{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 10.0,
  "eval_steps": 500,
  "global_step": 2820,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.03546099290780142,
      "grad_norm": 12.458266258239746,
      "learning_rate": 6.3829787234042555e-06,
      "loss": 1.452,
      "step": 10
    },
    {
      "epoch": 0.07092198581560284,
      "grad_norm": 2.127192974090576,
      "learning_rate": 1.347517730496454e-05,
      "loss": 0.5036,
      "step": 20
    },
    {
      "epoch": 0.10638297872340426,
      "grad_norm": 1.6429086923599243,
      "learning_rate": 2.0567375886524822e-05,
      "loss": 0.3388,
      "step": 30
    },
    {
      "epoch": 0.14184397163120568,
      "grad_norm": 1.6117417812347412,
      "learning_rate": 2.765957446808511e-05,
      "loss": 0.3197,
      "step": 40
    },
    {
      "epoch": 0.1773049645390071,
      "grad_norm": 1.5715168714523315,
      "learning_rate": 3.4751773049645395e-05,
      "loss": 0.244,
      "step": 50
    },
    {
      "epoch": 0.2127659574468085,
      "grad_norm": 1.6177239418029785,
      "learning_rate": 4.1843971631205674e-05,
      "loss": 0.2038,
      "step": 60
    },
    {
      "epoch": 0.24822695035460993,
      "grad_norm": 1.159698724746704,
      "learning_rate": 4.893617021276596e-05,
      "loss": 0.1898,
      "step": 70
    },
    {
      "epoch": 0.28368794326241137,
      "grad_norm": 1.3677239418029785,
      "learning_rate": 5.602836879432625e-05,
      "loss": 0.1535,
      "step": 80
    },
    {
      "epoch": 0.3191489361702128,
      "grad_norm": 0.9369192719459534,
      "learning_rate": 6.312056737588653e-05,
      "loss": 0.148,
      "step": 90
    },
    {
      "epoch": 0.3546099290780142,
      "grad_norm": 1.7364780902862549,
      "learning_rate": 7.021276595744681e-05,
      "loss": 0.1408,
      "step": 100
    },
    {
      "epoch": 0.3900709219858156,
      "grad_norm": 1.2476744651794434,
      "learning_rate": 7.73049645390071e-05,
      "loss": 0.1224,
      "step": 110
    },
    {
      "epoch": 0.425531914893617,
      "grad_norm": 1.190326452255249,
      "learning_rate": 8.439716312056739e-05,
      "loss": 0.1166,
      "step": 120
    },
    {
      "epoch": 0.46099290780141844,
      "grad_norm": 1.6969983577728271,
      "learning_rate": 9.148936170212766e-05,
      "loss": 0.1248,
      "step": 130
    },
    {
      "epoch": 0.49645390070921985,
      "grad_norm": 2.593116044998169,
      "learning_rate": 9.858156028368794e-05,
      "loss": 0.1202,
      "step": 140
    },
    {
      "epoch": 0.5319148936170213,
      "grad_norm": 1.28589928150177,
      "learning_rate": 9.999779975463078e-05,
      "loss": 0.1123,
      "step": 150
    },
    {
      "epoch": 0.5673758865248227,
      "grad_norm": 0.9029638171195984,
      "learning_rate": 9.9988861589696e-05,
      "loss": 0.1015,
      "step": 160
    },
    {
      "epoch": 0.6028368794326241,
      "grad_norm": 1.8789817094802856,
      "learning_rate": 9.99730492180442e-05,
      "loss": 0.1087,
      "step": 170
    },
    {
      "epoch": 0.6382978723404256,
      "grad_norm": 1.440181851387024,
      "learning_rate": 9.995036481411004e-05,
      "loss": 0.128,
      "step": 180
    },
    {
      "epoch": 0.6737588652482269,
      "grad_norm": 0.8681181073188782,
      "learning_rate": 9.992081149733404e-05,
      "loss": 0.1248,
      "step": 190
    },
    {
      "epoch": 0.7092198581560284,
      "grad_norm": 0.7192113399505615,
      "learning_rate": 9.988439333173374e-05,
      "loss": 0.1062,
      "step": 200
    },
    {
      "epoch": 0.7446808510638298,
      "grad_norm": 1.1343145370483398,
      "learning_rate": 9.98411153253447e-05,
      "loss": 0.1067,
      "step": 210
    },
    {
      "epoch": 0.7801418439716312,
      "grad_norm": 0.8589990139007568,
      "learning_rate": 9.979098342953199e-05,
      "loss": 0.1008,
      "step": 220
    },
    {
      "epoch": 0.8156028368794326,
      "grad_norm": 0.9460039734840393,
      "learning_rate": 9.97340045381716e-05,
      "loss": 0.0987,
      "step": 230
    },
    {
      "epoch": 0.851063829787234,
      "grad_norm": 1.0302399396896362,
      "learning_rate": 9.967018648670259e-05,
      "loss": 0.101,
      "step": 240
    },
    {
      "epoch": 0.8865248226950354,
      "grad_norm": 1.2364697456359863,
      "learning_rate": 9.959953805104952e-05,
      "loss": 0.0961,
      "step": 250
    },
    {
      "epoch": 0.9219858156028369,
      "grad_norm": 1.174009919166565,
      "learning_rate": 9.952206894641565e-05,
      "loss": 0.0933,
      "step": 260
    },
    {
      "epoch": 0.9574468085106383,
      "grad_norm": 1.2122607231140137,
      "learning_rate": 9.943778982594694e-05,
      "loss": 0.0892,
      "step": 270
    },
    {
      "epoch": 0.9929078014184397,
      "grad_norm": 1.4150590896606445,
      "learning_rate": 9.934671227926712e-05,
      "loss": 0.0856,
      "step": 280
    },
    {
      "epoch": 1.0283687943262412,
      "grad_norm": 0.9654158353805542,
      "learning_rate": 9.924884883088391e-05,
      "loss": 0.0918,
      "step": 290
    },
    {
      "epoch": 1.0638297872340425,
      "grad_norm": 1.0064071416854858,
      "learning_rate": 9.914421293846674e-05,
      "loss": 0.0861,
      "step": 300
    },
    {
      "epoch": 1.099290780141844,
      "grad_norm": 1.430910348892212,
      "learning_rate": 9.90328189909961e-05,
      "loss": 0.085,
      "step": 310
    },
    {
      "epoch": 1.1347517730496455,
      "grad_norm": 1.0088697671890259,
      "learning_rate": 9.891468230678487e-05,
      "loss": 0.0927,
      "step": 320
    },
    {
      "epoch": 1.1702127659574468,
      "grad_norm": 0.7953307032585144,
      "learning_rate": 9.878981913137179e-05,
      "loss": 0.0962,
      "step": 330
    },
    {
      "epoch": 1.2056737588652482,
      "grad_norm": 0.7745844721794128,
      "learning_rate": 9.86582466352875e-05,
      "loss": 0.0806,
      "step": 340
    },
    {
      "epoch": 1.2411347517730495,
      "grad_norm": 0.39872488379478455,
      "learning_rate": 9.85199829116933e-05,
      "loss": 0.0751,
      "step": 350
