{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 1148,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.008710801393728223,
      "grad_norm": 96.23190307617188,
      "learning_rate": 2.0000000000000002e-07,
      "loss": 10.808,
      "step": 10
    },
    {
      "epoch": 0.017421602787456445,
      "grad_norm": 463.5845642089844,
      "learning_rate": 4.0000000000000003e-07,
      "loss": 10.8147,
      "step": 20
    },
    {
      "epoch": 0.02613240418118467,
      "grad_norm": 0.10578063130378723,
      "learning_rate": 6.000000000000001e-07,
      "loss": 10.8077,
      "step": 30
    },
    {
      "epoch": 0.03484320557491289,
      "grad_norm": 0.09780663996934891,
      "learning_rate": 8.000000000000001e-07,
      "loss": 10.8218,
      "step": 40
    },
    {
      "epoch": 0.04355400696864112,
      "grad_norm": 1469.5677490234375,
      "learning_rate": 1.0000000000000002e-06,
      "loss": 10.8079,
      "step": 50
    },
    {
      "epoch": 0.05226480836236934,
      "grad_norm": 0.09817048907279968,
      "learning_rate": 1.2000000000000002e-06,
      "loss": 10.8204,
      "step": 60
    },
    {
      "epoch": 0.06097560975609756,
      "grad_norm": 0.09422990679740906,
      "learning_rate": 1.4000000000000001e-06,
      "loss": 10.8159,
      "step": 70
    },
    {
      "epoch": 0.06968641114982578,
      "grad_norm": 0.09545309096574783,
      "learning_rate": 1.6000000000000001e-06,
      "loss": 10.811,
      "step": 80
    },
    {
      "epoch": 0.078397212543554,
      "grad_norm": 0.08887256681919098,
      "learning_rate": 1.8000000000000001e-06,
      "loss": 10.8048,
      "step": 90
    },
    {
      "epoch": 0.08710801393728224,
      "grad_norm": 0.0909484326839447,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 10.8126,
      "step": 100
    },
    {
      "epoch": 0.09581881533101046,
      "grad_norm": 0.09697296470403671,
      "learning_rate": 2.2e-06,
      "loss": 10.8056,
      "step": 110
    },
    {
      "epoch": 0.10452961672473868,
      "grad_norm": 0.09927204251289368,
      "learning_rate": 2.4000000000000003e-06,
      "loss": 11.0541,
      "step": 120
    },
    {
      "epoch": 0.1132404181184669,
      "grad_norm": 0.09392143785953522,
      "learning_rate": 2.6e-06,
      "loss": 10.8073,
      "step": 130
    },
    {
      "epoch": 0.12195121951219512,
      "grad_norm": 0.09451345354318619,
      "learning_rate": 2.8000000000000003e-06,
      "loss": 10.8041,
      "step": 140
    },
    {
      "epoch": 0.13066202090592335,
      "grad_norm": 0.09947695583105087,
      "learning_rate": 3e-06,
      "loss": 10.8141,
      "step": 150
    },
    {
      "epoch": 0.13937282229965156,
      "grad_norm": 0.09803537279367447,
      "learning_rate": 3.2000000000000003e-06,
      "loss": 10.805,
      "step": 160
    },
    {
      "epoch": 0.1480836236933798,
      "grad_norm": 0.09703430533409119,
      "learning_rate": 3.4000000000000005e-06,
      "loss": 10.8193,
      "step": 170
    },
    {
      "epoch": 0.156794425087108,
      "grad_norm": 0.09272942692041397,
      "learning_rate": 3.6000000000000003e-06,
      "loss": 10.8151,
      "step": 180
    },
    {
      "epoch": 0.16550522648083624,
      "grad_norm": 0.0950935110449791,
      "learning_rate": 3.8000000000000005e-06,
