{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 2094,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.014344629729245113,
      "grad_norm": 3.3866311674749072,
      "learning_rate": 4.285714285714286e-06,
      "loss": 1.5266,
      "step": 10
    },
    {
      "epoch": 0.028689259458490227,
      "grad_norm": 1.8728384521077752,
      "learning_rate": 9.047619047619047e-06,
      "loss": 1.3066,
      "step": 20
    },
    {
      "epoch": 0.04303388918773534,
      "grad_norm": 1.1126841032649692,
      "learning_rate": 1.3809523809523811e-05,
      "loss": 1.1436,
      "step": 30
    },
    {
      "epoch": 0.05737851891698045,
      "grad_norm": 1.1207104465369435,
      "learning_rate": 1.8571428571428572e-05,
      "loss": 1.0535,
      "step": 40
    },
    {
      "epoch": 0.07172314864622557,
      "grad_norm": 1.1391226479086436,
      "learning_rate": 2.3333333333333336e-05,
      "loss": 1.0172,
      "step": 50
    },
    {
      "epoch": 0.08606777837547068,
      "grad_norm": 1.146545389971907,
      "learning_rate": 2.8095238095238096e-05,
      "loss": 0.9859,
      "step": 60
    },
    {
      "epoch": 0.1004124081047158,
      "grad_norm": 1.1888015259911366,
      "learning_rate": 3.285714285714286e-05,
      "loss": 0.9528,
      "step": 70
    },
    {
      "epoch": 0.1147570378339609,
      "grad_norm": 1.2903823343286538,
      "learning_rate": 3.761904761904762e-05,
      "loss": 0.9397,
      "step": 80
    },
    {
      "epoch": 0.129101667563206,
      "grad_norm": 1.2943713329149742,
      "learning_rate": 4.2380952380952385e-05,
      "loss": 0.9282,
      "step": 90
    },
    {
      "epoch": 0.14344629729245115,
      "grad_norm": 1.2718909502396083,
      "learning_rate": 4.714285714285714e-05,
      "loss": 0.9186,
      "step": 100
    },
    {
      "epoch": 0.15779092702169625,
      "grad_norm": 1.4791012823315328,
      "learning_rate": 5.1904761904761913e-05,
      "loss": 0.9074,
      "step": 110
    },
    {
      "epoch": 0.17213555675094136,
      "grad_norm": 1.336532460075517,
      "learning_rate": 5.666666666666667e-05,
      "loss": 0.9002,
      "step": 120
    },
    {
      "epoch": 0.1864801864801865,
      "grad_norm": 1.6094057366302217,
      "learning_rate": 6.142857142857143e-05,
      "loss": 0.8887,
      "step": 130
    },
    {
      "epoch": 0.2008248162094316,
      "grad_norm": 1.070539230983606,
      "learning_rate": 6.619047619047619e-05,
      "loss": 0.884,
      "step": 140
    },
    {
      "epoch": 0.2151694459386767,
      "grad_norm": 1.1200792510458928,
      "learning_rate": 7.095238095238096e-05,
      "loss": 0.8809,
      "step": 150
    },
    {
      "epoch": 0.2295140756679218,
      "grad_norm": 1.087537431311211,
      "learning_rate": 7.571428571428571e-05,
      "loss": 0.8799,
      "step": 160
    },
    {
      "epoch": 0.24385870539716695,
      "grad_norm": 0.9999716597796505,
      "learning_rate": 8.047619047619048e-05,
      "loss": 0.8802,
      "step": 170
    },
    {
      "epoch": 0.258203335126412,
      "grad_norm": 1.2050995451812405,
      "learning_rate": 8.523809523809524e-05,
      "loss": 0.8692,
      "step": 180
    },
    {
      "epoch": 0.2725479648556572,
      "grad_norm": 0.9483149209977663,
      "learning_rate": 9e-05,
      "loss": 0.8675,
      "step": 190
    },
    {
      "epoch": 0.2868925945849023,
      "grad_norm": 0.9047494721372505,
      "learning_rate": 9.476190476190476e-05,
      "loss": 0.8651,
      "step": 200
    },
    {
      "epoch": 0.3012372243141474,
      "grad_norm": 0.9614336061705705,
      "learning_rate": 9.952380952380953e-05,
      "loss": 0.8743,
      "step": 210
    },
    {
      "epoch": 0.3155818540433925,
      "grad_norm": 0.840798357905234,
      "learning_rate": 9.999436939807164e-05,
      "loss": 0.8831,
      "step": 220
    },
    {
      "epoch": 0.3299264837726376,
      "grad_norm": 0.894805005888125,
      "learning_rate": 9.99749072170404e-05,
      "loss": 0.8753,
      "step": 230
    },
    {
      "epoch": 0.3442711135018827,
      "grad_norm": 0.9621522988496558,
      "learning_rate": 9.994154935353517e-05,
      "loss": 0.8553,
      "step": 240
    },
    {
      "epoch": 0.3586157432311278,
      "grad_norm": 0.8338221919665476,
      "learning_rate": 9.98943050828164e-05,
      "loss": 0.8577,
      "step": 250
    },
    {
      "epoch": 0.372960372960373,
      "grad_norm": 0.8401540225438021,
      "learning_rate": 9.983318754130435e-05,
      "loss": 0.842,
      "step": 260
    },
    {
      "epoch": 0.3873050026896181,
