{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 99.2,
  "eval_steps": 50,
  "global_step": 6200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.16,
      "grad_norm": 0.730820894241333,
      "learning_rate": 0.00019967741935483872,
      "loss": 2.5389,
      "step": 10
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.7758501172065735,
      "learning_rate": 0.00019935483870967745,
      "loss": 1.8047,
      "step": 20
    },
    {
      "epoch": 0.48,
      "grad_norm": 4.077883720397949,
      "learning_rate": 0.00019906451612903227,
      "loss": 1.5731,
      "step": 30
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.27714163064956665,
      "learning_rate": 0.00019874193548387098,
      "loss": 1.508,
      "step": 40
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.1314336508512497,
      "learning_rate": 0.00019841935483870968,
      "loss": 1.4705,
      "step": 50
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.1689644455909729,
      "learning_rate": 0.0001980967741935484,
      "loss": 1.4734,
      "step": 60
    },
    {
      "epoch": 1.12,
      "grad_norm": 0.14989492297172546,
      "learning_rate": 0.00019777419354838712,
      "loss": 1.4617,
      "step": 70
    },
    {
      "epoch": 1.28,
      "grad_norm": 0.6179828643798828,
      "learning_rate": 0.00019745161290322583,
      "loss": 1.4413,
      "step": 80
    },
    {
      "epoch": 1.44,
      "grad_norm": 2.594003677368164,
      "learning_rate": 0.0001971290322580645,
      "loss": 1.353,
      "step": 90
    },
    {
      "epoch": 1.6,
      "grad_norm": 0.08125967532396317,
      "learning_rate": 0.00019680645161290324,
      "loss": 1.3227,
      "step": 100
    },
    {
      "epoch": 1.76,
      "grad_norm": 0.09471185505390167,
      "learning_rate": 0.00019648387096774195,
      "loss": 1.3214,
      "step": 110
    },
    {
      "epoch": 1.92,
      "grad_norm": 0.10813447833061218,
      "learning_rate": 0.00019616129032258065,
      "loss": 1.3236,
      "step": 120
    },
    {
      "epoch": 2.08,
      "grad_norm": 0.07840854674577713,
      "learning_rate": 0.0001958387096774194,
      "loss": 1.3222,
      "step": 130
    },
    {
      "epoch": 2.24,
      "grad_norm": 0.08169138431549072,
      "learning_rate": 0.00019551612903225807,
      "loss": 1.3196,
      "step": 140
    },
    {
      "epoch": 2.4,
      "grad_norm": 0.15425510704517365,
      "learning_rate": 0.00019519354838709677,
      "loss": 1.3191,
      "step": 150
    },
    {
      "epoch": 2.56,
      "grad_norm": 0.08724704384803772,
      "learning_rate": 0.0001948709677419355,
      "loss": 1.3223,
      "step": 160
    },
    {
      "epoch": 2.7199999999999998,
      "grad_norm": 0.1369275003671646,
      "learning_rate": 0.0001945483870967742,
      "loss": 1.3209,
      "step": 170
    },
    {
      "epoch": 2.88,
      "grad_norm": 0.12385846674442291,
      "learning_rate": 0.0001942258064516129,
      "loss": 1.3211,
      "step": 180
    },
    {
      "epoch": 3.04,
      "grad_norm": 0.1191711276769638,
      "learning_rate": 0.00019390322580645162,
      "loss": 1.3225,
      "step": 190
    },
    {
      "epoch": 3.2,
      "grad_norm": 0.0959768146276474,
      "learning_rate": 0.00019358064516129033,
      "loss": 1.3192,
      "step": 200
    },
    {
      "epoch": 3.36,
      "grad_norm": 0.08907590806484222,
      "learning_rate": 0.00019325806451612904,
      "loss": 1.319,
      "step": 210
    },
    {
      "epoch": 3.52,
      "grad_norm": 0.1829214245080948,
      "learning_rate": 0.00019293548387096777,
      "loss": 1.3198,
      "step": 220
    },
    {
      "epoch": 3.68,
      "grad_norm": 0.10891153663396835,
      "learning_rate": 0.00019261290322580645,
      "loss": 1.32,
      "step": 230
    },
    {
      "epoch": 3.84,
      "grad_norm": 0.10339193791151047,
      "learning_rate": 0.00019229032258064516,
      "loss": 1.3191,
      "step": 240
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.0916794016957283,
      "learning_rate": 0.0001919677419354839,
      "loss": 1.3223,
      "step": 250
    },
    {
      "epoch": 4.16,
      "grad_norm": 0.15707476437091827,
      "learning_rate": 0.0001916451612903226,
      "loss": 1.3196,
      "step": 260
    },
    {
      "epoch": 4.32,
      "grad_norm": 0.11287837475538254,
      "learning_rate": 0.0001913225806451613,
      "loss": 1.3168,
      "step": 270
    },
    {
      "epoch": 4.48,
      "grad_norm": 0.19220077991485596,
      "learning_rate": 0.000191,
      "loss": 1.3164,
      "step": 280
    },
    {
      "epoch": 4.64,
      "grad_norm": 0.08514147996902466,
      "learning_rate": 0.00019067741935483871,
      "loss": 1.3193,
      "step": 290
    },
    {
      "epoch": 4.8,
      "grad_norm": 0.12788636982440948,
      "learning_rate": 0.00019035483870967742,
      "loss": 1.3194,
      "step": 300
    },
    {
      "epoch": 4.96,
      "grad_norm": 0.14810492098331451,
      "learning_rate": 0.00019003225806451615,
      "loss": 1.3193,
      "step": 310
    },
    {
      "epoch": 5.12,
      "grad_norm": 0.16166192293167114,
      "learning_rate": 0.00018970967741935486,
      "loss": 1.3183,
      "step": 320
    },
    {
      "epoch": 5.28,
      "grad_norm": 0.23055078089237213,
      "learning_rate": 0.00018938709677419354,
      "loss": 1.3156,
      "step": 330
    },
    {
      "epoch": 5.44,
      "grad_norm": 0.12718935310840607,
      "learning_rate": 0.00018906451612903227,
      "loss": 1.3163,
      "step": 340
    },
    {
      "epoch": 5.6,
      "grad_norm": 0.20426005125045776,
      "learning_rate": 0.00018874193548387098,
      "loss": 1.3157,
      "step": 350
    },
    {
      "epoch": 5.76,
      "grad_norm": 0.11526554077863693,
      "learning_rate": 0.00018841935483870968,
      "loss": 1.3161,
      "step": 360
    },
    {
      "epoch": 5.92,
      "grad_norm": 0.1523907333612442,
      "learning_rate": 0.0001880967741935484,
      "loss": 1.3176,
      "step": 370
    },
    {
      "epoch": 6.08,
      "grad_norm": 0.2060098499059677,
      "learning_rate": 0.0001877741935483871,
      "loss": 1.3142,
      "step": 380
    },
    {
      "epoch": 6.24,
      "grad_norm": 0.24104587733745575,
      "learning_rate": 0.0001874516129032258,
      "loss": 1.3123,
      "step": 390
    },
    {
      "epoch": 6.4,
      "grad_norm": 0.17447544634342194,
      "learning_rate": 0.00018712903225806454,
      "loss": 1.31,
      "step": 400
    },
    {
      "epoch": 6.5600000000000005,
      "grad_norm": 0.1756139099597931,
      "learning_rate": 0.00018680645161290324,
      "loss": 1.3142,
      "step": 410
    },
    {
      "epoch": 6.72,
      "grad_norm": 0.17624247074127197,
      "learning_rate": 0.00018648387096774195,
      "loss": 1.3157,
      "step": 420
    },
    {
      "epoch": 6.88,
      "grad_norm": 0.1455477774143219,
      "learning_rate": 0.00018616129032258065,
      "loss": 1.3155,
      "step": 430
    },
    {
      "epoch": 7.04,
      "grad_norm": 0.21999657154083252,
      "learning_rate": 0.00018583870967741936,
      "loss": 1.3143,
      "step": 440
    },
    {
      "epoch": 7.2,
      "grad_norm": 0.24742542207241058,
      "learning_rate": 0.00018551612903225807,
      "loss": 1.3003,
      "step": 450
    },
    {
      "epoch": 7.36,
      "grad_norm": 0.2892099618911743,
      "learning_rate": 0.0001851935483870968,
      "loss": 1.3064,
      "step": 460
    },
    {
      "epoch": 7.52,
      "grad_norm": 0.2647784650325775,
      "learning_rate": 0.00018487096774193548,
      "loss": 1.3058,
      "step": 470
    },
    {
      "epoch": 7.68,
      "grad_norm": 0.2782827913761139,
      "learning_rate": 0.0001845483870967742,
      "loss": 1.3097,
      "step": 480
    },
    {
      "epoch": 7.84,
      "grad_norm": 0.17432355880737305,
      "learning_rate": 0.00018422580645161292,
      "loss": 1.3141,
      "step": 490
    },
    {
      "epoch": 8.0,
      "grad_norm": 0.20223839581012726,
      "learning_rate": 0.00018390322580645163,
      "loss": 1.3125,
      "step": 500
    },
    {
      "epoch": 8.16,
      "grad_norm": 0.39165350794792175,
      "learning_rate": 0.00018358064516129033,
      "loss": 1.2921,
      "step": 510
    },
    {
      "epoch": 8.32,
      "grad_norm": 0.35312631726264954,
      "learning_rate": 0.00018325806451612904,
      "loss": 1.2952,
      "step": 520
    },
    {
      "epoch": 8.48,
      "grad_norm": 0.34015020728111267,
      "learning_rate": 0.00018293548387096774,
      "loss": 1.297,
      "step": 530
    },
    {
      "epoch": 8.64,
      "grad_norm": 0.3401159346103668,
      "learning_rate": 0.00018261290322580648,
      "loss": 1.298,
      "step": 540
    },
    {
      "epoch": 8.8,
      "grad_norm": 0.3697248697280884,
      "learning_rate": 0.00018229032258064518,
      "loss": 1.3025,
      "step": 550
    },
    {
      "epoch": 8.96,
      "grad_norm": 0.31431111693382263,
      "learning_rate": 0.00018196774193548386,
      "loss": 1.3042,
      "step": 560
    },
    {
      "epoch": 9.12,
      "grad_norm": 0.5856003761291504,
      "learning_rate": 0.0001816451612903226,
      "loss": 1.2838,
      "step": 570
    },
    {
      "epoch": 9.28,
      "grad_norm": 0.5546128749847412,
      "learning_rate": 0.0001813225806451613,
      "loss": 1.2766,
      "step": 580
    },
    {
      "epoch": 9.44,
      "grad_norm": 0.5272495150566101,
      "learning_rate": 0.000181,
      "loss": 1.2825,
      "step": 590
    },
    {
      "epoch": 9.6,
      "grad_norm": 0.37607982754707336,
      "learning_rate": 0.00018067741935483874,
      "loss": 1.2855,
      "step": 600
    },
    {
      "epoch": 9.76,
      "grad_norm": 0.4203632175922394,
      "learning_rate": 0.00018035483870967742,
      "loss": 1.2905,
      "step": 610
    },
    {
      "epoch": 9.92,
      "grad_norm": 0.5385815501213074,
      "learning_rate": 0.00018003225806451613,
      "loss": 1.2862,
      "step": 620
    },
    {
      "epoch": 10.08,
      "grad_norm": 0.6651461124420166,
      "learning_rate": 0.00017970967741935486,
      "loss": 1.2753,
      "step": 630
    },
    {
      "epoch": 10.24,
      "grad_norm": 0.6366870999336243,
      "learning_rate": 0.00017938709677419357,
      "loss": 1.2522,
      "step": 640
    },
    {
      "epoch": 10.4,
      "grad_norm": 0.662529468536377,
      "learning_rate": 0.00017906451612903225,
      "loss": 1.2568,
      "step": 650
    },
    {
      "epoch": 10.56,
      "grad_norm": 0.569629967212677,
      "learning_rate": 0.00017874193548387098,
      "loss": 1.2596,
      "step": 660
    },
    {
      "epoch": 10.72,
      "grad_norm": 0.5770915150642395,
      "learning_rate": 0.00017841935483870969,
      "loss": 1.2699,
      "step": 670
    },
    {
      "epoch": 10.88,
      "grad_norm": 0.5522739887237549,
      "learning_rate": 0.0001780967741935484,
      "loss": 1.2682,
      "step": 680
    },
    {
      "epoch": 11.04,
      "grad_norm": 0.7272565960884094,
      "learning_rate": 0.00017777419354838712,
      "loss": 1.2536,
      "step": 690
    },
    {
      "epoch": 11.2,
      "grad_norm": 0.9644076824188232,
      "learning_rate": 0.0001774516129032258,
      "loss": 1.22,
      "step": 700
    },
    {
      "epoch": 11.36,
      "grad_norm": 0.7902132272720337,
      "learning_rate": 0.0001771290322580645,
      "loss": 1.2189,
      "step": 710
    },
    {
      "epoch": 11.52,
      "grad_norm": 0.7859879732131958,
      "learning_rate": 0.00017680645161290324,
      "loss": 1.231,
      "step": 720
    },
    {
      "epoch": 11.68,
      "grad_norm": 0.7887389659881592,
      "learning_rate": 0.00017648387096774195,
      "loss": 1.2358,
      "step": 730
    },
    {
      "epoch": 11.84,
      "grad_norm": 0.7551841139793396,
      "learning_rate": 0.00017616129032258066,
      "loss": 1.2444,
      "step": 740
    },
    {
      "epoch": 12.0,
      "grad_norm": 0.6560466885566711,
      "learning_rate": 0.00017583870967741936,
      "loss": 1.2326,
      "step": 750
    },
    {
      "epoch": 12.16,
      "grad_norm": 1.0130267143249512,
      "learning_rate": 0.00017551612903225807,
      "loss": 1.167,
      "step": 760
    },
    {
      "epoch": 12.32,
      "grad_norm": 0.8946852087974548,
      "learning_rate": 0.00017519354838709677,
      "loss": 1.1776,
      "step": 770
    },
    {
      "epoch": 12.48,
      "grad_norm": 1.0538991689682007,