      "loss": 10.8051,
      "step": 190
    },
    {
      "epoch": 0.17421602787456447,
      "grad_norm": 0.10270467400550842,
      "learning_rate": 4.000000000000001e-06,
      "loss": 10.8043,
      "step": 200
    },
    {
      "epoch": 0.18292682926829268,
      "grad_norm": 0.09881923347711563,
      "learning_rate": 4.2000000000000004e-06,
      "loss": 10.8044,
      "step": 210
    },
    {
      "epoch": 0.1916376306620209,
      "grad_norm": 0.10348394513130188,
      "learning_rate": 4.4e-06,
      "loss": 10.8036,
      "step": 220
    },
    {
      "epoch": 0.20034843205574912,
      "grad_norm": 0.09483418613672256,
      "learning_rate": 4.600000000000001e-06,
      "loss": 10.8046,
      "step": 230
    },
    {
      "epoch": 0.20905923344947736,
      "grad_norm": 0.09620082378387451,
      "learning_rate": 4.800000000000001e-06,
      "loss": 10.807,
      "step": 240
    },
    {
      "epoch": 0.21777003484320556,
      "grad_norm": 0.09709980338811874,
      "learning_rate": 5e-06,
      "loss": 10.8058,
      "step": 250
    },
    {
      "epoch": 0.2264808362369338,
      "grad_norm": 0.09373420476913452,
      "learning_rate": 5.2e-06,
      "loss": 10.8044,
      "step": 260
    },
    {
      "epoch": 0.23519163763066203,
      "grad_norm": 0.09700135141611099,
      "learning_rate": 5.400000000000001e-06,
      "loss": 10.8168,
      "step": 270
    },
    {
      "epoch": 0.24390243902439024,
      "grad_norm": 0.09450981020927429,
      "learning_rate": 5.600000000000001e-06,
      "loss": 10.8043,
      "step": 280
    },
    {
      "epoch": 0.25261324041811845,
      "grad_norm": 0.09983480721712112,
      "learning_rate": 5.8e-06,
      "loss": 10.8037,
      "step": 290
    },
    {
      "epoch": 0.2613240418118467,
      "grad_norm": 0.09153986722230911,
      "learning_rate": 6e-06,
      "loss": 10.8041,
      "step": 300
    },
    {
      "epoch": 0.2700348432055749,
      "grad_norm": 0.09664902836084366,
      "learning_rate": 6.200000000000001e-06,
      "loss": 10.805,
      "step": 310
    },
    {
      "epoch": 0.2787456445993031,
      "grad_norm": 0.08636524528265,
      "learning_rate": 6.4000000000000006e-06,
      "loss": 10.8049,
      "step": 320
    },
    {
      "epoch": 0.2874564459930314,
      "grad_norm": 0.0945158526301384,
      "learning_rate": 6.600000000000001e-06,
      "loss": 10.8047,
      "step": 330
    },
    {
      "epoch": 0.2961672473867596,
      "grad_norm": 0.09447965025901794,
      "learning_rate": 6.800000000000001e-06,
      "loss": 10.8044,
      "step": 340
    },
    {
      "epoch": 0.3048780487804878,
      "grad_norm": 0.0949367955327034,
      "learning_rate": 7e-06,
      "loss": 10.8038,
      "step": 350
    },
    {
      "epoch": 0.313588850174216,
      "grad_norm": 0.10188119113445282,
      "learning_rate": 7.2000000000000005e-06,
      "loss": 10.8037,
      "step": 360
    },
    {
      "epoch": 0.32229965156794427,
      "grad_norm": 0.09333271533250809,
      "learning_rate": 7.4e-06,
      "loss": 10.8029,
      "step": 370
    },
    {
      "epoch": 0.3310104529616725,
      "grad_norm": 0.10046978294849396,
      "learning_rate": 7.600000000000001e-06,
      "loss": 10.8037,
      "step": 380
    },
    {
      "epoch": 0.3397212543554007,
      "grad_norm": 0.09508766233921051,
