{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.2917653815780028,
  "eval_steps": 500,
  "global_step": 2250,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00574341302318903,
      "grad_norm": 3.515625,
      "learning_rate": 1.9920000000000002e-05,
      "loss": 1.3417,
      "step": 10
    },
    {
      "epoch": 0.01148682604637806,
      "grad_norm": 2.609375,
      "learning_rate": 1.9831111111111113e-05,
      "loss": 0.654,
      "step": 20
    },
    {
      "epoch": 0.01723023906956709,
      "grad_norm": 2.171875,
      "learning_rate": 1.9742222222222223e-05,
      "loss": 0.6043,
      "step": 30
    },
    {
      "epoch": 0.02297365209275612,
      "grad_norm": 2.296875,
      "learning_rate": 1.9653333333333334e-05,
      "loss": 0.5744,
      "step": 40
    },
    {
      "epoch": 0.02871706511594515,
      "grad_norm": 1.890625,
      "learning_rate": 1.9564444444444444e-05,
      "loss": 0.563,
      "step": 50
    },
    {
      "epoch": 0.03446047813913418,
      "grad_norm": 2.3125,
      "learning_rate": 1.9475555555555558e-05,
      "loss": 0.5506,
      "step": 60
    },
    {
      "epoch": 0.04020389116232321,
      "grad_norm": 1.984375,
      "learning_rate": 1.938666666666667e-05,
      "loss": 0.563,
      "step": 70
    },
    {
      "epoch": 0.04594730418551224,
      "grad_norm": 2.125,
      "learning_rate": 1.929777777777778e-05,
      "loss": 0.561,
      "step": 80
    },
    {
      "epoch": 0.05169071720870127,
      "grad_norm": 1.9765625,
      "learning_rate": 1.920888888888889e-05,
      "loss": 0.5288,
      "step": 90
    },
    {
      "epoch": 0.0574341302318903,
      "grad_norm": 2.109375,
      "learning_rate": 1.912e-05,
      "loss": 0.5693,
      "step": 100
    },
    {
      "epoch": 0.06317754325507934,
      "grad_norm": 2.21875,
      "learning_rate": 1.9031111111111114e-05,
      "loss": 0.5353,
      "step": 110
    },
    {
      "epoch": 0.06892095627826836,
      "grad_norm": 1.9140625,
      "learning_rate": 1.8942222222222224e-05,
      "loss": 0.5383,
      "step": 120
    },
    {
      "epoch": 0.0746643693014574,
      "grad_norm": 2.40625,
      "learning_rate": 1.8853333333333335e-05,
      "loss": 0.4849,
      "step": 130
    },
    {
      "epoch": 0.08040778232464642,
      "grad_norm": 1.8671875,
      "learning_rate": 1.8764444444444445e-05,
      "loss": 0.5155,
      "step": 140
    },
    {
      "epoch": 0.08615119534783545,
      "grad_norm": 2.265625,
      "learning_rate": 1.867555555555556e-05,
      "loss": 0.5071,
      "step": 150
    },
    {
      "epoch": 0.09189460837102448,
      "grad_norm": 2.03125,
      "learning_rate": 1.858666666666667e-05,
      "loss": 0.5065,
      "step": 160
    },
    {
      "epoch": 0.09763802139421351,
      "grad_norm": 1.84375,
      "learning_rate": 1.849777777777778e-05,
      "loss": 0.509,
      "step": 170
    },
    {
      "epoch": 0.10338143441740254,
      "grad_norm": 1.703125,
      "learning_rate": 1.840888888888889e-05,
      "loss": 0.5076,
      "step": 180
    },
    {
      "epoch": 0.10912484744059157,
      "grad_norm": 2.0,
      "learning_rate": 1.832e-05,
      "loss": 0.4804,
      "step": 190
    },
    {
      "epoch": 0.1148682604637806,
      "grad_norm": 1.8828125,
      "learning_rate": 1.823111111111111e-05,
      "loss": 0.4584,
      "step": 200
    },
    {
      "epoch": 0.12061167348696963,
      "grad_norm": 1.8046875,
      "learning_rate": 1.814222222222222e-05,
      "loss": 0.4951,
      "step": 210
    },
    {
      "epoch": 0.12635508651015867,
      "grad_norm": 1.7265625,
      "learning_rate": 1.8053333333333332e-05,
      "loss": 0.4758,
      "step": 220
    },
    {
      "epoch": 0.13209849953334768,
      "grad_norm": 2.1875,
      "learning_rate": 1.7964444444444446e-05,
      "loss": 0.5146,
      "step": 230
    },
    {
      "epoch": 0.13784191255653672,
      "grad_norm": 2.140625,
      "learning_rate": 1.7875555555555556e-05,
      "loss": 0.5004,
      "step": 240
    },
    {
      "epoch": 0.14358532557972575,
      "grad_norm": 2.234375,
      "learning_rate": 1.7786666666666667e-05,
      "loss": 0.4744,
      "step": 250
    },
    {
      "epoch": 0.1493287386029148,
      "grad_norm": 2.078125,
      "learning_rate": 1.7697777777777777e-05,
      "loss": 0.4943,
      "step": 260
    },
    {
      "epoch": 0.1550721516261038,
      "grad_norm": 2.046875,
      "learning_rate": 1.760888888888889e-05,
      "loss": 0.4751,
      "step": 270
    },
    {
      "epoch": 0.16081556464929284,
      "grad_norm": 2.09375,
