{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9237875288683602,
  "eval_steps": 500,
  "global_step": 2000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.004618937644341801,
      "grad_norm": 10.975909233093262,
      "learning_rate": 3.4642032332563515e-07,
      "loss": 2.4942,
      "step": 10
    },
    {
      "epoch": 0.009237875288683603,
      "grad_norm": 9.186948776245117,
      "learning_rate": 9.237875288683603e-07,
      "loss": 2.7582,
      "step": 20
    },
    {
      "epoch": 0.013856812933025405,
      "grad_norm": 10.907584190368652,
      "learning_rate": 1.443418013856813e-06,
      "loss": 2.6814,
      "step": 30
    },
    {
      "epoch": 0.018475750577367205,
      "grad_norm": 9.531168937683105,
      "learning_rate": 2.0207852193995383e-06,
      "loss": 2.6982,
      "step": 40
    },
    {
      "epoch": 0.023094688221709007,
      "grad_norm": 14.727725982666016,
      "learning_rate": 2.5981524249422633e-06,
      "loss": 2.0218,
      "step": 50
    },
    {
      "epoch": 0.02771362586605081,
      "grad_norm": 8.314309120178223,
      "learning_rate": 3.117782909930716e-06,
      "loss": 1.5595,
      "step": 60
    },
    {
      "epoch": 0.03233256351039261,
      "grad_norm": 4.944284915924072,
      "learning_rate": 3.6951501154734412e-06,
      "loss": 1.02,
      "step": 70
    },
    {
      "epoch": 0.03695150115473441,
      "grad_norm": 10.882843971252441,
      "learning_rate": 4.272517321016166e-06,
      "loss": 1.0419,
      "step": 80
    },
    {
      "epoch": 0.04157043879907621,
      "grad_norm": 12.310320854187012,
      "learning_rate": 4.849884526558892e-06,
      "loss": 1.0801,
      "step": 90
    },
    {
      "epoch": 0.046189376443418015,
      "grad_norm": 6.49992036819458,
      "learning_rate": 5.427251732101616e-06,
      "loss": 0.8444,
      "step": 100
    },
    {
      "epoch": 0.050808314087759814,
      "grad_norm": 2.89493465423584,
      "learning_rate": 6.004618937644342e-06,
      "loss": 0.8884,
      "step": 110
    },
    {
      "epoch": 0.05542725173210162,
      "grad_norm": 3.966763734817505,
      "learning_rate": 6.581986143187067e-06,
      "loss": 0.8672,
      "step": 120
    },
    {
      "epoch": 0.06004618937644342,
      "grad_norm": 4.442293167114258,
      "learning_rate": 7.159353348729793e-06,
      "loss": 0.8037,
      "step": 130
    },
    {
      "epoch": 0.06466512702078522,
      "grad_norm": 2.506918430328369,
      "learning_rate": 7.736720554272519e-06,
      "loss": 0.724,
      "step": 140
    },
    {
      "epoch": 0.06928406466512702,
      "grad_norm": 5.733686447143555,
      "learning_rate": 8.314087759815242e-06,
      "loss": 0.7692,
      "step": 150
    },
    {
      "epoch": 0.07390300230946882,
      "grad_norm": 4.161188125610352,
      "learning_rate": 8.891454965357968e-06,
      "loss": 0.7413,
      "step": 160
    },
    {
      "epoch": 0.07852193995381063,
      "grad_norm": 3.9434962272644043,
      "learning_rate": 9.468822170900693e-06,
      "loss": 0.7386,
      "step": 170
    },
    {
      "epoch": 0.08314087759815242,
      "grad_norm": 2.9100701808929443,
      "learning_rate": 1.0046189376443418e-05,
      "loss": 0.6942,
      "step": 180
    },
    {
      "epoch": 0.08775981524249422,
      "grad_norm": 5.367318153381348,
      "learning_rate": 1.0623556581986144e-05,
      "loss": 0.8011,
      "step": 190
    },
    {
      "epoch": 0.09237875288683603,
      "grad_norm": 3.1690614223480225,
      "learning_rate": 1.1200923787528869e-05,
      "loss": 0.6816,
      "step": 200
    },
    {
      "epoch": 0.09699769053117784,
      "grad_norm": 4.35976505279541,
      "learning_rate": 1.1778290993071595e-05,
      "loss": 0.7408,
      "step": 210
    },
    {
      "epoch": 0.10161662817551963,
      "grad_norm": 3.330937623977661,
      "learning_rate": 1.235565819861432e-05,
      "loss": 0.7159,
      "step": 220
    },
    {
      "epoch": 0.10623556581986143,
      "grad_norm": 5.761129379272461,
      "learning_rate": 1.2933025404157046e-05,
      "loss": 0.6838,
      "step": 230
    },
    {
      "epoch": 0.11085450346420324,
      "grad_norm": 7.05668830871582,
      "learning_rate": 1.351039260969977e-05,
      "loss": 0.7135,
      "step": 240
    },
    {
      "epoch": 0.11547344110854503,
      "grad_norm": 3.7135939598083496,
