_norm": 0.9017561078071594, + "learning_rate": 1.336103855024694e-06, + "loss": 0.20860724449157714, + "step": 12465, + "token_acc": 0.9392666281387784 + }, + { + "epoch": 0.7740053379678481, + "grad_norm": 0.8357691168785095, + "learning_rate": 1.3326138734825356e-06, + "loss": 0.21290783882141112, + "step": 12470, + "token_acc": 0.938938093648331 + }, + { + "epoch": 0.7743156849357582, + "grad_norm": 0.8243005871772766, + "learning_rate": 1.3291277549846555e-06, + "loss": 0.21257035732269286, + "step": 12475, + "token_acc": 0.9389084985066521 + }, + { + "epoch": 0.7746260319036683, + "grad_norm": 0.7856304049491882, + "learning_rate": 1.325645503203158e-06, + "loss": 0.21476407051086427, + "step": 12480, + "token_acc": 0.9390006532109607 + }, + { + "epoch": 0.7749363788715784, + "grad_norm": 0.8046197891235352, + "learning_rate": 1.3221671218060733e-06, + "loss": 0.20982093811035157, + "step": 12485, + "token_acc": 0.9399306105218722 + }, + { + "epoch": 0.7752467258394885, + "grad_norm": 0.8144133687019348, + "learning_rate": 1.3186926144573591e-06, + "loss": 0.2089989423751831, + "step": 12490, + "token_acc": 0.9401799824117114 + }, + { + "epoch": 0.7755570728073987, + "grad_norm": 0.8653635382652283, + "learning_rate": 1.3152219848168934e-06, + "loss": 0.21198296546936035, + "step": 12495, + "token_acc": 0.9389469384528217 + }, + { + "epoch": 0.7758674197753088, + "grad_norm": 0.8275713920593262, + "learning_rate": 1.3117552365404624e-06, + "loss": 0.21275269985198975, + "step": 12500, + "token_acc": 0.9396505420907795 + }, + { + "epoch": 0.7758674197753088, + "eval_loss": 0.23086296021938324, + "eval_runtime": 126.1682, + "eval_samples_per_second": 158.519, + "eval_steps_per_second": 0.214, + "eval_token_acc": 0.9344005932038235, + "step": 12500 + }, + { + "epoch": 0.7761777667432189, + "grad_norm": 0.8255384564399719, + "learning_rate": 1.3082923732797703e-06, + "loss": 0.20739402770996093, + "step": 12505, + "token_acc": 0.9406293304491837 + }, + { + "epoch": 0.776488113711129, + "grad_norm": 0.80074143409729, + "learning_rate": 1.3048333986824286e-06, + "loss": 0.20916342735290527, + "step": 12510, + "token_acc": 0.9406059219072876 + }, + { + "epoch": 0.7767984606790391, + "grad_norm": 0.7719592452049255, + "learning_rate": 1.301378316391948e-06, + "loss": 0.20980572700500488, + "step": 12515, + "token_acc": 0.9396963862365253 + }, + { + "epoch": 0.7771088076469493, + "grad_norm": 0.8368150591850281, + "learning_rate": 1.2979271300477454e-06, + "loss": 0.21203787326812745, + "step": 12520, + "token_acc": 0.9398304692204363 + }, + { + "epoch": 0.7774191546148594, + "grad_norm": 0.8214048743247986, + "learning_rate": 1.294479843285128e-06, + "loss": 0.2123859405517578, + "step": 12525, + "token_acc": 0.9389058893004778 + }, + { + "epoch": 0.7777295015827695, + "grad_norm": 0.8253695964813232, + "learning_rate": 1.2910364597352998e-06, + "loss": 0.20806477069854737, + "step": 12530, + "token_acc": 0.9407399505277265 + }, + { + "epoch": 0.7780398485506796, + "grad_norm": 0.7995487451553345, + "learning_rate": 1.2875969830253527e-06, + "loss": 0.20676717758178711, + "step": 12535, + "token_acc": 0.9404137578177069 + }, + { + "epoch": 0.7783501955185897, + "grad_norm": 0.8766912221908569, + "learning_rate": 1.2841614167782602e-06, + "loss": 0.2085273265838623, + "step": 12540, + "token_acc": 0.9412311883326135 + }, + { + "epoch": 0.7786605424864999, + "grad_norm": 0.8271338939666748, + "learning_rate": 1.2807297646128796e-06, + "loss": 0.21137304306030275, + "step": 12545, + "token_acc": 0.9394346788559137 + }, + { + "epoch": 0.7789708894544101, + "grad_norm": 0.8839739561080933, + "learning_rate": 1.277302030143947e-06, + "loss": 0.21183102130889891, + "step": 12550, + "token_acc": 0.9387301866735278 + }, + { + "epoch": 0.7792812364223202, + "grad_norm": 0.8562120199203491, + "learning_rate": 1.2738782169820673e-06, + "loss": 0.22059779167175292, + "step": 12555, + "token_acc": 0.9364110230595208 + }, + { + "epoch": 0.7795915833902303, + "grad_norm": 0.8185442686080933, + "learning_rate": 1.270458328733714e-06, + "loss": 0.22408900260925294, + "step": 12560, + "token_acc": 0.9361121014994854 + }, + { + "epoch": 0.7799019303581404, + "grad_norm": 0.8719221949577332, + "learning_rate": 1.2670423690012351e-06, + "loss": 0.23155388832092286, + "step": 12565, + "token_acc": 0.9341708277888302 + }, + { + "epoch": 0.7802122773260505, + "grad_norm": 0.7904311418533325, + "learning_rate": 1.2636303413828321e-06, + "loss": 0.22426278591156007, + "step": 12570, + "token_acc": 0.9356533417601517 + }, + { + "epoch": 0.7805226242939607, + "grad_norm": 0.772442102432251, + "learning_rate": 1.2602222494725658e-06, + "loss": 0.22647862434387206, + "step": 12575, + "token_acc": 0.9348761429758936 + }, + { + "epoch": 0.7808329712618708, + "grad_norm": 0.883744478225708, + "learning_rate": 1.2568180968603538e-06, + "loss": 0.2284019947052002, + "step": 12580, + "token_acc": 0.9342206853771686 + }, + { + "epoch": 0.7811433182297809, + "grad_norm": 0.8761411309242249, + "learning_rate": 1.2534178871319647e-06, + "loss": 0.22209057807922364, + "step": 12585, + "token_acc": 0.9360037844716731 + }, + { + "epoch": 0.781453665197691, + "grad_norm": 0.84617680311203, + "learning_rate": 1.2500216238690093e-06, + "loss": 0.22072720527648926, + "step": 12590, + "token_acc": 0.9364273998906565 + }, + { + "epoch": 0.7817640121656011, + "grad_norm": 0.8088759183883667, + "learning_rate": 1.2466293106489463e-06, + "loss": 0.2220153331756592, + "step": 12595, + "token_acc": 0.9361242242613907 + }, + { + "epoch": 0.7820743591335113, + "grad_norm": 0.8824402093887329, + "learning_rate": 1.2432409510450704e-06, + "loss": 0.22883250713348388, + "step": 12600, + "token_acc": 0.9349787507876285 + }, + { + "epoch": 0.7823847061014214, + "grad_norm": 0.8352385759353638, + "learning_rate": 1.2398565486265163e-06, + "loss": 0.22454757690429689, + "step": 12605, + "token_acc": 0.9356594368853564 + }, + { + "epoch": 0.7826950530693315, + "grad_norm": 0.835066020488739, + "learning_rate": 1.2364761069582449e-06, + "loss": 0.22360095977783204, + "step": 12610, + "token_acc": 0.9358114983882299 + }, + { + "epoch": 0.7830054000372416, + "grad_norm": 0.8175336122512817, + "learning_rate": 1.233099629601044e-06, + "loss": 0.22271137237548827, + "step": 12615, + "token_acc": 0.935845173017549 + }, + { + "epoch": 0.7833157470051517, + "grad_norm": 0.8464365005493164, + "learning_rate": 1.2297271201115345e-06, + "loss": 0.21818089485168457, + "step": 12620, + "token_acc": 0.9368458578846827 + }, + { + "epoch": 0.7836260939730619, + "grad_norm": 0.8442290425300598, + "learning_rate": 1.2263585820421504e-06, + "loss": 0.226743745803833, + "step": 12625, + "token_acc": 0.9346090381646883 + }, + { + "epoch": 0.783936440940972, + "grad_norm": 0.8281969428062439, + "learning_rate": 1.2229940189411416e-06, + "loss": 0.2211991310119629, + "step": 12630, + "token_acc": 0.9364523375915562 + }, + { + "epoch": 0.7842467879088821, + "grad_norm": 0.8892753720283508, + "learning_rate": 1.2196334343525757e-06, + "loss": 0.22523655891418456, + "step": 12635, + "token_acc": 0.9347965237130546 + }, + { + "epoch": 0.7845571348767922, + "grad_norm": 0.8055672645568848, + "learning_rate": 1.2162768318163276e-06, + "loss": 0.22533814907073973, + "step": 12640, + "token_acc": 0.9355450236966825 + }, + { + "epoch": 0.7848674818447023, + "grad_norm": 0.8877987861633301, + "learning_rate": 1.2129242148680758e-06, + "loss": 0.224007511138916, + "step": 12645, + "token_acc": 0.9364434438568212 + }, + { + "epoch": 0.7851778288126126, + "grad_norm": 0.9060133695602417, + "learning_rate": 1.2095755870393033e-06, + "loss": 0.22334764003753663, + "step": 12650, + "token_acc": 0.9366740309299056 + }, + { + "epoch": 0.7854881757805227, + "grad_norm": 0.8756243586540222, + "learning_rate": 1.2062309518572913e-06, + "loss": 0.22797393798828125, + "step": 12655, + "token_acc": 0.934730754367496 + }, + { + "epoch": 0.7857985227484328, + "grad_norm": 0.8132915496826172, + "learning_rate": 1.2028903128451114e-06, + "loss": 0.2247840404510498, + "step": 12660, + "token_acc": 0.9354079639776488 + }, + { + "epoch": 0.7861088697163429, + "grad_norm": 1.2830548286437988, + "learning_rate": 1.1995536735216312e-06, + "loss": 0.2256025791168213, + "step": 12665, + "token_acc": 0.9352856782284097 + }, + { + "epoch": 0.786419216684253, + "grad_norm": 0.8789248466491699, + "learning_rate": 1.1962210374014994e-06, + "loss": 0.2234166145324707, + "step": 12670, + "token_acc": 0.9361702127659575 + }, + { + "epoch": 0.7867295636521631, + "grad_norm": 0.8055316805839539, + "learning_rate": 1.192892407995153e-06, + "loss": 0.23173065185546876, + "step": 12675, + "token_acc": 0.9343665570890507 + }, + { + "epoch": 0.7870399106200733, + "grad_norm": 0.8655762076377869, + "learning_rate": 1.189567788808808e-06, + "loss": 0.22760238647460937, + "step": 12680, + "token_acc": 0.9353842227907945 + }, + { + "epoch": 0.7873502575879834, + "grad_norm": 0.8477614521980286, + "learning_rate": 1.1862471833444516e-06, + "loss": 0.22311749458312988, + "step": 12685, + "token_acc": 0.9353235139400315 + }, + { + "epoch": 0.7876606045558935, + "grad_norm": 0.8043866157531738, + "learning_rate": 1.1829305950998483e-06, + "loss": 0.2200601577758789, + "step": 12690, + "token_acc": 0.9368080813444128 + }, + { + "epoch": 0.7879709515238036, + "grad_norm": 0.8576344847679138, + "learning_rate": 1.1796180275685304e-06, + "loss": 0.22478463649749755, + "step": 12695, + "token_acc": 0.9363007361518967 + }, + { + "epoch": 0.7882812984917137, + "grad_norm": 0.8652995824813843, + "learning_rate": 1.1763094842397916e-06, + "loss": 0.22314186096191407, + "step": 12700, + "token_acc": 0.9359935833166232 + }, + { + "epoch": 0.7885916454596239, + "grad_norm": 0.8303631544113159, + "learning_rate": 1.1730049685986882e-06, + "loss": 0.21619052886962892, + "step": 12705, + "token_acc": 0.9371365624061154 + }, + { + "epoch": 0.788901992427534, + "grad_norm": 0.8847962021827698, + "learning_rate": 1.1697044841260363e-06, + "loss": 0.21924796104431152, + "step": 12710, + "token_acc": 0.936785716671231 + }, + { + "epoch": 0.7892123393954441, + "grad_norm": 0.8044585585594177, + "learning_rate": 1.1664080342984047e-06, + "loss": 0.2288738489151001, + "step": 12715, + "token_acc": 0.9339276554576562 + }, + { + "epoch": 0.7895226863633542, + "grad_norm": 0.7931410670280457, + "learning_rate": 1.1631156225881086e-06, + "loss": 0.22234642505645752, + "step": 12720, + "token_acc": 0.9358255945191375 + }, + { + "epoch": 0.7898330333312643, + "grad_norm": 0.8381511569023132, + "learning_rate": 1.1598272524632142e-06, + "loss": 0.22836451530456542, + "step": 12725, + "token_acc": 0.9347097323471505 + }, + { + "epoch": 0.7901433802991745, + "grad_norm": 0.8183347582817078, + "learning_rate": 1.15654292738753e-06, + "loss": 0.22361485958099364, + "step": 12730, + "token_acc": 0.9359785024715522 + }, + { + "epoch": 0.7904537272670846, + "grad_norm": 0.7894147634506226, + "learning_rate": 1.1532626508206024e-06, + "loss": 0.2195366621017456, + "step": 12735, + "token_acc": 0.9363521406635756 + }, + { + "epoch": 0.7907640742349947, + "grad_norm": 0.849263072013855, + "learning_rate": 1.1499864262177134e-06, + "loss": 0.2268310308456421, + "step": 12740, + "token_acc": 0.9353698759227266 + }, + { + "epoch": 0.7910744212029048, + "grad_norm": 0.8078927397727966, + "learning_rate": 1.1467142570298733e-06, + "loss": 0.22077901363372804, + "step": 12745, + "token_acc": 0.9364488440161804 + }, + { + "epoch": 0.7913847681708149, + "grad_norm": 0.8446094393730164, + "learning_rate": 1.1434461467038304e-06, + "loss": 0.22818379402160643, + "step": 12750, + "token_acc": 0.9349555052047417 + }, + { + "epoch": 0.7916951151387251, + "grad_norm": 0.8417384028434753, + "learning_rate": 1.1401820986820483e-06, + "loss": 0.22225720882415773, + "step": 12755, + "token_acc": 0.9363218922400978 + }, + { + "epoch": 0.7920054621066352, + "grad_norm": 0.8799223303794861, + "learning_rate": 1.1369221164027138e-06, + "loss": 0.22455615997314454, + "step": 12760, + "token_acc": 0.9364825264104888 + }, + { + "epoch": 0.7923158090745454, + "grad_norm": 0.8699793219566345, + "learning_rate": 1.1336662032997325e-06, + "loss": 0.21879632472991944, + "step": 12765, + "token_acc": 0.9372418938001226 + }, + { + "epoch": 0.7926261560424555, + "grad_norm": 0.8639557361602783, + "learning_rate": 1.130414362802726e-06, + "loss": 0.22421526908874512, + "step": 12770, + "token_acc": 0.9358479735141774 + }, + { + "epoch": 0.7929365030103656, + "grad_norm": 0.9079527258872986, + "learning_rate": 1.1271665983370189e-06, + "loss": 0.23095808029174805, + "step": 12775, + "token_acc": 0.9340278922554482 + }, + { + "epoch": 0.7932468499782757, + "grad_norm": 0.8535253405570984, + "learning_rate": 1.1239229133236495e-06, + "loss": 0.22228012084960938, + "step": 12780, + "token_acc": 0.9357533928393669 + }, + { + "epoch": 0.7935571969461859, + "grad_norm": 1.0384166240692139, + "learning_rate": 1.1206833111793563e-06, + "loss": 0.22688038349151612, + "step": 12785, + "token_acc": 0.9342392026578074 + }, + { + "epoch": 0.793867543914096, + "grad_norm": 0.821685254573822, + "learning_rate": 1.1174477953165751e-06, + "loss": 0.21769371032714843, + "step": 12790, + "token_acc": 0.9365463743676222 + }, + { + "epoch": 0.7941778908820061, + "grad_norm": 0.8394392132759094, + "learning_rate": 1.1142163691434415e-06, + "loss": 0.2210413932800293, + "step": 12795, + "token_acc": 0.9367459184972299 + }, + { + "epoch": 0.7944882378499162, + "grad_norm": 0.8492164015769958, + "learning_rate": 1.1109890360637787e-06, + "loss": 0.21926603317260743, + "step": 12800, + "token_acc": 0.9366479020447044 + }, + { + "epoch": 0.7947985848178263, + "grad_norm": 0.8305680751800537, + "learning_rate": 1.1077657994771012e-06, + "loss": 0.22715752124786376, + "step": 12805, + "token_acc": 0.9346157767732726 + }, + { + "epoch": 0.7951089317857365, + "grad_norm": 0.7976205348968506, + "learning_rate": 1.1045466627786105e-06, + "loss": 0.22399454116821288, + "step": 12810, + "token_acc": 0.9359469074089846 + }, + { + "epoch": 0.7954192787536466, + "grad_norm": 0.8076844811439514, + "learning_rate": 1.101331629359183e-06, + "loss": 0.2247994661331177, + "step": 12815, + "token_acc": 0.9357668383164658 + }, + { + "epoch": 0.7957296257215567, + "grad_norm": 0.7968108057975769, + "learning_rate": 1.0981207026053798e-06, + "loss": 0.22162671089172364, + "step": 12820, + "token_acc": 0.9367925210067471 + }, + { + "epoch": 0.7960399726894668, + "grad_norm": 0.8146686553955078, + "learning_rate": 1.094913885899433e-06, + "loss": 0.22734992504119872, + "step": 12825, + "token_acc": 0.9346577671018003 + }, + { + "epoch": 0.7963503196573769, + "grad_norm": 1.310233473777771, + "learning_rate": 1.0917111826192456e-06, + "loss": 0.22343831062316893, + "step": 12830, + "token_acc": 0.935617020147055 + }, + { + "epoch": 0.7966606666252871, + "grad_norm": 0.8626849055290222, + "learning_rate": 1.0885125961383846e-06, + "loss": 0.223510479927063, + "step": 12835, + "token_acc": 0.9351788193742333 + }, + { + "epoch": 0.7969710135931972, + "grad_norm": 0.8281062841415405, + "learning_rate": 1.085318129826089e-06, + "loss": 0.2221607208251953, + "step": 12840, + "token_acc": 0.9361774698376181 + }, + { + "epoch": 0.7972813605611073, + "grad_norm": 0.827797532081604, + "learning_rate": 1.0821277870472501e-06, + "loss": 0.22488794326782227, + "step": 12845, + "token_acc": 0.9353178245082 + }, + { + "epoch": 0.7975917075290174, + "grad_norm": 0.8393007516860962, + "learning_rate": 1.0789415711624167e-06, + "loss": 0.2286287784576416, + "step": 12850, + "token_acc": 0.9351252022343989 + }, + { + "epoch": 0.7979020544969275, + "grad_norm": 0.8443155288696289, + "learning_rate": 1.0757594855277937e-06, + "loss": 0.2271787166595459, + "step": 12855, + "token_acc": 0.9344780426517645 + }, + { + "epoch": 0.7982124014648377, + "grad_norm": 0.8940489888191223, + "learning_rate": 1.0725815334952345e-06, + "loss": 0.22351422309875488, + "step": 12860, + "token_acc": 0.9361468549149023 + }, + { + "epoch": 0.7985227484327478, + "grad_norm": 0.8047122359275818, + "learning_rate": 1.0694077184122354e-06, + "loss": 0.22614707946777343, + "step": 12865, + "token_acc": 0.934023174417988 + }, + { + "epoch": 0.7988330954006579, + "grad_norm": 0.8154724836349487, + "learning_rate": 1.0662380436219383e-06, + "loss": 0.2190253257751465, + "step": 12870, + "token_acc": 0.9368704638211464 + }, + { + "epoch": 0.799143442368568, + "grad_norm": 0.8031294941902161, + "learning_rate": 1.0630725124631225e-06, + "loss": 0.22470743656158448, + "step": 12875, + "token_acc": 0.9354843009577383 + }, + { + "epoch": 0.7994537893364781, + "grad_norm": 0.8470008969306946, + "learning_rate": 1.0599111282702045e-06, + "loss": 0.22684910297393798, + "step": 12880, + "token_acc": 0.9347399509102882 + }, + { + "epoch": 0.7997641363043883, + "grad_norm": 0.7995855212211609, + "learning_rate": 1.0567538943732303e-06, + "loss": 0.22510089874267578, + "step": 12885, + "token_acc": 0.9354354157374124 + }, + { + "epoch": 0.8000744832722985, + "grad_norm": 0.847175657749176, + "learning_rate": 1.0536008140978714e-06, + "loss": 0.22098641395568847, + "step": 12890, + "token_acc": 0.9362101942887823 + }, + { + "epoch": 0.8003848302402086, + "grad_norm": 0.8155221343040466, + "learning_rate": 1.050451890765433e-06, + "loss": 0.22893173694610597, + "step": 12895, + "token_acc": 0.9344546709001514 + }, + { + "epoch": 0.8006951772081187, + "grad_norm": 0.802585244178772, + "learning_rate": 1.0473071276928338e-06, + "loss": 0.2195821762084961, + "step": 12900, + "token_acc": 0.9364194533323577 + }, + { + "epoch": 0.8010055241760288, + "grad_norm": 0.8050913214683533, + "learning_rate": 1.044166528192611e-06, + "loss": 