{
  "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": 172.13922119140625,
      "learning_rate": 5.000000000000001e-07,
      "loss": 11.5329,
      "step": 10
    },
    {
      "epoch": 0.008888888888888889,
      "grad_norm": 179.96237182617188,
      "learning_rate": 1.5e-06,
      "loss": 11.3904,
      "step": 20
    },
    {
      "epoch": 0.013333333333333334,
      "grad_norm": 1364.69921875,
      "learning_rate": 2.5e-06,
      "loss": 10.7917,
      "step": 30
    },
    {
      "epoch": 0.017777777777777778,
      "grad_norm": 8219.5224609375,
      "learning_rate": 3.3e-06,
      "loss": 9.1512,
      "step": 40
    },
    {
      "epoch": 0.022222222222222223,
      "grad_norm": 12965.9892578125,
      "learning_rate": 4.2999999999999995e-06,
      "loss": 7.8213,
      "step": 50
    },
    {
      "epoch": 0.02666666666666667,
      "grad_norm": 31872.74609375,
      "learning_rate": 5.3e-06,
      "loss": 6.6272,
      "step": 60
    },
    {
      "epoch": 0.03111111111111111,
      "grad_norm": 788.8919677734375,
      "learning_rate": 6.300000000000001e-06,
      "loss": 6.3266,
      "step": 70
    },
    {
      "epoch": 0.035555555555555556,
      "grad_norm": 106.24903869628906,
      "learning_rate": 7.2999999999999996e-06,
      "loss": 5.7356,
      "step": 80
    },
    {
      "epoch": 0.04,
      "grad_norm": 409.2347412109375,
      "learning_rate": 8.3e-06,
      "loss": 5.018,
      "step": 90
    },
    {
      "epoch": 0.044444444444444446,
      "grad_norm": 817.6410522460938,
      "learning_rate": 9.3e-06,
      "loss": 5.422,
      "step": 100
    },
    {
      "epoch": 0.04888888888888889,
      "grad_norm": 589.453125,
      "learning_rate": 1.03e-05,
      "loss": 3.6104,
      "step": 110
    },
    {
      "epoch": 0.05333333333333334,
      "grad_norm": 299.5021667480469,
      "learning_rate": 1.13e-05,
      "loss": 3.3433,
      "step": 120
    },
    {
      "epoch": 0.057777777777777775,
      "grad_norm": 105.66629791259766,
      "learning_rate": 1.23e-05,
      "loss": 2.1184,
      "step": 130
    },
    {
      "epoch": 0.06222222222222222,
      "grad_norm": 94.36296844482422,
      "learning_rate": 1.3300000000000001e-05,
      "loss": 1.6441,
      "step": 140
    },
    {
      "epoch": 0.06666666666666667,
      "grad_norm": 99.46455383300781,
      "learning_rate": 1.43e-05,
      "loss": 1.0691,
      "step": 150
    },
    {
      "epoch": 0.07111111111111111,
      "grad_norm": 128.905517578125,
      "learning_rate": 1.53e-05,
      "loss": 1.347,
      "step": 160
    },
    {
      "epoch": 0.07555555555555556,
      "grad_norm": 10.669831275939941,
      "learning_rate": 1.63e-05,
      "loss": 0.8163,
      "step": 170
    },
    {
      "epoch": 0.08,
      "grad_norm": 17.684833526611328,
      "learning_rate": 1.73e-05,
      "loss": 1.0315,
      "step": 180
    },
    {
      "epoch": 0.08444444444444445,
      "grad_norm": 35.8447151184082,
      "learning_rate": 1.83e-05,
      "loss": 0.6946,
      "step": 190
    },
    {
      "epoch": 0.08888888888888889,
      "grad_norm": 88.678466796875,
      "learning_rate": 1.93e-05,
      "loss": 1.0834,
      "step": 200
    },
    {
      "epoch": 0.09333333333333334,
      "grad_norm": 50.5704231262207,
      "learning_rate": 2.0300000000000002e-05,
      "loss": 0.9062,
      "step": 210
    },
    {
      "epoch": 0.09777777777777778,
      "grad_norm": 5.56904935836792,
      "learning_rate": 2.13e-05,
      "loss": 0.9961,
      "step": 220
    },
    {
      "epoch": 0.10222222222222223,
      "grad_norm": 2.848574638366699,
      "learning_rate": 2.23e-05,
      "loss": 0.7547,
      "step": 230
    },
    {
      "epoch": 0.10666666666666667,
      "grad_norm": 51.646854400634766,
      "learning_rate": 2.3300000000000004e-05,
      "loss": 0.9434,
      "step": 240
    },
    {
      "epoch": 0.1111111111111111,
      "grad_norm": 4.968932151794434,
      "learning_rate": 2.43e-05,
      "loss": 0.9282,
      "step": 250
    },
    {
      "epoch": 0.11555555555555555,
      "grad_norm": 8.864336967468262,
      "learning_rate": 2.5300000000000002e-05,
      "loss": 1.1923,
      "step": 260
    },
    {
      "epoch": 0.12,
      "grad_norm": 6.225351333618164,
      "learning_rate": 2.6300000000000002e-05,
      "loss": 0.7519,
      "step": 270
    },
    {
      "epoch": 0.12444444444444444,
      "grad_norm": 16.043298721313477,
      "learning_rate": 2.7300000000000003e-05,
      "loss": 0.8965,
      "step": 280
    },
    {
      "epoch": 0.1288888888888889,
      "grad_norm": 13.324910163879395,
      "learning_rate": 2.83e-05,
      "loss": 0.8238,
      "step": 290
    },
    {
      "epoch": 0.13333333333333333,
