{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 700.0,
  "eval_steps": 500,
  "global_step": 1400,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.5,
      "grad_norm": 1.4533216953277588,
      "learning_rate": 0.0,
      "loss": 1.2846,
      "step": 1
    },
    {
      "epoch": 1.0,
      "grad_norm": 5.102683067321777,
      "learning_rate": 2e-05,
      "loss": 1.5633,
      "step": 2
    },
    {
      "epoch": 1.5,
      "grad_norm": 1.826399803161621,
      "learning_rate": 4e-05,
      "loss": 1.5518,
      "step": 3
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.120856761932373,
      "learning_rate": 6e-05,
      "loss": 1.2156,
      "step": 4
    },
    {
      "epoch": 2.5,
      "grad_norm": 4.832008361816406,
      "learning_rate": 8e-05,
      "loss": 1.6279,
      "step": 5
    },
    {
      "epoch": 3.0,
      "grad_norm": 4.409387588500977,
      "learning_rate": 0.0001,
      "loss": 1.5203,
      "step": 6
    },
    {
      "epoch": 3.5,
      "grad_norm": 6.504903316497803,
      "learning_rate": 9.999987320820351e-05,
      "loss": 1.7311,
      "step": 7
    },
    {
      "epoch": 4.0,
      "grad_norm": 8.893552780151367,
      "learning_rate": 9.999949283345713e-05,
      "loss": 1.5575,
      "step": 8
    },
    {
      "epoch": 4.5,
      "grad_norm": 3.4009008407592773,
      "learning_rate": 9.999885887768996e-05,
      "loss": 1.5584,
      "step": 9
    },
    {
      "epoch": 5.0,
      "grad_norm": 3.7018280029296875,
      "learning_rate": 9.999797134411723e-05,
      "loss": 1.7964,
      "step": 10
    },
    {
      "epoch": 5.5,
      "grad_norm": 1.724374771118164,
      "learning_rate": 9.99968302372402e-05,
      "loss": 1.6762,
      "step": 11
    },
    {
      "epoch": 6.0,
      "grad_norm": 16.11446189880371,
      "learning_rate": 9.999543556284623e-05,
      "loss": 3.4108,
      "step": 12
    },
    {
      "epoch": 6.5,
      "grad_norm": 1.2704060077667236,
      "learning_rate": 9.999378732800862e-05,
      "loss": 1.4583,
      "step": 13
    },
    {
      "epoch": 7.0,
      "grad_norm": 9.662576675415039,
      "learning_rate": 9.999188554108667e-05,
      "loss": 1.7856,
      "step": 14
    },
    {
      "epoch": 7.5,
      "grad_norm": 1.1054532527923584,
      "learning_rate": 9.998973021172563e-05,
      "loss": 1.5089,
      "step": 15
    },
    {
      "epoch": 8.0,
      "grad_norm": 1.939652919769287,
      "learning_rate": 9.998732135085665e-05,
      "loss": 1.2069,
      "step": 16
    },
    {
      "epoch": 8.5,
      "grad_norm": 1.4218322038650513,
      "learning_rate": 9.998465897069665e-05,
      "loss": 1.328,
      "step": 17
    },
    {
      "epoch": 9.0,
      "grad_norm": 2.588613271713257,
      "learning_rate": 9.998174308474836e-05,
      "loss": 1.0565,
      "step": 18
    },
    {
      "epoch": 9.5,
      "grad_norm": 0.9522063136100769,
      "learning_rate": 9.997857370780018e-05,
      "loss": 1.3767,
      "step": 19
    },
    {
      "epoch": 10.0,
      "grad_norm": 2.3010125160217285,
      "learning_rate": 9.997515085592617e-05,
      "loss": 1.108,
      "step": 20
    },
    {
      "epoch": 10.5,
      "grad_norm": 0.6938635110855103,
      "learning_rate": 9.99714745464859e-05,
      "loss": 1.3034,
      "step": 21
    },
    {
      "epoch": 11.0,
      "grad_norm": 1.6980719566345215,
      "learning_rate": 9.996754479812442e-05,
      "loss": 1.1537,
      "step": 22
    },
    {
      "epoch": 11.5,
      "grad_norm": 2.117846965789795,
      "learning_rate": 9.996336163077211e-05,
      "loss": 1.3536,
      "step": 23
    },
    {
      "epoch": 12.0,
      "grad_norm": 1.7240182161331177,
      "learning_rate": 9.995892506564462e-05,
      "loss": 1.3882,
      "step": 24
    },
    {
      "epoch": 12.5,
      "grad_norm": 1.4354215860366821,
      "learning_rate": 9.995423512524276e-05,
      "loss": 1.4039,
      "step": 25
    },
    {
      "epoch": 13.0,
      "grad_norm": 1.7549787759780884,
      "learning_rate": 9.994929183335238e-05,
      "loss": 1.0789,
      "step": 26
    },
    {
      "epoch": 13.5,
      "grad_norm": 1.525095820426941,
      "learning_rate": 9.99440952150442e-05,
      "loss": 1.3472,
      "step": 27
    },
    {
      "epoch": 14.0,
      "grad_norm": 1.1550884246826172,
      "learning_rate": 9.993864529667381e-05,
      "loss": 1.1391,
      "step": 28
    },
    {
      "epoch": 14.5,
      "grad_norm": 1.4837149381637573,
      "learning_rate": 9.993294210588135e-05,
      "loss": 1.6544,
      "step": 29
    },
    {
      "epoch": 15.0,
      "grad_norm": 2.167940616607666,
      "learning_rate": 9.992698567159159e-05,
      "loss": 1.0244,
      "step": 30
    },
    {
      "epoch": 15.5,
      "grad_norm": 1.03326416015625,
      "learning_rate": 9.992077602401357e-05,
      "loss": 1.3189,
      "step": 31
    },
    {
      "epoch": 16.0,
      "grad_norm": 2.2560083866119385,
      "learning_rate": 9.991431319464062e-05,
      "loss": 1.2672,
      "step": 32
    },
    {
      "epoch": 16.5,
      "grad_norm": 1.3611760139465332,
      "learning_rate": 9.990759721625005e-05,
      "loss": 1.1938,
      "step": 33
    },
    {
      "epoch": 17.0,
      "grad_norm": 1.8796628713607788,
      "learning_rate": 9.990062812290313e-05,
      "loss": 1.3249,
      "step": 34
    },
    {
      "epoch": 17.5,
      "grad_norm": 0.8938995003700256,
      "learning_rate": 9.989340594994479e-05,
      "loss": 1.3939,
      "step": 35
    },
    {
      "epoch": 18.0,
      "grad_norm": 1.7087706327438354,
      "learning_rate": 9.988593073400354e-05,
      "loss": 1.0114,
      "step": 36
    },
    {
      "epoch": 18.5,
      "grad_norm": 0.8916568756103516,
      "learning_rate": 9.987820251299122e-05,
      "loss": 1.4161,
      "step": 37
    },
    {
      "epoch": 19.0,
      "grad_norm": 1.37736177444458,
      "learning_rate": 9.987022132610282e-05,
      "loss": 0.9011,
      "step": 38
    },
    {
      "epoch": 19.5,
      "grad_norm": 0.8252999186515808,
      "learning_rate": 9.986198721381631e-05,
      "loss": 1.2672,
      "step": 39
    },
    {
      "epoch": 20.0,
      "grad_norm": 6.583056926727295,
      "learning_rate": 9.985350021789241e-05,
      "loss": 2.1181,
      "step": 40
    },
    {
      "epoch": 20.5,
      "grad_norm": 0.7365291714668274,
      "learning_rate": 9.984476038137436e-05,
      "loss": 1.2156,
      "step": 41
    },
    {
      "epoch": 21.0,
      "grad_norm": 2.4956235885620117,
      "learning_rate": 9.983576774858776e-05,
      "loss": 1.2126,
      "step": 42
    },
    {
      "epoch": 21.5,
      "grad_norm": 1.0655571222305298,
      "learning_rate": 9.98265223651403e-05,
      "loss": 1.2381,
      "step": 43
    },
    {
      "epoch": 22.0,
      "grad_norm": 2.3908939361572266,
      "learning_rate": 9.98170242779215e-05,
      "loss": 1.2759,
      "step": 44
    },
    {
      "epoch": 22.5,
      "grad_norm": 0.8008183836936951,
      "learning_rate": 9.980727353510257e-05,
      "loss": 1.2073,
      "step": 45
    },
    {
      "epoch": 23.0,
      "grad_norm": 2.6294734477996826,
      "learning_rate": 9.979727018613607e-05,
      "loss": 0.989,
      "step": 46
    },
    {
      "epoch": 23.5,
      "grad_norm": 1.5704054832458496,
      "learning_rate": 9.97870142817557e-05,
      "loss": 1.3178,
      "step": 47
    },
    {
      "epoch": 24.0,
      "grad_norm": 1.6115163564682007,
      "learning_rate": 9.977650587397604e-05,
      "loss": 0.9742,
      "step": 48
    },
    {
      "epoch": 24.5,
      "grad_norm": 0.6747032999992371,
      "learning_rate": 9.976574501609231e-05,
      "loss": 1.1176,
      "step": 49
    },
    {
      "epoch": 25.0,
      "grad_norm": 1.6712925434112549,
      "learning_rate": 9.975473176268e-05,
      "loss": 0.9116,
      "step": 50
    },
    {
      "epoch": 25.5,
      "grad_norm": 0.9749622941017151,
      "learning_rate": 9.974346616959476e-05,
      "loss": 1.0685,
      "step": 51
    },
    {
      "epoch": 26.0,
      "grad_norm": 1.4353482723236084,
      "learning_rate": 9.973194829397196e-05,
      "loss": 1.1886,
      "step": 52
    },
    {
      "epoch": 26.5,
      "grad_norm": 1.0750980377197266,
      "learning_rate": 9.97201781942265e-05,
      "loss": 1.091,
      "step": 53
    },
    {
      "epoch": 27.0,
      "grad_norm": 2.135573387145996,
      "learning_rate": 9.970815593005247e-05,
      "loss": 1.3116,
      "step": 54
    },
    {
      "epoch": 27.5,
      "grad_norm": 0.96718829870224,
      "learning_rate": 9.969588156242282e-05,
      "loss": 1.231,
      "step": 55
    },
    {
      "epoch": 28.0,
      "grad_norm": 2.673003673553467,
      "learning_rate": 9.968335515358916e-05,
      "loss": 1.0411,
      "step": 56
    },
    {
      "epoch": 28.5,
      "grad_norm": 0.9151428937911987,
      "learning_rate": 9.967057676708125e-05,
      "loss": 1.1175,
      "step": 57
    },
    {
      "epoch": 29.0,
      "grad_norm": 1.75728178024292,
      "learning_rate": 9.965754646770695e-05,
      "loss": 0.9549,
      "step": 58
    },
    {
      "epoch": 29.5,
      "grad_norm": 0.6613996624946594,
      "learning_rate": 9.964426432155162e-05,
      "loss": 1.123,
      "step": 59
    },
    {
      "epoch": 30.0,
      "grad_norm": 7.0879950523376465,
      "learning_rate": 9.963073039597797e-05,
      "loss": 1.57,
      "step": 60
    },
    {
      "epoch": 30.5,
      "grad_norm": 1.5590451955795288,
      "learning_rate": 9.961694475962562e-05,
      "loss": 1.0639,
      "step": 61
    },
    {
      "epoch": 31.0,
      "grad_norm": 4.407354354858398,
      "learning_rate": 9.960290748241079e-05,
      "loss": 1.2691,
      "step": 62
    },
    {
      "epoch": 31.5,
      "grad_norm": 1.5881550312042236,
      "learning_rate": 9.958861863552595e-05,
      "loss": 1.0861,
      "step": 63
    },
    {
      "epoch": 32.0,
      "grad_norm": 1.2596269845962524,
      "learning_rate": 9.957407829143943e-05,
      "loss": 0.8634,
      "step": 64
    },
    {
      "epoch": 32.5,
      "grad_norm": 0.823441207408905,
      "learning_rate": 9.95592865238951e-05,
      "loss": 0.9869,
      "step": 65
    },
    {
      "epoch": 33.0,