      "learning_rate": 1.752e-05,
      "loss": 0.4599,
      "step": 280
    },
    {
      "epoch": 0.16655897767248187,
      "grad_norm": 1.8984375,
      "learning_rate": 1.7431111111111112e-05,
      "loss": 0.4656,
      "step": 290
    },
    {
      "epoch": 0.1723023906956709,
      "grad_norm": 1.9609375,
      "learning_rate": 1.7342222222222226e-05,
      "loss": 0.4242,
      "step": 300
    },
    {
      "epoch": 0.17804580371885995,
      "grad_norm": 2.0625,
      "learning_rate": 1.7253333333333336e-05,
      "loss": 0.4804,
      "step": 310
    },
    {
      "epoch": 0.18378921674204896,
      "grad_norm": 1.8046875,
      "learning_rate": 1.7164444444444447e-05,
      "loss": 0.4736,
      "step": 320
    },
    {
      "epoch": 0.189532629765238,
      "grad_norm": 2.0625,
      "learning_rate": 1.7075555555555557e-05,
      "loss": 0.4317,
      "step": 330
    },
    {
      "epoch": 0.19527604278842703,
      "grad_norm": 2.46875,
      "learning_rate": 1.6986666666666668e-05,
      "loss": 0.4752,
      "step": 340
    },
    {
      "epoch": 0.20101945581161607,
      "grad_norm": 1.8359375,
      "learning_rate": 1.6897777777777778e-05,
      "loss": 0.3932,
      "step": 350
    },
    {
      "epoch": 0.20676286883480507,
      "grad_norm": 2.109375,
      "learning_rate": 1.680888888888889e-05,
      "loss": 0.4742,
      "step": 360
    },
    {
      "epoch": 0.2125062818579941,
      "grad_norm": 2.34375,
      "learning_rate": 1.672e-05,
      "loss": 0.4542,
      "step": 370
    },
    {
      "epoch": 0.21824969488118315,
      "grad_norm": 1.9140625,
      "learning_rate": 1.6631111111111113e-05,
      "loss": 0.4284,
      "step": 380
    },
    {
      "epoch": 0.22399310790437219,
      "grad_norm": 1.90625,
      "learning_rate": 1.6542222222222223e-05,
      "loss": 0.4232,
      "step": 390
    },
    {
      "epoch": 0.2297365209275612,
      "grad_norm": 1.890625,
      "learning_rate": 1.6453333333333334e-05,
      "loss": 0.4299,
      "step": 400
    },
    {
      "epoch": 0.23547993395075023,
      "grad_norm": 2.375,
      "learning_rate": 1.6364444444444444e-05,
      "loss": 0.4246,
      "step": 410
    },
    {
      "epoch": 0.24122334697393927,
      "grad_norm": 2.015625,
      "learning_rate": 1.6275555555555558e-05,
      "loss": 0.4215,
      "step": 420
    },
    {
      "epoch": 0.2469667599971283,
      "grad_norm": 1.9375,
      "learning_rate": 1.618666666666667e-05,
      "loss": 0.4398,
      "step": 430
    },
    {
      "epoch": 0.25271017302031734,
      "grad_norm": 2.1875,
      "learning_rate": 1.609777777777778e-05,
      "loss": 0.4319,
      "step": 440
    },
    {
      "epoch": 0.2584535860435064,
      "grad_norm": 2.109375,
      "learning_rate": 1.600888888888889e-05,
      "loss": 0.4227,
      "step": 450
    },
    {
      "epoch": 0.26419699906669536,
      "grad_norm": 2.390625,
      "learning_rate": 1.5920000000000003e-05,
      "loss": 0.431,
      "step": 460
    },
    {
      "epoch": 0.2699404120898844,
      "grad_norm": 2.421875,
      "learning_rate": 1.5831111111111114e-05,
      "loss": 0.4323,
      "step": 470
    },
    {
      "epoch": 0.27568382511307343,
      "grad_norm": 2.15625,
      "learning_rate": 1.5742222222222224e-05,
      "loss": 0.4417,
      "step": 480
    },
    {
      "epoch": 0.28142723813626247,
      "grad_norm": 1.8984375,
      "learning_rate": 1.5653333333333335e-05,
      "loss": 0.3823,
      "step": 490
    },
    {
      "epoch": 0.2871706511594515,
      "grad_norm": 1.9140625,
      "learning_rate": 1.5564444444444445e-05,
      "loss": 0.394,
      "step": 500
    },
    {
      "epoch": 0.29291406418264054,
      "grad_norm": 2.171875,
      "learning_rate": 1.5475555555555556e-05,
      "loss": 0.4017,
      "step": 510
    },
    {
      "epoch": 0.2986574772058296,
      "grad_norm": 2.4375,
      "learning_rate": 1.5386666666666666e-05,
      "loss": 0.4121,
      "step": 520
    },
    {
      "epoch": 0.3044008902290186,
      "grad_norm": 2.078125,
      "learning_rate": 1.5297777777777777e-05,
      "loss": 0.3843,
      "step": 530
    },
    {
      "epoch": 0.3101443032522076,
      "grad_norm": 2.34375,
      "learning_rate": 1.520888888888889e-05,
      "loss": 0.4395,
      "step": 540
    },
    {
      "epoch": 0.31588771627539664,
      "grad_norm": 2.25,
      "learning_rate": 1.5120000000000001e-05,
      "loss": 0.3883,
      "step": 550
    },
    {
      "epoch": 0.32163112929858567,
      "grad_norm": 1.6171875,
      "learning_rate": 1.5031111111111111e-05,
      "loss": 0.4089,