    },
    {
      "epoch": 1.2765957446808511,
      "grad_norm": 0.9581345915794373,
      "learning_rate": 9.837504697389311e-05,
      "loss": 0.0839,
      "step": 360
    },
    {
      "epoch": 1.3120567375886525,
      "grad_norm": 0.9904010891914368,
      "learning_rate": 9.822345875271883e-05,
      "loss": 0.0747,
      "step": 370
    },
    {
      "epoch": 1.3475177304964538,
      "grad_norm": 1.0076277256011963,
      "learning_rate": 9.806523909378956e-05,
      "loss": 0.072,
      "step": 380
    },
    {
      "epoch": 1.3829787234042552,
      "grad_norm": 1.1363801956176758,
      "learning_rate": 9.790040975464503e-05,
      "loss": 0.09,
      "step": 390
    },
    {
      "epoch": 1.4184397163120568,
      "grad_norm": 0.936143159866333,
      "learning_rate": 9.772899340175361e-05,
      "loss": 0.0956,
      "step": 400
    },
    {
      "epoch": 1.4539007092198581,
      "grad_norm": 0.5812869071960449,
      "learning_rate": 9.755101360739536e-05,
      "loss": 0.09,
      "step": 410
    },
    {
      "epoch": 1.4893617021276595,
      "grad_norm": 1.0588104724884033,
      "learning_rate": 9.736649484642043e-05,
      "loss": 0.0771,
      "step": 420
    },
    {
      "epoch": 1.524822695035461,
      "grad_norm": 1.1113085746765137,
      "learning_rate": 9.71754624928835e-05,
      "loss": 0.0681,
      "step": 430
    },
    {
      "epoch": 1.5602836879432624,
      "grad_norm": 0.7930032014846802,
      "learning_rate": 9.697794281655438e-05,
      "loss": 0.0752,
      "step": 440
    },
    {
      "epoch": 1.5957446808510638,
      "grad_norm": 0.8521007895469666,
      "learning_rate": 9.67739629793056e-05,
      "loss": 0.0754,
      "step": 450
    },
    {
      "epoch": 1.6312056737588652,
      "grad_norm": 0.7623393535614014,
      "learning_rate": 9.65635510313772e-05,
      "loss": 0.0858,
      "step": 460
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 1.1308192014694214,
      "learning_rate": 9.634673590751944e-05,
      "loss": 0.0819,
      "step": 470
    },
    {
      "epoch": 1.702127659574468,
      "grad_norm": 0.8282614350318909,
      "learning_rate": 9.612354742301382e-05,
      "loss": 0.0769,
      "step": 480
    },
    {
      "epoch": 1.7375886524822695,
      "grad_norm": 0.8726593852043152,
      "learning_rate": 9.589401626957308e-05,
      "loss": 0.0654,
      "step": 490
    },
    {
      "epoch": 1.773049645390071,
      "grad_norm": 0.6696294546127319,
      "learning_rate": 9.56581740111206e-05,
      "loss": 0.062,
      "step": 500
    },
    {
      "epoch": 1.8085106382978724,
      "grad_norm": 1.136682391166687,
      "learning_rate": 9.54160530794499e-05,
      "loss": 0.0733,
      "step": 510
    },
    {
      "epoch": 1.8439716312056738,
      "grad_norm": 0.6757348775863647,
      "learning_rate": 9.516768676976475e-05,
      "loss": 0.0786,
      "step": 520
    },
    {
      "epoch": 1.8794326241134751,
      "grad_norm": 0.6466293931007385,
      "learning_rate": 9.49131092361007e-05,
      "loss": 0.0707,
      "step": 530
    },
    {
      "epoch": 1.9148936170212765,
      "grad_norm": 0.49455875158309937,
      "learning_rate": 9.46523554866283e-05,
      "loss": 0.0719,
      "step": 540
    },
    {
      "epoch": 1.950354609929078,
      "grad_norm": 0.5616086721420288,
      "learning_rate": 9.438546137883897e-05,
      "loss": 0.0753,
      "step": 550
    },
    {
      "epoch": 1.9858156028368794,
      "grad_norm": 0.5382352471351624,
      "learning_rate": 9.41124636146141e-05,
      "loss": 0.0722,
      "step": 560
    },
    {
      "epoch": 2.021276595744681,
      "grad_norm": 0.5831167101860046,
      "learning_rate": 9.383339973517795e-05,
      "loss": 0.0628,
      "step": 570
    },
    {
      "epoch": 2.0567375886524824,
      "grad_norm": 0.6010406613349915,
      "learning_rate": 9.354830811593527e-05,
      "loss": 0.0643,
      "step": 580
    },
    {
      "epoch": 2.0921985815602837,
      "grad_norm": 0.6855683922767639,
      "learning_rate": 9.325722796119397e-05,
      "loss": 0.0734,
      "step": 590
    },
    {
      "epoch": 2.127659574468085,
      "grad_norm": 0.5798056125640869,
      "learning_rate": 9.29601992987741e-05,
      "loss": 0.0683,
      "step": 600
    },
    {
      "epoch": 2.1631205673758864,
      "grad_norm": 0.5442080497741699,
      "learning_rate": 9.265726297450333e-05,
      "loss": 0.0683,
      "step": 610
    },
    {
      "epoch": 2.198581560283688,
      "grad_norm": 0.5252788662910461,
      "learning_rate": 9.234846064660009e-05,
      "loss": 0.0776,
      "step": 620
    },
    {
      "epoch": 2.2340425531914896,
      "grad_norm": 0.7919511198997498,
      "learning_rate": 9.203383477994494e-05,
      "loss": 0.0611,
      "step": 630
    },
    {
      "epoch": 2.269503546099291,
      "grad_norm": 0.7639462947845459,
      "learning_rate": 9.171342864024101e-05,
      "loss": 0.0588,
      "step": 640
    },
    {
      "epoch": 2.3049645390070923,
      "grad_norm": 0.9514329433441162,
      "learning_rate": 9.138728628806439e-05,
      "loss": 0.0725,
      "step": 650
    },
    {
      "epoch": 2.3404255319148937,
      "grad_norm": 0.8672018051147461,
      "learning_rate": 9.105545257280501e-05,
      "loss": 0.0568,
      "step": 660
    },
    {
      "epoch": 2.375886524822695,
      "grad_norm": 0.7574223279953003,
      "learning_rate": 9.071797312649935e-05,
      "loss": 0.0752,
      "step": 670
    },
    {
      "epoch": 2.4113475177304964,
      "grad_norm": 1.0806403160095215,
      "learning_rate": 9.037489435755524e-05,
      "loss": 0.0682,
      "step": 680
    },
    {
      "epoch": 2.4468085106382977,
      "grad_norm": 0.8450857400894165,
      "learning_rate": 9.00262634443701e-05,
      "loss": 0.0669,
      "step": 690
    },
    {
      "epoch": 2.482269503546099,
      "grad_norm": 1.4907677173614502,
      "learning_rate": 8.96721283288432e-05,
      "loss": 0.0754,
      "step": 700
    },
    {
      "epoch": 2.5177304964539005,
      "grad_norm": 1.3826724290847778,
      "learning_rate": 8.93125377097828e-05,
      "loss": 0.07,