      "grad_norm": 3.09020733833313,
      "learning_rate": 9.954424340791196e-05,
      "loss": 0.919,
      "step": 66
    },
    {
      "epoch": 33.5,
      "grad_norm": 1.3963035345077515,
      "learning_rate": 9.952894901978371e-05,
      "loss": 1.221,
      "step": 67
    },
    {
      "epoch": 34.0,
      "grad_norm": 2.0274956226348877,
      "learning_rate": 9.951340343707852e-05,
      "loss": 0.8623,
      "step": 68
    },
    {
      "epoch": 34.5,
      "grad_norm": 0.9748483300209045,
      "learning_rate": 9.949760673863846e-05,
      "loss": 1.1341,
      "step": 69
    },
    {
      "epoch": 35.0,
      "grad_norm": 2.0322656631469727,
      "learning_rate": 9.94815590045792e-05,
      "loss": 1.3446,
      "step": 70
    },
    {
      "epoch": 35.5,
      "grad_norm": 0.6573627591133118,
      "learning_rate": 9.94652603162896e-05,
      "loss": 1.0673,
      "step": 71
    },
    {
      "epoch": 36.0,
      "grad_norm": 3.763916015625,
      "learning_rate": 9.944871075643124e-05,
      "loss": 1.2009,
      "step": 72
    },
    {
      "epoch": 36.5,
      "grad_norm": 1.1149028539657593,
      "learning_rate": 9.943191040893806e-05,
      "loss": 1.15,
      "step": 73
    },
    {
      "epoch": 37.0,
      "grad_norm": 2.46221661567688,
      "learning_rate": 9.941485935901592e-05,
      "loss": 0.8574,
      "step": 74
    },
    {
      "epoch": 37.5,
      "grad_norm": 1.0253549814224243,
      "learning_rate": 9.939755769314214e-05,
      "loss": 1.254,
      "step": 75
    },
    {
      "epoch": 38.0,
      "grad_norm": 2.1021811962127686,
      "learning_rate": 9.938000549906508e-05,
      "loss": 1.293,
      "step": 76
    },
    {
      "epoch": 38.5,
      "grad_norm": 1.5739225149154663,
      "learning_rate": 9.936220286580375e-05,
      "loss": 1.2263,
      "step": 77
    },
    {
      "epoch": 39.0,
      "grad_norm": 1.1300886869430542,
      "learning_rate": 9.934414988364722e-05,
      "loss": 1.1633,
      "step": 78
    },
    {
      "epoch": 39.5,
      "grad_norm": 0.8601087927818298,
      "learning_rate": 9.93258466441543e-05,
      "loss": 0.9503,
      "step": 79
    },
    {
      "epoch": 40.0,
      "grad_norm": 1.3174241781234741,
      "learning_rate": 9.930729324015304e-05,
      "loss": 1.152,
      "step": 80
    },
    {
      "epoch": 40.5,
      "grad_norm": 1.2089372873306274,
      "learning_rate": 9.928848976574019e-05,
      "loss": 1.0775,
      "step": 81
    },
    {
      "epoch": 41.0,
      "grad_norm": 2.567349672317505,
      "learning_rate": 9.926943631628082e-05,
      "loss": 0.9668,
      "step": 82
    },
    {
      "epoch": 41.5,
      "grad_norm": 0.914560854434967,
      "learning_rate": 9.925013298840776e-05,
      "loss": 1.1329,
      "step": 83
    },
    {
      "epoch": 42.0,
      "grad_norm": 1.183079481124878,
      "learning_rate": 9.923057988002117e-05,
      "loss": 0.7418,
      "step": 84
    },
    {
      "epoch": 42.5,
      "grad_norm": 0.7308894991874695,
      "learning_rate": 9.921077709028799e-05,
      "loss": 0.9932,
      "step": 85
    },
    {
      "epoch": 43.0,
      "grad_norm": 1.032607078552246,
      "learning_rate": 9.919072471964145e-05,
      "loss": 1.1528,
      "step": 86
    },
    {
      "epoch": 43.5,
      "grad_norm": 1.4365724325180054,
      "learning_rate": 9.917042286978064e-05,
      "loss": 0.9763,
      "step": 87
    },
    {
      "epoch": 44.0,
      "grad_norm": 1.7179330587387085,
      "learning_rate": 9.914987164366985e-05,
      "loss": 1.0806,
      "step": 88
    },
    {
      "epoch": 44.5,
      "grad_norm": 0.9357978105545044,
      "learning_rate": 9.912907114553816e-05,
      "loss": 1.1997,
      "step": 89
    },
    {
      "epoch": 45.0,
      "grad_norm": 2.325937032699585,
      "learning_rate": 9.910802148087887e-05,
      "loss": 0.7478,
      "step": 90
    },
    {
      "epoch": 45.5,
      "grad_norm": 1.1565285921096802,
      "learning_rate": 9.908672275644898e-05,
      "loss": 1.0437,
      "step": 91
    },
    {
      "epoch": 46.0,
      "grad_norm": 0.9706777930259705,
      "learning_rate": 9.906517508026862e-05,
      "loss": 0.8967,
      "step": 92
    },
    {
      "epoch": 46.5,
      "grad_norm": 1.1153693199157715,
      "learning_rate": 9.904337856162053e-05,
      "loss": 1.0401,
      "step": 93
    },
    {
      "epoch": 47.0,
      "grad_norm": 2.0540757179260254,
      "learning_rate": 9.902133331104952e-05,
      "loss": 0.7264,
      "step": 94
    },
    {
      "epoch": 47.5,
      "grad_norm": 0.8775630593299866,
      "learning_rate": 9.899903944036186e-05,
      "loss": 0.9697,
      "step": 95
    },
    {
      "epoch": 48.0,
      "grad_norm": 2.645958423614502,
      "learning_rate": 9.897649706262473e-05,
      "loss": 1.2249,
      "step": 96
    },
    {
      "epoch": 48.5,
      "grad_norm": 2.0274133682250977,
      "learning_rate": 9.895370629216569e-05,
      "loss": 0.9145,
      "step": 97
    },
    {
      "epoch": 49.0,
      "grad_norm": 0.8980880975723267,
      "learning_rate": 9.893066724457206e-05,
      "loss": 1.1964,
      "step": 98
    },
    {
      "epoch": 49.5,
      "grad_norm": 2.7316150665283203,
      "learning_rate": 9.890738003669029e-05,
      "loss": 0.8965,
      "step": 99
    },
    {
      "epoch": 50.0,
      "grad_norm": 1.2547019720077515,
      "learning_rate": 9.88838447866255e-05,
      "loss": 1.089,
      "step": 100
    },
    {
      "epoch": 50.5,
      "grad_norm": 1.416431188583374,
      "learning_rate": 9.88600616137407e-05,
      "loss": 1.0853,
      "step": 101
    },
    {
      "epoch": 51.0,
      "grad_norm": 1.23435640335083,
      "learning_rate": 9.883603063865641e-05,
      "loss": 0.8314,
      "step": 102
    },
    {
      "epoch": 51.5,
      "grad_norm": 0.7177619934082031,
      "learning_rate": 9.881175198324981e-05,
      "loss": 1.1313,
      "step": 103
    },
    {
      "epoch": 52.0,
      "grad_norm": 2.9981842041015625,
      "learning_rate": 9.878722577065426e-05,
      "loss": 0.9448,
      "step": 104
    },
    {
      "epoch": 52.5,
      "grad_norm": 2.073280096054077,
      "learning_rate": 9.876245212525871e-05,
      "loss": 0.9723,
      "step": 105
    },
    {
      "epoch": 53.0,
      "grad_norm": 2.2693674564361572,
      "learning_rate": 9.873743117270691e-05,
      "loss": 0.8206,
      "step": 106
    },
    {
      "epoch": 53.5,
      "grad_norm": 1.9117062091827393,
      "learning_rate": 9.871216303989694e-05,
      "loss": 1.0267,
      "step": 107
    },
    {
      "epoch": 54.0,
      "grad_norm": 4.710913181304932,
      "learning_rate": 9.86866478549805e-05,
      "loss": 0.886,
      "step": 108
    },
    {
      "epoch": 54.5,
      "grad_norm": 0.8019539713859558,
      "learning_rate": 9.86608857473622e-05,
      "loss": 0.9315,
      "step": 109
    },
    {
      "epoch": 55.0,
      "grad_norm": 1.7147183418273926,
      "learning_rate": 9.863487684769901e-05,
      "loss": 0.7433,
      "step": 110
    },
    {
      "epoch": 55.5,
      "grad_norm": 0.7925494909286499,
      "learning_rate": 9.860862128789953e-05,
      "loss": 0.9382,
      "step": 111
    },
    {
      "epoch": 56.0,
      "grad_norm": 11.186344146728516,
      "learning_rate": 9.858211920112336e-05,
      "loss": 1.6628,
      "step": 112
    },
    {
      "epoch": 56.5,
      "grad_norm": 1.3355357646942139,
      "learning_rate": 9.855537072178038e-05,
      "loss": 0.9532,
      "step": 113
    },
    {
      "epoch": 57.0,
      "grad_norm": 8.531452178955078,
      "learning_rate": 9.852837598553008e-05,
      "loss": 0.9091,
      "step": 114
    },
    {
      "epoch": 57.5,
      "grad_norm": 13.450335502624512,
      "learning_rate": 9.850113512928094e-05,
      "loss": 1.1558,
      "step": 115
    },
    {
      "epoch": 58.0,
      "grad_norm": 1.836601972579956,
      "learning_rate": 9.847364829118963e-05,
      "loss": 1.0967,
      "step": 116
    },
    {
      "epoch": 58.5,
      "grad_norm": 0.6922495365142822,
      "learning_rate": 9.844591561066034e-05,
      "loss": 0.9977,
      "step": 117
    },
    {
      "epoch": 59.0,
      "grad_norm": 1.8775699138641357,
      "learning_rate": 9.841793722834419e-05,
      "loss": 0.7858,
      "step": 118
    },
    {
      "epoch": 59.5,
      "grad_norm": 1.8492512702941895,
      "learning_rate": 9.838971328613828e-05,
      "loss": 1.1309,
      "step": 119
    },
    {
      "epoch": 60.0,
      "grad_norm": 2.903394937515259,
      "learning_rate": 9.836124392718526e-05,
      "loss": 0.6598,
      "step": 120
    },
    {
      "epoch": 60.5,
      "grad_norm": 0.8631826639175415,
      "learning_rate": 9.833252929587232e-05,
      "loss": 0.9875,
      "step": 121
    },
    {
      "epoch": 61.0,
      "grad_norm": 3.5877304077148438,
      "learning_rate": 9.830356953783066e-05,
      "loss": 0.8396,
      "step": 122
    },
    {
      "epoch": 61.5,
      "grad_norm": 0.880155622959137,
      "learning_rate": 9.827436479993469e-05,
      "loss": 0.9716,
      "step": 123
    },
    {
      "epoch": 62.0,
      "grad_norm": 1.8486472368240356,
      "learning_rate": 9.824491523030123e-05,
      "loss": 0.9779,
      "step": 124
    },
    {
      "epoch": 62.5,
      "grad_norm": 1.9419149160385132,
      "learning_rate": 9.821522097828884e-05,
      "loss": 0.8883,
      "step": 125
    },
    {
      "epoch": 63.0,
      "grad_norm": 1.4343353509902954,
      "learning_rate": 9.818528219449705e-05,
      "loss": 0.6191,
      "step": 126
    },
    {
      "epoch": 63.5,
      "grad_norm": 1.670262336730957,
      "learning_rate": 9.815509903076552e-05,
      "loss": 0.9135,
      "step": 127
    },
    {
      "epoch": 64.0,
      "grad_norm": 3.3399698734283447,
      "learning_rate": 9.812467164017337e-05,
      "loss": 1.4676,
      "step": 128
    },
    {
      "epoch": 64.5,
      "grad_norm": 0.6421170830726624,
      "learning_rate": 9.809400017703829e-05,
      "loss": 0.9426,
      "step": 129
    },
    {
      "epoch": 65.0,
      "grad_norm": 1.1922038793563843,
      "learning_rate": 9.806308479691595e-05,
      "loss": 1.2651,
      "step": 130
    },
    {
      "epoch": 65.5,