      "step": 560
    },
    {
      "epoch": 0.3273745423217747,
      "grad_norm": 1.734375,
      "learning_rate": 1.4942222222222222e-05,
      "loss": 0.3895,
      "step": 570
    },
    {
      "epoch": 0.33311795534496375,
      "grad_norm": 1.9453125,
      "learning_rate": 1.4853333333333336e-05,
      "loss": 0.4084,
      "step": 580
    },
    {
      "epoch": 0.3388613683681528,
      "grad_norm": 2.171875,
      "learning_rate": 1.4764444444444446e-05,
      "loss": 0.4083,
      "step": 590
    },
    {
      "epoch": 0.3446047813913418,
      "grad_norm": 2.109375,
      "learning_rate": 1.4675555555555557e-05,
      "loss": 0.4037,
      "step": 600
    },
    {
      "epoch": 0.35034819441453086,
      "grad_norm": 2.53125,
      "learning_rate": 1.4586666666666667e-05,
      "loss": 0.3538,
      "step": 610
    },
    {
      "epoch": 0.3560916074377199,
      "grad_norm": 2.015625,
      "learning_rate": 1.449777777777778e-05,
      "loss": 0.3661,
      "step": 620
    },
    {
      "epoch": 0.3618350204609089,
      "grad_norm": 1.984375,
      "learning_rate": 1.440888888888889e-05,
      "loss": 0.3992,
      "step": 630
    },
    {
      "epoch": 0.3675784334840979,
      "grad_norm": 2.328125,
      "learning_rate": 1.432e-05,
      "loss": 0.3721,
      "step": 640
    },
    {
      "epoch": 0.37332184650728695,
      "grad_norm": 2.453125,
      "learning_rate": 1.4231111111111114e-05,
      "loss": 0.4009,
      "step": 650
    },
    {
      "epoch": 0.379065259530476,
      "grad_norm": 2.25,
      "learning_rate": 1.4142222222222224e-05,
      "loss": 0.401,
      "step": 660
    },
    {
      "epoch": 0.384808672553665,
      "grad_norm": 2.46875,
      "learning_rate": 1.4053333333333335e-05,
      "loss": 0.3783,
      "step": 670
    },
    {
      "epoch": 0.39055208557685406,
      "grad_norm": 2.234375,
      "learning_rate": 1.3964444444444445e-05,
      "loss": 0.4021,
      "step": 680
    },
    {
      "epoch": 0.3962954986000431,
      "grad_norm": 2.15625,
      "learning_rate": 1.3875555555555557e-05,
      "loss": 0.3713,
      "step": 690
    },
    {
      "epoch": 0.40203891162323213,
      "grad_norm": 2.125,
      "learning_rate": 1.3786666666666668e-05,
      "loss": 0.4036,
      "step": 700
    },
    {
      "epoch": 0.4077823246464211,
      "grad_norm": 2.03125,
      "learning_rate": 1.3697777777777778e-05,
      "loss": 0.368,
      "step": 710
    },
    {
      "epoch": 0.41352573766961015,
      "grad_norm": 2.5625,
      "learning_rate": 1.3608888888888889e-05,
      "loss": 0.3578,
      "step": 720
    },
    {
      "epoch": 0.4192691506927992,
      "grad_norm": 2.390625,
      "learning_rate": 1.3520000000000003e-05,
      "loss": 0.3731,
      "step": 730
    },
    {
      "epoch": 0.4250125637159882,
      "grad_norm": 2.5,
      "learning_rate": 1.3431111111111113e-05,
      "loss": 0.3817,
      "step": 740
    },
    {
      "epoch": 0.43075597673917726,
      "grad_norm": 2.515625,
      "learning_rate": 1.3342222222222224e-05,
      "loss": 0.3717,
      "step": 750
    },
    {
      "epoch": 0.4364993897623663,
      "grad_norm": 2.203125,
      "learning_rate": 1.3253333333333334e-05,
      "loss": 0.3629,
      "step": 760
    },
    {
      "epoch": 0.44224280278555533,
      "grad_norm": 2.625,
      "learning_rate": 1.3164444444444446e-05,
      "loss": 0.3698,
      "step": 770
    },
    {
      "epoch": 0.44798621580874437,
      "grad_norm": 2.0,
      "learning_rate": 1.3075555555555557e-05,
      "loss": 0.3661,
      "step": 780
    },
    {
      "epoch": 0.45372962883193335,
      "grad_norm": 2.5,
      "learning_rate": 1.2986666666666667e-05,
      "loss": 0.342,
      "step": 790
    },
    {
      "epoch": 0.4594730418551224,
      "grad_norm": 2.046875,
      "learning_rate": 1.2897777777777778e-05,
      "loss": 0.3862,
      "step": 800
    },
    {
      "epoch": 0.4652164548783114,
      "grad_norm": 2.40625,
      "learning_rate": 1.2808888888888891e-05,
      "loss": 0.3436,
      "step": 810
    },
    {
      "epoch": 0.47095986790150046,
      "grad_norm": 2.3125,
      "learning_rate": 1.2720000000000002e-05,
      "loss": 0.3374,
      "step": 820
    },
    {
      "epoch": 0.4767032809246895,
      "grad_norm": 2.15625,
      "learning_rate": 1.2631111111111112e-05,
      "loss": 0.3648,
      "step": 830
    },
    {
      "epoch": 0.48244669394787854,
      "grad_norm": 2.25,
      "learning_rate": 1.2542222222222223e-05,
      "loss": 0.327,
      "step": 840
    },
    {
      "epoch": 0.48819010697106757,
