{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.21341158424755743,
  "eval_steps": 500,
  "global_step": 1600,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.001333822401547234,
      "grad_norm": 72.53171450734499,
      "learning_rate": 4e-06,
      "loss": 1.0439,
      "step": 10
    },
    {
      "epoch": 0.002667644803094468,
      "grad_norm": 7.72170877312484,
      "learning_rate": 8e-06,
      "loss": 1.0385,
      "step": 20
    },
    {
      "epoch": 0.004001467204641702,
      "grad_norm": 28.037294890647996,
      "learning_rate": 1.2e-05,
      "loss": 1.0328,
      "step": 30
    },
    {
      "epoch": 0.005335289606188936,
      "grad_norm": 10.980963099753343,
      "learning_rate": 1.6e-05,
      "loss": 0.9608,
      "step": 40
    },
    {
      "epoch": 0.00666911200773617,
      "grad_norm": 9.853684997615185,
      "learning_rate": 1.9999999999999998e-05,
      "loss": 0.8941,
      "step": 50
    },
    {
      "epoch": 0.008002934409283404,
      "grad_norm": 33.396994669126855,
      "learning_rate": 2.4e-05,
      "loss": 0.9174,
      "step": 60
    },
    {
      "epoch": 0.009336756810830639,
      "grad_norm": 5.621190624979587,
      "learning_rate": 2.8e-05,
      "loss": 0.8146,
      "step": 70
    },
    {
      "epoch": 0.010670579212377872,
      "grad_norm": 2.4646688601781843,
      "learning_rate": 2.9999966406213696e-05,
      "loss": 0.7774,
      "step": 80
    },
    {
      "epoch": 0.012004401613925107,
      "grad_norm": 2.1272107013027215,
      "learning_rate": 2.9999697656826056e-05,
      "loss": 0.7426,
      "step": 90
    },
    {
      "epoch": 0.01333822401547234,
      "grad_norm": 2.0001927521328575,
      "learning_rate": 2.9999160162865885e-05,
      "loss": 0.7251,
      "step": 100
    },
    {
      "epoch": 0.014672046417019574,
      "grad_norm": 4.278494734910313,
      "learning_rate": 2.9998353933963273e-05,
      "loss": 0.7283,
      "step": 110
    },
    {
      "epoch": 0.016005868818566808,
      "grad_norm": 84.74138622339441,
      "learning_rate": 2.999727898456315e-05,
      "loss": 0.721,
      "step": 120
    },
    {
      "epoch": 0.017339691220114042,
      "grad_norm": 2.5596285292704533,
      "learning_rate": 2.999593533392503e-05,
      "loss": 0.7476,
      "step": 130
    },
    {
      "epoch": 0.018673513621661277,
      "grad_norm": 3.167163073798878,
      "learning_rate": 2.9994323006122654e-05,
      "loss": 0.7701,
      "step": 140
    },
    {
      "epoch": 0.02000733602320851,
      "grad_norm": 12.240132620099228,
      "learning_rate": 2.9992442030043557e-05,
      "loss": 0.7042,
      "step": 150
    },
    {
      "epoch": 0.021341158424755743,
      "grad_norm": 2.266079140233148,
      "learning_rate": 2.9990292439388565e-05,
      "loss": 0.6945,
      "step": 160
    },
    {
      "epoch": 0.022674980826302978,
      "grad_norm": 4.493189214061043,
      "learning_rate": 2.9987874272671168e-05,
      "loss": 0.7811,
      "step": 170
    },
    {
      "epoch": 0.024008803227850213,
      "grad_norm": 3.8513915234854132,
      "learning_rate": 2.9985187573216855e-05,
      "loss": 0.7229,
      "step": 180
    },
    {
      "epoch": 0.025342625629397444,
      "grad_norm": 2.1061250870336896,
      "learning_rate": 2.998223238916232e-05,
      "loss": 0.7253,
      "step": 190
    },
    {
      "epoch": 0.02667644803094468,
      "grad_norm": 2.8326879963181373,
      "learning_rate": 2.9979008773454618e-05,
      "loss": 0.7576,