      "grad_norm": 1.680457592010498,
      "learning_rate": 9.803192565659896e-05,
      "loss": 1.0284,
      "step": 131
    },
    {
      "epoch": 66.0,
      "grad_norm": 1.808544635772705,
      "learning_rate": 9.80005229141163e-05,
      "loss": 0.7461,
      "step": 132
    },
    {
      "epoch": 66.5,
      "grad_norm": 1.045966625213623,
      "learning_rate": 9.796887672873232e-05,
      "loss": 1.0892,
      "step": 133
    },
    {
      "epoch": 67.0,
      "grad_norm": 3.1462604999542236,
      "learning_rate": 9.793698726094614e-05,
      "loss": 0.7814,
      "step": 134
    },
    {
      "epoch": 67.5,
      "grad_norm": 0.6181429028511047,
      "learning_rate": 9.790485467249065e-05,
      "loss": 0.9626,
      "step": 135
    },
    {
      "epoch": 68.0,
      "grad_norm": 1.4937608242034912,
      "learning_rate": 9.787247912633179e-05,
      "loss": 1.1897,
      "step": 136
    },
    {
      "epoch": 68.5,
      "grad_norm": 0.7545461654663086,
      "learning_rate": 9.783986078666772e-05,
      "loss": 0.8969,
      "step": 137
    },
    {
      "epoch": 69.0,
      "grad_norm": 1.2557651996612549,
      "learning_rate": 9.780699981892793e-05,
      "loss": 0.9322,
      "step": 138
    },
    {
      "epoch": 69.5,
      "grad_norm": 0.7036311626434326,
      "learning_rate": 9.777389638977251e-05,
      "loss": 0.8971,
      "step": 139
    },
    {
      "epoch": 70.0,
      "grad_norm": 1.5644367933273315,
      "learning_rate": 9.774055066709116e-05,
      "loss": 0.9557,
      "step": 140
    },
    {
      "epoch": 70.5,
      "grad_norm": 0.337898313999176,
      "learning_rate": 9.770696282000244e-05,
      "loss": 0.8245,
      "step": 141
    },
    {
      "epoch": 71.0,
      "grad_norm": 4.7719316482543945,
      "learning_rate": 9.767313301885292e-05,
      "loss": 0.9355,
      "step": 142
    },
    {
      "epoch": 71.5,
      "grad_norm": 1.0715103149414062,
      "learning_rate": 9.76390614352162e-05,
      "loss": 0.8867,
      "step": 143
    },
    {
      "epoch": 72.0,
      "grad_norm": 1.5713542699813843,
      "learning_rate": 9.760474824189222e-05,
      "loss": 0.6927,
      "step": 144
    },
    {
      "epoch": 72.5,
      "grad_norm": 1.0581064224243164,
      "learning_rate": 9.757019361290621e-05,
      "loss": 1.2562,
      "step": 145
    },
    {
      "epoch": 73.0,
      "grad_norm": 4.257506847381592,
      "learning_rate": 9.753539772350792e-05,
      "loss": 1.0295,
      "step": 146
    },
    {
      "epoch": 73.5,
      "grad_norm": 1.3396309614181519,
      "learning_rate": 9.750036075017069e-05,
      "loss": 0.8717,
      "step": 147
    },
    {
      "epoch": 74.0,
      "grad_norm": 1.8104369640350342,
      "learning_rate": 9.74650828705905e-05,
      "loss": 0.6148,
      "step": 148
    },
    {
      "epoch": 74.5,
      "grad_norm": 1.446446180343628,
      "learning_rate": 9.742956426368526e-05,
      "loss": 0.7584,
      "step": 149
    },
    {
      "epoch": 75.0,
      "grad_norm": 1.6324974298477173,
      "learning_rate": 9.739380510959364e-05,
      "loss": 0.8674,
      "step": 150
    },
    {
      "epoch": 75.5,
      "grad_norm": 0.7358859181404114,
      "learning_rate": 9.735780558967434e-05,
      "loss": 0.8579,
      "step": 151
    },
    {
      "epoch": 76.0,
      "grad_norm": 2.550913095474243,
      "learning_rate": 9.732156588650512e-05,
      "loss": 0.7567,
      "step": 152
    },
    {
      "epoch": 76.5,
      "grad_norm": 0.5961642861366272,
      "learning_rate": 9.728508618388186e-05,
      "loss": 0.9019,
      "step": 153
    },
    {
      "epoch": 77.0,
      "grad_norm": 2.259815216064453,
      "learning_rate": 9.724836666681765e-05,
      "loss": 0.6635,
      "step": 154
    },
    {
      "epoch": 77.5,
      "grad_norm": 0.6815266609191895,
      "learning_rate": 9.721140752154181e-05,
      "loss": 0.8679,
      "step": 155
    },
    {
      "epoch": 78.0,
      "grad_norm": 1.3151683807373047,
      "learning_rate": 9.717420893549902e-05,
      "loss": 1.1615,
      "step": 156
    },
    {
      "epoch": 78.5,
      "grad_norm": 0.8054331541061401,
      "learning_rate": 9.713677109734829e-05,
      "loss": 0.8818,
      "step": 157
    },
    {
      "epoch": 79.0,
      "grad_norm": 0.8047096729278564,
      "learning_rate": 9.709909419696204e-05,
      "loss": 1.1889,
      "step": 158
    },
    {
      "epoch": 79.5,
      "grad_norm": 1.7221477031707764,
      "learning_rate": 9.706117842542518e-05,
      "loss": 1.0062,
      "step": 159
    },
    {
      "epoch": 80.0,
      "grad_norm": 1.886643886566162,
      "learning_rate": 9.702302397503402e-05,
      "loss": 0.663,
      "step": 160
    },
    {
      "epoch": 80.5,
      "grad_norm": 0.7801350951194763,
      "learning_rate": 9.698463103929542e-05,
      "loss": 0.8081,
      "step": 161
    },
    {
      "epoch": 81.0,
      "grad_norm": 3.242637872695923,
      "learning_rate": 9.694599981292578e-05,
      "loss": 0.5923,
      "step": 162
    },
    {
      "epoch": 81.5,
      "grad_norm": 1.0234795808792114,
      "learning_rate": 9.690713049184999e-05,
      "loss": 0.8623,
      "step": 163
    },
    {
      "epoch": 82.0,
      "grad_norm": 1.984404444694519,
      "learning_rate": 9.686802327320049e-05,
      "loss": 0.626,
      "step": 164
    },
    {
      "epoch": 82.5,
      "grad_norm": 0.980360746383667,
      "learning_rate": 9.682867835531624e-05,
      "loss": 0.7681,
      "step": 165
    },
    {
      "epoch": 83.0,
      "grad_norm": 1.5169624090194702,
      "learning_rate": 9.67890959377418e-05,
      "loss": 0.513,
      "step": 166
    },
    {
      "epoch": 83.5,
      "grad_norm": 1.1487135887145996,
      "learning_rate": 9.674927622122614e-05,
      "loss": 0.8092,
      "step": 167
    },
    {
      "epoch": 84.0,
      "grad_norm": 0.9957097768783569,
      "learning_rate": 9.670921940772186e-05,
      "loss": 0.5238,
      "step": 168
    },
    {
      "epoch": 84.5,
      "grad_norm": 1.4312140941619873,
      "learning_rate": 9.666892570038393e-05,
      "loss": 0.8492,
      "step": 169
    },
    {
      "epoch": 85.0,
      "grad_norm": 3.7888259887695312,
      "learning_rate": 9.662839530356883e-05,
      "loss": 1.2785,
      "step": 170
    },
    {
      "epoch": 85.5,
      "grad_norm": 0.7849655151367188,
      "learning_rate": 9.658762842283343e-05,
      "loss": 0.8865,
      "step": 171
    },
    {
      "epoch": 86.0,
      "grad_norm": 3.767606019973755,
      "learning_rate": 9.654662526493397e-05,
      "loss": 0.7688,
      "step": 172
    },
    {
      "epoch": 86.5,
      "grad_norm": 1.7961673736572266,
      "learning_rate": 9.650538603782502e-05,
      "loss": 0.9244,
      "step": 173
    },
    {
      "epoch": 87.0,
      "grad_norm": 3.2236788272857666,
      "learning_rate": 9.646391095065839e-05,
      "loss": 0.8467,
      "step": 174
    },
    {
      "epoch": 87.5,
      "grad_norm": 1.0763840675354004,
      "learning_rate": 9.642220021378212e-05,
      "loss": 0.8121,
      "step": 175
    },
    {
      "epoch": 88.0,
      "grad_norm": 1.062946081161499,
      "learning_rate": 9.63802540387394e-05,
      "loss": 0.8216,
      "step": 176
    },
    {
      "epoch": 88.5,
      "grad_norm": 0.7908176779747009,
      "learning_rate": 9.633807263826744e-05,
      "loss": 0.9088,
      "step": 177
    },
    {
      "epoch": 89.0,
      "grad_norm": 1.363552212715149,
      "learning_rate": 9.629565622629646e-05,
      "loss": 0.7706,
      "step": 178
    },
    {
      "epoch": 89.5,
      "grad_norm": 1.743476152420044,
      "learning_rate": 9.625300501794861e-05,
      "loss": 0.8235,
      "step": 179
    },
    {
      "epoch": 90.0,
      "grad_norm": 1.644329309463501,
      "learning_rate": 9.62101192295368e-05,
      "loss": 1.098,
      "step": 180
    },
    {
      "epoch": 90.5,
      "grad_norm": 1.2850360870361328,
      "learning_rate": 9.616699907856368e-05,
      "loss": 0.861,
      "step": 181
    },
    {
      "epoch": 91.0,
      "grad_norm": 1.6298117637634277,
      "learning_rate": 9.612364478372052e-05,
      "loss": 0.7435,
      "step": 182
    },
    {
      "epoch": 91.5,
      "grad_norm": 0.7512349486351013,
      "learning_rate": 9.608005656488605e-05,
      "loss": 0.8419,
      "step": 183
    },
    {
      "epoch": 92.0,
      "grad_norm": 1.4328938722610474,
      "learning_rate": 9.603623464312545e-05,
      "loss": 1.0971,
      "step": 184
    },
    {
      "epoch": 92.5,
      "grad_norm": 0.9426387548446655,
      "learning_rate": 9.599217924068909e-05,
      "loss": 1.0244,
      "step": 185
    },
    {
      "epoch": 93.0,
      "grad_norm": 1.344358205795288,
      "learning_rate": 9.594789058101153e-05,
      "loss": 0.8422,
      "step": 186
    },
    {
      "epoch": 93.5,
      "grad_norm": 1.0947461128234863,
      "learning_rate": 9.590336888871034e-05,
      "loss": 0.7944,
      "step": 187
    },
    {
      "epoch": 94.0,
      "grad_norm": 1.1683955192565918,
      "learning_rate": 9.58586143895849e-05,
      "loss": 0.5585,
      "step": 188
    },
    {
      "epoch": 94.5,
      "grad_norm": 1.0564055442810059,
      "learning_rate": 9.581362731061536e-05,
      "loss": 0.8456,
      "step": 189
    },
    {
      "epoch": 95.0,
      "grad_norm": 1.0157687664031982,
      "learning_rate": 9.576840787996142e-05,
      "loss": 0.645,
      "step": 190
    },
    {
      "epoch": 95.5,
      "grad_norm": 1.7204697132110596,
      "learning_rate": 9.57229563269612e-05,
      "loss": 0.8123,
      "step": 191
    },
    {
      "epoch": 96.0,
      "grad_norm": 1.3989800214767456,
      "learning_rate": 9.567727288213005e-05,
      "loss": 0.7705,
      "step": 192
    },
    {
      "epoch": 96.5,
      "grad_norm": 2.053833246231079,
      "learning_rate": 9.563135777715941e-05,
      "loss": 0.8302,
      "step": 193
    },
    {
      "epoch": 97.0,
      "grad_norm": 3.6400094032287598,
      "learning_rate": 9.558521124491566e-05,
      "loss": 1.1909,
      "step": 194
    },
    {
      "epoch": 97.5,
      "grad_norm": 2.8067195415496826,
      "learning_rate": 9.553883351943882e-05,
      "loss": 1.0373,
      "step": 195
