{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.5,
  "eval_steps": 500,
  "global_step": 838,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "grad_norm": 1.0434608134535555,
      "learning_rate": 2e-05,
      "loss": 0.7234,
      "step": 1
    },
    {
      "epoch": 0.0,
      "grad_norm": 0.9514763639201144,
      "learning_rate": 4e-05,
      "loss": 0.5916,
      "step": 2
    },
    {
      "epoch": 0.0,
      "grad_norm": 0.8903336235741379,
      "learning_rate": 6e-05,
      "loss": 0.5827,
      "step": 3
    },
    {
      "epoch": 0.0,
      "grad_norm": 0.764603046688215,
      "learning_rate": 8e-05,
      "loss": 0.5555,
      "step": 4
    },
    {
      "epoch": 0.0,
      "grad_norm": 0.8956967835024867,
      "learning_rate": 0.0001,
      "loss": 0.6073,
      "step": 5
    },
    {
      "epoch": 0.0,
      "grad_norm": 1.378526599131456,
      "learning_rate": 0.00012,
      "loss": 0.6204,
      "step": 6
    },
    {
      "epoch": 0.0,
      "grad_norm": 1.1442512069955262,
      "learning_rate": 0.00014,
      "loss": 0.5452,
      "step": 7
    },
    {
      "epoch": 0.0,
      "grad_norm": 0.9100913748328603,
      "learning_rate": 0.00016,
      "loss": 0.5858,
      "step": 8
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.6859557442210851,
      "learning_rate": 0.00018,
      "loss": 0.5182,
      "step": 9
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.8288606055568941,
      "learning_rate": 0.0002,
      "loss": 0.6083,
      "step": 10
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.0417686973414615,
      "learning_rate": 0.0001999999558168346,
      "loss": 0.65,
      "step": 11
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.8756017006672853,
      "learning_rate": 0.00019999982326737747,
      "loss": 0.5837,
      "step": 12
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.7284541589078422,
      "learning_rate": 0.0001999996023517457,
      "loss": 0.5738,
      "step": 13
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.8602799120413903,
      "learning_rate": 0.0001999992930701345,
      "loss": 0.595,
      "step": 14
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.7938840633060059,
      "learning_rate": 0.00019999889542281728,
      "loss": 0.5907,
      "step": 15
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.7022632853348306,
      "learning_rate": 0.00019999840941014525,
      "loss": 0.5513,
      "step": 16
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.6531781263691616,
      "learning_rate": 0.00019999783503254803,
      "loss": 0.5475,
      "step": 17
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.5808528104992969,
      "learning_rate": 0.0001999971722905331,
      "loss": 0.519,
      "step": 18
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.5452643486331965,
      "learning_rate": 0.00019999642118468614,
      "loss": 0.5421,
      "step": 19
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.5893567274117093,
      "learning_rate": 0.00019999558171567082,
      "loss": 0.6016,
      "step": 20
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.5039905113559068,
      "learning_rate": 0.000199994653884229,
      "loss": 0.6096,
      "step": 21
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.5236847425188783,
      "learning_rate": 0.00019999363769118055,
      "loss": 0.5845,
      "step": 22
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.4403777461558745,
      "learning_rate": 0.00019999253313742344,
      "loss": 0.5657,
      "step": 23
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.411935443568472,
      "learning_rate": 0.00019999134022393375,
      "loss": 0.5619,
      "step": 24
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.3669921362459581,
      "learning_rate": 0.0001999900589517656,
      "loss": 0.6115,
      "step": 25
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.3613613842516578,
      "learning_rate": 0.0001999886893220512,
      "loss": 0.5286,
      "step": 26
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.378560233146017,
      "learning_rate": 0.0001999872313360008,
      "loss": 0.5887,
      "step": 27
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.3562687135057843,
      "learning_rate": 0.00019998568499490283,
      "loss": 0.5598,
      "step": 28
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.34581851507208355,
      "learning_rate": 0.00019998405030012371,
      "loss": 0.5772,
      "step": 29
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.3812400800902662,
      "learning_rate": 0.00019998232725310796,
      "loss": 0.6154,
      "step": 30
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.2876811822631032,
      "learning_rate": 0.00019998051585537818,
      "loss": 0.4949,
      "step": 31
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.3291934945038139,
      "learning_rate": 0.00019997861610853503,
      "loss": 0.5388,
      "step": 32
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.35220229516562385,
      "learning_rate": 0.00019997662801425725,
      "loss": 0.5801,
      "step": 33
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.3908550087374589,
      "learning_rate": 0.00019997455157430165,
      "loss": 0.5783,
      "step": 34
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.38564043955621646,
      "learning_rate": 0.00019997238679050308,
      "loss": 0.5628,
      "step": 35
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.3248792879576579,
      "learning_rate": 0.00019997013366477453,
      "loss": 0.5896,
      "step": 36
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.33243708146621687,
      "learning_rate": 0.00019996779219910696,
      "loss": 0.5618,
      "step": 37
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.3589452470436555,
      "learning_rate": 0.00019996536239556942,
      "loss": 0.5387,
      "step": 38
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.3635189263065437,
      "learning_rate": 0.0001999628442563091,
      "loss": 0.629,
      "step": 39
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.36761442942017947,
      "learning_rate": 0.00019996023778355113,
      "loss": 0.6133,
      "step": 40
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.30331424568033827,
      "learning_rate": 0.00019995754297959882,
      "loss": 0.5377,
      "step": 41
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.3157701200247212,
      "learning_rate": 0.0001999547598468334,
      "loss": 0.6249,
      "step": 42
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.3112819252913729,
      "learning_rate": 0.00019995188838771425,
      "loss": 0.5424,
      "step": 43
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.33384944369487113,
      "learning_rate": 0.0001999489286047788,
      "loss": 0.6014,
      "step": 44
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.3330564524921197,
      "learning_rate": 0.00019994588050064243,
      "loss": 0.5469,
      "step": 45
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.2889450580843479,
      "learning_rate": 0.00019994274407799872,
      "loss": 0.512,
      "step": 46
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.3273617072745067,
      "learning_rate": 0.00019993951933961913,
      "loss": 0.5456,
      "step": 47
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.3228418979730564,
      "learning_rate": 0.00019993620628835332,
      "loss": 0.5716,
      "step": 48
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.3439278828333003,
      "learning_rate": 0.0001999328049271289,
      "loss": 0.5177,
      "step": 49
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.3186271172935729,
      "learning_rate": 0.0001999293152589515,
      "loss": 0.5502,
      "step": 50
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.38357922086488366,
      "learning_rate": 0.0001999257372869048,
      "loss": 0.6178,
      "step": 51
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.36013497860303273,
      "learning_rate": 0.00019992207101415053,
      "loss": 0.6278,
      "step": 52
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.3146595251755829,
      "learning_rate": 0.00019991831644392848,
      "loss": 0.5348,
      "step": 53
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.3986948949803995,
      "learning_rate": 0.00019991447357955639,
      "loss": 0.6331,
      "step": 54
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.3194176204715625,
      "learning_rate": 0.00019991054242443008,
      "loss": 0.5817,
      "step": 55
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.29564051537116465,
      "learning_rate": 0.00019990652298202335,
      "loss": 0.545,
      "step": 56
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.2908246398716361,
      "learning_rate": 0.00019990241525588804,
      "loss": 0.5294,
      "step": 57
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.3480952622658696,
      "learning_rate": 0.000199898219249654,
      "loss": 0.6282,
      "step": 58
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.4278607132919695,
      "learning_rate": 0.00019989393496702907,
      "loss": 0.7008,
      "step": 59
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.3088760407735635,
      "learning_rate": 0.00019988956241179912,
      "loss": 0.5747,
      "step": 60
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.3549589308890128,
      "learning_rate": 0.00019988510158782804,
      "loss": 0.615,
      "step": 61
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.28349959678885256,
      "learning_rate": 0.00019988055249905767,
      "loss": 0.577,
      "step": 62
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.35001480138074803,
      "learning_rate": 0.00019987591514950787,
      "loss": 0.5551,
      "step": 63
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.32895401860950285,
      "learning_rate": 0.00019987118954327654,
      "loss": 0.5617,
      "step": 64
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.347007326906862,
      "learning_rate": 0.00019986637568453945,
      "loss": 0.5935,
      "step": 65
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.30223641676037666,
      "learning_rate": 0.00019986147357755048,
      "loss": 0.5355,
      "step": 66
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.302279046184156,
      "learning_rate": 0.00019985648322664145,
      "loss": 0.5571,
      "step": 67
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.3181910281320864,
      "learning_rate": 0.00019985140463622215,
      "loss": 0.5198,
      "step": 68
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.32334719229096776,
      "learning_rate": 0.0001998462378107803,
      "loss": 0.5063,
      "step": 69
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.31038499461943353,
      "learning_rate": 0.0001998409827548817,
      "loss": 0.5805,
      "step": 70
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.3032049786542595,
      "learning_rate": 0.00019983563947316996,
      "loss": 0.564,
      "step": 71
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.3345643555445713,
      "learning_rate": 0.00019983020797036683,
      "loss": 0.5442,
      "step": 72
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.32583720357675877,
      "learning_rate": 0.00019982468825127187,
      "loss": 0.5674,
      "step": 73
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.30278224625704836,
      "learning_rate": 0.0001998190803207627,
      "loss": 0.569,
      "step": 74
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.29996902392483177,
      "learning_rate": 0.0001998133841837948,
      "loss": 0.6142,
      "step": 75
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.2947151895973628,
      "learning_rate": 0.00019980759984540168,
      "loss": 0.5084,
      "step": 76
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.33309005605837944,
      "learning_rate": 0.0001998017273106947,
      "loss": 0.5807,
      "step": 77
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.31281343399912853,
      "learning_rate": 0.00019979576658486325,
      "loss": 0.6299,
      "step": 78
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.30980839190781245,
      "learning_rate": 0.00019978971767317457,
      "loss": 0.5521,
      "step": 79
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.2988115356408324,
      "learning_rate": 0.00019978358058097388,
      "loss": 0.5645,
      "step": 80
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.2864799366004751,
      "learning_rate": 0.0001997773553136843,
      "loss": 0.5604,
      "step": 81
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.28284272149262185,
      "learning_rate": 0.00019977104187680688,
      "loss": 0.5964,
      "step": 82
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.2807639116477172,
      "learning_rate": 0.00019976464027592053,
      "loss": 0.5441,
      "step": 83
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.3169919212395633,
      "learning_rate": 0.00019975815051668217,
      "loss": 0.5672,
      "step": 84
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.2827524094344841,
      "learning_rate": 0.0001997515726048265,
      "loss": 0.5631,
      "step": 85
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.29538492598974014,
      "learning_rate": 0.00019974490654616625,
      "loss": 0.609,
      "step": 86
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.29397994414783907,
      "learning_rate": 0.0001997381523465919,
      "loss": 0.5723,
      "step": 87
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.2824005824065347,
      "learning_rate": 0.00019973131001207195,
      "loss": 0.5209,
      "step": 88
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.2847018611508931,
      "learning_rate": 0.00019972437954865265,
      "loss": 0.5617,
      "step": 89
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.2908040007926844,
      "learning_rate": 0.00019971736096245825,
      "loss": 0.5624,
      "step": 90
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.27754759063410545,
      "learning_rate": 0.00019971025425969083,
      "loss": 0.5353,
      "step": 91
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.2885964599083646,
      "learning_rate": 0.0001997030594466303,
      "loss": 0.5181,
      "step": 92
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.29372725474160666,
      "learning_rate": 0.00019969577652963444,
      "loss": 0.5757,
      "step": 93
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.32149915053639194,
      "learning_rate": 0.0001996884055151389,
      "loss": 0.552,
      "step": 94
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.2816717380191918,
      "learning_rate": 0.00019968094640965717,
      "loss": 0.4968,
      "step": 95
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.2719020140724135,
      "learning_rate": 0.00019967339921978062,
      "loss": 0.5503,
      "step": 96
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.28166729851780475,
      "learning_rate": 0.00019966576395217837,
      "loss": 0.5546,
      "step": 97
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.27817598279558775,
      "learning_rate": 0.0001996580406135975,
      "loss": 0.6145,
      "step": 98
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.3471492732103861,
      "learning_rate": 0.00019965022921086275,
      "loss": 0.6464,
      "step": 99
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.2838977359279957,
      "learning_rate": 0.00019964232975087687,
      "loss": 0.5576,
      "step": 100
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.2989119777268752,
      "learning_rate": 0.00019963434224062025,
      "loss": 0.5747,
      "step": 101
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.26079447242968457,
      "learning_rate": 0.0001996262666871512,
      "loss": 0.5144,
      "step": 102
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.2904578974664885,
      "learning_rate": 0.00019961810309760577,
      "loss": 0.5623,
      "step": 103
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.2682051539483259,
      "learning_rate": 0.00019960985147919778,
      "loss": 0.5722,
      "step": 104
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.3004061592870477,
      "learning_rate": 0.00019960151183921897,
      "loss": 0.5526,
      "step": 105
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.27675115608209533,
      "learning_rate": 0.00019959308418503877,
      "loss": 0.5859,
      "step": 106