      "learning_rate": 0.0001748709677419355,
      "loss": 1.1792,
      "step": 780
    },
    {
      "epoch": 12.64,
      "grad_norm": 1.0721157789230347,
      "learning_rate": 0.00017454838709677421,
      "loss": 1.1915,
      "step": 790
    },
    {
      "epoch": 12.8,
      "grad_norm": 0.8850572109222412,
      "learning_rate": 0.0001742258064516129,
      "loss": 1.193,
      "step": 800
    },
    {
      "epoch": 12.96,
      "grad_norm": 1.0722297430038452,
      "learning_rate": 0.00017390322580645163,
      "loss": 1.1925,
      "step": 810
    },
    {
      "epoch": 13.12,
      "grad_norm": 1.2234816551208496,
      "learning_rate": 0.00017358064516129033,
      "loss": 1.1253,
      "step": 820
    },
    {
      "epoch": 13.28,
      "grad_norm": 1.289270043373108,
      "learning_rate": 0.00017325806451612904,
      "loss": 1.1006,
      "step": 830
    },
    {
      "epoch": 13.44,
      "grad_norm": 1.1833810806274414,
      "learning_rate": 0.00017293548387096775,
      "loss": 1.102,
      "step": 840
    },
    {
      "epoch": 13.6,
      "grad_norm": 1.2299394607543945,
      "learning_rate": 0.00017261290322580645,
      "loss": 1.1233,
      "step": 850
    },
    {
      "epoch": 13.76,
      "grad_norm": 1.1658111810684204,
      "learning_rate": 0.00017229032258064516,
      "loss": 1.1253,
      "step": 860
    },
    {
      "epoch": 13.92,
      "grad_norm": 1.1958309412002563,
      "learning_rate": 0.0001719677419354839,
      "loss": 1.1253,
      "step": 870
    },
    {
      "epoch": 14.08,
      "grad_norm": 1.418826699256897,
      "learning_rate": 0.0001716451612903226,
      "loss": 1.0708,
      "step": 880
    },
    {
      "epoch": 14.24,
      "grad_norm": 1.5377317667007446,
      "learning_rate": 0.0001713225806451613,
      "loss": 1.0089,
      "step": 890
    },
    {
      "epoch": 14.4,
      "grad_norm": 1.5082684755325317,
      "learning_rate": 0.000171,
      "loss": 1.0263,
      "step": 900
    },
    {
      "epoch": 14.56,
      "grad_norm": 1.5111769437789917,
      "learning_rate": 0.00017067741935483872,
      "loss": 1.0347,
      "step": 910
    },
    {
      "epoch": 14.72,
      "grad_norm": 1.4334949254989624,
      "learning_rate": 0.00017035483870967742,
      "loss": 1.037,
      "step": 920
    },
    {
      "epoch": 14.88,
      "grad_norm": 2.889636278152466,
      "learning_rate": 0.00017003225806451616,
      "loss": 1.0348,
      "step": 930
    },
    {
      "epoch": 15.04,
      "grad_norm": 1.4357563257217407,
      "learning_rate": 0.00016970967741935483,
      "loss": 1.0162,
      "step": 940
    },
    {
      "epoch": 15.2,
      "grad_norm": 1.679137110710144,
      "learning_rate": 0.00016938709677419357,
      "loss": 0.9209,
      "step": 950
    },
    {
      "epoch": 15.36,
      "grad_norm": 1.6809172630310059,
      "learning_rate": 0.00016906451612903227,
      "loss": 0.9224,
      "step": 960
    },
    {
      "epoch": 15.52,
      "grad_norm": 2.0711913108825684,
      "learning_rate": 0.00016874193548387098,
      "loss": 0.9374,
      "step": 970
    },
    {
      "epoch": 15.68,
      "grad_norm": 1.6750158071517944,
      "learning_rate": 0.00016841935483870969,
      "loss": 0.9386,
      "step": 980
    },
    {
      "epoch": 15.84,
      "grad_norm": 1.7292613983154297,
      "learning_rate": 0.0001680967741935484,
      "loss": 0.949,
      "step": 990
    },
    {
      "epoch": 16.0,
      "grad_norm": 1.7566314935684204,
      "learning_rate": 0.0001677741935483871,
      "loss": 0.9549,
      "step": 1000
    },
    {
      "epoch": 16.16,
      "grad_norm": 1.9804630279541016,
      "learning_rate": 0.00016745161290322583,
      "loss": 0.8478,
      "step": 1010
    },
    {
      "epoch": 16.32,
      "grad_norm": 1.8297134637832642,
      "learning_rate": 0.00016712903225806454,
      "loss": 0.813,
      "step": 1020
    },
    {
      "epoch": 16.48,
      "grad_norm": 2.0066633224487305,
      "learning_rate": 0.00016680645161290322,
      "loss": 0.835,
      "step": 1030
    },
    {
      "epoch": 16.64,
      "grad_norm": 1.9635342359542847,
      "learning_rate": 0.00016648387096774195,
      "loss": 0.8383,
      "step": 1040
    },
    {
      "epoch": 16.8,
      "grad_norm": 1.8945978879928589,
      "learning_rate": 0.00016616129032258066,
      "loss": 0.8675,
      "step": 1050
    },
    {
      "epoch": 16.96,
      "grad_norm": 1.917394757270813,
      "learning_rate": 0.00016583870967741936,
      "loss": 0.8529,
      "step": 1060
    },
    {
      "epoch": 17.12,
      "grad_norm": 1.9029055833816528,
      "learning_rate": 0.00016551612903225807,
      "loss": 0.7593,
      "step": 1070
    },
    {
      "epoch": 17.28,
      "grad_norm": 2.5159807205200195,
      "learning_rate": 0.00016519354838709678,
      "loss": 0.7418,
      "step": 1080
    },
    {
      "epoch": 17.44,
      "grad_norm": 1.9958096742630005,
      "learning_rate": 0.00016487096774193548,
      "loss": 0.7489,
      "step": 1090
    },
    {
      "epoch": 17.6,
      "grad_norm": 2.0814130306243896,
      "learning_rate": 0.00016454838709677421,
      "loss": 0.7502,
      "step": 1100
    },
    {
      "epoch": 17.76,
      "grad_norm": 2.4516828060150146,
      "learning_rate": 0.00016422580645161292,
      "loss": 0.7607,
      "step": 1110
    },
    {
      "epoch": 17.92,
      "grad_norm": 2.1293084621429443,
      "learning_rate": 0.0001639032258064516,
      "loss": 0.7671,
      "step": 1120
    },
    {
      "epoch": 18.08,
      "grad_norm": 2.431506872177124,
      "learning_rate": 0.00016358064516129033,
      "loss": 0.7117,
      "step": 1130
    },
    {
      "epoch": 18.24,
      "grad_norm": 2.156888723373413,
      "learning_rate": 0.00016325806451612904,
      "loss": 0.6516,
      "step": 1140
    },
    {
      "epoch": 18.4,
      "grad_norm": 2.404205322265625,
      "learning_rate": 0.00016293548387096775,
      "loss": 0.6595,
      "step": 1150
    },
    {
      "epoch": 18.56,
      "grad_norm": 2.5965211391448975,
      "learning_rate": 0.00016261290322580648,
      "loss": 0.6738,
      "step": 1160
    },
    {
      "epoch": 18.72,
      "grad_norm": 2.4526731967926025,
      "learning_rate": 0.00016229032258064516,
      "loss": 0.6581,
      "step": 1170
    },
    {
      "epoch": 18.88,
      "grad_norm": 2.5058436393737793,
      "learning_rate": 0.00016196774193548386,
      "loss": 0.6821,
      "step": 1180
    },
    {
      "epoch": 19.04,
      "grad_norm": 2.1841073036193848,
      "learning_rate": 0.0001616451612903226,
      "loss": 0.6644,
      "step": 1190
    },
    {
      "epoch": 19.2,
      "grad_norm": 2.2386417388916016,
      "learning_rate": 0.0001613225806451613,
      "loss": 0.572,
      "step": 1200
    },
    {
      "epoch": 19.36,
      "grad_norm": 2.338043689727783,
      "learning_rate": 0.000161,
      "loss": 0.5822,
      "step": 1210
    },
    {
      "epoch": 19.52,
      "grad_norm": 2.37507700920105,
      "learning_rate": 0.00016067741935483872,
      "loss": 0.5925,
      "step": 1220
    },
    {
      "epoch": 19.68,
      "grad_norm": 2.571474552154541,
      "learning_rate": 0.00016035483870967742,
      "loss": 0.6041,
      "step": 1230
    },
    {
      "epoch": 19.84,
      "grad_norm": 2.4885830879211426,
      "learning_rate": 0.00016003225806451613,
      "loss": 0.6237,
      "step": 1240
    },
    {
      "epoch": 20.0,
      "grad_norm": 2.9572579860687256,
      "learning_rate": 0.00015970967741935486,
      "loss": 0.6022,
      "step": 1250
    },
    {
      "epoch": 20.16,
      "grad_norm": 2.3328864574432373,
      "learning_rate": 0.00015938709677419354,
      "loss": 0.5042,
      "step": 1260
    },
    {
      "epoch": 20.32,
      "grad_norm": 2.6092636585235596,
      "learning_rate": 0.00015906451612903225,
      "loss": 0.5112,
      "step": 1270
    },
    {
      "epoch": 20.48,
      "grad_norm": 2.577423572540283,
      "learning_rate": 0.00015874193548387098,
      "loss": 0.5242,
      "step": 1280
    },
    {
      "epoch": 20.64,
      "grad_norm": 2.496828556060791,
      "learning_rate": 0.0001584193548387097,
      "loss": 0.5326,
      "step": 1290
    },
    {
      "epoch": 20.8,
      "grad_norm": 2.479008197784424,
      "learning_rate": 0.0001580967741935484,
      "loss": 0.5427,
      "step": 1300
    },
    {
      "epoch": 20.96,
      "grad_norm": 3.1138076782226562,
      "learning_rate": 0.0001577741935483871,
      "loss": 0.5466,
      "step": 1310
    },
    {
      "epoch": 21.12,
      "grad_norm": 2.6781411170959473,
      "learning_rate": 0.0001574516129032258,
      "loss": 0.4764,
      "step": 1320
    },
    {
      "epoch": 21.28,
      "grad_norm": 3.1711771488189697,
      "learning_rate": 0.0001571290322580645,
      "loss": 0.4623,
      "step": 1330
    },
    {
      "epoch": 21.44,
      "grad_norm": 2.4297163486480713,
      "learning_rate": 0.00015680645161290325,
      "loss": 0.4542,
      "step": 1340
    },
    {
      "epoch": 21.6,
      "grad_norm": 2.8714449405670166,
      "learning_rate": 0.00015648387096774195,
      "loss": 0.4752,
      "step": 1350
    },
    {
      "epoch": 21.76,
      "grad_norm": 2.6193573474884033,
      "learning_rate": 0.00015616129032258066,
      "loss": 0.484,
      "step": 1360
    },
    {
      "epoch": 21.92,
      "grad_norm": 3.0310218334198,
      "learning_rate": 0.00015583870967741936,
      "loss": 0.4792,
      "step": 1370
    },
    {
      "epoch": 22.08,
      "grad_norm": 2.4924027919769287,
      "learning_rate": 0.00015551612903225807,
      "loss": 0.4529,
      "step": 1380
    },
    {
      "epoch": 22.24,
      "grad_norm": 2.0856494903564453,
      "learning_rate": 0.00015519354838709678,
      "loss": 0.4077,
      "step": 1390
    },
    {
      "epoch": 22.4,
      "grad_norm": 2.794166088104248,
      "learning_rate": 0.0001548709677419355,
      "loss": 0.4142,
      "step": 1400
    },
    {
      "epoch": 22.56,
      "grad_norm": 2.5052692890167236,
      "learning_rate": 0.0001545483870967742,
      "loss": 0.4209,
      "step": 1410
    },
    {
      "epoch": 22.72,
      "grad_norm": 2.6941845417022705,
      "learning_rate": 0.00015422580645161292,
      "loss": 0.4299,
      "step": 1420
    },
    {
      "epoch": 22.88,
      "grad_norm": 3.0589232444763184,
      "learning_rate": 0.00015390322580645163,
      "loss": 0.4398,
      "step": 1430
    },
    {
      "epoch": 23.04,
      "grad_norm": 2.647005558013916,
      "learning_rate": 0.00015358064516129033,
      "loss": 0.4283,
      "step": 1440
    },
    {
      "epoch": 23.2,
      "grad_norm": 2.0591695308685303,
      "learning_rate": 0.00015325806451612904,
      "loss": 0.3667,
      "step": 1450
    },
    {
      "epoch": 23.36,
      "grad_norm": 2.557981014251709,
      "learning_rate": 0.00015293548387096775,
      "loss": 0.3748,
      "step": 1460
    },
    {
      "epoch": 23.52,
      "grad_norm": 2.7171032428741455,
      "learning_rate": 0.00015261290322580645,
      "loss": 0.3771,
      "step": 1470
    },
    {
      "epoch": 23.68,
      "grad_norm": 2.8117640018463135,
      "learning_rate": 0.00015229032258064516,
      "loss": 0.3857,
      "step": 1480
    },
    {
      "epoch": 23.84,
      "grad_norm": 2.635938882827759,
      "learning_rate": 0.0001519677419354839,
      "loss": 0.3955,
      "step": 1490
    },
    {
      "epoch": 24.0,
      "grad_norm": 2.530909299850464,
      "learning_rate": 0.00015164516129032257,
      "loss": 0.4003,
      "step": 1500
    },
    {
      "epoch": 24.16,
      "grad_norm": 2.716811418533325,
      "learning_rate": 0.0001513225806451613,
      "loss": 0.3257,
      "step": 1510
    },
    {
      "epoch": 24.32,
      "grad_norm": 2.6447346210479736,
      "learning_rate": 0.000151,
      "loss": 0.3342,
      "step": 1520
    },
    {
      "epoch": 24.48,
      "grad_norm": 2.419239044189453,
      "learning_rate": 0.00015067741935483872,
      "loss": 0.3443,
      "step": 1530
    },
    {
      "epoch": 24.64,
      "grad_norm": 2.9881439208984375,
      "learning_rate": 0.00015035483870967742,
      "loss": 0.351,
      "step": 1540
    },
    {
      "epoch": 24.8,
      "grad_norm": 2.9499566555023193,
      "learning_rate": 0.00015003225806451613,
      "loss": 0.3626,
      "step": 1550
    },
    {
      "epoch": 24.96,
      "grad_norm": 3.1362252235412598,
      "learning_rate": 0.00014970967741935484,
