{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.129305477131564,
  "eval_steps": 500,
  "global_step": 5000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00564652738565782,
      "grad_norm": 0.0635828971862793,
      "learning_rate": 1.2531328320802006e-05,
      "loss": 1.3836,
      "step": 25
    },
    {
      "epoch": 0.01129305477131564,
      "grad_norm": 0.2005225569009781,
      "learning_rate": 2.506265664160401e-05,
      "loss": 1.6497,
      "step": 50
    },
    {
      "epoch": 0.01693958215697346,
      "grad_norm": 0.07235410064458847,
      "learning_rate": 3.759398496240601e-05,
      "loss": 1.1768,
      "step": 75
    },
    {
      "epoch": 0.02258610954263128,
      "grad_norm": 0.17648495733737946,
      "learning_rate": 5.012531328320802e-05,
      "loss": 1.2146,
      "step": 100
    },
    {
      "epoch": 0.028232636928289104,
      "grad_norm": 0.06953724473714828,
      "learning_rate": 6.265664160401002e-05,
      "loss": 0.8725,
      "step": 125
    },
    {
      "epoch": 0.03387916431394692,
      "grad_norm": 0.12059248238801956,
      "learning_rate": 7.518796992481203e-05,
      "loss": 0.7548,
      "step": 150
    },
    {
      "epoch": 0.039525691699604744,
      "grad_norm": 0.0689113661646843,
      "learning_rate": 8.771929824561403e-05,
      "loss": 0.7917,
      "step": 175
    },
    {
      "epoch": 0.04517221908526256,
      "grad_norm": 0.16621273756027222,
      "learning_rate": 0.00010025062656641604,
      "loss": 0.6985,
      "step": 200
    },
    {
      "epoch": 0.050818746470920384,
      "grad_norm": 0.07838872820138931,
      "learning_rate": 0.00011278195488721806,
      "loss": 0.7874,
      "step": 225
    },
    {
      "epoch": 0.05646527385657821,
      "grad_norm": 0.11714337766170502,
      "learning_rate": 0.00012531328320802005,
      "loss": 0.689,
      "step": 250
    },
    {
      "epoch": 0.062111801242236024,
      "grad_norm": 0.07283364981412888,
      "learning_rate": 0.00013784461152882208,
      "loss": 0.7719,
      "step": 275
    },
    {
      "epoch": 0.06775832862789384,
      "grad_norm": 0.1499466449022293,
      "learning_rate": 0.00015037593984962405,
      "loss": 0.655,
      "step": 300
    },
    {
      "epoch": 0.07340485601355166,
      "grad_norm": 0.08195521682500839,
      "learning_rate": 0.00016290726817042608,
      "loss": 0.7508,
      "step": 325
    },
    {
      "epoch": 0.07905138339920949,
      "grad_norm": 0.1490202099084854,
      "learning_rate": 0.00017543859649122806,
      "loss": 0.6372,
      "step": 350
    },
    {
      "epoch": 0.08469791078486731,
      "grad_norm": 0.1617508977651596,
      "learning_rate": 0.00018796992481203009,
      "loss": 0.7396,
      "step": 375
    },
    {
      "epoch": 0.09034443817052512,
      "grad_norm": 0.180181622505188,
      "learning_rate": 0.00019999999702625888,
      "loss": 0.6318,
      "step": 400
    },
    {
      "epoch": 0.09599096555618294,
      "grad_norm": 0.08272858709096909,
      "learning_rate": 0.00019999798975772924,
      "loss": 0.724,
      "step": 425
    },
    {
      "epoch": 0.10163749294184077,
      "grad_norm": 0.120403952896595,
      "learning_rate": 0.00019999226539902187,
      "loss": 0.6271,
      "step": 450
    },
    {
      "epoch": 0.10728402032749859,
      "grad_norm": 0.07971920073032379,
      "learning_rate": 0.00019998282416292055,
      "loss": 0.7256,
      "step": 475
    },
    {
      "epoch": 0.11293054771315642,
      "grad_norm": 0.1328658014535904,
      "learning_rate": 0.00019996966640037166,
      "loss": 0.6231,
      "step": 500
    },
    {
      "epoch": 0.11857707509881422,
      "grad_norm": 0.07475866377353668,
      "learning_rate": 0.00019995279260047092,
      "loss": 0.7251,
      "step": 525
    },
    {
      "epoch": 0.12422360248447205,
      "grad_norm": 0.1371573656797409,
      "learning_rate": 0.00019993220339044524,
      "loss": 0.5907,
      "step": 550
    },
    {
      "epoch": 0.12987012987012986,
      "grad_norm": 0.07268711924552917,
      "learning_rate": 0.00019990789953562961,
      "loss": 0.7304,
      "step": 575
    },
    {
      "epoch": 0.13551665725578768,
      "grad_norm": 0.14224526286125183,
      "learning_rate": 0.0001998798819394383,
      "loss": 0.5931,
      "step": 600
    },
    {
      "epoch": 0.1411631846414455,
      "grad_norm": 0.0803467407822609,
