diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,5061 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.6454566185356987,
+  "eval_steps": 500,
+  "global_step": 72000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.000896467525744026,
+      "grad_norm": 0.5214600563049316,
+      "learning_rate": 0.0004999990104957794,
+      "loss": 4.1615,
+      "step": 100
+    },
+    {
+      "epoch": 0.001792935051488052,
+      "grad_norm": 0.5064386129379272,
+      "learning_rate": 0.0004999960380279981,
+      "loss": 3.6511,
+      "step": 200
+    },
+    {
+      "epoch": 0.002689402577232078,
+      "grad_norm": 0.5299622416496277,
+      "learning_rate": 0.0004999910826192423,
+      "loss": 3.537,
+      "step": 300
+    },
+    {
+      "epoch": 0.003585870102976104,
+      "grad_norm": 0.49445027112960815,
+      "learning_rate": 0.0004999841443088176,
+      "loss": 3.4766,
+      "step": 400
+    },
+    {
+      "epoch": 0.00448233762872013,
+      "grad_norm": 0.5226134061813354,
+      "learning_rate": 0.0004999752231517578,
+      "loss": 3.4303,
+      "step": 500
+    },
+    {
+      "epoch": 0.005378805154464156,
+      "grad_norm": 0.4533003568649292,
+      "learning_rate": 0.0004999643192188245,
+      "loss": 3.387,
+      "step": 600
+    },
+    {
+      "epoch": 0.0062752726802081825,
+      "grad_norm": 0.4646398723125458,
+      "learning_rate": 0.0004999514325965069,
+      "loss": 3.3508,
+      "step": 700
+    },
+    {
+      "epoch": 0.007171740205952208,
+      "grad_norm": 0.47208401560783386,
+      "learning_rate": 0.0004999365633870197,
+      "loss": 3.3181,
+      "step": 800
+    },
+    {
+      "epoch": 0.008068207731696234,
+      "grad_norm": 0.45212841033935547,
+      "learning_rate": 0.0004999197117083043,
+      "loss": 3.2928,
+      "step": 900
+    },
+    {
+      "epoch": 0.00896467525744026,
+      "grad_norm": 0.4529814124107361,
+      "learning_rate": 0.0004999008776940261,
+      "loss": 3.2734,
+      "step": 1000
+    },
+    {
+      "epoch": 0.009861142783184286,
+      "grad_norm": 0.4323786795139313,
+      "learning_rate": 0.0004998800614935745,
+      "loss": 3.2607,
+      "step": 1100
+    },
+    {
+      "epoch": 0.010757610308928313,
+      "grad_norm": 0.45383089780807495,
+      "learning_rate": 0.0004998572632720614,
+      "loss": 3.2526,
+      "step": 1200
+    },
+    {
+      "epoch": 0.011654077834672339,
+      "grad_norm": 0.4236689805984497,
+      "learning_rate": 0.0004998324832103195,
+      "loss": 3.216,
+      "step": 1300
+    },
+    {
+      "epoch": 0.012550545360416365,
+      "grad_norm": 0.3684336841106415,
+      "learning_rate": 0.0004998057215049018,
+      "loss": 3.2024,
+      "step": 1400
+    },
+    {
+      "epoch": 0.013447012886160391,
+      "grad_norm": 0.39953121542930603,
+      "learning_rate": 0.000499776978368079,
+      "loss": 3.1853,
+      "step": 1500
+    },
+    {
+      "epoch": 0.014343480411904416,
+      "grad_norm": 0.3904390335083008,
+      "learning_rate": 0.0004997462540278386,
+      "loss": 3.1799,
+      "step": 1600
+    },
+    {
+      "epoch": 0.015239947937648442,
+      "grad_norm": 0.3936985433101654,
+      "learning_rate": 0.0004997135487278825,
+      "loss": 3.1613,
+      "step": 1700
+    },
+    {
+      "epoch": 0.016136415463392468,
+      "grad_norm": 0.36848315596580505,
+      "learning_rate": 0.0004996788627276258,
+      "loss": 3.1438,
+      "step": 1800
+    },
+    {
+      "epoch": 0.017032882989136494,
+      "grad_norm": 0.3514796495437622,
+      "learning_rate": 0.0004996421963021939,
+      "loss": 3.1358,
+      "step": 1900
+    },
+    {
+      "epoch": 0.01792935051488052,
+      "grad_norm": 0.3697972893714905,
+      "learning_rate": 0.0004996035497424208,
+      "loss": 3.1413,
+      "step": 2000
+    },
+    {
+      "epoch": 0.018825818040624547,
+      "grad_norm": 0.3728257715702057,
+      "learning_rate": 0.0004995629233548467,
+      "loss": 3.1187,
+      "step": 2100
+    },
+    {
+      "epoch": 0.019722285566368573,
+      "grad_norm": 0.3539767861366272,
+      "learning_rate": 0.0004995203174617158,
+      "loss": 3.1038,
+      "step": 2200
+    },
+    {
+      "epoch": 0.0206187530921126,
+      "grad_norm": 0.3395736813545227,
+      "learning_rate": 0.0004994757324009732,
+      "loss": 3.0942,
+      "step": 2300
+    },
+    {
+      "epoch": 0.021515220617856625,
+      "grad_norm": 0.36709773540496826,
+      "learning_rate": 0.0004994291685262627,
+      "loss": 3.0961,
+      "step": 2400
+    },
+    {
+      "epoch": 0.02241168814360065,
+      "grad_norm": 0.32454609870910645,
+      "learning_rate": 0.0004993806262069239,
+      "loss": 3.0847,
+      "step": 2500
+    },
+    {
+      "epoch": 0.023308155669344677,
+      "grad_norm": 0.33812129497528076,
+      "learning_rate": 0.000499330105827989,
+      "loss": 3.0767,
+      "step": 2600
+    },
+    {
+      "epoch": 0.024204623195088704,
+      "grad_norm": 0.33684587478637695,
+      "learning_rate": 0.0004992776077901801,
+      "loss": 3.0641,
+      "step": 2700
+    },
+    {
+      "epoch": 0.02510109072083273,
+      "grad_norm": 0.3293500542640686,
+      "learning_rate": 0.0004992231325099059,
+      "loss": 3.0715,
+      "step": 2800
+    },
+    {
+      "epoch": 0.025997558246576756,
+      "grad_norm": 0.3323623836040497,
+      "learning_rate": 0.0004991666804192582,
+      "loss": 3.055,
+      "step": 2900
+    },
+    {
+      "epoch": 0.026894025772320782,
+      "grad_norm": 0.3230840861797333,
+      "learning_rate": 0.0004991082519660089,
+      "loss": 3.0366,
+      "step": 3000
+    },
+    {
+      "epoch": 0.02779049329806481,
+      "grad_norm": 0.31436941027641296,
+      "learning_rate": 0.0004990478476136059,
+      "loss": 3.035,
+      "step": 3100
+    },
+    {
+      "epoch": 0.02868696082380883,
+      "grad_norm": 0.32258903980255127,
+      "learning_rate": 0.0004989854678411699,
+      "loss": 3.0321,
+      "step": 3200
+    },
+    {
+      "epoch": 0.029583428349552857,
+      "grad_norm": 0.3148656487464905,
+      "learning_rate": 0.0004989211131434904,
+      "loss": 3.0146,
+      "step": 3300
+    },
+    {
+      "epoch": 0.030479895875296883,
+      "grad_norm": 0.31875699758529663,
+      "learning_rate": 0.0004988547840310217,
+      "loss": 3.0214,
+      "step": 3400
+    },
+    {
+      "epoch": 0.03137636340104091,
+      "grad_norm": 0.3068808913230896,
+      "learning_rate": 0.0004987864810298791,
+      "loss": 3.0135,
+      "step": 3500
+    },
+    {
+      "epoch": 0.032272830926784936,
+      "grad_norm": 0.3052406311035156,
+      "learning_rate": 0.0004987162046818343,
+      "loss": 3.0042,
+      "step": 3600
+    },
+    {
+      "epoch": 0.033169298452528966,
+      "grad_norm": 0.3155302405357361,
+      "learning_rate": 0.0004986439555443118,
+      "loss": 3.008,
+      "step": 3700
+    },
+    {
+      "epoch": 0.03406576597827299,
+      "grad_norm": 0.31643468141555786,
+      "learning_rate": 0.0004985697341903834,
+      "loss": 2.9999,
+      "step": 3800
+    },
+    {
+      "epoch": 0.03496223350401702,
+      "grad_norm": 0.29758599400520325,
+      "learning_rate": 0.0004984935412087651,
+      "loss": 2.9957,
+      "step": 3900
+    },
+    {
+      "epoch": 0.03585870102976104,
+      "grad_norm": 0.30581027269363403,
+      "learning_rate": 0.0004984153772038109,
+      "loss": 2.9873,
+      "step": 4000
+    },
+    {
+      "epoch": 0.03675516855550507,
+      "grad_norm": 0.30977991223335266,
+      "learning_rate": 0.0004983352427955092,
+      "loss": 2.9864,
+      "step": 4100
+    },
+    {
+      "epoch": 0.03765163608124909,
+      "grad_norm": 0.30028942227363586,
+      "learning_rate": 0.0004982531386194774,
+      "loss": 2.9801,
+      "step": 4200
+    },
+    {
+      "epoch": 0.038548103606993116,
+      "grad_norm": 0.2785912752151489,
+      "learning_rate": 0.0004981690653269564,
+      "loss": 2.9707,
+      "step": 4300
+    },
+    {
+      "epoch": 0.039444571132737145,
+      "grad_norm": 0.30985763669013977,
+      "learning_rate": 0.0004980830235848066,
+      "loss": 2.9673,
+      "step": 4400
+    },
+    {
+      "epoch": 0.04034103865848117,
+      "grad_norm": 0.2833133339881897,
+      "learning_rate": 0.0004979950140755015,
+      "loss": 2.9683,
+      "step": 4500
+    },
+    {
+      "epoch": 0.0412375061842252,
+      "grad_norm": 0.27662980556488037,
+      "learning_rate": 0.0004979050374971228,
+      "loss": 2.9525,
+      "step": 4600
+    },
+    {
+      "epoch": 0.04213397370996922,
+      "grad_norm": 0.284396231174469,
+      "learning_rate": 0.0004978130945633548,
+      "loss": 2.9643,
+      "step": 4700
+    },
+    {
+      "epoch": 0.04303044123571325,
+      "grad_norm": 0.2856046259403229,
+      "learning_rate": 0.0004977191860034787,
+      "loss": 2.9566,
+      "step": 4800
+    },
+    {
+      "epoch": 0.04392690876145727,
+      "grad_norm": 0.27780982851982117,
+      "learning_rate": 0.0004976233125623668,
+      "loss": 2.9501,
+      "step": 4900
+    },
+    {
+      "epoch": 0.0448233762872013,
+      "grad_norm": 0.282685786485672,
+      "learning_rate": 0.0004975254750004767,
+      "loss": 2.943,
+      "step": 5000
+    },
+    {
+      "epoch": 0.045719843812945325,
+      "grad_norm": 0.28598085045814514,
+      "learning_rate": 0.0004974256740938451,
+      "loss": 2.9507,
+      "step": 5100
+    },
+    {
+      "epoch": 0.046616311338689355,
+      "grad_norm": 0.2788136899471283,
+      "learning_rate": 0.0004973239106340817,
+      "loss": 2.9437,
+      "step": 5200
+    },
+    {
+      "epoch": 0.04751277886443338,
+      "grad_norm": 0.2748227119445801,
+      "learning_rate": 0.0004972201854283632,
+      "loss": 2.9439,
+      "step": 5300
+    },
+    {
+      "epoch": 0.04840924639017741,
+      "grad_norm": 0.27777764201164246,
+      "learning_rate": 0.0004971144992994265,
+      "loss": 2.9362,
+      "step": 5400
+    },
+    {
+      "epoch": 0.04930571391592143,
+      "grad_norm": 0.273762583732605,
+      "learning_rate": 0.0004970068530855621,
+      "loss": 2.9336,
+      "step": 5500
+    },
+    {
+      "epoch": 0.05020218144166546,
+      "grad_norm": 0.27693507075309753,
+      "learning_rate": 0.0004968972476406079,
+      "loss": 2.9285,
+      "step": 5600
+    },
+    {
+      "epoch": 0.05109864896740948,
+      "grad_norm": 0.26276636123657227,
+      "learning_rate": 0.0004967856838339424,
+      "loss": 2.9161,
+      "step": 5700
+    },
+    {
+      "epoch": 0.05199511649315351,
+      "grad_norm": 0.26518794894218445,
+      "learning_rate": 0.0004966721625504772,
+      "loss": 2.9268,
+      "step": 5800
+    },
+    {
+      "epoch": 0.052891584018897535,
+      "grad_norm": 0.2554951012134552,
+      "learning_rate": 0.0004965566846906506,
+      "loss": 2.922,
+      "step": 5900
+    },
+    {
+      "epoch": 0.053788051544641564,
+      "grad_norm": 0.25945886969566345,
+      "learning_rate": 0.0004964392511704205,
+      "loss": 2.9179,
+      "step": 6000
+    },
+    {
+      "epoch": 0.05468451907038559,
+      "grad_norm": 0.2560843527317047,
+      "learning_rate": 0.0004963198629212563,
+      "loss": 2.9127,
+      "step": 6100
+    },
+    {
+      "epoch": 0.05558098659612962,
+      "grad_norm": 0.29079943895339966,
+      "learning_rate": 0.0004961985208901326,
+      "loss": 2.9097,
+      "step": 6200
+    },
+    {
+      "epoch": 0.05647745412187364,
+      "grad_norm": 0.2522130608558655,
+      "learning_rate": 0.0004960752260395208,
+      "loss": 2.918,
+      "step": 6300
+    },
+    {
+      "epoch": 0.05737392164761766,
+      "grad_norm": 0.26022017002105713,
+      "learning_rate": 0.0004959499793473822,
+      "loss": 2.9088,
+      "step": 6400
+    },
+    {
+      "epoch": 0.05827038917336169,
+      "grad_norm": 0.25249192118644714,
+      "learning_rate": 0.0004958227818071597,
+      "loss": 2.8943,
+      "step": 6500
+    },
+    {
+      "epoch": 0.059166856699105715,
+      "grad_norm": 0.24774658679962158,
+      "learning_rate": 0.0004956936344277703,
+      "loss": 2.9003,
+      "step": 6600
+    },
+    {
+      "epoch": 0.060063324224849744,
+      "grad_norm": 0.243002787232399,
+      "learning_rate": 0.0004955625382335964,
+      "loss": 2.8918,
+      "step": 6700
+    },
+    {
+      "epoch": 0.06095979175059377,
+      "grad_norm": 0.2450036108493805,
+      "learning_rate": 0.0004954294942644788,
+      "loss": 2.8959,
+      "step": 6800
+    },
+    {
+      "epoch": 0.0618562592763378,
+      "grad_norm": 0.24282772839069366,
+      "learning_rate": 0.0004952945035757076,
+      "loss": 2.8954,
+      "step": 6900
+    },
+    {
+      "epoch": 0.06275272680208183,
+      "grad_norm": 0.24141670763492584,
+      "learning_rate": 0.0004951575672380139,
+      "loss": 2.8933,
+      "step": 7000
+    },
+    {
+      "epoch": 0.06364919432782584,
+      "grad_norm": 0.23969806730747223,
+      "learning_rate": 0.0004950186863375616,
+      "loss": 2.8874,
+      "step": 7100
+    },
+    {
+      "epoch": 0.06454566185356987,
+      "grad_norm": 0.24008285999298096,
+      "learning_rate": 0.0004948778619759389,
+      "loss": 2.8894,
+      "step": 7200
+    },
+    {
+      "epoch": 0.0654421293793139,
+      "grad_norm": 0.24920688569545746,
+      "learning_rate": 0.0004947350952701489,
+      "loss": 2.8913,
+      "step": 7300
+    },
+    {
+      "epoch": 0.06633859690505793,
+      "grad_norm": 0.23809005320072174,
+      "learning_rate": 0.0004945903873526013,
+      "loss": 2.8832,
+      "step": 7400
+    },
+    {
+      "epoch": 0.06723506443080195,
+      "grad_norm": 0.2459540218114853,
+      "learning_rate": 0.0004944437393711034,
+      "loss": 2.8836,
+      "step": 7500
+    },
+    {
+      "epoch": 0.06813153195654598,
+      "grad_norm": 0.2407006472349167,
+      "learning_rate": 0.0004942951524888509,
+      "loss": 2.8717,
+      "step": 7600
+    },
+    {
+      "epoch": 0.06902799948229,
+      "grad_norm": 0.22954203188419342,
+      "learning_rate": 0.0004941446278844185,
+      "loss": 2.8766,
+      "step": 7700
+    },
+    {
+      "epoch": 0.06992446700803404,
+      "grad_norm": 0.2322009801864624,
+      "learning_rate": 0.0004939921667517505,
+      "loss": 2.8691,
+      "step": 7800
+    },
+    {
+      "epoch": 0.07082093453377805,
+      "grad_norm": 0.2393633872270584,
+      "learning_rate": 0.0004938377703001521,
+      "loss": 2.874,
+      "step": 7900
+    },
+    {
+      "epoch": 0.07171740205952208,
+      "grad_norm": 0.23171070218086243,
+      "learning_rate": 0.0004936814397542788,
+      "loss": 2.8763,
+      "step": 8000
+    },
+    {
+      "epoch": 0.07261386958526611,
+      "grad_norm": 0.23274166882038116,
+      "learning_rate": 0.0004935231763541272,
+      "loss": 2.8668,
+      "step": 8100
+    },
+    {
+      "epoch": 0.07351033711101014,
+      "grad_norm": 0.23799176514148712,
+      "learning_rate": 0.0004933629813550251,
+      "loss": 2.8675,
+      "step": 8200
+    },
+    {
+      "epoch": 0.07440680463675416,
+      "grad_norm": 0.2373514026403427,
+      "learning_rate": 0.0004932008560276217,
+      "loss": 2.8648,
+      "step": 8300
+    },
+    {
+      "epoch": 0.07530327216249819,
+      "grad_norm": 0.22208142280578613,
+      "learning_rate": 0.0004930368016578769,
+      "loss": 2.8605,
+      "step": 8400
+    },
+    {
+      "epoch": 0.07619973968824222,
+      "grad_norm": 0.230060413479805,
+      "learning_rate": 0.0004928708195470521,
+      "loss": 2.8626,
+      "step": 8500
+    },
+    {
+      "epoch": 0.07709620721398623,
+      "grad_norm": 0.22975729405879974,
+      "learning_rate": 0.000492702911011699,
+      "loss": 2.8612,
+      "step": 8600
+    },
+    {
+      "epoch": 0.07799267473973026,
+      "grad_norm": 0.2305164784193039,
+      "learning_rate": 0.0004925330773836496,
+      "loss": 2.8615,
+      "step": 8700
+    },
+    {
+      "epoch": 0.07888914226547429,
+      "grad_norm": 0.22835880517959595,
+      "learning_rate": 0.0004923613200100054,
+      "loss": 2.8563,
+      "step": 8800
+    },
+    {
+      "epoch": 0.07978560979121832,
+      "grad_norm": 0.23437829315662384,
+      "learning_rate": 0.000492187640253127,
+      "loss": 2.8429,
+      "step": 8900
+    },
+    {
+      "epoch": 0.08068207731696234,
+      "grad_norm": 0.21821853518486023,