      "learning_rate": 1.4087759815242497e-05,
      "loss": 0.6385,
      "step": 250
    },
    {
      "epoch": 0.12009237875288684,
      "grad_norm": 5.477907657623291,
      "learning_rate": 1.4665127020785218e-05,
      "loss": 0.6292,
      "step": 260
    },
    {
      "epoch": 0.12471131639722864,
      "grad_norm": 6.577059268951416,
      "learning_rate": 1.5242494226327944e-05,
      "loss": 0.6921,
      "step": 270
    },
    {
      "epoch": 0.12933025404157045,
      "grad_norm": 3.6328892707824707,
      "learning_rate": 1.581986143187067e-05,
      "loss": 0.6621,
      "step": 280
    },
    {
      "epoch": 0.13394919168591224,
      "grad_norm": 4.084783554077148,
      "learning_rate": 1.6397228637413393e-05,
      "loss": 0.6667,
      "step": 290
    },
    {
      "epoch": 0.13856812933025403,
      "grad_norm": 3.8719701766967773,
      "learning_rate": 1.697459584295612e-05,
      "loss": 0.6692,
      "step": 300
    },
    {
      "epoch": 0.14318706697459585,
      "grad_norm": 7.860931873321533,
      "learning_rate": 1.7551963048498846e-05,
      "loss": 0.6251,
      "step": 310
    },
    {
      "epoch": 0.14780600461893764,
      "grad_norm": 4.381837368011475,
      "learning_rate": 1.812933025404157e-05,
      "loss": 0.6297,
      "step": 320
    },
    {
      "epoch": 0.15242494226327943,
      "grad_norm": 3.7145886421203613,
      "learning_rate": 1.8706697459584295e-05,
      "loss": 0.6483,
      "step": 330
    },
    {
      "epoch": 0.15704387990762125,
      "grad_norm": 2.609006643295288,
      "learning_rate": 1.9284064665127023e-05,
      "loss": 0.6149,
      "step": 340
    },
    {
      "epoch": 0.16166281755196305,
      "grad_norm": 4.774081230163574,
      "learning_rate": 1.9861431870669748e-05,
      "loss": 0.6034,
      "step": 350
    },
    {
      "epoch": 0.16628175519630484,
      "grad_norm": 7.305100440979004,
      "learning_rate": 2.0438799076212473e-05,
      "loss": 0.6496,
      "step": 360
    },
    {
      "epoch": 0.17090069284064666,
      "grad_norm": 5.507181644439697,
      "learning_rate": 2.1016166281755197e-05,
      "loss": 0.643,
      "step": 370
    },
    {
      "epoch": 0.17551963048498845,
      "grad_norm": 4.033135890960693,
      "learning_rate": 2.1593533487297922e-05,
      "loss": 0.6186,
      "step": 380
    },
    {
      "epoch": 0.18013856812933027,
      "grad_norm": 3.903007745742798,
      "learning_rate": 2.217090069284065e-05,
      "loss": 0.6041,
      "step": 390
    },
    {
      "epoch": 0.18475750577367206,
      "grad_norm": 4.785562992095947,
      "learning_rate": 2.2748267898383374e-05,
      "loss": 0.5527,
      "step": 400
    },
    {
      "epoch": 0.18937644341801385,
      "grad_norm": 3.4289231300354004,
      "learning_rate": 2.3325635103926096e-05,
      "loss": 0.5936,
      "step": 410
    },
    {
      "epoch": 0.19399538106235567,
      "grad_norm": 2.384840965270996,
      "learning_rate": 2.3903002309468824e-05,
      "loss": 0.5421,
      "step": 420
    },
    {
      "epoch": 0.19861431870669746,
      "grad_norm": 4.025755882263184,
      "learning_rate": 2.448036951501155e-05,
      "loss": 0.5839,
      "step": 430
    },
    {
      "epoch": 0.20323325635103925,
      "grad_norm": 4.832013130187988,
      "learning_rate": 2.5057736720554276e-05,
      "loss": 0.5938,
      "step": 440
    },
    {
      "epoch": 0.20785219399538107,
      "grad_norm": 3.66886305809021,
      "learning_rate": 2.5635103926096998e-05,
      "loss": 0.5607,
      "step": 450
    },
    {
      "epoch": 0.21247113163972287,
      "grad_norm": 3.7285852432250977,
      "learning_rate": 2.6212471131639726e-05,
      "loss": 0.5457,
      "step": 460
    },
    {
      "epoch": 0.21709006928406466,
      "grad_norm": 3.755711555480957,
      "learning_rate": 2.678983833718245e-05,
      "loss": 0.5721,
      "step": 470
    },
    {
      "epoch": 0.22170900692840648,
      "grad_norm": 4.016116619110107,
      "learning_rate": 2.7367205542725178e-05,
      "loss": 0.59,
      "step": 480
    },
    {
      "epoch": 0.22632794457274827,
      "grad_norm": 6.123377799987793,
      "learning_rate": 2.79445727482679e-05,
      "loss": 0.6236,
      "step": 490
    },
    {
      "epoch": 0.23094688221709006,
      "grad_norm": 3.77093505859375,
      "learning_rate": 2.8521939953810624e-05,