      "grad_norm": 3.661782741546631,
      "learning_rate": 2.93e-05,
      "loss": 0.6619,
      "step": 300
    },
    {
      "epoch": 0.13777777777777778,
      "grad_norm": 24.01588249206543,
      "learning_rate": 3.03e-05,
      "loss": 0.9874,
      "step": 310
    },
    {
      "epoch": 0.14222222222222222,
      "grad_norm": 3.4116015434265137,
      "learning_rate": 3.13e-05,
      "loss": 0.7866,
      "step": 320
    },
    {
      "epoch": 0.14666666666666667,
      "grad_norm": 5.378733158111572,
      "learning_rate": 3.2300000000000006e-05,
      "loss": 0.784,
      "step": 330
    },
    {
      "epoch": 0.1511111111111111,
      "grad_norm": 3.0666747093200684,
      "learning_rate": 3.33e-05,
      "loss": 0.8068,
      "step": 340
    },
    {
      "epoch": 0.15555555555555556,
      "grad_norm": 5.233002662658691,
      "learning_rate": 3.430000000000001e-05,
      "loss": 0.8338,
      "step": 350
    },
    {
      "epoch": 0.16,
      "grad_norm": 12.038925170898438,
      "learning_rate": 3.53e-05,
      "loss": 0.9978,
      "step": 360
    },
    {
      "epoch": 0.16444444444444445,
      "grad_norm": 4.42678165435791,
      "learning_rate": 3.63e-05,
      "loss": 0.9186,
      "step": 370
    },
    {
      "epoch": 0.1688888888888889,
      "grad_norm": 3.2533979415893555,
      "learning_rate": 3.73e-05,
      "loss": 0.853,
      "step": 380
    },
    {
      "epoch": 0.17333333333333334,
      "grad_norm": 6.089888095855713,
      "learning_rate": 3.83e-05,
      "loss": 0.6885,
      "step": 390
    },
    {
      "epoch": 0.17777777777777778,
      "grad_norm": 2.185434341430664,
      "learning_rate": 3.9300000000000007e-05,
      "loss": 0.7051,
      "step": 400
    },
    {
      "epoch": 0.18222222222222223,
      "grad_norm": 3.463395118713379,
      "learning_rate": 4.0300000000000004e-05,
      "loss": 0.7594,
      "step": 410
    },
    {
      "epoch": 0.18666666666666668,
      "grad_norm": 111.5975341796875,
      "learning_rate": 4.13e-05,
      "loss": 0.7679,
      "step": 420
    },
    {
      "epoch": 0.19111111111111112,
      "grad_norm": 2.420258045196533,
      "learning_rate": 4.23e-05,
      "loss": 0.9897,
      "step": 430
    },
    {
      "epoch": 0.19555555555555557,
      "grad_norm": 3.507974863052368,
      "learning_rate": 4.33e-05,
      "loss": 0.7764,
      "step": 440
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.6020970344543457,
      "learning_rate": 4.43e-05,
      "loss": 0.6163,
      "step": 450
    },
    {
      "epoch": 0.20444444444444446,
      "grad_norm": 1.9605637788772583,
      "learning_rate": 4.53e-05,
      "loss": 0.9071,
      "step": 460
    },
    {
      "epoch": 0.2088888888888889,
      "grad_norm": 6.756528377532959,
      "learning_rate": 4.630000000000001e-05,
      "loss": 0.939,
      "step": 470
    },
    {
      "epoch": 0.21333333333333335,
      "grad_norm": 6.34450101852417,
      "learning_rate": 4.73e-05,
      "loss": 0.832,
      "step": 480
    },
    {
      "epoch": 0.21777777777777776,
      "grad_norm": 3.8294358253479004,
      "learning_rate": 4.83e-05,
      "loss": 0.9205,
      "step": 490
    },
    {
      "epoch": 0.2222222222222222,
      "grad_norm": 2.1834263801574707,
      "learning_rate": 4.93e-05,
      "loss": 0.6335,
      "step": 500
    },
    {
      "epoch": 0.22666666666666666,
      "grad_norm": 2.837082624435425,
      "learning_rate": 4.9914285714285717e-05,
      "loss": 0.6933,
      "step": 510
    },
    {
      "epoch": 0.2311111111111111,
      "grad_norm": 1.8396414518356323,
      "learning_rate": 4.962857142857143e-05,
      "loss": 1.0322,
      "step": 520
    },
    {
      "epoch": 0.23555555555555555,
      "grad_norm": 4.330628871917725,
      "learning_rate": 4.934285714285715e-05,
      "loss": 0.7124,
      "step": 530
    },
    {
      "epoch": 0.24,
      "grad_norm": 2.490299701690674,
      "learning_rate": 4.905714285714286e-05,
      "loss": 0.7426,
      "step": 540
    },
    {
      "epoch": 0.24444444444444444,
      "grad_norm": 1.341838002204895,
      "learning_rate": 4.8771428571428574e-05,
      "loss": 0.7588,
      "step": 550
    },
    {
      "epoch": 0.24888888888888888,
      "grad_norm": 2.234884262084961,
      "learning_rate": 4.848571428571429e-05,
      "loss": 0.8314,
      "step": 560
    },
    {
      "epoch": 0.25333333333333335,
      "grad_norm": 2.8187923431396484,
      "learning_rate": 4.82e-05,
      "loss": 0.8109,
      "step": 570
    },
    {
      "epoch": 0.2577777777777778,
      "grad_norm": 0.9691933393478394,
      "learning_rate": 4.7914285714285715e-05,
      "loss": 0.6785,
      "step": 580
    },
    {
      "epoch": 0.26222222222222225,
      "grad_norm": 1.8897961378097534,