      "grad_norm": 0.7967631358112257,
      "learning_rate": 9.975821372292653e-05,
      "loss": 0.8346,
      "step": 270
    },
    {
      "epoch": 0.4016496324188632,
      "grad_norm": 0.7572726108984663,
      "learning_rate": 9.966940447439245e-05,
      "loss": 0.8272,
      "step": 280
    },
    {
      "epoch": 0.4159942621481083,
      "grad_norm": 0.7862046452006615,
      "learning_rate": 9.956678448939718e-05,
      "loss": 0.8161,
      "step": 290
    },
    {
      "epoch": 0.4303388918773534,
      "grad_norm": 0.8007930832524255,
      "learning_rate": 9.945038230175509e-05,
      "loss": 0.8263,
      "step": 300
    },
    {
      "epoch": 0.4446835216065985,
      "grad_norm": 0.6882798627741967,
      "learning_rate": 9.932023027746602e-05,
      "loss": 0.8119,
      "step": 310
    },
    {
      "epoch": 0.4590281513358436,
      "grad_norm": 0.7191586339334445,
      "learning_rate": 9.917636460571578e-05,
      "loss": 0.8195,
      "step": 320
    },
    {
      "epoch": 0.47337278106508873,
      "grad_norm": 0.6479226154771014,
      "learning_rate": 9.901882528881363e-05,
      "loss": 0.8152,
      "step": 330
    },
    {
      "epoch": 0.4877174107943339,
      "grad_norm": 0.7520756414640943,
      "learning_rate": 9.884765613106948e-05,
      "loss": 0.8056,
      "step": 340
    },
    {
      "epoch": 0.502062040523579,
      "grad_norm": 0.7496882676872738,
      "learning_rate": 9.866290472661406e-05,
      "loss": 0.8143,
      "step": 350
    },
    {
      "epoch": 0.516406670252824,
      "grad_norm": 0.8849980877887388,
      "learning_rate": 9.846462244616508e-05,
      "loss": 0.8139,
      "step": 360
    },
    {
      "epoch": 0.5307512999820692,
      "grad_norm": 0.689970676364689,
      "learning_rate": 9.825286442274357e-05,
      "loss": 0.8052,
      "step": 370
    },
    {
      "epoch": 0.5450959297113144,
      "grad_norm": 0.7192552702117557,
      "learning_rate": 9.802768953634388e-05,
      "loss": 0.7918,
      "step": 380
    },
    {
      "epoch": 0.5594405594405595,
      "grad_norm": 0.6975024662772501,
      "learning_rate": 9.778916039756193e-05,
      "loss": 0.7859,
      "step": 390
    },
    {
      "epoch": 0.5737851891698046,
      "grad_norm": 0.6553934247634706,
      "learning_rate": 9.753734333018616e-05,
      "loss": 0.776,
      "step": 400
    },
    {
      "epoch": 0.5881298188990497,
      "grad_norm": 0.757809076489788,
      "learning_rate": 9.727230835275598e-05,
      "loss": 0.7811,
      "step": 410
    },
    {
      "epoch": 0.6024744486282948,
      "grad_norm": 0.7082602077432336,
      "learning_rate": 9.699412915909284e-05,
      "loss": 0.7819,
      "step": 420
    },
    {
      "epoch": 0.6168190783575399,
      "grad_norm": 0.6348735880274488,
      "learning_rate": 9.670288309780953e-05,
      "loss": 0.7679,
      "step": 430
    },
    {
      "epoch": 0.631163708086785,
      "grad_norm": 0.6405240862226912,
      "learning_rate": 9.639865115080304e-05,
      "loss": 0.779,
      "step": 440
    },
    {
      "epoch": 0.6455083378160301,
      "grad_norm": 0.6634767779062684,
      "learning_rate": 9.608151791073737e-05,
      "loss": 0.7739,
      "step": 450
    },
    {
      "epoch": 0.6598529675452752,
      "grad_norm": 0.5759274474197833,
      "learning_rate": 9.575157155752222e-05,
      "loss": 0.7643,
      "step": 460
    },
    {
      "epoch": 0.6741975972745203,
      "grad_norm": 0.6640199963749864,
      "learning_rate": 9.54089038337943e-05,
      "loss": 0.7556,
      "step": 470
    },
    {
      "epoch": 0.6885422270037654,
      "grad_norm": 0.6203004765144989,
      "learning_rate": 9.5053610019408e-05,
      "loss": 0.7596,
      "step": 480
    },
    {
      "epoch": 0.7028868567330105,
      "grad_norm": 0.5775087991410168,
      "learning_rate": 9.468578890494256e-05,
      "loss": 0.7534,
      "step": 490
    },
    {
      "epoch": 0.7172314864622557,
      "grad_norm": 0.6714966253079345,
      "learning_rate": 9.430554276423292e-05,
      "loss": 0.7552,
      "step": 500
    },
    {
      "epoch": 0.7315761161915008,
      "grad_norm": 0.5888288333456809,
      "learning_rate": 9.391297732593229e-05,
      "loss": 0.7526,
      "step": 510
    },
    {
      "epoch": 0.745920745920746,
      "grad_norm": 0.6259226612543665,
      "learning_rate": 9.350820174411386e-05,
      "loss": 0.7537,
      "step": 520
    },
    {
      "epoch": 0.7602653756499911,
      "grad_norm": 0.5499968625723676,
      "learning_rate": 9.309132856792023e-05,
      "loss": 0.7499,