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.26526760651496173,
      "learning_rate": 0.00019958456852410433,
      "loss": 0.5395,
      "step": 107
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.29513224606753785,
      "learning_rate": 0.0001995759648639406,
      "loss": 0.59,
      "step": 108
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.27848396362609984,
      "learning_rate": 0.00019956727321215044,
      "loss": 0.6076,
      "step": 109
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.29804931563512865,
      "learning_rate": 0.00019955849357641424,
      "loss": 0.5555,
      "step": 110
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.2756038079003531,
      "learning_rate": 0.00019954962596449024,
      "loss": 0.5542,
      "step": 111
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.28433942136507906,
      "learning_rate": 0.0001995406703842145,
      "loss": 0.5527,
      "step": 112
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.646736414676863,
      "learning_rate": 0.0001995316268435007,
      "loss": 0.7024,
      "step": 113
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.25030459489112267,
      "learning_rate": 0.00019952249535034025,
      "loss": 0.4928,
      "step": 114
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.30977646038996587,
      "learning_rate": 0.00019951327591280236,
      "loss": 0.5883,
      "step": 115
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.2854791945432696,
      "learning_rate": 0.0001995039685390339,
      "loss": 0.6318,
      "step": 116
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.3264119003161445,
      "learning_rate": 0.00019949457323725946,
      "loss": 0.5654,
      "step": 117
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.266512900378873,
      "learning_rate": 0.0001994850900157813,
      "loss": 0.5457,
      "step": 118
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.30259231663288877,
      "learning_rate": 0.0001994755188829794,
      "loss": 0.5828,
      "step": 119
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.33798187117798184,
      "learning_rate": 0.00019946585984731142,
      "loss": 0.5669,
      "step": 120
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.2769767197030659,
      "learning_rate": 0.00019945611291731274,
      "loss": 0.5619,
      "step": 121
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.2657775119925365,
      "learning_rate": 0.00019944627810159632,
      "loss": 0.59,
      "step": 122
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.2918370192930011,
      "learning_rate": 0.00019943635540885279,
      "loss": 0.5816,
      "step": 123
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.33417238245851544,
      "learning_rate": 0.00019942634484785052,
      "loss": 0.5921,
      "step": 124
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.2482914745160091,
      "learning_rate": 0.00019941624642743548,
      "loss": 0.5113,
      "step": 125
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.28272991564412037,
      "learning_rate": 0.0001994060601565313,
      "loss": 0.5543,
      "step": 126
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.35389083143384653,
      "learning_rate": 0.00019939578604413912,
      "loss": 0.5921,
      "step": 127
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.3426787529883331,
      "learning_rate": 0.00019938542409933787,
      "loss": 0.6073,
      "step": 128
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.25621145606980844,
      "learning_rate": 0.000199374974331284,
      "loss": 0.5639,
      "step": 129
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.2996022840977615,
      "learning_rate": 0.00019936443674921158,
      "loss": 0.5874,
      "step": 130
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.3283726508140091,
      "learning_rate": 0.0001993538113624323,
      "loss": 0.6295,
      "step": 131
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.2878523809947236,
      "learning_rate": 0.00019934309818033544,
      "loss": 0.5565,
      "step": 132
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.2764010556696928,
      "learning_rate": 0.0001993322972123878,
      "loss": 0.554,
      "step": 133
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.28827215996506156,
      "learning_rate": 0.0001993214084681338,
      "loss": 0.5788,
      "step": 134
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.2945105689397871,
      "learning_rate": 0.00019931043195719548,
      "loss": 0.5197,
      "step": 135
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.25439488455073433,
      "learning_rate": 0.00019929936768927232,
      "loss": 0.509,
      "step": 136
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.2746742447613063,
      "learning_rate": 0.00019928821567414144,
      "loss": 0.5479,
      "step": 137
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.3033960913538536,
      "learning_rate": 0.00019927697592165747,
      "loss": 0.5859,
      "step": 138
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.28486244663010424,
      "learning_rate": 0.00019926564844175256,
      "loss": 0.5951,
      "step": 139
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.3208016816012168,
      "learning_rate": 0.00019925423324443638,
      "loss": 0.5823,
      "step": 140
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.3005362808247367,
      "learning_rate": 0.00019924273033979613,
      "loss": 0.5652,
      "step": 141
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.26910338931514155,
      "learning_rate": 0.0001992311397379965,
      "loss": 0.5463,
      "step": 142
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.28718562659091934,
      "learning_rate": 0.00019921946144927966,
      "loss": 0.5245,
      "step": 143
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.2616953117781411,
      "learning_rate": 0.0001992076954839653,
      "loss": 0.5358,
      "step": 144
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.3107257337368987,
      "learning_rate": 0.00019919584185245062,
      "loss": 0.5536,
      "step": 145
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.28427245738065265,
      "learning_rate": 0.00019918390056521018,
      "loss": 0.6126,
      "step": 146
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.25398310452304734,
      "learning_rate": 0.00019917187163279605,
      "loss": 0.5068,
      "step": 147
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.2556256469730818,
      "learning_rate": 0.00019915975506583778,
      "loss": 0.5416,
      "step": 148
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.2611702329742577,
      "learning_rate": 0.00019914755087504236,
      "loss": 0.5276,
      "step": 149
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.2789045987166575,
      "learning_rate": 0.00019913525907119418,
      "loss": 0.5591,
      "step": 150
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.26837658503581957,
      "learning_rate": 0.000199122879665155,
      "loss": 0.6581,
      "step": 151
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.34601396912277804,
      "learning_rate": 0.0001991104126678641,
      "loss": 0.5394,
      "step": 152
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.25684957257052443,
      "learning_rate": 0.00019909785809033806,
      "loss": 0.5392,
      "step": 153
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.2906797315813485,
      "learning_rate": 0.00019908521594367098,
      "loss": 0.5185,
      "step": 154
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.2852843546202924,
      "learning_rate": 0.0001990724862390342,
      "loss": 0.5436,
      "step": 155
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.2875355300862882,
      "learning_rate": 0.0001990596689876765,
      "loss": 0.6009,
      "step": 156
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.32052910212305513,
      "learning_rate": 0.00019904676420092404,
      "loss": 0.5831,
      "step": 157
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.266884162852661,
      "learning_rate": 0.00019903377189018024,
      "loss": 0.5459,
      "step": 158
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.2957365744895018,
      "learning_rate": 0.000199020692066926,
      "loss": 0.5211,
      "step": 159
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.24951992931808137,
      "learning_rate": 0.00019900752474271945,
      "loss": 0.497,
      "step": 160
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.30509964150122953,
      "learning_rate": 0.0001989942699291961,
      "loss": 0.5812,
      "step": 161
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.2790293776337124,
      "learning_rate": 0.0001989809276380687,
      "loss": 0.5856,
      "step": 162
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.24940387850774506,
      "learning_rate": 0.00019896749788112737,
      "loss": 0.5281,
      "step": 163
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.2664890107453781,
      "learning_rate": 0.0001989539806702395,
      "loss": 0.524,
      "step": 164
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.2896608423493073,
      "learning_rate": 0.0001989403760173497,
      "loss": 0.5171,
      "step": 165
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.2544937836162412,
      "learning_rate": 0.00019892668393447997,
      "loss": 0.5546,
      "step": 166
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.2626626371027326,
      "learning_rate": 0.00019891290443372944,
      "loss": 0.5498,
      "step": 167
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.281410490858952,
      "learning_rate": 0.0001988990375272746,
      "loss": 0.5377,
      "step": 168
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.3376943176164128,
      "learning_rate": 0.0001988850832273691,
      "loss": 0.5469,
      "step": 169
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.2507691377758427,
      "learning_rate": 0.0001988710415463439,
      "loss": 0.549,
      "step": 170
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.27178996752570117,
      "learning_rate": 0.00019885691249660702,
      "loss": 0.5636,
      "step": 171
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.3359421766962587,
      "learning_rate": 0.00019884269609064386,
      "loss": 0.5957,
      "step": 172
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.2638709645045905,
      "learning_rate": 0.0001988283923410169,
      "loss": 0.5793,
      "step": 173
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.25585919726912226,
      "learning_rate": 0.00019881400126036582,
      "loss": 0.5817,
      "step": 174
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.2905067973645414,
      "learning_rate": 0.00019879952286140754,
      "loss": 0.5585,
      "step": 175
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.24197399766587002,
      "learning_rate": 0.0001987849571569361,
      "loss": 0.507,
      "step": 176
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.28898034252731664,
      "learning_rate": 0.0001987703041598226,
      "loss": 0.5981,
      "step": 177
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.26516349701479863,
      "learning_rate": 0.00019875556388301543,
      "loss": 0.56,
      "step": 178
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.27235027968517367,
      "learning_rate": 0.00019874073633953997,
      "loss": 0.5872,
      "step": 179
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.2692241567318253,
      "learning_rate": 0.00019872582154249884,
      "loss": 0.5397,
      "step": 180
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.2560507155942398,
      "learning_rate": 0.00019871081950507163,
      "loss": 0.5431,
      "step": 181
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.26691224099103567,
      "learning_rate": 0.00019869573024051517,
      "loss": 0.5608,
      "step": 182
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.2961375506924155,
      "learning_rate": 0.00019868055376216323,
      "loss": 0.5784,
      "step": 183
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.26055755072015874,
      "learning_rate": 0.00019866529008342673,
      "loss": 0.5369,
      "step": 184
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.2525359310079611,
      "learning_rate": 0.00019864993921779361,
      "loss": 0.5438,
      "step": 185
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.249327141855566,
      "learning_rate": 0.0001986345011788289,
      "loss": 0.5668,
      "step": 186
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.2983950007732028,
      "learning_rate": 0.00019861897598017457,
      "loss": 0.5271,
      "step": 187
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.25610455444964525,
      "learning_rate": 0.00019860336363554973,
      "loss": 0.6012,
      "step": 188
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.24760444410184018,
      "learning_rate": 0.0001985876641587504,
      "loss": 0.5066,
      "step": 189
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.2614264060863463,
      "learning_rate": 0.00019857187756364958,
      "loss": 0.5792,
      "step": 190
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.27219045408444215,
      "learning_rate": 0.00019855600386419744,
      "loss": 0.543,
      "step": 191
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.24606131498871828,
      "learning_rate": 0.00019854004307442088,
      "loss": 0.5676,
      "step": 192
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.28394763236035964,
      "learning_rate": 0.0001985239952084239,
      "loss": 0.6032,
      "step": 193
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.28350170917034406,
      "learning_rate": 0.0001985078602803874,
      "loss": 0.6264,
      "step": 194
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.24011552907338696,
      "learning_rate": 0.00019849163830456922,
      "loss": 0.4793,
      "step": 195
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.2561209086280576,
      "learning_rate": 0.00019847532929530415,
      "loss": 0.6198,
      "step": 196
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.23712886628255178,
      "learning_rate": 0.00019845893326700384,
      "loss": 0.4989,
      "step": 197
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.26720592489417233,
      "learning_rate": 0.00019844245023415685,
      "loss": 0.4934,
      "step": 198
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.2753251397417421,
      "learning_rate": 0.0001984258802113287,
      "loss": 0.5544,
      "step": 199
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.2557869713293877,
      "learning_rate": 0.0001984092232131616,
      "loss": 0.5643,
      "step": 200
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.2669651919314609,
      "learning_rate": 0.0001983924792543748,
      "loss": 0.5879,
      "step": 201
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.25579187132615644,
      "learning_rate": 0.00019837564834976432,
      "loss": 0.5742,
      "step": 202
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.2550207949421237,
      "learning_rate": 0.000198358730514203,
      "loss": 0.574,
      "step": 203
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.23565090455665763,
      "learning_rate": 0.0001983417257626405,
      "loss": 0.5299,
      "step": 204
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.236980034600526,
      "learning_rate": 0.00019832463411010331,
      "loss": 0.5199,
      "step": 205
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.2434029841189093,
      "learning_rate": 0.0001983074555716947,
      "loss": 0.5477,
      "step": 206
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.24771511082227154,
      "learning_rate": 0.00019829019016259468,
      "loss": 0.5697,
      "step": 207
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.23705880771864213,
      "learning_rate": 0.00019827283789806011,
      "loss": 0.521,
      "step": 208
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.24167295291353477,
      "learning_rate": 0.0001982553987934245,
      "loss": 0.558,
      "step": 209
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.2535406245356529,
      "learning_rate": 0.0001982378728640982,
      "loss": 0.5693,
      "step": 210
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.24865334136075806,
      "learning_rate": 0.00019822026012556818,
      "loss": 0.5499,
