{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 5000,
  "global_step": 2250,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0044444444444444444,
      "grad_norm": 152.64707946777344,
      "learning_rate": 5.000000000000001e-07,
      "loss": 10.2908,
      "step": 10
    },
    {
      "epoch": 0.008888888888888889,
      "grad_norm": 211.05294799804688,
      "learning_rate": 1.4000000000000001e-06,
      "loss": 10.0664,
      "step": 20
    },
    {
      "epoch": 0.013333333333333334,
      "grad_norm": 2954.0419921875,
      "learning_rate": 2.4000000000000003e-06,
      "loss": 10.0624,
      "step": 30
    },
    {
      "epoch": 0.017777777777777778,
      "grad_norm": 1075.460205078125,
      "learning_rate": 3.4000000000000005e-06,
      "loss": 8.5938,
      "step": 40
    },
    {
      "epoch": 0.022222222222222223,
      "grad_norm": 7704.75048828125,
      "learning_rate": 4.4e-06,
      "loss": 7.8263,
      "step": 50
    },
    {
      "epoch": 0.02666666666666667,
      "grad_norm": 6002.2890625,
      "learning_rate": 5.4e-06,
      "loss": 6.2896,
      "step": 60
    },
    {
      "epoch": 0.03111111111111111,
      "grad_norm": 14311.9560546875,
      "learning_rate": 6.4000000000000006e-06,
      "loss": 5.4298,
      "step": 70
    },
    {
      "epoch": 0.035555555555555556,
      "grad_norm": 5250.40869140625,
      "learning_rate": 7.4e-06,
      "loss": 5.2735,
      "step": 80
    },
    {
      "epoch": 0.04,
      "grad_norm": 470.0221862792969,
      "learning_rate": 8.400000000000001e-06,
      "loss": 4.9127,
      "step": 90
    },
    {
      "epoch": 0.044444444444444446,
      "grad_norm": 19534.701171875,
      "learning_rate": 9.4e-06,
      "loss": 4.6077,
      "step": 100
    },
    {
      "epoch": 0.04888888888888889,
      "grad_norm": 492.7581481933594,
      "learning_rate": 1.04e-05,
      "loss": 4.5512,
      "step": 110
    },
    {
      "epoch": 0.05333333333333334,
      "grad_norm": 9192.6767578125,
      "learning_rate": 1.13e-05,
      "loss": 5.1424,
      "step": 120
    },
    {
      "epoch": 0.057777777777777775,
      "grad_norm": 108.0229263305664,
      "learning_rate": 1.23e-05,
      "loss": 3.7802,
      "step": 130
    },
    {
      "epoch": 0.06222222222222222,
      "grad_norm": 1061.415771484375,
      "learning_rate": 1.3300000000000001e-05,
      "loss": 3.3019,
      "step": 140
    },
    {
      "epoch": 0.06666666666666667,
      "grad_norm": 128.2921600341797,
      "learning_rate": 1.43e-05,
      "loss": 2.9124,
      "step": 150
    },
    {
      "epoch": 0.07111111111111111,
      "grad_norm": 29.611740112304688,
      "learning_rate": 1.53e-05,
      "loss": 2.6086,
      "step": 160
    },
    {
      "epoch": 0.07555555555555556,
      "grad_norm": 43.61715316772461,
      "learning_rate": 1.63e-05,
      "loss": 1.8848,
      "step": 170
    },
    {
      "epoch": 0.08,
      "grad_norm": 11403.16015625,
      "learning_rate": 1.73e-05,
      "loss": 1.9176,
      "step": 180
    },
    {
      "epoch": 0.08444444444444445,
      "grad_norm": 381.3476867675781,
      "learning_rate": 1.83e-05,
      "loss": 1.734,
      "step": 190
    },
    {
      "epoch": 0.08888888888888889,
      "grad_norm": 4.966193199157715,
      "learning_rate": 1.93e-05,
      "loss": 1.3579,
      "step": 200
    },
    {
      "epoch": 0.09333333333333334,
      "grad_norm": 4.789448261260986,
      "learning_rate": 2.0300000000000002e-05,
      "loss": 1.0905,
      "step": 210
    },
    {
      "epoch": 0.09777777777777778,
      "grad_norm": 2.323220729827881,
      "learning_rate": 2.13e-05,
      "loss": 1.0379,
      "step": 220
    },
    {
      "epoch": 0.10222222222222223,
      "grad_norm": 1.8111121654510498,
      "learning_rate": 2.23e-05,
      "loss": 0.9473,
      "step": 230
    },
    {
      "epoch": 0.10666666666666667,
      "grad_norm": 4.103352069854736,
      "learning_rate": 2.3300000000000004e-05,
      "loss": 1.0941,
      "step": 240
    },
    {
      "epoch": 0.1111111111111111,
      "grad_norm": 2.7268946170806885,
      "learning_rate": 2.43e-05,
      "loss": 1.1331,
      "step": 250
    },
    {
      "epoch": 0.11555555555555555,
      "grad_norm": 36.0265998840332,
      "learning_rate": 2.5300000000000002e-05,
      "loss": 1.3104,
      "step": 260
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.8667157888412476,
      "learning_rate": 2.6300000000000002e-05,
      "loss": 1.0259,
      "step": 270
    },
    {
      "epoch": 0.12444444444444444,
      "grad_norm": 5.474687099456787,
      "learning_rate": 2.7300000000000003e-05,
      "loss": 1.0916,
      "step": 280
    },
    {
      "epoch": 0.1288888888888889,
      "grad_norm": 2.022836208343506,
      "learning_rate": 2.83e-05,
