{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.1842105263157894,
  "eval_steps": 500,
  "global_step": 1800,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.006578947368421052,
      "grad_norm": 19.201007081270827,
      "learning_rate": 8.458646616541353e-07,
      "loss": 8.2522,
      "step": 10
    },
    {
      "epoch": 0.013157894736842105,
      "grad_norm": 18.114759300569585,
      "learning_rate": 1.7857142857142857e-06,
      "loss": 8.1346,
      "step": 20
    },
    {
      "epoch": 0.019736842105263157,
      "grad_norm": 9.479492840003324,
      "learning_rate": 2.725563909774436e-06,
      "loss": 7.5119,
      "step": 30
    },
    {
      "epoch": 0.02631578947368421,
      "grad_norm": 4.976331973381574,
      "learning_rate": 3.6654135338345866e-06,
      "loss": 6.7788,
      "step": 40
    },
    {
      "epoch": 0.03289473684210526,
      "grad_norm": 3.2804593438080754,
      "learning_rate": 4.605263157894737e-06,
      "loss": 6.2892,
      "step": 50
    },
    {
      "epoch": 0.039473684210526314,
      "grad_norm": 2.8583967169035027,
      "learning_rate": 5.5451127819548875e-06,
      "loss": 5.9528,
      "step": 60
    },
    {
      "epoch": 0.046052631578947366,
      "grad_norm": 3.5512069733934597,
      "learning_rate": 6.484962406015038e-06,
      "loss": 5.5755,
      "step": 70
    },
    {
      "epoch": 0.05263157894736842,
      "grad_norm": 3.731444843289522,
      "learning_rate": 7.424812030075189e-06,
      "loss": 4.9914,
      "step": 80
    },
    {
      "epoch": 0.05921052631578947,
      "grad_norm": 5.60127901105382,
      "learning_rate": 8.364661654135339e-06,
      "loss": 4.3233,
      "step": 90
    },
    {
      "epoch": 0.06578947368421052,
      "grad_norm": 3.7839180004348996,
      "learning_rate": 9.30451127819549e-06,
      "loss": 3.7305,
      "step": 100
    },
    {
      "epoch": 0.07236842105263158,
      "grad_norm": 5.421177691842472,
      "learning_rate": 1.024436090225564e-05,
      "loss": 3.3248,
      "step": 110
    },
    {
      "epoch": 0.07894736842105263,
      "grad_norm": 3.7158079343320694,
      "learning_rate": 1.118421052631579e-05,
      "loss": 3.092,
      "step": 120
    },
    {
      "epoch": 0.08552631578947369,
      "grad_norm": 5.042680618782553,
      "learning_rate": 1.212406015037594e-05,
      "loss": 2.9383,
      "step": 130
    },
    {
      "epoch": 0.09210526315789473,
      "grad_norm": 5.8029975922287615,
      "learning_rate": 1.3063909774436093e-05,
      "loss": 2.8108,
      "step": 140
    },
    {
      "epoch": 0.09868421052631579,
      "grad_norm": 7.443887556478386,
      "learning_rate": 1.4003759398496241e-05,
      "loss": 2.7256,
      "step": 150
    },
    {
      "epoch": 0.10526315789473684,
      "grad_norm": 6.942055828457074,
      "learning_rate": 1.4943609022556393e-05,
      "loss": 2.6465,
      "step": 160
    },
    {
      "epoch": 0.1118421052631579,
      "grad_norm": 20.211444588304172,
      "learning_rate": 1.5883458646616542e-05,
      "loss": 2.5756,
      "step": 170
    },
    {
      "epoch": 0.11842105263157894,
      "grad_norm": 23.482614308148037,
      "learning_rate": 1.6823308270676694e-05,
      "loss": 2.5262,
      "step": 180
    },
    {
      "epoch": 0.125,
      "grad_norm": 13.648396628618737,
      "learning_rate": 1.7763157894736842e-05,
      "loss": 2.4699,
      "step": 190
    },
    {
      "epoch": 0.13157894736842105,
      "grad_norm": 15.205800265082752,
      "learning_rate": 1.8703007518796993e-05,
      "loss": 2.4115,
      "step": 200
    },
    {
      "epoch": 0.13815789473684212,
      "grad_norm": 16.093588917694532,
      "learning_rate": 1.9642857142857145e-05,
      "loss": 2.3548,
      "step": 210
    },
    {
      "epoch": 0.14473684210526316,
      "grad_norm": 19.61654573509106,
      "learning_rate": 2.0582706766917297e-05,
      "loss": 2.3042,
      "step": 220
