{
  "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": 2.7659523487091064,
      "learning_rate": 4.9548736462093865e-05,
      "loss": 0.1292,
      "step": 10
    },
    {
      "epoch": 0.03605227579990987,
      "grad_norm": 1.665235161781311,
      "learning_rate": 4.909747292418773e-05,
      "loss": 0.0471,
      "step": 20
    },
    {
      "epoch": 0.054078413699864804,
      "grad_norm": 1.1857678890228271,
      "learning_rate": 4.864620938628159e-05,
      "loss": 0.0522,
      "step": 30
    },
    {
      "epoch": 0.07210455159981974,
      "grad_norm": 10.890019416809082,
      "learning_rate": 4.819494584837546e-05,
      "loss": 0.0968,
      "step": 40
    },
    {
      "epoch": 0.09013068949977468,
      "grad_norm": 7.386863708496094,
      "learning_rate": 4.7743682310469314e-05,
      "loss": 0.0392,
      "step": 50
    },
    {
      "epoch": 0.10815682739972961,
      "grad_norm": 8.243688583374023,
      "learning_rate": 4.7292418772563177e-05,
      "loss": 0.0603,
      "step": 60
    },
    {
      "epoch": 0.12618296529968454,
      "grad_norm": 0.6079480051994324,
      "learning_rate": 4.684115523465704e-05,
      "loss": 0.0417,
      "step": 70
    },
    {
      "epoch": 0.14420910319963948,
      "grad_norm": 3.4293453693389893,
      "learning_rate": 4.63898916967509e-05,
      "loss": 0.0479,
      "step": 80
    },
    {
      "epoch": 0.16223524109959442,
      "grad_norm": 0.9851968288421631,
      "learning_rate": 4.5938628158844764e-05,
      "loss": 0.0836,
      "step": 90
    },
    {
      "epoch": 0.18026137899954936,
      "grad_norm": 0.508313775062561,
      "learning_rate": 4.548736462093863e-05,
      "loss": 0.0207,
      "step": 100
    },
    {
      "epoch": 0.19828751689950427,
      "grad_norm": 0.2911165952682495,
      "learning_rate": 4.5036101083032495e-05,
      "loss": 0.0605,
      "step": 110
    },
    {
      "epoch": 0.21631365479945922,
      "grad_norm": 0.7530468106269836,
      "learning_rate": 4.458483754512636e-05,
      "loss": 0.059,
      "step": 120
    },
    {
      "epoch": 0.23433979269941416,
      "grad_norm": 0.6048161387443542,
      "learning_rate": 4.413357400722022e-05,
      "loss": 0.0426,
      "step": 130
    },
    {
      "epoch": 0.25236593059936907,
      "grad_norm": 7.237401485443115,
      "learning_rate": 4.368231046931408e-05,
      "loss": 0.0553,
      "step": 140
    },
    {
      "epoch": 0.270392068499324,
      "grad_norm": 4.217318058013916,
      "learning_rate": 4.3231046931407945e-05,
      "loss": 0.0314,
      "step": 150
    },
    {
      "epoch": 0.28841820639927895,
      "grad_norm": 3.7883381843566895,
      "learning_rate": 4.277978339350181e-05,
      "loss": 0.0422,
      "step": 160
    },
    {
      "epoch": 0.3064443442992339,
      "grad_norm": 7.496313571929932,
      "learning_rate": 4.232851985559567e-05,
      "loss": 0.0422,
      "step": 170
    },
    {
      "epoch": 0.32447048219918884,
      "grad_norm": 1.371532678604126,
      "learning_rate": 4.187725631768953e-05,
      "loss": 0.0574,
      "step": 180
    },
    {
      "epoch": 0.3424966200991438,
      "grad_norm": 1.0054638385772705,
      "learning_rate": 4.1425992779783394e-05,
      "loss": 0.0528,
      "step": 190
    },
    {