      "step": 211
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.2544751551481819,
      "learning_rate": 0.0001982025605933982,
      "loss": 0.5449,
      "step": 212
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.2599391794330939,
      "learning_rate": 0.0001981847742832287,
      "loss": 0.6222,
      "step": 213
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.23171934920449544,
      "learning_rate": 0.00019816690121077674,
      "loss": 0.5448,
      "step": 214
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.24380268930715565,
      "learning_rate": 0.00019814894139183614,
      "loss": 0.5773,
      "step": 215
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.2518755786374484,
      "learning_rate": 0.00019813089484227732,
      "loss": 0.5479,
      "step": 216
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.23133984467720642,
      "learning_rate": 0.00019811276157804733,
      "loss": 0.471,
      "step": 217
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.250968947574734,
      "learning_rate": 0.00019809454161516993,
      "loss": 0.5738,
      "step": 218
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.25976015596485974,
      "learning_rate": 0.00019807623496974537,
      "loss": 0.5592,
      "step": 219
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.24400199531999783,
      "learning_rate": 0.0001980578416579506,
      "loss": 0.5266,
      "step": 220
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.24001572180370875,
      "learning_rate": 0.00019803936169603912,
      "loss": 0.5843,
      "step": 221
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.22867208326764507,
      "learning_rate": 0.00019802079510034096,
      "loss": 0.518,
      "step": 222
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.2381724022911579,
      "learning_rate": 0.00019800214188726276,
      "loss": 0.5175,
      "step": 223
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.2700455397530704,
      "learning_rate": 0.00019798340207328766,
      "loss": 0.5804,
      "step": 224
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.24320219539003604,
      "learning_rate": 0.00019796457567497537,
      "loss": 0.5304,
      "step": 225
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.2370472610839002,
      "learning_rate": 0.0001979456627089621,
      "loss": 0.5671,
      "step": 226
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.26756212643991917,
      "learning_rate": 0.0001979266631919605,
      "loss": 0.5528,
      "step": 227
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.24929490389202372,
      "learning_rate": 0.00019790757714075979,
      "loss": 0.5407,
      "step": 228
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.23090595152280974,
      "learning_rate": 0.00019788840457222556,
      "loss": 0.5258,
      "step": 229
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.2979564627406142,
      "learning_rate": 0.0001978691455033,
      "loss": 0.5367,
      "step": 230
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.24228845587479894,
      "learning_rate": 0.0001978497999510015,
      "loss": 0.5344,
      "step": 231
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.25363482164729867,
      "learning_rate": 0.00019783036793242516,
      "loss": 0.5669,
      "step": 232
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.23622712417060854,
      "learning_rate": 0.00019781084946474226,
      "loss": 0.5797,
      "step": 233
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.21594682302559634,
      "learning_rate": 0.00019779124456520056,
      "loss": 0.5011,
      "step": 234
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.24211833950801223,
      "learning_rate": 0.0001977715532511242,
      "loss": 0.5164,
      "step": 235
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.2693820157715391,
      "learning_rate": 0.0001977517755399137,
      "loss": 0.5806,
      "step": 236
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.24734981937576542,
      "learning_rate": 0.00019773191144904586,
      "loss": 0.5233,
      "step": 237
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.2741663737268136,
      "learning_rate": 0.00019771196099607386,
      "loss": 0.5402,
      "step": 238
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.2521489075033339,
      "learning_rate": 0.00019769192419862716,
      "loss": 0.5862,
      "step": 239
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.2633255671236209,
      "learning_rate": 0.0001976718010744116,
      "loss": 0.548,
      "step": 240
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.2691399721238696,
      "learning_rate": 0.00019765159164120916,
      "loss": 0.5648,
      "step": 241
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.25501545746966797,
      "learning_rate": 0.00019763129591687827,
      "loss": 0.5602,
      "step": 242
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.3049976908839563,
      "learning_rate": 0.00019761091391935347,
      "loss": 0.5508,
      "step": 243
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.24467331400916031,
      "learning_rate": 0.00019759044566664558,
      "loss": 0.5229,
      "step": 244
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.27011009612786374,
      "learning_rate": 0.00019756989117684164,
      "loss": 0.5448,
      "step": 245
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.24427143044387528,
      "learning_rate": 0.00019754925046810493,
      "loss": 0.5435,
      "step": 246
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.22753961311031143,
      "learning_rate": 0.00019752852355867486,
      "loss": 0.5369,
      "step": 247
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.23865046003559778,
      "learning_rate": 0.00019750771046686704,
      "loss": 0.5354,
      "step": 248
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.2736283930569903,
      "learning_rate": 0.00019748681121107325,
      "loss": 0.5588,
      "step": 249
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.24727127426749082,
      "learning_rate": 0.00019746582580976136,
      "loss": 0.5753,
      "step": 250
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.2828829340227291,
      "learning_rate": 0.00019744475428147546,
      "loss": 0.6793,
      "step": 251
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.21818184663896711,
      "learning_rate": 0.00019742359664483563,
      "loss": 0.5248,
      "step": 252
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.2320708306833192,
      "learning_rate": 0.00019740235291853812,
      "loss": 0.5461,
      "step": 253
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.25703347930088793,
      "learning_rate": 0.00019738102312135523,
      "loss": 0.5713,
      "step": 254
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.24399588128874033,
      "learning_rate": 0.0001973596072721353,
      "loss": 0.5178,
      "step": 255
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.2229881452119291,
      "learning_rate": 0.00019733810538980281,
      "loss": 0.5144,
      "step": 256
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.23889465035265364,
      "learning_rate": 0.0001973165174933581,
      "loss": 0.5727,
      "step": 257
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.25790569964877214,
      "learning_rate": 0.0001972948436018776,
      "loss": 0.5659,
      "step": 258
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.22511338701135042,
      "learning_rate": 0.00019727308373451377,
      "loss": 0.5292,
      "step": 259
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.23111498863739158,
      "learning_rate": 0.000197251237910495,
      "loss": 0.5267,
      "step": 260
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.23740021982137896,
      "learning_rate": 0.00019722930614912563,
      "loss": 0.5499,
      "step": 261
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.24020258332985106,
      "learning_rate": 0.00019720728846978598,
      "loss": 0.5604,
      "step": 262
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.23947573439011133,
      "learning_rate": 0.00019718518489193225,
      "loss": 0.5638,
      "step": 263
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.23526187217481284,
      "learning_rate": 0.00019716299543509654,
      "loss": 0.5436,
      "step": 264
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.2505831003191156,
      "learning_rate": 0.00019714072011888686,
      "loss": 0.5039,
      "step": 265
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.21592487431784965,
      "learning_rate": 0.00019711835896298713,
      "loss": 0.484,
      "step": 266
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.26528309878122613,
      "learning_rate": 0.00019709591198715707,
      "loss": 0.539,
      "step": 267
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.21635789850987178,
      "learning_rate": 0.00019707337921123221,
      "loss": 0.5553,
      "step": 268
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.23671151623321054,
      "learning_rate": 0.00019705076065512398,
      "loss": 0.4968,
      "step": 269
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.25400871793456326,
      "learning_rate": 0.00019702805633881957,
      "loss": 0.5982,
      "step": 270
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.2622810971314154,
      "learning_rate": 0.0001970052662823819,
      "loss": 0.5879,
      "step": 271
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.22931814830456296,
      "learning_rate": 0.00019698239050594977,
      "loss": 0.5611,
      "step": 272
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.213695866193263,
      "learning_rate": 0.0001969594290297376,
      "loss": 0.5386,
      "step": 273
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.2431252328609808,
      "learning_rate": 0.00019693638187403563,
      "loss": 0.6039,
      "step": 274
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.23108667253454973,
      "learning_rate": 0.00019691324905920984,
      "loss": 0.5579,
      "step": 275
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.22718831415064272,
      "learning_rate": 0.0001968900306057018,
      "loss": 0.5196,
      "step": 276
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.23632362967796033,
      "learning_rate": 0.0001968667265340288,
      "loss": 0.5336,
      "step": 277
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.2421878973201691,
      "learning_rate": 0.00019684333686478383,
      "loss": 0.5928,
      "step": 278
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.225775487602821,
      "learning_rate": 0.00019681986161863542,
      "loss": 0.552,
      "step": 279
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.23037375759338816,
      "learning_rate": 0.00019679630081632782,
      "loss": 0.4983,
      "step": 280
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.24684136612832333,
      "learning_rate": 0.00019677265447868086,
      "loss": 0.5655,
      "step": 281
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.2412756534364674,
      "learning_rate": 0.0001967489226265899,
      "loss": 0.5063,
      "step": 282
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.22005765622474396,
      "learning_rate": 0.00019672510528102597,
      "loss": 0.5188,
      "step": 283
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.25071531725514384,
      "learning_rate": 0.0001967012024630355,
      "loss": 0.5938,
      "step": 284
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.22139592405512468,
      "learning_rate": 0.00019667721419374065,
      "loss": 0.4917,
      "step": 285
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.23067244251762076,
      "learning_rate": 0.00019665314049433888,
      "loss": 0.5584,
      "step": 286
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.23829875535152545,
      "learning_rate": 0.00019662898138610323,
      "loss": 0.5264,
      "step": 287
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.2641034663020514,
      "learning_rate": 0.00019660473689038228,
      "loss": 0.5805,
      "step": 288
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.22321690487140503,
      "learning_rate": 0.00019658040702859997,
      "loss": 0.5529,
      "step": 289
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.2502632163555589,
      "learning_rate": 0.00019655599182225565,
      "loss": 0.5347,
      "step": 290
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.2392608020883604,
      "learning_rate": 0.00019653149129292426,
      "loss": 0.5263,
      "step": 291
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.2539237490519494,
      "learning_rate": 0.00019650690546225592,
      "loss": 0.5156,
      "step": 292
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.21964099103511592,
      "learning_rate": 0.00019648223435197627,
      "loss": 0.5101,
      "step": 293
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.24992985700157416,
      "learning_rate": 0.00019645747798388628,
      "loss": 0.5621,
      "step": 294
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.2559439615345381,
      "learning_rate": 0.0001964326363798622,
      "loss": 0.5753,
      "step": 295
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.2504368010690795,
      "learning_rate": 0.00019640770956185567,
      "loss": 0.5558,
      "step": 296
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.21022187251218089,
      "learning_rate": 0.0001963826975518936,
      "loss": 0.5322,
      "step": 297
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.2422143856532352,
      "learning_rate": 0.00019635760037207817,
      "loss": 0.538,
      "step": 298
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.23174554470584352,
      "learning_rate": 0.00019633241804458687,
      "loss": 0.5545,
      "step": 299
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.267070426953347,
      "learning_rate": 0.00019630715059167238,
      "loss": 0.5632,
      "step": 300
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.22256523603127878,
      "learning_rate": 0.0001962817980356626,
      "loss": 0.545,
      "step": 301
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.24403300497950306,
      "learning_rate": 0.0001962563603989607,
      "loss": 0.5448,
      "step": 302
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.2193103621019292,
      "learning_rate": 0.00019623083770404492,
      "loss": 0.5064,
      "step": 303
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.23299992325798072,
      "learning_rate": 0.0001962052299734688,
      "loss": 0.5192,
      "step": 304
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.2371054150083945,
      "learning_rate": 0.00019617953722986096,
      "loss": 0.5157,
      "step": 305
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.2436064901189273,
      "learning_rate": 0.00019615375949592504,
      "loss": 0.5672,
      "step": 306
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.25098365347678436,
      "learning_rate": 0.00019612789679443997,
      "loss": 0.5548,
      "step": 307
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.2319425382974216,
      "learning_rate": 0.00019610194914825962,
      "loss": 0.5293,
      "step": 308
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.24156576209403272,
      "learning_rate": 0.000196075916580313,
      "loss": 0.5672,
      "step": 309
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.2337383575844323,
      "learning_rate": 0.0001960497991136041,
      "loss": 0.5509,
      "step": 310
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.23799692988502053,
      "learning_rate": 0.00019602359677121199,
      "loss": 0.5604,
      "step": 311
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.2296728275122706,
      "learning_rate": 0.0001959973095762907,
      "loss": 0.5371,
      "step": 312
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.22381626870035518,
      "learning_rate": 0.00019597093755206936,
      "loss": 0.5465,
      "step": 313
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.23335681761234933,
      "learning_rate": 0.00019594448072185182,
      "loss": 0.5386,
      "step": 314
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.22582265649304345,
      "learning_rate": 0.00019591793910901707,
      "loss": 0.543,
      "step": 315
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.2439330072441743,
      "learning_rate": 0.00019589131273701894,