      "learning_rate": 0.00019984815164333163,
      "loss": 0.7076,
      "step": 625
    },
    {
      "epoch": 0.14680971202710333,
      "grad_norm": 0.12445727735757828,
      "learning_rate": 0.00019981270982677698,
      "loss": 0.5566,
      "step": 650
    },
    {
      "epoch": 0.15245623941276115,
      "grad_norm": 0.07665792107582092,
      "learning_rate": 0.00019977355780720514,
      "loss": 0.6985,
      "step": 675
    },
    {
      "epoch": 0.15810276679841898,
      "grad_norm": 0.13053999841213226,
      "learning_rate": 0.00019973069703996125,
      "loss": 0.5901,
      "step": 700
    },
    {
      "epoch": 0.1637492941840768,
      "grad_norm": 0.07512692362070084,
      "learning_rate": 0.00019968412911825067,
      "loss": 0.7184,
      "step": 725
    },
    {
      "epoch": 0.16939582156973462,
      "grad_norm": 0.12033283710479736,
      "learning_rate": 0.00019963385577307987,
      "loss": 0.6013,
      "step": 750
    },
    {
      "epoch": 0.17504234895539245,
      "grad_norm": 0.0809774249792099,
      "learning_rate": 0.000199579878873192,
      "loss": 0.7054,
      "step": 775
    },
    {
      "epoch": 0.18068887634105024,
      "grad_norm": 0.12490582466125488,
      "learning_rate": 0.0001995222004249974,
      "loss": 0.5714,
      "step": 800
    },
    {
      "epoch": 0.18633540372670807,
      "grad_norm": 0.08289226144552231,
      "learning_rate": 0.00019946082257249912,
      "loss": 0.7304,
      "step": 825
    },
    {
      "epoch": 0.1919819311123659,
      "grad_norm": 0.14385385811328888,
      "learning_rate": 0.00019939574759721316,
      "loss": 0.5639,
      "step": 850
    },
    {
      "epoch": 0.1976284584980237,
      "grad_norm": 0.07623141258955002,
      "learning_rate": 0.00019932697791808366,
      "loss": 0.7126,
      "step": 875
    },
    {
      "epoch": 0.20327498588368154,
      "grad_norm": 0.11150185018777847,
      "learning_rate": 0.000199254516091393,
      "loss": 0.5903,
      "step": 900
    },
    {
      "epoch": 0.20892151326933936,
      "grad_norm": 0.07193930447101593,
      "learning_rate": 0.00019917836481066675,
      "loss": 0.6952,
      "step": 925
    },
    {
      "epoch": 0.21456804065499718,
      "grad_norm": 0.11242598295211792,
      "learning_rate": 0.00019909852690657359,
      "loss": 0.5853,
      "step": 950
    },
    {
      "epoch": 0.220214568040655,
      "grad_norm": 0.07508910447359085,
      "learning_rate": 0.0001990150053468201,
      "loss": 0.6969,
      "step": 975
    },
    {
      "epoch": 0.22586109542631283,
      "grad_norm": 0.11998436599969864,
      "learning_rate": 0.00019892780323604035,
      "loss": 0.5791,
      "step": 1000
    },
    {
      "epoch": 0.23150762281197063,
      "grad_norm": 0.07748444378376007,
      "learning_rate": 0.0001988369238156806,
      "loss": 0.7038,
      "step": 1025
    },
    {
      "epoch": 0.23715415019762845,
      "grad_norm": 0.19617140293121338,
      "learning_rate": 0.0001987423704638788,
      "loss": 0.5792,
      "step": 1050
    },
    {
      "epoch": 0.24280067758328627,
      "grad_norm": 0.08167731016874313,
      "learning_rate": 0.00019864414669533892,
      "loss": 0.6989,
      "step": 1075
    },
    {
      "epoch": 0.2484472049689441,
      "grad_norm": 0.11911099404096603,
      "learning_rate": 0.00019854225616120044,
      "loss": 0.5985,
      "step": 1100
    },
    {
      "epoch": 0.2540937323546019,
      "grad_norm": 0.0718325674533844,
      "learning_rate": 0.0001984367026489025,
      "loss": 0.6853,
      "step": 1125
    },
    {
      "epoch": 0.2597402597402597,
      "grad_norm": 0.12495870143175125,
      "learning_rate": 0.0001983274900820432,
      "loss": 0.5738,
      "step": 1150
    },
    {
      "epoch": 0.26538678712591757,
      "grad_norm": 0.08245467394590378,
      "learning_rate": 0.0001982146225202338,
      "loss": 0.6766,
      "step": 1175
    },
    {
      "epoch": 0.27103331451157536,
      "grad_norm": 0.11854438483715057,
      "learning_rate": 0.00019809810415894767,
      "loss": 0.5828,
      "step": 1200
    },
    {
      "epoch": 0.2766798418972332,
      "grad_norm": 0.07968372851610184,
      "learning_rate": 0.0001979779393293644,
      "loss": 0.6901,
      "step": 1225
    },
    {
      "epoch": 0.282326369282891,
      "grad_norm": 0.13610929250717163,
      "learning_rate": 0.00019785413249820893,