      "loss": 0.9872,
      "step": 290
    },
    {
      "epoch": 0.13333333333333333,
      "grad_norm": 1.9781649112701416,
      "learning_rate": 2.93e-05,
      "loss": 0.8844,
      "step": 300
    },
    {
      "epoch": 0.13777777777777778,
      "grad_norm": 1.9537264108657837,
      "learning_rate": 3.03e-05,
      "loss": 1.0466,
      "step": 310
    },
    {
      "epoch": 0.14222222222222222,
      "grad_norm": 2.3547990322113037,
      "learning_rate": 3.13e-05,
      "loss": 1.0016,
      "step": 320
    },
    {
      "epoch": 0.14666666666666667,
      "grad_norm": 2.5301690101623535,
      "learning_rate": 3.2300000000000006e-05,
      "loss": 1.015,
      "step": 330
    },
    {
      "epoch": 0.1511111111111111,
      "grad_norm": 1.698588252067566,
      "learning_rate": 3.33e-05,
      "loss": 0.994,
      "step": 340
    },
    {
      "epoch": 0.15555555555555556,
      "grad_norm": 2.17368221282959,
      "learning_rate": 3.430000000000001e-05,
      "loss": 1.0743,
      "step": 350
    },
    {
      "epoch": 0.16,
      "grad_norm": 2.106858968734741,
      "learning_rate": 3.53e-05,
      "loss": 1.161,
      "step": 360
    },
    {
      "epoch": 0.16444444444444445,
      "grad_norm": 2.4343013763427734,
      "learning_rate": 3.63e-05,
      "loss": 1.0599,
      "step": 370
    },
    {
      "epoch": 0.1688888888888889,
      "grad_norm": 2.284984588623047,
      "learning_rate": 3.73e-05,
      "loss": 0.9959,
      "step": 380
    },
    {
      "epoch": 0.17333333333333334,
      "grad_norm": 1.8927963972091675,
      "learning_rate": 3.83e-05,
      "loss": 1.0063,
      "step": 390
    },
    {
      "epoch": 0.17777777777777778,
      "grad_norm": 2.731109142303467,
      "learning_rate": 3.9300000000000007e-05,
      "loss": 1.0467,
      "step": 400
    },
    {
      "epoch": 0.18222222222222223,
      "grad_norm": 9.546753883361816,
      "learning_rate": 4.0300000000000004e-05,
      "loss": 1.0892,
      "step": 410
    },
    {
      "epoch": 0.18666666666666668,
      "grad_norm": 1.986333966255188,
      "learning_rate": 4.13e-05,
      "loss": 1.0233,
      "step": 420
    },
    {
      "epoch": 0.19111111111111112,
      "grad_norm": 2.203075408935547,
      "learning_rate": 4.23e-05,
      "loss": 1.268,
      "step": 430
    },
    {
      "epoch": 0.19555555555555557,
      "grad_norm": 2.235809564590454,
      "learning_rate": 4.33e-05,
      "loss": 0.9937,
      "step": 440
    },
    {
      "epoch": 0.2,
      "grad_norm": 3.598283290863037,
      "learning_rate": 4.43e-05,
      "loss": 0.8837,
      "step": 450
    },
    {
      "epoch": 0.20444444444444446,
      "grad_norm": 1.872710108757019,
      "learning_rate": 4.53e-05,
      "loss": 1.1982,
      "step": 460
    },
    {
      "epoch": 0.2088888888888889,
      "grad_norm": 3.0466055870056152,
      "learning_rate": 4.630000000000001e-05,
      "loss": 1.1995,
      "step": 470
    },
    {
      "epoch": 0.21333333333333335,
      "grad_norm": 2.7732715606689453,
      "learning_rate": 4.73e-05,
      "loss": 1.0711,
      "step": 480
    },
    {
      "epoch": 0.21777777777777776,
      "grad_norm": 2.3050129413604736,
      "learning_rate": 4.83e-05,
      "loss": 1.2792,
      "step": 490
    },
    {
      "epoch": 0.2222222222222222,
      "grad_norm": 2.128685235977173,
      "learning_rate": 4.93e-05,
      "loss": 1.1137,
      "step": 500
    },
    {
      "epoch": 0.22666666666666666,
      "grad_norm": 2.2330660820007324,
      "learning_rate": 4.9914285714285717e-05,
      "loss": 1.0332,
      "step": 510
    },
    {
      "epoch": 0.2311111111111111,
      "grad_norm": 2.049591541290283,
      "learning_rate": 4.962857142857143e-05,
      "loss": 1.1467,
      "step": 520
    },
    {
      "epoch": 0.23555555555555555,
      "grad_norm": 2.388408660888672,
      "learning_rate": 4.934285714285715e-05,
      "loss": 1.1017,
      "step": 530
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.9390537738800049,
      "learning_rate": 4.905714285714286e-05,
      "loss": 1.0874,
      "step": 540
    },
    {
      "epoch": 0.24444444444444444,
      "grad_norm": 1.530515432357788,
      "learning_rate": 4.8771428571428574e-05,
      "loss": 1.0236,
      "step": 550
    },
    {
      "epoch": 0.24888888888888888,
      "grad_norm": 2.425351142883301,
      "learning_rate": 4.848571428571429e-05,
      "loss": 1.1392,
      "step": 560
    },
    {
      "epoch": 0.25333333333333335,
      "grad_norm": 2.0615339279174805,
      "learning_rate": 4.82e-05,
      "loss": 1.035,
      "step": 570
    },
    {
      "epoch": 0.2577777777777778,
      "grad_norm": 1.8026305437088013,
      "learning_rate": 4.7914285714285715e-05,
      "loss": 0.9894,
      "step": 580