      "loss": 0.5177,
      "step": 316
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.2138593422237162,
      "learning_rate": 0.00019586460162938622,
      "loss": 0.5157,
      "step": 317
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.24003613679646058,
      "learning_rate": 0.00019583780580972253,
      "loss": 0.5611,
      "step": 318
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.2552582800734971,
      "learning_rate": 0.00019581092530170633,
      "loss": 0.5922,
      "step": 319
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.21898423827197905,
      "learning_rate": 0.00019578396012909092,
      "loss": 0.5272,
      "step": 320
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.22013525107478477,
      "learning_rate": 0.00019575691031570446,
      "loss": 0.5184,
      "step": 321
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.21113180640163418,
      "learning_rate": 0.00019572977588544986,
      "loss": 0.5134,
      "step": 322
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.22335860079943387,
      "learning_rate": 0.00019570255686230485,
      "loss": 0.5227,
      "step": 323
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.23006684721287293,
      "learning_rate": 0.00019567525327032187,
      "loss": 0.5885,
      "step": 324
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.21933564641390155,
      "learning_rate": 0.0001956478651336281,
      "loss": 0.5598,
      "step": 325
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.21770749652400337,
      "learning_rate": 0.00019562039247642546,
      "loss": 0.5082,
      "step": 326
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.22800344133653658,
      "learning_rate": 0.00019559283532299043,
      "loss": 0.5539,
      "step": 327
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.2385574924163353,
      "learning_rate": 0.00019556519369767438,
      "loss": 0.5497,
      "step": 328
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.23099538598172079,
      "learning_rate": 0.0001955374676249031,
      "loss": 0.5138,
      "step": 329
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.21517217725478144,
      "learning_rate": 0.0001955096571291772,
      "loss": 0.5051,
      "step": 330
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.21535771106277588,
      "learning_rate": 0.0001954817622350717,
      "loss": 0.524,
      "step": 331
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.20361747402971658,
      "learning_rate": 0.00019545378296723635,
      "loss": 0.4989,
      "step": 332
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.24644921068325687,
      "learning_rate": 0.0001954257193503954,
      "loss": 0.5927,
      "step": 333
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.24765268362172385,
      "learning_rate": 0.0001953975714093476,
      "loss": 0.5451,
      "step": 334
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.20846277824915477,
      "learning_rate": 0.00019536933916896633,
      "loss": 0.5259,
      "step": 335
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.2457371199220107,
      "learning_rate": 0.00019534102265419932,
      "loss": 0.5784,
      "step": 336
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.23029745387228598,
      "learning_rate": 0.00019531262189006882,
      "loss": 0.5918,
      "step": 337
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.2387820151516941,
      "learning_rate": 0.0001952841369016716,
      "loss": 0.5576,
      "step": 338
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.226451643448924,
      "learning_rate": 0.00019525556771417875,
      "loss": 0.5241,
      "step": 339
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.22086691724075064,
      "learning_rate": 0.00019522691435283585,
      "loss": 0.5392,
      "step": 340
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.2259720671796772,
      "learning_rate": 0.00019519817684296285,
      "loss": 0.516,
      "step": 341
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.2244741513315317,
      "learning_rate": 0.00019516935520995393,
      "loss": 0.569,
      "step": 342
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.23890602213836412,
      "learning_rate": 0.0001951404494792778,
      "loss": 0.524,
      "step": 343
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.22136745892767679,
      "learning_rate": 0.00019511145967647737,
      "loss": 0.5472,
      "step": 344
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.22275740066078306,
      "learning_rate": 0.00019508238582716984,
      "loss": 0.5553,
      "step": 345
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.21225155652808625,
      "learning_rate": 0.00019505322795704676,
      "loss": 0.5302,
      "step": 346
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.22704101844750724,
      "learning_rate": 0.0001950239860918738,
      "loss": 0.5485,
      "step": 347
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.2135110250199134,
      "learning_rate": 0.00019499466025749097,
      "loss": 0.5343,
      "step": 348
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.22772242632973722,
      "learning_rate": 0.00019496525047981242,
      "loss": 0.5159,
      "step": 349
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.4444297049160113,
      "learning_rate": 0.00019493575678482649,
      "loss": 0.5121,
      "step": 350
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.226632712040011,
      "learning_rate": 0.0001949061791985957,
      "loss": 0.5304,
      "step": 351
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.22132303156586286,
      "learning_rate": 0.00019487651774725663,
      "loss": 0.4817,
      "step": 352
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.23206581340772667,
      "learning_rate": 0.00019484677245702004,
      "loss": 0.5258,
      "step": 353
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.2374903834541946,
      "learning_rate": 0.0001948169433541708,
      "loss": 0.5318,
      "step": 354
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.22896458770920267,
      "learning_rate": 0.00019478703046506773,
      "loss": 0.4806,
      "step": 355
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.21040301706147688,
      "learning_rate": 0.00019475703381614375,
      "loss": 0.5144,
      "step": 356
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.21179618454444762,
      "learning_rate": 0.00019472695343390585,
      "loss": 0.524,
      "step": 357
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.20436614333218908,
      "learning_rate": 0.00019469678934493488,
      "loss": 0.501,
      "step": 358
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.2478256130980173,
      "learning_rate": 0.0001946665415758858,
      "loss": 0.5386,
      "step": 359
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.226116084636948,
      "learning_rate": 0.00019463621015348748,
      "loss": 0.5101,
      "step": 360
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.21838947264457534,
      "learning_rate": 0.00019460579510454263,
      "loss": 0.5296,
      "step": 361
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.2152879498375444,
      "learning_rate": 0.00019457529645592792,
      "loss": 0.512,
      "step": 362
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.22514971802642378,
      "learning_rate": 0.00019454471423459389,
      "loss": 0.5593,
      "step": 363
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.23402386101532432,
      "learning_rate": 0.00019451404846756494,
      "loss": 0.555,
      "step": 364
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.2244514913016572,
      "learning_rate": 0.00019448329918193927,
      "loss": 0.5689,
      "step": 365
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.22260707231596893,
      "learning_rate": 0.00019445246640488893,
      "loss": 0.6062,
      "step": 366
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.21791090145253736,
      "learning_rate": 0.00019442155016365965,
      "loss": 0.531,
      "step": 367
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.21895889257567258,
      "learning_rate": 0.00019439055048557101,
      "loss": 0.5538,
      "step": 368
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.21306696799872818,
      "learning_rate": 0.00019435946739801633,
      "loss": 0.5673,
      "step": 369
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.2294607768810707,
      "learning_rate": 0.00019432830092846253,
      "loss": 0.5855,
      "step": 370
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.22758409665267085,
      "learning_rate": 0.0001942970511044503,
      "loss": 0.5783,
      "step": 371
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.21334100614200935,
      "learning_rate": 0.00019426571795359398,
      "loss": 0.5056,
      "step": 372
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.24187624093875965,
      "learning_rate": 0.0001942343015035815,
      "loss": 0.543,
      "step": 373
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.2275714845035408,
      "learning_rate": 0.00019420280178217443,
      "loss": 0.5329,
      "step": 374
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.23237641477505608,
      "learning_rate": 0.00019417121881720793,
      "loss": 0.5134,
      "step": 375
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.25196886008416386,
      "learning_rate": 0.0001941395526365907,
      "loss": 0.6023,
      "step": 376
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.22418514390796682,
      "learning_rate": 0.00019410780326830498,
      "loss": 0.5529,
      "step": 377
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.21438856736265666,
      "learning_rate": 0.0001940759707404065,
      "loss": 0.5134,
      "step": 378
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.2331754234870151,
      "learning_rate": 0.00019404405508102455,
      "loss": 0.5406,
      "step": 379
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.24908239322819828,
      "learning_rate": 0.00019401205631836178,
      "loss": 0.5377,
      "step": 380
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.21332745391417657,
      "learning_rate": 0.00019397997448069435,
      "loss": 0.5025,
      "step": 381
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.20749658696001225,
      "learning_rate": 0.00019394780959637177,
      "loss": 0.5257,
      "step": 382
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.2237716482529178,
      "learning_rate": 0.000193915561693817,
      "loss": 0.5,
      "step": 383
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.25234282015654147,
      "learning_rate": 0.00019388323080152633,
      "loss": 0.5753,
      "step": 384
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.2252939326339829,
      "learning_rate": 0.00019385081694806936,
      "loss": 0.5662,
      "step": 385
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.21979629294660186,
      "learning_rate": 0.00019381832016208904,
      "loss": 0.5141,
      "step": 386
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.24762535901866153,
      "learning_rate": 0.0001937857404723016,
      "loss": 0.6193,
      "step": 387
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.25032044234085526,
      "learning_rate": 0.00019375307790749647,
      "loss": 0.5024,
      "step": 388
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.22892425302508923,
      "learning_rate": 0.0001937203324965364,
      "loss": 0.5401,
      "step": 389
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.2461599771002527,
      "learning_rate": 0.0001936875042683573,
      "loss": 0.5301,
      "step": 390
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.22363255721865732,
      "learning_rate": 0.00019365459325196825,
      "loss": 0.5538,
      "step": 391
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.22482667580972365,
      "learning_rate": 0.00019362159947645152,
      "loss": 0.4928,
      "step": 392
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.22869596173751142,
      "learning_rate": 0.00019358852297096253,
      "loss": 0.5546,
      "step": 393
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.2274546469780496,
      "learning_rate": 0.00019355536376472972,
      "loss": 0.5763,
      "step": 394
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.21284874650406885,
      "learning_rate": 0.0001935221218870547,
      "loss": 0.5778,
      "step": 395
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.23158847478661296,
      "learning_rate": 0.0001934887973673121,
      "loss": 0.5654,
      "step": 396
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.24510006704514478,
      "learning_rate": 0.0001934553902349496,
      "loss": 0.5053,
      "step": 397
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.20330878586204656,
      "learning_rate": 0.00019342190051948777,
      "loss": 0.5171,
      "step": 398
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.2131804710318274,
      "learning_rate": 0.0001933883282505203,
      "loss": 0.5286,
      "step": 399
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.23297933515492006,
      "learning_rate": 0.00019335467345771377,
      "loss": 0.5593,
      "step": 400
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.24611434220143105,
      "learning_rate": 0.0001933209361708077,
      "loss": 0.604,
      "step": 401
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.23281321736249425,
      "learning_rate": 0.00019328711641961445,
      "loss": 0.5579,
      "step": 402
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.21399822113981087,
      "learning_rate": 0.00019325321423401933,
      "loss": 0.5661,
      "step": 403
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.22113107520723113,
      "learning_rate": 0.00019321922964398046,
      "loss": 0.5789,
      "step": 404
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.23262880002546846,
      "learning_rate": 0.00019318516267952874,
      "loss": 0.5447,
      "step": 405
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.24962941770082592,
      "learning_rate": 0.00019315101337076792,
      "loss": 0.5512,
      "step": 406
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.22210049422713798,
      "learning_rate": 0.0001931167817478745,
      "loss": 0.5427,
      "step": 407
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.22647809883332484,
      "learning_rate": 0.0001930824678410977,
      "loss": 0.4888,
      "step": 408
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.23660763255678552,
      "learning_rate": 0.00019304807168075944,
      "loss": 0.5755,
      "step": 409
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.2354103448271752,
      "learning_rate": 0.00019301359329725436,
      "loss": 0.5265,
      "step": 410
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.24322261128085423,
      "learning_rate": 0.00019297903272104977,
      "loss": 0.5291,
      "step": 411
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.20525199182278092,
      "learning_rate": 0.00019294438998268554,
      "loss": 0.4996,
      "step": 412
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.24678535182755174,
      "learning_rate": 0.00019290966511277422,
      "loss": 0.567,
      "step": 413
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.22165331172413838,
      "learning_rate": 0.00019287485814200087,
      "loss": 0.5348,
      "step": 414
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.24541020782476444,
      "learning_rate": 0.00019283996910112318,
      "loss": 0.5432,
      "step": 415
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.2255959168063083,
      "learning_rate": 0.00019280499802097126,
      "loss": 0.5891,
      "step": 416
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.21159018099714821,
      "learning_rate": 0.0001927699449324478,
      "loss": 0.5003,
      "step": 417
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.21379995902020923,
      "learning_rate": 0.00019273480986652794,
      "loss": 0.5314,
      "step": 418
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.2853169518220406,
      "learning_rate": 0.0001926995928542592,
      "loss": 0.6108,
      "step": 419
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.22738285867292138,
      "learning_rate": 0.00019266429392676164,
      "loss": 0.5217,
      "step": 420
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.23835369502554374,