      "loss": 0.554,
      "step": 1250
    },
    {
      "epoch": 0.28797289666854886,
      "grad_norm": 0.08099998533725739,
      "learning_rate": 0.00019772668826758527,
      "loss": 0.6615,
      "step": 1275
    },
    {
      "epoch": 0.29361942405420666,
      "grad_norm": 0.10878365486860275,
      "learning_rate": 0.0001975956113748057,
      "loss": 0.547,
      "step": 1300
    },
    {
      "epoch": 0.2992659514398645,
      "grad_norm": 0.0784253254532814,
      "learning_rate": 0.0001974609066922144,
      "loss": 0.6589,
      "step": 1325
    },
    {
      "epoch": 0.3049124788255223,
      "grad_norm": 0.10678227245807648,
      "learning_rate": 0.00019732257922700655,
      "loss": 0.5443,
      "step": 1350
    },
    {
      "epoch": 0.3105590062111801,
      "grad_norm": 0.07781955599784851,
      "learning_rate": 0.00019718063412104222,
      "loss": 0.6798,
      "step": 1375
    },
    {
      "epoch": 0.31620553359683795,
      "grad_norm": 0.12910470366477966,
      "learning_rate": 0.00019703507665065498,
      "loss": 0.54,
      "step": 1400
    },
    {
      "epoch": 0.32185206098249575,
      "grad_norm": 0.08442296832799911,
      "learning_rate": 0.00019688591222645607,
      "loss": 0.6998,
      "step": 1425
    },
    {
      "epoch": 0.3274985883681536,
      "grad_norm": 0.10945732891559601,
      "learning_rate": 0.00019673314639313315,
      "loss": 0.5576,
      "step": 1450
    },
    {
      "epoch": 0.3331451157538114,
      "grad_norm": 0.07373673468828201,
      "learning_rate": 0.00019657678482924406,
      "loss": 0.709,
      "step": 1475
    },
    {
      "epoch": 0.33879164313946925,
      "grad_norm": 0.10061786323785782,
      "learning_rate": 0.00019641683334700608,
      "loss": 0.5329,
      "step": 1500
    },
    {
      "epoch": 0.34443817052512704,
      "grad_norm": 0.08854890614748001,
      "learning_rate": 0.00019625329789207949,
      "loss": 0.6753,
      "step": 1525
    },
    {
      "epoch": 0.3500846979107849,
      "grad_norm": 0.11869047582149506,
      "learning_rate": 0.00019608618454334685,
      "loss": 0.54,
      "step": 1550
    },
    {
      "epoch": 0.3557312252964427,
      "grad_norm": 0.0819987803697586,
      "learning_rate": 0.00019591549951268692,
      "loss": 0.683,
      "step": 1575
    },
    {
      "epoch": 0.3613777526821005,
      "grad_norm": 0.10941476374864578,
      "learning_rate": 0.00019574124914474374,
      "loss": 0.5713,
      "step": 1600
    },
    {
      "epoch": 0.36702428006775834,
      "grad_norm": 0.0801759734749794,
      "learning_rate": 0.00019556343991669083,
      "loss": 0.6494,
      "step": 1625
    },
    {
      "epoch": 0.37267080745341613,
      "grad_norm": 0.11291850358247757,
      "learning_rate": 0.0001953820784379904,
      "loss": 0.5428,
      "step": 1650
    },
    {
      "epoch": 0.378317334839074,
      "grad_norm": 0.0784342810511589,
      "learning_rate": 0.00019519717145014765,
      "loss": 0.6789,
      "step": 1675
    },
    {
      "epoch": 0.3839638622247318,
      "grad_norm": 0.11995444446802139,
      "learning_rate": 0.00019500872582646034,
      "loss": 0.5409,
      "step": 1700
    },
    {
      "epoch": 0.38961038961038963,
      "grad_norm": 0.10059577226638794,
      "learning_rate": 0.00019481674857176293,
      "loss": 0.6887,
      "step": 1725
    },
    {
      "epoch": 0.3952569169960474,
      "grad_norm": 0.11123017966747284,
      "learning_rate": 0.0001946212468221666,
      "loss": 0.5482,
      "step": 1750
    },
    {
      "epoch": 0.4009034443817053,
      "grad_norm": 0.0788169875741005,
      "learning_rate": 0.00019442222784479382,
      "loss": 0.6577,
      "step": 1775
    },
    {
      "epoch": 0.40654997176736307,
      "grad_norm": 0.09987206757068634,
      "learning_rate": 0.0001942196990375081,
      "loss": 0.5353,
      "step": 1800
    },
    {
      "epoch": 0.41219649915302087,
      "grad_norm": 0.081305131316185,
      "learning_rate": 0.00019401366792863914,
      "loss": 0.6757,
      "step": 1825
    },
    {
      "epoch": 0.4178430265386787,
      "grad_norm": 0.11595868319272995,
      "learning_rate": 0.00019380414217670309,
      "loss": 0.5306,
      "step": 1850
    },
    {
      "epoch": 0.4234895539243365,
      "grad_norm": 0.08652087301015854,
      "learning_rate": 0.00019359112957011764,
      "loss": 0.6634,
      "step": 1875