      "step": 710
    },
    {
      "epoch": 2.5531914893617023,
      "grad_norm": 1.0240296125411987,
      "learning_rate": 8.894754103620964e-05,
      "loss": 0.077,
      "step": 720
    },
    {
      "epoch": 2.5886524822695036,
      "grad_norm": 0.8244665265083313,
      "learning_rate": 8.857718850055662e-05,
      "loss": 0.0731,
      "step": 730
    },
    {
      "epoch": 2.624113475177305,
      "grad_norm": 0.5112609267234802,
      "learning_rate": 8.820153103176693e-05,
      "loss": 0.0737,
      "step": 740
    },
    {
      "epoch": 2.6595744680851063,
      "grad_norm": 0.6853227019309998,
      "learning_rate": 8.782062028829028e-05,
      "loss": 0.077,
      "step": 750
    },
    {
      "epoch": 2.6950354609929077,
      "grad_norm": 0.6311427354812622,
      "learning_rate": 8.743450865097929e-05,
      "loss": 0.0635,
      "step": 760
    },
    {
      "epoch": 2.7304964539007095,
      "grad_norm": 0.5705634355545044,
      "learning_rate": 8.704324921588632e-05,
      "loss": 0.0699,
      "step": 770
    },
    {
      "epoch": 2.7659574468085104,
      "grad_norm": 1.1000028848648071,
      "learning_rate": 8.664689578696188e-05,
      "loss": 0.0761,
      "step": 780
    },
    {
      "epoch": 2.801418439716312,
      "grad_norm": 0.7304648160934448,
      "learning_rate": 8.624550286865591e-05,
      "loss": 0.0679,
      "step": 790
    },
    {
      "epoch": 2.8368794326241136,
      "grad_norm": 0.6147809028625488,
      "learning_rate": 8.583912565842257e-05,
      "loss": 0.0675,
      "step": 800
    },
    {
      "epoch": 2.872340425531915,
      "grad_norm": 0.5900303721427917,
      "learning_rate": 8.542782003912972e-05,
      "loss": 0.0649,
      "step": 810
    },
    {
      "epoch": 2.9078014184397163,
      "grad_norm": 0.5605118274688721,
      "learning_rate": 8.50116425713743e-05,
      "loss": 0.0701,
      "step": 820
    },
    {
      "epoch": 2.9432624113475176,
      "grad_norm": 0.691639244556427,
      "learning_rate": 8.459065048570434e-05,
      "loss": 0.0679,
      "step": 830
    },
    {
      "epoch": 2.978723404255319,
      "grad_norm": 0.6846038103103638,
      "learning_rate": 8.416490167474895e-05,
      "loss": 0.0693,
      "step": 840
    },
    {
      "epoch": 3.0141843971631204,
      "grad_norm": 0.5968617796897888,
      "learning_rate": 8.373445468525719e-05,
      "loss": 0.0575,
      "step": 850
    },
    {
      "epoch": 3.049645390070922,
      "grad_norm": 1.1188421249389648,
      "learning_rate": 8.329936871004703e-05,
      "loss": 0.0715,
      "step": 860
    },
    {
      "epoch": 3.0851063829787235,
      "grad_norm": 0.888618528842926,
      "learning_rate": 8.285970357986559e-05,
      "loss": 0.057,
      "step": 870
    },
    {
      "epoch": 3.120567375886525,
      "grad_norm": 0.8040549159049988,
      "learning_rate": 8.241551975516133e-05,
      "loss": 0.068,
      "step": 880
    },
    {
      "epoch": 3.1560283687943262,
      "grad_norm": 0.8645780086517334,
      "learning_rate": 8.196687831776996e-05,
      "loss": 0.0666,
      "step": 890
    },
    {
      "epoch": 3.1914893617021276,
      "grad_norm": 0.9035763144493103,
      "learning_rate": 8.15138409625148e-05,
      "loss": 0.0616,
      "step": 900
    },
    {
      "epoch": 3.226950354609929,
      "grad_norm": 0.6809470057487488,
      "learning_rate": 8.105646998872274e-05,
      "loss": 0.0608,
      "step": 910
    },
    {
      "epoch": 3.2624113475177303,
      "grad_norm": 0.4224311113357544,
      "learning_rate": 8.059482829165728e-05,
      "loss": 0.0547,
      "step": 920
    },
    {
      "epoch": 3.297872340425532,
      "grad_norm": 0.5534777045249939,
      "learning_rate": 8.012897935386937e-05,
      "loss": 0.0578,
      "step": 930
    },
    {
      "epoch": 3.3333333333333335,
      "grad_norm": 0.6809782981872559,
      "learning_rate": 7.965898723646776e-05,
      "loss": 0.0537,
      "step": 940
    },
    {
      "epoch": 3.368794326241135,
      "grad_norm": 0.6213827729225159,
      "learning_rate": 7.918491657030955e-05,
      "loss": 0.0574,
      "step": 950
    },
    {
      "epoch": 3.404255319148936,
      "grad_norm": 1.0936675071716309,
      "learning_rate": 7.870683254711255e-05,
      "loss": 0.0655,
      "step": 960
    },
    {
      "epoch": 3.4397163120567376,
      "grad_norm": 0.879903256893158,
      "learning_rate": 7.822480091049041e-05,
      "loss": 0.0655,
      "step": 970
    },
    {
      "epoch": 3.475177304964539,
      "grad_norm": 0.576164722442627,
      "learning_rate": 7.773888794691191e-05,
      "loss": 0.0582,
      "step": 980
    },
    {
      "epoch": 3.5106382978723403,
      "grad_norm": 0.7445582747459412,
      "learning_rate": 7.724916047658568e-05,
      "loss": 0.0636,
      "step": 990
    },
    {
      "epoch": 3.546099290780142,
      "grad_norm": 0.6598033905029297,
      "learning_rate": 7.675568584427125e-05,
      "loss": 0.0642,
      "step": 1000
    },
    {
      "epoch": 3.581560283687943,
      "grad_norm": 0.8604206442832947,
      "learning_rate": 7.625853191001838e-05,
      "loss": 0.0665,
      "step": 1010
    },
    {
      "epoch": 3.617021276595745,
      "grad_norm": 0.7154328227043152,
      "learning_rate": 7.575776703983508e-05,
      "loss": 0.0557,
      "step": 1020
    },
    {
      "epoch": 3.652482269503546,
      "grad_norm": 0.49133405089378357,
      "learning_rate": 7.525346009628646e-05,
      "loss": 0.0551,
      "step": 1030
    },
    {
      "epoch": 3.6879432624113475,
      "grad_norm": 0.8421124815940857,
      "learning_rate": 7.474568042902496e-05,
      "loss": 0.0612,
      "step": 1040
    },
    {
      "epoch": 3.723404255319149,
      "grad_norm": 0.7970739603042603,
      "learning_rate": 7.42344978652539e-05,
      "loss": 0.0598,
      "step": 1050
    },
    {
      "epoch": 3.7588652482269502,
      "grad_norm": 0.8975237011909485,
      "learning_rate": 7.371998270012504e-05,
      "loss": 0.0574,
      "step": 1060
    },
    {
      "epoch": 3.794326241134752,
      "grad_norm": 0.7334581613540649,