      "loss": 0.6306,
      "step": 500
    },
    {
      "epoch": 0.23556581986143188,
      "grad_norm": 5.101199626922607,
      "learning_rate": 2.9099307159353352e-05,
      "loss": 0.5806,
      "step": 510
    },
    {
      "epoch": 0.24018475750577367,
      "grad_norm": 4.425400257110596,
      "learning_rate": 2.9676674364896073e-05,
      "loss": 0.5644,
      "step": 520
    },
    {
      "epoch": 0.24480369515011546,
      "grad_norm": 3.661381244659424,
      "learning_rate": 3.02540415704388e-05,
      "loss": 0.538,
      "step": 530
    },
    {
      "epoch": 0.24942263279445728,
      "grad_norm": 3.271655559539795,
      "learning_rate": 3.0831408775981526e-05,
      "loss": 0.5927,
      "step": 540
    },
    {
      "epoch": 0.2540415704387991,
      "grad_norm": 4.603795051574707,
      "learning_rate": 3.140877598152425e-05,
      "loss": 0.5772,
      "step": 550
    },
    {
      "epoch": 0.2586605080831409,
      "grad_norm": 3.379786968231201,
      "learning_rate": 3.1986143187066975e-05,
      "loss": 0.5577,
      "step": 560
    },
    {
      "epoch": 0.2632794457274827,
      "grad_norm": 3.3702409267425537,
      "learning_rate": 3.25635103926097e-05,
      "loss": 0.6258,
      "step": 570
    },
    {
      "epoch": 0.2678983833718245,
      "grad_norm": 3.1498706340789795,
      "learning_rate": 3.3140877598152424e-05,
      "loss": 0.5631,
      "step": 580
    },
    {
      "epoch": 0.27251732101616627,
      "grad_norm": 4.846761703491211,
      "learning_rate": 3.3718244803695156e-05,
      "loss": 0.5944,
      "step": 590
    },
    {
      "epoch": 0.27713625866050806,
      "grad_norm": 4.397818088531494,
      "learning_rate": 3.4295612009237874e-05,
      "loss": 0.5698,
      "step": 600
    },
    {
      "epoch": 0.2817551963048499,
      "grad_norm": 2.2724201679229736,
      "learning_rate": 3.4872979214780605e-05,
      "loss": 0.5468,
      "step": 610
    },
    {
      "epoch": 0.2863741339491917,
      "grad_norm": 4.73584508895874,
      "learning_rate": 3.545034642032333e-05,
      "loss": 0.5697,
      "step": 620
    },
    {
      "epoch": 0.2909930715935335,
      "grad_norm": 3.557711124420166,
      "learning_rate": 3.6027713625866054e-05,
      "loss": 0.5195,
      "step": 630
    },
    {
      "epoch": 0.2956120092378753,
      "grad_norm": 4.273311614990234,
      "learning_rate": 3.660508083140878e-05,
      "loss": 0.5561,
      "step": 640
    },
    {
      "epoch": 0.3002309468822171,
      "grad_norm": 3.6489686965942383,
      "learning_rate": 3.7182448036951504e-05,
      "loss": 0.566,
      "step": 650
    },
    {
      "epoch": 0.30484988452655887,
      "grad_norm": 3.4011149406433105,
      "learning_rate": 3.775981524249423e-05,
      "loss": 0.5468,
      "step": 660
    },
    {
      "epoch": 0.3094688221709007,
      "grad_norm": 3.517822742462158,
      "learning_rate": 3.833718244803695e-05,
      "loss": 0.5834,
      "step": 670
    },
    {
      "epoch": 0.3140877598152425,
      "grad_norm": 2.7577602863311768,
      "learning_rate": 3.891454965357968e-05,
      "loss": 0.5489,
      "step": 680
    },
    {
      "epoch": 0.3187066974595843,
      "grad_norm": 2.856598138809204,
      "learning_rate": 3.94919168591224e-05,
      "loss": 0.4906,
      "step": 690
    },
    {
      "epoch": 0.3233256351039261,
      "grad_norm": 4.136464595794678,
      "learning_rate": 4.0069284064665133e-05,
      "loss": 0.5705,
      "step": 700
    },
    {
      "epoch": 0.3279445727482679,
      "grad_norm": 4.121008396148682,
      "learning_rate": 4.064665127020785e-05,
      "loss": 0.6241,
      "step": 710
    },
    {
      "epoch": 0.3325635103926097,
      "grad_norm": 3.611814498901367,
      "learning_rate": 4.122401847575058e-05,
      "loss": 0.5034,
      "step": 720
    },
    {
      "epoch": 0.3371824480369515,
      "grad_norm": 3.4624834060668945,
      "learning_rate": 4.18013856812933e-05,
      "loss": 0.5643,
      "step": 730
    },
    {
      "epoch": 0.3418013856812933,
      "grad_norm": 3.3586976528167725,
      "learning_rate": 4.237875288683603e-05,
      "loss": 0.5439,
      "step": 740
    },
    {
      "epoch": 0.3464203233256351,
      "grad_norm": 2.5129313468933105,
      "learning_rate": 4.2956120092378757e-05,
      "loss": 0.5062,
      "step": 750