+      "learning_rate": 0.0004920120394906231,
+      "loss": 2.8556,
+      "step": 9000
+    },
+    {
+      "epoch": 0.08157854484270637,
+      "grad_norm": 0.22391639649868011,
+      "learning_rate": 0.0004918345191153395,
+      "loss": 2.8439,
+      "step": 9100
+    },
+    {
+      "epoch": 0.0824750123684504,
+      "grad_norm": 0.22824177145957947,
+      "learning_rate": 0.0004916550805353483,
+      "loss": 2.851,
+      "step": 9200
+    },
+    {
+      "epoch": 0.08337147989419443,
+      "grad_norm": 0.22843202948570251,
+      "learning_rate": 0.0004914737251739363,
+      "loss": 2.8493,
+      "step": 9300
+    },
+    {
+      "epoch": 0.08426794741993844,
+      "grad_norm": 0.22368155419826508,
+      "learning_rate": 0.0004912904544695944,
+      "loss": 2.8447,
+      "step": 9400
+    },
+    {
+      "epoch": 0.08516441494568247,
+      "grad_norm": 0.22078120708465576,
+      "learning_rate": 0.0004911052698760055,
+      "loss": 2.8498,
+      "step": 9500
+    },
+    {
+      "epoch": 0.0860608824714265,
+      "grad_norm": 0.21971289813518524,
+      "learning_rate": 0.0004909181728620333,
+      "loss": 2.846,
+      "step": 9600
+    },
+    {
+      "epoch": 0.08695734999717053,
+      "grad_norm": 0.217596173286438,
+      "learning_rate": 0.0004907291649117109,
+      "loss": 2.8415,
+      "step": 9700
+    },
+    {
+      "epoch": 0.08785381752291455,
+      "grad_norm": 0.2179151475429535,
+      "learning_rate": 0.0004905382475242282,
+      "loss": 2.8454,
+      "step": 9800
+    },
+    {
+      "epoch": 0.08875028504865858,
+      "grad_norm": 0.2107924073934555,
+      "learning_rate": 0.0004903454222139212,
+      "loss": 2.8413,
+      "step": 9900
+    },
+    {
+      "epoch": 0.0896467525744026,
+      "grad_norm": 0.20650827884674072,
+      "learning_rate": 0.0004901506905102585,
+      "loss": 2.825,
+      "step": 10000
+    },
+    {
+      "epoch": 0.09054322010014663,
+      "grad_norm": 0.2175520658493042,
+      "learning_rate": 0.0004899540539578308,
+      "loss": 2.8297,
+      "step": 10100
+    },
+    {
+      "epoch": 0.09143968762589065,
+      "grad_norm": 0.21441052854061127,
+      "learning_rate": 0.0004897555141163376,
+      "loss": 2.8265,
+      "step": 10200
+    },
+    {
+      "epoch": 0.09233615515163468,
+      "grad_norm": 0.21620461344718933,
+      "learning_rate": 0.0004895550725605748,
+      "loss": 2.8266,
+      "step": 10300
+    },
+    {
+      "epoch": 0.09323262267737871,
+      "grad_norm": 0.21449413895606995,
+      "learning_rate": 0.0004893527308804227,
+      "loss": 2.8308,
+      "step": 10400
+    },
+    {
+      "epoch": 0.09412909020312274,
+      "grad_norm": 0.2166638970375061,
+      "learning_rate": 0.0004891484906808332,
+      "loss": 2.8243,
+      "step": 10500
+    },
+    {
+      "epoch": 0.09502555772886676,
+      "grad_norm": 0.2123963087797165,
+      "learning_rate": 0.0004889423535818172,
+      "loss": 2.8224,
+      "step": 10600
+    },
+    {
+      "epoch": 0.09592202525461078,
+      "grad_norm": 0.20958571135997772,
+      "learning_rate": 0.0004887343212184311,
+      "loss": 2.8235,
+      "step": 10700
+    },
+    {
+      "epoch": 0.09681849278035481,
+      "grad_norm": 0.20609615743160248,
+      "learning_rate": 0.0004885243952407648,
+      "loss": 2.8304,
+      "step": 10800
+    },
+    {
+      "epoch": 0.09771496030609883,
+      "grad_norm": 0.20418639481067657,
+      "learning_rate": 0.000488312577313928,
+      "loss": 2.8251,
+      "step": 10900
+    },
+    {
+      "epoch": 0.09861142783184286,
+      "grad_norm": 0.20903444290161133,
+      "learning_rate": 0.00048809886911803705,
+      "loss": 2.8244,
+      "step": 11000
+    },
+    {
+      "epoch": 0.09950789535758689,
+      "grad_norm": 0.20328710973262787,
+      "learning_rate": 0.00048788327234820187,
+      "loss": 2.8212,
+      "step": 11100
+    },
+    {
+      "epoch": 0.10040436288333092,
+      "grad_norm": 0.21136407554149628,
+      "learning_rate": 0.0004876657887145121,
+      "loss": 2.8116,
+      "step": 11200
+    },
+    {
+      "epoch": 0.10130083040907494,
+      "grad_norm": 0.20518264174461365,
+      "learning_rate": 0.00048744641994202397,
+      "loss": 2.8183,
+      "step": 11300
+    },
+    {
+      "epoch": 0.10219729793481896,
+      "grad_norm": 0.20106618106365204,
+      "learning_rate": 0.0004872251677707464,
+      "loss": 2.8147,
+      "step": 11400
+    },
+    {
+      "epoch": 0.103093765460563,
+      "grad_norm": 0.20347937941551208,
+      "learning_rate": 0.00048700203395562715,
+      "loss": 2.8081,
+      "step": 11500
+    },
+    {
+      "epoch": 0.10399023298630702,
+      "grad_norm": 0.1995743364095688,
+      "learning_rate": 0.000486777020266539,
+      "loss": 2.8204,
+      "step": 11600
+    },
+    {
+      "epoch": 0.10488670051205104,
+      "grad_norm": 0.20354920625686646,
+      "learning_rate": 0.00048655012848826574,
+      "loss": 2.8077,
+      "step": 11700
+    },
+    {
+      "epoch": 0.10578316803779507,
+      "grad_norm": 0.1998705416917801,
+      "learning_rate": 0.00048632136042048784,
+      "loss": 2.8129,
+      "step": 11800
+    },
+    {
+      "epoch": 0.1066796355635391,
+      "grad_norm": 0.19472312927246094,
+      "learning_rate": 0.00048609071787776846,
+      "loss": 2.8141,
+      "step": 11900
+    },
+    {
+      "epoch": 0.10757610308928313,
+      "grad_norm": 0.1991991549730301,
+      "learning_rate": 0.0004858582026895387,
+      "loss": 2.8148,
+      "step": 12000
+    },
+    {
+      "epoch": 0.10847257061502714,
+      "grad_norm": 0.20118767023086548,
+      "learning_rate": 0.0004856238167000834,
+      "loss": 2.8126,
+      "step": 12100
+    },
+    {
+      "epoch": 0.10936903814077117,
+      "grad_norm": 0.20079652965068817,
+      "learning_rate": 0.0004853875617685263,
+      "loss": 2.8128,
+      "step": 12200
+    },
+    {
+      "epoch": 0.1102655056665152,
+      "grad_norm": 0.20117126405239105,
+      "learning_rate": 0.0004851494397688154,
+      "loss": 2.8065,
+      "step": 12300
+    },
+    {
+      "epoch": 0.11116197319225923,
+      "grad_norm": 0.19336390495300293,
+      "learning_rate": 0.00048490945258970833,
+      "loss": 2.8034,
+      "step": 12400
+    },
+    {
+      "epoch": 0.11205844071800325,
+      "grad_norm": 0.20018301904201508,
+      "learning_rate": 0.00048466760213475665,
+      "loss": 2.8033,
+      "step": 12500
+    },
+    {
+      "epoch": 0.11295490824374728,
+      "grad_norm": 0.19468124210834503,
+      "learning_rate": 0.00048442389032229163,
+      "loss": 2.7959,
+      "step": 12600
+    },
+    {
+      "epoch": 0.11385137576949131,
+      "grad_norm": 0.1926909238100052,
+      "learning_rate": 0.00048417831908540836,
+      "loss": 2.7999,
+      "step": 12700
+    },
+    {
+      "epoch": 0.11474784329523532,
+      "grad_norm": 0.19469407200813293,
+      "learning_rate": 0.00048393089037195085,
+      "loss": 2.8011,
+      "step": 12800
+    },
+    {
+      "epoch": 0.11564431082097935,
+      "grad_norm": 0.19478073716163635,
+      "learning_rate": 0.0004836816061444964,
+      "loss": 2.8136,
+      "step": 12900
+    },
+    {
+      "epoch": 0.11654077834672338,
+      "grad_norm": 0.19283322989940643,
+      "learning_rate": 0.00048343046838033975,
+      "loss": 2.7914,
+      "step": 13000
+    },
+    {
+      "epoch": 0.11743724587246741,
+      "grad_norm": 0.19198790192604065,
+      "learning_rate": 0.0004831774790714781,
+      "loss": 2.8133,
+      "step": 13100
+    },
+    {
+      "epoch": 0.11833371339821143,
+      "grad_norm": 0.18972335755825043,
+      "learning_rate": 0.0004829226402245948,
+      "loss": 2.7983,
+      "step": 13200
+    },
+    {
+      "epoch": 0.11923018092395546,
+      "grad_norm": 0.18818192183971405,
+      "learning_rate": 0.0004826659538610433,
+      "loss": 2.7911,
+      "step": 13300
+    },
+    {
+      "epoch": 0.12012664844969949,
+      "grad_norm": 0.19898109138011932,
+      "learning_rate": 0.0004824074220168315,
+      "loss": 2.7961,
+      "step": 13400
+    },
+    {
+      "epoch": 0.12102311597544352,
+      "grad_norm": 0.1899373084306717,
+      "learning_rate": 0.00048214704674260544,
+      "loss": 2.7923,
+      "step": 13500
+    },
+    {
+      "epoch": 0.12191958350118753,
+      "grad_norm": 0.19022135436534882,
+      "learning_rate": 0.00048188483010363315,
+      "loss": 2.7933,
+      "step": 13600
+    },
+    {
+      "epoch": 0.12281605102693156,
+      "grad_norm": 0.19756515324115753,
+      "learning_rate": 0.000481620774179788,
+      "loss": 2.7928,
+      "step": 13700
+    },
+    {
+      "epoch": 0.1237125185526756,
+      "grad_norm": 0.18114596605300903,
+      "learning_rate": 0.00048135488106553246,
+      "loss": 2.7896,
+      "step": 13800
+    },
+    {
+      "epoch": 0.12460898607841962,
+      "grad_norm": 0.19043464958667755,
+      "learning_rate": 0.0004810871528699013,
+      "loss": 2.8021,
+      "step": 13900
+    },
+    {
+      "epoch": 0.12550545360416365,
+      "grad_norm": 0.18425244092941284,
+      "learning_rate": 0.000480817591716485,
+      "loss": 2.7863,
+      "step": 14000
+    },
+    {
+      "epoch": 0.12640192112990767,
+      "grad_norm": 0.186711385846138,
+      "learning_rate": 0.00048054619974341293,
+      "loss": 2.7866,
+      "step": 14100
+    },
+    {
+      "epoch": 0.12729838865565168,
+      "grad_norm": 0.18656401336193085,
+      "learning_rate": 0.00048027297910333634,
+      "loss": 2.7888,
+      "step": 14200
+    },
+    {
+      "epoch": 0.12819485618139573,
+      "grad_norm": 0.19122888147830963,
+      "learning_rate": 0.00048000069146883305,
+      "loss": 2.7899,
+      "step": 14300
+    },
+    {
+      "epoch": 0.12909132370713974,
+      "grad_norm": 0.19195199012756348,
+      "learning_rate": 0.000479723838243038,
+      "loss": 2.7819,
+      "step": 14400
+    },
+    {
+      "epoch": 0.1299877912328838,
+      "grad_norm": 0.1867845505475998,
+      "learning_rate": 0.00047944516287311896,
+      "loss": 2.7853,
+      "step": 14500
+    },
+    {
+      "epoch": 0.1308842587586278,
+      "grad_norm": 0.18608401715755463,
+      "learning_rate": 0.0004791646675694982,
+      "loss": 2.7899,
+      "step": 14600
+    },
+    {
+      "epoch": 0.13178072628437182,
+      "grad_norm": 0.18311014771461487,
+      "learning_rate": 0.0004788823545570335,
+      "loss": 2.7835,
+      "step": 14700
+    },
+    {
+      "epoch": 0.13267719381011586,
+      "grad_norm": 0.18312698602676392,
+      "learning_rate": 0.00047859822607500046,
+      "loss": 2.7814,
+      "step": 14800
+    },
+    {
+      "epoch": 0.13357366133585988,
+      "grad_norm": 0.1876700073480606,
+      "learning_rate": 0.00047831228437707495,
+      "loss": 2.7925,
+      "step": 14900
+    },
+    {
+      "epoch": 0.1344701288616039,
+      "grad_norm": 0.18215568363666534,
+      "learning_rate": 0.0004780245317313149,
+      "loss": 2.7815,
+      "step": 15000
+    },
+    {
+      "epoch": 0.13536659638734794,
+      "grad_norm": 0.18465806543827057,
+      "learning_rate": 0.0004777349704201426,
+      "loss": 2.7845,
+      "step": 15100
+    },
+    {
+      "epoch": 0.13626306391309195,
+      "grad_norm": 0.18615123629570007,
+      "learning_rate": 0.00047744360274032616,
+      "loss": 2.7788,
+      "step": 15200
+    },
+    {
+      "epoch": 0.13715953143883597,
+      "grad_norm": 0.18580564856529236,
+      "learning_rate": 0.0004771504310029621,
+      "loss": 2.7727,
+      "step": 15300
+    },
+    {
+      "epoch": 0.13805599896458,
+      "grad_norm": 0.19075210392475128,
+      "learning_rate": 0.00047685545753345615,
+      "loss": 2.7777,
+      "step": 15400
+    },
+    {
+      "epoch": 0.13895246649032403,
+      "grad_norm": 0.17943909764289856,
+      "learning_rate": 0.00047655868467150534,
+      "loss": 2.7842,
+      "step": 15500
+    },
+    {
+      "epoch": 0.13984893401606807,
+      "grad_norm": 0.19626812636852264,
+      "learning_rate": 0.00047626011477107925,
+      "loss": 2.7736,
+      "step": 15600
+    },
+    {
+      "epoch": 0.1407454015418121,
+      "grad_norm": 0.18597416579723358,
+      "learning_rate": 0.0004759597502004014,
+      "loss": 2.7789,
+      "step": 15700
+    },
+    {
+      "epoch": 0.1416418690675561,
+      "grad_norm": 0.18150849640369415,
+      "learning_rate": 0.0004756575933419304,
+      "loss": 2.7771,
+      "step": 15800
+    },
+    {
+      "epoch": 0.14253833659330015,
+      "grad_norm": 0.17691569030284882,
+      "learning_rate": 0.00047535364659234095,
+      "loss": 2.7689,
+      "step": 15900
+    },
+    {
+      "epoch": 0.14343480411904416,
+      "grad_norm": 0.1855591982603073,
+      "learning_rate": 0.00047504791236250535,
+      "loss": 2.7755,
+      "step": 16000
+    },
+    {
+      "epoch": 0.14433127164478818,
+      "grad_norm": 0.18740758299827576,
+      "learning_rate": 0.00047474039307747354,
+      "loss": 2.7756,
+      "step": 16100
+    },
+    {
+      "epoch": 0.14522773917053222,
+      "grad_norm": 0.17985744774341583,
+      "learning_rate": 0.00047443109117645466,
+      "loss": 2.774,
+      "step": 16200
+    },
+    {
+      "epoch": 0.14612420669627624,
+      "grad_norm": 0.17564240097999573,
+      "learning_rate": 0.0004741200091127973,
+      "loss": 2.7719,
+      "step": 16300
+    },
+    {
+      "epoch": 0.14702067422202028,
+      "grad_norm": 0.17614321410655975,
+      "learning_rate": 0.00047380714935396986,
+      "loss": 2.7679,
+      "step": 16400
+    },
+    {
+      "epoch": 0.1479171417477643,
+      "grad_norm": 0.17652449011802673,
+      "learning_rate": 0.00047349251438154154,
+      "loss": 2.77,
+      "step": 16500
+    },
+    {
+      "epoch": 0.1488136092735083,
+      "grad_norm": 0.1796472817659378,
+      "learning_rate": 0.0004731761066911622,
+      "loss": 2.7652,
+      "step": 16600
+    },
+    {
+      "epoch": 0.14971007679925236,
+      "grad_norm": 0.17830465734004974,
+      "learning_rate": 0.00047285792879254274,
+      "loss": 2.7684,
+      "step": 16700
+    },
+    {
+      "epoch": 0.15060654432499637,
+      "grad_norm": 0.17882998287677765,
+      "learning_rate": 0.000472537983209435,
+      "loss": 2.7709,
+      "step": 16800
+    },
+    {
+      "epoch": 0.1515030118507404,
+      "grad_norm": 0.17409980297088623,
+      "learning_rate": 0.0004722162724796122,
+      "loss": 2.7692,
+      "step": 16900
+    },
+    {
+      "epoch": 0.15239947937648443,
+      "grad_norm": 0.1783556491136551,
+      "learning_rate": 0.00047189279915484816,
+      "loss": 2.7736,
+      "step": 17000
+    },
+    {
+      "epoch": 0.15329594690222845,
+      "grad_norm": 0.17761445045471191,
+      "learning_rate": 0.00047156756580089766,
+      "loss": 2.7668,
+      "step": 17100
+    },
+    {
+      "epoch": 0.15419241442797246,
+      "grad_norm": 0.17673403024673462,
+      "learning_rate": 0.00047124057499747573,
+      "loss": 2.7712,
+      "step": 17200
+    },
+    {
+      "epoch": 0.1550888819537165,
+      "grad_norm": 0.1734510213136673,
+      "learning_rate": 0.00047091182933823737,
+      "loss": 2.7579,
+      "step": 17300
+    },
+    {
+      "epoch": 0.15598534947946052,
+      "grad_norm": 0.17702840268611908,
+      "learning_rate": 0.0004705813314307569,
+      "loss": 2.7612,
+      "step": 17400
+    },
+    {
+      "epoch": 0.15688181700520457,
+      "grad_norm": 0.17082242667675018,
+      "learning_rate": 0.00047024908389650704,
+      "loss": 2.7597,
+      "step": 17500
+    },
+    {
+      "epoch": 0.15777828453094858,
+      "grad_norm": 0.18025079369544983,
+      "learning_rate": 0.00046991508937083875,
+      "loss": 2.7559,
+      "step": 17600
+    },
+    {
+      "epoch": 0.1586747520566926,
+      "grad_norm": 0.17427673935890198,
+      "learning_rate": 0.00046957935050295963,
+      "loss": 2.7577,
+      "step": 17700
+    },
+    {
+      "epoch": 0.15957121958243664,
+      "grad_norm": 0.17229335010051727,
+      "learning_rate": 0.0004692418699559134,
+      "loss": 2.7559,
+      "step": 17800
+    },
+    {
+      "epoch": 0.16046768710818066,
+      "grad_norm": 0.17168091237545013,
+      "learning_rate": 0.0004689026504065585,
+      "loss": 2.7582,
+      "step": 17900
+    },
+    {
+      "epoch": 0.16136415463392467,
+      "grad_norm": 0.17163404822349548,
+      "learning_rate": 0.0004685616945455469,
