{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.002227791701476,
  "eval_steps": 500,
  "global_step": 1800,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00556947925368978,
      "grad_norm": 19.463690451045245,
      "learning_rate": 7.154213036565979e-07,
      "loss": 8.2534,
      "step": 10
    },
    {
      "epoch": 0.01113895850737956,
      "grad_norm": 19.73356911714459,
      "learning_rate": 1.51033386327504e-06,
      "loss": 8.1778,
      "step": 20
    },
    {
      "epoch": 0.01670843776106934,
      "grad_norm": 10.029986538109684,
      "learning_rate": 2.3052464228934817e-06,
      "loss": 7.6937,
      "step": 30
    },
    {
      "epoch": 0.02227791701475912,
      "grad_norm": 7.379827099032999,
      "learning_rate": 3.100158982511924e-06,
      "loss": 6.9205,
      "step": 40
    },
    {
      "epoch": 0.0278473962684489,
      "grad_norm": 3.5728211270030625,
      "learning_rate": 3.895071542130366e-06,
      "loss": 6.3948,
      "step": 50
    },
    {
      "epoch": 0.03341687552213868,
      "grad_norm": 2.8392085477918747,
      "learning_rate": 4.689984101748808e-06,
      "loss": 6.0575,
      "step": 60
    },
    {
      "epoch": 0.03898635477582846,
      "grad_norm": 3.3563782303356495,
      "learning_rate": 5.48489666136725e-06,
      "loss": 5.7294,
      "step": 70
    },
    {
      "epoch": 0.04455583402951824,
      "grad_norm": 3.826365209612832,
      "learning_rate": 6.279809220985691e-06,
      "loss": 5.2614,
      "step": 80
    },
    {
      "epoch": 0.05012531328320802,
      "grad_norm": 3.965749199888462,
      "learning_rate": 7.0747217806041336e-06,
      "loss": 4.658,
      "step": 90
    },
    {
      "epoch": 0.0556947925368978,
      "grad_norm": 5.147891128586454,
      "learning_rate": 7.869634340222575e-06,
      "loss": 4.0477,
      "step": 100
    },
    {
      "epoch": 0.06126427179058758,
      "grad_norm": 3.695764472570234,
      "learning_rate": 8.664546899841019e-06,
      "loss": 3.5584,
      "step": 110
    },
    {
      "epoch": 0.06683375104427736,
      "grad_norm": 4.026692299224468,
      "learning_rate": 9.45945945945946e-06,
      "loss": 3.2227,
      "step": 120
    },
    {
      "epoch": 0.07240323029796714,
      "grad_norm": 5.389627945808325,
      "learning_rate": 1.02543720190779e-05,
      "loss": 3.0255,
      "step": 130
    },
    {
      "epoch": 0.07797270955165692,
      "grad_norm": 5.554941152889833,
      "learning_rate": 1.1049284578696344e-05,
      "loss": 2.8932,
      "step": 140
    },
    {
      "epoch": 0.0835421888053467,
      "grad_norm": 4.8986035107973995,
      "learning_rate": 1.1844197138314786e-05,
      "loss": 2.7959,
      "step": 150
    },
    {
      "epoch": 0.08911166805903648,
      "grad_norm": 4.56461997585786,
      "learning_rate": 1.2639109697933228e-05,
      "loss": 2.7067,
      "step": 160
    },
    {
      "epoch": 0.09468114731272625,
      "grad_norm": 9.192995633454027,
      "learning_rate": 1.3434022257551671e-05,
      "loss": 2.6194,
      "step": 170
    },
    {
      "epoch": 0.10025062656641603,
      "grad_norm": 8.520097501414543,
      "learning_rate": 1.4228934817170111e-05,
      "loss": 2.5516,
      "step": 180
    },
    {
      "epoch": 0.10582010582010581,
      "grad_norm": 14.274847317292076,
      "learning_rate": 1.5023847376788555e-05,
      "loss": 2.4789,
      "step": 190
    },
    {
      "epoch": 0.1113895850737956,
      "grad_norm": 19.214437222876512,
      "learning_rate": 1.5818759936406995e-05,
      "loss": 2.4363,
      "step": 200
    },
    {
      "epoch": 0.11695906432748537,
      "grad_norm": 16.394738746567327,
      "learning_rate": 1.6613672496025438e-05,
      "loss": 2.3902,
      "step": 210
    },
    {
      "epoch": 0.12252854358117515,
      "grad_norm": 18.247117883573495,
      "learning_rate": 1.740858505564388e-05,
      "loss": 2.3452,
      "step": 220
    },