      "loss": 0.3572,
      "step": 1560
    },
    {
      "epoch": 25.12,
      "grad_norm": 2.9108505249023438,
      "learning_rate": 0.00014938709677419357,
      "loss": 0.3141,
      "step": 1570
    },
    {
      "epoch": 25.28,
      "grad_norm": 2.4450314044952393,
      "learning_rate": 0.00014906451612903228,
      "loss": 0.3062,
      "step": 1580
    },
    {
      "epoch": 25.44,
      "grad_norm": 3.3903231620788574,
      "learning_rate": 0.00014874193548387095,
      "loss": 0.3137,
      "step": 1590
    },
    {
      "epoch": 25.6,
      "grad_norm": 2.7393417358398438,
      "learning_rate": 0.0001484193548387097,
      "loss": 0.3276,
      "step": 1600
    },
    {
      "epoch": 25.76,
      "grad_norm": 2.811481475830078,
      "learning_rate": 0.0001480967741935484,
      "loss": 0.3337,
      "step": 1610
    },
    {
      "epoch": 25.92,
      "grad_norm": 2.320997953414917,
      "learning_rate": 0.0001477741935483871,
      "loss": 0.3257,
      "step": 1620
    },
    {
      "epoch": 26.08,
      "grad_norm": 2.6050562858581543,
      "learning_rate": 0.00014745161290322583,
      "loss": 0.3058,
      "step": 1630
    },
    {
      "epoch": 26.24,
      "grad_norm": 2.153953790664673,
      "learning_rate": 0.0001471290322580645,
      "loss": 0.2795,
      "step": 1640
    },
    {
      "epoch": 26.4,
      "grad_norm": 2.5632269382476807,
      "learning_rate": 0.00014680645161290322,
      "loss": 0.284,
      "step": 1650
    },
    {
      "epoch": 26.56,
      "grad_norm": 2.355788230895996,
      "learning_rate": 0.00014648387096774195,
      "loss": 0.2955,
      "step": 1660
    },
    {
      "epoch": 26.72,
      "grad_norm": 2.563262462615967,
      "learning_rate": 0.00014616129032258066,
      "loss": 0.2965,
      "step": 1670
    },
    {
      "epoch": 26.88,
      "grad_norm": 2.3319389820098877,
      "learning_rate": 0.00014583870967741936,
      "loss": 0.3023,
      "step": 1680
    },
    {
      "epoch": 27.04,
      "grad_norm": 1.6385563611984253,
      "learning_rate": 0.00014551612903225807,
      "loss": 0.295,
      "step": 1690
    },
    {
      "epoch": 27.2,
      "grad_norm": 2.8859646320343018,
      "learning_rate": 0.00014519354838709678,
      "loss": 0.2576,
      "step": 1700
    },
    {
      "epoch": 27.36,
      "grad_norm": 2.1979823112487793,
      "learning_rate": 0.00014487096774193548,
      "loss": 0.2686,
      "step": 1710
    },
    {
      "epoch": 27.52,
      "grad_norm": 2.4276225566864014,
      "learning_rate": 0.00014454838709677422,
      "loss": 0.2667,
      "step": 1720
    },
    {
      "epoch": 27.68,
      "grad_norm": 2.81669282913208,
      "learning_rate": 0.0001442258064516129,
      "loss": 0.2726,
      "step": 1730
    },
    {
      "epoch": 27.84,
      "grad_norm": 3.1404976844787598,
      "learning_rate": 0.0001439032258064516,
      "loss": 0.2855,
      "step": 1740
    },
    {
      "epoch": 28.0,
      "grad_norm": 3.198063373565674,
      "learning_rate": 0.00014358064516129034,
      "loss": 0.2911,
      "step": 1750
    },
    {
      "epoch": 28.16,
      "grad_norm": 2.2350854873657227,
      "learning_rate": 0.00014325806451612904,
      "loss": 0.2378,
      "step": 1760
    },
    {
      "epoch": 28.32,
      "grad_norm": 2.4164915084838867,
      "learning_rate": 0.00014293548387096775,
      "loss": 0.2397,
      "step": 1770
    },
    {
      "epoch": 28.48,
      "grad_norm": 2.8860878944396973,
      "learning_rate": 0.00014261290322580645,
      "loss": 0.2505,
      "step": 1780
    },
    {
      "epoch": 28.64,
      "grad_norm": 2.7349557876586914,
      "learning_rate": 0.00014229032258064516,
      "loss": 0.262,
      "step": 1790
    },
    {
      "epoch": 28.8,
      "grad_norm": 2.604602813720703,
      "learning_rate": 0.00014196774193548387,
      "loss": 0.2639,
      "step": 1800
    },
    {
      "epoch": 28.96,
      "grad_norm": 2.5355823040008545,
      "learning_rate": 0.0001416451612903226,
      "loss": 0.2666,
      "step": 1810
    },
    {
      "epoch": 29.12,
      "grad_norm": 2.2154204845428467,
      "learning_rate": 0.0001413225806451613,
      "loss": 0.233,
      "step": 1820
    },
    {
      "epoch": 29.28,
      "grad_norm": 2.8737149238586426,
      "learning_rate": 0.000141,
      "loss": 0.2245,
      "step": 1830
    },
    {
      "epoch": 29.44,
      "grad_norm": 2.359257221221924,
      "learning_rate": 0.00014067741935483872,
      "loss": 0.2314,
      "step": 1840
    },
    {
      "epoch": 29.6,
      "grad_norm": 2.4809696674346924,
      "learning_rate": 0.00014035483870967742,
      "loss": 0.24,
      "step": 1850
    },
    {
      "epoch": 29.76,
      "grad_norm": 2.770080804824829,
      "learning_rate": 0.00014003225806451613,
      "loss": 0.2471,
      "step": 1860
    },
    {
      "epoch": 29.92,
      "grad_norm": 2.3980712890625,
      "learning_rate": 0.00013970967741935486,
      "loss": 0.2459,
      "step": 1870
    },
    {
      "epoch": 30.08,
      "grad_norm": 2.22361421585083,
      "learning_rate": 0.00013938709677419354,
      "loss": 0.2293,
      "step": 1880
    },
    {
      "epoch": 30.24,
      "grad_norm": 2.23351788520813,
      "learning_rate": 0.00013906451612903228,
      "loss": 0.2161,
      "step": 1890
    },
    {
      "epoch": 30.4,
      "grad_norm": 2.324384927749634,
      "learning_rate": 0.00013874193548387098,
      "loss": 0.2181,
      "step": 1900
    },
    {
      "epoch": 30.56,
      "grad_norm": 2.4072265625,
      "learning_rate": 0.0001384193548387097,
      "loss": 0.2252,
      "step": 1910
    },
    {
      "epoch": 30.72,
      "grad_norm": 2.428682565689087,
      "learning_rate": 0.0001380967741935484,
      "loss": 0.2332,
      "step": 1920
    },
    {
      "epoch": 30.88,
      "grad_norm": 2.5889081954956055,
      "learning_rate": 0.0001377741935483871,
      "loss": 0.234,
      "step": 1930
    },
    {
      "epoch": 31.04,
      "grad_norm": 2.178375005722046,
      "learning_rate": 0.0001374516129032258,
      "loss": 0.2221,
      "step": 1940
    },
    {
      "epoch": 31.2,
      "grad_norm": 2.2690091133117676,
      "learning_rate": 0.00013712903225806451,
      "loss": 0.1978,
      "step": 1950
    },
    {
      "epoch": 31.36,
      "grad_norm": 2.445535659790039,
      "learning_rate": 0.00013680645161290325,
      "loss": 0.2075,
      "step": 1960
    },
    {
      "epoch": 31.52,
      "grad_norm": 2.297614097595215,
      "learning_rate": 0.00013648387096774193,
      "loss": 0.2105,
      "step": 1970
    },
    {
      "epoch": 31.68,
      "grad_norm": 2.456094980239868,
      "learning_rate": 0.00013616129032258066,
      "loss": 0.2183,
      "step": 1980
    },
    {
      "epoch": 31.84,
      "grad_norm": 2.9073026180267334,
      "learning_rate": 0.00013583870967741937,
      "loss": 0.2219,
      "step": 1990
    },
    {
      "epoch": 32.0,
      "grad_norm": 2.5645761489868164,
      "learning_rate": 0.00013551612903225807,
      "loss": 0.2244,
      "step": 2000
    },
    {
      "epoch": 32.16,
      "grad_norm": 1.9080466032028198,
      "learning_rate": 0.00013519354838709678,
      "loss": 0.1852,
      "step": 2010
    },
    {
      "epoch": 32.32,
      "grad_norm": 2.4245429039001465,
      "learning_rate": 0.00013487096774193548,
      "loss": 0.1941,
      "step": 2020
    },
    {
      "epoch": 32.48,
      "grad_norm": 2.873189687728882,
      "learning_rate": 0.0001345483870967742,
      "loss": 0.198,
      "step": 2030
    },
    {
      "epoch": 32.64,
      "grad_norm": 3.123337984085083,
      "learning_rate": 0.00013422580645161292,
      "loss": 0.203,
      "step": 2040
    },
    {
      "epoch": 32.8,
      "grad_norm": 2.592850923538208,
      "learning_rate": 0.00013390322580645163,
      "loss": 0.2095,
      "step": 2050
    },
    {
      "epoch": 32.96,
      "grad_norm": 2.4933552742004395,
      "learning_rate": 0.0001335806451612903,
      "loss": 0.2053,
      "step": 2060
    },
    {
      "epoch": 33.12,
      "grad_norm": 1.862947940826416,
      "learning_rate": 0.00013325806451612904,
      "loss": 0.1886,
      "step": 2070
    },
    {
      "epoch": 33.28,
      "grad_norm": 2.4500813484191895,
      "learning_rate": 0.00013293548387096775,
      "loss": 0.1834,
      "step": 2080
    },
    {
      "epoch": 33.44,
      "grad_norm": 2.4213106632232666,
      "learning_rate": 0.00013261290322580645,
      "loss": 0.1916,
      "step": 2090
    },
    {
      "epoch": 33.6,
      "grad_norm": 2.4137024879455566,
      "learning_rate": 0.0001322903225806452,
      "loss": 0.1889,
      "step": 2100
    },
    {
      "epoch": 33.76,
      "grad_norm": 1.9225903749465942,
      "learning_rate": 0.00013196774193548387,
      "loss": 0.1932,
      "step": 2110
    },
    {
      "epoch": 33.92,
      "grad_norm": 2.627751350402832,
      "learning_rate": 0.00013164516129032257,
      "loss": 0.1994,
      "step": 2120
    },
    {
      "epoch": 34.08,
      "grad_norm": 2.1216776371002197,
      "learning_rate": 0.0001313225806451613,
      "loss": 0.1854,
      "step": 2130
    },
    {
      "epoch": 34.24,
      "grad_norm": 1.79988431930542,
      "learning_rate": 0.000131,
      "loss": 0.1764,
      "step": 2140
    },
    {
      "epoch": 34.4,
      "grad_norm": 2.725584030151367,
      "learning_rate": 0.00013067741935483872,
      "loss": 0.1724,
      "step": 2150
    },
    {
      "epoch": 34.56,
      "grad_norm": 3.0872607231140137,
      "learning_rate": 0.00013035483870967743,
      "loss": 0.1864,
      "step": 2160
    },
    {
      "epoch": 34.72,
      "grad_norm": 2.670351982116699,
      "learning_rate": 0.00013003225806451613,
      "loss": 0.189,
      "step": 2170
    },
    {
      "epoch": 34.88,
      "grad_norm": 2.467918872833252,
      "learning_rate": 0.00012970967741935484,
      "loss": 0.1892,
      "step": 2180
    },
    {
      "epoch": 35.04,
      "grad_norm": 1.591225504875183,
      "learning_rate": 0.00012938709677419357,
      "loss": 0.1862,
      "step": 2190
    },
    {
      "epoch": 35.2,
      "grad_norm": 2.1949949264526367,
      "learning_rate": 0.00012906451612903225,
      "loss": 0.1594,
      "step": 2200
    },
    {
      "epoch": 35.36,
      "grad_norm": 2.2106893062591553,
      "learning_rate": 0.00012874193548387096,
      "loss": 0.1699,
      "step": 2210
    },
    {
      "epoch": 35.52,
      "grad_norm": 1.835793375968933,
      "learning_rate": 0.0001284193548387097,
      "loss": 0.1699,
      "step": 2220
    },
    {
      "epoch": 35.68,
      "grad_norm": 1.770975947380066,
      "learning_rate": 0.0001280967741935484,
      "loss": 0.1786,
      "step": 2230
    },
    {
      "epoch": 35.84,
      "grad_norm": 2.7737338542938232,
      "learning_rate": 0.0001277741935483871,
      "loss": 0.1799,
      "step": 2240
    },
    {
      "epoch": 36.0,
      "grad_norm": 2.317680835723877,
      "learning_rate": 0.0001274516129032258,
      "loss": 0.1807,
      "step": 2250
    },
    {
      "epoch": 36.16,
      "grad_norm": 2.137967586517334,
      "learning_rate": 0.00012712903225806451,
      "loss": 0.155,
      "step": 2260
    },
    {
      "epoch": 36.32,
      "grad_norm": 1.764809012413025,
      "learning_rate": 0.00012680645161290322,
      "loss": 0.1602,
      "step": 2270
    },
    {
      "epoch": 36.48,
      "grad_norm": 2.4306418895721436,
      "learning_rate": 0.00012648387096774195,
      "loss": 0.1644,
      "step": 2280
    },
    {
      "epoch": 36.64,
      "grad_norm": 1.7215269804000854,
      "learning_rate": 0.00012616129032258066,
      "loss": 0.1675,
      "step": 2290
    },
    {
      "epoch": 36.8,
      "grad_norm": 2.1169686317443848,
      "learning_rate": 0.00012583870967741937,
      "loss": 0.169,
      "step": 2300
    },
    {
      "epoch": 36.96,
      "grad_norm": 2.576965093612671,
      "learning_rate": 0.00012551612903225807,
      "loss": 0.1759,
      "step": 2310
    },
    {
      "epoch": 37.12,
      "grad_norm": 2.349083185195923,
      "learning_rate": 0.00012519354838709678,
      "loss": 0.1569,
      "step": 2320
    },
    {
      "epoch": 37.28,
      "grad_norm": 1.698199987411499,
      "learning_rate": 0.00012487096774193549,
      "loss": 0.1523,
      "step": 2330
    },
    {
      "epoch": 37.44,
      "grad_norm": 2.2408599853515625,
      "learning_rate": 0.00012454838709677422,