      "epoch": 0.3605227579990987,
      "grad_norm": 6.67033052444458,
      "learning_rate": 4.0974729241877256e-05,
      "loss": 0.0528,
      "step": 200
    },
    {
      "epoch": 0.3785488958990536,
      "grad_norm": 2.814408779144287,
      "learning_rate": 4.052346570397112e-05,
      "loss": 0.0514,
      "step": 210
    },
    {
      "epoch": 0.39657503379900855,
      "grad_norm": 0.7789732217788696,
      "learning_rate": 4.007220216606498e-05,
      "loss": 0.0241,
      "step": 220
    },
    {
      "epoch": 0.4146011716989635,
      "grad_norm": 5.927615642547607,
      "learning_rate": 3.962093862815885e-05,
      "loss": 0.0742,
      "step": 230
    },
    {
      "epoch": 0.43262730959891843,
      "grad_norm": 4.2895307540893555,
      "learning_rate": 3.916967509025271e-05,
      "loss": 0.0487,
      "step": 240
    },
    {
      "epoch": 0.45065344749887337,
      "grad_norm": 0.24509297311306,
      "learning_rate": 3.8718411552346575e-05,
      "loss": 0.0185,
      "step": 250
    },
    {
      "epoch": 0.4686795853988283,
      "grad_norm": 0.12779287993907928,
      "learning_rate": 3.826714801444044e-05,
      "loss": 0.011,
      "step": 260
    },
    {
      "epoch": 0.48670572329878325,
      "grad_norm": 3.5602619647979736,
      "learning_rate": 3.78158844765343e-05,
      "loss": 0.0605,
      "step": 270
    },
    {
      "epoch": 0.5047318611987381,
      "grad_norm": 7.118227005004883,
      "learning_rate": 3.7364620938628155e-05,
      "loss": 0.062,
      "step": 280
    },
    {
      "epoch": 0.5227579990986931,
      "grad_norm": 0.6336535215377808,
      "learning_rate": 3.6913357400722025e-05,
      "loss": 0.0278,
      "step": 290
    },
    {
      "epoch": 0.540784136998648,
      "grad_norm": 3.6674723625183105,
      "learning_rate": 3.646209386281589e-05,
      "loss": 0.0519,
      "step": 300
    },
    {
      "epoch": 0.558810274898603,
      "grad_norm": 7.329596519470215,
      "learning_rate": 3.601083032490975e-05,
      "loss": 0.0337,
      "step": 310
    },
    {
      "epoch": 0.5768364127985579,
      "grad_norm": 1.3290072679519653,
      "learning_rate": 3.555956678700361e-05,
      "loss": 0.0528,
      "step": 320
    },
    {
      "epoch": 0.5948625506985128,
      "grad_norm": 1.0643926858901978,
      "learning_rate": 3.5108303249097474e-05,
      "loss": 0.0592,
      "step": 330
    },
    {
      "epoch": 0.6128886885984678,
      "grad_norm": 1.0425355434417725,
      "learning_rate": 3.4657039711191336e-05,
      "loss": 0.0485,
      "step": 340
    },
    {
      "epoch": 0.6309148264984227,
      "grad_norm": 3.1306347846984863,
      "learning_rate": 3.42057761732852e-05,
      "loss": 0.0221,
      "step": 350
    },
    {
      "epoch": 0.6489409643983777,
      "grad_norm": 0.37137073278427124,
      "learning_rate": 3.375451263537907e-05,
      "loss": 0.0401,
      "step": 360
    },
    {
      "epoch": 0.6669671022983326,
      "grad_norm": 0.29780980944633484,
      "learning_rate": 3.330324909747293e-05,
      "loss": 0.0374,
      "step": 370
    },
    {
      "epoch": 0.6849932401982876,
      "grad_norm": 1.7151790857315063,
      "learning_rate": 3.2851985559566786e-05,
      "loss": 0.0364,
      "step": 380
    },
    {
      "epoch": 0.7030193780982424,