      "step": 200
    },
    {
      "epoch": 0.028010270432491914,
      "grad_norm": 3.0294324007099402,
      "learning_rate": 2.997551678385019e-05,
      "loss": 0.7362,
      "step": 210
    },
    {
      "epoch": 0.02934409283403915,
      "grad_norm": 1.5691302199679913,
      "learning_rate": 2.997175648291384e-05,
      "loss": 0.7212,
      "step": 220
    },
    {
      "epoch": 0.03067791523558638,
      "grad_norm": 4.379709379404857,
      "learning_rate": 2.996772793801763e-05,
      "loss": 0.7382,
      "step": 230
    },
    {
      "epoch": 0.032011737637133615,
      "grad_norm": 1.899348727085658,
      "learning_rate": 2.996343122133965e-05,
      "loss": 0.7056,
      "step": 240
    },
    {
      "epoch": 0.033345560038680847,
      "grad_norm": 21.481584335801955,
      "learning_rate": 2.9958866409862745e-05,
      "loss": 0.697,
      "step": 250
    },
    {
      "epoch": 0.034679382440228085,
      "grad_norm": 3.022034785435926,
      "learning_rate": 2.9954033585373108e-05,
      "loss": 0.69,
      "step": 260
    },
    {
      "epoch": 0.036013204841775316,
      "grad_norm": 2.869274575136443,
      "learning_rate": 2.994893283445885e-05,
      "loss": 0.7026,
      "step": 270
    },
    {
      "epoch": 0.037347027243322554,
      "grad_norm": 7.8785627268474,
      "learning_rate": 2.9943564248508415e-05,
      "loss": 0.7021,
      "step": 280
    },
    {
      "epoch": 0.038680849644869786,
      "grad_norm": 4.450751706186503,
      "learning_rate": 2.9937927923708966e-05,
      "loss": 0.6814,
      "step": 290
    },
    {
      "epoch": 0.04001467204641702,
      "grad_norm": 11.613541400971123,
      "learning_rate": 2.993202396104465e-05,
      "loss": 0.7043,
      "step": 300
    },
    {
      "epoch": 0.041348494447964255,
      "grad_norm": 2.0602161594051536,
      "learning_rate": 2.9925852466294795e-05,
      "loss": 0.7073,
      "step": 310
    },
    {
      "epoch": 0.04268231684951149,
      "grad_norm": 19.383133688390824,
      "learning_rate": 2.9919413550032014e-05,
      "loss": 0.6965,
      "step": 320
    },
    {
      "epoch": 0.04401613925105872,
      "grad_norm": 7.9603948136759906,
      "learning_rate": 2.991270732762022e-05,
      "loss": 0.7029,
      "step": 330
    },
    {
      "epoch": 0.045349961652605957,
      "grad_norm": 2.810319267000959,
      "learning_rate": 2.990573391921255e-05,
      "loss": 0.7412,
      "step": 340
    },
    {
      "epoch": 0.04668378405415319,
      "grad_norm": 6.082487526538632,
      "learning_rate": 2.989849344974924e-05,
      "loss": 0.6913,
      "step": 350
    },
    {
      "epoch": 0.048017606455700426,
      "grad_norm": 1.3651265332328906,
      "learning_rate": 2.9890986048955368e-05,
      "loss": 0.6802,
      "step": 360
    },
    {
      "epoch": 0.04935142885724766,
      "grad_norm": 2.7766095141894023,
      "learning_rate": 2.9883211851338516e-05,
      "loss": 0.7254,
      "step": 370
    },
    {
      "epoch": 0.05068525125879489,
      "grad_norm": 2.2284551774435033,
      "learning_rate": 2.9875170996186392e-05,
      "loss": 0.7026,
      "step": 380
    },
    {
      "epoch": 0.05201907366034213,
      "grad_norm": 2.1609680967998957,
      "learning_rate": 2.986686362756431e-05,
      "loss": 0.7415,
      "step": 390
    },
    {
      "epoch": 0.05335289606188936,
      "grad_norm": 2.071743222471935,
      "learning_rate": 2.9858289894312617e-05,
      "loss": 0.7081,
      "step": 400
    },
    {