    },
    {
      "epoch": 0.42913608130999437,
      "grad_norm": 0.1181817501783371,
      "learning_rate": 0.00019337463802691264,
      "loss": 0.5465,
      "step": 1900
    },
    {
      "epoch": 0.43478260869565216,
      "grad_norm": 0.08145532757043839,
      "learning_rate": 0.00019315467559443574,
      "loss": 0.6639,
      "step": 1925
    },
    {
      "epoch": 0.44042913608131,
      "grad_norm": 0.10263626277446747,
      "learning_rate": 0.0001929312504490533,
      "loss": 0.5418,
      "step": 1950
    },
    {
      "epoch": 0.4460756634669678,
      "grad_norm": 0.07672405242919922,
      "learning_rate": 0.00019270437089584635,
      "loss": 0.6547,
      "step": 1975
    },
    {
      "epoch": 0.45172219085262566,
      "grad_norm": 0.0984482392668724,
      "learning_rate": 0.00019247404536830204,
      "loss": 0.5321,
      "step": 2000
    },
    {
      "epoch": 0.45736871823828346,
      "grad_norm": 0.08558057248592377,
      "learning_rate": 0.00019224028242800006,
      "loss": 0.6657,
      "step": 2025
    },
    {
      "epoch": 0.46301524562394125,
      "grad_norm": 0.10624836385250092,
      "learning_rate": 0.00019200309076429438,
      "loss": 0.5542,
      "step": 2050
    },
    {
      "epoch": 0.4686617730095991,
      "grad_norm": 0.08049538731575012,
      "learning_rate": 0.00019176247919399023,
      "loss": 0.6389,
      "step": 2075
    },
    {
      "epoch": 0.4743083003952569,
      "grad_norm": 0.11163744330406189,
      "learning_rate": 0.00019151845666101646,
      "loss": 0.5382,
      "step": 2100
    },
    {
      "epoch": 0.47995482778091475,
      "grad_norm": 0.07694784551858902,
      "learning_rate": 0.00019127103223609307,
      "loss": 0.6345,
      "step": 2125
    },
    {
      "epoch": 0.48560135516657255,
      "grad_norm": 0.11501342058181763,
      "learning_rate": 0.0001910202151163939,
      "loss": 0.5303,
      "step": 2150
    },
    {
      "epoch": 0.4912478825522304,
      "grad_norm": 0.08532056957483292,
      "learning_rate": 0.00019076601462520492,
      "loss": 0.666,
      "step": 2175
    },
    {
      "epoch": 0.4968944099378882,
      "grad_norm": 0.09018545597791672,
      "learning_rate": 0.00019050844021157752,
      "loss": 0.5363,
      "step": 2200
    },
    {
      "epoch": 0.502540937323546,
      "grad_norm": 0.08757849037647247,
      "learning_rate": 0.00019024750144997746,
      "loss": 0.6684,
      "step": 2225
    },
    {
      "epoch": 0.5081874647092038,
      "grad_norm": 0.09418553113937378,
      "learning_rate": 0.00018998320803992872,
      "loss": 0.565,
      "step": 2250
    },
    {
      "epoch": 0.5138339920948617,
      "grad_norm": 0.10246949642896652,
      "learning_rate": 0.00018971556980565329,
      "loss": 0.6524,
      "step": 2275
    },
    {
      "epoch": 0.5194805194805194,
      "grad_norm": 0.12446881830692291,
      "learning_rate": 0.00018944459669570555,
      "loss": 0.5447,
      "step": 2300
    },
    {
      "epoch": 0.5251270468661773,
      "grad_norm": 0.08439560234546661,
      "learning_rate": 0.00018917029878260294,
      "loss": 0.673,
      "step": 2325
    },
    {
      "epoch": 0.5307735742518351,
      "grad_norm": 0.1195409968495369,
      "learning_rate": 0.00018889268626245116,
      "loss": 0.51,
      "step": 2350
    },
    {
      "epoch": 0.536420101637493,
      "grad_norm": 0.09150119870901108,
      "learning_rate": 0.0001886117694545654,
      "loss": 0.6631,
      "step": 2375
    },
    {
      "epoch": 0.5420666290231507,
      "grad_norm": 0.10301247239112854,
      "learning_rate": 0.0001883275588010866,
      "loss": 0.5257,
      "step": 2400
    },
    {
      "epoch": 0.5477131564088086,
      "grad_norm": 0.08086048811674118,
      "learning_rate": 0.00018804006486659346,
      "loss": 0.6305,
      "step": 2425
    },
    {
      "epoch": 0.5533596837944664,
      "grad_norm": 0.10540500283241272,
      "learning_rate": 0.0001877492983377096,
      "loss": 0.5607,
      "step": 2450
    },
    {
      "epoch": 0.5590062111801242,
      "grad_norm": 0.0777876153588295,
      "learning_rate": 0.00018745527002270634,
      "loss": 0.6413,
      "step": 2475
    },
    {
      "epoch": 0.564652738565782,
      "grad_norm": 0.10727331042289734,
      "learning_rate": 0.00018715799085110112,
      "loss": 0.5206,
      "step": 2500
    },
    {
      "epoch": 0.5702992659514399,