      "grad_norm": 0.6182792782783508,
      "learning_rate": 3.240072202166065e-05,
      "loss": 0.0253,
      "step": 390
    },
    {
      "epoch": 0.7210455159981974,
      "grad_norm": 0.5171168446540833,
      "learning_rate": 3.194945848375451e-05,
      "loss": 0.0234,
      "step": 400
    },
    {
      "epoch": 0.7390716538981523,
      "grad_norm": 3.500291585922241,
      "learning_rate": 3.149819494584837e-05,
      "loss": 0.048,
      "step": 410
    },
    {
      "epoch": 0.7570977917981072,
      "grad_norm": 4.192553520202637,
      "learning_rate": 3.104693140794224e-05,
      "loss": 0.0191,
      "step": 420
    },
    {
      "epoch": 0.7751239296980622,
      "grad_norm": 0.2889997661113739,
      "learning_rate": 3.0595667870036104e-05,
      "loss": 0.073,
      "step": 430
    },
    {
      "epoch": 0.7931500675980171,
      "grad_norm": 0.731918215751648,
      "learning_rate": 3.0144404332129967e-05,
      "loss": 0.0269,
      "step": 440
    },
    {
      "epoch": 0.8111762054979721,
      "grad_norm": 3.4268507957458496,
      "learning_rate": 2.969314079422383e-05,
      "loss": 0.0634,
      "step": 450
    },
    {
      "epoch": 0.829202343397927,
      "grad_norm": 0.8300638794898987,
      "learning_rate": 2.924187725631769e-05,
      "loss": 0.0382,
      "step": 460
    },
    {
      "epoch": 0.847228481297882,
      "grad_norm": 0.6669813990592957,
      "learning_rate": 2.879061371841155e-05,
      "loss": 0.0678,
      "step": 470
    },
    {
      "epoch": 0.8652546191978369,
      "grad_norm": 1.2740308046340942,
      "learning_rate": 2.8339350180505413e-05,
      "loss": 0.0415,
      "step": 480
    },
    {
      "epoch": 0.8832807570977917,
      "grad_norm": 3.1045358180999756,
      "learning_rate": 2.7888086642599282e-05,
      "loss": 0.031,
      "step": 490
    },
    {
      "epoch": 0.9013068949977467,
      "grad_norm": 3.4434397220611572,
      "learning_rate": 2.7436823104693144e-05,
      "loss": 0.051,
      "step": 500
    },
    {
      "epoch": 0.9193330328977016,
      "grad_norm": 8.245412826538086,
      "learning_rate": 2.6985559566787007e-05,
      "loss": 0.0209,
      "step": 510
    },
    {
      "epoch": 0.9373591707976566,
      "grad_norm": 1.196775197982788,
      "learning_rate": 2.6534296028880866e-05,
      "loss": 0.0416,
      "step": 520
    },
    {
      "epoch": 0.9553853086976115,
      "grad_norm": 0.6041325330734253,
      "learning_rate": 2.6083032490974728e-05,
      "loss": 0.0503,
      "step": 530
    },
    {
      "epoch": 0.9734114465975665,
      "grad_norm": 0.4048420190811157,
      "learning_rate": 2.563176895306859e-05,
      "loss": 0.0298,
      "step": 540
    },
    {
      "epoch": 0.9914375844975214,
      "grad_norm": 0.48081880807876587,
      "learning_rate": 2.518050541516246e-05,
      "loss": 0.0507,
      "step": 550
    },
    {
      "epoch": 1.0,
      "eval_f1": 0.0,
      "eval_fn": 43,
      "eval_fp": 0,
      "eval_loss": 0.042322322726249695,
      "eval_precision": 0.0,
      "eval_recall": 0.0,
      "eval_runtime": 14.9507,
      "eval_samples_per_second": 312.56,
      "eval_steps_per_second": 39.129,
      "eval_tn": 4630,
      "eval_tp": 0,
      "step": 555
    },
    {