    },
    {
      "epoch": 0.26222222222222225,
      "grad_norm": 1.7569513320922852,
      "learning_rate": 4.762857142857143e-05,
      "loss": 1.1085,
      "step": 590
    },
    {
      "epoch": 0.26666666666666666,
      "grad_norm": 2.374699354171753,
      "learning_rate": 4.734285714285715e-05,
      "loss": 1.0259,
      "step": 600
    },
    {
      "epoch": 0.27111111111111114,
      "grad_norm": 2.0250742435455322,
      "learning_rate": 4.7057142857142864e-05,
      "loss": 1.0345,
      "step": 610
    },
    {
      "epoch": 0.27555555555555555,
      "grad_norm": 2.330720901489258,
      "learning_rate": 4.677142857142857e-05,
      "loss": 1.1081,
      "step": 620
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.5041579008102417,
      "learning_rate": 4.648571428571429e-05,
      "loss": 0.8348,
      "step": 630
    },
    {
      "epoch": 0.28444444444444444,
      "grad_norm": 2.397007703781128,
      "learning_rate": 4.6200000000000005e-05,
      "loss": 1.2422,
      "step": 640
    },
    {
      "epoch": 0.28888888888888886,
      "grad_norm": 1.4800223112106323,
      "learning_rate": 4.5914285714285714e-05,
      "loss": 0.995,
      "step": 650
    },
    {
      "epoch": 0.29333333333333333,
      "grad_norm": 1.6215323209762573,
      "learning_rate": 4.562857142857143e-05,
      "loss": 1.0207,
      "step": 660
    },
    {
      "epoch": 0.29777777777777775,
      "grad_norm": 4.243315696716309,
      "learning_rate": 4.534285714285714e-05,
      "loss": 1.0753,
      "step": 670
    },
    {
      "epoch": 0.3022222222222222,
      "grad_norm": 1.8909763097763062,
      "learning_rate": 4.5057142857142856e-05,
      "loss": 0.9546,
      "step": 680
    },
    {
      "epoch": 0.30666666666666664,
      "grad_norm": 1.5152394771575928,
      "learning_rate": 4.477142857142858e-05,
      "loss": 1.0375,
      "step": 690
    },
    {
      "epoch": 0.3111111111111111,
      "grad_norm": 1.9466438293457031,
      "learning_rate": 4.448571428571429e-05,
      "loss": 0.942,
      "step": 700
    },
    {
      "epoch": 0.31555555555555553,
      "grad_norm": 2.293703317642212,
      "learning_rate": 4.4200000000000004e-05,
      "loss": 1.1798,
      "step": 710
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.8559818267822266,
      "learning_rate": 4.391428571428572e-05,
      "loss": 0.9743,
      "step": 720
    },
    {
      "epoch": 0.3244444444444444,
      "grad_norm": 2.2498509883880615,
      "learning_rate": 4.362857142857143e-05,
      "loss": 1.0288,
      "step": 730
    },
    {
      "epoch": 0.3288888888888889,
      "grad_norm": 2.1050989627838135,
      "learning_rate": 4.3342857142857145e-05,
      "loss": 1.0123,
      "step": 740
    },
    {
      "epoch": 0.3333333333333333,
      "grad_norm": 1.5077266693115234,
      "learning_rate": 4.3057142857142854e-05,
      "loss": 0.9108,
      "step": 750
    },
    {
      "epoch": 0.3377777777777778,
      "grad_norm": 1.7779529094696045,
      "learning_rate": 4.277142857142857e-05,
      "loss": 1.1093,
      "step": 760
    },
    {
      "epoch": 0.3422222222222222,
      "grad_norm": 1.9781780242919922,
      "learning_rate": 4.2485714285714286e-05,
      "loss": 1.0721,
      "step": 770
    },
    {
      "epoch": 0.3466666666666667,
      "grad_norm": 1.94735848903656,
      "learning_rate": 4.22e-05,
      "loss": 1.1273,
      "step": 780
    },
    {
      "epoch": 0.3511111111111111,
      "grad_norm": 1.4685845375061035,
      "learning_rate": 4.191428571428572e-05,
      "loss": 0.9747,
      "step": 790
    },
    {
      "epoch": 0.35555555555555557,
      "grad_norm": 1.094089150428772,
      "learning_rate": 4.162857142857143e-05,
      "loss": 1.0063,
      "step": 800
    },
    {
      "epoch": 0.36,
      "grad_norm": 2.0618069171905518,
      "learning_rate": 4.1342857142857144e-05,
      "loss": 0.9391,
      "step": 810
    },
    {
      "epoch": 0.36444444444444446,
      "grad_norm": 1.430254578590393,
      "learning_rate": 4.105714285714286e-05,
      "loss": 0.842,
      "step": 820
    },
    {
      "epoch": 0.3688888888888889,
      "grad_norm": 2.3631067276000977,
      "learning_rate": 4.077142857142857e-05,
      "loss": 0.9372,
      "step": 830
    },
    {
      "epoch": 0.37333333333333335,
      "grad_norm": 2.09013032913208,
      "learning_rate": 4.0485714285714285e-05,
      "loss": 0.9428,
      "step": 840
    },
    {
      "epoch": 0.37777777777777777,
      "grad_norm": 2.336822748184204,
      "learning_rate": 4.02e-05,
      "loss": 1.0985,
      "step": 850
    },
    {
      "epoch": 0.38222222222222224,
      "grad_norm": 2.335042953491211,
      "learning_rate": 3.991428571428572e-05,
      "loss": 1.2887,
      "step": 860
    },
    {