    },
    {
      "epoch": 0.1513157894736842,
      "grad_norm": 12.268305710278183,
      "learning_rate": 2.1522556390977445e-05,
      "loss": 2.2494,
      "step": 230
    },
    {
      "epoch": 0.15789473684210525,
      "grad_norm": 12.399788763360808,
      "learning_rate": 2.2462406015037596e-05,
      "loss": 2.1853,
      "step": 240
    },
    {
      "epoch": 0.16447368421052633,
      "grad_norm": 7.463420275410263,
      "learning_rate": 2.3402255639097744e-05,
      "loss": 2.0998,
      "step": 250
    },
    {
      "epoch": 0.17105263157894737,
      "grad_norm": 12.087535187997679,
      "learning_rate": 2.4342105263157896e-05,
      "loss": 2.0075,
      "step": 260
    },
    {
      "epoch": 0.17763157894736842,
      "grad_norm": 6.62910919540051,
      "learning_rate": 2.5281954887218044e-05,
      "loss": 1.8974,
      "step": 270
    },
    {
      "epoch": 0.18421052631578946,
      "grad_norm": 8.989470792407845,
      "learning_rate": 2.62218045112782e-05,
      "loss": 1.7928,
      "step": 280
    },
    {
      "epoch": 0.19078947368421054,
      "grad_norm": 6.928524369383665,
      "learning_rate": 2.716165413533835e-05,
      "loss": 1.7057,
      "step": 290
    },
    {
      "epoch": 0.19736842105263158,
      "grad_norm": 10.311746680652767,
      "learning_rate": 2.8101503759398495e-05,
      "loss": 1.6251,
      "step": 300
    },
    {
      "epoch": 0.20394736842105263,
      "grad_norm": 7.107313244759928,
      "learning_rate": 2.9041353383458647e-05,
      "loss": 1.5522,
      "step": 310
    },
    {
      "epoch": 0.21052631578947367,
      "grad_norm": 5.8857302431981156,
      "learning_rate": 2.9981203007518798e-05,
      "loss": 1.4974,
      "step": 320
    },
    {
      "epoch": 0.21710526315789475,
      "grad_norm": 6.775917238582609,
      "learning_rate": 3.092105263157895e-05,
      "loss": 1.4355,
      "step": 330
    },
    {
      "epoch": 0.2236842105263158,
      "grad_norm": 9.938119729849578,
      "learning_rate": 3.1860902255639094e-05,
      "loss": 1.3895,
      "step": 340
    },
    {
      "epoch": 0.23026315789473684,
      "grad_norm": 4.870796794495703,
      "learning_rate": 3.2800751879699246e-05,
      "loss": 1.3432,
      "step": 350
    },
    {
      "epoch": 0.23684210526315788,
      "grad_norm": 5.353011232164181,
      "learning_rate": 3.37406015037594e-05,
      "loss": 1.3002,
      "step": 360
    },
    {
      "epoch": 0.24342105263157895,
      "grad_norm": 7.078794515741611,
      "learning_rate": 3.468045112781955e-05,
      "loss": 1.2641,
      "step": 370
    },
    {
      "epoch": 0.25,
      "grad_norm": 5.419600483372193,
      "learning_rate": 3.56203007518797e-05,
      "loss": 1.209,
      "step": 380
    },
    {
      "epoch": 0.2565789473684211,
      "grad_norm": 3.5327527407457926,
      "learning_rate": 3.656015037593985e-05,
      "loss": 1.1644,
      "step": 390
    },
    {
      "epoch": 0.2631578947368421,
      "grad_norm": 5.894523313050658,
      "learning_rate": 3.7500000000000003e-05,
      "loss": 1.1319,
      "step": 400
    },
    {
      "epoch": 0.26973684210526316,
      "grad_norm": 3.9863193805342316,
      "learning_rate": 3.8439849624060155e-05,
      "loss": 1.094,
      "step": 410
    },
    {
      "epoch": 0.27631578947368424,
      "grad_norm": 4.225678742387179,
      "learning_rate": 3.9379699248120307e-05,
      "loss": 1.0651,
      "step": 420
    },
    {
      "epoch": 0.28289473684210525,
      "grad_norm": 5.586389531542015,
      "learning_rate": 4.031954887218045e-05,
      "loss": 1.0247,
      "step": 430
    },
    {
      "epoch": 0.2894736842105263,
      "grad_norm": 7.0928476432315675,
      "learning_rate": 4.12593984962406e-05,
      "loss": 0.9973,
      "step": 440
    },
    {
      "epoch": 0.29605263157894735,
      "grad_norm": 5.797737179453841,
      "learning_rate": 4.2199248120300754e-05,