      "step": 530
    },
    {
      "epoch": 0.7746100053792362,
      "grad_norm": 0.6500285934875327,
      "learning_rate": 9.266247371026873e-05,
      "loss": 0.7447,
      "step": 540
    },
    {
      "epoch": 0.7889546351084813,
      "grad_norm": 0.5672171674827838,
      "learning_rate": 9.222175641562143e-05,
      "loss": 0.7496,
      "step": 550
    },
    {
      "epoch": 0.8032992648377264,
      "grad_norm": 0.6465733807975256,
      "learning_rate": 9.176929922682891e-05,
      "loss": 0.74,
      "step": 560
    },
    {
      "epoch": 0.8176438945669715,
      "grad_norm": 0.6101304041389433,
      "learning_rate": 9.130522795105676e-05,
      "loss": 0.7411,
      "step": 570
    },
    {
      "epoch": 0.8319885242962166,
      "grad_norm": 0.5489063036723607,
      "learning_rate": 9.082967162480459e-05,
      "loss": 0.7323,
      "step": 580
    },
    {
      "epoch": 0.8463331540254617,
      "grad_norm": 0.5516558559909698,
      "learning_rate": 9.034276247802688e-05,
      "loss": 0.7231,
      "step": 590
    },
    {
      "epoch": 0.8606777837547068,
      "grad_norm": 0.6477118047968917,
      "learning_rate": 8.984463589736614e-05,
      "loss": 0.7288,
      "step": 600
    },
    {
      "epoch": 0.8750224134839519,
      "grad_norm": 0.5962486006428408,
      "learning_rate": 8.933543038850816e-05,
      "loss": 0.732,
      "step": 610
    },
    {
      "epoch": 0.889367043213197,
      "grad_norm": 0.5713715183887016,
      "learning_rate": 8.881528753767007e-05,
      "loss": 0.728,
      "step": 620
    },
    {
      "epoch": 0.9037116729424421,
      "grad_norm": 0.6301745705273146,
      "learning_rate": 8.82843519722319e-05,
      "loss": 0.7263,
      "step": 630
    },
    {
      "epoch": 0.9180563026716873,
      "grad_norm": 0.5819941044165688,
      "learning_rate": 8.774277132052237e-05,
      "loss": 0.7189,
      "step": 640
    },
    {
      "epoch": 0.9324009324009324,
      "grad_norm": 0.5336149445366645,
      "learning_rate": 8.719069617077046e-05,
      "loss": 0.7254,
      "step": 650
    },
    {
      "epoch": 0.9467455621301775,
      "grad_norm": 0.6083401813961216,
      "learning_rate": 8.662828002923378e-05,
      "loss": 0.7218,
      "step": 660
    },
    {
      "epoch": 0.9610901918594227,
      "grad_norm": 0.5371262960983777,
      "learning_rate": 8.605567927751576e-05,
      "loss": 0.719,
      "step": 670
    },
    {
      "epoch": 0.9754348215886678,
      "grad_norm": 0.5551868136565808,
      "learning_rate": 8.547305312908318e-05,
      "loss": 0.7089,
      "step": 680
    },
    {
      "epoch": 0.9897794513179129,
      "grad_norm": 0.5899796664828114,
      "learning_rate": 8.48805635849964e-05,
      "loss": 0.7075,
      "step": 690
    },
    {
      "epoch": 1.002868925945849,
      "grad_norm": 0.608174630463498,
      "learning_rate": 8.427837538886437e-05,
      "loss": 0.6894,
      "step": 700
    },
    {
      "epoch": 1.017213555675094,
      "grad_norm": 0.6116990279567732,
      "learning_rate": 8.366665598103727e-05,
      "loss": 0.5865,
      "step": 710
    },
    {
      "epoch": 1.0315581854043392,
      "grad_norm": 0.5768126568975168,
      "learning_rate": 8.304557545204908e-05,
      "loss": 0.5931,
      "step": 720
    },
    {
      "epoch": 1.0459028151335843,
      "grad_norm": 0.5727241793491287,
      "learning_rate": 8.241530649532339e-05,
      "loss": 0.5859,
      "step": 730
    },
    {
      "epoch": 1.0602474448628294,
      "grad_norm": 0.5834066080830065,
      "learning_rate": 8.177602435915546e-05,
      "loss": 0.5833,
      "step": 740
    },
    {
      "epoch": 1.0745920745920745,
      "grad_norm": 0.6168394884857555,
      "learning_rate": 8.11279067979839e-05,
      "loss": 0.5805,
      "step": 750
    },
    {
      "epoch": 1.0889367043213196,
      "grad_norm": 0.4885078172344663,
      "learning_rate": 8.04711340229654e-05,
      "loss": 0.5733,
      "step": 760
    },
    {
      "epoch": 1.1032813340505647,
      "grad_norm": 0.5289947493535304,
      "learning_rate": 7.980588865186649e-05,
      "loss": 0.5812,
      "step": 770
    },
    {
      "epoch": 1.11762596377981,
      "grad_norm": 0.5672904537751536,
      "learning_rate": 7.913235565828613e-05,
      "loss": 0.5811,
      "step": 780
    },
    {
      "epoch": 1.1319705935090552,
      "grad_norm": 0.6107186114378813,
      "learning_rate": 7.845072232022311e-05,
      "loss": 0.5755,
      "step": 790
    },
    {
      "epoch": 1.1463152232383003,
      "grad_norm": 0.6696900109358123,