      "epoch": 0.38666666666666666,
      "grad_norm": 2.4629454612731934,
      "learning_rate": 3.9628571428571433e-05,
      "loss": 1.1101,
      "step": 870
    },
    {
      "epoch": 0.39111111111111113,
      "grad_norm": 2.509438991546631,
      "learning_rate": 3.934285714285714e-05,
      "loss": 1.0036,
      "step": 880
    },
    {
      "epoch": 0.39555555555555555,
      "grad_norm": 16.282512664794922,
      "learning_rate": 3.905714285714286e-05,
      "loss": 1.0398,
      "step": 890
    },
    {
      "epoch": 0.4,
      "grad_norm": 2.0597307682037354,
      "learning_rate": 3.8771428571428575e-05,
      "loss": 0.9606,
      "step": 900
    },
    {
      "epoch": 0.40444444444444444,
      "grad_norm": 1.9231626987457275,
      "learning_rate": 3.8485714285714284e-05,
      "loss": 0.9271,
      "step": 910
    },
    {
      "epoch": 0.4088888888888889,
      "grad_norm": 2.7093663215637207,
      "learning_rate": 3.82e-05,
      "loss": 1.0634,
      "step": 920
    },
    {
      "epoch": 0.41333333333333333,
      "grad_norm": 1.8224252462387085,
      "learning_rate": 3.7914285714285716e-05,
      "loss": 0.9241,
      "step": 930
    },
    {
      "epoch": 0.4177777777777778,
      "grad_norm": 1.1034265756607056,
      "learning_rate": 3.762857142857143e-05,
      "loss": 0.9692,
      "step": 940
    },
    {
      "epoch": 0.4222222222222222,
      "grad_norm": 1.8517080545425415,
      "learning_rate": 3.734285714285715e-05,
      "loss": 0.93,
      "step": 950
    },
    {
      "epoch": 0.4266666666666667,
      "grad_norm": 2.2137563228607178,
      "learning_rate": 3.705714285714286e-05,
      "loss": 0.9953,
      "step": 960
    },
    {
      "epoch": 0.4311111111111111,
      "grad_norm": 1.9600673913955688,
      "learning_rate": 3.6771428571428574e-05,
      "loss": 1.2618,
      "step": 970
    },
    {
      "epoch": 0.43555555555555553,
      "grad_norm": 2.1263670921325684,
      "learning_rate": 3.648571428571429e-05,
      "loss": 1.139,
      "step": 980
    },
    {
      "epoch": 0.44,
      "grad_norm": 15.621545791625977,
      "learning_rate": 3.62e-05,
      "loss": 1.1132,
      "step": 990
    },
    {
      "epoch": 0.4444444444444444,
      "grad_norm": 1.1507309675216675,
      "learning_rate": 3.5914285714285715e-05,
      "loss": 0.8175,
      "step": 1000
    },
    {
      "epoch": 0.4488888888888889,
      "grad_norm": 1.6997977495193481,
      "learning_rate": 3.562857142857143e-05,
      "loss": 0.9233,
      "step": 1010
    },
    {
      "epoch": 0.4533333333333333,
      "grad_norm": 4.499351501464844,
      "learning_rate": 3.534285714285715e-05,
      "loss": 1.2277,
      "step": 1020
    },
    {
      "epoch": 0.4577777777777778,
      "grad_norm": 2.199875593185425,
      "learning_rate": 3.505714285714286e-05,
      "loss": 1.122,
      "step": 1030
    },
    {
      "epoch": 0.4622222222222222,
      "grad_norm": 1.630294919013977,
      "learning_rate": 3.477142857142857e-05,
      "loss": 1.0278,
      "step": 1040
    },
    {
      "epoch": 0.4666666666666667,
      "grad_norm": 1.3660622835159302,
      "learning_rate": 3.448571428571429e-05,
      "loss": 0.8593,
      "step": 1050
    },
    {
      "epoch": 0.4711111111111111,
      "grad_norm": 1.1493386030197144,
      "learning_rate": 3.4200000000000005e-05,
      "loss": 1.1425,
      "step": 1060
    },
    {
      "epoch": 0.47555555555555556,
      "grad_norm": 1.2984066009521484,
      "learning_rate": 3.3914285714285714e-05,
      "loss": 1.0536,
      "step": 1070
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.9474918842315674,
      "learning_rate": 3.362857142857143e-05,
      "loss": 1.0463,
      "step": 1080
    },
    {
      "epoch": 0.48444444444444446,
      "grad_norm": 1.580550193786621,
      "learning_rate": 3.334285714285714e-05,
      "loss": 1.0603,
      "step": 1090
    },
    {
      "epoch": 0.4888888888888889,
      "grad_norm": 2.1200408935546875,
      "learning_rate": 3.305714285714286e-05,
      "loss": 1.0465,
      "step": 1100
    },
    {
      "epoch": 0.49333333333333335,
      "grad_norm": 1.9244203567504883,
      "learning_rate": 3.277142857142858e-05,
      "loss": 1.1546,
      "step": 1110
    },
    {
      "epoch": 0.49777777777777776,
      "grad_norm": 2.738420248031616,
      "learning_rate": 3.248571428571429e-05,
      "loss": 1.0352,
      "step": 1120
    },
    {
      "epoch": 0.5022222222222222,
      "grad_norm": 1.3462022542953491,
      "learning_rate": 3.2200000000000003e-05,
      "loss": 1.0471,
      "step": 1130
    },
    {
      "epoch": 0.5066666666666667,
      "grad_norm": 2.3860256671905518,
      "learning_rate": 3.191428571428571e-05,
      "loss": 1.2251,
      "step": 1140
    },
    {