      "epoch": 0.0546867184634366,
      "grad_norm": 7.349693198515479,
      "learning_rate": 2.9849449950044036e-05,
      "loss": 0.7212,
      "step": 410
    },
    {
      "epoch": 0.05602054086498383,
      "grad_norm": 12.990346405367537,
      "learning_rate": 2.984034395314088e-05,
      "loss": 0.7187,
      "step": 420
    },
    {
      "epoch": 0.05735436326653106,
      "grad_norm": 2.4106321189642723,
      "learning_rate": 2.983097206675227e-05,
      "loss": 0.6881,
      "step": 430
    },
    {
      "epoch": 0.0586881856680783,
      "grad_norm": 3.1647778114563163,
      "learning_rate": 2.9821334458791156e-05,
      "loss": 0.7529,
      "step": 440
    },
    {
      "epoch": 0.06002200806962553,
      "grad_norm": 2.7769936556882713,
      "learning_rate": 2.9811431301931344e-05,
      "loss": 0.7073,
      "step": 450
    },
    {
      "epoch": 0.06135583047117276,
      "grad_norm": 2.834916196253906,
      "learning_rate": 2.9801262773604377e-05,
      "loss": 0.6649,
      "step": 460
    },
    {
      "epoch": 0.06268965287271999,
      "grad_norm": 2.431783869557187,
      "learning_rate": 2.9790829055996398e-05,
      "loss": 0.6889,
      "step": 470
    },
    {
      "epoch": 0.06402347527426723,
      "grad_norm": 3.7325728788625825,
      "learning_rate": 2.978013033604483e-05,
      "loss": 0.6771,
      "step": 480
    },
    {
      "epoch": 0.06535729767581447,
      "grad_norm": 10.194734403509758,
      "learning_rate": 2.976916680543506e-05,
      "loss": 0.6917,
      "step": 490
    },
    {
      "epoch": 0.06669112007736169,
      "grad_norm": 6.4699962355353655,
      "learning_rate": 2.975793866059701e-05,
      "loss": 0.6656,
      "step": 500
    },
    {
      "epoch": 0.06802494247890893,
      "grad_norm": 7.664735661456941,
      "learning_rate": 2.9746446102701606e-05,
      "loss": 0.7268,
      "step": 510
    },
    {
      "epoch": 0.06935876488045617,
      "grad_norm": 2.926294421836929,
      "learning_rate": 2.9734689337657157e-05,
      "loss": 0.7498,
      "step": 520
    },
    {
      "epoch": 0.07069258728200341,
      "grad_norm": 8.95556138792813,
      "learning_rate": 2.9722668576105703e-05,
      "loss": 0.7237,
      "step": 530
    },
    {
      "epoch": 0.07202640968355063,
      "grad_norm": 2.725769364825004,
      "learning_rate": 2.971038403341921e-05,
      "loss": 0.701,
      "step": 540
    },
    {
      "epoch": 0.07336023208509787,
      "grad_norm": 2.458130413236654,
      "learning_rate": 2.9697835929695727e-05,
      "loss": 0.6775,
      "step": 550
    },
    {
      "epoch": 0.07469405448664511,
      "grad_norm": 1.908510150241819,
      "learning_rate": 2.968502448975544e-05,
      "loss": 0.7042,
      "step": 560
    },
    {
      "epoch": 0.07602787688819233,
      "grad_norm": 26.340984665432142,
      "learning_rate": 2.967194994313663e-05,
      "loss": 0.7234,
      "step": 570
    },
    {
      "epoch": 0.07736169928973957,
      "grad_norm": 5.677162249188698,
      "learning_rate": 2.9658612524091594e-05,
      "loss": 0.7159,
      "step": 580
    },
    {
      "epoch": 0.07869552169128681,
      "grad_norm": 1.7975075976512287,
      "learning_rate": 2.9645012471582406e-05,
      "loss": 0.7268,
      "step": 590
    },
    {
      "epoch": 0.08002934409283403,
      "grad_norm": 3.031655038731278,
      "learning_rate": 2.9631150029276662e-05,
      "loss": 0.6904,
      "step": 600
    },
    {
      "epoch": 0.08136316649438127,