0.22144131660461425, + "step": 12905, + "token_acc": 0.9368141295206055 + }, + { + "epoch": 0.8013158711439389, + "grad_norm": 0.8552621006965637, + "learning_rate": 1.0410300955729196e-06, + "loss": 0.2265634775161743, + "step": 12910, + "token_acc": 0.9348738930311432 + }, + { + "epoch": 0.8016262181118491, + "grad_norm": 0.8247837424278259, + "learning_rate": 1.0378978331375244e-06, + "loss": 0.21928126811981202, + "step": 12915, + "token_acc": 0.9374730630242349 + }, + { + "epoch": 0.8019365650797592, + "grad_norm": 0.912848711013794, + "learning_rate": 1.034769744185794e-06, + "loss": 0.22553062438964844, + "step": 12920, + "token_acc": 0.9359019922081928 + }, + { + "epoch": 0.8022469120476693, + "grad_norm": 0.8314116597175598, + "learning_rate": 1.0316458320127053e-06, + "loss": 0.22594618797302246, + "step": 12925, + "token_acc": 0.9357668818238918 + }, + { + "epoch": 0.8025572590155794, + "grad_norm": 0.8448149561882019, + "learning_rate": 1.0285260999088347e-06, + "loss": 0.22612357139587402, + "step": 12930, + "token_acc": 0.9355605208621632 + }, + { + "epoch": 0.8028676059834895, + "grad_norm": 0.7903830409049988, + "learning_rate": 1.0254105511603518e-06, + "loss": 0.22227208614349364, + "step": 12935, + "token_acc": 0.9356473525937063 + }, + { + "epoch": 0.8031779529513997, + "grad_norm": 0.7971529364585876, + "learning_rate": 1.0222991890490263e-06, + "loss": 0.21559555530548097, + "step": 12940, + "token_acc": 0.9379678081738093 + }, + { + "epoch": 0.8034882999193098, + "grad_norm": 0.8035684823989868, + "learning_rate": 1.0191920168522102e-06, + "loss": 0.21880548000335692, + "step": 12945, + "token_acc": 0.937204114924926 + }, + { + "epoch": 0.8037986468872199, + "grad_norm": 0.8464736938476562, + "learning_rate": 1.0160890378428478e-06, + "loss": 0.22502360343933106, + "step": 12950, + "token_acc": 0.9358651845217711 + }, + { + "epoch": 0.80410899385513, + "grad_norm": 1.0015594959259033, + "learning_rate": 1.012990255289466e-06, + "loss": 0.22450919151306153, + "step": 12955, + "token_acc": 0.9357565062790651 + }, + { + "epoch": 0.8044193408230401, + "grad_norm": 0.8086045980453491, + "learning_rate": 1.009895672456167e-06, + "loss": 0.2207709312438965, + "step": 12960, + "token_acc": 0.9361425190638969 + }, + { + "epoch": 0.8047296877909503, + "grad_norm": 0.8135299682617188, + "learning_rate": 1.0068052926026345e-06, + "loss": 0.23184528350830078, + "step": 12965, + "token_acc": 0.9343287588497802 + }, + { + "epoch": 0.8050400347588604, + "grad_norm": 0.816990852355957, + "learning_rate": 1.0037191189841232e-06, + "loss": 0.22390303611755372, + "step": 12970, + "token_acc": 0.935440396202838 + }, + { + "epoch": 0.8053503817267705, + "grad_norm": 0.8483110070228577, + "learning_rate": 1.0006371548514554e-06, + "loss": 0.23027479648590088, + "step": 12975, + "token_acc": 0.934013040742278 + }, + { + "epoch": 0.8056607286946806, + "grad_norm": 0.8023169040679932, + "learning_rate": 9.975594034510188e-07, + "loss": 0.22619359493255614, + "step": 12980, + "token_acc": 0.9348084073947098 + }, + { + "epoch": 0.8059710756625907, + "grad_norm": 0.8581125140190125, + "learning_rate": 9.944858680247693e-07, + "loss": 0.22907755374908448, + "step": 12985, + "token_acc": 0.9345430677840246 + }, + { + "epoch": 0.806281422630501, + "grad_norm": 0.8042314052581787, + "learning_rate": 9.91416551810217e-07, + "loss": 0.22220296859741212, + "step": 12990, + "token_acc": 0.936184294089727 + }, + { + "epoch": 0.8065917695984111, + "grad_norm": 0.8850866556167603, + "learning_rate": 9.883514580404264e-07, + "loss": 0.23333415985107422, + "step": 12995, + "token_acc": 0.9339679832654938 + }, + { + "epoch": 0.8069021165663212, + "grad_norm": 0.7955305576324463, + "learning_rate": 9.85290589944018e-07, + "loss": 0.2253568649291992, + "step": 13000, + "token_acc": 0.9349920432109771 + }, + { + "epoch": 0.8069021165663212, + "eval_loss": 0.23698461055755615, + "eval_runtime": 153.829, + "eval_samples_per_second": 130.014, + "eval_steps_per_second": 0.176, + "eval_token_acc": 0.9319708217613063, + "step": 13000 + }, + { + "epoch": 0.8072124635342313, + "grad_norm": 0.8070829510688782, + "learning_rate": 9.822339507451612e-07, + "loss": 0.22407312393188478, + "step": 13005, + "token_acc": 0.9358839831032906 + }, + { + "epoch": 0.8075228105021414, + "grad_norm": 0.8428082466125488, + "learning_rate": 9.79181543663567e-07, + "loss": 0.225250768661499, + "step": 13010, + "token_acc": 0.9350381915432542 + }, + { + "epoch": 0.8078331574700515, + "grad_norm": 0.7860952615737915, + "learning_rate": 9.761333719144922e-07, + "loss": 0.22355871200561522, + "step": 13015, + "token_acc": 0.9352677386828 + }, + { + "epoch": 0.8081435044379617, + "grad_norm": 0.9925312399864197, + "learning_rate": 9.730894387087314e-07, + "loss": 0.22820630073547363, + "step": 13020, + "token_acc": 0.9340691888476423 + }, + { + "epoch": 0.8084538514058718, + "grad_norm": 0.7808263301849365, + "learning_rate": 9.700497472526148e-07, + "loss": 0.22205171585083008, + "step": 13025, + "token_acc": 0.9369064863994849 + }, + { + "epoch": 0.8087641983737819, + "grad_norm": 0.8282653093338013, + "learning_rate": 9.670143007480032e-07, + "loss": 0.223274827003479, + "step": 13030, + "token_acc": 0.9355620789977197 + }, + { + "epoch": 0.809074545341692, + "grad_norm": 0.8384020328521729, + "learning_rate": 9.639831023922858e-07, + "loss": 0.22811155319213866, + "step": 13035, + "token_acc": 0.9355086524897845 + }, + { + "epoch": 0.8093848923096021, + "grad_norm": 0.8355771899223328, + "learning_rate": 9.60956155378378e-07, + "loss": 0.21945717334747314, + "step": 13040, + "token_acc": 0.9356338327489109 + }, + { + "epoch": 0.8096952392775123, + "grad_norm": 0.7694079279899597, + "learning_rate": 9.579334628947196e-07, + "loss": 0.21984879970550536, + "step": 13045, + "token_acc": 0.9366582308909648 + }, + { + "epoch": 0.8100055862454224, + "grad_norm": 0.8216609358787537, + "learning_rate": 9.549150281252633e-07, + "loss": 0.22556371688842775, + "step": 13050, + "token_acc": 0.9352886065688144 + }, + { + "epoch": 0.8103159332133325, + "grad_norm": 0.8095284104347229, + "learning_rate": 9.519008542494806e-07, + "loss": 0.22296245098114015, + "step": 13055, + "token_acc": 0.9355804105254117 + }, + { + "epoch": 0.8106262801812426, + "grad_norm": 0.8060946464538574, + "learning_rate": 9.488909444423561e-07, + "loss": 0.22604262828826904, + "step": 13060, + "token_acc": 0.9355944493885757 + }, + { + "epoch": 0.8109366271491527, + "grad_norm": 0.8428642153739929, + "learning_rate": 9.45885301874378e-07, + "loss": 0.22742850780487062, + "step": 13065, + "token_acc": 0.9342539936842523 + }, + { + "epoch": 0.8112469741170629, + "grad_norm": 0.7942268252372742, + "learning_rate": 9.428839297115455e-07, + "loss": 0.21590676307678222, + "step": 13070, + "token_acc": 0.9373856140304678 + }, + { + "epoch": 0.811557321084973, + "grad_norm": 0.871278703212738, + "learning_rate": 9.39886831115353e-07, + "loss": 0.2256011486053467, + "step": 13075, + "token_acc": 0.9359838095491705 + }, + { + "epoch": 0.8118676680528831, + "grad_norm": 0.8474001288414001, + "learning_rate": 9.368940092427986e-07, + "loss": 0.22525634765625, + "step": 13080, + "token_acc": 0.9351766011621074 + }, + { + "epoch": 0.8121780150207932, + "grad_norm": 0.8340663909912109, + "learning_rate": 9.339054672463749e-07, + "loss": 0.22260241508483886, + "step": 13085, + "token_acc": 0.9362283108654446 + }, + { + "epoch": 0.8124883619887033, + "grad_norm": 0.8412326574325562, + "learning_rate": 9.309212082740626e-07, + "loss": 0.21625075340270997, + "step": 13090, + "token_acc": 0.9378053147571332 + }, + { + "epoch": 0.8127987089566135, + "grad_norm": 0.8362616896629333, + "learning_rate": 9.279412354693351e-07, + "loss": 0.2259580135345459, + "step": 13095, + "token_acc": 0.9354080339377064 + }, + { + "epoch": 0.8131090559245236, + "grad_norm": 0.8110621571540833, + "learning_rate": 9.249655519711504e-07, + "loss": 0.22535967826843262, + "step": 13100, + "token_acc": 0.9347857661583152 + }, + { + "epoch": 0.8134194028924338, + "grad_norm": 0.7994263768196106, + "learning_rate": 9.219941609139455e-07, + "loss": 0.22051167488098145, + "step": 13105, + "token_acc": 0.9370036985574954 + }, + { + "epoch": 0.8137297498603439, + "grad_norm": 0.8089731931686401, + "learning_rate": 9.190270654276356e-07, + "loss": 0.2261592149734497, + "step": 13110, + "token_acc": 0.9363463185073577 + }, + { + "epoch": 0.814040096828254, + "grad_norm": 0.8257761001586914, + "learning_rate": 9.160642686376181e-07, + "loss": 0.22188720703125, + "step": 13115, + "token_acc": 0.9361754038905374 + }, + { + "epoch": 0.8143504437961641, + "grad_norm": 0.8402572870254517, + "learning_rate": 9.131057736647552e-07, + "loss": 0.21994290351867676, + "step": 13120, + "token_acc": 0.9379249029032107 + }, + { + "epoch": 0.8146607907640743, + "grad_norm": 0.8228302001953125, + "learning_rate": 9.101515836253783e-07, + "loss": 0.21983132362365723, + "step": 13125, + "token_acc": 0.9366517026744576 + }, + { + "epoch": 0.8149711377319844, + "grad_norm": 0.788607120513916, + "learning_rate": 9.072017016312878e-07, + "loss": 0.22102198600769044, + "step": 13130, + "token_acc": 0.9361422679870761 + }, + { + "epoch": 0.8152814846998945, + "grad_norm": 0.7838534116744995, + "learning_rate": 9.042561307897452e-07, + "loss": 0.21215338706970216, + "step": 13135, + "token_acc": 0.9386705567581376 + }, + { + "epoch": 0.8155918316678046, + "grad_norm": 0.8673654198646545, + "learning_rate": 9.013148742034683e-07, + "loss": 0.22652974128723144, + "step": 13140, + "token_acc": 0.934895441306411 + }, + { + "epoch": 0.8159021786357147, + "grad_norm": 0.8284367918968201, + "learning_rate": 8.983779349706323e-07, + "loss": 0.22522192001342772, + "step": 13145, + "token_acc": 0.9356681454838773 + }, + { + "epoch": 0.8162125256036249, + "grad_norm": 0.8746384382247925, + "learning_rate": 8.954453161848652e-07, + "loss": 0.22789402008056642, + "step": 13150, + "token_acc": 0.9349028083481397 + }, + { + "epoch": 0.816522872571535, + "grad_norm": 0.8089002966880798, + "learning_rate": 8.925170209352451e-07, + "loss": 0.22272417545318604, + "step": 13155, + "token_acc": 0.9354461566915248 + }, + { + "epoch": 0.8168332195394451, + "grad_norm": 0.794342577457428, + "learning_rate": 8.895930523062929e-07, + "loss": 0.2230823516845703, + "step": 13160, + "token_acc": 0.9358332386230688 + }, + { + "epoch": 0.8171435665073552, + "grad_norm": 0.8907174468040466, + "learning_rate": 8.866734133779703e-07, + "loss": 0.22559120655059814, + "step": 13165, + "token_acc": 0.9359468182603037 + }, + { + "epoch": 0.8174539134752653, + "grad_norm": 0.8276944160461426, + "learning_rate": 8.837581072256868e-07, + "loss": 0.22072415351867675, + "step": 13170, + "token_acc": 0.9363692572380978 + }, + { + "epoch": 0.8177642604431755, + "grad_norm": 0.8089633584022522, + "learning_rate": 8.808471369202803e-07, + "loss": 0.22107532024383544, + "step": 13175, + "token_acc": 0.9363081439758756 + }, + { + "epoch": 0.8180746074110856, + "grad_norm": 0.8073899149894714, + "learning_rate": 8.779405055280227e-07, + "loss": 0.21878132820129395, + "step": 13180, + "token_acc": 0.9359309395048873 + }, + { + "epoch": 0.8183849543789957, + "grad_norm": 0.7999584674835205, + "learning_rate": 8.750382161106169e-07, + "loss": 0.21567444801330565, + "step": 13185, + "token_acc": 0.938111420723962 + }, + { + "epoch": 0.8186953013469058, + "grad_norm": 0.9542261362075806, + "learning_rate": 8.721402717251947e-07, + "loss": 0.2199875831604004, + "step": 13190, + "token_acc": 0.936576007961554 + }, + { + "epoch": 0.8190056483148159, + "grad_norm": 0.806124210357666, + "learning_rate": 8.692466754243045e-07, + "loss": 0.22518625259399414, + "step": 13195, + "token_acc": 0.9352699279534165 + }, + { + "epoch": 0.8193159952827261, + "grad_norm": 0.7953113317489624, + "learning_rate": 8.663574302559202e-07, + "loss": 0.22359278202056884, + "step": 13200, + "token_acc": 0.9356582860232495 + }, + { + "epoch": 0.8196263422506362, + "grad_norm": 0.7611054182052612, + "learning_rate": 8.634725392634314e-07, + "loss": 0.21600065231323243, + "step": 13205, + "token_acc": 0.937046004842615 + }, + { + "epoch": 0.8199366892185463, + "grad_norm": 0.8153495788574219, + "learning_rate": 8.605920054856387e-07, + "loss": 0.22160396575927735, + "step": 13210, + "token_acc": 0.9364637245993178 + }, + { + "epoch": 0.8202470361864564, + "grad_norm": 1.145829677581787, + "learning_rate": 8.577158319567569e-07, + "loss": 0.2203451633453369, + "step": 13215, + "token_acc": 0.9365538040481742 + }, + { + "epoch": 0.8205573831543665, + "grad_norm": 0.8926712870597839, + "learning_rate": 8.548440217064035e-07, + "loss": 0.2226426362991333, + "step": 13220, + "token_acc": 0.9359264845613592 + }, + { + "epoch": 0.8208677301222767, + "grad_norm": 0.8159487843513489, + "learning_rate": 8.519765777596034e-07, + "loss": 0.22569289207458496, + "step": 13225, + "token_acc": 0.9358965002487974 + }, + { + "epoch": 0.8211780770901869, + "grad_norm": 1.3460496664047241, + "learning_rate": 8.491135031367825e-07, + "loss": 0.22686457633972168, + "step": 13230, + "token_acc": 0.9347065114147365 + }, + { + "epoch": 0.821488424058097, + "grad_norm": 0.7944508790969849, + "learning_rate": 8.462548008537597e-07, + "loss": 0.2235942602157593, + "step": 13235, + "token_acc": 0.9356265420756995 + }, + { + "epoch": 0.8217987710260071, + "grad_norm": 0.9820019006729126, + "learning_rate": 8.434004739217533e-07, + "loss": 0.22306313514709472, + "step": 13240, + "token_acc": 0.9360360003695053 + }, + { + "epoch": 0.8221091179939172, + "grad_norm": 0.8292422890663147, + "learning_rate": 8.405505253473723e-07, + "loss": 0.221091365814209, + "step": 13245, + "token_acc": 0.9365873068505869 + }, + { + "epoch": 0.8224194649618273, + "grad_norm": 0.8238874077796936, + "learning_rate": 8.377049581326113e-07, + "loss": 0.22572808265686034, + "step": 13250, + "token_acc": 0.9355928363936785 + }, + { + "epoch": 0.8227298119297375, + "grad_norm": 0.8497426509857178, + "learning_rate": 8.348637752748478e-07, + "loss": 0.2240577220916748, + "step": 13255, + "token_acc": 0.9360394506799771 + }, + { + "epoch": 0.8230401588976476, + "grad_norm": 0.8874006867408752, + "learning_rate": 8.320269797668495e-07, + "loss": 0.22554175853729247, + "step": 13260, + "token_acc": 0.9354543450229494 + }, + { + "epoch": 0.8233505058655577, + "grad_norm": 0.8152508735656738, + "learning_rate": 8.29194574596755e-07, + "loss": 0.22915589809417725, + "step": 13265, + "token_acc": 0.9347547462017035 + }, + { + "epoch": 0.8236608528334678, + "grad_norm": 0.7953078150749207, + "learning_rate": 8.263665627480794e-07, + "loss": 0.22665848731994628, + "step": 13270, + "token_acc": 0.9352198912808839 + }, + { + "epoch": 0.8239711998013779, + "grad_norm": 0.8562704920768738, + "learning_rate": 8.235429471997126e-07, + "loss": 0.21496641635894775, + "step": 13275, + "token_acc": 0.9380150222416684 + }, + { + "epoch": 0.8242815467692881, + "grad_norm": 0.8309566974639893, + "learning_rate": 8.20723730925913e-07, + "loss": 0.22288739681243896, + "step": 13280, + "token_acc": 0.9356210062040838 + }, + { + "epoch": 0.8245918937371982, + "grad_norm": 0.8409950137138367, + "learning_rate": 8.179089168963028e-07, + "loss": 0.2203953742980957, + "step": 13285, + "token_acc": 0.9380385309973944 + }, + { + "epoch": 0.8249022407051083, + "grad_norm": 0.8006001710891724, + "learning_rate": 8.1509850807587e-07, + "loss": 0.2180492401123047, + "step": 13290, + "token_acc": 0.9374203322188186 + }, + { + "epoch": 0.8252125876730184, + "grad_norm": 0.8422213196754456, + "learning_rate": 8.122925074249611e-07, + "loss": 0.22173280715942384, + "step": 13295, + "token_acc": 0.9362570665091109 + }, + { + "epoch": 0.8255229346409285, + "grad_norm": 0.8665030598640442, + "learning_rate": 8.094909178992799e-07, + "loss": 0.2230001926422119, + "step": 13300, + "token_acc": 0.9357678211774029 + }, + { + "epoch": 0.8258332816088387, + "grad_norm": 0.7943859100341797, + "learning_rate": 8.066937424498822e-07, + "loss": 0.21894078254699706, + "step": 13305, + "token_acc": 0.9361511906393365 + }, + { + "epoch": 0.8261436285767488, + "grad_norm": 0.8321238160133362, + "learning_rate": 8.039009840231726e-07, + "loss": 0.22183375358581542, + "step": 13310, + "token_acc": 0.9359273653353849 + }, + { + "epoch": 0.8264539755446589, + "grad_norm": 0.7969396710395813, + "learning_rate": 8.011126455609103e-07, + "loss": 0.22131800651550293, + "step": 13315, + "token_acc": 0.9370608860270101 + }, + { + "epoch": 0.826764322512569, + "grad_norm": 0.8205084800720215, + "learning_rate": 7.983287300001924e-07, + "loss": 0.22153158187866212, + "step": 13320, + "token_acc": 0.9356370368397219 + }, + { + "epoch": 0.8270746694804791, + "grad_norm": 0.8142780661582947, + "learning_rate": 7.955492402734561e-07, + "loss": 0.2312385082244873, + "step": 13325, + "token_acc": 0.9334572367329329 + }, + { + "epoch": 0.8273850164483894, + "grad_norm": 0.8653333187103271, + "learning_rate": 7.92774179308482e-07, + "loss": 0.22207512855529785, + "step": 13330, + "token_acc": 0.9367040372234813 + }, + { + "epoch": 0.8276953634162995, + "grad_norm": 0.8302726745605469, + "learning_rate": 7.900035500283826e-07, + "loss": 0.22133378982543944, + "step": 13335, + "token_acc": 0.9365659713897546 + }, + { + "epoch": 0.8280057103842096, + "grad_norm": 0.8355551362037659, + "learning_rate": 7.872373553516011e-07, + "loss": 0.22538514137268068, + "step": 13340, + "token_acc": 0.93590114032671 + }, + { + "epoch": 0.8283160573521197, + "grad_norm": 0.8474147319793701, + "learning_rate": 7.844755981919122e-07, + "loss": 0.2235948085784912, + "step": 13345, + "token_acc": 0.9357101460860529 + }, + { + "epoch": 0.8286264043200298, + "grad_norm": 