      "learning_rate": 4.762857142857143e-05,
      "loss": 0.9523,
      "step": 590
    },
    {
      "epoch": 0.26666666666666666,
      "grad_norm": 7.316945552825928,
      "learning_rate": 4.734285714285715e-05,
      "loss": 0.5996,
      "step": 600
    },
    {
      "epoch": 0.27111111111111114,
      "grad_norm": 2.491607427597046,
      "learning_rate": 4.7057142857142864e-05,
      "loss": 0.5426,
      "step": 610
    },
    {
      "epoch": 0.27555555555555555,
      "grad_norm": 2.7054076194763184,
      "learning_rate": 4.677142857142857e-05,
      "loss": 0.6294,
      "step": 620
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.5034236907958984,
      "learning_rate": 4.648571428571429e-05,
      "loss": 0.4571,
      "step": 630
    },
    {
      "epoch": 0.28444444444444444,
      "grad_norm": 5.013509750366211,
      "learning_rate": 4.6200000000000005e-05,
      "loss": 0.9696,
      "step": 640
    },
    {
      "epoch": 0.28888888888888886,
      "grad_norm": 1.4784796237945557,
      "learning_rate": 4.5914285714285714e-05,
      "loss": 0.79,
      "step": 650
    },
    {
      "epoch": 0.29333333333333333,
      "grad_norm": 1.679485559463501,
      "learning_rate": 4.562857142857143e-05,
      "loss": 0.5445,
      "step": 660
    },
    {
      "epoch": 0.29777777777777775,
      "grad_norm": 11.93566608428955,
      "learning_rate": 4.534285714285714e-05,
      "loss": 0.8077,
      "step": 670
    },
    {
      "epoch": 0.3022222222222222,
      "grad_norm": 2.100651264190674,
      "learning_rate": 4.5057142857142856e-05,
      "loss": 0.5643,
      "step": 680
    },
    {
      "epoch": 0.30666666666666664,
      "grad_norm": 2.368565797805786,
      "learning_rate": 4.477142857142858e-05,
      "loss": 1.0012,
      "step": 690
    },
    {
      "epoch": 0.3111111111111111,
      "grad_norm": 2.4255175590515137,
      "learning_rate": 4.448571428571429e-05,
      "loss": 0.6344,
      "step": 700
    },
    {
      "epoch": 0.31555555555555553,
      "grad_norm": 3.4809205532073975,
      "learning_rate": 4.4200000000000004e-05,
      "loss": 0.87,
      "step": 710
    },
    {
      "epoch": 0.32,
      "grad_norm": 12.76645565032959,
      "learning_rate": 4.391428571428572e-05,
      "loss": 0.673,
      "step": 720
    },
    {
      "epoch": 0.3244444444444444,
      "grad_norm": 3.417945623397827,
      "learning_rate": 4.362857142857143e-05,
      "loss": 0.6753,
      "step": 730
    },
    {
      "epoch": 0.3288888888888889,
      "grad_norm": 2.9550516605377197,
      "learning_rate": 4.3342857142857145e-05,
      "loss": 0.6621,
      "step": 740
    },
    {
      "epoch": 0.3333333333333333,
      "grad_norm": 1.9887003898620605,
      "learning_rate": 4.3057142857142854e-05,
      "loss": 0.671,
      "step": 750
    },
    {
      "epoch": 0.3377777777777778,
      "grad_norm": 1.826278567314148,
      "learning_rate": 4.277142857142857e-05,
      "loss": 0.7536,
      "step": 760
    },
    {
      "epoch": 0.3422222222222222,
      "grad_norm": 4.41416597366333,
      "learning_rate": 4.2485714285714286e-05,
      "loss": 0.7106,
      "step": 770
    },
    {
      "epoch": 0.3466666666666667,
      "grad_norm": 1.8409727811813354,
      "learning_rate": 4.22e-05,
      "loss": 0.7142,
      "step": 780
    },
    {
      "epoch": 0.3511111111111111,
      "grad_norm": 2.143974781036377,
      "learning_rate": 4.191428571428572e-05,
      "loss": 0.5623,
      "step": 790
    },
    {
      "epoch": 0.35555555555555557,
      "grad_norm": 1.1921825408935547,
      "learning_rate": 4.162857142857143e-05,
      "loss": 0.6638,
      "step": 800
    },
    {
      "epoch": 0.36,
      "grad_norm": 2.4289398193359375,
      "learning_rate": 4.1342857142857144e-05,
      "loss": 0.503,
      "step": 810
    },
    {
      "epoch": 0.36444444444444446,
      "grad_norm": 1.718072533607483,
      "learning_rate": 4.105714285714286e-05,
      "loss": 0.4197,
      "step": 820
    },
    {
      "epoch": 0.3688888888888889,
      "grad_norm": 2.278563976287842,
      "learning_rate": 4.077142857142857e-05,
      "loss": 0.611,
      "step": 830
    },
    {
      "epoch": 0.37333333333333335,
      "grad_norm": 2.060332775115967,
      "learning_rate": 4.0485714285714285e-05,
      "loss": 0.5883,
      "step": 840
    },
    {
      "epoch": 0.37777777777777777,
      "grad_norm": 2.5066239833831787,
      "learning_rate": 4.02e-05,
      "loss": 0.6179,
      "step": 850
    },
    {
      "epoch": 0.38222222222222224,
      "grad_norm": 5.195372104644775,
      "learning_rate": 3.991428571428572e-05,
      "loss": 0.7709,
      "step": 860
    },
    {
      "epoch": 0.38666666666666666,
      "grad_norm": 3.0029399394989014,