    },
    {
      "epoch": 0.3510392609699769,
      "grad_norm": 3.0052690505981445,
      "learning_rate": 4.353348729792148e-05,
      "loss": 0.5219,
      "step": 760
    },
    {
      "epoch": 0.3556581986143187,
      "grad_norm": 3.7070388793945312,
      "learning_rate": 4.4110854503464206e-05,
      "loss": 0.5425,
      "step": 770
    },
    {
      "epoch": 0.36027713625866054,
      "grad_norm": 2.8378756046295166,
      "learning_rate": 4.468822170900693e-05,
      "loss": 0.5208,
      "step": 780
    },
    {
      "epoch": 0.3648960739030023,
      "grad_norm": 2.3988196849823,
      "learning_rate": 4.5265588914549655e-05,
      "loss": 0.5149,
      "step": 790
    },
    {
      "epoch": 0.3695150115473441,
      "grad_norm": 2.2305569648742676,
      "learning_rate": 4.584295612009238e-05,
      "loss": 0.5028,
      "step": 800
    },
    {
      "epoch": 0.3741339491916859,
      "grad_norm": 3.0817391872406006,
      "learning_rate": 4.6420323325635104e-05,
      "loss": 0.5419,
      "step": 810
    },
    {
      "epoch": 0.3787528868360277,
      "grad_norm": 2.767381429672241,
      "learning_rate": 4.699769053117783e-05,
      "loss": 0.5036,
      "step": 820
    },
    {
      "epoch": 0.3833718244803695,
      "grad_norm": 2.8563129901885986,
      "learning_rate": 4.757505773672056e-05,
      "loss": 0.5079,
      "step": 830
    },
    {
      "epoch": 0.38799076212471134,
      "grad_norm": 4.459218978881836,
      "learning_rate": 4.815242494226328e-05,
      "loss": 0.4891,
      "step": 840
    },
    {
      "epoch": 0.39260969976905313,
      "grad_norm": 2.825631856918335,
      "learning_rate": 4.872979214780601e-05,
      "loss": 0.4921,
      "step": 850
    },
    {
      "epoch": 0.3972286374133949,
      "grad_norm": 2.835643768310547,
      "learning_rate": 4.9307159353348734e-05,
      "loss": 0.4736,
      "step": 860
    },
    {
      "epoch": 0.4018475750577367,
      "grad_norm": 3.237922430038452,
      "learning_rate": 4.988452655889146e-05,
      "loss": 0.5254,
      "step": 870
    },
    {
      "epoch": 0.4064665127020785,
      "grad_norm": 3.913771390914917,
      "learning_rate": 4.9999870022388165e-05,
      "loss": 0.5212,
      "step": 880
    },
    {
      "epoch": 0.4110854503464203,
      "grad_norm": 4.357654094696045,
      "learning_rate": 4.999934199065641e-05,
      "loss": 0.6004,
      "step": 890
    },
    {
      "epoch": 0.41570438799076215,
      "grad_norm": 2.1519246101379395,
      "learning_rate": 4.999840778977644e-05,
      "loss": 0.5106,
      "step": 900
    },
    {
      "epoch": 0.42032332563510394,
      "grad_norm": 3.196580410003662,
      "learning_rate": 4.9997067434926386e-05,
      "loss": 0.4708,
      "step": 910
    },
    {
      "epoch": 0.42494226327944573,
      "grad_norm": 3.0933518409729004,
      "learning_rate": 4.9995320947883265e-05,
      "loss": 0.5228,
      "step": 920
    },
    {
      "epoch": 0.4295612009237875,
      "grad_norm": 2.5883092880249023,
      "learning_rate": 4.999316835702259e-05,
      "loss": 0.54,
      "step": 930
    },
    {
      "epoch": 0.4341801385681293,
      "grad_norm": 2.738126277923584,
      "learning_rate": 4.9990609697317916e-05,
      "loss": 0.4965,
      "step": 940
    },
    {
      "epoch": 0.4387990762124711,
      "grad_norm": 2.903364419937134,
      "learning_rate": 4.998764501034028e-05,
      "loss": 0.4732,
      "step": 950
    },
    {
      "epoch": 0.44341801385681295,
      "grad_norm": 1.895765781402588,
      "learning_rate": 4.998427434425753e-05,
      "loss": 0.4857,
      "step": 960
    },
    {
      "epoch": 0.44803695150115475,
      "grad_norm": 5.765718936920166,
      "learning_rate": 4.998049775383353e-05,
      "loss": 0.4901,
      "step": 970
    },
    {
      "epoch": 0.45265588914549654,
      "grad_norm": 3.9257349967956543,
      "learning_rate": 4.997631530042727e-05,
      "loss": 0.493,
      "step": 980
    },
    {
      "epoch": 0.45727482678983833,
      "grad_norm": 2.3068912029266357,
      "learning_rate": 4.997172705199189e-05,
      "loss": 0.5061,
      "step": 990
    },
    {
      "epoch": 0.4618937644341801,
      "grad_norm": 2.966820001602173,
      "learning_rate": 4.996673308307355e-05,
      "loss": 0.4557,
      "step": 1000
    },
    {
      "epoch": 0.4665127020785219,
