{
  "best_global_step": 3200,
  "best_metric": 5.454287528991699,
  "best_model_checkpoint": "/root/odia_ocr/checkpoints_improved/checkpoint-3200",
  "epoch": 1.5016046018770814,
  "eval_steps": 200,
  "global_step": 6200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.004844081138359067,
      "grad_norm": 12.812041282653809,
      "learning_rate": 1.532258064516129e-06,
      "loss": 12.957037353515625,
      "step": 20
    },
    {
      "epoch": 0.009688162276718135,
      "grad_norm": 5.143731117248535,
      "learning_rate": 3.1451612903225806e-06,
      "loss": 11.34280776977539,
      "step": 40
    },
    {
      "epoch": 0.014532243415077202,
      "grad_norm": 1.9331496953964233,
      "learning_rate": 4.758064516129032e-06,
      "loss": 8.1042724609375,
      "step": 60
    },
    {
      "epoch": 0.01937632455343627,
      "grad_norm": 0.4644749164581299,
      "learning_rate": 6.370967741935484e-06,
      "loss": 5.809648132324218,
      "step": 80
    },
    {
      "epoch": 0.024220405691795337,
      "grad_norm": 0.11639782786369324,
      "learning_rate": 7.983870967741935e-06,
      "loss": 5.1223915100097654,
      "step": 100
    },
    {
      "epoch": 0.029064486830154404,
      "grad_norm": 0.05908145383000374,
      "learning_rate": 9.596774193548387e-06,
      "loss": 4.9434257507324215,
      "step": 120
    },
    {
      "epoch": 0.03390856796851347,
      "grad_norm": 0.04645635560154915,
      "learning_rate": 1.1209677419354839e-05,
      "loss": 4.949678802490235,
      "step": 140
    },
    {
      "epoch": 0.03875264910687254,
      "grad_norm": 0.05589432641863823,
      "learning_rate": 1.2822580645161291e-05,
      "loss": 4.812170791625976,
      "step": 160
    },
    {
      "epoch": 0.043596730245231606,
      "grad_norm": 0.05051477253437042,
      "learning_rate": 1.4435483870967741e-05,
      "loss": 4.792440032958984,
      "step": 180
    },
    {
      "epoch": 0.04844081138359067,
      "grad_norm": 0.0538371242582798,
      "learning_rate": 1.6048387096774197e-05,
      "loss": 5.007255172729492,
      "step": 200
    },
    {
      "epoch": 0.04844081138359067,
      "eval_loss": 5.5122504234313965,
      "eval_runtime": 481.6195,
      "eval_samples_per_second": 15.24,
      "eval_steps_per_second": 7.62,
      "step": 200
    },
    {
      "epoch": 0.05328489252194974,
      "grad_norm": 0.07251875847578049,
      "learning_rate": 1.7661290322580645e-05,
      "loss": 4.900541305541992,
      "step": 220
    },
    {
      "epoch": 0.05812897366030881,
      "grad_norm": 0.05386732146143913,
      "learning_rate": 1.9274193548387097e-05,
      "loss": 4.945923233032227,
      "step": 240
    },
    {
      "epoch": 0.06297305479866788,
      "grad_norm": 0.07287013530731201,
      "learning_rate": 2.088709677419355e-05,
      "loss": 4.906622314453125,
      "step": 260
    },
    {
      "epoch": 0.06781713593702694,
      "grad_norm": 0.06443014740943909,
      "learning_rate": 2.25e-05,
      "loss": 4.96992073059082,
      "step": 280
    },
    {
      "epoch": 0.07266121707538602,
      "grad_norm": 0.051735080778598785,
      "learning_rate": 2.4112903225806453e-05,
      "loss": 4.886655044555664,
      "step": 300
    },
    {
      "epoch": 0.07750529821374508,
      "grad_norm": 0.06519602984189987,
      "learning_rate": 2.5725806451612905e-05,
      "loss": 4.9916423797607425,
      "step": 320
    },
    {
      "epoch": 0.08234937935210415,
      "grad_norm": 0.06522039324045181,
      "learning_rate": 2.7338709677419354e-05,
      "loss": 4.85333137512207,
      "step": 340
    },
    {
      "epoch": 0.08719346049046321,
      "grad_norm": 0.09097781777381897,
      "learning_rate": 2.8951612903225806e-05,
      "loss": 5.001050186157227,
      "step": 360
    },
    {
      "epoch": 0.09203754162882229,
      "grad_norm": 0.06196173280477524,
      "learning_rate": 3.056451612903226e-05,
      "loss": 4.854343032836914,
      "step": 380
    },
    {
      "epoch": 0.09688162276718135,
      "grad_norm": 0.0945107564330101,
      "learning_rate": 3.217741935483871e-05,
      "loss": 4.964828872680664,
      "step": 400
    },
    {
      "epoch": 0.09688162276718135,
      "eval_loss": 5.484151363372803,
      "eval_runtime": 482.9864,
      "eval_samples_per_second": 15.197,
      "eval_steps_per_second": 7.599,
      "step": 400
    },
    {
      "epoch": 0.10172570390554042,
      "grad_norm": 0.0707496702671051,
      "learning_rate": 3.379032258064516e-05,
      "loss": 5.008393859863281,
      "step": 420
    },
    {
      "epoch": 0.10656978504389948,
      "grad_norm": 0.0661933496594429,
      "learning_rate": 3.5403225806451614e-05,
      "loss": 4.982975006103516,
      "step": 440
    },
    {
      "epoch": 0.11141386618225856,
      "grad_norm": 0.07057442516088486,
      "learning_rate": 3.701612903225807e-05,
      "loss": 4.927794647216797,
      "step": 460
    },
    {
      "epoch": 0.11625794732061762,
      "grad_norm": 0.05171314626932144,
      "learning_rate": 3.862903225806452e-05,
      "loss": 4.754631042480469,
      "step": 480
    },
    {
      "epoch": 0.12110202845897669,
      "grad_norm": 0.0782339945435524,
      "learning_rate": 4.024193548387097e-05,
      "loss": 4.914446640014648,
      "step": 500
    },
    {
      "epoch": 0.12594610959733576,
      "grad_norm": 0.07288593798875809,
      "learning_rate": 4.1854838709677415e-05,
      "loss": 4.860435485839844,
      "step": 520
    },
    {
      "epoch": 0.1307901907356948,
      "grad_norm": 0.05010804161429405,
      "learning_rate": 4.346774193548388e-05,
      "loss": 4.710846328735352,
      "step": 540
    },
    {
      "epoch": 0.13563427187405389,
      "grad_norm": 0.0925319567322731,
      "learning_rate": 4.5080645161290326e-05,
      "loss": 4.98031005859375,
      "step": 560
    },
    {
      "epoch": 0.14047835301241296,
      "grad_norm": 0.047361522912979126,
      "learning_rate": 4.6693548387096775e-05,
      "loss": 4.936359405517578,
      "step": 580
    },
    {
      "epoch": 0.14532243415077203,
      "grad_norm": 0.06307175010442734,
      "learning_rate": 4.830645161290323e-05,
      "loss": 4.697824478149414,
      "step": 600
    },
    {
      "epoch": 0.14532243415077203,
      "eval_loss": 5.4713640213012695,
      "eval_runtime": 483.5438,
      "eval_samples_per_second": 15.18,
      "eval_steps_per_second": 7.59,
      "step": 600
    },
    {
      "epoch": 0.15016651528913108,
      "grad_norm": 0.06868302077054977,
      "learning_rate": 4.991935483870968e-05,
      "loss": 4.9246562957763675,
      "step": 620
    },
    {
      "epoch": 0.15501059642749015,
      "grad_norm": 0.06361743062734604,
      "learning_rate": 4.9999678349284276e-05,
      "loss": 4.979264450073242,
      "step": 640
    },
    {
      "epoch": 0.15985467756584923,
      "grad_norm": 0.06285427510738373,
      "learning_rate": 4.999864479953565e-05,
      "loss": 4.902791976928711,
      "step": 660
    },
    {
      "epoch": 0.1646987587042083,
      "grad_norm": 0.048156049102544785,
      "learning_rate": 4.999689848922155e-05,
      "loss": 4.905753326416016,
      "step": 680
    },
    {
      "epoch": 0.16954283984256735,
      "grad_norm": 0.036559417843818665,
      "learning_rate": 4.999443946813271e-05,
      "loss": 4.906116485595703,
      "step": 700
    },
    {
      "epoch": 0.17438692098092642,
      "grad_norm": 0.04299989715218544,
      "learning_rate": 4.999126780638068e-05,
      "loss": 4.8410499572753904,
      "step": 720
    },
    {
      "epoch": 0.1792310021192855,
      "grad_norm": 0.04031865671277046,
      "learning_rate": 4.9987383594395764e-05,
      "loss": 4.942621231079102,
      "step": 740
    },
    {
      "epoch": 0.18407508325764457,
      "grad_norm": 0.0709218680858612,
      "learning_rate": 4.99827869429245e-05,
      "loss": 4.831058120727539,
      "step": 760
    },
    {
      "epoch": 0.18891916439600362,
      "grad_norm": 0.04334535449743271,
      "learning_rate": 4.9977477983026456e-05,
      "loss": 4.815376663208008,