+      "loss": 2.7621,
+      "step": 18000
+    },
+    {
+      "epoch": 0.16226062215966872,
+      "grad_norm": 0.16957524418830872,
+      "learning_rate": 0.00046821900507730275,
+      "loss": 2.7586,
+      "step": 18100
+    },
+    {
+      "epoch": 0.16315708968541273,
+      "grad_norm": 0.165022075176239,
+      "learning_rate": 0.0004678745847200012,
+      "loss": 2.7541,
+      "step": 18200
+    },
+    {
+      "epoch": 0.16405355721115678,
+      "grad_norm": 0.17016685009002686,
+      "learning_rate": 0.00046752843620554655,
+      "loss": 2.7522,
+      "step": 18300
+    },
+    {
+      "epoch": 0.1649500247369008,
+      "grad_norm": 0.16751013696193695,
+      "learning_rate": 0.00046718056227955043,
+      "loss": 2.7501,
+      "step": 18400
+    },
+    {
+      "epoch": 0.1658464922626448,
+      "grad_norm": 0.1747274398803711,
+      "learning_rate": 0.0004668309657013106,
+      "loss": 2.7551,
+      "step": 18500
+    },
+    {
+      "epoch": 0.16674295978838885,
+      "grad_norm": 0.1704045832157135,
+      "learning_rate": 0.0004664796492437884,
+      "loss": 2.7522,
+      "step": 18600
+    },
+    {
+      "epoch": 0.16763942731413287,
+      "grad_norm": 0.16978523135185242,
+      "learning_rate": 0.0004661266156935873,
+      "loss": 2.7494,
+      "step": 18700
+    },
+    {
+      "epoch": 0.16853589483987688,
+      "grad_norm": 0.1745961755514145,
+      "learning_rate": 0.0004657718678509303,
+      "loss": 2.7546,
+      "step": 18800
+    },
+    {
+      "epoch": 0.16943236236562093,
+      "grad_norm": 0.16892167925834656,
+      "learning_rate": 0.0004654154085296382,
+      "loss": 2.7597,
+      "step": 18900
+    },
+    {
+      "epoch": 0.17032882989136494,
+      "grad_norm": 0.17090590298175812,
+      "learning_rate": 0.000465057240557107,
+      "loss": 2.7545,
+      "step": 19000
+    },
+    {
+      "epoch": 0.17122529741710896,
+      "grad_norm": 0.1694989651441574,
+      "learning_rate": 0.00046469736677428556,
+      "loss": 2.7525,
+      "step": 19100
+    },
+    {
+      "epoch": 0.172121764942853,
+      "grad_norm": 0.16777247190475464,
+      "learning_rate": 0.00046433579003565286,
+      "loss": 2.7478,
+      "step": 19200
+    },
+    {
+      "epoch": 0.17301823246859702,
+      "grad_norm": 0.1667390614748001,
+      "learning_rate": 0.00046397251320919584,
+      "loss": 2.7606,
+      "step": 19300
+    },
+    {
+      "epoch": 0.17391469999434106,
+      "grad_norm": 0.16497531533241272,
+      "learning_rate": 0.00046360753917638604,
+      "loss": 2.7459,
+      "step": 19400
+    },
+    {
+      "epoch": 0.17481116752008508,
+      "grad_norm": 0.16512423753738403,
+      "learning_rate": 0.00046324087083215727,
+      "loss": 2.748,
+      "step": 19500
+    },
+    {
+      "epoch": 0.1757076350458291,
+      "grad_norm": 0.1689414232969284,
+      "learning_rate": 0.0004628725110848823,
+      "loss": 2.7553,
+      "step": 19600
+    },
+    {
+      "epoch": 0.17660410257157313,
+      "grad_norm": 0.17020919919013977,
+      "learning_rate": 0.00046250246285635,
+      "loss": 2.7498,
+      "step": 19700
+    },
+    {
+      "epoch": 0.17750057009731715,
+      "grad_norm": 0.16616366803646088,
+      "learning_rate": 0.00046213072908174213,
+      "loss": 2.7459,
+      "step": 19800
+    },
+    {
+      "epoch": 0.17839703762306117,
+      "grad_norm": 0.16394533216953278,
+      "learning_rate": 0.00046175731270961,
+      "loss": 2.7494,
+      "step": 19900
+    },
+    {
+      "epoch": 0.1792935051488052,
+      "grad_norm": 0.1853574514389038,
+      "learning_rate": 0.0004613822167018508,
+      "loss": 2.7406,
+      "step": 20000
+    },
+    {
+      "epoch": 0.18018997267454923,
+      "grad_norm": 0.16652604937553406,
+      "learning_rate": 0.00046100544403368483,
+      "loss": 2.7364,
+      "step": 20100
+    },
+    {
+      "epoch": 0.18108644020029327,
+      "grad_norm": 0.16384656727313995,
+      "learning_rate": 0.0004606269976936314,
+      "loss": 2.747,
+      "step": 20200
+    },
+    {
+      "epoch": 0.18198290772603729,
+      "grad_norm": 0.16769762337207794,
+      "learning_rate": 0.00046024688068348484,
+      "loss": 2.7465,
+      "step": 20300
+    },
+    {
+      "epoch": 0.1828793752517813,
+      "grad_norm": 0.15770003199577332,
+      "learning_rate": 0.0004598650960182915,
+      "loss": 2.7459,
+      "step": 20400
+    },
+    {
+      "epoch": 0.18377584277752534,
+      "grad_norm": 0.16936561465263367,
+      "learning_rate": 0.00045948164672632514,
+      "loss": 2.7496,
+      "step": 20500
+    },
+    {
+      "epoch": 0.18467231030326936,
+      "grad_norm": 0.16903486847877502,
+      "learning_rate": 0.0004590965358490632,
+      "loss": 2.7447,
+      "step": 20600
+    },
+    {
+      "epoch": 0.18556877782901338,
+      "grad_norm": 0.16535791754722595,
+      "learning_rate": 0.00045870976644116267,
+      "loss": 2.7375,
+      "step": 20700
+    },
+    {
+      "epoch": 0.18646524535475742,
+      "grad_norm": 0.16924481093883514,
+      "learning_rate": 0.0004583213415704358,
+      "loss": 2.7494,
+      "step": 20800
+    },
+    {
+      "epoch": 0.18736171288050144,
+      "grad_norm": 0.16382168233394623,
+      "learning_rate": 0.00045793126431782584,
+      "loss": 2.7411,
+      "step": 20900
+    },
+    {
+      "epoch": 0.18825818040624548,
+      "grad_norm": 0.16305121779441833,
+      "learning_rate": 0.00045753953777738233,
+      "loss": 2.7465,
+      "step": 21000
+    },
+    {
+      "epoch": 0.1891546479319895,
+      "grad_norm": 0.1712082326412201,
+      "learning_rate": 0.00045714616505623703,
+      "loss": 2.7368,
+      "step": 21100
+    },
+    {
+      "epoch": 0.1900511154577335,
+      "grad_norm": 0.15486562252044678,
+      "learning_rate": 0.00045675114927457895,
+      "loss": 2.7385,
+      "step": 21200
+    },
+    {
+      "epoch": 0.19094758298347755,
+      "grad_norm": 0.16223442554473877,
+      "learning_rate": 0.0004563544935656296,
+      "loss": 2.7404,
+      "step": 21300
+    },
+    {
+      "epoch": 0.19184405050922157,
+      "grad_norm": 0.16115334630012512,
+      "learning_rate": 0.0004559562010756185,
+      "loss": 2.7325,
+      "step": 21400
+    },
+    {
+      "epoch": 0.19274051803496559,
+      "grad_norm": 0.15933333337306976,
+      "learning_rate": 0.0004555562749637574,
+      "loss": 2.7438,
+      "step": 21500
+    },
+    {
+      "epoch": 0.19363698556070963,
+      "grad_norm": 0.16439130902290344,
+      "learning_rate": 0.00045515471840221623,
+      "loss": 2.7353,
+      "step": 21600
+    },
+    {
+      "epoch": 0.19453345308645364,
+      "grad_norm": 0.1617208570241928,
+      "learning_rate": 0.0004547515345760973,
+      "loss": 2.7307,
+      "step": 21700
+    },
+    {
+      "epoch": 0.19542992061219766,
+      "grad_norm": 0.16066919267177582,
+      "learning_rate": 0.00045434672668341017,
+      "loss": 2.7379,
+      "step": 21800
+    },
+    {
+      "epoch": 0.1963263881379417,
+      "grad_norm": 0.16751715540885925,
+      "learning_rate": 0.0004539402979350464,
+      "loss": 2.7363,
+      "step": 21900
+    },
+    {
+      "epoch": 0.19722285566368572,
+      "grad_norm": 0.16093355417251587,
+      "learning_rate": 0.00045353225155475384,
+      "loss": 2.7412,
+      "step": 22000
+    },
+    {
+      "epoch": 0.19811932318942976,
+      "grad_norm": 0.16404040157794952,
+      "learning_rate": 0.00045312259077911153,
+      "loss": 2.7436,
+      "step": 22100
+    },
+    {
+      "epoch": 0.19901579071517378,
+      "grad_norm": 0.15982191264629364,
+      "learning_rate": 0.0004527113188575034,
+      "loss": 2.735,
+      "step": 22200
+    },
+    {
+      "epoch": 0.1999122582409178,
+      "grad_norm": 0.15941210091114044,
+      "learning_rate": 0.00045229843905209287,
+      "loss": 2.7457,
+      "step": 22300
+    },
+    {
+      "epoch": 0.20080872576666184,
+      "grad_norm": 0.15950894355773926,
+      "learning_rate": 0.00045188395463779707,
+      "loss": 2.736,
+      "step": 22400
+    },
+    {
+      "epoch": 0.20170519329240585,
+      "grad_norm": 0.1595742106437683,
+      "learning_rate": 0.0004514678689022606,
+      "loss": 2.7345,
+      "step": 22500
+    },
+    {
+      "epoch": 0.20260166081814987,
+      "grad_norm": 0.16472984850406647,
+      "learning_rate": 0.00045105018514582953,
+      "loss": 2.7316,
+      "step": 22600
+    },
+    {
+      "epoch": 0.2034981283438939,
+      "grad_norm": 0.1674964427947998,
+      "learning_rate": 0.0004506309066815254,
+      "loss": 2.7273,
+      "step": 22700
+    },
+    {
+      "epoch": 0.20439459586963793,
+      "grad_norm": 0.159923255443573,
+      "learning_rate": 0.00045021003683501853,
+      "loss": 2.7275,
+      "step": 22800
+    },
+    {
+      "epoch": 0.20529106339538197,
+      "grad_norm": 0.16554875671863556,
+      "learning_rate": 0.0004497875789446023,
+      "loss": 2.7388,
+      "step": 22900
+    },
+    {
+      "epoch": 0.206187530921126,
+      "grad_norm": 0.16701550781726837,
+      "learning_rate": 0.0004493635363611659,
+      "loss": 2.7303,
+      "step": 23000
+    },
+    {
+      "epoch": 0.20708399844687,
+      "grad_norm": 0.16078519821166992,
+      "learning_rate": 0.0004489379124481683,
+      "loss": 2.7312,
+      "step": 23100
+    },
+    {
+      "epoch": 0.20798046597261405,
+      "grad_norm": 0.16200828552246094,
+      "learning_rate": 0.0004485107105816115,
+      "loss": 2.7358,
+      "step": 23200
+    },
+    {
+      "epoch": 0.20887693349835806,
+      "grad_norm": 0.15977811813354492,
+      "learning_rate": 0.00044808193415001344,
+      "loss": 2.7309,
+      "step": 23300
+    },
+    {
+      "epoch": 0.20977340102410208,
+      "grad_norm": 0.16767093539237976,
+      "learning_rate": 0.0004476515865543815,
+      "loss": 2.7315,
+      "step": 23400
+    },
+    {
+      "epoch": 0.21066986854984612,
+      "grad_norm": 0.15728136897087097,
+      "learning_rate": 0.0004472196712081852,
+      "loss": 2.735,
+      "step": 23500
+    },
+    {
+      "epoch": 0.21156633607559014,
+      "grad_norm": 0.15546581149101257,
+      "learning_rate": 0.0004467861915373295,
+      "loss": 2.7301,
+      "step": 23600
+    },
+    {
+      "epoch": 0.21246280360133415,
+      "grad_norm": 0.15766388177871704,
+      "learning_rate": 0.0004463511509801273,
+      "loss": 2.7248,
+      "step": 23700
+    },
+    {
+      "epoch": 0.2133592711270782,
+      "grad_norm": 0.15457496047019958,
+      "learning_rate": 0.00044591455298727213,
+      "loss": 2.7234,
+      "step": 23800
+    },
+    {
+      "epoch": 0.21425573865282221,
+      "grad_norm": 0.15532919764518738,
+      "learning_rate": 0.0004454764010218112,
+      "loss": 2.7234,
+      "step": 23900
+    },
+    {
+      "epoch": 0.21515220617856626,
+      "grad_norm": 0.15373288094997406,
+      "learning_rate": 0.00044503669855911756,
+      "loss": 2.7254,
+      "step": 24000
+    },
+    {
+      "epoch": 0.21604867370431027,
+      "grad_norm": 0.15703774988651276,
+      "learning_rate": 0.00044459544908686236,
+      "loss": 2.7221,
+      "step": 24100
+    },
+    {
+      "epoch": 0.2169451412300543,
+      "grad_norm": 0.15091052651405334,
+      "learning_rate": 0.000444152656104988,
+      "loss": 2.7288,
+      "step": 24200
+    },
+    {
+      "epoch": 0.21784160875579833,
+      "grad_norm": 0.15579521656036377,
+      "learning_rate": 0.0004437083231256794,
+      "loss": 2.7192,
+      "step": 24300
+    },
+    {
+      "epoch": 0.21873807628154235,
+      "grad_norm": 0.15548183023929596,
+      "learning_rate": 0.0004432624536733367,
+      "loss": 2.735,
+      "step": 24400
+    },
+    {
+      "epoch": 0.21963454380728636,
+      "grad_norm": 0.15607748925685883,
+      "learning_rate": 0.00044281505128454713,
+      "loss": 2.7218,
+      "step": 24500
+    },
+    {
+      "epoch": 0.2205310113330304,
+      "grad_norm": 0.16912153363227844,
+      "learning_rate": 0.00044236611950805707,
+      "loss": 2.7198,
+      "step": 24600
+    },
+    {
+      "epoch": 0.22142747885877442,
+      "grad_norm": 0.15691158175468445,
+      "learning_rate": 0.0004419156619047439,
+      "loss": 2.7233,
+      "step": 24700
+    },
+    {
+      "epoch": 0.22232394638451847,
+      "grad_norm": 0.1513551026582718,
+      "learning_rate": 0.0004414636820475875,
+      "loss": 2.7198,
+      "step": 24800
+    },
+    {
+      "epoch": 0.22322041391026248,
+      "grad_norm": 0.15347003936767578,
+      "learning_rate": 0.0004410101835216422,
+      "loss": 2.7185,
+      "step": 24900
+    },
+    {
+      "epoch": 0.2241168814360065,
+      "grad_norm": 0.15076322853565216,
+      "learning_rate": 0.00044055516992400827,
+      "loss": 2.7232,
+      "step": 25000
+    },
+    {
+      "epoch": 0.22501334896175054,
+      "grad_norm": 0.1548827737569809,
+      "learning_rate": 0.00044009864486380335,
+      "loss": 2.7195,
+      "step": 25100
+    },
+    {
+      "epoch": 0.22590981648749456,
+      "grad_norm": 0.15956267714500427,
+      "learning_rate": 0.0004396406119621338,
+      "loss": 2.7231,
+      "step": 25200
+    },
+    {
+      "epoch": 0.22680628401323857,
+      "grad_norm": 0.14837653934955597,
+      "learning_rate": 0.00043918107485206603,
+      "loss": 2.7211,
+      "step": 25300
+    },
+    {
+      "epoch": 0.22770275153898262,
+      "grad_norm": 0.15196016430854797,
+      "learning_rate": 0.0004387200371785977,
+      "loss": 2.7209,
+      "step": 25400
+    },
+    {
+      "epoch": 0.22859921906472663,
+      "grad_norm": 0.15298311412334442,
+      "learning_rate": 0.00043825750259862873,
+      "loss": 2.7132,
+      "step": 25500
+    },
+    {
+      "epoch": 0.22949568659047065,
+      "grad_norm": 0.16172164678573608,
+      "learning_rate": 0.00043779347478093225,
+      "loss": 2.72,
+      "step": 25600
+    },
+    {
+      "epoch": 0.2303921541162147,
+      "grad_norm": 0.15316250920295715,
+      "learning_rate": 0.0004373279574061258,
+      "loss": 2.7284,
+      "step": 25700
+    },
+    {
+      "epoch": 0.2312886216419587,
+      "grad_norm": 0.1559763252735138,
+      "learning_rate": 0.0004368609541666417,
+      "loss": 2.7287,
+      "step": 25800
+    },
+    {
+      "epoch": 0.23218508916770275,
+      "grad_norm": 0.15110167860984802,
+      "learning_rate": 0.0004363924687666982,
+      "loss": 2.7144,
+      "step": 25900
+    },
+    {
+      "epoch": 0.23308155669344677,
+      "grad_norm": 0.15237966179847717,
+      "learning_rate": 0.0004359225049222697,
+      "loss": 2.7236,
+      "step": 26000
+    },
+    {
+      "epoch": 0.23397802421919078,
+      "grad_norm": 0.1501353681087494,
+      "learning_rate": 0.0004354510663610576,
+      "loss": 2.7132,
+      "step": 26100
+    },
+    {
+      "epoch": 0.23487449174493483,
+      "grad_norm": 0.15224196016788483,
+      "learning_rate": 0.00043497815682246044,
+      "loss": 2.7213,
+      "step": 26200
+    },
+    {
+      "epoch": 0.23577095927067884,
+      "grad_norm": 0.1475629061460495,
+      "learning_rate": 0.00043450378005754453,
+      "loss": 2.7138,
+      "step": 26300
+    },
+    {
+      "epoch": 0.23666742679642286,
+      "grad_norm": 0.15193389356136322,
+      "learning_rate": 0.000434027939829014,
+      "loss": 2.7079,
+      "step": 26400
+    },
+    {
+      "epoch": 0.2375638943221669,
+      "grad_norm": 0.1544477641582489,
+      "learning_rate": 0.00043355063991118095,
+      "loss": 2.7213,
+      "step": 26500
+    },
+    {
+      "epoch": 0.23846036184791092,
+      "grad_norm": 0.14745964109897614,
+      "learning_rate": 0.0004330718840899357,
+      "loss": 2.7175,
+      "step": 26600
+    },
+    {
+      "epoch": 0.23935682937365496,
+      "grad_norm": 0.1513916403055191,
+      "learning_rate": 0.00043259167616271644,
+      "loss": 2.7168,
+      "step": 26700
+    },
+    {
+      "epoch": 0.24025329689939898,
+      "grad_norm": 0.15179474651813507,
+      "learning_rate": 0.00043211001993847967,
+      "loss": 2.7099,
+      "step": 26800
+    },
+    {
+      "epoch": 0.241149764425143,
+      "grad_norm": 0.1514696627855301,
+      "learning_rate": 0.00043162691923766917,
+      "loss": 2.7201,
+      "step": 26900
+    },
+    {
+      "epoch": 0.24204623195088704,
+      "grad_norm": 0.15021684765815735,
+      "learning_rate": 0.0004311423778921865,