      "learning_rate": 7.320220568707206e-05,
      "loss": 0.0571,
      "step": 1070
    },
    {
      "epoch": 3.829787234042553,
      "grad_norm": 0.3972456753253937,
      "learning_rate": 7.268123802808096e-05,
      "loss": 0.0564,
      "step": 1080
    },
    {
      "epoch": 3.8652482269503547,
      "grad_norm": 0.7282454967498779,
      "learning_rate": 7.215715136389862e-05,
      "loss": 0.0537,
      "step": 1090
    },
    {
      "epoch": 3.900709219858156,
      "grad_norm": 0.5621278285980225,
      "learning_rate": 7.163001776418122e-05,
      "loss": 0.0628,
      "step": 1100
    },
    {
      "epoch": 3.9361702127659575,
      "grad_norm": 0.4074684977531433,
      "learning_rate": 7.109990971758354e-05,
      "loss": 0.05,
      "step": 1110
    },
    {
      "epoch": 3.971631205673759,
      "grad_norm": 0.9387255311012268,
      "learning_rate": 7.056690012179077e-05,
      "loss": 0.0617,
      "step": 1120
    },
    {
      "epoch": 4.00709219858156,
      "grad_norm": 0.8795057535171509,
      "learning_rate": 7.003106227349399e-05,
      "loss": 0.0555,
      "step": 1130
    },
    {
      "epoch": 4.042553191489362,
      "grad_norm": 0.6610236763954163,
      "learning_rate": 6.949246985831068e-05,
      "loss": 0.0526,
      "step": 1140
    },
    {
      "epoch": 4.078014184397163,
      "grad_norm": 0.519904613494873,
      "learning_rate": 6.895119694065204e-05,
      "loss": 0.0508,
      "step": 1150
    },
    {
      "epoch": 4.113475177304965,
      "grad_norm": 0.7559794187545776,
      "learning_rate": 6.840731795353787e-05,
      "loss": 0.0589,
      "step": 1160
    },
    {
      "epoch": 4.148936170212766,
      "grad_norm": 0.5495615601539612,
      "learning_rate": 6.786090768836108e-05,
      "loss": 0.054,
      "step": 1170
    },
    {
      "epoch": 4.184397163120567,
      "grad_norm": 0.38352513313293457,
      "learning_rate": 6.731204128460264e-05,
      "loss": 0.0496,
      "step": 1180
    },
    {
      "epoch": 4.219858156028369,
      "grad_norm": 2.50418758392334,
      "learning_rate": 6.676079421949887e-05,
      "loss": 0.0542,
      "step": 1190
    },
    {
      "epoch": 4.25531914893617,
      "grad_norm": 0.6766555905342102,
      "learning_rate": 6.620724229766218e-05,
      "loss": 0.0512,
      "step": 1200
    },
    {
      "epoch": 4.290780141843972,
      "grad_norm": 0.6073281764984131,
      "learning_rate": 6.565146164065683e-05,
      "loss": 0.0515,
      "step": 1210
    },
    {
      "epoch": 4.326241134751773,
      "grad_norm": 0.5780677795410156,
      "learning_rate": 6.509352867653106e-05,
      "loss": 0.0601,
      "step": 1220
    },
    {
      "epoch": 4.361702127659575,
      "grad_norm": 1.2549879550933838,
      "learning_rate": 6.453352012930712e-05,
      "loss": 0.0538,
      "step": 1230
    },
    {
      "epoch": 4.397163120567376,
      "grad_norm": 1.9051744937896729,
      "learning_rate": 6.397151300843065e-05,
      "loss": 0.0544,
      "step": 1240
    },
    {
      "epoch": 4.432624113475177,
      "grad_norm": 0.48807811737060547,
      "learning_rate": 6.340758459818059e-05,
      "loss": 0.0514,
      "step": 1250
    },
    {
      "epoch": 4.468085106382979,
      "grad_norm": 0.7081537246704102,
      "learning_rate": 6.284181244704161e-05,
      "loss": 0.0595,
      "step": 1260
    },
    {
      "epoch": 4.50354609929078,
      "grad_norm": 0.6502603888511658,
      "learning_rate": 6.227427435703997e-05,
      "loss": 0.0629,
      "step": 1270
    },
    {
      "epoch": 4.539007092198582,
      "grad_norm": 0.9540507197380066,
      "learning_rate": 6.170504837304458e-05,
      "loss": 0.045,
      "step": 1280
    },
    {
      "epoch": 4.574468085106383,
      "grad_norm": 0.5669839382171631,
      "learning_rate": 6.11342127720347e-05,
      "loss": 0.059,
      "step": 1290
    },
    {
      "epoch": 4.609929078014185,
      "grad_norm": 0.7695325016975403,
      "learning_rate": 6.056184605233576e-05,
      "loss": 0.0558,
      "step": 1300
    },
    {
      "epoch": 4.6453900709219855,
      "grad_norm": 0.6503219604492188,
      "learning_rate": 5.9988026922824534e-05,
      "loss": 0.0528,
      "step": 1310
    },
    {
      "epoch": 4.680851063829787,
      "grad_norm": 0.8653897643089294,
      "learning_rate": 5.9412834292105676e-05,
      "loss": 0.0521,
      "step": 1320
    },
    {
      "epoch": 4.716312056737589,
      "grad_norm": 1.1591063737869263,
      "learning_rate": 5.883634725766048e-05,
      "loss": 0.0568,
      "step": 1330
    },
    {
      "epoch": 4.75177304964539,
      "grad_norm": 0.5562894940376282,
      "learning_rate": 5.825864509496991e-05,
      "loss": 0.0515,
      "step": 1340
    },
    {
      "epoch": 4.787234042553192,
      "grad_norm": 0.6316646337509155,
      "learning_rate": 5.767980724661295e-05,
      "loss": 0.045,
      "step": 1350
    },
    {
      "epoch": 4.822695035460993,
      "grad_norm": 0.8003104329109192,
      "learning_rate": 5.7099913311342234e-05,
      "loss": 0.0457,
      "step": 1360
    },
    {
      "epoch": 4.858156028368795,
      "grad_norm": 0.4407203793525696,
      "learning_rate": 5.651904303313784e-05,
      "loss": 0.055,
      "step": 1370
    },
    {
      "epoch": 4.8936170212765955,
      "grad_norm": 0.5941787362098694,
      "learning_rate": 5.593727629024148e-05,
      "loss": 0.0476,
      "step": 1380
    },
    {
      "epoch": 4.929078014184397,
      "grad_norm": 0.4574081599712372,
      "learning_rate": 5.535469308417198e-05,
      "loss": 0.0435,
      "step": 1390
    },
    {
      "epoch": 4.964539007092198,
      "grad_norm": 0.7371949553489685,
      "learning_rate": 5.4771373528723924e-05,
      "loss": 0.0576,
      "step": 1400
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.4565671682357788,
      "learning_rate": 5.418739783895078e-05,
      "loss": 0.0415,
      "step": 1410
    },
    {
      "epoch": 5.035460992907802,
      "grad_norm": 0.6150527000427246,
      "learning_rate": 5.360284632013421e-05,
      "loss": 0.0435,
      "step": 1420
    },
    {
      "epoch": 5.070921985815603,