      "loss": 0.9738,
      "step": 450
    },
    {
      "epoch": 0.3026315789473684,
      "grad_norm": 5.436160406452568,
      "learning_rate": 4.3139097744360906e-05,
      "loss": 0.9539,
      "step": 460
    },
    {
      "epoch": 0.3092105263157895,
      "grad_norm": 4.940473111611711,
      "learning_rate": 4.407894736842105e-05,
      "loss": 0.9197,
      "step": 470
    },
    {
      "epoch": 0.3157894736842105,
      "grad_norm": 5.844247327503255,
      "learning_rate": 4.50187969924812e-05,
      "loss": 0.8862,
      "step": 480
    },
    {
      "epoch": 0.3223684210526316,
      "grad_norm": 4.744124381908569,
      "learning_rate": 4.5958646616541354e-05,
      "loss": 0.8668,
      "step": 490
    },
    {
      "epoch": 0.32894736842105265,
      "grad_norm": 4.689028836800058,
      "learning_rate": 4.6898496240601505e-05,
      "loss": 0.8476,
      "step": 500
    },
    {
      "epoch": 0.3355263157894737,
      "grad_norm": 4.398308995209079,
      "learning_rate": 4.783834586466166e-05,
      "loss": 0.82,
      "step": 510
    },
    {
      "epoch": 0.34210526315789475,
      "grad_norm": 5.554811082819198,
      "learning_rate": 4.877819548872181e-05,
      "loss": 0.7974,
      "step": 520
    },
    {
      "epoch": 0.34868421052631576,
      "grad_norm": 3.7243976034524584,
      "learning_rate": 4.971804511278196e-05,
      "loss": 0.7715,
      "step": 530
    },
    {
      "epoch": 0.35526315789473684,
      "grad_norm": 4.879577435611285,
      "learning_rate": 5.0657894736842104e-05,
      "loss": 0.7514,
      "step": 540
    },
    {
      "epoch": 0.3618421052631579,
      "grad_norm": 5.516075231521318,
      "learning_rate": 5.1597744360902256e-05,
      "loss": 0.7291,
      "step": 550
    },
    {
      "epoch": 0.3684210526315789,
      "grad_norm": 5.494655015332771,
      "learning_rate": 5.253759398496241e-05,
      "loss": 0.6974,
      "step": 560
    },
    {
      "epoch": 0.375,
      "grad_norm": 5.0610034411984826,
      "learning_rate": 5.347744360902256e-05,
      "loss": 0.6779,
      "step": 570
    },
    {
      "epoch": 0.3815789473684211,
      "grad_norm": 6.241132985448258,
      "learning_rate": 5.441729323308271e-05,
      "loss": 0.6723,
      "step": 580
    },
    {
      "epoch": 0.3881578947368421,
      "grad_norm": 4.582816961569543,
      "learning_rate": 5.535714285714286e-05,
      "loss": 0.652,
      "step": 590
    },
    {
      "epoch": 0.39473684210526316,
      "grad_norm": 3.5845708789821042,
      "learning_rate": 5.6296992481203013e-05,
      "loss": 0.6154,
      "step": 600
    },
    {
      "epoch": 0.40131578947368424,
      "grad_norm": 4.328768044098044,
      "learning_rate": 5.723684210526316e-05,
      "loss": 0.6042,
      "step": 610
    },
    {
      "epoch": 0.40789473684210525,
      "grad_norm": 3.5473602851751793,
      "learning_rate": 5.817669172932331e-05,
      "loss": 0.5803,
      "step": 620
    },
    {
      "epoch": 0.4144736842105263,
      "grad_norm": 4.233181045605189,
      "learning_rate": 5.911654135338346e-05,
      "loss": 0.5688,
      "step": 630
    },
    {
      "epoch": 0.42105263157894735,
      "grad_norm": 4.0831509210740835,
      "learning_rate": 6.005639097744361e-05,
      "loss": 0.5385,
      "step": 640
    },
    {
      "epoch": 0.4276315789473684,
      "grad_norm": 3.8705204303765814,
      "learning_rate": 6.0996240601503764e-05,
      "loss": 0.5201,
      "step": 650
    },
    {
      "epoch": 0.4342105263157895,
      "grad_norm": 3.8269827275761337,
      "learning_rate": 6.193609022556392e-05,
      "loss": 0.504,
      "step": 660
    },
    {
      "epoch": 0.4407894736842105,
      "grad_norm": 3.7042333223437476,
      "learning_rate": 6.287593984962407e-05,
      "loss": 0.485,
      "step": 670
    },
    {
      "epoch": 0.4473684210526316,
      "grad_norm": 3.1454584376141725,
      "learning_rate": 6.381578947368422e-05,
