{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9998899122967965,
  "eval_steps": 500,
  "global_step": 3406,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0029356720854280577,
      "grad_norm": 2.3989222049713135,
      "learning_rate": 1.9999583060217186e-05,
      "loss": 1.4046,
      "step": 10
    },
    {
      "epoch": 0.005871344170856115,
      "grad_norm": 2.4013407230377197,
      "learning_rate": 1.9998315387870395e-05,
      "loss": 1.2805,
      "step": 20
    },
    {
      "epoch": 0.008807016256284174,
      "grad_norm": 2.5987584590911865,
      "learning_rate": 1.9996197048273697e-05,
      "loss": 1.2627,
      "step": 30
    },
    {
      "epoch": 0.01174268834171223,
      "grad_norm": 2.5626957416534424,
      "learning_rate": 1.999322822165767e-05,
      "loss": 1.34,
      "step": 40
    },
    {
      "epoch": 0.014678360427140288,
      "grad_norm": 2.4768199920654297,
      "learning_rate": 1.998940916061322e-05,
      "loss": 1.2963,
      "step": 50
    },
    {
      "epoch": 0.017614032512568347,
      "grad_norm": 2.1327669620513916,
      "learning_rate": 1.9984740190070102e-05,
      "loss": 1.3513,
      "step": 60
    },
    {
      "epoch": 0.020549704597996404,
      "grad_norm": 2.7165908813476562,
      "learning_rate": 1.9979221707269273e-05,
      "loss": 1.2629,
      "step": 70
    },
    {
      "epoch": 0.02348537668342446,
      "grad_norm": 2.5773496627807617,
      "learning_rate": 1.997285418172908e-05,
      "loss": 1.2504,
      "step": 80
    },
    {
      "epoch": 0.02642104876885252,
      "grad_norm": 2.7524304389953613,
      "learning_rate": 1.9965638155205335e-05,
      "loss": 1.2854,
      "step": 90
    },
    {
      "epoch": 0.029356720854280576,
      "grad_norm": 2.542572498321533,
      "learning_rate": 1.995757424164521e-05,
      "loss": 1.2583,
      "step": 100
    },
    {
      "epoch": 0.03229239293970863,
      "grad_norm": 3.595125198364258,
      "learning_rate": 1.9948663127135003e-05,
      "loss": 1.2612,
      "step": 110
    },
    {
      "epoch": 0.035228065025136694,
      "grad_norm": 2.8669538497924805,
      "learning_rate": 1.9938905569841754e-05,
      "loss": 1.237,
      "step": 120
    },
    {
      "epoch": 0.03816373711056475,
      "grad_norm": 3.1492984294891357,
      "learning_rate": 1.9928302399948767e-05,
      "loss": 1.2394,
      "step": 130
    },
    {
      "epoch": 0.04109940919599281,
      "grad_norm": 3.0048630237579346,
      "learning_rate": 1.991685451958495e-05,
      "loss": 1.1899,
      "step": 140
    },
    {
      "epoch": 0.04403508128142086,
      "grad_norm": 2.9907774925231934,
      "learning_rate": 1.990456290274808e-05,
      "loss": 1.1939,
      "step": 150
    },
    {
      "epoch": 0.04697075336684892,
      "grad_norm": 3.010820150375366,
      "learning_rate": 1.9891428595221914e-05,
      "loss": 1.2212,
      "step": 160
    },
    {
      "epoch": 0.049906425452276984,
      "grad_norm": 2.486607551574707,
      "learning_rate": 1.9877452714487232e-05,
      "loss": 1.1824,
      "step": 170
    },
    {
      "epoch": 0.05284209753770504,
      "grad_norm": 2.491534471511841,
      "learning_rate": 1.9862636449626752e-05,
      "loss": 1.2118,
      "step": 180
    },
    {
      "epoch": 0.0557777696231331,
      "grad_norm": 2.5148768424987793,
      "learning_rate": 1.9846981061223958e-05,
      "loss": 1.2377,
      "step": 190
    },
    {
      "epoch": 0.05871344170856115,
      "grad_norm": 2.640007734298706,
      "learning_rate": 1.9830487881255864e-05,
      "loss": 1.1995,
      "step": 200
    },
    {
      "epoch": 0.06164911379398921,
      "grad_norm": 2.6175191402435303,
      "learning_rate": 1.981315831297966e-05,
      "loss": 1.1114,
      "step": 210
    },
    {
      "epoch": 0.06458478587941727,
      "grad_norm": 2.8480780124664307,
      "learning_rate": 1.9794993830813358e-05,
      "loss": 1.1897,
      "step": 220
    },
    {
      "epoch": 0.06752045796484532,
      "grad_norm": 2.635657787322998,
      "learning_rate": 1.9775995980210306e-05,
      "loss": 1.1592,
      "step": 230
    },
    {
      "epoch": 0.07045613005027339,
      "grad_norm": 2.8263065814971924,
      "learning_rate": 1.9756166377527734e-05,
      "loss": 1.1712,
      "step": 240
    },
    {
      "epoch": 0.07339180213570144,
      "grad_norm": 3.2777886390686035,
      "learning_rate": 1.9735506709889213e-05,
      "loss": 1.1724,
      "step": 250
    },
    {
      "epoch": 0.0763274742211295,
      "grad_norm": 2.587409019470215,
      "learning_rate": 1.9714018735041125e-05,
      "loss": 1.1767,
      "step": 260
    },
    {
      "epoch": 0.07926314630655756,
      "grad_norm": 2.9432199001312256,
      "learning_rate": 1.9691704281203098e-05,
      "loss": 1.1553,
      "step": 270
    },
    {
      "epoch": 0.08219881839198562,
      "grad_norm": 3.1995699405670166,
      "learning_rate": 1.966856524691247e-05,
      "loss": 1.1577,
      "step": 280
    },
    {
      "epoch": 0.08513449047741367,
      "grad_norm": 2.805522918701172,
      "learning_rate": 1.9644603600862753e-05,
      "loss": 1.1627,
      "step": 290
    },
    {
      "epoch": 0.08807016256284173,
      "grad_norm": 2.6327457427978516,
      "learning_rate": 1.961982138173615e-05,
      "loss": 1.1483,
      "step": 300
    },
    {
      "epoch": 0.09100583464826979,
      "grad_norm": 4.197879314422607,
      "learning_rate": 1.959422069803007e-05,
      "loss": 1.2015,
      "step": 310
    },
    {
      "epoch": 0.09394150673369785,
      "grad_norm": 3.2353358268737793,
      "learning_rate": 1.956780372787777e-05,
      "loss": 1.1068,
      "step": 320
    },
    {
      "epoch": 0.0968771788191259,
      "grad_norm": 2.8246076107025146,
      "learning_rate": 1.9540572718863012e-05,
      "loss": 1.1424,
      "step": 330
    },
    {
      "epoch": 0.09981285090455397,
      "grad_norm": 4.32204532623291,
      "learning_rate": 1.9512529987828853e-05,
      "loss": 1.1508,
      "step": 340
    },
    {
      "epoch": 0.10274852298998202,
      "grad_norm": 2.8142335414886475,
      "learning_rate": 1.9483677920680512e-05,
      "loss": 1.1623,
      "step": 350
    },
    {
      "epoch": 0.10568419507541008,
      "grad_norm": 2.9336957931518555,
      "learning_rate": 1.9454018972182383e-05,
      "loss": 1.161,
      "step": 360
    },
    {
      "epoch": 0.10861986716083813,
      "grad_norm": 2.9903533458709717,
      "learning_rate": 1.9423555665749182e-05,
      "loss": 1.1444,
      "step": 370
    },
    {
      "epoch": 0.1115555392462662,
      "grad_norm": 2.8897149562835693,
      "learning_rate": 1.939229059323124e-05,
      "loss": 1.171,
      "step": 380
    },
    {
      "epoch": 0.11449121133169425,
      "grad_norm": 3.222294807434082,
      "learning_rate": 1.9360226414694008e-05,
      "loss": 1.1365,
      "step": 390
    },
    {
      "epoch": 0.1174268834171223,
      "grad_norm": 2.9000742435455322,
      "learning_rate": 1.932736585819171e-05,
      "loss": 1.1342,
      "step": 400
    },
    {
      "epoch": 0.12036255550255036,
      "grad_norm": 2.682969093322754,
      "learning_rate": 1.929371171953526e-05,
      "loss": 1.0428,
      "step": 410
    },
    {
      "epoch": 0.12329822758797843,
      "grad_norm": 2.8611629009246826,
      "learning_rate": 1.9259266862054366e-05,
      "loss": 1.1135,
      "step": 420
    },
    {