      "epoch": 1.0090130689499774,
      "grad_norm": 1.1511574983596802,
      "learning_rate": 2.472924187725632e-05,
      "loss": 0.0107,
      "step": 560
    },
    {
      "epoch": 1.0270392068499323,
      "grad_norm": 6.283339500427246,
      "learning_rate": 2.427797833935018e-05,
      "loss": 0.0327,
      "step": 570
    },
    {
      "epoch": 1.0450653447498874,
      "grad_norm": 1.1113625764846802,
      "learning_rate": 2.3826714801444043e-05,
      "loss": 0.0184,
      "step": 580
    },
    {
      "epoch": 1.0630914826498423,
      "grad_norm": 0.49993896484375,
      "learning_rate": 2.3375451263537906e-05,
      "loss": 0.0755,
      "step": 590
    },
    {
      "epoch": 1.0811176205497972,
      "grad_norm": 3.5307536125183105,
      "learning_rate": 2.292418772563177e-05,
      "loss": 0.0292,
      "step": 600
    },
    {
      "epoch": 1.099143758449752,
      "grad_norm": 5.070742607116699,
      "learning_rate": 2.2472924187725634e-05,
      "loss": 0.0226,
      "step": 610
    },
    {
      "epoch": 1.117169896349707,
      "grad_norm": 3.8874106407165527,
      "learning_rate": 2.2021660649819496e-05,
      "loss": 0.0494,
      "step": 620
    },
    {
      "epoch": 1.135196034249662,
      "grad_norm": 7.7703680992126465,
      "learning_rate": 2.157039711191336e-05,
      "loss": 0.0413,
      "step": 630
    },
    {
      "epoch": 1.153222172149617,
      "grad_norm": 3.534682273864746,
      "learning_rate": 2.111913357400722e-05,
      "loss": 0.0631,
      "step": 640
    },
    {
      "epoch": 1.1712483100495719,
      "grad_norm": 3.8025522232055664,
      "learning_rate": 2.0667870036101083e-05,
      "loss": 0.053,
      "step": 650
    },
    {
      "epoch": 1.1892744479495267,
      "grad_norm": 0.3169710040092468,
      "learning_rate": 2.0216606498194946e-05,
      "loss": 0.0254,
      "step": 660
    },
    {
      "epoch": 1.2073005858494819,
      "grad_norm": 3.912919282913208,
      "learning_rate": 1.976534296028881e-05,
      "loss": 0.0282,
      "step": 670
    },
    {
      "epoch": 1.2253267237494367,
      "grad_norm": 0.7752419710159302,
      "learning_rate": 1.9314079422382674e-05,
      "loss": 0.0503,
      "step": 680
    },
    {
      "epoch": 1.2433528616493916,
      "grad_norm": 0.21697914600372314,
      "learning_rate": 1.8862815884476533e-05,
      "loss": 0.0419,
      "step": 690
    },
    {
      "epoch": 1.2613789995493465,
      "grad_norm": 2.617170810699463,
      "learning_rate": 1.84115523465704e-05,
      "loss": 0.0182,
      "step": 700
    },
    {
      "epoch": 1.2794051374493014,
      "grad_norm": 1.5197302103042603,
      "learning_rate": 1.796028880866426e-05,
      "loss": 0.0559,
      "step": 710
    },
    {
      "epoch": 1.2974312753492563,
      "grad_norm": 0.32236000895500183,
      "learning_rate": 1.7509025270758123e-05,
      "loss": 0.0026,
      "step": 720
    },
    {
      "epoch": 1.3154574132492114,
      "grad_norm": 8.387882232666016,
      "learning_rate": 1.7057761732851986e-05,
      "loss": 0.0668,
      "step": 730
    },
    {
      "epoch": 1.3334835511491663,
      "grad_norm": 4.180727958679199,
      "learning_rate": 1.6606498194945848e-05,
      "loss": 0.0861,
      "step": 740
    },
    {
      "epoch": 1.3515096890491212,