      "step": 780
    },
    {
      "epoch": 0.1937632455343627,
      "grad_norm": 0.08114223182201385,
      "learning_rate": 4.997145686607054e-05,
      "loss": 4.922482681274414,
      "step": 800
    },
    {
      "epoch": 0.1937632455343627,
      "eval_loss": 5.467911720275879,
      "eval_runtime": 486.133,
      "eval_samples_per_second": 15.099,
      "eval_steps_per_second": 7.549,
      "step": 800
    },
    {
      "epoch": 0.19860732667272177,
      "grad_norm": 0.03837033733725548,
      "learning_rate": 4.996472376373064e-05,
      "loss": 4.721331787109375,
      "step": 820
    },
    {
      "epoch": 0.20345140781108084,
      "grad_norm": 0.042904820293188095,
      "learning_rate": 4.995727886798076e-05,
      "loss": 4.922665405273437,
      "step": 840
    },
    {
      "epoch": 0.20829548894943992,
      "grad_norm": 0.042334090918302536,
      "learning_rate": 4.994912239108953e-05,
      "loss": 4.739095306396484,
      "step": 860
    },
    {
      "epoch": 0.21313957008779896,
      "grad_norm": 0.05880925804376602,
      "learning_rate": 4.994025456561415e-05,
      "loss": 4.862358856201172,
      "step": 880
    },
    {
      "epoch": 0.21798365122615804,
      "grad_norm": 0.05243176966905594,
      "learning_rate": 4.99306756443938e-05,
      "loss": 4.801820755004883,
      "step": 900
    },
    {
      "epoch": 0.2228277323645171,
      "grad_norm": 0.04455817863345146,
      "learning_rate": 4.992038590054235e-05,
      "loss": 4.858511352539063,
      "step": 920
    },
    {
      "epoch": 0.22767181350287619,
      "grad_norm": 0.052609678357839584,
      "learning_rate": 4.990938562744067e-05,
      "loss": 4.80189323425293,
      "step": 940
    },
    {
      "epoch": 0.23251589464123523,
      "grad_norm": 0.03932235389947891,
      "learning_rate": 4.9897675138728194e-05,
      "loss": 4.971810150146484,
      "step": 960
    },
    {
      "epoch": 0.2373599757795943,
      "grad_norm": 0.034223828464746475,
      "learning_rate": 4.9885254768293993e-05,
      "loss": 4.843924713134766,
      "step": 980
    },
    {
      "epoch": 0.24220405691795338,
      "grad_norm": 0.052775464951992035,
      "learning_rate": 4.9872124870267277e-05,
      "loss": 5.065494537353516,
      "step": 1000
    },
    {
      "epoch": 0.24220405691795338,
      "eval_loss": 5.465411186218262,
      "eval_runtime": 482.3283,
      "eval_samples_per_second": 15.218,
      "eval_steps_per_second": 7.609,
      "step": 1000
    },
    {
      "epoch": 0.24704813805631246,
      "grad_norm": 0.02837868221104145,
      "learning_rate": 4.985828581900729e-05,
      "loss": 4.921312713623047,
      "step": 1020
    },
    {
      "epoch": 0.25189221919467153,
      "grad_norm": 0.03618283197283745,
      "learning_rate": 4.984373800909262e-05,
      "loss": 4.854024887084961,
      "step": 1040
    },
    {
      "epoch": 0.2567363003330306,
      "grad_norm": 0.03154360130429268,
      "learning_rate": 4.982848185530995e-05,
      "loss": 4.909141540527344,
      "step": 1060
    },
    {
      "epoch": 0.2615803814713896,
      "grad_norm": 0.03342705965042114,
      "learning_rate": 4.981251779264227e-05,
      "loss": 4.873151779174805,
      "step": 1080
    },
    {
      "epoch": 0.2664244626097487,
      "grad_norm": 0.038160618394613266,
      "learning_rate": 4.9795846276256395e-05,
      "loss": 4.937899780273438,
      "step": 1100
    },
    {
      "epoch": 0.27126854374810777,
      "grad_norm": 0.02787560038268566,
      "learning_rate": 4.97784677814901e-05,
      "loss": 4.889313888549805,
      "step": 1120
    },
    {
      "epoch": 0.2761126248864669,
      "grad_norm": 0.03252527490258217,
      "learning_rate": 4.976038280383847e-05,
      "loss": 5.092146301269532,
      "step": 1140
    },
    {
      "epoch": 0.2809567060248259,
      "grad_norm": 0.03215664252638817,
      "learning_rate": 4.97415918589398e-05,
      "loss": 4.881315994262695,
      "step": 1160
    },
    {
      "epoch": 0.28580078716318497,
      "grad_norm": 0.056495729833841324,
      "learning_rate": 4.972209548256093e-05,
      "loss": 4.913896179199218,
      "step": 1180
    },
    {
      "epoch": 0.29064486830154407,
      "grad_norm": 0.023406801745295525,
      "learning_rate": 4.97018942305819e-05,
      "loss": 5.040941619873047,
      "step": 1200
    },
    {
      "epoch": 0.29064486830154407,
      "eval_loss": 5.464733123779297,
      "eval_runtime": 479.1861,
      "eval_samples_per_second": 15.318,
      "eval_steps_per_second": 7.659,
      "step": 1200
    },
    {
      "epoch": 0.2954889494399031,
      "grad_norm": 0.03091084584593773,
      "learning_rate": 4.968098867898019e-05,
      "loss": 4.7054706573486325,
      "step": 1220
    },
    {
      "epoch": 0.30033303057826216,
      "grad_norm": 0.033658791333436966,
      "learning_rate": 4.965937942381421e-05,
      "loss": 4.971173858642578,
      "step": 1240
    },
    {
      "epoch": 0.30517711171662126,
      "grad_norm": 0.03186121955513954,
      "learning_rate": 4.963706708120635e-05,
      "loss": 4.918471908569336,
      "step": 1260
    },
    {
      "epoch": 0.3100211928549803,
      "grad_norm": 0.037234991788864136,
      "learning_rate": 4.9614052287325426e-05,
      "loss": 4.957798004150391,
      "step": 1280
    },
    {
      "epoch": 0.3148652739933394,
      "grad_norm": 0.04513128846883774,
      "learning_rate": 4.95903356983685e-05,
      "loss": 4.933160400390625,
      "step": 1300
    },
    {
      "epoch": 0.31970935513169846,
      "grad_norm": 0.021936912089586258,
      "learning_rate": 4.95659179905422e-05,
      "loss": 4.837236404418945,
      "step": 1320
    },
    {
      "epoch": 0.3245534362700575,
      "grad_norm": 0.047507770359516144,
      "learning_rate": 4.954079986004345e-05,
      "loss": 4.874291229248047,
      "step": 1340
    },
    {
      "epoch": 0.3293975174084166,
      "grad_norm": 0.02777005545794964,
      "learning_rate": 4.951498202303957e-05,
      "loss": 4.831377410888672,
      "step": 1360
    },
    {
      "epoch": 0.33424159854677565,
      "grad_norm": 0.04671608284115791,
      "learning_rate": 4.948846521564792e-05,
      "loss": 4.9074962615966795,
      "step": 1380
    },
    {
      "epoch": 0.3390856796851347,
      "grad_norm": 0.05117534101009369,
      "learning_rate": 4.946125019391486e-05,
      "loss": 4.955949783325195,
      "step": 1400
    },
    {
      "epoch": 0.3390856796851347,
      "eval_loss": 5.462268829345703,
      "eval_runtime": 480.5711,
      "eval_samples_per_second": 15.273,
      "eval_steps_per_second": 7.637,
      "step": 1400
    },
    {
      "epoch": 0.3439297608234938,
      "grad_norm": 0.04413221776485443,
      "learning_rate": 4.9433337733794225e-05,
      "loss": 4.930236434936523,
      "step": 1420
    },
    {
      "epoch": 0.34877384196185285,
      "grad_norm": 0.031841397285461426,
      "learning_rate": 4.94047286311252e-05,
      "loss": 4.9899742126464846,
      "step": 1440
    },
    {
      "epoch": 0.35361792310021195,
      "grad_norm": 0.03400244191288948,
      "learning_rate": 4.93754237016096e-05,
      "loss": 4.879816055297852,
      "step": 1460
    },
    {
      "epoch": 0.358462004238571,
      "grad_norm": 0.02796698734164238,
      "learning_rate": 4.9345423780788655e-05,
      "loss": 4.900162124633789,
      "step": 1480
    },
    {
      "epoch": 0.36330608537693004,
      "grad_norm": 0.033209074288606644,
      "learning_rate": 4.9314729724019135e-05,
      "loss": 4.824586868286133,
      "step": 1500
    },
    {
      "epoch": 0.36815016651528915,
      "grad_norm": 0.03357836231589317,
      "learning_rate": 4.928334240644903e-05,
      "loss": 4.901557540893554,
      "step": 1520
    },
    {
      "epoch": 0.3729942476536482,
      "grad_norm": 0.030524376779794693,
      "learning_rate": 4.925126272299253e-05,
      "loss": 4.955865097045899,
      "step": 1540
    },
    {
      "epoch": 0.37783832879200724,
      "grad_norm": 0.02066030353307724,
      "learning_rate": 4.921849158830455e-05,
      "loss": 4.921824264526367,