      "grad_norm": 1.273917317390442,
      "learning_rate": 5.301779935674087e-05,
      "loss": 0.0475,
      "step": 1430
    },
    {
      "epoch": 5.1063829787234045,
      "grad_norm": 0.5376979112625122,
      "learning_rate": 5.243233740136833e-05,
      "loss": 0.0476,
      "step": 1440
    },
    {
      "epoch": 5.141843971631205,
      "grad_norm": 0.4023731052875519,
      "learning_rate": 5.1846540963681725e-05,
      "loss": 0.0441,
      "step": 1450
    },
    {
      "epoch": 5.177304964539007,
      "grad_norm": 0.630713939666748,
      "learning_rate": 5.1260490599342395e-05,
      "loss": 0.0454,
      "step": 1460
    },
    {
      "epoch": 5.212765957446808,
      "grad_norm": 0.9711371660232544,
      "learning_rate": 5.067426689893042e-05,
      "loss": 0.0485,
      "step": 1470
    },
    {
      "epoch": 5.24822695035461,
      "grad_norm": 0.5721754431724548,
      "learning_rate": 5.00879504768621e-05,
      "loss": 0.0482,
      "step": 1480
    },
    {
      "epoch": 5.283687943262412,
      "grad_norm": 1.25078547000885,
      "learning_rate": 4.9501621960304323e-05,
      "loss": 0.0495,
      "step": 1490
    },
    {
      "epoch": 5.319148936170213,
      "grad_norm": 0.5160683989524841,
      "learning_rate": 4.891536197808719e-05,
      "loss": 0.0433,
      "step": 1500
    },
    {
      "epoch": 5.3546099290780145,
      "grad_norm": 0.5063398480415344,
      "learning_rate": 4.832925114961629e-05,
      "loss": 0.0512,
      "step": 1510
    },
    {
      "epoch": 5.390070921985815,
      "grad_norm": 0.5604034066200256,
      "learning_rate": 4.774337007378633e-05,
      "loss": 0.042,
      "step": 1520
    },
    {
      "epoch": 5.425531914893617,
      "grad_norm": 0.5415745377540588,
      "learning_rate": 4.715779931789776e-05,
      "loss": 0.0451,
      "step": 1530
    },
    {
      "epoch": 5.460992907801418,
      "grad_norm": 0.43733975291252136,
      "learning_rate": 4.657261940657732e-05,
      "loss": 0.0512,
      "step": 1540
    },
    {
      "epoch": 5.49645390070922,
      "grad_norm": 0.39489609003067017,
      "learning_rate": 4.5987910810704925e-05,
      "loss": 0.0408,
      "step": 1550
    },
    {
      "epoch": 5.531914893617021,
      "grad_norm": 0.5629368424415588,
      "learning_rate": 4.540375393634762e-05,
      "loss": 0.0425,
      "step": 1560
    },
    {
      "epoch": 5.567375886524823,
      "grad_norm": 0.764885425567627,
      "learning_rate": 4.48202291137026e-05,
      "loss": 0.0531,
      "step": 1570
    },
    {
      "epoch": 5.602836879432624,
      "grad_norm": 0.6770176291465759,
      "learning_rate": 4.423741658605066e-05,
      "loss": 0.0433,
      "step": 1580
    },
    {
      "epoch": 5.638297872340425,
      "grad_norm": 0.8399835228919983,
      "learning_rate": 4.365539649872146e-05,
      "loss": 0.0353,
      "step": 1590
    },
    {
      "epoch": 5.673758865248227,
      "grad_norm": 0.8981763124465942,
      "learning_rate": 4.307424888807242e-05,
      "loss": 0.0488,
      "step": 1600
    },
    {
      "epoch": 5.709219858156028,
      "grad_norm": 0.8662058711051941,
      "learning_rate": 4.249405367048254e-05,
      "loss": 0.0495,
      "step": 1610
    },
    {
      "epoch": 5.74468085106383,
      "grad_norm": 0.6840683221817017,
      "learning_rate": 4.191489063136274e-05,
      "loss": 0.0502,
      "step": 1620
    },
    {
      "epoch": 5.780141843971631,
      "grad_norm": 0.511677622795105,
      "learning_rate": 4.133683941418411e-05,
      "loss": 0.0419,
      "step": 1630
    },
    {
      "epoch": 5.815602836879433,
      "grad_norm": 1.7016257047653198,
      "learning_rate": 4.0759979509525824e-05,
      "loss": 0.0469,
      "step": 1640
    },
    {
      "epoch": 5.851063829787234,
      "grad_norm": 0.4668084681034088,
      "learning_rate": 4.018439024414398e-05,
      "loss": 0.049,
      "step": 1650
    },
    {
      "epoch": 5.886524822695035,
      "grad_norm": 0.8224870562553406,
      "learning_rate": 3.9610150770063015e-05,
      "loss": 0.044,
      "step": 1660
    },
    {
      "epoch": 5.921985815602837,
      "grad_norm": 0.41005393862724304,
      "learning_rate": 3.903734005369115e-05,
      "loss": 0.0443,
      "step": 1670
    },
    {
      "epoch": 5.957446808510638,
      "grad_norm": 0.3030180335044861,
      "learning_rate": 3.8466036864961314e-05,
      "loss": 0.0381,
      "step": 1680
    },
    {
      "epoch": 5.99290780141844,
      "grad_norm": 1.0716583728790283,
      "learning_rate": 3.789631976649907e-05,
      "loss": 0.0426,
      "step": 1690
    },
    {
      "epoch": 6.028368794326241,
      "grad_norm": 0.5316461324691772,
      "learning_rate": 3.7328267102819225e-05,
      "loss": 0.0535,
      "step": 1700
    },
    {
      "epoch": 6.0638297872340425,
      "grad_norm": 0.3643524944782257,
      "learning_rate": 3.676195698955214e-05,
      "loss": 0.0413,
      "step": 1710
    },
    {
      "epoch": 6.099290780141844,
      "grad_norm": 0.4204830527305603,
      "learning_rate": 3.619746730270185e-05,
      "loss": 0.0381,
      "step": 1720
    },
    {
      "epoch": 6.134751773049645,
      "grad_norm": 0.42800143361091614,
      "learning_rate": 3.56348756679368e-05,
      "loss": 0.0377,
      "step": 1730
    },
    {
      "epoch": 6.170212765957447,
      "grad_norm": 0.446856290102005,
      "learning_rate": 3.5074259449915284e-05,
      "loss": 0.0441,
      "step": 1740
    },
    {
      "epoch": 6.205673758865248,
      "grad_norm": 0.5602040886878967,
      "learning_rate": 3.4515695741646584e-05,
      "loss": 0.0489,
      "step": 1750
    },
    {
      "epoch": 6.24113475177305,
      "grad_norm": 0.8359407782554626,
      "learning_rate": 3.3959261353889605e-05,
      "loss": 0.0471,
      "step": 1760
    },
    {
      "epoch": 6.276595744680851,
      "grad_norm": 0.41129371523857117,
      "learning_rate": 3.3405032804590244e-05,
      "loss": 0.0411,
      "step": 1770
    },
    {
      "epoch": 6.3120567375886525,
      "grad_norm": 0.5542031526565552,
      "learning_rate": 3.2853086308359025e-05,
      "loss": 0.0419,