    {
      "epoch": 0.12809802283486493,
      "grad_norm": 15.142493870476338,
      "learning_rate": 1.820349761526232e-05,
      "loss": 2.3042,
      "step": 230
    },
    {
      "epoch": 0.1336675020885547,
      "grad_norm": 8.204095855953893,
      "learning_rate": 1.8998410174880765e-05,
      "loss": 2.2554,
      "step": 240
    },
    {
      "epoch": 0.1392369813422445,
      "grad_norm": 12.061502490559853,
      "learning_rate": 1.9793322734499205e-05,
      "loss": 2.2062,
      "step": 250
    },
    {
      "epoch": 0.14480646059593427,
      "grad_norm": 13.078924954972894,
      "learning_rate": 2.058823529411765e-05,
      "loss": 2.1414,
      "step": 260
    },
    {
      "epoch": 0.15037593984962405,
      "grad_norm": 13.325114907384508,
      "learning_rate": 2.1383147853736092e-05,
      "loss": 2.059,
      "step": 270
    },
    {
      "epoch": 0.15594541910331383,
      "grad_norm": 7.4709725407687255,
      "learning_rate": 2.2178060413354532e-05,
      "loss": 1.9772,
      "step": 280
    },
    {
      "epoch": 0.1615148983570036,
      "grad_norm": 19.25846753081156,
      "learning_rate": 2.2972972972972976e-05,
      "loss": 1.9011,
      "step": 290
    },
    {
      "epoch": 0.1670843776106934,
      "grad_norm": 13.01475345365111,
      "learning_rate": 2.3767885532591416e-05,
      "loss": 1.8118,
      "step": 300
    },
    {
      "epoch": 0.17265385686438317,
      "grad_norm": 6.508219489647502,
      "learning_rate": 2.4562798092209856e-05,
      "loss": 1.7365,
      "step": 310
    },
    {
      "epoch": 0.17822333611807295,
      "grad_norm": 8.638960149014848,
      "learning_rate": 2.53577106518283e-05,
      "loss": 1.6377,
      "step": 320
    },
    {
      "epoch": 0.18379281537176273,
      "grad_norm": 5.254562295866414,
      "learning_rate": 2.6152623211446743e-05,
      "loss": 1.5737,
      "step": 330
    },
    {
      "epoch": 0.1893622946254525,
      "grad_norm": 8.547148285512923,
      "learning_rate": 2.6947535771065186e-05,
      "loss": 1.5303,
      "step": 340
    },
    {
      "epoch": 0.1949317738791423,
      "grad_norm": 6.278158376274084,
      "learning_rate": 2.7742448330683623e-05,
      "loss": 1.4716,
      "step": 350
    },
    {
      "epoch": 0.20050125313283207,
      "grad_norm": 8.26795777275924,
      "learning_rate": 2.8537360890302066e-05,
      "loss": 1.4151,
      "step": 360
    },
    {
      "epoch": 0.20607073238652185,
      "grad_norm": 4.425290147294405,
      "learning_rate": 2.933227344992051e-05,
      "loss": 1.368,
      "step": 370
    },
    {
      "epoch": 0.21164021164021163,
      "grad_norm": 6.848350616045537,
      "learning_rate": 3.0127186009538953e-05,
      "loss": 1.3383,
      "step": 380
    },
    {
      "epoch": 0.2172096908939014,
      "grad_norm": 4.869312013250087,
      "learning_rate": 3.092209856915739e-05,
      "loss": 1.2751,
      "step": 390
    },
    {
      "epoch": 0.2227791701475912,
      "grad_norm": 7.264173089824176,
      "learning_rate": 3.171701112877583e-05,
      "loss": 1.2448,
      "step": 400
    },
    {
      "epoch": 0.22834864940128097,
      "grad_norm": 7.343411468837996,
      "learning_rate": 3.251192368839428e-05,
      "loss": 1.1956,
      "step": 410
    },
    {
      "epoch": 0.23391812865497075,
      "grad_norm": 8.139260855658039,
      "learning_rate": 3.330683624801272e-05,
      "loss": 1.1626,
      "step": 420
    },
    {
      "epoch": 0.23948760790866053,
      "grad_norm": 4.441795212879553,
      "learning_rate": 3.410174880763116e-05,
      "loss": 1.1313,
      "step": 430
    },
    {
      "epoch": 0.2450570871623503,
      "grad_norm": 7.0655482262903115,
      "learning_rate": 3.489666136724961e-05,
      "loss": 1.1016,
      "step": 440
    },
    {
      "epoch": 0.2506265664160401,
      "grad_norm": 5.343797555692575,
      "learning_rate": 3.569157392686804e-05,