    },
    {
      "epoch": 98.0,
      "grad_norm": 3.4638583660125732,
      "learning_rate": 9.549222483594153e-05,
      "loss": 0.7888,
      "step": 196
    },
    {
      "epoch": 98.5,
      "grad_norm": 2.022918939590454,
      "learning_rate": 9.544538543080773e-05,
      "loss": 1.1307,
      "step": 197
    },
    {
      "epoch": 99.0,
      "grad_norm": 2.333218574523926,
      "learning_rate": 9.539831554159151e-05,
      "loss": 0.8717,
      "step": 198
    },
    {
      "epoch": 99.5,
      "grad_norm": 1.954386591911316,
      "learning_rate": 9.53510154070159e-05,
      "loss": 0.8833,
      "step": 199
    },
    {
      "epoch": 100.0,
      "grad_norm": 1.3981575965881348,
      "learning_rate": 9.530348526697166e-05,
      "loss": 0.5786,
      "step": 200
    },
    {
      "epoch": 100.5,
      "grad_norm": 0.9758285284042358,
      "learning_rate": 9.525572536251607e-05,
      "loss": 0.8837,
      "step": 201
    },
    {
      "epoch": 101.0,
      "grad_norm": 1.1994894742965698,
      "learning_rate": 9.520773593587169e-05,
      "loss": 1.2143,
      "step": 202
    },
    {
      "epoch": 101.5,
      "grad_norm": 1.2751672267913818,
      "learning_rate": 9.515951723042513e-05,
      "loss": 0.8713,
      "step": 203
    },
    {
      "epoch": 102.0,
      "grad_norm": 10.766294479370117,
      "learning_rate": 9.511106949072588e-05,
      "loss": 1.7089,
      "step": 204
    },
    {
      "epoch": 102.5,
      "grad_norm": 1.484325647354126,
      "learning_rate": 9.506239296248496e-05,
      "loss": 0.8603,
      "step": 205
    },
    {
      "epoch": 103.0,
      "grad_norm": 2.784583568572998,
      "learning_rate": 9.501348789257373e-05,
      "loss": 0.8553,
      "step": 206
    },
    {
      "epoch": 103.5,
      "grad_norm": 0.8749284744262695,
      "learning_rate": 9.496435452902267e-05,
      "loss": 0.711,
      "step": 207
    },
    {
      "epoch": 104.0,
      "grad_norm": 2.875103235244751,
      "learning_rate": 9.491499312102008e-05,
      "loss": 0.9894,
      "step": 208
    },
    {
      "epoch": 104.5,
      "grad_norm": 0.9830334186553955,
      "learning_rate": 9.486540391891082e-05,
      "loss": 0.8073,
      "step": 209
    },
    {
      "epoch": 105.0,
      "grad_norm": 13.979928016662598,
      "learning_rate": 9.481558717419505e-05,
      "loss": 2.4878,
      "step": 210
    },
    {
      "epoch": 105.5,
      "grad_norm": 1.0810017585754395,
      "learning_rate": 9.476554313952697e-05,
      "loss": 0.6534,
      "step": 211
    },
    {
      "epoch": 106.0,
      "grad_norm": 1.807493805885315,
      "learning_rate": 9.471527206871348e-05,
      "loss": 0.621,
      "step": 212
    },
    {
      "epoch": 106.5,
      "grad_norm": 0.798584520816803,
      "learning_rate": 9.466477421671296e-05,
      "loss": 0.7576,
      "step": 213
    },
    {
      "epoch": 107.0,
      "grad_norm": 2.124471664428711,
      "learning_rate": 9.461404983963394e-05,
      "loss": 0.5049,
      "step": 214
    },
    {
      "epoch": 107.5,
      "grad_norm": 0.7614571452140808,
      "learning_rate": 9.456309919473384e-05,
      "loss": 0.7271,
      "step": 215
    },
    {
      "epoch": 108.0,
      "grad_norm": 1.4740674495697021,
      "learning_rate": 9.451192254041758e-05,
      "loss": 0.93,
      "step": 216
    },
    {
      "epoch": 108.5,
      "grad_norm": 0.6347062587738037,
      "learning_rate": 9.446052013623638e-05,
      "loss": 0.829,
      "step": 217
    },
    {
      "epoch": 109.0,
      "grad_norm": 4.417940139770508,
      "learning_rate": 9.440889224288635e-05,
      "loss": 0.5676,
      "step": 218
    },
    {
      "epoch": 109.5,
      "grad_norm": 0.9492636919021606,
      "learning_rate": 9.435703912220726e-05,
      "loss": 0.7662,
      "step": 219
    },
    {
      "epoch": 110.0,
      "grad_norm": 2.707926034927368,
      "learning_rate": 9.430496103718107e-05,
      "loss": 0.5767,
      "step": 220
    },
    {
      "epoch": 110.5,
      "grad_norm": 0.4781251549720764,
      "learning_rate": 9.425265825193076e-05,
      "loss": 0.7297,
      "step": 221
    },
    {
      "epoch": 111.0,
      "grad_norm": 1.0262559652328491,
      "learning_rate": 9.420013103171892e-05,
      "loss": 0.5205,
      "step": 222
    },
    {
      "epoch": 111.5,
      "grad_norm": 0.7944707870483398,
      "learning_rate": 9.414737964294636e-05,
      "loss": 0.8293,
      "step": 223
    },
    {
      "epoch": 112.0,
      "grad_norm": 1.5959267616271973,
      "learning_rate": 9.409440435315077e-05,
      "loss": 0.6237,
      "step": 224
    },
    {
      "epoch": 112.5,
      "grad_norm": 2.4740846157073975,
      "learning_rate": 9.404120543100552e-05,
      "loss": 0.8674,
      "step": 225
    },
    {
      "epoch": 113.0,
      "grad_norm": 1.3128033876419067,
      "learning_rate": 9.398778314631801e-05,
      "loss": 0.6412,
      "step": 226
    },
    {
      "epoch": 113.5,
      "grad_norm": 0.5115012526512146,
      "learning_rate": 9.393413777002858e-05,
      "loss": 0.8451,
      "step": 227
    },
    {
      "epoch": 114.0,
      "grad_norm": 2.7606518268585205,
      "learning_rate": 9.388026957420895e-05,
      "loss": 0.8272,
      "step": 228
    },
    {
      "epoch": 114.5,
      "grad_norm": 2.0180795192718506,
      "learning_rate": 9.382617883206095e-05,
      "loss": 0.8713,
      "step": 229
    },
    {
      "epoch": 115.0,
      "grad_norm": 3.012765884399414,
      "learning_rate": 9.377186581791508e-05,
      "loss": 0.646,
      "step": 230
    },
    {
      "epoch": 115.5,
      "grad_norm": 1.2706587314605713,
      "learning_rate": 9.371733080722911e-05,
      "loss": 0.8485,
      "step": 231
    },
    {
      "epoch": 116.0,
      "grad_norm": 2.9590582847595215,
      "learning_rate": 9.366257407658672e-05,
      "loss": 0.4839,
      "step": 232
    },
    {
      "epoch": 116.5,
      "grad_norm": 1.1685129404067993,
      "learning_rate": 9.36075959036961e-05,
      "loss": 0.798,
      "step": 233
    },
    {
      "epoch": 117.0,
      "grad_norm": 1.0731711387634277,
      "learning_rate": 9.355239656738848e-05,
      "loss": 0.4578,
      "step": 234
    },
    {
      "epoch": 117.5,
      "grad_norm": 0.9612593650817871,
      "learning_rate": 9.349697634761678e-05,
      "loss": 0.7873,
      "step": 235
    },
    {
      "epoch": 118.0,
      "grad_norm": 1.0949903726577759,
      "learning_rate": 9.34413355254542e-05,
      "loss": 0.4131,
      "step": 236
    },
    {
      "epoch": 118.5,
      "grad_norm": 1.7643836736679077,
      "learning_rate": 9.338547438309269e-05,
      "loss": 0.7952,
      "step": 237
    },
    {
      "epoch": 119.0,
      "grad_norm": 3.906822681427002,
      "learning_rate": 9.332939320384167e-05,
      "loss": 0.544,
      "step": 238
    },
    {
      "epoch": 119.5,
      "grad_norm": 1.5644744634628296,
      "learning_rate": 9.327309227212644e-05,
      "loss": 0.9862,
      "step": 239
    },
    {
      "epoch": 120.0,
      "grad_norm": 3.7236409187316895,
      "learning_rate": 9.321657187348689e-05,
      "loss": 0.8974,
      "step": 240
    },
    {
      "epoch": 120.5,
      "grad_norm": 0.5303816795349121,
      "learning_rate": 9.31598322945759e-05,
      "loss": 0.7485,
      "step": 241
    },
    {
      "epoch": 121.0,
      "grad_norm": 11.548822402954102,
      "learning_rate": 9.310287382315803e-05,
      "loss": 1.4675,
      "step": 242
    },
    {
      "epoch": 121.5,
      "grad_norm": 0.6730968356132507,
      "learning_rate": 9.304569674810794e-05,
      "loss": 0.6861,
      "step": 243
    },
    {
      "epoch": 122.0,
      "grad_norm": 1.9435992240905762,
      "learning_rate": 9.298830135940898e-05,
      "loss": 0.5761,
      "step": 244
    },
    {
      "epoch": 122.5,
      "grad_norm": 0.556621789932251,
      "learning_rate": 9.293068794815175e-05,
      "loss": 0.7334,
      "step": 245
    },
    {
      "epoch": 123.0,
      "grad_norm": 1.3812315464019775,
      "learning_rate": 9.287285680653254e-05,
      "loss": 0.6895,
      "step": 246
    },
    {
      "epoch": 123.5,
      "grad_norm": 0.7735423445701599,
      "learning_rate": 9.281480822785196e-05,
      "loss": 0.6797,
      "step": 247
    },
    {
      "epoch": 124.0,
      "grad_norm": 1.985374927520752,
      "learning_rate": 9.275654250651334e-05,
      "loss": 0.5939,
      "step": 248
    },
    {
      "epoch": 124.5,
      "grad_norm": 0.5469866394996643,
      "learning_rate": 9.269805993802129e-05,
      "loss": 0.79,
      "step": 249
    },
    {
      "epoch": 125.0,
      "grad_norm": 3.161343574523926,
      "learning_rate": 9.26393608189802e-05,
      "loss": 0.4819,
      "step": 250
    },
    {
      "epoch": 125.5,
      "grad_norm": 1.0311789512634277,
      "learning_rate": 9.258044544709274e-05,
      "loss": 0.6621,
      "step": 251
    },
    {
      "epoch": 126.0,
      "grad_norm": 2.7400898933410645,
      "learning_rate": 9.252131412115837e-05,
      "loss": 0.8655,
      "step": 252
    },
    {
      "epoch": 126.5,
      "grad_norm": 0.6111881732940674,
      "learning_rate": 9.246196714107176e-05,
      "loss": 0.6827,
      "step": 253
    },
    {
      "epoch": 127.0,
      "grad_norm": 0.8897466659545898,
      "learning_rate": 9.24024048078213e-05,
      "loss": 1.1724,
      "step": 254
    },
    {
      "epoch": 127.5,
      "grad_norm": 0.5455033779144287,
      "learning_rate": 9.234262742348763e-05,
      "loss": 0.739,
      "step": 255
    },
    {
      "epoch": 128.0,
      "grad_norm": 13.68809700012207,
      "learning_rate": 9.2282635291242e-05,
      "loss": 1.4461,
      "step": 256
    },
    {
      "epoch": 128.5,
      "grad_norm": 1.1735397577285767,
      "learning_rate": 9.222242871534484e-05,
      "loss": 0.7848,
      "step": 257
    },
    {
      "epoch": 129.0,
      "grad_norm": 4.6573991775512695,
      "learning_rate": 9.216200800114412e-05,
      "loss": 0.5311,
      "step": 258
    },
    {
      "epoch": 129.5,
      "grad_norm": 0.7705402374267578,
      "learning_rate": 9.21013734550739e-05,
      "loss": 0.8103,
      "step": 259
    },
    {
      "epoch": 130.0,