      "loss": 0.4692,
      "step": 680
    },
    {
      "epoch": 0.45394736842105265,
      "grad_norm": 3.6715281367823698,
      "learning_rate": 6.475563909774437e-05,
      "loss": 0.4474,
      "step": 690
    },
    {
      "epoch": 0.4605263157894737,
      "grad_norm": 3.3520165299765203,
      "learning_rate": 6.569548872180451e-05,
      "loss": 0.4335,
      "step": 700
    },
    {
      "epoch": 0.46710526315789475,
      "grad_norm": 3.4830856151143554,
      "learning_rate": 6.663533834586466e-05,
      "loss": 0.4156,
      "step": 710
    },
    {
      "epoch": 0.47368421052631576,
      "grad_norm": 3.4130338264227835,
      "learning_rate": 6.757518796992481e-05,
      "loss": 0.3984,
      "step": 720
    },
    {
      "epoch": 0.48026315789473684,
      "grad_norm": 3.6699720598135834,
      "learning_rate": 6.851503759398496e-05,
      "loss": 0.3884,
      "step": 730
    },
    {
      "epoch": 0.4868421052631579,
      "grad_norm": 3.2598730307422192,
      "learning_rate": 6.945488721804511e-05,
      "loss": 0.3727,
      "step": 740
    },
    {
      "epoch": 0.4934210526315789,
      "grad_norm": 3.0065146921519097,
      "learning_rate": 7.039473684210527e-05,
      "loss": 0.3521,
      "step": 750
    },
    {
      "epoch": 0.5,
      "grad_norm": 2.3801132893266956,
      "learning_rate": 7.133458646616542e-05,
      "loss": 0.3415,
      "step": 760
    },
    {
      "epoch": 0.506578947368421,
      "grad_norm": 2.5040976105308466,
      "learning_rate": 7.227443609022557e-05,
      "loss": 0.3303,
      "step": 770
    },
    {
      "epoch": 0.5131578947368421,
      "grad_norm": 2.3095873713136004,
      "learning_rate": 7.321428571428571e-05,
      "loss": 0.3155,
      "step": 780
    },
    {
      "epoch": 0.5197368421052632,
      "grad_norm": 2.221909098699267,
      "learning_rate": 7.415413533834586e-05,
      "loss": 0.3,
      "step": 790
    },
    {
      "epoch": 0.5263157894736842,
      "grad_norm": 2.3650008130107008,
      "learning_rate": 7.509398496240601e-05,
      "loss": 0.2881,
      "step": 800
    },
    {
      "epoch": 0.5328947368421053,
      "grad_norm": 2.92228181953594,
      "learning_rate": 7.603383458646618e-05,
      "loss": 0.2759,
      "step": 810
    },
    {
      "epoch": 0.5394736842105263,
      "grad_norm": 2.39576316770683,
      "learning_rate": 7.697368421052633e-05,
      "loss": 0.2611,
      "step": 820
    },
    {
      "epoch": 0.5460526315789473,
      "grad_norm": 1.9345517046210943,
      "learning_rate": 7.791353383458648e-05,
      "loss": 0.2472,
      "step": 830
    },
    {
      "epoch": 0.5526315789473685,
      "grad_norm": 2.321485582474392,
      "learning_rate": 7.885338345864663e-05,
      "loss": 0.2409,
      "step": 840
    },
    {
      "epoch": 0.5592105263157895,
      "grad_norm": 1.8690657258212422,
      "learning_rate": 7.979323308270678e-05,
      "loss": 0.2308,
      "step": 850
    },
    {
      "epoch": 0.5657894736842105,
      "grad_norm": 1.901049780468097,
      "learning_rate": 8.073308270676692e-05,
      "loss": 0.2171,
      "step": 860
    },
    {
      "epoch": 0.5723684210526315,
      "grad_norm": 2.2229311039801596,
      "learning_rate": 8.167293233082707e-05,
      "loss": 0.2072,
      "step": 870
    },
    {
      "epoch": 0.5789473684210527,
      "grad_norm": 2.213221418124993,
      "learning_rate": 8.261278195488722e-05,
      "loss": 0.1965,
      "step": 880
    },
    {
      "epoch": 0.5855263157894737,
      "grad_norm": 1.870540842703205,
      "learning_rate": 8.355263157894737e-05,
      "loss": 0.1877,
      "step": 890
    },
    {
      "epoch": 0.5921052631578947,
      "grad_norm": 1.311680351399902,
      "learning_rate": 8.449248120300753e-05,
      "loss": 0.1792,
      "step": 900
    },
    {
      "epoch": 0.5986842105263158,
      "grad_norm": 1.5798663811679752,