      "grad_norm": 2.3125,
      "learning_rate": 1.2453333333333335e-05,
      "loss": 0.3406,
      "step": 850
    },
    {
      "epoch": 0.4939335199942566,
      "grad_norm": 2.03125,
      "learning_rate": 1.2364444444444445e-05,
      "loss": 0.3441,
      "step": 860
    },
    {
      "epoch": 0.4996769330174456,
      "grad_norm": 2.609375,
      "learning_rate": 1.2275555555555556e-05,
      "loss": 0.3695,
      "step": 870
    },
    {
      "epoch": 0.5054203460406347,
      "grad_norm": 2.359375,
      "learning_rate": 1.2186666666666666e-05,
      "loss": 0.3386,
      "step": 880
    },
    {
      "epoch": 0.5111637590638237,
      "grad_norm": 2.328125,
      "learning_rate": 1.209777777777778e-05,
      "loss": 0.3478,
      "step": 890
    },
    {
      "epoch": 0.5169071720870128,
      "grad_norm": 2.078125,
      "learning_rate": 1.200888888888889e-05,
      "loss": 0.3274,
      "step": 900
    },
    {
      "epoch": 0.5226505851102017,
      "grad_norm": 2.3125,
      "learning_rate": 1.1920000000000001e-05,
      "loss": 0.3381,
      "step": 910
    },
    {
      "epoch": 0.5283939981333907,
      "grad_norm": 2.359375,
      "learning_rate": 1.1831111111111112e-05,
      "loss": 0.3438,
      "step": 920
    },
    {
      "epoch": 0.5341374111565798,
      "grad_norm": 2.390625,
      "learning_rate": 1.1742222222222224e-05,
      "loss": 0.3655,
      "step": 930
    },
    {
      "epoch": 0.5398808241797688,
      "grad_norm": 2.203125,
      "learning_rate": 1.1653333333333334e-05,
      "loss": 0.3219,
      "step": 940
    },
    {
      "epoch": 0.5456242372029578,
      "grad_norm": 2.453125,
      "learning_rate": 1.1564444444444445e-05,
      "loss": 0.3286,
      "step": 950
    },
    {
      "epoch": 0.5513676502261469,
      "grad_norm": 2.34375,
      "learning_rate": 1.1475555555555557e-05,
      "loss": 0.3547,
      "step": 960
    },
    {
      "epoch": 0.5571110632493359,
      "grad_norm": 2.578125,
      "learning_rate": 1.1386666666666669e-05,
      "loss": 0.3325,
      "step": 970
    },
    {
      "epoch": 0.5628544762725249,
      "grad_norm": 2.375,
      "learning_rate": 1.129777777777778e-05,
      "loss": 0.3711,
      "step": 980
    },
    {
      "epoch": 0.568597889295714,
      "grad_norm": 2.234375,
      "learning_rate": 1.120888888888889e-05,
      "loss": 0.3124,
      "step": 990
    },
    {
      "epoch": 0.574341302318903,
      "grad_norm": 2.234375,
      "learning_rate": 1.1120000000000002e-05,
      "loss": 0.3289,
      "step": 1000
    },
    {
      "epoch": 0.580084715342092,
      "grad_norm": 2.234375,
      "learning_rate": 1.1031111111111112e-05,
      "loss": 0.3218,
      "step": 1010
    },
    {
      "epoch": 0.5858281283652811,
      "grad_norm": 2.515625,
      "learning_rate": 1.0942222222222223e-05,
      "loss": 0.3503,
      "step": 1020
    },
    {
      "epoch": 0.5915715413884701,
      "grad_norm": 2.328125,
      "learning_rate": 1.0853333333333333e-05,
      "loss": 0.3551,
      "step": 1030
    },
    {
      "epoch": 0.5973149544116592,
      "grad_norm": 2.453125,
      "learning_rate": 1.0764444444444445e-05,
      "loss": 0.3194,
      "step": 1040
    },
    {
      "epoch": 0.6030583674348482,
      "grad_norm": 2.1875,
      "learning_rate": 1.0675555555555558e-05,
      "loss": 0.3478,
      "step": 1050
    },
    {
      "epoch": 0.6088017804580372,
      "grad_norm": 2.265625,
      "learning_rate": 1.0586666666666668e-05,
      "loss": 0.3202,
      "step": 1060
    },
    {
      "epoch": 0.6145451934812263,
      "grad_norm": 1.9140625,
      "learning_rate": 1.0497777777777779e-05,
      "loss": 0.3095,
      "step": 1070
    },
    {
      "epoch": 0.6202886065044152,
      "grad_norm": 2.71875,
      "learning_rate": 1.040888888888889e-05,
      "loss": 0.3112,
      "step": 1080
    },
    {
      "epoch": 0.6260320195276042,
      "grad_norm": 2.359375,
      "learning_rate": 1.0320000000000001e-05,
      "loss": 0.3115,
      "step": 1090
    },
    {
      "epoch": 0.6317754325507933,
      "grad_norm": 2.296875,
      "learning_rate": 1.0231111111111112e-05,
      "loss": 0.3296,
      "step": 1100
    },
    {
      "epoch": 0.6375188455739823,
      "grad_norm": 2.703125,
      "learning_rate": 1.0142222222222222e-05,
      "loss": 0.279,
      "step": 1110
    },
    {
      "epoch": 0.6432622585971713,
      "grad_norm": 2.71875,
      "learning_rate": 1.0053333333333334e-05,
      "loss": 0.3153,
      "step": 1120
    },
    {
      "epoch": 0.6490056716203604,
      "grad_norm": 1.953125,