      "grad_norm": 13.019769162242698,
      "learning_rate": 2.9617025445543114e-05,
      "loss": 0.683,
      "step": 610
    },
    {
      "epoch": 0.08269698889592851,
      "grad_norm": 5.982097939523946,
      "learning_rate": 2.9602638973447218e-05,
      "loss": 0.7511,
      "step": 620
    },
    {
      "epoch": 0.08403081129747574,
      "grad_norm": 2.7343884277014054,
      "learning_rate": 2.9587990870746574e-05,
      "loss": 0.7114,
      "step": 630
    },
    {
      "epoch": 0.08536463369902297,
      "grad_norm": 5.586409371459215,
      "learning_rate": 2.9573081399886356e-05,
      "loss": 0.6907,
      "step": 640
    },
    {
      "epoch": 0.08669845610057021,
      "grad_norm": 9.87255558712134,
      "learning_rate": 2.9557910827994568e-05,
      "loss": 0.7319,
      "step": 650
    },
    {
      "epoch": 0.08803227850211744,
      "grad_norm": 3.9867937712251664,
      "learning_rate": 2.9542479426877283e-05,
      "loss": 0.6759,
      "step": 660
    },
    {
      "epoch": 0.08936610090366467,
      "grad_norm": 3.111751395898768,
      "learning_rate": 2.9526787473013753e-05,
      "loss": 0.7086,
      "step": 670
    },
    {
      "epoch": 0.09069992330521191,
      "grad_norm": 3.0578210081080535,
      "learning_rate": 2.9510835247551485e-05,
      "loss": 0.7209,
      "step": 680
    },
    {
      "epoch": 0.09203374570675915,
      "grad_norm": 16.575220829800145,
      "learning_rate": 2.949462303630116e-05,
      "loss": 0.7125,
      "step": 690
    },
    {
      "epoch": 0.09336756810830638,
      "grad_norm": 1.8236956939037314,
      "learning_rate": 2.9478151129731567e-05,
      "loss": 0.6591,
      "step": 700
    },
    {
      "epoch": 0.09470139050985361,
      "grad_norm": 1.9746052093914703,
      "learning_rate": 2.9461419822964348e-05,
      "loss": 0.7048,
      "step": 710
    },
    {
      "epoch": 0.09603521291140085,
      "grad_norm": 5.9710918845264,
      "learning_rate": 2.9444429415768726e-05,
      "loss": 0.725,
      "step": 720
    },
    {
      "epoch": 0.09736903531294808,
      "grad_norm": 9.961068856364284,
      "learning_rate": 2.942718021255617e-05,
      "loss": 0.7155,
      "step": 730
    },
    {
      "epoch": 0.09870285771449532,
      "grad_norm": 9.195914431100427,
      "learning_rate": 2.940967252237488e-05,
      "loss": 0.7171,
      "step": 740
    },
    {
      "epoch": 0.10003668011604255,
      "grad_norm": 5.972259927477824,
      "learning_rate": 2.9391906658904296e-05,
      "loss": 0.7088,
      "step": 750
    },
    {
      "epoch": 0.10137050251758978,
      "grad_norm": 3.0581670197906226,
      "learning_rate": 2.937388294044946e-05,
      "loss": 0.6645,
      "step": 760
    },
    {
      "epoch": 0.10270432491913702,
      "grad_norm": 3.2614276687575945,
      "learning_rate": 2.9355601689935315e-05,
      "loss": 0.6997,
      "step": 770
    },
    {
      "epoch": 0.10403814732068425,
      "grad_norm": 3.212661762840432,
      "learning_rate": 2.933706323490092e-05,
      "loss": 0.7005,
      "step": 780
    },
    {
      "epoch": 0.10537196972223148,
      "grad_norm": 10.732038885056795,
      "learning_rate": 2.9318267907493583e-05,
      "loss": 0.6588,
      "step": 790
    },
    {
      "epoch": 0.10670579212377872,
      "grad_norm": 2.495380329507228,
      "learning_rate": 2.9299216044462903e-05,
      "loss": 0.6965,
      "step": 800
    },
    {
      "epoch": 0.10803961452532596,
      "grad_norm": 4.60295956813534,
