{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 4375,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00022857142857142857,
      "grad_norm": 1.563055157661438,
      "learning_rate": 0.0,
      "loss": 4.000716209411621,
      "step": 1
    },
    {
      "epoch": 0.00045714285714285713,
      "grad_norm": 1.3258074522018433,
      "learning_rate": 6.25e-06,
      "loss": 3.953533411026001,
      "step": 2
    },
    {
      "epoch": 0.0006857142857142857,
      "grad_norm": 1.1743125915527344,
      "learning_rate": 1.25e-05,
      "loss": 3.9451112747192383,
      "step": 3
    },
    {
      "epoch": 0.0009142857142857143,
      "grad_norm": 1.600172996520996,
      "learning_rate": 1.875e-05,
      "loss": 3.7971606254577637,
      "step": 4
    },
    {
      "epoch": 0.001142857142857143,
      "grad_norm": 1.2557759284973145,
      "learning_rate": 2.5e-05,
      "loss": 4.068172454833984,
      "step": 5
    },
    {
      "epoch": 0.0013714285714285714,
      "grad_norm": 1.1853646039962769,
      "learning_rate": 3.125e-05,
      "loss": 3.6892147064208984,
      "step": 6
    },
    {
      "epoch": 0.0016,
      "grad_norm": 1.0192577838897705,
      "learning_rate": 3.75e-05,
      "loss": 4.298388957977295,
      "step": 7
    },
    {
      "epoch": 0.0018285714285714285,
      "grad_norm": 0.9764495491981506,
      "learning_rate": 4.375e-05,
      "loss": 4.282886505126953,
      "step": 8
    },
    {
      "epoch": 0.002057142857142857,
      "grad_norm": 1.9685038328170776,
      "learning_rate": 5e-05,
      "loss": 3.4855551719665527,
      "step": 9
    },
    {
      "epoch": 0.002285714285714286,
      "grad_norm": 1.493470311164856,
      "learning_rate": 5.6250000000000005e-05,
      "loss": 3.6132140159606934,
      "step": 10
    },
    {
      "epoch": 0.002514285714285714,
      "grad_norm": 1.2944492101669312,
      "learning_rate": 6.25e-05,
      "loss": 3.7139229774475098,
      "step": 11
    },
    {
      "epoch": 0.002742857142857143,
      "grad_norm": 1.3209120035171509,
      "learning_rate": 6.875e-05,
      "loss": 3.9067726135253906,
      "step": 12
    },
    {
      "epoch": 0.0029714285714285715,
      "grad_norm": 1.6926525831222534,
      "learning_rate": 7.5e-05,
      "loss": 3.483598232269287,
      "step": 13
    },
    {
      "epoch": 0.0032,
      "grad_norm": 1.5207566022872925,
      "learning_rate": 8.125000000000001e-05,
      "loss": 3.5910162925720215,
      "step": 14
    },
    {
      "epoch": 0.0034285714285714284,
      "grad_norm": 1.9836430549621582,
      "learning_rate": 8.75e-05,
      "loss": 4.305594444274902,
      "step": 15
    },
    {
      "epoch": 0.003657142857142857,
      "grad_norm": 1.2500442266464233,
      "learning_rate": 9.375e-05,
      "loss": 3.976480007171631,
      "step": 16
    },
    {
      "epoch": 0.0038857142857142857,
      "grad_norm": 1.4551279544830322,
      "learning_rate": 0.0001,
      "loss": 3.731271266937256,
      "step": 17
    },
    {
      "epoch": 0.004114285714285714,
      "grad_norm": 2.811328649520874,
      "learning_rate": 0.00010625,
      "loss": 3.940420150756836,
      "step": 18
    },
    {
      "epoch": 0.004342857142857143,
      "grad_norm": 1.1207599639892578,
      "learning_rate": 0.00011250000000000001,
      "loss": 3.7734670639038086,
      "step": 19
    },
    {
      "epoch": 0.004571428571428572,
      "grad_norm": 1.3179820775985718,
      "learning_rate": 0.00011875,
      "loss": 4.211020469665527,
      "step": 20
    },
    {
      "epoch": 0.0048,
      "grad_norm": 1.089626669883728,
      "learning_rate": 0.000125,
      "loss": 4.106476783752441,
      "step": 21
    },
    {
      "epoch": 0.005028571428571428,
      "grad_norm": 1.1228468418121338,
      "learning_rate": 0.00013125000000000002,
      "loss": 3.4354937076568604,
      "step": 22
    },
    {
      "epoch": 0.005257142857142857,
      "grad_norm": 1.0387296676635742,
      "learning_rate": 0.0001375,
      "loss": 4.152472972869873,
      "step": 23
    },
    {
      "epoch": 0.005485714285714286,
      "grad_norm": 1.2175796031951904,
      "learning_rate": 0.00014375,
      "loss": 3.7258622646331787,
      "step": 24
    },
    {
      "epoch": 0.005714285714285714,
      "grad_norm": 1.0984045267105103,
      "learning_rate": 0.00015,
      "loss": 3.429095506668091,
      "step": 25
    },
    {
      "epoch": 0.005942857142857143,
      "grad_norm": 1.4311732053756714,
      "learning_rate": 0.00015625,
      "loss": 3.8544273376464844,
      "step": 26
    },
    {
      "epoch": 0.006171428571428572,
      "grad_norm": 1.157279372215271,
      "learning_rate": 0.00016250000000000002,
      "loss": 3.9421043395996094,
      "step": 27
    },
    {
      "epoch": 0.0064,
      "grad_norm": 1.2566580772399902,
      "learning_rate": 0.00016875,
      "loss": 3.3919973373413086,
      "step": 28
    },
    {
      "epoch": 0.006628571428571429,
      "grad_norm": 1.0717591047286987,
      "learning_rate": 0.000175,
      "loss": 3.7658205032348633,
      "step": 29
    },
    {
      "epoch": 0.006857142857142857,
      "grad_norm": 0.9253070950508118,
      "learning_rate": 0.00018125,
      "loss": 3.7739505767822266,
      "step": 30
    },
    {
      "epoch": 0.0070857142857142855,
      "grad_norm": 1.2166458368301392,
      "learning_rate": 0.0001875,
      "loss": 3.5364267826080322,
      "step": 31
    },
    {
      "epoch": 0.007314285714285714,
      "grad_norm": 1.256104588508606,
      "learning_rate": 0.00019375000000000002,
      "loss": 3.588336229324341,
      "step": 32
    },
    {
      "epoch": 0.007542857142857143,
      "grad_norm": 1.1018946170806885,
      "learning_rate": 0.0002,
      "loss": 4.002081871032715,
      "step": 33
    },
    {
      "epoch": 0.0077714285714285715,
      "grad_norm": 1.3787511587142944,
      "learning_rate": 0.00020625,
      "loss": 3.8752501010894775,
      "step": 34
    },
    {
      "epoch": 0.008,
      "grad_norm": 1.3544801473617554,
      "learning_rate": 0.0002125,
      "loss": 3.5342838764190674,
      "step": 35
    },
    {
      "epoch": 0.008228571428571429,
      "grad_norm": 1.112373948097229,
      "learning_rate": 0.00021875,
      "loss": 3.699503183364868,
      "step": 36
    },
    {
      "epoch": 0.008457142857142858,
      "grad_norm": 1.3061130046844482,
      "learning_rate": 0.00022500000000000002,
      "loss": 3.5054750442504883,
      "step": 37
    },
    {
      "epoch": 0.008685714285714286,
      "grad_norm": 1.248658299446106,
      "learning_rate": 0.00023125,
      "loss": 3.615384340286255,
      "step": 38
    },
    {
      "epoch": 0.008914285714285715,
      "grad_norm": 1.4349433183670044,
      "learning_rate": 0.0002375,
      "loss": 3.4871950149536133,
      "step": 39
    },
    {
      "epoch": 0.009142857142857144,
      "grad_norm": 1.2729493379592896,
      "learning_rate": 0.00024375,
      "loss": 3.9547719955444336,
      "step": 40
    },
    {
      "epoch": 0.009371428571428572,
      "grad_norm": 1.2393776178359985,
      "learning_rate": 0.00025,
      "loss": 3.5953311920166016,
      "step": 41
    },
    {
      "epoch": 0.0096,
      "grad_norm": 1.2098788022994995,
      "learning_rate": 0.00025624999999999997,
      "loss": 3.706698417663574,
      "step": 42
    },
    {
      "epoch": 0.009828571428571428,
      "grad_norm": 1.1835993528366089,
      "learning_rate": 0.00026250000000000004,
      "loss": 4.059277534484863,
      "step": 43
    },
    {
      "epoch": 0.010057142857142857,
      "grad_norm": 1.1306127309799194,
      "learning_rate": 0.00026875,
      "loss": 3.6649301052093506,
      "step": 44
    },
    {
      "epoch": 0.010285714285714285,
      "grad_norm": 1.3582308292388916,
      "learning_rate": 0.000275,
      "loss": 3.818805694580078,
      "step": 45
    },
    {
      "epoch": 0.010514285714285714,
      "grad_norm": 2.3792107105255127,
      "learning_rate": 0.00028125000000000003,
      "loss": 3.68064546585083,
      "step": 46
    },
    {
      "epoch": 0.010742857142857143,
      "grad_norm": 1.539760947227478,
      "learning_rate": 0.0002875,
      "loss": 3.5947117805480957,
      "step": 47
    },
    {
      "epoch": 0.010971428571428571,
      "grad_norm": 1.5554462671279907,
      "learning_rate": 0.00029375,
      "loss": 4.058520317077637,
      "step": 48
    },
    {
      "epoch": 0.0112,
      "grad_norm": 1.605898141860962,
      "learning_rate": 0.0003,
      "loss": 3.681978225708008,
      "step": 49
    },
    {
      "epoch": 0.011428571428571429,
      "grad_norm": 1.3231074810028076,
      "learning_rate": 0.00030625000000000004,
      "loss": 3.862358570098877,
      "step": 50
    },
    {
      "epoch": 0.011657142857142857,
      "grad_norm": 1.5183755159378052,
      "learning_rate": 0.0003125,
      "loss": 3.5748305320739746,
      "step": 51
    },
    {
      "epoch": 0.011885714285714286,
      "grad_norm": 1.4168895483016968,
      "learning_rate": 0.00031874999999999997,
      "loss": 3.836300849914551,
      "step": 52
    },
    {
      "epoch": 0.012114285714285715,
      "grad_norm": 1.5071827173233032,
      "learning_rate": 0.00032500000000000004,
      "loss": 4.031536102294922,
      "step": 53
    },
    {
      "epoch": 0.012342857142857143,
      "grad_norm": 1.308683156967163,
      "learning_rate": 0.00033125,
      "loss": 3.7215256690979004,
      "step": 54
    },
    {
      "epoch": 0.012571428571428572,
      "grad_norm": 1.4168165922164917,
      "learning_rate": 0.0003375,
      "loss": 3.590249538421631,
      "step": 55
    },
    {
      "epoch": 0.0128,
      "grad_norm": 1.235178828239441,
      "learning_rate": 0.00034375,
      "loss": 3.8189542293548584,
      "step": 56
    },
    {
      "epoch": 0.01302857142857143,
      "grad_norm": 1.1801726818084717,
      "learning_rate": 0.00035,
      "loss": 3.6454648971557617,
      "step": 57
    },
    {
      "epoch": 0.013257142857142858,
      "grad_norm": 1.5661216974258423,
      "learning_rate": 0.00035625,
      "loss": 3.9985687732696533,
      "step": 58
    },
    {
      "epoch": 0.013485714285714285,
      "grad_norm": 1.3641514778137207,
      "learning_rate": 0.0003625,
      "loss": 3.614236831665039,
      "step": 59
    },
    {
      "epoch": 0.013714285714285714,
      "grad_norm": 1.2128474712371826,
      "learning_rate": 0.00036875000000000005,
      "loss": 3.432009220123291,
      "step": 60
    },
    {
      "epoch": 0.013942857142857142,
      "grad_norm": 1.3696622848510742,
      "learning_rate": 0.000375,
      "loss": 3.9747397899627686,
      "step": 61
    },
    {
      "epoch": 0.014171428571428571,
      "grad_norm": 1.2073640823364258,
      "learning_rate": 0.00038124999999999997,
      "loss": 3.811964750289917,
      "step": 62
    },
    {
      "epoch": 0.0144,
      "grad_norm": 1.409297227859497,
      "learning_rate": 0.00038750000000000004,
      "loss": 3.5807268619537354,
      "step": 63
    },
    {
      "epoch": 0.014628571428571428,
      "grad_norm": 1.4627134799957275,
      "learning_rate": 0.00039375,
      "loss": 3.9213037490844727,
      "step": 64
    },
    {
      "epoch": 0.014857142857142857,
      "grad_norm": 1.7188282012939453,
      "learning_rate": 0.0004,
      "loss": 3.888178586959839,
      "step": 65
    },
    {
      "epoch": 0.015085714285714286,
      "grad_norm": 1.3482108116149902,
      "learning_rate": 0.00040625000000000004,
      "loss": 3.4543838500976562,
      "step": 66
    },
    {
      "epoch": 0.015314285714285714,
      "grad_norm": 1.335976004600525,
      "learning_rate": 0.0004125,
      "loss": 3.7046186923980713,
      "step": 67
    },
    {
      "epoch": 0.015542857142857143,
      "grad_norm": 1.289267897605896,
      "learning_rate": 0.00041875,
      "loss": 3.939502239227295,
      "step": 68
    },
    {
      "epoch": 0.01577142857142857,
      "grad_norm": 2.083914279937744,
      "learning_rate": 0.000425,
      "loss": 3.5302343368530273,
      "step": 69
    },
    {
      "epoch": 0.016,
      "grad_norm": 2.5294740200042725,
      "learning_rate": 0.00043125000000000005,
      "loss": 3.7287797927856445,
      "step": 70
    },
    {
      "epoch": 0.01622857142857143,
      "grad_norm": 1.5790722370147705,
      "learning_rate": 0.0004375,
      "loss": 3.8435840606689453,
      "step": 71
    },
    {
      "epoch": 0.016457142857142858,
      "grad_norm": 1.359749436378479,
      "learning_rate": 0.00044374999999999997,
      "loss": 3.7202670574188232,
      "step": 72
    },
    {
      "epoch": 0.016685714285714286,
      "grad_norm": 2.0596184730529785,
      "learning_rate": 0.00045000000000000004,
      "loss": 3.9328932762145996,
      "step": 73
    },
    {
      "epoch": 0.016914285714285715,
      "grad_norm": 1.3709889650344849,
      "learning_rate": 0.00045625,
      "loss": 3.814356803894043,
      "step": 74
    },
    {
      "epoch": 0.017142857142857144,
      "grad_norm": 1.8687207698822021,
      "learning_rate": 0.0004625,
      "loss": 4.310893535614014,
      "step": 75
    },
    {
      "epoch": 0.017371428571428572,
      "grad_norm": 1.153923511505127,
      "learning_rate": 0.00046875,
      "loss": 3.8177056312561035,
      "step": 76
    },
    {
      "epoch": 0.0176,
      "grad_norm": 1.5046632289886475,
      "learning_rate": 0.000475,
      "loss": 3.615830898284912,
      "step": 77
    },
    {
      "epoch": 0.01782857142857143,
      "grad_norm": 1.4608771800994873,
      "learning_rate": 0.00048125,
      "loss": 3.9287853240966797,
      "step": 78
    },
    {
      "epoch": 0.01805714285714286,
      "grad_norm": 1.7752225399017334,
      "learning_rate": 0.0004875,
      "loss": 3.922308921813965,
      "step": 79
    },
    {
      "epoch": 0.018285714285714287,
      "grad_norm": 1.3777114152908325,
      "learning_rate": 0.00049375,
      "loss": 3.29233980178833,
      "step": 80
    },
    {
      "epoch": 0.018514285714285716,
      "grad_norm": 1.2809195518493652,
      "learning_rate": 0.0005,
      "loss": 3.894735097885132,
      "step": 81
    },
    {
      "epoch": 0.018742857142857144,
      "grad_norm": 1.308272361755371,
      "learning_rate": 0.0005,
      "loss": 3.7446532249450684,
      "step": 82
    },
    {
      "epoch": 0.018971428571428573,
      "grad_norm": 1.4645415544509888,
      "learning_rate": 0.0005,
      "loss": 3.5220513343811035,
      "step": 83
    },
    {
      "epoch": 0.0192,
      "grad_norm": 1.4457155466079712,
      "learning_rate": 0.0005,
      "loss": 3.976024627685547,
      "step": 84
    },
    {
      "epoch": 0.019428571428571427,
      "grad_norm": 1.8887745141983032,
      "learning_rate": 0.0005,
      "loss": 3.8400659561157227,
      "step": 85
    },
    {
      "epoch": 0.019657142857142856,
      "grad_norm": 1.273095965385437,
      "learning_rate": 0.0005,
      "loss": 3.8880081176757812,
      "step": 86
    },
    {
      "epoch": 0.019885714285714284,
      "grad_norm": 1.3752671480178833,
      "learning_rate": 0.0005,
      "loss": 3.34137225151062,
      "step": 87
    },
    {
      "epoch": 0.020114285714285713,
      "grad_norm": 1.6106268167495728,
      "learning_rate": 0.0005,
      "loss": 3.6383838653564453,
      "step": 88
    },
    {
      "epoch": 0.02034285714285714,
      "grad_norm": 1.312519907951355,
      "learning_rate": 0.0005,
      "loss": 4.000592231750488,
      "step": 89
    },
    {
      "epoch": 0.02057142857142857,
      "grad_norm": 2.049669027328491,
      "learning_rate": 0.0005,
      "loss": 3.8618340492248535,
      "step": 90
    },
    {
      "epoch": 0.0208,
      "grad_norm": 1.25678551197052,
      "learning_rate": 0.0005,
      "loss": 3.582442283630371,
      "step": 91
    },
    {
      "epoch": 0.021028571428571428,
      "grad_norm": 1.126940131187439,
      "learning_rate": 0.0005,
      "loss": 3.6711580753326416,
      "step": 92
    },
    {
      "epoch": 0.021257142857142856,
      "grad_norm": 1.3346115350723267,
      "learning_rate": 0.0005,
      "loss": 4.0520124435424805,
      "step": 93
    },
    {
      "epoch": 0.021485714285714285,
      "grad_norm": 1.7092251777648926,
      "learning_rate": 0.0005,
      "loss": 3.9690990447998047,
      "step": 94
    },
    {
      "epoch": 0.021714285714285714,
      "grad_norm": 1.4492080211639404,
      "learning_rate": 0.0005,
      "loss": 3.7696619033813477,
      "step": 95
    },
    {
      "epoch": 0.021942857142857142,
      "grad_norm": 1.4325006008148193,
      "learning_rate": 0.0005,
      "loss": 3.5927414894104004,
      "step": 96
    },
    {
      "epoch": 0.02217142857142857,
      "grad_norm": 1.2093431949615479,
      "learning_rate": 0.0005,
      "loss": 3.970686197280884,
      "step": 97
    },
    {
      "epoch": 0.0224,
      "grad_norm": 1.3890284299850464,
      "learning_rate": 0.0005,
      "loss": 4.013213157653809,
      "step": 98
    },
    {
      "epoch": 0.02262857142857143,
      "grad_norm": 1.4635905027389526,
      "learning_rate": 0.0005,
      "loss": 3.9039266109466553,
      "step": 99
    },
    {
      "epoch": 0.022857142857142857,
      "grad_norm": 1.4292216300964355,
      "learning_rate": 0.0005,
      "loss": 3.796912670135498,
      "step": 100
    },
    {
      "epoch": 0.023085714285714286,
      "grad_norm": 1.6453803777694702,
      "learning_rate": 0.0005,
      "loss": 3.720942735671997,
      "step": 101
    },
    {
      "epoch": 0.023314285714285714,
      "grad_norm": 1.2871754169464111,
      "learning_rate": 0.0005,
      "loss": 4.04292106628418,
      "step": 102
    },
    {
      "epoch": 0.023542857142857143,
      "grad_norm": 1.338755488395691,
      "learning_rate": 0.0005,
      "loss": 3.957151412963867,
      "step": 103
    },
    {
      "epoch": 0.023771428571428572,
      "grad_norm": 2.0366430282592773,
      "learning_rate": 0.0005,
      "loss": 4.0411577224731445,
      "step": 104
    },
    {
      "epoch": 0.024,
      "grad_norm": 1.3768296241760254,
      "learning_rate": 0.0005,
      "loss": 3.9278650283813477,
      "step": 105
    },
    {
      "epoch": 0.02422857142857143,
      "grad_norm": 1.3399204015731812,
      "learning_rate": 0.0005,
      "loss": 3.977231025695801,
      "step": 106
    },
    {
      "epoch": 0.024457142857142858,
      "grad_norm": 1.357668399810791,
      "learning_rate": 0.0005,
      "loss": 3.676583766937256,
      "step": 107
    },
    {
      "epoch": 0.024685714285714287,
      "grad_norm": 1.545405626296997,
      "learning_rate": 0.0005,
      "loss": 3.5663325786590576,
      "step": 108
    },
    {
      "epoch": 0.024914285714285715,
      "grad_norm": 2.273883819580078,
      "learning_rate": 0.0005,
      "loss": 3.7200443744659424,
      "step": 109
    },
    {
      "epoch": 0.025142857142857144,
      "grad_norm": 1.6109744310379028,
      "learning_rate": 0.0005,
      "loss": 3.7702393531799316,
      "step": 110
    },
    {
      "epoch": 0.025371428571428573,
      "grad_norm": 1.2182873487472534,
      "learning_rate": 0.0005,
      "loss": 3.923719644546509,
      "step": 111
    },
    {
      "epoch": 0.0256,
      "grad_norm": 1.6417268514633179,
      "learning_rate": 0.0005,
      "loss": 3.812298536300659,
      "step": 112
    },
    {
      "epoch": 0.02582857142857143,
      "grad_norm": 1.689680576324463,
      "learning_rate": 0.0005,
      "loss": 4.318823337554932,
      "step": 113
    },
    {
      "epoch": 0.02605714285714286,
      "grad_norm": 1.5286896228790283,
      "learning_rate": 0.0005,
      "loss": 3.7507076263427734,
      "step": 114
    },
    {
      "epoch": 0.026285714285714287,
      "grad_norm": 1.1639305353164673,
      "learning_rate": 0.0005,
      "loss": 3.700477123260498,
      "step": 115
    },
    {
      "epoch": 0.026514285714285716,
      "grad_norm": 1.2797795534133911,
      "learning_rate": 0.0005,
      "loss": 3.6179392337799072,
      "step": 116
    },
    {
      "epoch": 0.02674285714285714,
      "grad_norm": 1.4961555004119873,
      "learning_rate": 0.0005,
      "loss": 4.115945816040039,
      "step": 117
    },
    {
      "epoch": 0.02697142857142857,
      "grad_norm": 1.220071792602539,
      "learning_rate": 0.0005,
      "loss": 3.7336971759796143,
      "step": 118
    },
    {
      "epoch": 0.0272,
      "grad_norm": 1.267430067062378,
      "learning_rate": 0.0005,
      "loss": 3.637418746948242,
      "step": 119
    },
    {
      "epoch": 0.027428571428571427,
      "grad_norm": 1.5194393396377563,
      "learning_rate": 0.0005,
      "loss": 3.757481336593628,
      "step": 120
    },
    {
      "epoch": 0.027657142857142856,
      "grad_norm": 1.199604868888855,
      "learning_rate": 0.0005,
      "loss": 3.811464548110962,
      "step": 121
    },
    {
      "epoch": 0.027885714285714285,
      "grad_norm": 1.1793861389160156,
      "learning_rate": 0.0005,
      "loss": 3.932215690612793,
      "step": 122
    },
    {
      "epoch": 0.028114285714285713,
      "grad_norm": 1.6691200733184814,
      "learning_rate": 0.0005,
      "loss": 3.5433497428894043,
      "step": 123
    },
    {
      "epoch": 0.028342857142857142,
      "grad_norm": 1.331429362297058,
      "learning_rate": 0.0005,
      "loss": 3.6708571910858154,
      "step": 124
    },
    {
      "epoch": 0.02857142857142857,
      "grad_norm": 1.350144863128662,
      "learning_rate": 0.0005,
      "loss": 3.609684944152832,
      "step": 125
    },
    {
      "epoch": 0.0288,
      "grad_norm": 1.1860101222991943,
      "learning_rate": 0.0005,
      "loss": 3.4407474994659424,
      "step": 126
    },
    {
      "epoch": 0.029028571428571428,
      "grad_norm": 1.6895368099212646,
      "learning_rate": 0.0005,
      "loss": 3.834057331085205,
      "step": 127
    },
    {
      "epoch": 0.029257142857142857,
      "grad_norm": 1.567077398300171,
      "learning_rate": 0.0005,
      "loss": 3.902327060699463,
      "step": 128
    },
    {
      "epoch": 0.029485714285714285,
      "grad_norm": 1.353013038635254,
      "learning_rate": 0.0005,
      "loss": 3.8366856575012207,
      "step": 129
    },
    {
      "epoch": 0.029714285714285714,
      "grad_norm": 1.2551172971725464,
      "learning_rate": 0.0005,
      "loss": 3.488246440887451,
      "step": 130
    },
    {
      "epoch": 0.029942857142857143,
      "grad_norm": 1.4480103254318237,
      "learning_rate": 0.0005,
      "loss": 3.575624942779541,
      "step": 131
    },
    {
      "epoch": 0.03017142857142857,
      "grad_norm": 1.373744249343872,
      "learning_rate": 0.0005,
      "loss": 4.03697395324707,
      "step": 132
    },
    {
      "epoch": 0.0304,
      "grad_norm": 1.2239898443222046,
      "learning_rate": 0.0005,
      "loss": 3.7191615104675293,
      "step": 133
    },
    {
      "epoch": 0.03062857142857143,
      "grad_norm": 1.2791234254837036,
      "learning_rate": 0.0005,
      "loss": 4.070167064666748,
      "step": 134
    },
    {
      "epoch": 0.030857142857142857,
      "grad_norm": 1.442443609237671,
      "learning_rate": 0.0005,
      "loss": 3.7328410148620605,
      "step": 135
    },
    {
      "epoch": 0.031085714285714286,
      "grad_norm": 1.3792619705200195,
      "learning_rate": 0.0005,
      "loss": 3.666830539703369,
      "step": 136
    },
    {
      "epoch": 0.03131428571428571,
      "grad_norm": 1.1862142086029053,
      "learning_rate": 0.0005,
      "loss": 3.8953182697296143,
      "step": 137
    },
    {
      "epoch": 0.03154285714285714,
      "grad_norm": 1.9386588335037231,
      "learning_rate": 0.0005,
      "loss": 4.045220375061035,
      "step": 138
    },
    {
      "epoch": 0.03177142857142857,
      "grad_norm": 1.4658712148666382,
      "learning_rate": 0.0005,
      "loss": 3.959496021270752,
      "step": 139
    },
    {
      "epoch": 0.032,
      "grad_norm": 1.3294095993041992,
      "learning_rate": 0.0005,
      "loss": 3.8833823204040527,
      "step": 140
    },
    {
      "epoch": 0.032228571428571426,
      "grad_norm": 1.5445894002914429,
      "learning_rate": 0.0005,
      "loss": 4.027015209197998,
      "step": 141
    },
    {
      "epoch": 0.03245714285714286,
      "grad_norm": 1.460942029953003,
      "learning_rate": 0.0005,
      "loss": 4.1589484214782715,
      "step": 142
    },
    {
      "epoch": 0.03268571428571428,
      "grad_norm": 1.2193883657455444,
      "learning_rate": 0.0005,
      "loss": 3.85184907913208,
      "step": 143
    },
    {
      "epoch": 0.032914285714285715,
      "grad_norm": 1.2044373750686646,
      "learning_rate": 0.0005,
      "loss": 3.8813719749450684,
      "step": 144
    },
    {
      "epoch": 0.03314285714285714,
      "grad_norm": 1.1442755460739136,
      "learning_rate": 0.0005,
      "loss": 3.8071022033691406,
      "step": 145
    },
    {
      "epoch": 0.03337142857142857,
      "grad_norm": 1.7308660745620728,
      "learning_rate": 0.0005,
      "loss": 3.9762532711029053,
      "step": 146
    },
    {
      "epoch": 0.0336,
      "grad_norm": 1.8234037160873413,
      "learning_rate": 0.0005,
      "loss": 3.357025623321533,
      "step": 147
    },
    {
      "epoch": 0.03382857142857143,
      "grad_norm": 2.080044746398926,
      "learning_rate": 0.0005,
      "loss": 3.9496524333953857,
      "step": 148
    },
    {
      "epoch": 0.034057142857142855,
      "grad_norm": 1.2784281969070435,
      "learning_rate": 0.0005,
      "loss": 3.740875720977783,
      "step": 149
    },
    {
      "epoch": 0.03428571428571429,
      "grad_norm": 1.7990317344665527,
      "learning_rate": 0.0005,
      "loss": 4.126687049865723,
      "step": 150
    },
    {
      "epoch": 0.03451428571428571,
      "grad_norm": 1.3636387586593628,
      "learning_rate": 0.0005,
      "loss": 3.542611837387085,
      "step": 151
    },
    {
      "epoch": 0.034742857142857145,
      "grad_norm": 1.3541808128356934,
      "learning_rate": 0.0005,
      "loss": 3.850137710571289,
      "step": 152
    },
    {
      "epoch": 0.03497142857142857,
      "grad_norm": 1.6047935485839844,
      "learning_rate": 0.0005,
      "loss": 3.638369083404541,
      "step": 153
    },
    {
      "epoch": 0.0352,
      "grad_norm": 1.5367470979690552,
      "learning_rate": 0.0005,
      "loss": 3.971201181411743,
      "step": 154
    },
    {
      "epoch": 0.03542857142857143,
      "grad_norm": 1.898033857345581,
      "learning_rate": 0.0005,
      "loss": 3.715546131134033,
      "step": 155
    },
    {
      "epoch": 0.03565714285714286,
      "grad_norm": 1.3745911121368408,
      "learning_rate": 0.0005,
      "loss": 3.9482340812683105,
      "step": 156
    },
    {
      "epoch": 0.035885714285714285,
      "grad_norm": 1.1635035276412964,
      "learning_rate": 0.0005,
      "loss": 3.5926194190979004,
      "step": 157
    },
    {
      "epoch": 0.03611428571428572,
      "grad_norm": 1.41801118850708,
      "learning_rate": 0.0005,
      "loss": 3.8779587745666504,
      "step": 158
    },
    {
      "epoch": 0.03634285714285714,
      "grad_norm": 1.6658660173416138,
      "learning_rate": 0.0005,
      "loss": 3.6878983974456787,
      "step": 159
    },
    {
      "epoch": 0.036571428571428574,
      "grad_norm": 1.215263843536377,
      "learning_rate": 0.0005,
      "loss": 3.5479817390441895,
      "step": 160
    },
    {
      "epoch": 0.0368,
      "grad_norm": 2.1402087211608887,
      "learning_rate": 0.0005,
      "loss": 3.3771119117736816,
      "step": 161
    },
    {
      "epoch": 0.03702857142857143,
      "grad_norm": 1.2970267534255981,
      "learning_rate": 0.0005,
      "loss": 3.601621627807617,
      "step": 162
    },
    {
      "epoch": 0.03725714285714286,
      "grad_norm": 1.631191372871399,
      "learning_rate": 0.0005,
      "loss": 3.610800266265869,
      "step": 163
    },
    {
      "epoch": 0.03748571428571429,
      "grad_norm": 1.6436046361923218,
      "learning_rate": 0.0005,
      "loss": 3.787407875061035,
      "step": 164
    },
    {
      "epoch": 0.037714285714285714,
      "grad_norm": 1.1235754489898682,
      "learning_rate": 0.0005,
      "loss": 3.7069811820983887,
      "step": 165
    },
    {
      "epoch": 0.037942857142857146,
      "grad_norm": 1.2559431791305542,
      "learning_rate": 0.0005,
      "loss": 3.422703742980957,
      "step": 166
    },
    {
      "epoch": 0.03817142857142857,
      "grad_norm": 1.588852882385254,
      "learning_rate": 0.0005,
      "loss": 3.5424790382385254,
      "step": 167
    },
    {
      "epoch": 0.0384,
      "grad_norm": 1.4449278116226196,
      "learning_rate": 0.0005,
      "loss": 3.4990057945251465,
      "step": 168
    },
    {
      "epoch": 0.03862857142857143,
      "grad_norm": 1.5056250095367432,
      "learning_rate": 0.0005,
      "loss": 3.7876057624816895,
      "step": 169
    },
    {
      "epoch": 0.038857142857142854,
      "grad_norm": 1.1734257936477661,
      "learning_rate": 0.0005,
      "loss": 3.878467321395874,
      "step": 170
    },
    {
      "epoch": 0.039085714285714286,
      "grad_norm": 1.418595314025879,
      "learning_rate": 0.0005,
      "loss": 3.913982391357422,
      "step": 171
    },
    {
      "epoch": 0.03931428571428571,
      "grad_norm": 1.3645241260528564,
      "learning_rate": 0.0005,
      "loss": 3.650054931640625,
      "step": 172
    },
    {
      "epoch": 0.039542857142857143,
      "grad_norm": 1.32620108127594,
      "learning_rate": 0.0005,
      "loss": 3.8011767864227295,
      "step": 173
    },
    {
      "epoch": 0.03977142857142857,
      "grad_norm": 1.3934112787246704,
      "learning_rate": 0.0005,
      "loss": 3.4510498046875,
      "step": 174
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.280582070350647,
      "learning_rate": 0.0005,
      "loss": 3.6263346672058105,
      "step": 175
    },
    {
      "epoch": 0.040228571428571426,
      "grad_norm": 1.401004433631897,
      "learning_rate": 0.0005,
      "loss": 3.4766197204589844,
      "step": 176
    },
    {
      "epoch": 0.04045714285714286,
      "grad_norm": 1.412644624710083,
      "learning_rate": 0.0005,
      "loss": 3.868211030960083,
      "step": 177
    },
    {
      "epoch": 0.04068571428571428,
      "grad_norm": 1.0835626125335693,
      "learning_rate": 0.0005,
      "loss": 3.881743907928467,
      "step": 178
    },
    {
      "epoch": 0.040914285714285716,
      "grad_norm": 1.1708346605300903,
      "learning_rate": 0.0005,
      "loss": 3.8134355545043945,
      "step": 179
    },
    {
      "epoch": 0.04114285714285714,
      "grad_norm": 1.8206517696380615,
      "learning_rate": 0.0005,
      "loss": 3.888252019882202,
      "step": 180
    },
    {
      "epoch": 0.04137142857142857,
      "grad_norm": 1.2196844816207886,
      "learning_rate": 0.0005,
      "loss": 3.787541389465332,
      "step": 181
    },
    {
      "epoch": 0.0416,
      "grad_norm": 1.4441864490509033,
      "learning_rate": 0.0005,
      "loss": 3.8336448669433594,
      "step": 182
    },
    {
      "epoch": 0.04182857142857143,
      "grad_norm": 1.4670696258544922,
      "learning_rate": 0.0005,
      "loss": 3.5021114349365234,
      "step": 183
    },
    {
      "epoch": 0.042057142857142855,
      "grad_norm": 1.1875871419906616,
      "learning_rate": 0.0005,
      "loss": 3.879133701324463,
      "step": 184
    },
    {
      "epoch": 0.04228571428571429,
      "grad_norm": 1.3518619537353516,
      "learning_rate": 0.0005,
      "loss": 3.9073567390441895,
      "step": 185
    },
    {
      "epoch": 0.04251428571428571,
      "grad_norm": 1.7341628074645996,
      "learning_rate": 0.0005,
      "loss": 4.259081840515137,
      "step": 186
    },
    {
      "epoch": 0.042742857142857145,
      "grad_norm": 1.4350199699401855,
      "learning_rate": 0.0005,
      "loss": 3.4714653491973877,
      "step": 187
    },
    {
      "epoch": 0.04297142857142857,
      "grad_norm": 1.4874223470687866,
      "learning_rate": 0.0005,
      "loss": 3.703986883163452,
      "step": 188
    },
    {
      "epoch": 0.0432,
      "grad_norm": 1.4151960611343384,
      "learning_rate": 0.0005,
      "loss": 3.6947708129882812,
      "step": 189
    },
    {
      "epoch": 0.04342857142857143,
      "grad_norm": 1.157325267791748,
      "learning_rate": 0.0005,
      "loss": 3.737499713897705,
      "step": 190
    },
    {
      "epoch": 0.04365714285714286,
      "grad_norm": 1.963815450668335,
      "learning_rate": 0.0005,
      "loss": 3.424140453338623,
      "step": 191
    },
    {
      "epoch": 0.043885714285714285,
      "grad_norm": 3.9101078510284424,
      "learning_rate": 0.0005,
      "loss": 3.53133487701416,
      "step": 192
    },
    {
      "epoch": 0.04411428571428572,
      "grad_norm": 1.4522440433502197,
      "learning_rate": 0.0005,
      "loss": 3.6986966133117676,
      "step": 193
    },
    {
      "epoch": 0.04434285714285714,
      "grad_norm": 1.2475334405899048,
      "learning_rate": 0.0005,
      "loss": 3.687687397003174,
      "step": 194
    },
    {
      "epoch": 0.044571428571428574,
      "grad_norm": 1.3344810009002686,
      "learning_rate": 0.0005,
      "loss": 3.8548412322998047,
      "step": 195
    },
    {
      "epoch": 0.0448,
      "grad_norm": 1.3346023559570312,
      "learning_rate": 0.0005,
      "loss": 3.802199602127075,
      "step": 196
    },
    {
      "epoch": 0.04502857142857143,
      "grad_norm": 1.4111653566360474,
      "learning_rate": 0.0005,
      "loss": 4.057753562927246,
      "step": 197
    },
    {
      "epoch": 0.04525714285714286,
      "grad_norm": 1.7387349605560303,
      "learning_rate": 0.0005,
      "loss": 3.5669050216674805,
      "step": 198
    },
    {
      "epoch": 0.04548571428571429,
      "grad_norm": 1.2928766012191772,
      "learning_rate": 0.0005,
      "loss": 3.9654312133789062,
      "step": 199
    },
    {
      "epoch": 0.045714285714285714,
      "grad_norm": 1.7153937816619873,
      "learning_rate": 0.0005,
      "loss": 3.6818463802337646,
      "step": 200
    },
    {
      "epoch": 0.04594285714285714,
      "grad_norm": 1.423218846321106,
      "learning_rate": 0.0005,
      "loss": 3.484154462814331,
      "step": 201
    },
    {
      "epoch": 0.04617142857142857,
      "grad_norm": 1.5035536289215088,
      "learning_rate": 0.0005,
      "loss": 3.517792224884033,
      "step": 202
    },
    {
      "epoch": 0.0464,
      "grad_norm": 1.79793381690979,
      "learning_rate": 0.0005,
      "loss": 3.4553561210632324,
      "step": 203
    },
    {
      "epoch": 0.04662857142857143,
      "grad_norm": 1.5017178058624268,
      "learning_rate": 0.0005,
      "loss": 3.707796096801758,
      "step": 204
    },
    {
      "epoch": 0.046857142857142854,
      "grad_norm": 1.6542028188705444,
      "learning_rate": 0.0005,
      "loss": 3.503617763519287,
      "step": 205
    },
    {
      "epoch": 0.047085714285714286,
      "grad_norm": 1.7485281229019165,
      "learning_rate": 0.0005,
      "loss": 3.892571449279785,
      "step": 206
    },
    {
      "epoch": 0.04731428571428571,
      "grad_norm": 1.3079618215560913,
      "learning_rate": 0.0005,
      "loss": 3.425382137298584,
      "step": 207
    },
    {
      "epoch": 0.047542857142857144,
      "grad_norm": 1.5893138647079468,
      "learning_rate": 0.0005,
      "loss": 3.848681926727295,
      "step": 208
    },
    {
      "epoch": 0.04777142857142857,
      "grad_norm": 1.2473441362380981,
      "learning_rate": 0.0005,
      "loss": 3.4193952083587646,
      "step": 209
    },
    {
      "epoch": 0.048,
      "grad_norm": 1.3488614559173584,
      "learning_rate": 0.0005,
      "loss": 3.8910179138183594,
      "step": 210
    },
    {
      "epoch": 0.048228571428571426,
      "grad_norm": 1.5454046726226807,
      "learning_rate": 0.0005,
      "loss": 3.888857841491699,
      "step": 211
    },
    {
      "epoch": 0.04845714285714286,
      "grad_norm": 1.5776426792144775,
      "learning_rate": 0.0005,
      "loss": 4.020862579345703,
      "step": 212
    },
    {
      "epoch": 0.048685714285714284,
      "grad_norm": 1.569182276725769,
      "learning_rate": 0.0005,
      "loss": 3.487306594848633,
      "step": 213
    },
    {
      "epoch": 0.048914285714285716,
      "grad_norm": 1.1732338666915894,
      "learning_rate": 0.0005,
      "loss": 3.8554978370666504,
      "step": 214
    },
    {
      "epoch": 0.04914285714285714,
      "grad_norm": 1.344928503036499,
      "learning_rate": 0.0005,
      "loss": 3.543522834777832,
      "step": 215
    },
    {
      "epoch": 0.04937142857142857,
      "grad_norm": 1.2798022031784058,
      "learning_rate": 0.0005,
      "loss": 4.118750095367432,
      "step": 216
    },
    {
      "epoch": 0.0496,
      "grad_norm": 1.0370323657989502,
      "learning_rate": 0.0005,
      "loss": 4.011840343475342,
      "step": 217
    },
    {
      "epoch": 0.04982857142857143,
      "grad_norm": 1.3434882164001465,
      "learning_rate": 0.0005,
      "loss": 3.704288959503174,
      "step": 218
    },
    {
      "epoch": 0.050057142857142856,
      "grad_norm": 1.4296915531158447,
      "learning_rate": 0.0005,
      "loss": 3.6497907638549805,
      "step": 219
    },
    {
      "epoch": 0.05028571428571429,
      "grad_norm": 2.890810012817383,
      "learning_rate": 0.0005,
      "loss": 3.7581899166107178,
      "step": 220
    },
    {
      "epoch": 0.05051428571428571,
      "grad_norm": 1.6530475616455078,
      "learning_rate": 0.0005,
      "loss": 3.6168501377105713,
      "step": 221
    },
    {
      "epoch": 0.050742857142857145,
      "grad_norm": 1.7184935808181763,
      "learning_rate": 0.0005,
      "loss": 3.667142391204834,
      "step": 222
    },
    {
      "epoch": 0.05097142857142857,
      "grad_norm": 1.349421739578247,
      "learning_rate": 0.0005,
      "loss": 3.526566982269287,
      "step": 223
    },
    {
      "epoch": 0.0512,
      "grad_norm": 1.5553141832351685,
      "learning_rate": 0.0005,
      "loss": 3.8120007514953613,
      "step": 224
    },
    {
      "epoch": 0.05142857142857143,
      "grad_norm": 1.6979200839996338,
      "learning_rate": 0.0005,
      "loss": 3.955761432647705,
      "step": 225
    },
    {
      "epoch": 0.05165714285714286,
      "grad_norm": 1.467490792274475,
      "learning_rate": 0.0005,
      "loss": 3.295105218887329,
      "step": 226
    },
    {
      "epoch": 0.051885714285714285,
      "grad_norm": 1.5187363624572754,
      "learning_rate": 0.0005,
      "loss": 3.9215359687805176,
      "step": 227
    },
    {
      "epoch": 0.05211428571428572,
      "grad_norm": 1.633123517036438,
      "learning_rate": 0.0005,
      "loss": 3.986494779586792,
      "step": 228
    },
    {
      "epoch": 0.05234285714285714,
      "grad_norm": 1.2633056640625,
      "learning_rate": 0.0005,
      "loss": 3.6235883235931396,
      "step": 229
    },
    {
      "epoch": 0.052571428571428575,
      "grad_norm": 1.2066441774368286,
      "learning_rate": 0.0005,
      "loss": 3.982962131500244,
      "step": 230
    },
    {
      "epoch": 0.0528,
      "grad_norm": 1.4105767011642456,
      "learning_rate": 0.0005,
      "loss": 3.339445114135742,
      "step": 231
    },
    {
      "epoch": 0.05302857142857143,
      "grad_norm": 1.4044400453567505,
      "learning_rate": 0.0005,
      "loss": 3.514371633529663,
      "step": 232
    },
    {
      "epoch": 0.05325714285714286,
      "grad_norm": 1.2505738735198975,
      "learning_rate": 0.0005,
      "loss": 3.604393482208252,
      "step": 233
    },
    {
      "epoch": 0.05348571428571428,
      "grad_norm": 1.620950698852539,
      "learning_rate": 0.0005,
      "loss": 3.904658555984497,
      "step": 234
    },
    {
      "epoch": 0.053714285714285714,
      "grad_norm": 1.230180025100708,
      "learning_rate": 0.0005,
      "loss": 3.8683152198791504,
      "step": 235
    },
    {
      "epoch": 0.05394285714285714,
      "grad_norm": 1.2296142578125,
      "learning_rate": 0.0005,
      "loss": 3.6946873664855957,
      "step": 236
    },
    {
      "epoch": 0.05417142857142857,
      "grad_norm": 1.408909559249878,
      "learning_rate": 0.0005,
      "loss": 3.6184024810791016,
      "step": 237
    },
    {
      "epoch": 0.0544,
      "grad_norm": 1.2091312408447266,
      "learning_rate": 0.0005,
      "loss": 3.8990769386291504,
      "step": 238
    },
    {
      "epoch": 0.05462857142857143,
      "grad_norm": 1.8112094402313232,
      "learning_rate": 0.0005,
      "loss": 3.708108425140381,
      "step": 239
    },
    {
      "epoch": 0.054857142857142854,
      "grad_norm": 1.5574519634246826,
      "learning_rate": 0.0005,
      "loss": 3.653510570526123,
      "step": 240
    },
    {
      "epoch": 0.055085714285714286,
      "grad_norm": 1.1765307188034058,
      "learning_rate": 0.0005,
      "loss": 3.885732889175415,
      "step": 241
    },
    {
      "epoch": 0.05531428571428571,
      "grad_norm": 1.4060622453689575,
      "learning_rate": 0.0005,
      "loss": 3.302473783493042,
      "step": 242
    },
    {
      "epoch": 0.055542857142857144,
      "grad_norm": 1.453696370124817,
      "learning_rate": 0.0005,
      "loss": 3.7380917072296143,
      "step": 243
    },
    {
      "epoch": 0.05577142857142857,
      "grad_norm": 1.3317773342132568,
      "learning_rate": 0.0005,
      "loss": 3.7037124633789062,
      "step": 244
    },
    {
      "epoch": 0.056,
      "grad_norm": 1.4232923984527588,
      "learning_rate": 0.0005,
      "loss": 3.923936367034912,
      "step": 245
    },
    {
      "epoch": 0.056228571428571426,
      "grad_norm": 1.2320622205734253,
      "learning_rate": 0.0005,
      "loss": 3.639322280883789,
      "step": 246
    },
    {
      "epoch": 0.05645714285714286,
      "grad_norm": 1.4088433980941772,
      "learning_rate": 0.0005,
      "loss": 4.350565433502197,
      "step": 247
    },
    {
      "epoch": 0.056685714285714284,
      "grad_norm": 1.5011903047561646,
      "learning_rate": 0.0005,
      "loss": 3.8753037452697754,
      "step": 248
    },
    {
      "epoch": 0.056914285714285716,
      "grad_norm": 1.173302412033081,
      "learning_rate": 0.0005,
      "loss": 4.067836761474609,
      "step": 249
    },
    {
      "epoch": 0.05714285714285714,
      "grad_norm": 2.2286272048950195,
      "learning_rate": 0.0005,
      "loss": 3.1932053565979004,
      "step": 250
    },
    {
      "epoch": 0.05737142857142857,
      "grad_norm": 1.7074507474899292,
      "learning_rate": 0.0005,
      "loss": 3.8071775436401367,
      "step": 251
    },
    {
      "epoch": 0.0576,
      "grad_norm": 1.3556994199752808,
      "learning_rate": 0.0005,
      "loss": 3.833752155303955,
      "step": 252
    },
    {
      "epoch": 0.05782857142857143,
      "grad_norm": 1.3615472316741943,
      "learning_rate": 0.0005,
      "loss": 3.8785057067871094,
      "step": 253
    },
    {
      "epoch": 0.058057142857142856,
      "grad_norm": 1.440555453300476,
      "learning_rate": 0.0005,
      "loss": 3.636813163757324,
      "step": 254
    },
    {
      "epoch": 0.05828571428571429,
      "grad_norm": 1.3739646673202515,
      "learning_rate": 0.0005,
      "loss": 4.061012268066406,
      "step": 255
    },
    {
      "epoch": 0.05851428571428571,
      "grad_norm": 1.2113922834396362,
      "learning_rate": 0.0005,
      "loss": 3.698965549468994,
      "step": 256
    },
    {
      "epoch": 0.058742857142857145,
      "grad_norm": 1.3796567916870117,
      "learning_rate": 0.0005,
      "loss": 3.645700454711914,
      "step": 257
    },
    {
      "epoch": 0.05897142857142857,
      "grad_norm": 1.5154588222503662,
      "learning_rate": 0.0005,
      "loss": 3.794344186782837,
      "step": 258
    },
    {
      "epoch": 0.0592,
      "grad_norm": 1.3863497972488403,
      "learning_rate": 0.0005,
      "loss": 3.2520129680633545,
      "step": 259
    },
    {
      "epoch": 0.05942857142857143,
      "grad_norm": 1.5560667514801025,
      "learning_rate": 0.0005,
      "loss": 3.8902390003204346,
      "step": 260
    },
    {
      "epoch": 0.05965714285714286,
      "grad_norm": 1.4990392923355103,
      "learning_rate": 0.0005,
      "loss": 3.1612653732299805,
      "step": 261
    },
    {
      "epoch": 0.059885714285714285,
      "grad_norm": 1.375090479850769,
      "learning_rate": 0.0005,
      "loss": 3.8060696125030518,
      "step": 262
    },
    {
      "epoch": 0.06011428571428572,
      "grad_norm": 1.252989411354065,
      "learning_rate": 0.0005,
      "loss": 3.3702549934387207,
      "step": 263
    },
    {
      "epoch": 0.06034285714285714,
      "grad_norm": 1.7148548364639282,
      "learning_rate": 0.0005,
      "loss": 3.7294089794158936,
      "step": 264
    },
    {
      "epoch": 0.060571428571428575,
      "grad_norm": 1.4030416011810303,
      "learning_rate": 0.0005,
      "loss": 3.277472972869873,
      "step": 265
    },
    {
      "epoch": 0.0608,
      "grad_norm": 1.521823763847351,
      "learning_rate": 0.0005,
      "loss": 3.8827903270721436,
      "step": 266
    },
    {
      "epoch": 0.06102857142857143,
      "grad_norm": 1.806654453277588,
      "learning_rate": 0.0005,
      "loss": 3.4610791206359863,
      "step": 267
    },
    {
      "epoch": 0.06125714285714286,
      "grad_norm": 1.4314630031585693,
      "learning_rate": 0.0005,
      "loss": 3.561121940612793,
      "step": 268
    },
    {
      "epoch": 0.06148571428571428,
      "grad_norm": 1.2702704668045044,
      "learning_rate": 0.0005,
      "loss": 3.3101296424865723,
      "step": 269
    },
    {
      "epoch": 0.061714285714285715,
      "grad_norm": 1.17933988571167,
      "learning_rate": 0.0005,
      "loss": 3.356336832046509,
      "step": 270
    },
    {
      "epoch": 0.06194285714285714,
      "grad_norm": 2.2510182857513428,
      "learning_rate": 0.0005,
      "loss": 3.678971767425537,
      "step": 271
    },
    {
      "epoch": 0.06217142857142857,
      "grad_norm": 1.6220695972442627,
      "learning_rate": 0.0005,
      "loss": 3.520798921585083,
      "step": 272
    },
    {
      "epoch": 0.0624,
      "grad_norm": 1.2107014656066895,
      "learning_rate": 0.0005,
      "loss": 3.8358020782470703,
      "step": 273
    },
    {
      "epoch": 0.06262857142857142,
      "grad_norm": 1.3935668468475342,
      "learning_rate": 0.0005,
      "loss": 3.723299741744995,
      "step": 274
    },
    {
      "epoch": 0.06285714285714286,
      "grad_norm": 1.3622251749038696,
      "learning_rate": 0.0005,
      "loss": 3.6107406616210938,
      "step": 275
    },
    {
      "epoch": 0.06308571428571429,
      "grad_norm": 1.6499896049499512,
      "learning_rate": 0.0005,
      "loss": 3.767772674560547,
      "step": 276
    },
    {
      "epoch": 0.06331428571428571,
      "grad_norm": 1.5945773124694824,
      "learning_rate": 0.0005,
      "loss": 3.9383435249328613,
      "step": 277
    },
    {
      "epoch": 0.06354285714285714,
      "grad_norm": 1.3597259521484375,
      "learning_rate": 0.0005,
      "loss": 3.4230875968933105,
      "step": 278
    },
    {
      "epoch": 0.06377142857142858,
      "grad_norm": 1.5023459196090698,
      "learning_rate": 0.0005,
      "loss": 3.7994801998138428,
      "step": 279
    },
    {
      "epoch": 0.064,
      "grad_norm": 1.4786256551742554,
      "learning_rate": 0.0005,
      "loss": 3.975897789001465,
      "step": 280
    },
    {
      "epoch": 0.06422857142857143,
      "grad_norm": 1.1918445825576782,
      "learning_rate": 0.0005,
      "loss": 4.020575523376465,
      "step": 281
    },
    {
      "epoch": 0.06445714285714285,
      "grad_norm": 1.6465709209442139,
      "learning_rate": 0.0005,
      "loss": 3.498065710067749,
      "step": 282
    },
    {
      "epoch": 0.06468571428571429,
      "grad_norm": 1.3304964303970337,
      "learning_rate": 0.0005,
      "loss": 3.911170482635498,
      "step": 283
    },
    {
      "epoch": 0.06491428571428572,
      "grad_norm": 1.456355094909668,
      "learning_rate": 0.0005,
      "loss": 3.738755226135254,
      "step": 284
    },
    {
      "epoch": 0.06514285714285714,
      "grad_norm": 1.2989567518234253,
      "learning_rate": 0.0005,
      "loss": 3.63086199760437,
      "step": 285
    },
    {
      "epoch": 0.06537142857142857,
      "grad_norm": 1.189714789390564,
      "learning_rate": 0.0005,
      "loss": 3.5887255668640137,
      "step": 286
    },
    {
      "epoch": 0.0656,
      "grad_norm": 1.437333583831787,
      "learning_rate": 0.0005,
      "loss": 3.6351752281188965,
      "step": 287
    },
    {
      "epoch": 0.06582857142857143,
      "grad_norm": 1.387915849685669,
      "learning_rate": 0.0005,
      "loss": 3.689579486846924,
      "step": 288
    },
    {
      "epoch": 0.06605714285714286,
      "grad_norm": 1.4056761264801025,
      "learning_rate": 0.0005,
      "loss": 3.5816879272460938,
      "step": 289
    },
    {
      "epoch": 0.06628571428571428,
      "grad_norm": 1.5918084383010864,
      "learning_rate": 0.0005,
      "loss": 3.8195278644561768,
      "step": 290
    },
    {
      "epoch": 0.06651428571428572,
      "grad_norm": 1.728590965270996,
      "learning_rate": 0.0005,
      "loss": 3.832324504852295,
      "step": 291
    },
    {
      "epoch": 0.06674285714285715,
      "grad_norm": 1.3873534202575684,
      "learning_rate": 0.0005,
      "loss": 3.986267566680908,
      "step": 292
    },
    {
      "epoch": 0.06697142857142857,
      "grad_norm": 1.542543888092041,
      "learning_rate": 0.0005,
      "loss": 3.8429737091064453,
      "step": 293
    },
    {
      "epoch": 0.0672,
      "grad_norm": 1.5978310108184814,
      "learning_rate": 0.0005,
      "loss": 3.7461585998535156,
      "step": 294
    },
    {
      "epoch": 0.06742857142857143,
      "grad_norm": 1.5184342861175537,
      "learning_rate": 0.0005,
      "loss": 3.557197093963623,
      "step": 295
    },
    {
      "epoch": 0.06765714285714286,
      "grad_norm": 1.206478238105774,
      "learning_rate": 0.0005,
      "loss": 3.8454251289367676,
      "step": 296
    },
    {
      "epoch": 0.06788571428571429,
      "grad_norm": 2.7141804695129395,
      "learning_rate": 0.0005,
      "loss": 3.874232769012451,
      "step": 297
    },
    {
      "epoch": 0.06811428571428571,
      "grad_norm": 2.2806174755096436,
      "learning_rate": 0.0005,
      "loss": 3.4598536491394043,
      "step": 298
    },
    {
      "epoch": 0.06834285714285715,
      "grad_norm": 1.6204638481140137,
      "learning_rate": 0.0005,
      "loss": 3.8401665687561035,
      "step": 299
    },
    {
      "epoch": 0.06857142857142857,
      "grad_norm": 1.3319571018218994,
      "learning_rate": 0.0005,
      "loss": 3.567629814147949,
      "step": 300
    },
    {
      "epoch": 0.0688,
      "grad_norm": 1.2134617567062378,
      "learning_rate": 0.0005,
      "loss": 3.7008156776428223,
      "step": 301
    },
    {
      "epoch": 0.06902857142857143,
      "grad_norm": 1.4116822481155396,
      "learning_rate": 0.0005,
      "loss": 3.9831738471984863,
      "step": 302
    },
    {
      "epoch": 0.06925714285714285,
      "grad_norm": 1.249320387840271,
      "learning_rate": 0.0005,
      "loss": 3.8751940727233887,
      "step": 303
    },
    {
      "epoch": 0.06948571428571429,
      "grad_norm": 1.2759509086608887,
      "learning_rate": 0.0005,
      "loss": 3.4942989349365234,
      "step": 304
    },
    {
      "epoch": 0.06971428571428571,
      "grad_norm": 1.2790813446044922,
      "learning_rate": 0.0005,
      "loss": 3.8463170528411865,
      "step": 305
    },
    {
      "epoch": 0.06994285714285714,
      "grad_norm": 1.568544864654541,
      "learning_rate": 0.0005,
      "loss": 3.6108551025390625,
      "step": 306
    },
    {
      "epoch": 0.07017142857142857,
      "grad_norm": 1.9920867681503296,
      "learning_rate": 0.0005,
      "loss": 3.7098164558410645,
      "step": 307
    },
    {
      "epoch": 0.0704,
      "grad_norm": 1.2100567817687988,
      "learning_rate": 0.0005,
      "loss": 3.679121494293213,
      "step": 308
    },
    {
      "epoch": 0.07062857142857143,
      "grad_norm": 1.2073389291763306,
      "learning_rate": 0.0005,
      "loss": 3.546339273452759,
      "step": 309
    },
    {
      "epoch": 0.07085714285714285,
      "grad_norm": 1.168912410736084,
      "learning_rate": 0.0005,
      "loss": 3.526641368865967,
      "step": 310
    },
    {
      "epoch": 0.07108571428571428,
      "grad_norm": 1.2348332405090332,
      "learning_rate": 0.0005,
      "loss": 3.6594772338867188,
      "step": 311
    },
    {
      "epoch": 0.07131428571428572,
      "grad_norm": 1.521415114402771,
      "learning_rate": 0.0005,
      "loss": 3.5009522438049316,
      "step": 312
    },
    {
      "epoch": 0.07154285714285714,
      "grad_norm": 1.4644889831542969,
      "learning_rate": 0.0005,
      "loss": 3.5840132236480713,
      "step": 313
    },
    {
      "epoch": 0.07177142857142857,
      "grad_norm": 1.4732965230941772,
      "learning_rate": 0.0005,
      "loss": 3.4639716148376465,
      "step": 314
    },
    {
      "epoch": 0.072,
      "grad_norm": 1.2747281789779663,
      "learning_rate": 0.0005,
      "loss": 3.496725082397461,
      "step": 315
    },
    {
      "epoch": 0.07222857142857143,
      "grad_norm": 1.300007700920105,
      "learning_rate": 0.0005,
      "loss": 3.7136588096618652,
      "step": 316
    },
    {
      "epoch": 0.07245714285714286,
      "grad_norm": 1.1505074501037598,
      "learning_rate": 0.0005,
      "loss": 3.754096746444702,
      "step": 317
    },
    {
      "epoch": 0.07268571428571428,
      "grad_norm": 1.3167377710342407,
      "learning_rate": 0.0005,
      "loss": 3.2982516288757324,
      "step": 318
    },
    {
      "epoch": 0.07291428571428571,
      "grad_norm": 1.1764270067214966,
      "learning_rate": 0.0005,
      "loss": 3.7673044204711914,
      "step": 319
    },
    {
      "epoch": 0.07314285714285715,
      "grad_norm": 1.3944252729415894,
      "learning_rate": 0.0005,
      "loss": 3.672677993774414,
      "step": 320
    },
    {
      "epoch": 0.07337142857142857,
      "grad_norm": 1.4638551473617554,
      "learning_rate": 0.0005,
      "loss": 3.653892993927002,
      "step": 321
    },
    {
      "epoch": 0.0736,
      "grad_norm": 1.5427955389022827,
      "learning_rate": 0.0005,
      "loss": 3.8954038619995117,
      "step": 322
    },
    {
      "epoch": 0.07382857142857142,
      "grad_norm": 3.2110636234283447,
      "learning_rate": 0.0005,
      "loss": 3.318498373031616,
      "step": 323
    },
    {
      "epoch": 0.07405714285714286,
      "grad_norm": 1.509854793548584,
      "learning_rate": 0.0005,
      "loss": 3.750149965286255,
      "step": 324
    },
    {
      "epoch": 0.07428571428571429,
      "grad_norm": 1.6242187023162842,
      "learning_rate": 0.0005,
      "loss": 3.7089345455169678,
      "step": 325
    },
    {
      "epoch": 0.07451428571428571,
      "grad_norm": 1.5572633743286133,
      "learning_rate": 0.0005,
      "loss": 3.527831554412842,
      "step": 326
    },
    {
      "epoch": 0.07474285714285714,
      "grad_norm": 1.3194293975830078,
      "learning_rate": 0.0005,
      "loss": 3.792769193649292,
      "step": 327
    },
    {
      "epoch": 0.07497142857142858,
      "grad_norm": 1.548492670059204,
      "learning_rate": 0.0005,
      "loss": 3.6935267448425293,
      "step": 328
    },
    {
      "epoch": 0.0752,
      "grad_norm": 1.4776279926300049,
      "learning_rate": 0.0005,
      "loss": 4.060778617858887,
      "step": 329
    },
    {
      "epoch": 0.07542857142857143,
      "grad_norm": 1.1801342964172363,
      "learning_rate": 0.0005,
      "loss": 3.5094523429870605,
      "step": 330
    },
    {
      "epoch": 0.07565714285714285,
      "grad_norm": 1.2085458040237427,
      "learning_rate": 0.0005,
      "loss": 3.759072780609131,
      "step": 331
    },
    {
      "epoch": 0.07588571428571429,
      "grad_norm": 1.3487896919250488,
      "learning_rate": 0.0005,
      "loss": 3.272772789001465,
      "step": 332
    },
    {
      "epoch": 0.07611428571428572,
      "grad_norm": 1.3175698518753052,
      "learning_rate": 0.0005,
      "loss": 3.5341169834136963,
      "step": 333
    },
    {
      "epoch": 0.07634285714285714,
      "grad_norm": 1.4361851215362549,
      "learning_rate": 0.0005,
      "loss": 3.4780938625335693,
      "step": 334
    },
    {
      "epoch": 0.07657142857142857,
      "grad_norm": 1.4139351844787598,
      "learning_rate": 0.0005,
      "loss": 3.6460390090942383,
      "step": 335
    },
    {
      "epoch": 0.0768,
      "grad_norm": 1.215040922164917,
      "learning_rate": 0.0005,
      "loss": 3.6368422508239746,
      "step": 336
    },
    {
      "epoch": 0.07702857142857143,
      "grad_norm": 1.6508798599243164,
      "learning_rate": 0.0005,
      "loss": 3.843050241470337,
      "step": 337
    },
    {
      "epoch": 0.07725714285714286,
      "grad_norm": 2.0438170433044434,
      "learning_rate": 0.0005,
      "loss": 3.6644577980041504,
      "step": 338
    },
    {
      "epoch": 0.07748571428571428,
      "grad_norm": 1.201583743095398,
      "learning_rate": 0.0005,
      "loss": 3.659614324569702,
      "step": 339
    },
    {
      "epoch": 0.07771428571428571,
      "grad_norm": 1.371103048324585,
      "learning_rate": 0.0005,
      "loss": 3.7200236320495605,
      "step": 340
    },
    {
      "epoch": 0.07794285714285715,
      "grad_norm": 1.1748371124267578,
      "learning_rate": 0.0005,
      "loss": 3.36399507522583,
      "step": 341
    },
    {
      "epoch": 0.07817142857142857,
      "grad_norm": 1.6423648595809937,
      "learning_rate": 0.0005,
      "loss": 3.4408302307128906,
      "step": 342
    },
    {
      "epoch": 0.0784,
      "grad_norm": 1.2023521661758423,
      "learning_rate": 0.0005,
      "loss": 3.7996175289154053,
      "step": 343
    },
    {
      "epoch": 0.07862857142857142,
      "grad_norm": 1.358262062072754,
      "learning_rate": 0.0005,
      "loss": 3.5375587940216064,
      "step": 344
    },
    {
      "epoch": 0.07885714285714286,
      "grad_norm": 1.2512046098709106,
      "learning_rate": 0.0005,
      "loss": 3.3768086433410645,
      "step": 345
    },
    {
      "epoch": 0.07908571428571429,
      "grad_norm": 1.3445848226547241,
      "learning_rate": 0.0005,
      "loss": 3.8082776069641113,
      "step": 346
    },
    {
      "epoch": 0.07931428571428571,
      "grad_norm": 2.059642791748047,
      "learning_rate": 0.0005,
      "loss": 3.5630273818969727,
      "step": 347
    },
    {
      "epoch": 0.07954285714285714,
      "grad_norm": 1.5330147743225098,
      "learning_rate": 0.0005,
      "loss": 3.6555371284484863,
      "step": 348
    },
    {
      "epoch": 0.07977142857142858,
      "grad_norm": 1.4341421127319336,
      "learning_rate": 0.0005,
      "loss": 3.7400732040405273,
      "step": 349
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.3215850591659546,
      "learning_rate": 0.0005,
      "loss": 3.435283899307251,
      "step": 350
    },
    {
      "epoch": 0.08022857142857143,
      "grad_norm": 1.8934437036514282,
      "learning_rate": 0.0005,
      "loss": 3.5575506687164307,
      "step": 351
    },
    {
      "epoch": 0.08045714285714285,
      "grad_norm": 1.282736897468567,
      "learning_rate": 0.0005,
      "loss": 3.8786985874176025,
      "step": 352
    },
    {
      "epoch": 0.08068571428571429,
      "grad_norm": 1.286582589149475,
      "learning_rate": 0.0005,
      "loss": 3.765965223312378,
      "step": 353
    },
    {
      "epoch": 0.08091428571428572,
      "grad_norm": 1.341104507446289,
      "learning_rate": 0.0005,
      "loss": 3.304577589035034,
      "step": 354
    },
    {
      "epoch": 0.08114285714285714,
      "grad_norm": 1.401099681854248,
      "learning_rate": 0.0005,
      "loss": 3.6086666584014893,
      "step": 355
    },
    {
      "epoch": 0.08137142857142857,
      "grad_norm": 1.3088657855987549,
      "learning_rate": 0.0005,
      "loss": 3.7311766147613525,
      "step": 356
    },
    {
      "epoch": 0.0816,
      "grad_norm": 1.1418997049331665,
      "learning_rate": 0.0005,
      "loss": 3.6190600395202637,
      "step": 357
    },
    {
      "epoch": 0.08182857142857143,
      "grad_norm": 1.392096757888794,
      "learning_rate": 0.0005,
      "loss": 3.5025148391723633,
      "step": 358
    },
    {
      "epoch": 0.08205714285714286,
      "grad_norm": 1.4622337818145752,
      "learning_rate": 0.0005,
      "loss": 3.811997175216675,
      "step": 359
    },
    {
      "epoch": 0.08228571428571428,
      "grad_norm": 1.587007999420166,
      "learning_rate": 0.0005,
      "loss": 3.5051519870758057,
      "step": 360
    },
    {
      "epoch": 0.08251428571428572,
      "grad_norm": 1.3204596042633057,
      "learning_rate": 0.0005,
      "loss": 3.6969099044799805,
      "step": 361
    },
    {
      "epoch": 0.08274285714285715,
      "grad_norm": 1.4071555137634277,
      "learning_rate": 0.0005,
      "loss": 3.6496968269348145,
      "step": 362
    },
    {
      "epoch": 0.08297142857142857,
      "grad_norm": 1.4153343439102173,
      "learning_rate": 0.0005,
      "loss": 3.4801621437072754,
      "step": 363
    },
    {
      "epoch": 0.0832,
      "grad_norm": 1.3331066370010376,
      "learning_rate": 0.0005,
      "loss": 3.7088382244110107,
      "step": 364
    },
    {
      "epoch": 0.08342857142857144,
      "grad_norm": 1.5735594034194946,
      "learning_rate": 0.0005,
      "loss": 3.5963356494903564,
      "step": 365
    },
    {
      "epoch": 0.08365714285714286,
      "grad_norm": 1.3685883283615112,
      "learning_rate": 0.0005,
      "loss": 3.402953624725342,
      "step": 366
    },
    {
      "epoch": 0.08388571428571429,
      "grad_norm": 1.2061632871627808,
      "learning_rate": 0.0005,
      "loss": 3.603895902633667,
      "step": 367
    },
    {
      "epoch": 0.08411428571428571,
      "grad_norm": 1.6633912324905396,
      "learning_rate": 0.0005,
      "loss": 3.8311686515808105,
      "step": 368
    },
    {
      "epoch": 0.08434285714285714,
      "grad_norm": 1.5278801918029785,
      "learning_rate": 0.0005,
      "loss": 4.043961524963379,
      "step": 369
    },
    {
      "epoch": 0.08457142857142858,
      "grad_norm": 1.4926207065582275,
      "learning_rate": 0.0005,
      "loss": 3.6231112480163574,
      "step": 370
    },
    {
      "epoch": 0.0848,
      "grad_norm": 1.3038722276687622,
      "learning_rate": 0.0005,
      "loss": 4.11037540435791,
      "step": 371
    },
    {
      "epoch": 0.08502857142857143,
      "grad_norm": 1.5589001178741455,
      "learning_rate": 0.0005,
      "loss": 3.5757462978363037,
      "step": 372
    },
    {
      "epoch": 0.08525714285714285,
      "grad_norm": 1.4059929847717285,
      "learning_rate": 0.0005,
      "loss": 3.675281047821045,
      "step": 373
    },
    {
      "epoch": 0.08548571428571429,
      "grad_norm": 1.1244523525238037,
      "learning_rate": 0.0005,
      "loss": 3.598385810852051,
      "step": 374
    },
    {
      "epoch": 0.08571428571428572,
      "grad_norm": 1.5446771383285522,
      "learning_rate": 0.0005,
      "loss": 3.5697579383850098,
      "step": 375
    },
    {
      "epoch": 0.08594285714285714,
      "grad_norm": 1.1067748069763184,
      "learning_rate": 0.0005,
      "loss": 3.7270100116729736,
      "step": 376
    },
    {
      "epoch": 0.08617142857142857,
      "grad_norm": 1.1611194610595703,
      "learning_rate": 0.0005,
      "loss": 3.286832809448242,
      "step": 377
    },
    {
      "epoch": 0.0864,
      "grad_norm": 1.350606083869934,
      "learning_rate": 0.0005,
      "loss": 3.6877050399780273,
      "step": 378
    },
    {
      "epoch": 0.08662857142857143,
      "grad_norm": 1.35396409034729,
      "learning_rate": 0.0005,
      "loss": 3.751121997833252,
      "step": 379
    },
    {
      "epoch": 0.08685714285714285,
      "grad_norm": 1.47518789768219,
      "learning_rate": 0.0005,
      "loss": 3.326115608215332,
      "step": 380
    },
    {
      "epoch": 0.08708571428571428,
      "grad_norm": 1.3288865089416504,
      "learning_rate": 0.0005,
      "loss": 3.515101909637451,
      "step": 381
    },
    {
      "epoch": 0.08731428571428572,
      "grad_norm": 1.2035319805145264,
      "learning_rate": 0.0005,
      "loss": 3.585296630859375,
      "step": 382
    },
    {
      "epoch": 0.08754285714285714,
      "grad_norm": 1.397428274154663,
      "learning_rate": 0.0005,
      "loss": 3.5142874717712402,
      "step": 383
    },
    {
      "epoch": 0.08777142857142857,
      "grad_norm": 1.2564878463745117,
      "learning_rate": 0.0005,
      "loss": 3.396972894668579,
      "step": 384
    },
    {
      "epoch": 0.088,
      "grad_norm": 1.3066142797470093,
      "learning_rate": 0.0005,
      "loss": 3.742716073989868,
      "step": 385
    },
    {
      "epoch": 0.08822857142857143,
      "grad_norm": 1.3552844524383545,
      "learning_rate": 0.0005,
      "loss": 3.1984140872955322,
      "step": 386
    },
    {
      "epoch": 0.08845714285714286,
      "grad_norm": 1.1990270614624023,
      "learning_rate": 0.0005,
      "loss": 3.4115896224975586,
      "step": 387
    },
    {
      "epoch": 0.08868571428571428,
      "grad_norm": 1.566868782043457,
      "learning_rate": 0.0005,
      "loss": 4.1977009773254395,
      "step": 388
    },
    {
      "epoch": 0.08891428571428571,
      "grad_norm": 1.1521899700164795,
      "learning_rate": 0.0005,
      "loss": 3.8816628456115723,
      "step": 389
    },
    {
      "epoch": 0.08914285714285715,
      "grad_norm": 1.339096188545227,
      "learning_rate": 0.0005,
      "loss": 3.8031508922576904,
      "step": 390
    },
    {
      "epoch": 0.08937142857142857,
      "grad_norm": 1.420404076576233,
      "learning_rate": 0.0005,
      "loss": 3.5656962394714355,
      "step": 391
    },
    {
      "epoch": 0.0896,
      "grad_norm": 1.2034275531768799,
      "learning_rate": 0.0005,
      "loss": 3.6832706928253174,
      "step": 392
    },
    {
      "epoch": 0.08982857142857142,
      "grad_norm": 1.146755576133728,
      "learning_rate": 0.0005,
      "loss": 3.7228288650512695,
      "step": 393
    },
    {
      "epoch": 0.09005714285714286,
      "grad_norm": 1.3703813552856445,
      "learning_rate": 0.0005,
      "loss": 3.6604435443878174,
      "step": 394
    },
    {
      "epoch": 0.09028571428571429,
      "grad_norm": 1.473766803741455,
      "learning_rate": 0.0005,
      "loss": 3.4228732585906982,
      "step": 395
    },
    {
      "epoch": 0.09051428571428571,
      "grad_norm": 1.4480396509170532,
      "learning_rate": 0.0005,
      "loss": 3.3679041862487793,
      "step": 396
    },
    {
      "epoch": 0.09074285714285714,
      "grad_norm": 1.3072739839553833,
      "learning_rate": 0.0005,
      "loss": 3.2406210899353027,
      "step": 397
    },
    {
      "epoch": 0.09097142857142858,
      "grad_norm": 1.1768566370010376,
      "learning_rate": 0.0005,
      "loss": 3.911329984664917,
      "step": 398
    },
    {
      "epoch": 0.0912,
      "grad_norm": 1.3759900331497192,
      "learning_rate": 0.0005,
      "loss": 3.472726345062256,
      "step": 399
    },
    {
      "epoch": 0.09142857142857143,
      "grad_norm": 1.3300721645355225,
      "learning_rate": 0.0005,
      "loss": 3.6093525886535645,
      "step": 400
    },
    {
      "epoch": 0.09165714285714285,
      "grad_norm": 1.3269250392913818,
      "learning_rate": 0.0005,
      "loss": 3.2099485397338867,
      "step": 401
    },
    {
      "epoch": 0.09188571428571428,
      "grad_norm": 1.2767302989959717,
      "learning_rate": 0.0005,
      "loss": 4.12113094329834,
      "step": 402
    },
    {
      "epoch": 0.09211428571428572,
      "grad_norm": 1.1523395776748657,
      "learning_rate": 0.0005,
      "loss": 3.8204140663146973,
      "step": 403
    },
    {
      "epoch": 0.09234285714285714,
      "grad_norm": 1.2688173055648804,
      "learning_rate": 0.0005,
      "loss": 3.75929856300354,
      "step": 404
    },
    {
      "epoch": 0.09257142857142857,
      "grad_norm": 1.470902681350708,
      "learning_rate": 0.0005,
      "loss": 3.668781042098999,
      "step": 405
    },
    {
      "epoch": 0.0928,
      "grad_norm": 1.2483205795288086,
      "learning_rate": 0.0005,
      "loss": 3.3948488235473633,
      "step": 406
    },
    {
      "epoch": 0.09302857142857143,
      "grad_norm": 1.323643684387207,
      "learning_rate": 0.0005,
      "loss": 3.5386648178100586,
      "step": 407
    },
    {
      "epoch": 0.09325714285714286,
      "grad_norm": 1.2335057258605957,
      "learning_rate": 0.0005,
      "loss": 3.7350637912750244,
      "step": 408
    },
    {
      "epoch": 0.09348571428571428,
      "grad_norm": 1.3203126192092896,
      "learning_rate": 0.0005,
      "loss": 3.3751420974731445,
      "step": 409
    },
    {
      "epoch": 0.09371428571428571,
      "grad_norm": 1.4195871353149414,
      "learning_rate": 0.0005,
      "loss": 3.8481462001800537,
      "step": 410
    },
    {
      "epoch": 0.09394285714285715,
      "grad_norm": 1.6018916368484497,
      "learning_rate": 0.0005,
      "loss": 3.40981388092041,
      "step": 411
    },
    {
      "epoch": 0.09417142857142857,
      "grad_norm": 1.3591737747192383,
      "learning_rate": 0.0005,
      "loss": 3.7409563064575195,
      "step": 412
    },
    {
      "epoch": 0.0944,
      "grad_norm": 1.3438103199005127,
      "learning_rate": 0.0005,
      "loss": 3.8454294204711914,
      "step": 413
    },
    {
      "epoch": 0.09462857142857142,
      "grad_norm": 1.3571646213531494,
      "learning_rate": 0.0005,
      "loss": 3.5231900215148926,
      "step": 414
    },
    {
      "epoch": 0.09485714285714286,
      "grad_norm": 1.2614967823028564,
      "learning_rate": 0.0005,
      "loss": 3.9322304725646973,
      "step": 415
    },
    {
      "epoch": 0.09508571428571429,
      "grad_norm": 1.403802514076233,
      "learning_rate": 0.0005,
      "loss": 3.4331212043762207,
      "step": 416
    },
    {
      "epoch": 0.09531428571428571,
      "grad_norm": 1.3498650789260864,
      "learning_rate": 0.0005,
      "loss": 3.6776788234710693,
      "step": 417
    },
    {
      "epoch": 0.09554285714285714,
      "grad_norm": 1.8535287380218506,
      "learning_rate": 0.0005,
      "loss": 3.429011344909668,
      "step": 418
    },
    {
      "epoch": 0.09577142857142858,
      "grad_norm": 2.349604845046997,
      "learning_rate": 0.0005,
      "loss": 3.904383659362793,
      "step": 419
    },
    {
      "epoch": 0.096,
      "grad_norm": 1.3664841651916504,
      "learning_rate": 0.0005,
      "loss": 3.832124710083008,
      "step": 420
    },
    {
      "epoch": 0.09622857142857143,
      "grad_norm": 1.544201135635376,
      "learning_rate": 0.0005,
      "loss": 3.449954032897949,
      "step": 421
    },
    {
      "epoch": 0.09645714285714285,
      "grad_norm": 1.4122731685638428,
      "learning_rate": 0.0005,
      "loss": 3.341754674911499,
      "step": 422
    },
    {
      "epoch": 0.09668571428571429,
      "grad_norm": 1.2019599676132202,
      "learning_rate": 0.0005,
      "loss": 3.266845703125,
      "step": 423
    },
    {
      "epoch": 0.09691428571428572,
      "grad_norm": 1.6952075958251953,
      "learning_rate": 0.0005,
      "loss": 3.6798112392425537,
      "step": 424
    },
    {
      "epoch": 0.09714285714285714,
      "grad_norm": 1.2490808963775635,
      "learning_rate": 0.0005,
      "loss": 3.8810343742370605,
      "step": 425
    },
    {
      "epoch": 0.09737142857142857,
      "grad_norm": 1.350743055343628,
      "learning_rate": 0.0005,
      "loss": 3.5673015117645264,
      "step": 426
    },
    {
      "epoch": 0.0976,
      "grad_norm": 1.677984595298767,
      "learning_rate": 0.0005,
      "loss": 3.458040714263916,
      "step": 427
    },
    {
      "epoch": 0.09782857142857143,
      "grad_norm": 1.2624021768569946,
      "learning_rate": 0.0005,
      "loss": 3.5245614051818848,
      "step": 428
    },
    {
      "epoch": 0.09805714285714286,
      "grad_norm": 1.2233283519744873,
      "learning_rate": 0.0005,
      "loss": 3.6560068130493164,
      "step": 429
    },
    {
      "epoch": 0.09828571428571428,
      "grad_norm": 1.2917431592941284,
      "learning_rate": 0.0005,
      "loss": 3.603595733642578,
      "step": 430
    },
    {
      "epoch": 0.09851428571428572,
      "grad_norm": 1.2832313776016235,
      "learning_rate": 0.0005,
      "loss": 3.6006157398223877,
      "step": 431
    },
    {
      "epoch": 0.09874285714285715,
      "grad_norm": 1.3349357843399048,
      "learning_rate": 0.0005,
      "loss": 3.6945619583129883,
      "step": 432
    },
    {
      "epoch": 0.09897142857142857,
      "grad_norm": 1.2243572473526,
      "learning_rate": 0.0005,
      "loss": 3.5599851608276367,
      "step": 433
    },
    {
      "epoch": 0.0992,
      "grad_norm": 1.3532084226608276,
      "learning_rate": 0.0005,
      "loss": 3.489680290222168,
      "step": 434
    },
    {
      "epoch": 0.09942857142857142,
      "grad_norm": 1.2197086811065674,
      "learning_rate": 0.0005,
      "loss": 3.2790753841400146,
      "step": 435
    },
    {
      "epoch": 0.09965714285714286,
      "grad_norm": 1.3335022926330566,
      "learning_rate": 0.0005,
      "loss": 3.558333158493042,
      "step": 436
    },
    {
      "epoch": 0.09988571428571429,
      "grad_norm": 1.6606781482696533,
      "learning_rate": 0.0005,
      "loss": 3.6792807579040527,
      "step": 437
    },
    {
      "epoch": 0.10011428571428571,
      "grad_norm": 1.0976309776306152,
      "learning_rate": 0.0005,
      "loss": 3.482938766479492,
      "step": 438
    },
    {
      "epoch": 0.10034285714285714,
      "grad_norm": 1.4876348972320557,
      "learning_rate": 0.0005,
      "loss": 3.641645908355713,
      "step": 439
    },
    {
      "epoch": 0.10057142857142858,
      "grad_norm": 1.22615385055542,
      "learning_rate": 0.0005,
      "loss": 3.2594499588012695,
      "step": 440
    },
    {
      "epoch": 0.1008,
      "grad_norm": 1.7769153118133545,
      "learning_rate": 0.0005,
      "loss": 3.5491228103637695,
      "step": 441
    },
    {
      "epoch": 0.10102857142857143,
      "grad_norm": 1.3555201292037964,
      "learning_rate": 0.0005,
      "loss": 3.1376261711120605,
      "step": 442
    },
    {
      "epoch": 0.10125714285714285,
      "grad_norm": 1.4866825342178345,
      "learning_rate": 0.0005,
      "loss": 3.8171348571777344,
      "step": 443
    },
    {
      "epoch": 0.10148571428571429,
      "grad_norm": 1.2722042798995972,
      "learning_rate": 0.0005,
      "loss": 3.6559505462646484,
      "step": 444
    },
    {
      "epoch": 0.10171428571428572,
      "grad_norm": 1.2002360820770264,
      "learning_rate": 0.0005,
      "loss": 3.5758872032165527,
      "step": 445
    },
    {
      "epoch": 0.10194285714285714,
      "grad_norm": 1.1575227975845337,
      "learning_rate": 0.0005,
      "loss": 3.688617706298828,
      "step": 446
    },
    {
      "epoch": 0.10217142857142857,
      "grad_norm": 1.3998205661773682,
      "learning_rate": 0.0005,
      "loss": 3.5195534229278564,
      "step": 447
    },
    {
      "epoch": 0.1024,
      "grad_norm": 1.3692439794540405,
      "learning_rate": 0.0005,
      "loss": 3.6461358070373535,
      "step": 448
    },
    {
      "epoch": 0.10262857142857143,
      "grad_norm": 1.3523601293563843,
      "learning_rate": 0.0005,
      "loss": 3.6538729667663574,
      "step": 449
    },
    {
      "epoch": 0.10285714285714286,
      "grad_norm": 1.5553241968154907,
      "learning_rate": 0.0005,
      "loss": 3.492299795150757,
      "step": 450
    },
    {
      "epoch": 0.10308571428571428,
      "grad_norm": 1.1943773031234741,
      "learning_rate": 0.0005,
      "loss": 3.272653579711914,
      "step": 451
    },
    {
      "epoch": 0.10331428571428572,
      "grad_norm": 1.1808604001998901,
      "learning_rate": 0.0005,
      "loss": 3.5652015209198,
      "step": 452
    },
    {
      "epoch": 0.10354285714285714,
      "grad_norm": 1.3018568754196167,
      "learning_rate": 0.0005,
      "loss": 3.6031107902526855,
      "step": 453
    },
    {
      "epoch": 0.10377142857142857,
      "grad_norm": 1.5600961446762085,
      "learning_rate": 0.0005,
      "loss": 3.303699016571045,
      "step": 454
    },
    {
      "epoch": 0.104,
      "grad_norm": 1.4959473609924316,
      "learning_rate": 0.0005,
      "loss": 3.8576085567474365,
      "step": 455
    },
    {
      "epoch": 0.10422857142857143,
      "grad_norm": 1.1918203830718994,
      "learning_rate": 0.0005,
      "loss": 3.940556526184082,
      "step": 456
    },
    {
      "epoch": 0.10445714285714286,
      "grad_norm": 1.4864777326583862,
      "learning_rate": 0.0005,
      "loss": 3.239314079284668,
      "step": 457
    },
    {
      "epoch": 0.10468571428571428,
      "grad_norm": 1.6245197057724,
      "learning_rate": 0.0005,
      "loss": 3.5924549102783203,
      "step": 458
    },
    {
      "epoch": 0.10491428571428571,
      "grad_norm": 1.3549028635025024,
      "learning_rate": 0.0005,
      "loss": 3.6259727478027344,
      "step": 459
    },
    {
      "epoch": 0.10514285714285715,
      "grad_norm": 1.4500203132629395,
      "learning_rate": 0.0005,
      "loss": 3.9808871746063232,
      "step": 460
    },
    {
      "epoch": 0.10537142857142857,
      "grad_norm": 1.2358193397521973,
      "learning_rate": 0.0005,
      "loss": 3.3773839473724365,
      "step": 461
    },
    {
      "epoch": 0.1056,
      "grad_norm": 2.2598726749420166,
      "learning_rate": 0.0005,
      "loss": 3.7704339027404785,
      "step": 462
    },
    {
      "epoch": 0.10582857142857142,
      "grad_norm": 1.264886498451233,
      "learning_rate": 0.0005,
      "loss": 3.6965627670288086,
      "step": 463
    },
    {
      "epoch": 0.10605714285714286,
      "grad_norm": 1.3942813873291016,
      "learning_rate": 0.0005,
      "loss": 3.3916878700256348,
      "step": 464
    },
    {
      "epoch": 0.10628571428571429,
      "grad_norm": 1.5215466022491455,
      "learning_rate": 0.0005,
      "loss": 3.7154946327209473,
      "step": 465
    },
    {
      "epoch": 0.10651428571428571,
      "grad_norm": 1.417777419090271,
      "learning_rate": 0.0005,
      "loss": 3.479729652404785,
      "step": 466
    },
    {
      "epoch": 0.10674285714285714,
      "grad_norm": 1.210738182067871,
      "learning_rate": 0.0005,
      "loss": 3.264779567718506,
      "step": 467
    },
    {
      "epoch": 0.10697142857142856,
      "grad_norm": 1.6876869201660156,
      "learning_rate": 0.0005,
      "loss": 3.88247013092041,
      "step": 468
    },
    {
      "epoch": 0.1072,
      "grad_norm": 1.4067280292510986,
      "learning_rate": 0.0005,
      "loss": 3.6955037117004395,
      "step": 469
    },
    {
      "epoch": 0.10742857142857143,
      "grad_norm": 1.3512946367263794,
      "learning_rate": 0.0005,
      "loss": 3.593568801879883,
      "step": 470
    },
    {
      "epoch": 0.10765714285714285,
      "grad_norm": 1.124022126197815,
      "learning_rate": 0.0005,
      "loss": 3.8063504695892334,
      "step": 471
    },
    {
      "epoch": 0.10788571428571428,
      "grad_norm": 1.7028707265853882,
      "learning_rate": 0.0005,
      "loss": 3.5206494331359863,
      "step": 472
    },
    {
      "epoch": 0.10811428571428572,
      "grad_norm": 1.6694047451019287,
      "learning_rate": 0.0005,
      "loss": 3.5685436725616455,
      "step": 473
    },
    {
      "epoch": 0.10834285714285714,
      "grad_norm": 1.4319220781326294,
      "learning_rate": 0.0005,
      "loss": 3.5232505798339844,
      "step": 474
    },
    {
      "epoch": 0.10857142857142857,
      "grad_norm": 1.538309931755066,
      "learning_rate": 0.0005,
      "loss": 3.9773221015930176,
      "step": 475
    },
    {
      "epoch": 0.1088,
      "grad_norm": 1.2790374755859375,
      "learning_rate": 0.0005,
      "loss": 3.4565889835357666,
      "step": 476
    },
    {
      "epoch": 0.10902857142857143,
      "grad_norm": 2.300703287124634,
      "learning_rate": 0.0005,
      "loss": 3.239528179168701,
      "step": 477
    },
    {
      "epoch": 0.10925714285714286,
      "grad_norm": 1.407217264175415,
      "learning_rate": 0.0005,
      "loss": 3.258190155029297,
      "step": 478
    },
    {
      "epoch": 0.10948571428571428,
      "grad_norm": 1.3425604104995728,
      "learning_rate": 0.0005,
      "loss": 3.4657769203186035,
      "step": 479
    },
    {
      "epoch": 0.10971428571428571,
      "grad_norm": 1.2434214353561401,
      "learning_rate": 0.0005,
      "loss": 3.6943063735961914,
      "step": 480
    },
    {
      "epoch": 0.10994285714285715,
      "grad_norm": 1.5595176219940186,
      "learning_rate": 0.0005,
      "loss": 3.971341609954834,
      "step": 481
    },
    {
      "epoch": 0.11017142857142857,
      "grad_norm": 1.1903125047683716,
      "learning_rate": 0.0005,
      "loss": 3.577483654022217,
      "step": 482
    },
    {
      "epoch": 0.1104,
      "grad_norm": 1.4645559787750244,
      "learning_rate": 0.0005,
      "loss": 3.903996229171753,
      "step": 483
    },
    {
      "epoch": 0.11062857142857142,
      "grad_norm": 1.4681087732315063,
      "learning_rate": 0.0005,
      "loss": 3.684515953063965,
      "step": 484
    },
    {
      "epoch": 0.11085714285714286,
      "grad_norm": 1.3509248495101929,
      "learning_rate": 0.0005,
      "loss": 3.715175151824951,
      "step": 485
    },
    {
      "epoch": 0.11108571428571429,
      "grad_norm": 1.2012195587158203,
      "learning_rate": 0.0005,
      "loss": 3.5327141284942627,
      "step": 486
    },
    {
      "epoch": 0.11131428571428571,
      "grad_norm": 1.2611082792282104,
      "learning_rate": 0.0005,
      "loss": 3.767380952835083,
      "step": 487
    },
    {
      "epoch": 0.11154285714285714,
      "grad_norm": 1.4301815032958984,
      "learning_rate": 0.0005,
      "loss": 3.400557518005371,
      "step": 488
    },
    {
      "epoch": 0.11177142857142858,
      "grad_norm": 1.4819960594177246,
      "learning_rate": 0.0005,
      "loss": 3.5852272510528564,
      "step": 489
    },
    {
      "epoch": 0.112,
      "grad_norm": 1.1634678840637207,
      "learning_rate": 0.0005,
      "loss": 3.5717461109161377,
      "step": 490
    },
    {
      "epoch": 0.11222857142857143,
      "grad_norm": 1.3186256885528564,
      "learning_rate": 0.0005,
      "loss": 3.6545400619506836,
      "step": 491
    },
    {
      "epoch": 0.11245714285714285,
      "grad_norm": 1.1591006517410278,
      "learning_rate": 0.0005,
      "loss": 3.551281452178955,
      "step": 492
    },
    {
      "epoch": 0.11268571428571429,
      "grad_norm": 1.2611786127090454,
      "learning_rate": 0.0005,
      "loss": 3.4868316650390625,
      "step": 493
    },
    {
      "epoch": 0.11291428571428572,
      "grad_norm": 1.3877400159835815,
      "learning_rate": 0.0005,
      "loss": 3.4854555130004883,
      "step": 494
    },
    {
      "epoch": 0.11314285714285714,
      "grad_norm": 1.8179017305374146,
      "learning_rate": 0.0005,
      "loss": 3.3458058834075928,
      "step": 495
    },
    {
      "epoch": 0.11337142857142857,
      "grad_norm": 1.0434560775756836,
      "learning_rate": 0.0005,
      "loss": 3.7137246131896973,
      "step": 496
    },
    {
      "epoch": 0.1136,
      "grad_norm": 1.5516504049301147,
      "learning_rate": 0.0005,
      "loss": 3.428447723388672,
      "step": 497
    },
    {
      "epoch": 0.11382857142857143,
      "grad_norm": 2.100184440612793,
      "learning_rate": 0.0005,
      "loss": 3.590832233428955,
      "step": 498
    },
    {
      "epoch": 0.11405714285714286,
      "grad_norm": 1.6004587411880493,
      "learning_rate": 0.0005,
      "loss": 3.876399040222168,
      "step": 499
    },
    {
      "epoch": 0.11428571428571428,
      "grad_norm": 1.2955355644226074,
      "learning_rate": 0.0005,
      "loss": 3.5771031379699707,
      "step": 500
    },
    {
      "epoch": 0.11451428571428571,
      "grad_norm": 1.2196674346923828,
      "learning_rate": 0.0005,
      "loss": 3.6854755878448486,
      "step": 501
    },
    {
      "epoch": 0.11474285714285715,
      "grad_norm": 1.3593248128890991,
      "learning_rate": 0.0005,
      "loss": 3.5146193504333496,
      "step": 502
    },
    {
      "epoch": 0.11497142857142857,
      "grad_norm": 1.954837441444397,
      "learning_rate": 0.0005,
      "loss": 3.1663906574249268,
      "step": 503
    },
    {
      "epoch": 0.1152,
      "grad_norm": 1.4953300952911377,
      "learning_rate": 0.0005,
      "loss": 3.8076229095458984,
      "step": 504
    },
    {
      "epoch": 0.11542857142857142,
      "grad_norm": 1.7837493419647217,
      "learning_rate": 0.0005,
      "loss": 3.4374921321868896,
      "step": 505
    },
    {
      "epoch": 0.11565714285714286,
      "grad_norm": 1.1906944513320923,
      "learning_rate": 0.0005,
      "loss": 3.1963043212890625,
      "step": 506
    },
    {
      "epoch": 0.11588571428571429,
      "grad_norm": 1.4154081344604492,
      "learning_rate": 0.0005,
      "loss": 3.6970205307006836,
      "step": 507
    },
    {
      "epoch": 0.11611428571428571,
      "grad_norm": 1.1819933652877808,
      "learning_rate": 0.0005,
      "loss": 3.496640682220459,
      "step": 508
    },
    {
      "epoch": 0.11634285714285714,
      "grad_norm": 1.2467728853225708,
      "learning_rate": 0.0005,
      "loss": 3.560436725616455,
      "step": 509
    },
    {
      "epoch": 0.11657142857142858,
      "grad_norm": 1.3246896266937256,
      "learning_rate": 0.0005,
      "loss": 3.5426385402679443,
      "step": 510
    },
    {
      "epoch": 0.1168,
      "grad_norm": 1.9025654792785645,
      "learning_rate": 0.0005,
      "loss": 3.528945207595825,
      "step": 511
    },
    {
      "epoch": 0.11702857142857143,
      "grad_norm": 1.6316934823989868,
      "learning_rate": 0.0005,
      "loss": 3.249025821685791,
      "step": 512
    },
    {
      "epoch": 0.11725714285714285,
      "grad_norm": 1.4845094680786133,
      "learning_rate": 0.0005,
      "loss": 3.878725528717041,
      "step": 513
    },
    {
      "epoch": 0.11748571428571429,
      "grad_norm": 1.4473185539245605,
      "learning_rate": 0.0005,
      "loss": 3.698549747467041,
      "step": 514
    },
    {
      "epoch": 0.11771428571428572,
      "grad_norm": 1.451786994934082,
      "learning_rate": 0.0005,
      "loss": 3.5434131622314453,
      "step": 515
    },
    {
      "epoch": 0.11794285714285714,
      "grad_norm": 1.9033275842666626,
      "learning_rate": 0.0005,
      "loss": 3.6143851280212402,
      "step": 516
    },
    {
      "epoch": 0.11817142857142857,
      "grad_norm": 1.3777387142181396,
      "learning_rate": 0.0005,
      "loss": 3.286029577255249,
      "step": 517
    },
    {
      "epoch": 0.1184,
      "grad_norm": 1.1582823991775513,
      "learning_rate": 0.0005,
      "loss": 3.314500331878662,
      "step": 518
    },
    {
      "epoch": 0.11862857142857143,
      "grad_norm": 1.509984016418457,
      "learning_rate": 0.0005,
      "loss": 3.3856379985809326,
      "step": 519
    },
    {
      "epoch": 0.11885714285714286,
      "grad_norm": 1.3719030618667603,
      "learning_rate": 0.0005,
      "loss": 4.189189434051514,
      "step": 520
    },
    {
      "epoch": 0.11908571428571428,
      "grad_norm": 1.466270089149475,
      "learning_rate": 0.0005,
      "loss": 3.101868152618408,
      "step": 521
    },
    {
      "epoch": 0.11931428571428572,
      "grad_norm": 1.3433579206466675,
      "learning_rate": 0.0005,
      "loss": 3.5034871101379395,
      "step": 522
    },
    {
      "epoch": 0.11954285714285715,
      "grad_norm": 1.273208498954773,
      "learning_rate": 0.0005,
      "loss": 3.9476265907287598,
      "step": 523
    },
    {
      "epoch": 0.11977142857142857,
      "grad_norm": 1.4041937589645386,
      "learning_rate": 0.0005,
      "loss": 3.6539759635925293,
      "step": 524
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.2310388088226318,
      "learning_rate": 0.0005,
      "loss": 3.363105297088623,
      "step": 525
    },
    {
      "epoch": 0.12022857142857143,
      "grad_norm": 2.119318723678589,
      "learning_rate": 0.0005,
      "loss": 3.3692965507507324,
      "step": 526
    },
    {
      "epoch": 0.12045714285714286,
      "grad_norm": 1.7262386083602905,
      "learning_rate": 0.0005,
      "loss": 3.7536230087280273,
      "step": 527
    },
    {
      "epoch": 0.12068571428571429,
      "grad_norm": 1.4317426681518555,
      "learning_rate": 0.0005,
      "loss": 3.716709613800049,
      "step": 528
    },
    {
      "epoch": 0.12091428571428571,
      "grad_norm": 1.4034680128097534,
      "learning_rate": 0.0005,
      "loss": 3.5209879875183105,
      "step": 529
    },
    {
      "epoch": 0.12114285714285715,
      "grad_norm": 1.6710659265518188,
      "learning_rate": 0.0005,
      "loss": 3.6880576610565186,
      "step": 530
    },
    {
      "epoch": 0.12137142857142857,
      "grad_norm": 1.4356098175048828,
      "learning_rate": 0.0005,
      "loss": 3.4448978900909424,
      "step": 531
    },
    {
      "epoch": 0.1216,
      "grad_norm": 1.8492984771728516,
      "learning_rate": 0.0005,
      "loss": 3.7822608947753906,
      "step": 532
    },
    {
      "epoch": 0.12182857142857142,
      "grad_norm": 1.439728021621704,
      "learning_rate": 0.0005,
      "loss": 3.602376937866211,
      "step": 533
    },
    {
      "epoch": 0.12205714285714286,
      "grad_norm": 1.5829676389694214,
      "learning_rate": 0.0005,
      "loss": 3.567934274673462,
      "step": 534
    },
    {
      "epoch": 0.12228571428571429,
      "grad_norm": 1.422768235206604,
      "learning_rate": 0.0005,
      "loss": 3.676254987716675,
      "step": 535
    },
    {
      "epoch": 0.12251428571428571,
      "grad_norm": 1.095672369003296,
      "learning_rate": 0.0005,
      "loss": 3.473499298095703,
      "step": 536
    },
    {
      "epoch": 0.12274285714285714,
      "grad_norm": 1.338233470916748,
      "learning_rate": 0.0005,
      "loss": 3.877812623977661,
      "step": 537
    },
    {
      "epoch": 0.12297142857142856,
      "grad_norm": 1.6799508333206177,
      "learning_rate": 0.0005,
      "loss": 3.560312271118164,
      "step": 538
    },
    {
      "epoch": 0.1232,
      "grad_norm": 1.4923204183578491,
      "learning_rate": 0.0005,
      "loss": 3.2903833389282227,
      "step": 539
    },
    {
      "epoch": 0.12342857142857143,
      "grad_norm": 1.3227003812789917,
      "learning_rate": 0.0005,
      "loss": 3.6803126335144043,
      "step": 540
    },
    {
      "epoch": 0.12365714285714285,
      "grad_norm": 1.2114008665084839,
      "learning_rate": 0.0005,
      "loss": 3.733154535293579,
      "step": 541
    },
    {
      "epoch": 0.12388571428571428,
      "grad_norm": 1.4413926601409912,
      "learning_rate": 0.0005,
      "loss": 3.622185230255127,
      "step": 542
    },
    {
      "epoch": 0.12411428571428572,
      "grad_norm": 1.2624928951263428,
      "learning_rate": 0.0005,
      "loss": 3.4513912200927734,
      "step": 543
    },
    {
      "epoch": 0.12434285714285714,
      "grad_norm": 1.1511212587356567,
      "learning_rate": 0.0005,
      "loss": 3.819765567779541,
      "step": 544
    },
    {
      "epoch": 0.12457142857142857,
      "grad_norm": 1.4617167711257935,
      "learning_rate": 0.0005,
      "loss": 3.3952951431274414,
      "step": 545
    },
    {
      "epoch": 0.1248,
      "grad_norm": 1.2303882837295532,
      "learning_rate": 0.0005,
      "loss": 3.5665202140808105,
      "step": 546
    },
    {
      "epoch": 0.12502857142857143,
      "grad_norm": 1.3206497430801392,
      "learning_rate": 0.0005,
      "loss": 3.508967399597168,
      "step": 547
    },
    {
      "epoch": 0.12525714285714284,
      "grad_norm": 1.1913822889328003,
      "learning_rate": 0.0005,
      "loss": 3.4345500469207764,
      "step": 548
    },
    {
      "epoch": 0.12548571428571428,
      "grad_norm": 1.365575909614563,
      "learning_rate": 0.0005,
      "loss": 3.6840996742248535,
      "step": 549
    },
    {
      "epoch": 0.12571428571428572,
      "grad_norm": 1.2756586074829102,
      "learning_rate": 0.0005,
      "loss": 3.589536190032959,
      "step": 550
    },
    {
      "epoch": 0.12594285714285713,
      "grad_norm": 1.0290449857711792,
      "learning_rate": 0.0005,
      "loss": 3.7677249908447266,
      "step": 551
    },
    {
      "epoch": 0.12617142857142857,
      "grad_norm": 1.7632781267166138,
      "learning_rate": 0.0005,
      "loss": 3.4048032760620117,
      "step": 552
    },
    {
      "epoch": 0.1264,
      "grad_norm": 1.1841031312942505,
      "learning_rate": 0.0005,
      "loss": 3.3941850662231445,
      "step": 553
    },
    {
      "epoch": 0.12662857142857142,
      "grad_norm": 1.2468888759613037,
      "learning_rate": 0.0005,
      "loss": 3.7258052825927734,
      "step": 554
    },
    {
      "epoch": 0.12685714285714286,
      "grad_norm": 1.232893466949463,
      "learning_rate": 0.0005,