0.824914813041687, + "learning_rate": 7.817182814584162e-07, + "loss": 0.22174663543701173, + "step": 13350, + "token_acc": 0.9359524040482426 + }, + { + "epoch": 0.8289367512879399, + "grad_norm": 0.8527224659919739, + "learning_rate": 7.789654080555326e-07, + "loss": 0.22278196811676027, + "step": 13355, + "token_acc": 0.9358262380081328 + }, + { + "epoch": 0.8292470982558501, + "grad_norm": 0.8250142335891724, + "learning_rate": 7.76216980883005e-07, + "loss": 0.22018790245056152, + "step": 13360, + "token_acc": 0.9370791688584955 + }, + { + "epoch": 0.8295574452237602, + "grad_norm": 0.8144986629486084, + "learning_rate": 7.734730028358895e-07, + "loss": 0.21473636627197265, + "step": 13365, + "token_acc": 0.93809836152465 + }, + { + "epoch": 0.8298677921916703, + "grad_norm": 0.8287633061408997, + "learning_rate": 7.707334768045588e-07, + "loss": 0.22041540145874022, + "step": 13370, + "token_acc": 0.9359269561976761 + }, + { + "epoch": 0.8301781391595804, + "grad_norm": 0.8692269325256348, + "learning_rate": 7.679984056746959e-07, + "loss": 0.2224153995513916, + "step": 13375, + "token_acc": 0.9359135593555141 + }, + { + "epoch": 0.8304884861274905, + "grad_norm": 0.8593611717224121, + "learning_rate": 7.652677923272883e-07, + "loss": 0.22484610080718995, + "step": 13380, + "token_acc": 0.9357974408580253 + }, + { + "epoch": 0.8307988330954007, + "grad_norm": 0.8387455940246582, + "learning_rate": 7.625416396386309e-07, + "loss": 0.2198361873626709, + "step": 13385, + "token_acc": 0.936971915989378 + }, + { + "epoch": 0.8311091800633108, + "grad_norm": 0.8691999316215515, + "learning_rate": 7.598199504803206e-07, + "loss": 0.21991174221038817, + "step": 13390, + "token_acc": 0.9361344206493669 + }, + { + "epoch": 0.8314195270312209, + "grad_norm": 0.8686413168907166, + "learning_rate": 7.571027277192484e-07, + "loss": 0.2294692039489746, + "step": 13395, + "token_acc": 0.9344842998472346 + }, + { + "epoch": 0.831729873999131, + "grad_norm": 0.8879052400588989, + "learning_rate": 7.543899742176036e-07, + "loss": 0.22949347496032715, + "step": 13400, + "token_acc": 0.9336858784175899 + }, + { + "epoch": 0.8320402209670411, + "grad_norm": 0.8343828320503235, + "learning_rate": 7.516816928328674e-07, + "loss": 0.2190619945526123, + "step": 13405, + "token_acc": 0.9368207661791063 + }, + { + "epoch": 0.8323505679349513, + "grad_norm": 0.823890745639801, + "learning_rate": 7.489778864178126e-07, + "loss": 0.22779510021209717, + "step": 13410, + "token_acc": 0.934825480873136 + }, + { + "epoch": 0.8326609149028614, + "grad_norm": 0.799058735370636, + "learning_rate": 7.462785578204923e-07, + "loss": 0.22777600288391114, + "step": 13415, + "token_acc": 0.9348459467739098 + }, + { + "epoch": 0.8329712618707715, + "grad_norm": 0.8648195862770081, + "learning_rate": 7.435837098842491e-07, + "loss": 0.22072944641113282, + "step": 13420, + "token_acc": 0.9365720090688303 + }, + { + "epoch": 0.8332816088386816, + "grad_norm": 0.8185339570045471, + "learning_rate": 7.408933454477029e-07, + "loss": 0.2284613609313965, + "step": 13425, + "token_acc": 0.9352369323251772 + }, + { + "epoch": 0.8335919558065917, + "grad_norm": 0.8096317648887634, + "learning_rate": 7.382074673447509e-07, + "loss": 0.2231419563293457, + "step": 13430, + "token_acc": 0.9350769333023163 + }, + { + "epoch": 0.833902302774502, + "grad_norm": 0.7852708101272583, + "learning_rate": 7.35526078404566e-07, + "loss": 0.22208774089813232, + "step": 13435, + "token_acc": 0.9364715262256728 + }, + { + "epoch": 0.834212649742412, + "grad_norm": 0.8249885439872742, + "learning_rate": 7.328491814515887e-07, + "loss": 0.22640738487243653, + "step": 13440, + "token_acc": 0.9356040059798643 + }, + { + "epoch": 0.8345229967103222, + "grad_norm": 0.8176678419113159, + "learning_rate": 7.301767793055348e-07, + "loss": 0.21846914291381836, + "step": 13445, + "token_acc": 0.9378596547314578 + }, + { + "epoch": 0.8348333436782323, + "grad_norm": 0.7965033650398254, + "learning_rate": 7.275088747813791e-07, + "loss": 0.22157640457153321, + "step": 13450, + "token_acc": 0.9359071109875415 + }, + { + "epoch": 0.8351436906461424, + "grad_norm": 0.8601400256156921, + "learning_rate": 7.24845470689361e-07, + "loss": 0.22265450954437255, + "step": 13455, + "token_acc": 0.9369433017939796 + }, + { + "epoch": 0.8354540376140525, + "grad_norm": 0.8442131280899048, + "learning_rate": 7.22186569834979e-07, + "loss": 0.2222574234008789, + "step": 13460, + "token_acc": 0.9365841215286402 + }, + { + "epoch": 0.8357643845819627, + "grad_norm": 0.8081040382385254, + "learning_rate": 7.195321750189905e-07, + "loss": 0.22888822555541993, + "step": 13465, + "token_acc": 0.9349260477410762 + }, + { + "epoch": 0.8360747315498728, + "grad_norm": 0.8395698666572571, + "learning_rate": 7.168822890374027e-07, + "loss": 0.22233219146728517, + "step": 13470, + "token_acc": 0.9357280659007616 + }, + { + "epoch": 0.8363850785177829, + "grad_norm": 0.8187286257743835, + "learning_rate": 7.142369146814754e-07, + "loss": 0.22893452644348145, + "step": 13475, + "token_acc": 0.9357236990987873 + }, + { + "epoch": 0.836695425485693, + "grad_norm": 0.882646918296814, + "learning_rate": 7.115960547377176e-07, + "loss": 0.226092529296875, + "step": 13480, + "token_acc": 0.9359507547456741 + }, + { + "epoch": 0.8370057724536031, + "grad_norm": 0.8039208650588989, + "learning_rate": 7.089597119878794e-07, + "loss": 0.22088911533355712, + "step": 13485, + "token_acc": 0.9365893360507126 + }, + { + "epoch": 0.8373161194215133, + "grad_norm": 0.9555528163909912, + "learning_rate": 7.063278892089559e-07, + "loss": 0.2275542736053467, + "step": 13490, + "token_acc": 0.9357159251122987 + }, + { + "epoch": 0.8376264663894234, + "grad_norm": 0.8214645385742188, + "learning_rate": 7.037005891731791e-07, + "loss": 0.22347598075866698, + "step": 13495, + "token_acc": 0.9362299554195223 + }, + { + "epoch": 0.8379368133573335, + "grad_norm": 0.9231308102607727, + "learning_rate": 7.010778146480174e-07, + "loss": 0.22857069969177246, + "step": 13500, + "token_acc": 0.9344727447114616 + }, + { + "epoch": 0.8379368133573335, + "eval_loss": 0.2366919368505478, + "eval_runtime": 99.1487, + "eval_samples_per_second": 201.717, + "eval_steps_per_second": 0.272, + "eval_token_acc": 0.9322839237656372, + "step": 13500 + }, + { + "epoch": 0.8382471603252436, + "grad_norm": 0.8349510431289673, + "learning_rate": 6.984595683961742e-07, + "loss": 0.22093005180358888, + "step": 13505, + "token_acc": 0.9361443300339686 + }, + { + "epoch": 0.8385575072931537, + "grad_norm": 0.8124993443489075, + "learning_rate": 6.958458531755791e-07, + "loss": 0.22126507759094238, + "step": 13510, + "token_acc": 0.9361781745527928 + }, + { + "epoch": 0.8388678542610639, + "grad_norm": 0.8156249523162842, + "learning_rate": 6.932366717393929e-07, + "loss": 0.22276418209075927, + "step": 13515, + "token_acc": 0.9358912286519648 + }, + { + "epoch": 0.839178201228974, + "grad_norm": 0.9059085249900818, + "learning_rate": 6.906320268359995e-07, + "loss": 0.22083280086517335, + "step": 13520, + "token_acc": 0.9367146797425682 + }, + { + "epoch": 0.8394885481968841, + "grad_norm": 0.8127454519271851, + "learning_rate": 6.88031921209003e-07, + "loss": 0.22016115188598634, + "step": 13525, + "token_acc": 0.9367149188367031 + }, + { + "epoch": 0.8397988951647942, + "grad_norm": 0.8526889681816101, + "learning_rate": 6.85436357597225e-07, + "loss": 0.2210853099822998, + "step": 13530, + "token_acc": 0.9363002439283062 + }, + { + "epoch": 0.8401092421327043, + "grad_norm": 0.8203629851341248, + "learning_rate": 6.828453387347084e-07, + "loss": 0.22853195667266846, + "step": 13535, + "token_acc": 0.9348351458656999 + }, + { + "epoch": 0.8404195891006145, + "grad_norm": 0.8334424495697021, + "learning_rate": 6.802588673507027e-07, + "loss": 0.21643648147583008, + "step": 13540, + "token_acc": 0.9370458610243989 + }, + { + "epoch": 0.8407299360685246, + "grad_norm": 0.8623334169387817, + "learning_rate": 6.776769461696697e-07, + "loss": 0.22779502868652343, + "step": 13545, + "token_acc": 0.9347989949748744 + }, + { + "epoch": 0.8410402830364347, + "grad_norm": 0.7998244166374207, + "learning_rate": 6.750995779112784e-07, + "loss": 0.22149572372436524, + "step": 13550, + "token_acc": 0.9361522115535967 + }, + { + "epoch": 0.8413506300043448, + "grad_norm": 0.795170783996582, + "learning_rate": 6.725267652904033e-07, + "loss": 0.2235417366027832, + "step": 13555, + "token_acc": 0.9351394870854989 + }, + { + "epoch": 0.841660976972255, + "grad_norm": 0.8157673478126526, + "learning_rate": 6.699585110171164e-07, + "loss": 0.2192737102508545, + "step": 13560, + "token_acc": 0.9373549731905243 + }, + { + "epoch": 0.841971323940165, + "grad_norm": 0.8219297528266907, + "learning_rate": 6.673948177966921e-07, + "loss": 0.2271054744720459, + "step": 13565, + "token_acc": 0.934590509263933 + }, + { + "epoch": 0.8422816709080753, + "grad_norm": 0.8130685687065125, + "learning_rate": 6.648356883295981e-07, + "loss": 0.2225255012512207, + "step": 13570, + "token_acc": 0.9358016801800003 + }, + { + "epoch": 0.8425920178759854, + "grad_norm": 0.811294436454773, + "learning_rate": 6.622811253114969e-07, + "loss": 0.22747445106506348, + "step": 13575, + "token_acc": 0.9352662133023719 + }, + { + "epoch": 0.8429023648438955, + "grad_norm": 0.8459377288818359, + "learning_rate": 6.597311314332389e-07, + "loss": 0.2207411050796509, + "step": 13580, + "token_acc": 0.9366051660516606 + }, + { + "epoch": 0.8432127118118056, + "grad_norm": 0.9162362217903137, + "learning_rate": 6.571857093808598e-07, + "loss": 0.22669684886932373, + "step": 13585, + "token_acc": 0.9352700088527492 + }, + { + "epoch": 0.8435230587797157, + "grad_norm": 0.8251834511756897, + "learning_rate": 6.546448618355866e-07, + "loss": 0.22489099502563475, + "step": 13590, + "token_acc": 0.9359348391440377 + }, + { + "epoch": 0.8438334057476259, + "grad_norm": 0.7978650331497192, + "learning_rate": 6.521085914738207e-07, + "loss": 0.21951837539672853, + "step": 13595, + "token_acc": 0.9373352737222606 + }, + { + "epoch": 0.844143752715536, + "grad_norm": 0.8423801064491272, + "learning_rate": 6.495769009671443e-07, + "loss": 0.22534689903259278, + "step": 13600, + "token_acc": 0.935541080226953 + }, + { + "epoch": 0.8444540996834461, + "grad_norm": 0.8170361518859863, + "learning_rate": 6.470497929823155e-07, + "loss": 0.2162562370300293, + "step": 13605, + "token_acc": 0.9375076533736071 + }, + { + "epoch": 0.8447644466513562, + "grad_norm": 0.8180153965950012, + "learning_rate": 6.445272701812683e-07, + "loss": 0.22078292369842528, + "step": 13610, + "token_acc": 0.9373455395817278 + }, + { + "epoch": 0.8450747936192663, + "grad_norm": 0.825749933719635, + "learning_rate": 6.420093352211009e-07, + "loss": 0.22528307437896727, + "step": 13615, + "token_acc": 0.9362145152651312 + }, + { + "epoch": 0.8453851405871765, + "grad_norm": 0.8397501111030579, + "learning_rate": 6.394959907540837e-07, + "loss": 0.2261364459991455, + "step": 13620, + "token_acc": 0.9350494705083521 + }, + { + "epoch": 0.8456954875550866, + "grad_norm": 1.0161067247390747, + "learning_rate": 6.36987239427651e-07, + "loss": 0.2186234712600708, + "step": 13625, + "token_acc": 0.9367113168917021 + }, + { + "epoch": 0.8460058345229967, + "grad_norm": 0.8152735829353333, + "learning_rate": 6.344830838843963e-07, + "loss": 0.2138432025909424, + "step": 13630, + "token_acc": 0.9381762183546405 + }, + { + "epoch": 0.8463161814909068, + "grad_norm": 0.8507735133171082, + "learning_rate": 6.319835267620766e-07, + "loss": 0.2202380657196045, + "step": 13635, + "token_acc": 0.9369348465964837 + }, + { + "epoch": 0.8466265284588169, + "grad_norm": 0.7839285135269165, + "learning_rate": 6.294885706935994e-07, + "loss": 0.22059693336486816, + "step": 13640, + "token_acc": 0.9372607030978072 + }, + { + "epoch": 0.8469368754267271, + "grad_norm": 0.8739960789680481, + "learning_rate": 6.2699821830703e-07, + "loss": 0.2228912115097046, + "step": 13645, + "token_acc": 0.9359732528098426 + }, + { + "epoch": 0.8472472223946372, + "grad_norm": 0.820176362991333, + "learning_rate": 6.245124722255846e-07, + "loss": 0.22721219062805176, + "step": 13650, + "token_acc": 0.9350567969708282 + }, + { + "epoch": 0.8475575693625473, + "grad_norm": 0.8305045962333679, + "learning_rate": 6.220313350676243e-07, + "loss": 0.2217088222503662, + "step": 13655, + "token_acc": 0.9356687940301277 + }, + { + "epoch": 0.8478679163304574, + "grad_norm": 0.8315295577049255, + "learning_rate": 6.195548094466569e-07, + "loss": 0.2183217763900757, + "step": 13660, + "token_acc": 0.9375252207852347 + }, + { + "epoch": 0.8481782632983675, + "grad_norm": 0.8000178933143616, + "learning_rate": 6.17082897971335e-07, + "loss": 0.22687458992004395, + "step": 13665, + "token_acc": 0.9350973605643478 + }, + { + "epoch": 0.8484886102662776, + "grad_norm": 0.8673816919326782, + "learning_rate": 6.146156032454464e-07, + "loss": 0.2211911916732788, + "step": 13670, + "token_acc": 0.9366187295665603 + }, + { + "epoch": 0.8487989572341879, + "grad_norm": 0.8671697378158569, + "learning_rate": 6.121529278679172e-07, + "loss": 0.22640559673309327, + "step": 13675, + "token_acc": 0.9342985442358894 + }, + { + "epoch": 0.849109304202098, + "grad_norm": 0.9152034521102905, + "learning_rate": 6.096948744328118e-07, + "loss": 0.2249903917312622, + "step": 13680, + "token_acc": 0.9355750996108075 + }, + { + "epoch": 0.8494196511700081, + "grad_norm": 0.8704900145530701, + "learning_rate": 6.072414455293212e-07, + "loss": 0.2296988248825073, + "step": 13685, + "token_acc": 0.9356885267797516 + }, + { + "epoch": 0.8497299981379182, + "grad_norm": 0.8247849941253662, + "learning_rate": 6.047926437417656e-07, + "loss": 0.2226785659790039, + "step": 13690, + "token_acc": 0.9357748547773252 + }, + { + "epoch": 0.8500403451058283, + "grad_norm": 0.83273845911026, + "learning_rate": 6.023484716495937e-07, + "loss": 0.22238051891326904, + "step": 13695, + "token_acc": 0.9359591743916957 + }, + { + "epoch": 0.8503506920737385, + "grad_norm": 0.8136412501335144, + "learning_rate": 5.999089318273771e-07, + "loss": 0.22826361656188965, + "step": 13700, + "token_acc": 0.93428001795664 + }, + { + "epoch": 0.8506610390416486, + "grad_norm": 0.7831549048423767, + "learning_rate": 5.974740268448054e-07, + "loss": 0.21985082626342772, + "step": 13705, + "token_acc": 0.9368616218874395 + }, + { + "epoch": 0.8509713860095587, + "grad_norm": 0.7810198068618774, + "learning_rate": 5.950437592666886e-07, + "loss": 0.22025504112243652, + "step": 13710, + "token_acc": 0.9363794747049735 + }, + { + "epoch": 0.8512817329774688, + "grad_norm": 0.824325680732727, + "learning_rate": 5.926181316529506e-07, + "loss": 0.22052376270294188, + "step": 13715, + "token_acc": 0.9367487554808295 + }, + { + "epoch": 0.8515920799453789, + "grad_norm": 0.7987380623817444, + "learning_rate": 5.901971465586304e-07, + "loss": 0.21893658638000488, + "step": 13720, + "token_acc": 0.9367648130844218 + }, + { + "epoch": 0.8519024269132891, + "grad_norm": 0.8551231026649475, + "learning_rate": 5.877808065338725e-07, + "loss": 0.21686863899230957, + "step": 13725, + "token_acc": 0.9374467971908916 + }, + { + "epoch": 0.8522127738811992, + "grad_norm": 0.8411312699317932, + "learning_rate": 5.853691141239309e-07, + "loss": 0.2234273910522461, + "step": 13730, + "token_acc": 0.9358976907530729 + }, + { + "epoch": 0.8525231208491093, + "grad_norm": 0.8116558790206909, + "learning_rate": 5.829620718691636e-07, + "loss": 0.22601895332336425, + "step": 13735, + "token_acc": 0.9346537949400799 + }, + { + "epoch": 0.8528334678170194, + "grad_norm": 0.7941246032714844, + "learning_rate": 5.80559682305033e-07, + "loss": 0.2221287727355957, + "step": 13740, + "token_acc": 0.9356447809770516 + }, + { + "epoch": 0.8531438147849295, + "grad_norm": 0.8664836883544922, + "learning_rate": 5.781619479620948e-07, + "loss": 0.21954770088195802, + "step": 13745, + "token_acc": 0.9360259551082566 + }, + { + "epoch": 0.8534541617528397, + "grad_norm": 0.8385396599769592, + "learning_rate": 5.757688713660059e-07, + "loss": 0.2279510498046875, + "step": 13750, + "token_acc": 0.9353648767813028 + }, + { + "epoch": 0.8537645087207498, + "grad_norm": 0.8263354897499084, + "learning_rate": 5.733804550375171e-07, + "loss": 0.22496931552886962, + "step": 13755, + "token_acc": 0.9346130734232717 + }, + { + "epoch": 0.8540748556886599, + "grad_norm": 0.8265362977981567, + "learning_rate": 5.709967014924667e-07, + "loss": 0.22928013801574706, + "step": 13760, + "token_acc": 0.9342013362676752 + }, + { + "epoch": 0.85438520265657, + "grad_norm": 0.8002598881721497, + "learning_rate": 5.686176132417848e-07, + "loss": 0.222098970413208, + "step": 13765, + "token_acc": 0.936331453338897 + }, + { + "epoch": 0.8546955496244801, + "grad_norm": 0.8581848740577698, + "learning_rate": 5.662431927914847e-07, + "loss": 0.2161628007888794, + "step": 13770, + "token_acc": 0.9362134688691233 + }, + { + "epoch": 0.8550058965923903, + "grad_norm": 0.8337386846542358, + "learning_rate": 5.638734426426651e-07, + "loss": 0.21344079971313476, + "step": 13775, + "token_acc": 0.9375675960853623 + }, + { + "epoch": 0.8553162435603004, + "grad_norm": 0.8187395930290222, + "learning_rate": 5.615083652915049e-07, + "loss": 0.22695527076721192, + "step": 13780, + "token_acc": 0.9352970711297071 + }, + { + "epoch": 0.8556265905282106, + "grad_norm": 1.5552362203598022, + "learning_rate": 5.591479632292585e-07, + "loss": 0.22024729251861572, + "step": 13785, + "token_acc": 0.9363651574473781 + }, + { + "epoch": 0.8559369374961207, + "grad_norm": 0.840758204460144, + "learning_rate": 5.567922389422586e-07, + "loss": 0.22336506843566895, + "step": 13790, + "token_acc": 0.935510395079885 + }, + { + "epoch": 