      "grad_norm": 0.09144891798496246,
      "learning_rate": 0.0001868574718732508,
      "loss": 0.6655,
      "step": 2525
    },
    {
      "epoch": 0.5759457933370977,
      "grad_norm": 0.08949702978134155,
      "learning_rate": 0.00018655372425994152,
      "loss": 0.5358,
      "step": 2550
    },
    {
      "epoch": 0.5815923207227555,
      "grad_norm": 0.08744482696056366,
      "learning_rate": 0.0001862467593019728,
      "loss": 0.6396,
      "step": 2575
    },
    {
      "epoch": 0.5872388481084133,
      "grad_norm": 0.10931294411420822,
      "learning_rate": 0.0001859365884097384,
      "loss": 0.5132,
      "step": 2600
    },
    {
      "epoch": 0.5928853754940712,
      "grad_norm": 0.07518015056848526,
      "learning_rate": 0.00018562322311280186,
      "loss": 0.6378,
      "step": 2625
    },
    {
      "epoch": 0.598531902879729,
      "grad_norm": 0.09787683933973312,
      "learning_rate": 0.000185306675059468,
      "loss": 0.5272,
      "step": 2650
    },
    {
      "epoch": 0.6041784302653868,
      "grad_norm": 0.08679146319627762,
      "learning_rate": 0.00018498695601634993,
      "loss": 0.6486,
      "step": 2675
    },
    {
      "epoch": 0.6098249576510446,
      "grad_norm": 0.10611116141080856,
      "learning_rate": 0.00018466407786793174,
      "loss": 0.5237,
      "step": 2700
    },
    {
      "epoch": 0.6154714850367025,
      "grad_norm": 0.0772860199213028,
      "learning_rate": 0.00018433805261612663,
      "loss": 0.6526,
      "step": 2725
    },
    {
      "epoch": 0.6211180124223602,
      "grad_norm": 0.1005689725279808,
      "learning_rate": 0.00018400889237983086,
      "loss": 0.5297,
      "step": 2750
    },
    {
      "epoch": 0.626764539808018,
      "grad_norm": 0.08322805911302567,
      "learning_rate": 0.00018367660939447316,
      "loss": 0.6448,
      "step": 2775
    },
    {
      "epoch": 0.6324110671936759,
      "grad_norm": 0.10492440313100815,
      "learning_rate": 0.00018334121601156002,
      "loss": 0.493,
      "step": 2800
    },
    {
      "epoch": 0.6380575945793338,
      "grad_norm": 0.08439356088638306,
      "learning_rate": 0.00018300272469821662,
      "loss": 0.6421,
      "step": 2825
    },
    {
      "epoch": 0.6437041219649915,
      "grad_norm": 0.11175478994846344,
      "learning_rate": 0.00018266114803672318,
      "loss": 0.5293,
      "step": 2850
    },
    {
      "epoch": 0.6493506493506493,
      "grad_norm": 0.07700271904468536,
      "learning_rate": 0.00018231649872404754,
      "loss": 0.6506,
      "step": 2875
    },
    {
      "epoch": 0.6549971767363072,
      "grad_norm": 0.08704936504364014,
      "learning_rate": 0.00018196878957137295,
      "loss": 0.5091,
      "step": 2900
    },
    {
      "epoch": 0.6606437041219649,
      "grad_norm": 0.0863596498966217,
      "learning_rate": 0.00018161803350362198,
      "loss": 0.6531,
      "step": 2925
    },
    {
      "epoch": 0.6662902315076228,
      "grad_norm": 0.11374282091856003,
      "learning_rate": 0.00018126424355897612,
      "loss": 0.5389,
      "step": 2950
    },
    {
      "epoch": 0.6719367588932806,
      "grad_norm": 0.08936896920204163,
      "learning_rate": 0.000180907432888391,
      "loss": 0.6267,
      "step": 2975
    },
    {
      "epoch": 0.6775832862789385,
      "grad_norm": 0.08924362808465958,
      "learning_rate": 0.0001805476147551076,
      "loss": 0.4882,
      "step": 3000
    },
    {
      "epoch": 0.6832298136645962,
      "grad_norm": 0.08161070942878723,
      "learning_rate": 0.0001801848025341593,
      "loss": 0.6502,
      "step": 3025
    },
    {
      "epoch": 0.6888763410502541,
      "grad_norm": 0.11261451989412308,
      "learning_rate": 0.00017981900971187465,
      "loss": 0.5241,
      "step": 3050
    },
    {
      "epoch": 0.6945228684359119,
      "grad_norm": 0.0859522670507431,
      "learning_rate": 0.00017945024988537603,
      "loss": 0.6413,
      "step": 3075
    },
    {
      "epoch": 0.7001693958215698,
      "grad_norm": 0.11650535464286804,
      "learning_rate": 0.0001790785367620743,
      "loss": 0.519,
      "step": 3100
    },
    {
      "epoch": 0.7058159232072275,
      "grad_norm": 0.09202492237091064,
      "learning_rate": 0.00017870388415915922,
      "loss": 0.6212,
      "step": 3125
    },
    {
      "epoch": 0.7114624505928854,
      "grad_norm": 0.1048663780093193,