      "epoch": 0.12623389967340648,
      "grad_norm": 2.9910166263580322,
      "learning_rate": 1.9224034216353947e-05,
      "loss": 1.1154,
      "step": 430
    },
    {
      "epoch": 0.12916957175883453,
      "grad_norm": 3.245227336883545,
      "learning_rate": 1.9188016780064768e-05,
      "loss": 1.2029,
      "step": 440
    },
    {
      "epoch": 0.1321052438442626,
      "grad_norm": 3.2252321243286133,
      "learning_rate": 1.9151217617588412e-05,
      "loss": 1.1272,
      "step": 450
    },
    {
      "epoch": 0.13504091592969064,
      "grad_norm": 3.23498272895813,
      "learning_rate": 1.9113639859836544e-05,
      "loss": 1.1421,
      "step": 460
    },
    {
      "epoch": 0.13797658801511872,
      "grad_norm": 3.0721843242645264,
      "learning_rate": 1.9075286703964554e-05,
      "loss": 1.14,
      "step": 470
    },
    {
      "epoch": 0.14091226010054678,
      "grad_norm": 3.130610704421997,
      "learning_rate": 1.9036161413099512e-05,
      "loss": 1.1699,
      "step": 480
    },
    {
      "epoch": 0.14384793218597483,
      "grad_norm": 3.054914712905884,
      "learning_rate": 1.899626731606255e-05,
      "loss": 1.0919,
      "step": 490
    },
    {
      "epoch": 0.14678360427140288,
      "grad_norm": 3.3167009353637695,
      "learning_rate": 1.895560780708565e-05,
      "loss": 1.0625,
      "step": 500
    },
    {
      "epoch": 0.14971927635683094,
      "grad_norm": 3.075392484664917,
      "learning_rate": 1.8914186345522846e-05,
      "loss": 1.0899,
      "step": 510
    },
    {
      "epoch": 0.152654948442259,
      "grad_norm": 3.1269266605377197,
      "learning_rate": 1.8872006455555906e-05,
      "loss": 1.1148,
      "step": 520
    },
    {
      "epoch": 0.15559062052768705,
      "grad_norm": 3.867361068725586,
      "learning_rate": 1.8829071725894483e-05,
      "loss": 1.02,
      "step": 530
    },
    {
      "epoch": 0.15852629261311513,
      "grad_norm": 3.529639720916748,
      "learning_rate": 1.87853858094708e-05,
      "loss": 1.1167,
      "step": 540
    },
    {
      "epoch": 0.16146196469854318,
      "grad_norm": 3.098249673843384,
      "learning_rate": 1.8740952423128842e-05,
      "loss": 1.0181,
      "step": 550
    },
    {
      "epoch": 0.16439763678397123,
      "grad_norm": 3.1614904403686523,
      "learning_rate": 1.869577534730812e-05,
      "loss": 1.1118,
      "step": 560
    },
    {
      "epoch": 0.1673333088693993,
      "grad_norm": 3.054616928100586,
      "learning_rate": 1.8649858425722033e-05,
      "loss": 1.0666,
      "step": 570
    },
    {
      "epoch": 0.17026898095482734,
      "grad_norm": 3.479527711868286,
      "learning_rate": 1.8603205565030846e-05,
      "loss": 1.108,
      "step": 580
    },
    {
      "epoch": 0.1732046530402554,
      "grad_norm": 2.9523024559020996,
      "learning_rate": 1.8555820734509297e-05,
      "loss": 1.0833,
      "step": 590
    },
    {
      "epoch": 0.17614032512568345,
      "grad_norm": 2.9907584190368652,
      "learning_rate": 1.8507707965708892e-05,
      "loss": 1.0283,
      "step": 600
    },
    {
      "epoch": 0.17907599721111153,
      "grad_norm": 3.2911376953125,
      "learning_rate": 1.8458871352114894e-05,
      "loss": 1.0747,
      "step": 610
    },
    {
      "epoch": 0.18201166929653959,
      "grad_norm": 3.1361849308013916,
      "learning_rate": 1.840931504879806e-05,
      "loss": 1.11,
      "step": 620
    },
    {
      "epoch": 0.18494734138196764,
      "grad_norm": 3.527332067489624,
      "learning_rate": 1.8359043272061086e-05,
      "loss": 1.0424,
      "step": 630
    },
    {
      "epoch": 0.1878830134673957,
      "grad_norm": 3.5494275093078613,
      "learning_rate": 1.8308060299079926e-05,
      "loss": 1.0818,
      "step": 640
    },
    {
      "epoch": 0.19081868555282375,
      "grad_norm": 3.4427106380462646,
      "learning_rate": 1.8256370467539847e-05,
      "loss": 1.0883,
      "step": 650
    },
    {
      "epoch": 0.1937543576382518,
      "grad_norm": 3.092515230178833,
      "learning_rate": 1.82039781752664e-05,
      "loss": 1.0285,
      "step": 660
    },
    {
      "epoch": 0.19669002972367985,
      "grad_norm": 2.667904853820801,
      "learning_rate": 1.815088787985124e-05,
      "loss": 0.9751,
      "step": 670
    },
    {
      "epoch": 0.19962570180910794,
      "grad_norm": 3.5892174243927,
      "learning_rate": 1.809710409827285e-05,
      "loss": 1.0603,
      "step": 680
    },
    {
      "epoch": 0.202561373894536,
      "grad_norm": 3.122434616088867,
      "learning_rate": 1.804263140651227e-05,
      "loss": 1.0919,
      "step": 690
    },
    {
      "epoch": 0.20549704597996404,
      "grad_norm": 2.9182698726654053,
      "learning_rate": 1.798747443916374e-05,
      "loss": 1.0553,
      "step": 700
    },
    {
      "epoch": 0.2084327180653921,
      "grad_norm": 3.260917901992798,
      "learning_rate": 1.793163788904038e-05,
      "loss": 1.0711,
      "step": 710
    },
    {
      "epoch": 0.21136839015082015,
      "grad_norm": 3.4142649173736572,
      "learning_rate": 1.7875126506774956e-05,
      "loss": 1.0423,
      "step": 720
    },
    {
      "epoch": 0.2143040622362482,
      "grad_norm": 3.0127294063568115,
      "learning_rate": 1.781794510041564e-05,
      "loss": 1.0679,
      "step": 730
    },
    {
      "epoch": 0.21723973432167626,
      "grad_norm": 3.398015022277832,
      "learning_rate": 1.776009853501698e-05,
      "loss": 1.0558,
      "step": 740
    },
    {
      "epoch": 0.2201754064071043,
      "grad_norm": 3.6017568111419678,
      "learning_rate": 1.770159173222595e-05,
      "loss": 1.0198,
      "step": 750
    },
    {
      "epoch": 0.2231110784925324,
      "grad_norm": 3.5204339027404785,
      "learning_rate": 1.7642429669863225e-05,
      "loss": 0.9951,
      "step": 760
    },
    {
      "epoch": 0.22604675057796045,
      "grad_norm": 3.3134777545928955,
      "learning_rate": 1.7582617381499655e-05,
      "loss": 0.9906,
      "step": 770
    },
    {
      "epoch": 0.2289824226633885,
      "grad_norm": 3.5017244815826416,
      "learning_rate": 1.7522159956028003e-05,
      "loss": 1.0711,
      "step": 780
    },
    {
      "epoch": 0.23191809474881656,
      "grad_norm": 3.1634137630462646,
      "learning_rate": 1.7461062537229987e-05,
      "loss": 0.9909,
      "step": 790
    },
    {
      "epoch": 0.2348537668342446,
      "grad_norm": 3.368623971939087,
      "learning_rate": 1.739933032333863e-05,
      "loss": 0.9815,
      "step": 800
    },
    {
      "epoch": 0.23778943891967266,
      "grad_norm": 3.1064817905426025,
      "learning_rate": 1.733696856659599e-05,
      "loss": 1.0191,
      "step": 810
    },
    {
      "epoch": 0.24072511100510072,
      "grad_norm": 3.2074899673461914,
      "learning_rate": 1.7273982572806303e-05,
      "loss": 1.0314,
      "step": 820
    },
    {
      "epoch": 0.2436607830905288,
      "grad_norm": 2.5882649421691895,
      "learning_rate": 1.721037770088455e-05,
      "loss": 0.958,
      "step": 830
    },
    {
      "epoch": 0.24659645517595685,
      "grad_norm": 3.730363130569458,
      "learning_rate": 1.7146159362400515e-05,
      "loss": 1.0272,
      "step": 840
    },
    {
      "epoch": 0.2495321272613849,
      "grad_norm": 2.940425395965576,
      "learning_rate": 1.708133302111837e-05,
      "loss": 1.0437,
      "step": 850
    },
    {