      "loss": 3.605984687805176,
      "step": 555
    },
    {
      "epoch": 0.12708571428571427,
      "grad_norm": 1.1070239543914795,
      "learning_rate": 0.0005,
      "loss": 4.07947301864624,
      "step": 556
    },
    {
      "epoch": 0.1273142857142857,
      "grad_norm": 1.484190583229065,
      "learning_rate": 0.0005,
      "loss": 3.422792673110962,
      "step": 557
    },
    {
      "epoch": 0.12754285714285715,
      "grad_norm": 1.2247365713119507,
      "learning_rate": 0.0005,
      "loss": 3.348809003829956,
      "step": 558
    },
    {
      "epoch": 0.12777142857142856,
      "grad_norm": 1.461350917816162,
      "learning_rate": 0.0005,
      "loss": 3.252192735671997,
      "step": 559
    },
    {
      "epoch": 0.128,
      "grad_norm": 1.3909591436386108,
      "learning_rate": 0.0005,
      "loss": 3.33501935005188,
      "step": 560
    },
    {
      "epoch": 0.12822857142857144,
      "grad_norm": 1.3328571319580078,
      "learning_rate": 0.0005,
      "loss": 3.3022258281707764,
      "step": 561
    },
    {
      "epoch": 0.12845714285714285,
      "grad_norm": 1.3177751302719116,
      "learning_rate": 0.0005,
      "loss": 3.742689847946167,
      "step": 562
    },
    {
      "epoch": 0.1286857142857143,
      "grad_norm": 1.46737539768219,
      "learning_rate": 0.0005,
      "loss": 3.3949739933013916,
      "step": 563
    },
    {
      "epoch": 0.1289142857142857,
      "grad_norm": 1.5060112476348877,
      "learning_rate": 0.0005,
      "loss": 3.814328670501709,
      "step": 564
    },
    {
      "epoch": 0.12914285714285714,
      "grad_norm": 1.4423567056655884,
      "learning_rate": 0.0005,
      "loss": 3.5666537284851074,
      "step": 565
    },
    {
      "epoch": 0.12937142857142858,
      "grad_norm": 1.4083375930786133,
      "learning_rate": 0.0005,
      "loss": 3.441295623779297,
      "step": 566
    },
    {
      "epoch": 0.1296,
      "grad_norm": 3.4402241706848145,
      "learning_rate": 0.0005,
      "loss": 4.0022125244140625,
      "step": 567
    },
    {
      "epoch": 0.12982857142857143,
      "grad_norm": 1.4908816814422607,
      "learning_rate": 0.0005,
      "loss": 3.6808273792266846,
      "step": 568
    },
    {
      "epoch": 0.13005714285714284,
      "grad_norm": 1.256359577178955,
      "learning_rate": 0.0005,
      "loss": 3.355713367462158,
      "step": 569
    },
    {
      "epoch": 0.13028571428571428,
      "grad_norm": 1.2205216884613037,
      "learning_rate": 0.0005,
      "loss": 3.871544361114502,
      "step": 570
    },
    {
      "epoch": 0.13051428571428572,
      "grad_norm": 1.6265537738800049,
      "learning_rate": 0.0005,
      "loss": 3.3898658752441406,
      "step": 571
    },
    {
      "epoch": 0.13074285714285713,
      "grad_norm": 1.2123340368270874,
      "learning_rate": 0.0005,
      "loss": 3.653550863265991,
      "step": 572
    },
    {
      "epoch": 0.13097142857142857,
      "grad_norm": 1.1956431865692139,
      "learning_rate": 0.0005,
      "loss": 3.5650687217712402,
      "step": 573
    },
    {
      "epoch": 0.1312,
      "grad_norm": 1.519844651222229,
      "learning_rate": 0.0005,
      "loss": 3.3032710552215576,
      "step": 574
    },
    {
      "epoch": 0.13142857142857142,
      "grad_norm": 1.3146233558654785,
      "learning_rate": 0.0005,
      "loss": 3.1596391201019287,
      "step": 575
    },
    {
      "epoch": 0.13165714285714286,
      "grad_norm": 1.2000623941421509,
      "learning_rate": 0.0005,
      "loss": 3.100778102874756,
      "step": 576
    },
    {
      "epoch": 0.13188571428571427,
      "grad_norm": 2.0096302032470703,
      "learning_rate": 0.0005,
      "loss": 3.1353976726531982,
      "step": 577
    },
    {
      "epoch": 0.1321142857142857,
      "grad_norm": 2.079554319381714,
      "learning_rate": 0.0005,
      "loss": 3.736290454864502,
      "step": 578
    },
    {
      "epoch": 0.13234285714285715,
      "grad_norm": 1.5446887016296387,
      "learning_rate": 0.0005,
      "loss": 3.3441972732543945,
      "step": 579
    },
    {
      "epoch": 0.13257142857142856,
      "grad_norm": 2.0336196422576904,
      "learning_rate": 0.0005,
      "loss": 3.2889602184295654,
      "step": 580
    },
    {
      "epoch": 0.1328,
      "grad_norm": 1.835951805114746,
      "learning_rate": 0.0005,
      "loss": 3.435185670852661,
      "step": 581
    },
    {
      "epoch": 0.13302857142857144,
      "grad_norm": 1.8990117311477661,
      "learning_rate": 0.0005,
      "loss": 3.7365307807922363,
      "step": 582
    },
    {
      "epoch": 0.13325714285714285,
      "grad_norm": 1.5874685049057007,
      "learning_rate": 0.0005,
      "loss": 3.3319506645202637,
      "step": 583
    },
    {
      "epoch": 0.1334857142857143,
      "grad_norm": 1.4472984075546265,
      "learning_rate": 0.0005,
      "loss": 3.6664698123931885,
      "step": 584
    },
    {
      "epoch": 0.1337142857142857,
      "grad_norm": 1.2923895120620728,
      "learning_rate": 0.0005,
      "loss": 3.379303455352783,
      "step": 585
    },
    {
      "epoch": 0.13394285714285714,
      "grad_norm": 1.359681487083435,
      "learning_rate": 0.0005,
      "loss": 3.6761245727539062,
      "step": 586
    },
    {
      "epoch": 0.13417142857142858,
      "grad_norm": 1.1012723445892334,
      "learning_rate": 0.0005,
      "loss": 3.6978912353515625,
      "step": 587
    },
    {
      "epoch": 0.1344,
      "grad_norm": 1.0022380352020264,
      "learning_rate": 0.0005,
      "loss": 3.5806684494018555,
      "step": 588
    },
    {
      "epoch": 0.13462857142857143,
      "grad_norm": 1.2099674940109253,
      "learning_rate": 0.0005,
      "loss": 3.3939716815948486,
      "step": 589
    },
    {
      "epoch": 0.13485714285714287,
      "grad_norm": 1.2985740900039673,
      "learning_rate": 0.0005,
      "loss": 3.536775588989258,
      "step": 590
    },
    {
      "epoch": 0.13508571428571428,
      "grad_norm": 1.459088921546936,
      "learning_rate": 0.0005,
      "loss": 3.310791254043579,
      "step": 591
    },
    {
      "epoch": 0.13531428571428572,
      "grad_norm": 1.1078500747680664,
      "learning_rate": 0.0005,
      "loss": 3.7230911254882812,
      "step": 592
    },
    {
      "epoch": 0.13554285714285713,
      "grad_norm": 1.3800822496414185,
      "learning_rate": 0.0005,
      "loss": 3.4981439113616943,
      "step": 593
    },
    {
      "epoch": 0.13577142857142857,
      "grad_norm": 1.3215930461883545,
      "learning_rate": 0.0005,
      "loss": 3.854546308517456,
      "step": 594
    },
    {
      "epoch": 0.136,
      "grad_norm": 1.4542104005813599,
      "learning_rate": 0.0005,
      "loss": 3.287318468093872,
      "step": 595
    },
    {
      "epoch": 0.13622857142857142,
      "grad_norm": 1.402163028717041,
      "learning_rate": 0.0005,
      "loss": 3.339559555053711,
      "step": 596
    },
    {
      "epoch": 0.13645714285714286,
      "grad_norm": 1.2258198261260986,
      "learning_rate": 0.0005,
      "loss": 3.5963294506073,
      "step": 597
    },
    {
      "epoch": 0.1366857142857143,
      "grad_norm": 1.283276081085205,
      "learning_rate": 0.0005,
      "loss": 3.4056408405303955,
      "step": 598
    },
    {
      "epoch": 0.1369142857142857,
      "grad_norm": 1.1975167989730835,
      "learning_rate": 0.0005,
      "loss": 3.5841054916381836,
      "step": 599
    },
    {
      "epoch": 0.13714285714285715,
      "grad_norm": 1.6987180709838867,
      "learning_rate": 0.0005,
      "loss": 4.043990135192871,
      "step": 600
    },
    {
      "epoch": 0.13737142857142856,
      "grad_norm": 1.226423978805542,
      "learning_rate": 0.0005,
      "loss": 3.4773378372192383,
      "step": 601
    },
    {
      "epoch": 0.1376,
      "grad_norm": 1.4880492687225342,
      "learning_rate": 0.0005,
      "loss": 3.8822214603424072,
      "step": 602
    },
    {
      "epoch": 0.13782857142857144,
      "grad_norm": 1.284185528755188,
      "learning_rate": 0.0005,
      "loss": 3.69071102142334,
      "step": 603
    },
    {
      "epoch": 0.13805714285714285,
      "grad_norm": 1.4033482074737549,
      "learning_rate": 0.0005,
      "loss": 3.494607448577881,
      "step": 604
    },
    {
      "epoch": 0.1382857142857143,
      "grad_norm": 1.437669038772583,
      "learning_rate": 0.0005,
      "loss": 3.5918118953704834,
      "step": 605
    },
    {
      "epoch": 0.1385142857142857,
      "grad_norm": 1.1165932416915894,
      "learning_rate": 0.0005,
      "loss": 3.7179677486419678,
      "step": 606
    },
    {
      "epoch": 0.13874285714285714,
      "grad_norm": 1.282517433166504,
      "learning_rate": 0.0005,
      "loss": 3.6873459815979004,
      "step": 607
    },
    {
      "epoch": 0.13897142857142858,
      "grad_norm": 1.2701224088668823,
      "learning_rate": 0.0005,
      "loss": 3.858180046081543,
      "step": 608
    },
    {
      "epoch": 0.1392,
      "grad_norm": 1.1099462509155273,
      "learning_rate": 0.0005,
      "loss": 3.352518081665039,
      "step": 609
    },
    {
      "epoch": 0.13942857142857143,
      "grad_norm": 1.3675005435943604,
      "learning_rate": 0.0005,
      "loss": 3.5248641967773438,
      "step": 610
    },
    {
      "epoch": 0.13965714285714287,
      "grad_norm": 1.2807540893554688,
      "learning_rate": 0.0005,
      "loss": 3.8029770851135254,
      "step": 611
    },
    {
      "epoch": 0.13988571428571428,
      "grad_norm": 1.6636391878128052,
      "learning_rate": 0.0005,
      "loss": 3.4376373291015625,
      "step": 612
    },
    {
      "epoch": 0.14011428571428572,
      "grad_norm": 1.547605276107788,
      "learning_rate": 0.0005,
      "loss": 3.79226016998291,
      "step": 613
    },
    {
      "epoch": 0.14034285714285713,
      "grad_norm": 1.2357211112976074,
      "learning_rate": 0.0005,
      "loss": 3.787445545196533,
      "step": 614
    },
    {
      "epoch": 0.14057142857142857,
      "grad_norm": 1.171739101409912,
      "learning_rate": 0.0005,
      "loss": 3.557544708251953,
      "step": 615
    },
    {
      "epoch": 0.1408,
      "grad_norm": 1.3470102548599243,
      "learning_rate": 0.0005,
      "loss": 3.5617923736572266,
      "step": 616
    },
    {
      "epoch": 0.14102857142857142,
      "grad_norm": 1.8703831434249878,
      "learning_rate": 0.0005,
      "loss": 3.566324234008789,
      "step": 617
    },
    {
      "epoch": 0.14125714285714286,
      "grad_norm": 2.5095255374908447,
      "learning_rate": 0.0005,
      "loss": 3.3058698177337646,
      "step": 618
    },
    {
      "epoch": 0.1414857142857143,
      "grad_norm": 1.2168569564819336,
      "learning_rate": 0.0005,
      "loss": 3.4743947982788086,
      "step": 619
    },
    {
      "epoch": 0.1417142857142857,
      "grad_norm": 1.2147020101547241,
      "learning_rate": 0.0005,
      "loss": 3.756235361099243,
      "step": 620
    },
    {
      "epoch": 0.14194285714285715,
      "grad_norm": 1.5132299661636353,
      "learning_rate": 0.0005,
      "loss": 3.5031731128692627,
      "step": 621
    },
    {
      "epoch": 0.14217142857142856,
      "grad_norm": 2.079705238342285,
      "learning_rate": 0.0005,
      "loss": 3.364931106567383,
      "step": 622
    },
    {
      "epoch": 0.1424,
      "grad_norm": 1.2520803213119507,
      "learning_rate": 0.0005,
      "loss": 3.7662441730499268,
      "step": 623
    },
    {
      "epoch": 0.14262857142857144,
      "grad_norm": 1.2035717964172363,
      "learning_rate": 0.0005,
      "loss": 3.3385846614837646,
      "step": 624
    },
    {
      "epoch": 0.14285714285714285,
      "grad_norm": 1.4913294315338135,
      "learning_rate": 0.0005,
      "loss": 3.6284894943237305,
      "step": 625
    },
    {
      "epoch": 0.1430857142857143,
      "grad_norm": 1.1868605613708496,
      "learning_rate": 0.0005,
      "loss": 3.476626396179199,
      "step": 626
    },
    {
      "epoch": 0.14331428571428573,
      "grad_norm": 1.2471282482147217,
      "learning_rate": 0.0005,
      "loss": 3.557309865951538,
      "step": 627
    },
    {
      "epoch": 0.14354285714285714,
      "grad_norm": 1.2036734819412231,
      "learning_rate": 0.0005,
      "loss": 3.304741382598877,
      "step": 628
    },
    {
      "epoch": 0.14377142857142858,
      "grad_norm": 1.2529958486557007,
      "learning_rate": 0.0005,
      "loss": 3.476818323135376,
      "step": 629
    },
    {
      "epoch": 0.144,
      "grad_norm": 1.2588236331939697,
      "learning_rate": 0.0005,
      "loss": 3.436291217803955,
      "step": 630
    },
    {
      "epoch": 0.14422857142857143,
      "grad_norm": 1.3026185035705566,
      "learning_rate": 0.0005,
      "loss": 3.476910352706909,
      "step": 631
    },
    {
      "epoch": 0.14445714285714287,
      "grad_norm": 1.2607369422912598,
      "learning_rate": 0.0005,
      "loss": 3.566436290740967,
      "step": 632
    },
    {
      "epoch": 0.14468571428571428,
      "grad_norm": 1.3003793954849243,
      "learning_rate": 0.0005,
      "loss": 3.770275354385376,
      "step": 633
    },
    {
      "epoch": 0.14491428571428572,
      "grad_norm": 1.3113476037979126,
      "learning_rate": 0.0005,
      "loss": 3.3048436641693115,
      "step": 634
    },
    {
      "epoch": 0.14514285714285713,
      "grad_norm": 1.2869431972503662,
      "learning_rate": 0.0005,
      "loss": 3.5115203857421875,
      "step": 635
    },
    {
      "epoch": 0.14537142857142857,
      "grad_norm": 1.4517159461975098,
      "learning_rate": 0.0005,
      "loss": 3.470391035079956,
      "step": 636
    },
    {
      "epoch": 0.1456,
      "grad_norm": 1.8329254388809204,
      "learning_rate": 0.0005,
      "loss": 3.224926471710205,
      "step": 637
    },
    {
      "epoch": 0.14582857142857142,
      "grad_norm": 2.1137211322784424,
      "learning_rate": 0.0005,
      "loss": 3.907107353210449,
      "step": 638
    },
    {
      "epoch": 0.14605714285714286,
      "grad_norm": 1.4789263010025024,
      "learning_rate": 0.0005,
      "loss": 3.7963924407958984,
      "step": 639
    },
    {
      "epoch": 0.1462857142857143,
      "grad_norm": 1.7204467058181763,
      "learning_rate": 0.0005,
      "loss": 3.412706136703491,
      "step": 640
    },
    {
      "epoch": 0.1465142857142857,
      "grad_norm": 1.799172282218933,
      "learning_rate": 0.0005,
      "loss": 3.0724806785583496,
      "step": 641
    },
    {
      "epoch": 0.14674285714285715,
      "grad_norm": 1.3268663883209229,
      "learning_rate": 0.0005,
      "loss": 3.6590752601623535,
      "step": 642
    },
    {
      "epoch": 0.14697142857142856,
      "grad_norm": 1.626786470413208,
      "learning_rate": 0.0005,
      "loss": 3.9136409759521484,
      "step": 643
    },
    {
      "epoch": 0.1472,
      "grad_norm": 1.3974729776382446,
      "learning_rate": 0.0005,
      "loss": 3.563929319381714,
      "step": 644
    },
    {
      "epoch": 0.14742857142857144,
      "grad_norm": 1.3189760446548462,
      "learning_rate": 0.0005,
      "loss": 3.6721994876861572,
      "step": 645
    },
    {
      "epoch": 0.14765714285714285,
      "grad_norm": 1.4159667491912842,
      "learning_rate": 0.0005,
      "loss": 3.8910281658172607,
      "step": 646
    },
    {
      "epoch": 0.1478857142857143,
      "grad_norm": 1.3670991659164429,
      "learning_rate": 0.0005,
      "loss": 3.6380555629730225,
      "step": 647
    },
    {
      "epoch": 0.14811428571428573,
      "grad_norm": 1.745355248451233,
      "learning_rate": 0.0005,
      "loss": 3.613377094268799,
      "step": 648
    },
    {
      "epoch": 0.14834285714285714,
      "grad_norm": 1.4147311449050903,
      "learning_rate": 0.0005,
      "loss": 3.5193424224853516,
      "step": 649
    },
    {
      "epoch": 0.14857142857142858,
      "grad_norm": 1.1216942071914673,
      "learning_rate": 0.0005,
      "loss": 3.655290126800537,
      "step": 650
    },
    {
      "epoch": 0.1488,
      "grad_norm": 1.5866591930389404,
      "learning_rate": 0.0005,
      "loss": 3.545962333679199,
      "step": 651
    },
    {
      "epoch": 0.14902857142857143,
      "grad_norm": 1.553415298461914,
      "learning_rate": 0.0005,
      "loss": 3.5763092041015625,
      "step": 652
    },
    {
      "epoch": 0.14925714285714287,
      "grad_norm": 1.2262020111083984,
      "learning_rate": 0.0005,
      "loss": 3.6269922256469727,
      "step": 653
    },
    {
      "epoch": 0.14948571428571428,
      "grad_norm": 1.1643424034118652,
      "learning_rate": 0.0005,
      "loss": 3.6347217559814453,
      "step": 654
    },
    {
      "epoch": 0.14971428571428572,
      "grad_norm": 1.275891661643982,
      "learning_rate": 0.0005,
      "loss": 3.59344482421875,
      "step": 655
    },
    {
      "epoch": 0.14994285714285716,
      "grad_norm": 1.5636616945266724,
      "learning_rate": 0.0005,
      "loss": 3.6251463890075684,
      "step": 656
    },
    {
      "epoch": 0.15017142857142857,
      "grad_norm": 1.479952335357666,
      "learning_rate": 0.0005,
      "loss": 3.274219036102295,
      "step": 657
    },
    {
      "epoch": 0.1504,
      "grad_norm": 1.421256184577942,
      "learning_rate": 0.0005,
      "loss": 3.636701822280884,
      "step": 658
    },
    {
      "epoch": 0.15062857142857142,
      "grad_norm": 1.1822140216827393,
      "learning_rate": 0.0005,
      "loss": 3.2289133071899414,
      "step": 659
    },
    {
      "epoch": 0.15085714285714286,
      "grad_norm": 1.3608454465866089,
      "learning_rate": 0.0005,
      "loss": 3.1614456176757812,
      "step": 660
    },
    {
      "epoch": 0.1510857142857143,
      "grad_norm": 1.8468095064163208,
      "learning_rate": 0.0005,
      "loss": 3.7463529109954834,
      "step": 661
    },
    {
      "epoch": 0.1513142857142857,
      "grad_norm": 1.2846994400024414,
      "learning_rate": 0.0005,
      "loss": 3.2610585689544678,
      "step": 662
    },
    {
      "epoch": 0.15154285714285715,
      "grad_norm": 1.5646328926086426,
      "learning_rate": 0.0005,
      "loss": 3.3279669284820557,
      "step": 663
    },
    {
      "epoch": 0.15177142857142858,
      "grad_norm": 1.2857310771942139,
      "learning_rate": 0.0005,
      "loss": 3.4528889656066895,
      "step": 664
    },
    {
      "epoch": 0.152,
      "grad_norm": 1.4016469717025757,
      "learning_rate": 0.0005,
      "loss": 3.4581847190856934,
      "step": 665
    },
    {
      "epoch": 0.15222857142857144,
      "grad_norm": 1.4908384084701538,
      "learning_rate": 0.0005,
      "loss": 3.3527612686157227,
      "step": 666
    },
    {
      "epoch": 0.15245714285714285,
      "grad_norm": 1.5399065017700195,
      "learning_rate": 0.0005,
      "loss": 3.362276315689087,
      "step": 667
    },
    {
      "epoch": 0.15268571428571429,
      "grad_norm": 1.952296495437622,
      "learning_rate": 0.0005,
      "loss": 3.3309807777404785,
      "step": 668
    },
    {
      "epoch": 0.15291428571428572,
      "grad_norm": 1.4551836252212524,
      "learning_rate": 0.0005,
      "loss": 3.311342477798462,
      "step": 669
    },
    {
      "epoch": 0.15314285714285714,
      "grad_norm": 1.3073878288269043,
      "learning_rate": 0.0005,
      "loss": 3.5546035766601562,
      "step": 670
    },
    {
      "epoch": 0.15337142857142858,
      "grad_norm": 1.0742716789245605,
      "learning_rate": 0.0005,
      "loss": 3.5776925086975098,
      "step": 671
    },
    {
      "epoch": 0.1536,
      "grad_norm": 1.268786907196045,
      "learning_rate": 0.0005,
      "loss": 3.3925466537475586,
      "step": 672
    },
    {
      "epoch": 0.15382857142857143,
      "grad_norm": 1.1807541847229004,
      "learning_rate": 0.0005,
      "loss": 3.8827123641967773,
      "step": 673
    },
    {
      "epoch": 0.15405714285714286,
      "grad_norm": 1.050044298171997,
      "learning_rate": 0.0005,
      "loss": 3.4081687927246094,
      "step": 674
    },
    {
      "epoch": 0.15428571428571428,
      "grad_norm": 1.81032133102417,
      "learning_rate": 0.0005,
      "loss": 3.3619461059570312,
      "step": 675
    },
    {
      "epoch": 0.15451428571428572,
      "grad_norm": 1.6925122737884521,
      "learning_rate": 0.0005,
      "loss": 3.3586626052856445,
      "step": 676
    },
    {
      "epoch": 0.15474285714285715,
      "grad_norm": 1.5579502582550049,
      "learning_rate": 0.0005,
      "loss": 3.569904327392578,
      "step": 677
    },
    {
      "epoch": 0.15497142857142857,
      "grad_norm": 1.3419843912124634,
      "learning_rate": 0.0005,
      "loss": 3.4758684635162354,
      "step": 678
    },
    {
      "epoch": 0.1552,
      "grad_norm": 1.4157066345214844,
      "learning_rate": 0.0005,
      "loss": 3.2922475337982178,
      "step": 679
    },
    {
      "epoch": 0.15542857142857142,
      "grad_norm": 1.4660831689834595,
      "learning_rate": 0.0005,
      "loss": 3.4694457054138184,
      "step": 680
    },
    {
      "epoch": 0.15565714285714286,
      "grad_norm": 1.0126621723175049,
      "learning_rate": 0.0005,
      "loss": 3.6261372566223145,
      "step": 681
    },
    {
      "epoch": 0.1558857142857143,
      "grad_norm": 1.3990116119384766,
      "learning_rate": 0.0005,
      "loss": 3.382201671600342,
      "step": 682
    },
    {
      "epoch": 0.1561142857142857,
      "grad_norm": 2.8307180404663086,
      "learning_rate": 0.0005,
      "loss": 3.6324968338012695,
      "step": 683
    },
    {
      "epoch": 0.15634285714285714,
      "grad_norm": 1.4766181707382202,
      "learning_rate": 0.0005,
      "loss": 3.0196213722229004,
      "step": 684
    },
    {
      "epoch": 0.15657142857142858,
      "grad_norm": 1.148485779762268,
      "learning_rate": 0.0005,
      "loss": 2.968916177749634,
      "step": 685
    },
    {
      "epoch": 0.1568,
      "grad_norm": 1.1039232015609741,
      "learning_rate": 0.0005,
      "loss": 3.6770195960998535,
      "step": 686
    },
    {
      "epoch": 0.15702857142857143,
      "grad_norm": 1.1811952590942383,
      "learning_rate": 0.0005,
      "loss": 3.347137451171875,
      "step": 687
    },
    {
      "epoch": 0.15725714285714285,
      "grad_norm": 1.2020772695541382,
      "learning_rate": 0.0005,
      "loss": 3.1085548400878906,
      "step": 688
    },
    {
      "epoch": 0.15748571428571428,
      "grad_norm": 1.4355370998382568,
      "learning_rate": 0.0005,
      "loss": 3.5485191345214844,
      "step": 689
    },
    {
      "epoch": 0.15771428571428572,
      "grad_norm": 1.0200873613357544,
      "learning_rate": 0.0005,
      "loss": 3.5324690341949463,
      "step": 690
    },
    {
      "epoch": 0.15794285714285713,
      "grad_norm": 1.4921095371246338,
      "learning_rate": 0.0005,
      "loss": 3.356628179550171,
      "step": 691
    },
    {
      "epoch": 0.15817142857142857,
      "grad_norm": 1.5638848543167114,
      "learning_rate": 0.0005,
      "loss": 3.608468532562256,
      "step": 692
    },
    {
      "epoch": 0.1584,
      "grad_norm": 1.4511183500289917,
      "learning_rate": 0.0005,
      "loss": 3.640770435333252,
      "step": 693
    },
    {
      "epoch": 0.15862857142857142,
      "grad_norm": 1.5956618785858154,
      "learning_rate": 0.0005,
      "loss": 3.6087586879730225,
      "step": 694
    },
    {
      "epoch": 0.15885714285714286,
      "grad_norm": 1.2740671634674072,
      "learning_rate": 0.0005,
      "loss": 2.9605512619018555,
      "step": 695
    },
    {
      "epoch": 0.15908571428571427,
      "grad_norm": 1.2319661378860474,
      "learning_rate": 0.0005,
      "loss": 3.509734630584717,
      "step": 696
    },
    {
      "epoch": 0.1593142857142857,
      "grad_norm": 1.4896284341812134,
      "learning_rate": 0.0005,
      "loss": 3.6550559997558594,
      "step": 697
    },
    {
      "epoch": 0.15954285714285715,
      "grad_norm": 1.5934420824050903,
      "learning_rate": 0.0005,
      "loss": 3.283398151397705,
      "step": 698
    },
    {
      "epoch": 0.15977142857142856,
      "grad_norm": 1.232232928276062,
      "learning_rate": 0.0005,
      "loss": 3.1732797622680664,
      "step": 699
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.4828215837478638,
      "learning_rate": 0.0005,
      "loss": 3.0725865364074707,
      "step": 700
    },
    {
      "epoch": 0.16022857142857144,
      "grad_norm": 1.4187902212142944,
      "learning_rate": 0.0005,
      "loss": 3.3093113899230957,
      "step": 701
    },
    {
      "epoch": 0.16045714285714285,
      "grad_norm": 1.4476622343063354,
      "learning_rate": 0.0005,
      "loss": 3.5364272594451904,
      "step": 702
    },
    {
      "epoch": 0.1606857142857143,
      "grad_norm": 1.4255255460739136,
      "learning_rate": 0.0005,
      "loss": 3.194794178009033,
      "step": 703
    },
    {
      "epoch": 0.1609142857142857,
      "grad_norm": 1.6625568866729736,
      "learning_rate": 0.0005,
      "loss": 3.8696727752685547,
      "step": 704
    },
    {
      "epoch": 0.16114285714285714,
      "grad_norm": 1.172251582145691,
      "learning_rate": 0.0005,
      "loss": 3.5154428482055664,
      "step": 705
    },
    {
      "epoch": 0.16137142857142858,
      "grad_norm": 1.4136682748794556,
      "learning_rate": 0.0005,
      "loss": 3.4626803398132324,
      "step": 706
    },
    {
      "epoch": 0.1616,
      "grad_norm": 1.151196837425232,
      "learning_rate": 0.0005,
      "loss": 3.3357467651367188,
      "step": 707
    },
    {
      "epoch": 0.16182857142857143,
      "grad_norm": 1.0954340696334839,
      "learning_rate": 0.0005,
      "loss": 3.3177647590637207,
      "step": 708
    },
    {
      "epoch": 0.16205714285714284,
      "grad_norm": 1.2075839042663574,
      "learning_rate": 0.0005,
      "loss": 3.324951171875,
      "step": 709
    },
    {
      "epoch": 0.16228571428571428,
      "grad_norm": 1.5330203771591187,
      "learning_rate": 0.0005,
      "loss": 3.5176308155059814,
      "step": 710
    },
    {
      "epoch": 0.16251428571428572,
      "grad_norm": 1.538003921508789,
      "learning_rate": 0.0005,
      "loss": 3.42223858833313,
      "step": 711
    },
    {
      "epoch": 0.16274285714285713,
      "grad_norm": 1.1425693035125732,
      "learning_rate": 0.0005,
      "loss": 3.2810497283935547,
      "step": 712
    },
    {
      "epoch": 0.16297142857142857,
      "grad_norm": 1.536577582359314,
      "learning_rate": 0.0005,
      "loss": 3.192049503326416,
      "step": 713
    },
    {
      "epoch": 0.1632,
      "grad_norm": 1.5075563192367554,
      "learning_rate": 0.0005,
      "loss": 3.274887800216675,
      "step": 714
    },
    {
      "epoch": 0.16342857142857142,
      "grad_norm": 1.5091917514801025,
      "learning_rate": 0.0005,
      "loss": 3.1419010162353516,
      "step": 715
    },
    {
      "epoch": 0.16365714285714286,
      "grad_norm": 1.3737633228302002,
      "learning_rate": 0.0005,
      "loss": 3.9571595191955566,
      "step": 716
    },
    {
      "epoch": 0.16388571428571427,
      "grad_norm": 1.2288603782653809,
      "learning_rate": 0.0005,
      "loss": 2.9504246711730957,
      "step": 717
    },
    {
      "epoch": 0.1641142857142857,
      "grad_norm": 1.3389612436294556,
      "learning_rate": 0.0005,
      "loss": 3.3575387001037598,
      "step": 718
    },
    {
      "epoch": 0.16434285714285715,
      "grad_norm": 1.3712737560272217,
      "learning_rate": 0.0005,
      "loss": 3.4121909141540527,
      "step": 719
    },
    {
      "epoch": 0.16457142857142856,
      "grad_norm": 1.3736587762832642,
      "learning_rate": 0.0005,
      "loss": 3.2728888988494873,
      "step": 720
    },
    {
      "epoch": 0.1648,
      "grad_norm": 1.422711730003357,
      "learning_rate": 0.0005,
      "loss": 3.309037208557129,
      "step": 721
    },
    {
      "epoch": 0.16502857142857144,
      "grad_norm": 1.4447541236877441,
      "learning_rate": 0.0005,
      "loss": 3.5878562927246094,
      "step": 722
    },
    {
      "epoch": 0.16525714285714285,
      "grad_norm": 1.1427092552185059,
      "learning_rate": 0.0005,
      "loss": 3.3669111728668213,
      "step": 723
    },
    {
      "epoch": 0.1654857142857143,
      "grad_norm": 2.0631566047668457,
      "learning_rate": 0.0005,
      "loss": 3.238676071166992,
      "step": 724
    },
    {
      "epoch": 0.1657142857142857,
      "grad_norm": 1.103036880493164,
      "learning_rate": 0.0005,
      "loss": 3.1033096313476562,
      "step": 725
    },
    {
      "epoch": 0.16594285714285714,
      "grad_norm": 1.471277117729187,
      "learning_rate": 0.0005,
      "loss": 3.763969898223877,
      "step": 726
    },
    {
      "epoch": 0.16617142857142858,
      "grad_norm": 1.2861872911453247,
      "learning_rate": 0.0005,
      "loss": 3.3704519271850586,
      "step": 727
    },
    {
      "epoch": 0.1664,
      "grad_norm": 1.6147068738937378,
      "learning_rate": 0.0005,
      "loss": 3.393894672393799,
      "step": 728
    },
    {
      "epoch": 0.16662857142857143,
      "grad_norm": 1.3040026426315308,
      "learning_rate": 0.0005,
      "loss": 3.8129780292510986,
      "step": 729
    },
    {
      "epoch": 0.16685714285714287,
      "grad_norm": 1.1040351390838623,
      "learning_rate": 0.0005,
      "loss": 3.30159592628479,
      "step": 730
    },
    {
      "epoch": 0.16708571428571428,
      "grad_norm": 1.5733098983764648,
      "learning_rate": 0.0005,
      "loss": 3.3921592235565186,
      "step": 731
    },
    {
      "epoch": 0.16731428571428572,
      "grad_norm": 1.2765247821807861,
      "learning_rate": 0.0005,
      "loss": 3.6721808910369873,
      "step": 732
    },
    {
      "epoch": 0.16754285714285713,
      "grad_norm": 1.1141701936721802,
      "learning_rate": 0.0005,
      "loss": 3.213315486907959,
      "step": 733
    },
    {
      "epoch": 0.16777142857142857,
      "grad_norm": 1.4024940729141235,
      "learning_rate": 0.0005,
      "loss": 3.4572978019714355,
      "step": 734
    },
    {
      "epoch": 0.168,
      "grad_norm": 1.3304640054702759,
      "learning_rate": 0.0005,
      "loss": 3.593047618865967,
      "step": 735
    },
    {
      "epoch": 0.16822857142857142,
      "grad_norm": 1.1333320140838623,
      "learning_rate": 0.0005,
      "loss": 3.5745930671691895,
      "step": 736
    },
    {
      "epoch": 0.16845714285714286,
      "grad_norm": 1.5142256021499634,
      "learning_rate": 0.0005,
      "loss": 3.28049898147583,
      "step": 737
    },
    {
      "epoch": 0.16868571428571427,
      "grad_norm": 1.6863270998001099,
      "learning_rate": 0.0005,
      "loss": 3.396970272064209,
      "step": 738
    },
    {
      "epoch": 0.1689142857142857,
      "grad_norm": 1.4536428451538086,
      "learning_rate": 0.0005,
      "loss": 3.3782172203063965,
      "step": 739
    },
    {
      "epoch": 0.16914285714285715,
      "grad_norm": 1.6713532209396362,
      "learning_rate": 0.0005,
      "loss": 4.00486946105957,
      "step": 740
    },
    {
      "epoch": 0.16937142857142856,
      "grad_norm": 1.3873757123947144,
      "learning_rate": 0.0005,
      "loss": 3.384594440460205,
      "step": 741
    },
    {
      "epoch": 0.1696,
      "grad_norm": 1.2142024040222168,
      "learning_rate": 0.0005,
      "loss": 3.52667498588562,
      "step": 742
    },
    {
      "epoch": 0.16982857142857144,
      "grad_norm": 1.7076416015625,
      "learning_rate": 0.0005,
      "loss": 3.0925724506378174,
      "step": 743
    },
    {
      "epoch": 0.17005714285714285,
      "grad_norm": 1.5104230642318726,
      "learning_rate": 0.0005,
      "loss": 3.462310552597046,
      "step": 744
    },
    {
      "epoch": 0.1702857142857143,
      "grad_norm": 1.2111618518829346,
      "learning_rate": 0.0005,
      "loss": 3.755767345428467,
      "step": 745
    },
    {
      "epoch": 0.1705142857142857,
      "grad_norm": 1.2379568815231323,
      "learning_rate": 0.0005,
      "loss": 3.4094300270080566,
      "step": 746
    },
    {
      "epoch": 0.17074285714285714,
      "grad_norm": 1.5253794193267822,
      "learning_rate": 0.0005,
      "loss": 3.5627565383911133,
      "step": 747
    },
    {
      "epoch": 0.17097142857142858,
      "grad_norm": 1.579489827156067,
      "learning_rate": 0.0005,
      "loss": 3.707374095916748,
      "step": 748
    },
    {
      "epoch": 0.1712,
      "grad_norm": 1.1579872369766235,
      "learning_rate": 0.0005,
      "loss": 3.3120527267456055,
      "step": 749
    },
    {
      "epoch": 0.17142857142857143,
      "grad_norm": 1.9357930421829224,
      "learning_rate": 0.0005,
      "loss": 3.3308768272399902,
      "step": 750
    },
    {
      "epoch": 0.17165714285714287,
      "grad_norm": 2.3919677734375,
      "learning_rate": 0.0005,
      "loss": 3.5178256034851074,
      "step": 751
    },
    {
      "epoch": 0.17188571428571428,
      "grad_norm": 1.310799241065979,
      "learning_rate": 0.0005,
      "loss": 3.1977059841156006,
      "step": 752
    },
    {
      "epoch": 0.17211428571428572,
      "grad_norm": 1.2285916805267334,
      "learning_rate": 0.0005,
      "loss": 3.620934009552002,
      "step": 753
    },
    {
      "epoch": 0.17234285714285713,
      "grad_norm": 1.2503358125686646,
      "learning_rate": 0.0005,
      "loss": 3.2486572265625,
      "step": 754
    },
    {
      "epoch": 0.17257142857142857,
      "grad_norm": 1.517996072769165,
      "learning_rate": 0.0005,
      "loss": 3.7959978580474854,
      "step": 755
    },
    {
      "epoch": 0.1728,
      "grad_norm": 1.4651169776916504,
      "learning_rate": 0.0005,
      "loss": 3.459908962249756,
      "step": 756
    },
    {
      "epoch": 0.17302857142857142,
      "grad_norm": 1.9417372941970825,
      "learning_rate": 0.0005,
      "loss": 3.078817129135132,
      "step": 757
    },
    {
      "epoch": 0.17325714285714286,
      "grad_norm": 1.2261064052581787,
      "learning_rate": 0.0005,
      "loss": 3.3783440589904785,
      "step": 758
    },
    {
      "epoch": 0.1734857142857143,
      "grad_norm": 1.2176097631454468,
      "learning_rate": 0.0005,
      "loss": 3.5626909732818604,
      "step": 759
    },
    {
      "epoch": 0.1737142857142857,
      "grad_norm": 1.210042953491211,
      "learning_rate": 0.0005,
      "loss": 3.242164134979248,
      "step": 760
    },
    {
      "epoch": 0.17394285714285715,
      "grad_norm": 1.033092975616455,
      "learning_rate": 0.0005,
      "loss": 3.416628837585449,
      "step": 761
    },
    {
      "epoch": 0.17417142857142856,
      "grad_norm": 1.0386143922805786,
      "learning_rate": 0.0005,
      "loss": 3.602151393890381,
      "step": 762
    },
    {
      "epoch": 0.1744,
      "grad_norm": 1.4977697134017944,
      "learning_rate": 0.0005,
      "loss": 3.3743457794189453,
      "step": 763
    },
    {
      "epoch": 0.17462857142857144,
      "grad_norm": 1.0918523073196411,
      "learning_rate": 0.0005,
      "loss": 3.503119468688965,
      "step": 764
    },
    {
      "epoch": 0.17485714285714285,
      "grad_norm": 1.2838093042373657,
      "learning_rate": 0.0005,
      "loss": 3.495863437652588,
      "step": 765
    },
    {
      "epoch": 0.1750857142857143,
      "grad_norm": 1.2377965450286865,
      "learning_rate": 0.0005,
      "loss": 3.2405810356140137,
      "step": 766
    },
    {
      "epoch": 0.17531428571428573,
      "grad_norm": 1.1284704208374023,
      "learning_rate": 0.0005,
      "loss": 3.410153388977051,
      "step": 767
    },
    {
      "epoch": 0.17554285714285714,
      "grad_norm": 1.2999035120010376,
      "learning_rate": 0.0005,
      "loss": 3.538094997406006,
      "step": 768
    },
    {
      "epoch": 0.17577142857142858,
      "grad_norm": 1.175499677658081,
      "learning_rate": 0.0005,
      "loss": 3.408811330795288,
      "step": 769
    },
    {
      "epoch": 0.176,
      "grad_norm": 1.2492890357971191,
      "learning_rate": 0.0005,
      "loss": 3.548922300338745,
      "step": 770
    },
    {
      "epoch": 0.17622857142857143,
      "grad_norm": 1.1940655708312988,
      "learning_rate": 0.0005,
      "loss": 3.3789560794830322,
      "step": 771
    },
    {
      "epoch": 0.17645714285714287,
      "grad_norm": 1.2049988508224487,
      "learning_rate": 0.0005,
      "loss": 3.2718453407287598,
      "step": 772
    },
    {
      "epoch": 0.17668571428571428,
      "grad_norm": 1.4327411651611328,
      "learning_rate": 0.0005,
      "loss": 3.3938565254211426,
      "step": 773
    },
    {
      "epoch": 0.17691428571428572,
      "grad_norm": 1.5136051177978516,
      "learning_rate": 0.0005,
      "loss": 3.642425060272217,
      "step": 774
    },
    {
      "epoch": 0.17714285714285713,
      "grad_norm": 1.4190528392791748,
      "learning_rate": 0.0005,
      "loss": 2.9305577278137207,
      "step": 775
    },
    {
      "epoch": 0.17737142857142857,
      "grad_norm": 1.310421109199524,
      "learning_rate": 0.0005,
      "loss": 4.066289901733398,
      "step": 776
    },
    {
      "epoch": 0.1776,
      "grad_norm": 1.214053988456726,
      "learning_rate": 0.0005,
      "loss": 3.4543051719665527,
      "step": 777
    },
    {
      "epoch": 0.17782857142857142,
      "grad_norm": 1.4018194675445557,
      "learning_rate": 0.0005,
      "loss": 3.579622507095337,
      "step": 778
    },
    {
      "epoch": 0.17805714285714286,
      "grad_norm": 2.8399224281311035,
      "learning_rate": 0.0005,
      "loss": 3.6045167446136475,
      "step": 779
    },
    {
      "epoch": 0.1782857142857143,
      "grad_norm": 2.061922311782837,
      "learning_rate": 0.0005,
      "loss": 3.46919584274292,
      "step": 780
    },
    {
      "epoch": 0.1785142857142857,
      "grad_norm": 2.031554698944092,
      "learning_rate": 0.0005,
      "loss": 3.261375904083252,
      "step": 781
    },
    {
      "epoch": 0.17874285714285715,
      "grad_norm": 1.7830232381820679,
      "learning_rate": 0.0005,
      "loss": 3.8172030448913574,
      "step": 782
    },
    {
      "epoch": 0.17897142857142856,
      "grad_norm": 1.3518677949905396,
      "learning_rate": 0.0005,
      "loss": 3.2652459144592285,
      "step": 783
    },
    {
      "epoch": 0.1792,
      "grad_norm": 1.1786571741104126,
      "learning_rate": 0.0005,
      "loss": 3.130833148956299,
      "step": 784
    },
    {
      "epoch": 0.17942857142857144,
      "grad_norm": 1.2926353216171265,
      "learning_rate": 0.0005,
      "loss": 3.7334749698638916,
      "step": 785
    },
    {
      "epoch": 0.17965714285714285,
      "grad_norm": 1.3999463319778442,
      "learning_rate": 0.0005,
      "loss": 3.176736831665039,
      "step": 786
    },
    {
      "epoch": 0.1798857142857143,
      "grad_norm": 1.1806037425994873,
      "learning_rate": 0.0005,
      "loss": 3.078035593032837,
      "step": 787
    },
    {
      "epoch": 0.18011428571428573,
      "grad_norm": 1.7273372411727905,
      "learning_rate": 0.0005,
      "loss": 3.068572521209717,
      "step": 788
    },
    {
      "epoch": 0.18034285714285714,
      "grad_norm": 2.1302719116210938,
      "learning_rate": 0.0005,
      "loss": 3.34804105758667,
      "step": 789
    },
    {
      "epoch": 0.18057142857142858,
      "grad_norm": 1.3298174142837524,
      "learning_rate": 0.0005,
      "loss": 3.565544605255127,
      "step": 790
    },
    {
      "epoch": 0.1808,
      "grad_norm": 1.4901196956634521,
      "learning_rate": 0.0005,
      "loss": 3.1334729194641113,
      "step": 791
    },
    {
      "epoch": 0.18102857142857143,
      "grad_norm": 1.1765938997268677,
      "learning_rate": 0.0005,
      "loss": 3.5784287452697754,
      "step": 792
    },
    {
      "epoch": 0.18125714285714287,
      "grad_norm": 1.4854779243469238,
      "learning_rate": 0.0005,
      "loss": 3.634225845336914,
      "step": 793
    },
    {
      "epoch": 0.18148571428571428,
      "grad_norm": 1.3171409368515015,
      "learning_rate": 0.0005,
      "loss": 3.40967059135437,
      "step": 794
    },
    {
      "epoch": 0.18171428571428572,
      "grad_norm": 1.5961982011795044,
      "learning_rate": 0.0005,
      "loss": 3.313776969909668,
      "step": 795
    },
    {
      "epoch": 0.18194285714285716,
      "grad_norm": 1.3275343179702759,
      "learning_rate": 0.0005,
      "loss": 2.8698010444641113,
      "step": 796
    },
    {
      "epoch": 0.18217142857142857,
      "grad_norm": 1.263198733329773,
      "learning_rate": 0.0005,
      "loss": 3.394155979156494,
      "step": 797
    },
    {
      "epoch": 0.1824,
      "grad_norm": 1.2064757347106934,
      "learning_rate": 0.0005,
      "loss": 3.2576937675476074,
      "step": 798
    },
    {
      "epoch": 0.18262857142857142,
      "grad_norm": 1.4544663429260254,
      "learning_rate": 0.0005,
      "loss": 3.088834762573242,
      "step": 799
    },
    {
      "epoch": 0.18285714285714286,
      "grad_norm": 1.1915557384490967,
      "learning_rate": 0.0005,
      "loss": 3.4302875995635986,
      "step": 800
    },
    {
      "epoch": 0.1830857142857143,
      "grad_norm": 1.4596924781799316,
      "learning_rate": 0.0005,
      "loss": 3.4414615631103516,
      "step": 801
    },
    {
      "epoch": 0.1833142857142857,
      "grad_norm": 1.2596290111541748,
      "learning_rate": 0.0005,
      "loss": 3.6663506031036377,
      "step": 802
    },
    {
      "epoch": 0.18354285714285715,
      "grad_norm": 1.151126742362976,
      "learning_rate": 0.0005,
      "loss": 3.328634738922119,
      "step": 803
    },
    {
      "epoch": 0.18377142857142856,
      "grad_norm": 1.226895809173584,
      "learning_rate": 0.0005,
      "loss": 3.433901786804199,
      "step": 804
    },
    {
      "epoch": 0.184,
      "grad_norm": 1.415116548538208,
      "learning_rate": 0.0005,
      "loss": 3.4474925994873047,
      "step": 805
    },
    {
      "epoch": 0.18422857142857144,
      "grad_norm": 1.3859137296676636,
      "learning_rate": 0.0005,
      "loss": 3.7253692150115967,
      "step": 806
    },
    {
      "epoch": 0.18445714285714285,
      "grad_norm": 1.2533220052719116,
      "learning_rate": 0.0005,
      "loss": 3.3100061416625977,
      "step": 807
    },
    {
      "epoch": 0.1846857142857143,
      "grad_norm": 1.1669399738311768,
      "learning_rate": 0.0005,
      "loss": 3.335317850112915,
      "step": 808
    },
    {
      "epoch": 0.18491428571428573,
      "grad_norm": 1.4628775119781494,
      "learning_rate": 0.0005,
      "loss": 4.032797813415527,
      "step": 809
    },
    {
      "epoch": 0.18514285714285714,
      "grad_norm": 1.4911201000213623,
      "learning_rate": 0.0005,
      "loss": 3.2063229084014893,
      "step": 810
    },
    {
      "epoch": 0.18537142857142858,
      "grad_norm": 1.2215319871902466,
      "learning_rate": 0.0005,
      "loss": 3.691534996032715,
      "step": 811
    },
    {
      "epoch": 0.1856,
      "grad_norm": 1.1892108917236328,
      "learning_rate": 0.0005,
      "loss": 3.400261878967285,
      "step": 812
    },
    {
      "epoch": 0.18582857142857143,
      "grad_norm": 1.235018014907837,
      "learning_rate": 0.0005,
      "loss": 3.645078659057617,
      "step": 813
    },
    {
      "epoch": 0.18605714285714287,
      "grad_norm": 1.4645570516586304,
      "learning_rate": 0.0005,
      "loss": 3.3295655250549316,
      "step": 814
    },
    {
      "epoch": 0.18628571428571428,
      "grad_norm": 1.3131418228149414,
      "learning_rate": 0.0005,
      "loss": 3.317436695098877,
      "step": 815
    },
    {
      "epoch": 0.18651428571428572,
      "grad_norm": 1.3780702352523804,
      "learning_rate": 0.0005,
      "loss": 3.5878562927246094,
      "step": 816
    },
    {
      "epoch": 0.18674285714285715,
      "grad_norm": 1.5116751194000244,
      "learning_rate": 0.0005,
      "loss": 3.4530115127563477,
      "step": 817
    },
    {
      "epoch": 0.18697142857142857,
      "grad_norm": 1.497693419456482,
      "learning_rate": 0.0005,
      "loss": 3.3805742263793945,
      "step": 818
    },
    {
      "epoch": 0.1872,
      "grad_norm": 1.2178157567977905,
      "learning_rate": 0.0005,
      "loss": 3.5300209522247314,
      "step": 819
    },
    {
      "epoch": 0.18742857142857142,
      "grad_norm": 1.1870453357696533,
      "learning_rate": 0.0005,
      "loss": 2.796513557434082,
      "step": 820
    },
    {
      "epoch": 0.18765714285714286,
      "grad_norm": 1.569715976715088,
      "learning_rate": 0.0005,
      "loss": 3.144012928009033,
      "step": 821
    },
    {
      "epoch": 0.1878857142857143,
      "grad_norm": 1.1166415214538574,
      "learning_rate": 0.0005,
      "loss": 3.072173595428467,
      "step": 822
    },
    {
      "epoch": 0.1881142857142857,
      "grad_norm": 1.1478090286254883,
      "learning_rate": 0.0005,
      "loss": 3.642240524291992,
      "step": 823
    },
    {
      "epoch": 0.18834285714285715,
      "grad_norm": 1.1041889190673828,
      "learning_rate": 0.0005,
      "loss": 3.4214258193969727,
      "step": 824
    },
    {
      "epoch": 0.18857142857142858,
      "grad_norm": 1.6530723571777344,
      "learning_rate": 0.0005,
      "loss": 3.2581381797790527,
      "step": 825
    },
    {
      "epoch": 0.1888,
      "grad_norm": 1.1681463718414307,
      "learning_rate": 0.0005,
      "loss": 3.3535332679748535,
      "step": 826
    },
    {
      "epoch": 0.18902857142857143,
      "grad_norm": 1.4680942296981812,
      "learning_rate": 0.0005,
      "loss": 3.8358235359191895,
      "step": 827
    },
    {
      "epoch": 0.18925714285714285,
      "grad_norm": 1.2672994136810303,
      "learning_rate": 0.0005,
      "loss": 3.742711067199707,
      "step": 828
    },
    {
      "epoch": 0.18948571428571429,
      "grad_norm": 1.5435974597930908,
      "learning_rate": 0.0005,
      "loss": 3.2735018730163574,
      "step": 829
    },
    {
      "epoch": 0.18971428571428572,
      "grad_norm": 1.1255220174789429,
      "learning_rate": 0.0005,
      "loss": 3.630632162094116,
      "step": 830
    },
    {
      "epoch": 0.18994285714285714,
      "grad_norm": 1.2306891679763794,
      "learning_rate": 0.0005,
      "loss": 3.584035634994507,
      "step": 831
    },
    {
      "epoch": 0.19017142857142857,
      "grad_norm": 1.163008213043213,
      "learning_rate": 0.0005,
      "loss": 3.439211368560791,
      "step": 832
    },
    {
      "epoch": 0.1904,
      "grad_norm": 1.6986901760101318,
      "learning_rate": 0.0005,
      "loss": 3.1426620483398438,
      "step": 833
    },
    {
      "epoch": 0.19062857142857143,
      "grad_norm": 1.247002124786377,
      "learning_rate": 0.0005,
      "loss": 3.28322696685791,
      "step": 834
    },
    {
      "epoch": 0.19085714285714286,
      "grad_norm": 1.4263784885406494,
      "learning_rate": 0.0005,
      "loss": 3.315540075302124,
      "step": 835
    },
    {
      "epoch": 0.19108571428571428,
      "grad_norm": 1.2361615896224976,
      "learning_rate": 0.0005,
      "loss": 3.0209550857543945,
      "step": 836
    },
    {
      "epoch": 0.19131428571428571,
      "grad_norm": 1.3899613618850708,
      "learning_rate": 0.0005,
      "loss": 3.211899757385254,
      "step": 837
    },
    {
      "epoch": 0.19154285714285715,
      "grad_norm": 1.285293459892273,
      "learning_rate": 0.0005,
      "loss": 3.051928997039795,
      "step": 838
    },
    {
      "epoch": 0.19177142857142856,
      "grad_norm": 1.4558545351028442,
      "learning_rate": 0.0005,
      "loss": 3.5843491554260254,
      "step": 839
    },
    {
      "epoch": 0.192,
      "grad_norm": 1.693636178970337,
      "learning_rate": 0.0005,
      "loss": 3.680375337600708,
      "step": 840
    },
    {
      "epoch": 0.19222857142857142,
      "grad_norm": 1.2738311290740967,
      "learning_rate": 0.0005,
      "loss": 3.3327090740203857,
      "step": 841
    },
    {
      "epoch": 0.19245714285714285,
      "grad_norm": 1.048026204109192,
      "learning_rate": 0.0005,
      "loss": 3.781127452850342,
      "step": 842
    },
    {
      "epoch": 0.1926857142857143,
      "grad_norm": 1.3493512868881226,
      "learning_rate": 0.0005,
      "loss": 3.208629608154297,
      "step": 843
    },
    {
      "epoch": 0.1929142857142857,
      "grad_norm": 1.3515114784240723,
      "learning_rate": 0.0005,
      "loss": 3.083198070526123,
      "step": 844
    },
    {
      "epoch": 0.19314285714285714,
      "grad_norm": 1.3895902633666992,
      "learning_rate": 0.0005,
      "loss": 3.458336353302002,
      "step": 845
    },
    {
      "epoch": 0.19337142857142858,
      "grad_norm": 1.2008075714111328,
      "learning_rate": 0.0005,
      "loss": 3.6520469188690186,
      "step": 846
    },
    {
      "epoch": 0.1936,
      "grad_norm": 1.3718748092651367,
      "learning_rate": 0.0005,
      "loss": 3.349013328552246,
      "step": 847
    },
    {
      "epoch": 0.19382857142857143,
      "grad_norm": 1.9208991527557373,
      "learning_rate": 0.0005,
      "loss": 3.121079683303833,
      "step": 848
    },
    {
      "epoch": 0.19405714285714284,
      "grad_norm": 1.444135308265686,
      "learning_rate": 0.0005,
      "loss": 3.357985496520996,
      "step": 849
    },
    {
      "epoch": 0.19428571428571428,
      "grad_norm": 1.1111077070236206,
      "learning_rate": 0.0005,
      "loss": 3.293792247772217,
      "step": 850
    },
    {
      "epoch": 0.19451428571428572,
      "grad_norm": 1.2541404962539673,
      "learning_rate": 0.0005,
      "loss": 3.4546680450439453,
      "step": 851
    },
    {
      "epoch": 0.19474285714285713,
      "grad_norm": 1.7054702043533325,
      "learning_rate": 0.0005,
      "loss": 3.1289000511169434,
      "step": 852
    },
    {
      "epoch": 0.19497142857142857,
      "grad_norm": 1.3415508270263672,
      "learning_rate": 0.0005,
      "loss": 3.743967056274414,
      "step": 853
    },
    {
      "epoch": 0.1952,
      "grad_norm": 1.2374346256256104,
      "learning_rate": 0.0005,
      "loss": 3.027939558029175,
      "step": 854
    },
    {
      "epoch": 0.19542857142857142,
      "grad_norm": 1.290990948677063,
      "learning_rate": 0.0005,
      "loss": 3.5765504837036133,
      "step": 855
    },
    {
      "epoch": 0.19565714285714286,
      "grad_norm": 1.564489722251892,
      "learning_rate": 0.0005,
      "loss": 3.5784921646118164,
      "step": 856
    },
    {
      "epoch": 0.19588571428571427,
      "grad_norm": 1.6282681226730347,
      "learning_rate": 0.0005,
      "loss": 3.1227574348449707,
      "step": 857
    },
    {
      "epoch": 0.1961142857142857,
      "grad_norm": 1.6442543268203735,
      "learning_rate": 0.0005,
      "loss": 3.30216383934021,
      "step": 858
    },
    {
      "epoch": 0.19634285714285715,
      "grad_norm": 1.509804129600525,
      "learning_rate": 0.0005,
      "loss": 3.1397106647491455,
      "step": 859
    },
    {
      "epoch": 0.19657142857142856,
      "grad_norm": 1.0371224880218506,
      "learning_rate": 0.0005,
      "loss": 3.085520029067993,
      "step": 860
    },
    {
      "epoch": 0.1968,
      "grad_norm": 1.3678643703460693,
      "learning_rate": 0.0005,
      "loss": 3.4616892337799072,
      "step": 861
    },
    {
      "epoch": 0.19702857142857144,
      "grad_norm": 1.36771821975708,
      "learning_rate": 0.0005,
      "loss": 3.5331802368164062,
      "step": 862
    },
    {
      "epoch": 0.19725714285714285,
      "grad_norm": 1.181589126586914,
      "learning_rate": 0.0005,
      "loss": 3.7664098739624023,
      "step": 863
    },
    {
      "epoch": 0.1974857142857143,
      "grad_norm": 1.381927251815796,
      "learning_rate": 0.0005,
      "loss": 3.33890962600708,
      "step": 864
    },
    {
      "epoch": 0.1977142857142857,
      "grad_norm": 1.3730456829071045,
      "learning_rate": 0.0005,
      "loss": 3.086278200149536,
      "step": 865
    },
    {
      "epoch": 0.19794285714285714,
      "grad_norm": 1.606675624847412,
      "learning_rate": 0.0005,
      "loss": 2.976083755493164,
      "step": 866
    },
    {
      "epoch": 0.19817142857142858,
      "grad_norm": 1.1358728408813477,
      "learning_rate": 0.0005,
      "loss": 3.2258710861206055,
      "step": 867
    },
    {
      "epoch": 0.1984,
      "grad_norm": 1.2353847026824951,
      "learning_rate": 0.0005,
      "loss": 3.4063148498535156,
      "step": 868
    },
    {
      "epoch": 0.19862857142857143,
      "grad_norm": 1.7992868423461914,
      "learning_rate": 0.0005,
      "loss": 3.4500744342803955,
      "step": 869
    },
    {
      "epoch": 0.19885714285714284,
      "grad_norm": 1.2144564390182495,
      "learning_rate": 0.0005,
      "loss": 3.5607681274414062,
      "step": 870
    },
    {
      "epoch": 0.19908571428571428,
      "grad_norm": 1.1863081455230713,
      "learning_rate": 0.0005,
      "loss": 3.594755172729492,
      "step": 871
    },
    {
      "epoch": 0.19931428571428572,
      "grad_norm": 1.4645036458969116,
      "learning_rate": 0.0005,
      "loss": 3.349208354949951,
      "step": 872
    },
    {
      "epoch": 0.19954285714285713,
      "grad_norm": 1.419419527053833,
      "learning_rate": 0.0005,
      "loss": 3.3570375442504883,
      "step": 873
    },
    {
      "epoch": 0.19977142857142857,
      "grad_norm": 1.3876653909683228,
      "learning_rate": 0.0005,
      "loss": 3.1916184425354004,
      "step": 874
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.4138531684875488,
      "learning_rate": 0.0005,
      "loss": 3.1438796520233154,
      "step": 875
    },
    {
      "epoch": 0.20022857142857142,
      "grad_norm": 1.2738038301467896,
      "learning_rate": 0.0005,
      "loss": 3.1906750202178955,
      "step": 876
    },
    {
      "epoch": 0.20045714285714286,
      "grad_norm": 1.463143229484558,
      "learning_rate": 0.0005,
      "loss": 3.1650230884552,
      "step": 877
    },
    {
      "epoch": 0.20068571428571427,
      "grad_norm": 1.1147704124450684,
      "learning_rate": 0.0005,
      "loss": 3.420649528503418,
      "step": 878
    },
    {
      "epoch": 0.2009142857142857,
      "grad_norm": 1.2013472318649292,
      "learning_rate": 0.0005,
      "loss": 3.5879695415496826,
      "step": 879
    },
    {
      "epoch": 0.20114285714285715,
      "grad_norm": 1.4077320098876953,
      "learning_rate": 0.0005,
      "loss": 3.34086012840271,
      "step": 880
    },
    {
      "epoch": 0.20137142857142856,
      "grad_norm": 1.2488815784454346,
      "learning_rate": 0.0005,
      "loss": 3.189689874649048,
      "step": 881
    },
    {
      "epoch": 0.2016,
      "grad_norm": 1.311488389968872,
      "learning_rate": 0.0005,
      "loss": 3.259587049484253,
      "step": 882
    },
    {
      "epoch": 0.20182857142857144,
      "grad_norm": 1.6804254055023193,
      "learning_rate": 0.0005,
      "loss": 3.2890408039093018,
      "step": 883
    },
    {
      "epoch": 0.20205714285714285,
      "grad_norm": 1.4658805131912231,
      "learning_rate": 0.0005,
      "loss": 2.770874500274658,
      "step": 884
    },
    {
      "epoch": 0.2022857142857143,
      "grad_norm": 1.553344488143921,
      "learning_rate": 0.0005,
      "loss": 3.5323731899261475,
      "step": 885
    },
    {
      "epoch": 0.2025142857142857,
      "grad_norm": 1.7214059829711914,
      "learning_rate": 0.0005,
      "loss": 3.197622060775757,
      "step": 886
    },
    {
      "epoch": 0.20274285714285714,
      "grad_norm": 1.3158998489379883,
      "learning_rate": 0.0005,
      "loss": 3.4128100872039795,
      "step": 887
    },
    {
      "epoch": 0.20297142857142858,
      "grad_norm": 1.3635317087173462,
      "learning_rate": 0.0005,
      "loss": 3.6824026107788086,
      "step": 888
    },
    {
      "epoch": 0.2032,
      "grad_norm": 1.6625443696975708,
      "learning_rate": 0.0005,
      "loss": 3.515289783477783,
      "step": 889
    },
    {
      "epoch": 0.20342857142857143,
      "grad_norm": 1.4266364574432373,
      "learning_rate": 0.0005,
      "loss": 3.4837799072265625,
      "step": 890
    },
    {
      "epoch": 0.20365714285714287,
      "grad_norm": 1.3793355226516724,
      "learning_rate": 0.0005,
      "loss": 3.533759832382202,
      "step": 891
    },
    {
      "epoch": 0.20388571428571428,
      "grad_norm": 1.2525748014450073,
      "learning_rate": 0.0005,
      "loss": 3.2774765491485596,
      "step": 892
    },
    {
      "epoch": 0.20411428571428572,
      "grad_norm": 1.387134313583374,
      "learning_rate": 0.0005,
      "loss": 3.443859338760376,
      "step": 893
    },
    {
      "epoch": 0.20434285714285713,
      "grad_norm": 1.1954214572906494,
      "learning_rate": 0.0005,
      "loss": 3.3295373916625977,
      "step": 894
    },
    {
      "epoch": 0.20457142857142857,
      "grad_norm": 1.2390857934951782,
      "learning_rate": 0.0005,
      "loss": 3.238279342651367,
      "step": 895
    },
    {
      "epoch": 0.2048,
      "grad_norm": 1.5448769330978394,
      "learning_rate": 0.0005,
      "loss": 3.30766224861145,
      "step": 896
    },
    {
      "epoch": 0.20502857142857142,
      "grad_norm": 1.2963709831237793,
      "learning_rate": 0.0005,
      "loss": 3.0762548446655273,
      "step": 897
    },
    {
      "epoch": 0.20525714285714286,
      "grad_norm": 1.5130460262298584,
      "learning_rate": 0.0005,
      "loss": 2.947910785675049,
      "step": 898
    },
    {
      "epoch": 0.2054857142857143,
      "grad_norm": 1.2429906129837036,
      "learning_rate": 0.0005,
      "loss": 3.4610280990600586,
      "step": 899
    },
    {
      "epoch": 0.2057142857142857,
      "grad_norm": 1.3201048374176025,
      "learning_rate": 0.0005,
      "loss": 3.542475461959839,
      "step": 900
    },
    {
      "epoch": 0.20594285714285715,
      "grad_norm": 1.2367074489593506,
      "learning_rate": 0.0005,
      "loss": 3.3496618270874023,
      "step": 901
    },
    {
      "epoch": 0.20617142857142856,
      "grad_norm": 1.300564169883728,
      "learning_rate": 0.0005,
      "loss": 3.5469589233398438,
      "step": 902
    },
    {
      "epoch": 0.2064,
      "grad_norm": 1.3340543508529663,
      "learning_rate": 0.0005,
      "loss": 3.3764352798461914,
      "step": 903
    },
    {
      "epoch": 0.20662857142857144,
      "grad_norm": 1.6681269407272339,
      "learning_rate": 0.0005,
      "loss": 3.323925733566284,
      "step": 904
    },
    {
      "epoch": 0.20685714285714285,
      "grad_norm": 1.3115147352218628,
      "learning_rate": 0.0005,
      "loss": 3.681365728378296,
      "step": 905
    },
    {
      "epoch": 0.2070857142857143,
      "grad_norm": 1.3942707777023315,
      "learning_rate": 0.0005,
      "loss": 3.251279354095459,
      "step": 906
    },
    {
      "epoch": 0.2073142857142857,
      "grad_norm": 1.2580039501190186,
      "learning_rate": 0.0005,
      "loss": 3.170325517654419,
      "step": 907
    },
    {
      "epoch": 0.20754285714285714,
      "grad_norm": 4.356743812561035,
      "learning_rate": 0.0005,
      "loss": 3.841113328933716,
      "step": 908
    },
    {
      "epoch": 0.20777142857142858,
      "grad_norm": 1.6080522537231445,
      "learning_rate": 0.0005,
      "loss": 3.232780694961548,
      "step": 909
    },
    {
      "epoch": 0.208,
      "grad_norm": 1.5155048370361328,
      "learning_rate": 0.0005,
      "loss": 3.335545539855957,
      "step": 910
    },
    {
      "epoch": 0.20822857142857143,
      "grad_norm": 1.0209485292434692,
      "learning_rate": 0.0005,
      "loss": 3.4842987060546875,
      "step": 911
    },
    {
      "epoch": 0.20845714285714287,
      "grad_norm": 1.1282379627227783,
      "learning_rate": 0.0005,
      "loss": 3.632443428039551,
      "step": 912
    },
    {
      "epoch": 0.20868571428571428,
      "grad_norm": 1.3485491275787354,
      "learning_rate": 0.0005,
      "loss": 3.079315185546875,
      "step": 913
    },
    {
      "epoch": 0.20891428571428572,
      "grad_norm": 1.4230161905288696,
      "learning_rate": 0.0005,
      "loss": 3.1175663471221924,
      "step": 914
    },
    {
      "epoch": 0.20914285714285713,
      "grad_norm": 1.4861295223236084,
      "learning_rate": 0.0005,
      "loss": 3.468182325363159,
      "step": 915
    },
    {
      "epoch": 0.20937142857142857,
      "grad_norm": 1.9710485935211182,
      "learning_rate": 0.0005,
      "loss": 3.702723979949951,
      "step": 916
    },
    {
      "epoch": 0.2096,
      "grad_norm": 1.2862919569015503,
      "learning_rate": 0.0005,
      "loss": 3.49647855758667,
      "step": 917
    },
    {
      "epoch": 0.20982857142857142,
      "grad_norm": 1.3741791248321533,
      "learning_rate": 0.0005,
      "loss": 3.497265338897705,
      "step": 918
    },
    {
      "epoch": 0.21005714285714286,
      "grad_norm": 1.1071445941925049,
      "learning_rate": 0.0005,
      "loss": 3.1454930305480957,
      "step": 919
    },
    {
      "epoch": 0.2102857142857143,
      "grad_norm": 1.8426306247711182,
      "learning_rate": 0.0005,
      "loss": 3.150722026824951,
      "step": 920
    },
    {
      "epoch": 0.2105142857142857,
      "grad_norm": 1.192992091178894,
      "learning_rate": 0.0005,
      "loss": 3.1612026691436768,
      "step": 921
    },
    {
      "epoch": 0.21074285714285715,
      "grad_norm": 1.3272584676742554,
      "learning_rate": 0.0005,
      "loss": 3.4927210807800293,
      "step": 922
    },
    {
      "epoch": 0.21097142857142856,
      "grad_norm": 1.3133400678634644,
      "learning_rate": 0.0005,
      "loss": 3.29447340965271,
      "step": 923
    },
    {
      "epoch": 0.2112,
      "grad_norm": 1.2461599111557007,
      "learning_rate": 0.0005,
      "loss": 3.3266329765319824,
      "step": 924
    },
    {
      "epoch": 0.21142857142857144,
      "grad_norm": 1.1262140274047852,
      "learning_rate": 0.0005,
      "loss": 3.2122530937194824,
      "step": 925
    },
    {
      "epoch": 0.21165714285714285,
      "grad_norm": 1.287468671798706,
      "learning_rate": 0.0005,
      "loss": 3.19608211517334,
      "step": 926
    },
    {
      "epoch": 0.2118857142857143,
      "grad_norm": 1.530120849609375,
      "learning_rate": 0.0005,
      "loss": 3.570788621902466,
      "step": 927
    },
    {
      "epoch": 0.21211428571428573,
      "grad_norm": 1.2718627452850342,
      "learning_rate": 0.0005,
      "loss": 3.0786972045898438,
      "step": 928
    },
    {
      "epoch": 0.21234285714285714,
      "grad_norm": 1.053559422492981,
      "learning_rate": 0.0005,
      "loss": 3.536553382873535,
      "step": 929
    },
    {
      "epoch": 0.21257142857142858,
      "grad_norm": 1.2699836492538452,
      "learning_rate": 0.0005,
      "loss": 3.476599931716919,
      "step": 930
    },
    {
      "epoch": 0.2128,
      "grad_norm": 1.750025749206543,
      "learning_rate": 0.0005,
      "loss": 3.294356346130371,
      "step": 931
    },
    {
      "epoch": 0.21302857142857143,
      "grad_norm": 1.3592047691345215,
      "learning_rate": 0.0005,
      "loss": 3.4708895683288574,
      "step": 932
    },
    {
      "epoch": 0.21325714285714287,
      "grad_norm": 1.2436221837997437,
      "learning_rate": 0.0005,
      "loss": 3.4358582496643066,
      "step": 933
    },
    {
      "epoch": 0.21348571428571428,
      "grad_norm": 1.278235673904419,
      "learning_rate": 0.0005,
      "loss": 3.424565315246582,
      "step": 934
    },
    {
      "epoch": 0.21371428571428572,
      "grad_norm": 1.3145415782928467,
      "learning_rate": 0.0005,
      "loss": 3.157191753387451,
      "step": 935
    },
    {
      "epoch": 0.21394285714285713,
      "grad_norm": 1.7816200256347656,
      "learning_rate": 0.0005,
      "loss": 3.3237833976745605,
      "step": 936
    },
    {
      "epoch": 0.21417142857142857,
      "grad_norm": 1.7594776153564453,
      "learning_rate": 0.0005,
      "loss": 3.1960244178771973,
      "step": 937
    },
    {
      "epoch": 0.2144,
      "grad_norm": 1.1474956274032593,
      "learning_rate": 0.0005,
      "loss": 3.242429733276367,
      "step": 938
    },
    {
      "epoch": 0.21462857142857142,
      "grad_norm": 1.1481449604034424,
      "learning_rate": 0.0005,
      "loss": 3.696629047393799,
      "step": 939
    },
    {
      "epoch": 0.21485714285714286,
      "grad_norm": 1.3046914339065552,
      "learning_rate": 0.0005,
      "loss": 3.5355124473571777,
      "step": 940
    },
    {
      "epoch": 0.2150857142857143,
      "grad_norm": 1.4470711946487427,
      "learning_rate": 0.0005,
      "loss": 3.1986422538757324,
      "step": 941
    },
    {
      "epoch": 0.2153142857142857,
      "grad_norm": 1.141608715057373,
      "learning_rate": 0.0005,
      "loss": 3.4446747303009033,
      "step": 942
    },
    {
      "epoch": 0.21554285714285715,
      "grad_norm": 1.219967007637024,
      "learning_rate": 0.0005,
      "loss": 3.1929678916931152,
      "step": 943
    },
    {
      "epoch": 0.21577142857142856,
      "grad_norm": 1.332938313484192,
      "learning_rate": 0.0005,
      "loss": 3.293193817138672,
      "step": 944
    },
    {
      "epoch": 0.216,
      "grad_norm": 1.4388915300369263,
      "learning_rate": 0.0005,
      "loss": 3.387953281402588,
      "step": 945
    },
    {
      "epoch": 0.21622857142857144,
      "grad_norm": 1.290223479270935,
      "learning_rate": 0.0005,
      "loss": 3.4952850341796875,
      "step": 946
    },
    {
      "epoch": 0.21645714285714285,
      "grad_norm": 1.3020589351654053,
      "learning_rate": 0.0005,
      "loss": 3.2684073448181152,
      "step": 947
    },
    {
      "epoch": 0.2166857142857143,
      "grad_norm": 1.2797067165374756,
      "learning_rate": 0.0005,
      "loss": 3.4004054069519043,
      "step": 948
    },
    {
      "epoch": 0.21691428571428573,
      "grad_norm": 1.327130675315857,
      "learning_rate": 0.0005,
      "loss": 3.465162754058838,
      "step": 949
    },
    {
      "epoch": 0.21714285714285714,
      "grad_norm": 1.1161973476409912,
      "learning_rate": 0.0005,
      "loss": 3.2881641387939453,
      "step": 950
    },
    {
      "epoch": 0.21737142857142858,
      "grad_norm": 1.4310871362686157,
      "learning_rate": 0.0005,
      "loss": 3.1153860092163086,
      "step": 951
    },
    {
      "epoch": 0.2176,
      "grad_norm": 2.7420144081115723,
      "learning_rate": 0.0005,
      "loss": 3.3858795166015625,
      "step": 952
    },
    {
      "epoch": 0.21782857142857143,
      "grad_norm": 1.432816982269287,
      "learning_rate": 0.0005,
      "loss": 3.468825578689575,
      "step": 953
    },
    {
      "epoch": 0.21805714285714287,
      "grad_norm": 1.1653063297271729,
      "learning_rate": 0.0005,
      "loss": 3.4850218296051025,
      "step": 954
    },
    {
      "epoch": 0.21828571428571428,
      "grad_norm": 1.2451757192611694,
      "learning_rate": 0.0005,
      "loss": 3.4840378761291504,
      "step": 955
    },
    {
      "epoch": 0.21851428571428572,
      "grad_norm": 1.0092865228652954,
      "learning_rate": 0.0005,
      "loss": 3.4823718070983887,
      "step": 956
    },
    {
      "epoch": 0.21874285714285716,
      "grad_norm": 1.792476773262024,
      "learning_rate": 0.0005,
      "loss": 3.44735050201416,
      "step": 957
    },
    {
      "epoch": 0.21897142857142857,
      "grad_norm": 1.57414972782135,
      "learning_rate": 0.0005,
      "loss": 3.461575984954834,
      "step": 958
    },
    {
      "epoch": 0.2192,
      "grad_norm": 1.0966614484786987,
      "learning_rate": 0.0005,
      "loss": 3.2901477813720703,
      "step": 959
    },
    {
      "epoch": 0.21942857142857142,
      "grad_norm": 1.8234330415725708,
      "learning_rate": 0.0005,
      "loss": 3.2414746284484863,
      "step": 960
    },
    {
      "epoch": 0.21965714285714286,
      "grad_norm": 1.2871724367141724,
      "learning_rate": 0.0005,
      "loss": 3.537618637084961,
      "step": 961
    },
    {
      "epoch": 0.2198857142857143,
      "grad_norm": 1.283299446105957,
      "learning_rate": 0.0005,
      "loss": 3.309619188308716,
      "step": 962
    },
    {
      "epoch": 0.2201142857142857,
      "grad_norm": 2.0795021057128906,
      "learning_rate": 0.0005,
      "loss": 3.481027841567993,
      "step": 963
    },
    {
      "epoch": 0.22034285714285715,
      "grad_norm": 1.506671667098999,
      "learning_rate": 0.0005,
      "loss": 3.3130710124969482,
      "step": 964
    },
    {
      "epoch": 0.22057142857142858,
      "grad_norm": 1.428529977798462,
      "learning_rate": 0.0005,
      "loss": 3.31412935256958,
      "step": 965
    },
    {
      "epoch": 0.2208,
      "grad_norm": 1.1831045150756836,
      "learning_rate": 0.0005,
      "loss": 3.4167375564575195,
      "step": 966
    },
    {
      "epoch": 0.22102857142857144,
      "grad_norm": 1.6533091068267822,
      "learning_rate": 0.0005,
      "loss": 3.3340415954589844,
      "step": 967
    },
    {
      "epoch": 0.22125714285714285,
      "grad_norm": 1.2294726371765137,
      "learning_rate": 0.0005,
      "loss": 3.2852418422698975,
      "step": 968
    },
    {
      "epoch": 0.22148571428571429,
      "grad_norm": 1.4905993938446045,
      "learning_rate": 0.0005,
      "loss": 3.0342960357666016,
      "step": 969
    },
    {
      "epoch": 0.22171428571428572,
      "grad_norm": 1.2359023094177246,
      "learning_rate": 0.0005,
      "loss": 3.64931058883667,
      "step": 970
    },
    {
      "epoch": 0.22194285714285714,
      "grad_norm": 1.4178574085235596,
      "learning_rate": 0.0005,
      "loss": 3.3429956436157227,
      "step": 971
    },
    {
      "epoch": 0.22217142857142858,
      "grad_norm": 1.4975005388259888,
      "learning_rate": 0.0005,
      "loss": 3.1111316680908203,
      "step": 972
    },
    {
      "epoch": 0.2224,
      "grad_norm": 1.4405171871185303,
      "learning_rate": 0.0005,
      "loss": 3.7677626609802246,
      "step": 973
    },
    {
      "epoch": 0.22262857142857143,
      "grad_norm": 1.1006829738616943,
      "learning_rate": 0.0005,
      "loss": 3.2604496479034424,
      "step": 974
    },
    {
      "epoch": 0.22285714285714286,
      "grad_norm": 1.4817878007888794,
      "learning_rate": 0.0005,
      "loss": 3.141881227493286,
      "step": 975
    },
    {
      "epoch": 0.22308571428571428,
      "grad_norm": 1.0974680185317993,
      "learning_rate": 0.0005,
      "loss": 3.190951347351074,
      "step": 976
    },
    {
      "epoch": 0.22331428571428572,
      "grad_norm": 1.0238566398620605,
      "learning_rate": 0.0005,
      "loss": 3.109053134918213,
      "step": 977
    },
    {
      "epoch": 0.22354285714285715,
      "grad_norm": 1.3014897108078003,
      "learning_rate": 0.0005,
      "loss": 3.3220434188842773,
      "step": 978
    },
    {
      "epoch": 0.22377142857142857,
      "grad_norm": 1.2971378564834595,
      "learning_rate": 0.0005,
      "loss": 3.761173725128174,
      "step": 979
    },
    {
      "epoch": 0.224,
      "grad_norm": 1.3167688846588135,
      "learning_rate": 0.0005,
      "loss": 3.2525930404663086,
      "step": 980
    },
    {
      "epoch": 0.22422857142857142,
      "grad_norm": 1.292825698852539,
      "learning_rate": 0.0005,
      "loss": 3.309990882873535,
      "step": 981
    },
    {
      "epoch": 0.22445714285714286,
      "grad_norm": 1.1253159046173096,
      "learning_rate": 0.0005,
      "loss": 3.5896050930023193,
      "step": 982
    },
    {
      "epoch": 0.2246857142857143,
      "grad_norm": 1.0891181230545044,
      "learning_rate": 0.0005,
      "loss": 3.4392266273498535,
      "step": 983
    },
    {
      "epoch": 0.2249142857142857,
      "grad_norm": 1.40304434299469,
      "learning_rate": 0.0005,
      "loss": 3.541728973388672,
      "step": 984
    },
    {
      "epoch": 0.22514285714285714,
      "grad_norm": 1.089601993560791,
      "learning_rate": 0.0005,
      "loss": 3.244215965270996,
      "step": 985
    },
    {
      "epoch": 0.22537142857142858,
      "grad_norm": 1.2911158800125122,
      "learning_rate": 0.0005,
      "loss": 3.3681163787841797,
      "step": 986
    },
    {
      "epoch": 0.2256,
      "grad_norm": 1.237251877784729,
      "learning_rate": 0.0005,
      "loss": 3.1701040267944336,
      "step": 987
    },
    {
      "epoch": 0.22582857142857143,
      "grad_norm": 1.1574287414550781,
      "learning_rate": 0.0005,
      "loss": 3.2123141288757324,
      "step": 988
    },
    {
      "epoch": 0.22605714285714285,
      "grad_norm": 1.4377480745315552,
      "learning_rate": 0.0005,
      "loss": 3.039074420928955,
      "step": 989
    },
    {
      "epoch": 0.22628571428571428,
      "grad_norm": 1.942594289779663,
      "learning_rate": 0.0005,
      "loss": 3.9310789108276367,
      "step": 990
    },
    {
      "epoch": 0.22651428571428572,
      "grad_norm": 1.1699734926223755,
      "learning_rate": 0.0005,
      "loss": 3.241823196411133,
      "step": 991
    },
    {
      "epoch": 0.22674285714285713,
      "grad_norm": 1.5157334804534912,
      "learning_rate": 0.0005,
      "loss": 3.470942735671997,
      "step": 992
    },
    {
      "epoch": 0.22697142857142857,
      "grad_norm": 1.1677191257476807,
      "learning_rate": 0.0005,
      "loss": 3.1240317821502686,
      "step": 993
    },
    {
      "epoch": 0.2272,
      "grad_norm": 1.1696906089782715,
      "learning_rate": 0.0005,
      "loss": 3.27163028717041,
      "step": 994
    },
    {
      "epoch": 0.22742857142857142,
      "grad_norm": 1.2639247179031372,
      "learning_rate": 0.0005,
      "loss": 3.429900884628296,
      "step": 995
    },
    {
      "epoch": 0.22765714285714286,
      "grad_norm": 1.1304646730422974,
      "learning_rate": 0.0005,
      "loss": 3.355325222015381,
      "step": 996
    },
    {
      "epoch": 0.22788571428571427,
      "grad_norm": 1.0709110498428345,
      "learning_rate": 0.0005,
      "loss": 3.1499080657958984,
      "step": 997
    },
    {
      "epoch": 0.22811428571428571,
      "grad_norm": 1.3820265531539917,
      "learning_rate": 0.0005,
      "loss": 3.3958117961883545,
      "step": 998
    },
    {
      "epoch": 0.22834285714285715,
      "grad_norm": 1.398696780204773,
      "learning_rate": 0.0005,
      "loss": 3.2592344284057617,
      "step": 999
    },
    {
      "epoch": 0.22857142857142856,
      "grad_norm": 1.366585612297058,
      "learning_rate": 0.0005,
      "loss": 3.369699716567993,
      "step": 1000
    },
    {
      "epoch": 0.2288,
      "grad_norm": 1.0245234966278076,
      "learning_rate": 0.0005,
      "loss": 3.029247522354126,
      "step": 1001
    },
    {
      "epoch": 0.22902857142857141,
      "grad_norm": 1.1805179119110107,
      "learning_rate": 0.0005,
      "loss": 3.4210612773895264,
      "step": 1002
    },
    {
      "epoch": 0.22925714285714285,
      "grad_norm": 1.2720112800598145,
      "learning_rate": 0.0005,
      "loss": 3.2773990631103516,
      "step": 1003
    },
    {
      "epoch": 0.2294857142857143,
      "grad_norm": 1.3282208442687988,
      "learning_rate": 0.0005,
      "loss": 3.689124345779419,
      "step": 1004
    },
    {
      "epoch": 0.2297142857142857,
      "grad_norm": 1.5877629518508911,
      "learning_rate": 0.0005,
      "loss": 3.5925371646881104,
      "step": 1005
    },
    {
      "epoch": 0.22994285714285714,
      "grad_norm": 1.4865071773529053,
      "learning_rate": 0.0005,
      "loss": 3.174323558807373,
      "step": 1006
    },
    {
      "epoch": 0.23017142857142858,
      "grad_norm": 1.0998824834823608,
      "learning_rate": 0.0005,
      "loss": 3.388476848602295,
      "step": 1007
    },
    {
      "epoch": 0.2304,
      "grad_norm": 1.4233752489089966,
      "learning_rate": 0.0005,
      "loss": 3.513059616088867,
      "step": 1008
    },
    {
      "epoch": 0.23062857142857143,
      "grad_norm": 1.356092929840088,
      "learning_rate": 0.0005,
      "loss": 3.067478656768799,
      "step": 1009
    },
    {
      "epoch": 0.23085714285714284,
      "grad_norm": 1.5959640741348267,
      "learning_rate": 0.0005,
      "loss": 3.2831077575683594,
      "step": 1010
    },
    {
      "epoch": 0.23108571428571428,
      "grad_norm": 1.1009615659713745,
      "learning_rate": 0.0005,
      "loss": 2.9214532375335693,
      "step": 1011
    },
    {
      "epoch": 0.23131428571428572,
      "grad_norm": 1.2052000761032104,
      "learning_rate": 0.0005,
      "loss": 3.6531081199645996,
      "step": 1012
    },
    {
      "epoch": 0.23154285714285713,
      "grad_norm": 1.093749761581421,
      "learning_rate": 0.0005,
      "loss": 3.355480194091797,
      "step": 1013
    },
    {
      "epoch": 0.23177142857142857,
      "grad_norm": 1.144089698791504,
      "learning_rate": 0.0005,
      "loss": 3.377493381500244,
      "step": 1014
    },
    {
      "epoch": 0.232,
      "grad_norm": 1.3334516286849976,
      "learning_rate": 0.0005,
      "loss": 3.0026450157165527,
      "step": 1015
    },
    {
      "epoch": 0.23222857142857142,
      "grad_norm": 1.2167794704437256,
      "learning_rate": 0.0005,
      "loss": 3.2708773612976074,
      "step": 1016
    },
    {
      "epoch": 0.23245714285714286,
      "grad_norm": 1.3945969343185425,
      "learning_rate": 0.0005,
      "loss": 3.4840574264526367,
      "step": 1017
    },
    {
      "epoch": 0.23268571428571427,
      "grad_norm": 1.219571828842163,
      "learning_rate": 0.0005,
      "loss": 3.4644999504089355,
      "step": 1018
    },
    {
      "epoch": 0.2329142857142857,
      "grad_norm": 1.2769314050674438,
      "learning_rate": 0.0005,
      "loss": 3.1751465797424316,
      "step": 1019
    },
    {
      "epoch": 0.23314285714285715,
      "grad_norm": 1.1252217292785645,
      "learning_rate": 0.0005,
      "loss": 3.045907497406006,
      "step": 1020
    },
    {
      "epoch": 0.23337142857142856,
      "grad_norm": 1.175329327583313,
      "learning_rate": 0.0005,
      "loss": 3.279174327850342,
      "step": 1021
    },
    {
      "epoch": 0.2336,
      "grad_norm": 1.5474909543991089,
      "learning_rate": 0.0005,
      "loss": 3.399862289428711,
      "step": 1022
    },
    {
      "epoch": 0.23382857142857144,
      "grad_norm": 1.2887436151504517,
      "learning_rate": 0.0005,
      "loss": 3.4017765522003174,
      "step": 1023
    },
    {
      "epoch": 0.23405714285714285,
      "grad_norm": 1.3629381656646729,
      "learning_rate": 0.0005,
      "loss": 3.088256597518921,
      "step": 1024
    },
    {
      "epoch": 0.2342857142857143,
      "grad_norm": 1.078724980354309,
      "learning_rate": 0.0005,
      "loss": 3.2623143196105957,
      "step": 1025
    },
    {
      "epoch": 0.2345142857142857,
      "grad_norm": 1.2772469520568848,
      "learning_rate": 0.0005,
      "loss": 3.1371090412139893,
      "step": 1026
    },
    {
      "epoch": 0.23474285714285714,
      "grad_norm": 1.370227575302124,
      "learning_rate": 0.0005,
      "loss": 3.12623929977417,
      "step": 1027
    },
    {
      "epoch": 0.23497142857142858,
      "grad_norm": 1.0460622310638428,
      "learning_rate": 0.0005,
      "loss": 3.317434310913086,
      "step": 1028
    },
    {
      "epoch": 0.2352,
      "grad_norm": 1.260230541229248,
      "learning_rate": 0.0005,
      "loss": 3.2756106853485107,
      "step": 1029
    },
    {
      "epoch": 0.23542857142857143,
      "grad_norm": 1.7227885723114014,
      "learning_rate": 0.0005,
      "loss": 3.327519416809082,
      "step": 1030
    },
    {
      "epoch": 0.23565714285714287,
      "grad_norm": 1.4532601833343506,
      "learning_rate": 0.0005,
      "loss": 3.2281863689422607,
      "step": 1031
    },
    {
      "epoch": 0.23588571428571428,
      "grad_norm": 1.6560367345809937,
      "learning_rate": 0.0005,
      "loss": 3.667605400085449,
      "step": 1032
    },
    {
      "epoch": 0.23611428571428572,
      "grad_norm": 1.1242958307266235,
      "learning_rate": 0.0005,
      "loss": 3.1918418407440186,
      "step": 1033
    },
    {
      "epoch": 0.23634285714285713,
      "grad_norm": 1.4235745668411255,
      "learning_rate": 0.0005,
      "loss": 3.270228147506714,
      "step": 1034
    },
    {
      "epoch": 0.23657142857142857,
      "grad_norm": 1.188766598701477,
      "learning_rate": 0.0005,
      "loss": 3.1743602752685547,
      "step": 1035
    },
    {
      "epoch": 0.2368,
      "grad_norm": 1.327566146850586,
      "learning_rate": 0.0005,
      "loss": 3.5720574855804443,
      "step": 1036
    },
    {
      "epoch": 0.23702857142857142,
      "grad_norm": 1.2445787191390991,
      "learning_rate": 0.0005,
      "loss": 3.1955833435058594,
      "step": 1037
    },
    {
      "epoch": 0.23725714285714286,
      "grad_norm": 1.1208595037460327,
      "learning_rate": 0.0005,
      "loss": 3.494938373565674,
      "step": 1038
    },
    {
      "epoch": 0.23748571428571427,
      "grad_norm": 1.5186210870742798,
      "learning_rate": 0.0005,
      "loss": 3.4812159538269043,
      "step": 1039
    },
    {
      "epoch": 0.2377142857142857,
      "grad_norm": 1.236395239830017,
      "learning_rate": 0.0005,
      "loss": 3.003695487976074,
      "step": 1040
    },
    {
      "epoch": 0.23794285714285715,
      "grad_norm": 1.1086400747299194,
      "learning_rate": 0.0005,
      "loss": 3.159146785736084,
      "step": 1041
    },
    {
      "epoch": 0.23817142857142856,
      "grad_norm": 1.0478602647781372,
      "learning_rate": 0.0005,
      "loss": 3.1260671615600586,
      "step": 1042
    },
    {
      "epoch": 0.2384,
      "grad_norm": 1.0767940282821655,
      "learning_rate": 0.0005,
      "loss": 2.9085021018981934,
      "step": 1043
    },
    {
      "epoch": 0.23862857142857144,
      "grad_norm": 1.508844256401062,
      "learning_rate": 0.0005,
      "loss": 2.923097610473633,
      "step": 1044
    },
    {
      "epoch": 0.23885714285714285,
      "grad_norm": 1.1138564348220825,
      "learning_rate": 0.0005,
      "loss": 3.4822139739990234,
      "step": 1045
    },
    {
      "epoch": 0.2390857142857143,
      "grad_norm": 1.2618496417999268,
      "learning_rate": 0.0005,
      "loss": 3.6111249923706055,
      "step": 1046
    },
    {
      "epoch": 0.2393142857142857,
      "grad_norm": 1.6572974920272827,
      "learning_rate": 0.0005,
      "loss": 3.228760004043579,
      "step": 1047
    },
    {
      "epoch": 0.23954285714285714,
      "grad_norm": 1.5468380451202393,
      "learning_rate": 0.0005,
      "loss": 3.187166690826416,
      "step": 1048
    },
    {
      "epoch": 0.23977142857142858,
      "grad_norm": 1.063461422920227,
      "learning_rate": 0.0005,
      "loss": 3.6370272636413574,
      "step": 1049
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.4060920476913452,
      "learning_rate": 0.0005,
      "loss": 2.8538870811462402,
      "step": 1050
    },
    {
      "epoch": 0.24022857142857143,
      "grad_norm": 1.3102666139602661,
      "learning_rate": 0.0005,
      "loss": 3.122936487197876,
      "step": 1051
    },
    {
      "epoch": 0.24045714285714287,
      "grad_norm": 1.112098217010498,
      "learning_rate": 0.0005,
      "loss": 3.4682259559631348,
      "step": 1052
    },
    {
      "epoch": 0.24068571428571428,
      "grad_norm": 1.0529614686965942,
      "learning_rate": 0.0005,
      "loss": 3.0480763912200928,
      "step": 1053
    },
    {
      "epoch": 0.24091428571428572,
      "grad_norm": 1.1301331520080566,
      "learning_rate": 0.0005,
      "loss": 3.6639583110809326,
      "step": 1054
    },
    {
      "epoch": 0.24114285714285713,
      "grad_norm": 1.2910820245742798,
      "learning_rate": 0.0005,
      "loss": 3.1352920532226562,
      "step": 1055
    },
    {
      "epoch": 0.24137142857142857,
      "grad_norm": 0.999413251876831,
      "learning_rate": 0.0005,
      "loss": 3.1765589714050293,
      "step": 1056
    },
    {
      "epoch": 0.2416,
      "grad_norm": 1.4716072082519531,
      "learning_rate": 0.0005,
      "loss": 3.34980845451355,
      "step": 1057
    },
    {
      "epoch": 0.24182857142857142,
      "grad_norm": 1.1370370388031006,
      "learning_rate": 0.0005,
      "loss": 3.511193037033081,
      "step": 1058
    },
    {
      "epoch": 0.24205714285714286,
      "grad_norm": 1.3039287328720093,
      "learning_rate": 0.0005,
      "loss": 3.3486995697021484,
      "step": 1059
    },
    {
      "epoch": 0.2422857142857143,
      "grad_norm": 1.2914133071899414,
      "learning_rate": 0.0005,
      "loss": 3.2442080974578857,
      "step": 1060
    },
    {
      "epoch": 0.2425142857142857,
      "grad_norm": 1.1517596244812012,
      "learning_rate": 0.0005,
      "loss": 3.100588321685791,
      "step": 1061
    },
    {
      "epoch": 0.24274285714285715,
      "grad_norm": 1.4714115858078003,
      "learning_rate": 0.0005,
      "loss": 3.386016607284546,
      "step": 1062
    },
    {
      "epoch": 0.24297142857142856,
      "grad_norm": 3.056370258331299,
      "learning_rate": 0.0005,
      "loss": 4.007850646972656,
      "step": 1063
    },
    {
      "epoch": 0.2432,
      "grad_norm": 1.06111478805542,
      "learning_rate": 0.0005,
      "loss": 3.3227665424346924,
      "step": 1064
    },
    {
      "epoch": 0.24342857142857144,
      "grad_norm": 1.2237426042556763,
      "learning_rate": 0.0005,
      "loss": 3.544290542602539,
      "step": 1065
    },
    {
      "epoch": 0.24365714285714285,
      "grad_norm": 1.0285568237304688,
      "learning_rate": 0.0005,
      "loss": 3.7509117126464844,
      "step": 1066
    },
    {
      "epoch": 0.2438857142857143,
      "grad_norm": 1.1693593263626099,
      "learning_rate": 0.0005,
      "loss": 3.169999122619629,
      "step": 1067
    },
    {
      "epoch": 0.24411428571428573,
      "grad_norm": 1.1824324131011963,
      "learning_rate": 0.0005,
      "loss": 3.138110637664795,
      "step": 1068
    },
    {
      "epoch": 0.24434285714285714,
      "grad_norm": 1.4366098642349243,
      "learning_rate": 0.0005,
      "loss": 3.209362030029297,
      "step": 1069
    },
    {
      "epoch": 0.24457142857142858,
      "grad_norm": 1.3557711839675903,
      "learning_rate": 0.0005,
      "loss": 2.8193423748016357,
      "step": 1070
    },
    {
      "epoch": 0.2448,
      "grad_norm": 1.4940201044082642,
      "learning_rate": 0.0005,
      "loss": 3.141659736633301,
      "step": 1071
    },
    {
      "epoch": 0.24502857142857143,
      "grad_norm": 1.298449993133545,
      "learning_rate": 0.0005,
      "loss": 3.301161050796509,
      "step": 1072
    },
    {
      "epoch": 0.24525714285714287,
      "grad_norm": 1.080451250076294,
      "learning_rate": 0.0005,
      "loss": 3.1611766815185547,
      "step": 1073
    },
    {
      "epoch": 0.24548571428571428,
      "grad_norm": 1.4809359312057495,
      "learning_rate": 0.0005,
      "loss": 3.1986567974090576,
      "step": 1074
    },
    {
      "epoch": 0.24571428571428572,
      "grad_norm": 1.218169927597046,
      "learning_rate": 0.0005,
      "loss": 3.031515121459961,
      "step": 1075
    },
    {
      "epoch": 0.24594285714285713,
      "grad_norm": 1.1946299076080322,
      "learning_rate": 0.0005,
      "loss": 3.136106491088867,
      "step": 1076
    },
    {
      "epoch": 0.24617142857142857,
      "grad_norm": 1.5050660371780396,
      "learning_rate": 0.0005,
      "loss": 3.441460609436035,
      "step": 1077
    },
    {
      "epoch": 0.2464,
      "grad_norm": 1.283355951309204,
      "learning_rate": 0.0005,
      "loss": 3.1331968307495117,
      "step": 1078
    },
    {
      "epoch": 0.24662857142857142,
      "grad_norm": 1.2644990682601929,
      "learning_rate": 0.0005,
      "loss": 3.183519124984741,
      "step": 1079
    },
    {
      "epoch": 0.24685714285714286,
      "grad_norm": 1.8514738082885742,
      "learning_rate": 0.0005,
      "loss": 3.0937118530273438,
      "step": 1080
    },
    {
      "epoch": 0.2470857142857143,
      "grad_norm": 1.2947914600372314,
      "learning_rate": 0.0005,
      "loss": 2.9615559577941895,
      "step": 1081
    },
    {
      "epoch": 0.2473142857142857,
      "grad_norm": 1.3044830560684204,
      "learning_rate": 0.0005,
      "loss": 3.2169041633605957,
      "step": 1082
    },
    {
      "epoch": 0.24754285714285715,
      "grad_norm": 1.049437165260315,
      "learning_rate": 0.0005,
      "loss": 3.1802525520324707,
      "step": 1083
    },
    {
      "epoch": 0.24777142857142856,
      "grad_norm": 1.6323089599609375,
      "learning_rate": 0.0005,
      "loss": 3.3326261043548584,
      "step": 1084
    },
    {
      "epoch": 0.248,
      "grad_norm": 1.482959270477295,
      "learning_rate": 0.0005,
      "loss": 3.1812081336975098,
      "step": 1085
    },
    {
      "epoch": 0.24822857142857144,
      "grad_norm": 1.5916515588760376,
      "learning_rate": 0.0005,
      "loss": 2.952970027923584,
      "step": 1086
    },
    {
      "epoch": 0.24845714285714285,
      "grad_norm": 1.402376413345337,
      "learning_rate": 0.0005,
      "loss": 3.154059886932373,
      "step": 1087
    },
    {
      "epoch": 0.2486857142857143,
      "grad_norm": 2.276156425476074,
      "learning_rate": 0.0005,
      "loss": 3.630647897720337,
      "step": 1088
    },
    {
      "epoch": 0.24891428571428573,
      "grad_norm": 1.116141676902771,
      "learning_rate": 0.0005,
      "loss": 3.217663288116455,
      "step": 1089
    },
    {
      "epoch": 0.24914285714285714,
      "grad_norm": 1.3568247556686401,
      "learning_rate": 0.0005,
      "loss": 3.201796531677246,
      "step": 1090
    },
    {
      "epoch": 0.24937142857142858,
      "grad_norm": 1.489665150642395,
      "learning_rate": 0.0005,
      "loss": 3.3749992847442627,
      "step": 1091
    },
    {
      "epoch": 0.2496,
      "grad_norm": 1.2992899417877197,
      "learning_rate": 0.0005,
      "loss": 3.197815418243408,
      "step": 1092
    },
    {
      "epoch": 0.24982857142857143,
      "grad_norm": 1.2536650896072388,
      "learning_rate": 0.0005,
      "loss": 2.944413661956787,
      "step": 1093
    },
    {
      "epoch": 0.25005714285714287,
      "grad_norm": 1.490009069442749,
      "learning_rate": 0.0005,
      "loss": 3.5163590908050537,
      "step": 1094
    },
    {
      "epoch": 0.2502857142857143,
      "grad_norm": 1.2291462421417236,
      "learning_rate": 0.0005,
      "loss": 3.3791630268096924,
      "step": 1095
    },
    {
      "epoch": 0.2505142857142857,
      "grad_norm": 1.2949362993240356,
      "learning_rate": 0.0005,
      "loss": 3.2930197715759277,
      "step": 1096
    },
    {
      "epoch": 0.25074285714285716,
      "grad_norm": 1.0617543458938599,
      "learning_rate": 0.0005,
      "loss": 3.3214781284332275,
      "step": 1097
    },
    {
      "epoch": 0.25097142857142857,
      "grad_norm": 2.0311288833618164,
      "learning_rate": 0.0005,
      "loss": 3.0656657218933105,
      "step": 1098
    },
    {
      "epoch": 0.2512,
      "grad_norm": 1.4291895627975464,
      "learning_rate": 0.0005,
      "loss": 3.439469814300537,
      "step": 1099
    },
    {
      "epoch": 0.25142857142857145,
      "grad_norm": 1.4490859508514404,
      "learning_rate": 0.0005,
      "loss": 3.5531740188598633,
      "step": 1100
    },
    {
      "epoch": 0.25165714285714286,
      "grad_norm": 1.137143611907959,
      "learning_rate": 0.0005,
      "loss": 3.135704517364502,
      "step": 1101
    },
    {
      "epoch": 0.25188571428571427,
      "grad_norm": 1.147591233253479,
      "learning_rate": 0.0005,
      "loss": 2.9652915000915527,
      "step": 1102
    },
    {
      "epoch": 0.25211428571428574,
      "grad_norm": 1.393416404724121,
      "learning_rate": 0.0005,
      "loss": 3.171722650527954,
      "step": 1103
    },
    {
      "epoch": 0.25234285714285715,
      "grad_norm": 1.4040108919143677,
      "learning_rate": 0.0005,
      "loss": 3.5368518829345703,
      "step": 1104
    },
    {
      "epoch": 0.25257142857142856,
      "grad_norm": 1.1544015407562256,
      "learning_rate": 0.0005,
      "loss": 3.3560948371887207,
      "step": 1105
    },
    {
      "epoch": 0.2528,
      "grad_norm": 1.052939534187317,
      "learning_rate": 0.0005,
      "loss": 2.918015718460083,
      "step": 1106
    },
    {
      "epoch": 0.25302857142857144,
      "grad_norm": 1.1486860513687134,
      "learning_rate": 0.0005,
      "loss": 3.0561976432800293,
      "step": 1107
    },
    {
      "epoch": 0.25325714285714285,
      "grad_norm": 2.568215847015381,
      "learning_rate": 0.0005,
      "loss": 3.589733123779297,
      "step": 1108
    },
    {
      "epoch": 0.25348571428571426,
      "grad_norm": 1.2699317932128906,
      "learning_rate": 0.0005,
      "loss": 3.6575610637664795,
      "step": 1109
    },
    {
      "epoch": 0.2537142857142857,
      "grad_norm": 2.448563575744629,
      "learning_rate": 0.0005,
      "loss": 3.06381893157959,
      "step": 1110
    },
    {
      "epoch": 0.25394285714285714,
      "grad_norm": 1.413048267364502,
      "learning_rate": 0.0005,
      "loss": 3.362950325012207,
      "step": 1111
    },
    {
      "epoch": 0.25417142857142855,
      "grad_norm": 2.7094597816467285,
      "learning_rate": 0.0005,
      "loss": 3.301680564880371,
      "step": 1112
    },
    {
      "epoch": 0.2544,
      "grad_norm": 1.8361951112747192,
      "learning_rate": 0.0005,
      "loss": 3.797544479370117,
      "step": 1113
    },
