{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 4.705882352941177,
  "eval_steps": 500,
  "global_step": 1200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0392156862745098,
      "grad_norm": 1.956024169921875,
      "learning_rate": 2.8125000000000003e-05,
      "loss": 2.2858,
      "step": 10
    },
    {
      "epoch": 0.0784313725490196,
      "grad_norm": 0.5354028940200806,
      "learning_rate": 5.9375e-05,
      "loss": 0.8601,
      "step": 20
    },
    {
      "epoch": 0.11764705882352941,
      "grad_norm": 0.41287925839424133,
      "learning_rate": 9.062500000000001e-05,
      "loss": 0.5216,
      "step": 30
    },
    {
      "epoch": 0.1568627450980392,
      "grad_norm": 0.40572163462638855,
      "learning_rate": 0.00012187500000000001,
      "loss": 0.4683,
      "step": 40
    },
    {
      "epoch": 0.19607843137254902,
      "grad_norm": 0.378238707780838,
      "learning_rate": 0.000153125,
      "loss": 0.4577,
      "step": 50
    },
    {
      "epoch": 0.23529411764705882,
      "grad_norm": 0.3841954469680786,
      "learning_rate": 0.000184375,
      "loss": 0.455,
      "step": 60
    },
    {
      "epoch": 0.27450980392156865,
      "grad_norm": 0.3890840411186218,
      "learning_rate": 0.00019999158768778033,
      "loss": 0.4401,
      "step": 70
    },
    {
      "epoch": 0.3137254901960784,
      "grad_norm": 0.34165677428245544,
      "learning_rate": 0.00019992429768182467,
      "loss": 0.4646,
      "step": 80
    },
    {
      "epoch": 0.35294117647058826,
      "grad_norm": 0.38149768114089966,
      "learning_rate": 0.00019978976295317172,
      "loss": 0.4144,
      "step": 90
    },
    {
      "epoch": 0.39215686274509803,
      "grad_norm": 0.36713501811027527,
      "learning_rate": 0.00019958807403786452,
      "loss": 0.4333,
      "step": 100
    },
    {
      "epoch": 0.43137254901960786,
      "grad_norm": 0.30455392599105835,
      "learning_rate": 0.00019931936666380416,
      "loss": 0.3978,
      "step": 110
    },
    {
      "epoch": 0.47058823529411764,
      "grad_norm": 0.31675583124160767,
      "learning_rate": 0.0001989838216594104,
      "loss": 0.4048,
      "step": 120
    },
    {
      "epoch": 0.5098039215686274,
      "grad_norm": 0.34656670689582825,
      "learning_rate": 0.0001985816648319323,
      "loss": 0.4188,
      "step": 130
    },
    {
      "epoch": 0.5490196078431373,
      "grad_norm": 0.3448728919029236,
      "learning_rate": 0.00019811316681548968,
      "loss": 0.4212,
      "step": 140
    },
    {
      "epoch": 0.5882352941176471,
      "grad_norm": 0.31498879194259644,
      "learning_rate": 0.0001975786428889479,
      "loss": 0.3918,
      "step": 150
    },
    {
      "epoch": 0.6274509803921569,
      "grad_norm": 0.364424467086792,
      "learning_rate": 0.00019697845276374902,
      "loss": 0.397,
      "step": 160
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 0.32824113965034485,
      "learning_rate": 0.00019631300034184153,
      "loss": 0.3923,
      "step": 170
    },
    {
      "epoch": 0.7058823529411765,
      "grad_norm": 0.34096255898475647,
      "learning_rate": 0.00019558273344387182,
      "loss": 0.3774,
      "step": 180
    },
    {
      "epoch": 0.7450980392156863,
      "grad_norm": 0.3589867949485779,
      "learning_rate": 0.00019478814350782037,
      "loss": 0.3919,
      "step": 190
    },
    {
      "epoch": 0.7843137254901961,
      "grad_norm": 0.3224334120750427,
      "learning_rate": 0.0001939297652582851,