      "learning_rate": 9.964444444444445e-06,
      "loss": 0.3039,
      "step": 1130
    },
    {
      "epoch": 0.6547490846435494,
      "grad_norm": 2.140625,
      "learning_rate": 9.875555555555557e-06,
      "loss": 0.3238,
      "step": 1140
    },
    {
      "epoch": 0.6604924976667385,
      "grad_norm": 2.25,
      "learning_rate": 9.786666666666667e-06,
      "loss": 0.3143,
      "step": 1150
    },
    {
      "epoch": 0.6662359106899275,
      "grad_norm": 2.546875,
      "learning_rate": 9.697777777777778e-06,
      "loss": 0.3145,
      "step": 1160
    },
    {
      "epoch": 0.6719793237131165,
      "grad_norm": 2.15625,
      "learning_rate": 9.60888888888889e-06,
      "loss": 0.3034,
      "step": 1170
    },
    {
      "epoch": 0.6777227367363056,
      "grad_norm": 3.875,
      "learning_rate": 9.52e-06,
      "loss": 0.2813,
      "step": 1180
    },
    {
      "epoch": 0.6834661497594946,
      "grad_norm": 2.046875,
      "learning_rate": 9.431111111111112e-06,
      "loss": 0.3211,
      "step": 1190
    },
    {
      "epoch": 0.6892095627826836,
      "grad_norm": 2.515625,
      "learning_rate": 9.342222222222223e-06,
      "loss": 0.2954,
      "step": 1200
    },
    {
      "epoch": 0.6949529758058727,
      "grad_norm": 2.34375,
      "learning_rate": 9.253333333333333e-06,
      "loss": 0.2969,
      "step": 1210
    },
    {
      "epoch": 0.7006963888290617,
      "grad_norm": 2.171875,
      "learning_rate": 9.164444444444446e-06,
      "loss": 0.2942,
      "step": 1220
    },
    {
      "epoch": 0.7064398018522507,
      "grad_norm": 2.0,
      "learning_rate": 9.075555555555556e-06,
      "loss": 0.321,
      "step": 1230
    },
    {
      "epoch": 0.7121832148754398,
      "grad_norm": 2.171875,
      "learning_rate": 8.986666666666666e-06,
      "loss": 0.2949,
      "step": 1240
    },
    {
      "epoch": 0.7179266278986287,
      "grad_norm": 2.609375,
      "learning_rate": 8.897777777777779e-06,
      "loss": 0.3111,
      "step": 1250
    },
    {
      "epoch": 0.7236700409218177,
      "grad_norm": 2.234375,
      "learning_rate": 8.80888888888889e-06,
      "loss": 0.3199,
      "step": 1260
    },
    {
      "epoch": 0.7294134539450068,
      "grad_norm": 3.171875,
      "learning_rate": 8.720000000000001e-06,
      "loss": 0.3239,
      "step": 1270
    },
    {
      "epoch": 0.7351568669681958,
      "grad_norm": 2.640625,
      "learning_rate": 8.631111111111112e-06,
      "loss": 0.3203,
      "step": 1280
    },
    {
      "epoch": 0.7409002799913849,
      "grad_norm": 2.3125,
      "learning_rate": 8.542222222222222e-06,
      "loss": 0.3071,
      "step": 1290
    },
    {
      "epoch": 0.7466436930145739,
      "grad_norm": 2.328125,
      "learning_rate": 8.453333333333334e-06,
      "loss": 0.3296,
      "step": 1300
    },
    {
      "epoch": 0.7523871060377629,
      "grad_norm": 2.296875,
      "learning_rate": 8.364444444444445e-06,
      "loss": 0.2994,
      "step": 1310
    },
    {
      "epoch": 0.758130519060952,
      "grad_norm": 2.390625,
      "learning_rate": 8.275555555555557e-06,
      "loss": 0.2661,
      "step": 1320
    },
    {
      "epoch": 0.763873932084141,
      "grad_norm": 2.34375,
      "learning_rate": 8.186666666666667e-06,
      "loss": 0.3194,
      "step": 1330
    },
    {
      "epoch": 0.76961734510733,
      "grad_norm": 2.28125,
      "learning_rate": 8.09777777777778e-06,
      "loss": 0.283,
      "step": 1340
    },
    {
      "epoch": 0.7753607581305191,
      "grad_norm": 2.734375,
      "learning_rate": 8.00888888888889e-06,
      "loss": 0.3358,
      "step": 1350
    },
    {
      "epoch": 0.7811041711537081,
      "grad_norm": 2.53125,
      "learning_rate": 7.92e-06,
      "loss": 0.3041,
      "step": 1360
    },
    {
      "epoch": 0.7868475841768972,
      "grad_norm": 2.671875,
      "learning_rate": 7.831111111111111e-06,
      "loss": 0.2701,
      "step": 1370
    },
    {
      "epoch": 0.7925909972000862,
      "grad_norm": 2.53125,
      "learning_rate": 7.742222222222223e-06,
      "loss": 0.2783,
      "step": 1380
    },
    {
      "epoch": 0.7983344102232752,
      "grad_norm": 2.296875,
      "learning_rate": 7.653333333333333e-06,
      "loss": 0.2989,
      "step": 1390
    },
    {
      "epoch": 0.8040778232464643,
      "grad_norm": 3.09375,
      "learning_rate": 7.564444444444446e-06,
      "loss": 0.2711,
      "step": 1400
    },
    {
      "epoch": 0.8098212362696532,
      "grad_norm": 2.21875,
      "learning_rate": 7.475555555555556e-06,
      "loss": 0.29,
      "step": 1410
    },