      "loss": 1.0682,
      "step": 450
    },
    {
      "epoch": 0.25619604566972987,
      "grad_norm": 4.837151670486665,
      "learning_rate": 3.648648648648649e-05,
      "loss": 1.0388,
      "step": 460
    },
    {
      "epoch": 0.26176552492341965,
      "grad_norm": 3.9144650767846345,
      "learning_rate": 3.7281399046104934e-05,
      "loss": 1.0,
      "step": 470
    },
    {
      "epoch": 0.2673350041771094,
      "grad_norm": 6.192111002424375,
      "learning_rate": 3.8076311605723374e-05,
      "loss": 0.9819,
      "step": 480
    },
    {
      "epoch": 0.2729044834307992,
      "grad_norm": 4.664894996941613,
      "learning_rate": 3.8871224165341814e-05,
      "loss": 0.9512,
      "step": 490
    },
    {
      "epoch": 0.278473962684489,
      "grad_norm": 5.728600939316023,
      "learning_rate": 3.9666136724960254e-05,
      "loss": 0.9292,
      "step": 500
    },
    {
      "epoch": 0.28404344193817876,
      "grad_norm": 7.358090778625861,
      "learning_rate": 4.04610492845787e-05,
      "loss": 0.9018,
      "step": 510
    },
    {
      "epoch": 0.28961292119186854,
      "grad_norm": 6.346761431472806,
      "learning_rate": 4.125596184419714e-05,
      "loss": 0.8805,
      "step": 520
    },
    {
      "epoch": 0.2951824004455583,
      "grad_norm": 3.586718245895629,
      "learning_rate": 4.205087440381558e-05,
      "loss": 0.8568,
      "step": 530
    },
    {
      "epoch": 0.3007518796992481,
      "grad_norm": 5.430039185536721,
      "learning_rate": 4.284578696343403e-05,
      "loss": 0.8284,
      "step": 540
    },
    {
      "epoch": 0.3063213589529379,
      "grad_norm": 6.430553831376545,
      "learning_rate": 4.364069952305246e-05,
      "loss": 0.8128,
      "step": 550
    },
    {
      "epoch": 0.31189083820662766,
      "grad_norm": 5.48500374978629,
      "learning_rate": 4.443561208267091e-05,
      "loss": 0.7842,
      "step": 560
    },
    {
      "epoch": 0.31746031746031744,
      "grad_norm": 5.0124287229570745,
      "learning_rate": 4.523052464228935e-05,
      "loss": 0.7803,
      "step": 570
    },
    {
      "epoch": 0.3230297967140072,
      "grad_norm": 4.960089415525465,
      "learning_rate": 4.6025437201907795e-05,
      "loss": 0.7489,
      "step": 580
    },
    {
      "epoch": 0.328599275967697,
      "grad_norm": 4.429208732066051,
      "learning_rate": 4.6820349761526235e-05,
      "loss": 0.7243,
      "step": 590
    },
    {
      "epoch": 0.3341687552213868,
      "grad_norm": 5.1268739971720905,
      "learning_rate": 4.7615262321144675e-05,
      "loss": 0.7044,
      "step": 600
    },
    {
      "epoch": 0.33973823447507656,
      "grad_norm": 4.915480885708807,
      "learning_rate": 4.841017488076312e-05,
      "loss": 0.6939,
      "step": 610
    },
    {
      "epoch": 0.34530771372876634,
      "grad_norm": 5.102518466488775,
      "learning_rate": 4.9205087440381556e-05,
      "loss": 0.6767,
      "step": 620
    },
    {
      "epoch": 0.3508771929824561,
      "grad_norm": 4.364437305811626,
      "learning_rate": 5e-05,
      "loss": 0.6467,
      "step": 630
    },
    {
      "epoch": 0.3564466722361459,
      "grad_norm": 4.372236216369743,
      "learning_rate": 5.079491255961845e-05,
      "loss": 0.6234,
      "step": 640
    },
    {
      "epoch": 0.3620161514898357,
      "grad_norm": 4.439347899189148,
      "learning_rate": 5.158982511923689e-05,
      "loss": 0.5961,
      "step": 650
    },
    {
      "epoch": 0.36758563074352546,
      "grad_norm": 3.1594832305447316,
      "learning_rate": 5.2384737678855336e-05,
      "loss": 0.5786,
      "step": 660
    },
    {
      "epoch": 0.37315510999721524,
      "grad_norm": 4.473968649061542,
      "learning_rate": 5.317965023847376e-05,
      "loss": 0.5711,
      "step": 670
    },
    {
      "epoch": 0.378724589250905,
      "grad_norm": 3.4873086420651456,
      "learning_rate": 5.397456279809221e-05,