      "epoch": 0.5111111111111111,
      "grad_norm": 2.3215584754943848,
      "learning_rate": 3.162857142857143e-05,
      "loss": 1.0449,
      "step": 1150
    },
    {
      "epoch": 0.5155555555555555,
      "grad_norm": 2.4864187240600586,
      "learning_rate": 3.1342857142857145e-05,
      "loss": 1.0363,
      "step": 1160
    },
    {
      "epoch": 0.52,
      "grad_norm": 2.4286997318267822,
      "learning_rate": 3.1057142857142854e-05,
      "loss": 1.0117,
      "step": 1170
    },
    {
      "epoch": 0.5244444444444445,
      "grad_norm": 1.4346647262573242,
      "learning_rate": 3.077142857142857e-05,
      "loss": 1.0378,
      "step": 1180
    },
    {
      "epoch": 0.5288888888888889,
      "grad_norm": 1.2675151824951172,
      "learning_rate": 3.048571428571429e-05,
      "loss": 1.2185,
      "step": 1190
    },
    {
      "epoch": 0.5333333333333333,
      "grad_norm": 1.820166826248169,
      "learning_rate": 3.02e-05,
      "loss": 0.9562,
      "step": 1200
    },
    {
      "epoch": 0.5377777777777778,
      "grad_norm": 2.047520875930786,
      "learning_rate": 2.9914285714285718e-05,
      "loss": 1.1071,
      "step": 1210
    },
    {
      "epoch": 0.5422222222222223,
      "grad_norm": 1.5641695261001587,
      "learning_rate": 2.9628571428571428e-05,
      "loss": 1.0808,
      "step": 1220
    },
    {
      "epoch": 0.5466666666666666,
      "grad_norm": 1.394386887550354,
      "learning_rate": 2.9342857142857144e-05,
      "loss": 1.1786,
      "step": 1230
    },
    {
      "epoch": 0.5511111111111111,
      "grad_norm": 1.5711551904678345,
      "learning_rate": 2.905714285714286e-05,
      "loss": 1.0592,
      "step": 1240
    },
    {
      "epoch": 0.5555555555555556,
      "grad_norm": 4.389777660369873,
      "learning_rate": 2.8771428571428572e-05,
      "loss": 0.9317,
      "step": 1250
    },
    {
      "epoch": 0.56,
      "grad_norm": 2.3850152492523193,
      "learning_rate": 2.848571428571429e-05,
      "loss": 1.2062,
      "step": 1260
    },
    {
      "epoch": 0.5644444444444444,
      "grad_norm": 1.9837779998779297,
      "learning_rate": 2.8199999999999998e-05,
      "loss": 1.0094,
      "step": 1270
    },
    {
      "epoch": 0.5688888888888889,
      "grad_norm": 1.7174725532531738,
      "learning_rate": 2.7914285714285714e-05,
      "loss": 0.8973,
      "step": 1280
    },
    {
      "epoch": 0.5733333333333334,
      "grad_norm": 3.0591206550598145,
      "learning_rate": 2.762857142857143e-05,
      "loss": 0.9418,
      "step": 1290
    },
    {
      "epoch": 0.5777777777777777,
      "grad_norm": 2.102701187133789,
      "learning_rate": 2.7342857142857142e-05,
      "loss": 0.8937,
      "step": 1300
    },
    {
      "epoch": 0.5822222222222222,
      "grad_norm": 1.4529622793197632,
      "learning_rate": 2.705714285714286e-05,
      "loss": 0.8687,
      "step": 1310
    },
    {
      "epoch": 0.5866666666666667,
      "grad_norm": 2.070000410079956,
      "learning_rate": 2.6771428571428575e-05,
      "loss": 0.9817,
      "step": 1320
    },
    {
      "epoch": 0.5911111111111111,
      "grad_norm": 1.9277245998382568,
      "learning_rate": 2.6485714285714287e-05,
      "loss": 1.0032,
      "step": 1330
    },
    {
      "epoch": 0.5955555555555555,
      "grad_norm": 1.843050241470337,
      "learning_rate": 2.6200000000000003e-05,
      "loss": 1.1426,
      "step": 1340
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.0302870273590088,
      "learning_rate": 2.5914285714285713e-05,
      "loss": 0.9906,
      "step": 1350
    },
    {
      "epoch": 0.6044444444444445,
      "grad_norm": 2.2761287689208984,
      "learning_rate": 2.562857142857143e-05,
      "loss": 1.1229,
      "step": 1360
    },
    {
      "epoch": 0.6088888888888889,
      "grad_norm": 3.804614782333374,
      "learning_rate": 2.5342857142857145e-05,
      "loss": 0.9621,
      "step": 1370
    },
    {
      "epoch": 0.6133333333333333,
      "grad_norm": 2.139857053756714,
      "learning_rate": 2.5057142857142857e-05,
      "loss": 0.9236,
      "step": 1380
    },
    {
      "epoch": 0.6177777777777778,
      "grad_norm": 2.6473701000213623,
      "learning_rate": 2.4771428571428573e-05,
      "loss": 0.9378,
      "step": 1390
    },
    {
      "epoch": 0.6222222222222222,
      "grad_norm": 2.084102153778076,
      "learning_rate": 2.4485714285714286e-05,
      "loss": 1.1248,
      "step": 1400
    },
    {
      "epoch": 0.6266666666666667,
      "grad_norm": 1.9559253454208374,
      "learning_rate": 2.4200000000000002e-05,
      "loss": 0.9016,
      "step": 1410
    },
    {
      "epoch": 0.6311111111111111,
      "grad_norm": 2.2711124420166016,
      "learning_rate": 2.3914285714285715e-05,
      "loss": 0.9147,
      "step": 1420