      "learning_rate": 8.543233082706768e-05,
      "loss": 0.1716,
      "step": 910
    },
    {
      "epoch": 0.6052631578947368,
      "grad_norm": 7.169134004434136,
      "learning_rate": 8.637218045112783e-05,
      "loss": 0.1721,
      "step": 920
    },
    {
      "epoch": 0.6118421052631579,
      "grad_norm": 1.5044155665130965,
      "learning_rate": 8.731203007518798e-05,
      "loss": 0.193,
      "step": 930
    },
    {
      "epoch": 0.618421052631579,
      "grad_norm": 1.305309283144098,
      "learning_rate": 8.825187969924813e-05,
      "loss": 0.1566,
      "step": 940
    },
    {
      "epoch": 0.625,
      "grad_norm": 1.23686939558907,
      "learning_rate": 8.919172932330827e-05,
      "loss": 0.1445,
      "step": 950
    },
    {
      "epoch": 0.631578947368421,
      "grad_norm": 1.0445490067006857,
      "learning_rate": 9.013157894736842e-05,
      "loss": 0.1356,
      "step": 960
    },
    {
      "epoch": 0.6381578947368421,
      "grad_norm": 1.3136634908844078,
      "learning_rate": 9.107142857142857e-05,
      "loss": 0.1308,
      "step": 970
    },
    {
      "epoch": 0.6447368421052632,
      "grad_norm": 1.7716809252302665,
      "learning_rate": 9.201127819548872e-05,
      "loss": 0.124,
      "step": 980
    },
    {
      "epoch": 0.6513157894736842,
      "grad_norm": 1.0772175167269311,
      "learning_rate": 9.295112781954888e-05,
      "loss": 0.1202,
      "step": 990
    },
    {
      "epoch": 0.6578947368421053,
      "grad_norm": 1.1878390185089365,
      "learning_rate": 9.389097744360903e-05,
      "loss": 0.114,
      "step": 1000
    },
    {
      "epoch": 0.6644736842105263,
      "grad_norm": 0.8915150933559234,
      "learning_rate": 9.483082706766918e-05,
      "loss": 0.1054,
      "step": 1010
    },
    {
      "epoch": 0.6710526315789473,
      "grad_norm": 0.7818448157917527,
      "learning_rate": 9.577067669172933e-05,
      "loss": 0.0996,
      "step": 1020
    },
    {
      "epoch": 0.6776315789473685,
      "grad_norm": 0.990430969229284,
      "learning_rate": 9.671052631578947e-05,
      "loss": 0.0974,
      "step": 1030
    },
    {
      "epoch": 0.6842105263157895,
      "grad_norm": 0.9276518666370865,
      "learning_rate": 9.765037593984962e-05,
      "loss": 0.0933,
      "step": 1040
    },
    {
      "epoch": 0.6907894736842105,
      "grad_norm": 0.6846215892309722,
      "learning_rate": 9.859022556390977e-05,
      "loss": 0.0874,
      "step": 1050
    },
    {
      "epoch": 0.6973684210526315,
      "grad_norm": 0.714553023496577,
      "learning_rate": 9.953007518796992e-05,
      "loss": 0.0814,
      "step": 1060
    },
    {
      "epoch": 0.7039473684210527,
      "grad_norm": 0.7546346945965402,
      "learning_rate": 9.999993273155467e-05,
      "loss": 0.0774,
      "step": 1070
    },
    {
      "epoch": 0.7105263157894737,
      "grad_norm": 0.571192300635228,
      "learning_rate": 9.999939458507816e-05,
      "loss": 0.0762,
      "step": 1080
    },
    {
      "epoch": 0.7171052631578947,
      "grad_norm": 0.8584193322720889,
      "learning_rate": 9.999831829791713e-05,
      "loss": 0.0756,
      "step": 1090
    },
    {
      "epoch": 0.7236842105263158,
      "grad_norm": 1.3952479453359226,
      "learning_rate": 9.999670388165563e-05,
      "loss": 0.075,
      "step": 1100
    },
    {
      "epoch": 0.7302631578947368,
      "grad_norm": 0.751465926189691,
      "learning_rate": 9.999455135366951e-05,
      "loss": 0.069,
      "step": 1110
    },
    {
      "epoch": 0.7368421052631579,
      "grad_norm": 0.4873110858150168,
      "learning_rate": 9.999186073712635e-05,
      "loss": 0.064,
      "step": 1120
    },
    {
      "epoch": 0.743421052631579,
      "grad_norm": 0.4495642582063532,
      "learning_rate": 9.998863206098507e-05,
      "loss": 0.0623,
      "step": 1130
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.45684969056385805,