    {
      "epoch": 0.2546285714285714,
      "grad_norm": 1.0711824893951416,
      "learning_rate": 0.0005,
      "loss": 3.320502996444702,
      "step": 1114
    },
    {
      "epoch": 0.25485714285714284,
      "grad_norm": 1.1091464757919312,
      "learning_rate": 0.0005,
      "loss": 3.2134299278259277,
      "step": 1115
    },
    {
      "epoch": 0.2550857142857143,
      "grad_norm": 1.4578301906585693,
      "learning_rate": 0.0005,
      "loss": 3.4238271713256836,
      "step": 1116
    },
    {
      "epoch": 0.2553142857142857,
      "grad_norm": 1.127748727798462,
      "learning_rate": 0.0005,
      "loss": 3.3750998973846436,
      "step": 1117
    },
    {
      "epoch": 0.2555428571428571,
      "grad_norm": 1.2757480144500732,
      "learning_rate": 0.0005,
      "loss": 3.45703125,
      "step": 1118
    },
    {
      "epoch": 0.2557714285714286,
      "grad_norm": 1.1706792116165161,
      "learning_rate": 0.0005,
      "loss": 3.122662305831909,
      "step": 1119
    },
    {
      "epoch": 0.256,
      "grad_norm": 1.2667875289916992,
      "learning_rate": 0.0005,
      "loss": 3.059969425201416,
      "step": 1120
    },
    {
      "epoch": 0.2562285714285714,
      "grad_norm": 46.932003021240234,
      "learning_rate": 0.0005,
      "loss": 3.7209739685058594,
      "step": 1121
    },
    {
      "epoch": 0.2564571428571429,
      "grad_norm": 1.389408826828003,
      "learning_rate": 0.0005,
      "loss": 3.0162241458892822,
      "step": 1122
    },
    {
      "epoch": 0.2566857142857143,
      "grad_norm": 1.5388566255569458,
      "learning_rate": 0.0005,
      "loss": 3.0397801399230957,
      "step": 1123
    },
    {
      "epoch": 0.2569142857142857,
      "grad_norm": 1.1583408117294312,
      "learning_rate": 0.0005,
      "loss": 3.2985968589782715,
      "step": 1124
    },
    {
      "epoch": 0.2571428571428571,
      "grad_norm": 1.1795613765716553,
      "learning_rate": 0.0005,
      "loss": 3.126218318939209,
      "step": 1125
    },
    {
      "epoch": 0.2573714285714286,
      "grad_norm": 1.185238242149353,
      "learning_rate": 0.0005,
      "loss": 3.5075881481170654,
      "step": 1126
    },
    {
      "epoch": 0.2576,
      "grad_norm": 1.1616437435150146,
      "learning_rate": 0.0005,
      "loss": 2.8733596801757812,
      "step": 1127
    },
    {
      "epoch": 0.2578285714285714,
      "grad_norm": 1.0861878395080566,
      "learning_rate": 0.0005,
      "loss": 3.116790771484375,
      "step": 1128
    },
    {
      "epoch": 0.2580571428571429,
      "grad_norm": 1.2400082349777222,
      "learning_rate": 0.0005,
      "loss": 3.2111291885375977,
      "step": 1129
    },
    {
      "epoch": 0.2582857142857143,
      "grad_norm": 1.3069864511489868,
      "learning_rate": 0.0005,
      "loss": 3.0875091552734375,
      "step": 1130
    },
    {
      "epoch": 0.2585142857142857,
      "grad_norm": 1.6492979526519775,
      "learning_rate": 0.0005,
      "loss": 3.4064362049102783,
      "step": 1131
    },
    {
      "epoch": 0.25874285714285716,
      "grad_norm": 1.309396743774414,
      "learning_rate": 0.0005,
      "loss": 3.146908760070801,
      "step": 1132
    },
    {
      "epoch": 0.2589714285714286,
      "grad_norm": 1.2321662902832031,
      "learning_rate": 0.0005,
      "loss": 3.1116604804992676,
      "step": 1133
    },
    {
      "epoch": 0.2592,
      "grad_norm": 1.4250468015670776,
      "learning_rate": 0.0005,
      "loss": 3.3208460807800293,
      "step": 1134
    },
    {
      "epoch": 0.25942857142857145,
      "grad_norm": 1.484666347503662,
      "learning_rate": 0.0005,
      "loss": 3.1113550662994385,
      "step": 1135
    },
    {
      "epoch": 0.25965714285714286,
      "grad_norm": 1.2868825197219849,
      "learning_rate": 0.0005,
      "loss": 3.3327150344848633,
      "step": 1136
    },
    {
      "epoch": 0.2598857142857143,
      "grad_norm": 1.3545584678649902,
      "learning_rate": 0.0005,
      "loss": 3.0791573524475098,
      "step": 1137
    },
    {
      "epoch": 0.2601142857142857,
      "grad_norm": 1.3333932161331177,
      "learning_rate": 0.0005,
      "loss": 2.9505791664123535,
      "step": 1138
    },
    {
      "epoch": 0.26034285714285715,
      "grad_norm": 1.6257904767990112,
      "learning_rate": 0.0005,
      "loss": 3.3608531951904297,
      "step": 1139
    },
    {
      "epoch": 0.26057142857142856,
      "grad_norm": 1.3766181468963623,
      "learning_rate": 0.0005,
      "loss": 3.3588671684265137,
      "step": 1140
    },
    {
      "epoch": 0.2608,
      "grad_norm": 1.5528831481933594,
      "learning_rate": 0.0005,
      "loss": 3.3001809120178223,
      "step": 1141
    },
    {
      "epoch": 0.26102857142857144,
      "grad_norm": 1.6035572290420532,
      "learning_rate": 0.0005,
      "loss": 3.5849714279174805,
      "step": 1142
    },
    {
      "epoch": 0.26125714285714285,
      "grad_norm": 1.2930563688278198,
      "learning_rate": 0.0005,
      "loss": 3.646935224533081,
      "step": 1143
    },
    {
      "epoch": 0.26148571428571427,
      "grad_norm": 1.6084007024765015,
      "learning_rate": 0.0005,
      "loss": 3.2002809047698975,
      "step": 1144
    },
    {
      "epoch": 0.26171428571428573,
      "grad_norm": 1.532105803489685,
      "learning_rate": 0.0005,
      "loss": 3.5510597229003906,
      "step": 1145
    },
    {
      "epoch": 0.26194285714285714,
      "grad_norm": 1.3049505949020386,
      "learning_rate": 0.0005,
      "loss": 3.2263622283935547,
      "step": 1146
    },
    {
      "epoch": 0.26217142857142856,
      "grad_norm": 1.1577874422073364,
      "learning_rate": 0.0005,
      "loss": 3.40742826461792,
      "step": 1147
    },
    {
      "epoch": 0.2624,
      "grad_norm": 1.8214713335037231,
      "learning_rate": 0.0005,
      "loss": 3.031341314315796,
      "step": 1148
    },
    {
      "epoch": 0.26262857142857143,
      "grad_norm": 1.387927532196045,
      "learning_rate": 0.0005,
      "loss": 3.2747416496276855,
      "step": 1149
    },
    {
      "epoch": 0.26285714285714284,
      "grad_norm": 1.3153258562088013,
      "learning_rate": 0.0005,
      "loss": 3.649111747741699,
      "step": 1150
    },
    {
      "epoch": 0.2630857142857143,
      "grad_norm": 1.158482551574707,
      "learning_rate": 0.0005,
      "loss": 3.228522300720215,
      "step": 1151
    },
    {
      "epoch": 0.2633142857142857,
      "grad_norm": 1.1298010349273682,
      "learning_rate": 0.0005,
      "loss": 3.5168495178222656,
      "step": 1152
    },
    {
      "epoch": 0.26354285714285713,
      "grad_norm": 6.449540615081787,
      "learning_rate": 0.0005,
      "loss": 3.889963150024414,
      "step": 1153
    },
    {
      "epoch": 0.26377142857142855,
      "grad_norm": 3.610342502593994,
      "learning_rate": 0.0005,
      "loss": 3.140906810760498,
      "step": 1154
    },
    {
      "epoch": 0.264,
      "grad_norm": 1.6602258682250977,
      "learning_rate": 0.0005,
      "loss": 3.0831499099731445,
      "step": 1155
    },
    {
      "epoch": 0.2642285714285714,
      "grad_norm": 1.2967257499694824,
      "learning_rate": 0.0005,
      "loss": 2.9632599353790283,
      "step": 1156
    },
    {
      "epoch": 0.26445714285714284,
      "grad_norm": 1.1181166172027588,
      "learning_rate": 0.0005,
      "loss": 3.3146517276763916,
      "step": 1157
    },
    {
      "epoch": 0.2646857142857143,
      "grad_norm": 1.705923080444336,
      "learning_rate": 0.0005,
      "loss": 3.1967062950134277,
      "step": 1158
    },
    {
      "epoch": 0.2649142857142857,
      "grad_norm": 1.2964123487472534,
      "learning_rate": 0.0005,
      "loss": 3.460747718811035,
      "step": 1159
    },
    {
      "epoch": 0.2651428571428571,
      "grad_norm": 1.6059463024139404,
      "learning_rate": 0.0005,
      "loss": 3.210099697113037,
      "step": 1160
    },
    {
      "epoch": 0.2653714285714286,
      "grad_norm": 1.4205687046051025,
      "learning_rate": 0.0005,
      "loss": 3.0680291652679443,
      "step": 1161
    },
    {
      "epoch": 0.2656,
      "grad_norm": 1.525685429573059,
      "learning_rate": 0.0005,
      "loss": 3.1556663513183594,
      "step": 1162
    },
    {
      "epoch": 0.2658285714285714,
      "grad_norm": 1.1425701379776,
      "learning_rate": 0.0005,
      "loss": 3.0472970008850098,
      "step": 1163
    },
    {
      "epoch": 0.2660571428571429,
      "grad_norm": 1.1569546461105347,
      "learning_rate": 0.0005,
      "loss": 2.8651583194732666,
      "step": 1164
    },
    {
      "epoch": 0.2662857142857143,
      "grad_norm": 1.263775110244751,
      "learning_rate": 0.0005,
      "loss": 3.198338747024536,
      "step": 1165
    },
    {
      "epoch": 0.2665142857142857,
      "grad_norm": 1.2670981884002686,
      "learning_rate": 0.0005,
      "loss": 3.3212804794311523,
      "step": 1166
    },
    {
      "epoch": 0.26674285714285717,
      "grad_norm": 1.0727367401123047,
      "learning_rate": 0.0005,
      "loss": 3.04150128364563,
      "step": 1167
    },
    {
      "epoch": 0.2669714285714286,
      "grad_norm": 1.1552351713180542,
      "learning_rate": 0.0005,
      "loss": 3.491931676864624,
      "step": 1168
    },
    {
      "epoch": 0.2672,
      "grad_norm": 1.18842613697052,
      "learning_rate": 0.0005,
      "loss": 3.3339600563049316,
      "step": 1169
    },
    {
      "epoch": 0.2674285714285714,
      "grad_norm": 1.3301336765289307,
      "learning_rate": 0.0005,
      "loss": 3.3171558380126953,
      "step": 1170
    },
    {
      "epoch": 0.26765714285714287,
      "grad_norm": 1.13605797290802,
      "learning_rate": 0.0005,
      "loss": 3.6724705696105957,
      "step": 1171
    },
    {
      "epoch": 0.2678857142857143,
      "grad_norm": 1.7229629755020142,
      "learning_rate": 0.0005,
      "loss": 3.8825063705444336,
      "step": 1172
    },
    {
      "epoch": 0.2681142857142857,
      "grad_norm": 1.130028486251831,
      "learning_rate": 0.0005,
      "loss": 3.378112316131592,
      "step": 1173
    },
    {
      "epoch": 0.26834285714285716,
      "grad_norm": 1.7026365995407104,
      "learning_rate": 0.0005,
      "loss": 3.357504367828369,
      "step": 1174
    },
    {
      "epoch": 0.26857142857142857,
      "grad_norm": 1.0940172672271729,
      "learning_rate": 0.0005,
      "loss": 3.4884300231933594,
      "step": 1175
    },
    {
      "epoch": 0.2688,
      "grad_norm": 1.2888609170913696,
      "learning_rate": 0.0005,
      "loss": 3.145930290222168,
      "step": 1176
    },
    {
      "epoch": 0.26902857142857145,
      "grad_norm": 1.6060230731964111,
      "learning_rate": 0.0005,
      "loss": 3.5598437786102295,
      "step": 1177
    },
    {
      "epoch": 0.26925714285714286,
      "grad_norm": 1.2235252857208252,
      "learning_rate": 0.0005,
      "loss": 3.2463154792785645,
      "step": 1178
    },
    {
      "epoch": 0.2694857142857143,
      "grad_norm": 1.3172038793563843,
      "learning_rate": 0.0005,
      "loss": 3.235992431640625,
      "step": 1179
    },
    {
      "epoch": 0.26971428571428574,
      "grad_norm": 1.2928788661956787,
      "learning_rate": 0.0005,
      "loss": 3.71016788482666,
      "step": 1180
    },
    {
      "epoch": 0.26994285714285715,
      "grad_norm": 1.0961594581604004,
      "learning_rate": 0.0005,
      "loss": 3.2724499702453613,
      "step": 1181
    },
    {
      "epoch": 0.27017142857142856,
      "grad_norm": 1.26394784450531,
      "learning_rate": 0.0005,
      "loss": 3.477984666824341,
      "step": 1182
    },
    {
      "epoch": 0.2704,
      "grad_norm": 1.0932642221450806,
      "learning_rate": 0.0005,
      "loss": 3.7292890548706055,
      "step": 1183
    },
    {
      "epoch": 0.27062857142857144,
      "grad_norm": 1.5928343534469604,
      "learning_rate": 0.0005,
      "loss": 3.3703560829162598,
      "step": 1184
    },
    {
      "epoch": 0.27085714285714285,
      "grad_norm": 1.2527251243591309,
      "learning_rate": 0.0005,
      "loss": 3.1498351097106934,
      "step": 1185
    },
    {
      "epoch": 0.27108571428571426,
      "grad_norm": 1.5908743143081665,
      "learning_rate": 0.0005,
      "loss": 3.775286912918091,
      "step": 1186
    },
    {
      "epoch": 0.27131428571428573,
      "grad_norm": 1.5693947076797485,
      "learning_rate": 0.0005,
      "loss": 3.555837392807007,
      "step": 1187
    },
    {
      "epoch": 0.27154285714285714,
      "grad_norm": 1.2112048864364624,
      "learning_rate": 0.0005,
      "loss": 3.0062990188598633,
      "step": 1188
    },
    {
      "epoch": 0.27177142857142855,
      "grad_norm": 1.2896339893341064,
      "learning_rate": 0.0005,
      "loss": 3.49302339553833,
      "step": 1189
    },
    {
      "epoch": 0.272,
      "grad_norm": 1.2455127239227295,
      "learning_rate": 0.0005,
      "loss": 3.0723752975463867,
      "step": 1190
    },
    {
      "epoch": 0.27222857142857143,
      "grad_norm": 1.4943790435791016,
      "learning_rate": 0.0005,
      "loss": 3.2497401237487793,
      "step": 1191
    },
    {
      "epoch": 0.27245714285714284,
      "grad_norm": 1.2298558950424194,
      "learning_rate": 0.0005,
      "loss": 3.1432204246520996,
      "step": 1192
    },
    {
      "epoch": 0.2726857142857143,
      "grad_norm": 1.2510228157043457,
      "learning_rate": 0.0005,
      "loss": 3.167954921722412,
      "step": 1193
    },
    {
      "epoch": 0.2729142857142857,
      "grad_norm": 1.3390601873397827,
      "learning_rate": 0.0005,
      "loss": 3.3851888179779053,
      "step": 1194
    },
    {
      "epoch": 0.27314285714285713,
      "grad_norm": 1.2618153095245361,
      "learning_rate": 0.0005,
      "loss": 3.1554629802703857,
      "step": 1195
    },
    {
      "epoch": 0.2733714285714286,
      "grad_norm": 1.2907336950302124,
      "learning_rate": 0.0005,
      "loss": 3.310366630554199,
      "step": 1196
    },
    {
      "epoch": 0.2736,
      "grad_norm": 1.1101067066192627,
      "learning_rate": 0.0005,
      "loss": 2.996065139770508,
      "step": 1197
    },
    {
      "epoch": 0.2738285714285714,
      "grad_norm": 1.8547704219818115,
      "learning_rate": 0.0005,
      "loss": 3.17760968208313,
      "step": 1198
    },
    {
      "epoch": 0.27405714285714283,
      "grad_norm": 1.1432685852050781,
      "learning_rate": 0.0005,
      "loss": 3.3847527503967285,
      "step": 1199
    },
    {
      "epoch": 0.2742857142857143,
      "grad_norm": 1.194034218788147,
      "learning_rate": 0.0005,
      "loss": 2.930324077606201,
      "step": 1200
    },
    {
      "epoch": 0.2745142857142857,
      "grad_norm": 1.3310165405273438,
      "learning_rate": 0.0005,
      "loss": 3.166701316833496,
      "step": 1201
    },
    {
      "epoch": 0.2747428571428571,
      "grad_norm": 1.2739993333816528,
      "learning_rate": 0.0005,
      "loss": 3.095787525177002,
      "step": 1202
    },
    {
      "epoch": 0.2749714285714286,
      "grad_norm": 1.540579080581665,
      "learning_rate": 0.0005,
      "loss": 3.5183167457580566,
      "step": 1203
    },
    {
      "epoch": 0.2752,
      "grad_norm": 1.0412670373916626,
      "learning_rate": 0.0005,
      "loss": 3.1010334491729736,
      "step": 1204
    },
    {
      "epoch": 0.2754285714285714,
      "grad_norm": 1.3032169342041016,
      "learning_rate": 0.0005,
      "loss": 3.6486775875091553,
      "step": 1205
    },
    {
      "epoch": 0.2756571428571429,
      "grad_norm": 1.3415813446044922,
      "learning_rate": 0.0005,
      "loss": 3.104462146759033,
      "step": 1206
    },
    {
      "epoch": 0.2758857142857143,
      "grad_norm": 1.323291301727295,
      "learning_rate": 0.0005,
      "loss": 3.2977476119995117,
      "step": 1207
    },
    {
      "epoch": 0.2761142857142857,
      "grad_norm": 1.1914621591567993,
      "learning_rate": 0.0005,
      "loss": 3.342111349105835,
      "step": 1208
    },
    {
      "epoch": 0.27634285714285717,
      "grad_norm": 1.1742202043533325,
      "learning_rate": 0.0005,
      "loss": 3.145495891571045,
      "step": 1209
    },
    {
      "epoch": 0.2765714285714286,
      "grad_norm": 1.315573811531067,
      "learning_rate": 0.0005,
      "loss": 3.146130084991455,
      "step": 1210
    },
    {
      "epoch": 0.2768,
      "grad_norm": 1.7448524236679077,
      "learning_rate": 0.0005,
      "loss": 3.213390588760376,
      "step": 1211
    },
    {
      "epoch": 0.2770285714285714,
      "grad_norm": 1.1753430366516113,
      "learning_rate": 0.0005,
      "loss": 3.2105846405029297,
      "step": 1212
    },
    {
      "epoch": 0.27725714285714287,
      "grad_norm": 1.5631451606750488,
      "learning_rate": 0.0005,
      "loss": 3.7732574939727783,
      "step": 1213
    },
    {
      "epoch": 0.2774857142857143,
      "grad_norm": 1.3066740036010742,
      "learning_rate": 0.0005,
      "loss": 2.9548697471618652,
      "step": 1214
    },
    {
      "epoch": 0.2777142857142857,
      "grad_norm": 1.0825923681259155,
      "learning_rate": 0.0005,
      "loss": 3.3694348335266113,
      "step": 1215
    },
    {
      "epoch": 0.27794285714285716,
      "grad_norm": 1.626889944076538,
      "learning_rate": 0.0005,
      "loss": 3.39821720123291,
      "step": 1216
    },
    {
      "epoch": 0.27817142857142857,
      "grad_norm": 2.4464962482452393,
      "learning_rate": 0.0005,
      "loss": 3.2429676055908203,
      "step": 1217
    },
    {
      "epoch": 0.2784,
      "grad_norm": 1.5415946245193481,
      "learning_rate": 0.0005,
      "loss": 3.2904446125030518,
      "step": 1218
    },
    {
      "epoch": 0.27862857142857145,
      "grad_norm": 1.7012736797332764,
      "learning_rate": 0.0005,
      "loss": 3.2663369178771973,
      "step": 1219
    },
    {
      "epoch": 0.27885714285714286,
      "grad_norm": 1.0456711053848267,
      "learning_rate": 0.0005,
      "loss": 3.15718150138855,
      "step": 1220
    },
    {
      "epoch": 0.27908571428571427,
      "grad_norm": 1.3481340408325195,
      "learning_rate": 0.0005,
      "loss": 2.996861457824707,
      "step": 1221
    },
    {
      "epoch": 0.27931428571428574,
      "grad_norm": 1.628391981124878,
      "learning_rate": 0.0005,
      "loss": 2.8585593700408936,
      "step": 1222
    },
    {
      "epoch": 0.27954285714285715,
      "grad_norm": 1.1095558404922485,
      "learning_rate": 0.0005,
      "loss": 3.334444046020508,
      "step": 1223
    },
    {
      "epoch": 0.27977142857142856,
      "grad_norm": 1.5706031322479248,
      "learning_rate": 0.0005,
      "loss": 3.1801161766052246,
      "step": 1224
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.7314481735229492,
      "learning_rate": 0.0005,
      "loss": 3.5005764961242676,
      "step": 1225
    },
    {
      "epoch": 0.28022857142857144,
      "grad_norm": 1.0176838636398315,
      "learning_rate": 0.0005,
      "loss": 3.03684139251709,
      "step": 1226
    },
    {
      "epoch": 0.28045714285714285,
      "grad_norm": 1.2707672119140625,
      "learning_rate": 0.0005,
      "loss": 3.462103843688965,
      "step": 1227
    },
    {
      "epoch": 0.28068571428571426,
      "grad_norm": 1.5014296770095825,
      "learning_rate": 0.0005,
      "loss": 3.239656448364258,
      "step": 1228
    },
    {
      "epoch": 0.2809142857142857,
      "grad_norm": 1.196157693862915,
      "learning_rate": 0.0005,
      "loss": 3.1693921089172363,
      "step": 1229
    },
    {
      "epoch": 0.28114285714285714,
      "grad_norm": 1.3977198600769043,
      "learning_rate": 0.0005,
      "loss": 2.975597381591797,
      "step": 1230
    },
    {
      "epoch": 0.28137142857142855,
      "grad_norm": 1.23699951171875,
      "learning_rate": 0.0005,
      "loss": 3.229715347290039,
      "step": 1231
    },
    {
      "epoch": 0.2816,
      "grad_norm": 1.26917564868927,
      "learning_rate": 0.0005,
      "loss": 2.7800967693328857,
      "step": 1232
    },
    {
      "epoch": 0.28182857142857143,
      "grad_norm": 1.2337307929992676,
      "learning_rate": 0.0005,
      "loss": 3.3331565856933594,
      "step": 1233
    },
    {
      "epoch": 0.28205714285714284,
      "grad_norm": 1.0826371908187866,
      "learning_rate": 0.0005,
      "loss": 3.1334805488586426,
      "step": 1234
    },
    {
      "epoch": 0.2822857142857143,
      "grad_norm": 1.2340575456619263,
      "learning_rate": 0.0005,
      "loss": 3.1149702072143555,
      "step": 1235
    },
    {
      "epoch": 0.2825142857142857,
      "grad_norm": 1.3242164850234985,
      "learning_rate": 0.0005,
      "loss": 3.2699742317199707,
      "step": 1236
    },
    {
      "epoch": 0.28274285714285713,
      "grad_norm": 2.1229469776153564,
      "learning_rate": 0.0005,
      "loss": 3.0191874504089355,
      "step": 1237
    },
    {
      "epoch": 0.2829714285714286,
      "grad_norm": 1.4605075120925903,
      "learning_rate": 0.0005,
      "loss": 3.3845860958099365,
      "step": 1238
    },
    {
      "epoch": 0.2832,
      "grad_norm": 1.4869647026062012,
      "learning_rate": 0.0005,
      "loss": 3.2715773582458496,
      "step": 1239
    },
    {
      "epoch": 0.2834285714285714,
      "grad_norm": 2.0444061756134033,
      "learning_rate": 0.0005,
      "loss": 3.234959602355957,
      "step": 1240
    },
    {
      "epoch": 0.28365714285714283,
      "grad_norm": 1.3293615579605103,
      "learning_rate": 0.0005,
      "loss": 3.5553812980651855,
      "step": 1241
    },
    {
      "epoch": 0.2838857142857143,
      "grad_norm": 1.2727874517440796,
      "learning_rate": 0.0005,
      "loss": 3.3695826530456543,
      "step": 1242
    },
    {
      "epoch": 0.2841142857142857,
      "grad_norm": 1.233292579650879,
      "learning_rate": 0.0005,
      "loss": 3.182534694671631,
      "step": 1243
    },
    {
      "epoch": 0.2843428571428571,
      "grad_norm": 1.1242637634277344,
      "learning_rate": 0.0005,
      "loss": 3.446258306503296,
      "step": 1244
    },
    {
      "epoch": 0.2845714285714286,
      "grad_norm": 1.38852059841156,
      "learning_rate": 0.0005,
      "loss": 2.9096875190734863,
      "step": 1245
    },
    {
      "epoch": 0.2848,
      "grad_norm": 1.362173080444336,
      "learning_rate": 0.0005,
      "loss": 3.1956071853637695,
      "step": 1246
    },
    {
      "epoch": 0.2850285714285714,
      "grad_norm": 1.2583366632461548,
      "learning_rate": 0.0005,
      "loss": 2.9728140830993652,
      "step": 1247
    },
    {
      "epoch": 0.2852571428571429,
      "grad_norm": 1.2362728118896484,
      "learning_rate": 0.0005,
      "loss": 3.2637948989868164,
      "step": 1248
    },
    {
      "epoch": 0.2854857142857143,
      "grad_norm": 1.7622458934783936,
      "learning_rate": 0.0005,
      "loss": 2.972169876098633,
      "step": 1249
    },
    {
      "epoch": 0.2857142857142857,
      "grad_norm": 1.8252391815185547,
      "learning_rate": 0.0005,
      "loss": 2.9050376415252686,
      "step": 1250
    },
    {
      "epoch": 0.28594285714285717,
      "grad_norm": 1.1345325708389282,
      "learning_rate": 0.0005,
      "loss": 3.2302629947662354,
      "step": 1251
    },
    {
      "epoch": 0.2861714285714286,
      "grad_norm": 1.049342393875122,
      "learning_rate": 0.0005,
      "loss": 3.6919939517974854,
      "step": 1252
    },
    {
      "epoch": 0.2864,
      "grad_norm": 1.6495188474655151,
      "learning_rate": 0.0005,
      "loss": 3.182962417602539,
      "step": 1253
    },
    {
      "epoch": 0.28662857142857145,
      "grad_norm": 1.3297065496444702,
      "learning_rate": 0.0005,
      "loss": 3.1617653369903564,
      "step": 1254
    },
    {
      "epoch": 0.28685714285714287,
      "grad_norm": 1.5409008264541626,
      "learning_rate": 0.0005,
      "loss": 2.8185343742370605,
      "step": 1255
    },
    {
      "epoch": 0.2870857142857143,
      "grad_norm": 1.3593114614486694,
      "learning_rate": 0.0005,
      "loss": 2.919954538345337,
      "step": 1256
    },
    {
      "epoch": 0.2873142857142857,
      "grad_norm": 1.3885464668273926,
      "learning_rate": 0.0005,
      "loss": 3.496601104736328,
      "step": 1257
    },
    {
      "epoch": 0.28754285714285716,
      "grad_norm": 1.1657496690750122,
      "learning_rate": 0.0005,
      "loss": 2.980564594268799,
      "step": 1258
    },
    {
      "epoch": 0.28777142857142857,
      "grad_norm": 1.4134432077407837,
      "learning_rate": 0.0005,
      "loss": 3.292083740234375,
      "step": 1259
    },
    {
      "epoch": 0.288,
      "grad_norm": 1.500281810760498,
      "learning_rate": 0.0005,
      "loss": 3.022618532180786,
      "step": 1260
    },
    {
      "epoch": 0.28822857142857145,
      "grad_norm": 1.2954200506210327,
      "learning_rate": 0.0005,
      "loss": 3.3744659423828125,
      "step": 1261
    },
    {
      "epoch": 0.28845714285714286,
      "grad_norm": 1.118239164352417,
      "learning_rate": 0.0005,
      "loss": 3.1151561737060547,
      "step": 1262
    },
    {
      "epoch": 0.28868571428571427,
      "grad_norm": 1.4089930057525635,
      "learning_rate": 0.0005,
      "loss": 2.664538621902466,
      "step": 1263
    },
    {
      "epoch": 0.28891428571428573,
      "grad_norm": 1.1355178356170654,
      "learning_rate": 0.0005,
      "loss": 3.011949062347412,
      "step": 1264
    },
    {
      "epoch": 0.28914285714285715,
      "grad_norm": 1.8647680282592773,
      "learning_rate": 0.0005,
      "loss": 3.191842794418335,
      "step": 1265
    },
    {
      "epoch": 0.28937142857142856,
      "grad_norm": 1.144972324371338,
      "learning_rate": 0.0005,
      "loss": 3.3134312629699707,
      "step": 1266
    },
    {
      "epoch": 0.2896,
      "grad_norm": 1.7355245351791382,
      "learning_rate": 0.0005,
      "loss": 3.192549467086792,
      "step": 1267
    },
    {
      "epoch": 0.28982857142857144,
      "grad_norm": 3.045879602432251,
      "learning_rate": 0.0005,
      "loss": 3.3347086906433105,
      "step": 1268
    },
    {
      "epoch": 0.29005714285714285,
      "grad_norm": 1.3617289066314697,
      "learning_rate": 0.0005,
      "loss": 3.1745083332061768,
      "step": 1269
    },
    {
      "epoch": 0.29028571428571426,
      "grad_norm": 1.4409037828445435,
      "learning_rate": 0.0005,
      "loss": 2.9237709045410156,
      "step": 1270
    },
    {
      "epoch": 0.2905142857142857,
      "grad_norm": 1.3802387714385986,
      "learning_rate": 0.0005,
      "loss": 2.793193817138672,
      "step": 1271
    },
    {
      "epoch": 0.29074285714285714,
      "grad_norm": 1.589816689491272,
      "learning_rate": 0.0005,
      "loss": 3.158074378967285,
      "step": 1272
    },
    {
      "epoch": 0.29097142857142855,
      "grad_norm": 1.283583641052246,
      "learning_rate": 0.0005,
      "loss": 3.1891212463378906,
      "step": 1273
    },
    {
      "epoch": 0.2912,
      "grad_norm": 1.1436586380004883,
      "learning_rate": 0.0005,
      "loss": 3.1666646003723145,
      "step": 1274
    },
    {
      "epoch": 0.2914285714285714,
      "grad_norm": 1.3009272813796997,
      "learning_rate": 0.0005,
      "loss": 2.7879042625427246,
      "step": 1275
    },
    {
      "epoch": 0.29165714285714284,
      "grad_norm": 1.1756597757339478,
      "learning_rate": 0.0005,
      "loss": 3.4321794509887695,
      "step": 1276
    },
    {
      "epoch": 0.2918857142857143,
      "grad_norm": 1.5920088291168213,
      "learning_rate": 0.0005,
      "loss": 3.12089467048645,
      "step": 1277
    },
    {
      "epoch": 0.2921142857142857,
      "grad_norm": 1.2097954750061035,
      "learning_rate": 0.0005,
      "loss": 3.0529661178588867,
      "step": 1278
    },
    {
      "epoch": 0.2923428571428571,
      "grad_norm": 1.5644657611846924,
      "learning_rate": 0.0005,
      "loss": 2.932223320007324,
      "step": 1279
    },
    {
      "epoch": 0.2925714285714286,
      "grad_norm": 1.296210527420044,
      "learning_rate": 0.0005,
      "loss": 3.0020875930786133,
      "step": 1280
    },
    {
      "epoch": 0.2928,
      "grad_norm": 1.4031929969787598,
      "learning_rate": 0.0005,
      "loss": 3.287095069885254,
      "step": 1281
    },
    {
      "epoch": 0.2930285714285714,
      "grad_norm": 1.0933328866958618,
      "learning_rate": 0.0005,
      "loss": 3.1920998096466064,
      "step": 1282
    },
    {
      "epoch": 0.2932571428571429,
      "grad_norm": 1.1517566442489624,
      "learning_rate": 0.0005,
      "loss": 3.467801570892334,
      "step": 1283
    },
    {
      "epoch": 0.2934857142857143,
      "grad_norm": 1.028583288192749,
      "learning_rate": 0.0005,
      "loss": 2.9619626998901367,
      "step": 1284
    },
    {
      "epoch": 0.2937142857142857,
      "grad_norm": 1.675561785697937,
      "learning_rate": 0.0005,
      "loss": 3.1903390884399414,
      "step": 1285
    },
    {
      "epoch": 0.2939428571428571,
      "grad_norm": 1.6062946319580078,
      "learning_rate": 0.0005,
      "loss": 3.260089874267578,
      "step": 1286
    },
    {
      "epoch": 0.2941714285714286,
      "grad_norm": 1.404426097869873,
      "learning_rate": 0.0005,
      "loss": 3.535794258117676,
      "step": 1287
    },
    {
      "epoch": 0.2944,
      "grad_norm": 1.8842953443527222,
      "learning_rate": 0.0005,
      "loss": 3.0890326499938965,
      "step": 1288
    },
    {
      "epoch": 0.2946285714285714,
      "grad_norm": 1.4955906867980957,
      "learning_rate": 0.0005,
      "loss": 2.783099412918091,
      "step": 1289
    },
    {
      "epoch": 0.2948571428571429,
      "grad_norm": 1.0883903503417969,
      "learning_rate": 0.0005,
      "loss": 3.12682843208313,
      "step": 1290
    },
    {
      "epoch": 0.2950857142857143,
      "grad_norm": 1.2491700649261475,
      "learning_rate": 0.0005,
      "loss": 3.26346755027771,
      "step": 1291
    },
    {
      "epoch": 0.2953142857142857,
      "grad_norm": 1.2952848672866821,
      "learning_rate": 0.0005,
      "loss": 2.9692776203155518,
      "step": 1292
    },
    {
      "epoch": 0.29554285714285716,
      "grad_norm": 1.381825566291809,
      "learning_rate": 0.0005,
      "loss": 3.445896863937378,
      "step": 1293
    },
    {
      "epoch": 0.2957714285714286,
      "grad_norm": 1.0524667501449585,
      "learning_rate": 0.0005,
      "loss": 2.8896260261535645,
      "step": 1294
    },
    {
      "epoch": 0.296,
      "grad_norm": 1.1858974695205688,
      "learning_rate": 0.0005,
      "loss": 3.1023669242858887,
      "step": 1295
    },
    {
      "epoch": 0.29622857142857145,
      "grad_norm": 0.9524094462394714,
      "learning_rate": 0.0005,
      "loss": 3.1190311908721924,
      "step": 1296
    },
    {
      "epoch": 0.29645714285714286,
      "grad_norm": 1.6224452257156372,
      "learning_rate": 0.0005,
      "loss": 3.422473430633545,
      "step": 1297
    },
    {
      "epoch": 0.2966857142857143,
      "grad_norm": 1.3847239017486572,
      "learning_rate": 0.0005,
      "loss": 2.8995299339294434,
      "step": 1298
    },
    {
      "epoch": 0.29691428571428574,
      "grad_norm": 1.7863752841949463,
      "learning_rate": 0.0005,
      "loss": 3.273808002471924,
      "step": 1299
    },
    {
      "epoch": 0.29714285714285715,
      "grad_norm": 1.6527682542800903,
      "learning_rate": 0.0005,
      "loss": 3.3800888061523438,
      "step": 1300
    },
    {
      "epoch": 0.29737142857142856,
      "grad_norm": 1.1588882207870483,
      "learning_rate": 0.0005,
      "loss": 3.2243034839630127,
      "step": 1301
    },
    {
      "epoch": 0.2976,
      "grad_norm": 1.1225181818008423,
      "learning_rate": 0.0005,
      "loss": 3.0309486389160156,
      "step": 1302
    },
    {
      "epoch": 0.29782857142857144,
      "grad_norm": 1.2973567247390747,
      "learning_rate": 0.0005,
      "loss": 3.2347402572631836,
      "step": 1303
    },
    {
      "epoch": 0.29805714285714285,
      "grad_norm": 1.6134244203567505,
      "learning_rate": 0.0005,
      "loss": 3.159984588623047,
      "step": 1304
    },
    {
      "epoch": 0.29828571428571427,
      "grad_norm": 1.201042890548706,
      "learning_rate": 0.0005,
      "loss": 3.045102834701538,
      "step": 1305
    },
    {
      "epoch": 0.29851428571428573,
      "grad_norm": 1.2970058917999268,
      "learning_rate": 0.0005,
      "loss": 3.1715540885925293,
      "step": 1306
    },
    {
      "epoch": 0.29874285714285714,
      "grad_norm": 1.7008657455444336,
      "learning_rate": 0.0005,
      "loss": 2.999441385269165,
      "step": 1307
    },
    {
      "epoch": 0.29897142857142855,
      "grad_norm": 1.0802770853042603,
      "learning_rate": 0.0005,
      "loss": 2.9398722648620605,
      "step": 1308
    },
    {
      "epoch": 0.2992,
      "grad_norm": 1.1034327745437622,
      "learning_rate": 0.0005,
      "loss": 3.093902111053467,
      "step": 1309
    },
    {
      "epoch": 0.29942857142857143,
      "grad_norm": 1.1061512231826782,
      "learning_rate": 0.0005,
      "loss": 3.1278185844421387,
      "step": 1310
    },
    {
      "epoch": 0.29965714285714284,
      "grad_norm": 1.2460627555847168,
      "learning_rate": 0.0005,
      "loss": 3.0442285537719727,
      "step": 1311
    },
    {
      "epoch": 0.2998857142857143,
      "grad_norm": 1.0190467834472656,
      "learning_rate": 0.0005,
      "loss": 2.769345283508301,
      "step": 1312
    },
    {
      "epoch": 0.3001142857142857,
      "grad_norm": 1.3092217445373535,
      "learning_rate": 0.0005,
      "loss": 3.3877856731414795,
      "step": 1313
    },
    {
      "epoch": 0.30034285714285713,
      "grad_norm": 1.3094990253448486,
      "learning_rate": 0.0005,
      "loss": 3.549006462097168,
      "step": 1314
    },
    {
      "epoch": 0.30057142857142854,
      "grad_norm": 1.2474793195724487,
      "learning_rate": 0.0005,
      "loss": 3.396033763885498,
      "step": 1315
    },
    {
      "epoch": 0.3008,
      "grad_norm": 0.9892023801803589,
      "learning_rate": 0.0005,
      "loss": 3.5365896224975586,
      "step": 1316
    },
    {
      "epoch": 0.3010285714285714,
      "grad_norm": 1.2790782451629639,
      "learning_rate": 0.0005,
      "loss": 3.2631583213806152,
      "step": 1317
    },
    {
      "epoch": 0.30125714285714283,
      "grad_norm": 1.088068962097168,
      "learning_rate": 0.0005,
      "loss": 3.684664011001587,
      "step": 1318
    },
    {
      "epoch": 0.3014857142857143,
      "grad_norm": 1.2268731594085693,
      "learning_rate": 0.0005,
      "loss": 3.340979814529419,
      "step": 1319
    },
    {
      "epoch": 0.3017142857142857,
      "grad_norm": 1.128939151763916,
      "learning_rate": 0.0005,
      "loss": 3.124622344970703,
      "step": 1320
    },
    {
      "epoch": 0.3019428571428571,
      "grad_norm": 1.1330783367156982,
      "learning_rate": 0.0005,
      "loss": 3.4505274295806885,
      "step": 1321
    },
    {
      "epoch": 0.3021714285714286,
      "grad_norm": 1.5845942497253418,
      "learning_rate": 0.0005,
      "loss": 3.457763195037842,
      "step": 1322
    },
    {
      "epoch": 0.3024,
      "grad_norm": 1.0656205415725708,
      "learning_rate": 0.0005,
      "loss": 2.958230972290039,
      "step": 1323
    },
    {
      "epoch": 0.3026285714285714,
      "grad_norm": 0.9824689030647278,
      "learning_rate": 0.0005,
      "loss": 3.1518049240112305,
      "step": 1324
    },
    {
      "epoch": 0.3028571428571429,
      "grad_norm": 1.211126446723938,
      "learning_rate": 0.0005,
      "loss": 3.197604179382324,
      "step": 1325
    },
    {
      "epoch": 0.3030857142857143,
      "grad_norm": 1.2327104806900024,
      "learning_rate": 0.0005,
      "loss": 3.2247157096862793,
      "step": 1326
    },
    {
      "epoch": 0.3033142857142857,
      "grad_norm": 1.0731837749481201,
      "learning_rate": 0.0005,
      "loss": 3.6098575592041016,
      "step": 1327
    },
    {
      "epoch": 0.30354285714285717,
      "grad_norm": 1.1752265691757202,
      "learning_rate": 0.0005,
      "loss": 3.389374256134033,
      "step": 1328
    },
    {
      "epoch": 0.3037714285714286,
      "grad_norm": 1.1723414659500122,
      "learning_rate": 0.0005,
      "loss": 2.879910469055176,
      "step": 1329
    },
    {
      "epoch": 0.304,
      "grad_norm": 1.6797932386398315,
      "learning_rate": 0.0005,
      "loss": 3.0986757278442383,
      "step": 1330
    },
    {
      "epoch": 0.3042285714285714,
      "grad_norm": 1.2295936346054077,
      "learning_rate": 0.0005,
      "loss": 3.1775941848754883,
      "step": 1331
    },
    {
      "epoch": 0.30445714285714287,
      "grad_norm": 1.4732534885406494,
      "learning_rate": 0.0005,
      "loss": 3.2420966625213623,
      "step": 1332
    },
    {
      "epoch": 0.3046857142857143,
      "grad_norm": 1.2302976846694946,
      "learning_rate": 0.0005,
      "loss": 3.4029245376586914,
      "step": 1333
    },
    {
      "epoch": 0.3049142857142857,
      "grad_norm": 1.233150839805603,
      "learning_rate": 0.0005,
      "loss": 2.8453574180603027,
      "step": 1334
    },
    {
      "epoch": 0.30514285714285716,
      "grad_norm": 1.0400515794754028,
      "learning_rate": 0.0005,
      "loss": 3.3577709197998047,
      "step": 1335
    },
    {
      "epoch": 0.30537142857142857,
      "grad_norm": 1.0919668674468994,
      "learning_rate": 0.0005,
      "loss": 3.183206081390381,
      "step": 1336
    },
    {
      "epoch": 0.3056,
      "grad_norm": 1.2815113067626953,
      "learning_rate": 0.0005,
      "loss": 3.038179397583008,
      "step": 1337
    },
    {
      "epoch": 0.30582857142857145,
      "grad_norm": 1.6908437013626099,
      "learning_rate": 0.0005,
      "loss": 3.3624632358551025,
      "step": 1338
    },
    {
      "epoch": 0.30605714285714286,
      "grad_norm": 1.3765575885772705,
      "learning_rate": 0.0005,
      "loss": 3.182565689086914,
      "step": 1339
    },
    {
      "epoch": 0.3062857142857143,
      "grad_norm": 1.2444515228271484,
      "learning_rate": 0.0005,
      "loss": 3.1496691703796387,
      "step": 1340
    },
    {
      "epoch": 0.30651428571428574,
      "grad_norm": 1.0785560607910156,
      "learning_rate": 0.0005,
      "loss": 3.077735424041748,
      "step": 1341
    },
    {
      "epoch": 0.30674285714285715,
      "grad_norm": 1.129799723625183,
      "learning_rate": 0.0005,
      "loss": 3.0206167697906494,
      "step": 1342
    },
    {
      "epoch": 0.30697142857142856,
      "grad_norm": 1.461406946182251,
      "learning_rate": 0.0005,
      "loss": 3.0937185287475586,
      "step": 1343
    },
    {
      "epoch": 0.3072,
      "grad_norm": 1.4530279636383057,
      "learning_rate": 0.0005,
      "loss": 2.9386744499206543,
      "step": 1344
    },
    {
      "epoch": 0.30742857142857144,
      "grad_norm": 1.6167030334472656,
      "learning_rate": 0.0005,
      "loss": 3.185572624206543,
      "step": 1345
    },
    {
      "epoch": 0.30765714285714285,
      "grad_norm": 1.43446946144104,
      "learning_rate": 0.0005,
      "loss": 3.2161612510681152,
      "step": 1346
    },
    {
      "epoch": 0.30788571428571426,
      "grad_norm": 1.3314180374145508,
      "learning_rate": 0.0005,
      "loss": 2.728403091430664,
      "step": 1347
    },
    {
      "epoch": 0.30811428571428573,
      "grad_norm": 1.1517152786254883,
      "learning_rate": 0.0005,
      "loss": 3.3404111862182617,
      "step": 1348
    },
    {
      "epoch": 0.30834285714285714,
      "grad_norm": 1.3414390087127686,
      "learning_rate": 0.0005,
      "loss": 3.2150325775146484,
      "step": 1349
    },
    {
      "epoch": 0.30857142857142855,
      "grad_norm": 1.4766056537628174,
      "learning_rate": 0.0005,
      "loss": 3.3983869552612305,
      "step": 1350
    },
    {
      "epoch": 0.3088,
      "grad_norm": 1.2971340417861938,
      "learning_rate": 0.0005,
      "loss": 2.9799509048461914,
      "step": 1351
    },
    {
      "epoch": 0.30902857142857143,
      "grad_norm": 1.5991512537002563,
      "learning_rate": 0.0005,
      "loss": 3.3001112937927246,
      "step": 1352
    },
    {
      "epoch": 0.30925714285714284,
      "grad_norm": 1.697301983833313,
      "learning_rate": 0.0005,
      "loss": 3.2989683151245117,
      "step": 1353
    },
    {
      "epoch": 0.3094857142857143,
      "grad_norm": 1.4995174407958984,
      "learning_rate": 0.0005,
      "loss": 3.4403700828552246,
      "step": 1354
    },
    {
      "epoch": 0.3097142857142857,
      "grad_norm": 1.50566828250885,
      "learning_rate": 0.0005,
      "loss": 3.582648754119873,
      "step": 1355
    },
    {
      "epoch": 0.30994285714285713,
      "grad_norm": 1.5203014612197876,
      "learning_rate": 0.0005,
      "loss": 3.238114356994629,
      "step": 1356
    },
    {
      "epoch": 0.3101714285714286,
      "grad_norm": 1.155557632446289,
      "learning_rate": 0.0005,
      "loss": 3.0030558109283447,
      "step": 1357
    },
    {
      "epoch": 0.3104,
      "grad_norm": 1.3219950199127197,
      "learning_rate": 0.0005,
      "loss": 3.2635607719421387,
      "step": 1358
    },
    {
      "epoch": 0.3106285714285714,
      "grad_norm": 1.2226790189743042,
      "learning_rate": 0.0005,
      "loss": 3.098021984100342,
      "step": 1359
    },
    {
      "epoch": 0.31085714285714283,
      "grad_norm": 1.097556710243225,
      "learning_rate": 0.0005,
      "loss": 3.190894603729248,
      "step": 1360
    },
    {
      "epoch": 0.3110857142857143,
      "grad_norm": 1.545290470123291,
      "learning_rate": 0.0005,
      "loss": 3.2028448581695557,
      "step": 1361
    },
    {
      "epoch": 0.3113142857142857,
      "grad_norm": 1.0197135210037231,
      "learning_rate": 0.0005,
      "loss": 3.202213764190674,
      "step": 1362
    },
    {
      "epoch": 0.3115428571428571,
      "grad_norm": 1.516146183013916,
      "learning_rate": 0.0005,
      "loss": 3.057041645050049,
      "step": 1363
    },
    {
      "epoch": 0.3117714285714286,
      "grad_norm": 1.322717308998108,
      "learning_rate": 0.0005,
      "loss": 3.057084560394287,
      "step": 1364
    },
    {
      "epoch": 0.312,
      "grad_norm": 1.2215323448181152,
      "learning_rate": 0.0005,
      "loss": 3.1287665367126465,
      "step": 1365
    },
    {
      "epoch": 0.3122285714285714,
      "grad_norm": 1.1227742433547974,
      "learning_rate": 0.0005,
      "loss": 3.0142524242401123,
      "step": 1366
    },
    {
      "epoch": 0.3124571428571429,
      "grad_norm": 1.4234375953674316,
      "learning_rate": 0.0005,
      "loss": 3.0867953300476074,
      "step": 1367
    },
    {
      "epoch": 0.3126857142857143,
      "grad_norm": 1.1381722688674927,
      "learning_rate": 0.0005,
      "loss": 2.9375715255737305,
      "step": 1368
    },
    {
      "epoch": 0.3129142857142857,
      "grad_norm": 1.6360914707183838,
      "learning_rate": 0.0005,
      "loss": 2.9607715606689453,
      "step": 1369
    },
    {
      "epoch": 0.31314285714285717,
      "grad_norm": 1.0941888093948364,
      "learning_rate": 0.0005,
      "loss": 3.419132947921753,
      "step": 1370
    },
    {
      "epoch": 0.3133714285714286,
      "grad_norm": 1.1085125207901,
      "learning_rate": 0.0005,
      "loss": 3.0498504638671875,
      "step": 1371
    },
    {
      "epoch": 0.3136,
      "grad_norm": 1.0800436735153198,
      "learning_rate": 0.0005,
      "loss": 3.335331916809082,
      "step": 1372
    },
    {
      "epoch": 0.3138285714285714,
      "grad_norm": 1.1820502281188965,
      "learning_rate": 0.0005,
      "loss": 3.461688995361328,
      "step": 1373
    },
    {
      "epoch": 0.31405714285714287,
      "grad_norm": 1.153828740119934,
      "learning_rate": 0.0005,
      "loss": 3.3034510612487793,
      "step": 1374
    },
    {
      "epoch": 0.3142857142857143,
      "grad_norm": 1.3893864154815674,
      "learning_rate": 0.0005,
      "loss": 2.781877279281616,
      "step": 1375
    },
    {
      "epoch": 0.3145142857142857,
      "grad_norm": 2.004297971725464,
      "learning_rate": 0.0005,
      "loss": 3.315354347229004,
      "step": 1376
    },
    {
      "epoch": 0.31474285714285716,
      "grad_norm": 1.144432783126831,
      "learning_rate": 0.0005,
      "loss": 2.9911866188049316,
      "step": 1377
    },
    {
      "epoch": 0.31497142857142857,
      "grad_norm": 1.2076307535171509,
      "learning_rate": 0.0005,
      "loss": 3.080955743789673,
      "step": 1378
    },
    {
      "epoch": 0.3152,
      "grad_norm": 1.5732918977737427,
      "learning_rate": 0.0005,
      "loss": 3.0440282821655273,
      "step": 1379
    },
    {
      "epoch": 0.31542857142857145,
      "grad_norm": 1.2424455881118774,
      "learning_rate": 0.0005,
      "loss": 3.1027722358703613,
      "step": 1380
    },
    {
      "epoch": 0.31565714285714286,
      "grad_norm": 1.3199373483657837,
      "learning_rate": 0.0005,
      "loss": 3.223367929458618,
      "step": 1381
    },
    {
      "epoch": 0.31588571428571427,
      "grad_norm": 1.0423082113265991,
      "learning_rate": 0.0005,
      "loss": 3.104689121246338,
      "step": 1382
    },
    {
      "epoch": 0.31611428571428574,
      "grad_norm": 1.1818991899490356,
      "learning_rate": 0.0005,
      "loss": 3.2064380645751953,
      "step": 1383
    },
    {
      "epoch": 0.31634285714285715,
      "grad_norm": 2.249645948410034,
      "learning_rate": 0.0005,
      "loss": 3.3569304943084717,
      "step": 1384
    },
    {
      "epoch": 0.31657142857142856,
      "grad_norm": 1.1605218648910522,
      "learning_rate": 0.0005,
      "loss": 3.217287302017212,
      "step": 1385
    },
    {
      "epoch": 0.3168,
      "grad_norm": 1.2419939041137695,
      "learning_rate": 0.0005,
      "loss": 3.1123743057250977,
      "step": 1386
    },
    {
      "epoch": 0.31702857142857144,
      "grad_norm": 1.5484821796417236,
      "learning_rate": 0.0005,
      "loss": 3.105515480041504,
      "step": 1387
    },
    {
      "epoch": 0.31725714285714285,
      "grad_norm": 1.515061855316162,
      "learning_rate": 0.0005,
      "loss": 3.0314016342163086,
      "step": 1388
    },
    {
      "epoch": 0.31748571428571426,
      "grad_norm": 1.3462635278701782,
      "learning_rate": 0.0005,
      "loss": 3.1509532928466797,
      "step": 1389
    },
    {
      "epoch": 0.3177142857142857,
      "grad_norm": 1.8057386875152588,
      "learning_rate": 0.0005,
      "loss": 3.3833227157592773,
      "step": 1390
    },
    {
      "epoch": 0.31794285714285714,
      "grad_norm": 1.4874948263168335,
      "learning_rate": 0.0005,
      "loss": 2.8343546390533447,
      "step": 1391
    },
    {
      "epoch": 0.31817142857142855,
      "grad_norm": 1.164695143699646,
      "learning_rate": 0.0005,
      "loss": 3.4169764518737793,
      "step": 1392
    },
    {
      "epoch": 0.3184,
      "grad_norm": 1.1509723663330078,
      "learning_rate": 0.0005,
      "loss": 2.6884078979492188,
      "step": 1393
    },
    {
      "epoch": 0.3186285714285714,
      "grad_norm": 1.1875051259994507,
      "learning_rate": 0.0005,
      "loss": 2.972165107727051,
      "step": 1394
    },
    {
      "epoch": 0.31885714285714284,
      "grad_norm": 1.1930793523788452,
      "learning_rate": 0.0005,
      "loss": 3.645357847213745,
      "step": 1395
    },
    {
      "epoch": 0.3190857142857143,
      "grad_norm": 1.3768837451934814,
      "learning_rate": 0.0005,
      "loss": 2.859269618988037,
      "step": 1396
    },
    {
      "epoch": 0.3193142857142857,
      "grad_norm": 1.3687316179275513,
      "learning_rate": 0.0005,
      "loss": 3.24880313873291,
      "step": 1397
    },
    {
      "epoch": 0.31954285714285713,
      "grad_norm": 1.2957872152328491,
      "learning_rate": 0.0005,
      "loss": 3.173781156539917,
      "step": 1398
    },
    {
      "epoch": 0.3197714285714286,
      "grad_norm": 1.4422671794891357,
      "learning_rate": 0.0005,
      "loss": 3.064967632293701,
      "step": 1399
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.384992241859436,
      "learning_rate": 0.0005,
      "loss": 3.4461681842803955,
      "step": 1400
    },
    {
      "epoch": 0.3202285714285714,
      "grad_norm": 1.0419827699661255,
      "learning_rate": 0.0005,
      "loss": 3.597015380859375,
      "step": 1401
    },
    {
      "epoch": 0.3204571428571429,
      "grad_norm": 1.1182477474212646,
      "learning_rate": 0.0005,
      "loss": 3.269979476928711,
      "step": 1402
    },
    {
      "epoch": 0.3206857142857143,
      "grad_norm": 1.3293821811676025,
      "learning_rate": 0.0005,
      "loss": 3.1145238876342773,
      "step": 1403
    },
    {
      "epoch": 0.3209142857142857,
      "grad_norm": 1.0571242570877075,
      "learning_rate": 0.0005,
      "loss": 3.4211578369140625,
      "step": 1404
    },
    {
      "epoch": 0.3211428571428571,
      "grad_norm": 1.0650944709777832,
      "learning_rate": 0.0005,
      "loss": 3.2511484622955322,
      "step": 1405
    },
    {
      "epoch": 0.3213714285714286,
      "grad_norm": 1.0603253841400146,
      "learning_rate": 0.0005,
      "loss": 2.8438167572021484,
      "step": 1406
    },
    {
      "epoch": 0.3216,
      "grad_norm": 1.6355230808258057,
      "learning_rate": 0.0005,
      "loss": 3.537585735321045,
      "step": 1407
    },
    {
      "epoch": 0.3218285714285714,
      "grad_norm": 1.132704734802246,
      "learning_rate": 0.0005,
      "loss": 3.4857943058013916,
      "step": 1408
    },
    {
      "epoch": 0.3220571428571429,
      "grad_norm": 1.0465151071548462,
      "learning_rate": 0.0005,
      "loss": 3.125454902648926,
      "step": 1409
    },
    {
      "epoch": 0.3222857142857143,
      "grad_norm": 1.1481142044067383,
      "learning_rate": 0.0005,
      "loss": 2.8308522701263428,
      "step": 1410
    },
    {
      "epoch": 0.3225142857142857,
      "grad_norm": 1.1348906755447388,
      "learning_rate": 0.0005,
      "loss": 3.4766998291015625,
      "step": 1411
    },
    {
      "epoch": 0.32274285714285716,
      "grad_norm": 1.3944166898727417,
      "learning_rate": 0.0005,
      "loss": 3.0533947944641113,
      "step": 1412
    },
    {
      "epoch": 0.3229714285714286,
      "grad_norm": 1.1369953155517578,
      "learning_rate": 0.0005,
      "loss": 2.8827097415924072,
      "step": 1413
    },
    {
      "epoch": 0.3232,
      "grad_norm": 1.662022352218628,
      "learning_rate": 0.0005,
      "loss": 3.27250337600708,
      "step": 1414
    },
    {
      "epoch": 0.32342857142857145,
      "grad_norm": 1.7585558891296387,
      "learning_rate": 0.0005,
      "loss": 3.0540621280670166,
      "step": 1415
    },
    {
      "epoch": 0.32365714285714287,
      "grad_norm": 1.8051854372024536,
      "learning_rate": 0.0005,
      "loss": 2.972360134124756,
      "step": 1416
    },
    {
      "epoch": 0.3238857142857143,
      "grad_norm": 1.0855604410171509,
      "learning_rate": 0.0005,
      "loss": 3.0212299823760986,
      "step": 1417
    },
    {
      "epoch": 0.3241142857142857,
      "grad_norm": 1.1538416147232056,
      "learning_rate": 0.0005,
      "loss": 3.2980613708496094,
      "step": 1418
    },
    {
      "epoch": 0.32434285714285715,
      "grad_norm": 1.6274232864379883,
      "learning_rate": 0.0005,
      "loss": 3.0901858806610107,
      "step": 1419
    },
    {
      "epoch": 0.32457142857142857,
      "grad_norm": 1.1027473211288452,
      "learning_rate": 0.0005,
      "loss": 3.33320951461792,
      "step": 1420
    },
    {
      "epoch": 0.3248,
      "grad_norm": 1.28219735622406,
      "learning_rate": 0.0005,
      "loss": 3.052783966064453,
      "step": 1421
    },
    {
      "epoch": 0.32502857142857144,
      "grad_norm": 1.3240286111831665,
      "learning_rate": 0.0005,
      "loss": 3.151001453399658,
      "step": 1422
    },
    {
      "epoch": 0.32525714285714286,
      "grad_norm": 1.2825063467025757,
      "learning_rate": 0.0005,
      "loss": 2.9578847885131836,
      "step": 1423
    },
    {
      "epoch": 0.32548571428571427,
      "grad_norm": 1.1304113864898682,
      "learning_rate": 0.0005,
      "loss": 3.0259649753570557,
      "step": 1424
    },
    {
      "epoch": 0.32571428571428573,
      "grad_norm": 1.1474049091339111,
      "learning_rate": 0.0005,
      "loss": 3.1530354022979736,
      "step": 1425
    },
    {
      "epoch": 0.32594285714285715,
      "grad_norm": 1.6830153465270996,
      "learning_rate": 0.0005,
      "loss": 2.977947950363159,
      "step": 1426
    },
    {
      "epoch": 0.32617142857142856,
      "grad_norm": 1.0397148132324219,
      "learning_rate": 0.0005,
      "loss": 3.1847643852233887,
      "step": 1427
    },
    {
      "epoch": 0.3264,
      "grad_norm": 1.179452896118164,
      "learning_rate": 0.0005,
      "loss": 3.5262842178344727,
      "step": 1428
    },
    {
      "epoch": 0.32662857142857143,
      "grad_norm": 1.409746527671814,
      "learning_rate": 0.0005,
      "loss": 2.9470620155334473,
      "step": 1429
    },
    {
      "epoch": 0.32685714285714285,
      "grad_norm": 0.9633854627609253,
      "learning_rate": 0.0005,
      "loss": 3.1498143672943115,
      "step": 1430
    },
    {
      "epoch": 0.3270857142857143,
      "grad_norm": 1.1488559246063232,
      "learning_rate": 0.0005,
      "loss": 3.1063754558563232,
      "step": 1431
    },
    {
      "epoch": 0.3273142857142857,
      "grad_norm": 1.0837615728378296,
      "learning_rate": 0.0005,
      "loss": 2.8326590061187744,
      "step": 1432
    },
    {
      "epoch": 0.32754285714285714,
      "grad_norm": 1.1347260475158691,
      "learning_rate": 0.0005,
      "loss": 3.067315101623535,
      "step": 1433
    },
    {
      "epoch": 0.32777142857142855,
      "grad_norm": 1.5531283617019653,
      "learning_rate": 0.0005,
      "loss": 3.246004104614258,
      "step": 1434
    },
    {
      "epoch": 0.328,
      "grad_norm": 1.3794217109680176,
      "learning_rate": 0.0005,
      "loss": 3.0059239864349365,
      "step": 1435
    },
    {
      "epoch": 0.3282285714285714,
      "grad_norm": 1.3847702741622925,
      "learning_rate": 0.0005,
      "loss": 3.0965261459350586,
      "step": 1436
    },
    {
      "epoch": 0.32845714285714284,
      "grad_norm": 0.9805406928062439,
      "learning_rate": 0.0005,
      "loss": 3.274094581604004,
      "step": 1437
    },
    {
      "epoch": 0.3286857142857143,
      "grad_norm": 1.0479416847229004,
      "learning_rate": 0.0005,
      "loss": 3.243706226348877,
      "step": 1438
    },
    {
      "epoch": 0.3289142857142857,
      "grad_norm": 1.3363605737686157,
      "learning_rate": 0.0005,
      "loss": 3.4622557163238525,
      "step": 1439
    },
    {
      "epoch": 0.3291428571428571,
      "grad_norm": 1.368344783782959,
      "learning_rate": 0.0005,
      "loss": 3.098651170730591,
      "step": 1440
    },
    {
      "epoch": 0.3293714285714286,
      "grad_norm": 1.4114303588867188,
      "learning_rate": 0.0005,
      "loss": 3.2914841175079346,
      "step": 1441
    },
    {
      "epoch": 0.3296,
      "grad_norm": 1.2116928100585938,
      "learning_rate": 0.0005,
      "loss": 3.511561870574951,
      "step": 1442
    },
    {
      "epoch": 0.3298285714285714,
      "grad_norm": 1.3186230659484863,
      "learning_rate": 0.0005,
      "loss": 2.7973248958587646,
      "step": 1443
    },
    {
      "epoch": 0.3300571428571429,
      "grad_norm": 1.1073414087295532,
      "learning_rate": 0.0005,
      "loss": 3.402292251586914,
      "step": 1444
    },
    {
      "epoch": 0.3302857142857143,
      "grad_norm": 1.063323736190796,
      "learning_rate": 0.0005,
      "loss": 3.5148866176605225,
      "step": 1445
    },
    {
      "epoch": 0.3305142857142857,
      "grad_norm": 1.1195132732391357,
      "learning_rate": 0.0005,
      "loss": 2.983470916748047,
      "step": 1446
    },
    {
      "epoch": 0.3307428571428571,
      "grad_norm": 1.000182032585144,
      "learning_rate": 0.0005,
      "loss": 3.2904603481292725,
      "step": 1447
    },
    {
      "epoch": 0.3309714285714286,
      "grad_norm": 1.321871042251587,
      "learning_rate": 0.0005,
      "loss": 3.2777671813964844,
      "step": 1448
    },
    {
      "epoch": 0.3312,
      "grad_norm": 1.1353954076766968,
      "learning_rate": 0.0005,
      "loss": 3.398280143737793,
      "step": 1449
    },
    {
      "epoch": 0.3314285714285714,
      "grad_norm": 1.2913039922714233,
      "learning_rate": 0.0005,
      "loss": 2.8804173469543457,
      "step": 1450
    },
    {
      "epoch": 0.3316571428571429,
      "grad_norm": 1.5643131732940674,
      "learning_rate": 0.0005,
      "loss": 2.9541735649108887,
      "step": 1451
    },
    {
      "epoch": 0.3318857142857143,
      "grad_norm": 1.2275824546813965,
      "learning_rate": 0.0005,
      "loss": 3.1878108978271484,
      "step": 1452
    },
    {
      "epoch": 0.3321142857142857,
      "grad_norm": 1.1011476516723633,
      "learning_rate": 0.0005,
      "loss": 2.6418967247009277,
      "step": 1453
    },
    {
      "epoch": 0.33234285714285716,
      "grad_norm": 1.0772278308868408,
      "learning_rate": 0.0005,
      "loss": 3.007988452911377,
      "step": 1454
    },
    {
      "epoch": 0.3325714285714286,
      "grad_norm": 1.4624085426330566,
      "learning_rate": 0.0005,
      "loss": 3.1157751083374023,
      "step": 1455
    },
    {
      "epoch": 0.3328,
      "grad_norm": 1.1049821376800537,
      "learning_rate": 0.0005,
      "loss": 3.0093021392822266,
      "step": 1456
    },
    {
      "epoch": 0.33302857142857145,
      "grad_norm": 1.1755868196487427,
      "learning_rate": 0.0005,
      "loss": 3.098945140838623,
      "step": 1457
    },
    {
      "epoch": 0.33325714285714286,
      "grad_norm": 1.2363975048065186,
      "learning_rate": 0.0005,
      "loss": 2.8327364921569824,
      "step": 1458
    },
    {
      "epoch": 0.3334857142857143,
      "grad_norm": 1.0498814582824707,
      "learning_rate": 0.0005,
      "loss": 3.078984022140503,
      "step": 1459
    },
    {
      "epoch": 0.33371428571428574,
      "grad_norm": 1.2851461172103882,
      "learning_rate": 0.0005,
      "loss": 3.171504497528076,
      "step": 1460
    },
    {
      "epoch": 0.33394285714285715,
      "grad_norm": 1.4365419149398804,
      "learning_rate": 0.0005,
      "loss": 3.0163567066192627,
      "step": 1461
    },
    {
      "epoch": 0.33417142857142856,
      "grad_norm": 1.5405396223068237,
      "learning_rate": 0.0005,
      "loss": 3.4664711952209473,
      "step": 1462
    },
    {
      "epoch": 0.3344,
      "grad_norm": 1.5877153873443604,
      "learning_rate": 0.0005,
      "loss": 3.161365509033203,
      "step": 1463
    },
    {
      "epoch": 0.33462857142857144,
      "grad_norm": 1.46943199634552,
      "learning_rate": 0.0005,
      "loss": 3.1301300525665283,
      "step": 1464
    },
    {
      "epoch": 0.33485714285714285,
      "grad_norm": 1.2365461587905884,
      "learning_rate": 0.0005,
      "loss": 3.2492287158966064,
      "step": 1465
    },
    {
      "epoch": 0.33508571428571426,
      "grad_norm": 1.2469853162765503,
      "learning_rate": 0.0005,
      "loss": 2.9323086738586426,
      "step": 1466
    },
    {
      "epoch": 0.33531428571428573,
      "grad_norm": 1.0540879964828491,
      "learning_rate": 0.0005,
      "loss": 3.076340436935425,
      "step": 1467
    },
    {
      "epoch": 0.33554285714285714,
      "grad_norm": 1.347556471824646,
      "learning_rate": 0.0005,
      "loss": 2.9597363471984863,
      "step": 1468
    },
    {
      "epoch": 0.33577142857142855,
      "grad_norm": 1.5497254133224487,
      "learning_rate": 0.0005,
      "loss": 3.4875285625457764,
      "step": 1469
    },
    {
      "epoch": 0.336,
      "grad_norm": 1.4359560012817383,
      "learning_rate": 0.0005,
      "loss": 3.1748106479644775,
      "step": 1470
    },
    {
      "epoch": 0.33622857142857143,
      "grad_norm": 1.3493833541870117,
      "learning_rate": 0.0005,
      "loss": 2.9081270694732666,
      "step": 1471
    },
    {
      "epoch": 0.33645714285714284,
      "grad_norm": 1.3964293003082275,
      "learning_rate": 0.0005,
      "loss": 3.06734561920166,
      "step": 1472
    },
    {
      "epoch": 0.3366857142857143,
      "grad_norm": 1.1086349487304688,
      "learning_rate": 0.0005,
      "loss": 3.2239503860473633,
      "step": 1473
    },
    {
      "epoch": 0.3369142857142857,
      "grad_norm": 0.9822249412536621,
      "learning_rate": 0.0005,
      "loss": 3.0340569019317627,
      "step": 1474
    },
    {
      "epoch": 0.33714285714285713,
      "grad_norm": 1.0283337831497192,
      "learning_rate": 0.0005,
      "loss": 3.0568771362304688,
      "step": 1475
    },
    {
      "epoch": 0.33737142857142854,
      "grad_norm": 2.1645429134368896,
      "learning_rate": 0.0005,
      "loss": 3.4917349815368652,
      "step": 1476
    },
    {
      "epoch": 0.3376,
      "grad_norm": 1.7463144063949585,
      "learning_rate": 0.0005,
      "loss": 3.3383219242095947,
      "step": 1477
    },
    {
      "epoch": 0.3378285714285714,
      "grad_norm": 1.1549912691116333,
      "learning_rate": 0.0005,
      "loss": 2.9099647998809814,
      "step": 1478
    },
    {
      "epoch": 0.33805714285714283,
      "grad_norm": 0.9665444493293762,
      "learning_rate": 0.0005,
      "loss": 3.056901693344116,
      "step": 1479
    },
    {
      "epoch": 0.3382857142857143,
      "grad_norm": 1.1496986150741577,
      "learning_rate": 0.0005,
      "loss": 3.23298978805542,
      "step": 1480
    },
    {
      "epoch": 0.3385142857142857,
      "grad_norm": 1.197044014930725,
      "learning_rate": 0.0005,
      "loss": 2.9651286602020264,
      "step": 1481
    },
    {
      "epoch": 0.3387428571428571,
      "grad_norm": 1.0992426872253418,
      "learning_rate": 0.0005,
      "loss": 2.87457013130188,
      "step": 1482
    },
    {
      "epoch": 0.3389714285714286,
      "grad_norm": 1.6564610004425049,
      "learning_rate": 0.0005,
      "loss": 3.2672181129455566,
      "step": 1483
    },
    {
      "epoch": 0.3392,
      "grad_norm": 1.227616310119629,
      "learning_rate": 0.0005,
      "loss": 3.430346727371216,
      "step": 1484
    },
    {
      "epoch": 0.3394285714285714,
      "grad_norm": 1.2955327033996582,
      "learning_rate": 0.0005,
      "loss": 3.3382816314697266,
      "step": 1485
    },
    {
      "epoch": 0.3396571428571429,
      "grad_norm": 2.1628823280334473,
      "learning_rate": 0.0005,
      "loss": 3.02040958404541,
      "step": 1486
    },
    {
      "epoch": 0.3398857142857143,
      "grad_norm": 2.716973304748535,
      "learning_rate": 0.0005,
      "loss": 3.6265311241149902,
      "step": 1487
    },
    {
      "epoch": 0.3401142857142857,
      "grad_norm": 1.8898112773895264,
      "learning_rate": 0.0005,
      "loss": 2.9270987510681152,
      "step": 1488
    },
    {
      "epoch": 0.34034285714285717,
      "grad_norm": 1.1531834602355957,
      "learning_rate": 0.0005,
      "loss": 2.779815196990967,
      "step": 1489
    },
    {
      "epoch": 0.3405714285714286,
      "grad_norm": 1.4023605585098267,
      "learning_rate": 0.0005,
      "loss": 3.443798542022705,
      "step": 1490
    },
    {
      "epoch": 0.3408,
      "grad_norm": 1.1286815404891968,
      "learning_rate": 0.0005,
      "loss": 3.375253915786743,
      "step": 1491
    },
    {
      "epoch": 0.3410285714285714,
      "grad_norm": 1.1725246906280518,
      "learning_rate": 0.0005,
      "loss": 3.0210962295532227,
      "step": 1492
    },
    {
      "epoch": 0.34125714285714287,
      "grad_norm": 1.5057566165924072,
      "learning_rate": 0.0005,
      "loss": 2.917236089706421,
      "step": 1493
    },
    {
      "epoch": 0.3414857142857143,
      "grad_norm": 1.8345794677734375,
      "learning_rate": 0.0005,
      "loss": 3.374966621398926,
      "step": 1494
    },
    {
      "epoch": 0.3417142857142857,
      "grad_norm": 2.0231735706329346,
      "learning_rate": 0.0005,
      "loss": 3.0984444618225098,
      "step": 1495
    },
    {
      "epoch": 0.34194285714285716,
      "grad_norm": 1.4853079319000244,
      "learning_rate": 0.0005,
      "loss": 3.1776602268218994,
      "step": 1496
    },
    {
      "epoch": 0.34217142857142857,
      "grad_norm": 1.4702025651931763,
      "learning_rate": 0.0005,
      "loss": 2.9002652168273926,
      "step": 1497
    },
    {
      "epoch": 0.3424,
      "grad_norm": 1.4146311283111572,
      "learning_rate": 0.0005,
      "loss": 3.637957811355591,
      "step": 1498
    },
    {
      "epoch": 0.34262857142857145,
      "grad_norm": 1.6286510229110718,
      "learning_rate": 0.0005,
      "loss": 3.0932068824768066,
      "step": 1499
    },
    {
      "epoch": 0.34285714285714286,
      "grad_norm": 1.8635056018829346,
      "learning_rate": 0.0005,
      "loss": 3.100038528442383,
      "step": 1500
    },
    {
      "epoch": 0.34308571428571427,
      "grad_norm": 1.0752707719802856,
      "learning_rate": 0.0005,
      "loss": 3.4639759063720703,
      "step": 1501
    },
    {
      "epoch": 0.34331428571428574,
      "grad_norm": 1.5707205533981323,
      "learning_rate": 0.0005,
      "loss": 3.2104105949401855,
      "step": 1502
    },
    {
      "epoch": 0.34354285714285715,
      "grad_norm": 1.1708844900131226,
      "learning_rate": 0.0005,
      "loss": 2.9831454753875732,
      "step": 1503
    },
    {
      "epoch": 0.34377142857142856,
      "grad_norm": 1.286978840827942,
      "learning_rate": 0.0005,
      "loss": 3.127631664276123,
      "step": 1504
    },
    {
      "epoch": 0.344,
      "grad_norm": 1.2354544401168823,
      "learning_rate": 0.0005,
      "loss": 3.245645761489868,
      "step": 1505
    },
    {
      "epoch": 0.34422857142857144,
      "grad_norm": 1.1794037818908691,
      "learning_rate": 0.0005,
      "loss": 3.062906265258789,
      "step": 1506
    },
    {
      "epoch": 0.34445714285714285,
      "grad_norm": 1.3049794435501099,
      "learning_rate": 0.0005,
      "loss": 3.1196365356445312,
      "step": 1507
    },
    {
      "epoch": 0.34468571428571426,
      "grad_norm": 1.2255795001983643,
      "learning_rate": 0.0005,
      "loss": 3.0585262775421143,
      "step": 1508
    },
    {
      "epoch": 0.34491428571428573,
      "grad_norm": 1.1102447509765625,
      "learning_rate": 0.0005,
      "loss": 3.180795907974243,
      "step": 1509
    },
    {
      "epoch": 0.34514285714285714,
      "grad_norm": 1.392860770225525,
      "learning_rate": 0.0005,
      "loss": 3.0589089393615723,
      "step": 1510
    },
    {
      "epoch": 0.34537142857142855,
      "grad_norm": 1.527366042137146,
      "learning_rate": 0.0005,
      "loss": 3.141733407974243,
      "step": 1511
    },
    {
      "epoch": 0.3456,
      "grad_norm": 1.4329783916473389,
      "learning_rate": 0.0005,
      "loss": 3.3561103343963623,
      "step": 1512
    },
    {
      "epoch": 0.34582857142857143,
      "grad_norm": 1.17677903175354,
      "learning_rate": 0.0005,
      "loss": 3.2288341522216797,
      "step": 1513
    },
    {
      "epoch": 0.34605714285714284,
      "grad_norm": 1.3530901670455933,
      "learning_rate": 0.0005,
      "loss": 3.7693629264831543,
      "step": 1514
    },
    {
      "epoch": 0.3462857142857143,
      "grad_norm": 1.4992903470993042,
      "learning_rate": 0.0005,
      "loss": 3.03387451171875,
      "step": 1515
    },
    {
      "epoch": 0.3465142857142857,
      "grad_norm": 1.3460344076156616,
      "learning_rate": 0.0005,
      "loss": 3.030808448791504,
      "step": 1516
    },
    {
      "epoch": 0.34674285714285713,
      "grad_norm": 1.6083482503890991,
      "learning_rate": 0.0005,
      "loss": 3.298140287399292,
      "step": 1517
    },
    {
      "epoch": 0.3469714285714286,
      "grad_norm": 1.130464792251587,
      "learning_rate": 0.0005,
      "loss": 3.124453544616699,
      "step": 1518
    },
    {
      "epoch": 0.3472,
      "grad_norm": 1.4121912717819214,
      "learning_rate": 0.0005,
      "loss": 3.3830246925354004,
      "step": 1519
    },
    {
      "epoch": 0.3474285714285714,
      "grad_norm": 1.5459245443344116,
      "learning_rate": 0.0005,
      "loss": 3.6011621952056885,
      "step": 1520
    },
    {
      "epoch": 0.34765714285714283,
      "grad_norm": 1.376255989074707,
      "learning_rate": 0.0005,
      "loss": 3.132066488265991,
      "step": 1521
    },
    {
      "epoch": 0.3478857142857143,
      "grad_norm": 1.2499432563781738,
      "learning_rate": 0.0005,
      "loss": 2.862654685974121,
      "step": 1522
    },
    {
      "epoch": 0.3481142857142857,
      "grad_norm": 1.0655391216278076,
      "learning_rate": 0.0005,
      "loss": 3.313263177871704,
      "step": 1523
    },
    {
      "epoch": 0.3483428571428571,
      "grad_norm": 1.3576215505599976,
      "learning_rate": 0.0005,
      "loss": 3.1183242797851562,
      "step": 1524
    },
    {
      "epoch": 0.3485714285714286,
      "grad_norm": 2.184209108352661,
      "learning_rate": 0.0005,
      "loss": 3.3371505737304688,
      "step": 1525
    },
    {
      "epoch": 0.3488,
      "grad_norm": 1.081516981124878,
      "learning_rate": 0.0005,
      "loss": 3.2336888313293457,
      "step": 1526
    },
    {
      "epoch": 0.3490285714285714,
      "grad_norm": 1.3522498607635498,
      "learning_rate": 0.0005,
      "loss": 3.375439167022705,
      "step": 1527
    },
    {
      "epoch": 0.3492571428571429,
      "grad_norm": 1.166731834411621,
      "learning_rate": 0.0005,
      "loss": 2.8967742919921875,
      "step": 1528
    },
    {
      "epoch": 0.3494857142857143,
      "grad_norm": 1.029475212097168,
      "learning_rate": 0.0005,
      "loss": 3.0708281993865967,
      "step": 1529
    },
    {
      "epoch": 0.3497142857142857,
      "grad_norm": 6.737969398498535,
      "learning_rate": 0.0005,
      "loss": 3.5321216583251953,
      "step": 1530
    },
    {
      "epoch": 0.34994285714285717,
      "grad_norm": 1.2734293937683105,
      "learning_rate": 0.0005,
      "loss": 3.0571181774139404,
      "step": 1531
    },
    {
      "epoch": 0.3501714285714286,
      "grad_norm": 1.0447072982788086,
      "learning_rate": 0.0005,
      "loss": 3.3661298751831055,
      "step": 1532
    },
    {
      "epoch": 0.3504,
      "grad_norm": 1.3441318273544312,
      "learning_rate": 0.0005,
      "loss": 2.8446714878082275,
      "step": 1533
    },
    {
      "epoch": 0.35062857142857146,
      "grad_norm": 1.3182196617126465,
      "learning_rate": 0.0005,
      "loss": 3.1556761264801025,
      "step": 1534
    },
    {
      "epoch": 0.35085714285714287,
      "grad_norm": 1.1335374116897583,
      "learning_rate": 0.0005,
      "loss": 2.9204421043395996,
      "step": 1535
    },
    {
      "epoch": 0.3510857142857143,
      "grad_norm": 1.1382598876953125,
      "learning_rate": 0.0005,
      "loss": 2.945805788040161,
      "step": 1536
    },
    {
      "epoch": 0.3513142857142857,
      "grad_norm": 1.7605595588684082,
      "learning_rate": 0.0005,
      "loss": 3.441171646118164,
      "step": 1537
    },
    {
      "epoch": 0.35154285714285716,
      "grad_norm": 1.4314618110656738,
      "learning_rate": 0.0005,
      "loss": 3.0069684982299805,
      "step": 1538
    },
    {
      "epoch": 0.35177142857142857,
      "grad_norm": 1.1811680793762207,
      "learning_rate": 0.0005,
      "loss": 3.0814545154571533,
      "step": 1539
    },
    {
      "epoch": 0.352,
      "grad_norm": 1.1010661125183105,
      "learning_rate": 0.0005,
      "loss": 2.928677797317505,
      "step": 1540
    },
    {
      "epoch": 0.35222857142857145,
      "grad_norm": 1.026636004447937,
      "learning_rate": 0.0005,
      "loss": 2.923400640487671,
      "step": 1541
    },
    {
      "epoch": 0.35245714285714286,
      "grad_norm": 1.0448089838027954,
      "learning_rate": 0.0005,
      "loss": 3.2129154205322266,
      "step": 1542
    },
    {
      "epoch": 0.35268571428571427,
      "grad_norm": 1.3825839757919312,
      "learning_rate": 0.0005,
      "loss": 2.9109532833099365,
      "step": 1543
    },
    {
      "epoch": 0.35291428571428574,
      "grad_norm": 1.0419598817825317,
      "learning_rate": 0.0005,
      "loss": 3.200833797454834,
      "step": 1544
    },
    {
      "epoch": 0.35314285714285715,
      "grad_norm": 1.7515592575073242,
      "learning_rate": 0.0005,
      "loss": 3.4257068634033203,
      "step": 1545
    },
    {
      "epoch": 0.35337142857142856,
      "grad_norm": 1.1635513305664062,
      "learning_rate": 0.0005,
      "loss": 3.4141104221343994,
      "step": 1546
    },
    {
      "epoch": 0.3536,
      "grad_norm": 0.958040177822113,
      "learning_rate": 0.0005,
      "loss": 2.9048283100128174,
      "step": 1547
    },
    {
      "epoch": 0.35382857142857144,
      "grad_norm": 1.011100172996521,
      "learning_rate": 0.0005,
      "loss": 3.4218590259552,
      "step": 1548
    },
    {
      "epoch": 0.35405714285714285,
      "grad_norm": 1.3057202100753784,
      "learning_rate": 0.0005,
      "loss": 3.218475818634033,
      "step": 1549
    },
    {
      "epoch": 0.35428571428571426,
      "grad_norm": 1.1131677627563477,
      "learning_rate": 0.0005,
      "loss": 3.001148223876953,
      "step": 1550
    },
    {
      "epoch": 0.3545142857142857,
      "grad_norm": 1.07765793800354,
      "learning_rate": 0.0005,
      "loss": 2.9367032051086426,
      "step": 1551
    },
    {
      "epoch": 0.35474285714285714,
      "grad_norm": 1.052375316619873,
      "learning_rate": 0.0005,
      "loss": 3.3938417434692383,
      "step": 1552
    },
    {
      "epoch": 0.35497142857142855,
      "grad_norm": 1.2066963911056519,
      "learning_rate": 0.0005,
      "loss": 2.8919084072113037,
      "step": 1553
    },
    {
      "epoch": 0.3552,
      "grad_norm": 1.156920075416565,
      "learning_rate": 0.0005,
      "loss": 2.805723190307617,
      "step": 1554
    },
    {
      "epoch": 0.3554285714285714,
      "grad_norm": 1.186169147491455,
      "learning_rate": 0.0005,
      "loss": 3.6866724491119385,
      "step": 1555
    },
    {
      "epoch": 0.35565714285714284,
      "grad_norm": 1.2721261978149414,
      "learning_rate": 0.0005,
      "loss": 3.260833501815796,
      "step": 1556
    },
    {
      "epoch": 0.3558857142857143,
      "grad_norm": 1.1024852991104126,
      "learning_rate": 0.0005,
      "loss": 3.1709134578704834,
      "step": 1557
    },
    {
      "epoch": 0.3561142857142857,
      "grad_norm": 1.0716303586959839,
      "learning_rate": 0.0005,
      "loss": 3.121380567550659,
      "step": 1558
    },
    {
      "epoch": 0.35634285714285713,
      "grad_norm": 1.1333616971969604,
      "learning_rate": 0.0005,
      "loss": 3.1244678497314453,
      "step": 1559
    },
    {
      "epoch": 0.3565714285714286,
      "grad_norm": 1.252811074256897,
      "learning_rate": 0.0005,
      "loss": 3.269632577896118,
      "step": 1560
    },
    {
      "epoch": 0.3568,
      "grad_norm": 1.4917107820510864,
      "learning_rate": 0.0005,
      "loss": 3.061126708984375,
      "step": 1561
    },
    {
      "epoch": 0.3570285714285714,
      "grad_norm": 1.078799843788147,
      "learning_rate": 0.0005,
      "loss": 3.162682056427002,
      "step": 1562
    },
    {
      "epoch": 0.3572571428571429,
      "grad_norm": 1.4843292236328125,
      "learning_rate": 0.0005,
      "loss": 2.910238742828369,
      "step": 1563
    },
    {
      "epoch": 0.3574857142857143,
      "grad_norm": 1.8405580520629883,
      "learning_rate": 0.0005,
      "loss": 3.374652624130249,
      "step": 1564
    },
    {
      "epoch": 0.3577142857142857,
      "grad_norm": 1.748340368270874,
      "learning_rate": 0.0005,
      "loss": 3.2390661239624023,
      "step": 1565
    },
    {
      "epoch": 0.3579428571428571,
      "grad_norm": 1.3139686584472656,
      "learning_rate": 0.0005,
      "loss": 2.946856737136841,
      "step": 1566
    },
    {
      "epoch": 0.3581714285714286,
      "grad_norm": 1.3156969547271729,
      "learning_rate": 0.0005,
      "loss": 3.109053611755371,
      "step": 1567
    },
    {
      "epoch": 0.3584,
      "grad_norm": 1.0489155054092407,
      "learning_rate": 0.0005,
      "loss": 3.4434714317321777,
      "step": 1568
    },
    {
      "epoch": 0.3586285714285714,
      "grad_norm": 0.979171872138977,
      "learning_rate": 0.0005,
      "loss": 3.3189635276794434,
      "step": 1569
    },
    {
      "epoch": 0.3588571428571429,
      "grad_norm": 1.0978304147720337,
      "learning_rate": 0.0005,
      "loss": 3.1540093421936035,
      "step": 1570
    },
    {
      "epoch": 0.3590857142857143,
      "grad_norm": 1.4407920837402344,
      "learning_rate": 0.0005,
      "loss": 2.97906756401062,
      "step": 1571
    },
    {
      "epoch": 0.3593142857142857,
      "grad_norm": 1.030862808227539,
      "learning_rate": 0.0005,
      "loss": 3.028770923614502,
      "step": 1572
    },
    {
      "epoch": 0.35954285714285716,
      "grad_norm": 1.0622620582580566,
      "learning_rate": 0.0005,
      "loss": 3.0124640464782715,
      "step": 1573
    },
    {
      "epoch": 0.3597714285714286,
      "grad_norm": 1.1256697177886963,
      "learning_rate": 0.0005,
      "loss": 3.1686272621154785,
      "step": 1574
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.1128168106079102,
      "learning_rate": 0.0005,
      "loss": 3.1530656814575195,
      "step": 1575
    },
    {
      "epoch": 0.36022857142857145,
      "grad_norm": 1.9827543497085571,
      "learning_rate": 0.0005,
      "loss": 3.0795295238494873,
      "step": 1576
    },
    {
      "epoch": 0.36045714285714286,
      "grad_norm": 1.2042571306228638,
      "learning_rate": 0.0005,
      "loss": 3.3377723693847656,
      "step": 1577
    },
    {
      "epoch": 0.3606857142857143,
      "grad_norm": 1.0781599283218384,
      "learning_rate": 0.0005,
      "loss": 2.745823860168457,
      "step": 1578
    },
    {
      "epoch": 0.3609142857142857,
      "grad_norm": 1.4969696998596191,
      "learning_rate": 0.0005,
      "loss": 3.6737475395202637,
      "step": 1579
    },
    {
      "epoch": 0.36114285714285715,
      "grad_norm": 1.2296594381332397,
      "learning_rate": 0.0005,
      "loss": 2.944016933441162,
      "step": 1580
    },
    {
      "epoch": 0.36137142857142857,
      "grad_norm": 1.1259784698486328,
      "learning_rate": 0.0005,
      "loss": 3.2017159461975098,
      "step": 1581
    },
    {
      "epoch": 0.3616,
      "grad_norm": 1.0051127672195435,
      "learning_rate": 0.0005,
      "loss": 2.932508945465088,
      "step": 1582
    },
    {
      "epoch": 0.36182857142857144,
      "grad_norm": 0.9591189622879028,
      "learning_rate": 0.0005,
      "loss": 2.845893383026123,
      "step": 1583
    },
    {
      "epoch": 0.36205714285714286,
      "grad_norm": 1.1572163105010986,
      "learning_rate": 0.0005,
      "loss": 3.0753674507141113,
      "step": 1584
    },
    {
      "epoch": 0.36228571428571427,
      "grad_norm": 1.5716701745986938,
      "learning_rate": 0.0005,
      "loss": 2.909480333328247,
      "step": 1585
    },
    {
      "epoch": 0.36251428571428573,
      "grad_norm": 1.3712736368179321,
      "learning_rate": 0.0005,
      "loss": 2.8982462882995605,
      "step": 1586
    },
    {
      "epoch": 0.36274285714285714,
      "grad_norm": 1.2068819999694824,
      "learning_rate": 0.0005,
      "loss": 3.3139052391052246,
      "step": 1587
    },
    {
      "epoch": 0.36297142857142856,
      "grad_norm": 0.9945864677429199,
      "learning_rate": 0.0005,
      "loss": 3.251258373260498,
      "step": 1588
    },
    {
      "epoch": 0.3632,
      "grad_norm": 1.2941992282867432,
      "learning_rate": 0.0005,
      "loss": 3.4359936714172363,
      "step": 1589
    },
    {
      "epoch": 0.36342857142857143,
      "grad_norm": 1.3099076747894287,
      "learning_rate": 0.0005,
      "loss": 3.2023258209228516,
      "step": 1590
    },
    {
      "epoch": 0.36365714285714285,
      "grad_norm": 1.0534359216690063,
      "learning_rate": 0.0005,
      "loss": 2.9697465896606445,
      "step": 1591
    },
    {
      "epoch": 0.3638857142857143,
      "grad_norm": 1.1097463369369507,
      "learning_rate": 0.0005,
      "loss": 2.7654013633728027,
      "step": 1592
    },
    {
      "epoch": 0.3641142857142857,
      "grad_norm": 1.5731333494186401,
      "learning_rate": 0.0005,
      "loss": 3.429358959197998,
      "step": 1593
    },
    {
      "epoch": 0.36434285714285713,
      "grad_norm": 1.298248529434204,
      "learning_rate": 0.0005,
      "loss": 3.0391836166381836,
      "step": 1594
    },
    {
      "epoch": 0.36457142857142855,
      "grad_norm": 1.3441133499145508,
      "learning_rate": 0.0005,
      "loss": 3.0404975414276123,
      "step": 1595
    },
    {
      "epoch": 0.3648,
      "grad_norm": 1.145740032196045,
      "learning_rate": 0.0005,
      "loss": 3.2136411666870117,
      "step": 1596
    },
    {
      "epoch": 0.3650285714285714,
      "grad_norm": 1.1504874229431152,
      "learning_rate": 0.0005,
      "loss": 3.3539905548095703,
      "step": 1597
    },
    {
      "epoch": 0.36525714285714284,
      "grad_norm": 1.128562569618225,
      "learning_rate": 0.0005,
      "loss": 3.0758888721466064,
      "step": 1598
    },
    {
      "epoch": 0.3654857142857143,
      "grad_norm": 1.4830386638641357,
      "learning_rate": 0.0005,
      "loss": 3.5342352390289307,
      "step": 1599
    },
    {
      "epoch": 0.3657142857142857,
      "grad_norm": 1.0829813480377197,
      "learning_rate": 0.0005,
      "loss": 3.2529749870300293,
      "step": 1600
    },
    {
      "epoch": 0.3659428571428571,
      "grad_norm": 1.2452634572982788,
      "learning_rate": 0.0005,
      "loss": 3.149690628051758,
      "step": 1601
    },
    {
      "epoch": 0.3661714285714286,
      "grad_norm": 1.3452415466308594,
      "learning_rate": 0.0005,
      "loss": 3.060279369354248,
      "step": 1602
    },
    {
      "epoch": 0.3664,
      "grad_norm": 1.0004597902297974,
      "learning_rate": 0.0005,
      "loss": 2.9827160835266113,
      "step": 1603
    },
    {
      "epoch": 0.3666285714285714,
      "grad_norm": 1.2606256008148193,
      "learning_rate": 0.0005,
      "loss": 3.3106260299682617,
      "step": 1604
    },
    {
      "epoch": 0.3668571428571429,
      "grad_norm": 0.9438570737838745,
      "learning_rate": 0.0005,
      "loss": 3.3894660472869873,
      "step": 1605
    },
    {
      "epoch": 0.3670857142857143,
      "grad_norm": 2.377408981323242,
      "learning_rate": 0.0005,
      "loss": 3.121826648712158,
      "step": 1606
    },
    {
      "epoch": 0.3673142857142857,
      "grad_norm": 1.190051794052124,
      "learning_rate": 0.0005,
      "loss": 3.0879921913146973,
      "step": 1607
    },
    {
      "epoch": 0.3675428571428571,
      "grad_norm": 1.287184238433838,
      "learning_rate": 0.0005,
      "loss": 3.3298120498657227,
      "step": 1608
    },
    {
      "epoch": 0.3677714285714286,
      "grad_norm": 1.3993229866027832,
      "learning_rate": 0.0005,
      "loss": 2.9950618743896484,
      "step": 1609
    },
    {
      "epoch": 0.368,
      "grad_norm": 1.169466257095337,
      "learning_rate": 0.0005,
      "loss": 2.9305405616760254,
      "step": 1610
    },
    {
      "epoch": 0.3682285714285714,
      "grad_norm": 1.1312934160232544,
      "learning_rate": 0.0005,
      "loss": 3.3461697101593018,
      "step": 1611
    },
    {
      "epoch": 0.36845714285714287,
      "grad_norm": 1.4305102825164795,
      "learning_rate": 0.0005,
      "loss": 3.1821539402008057,
      "step": 1612
    },
    {
      "epoch": 0.3686857142857143,
      "grad_norm": 1.129640817642212,
      "learning_rate": 0.0005,
      "loss": 3.264157772064209,
      "step": 1613
    },
    {
      "epoch": 0.3689142857142857,
      "grad_norm": 1.0432840585708618,
      "learning_rate": 0.0005,
      "loss": 3.34159517288208,
      "step": 1614
    },
    {
      "epoch": 0.36914285714285716,
      "grad_norm": 1.1601966619491577,
      "learning_rate": 0.0005,
      "loss": 3.269955635070801,
      "step": 1615
    },
    {
      "epoch": 0.3693714285714286,
      "grad_norm": 1.3728058338165283,
      "learning_rate": 0.0005,
      "loss": 2.985995292663574,
      "step": 1616
    },
    {
      "epoch": 0.3696,
      "grad_norm": 1.3169454336166382,
      "learning_rate": 0.0005,
      "loss": 2.956936836242676,
      "step": 1617
    },
    {
      "epoch": 0.36982857142857145,
      "grad_norm": 1.4092966318130493,
      "learning_rate": 0.0005,
      "loss": 3.1738293170928955,
      "step": 1618
    },
    {
      "epoch": 0.37005714285714286,
      "grad_norm": 1.531966209411621,
      "learning_rate": 0.0005,
      "loss": 2.798030376434326,
      "step": 1619
    },
    {
      "epoch": 0.3702857142857143,
      "grad_norm": 1.80563485622406,
      "learning_rate": 0.0005,
      "loss": 3.6457486152648926,
      "step": 1620
    },
    {
      "epoch": 0.37051428571428574,
      "grad_norm": 1.2786661386489868,
      "learning_rate": 0.0005,
      "loss": 3.310739278793335,
      "step": 1621
    },
    {
      "epoch": 0.37074285714285715,
      "grad_norm": 1.4274548292160034,
      "learning_rate": 0.0005,
      "loss": 3.0928993225097656,
      "step": 1622
    },
    {
      "epoch": 0.37097142857142856,
      "grad_norm": 0.9655894041061401,
      "learning_rate": 0.0005,
      "loss": 3.1363091468811035,
      "step": 1623
    },
    {
      "epoch": 0.3712,
      "grad_norm": 1.8133213520050049,
      "learning_rate": 0.0005,
      "loss": 3.10569429397583,
      "step": 1624
    },
    {
      "epoch": 0.37142857142857144,
      "grad_norm": 12.14417839050293,
      "learning_rate": 0.0005,
      "loss": 3.3236587047576904,
      "step": 1625
    },
    {
      "epoch": 0.37165714285714285,
      "grad_norm": 1.8388466835021973,
      "learning_rate": 0.0005,
      "loss": 3.154751777648926,
      "step": 1626
    },
    {
      "epoch": 0.37188571428571426,
      "grad_norm": 1.0915786027908325,
      "learning_rate": 0.0005,
      "loss": 2.9110355377197266,
      "step": 1627
    },
    {
      "epoch": 0.37211428571428573,
      "grad_norm": 1.0916337966918945,
      "learning_rate": 0.0005,
      "loss": 2.7645764350891113,
      "step": 1628
    },
    {
      "epoch": 0.37234285714285714,
      "grad_norm": 1.4303168058395386,
      "learning_rate": 0.0005,
      "loss": 2.863840103149414,
      "step": 1629
    },
    {
      "epoch": 0.37257142857142855,
      "grad_norm": 1.153329610824585,
      "learning_rate": 0.0005,
      "loss": 2.745715618133545,
      "step": 1630
    },
    {
      "epoch": 0.3728,
      "grad_norm": 1.3092950582504272,
      "learning_rate": 0.0005,
      "loss": 3.174400568008423,
      "step": 1631
    },
    {
      "epoch": 0.37302857142857143,
      "grad_norm": 1.3835136890411377,
      "learning_rate": 0.0005,
      "loss": 3.302156448364258,
      "step": 1632
    },
    {
      "epoch": 0.37325714285714284,
      "grad_norm": 1.1254572868347168,
      "learning_rate": 0.0005,
      "loss": 2.8945069313049316,
      "step": 1633
    },
    {
      "epoch": 0.3734857142857143,
      "grad_norm": 1.6931737661361694,
      "learning_rate": 0.0005,
      "loss": 2.7843339443206787,
      "step": 1634
    },
    {
      "epoch": 0.3737142857142857,
      "grad_norm": 1.0387784242630005,
      "learning_rate": 0.0005,
      "loss": 3.0659613609313965,
      "step": 1635
    },
    {
      "epoch": 0.37394285714285713,
      "grad_norm": 1.508204698562622,
      "learning_rate": 0.0005,
      "loss": 3.337646007537842,
      "step": 1636
    },
    {
      "epoch": 0.37417142857142854,
      "grad_norm": 1.3062958717346191,
      "learning_rate": 0.0005,
      "loss": 3.3714661598205566,
      "step": 1637
    },
    {
      "epoch": 0.3744,
      "grad_norm": 1.1202343702316284,
      "learning_rate": 0.0005,
      "loss": 3.0600523948669434,
      "step": 1638
    },
    {
      "epoch": 0.3746285714285714,
      "grad_norm": 1.2986711263656616,
      "learning_rate": 0.0005,
      "loss": 2.8411903381347656,
      "step": 1639
    },
    {
      "epoch": 0.37485714285714283,
      "grad_norm": 1.0932444334030151,
      "learning_rate": 0.0005,
      "loss": 3.1233723163604736,
      "step": 1640
    },
    {
      "epoch": 0.3750857142857143,
      "grad_norm": 1.1123546361923218,
      "learning_rate": 0.0005,
      "loss": 3.003856658935547,
      "step": 1641
    },
    {
      "epoch": 0.3753142857142857,
      "grad_norm": 1.0878180265426636,
      "learning_rate": 0.0005,
      "loss": 3.3568179607391357,
      "step": 1642
    },
    {
      "epoch": 0.3755428571428571,
      "grad_norm": 1.054125428199768,
      "learning_rate": 0.0005,
      "loss": 3.0374817848205566,
      "step": 1643
    },
    {
      "epoch": 0.3757714285714286,
      "grad_norm": 1.2352279424667358,
      "learning_rate": 0.0005,
      "loss": 3.5240206718444824,
      "step": 1644
    },
    {
      "epoch": 0.376,
      "grad_norm": 1.0706579685211182,
      "learning_rate": 0.0005,
      "loss": 3.230964422225952,
      "step": 1645
    },
    {
      "epoch": 0.3762285714285714,
      "grad_norm": 1.0672155618667603,
      "learning_rate": 0.0005,
      "loss": 3.113941192626953,
      "step": 1646
    },
    {
      "epoch": 0.3764571428571429,
      "grad_norm": 1.5107243061065674,
      "learning_rate": 0.0005,
      "loss": 3.028250217437744,
      "step": 1647
    },
    {
      "epoch": 0.3766857142857143,
      "grad_norm": 1.5323224067687988,
      "learning_rate": 0.0005,
      "loss": 3.331564426422119,
      "step": 1648
    },
    {
      "epoch": 0.3769142857142857,
      "grad_norm": 1.2645400762557983,
      "learning_rate": 0.0005,
      "loss": 2.646963596343994,
      "step": 1649
    },
    {
      "epoch": 0.37714285714285717,
      "grad_norm": 1.174921989440918,
      "learning_rate": 0.0005,
      "loss": 2.83797550201416,
      "step": 1650
    },
    {
      "epoch": 0.3773714285714286,
      "grad_norm": 0.9458812475204468,
      "learning_rate": 0.0005,
      "loss": 3.1252288818359375,
      "step": 1651
    },
    {
      "epoch": 0.3776,
      "grad_norm": 1.0655556917190552,
      "learning_rate": 0.0005,
      "loss": 2.9135210514068604,
      "step": 1652
    },
    {
      "epoch": 0.3778285714285714,
      "grad_norm": 1.381622552871704,
      "learning_rate": 0.0005,
      "loss": 3.0072641372680664,
      "step": 1653
    },
    {
      "epoch": 0.37805714285714287,
      "grad_norm": 1.2820175886154175,
      "learning_rate": 0.0005,
      "loss": 3.0381994247436523,
      "step": 1654
    },
    {
      "epoch": 0.3782857142857143,
      "grad_norm": 1.3524081707000732,
      "learning_rate": 0.0005,
      "loss": 3.0074241161346436,
      "step": 1655
    },
    {
      "epoch": 0.3785142857142857,
      "grad_norm": 1.411549687385559,
      "learning_rate": 0.0005,
      "loss": 2.9779341220855713,
      "step": 1656
    },
    {
      "epoch": 0.37874285714285716,
      "grad_norm": 1.3082727193832397,
      "learning_rate": 0.0005,
      "loss": 2.9284605979919434,
      "step": 1657
    },
    {
      "epoch": 0.37897142857142857,
      "grad_norm": 1.211290717124939,
      "learning_rate": 0.0005,
      "loss": 3.175473213195801,
      "step": 1658
    },
    {
      "epoch": 0.3792,
      "grad_norm": 1.835522174835205,
      "learning_rate": 0.0005,
      "loss": 3.043828010559082,
      "step": 1659
    },
    {
      "epoch": 0.37942857142857145,
      "grad_norm": 1.2949894666671753,
      "learning_rate": 0.0005,
      "loss": 2.9917609691619873,
      "step": 1660
    },
    {
      "epoch": 0.37965714285714286,
      "grad_norm": 1.1241908073425293,
      "learning_rate": 0.0005,
      "loss": 3.343291759490967,
      "step": 1661
    },
    {
      "epoch": 0.37988571428571427,
      "grad_norm": 1.0646331310272217,
      "learning_rate": 0.0005,
      "loss": 2.5915980339050293,
      "step": 1662
    },
    {
      "epoch": 0.38011428571428574,
      "grad_norm": 1.0535091161727905,
      "learning_rate": 0.0005,
      "loss": 2.789741039276123,
      "step": 1663
    },
    {
      "epoch": 0.38034285714285715,
      "grad_norm": 1.2794826030731201,
      "learning_rate": 0.0005,
      "loss": 3.53895902633667,
      "step": 1664
    },
    {
      "epoch": 0.38057142857142856,
      "grad_norm": 0.979448139667511,
      "learning_rate": 0.0005,
      "loss": 3.1220145225524902,
      "step": 1665
    },
    {
      "epoch": 0.3808,
      "grad_norm": 1.0960402488708496,
      "learning_rate": 0.0005,
      "loss": 3.178490400314331,
      "step": 1666
    },
    {
      "epoch": 0.38102857142857144,
      "grad_norm": 1.3108139038085938,
      "learning_rate": 0.0005,
      "loss": 3.1278016567230225,
      "step": 1667
    },
    {
      "epoch": 0.38125714285714285,
      "grad_norm": 1.179402470588684,
      "learning_rate": 0.0005,
      "loss": 2.979114532470703,
      "step": 1668
    },
    {
      "epoch": 0.38148571428571426,
      "grad_norm": 1.1703797578811646,
      "learning_rate": 0.0005,
      "loss": 3.140625476837158,
      "step": 1669
    },
    {
      "epoch": 0.38171428571428573,