      "learning_rate": 0.00017832630600308585,
      "loss": 0.5266,
      "step": 3150
    },
    {
      "epoch": 0.7171089779785432,
      "grad_norm": 0.08784987032413483,
      "learning_rate": 0.00017794581632905683,
      "loss": 0.6214,
      "step": 3175
    },
    {
      "epoch": 0.722755505364201,
      "grad_norm": 0.10895237326622009,
      "learning_rate": 0.00017756242928050085,
      "loss": 0.5059,
      "step": 3200
    },
    {
      "epoch": 0.7284020327498588,
      "grad_norm": 0.09365130960941315,
      "learning_rate": 0.0001771761591085467,
      "loss": 0.6529,
      "step": 3225
    },
    {
      "epoch": 0.7340485601355167,
      "grad_norm": 0.1045169085264206,
      "learning_rate": 0.0001767870201714936,
      "loss": 0.5229,
      "step": 3250
    },
    {
      "epoch": 0.7396950875211745,
      "grad_norm": 0.08621610701084137,
      "learning_rate": 0.0001763950269342776,
      "loss": 0.6489,
      "step": 3275
    },
    {
      "epoch": 0.7453416149068323,
      "grad_norm": 0.1039639338850975,
      "learning_rate": 0.00017600019396793367,
      "loss": 0.5116,
      "step": 3300
    },
    {
      "epoch": 0.7509881422924901,
      "grad_norm": 0.09034851938486099,
      "learning_rate": 0.00017560253594905425,
      "loss": 0.6587,
      "step": 3325
    },
    {
      "epoch": 0.756634669678148,
      "grad_norm": 0.10783170908689499,
      "learning_rate": 0.00017520206765924372,
      "loss": 0.5019,
      "step": 3350
    },
    {
      "epoch": 0.7622811970638057,
      "grad_norm": 0.08911153674125671,
      "learning_rate": 0.00017479880398456871,
      "loss": 0.6242,
      "step": 3375
    },
    {
      "epoch": 0.7679277244494636,
      "grad_norm": 0.10053360462188721,
      "learning_rate": 0.00017439275991500507,
      "loss": 0.4996,
      "step": 3400
    },
    {
      "epoch": 0.7735742518351214,
      "grad_norm": 0.09367632865905762,
      "learning_rate": 0.0001739839505438804,
      "loss": 0.6347,
      "step": 3425
    },
    {
      "epoch": 0.7792207792207793,
      "grad_norm": 0.09187789261341095,
      "learning_rate": 0.00017357239106731317,
      "loss": 0.4857,
      "step": 3450
    },
    {
      "epoch": 0.784867306606437,
      "grad_norm": 0.08425033092498779,
      "learning_rate": 0.00017315809678364777,
      "loss": 0.6305,
      "step": 3475
    },
    {
      "epoch": 0.7905138339920948,
      "grad_norm": 0.10775783658027649,
      "learning_rate": 0.00017274108309288594,
      "loss": 0.5013,
      "step": 3500
    },
    {
      "epoch": 0.7961603613777527,
      "grad_norm": 0.082928866147995,
      "learning_rate": 0.00017232136549611416,
      "loss": 0.6217,
      "step": 3525
    },
    {
      "epoch": 0.8018068887634106,
      "grad_norm": 0.09590538591146469,
      "learning_rate": 0.00017189895959492772,
      "loss": 0.4871,
      "step": 3550
    },
    {
      "epoch": 0.8074534161490683,
      "grad_norm": 0.08597232401371002,
      "learning_rate": 0.00017147388109085048,
      "loss": 0.6298,
      "step": 3575
    },
    {
      "epoch": 0.8130999435347261,
      "grad_norm": 0.10229629278182983,
      "learning_rate": 0.00017104614578475135,
      "loss": 0.5191,
      "step": 3600
    },
    {
      "epoch": 0.818746470920384,
      "grad_norm": 0.09689701348543167,
      "learning_rate": 0.0001706157695762571,
      "loss": 0.6087,
      "step": 3625
    },
    {
      "epoch": 0.8243929983060417,
      "grad_norm": 0.1155649945139885,
      "learning_rate": 0.000170182768463161,
      "loss": 0.5237,
      "step": 3650
    },
    {
      "epoch": 0.8300395256916996,
      "grad_norm": 0.08635739237070084,
      "learning_rate": 0.00016974715854082848,
      "loss": 0.6337,
      "step": 3675
    },
    {
      "epoch": 0.8356860530773574,
      "grad_norm": 0.08977790176868439,
      "learning_rate": 0.00016930895600159867,
      "loss": 0.4942,
      "step": 3700
    },
    {
      "epoch": 0.8413325804630153,
      "grad_norm": 0.08372893184423447,
      "learning_rate": 0.00016886817713418264,
      "loss": 0.6224,
      "step": 3725
    },
    {
      "epoch": 0.846979107848673,
      "grad_norm": 0.09409157931804657,
      "learning_rate": 0.00016842483832305765,
      "loss": 0.4948,
      "step": 3750
    },
    {
      "epoch": 0.8526256352343309,
      "grad_norm": 0.0824190303683281,
      "learning_rate": 0.00016797895604785842,