0.8562472844640308, + "grad_norm": 0.7998813986778259, + "learning_rate": 5.544411949119099e-07, + "loss": 0.22725446224212648, + "step": 13795, + "token_acc": 0.9352885362993029 + }, + { + "epoch": 0.8565576314319409, + "grad_norm": 0.7678705453872681, + "learning_rate": 5.520948336146853e-07, + "loss": 0.22968380451202391, + "step": 13800, + "token_acc": 0.9345927051066907 + }, + { + "epoch": 0.8568679783998511, + "grad_norm": 0.8189361095428467, + "learning_rate": 5.497531575221271e-07, + "loss": 0.22214803695678711, + "step": 13805, + "token_acc": 0.9358462851584985 + }, + { + "epoch": 0.8571783253677612, + "grad_norm": 0.9073184728622437, + "learning_rate": 5.474161691008428e-07, + "loss": 0.22661831378936767, + "step": 13810, + "token_acc": 0.9359261125104954 + }, + { + "epoch": 0.8574886723356713, + "grad_norm": 0.794543445110321, + "learning_rate": 5.450838708125006e-07, + "loss": 0.21825926303863524, + "step": 13815, + "token_acc": 0.9371886062129884 + }, + { + "epoch": 0.8577990193035814, + "grad_norm": 0.8126905560493469, + "learning_rate": 5.427562651138268e-07, + "loss": 0.22659952640533448, + "step": 13820, + "token_acc": 0.9356869184455391 + }, + { + "epoch": 0.8581093662714915, + "grad_norm": 0.8336213827133179, + "learning_rate": 5.404333544566087e-07, + "loss": 0.21632072925567628, + "step": 13825, + "token_acc": 0.9373384192981635 + }, + { + "epoch": 0.8584197132394017, + "grad_norm": 0.8209424018859863, + "learning_rate": 5.381151412876873e-07, + "loss": 0.2201441526412964, + "step": 13830, + "token_acc": 0.9363653035026591 + }, + { + "epoch": 0.8587300602073118, + "grad_norm": 0.8339968919754028, + "learning_rate": 5.358016280489514e-07, + "loss": 0.22497999668121338, + "step": 13835, + "token_acc": 0.9354937838614212 + }, + { + "epoch": 0.8590404071752219, + "grad_norm": 0.8450231552124023, + "learning_rate": 5.33492817177344e-07, + "loss": 0.2279956817626953, + "step": 13840, + "token_acc": 0.9351048481798356 + }, + { + "epoch": 0.859350754143132, + "grad_norm": 0.8172098398208618, + "learning_rate": 5.311887111048531e-07, + "loss": 0.21877048015594483, + "step": 13845, + "token_acc": 0.9369775473811907 + }, + { + "epoch": 0.8596611011110421, + "grad_norm": 0.994081437587738, + "learning_rate": 5.28889312258511e-07, + "loss": 0.21872148513793946, + "step": 13850, + "token_acc": 0.9368336604253965 + }, + { + "epoch": 0.8599714480789523, + "grad_norm": 0.8461717367172241, + "learning_rate": 5.265946230603914e-07, + "loss": 0.22129511833190918, + "step": 13855, + "token_acc": 0.9359140925361262 + }, + { + "epoch": 0.8602817950468624, + "grad_norm": 0.8423599600791931, + "learning_rate": 5.243046459276041e-07, + "loss": 0.22115736007690429, + "step": 13860, + "token_acc": 0.9366693043279999 + }, + { + "epoch": 0.8605921420147725, + "grad_norm": 1.3365287780761719, + "learning_rate": 5.22019383272303e-07, + "loss": 0.21376469135284423, + "step": 13865, + "token_acc": 0.9376749055426182 + }, + { + "epoch": 0.8609024889826826, + "grad_norm": 0.8363110423088074, + "learning_rate": 5.197388375016693e-07, + "loss": 0.22298383712768555, + "step": 13870, + "token_acc": 0.936169928617694 + }, + { + "epoch": 0.8612128359505927, + "grad_norm": 0.8174384236335754, + "learning_rate": 5.174630110179169e-07, + "loss": 0.2192277193069458, + "step": 13875, + "token_acc": 0.9364653243847875 + }, + { + "epoch": 0.8615231829185029, + "grad_norm": 0.7856926918029785, + "learning_rate": 5.151919062182909e-07, + "loss": 0.21927084922790527, + "step": 13880, + "token_acc": 0.9374889018960494 + }, + { + "epoch": 0.861833529886413, + "grad_norm": 0.7759402990341187, + "learning_rate": 5.129255254950616e-07, + "loss": 0.21959905624389647, + "step": 13885, + "token_acc": 0.9367281689202691 + }, + { + "epoch": 0.8621438768543231, + "grad_norm": 0.8142533898353577, + "learning_rate": 5.106638712355222e-07, + "loss": 0.22103066444396974, + "step": 13890, + "token_acc": 0.9360447495684486 + }, + { + "epoch": 0.8624542238222332, + "grad_norm": 0.8414297103881836, + "learning_rate": 5.084069458219882e-07, + "loss": 0.22990431785583496, + "step": 13895, + "token_acc": 0.9342858469477776 + }, + { + "epoch": 0.8627645707901433, + "grad_norm": 0.8227231502532959, + "learning_rate": 5.061547516317943e-07, + "loss": 0.22117996215820312, + "step": 13900, + "token_acc": 0.9357309445563448 + }, + { + "epoch": 0.8630749177580535, + "grad_norm": 0.8020862340927124, + "learning_rate": 5.039072910372911e-07, + "loss": 0.2226421356201172, + "step": 13905, + "token_acc": 0.9361263736263736 + }, + { + "epoch": 0.8633852647259637, + "grad_norm": 0.8681520223617554, + "learning_rate": 5.01664566405844e-07, + "loss": 0.2185288429260254, + "step": 13910, + "token_acc": 0.937235686402082 + }, + { + "epoch": 0.8636956116938738, + "grad_norm": 0.8243443369865417, + "learning_rate": 4.99426580099826e-07, + "loss": 0.2176241397857666, + "step": 13915, + "token_acc": 0.9376371940827596 + }, + { + "epoch": 0.8640059586617839, + "grad_norm": 0.7865133285522461, + "learning_rate": 4.971933344766244e-07, + "loss": 0.2170008659362793, + "step": 13920, + "token_acc": 0.9368930757863627 + }, + { + "epoch": 0.864316305629694, + "grad_norm": 0.8112902045249939, + "learning_rate": 4.949648318886308e-07, + "loss": 0.2254420042037964, + "step": 13925, + "token_acc": 0.9346399168096226 + }, + { + "epoch": 0.8646266525976041, + "grad_norm": 0.8493043184280396, + "learning_rate": 4.927410746832384e-07, + "loss": 0.22396745681762695, + "step": 13930, + "token_acc": 0.9353610389263359 + }, + { + "epoch": 0.8649369995655143, + "grad_norm": 0.8382201790809631, + "learning_rate": 4.905220652028442e-07, + "loss": 0.2188735008239746, + "step": 13935, + "token_acc": 0.9376233054899784 + }, + { + "epoch": 0.8652473465334244, + "grad_norm": 1.8398240804672241, + "learning_rate": 4.88307805784845e-07, + "loss": 0.23004064559936524, + "step": 13940, + "token_acc": 0.9343370727547071 + }, + { + "epoch": 0.8655576935013345, + "grad_norm": 0.8349360227584839, + "learning_rate": 4.860982987616325e-07, + "loss": 0.22324986457824708, + "step": 13945, + "token_acc": 0.9359973065395635 + }, + { + "epoch": 0.8658680404692446, + "grad_norm": 0.8304510116577148, + "learning_rate": 4.838935464605899e-07, + "loss": 0.22822217941284179, + "step": 13950, + "token_acc": 0.9347698856078744 + }, + { + "epoch": 0.8661783874371547, + "grad_norm": 0.864220380783081, + "learning_rate": 4.816935512040988e-07, + "loss": 0.2197864055633545, + "step": 13955, + "token_acc": 0.9363275698322141 + }, + { + "epoch": 0.8664887344050649, + "grad_norm": 0.9087817668914795, + "learning_rate": 4.794983153095251e-07, + "loss": 0.22802419662475587, + "step": 13960, + "token_acc": 0.9347578156447068 + }, + { + "epoch": 0.866799081372975, + "grad_norm": 0.8078750371932983, + "learning_rate": 4.773078410892201e-07, + "loss": 0.2186758279800415, + "step": 13965, + "token_acc": 0.9373711305921009 + }, + { + "epoch": 0.8671094283408851, + "grad_norm": 0.8475559949874878, + "learning_rate": 4.751221308505238e-07, + "loss": 0.22482919692993164, + "step": 13970, + "token_acc": 0.9350923341923596 + }, + { + "epoch": 0.8674197753087952, + "grad_norm": 0.8230874538421631, + "learning_rate": 4.729411868957562e-07, + "loss": 0.22475216388702393, + "step": 13975, + "token_acc": 0.9352758491088561 + }, + { + "epoch": 0.8677301222767053, + "grad_norm": 0.8507201075553894, + "learning_rate": 4.707650115222151e-07, + "loss": 0.228471040725708, + "step": 13980, + "token_acc": 0.9347124364523335 + }, + { + "epoch": 0.8680404692446155, + "grad_norm": 0.8171278238296509, + "learning_rate": 4.685936070221769e-07, + "loss": 0.2229231357574463, + "step": 13985, + "token_acc": 0.9358134940498836 + }, + { + "epoch": 0.8683508162125256, + "grad_norm": 1.1370609998703003, + "learning_rate": 4.6642697568289264e-07, + "loss": 0.22608165740966796, + "step": 13990, + "token_acc": 0.9355735959542187 + }, + { + "epoch": 0.8686611631804357, + "grad_norm": 0.8631123304367065, + "learning_rate": 4.642651197865866e-07, + "loss": 0.2222853422164917, + "step": 13995, + "token_acc": 0.9362928411543153 + }, + { + "epoch": 0.8689715101483458, + "grad_norm": 0.846705973148346, + "learning_rate": 4.6210804161045007e-07, + "loss": 0.22054131031036378, + "step": 14000, + "token_acc": 0.9363035642771326 + }, + { + "epoch": 0.8689715101483458, + "eval_loss": 0.23604150116443634, + "eval_runtime": 91.9745, + "eval_samples_per_second": 217.452, + "eval_steps_per_second": 0.294, + "eval_token_acc": 0.9323149949569066, + "step": 14000 + }, + { + "epoch": 0.8692818571162559, + "grad_norm": 0.8087241053581238, + "learning_rate": 4.599557434266416e-07, + "loss": 0.21888971328735352, + "step": 14005, + "token_acc": 0.9369004518908063 + }, + { + "epoch": 0.869592204084166, + "grad_norm": 0.8153994679450989, + "learning_rate": 4.578082275022905e-07, + "loss": 0.22527608871459961, + "step": 14010, + "token_acc": 0.9346741287137198 + }, + { + "epoch": 0.8699025510520763, + "grad_norm": 0.8240395784378052, + "learning_rate": 4.556654960994811e-07, + "loss": 0.22497911453247071, + "step": 14015, + "token_acc": 0.9354173985893036 + }, + { + "epoch": 0.8702128980199864, + "grad_norm": 0.8244780898094177, + "learning_rate": 4.5352755147526184e-07, + "loss": 0.21979053020477296, + "step": 14020, + "token_acc": 0.9373570495694085 + }, + { + "epoch": 0.8705232449878965, + "grad_norm": 0.8219349980354309, + "learning_rate": 4.5139439588163914e-07, + "loss": 0.22550661563873292, + "step": 14025, + "token_acc": 0.9351633215269579 + }, + { + "epoch": 0.8708335919558066, + "grad_norm": 0.8262487649917603, + "learning_rate": 4.49266031565575e-07, + "loss": 0.22323694229125976, + "step": 14030, + "token_acc": 0.9357293590081772 + }, + { + "epoch": 0.8711439389237167, + "grad_norm": 0.8308257460594177, + "learning_rate": 4.471424607689817e-07, + "loss": 0.22305445671081542, + "step": 14035, + "token_acc": 0.9368922924238563 + }, + { + "epoch": 0.8714542858916269, + "grad_norm": 0.811188817024231, + "learning_rate": 4.4502368572872544e-07, + "loss": 0.23076555728912354, + "step": 14040, + "token_acc": 0.9347120522461904 + }, + { + "epoch": 0.871764632859537, + "grad_norm": 0.8034340739250183, + "learning_rate": 4.4290970867662044e-07, + "loss": 0.22234387397766114, + "step": 14045, + "token_acc": 0.9362500578103424 + }, + { + "epoch": 0.8720749798274471, + "grad_norm": 0.809280276298523, + "learning_rate": 4.4080053183942537e-07, + "loss": 0.2181504487991333, + "step": 14050, + "token_acc": 0.9374040662447826 + }, + { + "epoch": 0.8723853267953572, + "grad_norm": 0.8466587662696838, + "learning_rate": 4.3869615743884376e-07, + "loss": 0.22360193729400635, + "step": 14055, + "token_acc": 0.9359935882816726 + }, + { + "epoch": 0.8726956737632673, + "grad_norm": 0.7931615114212036, + "learning_rate": 4.365965876915196e-07, + "loss": 0.21807851791381835, + "step": 14060, + "token_acc": 0.9373381771471008 + }, + { + "epoch": 0.8730060207311775, + "grad_norm": 0.8380593061447144, + "learning_rate": 4.3450182480903603e-07, + "loss": 0.22206435203552247, + "step": 14065, + "token_acc": 0.9363548762697573 + }, + { + "epoch": 0.8733163676990876, + "grad_norm": 0.8540844917297363, + "learning_rate": 4.324118709979158e-07, + "loss": 0.21670713424682617, + "step": 14070, + "token_acc": 0.9371944768103714 + }, + { + "epoch": 0.8736267146669977, + "grad_norm": 0.815854549407959, + "learning_rate": 4.303267284596108e-07, + "loss": 0.22212369441986085, + "step": 14075, + "token_acc": 0.936449861804299 + }, + { + "epoch": 0.8739370616349078, + "grad_norm": 0.7857194542884827, + "learning_rate": 4.282463993905084e-07, + "loss": 0.22327060699462892, + "step": 14080, + "token_acc": 0.9373234285560902 + }, + { + "epoch": 0.8742474086028179, + "grad_norm": 0.7979047894477844, + "learning_rate": 4.261708859819264e-07, + "loss": 0.21641452312469484, + "step": 14085, + "token_acc": 0.9382164747086954 + }, + { + "epoch": 0.8745577555707281, + "grad_norm": 0.8123712539672852, + "learning_rate": 4.2410019042010753e-07, + "loss": 0.21882843971252441, + "step": 14090, + "token_acc": 0.9377377430262046 + }, + { + "epoch": 0.8748681025386382, + "grad_norm": 0.8536764979362488, + "learning_rate": 4.220343148862194e-07, + "loss": 0.23130276203155517, + "step": 14095, + "token_acc": 0.9351280145916369 + }, + { + "epoch": 0.8751784495065483, + "grad_norm": 0.8347164988517761, + "learning_rate": 4.1997326155635564e-07, + "loss": 0.21999011039733887, + "step": 14100, + "token_acc": 0.9362239569735262 + }, + { + "epoch": 0.8754887964744584, + "grad_norm": 0.8113319277763367, + "learning_rate": 4.179170326015286e-07, + "loss": 0.21970810890197753, + "step": 14105, + "token_acc": 0.9363849749649243 + }, + { + "epoch": 0.8757991434423685, + "grad_norm": 0.8426684141159058, + "learning_rate": 4.158656301876668e-07, + "loss": 0.22234387397766114, + "step": 14110, + "token_acc": 0.9363777928232905 + }, + { + "epoch": 0.8761094904102787, + "grad_norm": 0.8175603151321411, + "learning_rate": 4.1381905647561837e-07, + "loss": 0.22120649814605714, + "step": 14115, + "token_acc": 0.936156308789069 + }, + { + "epoch": 0.8764198373781888, + "grad_norm": 0.8196847438812256, + "learning_rate": 4.117773136211439e-07, + "loss": 0.22477927207946777, + "step": 14120, + "token_acc": 0.9356704901771508 + }, + { + "epoch": 0.876730184346099, + "grad_norm": 0.8023508191108704, + "learning_rate": 4.0974040377491364e-07, + "loss": 0.2229152202606201, + "step": 14125, + "token_acc": 0.935361419660429 + }, + { + "epoch": 0.877040531314009, + "grad_norm": 0.8833409547805786, + "learning_rate": 4.0770832908250844e-07, + "loss": 0.2179572582244873, + "step": 14130, + "token_acc": 0.936572594310472 + }, + { + "epoch": 0.8773508782819192, + "grad_norm": 0.8328490853309631, + "learning_rate": 4.056810916844162e-07, + "loss": 0.21863164901733398, + "step": 14135, + "token_acc": 0.9373784128261891 + }, + { + "epoch": 0.8776612252498293, + "grad_norm": 0.8699842691421509, + "learning_rate": 4.036586937160303e-07, + "loss": 0.22139425277709962, + "step": 14140, + "token_acc": 0.9370080826818603 + }, + { + "epoch": 0.8779715722177395, + "grad_norm": 0.8552876710891724, + "learning_rate": 4.01641137307644e-07, + "loss": 0.22113661766052245, + "step": 14145, + "token_acc": 0.9360182006865171 + }, + { + "epoch": 0.8782819191856496, + "grad_norm": 0.8573318123817444, + "learning_rate": 3.996284245844512e-07, + "loss": 0.22212800979614258, + "step": 14150, + "token_acc": 0.9363608115880708 + }, + { + "epoch": 0.8785922661535597, + "grad_norm": 0.8009781241416931, + "learning_rate": 3.9762055766654594e-07, + "loss": 0.22340328693389894, + "step": 14155, + "token_acc": 0.9353219390885044 + }, + { + "epoch": 0.8789026131214698, + "grad_norm": 0.7970234751701355, + "learning_rate": 3.956175386689176e-07, + "loss": 0.22916715145111083, + "step": 14160, + "token_acc": 0.9347073547425041 + }, + { + "epoch": 0.8792129600893799, + "grad_norm": 0.8465226292610168, + "learning_rate": 3.936193697014451e-07, + "loss": 0.22166624069213867, + "step": 14165, + "token_acc": 0.9356337005124514 + }, + { + "epoch": 0.8795233070572901, + "grad_norm": 0.8103132843971252, + "learning_rate": 3.916260528689031e-07, + "loss": 0.22132353782653807, + "step": 14170, + "token_acc": 0.9363822125784573 + }, + { + "epoch": 0.8798336540252002, + "grad_norm": 0.8543841242790222, + "learning_rate": 3.896375902709543e-07, + "loss": 0.22148008346557618, + "step": 14175, + "token_acc": 0.9366451493425751 + }, + { + "epoch": 0.8801440009931103, + "grad_norm": 0.9252007603645325, + "learning_rate": 3.876539840021454e-07, + "loss": 0.22104721069335936, + "step": 14180, + "token_acc": 0.9367299190731523 + }, + { + "epoch": 0.8804543479610204, + "grad_norm": 0.8132889866828918, + "learning_rate": 3.8567523615191173e-07, + "loss": 0.22223396301269532, + "step": 14185, + "token_acc": 0.9355420135610677 + }, + { + "epoch": 0.8807646949289305, + "grad_norm": 0.826564371585846, + "learning_rate": 3.837013488045677e-07, + "loss": 0.22005805969238282, + "step": 14190, + "token_acc": 0.9365901502784552 + }, + { + "epoch": 0.8810750418968407, + "grad_norm": 0.7859030961990356, + "learning_rate": 3.817323240393089e-07, + "loss": 0.2187018632888794, + "step": 14195, + "token_acc": 0.9367512920845631 + }, + { + "epoch": 0.8813853888647508, + "grad_norm": 0.8400143980979919, + "learning_rate": 3.7976816393021145e-07, + "loss": 0.23415207862854004, + "step": 14200, + "token_acc": 0.9346424813882563 + }, + { + "epoch": 0.8816957358326609, + "grad_norm": 0.814201831817627, + "learning_rate": 3.7780887054622296e-07, + "loss": 0.22291197776794433, + "step": 14205, + "token_acc": 0.9360105085381872 + }, + { + "epoch": 0.882006082800571, + "grad_norm": 0.7668090462684631, + "learning_rate": 3.7585444595116707e-07, + "loss": 0.22440056800842284, + "step": 14210, + "token_acc": 0.935492825559646 + }, + { + "epoch": 0.8823164297684811, + "grad_norm": 0.8113758563995361, + "learning_rate": 3.7390489220373974e-07, + "loss": 0.2206037998199463, + "step": 14215, + "token_acc": 0.9361495798877487 + }, + { + "epoch": 0.8826267767363913, + "grad_norm": 0.798896074295044, + "learning_rate": 3.7196021135750384e-07, + "loss": 0.21548147201538087, + "step": 14220, + "token_acc": 0.9376771485106156 + }, + { + "epoch": 0.8829371237043014, + "grad_norm": 0.8520238995552063, + "learning_rate": 3.700204054608897e-07, + "loss": 0.22182600498199462, + "step": 14225, + "token_acc": 0.9358852311686583 + }, + { + "epoch": 0.8832474706722115, + "grad_norm": 0.8248578906059265, + "learning_rate": 3.680854765571967e-07, + "loss": 0.21766290664672852, + "step": 14230, + "token_acc": 0.9370466957312258 + }, + { + "epoch": 0.8835578176401216, + "grad_norm": 0.8016335964202881, + "learning_rate": 3.6615542668458126e-07, + "loss": 0.2297513961791992, + "step": 