      "step": 1780
    },
    {
      "epoch": 6.347517730496454,
      "grad_norm": 0.3114689886569977,
      "learning_rate": 3.230349776599044e-05,
      "loss": 0.0399,
      "step": 1790
    },
    {
      "epoch": 6.382978723404255,
      "grad_norm": 0.5866858959197998,
      "learning_rate": 3.1756342754025546e-05,
      "loss": 0.0392,
      "step": 1800
    },
    {
      "epoch": 6.418439716312057,
      "grad_norm": 0.7895340323448181,
      "learning_rate": 3.121169651435903e-05,
      "loss": 0.0364,
      "step": 1810
    },
    {
      "epoch": 6.453900709219858,
      "grad_norm": 0.6045717597007751,
      "learning_rate": 3.0669633943892295e-05,
      "loss": 0.0377,
      "step": 1820
    },
    {
      "epoch": 6.48936170212766,
      "grad_norm": 0.505060076713562,
      "learning_rate": 3.0130229584234115e-05,
      "loss": 0.0382,
      "step": 1830
    },
    {
      "epoch": 6.524822695035461,
      "grad_norm": 0.5746676325798035,
      "learning_rate": 2.9593557611450008e-05,
      "loss": 0.0332,
      "step": 1840
    },
    {
      "epoch": 6.560283687943262,
      "grad_norm": 0.3892854154109955,
      "learning_rate": 2.9059691825861922e-05,
      "loss": 0.0374,
      "step": 1850
    },
    {
      "epoch": 6.595744680851064,
      "grad_norm": 2.366185188293457,
      "learning_rate": 2.8528705641899667e-05,
      "loss": 0.0448,
      "step": 1860
    },
    {
      "epoch": 6.631205673758865,
      "grad_norm": 0.6133996248245239,
      "learning_rate": 2.8000672078005274e-05,
      "loss": 0.04,
      "step": 1870
    },
    {
      "epoch": 6.666666666666667,
      "grad_norm": 0.45607250928878784,
      "learning_rate": 2.7475663746591906e-05,
      "loss": 0.0465,
      "step": 1880
    },
    {
      "epoch": 6.702127659574468,
      "grad_norm": 0.6318036317825317,
      "learning_rate": 2.69537528440586e-05,
      "loss": 0.0382,
      "step": 1890
    },
    {
      "epoch": 6.73758865248227,
      "grad_norm": 0.409967839717865,
      "learning_rate": 2.6435011140862166e-05,
      "loss": 0.045,
      "step": 1900
    },
    {
      "epoch": 6.773049645390071,
      "grad_norm": 0.6712722778320312,
      "learning_rate": 2.59195099716477e-05,
      "loss": 0.0367,
      "step": 1910
    },
    {
      "epoch": 6.808510638297872,
      "grad_norm": 0.8447416424751282,
      "learning_rate": 2.5407320225439047e-05,
      "loss": 0.0379,
      "step": 1920
    },
    {
      "epoch": 6.843971631205674,
      "grad_norm": 1.3314627408981323,
      "learning_rate": 2.4898512335890422e-05,
      "loss": 0.0409,
      "step": 1930
    },
    {
      "epoch": 6.879432624113475,
      "grad_norm": 0.5814293026924133,
      "learning_rate": 2.439315627160085e-05,
      "loss": 0.0471,
      "step": 1940
    },
    {
      "epoch": 6.914893617021277,
      "grad_norm": 0.7060461640357971,
      "learning_rate": 2.3891321526492427e-05,
      "loss": 0.0429,
      "step": 1950
    },
    {
      "epoch": 6.950354609929078,
      "grad_norm": 0.8079804182052612,
      "learning_rate": 2.3393077110253835e-05,
      "loss": 0.0392,
      "step": 1960
    },
    {
      "epoch": 6.98581560283688,
      "grad_norm": 0.5584299564361572,
      "learning_rate": 2.2898491538850475e-05,
      "loss": 0.0432,
      "step": 1970
    },
    {
      "epoch": 7.0212765957446805,
      "grad_norm": 0.6511906385421753,
      "learning_rate": 2.2407632825102603e-05,
      "loss": 0.0376,
      "step": 1980
    },
    {
      "epoch": 7.056737588652482,
      "grad_norm": 0.30886492133140564,
      "learning_rate": 2.192056846933246e-05,
      "loss": 0.0338,
      "step": 1990
    },
    {
      "epoch": 7.092198581560283,
      "grad_norm": 0.432918518781662,
      "learning_rate": 2.1437365450082114e-05,
      "loss": 0.0436,
      "step": 2000
    },
    {
      "epoch": 7.127659574468085,
      "grad_norm": 0.3994809687137604,
      "learning_rate": 2.0958090214902733e-05,
      "loss": 0.0308,
      "step": 2010
    },
    {
      "epoch": 7.163120567375887,
      "grad_norm": 0.43582919239997864,
      "learning_rate": 2.0482808671217217e-05,
      "loss": 0.0375,
      "step": 2020
    },
    {
      "epoch": 7.198581560283688,
      "grad_norm": 0.5949472784996033,
      "learning_rate": 2.001158617725692e-05,
      "loss": 0.0387,
      "step": 2030
    },
    {
      "epoch": 7.23404255319149,
      "grad_norm": 0.3335230350494385,
      "learning_rate": 1.9544487533073887e-05,
      "loss": 0.0348,
      "step": 2040
    },
    {
      "epoch": 7.2695035460992905,
      "grad_norm": 0.6592429280281067,
      "learning_rate": 1.9081576971629927e-05,
      "loss": 0.0419,
      "step": 2050
    },
    {
      "epoch": 7.304964539007092,
      "grad_norm": 1.4891948699951172,
      "learning_rate": 1.8622918149963626e-05,
      "loss": 0.0347,
      "step": 2060
    },
    {
      "epoch": 7.340425531914893,
      "grad_norm": 0.45448461174964905,
      "learning_rate": 1.816857414043655e-05,
      "loss": 0.0367,
      "step": 2070
    },
    {
      "epoch": 7.375886524822695,
      "grad_norm": 0.38540035486221313,
      "learning_rate": 1.771860742205988e-05,
      "loss": 0.0319,
      "step": 2080
    },
    {
      "epoch": 7.411347517730497,
      "grad_norm": 1.0512962341308594,
      "learning_rate": 1.727307987190262e-05,
      "loss": 0.0425,
      "step": 2090
    },
    {
      "epoch": 7.446808510638298,
      "grad_norm": 0.8107235431671143,
      "learning_rate": 1.6832052756582584e-05,
      "loss": 0.038,
      "step": 2100
    },
    {
      "epoch": 7.4822695035460995,
      "grad_norm": 0.6066608428955078,
      "learning_rate": 1.6395586723841327e-05,
      "loss": 0.0326,
      "step": 2110
    },
    {
      "epoch": 7.5177304964539005,
      "grad_norm": 0.32806602120399475,
      "learning_rate": 1.5963741794204207e-05,
      "loss": 0.0328,
      "step": 2120
    },
    {
      "epoch": 7.553191489361702,
      "grad_norm": 0.5049278140068054,
      "learning_rate": 1.5536577352726607e-05,
      "loss": 0.0431,
      "step": 2130
    },
    {
      "epoch": 7.588652482269503,