      "loss": 0.6188,
      "step": 3775
    },
    {
      "epoch": 0.8582721626199887,
      "grad_norm": 0.11509312689304352,
      "learning_rate": 0.0001675305468827644,
      "loss": 0.513,
      "step": 3800
    },
    {
      "epoch": 0.8639186900056465,
      "grad_norm": 0.08958346396684647,
      "learning_rate": 0.0001670796274958837,
      "loss": 0.6206,
      "step": 3825
    },
    {
      "epoch": 0.8695652173913043,
      "grad_norm": 0.0959952101111412,
      "learning_rate": 0.00016662621464863338,
      "loss": 0.4885,
      "step": 3850
    },
    {
      "epoch": 0.8752117447769622,
      "grad_norm": 0.11436719447374344,
      "learning_rate": 0.00016617032519511686,
      "loss": 0.635,
      "step": 3875
    },
    {
      "epoch": 0.88085827216262,
      "grad_norm": 0.11012863367795944,
      "learning_rate": 0.00016571197608149674,
      "loss": 0.4982,
      "step": 3900
    },
    {
      "epoch": 0.8865047995482778,
      "grad_norm": 0.08522720634937286,
      "learning_rate": 0.00016525118434536546,
      "loss": 0.6105,
      "step": 3925
    },
    {
      "epoch": 0.8921513269339356,
      "grad_norm": 0.11074435710906982,
      "learning_rate": 0.00016478796711511171,
      "loss": 0.5007,
      "step": 3950
    },
    {
      "epoch": 0.8977978543195935,
      "grad_norm": 0.08319137990474701,
      "learning_rate": 0.00016432234160928378,
      "loss": 0.6008,
      "step": 3975
    },
    {
      "epoch": 0.9034443817052513,
      "grad_norm": 0.11340347677469254,
      "learning_rate": 0.00016385432513594953,
      "loss": 0.502,
      "step": 4000
    },
    {
      "epoch": 0.9090909090909091,
      "grad_norm": 0.08557943999767303,
      "learning_rate": 0.0001633839350920531,
      "loss": 0.598,
      "step": 4025
    },
    {
      "epoch": 0.9147374364765669,
      "grad_norm": 0.09813399612903595,
      "learning_rate": 0.00016291118896276803,
      "loss": 0.5139,
      "step": 4050
    },
    {
      "epoch": 0.9203839638622248,
      "grad_norm": 0.07974950224161148,
      "learning_rate": 0.00016243610432084755,
      "loss": 0.6227,
      "step": 4075
    },
    {
      "epoch": 0.9260304912478825,
      "grad_norm": 0.09820155799388885,
      "learning_rate": 0.0001619586988259712,
      "loss": 0.5044,
      "step": 4100
    },
    {
      "epoch": 0.9316770186335404,
      "grad_norm": 0.0881289690732956,
      "learning_rate": 0.00016147899022408852,
      "loss": 0.6192,
      "step": 4125
    },
    {
      "epoch": 0.9373235460191982,
      "grad_norm": 0.11012347787618637,
      "learning_rate": 0.00016099699634675925,
      "loss": 0.508,
      "step": 4150
    },
    {
      "epoch": 0.9429700734048561,
      "grad_norm": 0.0851251408457756,
      "learning_rate": 0.00016051273511049065,
      "loss": 0.5897,
      "step": 4175
    },
    {
      "epoch": 0.9486166007905138,
      "grad_norm": 0.10288140922784805,
      "learning_rate": 0.0001600262245160714,
      "loss": 0.4803,
      "step": 4200
    },
    {
      "epoch": 0.9542631281761716,
      "grad_norm": 0.08995950222015381,
      "learning_rate": 0.0001595374826479026,
      "loss": 0.6203,
      "step": 4225
    },
    {
      "epoch": 0.9599096555618295,
      "grad_norm": 0.10717281699180603,
      "learning_rate": 0.00015904652767332537,
      "loss": 0.5068,
      "step": 4250
    },
    {
      "epoch": 0.9655561829474872,
      "grad_norm": 0.08305976539850235,
      "learning_rate": 0.00015855337784194577,
      "loss": 0.5919,
      "step": 4275
    },
    {
      "epoch": 0.9712027103331451,
      "grad_norm": 0.09328145533800125,
      "learning_rate": 0.00015805805148495623,
      "loss": 0.4948,
      "step": 4300
    },
    {
      "epoch": 0.9768492377188029,
      "grad_norm": 0.09493458271026611,
      "learning_rate": 0.00015756056701445422,
      "loss": 0.6024,
      "step": 4325
    },
    {
      "epoch": 0.9824957651044608,
      "grad_norm": 0.1051265150308609,
      "learning_rate": 0.0001570609429227579,
      "loss": 0.4991,
      "step": 4350
    },
    {
      "epoch": 0.9881422924901185,
      "grad_norm": 0.09176123887300491,
      "learning_rate": 0.00015655919778171862,
      "loss": 0.6006,
      "step": 4375
    },
    {
      "epoch": 0.9937888198757764,
      "grad_norm": 0.0787501335144043,
      "learning_rate": 0.00015605535024203069,
      "loss": 0.5013,
      "step": 4400
    },
    {