      "learning_rate": 7.800000000000002e-06,
      "loss": 10.8065,
      "step": 390
    },
    {
      "epoch": 0.34843205574912894,
      "grad_norm": 0.09448560327291489,
      "learning_rate": 8.000000000000001e-06,
      "loss": 10.8058,
      "step": 400
    },
    {
      "epoch": 0.35714285714285715,
      "grad_norm": 0.09986421465873718,
      "learning_rate": 8.2e-06,
      "loss": 14.4299,
      "step": 410
    },
    {
      "epoch": 0.36585365853658536,
      "grad_norm": 0.10298220068216324,
      "learning_rate": 8.400000000000001e-06,
      "loss": 10.8035,
      "step": 420
    },
    {
      "epoch": 0.37456445993031356,
      "grad_norm": 0.09236953407526016,
      "learning_rate": 8.6e-06,
      "loss": 10.8309,
      "step": 430
    },
    {
      "epoch": 0.3832752613240418,
      "grad_norm": 0.09000707417726517,
      "learning_rate": 8.8e-06,
      "loss": 10.8027,
      "step": 440
    },
    {
      "epoch": 0.39198606271777003,
      "grad_norm": 0.09554922580718994,
      "learning_rate": 9e-06,
      "loss": 10.8045,
      "step": 450
    },
    {
      "epoch": 0.40069686411149824,
      "grad_norm": 130.35623168945312,
      "learning_rate": 9.200000000000002e-06,
      "loss": 10.8034,
      "step": 460
    },
    {
      "epoch": 0.4094076655052265,
      "grad_norm": 0.09580319374799728,
      "learning_rate": 9.4e-06,
      "loss": 10.8022,
      "step": 470
    },
    {
      "epoch": 0.4181184668989547,
      "grad_norm": 0.10328342765569687,
      "learning_rate": 9.600000000000001e-06,
      "loss": 10.8087,
      "step": 480
    },
    {
      "epoch": 0.4268292682926829,
      "grad_norm": 0.09383571892976761,
      "learning_rate": 9.800000000000001e-06,
      "loss": 10.8027,
      "step": 490
    },
    {
      "epoch": 0.4355400696864111,
      "grad_norm": 61532.49609375,
      "learning_rate": 1e-05,
      "loss": 15.4887,
      "step": 500
    },
    {
      "epoch": 0.4442508710801394,
      "grad_norm": 0.09285067766904831,
      "learning_rate": 9.966032608695653e-06,
      "loss": 10.8028,
      "step": 510
    },
    {
      "epoch": 0.4529616724738676,
      "grad_norm": 0.09329842031002045,
      "learning_rate": 9.932065217391306e-06,
      "loss": 10.8969,
      "step": 520
    },
    {
      "epoch": 0.4616724738675958,
      "grad_norm": 0.09399975091218948,
      "learning_rate": 9.898097826086957e-06,
      "loss": 10.8297,
      "step": 530
    },
    {
      "epoch": 0.47038327526132406,
      "grad_norm": 1130.7960205078125,
      "learning_rate": 9.86413043478261e-06,
      "loss": 10.8358,
      "step": 540
    },
    {
      "epoch": 0.47909407665505227,
      "grad_norm": 0.0971442461013794,
      "learning_rate": 9.830163043478262e-06,
      "loss": 10.8641,
      "step": 550
    },
    {
      "epoch": 0.4878048780487805,
      "grad_norm": 0.09969327598810196,
      "learning_rate": 9.796195652173915e-06,
      "loss": 10.8124,
      "step": 560
    },
    {
      "epoch": 0.4965156794425087,
      "grad_norm": 0.0888582393527031,
      "learning_rate": 9.762228260869566e-06,
      "loss": 10.9576,
      "step": 570
    },
    {
      "epoch": 0.5052264808362369,
      "grad_norm": 0.09847419708967209,
      "learning_rate": 9.728260869565218e-06,
      "loss": 10.8022,
      "step": 580