    {
      "epoch": 0.8155646492928422,
      "grad_norm": 2.359375,
      "learning_rate": 7.386666666666667e-06,
      "loss": 0.3208,
      "step": 1420
    },
    {
      "epoch": 0.8213080623160313,
      "grad_norm": 2.21875,
      "learning_rate": 7.297777777777778e-06,
      "loss": 0.2838,
      "step": 1430
    },
    {
      "epoch": 0.8270514753392203,
      "grad_norm": 2.828125,
      "learning_rate": 7.20888888888889e-06,
      "loss": 0.3104,
      "step": 1440
    },
    {
      "epoch": 0.8327948883624093,
      "grad_norm": 2.3125,
      "learning_rate": 7.1200000000000004e-06,
      "loss": 0.2805,
      "step": 1450
    },
    {
      "epoch": 0.8385383013855984,
      "grad_norm": 1.71875,
      "learning_rate": 7.031111111111112e-06,
      "loss": 0.2542,
      "step": 1460
    },
    {
      "epoch": 0.8442817144087874,
      "grad_norm": 2.1875,
      "learning_rate": 6.942222222222222e-06,
      "loss": 0.2698,
      "step": 1470
    },
    {
      "epoch": 0.8500251274319764,
      "grad_norm": 2.671875,
      "learning_rate": 6.853333333333334e-06,
      "loss": 0.2938,
      "step": 1480
    },
    {
      "epoch": 0.8557685404551655,
      "grad_norm": 2.484375,
      "learning_rate": 6.764444444444445e-06,
      "loss": 0.2657,
      "step": 1490
    },
    {
      "epoch": 0.8615119534783545,
      "grad_norm": 2.125,
      "learning_rate": 6.675555555555556e-06,
      "loss": 0.3194,
      "step": 1500
    },
    {
      "epoch": 0.8672553665015436,
      "grad_norm": 3.0,
      "learning_rate": 6.5866666666666666e-06,
      "loss": 0.2863,
      "step": 1510
    },
    {
      "epoch": 0.8729987795247326,
      "grad_norm": 2.640625,
      "learning_rate": 6.497777777777779e-06,
      "loss": 0.2636,
      "step": 1520
    },
    {
      "epoch": 0.8787421925479216,
      "grad_norm": 2.1875,
      "learning_rate": 6.408888888888889e-06,
      "loss": 0.2537,
      "step": 1530
    },
    {
      "epoch": 0.8844856055711107,
      "grad_norm": 2.265625,
      "learning_rate": 6.3200000000000005e-06,
      "loss": 0.2915,
      "step": 1540
    },
    {
      "epoch": 0.8902290185942997,
      "grad_norm": 2.53125,
      "learning_rate": 6.231111111111111e-06,
      "loss": 0.2814,
      "step": 1550
    },
    {
      "epoch": 0.8959724316174887,
      "grad_norm": 2.421875,
      "learning_rate": 6.142222222222223e-06,
      "loss": 0.2508,
      "step": 1560
    },
    {
      "epoch": 0.9017158446406778,
      "grad_norm": 1.8984375,
      "learning_rate": 6.0533333333333335e-06,
      "loss": 0.2499,
      "step": 1570
    },
    {
      "epoch": 0.9074592576638667,
      "grad_norm": 3.515625,
      "learning_rate": 5.964444444444445e-06,
      "loss": 0.2779,
      "step": 1580
    },
    {
      "epoch": 0.9132026706870557,
      "grad_norm": 2.09375,
      "learning_rate": 5.875555555555556e-06,
      "loss": 0.2703,
      "step": 1590
    },
    {
      "epoch": 0.9189460837102448,
      "grad_norm": 2.515625,
      "learning_rate": 5.7866666666666674e-06,
      "loss": 0.2851,
      "step": 1600
    },
    {
      "epoch": 0.9246894967334338,
      "grad_norm": 2.53125,
      "learning_rate": 5.697777777777779e-06,
      "loss": 0.2771,
      "step": 1610
    },
    {
      "epoch": 0.9304329097566229,
      "grad_norm": 2.515625,
      "learning_rate": 5.608888888888889e-06,
      "loss": 0.2545,
      "step": 1620
    },
    {
      "epoch": 0.9361763227798119,
      "grad_norm": 2.171875,
      "learning_rate": 5.5200000000000005e-06,
      "loss": 0.2534,
      "step": 1630
    },
    {
      "epoch": 0.9419197358030009,
      "grad_norm": 2.65625,
      "learning_rate": 5.431111111111112e-06,
      "loss": 0.2851,
      "step": 1640
    },
    {
      "epoch": 0.94766314882619,
      "grad_norm": 2.109375,
      "learning_rate": 5.342222222222223e-06,
      "loss": 0.3072,
      "step": 1650
    },
    {
      "epoch": 0.953406561849379,
      "grad_norm": 2.171875,
      "learning_rate": 5.2533333333333336e-06,
      "loss": 0.2709,
      "step": 1660
    },
    {
      "epoch": 0.959149974872568,
      "grad_norm": 2.359375,
      "learning_rate": 5.164444444444445e-06,
      "loss": 0.2876,
      "step": 1670
    },
    {
      "epoch": 0.9648933878957571,
      "grad_norm": 2.015625,
      "learning_rate": 5.075555555555556e-06,
      "loss": 0.2824,
      "step": 1680
    },
    {
      "epoch": 0.9706368009189461,
      "grad_norm": 2.515625,
      "learning_rate": 4.986666666666667e-06,
      "loss": 0.2867,
      "step": 1690
    },
    {
      "epoch": 0.9763802139421351,