      "grad_norm": 1.5643888711929321,
      "learning_rate": 0.0005,
      "loss": 2.942835807800293,
      "step": 1670
    },
    {
      "epoch": 0.38194285714285714,
      "grad_norm": 1.3202992677688599,
      "learning_rate": 0.0005,
      "loss": 3.3150877952575684,
      "step": 1671
    },
    {
      "epoch": 0.38217142857142855,
      "grad_norm": 2.297172784805298,
      "learning_rate": 0.0005,
      "loss": 3.437386989593506,
      "step": 1672
    },
    {
      "epoch": 0.3824,
      "grad_norm": 1.0966095924377441,
      "learning_rate": 0.0005,
      "loss": 3.023466110229492,
      "step": 1673
    },
    {
      "epoch": 0.38262857142857143,
      "grad_norm": 1.1001068353652954,
      "learning_rate": 0.0005,
      "loss": 2.9640159606933594,
      "step": 1674
    },
    {
      "epoch": 0.38285714285714284,
      "grad_norm": 1.4110063314437866,
      "learning_rate": 0.0005,
      "loss": 3.1446423530578613,
      "step": 1675
    },
    {
      "epoch": 0.3830857142857143,
      "grad_norm": 1.0005362033843994,
      "learning_rate": 0.0005,
      "loss": 3.2108442783355713,
      "step": 1676
    },
    {
      "epoch": 0.3833142857142857,
      "grad_norm": 1.0850847959518433,
      "learning_rate": 0.0005,
      "loss": 2.964341163635254,
      "step": 1677
    },
    {
      "epoch": 0.38354285714285713,
      "grad_norm": 1.0853465795516968,
      "learning_rate": 0.0005,
      "loss": 3.313673973083496,
      "step": 1678
    },
    {
      "epoch": 0.3837714285714286,
      "grad_norm": 1.2132513523101807,
      "learning_rate": 0.0005,
      "loss": 3.0591046810150146,
      "step": 1679
    },
    {
      "epoch": 0.384,
      "grad_norm": 1.31522536277771,
      "learning_rate": 0.0005,
      "loss": 3.2294342517852783,
      "step": 1680
    },
    {
      "epoch": 0.3842285714285714,
      "grad_norm": 1.7772595882415771,
      "learning_rate": 0.0005,
      "loss": 2.6371939182281494,
      "step": 1681
    },
    {
      "epoch": 0.38445714285714283,
      "grad_norm": 1.0648738145828247,
      "learning_rate": 0.0005,
      "loss": 3.256066083908081,
      "step": 1682
    },
    {
      "epoch": 0.3846857142857143,
      "grad_norm": 1.197371482849121,
      "learning_rate": 0.0005,
      "loss": 2.6888082027435303,
      "step": 1683
    },
    {
      "epoch": 0.3849142857142857,
      "grad_norm": 0.9239106774330139,
      "learning_rate": 0.0005,
      "loss": 3.0954766273498535,
      "step": 1684
    },
    {
      "epoch": 0.3851428571428571,
      "grad_norm": 1.2348413467407227,
      "learning_rate": 0.0005,
      "loss": 2.618298292160034,
      "step": 1685
    },
    {
      "epoch": 0.3853714285714286,
      "grad_norm": 1.407625436782837,
      "learning_rate": 0.0005,
      "loss": 2.8624284267425537,
      "step": 1686
    },
    {
      "epoch": 0.3856,
      "grad_norm": 1.2404918670654297,
      "learning_rate": 0.0005,
      "loss": 3.144500255584717,
      "step": 1687
    },
    {
      "epoch": 0.3858285714285714,
      "grad_norm": 1.2081140279769897,
      "learning_rate": 0.0005,
      "loss": 2.558907985687256,
      "step": 1688
    },
    {
      "epoch": 0.3860571428571429,
      "grad_norm": 1.3027929067611694,
      "learning_rate": 0.0005,
      "loss": 3.0224342346191406,
      "step": 1689
    },
    {
      "epoch": 0.3862857142857143,
      "grad_norm": 1.058828353881836,
      "learning_rate": 0.0005,
      "loss": 2.7136964797973633,
      "step": 1690
    },
    {
      "epoch": 0.3865142857142857,
      "grad_norm": 1.2552825212478638,
      "learning_rate": 0.0005,
      "loss": 2.8583450317382812,
      "step": 1691
    },
    {
      "epoch": 0.38674285714285717,
      "grad_norm": 1.8285820484161377,
      "learning_rate": 0.0005,
      "loss": 3.0686464309692383,
      "step": 1692
    },
    {
      "epoch": 0.3869714285714286,
      "grad_norm": 1.7247138023376465,
      "learning_rate": 0.0005,
      "loss": 3.0366861820220947,
      "step": 1693
    },
    {
      "epoch": 0.3872,
      "grad_norm": 1.2990522384643555,
      "learning_rate": 0.0005,
      "loss": 3.132293224334717,
      "step": 1694
    },
    {
      "epoch": 0.38742857142857146,
      "grad_norm": 1.2886770963668823,
      "learning_rate": 0.0005,
      "loss": 3.0126614570617676,
      "step": 1695
    },
    {
      "epoch": 0.38765714285714287,
      "grad_norm": 1.2396876811981201,
      "learning_rate": 0.0005,
      "loss": 3.244954824447632,
      "step": 1696
    },
    {
      "epoch": 0.3878857142857143,
      "grad_norm": 1.8922804594039917,
      "learning_rate": 0.0005,
      "loss": 2.8171792030334473,
      "step": 1697
    },
    {
      "epoch": 0.3881142857142857,
      "grad_norm": 1.35637629032135,
      "learning_rate": 0.0005,
      "loss": 3.0209298133850098,
      "step": 1698
    },
    {
      "epoch": 0.38834285714285716,
      "grad_norm": 1.4153265953063965,
      "learning_rate": 0.0005,
      "loss": 3.51302433013916,
      "step": 1699
    },
    {
      "epoch": 0.38857142857142857,
      "grad_norm": 1.2974916696548462,
      "learning_rate": 0.0005,
      "loss": 3.156860589981079,
      "step": 1700
    },
    {
      "epoch": 0.3888,
      "grad_norm": 1.0029048919677734,
      "learning_rate": 0.0005,
      "loss": 3.1026744842529297,
      "step": 1701
    },
    {
      "epoch": 0.38902857142857145,
      "grad_norm": 1.2353992462158203,
      "learning_rate": 0.0005,
      "loss": 3.234544277191162,
      "step": 1702
    },
    {
      "epoch": 0.38925714285714286,
      "grad_norm": 1.014015555381775,
      "learning_rate": 0.0005,
      "loss": 3.320169448852539,
      "step": 1703
    },
    {
      "epoch": 0.38948571428571427,
      "grad_norm": 1.554911732673645,
      "learning_rate": 0.0005,
      "loss": 2.9269144535064697,
      "step": 1704
    },
    {
      "epoch": 0.38971428571428574,
      "grad_norm": 1.3975452184677124,
      "learning_rate": 0.0005,
      "loss": 2.8722779750823975,
      "step": 1705
    },
    {
      "epoch": 0.38994285714285715,
      "grad_norm": 1.0101683139801025,
      "learning_rate": 0.0005,
      "loss": 2.927372932434082,
      "step": 1706
    },
    {
      "epoch": 0.39017142857142856,
      "grad_norm": 1.145767092704773,
      "learning_rate": 0.0005,
      "loss": 2.9479870796203613,
      "step": 1707
    },
    {
      "epoch": 0.3904,
      "grad_norm": 1.1972402334213257,
      "learning_rate": 0.0005,
      "loss": 3.111894130706787,
      "step": 1708
    },
    {
      "epoch": 0.39062857142857144,
      "grad_norm": 1.049681305885315,
      "learning_rate": 0.0005,
      "loss": 3.149815797805786,
      "step": 1709
    },
    {
      "epoch": 0.39085714285714285,
      "grad_norm": 1.4782707691192627,
      "learning_rate": 0.0005,
      "loss": 2.93636417388916,
      "step": 1710
    },
    {
      "epoch": 0.39108571428571426,
      "grad_norm": 0.8944515585899353,
      "learning_rate": 0.0005,
      "loss": 3.0989561080932617,
      "step": 1711
    },
    {
      "epoch": 0.3913142857142857,
      "grad_norm": 1.1621410846710205,
      "learning_rate": 0.0005,
      "loss": 3.174811601638794,
      "step": 1712
    },
    {
      "epoch": 0.39154285714285714,
      "grad_norm": 1.2289236783981323,
      "learning_rate": 0.0005,
      "loss": 3.0032660961151123,
      "step": 1713
    },
    {
      "epoch": 0.39177142857142855,
      "grad_norm": 1.0858454704284668,
      "learning_rate": 0.0005,
      "loss": 3.124305486679077,
      "step": 1714
    },
    {
      "epoch": 0.392,
      "grad_norm": 1.1166445016860962,
      "learning_rate": 0.0005,
      "loss": 2.8775768280029297,
      "step": 1715
    },
    {
      "epoch": 0.3922285714285714,
      "grad_norm": 1.1662284135818481,
      "learning_rate": 0.0005,
      "loss": 2.8913931846618652,
      "step": 1716
    },
    {
      "epoch": 0.39245714285714284,
      "grad_norm": 1.2255845069885254,
      "learning_rate": 0.0005,
      "loss": 2.9242470264434814,
      "step": 1717
    },
    {
      "epoch": 0.3926857142857143,
      "grad_norm": 1.1584420204162598,
      "learning_rate": 0.0005,
      "loss": 2.9100399017333984,
      "step": 1718
    },
    {
      "epoch": 0.3929142857142857,
      "grad_norm": 1.1021428108215332,
      "learning_rate": 0.0005,
      "loss": 3.2440600395202637,
      "step": 1719
    },
    {
      "epoch": 0.3931428571428571,
      "grad_norm": 1.5407698154449463,
      "learning_rate": 0.0005,
      "loss": 3.019806146621704,
      "step": 1720
    },
    {
      "epoch": 0.3933714285714286,
      "grad_norm": 1.152424931526184,
      "learning_rate": 0.0005,
      "loss": 3.1570801734924316,
      "step": 1721
    },
    {
      "epoch": 0.3936,
      "grad_norm": 1.1857545375823975,
      "learning_rate": 0.0005,
      "loss": 3.223879814147949,
      "step": 1722
    },
    {
      "epoch": 0.3938285714285714,
      "grad_norm": 1.071256399154663,
      "learning_rate": 0.0005,
      "loss": 2.929189682006836,
      "step": 1723
    },
    {
      "epoch": 0.3940571428571429,
      "grad_norm": 1.1882789134979248,
      "learning_rate": 0.0005,
      "loss": 2.7690505981445312,
      "step": 1724
    },
    {
      "epoch": 0.3942857142857143,
      "grad_norm": 6.6082963943481445,
      "learning_rate": 0.0005,
      "loss": 2.8537068367004395,
      "step": 1725
    },
    {
      "epoch": 0.3945142857142857,
      "grad_norm": 1.1095335483551025,
      "learning_rate": 0.0005,
      "loss": 2.948970317840576,
      "step": 1726
    },
    {
      "epoch": 0.3947428571428571,
      "grad_norm": 1.2800192832946777,
      "learning_rate": 0.0005,
      "loss": 2.7608838081359863,
      "step": 1727
    },
    {
      "epoch": 0.3949714285714286,
      "grad_norm": 1.2319830656051636,
      "learning_rate": 0.0005,
      "loss": 3.5316505432128906,
      "step": 1728
    },
    {
      "epoch": 0.3952,
      "grad_norm": 1.27193284034729,
      "learning_rate": 0.0005,
      "loss": 3.1911158561706543,
      "step": 1729
    },
    {
      "epoch": 0.3954285714285714,
      "grad_norm": 1.5129764080047607,
      "learning_rate": 0.0005,
      "loss": 2.842101812362671,
      "step": 1730
    },
    {
      "epoch": 0.3956571428571429,
      "grad_norm": 1.570762038230896,
      "learning_rate": 0.0005,
      "loss": 2.7052016258239746,
      "step": 1731
    },
    {
      "epoch": 0.3958857142857143,
      "grad_norm": 1.4352285861968994,
      "learning_rate": 0.0005,
      "loss": 3.0861239433288574,
      "step": 1732
    },
    {
      "epoch": 0.3961142857142857,
      "grad_norm": 1.0750855207443237,
      "learning_rate": 0.0005,
      "loss": 2.965676784515381,
      "step": 1733
    },
    {
      "epoch": 0.39634285714285716,
      "grad_norm": 1.534285545349121,
      "learning_rate": 0.0005,
      "loss": 3.1779050827026367,
      "step": 1734
    },
    {
      "epoch": 0.3965714285714286,
      "grad_norm": 1.1617469787597656,
      "learning_rate": 0.0005,
      "loss": 2.951464891433716,
      "step": 1735
    },
    {
      "epoch": 0.3968,
      "grad_norm": 1.1118855476379395,
      "learning_rate": 0.0005,
      "loss": 3.2747035026550293,
      "step": 1736
    },
    {
      "epoch": 0.39702857142857145,
      "grad_norm": 5.216573715209961,
      "learning_rate": 0.0005,
      "loss": 3.2005538940429688,
      "step": 1737
    },
    {
      "epoch": 0.39725714285714286,
      "grad_norm": 1.9995276927947998,
      "learning_rate": 0.0005,
      "loss": 3.0297703742980957,
      "step": 1738
    },
    {
      "epoch": 0.3974857142857143,
      "grad_norm": 1.8471225500106812,
      "learning_rate": 0.0005,
      "loss": 2.909975290298462,
      "step": 1739
    },
    {
      "epoch": 0.3977142857142857,
      "grad_norm": 1.2141897678375244,
      "learning_rate": 0.0005,
      "loss": 3.149533748626709,
      "step": 1740
    },
    {
      "epoch": 0.39794285714285715,
      "grad_norm": 1.1219862699508667,
      "learning_rate": 0.0005,
      "loss": 3.0785069465637207,
      "step": 1741
    },
    {
      "epoch": 0.39817142857142857,
      "grad_norm": 1.586148977279663,
      "learning_rate": 0.0005,
      "loss": 3.1865639686584473,
      "step": 1742
    },
    {
      "epoch": 0.3984,
      "grad_norm": 1.2140264511108398,
      "learning_rate": 0.0005,
      "loss": 2.9693753719329834,
      "step": 1743
    },
    {
      "epoch": 0.39862857142857144,
      "grad_norm": 1.260533094406128,
      "learning_rate": 0.0005,
      "loss": 3.422943592071533,
      "step": 1744
    },
    {
      "epoch": 0.39885714285714285,
      "grad_norm": 1.0628198385238647,
      "learning_rate": 0.0005,
      "loss": 2.855078935623169,
      "step": 1745
    },
    {
      "epoch": 0.39908571428571427,
      "grad_norm": 1.4485440254211426,
      "learning_rate": 0.0005,
      "loss": 2.9156856536865234,
      "step": 1746
    },
    {
      "epoch": 0.39931428571428573,
      "grad_norm": 1.3115261793136597,
      "learning_rate": 0.0005,
      "loss": 3.1338143348693848,
      "step": 1747
    },
    {
      "epoch": 0.39954285714285714,
      "grad_norm": 1.8493525981903076,
      "learning_rate": 0.0005,
      "loss": 3.344681739807129,
      "step": 1748
    },
    {
      "epoch": 0.39977142857142856,
      "grad_norm": 1.259926438331604,
      "learning_rate": 0.0005,
      "loss": 3.263110637664795,
      "step": 1749
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.4503288269042969,
      "learning_rate": 0.0005,
      "loss": 3.0295329093933105,
      "step": 1750
    },
    {
      "epoch": 0.40022857142857143,
      "grad_norm": 1.2086578607559204,
      "learning_rate": 0.0005,
      "loss": 3.0141258239746094,
      "step": 1751
    },
    {
      "epoch": 0.40045714285714284,
      "grad_norm": 1.0261542797088623,
      "learning_rate": 0.0005,
      "loss": 2.8167927265167236,
      "step": 1752
    },
    {
      "epoch": 0.4006857142857143,
      "grad_norm": 1.3956902027130127,
      "learning_rate": 0.0005,
      "loss": 2.8794546127319336,
      "step": 1753
    },
    {
      "epoch": 0.4009142857142857,
      "grad_norm": 1.2076886892318726,
      "learning_rate": 0.0005,
      "loss": 3.2045679092407227,
      "step": 1754
    },
    {
      "epoch": 0.40114285714285713,
      "grad_norm": 1.1552631855010986,
      "learning_rate": 0.0005,
      "loss": 3.159529209136963,
      "step": 1755
    },
    {
      "epoch": 0.40137142857142855,
      "grad_norm": 0.9207103252410889,
      "learning_rate": 0.0005,
      "loss": 3.4833197593688965,
      "step": 1756
    },
    {
      "epoch": 0.4016,
      "grad_norm": 1.0105464458465576,
      "learning_rate": 0.0005,
      "loss": 3.333667755126953,
      "step": 1757
    },
    {
      "epoch": 0.4018285714285714,
      "grad_norm": 1.2520822286605835,
      "learning_rate": 0.0005,
      "loss": 3.1898155212402344,
      "step": 1758
    },
    {
      "epoch": 0.40205714285714284,
      "grad_norm": 1.3654602766036987,
      "learning_rate": 0.0005,
      "loss": 3.148455858230591,
      "step": 1759
    },
    {
      "epoch": 0.4022857142857143,
      "grad_norm": 1.463035225868225,
      "learning_rate": 0.0005,
      "loss": 3.156409740447998,
      "step": 1760
    },
    {
      "epoch": 0.4025142857142857,
      "grad_norm": 1.3937323093414307,
      "learning_rate": 0.0005,
      "loss": 2.7118735313415527,
      "step": 1761
    },
    {
      "epoch": 0.4027428571428571,
      "grad_norm": 1.4327092170715332,
      "learning_rate": 0.0005,
      "loss": 3.5504698753356934,
      "step": 1762
    },
    {
      "epoch": 0.4029714285714286,
      "grad_norm": 1.2612189054489136,
      "learning_rate": 0.0005,
      "loss": 3.0370688438415527,
      "step": 1763
    },
    {
      "epoch": 0.4032,
      "grad_norm": 1.3623486757278442,
      "learning_rate": 0.0005,
      "loss": 2.935667037963867,
      "step": 1764
    },
    {
      "epoch": 0.4034285714285714,
      "grad_norm": 1.223456621170044,
      "learning_rate": 0.0005,
      "loss": 3.315638780593872,
      "step": 1765
    },
    {
      "epoch": 0.4036571428571429,
      "grad_norm": 1.6219834089279175,
      "learning_rate": 0.0005,
      "loss": 2.962202548980713,
      "step": 1766
    },
    {
      "epoch": 0.4038857142857143,
      "grad_norm": 1.4160561561584473,
      "learning_rate": 0.0005,
      "loss": 3.062069892883301,
      "step": 1767
    },
    {
      "epoch": 0.4041142857142857,
      "grad_norm": 1.5722668170928955,
      "learning_rate": 0.0005,
      "loss": 3.0364978313446045,
      "step": 1768
    },
    {
      "epoch": 0.40434285714285717,
      "grad_norm": 1.269467830657959,
      "learning_rate": 0.0005,
      "loss": 2.9930553436279297,
      "step": 1769
    },
    {
      "epoch": 0.4045714285714286,
      "grad_norm": 1.4530702829360962,
      "learning_rate": 0.0005,
      "loss": 3.1865556240081787,
      "step": 1770
    },
    {
      "epoch": 0.4048,
      "grad_norm": 1.5680711269378662,
      "learning_rate": 0.0005,
      "loss": 2.9538607597351074,
      "step": 1771
    },
    {
      "epoch": 0.4050285714285714,
      "grad_norm": 1.4399300813674927,
      "learning_rate": 0.0005,
      "loss": 3.088611364364624,
      "step": 1772
    },
    {
      "epoch": 0.40525714285714287,
      "grad_norm": 1.1369348764419556,
      "learning_rate": 0.0005,
      "loss": 3.149852752685547,
      "step": 1773
    },
    {
      "epoch": 0.4054857142857143,
      "grad_norm": 1.199944257736206,
      "learning_rate": 0.0005,
      "loss": 2.8947980403900146,
      "step": 1774
    },
    {
      "epoch": 0.4057142857142857,
      "grad_norm": 1.3382129669189453,
      "learning_rate": 0.0005,
      "loss": 3.505836009979248,
      "step": 1775
    },
    {
      "epoch": 0.40594285714285716,
      "grad_norm": 1.5157274007797241,
      "learning_rate": 0.0005,
      "loss": 3.0499613285064697,
      "step": 1776
    },
    {
      "epoch": 0.40617142857142857,
      "grad_norm": 1.0923304557800293,
      "learning_rate": 0.0005,
      "loss": 2.889235019683838,
      "step": 1777
    },
    {
      "epoch": 0.4064,
      "grad_norm": 2.3148460388183594,
      "learning_rate": 0.0005,
      "loss": 3.2145934104919434,
      "step": 1778
    },
    {
      "epoch": 0.40662857142857145,
      "grad_norm": 1.2151418924331665,
      "learning_rate": 0.0005,
      "loss": 2.6305816173553467,
      "step": 1779
    },
    {
      "epoch": 0.40685714285714286,
      "grad_norm": 1.1905282735824585,
      "learning_rate": 0.0005,
      "loss": 3.416290760040283,
      "step": 1780
    },
    {
      "epoch": 0.4070857142857143,
      "grad_norm": 1.1162817478179932,
      "learning_rate": 0.0005,
      "loss": 3.0209999084472656,
      "step": 1781
    },
    {
      "epoch": 0.40731428571428574,
      "grad_norm": 1.9235219955444336,
      "learning_rate": 0.0005,
      "loss": 3.1874685287475586,
      "step": 1782
    },
    {
      "epoch": 0.40754285714285715,
      "grad_norm": 3.157085418701172,
      "learning_rate": 0.0005,
      "loss": 2.830528736114502,
      "step": 1783
    },
    {
      "epoch": 0.40777142857142856,
      "grad_norm": 1.252623438835144,
      "learning_rate": 0.0005,
      "loss": 3.3792686462402344,
      "step": 1784
    },
    {
      "epoch": 0.408,
      "grad_norm": 1.5233415365219116,
      "learning_rate": 0.0005,
      "loss": 2.9880170822143555,
      "step": 1785
    },
    {
      "epoch": 0.40822857142857144,
      "grad_norm": 1.4111274480819702,
      "learning_rate": 0.0005,
      "loss": 2.6732938289642334,
      "step": 1786
    },
    {
      "epoch": 0.40845714285714285,
      "grad_norm": 1.048660397529602,
      "learning_rate": 0.0005,
      "loss": 2.6617798805236816,
      "step": 1787
    },
    {
      "epoch": 0.40868571428571426,
      "grad_norm": 1.6608299016952515,
      "learning_rate": 0.0005,
      "loss": 2.910874843597412,
      "step": 1788
    },
    {
      "epoch": 0.40891428571428573,
      "grad_norm": 1.3511492013931274,
      "learning_rate": 0.0005,
      "loss": 2.791200637817383,
      "step": 1789
    },
    {
      "epoch": 0.40914285714285714,
      "grad_norm": 1.178115963935852,
      "learning_rate": 0.0005,
      "loss": 2.8444669246673584,
      "step": 1790
    },
    {
      "epoch": 0.40937142857142855,
      "grad_norm": 1.2646795511245728,
      "learning_rate": 0.0005,
      "loss": 3.1046581268310547,
      "step": 1791
    },
    {
      "epoch": 0.4096,
      "grad_norm": 1.148437738418579,
      "learning_rate": 0.0005,
      "loss": 3.218348264694214,
      "step": 1792
    },
    {
      "epoch": 0.40982857142857143,
      "grad_norm": 1.5994254350662231,
      "learning_rate": 0.0005,
      "loss": 3.343510866165161,
      "step": 1793
    },
    {
      "epoch": 0.41005714285714284,
      "grad_norm": 1.1239230632781982,
      "learning_rate": 0.0005,
      "loss": 2.8547911643981934,
      "step": 1794
    },
    {
      "epoch": 0.4102857142857143,
      "grad_norm": 1.157077670097351,
      "learning_rate": 0.0005,
      "loss": 2.913123846054077,
      "step": 1795
    },
    {
      "epoch": 0.4105142857142857,
      "grad_norm": 1.347409725189209,
      "learning_rate": 0.0005,
      "loss": 2.9355993270874023,
      "step": 1796
    },
    {
      "epoch": 0.41074285714285713,
      "grad_norm": 1.1880521774291992,
      "learning_rate": 0.0005,
      "loss": 3.0023107528686523,
      "step": 1797
    },
    {
      "epoch": 0.4109714285714286,
      "grad_norm": 1.230881929397583,
      "learning_rate": 0.0005,
      "loss": 3.1432347297668457,
      "step": 1798
    },
    {
      "epoch": 0.4112,
      "grad_norm": 1.0782766342163086,
      "learning_rate": 0.0005,
      "loss": 2.9478914737701416,
      "step": 1799
    },
    {
      "epoch": 0.4114285714285714,
      "grad_norm": 1.1752763986587524,
      "learning_rate": 0.0005,
      "loss": 3.023087501525879,
      "step": 1800
    },
    {
      "epoch": 0.41165714285714283,
      "grad_norm": 1.092092752456665,
      "learning_rate": 0.0005,
      "loss": 2.694397449493408,
      "step": 1801
    },
    {
      "epoch": 0.4118857142857143,
      "grad_norm": 1.3175119161605835,
      "learning_rate": 0.0005,
      "loss": 3.149780750274658,
      "step": 1802
    },
    {
      "epoch": 0.4121142857142857,
      "grad_norm": 1.1446573734283447,
      "learning_rate": 0.0005,
      "loss": 3.100247859954834,
      "step": 1803
    },
    {
      "epoch": 0.4123428571428571,
      "grad_norm": 1.1093640327453613,
      "learning_rate": 0.0005,
      "loss": 2.963050365447998,
      "step": 1804
    },
    {
      "epoch": 0.4125714285714286,
      "grad_norm": 1.121651291847229,
      "learning_rate": 0.0005,
      "loss": 3.024230480194092,
      "step": 1805
    },
    {
      "epoch": 0.4128,
      "grad_norm": 1.2688931226730347,
      "learning_rate": 0.0005,
      "loss": 2.890868902206421,
      "step": 1806
    },
    {
      "epoch": 0.4130285714285714,
      "grad_norm": 1.0478527545928955,
      "learning_rate": 0.0005,
      "loss": 3.014232635498047,
      "step": 1807
    },
    {
      "epoch": 0.4132571428571429,
      "grad_norm": 1.0305782556533813,
      "learning_rate": 0.0005,
      "loss": 3.189478635787964,
      "step": 1808
    },
    {
      "epoch": 0.4134857142857143,
      "grad_norm": 1.1256428956985474,
      "learning_rate": 0.0005,
      "loss": 2.980988025665283,
      "step": 1809
    },
    {
      "epoch": 0.4137142857142857,
      "grad_norm": 2.0242342948913574,
      "learning_rate": 0.0005,
      "loss": 2.9823098182678223,
      "step": 1810
    },
    {
      "epoch": 0.41394285714285717,
      "grad_norm": 1.4543559551239014,
      "learning_rate": 0.0005,
      "loss": 2.991901397705078,
      "step": 1811
    },
    {
      "epoch": 0.4141714285714286,
      "grad_norm": 1.1060317754745483,
      "learning_rate": 0.0005,
      "loss": 2.707639694213867,
      "step": 1812
    },
    {
      "epoch": 0.4144,
      "grad_norm": 1.6122220754623413,
      "learning_rate": 0.0005,
      "loss": 2.6999127864837646,
      "step": 1813
    },
    {
      "epoch": 0.4146285714285714,
      "grad_norm": 1.0978622436523438,
      "learning_rate": 0.0005,
      "loss": 2.9183974266052246,
      "step": 1814
    },
    {
      "epoch": 0.41485714285714287,
      "grad_norm": 1.1201250553131104,
      "learning_rate": 0.0005,
      "loss": 2.677093505859375,
      "step": 1815
    },
    {
      "epoch": 0.4150857142857143,
      "grad_norm": 1.0537807941436768,
      "learning_rate": 0.0005,
      "loss": 2.8677845001220703,
      "step": 1816
    },
    {
      "epoch": 0.4153142857142857,
      "grad_norm": 1.0604145526885986,
      "learning_rate": 0.0005,
      "loss": 2.983475685119629,
      "step": 1817
    },
    {
      "epoch": 0.41554285714285716,
      "grad_norm": 1.054480791091919,
      "learning_rate": 0.0005,
      "loss": 2.889017105102539,
      "step": 1818
    },
    {
      "epoch": 0.41577142857142857,
      "grad_norm": 1.1182867288589478,
      "learning_rate": 0.0005,
      "loss": 2.8860297203063965,
      "step": 1819
    },
    {
      "epoch": 0.416,
      "grad_norm": 1.3863955736160278,
      "learning_rate": 0.0005,
      "loss": 2.705930471420288,
      "step": 1820
    },
    {
      "epoch": 0.41622857142857145,
      "grad_norm": 0.8574258089065552,
      "learning_rate": 0.0005,
      "loss": 3.0299267768859863,
      "step": 1821
    },
    {
      "epoch": 0.41645714285714286,
      "grad_norm": 1.137060523033142,
      "learning_rate": 0.0005,
      "loss": 3.0585126876831055,
      "step": 1822
    },
    {
      "epoch": 0.41668571428571427,
      "grad_norm": 1.1366617679595947,
      "learning_rate": 0.0005,
      "loss": 2.9812278747558594,
      "step": 1823
    },
    {
      "epoch": 0.41691428571428574,
      "grad_norm": 1.075867772102356,
      "learning_rate": 0.0005,
      "loss": 2.6107029914855957,
      "step": 1824
    },
    {
      "epoch": 0.41714285714285715,
      "grad_norm": 1.065071940422058,
      "learning_rate": 0.0005,
      "loss": 2.6966381072998047,
      "step": 1825
    },
    {
      "epoch": 0.41737142857142856,
      "grad_norm": 1.0913443565368652,
      "learning_rate": 0.0005,
      "loss": 3.1216602325439453,
      "step": 1826
    },
    {
      "epoch": 0.4176,
      "grad_norm": 1.309004545211792,
      "learning_rate": 0.0005,
      "loss": 2.732935905456543,
      "step": 1827
    },
    {
      "epoch": 0.41782857142857144,
      "grad_norm": 1.309272289276123,
      "learning_rate": 0.0005,
      "loss": 2.953993797302246,
      "step": 1828
    },
    {
      "epoch": 0.41805714285714285,
      "grad_norm": 1.1545753479003906,
      "learning_rate": 0.0005,
      "loss": 2.9931740760803223,
      "step": 1829
    },
    {
      "epoch": 0.41828571428571426,
      "grad_norm": 1.0444977283477783,
      "learning_rate": 0.0005,
      "loss": 2.7652199268341064,
      "step": 1830
    },
    {
      "epoch": 0.4185142857142857,
      "grad_norm": 1.2562587261199951,
      "learning_rate": 0.0005,
      "loss": 2.9467620849609375,
      "step": 1831
    },
    {
      "epoch": 0.41874285714285714,
      "grad_norm": 1.0883138179779053,
      "learning_rate": 0.0005,
      "loss": 3.078395128250122,
      "step": 1832
    },
    {
      "epoch": 0.41897142857142855,
      "grad_norm": 1.454852819442749,
      "learning_rate": 0.0005,
      "loss": 3.1551191806793213,
      "step": 1833
    },
    {
      "epoch": 0.4192,
      "grad_norm": 1.3340015411376953,
      "learning_rate": 0.0005,
      "loss": 2.9970076084136963,
      "step": 1834
    },
    {
      "epoch": 0.41942857142857143,
      "grad_norm": 1.450952410697937,
      "learning_rate": 0.0005,
      "loss": 3.164952278137207,
      "step": 1835
    },
    {
      "epoch": 0.41965714285714284,
      "grad_norm": 1.1990302801132202,
      "learning_rate": 0.0005,
      "loss": 3.0932183265686035,
      "step": 1836
    },
    {
      "epoch": 0.4198857142857143,
      "grad_norm": 1.4230189323425293,
      "learning_rate": 0.0005,
      "loss": 2.965982437133789,
      "step": 1837
    },
    {
      "epoch": 0.4201142857142857,
      "grad_norm": 1.6704199314117432,
      "learning_rate": 0.0005,
      "loss": 2.8150200843811035,
      "step": 1838
    },
    {
      "epoch": 0.42034285714285713,
      "grad_norm": 1.0902878046035767,
      "learning_rate": 0.0005,
      "loss": 3.1233654022216797,
      "step": 1839
    },
    {
      "epoch": 0.4205714285714286,
      "grad_norm": 1.0162495374679565,
      "learning_rate": 0.0005,
      "loss": 3.3985252380371094,
      "step": 1840
    },
    {
      "epoch": 0.4208,
      "grad_norm": 1.189666509628296,
      "learning_rate": 0.0005,
      "loss": 3.0702881813049316,
      "step": 1841
    },
    {
      "epoch": 0.4210285714285714,
      "grad_norm": 1.3451069593429565,
      "learning_rate": 0.0005,
      "loss": 2.9774065017700195,
      "step": 1842
    },
    {
      "epoch": 0.42125714285714283,
      "grad_norm": 1.3532849550247192,
      "learning_rate": 0.0005,
      "loss": 2.87384033203125,
      "step": 1843
    },
    {
      "epoch": 0.4214857142857143,
      "grad_norm": 1.0249249935150146,
      "learning_rate": 0.0005,
      "loss": 3.035400867462158,
      "step": 1844
    },
    {
      "epoch": 0.4217142857142857,
      "grad_norm": 1.3466509580612183,
      "learning_rate": 0.0005,
      "loss": 3.168699264526367,
      "step": 1845
    },
    {
      "epoch": 0.4219428571428571,
      "grad_norm": 0.9453490376472473,
      "learning_rate": 0.0005,
      "loss": 3.3242602348327637,
      "step": 1846
    },
    {
      "epoch": 0.4221714285714286,
      "grad_norm": 1.1328630447387695,
      "learning_rate": 0.0005,
      "loss": 2.925830841064453,
      "step": 1847
    },
    {
      "epoch": 0.4224,
      "grad_norm": 1.7239004373550415,
      "learning_rate": 0.0005,
      "loss": 3.0048253536224365,
      "step": 1848
    },
    {
      "epoch": 0.4226285714285714,
      "grad_norm": 1.053406834602356,
      "learning_rate": 0.0005,
      "loss": 2.9354348182678223,
      "step": 1849
    },
    {
      "epoch": 0.4228571428571429,
      "grad_norm": 1.2712016105651855,
      "learning_rate": 0.0005,
      "loss": 3.329010009765625,
      "step": 1850
    },
    {
      "epoch": 0.4230857142857143,
      "grad_norm": 1.0086538791656494,
      "learning_rate": 0.0005,
      "loss": 3.0110678672790527,
      "step": 1851
    },
    {
      "epoch": 0.4233142857142857,
      "grad_norm": 1.1986957788467407,
      "learning_rate": 0.0005,
      "loss": 3.3410370349884033,
      "step": 1852
    },
    {
      "epoch": 0.42354285714285717,
      "grad_norm": 1.181583285331726,
      "learning_rate": 0.0005,
      "loss": 2.98703670501709,
      "step": 1853
    },
    {
      "epoch": 0.4237714285714286,
      "grad_norm": 1.05763578414917,
      "learning_rate": 0.0005,
      "loss": 3.126708984375,
      "step": 1854
    },
    {
      "epoch": 0.424,
      "grad_norm": 1.2524933815002441,
      "learning_rate": 0.0005,
      "loss": 2.982515335083008,
      "step": 1855
    },
    {
      "epoch": 0.42422857142857145,
      "grad_norm": 1.248884916305542,
      "learning_rate": 0.0005,
      "loss": 3.0739173889160156,
      "step": 1856
    },
    {
      "epoch": 0.42445714285714287,
      "grad_norm": 2.9614930152893066,
      "learning_rate": 0.0005,
      "loss": 3.2339091300964355,
      "step": 1857
    },
    {
      "epoch": 0.4246857142857143,
      "grad_norm": 1.1697182655334473,
      "learning_rate": 0.0005,
      "loss": 3.1837871074676514,
      "step": 1858
    },
    {
      "epoch": 0.4249142857142857,
      "grad_norm": 1.135553240776062,
      "learning_rate": 0.0005,
      "loss": 3.190829277038574,
      "step": 1859
    },
    {
      "epoch": 0.42514285714285716,
      "grad_norm": 1.294765591621399,
      "learning_rate": 0.0005,
      "loss": 2.787782669067383,
      "step": 1860
    },
    {
      "epoch": 0.42537142857142857,
      "grad_norm": 1.9617348909378052,
      "learning_rate": 0.0005,
      "loss": 2.758664131164551,
      "step": 1861
    },
    {
      "epoch": 0.4256,
      "grad_norm": 1.2838331460952759,
      "learning_rate": 0.0005,
      "loss": 2.634580612182617,
      "step": 1862
    },
    {
      "epoch": 0.42582857142857145,
      "grad_norm": 1.2738935947418213,
      "learning_rate": 0.0005,
      "loss": 2.5835089683532715,
      "step": 1863
    },
    {
      "epoch": 0.42605714285714286,
      "grad_norm": 1.0383665561676025,
      "learning_rate": 0.0005,
      "loss": 2.9609851837158203,
      "step": 1864
    },
    {
      "epoch": 0.42628571428571427,
      "grad_norm": 1.221893548965454,
      "learning_rate": 0.0005,
      "loss": 3.0497183799743652,
      "step": 1865
    },
    {
      "epoch": 0.42651428571428573,
      "grad_norm": 1.2712173461914062,
      "learning_rate": 0.0005,
      "loss": 3.144604444503784,
      "step": 1866
    },
    {
      "epoch": 0.42674285714285715,
      "grad_norm": 1.2823083400726318,
      "learning_rate": 0.0005,
      "loss": 2.8227689266204834,
      "step": 1867
    },
    {
      "epoch": 0.42697142857142856,
      "grad_norm": 1.21002197265625,
      "learning_rate": 0.0005,
      "loss": 2.976879358291626,
      "step": 1868
    },
    {
      "epoch": 0.4272,
      "grad_norm": 1.213823676109314,
      "learning_rate": 0.0005,
      "loss": 3.0016775131225586,
      "step": 1869
    },
    {
      "epoch": 0.42742857142857144,
      "grad_norm": 1.0766321420669556,
      "learning_rate": 0.0005,
      "loss": 2.9864144325256348,
      "step": 1870
    },
    {
      "epoch": 0.42765714285714285,
      "grad_norm": 2.7720282077789307,
      "learning_rate": 0.0005,
      "loss": 2.9722516536712646,
      "step": 1871
    },
    {
      "epoch": 0.42788571428571426,
      "grad_norm": 1.3047471046447754,
      "learning_rate": 0.0005,
      "loss": 2.9005064964294434,
      "step": 1872
    },
    {
      "epoch": 0.4281142857142857,
      "grad_norm": 1.1854029893875122,
      "learning_rate": 0.0005,
      "loss": 2.904083728790283,
      "step": 1873
    },
    {
      "epoch": 0.42834285714285714,
      "grad_norm": 1.0875645875930786,
      "learning_rate": 0.0005,
      "loss": 3.174131155014038,
      "step": 1874
    },
    {
      "epoch": 0.42857142857142855,
      "grad_norm": 1.2962982654571533,
      "learning_rate": 0.0005,
      "loss": 2.7292251586914062,
      "step": 1875
    },
    {
      "epoch": 0.4288,
      "grad_norm": 1.0187926292419434,
      "learning_rate": 0.0005,
      "loss": 3.058030605316162,
      "step": 1876
    },
    {
      "epoch": 0.4290285714285714,
      "grad_norm": 1.038411021232605,
      "learning_rate": 0.0005,
      "loss": 3.0618340969085693,
      "step": 1877
    },
    {
      "epoch": 0.42925714285714284,
      "grad_norm": 1.3599883317947388,
      "learning_rate": 0.0005,
      "loss": 3.384536027908325,
      "step": 1878
    },
    {
      "epoch": 0.4294857142857143,
      "grad_norm": 1.0043224096298218,
      "learning_rate": 0.0005,
      "loss": 2.836566686630249,
      "step": 1879
    },
    {
      "epoch": 0.4297142857142857,
      "grad_norm": 1.0856108665466309,
      "learning_rate": 0.0005,
      "loss": 3.0602622032165527,
      "step": 1880
    },
    {
      "epoch": 0.4299428571428571,
      "grad_norm": 1.236599087715149,
      "learning_rate": 0.0005,
      "loss": 3.0284881591796875,
      "step": 1881
    },
    {
      "epoch": 0.4301714285714286,
      "grad_norm": 2.131490707397461,
      "learning_rate": 0.0005,
      "loss": 3.198554277420044,
      "step": 1882
    },
    {
      "epoch": 0.4304,
      "grad_norm": 1.2054136991500854,
      "learning_rate": 0.0005,
      "loss": 3.2092623710632324,
      "step": 1883
    },
    {
      "epoch": 0.4306285714285714,
      "grad_norm": 1.3735620975494385,
      "learning_rate": 0.0005,
      "loss": 3.235877752304077,
      "step": 1884
    },
    {
      "epoch": 0.4308571428571429,
      "grad_norm": 1.0466902256011963,
      "learning_rate": 0.0005,
      "loss": 2.3617210388183594,
      "step": 1885
    },
    {
      "epoch": 0.4310857142857143,
      "grad_norm": 2.06754732131958,
      "learning_rate": 0.0005,
      "loss": 2.918642044067383,
      "step": 1886
    },
    {
      "epoch": 0.4313142857142857,
      "grad_norm": 1.1729762554168701,
      "learning_rate": 0.0005,
      "loss": 3.200821876525879,
      "step": 1887
    },
    {
      "epoch": 0.4315428571428571,
      "grad_norm": 1.0195800065994263,
      "learning_rate": 0.0005,
      "loss": 3.155592918395996,
      "step": 1888
    },
    {
      "epoch": 0.4317714285714286,
      "grad_norm": 1.2161380052566528,
      "learning_rate": 0.0005,
      "loss": 3.05951189994812,
      "step": 1889
    },
    {
      "epoch": 0.432,
      "grad_norm": 1.5719741582870483,
      "learning_rate": 0.0005,
      "loss": 3.085491895675659,
      "step": 1890
    },
    {
      "epoch": 0.4322285714285714,
      "grad_norm": 1.294489860534668,
      "learning_rate": 0.0005,
      "loss": 2.5916266441345215,
      "step": 1891
    },
    {
      "epoch": 0.4324571428571429,
      "grad_norm": 1.1580524444580078,
      "learning_rate": 0.0005,
      "loss": 3.243215799331665,
      "step": 1892
    },
    {
      "epoch": 0.4326857142857143,
      "grad_norm": 1.188538908958435,
      "learning_rate": 0.0005,
      "loss": 2.8141634464263916,
      "step": 1893
    },
    {
      "epoch": 0.4329142857142857,
      "grad_norm": 1.0909593105316162,
      "learning_rate": 0.0005,
      "loss": 2.942080497741699,
      "step": 1894
    },
    {
      "epoch": 0.43314285714285716,
      "grad_norm": 1.1304138898849487,
      "learning_rate": 0.0005,
      "loss": 3.220900774002075,
      "step": 1895
    },
    {
      "epoch": 0.4333714285714286,
      "grad_norm": 1.6559048891067505,
      "learning_rate": 0.0005,
      "loss": 3.3067924976348877,
      "step": 1896
    },
    {
      "epoch": 0.4336,
      "grad_norm": 1.0975251197814941,
      "learning_rate": 0.0005,
      "loss": 3.119255542755127,
      "step": 1897
    },
    {
      "epoch": 0.43382857142857145,
      "grad_norm": 1.1279829740524292,
      "learning_rate": 0.0005,
      "loss": 2.8232226371765137,
      "step": 1898
    },
    {
      "epoch": 0.43405714285714286,
      "grad_norm": 2.0831782817840576,
      "learning_rate": 0.0005,
      "loss": 3.3985304832458496,
      "step": 1899
    },
    {
      "epoch": 0.4342857142857143,
      "grad_norm": 1.2783175706863403,
      "learning_rate": 0.0005,
      "loss": 2.802046298980713,
      "step": 1900
    },
    {
      "epoch": 0.43451428571428574,
      "grad_norm": 1.1005545854568481,
      "learning_rate": 0.0005,
      "loss": 2.9163897037506104,
      "step": 1901
    },
    {
      "epoch": 0.43474285714285715,
      "grad_norm": 1.2194066047668457,
      "learning_rate": 0.0005,
      "loss": 3.143111228942871,
      "step": 1902
    },
    {
      "epoch": 0.43497142857142856,
      "grad_norm": 1.0941858291625977,
      "learning_rate": 0.0005,
      "loss": 3.101776361465454,
      "step": 1903
    },
    {
      "epoch": 0.4352,
      "grad_norm": 1.0212527513504028,
      "learning_rate": 0.0005,
      "loss": 2.9262208938598633,
      "step": 1904
    },
    {
      "epoch": 0.43542857142857144,
      "grad_norm": 1.204108715057373,
      "learning_rate": 0.0005,
      "loss": 3.0760653018951416,
      "step": 1905
    },
    {
      "epoch": 0.43565714285714285,
      "grad_norm": 1.392800211906433,
      "learning_rate": 0.0005,
      "loss": 2.884378433227539,
      "step": 1906
    },
    {
      "epoch": 0.43588571428571427,
      "grad_norm": 1.0484946966171265,
      "learning_rate": 0.0005,
      "loss": 2.9994640350341797,
      "step": 1907
    },
    {
      "epoch": 0.43611428571428573,
      "grad_norm": 1.0637534856796265,
      "learning_rate": 0.0005,
      "loss": 3.0728249549865723,
      "step": 1908
    },
    {
      "epoch": 0.43634285714285714,
      "grad_norm": 1.2108771800994873,
      "learning_rate": 0.0005,
      "loss": 3.14804744720459,
      "step": 1909
    },
    {
      "epoch": 0.43657142857142855,
      "grad_norm": 1.0740388631820679,
      "learning_rate": 0.0005,
      "loss": 2.780651330947876,
      "step": 1910
    },
    {
      "epoch": 0.4368,
      "grad_norm": 0.9507365822792053,
      "learning_rate": 0.0005,
      "loss": 2.8107995986938477,
      "step": 1911
    },
    {
      "epoch": 0.43702857142857143,
      "grad_norm": 1.0013326406478882,
      "learning_rate": 0.0005,
      "loss": 3.0919952392578125,
      "step": 1912
    },
    {
      "epoch": 0.43725714285714284,
      "grad_norm": 1.1813687086105347,
      "learning_rate": 0.0005,
      "loss": 2.8334460258483887,
      "step": 1913
    },
    {
      "epoch": 0.4374857142857143,
      "grad_norm": 0.993360698223114,
      "learning_rate": 0.0005,
      "loss": 2.7867252826690674,
      "step": 1914
    },
    {
      "epoch": 0.4377142857142857,
      "grad_norm": 1.1217405796051025,
      "learning_rate": 0.0005,
      "loss": 2.880450963973999,
      "step": 1915
    },
    {
      "epoch": 0.43794285714285713,
      "grad_norm": 1.0903934240341187,
      "learning_rate": 0.0005,
      "loss": 3.01457142829895,
      "step": 1916
    },
    {
      "epoch": 0.43817142857142855,
      "grad_norm": 1.117553949356079,
      "learning_rate": 0.0005,
      "loss": 2.5807623863220215,
      "step": 1917
    },
    {
      "epoch": 0.4384,
      "grad_norm": 1.4466800689697266,
      "learning_rate": 0.0005,
      "loss": 2.8534388542175293,
      "step": 1918
    },
    {
      "epoch": 0.4386285714285714,
      "grad_norm": 1.4748234748840332,
      "learning_rate": 0.0005,
      "loss": 3.171708822250366,
      "step": 1919
    },
    {
      "epoch": 0.43885714285714283,
      "grad_norm": 1.0536441802978516,
      "learning_rate": 0.0005,
      "loss": 3.1995720863342285,
      "step": 1920
    },
    {
      "epoch": 0.4390857142857143,
      "grad_norm": 1.2735514640808105,
      "learning_rate": 0.0005,
      "loss": 2.9232473373413086,
      "step": 1921
    },
    {
      "epoch": 0.4393142857142857,
      "grad_norm": 1.9100946187973022,
      "learning_rate": 0.0005,
      "loss": 3.0309202671051025,
      "step": 1922
    },
    {
      "epoch": 0.4395428571428571,
      "grad_norm": 1.3062907457351685,
      "learning_rate": 0.0005,
      "loss": 3.169888496398926,
      "step": 1923
    },
    {
      "epoch": 0.4397714285714286,
      "grad_norm": 0.8938153982162476,
      "learning_rate": 0.0005,
      "loss": 2.8655266761779785,
      "step": 1924
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.1324445009231567,
      "learning_rate": 0.0005,
      "loss": 2.7279443740844727,
      "step": 1925
    },
    {
      "epoch": 0.4402285714285714,
      "grad_norm": 0.9562188982963562,
      "learning_rate": 0.0005,
      "loss": 2.7554478645324707,
      "step": 1926
    },
    {
      "epoch": 0.4404571428571429,
      "grad_norm": 1.0861856937408447,
      "learning_rate": 0.0005,
      "loss": 3.135450839996338,
      "step": 1927
    },
    {
      "epoch": 0.4406857142857143,
      "grad_norm": 0.9940224289894104,
      "learning_rate": 0.0005,
      "loss": 2.6850881576538086,
      "step": 1928
    },
    {
      "epoch": 0.4409142857142857,
      "grad_norm": 1.6141986846923828,
      "learning_rate": 0.0005,
      "loss": 2.8903770446777344,
      "step": 1929
    },
    {
      "epoch": 0.44114285714285717,
      "grad_norm": 1.3986964225769043,
      "learning_rate": 0.0005,
      "loss": 2.8664002418518066,
      "step": 1930
    },
    {
      "epoch": 0.4413714285714286,
      "grad_norm": 1.405385971069336,
      "learning_rate": 0.0005,
      "loss": 2.774930000305176,
      "step": 1931
    },
    {
      "epoch": 0.4416,
      "grad_norm": 1.3184726238250732,
      "learning_rate": 0.0005,
      "loss": 2.9528732299804688,
      "step": 1932
    },
    {
      "epoch": 0.4418285714285714,
      "grad_norm": 1.206641435623169,
      "learning_rate": 0.0005,
      "loss": 3.0579652786254883,
      "step": 1933
    },
    {
      "epoch": 0.44205714285714287,
      "grad_norm": 0.978466808795929,
      "learning_rate": 0.0005,
      "loss": 2.701784610748291,
      "step": 1934
    },
    {
      "epoch": 0.4422857142857143,
      "grad_norm": 1.204261302947998,
      "learning_rate": 0.0005,
      "loss": 3.0518665313720703,
      "step": 1935
    },
    {
      "epoch": 0.4425142857142857,
      "grad_norm": 1.2352176904678345,
      "learning_rate": 0.0005,
      "loss": 2.9334659576416016,
      "step": 1936
    },
    {
      "epoch": 0.44274285714285716,
      "grad_norm": 1.2340787649154663,
      "learning_rate": 0.0005,
      "loss": 3.074488401412964,
      "step": 1937
    },
    {
      "epoch": 0.44297142857142857,
      "grad_norm": 1.0938968658447266,
      "learning_rate": 0.0005,
      "loss": 2.9659156799316406,
      "step": 1938
    },
    {
      "epoch": 0.4432,
      "grad_norm": 1.107056975364685,
      "learning_rate": 0.0005,
      "loss": 3.1968934535980225,
      "step": 1939
    },
    {
      "epoch": 0.44342857142857145,
      "grad_norm": 1.0548354387283325,
      "learning_rate": 0.0005,
      "loss": 3.095398426055908,
      "step": 1940
    },
    {
      "epoch": 0.44365714285714286,
      "grad_norm": 1.3446511030197144,
      "learning_rate": 0.0005,
      "loss": 2.9347329139709473,
      "step": 1941
    },
    {
      "epoch": 0.4438857142857143,
      "grad_norm": 0.9890820384025574,
      "learning_rate": 0.0005,
      "loss": 2.737170934677124,
      "step": 1942
    },
    {
      "epoch": 0.44411428571428574,
      "grad_norm": 0.9485434293746948,
      "learning_rate": 0.0005,
      "loss": 3.1959939002990723,
      "step": 1943
    },
    {
      "epoch": 0.44434285714285715,
      "grad_norm": 1.0584365129470825,
      "learning_rate": 0.0005,
      "loss": 3.408478260040283,
      "step": 1944
    },
    {
      "epoch": 0.44457142857142856,
      "grad_norm": 1.2219572067260742,
      "learning_rate": 0.0005,
      "loss": 3.1022164821624756,
      "step": 1945
    },
    {
      "epoch": 0.4448,
      "grad_norm": 1.0274006128311157,
      "learning_rate": 0.0005,
      "loss": 2.9247660636901855,
      "step": 1946
    },
    {
      "epoch": 0.44502857142857144,
      "grad_norm": 1.1016639471054077,
      "learning_rate": 0.0005,
      "loss": 2.6417226791381836,
      "step": 1947
    },
    {
      "epoch": 0.44525714285714285,
      "grad_norm": 1.2179256677627563,
      "learning_rate": 0.0005,
      "loss": 2.528229236602783,
      "step": 1948
    },
    {
      "epoch": 0.44548571428571426,
      "grad_norm": 1.5647621154785156,
      "learning_rate": 0.0005,
      "loss": 2.706501007080078,
      "step": 1949
    },
    {
      "epoch": 0.44571428571428573,
      "grad_norm": 1.3237684965133667,
      "learning_rate": 0.0005,
      "loss": 2.8213582038879395,
      "step": 1950
    },
    {
      "epoch": 0.44594285714285714,
      "grad_norm": 1.2619692087173462,
      "learning_rate": 0.0005,
      "loss": 2.977476119995117,
      "step": 1951
    },
    {
      "epoch": 0.44617142857142855,
      "grad_norm": 1.11127507686615,
      "learning_rate": 0.0005,
      "loss": 2.8880677223205566,
      "step": 1952
    },
    {
      "epoch": 0.4464,
      "grad_norm": 1.0854510068893433,
      "learning_rate": 0.0005,
      "loss": 2.939920663833618,
      "step": 1953
    },
    {
      "epoch": 0.44662857142857143,
      "grad_norm": 1.0358002185821533,
      "learning_rate": 0.0005,
      "loss": 3.1646218299865723,
      "step": 1954
    },
    {
      "epoch": 0.44685714285714284,
      "grad_norm": 1.0829930305480957,
      "learning_rate": 0.0005,
      "loss": 2.8962817192077637,
      "step": 1955
    },
    {
      "epoch": 0.4470857142857143,
      "grad_norm": 1.1182864904403687,
      "learning_rate": 0.0005,
      "loss": 3.2040815353393555,
      "step": 1956
    },
    {
      "epoch": 0.4473142857142857,
      "grad_norm": 1.2124460935592651,
      "learning_rate": 0.0005,
      "loss": 2.8710215091705322,
      "step": 1957
    },
    {
      "epoch": 0.44754285714285713,
      "grad_norm": 1.1268060207366943,
      "learning_rate": 0.0005,
      "loss": 2.859461784362793,
      "step": 1958
    },
    {
      "epoch": 0.4477714285714286,
      "grad_norm": 1.0946097373962402,
      "learning_rate": 0.0005,
      "loss": 3.415198802947998,
      "step": 1959
    },
    {
      "epoch": 0.448,
      "grad_norm": 1.5253099203109741,
      "learning_rate": 0.0005,
      "loss": 2.853503704071045,
      "step": 1960
    },
    {
      "epoch": 0.4482285714285714,
      "grad_norm": 1.235680103302002,
      "learning_rate": 0.0005,
      "loss": 3.317392349243164,
      "step": 1961
    },
    {
      "epoch": 0.44845714285714283,
      "grad_norm": 1.4731889963150024,
      "learning_rate": 0.0005,
      "loss": 3.019909143447876,
      "step": 1962
    },
    {
      "epoch": 0.4486857142857143,
      "grad_norm": 1.449654459953308,
      "learning_rate": 0.0005,
      "loss": 3.343456268310547,
      "step": 1963
    },
    {
      "epoch": 0.4489142857142857,
      "grad_norm": 1.4664393663406372,
      "learning_rate": 0.0005,
      "loss": 2.8122901916503906,
      "step": 1964
    },
    {
      "epoch": 0.4491428571428571,
      "grad_norm": 1.008952021598816,
      "learning_rate": 0.0005,
      "loss": 3.090142011642456,
      "step": 1965
    },
    {
      "epoch": 0.4493714285714286,
      "grad_norm": 1.2543209791183472,
      "learning_rate": 0.0005,
      "loss": 2.67177677154541,
      "step": 1966
    },
    {
      "epoch": 0.4496,
      "grad_norm": 1.1045918464660645,
      "learning_rate": 0.0005,
      "loss": 3.0612869262695312,
      "step": 1967
    },
    {
      "epoch": 0.4498285714285714,
      "grad_norm": 1.200209379196167,
      "learning_rate": 0.0005,
      "loss": 2.896827220916748,
      "step": 1968
    },
    {
      "epoch": 0.4500571428571429,
      "grad_norm": 1.001721739768982,
      "learning_rate": 0.0005,
      "loss": 3.1211304664611816,
      "step": 1969
    },
    {
      "epoch": 0.4502857142857143,
      "grad_norm": 1.0761020183563232,
      "learning_rate": 0.0005,
      "loss": 3.050589084625244,
      "step": 1970
    },
    {
      "epoch": 0.4505142857142857,
      "grad_norm": 1.0922027826309204,
      "learning_rate": 0.0005,
      "loss": 3.2460572719573975,
      "step": 1971
    },
    {
      "epoch": 0.45074285714285717,
      "grad_norm": 1.0940217971801758,
      "learning_rate": 0.0005,
      "loss": 2.794628143310547,
      "step": 1972
    },
    {
      "epoch": 0.4509714285714286,
      "grad_norm": 1.0330743789672852,
      "learning_rate": 0.0005,
      "loss": 3.0079665184020996,
      "step": 1973
    },
    {
      "epoch": 0.4512,
      "grad_norm": 1.0939122438430786,
      "learning_rate": 0.0005,
      "loss": 2.8635058403015137,
      "step": 1974
    },
    {
      "epoch": 0.4514285714285714,
      "grad_norm": 1.166395902633667,
      "learning_rate": 0.0005,
      "loss": 2.7494523525238037,
      "step": 1975
    },
    {
      "epoch": 0.45165714285714287,
      "grad_norm": 1.170459508895874,
      "learning_rate": 0.0005,
      "loss": 2.7353146076202393,
      "step": 1976
    },
    {
      "epoch": 0.4518857142857143,
      "grad_norm": 1.3232191801071167,
      "learning_rate": 0.0005,
      "loss": 2.8759868144989014,
      "step": 1977
    },
    {
      "epoch": 0.4521142857142857,
      "grad_norm": 1.0664465427398682,
      "learning_rate": 0.0005,
      "loss": 2.9798240661621094,
      "step": 1978
    },
    {
      "epoch": 0.45234285714285716,
      "grad_norm": 1.4511864185333252,
      "learning_rate": 0.0005,
      "loss": 3.259528160095215,
      "step": 1979
    },
    {
      "epoch": 0.45257142857142857,
      "grad_norm": 1.2714059352874756,
      "learning_rate": 0.0005,
      "loss": 3.2013769149780273,
      "step": 1980
    },
    {
      "epoch": 0.4528,
      "grad_norm": 1.1145761013031006,
      "learning_rate": 0.0005,
      "loss": 2.9900124073028564,
      "step": 1981
    },
    {
      "epoch": 0.45302857142857145,
      "grad_norm": 1.0176268815994263,
      "learning_rate": 0.0005,
      "loss": 3.0600297451019287,
      "step": 1982
    },
    {
      "epoch": 0.45325714285714286,
      "grad_norm": 1.1262717247009277,
      "learning_rate": 0.0005,
      "loss": 2.953495979309082,
      "step": 1983
    },
    {
      "epoch": 0.45348571428571427,
      "grad_norm": 1.2640573978424072,
      "learning_rate": 0.0005,
      "loss": 2.546581506729126,
      "step": 1984
    },
    {
      "epoch": 0.45371428571428574,
      "grad_norm": 1.044439435005188,
      "learning_rate": 0.0005,
      "loss": 2.6287217140197754,
      "step": 1985
    },
    {
      "epoch": 0.45394285714285715,
      "grad_norm": 0.9097898602485657,
      "learning_rate": 0.0005,
      "loss": 3.090677261352539,
      "step": 1986
    },
    {
      "epoch": 0.45417142857142856,
      "grad_norm": 1.2312134504318237,
      "learning_rate": 0.0005,
      "loss": 2.8700952529907227,
      "step": 1987
    },
    {
      "epoch": 0.4544,
      "grad_norm": 1.8427051305770874,
      "learning_rate": 0.0005,
      "loss": 3.2476541996002197,
      "step": 1988
    },
    {
      "epoch": 0.45462857142857144,
      "grad_norm": 1.8278294801712036,
      "learning_rate": 0.0005,
      "loss": 2.9693398475646973,
      "step": 1989
    },
    {
      "epoch": 0.45485714285714285,
      "grad_norm": 1.4219194650650024,
      "learning_rate": 0.0005,
      "loss": 2.8187789916992188,
      "step": 1990
    },
    {
      "epoch": 0.45508571428571426,
      "grad_norm": 1.3854587078094482,
      "learning_rate": 0.0005,
      "loss": 2.9377036094665527,
      "step": 1991
    },
    {
      "epoch": 0.4553142857142857,
      "grad_norm": 1.1342018842697144,
      "learning_rate": 0.0005,
      "loss": 3.138406753540039,
      "step": 1992
    },
    {
      "epoch": 0.45554285714285714,
      "grad_norm": 1.0104385614395142,
      "learning_rate": 0.0005,
      "loss": 3.2009034156799316,
      "step": 1993
    },
    {
      "epoch": 0.45577142857142855,
      "grad_norm": 1.3162792921066284,
      "learning_rate": 0.0005,
      "loss": 3.1749048233032227,
      "step": 1994
    },
    {
      "epoch": 0.456,
      "grad_norm": 1.181699514389038,
      "learning_rate": 0.0005,
      "loss": 2.8328585624694824,
      "step": 1995
    },
    {
      "epoch": 0.45622857142857143,
      "grad_norm": 1.0334845781326294,
      "learning_rate": 0.0005,
      "loss": 2.8391458988189697,
      "step": 1996
    },
    {
      "epoch": 0.45645714285714284,
      "grad_norm": 1.0683225393295288,
      "learning_rate": 0.0005,
      "loss": 3.032393455505371,
      "step": 1997
    },
    {
      "epoch": 0.4566857142857143,
      "grad_norm": 1.1990995407104492,
      "learning_rate": 0.0005,
      "loss": 2.855405330657959,
      "step": 1998
    },
    {
      "epoch": 0.4569142857142857,
      "grad_norm": 1.3229023218154907,
      "learning_rate": 0.0005,
      "loss": 2.9292964935302734,
      "step": 1999
    },
    {
      "epoch": 0.45714285714285713,
      "grad_norm": 1.8142541646957397,
      "learning_rate": 0.0005,
      "loss": 2.921773910522461,
      "step": 2000
    },
    {
      "epoch": 0.4573714285714286,
      "grad_norm": 1.905449390411377,
      "learning_rate": 0.0005,
      "loss": 2.697697639465332,
      "step": 2001
    },
    {
      "epoch": 0.4576,
      "grad_norm": 1.1343357563018799,
      "learning_rate": 0.0005,
      "loss": 3.0776853561401367,
      "step": 2002
    },
    {
      "epoch": 0.4578285714285714,
      "grad_norm": 1.0003565549850464,
      "learning_rate": 0.0005,
      "loss": 2.9733943939208984,
      "step": 2003
    },
    {
      "epoch": 0.45805714285714283,
      "grad_norm": 2.4689440727233887,
      "learning_rate": 0.0005,
      "loss": 3.3121912479400635,
      "step": 2004
    },
    {
      "epoch": 0.4582857142857143,
      "grad_norm": 1.325075626373291,
      "learning_rate": 0.0005,
      "loss": 3.1121535301208496,
      "step": 2005
    },
    {
      "epoch": 0.4585142857142857,
      "grad_norm": 1.253161072731018,
      "learning_rate": 0.0005,
      "loss": 2.9637930393218994,
      "step": 2006
    },
    {
      "epoch": 0.4587428571428571,
      "grad_norm": 1.5103175640106201,
      "learning_rate": 0.0005,
      "loss": 2.9403481483459473,
      "step": 2007
    },
    {
      "epoch": 0.4589714285714286,
      "grad_norm": 1.1479383707046509,
      "learning_rate": 0.0005,
      "loss": 2.8141636848449707,
      "step": 2008
    },
    {
      "epoch": 0.4592,
      "grad_norm": 1.2081865072250366,
      "learning_rate": 0.0005,
      "loss": 2.8862903118133545,
      "step": 2009
    },
    {
      "epoch": 0.4594285714285714,
      "grad_norm": 1.2611087560653687,
      "learning_rate": 0.0005,
      "loss": 2.7532777786254883,
      "step": 2010
    },
    {
      "epoch": 0.4596571428571429,
      "grad_norm": 1.2369768619537354,
      "learning_rate": 0.0005,
      "loss": 3.114741563796997,
      "step": 2011
    },
    {
      "epoch": 0.4598857142857143,
      "grad_norm": 1.0425035953521729,
      "learning_rate": 0.0005,
      "loss": 2.8746275901794434,
      "step": 2012
    },
    {
      "epoch": 0.4601142857142857,
      "grad_norm": 1.3187310695648193,
      "learning_rate": 0.0005,
      "loss": 2.9179799556732178,
      "step": 2013
    },
    {
      "epoch": 0.46034285714285716,
      "grad_norm": 1.2960134744644165,
      "learning_rate": 0.0005,
      "loss": 2.9491395950317383,
      "step": 2014
    },
    {
      "epoch": 0.4605714285714286,
      "grad_norm": 1.3923490047454834,
      "learning_rate": 0.0005,
      "loss": 2.8143131732940674,
      "step": 2015
    },
    {
      "epoch": 0.4608,
      "grad_norm": 1.1124025583267212,
      "learning_rate": 0.0005,
      "loss": 3.0970940589904785,
      "step": 2016
    },
    {
      "epoch": 0.46102857142857145,
      "grad_norm": 1.3533897399902344,
      "learning_rate": 0.0005,
      "loss": 3.2929370403289795,
      "step": 2017
    },
    {
      "epoch": 0.46125714285714287,
      "grad_norm": 1.8183097839355469,
      "learning_rate": 0.0005,
      "loss": 2.9319376945495605,
      "step": 2018
    },
    {
      "epoch": 0.4614857142857143,
      "grad_norm": 1.7203856706619263,
      "learning_rate": 0.0005,
      "loss": 3.041823148727417,
      "step": 2019
    },
    {
      "epoch": 0.4617142857142857,
      "grad_norm": 1.3973462581634521,
      "learning_rate": 0.0005,
      "loss": 2.944213628768921,
      "step": 2020
    },
    {
      "epoch": 0.46194285714285716,
      "grad_norm": 1.2172430753707886,
      "learning_rate": 0.0005,
      "loss": 3.0028698444366455,
      "step": 2021
    },
    {
      "epoch": 0.46217142857142857,
      "grad_norm": 1.5062339305877686,
      "learning_rate": 0.0005,
      "loss": 2.7518768310546875,
      "step": 2022
    },
    {
      "epoch": 0.4624,
      "grad_norm": 1.27083158493042,
      "learning_rate": 0.0005,
      "loss": 3.085848331451416,
      "step": 2023
    },
    {
      "epoch": 0.46262857142857144,
      "grad_norm": 1.1439564228057861,
      "learning_rate": 0.0005,
      "loss": 2.5038866996765137,
      "step": 2024
    },
    {
      "epoch": 0.46285714285714286,
      "grad_norm": 1.1299179792404175,
      "learning_rate": 0.0005,
      "loss": 3.233346462249756,
      "step": 2025
    },
    {
      "epoch": 0.46308571428571427,
      "grad_norm": 1.2489113807678223,
      "learning_rate": 0.0005,
      "loss": 3.122361421585083,
      "step": 2026
    },
    {
      "epoch": 0.46331428571428573,
      "grad_norm": 0.9578064680099487,
      "learning_rate": 0.0005,
      "loss": 2.9972851276397705,
      "step": 2027
    },
    {
      "epoch": 0.46354285714285715,
      "grad_norm": 1.5972872972488403,
      "learning_rate": 0.0005,
      "loss": 2.99935245513916,
      "step": 2028
    },
    {
      "epoch": 0.46377142857142856,
      "grad_norm": 0.8698304891586304,
      "learning_rate": 0.0005,
      "loss": 2.5458481311798096,
      "step": 2029
    },
    {
      "epoch": 0.464,
      "grad_norm": 1.090773105621338,
      "learning_rate": 0.0005,
      "loss": 2.9262964725494385,
      "step": 2030
    },
    {
      "epoch": 0.46422857142857143,
      "grad_norm": 1.058463215827942,
      "learning_rate": 0.0005,
      "loss": 2.8728740215301514,
      "step": 2031
    },
    {
      "epoch": 0.46445714285714285,
      "grad_norm": 1.4908413887023926,
      "learning_rate": 0.0005,
      "loss": 2.79691219329834,
      "step": 2032
    },
    {
      "epoch": 0.4646857142857143,
      "grad_norm": 1.34389066696167,
      "learning_rate": 0.0005,
      "loss": 2.908141613006592,
      "step": 2033
    },
    {
      "epoch": 0.4649142857142857,
      "grad_norm": 1.1486543416976929,
      "learning_rate": 0.0005,
      "loss": 3.2732338905334473,
      "step": 2034
    },
    {
      "epoch": 0.46514285714285714,
      "grad_norm": 1.0301527976989746,
      "learning_rate": 0.0005,
      "loss": 2.9352235794067383,
      "step": 2035
    },
    {
      "epoch": 0.46537142857142855,
      "grad_norm": 1.0637743473052979,
      "learning_rate": 0.0005,
      "loss": 2.6078426837921143,
      "step": 2036
    },
    {
      "epoch": 0.4656,
      "grad_norm": 1.0562998056411743,
      "learning_rate": 0.0005,
      "loss": 2.642271041870117,
      "step": 2037
    },
    {
      "epoch": 0.4658285714285714,
      "grad_norm": 1.0190575122833252,
      "learning_rate": 0.0005,
      "loss": 2.7805376052856445,
      "step": 2038
    },
    {
      "epoch": 0.46605714285714284,
      "grad_norm": 2.096545696258545,
      "learning_rate": 0.0005,
      "loss": 3.095900535583496,
      "step": 2039
    },
    {
      "epoch": 0.4662857142857143,
      "grad_norm": 0.9539772868156433,
      "learning_rate": 0.0005,
      "loss": 2.8474414348602295,
      "step": 2040
    },
    {
      "epoch": 0.4665142857142857,
      "grad_norm": 2.4445736408233643,
      "learning_rate": 0.0005,
      "loss": 3.0244178771972656,
      "step": 2041
    },
    {
      "epoch": 0.4667428571428571,
      "grad_norm": 1.1268407106399536,
      "learning_rate": 0.0005,
      "loss": 3.3438615798950195,
      "step": 2042
    },
    {
      "epoch": 0.4669714285714286,
      "grad_norm": 1.0843162536621094,
      "learning_rate": 0.0005,
      "loss": 2.84893798828125,
      "step": 2043
    },
    {
      "epoch": 0.4672,
      "grad_norm": 1.0733481645584106,
      "learning_rate": 0.0005,
      "loss": 2.6957437992095947,
      "step": 2044
    },
    {
      "epoch": 0.4674285714285714,
      "grad_norm": 1.1581143140792847,
      "learning_rate": 0.0005,
      "loss": 2.504908561706543,
      "step": 2045
    },
    {
      "epoch": 0.4676571428571429,
      "grad_norm": 1.0642328262329102,
      "learning_rate": 0.0005,
      "loss": 2.9711251258850098,
      "step": 2046
    },
    {
      "epoch": 0.4678857142857143,
      "grad_norm": 1.2239199876785278,
      "learning_rate": 0.0005,
      "loss": 2.5384764671325684,
      "step": 2047
    },
    {
      "epoch": 0.4681142857142857,
      "grad_norm": 1.0262831449508667,
      "learning_rate": 0.0005,
      "loss": 2.852180242538452,
      "step": 2048
    },
    {
      "epoch": 0.4683428571428571,
      "grad_norm": 1.3428176641464233,
      "learning_rate": 0.0005,
      "loss": 2.706395387649536,
      "step": 2049
    },
    {
      "epoch": 0.4685714285714286,
      "grad_norm": 1.1855685710906982,
      "learning_rate": 0.0005,
      "loss": 3.2512245178222656,
      "step": 2050
    },
    {
      "epoch": 0.4688,
      "grad_norm": 1.1108365058898926,
      "learning_rate": 0.0005,
      "loss": 2.384066581726074,
      "step": 2051
    },
    {
      "epoch": 0.4690285714285714,
      "grad_norm": 1.089802622795105,
      "learning_rate": 0.0005,
      "loss": 2.659166097640991,
      "step": 2052
    },
    {
      "epoch": 0.4692571428571429,
      "grad_norm": 1.1063828468322754,
      "learning_rate": 0.0005,
      "loss": 3.2244434356689453,
      "step": 2053
    },
    {
      "epoch": 0.4694857142857143,
      "grad_norm": 1.6249045133590698,
      "learning_rate": 0.0005,
      "loss": 2.8851027488708496,
      "step": 2054
    },
    {
      "epoch": 0.4697142857142857,
      "grad_norm": 1.1565496921539307,
      "learning_rate": 0.0005,
      "loss": 2.828199863433838,
      "step": 2055
    },
    {
      "epoch": 0.46994285714285716,
      "grad_norm": 1.0507228374481201,
      "learning_rate": 0.0005,
      "loss": 2.998504161834717,
      "step": 2056
    },
    {
      "epoch": 0.4701714285714286,
      "grad_norm": 1.5198320150375366,
      "learning_rate": 0.0005,
      "loss": 2.8026108741760254,
      "step": 2057
    },
    {
      "epoch": 0.4704,
      "grad_norm": 0.9218361377716064,
      "learning_rate": 0.0005,
      "loss": 2.907386302947998,
      "step": 2058
    },
    {
      "epoch": 0.47062857142857145,
      "grad_norm": 1.087572693824768,
      "learning_rate": 0.0005,
      "loss": 2.886606216430664,
      "step": 2059
    },
    {
      "epoch": 0.47085714285714286,
      "grad_norm": 1.2448610067367554,
      "learning_rate": 0.0005,
      "loss": 2.96633243560791,
      "step": 2060
    },
    {
      "epoch": 0.4710857142857143,
      "grad_norm": 1.1032283306121826,
      "learning_rate": 0.0005,
      "loss": 3.164592742919922,
      "step": 2061
    },
    {
      "epoch": 0.47131428571428574,
      "grad_norm": 1.1853582859039307,
      "learning_rate": 0.0005,
      "loss": 2.73028564453125,
      "step": 2062
    },
    {
      "epoch": 0.47154285714285715,
      "grad_norm": 0.9104340672492981,
      "learning_rate": 0.0005,
      "loss": 2.9893505573272705,
      "step": 2063
    },
    {
      "epoch": 0.47177142857142856,
      "grad_norm": 1.0158157348632812,
      "learning_rate": 0.0005,
      "loss": 3.3995277881622314,
      "step": 2064
    },
    {
      "epoch": 0.472,
      "grad_norm": 1.1027246713638306,
      "learning_rate": 0.0005,
      "loss": 2.813937187194824,
      "step": 2065
    },
    {
      "epoch": 0.47222857142857144,
      "grad_norm": 0.9117674231529236,
      "learning_rate": 0.0005,
      "loss": 2.6704201698303223,
      "step": 2066
    },
    {
      "epoch": 0.47245714285714285,
      "grad_norm": 1.2297087907791138,
      "learning_rate": 0.0005,
      "loss": 2.9053361415863037,
      "step": 2067
    },
    {
      "epoch": 0.47268571428571426,
      "grad_norm": 0.9519703984260559,
      "learning_rate": 0.0005,
      "loss": 3.0339536666870117,
      "step": 2068
    },
    {
      "epoch": 0.47291428571428573,
      "grad_norm": 1.007095456123352,
      "learning_rate": 0.0005,
      "loss": 3.2147536277770996,
      "step": 2069
    },
    {
      "epoch": 0.47314285714285714,
      "grad_norm": 1.1081703901290894,
      "learning_rate": 0.0005,
      "loss": 3.0332207679748535,
      "step": 2070
    },
    {
      "epoch": 0.47337142857142855,
      "grad_norm": 1.3075851202011108,
      "learning_rate": 0.0005,
      "loss": 2.6855926513671875,
      "step": 2071
    },
    {
      "epoch": 0.4736,
      "grad_norm": 1.0418195724487305,
      "learning_rate": 0.0005,
      "loss": 3.0080316066741943,
      "step": 2072
    },
    {
      "epoch": 0.47382857142857143,
      "grad_norm": 1.016005277633667,
      "learning_rate": 0.0005,
      "loss": 2.8267369270324707,
      "step": 2073
    },
    {
      "epoch": 0.47405714285714284,
      "grad_norm": 1.154803991317749,
      "learning_rate": 0.0005,
      "loss": 2.8053624629974365,
      "step": 2074
    },
    {
      "epoch": 0.4742857142857143,
      "grad_norm": 1.471834421157837,
      "learning_rate": 0.0005,
      "loss": 3.1665709018707275,
      "step": 2075
    },
    {
      "epoch": 0.4745142857142857,
      "grad_norm": 0.9337404370307922,
      "learning_rate": 0.0005,
      "loss": 2.9253811836242676,
      "step": 2076
    },
    {
      "epoch": 0.47474285714285713,
      "grad_norm": 1.3249014616012573,
      "learning_rate": 0.0005,
      "loss": 3.2379069328308105,
      "step": 2077
    },
    {
      "epoch": 0.47497142857142854,
      "grad_norm": 2.0942342281341553,
      "learning_rate": 0.0005,
      "loss": 3.208157777786255,
      "step": 2078
    },
    {
      "epoch": 0.4752,
      "grad_norm": 1.7747223377227783,
      "learning_rate": 0.0005,
      "loss": 2.9058594703674316,
      "step": 2079
    },
    {
      "epoch": 0.4754285714285714,
      "grad_norm": 1.0357764959335327,
      "learning_rate": 0.0005,
      "loss": 3.066030263900757,
      "step": 2080
    },
    {
      "epoch": 0.47565714285714283,
      "grad_norm": 2.2764809131622314,
      "learning_rate": 0.0005,
      "loss": 3.0192370414733887,
      "step": 2081
    },
    {
      "epoch": 0.4758857142857143,
      "grad_norm": 1.8123594522476196,
      "learning_rate": 0.0005,
      "loss": 2.6113553047180176,
      "step": 2082
    },
    {
      "epoch": 0.4761142857142857,
      "grad_norm": 1.464027762413025,
      "learning_rate": 0.0005,
      "loss": 3.1661014556884766,
      "step": 2083
    },
    {
      "epoch": 0.4763428571428571,
      "grad_norm": 1.1135876178741455,
      "learning_rate": 0.0005,
      "loss": 2.291200876235962,
      "step": 2084
    },
    {
      "epoch": 0.4765714285714286,
      "grad_norm": 0.9903227686882019,
      "learning_rate": 0.0005,
      "loss": 3.017800807952881,
      "step": 2085
    },
    {
      "epoch": 0.4768,
      "grad_norm": 1.2058286666870117,
      "learning_rate": 0.0005,
      "loss": 3.100719928741455,
      "step": 2086
    },
    {
      "epoch": 0.4770285714285714,
      "grad_norm": 1.1278005838394165,
      "learning_rate": 0.0005,
      "loss": 2.869248867034912,
      "step": 2087
    },
    {
      "epoch": 0.4772571428571429,
      "grad_norm": 1.208213448524475,
      "learning_rate": 0.0005,
      "loss": 2.838024139404297,
      "step": 2088
    },
    {
      "epoch": 0.4774857142857143,
      "grad_norm": 1.276310682296753,
      "learning_rate": 0.0005,
      "loss": 2.6425833702087402,
      "step": 2089
    },
    {
      "epoch": 0.4777142857142857,
      "grad_norm": 1.2461782693862915,
      "learning_rate": 0.0005,
      "loss": 3.0004115104675293,
      "step": 2090
    },
    {
      "epoch": 0.47794285714285717,
      "grad_norm": 1.0322318077087402,
      "learning_rate": 0.0005,
      "loss": 2.9188408851623535,
      "step": 2091
    },
    {
      "epoch": 0.4781714285714286,
      "grad_norm": 0.9904196858406067,
      "learning_rate": 0.0005,
      "loss": 2.946981906890869,
      "step": 2092
    },
    {
      "epoch": 0.4784,
      "grad_norm": 0.9092068672180176,
      "learning_rate": 0.0005,
      "loss": 3.2065329551696777,
      "step": 2093
    },
    {
      "epoch": 0.4786285714285714,
      "grad_norm": 1.1344788074493408,
      "learning_rate": 0.0005,
      "loss": 2.437690496444702,
      "step": 2094
    },
    {
      "epoch": 0.47885714285714287,
      "grad_norm": 0.8632790446281433,
      "learning_rate": 0.0005,
      "loss": 2.5048019886016846,
      "step": 2095
    },
    {
      "epoch": 0.4790857142857143,
      "grad_norm": 1.1325874328613281,
      "learning_rate": 0.0005,
      "loss": 3.3501553535461426,
      "step": 2096
    },
    {
      "epoch": 0.4793142857142857,
      "grad_norm": 1.2048603296279907,
      "learning_rate": 0.0005,
      "loss": 3.1816887855529785,
      "step": 2097
    },
    {
      "epoch": 0.47954285714285716,
      "grad_norm": 1.091516375541687,
      "learning_rate": 0.0005,
      "loss": 3.201667070388794,
      "step": 2098
    },
    {
      "epoch": 0.47977142857142857,
      "grad_norm": 1.0999970436096191,
      "learning_rate": 0.0005,
      "loss": 2.6246414184570312,
      "step": 2099
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.5185405015945435,
      "learning_rate": 0.0005,
      "loss": 2.769996166229248,
      "step": 2100
    },
    {
      "epoch": 0.48022857142857145,
      "grad_norm": 1.1667159795761108,
      "learning_rate": 0.0005,
      "loss": 2.9619972705841064,
      "step": 2101
    },
    {
      "epoch": 0.48045714285714286,
      "grad_norm": 1.5629949569702148,
      "learning_rate": 0.0005,
      "loss": 2.681190013885498,
      "step": 2102
    },
    {
      "epoch": 0.48068571428571427,
      "grad_norm": 1.0694715976715088,
      "learning_rate": 0.0005,
      "loss": 2.773873805999756,
      "step": 2103
    },
    {
      "epoch": 0.48091428571428574,
      "grad_norm": 0.9453467726707458,
      "learning_rate": 0.0005,
      "loss": 3.0847935676574707,
      "step": 2104
    },
    {
      "epoch": 0.48114285714285715,
      "grad_norm": 1.062071442604065,
      "learning_rate": 0.0005,
      "loss": 2.9110946655273438,
      "step": 2105
    },
    {
      "epoch": 0.48137142857142856,
      "grad_norm": 1.363563060760498,
      "learning_rate": 0.0005,
      "loss": 3.026531219482422,
      "step": 2106
    },
    {
      "epoch": 0.4816,
      "grad_norm": 0.9698224067687988,
      "learning_rate": 0.0005,
      "loss": 2.7974026203155518,
      "step": 2107
    },
    {
      "epoch": 0.48182857142857144,
      "grad_norm": 1.0158414840698242,
      "learning_rate": 0.0005,
      "loss": 2.795421838760376,
      "step": 2108
    },
    {
      "epoch": 0.48205714285714285,
      "grad_norm": 1.3372174501419067,
      "learning_rate": 0.0005,
      "loss": 2.8332791328430176,
      "step": 2109
    },
    {
      "epoch": 0.48228571428571426,
      "grad_norm": 1.2650668621063232,
      "learning_rate": 0.0005,
      "loss": 2.626009464263916,
      "step": 2110
    },
    {
      "epoch": 0.48251428571428573,
      "grad_norm": 1.2563292980194092,
      "learning_rate": 0.0005,
      "loss": 3.0034732818603516,
      "step": 2111
    },
    {
      "epoch": 0.48274285714285714,
      "grad_norm": 1.627895712852478,
      "learning_rate": 0.0005,
      "loss": 3.1620514392852783,
      "step": 2112
    },
    {
      "epoch": 0.48297142857142855,
      "grad_norm": 1.0088369846343994,
      "learning_rate": 0.0005,
      "loss": 2.6040239334106445,
      "step": 2113
    },
    {
      "epoch": 0.4832,
      "grad_norm": 1.061987042427063,
      "learning_rate": 0.0005,
      "loss": 2.4349398612976074,
      "step": 2114
    },
    {
      "epoch": 0.48342857142857143,
      "grad_norm": 1.6710022687911987,
      "learning_rate": 0.0005,
      "loss": 3.014380693435669,
      "step": 2115
    },
    {
      "epoch": 0.48365714285714284,
      "grad_norm": 1.9425413608551025,
      "learning_rate": 0.0005,
      "loss": 2.7757556438446045,
      "step": 2116
    },
    {
      "epoch": 0.4838857142857143,
      "grad_norm": 1.134801983833313,
      "learning_rate": 0.0005,
      "loss": 2.968808174133301,
      "step": 2117
    },
    {
      "epoch": 0.4841142857142857,
      "grad_norm": 1.0350747108459473,
      "learning_rate": 0.0005,
      "loss": 3.008267402648926,
      "step": 2118
    },
    {
      "epoch": 0.48434285714285713,
      "grad_norm": 1.2255898714065552,
      "learning_rate": 0.0005,
      "loss": 3.1568050384521484,
      "step": 2119
    },
    {
      "epoch": 0.4845714285714286,
      "grad_norm": 1.1064051389694214,
      "learning_rate": 0.0005,
      "loss": 2.898507833480835,
      "step": 2120
    },
    {
      "epoch": 0.4848,
      "grad_norm": 0.9665022492408752,
      "learning_rate": 0.0005,
      "loss": 3.191889762878418,
      "step": 2121
    },
    {
      "epoch": 0.4850285714285714,
      "grad_norm": 0.989814817905426,
      "learning_rate": 0.0005,
      "loss": 2.9744253158569336,
      "step": 2122
    },
    {
      "epoch": 0.48525714285714283,
      "grad_norm": 0.9860565066337585,
      "learning_rate": 0.0005,
      "loss": 2.9359023571014404,
      "step": 2123
    },
    {
      "epoch": 0.4854857142857143,
      "grad_norm": 0.9588626623153687,
      "learning_rate": 0.0005,
      "loss": 3.181155204772949,
      "step": 2124
    },
    {
      "epoch": 0.4857142857142857,
      "grad_norm": 0.9765832424163818,
      "learning_rate": 0.0005,
      "loss": 3.185567855834961,
      "step": 2125
    },
    {
      "epoch": 0.4859428571428571,
      "grad_norm": 1.0974575281143188,
      "learning_rate": 0.0005,
      "loss": 3.2167606353759766,
      "step": 2126
    },
    {
      "epoch": 0.4861714285714286,
      "grad_norm": 1.1242786645889282,
      "learning_rate": 0.0005,
      "loss": 2.8623456954956055,
      "step": 2127
    },
    {
      "epoch": 0.4864,
      "grad_norm": 1.2589905261993408,
      "learning_rate": 0.0005,
      "loss": 3.160273313522339,
      "step": 2128
    },
    {
      "epoch": 0.4866285714285714,
      "grad_norm": 1.840396761894226,
      "learning_rate": 0.0005,
      "loss": 2.5852530002593994,
      "step": 2129
    },
    {
      "epoch": 0.4868571428571429,
      "grad_norm": 1.108350157737732,
      "learning_rate": 0.0005,
      "loss": 2.634293556213379,
      "step": 2130
    },
    {
      "epoch": 0.4870857142857143,
      "grad_norm": 1.1544275283813477,
      "learning_rate": 0.0005,
      "loss": 2.938804864883423,
      "step": 2131
    },
    {
      "epoch": 0.4873142857142857,
      "grad_norm": 0.8607510924339294,
      "learning_rate": 0.0005,
      "loss": 2.8526930809020996,
      "step": 2132
    },
    {
      "epoch": 0.48754285714285717,
      "grad_norm": 1.2364009618759155,
      "learning_rate": 0.0005,
      "loss": 2.7536656856536865,
      "step": 2133
    },
    {
      "epoch": 0.4877714285714286,
      "grad_norm": 1.2053115367889404,
      "learning_rate": 0.0005,
      "loss": 2.7159688472747803,
      "step": 2134
    },
    {
      "epoch": 0.488,
      "grad_norm": 1.4148956537246704,
      "learning_rate": 0.0005,
      "loss": 3.014864206314087,
      "step": 2135
    },
    {
      "epoch": 0.48822857142857146,
      "grad_norm": 0.9026050567626953,
      "learning_rate": 0.0005,
      "loss": 2.9891109466552734,
      "step": 2136
    },
    {
      "epoch": 0.48845714285714287,
      "grad_norm": 1.0953882932662964,
      "learning_rate": 0.0005,
      "loss": 3.0003156661987305,
      "step": 2137
    },
    {
      "epoch": 0.4886857142857143,
      "grad_norm": 0.9974609017372131,
      "learning_rate": 0.0005,
      "loss": 2.778686046600342,
      "step": 2138
    },
    {
      "epoch": 0.4889142857142857,
      "grad_norm": 2.2241783142089844,
      "learning_rate": 0.0005,
      "loss": 2.735637664794922,
      "step": 2139
    },
    {
      "epoch": 0.48914285714285716,
      "grad_norm": 1.0890350341796875,
      "learning_rate": 0.0005,
      "loss": 3.011718273162842,
      "step": 2140
    },
    {
      "epoch": 0.48937142857142857,
      "grad_norm": 1.2154532670974731,
      "learning_rate": 0.0005,
      "loss": 2.9591469764709473,
      "step": 2141
    },
    {
      "epoch": 0.4896,
      "grad_norm": 1.175551414489746,
      "learning_rate": 0.0005,
      "loss": 2.5464954376220703,
      "step": 2142
    },
    {
      "epoch": 0.48982857142857145,
      "grad_norm": 1.07098388671875,
      "learning_rate": 0.0005,
      "loss": 3.0169434547424316,
      "step": 2143
    },
    {
      "epoch": 0.49005714285714286,
      "grad_norm": 1.2181975841522217,
      "learning_rate": 0.0005,
      "loss": 2.885268211364746,
      "step": 2144
    },
    {
      "epoch": 0.49028571428571427,
      "grad_norm": 1.3362799882888794,
      "learning_rate": 0.0005,
      "loss": 2.6671805381774902,
      "step": 2145
    },
    {
      "epoch": 0.49051428571428574,
      "grad_norm": 1.2663980722427368,
      "learning_rate": 0.0005,
      "loss": 3.2674007415771484,
      "step": 2146
    },
    {
      "epoch": 0.49074285714285715,
      "grad_norm": 0.9688859581947327,
      "learning_rate": 0.0005,
      "loss": 2.9133107662200928,
      "step": 2147
    },
    {
      "epoch": 0.49097142857142856,
      "grad_norm": 1.0902554988861084,
      "learning_rate": 0.0005,
      "loss": 3.1034929752349854,
      "step": 2148
    },
    {
      "epoch": 0.4912,
      "grad_norm": 1.2256486415863037,
      "learning_rate": 0.0005,
      "loss": 2.9383325576782227,
      "step": 2149
    },
    {
      "epoch": 0.49142857142857144,
      "grad_norm": 1.026684284210205,
      "learning_rate": 0.0005,
      "loss": 3.0134706497192383,
      "step": 2150
    },
    {
      "epoch": 0.49165714285714285,
      "grad_norm": 1.006814956665039,
      "learning_rate": 0.0005,
      "loss": 2.502112865447998,
      "step": 2151
    },
    {
      "epoch": 0.49188571428571426,
      "grad_norm": 1.0363036394119263,
      "learning_rate": 0.0005,
      "loss": 2.8968825340270996,
      "step": 2152
    },
    {
      "epoch": 0.4921142857142857,
      "grad_norm": 1.1378326416015625,
      "learning_rate": 0.0005,
      "loss": 2.704730987548828,
      "step": 2153
    },
    {
      "epoch": 0.49234285714285714,
      "grad_norm": 1.5658230781555176,
      "learning_rate": 0.0005,
      "loss": 2.85793137550354,
      "step": 2154
    },
    {
      "epoch": 0.49257142857142855,
      "grad_norm": 1.0961730480194092,
      "learning_rate": 0.0005,
      "loss": 2.734113931655884,
      "step": 2155
    },
    {
      "epoch": 0.4928,
      "grad_norm": 1.0800373554229736,
      "learning_rate": 0.0005,
      "loss": 3.3808631896972656,
      "step": 2156
    },
    {
      "epoch": 0.4930285714285714,
      "grad_norm": 1.1621310710906982,
      "learning_rate": 0.0005,
      "loss": 2.9297492504119873,
      "step": 2157
    },
    {
      "epoch": 0.49325714285714284,
      "grad_norm": 1.6015881299972534,
      "learning_rate": 0.0005,
      "loss": 3.2199413776397705,
      "step": 2158
    },
    {
      "epoch": 0.4934857142857143,
      "grad_norm": 1.3391470909118652,
      "learning_rate": 0.0005,
      "loss": 2.9474294185638428,
      "step": 2159
    },
    {
      "epoch": 0.4937142857142857,
      "grad_norm": 0.9985405206680298,
      "learning_rate": 0.0005,
      "loss": 3.1840672492980957,
      "step": 2160
    },
    {
      "epoch": 0.49394285714285713,
      "grad_norm": 1.2184892892837524,
      "learning_rate": 0.0005,
      "loss": 2.9494457244873047,
      "step": 2161
    },
    {
      "epoch": 0.4941714285714286,
      "grad_norm": 0.8461372256278992,
      "learning_rate": 0.0005,
      "loss": 3.0748395919799805,
      "step": 2162
    },
    {
      "epoch": 0.4944,
      "grad_norm": 1.4123731851577759,
      "learning_rate": 0.0005,
      "loss": 2.75413179397583,
      "step": 2163
    },
    {
      "epoch": 0.4946285714285714,
      "grad_norm": 1.377363681793213,
      "learning_rate": 0.0005,
      "loss": 2.9635965824127197,
      "step": 2164
    },
    {
      "epoch": 0.4948571428571429,
      "grad_norm": 1.1998416185379028,
      "learning_rate": 0.0005,
      "loss": 2.5921099185943604,
      "step": 2165
    },
    {
      "epoch": 0.4950857142857143,
      "grad_norm": 1.13015615940094,
      "learning_rate": 0.0005,
      "loss": 2.7302403450012207,
      "step": 2166
    },
    {
      "epoch": 0.4953142857142857,
      "grad_norm": 1.2819406986236572,
      "learning_rate": 0.0005,
      "loss": 2.8555164337158203,
      "step": 2167
    },
    {
      "epoch": 0.4955428571428571,
      "grad_norm": 2.138951539993286,
      "learning_rate": 0.0005,
      "loss": 3.0503411293029785,
      "step": 2168
    },
    {
      "epoch": 0.4957714285714286,
      "grad_norm": 1.3030381202697754,
      "learning_rate": 0.0005,
      "loss": 3.1666955947875977,
      "step": 2169
    },
    {
      "epoch": 0.496,
      "grad_norm": 1.0499732494354248,
      "learning_rate": 0.0005,
      "loss": 2.8077807426452637,
      "step": 2170
    },
    {
      "epoch": 0.4962285714285714,
      "grad_norm": 1.5451655387878418,
      "learning_rate": 0.0005,
      "loss": 2.7401137351989746,
      "step": 2171
    },
    {
      "epoch": 0.4964571428571429,
      "grad_norm": 1.328005075454712,
      "learning_rate": 0.0005,
      "loss": 2.779768705368042,
      "step": 2172
    },
    {
      "epoch": 0.4966857142857143,
      "grad_norm": 1.1149321794509888,
      "learning_rate": 0.0005,
      "loss": 2.967249870300293,
      "step": 2173
    },
    {
      "epoch": 0.4969142857142857,
      "grad_norm": 1.1005622148513794,
      "learning_rate": 0.0005,
      "loss": 2.9511547088623047,
      "step": 2174
    },
    {
      "epoch": 0.49714285714285716,
      "grad_norm": 0.8832131028175354,
      "learning_rate": 0.0005,
      "loss": 3.102189302444458,
      "step": 2175
    },
    {
      "epoch": 0.4973714285714286,
      "grad_norm": 1.170576572418213,
      "learning_rate": 0.0005,
      "loss": 2.940843343734741,
      "step": 2176
    },
    {
      "epoch": 0.4976,
      "grad_norm": 1.2730666399002075,
      "learning_rate": 0.0005,
      "loss": 2.931602954864502,
      "step": 2177
    },
    {
      "epoch": 0.49782857142857145,
      "grad_norm": 1.1192364692687988,
      "learning_rate": 0.0005,
      "loss": 2.7026312351226807,
      "step": 2178
    },
    {
      "epoch": 0.49805714285714286,
      "grad_norm": 1.0760563611984253,
      "learning_rate": 0.0005,
      "loss": 2.9616012573242188,
      "step": 2179
    },
    {
      "epoch": 0.4982857142857143,
      "grad_norm": 1.307946801185608,
      "learning_rate": 0.0005,
      "loss": 2.6676716804504395,
      "step": 2180
    },
    {
      "epoch": 0.4985142857142857,
      "grad_norm": 1.3282026052474976,
      "learning_rate": 0.0005,
      "loss": 2.609964370727539,
      "step": 2181
    },
    {
      "epoch": 0.49874285714285715,
      "grad_norm": 1.5440411567687988,
      "learning_rate": 0.0005,
      "loss": 3.0421462059020996,
      "step": 2182
    },
    {
      "epoch": 0.49897142857142857,
      "grad_norm": 0.9639237523078918,
      "learning_rate": 0.0005,
      "loss": 2.7217326164245605,
      "step": 2183
    },
    {
      "epoch": 0.4992,
      "grad_norm": 1.3062618970870972,
      "learning_rate": 0.0005,
      "loss": 2.781674385070801,
      "step": 2184
    },
    {
      "epoch": 0.49942857142857144,
      "grad_norm": 1.184789776802063,
      "learning_rate": 0.0005,
      "loss": 2.6005523204803467,
      "step": 2185
    },
    {
      "epoch": 0.49965714285714286,
      "grad_norm": 1.2787749767303467,
      "learning_rate": 0.0005,
      "loss": 2.7199082374572754,
      "step": 2186
    },
    {
      "epoch": 0.49988571428571427,
      "grad_norm": 0.9930563569068909,
      "learning_rate": 0.0005,
      "loss": 2.8147811889648438,
      "step": 2187
    },
    {
      "epoch": 0.5001142857142857,
      "grad_norm": 1.26620614528656,
      "learning_rate": 0.0005,
      "loss": 3.064439296722412,
      "step": 2188
    },
    {
      "epoch": 0.5003428571428571,
      "grad_norm": 1.087458610534668,
      "learning_rate": 0.0005,
      "loss": 2.964138984680176,
      "step": 2189
    },
    {
      "epoch": 0.5005714285714286,
      "grad_norm": 1.2487578392028809,
      "learning_rate": 0.0005,
      "loss": 2.7761852741241455,
      "step": 2190
    },
    {
      "epoch": 0.5008,
      "grad_norm": 1.00693941116333,
      "learning_rate": 0.0005,
      "loss": 2.9560189247131348,
      "step": 2191
    },
    {
      "epoch": 0.5010285714285714,
      "grad_norm": 1.0206936597824097,
      "learning_rate": 0.0005,
      "loss": 3.111145496368408,
      "step": 2192
    },
    {
      "epoch": 0.5012571428571428,
      "grad_norm": 1.2054109573364258,
      "learning_rate": 0.0005,
      "loss": 2.7390339374542236,
      "step": 2193
    },
    {
      "epoch": 0.5014857142857143,
      "grad_norm": 1.0435824394226074,
      "learning_rate": 0.0005,
      "loss": 2.785518169403076,
      "step": 2194
    },
    {
      "epoch": 0.5017142857142857,
      "grad_norm": 1.0357264280319214,
      "learning_rate": 0.0005,
      "loss": 2.8834304809570312,
      "step": 2195
    },
    {
      "epoch": 0.5019428571428571,
      "grad_norm": 2.677783250808716,
      "learning_rate": 0.0005,
      "loss": 2.939009666442871,
      "step": 2196
    },
    {
      "epoch": 0.5021714285714286,
      "grad_norm": 1.979522705078125,
      "learning_rate": 0.0005,
      "loss": 2.7329299449920654,
      "step": 2197
    },
    {
      "epoch": 0.5024,
      "grad_norm": 1.5181193351745605,
      "learning_rate": 0.0005,
      "loss": 2.775308847427368,
      "step": 2198
    },
    {
      "epoch": 0.5026285714285714,
      "grad_norm": 1.0765254497528076,
      "learning_rate": 0.0005,
      "loss": 2.736177921295166,
      "step": 2199
    },
    {
      "epoch": 0.5028571428571429,
      "grad_norm": 1.6994905471801758,
      "learning_rate": 0.0005,
      "loss": 2.9904861450195312,
      "step": 2200
    },
    {
      "epoch": 0.5030857142857142,
      "grad_norm": 1.3191559314727783,
      "learning_rate": 0.0005,
      "loss": 2.8254377841949463,
      "step": 2201
    },
    {
      "epoch": 0.5033142857142857,
      "grad_norm": 0.9815040230751038,
      "learning_rate": 0.0005,
      "loss": 2.6913352012634277,
      "step": 2202
    },
    {
      "epoch": 0.5035428571428572,
      "grad_norm": 1.0914289951324463,
      "learning_rate": 0.0005,
      "loss": 3.069772481918335,
      "step": 2203
    },
    {
      "epoch": 0.5037714285714285,
      "grad_norm": 1.6077556610107422,
      "learning_rate": 0.0005,
      "loss": 2.911868095397949,
      "step": 2204
    },
    {
      "epoch": 0.504,
      "grad_norm": 1.1764423847198486,
      "learning_rate": 0.0005,
      "loss": 2.8101935386657715,
      "step": 2205
    },
    {
      "epoch": 0.5042285714285715,
      "grad_norm": 1.7515404224395752,
      "learning_rate": 0.0005,
      "loss": 3.330897331237793,
      "step": 2206
    },
    {
      "epoch": 0.5044571428571428,
      "grad_norm": 1.4117426872253418,
      "learning_rate": 0.0005,
      "loss": 3.0281901359558105,
      "step": 2207
    },
    {
      "epoch": 0.5046857142857143,
      "grad_norm": 1.8313554525375366,
      "learning_rate": 0.0005,
      "loss": 2.918630599975586,
      "step": 2208
    },
    {
      "epoch": 0.5049142857142858,
      "grad_norm": 1.3527181148529053,
      "learning_rate": 0.0005,
      "loss": 3.4367218017578125,
      "step": 2209
    },
    {
      "epoch": 0.5051428571428571,
      "grad_norm": 1.258136510848999,
      "learning_rate": 0.0005,
      "loss": 2.5861010551452637,
      "step": 2210
    },
    {
      "epoch": 0.5053714285714286,
      "grad_norm": 1.3705403804779053,
      "learning_rate": 0.0005,
      "loss": 3.3429203033447266,
      "step": 2211
    },
    {
      "epoch": 0.5056,
      "grad_norm": 1.1187000274658203,
      "learning_rate": 0.0005,
      "loss": 2.959308624267578,
      "step": 2212
    },
    {
      "epoch": 0.5058285714285714,
      "grad_norm": 1.2959271669387817,
      "learning_rate": 0.0005,
      "loss": 2.8648409843444824,
      "step": 2213
    },
    {
      "epoch": 0.5060571428571429,
      "grad_norm": 1.316481113433838,
      "learning_rate": 0.0005,
      "loss": 3.3899173736572266,
      "step": 2214
    },
    {
      "epoch": 0.5062857142857143,
      "grad_norm": 1.0510532855987549,
      "learning_rate": 0.0005,
      "loss": 3.425361394882202,
      "step": 2215
    },
    {
      "epoch": 0.5065142857142857,
      "grad_norm": 1.2787821292877197,
      "learning_rate": 0.0005,
      "loss": 2.4203872680664062,
      "step": 2216
    },
    {
      "epoch": 0.5067428571428572,
      "grad_norm": 1.2358291149139404,
      "learning_rate": 0.0005,
      "loss": 2.7673423290252686,
      "step": 2217
    },
    {
      "epoch": 0.5069714285714285,
      "grad_norm": 1.1675043106079102,
      "learning_rate": 0.0005,
      "loss": 2.9870941638946533,
      "step": 2218
    },
    {
      "epoch": 0.5072,
      "grad_norm": 1.205219030380249,
      "learning_rate": 0.0005,
      "loss": 2.8732991218566895,
      "step": 2219
    },
    {
      "epoch": 0.5074285714285715,
      "grad_norm": 1.2602368593215942,
      "learning_rate": 0.0005,
      "loss": 2.9966018199920654,
      "step": 2220
    },
    {
      "epoch": 0.5076571428571428,
      "grad_norm": 1.2297805547714233,
      "learning_rate": 0.0005,
      "loss": 2.7185449600219727,
      "step": 2221
    },
    {
      "epoch": 0.5078857142857143,
      "grad_norm": 1.229736089706421,
      "learning_rate": 0.0005,
      "loss": 2.8169174194335938,
      "step": 2222
    },
    {
      "epoch": 0.5081142857142857,
      "grad_norm": 1.12477707862854,
      "learning_rate": 0.0005,
      "loss": 3.0252389907836914,
      "step": 2223
    },
    {
      "epoch": 0.5083428571428571,
      "grad_norm": 1.5354493856430054,
      "learning_rate": 0.0005,
      "loss": 2.7854177951812744,
      "step": 2224
    },
    {
      "epoch": 0.5085714285714286,
      "grad_norm": 2.141786813735962,
      "learning_rate": 0.0005,
      "loss": 3.19732403755188,
      "step": 2225
    },
    {
      "epoch": 0.5088,
      "grad_norm": 1.1039025783538818,