      "grad_norm": 39.913673400878906,
      "learning_rate": 9.204052538465269e-05,
      "loss": 3.3263,
      "step": 260
    },
    {
      "epoch": 130.5,
      "grad_norm": 0.7834247946739197,
      "learning_rate": 9.197946409848194e-05,
      "loss": 0.6385,
      "step": 261
    },
    {
      "epoch": 131.0,
      "grad_norm": 1.4882315397262573,
      "learning_rate": 9.191818990624448e-05,
      "loss": 1.0064,
      "step": 262
    },
    {
      "epoch": 131.5,
      "grad_norm": 1.2116917371749878,
      "learning_rate": 9.185670311870286e-05,
      "loss": 0.7524,
      "step": 263
    },
    {
      "epoch": 132.0,
      "grad_norm": 2.26658296585083,
      "learning_rate": 9.179500404769792e-05,
      "loss": 0.5387,
      "step": 264
    },
    {
      "epoch": 132.5,
      "grad_norm": 0.5079756379127502,
      "learning_rate": 9.173309300614709e-05,
      "loss": 0.7884,
      "step": 265
    },
    {
      "epoch": 133.0,
      "grad_norm": 2.116687297821045,
      "learning_rate": 9.167097030804288e-05,
      "loss": 0.5047,
      "step": 266
    },
    {
      "epoch": 133.5,
      "grad_norm": 1.283274531364441,
      "learning_rate": 9.16086362684512e-05,
      "loss": 0.7568,
      "step": 267
    },
    {
      "epoch": 134.0,
      "grad_norm": 1.3869727849960327,
      "learning_rate": 9.154609120350987e-05,
      "loss": 0.4355,
      "step": 268
    },
    {
      "epoch": 134.5,
      "grad_norm": 1.6364341974258423,
      "learning_rate": 9.148333543042692e-05,
      "loss": 0.8209,
      "step": 269
    },
    {
      "epoch": 135.0,
      "grad_norm": 1.3697168827056885,
      "learning_rate": 9.142036926747904e-05,
      "loss": 0.4634,
      "step": 270
    },
    {
      "epoch": 135.5,
      "grad_norm": 1.1338863372802734,
      "learning_rate": 9.135719303400993e-05,
      "loss": 0.71,
      "step": 271
    },
    {
      "epoch": 136.0,
      "grad_norm": 0.8962456583976746,
      "learning_rate": 9.129380705042873e-05,
      "loss": 0.433,
      "step": 272
    },
    {
      "epoch": 136.5,
      "grad_norm": 0.6447909474372864,
      "learning_rate": 9.123021163820839e-05,
      "loss": 0.7221,
      "step": 273
    },
    {
      "epoch": 137.0,
      "grad_norm": 1.3329360485076904,
      "learning_rate": 9.116640711988391e-05,
      "loss": 0.5899,
      "step": 274
    },
    {
      "epoch": 137.5,
      "grad_norm": 1.0404627323150635,
      "learning_rate": 9.110239381905091e-05,
      "loss": 0.8401,
      "step": 275
    },
    {
      "epoch": 138.0,
      "grad_norm": 5.447935581207275,
      "learning_rate": 9.103817206036382e-05,
      "loss": 0.7054,
      "step": 276
    },
    {
      "epoch": 138.5,
      "grad_norm": 0.4726102948188782,
      "learning_rate": 9.097374216953434e-05,
      "loss": 0.7456,
      "step": 277
    },
    {
      "epoch": 139.0,
      "grad_norm": 1.4676084518432617,
      "learning_rate": 9.090910447332975e-05,
      "loss": 0.4661,
      "step": 278
    },
    {
      "epoch": 139.5,
      "grad_norm": 2.533809185028076,
      "learning_rate": 9.08442592995712e-05,
      "loss": 0.8539,
      "step": 279
    },
    {
      "epoch": 140.0,
      "grad_norm": 1.3265609741210938,
      "learning_rate": 9.077920697713215e-05,
      "loss": 0.4764,
      "step": 280
    },
    {
      "epoch": 140.5,
      "grad_norm": 1.6127049922943115,
      "learning_rate": 9.071394783593663e-05,
      "loss": 0.6786,
      "step": 281
    },
    {
      "epoch": 141.0,
      "grad_norm": 1.0567877292633057,
      "learning_rate": 9.064848220695761e-05,
      "loss": 1.0216,
      "step": 282
    },
    {
      "epoch": 141.5,
      "grad_norm": 0.530373215675354,
      "learning_rate": 9.058281042221524e-05,
      "loss": 0.739,
      "step": 283
    },
    {
      "epoch": 142.0,
      "grad_norm": 3.3256218433380127,
      "learning_rate": 9.05169328147753e-05,
      "loss": 0.5873,
      "step": 284
    },
    {
      "epoch": 142.5,
      "grad_norm": 0.9769861102104187,
      "learning_rate": 9.045084971874738e-05,
      "loss": 0.8339,
      "step": 285
    },
    {
      "epoch": 143.0,
      "grad_norm": 1.3872549533843994,
      "learning_rate": 9.038456146928325e-05,
      "loss": 1.0337,
      "step": 286
    },
    {
      "epoch": 143.5,
      "grad_norm": 1.6853076219558716,
      "learning_rate": 9.031806840257519e-05,
      "loss": 0.8355,
      "step": 287
    },
    {
      "epoch": 144.0,
      "grad_norm": 1.8478177785873413,
      "learning_rate": 9.025137085585418e-05,
      "loss": 0.721,
      "step": 288
    },
    {
      "epoch": 144.5,
      "grad_norm": 1.6464306116104126,
      "learning_rate": 9.01844691673883e-05,
      "loss": 0.921,
      "step": 289
    },
    {
      "epoch": 145.0,
      "grad_norm": 1.9151690006256104,
      "learning_rate": 9.011736367648098e-05,
      "loss": 0.9846,
      "step": 290
    },
    {
      "epoch": 145.5,
      "grad_norm": 0.9094486832618713,
      "learning_rate": 9.005005472346924e-05,
      "loss": 0.8639,
      "step": 291
    },
    {
      "epoch": 146.0,
      "grad_norm": 5.555516719818115,
      "learning_rate": 8.998254264972198e-05,
      "loss": 0.6165,
      "step": 292
    },
    {
      "epoch": 146.5,
      "grad_norm": 1.1589049100875854,
      "learning_rate": 8.991482779763832e-05,
      "loss": 0.6636,
      "step": 293
    },
    {
      "epoch": 147.0,
      "grad_norm": 2.788694381713867,
      "learning_rate": 8.984691051064575e-05,
      "loss": 0.4547,
      "step": 294
    },
    {
      "epoch": 147.5,
      "grad_norm": 1.821389079093933,
      "learning_rate": 8.977879113319847e-05,
      "loss": 0.8994,
      "step": 295
    },
    {
      "epoch": 148.0,
      "grad_norm": 3.1623356342315674,
      "learning_rate": 8.97104700107756e-05,
      "loss": 0.649,
      "step": 296
    },
    {
      "epoch": 148.5,
      "grad_norm": 1.8579434156417847,
      "learning_rate": 8.964194748987947e-05,
      "loss": 0.7382,
      "step": 297
    },
    {
      "epoch": 149.0,
      "grad_norm": 3.05380916595459,
      "learning_rate": 8.95732239180338e-05,
      "loss": 0.9142,
      "step": 298
    },
    {
      "epoch": 149.5,
      "grad_norm": 0.7432356476783752,
      "learning_rate": 8.950429964378202e-05,
      "loss": 0.7739,
      "step": 299
    },
    {
      "epoch": 150.0,
      "grad_norm": 1.6515918970108032,
      "learning_rate": 8.943517501668542e-05,
      "loss": 0.9694,
      "step": 300
    },
    {
      "epoch": 150.5,
      "grad_norm": 0.8094480037689209,
      "learning_rate": 8.936585038732141e-05,
      "loss": 0.6357,
      "step": 301
    },
    {
      "epoch": 151.0,
      "grad_norm": 5.245303630828857,
      "learning_rate": 8.929632610728178e-05,
      "loss": 0.962,
      "step": 302
    },
    {
      "epoch": 151.5,
      "grad_norm": 0.9996303915977478,
      "learning_rate": 8.922660252917087e-05,
      "loss": 0.7182,
      "step": 303
    },
    {
      "epoch": 152.0,
      "grad_norm": 6.265303611755371,
      "learning_rate": 8.915668000660377e-05,
      "loss": 1.092,
      "step": 304
    },
    {
      "epoch": 152.5,
      "grad_norm": 0.7588598728179932,
      "learning_rate": 8.908655889420459e-05,
      "loss": 0.6088,
      "step": 305
    },
    {
      "epoch": 153.0,
      "grad_norm": 3.630662441253662,
      "learning_rate": 8.90162395476046e-05,
      "loss": 0.9619,
      "step": 306
    },
    {
      "epoch": 153.5,
      "grad_norm": 1.2417455911636353,
      "learning_rate": 8.894572232344042e-05,
      "loss": 0.7174,
      "step": 307
    },
    {
      "epoch": 154.0,
      "grad_norm": 0.7752083539962769,
      "learning_rate": 8.887500757935233e-05,
      "loss": 0.3885,
      "step": 308
    },
    {
      "epoch": 154.5,
      "grad_norm": 0.5876250863075256,
      "learning_rate": 8.880409567398224e-05,
      "loss": 0.7832,
      "step": 309
    },
    {
      "epoch": 155.0,
      "grad_norm": 0.9627898335456848,
      "learning_rate": 8.873298696697212e-05,
      "loss": 1.0326,
      "step": 310
    },
    {
      "epoch": 155.5,
      "grad_norm": 0.49632003903388977,
      "learning_rate": 8.866168181896198e-05,
      "loss": 0.6986,
      "step": 311
    },
    {
      "epoch": 156.0,
      "grad_norm": 0.744441568851471,
      "learning_rate": 8.85901805915881e-05,
      "loss": 1.0803,
      "step": 312
    },
    {
      "epoch": 156.5,
      "grad_norm": 0.9367627501487732,
      "learning_rate": 8.85184836474813e-05,
      "loss": 0.6632,
      "step": 313
    },
    {
      "epoch": 157.0,
      "grad_norm": 0.9376718401908875,
      "learning_rate": 8.844659135026491e-05,
      "loss": 0.5301,
      "step": 314
    },
    {
      "epoch": 157.5,
      "grad_norm": 1.9632515907287598,
      "learning_rate": 8.83745040645531e-05,
      "loss": 0.7688,
      "step": 315
    },
    {
      "epoch": 158.0,
      "grad_norm": 0.915960431098938,
      "learning_rate": 8.83022221559489e-05,
      "loss": 1.0062,
      "step": 316
    },
    {
      "epoch": 158.5,
      "grad_norm": 0.7271867394447327,
      "learning_rate": 8.822974599104246e-05,
      "loss": 0.7535,
      "step": 317
    },
    {
      "epoch": 159.0,
      "grad_norm": 1.7722750902175903,
      "learning_rate": 8.81570759374091e-05,
      "loss": 0.4835,
      "step": 318
    },
    {
      "epoch": 159.5,
      "grad_norm": 0.6312844157218933,
      "learning_rate": 8.808421236360746e-05,
      "loss": 0.6573,
      "step": 319
    },
    {
      "epoch": 160.0,
      "grad_norm": 0.7766721248626709,
      "learning_rate": 8.80111556391777e-05,
      "loss": 0.3765,
      "step": 320
    },
    {
      "epoch": 160.5,
      "grad_norm": 0.9972564578056335,
      "learning_rate": 8.793790613463955e-05,
      "loss": 0.8579,
      "step": 321
    },
    {
      "epoch": 161.0,
      "grad_norm": 29.352285385131836,
      "learning_rate": 8.786446422149046e-05,
      "loss": 1.7202,
      "step": 322
    },
    {
      "epoch": 161.5,
      "grad_norm": 2.404219627380371,
      "learning_rate": 8.779083027220372e-05,
      "loss": 0.7875,
      "step": 323
    },
    {
      "epoch": 162.0,
      "grad_norm": 1.326485514640808,
      "learning_rate": 8.771700466022654e-05,