      "grad_norm": 3.704120397567749,
      "learning_rate": 4.996133347481021e-05,
      "loss": 0.5302,
      "step": 1010
    },
    {
      "epoch": 0.47113163972286376,
      "grad_norm": 2.066206216812134,
      "learning_rate": 4.9955528314930376e-05,
      "loss": 0.5035,
      "step": 1020
    },
    {
      "epoch": 0.47575057736720555,
      "grad_norm": 3.7750158309936523,
      "learning_rate": 4.9949317697751596e-05,
      "loss": 0.5327,
      "step": 1030
    },
    {
      "epoch": 0.48036951501154734,
      "grad_norm": 2.5127952098846436,
      "learning_rate": 4.9942701724178965e-05,
      "loss": 0.4929,
      "step": 1040
    },
    {
      "epoch": 0.48498845265588914,
      "grad_norm": 2.594442367553711,
      "learning_rate": 4.9935680501703485e-05,
      "loss": 0.4795,
      "step": 1050
    },
    {
      "epoch": 0.4896073903002309,
      "grad_norm": 4.830276966094971,
      "learning_rate": 4.992825414440032e-05,
      "loss": 0.5268,
      "step": 1060
    },
    {
      "epoch": 0.4942263279445728,
      "grad_norm": 2.404269218444824,
      "learning_rate": 4.9920422772926933e-05,
      "loss": 0.5199,
      "step": 1070
    },
    {
      "epoch": 0.49884526558891457,
      "grad_norm": 2.622856378555298,
      "learning_rate": 4.991218651452114e-05,
      "loss": 0.4928,
      "step": 1080
    },
    {
      "epoch": 0.5034642032332564,
      "grad_norm": 2.660313844680786,
      "learning_rate": 4.9903545502999014e-05,
      "loss": 0.516,
      "step": 1090
    },
    {
      "epoch": 0.5080831408775982,
      "grad_norm": 2.2060108184814453,
      "learning_rate": 4.9894499878752744e-05,
      "loss": 0.474,
      "step": 1100
    },
    {
      "epoch": 0.5127020785219399,
      "grad_norm": 2.774624824523926,
      "learning_rate": 4.988504978874835e-05,
      "loss": 0.5053,
      "step": 1110
    },
    {
      "epoch": 0.5173210161662818,
      "grad_norm": 2.0086095333099365,
      "learning_rate": 4.987519538652326e-05,
      "loss": 0.4478,
      "step": 1120
    },
    {
      "epoch": 0.5219399538106235,
      "grad_norm": 2.7462918758392334,
      "learning_rate": 4.986493683218386e-05,
      "loss": 0.4872,
      "step": 1130
    },
    {
      "epoch": 0.5265588914549654,
      "grad_norm": 2.2208642959594727,
      "learning_rate": 4.985427429240286e-05,
      "loss": 0.4617,
      "step": 1140
    },
    {
      "epoch": 0.5311778290993071,
      "grad_norm": 3.3714964389801025,
      "learning_rate": 4.984320794041662e-05,
      "loss": 0.5293,
      "step": 1150
    },
    {
      "epoch": 0.535796766743649,
      "grad_norm": 2.732804298400879,
      "learning_rate": 4.98317379560223e-05,
      "loss": 0.4582,
      "step": 1160
    },
    {
      "epoch": 0.5404157043879908,
      "grad_norm": 3.4411492347717285,
      "learning_rate": 4.981986452557495e-05,
      "loss": 0.5359,
      "step": 1170
    },
    {
      "epoch": 0.5450346420323325,
      "grad_norm": 2.5346171855926514,
      "learning_rate": 4.9807587841984494e-05,
      "loss": 0.438,
      "step": 1180
    },
    {
      "epoch": 0.5496535796766744,
      "grad_norm": 2.068873882293701,
      "learning_rate": 4.9794908104712586e-05,
      "loss": 0.4541,
      "step": 1190
    },
    {
      "epoch": 0.5542725173210161,
      "grad_norm": 3.7652342319488525,
      "learning_rate": 4.978182551976939e-05,
      "loss": 0.5372,
      "step": 1200
    },
    {
      "epoch": 0.558891454965358,
      "grad_norm": 2.7133548259735107,
      "learning_rate": 4.976834029971017e-05,
      "loss": 0.4949,
      "step": 1210
    },
    {
      "epoch": 0.5635103926096998,
      "grad_norm": 2.7093558311462402,
      "learning_rate": 4.975445266363191e-05,
      "loss": 0.4484,
      "step": 1220
    },
    {
      "epoch": 0.5681293302540416,
      "grad_norm": 2.551117420196533,
      "learning_rate": 4.974016283716974e-05,
      "loss": 0.4865,
      "step": 1230
    },
    {
      "epoch": 0.5727482678983834,
      "grad_norm": 3.5680108070373535,
      "learning_rate": 4.9725471052493225e-05,
      "loss": 0.512,
      "step": 1240
    },
    {
      "epoch": 0.5773672055427251,
      "grad_norm": 3.1270763874053955,
      "learning_rate": 4.9710377548302636e-05,
      "loss": 0.4658,
      "step": 1250
    },
    {
      "epoch": 0.581986143187067,
      "grad_norm": 2.7450294494628906,