      "grad_norm": 2.421875,
      "learning_rate": 4.897777777777778e-06,
      "loss": 0.2711,
      "step": 1700
    },
    {
      "epoch": 0.9821236269653242,
      "grad_norm": 2.375,
      "learning_rate": 4.808888888888889e-06,
      "loss": 0.3008,
      "step": 1710
    },
    {
      "epoch": 0.9878670399885132,
      "grad_norm": 1.9296875,
      "learning_rate": 4.7200000000000005e-06,
      "loss": 0.2649,
      "step": 1720
    },
    {
      "epoch": 0.9936104530117023,
      "grad_norm": 1.9296875,
      "learning_rate": 4.631111111111111e-06,
      "loss": 0.2625,
      "step": 1730
    },
    {
      "epoch": 0.9993538660348912,
      "grad_norm": 2.359375,
      "learning_rate": 4.542222222222223e-06,
      "loss": 0.2995,
      "step": 1740
    },
    {
      "epoch": 1.0045947304185512,
      "grad_norm": 2.078125,
      "learning_rate": 4.453333333333334e-06,
      "loss": 0.2862,
      "step": 1750
    },
    {
      "epoch": 1.0103381434417402,
      "grad_norm": 1.953125,
      "learning_rate": 4.364444444444445e-06,
      "loss": 0.24,
      "step": 1760
    },
    {
      "epoch": 1.0160815564649293,
      "grad_norm": 2.28125,
      "learning_rate": 4.275555555555556e-06,
      "loss": 0.2394,
      "step": 1770
    },
    {
      "epoch": 1.0218249694881183,
      "grad_norm": 1.796875,
      "learning_rate": 4.1866666666666675e-06,
      "loss": 0.2177,
      "step": 1780
    },
    {
      "epoch": 1.0275683825113073,
      "grad_norm": 2.28125,
      "learning_rate": 4.097777777777778e-06,
      "loss": 0.2514,
      "step": 1790
    },
    {
      "epoch": 1.0333117955344964,
      "grad_norm": 1.890625,
      "learning_rate": 4.008888888888889e-06,
      "loss": 0.2158,
      "step": 1800
    },
    {
      "epoch": 1.0390552085576854,
      "grad_norm": 2.015625,
      "learning_rate": 3.920000000000001e-06,
      "loss": 0.2377,
      "step": 1810
    },
    {
      "epoch": 1.0447986215808744,
      "grad_norm": 2.5,
      "learning_rate": 3.831111111111112e-06,
      "loss": 0.2193,
      "step": 1820
    },
    {
      "epoch": 1.0505420346040635,
      "grad_norm": 2.40625,
      "learning_rate": 3.7422222222222228e-06,
      "loss": 0.2354,
      "step": 1830
    },
    {
      "epoch": 1.0562854476272525,
      "grad_norm": 1.8984375,
      "learning_rate": 3.6533333333333336e-06,
      "loss": 0.2315,
      "step": 1840
    },
    {
      "epoch": 1.0620288606504416,
      "grad_norm": 2.484375,
      "learning_rate": 3.564444444444445e-06,
      "loss": 0.1987,
      "step": 1850
    },
    {
      "epoch": 1.0677722736736306,
      "grad_norm": 2.5,
      "learning_rate": 3.475555555555556e-06,
      "loss": 0.2276,
      "step": 1860
    },
    {
      "epoch": 1.0735156866968196,
      "grad_norm": 2.90625,
      "learning_rate": 3.386666666666667e-06,
      "loss": 0.225,
      "step": 1870
    },
    {
      "epoch": 1.0792590997200087,
      "grad_norm": 1.8046875,
      "learning_rate": 3.297777777777778e-06,
      "loss": 0.1992,
      "step": 1880
    },
    {
      "epoch": 1.0850025127431977,
      "grad_norm": 2.265625,
      "learning_rate": 3.2088888888888893e-06,
      "loss": 0.2068,
      "step": 1890
    },
    {
      "epoch": 1.0907459257663867,
      "grad_norm": 2.296875,
      "learning_rate": 3.12e-06,
      "loss": 0.225,
      "step": 1900
    },
    {
      "epoch": 1.0964893387895758,
      "grad_norm": 2.703125,
      "learning_rate": 3.0311111111111115e-06,
      "loss": 0.2173,
      "step": 1910
    },
    {
      "epoch": 1.1022327518127648,
      "grad_norm": 2.625,
      "learning_rate": 2.9422222222222224e-06,
      "loss": 0.2348,
      "step": 1920
    },
    {
      "epoch": 1.1079761648359538,
      "grad_norm": 1.796875,
      "learning_rate": 2.8533333333333337e-06,
      "loss": 0.2427,
      "step": 1930
    },
    {
      "epoch": 1.1137195778591429,
      "grad_norm": 2.0625,
      "learning_rate": 2.7644444444444446e-06,
      "loss": 0.2386,
      "step": 1940
    },
    {
      "epoch": 1.119462990882332,
      "grad_norm": 2.140625,
      "learning_rate": 2.675555555555556e-06,
      "loss": 0.2433,
      "step": 1950
    },
    {
      "epoch": 1.125206403905521,
      "grad_norm": 2.25,
      "learning_rate": 2.5866666666666667e-06,
      "loss": 0.2448,
      "step": 1960
    },
    {
      "epoch": 1.1309498169287098,
      "grad_norm": 2.140625,
      "learning_rate": 2.497777777777778e-06,
      "loss": 0.2304,
      "step": 1970
    },
    {
      "epoch": 1.136693229951899,
      "grad_norm": 2.09375,
      "learning_rate": 2.4088888888888894e-06,