      "loss": 0.8164,
      "step": 324
    },
    {
      "epoch": 162.5,
      "grad_norm": 1.1703271865844727,
      "learning_rate": 8.764298775997822e-05,
      "loss": 0.6902,
      "step": 325
    },
    {
      "epoch": 163.0,
      "grad_norm": 1.1087487936019897,
      "learning_rate": 8.756877994684818e-05,
      "loss": 0.5814,
      "step": 326
    },
    {
      "epoch": 163.5,
      "grad_norm": 0.8074220418930054,
      "learning_rate": 8.74943815971941e-05,
      "loss": 0.7178,
      "step": 327
    },
    {
      "epoch": 164.0,
      "grad_norm": 0.7630646228790283,
      "learning_rate": 8.741979308834001e-05,
      "loss": 0.4454,
      "step": 328
    },
    {
      "epoch": 164.5,
      "grad_norm": 0.6241486668586731,
      "learning_rate": 8.734501479857431e-05,
      "loss": 0.6656,
      "step": 329
    },
    {
      "epoch": 165.0,
      "grad_norm": 1.545361042022705,
      "learning_rate": 8.727004710714799e-05,
      "loss": 0.5126,
      "step": 330
    },
    {
      "epoch": 165.5,
      "grad_norm": 1.4805878400802612,
      "learning_rate": 8.719489039427256e-05,
      "loss": 0.7296,
      "step": 331
    },
    {
      "epoch": 166.0,
      "grad_norm": 0.8873409628868103,
      "learning_rate": 8.711954504111821e-05,
      "loss": 1.2059,
      "step": 332
    },
    {
      "epoch": 166.5,
      "grad_norm": 0.591742217540741,
      "learning_rate": 8.704401142981184e-05,
      "loss": 0.8018,
      "step": 333
    },
    {
      "epoch": 167.0,
      "grad_norm": 0.7729637622833252,
      "learning_rate": 8.696828994343517e-05,
      "loss": 1.1429,
      "step": 334
    },
    {
      "epoch": 167.5,
      "grad_norm": 0.6546936631202698,
      "learning_rate": 8.68923809660227e-05,
      "loss": 0.6517,
      "step": 335
    },
    {
      "epoch": 168.0,
      "grad_norm": 3.3543756008148193,
      "learning_rate": 8.681628488255986e-05,
      "loss": 1.0598,
      "step": 336
    },
    {
      "epoch": 168.5,
      "grad_norm": 3.4378745555877686,
      "learning_rate": 8.674000207898103e-05,
      "loss": 0.9474,
      "step": 337
    },
    {
      "epoch": 169.0,
      "grad_norm": 1.4147807359695435,
      "learning_rate": 8.666353294216756e-05,
      "loss": 1.1191,
      "step": 338
    },
    {
      "epoch": 169.5,
      "grad_norm": 0.7648679614067078,
      "learning_rate": 8.658687785994579e-05,
      "loss": 0.7762,
      "step": 339
    },
    {
      "epoch": 170.0,
      "grad_norm": 1.4035582542419434,
      "learning_rate": 8.651003722108517e-05,
      "loss": 0.8485,
      "step": 340
    },
    {
      "epoch": 170.5,
      "grad_norm": 0.5793386697769165,
      "learning_rate": 8.643301141529619e-05,
      "loss": 0.7562,
      "step": 341
    },
    {
      "epoch": 171.0,
      "grad_norm": 1.156826376914978,
      "learning_rate": 8.635580083322848e-05,
      "loss": 1.0532,
      "step": 342
    },
    {
      "epoch": 171.5,
      "grad_norm": 1.6001986265182495,
      "learning_rate": 8.627840586646876e-05,
      "loss": 0.7128,
      "step": 343
    },
    {
      "epoch": 172.0,
      "grad_norm": 1.0375338792800903,
      "learning_rate": 8.62008269075389e-05,
      "loss": 0.5411,
      "step": 344
    },
    {
      "epoch": 172.5,
      "grad_norm": 1.1245075464248657,
      "learning_rate": 8.612306434989395e-05,
      "loss": 0.6708,
      "step": 345
    },
    {
      "epoch": 173.0,
      "grad_norm": 1.2729289531707764,
      "learning_rate": 8.604511858792006e-05,
      "loss": 0.7413,
      "step": 346
    },
    {
      "epoch": 173.5,
      "grad_norm": 0.42658865451812744,
      "learning_rate": 8.596699001693255e-05,
      "loss": 0.5944,
      "step": 347
    },
    {
      "epoch": 174.0,
      "grad_norm": 1.5467840433120728,
      "learning_rate": 8.588867903317394e-05,
      "loss": 0.5171,
      "step": 348
    },
    {
      "epoch": 174.5,
      "grad_norm": 4.135446071624756,
      "learning_rate": 8.58101860338118e-05,
      "loss": 0.7633,
      "step": 349
    },
    {
      "epoch": 175.0,
      "grad_norm": 1.0240501165390015,
      "learning_rate": 8.573151141693689e-05,
      "loss": 0.3902,
      "step": 350
    },
    {
      "epoch": 175.5,
      "grad_norm": 0.7236465215682983,
      "learning_rate": 8.565265558156101e-05,
      "loss": 0.7579,
      "step": 351
    },
    {
      "epoch": 176.0,
      "grad_norm": 1.2244964838027954,
      "learning_rate": 8.557361892761513e-05,
      "loss": 0.4992,
      "step": 352
    },
    {
      "epoch": 176.5,
      "grad_norm": 1.2342673540115356,
      "learning_rate": 8.549440185594722e-05,
      "loss": 0.6983,
      "step": 353
    },
    {
      "epoch": 177.0,
      "grad_norm": 2.100895404815674,
      "learning_rate": 8.541500476832025e-05,
      "loss": 0.7569,
      "step": 354
    },
    {
      "epoch": 177.5,
      "grad_norm": 0.8631190061569214,
      "learning_rate": 8.53354280674102e-05,
      "loss": 0.699,
      "step": 355
    },
    {
      "epoch": 178.0,
      "grad_norm": 0.9449084401130676,
      "learning_rate": 8.525567215680398e-05,
      "loss": 0.4959,
      "step": 356
    },
    {
      "epoch": 178.5,
      "grad_norm": 0.5184299945831299,
      "learning_rate": 8.517573744099741e-05,
      "loss": 0.6175,
      "step": 357
    },
    {
      "epoch": 179.0,
      "grad_norm": 1.3312634229660034,
      "learning_rate": 8.509562432539313e-05,
      "loss": 0.4793,
      "step": 358
    },
    {
      "epoch": 179.5,
      "grad_norm": 1.3549054861068726,
      "learning_rate": 8.501533321629857e-05,
      "loss": 0.6387,
      "step": 359
    },
    {
      "epoch": 180.0,
      "grad_norm": 1.9658188819885254,
      "learning_rate": 8.493486452092392e-05,
      "loss": 0.6067,
      "step": 360
    },
    {
      "epoch": 180.5,
      "grad_norm": 1.1249042749404907,
      "learning_rate": 8.485421864737996e-05,
      "loss": 0.7603,
      "step": 361
    },
    {
      "epoch": 181.0,
      "grad_norm": 1.4607802629470825,
      "learning_rate": 8.477339600467611e-05,
      "loss": 0.5549,
      "step": 362
    },
    {
      "epoch": 181.5,
      "grad_norm": 0.9424148201942444,
      "learning_rate": 8.469239700271829e-05,
      "loss": 0.6994,
      "step": 363
    },
    {
      "epoch": 182.0,
      "grad_norm": 1.8076927661895752,
      "learning_rate": 8.461122205230689e-05,
      "loss": 0.5633,
      "step": 364
    },
    {
      "epoch": 182.5,
      "grad_norm": 0.8265047073364258,
      "learning_rate": 8.452987156513456e-05,
      "loss": 0.7273,
      "step": 365
    },
    {
      "epoch": 183.0,
      "grad_norm": 1.0330846309661865,
      "learning_rate": 8.444834595378434e-05,
      "loss": 0.4688,
      "step": 366
    },
    {
      "epoch": 183.5,
      "grad_norm": 1.6901235580444336,
      "learning_rate": 8.436664563172733e-05,
      "loss": 0.6996,
      "step": 367
    },
    {
      "epoch": 184.0,
      "grad_norm": 20.18507957458496,
      "learning_rate": 8.428477101332078e-05,
      "loss": 3.3022,
      "step": 368
    },
    {
      "epoch": 184.5,
      "grad_norm": 1.0538996458053589,
      "learning_rate": 8.420272251380588e-05,
      "loss": 0.781,
      "step": 369
    },
    {
      "epoch": 185.0,
      "grad_norm": 1.4863691329956055,
      "learning_rate": 8.412050054930571e-05,
      "loss": 1.076,
      "step": 370
    },
    {
      "epoch": 185.5,
      "grad_norm": 0.8447358012199402,
      "learning_rate": 8.403810553682307e-05,
      "loss": 0.6498,
      "step": 371
    },
    {
      "epoch": 186.0,
      "grad_norm": 1.0177277326583862,
      "learning_rate": 8.395553789423845e-05,
      "loss": 0.9183,
      "step": 372
    },
    {
      "epoch": 186.5,
      "grad_norm": 0.7734125256538391,
      "learning_rate": 8.387279804030781e-05,
      "loss": 0.6859,
      "step": 373
    },
    {
      "epoch": 187.0,
      "grad_norm": 1.395796775817871,
      "learning_rate": 8.378988639466057e-05,
      "loss": 0.5381,
      "step": 374
    },
    {
      "epoch": 187.5,
      "grad_norm": 0.6461387276649475,
      "learning_rate": 8.370680337779737e-05,
      "loss": 0.7091,
      "step": 375
    },
    {
      "epoch": 188.0,
      "grad_norm": 3.7634615898132324,
      "learning_rate": 8.362354941108802e-05,
      "loss": 0.7711,
      "step": 376
    },
    {
      "epoch": 188.5,
      "grad_norm": 0.8386605978012085,
      "learning_rate": 8.354012491676933e-05,
      "loss": 0.6811,
      "step": 377
    },
    {
      "epoch": 189.0,
      "grad_norm": 3.661172866821289,
      "learning_rate": 8.345653031794292e-05,
      "loss": 0.7433,
      "step": 378
    },
    {
      "epoch": 189.5,
      "grad_norm": 0.8018518090248108,
      "learning_rate": 8.33727660385732e-05,
      "loss": 0.6708,
      "step": 379
    },
    {
      "epoch": 190.0,
      "grad_norm": 1.4787194728851318,
      "learning_rate": 8.328883250348509e-05,
      "loss": 0.4978,
      "step": 380
    },
    {
      "epoch": 190.5,
      "grad_norm": 0.7550627589225769,
      "learning_rate": 8.320473013836196e-05,
      "loss": 0.6555,
      "step": 381
    },
    {
      "epoch": 191.0,
      "grad_norm": 6.441933631896973,
      "learning_rate": 8.312045936974339e-05,
      "loss": 1.4551,
      "step": 382
    },
    {
      "epoch": 191.5,
      "grad_norm": 0.9646530151367188,
      "learning_rate": 8.303602062502307e-05,
      "loss": 0.7053,
      "step": 383
    },
    {
      "epoch": 192.0,
      "grad_norm": 1.9152698516845703,
      "learning_rate": 8.295141433244659e-05,
      "loss": 0.7111,
      "step": 384
    },
    {
      "epoch": 192.5,
      "grad_norm": 1.044423222541809,
      "learning_rate": 8.286664092110935e-05,
      "loss": 0.7094,
      "step": 385
    },
    {
      "epoch": 193.0,
      "grad_norm": 0.8614663481712341,
      "learning_rate": 8.278170082095423e-05,
      "loss": 0.4728,
      "step": 386
    },
    {
      "epoch": 193.5,
      "grad_norm": 0.49174126982688904,
      "learning_rate": 8.269659446276955e-05,
      "loss": 0.5719,
      "step": 387
    },
    {
      "epoch": 194.0,
      "grad_norm": 0.9148985147476196,
      "learning_rate": 8.261132227818684e-05,
      "loss": 0.4294,