      "learning_rate": 7.776117816800288e-05,
      "loss": 0.5916,
      "step": 800
    },
    {
      "epoch": 1.1606598529675454,
      "grad_norm": 0.5412095989289454,
      "learning_rate": 7.706391493157805e-05,
      "loss": 0.5822,
      "step": 810
    },
    {
      "epoch": 1.1750044826967905,
      "grad_norm": 0.5408330429359739,
      "learning_rate": 7.635912648721718e-05,
      "loss": 0.5811,
      "step": 820
    },
    {
      "epoch": 1.1893491124260356,
      "grad_norm": 0.5952577265784902,
      "learning_rate": 7.564700880359696e-05,
      "loss": 0.5799,
      "step": 830
    },
    {
      "epoch": 1.2036937421552807,
      "grad_norm": 0.5026109808475954,
      "learning_rate": 7.492775988731243e-05,
      "loss": 0.5673,
      "step": 840
    },
    {
      "epoch": 1.2180383718845258,
      "grad_norm": 0.5302543029902266,
      "learning_rate": 7.420157972782063e-05,
      "loss": 0.5754,
      "step": 850
    },
    {
      "epoch": 1.232383001613771,
      "grad_norm": 0.5532266667153597,
      "learning_rate": 7.346867024183291e-05,
      "loss": 0.579,
      "step": 860
    },
    {
      "epoch": 1.246727631343016,
      "grad_norm": 0.5669382050811871,
      "learning_rate": 7.272923521717133e-05,
      "loss": 0.5756,
      "step": 870
    },
    {
      "epoch": 1.2610722610722611,
      "grad_norm": 0.5338735116713776,
      "learning_rate": 7.198348025610481e-05,
      "loss": 0.5769,
      "step": 880
    },
    {
      "epoch": 1.2754168908015062,
      "grad_norm": 0.5863122040109188,
      "learning_rate": 7.12316127181808e-05,
      "loss": 0.5685,
      "step": 890
    },
    {
      "epoch": 1.2897615205307513,
      "grad_norm": 0.5541827820189409,
      "learning_rate": 7.047384166256815e-05,
      "loss": 0.5801,
      "step": 900
    },
    {
      "epoch": 1.3041061502599964,
      "grad_norm": 0.5753955062923616,
      "learning_rate": 6.971037778992775e-05,
      "loss": 0.5795,
      "step": 910
    },
    {
      "epoch": 1.3184507799892415,
      "grad_norm": 0.47433414173616173,
      "learning_rate": 6.894143338382639e-05,
      "loss": 0.5768,
      "step": 920
    },
    {
      "epoch": 1.3327954097184866,
      "grad_norm": 0.5274791181928579,
      "learning_rate": 6.81672222517107e-05,
      "loss": 0.578,
      "step": 930
    },
    {
      "epoch": 1.3471400394477318,
      "grad_norm": 0.5774970902851654,
      "learning_rate": 6.73879596654573e-05,
      "loss": 0.5682,
      "step": 940
    },
    {
      "epoch": 1.3614846691769769,
      "grad_norm": 0.5662099451295628,
      "learning_rate": 6.660386230151571e-05,
      "loss": 0.5735,
      "step": 950
    },
    {
      "epoch": 1.375829298906222,
      "grad_norm": 0.5574084243877258,
      "learning_rate": 6.581514818066088e-05,
      "loss": 0.5739,
      "step": 960
    },
    {
      "epoch": 1.390173928635467,
      "grad_norm": 0.539589642853108,
      "learning_rate": 6.502203660737169e-05,
      "loss": 0.5688,
      "step": 970
    },
    {
      "epoch": 1.4045185583647122,
      "grad_norm": 0.49293926542133193,
      "learning_rate": 6.422474810885278e-05,
      "loss": 0.5669,
      "step": 980
    },
    {
      "epoch": 1.4188631880939573,
      "grad_norm": 0.49620916916034896,
      "learning_rate": 6.342350437371614e-05,
      "loss": 0.5702,
      "step": 990
    },
    {
      "epoch": 1.4332078178232024,
      "grad_norm": 0.5152950617397968,
      "learning_rate": 6.26185281903399e-05,
      "loss": 0.5685,
      "step": 1000
    },
    {
      "epoch": 1.4475524475524475,
      "grad_norm": 0.545911912027696,
      "learning_rate": 6.181004338492141e-05,
      "loss": 0.5651,
      "step": 1010
    },
    {
      "epoch": 1.4618970772816926,
      "grad_norm": 0.5074263813146201,
      "learning_rate": 6.09982747592415e-05,
      "loss": 0.5615,
      "step": 1020
    },
    {
      "epoch": 1.4762417070109377,
      "grad_norm": 0.5094909684862107,
      "learning_rate": 6.018344802815778e-05,
      "loss": 0.5679,
      "step": 1030
    },
    {
      "epoch": 1.4905863367401828,
      "grad_norm": 0.5103259058296251,
      "learning_rate": 5.936578975684378e-05,
      "loss": 0.5701,
      "step": 1040
    },
    {
      "epoch": 1.504930966469428,
      "grad_norm": 0.5278709818430807,
      "learning_rate": 5.854552729779184e-05,
      "loss": 0.5627,
      "step": 1050
    },
    {
      "epoch": 1.519275596198673,
      "grad_norm": 0.510686482301255,
      "learning_rate": 5.772288872759702e-05,
      "loss": 0.5625,
      "step": 1060
    },
    {