      "loss": 0.3969,
      "step": 200
    },
    {
      "epoch": 0.8235294117647058,
      "grad_norm": 0.31771010160446167,
      "learning_rate": 0.0001930081763466351,
      "loss": 0.3782,
      "step": 210
    },
    {
      "epoch": 0.8627450980392157,
      "grad_norm": 0.3161714971065521,
      "learning_rate": 0.00019202399696227626,
      "loss": 0.3691,
      "step": 220
    },
    {
      "epoch": 0.9019607843137255,
      "grad_norm": 0.3109596073627472,
      "learning_rate": 0.0001909778894152904,
      "loss": 0.3609,
      "step": 230
    },
    {
      "epoch": 0.9411764705882353,
      "grad_norm": 0.30899715423583984,
      "learning_rate": 0.00018987055769072972,
      "loss": 0.3829,
      "step": 240
    },
    {
      "epoch": 0.9803921568627451,
      "grad_norm": 0.39666712284088135,
      "learning_rate": 0.00018870274697486492,
      "loss": 0.3679,
      "step": 250
    },
    {
      "epoch": 1.0196078431372548,
      "grad_norm": 0.303822785615921,
      "learning_rate": 0.00018747524315370753,
      "loss": 0.3395,
      "step": 260
    },
    {
      "epoch": 1.0588235294117647,
      "grad_norm": 0.31603726744651794,
      "learning_rate": 0.00018618887228414232,
      "loss": 0.2691,
      "step": 270
    },
    {
      "epoch": 1.0980392156862746,
      "grad_norm": 0.3404744863510132,
      "learning_rate": 0.00018484450003802713,
      "loss": 0.2722,
      "step": 280
    },
    {
      "epoch": 1.1372549019607843,
      "grad_norm": 0.3062707781791687,
      "learning_rate": 0.00018344303111963292,
      "loss": 0.2741,
      "step": 290
    },
    {
      "epoch": 1.1764705882352942,
      "grad_norm": 0.2567201554775238,
      "learning_rate": 0.00018198540865681725,
      "loss": 0.2685,
      "step": 300
    },
    {
      "epoch": 1.215686274509804,
      "grad_norm": 0.33607518672943115,
      "learning_rate": 0.0001804726135663399,
      "loss": 0.2906,
      "step": 310
    },
    {
      "epoch": 1.2549019607843137,
      "grad_norm": 0.33044159412384033,
      "learning_rate": 0.00017890566389374855,
      "loss": 0.2655,
      "step": 320
    },
    {
      "epoch": 1.2941176470588236,
      "grad_norm": 0.3430589437484741,
      "learning_rate": 0.0001772856141282781,
      "loss": 0.2727,
      "step": 330
    },
    {
      "epoch": 1.3333333333333333,
      "grad_norm": 0.3410845994949341,
      "learning_rate": 0.00017561355449322526,
      "loss": 0.2912,
      "step": 340
    },
    {
      "epoch": 1.3725490196078431,
      "grad_norm": 0.3181304335594177,
      "learning_rate": 0.00017389061021227535,
      "loss": 0.2785,
      "step": 350
    },
    {
      "epoch": 1.4117647058823528,
      "grad_norm": 0.3184092938899994,
      "learning_rate": 0.0001721179407522756,
      "loss": 0.2753,
      "step": 360
    },
    {
      "epoch": 1.4509803921568627,
      "grad_norm": 0.3151929974555969,
      "learning_rate": 0.00017029673904296415,
      "loss": 0.2772,
      "step": 370
    },
    {
      "epoch": 1.4901960784313726,
      "grad_norm": 0.3399432599544525,
      "learning_rate": 0.00016842823067418018,
      "loss": 0.2661,
      "step": 380
    },
    {
      "epoch": 1.5294117647058822,
      "grad_norm": 0.2826154828071594,
      "learning_rate": 0.0001665136730710948,
      "loss": 0.2697,
      "step": 390
    },
    {
      "epoch": 1.5686274509803921,
      "grad_norm": 0.33506590127944946,
      "learning_rate": 0.00016455435464801866,
      "loss": 0.2664,