      "loss": 0.5503,
      "step": 680
    },
    {
      "epoch": 0.3842940685045948,
      "grad_norm": 3.831386918783317,
      "learning_rate": 5.4769475357710656e-05,
      "loss": 0.529,
      "step": 690
    },
    {
      "epoch": 0.3898635477582846,
      "grad_norm": 3.653243174902386,
      "learning_rate": 5.5564387917329096e-05,
      "loss": 0.5132,
      "step": 700
    },
    {
      "epoch": 0.39543302701197436,
      "grad_norm": 3.5521167023222047,
      "learning_rate": 5.635930047694754e-05,
      "loss": 0.4877,
      "step": 710
    },
    {
      "epoch": 0.40100250626566414,
      "grad_norm": 3.8712402877933934,
      "learning_rate": 5.715421303656598e-05,
      "loss": 0.4833,
      "step": 720
    },
    {
      "epoch": 0.4065719855193539,
      "grad_norm": 3.379905907827214,
      "learning_rate": 5.794912559618442e-05,
      "loss": 0.4615,
      "step": 730
    },
    {
      "epoch": 0.4121414647730437,
      "grad_norm": 2.5936051096516417,
      "learning_rate": 5.8744038155802863e-05,
      "loss": 0.4444,
      "step": 740
    },
    {
      "epoch": 0.4177109440267335,
      "grad_norm": 3.322838857318257,
      "learning_rate": 5.9538950715421304e-05,
      "loss": 0.4318,
      "step": 750
    },
    {
      "epoch": 0.42328042328042326,
      "grad_norm": 3.3843216521985253,
      "learning_rate": 6.033386327503975e-05,
      "loss": 0.4145,
      "step": 760
    },
    {
      "epoch": 0.42884990253411304,
      "grad_norm": 3.0846350440353434,
      "learning_rate": 6.11287758346582e-05,
      "loss": 0.3998,
      "step": 770
    },
    {
      "epoch": 0.4344193817878028,
      "grad_norm": 2.7361994149983566,
      "learning_rate": 6.192368839427664e-05,
      "loss": 0.3863,
      "step": 780
    },
    {
      "epoch": 0.4399888610414926,
      "grad_norm": 3.27515261932524,
      "learning_rate": 6.271860095389508e-05,
      "loss": 0.375,
      "step": 790
    },
    {
      "epoch": 0.4455583402951824,
      "grad_norm": 3.029805585678618,
      "learning_rate": 6.351351351351352e-05,
      "loss": 0.364,
      "step": 800
    },
    {
      "epoch": 0.45112781954887216,
      "grad_norm": 2.2590750215054287,
      "learning_rate": 6.430842607313196e-05,
      "loss": 0.3453,
      "step": 810
    },
    {
      "epoch": 0.45669729880256194,
      "grad_norm": 3.1086383281500805,
      "learning_rate": 6.51033386327504e-05,
      "loss": 0.3329,
      "step": 820
    },
    {
      "epoch": 0.4622667780562517,
      "grad_norm": 2.312028278866729,
      "learning_rate": 6.589825119236884e-05,
      "loss": 0.324,
      "step": 830
    },
    {
      "epoch": 0.4678362573099415,
      "grad_norm": 2.8653110069877226,
      "learning_rate": 6.669316375198729e-05,
      "loss": 0.3113,
      "step": 840
    },
    {
      "epoch": 0.4734057365636313,
      "grad_norm": 2.0346455435163224,
      "learning_rate": 6.748807631160573e-05,
      "loss": 0.2976,
      "step": 850
    },
    {
      "epoch": 0.47897521581732105,
      "grad_norm": 2.2187482142761854,
      "learning_rate": 6.828298887122417e-05,
      "loss": 0.2838,
      "step": 860
    },
    {
      "epoch": 0.48454469507101083,
      "grad_norm": 2.5760581999823926,
      "learning_rate": 6.907790143084261e-05,
      "loss": 0.2712,
      "step": 870
    },
    {
      "epoch": 0.4901141743247006,
      "grad_norm": 2.336167318842384,
      "learning_rate": 6.987281399046105e-05,
      "loss": 0.2659,
      "step": 880
    },
    {
      "epoch": 0.4956836535783904,
      "grad_norm": 1.942101679442901,
      "learning_rate": 7.066772655007949e-05,
      "loss": 0.2524,
      "step": 890
    },
    {
      "epoch": 0.5012531328320802,
      "grad_norm": 1.9676135441330864,
      "learning_rate": 7.146263910969793e-05,
      "loss": 0.242,
      "step": 900
    },
    {
      "epoch": 0.50682261208577,
      "grad_norm": 2.42452293659568,
      "learning_rate": 7.225755166931639e-05,