      "epoch": 1.5336202259279181,
      "grad_norm": 0.5278526048784437,
      "learning_rate": 5.6898102783539665e-05,
      "loss": 0.5606,
      "step": 1070
    },
    {
      "epoch": 1.5479648556571632,
      "grad_norm": 0.5559268236737518,
      "learning_rate": 5.607139879998427e-05,
      "loss": 0.5595,
      "step": 1080
    },
    {
      "epoch": 1.5623094853864083,
      "grad_norm": 0.5583305948616533,
      "learning_rate": 5.524300664461235e-05,
      "loss": 0.5575,
      "step": 1090
    },
    {
      "epoch": 1.5766541151156535,
      "grad_norm": 0.5186185461448056,
      "learning_rate": 5.441315665450697e-05,
      "loss": 0.5589,
      "step": 1100
    },
    {
      "epoch": 1.5909987448448986,
      "grad_norm": 0.5168193702732372,
      "learning_rate": 5.3582079572106794e-05,
      "loss": 0.5534,
      "step": 1110
    },
    {
      "epoch": 1.6053433745741437,
      "grad_norm": 0.5037158486433893,
      "learning_rate": 5.275000648104743e-05,
      "loss": 0.5569,
      "step": 1120
    },
    {
      "epoch": 1.6196880043033888,
      "grad_norm": 0.49884254214328216,
      "learning_rate": 5.191716874190785e-05,
      "loss": 0.5564,
      "step": 1130
    },
    {
      "epoch": 1.6340326340326339,
      "grad_norm": 0.5071669705468655,
      "learning_rate": 5.1083797927879896e-05,
      "loss": 0.5552,
      "step": 1140
    },
    {
      "epoch": 1.648377263761879,
      "grad_norm": 0.5410568223255828,
      "learning_rate": 5.025012576037855e-05,
      "loss": 0.564,
      "step": 1150
    },
    {
      "epoch": 1.6627218934911243,
      "grad_norm": 0.4850296619715352,
      "learning_rate": 4.9416384044611124e-05,
      "loss": 0.5579,
      "step": 1160
    },
    {
      "epoch": 1.6770665232203694,
      "grad_norm": 0.5039355867374112,
      "learning_rate": 4.858280460512302e-05,
      "loss": 0.5551,
      "step": 1170
    },
    {
      "epoch": 1.6914111529496145,
      "grad_norm": 0.5354526153870778,
      "learning_rate": 4.7749619221338227e-05,
      "loss": 0.5553,
      "step": 1180
    },
    {
      "epoch": 1.7057557826788596,
      "grad_norm": 0.5149848285267117,
      "learning_rate": 4.691705956311225e-05,
      "loss": 0.554,
      "step": 1190
    },
    {
      "epoch": 1.7201004124081047,
      "grad_norm": 0.5076867052953596,
      "learning_rate": 4.608535712631566e-05,
      "loss": 0.553,
      "step": 1200
    },
    {
      "epoch": 1.7344450421373498,
      "grad_norm": 0.5245034600250137,
      "learning_rate": 4.525474316846581e-05,
      "loss": 0.5455,
      "step": 1210
    },
    {
      "epoch": 1.748789671866595,
      "grad_norm": 0.5044035057693634,
      "learning_rate": 4.4425448644425066e-05,
      "loss": 0.5487,
      "step": 1220
    },
    {
      "epoch": 1.76313430159584,
      "grad_norm": 0.5182301310051464,
      "learning_rate": 4.359770414218296e-05,
      "loss": 0.5401,
      "step": 1230
    },
    {
      "epoch": 1.7774789313250852,
      "grad_norm": 0.5066048824304953,
      "learning_rate": 4.2771739818740565e-05,
      "loss": 0.5496,
      "step": 1240
    },
    {
      "epoch": 1.7918235610543303,
      "grad_norm": 0.5337687629976401,
      "learning_rate": 4.194778533611451e-05,
      "loss": 0.5413,
      "step": 1250
    },
    {
      "epoch": 1.8061681907835754,
      "grad_norm": 0.5207289329535878,
      "learning_rate": 4.112606979747881e-05,
      "loss": 0.544,
      "step": 1260
    },
    {
      "epoch": 1.8205128205128205,
      "grad_norm": 0.5450234045888507,
      "learning_rate": 4.030682168346192e-05,
      "loss": 0.5358,
      "step": 1270
    },
    {
      "epoch": 1.8348574502420656,
      "grad_norm": 0.48504143892984203,
      "learning_rate": 3.949026878861704e-05,
      "loss": 0.539,
      "step": 1280
    },
    {
      "epoch": 1.8492020799713107,
      "grad_norm": 0.49111142317889844,
      "learning_rate": 3.867663815808303e-05,
      "loss": 0.5389,
      "step": 1290
    },
    {
      "epoch": 1.8635467097005558,
      "grad_norm": 0.5008380643048362,
      "learning_rate": 3.78661560244539e-05,
      "loss": 0.5363,
      "step": 1300
    },
    {
      "epoch": 1.8778913394298011,
      "grad_norm": 0.4952709952315973,
      "learning_rate": 3.705904774487396e-05,
      "loss": 0.5319,
      "step": 1310
    },
    {
      "epoch": 1.8922359691590462,
      "grad_norm": 0.5164716516662802,
      "learning_rate": 3.6255537738376706e-05,
      "loss": 0.5402,
      "step": 1320
    },
    {
      "epoch": 1.9065805988882913,
      "grad_norm": 0.5240016137645628,