      "learning_rate": 3.9628571428571433e-05,
      "loss": 0.584,
      "step": 870
    },
    {
      "epoch": 0.39111111111111113,
      "grad_norm": 6.307543754577637,
      "learning_rate": 3.934285714285714e-05,
      "loss": 0.6633,
      "step": 880
    },
    {
      "epoch": 0.39555555555555555,
      "grad_norm": 6.635744571685791,
      "learning_rate": 3.905714285714286e-05,
      "loss": 0.85,
      "step": 890
    },
    {
      "epoch": 0.4,
      "grad_norm": 4.6865434646606445,
      "learning_rate": 3.8771428571428575e-05,
      "loss": 0.6508,
      "step": 900
    },
    {
      "epoch": 0.40444444444444444,
      "grad_norm": 1.8700509071350098,
      "learning_rate": 3.8485714285714284e-05,
      "loss": 0.5405,
      "step": 910
    },
    {
      "epoch": 0.4088888888888889,
      "grad_norm": 3.749453067779541,
      "learning_rate": 3.82e-05,
      "loss": 0.6869,
      "step": 920
    },
    {
      "epoch": 0.41333333333333333,
      "grad_norm": 2.443854570388794,
      "learning_rate": 3.7914285714285716e-05,
      "loss": 0.5013,
      "step": 930
    },
    {
      "epoch": 0.4177777777777778,
      "grad_norm": 1.066603183746338,
      "learning_rate": 3.762857142857143e-05,
      "loss": 0.5993,
      "step": 940
    },
    {
      "epoch": 0.4222222222222222,
      "grad_norm": 5.453496932983398,
      "learning_rate": 3.734285714285715e-05,
      "loss": 0.633,
      "step": 950
    },
    {
      "epoch": 0.4266666666666667,
      "grad_norm": 3.3490309715270996,
      "learning_rate": 3.705714285714286e-05,
      "loss": 0.7452,
      "step": 960
    },
    {
      "epoch": 0.4311111111111111,
      "grad_norm": 2.070544719696045,
      "learning_rate": 3.6771428571428574e-05,
      "loss": 0.7698,
      "step": 970
    },
    {
      "epoch": 0.43555555555555553,
      "grad_norm": 3.1262307167053223,
      "learning_rate": 3.648571428571429e-05,
      "loss": 0.7202,
      "step": 980
    },
    {
      "epoch": 0.44,
      "grad_norm": 4.966836929321289,
      "learning_rate": 3.62e-05,
      "loss": 0.6881,
      "step": 990
    },
    {
      "epoch": 0.4444444444444444,
      "grad_norm": 0.9398611783981323,
      "learning_rate": 3.5914285714285715e-05,
      "loss": 0.3937,
      "step": 1000
    },
    {
      "epoch": 0.4488888888888889,
      "grad_norm": 2.2783327102661133,
      "learning_rate": 3.562857142857143e-05,
      "loss": 0.5421,
      "step": 1010
    },
    {
      "epoch": 0.4533333333333333,
      "grad_norm": 1.5350534915924072,
      "learning_rate": 3.534285714285715e-05,
      "loss": 1.1952,
      "step": 1020
    },
    {
      "epoch": 0.4577777777777778,
      "grad_norm": 2.5785741806030273,
      "learning_rate": 3.505714285714286e-05,
      "loss": 0.7421,
      "step": 1030
    },
    {
      "epoch": 0.4622222222222222,
      "grad_norm": 3.4625842571258545,
      "learning_rate": 3.477142857142857e-05,
      "loss": 0.6769,
      "step": 1040
    },
    {
      "epoch": 0.4666666666666667,
      "grad_norm": 1.2348237037658691,
      "learning_rate": 3.448571428571429e-05,
      "loss": 0.4613,
      "step": 1050
    },
    {
      "epoch": 0.4711111111111111,
      "grad_norm": 1.4216806888580322,
      "learning_rate": 3.4200000000000005e-05,
      "loss": 0.6116,
      "step": 1060
    },
    {
      "epoch": 0.47555555555555556,
      "grad_norm": 1.089032769203186,
      "learning_rate": 3.3914285714285714e-05,
      "loss": 0.5686,
      "step": 1070
    },
    {
      "epoch": 0.48,
      "grad_norm": 2.135321855545044,
      "learning_rate": 3.362857142857143e-05,
      "loss": 0.5668,
      "step": 1080
    },
    {
      "epoch": 0.48444444444444446,
      "grad_norm": 1.846582055091858,
      "learning_rate": 3.334285714285714e-05,
      "loss": 0.62,
      "step": 1090
    },
    {
      "epoch": 0.4888888888888889,
      "grad_norm": 2.3308606147766113,
      "learning_rate": 3.305714285714286e-05,
      "loss": 0.7323,
      "step": 1100
    },
    {
      "epoch": 0.49333333333333335,
      "grad_norm": 10.256339073181152,
      "learning_rate": 3.277142857142858e-05,
      "loss": 0.6122,
      "step": 1110
    },
    {
      "epoch": 0.49777777777777776,
      "grad_norm": 5.60221529006958,
      "learning_rate": 3.248571428571429e-05,
      "loss": 0.7023,
      "step": 1120
    },
    {
      "epoch": 0.5022222222222222,
      "grad_norm": 1.3172476291656494,
      "learning_rate": 3.2200000000000003e-05,
      "loss": 0.6019,
      "step": 1130
    },
    {
      "epoch": 0.5066666666666667,
      "grad_norm": 2.984369993209839,
      "learning_rate": 3.191428571428571e-05,
      "loss": 0.8876,
      "step": 1140
    },
    {
      "epoch": 0.5111111111111111,