      "loss": 0.2303,
      "step": 910
    },
    {
      "epoch": 0.5123920913394597,
      "grad_norm": 2.0689938731425754,
      "learning_rate": 7.305246422893483e-05,
      "loss": 0.2204,
      "step": 920
    },
    {
      "epoch": 0.5179615705931495,
      "grad_norm": 2.171275305780617,
      "learning_rate": 7.384737678855325e-05,
      "loss": 0.2108,
      "step": 930
    },
    {
      "epoch": 0.5235310498468393,
      "grad_norm": 2.018182012710717,
      "learning_rate": 7.46422893481717e-05,
      "loss": 0.1988,
      "step": 940
    },
    {
      "epoch": 0.5291005291005291,
      "grad_norm": 1.8435919023282064,
      "learning_rate": 7.543720190779015e-05,
      "loss": 0.1951,
      "step": 950
    },
    {
      "epoch": 0.5346700083542189,
      "grad_norm": 1.6966065845617362,
      "learning_rate": 7.623211446740859e-05,
      "loss": 0.1868,
      "step": 960
    },
    {
      "epoch": 0.5402394876079086,
      "grad_norm": 2.107747858389665,
      "learning_rate": 7.702702702702703e-05,
      "loss": 0.1752,
      "step": 970
    },
    {
      "epoch": 0.5458089668615984,
      "grad_norm": 1.626232566427141,
      "learning_rate": 7.782193958664548e-05,
      "loss": 0.1726,
      "step": 980
    },
    {
      "epoch": 0.5513784461152882,
      "grad_norm": 1.710543294805638,
      "learning_rate": 7.861685214626392e-05,
      "loss": 0.1625,
      "step": 990
    },
    {
      "epoch": 0.556947925368978,
      "grad_norm": 1.568201961955694,
      "learning_rate": 7.941176470588235e-05,
      "loss": 0.1552,
      "step": 1000
    },
    {
      "epoch": 0.5625174046226677,
      "grad_norm": 1.4211321412129432,
      "learning_rate": 8.02066772655008e-05,
      "loss": 0.1485,
      "step": 1010
    },
    {
      "epoch": 0.5680868838763575,
      "grad_norm": 1.4970140750834295,
      "learning_rate": 8.100158982511924e-05,
      "loss": 0.1434,
      "step": 1020
    },
    {
      "epoch": 0.5736563631300473,
      "grad_norm": 1.853797381238809,
      "learning_rate": 8.179650238473768e-05,
      "loss": 0.1356,
      "step": 1030
    },
    {
      "epoch": 0.5792258423837371,
      "grad_norm": 1.475678872579458,
      "learning_rate": 8.259141494435613e-05,
      "loss": 0.1298,
      "step": 1040
    },
    {
      "epoch": 0.5847953216374269,
      "grad_norm": 1.690324091049369,
      "learning_rate": 8.338632750397457e-05,
      "loss": 0.1258,
      "step": 1050
    },
    {
      "epoch": 0.5903648008911166,
      "grad_norm": 1.908861076416452,
      "learning_rate": 8.418124006359301e-05,
      "loss": 0.1249,
      "step": 1060
    },
    {
      "epoch": 0.5959342801448064,
      "grad_norm": 1.236244782970751,
      "learning_rate": 8.497615262321145e-05,
      "loss": 0.1171,
      "step": 1070
    },
    {
      "epoch": 0.6015037593984962,
      "grad_norm": 1.2774164058752888,
      "learning_rate": 8.57710651828299e-05,
      "loss": 0.1114,
      "step": 1080
    },
    {
      "epoch": 0.607073238652186,
      "grad_norm": 0.8951848955749875,
      "learning_rate": 8.656597774244833e-05,
      "loss": 0.1059,
      "step": 1090
    },
    {
      "epoch": 0.6126427179058758,
      "grad_norm": 1.0572189677651873,
      "learning_rate": 8.736089030206677e-05,
      "loss": 0.1012,
      "step": 1100
    },
    {
      "epoch": 0.6182121971595655,
      "grad_norm": 0.9583280956588,
      "learning_rate": 8.815580286168523e-05,
      "loss": 0.0973,
      "step": 1110
    },
    {
      "epoch": 0.6237816764132553,
      "grad_norm": 0.966004358455493,
      "learning_rate": 8.895071542130367e-05,
      "loss": 0.098,
      "step": 1120
    },
    {
      "epoch": 0.6293511556669451,
      "grad_norm": 0.8285031685005042,
      "learning_rate": 8.97456279809221e-05,
      "loss": 0.0897,
      "step": 1130
    },
    {
      "epoch": 0.6349206349206349,
      "grad_norm": 1.3954391622958595,