      "step": 1560
    },
    {
      "epoch": 0.38268240993036634,
      "grad_norm": 0.03722696006298065,
      "learning_rate": 4.918502993675463e-05,
      "loss": 4.897643661499023,
      "step": 1580
    },
    {
      "epoch": 0.3875264910687254,
      "grad_norm": 0.03600393235683441,
      "learning_rate": 4.915087872240034e-05,
      "loss": 4.867633819580078,
      "step": 1600
    },
    {
      "epoch": 0.3875264910687254,
      "eval_loss": 5.463290691375732,
      "eval_runtime": 482.557,
      "eval_samples_per_second": 15.211,
      "eval_steps_per_second": 7.605,
      "step": 1600
    },
    {
      "epoch": 0.3923705722070845,
      "grad_norm": 0.01853497512638569,
      "learning_rate": 4.911603891896e-05,
      "loss": 4.990079116821289,
      "step": 1620
    },
    {
      "epoch": 0.39721465334544354,
      "grad_norm": 0.03120853751897812,
      "learning_rate": 4.9080511519784975e-05,
      "loss": 4.889575576782226,
      "step": 1640
    },
    {
      "epoch": 0.4020587344838026,
      "grad_norm": 0.04181843250989914,
      "learning_rate": 4.9044297537831355e-05,
      "loss": 4.894519424438476,
      "step": 1660
    },
    {
      "epoch": 0.4069028156221617,
      "grad_norm": 0.027956627309322357,
      "learning_rate": 4.900739800563103e-05,
      "loss": 5.025855255126953,
      "step": 1680
    },
    {
      "epoch": 0.41174689676052073,
      "grad_norm": 0.025699257850646973,
      "learning_rate": 4.896981397526229e-05,
      "loss": 4.933831405639649,
      "step": 1700
    },
    {
      "epoch": 0.41659097789887983,
      "grad_norm": 0.03355032578110695,
      "learning_rate": 4.893154651831981e-05,
      "loss": 4.883275985717773,
      "step": 1720
    },
    {
      "epoch": 0.4214350590372389,
      "grad_norm": 0.02468765340745449,
      "learning_rate": 4.8892596725884124e-05,
      "loss": 4.825000762939453,
      "step": 1740
    },
    {
      "epoch": 0.4262791401755979,
      "grad_norm": 0.04183134064078331,
      "learning_rate": 4.885296570849045e-05,
      "loss": 4.861746978759766,
      "step": 1760
    },
    {
      "epoch": 0.43112322131395703,
      "grad_norm": 0.028777016326785088,
      "learning_rate": 4.881265459609711e-05,
      "loss": 4.734410095214844,
      "step": 1780
    },
    {
      "epoch": 0.4359673024523161,
      "grad_norm": 0.11064556241035461,
      "learning_rate": 4.877166453805325e-05,
      "loss": 4.908732604980469,
      "step": 1800
    },
    {
      "epoch": 0.4359673024523161,
      "eval_loss": 5.462699890136719,
      "eval_runtime": 484.184,
      "eval_samples_per_second": 15.16,
      "eval_steps_per_second": 7.58,
      "step": 1800
    },
    {
      "epoch": 0.4408113835906751,
      "grad_norm": 0.03197271004319191,
      "learning_rate": 4.8729996703066117e-05,
      "loss": 4.928284072875977,
      "step": 1820
    },
    {
      "epoch": 0.4456554647290342,
      "grad_norm": 0.050927624106407166,
      "learning_rate": 4.868765227916769e-05,
      "loss": 4.886736297607422,
      "step": 1840
    },
    {
      "epoch": 0.45049954586739327,
      "grad_norm": 0.02937314845621586,
      "learning_rate": 4.864463247368082e-05,
      "loss": 4.795632934570312,
      "step": 1860
    },
    {
      "epoch": 0.45534362700575237,
      "grad_norm": 0.03579329699277878,
      "learning_rate": 4.8600938513184847e-05,
      "loss": 4.779296875,
      "step": 1880
    },
    {
      "epoch": 0.4601877081441114,
      "grad_norm": 0.03522449731826782,
      "learning_rate": 4.8556571643480585e-05,
      "loss": 4.904380416870117,
      "step": 1900
    },
    {
      "epoch": 0.46503178928247046,
      "grad_norm": 0.023437773808836937,
      "learning_rate": 4.8511533129554816e-05,
      "loss": 4.982876968383789,
      "step": 1920
    },
    {
      "epoch": 0.46987587042082957,
      "grad_norm": 0.015325312502682209,
      "learning_rate": 4.84658242555442e-05,
      "loss": 4.990073394775391,
      "step": 1940
    },
    {
      "epoch": 0.4747199515591886,
      "grad_norm": 0.030531391501426697,
      "learning_rate": 4.841944632469872e-05,
      "loss": 4.923136901855469,
      "step": 1960
    },
    {
      "epoch": 0.47956403269754766,
      "grad_norm": 0.0198193471878767,
      "learning_rate": 4.8372400659344444e-05,
      "loss": 4.956871414184571,
      "step": 1980
    },
    {
      "epoch": 0.48440811383590676,
      "grad_norm": 0.0377812422811985,
      "learning_rate": 4.83246886008459e-05,
      "loss": 4.910104370117187,
      "step": 2000
    },
    {
      "epoch": 0.48440811383590676,
      "eval_loss": 5.462129592895508,
      "eval_runtime": 481.7351,
      "eval_samples_per_second": 15.237,
      "eval_steps_per_second": 7.618,
      "step": 2000
    },
    {
      "epoch": 0.4892521949742658,
      "grad_norm": 0.03756606578826904,
      "learning_rate": 4.8276311509567785e-05,
      "loss": 4.961421203613281,
      "step": 2020
    },
    {
      "epoch": 0.4940962761126249,
      "grad_norm": 0.025465775281190872,
      "learning_rate": 4.8227270764836175e-05,
      "loss": 4.971789932250976,
      "step": 2040
    },
    {
      "epoch": 0.49894035725098396,
      "grad_norm": 0.04584196209907532,
      "learning_rate": 4.817756776489923e-05,
      "loss": 4.722394180297852,
      "step": 2060
    },
    {
      "epoch": 0.5037844383893431,
      "grad_norm": 0.024054571986198425,
      "learning_rate": 4.812720392688731e-05,
      "loss": 4.861309432983399,
      "step": 2080
    },
    {
      "epoch": 0.508628519527702,
      "grad_norm": 0.021322039887309074,
      "learning_rate": 4.807618068677254e-05,
      "loss": 4.906083679199218,
      "step": 2100
    },
    {
      "epoch": 0.5134726006660612,
      "grad_norm": 0.024236466735601425,
      "learning_rate": 4.802449949932796e-05,
      "loss": 4.774163818359375,
      "step": 2120
    },
    {
      "epoch": 0.5183166818044203,
      "grad_norm": 0.025150248780846596,
      "learning_rate": 4.79721618380859e-05,
      "loss": 5.06263427734375,
      "step": 2140
    },
    {
      "epoch": 0.5231607629427792,
      "grad_norm": 0.016574619337916374,
      "learning_rate": 4.791916919529611e-05,
      "loss": 4.989673233032226,
      "step": 2160
    },
    {
      "epoch": 0.5280048440811383,
      "grad_norm": 0.022480934858322144,
      "learning_rate": 4.786552308188314e-05,
      "loss": 4.921234130859375,
      "step": 2180
    },
    {
      "epoch": 0.5328489252194974,
      "grad_norm": 0.018986444920301437,
      "learning_rate": 4.781122502740329e-05,
      "loss": 4.958284378051758,
      "step": 2200
    },
    {
      "epoch": 0.5328489252194974,
      "eval_loss": 5.459774017333984,
      "eval_runtime": 481.5781,
      "eval_samples_per_second": 15.242,
      "eval_steps_per_second": 7.621,
      "step": 2200
    },
    {
      "epoch": 0.5376930063578564,
      "grad_norm": 0.024930085986852646,
      "learning_rate": 4.775627658000091e-05,
      "loss": 4.923190307617188,
      "step": 2220
    },
    {
      "epoch": 0.5425370874962155,
      "grad_norm": 0.016377661377191544,
      "learning_rate": 4.770067930636442e-05,
      "loss": 4.9806373596191404,
      "step": 2240
    },
    {
      "epoch": 0.5473811686345746,
      "grad_norm": 0.028611725196242332,
      "learning_rate": 4.764443479168151e-05,
      "loss": 4.843650054931641,
      "step": 2260
    },
    {
      "epoch": 0.5522252497729337,
      "grad_norm": 0.02169787883758545,
      "learning_rate": 4.758754463959401e-05,
      "loss": 4.820799255371094,
      "step": 2280
    },
    {
      "epoch": 0.5570693309112927,
      "grad_norm": 0.014351587742567062,
      "learning_rate": 4.75300104721521e-05,
      "loss": 4.888659286499023,
      "step": 2300
    },
    {
      "epoch": 0.5619134120496518,
      "grad_norm": 0.02870621345937252,
      "learning_rate": 4.747183392976815e-05,
      "loss": 4.703000259399414,
      "step": 2320
    },
    {
      "epoch": 0.5667574931880109,
      "grad_norm": 0.02605106681585312,
      "learning_rate": 4.741301667116987e-05,