    },
    {
      "epoch": 0.5139372822299652,
      "grad_norm": 0.0947929099202156,
      "learning_rate": 9.694293478260869e-06,
      "loss": 10.8011,
      "step": 590
    },
    {
      "epoch": 0.5226480836236934,
      "grad_norm": 9.681638717651367,
      "learning_rate": 9.660326086956523e-06,
      "loss": 10.8015,
      "step": 600
    },
    {
      "epoch": 0.5313588850174216,
      "grad_norm": 0.09560558199882507,
      "learning_rate": 9.626358695652174e-06,
      "loss": 13.2261,
      "step": 610
    },
    {
      "epoch": 0.5400696864111498,
      "grad_norm": 0.09392021596431732,
      "learning_rate": 9.592391304347827e-06,
      "loss": 10.802,
      "step": 620
    },
    {
      "epoch": 0.5487804878048781,
      "grad_norm": 0.10044202208518982,
      "learning_rate": 9.558423913043478e-06,
      "loss": 10.8005,
      "step": 630
    },
    {
      "epoch": 0.5574912891986062,
      "grad_norm": 0.09005896002054214,
      "learning_rate": 9.524456521739132e-06,
      "loss": 10.802,
      "step": 640
    },
    {
      "epoch": 0.5662020905923345,
      "grad_norm": 0.09686450660228729,
      "learning_rate": 9.490489130434783e-06,
      "loss": 10.801,
      "step": 650
    },
    {
      "epoch": 0.5749128919860628,
      "grad_norm": 0.09321983903646469,
      "learning_rate": 9.456521739130436e-06,
      "loss": 10.8008,
      "step": 660
    },
    {
      "epoch": 0.5836236933797909,
      "grad_norm": 0.0884283185005188,
      "learning_rate": 9.422554347826087e-06,
      "loss": 10.801,
      "step": 670
    },
    {
      "epoch": 0.5923344947735192,
      "grad_norm": 0.09928812086582184,
      "learning_rate": 9.388586956521741e-06,
      "loss": 10.7998,
      "step": 680
    },
    {
      "epoch": 0.6010452961672473,
      "grad_norm": 0.09434370696544647,
      "learning_rate": 9.354619565217392e-06,
      "loss": 10.8007,
      "step": 690
    },
    {
      "epoch": 0.6097560975609756,
      "grad_norm": 0.09228064864873886,
      "learning_rate": 9.320652173913044e-06,
      "loss": 10.8002,
      "step": 700
    },
    {
      "epoch": 0.6184668989547039,
      "grad_norm": 0.09412040561437607,
      "learning_rate": 9.286684782608695e-06,
      "loss": 10.8009,
      "step": 710
    },
    {
      "epoch": 0.627177700348432,
      "grad_norm": 0.0975504145026207,
      "learning_rate": 9.25271739130435e-06,
      "loss": 10.8002,
      "step": 720
    },
    {
      "epoch": 0.6358885017421603,
      "grad_norm": 0.10332240164279938,
      "learning_rate": 9.21875e-06,
      "loss": 10.8001,
      "step": 730
    },
    {
      "epoch": 0.6445993031358885,
      "grad_norm": 0.09318797290325165,
      "learning_rate": 9.184782608695653e-06,
      "loss": 10.8001,
      "step": 740
    },
    {
      "epoch": 0.6533101045296167,
      "grad_norm": 0.09171754121780396,
      "learning_rate": 9.150815217391306e-06,
      "loss": 10.7999,
      "step": 750
    },
    {
      "epoch": 0.662020905923345,
      "grad_norm": 0.0921366959810257,
      "learning_rate": 9.116847826086958e-06,
      "loss": 10.8003,
      "step": 760
    },
    {
      "epoch": 0.6707317073170732,
      "grad_norm": 0.09661433100700378,
      "learning_rate": 9.08288043478261e-06,
      "loss": 10.8001,
      "step": 770
    },
    {