      "loss": 0.1579,
      "step": 2340
    },
    {
      "epoch": 37.6,
      "grad_norm": 1.7537211179733276,
      "learning_rate": 0.0001242258064516129,
      "loss": 0.1613,
      "step": 2350
    },
    {
      "epoch": 37.76,
      "grad_norm": 2.2816600799560547,
      "learning_rate": 0.0001239032258064516,
      "loss": 0.1604,
      "step": 2360
    },
    {
      "epoch": 37.92,
      "grad_norm": 2.135903835296631,
      "learning_rate": 0.00012358064516129034,
      "loss": 0.1666,
      "step": 2370
    },
    {
      "epoch": 38.08,
      "grad_norm": 1.4444613456726074,
      "learning_rate": 0.00012325806451612904,
      "loss": 0.1574,
      "step": 2380
    },
    {
      "epoch": 38.24,
      "grad_norm": 2.1274943351745605,
      "learning_rate": 0.00012293548387096775,
      "loss": 0.1497,
      "step": 2390
    },
    {
      "epoch": 38.4,
      "grad_norm": 2.0582194328308105,
      "learning_rate": 0.00012261290322580646,
      "loss": 0.1502,
      "step": 2400
    },
    {
      "epoch": 38.56,
      "grad_norm": 1.99543035030365,
      "learning_rate": 0.00012229032258064516,
      "loss": 0.1527,
      "step": 2410
    },
    {
      "epoch": 38.72,
      "grad_norm": 2.4016642570495605,
      "learning_rate": 0.00012196774193548388,
      "loss": 0.1579,
      "step": 2420
    },
    {
      "epoch": 38.88,
      "grad_norm": 1.889532446861267,
      "learning_rate": 0.0001216451612903226,
      "loss": 0.1581,
      "step": 2430
    },
    {
      "epoch": 39.04,
      "grad_norm": 1.5487650632858276,
      "learning_rate": 0.0001213225806451613,
      "loss": 0.1557,
      "step": 2440
    },
    {
      "epoch": 39.2,
      "grad_norm": 1.7775896787643433,
      "learning_rate": 0.000121,
      "loss": 0.1433,
      "step": 2450
    },
    {
      "epoch": 39.36,
      "grad_norm": 1.629921555519104,
      "learning_rate": 0.00012067741935483872,
      "loss": 0.1477,
      "step": 2460
    },
    {
      "epoch": 39.52,
      "grad_norm": 1.9216187000274658,
      "learning_rate": 0.00012035483870967743,
      "loss": 0.1481,
      "step": 2470
    },
    {
      "epoch": 39.68,
      "grad_norm": 2.4521870613098145,
      "learning_rate": 0.00012003225806451615,
      "loss": 0.1548,
      "step": 2480
    },
    {
      "epoch": 39.84,
      "grad_norm": 2.893139600753784,
      "learning_rate": 0.00011970967741935484,
      "loss": 0.1542,
      "step": 2490
    },
    {
      "epoch": 40.0,
      "grad_norm": 1.7710884809494019,
      "learning_rate": 0.00011938709677419356,
      "loss": 0.1539,
      "step": 2500
    },
    {
      "epoch": 40.16,
      "grad_norm": 2.467780590057373,
      "learning_rate": 0.00011906451612903226,
      "loss": 0.1384,
      "step": 2510
    },
    {
      "epoch": 40.32,
      "grad_norm": 1.692098617553711,
      "learning_rate": 0.00011874193548387098,
      "loss": 0.1398,
      "step": 2520
    },
    {
      "epoch": 40.48,
      "grad_norm": 2.4998574256896973,
      "learning_rate": 0.00011841935483870968,
      "loss": 0.141,
      "step": 2530
    },
    {
      "epoch": 40.64,
      "grad_norm": 2.455094575881958,
      "learning_rate": 0.00011809677419354838,
      "loss": 0.1472,
      "step": 2540
    },
    {
      "epoch": 40.8,
      "grad_norm": 2.2501373291015625,
      "learning_rate": 0.0001177741935483871,
      "loss": 0.1478,
      "step": 2550
    },
    {
      "epoch": 40.96,
      "grad_norm": 1.7352842092514038,
      "learning_rate": 0.00011745161290322582,
      "loss": 0.1523,
      "step": 2560
    },
    {
      "epoch": 41.12,
      "grad_norm": 1.6160428524017334,
      "learning_rate": 0.00011712903225806453,
      "loss": 0.1383,
      "step": 2570
    },
    {
      "epoch": 41.28,
      "grad_norm": 1.7578778266906738,
      "learning_rate": 0.00011680645161290322,
      "loss": 0.1344,
      "step": 2580
    },
    {
      "epoch": 41.44,
      "grad_norm": 1.668540120124817,
      "learning_rate": 0.00011648387096774194,
      "loss": 0.137,
      "step": 2590
    },
    {
      "epoch": 41.6,
      "grad_norm": 1.9453370571136475,
      "learning_rate": 0.00011616129032258065,
      "loss": 0.1445,
      "step": 2600
    },
    {
      "epoch": 41.76,
      "grad_norm": 2.1329894065856934,
      "learning_rate": 0.00011583870967741937,
      "loss": 0.1401,
      "step": 2610
    },
    {
      "epoch": 41.92,
      "grad_norm": 2.9534623622894287,
      "learning_rate": 0.00011551612903225807,
      "loss": 0.1451,
      "step": 2620
    },
    {
      "epoch": 42.08,
      "grad_norm": 1.3531379699707031,
      "learning_rate": 0.00011519354838709677,
      "loss": 0.1375,
      "step": 2630
    },
    {
      "epoch": 42.24,
      "grad_norm": 2.283677816390991,
      "learning_rate": 0.00011487096774193549,
      "loss": 0.1307,
      "step": 2640
    },
    {
      "epoch": 42.4,
      "grad_norm": 1.6229647397994995,
      "learning_rate": 0.0001145483870967742,
      "loss": 0.1313,
      "step": 2650
    },
    {
      "epoch": 42.56,
      "grad_norm": 1.5721402168273926,
      "learning_rate": 0.00011422580645161291,
      "loss": 0.1322,
      "step": 2660
    },
    {
      "epoch": 42.72,
      "grad_norm": 1.6071618795394897,
      "learning_rate": 0.0001139032258064516,
      "loss": 0.1376,
      "step": 2670
    },
    {
      "epoch": 42.88,
      "grad_norm": 1.4502424001693726,
      "learning_rate": 0.00011358064516129032,
      "loss": 0.1368,
      "step": 2680
    },
    {
      "epoch": 43.04,
      "grad_norm": 1.2102515697479248,
      "learning_rate": 0.00011325806451612903,
      "loss": 0.1363,
      "step": 2690
    },
    {
      "epoch": 43.2,
      "grad_norm": 1.4724687337875366,
      "learning_rate": 0.00011293548387096775,
      "loss": 0.1241,
      "step": 2700
    },
    {
      "epoch": 43.36,
      "grad_norm": 2.126573085784912,
      "learning_rate": 0.00011261290322580647,
      "loss": 0.1281,
      "step": 2710
    },
    {
      "epoch": 43.52,
      "grad_norm": 2.0378658771514893,
      "learning_rate": 0.00011229032258064516,
      "loss": 0.1352,
      "step": 2720
    },
    {
      "epoch": 43.68,
      "grad_norm": 1.978238821029663,
      "learning_rate": 0.00011196774193548387,
      "loss": 0.1357,
      "step": 2730
    },
    {
      "epoch": 43.84,
      "grad_norm": 1.5167338848114014,
      "learning_rate": 0.00011164516129032259,
      "loss": 0.1378,
      "step": 2740
    },
    {
      "epoch": 44.0,
      "grad_norm": 2.6229848861694336,
      "learning_rate": 0.0001113225806451613,
      "loss": 0.1368,
      "step": 2750
    },
    {
      "epoch": 44.16,
      "grad_norm": 2.586956024169922,
      "learning_rate": 0.00011100000000000001,
      "loss": 0.1215,
      "step": 2760
    },
    {
      "epoch": 44.32,
      "grad_norm": 1.5434893369674683,
      "learning_rate": 0.00011067741935483871,
      "loss": 0.122,
      "step": 2770
    },
    {
      "epoch": 44.48,
      "grad_norm": 1.6991748809814453,
      "learning_rate": 0.00011035483870967743,
      "loss": 0.1288,
      "step": 2780
    },
    {
      "epoch": 44.64,
      "grad_norm": 1.6747218370437622,
      "learning_rate": 0.00011003225806451613,
      "loss": 0.1298,
      "step": 2790
    },
    {
      "epoch": 44.8,
      "grad_norm": 1.8977768421173096,
      "learning_rate": 0.00010970967741935485,
      "loss": 0.1274,
      "step": 2800
    },
    {
      "epoch": 44.96,
      "grad_norm": 2.1921539306640625,
      "learning_rate": 0.00010938709677419355,
      "loss": 0.1328,
      "step": 2810
    },
    {
      "epoch": 45.12,
      "grad_norm": 2.201087474822998,
      "learning_rate": 0.00010906451612903225,
      "loss": 0.1199,
      "step": 2820
    },
    {
      "epoch": 45.28,
      "grad_norm": 1.6816660165786743,
      "learning_rate": 0.00010874193548387097,
      "loss": 0.1214,
      "step": 2830
    },
    {
      "epoch": 45.44,
      "grad_norm": 1.0567231178283691,
      "learning_rate": 0.00010841935483870969,
      "loss": 0.1234,
      "step": 2840
    },
    {
      "epoch": 45.6,
      "grad_norm": 1.3555560111999512,
      "learning_rate": 0.0001080967741935484,
      "loss": 0.1236,
      "step": 2850
    },
    {
      "epoch": 45.76,
      "grad_norm": 2.097130537033081,
      "learning_rate": 0.00010777419354838709,
      "loss": 0.1275,
      "step": 2860
    },
    {
      "epoch": 45.92,
      "grad_norm": 2.127514123916626,
      "learning_rate": 0.00010745161290322581,
      "loss": 0.1286,
      "step": 2870
    },
    {
      "epoch": 46.08,
      "grad_norm": 1.4086848497390747,
      "learning_rate": 0.00010712903225806452,
      "loss": 0.1203,
      "step": 2880
    },
    {
      "epoch": 46.24,
      "grad_norm": 1.6010750532150269,
      "learning_rate": 0.00010680645161290324,
      "loss": 0.1175,
      "step": 2890
    },
    {
      "epoch": 46.4,
      "grad_norm": 1.5432184934616089,
      "learning_rate": 0.00010648387096774196,
      "loss": 0.1163,
      "step": 2900
    },
    {
      "epoch": 46.56,
      "grad_norm": 3.0002214908599854,
      "learning_rate": 0.00010616129032258065,
      "loss": 0.1257,
      "step": 2910
    },
    {
      "epoch": 46.72,
      "grad_norm": 1.6969698667526245,
      "learning_rate": 0.00010583870967741935,
      "loss": 0.1207,
      "step": 2920
    },
    {
      "epoch": 46.88,
      "grad_norm": 1.35691237449646,
      "learning_rate": 0.00010551612903225807,
      "loss": 0.1257,
      "step": 2930
    },
    {
      "epoch": 47.04,
      "grad_norm": 1.2951841354370117,
      "learning_rate": 0.00010519354838709678,
      "loss": 0.1226,
      "step": 2940
    },
    {
      "epoch": 47.2,
      "grad_norm": 1.980775237083435,
      "learning_rate": 0.0001048709677419355,
      "loss": 0.1121,
      "step": 2950
    },
    {
      "epoch": 47.36,
      "grad_norm": 1.7462553977966309,
      "learning_rate": 0.00010454838709677419,
      "loss": 0.1193,
      "step": 2960
    },
    {
      "epoch": 47.52,
      "grad_norm": 1.3529144525527954,
      "learning_rate": 0.00010422580645161291,
      "loss": 0.115,
      "step": 2970
    },
    {
      "epoch": 47.68,
      "grad_norm": 1.7577213048934937,
      "learning_rate": 0.00010390322580645162,
      "loss": 0.1201,
      "step": 2980
    },
    {
      "epoch": 47.84,
      "grad_norm": 1.2776310443878174,
      "learning_rate": 0.00010358064516129034,
      "loss": 0.1198,
      "step": 2990
    },
    {
      "epoch": 48.0,
      "grad_norm": 1.3981118202209473,
      "learning_rate": 0.00010325806451612903,
      "loss": 0.1204,
      "step": 3000
    },
    {
      "epoch": 48.16,
      "grad_norm": 1.208207130432129,
      "learning_rate": 0.00010293548387096774,
      "loss": 0.1063,
      "step": 3010
    },
    {
      "epoch": 48.32,
      "grad_norm": 1.3536474704742432,
      "learning_rate": 0.00010261290322580646,
      "loss": 0.1123,
      "step": 3020
    },
    {
      "epoch": 48.48,
      "grad_norm": 1.3631086349487305,
      "learning_rate": 0.00010229032258064516,
      "loss": 0.115,
      "step": 3030
    },
    {
      "epoch": 48.64,
      "grad_norm": 1.2178473472595215,
      "learning_rate": 0.00010196774193548388,
      "loss": 0.1143,
      "step": 3040
    },
    {
      "epoch": 48.8,
      "grad_norm": 2.1066434383392334,
      "learning_rate": 0.00010164516129032258,
      "loss": 0.1234,
      "step": 3050
    },
    {
      "epoch": 48.96,
      "grad_norm": 1.818852186203003,
      "learning_rate": 0.0001013225806451613,
      "loss": 0.1242,
      "step": 3060
    },
    {
      "epoch": 49.12,
      "grad_norm": 1.2515507936477661,
      "learning_rate": 0.000101,
      "loss": 0.1118,
      "step": 3070
    },
    {
      "epoch": 49.28,
      "grad_norm": 1.3625049591064453,
      "learning_rate": 0.00010067741935483872,
      "loss": 0.1097,
      "step": 3080
    },
    {
      "epoch": 49.44,
      "grad_norm": 1.7623980045318604,
      "learning_rate": 0.00010035483870967743,
      "loss": 0.1126,
      "step": 3090
    },
    {
      "epoch": 49.6,
      "grad_norm": 1.3493503332138062,
      "learning_rate": 0.00010003225806451612,
      "loss": 0.1139,
      "step": 3100
    },
    {
      "epoch": 49.76,
      "grad_norm": 1.5169587135314941,
      "learning_rate": 9.970967741935484e-05,
      "loss": 0.1154,
      "step": 3110
    },
    {
      "epoch": 49.92,
      "grad_norm": 2.1966283321380615,