      "learning_rate": 2.927990798715475e-05,
      "loss": 0.6959,
      "step": 810
    },
    {
      "epoch": 0.1093734369268732,
      "grad_norm": 5.95010483593304,
      "learning_rate": 2.926034408150513e-05,
      "loss": 0.7083,
      "step": 820
    },
    {
      "epoch": 0.11070725932842042,
      "grad_norm": 10.48663715000287,
      "learning_rate": 2.9240524678034016e-05,
      "loss": 0.6977,
      "step": 830
    },
    {
      "epoch": 0.11204108172996766,
      "grad_norm": 3.159424497856185,
      "learning_rate": 2.9220450131839037e-05,
      "loss": 0.7129,
      "step": 840
    },
    {
      "epoch": 0.1133749041315149,
      "grad_norm": 3.984530217756339,
      "learning_rate": 2.920012080258912e-05,
      "loss": 0.6618,
      "step": 850
    },
    {
      "epoch": 0.11470872653306212,
      "grad_norm": 4.212436954015185,
      "learning_rate": 2.9179537054518085e-05,
      "loss": 0.6958,
      "step": 860
    },
    {
      "epoch": 0.11604254893460936,
      "grad_norm": 2.5397356754848657,
      "learning_rate": 2.9158699256418056e-05,
      "loss": 0.6593,
      "step": 870
    },
    {
      "epoch": 0.1173763713361566,
      "grad_norm": 7.8192219616904115,
      "learning_rate": 2.9137607781632913e-05,
      "loss": 0.6932,
      "step": 880
    },
    {
      "epoch": 0.11871019373770382,
      "grad_norm": 3.890974772100691,
      "learning_rate": 2.911626300805155e-05,
      "loss": 0.674,
      "step": 890
    },
    {
      "epoch": 0.12004401613925106,
      "grad_norm": 2.241784461612369,
      "learning_rate": 2.9094665318101155e-05,
      "loss": 0.6808,
      "step": 900
    },
    {
      "epoch": 0.1213778385407983,
      "grad_norm": 2.8295407624221456,
      "learning_rate": 2.9072815098740326e-05,
      "loss": 0.7074,
      "step": 910
    },
    {
      "epoch": 0.12271166094234552,
      "grad_norm": 2.8191065019379455,
      "learning_rate": 2.9050712741452136e-05,
      "loss": 0.7218,
      "step": 920
    },
    {
      "epoch": 0.12404548334389276,
      "grad_norm": 2.1348105213857362,
      "learning_rate": 2.902835864223715e-05,
      "loss": 0.6901,
      "step": 930
    },
    {
      "epoch": 0.12537930574543998,
      "grad_norm": 2.547712785034466,
      "learning_rate": 2.9005753201606287e-05,
      "loss": 0.7024,
      "step": 940
    },
    {
      "epoch": 0.12671312814698724,
      "grad_norm": 6.0714243848770195,
      "learning_rate": 2.8982896824573678e-05,
      "loss": 0.665,
      "step": 950
    },
    {
      "epoch": 0.12804695054853446,
      "grad_norm": 2.928297684665459,
      "learning_rate": 2.8959789920649394e-05,
      "loss": 0.7178,
      "step": 960
    },
    {
      "epoch": 0.12938077295008168,
      "grad_norm": 9.456759373673547,
      "learning_rate": 2.893643290383212e-05,
      "loss": 0.6759,
      "step": 970
    },
    {
      "epoch": 0.13071459535162894,
      "grad_norm": 3.9700721280634648,
      "learning_rate": 2.891282619260172e-05,
      "loss": 0.6701,
      "step": 980
    },
    {
      "epoch": 0.13204841775317616,
      "grad_norm": 1.9216628539066036,
      "learning_rate": 2.8888970209911754e-05,
      "loss": 0.6725,
      "step": 990
    },
    {
      "epoch": 0.13338224015472339,
      "grad_norm": 6.51104150647371,
      "learning_rate": 2.8864865383181893e-05,
      "loss": 0.6811,
      "step": 1000
    },
    {
      "epoch": 0.13471606255627064,
      "grad_norm": 7.1765693144596705,
      "learning_rate": 2.8840512144290273e-05,
      "loss": 0.7085,