      "loss": 4.83842658996582,
      "step": 2340
    },
    {
      "epoch": 0.5716015743263699,
      "grad_norm": 0.029983455315232277,
      "learning_rate": 4.735356037335309e-05,
      "loss": 4.874066925048828,
      "step": 2360
    },
    {
      "epoch": 0.576445655464729,
      "grad_norm": 0.022755078971385956,
      "learning_rate": 4.729346673153386e-05,
      "loss": 4.92518424987793,
      "step": 2380
    },
    {
      "epoch": 0.5812897366030881,
      "grad_norm": 0.026459872722625732,
      "learning_rate": 4.7232737459100206e-05,
      "loss": 4.806256866455078,
      "step": 2400
    },
    {
      "epoch": 0.5812897366030881,
      "eval_loss": 5.461882591247559,
      "eval_runtime": 482.7564,
      "eval_samples_per_second": 15.204,
      "eval_steps_per_second": 7.602,
      "step": 2400
    },
    {
      "epoch": 0.5861338177414471,
      "grad_norm": 0.03340192884206772,
      "learning_rate": 4.7171374287563196e-05,
      "loss": 4.913430786132812,
      "step": 2420
    },
    {
      "epoch": 0.5909778988798062,
      "grad_norm": 0.027108795940876007,
      "learning_rate": 4.710937896650763e-05,
      "loss": 4.880179214477539,
      "step": 2440
    },
    {
      "epoch": 0.5958219800181653,
      "grad_norm": 0.026677660644054413,
      "learning_rate": 4.704675326354213e-05,
      "loss": 4.978963470458984,
      "step": 2460
    },
    {
      "epoch": 0.6006660611565243,
      "grad_norm": 0.023100655525922775,
      "learning_rate": 4.698349896424874e-05,
      "loss": 4.817289352416992,
      "step": 2480
    },
    {
      "epoch": 0.6055101422948834,
      "grad_norm": 0.02461540326476097,
      "learning_rate": 4.691961787213204e-05,
      "loss": 4.825991821289063,
      "step": 2500
    },
    {
      "epoch": 0.6103542234332425,
      "grad_norm": 0.02112703211605549,
      "learning_rate": 4.68551118085677e-05,
      "loss": 4.912199783325195,
      "step": 2520
    },
    {
      "epoch": 0.6151983045716016,
      "grad_norm": 0.028458120301365852,
      "learning_rate": 4.678998261275054e-05,
      "loss": 4.932154846191406,
      "step": 2540
    },
    {
      "epoch": 0.6200423857099606,
      "grad_norm": 0.014952564612030983,
      "learning_rate": 4.6724232141642133e-05,
      "loss": 4.791392517089844,
      "step": 2560
    },
    {
      "epoch": 0.6248864668483197,
      "grad_norm": 0.024178558960556984,
      "learning_rate": 4.6657862269917805e-05,
      "loss": 4.836280822753906,
      "step": 2580
    },
    {
      "epoch": 0.6297305479866788,
      "grad_norm": 0.03559703379869461,
      "learning_rate": 4.659087488991325e-05,
      "loss": 4.877938842773437,
      "step": 2600
    },
    {
      "epoch": 0.6297305479866788,
      "eval_loss": 5.461455821990967,
      "eval_runtime": 477.3623,
      "eval_samples_per_second": 15.376,
      "eval_steps_per_second": 7.688,
      "step": 2600
    },
    {
      "epoch": 0.6345746291250378,
      "grad_norm": 0.0435916930437088,
      "learning_rate": 4.6523271911570484e-05,
      "loss": 4.8993488311767575,
      "step": 2620
    },
    {
      "epoch": 0.6394187102633969,
      "grad_norm": 0.03119947761297226,
      "learning_rate": 4.6455055262383504e-05,
      "loss": 4.764545059204101,
      "step": 2640
    },
    {
      "epoch": 0.644262791401756,
      "grad_norm": 0.0388447642326355,
      "learning_rate": 4.6386226887343236e-05,
      "loss": 5.00665283203125,
      "step": 2660
    },
    {
      "epoch": 0.649106872540115,
      "grad_norm": 0.018733784556388855,
      "learning_rate": 4.6316788748882114e-05,
      "loss": 4.9001209259033205,
      "step": 2680
    },
    {
      "epoch": 0.6539509536784741,
      "grad_norm": 0.03436599299311638,
      "learning_rate": 4.6246742826818134e-05,
      "loss": 4.823849868774414,
      "step": 2700
    },
    {
      "epoch": 0.6587950348168332,
      "grad_norm": 0.014111216180026531,
      "learning_rate": 4.6176091118298406e-05,
      "loss": 4.89782943725586,
      "step": 2720
    },
    {
      "epoch": 0.6636391159551922,
      "grad_norm": 0.012464863248169422,
      "learning_rate": 4.610483563774218e-05,
      "loss": 4.9627235412597654,
      "step": 2740
    },
    {
      "epoch": 0.6684831970935513,
      "grad_norm": 0.027146505191922188,
      "learning_rate": 4.603297841678346e-05,
      "loss": 4.825485610961914,
      "step": 2760
    },
    {
      "epoch": 0.6733272782319104,
      "grad_norm": 0.025906676426529884,
      "learning_rate": 4.596052150421306e-05,
      "loss": 4.924727249145508,
      "step": 2780
    },
    {
      "epoch": 0.6781713593702694,
      "grad_norm": 0.030369574204087257,
      "learning_rate": 4.588746696592014e-05,
      "loss": 4.913991928100586,
      "step": 2800
    },
    {
      "epoch": 0.6781713593702694,
      "eval_loss": 5.460803508758545,
      "eval_runtime": 483.5598,
      "eval_samples_per_second": 15.179,
      "eval_steps_per_second": 7.59,
      "step": 2800
    },
    {
      "epoch": 0.6830154405086285,
      "grad_norm": 0.01944783516228199,
      "learning_rate": 4.5813816884833394e-05,
      "loss": 4.978294754028321,
      "step": 2820
    },
    {
      "epoch": 0.6878595216469876,
      "grad_norm": 0.023732610046863556,
      "learning_rate": 4.573957336086157e-05,
      "loss": 4.956566238403321,
      "step": 2840
    },
    {
      "epoch": 0.6927036027853467,
      "grad_norm": 0.01582464762032032,
      "learning_rate": 4.566473851083369e-05,
      "loss": 4.86058464050293,
      "step": 2860
    },
    {
      "epoch": 0.6975476839237057,
      "grad_norm": 0.01658930629491806,
      "learning_rate": 4.558931446843861e-05,
      "loss": 4.795158386230469,
      "step": 2880
    },
    {
      "epoch": 0.7023917650620648,
      "grad_norm": 0.0260655265301466,
      "learning_rate": 4.55133033841642e-05,
      "loss": 4.9587970733642575,
      "step": 2900
    },
    {
      "epoch": 0.7072358462004239,
      "grad_norm": 0.021988550201058388,
      "learning_rate": 4.543670742523611e-05,
      "loss": 4.91474723815918,
      "step": 2920
    },
    {
      "epoch": 0.7120799273387829,
      "grad_norm": 0.017105242237448692,
      "learning_rate": 4.535952877555587e-05,
      "loss": 4.704911804199218,
      "step": 2940
    },
    {
      "epoch": 0.716924008477142,
      "grad_norm": 0.018221445381641388,
      "learning_rate": 4.52817696356387e-05,
      "loss": 4.996733856201172,
      "step": 2960
    },
    {
      "epoch": 0.7217680896155011,
      "grad_norm": 0.022833779454231262,
      "learning_rate": 4.52034322225507e-05,
      "loss": 4.984310531616211,
      "step": 2980
    },
    {
      "epoch": 0.7266121707538601,
      "grad_norm": 0.0188573207706213,
      "learning_rate": 4.5124518769845715e-05,
      "loss": 4.887367630004883,
      "step": 3000
    },
    {
      "epoch": 0.7266121707538601,
      "eval_loss": 5.459744453430176,
      "eval_runtime": 482.3857,
      "eval_samples_per_second": 15.216,
      "eval_steps_per_second": 7.608,
      "step": 3000
    },
    {
      "epoch": 0.7314562518922192,
      "grad_norm": 0.033990390598773956,
      "learning_rate": 4.5045031527501604e-05,
      "loss": 4.899966049194336,
      "step": 3020
    },
    {
      "epoch": 0.7363003330305783,
      "grad_norm": 0.02148909866809845,
      "learning_rate": 4.4964972761856084e-05,
      "loss": 4.8198493957519535,
      "step": 3040
    },
    {
      "epoch": 0.7411444141689373,
      "grad_norm": 0.019260123372077942,
      "learning_rate": 4.4884344755542134e-05,
      "loss": 4.738033294677734,
      "step": 3060
    },
    {
      "epoch": 0.7459884953072964,
      "grad_norm": 0.024531004950404167,
      "learning_rate": 4.480314980742291e-05,
      "loss": 4.879784774780274,
      "step": 3080
    },
    {
      "epoch": 0.7508325764456555,
      "grad_norm": 0.022741004824638367,
      "learning_rate": 4.472139023252619e-05,
      "loss": 4.864042663574219,
      "step": 3100
    },
    {
      "epoch": 0.7556766575840145,
      "grad_norm": 0.03277403116226196,
      "learning_rate": 4.4639068361978375e-05,