      "learning_rate": 3.545584942348426e-05,
      "loss": 0.5309,
      "step": 1330
    },
    {
      "epoch": 1.9209252286175365,
      "grad_norm": 0.5058856022275582,
      "learning_rate": 3.466020515608525e-05,
      "loss": 0.5298,
      "step": 1340
    },
    {
      "epoch": 1.9352698583467816,
      "grad_norm": 0.5019184399441088,
      "learning_rate": 3.386882616760794e-05,
      "loss": 0.5319,
      "step": 1350
    },
    {
      "epoch": 1.9496144880760267,
      "grad_norm": 0.471920743317965,
      "learning_rate": 3.30819325035062e-05,
      "loss": 0.5287,
      "step": 1360
    },
    {
      "epoch": 1.9639591178052718,
      "grad_norm": 0.5094525188470136,
      "learning_rate": 3.229974296207513e-05,
      "loss": 0.5385,
      "step": 1370
    },
    {
      "epoch": 1.9783037475345169,
      "grad_norm": 0.5606826371060244,
      "learning_rate": 3.152247503361353e-05,
      "loss": 0.527,
      "step": 1380
    },
    {
      "epoch": 1.992648377263762,
      "grad_norm": 0.5085670056810321,
      "learning_rate": 3.075034483994997e-05,
      "loss": 0.5257,
      "step": 1390
    },
    {
      "epoch": 2.005737851891698,
      "grad_norm": 0.5719992437170487,
      "learning_rate": 2.998356707434947e-05,
      "loss": 0.4623,
      "step": 1400
    },
    {
      "epoch": 2.020082481620943,
      "grad_norm": 0.5387964274909228,
      "learning_rate": 2.9222354941817375e-05,
      "loss": 0.3606,
      "step": 1410
    },
    {
      "epoch": 2.034427111350188,
      "grad_norm": 0.5698349863645911,
      "learning_rate": 2.846692009981693e-05,
      "loss": 0.3527,
      "step": 1420
    },
    {
      "epoch": 2.0487717410794333,
      "grad_norm": 0.5324199413342471,
      "learning_rate": 2.771747259941734e-05,
      "loss": 0.3462,
      "step": 1430
    },
    {
      "epoch": 2.0631163708086784,
      "grad_norm": 0.53215786873291,
      "learning_rate": 2.6974220826888374e-05,
      "loss": 0.342,
      "step": 1440
    },
    {
      "epoch": 2.0774610005379235,
      "grad_norm": 0.5110033203467198,
      "learning_rate": 2.623737144575787e-05,
      "loss": 0.3462,
      "step": 1450
    },
    {
      "epoch": 2.0918056302671686,
      "grad_norm": 0.5246624681955846,
      "learning_rate": 2.5507129339348335e-05,
      "loss": 0.3487,
      "step": 1460
    },
    {
      "epoch": 2.1061502599964137,
      "grad_norm": 0.5070575905086573,
      "learning_rate": 2.478369755380839e-05,
      "loss": 0.3465,
      "step": 1470
    },
    {
      "epoch": 2.120494889725659,
      "grad_norm": 0.5132555087524301,
      "learning_rate": 2.406727724165524e-05,
      "loss": 0.3451,
      "step": 1480
    },
    {
      "epoch": 2.134839519454904,
      "grad_norm": 0.5143913782142077,
      "learning_rate": 2.3358067605843537e-05,
      "loss": 0.3442,
      "step": 1490
    },
    {
      "epoch": 2.149184149184149,
      "grad_norm": 0.4894675520933593,
      "learning_rate": 2.2656265844376367e-05,
      "loss": 0.3449,
      "step": 1500
    },
    {
      "epoch": 2.163528778913394,
      "grad_norm": 0.5118938379622366,
      "learning_rate": 2.1962067095473648e-05,
      "loss": 0.3443,
      "step": 1510
    },
    {
      "epoch": 2.1778734086426392,
      "grad_norm": 0.5264199925745078,
      "learning_rate": 2.127566438331345e-05,
      "loss": 0.3443,
      "step": 1520
    },
    {
      "epoch": 2.1922180383718843,
      "grad_norm": 0.5517034745554297,
      "learning_rate": 2.059724856436092e-05,
      "loss": 0.34,
      "step": 1530
    },
    {
      "epoch": 2.2065626681011294,
      "grad_norm": 0.5153912300862288,
      "learning_rate": 1.992700827430007e-05,
      "loss": 0.3447,
      "step": 1540
    },
    {
      "epoch": 2.2209072978303745,
      "grad_norm": 0.49999210085621454,
      "learning_rate": 1.9265129875582954e-05,
      "loss": 0.3402,
      "step": 1550
    },
    {
      "epoch": 2.23525192755962,
      "grad_norm": 0.5117332103404209,
      "learning_rate": 1.8611797405611097e-05,
      "loss": 0.3434,
      "step": 1560
    },
    {
      "epoch": 2.249596557288865,
      "grad_norm": 0.5334345620530477,
      "learning_rate": 1.7967192525563254e-05,
      "loss": 0.3429,
      "step": 1570
    },
    {
      "epoch": 2.2639411870181103,
      "grad_norm": 0.5052797787039551,
      "learning_rate": 1.733149446988394e-05,
      "loss": 0.3402,
      "step": 1580
    },
    {
      "epoch": 2.2782858167473554,
      "grad_norm": 0.5115033156660913,
      "learning_rate": 1.670487999644669e-05,
      "loss": 0.3413,
      "step": 1590
