{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.996845425867508,
  "eval_steps": 500,
  "global_step": 1108,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.018026137899954935,
      "grad_norm": 16.059541702270508,
      "learning_rate": 4.9548736462093865e-05,
      "loss": 0.6728,
      "step": 10
    },
    {
      "epoch": 0.03605227579990987,
      "grad_norm": 22.368194580078125,
      "learning_rate": 4.909747292418773e-05,
      "loss": 0.5752,
      "step": 20
    },
    {
      "epoch": 0.054078413699864804,
      "grad_norm": 21.088069915771484,
      "learning_rate": 4.864620938628159e-05,
      "loss": 0.5414,
      "step": 30
    },
    {
      "epoch": 0.07210455159981974,
      "grad_norm": 17.759164810180664,
      "learning_rate": 4.819494584837546e-05,
      "loss": 0.5483,
      "step": 40
    },
    {
      "epoch": 0.09013068949977468,
      "grad_norm": 35.45125961303711,
      "learning_rate": 4.7743682310469314e-05,
      "loss": 0.5284,
      "step": 50
    },
    {
      "epoch": 0.10815682739972961,
      "grad_norm": 34.69957733154297,
      "learning_rate": 4.7292418772563177e-05,
      "loss": 0.4792,
      "step": 60
    },
    {
      "epoch": 0.12618296529968454,
      "grad_norm": 31.315162658691406,
      "learning_rate": 4.684115523465704e-05,
      "loss": 0.505,
      "step": 70
    },
    {
      "epoch": 0.14420910319963948,
      "grad_norm": 28.01712417602539,
      "learning_rate": 4.63898916967509e-05,
      "loss": 0.4575,
      "step": 80
    },
    {
      "epoch": 0.16223524109959442,
      "grad_norm": 22.80365562438965,
      "learning_rate": 4.5938628158844764e-05,
      "loss": 0.4678,
      "step": 90
    },
    {
      "epoch": 0.18026137899954936,
      "grad_norm": 38.55846405029297,
      "learning_rate": 4.548736462093863e-05,
      "loss": 0.4724,
      "step": 100
    },
    {
      "epoch": 0.19828751689950427,
      "grad_norm": 29.91495704650879,
      "learning_rate": 4.5036101083032495e-05,
      "loss": 0.4891,
      "step": 110
    },
    {
      "epoch": 0.21631365479945922,
      "grad_norm": 26.530668258666992,
      "learning_rate": 4.458483754512636e-05,
      "loss": 0.4578,
      "step": 120
    },
    {
      "epoch": 0.23433979269941416,
      "grad_norm": 53.58538055419922,
      "learning_rate": 4.413357400722022e-05,
      "loss": 0.4735,
      "step": 130
    },
    {
      "epoch": 0.25236593059936907,
      "grad_norm": 22.026208877563477,
      "learning_rate": 4.368231046931408e-05,
      "loss": 0.4743,
      "step": 140
    },
    {
      "epoch": 0.270392068499324,
      "grad_norm": 28.057222366333008,
      "learning_rate": 4.3231046931407945e-05,
      "loss": 0.4559,
      "step": 150
    },
    {
      "epoch": 0.28841820639927895,
      "grad_norm": 22.690412521362305,
      "learning_rate": 4.277978339350181e-05,
      "loss": 0.4638,
      "step": 160
    },
    {
      "epoch": 0.3064443442992339,
      "grad_norm": 21.292953491210938,
      "learning_rate": 4.232851985559567e-05,
      "loss": 0.4877,
      "step": 170
    },
    {
      "epoch": 0.32447048219918884,
      "grad_norm": 36.56389236450195,
      "learning_rate": 4.187725631768953e-05,
      "loss": 0.4927,
      "step": 180
    },
    {
      "epoch": 0.3424966200991438,
      "grad_norm": 17.6762752532959,