      "loss": 4.944465255737304,
      "step": 3120
    },
    {
      "epoch": 0.7605207387223736,
      "grad_norm": 0.01316901296377182,
      "learning_rate": 4.455618654293801e-05,
      "loss": 4.936913299560547,
      "step": 3140
    },
    {
      "epoch": 0.7653648198607327,
      "grad_norm": 0.017750350758433342,
      "learning_rate": 4.44727471385289e-05,
      "loss": 4.916695785522461,
      "step": 3160
    },
    {
      "epoch": 0.7702089009990918,
      "grad_norm": 0.024009494110941887,
      "learning_rate": 4.4388752527772677e-05,
      "loss": 4.825696182250977,
      "step": 3180
    },
    {
      "epoch": 0.7750529821374508,
      "grad_norm": 0.019279325380921364,
      "learning_rate": 4.4304205105521015e-05,
      "loss": 4.900424575805664,
      "step": 3200
    },
    {
      "epoch": 0.7750529821374508,
      "eval_loss": 5.454287528991699,
      "eval_runtime": 483.1463,
      "eval_samples_per_second": 15.192,
      "eval_steps_per_second": 7.596,
      "step": 3200
    },
    {
      "epoch": 0.7798970632758099,
      "grad_norm": 0.011572799645364285,
      "learning_rate": 4.421910728238733e-05,
      "loss": 4.900468826293945,
      "step": 3220
    },
    {
      "epoch": 0.784741144414169,
      "grad_norm": 0.022734951227903366,
      "learning_rate": 4.413346148467804e-05,
      "loss": 4.889355850219727,
      "step": 3240
    },
    {
      "epoch": 0.789585225552528,
      "grad_norm": 0.010450748726725578,
      "learning_rate": 4.4047270154323425e-05,
      "loss": 4.801429748535156,
      "step": 3260
    },
    {
      "epoch": 0.7944293066908871,
      "grad_norm": 0.02443559840321541,
      "learning_rate": 4.396053574880794e-05,
      "loss": 4.850070190429688,
      "step": 3280
    },
    {
      "epoch": 0.7992733878292462,
      "grad_norm": 0.01849784515798092,
      "learning_rate": 4.38732607411002e-05,
      "loss": 4.954670333862305,
      "step": 3300
    },
    {
      "epoch": 0.8041174689676052,
      "grad_norm": 0.018099434673786163,
      "learning_rate": 4.378544761958244e-05,
      "loss": 4.958530044555664,
      "step": 3320
    },
    {
      "epoch": 0.8089615501059643,
      "grad_norm": 0.027655702084302902,
      "learning_rate": 4.36970988879796e-05,
      "loss": 5.005998611450195,
      "step": 3340
    },
    {
      "epoch": 0.8138056312443234,
      "grad_norm": 0.01677539013326168,
      "learning_rate": 4.3608217065287904e-05,
      "loss": 4.924520874023438,
      "step": 3360
    },
    {
      "epoch": 0.8186497123826824,
      "grad_norm": 0.03358837217092514,
      "learning_rate": 4.351880468570305e-05,
      "loss": 4.904378128051758,
      "step": 3380
    },
    {
      "epoch": 0.8234937935210415,
      "grad_norm": 0.02544589340686798,
      "learning_rate": 4.3428864298547965e-05,
      "loss": 4.933415222167969,
      "step": 3400
    },
    {
      "epoch": 0.8234937935210415,
      "eval_loss": 5.461795330047607,
      "eval_runtime": 483.8187,
      "eval_samples_per_second": 15.171,
      "eval_steps_per_second": 7.585,
      "step": 3400
    },
    {
      "epoch": 0.8283378746594006,
      "grad_norm": 0.023529496043920517,
      "learning_rate": 4.3338398468200104e-05,
      "loss": 4.9913169860839846,
      "step": 3420
    },
    {
      "epoch": 0.8331819557977597,
      "grad_norm": 0.014854267239570618,
      "learning_rate": 4.324740977401834e-05,
      "loss": 4.914155960083008,
      "step": 3440
    },
    {
      "epoch": 0.8380260369361187,
      "grad_norm": 0.028307568281888962,
      "learning_rate": 4.3155900810269445e-05,
      "loss": 4.852661514282227,
      "step": 3460
    },
    {
      "epoch": 0.8428701180744778,
      "grad_norm": 0.017718417569994926,
      "learning_rate": 4.306387418605407e-05,
      "loss": 4.959287643432617,
      "step": 3480
    },
    {
      "epoch": 0.8477141992128369,
      "grad_norm": 0.01476296130567789,
      "learning_rate": 4.297133252523239e-05,
      "loss": 4.727375411987305,
      "step": 3500
    },
    {
      "epoch": 0.8525582803511959,
      "grad_norm": 0.024725649505853653,
      "learning_rate": 4.2878278466349316e-05,
      "loss": 4.893677520751953,
      "step": 3520
    },
    {
      "epoch": 0.857402361489555,
      "grad_norm": 0.01822960376739502,
      "learning_rate": 4.27847146625592e-05,
      "loss": 4.89503288269043,
      "step": 3540
    },
    {
      "epoch": 0.8622464426279141,
      "grad_norm": 0.026599068194627762,
      "learning_rate": 4.2690643781550245e-05,
      "loss": 4.930253219604492,
      "step": 3560
    },
    {
      "epoch": 0.867090523766273,
      "grad_norm": 0.0182630755007267,
      "learning_rate": 4.259606850546841e-05,
      "loss": 4.800619888305664,
      "step": 3580
    },
    {
      "epoch": 0.8719346049046321,
      "grad_norm": 0.015945954248309135,
      "learning_rate": 4.250099153084097e-05,
      "loss": 4.828668594360352,
      "step": 3600
    },
    {
      "epoch": 0.8719346049046321,
      "eval_loss": 5.460173606872559,
      "eval_runtime": 484.3695,
      "eval_samples_per_second": 15.154,
      "eval_steps_per_second": 7.577,
      "step": 3600
    },
    {
      "epoch": 0.8767786860429913,
      "grad_norm": 0.014371720142662525,
      "learning_rate": 4.240541556849961e-05,
      "loss": 4.865261459350586,
      "step": 3620
    },
    {
      "epoch": 0.8816227671813502,
      "grad_norm": 0.01884741522371769,
      "learning_rate": 4.23093433435031e-05,
      "loss": 4.983065032958985,
      "step": 3640
    },
    {
      "epoch": 0.8864668483197093,
      "grad_norm": 0.0102775227278471,
      "learning_rate": 4.221277759505968e-05,
      "loss": 4.936088943481446,
      "step": 3660
    },
    {
      "epoch": 0.8913109294580684,
      "grad_norm": 0.02810431458055973,
      "learning_rate": 4.211572107644891e-05,
      "loss": 4.9662517547607425,
      "step": 3680
    },
    {
      "epoch": 0.8961550105964274,
      "grad_norm": 0.014943272806704044,
      "learning_rate": 4.201817655494315e-05,
      "loss": 4.909831237792969,
      "step": 3700
    },
    {
      "epoch": 0.9009990917347865,
      "grad_norm": 0.018770508468151093,
      "learning_rate": 4.192014681172869e-05,
      "loss": 4.761404418945313,
      "step": 3720
    },
    {
      "epoch": 0.9058431728731456,
      "grad_norm": 0.01450326107442379,
      "learning_rate": 4.182163464182646e-05,
      "loss": 4.860451507568359,
      "step": 3740
    },
    {
      "epoch": 0.9106872540115047,
      "grad_norm": 0.010839094407856464,
      "learning_rate": 4.172264285401232e-05,
      "loss": 4.847282028198242,
      "step": 3760
    },
    {
      "epoch": 0.9155313351498637,
      "grad_norm": 0.009730652906000614,
      "learning_rate": 4.162317427073695e-05,
      "loss": 4.987754821777344,
      "step": 3780
    },
    {
      "epoch": 0.9203754162882228,
      "grad_norm": 0.010914390906691551,
      "learning_rate": 4.1523231728045466e-05,
      "loss": 4.858595657348633,
      "step": 3800
    },
    {
      "epoch": 0.9203754162882228,
      "eval_loss": 5.4581074714660645,
      "eval_runtime": 490.4204,
      "eval_samples_per_second": 14.967,
      "eval_steps_per_second": 7.483,
      "step": 3800
    },
    {
      "epoch": 0.9252194974265819,
      "grad_norm": 0.017502611503005028,
      "learning_rate": 4.142281807549644e-05,
      "loss": 4.870289611816406,
      "step": 3820
    },
    {
      "epoch": 0.9300635785649409,
      "grad_norm": 0.017291737720370293,
      "learning_rate": 4.1321936176080755e-05,
      "loss": 4.889853286743164,
      "step": 3840
    },
    {
      "epoch": 0.9349076597033,
      "grad_norm": 0.017816677689552307,
      "learning_rate": 4.122058890613991e-05,
      "loss": 4.910069274902344,
      "step": 3860
    },
    {
      "epoch": 0.9397517408416591,
      "grad_norm": 0.027295643463730812,
      "learning_rate": 4.111877915528406e-05,
      "loss": 5.048255538940429,
      "step": 3880
    },
    {
      "epoch": 0.9445958219800181,
      "grad_norm": 0.035314708948135376,