      "step": 1010
    },
    {
      "epoch": 0.13604988495781786,
      "grad_norm": 4.136786758396937,
      "learning_rate": 2.8815910929565734e-05,
      "loss": 0.7128,
      "step": 1020
    },
    {
      "epoch": 0.1373837073593651,
      "grad_norm": 3.407893493053672,
      "learning_rate": 2.879106217978002e-05,
      "loss": 0.6782,
      "step": 1030
    },
    {
      "epoch": 0.13871752976091234,
      "grad_norm": 2.5759617833329544,
      "learning_rate": 2.8765966340139892e-05,
      "loss": 0.7387,
      "step": 1040
    },
    {
      "epoch": 0.14005135216245956,
      "grad_norm": 8.496713788438253,
      "learning_rate": 2.8740623860279116e-05,
      "loss": 0.6988,
      "step": 1050
    },
    {
      "epoch": 0.14138517456400682,
      "grad_norm": 1.438197837288145,
      "learning_rate": 2.871503519425044e-05,
      "loss": 0.6797,
      "step": 1060
    },
    {
      "epoch": 0.14271899696555404,
      "grad_norm": 3.305947931404651,
      "learning_rate": 2.8689200800517448e-05,
      "loss": 0.727,
      "step": 1070
    },
    {
      "epoch": 0.14405281936710126,
      "grad_norm": 4.384984204254061,
      "learning_rate": 2.866312114194634e-05,
      "loss": 0.7073,
      "step": 1080
    },
    {
      "epoch": 0.14538664176864852,
      "grad_norm": 3.0424755710445455,
      "learning_rate": 2.8636796685797657e-05,
      "loss": 0.6712,
      "step": 1090
    },
    {
      "epoch": 0.14672046417019574,
      "grad_norm": 3.7562374375017415,
      "learning_rate": 2.8610227903717876e-05,
      "loss": 0.6811,
      "step": 1100
    },
    {
      "epoch": 0.14805428657174297,
      "grad_norm": 18.148320631618972,
      "learning_rate": 2.8583415271730994e-05,
      "loss": 0.669,
      "step": 1110
    },
    {
      "epoch": 0.14938810897329022,
      "grad_norm": 10.095266883651389,
      "learning_rate": 2.855635927022998e-05,
      "loss": 0.7034,
      "step": 1120
    },
    {
      "epoch": 0.15072193137483744,
      "grad_norm": 6.298167661594497,
      "learning_rate": 2.8529060383968175e-05,
      "loss": 0.6736,
      "step": 1130
    },
    {
      "epoch": 0.15205575377638467,
      "grad_norm": 1.7112633381805116,
      "learning_rate": 2.850151910205061e-05,
      "loss": 0.7072,
      "step": 1140
    },
    {
      "epoch": 0.15338957617793192,
      "grad_norm": 4.478017946548926,
      "learning_rate": 2.847373591792523e-05,
      "loss": 0.679,
      "step": 1150
    },
    {
      "epoch": 0.15472339857947914,
      "grad_norm": 6.208223721690807,
      "learning_rate": 2.844571132937407e-05,
      "loss": 0.688,
      "step": 1160
    },
    {
      "epoch": 0.15605722098102637,
      "grad_norm": 3.5734211349612974,
      "learning_rate": 2.841744583850431e-05,
      "loss": 0.7054,
      "step": 1170
    },
    {
      "epoch": 0.15739104338257362,
      "grad_norm": 2.9418941665585874,
      "learning_rate": 2.838893995173932e-05,
      "loss": 0.7051,
      "step": 1180
    },
    {
      "epoch": 0.15872486578412084,
      "grad_norm": 2.8796543096657055,
      "learning_rate": 2.836019417980955e-05,
      "loss": 0.6856,
      "step": 1190
    },
    {
      "epoch": 0.16005868818566807,
      "grad_norm": 1.3218998593276705,
      "learning_rate": 2.8331209037743387e-05,
      "loss": 0.69,
      "step": 1200
    },
    {
      "epoch": 0.16139251058721532,
      "grad_norm": 1.566163155963456,
      "learning_rate": 2.8301985044857947e-05,
      "loss": 0.6384,
      "step": 1210
    },