      "step": 400
    },
    {
      "epoch": 1.607843137254902,
      "grad_norm": 0.2924242317676544,
      "learning_rate": 0.00016255159394135463,
      "loss": 0.2804,
      "step": 410
    },
    {
      "epoch": 1.6470588235294117,
      "grad_norm": 0.30420586466789246,
      "learning_rate": 0.0001605067387222803,
      "loss": 0.269,
      "step": 420
    },
    {
      "epoch": 1.6862745098039216,
      "grad_norm": 0.32626214623451233,
      "learning_rate": 0.00015842116508975608,
      "loss": 0.2867,
      "step": 430
    },
    {
      "epoch": 1.7254901960784315,
      "grad_norm": 0.3080560266971588,
      "learning_rate": 0.0001562962765444704,
      "loss": 0.2701,
      "step": 440
    },
    {
      "epoch": 1.7647058823529411,
      "grad_norm": 0.3522433042526245,
      "learning_rate": 0.0001541335030443444,
      "loss": 0.2689,
      "step": 450
    },
    {
      "epoch": 1.803921568627451,
      "grad_norm": 0.321441113948822,
      "learning_rate": 0.0001519343000422324,
      "loss": 0.2633,
      "step": 460
    },
    {
      "epoch": 1.843137254901961,
      "grad_norm": 0.31674930453300476,
      "learning_rate": 0.0001497001475064651,
      "loss": 0.2741,
      "step": 470
    },
    {
      "epoch": 1.8823529411764706,
      "grad_norm": 0.33134984970092773,
      "learning_rate": 0.0001474325489248949,
      "loss": 0.2587,
      "step": 480
    },
    {
      "epoch": 1.9215686274509802,
      "grad_norm": 0.3318138122558594,
      "learning_rate": 0.00014513303029311386,
      "loss": 0.2567,
      "step": 490
    },
    {
      "epoch": 1.9607843137254903,
      "grad_norm": 0.30877944827079773,
      "learning_rate": 0.0001428031390875249,
      "loss": 0.2639,
      "step": 500
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.3411860764026642,
      "learning_rate": 0.00014044444322395703,
      "loss": 0.2675,
      "step": 510
    },
    {
      "epoch": 2.0392156862745097,
      "grad_norm": 0.3239802420139313,
      "learning_rate": 0.00013805853000252585,
      "loss": 0.1614,
      "step": 520
    },
    {
      "epoch": 2.0784313725490198,
      "grad_norm": 0.3532658815383911,
      "learning_rate": 0.00013564700503944962,
      "loss": 0.1659,
      "step": 530
    },
    {
      "epoch": 2.1176470588235294,
      "grad_norm": 0.36013808846473694,
      "learning_rate": 0.00013321149118653903,
      "loss": 0.16,
      "step": 540
    },
    {
      "epoch": 2.156862745098039,
      "grad_norm": 0.27771538496017456,
      "learning_rate": 0.00013075362743908815,
      "loss": 0.1592,
      "step": 550
    },
    {
      "epoch": 2.196078431372549,
      "grad_norm": 0.26791512966156006,
      "learning_rate": 0.00012827506783290166,
      "loss": 0.1542,
      "step": 560
    },
    {
      "epoch": 2.235294117647059,
      "grad_norm": 0.30019891262054443,
      "learning_rate": 0.0001257774803312005,
      "loss": 0.1627,
      "step": 570
    },
    {
      "epoch": 2.2745098039215685,
      "grad_norm": 0.30322951078414917,
      "learning_rate": 0.00012326254570215467,
      "loss": 0.1555,
      "step": 580
    },
    {
      "epoch": 2.313725490196078,
      "grad_norm": 0.3170073628425598,
      "learning_rate": 0.00012073195638779943,
      "loss": 0.1645,
      "step": 590
    },
    {
      "epoch": 2.3529411764705883,
      "grad_norm": 0.27831751108169556,
      "learning_rate": 0.00011818741536509486,
      "loss": 0.1632,
      "step": 600
    },
    {
      "epoch": 2.392156862745098,