      "learning_rate": 4.101650982630956e-05,
      "loss": 4.915079498291016,
      "step": 3900
    },
    {
      "epoch": 0.9494399031183772,
      "grad_norm": 0.01960872672498226,
      "learning_rate": 4.091378383511628e-05,
      "loss": 4.810525894165039,
      "step": 3920
    },
    {
      "epoch": 0.9542839842567363,
      "grad_norm": 0.013185371644794941,
      "learning_rate": 4.08106041106244e-05,
      "loss": 4.84496841430664,
      "step": 3940
    },
    {
      "epoch": 0.9591280653950953,
      "grad_norm": 0.020618990063667297,
      "learning_rate": 4.0706973594690964e-05,
      "loss": 4.8005413055419925,
      "step": 3960
    },
    {
      "epoch": 0.9639721465334544,
      "grad_norm": 0.016433637589216232,
      "learning_rate": 4.060289524202594e-05,
      "loss": 4.936908721923828,
      "step": 3980
    },
    {
      "epoch": 0.9688162276718135,
      "grad_norm": 0.020894834771752357,
      "learning_rate": 4.0498372020108e-05,
      "loss": 4.909790802001953,
      "step": 4000
    },
    {
      "epoch": 0.9688162276718135,
      "eval_loss": 5.458165168762207,
      "eval_runtime": 480.7799,
      "eval_samples_per_second": 15.267,
      "eval_steps_per_second": 7.633,
      "step": 4000
    },
    {
      "epoch": 0.9736603088101726,
      "grad_norm": 0.016745753586292267,
      "learning_rate": 4.0393406909099925e-05,
      "loss": 4.82120361328125,
      "step": 4020
    },
    {
      "epoch": 0.9785043899485316,
      "grad_norm": 0.018331704661250114,
      "learning_rate": 4.028800290176362e-05,
      "loss": 4.880381393432617,
      "step": 4040
    },
    {
      "epoch": 0.9833484710868907,
      "grad_norm": 0.023544512689113617,
      "learning_rate": 4.01821630033748e-05,
      "loss": 4.834738922119141,
      "step": 4060
    },
    {
      "epoch": 0.9881925522252498,
      "grad_norm": 0.020899208262562752,
      "learning_rate": 4.007589023163727e-05,
      "loss": 4.852277374267578,
      "step": 4080
    },
    {
      "epoch": 0.9930366333636088,
      "grad_norm": 0.013782848604023457,
      "learning_rate": 3.996918761659694e-05,
      "loss": 4.841725921630859,
      "step": 4100
    },
    {
      "epoch": 0.9978807145019679,
      "grad_norm": 0.018252292647957802,
      "learning_rate": 3.986205820055534e-05,
      "loss": 4.945264434814453,
      "step": 4120
    },
    {
      "epoch": 1.0026642446260974,
      "grad_norm": 0.019553977996110916,
      "learning_rate": 3.9754505037983004e-05,
      "loss": 4.959503555297852,
      "step": 4140
    },
    {
      "epoch": 1.0075083257644566,
      "grad_norm": 0.015189483761787415,
      "learning_rate": 3.964653119543227e-05,
      "loss": 4.9404754638671875,
      "step": 4160
    },
    {
      "epoch": 1.0123524069028156,
      "grad_norm": 0.06231527775526047,
      "learning_rate": 3.953813975144991e-05,
      "loss": 4.842787551879883,
      "step": 4180
    },
    {
      "epoch": 1.0171964880411748,
      "grad_norm": 0.01840846799314022,
      "learning_rate": 3.942933379648932e-05,
      "loss": 4.911764144897461,
      "step": 4200
    },
    {
      "epoch": 1.0171964880411748,
      "eval_loss": 5.460827827453613,
      "eval_runtime": 484.9137,
      "eval_samples_per_second": 15.137,
      "eval_steps_per_second": 7.568,
      "step": 4200
    },
    {
      "epoch": 1.0220405691795338,
      "grad_norm": 0.021854782477021217,
      "learning_rate": 3.932011643282243e-05,
      "loss": 4.853986358642578,
      "step": 4220
    },
    {
      "epoch": 1.0268846503178928,
      "grad_norm": 0.017591115087270737,
      "learning_rate": 3.9210490774451235e-05,
      "loss": 4.8760326385498045,
      "step": 4240
    },
    {
      "epoch": 1.031728731456252,
      "grad_norm": 0.03734872490167618,
      "learning_rate": 3.910045994701903e-05,
      "loss": 4.93592529296875,
      "step": 4260
    },
    {
      "epoch": 1.036572812594611,
      "grad_norm": 0.009241367690265179,
      "learning_rate": 3.899002708772126e-05,
      "loss": 4.926780700683594,
      "step": 4280
    },
    {
      "epoch": 1.04141689373297,
      "grad_norm": 0.01902877353131771,
      "learning_rate": 3.88791953452161e-05,
      "loss": 5.037494277954101,
      "step": 4300
    },
    {
      "epoch": 1.0462609748713292,
      "grad_norm": 0.02036813460290432,
      "learning_rate": 3.876796787953466e-05,
      "loss": 4.970469665527344,
      "step": 4320
    },
    {
      "epoch": 1.0511050560096882,
      "grad_norm": 0.012490912340581417,
      "learning_rate": 3.865634786199092e-05,
      "loss": 4.780771636962891,
      "step": 4340
    },
    {
      "epoch": 1.0559491371480472,
      "grad_norm": 0.020513946190476418,
      "learning_rate": 3.854433847509126e-05,
      "loss": 4.8549247741699215,
      "step": 4360
    },
    {
      "epoch": 1.0607932182864064,
      "grad_norm": 0.018364248797297478,
      "learning_rate": 3.8431942912443754e-05,
      "loss": 4.861563873291016,
      "step": 4380
    },
    {
      "epoch": 1.0656372994247654,
      "grad_norm": 0.014563582837581635,
      "learning_rate": 3.831916437866711e-05,
      "loss": 4.935742568969727,
      "step": 4400
    },
    {
      "epoch": 1.0656372994247654,
      "eval_loss": 5.46129035949707,
      "eval_runtime": 488.8363,
      "eval_samples_per_second": 15.015,
      "eval_steps_per_second": 7.508,
      "step": 4400
    },
    {
      "epoch": 1.0704813805631244,
      "grad_norm": 0.013871117495000362,
      "learning_rate": 3.82060060892993e-05,
      "loss": 4.791946792602539,
      "step": 4420
    },
    {
      "epoch": 1.0753254617014836,
      "grad_norm": 0.016447104513645172,
      "learning_rate": 3.809247127070588e-05,
      "loss": 4.91345443725586,
      "step": 4440
    },
    {
      "epoch": 1.0801695428398426,
      "grad_norm": 0.01683361642062664,
      "learning_rate": 3.797856315998797e-05,
      "loss": 4.959566116333008,
      "step": 4460
    },
    {
      "epoch": 1.0850136239782016,
      "grad_norm": 0.013085919432342052,
      "learning_rate": 3.786428500489001e-05,
      "loss": 4.8983619689941404,
      "step": 4480
    },
    {
      "epoch": 1.0898577051165608,
      "grad_norm": 0.009764457121491432,
      "learning_rate": 3.7749640063707115e-05,
      "loss": 4.951383209228515,
      "step": 4500
    },
    {
      "epoch": 1.0947017862549198,
      "grad_norm": 0.015614672563970089,
      "learning_rate": 3.763463160519222e-05,
      "loss": 4.869015502929687,
      "step": 4520
    },
    {
      "epoch": 1.0995458673932788,
      "grad_norm": 0.024334661662578583,
      "learning_rate": 3.751926290846282e-05,
      "loss": 4.924750137329101,
      "step": 4540
    },
    {
      "epoch": 1.104389948531638,
      "grad_norm": 0.03006790392100811,
      "learning_rate": 3.740353726290753e-05,
      "loss": 4.894016265869141,
      "step": 4560
    },
    {
      "epoch": 1.109234029669997,
      "grad_norm": 0.016674896702170372,
      "learning_rate": 3.7287457968092286e-05,
      "loss": 4.868609237670898,
      "step": 4580
    },
    {
      "epoch": 1.114078110808356,
      "grad_norm": 0.01684091053903103,
      "learning_rate": 3.717102833366624e-05,
      "loss": 4.794527816772461,
      "step": 4600
    },
    {
      "epoch": 1.114078110808356,
      "eval_loss": 5.460610389709473,
      "eval_runtime": 482.5493,
      "eval_samples_per_second": 15.211,
      "eval_steps_per_second": 7.605,
      "step": 4600
    },
    {
      "epoch": 1.1189221919467152,
      "grad_norm": 0.007340766489505768,
      "learning_rate": 3.705425167926743e-05,
      "loss": 4.731179046630859,
      "step": 4620
    },
    {
      "epoch": 1.1237662730850742,
      "grad_norm": 0.014017272740602493,
      "learning_rate": 3.69371313344281e-05,
      "loss": 5.037481307983398,
      "step": 4640
    },
    {
      "epoch": 1.1286103542234334,
      "grad_norm": 0.01410000678151846,
      "learning_rate": 3.681967063847981e-05,
      "loss": 4.686013793945312,
      "step": 4660
    },
    {
      "epoch": 1.1334544353617924,
      "grad_norm": 0.0066877431236207485,