      "learning_rate": 9.938709677419356e-05,
      "loss": 0.1171,
      "step": 3120
    },
    {
      "epoch": 50.08,
      "grad_norm": 1.6833268404006958,
      "learning_rate": 9.906451612903225e-05,
      "loss": 0.113,
      "step": 3130
    },
    {
      "epoch": 50.24,
      "grad_norm": 1.4803876876831055,
      "learning_rate": 9.874193548387097e-05,
      "loss": 0.1054,
      "step": 3140
    },
    {
      "epoch": 50.4,
      "grad_norm": 1.3878694772720337,
      "learning_rate": 9.841935483870969e-05,
      "loss": 0.1079,
      "step": 3150
    },
    {
      "epoch": 50.56,
      "grad_norm": 1.64521062374115,
      "learning_rate": 9.809677419354838e-05,
      "loss": 0.1113,
      "step": 3160
    },
    {
      "epoch": 50.72,
      "grad_norm": 1.4510949850082397,
      "learning_rate": 9.77741935483871e-05,
      "loss": 0.1147,
      "step": 3170
    },
    {
      "epoch": 50.88,
      "grad_norm": 1.4455626010894775,
      "learning_rate": 9.745161290322581e-05,
      "loss": 0.1137,
      "step": 3180
    },
    {
      "epoch": 51.04,
      "grad_norm": 1.1447566747665405,
      "learning_rate": 9.712903225806452e-05,
      "loss": 0.1111,
      "step": 3190
    },
    {
      "epoch": 51.2,
      "grad_norm": 1.1405003070831299,
      "learning_rate": 9.680645161290322e-05,
      "loss": 0.1049,
      "step": 3200
    },
    {
      "epoch": 51.36,
      "grad_norm": 1.3528752326965332,
      "learning_rate": 9.648387096774194e-05,
      "loss": 0.1081,
      "step": 3210
    },
    {
      "epoch": 51.52,
      "grad_norm": 1.6244195699691772,
      "learning_rate": 9.616129032258065e-05,
      "loss": 0.1083,
      "step": 3220
    },
    {
      "epoch": 51.68,
      "grad_norm": 1.4993714094161987,
      "learning_rate": 9.583870967741936e-05,
      "loss": 0.1082,
      "step": 3230
    },
    {
      "epoch": 51.84,
      "grad_norm": 1.1913836002349854,
      "learning_rate": 9.551612903225808e-05,
      "loss": 0.1101,
      "step": 3240
    },
    {
      "epoch": 52.0,
      "grad_norm": 1.444038987159729,
      "learning_rate": 9.519354838709678e-05,
      "loss": 0.1136,
      "step": 3250
    },
    {
      "epoch": 52.16,
      "grad_norm": 1.5950225591659546,
      "learning_rate": 9.487096774193549e-05,
      "loss": 0.1015,
      "step": 3260
    },
    {
      "epoch": 52.32,
      "grad_norm": 1.5138227939605713,
      "learning_rate": 9.45483870967742e-05,
      "loss": 0.1044,
      "step": 3270
    },
    {
      "epoch": 52.48,
      "grad_norm": 1.7190989255905151,
      "learning_rate": 9.422580645161291e-05,
      "loss": 0.1049,
      "step": 3280
    },
    {
      "epoch": 52.64,
      "grad_norm": 2.0404067039489746,
      "learning_rate": 9.390322580645162e-05,
      "loss": 0.1084,
      "step": 3290
    },
    {
      "epoch": 52.8,
      "grad_norm": 0.8778730630874634,
      "learning_rate": 9.358064516129033e-05,
      "loss": 0.1089,
      "step": 3300
    },
    {
      "epoch": 52.96,
      "grad_norm": 1.302156686782837,
      "learning_rate": 9.325806451612905e-05,
      "loss": 0.1064,
      "step": 3310
    },
    {
      "epoch": 53.12,
      "grad_norm": 1.0487626791000366,
      "learning_rate": 9.293548387096774e-05,
      "loss": 0.1041,
      "step": 3320
    },
    {
      "epoch": 53.28,
      "grad_norm": 1.5156755447387695,
      "learning_rate": 9.261290322580646e-05,
      "loss": 0.1025,
      "step": 3330
    },
    {
      "epoch": 53.44,
      "grad_norm": 1.1233017444610596,
      "learning_rate": 9.229032258064516e-05,
      "loss": 0.1061,
      "step": 3340
    },
    {
      "epoch": 53.6,
      "grad_norm": 1.3437849283218384,
      "learning_rate": 9.196774193548387e-05,
      "loss": 0.1072,
      "step": 3350
    },
    {
      "epoch": 53.76,
      "grad_norm": 1.4854118824005127,
      "learning_rate": 9.164516129032259e-05,
      "loss": 0.1047,
      "step": 3360
    },
    {
      "epoch": 53.92,
      "grad_norm": 1.4864035844802856,
      "learning_rate": 9.13225806451613e-05,
      "loss": 0.1068,
      "step": 3370
    },
    {
      "epoch": 54.08,
      "grad_norm": 1.0401661396026611,
      "learning_rate": 9.1e-05,
      "loss": 0.1034,
      "step": 3380
    },
    {
      "epoch": 54.24,
      "grad_norm": 0.8670012354850769,
      "learning_rate": 9.067741935483871e-05,
      "loss": 0.0974,
      "step": 3390
    },
    {
      "epoch": 54.4,
      "grad_norm": 1.1909117698669434,
      "learning_rate": 9.035483870967743e-05,
      "loss": 0.1007,
      "step": 3400
    },
    {
      "epoch": 54.56,
      "grad_norm": 1.3757662773132324,
      "learning_rate": 9.003225806451614e-05,
      "loss": 0.1038,
      "step": 3410
    },
    {
      "epoch": 54.72,
      "grad_norm": 0.9391764402389526,
      "learning_rate": 8.970967741935484e-05,
      "loss": 0.1056,
      "step": 3420
    },
    {
      "epoch": 54.88,
      "grad_norm": 1.0170888900756836,
      "learning_rate": 8.938709677419356e-05,
      "loss": 0.1023,
      "step": 3430
    },
    {
      "epoch": 55.04,
      "grad_norm": 1.0093938112258911,
      "learning_rate": 8.906451612903227e-05,
      "loss": 0.1058,
      "step": 3440
    },
    {
      "epoch": 55.2,
      "grad_norm": 1.8801138401031494,
      "learning_rate": 8.874193548387097e-05,
      "loss": 0.099,
      "step": 3450
    },
    {
      "epoch": 55.36,
      "grad_norm": 1.5623430013656616,
      "learning_rate": 8.841935483870968e-05,
      "loss": 0.1046,
      "step": 3460
    },
    {
      "epoch": 55.52,
      "grad_norm": 1.7789967060089111,
      "learning_rate": 8.809677419354839e-05,
      "loss": 0.1038,
      "step": 3470
    },
    {
      "epoch": 55.68,
      "grad_norm": 1.797473669052124,
      "learning_rate": 8.777419354838709e-05,
      "loss": 0.1062,
      "step": 3480
    },
    {
      "epoch": 55.84,
      "grad_norm": 2.0963029861450195,
      "learning_rate": 8.745161290322581e-05,
      "loss": 0.104,
      "step": 3490
    },
    {
      "epoch": 56.0,
      "grad_norm": 1.711259126663208,
      "learning_rate": 8.712903225806452e-05,
      "loss": 0.107,
      "step": 3500
    },
    {
      "epoch": 56.16,
      "grad_norm": 1.2828420400619507,
      "learning_rate": 8.680645161290322e-05,
      "loss": 0.0967,
      "step": 3510
    },
    {
      "epoch": 56.32,
      "grad_norm": 1.2137057781219482,
      "learning_rate": 8.648387096774194e-05,
      "loss": 0.0971,
      "step": 3520
    },
    {
      "epoch": 56.48,
      "grad_norm": 0.9034223556518555,
      "learning_rate": 8.616129032258065e-05,
      "loss": 0.1006,
      "step": 3530
    },
    {
      "epoch": 56.64,
      "grad_norm": 1.3555456399917603,
      "learning_rate": 8.583870967741936e-05,
      "loss": 0.1018,
      "step": 3540
    },
    {
      "epoch": 56.8,
      "grad_norm": 1.7107219696044922,
      "learning_rate": 8.551612903225806e-05,
      "loss": 0.1041,
      "step": 3550
    },
    {
      "epoch": 56.96,
      "grad_norm": 1.4730465412139893,
      "learning_rate": 8.519354838709678e-05,
      "loss": 0.1051,
      "step": 3560
    },
    {
      "epoch": 57.12,
      "grad_norm": 1.1063456535339355,
      "learning_rate": 8.487096774193549e-05,
      "loss": 0.0968,
      "step": 3570
    },
    {
      "epoch": 57.28,
      "grad_norm": 0.9461372494697571,
      "learning_rate": 8.45483870967742e-05,
      "loss": 0.0958,
      "step": 3580
    },
    {
      "epoch": 57.44,
      "grad_norm": 0.8849135041236877,
      "learning_rate": 8.422580645161291e-05,
      "loss": 0.0984,
      "step": 3590
    },
    {
      "epoch": 57.6,
      "grad_norm": 0.8585776090621948,
      "learning_rate": 8.390322580645161e-05,
      "loss": 0.1,
      "step": 3600
    },
    {
      "epoch": 57.76,
      "grad_norm": 1.3201136589050293,
      "learning_rate": 8.358064516129033e-05,
      "loss": 0.1013,
      "step": 3610
    },
    {
      "epoch": 57.92,
      "grad_norm": 0.9458446502685547,
      "learning_rate": 8.325806451612905e-05,
      "loss": 0.1042,
      "step": 3620
    },
    {
      "epoch": 58.08,
      "grad_norm": 1.1642829179763794,
      "learning_rate": 8.293548387096774e-05,
      "loss": 0.0974,
      "step": 3630
    },
    {
      "epoch": 58.24,
      "grad_norm": 0.7290544509887695,
      "learning_rate": 8.261290322580646e-05,
      "loss": 0.0947,
      "step": 3640
    },
    {
      "epoch": 58.4,
      "grad_norm": 1.4136425256729126,
      "learning_rate": 8.229032258064517e-05,
      "loss": 0.0966,
      "step": 3650
    },
    {
      "epoch": 58.56,
      "grad_norm": 1.6335221529006958,
      "learning_rate": 8.196774193548387e-05,
      "loss": 0.0976,
      "step": 3660
    },
    {
      "epoch": 58.72,
      "grad_norm": 0.7871268391609192,
      "learning_rate": 8.164516129032258e-05,
      "loss": 0.0996,
      "step": 3670
    },
    {
      "epoch": 58.88,
      "grad_norm": 1.2082417011260986,
      "learning_rate": 8.13225806451613e-05,
      "loss": 0.1004,
      "step": 3680
    },
    {
      "epoch": 59.04,
      "grad_norm": 0.9343645572662354,
      "learning_rate": 8.1e-05,
      "loss": 0.0991,
      "step": 3690
    },
    {
      "epoch": 59.2,
      "grad_norm": 0.9338416457176208,
      "learning_rate": 8.067741935483871e-05,
      "loss": 0.0937,
      "step": 3700
    },
    {
      "epoch": 59.36,
      "grad_norm": 0.8510011434555054,
      "learning_rate": 8.035483870967743e-05,
      "loss": 0.0956,
      "step": 3710
    },
    {
      "epoch": 59.52,
      "grad_norm": 0.7901081442832947,
      "learning_rate": 8.003225806451614e-05,
      "loss": 0.0961,
      "step": 3720
    },
    {
      "epoch": 59.68,
      "grad_norm": 0.8202269673347473,
      "learning_rate": 7.970967741935484e-05,
      "loss": 0.0971,
      "step": 3730
    },
    {
      "epoch": 59.84,
      "grad_norm": 0.5913488268852234,
      "learning_rate": 7.938709677419355e-05,
      "loss": 0.0986,
      "step": 3740
    },
    {
      "epoch": 60.0,
      "grad_norm": 0.8020423650741577,
      "learning_rate": 7.906451612903227e-05,
      "loss": 0.1004,
      "step": 3750
    },
    {
      "epoch": 60.16,
      "grad_norm": 1.2948077917099,
      "learning_rate": 7.874193548387097e-05,
      "loss": 0.092,
      "step": 3760
    },
    {
      "epoch": 60.32,
      "grad_norm": 0.8374224305152893,
      "learning_rate": 7.841935483870968e-05,
      "loss": 0.0954,
      "step": 3770
    },
    {
      "epoch": 60.48,
      "grad_norm": 1.1318821907043457,
      "learning_rate": 7.80967741935484e-05,
      "loss": 0.0961,
      "step": 3780
    },
    {
      "epoch": 60.64,
      "grad_norm": 1.1860573291778564,
      "learning_rate": 7.777419354838709e-05,
      "loss": 0.0958,
      "step": 3790
    },
    {
      "epoch": 60.8,
      "grad_norm": 1.1659208536148071,
      "learning_rate": 7.745161290322581e-05,
      "loss": 0.0975,
      "step": 3800
    },
    {
      "epoch": 60.96,
      "grad_norm": 1.3710063695907593,
      "learning_rate": 7.712903225806452e-05,
      "loss": 0.0997,
      "step": 3810
    },
    {
      "epoch": 61.12,
      "grad_norm": 0.7192108035087585,
      "learning_rate": 7.680645161290323e-05,
      "loss": 0.0931,
      "step": 3820
    },
    {
      "epoch": 61.28,
      "grad_norm": 0.9875419735908508,
      "learning_rate": 7.648387096774194e-05,
      "loss": 0.0933,
      "step": 3830
    },
    {
      "epoch": 61.44,
      "grad_norm": 1.3378856182098389,
      "learning_rate": 7.616129032258065e-05,
      "loss": 0.0939,
      "step": 3840
    },
    {
      "epoch": 61.6,
      "grad_norm": 0.9775927662849426,
      "learning_rate": 7.583870967741936e-05,
      "loss": 0.0958,
      "step": 3850
    },
    {
      "epoch": 61.76,
      "grad_norm": 0.815967857837677,
      "learning_rate": 7.551612903225806e-05,
      "loss": 0.0971,
      "step": 3860
    },
    {
      "epoch": 61.92,
      "grad_norm": 0.711142897605896,
      "learning_rate": 7.519354838709678e-05,
      "loss": 0.098,
      "step": 3870
    },
    {
      "epoch": 62.08,
      "grad_norm": 0.757862389087677,
      "learning_rate": 7.487096774193548e-05,
      "loss": 0.0948,
      "step": 3880
    },
    {
      "epoch": 62.24,
      "grad_norm": 0.9134785532951355,
      "learning_rate": 7.45483870967742e-05,
      "loss": 0.0913,
      "step": 3890
    },
    {
      "epoch": 62.4,
      "grad_norm": 1.1586359739303589,
      "learning_rate": 7.422580645161292e-05,