    },
    {
      "epoch": 2.2926304464766005,
      "grad_norm": 0.4961174768714891,
      "learning_rate": 1.6087523337406024e-05,
      "loss": 0.3407,
      "step": 1600
    },
    {
      "epoch": 2.3069750762058456,
      "grad_norm": 0.5181202897712733,
      "learning_rate": 1.547959615075164e-05,
      "loss": 0.3343,
      "step": 1610
    },
    {
      "epoch": 2.3213197059350907,
      "grad_norm": 0.5076475142323975,
      "learning_rate": 1.4881267472578325e-05,
      "loss": 0.3415,
      "step": 1620
    },
    {
      "epoch": 2.335664335664336,
      "grad_norm": 0.4921735423850851,
      "learning_rate": 1.4292703670084916e-05,
      "loss": 0.3392,
      "step": 1630
    },
    {
      "epoch": 2.350008965393581,
      "grad_norm": 0.5151996030919073,
      "learning_rate": 1.3714068395315427e-05,
      "loss": 0.3349,
      "step": 1640
    },
    {
      "epoch": 2.364353595122826,
      "grad_norm": 0.5104074354031792,
      "learning_rate": 1.3145522539655041e-05,
      "loss": 0.3354,
      "step": 1650
    },
    {
      "epoch": 2.378698224852071,
      "grad_norm": 0.5380960601841327,
      "learning_rate": 1.2587224189093755e-05,
      "loss": 0.3302,
      "step": 1660
    },
    {
      "epoch": 2.3930428545813163,
      "grad_norm": 0.5049847193289703,
      "learning_rate": 1.2039328580270065e-05,
      "loss": 0.3333,
      "step": 1670
    },
    {
      "epoch": 2.4073874843105614,
      "grad_norm": 0.5040716684214055,
      "learning_rate": 1.150198805730689e-05,
      "loss": 0.3329,
      "step": 1680
    },
    {
      "epoch": 2.4217321140398065,
      "grad_norm": 0.5040857331959766,
      "learning_rate": 1.0975352029451863e-05,
      "loss": 0.3337,
      "step": 1690
    },
    {
      "epoch": 2.4360767437690516,
      "grad_norm": 0.5091871791617659,
      "learning_rate": 1.0459566929533588e-05,
      "loss": 0.3345,
      "step": 1700
    },
    {
      "epoch": 2.4504213734982967,
      "grad_norm": 0.4986671760440932,
      "learning_rate": 9.954776173245511e-06,
      "loss": 0.3369,
      "step": 1710
    },
    {
      "epoch": 2.464766003227542,
      "grad_norm": 0.5146311314437344,
      "learning_rate": 9.461120119268713e-06,
      "loss": 0.3338,
      "step": 1720
    },
    {
      "epoch": 2.479110632956787,
      "grad_norm": 0.5052152754810867,
      "learning_rate": 8.978736030244783e-06,
      "loss": 0.331,
      "step": 1730
    },
    {
      "epoch": 2.493455262686032,
      "grad_norm": 0.5060074098914322,
      "learning_rate": 8.50775803460948e-06,
      "loss": 0.3336,
      "step": 1740
    },
    {
      "epoch": 2.507799892415277,
      "grad_norm": 0.5047637166351485,
      "learning_rate": 8.048317089297875e-06,
      "loss": 0.331,
      "step": 1750
    },
    {
      "epoch": 2.5221445221445222,
      "grad_norm": 0.5051085483525463,
      "learning_rate": 7.600540943331347e-06,
      "loss": 0.3316,
      "step": 1760
    },
    {
      "epoch": 2.5364891518737673,
      "grad_norm": 0.5291232815833027,
      "learning_rate": 7.164554102296617e-06,
      "loss": 0.3322,
      "step": 1770
    },
    {
      "epoch": 2.5508337816030124,
      "grad_norm": 0.4950112167812311,
      "learning_rate": 6.740477793726529e-06,
      "loss": 0.3302,
      "step": 1780
    },
    {
      "epoch": 2.5651784113322575,
      "grad_norm": 0.5171790095780296,
      "learning_rate": 6.32842993339236e-06,
      "loss": 0.3278,
      "step": 1790
    },
    {
      "epoch": 2.5795230410615027,
      "grad_norm": 0.5041803659831317,
      "learning_rate": 5.928525092516934e-06,
      "loss": 0.3334,
      "step": 1800
    },
    {
      "epoch": 2.5938676707907478,
      "grad_norm": 0.5073754782012652,
      "learning_rate": 5.540874465917778e-06,
      "loss": 0.3325,
      "step": 1810
    },
    {
      "epoch": 2.608212300519993,
      "grad_norm": 0.48715355421051987,
      "learning_rate": 5.165585841089021e-06,
      "loss": 0.3246,
      "step": 1820
    },
    {
      "epoch": 2.622556930249238,
      "grad_norm": 0.5162975920861064,
      "learning_rate": 4.8027635682307445e-06,
      "loss": 0.325,
      "step": 1830
    },
    {
      "epoch": 2.636901559978483,
      "grad_norm": 0.5030085273093482,
      "learning_rate": 4.45250853123404e-06,
      "loss": 0.3267,
      "step": 1840
    },
    {
      "epoch": 2.651246189707728,
      "grad_norm": 0.5261872357667398,
      "learning_rate": 4.1149181196299905e-06,
      "loss": 0.3277,
      "step": 1850
    },
    {
      "epoch": 2.6655908194369733,
      "grad_norm": 0.49678770417028423,