      "epoch": 0.25246779934681296,
      "grad_norm": 4.833681106567383,
      "learning_rate": 1.7015904192531814e-05,
      "loss": 1.0393,
      "step": 860
    },
    {
      "epoch": 0.255403471432241,
      "grad_norm": 3.417707681655884,
      "learning_rate": 1.694987844339479e-05,
      "loss": 1.0602,
      "step": 870
    },
    {
      "epoch": 0.25833914351766907,
      "grad_norm": 3.239388942718506,
      "learning_rate": 1.6883261391247888e-05,
      "loss": 0.9515,
      "step": 880
    },
    {
      "epoch": 0.2612748156030971,
      "grad_norm": 3.1867291927337646,
      "learning_rate": 1.6816058703940366e-05,
      "loss": 0.9961,
      "step": 890
    },
    {
      "epoch": 0.2642104876885252,
      "grad_norm": 3.193343162536621,
      "learning_rate": 1.6748276099147952e-05,
      "loss": 1.0066,
      "step": 900
    },
    {
      "epoch": 0.26714615977395323,
      "grad_norm": 3.1413753032684326,
      "learning_rate": 1.6679919343886376e-05,
      "loss": 0.9714,
      "step": 910
    },
    {
      "epoch": 0.2700818318593813,
      "grad_norm": 3.0826566219329834,
      "learning_rate": 1.661099425402067e-05,
      "loss": 0.9689,
      "step": 920
    },
    {
      "epoch": 0.2730175039448094,
      "grad_norm": 3.5959160327911377,
      "learning_rate": 1.6541506693770403e-05,
      "loss": 0.9867,
      "step": 930
    },
    {
      "epoch": 0.27595317603023745,
      "grad_norm": 3.8435122966766357,
      "learning_rate": 1.647146257521071e-05,
      "loss": 1.0281,
      "step": 940
    },
    {
      "epoch": 0.2788888481156655,
      "grad_norm": 3.396488904953003,
      "learning_rate": 1.6400867857769287e-05,
      "loss": 0.975,
      "step": 950
    },
    {
      "epoch": 0.28182452020109355,
      "grad_norm": 3.2766590118408203,
      "learning_rate": 1.6329728547719375e-05,
      "loss": 0.9373,
      "step": 960
    },
    {
      "epoch": 0.2847601922865216,
      "grad_norm": 3.673755645751953,
      "learning_rate": 1.625805069766873e-05,
      "loss": 0.9651,
      "step": 970
    },
    {
      "epoch": 0.28769586437194966,
      "grad_norm": 3.8751864433288574,
      "learning_rate": 1.6185840406044657e-05,
      "loss": 0.9262,
      "step": 980
    },
    {
      "epoch": 0.2906315364573777,
      "grad_norm": 3.708500623703003,
      "learning_rate": 1.611310381657515e-05,
      "loss": 0.9972,
      "step": 990
    },
    {
      "epoch": 0.29356720854280577,
      "grad_norm": 3.4258711338043213,
      "learning_rate": 1.60398471177662e-05,
      "loss": 0.9331,
      "step": 1000
    },
    {
      "epoch": 0.2965028806282338,
      "grad_norm": 3.4662258625030518,
      "learning_rate": 1.596607654237522e-05,
      "loss": 0.9592,
      "step": 1010
    },
    {
      "epoch": 0.2994385527136619,
      "grad_norm": 2.938396453857422,
      "learning_rate": 1.589179836688081e-05,
      "loss": 0.9568,
      "step": 1020
    },
    {
      "epoch": 0.30237422479908993,
      "grad_norm": 3.248762845993042,
      "learning_rate": 1.5817018910948712e-05,
      "loss": 0.9928,
      "step": 1030
    },
    {
      "epoch": 0.305309896884518,
      "grad_norm": 3.423213243484497,
      "learning_rate": 1.574174453689415e-05,
      "loss": 0.9387,
      "step": 1040
    },
    {
      "epoch": 0.30824556896994604,
      "grad_norm": 3.249216318130493,
      "learning_rate": 1.566598164914049e-05,
      "loss": 0.8925,
      "step": 1050
    },
    {
      "epoch": 0.3111812410553741,
      "grad_norm": 3.6318016052246094,
      "learning_rate": 1.5589736693674372e-05,
      "loss": 1.0153,
      "step": 1060
    },
    {
      "epoch": 0.31411691314080215,
      "grad_norm": 3.9752533435821533,
      "learning_rate": 1.551301615749726e-05,
      "loss": 0.9323,
      "step": 1070
    },
    {
      "epoch": 0.31705258522623025,
      "grad_norm": 3.46864914894104,
      "learning_rate": 1.5435826568073532e-05,
      "loss": 0.8901,
      "step": 1080
    },
    {
      "epoch": 0.3199882573116583,
      "grad_norm": 4.399304389953613,
      "learning_rate": 1.535817449277511e-05,
      "loss": 0.9118,
      "step": 1090
    },
    {
      "epoch": 0.32292392939708636,
      "grad_norm": 3.2890026569366455,
      "learning_rate": 1.5280066538322703e-05,
      "loss": 0.8655,
      "step": 1100
    },
    {
      "epoch": 0.3258596014825144,
      "grad_norm": 3.491983652114868,
      "learning_rate": 1.5201509350223708e-05,
      "loss": 0.9217,
      "step": 1110
    },
    {
      "epoch": 0.32879527356794247,
      "grad_norm": 5.0014824867248535,
      "learning_rate": 1.5122509612206785e-05,
      "loss": 0.9362,
      "step": 1120
    },
    {
      "epoch": 0.3317309456533705,
      "grad_norm": 4.092339038848877,
      "learning_rate": 1.5043074045653215e-05,
      "loss": 0.9262,
      "step": 1130
    },
    {
      "epoch": 0.3346666177387986,
      "grad_norm": 3.286433219909668,
      "learning_rate": 1.496320940902503e-05,
      "loss": 0.8891,
      "step": 1140
    },
    {
      "epoch": 0.33760228982422663,
      "grad_norm": 3.6521873474121094,
      "learning_rate": 1.4882922497290007e-05,
      "loss": 0.9281,
      "step": 1150
    },
    {
      "epoch": 0.3405379619096547,
      "grad_norm": 3.8015809059143066,
      "learning_rate": 1.4802220141343516e-05,
      "loss": 0.8949,
      "step": 1160
    },
    {
      "epoch": 0.34347363399508274,
      "grad_norm": 4.149661064147949,
      "learning_rate": 1.472110920742738e-05,
      "loss": 0.8889,
      "step": 1170
    },
    {
      "epoch": 0.3464093060805108,
      "grad_norm": 3.5252785682678223,
      "learning_rate": 1.4639596596545656e-05,
      "loss": 0.8397,
      "step": 1180
    },
    {
      "epoch": 0.34934497816593885,
      "grad_norm": 3.6884541511535645,
      "learning_rate": 1.4557689243877507e-05,
      "loss": 0.9142,
      "step": 1190
    },
    {
      "epoch": 0.3522806502513669,
      "grad_norm": 3.9577550888061523,
      "learning_rate": 1.4475394118187146e-05,
      "loss": 0.9809,
      "step": 1200
    },
    {
      "epoch": 0.35521632233679495,
      "grad_norm": 3.6897339820861816,
      "learning_rate": 1.4392718221230917e-05,
      "loss": 0.9141,
      "step": 1210
    },
    {
      "epoch": 0.35815199442222306,
      "grad_norm": 3.061516046524048,
      "learning_rate": 1.4309668587161596e-05,
      "loss": 0.8669,
      "step": 1220
    },
    {
      "epoch": 0.3610876665076511,
      "grad_norm": 3.1191623210906982,
      "learning_rate": 1.4226252281929902e-05,
      "loss": 0.8384,
      "step": 1230
    },
    {
      "epoch": 0.36402333859307917,
      "grad_norm": 4.198310852050781,
      "learning_rate": 1.4142476402683327e-05,
      "loss": 0.8971,
      "step": 1240
    },
    {
      "epoch": 0.3669590106785072,
      "grad_norm": 3.8184337615966797,
      "learning_rate": 1.4058348077162301e-05,
      "loss": 0.8783,
      "step": 1250
    },
    {
      "epoch": 0.3698946827639353,
      "grad_norm": 3.842637777328491,
      "learning_rate": 1.3973874463093747e-05,
      "loss": 0.9623,
      "step": 1260
    },
    {
      "epoch": 0.37283035484936333,
      "grad_norm": 3.5173559188842773,
      "learning_rate": 1.3889062747582118e-05,
      "loss": 0.8092,
      "step": 1270
    },
    {
      "epoch": 0.3757660269347914,
      "grad_norm": 3.8953890800476074,
      "learning_rate": 1.3803920146497887e-05,
      "loss": 0.8762,
      "step": 1280