    },
    {
      "epoch": 0.6355555555555555,
      "grad_norm": 1.9618175029754639,
      "learning_rate": 2.362857142857143e-05,
      "loss": 1.0164,
      "step": 1430
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.52959144115448,
      "learning_rate": 2.3342857142857143e-05,
      "loss": 0.9068,
      "step": 1440
    },
    {
      "epoch": 0.6444444444444445,
      "grad_norm": 1.666641354560852,
      "learning_rate": 2.3057142857142856e-05,
      "loss": 0.9525,
      "step": 1450
    },
    {
      "epoch": 0.6488888888888888,
      "grad_norm": 2.2008984088897705,
      "learning_rate": 2.2771428571428572e-05,
      "loss": 1.2026,
      "step": 1460
    },
    {
      "epoch": 0.6533333333333333,
      "grad_norm": 1.7555994987487793,
      "learning_rate": 2.2485714285714288e-05,
      "loss": 1.0945,
      "step": 1470
    },
    {
      "epoch": 0.6577777777777778,
      "grad_norm": 2.01448392868042,
      "learning_rate": 2.22e-05,
      "loss": 1.0728,
      "step": 1480
    },
    {
      "epoch": 0.6622222222222223,
      "grad_norm": 3.834198474884033,
      "learning_rate": 2.1914285714285714e-05,
      "loss": 1.1607,
      "step": 1490
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 1.5447688102722168,
      "learning_rate": 2.162857142857143e-05,
      "loss": 0.9704,
      "step": 1500
    },
    {
      "epoch": 0.6711111111111111,
      "grad_norm": 1.6624338626861572,
      "learning_rate": 2.1342857142857146e-05,
      "loss": 0.8271,
      "step": 1510
    },
    {
      "epoch": 0.6755555555555556,
      "grad_norm": 1.5587396621704102,
      "learning_rate": 2.105714285714286e-05,
      "loss": 1.0568,
      "step": 1520
    },
    {
      "epoch": 0.68,
      "grad_norm": 1.042724370956421,
      "learning_rate": 2.077142857142857e-05,
      "loss": 1.1686,
      "step": 1530
    },
    {
      "epoch": 0.6844444444444444,
      "grad_norm": 1.855294942855835,
      "learning_rate": 2.0485714285714287e-05,
      "loss": 0.9368,
      "step": 1540
    },
    {
      "epoch": 0.6888888888888889,
      "grad_norm": 1.5107423067092896,
      "learning_rate": 2.0200000000000003e-05,
      "loss": 1.0277,
      "step": 1550
    },
    {
      "epoch": 0.6933333333333334,
      "grad_norm": 1.0451265573501587,
      "learning_rate": 1.9914285714285716e-05,
      "loss": 1.0994,
      "step": 1560
    },
    {
      "epoch": 0.6977777777777778,
      "grad_norm": 2.220353126525879,
      "learning_rate": 1.962857142857143e-05,
      "loss": 1.1676,
      "step": 1570
    },
    {
      "epoch": 0.7022222222222222,
      "grad_norm": 4.320748805999756,
      "learning_rate": 1.9342857142857144e-05,
      "loss": 1.0167,
      "step": 1580
    },
    {
      "epoch": 0.7066666666666667,
      "grad_norm": 1.5187314748764038,
      "learning_rate": 1.9057142857142857e-05,
      "loss": 1.0061,
      "step": 1590
    },
    {
      "epoch": 0.7111111111111111,
      "grad_norm": 2.62479305267334,
      "learning_rate": 1.8771428571428573e-05,
      "loss": 1.0231,
      "step": 1600
    },
    {
      "epoch": 0.7155555555555555,
      "grad_norm": 37.25562286376953,
      "learning_rate": 1.8485714285714286e-05,
      "loss": 1.0258,
      "step": 1610
    },
    {
      "epoch": 0.72,
      "grad_norm": 7.708355903625488,
      "learning_rate": 1.8200000000000002e-05,
      "loss": 0.8673,
      "step": 1620
    },
    {
      "epoch": 0.7244444444444444,
      "grad_norm": 7.80335807800293,
      "learning_rate": 1.7914285714285715e-05,
      "loss": 1.1722,
      "step": 1630
    },
    {
      "epoch": 0.7288888888888889,
      "grad_norm": 4.959846496582031,
      "learning_rate": 1.762857142857143e-05,
      "loss": 1.0197,
      "step": 1640
    },
    {
      "epoch": 0.7333333333333333,
      "grad_norm": 1.8894150257110596,
      "learning_rate": 1.7342857142857143e-05,
      "loss": 0.9571,
      "step": 1650
    },
    {
      "epoch": 0.7377777777777778,
      "grad_norm": 5.9880828857421875,
      "learning_rate": 1.7057142857142856e-05,
      "loss": 1.1879,
      "step": 1660
    },
    {
      "epoch": 0.7422222222222222,
      "grad_norm": 1.9994230270385742,
      "learning_rate": 1.6771428571428572e-05,
      "loss": 1.01,
      "step": 1670
    },
    {
      "epoch": 0.7466666666666667,
      "grad_norm": 1.1010164022445679,
      "learning_rate": 1.6485714285714288e-05,
      "loss": 0.8837,
      "step": 1680
    },
    {
      "epoch": 0.7511111111111111,
      "grad_norm": 4.9511399269104,
      "learning_rate": 1.62e-05,
      "loss": 0.9419,
      "step": 1690
    },
    {
      "epoch": 0.7555555555555555,
      "grad_norm": 1.8997151851654053,
      "learning_rate": 1.5914285714285713e-05,
      "loss": 0.9052,
      "step": 1700
    },
    {