      "learning_rate": 9.998486535999573e-05,
      "loss": 0.059,
      "step": 1140
    },
    {
      "epoch": 0.756578947368421,
      "grad_norm": 0.4304228051311271,
      "learning_rate": 9.998056067469915e-05,
      "loss": 0.0574,
      "step": 1150
    },
    {
      "epoch": 0.7631578947368421,
      "grad_norm": 0.38692754350368835,
      "learning_rate": 9.997571805142639e-05,
      "loss": 0.0568,
      "step": 1160
    },
    {
      "epoch": 0.7697368421052632,
      "grad_norm": 0.44697102978905795,
      "learning_rate": 9.997033754229835e-05,
      "loss": 0.054,
      "step": 1170
    },
    {
      "epoch": 0.7763157894736842,
      "grad_norm": 0.3475319362475707,
      "learning_rate": 9.996441920522512e-05,
      "loss": 0.0512,
      "step": 1180
    },
    {
      "epoch": 0.7828947368421053,
      "grad_norm": 0.3776964171490181,
      "learning_rate": 9.995796310390547e-05,
      "loss": 0.0496,
      "step": 1190
    },
    {
      "epoch": 0.7894736842105263,
      "grad_norm": 0.3599383764053137,
      "learning_rate": 9.995096930782604e-05,
      "loss": 0.049,
      "step": 1200
    },
    {
      "epoch": 0.7960526315789473,
      "grad_norm": 0.4004701605242826,
      "learning_rate": 9.994343789226066e-05,
      "loss": 0.0468,
      "step": 1210
    },
    {
      "epoch": 0.8026315789473685,
      "grad_norm": 0.3822329385724864,
      "learning_rate": 9.993536893826954e-05,
      "loss": 0.0457,
      "step": 1220
    },
    {
      "epoch": 0.8092105263157895,
      "grad_norm": 0.3620248875103048,
      "learning_rate": 9.992676253269837e-05,
      "loss": 0.0446,
      "step": 1230
    },
    {
      "epoch": 0.8157894736842105,
      "grad_norm": 0.28358957764617276,
      "learning_rate": 9.991761876817743e-05,
      "loss": 0.0429,
      "step": 1240
    },
    {
      "epoch": 0.8223684210526315,
      "grad_norm": 0.32929903528232773,
      "learning_rate": 9.990793774312053e-05,
      "loss": 0.0421,
      "step": 1250
    },
    {
      "epoch": 0.8289473684210527,
      "grad_norm": 0.34141322063560897,
      "learning_rate": 9.9897719561724e-05,
      "loss": 0.0418,
      "step": 1260
    },
    {
      "epoch": 0.8355263157894737,
      "grad_norm": 0.28296527941759625,
      "learning_rate": 9.988696433396557e-05,
      "loss": 0.0403,
      "step": 1270
    },
    {
      "epoch": 0.8421052631578947,
      "grad_norm": 0.3686897659419288,
      "learning_rate": 9.987567217560314e-05,
      "loss": 0.0395,
      "step": 1280
    },
    {
      "epoch": 0.8486842105263158,
      "grad_norm": 0.3066999773216274,
      "learning_rate": 9.986384320817359e-05,
      "loss": 0.0395,
      "step": 1290
    },
    {
      "epoch": 0.8552631578947368,
      "grad_norm": 0.33225917789321974,
      "learning_rate": 9.985147755899142e-05,
      "loss": 0.0385,
      "step": 1300
    },
    {
      "epoch": 0.8618421052631579,
      "grad_norm": 0.3004075266809671,
      "learning_rate": 9.983857536114743e-05,
      "loss": 0.0377,
      "step": 1310
    },
    {
      "epoch": 0.868421052631579,
      "grad_norm": 0.2698302147467991,
      "learning_rate": 9.982513675350724e-05,
      "loss": 0.0367,
      "step": 1320
    },
    {
      "epoch": 0.875,
      "grad_norm": 0.26807280624783797,
      "learning_rate": 9.981116188070984e-05,
      "loss": 0.0356,
      "step": 1330
    },
    {
      "epoch": 0.881578947368421,
      "grad_norm": 0.3127969801843736,
      "learning_rate": 9.979665089316602e-05,
      "loss": 0.035,
      "step": 1340
    },
    {
      "epoch": 0.8881578947368421,
      "grad_norm": 0.28653732054861003,
      "learning_rate": 9.978160394705668e-05,
      "loss": 0.0343,
      "step": 1350
    },
    {
      "epoch": 0.8947368421052632,
      "grad_norm": 0.2504478087548026,
      "learning_rate": 9.976602120433133e-05,
      "loss": 0.0341,
      "step": 1360
    },
    {
      "epoch": 0.9013157894736842,