14235, + "token_acc": 0.9335914430080714 + }, + { + "epoch": 0.8838681646080317, + "grad_norm": 0.8269205689430237, + "learning_rate": 3.642302578760637e-07, + "loss": 0.22135515213012696, + "step": 14240, + "token_acc": 0.9365124811068868 + }, + { + "epoch": 0.8841785115759419, + "grad_norm": 0.7941378355026245, + "learning_rate": 3.6230997215952137e-07, + "loss": 0.2158459186553955, + "step": 14245, + "token_acc": 0.937541377165823 + }, + { + "epoch": 0.8844888585438521, + "grad_norm": 0.8167436122894287, + "learning_rate": 3.603945715576912e-07, + "loss": 0.22228987216949464, + "step": 14250, + "token_acc": 0.935736582418689 + }, + { + "epoch": 0.8847992055117622, + "grad_norm": 0.8525120615959167, + "learning_rate": 3.584840580881593e-07, + "loss": 0.22433433532714844, + "step": 14255, + "token_acc": 0.9354451054336856 + }, + { + "epoch": 0.8851095524796723, + "grad_norm": 0.8139162659645081, + "learning_rate": 3.565784337633682e-07, + "loss": 0.22412440776824952, + "step": 14260, + "token_acc": 0.9353323986137932 + }, + { + "epoch": 0.8854198994475824, + "grad_norm": 0.8411745429039001, + "learning_rate": 3.5467770059060825e-07, + "loss": 0.22113466262817383, + "step": 14265, + "token_acc": 0.9359264721373644 + }, + { + "epoch": 0.8857302464154925, + "grad_norm": 0.8975212574005127, + "learning_rate": 3.527818605720196e-07, + "loss": 0.23154678344726562, + "step": 14270, + "token_acc": 0.9341560493629618 + }, + { + "epoch": 0.8860405933834027, + "grad_norm": 0.8223572969436646, + "learning_rate": 3.508909157045859e-07, + "loss": 0.21809000968933107, + "step": 14275, + "token_acc": 0.9372870169578277 + }, + { + "epoch": 0.8863509403513128, + "grad_norm": 0.825225293636322, + "learning_rate": 3.490048679801344e-07, + "loss": 0.21957144737243653, + "step": 14280, + "token_acc": 0.936695194206715 + }, + { + "epoch": 0.8866612873192229, + "grad_norm": 0.808113157749176, + "learning_rate": 3.471237193853372e-07, + "loss": 0.22571654319763185, + "step": 14285, + "token_acc": 0.9352347672999594 + }, + { + "epoch": 0.886971634287133, + "grad_norm": 0.8071355819702148, + "learning_rate": 3.452474719017035e-07, + "loss": 0.22090961933135986, + "step": 14290, + "token_acc": 0.9364162996468675 + }, + { + "epoch": 0.8872819812550431, + "grad_norm": 0.8311376571655273, + "learning_rate": 3.433761275055786e-07, + "loss": 0.22401165962219238, + "step": 14295, + "token_acc": 0.935951617604197 + }, + { + "epoch": 0.8875923282229533, + "grad_norm": 0.80216383934021, + "learning_rate": 3.415096881681451e-07, + "loss": 0.21582574844360353, + "step": 14300, + "token_acc": 0.9377136067499733 + }, + { + "epoch": 0.8879026751908634, + "grad_norm": 0.923351526260376, + "learning_rate": 3.3964815585542046e-07, + "loss": 0.22542135715484618, + "step": 14305, + "token_acc": 0.9347344318052477 + }, + { + "epoch": 0.8882130221587735, + "grad_norm": 0.8169431686401367, + "learning_rate": 3.377915325282488e-07, + "loss": 0.22918984889984131, + "step": 14310, + "token_acc": 0.9345305751307115 + }, + { + "epoch": 0.8885233691266836, + "grad_norm": 1.1277070045471191, + "learning_rate": 3.359398201423064e-07, + "loss": 0.21828999519348144, + "step": 14315, + "token_acc": 0.9379924835370969 + }, + { + "epoch": 0.8888337160945937, + "grad_norm": 0.8301405310630798, + "learning_rate": 3.34093020648098e-07, + "loss": 0.2268648624420166, + "step": 14320, + "token_acc": 0.9349824781142987 + }, + { + "epoch": 0.8891440630625039, + "grad_norm": 0.8314269185066223, + "learning_rate": 3.3225113599094915e-07, + "loss": 0.22644612789154053, + "step": 14325, + "token_acc": 0.9355607638426505 + }, + { + "epoch": 0.889454410030414, + "grad_norm": 0.8027136325836182, + "learning_rate": 3.304141681110118e-07, + "loss": 0.2237917423248291, + "step": 14330, + "token_acc": 0.9359336066762468 + }, + { + "epoch": 0.8897647569983241, + "grad_norm": 0.876594603061676, + "learning_rate": 3.28582118943257e-07, + "loss": 0.22314050197601318, + "step": 14335, + "token_acc": 0.9364445275758871 + }, + { + "epoch": 0.8900751039662342, + "grad_norm": 0.8254295587539673, + "learning_rate": 3.2675499041747494e-07, + "loss": 0.21964771747589112, + "step": 14340, + "token_acc": 0.9364718629018406 + }, + { + "epoch": 0.8903854509341443, + "grad_norm": 0.8167744874954224, + "learning_rate": 3.24932784458275e-07, + "loss": 0.21919288635253906, + "step": 14345, + "token_acc": 0.9368875797741777 + }, + { + "epoch": 0.8906957979020544, + "grad_norm": 0.8570823073387146, + "learning_rate": 3.2311550298507686e-07, + "loss": 0.21904873847961426, + "step": 14350, + "token_acc": 0.9371303205500097 + }, + { + "epoch": 0.8910061448699647, + "grad_norm": 1.0818060636520386, + "learning_rate": 3.2130314791211645e-07, + "loss": 0.2173046350479126, + "step": 14355, + "token_acc": 0.9373009633705979 + }, + { + "epoch": 0.8913164918378748, + "grad_norm": 0.8769660592079163, + "learning_rate": 3.1949572114844063e-07, + "loss": 0.21947035789489747, + "step": 14360, + "token_acc": 0.937330200937094 + }, + { + "epoch": 0.8916268388057849, + "grad_norm": 0.8245081901550293, + "learning_rate": 3.1769322459790306e-07, + "loss": 0.22816734313964843, + "step": 14365, + "token_acc": 0.9345198097636398 + }, + { + "epoch": 0.891937185773695, + "grad_norm": 0.8592255115509033, + "learning_rate": 3.158956601591634e-07, + "loss": 0.2286104679107666, + "step": 14370, + "token_acc": 0.9342317736887223 + }, + { + "epoch": 0.8922475327416051, + "grad_norm": 0.8381128907203674, + "learning_rate": 3.1410302972569084e-07, + "loss": 0.22489356994628906, + "step": 14375, + "token_acc": 0.9353818315852964 + }, + { + "epoch": 0.8925578797095153, + "grad_norm": 0.846595287322998, + "learning_rate": 3.123153351857522e-07, + "loss": 0.22690954208374023, + "step": 14380, + "token_acc": 0.9342258363083432 + }, + { + "epoch": 0.8928682266774254, + "grad_norm": 0.8364216685295105, + "learning_rate": 3.105325784224161e-07, + "loss": 0.22323029041290282, + "step": 14385, + "token_acc": 0.9355471329336257 + }, + { + "epoch": 0.8931785736453355, + "grad_norm": 0.8729968667030334, + "learning_rate": 3.0875476131355277e-07, + "loss": 0.22561440467834473, + "step": 14390, + "token_acc": 0.9354543479262825 + }, + { + "epoch": 0.8934889206132456, + "grad_norm": 0.7856855988502502, + "learning_rate": 3.0698188573182744e-07, + "loss": 0.2232806921005249, + "step": 14395, + "token_acc": 0.9358584425780657 + }, + { + "epoch": 0.8937992675811557, + "grad_norm": 0.8003990650177002, + "learning_rate": 3.052139535446985e-07, + "loss": 0.21787471771240235, + "step": 14400, + "token_acc": 0.937230749177176 + }, + { + "epoch": 0.8941096145490659, + "grad_norm": 0.8133044838905334, + "learning_rate": 3.0345096661441974e-07, + "loss": 0.21910924911499025, + "step": 14405, + "token_acc": 0.9370250309126759 + }, + { + "epoch": 0.894419961516976, + "grad_norm": 0.8641524314880371, + "learning_rate": 3.0169292679803455e-07, + "loss": 0.22184245586395263, + "step": 14410, + "token_acc": 0.936180496351541 + }, + { + "epoch": 0.8947303084848861, + "grad_norm": 0.8860271573066711, + "learning_rate": 2.999398359473771e-07, + "loss": 0.22291710376739501, + "step": 14415, + "token_acc": 0.9362331025603973 + }, + { + "epoch": 0.8950406554527962, + "grad_norm": 0.8459035754203796, + "learning_rate": 2.9819169590906583e-07, + "loss": 0.22820255756378174, + "step": 14420, + "token_acc": 0.9344099674628997 + }, + { + "epoch": 0.8953510024207063, + "grad_norm": 0.7979179620742798, + "learning_rate": 2.964485085245039e-07, + "loss": 0.21536626815795898, + "step": 14425, + "token_acc": 0.9377832161249983 + }, + { + "epoch": 0.8956613493886165, + "grad_norm": 0.8881822228431702, + "learning_rate": 2.947102756298831e-07, + "loss": 0.22440438270568847, + "step": 14430, + "token_acc": 0.9356905632527089 + }, + { + "epoch": 0.8959716963565266, + "grad_norm": 0.8861150145530701, + "learning_rate": 2.929769990561704e-07, + "loss": 0.21970758438110352, + "step": 14435, + "token_acc": 0.9368858629395972 + }, + { + "epoch": 0.8962820433244367, + "grad_norm": 0.8024405837059021, + "learning_rate": 2.912486806291132e-07, + "loss": 0.2186105728149414, + "step": 14440, + "token_acc": 0.936586614845258 + }, + { + "epoch": 0.8965923902923468, + "grad_norm": 0.8578680753707886, + "learning_rate": 2.895253221692379e-07, + "loss": 0.221679949760437, + "step": 14445, + "token_acc": 0.9358649789029536 + }, + { + "epoch": 0.8969027372602569, + "grad_norm": 0.8500242233276367, + "learning_rate": 2.87806925491847e-07, + "loss": 0.22474842071533202, + "step": 14450, + "token_acc": 0.935863067226618 + }, + { + "epoch": 0.8972130842281671, + "grad_norm": 0.8253446221351624, + "learning_rate": 2.86093492407013e-07, + "loss": 0.22402801513671874, + "step": 14455, + "token_acc": 0.9369696483379807 + }, + { + "epoch": 0.8975234311960772, + "grad_norm": 0.866589367389679, + "learning_rate": 2.8438502471958283e-07, + "loss": 0.22346680164337157, + "step": 14460, + "token_acc": 0.9357250619753693 + }, + { + "epoch": 0.8978337781639874, + "grad_norm": 0.8102363348007202, + "learning_rate": 2.826815242291736e-07, + "loss": 0.2292325973510742, + "step": 14465, + "token_acc": 0.9341550695825049 + }, + { + "epoch": 0.8981441251318975, + "grad_norm": 0.872162401676178, + "learning_rate": 2.8098299273016714e-07, + "loss": 0.2273706912994385, + "step": 14470, + "token_acc": 0.9353147783513384 + }, + { + "epoch": 0.8984544720998076, + "grad_norm": 0.8664935231208801, + "learning_rate": 2.7928943201171453e-07, + "loss": 0.221476411819458, + "step": 14475, + "token_acc": 0.9357775704521243 + }, + { + "epoch": 0.8987648190677177, + "grad_norm": 0.999637246131897, + "learning_rate": 2.776008438577277e-07, + "loss": 0.2201094150543213, + "step": 14480, + "token_acc": 0.9362012079929756 + }, + { + "epoch": 0.8990751660356279, + "grad_norm": 0.8095411062240601, + "learning_rate": 2.7591723004688376e-07, + "loss": 0.22244060039520264, + "step": 14485, + "token_acc": 0.9358760760046015 + }, + { + "epoch": 0.899385513003538, + "grad_norm": 0.8493530750274658, + "learning_rate": 2.742385923526192e-07, + "loss": 0.22095117568969727, + "step": 14490, + "token_acc": 0.9366565244312469 + }, + { + "epoch": 0.8996958599714481, + "grad_norm": 0.8040181398391724, + "learning_rate": 2.725649325431268e-07, + "loss": 0.2164665699005127, + "step": 14495, + "token_acc": 0.9373950682056663 + }, + { + "epoch": 0.9000062069393582, + "grad_norm": 0.8571121096611023, + "learning_rate": 2.708962523813585e-07, + "loss": 0.22633914947509765, + "step": 14500, + "token_acc": 0.9344589318407014 + }, + { + "epoch": 0.9000062069393582, + "eval_loss": 0.23566323518753052, + "eval_runtime": 103.9079, + "eval_samples_per_second": 192.478, + "eval_steps_per_second": 0.26, + "eval_token_acc": 0.9323365057816316, + "step": 14500 + }, + { + "epoch": 0.9003165539072683, + "grad_norm": 0.7956224083900452, + "learning_rate": 2.692325536250201e-07, + "loss": 0.21967792510986328, + "step": 14505, + "token_acc": 0.9369756386603283 + }, + { + "epoch": 0.9006269008751785, + "grad_norm": 0.837196946144104, + "learning_rate": 2.675738380265702e-07, + "loss": 0.22492718696594238, + "step": 14510, + "token_acc": 0.9354782398517544 + }, + { + "epoch": 0.9009372478430886, + "grad_norm": 0.8757438063621521, + "learning_rate": 2.6592010733321737e-07, + "loss": 0.2260443925857544, + "step": 14515, + "token_acc": 0.9349513596068927 + }, + { + "epoch": 0.9012475948109987, + "grad_norm": 0.8186724781990051, + "learning_rate": 2.642713632869209e-07, + "loss": 0.21876153945922852, + "step": 14520, + "token_acc": 0.9368021894681308 + }, + { + "epoch": 0.9015579417789088, + "grad_norm": 0.8073204159736633, + "learning_rate": 2.6262760762438665e-07, + "loss": 0.2218660831451416, + "step": 14525, + "token_acc": 0.9358341822636811 + }, + { + "epoch": 0.9018682887468189, + "grad_norm": 0.9101963639259338, + "learning_rate": 2.6098884207706564e-07, + "loss": 0.2245234251022339, + "step": 14530, + "token_acc": 0.9354048748237548 + }, + { + "epoch": 0.9021786357147291, + "grad_norm": 0.7929571270942688, + "learning_rate": 2.5935506837115376e-07, + "loss": 0.21697502136230468, + "step": 14535, + "token_acc": 0.9370374549485269 + }, + { + "epoch": 0.9024889826826392, + "grad_norm": 0.818593442440033, + "learning_rate": 2.577262882275866e-07, + "loss": 0.21823678016662598, + "step": 14540, + "token_acc": 0.9370296976675683 + }, + { + "epoch": 0.9027993296505493, + "grad_norm": 0.8255177140235901, + "learning_rate": 2.561025033620435e-07, + "loss": 0.2217839241027832, + "step": 14545, + "token_acc": 0.9361010806579977 + }, + { + "epoch": 0.9031096766184594, + "grad_norm": 0.7913464903831482, + "learning_rate": 2.5448371548493723e-07, + "loss": 0.22095584869384766, + "step": 14550, + "token_acc": 0.9356015701151242 + }, + { + "epoch": 0.9034200235863695, + "grad_norm": 0.8055823445320129, + "learning_rate": 2.528699263014195e-07, + "loss": 0.21829023361206054, + "step": 14555, + "token_acc": 0.9367426727542946 + }, + { + "epoch": 0.9037303705542797, + "grad_norm": 0.803391695022583, + "learning_rate": 2.512611375113783e-07, + "loss": 0.21504998207092285, + "step": 14560, + "token_acc": 0.9373572399952468 + }, + { + "epoch": 0.9040407175221898, + "grad_norm": 0.8690827488899231, + "learning_rate": 2.496573508094313e-07, + "loss": 0.22011799812316896, + "step": 14565, + "token_acc": 0.9370651889520257 + }, + { + "epoch": 0.9043510644900999, + "grad_norm": 0.8185088038444519, + "learning_rate": 2.48058567884929e-07, + "loss": 0.22260572910308837, + "step": 14570, + "token_acc": 0.9358899684943475 + }, + { + "epoch": 0.90466141145801, + "grad_norm": 0.8033232092857361, + "learning_rate": 2.464647904219497e-07, + "loss": 0.2223740577697754, + "step": 14575, + "token_acc": 0.9368577392421323 + }, + { + "epoch": 0.9049717584259201, + "grad_norm": 0.8143609762191772, + "learning_rate": 2.4487602009930225e-07, + "loss": 0.22102956771850585, + "step": 14580, + "token_acc": 0.9366127786507578 + }, + { + "epoch": 0.9052821053938302, + "grad_norm": 0.7892504334449768, + "learning_rate": 2.4329225859051785e-07, + "loss": 0.22567241191864013, + "step": 14585, + "token_acc": 0.9351482869731045 + }, + { + "epoch": 0.9055924523617405, + "grad_norm": 0.8111969232559204, + "learning_rate": 2.417135075638527e-07, + "loss": 0.22044172286987304, + "step": 14590, + "token_acc": 0.936224205425467 + }, + { + "epoch": 0.9059027993296506, + "grad_norm": 0.9308971762657166, + "learning_rate": 2.4013976868228703e-07, + "loss": 0.22542757987976075, + "step": 14595, + "token_acc": 0.9347193457950149 + }, + { + "epoch": 0.9062131462975607, + "grad_norm": 0.813959002494812, + "learning_rate": 2.3857104360351825e-07, + "loss": 0.21785888671875, + "step": 14600, + "token_acc": 0.937396684613393 + }, + { + "epoch": 0.9065234932654708, + "grad_norm": 0.8302735686302185, + "learning_rate": 2.370073339799661e-07, + "loss": 0.2217571496963501, + "step": 14605, + "token_acc": 0.9354015445295151 + }, + { + "epoch": 0.9068338402333809, + "grad_norm": 0.784950852394104, + "learning_rate": 2.3544864145876366e-07, + "loss": 0.2144294023513794, + "step": 14610, + "token_acc": 0.937896654456012 + }, + { + "epoch": 0.9071441872012911, + "grad_norm": 0.8525421023368835, + "learning_rate": 2.3389496768176144e-07, + "loss": 0.2239243984222412, + "step": 14615, + "token_acc": 0.9358435435136302 + }, + { + "epoch": 0.9074545341692012, + "grad_norm": 0.8063422441482544, + "learning_rate": 2.3234631428552423e-07, + "loss": 0.22260589599609376, + "step": 14620, + "token_acc": 0.9359237998271831 + }, + { + "epoch": 0.9077648811371113, + "grad_norm": 0.8454898595809937, + "learning_rate": 2.30802682901326e-07, + "loss": 0.22190825939178466, + "step": 14625, + "token_acc": 0.9367863076557926 + }, + { + "epoch": 0.9080752281050214, + "grad_norm": 0.7919603586196899, + "learning_rate": 2.292640751551528e-07, + "loss": 0.22093782424926758, + "step": 14630, + "token_acc": 0.9365748121156912 + }, + { + "epoch": 0.9083855750729315, + "grad_norm": 0.819851279258728, + "learning_rate": 2.2773049266769876e-07, + "loss": 0.2225898027420044, + "step": 14635, + "token_acc": 0.9361507121496953 + }, + { + "epoch": 0.9086959220408417, + "grad_norm": 0.8274653553962708, + "learning_rate": 2.2620193705436388e-07, + "loss": 0.22367024421691895, + "step": 14640, + "token_acc": 0.9350973975796751 + }, + { + "epoch": 0.9090062690087518, + "grad_norm": 0.8254303932189941, + "learning_rate": 2.2467840992525247e-07, + "loss": 0.22579972743988036, + "step": 14645, + "token_acc": 0.9351789690584599 + }, + { + "epoch": 0.9093166159766619, + "grad_norm": 0.8595606088638306, + "learning_rate": 2.231599128851747e-07, + "loss": 0.2166797161102295, + "step": 14650, + "token_acc": 0.9376487007376886 + }, + { + "epoch": 0.909626962944572, + "grad_norm": 0.7773082256317139, + "learning_rate": 2.2164644753364106e-07, + "loss": 0.21489458084106444, + "step": 14655, + "token_acc": 0.9379349307101791 + }, + { + "epoch": 0.9099373099124821, + "grad_norm": 0.7872897982597351, + "learning_rate": 2.2013801546485913e-07, + "loss": 0.22157564163208007, + "step": 14660, + "token_acc": 0.9357654328307082 + }, + { + "epoch": 0.9102476568803923, + "grad_norm": 0.8695574998855591, + "learning_rate": 2.1863461826773903e-07, + "loss": 0.22341136932373046, + "step": 14665, + "token_acc": 0.9353798495156326 + }, + { + "epoch": 0.9105580038483024, + "grad_norm": 0.8586180806159973, + "learning_rate": 2.1713625752588562e-07, + "loss": 0.21971554756164552, + "step": 14670, + "token_acc": 0.9366916986422182 + }, + { + "epoch": 0.9108683508162125, + "grad_norm": 0.8224927186965942, + "learning_rate": 2.1564293481759757e-07, + "loss": 0.22295675277709961, + "step": 14675, + "token_acc": 0.9359615152692725 + }, + { + "epoch": 0.9111786977841226, + "grad_norm": 0.8339778780937195, + "learning_rate": 