      "grad_norm": 0.5972107648849487,
      "learning_rate": 1.5114152140827742e-05,
      "loss": 0.0339,
      "step": 2140
    },
    {
      "epoch": 7.624113475177305,
      "grad_norm": 0.5267333984375,
      "learning_rate": 1.4696524248212746e-05,
      "loss": 0.034,
      "step": 2150
    },
    {
      "epoch": 7.659574468085106,
      "grad_norm": 0.45704758167266846,
      "learning_rate": 1.4283751104884446e-05,
      "loss": 0.03,
      "step": 2160
    },
    {
      "epoch": 7.695035460992908,
      "grad_norm": 0.25589045882225037,
      "learning_rate": 1.3875889473245995e-05,
      "loss": 0.0324,
      "step": 2170
    },
    {
      "epoch": 7.7304964539007095,
      "grad_norm": 0.3722244203090668,
      "learning_rate": 1.3472995440295182e-05,
      "loss": 0.0309,
      "step": 2180
    },
    {
      "epoch": 7.76595744680851,
      "grad_norm": 0.4319707155227661,
      "learning_rate": 1.3075124409911581e-05,
      "loss": 0.0354,
      "step": 2190
    },
    {
      "epoch": 7.801418439716312,
      "grad_norm": 0.5372457504272461,
      "learning_rate": 1.2682331095237698e-05,
      "loss": 0.0374,
      "step": 2200
    },
    {
      "epoch": 7.836879432624113,
      "grad_norm": 0.3582875430583954,
      "learning_rate": 1.2294669511155193e-05,
      "loss": 0.0366,
      "step": 2210
    },
    {
      "epoch": 7.872340425531915,
      "grad_norm": 0.4589178264141083,
      "learning_rate": 1.1912192966856961e-05,
      "loss": 0.0356,
      "step": 2220
    },
    {
      "epoch": 7.907801418439716,
      "grad_norm": 0.4580075442790985,
      "learning_rate": 1.1534954058516356e-05,
      "loss": 0.0403,
      "step": 2230
    },
    {
      "epoch": 7.943262411347518,
      "grad_norm": 0.8937917351722717,
      "learning_rate": 1.1163004662054432e-05,
      "loss": 0.0308,
      "step": 2240
    },
    {
      "epoch": 7.9787234042553195,
      "grad_norm": 0.4087877869606018,
      "learning_rate": 1.0796395926006259e-05,
      "loss": 0.0366,
      "step": 2250
    },
    {
      "epoch": 8.01418439716312,
      "grad_norm": 0.4792381823062897,
      "learning_rate": 1.0435178264487206e-05,
      "loss": 0.0321,
      "step": 2260
    },
    {
      "epoch": 8.049645390070921,
      "grad_norm": 0.49857097864151,
      "learning_rate": 1.0079401350260287e-05,
      "loss": 0.0309,
      "step": 2270
    },
    {
      "epoch": 8.085106382978724,
      "grad_norm": 1.709592342376709,
      "learning_rate": 9.729114107905417e-06,
      "loss": 0.0379,
      "step": 2280
    },
    {
      "epoch": 8.120567375886525,
      "grad_norm": 0.8211184144020081,
      "learning_rate": 9.384364707091559e-06,
      "loss": 0.0375,
      "step": 2290
    },
    {
      "epoch": 8.156028368794326,
      "grad_norm": 0.4921300411224365,
      "learning_rate": 9.04520055595266e-06,
      "loss": 0.0385,
      "step": 2300
    },
    {
      "epoch": 8.191489361702128,
      "grad_norm": 0.9879147410392761,
      "learning_rate": 8.71166829456837e-06,
      "loss": 0.0295,
      "step": 2310
    },
    {
      "epoch": 8.22695035460993,
      "grad_norm": 0.9006680846214294,
      "learning_rate": 8.383813788550326e-06,
      "loss": 0.0368,
      "step": 2320
    },
    {
      "epoch": 8.26241134751773,
      "grad_norm": 0.5323331952095032,
      "learning_rate": 8.061682122734937e-06,
      "loss": 0.0418,
      "step": 2330
    },
    {
      "epoch": 8.297872340425531,
      "grad_norm": 0.37056395411491394,
      "learning_rate": 7.745317594983598e-06,
      "loss": 0.0347,
      "step": 2340
    },
    {
      "epoch": 8.333333333333334,
      "grad_norm": 0.399641215801239,
      "learning_rate": 7.434763710090992e-06,
      "loss": 0.0335,
      "step": 2350
    },
    {
      "epoch": 8.368794326241135,
      "grad_norm": 0.36943164467811584,
      "learning_rate": 7.130063173802637e-06,
      "loss": 0.0304,
      "step": 2360
    },
    {
      "epoch": 8.404255319148936,
      "grad_norm": 0.8294058442115784,
      "learning_rate": 6.8312578869422e-06,
      "loss": 0.0348,
      "step": 2370
    },
    {
      "epoch": 8.439716312056738,
      "grad_norm": 0.8177503347396851,
      "learning_rate": 6.538388939649443e-06,
      "loss": 0.0307,
      "step": 2380
    },
    {
      "epoch": 8.47517730496454,
      "grad_norm": 0.5157350301742554,
      "learning_rate": 6.251496605729773e-06,
      "loss": 0.0308,
      "step": 2390
    },
    {
      "epoch": 8.51063829787234,
      "grad_norm": 0.6393640041351318,
      "learning_rate": 5.970620337116012e-06,
      "loss": 0.0323,
      "step": 2400
    },
    {
      "epoch": 8.546099290780141,
      "grad_norm": 0.6659883260726929,
      "learning_rate": 5.695798758443133e-06,
      "loss": 0.0359,
      "step": 2410
    },
    {
      "epoch": 8.581560283687944,
      "grad_norm": 0.40989336371421814,
      "learning_rate": 5.427069661736872e-06,
      "loss": 0.034,
      "step": 2420
    },
    {
      "epoch": 8.617021276595745,
      "grad_norm": 0.6710225343704224,
      "learning_rate": 5.164470001216659e-06,
      "loss": 0.0368,
      "step": 2430
    },
    {
      "epoch": 8.652482269503546,
      "grad_norm": 0.3472451865673065,
      "learning_rate": 4.908035888213963e-06,
      "loss": 0.0335,
      "step": 2440
    },
    {
      "epoch": 8.687943262411348,
      "grad_norm": 0.5614815950393677,
      "learning_rate": 4.657802586206411e-06,
      "loss": 0.0333,
      "step": 2450
    },
    {
      "epoch": 8.72340425531915,
      "grad_norm": 0.8335484266281128,
      "learning_rate": 4.413804505968533e-06,
      "loss": 0.0364,
      "step": 2460
    },
    {
      "epoch": 8.75886524822695,
      "grad_norm": 0.3221122920513153,
      "learning_rate": 4.17607520083979e-06,
      "loss": 0.0342,
      "step": 2470
    },
    {
      "epoch": 8.794326241134751,
      "grad_norm": 0.5895943641662598,
      "learning_rate": 3.944647362110487e-06,
      "loss": 0.0305,
      "step": 2480
    },
    {
      "epoch": 8.829787234042554,
      "grad_norm": 0.9525559544563293,
      "learning_rate": 3.7195528145262337e-06,
      "loss": 0.0273,