+      "loss": 2.7178,
+      "step": 27000
+    },
+    {
+      "epoch": 0.24294269947663105,
+      "grad_norm": 0.15158309042453766,
+      "learning_rate": 0.0004306563997453601,
+      "loss": 2.7076,
+      "step": 27100
+    },
+    {
+      "epoch": 0.24383916700237507,
+      "grad_norm": 0.14725959300994873,
+      "learning_rate": 0.00043016898865191487,
+      "loss": 2.7078,
+      "step": 27200
+    },
+    {
+      "epoch": 0.2447356345281191,
+      "grad_norm": 0.14671307802200317,
+      "learning_rate": 0.0004296801484779419,
+      "loss": 2.7109,
+      "step": 27300
+    },
+    {
+      "epoch": 0.24563210205386313,
+      "grad_norm": 0.1534958928823471,
+      "learning_rate": 0.0004291898831008675,
+      "loss": 2.7154,
+      "step": 27400
+    },
+    {
+      "epoch": 0.24652856957960714,
+      "grad_norm": 0.14696376025676727,
+      "learning_rate": 0.00042869819640942245,
+      "loss": 2.704,
+      "step": 27500
+    },
+    {
+      "epoch": 0.2474250371053512,
+      "grad_norm": 0.15045034885406494,
+      "learning_rate": 0.00042820509230361134,
+      "loss": 2.7145,
+      "step": 27600
+    },
+    {
+      "epoch": 0.2483215046310952,
+      "grad_norm": 0.1504196673631668,
+      "learning_rate": 0.0004277105746946814,
+      "loss": 2.7076,
+      "step": 27700
+    },
+    {
+      "epoch": 0.24921797215683925,
+      "grad_norm": 0.14724665880203247,
+      "learning_rate": 0.00042721464750509195,
+      "loss": 2.7034,
+      "step": 27800
+    },
+    {
+      "epoch": 0.25011443968258323,
+      "grad_norm": 0.14712798595428467,
+      "learning_rate": 0.00042671731466848253,
+      "loss": 2.7139,
+      "step": 27900
+    },
+    {
+      "epoch": 0.2510109072083273,
+      "grad_norm": 0.15577539801597595,
+      "learning_rate": 0.0004262185801296422,
+      "loss": 2.7077,
+      "step": 28000
+    },
+    {
+      "epoch": 0.2519073747340713,
+      "grad_norm": 0.14991851150989532,
+      "learning_rate": 0.0004257184478444785,
+      "loss": 2.7028,
+      "step": 28100
+    },
+    {
+      "epoch": 0.25280384225981534,
+      "grad_norm": 0.14650234580039978,
+      "learning_rate": 0.00042521692177998537,
+      "loss": 2.7069,
+      "step": 28200
+    },
+    {
+      "epoch": 0.25370030978555935,
+      "grad_norm": 0.1475357860326767,
+      "learning_rate": 0.0004247140059142123,
+      "loss": 2.7074,
+      "step": 28300
+    },
+    {
+      "epoch": 0.25459677731130337,
+      "grad_norm": 0.14877575635910034,
+      "learning_rate": 0.0004242097042362322,
+      "loss": 2.6997,
+      "step": 28400
+    },
+    {
+      "epoch": 0.25549324483704744,
+      "grad_norm": 0.15285931527614594,
+      "learning_rate": 0.0004237040207461104,
+      "loss": 2.7016,
+      "step": 28500
+    },
+    {
+      "epoch": 0.25638971236279146,
+      "grad_norm": 0.1494821012020111,
+      "learning_rate": 0.00042319695945487257,
+      "loss": 2.7093,
+      "step": 28600
+    },
+    {
+      "epoch": 0.25728617988853547,
+      "grad_norm": 0.14891798794269562,
+      "learning_rate": 0.00042268852438447297,
+      "loss": 2.7037,
+      "step": 28700
+    },
+    {
+      "epoch": 0.2581826474142795,
+      "grad_norm": 0.14432930946350098,
+      "learning_rate": 0.0004221787195677623,
+      "loss": 2.6977,
+      "step": 28800
+    },
+    {
+      "epoch": 0.2590791149400235,
+      "grad_norm": 0.1493179351091385,
+      "learning_rate": 0.0004216675490484561,
+      "loss": 2.7135,
+      "step": 28900
+    },
+    {
+      "epoch": 0.2599755824657676,
+      "grad_norm": 0.14740417897701263,
+      "learning_rate": 0.00042115501688110257,
+      "loss": 2.7079,
+      "step": 29000
+    },
+    {
+      "epoch": 0.2608720499915116,
+      "grad_norm": 0.1458692103624344,
+      "learning_rate": 0.0004206411271310502,
+      "loss": 2.7029,
+      "step": 29100
+    },
+    {
+      "epoch": 0.2617685175172556,
+      "grad_norm": 0.14873549342155457,
+      "learning_rate": 0.00042012588387441586,
+      "loss": 2.7096,
+      "step": 29200
+    },
+    {
+      "epoch": 0.2626649850429996,
+      "grad_norm": 0.14170105755329132,
+      "learning_rate": 0.00041960929119805215,
+      "loss": 2.7045,
+      "step": 29300
+    },
+    {
+      "epoch": 0.26356145256874364,
+      "grad_norm": 0.14562779664993286,
+      "learning_rate": 0.00041909135319951495,
+      "loss": 2.7044,
+      "step": 29400
+    },
+    {
+      "epoch": 0.26445792009448765,
+      "grad_norm": 0.14775702357292175,
+      "learning_rate": 0.00041857207398703154,
+      "loss": 2.7016,
+      "step": 29500
+    },
+    {
+      "epoch": 0.2653543876202317,
+      "grad_norm": 0.14575980603694916,
+      "learning_rate": 0.0004180514576794673,
+      "loss": 2.7095,
+      "step": 29600
+    },
+    {
+      "epoch": 0.26625085514597574,
+      "grad_norm": 0.14411023259162903,
+      "learning_rate": 0.0004175295084062931,
+      "loss": 2.7031,
+      "step": 29700
+    },
+    {
+      "epoch": 0.26714732267171976,
+      "grad_norm": 0.15136118233203888,
+      "learning_rate": 0.0004170062303075531,
+      "loss": 2.7052,
+      "step": 29800
+    },
+    {
+      "epoch": 0.26804379019746377,
+      "grad_norm": 0.14870133996009827,
+      "learning_rate": 0.00041648162753383144,
+      "loss": 2.7044,
+      "step": 29900
+    },
+    {
+      "epoch": 0.2689402577232078,
+      "grad_norm": 0.1507934033870697,
+      "learning_rate": 0.0004159557042462193,
+      "loss": 2.7113,
+      "step": 30000
+    },
+    {
+      "epoch": 0.26983672524895186,
+      "grad_norm": 0.14786238968372345,
+      "learning_rate": 0.0004154284646162822,
+      "loss": 2.6978,
+      "step": 30100
+    },
+    {
+      "epoch": 0.2707331927746959,
+      "grad_norm": 0.14871017634868622,
+      "learning_rate": 0.00041489991282602667,
+      "loss": 2.6969,
+      "step": 30200
+    },
+    {
+      "epoch": 0.2716296603004399,
+      "grad_norm": 0.1434573084115982,
+      "learning_rate": 0.00041437005306786716,
+      "loss": 2.7015,
+      "step": 30300
+    },
+    {
+      "epoch": 0.2725261278261839,
+      "grad_norm": 0.14657068252563477,
+      "learning_rate": 0.0004138388895445928,
+      "loss": 2.6979,
+      "step": 30400
+    },
+    {
+      "epoch": 0.2734225953519279,
+      "grad_norm": 0.14329074323177338,
+      "learning_rate": 0.00041330642646933397,
+      "loss": 2.7004,
+      "step": 30500
+    },
+    {
+      "epoch": 0.27431906287767194,
+      "grad_norm": 0.14054018259048462,
+      "learning_rate": 0.00041277266806552906,
+      "loss": 2.7023,
+      "step": 30600
+    },
+    {
+      "epoch": 0.275215530403416,
+      "grad_norm": 0.1455359160900116,
+      "learning_rate": 0.00041223761856689067,
+      "loss": 2.6972,
+      "step": 30700
+    },
+    {
+      "epoch": 0.27611199792916,
+      "grad_norm": 0.14418181777000427,
+      "learning_rate": 0.0004117012822173725,
+      "loss": 2.7018,
+      "step": 30800
+    },
+    {
+      "epoch": 0.27700846545490404,
+      "grad_norm": 0.14956611394882202,
+      "learning_rate": 0.0004111636632711353,
+      "loss": 2.7062,
+      "step": 30900
+    },
+    {
+      "epoch": 0.27790493298064806,
+      "grad_norm": 0.15083587169647217,
+      "learning_rate": 0.00041062476599251297,
+      "loss": 2.6939,
+      "step": 31000
+    },
+    {
+      "epoch": 0.27880140050639207,
+      "grad_norm": 0.14435066282749176,
+      "learning_rate": 0.00041008459465597947,
+      "loss": 2.699,
+      "step": 31100
+    },
+    {
+      "epoch": 0.27969786803213614,
+      "grad_norm": 0.14453698694705963,
+      "learning_rate": 0.00040954315354611403,
+      "loss": 2.7017,
+      "step": 31200
+    },
+    {
+      "epoch": 0.28059433555788016,
+      "grad_norm": 0.14497888088226318,
+      "learning_rate": 0.0004090004469575679,
+      "loss": 2.7007,
+      "step": 31300
+    },
+    {
+      "epoch": 0.2814908030836242,
+      "grad_norm": 0.1438531130552292,
+      "learning_rate": 0.0004084564791950298,
+      "loss": 2.7045,
+      "step": 31400
+    },
+    {
+      "epoch": 0.2823872706093682,
+      "grad_norm": 0.14402426779270172,
+      "learning_rate": 0.0004079112545731919,
+      "loss": 2.6936,
+      "step": 31500
+    },
+    {
+      "epoch": 0.2832837381351122,
+      "grad_norm": 0.14424774050712585,
+      "learning_rate": 0.00040736477741671576,
+      "loss": 2.6976,
+      "step": 31600
+    },
+    {
+      "epoch": 0.2841802056608562,
+      "grad_norm": 0.1417878419160843,
+      "learning_rate": 0.0004068170520601978,
+      "loss": 2.6965,
+      "step": 31700
+    },
+    {
+      "epoch": 0.2850766731866003,
+      "grad_norm": 0.14787960052490234,
+      "learning_rate": 0.0004062680828481352,
+      "loss": 2.6921,
+      "step": 31800
+    },
+    {
+      "epoch": 0.2859731407123443,
+      "grad_norm": 0.14073611795902252,
+      "learning_rate": 0.00040571787413489104,
+      "loss": 2.6997,
+      "step": 31900
+    },
+    {
+      "epoch": 0.2868696082380883,
+      "grad_norm": 0.13744498789310455,
+      "learning_rate": 0.0004051664302846601,
+      "loss": 2.6971,
+      "step": 32000
+    },
+    {
+      "epoch": 0.28776607576383234,
+      "grad_norm": 0.1451827883720398,
+      "learning_rate": 0.00040461375567143413,
+      "loss": 2.6972,
+      "step": 32100
+    },
+    {
+      "epoch": 0.28866254328957636,
+      "grad_norm": 0.14334948360919952,
+      "learning_rate": 0.0004040598546789672,
+      "loss": 2.6915,
+      "step": 32200
+    },
+    {
+      "epoch": 0.28955901081532043,
+      "grad_norm": 0.14681456983089447,
+      "learning_rate": 0.00040350473170074075,
+      "loss": 2.6943,
+      "step": 32300
+    },
+    {
+      "epoch": 0.29045547834106444,
+      "grad_norm": 0.14370225369930267,
+      "learning_rate": 0.00040294839113992907,
+      "loss": 2.6919,
+      "step": 32400
+    },
+    {
+      "epoch": 0.29135194586680846,
+      "grad_norm": 0.14142639935016632,
+      "learning_rate": 0.000402390837409364,
+      "loss": 2.6997,
+      "step": 32500
+    },
+    {
+      "epoch": 0.2922484133925525,
+      "grad_norm": 0.14680539071559906,
+      "learning_rate": 0.0004018320749315001,
+      "loss": 2.6939,
+      "step": 32600
+    },
+    {
+      "epoch": 0.2931448809182965,
+      "grad_norm": 0.14420635998249054,
+      "learning_rate": 0.00040127210813837975,
+      "loss": 2.6892,
+      "step": 32700
+    },
+    {
+      "epoch": 0.29404134844404056,
+      "grad_norm": 0.1437712162733078,
+      "learning_rate": 0.00040071094147159776,
+      "loss": 2.695,
+      "step": 32800
+    },
+    {
+      "epoch": 0.2949378159697846,
+      "grad_norm": 0.14098462462425232,
+      "learning_rate": 0.00040014857938226614,
+      "loss": 2.6851,
+      "step": 32900
+    },
+    {
+      "epoch": 0.2958342834955286,
+      "grad_norm": 0.13981753587722778,
+      "learning_rate": 0.00039958502633097895,
+      "loss": 2.6938,
+      "step": 33000
+    },
+    {
+      "epoch": 0.2967307510212726,
+      "grad_norm": 0.13768906891345978,
+      "learning_rate": 0.0003990202867877766,
+      "loss": 2.6834,
+      "step": 33100
+    },
+    {
+      "epoch": 0.2976272185470166,
+      "grad_norm": 0.13921628892421722,
+      "learning_rate": 0.000398454365232111,
+      "loss": 2.6977,
+      "step": 33200
+    },
+    {
+      "epoch": 0.29852368607276064,
+      "grad_norm": 0.14550244808197021,
+      "learning_rate": 0.0003978872661528094,
+      "loss": 2.6897,
+      "step": 33300
+    },
+    {
+      "epoch": 0.2994201535985047,
+      "grad_norm": 0.13877898454666138,
+      "learning_rate": 0.00039731899404803905,
+      "loss": 2.7007,
+      "step": 33400
+    },
+    {
+      "epoch": 0.30031662112424873,
+      "grad_norm": 0.14092972874641418,
+      "learning_rate": 0.00039674955342527165,
+      "loss": 2.6964,
+      "step": 33500
+    },
+    {
+      "epoch": 0.30121308864999274,
+      "grad_norm": 0.13762860000133514,
+      "learning_rate": 0.00039617894880124716,
+      "loss": 2.6908,
+      "step": 33600
+    },
+    {
+      "epoch": 0.30210955617573676,
+      "grad_norm": 0.14549992978572845,
+      "learning_rate": 0.00039560718470193866,
+      "loss": 2.6874,
+      "step": 33700
+    },
+    {
+      "epoch": 0.3030060237014808,
+      "grad_norm": 0.1463141292333603,
+      "learning_rate": 0.00039503426566251575,
+      "loss": 2.6867,
+      "step": 33800
+    },
+    {
+      "epoch": 0.30390249122722485,
+      "grad_norm": 0.13913968205451965,
+      "learning_rate": 0.0003944601962273091,
+      "loss": 2.6949,
+      "step": 33900
+    },
+    {
+      "epoch": 0.30479895875296886,
+      "grad_norm": 0.14150671660900116,
+      "learning_rate": 0.000393884980949774,
+      "loss": 2.6921,
+      "step": 34000
+    },
+    {
+      "epoch": 0.3056954262787129,
+      "grad_norm": 0.14346466958522797,
+      "learning_rate": 0.0003933086243924545,
+      "loss": 2.6935,
+      "step": 34100
+    },
+    {
+      "epoch": 0.3065918938044569,
+      "grad_norm": 0.14826858043670654,
+      "learning_rate": 0.00039273113112694736,
+      "loss": 2.6841,
+      "step": 34200
+    },
+    {
+      "epoch": 0.3074883613302009,
+      "grad_norm": 0.13878753781318665,
+      "learning_rate": 0.0003921525057338652,
+      "loss": 2.6858,
+      "step": 34300
+    },
+    {
+      "epoch": 0.3083848288559449,
+      "grad_norm": 0.1484604924917221,
+      "learning_rate": 0.0003915727528028009,
+      "loss": 2.6928,
+      "step": 34400
+    },
+    {
+      "epoch": 0.309281296381689,
+      "grad_norm": 0.14599835872650146,
+      "learning_rate": 0.00039099187693229066,
+      "loss": 2.6878,
+      "step": 34500
+    },
+    {
+      "epoch": 0.310177763907433,
+      "grad_norm": 0.14739172160625458,
+      "learning_rate": 0.0003904098827297777,
+      "loss": 2.6846,
+      "step": 34600
+    },
+    {
+      "epoch": 0.31107423143317703,
+      "grad_norm": 0.14112310111522675,
+      "learning_rate": 0.0003898267748115759,
+      "loss": 2.6879,
+      "step": 34700
+    },
+    {
+      "epoch": 0.31197069895892104,
+      "grad_norm": 0.13632678985595703,
+      "learning_rate": 0.00038924255780283277,
+      "loss": 2.6899,
+      "step": 34800
+    },
+    {
+      "epoch": 0.31286716648466506,
+      "grad_norm": 0.14148631691932678,
+      "learning_rate": 0.0003886572363374933,
+      "loss": 2.6901,
+      "step": 34900
+    },
+    {
+      "epoch": 0.31376363401040913,
+      "grad_norm": 0.14057014882564545,
+      "learning_rate": 0.0003880708150582626,
+      "loss": 2.6896,
+      "step": 35000
+    },
+    {
+      "epoch": 0.31466010153615315,
+      "grad_norm": 0.1407003551721573,
+      "learning_rate": 0.00038748329861656945,
+      "loss": 2.6833,
+      "step": 35100
+    },
+    {
+      "epoch": 0.31555656906189716,
+      "grad_norm": 0.1437043398618698,
+      "learning_rate": 0.00038689469167252957,
+      "loss": 2.6878,
+      "step": 35200
+    },
+    {
+      "epoch": 0.3164530365876412,
+      "grad_norm": 0.13795500993728638,
+      "learning_rate": 0.0003863049988949081,
+      "loss": 2.6831,
+      "step": 35300
+    },
+    {
+      "epoch": 0.3173495041133852,
+      "grad_norm": 0.13427403569221497,
+      "learning_rate": 0.0003857142249610833,
+      "loss": 2.6855,
+      "step": 35400
+    },
+    {
+      "epoch": 0.31824597163912927,
+      "grad_norm": 0.14448797702789307,
+      "learning_rate": 0.0003851223745570085,
+      "loss": 2.6934,
+      "step": 35500
+    },
+    {
+      "epoch": 0.3191424391648733,
+      "grad_norm": 0.13982614874839783,
+      "learning_rate": 0.00038452945237717597,
+      "loss": 2.6853,
+      "step": 35600
+    },
+    {
+      "epoch": 0.3200389066906173,
+      "grad_norm": 0.1487363576889038,
+      "learning_rate": 0.0003839354631245789,
+      "loss": 2.6812,
+      "step": 35700
+    },
+    {
+      "epoch": 0.3209353742163613,
+      "grad_norm": 0.1380891501903534,
+      "learning_rate": 0.00038334041151067447,
+      "loss": 2.6825,
+      "step": 35800
+    },
+    {
+      "epoch": 0.32183184174210533,
+      "grad_norm": 0.14058321714401245,
+      "learning_rate": 0.00038274430225534637,
+      "loss": 2.6868,
+      "step": 35900
+    },
+    {
+      "epoch": 0.32272830926784934,
+      "grad_norm": 0.1391235589981079,
+      "learning_rate": 0.00038214714008686745,
+      "loss": 2.6914,