      "learning_rate": 4.1425992779783394e-05,
      "loss": 0.491,
      "step": 190
    },
    {
      "epoch": 0.3605227579990987,
      "grad_norm": 19.962385177612305,
      "learning_rate": 4.0974729241877256e-05,
      "loss": 0.4428,
      "step": 200
    },
    {
      "epoch": 0.3785488958990536,
      "grad_norm": 23.71368408203125,
      "learning_rate": 4.052346570397112e-05,
      "loss": 0.4318,
      "step": 210
    },
    {
      "epoch": 0.39657503379900855,
      "grad_norm": 24.030736923217773,
      "learning_rate": 4.007220216606498e-05,
      "loss": 0.4532,
      "step": 220
    },
    {
      "epoch": 0.4146011716989635,
      "grad_norm": 23.85675811767578,
      "learning_rate": 3.962093862815885e-05,
      "loss": 0.4427,
      "step": 230
    },
    {
      "epoch": 0.43262730959891843,
      "grad_norm": 32.091190338134766,
      "learning_rate": 3.916967509025271e-05,
      "loss": 0.4432,
      "step": 240
    },
    {
      "epoch": 0.45065344749887337,
      "grad_norm": 32.51976776123047,
      "learning_rate": 3.8718411552346575e-05,
      "loss": 0.4289,
      "step": 250
    },
    {
      "epoch": 0.4686795853988283,
      "grad_norm": 17.93523406982422,
      "learning_rate": 3.826714801444044e-05,
      "loss": 0.4485,
      "step": 260
    },
    {
      "epoch": 0.48670572329878325,
      "grad_norm": 34.29135513305664,
      "learning_rate": 3.78158844765343e-05,
      "loss": 0.4373,
      "step": 270
    },
    {
      "epoch": 0.5047318611987381,
      "grad_norm": 21.879440307617188,
      "learning_rate": 3.7364620938628155e-05,
      "loss": 0.4367,
      "step": 280
    },
    {
      "epoch": 0.5227579990986931,
      "grad_norm": 16.587345123291016,
      "learning_rate": 3.6913357400722025e-05,
      "loss": 0.4476,
      "step": 290
    },
    {
      "epoch": 0.540784136998648,
      "grad_norm": 21.923376083374023,
      "learning_rate": 3.646209386281589e-05,
      "loss": 0.4186,
      "step": 300
    },
    {
      "epoch": 0.558810274898603,
      "grad_norm": 25.472400665283203,
      "learning_rate": 3.601083032490975e-05,
      "loss": 0.4317,
      "step": 310
    },
    {
      "epoch": 0.5768364127985579,
      "grad_norm": 36.07789611816406,
      "learning_rate": 3.555956678700361e-05,
      "loss": 0.4281,
      "step": 320
    },
    {
      "epoch": 0.5948625506985128,
      "grad_norm": 32.51295852661133,
      "learning_rate": 3.5108303249097474e-05,
      "loss": 0.4391,
      "step": 330
    },
    {
      "epoch": 0.6128886885984678,
      "grad_norm": 19.069826126098633,
      "learning_rate": 3.4657039711191336e-05,
      "loss": 0.4249,
      "step": 340
    },
    {
      "epoch": 0.6309148264984227,
      "grad_norm": 22.865673065185547,
      "learning_rate": 3.42057761732852e-05,
      "loss": 0.4394,
      "step": 350
    },
    {
      "epoch": 0.6489409643983777,
      "grad_norm": 35.98612594604492,
      "learning_rate": 3.375451263537907e-05,
      "loss": 0.407,
      "step": 360
    },
    {
      "epoch": 0.6669671022983326,
      "grad_norm": 19.89366340637207,
      "learning_rate": 3.330324909747293e-05,
      "loss": 0.434,
      "step": 370
    },
    {
      "epoch": 0.6849932401982876,
      "grad_norm": 20.73752212524414,
      "learning_rate": 3.2851985559566786e-05,
      "loss": 0.3815,
      "step": 380
    },
    {