      "step": 2490
    },
    {
      "epoch": 8.865248226950355,
      "grad_norm": 1.2005960941314697,
      "learning_rate": 3.5008225119115777e-06,
      "loss": 0.0447,
      "step": 2500
    },
    {
      "epoch": 8.900709219858156,
      "grad_norm": 0.5620400309562683,
      "learning_rate": 3.2884865329133986e-06,
      "loss": 0.0283,
      "step": 2510
    },
    {
      "epoch": 8.936170212765958,
      "grad_norm": 0.8444902896881104,
      "learning_rate": 3.082574076864636e-06,
      "loss": 0.0323,
      "step": 2520
    },
    {
      "epoch": 8.97163120567376,
      "grad_norm": 0.21826264262199402,
      "learning_rate": 2.88311345976896e-06,
      "loss": 0.0312,
      "step": 2530
    },
    {
      "epoch": 9.00709219858156,
      "grad_norm": 0.6866464614868164,
      "learning_rate": 2.6901321104069287e-06,
      "loss": 0.0279,
      "step": 2540
    },
    {
      "epoch": 9.042553191489361,
      "grad_norm": 0.5157074332237244,
      "learning_rate": 2.5036565665640444e-06,
      "loss": 0.0314,
      "step": 2550
    },
    {
      "epoch": 9.078014184397164,
      "grad_norm": 0.4928205907344818,
      "learning_rate": 2.3237124713815285e-06,
      "loss": 0.0304,
      "step": 2560
    },
    {
      "epoch": 9.113475177304965,
      "grad_norm": 0.6988965272903442,
      "learning_rate": 2.1503245698299314e-06,
      "loss": 0.035,
      "step": 2570
    },
    {
      "epoch": 9.148936170212766,
      "grad_norm": 0.5993123650550842,
      "learning_rate": 1.9835167053063376e-06,
      "loss": 0.036,
      "step": 2580
    },
    {
      "epoch": 9.184397163120567,
      "grad_norm": 0.429684579372406,
      "learning_rate": 1.8233118163555828e-06,
      "loss": 0.0333,
      "step": 2590
    },
    {
      "epoch": 9.21985815602837,
      "grad_norm": 0.3981442153453827,
      "learning_rate": 1.6697319335158613e-06,
      "loss": 0.0275,
      "step": 2600
    },
    {
      "epoch": 9.25531914893617,
      "grad_norm": 0.4685744345188141,
      "learning_rate": 1.5227981762891586e-06,
      "loss": 0.0322,
      "step": 2610
    },
    {
      "epoch": 9.290780141843971,
      "grad_norm": 0.32031360268592834,
      "learning_rate": 1.3825307502370488e-06,
      "loss": 0.0275,
      "step": 2620
    },
    {
      "epoch": 9.326241134751774,
      "grad_norm": 0.316394567489624,
      "learning_rate": 1.2489489442021274e-06,
      "loss": 0.0282,
      "step": 2630
    },
    {
      "epoch": 9.361702127659575,
      "grad_norm": 0.5381686091423035,
      "learning_rate": 1.1220711276554773e-06,
      "loss": 0.0342,
      "step": 2640
    },
    {
      "epoch": 9.397163120567376,
      "grad_norm": 0.5553480386734009,
      "learning_rate": 1.0019147481706625e-06,
      "loss": 0.0284,
      "step": 2650
    },
    {
      "epoch": 9.432624113475176,
      "grad_norm": 0.868281364440918,
      "learning_rate": 8.884963290243408e-07,
      "loss": 0.0314,
      "step": 2660
    },
    {
      "epoch": 9.46808510638298,
      "grad_norm": 0.31793954968452454,
      "learning_rate": 7.818314669241544e-07,
      "loss": 0.0319,
      "step": 2670
    },
    {
      "epoch": 9.50354609929078,
      "grad_norm": 0.6224974989891052,
      "learning_rate": 6.819348298638839e-07,
      "loss": 0.0291,
      "step": 2680
    },
    {
      "epoch": 9.539007092198581,
      "grad_norm": 0.646937370300293,
      "learning_rate": 5.888201551064288e-07,
      "loss": 0.0325,
      "step": 2690
    },
    {
      "epoch": 9.574468085106384,
      "grad_norm": 0.3932403326034546,
      "learning_rate": 5.025002472947071e-07,
      "loss": 0.032,
      "step": 2700
    },
    {
      "epoch": 9.609929078014185,
      "grad_norm": 0.3964429795742035,
      "learning_rate": 4.2298697669084785e-07,
      "loss": 0.0329,
      "step": 2710
    },
    {
      "epoch": 9.645390070921986,
      "grad_norm": 0.9335715770721436,
      "learning_rate": 3.502912775438183e-07,
      "loss": 0.0267,
      "step": 2720
    },
    {
      "epoch": 9.680851063829786,
      "grad_norm": 0.4232543408870697,
      "learning_rate": 2.8442314658584936e-07,
      "loss": 0.0398,
      "step": 2730
    },
    {
      "epoch": 9.71631205673759,
      "grad_norm": 0.8966019153594971,
      "learning_rate": 2.2539164165770177e-07,
      "loss": 0.03,
      "step": 2740
    },
    {
      "epoch": 9.75177304964539,
      "grad_norm": 1.149170160293579,
      "learning_rate": 1.732048804630959e-07,
      "loss": 0.0275,
      "step": 2750
    },
    {
      "epoch": 9.787234042553191,
      "grad_norm": 0.6613313555717468,
      "learning_rate": 1.2787003945239906e-07,
      "loss": 0.0251,
      "step": 2760
    },
    {
      "epoch": 9.822695035460994,
      "grad_norm": 0.5298823118209839,
      "learning_rate": 8.939335283577599e-08,
      "loss": 0.0296,
      "step": 2770
    },
    {
      "epoch": 9.858156028368795,
      "grad_norm": 0.3377326428890228,
      "learning_rate": 5.778011172586362e-08,
      "loss": 0.0348,
      "step": 2780
    },
    {
      "epoch": 9.893617021276595,
      "grad_norm": 0.6410892009735107,
      "learning_rate": 3.3034663410191904e-08,
      "loss": 0.0287,
      "step": 2790
    },
    {
      "epoch": 9.929078014184396,
      "grad_norm": 1.0103349685668945,
      "learning_rate": 1.51604107533454e-08,
      "loss": 0.0292,
      "step": 2800
    },
    {
      "epoch": 9.964539007092199,
      "grad_norm": 0.25607502460479736,
      "learning_rate": 4.159811729037566e-09,
      "loss": 0.0289,
      "step": 2810
    },
    {
      "epoch": 10.0,
      "grad_norm": 0.471729040145874,
      "learning_rate": 3.437908209780183e-11,
      "loss": 0.0342,
      "step": 2820
    },
    {
      "epoch": 10.0,
      "step": 2820,
      "total_flos": 0.0,
      "train_loss": 0.0648594191099735,
      "train_runtime": 2953.1296,
      "train_samples_per_second": 46.774,
      "train_steps_per_second": 0.955
    }
  ],
  "logging_steps": 10,
  "max_steps": 2820,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 1000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 49,
  "trial_name": null,
  "trial_params": null
}