      "epoch": 0.6794425087108014,
      "grad_norm": 0.09323884546756744,
      "learning_rate": 9.048913043478262e-06,
      "loss": 10.7997,
      "step": 780
    },
    {
      "epoch": 0.6881533101045296,
      "grad_norm": 0.08752495795488358,
      "learning_rate": 9.014945652173914e-06,
      "loss": 10.8,
      "step": 790
    },
    {
      "epoch": 0.6968641114982579,
      "grad_norm": 0.09087631851434708,
      "learning_rate": 8.980978260869567e-06,
      "loss": 10.798,
      "step": 800
    },
    {
      "epoch": 0.705574912891986,
      "grad_norm": 0.09523475170135498,
      "learning_rate": 8.947010869565218e-06,
      "loss": 10.7994,
      "step": 810
    },
    {
      "epoch": 0.7142857142857143,
      "grad_norm": 0.09918010234832764,
      "learning_rate": 8.91304347826087e-06,
      "loss": 10.7996,
      "step": 820
    },
    {
      "epoch": 0.7229965156794426,
      "grad_norm": 0.09539210051298141,
      "learning_rate": 8.879076086956523e-06,
      "loss": 10.7993,
      "step": 830
    },
    {
      "epoch": 0.7317073170731707,
      "grad_norm": 0.09296409040689468,
      "learning_rate": 8.845108695652174e-06,
      "loss": 10.7993,
      "step": 840
    },
    {
      "epoch": 0.740418118466899,
      "grad_norm": 0.08845611661672592,
      "learning_rate": 8.811141304347827e-06,
      "loss": 10.7993,
      "step": 850
    },
    {
      "epoch": 0.7491289198606271,
      "grad_norm": 0.0954870954155922,
      "learning_rate": 8.77717391304348e-06,
      "loss": 10.7981,
      "step": 860
    },
    {
      "epoch": 0.7578397212543554,
      "grad_norm": 0.09371493011713028,
      "learning_rate": 8.743206521739132e-06,
      "loss": 10.7984,
      "step": 870
    },
    {
      "epoch": 0.7665505226480837,
      "grad_norm": 0.09903618693351746,
      "learning_rate": 8.709239130434783e-06,
      "loss": 10.7987,
      "step": 880
    },
    {
      "epoch": 0.7752613240418118,
      "grad_norm": 0.08991510421037674,
      "learning_rate": 8.675271739130435e-06,
      "loss": 10.7976,
      "step": 890
    },
    {
      "epoch": 0.7839721254355401,
      "grad_norm": 0.09141765534877777,
      "learning_rate": 8.641304347826088e-06,
      "loss": 10.799,
      "step": 900
    },
    {
      "epoch": 0.7926829268292683,
      "grad_norm": 0.09670023620128632,
      "learning_rate": 8.60733695652174e-06,
      "loss": 10.7973,
      "step": 910
    },
    {
      "epoch": 0.8013937282229965,
      "grad_norm": 0.09337509423494339,
      "learning_rate": 8.573369565217391e-06,
      "loss": 10.7985,
      "step": 920
    },
    {
      "epoch": 0.8101045296167247,
      "grad_norm": 0.09618163853883743,
      "learning_rate": 8.539402173913044e-06,
      "loss": 10.7989,
      "step": 930
    },
    {
      "epoch": 0.818815331010453,
      "grad_norm": 0.09537822753190994,
      "learning_rate": 8.505434782608697e-06,
      "loss": 10.7981,
      "step": 940
    },
    {
      "epoch": 0.8275261324041812,
      "grad_norm": 0.09312313050031662,
      "learning_rate": 8.47146739130435e-06,
      "loss": 10.7984,
      "step": 950
    },
    {
      "epoch": 0.8362369337979094,
      "grad_norm": 0.09118187427520752,
      "learning_rate": 8.4375e-06,
      "loss": 10.7981,
      "step": 960
    },
    {
      "epoch": 0.8449477351916377,