+      "step": 36000
+    },
+    {
+      "epoch": 0.3236247767935934,
+      "grad_norm": 0.13752683997154236,
+      "learning_rate": 0.0003815489297418621,
+      "loss": 2.6854,
+      "step": 36100
+    },
+    {
+      "epoch": 0.32452124431933743,
+      "grad_norm": 0.14746305346488953,
+      "learning_rate": 0.00038094967596526873,
+      "loss": 2.6827,
+      "step": 36200
+    },
+    {
+      "epoch": 0.32541771184508145,
+      "grad_norm": 0.14420130848884583,
+      "learning_rate": 0.00038034938351030206,
+      "loss": 2.6876,
+      "step": 36300
+    },
+    {
+      "epoch": 0.32631417937082546,
+      "grad_norm": 0.1397886425256729,
+      "learning_rate": 0.0003797480571384157,
+      "loss": 2.6798,
+      "step": 36400
+    },
+    {
+      "epoch": 0.3272106468965695,
+      "grad_norm": 0.14020013809204102,
+      "learning_rate": 0.00037914570161926405,
+      "loss": 2.6868,
+      "step": 36500
+    },
+    {
+      "epoch": 0.32810711442231355,
+      "grad_norm": 0.13931307196617126,
+      "learning_rate": 0.0003785423217306645,
+      "loss": 2.6836,
+      "step": 36600
+    },
+    {
+      "epoch": 0.32900358194805757,
+      "grad_norm": 0.14019936323165894,
+      "learning_rate": 0.00037793792225855973,
+      "loss": 2.6871,
+      "step": 36700
+    },
+    {
+      "epoch": 0.3299000494738016,
+      "grad_norm": 0.13788673281669617,
+      "learning_rate": 0.0003773325079969796,
+      "loss": 2.6785,
+      "step": 36800
+    },
+    {
+      "epoch": 0.3307965169995456,
+      "grad_norm": 0.13793878257274628,
+      "learning_rate": 0.0003767260837480032,
+      "loss": 2.6785,
+      "step": 36900
+    },
+    {
+      "epoch": 0.3316929845252896,
+      "grad_norm": 0.13882067799568176,
+      "learning_rate": 0.0003761186543217209,
+      "loss": 2.6791,
+      "step": 37000
+    },
+    {
+      "epoch": 0.33258945205103363,
+      "grad_norm": 0.1409679502248764,
+      "learning_rate": 0.00037551022453619564,
+      "loss": 2.6809,
+      "step": 37100
+    },
+    {
+      "epoch": 0.3334859195767777,
+      "grad_norm": 0.1374359279870987,
+      "learning_rate": 0.0003749007992174254,
+      "loss": 2.6836,
+      "step": 37200
+    },
+    {
+      "epoch": 0.3343823871025217,
+      "grad_norm": 0.14376886188983917,
+      "learning_rate": 0.00037429038319930453,
+      "loss": 2.6798,
+      "step": 37300
+    },
+    {
+      "epoch": 0.33527885462826573,
+      "grad_norm": 0.1401790827512741,
+      "learning_rate": 0.00037367898132358544,
+      "loss": 2.6711,
+      "step": 37400
+    },
+    {
+      "epoch": 0.33617532215400975,
+      "grad_norm": 0.13772746920585632,
+      "learning_rate": 0.0003730665984398404,
+      "loss": 2.6768,
+      "step": 37500
+    },
+    {
+      "epoch": 0.33707178967975376,
+      "grad_norm": 0.14339695870876312,
+      "learning_rate": 0.0003724532394054228,
+      "loss": 2.6825,
+      "step": 37600
+    },
+    {
+      "epoch": 0.33796825720549784,
+      "grad_norm": 0.13825613260269165,
+      "learning_rate": 0.0003718389090854287,
+      "loss": 2.6796,
+      "step": 37700
+    },
+    {
+      "epoch": 0.33886472473124185,
+      "grad_norm": 0.13653145730495453,
+      "learning_rate": 0.00037122361235265855,
+      "loss": 2.6715,
+      "step": 37800
+    },
+    {
+      "epoch": 0.33976119225698587,
+      "grad_norm": 0.13996466994285583,
+      "learning_rate": 0.00037060735408757795,
+      "loss": 2.6852,
+      "step": 37900
+    },
+    {
+      "epoch": 0.3406576597827299,
+      "grad_norm": 0.13708461821079254,
+      "learning_rate": 0.0003699901391782795,
+      "loss": 2.6801,
+      "step": 38000
+    },
+    {
+      "epoch": 0.3415541273084739,
+      "grad_norm": 0.1354062259197235,
+      "learning_rate": 0.00036937197252044374,
+      "loss": 2.6748,
+      "step": 38100
+    },
+    {
+      "epoch": 0.3424505948342179,
+      "grad_norm": 0.13771981000900269,
+      "learning_rate": 0.00036875285901730035,
+      "loss": 2.6823,
+      "step": 38200
+    },
+    {
+      "epoch": 0.343347062359962,
+      "grad_norm": 0.13679315149784088,
+      "learning_rate": 0.0003681328035795892,
+      "loss": 2.6733,
+      "step": 38300
+    },
+    {
+      "epoch": 0.344243529885706,
+      "grad_norm": 0.14163535833358765,
+      "learning_rate": 0.0003675118111255216,
+      "loss": 2.6796,
+      "step": 38400
+    },
+    {
+      "epoch": 0.34513999741145,
+      "grad_norm": 0.13825653493404388,
+      "learning_rate": 0.00036688988658074124,
+      "loss": 2.6762,
+      "step": 38500
+    },
+    {
+      "epoch": 0.34603646493719403,
+      "grad_norm": 0.14156493544578552,
+      "learning_rate": 0.00036626703487828487,
+      "loss": 2.6694,
+      "step": 38600
+    },
+    {
+      "epoch": 0.34693293246293805,
+      "grad_norm": 0.13615700602531433,
+      "learning_rate": 0.0003656432609585435,
+      "loss": 2.6779,
+      "step": 38700
+    },
+    {
+      "epoch": 0.3478293999886821,
+      "grad_norm": 0.13561411201953888,
+      "learning_rate": 0.0003650185697692229,
+      "loss": 2.6838,
+      "step": 38800
+    },
+    {
+      "epoch": 0.34872586751442614,
+      "grad_norm": 0.1366826444864273,
+      "learning_rate": 0.0003643929662653046,
+      "loss": 2.6663,
+      "step": 38900
+    },
+    {
+      "epoch": 0.34962233504017015,
+      "grad_norm": 0.14089661836624146,
+      "learning_rate": 0.00036376645540900663,
+      "loss": 2.6708,
+      "step": 39000
+    },
+    {
+      "epoch": 0.35051880256591417,
+      "grad_norm": 0.13665175437927246,
+      "learning_rate": 0.000363139042169744,
+      "loss": 2.6705,
+      "step": 39100
+    },
+    {
+      "epoch": 0.3514152700916582,
+      "grad_norm": 0.13564077019691467,
+      "learning_rate": 0.0003625107315240891,
+      "loss": 2.6698,
+      "step": 39200
+    },
+    {
+      "epoch": 0.35231173761740225,
+      "grad_norm": 0.14218759536743164,
+      "learning_rate": 0.0003618815284557326,
+      "loss": 2.6816,
+      "step": 39300
+    },
+    {
+      "epoch": 0.35320820514314627,
+      "grad_norm": 0.14430591464042664,
+      "learning_rate": 0.00036125143795544387,
+      "loss": 2.6763,
+      "step": 39400
+    },
+    {
+      "epoch": 0.3541046726688903,
+      "grad_norm": 0.13792584836483002,
+      "learning_rate": 0.0003606204650210312,
+      "loss": 2.6674,
+      "step": 39500
+    },
+    {
+      "epoch": 0.3550011401946343,
+      "grad_norm": 0.136796236038208,
+      "learning_rate": 0.0003599886146573022,
+      "loss": 2.6718,
+      "step": 39600
+    },
+    {
+      "epoch": 0.3558976077203783,
+      "grad_norm": 0.13962402939796448,
+      "learning_rate": 0.00035935589187602426,
+      "loss": 2.6776,
+      "step": 39700
+    },
+    {
+      "epoch": 0.35679407524612233,
+      "grad_norm": 0.1402391791343689,
+      "learning_rate": 0.0003587223016958845,
+      "loss": 2.6739,
+      "step": 39800
+    },
+    {
+      "epoch": 0.3576905427718664,
+      "grad_norm": 0.13909810781478882,
+      "learning_rate": 0.0003580878491424504,
+      "loss": 2.6799,
+      "step": 39900
+    },
+    {
+      "epoch": 0.3585870102976104,
+      "grad_norm": 0.13748426735401154,
+      "learning_rate": 0.0003574525392481295,
+      "loss": 2.6765,
+      "step": 40000
+    },
+    {
+      "epoch": 0.35948347782335444,
+      "grad_norm": 0.132966548204422,
+      "learning_rate": 0.0003568163770521299,
+      "loss": 2.6702,
+      "step": 40100
+    },
+    {
+      "epoch": 0.36037994534909845,
+      "grad_norm": 0.13667432963848114,
+      "learning_rate": 0.00035617936760041976,
+      "loss": 2.6645,
+      "step": 40200
+    },
+    {
+      "epoch": 0.36127641287484247,
+      "grad_norm": 0.14251597225666046,
+      "learning_rate": 0.00035554151594568767,
+      "loss": 2.6719,
+      "step": 40300
+    },
+    {
+      "epoch": 0.36217288040058654,
+      "grad_norm": 0.1350788176059723,
+      "learning_rate": 0.00035490282714730273,
+      "loss": 2.6701,
+      "step": 40400
+    },
+    {
+      "epoch": 0.36306934792633055,
+      "grad_norm": 0.1342398077249527,
+      "learning_rate": 0.0003542633062712738,
+      "loss": 2.6653,
+      "step": 40500
+    },
+    {
+      "epoch": 0.36396581545207457,
+      "grad_norm": 0.13306699693202972,
+      "learning_rate": 0.00035362295839020996,
+      "loss": 2.6724,
+      "step": 40600
+    },
+    {
+      "epoch": 0.3648622829778186,
+      "grad_norm": 0.13337896764278412,
+      "learning_rate": 0.0003529817885832799,
+      "loss": 2.6633,
+      "step": 40700
+    },
+    {
+      "epoch": 0.3657587505035626,
+      "grad_norm": 0.13504943251609802,
+      "learning_rate": 0.00035233980193617184,
+      "loss": 2.6678,
+      "step": 40800
+    },
+    {
+      "epoch": 0.3666552180293066,
+      "grad_norm": 0.1362612545490265,
+      "learning_rate": 0.0003516970035410527,
+      "loss": 2.6669,
+      "step": 40900
+    },
+    {
+      "epoch": 0.3675516855550507,
+      "grad_norm": 0.1384200006723404,
+      "learning_rate": 0.00035105339849652873,
+      "loss": 2.6706,
+      "step": 41000
+    },
+    {
+      "epoch": 0.3684481530807947,
+      "grad_norm": 0.1365375816822052,
+      "learning_rate": 0.0003504089919076039,
+      "loss": 2.6647,
+      "step": 41100
+    },
+    {
+      "epoch": 0.3693446206065387,
+      "grad_norm": 0.13598588109016418,
+      "learning_rate": 0.00034976378888564014,
+      "loss": 2.6647,
+      "step": 41200
+    },
+    {
+      "epoch": 0.37024108813228274,
+      "grad_norm": 0.13357794284820557,
+      "learning_rate": 0.00034911779454831665,
+      "loss": 2.6694,
+      "step": 41300
+    },
+    {
+      "epoch": 0.37113755565802675,
+      "grad_norm": 0.13431541621685028,
+      "learning_rate": 0.00034847101401958914,
+      "loss": 2.6644,
+      "step": 41400
+    },
+    {
+      "epoch": 0.3720340231837708,
+      "grad_norm": 0.13836424052715302,
+      "learning_rate": 0.0003478234524296494,
+      "loss": 2.6752,
+      "step": 41500
+    },
+    {
+      "epoch": 0.37293049070951484,
+      "grad_norm": 0.13321442902088165,
+      "learning_rate": 0.00034717511491488454,
+      "loss": 2.6755,
+      "step": 41600
+    },
+    {
+      "epoch": 0.37382695823525885,
+      "grad_norm": 0.13987848162651062,
+      "learning_rate": 0.00034652600661783594,
+      "loss": 2.6666,
+      "step": 41700
+    },
+    {
+      "epoch": 0.37472342576100287,
+      "grad_norm": 0.1362723857164383,
+      "learning_rate": 0.00034587613268715917,
+      "loss": 2.6736,
+      "step": 41800
+    },
+    {
+      "epoch": 0.3756198932867469,
+      "grad_norm": 0.13275477290153503,
+      "learning_rate": 0.0003452254982775824,
+      "loss": 2.6658,
+      "step": 41900
+    },
+    {
+      "epoch": 0.37651636081249096,
+      "grad_norm": 0.13551685214042664,
+      "learning_rate": 0.0003445741085498659,
+      "loss": 2.6666,
+      "step": 42000
+    },
+    {
+      "epoch": 0.377412828338235,
+      "grad_norm": 0.13056619465351105,
+      "learning_rate": 0.00034392196867076124,
+      "loss": 2.6638,
+      "step": 42100
+    },
+    {
+      "epoch": 0.378309295863979,
+      "grad_norm": 0.1297016590833664,
+      "learning_rate": 0.0003432690838129698,
+      "loss": 2.6666,
+      "step": 42200
+    },
+    {
+      "epoch": 0.379205763389723,
+      "grad_norm": 0.1312173455953598,
+      "learning_rate": 0.00034261545915510223,
+      "loss": 2.6665,
+      "step": 42300
+    },
+    {
+      "epoch": 0.380102230915467,
+      "grad_norm": 0.13310836255550385,
+      "learning_rate": 0.00034196109988163714,
+      "loss": 2.6629,
+      "step": 42400
+    },
+    {
+      "epoch": 0.38099869844121104,
+      "grad_norm": 0.13753747940063477,
+      "learning_rate": 0.0003413060111828801,
+      "loss": 2.6682,
+      "step": 42500
+    },
+    {
+      "epoch": 0.3818951659669551,
+      "grad_norm": 0.13357776403427124,
+      "learning_rate": 0.00034065019825492237,
+      "loss": 2.6657,
+      "step": 42600
+    },
+    {
+      "epoch": 0.3827916334926991,
+      "grad_norm": 0.13426382839679718,
+      "learning_rate": 0.00033999366629959956,
+      "loss": 2.6663,
+      "step": 42700
+    },
+    {
+      "epoch": 0.38368810101844314,
+      "grad_norm": 0.14345508813858032,
+      "learning_rate": 0.0003393364205244508,
+      "loss": 2.6652,
+      "step": 42800
+    },
+    {
+      "epoch": 0.38458456854418716,
+      "grad_norm": 0.1440078616142273,
+      "learning_rate": 0.00033867846614267695,
+      "loss": 2.6729,
+      "step": 42900
+    },
+    {
+      "epoch": 0.38548103606993117,
+      "grad_norm": 0.1365692913532257,
+      "learning_rate": 0.00033801980837309953,
+      "loss": 2.6725,
+      "step": 43000
+    },
+    {
+      "epoch": 0.38637750359567524,
+      "grad_norm": 0.13232523202896118,
+      "learning_rate": 0.0003373604524401193,
+      "loss": 2.6655,
+      "step": 43100
+    },
+    {
+      "epoch": 0.38727397112141926,
+      "grad_norm": 0.13531488180160522,
+      "learning_rate": 0.0003367004035736747,
+      "loss": 2.6694,
+      "step": 43200
+    },
+    {
+      "epoch": 0.3881704386471633,
+      "grad_norm": 0.1343296617269516,
+      "learning_rate": 0.0003360396670092004,
+      "loss": 2.6576,
+      "step": 43300
+    },
+    {
+      "epoch": 0.3890669061729073,
+      "grad_norm": 0.13563938438892365,
+      "learning_rate": 0.00033537824798758603,
+      "loss": 2.6682,
+      "step": 43400
+    },
+    {
+      "epoch": 0.3899633736986513,
+      "grad_norm": 0.13184553384780884,
+      "learning_rate": 0.00033471615175513416,
+      "loss": 2.6544,
+      "step": 43500
+    },
+    {
+      "epoch": 0.3908598412243953,
+      "grad_norm": 0.13147491216659546,
+      "learning_rate": 0.00033405338356351906,
+      "loss": 2.659,
+      "step": 43600
+    },
+    {
+      "epoch": 0.3917563087501394,
+      "grad_norm": 0.128835991024971,
+      "learning_rate": 0.00033338994866974477,
+      "loss": 2.668,
+      "step": 43700
+    },
+    {
+      "epoch": 0.3926527762758834,
+      "grad_norm": 0.13172951340675354,
+      "learning_rate": 0.00033272585233610363,
+      "loss": 2.668,
+      "step": 43800
+    },
+    {
+      "epoch": 0.3935492438016274,
+      "grad_norm": 0.13852158188819885,
+      "learning_rate": 0.00033206109983013464,
+      "loss": 2.6618,
+      "step": 43900
+    },
+    {
+      "epoch": 0.39444571132737144,
+      "grad_norm": 0.13343538343906403,
+      "learning_rate": 0.00033139569642458104,
+      "loss": 2.6694,
+      "step": 44000
+    },
+    {
+      "epoch": 0.39534217885311546,
+      "grad_norm": 0.13819359242916107,
+      "learning_rate": 0.00033072964739734936,
+      "loss": 2.6645,
+      "step": 44100
+    },
+    {
+      "epoch": 0.3962386463788595,
+      "grad_norm": 0.14136740565299988,
+      "learning_rate": 0.0003300629580314668,
+      "loss": 2.6596,
+      "step": 44200
+    },
+    {
+      "epoch": 0.39713511390460354,
+      "grad_norm": 0.13478617370128632,
+      "learning_rate": 0.00032939563361503995,
+      "loss": 2.6682,
+      "step": 44300
+    },
+    {
+      "epoch": 0.39803158143034756,
+      "grad_norm": 0.13910335302352905,
+      "learning_rate": 0.00032872767944121234,
+      "loss": 2.6561,
+      "step": 44400
+    },
+    {
+      "epoch": 0.3989280489560916,
+      "grad_norm": 0.12869343161582947,
+      "learning_rate": 0.0003280591008081227,
+      "loss": 2.6698,
+      "step": 44500
+    },
+    {
+      "epoch": 0.3998245164818356,
+      "grad_norm": 0.13684409856796265,
+      "learning_rate": 0.00032738990301886306,
+      "loss": 2.6606,
+      "step": 44600
+    },
+    {
+      "epoch": 0.4007209840075796,
+      "grad_norm": 0.13658647239208221,
+      "learning_rate": 0.00032672009138143634,
+      "loss": 2.6641,
+      "step": 44700
+    },
+    {
+      "epoch": 0.4016174515333237,
+      "grad_norm": 0.13844211399555206,
+      "learning_rate": 0.00032604967120871444,
+      "loss": 2.6548,
+      "step": 44800
+    },
+    {
+      "epoch": 0.4025139190590677,
+      "grad_norm": 0.12905430793762207,
+      "learning_rate": 0.0003253786478183963,
+      "loss": 2.6601,
+      "step": 44900
+    },
+    {
+      "epoch": 0.4034103865848117,
+      "grad_norm": 0.1291283369064331,
+      "learning_rate": 0.00032470702653296515,
+      "loss": 2.659,
+      "step": 45000
+    },
+    {