    },
    {
      "epoch": 0.37870169902021944,
      "grad_norm": 3.0550928115844727,
      "learning_rate": 1.3718453903863616e-05,
      "loss": 0.8321,
      "step": 1290
    },
    {
      "epoch": 0.3816373711056475,
      "grad_norm": 3.9677445888519287,
      "learning_rate": 1.3632671291237645e-05,
      "loss": 0.8566,
      "step": 1300
    },
    {
      "epoch": 0.38457304319107555,
      "grad_norm": 3.887268304824829,
      "learning_rate": 1.35465796070954e-05,
      "loss": 0.8944,
      "step": 1310
    },
    {
      "epoch": 0.3875087152765036,
      "grad_norm": 3.1006393432617188,
      "learning_rate": 1.3460186176208439e-05,
      "loss": 0.7583,
      "step": 1320
    },
    {
      "epoch": 0.39044438736193166,
      "grad_norm": 3.7594895362854004,
      "learning_rate": 1.337349834902125e-05,
      "loss": 0.814,
      "step": 1330
    },
    {
      "epoch": 0.3933800594473597,
      "grad_norm": 4.34951114654541,
      "learning_rate": 1.328652350102588e-05,
      "loss": 0.8006,
      "step": 1340
    },
    {
      "epoch": 0.39631573153278776,
      "grad_norm": 2.9645636081695557,
      "learning_rate": 1.3199269032134395e-05,
      "loss": 0.8129,
      "step": 1350
    },
    {
      "epoch": 0.39925140361821587,
      "grad_norm": 3.858602285385132,
      "learning_rate": 1.3111742366049317e-05,
      "loss": 0.8366,
      "step": 1360
    },
    {
      "epoch": 0.4021870757036439,
      "grad_norm": 3.2778103351593018,
      "learning_rate": 1.3023950949631979e-05,
      "loss": 0.8551,
      "step": 1370
    },
    {
      "epoch": 0.405122747789072,
      "grad_norm": 3.2402875423431396,
      "learning_rate": 1.2935902252268965e-05,
      "loss": 0.8398,
      "step": 1380
    },
    {
      "epoch": 0.40805841987450003,
      "grad_norm": 4.325957775115967,
      "learning_rate": 1.2847603765236589e-05,
      "loss": 0.836,
      "step": 1390
    },
    {
      "epoch": 0.4109940919599281,
      "grad_norm": 3.5310022830963135,
      "learning_rate": 1.2759063001063531e-05,
      "loss": 0.8369,
      "step": 1400
    },
    {
      "epoch": 0.41392976404535614,
      "grad_norm": 3.5352087020874023,
      "learning_rate": 1.2670287492891675e-05,
      "loss": 0.8988,
      "step": 1410
    },
    {
      "epoch": 0.4168654361307842,
      "grad_norm": 3.190788745880127,
      "learning_rate": 1.258128479383516e-05,
      "loss": 0.8352,
      "step": 1420
    },
    {
      "epoch": 0.41980110821621225,
      "grad_norm": 3.459728240966797,
      "learning_rate": 1.249206247633778e-05,
      "loss": 0.8295,
      "step": 1430
    },
    {
      "epoch": 0.4227367803016403,
      "grad_norm": 3.5794529914855957,
      "learning_rate": 1.2402628131528686e-05,
      "loss": 0.8103,
      "step": 1440
    },
    {
      "epoch": 0.42567245238706836,
      "grad_norm": 4.169612407684326,
      "learning_rate": 1.2312989368576547e-05,
      "loss": 0.7757,
      "step": 1450
    },
    {
      "epoch": 0.4286081244724964,
      "grad_norm": 3.301011562347412,
      "learning_rate": 1.2223153814042137e-05,
      "loss": 0.7871,
      "step": 1460
    },
    {
      "epoch": 0.43154379655792446,
      "grad_norm": 4.524185657501221,
      "learning_rate": 1.2133129111229466e-05,
      "loss": 0.851,
      "step": 1470
    },
    {
      "epoch": 0.4344794686433525,
      "grad_norm": 3.72041392326355,
      "learning_rate": 1.2042922919535484e-05,
      "loss": 0.803,
      "step": 1480
    },
    {
      "epoch": 0.43741514072878057,
      "grad_norm": 3.926424503326416,
      "learning_rate": 1.1952542913798406e-05,
      "loss": 0.761,
      "step": 1490
    },
    {
      "epoch": 0.4403508128142086,
      "grad_norm": 3.5725414752960205,
      "learning_rate": 1.1861996783644727e-05,
      "loss": 0.8086,
      "step": 1500
    },
    {
      "epoch": 0.44328648489963673,
      "grad_norm": 4.109748363494873,
      "learning_rate": 1.1771292232834983e-05,
      "loss": 0.8483,
      "step": 1510
    },
    {
      "epoch": 0.4462221569850648,
      "grad_norm": 3.673794984817505,
      "learning_rate": 1.1680436978608314e-05,
      "loss": 0.738,
      "step": 1520
    },
    {
      "epoch": 0.44915782907049284,
      "grad_norm": 3.831571102142334,
      "learning_rate": 1.1589438751025852e-05,
      "loss": 0.7462,
      "step": 1530
    },
    {
      "epoch": 0.4520935011559209,
      "grad_norm": 4.181507587432861,
      "learning_rate": 1.149830529231307e-05,
      "loss": 0.7707,
      "step": 1540
    },
    {
      "epoch": 0.45502917324134895,
      "grad_norm": 3.3295936584472656,
      "learning_rate": 1.140704435620104e-05,
      "loss": 0.7832,
      "step": 1550
    },
    {
      "epoch": 0.457964845326777,
      "grad_norm": 4.025683403015137,
      "learning_rate": 1.1315663707266742e-05,
      "loss": 0.74,
      "step": 1560
    },
    {
      "epoch": 0.46090051741220506,
      "grad_norm": 3.792701244354248,
      "learning_rate": 1.1224171120272455e-05,
      "loss": 0.6698,
      "step": 1570
    },
    {
      "epoch": 0.4638361894976331,
      "grad_norm": 3.7220959663391113,
      "learning_rate": 1.1132574379504269e-05,
      "loss": 0.7604,
      "step": 1580
    },
    {
      "epoch": 0.46677186158306117,
      "grad_norm": 4.423033714294434,
      "learning_rate": 1.1040881278109784e-05,
      "loss": 0.7466,
      "step": 1590
    },
    {
      "epoch": 0.4697075336684892,
      "grad_norm": 3.633347272872925,
      "learning_rate": 1.0949099617435062e-05,
      "loss": 0.7452,
      "step": 1600
    },
    {
      "epoch": 0.4726432057539173,
      "grad_norm": 3.661238431930542,
      "learning_rate": 1.0857237206360885e-05,
      "loss": 0.7637,
      "step": 1610
    },
    {
      "epoch": 0.4755788778393453,
      "grad_norm": 4.33590030670166,
      "learning_rate": 1.0765301860638364e-05,
      "loss": 0.7364,
      "step": 1620
    },
    {
      "epoch": 0.4785145499247734,
      "grad_norm": 3.7030036449432373,
      "learning_rate": 1.0673301402223964e-05,
      "loss": 0.7356,
      "step": 1630
    },
    {
      "epoch": 0.48145022201020143,
      "grad_norm": 4.784999847412109,
      "learning_rate": 1.0581243658614013e-05,
      "loss": 0.765,
      "step": 1640
    },
    {
      "epoch": 0.48438589409562954,
      "grad_norm": 3.2158679962158203,
      "learning_rate": 1.0489136462178718e-05,
      "loss": 0.75,
      "step": 1650
    },
    {
      "epoch": 0.4873215661810576,
      "grad_norm": 4.584315299987793,
      "learning_rate": 1.039698764949579e-05,
      "loss": 0.7347,
      "step": 1660
    },
    {
      "epoch": 0.49025723826648565,
      "grad_norm": 3.4453585147857666,
      "learning_rate": 1.0304805060683692e-05,
      "loss": 0.7887,
      "step": 1670
    },
    {
      "epoch": 0.4931929103519137,
      "grad_norm": 3.9263744354248047,
      "learning_rate": 1.021259653873459e-05,
      "loss": 0.7492,
      "step": 1680
    },
    {
      "epoch": 0.49612858243734176,
      "grad_norm": 4.6535539627075195,
      "learning_rate": 1.012036992884708e-05,
      "loss": 0.7676,
      "step": 1690
    },
    {
      "epoch": 0.4990642545227698,
      "grad_norm": 4.22018575668335,
      "learning_rate": 1.0028133077758688e-05,
      "loss": 0.7088,
      "step": 1700
    },
    {
      "epoch": 0.5019999266081978,
      "grad_norm": 4.408539295196533,
      "learning_rate": 9.935893833078284e-06,
      "loss": 0.7646,