      "loss": 0.0933,
      "step": 3900
    },
    {
      "epoch": 62.56,
      "grad_norm": 0.93504399061203,
      "learning_rate": 7.390322580645161e-05,
      "loss": 0.0946,
      "step": 3910
    },
    {
      "epoch": 62.72,
      "grad_norm": 1.2178099155426025,
      "learning_rate": 7.358064516129033e-05,
      "loss": 0.0972,
      "step": 3920
    },
    {
      "epoch": 62.88,
      "grad_norm": 1.491281270980835,
      "learning_rate": 7.325806451612903e-05,
      "loss": 0.095,
      "step": 3930
    },
    {
      "epoch": 63.04,
      "grad_norm": 0.789310097694397,
      "learning_rate": 7.293548387096774e-05,
      "loss": 0.0949,
      "step": 3940
    },
    {
      "epoch": 63.2,
      "grad_norm": 0.8426668643951416,
      "learning_rate": 7.261290322580645e-05,
      "loss": 0.0925,
      "step": 3950
    },
    {
      "epoch": 63.36,
      "grad_norm": 0.6679518818855286,
      "learning_rate": 7.229032258064517e-05,
      "loss": 0.0935,
      "step": 3960
    },
    {
      "epoch": 63.52,
      "grad_norm": 1.1688344478607178,
      "learning_rate": 7.196774193548387e-05,
      "loss": 0.095,
      "step": 3970
    },
    {
      "epoch": 63.68,
      "grad_norm": 0.9475545883178711,
      "learning_rate": 7.164516129032258e-05,
      "loss": 0.0954,
      "step": 3980
    },
    {
      "epoch": 63.84,
      "grad_norm": 0.7896549701690674,
      "learning_rate": 7.13225806451613e-05,
      "loss": 0.0973,
      "step": 3990
    },
    {
      "epoch": 64.0,
      "grad_norm": 0.9919887185096741,
      "learning_rate": 7.1e-05,
      "loss": 0.0984,
      "step": 4000
    },
    {
      "epoch": 64.16,
      "grad_norm": 0.8883840441703796,
      "learning_rate": 7.067741935483871e-05,
      "loss": 0.0909,
      "step": 4010
    },
    {
      "epoch": 64.32,
      "grad_norm": 1.3879891633987427,
      "learning_rate": 7.035483870967742e-05,
      "loss": 0.0924,
      "step": 4020
    },
    {
      "epoch": 64.48,
      "grad_norm": 0.6254050135612488,
      "learning_rate": 7.003225806451614e-05,
      "loss": 0.0937,
      "step": 4030
    },
    {
      "epoch": 64.64,
      "grad_norm": 0.8522732257843018,
      "learning_rate": 6.970967741935484e-05,
      "loss": 0.0944,
      "step": 4040
    },
    {
      "epoch": 64.8,
      "grad_norm": 0.8993662595748901,
      "learning_rate": 6.938709677419355e-05,
      "loss": 0.0946,
      "step": 4050
    },
    {
      "epoch": 64.96,
      "grad_norm": 0.6210134029388428,
      "learning_rate": 6.906451612903227e-05,
      "loss": 0.0961,
      "step": 4060
    },
    {
      "epoch": 65.12,
      "grad_norm": 1.4237314462661743,
      "learning_rate": 6.874193548387096e-05,
      "loss": 0.0915,
      "step": 4070
    },
    {
      "epoch": 65.28,
      "grad_norm": 0.590337336063385,
      "learning_rate": 6.841935483870968e-05,
      "loss": 0.0903,
      "step": 4080
    },
    {
      "epoch": 65.44,
      "grad_norm": 0.8954355120658875,
      "learning_rate": 6.809677419354839e-05,
      "loss": 0.0915,
      "step": 4090
    },
    {
      "epoch": 65.6,
      "grad_norm": 1.272484302520752,
      "learning_rate": 6.77741935483871e-05,
      "loss": 0.0919,
      "step": 4100
    },
    {
      "epoch": 65.76,
      "grad_norm": 0.5799235105514526,
      "learning_rate": 6.745161290322581e-05,
      "loss": 0.0953,
      "step": 4110
    },
    {
      "epoch": 65.92,
      "grad_norm": 0.5489311814308167,
      "learning_rate": 6.712903225806452e-05,
      "loss": 0.0944,
      "step": 4120
    },
    {
      "epoch": 66.08,
      "grad_norm": 0.454113632440567,
      "learning_rate": 6.680645161290323e-05,
      "loss": 0.0925,
      "step": 4130
    },
    {
      "epoch": 66.24,
      "grad_norm": 0.46685802936553955,
      "learning_rate": 6.648387096774193e-05,
      "loss": 0.09,
      "step": 4140
    },
    {
      "epoch": 66.4,
      "grad_norm": 0.5726438164710999,
      "learning_rate": 6.616129032258065e-05,
      "loss": 0.0921,
      "step": 4150
    },
    {
      "epoch": 66.56,
      "grad_norm": 0.9260037541389465,
      "learning_rate": 6.583870967741936e-05,
      "loss": 0.0937,
      "step": 4160
    },
    {
      "epoch": 66.72,
      "grad_norm": 0.6579046249389648,
      "learning_rate": 6.551612903225806e-05,
      "loss": 0.0927,
      "step": 4170
    },
    {
      "epoch": 66.88,
      "grad_norm": 1.529139518737793,
      "learning_rate": 6.519354838709678e-05,
      "loss": 0.0941,
      "step": 4180
    },
    {
      "epoch": 67.04,
      "grad_norm": 0.49119099974632263,
      "learning_rate": 6.487096774193549e-05,
      "loss": 0.0922,
      "step": 4190
    },
    {
      "epoch": 67.2,
      "grad_norm": 0.5991088151931763,
      "learning_rate": 6.45483870967742e-05,
      "loss": 0.0891,
      "step": 4200
    },
    {
      "epoch": 67.36,
      "grad_norm": 0.5796908140182495,
      "learning_rate": 6.42258064516129e-05,
      "loss": 0.0897,
      "step": 4210
    },
    {
      "epoch": 67.52,
      "grad_norm": 0.4511209726333618,
      "learning_rate": 6.390322580645162e-05,
      "loss": 0.0907,
      "step": 4220
    },
    {
      "epoch": 67.68,
      "grad_norm": 0.35779204964637756,
      "learning_rate": 6.358064516129033e-05,
      "loss": 0.0924,
      "step": 4230
    },
    {
      "epoch": 67.84,
      "grad_norm": 0.5029581189155579,
      "learning_rate": 6.325806451612903e-05,
      "loss": 0.093,
      "step": 4240
    },
    {
      "epoch": 68.0,
      "grad_norm": 0.5110875964164734,
      "learning_rate": 6.293548387096775e-05,
      "loss": 0.0937,
      "step": 4250
    },
    {
      "epoch": 68.16,
      "grad_norm": 0.390669584274292,
      "learning_rate": 6.261290322580645e-05,
      "loss": 0.0886,
      "step": 4260
    },
    {
      "epoch": 68.32,
      "grad_norm": 0.6021801233291626,
      "learning_rate": 6.229032258064517e-05,
      "loss": 0.0876,
      "step": 4270
    },
    {
      "epoch": 68.48,
      "grad_norm": 0.41805922985076904,
      "learning_rate": 6.196774193548387e-05,
      "loss": 0.0898,
      "step": 4280
    },
    {
      "epoch": 68.64,
      "grad_norm": 0.40193480253219604,
      "learning_rate": 6.164516129032258e-05,
      "loss": 0.0911,
      "step": 4290
    },
    {
      "epoch": 68.8,
      "grad_norm": 1.7792600393295288,
      "learning_rate": 6.13225806451613e-05,
      "loss": 0.0931,
      "step": 4300
    },
    {
      "epoch": 68.96,
      "grad_norm": 0.4636439383029938,
      "learning_rate": 6.1e-05,
      "loss": 0.0924,
      "step": 4310
    },
    {
      "epoch": 69.12,
      "grad_norm": 0.8066326975822449,
      "learning_rate": 6.067741935483872e-05,
      "loss": 0.0896,
      "step": 4320
    },
    {
      "epoch": 69.28,
      "grad_norm": 0.5064697265625,
      "learning_rate": 6.035483870967742e-05,
      "loss": 0.088,
      "step": 4330
    },
    {
      "epoch": 69.44,
      "grad_norm": 0.40276482701301575,
      "learning_rate": 6.003225806451613e-05,
      "loss": 0.0896,
      "step": 4340
    },
    {
      "epoch": 69.6,
      "grad_norm": 0.39066290855407715,
      "learning_rate": 5.970967741935484e-05,
      "loss": 0.0912,
      "step": 4350
    },
    {
      "epoch": 69.76,
      "grad_norm": 0.5445528030395508,
      "learning_rate": 5.938709677419355e-05,
      "loss": 0.0932,
      "step": 4360
    },
    {
      "epoch": 69.92,
      "grad_norm": 0.44867417216300964,
      "learning_rate": 5.906451612903226e-05,
      "loss": 0.0933,
      "step": 4370
    },
    {
      "epoch": 70.08,
      "grad_norm": 1.0833595991134644,
      "learning_rate": 5.874193548387097e-05,
      "loss": 0.0907,
      "step": 4380
    },
    {
      "epoch": 70.24,
      "grad_norm": 0.5424122214317322,
      "learning_rate": 5.841935483870968e-05,
      "loss": 0.0883,
      "step": 4390
    },
    {
      "epoch": 70.4,
      "grad_norm": 0.46907031536102295,
      "learning_rate": 5.809677419354839e-05,
      "loss": 0.0887,
      "step": 4400
    },
    {
      "epoch": 70.56,
      "grad_norm": 0.40855908393859863,
      "learning_rate": 5.77741935483871e-05,
      "loss": 0.089,
      "step": 4410
    },
    {
      "epoch": 70.72,
      "grad_norm": 0.37600624561309814,
      "learning_rate": 5.745161290322581e-05,
      "loss": 0.0915,
      "step": 4420
    },
    {
      "epoch": 70.88,
      "grad_norm": 0.6693145036697388,
      "learning_rate": 5.712903225806452e-05,
      "loss": 0.0919,
      "step": 4430
    },
    {
      "epoch": 71.04,
      "grad_norm": 0.4039818346500397,
      "learning_rate": 5.6806451612903234e-05,
      "loss": 0.0909,
      "step": 4440
    },
    {
      "epoch": 71.2,
      "grad_norm": 0.3611462414264679,
      "learning_rate": 5.648387096774193e-05,
      "loss": 0.0874,
      "step": 4450
    },
    {
      "epoch": 71.36,
      "grad_norm": 0.402322381734848,
      "learning_rate": 5.616129032258065e-05,
      "loss": 0.0877,
      "step": 4460
    },
    {
      "epoch": 71.52,
      "grad_norm": 0.48968443274497986,
      "learning_rate": 5.583870967741935e-05,
      "loss": 0.0893,
      "step": 4470
    },
    {
      "epoch": 71.68,
      "grad_norm": 0.4324292242527008,
      "learning_rate": 5.5516129032258065e-05,
      "loss": 0.0898,
      "step": 4480
    },
    {
      "epoch": 71.84,
      "grad_norm": 0.44097113609313965,
      "learning_rate": 5.519354838709677e-05,
      "loss": 0.0903,
      "step": 4490
    },
    {
      "epoch": 72.0,
      "grad_norm": 0.38512125611305237,
      "learning_rate": 5.4870967741935484e-05,
      "loss": 0.0921,
      "step": 4500
    },
    {
      "epoch": 72.16,
      "grad_norm": 0.37481260299682617,
      "learning_rate": 5.45483870967742e-05,
      "loss": 0.0867,
      "step": 4510
    },
    {
      "epoch": 72.32,
      "grad_norm": 0.4293956160545349,
      "learning_rate": 5.4225806451612904e-05,
      "loss": 0.0876,
      "step": 4520
    },
    {
      "epoch": 72.48,
      "grad_norm": 0.4083200991153717,
      "learning_rate": 5.390322580645162e-05,
      "loss": 0.0886,
      "step": 4530
    },
    {
      "epoch": 72.64,
      "grad_norm": 0.3753615617752075,
      "learning_rate": 5.358064516129032e-05,
      "loss": 0.09,
      "step": 4540
    },
    {
      "epoch": 72.8,
      "grad_norm": 0.3720114529132843,
      "learning_rate": 5.3258064516129036e-05,
      "loss": 0.0892,
      "step": 4550
    },
    {
      "epoch": 72.96,
      "grad_norm": 0.3209419250488281,
      "learning_rate": 5.293548387096774e-05,
      "loss": 0.0915,
      "step": 4560
    },
    {
      "epoch": 73.12,
      "grad_norm": 0.35520559549331665,
      "learning_rate": 5.2612903225806455e-05,
      "loss": 0.0871,
      "step": 4570
    },
    {
      "epoch": 73.28,
      "grad_norm": 0.3513979911804199,
      "learning_rate": 5.229032258064517e-05,
      "loss": 0.087,
      "step": 4580
    },
    {
      "epoch": 73.44,
      "grad_norm": 0.5205033421516418,
      "learning_rate": 5.1967741935483874e-05,
      "loss": 0.088,
      "step": 4590
    },
    {
      "epoch": 73.6,
      "grad_norm": 0.3683389127254486,
      "learning_rate": 5.164516129032259e-05,
      "loss": 0.0886,
      "step": 4600
    },
    {
      "epoch": 73.76,
      "grad_norm": 0.45570969581604004,
      "learning_rate": 5.132258064516129e-05,
      "loss": 0.0902,
      "step": 4610
    },
    {
      "epoch": 73.92,
      "grad_norm": 0.34694406390190125,
      "learning_rate": 5.1000000000000006e-05,
      "loss": 0.0909,
      "step": 4620
    },
    {
      "epoch": 74.08,
      "grad_norm": 0.3391146957874298,
      "learning_rate": 5.0677419354838706e-05,
      "loss": 0.0888,
      "step": 4630
    },
    {
      "epoch": 74.24,
      "grad_norm": 0.45434221625328064,
      "learning_rate": 5.035483870967742e-05,
      "loss": 0.0866,
      "step": 4640
    },
    {
      "epoch": 74.4,
      "grad_norm": 0.4046492576599121,
      "learning_rate": 5.003225806451614e-05,
      "loss": 0.0881,
      "step": 4650
    },
    {
      "epoch": 74.56,
      "grad_norm": 0.36426255106925964,
      "learning_rate": 4.970967741935484e-05,
      "loss": 0.0887,
      "step": 4660
    },
    {
      "epoch": 74.72,
      "grad_norm": 0.38272616267204285,
      "learning_rate": 4.938709677419355e-05,
      "loss": 0.0884,
      "step": 4670
    },
    {
      "epoch": 74.88,
      "grad_norm": 0.4772709012031555,
      "learning_rate": 4.9064516129032264e-05,