      "grad_norm": 2.645580291748047,
      "learning_rate": 3.162857142857143e-05,
      "loss": 0.6198,
      "step": 1150
    },
    {
      "epoch": 0.5155555555555555,
      "grad_norm": 6.525484085083008,
      "learning_rate": 3.1342857142857145e-05,
      "loss": 0.7378,
      "step": 1160
    },
    {
      "epoch": 0.52,
      "grad_norm": 2.520108938217163,
      "learning_rate": 3.1057142857142854e-05,
      "loss": 0.6144,
      "step": 1170
    },
    {
      "epoch": 0.5244444444444445,
      "grad_norm": 1.4671144485473633,
      "learning_rate": 3.077142857142857e-05,
      "loss": 0.5624,
      "step": 1180
    },
    {
      "epoch": 0.5288888888888889,
      "grad_norm": 1.6521360874176025,
      "learning_rate": 3.048571428571429e-05,
      "loss": 0.8754,
      "step": 1190
    },
    {
      "epoch": 0.5333333333333333,
      "grad_norm": 1.7824636697769165,
      "learning_rate": 3.02e-05,
      "loss": 0.6161,
      "step": 1200
    },
    {
      "epoch": 0.5377777777777778,
      "grad_norm": 3.1892054080963135,
      "learning_rate": 2.9914285714285718e-05,
      "loss": 0.6746,
      "step": 1210
    },
    {
      "epoch": 0.5422222222222223,
      "grad_norm": 2.471684217453003,
      "learning_rate": 2.9628571428571428e-05,
      "loss": 0.7047,
      "step": 1220
    },
    {
      "epoch": 0.5466666666666666,
      "grad_norm": 1.1938486099243164,
      "learning_rate": 2.9342857142857144e-05,
      "loss": 0.7613,
      "step": 1230
    },
    {
      "epoch": 0.5511111111111111,
      "grad_norm": 4.988212585449219,
      "learning_rate": 2.905714285714286e-05,
      "loss": 0.7503,
      "step": 1240
    },
    {
      "epoch": 0.5555555555555556,
      "grad_norm": 6.780579566955566,
      "learning_rate": 2.8771428571428572e-05,
      "loss": 0.5375,
      "step": 1250
    },
    {
      "epoch": 0.56,
      "grad_norm": 2.5735678672790527,
      "learning_rate": 2.848571428571429e-05,
      "loss": 0.703,
      "step": 1260
    },
    {
      "epoch": 0.5644444444444444,
      "grad_norm": 3.509777069091797,
      "learning_rate": 2.8199999999999998e-05,
      "loss": 0.5273,
      "step": 1270
    },
    {
      "epoch": 0.5688888888888889,
      "grad_norm": 1.5865542888641357,
      "learning_rate": 2.7914285714285714e-05,
      "loss": 0.4821,
      "step": 1280
    },
    {
      "epoch": 0.5733333333333334,
      "grad_norm": 1.2859346866607666,
      "learning_rate": 2.762857142857143e-05,
      "loss": 0.6347,
      "step": 1290
    },
    {
      "epoch": 0.5777777777777777,
      "grad_norm": 2.0483293533325195,
      "learning_rate": 2.7342857142857142e-05,
      "loss": 0.4798,
      "step": 1300
    },
    {
      "epoch": 0.5822222222222222,
      "grad_norm": 1.2751628160476685,
      "learning_rate": 2.705714285714286e-05,
      "loss": 0.4283,
      "step": 1310
    },
    {
      "epoch": 0.5866666666666667,
      "grad_norm": 1.7461509704589844,
      "learning_rate": 2.6771428571428575e-05,
      "loss": 0.577,
      "step": 1320
    },
    {
      "epoch": 0.5911111111111111,
      "grad_norm": 2.4847190380096436,
      "learning_rate": 2.6485714285714287e-05,
      "loss": 0.6608,
      "step": 1330
    },
    {
      "epoch": 0.5955555555555555,
      "grad_norm": 7.046814441680908,
      "learning_rate": 2.6200000000000003e-05,
      "loss": 0.7501,
      "step": 1340
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.8862992525100708,
      "learning_rate": 2.5914285714285713e-05,
      "loss": 0.5347,
      "step": 1350
    },
    {
      "epoch": 0.6044444444444445,
      "grad_norm": 1.941525936126709,
      "learning_rate": 2.562857142857143e-05,
      "loss": 0.6533,
      "step": 1360
    },
    {
      "epoch": 0.6088888888888889,
      "grad_norm": 1.9246207475662231,
      "learning_rate": 2.5342857142857145e-05,
      "loss": 0.4619,
      "step": 1370
    },
    {
      "epoch": 0.6133333333333333,
      "grad_norm": 2.3284366130828857,
      "learning_rate": 2.5057142857142857e-05,
      "loss": 0.5442,
      "step": 1380
    },
    {
      "epoch": 0.6177777777777778,
      "grad_norm": 3.2495412826538086,
      "learning_rate": 2.4771428571428573e-05,
      "loss": 0.5538,
      "step": 1390
    },
    {
      "epoch": 0.6222222222222222,
      "grad_norm": 1.9646495580673218,
      "learning_rate": 2.4485714285714286e-05,
      "loss": 0.6764,
      "step": 1400
    },
    {
      "epoch": 0.6266666666666667,
      "grad_norm": 2.3741135597229004,
      "learning_rate": 2.4200000000000002e-05,
      "loss": 0.4923,
      "step": 1410
    },
    {
      "epoch": 0.6311111111111111,
      "grad_norm": 3.175323486328125,
      "learning_rate": 2.3914285714285715e-05,
      "loss": 0.5318,
      "step": 1420