      "learning_rate": 0.0005,
      "loss": 3.013502597808838,
      "step": 2226
    },
    {
      "epoch": 0.5090285714285714,
      "grad_norm": 1.1825191974639893,
      "learning_rate": 0.0005,
      "loss": 2.9028263092041016,
      "step": 2227
    },
    {
      "epoch": 0.5092571428571429,
      "grad_norm": 1.0709627866744995,
      "learning_rate": 0.0005,
      "loss": 2.356444835662842,
      "step": 2228
    },
    {
      "epoch": 0.5094857142857143,
      "grad_norm": 0.9142615795135498,
      "learning_rate": 0.0005,
      "loss": 2.7494924068450928,
      "step": 2229
    },
    {
      "epoch": 0.5097142857142857,
      "grad_norm": 1.118710994720459,
      "learning_rate": 0.0005,
      "loss": 3.1397762298583984,
      "step": 2230
    },
    {
      "epoch": 0.5099428571428571,
      "grad_norm": 1.1802990436553955,
      "learning_rate": 0.0005,
      "loss": 2.941362142562866,
      "step": 2231
    },
    {
      "epoch": 0.5101714285714286,
      "grad_norm": 0.9576511383056641,
      "learning_rate": 0.0005,
      "loss": 2.6457877159118652,
      "step": 2232
    },
    {
      "epoch": 0.5104,
      "grad_norm": 1.2065554857254028,
      "learning_rate": 0.0005,
      "loss": 3.2672576904296875,
      "step": 2233
    },
    {
      "epoch": 0.5106285714285714,
      "grad_norm": 1.1170082092285156,
      "learning_rate": 0.0005,
      "loss": 2.843250274658203,
      "step": 2234
    },
    {
      "epoch": 0.5108571428571429,
      "grad_norm": 1.3490046262741089,
      "learning_rate": 0.0005,
      "loss": 2.5515596866607666,
      "step": 2235
    },
    {
      "epoch": 0.5110857142857143,
      "grad_norm": 0.897946834564209,
      "learning_rate": 0.0005,
      "loss": 3.052023410797119,
      "step": 2236
    },
    {
      "epoch": 0.5113142857142857,
      "grad_norm": 1.1528115272521973,
      "learning_rate": 0.0005,
      "loss": 2.9298505783081055,
      "step": 2237
    },
    {
      "epoch": 0.5115428571428572,
      "grad_norm": 1.2759685516357422,
      "learning_rate": 0.0005,
      "loss": 3.037059783935547,
      "step": 2238
    },
    {
      "epoch": 0.5117714285714285,
      "grad_norm": 2.3481969833374023,
      "learning_rate": 0.0005,
      "loss": 2.806366443634033,
      "step": 2239
    },
    {
      "epoch": 0.512,
      "grad_norm": 0.9838137626647949,
      "learning_rate": 0.0005,
      "loss": 2.900705099105835,
      "step": 2240
    },
    {
      "epoch": 0.5122285714285715,
      "grad_norm": 1.3351032733917236,
      "learning_rate": 0.0005,
      "loss": 2.9622702598571777,
      "step": 2241
    },
    {
      "epoch": 0.5124571428571428,
      "grad_norm": 1.1008498668670654,
      "learning_rate": 0.0005,
      "loss": 2.808263063430786,
      "step": 2242
    },
    {
      "epoch": 0.5126857142857143,
      "grad_norm": 0.9465318322181702,
      "learning_rate": 0.0005,
      "loss": 2.8910582065582275,
      "step": 2243
    },
    {
      "epoch": 0.5129142857142858,
      "grad_norm": 1.0929453372955322,
      "learning_rate": 0.0005,
      "loss": 3.019218683242798,
      "step": 2244
    },
    {
      "epoch": 0.5131428571428571,
      "grad_norm": 1.5734915733337402,
      "learning_rate": 0.0005,
      "loss": 2.8747425079345703,
      "step": 2245
    },
    {
      "epoch": 0.5133714285714286,
      "grad_norm": 1.2329927682876587,
      "learning_rate": 0.0005,
      "loss": 3.0534677505493164,
      "step": 2246
    },
    {
      "epoch": 0.5136,
      "grad_norm": 1.2094919681549072,
      "learning_rate": 0.0005,
      "loss": 2.6798205375671387,
      "step": 2247
    },
    {
      "epoch": 0.5138285714285714,
      "grad_norm": 1.4008370637893677,
      "learning_rate": 0.0005,
      "loss": 2.77504301071167,
      "step": 2248
    },
    {
      "epoch": 0.5140571428571429,
      "grad_norm": 0.9199010729789734,
      "learning_rate": 0.0005,
      "loss": 2.910994052886963,
      "step": 2249
    },
    {
      "epoch": 0.5142857142857142,
      "grad_norm": 1.2125234603881836,
      "learning_rate": 0.0005,
      "loss": 2.742030143737793,
      "step": 2250
    },
    {
      "epoch": 0.5145142857142857,
      "grad_norm": 1.1698532104492188,
      "learning_rate": 0.0005,
      "loss": 3.0854406356811523,
      "step": 2251
    },
    {
      "epoch": 0.5147428571428572,
      "grad_norm": 1.2482138872146606,
      "learning_rate": 0.0005,
      "loss": 2.584109306335449,
      "step": 2252
    },
    {
      "epoch": 0.5149714285714285,
      "grad_norm": 1.146425485610962,
      "learning_rate": 0.0005,
      "loss": 2.95815372467041,
      "step": 2253
    },
    {
      "epoch": 0.5152,
      "grad_norm": 1.1187580823898315,
      "learning_rate": 0.0005,
      "loss": 3.0993242263793945,
      "step": 2254
    },
    {
      "epoch": 0.5154285714285715,
      "grad_norm": 1.0099658966064453,
      "learning_rate": 0.0005,
      "loss": 3.103743076324463,
      "step": 2255
    },
    {
      "epoch": 0.5156571428571428,
      "grad_norm": 1.49092435836792,
      "learning_rate": 0.0005,
      "loss": 2.6847989559173584,
      "step": 2256
    },
    {
      "epoch": 0.5158857142857143,
      "grad_norm": 1.26321542263031,
      "learning_rate": 0.0005,
      "loss": 2.626406669616699,
      "step": 2257
    },
    {
      "epoch": 0.5161142857142857,
      "grad_norm": 0.9596143960952759,
      "learning_rate": 0.0005,
      "loss": 2.450892448425293,
      "step": 2258
    },
    {
      "epoch": 0.5163428571428571,
      "grad_norm": 0.9765304923057556,
      "learning_rate": 0.0005,
      "loss": 2.7014594078063965,
      "step": 2259
    },
    {
      "epoch": 0.5165714285714286,
      "grad_norm": 1.553956389427185,
      "learning_rate": 0.0005,
      "loss": 2.904545545578003,
      "step": 2260
    },
    {
      "epoch": 0.5168,
      "grad_norm": 0.8841590881347656,
      "learning_rate": 0.0005,
      "loss": 3.007059097290039,
      "step": 2261
    },
    {
      "epoch": 0.5170285714285714,
      "grad_norm": 1.101099967956543,
      "learning_rate": 0.0005,
      "loss": 2.7518317699432373,
      "step": 2262
    },
    {
      "epoch": 0.5172571428571429,
      "grad_norm": 1.2401846647262573,
      "learning_rate": 0.0005,
      "loss": 2.62814998626709,
      "step": 2263
    },
    {
      "epoch": 0.5174857142857143,
      "grad_norm": 1.0906975269317627,
      "learning_rate": 0.0005,
      "loss": 2.7818901538848877,
      "step": 2264
    },
    {
      "epoch": 0.5177142857142857,
      "grad_norm": 1.1118040084838867,
      "learning_rate": 0.0005,
      "loss": 2.8329100608825684,
      "step": 2265
    },
    {
      "epoch": 0.5179428571428571,
      "grad_norm": 1.3465710878372192,
      "learning_rate": 0.0005,
      "loss": 2.8109612464904785,
      "step": 2266
    },
    {
      "epoch": 0.5181714285714286,
      "grad_norm": 1.5130290985107422,
      "learning_rate": 0.0005,
      "loss": 2.715261459350586,
      "step": 2267
    },
    {
      "epoch": 0.5184,
      "grad_norm": 1.2257500886917114,
      "learning_rate": 0.0005,
      "loss": 2.810760021209717,
      "step": 2268
    },
    {
      "epoch": 0.5186285714285714,
      "grad_norm": 0.9792657494544983,
      "learning_rate": 0.0005,
      "loss": 3.189577341079712,
      "step": 2269
    },
    {
      "epoch": 0.5188571428571429,
      "grad_norm": 1.0624284744262695,
      "learning_rate": 0.0005,
      "loss": 2.677356481552124,
      "step": 2270
    },
    {
      "epoch": 0.5190857142857143,
      "grad_norm": 0.9267731308937073,
      "learning_rate": 0.0005,
      "loss": 3.0198545455932617,
      "step": 2271
    },
    {
      "epoch": 0.5193142857142857,
      "grad_norm": 1.532902717590332,
      "learning_rate": 0.0005,
      "loss": 3.1008944511413574,
      "step": 2272
    },
    {
      "epoch": 0.5195428571428572,
      "grad_norm": 1.041614055633545,
      "learning_rate": 0.0005,
      "loss": 2.869473695755005,
      "step": 2273
    },
    {
      "epoch": 0.5197714285714286,
      "grad_norm": 2.542592763900757,
      "learning_rate": 0.0005,
      "loss": 2.762134313583374,
      "step": 2274
    },
    {
      "epoch": 0.52,
      "grad_norm": 1.173840880393982,
      "learning_rate": 0.0005,
      "loss": 3.214323043823242,
      "step": 2275
    },
    {
      "epoch": 0.5202285714285714,
      "grad_norm": 0.9963619709014893,
      "learning_rate": 0.0005,
      "loss": 2.6838083267211914,
      "step": 2276
    },
    {
      "epoch": 0.5204571428571428,
      "grad_norm": 1.457248330116272,
      "learning_rate": 0.0005,
      "loss": 3.0030999183654785,
      "step": 2277
    },
    {
      "epoch": 0.5206857142857143,
      "grad_norm": 1.2575262784957886,
      "learning_rate": 0.0005,
      "loss": 3.063668727874756,
      "step": 2278
    },
    {
      "epoch": 0.5209142857142857,
      "grad_norm": 1.3968735933303833,
      "learning_rate": 0.0005,
      "loss": 3.7496094703674316,
      "step": 2279
    },
    {
      "epoch": 0.5211428571428571,
      "grad_norm": 1.8294477462768555,
      "learning_rate": 0.0005,
      "loss": 2.942953586578369,
      "step": 2280
    },
    {
      "epoch": 0.5213714285714286,
      "grad_norm": 1.1347235441207886,
      "learning_rate": 0.0005,
      "loss": 3.0299410820007324,
      "step": 2281
    },
    {
      "epoch": 0.5216,
      "grad_norm": 0.8524429798126221,
      "learning_rate": 0.0005,
      "loss": 2.6055173873901367,
      "step": 2282
    },
    {
      "epoch": 0.5218285714285714,
      "grad_norm": 0.9735625386238098,
      "learning_rate": 0.0005,
      "loss": 3.0647192001342773,
      "step": 2283
    },
    {
      "epoch": 0.5220571428571429,
      "grad_norm": 1.03725004196167,
      "learning_rate": 0.0005,
      "loss": 2.7529375553131104,
      "step": 2284
    },
    {
      "epoch": 0.5222857142857142,
      "grad_norm": 1.0319408178329468,
      "learning_rate": 0.0005,
      "loss": 2.796015739440918,
      "step": 2285
    },
    {
      "epoch": 0.5225142857142857,
      "grad_norm": 0.9797798991203308,
      "learning_rate": 0.0005,
      "loss": 2.9722254276275635,
      "step": 2286
    },
    {
      "epoch": 0.5227428571428572,
      "grad_norm": 1.016547679901123,
      "learning_rate": 0.0005,
      "loss": 3.175347328186035,
      "step": 2287
    },
    {
      "epoch": 0.5229714285714285,
      "grad_norm": 4.910841464996338,
      "learning_rate": 0.0005,
      "loss": 2.6637449264526367,
      "step": 2288
    },
    {
      "epoch": 0.5232,
      "grad_norm": 1.2331172227859497,
      "learning_rate": 0.0005,
      "loss": 3.054776430130005,
      "step": 2289
    },
    {
      "epoch": 0.5234285714285715,
      "grad_norm": 1.0850924253463745,
      "learning_rate": 0.0005,
      "loss": 3.0862064361572266,
      "step": 2290
    },
    {
      "epoch": 0.5236571428571428,
      "grad_norm": 1.3584768772125244,
      "learning_rate": 0.0005,
      "loss": 3.067441463470459,
      "step": 2291
    },
    {
      "epoch": 0.5238857142857143,
      "grad_norm": 1.3919991254806519,
      "learning_rate": 0.0005,
      "loss": 2.8864712715148926,
      "step": 2292
    },
    {
      "epoch": 0.5241142857142858,
      "grad_norm": 0.8273470997810364,
      "learning_rate": 0.0005,
      "loss": 2.577711582183838,
      "step": 2293
    },
    {
      "epoch": 0.5243428571428571,
      "grad_norm": 1.0425598621368408,
      "learning_rate": 0.0005,
      "loss": 2.82565975189209,
      "step": 2294
    },
    {
      "epoch": 0.5245714285714286,
      "grad_norm": 0.8476508855819702,
      "learning_rate": 0.0005,
      "loss": 2.7813990116119385,
      "step": 2295
    },
    {
      "epoch": 0.5248,
      "grad_norm": 0.9509479403495789,
      "learning_rate": 0.0005,
      "loss": 2.7774734497070312,
      "step": 2296
    },
    {
      "epoch": 0.5250285714285714,
      "grad_norm": 1.7850254774093628,
      "learning_rate": 0.0005,
      "loss": 2.7255444526672363,
      "step": 2297
    },
    {
      "epoch": 0.5252571428571429,
      "grad_norm": 1.2460561990737915,
      "learning_rate": 0.0005,
      "loss": 2.7475247383117676,
      "step": 2298
    },
    {
      "epoch": 0.5254857142857143,
      "grad_norm": 1.1456665992736816,
      "learning_rate": 0.0005,
      "loss": 2.9704136848449707,
      "step": 2299
    },
    {
      "epoch": 0.5257142857142857,
      "grad_norm": 0.921469509601593,
      "learning_rate": 0.0005,
      "loss": 2.8085622787475586,
      "step": 2300
    },
    {
      "epoch": 0.5259428571428572,
      "grad_norm": 1.1919097900390625,
      "learning_rate": 0.0005,
      "loss": 2.6433091163635254,
      "step": 2301
    },
    {
      "epoch": 0.5261714285714286,
      "grad_norm": 1.0512806177139282,
      "learning_rate": 0.0005,
      "loss": 2.6684110164642334,
      "step": 2302
    },
    {
      "epoch": 0.5264,
      "grad_norm": 1.0523476600646973,
      "learning_rate": 0.0005,
      "loss": 2.763711452484131,
      "step": 2303
    },
    {
      "epoch": 0.5266285714285714,
      "grad_norm": 1.805392861366272,
      "learning_rate": 0.0005,
      "loss": 3.181675910949707,
      "step": 2304
    },
    {
      "epoch": 0.5268571428571428,
      "grad_norm": 1.2150219678878784,
      "learning_rate": 0.0005,
      "loss": 2.4997267723083496,
      "step": 2305
    },
    {
      "epoch": 0.5270857142857143,
      "grad_norm": 1.320425033569336,
      "learning_rate": 0.0005,
      "loss": 3.010371208190918,
      "step": 2306
    },
    {
      "epoch": 0.5273142857142857,
      "grad_norm": 1.1608902215957642,
      "learning_rate": 0.0005,
      "loss": 3.014477014541626,
      "step": 2307
    },
    {
      "epoch": 0.5275428571428571,
      "grad_norm": 1.1953586339950562,
      "learning_rate": 0.0005,
      "loss": 3.0440008640289307,
      "step": 2308
    },
    {
      "epoch": 0.5277714285714286,
      "grad_norm": 1.1195183992385864,
      "learning_rate": 0.0005,
      "loss": 3.068117141723633,
      "step": 2309
    },
    {
      "epoch": 0.528,
      "grad_norm": 1.3944482803344727,
      "learning_rate": 0.0005,
      "loss": 2.5937905311584473,
      "step": 2310
    },
    {
      "epoch": 0.5282285714285714,
      "grad_norm": 1.0009676218032837,
      "learning_rate": 0.0005,
      "loss": 2.540524482727051,
      "step": 2311
    },
    {
      "epoch": 0.5284571428571428,
      "grad_norm": 0.8667576313018799,
      "learning_rate": 0.0005,
      "loss": 2.5600085258483887,
      "step": 2312
    },
    {
      "epoch": 0.5286857142857143,
      "grad_norm": 0.8265567421913147,
      "learning_rate": 0.0005,
      "loss": 2.762564182281494,
      "step": 2313
    },
    {
      "epoch": 0.5289142857142857,
      "grad_norm": 0.9034696817398071,
      "learning_rate": 0.0005,
      "loss": 2.569897413253784,
      "step": 2314
    },
    {
      "epoch": 0.5291428571428571,
      "grad_norm": 1.2535371780395508,
      "learning_rate": 0.0005,
      "loss": 2.921856164932251,
      "step": 2315
    },
    {
      "epoch": 0.5293714285714286,
      "grad_norm": 1.0357974767684937,
      "learning_rate": 0.0005,
      "loss": 2.6120150089263916,
      "step": 2316
    },
    {
      "epoch": 0.5296,
      "grad_norm": 1.196892261505127,
      "learning_rate": 0.0005,
      "loss": 2.501011610031128,
      "step": 2317
    },
    {
      "epoch": 0.5298285714285714,
      "grad_norm": 1.2543731927871704,
      "learning_rate": 0.0005,
      "loss": 2.9814043045043945,
      "step": 2318
    },
    {
      "epoch": 0.5300571428571429,
      "grad_norm": 1.0521094799041748,
      "learning_rate": 0.0005,
      "loss": 2.6471238136291504,
      "step": 2319
    },
    {
      "epoch": 0.5302857142857142,
      "grad_norm": 1.4404727220535278,
      "learning_rate": 0.0005,
      "loss": 3.086440086364746,
      "step": 2320
    },
    {
      "epoch": 0.5305142857142857,
      "grad_norm": 1.0892621278762817,
      "learning_rate": 0.0005,
      "loss": 3.0075838565826416,
      "step": 2321
    },
    {
      "epoch": 0.5307428571428572,
      "grad_norm": 1.2518662214279175,
      "learning_rate": 0.0005,
      "loss": 2.8125104904174805,
      "step": 2322
    },
    {
      "epoch": 0.5309714285714285,
      "grad_norm": 1.0972055196762085,
      "learning_rate": 0.0005,
      "loss": 2.9284610748291016,
      "step": 2323
    },
    {
      "epoch": 0.5312,
      "grad_norm": 1.043396234512329,
      "learning_rate": 0.0005,
      "loss": 3.0609443187713623,
      "step": 2324
    },
    {
      "epoch": 0.5314285714285715,
      "grad_norm": 0.9953218102455139,
      "learning_rate": 0.0005,
      "loss": 2.643761157989502,
      "step": 2325
    },
    {
      "epoch": 0.5316571428571428,
      "grad_norm": 1.2932618856430054,
      "learning_rate": 0.0005,
      "loss": 2.438591480255127,
      "step": 2326
    },
    {
      "epoch": 0.5318857142857143,
      "grad_norm": 1.2396037578582764,
      "learning_rate": 0.0005,
      "loss": 2.8505759239196777,
      "step": 2327
    },
    {
      "epoch": 0.5321142857142858,
      "grad_norm": 1.1306699514389038,
      "learning_rate": 0.0005,
      "loss": 2.7068886756896973,
      "step": 2328
    },
    {
      "epoch": 0.5323428571428571,
      "grad_norm": 1.1648987531661987,
      "learning_rate": 0.0005,
      "loss": 2.8057734966278076,
      "step": 2329
    },
    {
      "epoch": 0.5325714285714286,
      "grad_norm": 1.1138622760772705,
      "learning_rate": 0.0005,
      "loss": 2.850160598754883,
      "step": 2330
    },
    {
      "epoch": 0.5328,
      "grad_norm": 1.007527470588684,
      "learning_rate": 0.0005,
      "loss": 3.1688175201416016,
      "step": 2331
    },
    {
      "epoch": 0.5330285714285714,
      "grad_norm": 1.2494838237762451,
      "learning_rate": 0.0005,
      "loss": 2.5782408714294434,
      "step": 2332
    },
    {
      "epoch": 0.5332571428571429,
      "grad_norm": 0.98566073179245,
      "learning_rate": 0.0005,
      "loss": 2.8809080123901367,
      "step": 2333
    },
    {
      "epoch": 0.5334857142857143,
      "grad_norm": 1.1162209510803223,
      "learning_rate": 0.0005,
      "loss": 2.8471004962921143,
      "step": 2334
    },
    {
      "epoch": 0.5337142857142857,
      "grad_norm": 1.2914553880691528,
      "learning_rate": 0.0005,
      "loss": 2.6473541259765625,
      "step": 2335
    },
    {
      "epoch": 0.5339428571428572,
      "grad_norm": 1.1533007621765137,
      "learning_rate": 0.0005,
      "loss": 2.825568199157715,
      "step": 2336
    },
    {
      "epoch": 0.5341714285714285,
      "grad_norm": 1.197556972503662,
      "learning_rate": 0.0005,
      "loss": 2.508902072906494,
      "step": 2337
    },
    {
      "epoch": 0.5344,
      "grad_norm": 1.372479796409607,
      "learning_rate": 0.0005,
      "loss": 2.7820000648498535,
      "step": 2338
    },
    {
      "epoch": 0.5346285714285715,
      "grad_norm": 0.980650782585144,
      "learning_rate": 0.0005,
      "loss": 2.797452211380005,
      "step": 2339
    },
    {
      "epoch": 0.5348571428571428,
      "grad_norm": 1.3568557500839233,
      "learning_rate": 0.0005,
      "loss": 2.7339372634887695,
      "step": 2340
    },
    {
      "epoch": 0.5350857142857143,
      "grad_norm": 1.6078288555145264,
      "learning_rate": 0.0005,
      "loss": 2.332304000854492,
      "step": 2341
    },
    {
      "epoch": 0.5353142857142857,
      "grad_norm": 1.0202265977859497,
      "learning_rate": 0.0005,
      "loss": 2.8736765384674072,
      "step": 2342
    },
    {
      "epoch": 0.5355428571428571,
      "grad_norm": 3.145472288131714,
      "learning_rate": 0.0005,
      "loss": 3.0693163871765137,
      "step": 2343
    },
    {
      "epoch": 0.5357714285714286,
      "grad_norm": 1.1138503551483154,
      "learning_rate": 0.0005,
      "loss": 3.0663633346557617,
      "step": 2344
    },
    {
      "epoch": 0.536,
      "grad_norm": 1.2239305973052979,
      "learning_rate": 0.0005,
      "loss": 3.0679678916931152,
      "step": 2345
    },
    {
      "epoch": 0.5362285714285714,
      "grad_norm": 1.00180983543396,
      "learning_rate": 0.0005,
      "loss": 2.6787073612213135,
      "step": 2346
    },
    {
      "epoch": 0.5364571428571429,
      "grad_norm": 1.283716082572937,
      "learning_rate": 0.0005,
      "loss": 2.953575611114502,
      "step": 2347
    },
    {
      "epoch": 0.5366857142857143,
      "grad_norm": 1.0913519859313965,
      "learning_rate": 0.0005,
      "loss": 2.8432679176330566,
      "step": 2348
    },
    {
      "epoch": 0.5369142857142857,
      "grad_norm": 1.7029579877853394,
      "learning_rate": 0.0005,
      "loss": 3.0864710807800293,
      "step": 2349
    },
    {
      "epoch": 0.5371428571428571,
      "grad_norm": 1.083850383758545,
      "learning_rate": 0.0005,
      "loss": 2.8018150329589844,
      "step": 2350
    },
    {
      "epoch": 0.5373714285714286,
      "grad_norm": 0.9853575229644775,
      "learning_rate": 0.0005,
      "loss": 2.609340190887451,
      "step": 2351
    },
    {
      "epoch": 0.5376,
      "grad_norm": 1.897676706314087,
      "learning_rate": 0.0005,
      "loss": 3.0458850860595703,
      "step": 2352
    },
    {
      "epoch": 0.5378285714285714,
      "grad_norm": 1.092680811882019,
      "learning_rate": 0.0005,
      "loss": 2.7061550617218018,
      "step": 2353
    },
    {
      "epoch": 0.5380571428571429,
      "grad_norm": 0.9781025648117065,
      "learning_rate": 0.0005,
      "loss": 2.7401819229125977,
      "step": 2354
    },
    {
      "epoch": 0.5382857142857143,
      "grad_norm": 1.1701653003692627,
      "learning_rate": 0.0005,
      "loss": 3.0111658573150635,
      "step": 2355
    },
    {
      "epoch": 0.5385142857142857,
      "grad_norm": 1.4891239404678345,
      "learning_rate": 0.0005,
      "loss": 2.5961880683898926,
      "step": 2356
    },
    {
      "epoch": 0.5387428571428572,
      "grad_norm": 1.0203198194503784,
      "learning_rate": 0.0005,
      "loss": 2.8762927055358887,
      "step": 2357
    },
    {
      "epoch": 0.5389714285714285,
      "grad_norm": 1.08074152469635,
      "learning_rate": 0.0005,
      "loss": 2.929867744445801,
      "step": 2358
    },
    {
      "epoch": 0.5392,
      "grad_norm": 0.9950660467147827,
      "learning_rate": 0.0005,
      "loss": 2.6724653244018555,
      "step": 2359
    },
    {
      "epoch": 0.5394285714285715,
      "grad_norm": 1.079459309577942,
      "learning_rate": 0.0005,
      "loss": 3.100395441055298,
      "step": 2360
    },
    {
      "epoch": 0.5396571428571428,
      "grad_norm": 0.8820501565933228,
      "learning_rate": 0.0005,
      "loss": 2.506866693496704,
      "step": 2361
    },
    {
      "epoch": 0.5398857142857143,
      "grad_norm": 1.0452287197113037,
      "learning_rate": 0.0005,
      "loss": 2.841625452041626,
      "step": 2362
    },
    {
      "epoch": 0.5401142857142858,
      "grad_norm": 1.3115507364273071,
      "learning_rate": 0.0005,
      "loss": 3.095167875289917,
      "step": 2363
    },
    {
      "epoch": 0.5403428571428571,
      "grad_norm": 0.9970411062240601,
      "learning_rate": 0.0005,
      "loss": 2.8448143005371094,
      "step": 2364
    },
    {
      "epoch": 0.5405714285714286,
      "grad_norm": 1.9352970123291016,
      "learning_rate": 0.0005,
      "loss": 3.0664496421813965,
      "step": 2365
    },
    {
      "epoch": 0.5408,
      "grad_norm": 0.9522634148597717,
      "learning_rate": 0.0005,
      "loss": 2.7539892196655273,
      "step": 2366
    },
    {
      "epoch": 0.5410285714285714,
      "grad_norm": 1.2160826921463013,
      "learning_rate": 0.0005,
      "loss": 2.86796236038208,
      "step": 2367
    },
    {
      "epoch": 0.5412571428571429,
      "grad_norm": 0.9042659997940063,
      "learning_rate": 0.0005,
      "loss": 2.62961745262146,
      "step": 2368
    },
    {
      "epoch": 0.5414857142857142,
      "grad_norm": 1.0605031251907349,
      "learning_rate": 0.0005,
      "loss": 2.4627418518066406,
      "step": 2369
    },
    {
      "epoch": 0.5417142857142857,
      "grad_norm": 2.1240427494049072,
      "learning_rate": 0.0005,
      "loss": 2.865752696990967,
      "step": 2370
    },
    {
      "epoch": 0.5419428571428572,
      "grad_norm": 1.2296141386032104,
      "learning_rate": 0.0005,
      "loss": 3.0422568321228027,
      "step": 2371
    },
    {
      "epoch": 0.5421714285714285,
      "grad_norm": 1.4101399183273315,
      "learning_rate": 0.0005,
      "loss": 2.7590291500091553,
      "step": 2372
    },
    {
      "epoch": 0.5424,
      "grad_norm": 1.1107518672943115,
      "learning_rate": 0.0005,
      "loss": 2.7693986892700195,
      "step": 2373
    },
    {
      "epoch": 0.5426285714285715,
      "grad_norm": 0.8666242361068726,
      "learning_rate": 0.0005,
      "loss": 2.9429519176483154,
      "step": 2374
    },
    {
      "epoch": 0.5428571428571428,
      "grad_norm": 1.2056970596313477,
      "learning_rate": 0.0005,
      "loss": 2.866971969604492,
      "step": 2375
    },
    {
      "epoch": 0.5430857142857143,
      "grad_norm": 1.2211532592773438,
      "learning_rate": 0.0005,
      "loss": 2.6554410457611084,
      "step": 2376
    },
    {
      "epoch": 0.5433142857142857,
      "grad_norm": 1.0387159585952759,
      "learning_rate": 0.0005,
      "loss": 2.9133801460266113,
      "step": 2377
    },
    {
      "epoch": 0.5435428571428571,
      "grad_norm": 1.7164644002914429,
      "learning_rate": 0.0005,
      "loss": 2.980454444885254,
      "step": 2378
    },
    {
      "epoch": 0.5437714285714286,
      "grad_norm": 1.5648506879806519,
      "learning_rate": 0.0005,
      "loss": 2.7752819061279297,
      "step": 2379
    },
    {
      "epoch": 0.544,
      "grad_norm": 1.1614811420440674,
      "learning_rate": 0.0005,
      "loss": 2.4990437030792236,
      "step": 2380
    },
    {
      "epoch": 0.5442285714285714,
      "grad_norm": 1.3805689811706543,
      "learning_rate": 0.0005,
      "loss": 2.7586669921875,
      "step": 2381
    },
    {
      "epoch": 0.5444571428571429,
      "grad_norm": 1.0469841957092285,
      "learning_rate": 0.0005,
      "loss": 2.9292006492614746,
      "step": 2382
    },
    {
      "epoch": 0.5446857142857143,
      "grad_norm": 1.047481656074524,
      "learning_rate": 0.0005,
      "loss": 2.6046204566955566,
      "step": 2383
    },
    {
      "epoch": 0.5449142857142857,
      "grad_norm": 1.1044118404388428,
      "learning_rate": 0.0005,
      "loss": 2.967336893081665,
      "step": 2384
    },
    {
      "epoch": 0.5451428571428572,
      "grad_norm": 1.4267038106918335,
      "learning_rate": 0.0005,
      "loss": 3.911120891571045,
      "step": 2385
    },
    {
      "epoch": 0.5453714285714286,
      "grad_norm": 1.3023123741149902,
      "learning_rate": 0.0005,
      "loss": 2.5905380249023438,
      "step": 2386
    },
    {
      "epoch": 0.5456,
      "grad_norm": 0.8859487771987915,
      "learning_rate": 0.0005,
      "loss": 2.8588459491729736,
      "step": 2387
    },
    {
      "epoch": 0.5458285714285714,
      "grad_norm": 1.2277289628982544,
      "learning_rate": 0.0005,
      "loss": 3.2615203857421875,
      "step": 2388
    },
    {
      "epoch": 0.5460571428571429,
      "grad_norm": 1.0937869548797607,
      "learning_rate": 0.0005,
      "loss": 2.8311214447021484,
      "step": 2389
    },
    {
      "epoch": 0.5462857142857143,
      "grad_norm": 0.9980322122573853,
      "learning_rate": 0.0005,
      "loss": 3.099616050720215,
      "step": 2390
    },
    {
      "epoch": 0.5465142857142857,
      "grad_norm": 1.5096570253372192,
      "learning_rate": 0.0005,
      "loss": 3.4401676654815674,
      "step": 2391
    },
    {
      "epoch": 0.5467428571428572,
      "grad_norm": 1.226395845413208,
      "learning_rate": 0.0005,
      "loss": 2.8233044147491455,
      "step": 2392
    },
    {
      "epoch": 0.5469714285714286,
      "grad_norm": 1.103659749031067,
      "learning_rate": 0.0005,
      "loss": 2.721977710723877,
      "step": 2393
    },
    {
      "epoch": 0.5472,
      "grad_norm": 0.8914397954940796,
      "learning_rate": 0.0005,
      "loss": 2.5642714500427246,
      "step": 2394
    },
    {
      "epoch": 0.5474285714285714,
      "grad_norm": 0.975033164024353,
      "learning_rate": 0.0005,
      "loss": 2.9694814682006836,
      "step": 2395
    },
    {
      "epoch": 0.5476571428571428,
      "grad_norm": 0.9214494824409485,
      "learning_rate": 0.0005,
      "loss": 2.8211655616760254,
      "step": 2396
    },
    {
      "epoch": 0.5478857142857143,
      "grad_norm": 1.223818063735962,
      "learning_rate": 0.0005,
      "loss": 2.8794994354248047,
      "step": 2397
    },
    {
      "epoch": 0.5481142857142857,
      "grad_norm": 1.5301364660263062,
      "learning_rate": 0.0005,
      "loss": 2.8099989891052246,
      "step": 2398
    },
    {
      "epoch": 0.5483428571428571,
      "grad_norm": 0.8609481453895569,
      "learning_rate": 0.0005,
      "loss": 2.615920066833496,
      "step": 2399
    },
    {
      "epoch": 0.5485714285714286,
      "grad_norm": 1.0191479921340942,
      "learning_rate": 0.0005,
      "loss": 2.902961492538452,
      "step": 2400
    },
    {
      "epoch": 0.5488,
      "grad_norm": 1.0757558345794678,
      "learning_rate": 0.0005,
      "loss": 2.809776782989502,
      "step": 2401
    },
    {
      "epoch": 0.5490285714285714,
      "grad_norm": 1.3306710720062256,
      "learning_rate": 0.0005,
      "loss": 3.0726284980773926,
      "step": 2402
    },
    {
      "epoch": 0.5492571428571429,
      "grad_norm": 1.0826489925384521,
      "learning_rate": 0.0005,
      "loss": 3.0139622688293457,
      "step": 2403
    },
    {
      "epoch": 0.5494857142857142,
      "grad_norm": 1.5200437307357788,
      "learning_rate": 0.0005,
      "loss": 3.050940752029419,
      "step": 2404
    },
    {
      "epoch": 0.5497142857142857,
      "grad_norm": 1.1961097717285156,
      "learning_rate": 0.0005,
      "loss": 2.791792392730713,
      "step": 2405
    },
    {
      "epoch": 0.5499428571428572,
      "grad_norm": 0.9393734931945801,
      "learning_rate": 0.0005,
      "loss": 2.960160255432129,
      "step": 2406
    },
    {
      "epoch": 0.5501714285714285,
      "grad_norm": 1.0568232536315918,
      "learning_rate": 0.0005,
      "loss": 2.606109380722046,
      "step": 2407
    },
    {
      "epoch": 0.5504,
      "grad_norm": 1.1598303318023682,
      "learning_rate": 0.0005,
      "loss": 2.8611903190612793,
      "step": 2408
    },
    {
      "epoch": 0.5506285714285715,
      "grad_norm": 1.2960666418075562,
      "learning_rate": 0.0005,
      "loss": 2.8343396186828613,
      "step": 2409
    },
    {
      "epoch": 0.5508571428571428,
      "grad_norm": 1.1804485321044922,
      "learning_rate": 0.0005,
      "loss": 3.063138484954834,
      "step": 2410
    },
    {
      "epoch": 0.5510857142857143,
      "grad_norm": 1.0013585090637207,
      "learning_rate": 0.0005,
      "loss": 2.7269444465637207,
      "step": 2411
    },
    {
      "epoch": 0.5513142857142858,
      "grad_norm": 1.179592490196228,
      "learning_rate": 0.0005,
      "loss": 2.9025516510009766,
      "step": 2412
    },
    {
      "epoch": 0.5515428571428571,
      "grad_norm": 0.9979792237281799,
      "learning_rate": 0.0005,
      "loss": 2.7167794704437256,
      "step": 2413
    },
    {
      "epoch": 0.5517714285714286,
      "grad_norm": 1.0481884479522705,
      "learning_rate": 0.0005,
      "loss": 2.7746167182922363,
      "step": 2414
    },
    {
      "epoch": 0.552,
      "grad_norm": 1.0900462865829468,
      "learning_rate": 0.0005,
      "loss": 2.708580732345581,
      "step": 2415
    },
    {
      "epoch": 0.5522285714285714,
      "grad_norm": 1.100218415260315,
      "learning_rate": 0.0005,
      "loss": 2.975940227508545,
      "step": 2416
    },
    {
      "epoch": 0.5524571428571429,
      "grad_norm": 1.1770559549331665,
      "learning_rate": 0.0005,
      "loss": 2.529022216796875,
      "step": 2417
    },
    {
      "epoch": 0.5526857142857143,
      "grad_norm": 3.2121407985687256,
      "learning_rate": 0.0005,
      "loss": 2.82596492767334,
      "step": 2418
    },
    {
      "epoch": 0.5529142857142857,
      "grad_norm": 1.2571722269058228,
      "learning_rate": 0.0005,
      "loss": 3.144740581512451,
      "step": 2419
    },
    {
      "epoch": 0.5531428571428572,
      "grad_norm": 1.0147249698638916,
      "learning_rate": 0.0005,
      "loss": 2.8437161445617676,
      "step": 2420
    },
    {
      "epoch": 0.5533714285714286,
      "grad_norm": 1.3663948774337769,
      "learning_rate": 0.0005,
      "loss": 2.815568685531616,
      "step": 2421
    },
    {
      "epoch": 0.5536,
      "grad_norm": 1.2378647327423096,
      "learning_rate": 0.0005,
      "loss": 2.5944390296936035,
      "step": 2422
    },
    {
      "epoch": 0.5538285714285714,
      "grad_norm": 1.0220671892166138,
      "learning_rate": 0.0005,
      "loss": 2.7602715492248535,
      "step": 2423
    },
    {
      "epoch": 0.5540571428571428,
      "grad_norm": 1.1450986862182617,
      "learning_rate": 0.0005,
      "loss": 2.8737308979034424,
      "step": 2424
    },
    {
      "epoch": 0.5542857142857143,
      "grad_norm": 1.2467505931854248,
      "learning_rate": 0.0005,
      "loss": 2.3652191162109375,
      "step": 2425
    },
    {
      "epoch": 0.5545142857142857,
      "grad_norm": 1.186907172203064,
      "learning_rate": 0.0005,
      "loss": 2.8278250694274902,
      "step": 2426
    },
    {
      "epoch": 0.5547428571428571,
      "grad_norm": 0.9444205164909363,
      "learning_rate": 0.0005,
      "loss": 2.507387638092041,
      "step": 2427
    },
    {
      "epoch": 0.5549714285714286,
      "grad_norm": 1.0288714170455933,
      "learning_rate": 0.0005,
      "loss": 2.8631935119628906,
      "step": 2428
    },
    {
      "epoch": 0.5552,
      "grad_norm": 1.0584696531295776,
      "learning_rate": 0.0005,
      "loss": 2.7150721549987793,
      "step": 2429
    },
    {
      "epoch": 0.5554285714285714,
      "grad_norm": 1.1720750331878662,
      "learning_rate": 0.0005,
      "loss": 2.947399616241455,
      "step": 2430
    },
    {
      "epoch": 0.5556571428571428,
      "grad_norm": 0.9211481213569641,
      "learning_rate": 0.0005,
      "loss": 2.9664926528930664,
      "step": 2431
    },
    {
      "epoch": 0.5558857142857143,
      "grad_norm": 1.038833498954773,
      "learning_rate": 0.0005,
      "loss": 2.8770689964294434,
      "step": 2432
    },
    {
      "epoch": 0.5561142857142857,
      "grad_norm": 1.609456181526184,
      "learning_rate": 0.0005,
      "loss": 2.8440561294555664,
      "step": 2433
    },
    {
      "epoch": 0.5563428571428571,
      "grad_norm": 1.098931074142456,
      "learning_rate": 0.0005,
      "loss": 2.9380102157592773,
      "step": 2434
    },
    {
      "epoch": 0.5565714285714286,
      "grad_norm": 1.3180711269378662,
      "learning_rate": 0.0005,
      "loss": 2.48996639251709,
      "step": 2435
    },
    {
      "epoch": 0.5568,
      "grad_norm": 1.3691238164901733,
      "learning_rate": 0.0005,
      "loss": 2.7202072143554688,
      "step": 2436
    },
    {
      "epoch": 0.5570285714285714,
      "grad_norm": 1.0131725072860718,
      "learning_rate": 0.0005,
      "loss": 2.929358959197998,
      "step": 2437
    },
    {
      "epoch": 0.5572571428571429,
      "grad_norm": 0.8409262895584106,
      "learning_rate": 0.0005,
      "loss": 2.98461651802063,
      "step": 2438
    },
    {
      "epoch": 0.5574857142857143,
      "grad_norm": 0.9651982188224792,
      "learning_rate": 0.0005,
      "loss": 3.202310562133789,
      "step": 2439
    },
    {
      "epoch": 0.5577142857142857,
      "grad_norm": 1.1925654411315918,
      "learning_rate": 0.0005,
      "loss": 2.68837833404541,
      "step": 2440
    },
    {
      "epoch": 0.5579428571428572,
      "grad_norm": 1.126657485961914,
      "learning_rate": 0.0005,
      "loss": 3.01814603805542,
      "step": 2441
    },
    {
      "epoch": 0.5581714285714285,
      "grad_norm": 1.0265660285949707,
      "learning_rate": 0.0005,
      "loss": 2.7608752250671387,
      "step": 2442
    },
    {
      "epoch": 0.5584,
      "grad_norm": 1.3027302026748657,
      "learning_rate": 0.0005,
      "loss": 2.656996011734009,
      "step": 2443
    },
    {
      "epoch": 0.5586285714285715,
      "grad_norm": 1.6584821939468384,
      "learning_rate": 0.0005,
      "loss": 2.752786159515381,
      "step": 2444
    },
    {
      "epoch": 0.5588571428571428,
      "grad_norm": 1.4662657976150513,
      "learning_rate": 0.0005,
      "loss": 2.627044677734375,
      "step": 2445
    },
    {
      "epoch": 0.5590857142857143,
      "grad_norm": 1.1107513904571533,
      "learning_rate": 0.0005,
      "loss": 2.811903953552246,
      "step": 2446
    },
    {
      "epoch": 0.5593142857142858,
      "grad_norm": 1.0800666809082031,
      "learning_rate": 0.0005,
      "loss": 2.7484993934631348,
      "step": 2447
    },
    {
      "epoch": 0.5595428571428571,
      "grad_norm": 1.2942233085632324,
      "learning_rate": 0.0005,
      "loss": 2.925264358520508,
      "step": 2448
    },
    {
      "epoch": 0.5597714285714286,
      "grad_norm": 1.1515365839004517,
      "learning_rate": 0.0005,
      "loss": 2.8191661834716797,
      "step": 2449
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.0621240139007568,
      "learning_rate": 0.0005,
      "loss": 2.8168482780456543,
      "step": 2450
    },
    {
      "epoch": 0.5602285714285714,
      "grad_norm": 1.0926040410995483,
      "learning_rate": 0.0005,
      "loss": 2.670497179031372,
      "step": 2451
    },
    {
      "epoch": 0.5604571428571429,
      "grad_norm": 1.0487629175186157,
      "learning_rate": 0.0005,
      "loss": 2.7432620525360107,
      "step": 2452
    },
    {
      "epoch": 0.5606857142857142,
      "grad_norm": 1.1874114274978638,
      "learning_rate": 0.0005,
      "loss": 2.8376595973968506,
      "step": 2453
    },
    {
      "epoch": 0.5609142857142857,
      "grad_norm": 0.9625793695449829,
      "learning_rate": 0.0005,
      "loss": 2.8254661560058594,
      "step": 2454
    },
    {
      "epoch": 0.5611428571428572,
      "grad_norm": 1.1082669496536255,
      "learning_rate": 0.0005,
      "loss": 2.689910888671875,
      "step": 2455
    },
    {
      "epoch": 0.5613714285714285,
      "grad_norm": 0.9866765737533569,
      "learning_rate": 0.0005,
      "loss": 2.5841431617736816,
      "step": 2456
    },
    {
      "epoch": 0.5616,
      "grad_norm": 0.9758598208427429,
      "learning_rate": 0.0005,
      "loss": 2.585317373275757,
      "step": 2457
    },
    {
      "epoch": 0.5618285714285715,
      "grad_norm": 1.257479190826416,
      "learning_rate": 0.0005,
      "loss": 2.8544199466705322,
      "step": 2458
    },
    {
      "epoch": 0.5620571428571428,
      "grad_norm": 1.278424620628357,
      "learning_rate": 0.0005,
      "loss": 2.9239025115966797,
      "step": 2459
    },
    {
      "epoch": 0.5622857142857143,
      "grad_norm": 1.1312479972839355,
      "learning_rate": 0.0005,
      "loss": 2.78668212890625,
      "step": 2460
    },
    {
      "epoch": 0.5625142857142857,
      "grad_norm": 1.125455617904663,
      "learning_rate": 0.0005,
      "loss": 2.8266279697418213,
      "step": 2461
    },
    {
      "epoch": 0.5627428571428571,
      "grad_norm": 1.3686450719833374,
      "learning_rate": 0.0005,
      "loss": 2.469559669494629,
      "step": 2462
    },
    {
      "epoch": 0.5629714285714286,
      "grad_norm": 1.4246948957443237,
      "learning_rate": 0.0005,
      "loss": 2.9085264205932617,
      "step": 2463
    },
    {
      "epoch": 0.5632,
      "grad_norm": 1.4032986164093018,
      "learning_rate": 0.0005,
      "loss": 2.895310401916504,
      "step": 2464
    },
    {
      "epoch": 0.5634285714285714,
      "grad_norm": 1.22219979763031,
      "learning_rate": 0.0005,
      "loss": 2.6934237480163574,
      "step": 2465
    },
    {
      "epoch": 0.5636571428571429,
      "grad_norm": 1.3186321258544922,
      "learning_rate": 0.0005,
      "loss": 2.5078024864196777,
      "step": 2466
    },
    {
      "epoch": 0.5638857142857143,
      "grad_norm": 1.274919867515564,
      "learning_rate": 0.0005,
      "loss": 2.853695869445801,
      "step": 2467
    },
    {
      "epoch": 0.5641142857142857,
      "grad_norm": 1.088996171951294,
      "learning_rate": 0.0005,
      "loss": 2.7579240798950195,
      "step": 2468
    },
    {
      "epoch": 0.5643428571428571,
      "grad_norm": 1.001536250114441,
      "learning_rate": 0.0005,
      "loss": 3.1009557247161865,
      "step": 2469
    },
    {
      "epoch": 0.5645714285714286,
      "grad_norm": 1.1451138257980347,
      "learning_rate": 0.0005,
      "loss": 2.9117631912231445,
      "step": 2470
    },
    {
      "epoch": 0.5648,
      "grad_norm": 1.3621232509613037,
      "learning_rate": 0.0005,
      "loss": 2.7192330360412598,
      "step": 2471
    },
    {
      "epoch": 0.5650285714285714,
      "grad_norm": 1.1698012351989746,
      "learning_rate": 0.0005,
      "loss": 2.8947863578796387,
      "step": 2472
    },
    {
      "epoch": 0.5652571428571429,
      "grad_norm": 1.2566853761672974,
      "learning_rate": 0.0005,
      "loss": 2.88254976272583,
      "step": 2473
    },
    {
      "epoch": 0.5654857142857143,
      "grad_norm": 1.1771340370178223,
      "learning_rate": 0.0005,
      "loss": 2.4901294708251953,
      "step": 2474
    },
    {
      "epoch": 0.5657142857142857,
      "grad_norm": 0.909431517124176,
      "learning_rate": 0.0005,
      "loss": 2.834069013595581,
      "step": 2475
    },
    {
      "epoch": 0.5659428571428572,
      "grad_norm": 1.0662184953689575,
      "learning_rate": 0.0005,
      "loss": 2.891664505004883,
      "step": 2476
    },
    {
      "epoch": 0.5661714285714285,
      "grad_norm": 1.01483154296875,
      "learning_rate": 0.0005,
      "loss": 2.746246814727783,
      "step": 2477
    },
    {
      "epoch": 0.5664,
      "grad_norm": 0.9924781918525696,
      "learning_rate": 0.0005,
      "loss": 2.895620822906494,
      "step": 2478
    },
    {
      "epoch": 0.5666285714285715,
      "grad_norm": 0.8720110058784485,
      "learning_rate": 0.0005,
      "loss": 2.744494676589966,
      "step": 2479
    },
    {
      "epoch": 0.5668571428571428,
      "grad_norm": 1.5711382627487183,
      "learning_rate": 0.0005,
      "loss": 2.6128716468811035,
      "step": 2480
    },
    {
      "epoch": 0.5670857142857143,
      "grad_norm": 1.3627252578735352,
      "learning_rate": 0.0005,
      "loss": 2.842881202697754,
      "step": 2481
    },
    {
      "epoch": 0.5673142857142857,
      "grad_norm": 1.0850392580032349,
      "learning_rate": 0.0005,
      "loss": 2.492504358291626,
      "step": 2482
    },
    {
      "epoch": 0.5675428571428571,
      "grad_norm": 0.8934386372566223,
      "learning_rate": 0.0005,
      "loss": 2.971247434616089,
      "step": 2483
    },
    {
      "epoch": 0.5677714285714286,
      "grad_norm": 0.8581506609916687,
      "learning_rate": 0.0005,
      "loss": 2.857992172241211,
      "step": 2484
    },
    {
      "epoch": 0.568,
      "grad_norm": 0.9694806337356567,
      "learning_rate": 0.0005,
      "loss": 3.1430766582489014,
      "step": 2485
    },
    {
      "epoch": 0.5682285714285714,
      "grad_norm": 0.8162602186203003,
      "learning_rate": 0.0005,
      "loss": 3.164956569671631,
      "step": 2486
    },
    {
      "epoch": 0.5684571428571429,
      "grad_norm": 1.211520791053772,
      "learning_rate": 0.0005,
      "loss": 3.010584592819214,
      "step": 2487
    },
    {
      "epoch": 0.5686857142857142,
      "grad_norm": 0.9409313201904297,
      "learning_rate": 0.0005,
      "loss": 2.6803689002990723,
      "step": 2488
    },
    {
      "epoch": 0.5689142857142857,
      "grad_norm": 1.2315579652786255,
      "learning_rate": 0.0005,
      "loss": 2.736191749572754,
      "step": 2489
    },
    {
      "epoch": 0.5691428571428572,
      "grad_norm": 0.9369608759880066,
      "learning_rate": 0.0005,
      "loss": 2.882596015930176,
      "step": 2490
    },
    {
      "epoch": 0.5693714285714285,
      "grad_norm": 1.1413211822509766,
      "learning_rate": 0.0005,
      "loss": 2.842111349105835,
      "step": 2491
    },
    {
      "epoch": 0.5696,
      "grad_norm": 0.8744032979011536,
      "learning_rate": 0.0005,
      "loss": 2.7904696464538574,
      "step": 2492
    },
    {
      "epoch": 0.5698285714285715,
      "grad_norm": 1.194475531578064,
      "learning_rate": 0.0005,
      "loss": 3.060345411300659,
      "step": 2493
    },
    {
      "epoch": 0.5700571428571428,
      "grad_norm": 1.1744786500930786,
      "learning_rate": 0.0005,
      "loss": 3.022722005844116,
      "step": 2494
    },
    {
      "epoch": 0.5702857142857143,
      "grad_norm": 1.0159810781478882,
      "learning_rate": 0.0005,
      "loss": 2.6926205158233643,
      "step": 2495
    },
    {
      "epoch": 0.5705142857142858,
      "grad_norm": 1.1784015893936157,
      "learning_rate": 0.0005,
      "loss": 3.122732639312744,
      "step": 2496
    },
    {
      "epoch": 0.5707428571428571,
      "grad_norm": 1.185684084892273,
      "learning_rate": 0.0005,
      "loss": 2.8216705322265625,
      "step": 2497
    },
    {
      "epoch": 0.5709714285714286,
      "grad_norm": 0.8583148121833801,
      "learning_rate": 0.0005,
      "loss": 2.555966854095459,
      "step": 2498
    },
    {
      "epoch": 0.5712,
      "grad_norm": 0.9489954113960266,
      "learning_rate": 0.0005,
      "loss": 2.566906452178955,
      "step": 2499
    },
    {
      "epoch": 0.5714285714285714,
      "grad_norm": 1.0634307861328125,
      "learning_rate": 0.0005,
      "loss": 2.416923999786377,
      "step": 2500
    },
    {
      "epoch": 0.5716571428571429,
      "grad_norm": 1.104116439819336,
      "learning_rate": 0.0005,
      "loss": 2.5301969051361084,
      "step": 2501
    },
    {
      "epoch": 0.5718857142857143,
      "grad_norm": 1.0325974225997925,
      "learning_rate": 0.0005,
      "loss": 2.619415760040283,
      "step": 2502
    },
    {
      "epoch": 0.5721142857142857,
      "grad_norm": 1.493436336517334,
      "learning_rate": 0.0005,
      "loss": 2.9205777645111084,
      "step": 2503
    },
    {
      "epoch": 0.5723428571428572,
      "grad_norm": 1.1331285238265991,
      "learning_rate": 0.0005,
      "loss": 2.7420408725738525,
      "step": 2504
    },
    {
      "epoch": 0.5725714285714286,
      "grad_norm": 1.0094938278198242,
      "learning_rate": 0.0005,
      "loss": 3.002028465270996,
      "step": 2505
    },
    {
      "epoch": 0.5728,
      "grad_norm": 1.3934123516082764,
      "learning_rate": 0.0005,
      "loss": 2.540925979614258,
      "step": 2506
    },
    {
      "epoch": 0.5730285714285714,
      "grad_norm": 0.9522922039031982,
      "learning_rate": 0.0005,
      "loss": 2.7596943378448486,
      "step": 2507
    },
    {
      "epoch": 0.5732571428571429,
      "grad_norm": 1.277550458908081,
      "learning_rate": 0.0005,
      "loss": 2.9397025108337402,
      "step": 2508
    },
    {
      "epoch": 0.5734857142857143,
      "grad_norm": 1.0138503313064575,
      "learning_rate": 0.0005,
      "loss": 3.029933452606201,
      "step": 2509
    },
    {
      "epoch": 0.5737142857142857,
      "grad_norm": 0.9103339910507202,
      "learning_rate": 0.0005,
      "loss": 2.7041897773742676,
      "step": 2510
    },
    {
      "epoch": 0.5739428571428571,
      "grad_norm": 1.3657714128494263,
      "learning_rate": 0.0005,
      "loss": 2.705540418624878,
      "step": 2511
    },
    {
      "epoch": 0.5741714285714286,
      "grad_norm": 1.0962432622909546,
      "learning_rate": 0.0005,
      "loss": 2.808222770690918,
      "step": 2512
    },
    {
      "epoch": 0.5744,
      "grad_norm": 1.046663522720337,
      "learning_rate": 0.0005,
      "loss": 2.7153024673461914,
      "step": 2513
    },
    {
      "epoch": 0.5746285714285714,
      "grad_norm": 0.9595842957496643,
      "learning_rate": 0.0005,
      "loss": 2.902432680130005,
      "step": 2514
    },
    {
      "epoch": 0.5748571428571428,
      "grad_norm": 1.2544478178024292,
      "learning_rate": 0.0005,
      "loss": 3.25951886177063,
      "step": 2515
    },
    {
      "epoch": 0.5750857142857143,
      "grad_norm": 1.0220017433166504,
      "learning_rate": 0.0005,
      "loss": 2.519317626953125,
      "step": 2516
    },
    {
      "epoch": 0.5753142857142857,
      "grad_norm": 0.9704194664955139,
      "learning_rate": 0.0005,
      "loss": 2.971221446990967,
      "step": 2517
    },
    {
      "epoch": 0.5755428571428571,
      "grad_norm": 0.9734841585159302,
      "learning_rate": 0.0005,
      "loss": 2.8080291748046875,
      "step": 2518
    },
    {
      "epoch": 0.5757714285714286,
      "grad_norm": 1.1605772972106934,
      "learning_rate": 0.0005,
      "loss": 2.5933339595794678,
      "step": 2519
    },
    {
      "epoch": 0.576,
      "grad_norm": 1.0362811088562012,
      "learning_rate": 0.0005,
      "loss": 3.034341335296631,
      "step": 2520
    },
    {
      "epoch": 0.5762285714285714,
      "grad_norm": 0.8197253942489624,
      "learning_rate": 0.0005,
      "loss": 2.534780263900757,
      "step": 2521
    },
    {
      "epoch": 0.5764571428571429,
      "grad_norm": 1.3247524499893188,
      "learning_rate": 0.0005,
      "loss": 2.9309492111206055,
      "step": 2522
    },
    {
      "epoch": 0.5766857142857142,
      "grad_norm": 1.331204891204834,
      "learning_rate": 0.0005,
      "loss": 2.6190547943115234,
      "step": 2523
    },
    {
      "epoch": 0.5769142857142857,
      "grad_norm": 0.9638024568557739,
      "learning_rate": 0.0005,
      "loss": 2.6653013229370117,
      "step": 2524
    },
    {
      "epoch": 0.5771428571428572,
      "grad_norm": 0.9957114458084106,
      "learning_rate": 0.0005,
      "loss": 2.5488827228546143,
      "step": 2525
    },
    {
      "epoch": 0.5773714285714285,
      "grad_norm": 0.9864479899406433,
      "learning_rate": 0.0005,
      "loss": 2.7737574577331543,
      "step": 2526
    },
    {
      "epoch": 0.5776,
      "grad_norm": 1.137168049812317,
      "learning_rate": 0.0005,
      "loss": 2.74985408782959,
      "step": 2527
    },
    {
      "epoch": 0.5778285714285715,
      "grad_norm": 1.5384188890457153,
      "learning_rate": 0.0005,
      "loss": 2.956913471221924,
      "step": 2528
    },
    {
      "epoch": 0.5780571428571428,
      "grad_norm": 1.2063122987747192,
      "learning_rate": 0.0005,
      "loss": 2.5503368377685547,
      "step": 2529
    },
    {
      "epoch": 0.5782857142857143,
      "grad_norm": 1.0634665489196777,
      "learning_rate": 0.0005,
      "loss": 2.979573965072632,
      "step": 2530
    },
    {
      "epoch": 0.5785142857142858,
      "grad_norm": 1.006244421005249,
      "learning_rate": 0.0005,
      "loss": 2.635331153869629,
      "step": 2531
    },
    {
      "epoch": 0.5787428571428571,
      "grad_norm": 1.2637819051742554,
      "learning_rate": 0.0005,
      "loss": 2.6317803859710693,
      "step": 2532
    },
    {
      "epoch": 0.5789714285714286,
      "grad_norm": 1.0395177602767944,
      "learning_rate": 0.0005,
      "loss": 2.751710891723633,
      "step": 2533
    },
    {
      "epoch": 0.5792,
      "grad_norm": 0.8794528841972351,
      "learning_rate": 0.0005,
      "loss": 2.649064779281616,
      "step": 2534
    },
    {
      "epoch": 0.5794285714285714,
      "grad_norm": 1.2481119632720947,
      "learning_rate": 0.0005,
      "loss": 2.8555221557617188,
      "step": 2535
    },
    {
      "epoch": 0.5796571428571429,
      "grad_norm": 0.935827374458313,
      "learning_rate": 0.0005,
      "loss": 3.044037342071533,
      "step": 2536
    },
    {
      "epoch": 0.5798857142857143,
      "grad_norm": 1.062294840812683,
      "learning_rate": 0.0005,
      "loss": 2.5561184883117676,
      "step": 2537
    },
    {
      "epoch": 0.5801142857142857,
      "grad_norm": 2.4962873458862305,
      "learning_rate": 0.0005,
      "loss": 2.561983823776245,
      "step": 2538
    },
    {
      "epoch": 0.5803428571428572,
      "grad_norm": 1.0807193517684937,
      "learning_rate": 0.0005,
      "loss": 2.7299509048461914,
      "step": 2539
    },
    {
      "epoch": 0.5805714285714285,
      "grad_norm": 1.2401920557022095,
      "learning_rate": 0.0005,
      "loss": 3.078831911087036,
      "step": 2540
    },
    {
      "epoch": 0.5808,
      "grad_norm": 1.0121734142303467,
      "learning_rate": 0.0005,
      "loss": 2.7671709060668945,
      "step": 2541
    },
    {
      "epoch": 0.5810285714285714,
      "grad_norm": 0.8301960229873657,
      "learning_rate": 0.0005,
      "loss": 2.8279318809509277,
      "step": 2542
    },
    {
      "epoch": 0.5812571428571428,
      "grad_norm": 1.019629716873169,
      "learning_rate": 0.0005,
      "loss": 2.678849697113037,
      "step": 2543
    },
    {
      "epoch": 0.5814857142857143,
      "grad_norm": 1.1671866178512573,
      "learning_rate": 0.0005,
      "loss": 2.8042373657226562,
      "step": 2544
    },
    {
      "epoch": 0.5817142857142857,
      "grad_norm": 1.04668128490448,
      "learning_rate": 0.0005,
      "loss": 2.853929281234741,
      "step": 2545
    },
    {
      "epoch": 0.5819428571428571,
      "grad_norm": 1.2473196983337402,
      "learning_rate": 0.0005,
      "loss": 2.985403299331665,
      "step": 2546
    },
    {
      "epoch": 0.5821714285714286,
      "grad_norm": 1.215315818786621,
      "learning_rate": 0.0005,
      "loss": 2.746609926223755,
      "step": 2547
    },
    {
      "epoch": 0.5824,
      "grad_norm": 0.98944491147995,
      "learning_rate": 0.0005,
      "loss": 2.7652454376220703,
      "step": 2548
    },
    {
      "epoch": 0.5826285714285714,
      "grad_norm": 1.0186779499053955,
      "learning_rate": 0.0005,
      "loss": 2.5600979328155518,
      "step": 2549
    },
    {
      "epoch": 0.5828571428571429,
      "grad_norm": 1.0214086771011353,
      "learning_rate": 0.0005,
      "loss": 2.8744287490844727,
      "step": 2550
    },
    {
      "epoch": 0.5830857142857143,
      "grad_norm": 1.2415932416915894,
      "learning_rate": 0.0005,
      "loss": 2.804683208465576,
      "step": 2551
    },
    {
      "epoch": 0.5833142857142857,
      "grad_norm": 1.089855670928955,
      "learning_rate": 0.0005,
      "loss": 2.70982027053833,
      "step": 2552
    },
    {
      "epoch": 0.5835428571428571,
      "grad_norm": 1.0567108392715454,
      "learning_rate": 0.0005,
      "loss": 2.8029332160949707,
      "step": 2553
    },
    {
      "epoch": 0.5837714285714286,
      "grad_norm": 1.5509167909622192,
      "learning_rate": 0.0005,
      "loss": 2.8528380393981934,
      "step": 2554
    },
    {
      "epoch": 0.584,
      "grad_norm": 1.3149017095565796,
      "learning_rate": 0.0005,
      "loss": 2.9309537410736084,
      "step": 2555
    },
    {
      "epoch": 0.5842285714285714,
      "grad_norm": 1.0698519945144653,
      "learning_rate": 0.0005,
      "loss": 2.663184642791748,
      "step": 2556
    },
    {
      "epoch": 0.5844571428571429,
      "grad_norm": 1.0651476383209229,
      "learning_rate": 0.0005,
      "loss": 2.7947068214416504,
      "step": 2557
    },
    {
      "epoch": 0.5846857142857143,
      "grad_norm": 1.1790730953216553,
      "learning_rate": 0.0005,
      "loss": 2.7913923263549805,
      "step": 2558
    },
    {
      "epoch": 0.5849142857142857,
      "grad_norm": 1.061249017715454,
      "learning_rate": 0.0005,
      "loss": 2.90802264213562,
      "step": 2559
    },
    {
      "epoch": 0.5851428571428572,
      "grad_norm": 1.0713739395141602,
      "learning_rate": 0.0005,
      "loss": 2.535578727722168,
      "step": 2560
    },
    {
      "epoch": 0.5853714285714285,
      "grad_norm": 1.0586469173431396,
      "learning_rate": 0.0005,
      "loss": 2.6725621223449707,
      "step": 2561
    },
    {
      "epoch": 0.5856,
      "grad_norm": 1.0657451152801514,
      "learning_rate": 0.0005,
      "loss": 2.8481833934783936,
      "step": 2562
    },
    {
      "epoch": 0.5858285714285715,
      "grad_norm": 1.7469829320907593,
      "learning_rate": 0.0005,
      "loss": 2.666635513305664,
      "step": 2563
    },
    {
      "epoch": 0.5860571428571428,
      "grad_norm": 1.1063944101333618,
      "learning_rate": 0.0005,
      "loss": 2.8683900833129883,
      "step": 2564
    },
    {
      "epoch": 0.5862857142857143,
      "grad_norm": 1.1216539144515991,
      "learning_rate": 0.0005,
      "loss": 2.5251224040985107,
      "step": 2565
    },
    {
      "epoch": 0.5865142857142858,
      "grad_norm": 1.1393153667449951,
      "learning_rate": 0.0005,
      "loss": 2.6109697818756104,
      "step": 2566
    },
    {
      "epoch": 0.5867428571428571,
      "grad_norm": 1.1957645416259766,
      "learning_rate": 0.0005,
      "loss": 2.3239002227783203,
      "step": 2567
    },
    {
      "epoch": 0.5869714285714286,
      "grad_norm": 1.1490391492843628,
      "learning_rate": 0.0005,
      "loss": 3.093456268310547,
      "step": 2568
    },
    {
      "epoch": 0.5872,
      "grad_norm": 0.8580588698387146,
      "learning_rate": 0.0005,
      "loss": 3.03347110748291,
      "step": 2569
    },
    {
      "epoch": 0.5874285714285714,
      "grad_norm": 1.0897935628890991,
      "learning_rate": 0.0005,
      "loss": 3.076599597930908,
      "step": 2570
    },
    {
      "epoch": 0.5876571428571429,
      "grad_norm": 0.8637686371803284,
      "learning_rate": 0.0005,
      "loss": 2.9067554473876953,
      "step": 2571
    },
    {
      "epoch": 0.5878857142857142,
      "grad_norm": 1.2127255201339722,
      "learning_rate": 0.0005,
      "loss": 2.572481155395508,
      "step": 2572
    },
    {
      "epoch": 0.5881142857142857,
      "grad_norm": 1.1518263816833496,
      "learning_rate": 0.0005,
      "loss": 2.804595708847046,
      "step": 2573
    },
    {
      "epoch": 0.5883428571428572,
      "grad_norm": 1.7113498449325562,
      "learning_rate": 0.0005,
      "loss": 2.627100944519043,
      "step": 2574
    },
    {
      "epoch": 0.5885714285714285,
      "grad_norm": 1.3819218873977661,
      "learning_rate": 0.0005,
      "loss": 2.983541250228882,
      "step": 2575
    },
    {
      "epoch": 0.5888,
      "grad_norm": 3.172053098678589,
      "learning_rate": 0.0005,
      "loss": 2.72743558883667,
      "step": 2576
    },
    {
      "epoch": 0.5890285714285715,
      "grad_norm": 1.0270369052886963,
      "learning_rate": 0.0005,
      "loss": 2.7891006469726562,
      "step": 2577
    },
    {
      "epoch": 0.5892571428571428,
      "grad_norm": 1.1235390901565552,
      "learning_rate": 0.0005,
      "loss": 2.975710391998291,
      "step": 2578
    },
    {
      "epoch": 0.5894857142857143,
      "grad_norm": 1.167035460472107,
      "learning_rate": 0.0005,
      "loss": 2.3380777835845947,
      "step": 2579
    },
    {
      "epoch": 0.5897142857142857,
      "grad_norm": 1.36016845703125,
      "learning_rate": 0.0005,
      "loss": 2.438394546508789,
      "step": 2580
    },
    {
      "epoch": 0.5899428571428571,
      "grad_norm": 1.1244990825653076,
      "learning_rate": 0.0005,
      "loss": 3.0440330505371094,
      "step": 2581
    },
    {
      "epoch": 0.5901714285714286,
      "grad_norm": 1.0587376356124878,
      "learning_rate": 0.0005,
      "loss": 2.937849760055542,
      "step": 2582
    },
    {
      "epoch": 0.5904,
      "grad_norm": 1.252058982849121,
      "learning_rate": 0.0005,
      "loss": 2.941439151763916,
      "step": 2583
    },
    {
      "epoch": 0.5906285714285714,
      "grad_norm": 1.031575083732605,
      "learning_rate": 0.0005,
      "loss": 2.7804741859436035,
      "step": 2584
    },
    {
      "epoch": 0.5908571428571429,
      "grad_norm": 1.0850259065628052,
      "learning_rate": 0.0005,
      "loss": 2.8307995796203613,
      "step": 2585
    },
    {
      "epoch": 0.5910857142857143,
      "grad_norm": 1.3360753059387207,
      "learning_rate": 0.0005,
      "loss": 2.5225131511688232,
      "step": 2586
    },
    {
      "epoch": 0.5913142857142857,
      "grad_norm": 1.0145059823989868,
      "learning_rate": 0.0005,
      "loss": 2.4838781356811523,
      "step": 2587
    },
    {
      "epoch": 0.5915428571428571,
      "grad_norm": 0.91586834192276,
      "learning_rate": 0.0005,
      "loss": 2.942129135131836,
      "step": 2588
    },
    {
      "epoch": 0.5917714285714286,
      "grad_norm": 1.1929264068603516,
      "learning_rate": 0.0005,
      "loss": 2.5792133808135986,
      "step": 2589
    },
    {
      "epoch": 0.592,
      "grad_norm": 0.9214009642601013,
      "learning_rate": 0.0005,
      "loss": 2.616100311279297,
      "step": 2590
    },
    {
      "epoch": 0.5922285714285714,
      "grad_norm": 1.1795557737350464,
      "learning_rate": 0.0005,
      "loss": 3.2579078674316406,
      "step": 2591
    },
    {
      "epoch": 0.5924571428571429,
      "grad_norm": 1.026687741279602,
      "learning_rate": 0.0005,
      "loss": 2.6121997833251953,
      "step": 2592
    },
    {
      "epoch": 0.5926857142857143,
      "grad_norm": 1.440995216369629,
      "learning_rate": 0.0005,
      "loss": 2.5963377952575684,
      "step": 2593
    },
    {
      "epoch": 0.5929142857142857,
      "grad_norm": 1.1587212085723877,
      "learning_rate": 0.0005,
      "loss": 2.5899033546447754,
      "step": 2594
    },
    {
      "epoch": 0.5931428571428572,
      "grad_norm": 1.1080833673477173,
      "learning_rate": 0.0005,
      "loss": 2.6562023162841797,
      "step": 2595
    },
    {
      "epoch": 0.5933714285714285,
      "grad_norm": 1.2391105890274048,
      "learning_rate": 0.0005,
      "loss": 2.8049826622009277,
      "step": 2596
    },
    {
      "epoch": 0.5936,
      "grad_norm": 1.174601435661316,
      "learning_rate": 0.0005,
      "loss": 2.4367926120758057,
      "step": 2597
    },
    {
      "epoch": 0.5938285714285715,
      "grad_norm": 1.1242254972457886,
      "learning_rate": 0.0005,
      "loss": 2.804141044616699,
      "step": 2598
    },
    {
      "epoch": 0.5940571428571428,
      "grad_norm": 1.024918794631958,
      "learning_rate": 0.0005,
      "loss": 2.9204273223876953,
      "step": 2599
    },
    {
      "epoch": 0.5942857142857143,
      "grad_norm": 1.038064956665039,
      "learning_rate": 0.0005,
      "loss": 2.6458334922790527,
      "step": 2600
    },
    {
      "epoch": 0.5945142857142857,
      "grad_norm": 1.1607803106307983,
      "learning_rate": 0.0005,
      "loss": 2.8570804595947266,
      "step": 2601
    },
    {
      "epoch": 0.5947428571428571,
      "grad_norm": 0.9816292524337769,
      "learning_rate": 0.0005,
      "loss": 3.010739803314209,
      "step": 2602
    },
    {
      "epoch": 0.5949714285714286,
      "grad_norm": 1.3369362354278564,
      "learning_rate": 0.0005,
      "loss": 2.9747748374938965,
      "step": 2603
    },
    {
      "epoch": 0.5952,
      "grad_norm": 0.9709200263023376,
      "learning_rate": 0.0005,
      "loss": 2.913910388946533,
      "step": 2604
    },
    {
      "epoch": 0.5954285714285714,
      "grad_norm": 1.0588641166687012,
      "learning_rate": 0.0005,
      "loss": 2.891529083251953,
      "step": 2605
    },
    {
      "epoch": 0.5956571428571429,
      "grad_norm": 1.0958967208862305,
      "learning_rate": 0.0005,
      "loss": 3.323105812072754,
      "step": 2606
    },
    {
      "epoch": 0.5958857142857142,
      "grad_norm": 1.0656274557113647,
      "learning_rate": 0.0005,
      "loss": 2.5466883182525635,
      "step": 2607
    },
    {
      "epoch": 0.5961142857142857,
      "grad_norm": 0.9093886613845825,
      "learning_rate": 0.0005,
      "loss": 2.9410080909729004,
      "step": 2608
    },
    {
      "epoch": 0.5963428571428572,
      "grad_norm": 0.9849008321762085,
      "learning_rate": 0.0005,
      "loss": 2.2865777015686035,
      "step": 2609
    },
    {
      "epoch": 0.5965714285714285,
      "grad_norm": 1.090868353843689,
      "learning_rate": 0.0005,
      "loss": 2.7366178035736084,
      "step": 2610
    },
    {
      "epoch": 0.5968,
      "grad_norm": 0.978985607624054,
      "learning_rate": 0.0005,
      "loss": 2.7279481887817383,
      "step": 2611
    },
    {
      "epoch": 0.5970285714285715,
      "grad_norm": 1.1241368055343628,
      "learning_rate": 0.0005,
      "loss": 2.6879029273986816,
      "step": 2612
    },
    {
      "epoch": 0.5972571428571428,
      "grad_norm": 1.0351948738098145,
      "learning_rate": 0.0005,
      "loss": 3.196737766265869,
      "step": 2613
    },
    {
      "epoch": 0.5974857142857143,
      "grad_norm": 1.362951397895813,
      "learning_rate": 0.0005,
      "loss": 2.7821576595306396,
      "step": 2614
    },
    {
      "epoch": 0.5977142857142858,
      "grad_norm": 0.9622781872749329,
      "learning_rate": 0.0005,
      "loss": 2.773369789123535,
      "step": 2615
    },
    {
      "epoch": 0.5979428571428571,
      "grad_norm": 1.0090250968933105,
      "learning_rate": 0.0005,
      "loss": 2.9872894287109375,
      "step": 2616
    },
    {
      "epoch": 0.5981714285714286,
      "grad_norm": 1.1563050746917725,
      "learning_rate": 0.0005,
      "loss": 2.438824415206909,
      "step": 2617
    },
    {
      "epoch": 0.5984,
      "grad_norm": 1.115014672279358,
      "learning_rate": 0.0005,
      "loss": 2.710235595703125,
      "step": 2618
    },
    {
      "epoch": 0.5986285714285714,
      "grad_norm": 1.085199236869812,
      "learning_rate": 0.0005,
      "loss": 2.7934963703155518,
      "step": 2619
    },
    {
      "epoch": 0.5988571428571429,
      "grad_norm": 1.3107101917266846,
      "learning_rate": 0.0005,
      "loss": 2.690143585205078,
      "step": 2620
    },
    {
      "epoch": 0.5990857142857143,
      "grad_norm": 0.9286937117576599,
      "learning_rate": 0.0005,
      "loss": 3.3080496788024902,
      "step": 2621
    },
    {
      "epoch": 0.5993142857142857,
      "grad_norm": 0.9911589622497559,
      "learning_rate": 0.0005,
      "loss": 2.5875794887542725,
      "step": 2622
    },
    {
      "epoch": 0.5995428571428572,
      "grad_norm": 1.2047631740570068,
      "learning_rate": 0.0005,
      "loss": 2.685812473297119,
      "step": 2623
    },
    {
      "epoch": 0.5997714285714286,
      "grad_norm": 0.942592442035675,
      "learning_rate": 0.0005,
      "loss": 2.809325933456421,
      "step": 2624
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.1168811321258545,
      "learning_rate": 0.0005,
      "loss": 2.665541648864746,
      "step": 2625
    },
    {
      "epoch": 0.6002285714285714,
      "grad_norm": 1.1907986402511597,
      "learning_rate": 0.0005,
      "loss": 2.4224753379821777,
      "step": 2626
    },
    {
      "epoch": 0.6004571428571429,
      "grad_norm": 1.4494054317474365,
      "learning_rate": 0.0005,
      "loss": 2.731999397277832,
      "step": 2627
    },
    {
      "epoch": 0.6006857142857143,
      "grad_norm": 1.3482393026351929,
      "learning_rate": 0.0005,
      "loss": 2.554752826690674,
      "step": 2628
    },
    {
      "epoch": 0.6009142857142857,
      "grad_norm": 1.0977866649627686,
      "learning_rate": 0.0005,
      "loss": 3.301215648651123,
      "step": 2629
    },
    {
      "epoch": 0.6011428571428571,
      "grad_norm": 1.0075987577438354,
      "learning_rate": 0.0005,
      "loss": 2.5740151405334473,
      "step": 2630
    },
    {
      "epoch": 0.6013714285714286,
      "grad_norm": 1.139339804649353,
      "learning_rate": 0.0005,
      "loss": 2.8200368881225586,
      "step": 2631
    },
    {
      "epoch": 0.6016,
      "grad_norm": 0.9457255005836487,
      "learning_rate": 0.0005,
      "loss": 2.755483627319336,
      "step": 2632
    },
    {
      "epoch": 0.6018285714285714,
      "grad_norm": 0.9844979047775269,
      "learning_rate": 0.0005,
      "loss": 2.8215599060058594,
      "step": 2633
    },
    {
      "epoch": 0.6020571428571428,
      "grad_norm": 1.005347728729248,
      "learning_rate": 0.0005,
      "loss": 2.8926286697387695,
      "step": 2634
    },
    {
      "epoch": 0.6022857142857143,
      "grad_norm": 0.9954663515090942,
      "learning_rate": 0.0005,
      "loss": 3.1469337940216064,
      "step": 2635
    },
    {
      "epoch": 0.6025142857142857,
      "grad_norm": 1.1135236024856567,
      "learning_rate": 0.0005,
      "loss": 2.842529773712158,
      "step": 2636
    },
    {
      "epoch": 0.6027428571428571,
      "grad_norm": 1.590264916419983,
      "learning_rate": 0.0005,
      "loss": 2.957552433013916,
      "step": 2637
    },
    {
      "epoch": 0.6029714285714286,
      "grad_norm": 1.2171010971069336,
      "learning_rate": 0.0005,
      "loss": 2.5632810592651367,
      "step": 2638
    },
    {
      "epoch": 0.6032,
      "grad_norm": 1.3541011810302734,
      "learning_rate": 0.0005,
      "loss": 3.2012686729431152,
      "step": 2639
    },
    {
      "epoch": 0.6034285714285714,
      "grad_norm": 1.1498041152954102,
      "learning_rate": 0.0005,
      "loss": 2.887477397918701,
      "step": 2640
    },
    {
      "epoch": 0.6036571428571429,
      "grad_norm": 1.2353607416152954,
      "learning_rate": 0.0005,
      "loss": 3.1751132011413574,
      "step": 2641
    },
    {
      "epoch": 0.6038857142857142,
      "grad_norm": 0.8874204158782959,
      "learning_rate": 0.0005,
      "loss": 2.9281678199768066,
      "step": 2642
    },
    {
      "epoch": 0.6041142857142857,
      "grad_norm": 1.637192726135254,
      "learning_rate": 0.0005,
      "loss": 2.9949138164520264,
      "step": 2643
    },
    {
      "epoch": 0.6043428571428572,
      "grad_norm": 1.4847768545150757,
      "learning_rate": 0.0005,
      "loss": 2.7144834995269775,
      "step": 2644
    },
    {
      "epoch": 0.6045714285714285,
      "grad_norm": 1.4501646757125854,
      "learning_rate": 0.0005,
      "loss": 2.7250428199768066,
      "step": 2645
    },
    {
      "epoch": 0.6048,
      "grad_norm": 1.102194905281067,
      "learning_rate": 0.0005,
      "loss": 2.9458742141723633,
      "step": 2646
    },
    {
      "epoch": 0.6050285714285715,
      "grad_norm": 1.1724406480789185,
      "learning_rate": 0.0005,
      "loss": 2.6190147399902344,
      "step": 2647
    },
    {
      "epoch": 0.6052571428571428,
      "grad_norm": 1.503197431564331,
      "learning_rate": 0.0005,
      "loss": 2.7856736183166504,
      "step": 2648
    },
    {
      "epoch": 0.6054857142857143,
      "grad_norm": 0.906467854976654,
      "learning_rate": 0.0005,
      "loss": 2.6667277812957764,
      "step": 2649
    },
    {
      "epoch": 0.6057142857142858,
      "grad_norm": 1.2421923875808716,
      "learning_rate": 0.0005,
      "loss": 2.6321299076080322,
      "step": 2650
    },
    {
      "epoch": 0.6059428571428571,
      "grad_norm": 0.9919896125793457,
      "learning_rate": 0.0005,
      "loss": 2.621173620223999,
      "step": 2651
    },
    {
      "epoch": 0.6061714285714286,
      "grad_norm": 1.8802528381347656,
      "learning_rate": 0.0005,
      "loss": 3.4518141746520996,
      "step": 2652
    },
    {
      "epoch": 0.6064,
      "grad_norm": 1.231364130973816,
      "learning_rate": 0.0005,
      "loss": 2.8088669776916504,
      "step": 2653
    },
    {
      "epoch": 0.6066285714285714,
      "grad_norm": 1.1149752140045166,
      "learning_rate": 0.0005,
      "loss": 2.960988759994507,
      "step": 2654
    },
    {
      "epoch": 0.6068571428571429,
      "grad_norm": 1.0993746519088745,
      "learning_rate": 0.0005,
      "loss": 2.525003433227539,
      "step": 2655
    },
    {
      "epoch": 0.6070857142857143,
      "grad_norm": 0.9898205399513245,
      "learning_rate": 0.0005,
      "loss": 2.4947304725646973,
      "step": 2656
    },
    {
      "epoch": 0.6073142857142857,
      "grad_norm": 1.5261179208755493,
      "learning_rate": 0.0005,
      "loss": 2.9285573959350586,
      "step": 2657
    },
    {
      "epoch": 0.6075428571428572,
      "grad_norm": 1.1345322132110596,
      "learning_rate": 0.0005,
      "loss": 2.8277835845947266,
      "step": 2658
    },
    {
      "epoch": 0.6077714285714285,
      "grad_norm": 0.9293845295906067,
      "learning_rate": 0.0005,
      "loss": 2.675576686859131,
      "step": 2659
    },
    {
      "epoch": 0.608,
      "grad_norm": 0.9485197067260742,
      "learning_rate": 0.0005,
      "loss": 2.6366281509399414,
      "step": 2660
    },
    {
      "epoch": 0.6082285714285715,
      "grad_norm": 1.1635820865631104,
      "learning_rate": 0.0005,
      "loss": 2.9550085067749023,
      "step": 2661
    },
    {
      "epoch": 0.6084571428571428,
      "grad_norm": 1.2395902872085571,
      "learning_rate": 0.0005,
      "loss": 2.935727119445801,
      "step": 2662
    },
    {
      "epoch": 0.6086857142857143,
      "grad_norm": 1.1078916788101196,
      "learning_rate": 0.0005,
      "loss": 2.788031578063965,
      "step": 2663
    },
    {
      "epoch": 0.6089142857142857,
      "grad_norm": 1.3232717514038086,
      "learning_rate": 0.0005,
      "loss": 2.5896153450012207,
      "step": 2664
    },
    {
      "epoch": 0.6091428571428571,
      "grad_norm": 1.6357756853103638,
      "learning_rate": 0.0005,
      "loss": 2.6813840866088867,
      "step": 2665
    },
    {
      "epoch": 0.6093714285714286,
      "grad_norm": 1.7513736486434937,
      "learning_rate": 0.0005,
      "loss": 2.7697200775146484,
      "step": 2666
    },
    {
      "epoch": 0.6096,
      "grad_norm": 1.1043199300765991,
      "learning_rate": 0.0005,
      "loss": 2.667269229888916,
      "step": 2667
    },
    {
      "epoch": 0.6098285714285714,
      "grad_norm": 0.9477903842926025,
      "learning_rate": 0.0005,
      "loss": 2.8703370094299316,
      "step": 2668
    },
    {
      "epoch": 0.6100571428571429,
      "grad_norm": 1.3467923402786255,
      "learning_rate": 0.0005,
      "loss": 2.80188250541687,
      "step": 2669
    },
    {
      "epoch": 0.6102857142857143,
      "grad_norm": 1.1872963905334473,
      "learning_rate": 0.0005,
      "loss": 2.5043530464172363,
      "step": 2670
    },
    {
      "epoch": 0.6105142857142857,
      "grad_norm": 1.021425724029541,
      "learning_rate": 0.0005,
      "loss": 2.8134238719940186,
      "step": 2671
    },
    {
      "epoch": 0.6107428571428571,
      "grad_norm": 1.0133241415023804,
      "learning_rate": 0.0005,
      "loss": 2.6448488235473633,
      "step": 2672
    },
    {
      "epoch": 0.6109714285714286,
      "grad_norm": 1.4722774028778076,
      "learning_rate": 0.0005,
      "loss": 2.2809336185455322,
      "step": 2673
    },
    {
      "epoch": 0.6112,
      "grad_norm": 1.409346103668213,
      "learning_rate": 0.0005,
      "loss": 2.8288798332214355,
      "step": 2674
    },
    {
      "epoch": 0.6114285714285714,
      "grad_norm": 1.2745134830474854,
      "learning_rate": 0.0005,
      "loss": 2.6827392578125,
      "step": 2675
    },
    {
      "epoch": 0.6116571428571429,
      "grad_norm": 0.925715446472168,
      "learning_rate": 0.0005,
      "loss": 2.782655954360962,
      "step": 2676
    },
    {
      "epoch": 0.6118857142857143,
      "grad_norm": 1.0031284093856812,
      "learning_rate": 0.0005,
      "loss": 2.3910577297210693,
      "step": 2677
    },
    {
      "epoch": 0.6121142857142857,
      "grad_norm": 0.9172390103340149,
      "learning_rate": 0.0005,
      "loss": 2.7802538871765137,
      "step": 2678
    },
    {
      "epoch": 0.6123428571428572,
      "grad_norm": 1.183843970298767,
      "learning_rate": 0.0005,
      "loss": 2.863856554031372,
      "step": 2679
    },
    {
      "epoch": 0.6125714285714285,
      "grad_norm": 1.1570121049880981,
      "learning_rate": 0.0005,
      "loss": 2.6190221309661865,
      "step": 2680
    },
    {
      "epoch": 0.6128,
      "grad_norm": 1.203328251838684,
      "learning_rate": 0.0005,
      "loss": 2.635751962661743,
      "step": 2681
    },
    {
      "epoch": 0.6130285714285715,
      "grad_norm": 1.2338411808013916,
      "learning_rate": 0.0005,
      "loss": 2.4252376556396484,
      "step": 2682
    },
    {
      "epoch": 0.6132571428571428,
      "grad_norm": 1.0590232610702515,
      "learning_rate": 0.0005,
      "loss": 2.6997947692871094,
      "step": 2683
    },
    {
      "epoch": 0.6134857142857143,
      "grad_norm": 1.1957190036773682,
      "learning_rate": 0.0005,
      "loss": 2.6158690452575684,
      "step": 2684
    },
    {
      "epoch": 0.6137142857142858,
      "grad_norm": 1.2704970836639404,
      "learning_rate": 0.0005,
      "loss": 2.8965649604797363,
      "step": 2685
    },
    {
      "epoch": 0.6139428571428571,
      "grad_norm": 0.9491807818412781,
      "learning_rate": 0.0005,
      "loss": 2.274264097213745,
      "step": 2686
    },
    {
      "epoch": 0.6141714285714286,
      "grad_norm": 0.9981251358985901,
      "learning_rate": 0.0005,
      "loss": 2.6985721588134766,
      "step": 2687
    },
    {
      "epoch": 0.6144,
      "grad_norm": 1.0165199041366577,
      "learning_rate": 0.0005,
      "loss": 2.858560800552368,
      "step": 2688
    },
    {
      "epoch": 0.6146285714285714,
      "grad_norm": 1.130845308303833,
      "learning_rate": 0.0005,
      "loss": 2.7253732681274414,
      "step": 2689
    },
    {
      "epoch": 0.6148571428571429,
      "grad_norm": 0.9465959072113037,
      "learning_rate": 0.0005,
      "loss": 2.6809067726135254,
      "step": 2690
    },
    {
      "epoch": 0.6150857142857142,
      "grad_norm": 1.186848521232605,
      "learning_rate": 0.0005,
      "loss": 2.8655524253845215,
      "step": 2691
    },
    {
      "epoch": 0.6153142857142857,
      "grad_norm": 1.1355030536651611,
      "learning_rate": 0.0005,
      "loss": 2.654719829559326,
      "step": 2692
    },
    {
      "epoch": 0.6155428571428572,
      "grad_norm": 0.9500078558921814,
      "learning_rate": 0.0005,
      "loss": 2.778301954269409,
      "step": 2693
    },
    {
      "epoch": 0.6157714285714285,
      "grad_norm": 1.6123961210250854,
      "learning_rate": 0.0005,
      "loss": 2.817284107208252,
      "step": 2694
    },
    {
      "epoch": 0.616,
      "grad_norm": 1.0339144468307495,
      "learning_rate": 0.0005,
      "loss": 2.9218029975891113,
      "step": 2695
    },
    {
      "epoch": 0.6162285714285715,
      "grad_norm": 1.0276386737823486,
      "learning_rate": 0.0005,
      "loss": 2.7054696083068848,
      "step": 2696
    },
    {
      "epoch": 0.6164571428571428,
      "grad_norm": 0.9717345237731934,
      "learning_rate": 0.0005,
      "loss": 2.666508197784424,
      "step": 2697
    },
    {
      "epoch": 0.6166857142857143,
      "grad_norm": 0.9398512244224548,
      "learning_rate": 0.0005,
      "loss": 2.9599533081054688,
      "step": 2698
    },
    {
      "epoch": 0.6169142857142857,
      "grad_norm": 0.869759202003479,
      "learning_rate": 0.0005,
      "loss": 2.6809606552124023,
      "step": 2699
    },
    {
      "epoch": 0.6171428571428571,
      "grad_norm": 1.6891019344329834,
      "learning_rate": 0.0005,
      "loss": 2.662895679473877,
      "step": 2700
    },
    {
      "epoch": 0.6173714285714286,
      "grad_norm": 1.004440426826477,
      "learning_rate": 0.0005,
      "loss": 2.6836836338043213,
      "step": 2701
    },
    {
      "epoch": 0.6176,
      "grad_norm": 0.8430218696594238,
      "learning_rate": 0.0005,
      "loss": 2.7802324295043945,
      "step": 2702
    },
    {
      "epoch": 0.6178285714285714,
      "grad_norm": 1.0037691593170166,
      "learning_rate": 0.0005,
      "loss": 2.8184046745300293,
      "step": 2703
    },
    {
      "epoch": 0.6180571428571429,
      "grad_norm": 1.0846538543701172,
      "learning_rate": 0.0005,
      "loss": 2.5470223426818848,
      "step": 2704
    },
    {
      "epoch": 0.6182857142857143,
      "grad_norm": 1.159996747970581,
      "learning_rate": 0.0005,
      "loss": 2.656322956085205,
      "step": 2705
    },
    {
      "epoch": 0.6185142857142857,
      "grad_norm": 1.112176537513733,
      "learning_rate": 0.0005,
      "loss": 2.6373348236083984,
      "step": 2706
    },
    {
      "epoch": 0.6187428571428571,
      "grad_norm": 1.3966256380081177,
      "learning_rate": 0.0005,
      "loss": 2.910946846008301,
      "step": 2707
    },
    {
      "epoch": 0.6189714285714286,
      "grad_norm": 0.9568463563919067,
      "learning_rate": 0.0005,
      "loss": 2.422957420349121,
      "step": 2708
    },
    {
      "epoch": 0.6192,
      "grad_norm": 0.9116679430007935,
      "learning_rate": 0.0005,
      "loss": 3.005927085876465,
      "step": 2709
    },
    {
      "epoch": 0.6194285714285714,
      "grad_norm": 0.983733594417572,
      "learning_rate": 0.0005,
      "loss": 3.072917938232422,
      "step": 2710
    },
    {
      "epoch": 0.6196571428571429,
      "grad_norm": 0.9590238332748413,
      "learning_rate": 0.0005,
      "loss": 2.7328052520751953,
      "step": 2711
    },
    {
      "epoch": 0.6198857142857143,
      "grad_norm": 1.2029904127120972,
      "learning_rate": 0.0005,
      "loss": 2.806469202041626,
      "step": 2712
    },
    {
      "epoch": 0.6201142857142857,
      "grad_norm": 0.8911043405532837,
      "learning_rate": 0.0005,
      "loss": 2.833008050918579,
      "step": 2713
    },
    {
      "epoch": 0.6203428571428572,
      "grad_norm": 0.9771230816841125,
      "learning_rate": 0.0005,
      "loss": 3.0256896018981934,
      "step": 2714
    },
    {
      "epoch": 0.6205714285714286,
      "grad_norm": 1.048836350440979,
      "learning_rate": 0.0005,
      "loss": 3.042185068130493,
      "step": 2715
    },
    {
      "epoch": 0.6208,
      "grad_norm": 1.2450748682022095,
      "learning_rate": 0.0005,
      "loss": 2.8776392936706543,
      "step": 2716
    },
    {
      "epoch": 0.6210285714285714,
      "grad_norm": 1.0893094539642334,
      "learning_rate": 0.0005,
      "loss": 2.65390682220459,
      "step": 2717
    },
    {
      "epoch": 0.6212571428571428,
      "grad_norm": 1.793266773223877,
      "learning_rate": 0.0005,
      "loss": 2.734698534011841,
      "step": 2718
    },
    {
      "epoch": 0.6214857142857143,
      "grad_norm": 0.9183365106582642,
      "learning_rate": 0.0005,
      "loss": 2.9818954467773438,
      "step": 2719
    },
    {
      "epoch": 0.6217142857142857,
      "grad_norm": 0.9068136811256409,
      "learning_rate": 0.0005,
      "loss": 2.862483263015747,
      "step": 2720
    },
    {
      "epoch": 0.6219428571428571,
      "grad_norm": 0.9479925036430359,
      "learning_rate": 0.0005,
      "loss": 2.99727725982666,
      "step": 2721
    },
    {
      "epoch": 0.6221714285714286,
      "grad_norm": 0.9918519258499146,
      "learning_rate": 0.0005,
      "loss": 2.8129358291625977,
      "step": 2722
    },
    {
      "epoch": 0.6224,
      "grad_norm": 1.3186103105545044,
      "learning_rate": 0.0005,
      "loss": 2.211362838745117,
      "step": 2723
    },
    {
      "epoch": 0.6226285714285714,
      "grad_norm": 1.193631887435913,
      "learning_rate": 0.0005,
      "loss": 2.787034511566162,
      "step": 2724
    },
    {
      "epoch": 0.6228571428571429,
      "grad_norm": 0.9713414311408997,
      "learning_rate": 0.0005,
      "loss": 2.5339486598968506,
      "step": 2725
    },
    {
      "epoch": 0.6230857142857142,
      "grad_norm": 0.9814707040786743,
      "learning_rate": 0.0005,
      "loss": 2.8524129390716553,
      "step": 2726
    },
    {
      "epoch": 0.6233142857142857,
      "grad_norm": 1.0878506898880005,
      "learning_rate": 0.0005,
      "loss": 2.725560188293457,
      "step": 2727
    },
    {
      "epoch": 0.6235428571428572,
      "grad_norm": 1.087360143661499,
      "learning_rate": 0.0005,
      "loss": 3.1331541538238525,
      "step": 2728
    },
    {
      "epoch": 0.6237714285714285,
      "grad_norm": 0.9973111152648926,
      "learning_rate": 0.0005,
      "loss": 2.73483943939209,
      "step": 2729
    },
    {
      "epoch": 0.624,
      "grad_norm": 0.8238992691040039,
      "learning_rate": 0.0005,
      "loss": 2.332069158554077,
      "step": 2730
    },
    {
      "epoch": 0.6242285714285715,
      "grad_norm": 1.1895751953125,
      "learning_rate": 0.0005,
      "loss": 2.7402658462524414,
      "step": 2731
    },
    {
      "epoch": 0.6244571428571428,
      "grad_norm": 1.1468732357025146,
      "learning_rate": 0.0005,
      "loss": 2.980731248855591,
      "step": 2732
    },
    {
      "epoch": 0.6246857142857143,
      "grad_norm": 1.3373439311981201,
      "learning_rate": 0.0005,
      "loss": 2.9045538902282715,
      "step": 2733
    },
    {
      "epoch": 0.6249142857142858,
      "grad_norm": 1.0974812507629395,
      "learning_rate": 0.0005,
      "loss": 2.9755859375,
      "step": 2734
    },
    {
      "epoch": 0.6251428571428571,
      "grad_norm": 1.6932880878448486,
      "learning_rate": 0.0005,
      "loss": 2.8272879123687744,
      "step": 2735
    },
    {
      "epoch": 0.6253714285714286,
      "grad_norm": 1.3528918027877808,
      "learning_rate": 0.0005,
      "loss": 2.9419686794281006,
      "step": 2736
    },
    {
      "epoch": 0.6256,
      "grad_norm": 0.914581298828125,
      "learning_rate": 0.0005,
      "loss": 2.7527287006378174,
      "step": 2737
    },
    {
      "epoch": 0.6258285714285714,
      "grad_norm": 1.0116302967071533,
      "learning_rate": 0.0005,
      "loss": 2.5869345664978027,
      "step": 2738
    },
    {
      "epoch": 0.6260571428571429,
      "grad_norm": 1.2003856897354126,
      "learning_rate": 0.0005,
      "loss": 2.6490182876586914,
      "step": 2739
    },
    {
      "epoch": 0.6262857142857143,
      "grad_norm": 0.8315615653991699,
      "learning_rate": 0.0005,
      "loss": 2.9392690658569336,
      "step": 2740
    },
    {
      "epoch": 0.6265142857142857,
      "grad_norm": 0.9842820167541504,
      "learning_rate": 0.0005,
      "loss": 2.6569745540618896,
      "step": 2741
    },
    {
      "epoch": 0.6267428571428572,
      "grad_norm": 1.3147218227386475,
      "learning_rate": 0.0005,
      "loss": 2.760671615600586,
      "step": 2742
    },
    {
      "epoch": 0.6269714285714286,
      "grad_norm": 1.7823339700698853,
      "learning_rate": 0.0005,
      "loss": 2.807281017303467,
      "step": 2743
    },
    {
      "epoch": 0.6272,
      "grad_norm": 1.1948202848434448,
      "learning_rate": 0.0005,
      "loss": 3.04805850982666,
      "step": 2744
    },
    {
      "epoch": 0.6274285714285714,
      "grad_norm": 1.386938214302063,
      "learning_rate": 0.0005,
      "loss": 2.6835989952087402,
      "step": 2745
    },
    {
      "epoch": 0.6276571428571428,
      "grad_norm": 0.9171589016914368,
      "learning_rate": 0.0005,
      "loss": 2.6067025661468506,
      "step": 2746
    },
    {
      "epoch": 0.6278857142857143,
      "grad_norm": 1.2875065803527832,
      "learning_rate": 0.0005,
      "loss": 2.8189024925231934,
      "step": 2747
    },
    {
      "epoch": 0.6281142857142857,
      "grad_norm": 1.026686191558838,
      "learning_rate": 0.0005,
      "loss": 2.5649588108062744,
      "step": 2748
    },
    {
      "epoch": 0.6283428571428571,
      "grad_norm": 1.284875512123108,
      "learning_rate": 0.0005,
      "loss": 2.663296699523926,
      "step": 2749
    },
    {
      "epoch": 0.6285714285714286,
      "grad_norm": 0.8655788898468018,
      "learning_rate": 0.0005,
      "loss": 2.615895986557007,
      "step": 2750
    },
    {
      "epoch": 0.6288,
      "grad_norm": 1.132967472076416,
      "learning_rate": 0.0005,
      "loss": 2.294379472732544,
      "step": 2751
    },
    {
      "epoch": 0.6290285714285714,
      "grad_norm": 0.8795512914657593,
      "learning_rate": 0.0005,
      "loss": 2.7443366050720215,
      "step": 2752
    },
    {
      "epoch": 0.6292571428571428,
      "grad_norm": 1.030795931816101,
      "learning_rate": 0.0005,
      "loss": 2.433955192565918,
      "step": 2753
    },
    {
      "epoch": 0.6294857142857143,
      "grad_norm": 0.9273151159286499,
      "learning_rate": 0.0005,
      "loss": 2.7828707695007324,
      "step": 2754
    },
    {
      "epoch": 0.6297142857142857,
      "grad_norm": 1.0019116401672363,
      "learning_rate": 0.0005,
      "loss": 2.542306900024414,
      "step": 2755
    },
    {
      "epoch": 0.6299428571428571,
      "grad_norm": 0.8631278276443481,
      "learning_rate": 0.0005,
      "loss": 2.794285297393799,
      "step": 2756
    },
    {
      "epoch": 0.6301714285714286,
      "grad_norm": 0.9370288848876953,
      "learning_rate": 0.0005,
      "loss": 2.5814740657806396,
      "step": 2757
    },
    {
      "epoch": 0.6304,
      "grad_norm": 1.008039116859436,
      "learning_rate": 0.0005,
      "loss": 2.9822335243225098,
      "step": 2758
    },
    {
      "epoch": 0.6306285714285714,
      "grad_norm": 2.220532178878784,
      "learning_rate": 0.0005,
      "loss": 2.402564525604248,
      "step": 2759
    },
    {
      "epoch": 0.6308571428571429,
      "grad_norm": 1.3654459714889526,
      "learning_rate": 0.0005,
      "loss": 2.3875746726989746,
      "step": 2760
    },
    {
      "epoch": 0.6310857142857142,
      "grad_norm": 1.503889560699463,
      "learning_rate": 0.0005,
      "loss": 2.6018857955932617,
      "step": 2761
    },
    {
      "epoch": 0.6313142857142857,
      "grad_norm": 0.9213805198669434,
      "learning_rate": 0.0005,
      "loss": 2.375352382659912,
      "step": 2762
    },
    {
      "epoch": 0.6315428571428572,
      "grad_norm": 1.2742810249328613,
      "learning_rate": 0.0005,
      "loss": 2.8850364685058594,
      "step": 2763
    },
    {
      "epoch": 0.6317714285714285,
      "grad_norm": 0.8415848612785339,
      "learning_rate": 0.0005,
      "loss": 2.6621553897857666,
      "step": 2764
    },
    {
      "epoch": 0.632,
      "grad_norm": 1.0189452171325684,
      "learning_rate": 0.0005,
      "loss": 2.4817514419555664,
      "step": 2765
    },
    {
      "epoch": 0.6322285714285715,
      "grad_norm": 0.8884305953979492,
      "learning_rate": 0.0005,
      "loss": 2.388981342315674,
      "step": 2766
    },
    {
      "epoch": 0.6324571428571428,
      "grad_norm": 1.3583850860595703,
      "learning_rate": 0.0005,
      "loss": 2.8706166744232178,
      "step": 2767
    },
    {
      "epoch": 0.6326857142857143,
      "grad_norm": 1.0675907135009766,
      "learning_rate": 0.0005,
      "loss": 2.937103271484375,
      "step": 2768
    },
    {
      "epoch": 0.6329142857142858,
      "grad_norm": 1.9284464120864868,
      "learning_rate": 0.0005,
      "loss": 2.6366171836853027,
      "step": 2769
    },
    {
      "epoch": 0.6331428571428571,
      "grad_norm": 1.058510422706604,
      "learning_rate": 0.0005,
      "loss": 2.749692916870117,
      "step": 2770
    },
    {
      "epoch": 0.6333714285714286,
      "grad_norm": 1.3095366954803467,
      "learning_rate": 0.0005,
      "loss": 2.731503486633301,
      "step": 2771
    },
    {
      "epoch": 0.6336,
      "grad_norm": 0.9454983472824097,
      "learning_rate": 0.0005,
      "loss": 3.0668087005615234,
      "step": 2772
    },
    {
      "epoch": 0.6338285714285714,
      "grad_norm": 1.152617335319519,
      "learning_rate": 0.0005,
      "loss": 2.8977558612823486,
      "step": 2773
    },
    {
      "epoch": 0.6340571428571429,
      "grad_norm": 1.4637885093688965,
      "learning_rate": 0.0005,
      "loss": 2.6440091133117676,
      "step": 2774
    },
    {
      "epoch": 0.6342857142857142,
      "grad_norm": 1.1243667602539062,
      "learning_rate": 0.0005,
      "loss": 2.454723834991455,
      "step": 2775
    },
    {
      "epoch": 0.6345142857142857,
      "grad_norm": 0.8616605997085571,
      "learning_rate": 0.0005,
      "loss": 2.5571632385253906,
      "step": 2776
    },
    {
      "epoch": 0.6347428571428572,
      "grad_norm": 1.1571341753005981,
      "learning_rate": 0.0005,
      "loss": 3.2612342834472656,
      "step": 2777
    },
    {
      "epoch": 0.6349714285714285,
      "grad_norm": 0.8980388045310974,
      "learning_rate": 0.0005,
      "loss": 2.8708229064941406,
      "step": 2778
    },
    {
      "epoch": 0.6352,
      "grad_norm": 1.2777713537216187,
      "learning_rate": 0.0005,
      "loss": 2.8272814750671387,
      "step": 2779
    },
    {
      "epoch": 0.6354285714285715,
      "grad_norm": 1.171330213546753,
      "learning_rate": 0.0005,
      "loss": 2.650953769683838,
      "step": 2780
    },
    {
      "epoch": 0.6356571428571428,
      "grad_norm": 1.4271024465560913,
      "learning_rate": 0.0005,
      "loss": 3.27408766746521,
      "step": 2781
    },
    {
      "epoch": 0.6358857142857143,
      "grad_norm": 1.0707908868789673,
      "learning_rate": 0.0005,
      "loss": 2.7242519855499268,
      "step": 2782
    },
    {
      "epoch": 0.6361142857142857,
      "grad_norm": 1.290899395942688,
      "learning_rate": 0.0005,
      "loss": 2.7010278701782227,
      "step": 2783
    },
    {
      "epoch": 0.6363428571428571,