      "grad_norm": 0.2873671054840088,
      "learning_rate": 0.00011563063499989624,
      "loss": 0.1646,
      "step": 610
    },
    {
      "epoch": 2.431372549019608,
      "grad_norm": 0.2752895951271057,
      "learning_rate": 0.00011306333589460603,
      "loss": 0.1716,
      "step": 620
    },
    {
      "epoch": 2.4705882352941178,
      "grad_norm": 0.2440868765115738,
      "learning_rate": 0.00011048724573028295,
      "loss": 0.1575,
      "step": 630
    },
    {
      "epoch": 2.5098039215686274,
      "grad_norm": 0.3065614700317383,
      "learning_rate": 0.00010790409810398768,
      "loss": 0.1606,
      "step": 640
    },
    {
      "epoch": 2.549019607843137,
      "grad_norm": 0.24554114043712616,
      "learning_rate": 0.00010531563136214684,
      "loss": 0.1644,
      "step": 650
    },
    {
      "epoch": 2.588235294117647,
      "grad_norm": 0.3355470895767212,
      "learning_rate": 0.00010272358743072152,
      "loss": 0.1583,
      "step": 660
    },
    {
      "epoch": 2.627450980392157,
      "grad_norm": 0.26972317695617676,
      "learning_rate": 0.00010012971064296626,
      "loss": 0.1579,
      "step": 670
    },
    {
      "epoch": 2.6666666666666665,
      "grad_norm": 0.3012099266052246,
      "learning_rate": 9.753574656556854e-05,
      "loss": 0.1559,
      "step": 680
    },
    {
      "epoch": 2.7058823529411766,
      "grad_norm": 0.2840341627597809,
      "learning_rate": 9.494344082395799e-05,
      "loss": 0.1599,
      "step": 690
    },
    {
      "epoch": 2.7450980392156863,
      "grad_norm": 0.3037700355052948,
      "learning_rate": 9.235453792757609e-05,
      "loss": 0.1581,
      "step": 700
    },
    {
      "epoch": 2.784313725490196,
      "grad_norm": 0.2722160220146179,
      "learning_rate": 8.9770780095897e-05,
      "loss": 0.1623,
      "step": 710
    },
    {
      "epoch": 2.8235294117647056,
      "grad_norm": 0.26960518956184387,
      "learning_rate": 8.719390608598917e-05,
      "loss": 0.161,
      "step": 720
    },
    {
      "epoch": 2.8627450980392157,
      "grad_norm": 0.30165761709213257,
      "learning_rate": 8.462565002240732e-05,
      "loss": 0.1661,
      "step": 730
    },
    {
      "epoch": 2.9019607843137254,
      "grad_norm": 0.2838638126850128,
      "learning_rate": 8.206774023020168e-05,
      "loss": 0.1518,
      "step": 740
    },
    {
      "epoch": 2.9411764705882355,
      "grad_norm": 0.33738985657691956,
      "learning_rate": 7.952189807183017e-05,
      "loss": 0.1574,
      "step": 750
    },
    {
      "epoch": 2.980392156862745,
      "grad_norm": 0.29775676131248474,
      "learning_rate": 7.698983678875601e-05,
      "loss": 0.1542,
      "step": 760
    },
    {
      "epoch": 3.019607843137255,
      "grad_norm": 0.25064918398857117,
      "learning_rate": 7.447326034851041e-05,
      "loss": 0.131,
      "step": 770
    },
    {
      "epoch": 3.0588235294117645,
      "grad_norm": 0.26662278175354004,
      "learning_rate": 7.19738622979964e-05,
      "loss": 0.0906,
      "step": 780
    },
    {
      "epoch": 3.0980392156862746,
      "grad_norm": 0.18513642251491547,
      "learning_rate": 6.949332462380494e-05,
      "loss": 0.0865,
      "step": 790
    },
    {
      "epoch": 3.1372549019607843,
      "grad_norm": 0.20468716323375702,
      "learning_rate": 6.703331662031098e-05,
      "loss": 0.0901,
      "step": 800
    },
    {
      "epoch": 3.176470588235294,
      "grad_norm": 0.2853894531726837,