      "epoch": 0.76,
      "grad_norm": 1.6055902242660522,
      "learning_rate": 1.562857142857143e-05,
      "loss": 0.9279,
      "step": 1710
    },
    {
      "epoch": 0.7644444444444445,
      "grad_norm": 1.4079903364181519,
      "learning_rate": 1.5342857142857146e-05,
      "loss": 1.0294,
      "step": 1720
    },
    {
      "epoch": 0.7688888888888888,
      "grad_norm": 1.1559503078460693,
      "learning_rate": 1.5057142857142858e-05,
      "loss": 0.9437,
      "step": 1730
    },
    {
      "epoch": 0.7733333333333333,
      "grad_norm": 2.20170259475708,
      "learning_rate": 1.4771428571428573e-05,
      "loss": 1.1114,
      "step": 1740
    },
    {
      "epoch": 0.7777777777777778,
      "grad_norm": 1.4884487390518188,
      "learning_rate": 1.4485714285714285e-05,
      "loss": 1.0683,
      "step": 1750
    },
    {
      "epoch": 0.7822222222222223,
      "grad_norm": 1.7694205045700073,
      "learning_rate": 1.42e-05,
      "loss": 0.683,
      "step": 1760
    },
    {
      "epoch": 0.7866666666666666,
      "grad_norm": 7.573609352111816,
      "learning_rate": 1.3914285714285716e-05,
      "loss": 0.9765,
      "step": 1770
    },
    {
      "epoch": 0.7911111111111111,
      "grad_norm": 1.1830403804779053,
      "learning_rate": 1.362857142857143e-05,
      "loss": 1.0293,
      "step": 1780
    },
    {
      "epoch": 0.7955555555555556,
      "grad_norm": 2.407702922821045,
      "learning_rate": 1.3342857142857143e-05,
      "loss": 1.092,
      "step": 1790
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.5534087419509888,
      "learning_rate": 1.3057142857142857e-05,
      "loss": 0.9226,
      "step": 1800
    },
    {
      "epoch": 0.8044444444444444,
      "grad_norm": 1.5992074012756348,
      "learning_rate": 1.2771428571428573e-05,
      "loss": 1.024,
      "step": 1810
    },
    {
      "epoch": 0.8088888888888889,
      "grad_norm": 4.057394027709961,
      "learning_rate": 1.2485714285714287e-05,
      "loss": 1.0755,
      "step": 1820
    },
    {
      "epoch": 0.8133333333333334,
      "grad_norm": 1.3395154476165771,
      "learning_rate": 1.22e-05,
      "loss": 0.979,
      "step": 1830
    },
    {
      "epoch": 0.8177777777777778,
      "grad_norm": 8.083459854125977,
      "learning_rate": 1.1914285714285716e-05,
      "loss": 0.9691,
      "step": 1840
    },
    {
      "epoch": 0.8222222222222222,
      "grad_norm": 1.3631497621536255,
      "learning_rate": 1.1628571428571429e-05,
      "loss": 0.9148,
      "step": 1850
    },
    {
      "epoch": 0.8266666666666667,
      "grad_norm": 1.3387725353240967,
      "learning_rate": 1.1342857142857143e-05,
      "loss": 1.058,
      "step": 1860
    },
    {
      "epoch": 0.8311111111111111,
      "grad_norm": 5.362998008728027,
      "learning_rate": 1.1057142857142858e-05,
      "loss": 1.1844,
      "step": 1870
    },
    {
      "epoch": 0.8355555555555556,
      "grad_norm": 1.5291681289672852,
      "learning_rate": 1.0771428571428572e-05,
      "loss": 1.0351,
      "step": 1880
    },
    {
      "epoch": 0.84,
      "grad_norm": 2.036616563796997,
      "learning_rate": 1.0485714285714286e-05,
      "loss": 1.1685,
      "step": 1890
    },
    {
      "epoch": 0.8444444444444444,
      "grad_norm": 10.672110557556152,
      "learning_rate": 1.02e-05,
      "loss": 1.186,
      "step": 1900
    },
    {
      "epoch": 0.8488888888888889,
      "grad_norm": 7.35708475112915,
      "learning_rate": 9.914285714285715e-06,
      "loss": 1.0015,
      "step": 1910
    },
    {
      "epoch": 0.8533333333333334,
      "grad_norm": 3.0570969581604004,
      "learning_rate": 9.628571428571428e-06,
      "loss": 1.0048,
      "step": 1920
    },
    {
      "epoch": 0.8577777777777778,
      "grad_norm": 2.548383951187134,
      "learning_rate": 9.342857142857144e-06,
      "loss": 0.9308,
      "step": 1930
    },
    {
      "epoch": 0.8622222222222222,
      "grad_norm": 2.9634547233581543,
      "learning_rate": 9.057142857142856e-06,
      "loss": 1.108,
      "step": 1940
    },
    {
      "epoch": 0.8666666666666667,
      "grad_norm": 1.768025279045105,
      "learning_rate": 8.771428571428572e-06,
      "loss": 1.1125,
      "step": 1950
    },
    {
      "epoch": 0.8711111111111111,
      "grad_norm": 1.4923690557479858,
      "learning_rate": 8.485714285714285e-06,
      "loss": 1.0878,
      "step": 1960
    },
    {
      "epoch": 0.8755555555555555,
      "grad_norm": 1.8677984476089478,
      "learning_rate": 8.200000000000001e-06,
      "loss": 0.8959,
      "step": 1970
    },
    {
      "epoch": 0.88,
      "grad_norm": 4.373391151428223,
      "learning_rate": 7.914285714285714e-06,
      "loss": 0.9732,
      "step": 1980
    },
    {
      "epoch": 0.8844444444444445,