    },
    {
      "epoch": 0.6355555555555555,
      "grad_norm": 4.018131732940674,
      "learning_rate": 2.362857142857143e-05,
      "loss": 0.6042,
      "step": 1430
    },
    {
      "epoch": 0.64,
      "grad_norm": 5.065487384796143,
      "learning_rate": 2.3342857142857143e-05,
      "loss": 0.4944,
      "step": 1440
    },
    {
      "epoch": 0.6444444444444445,
      "grad_norm": 4.733597278594971,
      "learning_rate": 2.3057142857142856e-05,
      "loss": 0.5209,
      "step": 1450
    },
    {
      "epoch": 0.6488888888888888,
      "grad_norm": 2.8453125953674316,
      "learning_rate": 2.2771428571428572e-05,
      "loss": 0.6981,
      "step": 1460
    },
    {
      "epoch": 0.6533333333333333,
      "grad_norm": 5.391908168792725,
      "learning_rate": 2.2485714285714288e-05,
      "loss": 0.6173,
      "step": 1470
    },
    {
      "epoch": 0.6577777777777778,
      "grad_norm": 3.4642038345336914,
      "learning_rate": 2.22e-05,
      "loss": 0.5644,
      "step": 1480
    },
    {
      "epoch": 0.6622222222222223,
      "grad_norm": 5.350740909576416,
      "learning_rate": 2.1914285714285714e-05,
      "loss": 0.6647,
      "step": 1490
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 1.7532992362976074,
      "learning_rate": 2.162857142857143e-05,
      "loss": 0.5771,
      "step": 1500
    },
    {
      "epoch": 0.6711111111111111,
      "grad_norm": 1.6923083066940308,
      "learning_rate": 2.1342857142857146e-05,
      "loss": 0.4541,
      "step": 1510
    },
    {
      "epoch": 0.6755555555555556,
      "grad_norm": 3.4142796993255615,
      "learning_rate": 2.105714285714286e-05,
      "loss": 0.6353,
      "step": 1520
    },
    {
      "epoch": 0.68,
      "grad_norm": 1.2692922353744507,
      "learning_rate": 2.077142857142857e-05,
      "loss": 0.6471,
      "step": 1530
    },
    {
      "epoch": 0.6844444444444444,
      "grad_norm": 6.867006778717041,
      "learning_rate": 2.0485714285714287e-05,
      "loss": 0.5668,
      "step": 1540
    },
    {
      "epoch": 0.6888888888888889,
      "grad_norm": 1.1940428018569946,
      "learning_rate": 2.0200000000000003e-05,
      "loss": 0.7113,
      "step": 1550
    },
    {
      "epoch": 0.6933333333333334,
      "grad_norm": 0.9745553731918335,
      "learning_rate": 1.9914285714285716e-05,
      "loss": 0.6865,
      "step": 1560
    },
    {
      "epoch": 0.6977777777777778,
      "grad_norm": 1.4725788831710815,
      "learning_rate": 1.962857142857143e-05,
      "loss": 0.7809,
      "step": 1570
    },
    {
      "epoch": 0.7022222222222222,
      "grad_norm": 2.4500861167907715,
      "learning_rate": 1.9342857142857144e-05,
      "loss": 0.5468,
      "step": 1580
    },
    {
      "epoch": 0.7066666666666667,
      "grad_norm": 13.177860260009766,
      "learning_rate": 1.9057142857142857e-05,
      "loss": 0.5427,
      "step": 1590
    },
    {
      "epoch": 0.7111111111111111,
      "grad_norm": 5.718009948730469,
      "learning_rate": 1.8771428571428573e-05,
      "loss": 0.6003,
      "step": 1600
    },
    {
      "epoch": 0.7155555555555555,
      "grad_norm": 2.4186694622039795,
      "learning_rate": 1.8485714285714286e-05,
      "loss": 0.4843,
      "step": 1610
    },
    {
      "epoch": 0.72,
      "grad_norm": 1.150259256362915,
      "learning_rate": 1.8200000000000002e-05,
      "loss": 0.4339,
      "step": 1620
    },
    {
      "epoch": 0.7244444444444444,
      "grad_norm": 2.5581448078155518,
      "learning_rate": 1.7914285714285715e-05,
      "loss": 0.7765,
      "step": 1630
    },
    {
      "epoch": 0.7288888888888889,
      "grad_norm": 2.4859399795532227,
      "learning_rate": 1.762857142857143e-05,
      "loss": 0.5869,
      "step": 1640
    },
    {
      "epoch": 0.7333333333333333,
      "grad_norm": 1.6887503862380981,
      "learning_rate": 1.7342857142857143e-05,
      "loss": 0.4651,
      "step": 1650
    },
    {
      "epoch": 0.7377777777777778,
      "grad_norm": 2.0939149856567383,
      "learning_rate": 1.7057142857142856e-05,
      "loss": 0.7603,
      "step": 1660
    },
    {
      "epoch": 0.7422222222222222,
      "grad_norm": 1.9502744674682617,
      "learning_rate": 1.6771428571428572e-05,
      "loss": 0.5475,
      "step": 1670
    },
    {
      "epoch": 0.7466666666666667,
      "grad_norm": 0.8679394721984863,
      "learning_rate": 1.6485714285714288e-05,
      "loss": 0.4752,
      "step": 1680
    },
    {
      "epoch": 0.7511111111111111,
      "grad_norm": 1.2583119869232178,
      "learning_rate": 1.62e-05,
      "loss": 0.5677,
      "step": 1690
    },
    {
      "epoch": 0.7555555555555555,
      "grad_norm": 2.252943754196167,
      "learning_rate": 1.5914285714285713e-05,