      "epoch": 0.7030193780982424,
      "grad_norm": 20.23769187927246,
      "learning_rate": 3.240072202166065e-05,
      "loss": 0.4159,
      "step": 390
    },
    {
      "epoch": 0.7210455159981974,
      "grad_norm": 40.59239196777344,
      "learning_rate": 3.194945848375451e-05,
      "loss": 0.4582,
      "step": 400
    },
    {
      "epoch": 0.7390716538981523,
      "grad_norm": 22.728296279907227,
      "learning_rate": 3.149819494584837e-05,
      "loss": 0.4387,
      "step": 410
    },
    {
      "epoch": 0.7570977917981072,
      "grad_norm": 20.43068504333496,
      "learning_rate": 3.104693140794224e-05,
      "loss": 0.4302,
      "step": 420
    },
    {
      "epoch": 0.7751239296980622,
      "grad_norm": 30.140209197998047,
      "learning_rate": 3.0595667870036104e-05,
      "loss": 0.3858,
      "step": 430
    },
    {
      "epoch": 0.7931500675980171,
      "grad_norm": 15.338220596313477,
      "learning_rate": 3.0144404332129967e-05,
      "loss": 0.4634,
      "step": 440
    },
    {
      "epoch": 0.8111762054979721,
      "grad_norm": 24.517107009887695,
      "learning_rate": 2.969314079422383e-05,
      "loss": 0.4079,
      "step": 450
    },
    {
      "epoch": 0.829202343397927,
      "grad_norm": 25.604442596435547,
      "learning_rate": 2.924187725631769e-05,
      "loss": 0.3903,
      "step": 460
    },
    {
      "epoch": 0.847228481297882,
      "grad_norm": 26.43514060974121,
      "learning_rate": 2.879061371841155e-05,
      "loss": 0.4037,
      "step": 470
    },
    {
      "epoch": 0.8652546191978369,
      "grad_norm": 18.30228042602539,
      "learning_rate": 2.8339350180505413e-05,
      "loss": 0.4374,
      "step": 480
    },
    {
      "epoch": 0.8832807570977917,
      "grad_norm": 16.212318420410156,
      "learning_rate": 2.7888086642599282e-05,
      "loss": 0.4219,
      "step": 490
    },
    {
      "epoch": 0.9013068949977467,
      "grad_norm": 23.251705169677734,
      "learning_rate": 2.7436823104693144e-05,
      "loss": 0.3784,
      "step": 500
    },
    {
      "epoch": 0.9193330328977016,
      "grad_norm": 37.26618194580078,
      "learning_rate": 2.6985559566787007e-05,
      "loss": 0.4144,
      "step": 510
    },
    {
      "epoch": 0.9373591707976566,
      "grad_norm": 22.583784103393555,
      "learning_rate": 2.6534296028880866e-05,
      "loss": 0.4123,
      "step": 520
    },
    {
      "epoch": 0.9553853086976115,
      "grad_norm": 18.812397003173828,
      "learning_rate": 2.6083032490974728e-05,
      "loss": 0.422,
      "step": 530
    },
    {
      "epoch": 0.9734114465975665,
      "grad_norm": 20.292518615722656,
      "learning_rate": 2.563176895306859e-05,
      "loss": 0.4346,
      "step": 540
    },
    {
      "epoch": 0.9914375844975214,
      "grad_norm": 25.314495086669922,
      "learning_rate": 2.518050541516246e-05,
      "loss": 0.393,
      "step": 550
    },
    {
      "epoch": 1.0,
      "eval_f1": 0.8425531914893617,
      "eval_fn": 307,
      "eval_fp": 433,
      "eval_loss": 0.36401602625846863,
      "eval_precision": 0.8205553253211769,
      "eval_recall": 0.8657630083078268,
      "eval_runtime": 13.0048,
      "eval_samples_per_second": 359.329,
      "eval_steps_per_second": 44.983,
      "eval_tn": 1953,
      "eval_tp": 1980,
      "step": 555
    },
    {