      "step": 388
    },
    {
      "epoch": 194.5,
      "grad_norm": 1.0272809267044067,
      "learning_rate": 8.252588469967866e-05,
      "loss": 0.6664,
      "step": 389
    },
    {
      "epoch": 195.0,
      "grad_norm": 1.2198545932769775,
      "learning_rate": 8.244028216055634e-05,
      "loss": 0.4022,
      "step": 390
    },
    {
      "epoch": 195.5,
      "grad_norm": 1.267242193222046,
      "learning_rate": 8.235451509496789e-05,
      "loss": 0.6551,
      "step": 391
    },
    {
      "epoch": 196.0,
      "grad_norm": 1.340940237045288,
      "learning_rate": 8.226858393789571e-05,
      "loss": 0.788,
      "step": 392
    },
    {
      "epoch": 196.5,
      "grad_norm": 0.8046751618385315,
      "learning_rate": 8.218248912515442e-05,
      "loss": 0.6895,
      "step": 393
    },
    {
      "epoch": 197.0,
      "grad_norm": 1.768873929977417,
      "learning_rate": 8.209623109338871e-05,
      "loss": 0.5808,
      "step": 394
    },
    {
      "epoch": 197.5,
      "grad_norm": 0.5974808931350708,
      "learning_rate": 8.200981028007095e-05,
      "loss": 0.5866,
      "step": 395
    },
    {
      "epoch": 198.0,
      "grad_norm": 14.19038200378418,
      "learning_rate": 8.192322712349917e-05,
      "loss": 4.0742,
      "step": 396
    },
    {
      "epoch": 198.5,
      "grad_norm": 0.7362140417098999,
      "learning_rate": 8.183648206279475e-05,
      "loss": 0.7768,
      "step": 397
    },
    {
      "epoch": 199.0,
      "grad_norm": 1.8577327728271484,
      "learning_rate": 8.174957553790014e-05,
      "loss": 0.6989,
      "step": 398
    },
    {
      "epoch": 199.5,
      "grad_norm": 0.8037870526313782,
      "learning_rate": 8.166250798957676e-05,
      "loss": 0.6342,
      "step": 399
    },
    {
      "epoch": 200.0,
      "grad_norm": 1.4529780149459839,
      "learning_rate": 8.157527985940258e-05,
      "loss": 1.0457,
      "step": 400
    },
    {
      "epoch": 200.5,
      "grad_norm": 1.0832105875015259,
      "learning_rate": 8.148789158977013e-05,
      "loss": 0.7174,
      "step": 401
    },
    {
      "epoch": 201.0,
      "grad_norm": 0.9042577743530273,
      "learning_rate": 8.140034362388399e-05,
      "loss": 0.45,
      "step": 402
    },
    {
      "epoch": 201.5,
      "grad_norm": 0.46869227290153503,
      "learning_rate": 8.13126364057587e-05,
      "loss": 0.6832,
      "step": 403
    },
    {
      "epoch": 202.0,
      "grad_norm": 2.073945999145508,
      "learning_rate": 8.122477038021654e-05,
      "loss": 0.7281,
      "step": 404
    },
    {
      "epoch": 202.5,
      "grad_norm": 0.5627115368843079,
      "learning_rate": 8.11367459928851e-05,
      "loss": 0.781,
      "step": 405
    },
    {
      "epoch": 203.0,
      "grad_norm": 1.5018227100372314,
      "learning_rate": 8.104856369019524e-05,
      "loss": 0.4228,
      "step": 406
    },
    {
      "epoch": 203.5,
      "grad_norm": 0.451335608959198,
      "learning_rate": 8.096022391937866e-05,
      "loss": 0.5945,
      "step": 407
    },
    {
      "epoch": 204.0,
      "grad_norm": 1.4502065181732178,
      "learning_rate": 8.087172712846565e-05,
      "loss": 0.4664,
      "step": 408
    },
    {
      "epoch": 204.5,
      "grad_norm": 0.6502431631088257,
      "learning_rate": 8.07830737662829e-05,
      "loss": 0.7066,
      "step": 409
    },
    {
      "epoch": 205.0,
      "grad_norm": 1.70822012424469,
      "learning_rate": 8.069426428245124e-05,
      "loss": 0.7412,
      "step": 410
    },
    {
      "epoch": 205.5,
      "grad_norm": 0.9970735311508179,
      "learning_rate": 8.060529912738315e-05,
      "loss": 0.6078,
      "step": 411
    },
    {
      "epoch": 206.0,
      "grad_norm": 0.9485369324684143,
      "learning_rate": 8.051617875228072e-05,
      "loss": 0.4414,
      "step": 412
    },
    {
      "epoch": 206.5,
      "grad_norm": 2.461491584777832,
      "learning_rate": 8.04269036091333e-05,
      "loss": 0.7459,
      "step": 413
    },
    {
      "epoch": 207.0,
      "grad_norm": 1.0793882608413696,
      "learning_rate": 8.033747415071507e-05,
      "loss": 0.3834,
      "step": 414
    },
    {
      "epoch": 207.5,
      "grad_norm": 0.3889921307563782,
      "learning_rate": 8.024789083058289e-05,
      "loss": 0.7006,
      "step": 415
    },
    {
      "epoch": 208.0,
      "grad_norm": 4.822055816650391,
      "learning_rate": 8.015815410307398e-05,
      "loss": 0.6246,
      "step": 416
    },
    {
      "epoch": 208.5,
      "grad_norm": 0.41584470868110657,
      "learning_rate": 8.006826442330361e-05,
      "loss": 0.6116,
      "step": 417
    },
    {
      "epoch": 209.0,
      "grad_norm": 1.2937344312667847,
      "learning_rate": 7.997822224716269e-05,
      "loss": 0.5201,
      "step": 418
    },
    {
      "epoch": 209.5,
      "grad_norm": 0.8473069667816162,
      "learning_rate": 7.988802803131562e-05,
      "loss": 0.6621,
      "step": 419
    },
    {
      "epoch": 210.0,
      "grad_norm": 1.9003219604492188,
      "learning_rate": 7.979768223319785e-05,
      "loss": 0.5304,
      "step": 420
    },
    {
      "epoch": 210.5,
      "grad_norm": 0.6936240792274475,
      "learning_rate": 7.970718531101365e-05,
      "loss": 0.7693,
      "step": 421
    },
    {
      "epoch": 211.0,
      "grad_norm": 1.8913812637329102,
      "learning_rate": 7.961653772373365e-05,
      "loss": 0.5842,
      "step": 422
    },
    {
      "epoch": 211.5,
      "grad_norm": 0.7283148765563965,
      "learning_rate": 7.952573993109274e-05,
      "loss": 0.7449,
      "step": 423
    },
    {
      "epoch": 212.0,
      "grad_norm": 1.1354438066482544,
      "learning_rate": 7.943479239358748e-05,
      "loss": 0.4055,
      "step": 424
    },
    {
      "epoch": 212.5,
      "grad_norm": 0.7347318530082703,
      "learning_rate": 7.934369557247397e-05,
      "loss": 0.6287,
      "step": 425
    },
    {
      "epoch": 213.0,
      "grad_norm": 3.0338501930236816,
      "learning_rate": 7.925244992976538e-05,
      "loss": 0.4772,
      "step": 426
    },
    {
      "epoch": 213.5,
      "grad_norm": 0.8360968232154846,
      "learning_rate": 7.916105592822967e-05,
      "loss": 0.6218,
      "step": 427
    },
    {
      "epoch": 214.0,
      "grad_norm": 16.154359817504883,
      "learning_rate": 7.906951403138722e-05,
      "loss": 1.598,
      "step": 428
    },
    {
      "epoch": 214.5,
      "grad_norm": 1.0661816596984863,
      "learning_rate": 7.89778247035085e-05,
      "loss": 0.712,
      "step": 429
    },
    {
      "epoch": 215.0,
      "grad_norm": 0.785302996635437,
      "learning_rate": 7.888598840961169e-05,
      "loss": 1.0996,
      "step": 430
    },
    {
      "epoch": 215.5,
      "grad_norm": 0.9986799359321594,
      "learning_rate": 7.879400561546033e-05,
      "loss": 0.6672,
      "step": 431
    },
    {
      "epoch": 216.0,
      "grad_norm": 2.153475761413574,
      "learning_rate": 7.870187678756098e-05,
      "loss": 0.6346,
      "step": 432
    },
    {
      "epoch": 216.5,
      "grad_norm": 0.6204362511634827,
      "learning_rate": 7.860960239316083e-05,
      "loss": 0.6348,
      "step": 433
    },
    {
      "epoch": 217.0,
      "grad_norm": 1.1533006429672241,
      "learning_rate": 7.85171829002453e-05,
      "loss": 0.4502,
      "step": 434
    },
    {
      "epoch": 217.5,
      "grad_norm": 0.7941988110542297,
      "learning_rate": 7.842461877753575e-05,
      "loss": 0.4828,
      "step": 435
    },
    {
      "epoch": 218.0,
      "grad_norm": 3.4231910705566406,
      "learning_rate": 7.833191049448706e-05,
      "loss": 0.9052,
      "step": 436
    },
    {
      "epoch": 218.5,
      "grad_norm": 0.8901759386062622,
      "learning_rate": 7.82390585212852e-05,
      "loss": 0.5493,
      "step": 437
    },
    {
      "epoch": 219.0,
      "grad_norm": 2.2517404556274414,
      "learning_rate": 7.814606332884489e-05,
      "loss": 1.2113,
      "step": 438
    },
    {
      "epoch": 219.5,
      "grad_norm": 0.9658966064453125,
      "learning_rate": 7.805292538880724e-05,
      "loss": 0.7854,
      "step": 439
    },
    {
      "epoch": 220.0,
      "grad_norm": 1.139497995376587,
      "learning_rate": 7.795964517353735e-05,
      "loss": 1.0373,
      "step": 440
    },
    {
      "epoch": 220.5,
      "grad_norm": 0.5400660634040833,
      "learning_rate": 7.786622315612183e-05,
      "loss": 0.7205,
      "step": 441
    },
    {
      "epoch": 221.0,
      "grad_norm": 1.1441700458526611,
      "learning_rate": 7.77726598103665e-05,
      "loss": 0.4028,
      "step": 442
    },
    {
      "epoch": 221.5,
      "grad_norm": 1.1220710277557373,
      "learning_rate": 7.767895561079397e-05,
      "loss": 0.621,
      "step": 443
    },
    {
      "epoch": 222.0,
      "grad_norm": 1.529765248298645,
      "learning_rate": 7.758511103264117e-05,
      "loss": 0.4412,
      "step": 444
    },
    {
      "epoch": 222.5,
      "grad_norm": 2.0216169357299805,
      "learning_rate": 7.7491126551857e-05,
      "loss": 0.7247,
      "step": 445
    },
    {
      "epoch": 223.0,
      "grad_norm": 1.8950647115707397,
      "learning_rate": 7.739700264509992e-05,
      "loss": 0.4145,
      "step": 446
    },
    {
      "epoch": 223.5,
      "grad_norm": 0.8040174841880798,
      "learning_rate": 7.730273978973552e-05,
      "loss": 0.6753,
      "step": 447
    },
    {
      "epoch": 224.0,
      "grad_norm": 1.3904660940170288,
      "learning_rate": 7.720833846383402e-05,
      "loss": 0.3783,
      "step": 448
    },
    {
      "epoch": 224.5,
      "grad_norm": 2.2562403678894043,
      "learning_rate": 7.711379914616803e-05,
      "loss": 0.6707,
      "step": 449
    },
    {
      "epoch": 225.0,
      "grad_norm": 2.084462881088257,
      "learning_rate": 7.70191223162099e-05,
      "loss": 0.423,
      "step": 450
    },
    {
      "epoch": 225.5,
      "grad_norm": 1.4259127378463745,
      "learning_rate": 7.692430845412945e-05,
      "loss": 0.6086,
      "step": 451
    },
    {
      "epoch": 226.0,
      "grad_norm": 1.279314398765564,
      "learning_rate": 7.68293580407915e-05,
      "loss": 0.4839,
      "step": 452
    },
    {
      "epoch": 226.5,