      "loss": 0.4604,
      "step": 1700
    },
    {
      "epoch": 0.76,
      "grad_norm": 1.150745153427124,
      "learning_rate": 1.562857142857143e-05,
      "loss": 0.5713,
      "step": 1710
    },
    {
      "epoch": 0.7644444444444445,
      "grad_norm": 1.7483420372009277,
      "learning_rate": 1.5342857142857146e-05,
      "loss": 0.5989,
      "step": 1720
    },
    {
      "epoch": 0.7688888888888888,
      "grad_norm": 0.8522917628288269,
      "learning_rate": 1.5057142857142858e-05,
      "loss": 0.5211,
      "step": 1730
    },
    {
      "epoch": 0.7733333333333333,
      "grad_norm": 2.067723035812378,
      "learning_rate": 1.4771428571428573e-05,
      "loss": 0.6904,
      "step": 1740
    },
    {
      "epoch": 0.7777777777777778,
      "grad_norm": 1.3494257926940918,
      "learning_rate": 1.4485714285714285e-05,
      "loss": 0.6667,
      "step": 1750
    },
    {
      "epoch": 0.7822222222222223,
      "grad_norm": 1.3629013299942017,
      "learning_rate": 1.42e-05,
      "loss": 0.3367,
      "step": 1760
    },
    {
      "epoch": 0.7866666666666666,
      "grad_norm": 3.2202422618865967,
      "learning_rate": 1.3914285714285716e-05,
      "loss": 0.6033,
      "step": 1770
    },
    {
      "epoch": 0.7911111111111111,
      "grad_norm": 1.038692593574524,
      "learning_rate": 1.362857142857143e-05,
      "loss": 0.5305,
      "step": 1780
    },
    {
      "epoch": 0.7955555555555556,
      "grad_norm": 2.57888126373291,
      "learning_rate": 1.3342857142857143e-05,
      "loss": 0.5989,
      "step": 1790
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.118715524673462,
      "learning_rate": 1.3057142857142857e-05,
      "loss": 0.4963,
      "step": 1800
    },
    {
      "epoch": 0.8044444444444444,
      "grad_norm": 2.586069107055664,
      "learning_rate": 1.2771428571428573e-05,
      "loss": 0.6829,
      "step": 1810
    },
    {
      "epoch": 0.8088888888888889,
      "grad_norm": 2.27518630027771,
      "learning_rate": 1.2485714285714287e-05,
      "loss": 0.7121,
      "step": 1820
    },
    {
      "epoch": 0.8133333333333334,
      "grad_norm": 1.1266392469406128,
      "learning_rate": 1.22e-05,
      "loss": 0.5293,
      "step": 1830
    },
    {
      "epoch": 0.8177777777777778,
      "grad_norm": 1.4908055067062378,
      "learning_rate": 1.1914285714285716e-05,
      "loss": 0.5562,
      "step": 1840
    },
    {
      "epoch": 0.8222222222222222,
      "grad_norm": 1.4283602237701416,
      "learning_rate": 1.1628571428571429e-05,
      "loss": 0.4831,
      "step": 1850
    },
    {
      "epoch": 0.8266666666666667,
      "grad_norm": 1.0552829504013062,
      "learning_rate": 1.1342857142857143e-05,
      "loss": 0.5761,
      "step": 1860
    },
    {
      "epoch": 0.8311111111111111,
      "grad_norm": 6.45361852645874,
      "learning_rate": 1.1057142857142858e-05,
      "loss": 0.6874,
      "step": 1870
    },
    {
      "epoch": 0.8355555555555556,
      "grad_norm": 1.207909345626831,
      "learning_rate": 1.0771428571428572e-05,
      "loss": 0.6289,
      "step": 1880
    },
    {
      "epoch": 0.84,
      "grad_norm": 3.6151106357574463,
      "learning_rate": 1.0485714285714286e-05,
      "loss": 0.78,
      "step": 1890
    },
    {
      "epoch": 0.8444444444444444,
      "grad_norm": 2.560347318649292,
      "learning_rate": 1.02e-05,
      "loss": 0.7657,
      "step": 1900
    },
    {
      "epoch": 0.8488888888888889,
      "grad_norm": 1.1826066970825195,
      "learning_rate": 9.914285714285715e-06,
      "loss": 0.5419,
      "step": 1910
    },
    {
      "epoch": 0.8533333333333334,
      "grad_norm": 2.4970545768737793,
      "learning_rate": 9.628571428571428e-06,
      "loss": 0.5137,
      "step": 1920
    },
    {
      "epoch": 0.8577777777777778,
      "grad_norm": 1.6849817037582397,
      "learning_rate": 9.342857142857144e-06,
      "loss": 0.4476,
      "step": 1930
    },
    {
      "epoch": 0.8622222222222222,
      "grad_norm": 3.5557351112365723,
      "learning_rate": 9.057142857142856e-06,
      "loss": 0.6422,
      "step": 1940
    },
    {
      "epoch": 0.8666666666666667,
      "grad_norm": 1.7327756881713867,
      "learning_rate": 8.771428571428572e-06,
      "loss": 0.656,
      "step": 1950
    },
    {
      "epoch": 0.8711111111111111,
      "grad_norm": 1.3222990036010742,
      "learning_rate": 8.485714285714285e-06,
      "loss": 0.6862,
      "step": 1960
    },
    {
      "epoch": 0.8755555555555555,
      "grad_norm": 1.3732144832611084,
      "learning_rate": 8.200000000000001e-06,
      "loss": 0.5173,
      "step": 1970
    },
    {
      "epoch": 0.88,
      "grad_norm": 5.491960048675537,
      "learning_rate": 7.914285714285714e-06,