      "learning_rate": 6.459549376631065e-05,
      "loss": 0.0924,
      "step": 810
    },
    {
      "epoch": 3.215686274509804,
      "grad_norm": 0.21128232777118683,
      "learning_rate": 6.218149661095573e-05,
      "loss": 0.0884,
      "step": 820
    },
    {
      "epoch": 3.2549019607843137,
      "grad_norm": 0.2623343765735626,
      "learning_rate": 5.9792949669735096e-05,
      "loss": 0.0923,
      "step": 830
    },
    {
      "epoch": 3.2941176470588234,
      "grad_norm": 0.19632554054260254,
      "learning_rate": 5.743146033124633e-05,
      "loss": 0.0857,
      "step": 840
    },
    {
      "epoch": 3.3333333333333335,
      "grad_norm": 0.24294443428516388,
      "learning_rate": 5.50986177754927e-05,
      "loss": 0.0895,
      "step": 850
    },
    {
      "epoch": 3.372549019607843,
      "grad_norm": 0.25959959626197815,
      "learning_rate": 5.279599190443401e-05,
      "loss": 0.0922,
      "step": 860
    },
    {
      "epoch": 3.411764705882353,
      "grad_norm": 0.20078334212303162,
      "learning_rate": 5.052513228551048e-05,
      "loss": 0.0931,
      "step": 870
    },
    {
      "epoch": 3.450980392156863,
      "grad_norm": 0.2411254346370697,
      "learning_rate": 4.828756710885098e-05,
      "loss": 0.0914,
      "step": 880
    },
    {
      "epoch": 3.4901960784313726,
      "grad_norm": 0.24628736078739166,
      "learning_rate": 4.60848021588675e-05,
      "loss": 0.0902,
      "step": 890
    },
    {
      "epoch": 3.5294117647058822,
      "grad_norm": 0.21074128150939941,
      "learning_rate": 4.3918319800927285e-05,
      "loss": 0.0927,
      "step": 900
    },
    {
      "epoch": 3.568627450980392,
      "grad_norm": 0.22155368328094482,
      "learning_rate": 4.1789577983785164e-05,
      "loss": 0.092,
      "step": 910
    },
    {
      "epoch": 3.607843137254902,
      "grad_norm": 0.21323789656162262,
      "learning_rate": 3.970000925844737e-05,
      "loss": 0.0904,
      "step": 920
    },
    {
      "epoch": 3.6470588235294117,
      "grad_norm": 0.2426736205816269,
      "learning_rate": 3.7651019814126654e-05,
      "loss": 0.0922,
      "step": 930
    },
    {
      "epoch": 3.686274509803922,
      "grad_norm": 0.22919663786888123,
      "learning_rate": 3.5643988531937924e-05,
      "loss": 0.0912,
      "step": 940
    },
    {
      "epoch": 3.7254901960784315,
      "grad_norm": 0.1965906172990799,
      "learning_rate": 3.3680266056970965e-05,
      "loss": 0.0872,
      "step": 950
    },
    {
      "epoch": 3.764705882352941,
      "grad_norm": 0.20478536188602448,
      "learning_rate": 3.176117388936487e-05,
      "loss": 0.0877,
      "step": 960
    },
    {
      "epoch": 3.803921568627451,
      "grad_norm": 0.23426368832588196,
      "learning_rate": 2.9888003494995677e-05,
      "loss": 0.0897,
      "step": 970
    },
    {
      "epoch": 3.843137254901961,
      "grad_norm": 0.24720321595668793,
      "learning_rate": 2.806201543637571e-05,
      "loss": 0.0889,
      "step": 980
    },
    {
      "epoch": 3.8823529411764706,
      "grad_norm": 0.21011672914028168,
      "learning_rate": 2.628443852434981e-05,
      "loss": 0.0889,
      "step": 990
    },
    {
      "epoch": 3.9215686274509802,
      "grad_norm": 0.19164212048053741,
      "learning_rate": 2.4556468991158754e-05,
      "loss": 0.0874,
      "step": 1000
    },
    {
      "epoch": 3.9607843137254903,
      "grad_norm": 0.2355085164308548,
      "learning_rate": 2.287926968542674e-05,