      "learning_rate": 3.6701872940458186e-05,
      "loss": 4.932219696044922,
      "step": 4680
    },
    {
      "epoch": 1.1382985165001513,
      "grad_norm": 0.01702740788459778,
      "learning_rate": 3.658374159900744e-05,
      "loss": 4.854273986816406,
      "step": 4700
    },
    {
      "epoch": 1.1431425976385103,
      "grad_norm": 0.014306023716926575,
      "learning_rate": 3.646527998228464e-05,
      "loss": 4.978992462158203,
      "step": 4720
    },
    {
      "epoch": 1.1479866787768696,
      "grad_norm": 0.013222994282841682,
      "learning_rate": 3.634649146786361e-05,
      "loss": 4.848846817016602,
      "step": 4740
    },
    {
      "epoch": 1.1528307599152285,
      "grad_norm": 0.005920600611716509,
      "learning_rate": 3.622737944263872e-05,
      "loss": 4.912345123291016,
      "step": 4760
    },
    {
      "epoch": 1.1576748410535878,
      "grad_norm": 0.013458259403705597,
      "learning_rate": 3.6107947302728236e-05,
      "loss": 4.936886596679687,
      "step": 4780
    },
    {
      "epoch": 1.1625189221919467,
      "grad_norm": 0.013781394809484482,
      "learning_rate": 3.598819845337752e-05,
      "loss": 4.750154113769531,
      "step": 4800
    },
    {
      "epoch": 1.1625189221919467,
      "eval_loss": 5.460769176483154,
      "eval_runtime": 482.0557,
      "eval_samples_per_second": 15.226,
      "eval_steps_per_second": 7.613,
      "step": 4800
    },
    {
      "epoch": 1.1673630033303057,
      "grad_norm": 0.014957944862544537,
      "learning_rate": 3.5868136308861973e-05,
      "loss": 4.829911804199218,
      "step": 4820
    },
    {
      "epoch": 1.1722070844686647,
      "grad_norm": 0.016118552535772324,
      "learning_rate": 3.5747764292389626e-05,
      "loss": 4.940299606323242,
      "step": 4840
    },
    {
      "epoch": 1.177051165607024,
      "grad_norm": 0.01738092489540577,
      "learning_rate": 3.56270858360036e-05,
      "loss": 4.805704116821289,
      "step": 4860
    },
    {
      "epoch": 1.181895246745383,
      "grad_norm": 0.013412545435130596,
      "learning_rate": 3.550610438048419e-05,
      "loss": 4.848656463623047,
      "step": 4880
    },
    {
      "epoch": 1.1867393278837421,
      "grad_norm": 0.020154351368546486,
      "learning_rate": 3.5384823375250825e-05,
      "loss": 4.836146545410156,
      "step": 4900
    },
    {
      "epoch": 1.1915834090221011,
      "grad_norm": 0.012981275096535683,
      "learning_rate": 3.526324627826366e-05,
      "loss": 4.94176139831543,
      "step": 4920
    },
    {
      "epoch": 1.1964274901604601,
      "grad_norm": 0.02052425593137741,
      "learning_rate": 3.514137655592501e-05,
      "loss": 4.940605545043946,
      "step": 4940
    },
    {
      "epoch": 1.2012715712988193,
      "grad_norm": 0.009783861227333546,
      "learning_rate": 3.501921768298054e-05,
      "loss": 4.762470245361328,
      "step": 4960
    },
    {
      "epoch": 1.2061156524371783,
      "grad_norm": 0.01635117270052433,
      "learning_rate": 3.4896773142420134e-05,
      "loss": 4.848318099975586,
      "step": 4980
    },
    {
      "epoch": 1.2109597335755373,
      "grad_norm": 0.013676351867616177,
      "learning_rate": 3.477404642537865e-05,
      "loss": 4.775520324707031,
      "step": 5000
    },
    {
      "epoch": 1.2109597335755373,
      "eval_loss": 5.459352493286133,
      "eval_runtime": 481.9172,
      "eval_samples_per_second": 15.231,
      "eval_steps_per_second": 7.615,
      "step": 5000
    },
    {
      "epoch": 1.2158038147138965,
      "grad_norm": 0.025599844753742218,
      "learning_rate": 3.465104103103633e-05,
      "loss": 5.030489730834961,
      "step": 5020
    },
    {
      "epoch": 1.2206478958522555,
      "grad_norm": 0.010839181020855904,
      "learning_rate": 3.4527760466519065e-05,
      "loss": 4.919123840332031,
      "step": 5040
    },
    {
      "epoch": 1.2254919769906145,
      "grad_norm": 0.01727098785340786,
      "learning_rate": 3.440420824679841e-05,
      "loss": 4.916360092163086,
      "step": 5060
    },
    {
      "epoch": 1.2303360581289737,
      "grad_norm": 0.015268294140696526,
      "learning_rate": 3.42803878945913e-05,
      "loss": 4.849533462524414,
      "step": 5080
    },
    {
      "epoch": 1.2351801392673327,
      "grad_norm": 0.017013756558299065,
      "learning_rate": 3.4156302940259707e-05,
      "loss": 4.718303680419922,
      "step": 5100
    },
    {
      "epoch": 1.2400242204056917,
      "grad_norm": 0.016048861667513847,
      "learning_rate": 3.403195692170987e-05,
      "loss": 4.887918090820312,
      "step": 5120
    },
    {
      "epoch": 1.244868301544051,
      "grad_norm": 0.018063582479953766,
      "learning_rate": 3.390735338429153e-05,
      "loss": 4.8720966339111325,
      "step": 5140
    },
    {
      "epoch": 1.24971238268241,
      "grad_norm": 0.020126111805438995,
      "learning_rate": 3.378249588069679e-05,
      "loss": 4.978098297119141,
      "step": 5160
    },
    {
      "epoch": 1.254556463820769,
      "grad_norm": 0.013375598937273026,
      "learning_rate": 3.365738797085881e-05,
      "loss": 4.952982330322266,
      "step": 5180
    },
    {
      "epoch": 1.2594005449591281,
      "grad_norm": 0.015132354572415352,
      "learning_rate": 3.353203322185035e-05,
      "loss": 4.908714294433594,
      "step": 5200
    },
    {
      "epoch": 1.2594005449591281,
      "eval_loss": 5.45740270614624,
      "eval_runtime": 481.4693,
      "eval_samples_per_second": 15.245,
      "eval_steps_per_second": 7.623,
      "step": 5200
    },
    {
      "epoch": 1.2642446260974871,
      "grad_norm": 0.012495392933487892,
      "learning_rate": 3.340643520778201e-05,
      "loss": 4.851836776733398,
      "step": 5220
    },
    {
      "epoch": 1.2690887072358463,
      "grad_norm": 0.01966176927089691,
      "learning_rate": 3.328059750970038e-05,
      "loss": 4.94261474609375,
      "step": 5240
    },
    {
      "epoch": 1.2739327883742053,
      "grad_norm": 0.022419588640332222,
      "learning_rate": 3.315452371548592e-05,
      "loss": 4.894690704345703,
      "step": 5260
    },
    {
      "epoch": 1.2787768695125643,
      "grad_norm": 0.008414322510361671,
      "learning_rate": 3.302821741975061e-05,
      "loss": 4.914752197265625,
      "step": 5280
    },
    {
      "epoch": 1.2836209506509233,
      "grad_norm": 0.014880464412271976,
      "learning_rate": 3.290168222373557e-05,
      "loss": 4.73625602722168,
      "step": 5300
    },
    {
      "epoch": 1.2884650317892825,
      "grad_norm": 0.012800327502191067,
      "learning_rate": 3.2774921735208255e-05,
      "loss": 4.73109245300293,
      "step": 5320
    },
    {
      "epoch": 1.2933091129276415,
      "grad_norm": 0.0207191351801157,
      "learning_rate": 3.26479395683597e-05,
      "loss": 4.839684295654297,
      "step": 5340
    },
    {
      "epoch": 1.2981531940660007,
      "grad_norm": 0.016082677990198135,
      "learning_rate": 3.2520739343701415e-05,
      "loss": 4.872611999511719,
      "step": 5360
    },
    {
      "epoch": 1.3029972752043597,
      "grad_norm": 0.012835017405450344,
      "learning_rate": 3.2393324687962145e-05,
      "loss": 4.883394241333008,
      "step": 5380
    },
    {
      "epoch": 1.3078413563427187,
      "grad_norm": 0.01141536794602871,
      "learning_rate": 3.2265699233984506e-05,
      "loss": 4.8918407440185545,
      "step": 5400
    },
    {
      "epoch": 1.3078413563427187,
      "eval_loss": 5.4619011878967285,
      "eval_runtime": 484.2168,
      "eval_samples_per_second": 15.158,
      "eval_steps_per_second": 7.579,
      "step": 5400
    },
    {
      "epoch": 1.3126854374810777,
      "grad_norm": 0.011928900144994259,
      "learning_rate": 3.213786662062137e-05,
      "loss": 4.9856517791748045,
      "step": 5420
    },
    {
      "epoch": 1.317529518619437,
      "grad_norm": 0.027329420670866966,
      "learning_rate": 3.2009830492632175e-05,
      "loss": 4.990286254882813,
      "step": 5440
    },
    {
      "epoch": 1.322373599757796,
      "grad_norm": 0.014349824748933315,