      "loss": 0.5735,
      "step": 1980
    },
    {
      "epoch": 0.8844444444444445,
      "grad_norm": 1.9830958843231201,
      "learning_rate": 7.628571428571429e-06,
      "loss": 0.4936,
      "step": 1990
    },
    {
      "epoch": 0.8888888888888888,
      "grad_norm": 3.4833927154541016,
      "learning_rate": 7.342857142857143e-06,
      "loss": 0.5099,
      "step": 2000
    },
    {
      "epoch": 0.8933333333333333,
      "grad_norm": 1.197811484336853,
      "learning_rate": 7.057142857142858e-06,
      "loss": 0.4711,
      "step": 2010
    },
    {
      "epoch": 0.8977777777777778,
      "grad_norm": 2.15141224861145,
      "learning_rate": 6.771428571428571e-06,
      "loss": 0.5555,
      "step": 2020
    },
    {
      "epoch": 0.9022222222222223,
      "grad_norm": 1.7254974842071533,
      "learning_rate": 6.485714285714286e-06,
      "loss": 0.5582,
      "step": 2030
    },
    {
      "epoch": 0.9066666666666666,
      "grad_norm": 9.579049110412598,
      "learning_rate": 6.2e-06,
      "loss": 0.4571,
      "step": 2040
    },
    {
      "epoch": 0.9111111111111111,
      "grad_norm": 2.882373094558716,
      "learning_rate": 5.914285714285714e-06,
      "loss": 0.6287,
      "step": 2050
    },
    {
      "epoch": 0.9155555555555556,
      "grad_norm": 5.913229942321777,
      "learning_rate": 5.628571428571429e-06,
      "loss": 0.5779,
      "step": 2060
    },
    {
      "epoch": 0.92,
      "grad_norm": 4.352343559265137,
      "learning_rate": 5.342857142857143e-06,
      "loss": 0.5565,
      "step": 2070
    },
    {
      "epoch": 0.9244444444444444,
      "grad_norm": 2.8510098457336426,
      "learning_rate": 5.057142857142857e-06,
      "loss": 0.5999,
      "step": 2080
    },
    {
      "epoch": 0.9288888888888889,
      "grad_norm": 2.46824049949646,
      "learning_rate": 4.771428571428572e-06,
      "loss": 0.4779,
      "step": 2090
    },
    {
      "epoch": 0.9333333333333333,
      "grad_norm": 1.5886437892913818,
      "learning_rate": 4.485714285714286e-06,
      "loss": 0.5774,
      "step": 2100
    },
    {
      "epoch": 0.9377777777777778,
      "grad_norm": 1.5184223651885986,
      "learning_rate": 4.2000000000000004e-06,
      "loss": 0.4278,
      "step": 2110
    },
    {
      "epoch": 0.9422222222222222,
      "grad_norm": 4.225131511688232,
      "learning_rate": 3.914285714285715e-06,
      "loss": 0.6479,
      "step": 2120
    },
    {
      "epoch": 0.9466666666666667,
      "grad_norm": 2.665025234222412,
      "learning_rate": 3.6285714285714283e-06,
      "loss": 0.4907,
      "step": 2130
    },
    {
      "epoch": 0.9511111111111111,
      "grad_norm": 1.4222490787506104,
      "learning_rate": 3.3428571428571427e-06,
      "loss": 0.4263,
      "step": 2140
    },
    {
      "epoch": 0.9555555555555556,
      "grad_norm": 1.8098088502883911,
      "learning_rate": 3.0571428571428575e-06,
      "loss": 0.512,
      "step": 2150
    },
    {
      "epoch": 0.96,
      "grad_norm": 1.5185695886611938,
      "learning_rate": 2.771428571428572e-06,
      "loss": 0.4707,
      "step": 2160
    },
    {
      "epoch": 0.9644444444444444,
      "grad_norm": 3.510244846343994,
      "learning_rate": 2.4857142857142858e-06,
      "loss": 0.5985,
      "step": 2170
    },
    {
      "epoch": 0.9688888888888889,
      "grad_norm": 2.4858155250549316,
      "learning_rate": 2.2e-06,
      "loss": 0.5148,
      "step": 2180
    },
    {
      "epoch": 0.9733333333333334,
      "grad_norm": 2.5503909587860107,
      "learning_rate": 1.9142857142857145e-06,
      "loss": 0.5905,
      "step": 2190
    },
    {
      "epoch": 0.9777777777777777,
      "grad_norm": 1.640555739402771,
      "learning_rate": 1.6285714285714286e-06,
      "loss": 0.4983,
      "step": 2200
    },
    {
      "epoch": 0.9822222222222222,
      "grad_norm": 3.124906539916992,
      "learning_rate": 1.342857142857143e-06,
      "loss": 0.4201,
      "step": 2210
    },
    {
      "epoch": 0.9866666666666667,
      "grad_norm": 2.2014262676239014,
      "learning_rate": 1.0571428571428573e-06,
      "loss": 0.5991,
      "step": 2220
    },
    {
      "epoch": 0.9911111111111112,
      "grad_norm": 1.412158727645874,
      "learning_rate": 7.714285714285715e-07,
      "loss": 0.5062,
      "step": 2230
    },
    {
      "epoch": 0.9955555555555555,
      "grad_norm": 2.74906325340271,
      "learning_rate": 4.857142857142857e-07,
      "loss": 0.6499,
      "step": 2240
    },
    {
      "epoch": 1.0,
      "grad_norm": 1.1791956424713135,
      "learning_rate": 2.0000000000000002e-07,
      "loss": 0.521,
      "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": 3.20270921367552e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}