      "learning_rate": 3.7900862015101457e-06,
      "loss": 0.3285,
      "step": 1860
    },
    {
      "epoch": 2.6799354491662184,
      "grad_norm": 0.4889164865698146,
      "learning_rate": 3.4781030974262108e-06,
      "loss": 0.3268,
      "step": 1870
    },
    {
      "epoch": 2.6942800788954635,
      "grad_norm": 0.4884408260663131,
      "learning_rate": 3.1790555552761615e-06,
      "loss": 0.3246,
      "step": 1880
    },
    {
      "epoch": 2.7086247086247086,
      "grad_norm": 0.49292610830996564,
      "learning_rate": 2.8930267261836395e-06,
      "loss": 0.3231,
      "step": 1890
    },
    {
      "epoch": 2.7229693383539537,
      "grad_norm": 0.4890311152459734,
      "learning_rate": 2.6200961413776094e-06,
      "loss": 0.324,
      "step": 1900
    },
    {
      "epoch": 2.737313968083199,
      "grad_norm": 0.5165429385071639,
      "learning_rate": 2.3603396900783724e-06,
      "loss": 0.3251,
      "step": 1910
    },
    {
      "epoch": 2.751658597812444,
      "grad_norm": 0.5024981727502739,
      "learning_rate": 2.113829598396383e-06,
      "loss": 0.322,
      "step": 1920
    },
    {
      "epoch": 2.766003227541689,
      "grad_norm": 0.49038298209467507,
      "learning_rate": 1.8806344092494932e-06,
      "loss": 0.3248,
      "step": 1930
    },
    {
      "epoch": 2.780347857270934,
      "grad_norm": 0.4917027026533285,
      "learning_rate": 1.6608189633044113e-06,
      "loss": 0.325,
      "step": 1940
    },
    {
      "epoch": 2.7946924870001792,
      "grad_norm": 0.4869252686371276,
      "learning_rate": 1.4544443809475561e-06,
      "loss": 0.3247,
      "step": 1950
    },
    {
      "epoch": 2.8090371167294244,
      "grad_norm": 0.48253239623178734,
      "learning_rate": 1.2615680452903067e-06,
      "loss": 0.3234,
      "step": 1960
    },
    {
      "epoch": 2.8233817464586695,
      "grad_norm": 0.5096600620224961,
      "learning_rate": 1.082243586213455e-06,
      "loss": 0.3219,
      "step": 1970
    },
    {
      "epoch": 2.8377263761879146,
      "grad_norm": 0.48655701646749633,
      "learning_rate": 9.165208654552671e-07,
      "loss": 0.321,
      "step": 1980
    },
    {
      "epoch": 2.8520710059171597,
      "grad_norm": 0.47440137878879957,
      "learning_rate": 7.6444596274724e-07,
      "loss": 0.3229,
      "step": 1990
    },
    {
      "epoch": 2.866415635646405,
      "grad_norm": 0.5034996226878241,
      "learning_rate": 6.260611630015067e-07,
      "loss": 0.3221,
      "step": 2000
    },
    {
      "epoch": 2.88076026537565,
      "grad_norm": 0.47683170425958754,
      "learning_rate": 5.01404944553363e-07,
      "loss": 0.3219,
      "step": 2010
    },
    {
      "epoch": 2.895104895104895,
      "grad_norm": 0.4909218835510976,
      "learning_rate": 3.9051196846225935e-07,
      "loss": 0.3251,
      "step": 2020
    },
    {
      "epoch": 2.90944952483414,
      "grad_norm": 0.48456217220790176,
      "learning_rate": 2.9341306887417653e-07,
      "loss": 0.3213,
      "step": 2030
    },
    {
      "epoch": 2.923794154563385,
      "grad_norm": 0.4779578566731188,
      "learning_rate": 2.101352444480842e-07,
      "loss": 0.3241,
      "step": 2040
    },
    {
      "epoch": 2.9381387842926303,
      "grad_norm": 0.49966345259743034,
      "learning_rate": 1.4070165084889008e-07,
      "loss": 0.3199,
      "step": 2050
    },
    {
      "epoch": 2.9524834140218754,
      "grad_norm": 0.4754785997240394,
      "learning_rate": 8.513159430892925e-08,
      "loss": 0.3225,
      "step": 2060
    },
    {
      "epoch": 2.9668280437511205,
      "grad_norm": 0.4903628137644153,
      "learning_rate": 4.344052625981365e-08,
      "loss": 0.3219,
      "step": 2070
    },
    {
      "epoch": 2.9811726734803656,
      "grad_norm": 0.4992880216120249,
      "learning_rate": 1.5640039036085575e-08,
      "loss": 0.3187,
      "step": 2080
    },
    {
      "epoch": 2.9955173032096107,
      "grad_norm": 0.4963785468111048,
      "learning_rate": 1.7378626519626296e-09,
      "loss": 0.3238,
      "step": 2090
    },
    {
      "epoch": 3.0,
      "step": 2094,
      "total_flos": 306663055425536.0,
      "train_loss": 0.5762376924754328,
      "train_runtime": 90627.2697,
      "train_samples_per_second": 1.477,
      "train_steps_per_second": 0.023
    }
  ],
  "logging_steps": 10,
  "max_steps": 2094,
  "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": true
      },
      "attributes": {}
    }
  },
  "total_flos": 306663055425536.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}