      "loss": 0.0902,
      "step": 4680
    },
    {
      "epoch": 75.04,
      "grad_norm": 0.32595178484916687,
      "learning_rate": 4.874193548387097e-05,
      "loss": 0.0893,
      "step": 4690
    },
    {
      "epoch": 75.2,
      "grad_norm": 0.4148579239845276,
      "learning_rate": 4.8419354838709676e-05,
      "loss": 0.0856,
      "step": 4700
    },
    {
      "epoch": 75.36,
      "grad_norm": 0.43145865201950073,
      "learning_rate": 4.809677419354839e-05,
      "loss": 0.0866,
      "step": 4710
    },
    {
      "epoch": 75.52,
      "grad_norm": 0.4075751006603241,
      "learning_rate": 4.7774193548387096e-05,
      "loss": 0.0881,
      "step": 4720
    },
    {
      "epoch": 75.68,
      "grad_norm": 0.33043548464775085,
      "learning_rate": 4.745161290322581e-05,
      "loss": 0.0898,
      "step": 4730
    },
    {
      "epoch": 75.84,
      "grad_norm": 0.36957836151123047,
      "learning_rate": 4.712903225806452e-05,
      "loss": 0.0904,
      "step": 4740
    },
    {
      "epoch": 76.0,
      "grad_norm": 0.3838096857070923,
      "learning_rate": 4.680645161290323e-05,
      "loss": 0.0896,
      "step": 4750
    },
    {
      "epoch": 76.16,
      "grad_norm": 0.42275360226631165,
      "learning_rate": 4.648387096774194e-05,
      "loss": 0.0867,
      "step": 4760
    },
    {
      "epoch": 76.32,
      "grad_norm": 0.5027480721473694,
      "learning_rate": 4.616129032258065e-05,
      "loss": 0.0864,
      "step": 4770
    },
    {
      "epoch": 76.48,
      "grad_norm": 0.36403143405914307,
      "learning_rate": 4.583870967741935e-05,
      "loss": 0.088,
      "step": 4780
    },
    {
      "epoch": 76.64,
      "grad_norm": 0.3124987781047821,
      "learning_rate": 4.5516129032258066e-05,
      "loss": 0.0881,
      "step": 4790
    },
    {
      "epoch": 76.8,
      "grad_norm": 0.38175061345100403,
      "learning_rate": 4.519354838709678e-05,
      "loss": 0.0891,
      "step": 4800
    },
    {
      "epoch": 76.96,
      "grad_norm": 0.41596823930740356,
      "learning_rate": 4.4870967741935485e-05,
      "loss": 0.0898,
      "step": 4810
    },
    {
      "epoch": 77.12,
      "grad_norm": 0.31852710247039795,
      "learning_rate": 4.45483870967742e-05,
      "loss": 0.0864,
      "step": 4820
    },
    {
      "epoch": 77.28,
      "grad_norm": 0.4198415279388428,
      "learning_rate": 4.4225806451612905e-05,
      "loss": 0.0857,
      "step": 4830
    },
    {
      "epoch": 77.44,
      "grad_norm": 0.392374724149704,
      "learning_rate": 4.390322580645162e-05,
      "loss": 0.0864,
      "step": 4840
    },
    {
      "epoch": 77.6,
      "grad_norm": 0.4310876429080963,
      "learning_rate": 4.3580645161290324e-05,
      "loss": 0.0893,
      "step": 4850
    },
    {
      "epoch": 77.76,
      "grad_norm": 0.3877263367176056,
      "learning_rate": 4.325806451612903e-05,
      "loss": 0.0888,
      "step": 4860
    },
    {
      "epoch": 77.92,
      "grad_norm": 0.31759241223335266,
      "learning_rate": 4.293548387096775e-05,
      "loss": 0.0902,
      "step": 4870
    },
    {
      "epoch": 78.08,
      "grad_norm": 0.3832705616950989,
      "learning_rate": 4.2612903225806456e-05,
      "loss": 0.0883,
      "step": 4880
    },
    {
      "epoch": 78.24,
      "grad_norm": 0.3763550817966461,
      "learning_rate": 4.229032258064516e-05,
      "loss": 0.086,
      "step": 4890
    },
    {
      "epoch": 78.4,
      "grad_norm": 0.45972681045532227,
      "learning_rate": 4.1967741935483875e-05,
      "loss": 0.087,
      "step": 4900
    },
    {
      "epoch": 78.56,
      "grad_norm": 0.4792405664920807,
      "learning_rate": 4.164516129032258e-05,
      "loss": 0.0886,
      "step": 4910
    },
    {
      "epoch": 78.72,
      "grad_norm": 0.34303322434425354,
      "learning_rate": 4.132258064516129e-05,
      "loss": 0.0884,
      "step": 4920
    },
    {
      "epoch": 78.88,
      "grad_norm": 0.33531343936920166,
      "learning_rate": 4.1e-05,
      "loss": 0.0892,
      "step": 4930
    },
    {
      "epoch": 79.04,
      "grad_norm": 0.3543599545955658,
      "learning_rate": 4.0677419354838713e-05,
      "loss": 0.0891,
      "step": 4940
    },
    {
      "epoch": 79.2,
      "grad_norm": 0.3615137040615082,
      "learning_rate": 4.035483870967742e-05,
      "loss": 0.0853,
      "step": 4950
    },
    {
      "epoch": 79.36,
      "grad_norm": 0.40525078773498535,
      "learning_rate": 4.003225806451613e-05,
      "loss": 0.0864,
      "step": 4960
    },
    {
      "epoch": 79.52,
      "grad_norm": 0.3445266783237457,
      "learning_rate": 3.970967741935484e-05,
      "loss": 0.088,
      "step": 4970
    },
    {
      "epoch": 79.68,
      "grad_norm": 0.4922199547290802,
      "learning_rate": 3.938709677419355e-05,
      "loss": 0.0885,
      "step": 4980
    },
    {
      "epoch": 79.84,
      "grad_norm": 0.3156628906726837,
      "learning_rate": 3.906451612903226e-05,
      "loss": 0.0888,
      "step": 4990
    },
    {
      "epoch": 80.0,
      "grad_norm": 0.32096874713897705,
      "learning_rate": 3.8741935483870964e-05,
      "loss": 0.0889,
      "step": 5000
    },
    {
      "epoch": 80.16,
      "grad_norm": 0.3683740794658661,
      "learning_rate": 3.8419354838709684e-05,
      "loss": 0.0853,
      "step": 5010
    },
    {
      "epoch": 80.32,
      "grad_norm": 0.43887290358543396,
      "learning_rate": 3.809677419354839e-05,
      "loss": 0.0855,
      "step": 5020
    },
    {
      "epoch": 80.48,
      "grad_norm": 0.4039384424686432,
      "learning_rate": 3.7774193548387096e-05,
      "loss": 0.0865,
      "step": 5030
    },
    {
      "epoch": 80.64,
      "grad_norm": 0.3594481647014618,
      "learning_rate": 3.745161290322581e-05,
      "loss": 0.0871,
      "step": 5040
    },
    {
      "epoch": 80.8,
      "grad_norm": 0.3050839304924011,
      "learning_rate": 3.7129032258064516e-05,
      "loss": 0.0892,
      "step": 5050
    },
    {
      "epoch": 80.96,
      "grad_norm": 0.3346671164035797,
      "learning_rate": 3.680645161290323e-05,
      "loss": 0.0898,
      "step": 5060
    },
    {
      "epoch": 81.12,
      "grad_norm": 0.32323309779167175,
      "learning_rate": 3.648387096774194e-05,
      "loss": 0.0871,
      "step": 5070
    },
    {
      "epoch": 81.28,
      "grad_norm": 0.5844283699989319,
      "learning_rate": 3.616129032258065e-05,
      "loss": 0.085,
      "step": 5080
    },
    {
      "epoch": 81.44,
      "grad_norm": 0.39134910702705383,
      "learning_rate": 3.583870967741936e-05,
      "loss": 0.0868,
      "step": 5090
    },
    {
      "epoch": 81.6,
      "grad_norm": 0.3730677366256714,
      "learning_rate": 3.551612903225807e-05,
      "loss": 0.0879,
      "step": 5100
    },
    {
      "epoch": 81.76,
      "grad_norm": 0.40242692828178406,
      "learning_rate": 3.519354838709677e-05,
      "loss": 0.0878,
      "step": 5110
    },
    {
      "epoch": 81.92,
      "grad_norm": 0.40863659977912903,
      "learning_rate": 3.4870967741935486e-05,
      "loss": 0.0887,
      "step": 5120
    },
    {
      "epoch": 82.08,
      "grad_norm": 0.3297955095767975,
      "learning_rate": 3.454838709677419e-05,
      "loss": 0.0871,
      "step": 5130
    },
    {
      "epoch": 82.24,
      "grad_norm": 0.39270907640457153,
      "learning_rate": 3.4225806451612905e-05,
      "loss": 0.0851,
      "step": 5140
    },
    {
      "epoch": 82.4,
      "grad_norm": 0.35612475872039795,
      "learning_rate": 3.390322580645162e-05,
      "loss": 0.0864,
      "step": 5150
    },
    {
      "epoch": 82.56,
      "grad_norm": 0.419009804725647,
      "learning_rate": 3.3580645161290325e-05,
      "loss": 0.0872,
      "step": 5160
    },
    {
      "epoch": 82.72,
      "grad_norm": 0.44817715883255005,
      "learning_rate": 3.325806451612903e-05,
      "loss": 0.0882,
      "step": 5170
    },
    {
      "epoch": 82.88,
      "grad_norm": 0.36784857511520386,
      "learning_rate": 3.2935483870967744e-05,
      "loss": 0.0878,
      "step": 5180
    },
    {
      "epoch": 83.04,
      "grad_norm": 0.42855191230773926,
      "learning_rate": 3.261290322580645e-05,
      "loss": 0.088,
      "step": 5190
    },
    {
      "epoch": 83.2,
      "grad_norm": 0.34956803917884827,
      "learning_rate": 3.229032258064516e-05,
      "loss": 0.0848,
      "step": 5200
    },
    {
      "epoch": 83.36,
      "grad_norm": 0.3654519021511078,
      "learning_rate": 3.1967741935483876e-05,
      "loss": 0.0862,
      "step": 5210
    },
    {
      "epoch": 83.52,
      "grad_norm": 0.3838617205619812,
      "learning_rate": 3.164516129032258e-05,
      "loss": 0.0862,
      "step": 5220
    },
    {
      "epoch": 83.68,
      "grad_norm": 0.36475858092308044,
      "learning_rate": 3.1322580645161295e-05,
      "loss": 0.0876,
      "step": 5230
    },
    {
      "epoch": 83.84,
      "grad_norm": 0.39168354868888855,
      "learning_rate": 3.1e-05,
      "loss": 0.0882,
      "step": 5240
    },
    {
      "epoch": 84.0,
      "grad_norm": 0.38403555750846863,
      "learning_rate": 3.067741935483871e-05,
      "loss": 0.0885,
      "step": 5250
    },
    {
      "epoch": 84.16,
      "grad_norm": 0.3362351059913635,
      "learning_rate": 3.035483870967742e-05,
      "loss": 0.0852,
      "step": 5260
    },
    {
      "epoch": 84.32,
      "grad_norm": 0.37289589643478394,
      "learning_rate": 3.0032258064516127e-05,
      "loss": 0.0852,
      "step": 5270
    },
    {
      "epoch": 84.48,
      "grad_norm": 0.3794906437397003,
      "learning_rate": 2.9709677419354843e-05,
      "loss": 0.0866,
      "step": 5280
    },
    {
      "epoch": 84.64,
      "grad_norm": 0.46871262788772583,
      "learning_rate": 2.938709677419355e-05,
      "loss": 0.0871,
      "step": 5290
    },
    {
      "epoch": 84.8,
      "grad_norm": 0.3682585060596466,
      "learning_rate": 2.906451612903226e-05,
      "loss": 0.0879,
      "step": 5300
    },
    {
      "epoch": 84.96,
      "grad_norm": 0.36091622710227966,
      "learning_rate": 2.874193548387097e-05,
      "loss": 0.0884,
      "step": 5310
    },
    {
      "epoch": 85.12,
      "grad_norm": 0.3225698173046112,
      "learning_rate": 2.8419354838709678e-05,
      "loss": 0.0858,
      "step": 5320
    },
    {
      "epoch": 85.28,
      "grad_norm": 0.36931437253952026,
      "learning_rate": 2.8096774193548388e-05,
      "loss": 0.0853,
      "step": 5330
    },
    {
      "epoch": 85.44,
      "grad_norm": 0.3807561993598938,
      "learning_rate": 2.77741935483871e-05,
      "loss": 0.0859,
      "step": 5340
    },
    {
      "epoch": 85.6,
      "grad_norm": 0.4048006534576416,
      "learning_rate": 2.745161290322581e-05,
      "loss": 0.0865,
      "step": 5350
    },
    {
      "epoch": 85.76,
      "grad_norm": 0.3060702979564667,
      "learning_rate": 2.712903225806452e-05,
      "loss": 0.0874,
      "step": 5360
    },
    {
      "epoch": 85.92,
      "grad_norm": 0.3515985608100891,
      "learning_rate": 2.6806451612903226e-05,
      "loss": 0.088,
      "step": 5370
    },
    {
      "epoch": 86.08,
      "grad_norm": 0.3089566230773926,
      "learning_rate": 2.6483870967741936e-05,
      "loss": 0.0871,
      "step": 5380
    },
    {
      "epoch": 86.24,
      "grad_norm": 0.42711663246154785,
      "learning_rate": 2.6161290322580645e-05,
      "loss": 0.0844,
      "step": 5390
    },
    {
      "epoch": 86.4,
      "grad_norm": 0.37982869148254395,
      "learning_rate": 2.5838709677419355e-05,
      "loss": 0.0861,
      "step": 5400
    },
    {
      "epoch": 86.56,
      "grad_norm": 0.38406598567962646,
      "learning_rate": 2.5516129032258068e-05,
      "loss": 0.0866,
      "step": 5410
    },
    {
      "epoch": 86.72,
      "grad_norm": 0.37800315022468567,
      "learning_rate": 2.5193548387096777e-05,
      "loss": 0.0868,
      "step": 5420
    },
    {
      "epoch": 86.88,
      "grad_norm": 0.39667201042175293,
      "learning_rate": 2.4870967741935487e-05,
      "loss": 0.0874,
      "step": 5430
    },
    {
      "epoch": 87.04,
      "grad_norm": 0.30831629037857056,
      "learning_rate": 2.4548387096774193e-05,
      "loss": 0.087,
      "step": 5440
    },
    {
      "epoch": 87.2,
      "grad_norm": 0.3226090669631958,
      "learning_rate": 2.4225806451612903e-05,
      "loss": 0.0851,
      "step": 5450
    },
    {
      "epoch": 87.36,
      "grad_norm": 0.3767179250717163,