+      "epoch": 0.4043068541105557,
+      "grad_norm": 0.1292611062526703,
+      "learning_rate": 0.0003240348126796471,
+      "loss": 2.6575,
+      "step": 45100
+    },
+    {
+      "epoch": 0.40520332163629974,
+      "grad_norm": 0.13455387949943542,
+      "learning_rate": 0.00032336201159036815,
+      "loss": 2.6514,
+      "step": 45200
+    },
+    {
+      "epoch": 0.4060997891620438,
+      "grad_norm": 0.12952381372451782,
+      "learning_rate": 0.00032268862860171214,
+      "loss": 2.6559,
+      "step": 45300
+    },
+    {
+      "epoch": 0.4069962566877878,
+      "grad_norm": 0.13285262882709503,
+      "learning_rate": 0.00032201466905487885,
+      "loss": 2.66,
+      "step": 45400
+    },
+    {
+      "epoch": 0.40789272421353184,
+      "grad_norm": 0.13195694983005524,
+      "learning_rate": 0.0003213401382956408,
+      "loss": 2.6638,
+      "step": 45500
+    },
+    {
+      "epoch": 0.40878919173927586,
+      "grad_norm": 0.12948913872241974,
+      "learning_rate": 0.0003206650416743017,
+      "loss": 2.6568,
+      "step": 45600
+    },
+    {
+      "epoch": 0.4096856592650199,
+      "grad_norm": 0.13518835604190826,
+      "learning_rate": 0.0003199893845456531,
+      "loss": 2.6584,
+      "step": 45700
+    },
+    {
+      "epoch": 0.41058212679076395,
+      "grad_norm": 0.13390909135341644,
+      "learning_rate": 0.00031931317226893295,
+      "loss": 2.6477,
+      "step": 45800
+    },
+    {
+      "epoch": 0.41147859431650796,
+      "grad_norm": 0.13253462314605713,
+      "learning_rate": 0.00031863641020778247,
+      "loss": 2.6549,
+      "step": 45900
+    },
+    {
+      "epoch": 0.412375061842252,
+      "grad_norm": 0.13521717488765717,
+      "learning_rate": 0.0003179591037302035,
+      "loss": 2.6527,
+      "step": 46000
+    },
+    {
+      "epoch": 0.413271529367996,
+      "grad_norm": 0.14094848930835724,
+      "learning_rate": 0.0003172812582085163,
+      "loss": 2.6608,
+      "step": 46100
+    },
+    {
+      "epoch": 0.41416799689374,
+      "grad_norm": 0.1350257843732834,
+      "learning_rate": 0.0003166028790193166,
+      "loss": 2.6568,
+      "step": 46200
+    },
+    {
+      "epoch": 0.415064464419484,
+      "grad_norm": 0.1422998309135437,
+      "learning_rate": 0.00031592397154343334,
+      "loss": 2.6567,
+      "step": 46300
+    },
+    {
+      "epoch": 0.4159609319452281,
+      "grad_norm": 0.12808671593666077,
+      "learning_rate": 0.0003152445411658856,
+      "loss": 2.6528,
+      "step": 46400
+    },
+    {
+      "epoch": 0.4168573994709721,
+      "grad_norm": 0.12911584973335266,
+      "learning_rate": 0.00031456459327584027,
+      "loss": 2.6641,
+      "step": 46500
+    },
+    {
+      "epoch": 0.41775386699671613,
+      "grad_norm": 0.13506324589252472,
+      "learning_rate": 0.00031388413326656885,
+      "loss": 2.655,
+      "step": 46600
+    },
+    {
+      "epoch": 0.41865033452246014,
+      "grad_norm": 0.13051171600818634,
+      "learning_rate": 0.00031320316653540495,
+      "loss": 2.6625,
+      "step": 46700
+    },
+    {
+      "epoch": 0.41954680204820416,
+      "grad_norm": 0.1310121864080429,
+      "learning_rate": 0.00031252169848370155,
+      "loss": 2.6587,
+      "step": 46800
+    },
+    {
+      "epoch": 0.42044326957394823,
+      "grad_norm": 0.13700339198112488,
+      "learning_rate": 0.00031183973451678806,
+      "loss": 2.6501,
+      "step": 46900
+    },
+    {
+      "epoch": 0.42133973709969225,
+      "grad_norm": 0.13302241265773773,
+      "learning_rate": 0.0003111572800439273,
+      "loss": 2.6557,
+      "step": 47000
+    },
+    {
+      "epoch": 0.42223620462543626,
+      "grad_norm": 0.13645824790000916,
+      "learning_rate": 0.00031047434047827294,
+      "loss": 2.655,
+      "step": 47100
+    },
+    {
+      "epoch": 0.4231326721511803,
+      "grad_norm": 0.13405530154705048,
+      "learning_rate": 0.0003097909212368261,
+      "loss": 2.6616,
+      "step": 47200
+    },
+    {
+      "epoch": 0.4240291396769243,
+      "grad_norm": 0.1315753012895584,
+      "learning_rate": 0.0003091070277403927,
+      "loss": 2.6516,
+      "step": 47300
+    },
+    {
+      "epoch": 0.4249256072026683,
+      "grad_norm": 0.13063915073871613,
+      "learning_rate": 0.0003084226654135406,
+      "loss": 2.6581,
+      "step": 47400
+    },
+    {
+      "epoch": 0.4258220747284124,
+      "grad_norm": 0.1367933750152588,
+      "learning_rate": 0.00030773783968455614,
+      "loss": 2.6541,
+      "step": 47500
+    },
+    {
+      "epoch": 0.4267185422541564,
+      "grad_norm": 0.13026690483093262,
+      "learning_rate": 0.0003070525559854015,
+      "loss": 2.651,
+      "step": 47600
+    },
+    {
+      "epoch": 0.4276150097799004,
+      "grad_norm": 0.13831181824207306,
+      "learning_rate": 0.00030636681975167114,
+      "loss": 2.6596,
+      "step": 47700
+    },
+    {
+      "epoch": 0.42851147730564443,
+      "grad_norm": 0.13564659655094147,
+      "learning_rate": 0.0003056806364225493,
+      "loss": 2.6502,
+      "step": 47800
+    },
+    {
+      "epoch": 0.42940794483138844,
+      "grad_norm": 0.13234366476535797,
+      "learning_rate": 0.00030499401144076636,
+      "loss": 2.647,
+      "step": 47900
+    },
+    {
+      "epoch": 0.4303044123571325,
+      "grad_norm": 0.13098286092281342,
+      "learning_rate": 0.00030430695025255596,
+      "loss": 2.6519,
+      "step": 48000
+    },
+    {
+      "epoch": 0.43120087988287653,
+      "grad_norm": 0.13402335345745087,
+      "learning_rate": 0.00030361945830761146,
+      "loss": 2.6581,
+      "step": 48100
+    },
+    {
+      "epoch": 0.43209734740862055,
+      "grad_norm": 0.13898345828056335,
+      "learning_rate": 0.0003029315410590431,
+      "loss": 2.652,
+      "step": 48200
+    },
+    {
+      "epoch": 0.43299381493436456,
+      "grad_norm": 0.1286890059709549,
+      "learning_rate": 0.00030224320396333456,
+      "loss": 2.6529,
+      "step": 48300
+    },
+    {
+      "epoch": 0.4338902824601086,
+      "grad_norm": 0.13320975005626678,
+      "learning_rate": 0.00030155445248029975,
+      "loss": 2.6533,
+      "step": 48400
+    },
+    {
+      "epoch": 0.4347867499858526,
+      "grad_norm": 0.128444641828537,
+      "learning_rate": 0.00030086529207303935,
+      "loss": 2.6537,
+      "step": 48500
+    },
+    {
+      "epoch": 0.43568321751159667,
+      "grad_norm": 0.1326085776090622,
+      "learning_rate": 0.00030017572820789765,
+      "loss": 2.6522,
+      "step": 48600
+    },
+    {
+      "epoch": 0.4365796850373407,
+      "grad_norm": 0.13274775445461273,
+      "learning_rate": 0.00029948576635441905,
+      "loss": 2.6506,
+      "step": 48700
+    },
+    {
+      "epoch": 0.4374761525630847,
+      "grad_norm": 0.13352380692958832,
+      "learning_rate": 0.0002987954119853048,
+      "loss": 2.6552,
+      "step": 48800
+    },
+    {
+      "epoch": 0.4383726200888287,
+      "grad_norm": 0.13431380689144135,
+      "learning_rate": 0.0002981046705763696,
+      "loss": 2.6532,
+      "step": 48900
+    },
+    {
+      "epoch": 0.43926908761457273,
+      "grad_norm": 0.1325884610414505,
+      "learning_rate": 0.0002974135476064981,
+      "loss": 2.6586,
+      "step": 49000
+    },
+    {
+      "epoch": 0.4401655551403168,
+      "grad_norm": 0.13121579587459564,
+      "learning_rate": 0.0002967220485576013,
+      "loss": 2.6486,
+      "step": 49100
+    },
+    {
+      "epoch": 0.4410620226660608,
+      "grad_norm": 0.12927518784999847,
+      "learning_rate": 0.0002960301789145733,
+      "loss": 2.6553,
+      "step": 49200
+    },
+    {
+      "epoch": 0.44195849019180483,
+      "grad_norm": 0.1333087980747223,
+      "learning_rate": 0.0002953379441652478,
+      "loss": 2.6525,
+      "step": 49300
+    },
+    {
+      "epoch": 0.44285495771754885,
+      "grad_norm": 0.13250477612018585,
+      "learning_rate": 0.0002946453498003543,
+      "loss": 2.6441,
+      "step": 49400
+    },
+    {
+      "epoch": 0.44375142524329286,
+      "grad_norm": 0.13053111732006073,
+      "learning_rate": 0.00029395240131347507,
+      "loss": 2.6549,
+      "step": 49500
+    },
+    {
+      "epoch": 0.44464789276903693,
+      "grad_norm": 0.1300216168165207,
+      "learning_rate": 0.00029325910420100083,
+      "loss": 2.6509,
+      "step": 49600
+    },
+    {
+      "epoch": 0.44554436029478095,
+      "grad_norm": 0.13556109368801117,
+      "learning_rate": 0.00029256546396208766,
+      "loss": 2.6403,
+      "step": 49700
+    },
+    {
+      "epoch": 0.44644082782052497,
+      "grad_norm": 0.12819913029670715,
+      "learning_rate": 0.00029187148609861353,
+      "loss": 2.6435,
+      "step": 49800
+    },
+    {
+      "epoch": 0.447337295346269,
+      "grad_norm": 0.1305069774389267,
+      "learning_rate": 0.0002911771761151342,
+      "loss": 2.654,
+      "step": 49900
+    },
+    {
+      "epoch": 0.448233762872013,
+      "grad_norm": 0.1274418979883194,
+      "learning_rate": 0.0002904825395188397,
+      "loss": 2.65,
+      "step": 50000
+    },
+    {
+      "epoch": 0.449130230397757,
+      "grad_norm": 0.13119454681873322,
+      "learning_rate": 0.0002897875818195111,
+      "loss": 2.6608,
+      "step": 50100
+    },
+    {
+      "epoch": 0.4500266979235011,
+      "grad_norm": 0.13181808590888977,
+      "learning_rate": 0.00028909230852947575,
+      "loss": 2.6457,
+      "step": 50200
+    },
+    {
+      "epoch": 0.4509231654492451,
+      "grad_norm": 0.1294698864221573,
+      "learning_rate": 0.00028839672516356495,
+      "loss": 2.6537,
+      "step": 50300
+    },
+    {
+      "epoch": 0.4518196329749891,
+      "grad_norm": 0.13653060793876648,
+      "learning_rate": 0.00028770083723906904,
+      "loss": 2.6575,
+      "step": 50400
+    },
+    {
+      "epoch": 0.45271610050073313,
+      "grad_norm": 0.12975502014160156,
+      "learning_rate": 0.0002870046502756942,
+      "loss": 2.645,
+      "step": 50500
+    },
+    {
+      "epoch": 0.45361256802647715,
+      "grad_norm": 0.13355258107185364,
+      "learning_rate": 0.0002863081697955187,
+      "loss": 2.6435,
+      "step": 50600
+    },
+    {
+      "epoch": 0.4545090355522212,
+      "grad_norm": 0.1305021345615387,
+      "learning_rate": 0.00028561140132294863,
+      "loss": 2.6454,
+      "step": 50700
+    },
+    {
+      "epoch": 0.45540550307796523,
+      "grad_norm": 0.13262712955474854,
+      "learning_rate": 0.00028491435038467466,
+      "loss": 2.6534,
+      "step": 50800
+    },
+    {
+      "epoch": 0.45630197060370925,
+      "grad_norm": 0.13376927375793457,
+      "learning_rate": 0.00028421702250962786,
+      "loss": 2.646,
+      "step": 50900
+    },
+    {
+      "epoch": 0.45719843812945327,
+      "grad_norm": 0.13164885342121124,
+      "learning_rate": 0.0002835194232289361,
+      "loss": 2.6567,
+      "step": 51000
+    },
+    {
+      "epoch": 0.4580949056551973,
+      "grad_norm": 0.13637390732765198,
+      "learning_rate": 0.0002828215580758798,
+      "loss": 2.6466,
+      "step": 51100
+    },
+    {
+      "epoch": 0.4589913731809413,
+      "grad_norm": 0.13366563618183136,
+      "learning_rate": 0.0002821234325858482,
+      "loss": 2.6475,
+      "step": 51200
+    },
+    {
+      "epoch": 0.45988784070668537,
+      "grad_norm": 0.1311609447002411,
+      "learning_rate": 0.0002814250522962956,
+      "loss": 2.653,
+      "step": 51300
+    },
+    {
+      "epoch": 0.4607843082324294,
+      "grad_norm": 0.13002333045005798,
+      "learning_rate": 0.0002807264227466975,
+      "loss": 2.6472,
+      "step": 51400
+    },
+    {
+      "epoch": 0.4616807757581734,
+      "grad_norm": 0.12868809700012207,
+      "learning_rate": 0.00028002754947850623,
+      "loss": 2.6347,
+      "step": 51500
+    },
+    {
+      "epoch": 0.4625772432839174,
+      "grad_norm": 0.1310162991285324,
+      "learning_rate": 0.00027932843803510755,
+      "loss": 2.6414,
+      "step": 51600
+    },
+    {
+      "epoch": 0.46347371080966143,
+      "grad_norm": 0.12595106661319733,
+      "learning_rate": 0.00027862909396177615,
+      "loss": 2.6427,
+      "step": 51700
+    },
+    {
+      "epoch": 0.4643701783354055,
+      "grad_norm": 0.130056232213974,
+      "learning_rate": 0.000277929522805632,
+      "loss": 2.6461,
+      "step": 51800
+    },
+    {
+      "epoch": 0.4652666458611495,
+      "grad_norm": 0.12970997393131256,
+      "learning_rate": 0.00027722973011559633,
+      "loss": 2.6483,
+      "step": 51900
+    },
+    {
+      "epoch": 0.46616311338689354,
+      "grad_norm": 0.13723085820674896,
+      "learning_rate": 0.00027652972144234745,
+      "loss": 2.6384,
+      "step": 52000
+    },
+    {
+      "epoch": 0.46705958091263755,
+      "grad_norm": 0.13320958614349365,
+      "learning_rate": 0.000275829502338277,
+      "loss": 2.6573,
+      "step": 52100
+    },
+    {
+      "epoch": 0.46795604843838157,
+      "grad_norm": 0.1280314177274704,
+      "learning_rate": 0.00027512907835744547,
+      "loss": 2.6469,
+      "step": 52200
+    },
+    {
+      "epoch": 0.46885251596412564,
+      "grad_norm": 0.13650980591773987,
+      "learning_rate": 0.0002744284550555385,
+      "loss": 2.6457,
+      "step": 52300
+    },
+    {
+      "epoch": 0.46974898348986965,
+      "grad_norm": 0.1296728104352951,
+      "learning_rate": 0.0002737276379898229,
+      "loss": 2.6386,
+      "step": 52400
+    },
+    {
+      "epoch": 0.47064545101561367,
+      "grad_norm": 0.13340330123901367,
+      "learning_rate": 0.0002730266327191023,
+      "loss": 2.6419,
+      "step": 52500
+    },
+    {
+      "epoch": 0.4715419185413577,
+      "grad_norm": 0.1351691633462906,
+      "learning_rate": 0.0002723254448036731,
+      "loss": 2.6474,
+      "step": 52600
+    },
+    {
+      "epoch": 0.4724383860671017,
+      "grad_norm": 0.135370671749115,
+      "learning_rate": 0.00027162407980528037,
+      "loss": 2.6486,
+      "step": 52700
+    },
+    {
+      "epoch": 0.4733348535928457,
+      "grad_norm": 0.12964408099651337,
+      "learning_rate": 0.0002709225432870738,
+      "loss": 2.6437,
+      "step": 52800
+    },
+    {
+      "epoch": 0.4742313211185898,
+      "grad_norm": 0.12840570509433746,
+      "learning_rate": 0.0002702208408135637,
+      "loss": 2.6409,
+      "step": 52900
+    },
+    {
+      "epoch": 0.4751277886443338,
+      "grad_norm": 0.13321109116077423,
+      "learning_rate": 0.00026951897795057653,
+      "loss": 2.6443,
+      "step": 53000
+    },
+    {
+      "epoch": 0.4760242561700778,
+      "grad_norm": 0.13473448157310486,
+      "learning_rate": 0.0002688169602652113,
+      "loss": 2.6404,
+      "step": 53100
+    },
+    {
+      "epoch": 0.47692072369582184,
+      "grad_norm": 0.1295260488986969,
+      "learning_rate": 0.00026811479332579445,
+      "loss": 2.6359,
+      "step": 53200
+    },
+    {
+      "epoch": 0.47781719122156585,
+      "grad_norm": 0.13443715870380402,
+      "learning_rate": 0.0002674124827018368,
+      "loss": 2.6423,
+      "step": 53300
+    },
+    {
+      "epoch": 0.4787136587473099,
+      "grad_norm": 0.13394586741924286,
+      "learning_rate": 0.0002667100339639886,
+      "loss": 2.6353,
+      "step": 53400
+    },
+    {
+      "epoch": 0.47961012627305394,
+      "grad_norm": 0.1299637407064438,
+      "learning_rate": 0.00026600745268399583,
+      "loss": 2.6376,
+      "step": 53500
+    },
+    {
+      "epoch": 0.48050659379879795,
+      "grad_norm": 0.12686701118946075,
+      "learning_rate": 0.0002653047444346556,
+      "loss": 2.6426,
+      "step": 53600
+    },
+    {
+      "epoch": 0.48140306132454197,
+      "grad_norm": 0.13109557330608368,
+      "learning_rate": 0.00026460191478977203,
+      "loss": 2.6413,
+      "step": 53700
+    },
+    {
+      "epoch": 0.482299528850286,
+      "grad_norm": 0.12681497633457184,
+      "learning_rate": 0.00026389896932411234,
+      "loss": 2.6469,
+      "step": 53800
+    },
+    {
+      "epoch": 0.48319599637603,
+      "grad_norm": 0.12837442755699158,
+      "learning_rate": 0.0002631959136133624,
+      "loss": 2.6444,
+      "step": 53900
+    },
+    {
+      "epoch": 0.4840924639017741,
+      "grad_norm": 0.13397662341594696,
+      "learning_rate": 0.0002624927532340825,
+      "loss": 2.6385,
+      "step": 54000
+    },
+    {
+      "epoch": 0.4849889314275181,
+      "grad_norm": 0.1271597146987915,