      "step": 1710
    },
    {
      "epoch": 0.5049355986936259,
      "grad_norm": 5.264422416687012,
      "learning_rate": 9.843660042618372e-06,
      "loss": 0.8147,
      "step": 1720
    },
    {
      "epoch": 0.507871270779054,
      "grad_norm": 4.2693047523498535,
      "learning_rate": 9.75143955372742e-06,
      "loss": 0.7104,
      "step": 1730
    },
    {
      "epoch": 0.510806942864482,
      "grad_norm": 4.856871128082275,
      "learning_rate": 9.659240212622175e-06,
      "loss": 0.7367,
      "step": 1740
    },
    {
      "epoch": 0.5137426149499101,
      "grad_norm": 2.8976457118988037,
      "learning_rate": 9.567069863720113e-06,
      "loss": 0.7564,
      "step": 1750
    },
    {
      "epoch": 0.5166782870353381,
      "grad_norm": 5.992892742156982,
      "learning_rate": 9.474936348972021e-06,
      "loss": 0.7735,
      "step": 1760
    },
    {
      "epoch": 0.5196139591207662,
      "grad_norm": 3.6526339054107666,
      "learning_rate": 9.382847507194797e-06,
      "loss": 0.7035,
      "step": 1770
    },
    {
      "epoch": 0.5225496312061942,
      "grad_norm": 4.040701389312744,
      "learning_rate": 9.290811173404513e-06,
      "loss": 0.6347,
      "step": 1780
    },
    {
      "epoch": 0.5254853032916224,
      "grad_norm": 3.848483085632324,
      "learning_rate": 9.198835178149807e-06,
      "loss": 0.6359,
      "step": 1790
    },
    {
      "epoch": 0.5284209753770504,
      "grad_norm": 3.2821764945983887,
      "learning_rate": 9.106927346845663e-06,
      "loss": 0.7137,
      "step": 1800
    },
    {
      "epoch": 0.5313566474624785,
      "grad_norm": 4.672881603240967,
      "learning_rate": 9.015095499107578e-06,
      "loss": 0.7085,
      "step": 1810
    },
    {
      "epoch": 0.5342923195479065,
      "grad_norm": 3.976231098175049,
      "learning_rate": 8.923347448086311e-06,
      "loss": 0.6501,
      "step": 1820
    },
    {
      "epoch": 0.5372279916333346,
      "grad_norm": 4.726049423217773,
      "learning_rate": 8.831690999803101e-06,
      "loss": 0.8129,
      "step": 1830
    },
    {
      "epoch": 0.5401636637187626,
      "grad_norm": 6.278385162353516,
      "learning_rate": 8.740133952485515e-06,
      "loss": 0.6732,
      "step": 1840
    },
    {
      "epoch": 0.5430993358041907,
      "grad_norm": 4.620763301849365,
      "learning_rate": 8.648684095904001e-06,
      "loss": 0.6872,
      "step": 1850
    },
    {
      "epoch": 0.5460350078896188,
      "grad_norm": 4.494777679443359,
      "learning_rate": 8.557349210709098e-06,
      "loss": 0.6686,
      "step": 1860
    },
    {
      "epoch": 0.5489706799750468,
      "grad_norm": 4.2295637130737305,
      "learning_rate": 8.46613706776945e-06,
      "loss": 0.6853,
      "step": 1870
    },
    {
      "epoch": 0.5519063520604749,
      "grad_norm": 3.5783040523529053,
      "learning_rate": 8.375055427510673e-06,
      "loss": 0.6923,
      "step": 1880
    },
    {
      "epoch": 0.5548420241459029,
      "grad_norm": 3.5585546493530273,
      "learning_rate": 8.284112039255071e-06,
      "loss": 0.6744,
      "step": 1890
    },
    {
      "epoch": 0.557777696231331,
      "grad_norm": 3.939253330230713,
      "learning_rate": 8.193314640562315e-06,
      "loss": 0.627,
      "step": 1900
    },
    {
      "epoch": 0.560713368316759,
      "grad_norm": 3.630519390106201,
      "learning_rate": 8.102670956571139e-06,
      "loss": 0.6627,
      "step": 1910
    },
    {
      "epoch": 0.5636490404021871,
      "grad_norm": 11.943046569824219,
      "learning_rate": 8.012188699342072e-06,
      "loss": 0.6476,
      "step": 1920
    },
    {
      "epoch": 0.5665847124876151,
      "grad_norm": 5.358550071716309,
      "learning_rate": 7.92187556720126e-06,
      "loss": 0.6968,
      "step": 1930
    },
    {
      "epoch": 0.5695203845730432,
      "grad_norm": 3.8031585216522217,
      "learning_rate": 7.831739244085534e-06,
      "loss": 0.6811,
      "step": 1940
    },
    {
      "epoch": 0.5724560566584712,
      "grad_norm": 3.1659951210021973,
      "learning_rate": 7.741787398888617e-06,
      "loss": 0.6501,
      "step": 1950
    },
    {
      "epoch": 0.5753917287438993,
      "grad_norm": 3.7877001762390137,
      "learning_rate": 7.652027684808644e-06,
      "loss": 0.6496,
      "step": 1960
    },
    {
      "epoch": 0.5783274008293273,
      "grad_norm": 4.701345920562744,
      "learning_rate": 7.56246773869705e-06,
      "loss": 0.659,
      "step": 1970
    },
    {
      "epoch": 0.5812630729147554,
      "grad_norm": 4.617175579071045,
      "learning_rate": 7.47311518040879e-06,
      "loss": 0.6429,
      "step": 1980
    },
    {
      "epoch": 0.5841987450001834,
      "grad_norm": 5.269269943237305,
      "learning_rate": 7.3839776121540385e-06,
      "loss": 0.6845,
      "step": 1990
    },
    {
      "epoch": 0.5871344170856115,
      "grad_norm": 3.911558151245117,
      "learning_rate": 7.2950626178514e-06,
      "loss": 0.6536,
      "step": 2000
    },
    {
      "epoch": 0.5900700891710396,
      "grad_norm": NaN,
      "learning_rate": 7.215235676567183e-06,
      "loss": 0.6691,
      "step": 2010
    },
    {
      "epoch": 0.5930057612564676,
      "grad_norm": 5.29760217666626,
      "learning_rate": 7.126764398128368e-06,
      "loss": 0.6483,
      "step": 2020
    },
    {
      "epoch": 0.5959414333418958,
      "grad_norm": 3.4294636249542236,
      "learning_rate": 7.038537577614009e-06,
      "loss": 0.5965,
      "step": 2030
    },
    {
      "epoch": 0.5988771054273238,
      "grad_norm": 3.6569931507110596,
      "learning_rate": 6.950562721455325e-06,
      "loss": 0.5782,
      "step": 2040
    },
    {
      "epoch": 0.6018127775127519,
      "grad_norm": 3.845431089401245,
      "learning_rate": 6.86284731464614e-06,
      "loss": 0.6419,
      "step": 2050
    },
    {
      "epoch": 0.6047484495981799,
      "grad_norm": 3.8947107791900635,
      "learning_rate": 6.775398820106065e-06,
      "loss": 0.5942,
      "step": 2060
    },
    {
      "epoch": 0.607684121683608,
      "grad_norm": 5.501591682434082,
      "learning_rate": 6.688224678045507e-06,
      "loss": 0.5874,
      "step": 2070
    },
    {
      "epoch": 0.610619793769036,
      "grad_norm": 4.684408187866211,
      "learning_rate": 6.6013323053327065e-06,
      "loss": 0.6178,
      "step": 2080
    },
    {
      "epoch": 0.6135554658544641,
      "grad_norm": 4.132544040679932,
      "learning_rate": 6.5147290948626365e-06,
      "loss": 0.5972,
      "step": 2090
    },
    {
      "epoch": 0.6164911379398921,
      "grad_norm": 3.2844135761260986,
      "learning_rate": 6.428422414928066e-06,
      "loss": 0.5808,
      "step": 2100
    },
    {
      "epoch": 0.6194268100253202,
      "grad_norm": 4.8152289390563965,
      "learning_rate": 6.342419608592626e-06,
      "loss": 0.6407,
      "step": 2110
    },
    {
      "epoch": 0.6223624821107482,
      "grad_norm": 4.975841999053955,
      "learning_rate": 6.25672799306605e-06,
      "loss": 0.5792,
      "step": 2120
    },
    {
      "epoch": 0.6252981541961763,
      "grad_norm": 3.772268772125244,
      "learning_rate": 6.171354859081639e-06,
      "loss": 0.7062,
      "step": 2130
    },
    {
      "epoch": 0.6282338262816043,
      "grad_norm": 3.6091275215148926,
      "learning_rate": 6.086307470275947e-06,
      "loss": 0.6015,
      "step": 2140
    },
    {