      "learning_rate": 4.9694882569825045e-05,
      "loss": 0.5726,
      "step": 1260
    },
    {
      "epoch": 0.5866050808314087,
      "grad_norm": 2.192270040512085,
      "learning_rate": 4.967898636881039e-05,
      "loss": 0.456,
      "step": 1270
    },
    {
      "epoch": 0.5912240184757506,
      "grad_norm": 3.5508313179016113,
      "learning_rate": 4.9662689203527304e-05,
      "loss": 0.4363,
      "step": 1280
    },
    {
      "epoch": 0.5958429561200924,
      "grad_norm": 3.789588212966919,
      "learning_rate": 4.964599133875899e-05,
      "loss": 0.3894,
      "step": 1290
    },
    {
      "epoch": 0.6004618937644342,
      "grad_norm": 2.185136556625366,
      "learning_rate": 4.9628893045798905e-05,
      "loss": 0.4388,
      "step": 1300
    },
    {
      "epoch": 0.605080831408776,
      "grad_norm": 3.9445693492889404,
      "learning_rate": 4.961139460244631e-05,
      "loss": 0.4761,
      "step": 1310
    },
    {
      "epoch": 0.6096997690531177,
      "grad_norm": 2.027426242828369,
      "learning_rate": 4.95934962930018e-05,
      "loss": 0.4726,
      "step": 1320
    },
    {
      "epoch": 0.6143187066974596,
      "grad_norm": 2.555227518081665,
      "learning_rate": 4.957519840826268e-05,
      "loss": 0.4546,
      "step": 1330
    },
    {
      "epoch": 0.6189376443418014,
      "grad_norm": 3.800353765487671,
      "learning_rate": 4.955650124551823e-05,
      "loss": 0.4657,
      "step": 1340
    },
    {
      "epoch": 0.6235565819861432,
      "grad_norm": 1.877545952796936,
      "learning_rate": 4.953740510854485e-05,
      "loss": 0.46,
      "step": 1350
    },
    {
      "epoch": 0.628175519630485,
      "grad_norm": 2.205172538757324,
      "learning_rate": 4.9517910307601204e-05,
      "loss": 0.4845,
      "step": 1360
    },
    {
      "epoch": 0.6327944572748267,
      "grad_norm": 4.207221031188965,
      "learning_rate": 4.949801715942306e-05,
      "loss": 0.4784,
      "step": 1370
    },
    {
      "epoch": 0.6374133949191686,
      "grad_norm": 2.3010756969451904,
      "learning_rate": 4.947772598721828e-05,
      "loss": 0.4679,
      "step": 1380
    },
    {
      "epoch": 0.6420323325635104,
      "grad_norm": 1.776963233947754,
      "learning_rate": 4.9457037120661455e-05,
      "loss": 0.4405,
      "step": 1390
    },
    {
      "epoch": 0.6466512702078522,
      "grad_norm": 2.8734822273254395,
      "learning_rate": 4.9435950895888604e-05,
      "loss": 0.474,
      "step": 1400
    },
    {
      "epoch": 0.651270207852194,
      "grad_norm": 3.049509286880493,
      "learning_rate": 4.9414467655491695e-05,
      "loss": 0.4334,
      "step": 1410
    },
    {
      "epoch": 0.6558891454965358,
      "grad_norm": 3.274083375930786,
      "learning_rate": 4.9392587748513105e-05,
      "loss": 0.4519,
      "step": 1420
    },
    {
      "epoch": 0.6605080831408776,
      "grad_norm": 2.149959087371826,
      "learning_rate": 4.937031153043991e-05,
      "loss": 0.4625,
      "step": 1430
    },
    {
      "epoch": 0.6651270207852193,
      "grad_norm": 2.3562982082366943,
      "learning_rate": 4.934763936319814e-05,
      "loss": 0.4487,
      "step": 1440
    },
    {
      "epoch": 0.6697459584295612,
      "grad_norm": 2.5950613021850586,
      "learning_rate": 4.932457161514689e-05,
      "loss": 0.4979,
      "step": 1450
    },
    {
      "epoch": 0.674364896073903,
      "grad_norm": 2.4045233726501465,
      "learning_rate": 4.9301108661072315e-05,
      "loss": 0.4442,
      "step": 1460
    },
    {
      "epoch": 0.6789838337182448,
      "grad_norm": 2.833594799041748,
      "learning_rate": 4.9277250882181575e-05,
      "loss": 0.4408,
      "step": 1470
    },
    {
      "epoch": 0.6836027713625866,
      "grad_norm": 1.7960981130599976,
      "learning_rate": 4.9252998666096625e-05,
      "loss": 0.4672,
      "step": 1480
    },
    {
      "epoch": 0.6882217090069284,
      "grad_norm": 2.261974811553955,
      "learning_rate": 4.922835240684792e-05,
      "loss": 0.3947,
      "step": 1490
    },
    {
      "epoch": 0.6928406466512702,
      "grad_norm": 2.966799020767212,
      "learning_rate": 4.9203312504867994e-05,
      "loss": 0.4546,
      "step": 1500
    },
    {
      "epoch": 0.6974595842956121,
      "grad_norm": 2.1086158752441406,