+      "learning_rate": 0.00026178949376366316,
+      "loss": 2.637,
+      "step": 54100
+    },
+    {
+      "epoch": 0.4858853989532621,
+      "grad_norm": 0.12960323691368103,
+      "learning_rate": 0.00026108614078028077,
+      "loss": 2.6347,
+      "step": 54200
+    },
+    {
+      "epoch": 0.4867818664790061,
+      "grad_norm": 0.13047392666339874,
+      "learning_rate": 0.0002603826998628536,
+      "loss": 2.6414,
+      "step": 54300
+    },
+    {
+      "epoch": 0.48767833400475014,
+      "grad_norm": 0.1311124563217163,
+      "learning_rate": 0.0002596791765909973,
+      "loss": 2.6422,
+      "step": 54400
+    },
+    {
+      "epoch": 0.4885748015304942,
+      "grad_norm": 0.13057972490787506,
+      "learning_rate": 0.00025897557654498086,
+      "loss": 2.6416,
+      "step": 54500
+    },
+    {
+      "epoch": 0.4894712690562382,
+      "grad_norm": 0.12886947393417358,
+      "learning_rate": 0.0002582719053056822,
+      "loss": 2.6402,
+      "step": 54600
+    },
+    {
+      "epoch": 0.49036773658198224,
+      "grad_norm": 0.12779085338115692,
+      "learning_rate": 0.00025756816845454384,
+      "loss": 2.6349,
+      "step": 54700
+    },
+    {
+      "epoch": 0.49126420410772625,
+      "grad_norm": 0.12994657456874847,
+      "learning_rate": 0.0002568643715735288,
+      "loss": 2.6384,
+      "step": 54800
+    },
+    {
+      "epoch": 0.49216067163347027,
+      "grad_norm": 0.1322993040084839,
+      "learning_rate": 0.00025616052024507625,
+      "loss": 2.6413,
+      "step": 54900
+    },
+    {
+      "epoch": 0.4930571391592143,
+      "grad_norm": 0.1368289738893509,
+      "learning_rate": 0.00025545662005205716,
+      "loss": 2.641,
+      "step": 55000
+    },
+    {
+      "epoch": 0.49395360668495836,
+      "grad_norm": 0.1309647113084793,
+      "learning_rate": 0.00025475267657773027,
+      "loss": 2.6389,
+      "step": 55100
+    },
+    {
+      "epoch": 0.4948500742107024,
+      "grad_norm": 0.13188649713993073,
+      "learning_rate": 0.0002540486954056975,
+      "loss": 2.6445,
+      "step": 55200
+    },
+    {
+      "epoch": 0.4957465417364464,
+      "grad_norm": 0.13085578382015228,
+      "learning_rate": 0.0002533446821198597,
+      "loss": 2.6373,
+      "step": 55300
+    },
+    {
+      "epoch": 0.4966430092621904,
+      "grad_norm": 0.13119032979011536,
+      "learning_rate": 0.0002526406423043725,
+      "loss": 2.6449,
+      "step": 55400
+    },
+    {
+      "epoch": 0.4975394767879344,
+      "grad_norm": 0.1324346512556076,
+      "learning_rate": 0.0002519365815436021,
+      "loss": 2.6332,
+      "step": 55500
+    },
+    {
+      "epoch": 0.4984359443136785,
+      "grad_norm": 0.13353855907917023,
+      "learning_rate": 0.0002512325054220807,
+      "loss": 2.6365,
+      "step": 55600
+    },
+    {
+      "epoch": 0.4993324118394225,
+      "grad_norm": 0.13728708028793335,
+      "learning_rate": 0.0002505284195244624,
+      "loss": 2.6378,
+      "step": 55700
+    },
+    {
+      "epoch": 0.5002288793651665,
+      "grad_norm": 0.12542614340782166,
+      "learning_rate": 0.0002498243294354787,
+      "loss": 2.6436,
+      "step": 55800
+    },
+    {
+      "epoch": 0.5011253468909106,
+      "grad_norm": 0.12808193266391754,
+      "learning_rate": 0.0002491202407398945,
+      "loss": 2.6439,
+      "step": 55900
+    },
+    {
+      "epoch": 0.5020218144166546,
+      "grad_norm": 0.1336560845375061,
+      "learning_rate": 0.0002484161590224637,
+      "loss": 2.6317,
+      "step": 56000
+    },
+    {
+      "epoch": 0.5029182819423986,
+      "grad_norm": 0.1294037103652954,
+      "learning_rate": 0.0002477120898678847,
+      "loss": 2.6397,
+      "step": 56100
+    },
+    {
+      "epoch": 0.5038147494681426,
+      "grad_norm": 0.12634062767028809,
+      "learning_rate": 0.0002470080388607563,
+      "loss": 2.6345,
+      "step": 56200
+    },
+    {
+      "epoch": 0.5047112169938867,
+      "grad_norm": 0.12850134074687958,
+      "learning_rate": 0.0002463040115855333,
+      "loss": 2.6355,
+      "step": 56300
+    },
+    {
+      "epoch": 0.5056076845196307,
+      "grad_norm": 0.13224093616008759,
+      "learning_rate": 0.00024560001362648233,
+      "loss": 2.6333,
+      "step": 56400
+    },
+    {
+      "epoch": 0.5065041520453747,
+      "grad_norm": 0.129663348197937,
+      "learning_rate": 0.00024489605056763757,
+      "loss": 2.6363,
+      "step": 56500
+    },
+    {
+      "epoch": 0.5074006195711187,
+      "grad_norm": 0.1328994631767273,
+      "learning_rate": 0.00024419212799275627,
+      "loss": 2.6307,
+      "step": 56600
+    },
+    {
+      "epoch": 0.5082970870968627,
+      "grad_norm": 0.12859898805618286,
+      "learning_rate": 0.0002434882514852745,
+      "loss": 2.638,
+      "step": 56700
+    },
+    {
+      "epoch": 0.5091935546226067,
+      "grad_norm": 0.13029265403747559,
+      "learning_rate": 0.0002427844266282631,
+      "loss": 2.6369,
+      "step": 56800
+    },
+    {
+      "epoch": 0.5100900221483509,
+      "grad_norm": 0.1283547580242157,
+      "learning_rate": 0.00024208065900438309,
+      "loss": 2.634,
+      "step": 56900
+    },
+    {
+      "epoch": 0.5109864896740949,
+      "grad_norm": 0.12610678374767303,
+      "learning_rate": 0.00024137695419584163,
+      "loss": 2.6389,
+      "step": 57000
+    },
+    {
+      "epoch": 0.5118829571998389,
+      "grad_norm": 0.12752105295658112,
+      "learning_rate": 0.0002406733177843475,
+      "loss": 2.6364,
+      "step": 57100
+    },
+    {
+      "epoch": 0.5127794247255829,
+      "grad_norm": 0.13300961256027222,
+      "learning_rate": 0.0002399697553510671,
+      "loss": 2.6272,
+      "step": 57200
+    },
+    {
+      "epoch": 0.5136758922513269,
+      "grad_norm": 0.1362060010433197,
+      "learning_rate": 0.00023926627247657987,
+      "loss": 2.6327,
+      "step": 57300
+    },
+    {
+      "epoch": 0.5145723597770709,
+      "grad_norm": 0.12719836831092834,
+      "learning_rate": 0.00023856287474083437,
+      "loss": 2.6335,
+      "step": 57400
+    },
+    {
+      "epoch": 0.515468827302815,
+      "grad_norm": 0.13754625618457794,
+      "learning_rate": 0.0002378595677231038,
+      "loss": 2.627,
+      "step": 57500
+    },
+    {
+      "epoch": 0.516365294828559,
+      "grad_norm": 0.13351771235466003,
+      "learning_rate": 0.00023715635700194185,
+      "loss": 2.641,
+      "step": 57600
+    },
+    {
+      "epoch": 0.517261762354303,
+      "grad_norm": 0.12546195089817047,
+      "learning_rate": 0.0002364532481551382,
+      "loss": 2.639,
+      "step": 57700
+    },
+    {
+      "epoch": 0.518158229880047,
+      "grad_norm": 0.13150018453598022,
+      "learning_rate": 0.00023575024675967464,
+      "loss": 2.6328,
+      "step": 57800
+    },
+    {
+      "epoch": 0.519054697405791,
+      "grad_norm": 0.12908001244068146,
+      "learning_rate": 0.00023504735839168062,
+      "loss": 2.6236,
+      "step": 57900
+    },
+    {
+      "epoch": 0.5199511649315351,
+      "grad_norm": 0.12657436728477478,
+      "learning_rate": 0.00023434458862638919,
+      "loss": 2.6363,
+      "step": 58000
+    },
+    {
+      "epoch": 0.5208476324572792,
+      "grad_norm": 0.13314677774906158,
+      "learning_rate": 0.0002336419430380924,
+      "loss": 2.6304,
+      "step": 58100
+    },
+    {
+      "epoch": 0.5217440999830232,
+      "grad_norm": 0.1295042335987091,
+      "learning_rate": 0.0002329394272000976,
+      "loss": 2.6269,
+      "step": 58200
+    },
+    {
+      "epoch": 0.5226405675087672,
+      "grad_norm": 0.13147129118442535,
+      "learning_rate": 0.00023223704668468275,
+      "loss": 2.6355,
+      "step": 58300
+    },
+    {
+      "epoch": 0.5235370350345112,
+      "grad_norm": 0.13861297070980072,
+      "learning_rate": 0.0002315348070630526,
+      "loss": 2.6307,
+      "step": 58400
+    },
+    {
+      "epoch": 0.5244335025602552,
+      "grad_norm": 0.12782630324363708,
+      "learning_rate": 0.0002308327139052943,
+      "loss": 2.6396,
+      "step": 58500
+    },
+    {
+      "epoch": 0.5253299700859992,
+      "grad_norm": 0.1270517110824585,
+      "learning_rate": 0.00023013077278033334,
+      "loss": 2.6343,
+      "step": 58600
+    },
+    {
+      "epoch": 0.5262264376117433,
+      "grad_norm": 0.12780235707759857,
+      "learning_rate": 0.00022942898925588912,
+      "loss": 2.635,
+      "step": 58700
+    },
+    {
+      "epoch": 0.5271229051374873,
+      "grad_norm": 0.12590976059436798,
+      "learning_rate": 0.00022872736889843116,
+      "loss": 2.6274,
+      "step": 58800
+    },
+    {
+      "epoch": 0.5280193726632313,
+      "grad_norm": 0.1275419443845749,
+      "learning_rate": 0.0002280259172731346,
+      "loss": 2.6303,
+      "step": 58900
+    },
+    {
+      "epoch": 0.5289158401889753,
+      "grad_norm": 0.12726274132728577,
+      "learning_rate": 0.00022732463994383641,
+      "loss": 2.6297,
+      "step": 59000
+    },
+    {
+      "epoch": 0.5298123077147194,
+      "grad_norm": 0.13650214672088623,
+      "learning_rate": 0.00022662354247299086,
+      "loss": 2.6233,
+      "step": 59100
+    },
+    {
+      "epoch": 0.5307087752404634,
+      "grad_norm": 0.12631866335868835,
+      "learning_rate": 0.00022592263042162577,
+      "loss": 2.6333,
+      "step": 59200
+    },
+    {
+      "epoch": 0.5316052427662075,
+      "grad_norm": 0.12630638480186462,
+      "learning_rate": 0.00022522190934929801,
+      "loss": 2.6275,
+      "step": 59300
+    },
+    {
+      "epoch": 0.5325017102919515,
+      "grad_norm": 0.12765736877918243,
+      "learning_rate": 0.00022452138481404986,
+      "loss": 2.6306,
+      "step": 59400
+    },
+    {
+      "epoch": 0.5333981778176955,
+      "grad_norm": 0.12750397622585297,
+      "learning_rate": 0.00022382106237236457,
+      "loss": 2.6271,
+      "step": 59500
+    },
+    {
+      "epoch": 0.5342946453434395,
+      "grad_norm": 0.13162820041179657,
+      "learning_rate": 0.0002231209475791225,
+      "loss": 2.6287,
+      "step": 59600
+    },
+    {
+      "epoch": 0.5351911128691835,
+      "grad_norm": 0.13135115802288055,
+      "learning_rate": 0.0002224210459875568,
+      "loss": 2.6322,
+      "step": 59700
+    },
+    {
+      "epoch": 0.5360875803949275,
+      "grad_norm": 0.12867547571659088,
+      "learning_rate": 0.00022172136314920968,
+      "loss": 2.6353,
+      "step": 59800
+    },
+    {
+      "epoch": 0.5369840479206716,
+      "grad_norm": 0.1288844645023346,
+      "learning_rate": 0.00022102190461388818,
+      "loss": 2.6355,
+      "step": 59900
+    },
+    {
+      "epoch": 0.5378805154464156,
+      "grad_norm": 0.12920548021793365,
+      "learning_rate": 0.00022032267592962023,
+      "loss": 2.6274,
+      "step": 60000
+    },
+    {
+      "epoch": 0.5387769829721596,
+      "grad_norm": 0.1328190267086029,
+      "learning_rate": 0.0002196236826426104,
+      "loss": 2.6318,
+      "step": 60100
+    },
+    {
+      "epoch": 0.5396734504979037,
+      "grad_norm": 0.12695330381393433,
+      "learning_rate": 0.00021892493029719652,
+      "loss": 2.6287,
+      "step": 60200
+    },
+    {
+      "epoch": 0.5405699180236477,
+      "grad_norm": 0.13476692140102386,
+      "learning_rate": 0.00021822642443580483,
+      "loss": 2.6335,
+      "step": 60300
+    },
+    {
+      "epoch": 0.5414663855493917,
+      "grad_norm": 0.13216422498226166,
+      "learning_rate": 0.00021752817059890673,
+      "loss": 2.6262,
+      "step": 60400
+    },
+    {
+      "epoch": 0.5423628530751358,
+      "grad_norm": 0.13027572631835938,
+      "learning_rate": 0.00021683017432497464,
+      "loss": 2.6189,
+      "step": 60500
+    },
+    {
+      "epoch": 0.5432593206008798,
+      "grad_norm": 0.12778045237064362,
+      "learning_rate": 0.00021613244115043802,
+      "loss": 2.6295,
+      "step": 60600
+    },
+    {
+      "epoch": 0.5441557881266238,
+      "grad_norm": 0.13152672350406647,
+      "learning_rate": 0.00021543497660963924,
+      "loss": 2.6303,
+      "step": 60700
+    },
+    {
+      "epoch": 0.5450522556523678,
+      "grad_norm": 0.12792283296585083,
+      "learning_rate": 0.00021473778623479006,
+      "loss": 2.6272,
+      "step": 60800
+    },
+    {
+      "epoch": 0.5459487231781118,
+      "grad_norm": 0.1273530125617981,
+      "learning_rate": 0.00021404087555592762,
+      "loss": 2.625,
+      "step": 60900
+    },
+    {
+      "epoch": 0.5468451907038558,
+      "grad_norm": 0.1318267434835434,
+      "learning_rate": 0.0002133442501008705,
+      "loss": 2.6254,
+      "step": 61000
+    },
+    {
+      "epoch": 0.5477416582295999,
+      "grad_norm": 0.13691310584545135,
+      "learning_rate": 0.00021264791539517486,
+      "loss": 2.6211,
+      "step": 61100
+    },
+    {
+      "epoch": 0.5486381257553439,
+      "grad_norm": 0.12379534542560577,
+      "learning_rate": 0.00021195187696209072,
+      "loss": 2.6248,
+      "step": 61200
+    },
+    {
+      "epoch": 0.549534593281088,
+      "grad_norm": 0.13373887538909912,
+      "learning_rate": 0.00021125614032251822,
+      "loss": 2.6304,
+      "step": 61300
+    },
+    {
+      "epoch": 0.550431060806832,
+      "grad_norm": 0.13529092073440552,
+      "learning_rate": 0.00021056071099496333,
+      "loss": 2.6268,
+      "step": 61400
+    },
+    {
+      "epoch": 0.551327528332576,
+      "grad_norm": 0.13044504821300507,
+      "learning_rate": 0.0002098655944954948,
+      "loss": 2.6265,
+      "step": 61500
+    },
+    {
+      "epoch": 0.55222399585832,
+      "grad_norm": 0.12602747976779938,
+      "learning_rate": 0.00020917079633770005,
+      "loss": 2.6248,
+      "step": 61600
+    },
+    {
+      "epoch": 0.5531204633840641,
+      "grad_norm": 0.12802156805992126,
+      "learning_rate": 0.00020847632203264128,
+      "loss": 2.6194,
+      "step": 61700
+    },
+    {
+      "epoch": 0.5540169309098081,
+      "grad_norm": 0.1250976324081421,
+      "learning_rate": 0.000207782177088812,
+      "loss": 2.6297,
+      "step": 61800
+    },
+    {
+      "epoch": 0.5549133984355521,
+      "grad_norm": 0.13608118891716003,
+      "learning_rate": 0.00020708836701209332,
+      "loss": 2.6286,
+      "step": 61900
+    },
+    {
+      "epoch": 0.5558098659612961,
+      "grad_norm": 0.1274801641702652,
+      "learning_rate": 0.00020639489730571014,
+      "loss": 2.6283,
+      "step": 62000
+    },
+    {
+      "epoch": 0.5567063334870401,
+      "grad_norm": 0.12696389853954315,
+      "learning_rate": 0.0002057017734701877,
+      "loss": 2.6262,
+      "step": 62100
+    },
+    {
+      "epoch": 0.5576028010127841,
+      "grad_norm": 0.12838256359100342,
+      "learning_rate": 0.00020501592697065653,
+      "loss": 2.6261,
+      "step": 62200
+    },
+    {
+      "epoch": 0.5584992685385282,
+      "grad_norm": 0.12481874227523804,
+      "learning_rate": 0.00020432350777158638,
+      "loss": 2.6197,
+      "step": 62300
+    },
+    {
+      "epoch": 0.5593957360642723,
+      "grad_norm": 0.13098376989364624,
+      "learning_rate": 0.00020363145087340989,
+      "loss": 2.6218,
+      "step": 62400
+    },
+    {
+      "epoch": 0.5602922035900163,
+      "grad_norm": 0.1315186321735382,
+      "learning_rate": 0.0002029397617654461,
+      "loss": 2.6229,
+      "step": 62500
+    },
+    {
+      "epoch": 0.5611886711157603,
+      "grad_norm": 0.13062401115894318,
+      "learning_rate": 0.00020224844593409665,
+      "loss": 2.6224,
+      "step": 62600
+    },
+    {
+      "epoch": 0.5620851386415043,
+      "grad_norm": 0.12941262125968933,
+      "learning_rate": 0.00020155750886280245,
+      "loss": 2.6264,
+      "step": 62700
+    },
+    {
+      "epoch": 0.5629816061672483,
+      "grad_norm": 0.12870369851589203,
+      "learning_rate": 0.00020086695603199994,
+      "loss": 2.6176,
+      "step": 62800
+    },
+    {
+      "epoch": 0.5638780736929924,
+      "grad_norm": 0.12999162077903748,
+      "learning_rate": 0.00020017679291907804,
+      "loss": 2.6217,
+      "step": 62900
+    },
+    {
+      "epoch": 0.5647745412187364,
+      "grad_norm": 0.12885361909866333,
+      "learning_rate": 0.00019948702499833436,
+      "loss": 2.6248,
+      "step": 63000
+    },
+    {
+      "epoch": 0.5656710087444804,
+      "grad_norm": 0.12850242853164673,
+      "learning_rate": 0.00019879765774093188,