      "epoch": 0.6311694983670324,
      "grad_norm": 3.9650683403015137,
      "learning_rate": 6.001593062570776e-06,
      "loss": 0.699,
      "step": 2150
    },
    {
      "epoch": 0.6341051704524605,
      "grad_norm": 3.4142041206359863,
      "learning_rate": 5.917218843557551e-06,
      "loss": 0.5912,
      "step": 2160
    },
    {
      "epoch": 0.6370408425378885,
      "grad_norm": 3.6262595653533936,
      "learning_rate": 5.8415788415375744e-06,
      "loss": 0.6029,
      "step": 2170
    },
    {
      "epoch": 0.6399765146233166,
      "grad_norm": 3.612025737762451,
      "learning_rate": 5.757870733799642e-06,
      "loss": 0.6054,
      "step": 2180
    },
    {
      "epoch": 0.6429121867087446,
      "grad_norm": 3.7721731662750244,
      "learning_rate": 5.6745235509072135e-06,
      "loss": 0.5703,
      "step": 2190
    },
    {
      "epoch": 0.6458478587941727,
      "grad_norm": 4.44386100769043,
      "learning_rate": 5.591544384126769e-06,
      "loss": 0.6101,
      "step": 2200
    },
    {
      "epoch": 0.6487835308796007,
      "grad_norm": 3.6553893089294434,
      "learning_rate": 5.508940293413603e-06,
      "loss": 0.6131,
      "step": 2210
    },
    {
      "epoch": 0.6517192029650288,
      "grad_norm": 4.550465106964111,
      "learning_rate": 5.426718306811134e-06,
      "loss": 0.5761,
      "step": 2220
    },
    {
      "epoch": 0.6546548750504568,
      "grad_norm": 3.433598279953003,
      "learning_rate": 5.344885419852961e-06,
      "loss": 0.6456,
      "step": 2230
    },
    {
      "epoch": 0.6575905471358849,
      "grad_norm": 5.087676048278809,
      "learning_rate": 5.263448594967673e-06,
      "loss": 0.657,
      "step": 2240
    },
    {
      "epoch": 0.6605262192213129,
      "grad_norm": 4.578396797180176,
      "learning_rate": 5.182414760886484e-06,
      "loss": 0.6083,
      "step": 2250
    },
    {
      "epoch": 0.663461891306741,
      "grad_norm": 6.043960094451904,
      "learning_rate": 5.1017908120537105e-06,
      "loss": 0.5721,
      "step": 2260
    },
    {
      "epoch": 0.666397563392169,
      "grad_norm": 3.624394178390503,
      "learning_rate": 5.021583608040208e-06,
      "loss": 0.5952,
      "step": 2270
    },
    {
      "epoch": 0.6693332354775972,
      "grad_norm": 2.965820074081421,
      "learning_rate": 4.941799972959752e-06,
      "loss": 0.5074,
      "step": 2280
    },
    {
      "epoch": 0.6722689075630253,
      "grad_norm": 5.590756416320801,
      "learning_rate": 4.862446694888403e-06,
      "loss": 0.5274,
      "step": 2290
    },
    {
      "epoch": 0.6752045796484533,
      "grad_norm": 4.188043594360352,
      "learning_rate": 4.783530525287006e-06,
      "loss": 0.5694,
      "step": 2300
    },
    {
      "epoch": 0.6781402517338814,
      "grad_norm": 3.925184488296509,
      "learning_rate": 4.705058178426753e-06,
      "loss": 0.55,
      "step": 2310
    },
    {
      "epoch": 0.6810759238193094,
      "grad_norm": 4.226954936981201,
      "learning_rate": 4.627036330817926e-06,
      "loss": 0.5432,
      "step": 2320
    },
    {
      "epoch": 0.6840115959047375,
      "grad_norm": 3.9109609127044678,
      "learning_rate": 4.5494716206418555e-06,
      "loss": 0.5332,
      "step": 2330
    },
    {
      "epoch": 0.6869472679901655,
      "grad_norm": 4.698592662811279,
      "learning_rate": 4.4723706471861385e-06,
      "loss": 0.5744,
      "step": 2340
    },
    {
      "epoch": 0.6898829400755936,
      "grad_norm": 4.461889266967773,
      "learning_rate": 4.3957399702831505e-06,
      "loss": 0.5314,
      "step": 2350
    },
    {
      "epoch": 0.6928186121610216,
      "grad_norm": 4.412221908569336,
      "learning_rate": 4.31958610975195e-06,
      "loss": 0.554,
      "step": 2360
    },
    {
      "epoch": 0.6957542842464497,
      "grad_norm": 3.776421308517456,
      "learning_rate": 4.243915544843549e-06,
      "loss": 0.4857,
      "step": 2370
    },
    {
      "epoch": 0.6986899563318777,
      "grad_norm": 4.851159572601318,
      "learning_rate": 4.168734713689658e-06,
      "loss": 0.5484,
      "step": 2380
    },
    {
      "epoch": 0.7016256284173058,
      "grad_norm": 3.8917558193206787,
      "learning_rate": 4.094050012754925e-06,
      "loss": 0.4888,
      "step": 2390
    },
    {
      "epoch": 0.7045613005027338,
      "grad_norm": 4.396358966827393,
      "learning_rate": 4.019867796292709e-06,
      "loss": 0.5125,
      "step": 2400
    },
    {
      "epoch": 0.7074969725881619,
      "grad_norm": 4.374291896820068,
      "learning_rate": 3.946194375804452e-06,
      "loss": 0.5262,
      "step": 2410
    },
    {
      "epoch": 0.7104326446735899,
      "grad_norm": 4.330350875854492,
      "learning_rate": 3.873036019502716e-06,
      "loss": 0.5581,
      "step": 2420
    },
    {
      "epoch": 0.713368316759018,
      "grad_norm": 4.86287784576416,
      "learning_rate": 3.800398951777845e-06,
      "loss": 0.5687,
      "step": 2430
    },
    {
      "epoch": 0.7163039888444461,
      "grad_norm": 5.004453659057617,
      "learning_rate": 3.7282893526683914e-06,
      "loss": 0.5136,
      "step": 2440
    },
    {
      "epoch": 0.7192396609298741,
      "grad_norm": 3.035261631011963,
      "learning_rate": 3.656713357335334e-06,
      "loss": 0.5358,
      "step": 2450
    },
    {
      "epoch": 0.7221753330153022,
      "grad_norm": 3.3477425575256348,
      "learning_rate": 3.585677055540072e-06,
      "loss": 0.5214,
      "step": 2460
    },
    {
      "epoch": 0.7251110051007302,
      "grad_norm": 3.495814323425293,
      "learning_rate": 3.5151864911263066e-06,
      "loss": 0.5048,
      "step": 2470
    },
    {
      "epoch": 0.7280466771861583,
      "grad_norm": 3.35532283782959,
      "learning_rate": 3.4452476615058316e-06,
      "loss": 0.509,
      "step": 2480
    },
    {
      "epoch": 0.7309823492715863,
      "grad_norm": 3.0357613563537598,
      "learning_rate": 3.3758665171482474e-06,
      "loss": 0.5361,
      "step": 2490
    },
    {
      "epoch": 0.7339180213570144,
      "grad_norm": 4.02761173248291,
      "learning_rate": 3.3070489610747146e-06,
      "loss": 0.5033,
      "step": 2500
    },
    {
      "epoch": 0.7368536934424424,
      "grad_norm": 4.085331916809082,
      "learning_rate": 3.238800848355702e-06,
      "loss": 0.526,
      "step": 2510
    },
    {
      "epoch": 0.7397893655278706,
      "grad_norm": 4.161253929138184,
      "learning_rate": 3.1711279856128387e-06,
      "loss": 0.5014,
      "step": 2520
    },
    {
      "epoch": 0.7427250376132986,
      "grad_norm": 3.7220897674560547,
      "learning_rate": 3.10403613052487e-06,
      "loss": 0.4514,
      "step": 2530
    },
    {
      "epoch": 0.7456607096987267,
      "grad_norm": 4.337230682373047,
      "learning_rate": 3.037530991337807e-06,
      "loss": 0.5645,
      "step": 2540
    },
    {
      "epoch": 0.7485963817841547,
      "grad_norm": 4.30481481552124,
      "learning_rate": 2.9716182263792314e-06,
      "loss": 0.5026,
      "step": 2550
    },
    {
      "epoch": 0.7515320538695828,
      "grad_norm": 3.3447349071502686,
      "learning_rate": 2.9063034435769242e-06,
      "loss": 0.5318,
      "step": 2560
    },
    {
      "epoch": 0.7544677259550108,
      "grad_norm": 3.936032295227051,
      "learning_rate": 2.8415921999816966e-06,
      "loss": 0.5106,
      "step": 2570
    },
    {
      "epoch": 0.7574033980404389,