      "grad_norm": 0.9598969221115112,
      "learning_rate": 0.0005,
      "loss": 2.7317395210266113,
      "step": 2784
    },
    {
      "epoch": 0.6365714285714286,
      "grad_norm": 1.1745561361312866,
      "learning_rate": 0.0005,
      "loss": 2.7306838035583496,
      "step": 2785
    },
    {
      "epoch": 0.6368,
      "grad_norm": 1.153132438659668,
      "learning_rate": 0.0005,
      "loss": 3.140589714050293,
      "step": 2786
    },
    {
      "epoch": 0.6370285714285714,
      "grad_norm": 1.1046820878982544,
      "learning_rate": 0.0005,
      "loss": 2.675258159637451,
      "step": 2787
    },
    {
      "epoch": 0.6372571428571429,
      "grad_norm": 1.161462426185608,
      "learning_rate": 0.0005,
      "loss": 2.4959464073181152,
      "step": 2788
    },
    {
      "epoch": 0.6374857142857143,
      "grad_norm": 1.1918350458145142,
      "learning_rate": 0.0005,
      "loss": 2.973527431488037,
      "step": 2789
    },
    {
      "epoch": 0.6377142857142857,
      "grad_norm": 1.06147301197052,
      "learning_rate": 0.0005,
      "loss": 2.4708621501922607,
      "step": 2790
    },
    {
      "epoch": 0.6379428571428571,
      "grad_norm": 1.0073109865188599,
      "learning_rate": 0.0005,
      "loss": 2.7125284671783447,
      "step": 2791
    },
    {
      "epoch": 0.6381714285714286,
      "grad_norm": 1.0350368022918701,
      "learning_rate": 0.0005,
      "loss": 2.6301629543304443,
      "step": 2792
    },
    {
      "epoch": 0.6384,
      "grad_norm": 1.1839325428009033,
      "learning_rate": 0.0005,
      "loss": 2.4369325637817383,
      "step": 2793
    },
    {
      "epoch": 0.6386285714285714,
      "grad_norm": 1.072524905204773,
      "learning_rate": 0.0005,
      "loss": 2.48451566696167,
      "step": 2794
    },
    {
      "epoch": 0.6388571428571429,
      "grad_norm": 0.991980254650116,
      "learning_rate": 0.0005,
      "loss": 2.6067323684692383,
      "step": 2795
    },
    {
      "epoch": 0.6390857142857143,
      "grad_norm": 1.1456986665725708,
      "learning_rate": 0.0005,
      "loss": 2.6332712173461914,
      "step": 2796
    },
    {
      "epoch": 0.6393142857142857,
      "grad_norm": 0.9831145405769348,
      "learning_rate": 0.0005,
      "loss": 3.315967082977295,
      "step": 2797
    },
    {
      "epoch": 0.6395428571428572,
      "grad_norm": 1.6073700189590454,
      "learning_rate": 0.0005,
      "loss": 2.536548614501953,
      "step": 2798
    },
    {
      "epoch": 0.6397714285714285,
      "grad_norm": 1.0277901887893677,
      "learning_rate": 0.0005,
      "loss": 2.696169853210449,
      "step": 2799
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.0485472679138184,
      "learning_rate": 0.0005,
      "loss": 2.8560404777526855,
      "step": 2800
    },
    {
      "epoch": 0.6402285714285715,
      "grad_norm": 1.0864770412445068,
      "learning_rate": 0.0005,
      "loss": 2.6256372928619385,
      "step": 2801
    },
    {
      "epoch": 0.6404571428571428,
      "grad_norm": 1.1533026695251465,
      "learning_rate": 0.0005,
      "loss": 2.92008638381958,
      "step": 2802
    },
    {
      "epoch": 0.6406857142857143,
      "grad_norm": 0.9478782415390015,
      "learning_rate": 0.0005,
      "loss": 2.691354990005493,
      "step": 2803
    },
    {
      "epoch": 0.6409142857142858,
      "grad_norm": 1.2058606147766113,
      "learning_rate": 0.0005,
      "loss": 2.608834743499756,
      "step": 2804
    },
    {
      "epoch": 0.6411428571428571,
      "grad_norm": 1.291283369064331,
      "learning_rate": 0.0005,
      "loss": 2.943145275115967,
      "step": 2805
    },
    {
      "epoch": 0.6413714285714286,
      "grad_norm": 0.927727222442627,
      "learning_rate": 0.0005,
      "loss": 2.6112918853759766,
      "step": 2806
    },
    {
      "epoch": 0.6416,
      "grad_norm": 1.110029935836792,
      "learning_rate": 0.0005,
      "loss": 2.831657648086548,
      "step": 2807
    },
    {
      "epoch": 0.6418285714285714,
      "grad_norm": 1.007123351097107,
      "learning_rate": 0.0005,
      "loss": 2.853161334991455,
      "step": 2808
    },
    {
      "epoch": 0.6420571428571429,
      "grad_norm": 0.9524242877960205,
      "learning_rate": 0.0005,
      "loss": 2.9447977542877197,
      "step": 2809
    },
    {
      "epoch": 0.6422857142857142,
      "grad_norm": 0.9220031499862671,
      "learning_rate": 0.0005,
      "loss": 2.7120161056518555,
      "step": 2810
    },
    {
      "epoch": 0.6425142857142857,
      "grad_norm": 0.85962975025177,
      "learning_rate": 0.0005,
      "loss": 2.742709159851074,
      "step": 2811
    },
    {
      "epoch": 0.6427428571428572,
      "grad_norm": 1.619490623474121,
      "learning_rate": 0.0005,
      "loss": 2.835948944091797,
      "step": 2812
    },
    {
      "epoch": 0.6429714285714285,
      "grad_norm": 1.1548343896865845,
      "learning_rate": 0.0005,
      "loss": 2.938406467437744,
      "step": 2813
    },
    {
      "epoch": 0.6432,
      "grad_norm": 1.2743077278137207,
      "learning_rate": 0.0005,
      "loss": 2.6027042865753174,
      "step": 2814
    },
    {
      "epoch": 0.6434285714285715,
      "grad_norm": 1.06889808177948,
      "learning_rate": 0.0005,
      "loss": 2.8347272872924805,
      "step": 2815
    },
    {
      "epoch": 0.6436571428571428,
      "grad_norm": 0.8621300458908081,
      "learning_rate": 0.0005,
      "loss": 2.344574451446533,
      "step": 2816
    },
    {
      "epoch": 0.6438857142857143,
      "grad_norm": 0.9855267405509949,
      "learning_rate": 0.0005,
      "loss": 2.65535306930542,
      "step": 2817
    },
    {
      "epoch": 0.6441142857142858,
      "grad_norm": 1.6735728979110718,
      "learning_rate": 0.0005,
      "loss": 2.7999870777130127,
      "step": 2818
    },
    {
      "epoch": 0.6443428571428571,
      "grad_norm": 1.0891761779785156,
      "learning_rate": 0.0005,
      "loss": 2.94193172454834,
      "step": 2819
    },
    {
      "epoch": 0.6445714285714286,
      "grad_norm": 1.0905957221984863,
      "learning_rate": 0.0005,
      "loss": 2.91347074508667,
      "step": 2820
    },
    {
      "epoch": 0.6448,
      "grad_norm": 1.2576262950897217,
      "learning_rate": 0.0005,
      "loss": 2.9588205814361572,
      "step": 2821
    },
    {
      "epoch": 0.6450285714285714,
      "grad_norm": 1.035155177116394,
      "learning_rate": 0.0005,
      "loss": 2.412008047103882,
      "step": 2822
    },
    {
      "epoch": 0.6452571428571429,
      "grad_norm": 1.0941061973571777,
      "learning_rate": 0.0005,
      "loss": 2.5129995346069336,
      "step": 2823
    },
    {
      "epoch": 0.6454857142857143,
      "grad_norm": 1.1503846645355225,
      "learning_rate": 0.0005,
      "loss": 2.9490015506744385,
      "step": 2824
    },
    {
      "epoch": 0.6457142857142857,
      "grad_norm": 1.1951316595077515,
      "learning_rate": 0.0005,
      "loss": 2.778118371963501,
      "step": 2825
    },
    {
      "epoch": 0.6459428571428572,
      "grad_norm": 1.0143218040466309,
      "learning_rate": 0.0005,
      "loss": 2.570094347000122,
      "step": 2826
    },
    {
      "epoch": 0.6461714285714286,
      "grad_norm": 1.003889799118042,
      "learning_rate": 0.0005,
      "loss": 2.7312393188476562,
      "step": 2827
    },
    {
      "epoch": 0.6464,
      "grad_norm": 0.9688799381256104,
      "learning_rate": 0.0005,
      "loss": 2.530651330947876,
      "step": 2828
    },
    {
      "epoch": 0.6466285714285714,
      "grad_norm": 1.3223224878311157,
      "learning_rate": 0.0005,
      "loss": 2.859189510345459,
      "step": 2829
    },
    {
      "epoch": 0.6468571428571429,
      "grad_norm": 1.125491976737976,
      "learning_rate": 0.0005,
      "loss": 2.507276773452759,
      "step": 2830
    },
    {
      "epoch": 0.6470857142857143,
      "grad_norm": 1.2768105268478394,
      "learning_rate": 0.0005,
      "loss": 2.6171231269836426,
      "step": 2831
    },
    {
      "epoch": 0.6473142857142857,
      "grad_norm": 1.3132787942886353,
      "learning_rate": 0.0005,
      "loss": 2.494591236114502,
      "step": 2832
    },
    {
      "epoch": 0.6475428571428572,
      "grad_norm": 0.9283239841461182,
      "learning_rate": 0.0005,
      "loss": 2.5898780822753906,
      "step": 2833
    },
    {
      "epoch": 0.6477714285714286,
      "grad_norm": 1.0090315341949463,
      "learning_rate": 0.0005,
      "loss": 2.7044286727905273,
      "step": 2834
    },
    {
      "epoch": 0.648,
      "grad_norm": 0.8721749186515808,
      "learning_rate": 0.0005,
      "loss": 2.729158878326416,
      "step": 2835
    },
    {
      "epoch": 0.6482285714285714,
      "grad_norm": 1.0950640439987183,
      "learning_rate": 0.0005,
      "loss": 2.7462852001190186,
      "step": 2836
    },
    {
      "epoch": 0.6484571428571428,
      "grad_norm": 1.0922203063964844,
      "learning_rate": 0.0005,
      "loss": 2.9687952995300293,
      "step": 2837
    },
    {
      "epoch": 0.6486857142857143,
      "grad_norm": 1.2561763525009155,
      "learning_rate": 0.0005,
      "loss": 2.615004539489746,
      "step": 2838
    },
    {
      "epoch": 0.6489142857142857,
      "grad_norm": 1.0563859939575195,
      "learning_rate": 0.0005,
      "loss": 2.8810479640960693,
      "step": 2839
    },
    {
      "epoch": 0.6491428571428571,
      "grad_norm": 1.162387728691101,
      "learning_rate": 0.0005,
      "loss": 2.851699113845825,
      "step": 2840
    },
    {
      "epoch": 0.6493714285714286,
      "grad_norm": 0.9139289259910583,
      "learning_rate": 0.0005,
      "loss": 2.74631404876709,
      "step": 2841
    },
    {
      "epoch": 0.6496,
      "grad_norm": 1.302286982536316,
      "learning_rate": 0.0005,
      "loss": 3.0258188247680664,
      "step": 2842
    },
    {
      "epoch": 0.6498285714285714,
      "grad_norm": 1.0619992017745972,
      "learning_rate": 0.0005,
      "loss": 2.5139894485473633,
      "step": 2843
    },
    {
      "epoch": 0.6500571428571429,
      "grad_norm": 0.8810482621192932,
      "learning_rate": 0.0005,
      "loss": 2.670795440673828,
      "step": 2844
    },
    {
      "epoch": 0.6502857142857142,
      "grad_norm": 1.4642388820648193,
      "learning_rate": 0.0005,
      "loss": 2.8422160148620605,
      "step": 2845
    },
    {
      "epoch": 0.6505142857142857,
      "grad_norm": 0.9242563247680664,
      "learning_rate": 0.0005,
      "loss": 2.6948976516723633,
      "step": 2846
    },
    {
      "epoch": 0.6507428571428572,
      "grad_norm": 1.2676653861999512,
      "learning_rate": 0.0005,
      "loss": 2.4798755645751953,
      "step": 2847
    },
    {
      "epoch": 0.6509714285714285,
      "grad_norm": 1.3019726276397705,
      "learning_rate": 0.0005,
      "loss": 2.727142810821533,
      "step": 2848
    },
    {
      "epoch": 0.6512,
      "grad_norm": 0.9409902691841125,
      "learning_rate": 0.0005,
      "loss": 2.3482882976531982,
      "step": 2849
    },
    {
      "epoch": 0.6514285714285715,
      "grad_norm": 1.8152731657028198,
      "learning_rate": 0.0005,
      "loss": 2.5331790447235107,
      "step": 2850
    },
    {
      "epoch": 0.6516571428571428,
      "grad_norm": 1.2538394927978516,
      "learning_rate": 0.0005,
      "loss": 2.5952324867248535,
      "step": 2851
    },
    {
      "epoch": 0.6518857142857143,
      "grad_norm": 0.9208309650421143,
      "learning_rate": 0.0005,
      "loss": 3.1170849800109863,
      "step": 2852
    },
    {
      "epoch": 0.6521142857142858,
      "grad_norm": 1.004620909690857,
      "learning_rate": 0.0005,
      "loss": 2.4686107635498047,
      "step": 2853
    },
    {
      "epoch": 0.6523428571428571,
      "grad_norm": 1.0292760133743286,
      "learning_rate": 0.0005,
      "loss": 2.758262872695923,
      "step": 2854
    },
    {
      "epoch": 0.6525714285714286,
      "grad_norm": 0.9802204370498657,
      "learning_rate": 0.0005,
      "loss": 2.461890697479248,
      "step": 2855
    },
    {
      "epoch": 0.6528,
      "grad_norm": 1.2699466943740845,
      "learning_rate": 0.0005,
      "loss": 2.836012363433838,
      "step": 2856
    },
    {
      "epoch": 0.6530285714285714,
      "grad_norm": 1.0034750699996948,
      "learning_rate": 0.0005,
      "loss": 2.65960431098938,
      "step": 2857
    },
    {
      "epoch": 0.6532571428571429,
      "grad_norm": 2.7474000453948975,
      "learning_rate": 0.0005,
      "loss": 2.7962276935577393,
      "step": 2858
    },
    {
      "epoch": 0.6534857142857143,
      "grad_norm": 1.1539735794067383,
      "learning_rate": 0.0005,
      "loss": 2.5161783695220947,
      "step": 2859
    },
    {
      "epoch": 0.6537142857142857,
      "grad_norm": 1.0545772314071655,
      "learning_rate": 0.0005,
      "loss": 2.8027052879333496,
      "step": 2860
    },
    {
      "epoch": 0.6539428571428572,
      "grad_norm": 0.9302495718002319,
      "learning_rate": 0.0005,
      "loss": 2.4874393939971924,
      "step": 2861
    },
    {
      "epoch": 0.6541714285714286,
      "grad_norm": 0.9596552848815918,
      "learning_rate": 0.0005,
      "loss": 2.8380320072174072,
      "step": 2862
    },
    {
      "epoch": 0.6544,
      "grad_norm": 1.4816275835037231,
      "learning_rate": 0.0005,
      "loss": 3.0040557384490967,
      "step": 2863
    },
    {
      "epoch": 0.6546285714285714,
      "grad_norm": 1.0909383296966553,
      "learning_rate": 0.0005,
      "loss": 3.0113301277160645,
      "step": 2864
    },
    {
      "epoch": 0.6548571428571428,
      "grad_norm": 0.8834699988365173,
      "learning_rate": 0.0005,
      "loss": 2.413079023361206,
      "step": 2865
    },
    {
      "epoch": 0.6550857142857143,
      "grad_norm": 0.9896523356437683,
      "learning_rate": 0.0005,
      "loss": 2.761068820953369,
      "step": 2866
    },
    {
      "epoch": 0.6553142857142857,
      "grad_norm": 0.9627008438110352,
      "learning_rate": 0.0005,
      "loss": 2.6829357147216797,
      "step": 2867
    },
    {
      "epoch": 0.6555428571428571,
      "grad_norm": 0.840600848197937,
      "learning_rate": 0.0005,
      "loss": 2.56561017036438,
      "step": 2868
    },
    {
      "epoch": 0.6557714285714286,
      "grad_norm": 1.2137178182601929,
      "learning_rate": 0.0005,
      "loss": 2.829895496368408,
      "step": 2869
    },
    {
      "epoch": 0.656,
      "grad_norm": 1.2282193899154663,
      "learning_rate": 0.0005,
      "loss": 2.5892069339752197,
      "step": 2870
    },
    {
      "epoch": 0.6562285714285714,
      "grad_norm": 1.3390543460845947,
      "learning_rate": 0.0005,
      "loss": 2.7939984798431396,
      "step": 2871
    },
    {
      "epoch": 0.6564571428571429,
      "grad_norm": 1.192378282546997,
      "learning_rate": 0.0005,
      "loss": 3.04768443107605,
      "step": 2872
    },
    {
      "epoch": 0.6566857142857143,
      "grad_norm": 1.3893944025039673,
      "learning_rate": 0.0005,
      "loss": 2.9152231216430664,
      "step": 2873
    },
    {
      "epoch": 0.6569142857142857,
      "grad_norm": 1.3841978311538696,
      "learning_rate": 0.0005,
      "loss": 2.5129387378692627,
      "step": 2874
    },
    {
      "epoch": 0.6571428571428571,
      "grad_norm": 1.0922385454177856,
      "learning_rate": 0.0005,
      "loss": 2.7267656326293945,
      "step": 2875
    },
    {
      "epoch": 0.6573714285714286,
      "grad_norm": 1.0388587713241577,
      "learning_rate": 0.0005,
      "loss": 2.6623353958129883,
      "step": 2876
    },
    {
      "epoch": 0.6576,
      "grad_norm": 0.9397402405738831,
      "learning_rate": 0.0005,
      "loss": 2.5531115531921387,
      "step": 2877
    },
    {
      "epoch": 0.6578285714285714,
      "grad_norm": 1.6256840229034424,
      "learning_rate": 0.0005,
      "loss": 2.833449363708496,
      "step": 2878
    },
    {
      "epoch": 0.6580571428571429,
      "grad_norm": 1.0686994791030884,
      "learning_rate": 0.0005,
      "loss": 2.662264347076416,
      "step": 2879
    },
    {
      "epoch": 0.6582857142857143,
      "grad_norm": 1.4175405502319336,
      "learning_rate": 0.0005,
      "loss": 2.615750312805176,
      "step": 2880
    },
    {
      "epoch": 0.6585142857142857,
      "grad_norm": 1.1181353330612183,
      "learning_rate": 0.0005,
      "loss": 2.5775463581085205,
      "step": 2881
    },
    {
      "epoch": 0.6587428571428572,
      "grad_norm": 0.968070387840271,
      "learning_rate": 0.0005,
      "loss": 2.6895337104797363,
      "step": 2882
    },
    {
      "epoch": 0.6589714285714285,
      "grad_norm": 0.8496088981628418,
      "learning_rate": 0.0005,
      "loss": 2.7631702423095703,
      "step": 2883
    },
    {
      "epoch": 0.6592,
      "grad_norm": 1.0397616624832153,
      "learning_rate": 0.0005,
      "loss": 2.5683047771453857,
      "step": 2884
    },
    {
      "epoch": 0.6594285714285715,
      "grad_norm": 1.1572366952896118,
      "learning_rate": 0.0005,
      "loss": 2.9470343589782715,
      "step": 2885
    },
    {
      "epoch": 0.6596571428571428,
      "grad_norm": 1.369352102279663,
      "learning_rate": 0.0005,
      "loss": 2.8550946712493896,
      "step": 2886
    },
    {
      "epoch": 0.6598857142857143,
      "grad_norm": 1.1781920194625854,
      "learning_rate": 0.0005,
      "loss": 2.6787595748901367,
      "step": 2887
    },
    {
      "epoch": 0.6601142857142858,
      "grad_norm": 0.9461081027984619,
      "learning_rate": 0.0005,
      "loss": 2.510788917541504,
      "step": 2888
    },
    {
      "epoch": 0.6603428571428571,
      "grad_norm": 0.8895004391670227,
      "learning_rate": 0.0005,
      "loss": 2.75337553024292,
      "step": 2889
    },
    {
      "epoch": 0.6605714285714286,
      "grad_norm": 0.9470312595367432,
      "learning_rate": 0.0005,
      "loss": 2.865352153778076,
      "step": 2890
    },
    {
      "epoch": 0.6608,
      "grad_norm": 0.9338122606277466,
      "learning_rate": 0.0005,
      "loss": 2.655808925628662,
      "step": 2891
    },
    {
      "epoch": 0.6610285714285714,
      "grad_norm": 0.8626226782798767,
      "learning_rate": 0.0005,
      "loss": 2.5886306762695312,
      "step": 2892
    },
    {
      "epoch": 0.6612571428571429,
      "grad_norm": 1.1060367822647095,
      "learning_rate": 0.0005,
      "loss": 2.5246224403381348,
      "step": 2893
    },
    {
      "epoch": 0.6614857142857142,
      "grad_norm": 1.072784185409546,
      "learning_rate": 0.0005,
      "loss": 2.5433690547943115,
      "step": 2894
    },
    {
      "epoch": 0.6617142857142857,
      "grad_norm": 0.8231463432312012,
      "learning_rate": 0.0005,
      "loss": 2.7527976036071777,
      "step": 2895
    },
    {
      "epoch": 0.6619428571428572,
      "grad_norm": 1.4054096937179565,
      "learning_rate": 0.0005,
      "loss": 2.6058363914489746,
      "step": 2896
    },
    {
      "epoch": 0.6621714285714285,
      "grad_norm": 1.1313382387161255,
      "learning_rate": 0.0005,
      "loss": 2.7727766036987305,
      "step": 2897
    },
    {
      "epoch": 0.6624,
      "grad_norm": 1.2944936752319336,
      "learning_rate": 0.0005,
      "loss": 2.5723814964294434,
      "step": 2898
    },
    {
      "epoch": 0.6626285714285715,
      "grad_norm": 1.1034635305404663,
      "learning_rate": 0.0005,
      "loss": 2.4401540756225586,
      "step": 2899
    },
    {
      "epoch": 0.6628571428571428,
      "grad_norm": 1.0272835493087769,
      "learning_rate": 0.0005,
      "loss": 2.738875150680542,
      "step": 2900
    },
    {
      "epoch": 0.6630857142857143,
      "grad_norm": 0.9859132766723633,
      "learning_rate": 0.0005,
      "loss": 2.4954495429992676,
      "step": 2901
    },
    {
      "epoch": 0.6633142857142857,
      "grad_norm": 1.3225005865097046,
      "learning_rate": 0.0005,
      "loss": 2.531430959701538,
      "step": 2902
    },
    {
      "epoch": 0.6635428571428571,
      "grad_norm": 0.8809033036231995,
      "learning_rate": 0.0005,
      "loss": 2.632750988006592,
      "step": 2903
    },
    {
      "epoch": 0.6637714285714286,
      "grad_norm": 0.9004866480827332,
      "learning_rate": 0.0005,
      "loss": 2.469348907470703,
      "step": 2904
    },
    {
      "epoch": 0.664,
      "grad_norm": 0.9104151129722595,
      "learning_rate": 0.0005,
      "loss": 2.686624526977539,
      "step": 2905
    },
    {
      "epoch": 0.6642285714285714,
      "grad_norm": 0.847822368144989,
      "learning_rate": 0.0005,
      "loss": 2.8502871990203857,
      "step": 2906
    },
    {
      "epoch": 0.6644571428571429,
      "grad_norm": 1.1077237129211426,
      "learning_rate": 0.0005,
      "loss": 2.5938832759857178,
      "step": 2907
    },
    {
      "epoch": 0.6646857142857143,
      "grad_norm": 1.071722388267517,
      "learning_rate": 0.0005,
      "loss": 2.6901893615722656,
      "step": 2908
    },
    {
      "epoch": 0.6649142857142857,
      "grad_norm": 0.981740415096283,
      "learning_rate": 0.0005,
      "loss": 2.5605506896972656,
      "step": 2909
    },
    {
      "epoch": 0.6651428571428571,
      "grad_norm": 1.1148350238800049,
      "learning_rate": 0.0005,
      "loss": 2.9514284133911133,
      "step": 2910
    },
    {
      "epoch": 0.6653714285714286,
      "grad_norm": 1.0307869911193848,
      "learning_rate": 0.0005,
      "loss": 2.610912322998047,
      "step": 2911
    },
    {
      "epoch": 0.6656,
      "grad_norm": 1.0932523012161255,
      "learning_rate": 0.0005,
      "loss": 2.797459125518799,
      "step": 2912
    },
    {
      "epoch": 0.6658285714285714,
      "grad_norm": 1.2968508005142212,
      "learning_rate": 0.0005,
      "loss": 2.7888431549072266,
      "step": 2913
    },
    {
      "epoch": 0.6660571428571429,
      "grad_norm": 3.497462272644043,
      "learning_rate": 0.0005,
      "loss": 2.8840243816375732,
      "step": 2914
    },
    {
      "epoch": 0.6662857142857143,
      "grad_norm": 0.9831016659736633,
      "learning_rate": 0.0005,
      "loss": 2.048450469970703,
      "step": 2915
    },
    {
      "epoch": 0.6665142857142857,
      "grad_norm": 1.0073193311691284,
      "learning_rate": 0.0005,
      "loss": 2.5392470359802246,
      "step": 2916
    },
    {
      "epoch": 0.6667428571428572,
      "grad_norm": 0.9904738664627075,
      "learning_rate": 0.0005,
      "loss": 2.6433215141296387,
      "step": 2917
    },
    {
      "epoch": 0.6669714285714285,
      "grad_norm": 0.8652747869491577,
      "learning_rate": 0.0005,
      "loss": 2.9691641330718994,
      "step": 2918
    },
    {
      "epoch": 0.6672,
      "grad_norm": 1.0660065412521362,
      "learning_rate": 0.0005,
      "loss": 2.630642890930176,
      "step": 2919
    },
    {
      "epoch": 0.6674285714285715,
      "grad_norm": 1.207151174545288,
      "learning_rate": 0.0005,
      "loss": 2.642967700958252,
      "step": 2920
    },
    {
      "epoch": 0.6676571428571428,
      "grad_norm": 0.9362768530845642,
      "learning_rate": 0.0005,
      "loss": 2.8615899085998535,
      "step": 2921
    },
    {
      "epoch": 0.6678857142857143,
      "grad_norm": 0.9209977388381958,
      "learning_rate": 0.0005,
      "loss": 2.889920473098755,
      "step": 2922
    },
    {
      "epoch": 0.6681142857142857,
      "grad_norm": 0.9105929136276245,
      "learning_rate": 0.0005,
      "loss": 2.4978339672088623,
      "step": 2923
    },
    {
      "epoch": 0.6683428571428571,
      "grad_norm": 0.9706485271453857,
      "learning_rate": 0.0005,
      "loss": 2.6873035430908203,
      "step": 2924
    },
    {
      "epoch": 0.6685714285714286,
      "grad_norm": 1.1892011165618896,
      "learning_rate": 0.0005,
      "loss": 2.818824052810669,
      "step": 2925
    },
    {
      "epoch": 0.6688,
      "grad_norm": 1.212087869644165,
      "learning_rate": 0.0005,
      "loss": 2.811040163040161,
      "step": 2926
    },
    {
      "epoch": 0.6690285714285714,
      "grad_norm": 1.9766305685043335,
      "learning_rate": 0.0005,
      "loss": 2.4662094116210938,
      "step": 2927
    },
    {
      "epoch": 0.6692571428571429,
      "grad_norm": 0.9685038924217224,
      "learning_rate": 0.0005,
      "loss": 3.023803234100342,
      "step": 2928
    },
    {
      "epoch": 0.6694857142857142,
      "grad_norm": 1.0114648342132568,
      "learning_rate": 0.0005,
      "loss": 2.7058892250061035,
      "step": 2929
    },
    {
      "epoch": 0.6697142857142857,
      "grad_norm": 1.0060362815856934,
      "learning_rate": 0.0005,
      "loss": 2.9307429790496826,
      "step": 2930
    },
    {
      "epoch": 0.6699428571428572,
      "grad_norm": 0.8845807909965515,
      "learning_rate": 0.0005,
      "loss": 2.7098982334136963,
      "step": 2931
    },
    {
      "epoch": 0.6701714285714285,
      "grad_norm": 1.1731393337249756,
      "learning_rate": 0.0005,
      "loss": 2.4427032470703125,
      "step": 2932
    },
    {
      "epoch": 0.6704,
      "grad_norm": 1.059842824935913,
      "learning_rate": 0.0005,
      "loss": 2.7198963165283203,
      "step": 2933
    },
    {
      "epoch": 0.6706285714285715,
      "grad_norm": 1.1527998447418213,
      "learning_rate": 0.0005,
      "loss": 2.6738901138305664,
      "step": 2934
    },
    {
      "epoch": 0.6708571428571428,
      "grad_norm": 0.886830747127533,
      "learning_rate": 0.0005,
      "loss": 2.575073719024658,
      "step": 2935
    },
    {
      "epoch": 0.6710857142857143,
      "grad_norm": 0.8666631579399109,
      "learning_rate": 0.0005,
      "loss": 2.6585795879364014,
      "step": 2936
    },
    {
      "epoch": 0.6713142857142858,
      "grad_norm": 1.0520401000976562,
      "learning_rate": 0.0005,
      "loss": 2.5703885555267334,
      "step": 2937
    },
    {
      "epoch": 0.6715428571428571,
      "grad_norm": 1.2411590814590454,
      "learning_rate": 0.0005,
      "loss": 2.2901480197906494,
      "step": 2938
    },
    {
      "epoch": 0.6717714285714286,
      "grad_norm": 1.078282117843628,
      "learning_rate": 0.0005,
      "loss": 2.68522572517395,
      "step": 2939
    },
    {
      "epoch": 0.672,
      "grad_norm": 1.2027946710586548,
      "learning_rate": 0.0005,
      "loss": 2.7762250900268555,
      "step": 2940
    },
    {
      "epoch": 0.6722285714285714,
      "grad_norm": 0.8654278516769409,
      "learning_rate": 0.0005,
      "loss": 2.824462413787842,
      "step": 2941
    },
    {
      "epoch": 0.6724571428571429,
      "grad_norm": 1.102334976196289,
      "learning_rate": 0.0005,
      "loss": 2.642012596130371,
      "step": 2942
    },
    {
      "epoch": 0.6726857142857143,
      "grad_norm": 1.0339983701705933,
      "learning_rate": 0.0005,
      "loss": 2.886962890625,
      "step": 2943
    },
    {
      "epoch": 0.6729142857142857,
      "grad_norm": 1.0527592897415161,
      "learning_rate": 0.0005,
      "loss": 2.756866455078125,
      "step": 2944
    },
    {
      "epoch": 0.6731428571428572,
      "grad_norm": 2.427387237548828,
      "learning_rate": 0.0005,
      "loss": 2.570894718170166,
      "step": 2945
    },
    {
      "epoch": 0.6733714285714286,
      "grad_norm": 1.1500786542892456,
      "learning_rate": 0.0005,
      "loss": 2.664155960083008,
      "step": 2946
    },
    {
      "epoch": 0.6736,
      "grad_norm": 0.9669901728630066,
      "learning_rate": 0.0005,
      "loss": 2.635955572128296,
      "step": 2947
    },
    {
      "epoch": 0.6738285714285714,
      "grad_norm": 1.0884662866592407,
      "learning_rate": 0.0005,
      "loss": 2.744788646697998,
      "step": 2948
    },
    {
      "epoch": 0.6740571428571429,
      "grad_norm": 1.7180957794189453,
      "learning_rate": 0.0005,
      "loss": 2.3721609115600586,
      "step": 2949
    },
    {
      "epoch": 0.6742857142857143,
      "grad_norm": 1.0797576904296875,
      "learning_rate": 0.0005,
      "loss": 3.049055576324463,
      "step": 2950
    },
    {
      "epoch": 0.6745142857142857,
      "grad_norm": 0.7843718528747559,
      "learning_rate": 0.0005,
      "loss": 2.5361831188201904,
      "step": 2951
    },
    {
      "epoch": 0.6747428571428571,
      "grad_norm": 0.9525741934776306,
      "learning_rate": 0.0005,
      "loss": 2.5590667724609375,
      "step": 2952
    },
    {
      "epoch": 0.6749714285714286,
      "grad_norm": 1.388689637184143,
      "learning_rate": 0.0005,
      "loss": 2.631436824798584,
      "step": 2953
    },
    {
      "epoch": 0.6752,
      "grad_norm": 1.0536357164382935,
      "learning_rate": 0.0005,
      "loss": 2.8747777938842773,
      "step": 2954
    },
    {
      "epoch": 0.6754285714285714,
      "grad_norm": 1.0451740026474,
      "learning_rate": 0.0005,
      "loss": 2.92897367477417,
      "step": 2955
    },
    {
      "epoch": 0.6756571428571428,
      "grad_norm": 1.1056067943572998,
      "learning_rate": 0.0005,
      "loss": 2.4851021766662598,
      "step": 2956
    },
    {
      "epoch": 0.6758857142857143,
      "grad_norm": 1.0864462852478027,
      "learning_rate": 0.0005,
      "loss": 2.800198554992676,
      "step": 2957
    },
    {
      "epoch": 0.6761142857142857,
      "grad_norm": 1.0794481039047241,
      "learning_rate": 0.0005,
      "loss": 3.0205447673797607,
      "step": 2958
    },
    {
      "epoch": 0.6763428571428571,
      "grad_norm": 1.0217063426971436,
      "learning_rate": 0.0005,
      "loss": 2.6145801544189453,
      "step": 2959
    },
    {
      "epoch": 0.6765714285714286,
      "grad_norm": 0.9964876770973206,
      "learning_rate": 0.0005,
      "loss": 2.8015589714050293,
      "step": 2960
    },
    {
      "epoch": 0.6768,
      "grad_norm": 1.1702009439468384,
      "learning_rate": 0.0005,
      "loss": 2.37846302986145,
      "step": 2961
    },
    {
      "epoch": 0.6770285714285714,
      "grad_norm": 1.0594127178192139,
      "learning_rate": 0.0005,
      "loss": 2.5971503257751465,
      "step": 2962
    },
    {
      "epoch": 0.6772571428571429,
      "grad_norm": 0.8469681739807129,
      "learning_rate": 0.0005,
      "loss": 2.652935028076172,
      "step": 2963
    },
    {
      "epoch": 0.6774857142857142,
      "grad_norm": 1.3820602893829346,
      "learning_rate": 0.0005,
      "loss": 2.7699942588806152,
      "step": 2964
    },
    {
      "epoch": 0.6777142857142857,
      "grad_norm": 1.679200530052185,
      "learning_rate": 0.0005,
      "loss": 2.861600160598755,
      "step": 2965
    },
    {
      "epoch": 0.6779428571428572,
      "grad_norm": 0.8910651803016663,
      "learning_rate": 0.0005,
      "loss": 2.7423501014709473,
      "step": 2966
    },
    {
      "epoch": 0.6781714285714285,
      "grad_norm": 1.1634992361068726,
      "learning_rate": 0.0005,
      "loss": 2.906547784805298,
      "step": 2967
    },
    {
      "epoch": 0.6784,
      "grad_norm": 1.1666896343231201,
      "learning_rate": 0.0005,
      "loss": 2.6760902404785156,
      "step": 2968
    },
    {
      "epoch": 0.6786285714285715,
      "grad_norm": 0.8417664170265198,
      "learning_rate": 0.0005,
      "loss": 2.934316396713257,
      "step": 2969
    },
    {
      "epoch": 0.6788571428571428,
      "grad_norm": 1.2062968015670776,
      "learning_rate": 0.0005,
      "loss": 2.762714385986328,
      "step": 2970
    },
    {
      "epoch": 0.6790857142857143,
      "grad_norm": 1.0148860216140747,
      "learning_rate": 0.0005,
      "loss": 2.6541643142700195,
      "step": 2971
    },
    {
      "epoch": 0.6793142857142858,
      "grad_norm": 1.421179175376892,
      "learning_rate": 0.0005,
      "loss": 2.5684947967529297,
      "step": 2972
    },
    {
      "epoch": 0.6795428571428571,
      "grad_norm": 1.4908174276351929,
      "learning_rate": 0.0005,
      "loss": 2.740501880645752,
      "step": 2973
    },
    {
      "epoch": 0.6797714285714286,
      "grad_norm": 1.1309980154037476,
      "learning_rate": 0.0005,
      "loss": 2.7032337188720703,
      "step": 2974
    },
    {
      "epoch": 0.68,
      "grad_norm": 1.0380780696868896,
      "learning_rate": 0.0005,
      "loss": 2.9678115844726562,
      "step": 2975
    },
    {
      "epoch": 0.6802285714285714,
      "grad_norm": 1.0521490573883057,
      "learning_rate": 0.0005,
      "loss": 2.3791842460632324,
      "step": 2976
    },
    {
      "epoch": 0.6804571428571429,
      "grad_norm": 1.0862442255020142,
      "learning_rate": 0.0005,
      "loss": 2.3420944213867188,
      "step": 2977
    },
    {
      "epoch": 0.6806857142857143,
      "grad_norm": 1.0614124536514282,
      "learning_rate": 0.0005,
      "loss": 2.795271873474121,
      "step": 2978
    },
    {
      "epoch": 0.6809142857142857,
      "grad_norm": 1.9327149391174316,
      "learning_rate": 0.0005,
      "loss": 2.2864341735839844,
      "step": 2979
    },
    {
      "epoch": 0.6811428571428572,
      "grad_norm": 1.191240668296814,
      "learning_rate": 0.0005,
      "loss": 2.8323287963867188,
      "step": 2980
    },
    {
      "epoch": 0.6813714285714285,
      "grad_norm": 0.9724410772323608,
      "learning_rate": 0.0005,
      "loss": 2.5954136848449707,
      "step": 2981
    },
    {
      "epoch": 0.6816,
      "grad_norm": 1.285101056098938,
      "learning_rate": 0.0005,
      "loss": 2.627382278442383,
      "step": 2982
    },
    {
      "epoch": 0.6818285714285715,
      "grad_norm": 1.39792799949646,
      "learning_rate": 0.0005,
      "loss": 2.508798122406006,
      "step": 2983
    },
    {
      "epoch": 0.6820571428571428,
      "grad_norm": 1.0093038082122803,
      "learning_rate": 0.0005,
      "loss": 2.6936659812927246,
      "step": 2984
    },
    {
      "epoch": 0.6822857142857143,
      "grad_norm": 1.0831307172775269,
      "learning_rate": 0.0005,
      "loss": 2.4935190677642822,
      "step": 2985
    },
    {
      "epoch": 0.6825142857142857,
      "grad_norm": 1.07633376121521,
      "learning_rate": 0.0005,
      "loss": 2.6377058029174805,
      "step": 2986
    },
    {
      "epoch": 0.6827428571428571,
      "grad_norm": 1.0891820192337036,
      "learning_rate": 0.0005,
      "loss": 2.7507669925689697,
      "step": 2987
    },
    {
      "epoch": 0.6829714285714286,
      "grad_norm": 1.3848367929458618,
      "learning_rate": 0.0005,
      "loss": 2.5820531845092773,
      "step": 2988
    },
    {
      "epoch": 0.6832,
      "grad_norm": 1.2477999925613403,
      "learning_rate": 0.0005,
      "loss": 2.820690155029297,
      "step": 2989
    },
    {
      "epoch": 0.6834285714285714,
      "grad_norm": 1.1009516716003418,
      "learning_rate": 0.0005,
      "loss": 2.304945230484009,
      "step": 2990
    },
    {
      "epoch": 0.6836571428571429,
      "grad_norm": 1.1305897235870361,
      "learning_rate": 0.0005,
      "loss": 2.4227373600006104,
      "step": 2991
    },
    {
      "epoch": 0.6838857142857143,
      "grad_norm": 1.0582659244537354,
      "learning_rate": 0.0005,
      "loss": 2.309904098510742,
      "step": 2992
    },
    {
      "epoch": 0.6841142857142857,
      "grad_norm": 0.9538050889968872,
      "learning_rate": 0.0005,
      "loss": 2.515275478363037,
      "step": 2993
    },
    {
      "epoch": 0.6843428571428571,
      "grad_norm": 1.1273473501205444,
      "learning_rate": 0.0005,
      "loss": 2.5929486751556396,
      "step": 2994
    },
    {
      "epoch": 0.6845714285714286,
      "grad_norm": 1.415388584136963,
      "learning_rate": 0.0005,
      "loss": 2.9285240173339844,
      "step": 2995
    },
    {
      "epoch": 0.6848,
      "grad_norm": 1.3613276481628418,
      "learning_rate": 0.0005,
      "loss": 2.729349374771118,
      "step": 2996
    },
    {
      "epoch": 0.6850285714285714,
      "grad_norm": 0.8635530471801758,
      "learning_rate": 0.0005,
      "loss": 2.5550239086151123,
      "step": 2997
    },
    {
      "epoch": 0.6852571428571429,
      "grad_norm": 1.142272710800171,
      "learning_rate": 0.0005,
      "loss": 2.551535129547119,
      "step": 2998
    },
    {
      "epoch": 0.6854857142857143,
      "grad_norm": 1.566536545753479,
      "learning_rate": 0.0005,
      "loss": 2.685544490814209,
      "step": 2999
    },
    {
      "epoch": 0.6857142857142857,
      "grad_norm": 0.9630842208862305,
      "learning_rate": 0.0005,
      "loss": 2.5840272903442383,
      "step": 3000
    },
    {
      "epoch": 0.6859428571428572,
      "grad_norm": 1.3259403705596924,
      "learning_rate": 0.0005,
      "loss": 2.8474836349487305,
      "step": 3001
    },
    {
      "epoch": 0.6861714285714285,
      "grad_norm": 1.0683013200759888,
      "learning_rate": 0.0005,
      "loss": 2.513185501098633,
      "step": 3002
    },
    {
      "epoch": 0.6864,
      "grad_norm": 1.621461272239685,
      "learning_rate": 0.0005,
      "loss": 2.768850326538086,
      "step": 3003
    },
    {
      "epoch": 0.6866285714285715,
      "grad_norm": 1.0987175703048706,
      "learning_rate": 0.0005,
      "loss": 2.497750997543335,
      "step": 3004
    },
    {
      "epoch": 0.6868571428571428,
      "grad_norm": 0.9377861022949219,
      "learning_rate": 0.0005,
      "loss": 2.932436466217041,
      "step": 3005
    },
    {
      "epoch": 0.6870857142857143,
      "grad_norm": 1.2197401523590088,
      "learning_rate": 0.0005,
      "loss": 3.0159997940063477,
      "step": 3006
    },
    {
      "epoch": 0.6873142857142858,
      "grad_norm": 1.0797219276428223,
      "learning_rate": 0.0005,
      "loss": 2.910827875137329,
      "step": 3007
    },
    {
      "epoch": 0.6875428571428571,
      "grad_norm": 1.342504858970642,
      "learning_rate": 0.0005,
      "loss": 2.7704832553863525,
      "step": 3008
    },
    {
      "epoch": 0.6877714285714286,
      "grad_norm": 1.0132254362106323,
      "learning_rate": 0.0005,
      "loss": 2.829195737838745,
      "step": 3009
    },
    {
      "epoch": 0.688,
      "grad_norm": 0.925161600112915,
      "learning_rate": 0.0005,
      "loss": 2.519608497619629,
      "step": 3010
    },
    {
      "epoch": 0.6882285714285714,
      "grad_norm": 0.9195889234542847,
      "learning_rate": 0.0005,
      "loss": 2.570080280303955,
      "step": 3011
    },
    {
      "epoch": 0.6884571428571429,
      "grad_norm": 0.9389296174049377,
      "learning_rate": 0.0005,
      "loss": 2.207716464996338,
      "step": 3012
    },
    {
      "epoch": 0.6886857142857142,
      "grad_norm": 1.4023596048355103,
      "learning_rate": 0.0005,
      "loss": 2.825148820877075,
      "step": 3013
    },
    {
      "epoch": 0.6889142857142857,
      "grad_norm": 1.1924279928207397,
      "learning_rate": 0.0005,
      "loss": 2.5332021713256836,
      "step": 3014
    },
    {
      "epoch": 0.6891428571428572,
      "grad_norm": 0.916262149810791,
      "learning_rate": 0.0005,
      "loss": 2.5833048820495605,
      "step": 3015
    },
    {
      "epoch": 0.6893714285714285,
      "grad_norm": 1.2988665103912354,
      "learning_rate": 0.0005,
      "loss": 2.8290584087371826,
      "step": 3016
    },
    {
      "epoch": 0.6896,
      "grad_norm": 0.9308931231498718,
      "learning_rate": 0.0005,
      "loss": 2.987727165222168,
      "step": 3017
    },
    {
      "epoch": 0.6898285714285715,
      "grad_norm": 1.0365182161331177,
      "learning_rate": 0.0005,
      "loss": 2.5104384422302246,
      "step": 3018
    },
    {
      "epoch": 0.6900571428571428,
      "grad_norm": 0.8432320356369019,
      "learning_rate": 0.0005,
      "loss": 2.867231845855713,
      "step": 3019
    },
    {
      "epoch": 0.6902857142857143,
      "grad_norm": 1.4421619176864624,
      "learning_rate": 0.0005,
      "loss": 2.9753923416137695,
      "step": 3020
    },
    {
      "epoch": 0.6905142857142857,
      "grad_norm": 1.0069139003753662,
      "learning_rate": 0.0005,
      "loss": 2.3022170066833496,
      "step": 3021
    },
    {
      "epoch": 0.6907428571428571,
      "grad_norm": 1.0769405364990234,
      "learning_rate": 0.0005,
      "loss": 2.3901891708374023,
      "step": 3022
    },
    {
      "epoch": 0.6909714285714286,
      "grad_norm": 0.9051018357276917,
      "learning_rate": 0.0005,
      "loss": 2.845259189605713,
      "step": 3023
    },
    {
      "epoch": 0.6912,
      "grad_norm": 1.298079013824463,
      "learning_rate": 0.0005,
      "loss": 2.2959909439086914,
      "step": 3024
    },
    {
      "epoch": 0.6914285714285714,
      "grad_norm": 1.2264211177825928,
      "learning_rate": 0.0005,
      "loss": 2.319948673248291,
      "step": 3025
    },
    {
      "epoch": 0.6916571428571429,
      "grad_norm": 1.19206702709198,
      "learning_rate": 0.0005,
      "loss": 2.9250364303588867,
      "step": 3026
    },
    {
      "epoch": 0.6918857142857143,
      "grad_norm": 0.9684996604919434,
      "learning_rate": 0.0005,
      "loss": 2.894965648651123,
      "step": 3027
    },
    {
      "epoch": 0.6921142857142857,
      "grad_norm": 1.0107241868972778,
      "learning_rate": 0.0005,
      "loss": 2.417898654937744,
      "step": 3028
    },
    {
      "epoch": 0.6923428571428571,
      "grad_norm": 1.3176053762435913,
      "learning_rate": 0.0005,
      "loss": 2.7324626445770264,
      "step": 3029
    },
    {
      "epoch": 0.6925714285714286,
      "grad_norm": 1.9062138795852661,
      "learning_rate": 0.0005,
      "loss": 2.8214774131774902,
      "step": 3030
    },
    {
      "epoch": 0.6928,
      "grad_norm": 0.8611826300621033,
      "learning_rate": 0.0005,
      "loss": 2.4743127822875977,
      "step": 3031
    },
    {
      "epoch": 0.6930285714285714,
      "grad_norm": 0.8173714876174927,
      "learning_rate": 0.0005,
      "loss": 2.8423562049865723,
      "step": 3032
    },
    {
      "epoch": 0.6932571428571429,
      "grad_norm": 0.9548685550689697,
      "learning_rate": 0.0005,
      "loss": 2.8507275581359863,
      "step": 3033
    },
    {
      "epoch": 0.6934857142857143,
      "grad_norm": 0.9457232356071472,
      "learning_rate": 0.0005,
      "loss": 2.5583834648132324,
      "step": 3034
    },
    {
      "epoch": 0.6937142857142857,
      "grad_norm": 1.0479254722595215,
      "learning_rate": 0.0005,
      "loss": 2.570568084716797,
      "step": 3035
    },
    {
      "epoch": 0.6939428571428572,
      "grad_norm": 0.8748887181282043,
      "learning_rate": 0.0005,
      "loss": 2.8691089153289795,
      "step": 3036
    },
    {
      "epoch": 0.6941714285714286,
      "grad_norm": 1.319960594177246,
      "learning_rate": 0.0005,
      "loss": 2.559349775314331,
      "step": 3037
    },
    {
      "epoch": 0.6944,
      "grad_norm": 1.0546531677246094,
      "learning_rate": 0.0005,
      "loss": 2.8032760620117188,
      "step": 3038
    },
    {
      "epoch": 0.6946285714285715,
      "grad_norm": 1.3483316898345947,
      "learning_rate": 0.0005,
      "loss": 2.8124771118164062,
      "step": 3039
    },
    {
      "epoch": 0.6948571428571428,
      "grad_norm": 1.0483781099319458,
      "learning_rate": 0.0005,
      "loss": 2.5476772785186768,
      "step": 3040
    },
    {
      "epoch": 0.6950857142857143,
      "grad_norm": 0.9184999465942383,
      "learning_rate": 0.0005,
      "loss": 2.4992008209228516,
      "step": 3041
    },
    {
      "epoch": 0.6953142857142857,
      "grad_norm": 0.8884875774383545,
      "learning_rate": 0.0005,
      "loss": 2.705735683441162,
      "step": 3042
    },
    {
      "epoch": 0.6955428571428571,
      "grad_norm": 1.064902901649475,
      "learning_rate": 0.0005,
      "loss": 2.408229351043701,
      "step": 3043
    },
    {
      "epoch": 0.6957714285714286,
      "grad_norm": 0.998723566532135,
      "learning_rate": 0.0005,
      "loss": 2.5039358139038086,
      "step": 3044
    },
    {
      "epoch": 0.696,
      "grad_norm": 0.877332329750061,
      "learning_rate": 0.0005,
      "loss": 2.7484891414642334,
      "step": 3045
    },
    {
      "epoch": 0.6962285714285714,
      "grad_norm": 0.9590227603912354,
      "learning_rate": 0.0005,
      "loss": 2.9686169624328613,
      "step": 3046
    },
    {
      "epoch": 0.6964571428571429,
      "grad_norm": 0.9953233003616333,
      "learning_rate": 0.0005,
      "loss": 2.8694629669189453,
      "step": 3047
    },
    {
      "epoch": 0.6966857142857142,
      "grad_norm": 0.9866119623184204,
      "learning_rate": 0.0005,
      "loss": 2.511382579803467,
      "step": 3048
    },
    {
      "epoch": 0.6969142857142857,
      "grad_norm": 1.2598421573638916,
      "learning_rate": 0.0005,
      "loss": 2.5493602752685547,
      "step": 3049
    },
    {
      "epoch": 0.6971428571428572,
      "grad_norm": 0.9718234539031982,
      "learning_rate": 0.0005,
      "loss": 2.8709919452667236,
      "step": 3050
    },
    {
      "epoch": 0.6973714285714285,
      "grad_norm": 1.153194785118103,
      "learning_rate": 0.0005,
      "loss": 2.4192233085632324,
      "step": 3051
    },
    {
      "epoch": 0.6976,
      "grad_norm": 1.0863271951675415,
      "learning_rate": 0.0005,
      "loss": 2.733349323272705,
      "step": 3052
    },
    {
      "epoch": 0.6978285714285715,
      "grad_norm": 1.1631453037261963,
      "learning_rate": 0.0005,
      "loss": 2.6485815048217773,
      "step": 3053
    },
    {
      "epoch": 0.6980571428571428,
      "grad_norm": 1.3928955793380737,
      "learning_rate": 0.0005,
      "loss": 2.538196563720703,
      "step": 3054
    },
    {
      "epoch": 0.6982857142857143,
      "grad_norm": 1.2053298950195312,
      "learning_rate": 0.0005,
      "loss": 2.730569839477539,
      "step": 3055
    },
    {
      "epoch": 0.6985142857142858,
      "grad_norm": 1.023911476135254,
      "learning_rate": 0.0005,
      "loss": 2.309805393218994,
      "step": 3056
    },
    {
      "epoch": 0.6987428571428571,
      "grad_norm": 0.9301089644432068,
      "learning_rate": 0.0005,
      "loss": 2.6904780864715576,
      "step": 3057
    },
    {
      "epoch": 0.6989714285714286,
      "grad_norm": 1.2331774234771729,
      "learning_rate": 0.0005,
      "loss": 2.9185445308685303,
      "step": 3058
    },
    {
      "epoch": 0.6992,
      "grad_norm": 1.1102098226547241,
      "learning_rate": 0.0005,
      "loss": 2.393494129180908,
      "step": 3059
    },
    {
      "epoch": 0.6994285714285714,
      "grad_norm": 0.930248498916626,
      "learning_rate": 0.0005,
      "loss": 2.5940260887145996,
      "step": 3060
    },
    {
      "epoch": 0.6996571428571429,
      "grad_norm": 1.0457801818847656,
      "learning_rate": 0.0005,
      "loss": 2.787209987640381,
      "step": 3061
    },
    {
      "epoch": 0.6998857142857143,
      "grad_norm": 0.9191267490386963,
      "learning_rate": 0.0005,
      "loss": 2.4173390865325928,
      "step": 3062
    },
    {
      "epoch": 0.7001142857142857,
      "grad_norm": 1.1547424793243408,
      "learning_rate": 0.0005,
      "loss": 2.7667877674102783,
      "step": 3063
    },
    {
      "epoch": 0.7003428571428572,
      "grad_norm": 1.197133183479309,
      "learning_rate": 0.0005,
      "loss": 2.654831886291504,
      "step": 3064
    },
    {
      "epoch": 0.7005714285714286,
      "grad_norm": 1.1436057090759277,
      "learning_rate": 0.0005,
      "loss": 2.7461490631103516,
      "step": 3065
    },
    {
      "epoch": 0.7008,
      "grad_norm": 0.962299108505249,
      "learning_rate": 0.0005,
      "loss": 2.588120460510254,
      "step": 3066
    },
    {
      "epoch": 0.7010285714285714,
      "grad_norm": 1.5657068490982056,
      "learning_rate": 0.0005,
      "loss": 3.1481380462646484,
      "step": 3067
    },
    {
      "epoch": 0.7012571428571429,
      "grad_norm": 0.905645489692688,
      "learning_rate": 0.0005,
      "loss": 3.1790504455566406,
      "step": 3068
    },
    {
      "epoch": 0.7014857142857143,
      "grad_norm": 0.8789994120597839,
      "learning_rate": 0.0005,
      "loss": 2.623568058013916,
      "step": 3069
    },
    {
      "epoch": 0.7017142857142857,
      "grad_norm": 0.9144924879074097,
      "learning_rate": 0.0005,
      "loss": 2.696415901184082,
      "step": 3070
    },
    {
      "epoch": 0.7019428571428571,
      "grad_norm": 0.9713718295097351,
      "learning_rate": 0.0005,
      "loss": 2.6451821327209473,
      "step": 3071
    },
    {
      "epoch": 0.7021714285714286,
      "grad_norm": 0.917482852935791,
      "learning_rate": 0.0005,
      "loss": 2.5859274864196777,
      "step": 3072
    },
    {
      "epoch": 0.7024,
      "grad_norm": 0.9297755360603333,
      "learning_rate": 0.0005,
      "loss": 2.660524845123291,
      "step": 3073
    },
    {
      "epoch": 0.7026285714285714,
      "grad_norm": 1.3152576684951782,
      "learning_rate": 0.0005,
      "loss": 2.4931998252868652,
      "step": 3074
    },
    {
      "epoch": 0.7028571428571428,
      "grad_norm": 2.030756711959839,
      "learning_rate": 0.0005,
      "loss": 2.801398992538452,
      "step": 3075
    },
    {
      "epoch": 0.7030857142857143,
      "grad_norm": 0.8736261129379272,
      "learning_rate": 0.0005,
      "loss": 2.6372601985931396,
      "step": 3076
    },
    {
      "epoch": 0.7033142857142857,
      "grad_norm": 1.0876322984695435,
      "learning_rate": 0.0005,
      "loss": 2.6100943088531494,
      "step": 3077
    },
    {
      "epoch": 0.7035428571428571,
      "grad_norm": 0.8957040905952454,
      "learning_rate": 0.0005,
      "loss": 2.7355313301086426,
      "step": 3078
    },
    {
      "epoch": 0.7037714285714286,
      "grad_norm": 0.8959400057792664,
      "learning_rate": 0.0005,
      "loss": 2.720902442932129,
      "step": 3079
    },
    {
      "epoch": 0.704,
      "grad_norm": 0.8738780617713928,
      "learning_rate": 0.0005,
      "loss": 2.6232101917266846,
      "step": 3080
    },
    {
      "epoch": 0.7042285714285714,
      "grad_norm": 0.9575526118278503,
      "learning_rate": 0.0005,
      "loss": 2.4664204120635986,
      "step": 3081
    },
    {
      "epoch": 0.7044571428571429,
      "grad_norm": 1.0880801677703857,
      "learning_rate": 0.0005,
      "loss": 2.4593443870544434,
      "step": 3082
    },
    {
      "epoch": 0.7046857142857142,
      "grad_norm": 1.293722152709961,
      "learning_rate": 0.0005,
      "loss": 2.704510450363159,
      "step": 3083
    },
    {
      "epoch": 0.7049142857142857,
      "grad_norm": 1.1944226026535034,
      "learning_rate": 0.0005,
      "loss": 2.5861668586730957,
      "step": 3084
    },
    {
      "epoch": 0.7051428571428572,
      "grad_norm": 1.053284764289856,
      "learning_rate": 0.0005,
      "loss": 2.9142706394195557,
      "step": 3085
    },
    {
      "epoch": 0.7053714285714285,
      "grad_norm": 0.8607789874076843,
      "learning_rate": 0.0005,
      "loss": 2.4769256114959717,
      "step": 3086
    },
    {
      "epoch": 0.7056,
      "grad_norm": 1.0794126987457275,
      "learning_rate": 0.0005,
      "loss": 2.5288825035095215,
      "step": 3087
    },
    {
      "epoch": 0.7058285714285715,
      "grad_norm": 1.0345590114593506,
      "learning_rate": 0.0005,
      "loss": 2.7361044883728027,
      "step": 3088
    },
    {
      "epoch": 0.7060571428571428,
      "grad_norm": 0.9471300840377808,
      "learning_rate": 0.0005,
      "loss": 2.7933926582336426,
      "step": 3089
    },
    {
      "epoch": 0.7062857142857143,
      "grad_norm": 1.1634207963943481,
      "learning_rate": 0.0005,
      "loss": 2.5582351684570312,
      "step": 3090
    },
    {
      "epoch": 0.7065142857142858,
      "grad_norm": 1.1915229558944702,
      "learning_rate": 0.0005,
      "loss": 2.5661535263061523,
      "step": 3091
    },
    {
      "epoch": 0.7067428571428571,
      "grad_norm": 1.7702410221099854,
      "learning_rate": 0.0005,
      "loss": 2.657491683959961,
      "step": 3092
    },
    {
      "epoch": 0.7069714285714286,
      "grad_norm": 1.2045695781707764,
      "learning_rate": 0.0005,
      "loss": 3.0230863094329834,
      "step": 3093
    },
    {
      "epoch": 0.7072,
      "grad_norm": 1.081889033317566,
      "learning_rate": 0.0005,
      "loss": 2.9436230659484863,
      "step": 3094
    },
    {
      "epoch": 0.7074285714285714,
      "grad_norm": 0.8807134628295898,
      "learning_rate": 0.0005,
      "loss": 2.3837881088256836,
      "step": 3095
    },
    {
      "epoch": 0.7076571428571429,
      "grad_norm": 2.4542055130004883,
      "learning_rate": 0.0005,
      "loss": 2.4316799640655518,
      "step": 3096
    },
    {
      "epoch": 0.7078857142857143,
      "grad_norm": 0.9579233527183533,
      "learning_rate": 0.0005,
      "loss": 2.7521448135375977,
      "step": 3097
    },
    {
      "epoch": 0.7081142857142857,
      "grad_norm": 1.2856253385543823,
      "learning_rate": 0.0005,
      "loss": 2.855926036834717,
      "step": 3098
    },
    {
      "epoch": 0.7083428571428572,
      "grad_norm": 1.1742630004882812,
      "learning_rate": 0.0005,
      "loss": 2.665388584136963,
      "step": 3099
    },
    {
      "epoch": 0.7085714285714285,
      "grad_norm": 1.0938353538513184,
      "learning_rate": 0.0005,
      "loss": 2.772966146469116,
      "step": 3100
    },
    {
      "epoch": 0.7088,
      "grad_norm": 1.3174363374710083,
      "learning_rate": 0.0005,
      "loss": 3.3276727199554443,
      "step": 3101
    },
    {
      "epoch": 0.7090285714285715,
      "grad_norm": 1.262284517288208,
      "learning_rate": 0.0005,
      "loss": 2.5969972610473633,
      "step": 3102
    },
    {
      "epoch": 0.7092571428571428,
      "grad_norm": 1.0637216567993164,
      "learning_rate": 0.0005,
      "loss": 3.0604166984558105,
      "step": 3103
    },
    {
      "epoch": 0.7094857142857143,
      "grad_norm": 1.0801109075546265,
      "learning_rate": 0.0005,
      "loss": 2.4941253662109375,
      "step": 3104
    },
    {
      "epoch": 0.7097142857142857,
      "grad_norm": 1.118161916732788,
      "learning_rate": 0.0005,
      "loss": 2.7511067390441895,
      "step": 3105
    },
    {
      "epoch": 0.7099428571428571,
      "grad_norm": 1.0511454343795776,
      "learning_rate": 0.0005,
      "loss": 2.742560625076294,
      "step": 3106
    },
    {
      "epoch": 0.7101714285714286,
      "grad_norm": 1.2513481378555298,
      "learning_rate": 0.0005,
      "loss": 3.1831629276275635,
      "step": 3107
    },
    {
      "epoch": 0.7104,
      "grad_norm": 1.5711634159088135,
      "learning_rate": 0.0005,
      "loss": 2.922830104827881,
      "step": 3108
    },
    {
      "epoch": 0.7106285714285714,
      "grad_norm": 1.093149185180664,
      "learning_rate": 0.0005,
      "loss": 2.6469688415527344,
      "step": 3109
    },
    {
      "epoch": 0.7108571428571429,
      "grad_norm": 0.9019291400909424,
      "learning_rate": 0.0005,
      "loss": 2.4540061950683594,
      "step": 3110
    },
    {
      "epoch": 0.7110857142857143,
      "grad_norm": 0.9688547253608704,
      "learning_rate": 0.0005,
      "loss": 2.5792548656463623,
      "step": 3111
    },
    {
      "epoch": 0.7113142857142857,
      "grad_norm": 1.0261127948760986,
      "learning_rate": 0.0005,
      "loss": 2.514652729034424,
      "step": 3112
    },
    {
      "epoch": 0.7115428571428571,
      "grad_norm": 1.153064250946045,
      "learning_rate": 0.0005,
      "loss": 2.694981098175049,
      "step": 3113
    },
    {
      "epoch": 0.7117714285714286,
      "grad_norm": 1.086466908454895,
      "learning_rate": 0.0005,
      "loss": 2.5227646827697754,
      "step": 3114
    },
    {
      "epoch": 0.712,
      "grad_norm": 0.9846951961517334,
      "learning_rate": 0.0005,
      "loss": 2.3066911697387695,
      "step": 3115
    },
    {
      "epoch": 0.7122285714285714,
      "grad_norm": 1.0126867294311523,
      "learning_rate": 0.0005,
      "loss": 2.951955795288086,
      "step": 3116
    },
    {
      "epoch": 0.7124571428571429,
      "grad_norm": 1.0153822898864746,
      "learning_rate": 0.0005,
      "loss": 2.5232789516448975,
      "step": 3117
    },
    {
      "epoch": 0.7126857142857143,
      "grad_norm": 1.349198579788208,
      "learning_rate": 0.0005,
      "loss": 2.6335887908935547,
      "step": 3118
    },
    {
      "epoch": 0.7129142857142857,
      "grad_norm": 1.1658121347427368,
      "learning_rate": 0.0005,
      "loss": 2.4365127086639404,
      "step": 3119
    },
    {
      "epoch": 0.7131428571428572,
      "grad_norm": 1.225376844406128,
      "learning_rate": 0.0005,
      "loss": 3.0453364849090576,
      "step": 3120
    },
    {
      "epoch": 0.7133714285714285,
      "grad_norm": 1.3115241527557373,
      "learning_rate": 0.0005,
      "loss": 2.6902050971984863,
      "step": 3121
    },
    {
      "epoch": 0.7136,
      "grad_norm": 2.4093985557556152,
      "learning_rate": 0.0005,
      "loss": 2.5318922996520996,
      "step": 3122
    },
    {
      "epoch": 0.7138285714285715,
      "grad_norm": 1.0641001462936401,
      "learning_rate": 0.0005,
      "loss": 2.5151190757751465,
      "step": 3123
    },
    {
      "epoch": 0.7140571428571428,
      "grad_norm": 0.9564777612686157,
      "learning_rate": 0.0005,
      "loss": 2.9967308044433594,
      "step": 3124
    },
    {
      "epoch": 0.7142857142857143,
      "grad_norm": 0.9358117580413818,
      "learning_rate": 0.0005,
      "loss": 2.6489663124084473,
      "step": 3125
    },
    {
      "epoch": 0.7145142857142858,
      "grad_norm": 1.1270678043365479,
      "learning_rate": 0.0005,
      "loss": 2.5478639602661133,
      "step": 3126
    },
    {
      "epoch": 0.7147428571428571,
      "grad_norm": 0.918973445892334,
      "learning_rate": 0.0005,
      "loss": 2.506213426589966,
      "step": 3127
    },
    {
      "epoch": 0.7149714285714286,
      "grad_norm": 1.2052969932556152,
      "learning_rate": 0.0005,
      "loss": 2.67466139793396,
      "step": 3128
    },
    {
      "epoch": 0.7152,
      "grad_norm": 1.499547004699707,
      "learning_rate": 0.0005,
      "loss": 2.757944107055664,
      "step": 3129
    },
    {
      "epoch": 0.7154285714285714,
      "grad_norm": 1.1686278581619263,
      "learning_rate": 0.0005,
      "loss": 2.521237850189209,
      "step": 3130
    },
    {
      "epoch": 0.7156571428571429,
      "grad_norm": 1.109156847000122,
      "learning_rate": 0.0005,
      "loss": 2.884305953979492,
      "step": 3131
    },
    {
      "epoch": 0.7158857142857142,
      "grad_norm": 0.9701772928237915,
      "learning_rate": 0.0005,
      "loss": 2.5170352458953857,
      "step": 3132
    },
    {
      "epoch": 0.7161142857142857,
      "grad_norm": 1.3740193843841553,
      "learning_rate": 0.0005,
      "loss": 2.767526149749756,
      "step": 3133
    },
    {
      "epoch": 0.7163428571428572,
      "grad_norm": 1.105994701385498,
      "learning_rate": 0.0005,
      "loss": 3.0000345706939697,
      "step": 3134
    },
    {
      "epoch": 0.7165714285714285,
      "grad_norm": 1.0293575525283813,
      "learning_rate": 0.0005,
      "loss": 3.213104009628296,
      "step": 3135
    },
    {
      "epoch": 0.7168,
      "grad_norm": 1.3198822736740112,
      "learning_rate": 0.0005,
      "loss": 2.543774127960205,
      "step": 3136
    },
    {
      "epoch": 0.7170285714285715,
      "grad_norm": 1.0724507570266724,
      "learning_rate": 0.0005,
      "loss": 2.7147793769836426,
      "step": 3137
    },
    {
      "epoch": 0.7172571428571428,
      "grad_norm": 0.9429298639297485,
      "learning_rate": 0.0005,
      "loss": 2.5982353687286377,
      "step": 3138
    },
    {
      "epoch": 0.7174857142857143,
      "grad_norm": 1.4142601490020752,
      "learning_rate": 0.0005,
      "loss": 2.707367420196533,
      "step": 3139
    },
    {
      "epoch": 0.7177142857142857,
      "grad_norm": 1.1625850200653076,
      "learning_rate": 0.0005,
      "loss": 2.745265007019043,
      "step": 3140
    },
    {
      "epoch": 0.7179428571428571,
      "grad_norm": 0.9815985560417175,
      "learning_rate": 0.0005,
      "loss": 2.6423254013061523,
      "step": 3141
    },
    {
      "epoch": 0.7181714285714286,
      "grad_norm": 1.2368768453598022,
      "learning_rate": 0.0005,
      "loss": 2.5675785541534424,
      "step": 3142
    },
    {
      "epoch": 0.7184,
      "grad_norm": 2.2012948989868164,
      "learning_rate": 0.0005,
      "loss": 3.0975534915924072,
      "step": 3143
    },
    {
      "epoch": 0.7186285714285714,
      "grad_norm": 1.2949227094650269,
      "learning_rate": 0.0005,
      "loss": 2.823798179626465,
      "step": 3144
    },
    {
      "epoch": 0.7188571428571429,
      "grad_norm": 1.0995458364486694,
      "learning_rate": 0.0005,
      "loss": 2.8196096420288086,
      "step": 3145
    },
    {
      "epoch": 0.7190857142857143,
      "grad_norm": 0.9685551524162292,
      "learning_rate": 0.0005,
      "loss": 2.5365517139434814,
      "step": 3146
    },
    {
      "epoch": 0.7193142857142857,
      "grad_norm": 0.9720862507820129,
      "learning_rate": 0.0005,
      "loss": 2.6800074577331543,
      "step": 3147
    },
    {
      "epoch": 0.7195428571428572,
      "grad_norm": 0.9080411791801453,
      "learning_rate": 0.0005,
      "loss": 2.5979790687561035,
      "step": 3148
    },
    {
      "epoch": 0.7197714285714286,
      "grad_norm": 0.8975290656089783,
      "learning_rate": 0.0005,
      "loss": 2.952521800994873,
      "step": 3149
    },
    {
      "epoch": 0.72,
      "grad_norm": 1.0799779891967773,
      "learning_rate": 0.0005,
      "loss": 2.518254041671753,
      "step": 3150
    },
    {
      "epoch": 0.7202285714285714,
      "grad_norm": 1.235319972038269,
      "learning_rate": 0.0005,
      "loss": 2.9652342796325684,
      "step": 3151
    },
    {
      "epoch": 0.7204571428571429,
      "grad_norm": 0.8586491346359253,
      "learning_rate": 0.0005,
      "loss": 2.788663387298584,
      "step": 3152
    },
    {
      "epoch": 0.7206857142857143,
      "grad_norm": 0.9238666892051697,
      "learning_rate": 0.0005,
      "loss": 2.507131814956665,
      "step": 3153
    },
    {
      "epoch": 0.7209142857142857,
      "grad_norm": 1.1985585689544678,
      "learning_rate": 0.0005,
      "loss": 2.526618480682373,
      "step": 3154
    },
    {
      "epoch": 0.7211428571428572,
      "grad_norm": 1.4318288564682007,
      "learning_rate": 0.0005,
      "loss": 2.3965249061584473,
      "step": 3155
    },
    {
      "epoch": 0.7213714285714286,
      "grad_norm": 1.7806596755981445,
      "learning_rate": 0.0005,
      "loss": 2.530630350112915,
      "step": 3156
    },
    {
      "epoch": 0.7216,
      "grad_norm": 0.9599692821502686,
      "learning_rate": 0.0005,
      "loss": 2.676072120666504,
      "step": 3157
    },
    {
      "epoch": 0.7218285714285714,
      "grad_norm": 0.8745256662368774,
      "learning_rate": 0.0005,
      "loss": 2.669127941131592,
      "step": 3158
    },
    {
      "epoch": 0.7220571428571428,
      "grad_norm": 1.2634310722351074,
      "learning_rate": 0.0005,
      "loss": 2.302950620651245,
      "step": 3159
    },
    {
      "epoch": 0.7222857142857143,
      "grad_norm": 0.9290609955787659,
      "learning_rate": 0.0005,
      "loss": 2.6788322925567627,
      "step": 3160
    },
    {
      "epoch": 0.7225142857142857,
      "grad_norm": 0.9576400518417358,
      "learning_rate": 0.0005,
      "loss": 2.9544243812561035,
      "step": 3161
    },
    {
      "epoch": 0.7227428571428571,
      "grad_norm": 0.8347444534301758,
      "learning_rate": 0.0005,
      "loss": 2.5604400634765625,
      "step": 3162
    },
    {
      "epoch": 0.7229714285714286,
      "grad_norm": 1.2678744792938232,
      "learning_rate": 0.0005,
      "loss": 2.4674015045166016,
      "step": 3163
    },
    {
      "epoch": 0.7232,
      "grad_norm": 1.0336986780166626,
      "learning_rate": 0.0005,
      "loss": 2.8199238777160645,
      "step": 3164
    },
    {
      "epoch": 0.7234285714285714,
      "grad_norm": 1.1051037311553955,
      "learning_rate": 0.0005,
      "loss": 2.65210223197937,
      "step": 3165
    },
    {
      "epoch": 0.7236571428571429,
      "grad_norm": 1.188738465309143,
      "learning_rate": 0.0005,
      "loss": 2.4840402603149414,
      "step": 3166
    },
    {
      "epoch": 0.7238857142857142,
      "grad_norm": 7.5810394287109375,
      "learning_rate": 0.0005,
      "loss": 2.5652170181274414,
      "step": 3167
    },
    {
      "epoch": 0.7241142857142857,
      "grad_norm": 1.3711317777633667,
      "learning_rate": 0.0005,
      "loss": 2.363586902618408,
      "step": 3168
    },
    {
      "epoch": 0.7243428571428572,
      "grad_norm": 1.2206050157546997,
      "learning_rate": 0.0005,
      "loss": 2.8481011390686035,
      "step": 3169
    },
    {
      "epoch": 0.7245714285714285,
      "grad_norm": 1.1225011348724365,
      "learning_rate": 0.0005,
      "loss": 2.2362146377563477,
      "step": 3170
    },
    {
      "epoch": 0.7248,
      "grad_norm": 1.3610490560531616,
      "learning_rate": 0.0005,
      "loss": 2.6403729915618896,
      "step": 3171
    },
    {
      "epoch": 0.7250285714285715,
      "grad_norm": 0.9474913477897644,
      "learning_rate": 0.0005,
      "loss": 2.6627609729766846,
      "step": 3172
    },
    {
      "epoch": 0.7252571428571428,
      "grad_norm": 0.9736807346343994,
      "learning_rate": 0.0005,
      "loss": 2.5964863300323486,
      "step": 3173
    },
    {
      "epoch": 0.7254857142857143,
      "grad_norm": 1.2953132390975952,
      "learning_rate": 0.0005,
      "loss": 2.7966036796569824,
      "step": 3174
    },
    {
      "epoch": 0.7257142857142858,
      "grad_norm": 1.0741297006607056,
      "learning_rate": 0.0005,
      "loss": 2.5153205394744873,
      "step": 3175
    },
    {
      "epoch": 0.7259428571428571,
      "grad_norm": 1.0092099905014038,
      "learning_rate": 0.0005,
      "loss": 2.9028568267822266,
      "step": 3176
    },
    {
      "epoch": 0.7261714285714286,
      "grad_norm": 1.0172481536865234,
      "learning_rate": 0.0005,
      "loss": 2.7880301475524902,
      "step": 3177
    },
    {
      "epoch": 0.7264,
      "grad_norm": 1.1734594106674194,
      "learning_rate": 0.0005,
      "loss": 2.413945198059082,
      "step": 3178
    },
    {
      "epoch": 0.7266285714285714,
      "grad_norm": 0.9514299631118774,
      "learning_rate": 0.0005,
      "loss": 2.55880069732666,
      "step": 3179
    },
    {
      "epoch": 0.7268571428571429,
      "grad_norm": 0.9547292590141296,
      "learning_rate": 0.0005,
      "loss": 2.636629104614258,
      "step": 3180
    },
    {
      "epoch": 0.7270857142857143,
      "grad_norm": 1.0140061378479004,
      "learning_rate": 0.0005,
      "loss": 2.6786506175994873,
      "step": 3181
    },
    {
      "epoch": 0.7273142857142857,
      "grad_norm": 0.9590926170349121,
      "learning_rate": 0.0005,
      "loss": 2.8840928077697754,
      "step": 3182
    },
    {
      "epoch": 0.7275428571428572,
      "grad_norm": 0.9087859988212585,
      "learning_rate": 0.0005,
      "loss": 2.3056068420410156,
      "step": 3183
    },
    {
      "epoch": 0.7277714285714286,
      "grad_norm": 1.2817615270614624,
      "learning_rate": 0.0005,
      "loss": 2.8137316703796387,
      "step": 3184
    },
    {
      "epoch": 0.728,
      "grad_norm": 0.9628188014030457,
      "learning_rate": 0.0005,
      "loss": 2.551292896270752,
      "step": 3185
    },
    {
      "epoch": 0.7282285714285714,
      "grad_norm": 1.3467330932617188,
      "learning_rate": 0.0005,
      "loss": 2.5249528884887695,
      "step": 3186
    },
    {
      "epoch": 0.7284571428571428,
      "grad_norm": 1.002844214439392,
      "learning_rate": 0.0005,
      "loss": 2.423614978790283,
      "step": 3187
    },
    {
      "epoch": 0.7286857142857143,
      "grad_norm": 1.0327469110488892,
      "learning_rate": 0.0005,
      "loss": 2.6039328575134277,
      "step": 3188
    },
    {
      "epoch": 0.7289142857142857,
      "grad_norm": 1.0893335342407227,
      "learning_rate": 0.0005,
      "loss": 2.566831588745117,
      "step": 3189
    },
    {
      "epoch": 0.7291428571428571,
      "grad_norm": 0.9945144653320312,
      "learning_rate": 0.0005,
      "loss": 2.7840960025787354,
      "step": 3190
    },
    {
      "epoch": 0.7293714285714286,
      "grad_norm": 0.9706355929374695,
      "learning_rate": 0.0005,
      "loss": 2.7527263164520264,
      "step": 3191
    },
    {
      "epoch": 0.7296,
      "grad_norm": 14.401847839355469,
      "learning_rate": 0.0005,
      "loss": 2.4786133766174316,
      "step": 3192
    },
    {
      "epoch": 0.7298285714285714,
      "grad_norm": 1.0977438688278198,
      "learning_rate": 0.0005,
      "loss": 2.4768729209899902,
      "step": 3193
    },
    {
      "epoch": 0.7300571428571428,
      "grad_norm": 0.966454029083252,
      "learning_rate": 0.0005,
      "loss": 2.843441963195801,
      "step": 3194
    },
    {
      "epoch": 0.7302857142857143,
      "grad_norm": 0.9500324726104736,
      "learning_rate": 0.0005,
      "loss": 2.7516446113586426,
      "step": 3195
    },
    {
      "epoch": 0.7305142857142857,
      "grad_norm": 1.0156123638153076,
      "learning_rate": 0.0005,
      "loss": 2.6281495094299316,
      "step": 3196
    },
    {
      "epoch": 0.7307428571428571,
      "grad_norm": 0.8651865124702454,
      "learning_rate": 0.0005,
      "loss": 2.938154458999634,
      "step": 3197
    },
    {
      "epoch": 0.7309714285714286,
      "grad_norm": 0.9782639145851135,
      "learning_rate": 0.0005,
      "loss": 2.4339842796325684,
      "step": 3198
    },
    {
      "epoch": 0.7312,
      "grad_norm": 1.1713021993637085,
      "learning_rate": 0.0005,
      "loss": 2.621730089187622,
      "step": 3199
    },
    {
      "epoch": 0.7314285714285714,
      "grad_norm": 0.8657671213150024,
      "learning_rate": 0.0005,
      "loss": 2.301931858062744,
      "step": 3200
    },
    {
      "epoch": 0.7316571428571429,
      "grad_norm": 0.8780480623245239,
      "learning_rate": 0.0005,
      "loss": 2.8510358333587646,
      "step": 3201
    },
    {
      "epoch": 0.7318857142857143,
      "grad_norm": 1.1188924312591553,
      "learning_rate": 0.0005,
      "loss": 2.5575551986694336,
      "step": 3202
    },
    {
      "epoch": 0.7321142857142857,
      "grad_norm": 1.129006028175354,
      "learning_rate": 0.0005,
      "loss": 2.6137003898620605,
      "step": 3203
    },
    {
      "epoch": 0.7323428571428572,
      "grad_norm": 1.1324468851089478,
      "learning_rate": 0.0005,
      "loss": 2.701828956604004,
      "step": 3204
    },
    {
      "epoch": 0.7325714285714285,
      "grad_norm": 1.1030044555664062,
      "learning_rate": 0.0005,
      "loss": 3.056241035461426,
      "step": 3205
    },
    {
      "epoch": 0.7328,
      "grad_norm": 1.4738460779190063,
      "learning_rate": 0.0005,
      "loss": 2.4800119400024414,
      "step": 3206
    },
    {
      "epoch": 0.7330285714285715,
      "grad_norm": 1.2837975025177002,
      "learning_rate": 0.0005,
      "loss": 3.0179672241210938,
      "step": 3207
    },
    {
      "epoch": 0.7332571428571428,
      "grad_norm": 0.9803382754325867,
      "learning_rate": 0.0005,
      "loss": 2.5418710708618164,
      "step": 3208
    },
    {
      "epoch": 0.7334857142857143,
      "grad_norm": 0.9082301259040833,
      "learning_rate": 0.0005,
      "loss": 2.739387035369873,
      "step": 3209
    },
    {
      "epoch": 0.7337142857142858,
      "grad_norm": 1.0131304264068604,
      "learning_rate": 0.0005,
      "loss": 2.7272896766662598,
      "step": 3210
    },
    {
      "epoch": 0.7339428571428571,
      "grad_norm": 1.014064908027649,
      "learning_rate": 0.0005,
      "loss": 2.547926425933838,
      "step": 3211
    },
    {
      "epoch": 0.7341714285714286,
      "grad_norm": 1.0339009761810303,
      "learning_rate": 0.0005,
      "loss": 2.5361714363098145,
      "step": 3212
    },
    {
      "epoch": 0.7344,
      "grad_norm": 1.0866221189498901,
      "learning_rate": 0.0005,
      "loss": 2.5097203254699707,
      "step": 3213
    },
    {
      "epoch": 0.7346285714285714,
      "grad_norm": 1.0971990823745728,
      "learning_rate": 0.0005,
      "loss": 2.8583080768585205,
      "step": 3214
    },
    {
      "epoch": 0.7348571428571429,
      "grad_norm": 0.8280608654022217,
      "learning_rate": 0.0005,
      "loss": 2.3893699645996094,
      "step": 3215
    },
    {
      "epoch": 0.7350857142857142,
      "grad_norm": 0.8767127394676208,
      "learning_rate": 0.0005,
      "loss": 2.668494462966919,
      "step": 3216
    },
    {
      "epoch": 0.7353142857142857,
      "grad_norm": 1.1535698175430298,
      "learning_rate": 0.0005,
      "loss": 2.5746350288391113,
      "step": 3217
    },
    {
      "epoch": 0.7355428571428572,
      "grad_norm": 0.9180912971496582,
      "learning_rate": 0.0005,
      "loss": 2.446600914001465,
      "step": 3218
    },
    {
      "epoch": 0.7357714285714285,
      "grad_norm": 0.9868713617324829,
      "learning_rate": 0.0005,
      "loss": 2.525557279586792,
      "step": 3219
    },
    {
      "epoch": 0.736,
      "grad_norm": 0.9171088933944702,
      "learning_rate": 0.0005,
      "loss": 2.5216927528381348,
      "step": 3220
    },
    {
      "epoch": 0.7362285714285715,
      "grad_norm": 1.0168848037719727,
      "learning_rate": 0.0005,
      "loss": 2.750199317932129,
      "step": 3221
    },
    {
      "epoch": 0.7364571428571428,
      "grad_norm": 0.9519598484039307,
      "learning_rate": 0.0005,
      "loss": 2.746554374694824,
      "step": 3222
    },
    {
      "epoch": 0.7366857142857143,
      "grad_norm": 0.8987480401992798,
      "learning_rate": 0.0005,
      "loss": 2.635791778564453,
      "step": 3223
    },
    {
      "epoch": 0.7369142857142857,
      "grad_norm": 1.165555715560913,
      "learning_rate": 0.0005,
      "loss": 2.7588696479797363,
      "step": 3224
    },
    {
      "epoch": 0.7371428571428571,
      "grad_norm": 1.1803957223892212,
      "learning_rate": 0.0005,
      "loss": 2.6628730297088623,
      "step": 3225
    },
    {
      "epoch": 0.7373714285714286,
      "grad_norm": 1.136405348777771,
      "learning_rate": 0.0005,
      "loss": 2.6527299880981445,
      "step": 3226
    },
    {
      "epoch": 0.7376,
      "grad_norm": 1.3123023509979248,
      "learning_rate": 0.0005,
      "loss": 2.444774627685547,
      "step": 3227
    },
    {
      "epoch": 0.7378285714285714,
      "grad_norm": 1.069071888923645,
      "learning_rate": 0.0005,
      "loss": 3.042283535003662,
      "step": 3228
    },
    {
      "epoch": 0.7380571428571429,
      "grad_norm": 1.000360369682312,
      "learning_rate": 0.0005,
      "loss": 2.5023770332336426,
      "step": 3229
    },
    {
      "epoch": 0.7382857142857143,
      "grad_norm": 1.0603898763656616,
      "learning_rate": 0.0005,
      "loss": 2.9055910110473633,
      "step": 3230
    },
    {
      "epoch": 0.7385142857142857,
      "grad_norm": 1.0631928443908691,
      "learning_rate": 0.0005,
      "loss": 2.5291833877563477,
      "step": 3231
    },
    {
      "epoch": 0.7387428571428571,
      "grad_norm": 1.05446457862854,
      "learning_rate": 0.0005,
      "loss": 2.770555019378662,
      "step": 3232
    },
    {
      "epoch": 0.7389714285714286,
      "grad_norm": 1.0099992752075195,
      "learning_rate": 0.0005,
      "loss": 2.517070770263672,
      "step": 3233
    },
    {
      "epoch": 0.7392,
      "grad_norm": 1.0933879613876343,
      "learning_rate": 0.0005,
      "loss": 2.5011632442474365,
      "step": 3234
    },
    {
      "epoch": 0.7394285714285714,
      "grad_norm": 0.9749411344528198,
      "learning_rate": 0.0005,
      "loss": 2.8810842037200928,
      "step": 3235
    },
    {
      "epoch": 0.7396571428571429,
      "grad_norm": 1.0535091161727905,
      "learning_rate": 0.0005,
      "loss": 2.814072847366333,
      "step": 3236
    },
    {
      "epoch": 0.7398857142857143,
      "grad_norm": 0.9299401044845581,
      "learning_rate": 0.0005,
      "loss": 2.403225898742676,
      "step": 3237
    },
    {
      "epoch": 0.7401142857142857,
      "grad_norm": 0.9107871055603027,
      "learning_rate": 0.0005,
      "loss": 2.2954049110412598,
      "step": 3238
    },
    {
      "epoch": 0.7403428571428572,
      "grad_norm": 1.0832593441009521,
      "learning_rate": 0.0005,
      "loss": 2.6313273906707764,
      "step": 3239
    },
    {
      "epoch": 0.7405714285714285,
      "grad_norm": 1.0873947143554688,
      "learning_rate": 0.0005,
      "loss": 2.7271153926849365,
      "step": 3240
    },
    {
      "epoch": 0.7408,
      "grad_norm": 1.171940803527832,
      "learning_rate": 0.0005,
      "loss": 2.762080192565918,
      "step": 3241
    },
    {
      "epoch": 0.7410285714285715,
      "grad_norm": 1.2538663148880005,
      "learning_rate": 0.0005,
      "loss": 2.4101805686950684,
      "step": 3242
    },
    {
      "epoch": 0.7412571428571428,
      "grad_norm": 1.148500680923462,
      "learning_rate": 0.0005,
      "loss": 2.717398166656494,
      "step": 3243
    },
    {
      "epoch": 0.7414857142857143,
      "grad_norm": 1.3500258922576904,
      "learning_rate": 0.0005,
      "loss": 2.207852840423584,
      "step": 3244
    },
    {
      "epoch": 0.7417142857142857,
      "grad_norm": 0.9361358880996704,
      "learning_rate": 0.0005,
      "loss": 2.6542201042175293,
      "step": 3245
    },
    {
      "epoch": 0.7419428571428571,
      "grad_norm": 0.8729842901229858,
      "learning_rate": 0.0005,
      "loss": 2.376053810119629,
      "step": 3246
    },
    {
      "epoch": 0.7421714285714286,
      "grad_norm": 0.8898550271987915,
      "learning_rate": 0.0005,
      "loss": 2.6270713806152344,
      "step": 3247
    },
    {
      "epoch": 0.7424,
      "grad_norm": 0.8303256034851074,
      "learning_rate": 0.0005,
      "loss": 2.8432273864746094,
      "step": 3248
    },
    {
      "epoch": 0.7426285714285714,
      "grad_norm": 0.9427400827407837,
      "learning_rate": 0.0005,
      "loss": 2.3577749729156494,
      "step": 3249
    },
    {
      "epoch": 0.7428571428571429,
      "grad_norm": 0.9862282872200012,
      "learning_rate": 0.0005,
      "loss": 2.7315621376037598,
      "step": 3250
    },
    {
      "epoch": 0.7430857142857142,
      "grad_norm": 1.2026194334030151,
      "learning_rate": 0.0005,
      "loss": 2.583444833755493,
      "step": 3251
    },
    {
      "epoch": 0.7433142857142857,
      "grad_norm": 1.2627228498458862,
      "learning_rate": 0.0005,
      "loss": 2.8469414710998535,
      "step": 3252
    },
    {
      "epoch": 0.7435428571428572,
      "grad_norm": 0.9257164597511292,
      "learning_rate": 0.0005,
      "loss": 2.6901607513427734,
      "step": 3253
    },
    {
      "epoch": 0.7437714285714285,
      "grad_norm": 0.7530515789985657,
      "learning_rate": 0.0005,
      "loss": 2.7365899085998535,
      "step": 3254
    },
    {
      "epoch": 0.744,
      "grad_norm": 1.1915996074676514,
      "learning_rate": 0.0005,
      "loss": 2.6745693683624268,
      "step": 3255
    },
    {
      "epoch": 0.7442285714285715,
      "grad_norm": 1.1296669244766235,
      "learning_rate": 0.0005,
      "loss": 2.676387071609497,
      "step": 3256
    },
    {
      "epoch": 0.7444571428571428,
      "grad_norm": 0.8727503418922424,
      "learning_rate": 0.0005,
      "loss": 2.292778253555298,
      "step": 3257
    },
    {
      "epoch": 0.7446857142857143,
      "grad_norm": 0.867588222026825,
      "learning_rate": 0.0005,
      "loss": 2.8078927993774414,
      "step": 3258
    },
    {
      "epoch": 0.7449142857142858,
      "grad_norm": 0.8194257616996765,
      "learning_rate": 0.0005,
      "loss": 2.384207248687744,
      "step": 3259
    },
    {
      "epoch": 0.7451428571428571,
      "grad_norm": 1.2630435228347778,
      "learning_rate": 0.0005,
      "loss": 2.6532983779907227,
      "step": 3260
    },
    {
      "epoch": 0.7453714285714286,
      "grad_norm": 0.9258761405944824,
      "learning_rate": 0.0005,
      "loss": 2.3819775581359863,
      "step": 3261
    },
    {
      "epoch": 0.7456,
      "grad_norm": 1.1843903064727783,
      "learning_rate": 0.0005,
      "loss": 2.8823869228363037,
      "step": 3262
    },
    {
      "epoch": 0.7458285714285714,
      "grad_norm": 1.0554161071777344,
      "learning_rate": 0.0005,
      "loss": 2.518850326538086,
      "step": 3263
    },
    {
      "epoch": 0.7460571428571429,
      "grad_norm": 1.074902057647705,
      "learning_rate": 0.0005,
      "loss": 2.463817596435547,
      "step": 3264
    },
    {
      "epoch": 0.7462857142857143,
      "grad_norm": 1.1966898441314697,
      "learning_rate": 0.0005,
      "loss": 2.7594144344329834,
      "step": 3265
    },
    {
      "epoch": 0.7465142857142857,
      "grad_norm": 0.7888708114624023,
      "learning_rate": 0.0005,
      "loss": 2.2742748260498047,
      "step": 3266
    },
    {
      "epoch": 0.7467428571428572,
      "grad_norm": 1.1241718530654907,
      "learning_rate": 0.0005,
      "loss": 2.4328737258911133,
      "step": 3267
    },
    {
      "epoch": 0.7469714285714286,
      "grad_norm": 0.8343415856361389,
      "learning_rate": 0.0005,
      "loss": 2.3739919662475586,
      "step": 3268
    },
    {
      "epoch": 0.7472,
      "grad_norm": 0.8571866750717163,
      "learning_rate": 0.0005,
      "loss": 2.64678692817688,
      "step": 3269
    },
    {
      "epoch": 0.7474285714285714,
      "grad_norm": 0.9262595176696777,
      "learning_rate": 0.0005,
      "loss": 2.4975578784942627,
      "step": 3270
    },
    {
      "epoch": 0.7476571428571429,
      "grad_norm": 0.9086957573890686,
      "learning_rate": 0.0005,
      "loss": 2.650179862976074,
      "step": 3271
    },
    {
      "epoch": 0.7478857142857143,
      "grad_norm": 0.8719368577003479,
      "learning_rate": 0.0005,
      "loss": 2.283245801925659,
      "step": 3272
    },
    {
      "epoch": 0.7481142857142857,
      "grad_norm": 1.318085789680481,
      "learning_rate": 0.0005,
      "loss": 2.7498080730438232,
      "step": 3273
    },
    {
      "epoch": 0.7483428571428571,
      "grad_norm": 0.8838558197021484,
      "learning_rate": 0.0005,
      "loss": 2.546095371246338,
      "step": 3274
    },
    {
      "epoch": 0.7485714285714286,
      "grad_norm": 1.1075750589370728,
      "learning_rate": 0.0005,
      "loss": 2.610161781311035,
      "step": 3275
    },
    {
      "epoch": 0.7488,
      "grad_norm": 0.9923381209373474,
      "learning_rate": 0.0005,
      "loss": 2.2981581687927246,
      "step": 3276
    },
    {
      "epoch": 0.7490285714285714,
      "grad_norm": 0.9819334149360657,
      "learning_rate": 0.0005,
      "loss": 2.4720089435577393,
      "step": 3277
    },
    {
      "epoch": 0.7492571428571428,
      "grad_norm": 1.3856918811798096,
      "learning_rate": 0.0005,
      "loss": 2.4493565559387207,
      "step": 3278
    },
    {
      "epoch": 0.7494857142857143,
      "grad_norm": 0.7386382222175598,
      "learning_rate": 0.0005,
      "loss": 2.4361860752105713,
      "step": 3279
    },
    {
      "epoch": 0.7497142857142857,
      "grad_norm": 1.0708332061767578,
      "learning_rate": 0.0005,
      "loss": 2.4904532432556152,
      "step": 3280
    },
    {
      "epoch": 0.7499428571428571,
      "grad_norm": 1.224338173866272,
      "learning_rate": 0.0005,
      "loss": 2.6944124698638916,
      "step": 3281
    },
    {
      "epoch": 0.7501714285714286,
      "grad_norm": 0.9807581305503845,
      "learning_rate": 0.0005,
      "loss": 2.353611707687378,
      "step": 3282
    },
    {
      "epoch": 0.7504,
      "grad_norm": 0.9606974720954895,
      "learning_rate": 0.0005,
      "loss": 2.705677032470703,
      "step": 3283
    },
    {
      "epoch": 0.7506285714285714,
      "grad_norm": 0.9067217111587524,
      "learning_rate": 0.0005,
      "loss": 2.790090560913086,
      "step": 3284
    },
    {
      "epoch": 0.7508571428571429,
      "grad_norm": 1.0553563833236694,
      "learning_rate": 0.0005,
      "loss": 2.6900582313537598,
      "step": 3285
    },
    {
      "epoch": 0.7510857142857142,
      "grad_norm": 0.9563918113708496,
      "learning_rate": 0.0005,
      "loss": 2.6751255989074707,
      "step": 3286
    },
    {
      "epoch": 0.7513142857142857,
      "grad_norm": 0.9582123160362244,
      "learning_rate": 0.0005,
      "loss": 2.634918689727783,
      "step": 3287
    },
    {
      "epoch": 0.7515428571428572,
      "grad_norm": 1.0734041929244995,
      "learning_rate": 0.0005,
      "loss": 2.7479076385498047,
      "step": 3288
    },
    {
      "epoch": 0.7517714285714285,
      "grad_norm": 0.9730418920516968,
      "learning_rate": 0.0005,
      "loss": 2.83658504486084,
      "step": 3289
    },
    {
      "epoch": 0.752,
      "grad_norm": 1.0583702325820923,
      "learning_rate": 0.0005,
      "loss": 2.6827101707458496,
      "step": 3290
    },
    {
      "epoch": 0.7522285714285715,
      "grad_norm": 1.0708080530166626,
      "learning_rate": 0.0005,
      "loss": 2.1554431915283203,
      "step": 3291
    },
    {
      "epoch": 0.7524571428571428,
      "grad_norm": 0.8171729445457458,
      "learning_rate": 0.0005,
      "loss": 2.550048351287842,
      "step": 3292
    },
    {
      "epoch": 0.7526857142857143,
      "grad_norm": 0.8684660196304321,
      "learning_rate": 0.0005,
      "loss": 2.6475226879119873,
      "step": 3293
    },
    {
      "epoch": 0.7529142857142858,
      "grad_norm": 1.0507606267929077,
      "learning_rate": 0.0005,
      "loss": 2.264495849609375,
      "step": 3294
    },
    {
      "epoch": 0.7531428571428571,
      "grad_norm": 1.1157907247543335,
      "learning_rate": 0.0005,
      "loss": 2.601646900177002,
      "step": 3295
    },
    {
      "epoch": 0.7533714285714286,
      "grad_norm": 1.2778984308242798,
      "learning_rate": 0.0005,
      "loss": 2.823664903640747,
      "step": 3296
    },
    {
      "epoch": 0.7536,
      "grad_norm": 0.8865649104118347,
      "learning_rate": 0.0005,
      "loss": 2.885889768600464,
      "step": 3297
    },
    {
      "epoch": 0.7538285714285714,
      "grad_norm": 0.9584751725196838,
      "learning_rate": 0.0005,
      "loss": 2.485468864440918,
      "step": 3298
    },
    {
      "epoch": 0.7540571428571429,
      "grad_norm": 1.0218071937561035,
      "learning_rate": 0.0005,
      "loss": 2.6104788780212402,
      "step": 3299
    },
    {
      "epoch": 0.7542857142857143,
      "grad_norm": 1.0950052738189697,
      "learning_rate": 0.0005,
      "loss": 3.0129778385162354,
      "step": 3300
    },
    {
      "epoch": 0.7545142857142857,
      "grad_norm": 1.119939923286438,
      "learning_rate": 0.0005,
      "loss": 3.0209767818450928,
      "step": 3301
    },
    {
      "epoch": 0.7547428571428572,
      "grad_norm": 1.0613102912902832,
      "learning_rate": 0.0005,
      "loss": 2.4120092391967773,
      "step": 3302
    },
    {
      "epoch": 0.7549714285714286,
      "grad_norm": 0.8681820034980774,
      "learning_rate": 0.0005,
      "loss": 2.4325971603393555,
      "step": 3303
    },
    {
      "epoch": 0.7552,
      "grad_norm": 1.267486572265625,
      "learning_rate": 0.0005,
      "loss": 2.446160316467285,
      "step": 3304
    },
    {
      "epoch": 0.7554285714285714,
      "grad_norm": 1.6867809295654297,
      "learning_rate": 0.0005,
      "loss": 2.569206476211548,
      "step": 3305
    },
    {
      "epoch": 0.7556571428571428,
      "grad_norm": 0.980036199092865,
      "learning_rate": 0.0005,
      "loss": 2.5036067962646484,
      "step": 3306
    },
    {
      "epoch": 0.7558857142857143,
      "grad_norm": 0.8857386708259583,
      "learning_rate": 0.0005,
      "loss": 2.434992790222168,
      "step": 3307
    },
    {
      "epoch": 0.7561142857142857,
      "grad_norm": 0.9757887721061707,
      "learning_rate": 0.0005,
      "loss": 2.6444389820098877,
      "step": 3308
    },
    {
      "epoch": 0.7563428571428571,
      "grad_norm": 1.1554207801818848,
      "learning_rate": 0.0005,
      "loss": 2.5749545097351074,
      "step": 3309
    },
    {
      "epoch": 0.7565714285714286,
      "grad_norm": 1.2820571660995483,
      "learning_rate": 0.0005,
      "loss": 2.787961959838867,
      "step": 3310
    },
    {
      "epoch": 0.7568,
      "grad_norm": 1.2380133867263794,
      "learning_rate": 0.0005,
      "loss": 2.4599862098693848,
      "step": 3311
    },
    {
      "epoch": 0.7570285714285714,
      "grad_norm": 0.9414105415344238,
      "learning_rate": 0.0005,
      "loss": 2.6963553428649902,
      "step": 3312
    },
    {
      "epoch": 0.7572571428571429,
      "grad_norm": 1.1859304904937744,
      "learning_rate": 0.0005,
      "loss": 2.5848581790924072,
      "step": 3313
    },
    {
      "epoch": 0.7574857142857143,
      "grad_norm": 0.9595261812210083,
      "learning_rate": 0.0005,
      "loss": 2.5442748069763184,
      "step": 3314
    },
    {
      "epoch": 0.7577142857142857,
      "grad_norm": 0.9201557040214539,
      "learning_rate": 0.0005,
      "loss": 2.947746753692627,
      "step": 3315
    },
    {
      "epoch": 0.7579428571428571,
      "grad_norm": 0.9259023070335388,
      "learning_rate": 0.0005,
      "loss": 2.805039882659912,
      "step": 3316
    },
    {
      "epoch": 0.7581714285714286,
      "grad_norm": 0.9017371535301208,
      "learning_rate": 0.0005,
      "loss": 2.646221399307251,
      "step": 3317
    },
    {
      "epoch": 0.7584,
      "grad_norm": 0.9873809814453125,
      "learning_rate": 0.0005,
      "loss": 2.6708431243896484,
      "step": 3318
    },
    {
      "epoch": 0.7586285714285714,
      "grad_norm": 0.928821861743927,
      "learning_rate": 0.0005,
      "loss": 2.510650157928467,
      "step": 3319
    },
    {
      "epoch": 0.7588571428571429,
      "grad_norm": 0.8688258528709412,
      "learning_rate": 0.0005,
      "loss": 2.7537283897399902,
      "step": 3320
    },
    {
      "epoch": 0.7590857142857143,
      "grad_norm": 0.8935152888298035,
      "learning_rate": 0.0005,
      "loss": 2.266035556793213,
      "step": 3321
    },
    {
      "epoch": 0.7593142857142857,
      "grad_norm": 0.8762940168380737,
      "learning_rate": 0.0005,
      "loss": 2.5960166454315186,
      "step": 3322
    },
    {
      "epoch": 0.7595428571428572,
      "grad_norm": 1.0388795137405396,
      "learning_rate": 0.0005,
      "loss": 2.6718878746032715,
      "step": 3323
    },
    {
      "epoch": 0.7597714285714285,
      "grad_norm": 1.6972371339797974,
      "learning_rate": 0.0005,
      "loss": 2.7728452682495117,
      "step": 3324
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.8454287052154541,
      "learning_rate": 0.0005,
      "loss": 2.852518081665039,
      "step": 3325
    },
    {
      "epoch": 0.7602285714285715,
      "grad_norm": 0.7640851736068726,
      "learning_rate": 0.0005,
      "loss": 2.52463960647583,
      "step": 3326
    },
    {
      "epoch": 0.7604571428571428,
      "grad_norm": 1.1308248043060303,
      "learning_rate": 0.0005,
      "loss": 2.260491371154785,
      "step": 3327
    },
    {
      "epoch": 0.7606857142857143,
      "grad_norm": 1.235170602798462,
      "learning_rate": 0.0005,
      "loss": 2.507427453994751,
      "step": 3328
    },
    {
      "epoch": 0.7609142857142858,
      "grad_norm": 0.9385919570922852,
      "learning_rate": 0.0005,
      "loss": 2.5176687240600586,
      "step": 3329
    },
    {
      "epoch": 0.7611428571428571,
      "grad_norm": 1.1187645196914673,
      "learning_rate": 0.0005,
      "loss": 2.6946685314178467,
      "step": 3330
    },
    {
      "epoch": 0.7613714285714286,
      "grad_norm": 0.9455365538597107,
      "learning_rate": 0.0005,
      "loss": 2.704591751098633,
      "step": 3331
    },
    {
      "epoch": 0.7616,
      "grad_norm": 1.1438757181167603,
      "learning_rate": 0.0005,
      "loss": 2.391693115234375,
      "step": 3332
    },
    {
      "epoch": 0.7618285714285714,
      "grad_norm": 0.8643271327018738,
      "learning_rate": 0.0005,
      "loss": 2.3163328170776367,
      "step": 3333
    },
    {
      "epoch": 0.7620571428571429,
      "grad_norm": 0.9830561876296997,
      "learning_rate": 0.0005,
      "loss": 2.588460922241211,
      "step": 3334
    },
    {
      "epoch": 0.7622857142857142,
      "grad_norm": 1.0159375667572021,
      "learning_rate": 0.0005,
      "loss": 2.4640450477600098,
      "step": 3335
    },
    {
      "epoch": 0.7625142857142857,
      "grad_norm": 0.877434492111206,
      "learning_rate": 0.0005,
      "loss": 2.2699427604675293,
      "step": 3336
    },
    {
      "epoch": 0.7627428571428572,
      "grad_norm": 1.0287258625030518,
      "learning_rate": 0.0005,
      "loss": 2.933094024658203,
      "step": 3337
    },
    {
      "epoch": 0.7629714285714285,
      "grad_norm": 1.0653072595596313,
      "learning_rate": 0.0005,
      "loss": 2.9960296154022217,
      "step": 3338
    },
    {
      "epoch": 0.7632,
      "grad_norm": 0.9876996874809265,
      "learning_rate": 0.0005,
      "loss": 2.616177797317505,
      "step": 3339
    },
    {
      "epoch": 0.7634285714285715,
      "grad_norm": 0.9731858968734741,
      "learning_rate": 0.0005,
      "loss": 2.662153720855713,
      "step": 3340
    },
    {
      "epoch": 0.7636571428571428,
      "grad_norm": 0.8066849112510681,
      "learning_rate": 0.0005,
      "loss": 2.674673318862915,