+      "loss": 2.6276,
+      "step": 63100
+    },
+    {
+      "epoch": 0.5665674762702244,
+      "grad_norm": 0.12949827313423157,
+      "learning_rate": 0.0001981086966148556,
+      "loss": 2.6191,
+      "step": 63200
+    },
+    {
+      "epoch": 0.5674639437959684,
+      "grad_norm": 0.1278415471315384,
+      "learning_rate": 0.00019742014708486904,
+      "loss": 2.6321,
+      "step": 63300
+    },
+    {
+      "epoch": 0.5683604113217124,
+      "grad_norm": 0.12578001618385315,
+      "learning_rate": 0.00019673201461247125,
+      "loss": 2.616,
+      "step": 63400
+    },
+    {
+      "epoch": 0.5692568788474566,
+      "grad_norm": 0.12869875133037567,
+      "learning_rate": 0.00019604430465585277,
+      "loss": 2.6195,
+      "step": 63500
+    },
+    {
+      "epoch": 0.5701533463732006,
+      "grad_norm": 0.12808062136173248,
+      "learning_rate": 0.0001953570226698532,
+      "loss": 2.6231,
+      "step": 63600
+    },
+    {
+      "epoch": 0.5710498138989446,
+      "grad_norm": 0.13366073369979858,
+      "learning_rate": 0.0001946701741059174,
+      "loss": 2.6234,
+      "step": 63700
+    },
+    {
+      "epoch": 0.5719462814246886,
+      "grad_norm": 0.12796130776405334,
+      "learning_rate": 0.00019398376441205227,
+      "loss": 2.6268,
+      "step": 63800
+    },
+    {
+      "epoch": 0.5728427489504326,
+      "grad_norm": 0.13759686052799225,
+      "learning_rate": 0.00019329779903278375,
+      "loss": 2.619,
+      "step": 63900
+    },
+    {
+      "epoch": 0.5737392164761766,
+      "grad_norm": 0.12661823630332947,
+      "learning_rate": 0.00019261228340911352,
+      "loss": 2.6152,
+      "step": 64000
+    },
+    {
+      "epoch": 0.5746356840019207,
+      "grad_norm": 0.13385765254497528,
+      "learning_rate": 0.00019192722297847587,
+      "loss": 2.6217,
+      "step": 64100
+    },
+    {
+      "epoch": 0.5755321515276647,
+      "grad_norm": 0.1298641860485077,
+      "learning_rate": 0.00019124262317469443,
+      "loss": 2.622,
+      "step": 64200
+    },
+    {
+      "epoch": 0.5764286190534087,
+      "grad_norm": 0.12806908786296844,
+      "learning_rate": 0.00019056532844060364,
+      "loss": 2.6181,
+      "step": 64300
+    },
+    {
+      "epoch": 0.5773250865791527,
+      "grad_norm": 0.12761425971984863,
+      "learning_rate": 0.00018988849575433026,
+      "loss": 2.6178,
+      "step": 64400
+    },
+    {
+      "epoch": 0.5782215541048968,
+      "grad_norm": 0.1286136656999588,
+      "learning_rate": 0.00018920530080606497,
+      "loss": 2.6202,
+      "step": 64500
+    },
+    {
+      "epoch": 0.5791180216306409,
+      "grad_norm": 0.12874256074428558,
+      "learning_rate": 0.0001885225880746421,
+      "loss": 2.6064,
+      "step": 64600
+    },
+    {
+      "epoch": 0.5800144891563849,
+      "grad_norm": 0.12678521871566772,
+      "learning_rate": 0.00018784036297526366,
+      "loss": 2.6256,
+      "step": 64700
+    },
+    {
+      "epoch": 0.5809109566821289,
+      "grad_norm": 0.1318819671869278,
+      "learning_rate": 0.0001871586309192639,
+      "loss": 2.6207,
+      "step": 64800
+    },
+    {
+      "epoch": 0.5818074242078729,
+      "grad_norm": 0.13057227432727814,
+      "learning_rate": 0.00018647739731406606,
+      "loss": 2.6188,
+      "step": 64900
+    },
+    {
+      "epoch": 0.5827038917336169,
+      "grad_norm": 0.1309261918067932,
+      "learning_rate": 0.00018579666756314,
+      "loss": 2.6225,
+      "step": 65000
+    },
+    {
+      "epoch": 0.5836003592593609,
+      "grad_norm": 0.12833261489868164,
+      "learning_rate": 0.0001851164470659589,
+      "loss": 2.6151,
+      "step": 65100
+    },
+    {
+      "epoch": 0.584496826785105,
+      "grad_norm": 0.1268189400434494,
+      "learning_rate": 0.00018443674121795678,
+      "loss": 2.6177,
+      "step": 65200
+    },
+    {
+      "epoch": 0.585393294310849,
+      "grad_norm": 0.1255330890417099,
+      "learning_rate": 0.00018375755541048526,
+      "loss": 2.6206,
+      "step": 65300
+    },
+    {
+      "epoch": 0.586289761836593,
+      "grad_norm": 0.1327618807554245,
+      "learning_rate": 0.0001830788950307713,
+      "loss": 2.6236,
+      "step": 65400
+    },
+    {
+      "epoch": 0.587186229362337,
+      "grad_norm": 0.12842676043510437,
+      "learning_rate": 0.0001824007654618739,
+      "loss": 2.6176,
+      "step": 65500
+    },
+    {
+      "epoch": 0.5880826968880811,
+      "grad_norm": 0.1308692991733551,
+      "learning_rate": 0.00018172317208264212,
+      "loss": 2.6171,
+      "step": 65600
+    },
+    {
+      "epoch": 0.5889791644138251,
+      "grad_norm": 0.12851682305335999,
+      "learning_rate": 0.0001810461202676717,
+      "loss": 2.6164,
+      "step": 65700
+    },
+    {
+      "epoch": 0.5898756319395692,
+      "grad_norm": 0.14155201613903046,
+      "learning_rate": 0.00018036961538726314,
+      "loss": 2.6198,
+      "step": 65800
+    },
+    {
+      "epoch": 0.5907720994653132,
+      "grad_norm": 0.1263512820005417,
+      "learning_rate": 0.0001796936628073782,
+      "loss": 2.6202,
+      "step": 65900
+    },
+    {
+      "epoch": 0.5916685669910572,
+      "grad_norm": 0.12849842011928558,
+      "learning_rate": 0.00017901826788959825,
+      "loss": 2.6154,
+      "step": 66000
+    },
+    {
+      "epoch": 0.5925650345168012,
+      "grad_norm": 0.12791509926319122,
+      "learning_rate": 0.0001783434359910811,
+      "loss": 2.6201,
+      "step": 66100
+    },
+    {
+      "epoch": 0.5934615020425452,
+      "grad_norm": 0.12755149602890015,
+      "learning_rate": 0.00017766917246451902,
+      "loss": 2.6166,
+      "step": 66200
+    },
+    {
+      "epoch": 0.5943579695682892,
+      "grad_norm": 0.13783405721187592,
+      "learning_rate": 0.00017699548265809578,
+      "loss": 2.6165,
+      "step": 66300
+    },
+    {
+      "epoch": 0.5952544370940333,
+      "grad_norm": 0.1294228881597519,
+      "learning_rate": 0.00017632237191544462,
+      "loss": 2.6155,
+      "step": 66400
+    },
+    {
+      "epoch": 0.5961509046197773,
+      "grad_norm": 0.13272565603256226,
+      "learning_rate": 0.00017564984557560555,
+      "loss": 2.6116,
+      "step": 66500
+    },
+    {
+      "epoch": 0.5970473721455213,
+      "grad_norm": 0.13658390939235687,
+      "learning_rate": 0.00017497790897298327,
+      "loss": 2.6159,
+      "step": 66600
+    },
+    {
+      "epoch": 0.5979438396712654,
+      "grad_norm": 0.13049334287643433,
+      "learning_rate": 0.00017430656743730477,
+      "loss": 2.6191,
+      "step": 66700
+    },
+    {
+      "epoch": 0.5988403071970094,
+      "grad_norm": 0.13098062574863434,
+      "learning_rate": 0.00017363582629357705,
+      "loss": 2.618,
+      "step": 66800
+    },
+    {
+      "epoch": 0.5997367747227534,
+      "grad_norm": 0.12643533945083618,
+      "learning_rate": 0.00017296569086204479,
+      "loss": 2.6172,
+      "step": 66900
+    },
+    {
+      "epoch": 0.6006332422484975,
+      "grad_norm": 0.12678323686122894,
+      "learning_rate": 0.00017229616645814813,
+      "loss": 2.6185,
+      "step": 67000
+    },
+    {
+      "epoch": 0.6015297097742415,
+      "grad_norm": 0.12956561148166656,
+      "learning_rate": 0.00017162725839248077,
+      "loss": 2.6089,
+      "step": 67100
+    },
+    {
+      "epoch": 0.6024261772999855,
+      "grad_norm": 0.129245787858963,
+      "learning_rate": 0.00017095897197074773,
+      "loss": 2.6043,
+      "step": 67200
+    },
+    {
+      "epoch": 0.6033226448257295,
+      "grad_norm": 0.12439849972724915,
+      "learning_rate": 0.00017029131249372305,
+      "loss": 2.6174,
+      "step": 67300
+    },
+    {
+      "epoch": 0.6042191123514735,
+      "grad_norm": 0.1366392970085144,
+      "learning_rate": 0.00016962428525720807,
+      "loss": 2.623,
+      "step": 67400
+    },
+    {
+      "epoch": 0.6051155798772175,
+      "grad_norm": 0.12857592105865479,
+      "learning_rate": 0.00016895789555198908,
+      "loss": 2.61,
+      "step": 67500
+    },
+    {
+      "epoch": 0.6060120474029616,
+      "grad_norm": 0.1319228857755661,
+      "learning_rate": 0.00016829214866379572,
+      "loss": 2.611,
+      "step": 67600
+    },
+    {
+      "epoch": 0.6069085149287056,
+      "grad_norm": 0.1241278126835823,
+      "learning_rate": 0.00016762704987325884,
+      "loss": 2.6049,
+      "step": 67700
+    },
+    {
+      "epoch": 0.6078049824544497,
+      "grad_norm": 0.13301433622837067,
+      "learning_rate": 0.00016696260445586865,
+      "loss": 2.6089,
+      "step": 67800
+    },
+    {
+      "epoch": 0.6087014499801937,
+      "grad_norm": 0.1289006769657135,
+      "learning_rate": 0.00016629881768193271,
+      "loss": 2.6123,
+      "step": 67900
+    },
+    {
+      "epoch": 0.6095979175059377,
+      "grad_norm": 0.1281334012746811,
+      "learning_rate": 0.00016563569481653457,
+      "loss": 2.6086,
+      "step": 68000
+    },
+    {
+      "epoch": 0.6104943850316817,
+      "grad_norm": 0.13358446955680847,
+      "learning_rate": 0.00016497324111949146,
+      "loss": 2.6162,
+      "step": 68100
+    },
+    {
+      "epoch": 0.6113908525574258,
+      "grad_norm": 0.12993699312210083,
+      "learning_rate": 0.00016431146184531304,
+      "loss": 2.6077,
+      "step": 68200
+    },
+    {
+      "epoch": 0.6122873200831698,
+      "grad_norm": 0.12562768161296844,
+      "learning_rate": 0.00016365036224315948,
+      "loss": 2.6099,
+      "step": 68300
+    },
+    {
+      "epoch": 0.6131837876089138,
+      "grad_norm": 0.12500709295272827,
+      "learning_rate": 0.00016298994755679986,
+      "loss": 2.6186,
+      "step": 68400
+    },
+    {
+      "epoch": 0.6140802551346578,
+      "grad_norm": 0.13538555800914764,
+      "learning_rate": 0.00016233681683644314,
+      "loss": 2.6057,
+      "step": 68500
+    },
+    {
+      "epoch": 0.6149767226604018,
+      "grad_norm": 0.13071191310882568,
+      "learning_rate": 0.00016167778071145164,
+      "loss": 2.6108,
+      "step": 68600
+    },
+    {
+      "epoch": 0.6158731901861458,
+      "grad_norm": 0.12502072751522064,
+      "learning_rate": 0.00016101944514855299,
+      "loss": 2.6098,
+      "step": 68700
+    },
+    {
+      "epoch": 0.6167696577118899,
+      "grad_norm": 0.130848690867424,
+      "learning_rate": 0.0001603618153695922,
+      "loss": 2.6067,
+      "step": 68800
+    },
+    {
+      "epoch": 0.617666125237634,
+      "grad_norm": 0.12595491111278534,
+      "learning_rate": 0.0001597048965908161,
+      "loss": 2.6009,
+      "step": 68900
+    },
+    {
+      "epoch": 0.618562592763378,
+      "grad_norm": 0.13049688935279846,
+      "learning_rate": 0.0001590486940228319,
+      "loss": 2.6069,
+      "step": 69000
+    },
+    {
+      "epoch": 0.619459060289122,
+      "grad_norm": 0.1312711089849472,
+      "learning_rate": 0.00015839321287056585,
+      "loss": 2.6088,
+      "step": 69100
+    },
+    {
+      "epoch": 0.620355527814866,
+      "grad_norm": 0.13168184459209442,
+      "learning_rate": 0.00015773845833322208,
+      "loss": 2.6188,
+      "step": 69200
+    },
+    {
+      "epoch": 0.62125199534061,
+      "grad_norm": 0.12749746441841125,
+      "learning_rate": 0.00015708443560424141,
+      "loss": 2.6151,
+      "step": 69300
+    },
+    {
+      "epoch": 0.6221484628663541,
+      "grad_norm": 0.13357709348201752,
+      "learning_rate": 0.0001564311498712599,
+      "loss": 2.6064,
+      "step": 69400
+    },
+    {
+      "epoch": 0.6230449303920981,
+      "grad_norm": 0.12775705754756927,
+      "learning_rate": 0.00015577860631606783,
+      "loss": 2.6118,
+      "step": 69500
+    },
+    {
+      "epoch": 0.6239413979178421,
+      "grad_norm": 0.13166511058807373,
+      "learning_rate": 0.00015512681011456859,
+      "loss": 2.602,
+      "step": 69600
+    },
+    {
+      "epoch": 0.6248378654435861,
+      "grad_norm": 0.1261209398508072,
+      "learning_rate": 0.00015447576643673778,
+      "loss": 2.61,
+      "step": 69700
+    },
+    {
+      "epoch": 0.6257343329693301,
+      "grad_norm": 0.12618456780910492,
+      "learning_rate": 0.00015382548044658194,
+      "loss": 2.6102,
+      "step": 69800
+    },
+    {
+      "epoch": 0.6266308004950741,
+      "grad_norm": 0.12650653719902039,
+      "learning_rate": 0.00015317595730209783,
+      "loss": 2.6137,
+      "step": 69900
+    },
+    {
+      "epoch": 0.6275272680208183,
+      "grad_norm": 0.12881776690483093,
+      "learning_rate": 0.0001525272021552313,
+      "loss": 2.6134,
+      "step": 70000
+    },
+    {
+      "epoch": 0.6284237355465623,
+      "grad_norm": 0.1257714331150055,
+      "learning_rate": 0.00015187922015183638,
+      "loss": 2.6172,
+      "step": 70100
+    },
+    {
+      "epoch": 0.6293202030723063,
+      "grad_norm": 0.12860926985740662,
+      "learning_rate": 0.00015123201643163495,
+      "loss": 2.6093,
+      "step": 70200
+    },
+    {
+      "epoch": 0.6302166705980503,
+      "grad_norm": 0.1268460750579834,
+      "learning_rate": 0.0001505855961281754,
+      "loss": 2.6065,
+      "step": 70300
+    },
+    {
+      "epoch": 0.6311131381237943,
+      "grad_norm": 0.1278506964445114,
+      "learning_rate": 0.00014993996436879232,
+      "loss": 2.6117,
+      "step": 70400
+    },
+    {
+      "epoch": 0.6320096056495383,
+      "grad_norm": 0.1345282644033432,
+      "learning_rate": 0.00014930157071006556,
+      "loss": 2.6175,
+      "step": 70500
+    },
+    {
+      "epoch": 0.6329060731752824,
+      "grad_norm": 0.13263201713562012,
+      "learning_rate": 0.0001486575233826827,
+      "loss": 2.6045,
+      "step": 70600
+    },
+    {
+      "epoch": 0.6338025407010264,
+      "grad_norm": 0.12888827919960022,
+      "learning_rate": 0.00014801427989263638,
+      "loss": 2.606,
+      "step": 70700
+    },
+    {
+      "epoch": 0.6346990082267704,
+      "grad_norm": 0.12627027928829193,
+      "learning_rate": 0.00014737184534206293,
+      "loss": 2.6056,
+      "step": 70800
+    },
+    {
+      "epoch": 0.6355954757525144,
+      "grad_norm": 0.12864083051681519,
+      "learning_rate": 0.00014673022482668242,
+      "loss": 2.6085,
+      "step": 70900
+    },
+    {
+      "epoch": 0.6364919432782585,
+      "grad_norm": 0.1318010538816452,
+      "learning_rate": 0.00014608942343575788,
+      "loss": 2.6087,
+      "step": 71000
+    },
+    {
+      "epoch": 0.6373884108040025,
+      "grad_norm": 0.12685802578926086,
+      "learning_rate": 0.0001454494462520551,
+      "loss": 2.6069,
+      "step": 71100
+    },
+    {
+      "epoch": 0.6382848783297466,
+      "grad_norm": 0.12846295535564423,
+      "learning_rate": 0.0001448102983518025,
+      "loss": 2.6005,
+      "step": 71200
+    },
+    {
+      "epoch": 0.6391813458554906,
+      "grad_norm": 0.1338125467300415,
+      "learning_rate": 0.00014417198480465068,
+      "loss": 2.6043,
+      "step": 71300
+    },
+    {
+      "epoch": 0.6400778133812346,
+      "grad_norm": 0.1277831345796585,
+      "learning_rate": 0.00014353451067363227,
+      "loss": 2.6048,
+      "step": 71400
+    },
+    {
+      "epoch": 0.6409742809069786,
+      "grad_norm": 0.1271323561668396,
+      "learning_rate": 0.0001428978810151216,
+      "loss": 2.6066,
+      "step": 71500
+    },
+    {
+      "epoch": 0.6418707484327226,
+      "grad_norm": 0.12640222907066345,
+      "learning_rate": 0.00014226210087879494,
+      "loss": 2.6082,
+      "step": 71600
+    },
+    {
+      "epoch": 0.6427672159584666,
+      "grad_norm": 0.12706422805786133,
+      "learning_rate": 0.0001416271753075902,
+      "loss": 2.6041,
+      "step": 71700
+    },
+    {
+      "epoch": 0.6436636834842107,
+      "grad_norm": 0.12946538627147675,
+      "learning_rate": 0.0001409931093376667,
+      "loss": 2.6042,
+      "step": 71800
+    },
+    {
+      "epoch": 0.6445601510099547,
+      "grad_norm": 0.13379494845867157,
+      "learning_rate": 0.00014035990799836584,
+      "loss": 2.6064,
+      "step": 71900
+    },
+    {
+      "epoch": 0.6454566185356987,
+      "grad_norm": 0.13055512309074402,
+      "learning_rate": 0.00013972757631217075,
+      "loss": 2.6067,
+      "step": 72000
+    }
+  ],
+  "logging_steps": 100,
+  "max_steps": 111548,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 2000,
+  "total_flos": 3.857626077069312e+18,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}