      "learning_rate": 2.3903225806451616e-05,
      "loss": 0.0847,
      "step": 5460
    },
    {
      "epoch": 87.52,
      "grad_norm": 0.4611583650112152,
      "learning_rate": 2.3580645161290325e-05,
      "loss": 0.086,
      "step": 5470
    },
    {
      "epoch": 87.68,
      "grad_norm": 0.4075451195240021,
      "learning_rate": 2.325806451612903e-05,
      "loss": 0.087,
      "step": 5480
    },
    {
      "epoch": 87.84,
      "grad_norm": 0.3248244822025299,
      "learning_rate": 2.293548387096774e-05,
      "loss": 0.0874,
      "step": 5490
    },
    {
      "epoch": 88.0,
      "grad_norm": 0.3889995813369751,
      "learning_rate": 2.2612903225806454e-05,
      "loss": 0.0876,
      "step": 5500
    },
    {
      "epoch": 88.16,
      "grad_norm": 0.3838357925415039,
      "learning_rate": 2.229032258064516e-05,
      "loss": 0.0847,
      "step": 5510
    },
    {
      "epoch": 88.32,
      "grad_norm": 0.35209789872169495,
      "learning_rate": 2.196774193548387e-05,
      "loss": 0.0843,
      "step": 5520
    },
    {
      "epoch": 88.48,
      "grad_norm": 0.3327479958534241,
      "learning_rate": 2.1645161290322583e-05,
      "loss": 0.086,
      "step": 5530
    },
    {
      "epoch": 88.64,
      "grad_norm": 0.36551928520202637,
      "learning_rate": 2.1322580645161293e-05,
      "loss": 0.0863,
      "step": 5540
    },
    {
      "epoch": 88.8,
      "grad_norm": 0.38019606471061707,
      "learning_rate": 2.1e-05,
      "loss": 0.0866,
      "step": 5550
    },
    {
      "epoch": 88.96,
      "grad_norm": 0.3546133041381836,
      "learning_rate": 2.0677419354838712e-05,
      "loss": 0.0877,
      "step": 5560
    },
    {
      "epoch": 89.12,
      "grad_norm": 0.276177316904068,
      "learning_rate": 2.035483870967742e-05,
      "loss": 0.0856,
      "step": 5570
    },
    {
      "epoch": 89.28,
      "grad_norm": 0.36877840757369995,
      "learning_rate": 2.003225806451613e-05,
      "loss": 0.0845,
      "step": 5580
    },
    {
      "epoch": 89.44,
      "grad_norm": 0.3937029540538788,
      "learning_rate": 1.9709677419354837e-05,
      "loss": 0.0853,
      "step": 5590
    },
    {
      "epoch": 89.6,
      "grad_norm": 0.3682461977005005,
      "learning_rate": 1.938709677419355e-05,
      "loss": 0.0861,
      "step": 5600
    },
    {
      "epoch": 89.76,
      "grad_norm": 0.3939973711967468,
      "learning_rate": 1.906451612903226e-05,
      "loss": 0.0863,
      "step": 5610
    },
    {
      "epoch": 89.92,
      "grad_norm": 0.35440462827682495,
      "learning_rate": 1.8741935483870966e-05,
      "loss": 0.0871,
      "step": 5620
    },
    {
      "epoch": 90.08,
      "grad_norm": 0.3386973738670349,
      "learning_rate": 1.841935483870968e-05,
      "loss": 0.0852,
      "step": 5630
    },
    {
      "epoch": 90.24,
      "grad_norm": 0.359678715467453,
      "learning_rate": 1.809677419354839e-05,
      "loss": 0.084,
      "step": 5640
    },
    {
      "epoch": 90.4,
      "grad_norm": 0.38088712096214294,
      "learning_rate": 1.7774193548387098e-05,
      "loss": 0.0851,
      "step": 5650
    },
    {
      "epoch": 90.56,
      "grad_norm": 0.4327407479286194,
      "learning_rate": 1.7451612903225808e-05,
      "loss": 0.0861,
      "step": 5660
    },
    {
      "epoch": 90.72,
      "grad_norm": 0.3494204580783844,
      "learning_rate": 1.7129032258064517e-05,
      "loss": 0.086,
      "step": 5670
    },
    {
      "epoch": 90.88,
      "grad_norm": 0.3655516803264618,
      "learning_rate": 1.6806451612903227e-05,
      "loss": 0.0867,
      "step": 5680
    },
    {
      "epoch": 91.04,
      "grad_norm": 0.3738172948360443,
      "learning_rate": 1.6483870967741937e-05,
      "loss": 0.0867,
      "step": 5690
    },
    {
      "epoch": 91.2,
      "grad_norm": 0.33062541484832764,
      "learning_rate": 1.6161290322580646e-05,
      "loss": 0.0841,
      "step": 5700
    },
    {
      "epoch": 91.36,
      "grad_norm": 0.3357619345188141,
      "learning_rate": 1.5838709677419356e-05,
      "loss": 0.0845,
      "step": 5710
    },
    {
      "epoch": 91.52,
      "grad_norm": 0.3250825107097626,
      "learning_rate": 1.5516129032258065e-05,
      "loss": 0.0858,
      "step": 5720
    },
    {
      "epoch": 91.68,
      "grad_norm": 0.4149022102355957,
      "learning_rate": 1.5193548387096777e-05,
      "loss": 0.0857,
      "step": 5730
    },
    {
      "epoch": 91.84,
      "grad_norm": 0.3375944495201111,
      "learning_rate": 1.4870967741935485e-05,
      "loss": 0.0863,
      "step": 5740
    },
    {
      "epoch": 92.0,
      "grad_norm": 0.39512765407562256,
      "learning_rate": 1.4548387096774194e-05,
      "loss": 0.0867,
      "step": 5750
    },
    {
      "epoch": 92.16,
      "grad_norm": 0.36112552881240845,
      "learning_rate": 1.4225806451612905e-05,
      "loss": 0.0838,
      "step": 5760
    },
    {
      "epoch": 92.32,
      "grad_norm": 0.3789423704147339,
      "learning_rate": 1.3903225806451613e-05,
      "loss": 0.0839,
      "step": 5770
    },
    {
      "epoch": 92.48,
      "grad_norm": 0.4131874740123749,
      "learning_rate": 1.3580645161290323e-05,
      "loss": 0.0845,
      "step": 5780
    },
    {
      "epoch": 92.64,
      "grad_norm": 0.4238837659358978,
      "learning_rate": 1.3258064516129033e-05,
      "loss": 0.0859,
      "step": 5790
    },
    {
      "epoch": 92.8,
      "grad_norm": 0.4758787751197815,
      "learning_rate": 1.2935483870967744e-05,
      "loss": 0.0862,
      "step": 5800
    },
    {
      "epoch": 92.96,
      "grad_norm": 0.4589889943599701,
      "learning_rate": 1.2612903225806452e-05,
      "loss": 0.0867,
      "step": 5810
    },
    {
      "epoch": 93.12,
      "grad_norm": 0.3451363146305084,
      "learning_rate": 1.2290322580645163e-05,
      "loss": 0.0844,
      "step": 5820
    },
    {
      "epoch": 93.28,
      "grad_norm": 0.3493219316005707,
      "learning_rate": 1.1967741935483871e-05,
      "loss": 0.0847,
      "step": 5830
    },
    {
      "epoch": 93.44,
      "grad_norm": 0.3858239948749542,
      "learning_rate": 1.1645161290322582e-05,
      "loss": 0.085,
      "step": 5840
    },
    {
      "epoch": 93.6,
      "grad_norm": 0.4442474842071533,
      "learning_rate": 1.132258064516129e-05,
      "loss": 0.0851,
      "step": 5850
    },
    {
      "epoch": 93.76,
      "grad_norm": 0.41025981307029724,
      "learning_rate": 1.1000000000000001e-05,
      "loss": 0.0858,
      "step": 5860
    },
    {
      "epoch": 93.92,
      "grad_norm": 0.39134228229522705,
      "learning_rate": 1.0677419354838711e-05,
      "loss": 0.0855,
      "step": 5870
    },
    {
      "epoch": 94.08,
      "grad_norm": 0.39811140298843384,
      "learning_rate": 1.0354838709677419e-05,
      "loss": 0.0848,
      "step": 5880
    },
    {
      "epoch": 94.24,
      "grad_norm": 0.37056440114974976,
      "learning_rate": 1.003225806451613e-05,
      "loss": 0.0834,
      "step": 5890
    },
    {
      "epoch": 94.4,
      "grad_norm": 0.41117194294929504,
      "learning_rate": 9.709677419354838e-06,
      "loss": 0.0846,
      "step": 5900
    },
    {
      "epoch": 94.56,
      "grad_norm": 0.3775732219219208,
      "learning_rate": 9.38709677419355e-06,
      "loss": 0.0846,
      "step": 5910
    },
    {
      "epoch": 94.72,
      "grad_norm": 0.4182074964046478,
      "learning_rate": 9.064516129032259e-06,
      "loss": 0.085,
      "step": 5920
    },
    {
      "epoch": 94.88,
      "grad_norm": 0.3670865595340729,
      "learning_rate": 8.741935483870969e-06,
      "loss": 0.0854,
      "step": 5930
    },
    {
      "epoch": 95.04,
      "grad_norm": 0.3670247793197632,
      "learning_rate": 8.419354838709678e-06,
      "loss": 0.0853,
      "step": 5940
    },
    {
      "epoch": 95.2,
      "grad_norm": 0.3669385313987732,
      "learning_rate": 8.096774193548388e-06,
      "loss": 0.0839,
      "step": 5950
    },
    {
      "epoch": 95.36,
      "grad_norm": 0.41123315691947937,
      "learning_rate": 7.774193548387097e-06,
      "loss": 0.0838,
      "step": 5960
    },
    {
      "epoch": 95.52,
      "grad_norm": 0.4040362238883972,
      "learning_rate": 7.451612903225806e-06,
      "loss": 0.0847,
      "step": 5970
    },
    {
      "epoch": 95.68,
      "grad_norm": 0.40712663531303406,
      "learning_rate": 7.1290322580645166e-06,
      "loss": 0.0851,
      "step": 5980
    },
    {
      "epoch": 95.84,
      "grad_norm": 0.46202871203422546,
      "learning_rate": 6.806451612903226e-06,
      "loss": 0.0853,
      "step": 5990
    },
    {
      "epoch": 96.0,
      "grad_norm": 0.36839255690574646,
      "learning_rate": 6.483870967741936e-06,
      "loss": 0.0856,
      "step": 6000
    },
    {
      "epoch": 96.16,
      "grad_norm": 0.34025129675865173,
      "learning_rate": 6.161290322580645e-06,
      "loss": 0.084,
      "step": 6010
    },
    {
      "epoch": 96.32,
      "grad_norm": 0.37305212020874023,
      "learning_rate": 5.838709677419355e-06,
      "loss": 0.0838,
      "step": 6020
    },
    {
      "epoch": 96.48,
      "grad_norm": 0.3265480101108551,
      "learning_rate": 5.5161290322580645e-06,
      "loss": 0.0843,
      "step": 6030
    },
    {
      "epoch": 96.64,
      "grad_norm": 0.42518019676208496,
      "learning_rate": 5.193548387096774e-06,
      "loss": 0.0843,
      "step": 6040
    },
    {
      "epoch": 96.8,
      "grad_norm": 0.40051665902137756,
      "learning_rate": 4.870967741935484e-06,
      "loss": 0.0847,
      "step": 6050
    },
    {
      "epoch": 96.96,
      "grad_norm": 0.4043062627315521,
      "learning_rate": 4.548387096774194e-06,
      "loss": 0.0849,
      "step": 6060
    },
    {
      "epoch": 97.12,
      "grad_norm": 0.350195050239563,
      "learning_rate": 4.225806451612904e-06,
      "loss": 0.0845,
      "step": 6070
    },
    {
      "epoch": 97.28,
      "grad_norm": 0.4239977300167084,
      "learning_rate": 3.903225806451613e-06,
      "loss": 0.084,
      "step": 6080
    },
    {
      "epoch": 97.44,
      "grad_norm": 0.3817554712295532,
      "learning_rate": 3.5806451612903225e-06,
      "loss": 0.0842,
      "step": 6090
    },
    {
      "epoch": 97.6,
      "grad_norm": 0.3956522047519684,
      "learning_rate": 3.258064516129032e-06,
      "loss": 0.0837,
      "step": 6100
    },
    {
      "epoch": 97.76,
      "grad_norm": 0.4088551104068756,
      "learning_rate": 2.935483870967742e-06,
      "loss": 0.0842,
      "step": 6110
    },
    {
      "epoch": 97.92,
      "grad_norm": 0.4062351882457733,
      "learning_rate": 2.6129032258064518e-06,
      "loss": 0.084,
      "step": 6120
    },
    {
      "epoch": 98.08,
      "grad_norm": 0.3595307767391205,
      "learning_rate": 2.2903225806451614e-06,
      "loss": 0.0836,
      "step": 6130
    },
    {
      "epoch": 98.24,
      "grad_norm": 0.4038846790790558,
      "learning_rate": 1.967741935483871e-06,
      "loss": 0.0831,
      "step": 6140
    },
    {
      "epoch": 98.4,
      "grad_norm": 0.411885529756546,
      "learning_rate": 1.6451612903225808e-06,
      "loss": 0.0839,
      "step": 6150
    },
    {
      "epoch": 98.56,
      "grad_norm": 0.3277604579925537,
      "learning_rate": 1.3225806451612904e-06,
      "loss": 0.084,
      "step": 6160
    },
    {
      "epoch": 98.72,
      "grad_norm": 0.3658997118473053,
      "learning_rate": 1.0000000000000002e-06,
      "loss": 0.0843,
      "step": 6170
    },
    {
      "epoch": 98.88,
      "grad_norm": 0.4060486853122711,
      "learning_rate": 6.774193548387097e-07,
      "loss": 0.0838,
      "step": 6180
    },
    {
      "epoch": 99.04,
      "grad_norm": 0.3686084449291229,
      "learning_rate": 3.548387096774194e-07,
      "loss": 0.0835,
      "step": 6190
    },
    {
      "epoch": 99.2,
      "grad_norm": 0.36468222737312317,
      "learning_rate": 3.2258064516129035e-08,
      "loss": 0.0837,
      "step": 6200
    },
    {
      "epoch": 99.2,
      "step": 6200,
      "total_flos": 5.286498354069504e+17,
      "train_loss": 0.3455863463590222,
      "train_runtime": 16034.7635,
      "train_samples_per_second": 6.236,
      "train_steps_per_second": 0.387
    }
  ],
  "logging_steps": 10,
  "max_steps": 6200,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 100,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 5.286498354069504e+17,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}