      "grad_norm": 0.0919727087020874,
      "learning_rate": 8.403532608695653e-06,
      "loss": 10.7978,
      "step": 970
    },
    {
      "epoch": 0.8536585365853658,
      "grad_norm": 0.09706488996744156,
      "learning_rate": 8.369565217391305e-06,
      "loss": 10.7979,
      "step": 980
    },
    {
      "epoch": 0.8623693379790941,
      "grad_norm": 0.0944119319319725,
      "learning_rate": 8.335597826086958e-06,
      "loss": 10.7983,
      "step": 990
    },
    {
      "epoch": 0.8710801393728222,
      "grad_norm": 0.1016387864947319,
      "learning_rate": 8.301630434782609e-06,
      "loss": 10.7973,
      "step": 1000
    },
    {
      "epoch": 0.8797909407665505,
      "grad_norm": 0.10198845714330673,
      "learning_rate": 8.267663043478261e-06,
      "loss": 10.7984,
      "step": 1010
    },
    {
      "epoch": 0.8885017421602788,
      "grad_norm": 0.10102390497922897,
      "learning_rate": 8.233695652173914e-06,
      "loss": 10.7971,
      "step": 1020
    },
    {
      "epoch": 0.8972125435540069,
      "grad_norm": 0.0972791463136673,
      "learning_rate": 8.199728260869567e-06,
      "loss": 10.7971,
      "step": 1030
    },
    {
      "epoch": 0.9059233449477352,
      "grad_norm": 0.09343789517879486,
      "learning_rate": 8.165760869565218e-06,
      "loss": 10.7978,
      "step": 1040
    },
    {
      "epoch": 0.9146341463414634,
      "grad_norm": 0.09384151548147202,
      "learning_rate": 8.13179347826087e-06,
      "loss": 10.797,
      "step": 1050
    },
    {
      "epoch": 0.9233449477351916,
      "grad_norm": 0.09324201196432114,
      "learning_rate": 8.097826086956523e-06,
      "loss": 10.7969,
      "step": 1060
    },
    {
      "epoch": 0.9320557491289199,
      "grad_norm": 0.09341388940811157,
      "learning_rate": 8.063858695652175e-06,
      "loss": 10.7972,
      "step": 1070
    },
    {
      "epoch": 0.9407665505226481,
      "grad_norm": 0.09563016146421432,
      "learning_rate": 8.029891304347826e-06,
      "loss": 10.7978,
      "step": 1080
    },
    {
      "epoch": 0.9494773519163763,
      "grad_norm": 0.09752058237791061,
      "learning_rate": 7.995923913043479e-06,
      "loss": 10.7962,
      "step": 1090
    },
    {
      "epoch": 0.9581881533101045,
      "grad_norm": 0.09652635455131531,
      "learning_rate": 7.961956521739131e-06,
      "loss": 10.7968,
      "step": 1100
    },
    {
      "epoch": 0.9668989547038328,
      "grad_norm": 0.097730852663517,
      "learning_rate": 7.927989130434784e-06,
      "loss": 10.7967,
      "step": 1110
    },
    {
      "epoch": 0.975609756097561,
      "grad_norm": 0.09178122133016586,
      "learning_rate": 7.894021739130435e-06,
      "loss": 10.7963,
      "step": 1120
    },
    {
      "epoch": 0.9843205574912892,
      "grad_norm": 0.09718208760023117,
      "learning_rate": 7.860054347826088e-06,
      "loss": 10.7967,
      "step": 1130
    },
    {
      "epoch": 0.9930313588850174,
      "grad_norm": 0.09228435158729553,
      "learning_rate": 7.82608695652174e-06,
      "loss": 10.7973,
      "step": 1140
    }
  ],
  "logging_steps": 10,
  "max_steps": 3444,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "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": 565862839314432.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}