      "grad_norm": 1.9039726257324219,
      "learning_rate": 7.628571428571429e-06,
      "loss": 0.9692,
      "step": 1990
    },
    {
      "epoch": 0.8888888888888888,
      "grad_norm": 4.483780384063721,
      "learning_rate": 7.342857142857143e-06,
      "loss": 0.9734,
      "step": 2000
    },
    {
      "epoch": 0.8933333333333333,
      "grad_norm": 2.355618476867676,
      "learning_rate": 7.057142857142858e-06,
      "loss": 0.7923,
      "step": 2010
    },
    {
      "epoch": 0.8977777777777778,
      "grad_norm": 1.768234133720398,
      "learning_rate": 6.771428571428571e-06,
      "loss": 1.0397,
      "step": 2020
    },
    {
      "epoch": 0.9022222222222223,
      "grad_norm": 1.9736918210983276,
      "learning_rate": 6.485714285714286e-06,
      "loss": 0.9889,
      "step": 2030
    },
    {
      "epoch": 0.9066666666666666,
      "grad_norm": 2.7325940132141113,
      "learning_rate": 6.2e-06,
      "loss": 0.9604,
      "step": 2040
    },
    {
      "epoch": 0.9111111111111111,
      "grad_norm": 37.42255401611328,
      "learning_rate": 5.914285714285714e-06,
      "loss": 1.06,
      "step": 2050
    },
    {
      "epoch": 0.9155555555555556,
      "grad_norm": 3.5316126346588135,
      "learning_rate": 5.628571428571429e-06,
      "loss": 1.0478,
      "step": 2060
    },
    {
      "epoch": 0.92,
      "grad_norm": 4.869263648986816,
      "learning_rate": 5.342857142857143e-06,
      "loss": 0.8253,
      "step": 2070
    },
    {
      "epoch": 0.9244444444444444,
      "grad_norm": 71.07227325439453,
      "learning_rate": 5.057142857142857e-06,
      "loss": 1.1752,
      "step": 2080
    },
    {
      "epoch": 0.9288888888888889,
      "grad_norm": 5.730470657348633,
      "learning_rate": 4.771428571428572e-06,
      "loss": 0.9383,
      "step": 2090
    },
    {
      "epoch": 0.9333333333333333,
      "grad_norm": 24.144546508789062,
      "learning_rate": 4.485714285714286e-06,
      "loss": 0.9968,
      "step": 2100
    },
    {
      "epoch": 0.9377777777777778,
      "grad_norm": 3.403139352798462,
      "learning_rate": 4.2000000000000004e-06,
      "loss": 0.8893,
      "step": 2110
    },
    {
      "epoch": 0.9422222222222222,
      "grad_norm": 3.723447322845459,
      "learning_rate": 3.914285714285715e-06,
      "loss": 1.0404,
      "step": 2120
    },
    {
      "epoch": 0.9466666666666667,
      "grad_norm": 9.551548957824707,
      "learning_rate": 3.6285714285714283e-06,
      "loss": 0.9837,
      "step": 2130
    },
    {
      "epoch": 0.9511111111111111,
      "grad_norm": 6.35592794418335,
      "learning_rate": 3.3428571428571427e-06,
      "loss": 0.8563,
      "step": 2140
    },
    {
      "epoch": 0.9555555555555556,
      "grad_norm": 3.69284987449646,
      "learning_rate": 3.0571428571428575e-06,
      "loss": 1.0082,
      "step": 2150
    },
    {
      "epoch": 0.96,
      "grad_norm": 2.6728179454803467,
      "learning_rate": 2.771428571428572e-06,
      "loss": 0.9613,
      "step": 2160
    },
    {
      "epoch": 0.9644444444444444,
      "grad_norm": 6.068182945251465,
      "learning_rate": 2.4857142857142858e-06,
      "loss": 0.9627,
      "step": 2170
    },
    {
      "epoch": 0.9688888888888889,
      "grad_norm": 28.534027099609375,
      "learning_rate": 2.2e-06,
      "loss": 0.9503,
      "step": 2180
    },
    {
      "epoch": 0.9733333333333334,
      "grad_norm": 7.36533260345459,
      "learning_rate": 1.9142857142857145e-06,
      "loss": 1.0315,
      "step": 2190
    },
    {
      "epoch": 0.9777777777777777,
      "grad_norm": 126.33111572265625,
      "learning_rate": 1.6285714285714286e-06,
      "loss": 1.015,
      "step": 2200
    },
    {
      "epoch": 0.9822222222222222,
      "grad_norm": 3.259016990661621,
      "learning_rate": 1.342857142857143e-06,
      "loss": 0.8513,
      "step": 2210
    },
    {
      "epoch": 0.9866666666666667,
      "grad_norm": 3.191985607147217,
      "learning_rate": 1.0571428571428573e-06,
      "loss": 1.0846,
      "step": 2220
    },
    {
      "epoch": 0.9911111111111112,
      "grad_norm": 3.515030860900879,
      "learning_rate": 7.714285714285715e-07,
      "loss": 0.9536,
      "step": 2230
    },
    {
      "epoch": 0.9955555555555555,
      "grad_norm": 3.0338504314422607,
      "learning_rate": 4.857142857142857e-07,
      "loss": 1.0406,
      "step": 2240
    },
    {
      "epoch": 1.0,
      "grad_norm": 6.6893110275268555,
      "learning_rate": 2.285714285714286e-07,
      "loss": 1.0769,
      "step": 2250
    }
  ],
  "logging_steps": 10,
  "max_steps": 2250,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 5000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 8.0067730341888e+16,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}