      "grad_norm": 3.9542150497436523,
      "learning_rate": 2.7774900012946037e-06,
      "loss": 0.5299,
      "step": 2580
    },
    {
      "epoch": 0.760339070125867,
      "grad_norm": 4.351448059082031,
      "learning_rate": 2.714002301398524e-06,
      "loss": 0.5211,
      "step": 2590
    },
    {
      "epoch": 0.763274742211295,
      "grad_norm": 4.167295932769775,
      "learning_rate": 2.6511345018941225e-06,
      "loss": 0.5071,
      "step": 2600
    },
    {
      "epoch": 0.7662104142967231,
      "grad_norm": 5.125722408294678,
      "learning_rate": 2.588891951640288e-06,
      "loss": 0.5199,
      "step": 2610
    },
    {
      "epoch": 0.7691460863821511,
      "grad_norm": 4.25960111618042,
      "learning_rate": 2.527279946299037e-06,
      "loss": 0.4537,
      "step": 2620
    },
    {
      "epoch": 0.7720817584675792,
      "grad_norm": 3.875459909439087,
      "learning_rate": 2.4663037278849665e-06,
      "loss": 0.4993,
      "step": 2630
    },
    {
      "epoch": 0.7750174305530072,
      "grad_norm": 4.285188674926758,
      "learning_rate": 2.405968484319231e-06,
      "loss": 0.4596,
      "step": 2640
    },
    {
      "epoch": 0.7779531026384353,
      "grad_norm": 3.827913284301758,
      "learning_rate": 2.3462793489881884e-06,
      "loss": 0.5141,
      "step": 2650
    },
    {
      "epoch": 0.7808887747238633,
      "grad_norm": 4.174901485443115,
      "learning_rate": 2.2872414003066146e-06,
      "loss": 0.4483,
      "step": 2660
    },
    {
      "epoch": 0.7838244468092914,
      "grad_norm": 3.4712812900543213,
      "learning_rate": 2.2288596612856306e-06,
      "loss": 0.4834,
      "step": 2670
    },
    {
      "epoch": 0.7867601188947194,
      "grad_norm": 4.7577972412109375,
      "learning_rate": 2.1711390991053547e-06,
      "loss": 0.4913,
      "step": 2680
    },
    {
      "epoch": 0.7896957909801475,
      "grad_norm": 3.8983209133148193,
      "learning_rate": 2.1140846246922774e-06,
      "loss": 0.4748,
      "step": 2690
    },
    {
      "epoch": 0.7926314630655755,
      "grad_norm": 3.3365228176116943,
      "learning_rate": 2.0577010923014353e-06,
      "loss": 0.5014,
      "step": 2700
    },
    {
      "epoch": 0.7955671351510036,
      "grad_norm": 3.7394635677337646,
      "learning_rate": 2.001993299103411e-06,
      "loss": 0.4524,
      "step": 2710
    },
    {
      "epoch": 0.7985028072364317,
      "grad_norm": 3.325190544128418,
      "learning_rate": 1.946965984776181e-06,
      "loss": 0.486,
      "step": 2720
    },
    {
      "epoch": 0.8014384793218597,
      "grad_norm": 3.337636947631836,
      "learning_rate": 1.8926238311018551e-06,
      "loss": 0.4112,
      "step": 2730
    },
    {
      "epoch": 0.8043741514072879,
      "grad_norm": 4.279343128204346,
      "learning_rate": 1.8443055276768218e-06,
      "loss": 0.4758,
      "step": 2740
    },
    {
      "epoch": 0.8073098234927159,
      "grad_norm": 3.3319621086120605,
      "learning_rate": 1.7912778684550137e-06,
      "loss": 0.4209,
      "step": 2750
    },
    {
      "epoch": 0.810245495578144,
      "grad_norm": 4.332451343536377,
      "learning_rate": 1.7389486159957436e-06,
      "loss": 0.4913,
      "step": 2760
    },
    {
      "epoch": 0.813181167663572,
      "grad_norm": 5.22000789642334,
      "learning_rate": 1.6873222225271656e-06,
      "loss": 0.4523,
      "step": 2770
    },
    {
      "epoch": 0.8161168397490001,
      "grad_norm": 4.080671787261963,
      "learning_rate": 1.63640308047745e-06,
      "loss": 0.4545,
      "step": 2780
    },
    {
      "epoch": 0.8190525118344281,
      "grad_norm": 5.591613292694092,
      "learning_rate": 1.5861955221010671e-06,
      "loss": 0.5272,
      "step": 2790
    },
    {
      "epoch": 0.8219881839198562,
      "grad_norm": 3.4515106678009033,
      "learning_rate": 1.536703819110198e-06,
      "loss": 0.4166,
      "step": 2800
    },
    {
      "epoch": 0.8249238560052842,
      "grad_norm": 3.1613569259643555,
      "learning_rate": 1.4879321823112802e-06,
      "loss": 0.5194,
      "step": 2810
    },
    {
      "epoch": 0.8278595280907123,
      "grad_norm": 4.931222915649414,
      "learning_rate": 1.4398847612467736e-06,
      "loss": 0.4626,
      "step": 2820
    },
    {
      "epoch": 0.8307952001761403,
      "grad_norm": 3.088315963745117,
      "learning_rate": 1.3925656438420876e-06,
      "loss": 0.4246,
      "step": 2830
    },
    {
      "epoch": 0.8337308722615684,
      "grad_norm": 3.7036452293395996,
      "learning_rate": 1.3459788560577847e-06,
      "loss": 0.431,
      "step": 2840
    },
    {
      "epoch": 0.8366665443469964,
      "grad_norm": 4.452617168426514,
      "learning_rate": 1.3001283615470517e-06,
      "loss": 0.4478,
      "step": 2850
    },
    {
      "epoch": 0.8396022164324245,
      "grad_norm": 3.2161977291107178,
      "learning_rate": 1.255018061318467e-06,
      "loss": 0.4432,
      "step": 2860
    },
    {
      "epoch": 0.8425378885178526,
      "grad_norm": 4.302596092224121,
      "learning_rate": 1.2106517934040917e-06,
      "loss": 0.4598,
      "step": 2870
    },
    {
      "epoch": 0.8454735606032806,
      "grad_norm": 4.297342300415039,
      "learning_rate": 1.1670333325329353e-06,
      "loss": 0.4908,
      "step": 2880
    },
    {
      "epoch": 0.8484092326887087,
      "grad_norm": 3.9199209213256836,
      "learning_rate": 1.1241663898097865e-06,
      "loss": 0.4239,
      "step": 2890
    },
    {
      "epoch": 0.8513449047741367,
      "grad_norm": 4.693470001220703,
      "learning_rate": 1.08205461239948e-06,
      "loss": 0.4636,
      "step": 2900
    },
    {
      "epoch": 0.8542805768595648,
      "grad_norm": 4.2040510177612305,
      "learning_rate": 1.04070158321659e-06,
      "loss": 0.4595,
      "step": 2910
    },
    {
      "epoch": 0.8572162489449928,
      "grad_norm": 4.8676252365112305,
      "learning_rate": 1.00011082062058e-06,
      "loss": 0.4699,
      "step": 2920
    },
    {
      "epoch": 0.8601519210304209,
      "grad_norm": 3.176576852798462,
      "learning_rate": 9.602857781164721e-07,
      "loss": 0.4599,
      "step": 2930
    },
    {
      "epoch": 0.8630875931158489,
      "grad_norm": 4.111423969268799,
      "learning_rate": 9.212298440610101e-07,
      "loss": 0.4601,
      "step": 2940
    },
    {
      "epoch": 0.866023265201277,
      "grad_norm": 3.106792449951172,
      "learning_rate": 8.829463413743811e-07,
      "loss": 0.453,
      "step": 2950
    },
    {
      "epoch": 0.868958937286705,
      "grad_norm": 4.940300941467285,
      "learning_rate": 8.454385272574906e-07,
      "loss": 0.4298,
      "step": 2960
    },
    {
      "epoch": 0.8718946093721331,
      "grad_norm": 4.7473249435424805,
      "learning_rate": 8.087095929148436e-07,
      "loss": 0.457,
      "step": 2970
    },
    {
      "epoch": 0.8748302814575611,
      "grad_norm": 4.263439655303955,
      "learning_rate": 7.727626632830221e-07,
      "loss": 0.4194,
      "step": 2980
    },
    {
      "epoch": 0.8777659535429893,
      "grad_norm": 4.69775390625,
      "learning_rate": 7.376007967648302e-07,
      "loss": 0.4457,
      "step": 2990
    },
    {
      "epoch": 0.8807016256284173,
      "grad_norm": 4.177097320556641,
      "learning_rate": 7.032269849690654e-07,
      "loss": 0.4532,
      "step": 3000
    },
    {
      "epoch": 0.8836372977138454,