      "learning_rate": 9.054054054054055e-05,
      "loss": 0.087,
      "step": 1140
    },
    {
      "epoch": 0.6404901141743247,
      "grad_norm": 0.6608729564295608,
      "learning_rate": 9.133545310015899e-05,
      "loss": 0.0825,
      "step": 1150
    },
    {
      "epoch": 0.6460595934280144,
      "grad_norm": 0.8111382270197721,
      "learning_rate": 9.213036565977743e-05,
      "loss": 0.0764,
      "step": 1160
    },
    {
      "epoch": 0.6516290726817042,
      "grad_norm": 0.6891511752919594,
      "learning_rate": 9.292527821939587e-05,
      "loss": 0.0745,
      "step": 1170
    },
    {
      "epoch": 0.657198551935394,
      "grad_norm": 0.8999745942051316,
      "learning_rate": 9.372019077901432e-05,
      "loss": 0.0718,
      "step": 1180
    },
    {
      "epoch": 0.6627680311890838,
      "grad_norm": 0.7165073695545044,
      "learning_rate": 9.451510333863276e-05,
      "loss": 0.0699,
      "step": 1190
    },
    {
      "epoch": 0.6683375104427736,
      "grad_norm": 0.5691546082373657,
      "learning_rate": 9.531001589825119e-05,
      "loss": 0.0663,
      "step": 1200
    },
    {
      "epoch": 0.6739069896964633,
      "grad_norm": 0.618154476760873,
      "learning_rate": 9.610492845786964e-05,
      "loss": 0.065,
      "step": 1210
    },
    {
      "epoch": 0.6794764689501531,
      "grad_norm": 0.5460710019961567,
      "learning_rate": 9.689984101748808e-05,
      "loss": 0.0622,
      "step": 1220
    },
    {
      "epoch": 0.6850459482038429,
      "grad_norm": 0.44149906211171397,
      "learning_rate": 9.769475357710652e-05,
      "loss": 0.0582,
      "step": 1230
    },
    {
      "epoch": 0.6906154274575327,
      "grad_norm": 0.5860441580160569,
      "learning_rate": 9.848966613672496e-05,
      "loss": 0.0567,
      "step": 1240
    },
    {
      "epoch": 0.6961849067112225,
      "grad_norm": 0.8453684625264357,
      "learning_rate": 9.928457869634342e-05,
      "loss": 0.0564,
      "step": 1250
    },
    {
      "epoch": 0.7017543859649122,
      "grad_norm": 2.1379268435693555,
      "learning_rate": 9.999999807244224e-05,
      "loss": 0.0564,
      "step": 1260
    },
    {
      "epoch": 0.707323865218602,
      "grad_norm": 3.3074045726639083,
      "learning_rate": 9.999976676568994e-05,
      "loss": 0.1234,
      "step": 1270
    },
    {
      "epoch": 0.7128933444722918,
      "grad_norm": 0.6698300987807628,
      "learning_rate": 9.99991499494276e-05,
      "loss": 0.0776,
      "step": 1280
    },
    {
      "epoch": 0.7184628237259816,
      "grad_norm": 0.40986381929677007,
      "learning_rate": 9.999814762841102e-05,
      "loss": 0.0563,
      "step": 1290
    },
    {
      "epoch": 0.7240323029796714,
      "grad_norm": 0.5823587617428915,
      "learning_rate": 9.999675981036832e-05,
      "loss": 0.05,
      "step": 1300
    },
    {
      "epoch": 0.7296017822333611,
      "grad_norm": 0.36554660320089016,
      "learning_rate": 9.999498650599987e-05,
      "loss": 0.0505,
      "step": 1310
    },
    {
      "epoch": 0.7351712614870509,
      "grad_norm": 0.3490628453059406,
      "learning_rate": 9.999282772897828e-05,
      "loss": 0.0474,
      "step": 1320
    },
    {
      "epoch": 0.7407407407407407,
      "grad_norm": 0.3963213598340753,
      "learning_rate": 9.99902834959482e-05,
      "loss": 0.0462,
      "step": 1330
    },
    {
      "epoch": 0.7463102199944305,
      "grad_norm": 0.5565743717844861,
      "learning_rate": 9.998735382652624e-05,
      "loss": 0.0442,
      "step": 1340
    },
    {
      "epoch": 0.7518796992481203,
      "grad_norm": 0.29567277754356086,
      "learning_rate": 9.99840387433008e-05,
      "loss": 0.0439,
      "step": 1350
    },
    {
      "epoch": 0.75744917850181,
      "grad_norm": 0.42840154636629013,
      "learning_rate": 9.998033827183193e-05,
      "loss": 0.0432,
      "step": 1360
    },
    {
      "epoch": 0.7630186577554998,