      "learning_rate": 0.00019262891311522755,
      "loss": 0.5318,
      "step": 421
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.20671557324330114,
      "learning_rate": 0.0001925934504509217,
      "loss": 0.5234,
      "step": 422
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.205212164360302,
      "learning_rate": 0.00019255790596518112,
      "loss": 0.5023,
      "step": 423
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.21664090577036152,
      "learning_rate": 0.00019252227968941522,
      "loss": 0.5452,
      "step": 424
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.22146041084684798,
      "learning_rate": 0.00019248657165510556,
      "loss": 0.5474,
      "step": 425
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.2338997589574809,
      "learning_rate": 0.00019245078189380604,
      "loss": 0.5516,
      "step": 426
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.2313978280927526,
      "learning_rate": 0.0001924149104371428,
      "loss": 0.5831,
      "step": 427
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.2098577112814155,
      "learning_rate": 0.00019237895731681408,
      "loss": 0.5452,
      "step": 428
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.26497439164374026,
      "learning_rate": 0.0001923429225645904,
      "loss": 0.5666,
      "step": 429
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.21859970576834997,
      "learning_rate": 0.00019230680621231425,
      "loss": 0.5069,
      "step": 430
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.20509380886351694,
      "learning_rate": 0.0001922706082919004,
      "loss": 0.4573,
      "step": 431
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.2182328366507935,
      "learning_rate": 0.0001922343288353356,
      "loss": 0.6133,
      "step": 432
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.2822350271273954,
      "learning_rate": 0.00019219796787467867,
      "loss": 0.5709,
      "step": 433
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.24487543268473794,
      "learning_rate": 0.00019216152544206049,
      "loss": 0.546,
      "step": 434
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.24221176090281485,
      "learning_rate": 0.00019212500156968383,
      "loss": 0.5507,
      "step": 435
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.22053929296251015,
      "learning_rate": 0.00019208839628982358,
      "loss": 0.5473,
      "step": 436
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.22975415570737245,
      "learning_rate": 0.00019205170963482643,
      "loss": 0.5181,
      "step": 437
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.22969105575505203,
      "learning_rate": 0.00019201494163711104,
      "loss": 0.5463,
      "step": 438
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.23764087103158363,
      "learning_rate": 0.00019197809232916795,
      "loss": 0.55,
      "step": 439
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.21997498488474826,
      "learning_rate": 0.00019194116174355954,
      "loss": 0.5421,
      "step": 440
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.22225824990596896,
      "learning_rate": 0.00019190414991291998,
      "loss": 0.5439,
      "step": 441
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.243391488050543,
      "learning_rate": 0.00019186705686995533,
      "loss": 0.6289,
      "step": 442
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.222494273038652,
      "learning_rate": 0.0001918298826474433,
      "loss": 0.5088,
      "step": 443
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.22114450997419682,
      "learning_rate": 0.0001917926272782334,
      "loss": 0.5624,
      "step": 444
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.21964760504534894,
      "learning_rate": 0.00019175529079524687,
      "loss": 0.5289,
      "step": 445
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.3042847973140014,
      "learning_rate": 0.00019171787323147654,
      "loss": 0.5328,
      "step": 446
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.22425571202210934,
      "learning_rate": 0.00019168037461998695,
      "loss": 0.5699,
      "step": 447
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.23406959191320909,
      "learning_rate": 0.00019164279499391427,
      "loss": 0.5147,
      "step": 448
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.3604500123158513,
      "learning_rate": 0.00019160513438646617,
      "loss": 0.5697,
      "step": 449
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.2501436029131694,
      "learning_rate": 0.00019156739283092205,
      "loss": 0.6015,
      "step": 450
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.21928141490521824,
      "learning_rate": 0.00019152957036063265,
      "loss": 0.5111,
      "step": 451
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.257908225365161,
      "learning_rate": 0.00019149166700902032,
      "loss": 0.5132,
      "step": 452
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.2713678867101362,
      "learning_rate": 0.0001914536828095789,
      "loss": 0.5995,
      "step": 453
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.2398794022246256,
      "learning_rate": 0.0001914156177958736,
      "loss": 0.4993,
      "step": 454
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.2373981477389832,
      "learning_rate": 0.0001913774720015411,
      "loss": 0.5064,
      "step": 455
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.2188011093608266,
      "learning_rate": 0.00019133924546028942,
      "loss": 0.5606,
      "step": 456
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.24077263566935142,
      "learning_rate": 0.00019130093820589791,
      "loss": 0.5606,
      "step": 457
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.23519919814487683,
      "learning_rate": 0.00019126255027221735,
      "loss": 0.5307,
      "step": 458
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.21480730775028578,
      "learning_rate": 0.00019122408169316976,
      "loss": 0.526,
      "step": 459
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.2161668548042441,
      "learning_rate": 0.00019118553250274832,
      "loss": 0.5657,
      "step": 460
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.22318400428439122,
      "learning_rate": 0.00019114690273501765,
      "loss": 0.513,
      "step": 461
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.22252447744680176,
      "learning_rate": 0.00019110819242411337,
      "loss": 0.5247,
      "step": 462
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.21358818358042153,
      "learning_rate": 0.00019106940160424244,
      "loss": 0.556,
      "step": 463
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.2121229259271081,
      "learning_rate": 0.0001910305303096828,
      "loss": 0.5138,
      "step": 464
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.22636146624511622,
      "learning_rate": 0.0001909915785747836,
      "loss": 0.5111,
      "step": 465
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.20571954917028099,
      "learning_rate": 0.00019095254643396512,
      "loss": 0.5125,
      "step": 466
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.21968966730793454,
      "learning_rate": 0.0001909134339217186,
      "loss": 0.5358,
      "step": 467
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.21910723327372644,
      "learning_rate": 0.00019087424107260627,
      "loss": 0.5382,
      "step": 468
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.2153841373499183,
      "learning_rate": 0.00019083496792126153,
      "loss": 0.5375,
      "step": 469
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.23479205084160673,
      "learning_rate": 0.00019079561450238854,
      "loss": 0.5984,
      "step": 470
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.21595571362737268,
      "learning_rate": 0.00019075618085076247,
      "loss": 0.5417,
      "step": 471
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.24550770571804625,
      "learning_rate": 0.00019071666700122946,
      "loss": 0.5306,
      "step": 472
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.21802243564456578,
      "learning_rate": 0.00019067707298870638,
      "loss": 0.5157,
      "step": 473
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.2068796190094572,
      "learning_rate": 0.00019063739884818103,
      "loss": 0.5254,
      "step": 474
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.24034732867281272,
      "learning_rate": 0.000190597644614712,
      "loss": 0.6204,
      "step": 475
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.2260836607650634,
      "learning_rate": 0.00019055781032342864,
      "loss": 0.5492,
      "step": 476
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.2476351525598878,
      "learning_rate": 0.00019051789600953102,
      "loss": 0.5157,
      "step": 477
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.2280151093681579,
      "learning_rate": 0.00019047790170829003,
      "loss": 0.4984,
      "step": 478
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.2217333524292061,
      "learning_rate": 0.00019043782745504711,
      "loss": 0.5149,
      "step": 479
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.2356369467654302,
      "learning_rate": 0.00019039767328521442,
      "loss": 0.5724,
      "step": 480
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.21541809863677616,
      "learning_rate": 0.0001903574392342747,
      "loss": 0.5138,
      "step": 481
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.21722431891543054,
      "learning_rate": 0.00019031712533778137,
      "loss": 0.5536,
      "step": 482
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.2370708268417489,
      "learning_rate": 0.00019027673163135827,
      "loss": 0.5038,
      "step": 483
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.22809310323516838,
      "learning_rate": 0.00019023625815069989,
      "loss": 0.5713,
      "step": 484
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.22374988575329294,
      "learning_rate": 0.00019019570493157114,
      "loss": 0.5549,
      "step": 485
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.20510711707245072,
      "learning_rate": 0.0001901550720098074,
      "loss": 0.46,
      "step": 486
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.2621551195786783,
      "learning_rate": 0.00019011435942131448,
      "loss": 0.5546,
      "step": 487
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.20503054358781417,
      "learning_rate": 0.00019007356720206865,
      "loss": 0.5547,
      "step": 488
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.23586140447856616,
      "learning_rate": 0.00019003269538811647,
      "loss": 0.6075,
      "step": 489
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.2828040872125889,
      "learning_rate": 0.00018999174401557488,
      "loss": 0.602,
      "step": 490
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.2023429982220119,
      "learning_rate": 0.00018995071312063105,
      "loss": 0.4975,
      "step": 491
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.2054777673202953,
      "learning_rate": 0.00018990960273954254,
      "loss": 0.5295,
      "step": 492
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.1982185225446849,
      "learning_rate": 0.00018986841290863704,
      "loss": 0.5461,
      "step": 493
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.23248022218099268,
      "learning_rate": 0.0001898271436643125,
      "loss": 0.5924,
      "step": 494
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.2235279893303581,
      "learning_rate": 0.00018978579504303706,
      "loss": 0.5598,
      "step": 495
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.21675084465821123,
      "learning_rate": 0.000189744367081349,
      "loss": 0.5012,
      "step": 496
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.2041881848681654,
      "learning_rate": 0.00018970285981585662,
      "loss": 0.526,
      "step": 497
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.23258761727278376,
      "learning_rate": 0.00018966127328323842,
      "loss": 0.553,
      "step": 498
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.23066266735191,
      "learning_rate": 0.00018961960752024288,
      "loss": 0.5506,
      "step": 499
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.20634958879584178,
      "learning_rate": 0.0001895778625636885,
      "loss": 0.5006,
      "step": 500
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.21082421656186934,
      "learning_rate": 0.00018953603845046378,
      "loss": 0.5279,
      "step": 501
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.2057560041730304,
      "learning_rate": 0.00018949413521752713,
      "loss": 0.5598,
      "step": 502
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.2096114347066206,
      "learning_rate": 0.00018945215290190693,
      "loss": 0.5113,
      "step": 503
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.23218477255443984,
      "learning_rate": 0.00018941009154070136,
      "loss": 0.5169,
      "step": 504
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.20857717653678057,
      "learning_rate": 0.00018936795117107855,
      "loss": 0.5149,
      "step": 505
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.24006448825761761,
      "learning_rate": 0.0001893257318302764,
      "loss": 0.5228,
      "step": 506
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.2146671098435255,
      "learning_rate": 0.00018928343355560258,
      "loss": 0.5257,
      "step": 507
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.20608859556559073,
      "learning_rate": 0.00018924105638443452,
      "loss": 0.527,
      "step": 508
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.2336814919363686,
      "learning_rate": 0.0001891986003542194,
      "loss": 0.5461,
      "step": 509
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.2409130946928026,
      "learning_rate": 0.00018915606550247397,
      "loss": 0.5493,
      "step": 510
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.21371348825911873,
      "learning_rate": 0.0001891134518667848,
      "loss": 0.572,
      "step": 511
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.2014364828041311,
      "learning_rate": 0.000189070759484808,
      "loss": 0.5109,
      "step": 512
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.2290945612613713,
      "learning_rate": 0.0001890279883942692,
      "loss": 0.5493,
      "step": 513
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.22127732127756986,
      "learning_rate": 0.0001889851386329637,
      "loss": 0.5387,
      "step": 514
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.20564079598559082,
      "learning_rate": 0.00018894221023875622,
      "loss": 0.5192,
      "step": 515
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.213993086214796,
      "learning_rate": 0.00018889920324958106,
      "loss": 0.5044,
      "step": 516
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.21506249939577854,
      "learning_rate": 0.00018885611770344185,
      "loss": 0.4969,
      "step": 517
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.22792164808811663,
      "learning_rate": 0.00018881295363841174,
      "loss": 0.5564,
      "step": 518
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.1978731923118128,
      "learning_rate": 0.00018876971109263324,
      "loss": 0.4898,
      "step": 519
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.22394451521984352,
      "learning_rate": 0.00018872639010431822,
      "loss": 0.5586,
      "step": 520
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.20009625678598073,
      "learning_rate": 0.0001886829907117478,
      "loss": 0.5399,
      "step": 521
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.20448355507434923,
      "learning_rate": 0.00018863951295327244,
      "loss": 0.5263,
      "step": 522
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.1967777231547204,
      "learning_rate": 0.00018859595686731187,
      "loss": 0.4904,
      "step": 523
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.2052388343929957,
      "learning_rate": 0.00018855232249235498,
      "loss": 0.4951,
      "step": 524
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.1970956590240829,
      "learning_rate": 0.00018850860986695985,
      "loss": 0.5112,
      "step": 525