2.1415465171586712e-07, + "loss": 0.2200387954711914, + "step": 14680, + "token_acc": 0.9372668904518329 + }, + { + "epoch": 0.9114890447520327, + "grad_norm": 0.8161564469337463, + "learning_rate": 2.126714097883803e-07, + "loss": 0.22355222702026367, + "step": 14685, + "token_acc": 0.936147666386293 + }, + { + "epoch": 0.9117993917199428, + "grad_norm": 0.8717086911201477, + "learning_rate": 2.111932105975101e-07, + "loss": 0.21718392372131348, + "step": 14690, + "token_acc": 0.937369337979094 + }, + { + "epoch": 0.9121097386878531, + "grad_norm": 0.8104777932167053, + "learning_rate": 2.0972005570031984e-07, + "loss": 0.22424962520599365, + "step": 14695, + "token_acc": 0.936106329316626 + }, + { + "epoch": 0.9124200856557632, + "grad_norm": 0.9168269038200378, + "learning_rate": 2.0825194664855552e-07, + "loss": 0.2267009735107422, + "step": 14700, + "token_acc": 0.9343678806088421 + }, + { + "epoch": 0.9127304326236733, + "grad_norm": 0.8220688700675964, + "learning_rate": 2.0678888498865512e-07, + "loss": 0.23117473125457763, + "step": 14705, + "token_acc": 0.9337462201666105 + }, + { + "epoch": 0.9130407795915834, + "grad_norm": 0.7989029884338379, + "learning_rate": 2.053308722617342e-07, + "loss": 0.2226240634918213, + "step": 14710, + "token_acc": 0.9363551976152811 + }, + { + "epoch": 0.9133511265594935, + "grad_norm": 0.7994082570075989, + "learning_rate": 2.0387791000359037e-07, + "loss": 0.22126126289367676, + "step": 14715, + "token_acc": 0.9365286253135059 + }, + { + "epoch": 0.9136614735274037, + "grad_norm": 0.8369300365447998, + "learning_rate": 2.0242999974470379e-07, + "loss": 0.22027587890625, + "step": 14720, + "token_acc": 0.9367703316088878 + }, + { + "epoch": 0.9139718204953138, + "grad_norm": 0.8798187971115112, + "learning_rate": 2.009871430102328e-07, + "loss": 0.22298069000244142, + "step": 14725, + "token_acc": 0.9354777262516257 + }, + { + "epoch": 0.9142821674632239, + "grad_norm": 0.836509108543396, + "learning_rate": 1.9954934132001003e-07, + "loss": 0.22409319877624512, + "step": 14730, + "token_acc": 0.9355357893630439 + }, + { + "epoch": 0.914592514431134, + "grad_norm": 0.7939102649688721, + "learning_rate": 1.9811659618854616e-07, + "loss": 0.21672677993774414, + "step": 14735, + "token_acc": 0.9377436783808211 + }, + { + "epoch": 0.9149028613990441, + "grad_norm": 0.8172755241394043, + "learning_rate": 1.9668890912502514e-07, + "loss": 0.2224776029586792, + "step": 14740, + "token_acc": 0.9357937609801354 + }, + { + "epoch": 0.9152132083669543, + "grad_norm": 0.8581714034080505, + "learning_rate": 1.9526628163330063e-07, + "loss": 0.2262043237686157, + "step": 14745, + "token_acc": 0.9348848481672649 + }, + { + "epoch": 0.9155235553348644, + "grad_norm": 0.8044265508651733, + "learning_rate": 1.9384871521190006e-07, + "loss": 0.22104787826538086, + "step": 14750, + "token_acc": 0.9366974424281015 + }, + { + "epoch": 0.9158339023027745, + "grad_norm": 0.851483166217804, + "learning_rate": 1.924362113540168e-07, + "loss": 0.22310147285461426, + "step": 14755, + "token_acc": 0.9360927196253002 + }, + { + "epoch": 0.9161442492706846, + "grad_norm": 1.061468243598938, + "learning_rate": 1.9102877154751341e-07, + "loss": 0.22409377098083497, + "step": 14760, + "token_acc": 0.9354903496615854 + }, + { + "epoch": 0.9164545962385947, + "grad_norm": 0.8129377961158752, + "learning_rate": 1.8962639727491904e-07, + "loss": 0.21659207344055176, + "step": 14765, + "token_acc": 0.9369632085646946 + }, + { + "epoch": 0.9167649432065049, + "grad_norm": 0.8413456678390503, + "learning_rate": 1.8822909001342427e-07, + "loss": 0.22525689601898194, + "step": 14770, + "token_acc": 0.935409082454229 + }, + { + "epoch": 0.917075290174415, + "grad_norm": 0.8332679271697998, + "learning_rate": 1.8683685123488505e-07, + "loss": 0.22126288414001466, + "step": 14775, + "token_acc": 0.935311849424374 + }, + { + "epoch": 0.9173856371423251, + "grad_norm": 0.7876655459403992, + "learning_rate": 1.8544968240581663e-07, + "loss": 0.22011280059814453, + "step": 14780, + "token_acc": 0.9368950496634625 + }, + { + "epoch": 0.9176959841102352, + "grad_norm": 0.8535305261611938, + "learning_rate": 1.8406758498739462e-07, + "loss": 0.22580337524414062, + "step": 14785, + "token_acc": 0.935329933632311 + }, + { + "epoch": 0.9180063310781453, + "grad_norm": 0.8815336227416992, + "learning_rate": 1.826905604354512e-07, + "loss": 0.2286771297454834, + "step": 14790, + "token_acc": 0.9344937549827266 + }, + { + "epoch": 0.9183166780460554, + "grad_norm": 0.8448968529701233, + "learning_rate": 1.8131861020047936e-07, + "loss": 0.22403311729431152, + "step": 14795, + "token_acc": 0.9361213999295115 + }, + { + "epoch": 0.9186270250139656, + "grad_norm": 0.804602324962616, + "learning_rate": 1.7995173572762149e-07, + "loss": 0.22575454711914061, + "step": 14800, + "token_acc": 0.935919145234443 + }, + { + "epoch": 0.9189373719818758, + "grad_norm": 0.8726068139076233, + "learning_rate": 1.7858993845667694e-07, + "loss": 0.22514305114746094, + "step": 14805, + "token_acc": 0.935413774370737 + }, + { + "epoch": 0.9192477189497859, + "grad_norm": 0.9023924469947815, + "learning_rate": 1.7723321982209496e-07, + "loss": 0.22989070415496826, + "step": 14810, + "token_acc": 0.933703591504258 + }, + { + "epoch": 0.919558065917696, + "grad_norm": 0.8202250599861145, + "learning_rate": 1.758815812529785e-07, + "loss": 0.21946988105773926, + "step": 14815, + "token_acc": 0.9369790512267693 + }, + { + "epoch": 0.9198684128856061, + "grad_norm": 0.9518576860427856, + "learning_rate": 1.7453502417307477e-07, + "loss": 0.22705843448638915, + "step": 14820, + "token_acc": 0.9346299797702 + }, + { + "epoch": 0.9201787598535163, + "grad_norm": 0.8006566762924194, + "learning_rate": 1.7319355000078196e-07, + "loss": 0.22328073978424073, + "step": 14825, + "token_acc": 0.9353801303378718 + }, + { + "epoch": 0.9204891068214264, + "grad_norm": 0.8266062140464783, + "learning_rate": 1.718571601491431e-07, + "loss": 0.22194862365722656, + "step": 14830, + "token_acc": 0.9363865484748333 + }, + { + "epoch": 0.9207994537893365, + "grad_norm": 0.8038060069084167, + "learning_rate": 1.7052585602584603e-07, + "loss": 0.21702494621276855, + "step": 14835, + "token_acc": 0.9369817469808271 + }, + { + "epoch": 0.9211098007572466, + "grad_norm": 0.8460286259651184, + "learning_rate": 1.6919963903322012e-07, + "loss": 0.23009960651397704, + "step": 14840, + "token_acc": 0.934268639503034 + }, + { + "epoch": 0.9214201477251567, + "grad_norm": 0.8133052587509155, + "learning_rate": 1.6787851056823678e-07, + "loss": 0.22354378700256347, + "step": 14845, + "token_acc": 0.9358612903012816 + }, + { + "epoch": 0.9217304946930669, + "grad_norm": 0.8138416409492493, + "learning_rate": 1.6656247202250897e-07, + "loss": 0.22208771705627442, + "step": 14850, + "token_acc": 0.935843774904067 + }, + { + "epoch": 0.922040841660977, + "grad_norm": 0.8014063239097595, + "learning_rate": 1.6525152478228667e-07, + "loss": 0.21870801448822022, + "step": 14855, + "token_acc": 0.9365858211442819 + }, + { + "epoch": 0.9223511886288871, + "grad_norm": 0.8212425708770752, + "learning_rate": 1.63945670228457e-07, + "loss": 0.2198044776916504, + "step": 14860, + "token_acc": 0.9370388326946449 + }, + { + "epoch": 0.9226615355967972, + "grad_norm": 0.9671867489814758, + "learning_rate": 1.626449097365429e-07, + "loss": 0.2257516622543335, + "step": 14865, + "token_acc": 0.9355984617742914 + }, + { + "epoch": 0.9229718825647073, + "grad_norm": 0.8803640007972717, + "learning_rate": 1.6134924467670233e-07, + "loss": 0.22582755088806153, + "step": 14870, + "token_acc": 0.9348682257479302 + }, + { + "epoch": 0.9232822295326175, + "grad_norm": 0.8822341561317444, + "learning_rate": 1.600586764137241e-07, + "loss": 0.2316286563873291, + "step": 14875, + "token_acc": 0.9342431843468758 + }, + { + "epoch": 0.9235925765005276, + "grad_norm": 0.8386096954345703, + "learning_rate": 1.587732063070302e-07, + "loss": 0.2243124485015869, + "step": 14880, + "token_acc": 0.9359058164354322 + }, + { + "epoch": 0.9239029234684377, + "grad_norm": 0.8329931497573853, + "learning_rate": 1.574928357106703e-07, + "loss": 0.22314009666442872, + "step": 14885, + "token_acc": 0.9356135219562666 + }, + { + "epoch": 0.9242132704363478, + "grad_norm": 0.8203940987586975, + "learning_rate": 1.562175659733256e-07, + "loss": 0.2266066551208496, + "step": 14890, + "token_acc": 0.9356761512864108 + }, + { + "epoch": 0.9245236174042579, + "grad_norm": 0.8000124096870422, + "learning_rate": 1.549473984383021e-07, + "loss": 0.22768850326538087, + "step": 14895, + "token_acc": 0.935542375044501 + }, + { + "epoch": 0.9248339643721681, + "grad_norm": 0.8235083222389221, + "learning_rate": 1.5368233444353065e-07, + "loss": 0.22810401916503906, + "step": 14900, + "token_acc": 0.9345528775467339 + }, + { + "epoch": 0.9251443113400782, + "grad_norm": 0.8292826414108276, + "learning_rate": 1.5242237532156867e-07, + "loss": 0.22213082313537597, + "step": 14905, + "token_acc": 0.9360719091590048 + }, + { + "epoch": 0.9254546583079883, + "grad_norm": 0.8575629591941833, + "learning_rate": 1.5116752239959564e-07, + "loss": 0.22305045127868653, + "step": 14910, + "token_acc": 0.935664856312861 + }, + { + "epoch": 0.9257650052758984, + "grad_norm": 0.838779091835022, + "learning_rate": 1.499177769994109e-07, + "loss": 0.21985559463500975, + "step": 14915, + "token_acc": 0.9370169913190769 + }, + { + "epoch": 0.9260753522438085, + "grad_norm": 0.8374066948890686, + "learning_rate": 1.486731404374342e-07, + "loss": 0.22093665599822998, + "step": 14920, + "token_acc": 0.936477144704651 + }, + { + "epoch": 0.9263856992117186, + "grad_norm": 0.8346973061561584, + "learning_rate": 1.4743361402470623e-07, + "loss": 0.21868348121643066, + "step": 14925, + "token_acc": 0.9370967098173303 + }, + { + "epoch": 0.9266960461796289, + "grad_norm": 0.8514022827148438, + "learning_rate": 1.4619919906688263e-07, + "loss": 0.22188901901245117, + "step": 14930, + "token_acc": 0.9364490782514769 + }, + { + "epoch": 0.927006393147539, + "grad_norm": 0.8143284320831299, + "learning_rate": 1.4496989686423436e-07, + "loss": 0.22571344375610353, + "step": 14935, + "token_acc": 0.9354455783205226 + }, + { + "epoch": 0.9273167401154491, + "grad_norm": 0.7727757096290588, + "learning_rate": 1.4374570871164894e-07, + "loss": 0.22196607589721679, + "step": 14940, + "token_acc": 0.9354693796657197 + }, + { + "epoch": 0.9276270870833592, + "grad_norm": 0.8324426412582397, + "learning_rate": 1.425266358986266e-07, + "loss": 0.2241354465484619, + "step": 14945, + "token_acc": 0.9355714638360898 + }, + { + "epoch": 0.9279374340512693, + "grad_norm": 0.7846515774726868, + "learning_rate": 1.4131267970927677e-07, + "loss": 0.22103846073150635, + "step": 14950, + "token_acc": 0.9359715724961238 + }, + { + "epoch": 0.9282477810191795, + "grad_norm": 0.8250550031661987, + "learning_rate": 1.401038414223227e-07, + "loss": 0.22910633087158203, + "step": 14955, + "token_acc": 0.9341488101932403 + }, + { + "epoch": 0.9285581279870896, + "grad_norm": 0.8587515354156494, + "learning_rate": 1.389001223110953e-07, + "loss": 0.2300647735595703, + "step": 14960, + "token_acc": 0.9335307755189558 + }, + { + "epoch": 0.9288684749549997, + "grad_norm": 0.8877548575401306, + "learning_rate": 1.3770152364353306e-07, + "loss": 0.22652974128723144, + "step": 14965, + "token_acc": 0.9347390609861479 + }, + { + "epoch": 0.9291788219229098, + "grad_norm": 0.8566204309463501, + "learning_rate": 1.3650804668217998e-07, + "loss": 0.22847704887390136, + "step": 14970, + "token_acc": 0.9352512697139802 + }, + { + "epoch": 0.9294891688908199, + "grad_norm": 0.865992546081543, + "learning_rate": 1.3531969268418598e-07, + "loss": 0.22352111339569092, + "step": 14975, + "token_acc": 0.9355335577376329 + }, + { + "epoch": 0.9297995158587301, + "grad_norm": 0.8228665590286255, + "learning_rate": 1.3413646290130644e-07, + "loss": 0.23345065116882324, + "step": 14980, + "token_acc": 0.9334855403348554 + }, + { + "epoch": 0.9301098628266402, + "grad_norm": 0.83062344789505, + "learning_rate": 1.329583585798966e-07, + "loss": 0.22519009113311766, + "step": 14985, + "token_acc": 0.9355026770577158 + }, + { + "epoch": 0.9304202097945503, + "grad_norm": 0.8632986545562744, + "learning_rate": 1.3178538096091265e-07, + "loss": 0.22674951553344727, + "step": 14990, + "token_acc": 0.9356594538208468 + }, + { + "epoch": 0.9307305567624604, + "grad_norm": 0.8211206793785095, + "learning_rate": 1.306175312799124e-07, + "loss": 0.21751713752746582, + "step": 14995, + "token_acc": 0.9365848799760949 + }, + { + "epoch": 0.9310409037303705, + "grad_norm": 1.143462896347046, + "learning_rate": 1.2945481076705124e-07, + "loss": 0.22513751983642577, + "step": 15000, + "token_acc": 0.9355509901347528 + }, + { + "epoch": 0.9310409037303705, + "eval_loss": 0.2354583442211151, + "eval_runtime": 102.1542, + "eval_samples_per_second": 195.782, + "eval_steps_per_second": 0.264, + "eval_token_acc": 0.9324655707299818, + "step": 15000 + }, + { + "epoch": 0.9313512506982807, + "grad_norm": 0.7983279824256897, + "learning_rate": 1.2829722064708172e-07, + "loss": 0.22428507804870607, + "step": 15005, + "token_acc": 0.9354113582091543 + }, + { + "epoch": 0.9316615976661908, + "grad_norm": 0.8185499310493469, + "learning_rate": 1.2714476213935288e-07, + "loss": 0.22483978271484376, + "step": 15010, + "token_acc": 0.9357483046196996 + }, + { + "epoch": 0.9319719446341009, + "grad_norm": 0.817832350730896, + "learning_rate": 1.259974364578076e-07, + "loss": 0.22232589721679688, + "step": 15015, + "token_acc": 0.936296335676976 + }, + { + "epoch": 0.932282291602011, + "grad_norm": 0.8298930525779724, + "learning_rate": 1.2485524481098132e-07, + "loss": 0.22351136207580566, + "step": 15020, + "token_acc": 0.9354666167444473 + }, + { + "epoch": 0.9325926385699211, + "grad_norm": 0.8336822986602783, + "learning_rate": 1.2371818840200445e-07, + "loss": 0.22703261375427247, + "step": 15025, + "token_acc": 0.9351351351351351 + }, + { + "epoch": 0.9329029855378312, + "grad_norm": 0.8150284290313721, + "learning_rate": 1.2258626842859444e-07, + "loss": 0.2251688003540039, + "step": 15030, + "token_acc": 0.9354440653807831 + }, + { + "epoch": 0.9332133325057415, + "grad_norm": 0.8681549429893494, + "learning_rate": 1.2145948608306147e-07, + "loss": 0.21708149909973146, + "step": 15035, + "token_acc": 0.9369058488640636 + }, + { + "epoch": 0.9335236794736516, + "grad_norm": 0.8482911586761475, + "learning_rate": 1.2033784255230218e-07, + "loss": 0.22627925872802734, + "step": 15040, + "token_acc": 0.9352258270779354 + }, + { + "epoch": 0.9338340264415617, + "grad_norm": 0.8087306022644043, + "learning_rate": 1.1922133901780042e-07, + "loss": 0.22343907356262208, + "step": 15045, + "token_acc": 0.9361284285168247 + }, + { + "epoch": 0.9341443734094718, + "grad_norm": 0.8197100162506104, + "learning_rate": 1.1810997665562651e-07, + "loss": 0.21994640827178955, + "step": 15050, + "token_acc": 0.9363516018003707 + }, + { + "epoch": 0.9344547203773819, + "grad_norm": 0.8434872627258301, + "learning_rate": 1.1700375663643571e-07, + "loss": 0.23080043792724608, + "step": 15055, + "token_acc": 0.9345930617761641 + }, + { + "epoch": 0.9347650673452921, + "grad_norm": 0.8584432005882263, + "learning_rate": 1.1590268012546479e-07, + "loss": 0.22061944007873535, + "step": 15060, + "token_acc": 0.936584303476677 + }, + { + "epoch": 0.9350754143132022, + "grad_norm": 0.7802200317382812, + "learning_rate": 1.1480674828253324e-07, + "loss": 0.22167396545410156, + "step": 15065, + "token_acc": 0.9374548229057907 + }, + { + "epoch": 0.9353857612811123, + "grad_norm": 0.8954899311065674, + "learning_rate": 1.1371596226204262e-07, + "loss": 0.2254354476928711, + "step": 15070, + "token_acc": 0.9367627230040586 + }, + { + "epoch": 0.9356961082490224, + "grad_norm": 0.8148626089096069, + "learning_rate": 1.1263032321297385e-07, + "loss": 0.22028803825378418, + "step": 15075, + "token_acc": 0.9367313267029955 + }, + { + "epoch": 0.9360064552169325, + "grad_norm": 0.8108525276184082, + "learning_rate": 1.1154983227888438e-07, + "loss": 0.22486391067504882, + "step": 15080, + "token_acc": 0.9353436241075346 + }, + { + "epoch": 0.9363168021848427, + "grad_norm": 0.8741359114646912, + "learning_rate": 1.1047449059791049e-07, + "loss": 0.22285428047180175, + "step": 15085, + "token_acc": 0.9357492641419141 + }, + { + "epoch": 0.9366271491527528, + "grad_norm": 0.7960735559463501, + "learning_rate": 1.0940429930276497e-07, + "loss": 0.22872886657714844, + "step": 15090, + "token_acc": 0.9342841219641574 + }, + { + "epoch": 0.9369374961206629, + "grad_norm": 0.8427518606185913, + "learning_rate": 1.0833925952073443e-07, + "loss": 0.22136073112487792, + "step": 15095, + "token_acc": 0.9369343986261212 + }, + { + "epoch": 0.937247843088573, + "grad_norm": 0.7975824475288391, + "learning_rate": 1.0727937237367924e-07, + "loss": 0.22517714500427247, + "step": 15100, + "token_acc": 0.9362412493268712 + }, + { + "epoch": 0.9375581900564831, + "grad_norm": 0.863105297088623, + "learning_rate": 1.0622463897803248e-07, + "loss": 0.22664155960083007, + "step": 15105, + "token_acc": 0.9350580533513988 + }, + { + "epoch": 0.9378685370243933, + "grad_norm": 0.8479633927345276, + "learning_rate": 1.0517506044479875e-07, + "loss": 0.22400825023651122, + "step": 15110, + "token_acc": 0.935297683367156 + }, + { + "epoch": 0.9381788839923034, + "grad_norm": 0.8416533470153809, + "learning_rate": 1.0413063787955258e-07, + "loss": 0.22233805656433106, + "step": 15115, + "token_acc": 0.9358511672504028 + }, + { + "epoch": 0.9384892309602135, + "grad_norm": 0.8232436180114746, + "learning_rate": 1.030913723824356e-07, + "loss": 0.2247161626815796, + "step": 