      "step": 3341
    },
    {
      "epoch": 0.7638857142857143,
      "grad_norm": 0.9457006454467773,
      "learning_rate": 0.0005,
      "loss": 2.752453565597534,
      "step": 3342
    },
    {
      "epoch": 0.7641142857142857,
      "grad_norm": 0.8284391760826111,
      "learning_rate": 0.0005,
      "loss": 2.9396629333496094,
      "step": 3343
    },
    {
      "epoch": 0.7643428571428571,
      "grad_norm": 0.8833647966384888,
      "learning_rate": 0.0005,
      "loss": 2.3079891204833984,
      "step": 3344
    },
    {
      "epoch": 0.7645714285714286,
      "grad_norm": 0.9292017817497253,
      "learning_rate": 0.0005,
      "loss": 2.5243194103240967,
      "step": 3345
    },
    {
      "epoch": 0.7648,
      "grad_norm": 1.0876922607421875,
      "learning_rate": 0.0005,
      "loss": 2.5919761657714844,
      "step": 3346
    },
    {
      "epoch": 0.7650285714285714,
      "grad_norm": 1.0131901502609253,
      "learning_rate": 0.0005,
      "loss": 2.665390968322754,
      "step": 3347
    },
    {
      "epoch": 0.7652571428571429,
      "grad_norm": 0.8501442670822144,
      "learning_rate": 0.0005,
      "loss": 2.637007236480713,
      "step": 3348
    },
    {
      "epoch": 0.7654857142857143,
      "grad_norm": 0.8534291386604309,
      "learning_rate": 0.0005,
      "loss": 2.444477081298828,
      "step": 3349
    },
    {
      "epoch": 0.7657142857142857,
      "grad_norm": 0.8584223389625549,
      "learning_rate": 0.0005,
      "loss": 2.7890000343322754,
      "step": 3350
    },
    {
      "epoch": 0.7659428571428571,
      "grad_norm": 1.0396770238876343,
      "learning_rate": 0.0005,
      "loss": 2.651972770690918,
      "step": 3351
    },
    {
      "epoch": 0.7661714285714286,
      "grad_norm": 1.2438801527023315,
      "learning_rate": 0.0005,
      "loss": 2.6188764572143555,
      "step": 3352
    },
    {
      "epoch": 0.7664,
      "grad_norm": 1.4405186176300049,
      "learning_rate": 0.0005,
      "loss": 2.0316665172576904,
      "step": 3353
    },
    {
      "epoch": 0.7666285714285714,
      "grad_norm": 1.069464087486267,
      "learning_rate": 0.0005,
      "loss": 2.6184017658233643,
      "step": 3354
    },
    {
      "epoch": 0.7668571428571429,
      "grad_norm": 1.3306028842926025,
      "learning_rate": 0.0005,
      "loss": 2.5980305671691895,
      "step": 3355
    },
    {
      "epoch": 0.7670857142857143,
      "grad_norm": 1.1725167036056519,
      "learning_rate": 0.0005,
      "loss": 2.407914161682129,
      "step": 3356
    },
    {
      "epoch": 0.7673142857142857,
      "grad_norm": 0.9133893847465515,
      "learning_rate": 0.0005,
      "loss": 2.260901927947998,
      "step": 3357
    },
    {
      "epoch": 0.7675428571428572,
      "grad_norm": 0.9508811235427856,
      "learning_rate": 0.0005,
      "loss": 2.7164862155914307,
      "step": 3358
    },
    {
      "epoch": 0.7677714285714285,
      "grad_norm": 1.0750768184661865,
      "learning_rate": 0.0005,
      "loss": 2.582045078277588,
      "step": 3359
    },
    {
      "epoch": 0.768,
      "grad_norm": 1.1337265968322754,
      "learning_rate": 0.0005,
      "loss": 2.9282898902893066,
      "step": 3360
    },
    {
      "epoch": 0.7682285714285715,
      "grad_norm": 1.2693594694137573,
      "learning_rate": 0.0005,
      "loss": 2.4490737915039062,
      "step": 3361
    },
    {
      "epoch": 0.7684571428571428,
      "grad_norm": 1.1848417520523071,
      "learning_rate": 0.0005,
      "loss": 2.6116883754730225,
      "step": 3362
    },
    {
      "epoch": 0.7686857142857143,
      "grad_norm": 0.9453176856040955,
      "learning_rate": 0.0005,
      "loss": 2.700134038925171,
      "step": 3363
    },
    {
      "epoch": 0.7689142857142857,
      "grad_norm": 0.8957961201667786,
      "learning_rate": 0.0005,
      "loss": 2.3697235584259033,
      "step": 3364
    },
    {
      "epoch": 0.7691428571428571,
      "grad_norm": 0.9336755871772766,
      "learning_rate": 0.0005,
      "loss": 2.690866231918335,
      "step": 3365
    },
    {
      "epoch": 0.7693714285714286,
      "grad_norm": 0.9351866245269775,
      "learning_rate": 0.0005,
      "loss": 2.68973970413208,
      "step": 3366
    },
    {
      "epoch": 0.7696,
      "grad_norm": 1.246477723121643,
      "learning_rate": 0.0005,
      "loss": 2.647789239883423,
      "step": 3367
    },
    {
      "epoch": 0.7698285714285714,
      "grad_norm": 1.1127002239227295,
      "learning_rate": 0.0005,
      "loss": 2.499415397644043,
      "step": 3368
    },
    {
      "epoch": 0.7700571428571429,
      "grad_norm": 0.8386632800102234,
      "learning_rate": 0.0005,
      "loss": 2.4095513820648193,
      "step": 3369
    },
    {
      "epoch": 0.7702857142857142,
      "grad_norm": 1.0416630506515503,
      "learning_rate": 0.0005,
      "loss": 2.73007869720459,
      "step": 3370
    },
    {
      "epoch": 0.7705142857142857,
      "grad_norm": 1.0573363304138184,
      "learning_rate": 0.0005,
      "loss": 2.714855194091797,
      "step": 3371
    },
    {
      "epoch": 0.7707428571428572,
      "grad_norm": 1.1380459070205688,
      "learning_rate": 0.0005,
      "loss": 2.5787434577941895,
      "step": 3372
    },
    {
      "epoch": 0.7709714285714285,
      "grad_norm": 0.9468693733215332,
      "learning_rate": 0.0005,
      "loss": 2.951738119125366,
      "step": 3373
    },
    {
      "epoch": 0.7712,
      "grad_norm": 1.0712273120880127,
      "learning_rate": 0.0005,
      "loss": 2.4618091583251953,
      "step": 3374
    },
    {
      "epoch": 0.7714285714285715,
      "grad_norm": 0.9430662393569946,
      "learning_rate": 0.0005,
      "loss": 2.427151918411255,
      "step": 3375
    },
    {
      "epoch": 0.7716571428571428,
      "grad_norm": 1.234299659729004,
      "learning_rate": 0.0005,
      "loss": 2.574103593826294,
      "step": 3376
    },
    {
      "epoch": 0.7718857142857143,
      "grad_norm": 1.7269846200942993,
      "learning_rate": 0.0005,
      "loss": 2.640848159790039,
      "step": 3377
    },
    {
      "epoch": 0.7721142857142858,
      "grad_norm": 1.797945261001587,
      "learning_rate": 0.0005,
      "loss": 2.8725967407226562,
      "step": 3378
    },
    {
      "epoch": 0.7723428571428571,
      "grad_norm": 0.9883133172988892,
      "learning_rate": 0.0005,
      "loss": 2.4199612140655518,
      "step": 3379
    },
    {
      "epoch": 0.7725714285714286,
      "grad_norm": 0.9313606023788452,
      "learning_rate": 0.0005,
      "loss": 2.2660491466522217,
      "step": 3380
    },
    {
      "epoch": 0.7728,
      "grad_norm": 0.9930185079574585,
      "learning_rate": 0.0005,
      "loss": 2.473541259765625,
      "step": 3381
    },
    {
      "epoch": 0.7730285714285714,
      "grad_norm": 1.0760306119918823,
      "learning_rate": 0.0005,
      "loss": 2.196193218231201,
      "step": 3382
    },
    {
      "epoch": 0.7732571428571429,
      "grad_norm": 0.8226915001869202,
      "learning_rate": 0.0005,
      "loss": 2.6165285110473633,
      "step": 3383
    },
    {
      "epoch": 0.7734857142857143,
      "grad_norm": 0.9821994304656982,
      "learning_rate": 0.0005,
      "loss": 3.192620277404785,
      "step": 3384
    },
    {
      "epoch": 0.7737142857142857,
      "grad_norm": 0.9969742894172668,
      "learning_rate": 0.0005,
      "loss": 2.6757211685180664,
      "step": 3385
    },
    {
      "epoch": 0.7739428571428572,
      "grad_norm": 0.9806506037712097,
      "learning_rate": 0.0005,
      "loss": 2.329775810241699,
      "step": 3386
    },
    {
      "epoch": 0.7741714285714286,
      "grad_norm": 1.0003595352172852,
      "learning_rate": 0.0005,
      "loss": 2.593360185623169,
      "step": 3387
    },
    {
      "epoch": 0.7744,
      "grad_norm": 0.8433234691619873,
      "learning_rate": 0.0005,
      "loss": 2.627596855163574,
      "step": 3388
    },
    {
      "epoch": 0.7746285714285714,
      "grad_norm": 1.045478343963623,
      "learning_rate": 0.0005,
      "loss": 2.5134031772613525,
      "step": 3389
    },
    {
      "epoch": 0.7748571428571429,
      "grad_norm": 1.0123597383499146,
      "learning_rate": 0.0005,
      "loss": 2.548252582550049,
      "step": 3390
    },
    {
      "epoch": 0.7750857142857143,
      "grad_norm": 0.9077357649803162,
      "learning_rate": 0.0005,
      "loss": 2.360924243927002,
      "step": 3391
    },
    {
      "epoch": 0.7753142857142857,
      "grad_norm": 0.9298978447914124,
      "learning_rate": 0.0005,
      "loss": 2.791355848312378,
      "step": 3392
    },
    {
      "epoch": 0.7755428571428571,
      "grad_norm": 0.9233838319778442,
      "learning_rate": 0.0005,
      "loss": 2.655970573425293,
      "step": 3393
    },
    {
      "epoch": 0.7757714285714286,
      "grad_norm": 0.9169749617576599,
      "learning_rate": 0.0005,
      "loss": 2.594390392303467,
      "step": 3394
    },
    {
      "epoch": 0.776,
      "grad_norm": 0.8584573268890381,
      "learning_rate": 0.0005,
      "loss": 2.9621129035949707,
      "step": 3395
    },
    {
      "epoch": 0.7762285714285714,
      "grad_norm": 1.3054392337799072,
      "learning_rate": 0.0005,
      "loss": 2.6692099571228027,
      "step": 3396
    },
    {
      "epoch": 0.7764571428571428,
      "grad_norm": 1.0542711019515991,
      "learning_rate": 0.0005,
      "loss": 2.096263885498047,
      "step": 3397
    },
    {
      "epoch": 0.7766857142857143,
      "grad_norm": 1.18557870388031,
      "learning_rate": 0.0005,
      "loss": 2.6555614471435547,
      "step": 3398
    },
    {
      "epoch": 0.7769142857142857,
      "grad_norm": 1.9743332862854004,
      "learning_rate": 0.0005,
      "loss": 2.7927567958831787,
      "step": 3399
    },
    {
      "epoch": 0.7771428571428571,
      "grad_norm": 0.9077439904212952,
      "learning_rate": 0.0005,
      "loss": 2.7504074573516846,
      "step": 3400
    },
    {
      "epoch": 0.7773714285714286,
      "grad_norm": 1.0270848274230957,
      "learning_rate": 0.0005,
      "loss": 2.5692174434661865,
      "step": 3401
    },
    {
      "epoch": 0.7776,
      "grad_norm": 1.2058745622634888,
      "learning_rate": 0.0005,
      "loss": 2.8263018131256104,
      "step": 3402
    },
    {
      "epoch": 0.7778285714285714,
      "grad_norm": 0.9339042901992798,
      "learning_rate": 0.0005,
      "loss": 2.4782724380493164,
      "step": 3403
    },
    {
      "epoch": 0.7780571428571429,
      "grad_norm": 0.9201182126998901,
      "learning_rate": 0.0005,
      "loss": 2.187889575958252,
      "step": 3404
    },
    {
      "epoch": 0.7782857142857142,
      "grad_norm": 0.917798638343811,
      "learning_rate": 0.0005,
      "loss": 2.5829837322235107,
      "step": 3405
    },
    {
      "epoch": 0.7785142857142857,
      "grad_norm": 0.8531247973442078,
      "learning_rate": 0.0005,
      "loss": 2.571739673614502,
      "step": 3406
    },
    {
      "epoch": 0.7787428571428572,
      "grad_norm": 0.9995288252830505,
      "learning_rate": 0.0005,
      "loss": 2.609532117843628,
      "step": 3407
    },
    {
      "epoch": 0.7789714285714285,
      "grad_norm": 0.9471408128738403,
      "learning_rate": 0.0005,
      "loss": 2.434926748275757,
      "step": 3408
    },
    {
      "epoch": 0.7792,
      "grad_norm": 1.047118067741394,
      "learning_rate": 0.0005,
      "loss": 2.331191062927246,
      "step": 3409
    },
    {
      "epoch": 0.7794285714285715,
      "grad_norm": 1.1616026163101196,
      "learning_rate": 0.0005,
      "loss": 2.5283803939819336,
      "step": 3410
    },
    {
      "epoch": 0.7796571428571428,
      "grad_norm": 0.9135298132896423,
      "learning_rate": 0.0005,
      "loss": 2.6750614643096924,
      "step": 3411
    },
    {
      "epoch": 0.7798857142857143,
      "grad_norm": 0.9603934288024902,
      "learning_rate": 0.0005,
      "loss": 2.927525520324707,
      "step": 3412
    },
    {
      "epoch": 0.7801142857142858,
      "grad_norm": 1.0010472536087036,
      "learning_rate": 0.0005,
      "loss": 2.756986618041992,
      "step": 3413
    },
    {
      "epoch": 0.7803428571428571,
      "grad_norm": 1.0229555368423462,
      "learning_rate": 0.0005,
      "loss": 2.4497907161712646,
      "step": 3414
    },
    {
      "epoch": 0.7805714285714286,
      "grad_norm": 0.9908697605133057,
      "learning_rate": 0.0005,
      "loss": 2.9064276218414307,
      "step": 3415
    },
    {
      "epoch": 0.7808,
      "grad_norm": 0.9008356928825378,
      "learning_rate": 0.0005,
      "loss": 2.6794495582580566,
      "step": 3416
    },
    {
      "epoch": 0.7810285714285714,
      "grad_norm": 1.1521250009536743,
      "learning_rate": 0.0005,
      "loss": 2.6518490314483643,
      "step": 3417
    },
    {
      "epoch": 0.7812571428571429,
      "grad_norm": 1.059922456741333,
      "learning_rate": 0.0005,
      "loss": 2.6507492065429688,
      "step": 3418
    },
    {
      "epoch": 0.7814857142857143,
      "grad_norm": 0.9861302375793457,
      "learning_rate": 0.0005,
      "loss": 2.767636299133301,
      "step": 3419
    },
    {
      "epoch": 0.7817142857142857,
      "grad_norm": 0.8416023254394531,
      "learning_rate": 0.0005,
      "loss": 2.4148263931274414,
      "step": 3420
    },
    {
      "epoch": 0.7819428571428572,
      "grad_norm": 0.7988889813423157,
      "learning_rate": 0.0005,
      "loss": 2.624342441558838,
      "step": 3421
    },
    {
      "epoch": 0.7821714285714285,
      "grad_norm": 0.9244321584701538,
      "learning_rate": 0.0005,
      "loss": 2.7188377380371094,
      "step": 3422
    },
    {
      "epoch": 0.7824,
      "grad_norm": 1.2396728992462158,
      "learning_rate": 0.0005,
      "loss": 2.85270357131958,
      "step": 3423
    },
    {
      "epoch": 0.7826285714285715,
      "grad_norm": 1.1015160083770752,
      "learning_rate": 0.0005,
      "loss": 2.5608527660369873,
      "step": 3424
    },
    {
      "epoch": 0.7828571428571428,
      "grad_norm": 1.0224246978759766,
      "learning_rate": 0.0005,
      "loss": 2.5866549015045166,
      "step": 3425
    },
    {
      "epoch": 0.7830857142857143,
      "grad_norm": 1.0477739572525024,
      "learning_rate": 0.0005,
      "loss": 2.98353910446167,
      "step": 3426
    },
    {
      "epoch": 0.7833142857142857,
      "grad_norm": 1.0656532049179077,
      "learning_rate": 0.0005,
      "loss": 2.4209060668945312,
      "step": 3427
    },
    {
      "epoch": 0.7835428571428571,
      "grad_norm": 1.0166218280792236,
      "learning_rate": 0.0005,
      "loss": 2.7954142093658447,
      "step": 3428
    },
    {
      "epoch": 0.7837714285714286,
      "grad_norm": 1.4163610935211182,
      "learning_rate": 0.0005,
      "loss": 2.52091646194458,
      "step": 3429
    },
    {
      "epoch": 0.784,
      "grad_norm": 1.025121808052063,
      "learning_rate": 0.0005,
      "loss": 2.5820815563201904,
      "step": 3430
    },
    {
      "epoch": 0.7842285714285714,
      "grad_norm": 0.8907953500747681,
      "learning_rate": 0.0005,
      "loss": 2.5743699073791504,
      "step": 3431
    },
    {
      "epoch": 0.7844571428571429,
      "grad_norm": 2.6309428215026855,
      "learning_rate": 0.0005,
      "loss": 2.695704460144043,
      "step": 3432
    },
    {
      "epoch": 0.7846857142857143,
      "grad_norm": 1.0577490329742432,
      "learning_rate": 0.0005,
      "loss": 2.711793899536133,
      "step": 3433
    },
    {
      "epoch": 0.7849142857142857,
      "grad_norm": 1.0130949020385742,
      "learning_rate": 0.0005,
      "loss": 2.4927427768707275,
      "step": 3434
    },
    {
      "epoch": 0.7851428571428571,
      "grad_norm": 0.8516344428062439,
      "learning_rate": 0.0005,
      "loss": 2.4344866275787354,
      "step": 3435
    },
    {
      "epoch": 0.7853714285714286,
      "grad_norm": 0.9230930209159851,
      "learning_rate": 0.0005,
      "loss": 2.5495967864990234,
      "step": 3436
    },
    {
      "epoch": 0.7856,
      "grad_norm": 0.9264714121818542,
      "learning_rate": 0.0005,
      "loss": 2.401949882507324,
      "step": 3437
    },
    {
      "epoch": 0.7858285714285714,
      "grad_norm": 0.874267041683197,
      "learning_rate": 0.0005,
      "loss": 2.322890520095825,
      "step": 3438
    },
    {
      "epoch": 0.7860571428571429,
      "grad_norm": 0.9116944670677185,
      "learning_rate": 0.0005,
      "loss": 2.641620635986328,
      "step": 3439
    },
    {
      "epoch": 0.7862857142857143,
      "grad_norm": 0.8626567125320435,
      "learning_rate": 0.0005,
      "loss": 2.5192294120788574,
      "step": 3440
    },
    {
      "epoch": 0.7865142857142857,
      "grad_norm": 0.9761642217636108,
      "learning_rate": 0.0005,
      "loss": 2.564131736755371,
      "step": 3441
    },
    {
      "epoch": 0.7867428571428572,
      "grad_norm": 0.8232711553573608,
      "learning_rate": 0.0005,
      "loss": 2.1660661697387695,
      "step": 3442
    },
    {
      "epoch": 0.7869714285714285,
      "grad_norm": 1.101946234703064,
      "learning_rate": 0.0005,
      "loss": 2.240633249282837,
      "step": 3443
    },
    {
      "epoch": 0.7872,
      "grad_norm": 1.228963017463684,
      "learning_rate": 0.0005,
      "loss": 2.2643771171569824,
      "step": 3444
    },
    {
      "epoch": 0.7874285714285715,
      "grad_norm": 1.1615384817123413,
      "learning_rate": 0.0005,
      "loss": 2.5384774208068848,
      "step": 3445
    },
    {
      "epoch": 0.7876571428571428,
      "grad_norm": 1.0266504287719727,
      "learning_rate": 0.0005,
      "loss": 2.4437572956085205,
      "step": 3446
    },
    {
      "epoch": 0.7878857142857143,
      "grad_norm": 0.9407846927642822,
      "learning_rate": 0.0005,
      "loss": 2.639653205871582,
      "step": 3447
    },
    {
      "epoch": 0.7881142857142858,
      "grad_norm": 0.9017980098724365,
      "learning_rate": 0.0005,
      "loss": 2.695439338684082,
      "step": 3448
    },
    {
      "epoch": 0.7883428571428571,
      "grad_norm": 0.9440131783485413,
      "learning_rate": 0.0005,
      "loss": 2.484224796295166,
      "step": 3449
    },
    {
      "epoch": 0.7885714285714286,
      "grad_norm": 1.1344900131225586,
      "learning_rate": 0.0005,
      "loss": 2.509153127670288,
      "step": 3450
    },
    {
      "epoch": 0.7888,
      "grad_norm": 1.0504850149154663,
      "learning_rate": 0.0005,
      "loss": 2.538909435272217,
      "step": 3451
    },
    {
      "epoch": 0.7890285714285714,
      "grad_norm": 0.9562268853187561,
      "learning_rate": 0.0005,
      "loss": 2.495973587036133,
      "step": 3452
    },
    {
      "epoch": 0.7892571428571429,
      "grad_norm": 0.8904072642326355,
      "learning_rate": 0.0005,
      "loss": 2.345353603363037,
      "step": 3453
    },
    {
      "epoch": 0.7894857142857142,
      "grad_norm": 0.9968966245651245,
      "learning_rate": 0.0005,
      "loss": 2.443380832672119,
      "step": 3454
    },
    {
      "epoch": 0.7897142857142857,
      "grad_norm": 0.8171984553337097,
      "learning_rate": 0.0005,
      "loss": 2.711686134338379,
      "step": 3455
    },
    {
      "epoch": 0.7899428571428572,
      "grad_norm": 0.9639772772789001,
      "learning_rate": 0.0005,
      "loss": 2.439023017883301,
      "step": 3456
    },
    {
      "epoch": 0.7901714285714285,
      "grad_norm": 0.8085919618606567,
      "learning_rate": 0.0005,
      "loss": 2.591012954711914,
      "step": 3457
    },
    {
      "epoch": 0.7904,
      "grad_norm": 0.8805829286575317,
      "learning_rate": 0.0005,
      "loss": 2.3732919692993164,
      "step": 3458
    },
    {
      "epoch": 0.7906285714285715,
      "grad_norm": 1.231400728225708,
      "learning_rate": 0.0005,
      "loss": 2.703495740890503,
      "step": 3459
    },
    {
      "epoch": 0.7908571428571428,
      "grad_norm": 0.8943395018577576,
      "learning_rate": 0.0005,
      "loss": 2.595201015472412,
      "step": 3460
    },
    {
      "epoch": 0.7910857142857143,
      "grad_norm": 1.0072506666183472,
      "learning_rate": 0.0005,
      "loss": 2.874647617340088,
      "step": 3461
    },
    {
      "epoch": 0.7913142857142857,
      "grad_norm": 1.0038104057312012,
      "learning_rate": 0.0005,
      "loss": 2.678380012512207,
      "step": 3462
    },
    {
      "epoch": 0.7915428571428571,
      "grad_norm": 1.071566104888916,
      "learning_rate": 0.0005,
      "loss": 2.574726104736328,
      "step": 3463
    },
    {
      "epoch": 0.7917714285714286,
      "grad_norm": 3.443279981613159,
      "learning_rate": 0.0005,
      "loss": 2.7216033935546875,
      "step": 3464
    },
    {
      "epoch": 0.792,
      "grad_norm": 1.0231815576553345,
      "learning_rate": 0.0005,
      "loss": 2.3638744354248047,
      "step": 3465
    },
    {
      "epoch": 0.7922285714285714,
      "grad_norm": 0.970278799533844,
      "learning_rate": 0.0005,
      "loss": 2.6222124099731445,
      "step": 3466
    },
    {
      "epoch": 0.7924571428571429,
      "grad_norm": 1.0507868528366089,
      "learning_rate": 0.0005,
      "loss": 2.6106114387512207,
      "step": 3467
    },
    {
      "epoch": 0.7926857142857143,
      "grad_norm": 1.0454611778259277,
      "learning_rate": 0.0005,
      "loss": 2.862518072128296,
      "step": 3468
    },
    {
      "epoch": 0.7929142857142857,
      "grad_norm": 0.9361552596092224,
      "learning_rate": 0.0005,
      "loss": 2.669203758239746,
      "step": 3469
    },
    {
      "epoch": 0.7931428571428571,
      "grad_norm": 0.9969595670700073,
      "learning_rate": 0.0005,
      "loss": 2.670088291168213,
      "step": 3470
    },
    {
      "epoch": 0.7933714285714286,
      "grad_norm": 0.7825395464897156,
      "learning_rate": 0.0005,
      "loss": 2.581000804901123,
      "step": 3471
    },
    {
      "epoch": 0.7936,
      "grad_norm": 0.9642606973648071,
      "learning_rate": 0.0005,
      "loss": 2.4494740962982178,
      "step": 3472
    },
    {
      "epoch": 0.7938285714285714,
      "grad_norm": 0.92669278383255,
      "learning_rate": 0.0005,
      "loss": 2.69963002204895,
      "step": 3473
    },
    {
      "epoch": 0.7940571428571429,
      "grad_norm": 0.8999539017677307,
      "learning_rate": 0.0005,
      "loss": 2.700453758239746,
      "step": 3474
    },
    {
      "epoch": 0.7942857142857143,
      "grad_norm": 1.0189076662063599,
      "learning_rate": 0.0005,
      "loss": 2.527930736541748,
      "step": 3475
    },
    {
      "epoch": 0.7945142857142857,
      "grad_norm": 1.0111422538757324,
      "learning_rate": 0.0005,
      "loss": 2.651087522506714,
      "step": 3476
    },
    {
      "epoch": 0.7947428571428572,
      "grad_norm": 1.0056544542312622,
      "learning_rate": 0.0005,
      "loss": 2.9672133922576904,
      "step": 3477
    },
    {
      "epoch": 0.7949714285714286,
      "grad_norm": 1.0733054876327515,
      "learning_rate": 0.0005,
      "loss": 2.4653468132019043,
      "step": 3478
    },
    {
      "epoch": 0.7952,
      "grad_norm": 1.3245974779129028,
      "learning_rate": 0.0005,
      "loss": 2.574251174926758,
      "step": 3479
    },
    {
      "epoch": 0.7954285714285714,
      "grad_norm": 1.0980571508407593,
      "learning_rate": 0.0005,
      "loss": 2.8513081073760986,
      "step": 3480
    },
    {
      "epoch": 0.7956571428571428,
      "grad_norm": 1.2129685878753662,
      "learning_rate": 0.0005,
      "loss": 2.7205233573913574,
      "step": 3481
    },
    {
      "epoch": 0.7958857142857143,
      "grad_norm": 0.9259638786315918,
      "learning_rate": 0.0005,
      "loss": 2.966466188430786,
      "step": 3482
    },
    {
      "epoch": 0.7961142857142857,
      "grad_norm": 1.15169358253479,
      "learning_rate": 0.0005,
      "loss": 2.359104633331299,
      "step": 3483
    },
    {
      "epoch": 0.7963428571428571,
      "grad_norm": 0.8839331865310669,
      "learning_rate": 0.0005,
      "loss": 2.250837802886963,
      "step": 3484
    },
    {
      "epoch": 0.7965714285714286,
      "grad_norm": 0.8854320049285889,
      "learning_rate": 0.0005,
      "loss": 2.52363657951355,
      "step": 3485
    },
    {
      "epoch": 0.7968,
      "grad_norm": 1.0360883474349976,
      "learning_rate": 0.0005,
      "loss": 2.77638578414917,
      "step": 3486
    },
    {
      "epoch": 0.7970285714285714,
      "grad_norm": 0.9082557559013367,
      "learning_rate": 0.0005,
      "loss": 2.611889600753784,
      "step": 3487
    },
    {
      "epoch": 0.7972571428571429,
      "grad_norm": 1.0841282606124878,
      "learning_rate": 0.0005,
      "loss": 2.6551780700683594,
      "step": 3488
    },
    {
      "epoch": 0.7974857142857142,
      "grad_norm": 1.1844466924667358,
      "learning_rate": 0.0005,
      "loss": 2.6026995182037354,
      "step": 3489
    },
    {
      "epoch": 0.7977142857142857,
      "grad_norm": 1.0773898363113403,
      "learning_rate": 0.0005,
      "loss": 2.8073441982269287,
      "step": 3490
    },
    {
      "epoch": 0.7979428571428572,
      "grad_norm": 0.9980083703994751,
      "learning_rate": 0.0005,
      "loss": 2.7433533668518066,
      "step": 3491
    },
    {
      "epoch": 0.7981714285714285,
      "grad_norm": 1.161808967590332,
      "learning_rate": 0.0005,
      "loss": 2.4171390533447266,
      "step": 3492
    },
    {
      "epoch": 0.7984,
      "grad_norm": 1.1303681135177612,
      "learning_rate": 0.0005,
      "loss": 2.5076961517333984,
      "step": 3493
    },
    {
      "epoch": 0.7986285714285715,
      "grad_norm": 0.9512962698936462,
      "learning_rate": 0.0005,
      "loss": 2.7235944271087646,
      "step": 3494
    },
    {
      "epoch": 0.7988571428571428,
      "grad_norm": 0.8970752954483032,
      "learning_rate": 0.0005,
      "loss": 2.4733614921569824,
      "step": 3495
    },
    {
      "epoch": 0.7990857142857143,
      "grad_norm": 0.9416807293891907,
      "learning_rate": 0.0005,
      "loss": 2.5888900756835938,
      "step": 3496
    },
    {
      "epoch": 0.7993142857142858,
      "grad_norm": 0.9981902241706848,
      "learning_rate": 0.0005,
      "loss": 2.4608681201934814,
      "step": 3497
    },
    {
      "epoch": 0.7995428571428571,
      "grad_norm": 0.9270004034042358,
      "learning_rate": 0.0005,
      "loss": 2.714918613433838,
      "step": 3498
    },
    {
      "epoch": 0.7997714285714286,
      "grad_norm": 2.4416189193725586,
      "learning_rate": 0.0005,
      "loss": 2.5250418186187744,
      "step": 3499
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.8995660543441772,
      "learning_rate": 0.0005,
      "loss": 2.4963736534118652,
      "step": 3500
    },
    {
      "epoch": 0.8002285714285714,
      "grad_norm": 0.9400513768196106,
      "learning_rate": 0.0005,
      "loss": 2.861433506011963,
      "step": 3501
    },
    {
      "epoch": 0.8004571428571429,
      "grad_norm": 1.10733163356781,
      "learning_rate": 0.0004999983886377472,
      "loss": 2.343573570251465,
      "step": 3502
    },
    {
      "epoch": 0.8006857142857143,
      "grad_norm": 2.5315773487091064,
      "learning_rate": 0.0004999935545717605,
      "loss": 2.354417324066162,
      "step": 3503
    },
    {
      "epoch": 0.8009142857142857,
      "grad_norm": 0.9891689419746399,
      "learning_rate": 0.0004999854978643555,
      "loss": 2.293482780456543,
      "step": 3504
    },
    {
      "epoch": 0.8011428571428572,
      "grad_norm": 1.1317551136016846,
      "learning_rate": 0.0004999742186193903,
      "loss": 2.4704666137695312,
      "step": 3505
    },
    {
      "epoch": 0.8013714285714286,
      "grad_norm": 1.0177383422851562,
      "learning_rate": 0.0004999597169822646,
      "loss": 2.595102310180664,
      "step": 3506
    },
    {
      "epoch": 0.8016,
      "grad_norm": 1.0432238578796387,
      "learning_rate": 0.0004999419931399174,
      "loss": 2.582101821899414,
      "step": 3507
    },
    {
      "epoch": 0.8018285714285714,
      "grad_norm": 0.756969153881073,
      "learning_rate": 0.000499921047320825,
      "loss": 2.5390539169311523,
      "step": 3508
    },
    {
      "epoch": 0.8020571428571428,
      "grad_norm": 0.9723111391067505,
      "learning_rate": 0.0004998968797949979,
      "loss": 2.659449338912964,
      "step": 3509
    },
    {
      "epoch": 0.8022857142857143,
      "grad_norm": 0.9892172813415527,
      "learning_rate": 0.000499869490873977,
      "loss": 2.6501283645629883,
      "step": 3510
    },
    {
      "epoch": 0.8025142857142857,
      "grad_norm": 1.1489368677139282,
      "learning_rate": 0.0004998388809108303,
      "loss": 2.4652974605560303,
      "step": 3511
    },
    {
      "epoch": 0.8027428571428571,
      "grad_norm": 1.1062411069869995,
      "learning_rate": 0.0004998050503001477,
      "loss": 2.6238508224487305,
      "step": 3512
    },
    {
      "epoch": 0.8029714285714286,
      "grad_norm": 1.0919049978256226,
      "learning_rate": 0.0004997679994780361,
      "loss": 2.4636316299438477,
      "step": 3513
    },
    {
      "epoch": 0.8032,
      "grad_norm": 0.8983208537101746,
      "learning_rate": 0.0004997277289221137,
      "loss": 2.2434208393096924,
      "step": 3514
    },
    {
      "epoch": 0.8034285714285714,
      "grad_norm": 0.9291174411773682,
      "learning_rate": 0.0004996842391515044,
      "loss": 2.770772933959961,
      "step": 3515
    },
    {
      "epoch": 0.8036571428571428,
      "grad_norm": 1.1933844089508057,
      "learning_rate": 0.0004996375307268303,
      "loss": 2.3542633056640625,
      "step": 3516
    },
    {
      "epoch": 0.8038857142857143,
      "grad_norm": 1.1344382762908936,
      "learning_rate": 0.0004995876042502048,
      "loss": 2.135617733001709,
      "step": 3517
    },
    {
      "epoch": 0.8041142857142857,
      "grad_norm": 1.1032919883728027,
      "learning_rate": 0.0004995344603652253,
      "loss": 2.5921685695648193,
      "step": 3518
    },
    {
      "epoch": 0.8043428571428571,
      "grad_norm": 0.9483276605606079,
      "learning_rate": 0.0004994780997569639,
      "loss": 2.4072866439819336,
      "step": 3519
    },
    {
      "epoch": 0.8045714285714286,
      "grad_norm": 1.036401391029358,
      "learning_rate": 0.0004994185231519596,
      "loss": 2.940797805786133,
      "step": 3520
    },
    {
      "epoch": 0.8048,
      "grad_norm": 1.017916202545166,
      "learning_rate": 0.0004993557313182085,
      "loss": 2.7495150566101074,
      "step": 3521
    },
    {
      "epoch": 0.8050285714285714,
      "grad_norm": 1.091638207435608,
      "learning_rate": 0.0004992897250651535,
      "loss": 2.621140480041504,
      "step": 3522
    },
    {
      "epoch": 0.8052571428571429,
      "grad_norm": 1.0136083364486694,
      "learning_rate": 0.0004992205052436746,
      "loss": 2.325887680053711,
      "step": 3523
    },
    {
      "epoch": 0.8054857142857142,
      "grad_norm": 0.995622992515564,
      "learning_rate": 0.0004991480727460772,
      "loss": 2.6348674297332764,
      "step": 3524
    },
    {
      "epoch": 0.8057142857142857,
      "grad_norm": 0.9032522439956665,
      "learning_rate": 0.0004990724285060816,
      "loss": 2.5892748832702637,
      "step": 3525
    },
    {
      "epoch": 0.8059428571428572,
      "grad_norm": 1.1997171640396118,
      "learning_rate": 0.0004989935734988098,
      "loss": 2.550546646118164,
      "step": 3526
    },
    {
      "epoch": 0.8061714285714285,
      "grad_norm": 1.0518417358398438,
      "learning_rate": 0.0004989115087407736,
      "loss": 2.348337173461914,
      "step": 3527
    },
    {
      "epoch": 0.8064,
      "grad_norm": 1.1299042701721191,
      "learning_rate": 0.0004988262352898616,
      "loss": 2.6498546600341797,
      "step": 3528
    },
    {
      "epoch": 0.8066285714285715,
      "grad_norm": 1.2243552207946777,
      "learning_rate": 0.0004987377542453251,
      "loss": 2.720813274383545,
      "step": 3529
    },
    {
      "epoch": 0.8068571428571428,
      "grad_norm": 0.9944297671318054,
      "learning_rate": 0.000498646066747764,
      "loss": 2.475344657897949,
      "step": 3530
    },
    {
      "epoch": 0.8070857142857143,
      "grad_norm": 1.1082712411880493,
      "learning_rate": 0.0004985511739791129,
      "loss": 2.608341693878174,
      "step": 3531
    },
    {
      "epoch": 0.8073142857142858,
      "grad_norm": 0.9035975337028503,
      "learning_rate": 0.0004984530771626244,
      "loss": 2.366874933242798,
      "step": 3532
    },
    {
      "epoch": 0.8075428571428571,
      "grad_norm": 1.0480999946594238,
      "learning_rate": 0.0004983517775628549,
      "loss": 2.491131544113159,
      "step": 3533
    },
    {
      "epoch": 0.8077714285714286,
      "grad_norm": 1.2523257732391357,
      "learning_rate": 0.0004982472764856469,
      "loss": 2.0038294792175293,
      "step": 3534
    },
    {
      "epoch": 0.808,
      "grad_norm": 1.1123839616775513,
      "learning_rate": 0.0004981395752781134,
      "loss": 2.7273781299591064,
      "step": 3535
    },
    {
      "epoch": 0.8082285714285714,
      "grad_norm": 0.9801051020622253,
      "learning_rate": 0.0004980286753286195,
      "loss": 2.467543125152588,
      "step": 3536
    },
    {
      "epoch": 0.8084571428571429,
      "grad_norm": 1.288655400276184,
      "learning_rate": 0.0004979145780667652,
      "loss": 2.658557891845703,
      "step": 3537
    },
    {
      "epoch": 0.8086857142857143,
      "grad_norm": 1.1749600172042847,
      "learning_rate": 0.0004977972849633666,
      "loss": 2.6761271953582764,
      "step": 3538
    },
    {
      "epoch": 0.8089142857142857,
      "grad_norm": 1.0374549627304077,
      "learning_rate": 0.0004976767975304373,
      "loss": 2.6834607124328613,
      "step": 3539
    },
    {
      "epoch": 0.8091428571428572,
      "grad_norm": 1.1433669328689575,
      "learning_rate": 0.0004975531173211683,
      "loss": 2.584167003631592,
      "step": 3540
    },
    {
      "epoch": 0.8093714285714285,
      "grad_norm": 1.1261531114578247,
      "learning_rate": 0.0004974262459299087,
      "loss": 2.534224033355713,
      "step": 3541
    },
    {
      "epoch": 0.8096,
      "grad_norm": 1.0067051649093628,
      "learning_rate": 0.0004972961849921447,
      "loss": 2.509951591491699,
      "step": 3542
    },
    {
      "epoch": 0.8098285714285715,
      "grad_norm": 1.0079445838928223,
      "learning_rate": 0.0004971629361844785,
      "loss": 2.7700753211975098,
      "step": 3543
    },
    {
      "epoch": 0.8100571428571428,
      "grad_norm": 0.8959079384803772,
      "learning_rate": 0.0004970265012246069,
      "loss": 2.2554025650024414,
      "step": 3544
    },
    {
      "epoch": 0.8102857142857143,
      "grad_norm": 1.2946646213531494,
      "learning_rate": 0.0004968868818712992,
      "loss": 2.633452892303467,
      "step": 3545
    },
    {
      "epoch": 0.8105142857142857,
      "grad_norm": 1.0685063600540161,
      "learning_rate": 0.0004967440799243739,
      "loss": 2.637077808380127,
      "step": 3546
    },
    {
      "epoch": 0.8107428571428571,
      "grad_norm": 0.9974644780158997,
      "learning_rate": 0.0004965980972246767,
      "loss": 2.4366726875305176,
      "step": 3547
    },
    {
      "epoch": 0.8109714285714286,
      "grad_norm": 0.9632290601730347,
      "learning_rate": 0.0004964489356540555,
      "loss": 2.456230878829956,
      "step": 3548
    },
    {
      "epoch": 0.8112,
      "grad_norm": 0.8479512929916382,
      "learning_rate": 0.000496296597135337,
      "loss": 2.644582748413086,
      "step": 3549
    },
    {
      "epoch": 0.8114285714285714,
      "grad_norm": 0.7611117362976074,
      "learning_rate": 0.0004961410836323014,
      "loss": 2.200568675994873,
      "step": 3550
    },
    {
      "epoch": 0.8116571428571429,
      "grad_norm": 0.8606249094009399,
      "learning_rate": 0.0004959823971496574,
      "loss": 2.5033392906188965,
      "step": 3551
    },
    {
      "epoch": 0.8118857142857143,
      "grad_norm": 1.0118730068206787,
      "learning_rate": 0.0004958205397330164,
      "loss": 2.3400020599365234,
      "step": 3552
    },
    {
      "epoch": 0.8121142857142857,
      "grad_norm": 1.0510293245315552,
      "learning_rate": 0.0004956555134688659,
      "loss": 2.4576430320739746,
      "step": 3553
    },
    {
      "epoch": 0.8123428571428571,
      "grad_norm": 1.0720539093017578,
      "learning_rate": 0.0004954873204845423,
      "loss": 2.36843204498291,
      "step": 3554
    },
    {
      "epoch": 0.8125714285714286,
      "grad_norm": 0.9073250889778137,
      "learning_rate": 0.0004953159629482045,
      "loss": 2.227487802505493,
      "step": 3555
    },
    {
      "epoch": 0.8128,
      "grad_norm": 1.1289646625518799,
      "learning_rate": 0.000495141443068805,
      "loss": 2.453900098800659,
      "step": 3556
    },
    {
      "epoch": 0.8130285714285714,
      "grad_norm": 0.8693168759346008,
      "learning_rate": 0.0004949637630960618,
      "loss": 2.4525485038757324,
      "step": 3557
    },
    {
      "epoch": 0.8132571428571429,
      "grad_norm": 1.149539589881897,
      "learning_rate": 0.000494782925320429,
      "loss": 2.692354202270508,
      "step": 3558
    },
    {
      "epoch": 0.8134857142857143,
      "grad_norm": 1.4070053100585938,
      "learning_rate": 0.0004945989320730684,
      "loss": 2.7279982566833496,
      "step": 3559
    },
    {
      "epoch": 0.8137142857142857,
      "grad_norm": 1.9433865547180176,
      "learning_rate": 0.0004944117857258179,
      "loss": 2.586851119995117,
      "step": 3560
    },
    {
      "epoch": 0.8139428571428572,
      "grad_norm": 1.4222851991653442,
      "learning_rate": 0.000494221488691162,
      "loss": 2.4875495433807373,
      "step": 3561
    },
    {
      "epoch": 0.8141714285714285,
      "grad_norm": 1.153333306312561,
      "learning_rate": 0.0004940280434222003,
      "loss": 2.4005165100097656,
      "step": 3562
    },
    {
      "epoch": 0.8144,
      "grad_norm": 0.794992208480835,
      "learning_rate": 0.0004938314524126162,
      "loss": 2.5777788162231445,
      "step": 3563
    },
    {
      "epoch": 0.8146285714285715,
      "grad_norm": 0.9549087882041931,
      "learning_rate": 0.0004936317181966443,
      "loss": 2.7473411560058594,
      "step": 3564
    },
    {
      "epoch": 0.8148571428571428,
      "grad_norm": 1.0159670114517212,
      "learning_rate": 0.000493428843349038,
      "loss": 3.3631293773651123,
      "step": 3565
    },
    {
      "epoch": 0.8150857142857143,
      "grad_norm": 1.0269805192947388,
      "learning_rate": 0.0004932228304850363,
      "loss": 2.6795449256896973,
      "step": 3566
    },
    {
      "epoch": 0.8153142857142858,
      "grad_norm": 0.9650386571884155,
      "learning_rate": 0.0004930136822603299,
      "loss": 2.5585408210754395,
      "step": 3567
    },
    {
      "epoch": 0.8155428571428571,
      "grad_norm": 0.9664074778556824,
      "learning_rate": 0.0004928014013710274,
      "loss": 2.5565390586853027,
      "step": 3568
    },
    {
      "epoch": 0.8157714285714286,
      "grad_norm": 0.8322105407714844,
      "learning_rate": 0.00049258599055362,
      "loss": 2.51846981048584,
      "step": 3569
    },
    {
      "epoch": 0.816,
      "grad_norm": 0.8779276013374329,
      "learning_rate": 0.0004923674525849466,
      "loss": 2.707570791244507,
      "step": 3570
    },
    {
      "epoch": 0.8162285714285714,
      "grad_norm": 1.2742432355880737,
      "learning_rate": 0.0004921457902821578,
      "loss": 2.791362762451172,
      "step": 3571
    },
    {
      "epoch": 0.8164571428571429,
      "grad_norm": 0.8030603528022766,
      "learning_rate": 0.0004919210065026799,
      "loss": 2.7675552368164062,
      "step": 3572
    },
    {
      "epoch": 0.8166857142857142,
      "grad_norm": 1.1097166538238525,
      "learning_rate": 0.0004916931041441775,
      "loss": 2.6555495262145996,
      "step": 3573
    },
    {
      "epoch": 0.8169142857142857,
      "grad_norm": 1.0153530836105347,
      "learning_rate": 0.0004914620861445168,
      "loss": 2.3006539344787598,
      "step": 3574
    },
    {
      "epoch": 0.8171428571428572,
      "grad_norm": 1.058773159980774,
      "learning_rate": 0.0004912279554817273,
      "loss": 2.9586710929870605,
      "step": 3575
    },
    {
      "epoch": 0.8173714285714285,
      "grad_norm": 1.2076873779296875,
      "learning_rate": 0.0004909907151739633,
      "loss": 2.589195728302002,
      "step": 3576
    },
    {
      "epoch": 0.8176,
      "grad_norm": 1.2743281126022339,
      "learning_rate": 0.0004907503682794656,
      "loss": 2.3617660999298096,
      "step": 3577
    },
    {
      "epoch": 0.8178285714285715,
      "grad_norm": 0.7095112800598145,
      "learning_rate": 0.0004905069178965214,
      "loss": 2.515554189682007,
      "step": 3578
    },
    {
      "epoch": 0.8180571428571428,
      "grad_norm": 0.9168731570243835,
      "learning_rate": 0.000490260367163425,
      "loss": 2.215104341506958,
      "step": 3579
    },
    {
      "epoch": 0.8182857142857143,
      "grad_norm": 1.1349263191223145,
      "learning_rate": 0.0004900107192584364,
      "loss": 2.295797824859619,
      "step": 3580
    },
    {
      "epoch": 0.8185142857142857,
      "grad_norm": 0.9421923756599426,
      "learning_rate": 0.0004897579773997414,
      "loss": 2.567477226257324,
      "step": 3581
    },
    {
      "epoch": 0.8187428571428571,
      "grad_norm": 1.2354530096054077,
      "learning_rate": 0.0004895021448454097,
      "loss": 2.6950578689575195,
      "step": 3582
    },
    {
      "epoch": 0.8189714285714286,
      "grad_norm": 1.0551433563232422,
      "learning_rate": 0.0004892432248933525,
      "loss": 2.521472930908203,
      "step": 3583
    },
    {
      "epoch": 0.8192,
      "grad_norm": 1.1117446422576904,
      "learning_rate": 0.0004889812208812805,
      "loss": 2.418936014175415,
      "step": 3584
    },
    {
      "epoch": 0.8194285714285714,
      "grad_norm": 1.3595914840698242,
      "learning_rate": 0.0004887161361866607,
      "loss": 2.434295892715454,
      "step": 3585
    },
    {
      "epoch": 0.8196571428571429,
      "grad_norm": 1.028976559638977,
      "learning_rate": 0.000488447974226673,
      "loss": 2.1975889205932617,
      "step": 3586
    },
    {
      "epoch": 0.8198857142857143,
      "grad_norm": 1.0496193170547485,
      "learning_rate": 0.0004881767384581658,
      "loss": 2.3323206901550293,
      "step": 3587
    },
    {
      "epoch": 0.8201142857142857,
      "grad_norm": 0.9148577451705933,
      "learning_rate": 0.00048790243237761165,
      "loss": 2.7798733711242676,
      "step": 3588
    },
    {
      "epoch": 0.8203428571428572,
      "grad_norm": 0.8512684106826782,
      "learning_rate": 0.0004876250595210624,
      "loss": 2.536036968231201,
      "step": 3589
    },
    {
      "epoch": 0.8205714285714286,
      "grad_norm": 1.1659409999847412,
      "learning_rate": 0.00048734462346410317,
      "loss": 2.5445775985717773,
      "step": 3590
    },
    {
      "epoch": 0.8208,
      "grad_norm": 0.8995631337165833,
      "learning_rate": 0.0004870611278218066,
      "loss": 2.613753318786621,
      "step": 3591
    },
    {
      "epoch": 0.8210285714285714,
      "grad_norm": 1.0549442768096924,
      "learning_rate": 0.0004867745762486861,
      "loss": 2.3998334407806396,
      "step": 3592
    },
    {
      "epoch": 0.8212571428571429,
      "grad_norm": 1.3176356554031372,
      "learning_rate": 0.00048648497243864873,
      "loss": 2.70784592628479,
      "step": 3593
    },
    {
      "epoch": 0.8214857142857143,
      "grad_norm": 1.041659951210022,
      "learning_rate": 0.00048619232012494775,
      "loss": 2.3167309761047363,
      "step": 3594
    },
    {
      "epoch": 0.8217142857142857,
      "grad_norm": 0.9444294571876526,
      "learning_rate": 0.0004858966230801343,
      "loss": 2.578348398208618,
      "step": 3595
    },
    {
      "epoch": 0.8219428571428572,
      "grad_norm": 1.1938420534133911,
      "learning_rate": 0.00048559788511600875,
      "loss": 2.51796293258667,
      "step": 3596
    },
    {
      "epoch": 0.8221714285714286,
      "grad_norm": 0.9673129320144653,
      "learning_rate": 0.00048529611008357173,
      "loss": 2.6836204528808594,
      "step": 3597
    },
    {
      "epoch": 0.8224,
      "grad_norm": 1.2784783840179443,
      "learning_rate": 0.00048499130187297447,
      "loss": 2.515528678894043,
      "step": 3598
    },
    {
      "epoch": 0.8226285714285714,
      "grad_norm": 1.0411213636398315,
      "learning_rate": 0.00048468346441346853,
      "loss": 2.6491355895996094,
      "step": 3599
    },
    {
      "epoch": 0.8228571428571428,
      "grad_norm": 0.9567443132400513,
      "learning_rate": 0.0004843726016733552,
      "loss": 2.613990306854248,
      "step": 3600
    },
    {
      "epoch": 0.8230857142857143,
      "grad_norm": 1.0416834354400635,
      "learning_rate": 0.0004840587176599343,
      "loss": 2.0582833290100098,
      "step": 3601
    },
    {
      "epoch": 0.8233142857142857,
      "grad_norm": 1.396893858909607,
      "learning_rate": 0.0004837418164194527,
      "loss": 2.932270050048828,
      "step": 3602
    },
    {
      "epoch": 0.8235428571428571,
      "grad_norm": 0.9355141520500183,
      "learning_rate": 0.000483421902037052,
      "loss": 2.5989255905151367,
      "step": 3603
    },
    {
      "epoch": 0.8237714285714286,
      "grad_norm": 1.1451259851455688,
      "learning_rate": 0.00048309897863671583,
      "loss": 2.381955862045288,
      "step": 3604
    },
    {
      "epoch": 0.824,
      "grad_norm": 0.9416282176971436,
      "learning_rate": 0.0004827730503812168,
      "loss": 2.5005924701690674,
      "step": 3605
    },
    {
      "epoch": 0.8242285714285714,
      "grad_norm": 1.0093828439712524,
      "learning_rate": 0.00048244412147206283,
      "loss": 2.447427988052368,
      "step": 3606
    },
    {
      "epoch": 0.8244571428571429,
      "grad_norm": 1.1709988117218018,
      "learning_rate": 0.000482112196149443,
      "loss": 2.693192720413208,
      "step": 3607
    },
    {
      "epoch": 0.8246857142857142,
      "grad_norm": 0.9878803491592407,
      "learning_rate": 0.0004817772786921727,
      "loss": 2.751168966293335,
      "step": 3608
    },
    {
      "epoch": 0.8249142857142857,
      "grad_norm": 0.894743025302887,
      "learning_rate": 0.00048143937341763883,
      "loss": 2.595012664794922,
      "step": 3609
    },
    {
      "epoch": 0.8251428571428572,
      "grad_norm": 1.00009024143219,
      "learning_rate": 0.00048109848468174364,
      "loss": 2.302870273590088,
      "step": 3610
    },
    {
      "epoch": 0.8253714285714285,
      "grad_norm": 1.0925135612487793,
      "learning_rate": 0.00048075461687884934,
      "loss": 2.5080814361572266,
      "step": 3611
    },
    {
      "epoch": 0.8256,
      "grad_norm": 1.0065100193023682,
      "learning_rate": 0.00048040777444172047,
      "loss": 2.879535675048828,
      "step": 3612
    },
    {
      "epoch": 0.8258285714285715,
      "grad_norm": 0.9271668195724487,
      "learning_rate": 0.0004800579618414676,
      "loss": 2.57796049118042,
      "step": 3613
    },
    {
      "epoch": 0.8260571428571428,
      "grad_norm": 0.8602476119995117,
      "learning_rate": 0.00047970518358748926,
      "loss": 2.7210469245910645,
      "step": 3614
    },
    {
      "epoch": 0.8262857142857143,
      "grad_norm": 0.728459358215332,
      "learning_rate": 0.000479349444227414,
      "loss": 2.293386459350586,
      "step": 3615
    },
    {
      "epoch": 0.8265142857142858,
      "grad_norm": 0.8982915282249451,
      "learning_rate": 0.00047899074834704166,
      "loss": 2.8955347537994385,
      "step": 3616
    },
    {
      "epoch": 0.8267428571428571,
      "grad_norm": 1.0119596719741821,
      "learning_rate": 0.00047862910057028414,
      "loss": 2.7901358604431152,
      "step": 3617
    },
    {
      "epoch": 0.8269714285714286,
      "grad_norm": 0.8720492124557495,
      "learning_rate": 0.0004782645055591062,
      "loss": 2.5741186141967773,
      "step": 3618
    },
    {
      "epoch": 0.8272,
      "grad_norm": 1.0028767585754395,
      "learning_rate": 0.0004778969680134647,
      "loss": 2.619542360305786,
      "step": 3619
    },
    {
      "epoch": 0.8274285714285714,
      "grad_norm": 1.1666746139526367,
      "learning_rate": 0.00047752649267124894,
      "loss": 2.6033122539520264,
      "step": 3620
    },
    {
      "epoch": 0.8276571428571429,
      "grad_norm": 0.8286680579185486,
      "learning_rate": 0.0004771530843082186,
      "loss": 2.6547000408172607,
      "step": 3621
    },
    {
      "epoch": 0.8278857142857143,
      "grad_norm": 0.8922433853149414,
      "learning_rate": 0.0004767767477379429,
      "loss": 2.33679461479187,
      "step": 3622
    },
    {
      "epoch": 0.8281142857142857,
      "grad_norm": 0.8373787999153137,
      "learning_rate": 0.0004763974878117382,
      "loss": 2.5359292030334473,
      "step": 3623
    },
    {
      "epoch": 0.8283428571428572,
      "grad_norm": 0.9335997700691223,
      "learning_rate": 0.00047601530941860536,
      "loss": 2.213676691055298,
      "step": 3624
    },
    {
      "epoch": 0.8285714285714286,
      "grad_norm": 0.9479578733444214,
      "learning_rate": 0.00047563021748516735,
      "loss": 2.493492841720581,
      "step": 3625
    },
    {
      "epoch": 0.8288,
      "grad_norm": 1.6195470094680786,
      "learning_rate": 0.00047524221697560476,
      "loss": 2.7678062915802,
      "step": 3626
    },
    {
      "epoch": 0.8290285714285714,
      "grad_norm": 1.0669848918914795,
      "learning_rate": 0.0004748513128915928,
      "loss": 2.7951011657714844,
      "step": 3627
    },
    {
      "epoch": 0.8292571428571428,
      "grad_norm": 0.8400890827178955,
      "learning_rate": 0.00047445751027223594,
      "loss": 2.161846876144409,
      "step": 3628
    },
    {
      "epoch": 0.8294857142857143,
      "grad_norm": 1.005430817604065,
      "learning_rate": 0.0004740608141940038,
      "loss": 2.9706008434295654,
      "step": 3629
    },
    {
      "epoch": 0.8297142857142857,
      "grad_norm": 1.0195523500442505,
      "learning_rate": 0.0004736612297706649,
      "loss": 2.511512041091919,
      "step": 3630
    },
    {
      "epoch": 0.8299428571428571,
      "grad_norm": 0.7673805356025696,
      "learning_rate": 0.0004732587621532214,
      "loss": 2.3640687465667725,
      "step": 3631
    },
    {
      "epoch": 0.8301714285714286,
      "grad_norm": 1.0310426950454712,
      "learning_rate": 0.0004728534165298422,
      "loss": 2.4720520973205566,
      "step": 3632
    },
    {
      "epoch": 0.8304,
      "grad_norm": 0.7894354462623596,
      "learning_rate": 0.0004724451981257966,
      "loss": 2.475507974624634,
      "step": 3633
    },
    {
      "epoch": 0.8306285714285714,
      "grad_norm": 0.9740471839904785,
      "learning_rate": 0.00047203411220338615,
      "loss": 2.67801833152771,
      "step": 3634
    },
    {
      "epoch": 0.8308571428571428,
      "grad_norm": 0.9856567978858948,
      "learning_rate": 0.00047162016406187776,
      "loss": 2.55776309967041,
      "step": 3635
    },
    {
      "epoch": 0.8310857142857143,
      "grad_norm": 1.0786347389221191,
      "learning_rate": 0.0004712033590374346,
      "loss": 2.548703193664551,
      "step": 3636
    },
    {
      "epoch": 0.8313142857142857,
      "grad_norm": 1.147274374961853,
      "learning_rate": 0.0004707837025030478,
      "loss": 2.7671782970428467,
      "step": 3637
    },
    {
      "epoch": 0.8315428571428571,
      "grad_norm": 1.1316319704055786,
      "learning_rate": 0.0004703611998684669,
      "loss": 2.383859634399414,
      "step": 3638
    },
    {
      "epoch": 0.8317714285714286,
      "grad_norm": 0.7317661643028259,
      "learning_rate": 0.0004699358565801304,
      "loss": 2.443634271621704,
      "step": 3639
    },
    {
      "epoch": 0.832,
      "grad_norm": 0.8634117245674133,
      "learning_rate": 0.00046950767812109506,
      "loss": 2.3526992797851562,
      "step": 3640
    },
    {
      "epoch": 0.8322285714285714,
      "grad_norm": 1.0810785293579102,
      "learning_rate": 0.0004690766700109659,
      "loss": 2.1130053997039795,
      "step": 3641
    },
    {
      "epoch": 0.8324571428571429,
      "grad_norm": 0.8252564668655396,
      "learning_rate": 0.0004686428378058244,
      "loss": 2.479208469390869,
      "step": 3642
    },
    {
      "epoch": 0.8326857142857143,
      "grad_norm": 1.5142585039138794,
      "learning_rate": 0.0004682061870981573,
      "loss": 2.639312744140625,
      "step": 3643
    },
    {
      "epoch": 0.8329142857142857,
      "grad_norm": 1.0622012615203857,
      "learning_rate": 0.0004677667235167844,
      "loss": 2.4088945388793945,
      "step": 3644
    },
    {
      "epoch": 0.8331428571428572,
      "grad_norm": 1.1019221544265747,
      "learning_rate": 0.00046732445272678594,
      "loss": 2.4368975162506104,
      "step": 3645
    },
    {
      "epoch": 0.8333714285714285,
      "grad_norm": 0.9757528901100159,
      "learning_rate": 0.0004668793804294294,
      "loss": 2.2880663871765137,
      "step": 3646
    },
    {
      "epoch": 0.8336,
      "grad_norm": 1.148924708366394,
      "learning_rate": 0.00046643151236209644,
      "loss": 2.5474820137023926,
      "step": 3647
    },
    {
      "epoch": 0.8338285714285715,
      "grad_norm": 0.9878535866737366,
      "learning_rate": 0.0004659808542982088,
      "loss": 2.617013454437256,
      "step": 3648
    },
    {
      "epoch": 0.8340571428571428,
      "grad_norm": 0.9210923910140991,
      "learning_rate": 0.00046552741204715354,
      "loss": 2.6396937370300293,
      "step": 3649
    },
    {
      "epoch": 0.8342857142857143,
      "grad_norm": 0.9137697219848633,
      "learning_rate": 0.0004650711914542084,
      "loss": 2.2838027477264404,
      "step": 3650
    },
    {
      "epoch": 0.8345142857142858,
      "grad_norm": 1.265670895576477,
      "learning_rate": 0.0004646121984004665,
      "loss": 2.849912166595459,
      "step": 3651
    },
    {
      "epoch": 0.8347428571428571,
      "grad_norm": 0.9444496035575867,
      "learning_rate": 0.00046415043880276065,
      "loss": 2.447004556655884,
      "step": 3652
    },
    {
      "epoch": 0.8349714285714286,
      "grad_norm": 1.0097342729568481,
      "learning_rate": 0.00046368591861358665,
      "loss": 2.4880290031433105,
      "step": 3653
    },
    {
      "epoch": 0.8352,
      "grad_norm": 1.2965947389602661,
      "learning_rate": 0.00046321864382102687,
      "loss": 2.483927011489868,
      "step": 3654
    },
    {
      "epoch": 0.8354285714285714,
      "grad_norm": 1.4455236196517944,
      "learning_rate": 0.000462748620448673,
      "loss": 2.664431095123291,
      "step": 3655
    },
    {
      "epoch": 0.8356571428571429,
      "grad_norm": 1.0472943782806396,
      "learning_rate": 0.00046227585455554845,
      "loss": 2.4126288890838623,
      "step": 3656
    },
    {
      "epoch": 0.8358857142857142,
      "grad_norm": 0.752182126045227,
      "learning_rate": 0.00046180035223603,
      "loss": 2.605991840362549,
      "step": 3657
    },
    {
      "epoch": 0.8361142857142857,
      "grad_norm": 1.1797536611557007,
      "learning_rate": 0.0004613221196197698,
      "loss": 2.705195665359497,
      "step": 3658
    },
    {
      "epoch": 0.8363428571428572,
      "grad_norm": 1.4116535186767578,
      "learning_rate": 0.0004608411628716156,
      "loss": 2.8704867362976074,
      "step": 3659
    },
    {
      "epoch": 0.8365714285714285,
      "grad_norm": 1.1127145290374756,
      "learning_rate": 0.0004603574881915318,
      "loss": 2.780813217163086,
      "step": 3660
    },
    {
      "epoch": 0.8368,
      "grad_norm": 1.0819095373153687,
      "learning_rate": 0.0004598711018145193,
      "loss": 2.6210758686065674,
      "step": 3661
    },
    {
      "epoch": 0.8370285714285715,
      "grad_norm": 3.278501033782959,
      "learning_rate": 0.00045938201001053546,
      "loss": 2.168306827545166,
      "step": 3662
    },
    {
      "epoch": 0.8372571428571428,
      "grad_norm": 0.7496480345726013,
      "learning_rate": 0.0004588902190844127,
      "loss": 2.585871696472168,
      "step": 3663
    },
    {
      "epoch": 0.8374857142857143,
      "grad_norm": 1.1257266998291016,
      "learning_rate": 0.0004583957353757778,
      "loss": 2.7146990299224854,
      "step": 3664
    },
    {
      "epoch": 0.8377142857142857,
      "grad_norm": 1.08980393409729,
      "learning_rate": 0.0004578985652589697,
      "loss": 2.420363187789917,
      "step": 3665
    },
    {
      "epoch": 0.8379428571428571,
      "grad_norm": 1.0278162956237793,
      "learning_rate": 0.00045739871514295785,
      "loss": 2.4809913635253906,
      "step": 3666
    },
    {
      "epoch": 0.8381714285714286,
      "grad_norm": 1.5557708740234375,
      "learning_rate": 0.0004568961914712589,
      "loss": 2.791402816772461,
      "step": 3667
    },
    {
      "epoch": 0.8384,
      "grad_norm": 1.49166738986969,
      "learning_rate": 0.0004563910007218545,
      "loss": 2.4340600967407227,
      "step": 3668
    },
    {
      "epoch": 0.8386285714285714,
      "grad_norm": 1.0001918077468872,
      "learning_rate": 0.00045588314940710683,
      "loss": 2.685807228088379,
      "step": 3669
    },
    {
      "epoch": 0.8388571428571429,
      "grad_norm": 1.4084997177124023,
      "learning_rate": 0.00045537264407367564,
      "loss": 2.2205097675323486,
      "step": 3670
    },
    {
      "epoch": 0.8390857142857143,
      "grad_norm": 0.828496515750885,
      "learning_rate": 0.000454859491302433,
      "loss": 2.6396796703338623,
      "step": 3671
    },
    {
      "epoch": 0.8393142857142857,
      "grad_norm": 1.6874476671218872,
      "learning_rate": 0.00045434369770837893,
      "loss": 2.7649970054626465,
      "step": 3672
    },
    {
      "epoch": 0.8395428571428571,
      "grad_norm": 0.9436206817626953,
      "learning_rate": 0.0004538252699405561,
      "loss": 2.500178813934326,
      "step": 3673
    },
    {
      "epoch": 0.8397714285714286,
      "grad_norm": 1.1136940717697144,
      "learning_rate": 0.000453304214681964,
      "loss": 2.531984329223633,
      "step": 3674
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.8992746472358704,
      "learning_rate": 0.0004527805386494728,
      "loss": 2.3695268630981445,
      "step": 3675
    },
    {
      "epoch": 0.8402285714285714,
      "grad_norm": 0.8919164538383484,
      "learning_rate": 0.0004522542485937369,
      "loss": 2.544405698776245,
      "step": 3676
    },
    {
      "epoch": 0.8404571428571429,
      "grad_norm": 0.8945551514625549,
      "learning_rate": 0.00045172535129910763,
      "loss": 2.743925094604492,
      "step": 3677
    },
    {
      "epoch": 0.8406857142857143,
      "grad_norm": 1.0279937982559204,
      "learning_rate": 0.00045119385358354615,
      "loss": 2.377377510070801,
      "step": 3678
    },
    {
      "epoch": 0.8409142857142857,
      "grad_norm": 1.0665478706359863,
      "learning_rate": 0.00045065976229853534,
      "loss": 2.5833821296691895,
      "step": 3679
    },
    {
      "epoch": 0.8411428571428572,
      "grad_norm": 0.8578456044197083,
      "learning_rate": 0.0004501230843289914,
      "loss": 2.670318603515625,
      "step": 3680
    },
    {
      "epoch": 0.8413714285714285,
      "grad_norm": 0.9749974012374878,
      "learning_rate": 0.00044958382659317534,
      "loss": 2.1335558891296387,
      "step": 3681
    },
    {
      "epoch": 0.8416,
      "grad_norm": 1.017915964126587,
      "learning_rate": 0.00044904199604260377,
      "loss": 2.533872604370117,
      "step": 3682
    },
    {
      "epoch": 0.8418285714285715,
      "grad_norm": 0.9279276728630066,
      "learning_rate": 0.0004484975996619589,
      "loss": 2.2260611057281494,
      "step": 3683
    },
    {
      "epoch": 0.8420571428571428,
      "grad_norm": 1.005971074104309,
      "learning_rate": 0.0004479506444689989,
      "loss": 2.417163610458374,
      "step": 3684
    },
    {
      "epoch": 0.8422857142857143,
      "grad_norm": 1.1158115863800049,
      "learning_rate": 0.00044740113751446764,
      "loss": 2.4975156784057617,
      "step": 3685
    },
    {
      "epoch": 0.8425142857142857,
      "grad_norm": 1.3966678380966187,
      "learning_rate": 0.00044684908588200304,
      "loss": 2.654186725616455,
      "step": 3686
    },
    {
      "epoch": 0.8427428571428571,
      "grad_norm": 1.0246710777282715,
      "learning_rate": 0.0004462944966880464,
      "loss": 2.3696541786193848,
      "step": 3687
    },
    {
      "epoch": 0.8429714285714286,
      "grad_norm": 0.9302844405174255,
      "learning_rate": 0.00044573737708175057,
      "loss": 2.566072463989258,
      "step": 3688
    },
    {
      "epoch": 0.8432,
      "grad_norm": 1.3675822019577026,
      "learning_rate": 0.0004451777342448874,
      "loss": 2.661126136779785,
      "step": 3689
    },
    {
      "epoch": 0.8434285714285714,
      "grad_norm": 1.4229824542999268,
      "learning_rate": 0.00044461557539175587,
      "loss": 2.6596288681030273,
      "step": 3690
    },
    {
      "epoch": 0.8436571428571429,
      "grad_norm": 1.1317540407180786,
      "learning_rate": 0.0004440509077690883,
      "loss": 2.8628368377685547,
      "step": 3691
    },
    {
      "epoch": 0.8438857142857142,
      "grad_norm": 0.9393765330314636,
      "learning_rate": 0.00044348373865595736,
      "loss": 2.468172311782837,
      "step": 3692
    },
    {
      "epoch": 0.8441142857142857,
      "grad_norm": 1.015507459640503,
      "learning_rate": 0.0004429140753636824,
      "loss": 2.5025901794433594,
      "step": 3693
    },
    {
      "epoch": 0.8443428571428572,
      "grad_norm": 0.7718920111656189,
      "learning_rate": 0.00044234192523573467,
      "loss": 2.870743989944458,
      "step": 3694
    },
    {
      "epoch": 0.8445714285714285,
      "grad_norm": 1.3608908653259277,
      "learning_rate": 0.0004417672956476433,
      "loss": 2.4718871116638184,
      "step": 3695
    },
    {
      "epoch": 0.8448,
      "grad_norm": 1.1137486696243286,
      "learning_rate": 0.0004411901940068996,
      "loss": 2.2912793159484863,
      "step": 3696
    },
    {
      "epoch": 0.8450285714285715,
      "grad_norm": 1.0257718563079834,
      "learning_rate": 0.000440610627752862,
      "loss": 2.445267677307129,
      "step": 3697
    },
    {
      "epoch": 0.8452571428571428,
      "grad_norm": 0.9120813608169556,
      "learning_rate": 0.0004400286043566599,
      "loss": 2.7082624435424805,
      "step": 3698
    },
    {
      "epoch": 0.8454857142857143,
      "grad_norm": 1.2574001550674438,
      "learning_rate": 0.00043944413132109765,
      "loss": 2.4756112098693848,
      "step": 3699
    },
    {
      "epoch": 0.8457142857142858,
      "grad_norm": 1.5910638570785522,
      "learning_rate": 0.0004388572161805575,
      "loss": 2.522878646850586,
      "step": 3700
    },
    {
      "epoch": 0.8459428571428571,
      "grad_norm": 1.5524553060531616,
      "learning_rate": 0.00043826786650090276,
      "loss": 2.454291820526123,
      "step": 3701
    },
    {
      "epoch": 0.8461714285714286,
      "grad_norm": 0.847944438457489,
      "learning_rate": 0.0004376760898793799,
      "loss": 2.5324089527130127,
      "step": 3702
    },
    {
      "epoch": 0.8464,
      "grad_norm": 0.9555423855781555,
      "learning_rate": 0.00043708189394452113,
      "loss": 2.32401967048645,
      "step": 3703
    },
    {
      "epoch": 0.8466285714285714,
      "grad_norm": 1.3277225494384766,
      "learning_rate": 0.00043648528635604556,
      "loss": 2.708329439163208,
      "step": 3704
    },
    {
      "epoch": 0.8468571428571429,
      "grad_norm": 0.910297155380249,
      "learning_rate": 0.00043588627480476084,
      "loss": 2.2608978748321533,
      "step": 3705
    },
    {
      "epoch": 0.8470857142857143,
      "grad_norm": 0.9326587319374084,
      "learning_rate": 0.00043528486701246375,
      "loss": 2.408083200454712,
      "step": 3706
    },
    {
      "epoch": 0.8473142857142857,
      "grad_norm": 1.151759147644043,
      "learning_rate": 0.0004346810707318408,
      "loss": 2.541762351989746,
      "step": 3707
    },
    {
      "epoch": 0.8475428571428572,
      "grad_norm": 1.4562278985977173,
      "learning_rate": 0.0004340748937463683,
      "loss": 2.6226134300231934,
      "step": 3708
    },
    {
      "epoch": 0.8477714285714286,
      "grad_norm": 0.9256459474563599,
      "learning_rate": 0.000433466343870212,
      "loss": 2.714353322982788,
      "step": 3709
    },
    {
      "epoch": 0.848,
      "grad_norm": 1.0318273305892944,
      "learning_rate": 0.0004328554289481262,
      "loss": 2.8134424686431885,
      "step": 3710
    },
    {
      "epoch": 0.8482285714285714,
      "grad_norm": 0.8530421853065491,
      "learning_rate": 0.00043224215685535287,
      "loss": 2.662541389465332,
      "step": 3711
    },
    {
      "epoch": 0.8484571428571429,
      "grad_norm": 1.1564353704452515,
      "learning_rate": 0.00043162653549752014,
      "loss": 2.5817513465881348,
      "step": 3712
    },
    {
      "epoch": 0.8486857142857143,
      "grad_norm": 0.9361862540245056,
      "learning_rate": 0.00043100857281053997,
      "loss": 2.3645479679107666,
      "step": 3713
    },
    {
      "epoch": 0.8489142857142857,
      "grad_norm": 0.94491046667099,
      "learning_rate": 0.0004303882767605065,
      "loss": 2.455913543701172,
      "step": 3714
    },
    {
      "epoch": 0.8491428571428571,
      "grad_norm": 1.1864941120147705,
      "learning_rate": 0.0004297656553435928,
      "loss": 2.5820446014404297,
      "step": 3715
    },
    {
      "epoch": 0.8493714285714286,
      "grad_norm": 1.0270113945007324,
      "learning_rate": 0.0004291407165859481,
      "loss": 2.641120433807373,
      "step": 3716
    },
    {
      "epoch": 0.8496,
      "grad_norm": 1.029537558555603,
      "learning_rate": 0.00042851346854359405,
      "loss": 2.666703939437866,
      "step": 3717
    },
    {
      "epoch": 0.8498285714285714,
      "grad_norm": 0.8838146924972534,
      "learning_rate": 0.0004278839193023214,
      "loss": 2.786262035369873,
      "step": 3718
    },
    {
      "epoch": 0.8500571428571428,
      "grad_norm": 0.924292802810669,
      "learning_rate": 0.000427252076977585,
      "loss": 2.665875196456909,
      "step": 3719
    },
    {
      "epoch": 0.8502857142857143,
      "grad_norm": 1.2361332178115845,
      "learning_rate": 0.00042661794971440007,
      "loss": 2.446784019470215,
      "step": 3720
    },
    {
      "epoch": 0.8505142857142857,
      "grad_norm": 0.9689137935638428,
      "learning_rate": 0.00042598154568723626,
      "loss": 2.711400032043457,
      "step": 3721
    },
    {
      "epoch": 0.8507428571428571,
      "grad_norm": 0.8671380281448364,
      "learning_rate": 0.00042534287309991316,
      "loss": 2.0124285221099854,
      "step": 3722
    },
    {
      "epoch": 0.8509714285714286,
      "grad_norm": 1.0511078834533691,
      "learning_rate": 0.0004247019401854938,
      "loss": 2.291045665740967,
      "step": 3723
    },
    {
      "epoch": 0.8512,
      "grad_norm": 1.1492862701416016,
      "learning_rate": 0.0004240587552061792,
      "loss": 2.4300272464752197,
      "step": 3724
    },
    {
      "epoch": 0.8514285714285714,
      "grad_norm": 0.871580958366394,
      "learning_rate": 0.00042341332645320126,
      "loss": 2.532348394393921,
      "step": 3725
    },
    {
      "epoch": 0.8516571428571429,
      "grad_norm": 0.8765734434127808,
      "learning_rate": 0.0004227656622467162,
      "loss": 2.180854320526123,
      "step": 3726
    },
    {
      "epoch": 0.8518857142857142,
      "grad_norm": 0.9441112875938416,
      "learning_rate": 0.00042211577093569726,
      "loss": 2.5541467666625977,
      "step": 3727
    },
    {
      "epoch": 0.8521142857142857,
      "grad_norm": 1.0676400661468506,
      "learning_rate": 0.000421463660897827,
      "loss": 2.768430233001709,
      "step": 3728
    },
    {
      "epoch": 0.8523428571428572,
      "grad_norm": 1.0490084886550903,
      "learning_rate": 0.00042080934053938956,
      "loss": 2.7349066734313965,
      "step": 3729
    },
    {
      "epoch": 0.8525714285714285,
      "grad_norm": 1.1400868892669678,
      "learning_rate": 0.00042015281829516185,
      "loss": 2.4057934284210205,
      "step": 3730
    },
    {
      "epoch": 0.8528,
      "grad_norm": 0.9881770610809326,
      "learning_rate": 0.00041949410262830523,
      "loss": 2.531019449234009,
      "step": 3731
    },
    {
      "epoch": 0.8530285714285715,
      "grad_norm": 1.0050491094589233,
      "learning_rate": 0.0004188332020302561,
      "loss": 2.242699146270752,
      "step": 3732
    },
    {
      "epoch": 0.8532571428571428,
      "grad_norm": 1.2203141450881958,
      "learning_rate": 0.0004181701250206168,
      "loss": 2.7151589393615723,
      "step": 3733
    },
    {
      "epoch": 0.8534857142857143,
      "grad_norm": 0.9660093784332275,
      "learning_rate": 0.00041750488014704527,
      "loss": 2.4097678661346436,
      "step": 3734
    },
    {
      "epoch": 0.8537142857142858,
      "grad_norm": 1.3167186975479126,
      "learning_rate": 0.00041683747598514554,
      "loss": 2.5762643814086914,
      "step": 3735
    },
    {
      "epoch": 0.8539428571428571,
      "grad_norm": 1.230282187461853,
      "learning_rate": 0.00041616792113835646,
      "loss": 2.3814353942871094,
      "step": 3736
    },
    {
      "epoch": 0.8541714285714286,
      "grad_norm": 1.2041822671890259,
      "learning_rate": 0.0004154962242378413,
      "loss": 2.4705958366394043,
      "step": 3737
    },
    {
      "epoch": 0.8544,
      "grad_norm": 0.8807657957077026,
      "learning_rate": 0.0004148223939423763,
      "loss": 2.586179733276367,
      "step": 3738
    },
    {
      "epoch": 0.8546285714285714,
      "grad_norm": 1.174916386604309,
      "learning_rate": 0.0004141464389382391,
      "loss": 2.088823080062866,
      "step": 3739
    },
    {
      "epoch": 0.8548571428571429,
      "grad_norm": 1.0929746627807617,
      "learning_rate": 0.0004134683679390968,
      "loss": 2.780383825302124,
      "step": 3740
    },
    {
      "epoch": 0.8550857142857143,
      "grad_norm": 1.6432968378067017,
      "learning_rate": 0.0004127881896858934,
      "loss": 2.5330824851989746,
      "step": 3741
    },
    {
      "epoch": 0.8553142857142857,
      "grad_norm": 0.9510151147842407,
      "learning_rate": 0.0004121059129467374,
      "loss": 2.742108106613159,
      "step": 3742
    },
    {
      "epoch": 0.8555428571428572,
      "grad_norm": 0.9808242917060852,
      "learning_rate": 0.0004114215465167888,
      "loss": 2.315603733062744,
      "step": 3743
    },
    {
      "epoch": 0.8557714285714285,
      "grad_norm": 0.821926474571228,
      "learning_rate": 0.00041073509921814527,
      "loss": 2.4862194061279297,
      "step": 3744
    },
    {
      "epoch": 0.856,
      "grad_norm": 1.0754958391189575,
      "learning_rate": 0.00041004657989972905,
      "loss": 2.9035706520080566,
      "step": 3745
    },
    {
      "epoch": 0.8562285714285714,
      "grad_norm": 1.2073191404342651,
      "learning_rate": 0.00040935599743717243,
      "loss": 2.6940712928771973,
      "step": 3746
    },
    {
      "epoch": 0.8564571428571428,
      "grad_norm": 1.0464332103729248,
      "learning_rate": 0.00040866336073270356,
      "loss": 2.5602803230285645,
      "step": 3747
    },
    {
      "epoch": 0.8566857142857143,
      "grad_norm": 1.0192575454711914,
      "learning_rate": 0.0004079686787150315,
      "loss": 2.3645424842834473,
      "step": 3748
    },
    {
      "epoch": 0.8569142857142857,
      "grad_norm": 0.8620522022247314,
      "learning_rate": 0.0004072719603392313,
      "loss": 2.650883674621582,
      "step": 3749
    },
    {
      "epoch": 0.8571428571428571,
      "grad_norm": 0.9424662590026855,
      "learning_rate": 0.00040657321458662855,
      "loss": 2.4095749855041504,
      "step": 3750
    },
    {
      "epoch": 0.8573714285714286,
      "grad_norm": 1.1395097970962524,
      "learning_rate": 0.0004058724504646834,
      "loss": 2.578301191329956,
      "step": 3751
    },
    {
      "epoch": 0.8576,
      "grad_norm": 0.8617632389068604,
      "learning_rate": 0.0004051696770068748,
      "loss": 2.597322940826416,
      "step": 3752
    },
    {
      "epoch": 0.8578285714285714,
      "grad_norm": 0.7882718443870544,
      "learning_rate": 0.0004044649032725836,
      "loss": 2.620663642883301,
      "step": 3753
    },
    {
      "epoch": 0.8580571428571429,
      "grad_norm": 0.9949784278869629,
      "learning_rate": 0.0004037581383469763,
      "loss": 2.6337175369262695,
      "step": 3754
    },
    {
      "epoch": 0.8582857142857143,
      "grad_norm": 0.8993661999702454,
      "learning_rate": 0.00040304939134088726,
      "loss": 2.4311277866363525,
      "step": 3755
    },
    {
      "epoch": 0.8585142857142857,
      "grad_norm": 0.8750917315483093,
      "learning_rate": 0.00040233867139070206,
      "loss": 2.326312303543091,
      "step": 3756
    },
    {
      "epoch": 0.8587428571428571,
      "grad_norm": 1.0355907678604126,
      "learning_rate": 0.000401625987658239,
      "loss": 2.7356557846069336,
      "step": 3757
    },
    {
      "epoch": 0.8589714285714286,
      "grad_norm": 1.2479511499404907,
      "learning_rate": 0.0004009113493306314,
      "loss": 2.816699981689453,
      "step": 3758
    },
    {
      "epoch": 0.8592,
      "grad_norm": 0.8167040944099426,
      "learning_rate": 0.0004001947656202092,
      "loss": 2.340200185775757,
      "step": 3759
    },
    {
      "epoch": 0.8594285714285714,
      "grad_norm": 1.3641319274902344,
      "learning_rate": 0.0003994762457643797,
      "loss": 2.6132256984710693,
      "step": 3760
    },
    {
      "epoch": 0.8596571428571429,
      "grad_norm": 0.9177221655845642,
      "learning_rate": 0.0003987557990255093,
      "loss": 2.3341891765594482,
      "step": 3761
    },
    {
      "epoch": 0.8598857142857143,
      "grad_norm": 1.242415428161621,
      "learning_rate": 0.0003980334346908033,
      "loss": 2.358515977859497,
      "step": 3762
    },
    {
      "epoch": 0.8601142857142857,
      "grad_norm": 0.8579970598220825,
      "learning_rate": 0.0003973091620721868,
      "loss": 2.470602035522461,
      "step": 3763
    },
    {
      "epoch": 0.8603428571428572,
      "grad_norm": 1.031916856765747,
      "learning_rate": 0.0003965829905061841,
      "loss": 2.4790468215942383,
      "step": 3764
    },
    {
      "epoch": 0.8605714285714285,
      "grad_norm": 0.8457033634185791,
      "learning_rate": 0.00039585492935379895,
      "loss": 2.2959794998168945,
      "step": 3765
    },
    {
      "epoch": 0.8608,
      "grad_norm": 0.8820912837982178,
      "learning_rate": 0.00039512498800039336,
      "loss": 2.563701629638672,
      "step": 3766
    },
    {
      "epoch": 0.8610285714285715,
      "grad_norm": 1.0184649229049683,
      "learning_rate": 0.0003943931758555669,
      "loss": 2.5384957790374756,
      "step": 3767
    },
    {
      "epoch": 0.8612571428571428,
      "grad_norm": 0.8972774147987366,
      "learning_rate": 0.0003936595023530353,
      "loss": 2.385929822921753,
      "step": 3768
    },
    {
      "epoch": 0.8614857142857143,
      "grad_norm": 1.0236729383468628,
      "learning_rate": 0.00039292397695050887,
      "loss": 2.4815356731414795,
      "step": 3769
    },
    {
      "epoch": 0.8617142857142858,
      "grad_norm": 1.0012543201446533,
      "learning_rate": 0.0003921866091295705,
      "loss": 2.5879006385803223,
      "step": 3770
    },
    {
      "epoch": 0.8619428571428571,
      "grad_norm": 0.9029208421707153,
      "learning_rate": 0.00039144740839555366,
      "loss": 2.629206418991089,
      "step": 3771
    },
    {
      "epoch": 0.8621714285714286,
      "grad_norm": 1.0773470401763916,
      "learning_rate": 0.00039070638427741956,
      "loss": 2.6445751190185547,
      "step": 3772
    },
    {
      "epoch": 0.8624,
      "grad_norm": 0.9446911215782166,
      "learning_rate": 0.00038996354632763467,
      "loss": 2.740169048309326,
      "step": 3773
    },
    {
      "epoch": 0.8626285714285714,
      "grad_norm": 1.0810444355010986,
      "learning_rate": 0.000389218904122047,
      "loss": 2.328369140625,
      "step": 3774
    },
    {
      "epoch": 0.8628571428571429,
      "grad_norm": 0.8575284481048584,
      "learning_rate": 0.00038847246725976355,
      "loss": 2.556790351867676,
      "step": 3775
    },
    {
      "epoch": 0.8630857142857142,
      "grad_norm": 1.0386083126068115,
      "learning_rate": 0.0003877242453630256,
      "loss": 2.5671603679656982,
      "step": 3776
    },
    {
      "epoch": 0.8633142857142857,
      "grad_norm": 0.8255072236061096,
      "learning_rate": 0.00038697424807708546,
      "loss": 2.5294878482818604,
      "step": 3777
    },
    {
      "epoch": 0.8635428571428572,
      "grad_norm": 0.9087710380554199,
      "learning_rate": 0.0003862224850700816,
      "loss": 2.4265851974487305,
      "step": 3778
    },
    {
      "epoch": 0.8637714285714285,
      "grad_norm": 0.923771858215332,
      "learning_rate": 0.00038546896603291434,
      "loss": 2.6910572052001953,
      "step": 3779
    },
    {
      "epoch": 0.864,
      "grad_norm": 0.8677828907966614,
      "learning_rate": 0.0003847137006791206,
      "loss": 2.45001220703125,
      "step": 3780
    },
    {
      "epoch": 0.8642285714285715,
      "grad_norm": 0.9780474901199341,
      "learning_rate": 0.00038395669874474915,
      "loss": 2.3733930587768555,
      "step": 3781
    },
    {
      "epoch": 0.8644571428571428,
      "grad_norm": 1.0935029983520508,
      "learning_rate": 0.00038319796998823475,
      "loss": 2.2450413703918457,
      "step": 3782
    },
    {
      "epoch": 0.8646857142857143,
      "grad_norm": 1.2215365171432495,
      "learning_rate": 0.0003824375241902723,
      "loss": 2.52746319770813,
      "step": 3783
    },
    {
      "epoch": 0.8649142857142857,
      "grad_norm": 1.0444327592849731,
      "learning_rate": 0.00038167537115369116,
      "loss": 2.6679482460021973,
      "step": 3784
    },
    {
      "epoch": 0.8651428571428571,
      "grad_norm": 0.9340086579322815,
      "learning_rate": 0.00038091152070332833,
      "loss": 2.6821844577789307,
      "step": 3785
    },
    {
      "epoch": 0.8653714285714286,
      "grad_norm": 0.9361206293106079,
      "learning_rate": 0.0003801459826859022,
      "loss": 2.3384792804718018,
      "step": 3786
    },
    {
      "epoch": 0.8656,
      "grad_norm": 0.9260632395744324,
      "learning_rate": 0.00037937876696988505,
      "loss": 2.606501579284668,
      "step": 3787
    },
    {
      "epoch": 0.8658285714285714,
      "grad_norm": 0.9608365297317505,
      "learning_rate": 0.0003786098834453766,
      "loss": 2.3597469329833984,
      "step": 3788
    },
    {
      "epoch": 0.8660571428571429,
      "grad_norm": 0.8686981797218323,
      "learning_rate": 0.00037783934202397597,
      "loss": 2.550694227218628,
      "step": 3789
    },
    {
      "epoch": 0.8662857142857143,
      "grad_norm": 0.87773597240448,
      "learning_rate": 0.00037706715263865397,
      "loss": 2.450477123260498,
      "step": 3790
    },
    {
      "epoch": 0.8665142857142857,
      "grad_norm": 0.9927385449409485,
      "learning_rate": 0.0003762933252436253,
      "loss": 2.4100446701049805,
      "step": 3791
    },
    {
      "epoch": 0.8667428571428571,
      "grad_norm": 1.1382428407669067,
      "learning_rate": 0.00037551786981421985,
      "loss": 2.650754928588867,
      "step": 3792
    },
    {
      "epoch": 0.8669714285714286,
      "grad_norm": 1.2014020681381226,
      "learning_rate": 0.00037474079634675463,
      "loss": 2.6053647994995117,
      "step": 3793
    },
    {
      "epoch": 0.8672,
      "grad_norm": 1.2577085494995117,
      "learning_rate": 0.00037396211485840445,
      "loss": 2.536386489868164,
      "step": 3794
    },
    {
      "epoch": 0.8674285714285714,
      "grad_norm": 0.9396008253097534,
      "learning_rate": 0.0003731818353870729,
      "loss": 2.4606504440307617,
      "step": 3795
    },
    {
      "epoch": 0.8676571428571429,
      "grad_norm": 0.8783560991287231,
      "learning_rate": 0.00037239996799126314,
      "loss": 2.76385760307312,
      "step": 3796
    },
    {
      "epoch": 0.8678857142857143,
      "grad_norm": 0.8167362809181213,
      "learning_rate": 0.00037161652274994794,
      "loss": 2.1169447898864746,
      "step": 3797
    },
    {
      "epoch": 0.8681142857142857,
      "grad_norm": 1.4646674394607544,
      "learning_rate": 0.0003708315097624401,
      "loss": 2.3064162731170654,
      "step": 3798
    },
    {
      "epoch": 0.8683428571428572,
      "grad_norm": 0.7623499631881714,
      "learning_rate": 0.00037004493914826196,
      "loss": 2.4147789478302,
      "step": 3799
    },
    {
      "epoch": 0.8685714285714285,
      "grad_norm": 1.076818823814392,
      "learning_rate": 0.0003692568210470152,
      "loss": 2.541144847869873,
      "step": 3800
    },
    {
      "epoch": 0.8688,
      "grad_norm": 0.8416025638580322,
      "learning_rate": 0.00036846716561824967,
      "loss": 2.7181501388549805,
      "step": 3801
    },
    {
      "epoch": 0.8690285714285715,
      "grad_norm": 0.8453372716903687,
      "learning_rate": 0.0003676759830413332,
      "loss": 2.25830078125,
      "step": 3802
    },
    {
      "epoch": 0.8692571428571428,
      "grad_norm": 0.8744320273399353,
      "learning_rate": 0.00036688328351531954,
      "loss": 2.3588924407958984,
      "step": 3803
    },
    {
      "epoch": 0.8694857142857143,
      "grad_norm": 0.9895479083061218,
      "learning_rate": 0.00036608907725881745,
      "loss": 2.119791030883789,
      "step": 3804
    },
    {
      "epoch": 0.8697142857142857,
      "grad_norm": 0.7937715649604797,
      "learning_rate": 0.00036529337450985893,
      "loss": 2.5799479484558105,
      "step": 3805
    },
    {
      "epoch": 0.8699428571428571,
      "grad_norm": 0.8615305423736572,
      "learning_rate": 0.0003644961855257669,
      "loss": 2.7783336639404297,
      "step": 3806
    },
    {
      "epoch": 0.8701714285714286,
      "grad_norm": 0.9604050517082214,
      "learning_rate": 0.00036369752058302326,
      "loss": 2.437610387802124,
      "step": 3807
    },
    {
      "epoch": 0.8704,
      "grad_norm": 0.8239616751670837,
      "learning_rate": 0.0003628973899771363,
      "loss": 2.5707833766937256,
      "step": 3808
    },
    {
      "epoch": 0.8706285714285714,
      "grad_norm": 0.8653541207313538,
      "learning_rate": 0.0003620958040225081,
      "loss": 2.3925960063934326,
      "step": 3809
    },
    {
      "epoch": 0.8708571428571429,
      "grad_norm": 1.2359319925308228,
      "learning_rate": 0.0003612927730523014,
      "loss": 2.567819118499756,
      "step": 3810
    },
    {
      "epoch": 0.8710857142857142,
      "grad_norm": 0.7460139989852905,
      "learning_rate": 0.00036048830741830675,
      "loss": 2.462766408920288,
      "step": 3811
    },
    {
      "epoch": 0.8713142857142857,
      "grad_norm": 0.677042543888092,
      "learning_rate": 0.0003596824174908083,
      "loss": 2.391724109649658,
      "step": 3812
    },
    {
      "epoch": 0.8715428571428572,
      "grad_norm": 0.9056562781333923,
      "learning_rate": 0.0003588751136584512,
      "loss": 2.610745906829834,
      "step": 3813
    },
    {
      "epoch": 0.8717714285714285,
      "grad_norm": 0.8798251748085022,
      "learning_rate": 0.0003580664063281066,
      "loss": 2.822805404663086,
      "step": 3814
    },
    {
      "epoch": 0.872,
      "grad_norm": 1.054552435874939,
      "learning_rate": 0.0003572563059247384,
      "loss": 3.0404911041259766,
      "step": 3815
    },
    {
      "epoch": 0.8722285714285715,
      "grad_norm": 0.865313708782196,
      "learning_rate": 0.0003564448228912682,
      "loss": 2.5239739418029785,
      "step": 3816
    },
    {
      "epoch": 0.8724571428571428,
      "grad_norm": 0.7919363379478455,
      "learning_rate": 0.00035563196768844096,
      "loss": 2.5544095039367676,
      "step": 3817
    },
    {
      "epoch": 0.8726857142857143,
      "grad_norm": 0.9072160720825195,
      "learning_rate": 0.00035481775079469047,
      "loss": 2.6043951511383057,
      "step": 3818
    },
    {
      "epoch": 0.8729142857142858,
      "grad_norm": 1.000880479812622,
      "learning_rate": 0.0003540021827060034,
      "loss": 2.3962087631225586,
      "step": 3819
    },
    {
      "epoch": 0.8731428571428571,
      "grad_norm": 0.9536701440811157,
      "learning_rate": 0.0003531852739357849,
      "loss": 2.4747633934020996,
      "step": 3820
    },
    {
      "epoch": 0.8733714285714286,
      "grad_norm": 0.7937532067298889,
      "learning_rate": 0.00035236703501472264,
      "loss": 2.489603042602539,
      "step": 3821
    },
    {
      "epoch": 0.8736,
      "grad_norm": 0.9851025938987732,
      "learning_rate": 0.0003515474764906511,
      "loss": 2.4766244888305664,
      "step": 3822
    },
    {
      "epoch": 0.8738285714285714,
      "grad_norm": 0.7430683970451355,
      "learning_rate": 0.00035072660892841566,
      "loss": 2.4776530265808105,
      "step": 3823
    },
    {
      "epoch": 0.8740571428571429,
      "grad_norm": 0.6643300652503967,
      "learning_rate": 0.00034990444290973634,
      "loss": 2.195923089981079,
      "step": 3824
    },
    {
      "epoch": 0.8742857142857143,
      "grad_norm": 1.0037946701049805,
      "learning_rate": 0.00034908098903307146,
      "loss": 2.4599857330322266,
      "step": 3825
    },
    {
      "epoch": 0.8745142857142857,
      "grad_norm": 0.9368773698806763,
      "learning_rate": 0.0003482562579134809,
      "loss": 2.2265686988830566,
      "step": 3826
    },
    {
      "epoch": 0.8747428571428572,
      "grad_norm": 1.0355628728866577,
      "learning_rate": 0.00034743026018248957,
      "loss": 2.605515241622925,
      "step": 3827
    },
    {
      "epoch": 0.8749714285714286,
      "grad_norm": 1.032542109489441,
      "learning_rate": 0.00034660300648794973,
      "loss": 2.699228048324585,
      "step": 3828
    },
    {
      "epoch": 0.8752,
      "grad_norm": 1.019685983657837,
      "learning_rate": 0.00034577450749390473,
      "loss": 2.6104342937469482,
      "step": 3829
    },
    {
      "epoch": 0.8754285714285714,
      "grad_norm": 1.0236766338348389,
      "learning_rate": 0.0003449447738804503,
      "loss": 2.498319149017334,
      "step": 3830
    },
    {
      "epoch": 0.8756571428571429,
      "grad_norm": 0.9182736873626709,
      "learning_rate": 0.00034411381634359803,
      "loss": 2.28558611869812,
      "step": 3831
    },
    {
      "epoch": 0.8758857142857143,
      "grad_norm": 1.1878662109375,
      "learning_rate": 0.0003432816455951367,
      "loss": 2.3725056648254395,
      "step": 3832
    },
    {
      "epoch": 0.8761142857142857,
      "grad_norm": 1.007745623588562,
      "learning_rate": 0.00034244827236249467,
      "loss": 2.363173246383667,
      "step": 3833
    },
    {
      "epoch": 0.8763428571428571,
      "grad_norm": 1.0789525508880615,
      "learning_rate": 0.00034161370738860114,
      "loss": 2.737518787384033,
      "step": 3834
    },
    {
      "epoch": 0.8765714285714286,
      "grad_norm": 1.010540246963501,
      "learning_rate": 0.00034077796143174824,
      "loss": 2.4555187225341797,
      "step": 3835
    },
    {
      "epoch": 0.8768,
      "grad_norm": 0.8281636834144592,
      "learning_rate": 0.0003399410452654518,
      "loss": 2.6006276607513428,
      "step": 3836
    },
    {
      "epoch": 0.8770285714285714,
      "grad_norm": 1.2354499101638794,
      "learning_rate": 0.00033910296967831267,
      "loss": 2.5937671661376953,
      "step": 3837
    },
    {
      "epoch": 0.8772571428571428,
      "grad_norm": 0.8433457016944885,
      "learning_rate": 0.000338263745473878,
      "loss": 2.461045742034912,
      "step": 3838
    },
    {
      "epoch": 0.8774857142857143,
      "grad_norm": 1.0897961854934692,
      "learning_rate": 0.0003374233834705013,
      "loss": 2.4234678745269775,
      "step": 3839
    },
    {
      "epoch": 0.8777142857142857,
      "grad_norm": 0.7300510406494141,
      "learning_rate": 0.00033658189450120344,
      "loss": 2.239096164703369,
      "step": 3840
    },
    {
      "epoch": 0.8779428571428571,
      "grad_norm": 0.8505579829216003,
      "learning_rate": 0.0003357392894135329,
      "loss": 2.4669852256774902,
      "step": 3841
    },
    {
      "epoch": 0.8781714285714286,
      "grad_norm": 0.7502102851867676,
      "learning_rate": 0.00033489557906942597,
      "loss": 2.570010185241699,
      "step": 3842
    },
    {
      "epoch": 0.8784,
      "grad_norm": 0.9644457101821899,
      "learning_rate": 0.00033405077434506676,
      "loss": 2.5414915084838867,
      "step": 3843
    },
    {
      "epoch": 0.8786285714285714,
      "grad_norm": 0.8327218294143677,
      "learning_rate": 0.00033320488613074666,
      "loss": 2.386323928833008,
      "step": 3844
    },
    {
      "epoch": 0.8788571428571429,
      "grad_norm": 0.9712873101234436,
      "learning_rate": 0.0003323579253307245,
      "loss": 2.6814656257629395,
      "step": 3845
    },
    {
      "epoch": 0.8790857142857142,
      "grad_norm": 0.9355666637420654,
      "learning_rate": 0.00033150990286308547,
      "loss": 2.200161933898926,
      "step": 3846
    },
    {
      "epoch": 0.8793142857142857,
      "grad_norm": 1.3419463634490967,
      "learning_rate": 0.0003306608296596008,
      "loss": 2.6125035285949707,
      "step": 3847
    },
    {
      "epoch": 0.8795428571428572,
      "grad_norm": 1.0062332153320312,
      "learning_rate": 0.00032981071666558656,
      "loss": 2.189284324645996,
      "step": 3848
    },
    {
      "epoch": 0.8797714285714285,
      "grad_norm": 1.0318398475646973,
      "learning_rate": 0.0003289595748397627,
      "loss": 2.604152202606201,
      "step": 3849
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.8423863649368286,
      "learning_rate": 0.00032810741515411156,
      "loss": 2.453373908996582,
      "step": 3850
    },
    {
      "epoch": 0.8802285714285715,
      "grad_norm": 1.0360782146453857,
      "learning_rate": 0.00032725424859373687,
      "loss": 2.357020378112793,
      "step": 3851
    },
    {
      "epoch": 0.8804571428571428,
      "grad_norm": 0.9003981351852417,
      "learning_rate": 0.00032640008615672175,
      "loss": 2.436629295349121,
      "step": 3852
    },
    {
      "epoch": 0.8806857142857143,
      "grad_norm": 1.0384968519210815,
      "learning_rate": 0.000325544938853987,
      "loss": 2.3721530437469482,
      "step": 3853
    },
    {
      "epoch": 0.8809142857142858,
      "grad_norm": 0.99648118019104,
      "learning_rate": 0.00032468881770914937,
      "loss": 2.3611679077148438,
      "step": 3854
    },
    {
      "epoch": 0.8811428571428571,
      "grad_norm": 0.958049476146698,
      "learning_rate": 0.0003238317337583792,
      "loss": 2.5305800437927246,
      "step": 3855
    },
    {
      "epoch": 0.8813714285714286,
      "grad_norm": 1.1329429149627686,
      "learning_rate": 0.0003229736980502584,
      "loss": 2.546220302581787,
      "step": 3856
    },
    {
      "epoch": 0.8816,
      "grad_norm": 0.7734110951423645,
      "learning_rate": 0.00032211472164563753,
      "loss": 2.5996313095092773,
      "step": 3857
    },
    {
      "epoch": 0.8818285714285714,
      "grad_norm": 0.9827597141265869,
      "learning_rate": 0.00032125481561749405,
      "loss": 2.4411139488220215,
      "step": 3858
    },
    {
      "epoch": 0.8820571428571429,
      "grad_norm": 1.3960602283477783,
      "learning_rate": 0.00032039399105078883,
      "loss": 2.590094804763794,
      "step": 3859
    },
    {
      "epoch": 0.8822857142857143,
      "grad_norm": 1.5087692737579346,
      "learning_rate": 0.0003195322590423235,
      "loss": 2.3681583404541016,
      "step": 3860
    },
    {
      "epoch": 0.8825142857142857,
      "grad_norm": 0.861089289188385,
      "learning_rate": 0.0003186696307005976,
      "loss": 2.5585203170776367,
      "step": 3861
    },
    {
      "epoch": 0.8827428571428572,
      "grad_norm": 1.3982013463974,
      "learning_rate": 0.00031780611714566503,
      "loss": 2.2885868549346924,
      "step": 3862
    },
    {
      "epoch": 0.8829714285714285,
      "grad_norm": 0.9931345582008362,
      "learning_rate": 0.0003169417295089911,
      "loss": 2.356800079345703,
      "step": 3863
    },
    {
      "epoch": 0.8832,
      "grad_norm": 1.016614556312561,
      "learning_rate": 0.0003160764789333084,
      "loss": 2.3262643814086914,
      "step": 3864
    },
    {
      "epoch": 0.8834285714285715,
      "grad_norm": 1.2012379169464111,
      "learning_rate": 0.0003152103765724743,
      "loss": 2.6564931869506836,
      "step": 3865
    },
    {
      "epoch": 0.8836571428571428,
      "grad_norm": 1.1062045097351074,