      "loss": 0.0859,
      "step": 1010
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.14570626616477966,
      "learning_rate": 2.1253969289614807e-05,
      "loss": 0.0849,
      "step": 1020
    },
    {
      "epoch": 4.03921568627451,
      "grad_norm": 0.14226603507995605,
      "learning_rate": 1.9681661560466324e-05,
      "loss": 0.0634,
      "step": 1030
    },
    {
      "epoch": 4.078431372549019,
      "grad_norm": 0.17311133444309235,
      "learning_rate": 1.8163404592956145e-05,
      "loss": 0.0635,
      "step": 1040
    },
    {
      "epoch": 4.117647058823529,
      "grad_norm": 0.15298259258270264,
      "learning_rate": 1.6700220108238486e-05,
      "loss": 0.0598,
      "step": 1050
    },
    {
      "epoch": 4.1568627450980395,
      "grad_norm": 0.1494395136833191,
      "learning_rate": 1.5293092766072938e-05,
      "loss": 0.0619,
      "step": 1060
    },
    {
      "epoch": 4.196078431372549,
      "grad_norm": 0.13057436048984528,
      "learning_rate": 1.3942969502190983e-05,
      "loss": 0.0617,
      "step": 1070
    },
    {
      "epoch": 4.235294117647059,
      "grad_norm": 0.1499718278646469,
      "learning_rate": 1.2650758891049463e-05,
      "loss": 0.0623,
      "step": 1080
    },
    {
      "epoch": 4.2745098039215685,
      "grad_norm": 0.14512383937835693,
      "learning_rate": 1.1417330534399162e-05,
      "loss": 0.0623,
      "step": 1090
    },
    {
      "epoch": 4.313725490196078,
      "grad_norm": 0.15748609602451324,
      "learning_rate": 1.0243514476080629e-05,
      "loss": 0.0608,
      "step": 1100
    },
    {
      "epoch": 4.352941176470588,
      "grad_norm": 0.1674209088087082,
      "learning_rate": 9.130100643440587e-06,
      "loss": 0.0626,
      "step": 1110
    },
    {
      "epoch": 4.392156862745098,
      "grad_norm": 0.16120444238185883,
      "learning_rate": 8.07783831574499e-06,
      "loss": 0.0619,
      "step": 1120
    },
    {
      "epoch": 4.431372549019608,
      "grad_norm": 0.14218486845493317,
      "learning_rate": 7.087435619946692e-06,
      "loss": 0.062,
      "step": 1130
    },
    {
      "epoch": 4.470588235294118,
      "grad_norm": 0.1679692566394806,
      "learning_rate": 6.15955905414658e-06,
      "loss": 0.0602,
      "step": 1140
    },
    {
      "epoch": 4.509803921568627,
      "grad_norm": 0.16170984506607056,
      "learning_rate": 5.294833039069269e-06,
      "loss": 0.0607,
      "step": 1150
    },
    {
      "epoch": 4.549019607843137,
      "grad_norm": 0.19045813381671906,
      "learning_rate": 4.4938394978550015e-06,
      "loss": 0.0627,
      "step": 1160
    },
    {
      "epoch": 4.588235294117647,
      "grad_norm": 0.1651981770992279,
      "learning_rate": 3.757117464450566e-06,
      "loss": 0.0613,
      "step": 1170
    },
    {
      "epoch": 4.627450980392156,
      "grad_norm": 0.15379509329795837,
      "learning_rate": 3.085162720862844e-06,
      "loss": 0.0614,
      "step": 1180
    },
    {
      "epoch": 4.666666666666667,
      "grad_norm": 0.14703813195228577,
      "learning_rate": 2.4784274635190174e-06,
      "loss": 0.0596,
      "step": 1190
    },
    {
      "epoch": 4.705882352941177,
      "grad_norm": 0.14559875428676605,
      "learning_rate": 1.937319998957887e-06,
      "loss": 0.0607,
      "step": 1200
    }
  ],
  "logging_steps": 10,
  "max_steps": 1275,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.694189888282788e+17,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}