    {
      "epoch": 0.16272633298876255,
      "grad_norm": 6.731577281026319,
      "learning_rate": 2.8272522724749743e-05,
      "loss": 0.7241,
      "step": 1220
    },
    {
      "epoch": 0.16406015539030977,
      "grad_norm": 3.3521245738531813,
      "learning_rate": 2.8242822605285323e-05,
      "loss": 0.6931,
      "step": 1230
    },
    {
      "epoch": 0.16539397779185702,
      "grad_norm": 11.826633616347152,
      "learning_rate": 2.8212885218591812e-05,
      "loss": 0.702,
      "step": 1240
    },
    {
      "epoch": 0.16672780019340425,
      "grad_norm": 1.826627133990473,
      "learning_rate": 2.8182711101047362e-05,
      "loss": 0.7061,
      "step": 1250
    },
    {
      "epoch": 0.16806162259495147,
      "grad_norm": 2.573810458184169,
      "learning_rate": 2.815230079327156e-05,
      "loss": 0.7369,
      "step": 1260
    },
    {
      "epoch": 0.16939544499649872,
      "grad_norm": 3.2839708945300665,
      "learning_rate": 2.8121654840115734e-05,
      "loss": 0.6966,
      "step": 1270
    },
    {
      "epoch": 0.17072926739804595,
      "grad_norm": 2.3799212816187136,
      "learning_rate": 2.809077379065319e-05,
      "loss": 0.6586,
      "step": 1280
    },
    {
      "epoch": 0.17206308979959317,
      "grad_norm": 8.449506341181594,
      "learning_rate": 2.805965819816937e-05,
      "loss": 0.6759,
      "step": 1290
    },
    {
      "epoch": 0.17339691220114042,
      "grad_norm": 19.59186920922371,
      "learning_rate": 2.802830862015196e-05,
      "loss": 0.6537,
      "step": 1300
    },
    {
      "epoch": 0.17473073460268765,
      "grad_norm": 1.9822479021376653,
      "learning_rate": 2.799672561828087e-05,
      "loss": 0.6854,
      "step": 1310
    },
    {
      "epoch": 0.17606455700423487,
      "grad_norm": 34.5695029919346,
      "learning_rate": 2.79649097584182e-05,
      "loss": 0.696,
      "step": 1320
    },
    {
      "epoch": 0.17739837940578213,
      "grad_norm": 1.886275024045834,
      "learning_rate": 2.7932861610598077e-05,
      "loss": 0.7308,
      "step": 1330
    },
    {
      "epoch": 0.17873220180732935,
      "grad_norm": 12.128310547971918,
      "learning_rate": 2.7900581749016466e-05,
      "loss": 0.6775,
      "step": 1340
    },
    {
      "epoch": 0.1800660242088766,
      "grad_norm": 5.164562576250181,
      "learning_rate": 2.7868070752020865e-05,
      "loss": 0.7096,
      "step": 1350
    },
    {
      "epoch": 0.18139984661042383,
      "grad_norm": 10.603161437547913,
      "learning_rate": 2.7835329202099944e-05,
      "loss": 0.7128,
      "step": 1360
    },
    {
      "epoch": 0.18273366901197105,
      "grad_norm": 4.184026615605208,
      "learning_rate": 2.7802357685873117e-05,
      "loss": 0.6958,
      "step": 1370
    },
    {
      "epoch": 0.1840674914135183,
      "grad_norm": 4.135408851494439,
      "learning_rate": 2.7769156794080033e-05,
      "loss": 0.6933,
      "step": 1380
    },
    {
      "epoch": 0.18540131381506553,
      "grad_norm": 1.7535535971850658,
      "learning_rate": 2.7735727121569967e-05,
      "loss": 0.7088,
      "step": 1390
    },
    {
      "epoch": 0.18673513621661275,
      "grad_norm": 5.354882317397714,
      "learning_rate": 2.770206926729121e-05,
      "loss": 0.6777,
      "step": 1400
    },
    {
      "epoch": 0.18806895861816,
      "grad_norm": 3.2593978573810523,
      "learning_rate": 2.7668183834280284e-05,
      "loss": 0.6535,
      "step": 1410
    },
    {
      "epoch": 0.18940278101970723,