15120, + "token_acc": 0.9359148417055271 + }, + { + "epoch": 0.9387995779281236, + "grad_norm": 0.9780632257461548, + "learning_rate": 1.0205726504816215e-07, + "loss": 0.22240190505981444, + "step": 15125, + "token_acc": 0.9369569986107141 + }, + { + "epoch": 0.9391099248960337, + "grad_norm": 0.8231025338172913, + "learning_rate": 1.0102831696600757e-07, + "loss": 0.2330986499786377, + "step": 15130, + "token_acc": 0.9337647229896757 + }, + { + "epoch": 0.9394202718639438, + "grad_norm": 0.8251264691352844, + "learning_rate": 1.0000452921981596e-07, + "loss": 0.23132362365722656, + "step": 15135, + "token_acc": 0.9339897323311089 + }, + { + "epoch": 0.939730618831854, + "grad_norm": 0.8271506428718567, + "learning_rate": 9.898590288799526e-08, + "loss": 0.22400758266448975, + "step": 15140, + "token_acc": 0.9358257713248639 + }, + { + "epoch": 0.9400409657997641, + "grad_norm": 0.8518717288970947, + "learning_rate": 9.797243904351606e-08, + "loss": 0.21921846866607667, + "step": 15145, + "token_acc": 0.9365768270056579 + }, + { + "epoch": 0.9403513127676743, + "grad_norm": 0.9078429341316223, + "learning_rate": 9.696413875391109e-08, + "loss": 0.22769925594329835, + "step": 15150, + "token_acc": 0.9350705480271819 + }, + { + "epoch": 0.9406616597355844, + "grad_norm": 0.808941662311554, + "learning_rate": 9.59610030812752e-08, + "loss": 0.22335097789764405, + "step": 15155, + "token_acc": 0.936111546704302 + }, + { + "epoch": 0.9409720067034945, + "grad_norm": 0.8358493447303772, + "learning_rate": 9.496303308226151e-08, + "loss": 0.2244410037994385, + "step": 15160, + "token_acc": 0.9355537271680106 + }, + { + "epoch": 0.9412823536714047, + "grad_norm": 0.8067222833633423, + "learning_rate": 9.397022980808245e-08, + "loss": 0.22346572875976561, + "step": 15165, + "token_acc": 0.9350882001029391 + }, + { + "epoch": 0.9415927006393148, + "grad_norm": 0.8393672704696655, + "learning_rate": 9.298259430450873e-08, + "loss": 0.22526037693023682, + "step": 15170, + "token_acc": 0.9355808747368772 + }, + { + "epoch": 0.9419030476072249, + "grad_norm": 0.8061415553092957, + "learning_rate": 9.200012761186649e-08, + "loss": 0.2261514186859131, + "step": 15175, + "token_acc": 0.9360237864564194 + }, + { + "epoch": 0.942213394575135, + "grad_norm": 0.7989704012870789, + "learning_rate": 9.102283076503849e-08, + "loss": 0.22367300987243652, + "step": 15180, + "token_acc": 0.936292727842907 + }, + { + "epoch": 0.9425237415430451, + "grad_norm": 5.452362060546875, + "learning_rate": 9.005070479346067e-08, + "loss": 0.22573423385620117, + "step": 15185, + "token_acc": 0.9352249244953108 + }, + { + "epoch": 0.9428340885109553, + "grad_norm": 1.2356898784637451, + "learning_rate": 8.908375072112285e-08, + "loss": 0.2263432502746582, + "step": 15190, + "token_acc": 0.9348791021300419 + }, + { + "epoch": 0.9431444354788654, + "grad_norm": 0.8187481164932251, + "learning_rate": 8.812196956656694e-08, + "loss": 0.2323697566986084, + "step": 15195, + "token_acc": 0.9340425251272853 + }, + { + "epoch": 0.9434547824467755, + "grad_norm": 0.8493360280990601, + "learning_rate": 8.716536234288642e-08, + "loss": 0.22659645080566407, + "step": 15200, + "token_acc": 0.9348433074949143 + }, + { + "epoch": 0.9437651294146856, + "grad_norm": 0.8390522003173828, + "learning_rate": 8.62139300577236e-08, + "loss": 0.22274446487426758, + "step": 15205, + "token_acc": 0.9362010600563745 + }, + { + "epoch": 0.9440754763825957, + "grad_norm": 0.814529299736023, + "learning_rate": 8.526767371327016e-08, + "loss": 0.2261338233947754, + "step": 15210, + "token_acc": 0.9356428011234524 + }, + { + "epoch": 0.9443858233505059, + "grad_norm": 0.8201233148574829, + "learning_rate": 8.432659430626711e-08, + "loss": 0.22249193191528321, + "step": 15215, + "token_acc": 0.9357081317516072 + }, + { + "epoch": 0.944696170318416, + "grad_norm": 0.8114063143730164, + "learning_rate": 8.339069282800094e-08, + "loss": 0.2166703462600708, + "step": 15220, + "token_acc": 0.9372107251157099 + }, + { + "epoch": 0.9450065172863261, + "grad_norm": 0.7758262157440186, + "learning_rate": 8.245997026430363e-08, + "loss": 0.21614983081817626, + "step": 15225, + "token_acc": 0.9364209463454167 + }, + { + "epoch": 0.9453168642542362, + "grad_norm": 0.9567193388938904, + "learning_rate": 8.153442759555319e-08, + "loss": 0.22854628562927246, + "step": 15230, + "token_acc": 0.9352720757011974 + }, + { + "epoch": 0.9456272112221463, + "grad_norm": 0.8302553296089172, + "learning_rate": 8.061406579667031e-08, + "loss": 0.22727298736572266, + "step": 15235, + "token_acc": 0.9348053296873049 + }, + { + "epoch": 0.9459375581900565, + "grad_norm": 0.8652146458625793, + "learning_rate": 7.969888583711948e-08, + "loss": 0.22525861263275146, + "step": 15240, + "token_acc": 0.9356488529231098 + }, + { + "epoch": 0.9462479051579666, + "grad_norm": 0.8045849204063416, + "learning_rate": 7.878888868090629e-08, + "loss": 0.22467689514160155, + "step": 15245, + "token_acc": 0.935139955544316 + }, + { + "epoch": 0.9465582521258767, + "grad_norm": 0.8361722826957703, + "learning_rate": 7.78840752865756e-08, + "loss": 0.22995285987854003, + "step": 15250, + "token_acc": 0.9345721413776445 + }, + { + "epoch": 0.9468685990937868, + "grad_norm": 1.2566227912902832, + "learning_rate": 7.698444660721505e-08, + "loss": 0.22309591770172119, + "step": 15255, + "token_acc": 0.9357952133305031 + }, + { + "epoch": 0.947178946061697, + "grad_norm": 0.8758302330970764, + "learning_rate": 7.609000359044883e-08, + "loss": 0.2246638774871826, + "step": 15260, + "token_acc": 0.935675532618764 + }, + { + "epoch": 0.947489293029607, + "grad_norm": 0.7966662049293518, + "learning_rate": 7.520074717843884e-08, + "loss": 0.2244575023651123, + "step": 15265, + "token_acc": 0.93543662942494 + }, + { + "epoch": 0.9477996399975173, + "grad_norm": 0.8384653329849243, + "learning_rate": 7.431667830788414e-08, + "loss": 0.21775059700012206, + "step": 15270, + "token_acc": 0.9373261004463839 + }, + { + "epoch": 0.9481099869654274, + "grad_norm": 0.8193512558937073, + "learning_rate": 7.343779791002037e-08, + "loss": 0.22500150203704833, + "step": 15275, + "token_acc": 0.9364104622382768 + }, + { + "epoch": 0.9484203339333375, + "grad_norm": 0.8730255961418152, + "learning_rate": 7.256410691061533e-08, + "loss": 0.23197007179260254, + "step": 15280, + "token_acc": 0.9343172182023043 + }, + { + "epoch": 0.9487306809012476, + "grad_norm": 0.8121325969696045, + "learning_rate": 7.16956062299734e-08, + "loss": 0.2226422071456909, + "step": 15285, + "token_acc": 0.9363289382373352 + }, + { + "epoch": 0.9490410278691577, + "grad_norm": 0.8645696043968201, + "learning_rate": 7.083229678293002e-08, + "loss": 0.22263357639312745, + "step": 15290, + "token_acc": 0.9356545942365465 + }, + { + "epoch": 0.9493513748370679, + "grad_norm": 0.8563567399978638, + "learning_rate": 6.997417947885277e-08, + "loss": 0.22628865242004395, + "step": 15295, + "token_acc": 0.9349733890061857 + }, + { + "epoch": 0.949661721804978, + "grad_norm": 0.8714470863342285, + "learning_rate": 6.912125522164082e-08, + "loss": 0.21658577919006347, + "step": 15300, + "token_acc": 0.9377701806374118 + }, + { + "epoch": 0.9499720687728881, + "grad_norm": 0.8267343640327454, + "learning_rate": 6.827352490972216e-08, + "loss": 0.21735353469848634, + "step": 15305, + "token_acc": 0.9371506457933331 + }, + { + "epoch": 0.9502824157407982, + "grad_norm": 0.8219081163406372, + "learning_rate": 6.743098943605419e-08, + "loss": 0.23012862205505372, + "step": 15310, + "token_acc": 0.933909618850544 + }, + { + "epoch": 0.9505927627087083, + "grad_norm": 0.8483142852783203, + "learning_rate": 6.65936496881231e-08, + "loss": 0.22592477798461913, + "step": 15315, + "token_acc": 0.935226547543076 + }, + { + "epoch": 0.9509031096766185, + "grad_norm": 0.8228432536125183, + "learning_rate": 6.576150654794056e-08, + "loss": 0.22719712257385255, + "step": 15320, + "token_acc": 0.9351193330136913 + }, + { + "epoch": 0.9512134566445286, + "grad_norm": 0.9031339287757874, + "learning_rate": 6.493456089204541e-08, + "loss": 0.226958703994751, + "step": 15325, + "token_acc": 0.9349737200181425 + }, + { + "epoch": 0.9515238036124387, + "grad_norm": 0.7835374474525452, + "learning_rate": 6.411281359150202e-08, + "loss": 0.2203263521194458, + "step": 15330, + "token_acc": 0.9358614013419234 + }, + { + "epoch": 0.9518341505803488, + "grad_norm": 0.7952069640159607, + "learning_rate": 6.32962655118985e-08, + "loss": 0.22332868576049805, + "step": 15335, + "token_acc": 0.9358081620177805 + }, + { + "epoch": 0.9521444975482589, + "grad_norm": 2.0506751537323, + "learning_rate": 6.248491751334517e-08, + "loss": 0.223779034614563, + "step": 15340, + "token_acc": 0.9352412122971752 + }, + { + "epoch": 0.9524548445161691, + "grad_norm": 0.8377223610877991, + "learning_rate": 6.167877045047844e-08, + "loss": 0.22706923484802247, + "step": 15345, + "token_acc": 0.9351074104912573 + }, + { + "epoch": 0.9527651914840792, + "grad_norm": 0.7963693141937256, + "learning_rate": 6.087782517245233e-08, + "loss": 0.23039445877075196, + "step": 15350, + "token_acc": 0.9345963102559016 + }, + { + "epoch": 0.9530755384519893, + "grad_norm": 0.7980961203575134, + "learning_rate": 6.008208252294368e-08, + "loss": 0.2203824520111084, + "step": 15355, + "token_acc": 0.9367820650090547 + }, + { + "epoch": 0.9533858854198994, + "grad_norm": 0.7977734804153442, + "learning_rate": 5.929154334014864e-08, + "loss": 0.2187025785446167, + "step": 15360, + "token_acc": 0.9369687493788964 + }, + { + "epoch": 0.9536962323878095, + "grad_norm": 0.808886706829071, + "learning_rate": 5.850620845678223e-08, + "loss": 0.22373590469360352, + "step": 15365, + "token_acc": 0.9355298949277295 + }, + { + "epoch": 0.9540065793557196, + "grad_norm": 0.7952503561973572, + "learning_rate": 5.7726078700077736e-08, + "loss": 0.22443680763244628, + "step": 15370, + "token_acc": 0.9355510873099613 + }, + { + "epoch": 0.9543169263236299, + "grad_norm": 0.809476912021637, + "learning_rate": 5.6951154891785597e-08, + "loss": 0.22569818496704103, + "step": 15375, + "token_acc": 0.935259868596763 + }, + { + "epoch": 0.95462727329154, + "grad_norm": 0.8188254833221436, + "learning_rate": 5.6181437848172315e-08, + "loss": 0.22699193954467772, + "step": 15380, + "token_acc": 0.9347692799829392 + }, + { + "epoch": 0.9549376202594501, + "grad_norm": 0.8684312701225281, + "learning_rate": 5.541692838002044e-08, + "loss": 0.2206878900527954, + "step": 15385, + "token_acc": 0.936507406913119 + }, + { + "epoch": 0.9552479672273602, + "grad_norm": 0.8066678047180176, + "learning_rate": 5.465762729262636e-08, + "loss": 0.22271337509155273, + "step": 15390, + "token_acc": 0.9356075884916253 + }, + { + "epoch": 0.9555583141952703, + "grad_norm": 0.8246888518333435, + "learning_rate": 5.390353538579973e-08, + "loss": 0.2283928871154785, + "step": 15395, + "token_acc": 0.9347990510113711 + }, + { + "epoch": 0.9558686611631805, + "grad_norm": 0.8588503003120422, + "learning_rate": 5.315465345386517e-08, + "loss": 0.22897400856018066, + "step": 15400, + "token_acc": 0.9339744751352611 + }, + { + "epoch": 0.9561790081310906, + "grad_norm": 0.82988041639328, + "learning_rate": 5.241098228565722e-08, + "loss": 0.22697460651397705, + "step": 15405, + "token_acc": 0.9345906637279098 + }, + { + "epoch": 0.9564893550990007, + "grad_norm": 0.8785929083824158, + "learning_rate": 5.167252266452261e-08, + "loss": 0.22281122207641602, + "step": 15410, + "token_acc": 0.9360419190057402 + }, + { + "epoch": 0.9567997020669108, + "grad_norm": 0.8396813869476318, + "learning_rate": 5.093927536831911e-08, + "loss": 0.2242985963821411, + "step": 15415, + "token_acc": 0.9355860612460402 + }, + { + "epoch": 0.9571100490348209, + "grad_norm": 0.8194007873535156, + "learning_rate": 5.021124116941278e-08, + "loss": 0.22291479110717774, + "step": 15420, + "token_acc": 0.9363231220138917 + }, + { + "epoch": 0.9574203960027311, + "grad_norm": 0.7992196083068848, + "learning_rate": 4.9488420834679085e-08, + "loss": 0.2107405662536621, + "step": 15425, + "token_acc": 0.9386169590451089 + }, + { + "epoch": 0.9577307429706412, + "grad_norm": 0.8669182658195496, + "learning_rate": 4.877081512550119e-08, + "loss": 0.22837629318237304, + "step": 15430, + "token_acc": 0.9343999675548526 + }, + { + "epoch": 0.9580410899385513, + "grad_norm": 0.931969404220581, + "learning_rate": 4.8058424797770565e-08, + "loss": 0.23075571060180664, + "step": 15435, + "token_acc": 0.9345421810699589 + }, + { + "epoch": 0.9583514369064614, + "grad_norm": 0.9046128988265991, + "learning_rate": 4.735125060188361e-08, + "loss": 0.22263517379760742, + "step": 15440, + "token_acc": 0.9363239540662145 + }, + { + "epoch": 0.9586617838743715, + "grad_norm": 0.8204918503761292, + "learning_rate": 4.6649293282742814e-08, + "loss": 0.22345831394195556, + "step": 15445, + "token_acc": 0.9355371346108668 + }, + { + "epoch": 0.9589721308422817, + "grad_norm": 0.8200190663337708, + "learning_rate": 4.5952553579755586e-08, + "loss": 0.22515597343444824, + "step": 15450, + "token_acc": 0.9358677532876458 + }, + { + "epoch": 0.9592824778101918, + "grad_norm": 0.7968879342079163, + "learning_rate": 4.5261032226832645e-08, + "loss": 0.22422480583190918, + "step": 15455, + "token_acc": 0.9357563654518223 + }, + { + "epoch": 0.9595928247781019, + "grad_norm": 0.7624355554580688, + "learning_rate": 4.4574729952390204e-08, + "loss": 0.22359352111816405, + "step": 15460, + "token_acc": 0.9360458897149275 + }, + { + "epoch": 0.959903171746012, + "grad_norm": 0.819399356842041, + "learning_rate": 4.389364747934388e-08, + "loss": 0.22677216529846192, + "step": 15465, + "token_acc": 0.9355797819623389 + }, + { + "epoch": 0.9602135187139221, + "grad_norm": 0.8478218913078308, + "learning_rate": 4.3217785525112574e-08, + "loss": 0.22739701271057128, + "step": 15470, + "token_acc": 0.9345117966854897 + }, + { + "epoch": 0.9605238656818322, + "grad_norm": 0.8312594890594482, + "learning_rate": 4.254714480161681e-08, + "loss": 0.2292851448059082, + "step": 15475, + "token_acc": 0.9343431362906134 + }, + { + "epoch": 0.9608342126497424, + "grad_norm": 0.7991034984588623, + "learning_rate": 4.188172601527651e-08, + "loss": 0.21993248462677, + "step": 15480, + "token_acc": 0.9359837427434783 + }, + { + "epoch": 0.9611445596176525, + "grad_norm": 0.8233107924461365, + "learning_rate": 4.122152986700989e-08, + "loss": 0.21606411933898925, + "step": 15485, + "token_acc": 0.9368456085466104 + }, + { + "epoch": 0.9614549065855627, + "grad_norm": 1.1857753992080688, + "learning_rate": 4.0566557052236776e-08, + "loss": 0.22082462310791015, + "step": 15490, + "token_acc": 0.9364271457085829 + }, + { + "epoch": 0.9617652535534728, + "grad_norm": 0.8240007162094116, + "learning_rate": 3.9916808260871964e-08, + "loss": 0.22246360778808594, + "step": 15495, + "token_acc": 0.9357665418058392 + }, + { + "epoch": 0.9620756005213829, + "grad_norm": 0.8382442593574524, + "learning_rate": 3.92722841773302e-08, + "loss": 0.2287151336669922, + "step": 15500, + "token_acc": 0.9353692034254322 + }, + { + "epoch": 0.9620756005213829, + "eval_loss": 0.2353387475013733, + "eval_runtime": 105.9865, + "eval_samples_per_second": 188.703, + "eval_steps_per_second": 0.255, + "eval_token_acc": 0.9324512301801651, + "step": 15500 + }, + { + "epoch": 0.9623859474892931, + "grad_norm": 0.8209149837493896, + "learning_rate": 3.863298548052008e-08, + "loss": 0.22818026542663575, + "step": 15505, + "token_acc": 0.9340887478067532 + }, + { + "epoch": 0.9626962944572032, + "grad_norm": 0.819057822227478, + "learning_rate": 3.799891284384849e-08, + "loss": 0.2238163948059082, + "step": 15510, + "token_acc": 0.9355069867431028 + }, + { + "epoch": 0.9630066414251133, + "grad_norm": 0.799403190612793, + "learning_rate": 3.737006693521561e-08, + "loss": 0.22618515491485597, + "step": 15515, + "token_acc": 0.9349316202523262 + }, + { + "epoch": 0.9633169883930234, + "grad_norm": 0.7988102436065674, + "learning_rate": 3.674644841701769e-08, + "loss": 0.22670292854309082, + "step": 15520, + "token_acc": 0.9346353024361181 + }, + { + "epoch": 0.9636273353609335, + "grad_norm": 0.8185102939605713, + "learning_rate": 3.6128057946142604e-08, + "loss": 0.22490134239196777, + "step": 15525, + "token_acc": 0.9358514946448553 + }, + { + "epoch": 0.9639376823288437, + "grad_norm": 0.8071334958076477, + "learning_rate": 3.551489617397374e-08, + "loss": 0.2241300106048584, + "step": 15530, + "token_acc": 0.9355907790588345 + }, + { + "epoch": 0.9642480292967538, + "grad_norm": 0.8759570121765137, + "learning_rate": 3.490696374638558e-08, + "loss": 0.2264106035232544, + "step": 15535, + "token_acc": 0.9351200621698046 + }, + { + "epoch": 0.9645583762646639, + "grad_norm": 0.8114751577377319, + "learning_rate": 3.4304261303742535e-08, + "loss": 0.22525432109832763, + "step": 15540, + "token_acc": 0.9355221988989663 + }, + { + "epoch": 0.964868723232574, + "grad_norm": 0.8177556395530701, + "learning_rate": 3.3706789480903446e-08, + "loss": 0.2300882577896118, + "step": 15545, + "token_acc": 0.933638321851805 + }, + { + "epoch": 0.9651790702004841, + "grad_norm": 0.830560028553009, + "learning_rate": 3.3114548907215436e-08, + "loss": 0.214280366897583, + "step": 15550, + "token_acc": 0.9376256339940627 + }, + { + "epoch": 0.9654894171683943, + "grad_norm": 0.9937106370925903, + "learning_rate": 3.2527540206515054e-08, + "loss": 0.22566475868225097, + "step": 15555, + "token_acc": 0.9349657942592975 + }, + { + "epoch": 0.9657997641363044, + "grad_norm": 0.7684742212295532, + "learning_rate": 3.1945763997129344e-08, + "loss": 0.22068114280700685, + "step": 15560, + "token_acc": 0.9360149163608857 + }, + { + "epoch": 0.9661101111042145, + "grad_norm": 0.8296353816986084, + "learning_rate": 