      "grad_norm": 0.24532062538161709,
      "learning_rate": 9.97499028327061e-05,
      "loss": 0.0331,
      "step": 1370
    },
    {
      "epoch": 0.9078947368421053,
      "grad_norm": 0.5595503828448865,
      "learning_rate": 9.973324900566213e-05,
      "loss": 0.041,
      "step": 1380
    },
    {
      "epoch": 0.9144736842105263,
      "grad_norm": 0.32103277633584737,
      "learning_rate": 9.971605990244365e-05,
      "loss": 0.0396,
      "step": 1390
    },
    {
      "epoch": 0.9210526315789473,
      "grad_norm": 0.39243105455264066,
      "learning_rate": 9.969833570805599e-05,
      "loss": 0.0345,
      "step": 1400
    },
    {
      "epoch": 0.9276315789473685,
      "grad_norm": 0.24537350120004972,
      "learning_rate": 9.968007661326369e-05,
      "loss": 0.0323,
      "step": 1410
    },
    {
      "epoch": 0.9342105263157895,
      "grad_norm": 0.21282115870790244,
      "learning_rate": 9.966128281458832e-05,
      "loss": 0.0309,
      "step": 1420
    },
    {
      "epoch": 0.9407894736842105,
      "grad_norm": 0.22792100873990018,
      "learning_rate": 9.964195451430653e-05,
      "loss": 0.0302,
      "step": 1430
    },
    {
      "epoch": 0.9473684210526315,
      "grad_norm": 0.2513211642537743,
      "learning_rate": 9.962209192044772e-05,
      "loss": 0.0298,
      "step": 1440
    },
    {
      "epoch": 0.9539473684210527,
      "grad_norm": 0.25009559541845094,
      "learning_rate": 9.960169524679187e-05,
      "loss": 0.0293,
      "step": 1450
    },
    {
      "epoch": 0.9605263157894737,
      "grad_norm": 0.23293995789061867,
      "learning_rate": 9.958076471286724e-05,
      "loss": 0.0289,
      "step": 1460
    },
    {
      "epoch": 0.9671052631578947,
      "grad_norm": 0.22361624312785255,
      "learning_rate": 9.955930054394798e-05,
      "loss": 0.0282,
      "step": 1470
    },
    {
      "epoch": 0.9736842105263158,
      "grad_norm": 0.23906361425178743,
      "learning_rate": 9.953730297105175e-05,
      "loss": 0.0281,
      "step": 1480
    },
    {
      "epoch": 0.9802631578947368,
      "grad_norm": 0.23143391534241473,
      "learning_rate": 9.951477223093722e-05,
      "loss": 0.0275,
      "step": 1490
    },
    {
      "epoch": 0.9868421052631579,
      "grad_norm": 0.2147052689574608,
      "learning_rate": 9.949170856610144e-05,
      "loss": 0.0271,
      "step": 1500
    },
    {
      "epoch": 0.993421052631579,
      "grad_norm": 0.20377711124649098,
      "learning_rate": 9.946811222477737e-05,
      "loss": 0.0263,
      "step": 1510
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.2135903768894096,
      "learning_rate": 9.944398346093109e-05,
      "loss": 0.0262,
      "step": 1520
    },
    {
      "epoch": 1.006578947368421,
      "grad_norm": 0.20254766138767108,
      "learning_rate": 9.941932253425917e-05,
      "loss": 0.0248,
      "step": 1530
    },
    {
      "epoch": 1.013157894736842,
      "grad_norm": 0.2197030932035537,
      "learning_rate": 9.939412971018574e-05,
      "loss": 0.0246,
      "step": 1540
    },
    {
      "epoch": 1.019736842105263,
      "grad_norm": 0.1933628837351623,
      "learning_rate": 9.936840525985978e-05,
      "loss": 0.0242,
      "step": 1550
    },
    {
      "epoch": 1.0263157894736843,
      "grad_norm": 0.19721213696918533,
      "learning_rate": 9.934214946015213e-05,
      "loss": 0.0238,
      "step": 1560
    },
    {
      "epoch": 1.0328947368421053,
      "grad_norm": 0.20774618691091581,
      "learning_rate": 9.931536259365244e-05,
      "loss": 0.0234,
      "step": 1570
    },
    {
      "epoch": 1.0394736842105263,
      "grad_norm": 0.19785742831408606,
      "learning_rate": 9.928804494866629e-05,
      "loss": 0.0228,
      "step": 1580
    },
    {
      "epoch": 1.0460526315789473,
      "grad_norm": 0.2035358859489929,
      "learning_rate": 9.926019681921196e-05,
      "loss": 0.0227,