      "epoch": 1.0090130689499774,
      "grad_norm": 17.186704635620117,
      "learning_rate": 2.472924187725632e-05,
      "loss": 0.362,
      "step": 560
    },
    {
      "epoch": 1.0270392068499323,
      "grad_norm": 18.968002319335938,
      "learning_rate": 2.427797833935018e-05,
      "loss": 0.3115,
      "step": 570
    },
    {
      "epoch": 1.0450653447498874,
      "grad_norm": 20.980627059936523,
      "learning_rate": 2.3826714801444043e-05,
      "loss": 0.3436,
      "step": 580
    },
    {
      "epoch": 1.0630914826498423,
      "grad_norm": 26.233989715576172,
      "learning_rate": 2.3375451263537906e-05,
      "loss": 0.3174,
      "step": 590
    },
    {
      "epoch": 1.0811176205497972,
      "grad_norm": 18.767547607421875,
      "learning_rate": 2.292418772563177e-05,
      "loss": 0.3425,
      "step": 600
    },
    {
      "epoch": 1.099143758449752,
      "grad_norm": 24.689531326293945,
      "learning_rate": 2.2472924187725634e-05,
      "loss": 0.3261,
      "step": 610
    },
    {
      "epoch": 1.117169896349707,
      "grad_norm": 27.74945068359375,
      "learning_rate": 2.2021660649819496e-05,
      "loss": 0.3242,
      "step": 620
    },
    {
      "epoch": 1.135196034249662,
      "grad_norm": 24.07059669494629,
      "learning_rate": 2.157039711191336e-05,
      "loss": 0.2988,
      "step": 630
    },
    {
      "epoch": 1.153222172149617,
      "grad_norm": 22.548677444458008,
      "learning_rate": 2.111913357400722e-05,
      "loss": 0.3161,
      "step": 640
    },
    {
      "epoch": 1.1712483100495719,
      "grad_norm": 24.466567993164062,
      "learning_rate": 2.0667870036101083e-05,
      "loss": 0.253,
      "step": 650
    },
    {
      "epoch": 1.1892744479495267,
      "grad_norm": 22.701765060424805,
      "learning_rate": 2.0216606498194946e-05,
      "loss": 0.32,
      "step": 660
    },
    {
      "epoch": 1.2073005858494819,
      "grad_norm": 27.035432815551758,
      "learning_rate": 1.976534296028881e-05,
      "loss": 0.3208,
      "step": 670
    },
    {
      "epoch": 1.2253267237494367,
      "grad_norm": 23.99417495727539,
      "learning_rate": 1.9314079422382674e-05,
      "loss": 0.2686,
      "step": 680
    },
    {
      "epoch": 1.2433528616493916,
      "grad_norm": 34.936580657958984,
      "learning_rate": 1.8862815884476533e-05,
      "loss": 0.3097,
      "step": 690
    },
    {
      "epoch": 1.2613789995493465,
      "grad_norm": 59.716670989990234,
      "learning_rate": 1.84115523465704e-05,
      "loss": 0.3137,
      "step": 700
    },
    {
      "epoch": 1.2794051374493014,
      "grad_norm": 29.52703285217285,
      "learning_rate": 1.796028880866426e-05,
      "loss": 0.286,
      "step": 710
    },
    {
      "epoch": 1.2974312753492563,
      "grad_norm": 35.39006423950195,
      "learning_rate": 1.7509025270758123e-05,
      "loss": 0.3122,
      "step": 720
    },
    {
      "epoch": 1.3154574132492114,
      "grad_norm": 39.63450622558594,
      "learning_rate": 1.7057761732851986e-05,
      "loss": 0.3031,
      "step": 730
    },
    {
      "epoch": 1.3334835511491663,
      "grad_norm": 36.622711181640625,
      "learning_rate": 1.6606498194945848e-05,
      "loss": 0.3165,
      "step": 740
    },
    {
      "epoch": 1.3515096890491212,
      "grad_norm": 21.856231689453125,