      "grad_norm": 4.312076568603516,
      "learning_rate": 6.696441524559983e-07,
      "loss": 0.4772,
      "step": 3010
    },
    {
      "epoch": 0.8865729697992735,
      "grad_norm": 4.328220367431641,
      "learning_rate": 6.368551564885439e-07,
      "loss": 0.4239,
      "step": 3020
    },
    {
      "epoch": 0.8895086418847015,
      "grad_norm": 3.3847310543060303,
      "learning_rate": 6.048627867891665e-07,
      "loss": 0.4564,
      "step": 3030
    },
    {
      "epoch": 0.8924443139701296,
      "grad_norm": 3.408613681793213,
      "learning_rate": 5.736697653025192e-07,
      "loss": 0.4206,
      "step": 3040
    },
    {
      "epoch": 0.8953799860555576,
      "grad_norm": 4.045165061950684,
      "learning_rate": 5.432787459638722e-07,
      "loss": 0.4751,
      "step": 3050
    },
    {
      "epoch": 0.8983156581409857,
      "grad_norm": 3.965830087661743,
      "learning_rate": 5.136923144732997e-07,
      "loss": 0.4273,
      "step": 3060
    },
    {
      "epoch": 0.9012513302264137,
      "grad_norm": 3.462986707687378,
      "learning_rate": 4.849129880756886e-07,
      "loss": 0.472,
      "step": 3070
    },
    {
      "epoch": 0.9041870023118418,
      "grad_norm": 5.114270210266113,
      "learning_rate": 4.569432153465736e-07,
      "loss": 0.5233,
      "step": 3080
    },
    {
      "epoch": 0.9071226743972698,
      "grad_norm": 4.655681610107422,
      "learning_rate": 4.297853759838055e-07,
      "loss": 0.4543,
      "step": 3090
    },
    {
      "epoch": 0.9100583464826979,
      "grad_norm": 4.586308002471924,
      "learning_rate": 4.034417806050872e-07,
      "loss": 0.4383,
      "step": 3100
    },
    {
      "epoch": 0.9129940185681259,
      "grad_norm": 3.405298948287964,
      "learning_rate": 3.779146705513814e-07,
      "loss": 0.4644,
      "step": 3110
    },
    {
      "epoch": 0.915929690653554,
      "grad_norm": 3.8995399475097656,
      "learning_rate": 3.532062176962159e-07,
      "loss": 0.4591,
      "step": 3120
    },
    {
      "epoch": 0.918865362738982,
      "grad_norm": 3.6494014263153076,
      "learning_rate": 3.293185242608954e-07,
      "loss": 0.4354,
      "step": 3130
    },
    {
      "epoch": 0.9218010348244101,
      "grad_norm": 4.192446708679199,
      "learning_rate": 3.062536226356472e-07,
      "loss": 0.4466,
      "step": 3140
    },
    {
      "epoch": 0.9247367069098382,
      "grad_norm": 3.4682350158691406,
      "learning_rate": 2.8401347520670253e-07,
      "loss": 0.4629,
      "step": 3150
    },
    {
      "epoch": 0.9276723789952662,
      "grad_norm": 3.987903594970703,
      "learning_rate": 2.625999741893304e-07,
      "loss": 0.5156,
      "step": 3160
    },
    {
      "epoch": 0.9306080510806943,
      "grad_norm": 4.3802103996276855,
      "learning_rate": 2.420149414668493e-07,
      "loss": 0.42,
      "step": 3170
    },
    {
      "epoch": 0.9335437231661223,
      "grad_norm": 4.355963230133057,
      "learning_rate": 2.222601284356185e-07,
      "loss": 0.4408,
      "step": 3180
    },
    {
      "epoch": 0.9364793952515504,
      "grad_norm": 5.095834255218506,
      "learning_rate": 2.0333721585602984e-07,
      "loss": 0.4558,
      "step": 3190
    },
    {
      "epoch": 0.9394150673369784,
      "grad_norm": 3.3932993412017822,
      "learning_rate": 1.8524781370950262e-07,
      "loss": 0.4475,
      "step": 3200
    },
    {
      "epoch": 0.9423507394224065,
      "grad_norm": 5.2142014503479,
      "learning_rate": 1.679934610615064e-07,
      "loss": 0.4351,
      "step": 3210
    },
    {
      "epoch": 0.9452864115078345,
      "grad_norm": 4.271505832672119,
      "learning_rate": 1.515756259306178e-07,
      "loss": 0.4431,
      "step": 3220
    },
    {
      "epoch": 0.9482220835932627,
      "grad_norm": 4.888089656829834,
      "learning_rate": 1.3599570516361737e-07,
      "loss": 0.4256,
      "step": 3230
    },
    {
      "epoch": 0.9511577556786907,
      "grad_norm": 4.216527938842773,
      "learning_rate": 1.212550243166455e-07,
      "loss": 0.4811,
      "step": 3240
    },
    {
      "epoch": 0.9540934277641188,
      "grad_norm": 4.079187393188477,
      "learning_rate": 1.0735483754242049e-07,
      "loss": 0.4435,
      "step": 3250
    },
    {
      "epoch": 0.9570290998495468,
      "grad_norm": 4.337707042694092,
      "learning_rate": 9.429632748354068e-08,
      "loss": 0.4152,
      "step": 3260
    },
    {
      "epoch": 0.9599647719349749,
      "grad_norm": 3.4174439907073975,
      "learning_rate": 8.208060517185146e-08,
      "loss": 0.4579,
      "step": 3270
    },
    {
      "epoch": 0.9629004440204029,
      "grad_norm": 4.021118640899658,
      "learning_rate": 7.070870993393209e-08,
      "loss": 0.4531,
      "step": 3280
    },
    {
      "epoch": 0.965836116105831,
      "grad_norm": 3.2883527278900146,
      "learning_rate": 6.01816093026586e-08,
      "loss": 0.4384,
      "step": 3290
    },
    {
      "epoch": 0.9687717881912591,
      "grad_norm": 3.5450029373168945,
      "learning_rate": 5.0500198934889665e-08,
      "loss": 0.4028,
      "step": 3300
    },
    {
      "epoch": 0.9717074602766871,
      "grad_norm": 3.9169414043426514,
      "learning_rate": 4.16653025352598e-08,
      "loss": 0.4894,
      "step": 3310
    },
    {
      "epoch": 0.9746431323621152,
      "grad_norm": 4.527153015136719,
      "learning_rate": 3.367767178609982e-08,
      "loss": 0.4403,
      "step": 3320
    },
    {
      "epoch": 0.9775788044475432,
      "grad_norm": 4.728188514709473,
      "learning_rate": 2.6537986283485805e-08,
      "loss": 0.4123,
      "step": 3330
    },
    {
      "epoch": 0.9805144765329713,
      "grad_norm": 4.7194037437438965,
      "learning_rate": 2.024685347941202e-08,
      "loss": 0.4456,
      "step": 3340
    },
    {
      "epoch": 0.9834501486183993,
      "grad_norm": 4.34934139251709,
      "learning_rate": 1.4804808630112244e-08,
      "loss": 0.4383,
      "step": 3350
    },
    {
      "epoch": 0.9863858207038274,
      "grad_norm": 4.218658924102783,
      "learning_rate": 1.0212314750518426e-08,
      "loss": 0.4368,
      "step": 3360
    },
    {
      "epoch": 0.9893214927892554,
      "grad_norm": 4.213326454162598,
      "learning_rate": 6.469762574868866e-09,
      "loss": 0.4769,
      "step": 3370
    },
    {
      "epoch": 0.9922571648746835,
      "grad_norm": 3.9344699382781982,
      "learning_rate": 3.5774705234625783e-09,
      "loss": 0.4412,
      "step": 3380
    },
    {
      "epoch": 0.9951928369601115,
      "grad_norm": 3.622770071029663,
      "learning_rate": 1.5356846755654187e-09,
      "loss": 0.5059,
      "step": 3390
    },
    {
      "epoch": 0.9981285090455396,
      "grad_norm": 5.872491359710693,
      "learning_rate": 3.4457874847793063e-10,
      "loss": 0.4697,
      "step": 3400
    },
    {
      "epoch": 0.9998899122967965,
      "step": 3406,
      "total_flos": 3.2891568428128666e+18,
      "train_loss": 0.7668657067657847,
      "train_runtime": 24228.5265,
      "train_samples_per_second": 4.499,
      "train_steps_per_second": 0.141
    }
  ],
  "logging_steps": 10,
  "max_steps": 3406,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 100000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": false,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 3.2891568428128666e+18,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}