      "learning_rate": 4.9177879366984985e-05,
      "loss": 0.4226,
      "step": 1510
    },
    {
      "epoch": 0.7020785219399538,
      "grad_norm": 2.4200305938720703,
      "learning_rate": 4.915205340641601e-05,
      "loss": 0.4481,
      "step": 1520
    },
    {
      "epoch": 0.7066974595842956,
      "grad_norm": 2.486839771270752,
      "learning_rate": 4.912583504276045e-05,
      "loss": 0.4594,
      "step": 1530
    },
    {
      "epoch": 0.7113163972286374,
      "grad_norm": 1.7007755041122437,
      "learning_rate": 4.9099224701993115e-05,
      "loss": 0.4472,
      "step": 1540
    },
    {
      "epoch": 0.7159353348729792,
      "grad_norm": 1.6047892570495605,
      "learning_rate": 4.907222281645739e-05,
      "loss": 0.4734,
      "step": 1550
    },
    {
      "epoch": 0.7205542725173211,
      "grad_norm": 3.141820192337036,
      "learning_rate": 4.904482982485813e-05,
      "loss": 0.4777,
      "step": 1560
    },
    {
      "epoch": 0.7251732101616628,
      "grad_norm": 2.6380691528320312,
      "learning_rate": 4.901704617225455e-05,
      "loss": 0.4438,
      "step": 1570
    },
    {
      "epoch": 0.7297921478060047,
      "grad_norm": 2.238063335418701,
      "learning_rate": 4.898887231005306e-05,
      "loss": 0.4424,
      "step": 1580
    },
    {
      "epoch": 0.7344110854503464,
      "grad_norm": 2.1911888122558594,
      "learning_rate": 4.896030869599983e-05,
      "loss": 0.437,
      "step": 1590
    },
    {
      "epoch": 0.7390300230946882,
      "grad_norm": 2.0892059803009033,
      "learning_rate": 4.893135579417345e-05,
      "loss": 0.4341,
      "step": 1600
    },
    {
      "epoch": 0.74364896073903,
      "grad_norm": 1.727277398109436,
      "learning_rate": 4.8902014074977295e-05,
      "loss": 0.4428,
      "step": 1610
    },
    {
      "epoch": 0.7482678983833718,
      "grad_norm": 2.6759467124938965,
      "learning_rate": 4.8872284015131965e-05,
      "loss": 0.411,
      "step": 1620
    },
    {
      "epoch": 0.7528868360277137,
      "grad_norm": 3.616370439529419,
      "learning_rate": 4.8842166097667505e-05,
      "loss": 0.4376,
      "step": 1630
    },
    {
      "epoch": 0.7575057736720554,
      "grad_norm": 1.77035653591156,
      "learning_rate": 4.881166081191555e-05,
      "loss": 0.4217,
      "step": 1640
    },
    {
      "epoch": 0.7621247113163973,
      "grad_norm": 2.7078516483306885,
      "learning_rate": 4.878076865350136e-05,
      "loss": 0.4451,
      "step": 1650
    },
    {
      "epoch": 0.766743648960739,
      "grad_norm": 1.7627867460250854,
      "learning_rate": 4.874949012433584e-05,
      "loss": 0.4399,
      "step": 1660
    },
    {
      "epoch": 0.7713625866050808,
      "grad_norm": 1.7273634672164917,
      "learning_rate": 4.871782573260729e-05,
      "loss": 0.4202,
      "step": 1670
    },
    {
      "epoch": 0.7759815242494227,
      "grad_norm": 2.0963988304138184,
      "learning_rate": 4.868577599277322e-05,
      "loss": 0.4447,
      "step": 1680
    },
    {
      "epoch": 0.7806004618937644,
      "grad_norm": 2.144355297088623,
      "learning_rate": 4.865334142555196e-05,
      "loss": 0.4387,
      "step": 1690
    },
    {
      "epoch": 0.7852193995381063,
      "grad_norm": 2.0202884674072266,
      "learning_rate": 4.862052255791419e-05,
      "loss": 0.4124,
      "step": 1700
    },
    {
      "epoch": 0.789838337182448,
      "grad_norm": 1.6518304347991943,
      "learning_rate": 4.858731992307441e-05,
      "loss": 0.4222,
      "step": 1710
    },
    {
      "epoch": 0.7944572748267898,
      "grad_norm": 1.3399207592010498,
      "learning_rate": 4.855373406048226e-05,
      "loss": 0.4713,
      "step": 1720
    },
    {
      "epoch": 0.7990762124711316,
      "grad_norm": 1.7980890274047852,
      "learning_rate": 4.8519765515813744e-05,
      "loss": 0.4484,
      "step": 1730
    },
    {
      "epoch": 0.8036951501154734,
      "grad_norm": 3.4402995109558105,
      "learning_rate": 4.8485414840962384e-05,
      "loss": 0.3777,
      "step": 1740
    },
    {
      "epoch": 0.8083140877598153,
      "grad_norm": 1.9399069547653198,
      "learning_rate": 4.845068259403024e-05,
      "loss": 0.3942,
      "step": 1750
    },
    {
      "epoch": 0.812933025404157,
      "grad_norm": 1.8960610628128052,
      "learning_rate": 4.841556933931886e-05,
      "loss": 0.3738,