      "learning_rate": 1.615523465703971e-05,
      "loss": 0.3052,
      "step": 750
    },
    {
      "epoch": 1.3695358269490763,
      "grad_norm": 32.4267578125,
      "learning_rate": 1.5703971119133576e-05,
      "loss": 0.2838,
      "step": 760
    },
    {
      "epoch": 1.3875619648490312,
      "grad_norm": 26.172542572021484,
      "learning_rate": 1.5252707581227438e-05,
      "loss": 0.3322,
      "step": 770
    },
    {
      "epoch": 1.405588102748986,
      "grad_norm": 30.22015380859375,
      "learning_rate": 1.4801444043321299e-05,
      "loss": 0.3202,
      "step": 780
    },
    {
      "epoch": 1.423614240648941,
      "grad_norm": 20.275634765625,
      "learning_rate": 1.4350180505415165e-05,
      "loss": 0.3124,
      "step": 790
    },
    {
      "epoch": 1.4416403785488958,
      "grad_norm": 28.524124145507812,
      "learning_rate": 1.3898916967509026e-05,
      "loss": 0.2777,
      "step": 800
    },
    {
      "epoch": 1.4596665164488507,
      "grad_norm": 25.33469581604004,
      "learning_rate": 1.3447653429602888e-05,
      "loss": 0.2701,
      "step": 810
    },
    {
      "epoch": 1.4776926543488058,
      "grad_norm": 29.28204917907715,
      "learning_rate": 1.299638989169675e-05,
      "loss": 0.3033,
      "step": 820
    },
    {
      "epoch": 1.4957187922487607,
      "grad_norm": 48.7986946105957,
      "learning_rate": 1.2545126353790614e-05,
      "loss": 0.3141,
      "step": 830
    },
    {
      "epoch": 1.5137449301487156,
      "grad_norm": 21.521024703979492,
      "learning_rate": 1.2093862815884477e-05,
      "loss": 0.3042,
      "step": 840
    },
    {
      "epoch": 1.5317710680486707,
      "grad_norm": 23.701133728027344,
      "learning_rate": 1.164259927797834e-05,
      "loss": 0.3026,
      "step": 850
    },
    {
      "epoch": 1.5497972059486256,
      "grad_norm": 28.199644088745117,
      "learning_rate": 1.1191335740072201e-05,
      "loss": 0.3193,
      "step": 860
    },
    {
      "epoch": 1.5678233438485805,
      "grad_norm": 38.08968734741211,
      "learning_rate": 1.0740072202166066e-05,
      "loss": 0.3152,
      "step": 870
    },
    {
      "epoch": 1.5858494817485354,
      "grad_norm": 31.959123611450195,
      "learning_rate": 1.028880866425993e-05,
      "loss": 0.2985,
      "step": 880
    },
    {
      "epoch": 1.6038756196484902,
      "grad_norm": 19.682336807250977,
      "learning_rate": 9.83754512635379e-06,
      "loss": 0.2885,
      "step": 890
    },
    {
      "epoch": 1.6219017575484451,
      "grad_norm": 31.91750144958496,
      "learning_rate": 9.386281588447654e-06,
      "loss": 0.3419,
      "step": 900
    },
    {
      "epoch": 1.6399278954484002,
      "grad_norm": 22.43979835510254,
      "learning_rate": 8.935018050541517e-06,
      "loss": 0.3064,
      "step": 910
    },
    {
      "epoch": 1.6579540333483551,
      "grad_norm": 28.073787689208984,
      "learning_rate": 8.483754512635379e-06,
      "loss": 0.2712,
      "step": 920
    },
    {
      "epoch": 1.67598017124831,
      "grad_norm": 20.845012664794922,
      "learning_rate": 8.032490974729243e-06,
      "loss": 0.3087,
      "step": 930
    },
    {
      "epoch": 1.694006309148265,
      "grad_norm": 31.37566375732422,
      "learning_rate": 7.581227436823105e-06,
      "loss": 0.3132,
      "step": 940
    },
    {