3.136922089187144e-08, + "loss": 0.22519922256469727, + "step": 15565, + "token_acc": 0.9343289729808734 + }, + { + "epoch": 0.9664204580721246, + "grad_norm": 0.8396661877632141, + "learning_rate": 3.079791149804445e-08, + "loss": 0.22880682945251465, + "step": 15570, + "token_acc": 0.934270228977471 + }, + { + "epoch": 0.9667308050400347, + "grad_norm": 0.8300756812095642, + "learning_rate": 3.0231836417438095e-08, + "loss": 0.22211871147155762, + "step": 15575, + "token_acc": 0.9368708628071792 + }, + { + "epoch": 0.9670411520079449, + "grad_norm": 0.8245381712913513, + "learning_rate": 2.9670996246326522e-08, + "loss": 0.227608060836792, + "step": 15580, + "token_acc": 0.9348357663011074 + }, + { + "epoch": 0.967351498975855, + "grad_norm": 0.8507952690124512, + "learning_rate": 2.911539157547216e-08, + "loss": 0.22475545406341552, + "step": 15585, + "token_acc": 0.9353679954932564 + }, + { + "epoch": 0.9676618459437651, + "grad_norm": 0.8144983053207397, + "learning_rate": 2.856502299012187e-08, + "loss": 0.21740415096282958, + "step": 15590, + "token_acc": 0.9371765768005391 + }, + { + "epoch": 0.9679721929116752, + "grad_norm": 0.8377090692520142, + "learning_rate": 2.8019891070006354e-08, + "loss": 0.22457399368286132, + "step": 15595, + "token_acc": 0.9354453712421594 + }, + { + "epoch": 0.9682825398795853, + "grad_norm": 0.8060710430145264, + "learning_rate": 2.7479996389340736e-08, + "loss": 0.21875674724578859, + "step": 15600, + "token_acc": 0.9366500696005996 + }, + { + "epoch": 0.9685928868474954, + "grad_norm": 0.8571235537528992, + "learning_rate": 2.6945339516823988e-08, + "loss": 0.2335214376449585, + "step": 15605, + "token_acc": 0.9331404882784315 + }, + { + "epoch": 0.9689032338154057, + "grad_norm": 0.8052919507026672, + "learning_rate": 2.641592101563728e-08, + "loss": 0.2210322856903076, + "step": 15610, + "token_acc": 0.9358845385162233 + }, + { + "epoch": 0.9692135807833158, + "grad_norm": 0.8269677758216858, + "learning_rate": 2.5891741443443418e-08, + "loss": 0.22711396217346191, + "step": 15615, + "token_acc": 0.9337541308658295 + }, + { + "epoch": 0.9695239277512259, + "grad_norm": 0.8817762732505798, + "learning_rate": 2.537280135238851e-08, + "loss": 0.22270312309265136, + "step": 15620, + "token_acc": 0.9359517860939826 + }, + { + "epoch": 0.969834274719136, + "grad_norm": 0.8504263162612915, + "learning_rate": 2.485910128909752e-08, + "loss": 0.22194621562957764, + "step": 15625, + "token_acc": 0.9360558194140562 + }, + { + "epoch": 0.9701446216870461, + "grad_norm": 0.8006669282913208, + "learning_rate": 2.4350641794677055e-08, + "loss": 0.22227416038513184, + "step": 15630, + "token_acc": 0.9359969702707821 + }, + { + "epoch": 0.9704549686549563, + "grad_norm": 0.8290884494781494, + "learning_rate": 2.3847423404713686e-08, + "loss": 0.2239783763885498, + "step": 15635, + "token_acc": 0.9353793213828425 + }, + { + "epoch": 0.9707653156228664, + "grad_norm": 0.794661819934845, + "learning_rate": 2.3349446649272856e-08, + "loss": 0.22291333675384523, + "step": 15640, + "token_acc": 0.9355991882702361 + }, + { + "epoch": 0.9710756625907765, + "grad_norm": 0.794541597366333, + "learning_rate": 2.2856712052898854e-08, + "loss": 0.22443385124206544, + "step": 15645, + "token_acc": 0.9350375141092889 + }, + { + "epoch": 0.9713860095586866, + "grad_norm": 0.8332598805427551, + "learning_rate": 2.2369220134613732e-08, + "loss": 0.21990303993225097, + "step": 15650, + "token_acc": 0.9365769936936218 + }, + { + "epoch": 0.9716963565265967, + "grad_norm": 0.8595775961875916, + "learning_rate": 2.1886971407917846e-08, + "loss": 0.2271794319152832, + "step": 15655, + "token_acc": 0.9349701215880727 + }, + { + "epoch": 0.9720067034945069, + "grad_norm": 0.8646977543830872, + "learning_rate": 2.140996638078818e-08, + "loss": 0.21491079330444335, + "step": 15660, + "token_acc": 0.9382688854623492 + }, + { + "epoch": 0.972317050462417, + "grad_norm": 0.8218981027603149, + "learning_rate": 2.093820555567838e-08, + "loss": 0.2168516159057617, + "step": 15665, + "token_acc": 0.9374552636462448 + }, + { + "epoch": 0.9726273974303271, + "grad_norm": 0.8405992984771729, + "learning_rate": 2.0471689429518162e-08, + "loss": 0.22788629531860352, + "step": 15670, + "token_acc": 0.934632402769104 + }, + { + "epoch": 0.9729377443982372, + "grad_norm": 0.8392440676689148, + "learning_rate": 2.0010418493712214e-08, + "loss": 0.22279725074768067, + "step": 15675, + "token_acc": 0.9365945127968945 + }, + { + "epoch": 0.9732480913661473, + "grad_norm": 0.8152569532394409, + "learning_rate": 1.9554393234141877e-08, + "loss": 0.22373075485229493, + "step": 15680, + "token_acc": 0.9357038512616201 + }, + { + "epoch": 0.9735584383340575, + "grad_norm": 0.8252231478691101, + "learning_rate": 1.910361413116013e-08, + "loss": 0.21790575981140137, + "step": 15685, + "token_acc": 0.936180811072885 + }, + { + "epoch": 0.9738687853019676, + "grad_norm": 0.8161397576332092, + "learning_rate": 1.8658081659596593e-08, + "loss": 0.22199602127075196, + "step": 15690, + "token_acc": 0.9358669196677014 + }, + { + "epoch": 0.9741791322698777, + "grad_norm": 0.8439284563064575, + "learning_rate": 1.8217796288753643e-08, + "loss": 0.2254786729812622, + "step": 15695, + "token_acc": 0.9353280183141895 + }, + { + "epoch": 0.9744894792377878, + "grad_norm": 0.8505290150642395, + "learning_rate": 1.778275848240474e-08, + "loss": 0.22276694774627687, + "step": 15700, + "token_acc": 0.9359814296261315 + }, + { + "epoch": 0.9747998262056979, + "grad_norm": 0.9368773102760315, + "learning_rate": 1.735296869879888e-08, + "loss": 0.2216945171356201, + "step": 15705, + "token_acc": 0.9364060560603747 + }, + { + "epoch": 0.975110173173608, + "grad_norm": 0.8455086350440979, + "learning_rate": 1.692842739065448e-08, + "loss": 0.2272592306137085, + "step": 15710, + "token_acc": 0.9356848293410547 + }, + { + "epoch": 0.9754205201415183, + "grad_norm": 0.8515515923500061, + "learning_rate": 1.6509135005162715e-08, + "loss": 0.22269172668457032, + "step": 15715, + "token_acc": 0.9366872299013624 + }, + { + "epoch": 0.9757308671094284, + "grad_norm": 0.8456263542175293, + "learning_rate": 1.6095091983986398e-08, + "loss": 0.2282790422439575, + "step": 15720, + "token_acc": 0.9341726234930119 + }, + { + "epoch": 0.9760412140773385, + "grad_norm": 0.8005826473236084, + "learning_rate": 1.568629876325667e-08, + "loss": 0.22884807586669922, + "step": 15725, + "token_acc": 0.9342829933603713 + }, + { + "epoch": 0.9763515610452486, + "grad_norm": 0.8603096604347229, + "learning_rate": 1.528275577357685e-08, + "loss": 0.2260124683380127, + "step": 15730, + "token_acc": 0.9360232593972119 + }, + { + "epoch": 0.9766619080131587, + "grad_norm": 0.8844029903411865, + "learning_rate": 1.4884463440020259e-08, + "loss": 0.22288291454315184, + "step": 15735, + "token_acc": 0.9361955516523289 + }, + { + "epoch": 0.9769722549810689, + "grad_norm": 1.0851752758026123, + "learning_rate": 1.4491422182126847e-08, + "loss": 0.2216958999633789, + "step": 15740, + "token_acc": 0.936319567693571 + }, + { + "epoch": 0.977282601948979, + "grad_norm": 0.8752528429031372, + "learning_rate": 1.4103632413907664e-08, + "loss": 0.22626588344573975, + "step": 15745, + "token_acc": 0.9346526682990641 + }, + { + "epoch": 0.9775929489168891, + "grad_norm": 0.8206929564476013, + "learning_rate": 1.3721094543841518e-08, + "loss": 0.22306499481201172, + "step": 15750, + "token_acc": 0.936111537716762 + }, + { + "epoch": 0.9779032958847992, + "grad_norm": 0.8145995140075684, + "learning_rate": 1.3343808974875528e-08, + "loss": 0.22847816944122315, + "step": 15755, + "token_acc": 0.9355646660251256 + }, + { + "epoch": 0.9782136428527093, + "grad_norm": 0.8176822662353516, + "learning_rate": 1.297177610442235e-08, + "loss": 0.216821026802063, + "step": 15760, + "token_acc": 0.9367052521314823 + }, + { + "epoch": 0.9785239898206195, + "grad_norm": 0.8210958242416382, + "learning_rate": 1.2604996324364072e-08, + "loss": 0.22356324195861815, + "step": 15765, + "token_acc": 0.935554312801257 + }, + { + "epoch": 0.9788343367885296, + "grad_norm": 0.8578672409057617, + "learning_rate": 1.2243470021048309e-08, + "loss": 0.22584085464477538, + "step": 15770, + "token_acc": 0.935310116793771 + }, + { + "epoch": 0.9791446837564397, + "grad_norm": 0.8195298910140991, + "learning_rate": 1.1887197575289333e-08, + "loss": 0.22321221828460694, + "step": 15775, + "token_acc": 0.9361123095814942 + }, + { + "epoch": 0.9794550307243498, + "grad_norm": 0.8077801465988159, + "learning_rate": 1.1536179362366395e-08, + "loss": 0.21732738018035888, + "step": 15780, + "token_acc": 0.9365898239084509 + }, + { + "epoch": 0.9797653776922599, + "grad_norm": 0.7875742316246033, + "learning_rate": 1.1190415752025951e-08, + "loss": 0.2237471580505371, + "step": 15785, + "token_acc": 0.9352756535016153 + }, + { + "epoch": 0.9800757246601701, + "grad_norm": 0.8246194124221802, + "learning_rate": 1.0849907108476665e-08, + "loss": 0.2250540018081665, + "step": 15790, + "token_acc": 0.9359832774810046 + }, + { + "epoch": 0.9803860716280802, + "grad_norm": 0.8452962636947632, + "learning_rate": 1.0514653790395512e-08, + "loss": 0.22398147583007813, + "step": 15795, + "token_acc": 0.9353874795113758 + }, + { + "epoch": 0.9806964185959903, + "grad_norm": 0.8405020236968994, + "learning_rate": 1.018465615092057e-08, + "loss": 0.22321243286132814, + "step": 15800, + "token_acc": 0.9356602326899357 + }, + { + "epoch": 0.9810067655639004, + "grad_norm": 0.8672464489936829, + "learning_rate": 9.859914537655446e-09, + "loss": 0.2298654317855835, + "step": 15805, + "token_acc": 0.9343468103271408 + }, + { + "epoch": 0.9813171125318105, + "grad_norm": 0.828220009803772, + "learning_rate": 9.540429292667075e-09, + "loss": 0.2228759765625, + "step": 15810, + "token_acc": 0.9365199004809199 + }, + { + "epoch": 0.9816274594997206, + "grad_norm": 0.8338759541511536, + "learning_rate": 9.226200752485704e-09, + "loss": 0.22192158699035644, + "step": 15815, + "token_acc": 0.9357977166662272 + }, + { + "epoch": 0.9819378064676308, + "grad_norm": 0.8527135848999023, + "learning_rate": 8.917229248103787e-09, + "loss": 0.22875962257385254, + "step": 15820, + "token_acc": 0.9344708682494577 + }, + { + "epoch": 0.982248153435541, + "grad_norm": 0.9176541566848755, + "learning_rate": 8.613515104977099e-09, + "loss": 0.22667813301086426, + "step": 15825, + "token_acc": 0.9335897300577882 + }, + { + "epoch": 0.982558500403451, + "grad_norm": 0.8127972483634949, + "learning_rate": 8.315058643023066e-09, + "loss": 0.22446818351745607, + "step": 15830, + "token_acc": 0.9348869133651441 + }, + { + "epoch": 0.9828688473713612, + "grad_norm": 0.8317758440971375, + "learning_rate": 8.021860176620766e-09, + "loss": 0.22623372077941895, + "step": 15835, + "token_acc": 0.9345661169296561 + }, + { + "epoch": 0.9831791943392713, + "grad_norm": 0.8076893091201782, + "learning_rate": 7.733920014612039e-09, + "loss": 0.22411203384399414, + "step": 15840, + "token_acc": 0.935538294606468 + }, + { + "epoch": 0.9834895413071815, + "grad_norm": 0.9066287875175476, + "learning_rate": 7.451238460297605e-09, + "loss": 0.22220664024353026, + "step": 15845, + "token_acc": 0.9351101578222646 + }, + { + "epoch": 0.9837998882750916, + "grad_norm": 0.8255732655525208, + "learning_rate": 7.173815811440943e-09, + "loss": 0.2199860095977783, + "step": 15850, + "token_acc": 0.9370451527224436 + }, + { + "epoch": 0.9841102352430017, + "grad_norm": 0.879515528678894, + "learning_rate": 6.90165236026552e-09, + "loss": 0.22831313610076903, + "step": 15855, + "token_acc": 0.9346132331788397 + }, + { + "epoch": 0.9844205822109118, + "grad_norm": 0.7682735323905945, + "learning_rate": 6.634748393454793e-09, + "loss": 0.21818408966064454, + "step": 15860, + "token_acc": 0.9369266055045872 + }, + { + "epoch": 0.9847309291788219, + "grad_norm": 0.8491484522819519, + "learning_rate": 6.373104192152202e-09, + "loss": 0.22478189468383789, + "step": 15865, + "token_acc": 0.9352842653763841 + }, + { + "epoch": 0.9850412761467321, + "grad_norm": 0.8217035531997681, + "learning_rate": 6.1167200319611764e-09, + "loss": 0.2260317325592041, + "step": 15870, + "token_acc": 0.9344038378252324 + }, + { + "epoch": 0.9853516231146422, + "grad_norm": 0.8207763433456421, + "learning_rate": 5.865596182943467e-09, + "loss": 0.22303457260131837, + "step": 15875, + "token_acc": 0.9353655576041099 + }, + { + "epoch": 0.9856619700825523, + "grad_norm": 0.7968692779541016, + "learning_rate": 5.619732909621367e-09, + "loss": 0.22384588718414306, + "step": 15880, + "token_acc": 0.9358533791523482 + }, + { + "epoch": 0.9859723170504624, + "grad_norm": 0.8354984521865845, + "learning_rate": 5.379130470975491e-09, + "loss": 0.22613003253936767, + "step": 15885, + "token_acc": 0.9345577147222094 + }, + { + "epoch": 0.9862826640183725, + "grad_norm": 0.860225260257721, + "learning_rate": 5.14378912044311e-09, + "loss": 0.22345328330993652, + "step": 15890, + "token_acc": 0.9360545122153897 + }, + { + "epoch": 0.9865930109862827, + "grad_norm": 0.9035174250602722, + "learning_rate": 4.913709105923148e-09, + "loss": 0.22886548042297364, + "step": 15895, + "token_acc": 0.9336743471988777 + }, + { + "epoch": 0.9869033579541928, + "grad_norm": 0.8488447070121765, + "learning_rate": 4.688890669768964e-09, + "loss": 0.22817702293395997, + "step": 15900, + "token_acc": 0.9341405439478903 + }, + { + "epoch": 0.9872137049221029, + "grad_norm": 0.8325760960578918, + "learning_rate": 4.469334048793905e-09, + "loss": 0.220343017578125, + "step": 15905, + "token_acc": 0.9374274849604647 + }, + { + "epoch": 0.987524051890013, + "grad_norm": 0.9522079825401306, + "learning_rate": 4.2550394742679746e-09, + "loss": 0.21856536865234374, + "step": 15910, + "token_acc": 0.9375141393306274 + }, + { + "epoch": 0.9878343988579231, + "grad_norm": 0.8007991313934326, + "learning_rate": 4.046007171918942e-09, + "loss": 0.22292826175689698, + "step": 15915, + "token_acc": 0.9359938664093379 + }, + { + "epoch": 0.9881447458258332, + "grad_norm": 0.8300456404685974, + "learning_rate": 3.8422373619312335e-09, + "loss": 0.2224909782409668, + "step": 15920, + "token_acc": 0.9349102117394801 + }, + { + "epoch": 0.9884550927937434, + "grad_norm": 0.827904462814331, + "learning_rate": 3.643730258944822e-09, + "loss": 0.21944031715393067, + "step": 15925, + "token_acc": 0.9362928532976195 + }, + { + "epoch": 0.9887654397616535, + "grad_norm": 0.8470441699028015, + "learning_rate": 3.4504860720585566e-09, + "loss": 0.229913592338562, + "step": 15930, + "token_acc": 0.9339963356927254 + }, + { + "epoch": 0.9890757867295636, + "grad_norm": 0.8030912280082703, + "learning_rate": 3.2625050048262773e-09, + "loss": 0.22981815338134765, + "step": 15935, + "token_acc": 0.9345039783639713 + }, + { + "epoch": 0.9893861336974737, + "grad_norm": 0.9738321900367737, + "learning_rate": 3.0797872552573715e-09, + "loss": 0.22450294494628906, + "step": 15940, + "token_acc": 0.9355386541637807 + }, + { + "epoch": 0.9896964806653838, + "grad_norm": 0.8059338331222534, + "learning_rate": 2.902333015817327e-09, + "loss": 0.22481191158294678, + "step": 15945, + "token_acc": 0.9348184709593127 + }, + { + "epoch": 0.9900068276332941, + "grad_norm": 0.8449530601501465, + "learning_rate": 2.730142473429398e-09, + "loss": 0.22862575054168702, + "step": 15950, + "token_acc": 0.9343673523381176 + }, + { + "epoch": 0.9903171746012042, + "grad_norm": 0.8513461351394653, + "learning_rate": 2.563215809469055e-09, + "loss": 0.22250065803527833, + "step": 15955, + "token_acc": 0.9365558111323969 + }, + { + "epoch": 0.9906275215691143, + "grad_norm": 0.8042654395103455, + "learning_rate": 2.401553199769535e-09, + "loss": 0.2216949462890625, + "step": 15960, + "token_acc": 0.9367010873555699 + }, + { + "epoch": 0.9909378685370244, + "grad_norm": 0.8199389576911926, + "learning_rate": 2.2451548146179558e-09, + "loss": 0.22191257476806642, + "step": 15965, + "token_acc": 0.9364831038798498 + }, + { + "epoch": 0.9912482155049345, + "grad_norm": 1.0518592596054077, + "learning_rate": 2.094020818756426e-09, + "loss": 0.22218399047851561, + "step": 15970, + "token_acc": 0.9361805101373447 + }, + { + "epoch": 0.9915585624728447, + "grad_norm": 0.8628588318824768, + "learning_rate": 1.9481513713826004e-09, + "loss": 0.22758188247680664, + "step": 15975, + "token_acc": 0.9348076846585259 + }, + { + "epoch": 0.9918689094407548, + "grad_norm": 0.8437245488166809, + "learning_rate": 1.8075466261480156e-09, + "loss": 0.2203207492828369, + "step": 15980, + "token_acc": 0.9361688036452263 + }, + { + "epoch": 0.9921792564086649, + "grad_norm": 0.8273714184761047, + "learning_rate": 1.6722067311586431e-09, + "loss": 0.2229374408721924, + "step": 15985, + "token_acc": 0.9365461955244624 + }, + { + "epoch": 0.992489603376575, + "grad_norm": 0.8367292881011963, + "learning_rate": 1.542131828974891e-09, + "loss": 0.2309025526046753, + "step": 15990, + "token_acc": 0.934344579277237 + }, + { + "epoch": 0.9927999503444851, + "grad_norm": 0.8372355103492737, + "learning_rate": 1.417322056611048e-09, + "loss": 0.22440309524536134, + "step": 15995, + "token_acc": 0.934622547102914 + }, + { + "epoch": 0.9931102973123953, + "grad_norm": 0.8935984969139099, + "learning_rate": 1.297777545536394e-09, + "loss": 0.23045065402984619, + "step": 16000, + "token_acc": 0.9341573850679181 + }, + { + "epoch": 0.9931102973123953, + "eval_loss": 0.23542769253253937, + "eval_runtime": 101.9756, + "eval_samples_per_second": 196.125, + "eval_steps_per_second": 0.265, + "eval_token_acc": 0.932491861737979, + "step": 16000 + } + ], + "logging_steps": 5, + "max_steps": 16111, + "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": 2.4833701620433814e+19, + "train_batch_size": 96, + "trial_name": null, + "trial_params": null +}