      "grad_norm": 0.9219232201576233,
      "learning_rate": 1.615523465703971e-05,
      "loss": 0.0315,
      "step": 750
    },
    {
      "epoch": 1.3695358269490763,
      "grad_norm": 0.45807236433029175,
      "learning_rate": 1.5703971119133576e-05,
      "loss": 0.0141,
      "step": 760
    },
    {
      "epoch": 1.3875619648490312,
      "grad_norm": 4.011260032653809,
      "learning_rate": 1.5252707581227438e-05,
      "loss": 0.033,
      "step": 770
    },
    {
      "epoch": 1.405588102748986,
      "grad_norm": 0.1898992657661438,
      "learning_rate": 1.4801444043321299e-05,
      "loss": 0.028,
      "step": 780
    },
    {
      "epoch": 1.423614240648941,
      "grad_norm": 0.4946901500225067,
      "learning_rate": 1.4350180505415165e-05,
      "loss": 0.0645,
      "step": 790
    },
    {
      "epoch": 1.4416403785488958,
      "grad_norm": 4.39424991607666,
      "learning_rate": 1.3898916967509026e-05,
      "loss": 0.0789,
      "step": 800
    },
    {
      "epoch": 1.4596665164488507,
      "grad_norm": 0.3170035481452942,
      "learning_rate": 1.3447653429602888e-05,
      "loss": 0.0022,
      "step": 810
    },
    {
      "epoch": 1.4776926543488058,
      "grad_norm": 7.535902500152588,
      "learning_rate": 1.299638989169675e-05,
      "loss": 0.0533,
      "step": 820
    },
    {
      "epoch": 1.4957187922487607,
      "grad_norm": 0.3619483709335327,
      "learning_rate": 1.2545126353790614e-05,
      "loss": 0.0333,
      "step": 830
    },
    {
      "epoch": 1.5137449301487156,
      "grad_norm": 3.136875629425049,
      "learning_rate": 1.2093862815884477e-05,
      "loss": 0.0208,
      "step": 840
    },
    {
      "epoch": 1.5317710680486707,
      "grad_norm": 4.279853820800781,
      "learning_rate": 1.164259927797834e-05,
      "loss": 0.0307,
      "step": 850
    },
    {
      "epoch": 1.5497972059486256,
      "grad_norm": 0.1678500771522522,
      "learning_rate": 1.1191335740072201e-05,
      "loss": 0.041,
      "step": 860
    },
    {
      "epoch": 1.5678233438485805,
      "grad_norm": 3.2561416625976562,
      "learning_rate": 1.0740072202166066e-05,
      "loss": 0.0584,
      "step": 870
    },
    {
      "epoch": 1.5858494817485354,
      "grad_norm": 7.7148847579956055,
      "learning_rate": 1.028880866425993e-05,
      "loss": 0.0505,
      "step": 880
    },
    {
      "epoch": 1.6038756196484902,
      "grad_norm": 3.3185997009277344,
      "learning_rate": 9.83754512635379e-06,
      "loss": 0.0402,
      "step": 890
    },
    {
      "epoch": 1.6219017575484451,
      "grad_norm": 4.2412824630737305,
      "learning_rate": 9.386281588447654e-06,
      "loss": 0.0469,
      "step": 900
    },
    {
      "epoch": 1.6399278954484002,
      "grad_norm": 3.6402697563171387,
      "learning_rate": 8.935018050541517e-06,
      "loss": 0.0241,
      "step": 910
    },
    {
      "epoch": 1.6579540333483551,
      "grad_norm": 0.7010020613670349,
      "learning_rate": 8.483754512635379e-06,
      "loss": 0.0254,
      "step": 920
    },
    {
      "epoch": 1.67598017124831,
      "grad_norm": 0.4133553206920624,
      "learning_rate": 8.032490974729243e-06,
      "loss": 0.0243,
      "step": 930
    },
    {
      "epoch": 1.694006309148265,