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.2102143499682878,
      "learning_rate": 0.00018846481902975377,
      "loss": 0.5234,
      "step": 526
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.23384214794287286,
      "learning_rate": 0.00018842095001943306,
      "loss": 0.5387,
      "step": 527
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.20133953340775343,
      "learning_rate": 0.00018837700287476316,
      "loss": 0.4995,
      "step": 528
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.2238467486071384,
      "learning_rate": 0.00018833297763457858,
      "loss": 0.5709,
      "step": 529
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.26170161234282546,
      "learning_rate": 0.00018828887433778278,
      "loss": 0.6314,
      "step": 530
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.2317819906199683,
      "learning_rate": 0.00018824469302334822,
      "loss": 0.5333,
      "step": 531
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.21538390925414544,
      "learning_rate": 0.0001882004337303163,
      "loss": 0.5603,
      "step": 532
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.23053571801246284,
      "learning_rate": 0.0001881560964977974,
      "loss": 0.593,
      "step": 533
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.21173642276584706,
      "learning_rate": 0.0001881116813649706,
      "loss": 0.5539,
      "step": 534
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.24587290888576793,
      "learning_rate": 0.00018806718837108402,
      "loss": 0.5408,
      "step": 535
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.22324082101473863,
      "learning_rate": 0.00018802261755545443,
      "loss": 0.5857,
      "step": 536
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.21827653101692504,
      "learning_rate": 0.0001879779689574674,
      "loss": 0.5451,
      "step": 537
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.2146222856243753,
      "learning_rate": 0.00018793324261657737,
      "loss": 0.5007,
      "step": 538
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.20994383183759666,
      "learning_rate": 0.00018788843857230726,
      "loss": 0.5039,
      "step": 539
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.23384168426304514,
      "learning_rate": 0.00018784355686424876,
      "loss": 0.5329,
      "step": 540
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.20284382518697272,
      "learning_rate": 0.00018779859753206225,
      "loss": 0.5383,
      "step": 541
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.22307014132513725,
      "learning_rate": 0.00018775356061547662,
      "loss": 0.5766,
      "step": 542
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.21675879523474215,
      "learning_rate": 0.00018770844615428932,
      "loss": 0.4994,
      "step": 543
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.2200785983728407,
      "learning_rate": 0.00018766325418836637,
      "loss": 0.5615,
      "step": 544
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.20895654400479502,
      "learning_rate": 0.00018761798475764224,
      "loss": 0.4993,
      "step": 545
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.22152937631276676,
      "learning_rate": 0.00018757263790211988,
      "loss": 0.5275,
      "step": 546
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.209333487906431,
      "learning_rate": 0.0001875272136618706,
      "loss": 0.4911,
      "step": 547
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.2123519912763275,
      "learning_rate": 0.00018748171207703417,
      "loss": 0.5662,
      "step": 548
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.2147346642469028,
      "learning_rate": 0.00018743613318781868,
      "loss": 0.5651,
      "step": 549
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.2017789732342509,
      "learning_rate": 0.00018739047703450048,
      "loss": 0.5573,
      "step": 550
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.2084087089737107,
      "learning_rate": 0.00018734474365742428,
      "loss": 0.562,
      "step": 551
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.22130968599178,
      "learning_rate": 0.00018729893309700295,
      "loss": 0.5729,
      "step": 552
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.22736172090948445,
      "learning_rate": 0.0001872530453937176,
      "loss": 0.5548,
      "step": 553
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.21738577850339916,
      "learning_rate": 0.0001872070805881176,
      "loss": 0.5191,
      "step": 554
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.20994273135857797,
      "learning_rate": 0.00018716103872082026,
      "loss": 0.5153,
      "step": 555
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.25944295362906805,
      "learning_rate": 0.00018711491983251113,
      "loss": 0.5471,
      "step": 556
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.2138519097360962,
      "learning_rate": 0.00018706872396394376,
      "loss": 0.4875,
      "step": 557
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.23586915663527888,
      "learning_rate": 0.00018702245115593974,
      "loss": 0.5224,
      "step": 558
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.20477148046499385,
      "learning_rate": 0.0001869761014493887,
      "loss": 0.5466,
      "step": 559
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.21783175505387284,
      "learning_rate": 0.00018692967488524812,
      "loss": 0.5557,
      "step": 560
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.20442177589984145,
      "learning_rate": 0.0001868831715045435,
      "loss": 0.507,
      "step": 561
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.21291324212369495,
      "learning_rate": 0.00018683659134836813,
      "loss": 0.5779,
      "step": 562
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.22670486875141618,
      "learning_rate": 0.00018678993445788323,
      "loss": 0.5831,
      "step": 563
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.2431493116309222,
      "learning_rate": 0.00018674320087431768,
      "loss": 0.5389,
      "step": 564
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.22102091260855142,
      "learning_rate": 0.00018669639063896836,
      "loss": 0.5569,
      "step": 565
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.20001951850669827,
      "learning_rate": 0.0001866495037931997,
      "loss": 0.486,
      "step": 566
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.22781103196427857,
      "learning_rate": 0.00018660254037844388,
      "loss": 0.4973,
      "step": 567
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.21129685691062433,
      "learning_rate": 0.00018655550043620073,
      "loss": 0.5459,
      "step": 568
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.20363805081315986,
      "learning_rate": 0.0001865083840080378,
      "loss": 0.4997,
      "step": 569
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.22269838654252982,
      "learning_rate": 0.00018646119113559006,
      "loss": 0.5406,
      "step": 570
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.20307002281681275,
      "learning_rate": 0.00018641392186056016,
      "loss": 0.4861,
      "step": 571
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.20146261628709675,
      "learning_rate": 0.0001863665762247182,
      "loss": 0.561,
      "step": 572
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.21049257054009352,
      "learning_rate": 0.00018631915426990184,
      "loss": 0.5257,
      "step": 573
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.2245482792823418,
      "learning_rate": 0.00018627165603801605,
      "loss": 0.5441,
      "step": 574
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.2106578436256788,
      "learning_rate": 0.0001862240815710333,
      "loss": 0.5125,
      "step": 575
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.2091435884054145,
      "learning_rate": 0.0001861764309109934,
      "loss": 0.523,
      "step": 576
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.21256854318600532,
      "learning_rate": 0.00018612870410000354,
      "loss": 0.4851,
      "step": 577
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.24387962798982954,
      "learning_rate": 0.00018608090118023808,
      "loss": 0.5423,
      "step": 578
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.2357478920855788,
      "learning_rate": 0.00018603302219393874,
      "loss": 0.5386,
      "step": 579
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.21267780857117077,
      "learning_rate": 0.0001859850671834144,
      "loss": 0.5545,
      "step": 580
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.25049614581715324,
      "learning_rate": 0.0001859370361910412,
      "loss": 0.5241,
      "step": 581
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.1937807494598699,
      "learning_rate": 0.00018588892925926228,
      "loss": 0.5533,
      "step": 582
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.21209972240968475,
      "learning_rate": 0.00018584074643058807,
      "loss": 0.538,
      "step": 583
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.22281277082523665,
      "learning_rate": 0.00018579248774759586,
      "loss": 0.5456,
      "step": 584
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.22156542955128883,
      "learning_rate": 0.00018574415325293018,
      "loss": 0.5622,
      "step": 585
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.20068342929250654,
      "learning_rate": 0.00018569574298930237,
      "loss": 0.5372,
      "step": 586
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.21693418845369525,
      "learning_rate": 0.00018564725699949083,
      "loss": 0.4874,
      "step": 587
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.2060622909003744,
      "learning_rate": 0.0001855986953263409,
      "loss": 0.5331,
      "step": 588
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.20007419545283933,
      "learning_rate": 0.00018555005801276463,
      "loss": 0.5131,
      "step": 589
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.21905328017125653,
      "learning_rate": 0.00018550134510174115,
      "loss": 0.5572,
      "step": 590
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.21213287568506015,
      "learning_rate": 0.0001854525566363162,
      "loss": 0.5359,
      "step": 591
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.20066093050756748,
      "learning_rate": 0.00018540369265960242,
      "loss": 0.5334,
      "step": 592
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.2068811720002483,
      "learning_rate": 0.00018535475321477906,
      "loss": 0.5558,
      "step": 593
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.2025287668887073,
      "learning_rate": 0.00018530573834509215,
      "loss": 0.5098,
      "step": 594
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.20807380346718593,
      "learning_rate": 0.0001852566480938543,
      "loss": 0.5211,
      "step": 595
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.2049943719782544,
      "learning_rate": 0.00018520748250444474,
      "loss": 0.5379,
      "step": 596
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.8558508208219735,
      "learning_rate": 0.00018515824162030934,
      "loss": 0.5403,
      "step": 597
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.25414317775682305,
      "learning_rate": 0.00018510892548496047,
      "loss": 0.5804,
      "step": 598
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.20806597400748386,
      "learning_rate": 0.00018505953414197696,
      "loss": 0.5419,
      "step": 599
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.1950528976937739,
      "learning_rate": 0.00018501006763500414,
      "loss": 0.4956,
      "step": 600
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.20652545713558523,
      "learning_rate": 0.00018496052600775376,
      "loss": 0.4942,
      "step": 601
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.20955886781649663,
      "learning_rate": 0.0001849109093040039,
      "loss": 0.5177,
      "step": 602
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.21093362015684414,
      "learning_rate": 0.00018486121756759906,
      "loss": 0.5672,
      "step": 603
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.22033088091533184,
      "learning_rate": 0.00018481145084245002,
      "loss": 0.5691,
      "step": 604
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.20322111965044637,
      "learning_rate": 0.00018476160917253373,
      "loss": 0.5425,
      "step": 605
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.2028788101278272,
      "learning_rate": 0.0001847116926018935,
      "loss": 0.5176,
      "step": 606
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.19551140156538951,
      "learning_rate": 0.0001846617011746388,
      "loss": 0.5115,
      "step": 607
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.21944694996534547,
      "learning_rate": 0.00018461163493494517,
      "loss": 0.5496,
      "step": 608
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.21506814147924705,
      "learning_rate": 0.0001845614939270543,
      "loss": 0.5823,
      "step": 609
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.2220938137588105,
      "learning_rate": 0.00018451127819527402,
      "loss": 0.5731,
      "step": 610
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.21590208362786933,
      "learning_rate": 0.00018446098778397807,
      "loss": 0.6063,
      "step": 611
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.20084594317065918,
      "learning_rate": 0.00018441062273760628,
      "loss": 0.5286,
      "step": 612
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.21847304705653886,
      "learning_rate": 0.00018436018310066435,
      "loss": 0.5721,
      "step": 613
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.2467936487351411,
      "learning_rate": 0.000184309668917724,
      "loss": 0.571,
      "step": 614
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.21666156526926003,
      "learning_rate": 0.0001842590802334227,
      "loss": 0.5244,
      "step": 615
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.21336859433357677,
      "learning_rate": 0.00018420841709246383,
      "loss": 0.5724,
      "step": 616
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.1933070755110986,
      "learning_rate": 0.0001841576795396166,
      "loss": 0.5347,
      "step": 617
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.2332186369470874,
      "learning_rate": 0.00018410686761971586,
      "loss": 0.5474,
      "step": 618
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.1996293438855639,
      "learning_rate": 0.00018405598137766224,
      "loss": 0.5421,
      "step": 619
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.2012759807756364,
      "learning_rate": 0.00018400502085842208,
      "loss": 0.519,
      "step": 620
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.24355300568180752,
      "learning_rate": 0.00018395398610702733,
      "loss": 0.597,
      "step": 621
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.2136711983483761,
      "learning_rate": 0.00018390287716857546,
      "loss": 0.5398,
      "step": 622
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.22275088525970024,
      "learning_rate": 0.00018385169408822964,
      "loss": 0.5597,
      "step": 623
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.20011931485707388,
      "learning_rate": 0.0001838004369112184,
      "loss": 0.4901,
      "step": 624
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.19544716159187206,
      "learning_rate": 0.00018374910568283594,
      "loss": 0.4726,
      "step": 625
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.2176067620544374,
      "learning_rate": 0.00018369770044844168,
      "loss": 0.5369,
      "step": 626
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.2005629047810257,
      "learning_rate": 0.00018364622125346055,
      "loss": 0.4914,
      "step": 627
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.21497281608823432,
      "learning_rate": 0.0001835946681433829,
      "loss": 0.5559,
      "step": 628
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.20354723273049724,
      "learning_rate": 0.00018354304116376425,
      "loss": 0.5083,
      "step": 629
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.23536026550959782,
      "learning_rate": 0.0001834913403602255,
      "loss": 0.5449,