      "loss": 0.2304,
      "step": 1980
    },
    {
      "epoch": 1.1424366429750878,
      "grad_norm": 2.515625,
      "learning_rate": 2.3200000000000002e-06,
      "loss": 0.236,
      "step": 1990
    },
    {
      "epoch": 1.1481800559982769,
      "grad_norm": 2.125,
      "learning_rate": 2.2311111111111115e-06,
      "loss": 0.2367,
      "step": 2000
    },
    {
      "epoch": 1.153923469021466,
      "grad_norm": 2.484375,
      "learning_rate": 2.1422222222222224e-06,
      "loss": 0.2178,
      "step": 2010
    },
    {
      "epoch": 1.159666882044655,
      "grad_norm": 2.53125,
      "learning_rate": 2.0533333333333337e-06,
      "loss": 0.2333,
      "step": 2020
    },
    {
      "epoch": 1.165410295067844,
      "grad_norm": 3.09375,
      "learning_rate": 1.9644444444444446e-06,
      "loss": 0.2586,
      "step": 2030
    },
    {
      "epoch": 1.171153708091033,
      "grad_norm": 2.953125,
      "learning_rate": 1.8755555555555557e-06,
      "loss": 0.2483,
      "step": 2040
    },
    {
      "epoch": 1.176897121114222,
      "grad_norm": 2.25,
      "learning_rate": 1.7866666666666668e-06,
      "loss": 0.2505,
      "step": 2050
    },
    {
      "epoch": 1.182640534137411,
      "grad_norm": 1.96875,
      "learning_rate": 1.6977777777777779e-06,
      "loss": 0.208,
      "step": 2060
    },
    {
      "epoch": 1.1883839471606001,
      "grad_norm": 2.390625,
      "learning_rate": 1.608888888888889e-06,
      "loss": 0.2279,
      "step": 2070
    },
    {
      "epoch": 1.1941273601837892,
      "grad_norm": 2.171875,
      "learning_rate": 1.52e-06,
      "loss": 0.2002,
      "step": 2080
    },
    {
      "epoch": 1.1998707732069782,
      "grad_norm": 2.28125,
      "learning_rate": 1.4311111111111111e-06,
      "loss": 0.2204,
      "step": 2090
    },
    {
      "epoch": 1.2056141862301673,
      "grad_norm": 2.453125,
      "learning_rate": 1.3422222222222222e-06,
      "loss": 0.2339,
      "step": 2100
    },
    {
      "epoch": 1.2113575992533563,
      "grad_norm": 1.875,
      "learning_rate": 1.2533333333333333e-06,
      "loss": 0.2362,
      "step": 2110
    },
    {
      "epoch": 1.2171010122765453,
      "grad_norm": 2.484375,
      "learning_rate": 1.1644444444444446e-06,
      "loss": 0.2324,
      "step": 2120
    },
    {
      "epoch": 1.2228444252997344,
      "grad_norm": 2.640625,
      "learning_rate": 1.0755555555555557e-06,
      "loss": 0.2404,
      "step": 2130
    },
    {
      "epoch": 1.2285878383229234,
      "grad_norm": 2.265625,
      "learning_rate": 9.866666666666668e-07,
      "loss": 0.202,
      "step": 2140
    },
    {
      "epoch": 1.2343312513461124,
      "grad_norm": 2.421875,
      "learning_rate": 8.977777777777778e-07,
      "loss": 0.2306,
      "step": 2150
    },
    {
      "epoch": 1.2400746643693015,
      "grad_norm": 2.46875,
      "learning_rate": 8.088888888888889e-07,
      "loss": 0.2094,
      "step": 2160
    },
    {
      "epoch": 1.2458180773924905,
      "grad_norm": 2.09375,
      "learning_rate": 7.2e-07,
      "loss": 0.2177,
      "step": 2170
    },
    {
      "epoch": 1.2515614904156795,
      "grad_norm": 2.515625,
      "learning_rate": 6.311111111111112e-07,
      "loss": 0.2508,
      "step": 2180
    },
    {
      "epoch": 1.2573049034388686,
      "grad_norm": 2.15625,
      "learning_rate": 5.422222222222223e-07,
      "loss": 0.2231,
      "step": 2190
    },
    {
      "epoch": 1.2630483164620576,
      "grad_norm": 1.7109375,
      "learning_rate": 4.533333333333334e-07,
      "loss": 0.2178,
      "step": 2200
    },
    {
      "epoch": 1.2687917294852467,
      "grad_norm": 2.5625,
      "learning_rate": 3.644444444444445e-07,
      "loss": 0.194,
      "step": 2210
    },
    {
      "epoch": 1.2745351425084357,
      "grad_norm": 2.125,
      "learning_rate": 2.7555555555555555e-07,
      "loss": 0.2302,
      "step": 2220
    },
    {
      "epoch": 1.2802785555316247,
      "grad_norm": 2.328125,
      "learning_rate": 1.866666666666667e-07,
      "loss": 0.2347,
      "step": 2230
    },
    {
      "epoch": 1.2860219685548138,
      "grad_norm": 2.1875,
      "learning_rate": 9.777777777777779e-08,
      "loss": 0.2061,
      "step": 2240
    },
    {
      "epoch": 1.2917653815780028,
      "grad_norm": 2.828125,
      "learning_rate": 8.88888888888889e-09,
      "loss": 0.2463,
      "step": 2250
    }
  ],
  "logging_steps": 10,
  "max_steps": 2250,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.0743805199693906e+18,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}