      "learning_rate": 0.0003143434335913256,
      "loss": 2.5818655490875244,
      "step": 3866
    },
    {
      "epoch": 0.8838857142857143,
      "grad_norm": 1.1803401708602905,
      "learning_rate": 0.0003134756611655362,
      "loss": 2.1837401390075684,
      "step": 3867
    },
    {
      "epoch": 0.8841142857142857,
      "grad_norm": 0.9568623900413513,
      "learning_rate": 0.00031260707048147164,
      "loss": 2.388822555541992,
      "step": 3868
    },
    {
      "epoch": 0.8843428571428571,
      "grad_norm": 0.8225594162940979,
      "learning_rate": 0.000311737672736046,
      "loss": 2.4347071647644043,
      "step": 3869
    },
    {
      "epoch": 0.8845714285714286,
      "grad_norm": 1.165232539176941,
      "learning_rate": 0.000310867479136577,
      "loss": 2.53195858001709,
      "step": 3870
    },
    {
      "epoch": 0.8848,
      "grad_norm": 1.216845989227295,
      "learning_rate": 0.0003099965009006415,
      "loss": 2.263510227203369,
      "step": 3871
    },
    {
      "epoch": 0.8850285714285714,
      "grad_norm": 0.9292066097259521,
      "learning_rate": 0.0003091247492559312,
      "loss": 2.340485095977783,
      "step": 3872
    },
    {
      "epoch": 0.8852571428571429,
      "grad_norm": 0.8757928609848022,
      "learning_rate": 0.0003082522354401076,
      "loss": 2.1381912231445312,
      "step": 3873
    },
    {
      "epoch": 0.8854857142857143,
      "grad_norm": 0.899720311164856,
      "learning_rate": 0.00030737897070065736,
      "loss": 2.3413100242614746,
      "step": 3874
    },
    {
      "epoch": 0.8857142857142857,
      "grad_norm": 0.8978779315948486,
      "learning_rate": 0.0003065049662947471,
      "loss": 2.7879831790924072,
      "step": 3875
    },
    {
      "epoch": 0.8859428571428571,
      "grad_norm": 1.1066943407058716,
      "learning_rate": 0.0003056302334890786,
      "loss": 2.568556785583496,
      "step": 3876
    },
    {
      "epoch": 0.8861714285714286,
      "grad_norm": 1.0388405323028564,
      "learning_rate": 0.0003047547835597432,
      "loss": 2.318281888961792,
      "step": 3877
    },
    {
      "epoch": 0.8864,
      "grad_norm": 0.9656121134757996,
      "learning_rate": 0.00030387862779207667,
      "loss": 2.6062397956848145,
      "step": 3878
    },
    {
      "epoch": 0.8866285714285714,
      "grad_norm": 0.8901590704917908,
      "learning_rate": 0.00030300177748051373,
      "loss": 2.1722731590270996,
      "step": 3879
    },
    {
      "epoch": 0.8868571428571429,
      "grad_norm": 0.7448124289512634,
      "learning_rate": 0.00030212424392844214,
      "loss": 2.3150644302368164,
      "step": 3880
    },
    {
      "epoch": 0.8870857142857143,
      "grad_norm": 1.0838897228240967,
      "learning_rate": 0.00030124603844805764,
      "loss": 2.3121936321258545,
      "step": 3881
    },
    {
      "epoch": 0.8873142857142857,
      "grad_norm": 0.9222500324249268,
      "learning_rate": 0.0003003671723602173,
      "loss": 2.4406399726867676,
      "step": 3882
    },
    {
      "epoch": 0.8875428571428572,
      "grad_norm": 0.8245007395744324,
      "learning_rate": 0.0002994876569942945,
      "loss": 2.320772171020508,
      "step": 3883
    },
    {
      "epoch": 0.8877714285714285,
      "grad_norm": 0.8307053446769714,
      "learning_rate": 0.0002986075036880319,
      "loss": 2.3858957290649414,
      "step": 3884
    },
    {
      "epoch": 0.888,
      "grad_norm": 1.0476921796798706,
      "learning_rate": 0.00029772672378739623,
      "loss": 2.2990474700927734,
      "step": 3885
    },
    {
      "epoch": 0.8882285714285715,
      "grad_norm": 0.9564490914344788,
      "learning_rate": 0.0002968453286464312,
      "loss": 2.4108712673187256,
      "step": 3886
    },
    {
      "epoch": 0.8884571428571428,
      "grad_norm": 0.9680846333503723,
      "learning_rate": 0.0002959633296271117,
      "loss": 2.5949347019195557,
      "step": 3887
    },
    {
      "epoch": 0.8886857142857143,
      "grad_norm": 0.8563591241836548,
      "learning_rate": 0.00029508073809919724,
      "loss": 2.1141653060913086,
      "step": 3888
    },
    {
      "epoch": 0.8889142857142858,
      "grad_norm": 0.9561415910720825,
      "learning_rate": 0.00029419756544008506,
      "loss": 2.4522576332092285,
      "step": 3889
    },
    {
      "epoch": 0.8891428571428571,
      "grad_norm": 0.6990514397621155,
      "learning_rate": 0.00029331382303466404,
      "loss": 2.1766438484191895,
      "step": 3890
    },
    {
      "epoch": 0.8893714285714286,
      "grad_norm": 0.7795295119285583,
      "learning_rate": 0.0002924295222751672,
      "loss": 2.3053486347198486,
      "step": 3891
    },
    {
      "epoch": 0.8896,
      "grad_norm": 1.013709306716919,
      "learning_rate": 0.00029154467456102566,
      "loss": 2.4958953857421875,
      "step": 3892
    },
    {
      "epoch": 0.8898285714285714,
      "grad_norm": 1.0947858095169067,
      "learning_rate": 0.00029065929129872095,
      "loss": 2.338221549987793,
      "step": 3893
    },
    {
      "epoch": 0.8900571428571429,
      "grad_norm": 0.9765961170196533,
      "learning_rate": 0.0002897733839016384,
      "loss": 2.480442523956299,
      "step": 3894
    },
    {
      "epoch": 0.8902857142857142,
      "grad_norm": 0.8938852548599243,
      "learning_rate": 0.00028888696378991986,
      "loss": 2.5504636764526367,
      "step": 3895
    },
    {
      "epoch": 0.8905142857142857,
      "grad_norm": 1.0532257556915283,
      "learning_rate": 0.00028800004239031685,
      "loss": 2.4839844703674316,
      "step": 3896
    },
    {
      "epoch": 0.8907428571428572,
      "grad_norm": 0.9302260875701904,
      "learning_rate": 0.0002871126311360424,
      "loss": 2.0072221755981445,
      "step": 3897
    },
    {
      "epoch": 0.8909714285714285,
      "grad_norm": 1.3753728866577148,
      "learning_rate": 0.00028622474146662466,
      "loss": 2.6038589477539062,
      "step": 3898
    },
    {
      "epoch": 0.8912,
      "grad_norm": 0.9702432155609131,
      "learning_rate": 0.00028533638482775876,
      "loss": 2.597771644592285,
      "step": 3899
    },
    {
      "epoch": 0.8914285714285715,
      "grad_norm": 0.922528088092804,
      "learning_rate": 0.0002844475726711595,
      "loss": 2.4003310203552246,
      "step": 3900
    },
    {
      "epoch": 0.8916571428571428,
      "grad_norm": 1.054827332496643,
      "learning_rate": 0.0002835583164544139,
      "loss": 2.761160373687744,
      "step": 3901
    },
    {
      "epoch": 0.8918857142857143,
      "grad_norm": 0.8035948872566223,
      "learning_rate": 0.000282668627640833,
      "loss": 2.275892972946167,
      "step": 3902
    },
    {
      "epoch": 0.8921142857142857,
      "grad_norm": 0.7125060558319092,
      "learning_rate": 0.0002817785176993047,
      "loss": 2.119058132171631,
      "step": 3903
    },
    {
      "epoch": 0.8923428571428571,
      "grad_norm": 0.9174073934555054,
      "learning_rate": 0.00028088799810414535,
      "loss": 2.324014186859131,
      "step": 3904
    },
    {
      "epoch": 0.8925714285714286,
      "grad_norm": 0.8782837390899658,
      "learning_rate": 0.0002799970803349523,
      "loss": 2.6083483695983887,
      "step": 3905
    },
    {
      "epoch": 0.8928,
      "grad_norm": 0.84846031665802,
      "learning_rate": 0.00027910577587645567,
      "loss": 2.476255416870117,
      "step": 3906
    },
    {
      "epoch": 0.8930285714285714,
      "grad_norm": 1.0361802577972412,
      "learning_rate": 0.0002782140962183704,
      "loss": 3.022219657897949,
      "step": 3907
    },
    {
      "epoch": 0.8932571428571429,
      "grad_norm": 1.0475831031799316,
      "learning_rate": 0.000277322052855248,
      "loss": 2.4415645599365234,
      "step": 3908
    },
    {
      "epoch": 0.8934857142857143,
      "grad_norm": 1.2914146184921265,
      "learning_rate": 0.0002764296572863284,
      "loss": 2.7241673469543457,
      "step": 3909
    },
    {
      "epoch": 0.8937142857142857,
      "grad_norm": 1.0024330615997314,
      "learning_rate": 0.000275536921015392,
      "loss": 2.458156108856201,
      "step": 3910
    },
    {
      "epoch": 0.8939428571428572,
      "grad_norm": 1.0608261823654175,
      "learning_rate": 0.0002746438555506109,
      "loss": 2.5056405067443848,
      "step": 3911
    },
    {
      "epoch": 0.8941714285714286,
      "grad_norm": 0.8409633636474609,
      "learning_rate": 0.00027375047240440104,
      "loss": 2.493762731552124,
      "step": 3912
    },
    {
      "epoch": 0.8944,
      "grad_norm": 0.6956303715705872,
      "learning_rate": 0.00027285678309327336,
      "loss": 2.284900665283203,
      "step": 3913
    },
    {
      "epoch": 0.8946285714285714,
      "grad_norm": 0.8261454105377197,
      "learning_rate": 0.00027196279913768587,
      "loss": 2.2236056327819824,
      "step": 3914
    },
    {
      "epoch": 0.8948571428571429,
      "grad_norm": 1.1243560314178467,
      "learning_rate": 0.00027106853206189423,
      "loss": 2.5355007648468018,
      "step": 3915
    },
    {
      "epoch": 0.8950857142857143,
      "grad_norm": 0.7332906126976013,
      "learning_rate": 0.00027017399339380435,
      "loss": 2.2439217567443848,
      "step": 3916
    },
    {
      "epoch": 0.8953142857142857,
      "grad_norm": 1.013310194015503,
      "learning_rate": 0.0002692791946648229,
      "loss": 2.530683994293213,
      "step": 3917
    },
    {
      "epoch": 0.8955428571428572,
      "grad_norm": 1.0587997436523438,
      "learning_rate": 0.000268384147409709,
      "loss": 2.297908306121826,
      "step": 3918
    },
    {
      "epoch": 0.8957714285714286,
      "grad_norm": 1.0844494104385376,
      "learning_rate": 0.00026748886316642573,
      "loss": 2.334465265274048,
      "step": 3919
    },
    {
      "epoch": 0.896,
      "grad_norm": 0.8452267646789551,
      "learning_rate": 0.0002665933534759908,
      "loss": 2.075465440750122,
      "step": 3920
    },
    {
      "epoch": 0.8962285714285714,
      "grad_norm": 0.7570284605026245,
      "learning_rate": 0.0002656976298823284,
      "loss": 2.2906134128570557,
      "step": 3921
    },
    {
      "epoch": 0.8964571428571428,
      "grad_norm": 1.0567388534545898,
      "learning_rate": 0.0002648017039321199,
      "loss": 2.3196933269500732,
      "step": 3922
    },
    {
      "epoch": 0.8966857142857143,
      "grad_norm": 0.8485323190689087,
      "learning_rate": 0.00026390558717465546,
      "loss": 2.405189275741577,
      "step": 3923
    },
    {
      "epoch": 0.8969142857142857,
      "grad_norm": 1.0184311866760254,
      "learning_rate": 0.0002630092911616848,
      "loss": 2.5050806999206543,
      "step": 3924
    },
    {
      "epoch": 0.8971428571428571,
      "grad_norm": 1.159729242324829,
      "learning_rate": 0.0002621128274472684,
      "loss": 2.6244583129882812,
      "step": 3925
    },
    {
      "epoch": 0.8973714285714286,
      "grad_norm": 1.1844466924667358,
      "learning_rate": 0.00026121620758762877,
      "loss": 2.610020160675049,
      "step": 3926
    },
    {
      "epoch": 0.8976,
      "grad_norm": 0.9601858854293823,
      "learning_rate": 0.00026031944314100074,
      "loss": 2.3717613220214844,
      "step": 3927
    },
    {
      "epoch": 0.8978285714285714,
      "grad_norm": 1.025377631187439,
      "learning_rate": 0.0002594225456674837,
      "loss": 2.8128857612609863,
      "step": 3928
    },
    {
      "epoch": 0.8980571428571429,
      "grad_norm": 0.9653069972991943,
      "learning_rate": 0.0002585255267288912,
      "loss": 2.569420337677002,
      "step": 3929
    },
    {
      "epoch": 0.8982857142857142,
      "grad_norm": 1.051954984664917,
      "learning_rate": 0.00025762839788860307,
      "loss": 2.353025197982788,
      "step": 3930
    },
    {
      "epoch": 0.8985142857142857,
      "grad_norm": 0.9138957858085632,
      "learning_rate": 0.0002567311707114157,
      "loss": 2.265282392501831,
      "step": 3931
    },
    {
      "epoch": 0.8987428571428572,
      "grad_norm": 0.7849396467208862,
      "learning_rate": 0.00025583385676339316,
      "loss": 2.4859724044799805,
      "step": 3932
    },
    {
      "epoch": 0.8989714285714285,
      "grad_norm": 0.9586713910102844,
      "learning_rate": 0.00025493646761171794,
      "loss": 2.4317121505737305,
      "step": 3933
    },
    {
      "epoch": 0.8992,
      "grad_norm": 0.9305214285850525,
      "learning_rate": 0.00025403901482454216,
      "loss": 2.417233943939209,
      "step": 3934
    },
    {
      "epoch": 0.8994285714285715,
      "grad_norm": 1.43428635597229,
      "learning_rate": 0.0002531415099708382,
      "loss": 2.0631260871887207,
      "step": 3935
    },
    {
      "epoch": 0.8996571428571428,
      "grad_norm": 0.9271275997161865,
      "learning_rate": 0.00025224396462024943,
      "loss": 2.244919538497925,
      "step": 3936
    },
    {
      "epoch": 0.8998857142857143,
      "grad_norm": 0.8392537236213684,
      "learning_rate": 0.0002513463903429418,
      "loss": 2.225341796875,
      "step": 3937
    },
    {
      "epoch": 0.9001142857142858,
      "grad_norm": 1.079267144203186,
      "learning_rate": 0.00025044879870945335,
      "loss": 2.584639549255371,
      "step": 3938
    },
    {
      "epoch": 0.9003428571428571,
      "grad_norm": 1.2713905572891235,
      "learning_rate": 0.00024955120129054666,
      "loss": 2.5301380157470703,
      "step": 3939
    },
    {
      "epoch": 0.9005714285714286,
      "grad_norm": 1.0413649082183838,
      "learning_rate": 0.00024865360965705834,
      "loss": 2.736823797225952,
      "step": 3940
    },
    {
      "epoch": 0.9008,
      "grad_norm": 1.0848277807235718,
      "learning_rate": 0.0002477560353797505,
      "loss": 2.3798904418945312,
      "step": 3941
    },
    {
      "epoch": 0.9010285714285714,
      "grad_norm": 0.6877342462539673,
      "learning_rate": 0.0002468584900291618,
      "loss": 2.5109972953796387,
      "step": 3942
    },
    {
      "epoch": 0.9012571428571429,
      "grad_norm": 0.7634028792381287,
      "learning_rate": 0.0002459609851754579,
      "loss": 2.486176013946533,
      "step": 3943
    },
    {
      "epoch": 0.9014857142857143,
      "grad_norm": 0.9559112191200256,
      "learning_rate": 0.00024506353238828207,
      "loss": 2.3006539344787598,
      "step": 3944
    },
    {
      "epoch": 0.9017142857142857,
      "grad_norm": 1.112424612045288,
      "learning_rate": 0.00024416614323660685,
      "loss": 2.1733760833740234,
      "step": 3945
    },
    {
      "epoch": 0.9019428571428572,
      "grad_norm": 0.7271722555160522,
      "learning_rate": 0.00024326882928858435,
      "loss": 2.3603501319885254,
      "step": 3946
    },
    {
      "epoch": 0.9021714285714286,
      "grad_norm": 0.8452682495117188,
      "learning_rate": 0.00024237160211139694,
      "loss": 2.7863693237304688,
      "step": 3947
    },
    {
      "epoch": 0.9024,
      "grad_norm": 0.7631592154502869,
      "learning_rate": 0.00024147447327110884,
      "loss": 1.9289214611053467,
      "step": 3948
    },
    {
      "epoch": 0.9026285714285714,
      "grad_norm": 0.9622635245323181,
      "learning_rate": 0.00024057745433251636,
      "loss": 2.6968278884887695,
      "step": 3949
    },
    {
      "epoch": 0.9028571428571428,
      "grad_norm": 1.0272561311721802,
      "learning_rate": 0.00023968055685899924,
      "loss": 2.406121015548706,
      "step": 3950
    },
    {
      "epoch": 0.9030857142857143,
      "grad_norm": 0.7079058289527893,
      "learning_rate": 0.00023878379241237135,
      "loss": 2.231358528137207,
      "step": 3951
    },
    {
      "epoch": 0.9033142857142857,
      "grad_norm": 1.6693745851516724,
      "learning_rate": 0.00023788717255273154,
      "loss": 2.5318539142608643,
      "step": 3952
    },
    {
      "epoch": 0.9035428571428571,
      "grad_norm": 0.8991554975509644,
      "learning_rate": 0.0002369907088383152,
      "loss": 2.3007736206054688,
      "step": 3953
    },
    {
      "epoch": 0.9037714285714286,
      "grad_norm": 0.8079550862312317,
      "learning_rate": 0.00023609441282534458,
      "loss": 2.318856716156006,
      "step": 3954
    },
    {
      "epoch": 0.904,
      "grad_norm": 0.7997955679893494,
      "learning_rate": 0.00023519829606788006,
      "loss": 2.3069446086883545,
      "step": 3955
    },
    {
      "epoch": 0.9042285714285714,
      "grad_norm": 0.989905595779419,
      "learning_rate": 0.00023430237011767165,
      "loss": 2.3688533306121826,
      "step": 3956
    },
    {
      "epoch": 0.9044571428571428,
      "grad_norm": 0.8352833986282349,
      "learning_rate": 0.00023340664652400926,
      "loss": 2.4327638149261475,
      "step": 3957
    },
    {
      "epoch": 0.9046857142857143,
      "grad_norm": 0.8230957984924316,
      "learning_rate": 0.00023251113683357433,
      "loss": 2.5779075622558594,
      "step": 3958
    },
    {
      "epoch": 0.9049142857142857,
      "grad_norm": 1.0939419269561768,
      "learning_rate": 0.00023161585259029094,
      "loss": 2.337880849838257,
      "step": 3959
    },
    {
      "epoch": 0.9051428571428571,
      "grad_norm": 1.0720669031143188,
      "learning_rate": 0.0002307208053351771,
      "loss": 2.0460011959075928,
      "step": 3960
    },
    {
      "epoch": 0.9053714285714286,
      "grad_norm": 0.6544449925422668,
      "learning_rate": 0.0002298260066061957,
      "loss": 2.283308982849121,
      "step": 3961
    },
    {
      "epoch": 0.9056,
      "grad_norm": 1.23578679561615,
      "learning_rate": 0.0002289314679381059,
      "loss": 2.6354641914367676,
      "step": 3962
    },
    {
      "epoch": 0.9058285714285714,
      "grad_norm": 0.8890902400016785,
      "learning_rate": 0.00022803720086231422,
      "loss": 2.4197185039520264,
      "step": 3963
    },
    {
      "epoch": 0.9060571428571429,
      "grad_norm": 0.8863524794578552,
      "learning_rate": 0.00022714321690672665,
      "loss": 2.3518199920654297,
      "step": 3964
    },
    {
      "epoch": 0.9062857142857143,
      "grad_norm": 0.8962447643280029,
      "learning_rate": 0.000226249527595599,
      "loss": 2.4859046936035156,
      "step": 3965
    },
    {
      "epoch": 0.9065142857142857,
      "grad_norm": 0.8434134125709534,
      "learning_rate": 0.00022535614444938912,
      "loss": 2.246644973754883,
      "step": 3966
    },
    {
      "epoch": 0.9067428571428572,
      "grad_norm": 0.9095499515533447,
      "learning_rate": 0.00022446307898460805,
      "loss": 2.102403402328491,
      "step": 3967
    },
    {
      "epoch": 0.9069714285714285,
      "grad_norm": 0.8596362471580505,
      "learning_rate": 0.00022357034271367165,
      "loss": 2.4859039783477783,
      "step": 3968
    },
    {
      "epoch": 0.9072,
      "grad_norm": 2.5734164714813232,
      "learning_rate": 0.0002226779471447521,
      "loss": 2.4960479736328125,
      "step": 3969
    },
    {
      "epoch": 0.9074285714285715,
      "grad_norm": 0.7999812364578247,
      "learning_rate": 0.00022178590378162956,
      "loss": 2.254335403442383,
      "step": 3970
    },
    {
      "epoch": 0.9076571428571428,
      "grad_norm": 0.9631752967834473,
      "learning_rate": 0.0002208942241235443,
      "loss": 2.635525703430176,
      "step": 3971
    },
    {
      "epoch": 0.9078857142857143,
      "grad_norm": 1.0830048322677612,
      "learning_rate": 0.00022000291966504771,
      "loss": 2.2836694717407227,
      "step": 3972
    },
    {
      "epoch": 0.9081142857142858,
      "grad_norm": 0.9416781067848206,
      "learning_rate": 0.00021911200189585463,
      "loss": 2.4624974727630615,
      "step": 3973
    },
    {
      "epoch": 0.9083428571428571,
      "grad_norm": 1.1867402791976929,
      "learning_rate": 0.00021822148230069536,
      "loss": 2.6794936656951904,
      "step": 3974
    },
    {
      "epoch": 0.9085714285714286,
      "grad_norm": 0.9254763722419739,
      "learning_rate": 0.00021733137235916707,
      "loss": 2.1426188945770264,
      "step": 3975
    },
    {
      "epoch": 0.9088,
      "grad_norm": 0.8052186965942383,
      "learning_rate": 0.0002164416835455862,
      "loss": 2.318835735321045,
      "step": 3976
    },
    {
      "epoch": 0.9090285714285714,
      "grad_norm": 0.7510400414466858,
      "learning_rate": 0.0002155524273288405,
      "loss": 2.636512279510498,
      "step": 3977
    },
    {
      "epoch": 0.9092571428571429,
      "grad_norm": 0.8755637407302856,
      "learning_rate": 0.00021466361517224125,
      "loss": 2.24038028717041,
      "step": 3978
    },
    {
      "epoch": 0.9094857142857142,
      "grad_norm": 0.9088058471679688,
      "learning_rate": 0.0002137752585333754,
      "loss": 2.672713279724121,
      "step": 3979
    },
    {
      "epoch": 0.9097142857142857,
      "grad_norm": 0.8010004162788391,
      "learning_rate": 0.0002128873688639577,
      "loss": 2.4937987327575684,
      "step": 3980
    },
    {
      "epoch": 0.9099428571428572,
      "grad_norm": 1.1212283372879028,
      "learning_rate": 0.00021199995760968321,
      "loss": 2.300558567047119,
      "step": 3981
    },
    {
      "epoch": 0.9101714285714285,
      "grad_norm": 0.7318730354309082,
      "learning_rate": 0.00021111303621008015,
      "loss": 2.1496212482452393,
      "step": 3982
    },
    {
      "epoch": 0.9104,
      "grad_norm": 0.9009625911712646,
      "learning_rate": 0.0002102266160983617,
      "loss": 2.2337536811828613,
      "step": 3983
    },
    {
      "epoch": 0.9106285714285715,
      "grad_norm": 0.8393200039863586,
      "learning_rate": 0.0002093407087012791,
      "loss": 2.217409133911133,
      "step": 3984
    },
    {
      "epoch": 0.9108571428571428,
      "grad_norm": 1.0918875932693481,
      "learning_rate": 0.00020845532543897432,
      "loss": 2.385373830795288,
      "step": 3985
    },
    {
      "epoch": 0.9110857142857143,
      "grad_norm": 0.7653070688247681,
      "learning_rate": 0.00020757047772483278,
      "loss": 2.3265938758850098,
      "step": 3986
    },
    {
      "epoch": 0.9113142857142857,
      "grad_norm": 0.8826605677604675,
      "learning_rate": 0.00020668617696533603,
      "loss": 2.161219596862793,
      "step": 3987
    },
    {
      "epoch": 0.9115428571428571,
      "grad_norm": 1.0603606700897217,
      "learning_rate": 0.00020580243455991493,
      "loss": 2.449334144592285,
      "step": 3988
    },
    {
      "epoch": 0.9117714285714286,
      "grad_norm": 0.7682347297668457,
      "learning_rate": 0.00020491926190080285,
      "loss": 2.4253010749816895,
      "step": 3989
    },
    {
      "epoch": 0.912,
      "grad_norm": 1.0122877359390259,
      "learning_rate": 0.00020403667037288833,
      "loss": 2.507545232772827,
      "step": 3990
    },
    {
      "epoch": 0.9122285714285714,
      "grad_norm": 0.8381178975105286,
      "learning_rate": 0.0002031546713535688,
      "loss": 2.626365900039673,
      "step": 3991
    },
    {
      "epoch": 0.9124571428571429,
      "grad_norm": 0.8831295967102051,
      "learning_rate": 0.0002022732762126038,
      "loss": 2.6564908027648926,
      "step": 3992
    },
    {
      "epoch": 0.9126857142857143,
      "grad_norm": 0.8767942786216736,
      "learning_rate": 0.0002013924963119681,
      "loss": 2.6848864555358887,
      "step": 3993
    },
    {
      "epoch": 0.9129142857142857,
      "grad_norm": 0.8901404738426208,
      "learning_rate": 0.0002005123430057056,
      "loss": 2.4400768280029297,
      "step": 3994
    },
    {
      "epoch": 0.9131428571428571,
      "grad_norm": 1.1278276443481445,
      "learning_rate": 0.00019963282763978268,
      "loss": 2.179935932159424,
      "step": 3995
    },
    {
      "epoch": 0.9133714285714286,
      "grad_norm": 0.844079315662384,
      "learning_rate": 0.00019875396155194242,
      "loss": 2.4076547622680664,
      "step": 3996
    },
    {
      "epoch": 0.9136,
      "grad_norm": 0.9365500211715698,
      "learning_rate": 0.00019787575607155793,
      "loss": 2.6449801921844482,
      "step": 3997
    },
    {
      "epoch": 0.9138285714285714,
      "grad_norm": 0.8068600296974182,
      "learning_rate": 0.0001969982225194864,
      "loss": 2.5972347259521484,
      "step": 3998
    },
    {
      "epoch": 0.9140571428571429,
      "grad_norm": 0.9563422799110413,
      "learning_rate": 0.0001961213722079233,
      "loss": 2.363255023956299,
      "step": 3999
    },
    {
      "epoch": 0.9142857142857143,
      "grad_norm": 0.8940248489379883,
      "learning_rate": 0.00019524521644025684,
      "loss": 2.3891992568969727,
      "step": 4000
    },
    {
      "epoch": 0.9145142857142857,
      "grad_norm": 1.132643461227417,
      "learning_rate": 0.00019436976651092142,
      "loss": 2.3753082752227783,
      "step": 4001
    },
    {
      "epoch": 0.9147428571428572,
      "grad_norm": 0.7609168887138367,
      "learning_rate": 0.0001934950337052529,
      "loss": 2.3255772590637207,
      "step": 4002
    },
    {
      "epoch": 0.9149714285714285,
      "grad_norm": 0.9986788630485535,
      "learning_rate": 0.00019262102929934265,
      "loss": 2.5992751121520996,
      "step": 4003
    },
    {
      "epoch": 0.9152,
      "grad_norm": 1.032705307006836,
      "learning_rate": 0.00019174776455989245,
      "loss": 2.2982308864593506,
      "step": 4004
    },
    {
      "epoch": 0.9154285714285715,
      "grad_norm": 0.7303146719932556,
      "learning_rate": 0.0001908752507440689,
      "loss": 2.4449729919433594,
      "step": 4005
    },
    {
      "epoch": 0.9156571428571428,
      "grad_norm": 0.7767391204833984,
      "learning_rate": 0.00019000349909935853,
      "loss": 2.3875913619995117,
      "step": 4006
    },
    {
      "epoch": 0.9158857142857143,
      "grad_norm": 0.7868723273277283,
      "learning_rate": 0.0001891325208634231,
      "loss": 2.3930962085723877,
      "step": 4007
    },
    {
      "epoch": 0.9161142857142857,
      "grad_norm": 0.7586782574653625,
      "learning_rate": 0.00018826232726395408,
      "loss": 2.418017864227295,
      "step": 4008
    },
    {
      "epoch": 0.9163428571428571,
      "grad_norm": 0.8111346364021301,
      "learning_rate": 0.00018739292951852848,
      "loss": 2.282848834991455,
      "step": 4009
    },
    {
      "epoch": 0.9165714285714286,
      "grad_norm": 1.048263669013977,
      "learning_rate": 0.0001865243388344639,
      "loss": 2.1864376068115234,
      "step": 4010
    },
    {
      "epoch": 0.9168,
      "grad_norm": 0.8866171836853027,
      "learning_rate": 0.00018565656640867447,
      "loss": 2.5920779705047607,
      "step": 4011
    },
    {
      "epoch": 0.9170285714285714,
      "grad_norm": 0.9185689687728882,
      "learning_rate": 0.00018478962342752584,
      "loss": 2.465405225753784,
      "step": 4012
    },
    {
      "epoch": 0.9172571428571429,
      "grad_norm": 1.0608423948287964,
      "learning_rate": 0.0001839235210666916,
      "loss": 2.2137374877929688,
      "step": 4013
    },
    {
      "epoch": 0.9174857142857142,
      "grad_norm": 5.751997470855713,
      "learning_rate": 0.00018305827049100898,
      "loss": 2.5906083583831787,
      "step": 4014
    },
    {
      "epoch": 0.9177142857142857,
      "grad_norm": 0.9384812116622925,
      "learning_rate": 0.00018219388285433503,
      "loss": 2.574465751647949,
      "step": 4015
    },
    {
      "epoch": 0.9179428571428572,
      "grad_norm": 0.8990826606750488,
      "learning_rate": 0.0001813303692994025,
      "loss": 2.430551767349243,
      "step": 4016
    },
    {
      "epoch": 0.9181714285714285,
      "grad_norm": 0.7459377646446228,
      "learning_rate": 0.0001804677409576765,
      "loss": 2.1980886459350586,
      "step": 4017
    },
    {
      "epoch": 0.9184,
      "grad_norm": 0.9729992747306824,
      "learning_rate": 0.00017960600894921126,
      "loss": 2.817126750946045,
      "step": 4018
    },
    {
      "epoch": 0.9186285714285715,
      "grad_norm": 1.0766651630401611,
      "learning_rate": 0.00017874518438250596,
      "loss": 2.3396570682525635,
      "step": 4019
    },
    {
      "epoch": 0.9188571428571428,
      "grad_norm": 1.020684838294983,
      "learning_rate": 0.00017788527835436248,
      "loss": 2.406552791595459,
      "step": 4020
    },
    {
      "epoch": 0.9190857142857143,
      "grad_norm": 0.7695953249931335,
      "learning_rate": 0.00017702630194974168,
      "loss": 2.369671106338501,
      "step": 4021
    },
    {
      "epoch": 0.9193142857142858,
      "grad_norm": 0.744380533695221,
      "learning_rate": 0.0001761682662416208,
      "loss": 2.307199478149414,
      "step": 4022
    },
    {
      "epoch": 0.9195428571428571,
      "grad_norm": 0.9786500930786133,
      "learning_rate": 0.0001753111822908507,
      "loss": 2.2919182777404785,
      "step": 4023
    },
    {
      "epoch": 0.9197714285714286,
      "grad_norm": 0.8593237996101379,
      "learning_rate": 0.000174455061146013,
      "loss": 2.2998502254486084,
      "step": 4024
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.9072189331054688,
      "learning_rate": 0.00017359991384327831,
      "loss": 2.139403820037842,
      "step": 4025
    },
    {
      "epoch": 0.9202285714285714,
      "grad_norm": 0.9276069402694702,
      "learning_rate": 0.00017274575140626317,
      "loss": 2.2576217651367188,
      "step": 4026
    },
    {
      "epoch": 0.9204571428571429,
      "grad_norm": 1.1437565088272095,
      "learning_rate": 0.00017189258484588853,
      "loss": 2.5847930908203125,
      "step": 4027
    },
    {
      "epoch": 0.9206857142857143,
      "grad_norm": 0.8358618021011353,
      "learning_rate": 0.00017104042516023735,
      "loss": 2.281789541244507,
      "step": 4028
    },
    {
      "epoch": 0.9209142857142857,
      "grad_norm": 0.7668153643608093,
      "learning_rate": 0.00017018928333441347,
      "loss": 2.4084067344665527,
      "step": 4029
    },
    {
      "epoch": 0.9211428571428572,
      "grad_norm": 0.7179484963417053,
      "learning_rate": 0.00016933917034039927,
      "loss": 2.2160041332244873,
      "step": 4030
    },
    {
      "epoch": 0.9213714285714286,
      "grad_norm": 0.8012107610702515,
      "learning_rate": 0.00016849009713691454,
      "loss": 2.0939292907714844,
      "step": 4031
    },
    {
      "epoch": 0.9216,
      "grad_norm": 0.8473196625709534,
      "learning_rate": 0.0001676420746692755,
      "loss": 2.299546718597412,
      "step": 4032
    },
    {
      "epoch": 0.9218285714285714,
      "grad_norm": 1.1303839683532715,
      "learning_rate": 0.00016679511386925337,
      "loss": 2.4380435943603516,
      "step": 4033
    },
    {
      "epoch": 0.9220571428571429,
      "grad_norm": 0.9459303021430969,
      "learning_rate": 0.0001659492256549333,
      "loss": 2.445286750793457,
      "step": 4034
    },
    {
      "epoch": 0.9222857142857143,
      "grad_norm": 0.9648786187171936,
      "learning_rate": 0.000165104420930574,
      "loss": 2.2518699169158936,
      "step": 4035
    },
    {
      "epoch": 0.9225142857142857,
      "grad_norm": 0.9451817870140076,
      "learning_rate": 0.00016426071058646717,
      "loss": 2.6671695709228516,
      "step": 4036
    },
    {
      "epoch": 0.9227428571428572,
      "grad_norm": 0.8533628582954407,
      "learning_rate": 0.00016341810549879666,
      "loss": 2.4448466300964355,
      "step": 4037
    },
    {
      "epoch": 0.9229714285714286,
      "grad_norm": 0.8332159519195557,
      "learning_rate": 0.00016257661652949876,
      "loss": 2.55609130859375,
      "step": 4038
    },
    {
      "epoch": 0.9232,
      "grad_norm": 0.8477275967597961,
      "learning_rate": 0.00016173625452612202,
      "loss": 2.5962350368499756,
      "step": 4039
    },
    {
      "epoch": 0.9234285714285714,
      "grad_norm": 1.2214335203170776,
      "learning_rate": 0.00016089703032168734,
      "loss": 2.3077552318573,
      "step": 4040
    },
    {
      "epoch": 0.9236571428571428,
      "grad_norm": 0.8438049554824829,
      "learning_rate": 0.00016005895473454834,
      "loss": 2.4624783992767334,
      "step": 4041
    },
    {
      "epoch": 0.9238857142857143,
      "grad_norm": 1.7916700839996338,
      "learning_rate": 0.0001592220385682518,
      "loss": 2.903526782989502,
      "step": 4042
    },
    {
      "epoch": 0.9241142857142857,
      "grad_norm": 0.8269637227058411,
      "learning_rate": 0.00015838629261139892,
      "loss": 2.7185049057006836,
      "step": 4043
    },
    {
      "epoch": 0.9243428571428571,
      "grad_norm": 1.0357486009597778,
      "learning_rate": 0.00015755172763750542,
      "loss": 2.6937804222106934,
      "step": 4044
    },
    {
      "epoch": 0.9245714285714286,
      "grad_norm": 0.851421058177948,
      "learning_rate": 0.00015671835440486337,
      "loss": 2.4877171516418457,
      "step": 4045
    },
    {
      "epoch": 0.9248,
      "grad_norm": 0.8602510690689087,
      "learning_rate": 0.000155886183656402,
      "loss": 2.3517823219299316,
      "step": 4046
    },
    {
      "epoch": 0.9250285714285714,
      "grad_norm": 0.7836548089981079,
      "learning_rate": 0.00015505522611954976,
      "loss": 2.3503599166870117,
      "step": 4047
    },
    {
      "epoch": 0.9252571428571429,
      "grad_norm": 0.7246906161308289,
      "learning_rate": 0.00015422549250609542,
      "loss": 2.481053590774536,
      "step": 4048
    },
    {
      "epoch": 0.9254857142857142,
      "grad_norm": 1.5054020881652832,
      "learning_rate": 0.00015339699351205025,
      "loss": 2.516941547393799,
      "step": 4049
    },
    {
      "epoch": 0.9257142857142857,
      "grad_norm": 0.8777369856834412,
      "learning_rate": 0.0001525697398175105,
      "loss": 2.346561908721924,
      "step": 4050
    },
    {
      "epoch": 0.9259428571428572,
      "grad_norm": 0.957566499710083,
      "learning_rate": 0.0001517437420865191,
      "loss": 2.564037322998047,
      "step": 4051
    },
    {
      "epoch": 0.9261714285714285,
      "grad_norm": 0.7308264970779419,
      "learning_rate": 0.00015091901096692863,
      "loss": 2.316946506500244,
      "step": 4052
    },
    {
      "epoch": 0.9264,
      "grad_norm": 1.0200690031051636,
      "learning_rate": 0.00015009555709026364,
      "loss": 2.265915870666504,
      "step": 4053
    },
    {
      "epoch": 0.9266285714285715,
      "grad_norm": 1.0989201068878174,
      "learning_rate": 0.00014927339107158436,
      "loss": 2.2415027618408203,
      "step": 4054
    },
    {
      "epoch": 0.9268571428571428,
      "grad_norm": 1.1814255714416504,
      "learning_rate": 0.00014845252350934892,
      "loss": 2.6238152980804443,
      "step": 4055
    },
    {
      "epoch": 0.9270857142857143,
      "grad_norm": 1.0073248147964478,
      "learning_rate": 0.00014763296498527742,
      "loss": 2.472717761993408,
      "step": 4056
    },
    {
      "epoch": 0.9273142857142858,
      "grad_norm": 0.8806115984916687,
      "learning_rate": 0.00014681472606421512,
      "loss": 2.294185161590576,
      "step": 4057
    },
    {
      "epoch": 0.9275428571428571,
      "grad_norm": 0.6753144264221191,
      "learning_rate": 0.00014599781729399667,
      "loss": 2.193727970123291,
      "step": 4058
    },
    {
      "epoch": 0.9277714285714286,
      "grad_norm": 0.8535948395729065,
      "learning_rate": 0.0001451822492053096,
      "loss": 2.5586745738983154,
      "step": 4059
    },
    {
      "epoch": 0.928,
      "grad_norm": 0.8999954462051392,
      "learning_rate": 0.00014436803231155892,
      "loss": 2.0609047412872314,
      "step": 4060
    },
    {
      "epoch": 0.9282285714285714,
      "grad_norm": 0.8234038949012756,
      "learning_rate": 0.00014355517710873183,
      "loss": 2.5127782821655273,
      "step": 4061
    },
    {
      "epoch": 0.9284571428571429,
      "grad_norm": 0.8933808207511902,
      "learning_rate": 0.00014274369407526167,
      "loss": 2.5990042686462402,
      "step": 4062
    },
    {
      "epoch": 0.9286857142857143,
      "grad_norm": 0.8692030906677246,
      "learning_rate": 0.00014193359367189345,
      "loss": 2.327260971069336,
      "step": 4063
    },
    {
      "epoch": 0.9289142857142857,
      "grad_norm": 1.187094807624817,
      "learning_rate": 0.00014112488634154886,
      "loss": 2.454547882080078,
      "step": 4064
    },
    {
      "epoch": 0.9291428571428572,
      "grad_norm": 0.850171685218811,
      "learning_rate": 0.00014031758250919173,
      "loss": 2.56588077545166,
      "step": 4065
    },
    {
      "epoch": 0.9293714285714286,
      "grad_norm": 1.222370982170105,
      "learning_rate": 0.00013951169258169337,
      "loss": 2.5731751918792725,
      "step": 4066
    },
    {
      "epoch": 0.9296,
      "grad_norm": 0.8080718517303467,
      "learning_rate": 0.00013870722694769857,
      "loss": 2.0196468830108643,
      "step": 4067
    },
    {
      "epoch": 0.9298285714285714,
      "grad_norm": 0.8675833344459534,
      "learning_rate": 0.000137904195977492,
      "loss": 2.7812068462371826,
      "step": 4068
    },
    {
      "epoch": 0.9300571428571428,
      "grad_norm": 0.7879951596260071,
      "learning_rate": 0.00013710261002286382,
      "loss": 2.163550853729248,
      "step": 4069
    },
    {
      "epoch": 0.9302857142857143,
      "grad_norm": 0.9352191686630249,
      "learning_rate": 0.0001363024794169768,
      "loss": 2.4731109142303467,
      "step": 4070
    },
    {
      "epoch": 0.9305142857142857,
      "grad_norm": 0.8116070032119751,
      "learning_rate": 0.00013550381447423315,
      "loss": 2.2720298767089844,
      "step": 4071
    },
    {
      "epoch": 0.9307428571428571,
      "grad_norm": 0.9803361892700195,
      "learning_rate": 0.00013470662549014103,
      "loss": 2.369901657104492,
      "step": 4072
    },
    {
      "epoch": 0.9309714285714286,
      "grad_norm": 0.8042914271354675,
      "learning_rate": 0.00013391092274118256,
      "loss": 2.204988479614258,
      "step": 4073
    },
    {
      "epoch": 0.9312,
      "grad_norm": 0.7984622120857239,
      "learning_rate": 0.00013311671648468055,
      "loss": 2.4570953845977783,
      "step": 4074
    },
    {
      "epoch": 0.9314285714285714,
      "grad_norm": 0.6965162754058838,
      "learning_rate": 0.00013232401695866685,
      "loss": 2.2781190872192383,
      "step": 4075
    },
    {
      "epoch": 0.9316571428571429,
      "grad_norm": 0.9965574145317078,
      "learning_rate": 0.00013153283438175034,
      "loss": 2.4924612045288086,
      "step": 4076
    },
    {
      "epoch": 0.9318857142857143,
      "grad_norm": 0.9764564037322998,
      "learning_rate": 0.0001307431789529849,
      "loss": 2.7417101860046387,
      "step": 4077
    },
    {
      "epoch": 0.9321142857142857,
      "grad_norm": 0.7986329197883606,
      "learning_rate": 0.000129955060851738,
      "loss": 2.0420608520507812,
      "step": 4078
    },
    {
      "epoch": 0.9323428571428571,
      "grad_norm": 1.139923095703125,
      "learning_rate": 0.00012916849023755992,
      "loss": 2.4049439430236816,
      "step": 4079
    },
    {
      "epoch": 0.9325714285714286,
      "grad_norm": 1.9383529424667358,
      "learning_rate": 0.00012838347725005212,
      "loss": 2.269784450531006,
      "step": 4080
    },
    {
      "epoch": 0.9328,
      "grad_norm": 0.8943877220153809,
      "learning_rate": 0.00012760003200873698,
      "loss": 2.3598670959472656,
      "step": 4081
    },
    {
      "epoch": 0.9330285714285714,
      "grad_norm": 0.7517910599708557,
      "learning_rate": 0.00012681816461292713,
      "loss": 2.4978208541870117,
      "step": 4082
    },
    {
      "epoch": 0.9332571428571429,
      "grad_norm": 0.8110870122909546,
      "learning_rate": 0.00012603788514159553,
      "loss": 2.3588309288024902,
      "step": 4083
    },
    {
      "epoch": 0.9334857142857143,
      "grad_norm": 1.1071209907531738,
      "learning_rate": 0.00012525920365324535,
      "loss": 2.2507481575012207,
      "step": 4084
    },
    {
      "epoch": 0.9337142857142857,
      "grad_norm": 0.8265325427055359,
      "learning_rate": 0.00012448213018578008,
      "loss": 2.6611132621765137,
      "step": 4085
    },
    {
      "epoch": 0.9339428571428572,
      "grad_norm": 1.0104787349700928,
      "learning_rate": 0.00012370667475637473,
      "loss": 2.392538070678711,
      "step": 4086
    },
    {
      "epoch": 0.9341714285714285,
      "grad_norm": 0.5885429978370667,
      "learning_rate": 0.00012293284736134604,
      "loss": 2.283780097961426,
      "step": 4087
    },
    {
      "epoch": 0.9344,
      "grad_norm": 0.8644852042198181,
      "learning_rate": 0.0001221606579760241,
      "loss": 2.506540298461914,
      "step": 4088
    },
    {
      "epoch": 0.9346285714285715,
      "grad_norm": 0.8145380020141602,
      "learning_rate": 0.00012139011655462338,
      "loss": 2.0192952156066895,
      "step": 4089
    },
    {
      "epoch": 0.9348571428571428,
      "grad_norm": 0.834328293800354,
      "learning_rate": 0.00012062123303011499,
      "loss": 2.4456534385681152,
      "step": 4090
    },
    {
      "epoch": 0.9350857142857143,
      "grad_norm": 0.8666711449623108,
      "learning_rate": 0.00011985401731409792,
      "loss": 2.347778558731079,
      "step": 4091
    },
    {
      "epoch": 0.9353142857142858,
      "grad_norm": 0.8391051888465881,
      "learning_rate": 0.00011908847929667174,
      "loss": 2.4074325561523438,
      "step": 4092
    },
    {
      "epoch": 0.9355428571428571,
      "grad_norm": 0.8293136954307556,
      "learning_rate": 0.00011832462884630888,
      "loss": 2.197774648666382,
      "step": 4093
    },
    {
      "epoch": 0.9357714285714286,
      "grad_norm": 1.0032057762145996,
      "learning_rate": 0.00011756247580972776,
      "loss": 2.465630531311035,
      "step": 4094
    },
    {
      "epoch": 0.936,
      "grad_norm": 0.9053379893302917,
      "learning_rate": 0.00011680203001176529,
      "loss": 2.126312494277954,
      "step": 4095
    },
    {
      "epoch": 0.9362285714285714,
      "grad_norm": 0.9099084734916687,
      "learning_rate": 0.00011604330125525078,
      "loss": 2.281552791595459,
      "step": 4096
    },
    {
      "epoch": 0.9364571428571429,
      "grad_norm": 0.9194054007530212,
      "learning_rate": 0.0001152862993208794,
      "loss": 2.347642421722412,
      "step": 4097
    },
    {
      "epoch": 0.9366857142857142,
      "grad_norm": 0.7057921290397644,
      "learning_rate": 0.0001145310339670857,
      "loss": 2.4390883445739746,
      "step": 4098
    },
    {
      "epoch": 0.9369142857142857,
      "grad_norm": 0.8749399185180664,
      "learning_rate": 0.00011377751492991841,
      "loss": 2.688478708267212,
      "step": 4099
    },
    {
      "epoch": 0.9371428571428572,
      "grad_norm": 1.5993571281433105,
      "learning_rate": 0.0001130257519229145,
      "loss": 2.4041852951049805,
      "step": 4100
    },
    {
      "epoch": 0.9373714285714285,
      "grad_norm": 0.7800861597061157,
      "learning_rate": 0.0001122757546369744,
      "loss": 2.473782539367676,
      "step": 4101
    },
    {
      "epoch": 0.9376,
      "grad_norm": 0.94137042760849,
      "learning_rate": 0.00011152753274023652,
      "loss": 2.4826560020446777,
      "step": 4102
    },
    {
      "epoch": 0.9378285714285715,
      "grad_norm": 0.7894044518470764,
      "learning_rate": 0.0001107810958779531,
      "loss": 2.164605140686035,
      "step": 4103
    },
    {
      "epoch": 0.9380571428571428,
      "grad_norm": 0.9609920382499695,
      "learning_rate": 0.00011003645367236545,
      "loss": 2.4090170860290527,
      "step": 4104
    },
    {
      "epoch": 0.9382857142857143,
      "grad_norm": 0.9201875329017639,
      "learning_rate": 0.0001092936157225805,
      "loss": 2.3377652168273926,
      "step": 4105
    },
    {
      "epoch": 0.9385142857142857,
      "grad_norm": 0.9669972062110901,
      "learning_rate": 0.0001085525916044464,
      "loss": 2.4848029613494873,
      "step": 4106
    },
    {
      "epoch": 0.9387428571428571,
      "grad_norm": 0.8896302580833435,
      "learning_rate": 0.00010781339087042955,
      "loss": 2.5120744705200195,
      "step": 4107
    },
    {
      "epoch": 0.9389714285714286,
      "grad_norm": 0.839582085609436,
      "learning_rate": 0.00010707602304949115,
      "loss": 2.296685218811035,
      "step": 4108
    },
    {
      "epoch": 0.9392,
      "grad_norm": 0.7670115232467651,
      "learning_rate": 0.00010634049764696471,
      "loss": 2.4977598190307617,
      "step": 4109
    },
    {
      "epoch": 0.9394285714285714,
      "grad_norm": 1.0466090440750122,
      "learning_rate": 0.00010560682414443315,
      "loss": 2.76491117477417,
      "step": 4110
    },
    {
      "epoch": 0.9396571428571429,
      "grad_norm": 0.9849022626876831,
      "learning_rate": 0.00010487501199960662,
      "loss": 2.187324047088623,
      "step": 4111
    },
    {
      "epoch": 0.9398857142857143,
      "grad_norm": 0.8419234156608582,
      "learning_rate": 0.00010414507064620107,
      "loss": 2.409064292907715,
      "step": 4112
    },
    {
      "epoch": 0.9401142857142857,
      "grad_norm": 0.730537474155426,
      "learning_rate": 0.00010341700949381593,
      "loss": 2.275254011154175,
      "step": 4113
    },
    {
      "epoch": 0.9403428571428571,
      "grad_norm": 1.2322423458099365,
      "learning_rate": 0.00010269083792781323,
      "loss": 2.17710542678833,
      "step": 4114
    },
    {
      "epoch": 0.9405714285714286,
      "grad_norm": 0.9095158576965332,
      "learning_rate": 0.0001019665653091967,
      "loss": 2.278064250946045,
      "step": 4115
    },
    {
      "epoch": 0.9408,
      "grad_norm": 0.9718487858772278,
      "learning_rate": 0.00010124420097449077,
      "loss": 2.5085482597351074,
      "step": 4116
    },
    {
      "epoch": 0.9410285714285714,
      "grad_norm": 1.404073715209961,
      "learning_rate": 0.00010052375423562038,
      "loss": 2.133132219314575,
      "step": 4117
    },
    {
      "epoch": 0.9412571428571429,
      "grad_norm": 0.9315035939216614,
      "learning_rate": 9.980523437979085e-05,
      "loss": 2.5390799045562744,
      "step": 4118
    },
    {
      "epoch": 0.9414857142857143,
      "grad_norm": 0.7175660729408264,
      "learning_rate": 9.908865066936853e-05,
      "loss": 2.2533316612243652,
      "step": 4119
    },
    {
      "epoch": 0.9417142857142857,
      "grad_norm": 0.8010525703430176,
      "learning_rate": 9.837401234176103e-05,
      "loss": 2.555147647857666,
      "step": 4120
    },
    {
      "epoch": 0.9419428571428572,
      "grad_norm": 0.9625018239021301,
      "learning_rate": 9.766132860929799e-05,
      "loss": 2.2838330268859863,
      "step": 4121
    },
    {
      "epoch": 0.9421714285714285,
      "grad_norm": 0.9006732106208801,
      "learning_rate": 9.695060865911273e-05,
      "loss": 2.2641761302948,
      "step": 4122
    },
    {
      "epoch": 0.9424,
      "grad_norm": 0.7125609517097473,
      "learning_rate": 9.624186165302378e-05,
      "loss": 2.1159963607788086,
      "step": 4123
    },
    {
      "epoch": 0.9426285714285715,
      "grad_norm": 1.0337040424346924,
      "learning_rate": 9.553509672741645e-05,
      "loss": 2.421980381011963,
      "step": 4124
    },
    {
      "epoch": 0.9428571428571428,
      "grad_norm": 0.7667618989944458,
      "learning_rate": 9.483032299312525e-05,
      "loss": 2.5052080154418945,
      "step": 4125
    },
    {
      "epoch": 0.9430857142857143,
      "grad_norm": 0.733282744884491,
      "learning_rate": 9.412754953531663e-05,
      "loss": 2.308384418487549,
      "step": 4126
    },
    {
      "epoch": 0.9433142857142857,
      "grad_norm": 0.7600447535514832,
      "learning_rate": 9.342678541337154e-05,
      "loss": 2.2663519382476807,
      "step": 4127
    },
    {
      "epoch": 0.9435428571428571,
      "grad_norm": 1.2083184719085693,
      "learning_rate": 9.272803966076882e-05,
      "loss": 2.6560750007629395,
      "step": 4128
    },
    {
      "epoch": 0.9437714285714286,
      "grad_norm": 0.935817539691925,
      "learning_rate": 9.203132128496857e-05,
      "loss": 2.355966091156006,
      "step": 4129
    },
    {
      "epoch": 0.944,
      "grad_norm": 0.8715494275093079,
      "learning_rate": 9.133663926729652e-05,
      "loss": 2.507673978805542,
      "step": 4130
    },
    {
      "epoch": 0.9442285714285714,
      "grad_norm": 0.8473942279815674,
      "learning_rate": 9.064400256282756e-05,
      "loss": 2.4564852714538574,
      "step": 4131
    },
    {
      "epoch": 0.9444571428571429,
      "grad_norm": 0.8575993776321411,
      "learning_rate": 8.99534201002709e-05,
      "loss": 2.441746234893799,
      "step": 4132
    },
    {
      "epoch": 0.9446857142857142,
      "grad_norm": 0.9770210385322571,
      "learning_rate": 8.926490078185472e-05,
      "loss": 2.2477941513061523,
      "step": 4133
    },
    {
      "epoch": 0.9449142857142857,
      "grad_norm": 0.7561991214752197,
      "learning_rate": 8.857845348321125e-05,
      "loss": 2.304856300354004,
      "step": 4134
    },
    {
      "epoch": 0.9451428571428572,
      "grad_norm": 0.767754077911377,
      "learning_rate": 8.789408705326262e-05,
      "loss": 2.2504658699035645,
      "step": 4135
    },
    {
      "epoch": 0.9453714285714285,
      "grad_norm": 0.9333015084266663,
      "learning_rate": 8.72118103141066e-05,
      "loss": 2.342519760131836,
      "step": 4136
    },
    {
      "epoch": 0.9456,
      "grad_norm": 0.8814654350280762,
      "learning_rate": 8.653163206090325e-05,
      "loss": 2.291099786758423,
      "step": 4137
    },
    {
      "epoch": 0.9458285714285715,
      "grad_norm": 0.8094938397407532,
      "learning_rate": 8.585356106176093e-05,
      "loss": 2.581045627593994,
      "step": 4138
    },
    {
      "epoch": 0.9460571428571428,
      "grad_norm": 0.7125369310379028,
      "learning_rate": 8.517760605762378e-05,
      "loss": 2.451777458190918,
      "step": 4139
    },
    {
      "epoch": 0.9462857142857143,
      "grad_norm": 0.7890687584877014,
      "learning_rate": 8.450377576215876e-05,
      "loss": 2.220702886581421,
      "step": 4140
    },
    {
      "epoch": 0.9465142857142858,
      "grad_norm": 0.9261327385902405,
      "learning_rate": 8.383207886164365e-05,
      "loss": 2.3283791542053223,
      "step": 4141
    },
    {
      "epoch": 0.9467428571428571,
      "grad_norm": 0.934323251247406,
      "learning_rate": 8.31625240148545e-05,
      "loss": 2.4670796394348145,
      "step": 4142
    },
    {
      "epoch": 0.9469714285714286,
      "grad_norm": 0.7736852765083313,
      "learning_rate": 8.249511985295474e-05,
      "loss": 2.797874927520752,
      "step": 4143
    },
    {
      "epoch": 0.9472,
      "grad_norm": 0.8015835881233215,
      "learning_rate": 8.182987497938321e-05,
      "loss": 2.0399224758148193,
      "step": 4144
    },
    {
      "epoch": 0.9474285714285714,
      "grad_norm": 1.2334039211273193,
      "learning_rate": 8.116679796974389e-05,
      "loss": 2.1894595623016357,
      "step": 4145
    },
    {
      "epoch": 0.9476571428571429,
      "grad_norm": 0.7082759141921997,
      "learning_rate": 8.050589737169483e-05,
      "loss": 2.193028688430786,
      "step": 4146
    },
    {
      "epoch": 0.9478857142857143,
      "grad_norm": 1.144126296043396,
      "learning_rate": 7.984718170483813e-05,
      "loss": 2.478142261505127,
      "step": 4147
    },
    {
      "epoch": 0.9481142857142857,
      "grad_norm": 0.8239220380783081,
      "learning_rate": 7.919065946061043e-05,
      "loss": 2.561203956604004,
      "step": 4148
    },
    {
      "epoch": 0.9483428571428572,
      "grad_norm": 0.7284172773361206,
      "learning_rate": 7.853633910217303e-05,
      "loss": 2.5751166343688965,
      "step": 4149
    },
    {
      "epoch": 0.9485714285714286,
      "grad_norm": 0.7008686065673828,
      "learning_rate": 7.788422906430285e-05,
      "loss": 2.4091763496398926,
      "step": 4150
    },
    {
      "epoch": 0.9488,
      "grad_norm": 0.9741643667221069,
      "learning_rate": 7.723433775328384e-05,
      "loss": 2.2177367210388184,
      "step": 4151
    },
    {
      "epoch": 0.9490285714285714,
      "grad_norm": 0.716494619846344,
      "learning_rate": 7.65866735467988e-05,
      "loss": 2.0078837871551514,
      "step": 4152
    },
    {
      "epoch": 0.9492571428571429,
      "grad_norm": 1.3863645792007446,
      "learning_rate": 7.594124479382086e-05,
      "loss": 2.5040740966796875,
      "step": 4153
    },
    {
      "epoch": 0.9494857142857143,
      "grad_norm": 0.654206395149231,
      "learning_rate": 7.529805981450621e-05,
      "loss": 2.1617372035980225,
      "step": 4154
    },
    {
      "epoch": 0.9497142857142857,
      "grad_norm": 1.0160290002822876,
      "learning_rate": 7.465712690008686e-05,
      "loss": 2.5037240982055664,
      "step": 4155
    },
    {
      "epoch": 0.9499428571428571,
      "grad_norm": 0.884977400302887,
      "learning_rate": 7.401845431276377e-05,
      "loss": 2.4770407676696777,
      "step": 4156
    },
    {
      "epoch": 0.9501714285714286,
      "grad_norm": 0.786818265914917,
      "learning_rate": 7.338205028560002e-05,
      "loss": 2.3136143684387207,
      "step": 4157
    },
    {
      "epoch": 0.9504,
      "grad_norm": 0.9444769024848938,
      "learning_rate": 7.274792302241498e-05,
      "loss": 2.4006667137145996,
      "step": 4158
    },
    {
      "epoch": 0.9506285714285714,
      "grad_norm": 0.8820931911468506,
      "learning_rate": 7.211608069767867e-05,
      "loss": 2.734591245651245,
      "step": 4159
    },
    {
      "epoch": 0.9508571428571428,
      "grad_norm": 0.8113405108451843,
      "learning_rate": 7.1486531456406e-05,
      "loss": 2.713456869125366,
      "step": 4160
    },
    {
      "epoch": 0.9510857142857143,
      "grad_norm": 1.7557485103607178,
      "learning_rate": 7.085928341405193e-05,
      "loss": 2.57908296585083,
      "step": 4161
    },
    {
      "epoch": 0.9513142857142857,
      "grad_norm": 0.8909356594085693,
      "learning_rate": 7.02343446564072e-05,
      "loss": 2.737391710281372,
      "step": 4162
    },
    {
      "epoch": 0.9515428571428571,
      "grad_norm": 0.9573034048080444,
      "learning_rate": 6.961172323949353e-05,
      "loss": 2.3584892749786377,
      "step": 4163
    },
    {
      "epoch": 0.9517714285714286,
      "grad_norm": 0.6470253467559814,
      "learning_rate": 6.899142718946008e-05,
      "loss": 1.9702656269073486,
      "step": 4164
    },
    {
      "epoch": 0.952,
      "grad_norm": 0.7859966158866882,
      "learning_rate": 6.837346450247991e-05,
      "loss": 2.523341178894043,
      "step": 4165
    },
    {
      "epoch": 0.9522285714285714,
      "grad_norm": 0.7781214714050293,
      "learning_rate": 6.775784314464717e-05,
      "loss": 2.4332315921783447,
      "step": 4166
    },
    {
      "epoch": 0.9524571428571429,
      "grad_norm": 0.7646070718765259,
      "learning_rate": 6.714457105187383e-05,
      "loss": 2.2089195251464844,
      "step": 4167
    },
    {
      "epoch": 0.9526857142857142,
      "grad_norm": 0.9853861927986145,
      "learning_rate": 6.653365612978807e-05,
      "loss": 2.294888973236084,
      "step": 4168
    },
    {
      "epoch": 0.9529142857142857,
      "grad_norm": 0.9995832443237305,
      "learning_rate": 6.592510625363168e-05,
      "loss": 2.5007691383361816,
      "step": 4169
    },
    {
      "epoch": 0.9531428571428572,
      "grad_norm": 0.9752095937728882,
      "learning_rate": 6.531892926815925e-05,
      "loss": 2.3192267417907715,
      "step": 4170
    },
    {
      "epoch": 0.9533714285714285,
      "grad_norm": 1.108381748199463,
      "learning_rate": 6.471513298753634e-05,
      "loss": 2.138516664505005,
      "step": 4171
    },
    {
      "epoch": 0.9536,
      "grad_norm": 0.9743977189064026,
      "learning_rate": 6.41137251952392e-05,
      "loss": 2.453087329864502,
      "step": 4172
    },
    {
      "epoch": 0.9538285714285715,
      "grad_norm": 0.7523196339607239,
      "learning_rate": 6.351471364395448e-05,
      "loss": 2.4552533626556396,
      "step": 4173
    },
    {
      "epoch": 0.9540571428571428,
      "grad_norm": 0.7860829830169678,
      "learning_rate": 6.291810605547893e-05,
      "loss": 2.2350494861602783,
      "step": 4174
    },
    {
      "epoch": 0.9542857142857143,
      "grad_norm": 0.7984284162521362,
      "learning_rate": 6.232391012062016e-05,
      "loss": 2.569225311279297,
      "step": 4175
    },
    {
      "epoch": 0.9545142857142858,
      "grad_norm": 0.798438549041748,
      "learning_rate": 6.173213349909729e-05,
      "loss": 2.365705966949463,
      "step": 4176
    },
    {
      "epoch": 0.9547428571428571,
      "grad_norm": 1.0954794883728027,
      "learning_rate": 6.114278381944252e-05,
      "loss": 2.481187343597412,
      "step": 4177
    },
    {
      "epoch": 0.9549714285714286,
      "grad_norm": 0.8175252079963684,
      "learning_rate": 6.0555868678902335e-05,
      "loss": 2.4795260429382324,
      "step": 4178
    },
    {
      "epoch": 0.9552,
      "grad_norm": 0.8686597347259521,
      "learning_rate": 5.997139564334012e-05,
      "loss": 2.28759765625,
      "step": 4179
    },
    {
      "epoch": 0.9554285714285714,
      "grad_norm": 1.0643880367279053,
      "learning_rate": 5.9389372247138004e-05,
      "loss": 2.597892999649048,
      "step": 4180
    },
    {
      "epoch": 0.9556571428571429,
      "grad_norm": 0.8287124633789062,
      "learning_rate": 5.88098059931004e-05,
      "loss": 2.339578151702881,
      "step": 4181
    },
    {
      "epoch": 0.9558857142857143,
      "grad_norm": 0.8941729068756104,
      "learning_rate": 5.823270435235673e-05,
      "loss": 2.1930184364318848,
      "step": 4182
    },
    {
      "epoch": 0.9561142857142857,
      "grad_norm": 0.8289956450462341,
      "learning_rate": 5.765807476426527e-05,
      "loss": 2.2533695697784424,
      "step": 4183
    },
    {
      "epoch": 0.9563428571428572,
      "grad_norm": 0.8046292066574097,
      "learning_rate": 5.708592463631762e-05,
      "loss": 2.2636306285858154,
      "step": 4184
    },
    {
      "epoch": 0.9565714285714285,
      "grad_norm": 1.3272066116333008,
      "learning_rate": 5.651626134404267e-05,
      "loss": 2.3352274894714355,
      "step": 4185
    },
    {
      "epoch": 0.9568,
      "grad_norm": 0.7267163991928101,
      "learning_rate": 5.5949092230911794e-05,
      "loss": 1.9679417610168457,
      "step": 4186
    },
    {
      "epoch": 0.9570285714285715,
      "grad_norm": 1.0368057489395142,
      "learning_rate": 5.5384424608244165e-05,
      "loss": 2.095764636993408,
      "step": 4187
    },
    {
      "epoch": 0.9572571428571428,
      "grad_norm": 0.9301378726959229,
      "learning_rate": 5.482226575511262e-05,
      "loss": 2.292783737182617,
      "step": 4188
    },
    {
      "epoch": 0.9574857142857143,
      "grad_norm": 0.7853825688362122,
      "learning_rate": 5.4262622918249544e-05,
      "loss": 2.063793420791626,
      "step": 4189
    },
    {
      "epoch": 0.9577142857142857,
      "grad_norm": 0.9876095652580261,
      "learning_rate": 5.3705503311953646e-05,
      "loss": 2.3825173377990723,
      "step": 4190
    },
    {
      "epoch": 0.9579428571428571,
      "grad_norm": 1.071391224861145,
      "learning_rate": 5.315091411799699e-05,
      "loss": 2.8472647666931152,
      "step": 4191
    },
    {
      "epoch": 0.9581714285714286,
      "grad_norm": 1.7005032300949097,
      "learning_rate": 5.2598862485532416e-05,
      "loss": 2.6988353729248047,
      "step": 4192
    },
    {
      "epoch": 0.9584,
      "grad_norm": 0.8285865783691406,
      "learning_rate": 5.20493555310011e-05,
      "loss": 2.123180389404297,
      "step": 4193
    },
    {
      "epoch": 0.9586285714285714,
      "grad_norm": 0.7770256996154785,
      "learning_rate": 5.1502400338041156e-05,
      "loss": 2.50742769241333,
      "step": 4194
    },
    {
      "epoch": 0.9588571428571429,
      "grad_norm": 0.75770503282547,
      "learning_rate": 5.0958003957396276e-05,
      "loss": 2.341017961502075,
      "step": 4195
    },
    {
      "epoch": 0.9590857142857143,
      "grad_norm": 0.767578661441803,
      "learning_rate": 5.0416173406824666e-05,
      "loss": 2.1537106037139893,
      "step": 4196
    },
    {
      "epoch": 0.9593142857142857,
      "grad_norm": 0.9188908338546753,
      "learning_rate": 4.987691567100866e-05,
      "loss": 2.2111923694610596,
      "step": 4197
    },
    {
      "epoch": 0.9595428571428571,
      "grad_norm": 0.8361859321594238,
      "learning_rate": 4.9340237701464704e-05,
      "loss": 2.18149995803833,
      "step": 4198
    },
    {
      "epoch": 0.9597714285714286,
      "grad_norm": 0.862533688545227,
      "learning_rate": 4.880614641645389e-05,
      "loss": 2.5979225635528564,
      "step": 4199
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.6839424967765808,
      "learning_rate": 4.827464870089246e-05,
      "loss": 2.20473051071167,
      "step": 4200
    },
    {
      "epoch": 0.9602285714285714,
      "grad_norm": 0.8250691294670105,
      "learning_rate": 4.7745751406263163e-05,
      "loss": 2.3391149044036865,
      "step": 4201
    },
    {
      "epoch": 0.9604571428571429,
      "grad_norm": 0.8278523087501526,
      "learning_rate": 4.7219461350527235e-05,
      "loss": 2.405252456665039,
      "step": 4202
    },
    {
      "epoch": 0.9606857142857143,
      "grad_norm": 0.8121073842048645,
      "learning_rate": 4.6695785318036e-05,
      "loss": 2.2112042903900146,
      "step": 4203
    },
    {
      "epoch": 0.9609142857142857,
      "grad_norm": 1.0544458627700806,
      "learning_rate": 4.617473005944392e-05,
      "loss": 2.2535390853881836,
      "step": 4204
    },
    {
      "epoch": 0.9611428571428572,
      "grad_norm": 0.7789956331253052,
      "learning_rate": 4.5656302291621074e-05,
      "loss": 2.1782851219177246,
      "step": 4205
    },
    {
      "epoch": 0.9613714285714285,
      "grad_norm": 0.8562607169151306,
      "learning_rate": 4.5140508697567026e-05,
      "loss": 2.6605777740478516,
      "step": 4206
    },
    {
      "epoch": 0.9616,
      "grad_norm": 0.8653267025947571,
      "learning_rate": 4.462735592632439e-05,
      "loss": 2.368600845336914,
      "step": 4207
    },
    {
      "epoch": 0.9618285714285715,
      "grad_norm": 0.9979888796806335,
      "learning_rate": 4.411685059289314e-05,
      "loss": 2.20402193069458,
      "step": 4208
    },
    {
      "epoch": 0.9620571428571428,
      "grad_norm": 0.8527805805206299,
      "learning_rate": 4.360899927814555e-05,
      "loss": 2.7006454467773438,
      "step": 4209
    },
    {
      "epoch": 0.9622857142857143,
      "grad_norm": 0.8580187559127808,
      "learning_rate": 4.310380852874113e-05,
      "loss": 2.6834607124328613,
      "step": 4210
    },
    {
      "epoch": 0.9625142857142858,
      "grad_norm": 0.8418498635292053,
      "learning_rate": 4.260128485704226e-05,
      "loss": 2.190906524658203,
      "step": 4211
    },
    {
      "epoch": 0.9627428571428571,
      "grad_norm": 0.9905149936676025,
      "learning_rate": 4.2101434741030314e-05,
      "loss": 2.5000123977661133,
      "step": 4212
    },
    {
      "epoch": 0.9629714285714286,
      "grad_norm": 1.352835774421692,
      "learning_rate": 4.1604264624222285e-05,
      "loss": 2.5810484886169434,
      "step": 4213
    },
    {
      "epoch": 0.9632,
      "grad_norm": 0.8656988143920898,
      "learning_rate": 4.110978091558731e-05,
      "loss": 2.511427879333496,
      "step": 4214
    },
    {
      "epoch": 0.9634285714285714,
      "grad_norm": 0.872748076915741,
      "learning_rate": 4.0617989989464586e-05,
      "loss": 2.4156932830810547,
      "step": 4215
    },
    {
      "epoch": 0.9636571428571429,
      "grad_norm": 0.8131033182144165,
      "learning_rate": 4.012889818548068e-05,
      "loss": 2.4750351905822754,
      "step": 4216
    },
    {
      "epoch": 0.9638857142857142,
      "grad_norm": 0.7514214515686035,
      "learning_rate": 3.964251180846826e-05,
      "loss": 2.7563915252685547,
      "step": 4217
    },
    {
      "epoch": 0.9641142857142857,
      "grad_norm": 0.7351729273796082,
      "learning_rate": 3.915883712838444e-05,
      "loss": 2.534475326538086,
      "step": 4218
    },
    {
      "epoch": 0.9643428571428572,
      "grad_norm": 0.9124327301979065,
      "learning_rate": 3.867788038023018e-05,
      "loss": 2.3424129486083984,
      "step": 4219
    },
    {
      "epoch": 0.9645714285714285,
      "grad_norm": 0.9313555955886841,
      "learning_rate": 3.819964776396998e-05,
      "loss": 2.6452813148498535,
      "step": 4220
    },
    {
      "epoch": 0.9648,
      "grad_norm": 0.9557833671569824,
      "learning_rate": 3.772414544445163e-05,
      "loss": 2.4574167728424072,
      "step": 4221
    },
    {
      "epoch": 0.9650285714285715,
      "grad_norm": 0.8833873271942139,
      "learning_rate": 3.725137955132707e-05,
      "loss": 2.2242562770843506,
      "step": 4222
    },
    {
      "epoch": 0.9652571428571428,
      "grad_norm": 0.8217985033988953,
      "learning_rate": 3.678135617897318e-05,
      "loss": 2.42657470703125,
      "step": 4223
    },
    {
      "epoch": 0.9654857142857143,
      "grad_norm": 0.9071516394615173,
      "learning_rate": 3.63140813864134e-05,
      "loss": 2.78098726272583,
      "step": 4224
    },
    {
      "epoch": 0.9657142857142857,
      "grad_norm": 0.7697664499282837,
      "learning_rate": 3.58495611972394e-05,
      "loss": 2.336276054382324,
      "step": 4225
    },
    {
      "epoch": 0.9659428571428571,
      "grad_norm": 0.8726691007614136,
      "learning_rate": 3.5387801599533474e-05,
      "loss": 2.6282906532287598,
      "step": 4226
    },
    {
      "epoch": 0.9661714285714286,
      "grad_norm": 0.9830082058906555,
      "learning_rate": 3.4928808545791614e-05,
      "loss": 2.1199114322662354,
      "step": 4227
    },
    {
      "epoch": 0.9664,
      "grad_norm": 2.107774019241333,
      "learning_rate": 3.4472587952846496e-05,
      "loss": 2.9637856483459473,
      "step": 4228
    },
    {
      "epoch": 0.9666285714285714,
      "grad_norm": 0.7147778868675232,
      "learning_rate": 3.401914570179118e-05,
      "loss": 2.1531670093536377,
      "step": 4229
    },
    {
      "epoch": 0.9668571428571429,
      "grad_norm": 0.9241762161254883,
      "learning_rate": 3.356848763790349e-05,
      "loss": 2.1355700492858887,
      "step": 4230
    },
    {
      "epoch": 0.9670857142857143,
      "grad_norm": 0.9031846523284912,
      "learning_rate": 3.312061957057061e-05,
      "loss": 2.401029109954834,
      "step": 4231
    },
    {
      "epoch": 0.9673142857142857,
      "grad_norm": 0.8230318427085876,
      "learning_rate": 3.267554727321415e-05,
      "loss": 2.248922824859619,
      "step": 4232
    },
    {
      "epoch": 0.9675428571428571,
      "grad_norm": 0.9139852523803711,
      "learning_rate": 3.223327648321561e-05,
      "loss": 2.146406888961792,
      "step": 4233
    },
    {
      "epoch": 0.9677714285714286,
      "grad_norm": 0.8311507701873779,
      "learning_rate": 3.179381290184266e-05,
      "loss": 2.2313761711120605,
      "step": 4234
    },
    {
      "epoch": 0.968,
      "grad_norm": 0.9585351943969727,
      "learning_rate": 3.1357162194175616e-05,
      "loss": 2.3341569900512695,
      "step": 4235
    },
    {
      "epoch": 0.9682285714285714,
      "grad_norm": 0.7013172507286072,
      "learning_rate": 3.092332998903416e-05,
      "loss": 2.142116069793701,
      "step": 4236
    },
    {
      "epoch": 0.9684571428571429,
      "grad_norm": 0.7995169758796692,
      "learning_rate": 3.049232187890491e-05,
      "loss": 1.813408613204956,
      "step": 4237
    },
    {
      "epoch": 0.9686857142857143,
      "grad_norm": 0.7889042496681213,
      "learning_rate": 3.0064143419869627e-05,
      "loss": 2.5309579372406006,
      "step": 4238
    },
    {
      "epoch": 0.9689142857142857,
      "grad_norm": 0.9062716960906982,
      "learning_rate": 2.9638800131533027e-05,
      "loss": 2.441260814666748,
      "step": 4239
    },
    {
      "epoch": 0.9691428571428572,
      "grad_norm": 0.855225682258606,
      "learning_rate": 2.9216297496952206e-05,
      "loss": 1.8440110683441162,
      "step": 4240
    },
    {
      "epoch": 0.9693714285714286,
      "grad_norm": 0.9266871809959412,
      "learning_rate": 2.8796640962565372e-05,
      "loss": 2.553401470184326,
      "step": 4241
    },
    {
      "epoch": 0.9696,
      "grad_norm": 0.993186354637146,
      "learning_rate": 2.837983593812227e-05,
      "loss": 2.4229342937469482,
      "step": 4242
    },
    {
      "epoch": 0.9698285714285714,
      "grad_norm": 0.8699562549591064,
      "learning_rate": 2.796588779661388e-05,
      "loss": 2.5702738761901855,
      "step": 4243
    },
    {
      "epoch": 0.9700571428571428,
      "grad_norm": 0.8018946647644043,
      "learning_rate": 2.7554801874203506e-05,
      "loss": 2.2131402492523193,
      "step": 4244
    },
    {
      "epoch": 0.9702857142857143,
      "grad_norm": 1.0314863920211792,
      "learning_rate": 2.7146583470157803e-05,
      "loss": 2.3922054767608643,
      "step": 4245
    },
    {
      "epoch": 0.9705142857142857,
      "grad_norm": 0.7510316371917725,
      "learning_rate": 2.6741237846778677e-05,
      "loss": 2.429168701171875,
      "step": 4246
    },
    {
      "epoch": 0.9707428571428571,
      "grad_norm": 0.728290855884552,
      "learning_rate": 2.6338770229335174e-05,
      "loss": 2.3388748168945312,
      "step": 4247
    },
    {
      "epoch": 0.9709714285714286,
      "grad_norm": 0.904453456401825,
      "learning_rate": 2.5939185805996247e-05,
      "loss": 2.847438335418701,
      "step": 4248
    },
    {
      "epoch": 0.9712,
      "grad_norm": 1.0390218496322632,
      "learning_rate": 2.5542489727764072e-05,
      "loss": 1.993369460105896,
      "step": 4249
    },
    {
      "epoch": 0.9714285714285714,
      "grad_norm": 0.9634931087493896,
      "learning_rate": 2.514868710840723e-05,
      "loss": 2.256762981414795,
      "step": 4250
    },
    {
      "epoch": 0.9716571428571429,
      "grad_norm": 0.797387421131134,
      "learning_rate": 2.4757783024395242e-05,
      "loss": 2.4893760681152344,
      "step": 4251
    },
    {
      "epoch": 0.9718857142857142,
      "grad_norm": 0.8519687652587891,
      "learning_rate": 2.4369782514832655e-05,
      "loss": 2.2832579612731934,
      "step": 4252
    },
    {
      "epoch": 0.9721142857142857,
      "grad_norm": 0.7261501550674438,
      "learning_rate": 2.3984690581394603e-05,
      "loss": 2.5130224227905273,
      "step": 4253
    },
    {
      "epoch": 0.9723428571428572,
      "grad_norm": 0.6618391871452332,
      "learning_rate": 2.3602512188261844e-05,
      "loss": 2.3763580322265625,
      "step": 4254
    },
    {
      "epoch": 0.9725714285714285,
      "grad_norm": 0.8828823566436768,
      "learning_rate": 2.322325226205707e-05,
      "loss": 2.7498247623443604,
      "step": 4255
    },
    {
      "epoch": 0.9728,
      "grad_norm": 0.7459583282470703,
      "learning_rate": 2.2846915691781377e-05,
      "loss": 2.4620797634124756,
      "step": 4256
    },
    {
      "epoch": 0.9730285714285715,
      "grad_norm": 0.6890153288841248,
      "learning_rate": 2.2473507328751085e-05,
      "loss": 2.0447158813476562,
      "step": 4257
    },
    {
      "epoch": 0.9732571428571428,
      "grad_norm": 0.962544858455658,
      "learning_rate": 2.2103031986535323e-05,
      "loss": 2.252401828765869,
      "step": 4258
    },
    {
      "epoch": 0.9734857142857143,
      "grad_norm": 0.8952361941337585,
      "learning_rate": 2.17354944408939e-05,
      "loss": 2.4719104766845703,
      "step": 4259
    },
    {
      "epoch": 0.9737142857142858,
      "grad_norm": 1.988046407699585,
      "learning_rate": 2.13708994297159e-05,
      "loss": 2.286290407180786,
      "step": 4260
    },
    {
      "epoch": 0.9739428571428571,
      "grad_norm": 0.853951096534729,
      "learning_rate": 2.1009251652958388e-05,
      "loss": 2.553867816925049,
      "step": 4261
    },
    {
      "epoch": 0.9741714285714286,
      "grad_norm": 1.0330442190170288,
      "learning_rate": 2.0650555772585984e-05,
      "loss": 2.525975465774536,
      "step": 4262
    },
    {
      "epoch": 0.9744,
      "grad_norm": 1.410566806793213,
      "learning_rate": 2.0294816412510715e-05,
      "loss": 2.6928164958953857,
      "step": 4263
    },
    {
      "epoch": 0.9746285714285714,
      "grad_norm": 0.7999207377433777,
      "learning_rate": 1.9942038158532405e-05,
      "loss": 2.2980246543884277,
      "step": 4264
    },
    {
      "epoch": 0.9748571428571429,
      "grad_norm": 1.0216820240020752,
      "learning_rate": 1.959222555827955e-05,
      "loss": 2.669675350189209,
      "step": 4265
    },
    {
      "epoch": 0.9750857142857143,
      "grad_norm": 0.8256901502609253,
      "learning_rate": 1.924538312115068e-05,
      "loss": 2.6989684104919434,
      "step": 4266
    },
    {
      "epoch": 0.9753142857142857,
      "grad_norm": 1.0577876567840576,
      "learning_rate": 1.890151531825632e-05,
      "loss": 2.7043495178222656,
      "step": 4267
    },
    {
      "epoch": 0.9755428571428572,
      "grad_norm": 1.1108407974243164,
      "learning_rate": 1.856062658236121e-05,
      "loss": 2.565814733505249,
      "step": 4268
    },
    {
      "epoch": 0.9757714285714286,
      "grad_norm": 0.8701914548873901,
      "learning_rate": 1.8222721307827316e-05,
      "loss": 2.0539751052856445,
      "step": 4269
    },
    {
      "epoch": 0.976,
      "grad_norm": 1.6833677291870117,
      "learning_rate": 1.788780385055702e-05,
      "loss": 2.1637468338012695,
      "step": 4270
    },
    {
      "epoch": 0.9762285714285714,
      "grad_norm": 0.8719700574874878,
      "learning_rate": 1.7555878527937163e-05,
      "loss": 2.7337114810943604,
      "step": 4271
    },
    {
      "epoch": 0.9764571428571429,
      "grad_norm": 0.8190918564796448,
      "learning_rate": 1.722694961878324e-05,
      "loss": 2.3628735542297363,
      "step": 4272
    },
    {
      "epoch": 0.9766857142857143,
      "grad_norm": 0.8047924637794495,
      "learning_rate": 1.69010213632842e-05,
      "loss": 2.6722869873046875,
      "step": 4273
    },
    {
      "epoch": 0.9769142857142857,
      "grad_norm": 0.8145237565040588,
      "learning_rate": 1.6578097962948012e-05,
      "loss": 2.4263477325439453,
      "step": 4274
    },
    {
      "epoch": 0.9771428571428571,
      "grad_norm": 0.7969304919242859,
      "learning_rate": 1.6258183580547288e-05,
      "loss": 2.3497257232666016,
      "step": 4275
    },
    {
      "epoch": 0.9773714285714286,
      "grad_norm": 1.007598876953125,
      "learning_rate": 1.59412823400657e-05,
      "loss": 2.523102283477783,
      "step": 4276
    },
    {
      "epoch": 0.9776,
      "grad_norm": 0.8122606873512268,
      "learning_rate": 1.562739832664481e-05,
      "loss": 2.2069008350372314,
      "step": 4277
    },
    {
      "epoch": 0.9778285714285714,
      "grad_norm": 5.664660453796387,
      "learning_rate": 1.5316535586531482e-05,
      "loss": 2.7775516510009766,
      "step": 4278
    },
    {
      "epoch": 0.9780571428571428,
      "grad_norm": 0.7192355394363403,
      "learning_rate": 1.5008698127025544e-05,
      "loss": 2.2167978286743164,
      "step": 4279
    },
    {
      "epoch": 0.9782857142857143,
      "grad_norm": 0.9955984354019165,
      "learning_rate": 1.470388991642832e-05,
      "loss": 1.8872826099395752,
      "step": 4280
    },
    {
      "epoch": 0.9785142857142857,
      "grad_norm": 0.8365027904510498,
      "learning_rate": 1.4402114883991318e-05,
      "loss": 2.2315592765808105,
      "step": 4281
    },
    {
      "epoch": 0.9787428571428571,
      "grad_norm": 0.7317114472389221,
      "learning_rate": 1.4103376919865769e-05,
      "loss": 2.7112083435058594,
      "step": 4282
    },
    {
      "epoch": 0.9789714285714286,
      "grad_norm": 2.690063238143921,
      "learning_rate": 1.380767987505227e-05,
      "loss": 2.06575608253479,
      "step": 4283
    },
    {
      "epoch": 0.9792,
      "grad_norm": 1.0383996963500977,
      "learning_rate": 1.3515027561351285e-05,
      "loss": 2.4334206581115723,
      "step": 4284
    },
    {
      "epoch": 0.9794285714285714,
      "grad_norm": 1.0624008178710938,
      "learning_rate": 1.3225423751313942e-05,
      "loss": 2.266042947769165,
      "step": 4285
    },
    {
      "epoch": 0.9796571428571429,
      "grad_norm": 0.7746022343635559,
      "learning_rate": 1.2938872178193395e-05,
      "loss": 2.381157875061035,
      "step": 4286
    },
    {
      "epoch": 0.9798857142857142,
      "grad_norm": 0.7052043676376343,
      "learning_rate": 1.2655376535896851e-05,
      "loss": 2.457979440689087,
      "step": 4287
    },
    {
      "epoch": 0.9801142857142857,
      "grad_norm": 0.7743537425994873,
      "learning_rate": 1.2374940478937585e-05,
      "loss": 2.4029626846313477,
      "step": 4288
    },
    {
      "epoch": 0.9803428571428572,
      "grad_norm": 0.7237978577613831,
      "learning_rate": 1.2097567622388311e-05,
      "loss": 2.4557249546051025,
      "step": 4289
    },
    {
      "epoch": 0.9805714285714285,
      "grad_norm": 1.1614879369735718,
      "learning_rate": 1.1823261541834217e-05,
      "loss": 2.637601852416992,
      "step": 4290
    },
    {
      "epoch": 0.9808,
      "grad_norm": 0.8645465970039368,
      "learning_rate": 1.1552025773327007e-05,
      "loss": 2.219975471496582,
      "step": 4291
    },
    {
      "epoch": 0.9810285714285715,
      "grad_norm": 0.9520664215087891,
      "learning_rate": 1.1283863813339262e-05,
      "loss": 2.3600258827209473,
      "step": 4292
    },
    {
      "epoch": 0.9812571428571428,
      "grad_norm": 0.8261463642120361,
      "learning_rate": 1.1018779118719541e-05,
      "loss": 2.2766406536102295,
      "step": 4293
    },
    {
      "epoch": 0.9814857142857143,
      "grad_norm": 0.7261652946472168,
      "learning_rate": 1.0756775106647576e-05,
      "loss": 2.0963613986968994,
      "step": 4294
    },
    {
      "epoch": 0.9817142857142858,
      "grad_norm": 1.075636863708496,
      "learning_rate": 1.0497855154590342e-05,
      "loss": 2.3400349617004395,
      "step": 4295
    },
    {
      "epoch": 0.9819428571428571,
      "grad_norm": 0.7659756541252136,
      "learning_rate": 1.024202260025861e-05,
      "loss": 2.1917858123779297,
      "step": 4296
    },
    {
      "epoch": 0.9821714285714286,
      "grad_norm": 0.8116034269332886,
      "learning_rate": 9.989280741563689e-06,
      "loss": 2.6491165161132812,
      "step": 4297
    },
    {
      "epoch": 0.9824,
      "grad_norm": 1.0358010530471802,
      "learning_rate": 9.739632836575085e-06,
      "loss": 2.4408583641052246,
      "step": 4298
    },
    {
      "epoch": 0.9826285714285714,
      "grad_norm": 0.7110686898231506,
      "learning_rate": 9.493082103478518e-06,
      "loss": 2.195950508117676,
      "step": 4299
    },
    {
      "epoch": 0.9828571428571429,
      "grad_norm": 0.7188416123390198,
      "learning_rate": 9.249631720534362e-06,
      "loss": 2.331231117248535,
      "step": 4300
    },
    {
      "epoch": 0.9830857142857143,
      "grad_norm": 2.5726966857910156,
      "learning_rate": 9.00928482603669e-06,
      "loss": 2.737144947052002,
      "step": 4301
    },
    {
      "epoch": 0.9833142857142857,
      "grad_norm": 1.35175621509552,
      "learning_rate": 8.772044518272737e-06,
      "loss": 2.4907917976379395,
      "step": 4302
    },
    {
      "epoch": 0.9835428571428572,
      "grad_norm": 1.0570306777954102,
      "learning_rate": 8.537913855483165e-06,
      "loss": 2.6298954486846924,
      "step": 4303
    },
    {
      "epoch": 0.9837714285714285,
      "grad_norm": 0.9013113975524902,
      "learning_rate": 8.306895855822504e-06,
      "loss": 2.6776833534240723,
      "step": 4304
    },
    {
      "epoch": 0.984,
      "grad_norm": 1.308972716331482,
      "learning_rate": 8.078993497320165e-06,
      "loss": 2.5454554557800293,
      "step": 4305
    },
    {
      "epoch": 0.9842285714285715,
      "grad_norm": 0.824654221534729,
      "learning_rate": 7.854209717842232e-06,
      "loss": 2.5783257484436035,
      "step": 4306
    },
    {
      "epoch": 0.9844571428571428,
      "grad_norm": 0.8983831405639648,
      "learning_rate": 7.632547415053482e-06,
      "loss": 2.3404128551483154,
      "step": 4307
    },
    {
      "epoch": 0.9846857142857143,
      "grad_norm": 0.8850611448287964,
      "learning_rate": 7.414009446380099e-06,
      "loss": 2.146860122680664,
      "step": 4308
    },
    {
      "epoch": 0.9849142857142857,
      "grad_norm": 0.9134520888328552,
      "learning_rate": 7.198598628972653e-06,
      "loss": 2.4973936080932617,
      "step": 4309
    },
    {
      "epoch": 0.9851428571428571,
      "grad_norm": 0.7257719039916992,
      "learning_rate": 6.986317739670151e-06,
      "loss": 2.153285264968872,
      "step": 4310
    },
    {
      "epoch": 0.9853714285714286,
      "grad_norm": 0.7529383897781372,
      "learning_rate": 6.777169514963765e-06,
      "loss": 2.382678985595703,
      "step": 4311
    },
    {
      "epoch": 0.9856,
      "grad_norm": 0.9569660425186157,
      "learning_rate": 6.5711566509620295e-06,
      "loss": 2.798175811767578,
      "step": 4312
    },
    {
      "epoch": 0.9858285714285714,
      "grad_norm": 1.2803561687469482,
      "learning_rate": 6.368281803355691e-06,
      "loss": 2.328413724899292,
      "step": 4313
    },
    {
      "epoch": 0.9860571428571429,
      "grad_norm": 1.100085973739624,
      "learning_rate": 6.1685475873838044e-06,
      "loss": 2.242341995239258,
      "step": 4314
    },
    {
      "epoch": 0.9862857142857143,
      "grad_norm": 0.7768242955207825,
      "learning_rate": 5.971956577799725e-06,
      "loss": 2.3546223640441895,
      "step": 4315
    },
    {
      "epoch": 0.9865142857142857,
      "grad_norm": 0.7589062452316284,
      "learning_rate": 5.778511308838108e-06,
      "loss": 2.189220905303955,
      "step": 4316
    },
    {
      "epoch": 0.9867428571428571,
      "grad_norm": 0.7468703389167786,
      "learning_rate": 5.588214274182157e-06,
      "loss": 2.5350396633148193,
      "step": 4317
    },
    {
      "epoch": 0.9869714285714286,
      "grad_norm": 0.9406833648681641,
      "learning_rate": 5.401067926931624e-06,
      "loss": 2.454659938812256,
      "step": 4318
    },
    {
      "epoch": 0.9872,
      "grad_norm": 1.1347651481628418,
      "learning_rate": 5.217074679570943e-06,
      "loss": 2.3944337368011475,
      "step": 4319
    },
    {
      "epoch": 0.9874285714285714,
      "grad_norm": 0.9365127682685852,
      "learning_rate": 5.0362369039382845e-06,
      "loss": 2.4983654022216797,
      "step": 4320
    },
    {
      "epoch": 0.9876571428571429,
      "grad_norm": 0.8327993750572205,
      "learning_rate": 4.858556931194996e-06,
      "loss": 2.2017955780029297,
      "step": 4321
    },
    {
      "epoch": 0.9878857142857143,
      "grad_norm": 0.6799570918083191,
      "learning_rate": 4.684037051795487e-06,
      "loss": 2.3199617862701416,
      "step": 4322
    },
    {
      "epoch": 0.9881142857142857,
      "grad_norm": 0.9736019968986511,
      "learning_rate": 4.512679515457723e-06,
      "loss": 2.4450321197509766,
      "step": 4323
    },
    {
      "epoch": 0.9883428571428572,
      "grad_norm": 0.6809149980545044,
      "learning_rate": 4.3444865311342005e-06,
      "loss": 2.536651849746704,
      "step": 4324
    },
    {
      "epoch": 0.9885714285714285,
      "grad_norm": 0.9556301236152649,
      "learning_rate": 4.1794602669835976e-06,
      "loss": 2.2875006198883057,
      "step": 4325
    },
    {
      "epoch": 0.9888,
      "grad_norm": 0.8944602608680725,
      "learning_rate": 4.017602850342584e-06,
      "loss": 2.572681188583374,
      "step": 4326
    },
    {
      "epoch": 0.9890285714285715,
      "grad_norm": 1.2885172367095947,
      "learning_rate": 3.858916367698667e-06,
      "loss": 2.3619532585144043,
      "step": 4327
    },
    {
      "epoch": 0.9892571428571428,
      "grad_norm": 0.6782890558242798,
      "learning_rate": 3.703402864663058e-06,
      "loss": 2.347378969192505,
      "step": 4328
    },
    {
      "epoch": 0.9894857142857143,
      "grad_norm": 0.9873284697532654,
      "learning_rate": 3.5510643459445158e-06,
      "loss": 2.3571181297302246,
      "step": 4329
    },
    {
      "epoch": 0.9897142857142858,
      "grad_norm": 0.7358430027961731,
      "learning_rate": 3.4019027753233455e-06,
      "loss": 2.1870741844177246,
      "step": 4330
    },
    {
      "epoch": 0.9899428571428571,
      "grad_norm": 0.677939236164093,
      "learning_rate": 3.255920075626084e-06,
      "loss": 2.1365251541137695,
      "step": 4331
    },
    {
      "epoch": 0.9901714285714286,
      "grad_norm": 0.9645432829856873,
      "learning_rate": 3.113118128700909e-06,
      "loss": 2.583588123321533,
      "step": 4332
    },
    {
      "epoch": 0.9904,
      "grad_norm": 0.7994803786277771,
      "learning_rate": 2.973498775393102e-06,
      "loss": 2.117969512939453,
      "step": 4333
    },
    {
      "epoch": 0.9906285714285714,
      "grad_norm": 0.7625970840454102,
      "learning_rate": 2.8370638155215123e-06,
      "loss": 2.4162588119506836,
      "step": 4334
    },
    {
      "epoch": 0.9908571428571429,
      "grad_norm": 0.9249416589736938,
      "learning_rate": 2.703815007855298e-06,
      "loss": 2.5350770950317383,
      "step": 4335
    },
    {
      "epoch": 0.9910857142857142,
      "grad_norm": 0.8324594497680664,
      "learning_rate": 2.5737540700912774e-06,
      "loss": 2.1165356636047363,
      "step": 4336
    },
    {
      "epoch": 0.9913142857142857,
      "grad_norm": 0.84002685546875,
      "learning_rate": 2.446882678831697e-06,
      "loss": 2.7876474857330322,
      "step": 4337
    },
    {
      "epoch": 0.9915428571428572,
      "grad_norm": 0.8020087480545044,
      "learning_rate": 2.323202469562774e-06,
      "loss": 2.542198657989502,
      "step": 4338
    },
    {
      "epoch": 0.9917714285714285,
      "grad_norm": 0.8454816937446594,
      "learning_rate": 2.2027150366334125e-06,
      "loss": 2.266338348388672,
      "step": 4339
    },
    {
      "epoch": 0.992,
      "grad_norm": 0.8436320424079895,
      "learning_rate": 2.0854219332348545e-06,
      "loss": 2.2482993602752686,
      "step": 4340
    },
    {
      "epoch": 0.9922285714285715,
      "grad_norm": 1.1458792686462402,
      "learning_rate": 1.9713246713805587e-06,
      "loss": 2.272585868835449,
      "step": 4341
    },
    {
      "epoch": 0.9924571428571428,
      "grad_norm": 0.8911418914794922,
      "learning_rate": 1.8604247218866332e-06,
      "loss": 2.424161195755005,
      "step": 4342
    },
    {
      "epoch": 0.9926857142857143,
      "grad_norm": 1.0696945190429688,
      "learning_rate": 1.7527235143530718e-06,
      "loss": 2.420964241027832,
      "step": 4343
    },
    {
      "epoch": 0.9929142857142857,
      "grad_norm": 0.8531903028488159,
      "learning_rate": 1.6482224371451303e-06,
      "loss": 2.097731590270996,
      "step": 4344
    },
    {
      "epoch": 0.9931428571428571,
      "grad_norm": 0.7035436034202576,
      "learning_rate": 1.5469228373755362e-06,
      "loss": 2.5119457244873047,
      "step": 4345
    },
    {
      "epoch": 0.9933714285714286,
      "grad_norm": 0.9395299553871155,
      "learning_rate": 1.4488260208871395e-06,
      "loss": 2.4358959197998047,
      "step": 4346
    },
    {
      "epoch": 0.9936,
      "grad_norm": 1.036234736442566,
      "learning_rate": 1.3539332522359283e-06,
      "loss": 2.3983373641967773,
      "step": 4347
    },
    {
      "epoch": 0.9938285714285714,
      "grad_norm": 0.9232961535453796,
      "learning_rate": 1.2622457546749566e-06,
      "loss": 2.156259059906006,
      "step": 4348
    },
    {
      "epoch": 0.9940571428571429,
      "grad_norm": 0.8632964491844177,
      "learning_rate": 1.1737647101383863e-06,
      "loss": 2.692436695098877,
      "step": 4349
    },
    {
      "epoch": 0.9942857142857143,
      "grad_norm": 0.9203948974609375,
      "learning_rate": 1.0884912592263873e-06,
      "loss": 2.4156394004821777,
      "step": 4350
    },
    {
      "epoch": 0.9945142857142857,
      "grad_norm": 0.7897195816040039,
      "learning_rate": 1.006426501190233e-06,
      "loss": 2.253957986831665,
      "step": 4351
    },
    {
      "epoch": 0.9947428571428572,
      "grad_norm": 1.1531033515930176,
      "learning_rate": 9.275714939184221e-07,
      "loss": 2.6095309257507324,
      "step": 4352
    },
    {
      "epoch": 0.9949714285714286,
      "grad_norm": 1.0385409593582153,
      "learning_rate": 8.51927253922774e-07,
      "loss": 2.4509081840515137,
      "step": 4353
    },
    {
      "epoch": 0.9952,
      "grad_norm": 0.8101956844329834,
      "learning_rate": 7.79494756325494e-07,
      "loss": 2.390984058380127,
      "step": 4354
    },
    {
      "epoch": 0.9954285714285714,
      "grad_norm": 0.9755648374557495,
      "learning_rate": 7.102749348465165e-07,
      "loss": 2.432678699493408,
      "step": 4355
    },
    {
      "epoch": 0.9956571428571429,
      "grad_norm": 0.9269997477531433,
      "learning_rate": 6.442686817914878e-07,
      "loss": 2.264615535736084,
      "step": 4356
    },
    {
      "epoch": 0.9958857142857143,
      "grad_norm": 0.9389836192131042,
      "learning_rate": 5.814768480403021e-07,
      "loss": 2.283946990966797,
      "step": 4357
    },
    {
      "epoch": 0.9961142857142857,
      "grad_norm": 0.7383842468261719,
      "learning_rate": 5.21900243036083e-07,
      "loss": 2.1721200942993164,
      "step": 4358
    },
    {
      "epoch": 0.9963428571428572,
      "grad_norm": 1.0614696741104126,
      "learning_rate": 4.655396347747476e-07,
      "loss": 2.294919490814209,
      "step": 4359
    },
    {
      "epoch": 0.9965714285714286,
      "grad_norm": 1.0823782682418823,
      "learning_rate": 4.1239574979518024e-07,
      "loss": 2.7723255157470703,
      "step": 4360
    },
    {
      "epoch": 0.9968,
      "grad_norm": 0.6759255528450012,
      "learning_rate": 3.6246927316976875e-07,
      "loss": 2.444321632385254,
      "step": 4361
    },
    {
      "epoch": 0.9970285714285714,
      "grad_norm": 0.782198429107666,
      "learning_rate": 3.1576084849563315e-07,
      "loss": 2.4090681076049805,
      "step": 4362
    },
    {
      "epoch": 0.9972571428571428,
      "grad_norm": 0.7655799984931946,
      "learning_rate": 2.7227107788629915e-07,
      "loss": 2.0535831451416016,
      "step": 4363
    },
    {
      "epoch": 0.9974857142857143,
      "grad_norm": 0.7879562973976135,
      "learning_rate": 2.3200052196398202e-07,
      "loss": 2.322883367538452,
      "step": 4364
    },
    {
      "epoch": 0.9977142857142857,
      "grad_norm": 0.8462972044944763,
      "learning_rate": 1.949496998523148e-07,
      "loss": 2.5154497623443604,
      "step": 4365
    },
    {
      "epoch": 0.9979428571428571,
      "grad_norm": 4.999929428100586,
      "learning_rate": 1.61119089169659e-07,
      "loss": 2.5251517295837402,
      "step": 4366
    },
    {
      "epoch": 0.9981714285714286,
      "grad_norm": 0.9058965444564819,
      "learning_rate": 1.3050912602297071e-07,
      "loss": 2.388878345489502,
      "step": 4367
    },
    {
      "epoch": 0.9984,
      "grad_norm": 0.9303621649742126,
      "learning_rate": 1.0312020500216624e-07,
      "loss": 2.210564613342285,
      "step": 4368
    },
    {
      "epoch": 0.9986285714285714,
      "grad_norm": 0.8075791001319885,
      "learning_rate": 7.895267917501503e-08,
      "loss": 2.774789810180664,
      "step": 4369
    },
    {
      "epoch": 0.9988571428571429,
      "grad_norm": 0.6800602078437805,
      "learning_rate": 5.800686008264333e-08,
      "loss": 2.3129634857177734,
      "step": 4370
    },
    {
      "epoch": 0.9990857142857142,
      "grad_norm": 0.690558671951294,
      "learning_rate": 4.0283017735454064e-08,
      "loss": 2.307908535003662,
      "step": 4371
    },
    {
      "epoch": 0.9993142857142857,
      "grad_norm": 0.8091114163398743,
      "learning_rate": 2.5781380609712914e-08,
      "loss": 2.328756332397461,
      "step": 4372
    },
    {
      "epoch": 0.9995428571428572,
      "grad_norm": 0.9487956166267395,
      "learning_rate": 1.4502135644522962e-08,
      "loss": 2.4565272331237793,
      "step": 4373
    },
    {
      "epoch": 0.9997714285714285,
      "grad_norm": 0.7997180223464966,
      "learning_rate": 6.44542823952099e-09,
      "loss": 2.4178028106689453,
      "step": 4374
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.8936803936958313,
      "learning_rate": 1.6113622528513005e-09,
      "loss": 2.2666845321655273,
      "step": 4375
    }
  ],
  "logging_steps": 1,
  "max_steps": 4375,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 438,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 7.26934044957003e+18,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}