      "epoch": 0.9994353472614342,
      "grad_norm": 0.10489033907651901,
      "learning_rate": 0.00015554941903253797,
      "loss": 0.5548,
      "step": 4425
    },
    {
      "epoch": 1.005081874647092,
      "grad_norm": 0.08116624504327774,
      "learning_rate": 0.00015504142295953783,
      "loss": 0.5719,
      "step": 4450
    },
    {
      "epoch": 1.0107284020327498,
      "grad_norm": 0.09656750410795212,
      "learning_rate": 0.000154531380906082,
      "loss": 0.4752,
      "step": 4475
    },
    {
      "epoch": 1.0163749294184077,
      "grad_norm": 0.09433398395776749,
      "learning_rate": 0.0001540193118312747,
      "loss": 0.5796,
      "step": 4500
    },
    {
      "epoch": 1.0220214568040655,
      "grad_norm": 0.10783005505800247,
      "learning_rate": 0.0001535052347695678,
      "loss": 0.5077,
      "step": 4525
    },
    {
      "epoch": 1.0276679841897234,
      "grad_norm": 0.08784560114145279,
      "learning_rate": 0.00015298916883005342,
      "loss": 0.5571,
      "step": 4550
    },
    {
      "epoch": 1.0333145115753812,
      "grad_norm": 0.11868823319673538,
      "learning_rate": 0.00015247113319575358,
      "loss": 0.5223,
      "step": 4575
    },
    {
      "epoch": 1.0389610389610389,
      "grad_norm": 0.08813630044460297,
      "learning_rate": 0.000151951147122907,
      "loss": 0.561,
      "step": 4600
    },
    {
      "epoch": 1.0446075663466967,
      "grad_norm": 0.09996681660413742,
      "learning_rate": 0.0001514292299402535,
      "loss": 0.5138,
      "step": 4625
    },
    {
      "epoch": 1.0502540937323546,
      "grad_norm": 0.07442251592874527,
      "learning_rate": 0.00015090540104831539,
      "loss": 0.5698,
      "step": 4650
    },
    {
      "epoch": 1.0559006211180124,
      "grad_norm": 0.09993384033441544,
      "learning_rate": 0.00015037967991867642,
      "loss": 0.5093,
      "step": 4675
    },
    {
      "epoch": 1.0615471485036703,
      "grad_norm": 0.0890798568725586,
      "learning_rate": 0.0001498520860932579,
      "loss": 0.5481,
      "step": 4700
    },
    {
      "epoch": 1.0671936758893281,
      "grad_norm": 0.1006857305765152,
      "learning_rate": 0.00014932263918359228,
      "loss": 0.5045,
      "step": 4725
    },
    {
      "epoch": 1.072840203274986,
      "grad_norm": 0.08858868479728699,
      "learning_rate": 0.00014879135887009435,
      "loss": 0.5772,
      "step": 4750
    },
    {
      "epoch": 1.0784867306606438,
      "grad_norm": 0.11783988773822784,
      "learning_rate": 0.00014825826490132938,
      "loss": 0.4937,
      "step": 4775
    },
    {
      "epoch": 1.0841332580463015,
      "grad_norm": 0.0875164121389389,
      "learning_rate": 0.00014772337709327923,
      "loss": 0.5554,
      "step": 4800
    },
    {
      "epoch": 1.0897797854319593,
      "grad_norm": 0.112996406853199,
      "learning_rate": 0.00014718671532860592,
      "loss": 0.5126,
      "step": 4825
    },
    {
      "epoch": 1.0954263128176172,
      "grad_norm": 0.08834437280893326,
      "learning_rate": 0.000146648299555912,
      "loss": 0.5885,
      "step": 4850
    },
    {
      "epoch": 1.101072840203275,
      "grad_norm": 0.11341985315084457,
      "learning_rate": 0.00014610814978899983,
      "loss": 0.4871,
      "step": 4875
    },
    {
      "epoch": 1.1067193675889329,
      "grad_norm": 0.09301973134279251,
      "learning_rate": 0.00014556628610612677,
      "loss": 0.5839,
      "step": 4900
    },
    {
      "epoch": 1.1123658949745907,
      "grad_norm": 0.11545081436634064,
      "learning_rate": 0.00014502272864925955,
      "loss": 0.4868,
      "step": 4925
    },
    {
      "epoch": 1.1180124223602483,
      "grad_norm": 0.08969846367835999,
      "learning_rate": 0.00014447749762332515,
      "loss": 0.5472,
      "step": 4950
    },
    {
      "epoch": 1.1236589497459062,
      "grad_norm": 0.12094131112098694,
      "learning_rate": 0.00014393061329545992,
      "loss": 0.5234,
      "step": 4975
    },
    {
      "epoch": 1.129305477131564,
      "grad_norm": 0.0875294953584671,
      "learning_rate": 0.0001433820959942561,
      "loss": 0.5852,
      "step": 5000
    }
  ],
  "logging_steps": 25,
  "max_steps": 13281,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 200,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 2.946704435084667e+18,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}