      "grad_norm": 0.443309485912323,
      "learning_rate": 7.581227436823105e-06,
      "loss": 0.0204,
      "step": 940
    },
    {
      "epoch": 1.71203244704822,
      "grad_norm": 0.9218553304672241,
      "learning_rate": 7.129963898916968e-06,
      "loss": 0.021,
      "step": 950
    },
    {
      "epoch": 1.7300585849481749,
      "grad_norm": 0.4528704881668091,
      "learning_rate": 6.678700361010831e-06,
      "loss": 0.0817,
      "step": 960
    },
    {
      "epoch": 1.7480847228481298,
      "grad_norm": 0.191946342587471,
      "learning_rate": 6.227436823104693e-06,
      "loss": 0.0289,
      "step": 970
    },
    {
      "epoch": 1.7661108607480847,
      "grad_norm": 0.4200621545314789,
      "learning_rate": 5.776173285198557e-06,
      "loss": 0.026,
      "step": 980
    },
    {
      "epoch": 1.7841369986480395,
      "grad_norm": 0.16432328522205353,
      "learning_rate": 5.324909747292419e-06,
      "loss": 0.055,
      "step": 990
    },
    {
      "epoch": 1.8021631365479944,
      "grad_norm": 3.6078760623931885,
      "learning_rate": 4.873646209386281e-06,
      "loss": 0.0416,
      "step": 1000
    },
    {
      "epoch": 1.8201892744479495,
      "grad_norm": 0.8375062346458435,
      "learning_rate": 4.422382671480145e-06,
      "loss": 0.0253,
      "step": 1010
    },
    {
      "epoch": 1.8382154123479044,
      "grad_norm": 0.4356284737586975,
      "learning_rate": 3.971119133574008e-06,
      "loss": 0.0344,
      "step": 1020
    },
    {
      "epoch": 1.8562415502478595,
      "grad_norm": 3.238054037094116,
      "learning_rate": 3.51985559566787e-06,
      "loss": 0.0319,
      "step": 1030
    },
    {
      "epoch": 1.8742676881478144,
      "grad_norm": 3.837254524230957,
      "learning_rate": 3.068592057761733e-06,
      "loss": 0.0594,
      "step": 1040
    },
    {
      "epoch": 1.8922938260477693,
      "grad_norm": 0.2289639711380005,
      "learning_rate": 2.6173285198555957e-06,
      "loss": 0.0463,
      "step": 1050
    },
    {
      "epoch": 1.9103199639477242,
      "grad_norm": 0.4962206482887268,
      "learning_rate": 2.1660649819494585e-06,
      "loss": 0.0103,
      "step": 1060
    },
    {
      "epoch": 1.928346101847679,
      "grad_norm": 2.8832972049713135,
      "learning_rate": 1.7148014440433215e-06,
      "loss": 0.0144,
      "step": 1070
    },
    {
      "epoch": 1.946372239747634,
      "grad_norm": 0.40793848037719727,
      "learning_rate": 1.263537906137184e-06,
      "loss": 0.037,
      "step": 1080
    },
    {
      "epoch": 1.9643983776475888,
      "grad_norm": 1.6269336938858032,
      "learning_rate": 8.122743682310469e-07,
      "loss": 0.0203,
      "step": 1090
    },
    {
      "epoch": 1.982424515547544,
      "grad_norm": 1.4103859663009644,
      "learning_rate": 3.610108303249098e-07,
      "loss": 0.0416,
      "step": 1100
    },
    {
      "epoch": 1.996845425867508,
      "eval_f1": 0.0,
      "eval_fn": 43,
      "eval_fp": 0,
      "eval_loss": 0.039656173437833786,
      "eval_precision": 0.0,
      "eval_recall": 0.0,
      "eval_runtime": 20.0815,
      "eval_samples_per_second": 232.702,
      "eval_steps_per_second": 29.131,
      "eval_tn": 4630,
      "eval_tp": 0,
      "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
}