      "step": 630
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.20887211237530257,
      "learning_rate": 0.00018343956577845276,
      "loss": 0.5131,
      "step": 631
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.21728763678777088,
      "learning_rate": 0.00018338771746419726,
      "loss": 0.5484,
      "step": 632
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.21910570476522437,
      "learning_rate": 0.00018333579546327556,
      "loss": 0.5452,
      "step": 633
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.21247350127543838,
      "learning_rate": 0.00018328379982156915,
      "loss": 0.5232,
      "step": 634
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.21706686115456897,
      "learning_rate": 0.00018323173058502472,
      "loss": 0.5353,
      "step": 635
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.19529494853666482,
      "learning_rate": 0.00018317958779965387,
      "loss": 0.4611,
      "step": 636
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.2194890381897013,
      "learning_rate": 0.00018312737151153334,
      "loss": 0.4884,
      "step": 637
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.24336065627870296,
      "learning_rate": 0.00018307508176680472,
      "loss": 0.5708,
      "step": 638
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.22638828434923797,
      "learning_rate": 0.00018302271861167456,
      "loss": 0.5795,
      "step": 639
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.20501380703607638,
      "learning_rate": 0.0001829702820924142,
      "loss": 0.5645,
      "step": 640
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.22705979847255006,
      "learning_rate": 0.00018291777225535994,
      "loss": 0.4974,
      "step": 641
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.22629645320684777,
      "learning_rate": 0.00018286518914691272,
      "loss": 0.5587,
      "step": 642
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.21772563640763765,
      "learning_rate": 0.00018281253281353838,
      "loss": 0.5219,
      "step": 643
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.20447194133414195,
      "learning_rate": 0.00018275980330176737,
      "loss": 0.5425,
      "step": 644
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.24126870503035064,
      "learning_rate": 0.00018270700065819477,
      "loss": 0.5119,
      "step": 645
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.23269297218381896,
      "learning_rate": 0.00018265412492948042,
      "loss": 0.5507,
      "step": 646
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.23416570398912578,
      "learning_rate": 0.0001826011761623486,
      "loss": 0.5947,
      "step": 647
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.2186560086983282,
      "learning_rate": 0.0001825481544035882,
      "loss": 0.5204,
      "step": 648
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.20624707271501935,
      "learning_rate": 0.00018249505970005262,
      "loss": 0.4785,
      "step": 649
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.23418189558532218,
      "learning_rate": 0.00018244189209865974,
      "loss": 0.4976,
      "step": 650
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.21372290734059424,
      "learning_rate": 0.00018238865164639173,
      "loss": 0.5237,
      "step": 651
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.1986689651795865,
      "learning_rate": 0.0001823353383902953,
      "loss": 0.5354,
      "step": 652
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.21154599437074698,
      "learning_rate": 0.0001822819523774814,
      "loss": 0.5292,
      "step": 653
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.21348268586605149,
      "learning_rate": 0.00018222849365512523,
      "loss": 0.5249,
      "step": 654
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.22296243039072478,
      "learning_rate": 0.0001821749622704664,
      "loss": 0.5458,
      "step": 655
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.22596567506529938,
      "learning_rate": 0.00018212135827080857,
      "loss": 0.5085,
      "step": 656
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.19012132806019622,
      "learning_rate": 0.00018206768170351962,
      "loss": 0.4977,
      "step": 657
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.2125366600531234,
      "learning_rate": 0.0001820139326160316,
      "loss": 0.5051,
      "step": 658
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.23677835317412968,
      "learning_rate": 0.00018196011105584058,
      "loss": 0.575,
      "step": 659
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.2262210065848097,
      "learning_rate": 0.00018190621707050671,
      "loss": 0.5744,
      "step": 660
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.21618545867420894,
      "learning_rate": 0.0001818522507076541,
      "loss": 0.5715,
      "step": 661
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.2050215711297079,
      "learning_rate": 0.00018179821201497092,
      "loss": 0.5201,
      "step": 662
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.20218467055707082,
      "learning_rate": 0.0001817441010402091,
      "loss": 0.5058,
      "step": 663
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.20940987275867923,
      "learning_rate": 0.00018168991783118452,
      "loss": 0.5095,
      "step": 664
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.21341822518403558,
      "learning_rate": 0.00018163566243577697,
      "loss": 0.5599,
      "step": 665
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.20028205017927186,
      "learning_rate": 0.0001815813349019299,
      "loss": 0.5318,
      "step": 666
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.20184912350066175,
      "learning_rate": 0.00018152693527765057,
      "loss": 0.5643,
      "step": 667
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.20882160405967118,
      "learning_rate": 0.0001814724636110099,
      "loss": 0.542,
      "step": 668
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.20252144356881077,
      "learning_rate": 0.00018141791995014255,
      "loss": 0.4496,
      "step": 669
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.1956328371434174,
      "learning_rate": 0.00018136330434324674,
      "loss": 0.56,
      "step": 670
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.20691128111503362,
      "learning_rate": 0.00018130861683858426,
      "loss": 0.5726,
      "step": 671
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.2258004454621585,
      "learning_rate": 0.00018125385748448048,
      "loss": 0.583,
      "step": 672
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.22330059205477634,
      "learning_rate": 0.00018119902632932416,
      "loss": 0.5288,
      "step": 673
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.20473079466150892,
      "learning_rate": 0.0001811441234215677,
      "loss": 0.5085,
      "step": 674
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.19439333859223318,
      "learning_rate": 0.0001810891488097267,
      "loss": 0.5147,
      "step": 675
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.2037181989313857,
      "learning_rate": 0.00018103410254238021,
      "loss": 0.5228,
      "step": 676
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.21580635559566858,
      "learning_rate": 0.0001809789846681706,
      "loss": 0.5034,
      "step": 677
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.21490060304667385,
      "learning_rate": 0.00018092379523580357,
      "loss": 0.5347,
      "step": 678
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.20927738857723482,
      "learning_rate": 0.00018086853429404793,
      "loss": 0.5039,
      "step": 679
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.21391199422702836,
      "learning_rate": 0.00018081320189173577,
      "loss": 0.5404,
      "step": 680
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.22355130583819918,
      "learning_rate": 0.0001807577980777623,
      "loss": 0.5147,
      "step": 681
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.21899190720848985,
      "learning_rate": 0.00018070232290108584,
      "loss": 0.5195,
      "step": 682
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.20636096645560792,
      "learning_rate": 0.00018064677641072775,
      "loss": 0.5158,
      "step": 683
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.20462410706105155,
      "learning_rate": 0.00018059115865577249,
      "loss": 0.5194,
      "step": 684
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.21978634315593423,
      "learning_rate": 0.00018053546968536735,
      "loss": 0.4986,
      "step": 685
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.2203882917140438,
      "learning_rate": 0.00018047970954872264,
      "loss": 0.5855,
      "step": 686
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.20144829000454462,
      "learning_rate": 0.0001804238782951116,
      "loss": 0.5212,
      "step": 687
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.21142991796239274,
      "learning_rate": 0.00018036797597387023,
      "loss": 0.495,
      "step": 688
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.21275432668758548,
      "learning_rate": 0.00018031200263439736,
      "loss": 0.5694,
      "step": 689
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.2035189446424034,
      "learning_rate": 0.00018025595832615459,
      "loss": 0.55,
      "step": 690
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.20030837247360464,
      "learning_rate": 0.00018019984309866619,
      "loss": 0.4748,
      "step": 691
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.20366715425572,
      "learning_rate": 0.00018014365700151912,
      "loss": 0.5792,
      "step": 692
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.2082468197583491,
      "learning_rate": 0.000180087400084363,
      "loss": 0.4973,
      "step": 693
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.21820027454676755,
      "learning_rate": 0.00018003107239691004,
      "loss": 0.5512,
      "step": 694
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.2085678250499903,
      "learning_rate": 0.00017997467398893488,
      "loss": 0.5148,
      "step": 695
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.20422653056781329,
      "learning_rate": 0.00017991820491027472,
      "loss": 0.4968,
      "step": 696
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.1875899162050169,
      "learning_rate": 0.0001798616652108293,
      "loss": 0.5061,
      "step": 697
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.20869663705218836,
      "learning_rate": 0.00017980505494056062,
      "loss": 0.5182,
      "step": 698
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.19250179476147736,
      "learning_rate": 0.00017974837414949307,
      "loss": 0.5184,
      "step": 699
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.21732108838463451,
      "learning_rate": 0.00017969162288771347,
      "loss": 0.5524,
      "step": 700
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.20200315361578813,
      "learning_rate": 0.0001796348012053707,
      "loss": 0.5386,
      "step": 701
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.20242537832049035,
      "learning_rate": 0.00017957790915267615,
      "loss": 0.5656,
      "step": 702
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.1889172192023988,
      "learning_rate": 0.0001795209467799031,
      "loss": 0.5115,
      "step": 703
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.19623435201373893,
      "learning_rate": 0.0001794639141373872,
      "loss": 0.497,
      "step": 704
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.22372809637554478,
      "learning_rate": 0.00017940681127552604,
      "loss": 0.5579,
      "step": 705
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.1968536923376666,
      "learning_rate": 0.0001793496382447794,
      "loss": 0.4891,
      "step": 706
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.1990723573146364,
      "learning_rate": 0.00017929239509566894,
      "loss": 0.5921,
      "step": 707
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.20388703819339077,
      "learning_rate": 0.00017923508187877834,
      "loss": 0.5414,
      "step": 708
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.23657852979478725,
      "learning_rate": 0.00017917769864475314,
      "loss": 0.5672,
      "step": 709
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.22888252332289927,
      "learning_rate": 0.00017912024544430088,
      "loss": 0.5459,
      "step": 710
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.19383907969249117,
      "learning_rate": 0.0001790627223281908,
      "loss": 0.5509,
      "step": 711
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.2154263629956836,
      "learning_rate": 0.00017900512934725397,
      "loss": 0.5629,
      "step": 712
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.19802419635693494,
      "learning_rate": 0.0001789474665523832,
      "loss": 0.5128,
      "step": 713
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.19783321602266912,
      "learning_rate": 0.00017888973399453296,
      "loss": 0.5064,
      "step": 714
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.19864882371614528,
      "learning_rate": 0.00017883193172471944,
      "loss": 0.5458,
      "step": 715
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.23609512585527,
      "learning_rate": 0.00017877405979402038,
      "loss": 0.5069,
      "step": 716
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.19894144678524353,
      "learning_rate": 0.00017871611825357502,
      "loss": 0.5812,
      "step": 717
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.19598819977852033,
      "learning_rate": 0.00017865810715458427,
      "loss": 0.5223,
      "step": 718
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.23274847505011953,
      "learning_rate": 0.00017860002654831032,
      "loss": 0.5703,
      "step": 719
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.19794477486450376,
      "learning_rate": 0.00017854187648607694,
      "loss": 0.5538,
      "step": 720
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.2091737019131215,
      "learning_rate": 0.00017848365701926913,
      "loss": 0.4962,
      "step": 721
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.21890749511490995,
      "learning_rate": 0.00017842536819933337,
      "loss": 0.5074,
      "step": 722
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.22746821737803938,
      "learning_rate": 0.0001783670100777773,
      "loss": 0.5849,
      "step": 723
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.20967916540656184,
      "learning_rate": 0.0001783085827061699,
      "loss": 0.5246,
      "step": 724
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.19798059353181535,
      "learning_rate": 0.00017825008613614127,
      "loss": 0.4667,
      "step": 725
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.1992664047298655,
      "learning_rate": 0.00017819152041938265,
      "loss": 0.5247,
      "step": 726
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.22025628624147217,
      "learning_rate": 0.00017813288560764647,
      "loss": 0.5291,
      "step": 727
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.20405038516624363,
      "learning_rate": 0.00017807418175274612,
      "loss": 0.5235,
      "step": 728
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.20626127985692586,
      "learning_rate": 0.00017801540890655609,
      "loss": 0.5103,
      "step": 729
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.2187527308725265,
      "learning_rate": 0.00017795656712101172,
      "loss": 0.5515,
      "step": 730
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.20386714530070776,
      "learning_rate": 0.00017789765644810935,
      "loss": 0.5109,
      "step": 731
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.1990293686392052,
      "learning_rate": 0.00017783867693990624,
      "loss": 0.5208,
      "step": 732
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.19601721442767256,
      "learning_rate": 0.0001777796286485204,
      "loss": 0.5318,
      "step": 733
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.20542580410660244,
      "learning_rate": 0.0001777205116261306,
      "loss": 0.5198,
      "step": 734
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.20998518101289002,