      "learning_rate": 3.18815945005789e-05,
      "loss": 4.962379837036133,
      "step": 5460
    },
    {
      "epoch": 1.327217680896155,
      "grad_norm": 0.0324920155107975,
      "learning_rate": 3.1753162300722096e-05,
      "loss": 4.6757560729980465,
      "step": 5480
    },
    {
      "epoch": 1.332061762034514,
      "grad_norm": 0.014296727254986763,
      "learning_rate": 3.162453755491655e-05,
      "loss": 4.982074356079101,
      "step": 5500
    },
    {
      "epoch": 1.336905843172873,
      "grad_norm": 0.03129876032471657,
      "learning_rate": 3.149572393050691e-05,
      "loss": 5.023649215698242,
      "step": 5520
    },
    {
      "epoch": 1.341749924311232,
      "grad_norm": 0.02015358954668045,
      "learning_rate": 3.136672510022316e-05,
      "loss": 4.832314300537109,
      "step": 5540
    },
    {
      "epoch": 1.3465940054495913,
      "grad_norm": 0.010978948324918747,
      "learning_rate": 3.123754474207583e-05,
      "loss": 4.861849594116211,
      "step": 5560
    },
    {
      "epoch": 1.3514380865879503,
      "grad_norm": 0.028877103701233864,
      "learning_rate": 3.1108186539251196e-05,
      "loss": 4.910951232910156,
      "step": 5580
    },
    {
      "epoch": 1.3562821677263095,
      "grad_norm": 0.013014769181609154,
      "learning_rate": 3.0978654180006214e-05,
      "loss": 4.845300674438477,
      "step": 5600
    },
    {
      "epoch": 1.3562821677263095,
      "eval_loss": 5.460139751434326,
      "eval_runtime": 490.1706,
      "eval_samples_per_second": 14.974,
      "eval_steps_per_second": 7.487,
      "step": 5600
    },
    {
      "epoch": 1.3611262488646685,
      "grad_norm": 0.014915244653820992,
      "learning_rate": 3.08489513575634e-05,
      "loss": 4.975334548950196,
      "step": 5620
    },
    {
      "epoch": 1.3659703300030275,
      "grad_norm": 0.009035931900143623,
      "learning_rate": 3.0719081770005496e-05,
      "loss": 4.825616455078125,
      "step": 5640
    },
    {
      "epoch": 1.3708144111413867,
      "grad_norm": 0.01582302153110504,
      "learning_rate": 3.058904912017006e-05,
      "loss": 4.912509536743164,
      "step": 5660
    },
    {
      "epoch": 1.3756584922797457,
      "grad_norm": 0.017856527119874954,
      "learning_rate": 3.0458857115543856e-05,
      "loss": 4.868062973022461,
      "step": 5680
    },
    {
      "epoch": 1.3805025734181047,
      "grad_norm": 0.01170275080949068,
      "learning_rate": 3.0328509468157177e-05,
      "loss": 4.956634521484375,
      "step": 5700
    },
    {
      "epoch": 1.3853466545564639,
      "grad_norm": 0.025918355211615562,
      "learning_rate": 3.0198009894478013e-05,
      "loss": 4.850352478027344,
      "step": 5720
    },
    {
      "epoch": 1.3901907356948229,
      "grad_norm": 0.022060878574848175,
      "learning_rate": 3.006736211530606e-05,
      "loss": 4.954733657836914,
      "step": 5740
    },
    {
      "epoch": 1.3950348168331819,
      "grad_norm": 0.010089868679642677,
      "learning_rate": 2.993656985566663e-05,
      "loss": 4.827342987060547,
      "step": 5760
    },
    {
      "epoch": 1.399878897971541,
      "grad_norm": 0.011502007022500038,
      "learning_rate": 2.9805636844704477e-05,
      "loss": 4.790507507324219,
      "step": 5780
    },
    {
      "epoch": 1.4047229791099,
      "grad_norm": 0.01932416670024395,
      "learning_rate": 2.9674566815577443e-05,
      "loss": 4.710316467285156,
      "step": 5800
    },
    {
      "epoch": 1.4047229791099,
      "eval_loss": 5.4579668045043945,
      "eval_runtime": 483.7605,
      "eval_samples_per_second": 15.173,
      "eval_steps_per_second": 7.586,
      "step": 5800
    },
    {
      "epoch": 1.4095670602482593,
      "grad_norm": 0.020764078944921494,
      "learning_rate": 2.954336350535003e-05,
      "loss": 4.882500839233399,
      "step": 5820
    },
    {
      "epoch": 1.4144111413866183,
      "grad_norm": 0.01630806177854538,
      "learning_rate": 2.9412030654886858e-05,
      "loss": 4.921166610717774,
      "step": 5840
    },
    {
      "epoch": 1.4192552225249773,
      "grad_norm": 0.012430929578840733,
      "learning_rate": 2.9280572008745966e-05,
      "loss": 4.942293548583985,
      "step": 5860
    },
    {
      "epoch": 1.4240993036633363,
      "grad_norm": 0.01300967950373888,
      "learning_rate": 2.91489913150721e-05,
      "loss": 4.8684028625488285,
      "step": 5880
    },
    {
      "epoch": 1.4289433848016955,
      "grad_norm": 0.017518101260066032,
      "learning_rate": 2.901729232548981e-05,
      "loss": 4.943297958374023,
      "step": 5900
    },
    {
      "epoch": 1.4337874659400545,
      "grad_norm": 0.02635425142943859,
      "learning_rate": 2.8885478794996494e-05,
      "loss": 4.939146041870117,
      "step": 5920
    },
    {
      "epoch": 1.4386315470784137,
      "grad_norm": 0.01673278398811817,
      "learning_rate": 2.875355448185536e-05,
      "loss": 4.740990447998047,
      "step": 5940
    },
    {
      "epoch": 1.4434756282167727,
      "grad_norm": 0.013996589928865433,
      "learning_rate": 2.8621523147488206e-05,
      "loss": 4.899109649658203,
      "step": 5960
    },
    {
      "epoch": 1.4483197093551317,
      "grad_norm": 0.012060189619660378,
      "learning_rate": 2.8489388556368258e-05,
      "loss": 4.813941574096679,
      "step": 5980
    },
    {
      "epoch": 1.4531637904934906,
      "grad_norm": 0.01745065301656723,
      "learning_rate": 2.8357154475912757e-05,
      "loss": 4.8057598114013675,
      "step": 6000
    },
    {
      "epoch": 1.4531637904934906,
      "eval_loss": 5.459037780761719,
      "eval_runtime": 481.1772,
      "eval_samples_per_second": 15.254,
      "eval_steps_per_second": 7.627,
      "step": 6000
    },
    {
      "epoch": 1.4580078716318499,
      "grad_norm": 0.015925351530313492,
      "learning_rate": 2.8224824676375612e-05,
      "loss": 4.829607772827148,
      "step": 6020
    },
    {
      "epoch": 1.4628519527702089,
      "grad_norm": 0.028525961562991142,
      "learning_rate": 2.8092402930739854e-05,
      "loss": 4.844269943237305,
      "step": 6040
    },
    {
      "epoch": 1.467696033908568,
      "grad_norm": 0.008764233440160751,
      "learning_rate": 2.795989301461009e-05,
      "loss": 4.854158782958985,
      "step": 6060
    },
    {
      "epoch": 1.472540115046927,
      "grad_norm": 0.02136807143688202,
      "learning_rate": 2.7827298706104827e-05,
      "loss": 4.853776550292968,
      "step": 6080
    },
    {
      "epoch": 1.477384196185286,
      "grad_norm": 0.01663500815629959,
      "learning_rate": 2.769462378574877e-05,
      "loss": 4.822975921630859,
      "step": 6100
    },
    {
      "epoch": 1.482228277323645,
      "grad_norm": 0.016130829229950905,
      "learning_rate": 2.7561872036365033e-05,
      "loss": 5.028213119506836,
      "step": 6120
    },
    {
      "epoch": 1.4870723584620043,
      "grad_norm": 0.023693716153502464,
      "learning_rate": 2.742904724296727e-05,
      "loss": 4.8783924102783205,
      "step": 6140
    },
    {
      "epoch": 1.4919164396003632,
      "grad_norm": 0.015705805271863937,
      "learning_rate": 2.7296153192651775e-05,
      "loss": 4.831390762329102,
      "step": 6160
    },
    {
      "epoch": 1.4967605207387225,
      "grad_norm": 0.024449199438095093,
      "learning_rate": 2.7163193674489475e-05,
      "loss": 4.918373107910156,
      "step": 6180
    },
    {
      "epoch": 1.5016046018770814,
      "grad_norm": 0.01422477513551712,
      "learning_rate": 2.7030172479417925e-05,
      "loss": 5.008320236206055,
      "step": 6200
    },
    {
      "epoch": 1.5016046018770814,
      "eval_loss": 5.461097717285156,
      "eval_runtime": 481.5246,
      "eval_samples_per_second": 15.243,
      "eval_steps_per_second": 7.622,
      "step": 6200
    }
  ],
  "logging_steps": 20,
  "max_steps": 12387,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 200,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.1428451608248975e+18,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}