      "step": 1760
    },
    {
      "epoch": 0.8175519630484989,
      "grad_norm": 2.6203131675720215,
      "learning_rate": 4.838007564732008e-05,
      "loss": 0.4725,
      "step": 1770
    },
    {
      "epoch": 0.8221709006928406,
      "grad_norm": 2.441143274307251,
      "learning_rate": 4.834420209470679e-05,
      "loss": 0.4287,
      "step": 1780
    },
    {
      "epoch": 0.8267898383371824,
      "grad_norm": 3.31255841255188,
      "learning_rate": 4.830794926432355e-05,
      "loss": 0.4215,
      "step": 1790
    },
    {
      "epoch": 0.8314087759815243,
      "grad_norm": 2.7491471767425537,
      "learning_rate": 4.827131774517709e-05,
      "loss": 0.4301,
      "step": 1800
    },
    {
      "epoch": 0.836027713625866,
      "grad_norm": 1.9549751281738281,
      "learning_rate": 4.8234308132426807e-05,
      "loss": 0.4379,
      "step": 1810
    },
    {
      "epoch": 0.8406466512702079,
      "grad_norm": 2.4840948581695557,
      "learning_rate": 4.819692102737505e-05,
      "loss": 0.4474,
      "step": 1820
    },
    {
      "epoch": 0.8452655889145496,
      "grad_norm": 2.865260601043701,
      "learning_rate": 4.815915703745734e-05,
      "loss": 0.4608,
      "step": 1830
    },
    {
      "epoch": 0.8498845265588915,
      "grad_norm": 2.3620669841766357,
      "learning_rate": 4.812101677623254e-05,
      "loss": 0.4311,
      "step": 1840
    },
    {
      "epoch": 0.8545034642032333,
      "grad_norm": 3.4287946224212646,
      "learning_rate": 4.808250086337284e-05,
      "loss": 0.4035,
      "step": 1850
    },
    {
      "epoch": 0.859122401847575,
      "grad_norm": 1.6560109853744507,
      "learning_rate": 4.8043609924653745e-05,
      "loss": 0.4004,
      "step": 1860
    },
    {
      "epoch": 0.8637413394919169,
      "grad_norm": 2.310784339904785,
      "learning_rate": 4.800434459194386e-05,
      "loss": 0.4219,
      "step": 1870
    },
    {
      "epoch": 0.8683602771362586,
      "grad_norm": 1.660062313079834,
      "learning_rate": 4.796470550319465e-05,
      "loss": 0.4102,
      "step": 1880
    },
    {
      "epoch": 0.8729792147806005,
      "grad_norm": 2.578974723815918,
      "learning_rate": 4.792469330243007e-05,
      "loss": 0.4375,
      "step": 1890
    },
    {
      "epoch": 0.8775981524249422,
      "grad_norm": 2.095388412475586,
      "learning_rate": 4.788430863973607e-05,
      "loss": 0.4385,
      "step": 1900
    },
    {
      "epoch": 0.8822170900692841,
      "grad_norm": 2.4962880611419678,
      "learning_rate": 4.7843552171250085e-05,
      "loss": 0.4399,
      "step": 1910
    },
    {
      "epoch": 0.8868360277136259,
      "grad_norm": 2.8000426292419434,
      "learning_rate": 4.780242455915036e-05,
      "loss": 0.4344,
      "step": 1920
    },
    {
      "epoch": 0.8914549653579676,
      "grad_norm": 2.0987462997436523,
      "learning_rate": 4.776092647164516e-05,
      "loss": 0.3961,
      "step": 1930
    },
    {
      "epoch": 0.8960739030023095,
      "grad_norm": 3.531191825866699,
      "learning_rate": 4.771905858296195e-05,
      "loss": 0.4463,
      "step": 1940
    },
    {
      "epoch": 0.9006928406466512,
      "grad_norm": 2.8484129905700684,
      "learning_rate": 4.767682157333643e-05,
      "loss": 0.4993,
      "step": 1950
    },
    {
      "epoch": 0.9053117782909931,
      "grad_norm": 2.164492607116699,
      "learning_rate": 4.7634216129001453e-05,
      "loss": 0.4545,
      "step": 1960
    },
    {
      "epoch": 0.9099307159353349,
      "grad_norm": 2.37625789642334,
      "learning_rate": 4.7591242942175924e-05,
      "loss": 0.3965,
      "step": 1970
    },
    {
      "epoch": 0.9145496535796767,
      "grad_norm": 2.488903284072876,
      "learning_rate": 4.7547902711053535e-05,
      "loss": 0.4459,
      "step": 1980
    },
    {
      "epoch": 0.9191685912240185,
      "grad_norm": 3.028672933578491,
      "learning_rate": 4.7504196139791405e-05,
      "loss": 0.452,
      "step": 1990
    },
    {
      "epoch": 0.9237875288683602,
      "grad_norm": 2.084078073501587,
      "learning_rate": 4.746012393849866e-05,
      "loss": 0.4078,
      "step": 2000
    }
  ],
  "logging_steps": 10,
  "max_steps": 8660,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 5.4846778836097434e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}