      "learning_rate": 0.0001776613259249764,
      "loss": 0.5384,
      "step": 735
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.20134476803418952,
      "learning_rate": 0.00017760207159735805,
      "loss": 0.5448,
      "step": 736
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.22396912180134018,
      "learning_rate": 0.00017754274869563637,
      "loss": 0.59,
      "step": 737
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.2044555533666512,
      "learning_rate": 0.00017748335727223294,
      "loss": 0.5152,
      "step": 738
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.2106748606361736,
      "learning_rate": 0.00017742389737962966,
      "loss": 0.5233,
      "step": 739
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.19348093577281505,
      "learning_rate": 0.0001773643690703691,
      "loss": 0.5181,
      "step": 740
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.20393569458294794,
      "learning_rate": 0.00017730477239705428,
      "loss": 0.5671,
      "step": 741
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.19728761757057783,
      "learning_rate": 0.00017724510741234858,
      "loss": 0.4919,
      "step": 742
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.2025575313201386,
      "learning_rate": 0.0001771853741689757,
      "loss": 0.5452,
      "step": 743
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.19153867099886435,
      "learning_rate": 0.0001771255727197198,
      "loss": 0.4951,
      "step": 744
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.2220125331576081,
      "learning_rate": 0.00017706570311742516,
      "loss": 0.5521,
      "step": 745
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.22704666961693065,
      "learning_rate": 0.0001770057654149964,
      "loss": 0.5184,
      "step": 746
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.20871880228168335,
      "learning_rate": 0.00017694575966539823,
      "loss": 0.5205,
      "step": 747
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.2105924919088961,
      "learning_rate": 0.00017688568592165552,
      "loss": 0.5448,
      "step": 748
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.19780662201378688,
      "learning_rate": 0.00017682554423685329,
      "loss": 0.6037,
      "step": 749
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.23105576261963792,
      "learning_rate": 0.0001767653346641365,
      "loss": 0.7225,
      "step": 750
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.21997563912032173,
      "learning_rate": 0.00017670505725671013,
      "loss": 0.552,
      "step": 751
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.2033859052649398,
      "learning_rate": 0.00017664471206783915,
      "loss": 0.5315,
      "step": 752
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.19979214467824102,
      "learning_rate": 0.00017658429915084835,
      "loss": 0.5697,
      "step": 753
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.20567412732571028,
      "learning_rate": 0.00017652381855912247,
      "loss": 0.5051,
      "step": 754
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.20563597140752976,
      "learning_rate": 0.0001764632703461059,
      "loss": 0.5141,
      "step": 755
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.1979658869623221,
      "learning_rate": 0.00017640265456530293,
      "loss": 0.5257,
      "step": 756
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.2241077787834463,
      "learning_rate": 0.0001763419712702775,
      "loss": 0.5203,
      "step": 757
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.2197932846972142,
      "learning_rate": 0.00017628122051465322,
      "loss": 0.5847,
      "step": 758
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.1990944255813207,
      "learning_rate": 0.00017622040235211326,
      "loss": 0.4962,
      "step": 759
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.22179309744704687,
      "learning_rate": 0.00017615951683640045,
      "loss": 0.5635,
      "step": 760
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.20505896786344424,
      "learning_rate": 0.00017609856402131703,
      "loss": 0.4968,
      "step": 761
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.21771157401053975,
      "learning_rate": 0.00017603754396072483,
      "loss": 0.4858,
      "step": 762
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.23357401076131715,
      "learning_rate": 0.000175976456708545,
      "loss": 0.5766,
      "step": 763
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.21488993425737504,
      "learning_rate": 0.0001759153023187581,
      "loss": 0.5419,
      "step": 764
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.2035555999534868,
      "learning_rate": 0.00017585408084540405,
      "loss": 0.5272,
      "step": 765
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.20066829451010718,
      "learning_rate": 0.00017579279234258198,
      "loss": 0.5013,
      "step": 766
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.2052255359730049,
      "learning_rate": 0.00017573143686445034,
      "loss": 0.5383,
      "step": 767
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.19180058672325329,
      "learning_rate": 0.00017567001446522665,
      "loss": 0.5108,
      "step": 768
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.22862029943228582,
      "learning_rate": 0.0001756085251991877,
      "loss": 0.531,
      "step": 769
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.2180888066741993,
      "learning_rate": 0.00017554696912066924,
      "loss": 0.5938,
      "step": 770
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.19823263656993223,
      "learning_rate": 0.00017548534628406616,
      "loss": 0.5158,
      "step": 771
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.18700255356016454,
      "learning_rate": 0.00017542365674383227,
      "loss": 0.517,
      "step": 772
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.22948411236460914,
      "learning_rate": 0.00017536190055448037,
      "loss": 0.5464,
      "step": 773
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.21370070213829387,
      "learning_rate": 0.00017530007777058213,
      "loss": 0.5158,
      "step": 774
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.19174674116457566,
      "learning_rate": 0.0001752381884467681,
      "loss": 0.5035,
      "step": 775
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.19069115110218368,
      "learning_rate": 0.00017517623263772758,
      "loss": 0.5341,
      "step": 776
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.2401612943495333,
      "learning_rate": 0.00017511421039820863,
      "loss": 0.578,
      "step": 777
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.20371659209716964,
      "learning_rate": 0.00017505212178301805,
      "loss": 0.5103,
      "step": 778
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.2029847143168681,
      "learning_rate": 0.00017498996684702132,
      "loss": 0.537,
      "step": 779
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.1904915669257304,
      "learning_rate": 0.00017492774564514235,
      "loss": 0.5129,
      "step": 780
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.20640027525482552,
      "learning_rate": 0.00017486545823236385,
      "loss": 0.5585,
      "step": 781
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.23582084058854208,
      "learning_rate": 0.00017480310466372686,
      "loss": 0.5648,
      "step": 782
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.2219618762642625,
      "learning_rate": 0.00017474068499433098,
      "loss": 0.5365,
      "step": 783
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.2021980496149104,
      "learning_rate": 0.00017467819927933416,
      "loss": 0.5232,
      "step": 784
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.22350007413119136,
      "learning_rate": 0.00017461564757395272,
      "loss": 0.571,
      "step": 785
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.1982267515659923,
      "learning_rate": 0.00017455302993346134,
      "loss": 0.5228,
      "step": 786
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.1981641437638338,
      "learning_rate": 0.00017449034641319288,
      "loss": 0.5233,
      "step": 787
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.21019781303279997,
      "learning_rate": 0.00017442759706853855,
      "loss": 0.5207,
      "step": 788
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.2060312831458839,
      "learning_rate": 0.00017436478195494756,
      "loss": 0.5262,
      "step": 789
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.21829001169718656,
      "learning_rate": 0.00017430190112792737,
      "loss": 0.563,
      "step": 790
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.18511782073951058,
      "learning_rate": 0.00017423895464304342,
      "loss": 0.5017,
      "step": 791
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.1883852134929889,
      "learning_rate": 0.00017417594255591927,
      "loss": 0.4598,
      "step": 792
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.18093236424530848,
      "learning_rate": 0.00017411286492223632,
      "loss": 0.4834,
      "step": 793
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.18428120434597678,
      "learning_rate": 0.000174049721797734,
      "loss": 0.5032,
      "step": 794
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.20829275110131446,
      "learning_rate": 0.00017398651323820958,
      "loss": 0.5844,
      "step": 795
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.20484798677763622,
      "learning_rate": 0.00017392323929951812,
      "loss": 0.5674,
      "step": 796
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.24390628538267795,
      "learning_rate": 0.0001738599000375725,
      "loss": 0.5415,
      "step": 797
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.7159631217821198,
      "learning_rate": 0.00017379649550834327,
      "loss": 0.5248,
      "step": 798
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.2153929799459398,
      "learning_rate": 0.00017373302576785874,
      "loss": 0.5362,
      "step": 799
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.18434502268826083,
      "learning_rate": 0.00017366949087220472,
      "loss": 0.5179,
      "step": 800
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.1993458339623336,
      "learning_rate": 0.0001736058908775247,
      "loss": 0.5378,
      "step": 801
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.20482117487035287,
      "learning_rate": 0.0001735422258400197,
      "loss": 0.5066,
      "step": 802
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.22481951556352617,
      "learning_rate": 0.0001734784958159481,
      "loss": 0.5504,
      "step": 803
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.20893222575857784,
      "learning_rate": 0.00017341470086162586,
      "loss": 0.5558,
      "step": 804
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.21011978723049574,
      "learning_rate": 0.0001733508410334262,
      "loss": 0.5164,
      "step": 805
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.19493427746334713,
      "learning_rate": 0.0001732869163877797,
      "loss": 0.4928,
      "step": 806
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.21183476672026114,
      "learning_rate": 0.00017322292698117425,
      "loss": 0.539,
      "step": 807
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.19833380077404217,
      "learning_rate": 0.00017315887287015492,
      "loss": 0.5271,
      "step": 808
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.1914374518219283,
      "learning_rate": 0.000173094754111324,
      "loss": 0.5408,
      "step": 809
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.2086814492670768,
      "learning_rate": 0.00017303057076134085,
      "loss": 0.5289,
      "step": 810
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.20957903788826676,
      "learning_rate": 0.000172966322876922,
      "loss": 0.4998,
      "step": 811
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.20998255172298386,
      "learning_rate": 0.00017290201051484085,
      "loss": 0.5481,
      "step": 812
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.2071975609134585,
      "learning_rate": 0.00017283763373192798,
      "loss": 0.5183,
      "step": 813
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.21738328519054306,
      "learning_rate": 0.00017277319258507073,
      "loss": 0.539,
      "step": 814
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.20518040499899,
      "learning_rate": 0.0001727086871312134,
      "loss": 0.5109,
      "step": 815
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.19341379491491822,
      "learning_rate": 0.00017264411742735707,
      "loss": 0.4882,
      "step": 816
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.23128359760674316,
      "learning_rate": 0.00017257948353055963,
      "loss": 0.547,
      "step": 817
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.1960131633047162,
      "learning_rate": 0.0001725147854979357,
      "loss": 0.5467,
      "step": 818
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.21053602560644855,
      "learning_rate": 0.00017245002338665656,
      "loss": 0.5644,
      "step": 819
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.19602457133539752,
      "learning_rate": 0.00017238519725395007,
      "loss": 0.5121,
      "step": 820
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.1923459024283483,
      "learning_rate": 0.00017232030715710076,
      "loss": 0.5335,
      "step": 821
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.19919783133579333,
      "learning_rate": 0.00017225535315344955,
      "loss": 0.5076,
      "step": 822
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.23727428892467575,
      "learning_rate": 0.00017219033530039397,
      "loss": 0.5396,
      "step": 823
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.18048505778792392,
      "learning_rate": 0.00017212525365538792,
      "loss": 0.467,
      "step": 824
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.20071702267002645,
      "learning_rate": 0.00017206010827594163,
      "loss": 0.5217,
      "step": 825
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.20315216612025339,
      "learning_rate": 0.0001719948992196217,
      "loss": 0.4975,
      "step": 826
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.2142259292765235,
      "learning_rate": 0.00017192962654405096,
      "loss": 0.5148,
      "step": 827
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.19450012283752555,
      "learning_rate": 0.00017186429030690848,
      "loss": 0.5297,
      "step": 828
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.19853162923467543,
      "learning_rate": 0.00017179889056592954,
      "loss": 0.547,
      "step": 829
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.24873174470750906,
      "learning_rate": 0.00017173342737890544,
      "loss": 0.563,
      "step": 830
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.18593730182623175,
      "learning_rate": 0.00017166790080368357,
      "loss": 0.4647,
      "step": 831
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.19387710879340586,
      "learning_rate": 0.00017160231089816748,
      "loss": 0.5313,
      "step": 832
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.20818447206363588,
      "learning_rate": 0.00017153665772031643,
      "loss": 0.5333,
      "step": 833
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.17584822143732362,
      "learning_rate": 0.0001714709413281458,
      "loss": 0.4467,
      "step": 834
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.19622166504995672,
      "learning_rate": 0.00017140516177972676,
      "loss": 0.5129,
      "step": 835
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.18822988249157332,
      "learning_rate": 0.00017133931913318625,
      "loss": 0.5186,
      "step": 836
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.2021164051164271,
      "learning_rate": 0.00017127341344670696,
      "loss": 0.551,
      "step": 837
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.19354327476685654,
      "learning_rate": 0.00017120744477852745,
      "loss": 0.5001,
      "step": 838
    }
  ],
  "logging_steps": 1,
  "max_steps": 3352,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 838,
  "total_flos": 829560862015488.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}