      "step": 1590
    },
    {
      "epoch": 1.0526315789473684,
      "grad_norm": 0.20629560858059492,
      "learning_rate": 9.92318185050173e-05,
      "loss": 0.0227,
      "step": 1600
    },
    {
      "epoch": 1.0592105263157894,
      "grad_norm": 0.17943582514083126,
      "learning_rate": 9.920291031151652e-05,
      "loss": 0.0223,
      "step": 1610
    },
    {
      "epoch": 1.0657894736842106,
      "grad_norm": 0.2069047937737935,
      "learning_rate": 9.917347254984688e-05,
      "loss": 0.0219,
      "step": 1620
    },
    {
      "epoch": 1.0723684210526316,
      "grad_norm": 0.18149990279040348,
      "learning_rate": 9.91435055368454e-05,
      "loss": 0.0215,
      "step": 1630
    },
    {
      "epoch": 1.0789473684210527,
      "grad_norm": 0.21588259501861692,
      "learning_rate": 9.91130095950453e-05,
      "loss": 0.0216,
      "step": 1640
    },
    {
      "epoch": 1.0855263157894737,
      "grad_norm": 0.19248955273518795,
      "learning_rate": 9.908198505267273e-05,
      "loss": 0.0211,
      "step": 1650
    },
    {
      "epoch": 1.0921052631578947,
      "grad_norm": 0.21323425974286558,
      "learning_rate": 9.90504322436431e-05,
      "loss": 0.0212,
      "step": 1660
    },
    {
      "epoch": 1.0986842105263157,
      "grad_norm": 0.19659430552879165,
      "learning_rate": 9.901835150755754e-05,
      "loss": 0.0204,
      "step": 1670
    },
    {
      "epoch": 1.1052631578947367,
      "grad_norm": 0.18723287056746327,
      "learning_rate": 9.898574318969918e-05,
      "loss": 0.0202,
      "step": 1680
    },
    {
      "epoch": 1.111842105263158,
      "grad_norm": 0.18512625335755417,
      "learning_rate": 9.895260764102956e-05,
      "loss": 0.0198,
      "step": 1690
    },
    {
      "epoch": 1.118421052631579,
      "grad_norm": 0.19124114649058113,
      "learning_rate": 9.89189452181847e-05,
      "loss": 0.0201,
      "step": 1700
    },
    {
      "epoch": 1.125,
      "grad_norm": 0.18895302686535878,
      "learning_rate": 9.888475628347142e-05,
      "loss": 0.0195,
      "step": 1710
    },
    {
      "epoch": 1.131578947368421,
      "grad_norm": 0.1772267382799342,
      "learning_rate": 9.885004120486326e-05,
      "loss": 0.0191,
      "step": 1720
    },
    {
      "epoch": 1.138157894736842,
      "grad_norm": 0.17908235775562148,
      "learning_rate": 9.881480035599667e-05,
      "loss": 0.0189,
      "step": 1730
    },
    {
      "epoch": 1.1447368421052633,
      "grad_norm": 0.17580586627136524,
      "learning_rate": 9.877903411616697e-05,
      "loss": 0.0189,
      "step": 1740
    },
    {
      "epoch": 1.1513157894736843,
      "grad_norm": 0.18518601084678493,
      "learning_rate": 9.874274287032417e-05,
      "loss": 0.0186,
      "step": 1750
    },
    {
      "epoch": 1.1578947368421053,
      "grad_norm": 0.18048608002452143,
      "learning_rate": 9.870592700906892e-05,
      "loss": 0.0183,
      "step": 1760
    },
    {
      "epoch": 1.1644736842105263,
      "grad_norm": 0.15777383982627308,
      "learning_rate": 9.866858692864829e-05,
      "loss": 0.018,
      "step": 1770
    },
    {
      "epoch": 1.1710526315789473,
      "grad_norm": 0.20607208990685497,
      "learning_rate": 9.863072303095144e-05,
      "loss": 0.0175,
      "step": 1780
    },
    {
      "epoch": 1.1776315789473684,
      "grad_norm": 0.1611318675087161,
      "learning_rate": 9.859233572350541e-05,
      "loss": 0.0171,
      "step": 1790
    },
    {
      "epoch": 1.1842105263157894,
      "grad_norm": 0.17589801094486104,
      "learning_rate": 9.855342541947065e-05,
      "loss": 0.0171,
      "step": 1800
    }
  ],
  "logging_steps": 10,
  "max_steps": 10640,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 7,
  "save_steps": 1800,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 143324589588480.0,
  "train_batch_size": 128,
  "trial_name": null,
  "trial_params": null
}