      "epoch": 1.71203244704822,
      "grad_norm": 32.558406829833984,
      "learning_rate": 7.129963898916968e-06,
      "loss": 0.3196,
      "step": 950
    },
    {
      "epoch": 1.7300585849481749,
      "grad_norm": 24.960054397583008,
      "learning_rate": 6.678700361010831e-06,
      "loss": 0.2834,
      "step": 960
    },
    {
      "epoch": 1.7480847228481298,
      "grad_norm": 33.798545837402344,
      "learning_rate": 6.227436823104693e-06,
      "loss": 0.3007,
      "step": 970
    },
    {
      "epoch": 1.7661108607480847,
      "grad_norm": 26.20290184020996,
      "learning_rate": 5.776173285198557e-06,
      "loss": 0.3076,
      "step": 980
    },
    {
      "epoch": 1.7841369986480395,
      "grad_norm": 30.72686004638672,
      "learning_rate": 5.324909747292419e-06,
      "loss": 0.3332,
      "step": 990
    },
    {
      "epoch": 1.8021631365479944,
      "grad_norm": 26.3869686126709,
      "learning_rate": 4.873646209386281e-06,
      "loss": 0.2907,
      "step": 1000
    },
    {
      "epoch": 1.8201892744479495,
      "grad_norm": 30.819374084472656,
      "learning_rate": 4.422382671480145e-06,
      "loss": 0.2888,
      "step": 1010
    },
    {
      "epoch": 1.8382154123479044,
      "grad_norm": 22.17605972290039,
      "learning_rate": 3.971119133574008e-06,
      "loss": 0.2945,
      "step": 1020
    },
    {
      "epoch": 1.8562415502478595,
      "grad_norm": 28.60884666442871,
      "learning_rate": 3.51985559566787e-06,
      "loss": 0.2902,
      "step": 1030
    },
    {
      "epoch": 1.8742676881478144,
      "grad_norm": 34.39057540893555,
      "learning_rate": 3.068592057761733e-06,
      "loss": 0.3129,
      "step": 1040
    },
    {
      "epoch": 1.8922938260477693,
      "grad_norm": 33.24583053588867,
      "learning_rate": 2.6173285198555957e-06,
      "loss": 0.2921,
      "step": 1050
    },
    {
      "epoch": 1.9103199639477242,
      "grad_norm": 22.568416595458984,
      "learning_rate": 2.1660649819494585e-06,
      "loss": 0.2668,
      "step": 1060
    },
    {
      "epoch": 1.928346101847679,
      "grad_norm": 24.128253936767578,
      "learning_rate": 1.7148014440433215e-06,
      "loss": 0.2804,
      "step": 1070
    },
    {
      "epoch": 1.946372239747634,
      "grad_norm": 20.853330612182617,
      "learning_rate": 1.263537906137184e-06,
      "loss": 0.3116,
      "step": 1080
    },
    {
      "epoch": 1.9643983776475888,
      "grad_norm": 29.551576614379883,
      "learning_rate": 8.122743682310469e-07,
      "loss": 0.2975,
      "step": 1090
    },
    {
      "epoch": 1.982424515547544,
      "grad_norm": 30.593189239501953,
      "learning_rate": 3.610108303249098e-07,
      "loss": 0.2876,
      "step": 1100
    },
    {
      "epoch": 1.996845425867508,
      "eval_f1": 0.8350383631713555,
      "eval_fn": 328,
      "eval_fp": 446,
      "eval_loss": 0.3849656879901886,
      "eval_precision": 0.8145530145530145,
      "eval_recall": 0.8565806733712287,
      "eval_runtime": 18.4483,
      "eval_samples_per_second": 253.303,
      "eval_steps_per_second": 31.71,
      "eval_tn": 1940,
      "eval_tp": 1959,
      "step": 1108
    }
  ],
  "logging_steps": 10,
  "max_steps": 1108,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 9326760690401280.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}