      "grad_norm": 0.3843192755820257,
      "learning_rate": 9.997625244065113e-05,
      "loss": 0.0419,
      "step": 1370
    },
    {
      "epoch": 0.7685881370091896,
      "grad_norm": 0.2794549966082612,
      "learning_rate": 9.997178128126101e-05,
      "loss": 0.04,
      "step": 1380
    },
    {
      "epoch": 0.7741576162628794,
      "grad_norm": 0.3169631435418986,
      "learning_rate": 9.99669248281353e-05,
      "loss": 0.0395,
      "step": 1390
    },
    {
      "epoch": 0.7797270955165692,
      "grad_norm": 0.2751145242333607,
      "learning_rate": 9.996168311871832e-05,
      "loss": 0.039,
      "step": 1400
    },
    {
      "epoch": 0.7852965747702589,
      "grad_norm": 0.2873246557323452,
      "learning_rate": 9.995605619342484e-05,
      "loss": 0.0386,
      "step": 1410
    },
    {
      "epoch": 0.7908660540239487,
      "grad_norm": 0.26158401303739276,
      "learning_rate": 9.995004409563969e-05,
      "loss": 0.0364,
      "step": 1420
    },
    {
      "epoch": 0.7964355332776385,
      "grad_norm": 0.3512702948952268,
      "learning_rate": 9.994364687171756e-05,
      "loss": 0.0358,
      "step": 1430
    },
    {
      "epoch": 0.8020050125313283,
      "grad_norm": 0.30233442060518095,
      "learning_rate": 9.993686457098246e-05,
      "loss": 0.0355,
      "step": 1440
    },
    {
      "epoch": 0.8075744917850181,
      "grad_norm": 0.31354227279951397,
      "learning_rate": 9.992969724572746e-05,
      "loss": 0.0355,
      "step": 1450
    },
    {
      "epoch": 0.8131439710387078,
      "grad_norm": 0.257194967469868,
      "learning_rate": 9.992214495121428e-05,
      "loss": 0.0357,
      "step": 1460
    },
    {
      "epoch": 0.8187134502923976,
      "grad_norm": 0.3479392485987129,
      "learning_rate": 9.99142077456728e-05,
      "loss": 0.0345,
      "step": 1470
    },
    {
      "epoch": 0.8242829295460874,
      "grad_norm": 0.24684398703549518,
      "learning_rate": 9.990588569030068e-05,
      "loss": 0.0337,
      "step": 1480
    },
    {
      "epoch": 0.8298524087997772,
      "grad_norm": 0.23435105711158805,
      "learning_rate": 9.989717884926286e-05,
      "loss": 0.0328,
      "step": 1490
    },
    {
      "epoch": 0.835421888053467,
      "grad_norm": 0.226826410374761,
      "learning_rate": 9.988808728969104e-05,
      "loss": 0.0325,
      "step": 1500
    },
    {
      "epoch": 0.8409913673071567,
      "grad_norm": 0.23919961520195954,
      "learning_rate": 9.987861108168317e-05,
      "loss": 0.0319,
      "step": 1510
    },
    {
      "epoch": 0.8465608465608465,
      "grad_norm": 0.262516092950579,
      "learning_rate": 9.986875029830302e-05,
      "loss": 0.0311,
      "step": 1520
    },
    {
      "epoch": 0.8521303258145363,
      "grad_norm": 0.27505170240477905,
      "learning_rate": 9.985850501557941e-05,
      "loss": 0.0306,
      "step": 1530
    },
    {
      "epoch": 0.8576998050682261,
      "grad_norm": 0.2136353900303048,
      "learning_rate": 9.984787531250581e-05,
      "loss": 0.0306,
      "step": 1540
    },
    {
      "epoch": 0.8632692843219159,
      "grad_norm": 0.2681884993661502,
      "learning_rate": 9.983686127103963e-05,
      "loss": 0.0304,
      "step": 1550
    },
    {
      "epoch": 0.8688387635756056,
      "grad_norm": 0.27064206276631336,
      "learning_rate": 9.982546297610163e-05,
      "loss": 0.0302,
      "step": 1560
    },
    {
      "epoch": 0.8744082428292954,
      "grad_norm": 0.22756357273919053,
      "learning_rate": 9.981368051557522e-05,
      "loss": 0.0296,
      "step": 1570
    },
    {
      "epoch": 0.8799777220829852,
      "grad_norm": 0.23817203224591024,
      "learning_rate": 9.980151398030586e-05,
      "loss": 0.029,
      "step": 1580
    },
    {
      "epoch": 0.885547201336675,
      "grad_norm": 0.23612451468006537,
      "learning_rate": 9.978896346410025e-05,
      "loss": 0.0283,
      "step": 1590