      "grad_norm": 2.48929554913285,
      "learning_rate": 2.763407142965117e-05,
      "loss": 0.6853,
      "step": 1420
    },
    {
      "epoch": 0.19073660342125445,
      "grad_norm": 7.550246976134938,
      "learning_rate": 2.759973266458444e-05,
      "loss": 0.679,
      "step": 1430
    },
    {
      "epoch": 0.1920704258228017,
      "grad_norm": 5.033109901639584,
      "learning_rate": 2.756516815431627e-05,
      "loss": 0.6875,
      "step": 1440
    },
    {
      "epoch": 0.19340424822434893,
      "grad_norm": 6.1146492204468545,
      "learning_rate": 2.7530378518127445e-05,
      "loss": 0.6762,
      "step": 1450
    },
    {
      "epoch": 0.19473807062589615,
      "grad_norm": 72.48946246157955,
      "learning_rate": 2.7495364379332256e-05,
      "loss": 0.6766,
      "step": 1460
    },
    {
      "epoch": 0.1960718930274434,
      "grad_norm": 2.9676182245937426,
      "learning_rate": 2.7460126365267335e-05,
      "loss": 0.6731,
      "step": 1470
    },
    {
      "epoch": 0.19740571542899063,
      "grad_norm": 2.31300235137622,
      "learning_rate": 2.7424665107280402e-05,
      "loss": 0.6936,
      "step": 1480
    },
    {
      "epoch": 0.19873953783053785,
      "grad_norm": 2.594823775876976,
      "learning_rate": 2.738898124071898e-05,
      "loss": 0.6703,
      "step": 1490
    },
    {
      "epoch": 0.2000733602320851,
      "grad_norm": 2.8354629666044575,
      "learning_rate": 2.735307540491898e-05,
      "loss": 0.7022,
      "step": 1500
    },
    {
      "epoch": 0.20140718263363233,
      "grad_norm": 4.553033073507926,
      "learning_rate": 2.7316948243193273e-05,
      "loss": 0.6948,
      "step": 1510
    },
    {
      "epoch": 0.20274100503517956,
      "grad_norm": 10.052746707722667,
      "learning_rate": 2.7280600402820146e-05,
      "loss": 0.7216,
      "step": 1520
    },
    {
      "epoch": 0.2040748274367268,
      "grad_norm": 2.3783106567256977,
      "learning_rate": 2.724403253503171e-05,
      "loss": 0.6955,
      "step": 1530
    },
    {
      "epoch": 0.20540864983827403,
      "grad_norm": 2.26171864458522,
      "learning_rate": 2.7207245295002242e-05,
      "loss": 0.7062,
      "step": 1540
    },
    {
      "epoch": 0.20674247223982126,
      "grad_norm": 3.223919319241339,
      "learning_rate": 2.7170239341836436e-05,
      "loss": 0.6659,
      "step": 1550
    },
    {
      "epoch": 0.2080762946413685,
      "grad_norm": 1.6810934171118763,
      "learning_rate": 2.7133015338557585e-05,
      "loss": 0.7069,
      "step": 1560
    },
    {
      "epoch": 0.20941011704291573,
      "grad_norm": 2.1432477935562515,
      "learning_rate": 2.7095573952095727e-05,
      "loss": 0.703,
      "step": 1570
    },
    {
      "epoch": 0.21074393944446296,
      "grad_norm": 1.5654642597457342,
      "learning_rate": 2.705791585327568e-05,
      "loss": 0.7037,
      "step": 1580
    },
    {
      "epoch": 0.2120777618460102,
      "grad_norm": 2.5797486712147992,
      "learning_rate": 2.7020041716805014e-05,
      "loss": 0.6945,
      "step": 1590
    },
    {
      "epoch": 0.21341158424755743,
      "grad_norm": 6.993511462468879,
      "learning_rate": 2.6981952221261986e-05,
      "loss": 0.6965,
      "step": 1600
    }
  ],
  "logging_steps": 10,
  "max_steps": 7497,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 400,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 7.289521573986304e+18,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}