    },
    {
      "epoch": 0.8911166805903648,
      "grad_norm": 0.21633179926653082,
      "learning_rate": 9.977602906372577e-05,
      "loss": 0.0284,
      "step": 1600
    },
    {
      "epoch": 0.8966861598440545,
      "grad_norm": 0.3007834433966076,
      "learning_rate": 9.97627108789095e-05,
      "loss": 0.0276,
      "step": 1610
    },
    {
      "epoch": 0.9022556390977443,
      "grad_norm": 0.22874451896032774,
      "learning_rate": 9.974900901233772e-05,
      "loss": 0.0275,
      "step": 1620
    },
    {
      "epoch": 0.9078251183514341,
      "grad_norm": 0.21229440559613427,
      "learning_rate": 9.973492356965488e-05,
      "loss": 0.0272,
      "step": 1630
    },
    {
      "epoch": 0.9133945976051239,
      "grad_norm": 0.21316799793517352,
      "learning_rate": 9.972045465946295e-05,
      "loss": 0.0266,
      "step": 1640
    },
    {
      "epoch": 0.9189640768588136,
      "grad_norm": 0.21840470782406107,
      "learning_rate": 9.97056023933205e-05,
      "loss": 0.0262,
      "step": 1650
    },
    {
      "epoch": 0.9245335561125034,
      "grad_norm": 0.2296202067200069,
      "learning_rate": 9.969036688574184e-05,
      "loss": 0.0262,
      "step": 1660
    },
    {
      "epoch": 0.9301030353661932,
      "grad_norm": 0.20035916457466751,
      "learning_rate": 9.967474825419619e-05,
      "loss": 0.0256,
      "step": 1670
    },
    {
      "epoch": 0.935672514619883,
      "grad_norm": 0.2392722746437584,
      "learning_rate": 9.965874661910675e-05,
      "loss": 0.0252,
      "step": 1680
    },
    {
      "epoch": 0.9412419938735728,
      "grad_norm": 0.2373750894769379,
      "learning_rate": 9.964236210384972e-05,
      "loss": 0.025,
      "step": 1690
    },
    {
      "epoch": 0.9468114731272625,
      "grad_norm": 0.23565128407439787,
      "learning_rate": 9.962559483475342e-05,
      "loss": 0.0245,
      "step": 1700
    },
    {
      "epoch": 0.9523809523809523,
      "grad_norm": 0.17651355447696596,
      "learning_rate": 9.960844494109731e-05,
      "loss": 0.024,
      "step": 1710
    },
    {
      "epoch": 0.9579504316346421,
      "grad_norm": 0.19685563495858988,
      "learning_rate": 9.959091255511093e-05,
      "loss": 0.0237,
      "step": 1720
    },
    {
      "epoch": 0.9635199108883319,
      "grad_norm": 0.1904961294977808,
      "learning_rate": 9.957299781197295e-05,
      "loss": 0.0235,
      "step": 1730
    },
    {
      "epoch": 0.9690893901420217,
      "grad_norm": 0.21714296341973274,
      "learning_rate": 9.955470084981009e-05,
      "loss": 0.0232,
      "step": 1740
    },
    {
      "epoch": 0.9746588693957114,
      "grad_norm": 0.21953603915286493,
      "learning_rate": 9.953602180969606e-05,
      "loss": 0.0227,
      "step": 1750
    },
    {
      "epoch": 0.9802283486494012,
      "grad_norm": 0.21497187075207344,
      "learning_rate": 9.951696083565048e-05,
      "loss": 0.0224,
      "step": 1760
    },
    {
      "epoch": 0.985797827903091,
      "grad_norm": 0.18608055976176674,
      "learning_rate": 9.949751807463777e-05,
      "loss": 0.0226,
      "step": 1770
    },
    {
      "epoch": 0.9913673071567808,
      "grad_norm": 0.20352070118783105,
      "learning_rate": 9.947769367656603e-05,
      "loss": 0.0223,
      "step": 1780
    },
    {
      "epoch": 0.9969367864104706,
      "grad_norm": 0.24791255800751252,
      "learning_rate": 9.945748779428585e-05,
      "loss": 0.0217,
      "step": 1790
    },
    {
      "epoch": 1.002227791701476,
      "grad_norm": 0.22472692732126306,
      "learning_rate": 9.943690058358915e-05,
      "loss": 0.0217,
      "step": 1800
    }
  ],
  "logging_steps": 10,
  "max_steps": 12572,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 7,
  "save_steps": 1800,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 149526216179712.0,
  "train_batch_size": 128,
  "trial_name": null,
  "trial_params": null
}