      "grad_norm": 0.7970501780509949,
      "learning_rate": 7.673427155775336e-05,
      "loss": 0.5556,
      "step": 453
    },
    {
      "epoch": 227.0,
      "grad_norm": 1.4189525842666626,
      "learning_rate": 7.66390494872625e-05,
      "loss": 0.4023,
      "step": 454
    },
    {
      "epoch": 227.5,
      "grad_norm": 1.2644344568252563,
      "learning_rate": 7.654369231225398e-05,
      "loss": 0.6492,
      "step": 455
    },
    {
      "epoch": 228.0,
      "grad_norm": 15.089214324951172,
      "learning_rate": 7.644820051634812e-05,
      "loss": 1.8448,
      "step": 456
    },
    {
      "epoch": 228.5,
      "grad_norm": 1.3588922023773193,
      "learning_rate": 7.635257458384799e-05,
      "loss": 0.6282,
      "step": 457
    },
    {
      "epoch": 229.0,
      "grad_norm": 1.6629581451416016,
      "learning_rate": 7.625681499973692e-05,
      "loss": 0.3983,
      "step": 458
    },
    {
      "epoch": 229.5,
      "grad_norm": 0.8579509258270264,
      "learning_rate": 7.61609222496761e-05,
      "loss": 0.5012,
      "step": 459
    },
    {
      "epoch": 230.0,
      "grad_norm": 2.2509641647338867,
      "learning_rate": 7.606489682000211e-05,
      "loss": 0.5443,
      "step": 460
    },
    {
      "epoch": 230.5,
      "grad_norm": 1.495695948600769,
      "learning_rate": 7.596873919772438e-05,
      "loss": 0.6166,
      "step": 461
    },
    {
      "epoch": 231.0,
      "grad_norm": 3.4327473640441895,
      "learning_rate": 7.587244987052287e-05,
      "loss": 0.5054,
      "step": 462
    },
    {
      "epoch": 231.5,
      "grad_norm": 1.2567977905273438,
      "learning_rate": 7.57760293267454e-05,
      "loss": 0.6669,
      "step": 463
    },
    {
      "epoch": 232.0,
      "grad_norm": 1.0312711000442505,
      "learning_rate": 7.567947805540537e-05,
      "loss": 0.9708,
      "step": 464
    },
    {
      "epoch": 232.5,
      "grad_norm": 8.136346817016602,
      "learning_rate": 7.558279654617912e-05,
      "loss": 0.7832,
      "step": 465
    },
    {
      "epoch": 233.0,
      "grad_norm": 1.1686395406723022,
      "learning_rate": 7.548598528940353e-05,
      "loss": 0.3975,
      "step": 466
    },
    {
      "epoch": 233.5,
      "grad_norm": 2.484499454498291,
      "learning_rate": 7.538904477607356e-05,
      "loss": 0.716,
      "step": 467
    },
    {
      "epoch": 234.0,
      "grad_norm": 2.3993451595306396,
      "learning_rate": 7.529197549783967e-05,
      "loss": 0.4798,
      "step": 468
    },
    {
      "epoch": 234.5,
      "grad_norm": 2.4435184001922607,
      "learning_rate": 7.519477794700538e-05,
      "loss": 0.8925,
      "step": 469
    },
    {
      "epoch": 235.0,
      "grad_norm": 1.2841492891311646,
      "learning_rate": 7.509745261652479e-05,
      "loss": 0.4291,
      "step": 470
    },
    {
      "epoch": 235.5,
      "grad_norm": 1.3279870748519897,
      "learning_rate": 7.500000000000001e-05,
      "loss": 0.6263,
      "step": 471
    },
    {
      "epoch": 236.0,
      "grad_norm": 1.6670854091644287,
      "learning_rate": 7.490242059167876e-05,
      "loss": 0.9435,
      "step": 472
    },
    {
      "epoch": 236.5,
      "grad_norm": 2.0598576068878174,
      "learning_rate": 7.480471488645175e-05,
      "loss": 0.7599,
      "step": 473
    },
    {
      "epoch": 237.0,
      "grad_norm": 1.1438533067703247,
      "learning_rate": 7.47068833798503e-05,
      "loss": 0.9044,
      "step": 474
    },
    {
      "epoch": 237.5,
      "grad_norm": 1.4255588054656982,
      "learning_rate": 7.460892656804366e-05,
      "loss": 0.6228,
      "step": 475
    },
    {
      "epoch": 238.0,
      "grad_norm": 2.163013219833374,
      "learning_rate": 7.451084494783668e-05,
      "loss": 0.5067,
      "step": 476
    },
    {
      "epoch": 238.5,
      "grad_norm": 0.5647773742675781,
      "learning_rate": 7.44126390166671e-05,
      "loss": 0.6117,
      "step": 477
    },
    {
      "epoch": 239.0,
      "grad_norm": 6.719013214111328,
      "learning_rate": 7.431430927260326e-05,
      "loss": 0.863,
      "step": 478
    },
    {
      "epoch": 239.5,
      "grad_norm": 0.6916800141334534,
      "learning_rate": 7.421585621434127e-05,
      "loss": 0.6143,
      "step": 479
    },
    {
      "epoch": 240.0,
      "grad_norm": 0.9313693642616272,
      "learning_rate": 7.411728034120279e-05,
      "loss": 0.383,
      "step": 480
    },
    {
      "epoch": 240.5,
      "grad_norm": 0.5331056118011475,
      "learning_rate": 7.401858215313228e-05,
      "loss": 0.5993,
      "step": 481
    },
    {
      "epoch": 241.0,
      "grad_norm": 1.4826580286026,
      "learning_rate": 7.391976215069456e-05,
      "loss": 0.3796,
      "step": 482
    },
    {
      "epoch": 241.5,
      "grad_norm": 0.5266751050949097,
      "learning_rate": 7.382082083507226e-05,
      "loss": 0.5729,
      "step": 483
    },
    {
      "epoch": 242.0,
      "grad_norm": 0.9854914546012878,
      "learning_rate": 7.372175870806328e-05,
      "loss": 1.1022,
      "step": 484
    },
    {
      "epoch": 242.5,
      "grad_norm": 0.43273159861564636,
      "learning_rate": 7.362257627207819e-05,
      "loss": 0.5501,
      "step": 485
    },
    {
      "epoch": 243.0,
      "grad_norm": 1.200194239616394,
      "learning_rate": 7.35232740301378e-05,
      "loss": 0.4598,
      "step": 486
    },
    {
      "epoch": 243.5,
      "grad_norm": 2.9032397270202637,
      "learning_rate": 7.342385248587043e-05,
      "loss": 0.7681,
      "step": 487
    },
    {
      "epoch": 244.0,
      "grad_norm": 1.0617011785507202,
      "learning_rate": 7.332431214350963e-05,
      "loss": 0.7239,
      "step": 488
    },
    {
      "epoch": 244.5,
      "grad_norm": 0.9474357962608337,
      "learning_rate": 7.322465350789126e-05,
      "loss": 0.6409,
      "step": 489
    },
    {
      "epoch": 245.0,
      "grad_norm": 2.160486936569214,
      "learning_rate": 7.312487708445128e-05,
      "loss": 0.4184,
      "step": 490
    },
    {
      "epoch": 245.5,
      "grad_norm": 0.7519797682762146,
      "learning_rate": 7.302498337922293e-05,
      "loss": 0.5701,
      "step": 491
    },
    {
      "epoch": 246.0,
      "grad_norm": 0.9340056777000427,
      "learning_rate": 7.292497289883433e-05,
      "loss": 0.4788,
      "step": 492
    },
    {
      "epoch": 246.5,
      "grad_norm": 0.6592388153076172,
      "learning_rate": 7.282484615050577e-05,
      "loss": 0.592,
      "step": 493
    },
    {
      "epoch": 247.0,
      "grad_norm": 2.502549409866333,
      "learning_rate": 7.272460364204733e-05,
      "loss": 0.5425,
      "step": 494
    },
    {
      "epoch": 247.5,
      "grad_norm": 0.40951523184776306,
      "learning_rate": 7.26242458818561e-05,
      "loss": 0.6294,
      "step": 495
    },
    {
      "epoch": 248.0,
      "grad_norm": 1.9842380285263062,
      "learning_rate": 7.252377337891365e-05,
      "loss": 0.5838,
      "step": 496
    },
    {
      "epoch": 248.5,
      "grad_norm": 3.001387357711792,
      "learning_rate": 7.242318664278362e-05,
      "loss": 0.7478,
      "step": 497
    },
    {
      "epoch": 249.0,
      "grad_norm": 3.081714391708374,
      "learning_rate": 7.232248618360889e-05,
      "loss": 0.4092,
      "step": 498
    },
    {
      "epoch": 249.5,
      "grad_norm": 4.5948662757873535,
      "learning_rate": 7.222167251210916e-05,
      "loss": 0.7334,
      "step": 499
    },
    {
      "epoch": 250.0,
      "grad_norm": 1.7451257705688477,
      "learning_rate": 7.212074613957828e-05,
      "loss": 0.8903,
      "step": 500
    },
    {
      "epoch": 250.5,
      "grad_norm": 0.5748724341392517,
      "learning_rate": 7.201970757788172e-05,
      "loss": 0.5369,
      "step": 501
    },
    {
      "epoch": 251.0,
      "grad_norm": 1.3469123840332031,
      "learning_rate": 7.191855733945387e-05,
      "loss": 0.4342,
      "step": 502
    },
    {
      "epoch": 251.5,
      "grad_norm": 1.7363576889038086,
      "learning_rate": 7.181729593729558e-05,
      "loss": 0.8242,
      "step": 503
    },
    {
      "epoch": 252.0,
      "grad_norm": 0.9408396482467651,
      "learning_rate": 7.171592388497144e-05,
      "loss": 0.3737,
      "step": 504
    },
    {
      "epoch": 252.5,
      "grad_norm": 1.0432978868484497,
      "learning_rate": 7.161444169660722e-05,
      "loss": 0.7717,
      "step": 505
    },
    {
      "epoch": 253.0,
      "grad_norm": 1.599176049232483,
      "learning_rate": 7.15128498868873e-05,
      "loss": 0.4873,
      "step": 506
    },
    {
      "epoch": 253.5,
      "grad_norm": 0.40978094935417175,
      "learning_rate": 7.141114897105202e-05,
      "loss": 0.7089,
      "step": 507
    },
    {
      "epoch": 254.0,
      "grad_norm": 2.1264853477478027,
      "learning_rate": 7.1309339464895e-05,
      "loss": 0.5753,
      "step": 508
    },
    {
      "epoch": 254.5,
      "grad_norm": 1.3448783159255981,
      "learning_rate": 7.12074218847607e-05,
      "loss": 0.7606,
      "step": 509
    },
    {
      "epoch": 255.0,
      "grad_norm": 1.1680799722671509,
      "learning_rate": 7.11053967475416e-05,
      "loss": 0.5901,
      "step": 510
    },
    {
      "epoch": 255.5,
      "grad_norm": 0.7040720582008362,
      "learning_rate": 7.100326457067575e-05,
      "loss": 0.6231,
      "step": 511
    },
    {
      "epoch": 256.0,
      "grad_norm": 1.1731950044631958,
      "learning_rate": 7.090102587214402e-05,
      "loss": 0.4883,
      "step": 512
    },
    {
      "epoch": 256.5,
      "grad_norm": 0.6589009165763855,
      "learning_rate": 7.079868117046755e-05,
      "loss": 0.6568,
      "step": 513
    },
    {
      "epoch": 257.0,
      "grad_norm": 1.5691967010498047,
      "learning_rate": 7.069623098470508e-05,
      "loss": 0.5694,
      "step": 514
    },
    {
      "epoch": 257.5,
      "grad_norm": 0.5299769043922424,
      "learning_rate": 7.059367583445034e-05,
      "loss": 0.5316,
      "step": 515
    },
    {
      "epoch": 258.0,
      "grad_norm": 2.020585775375366,
      "learning_rate": 7.049101623982937e-05,
      "loss": 0.3606,
      "step": 516
    },
    {
      "epoch": 258.5,
      "grad_norm": 0.6764979958534241,
      "learning_rate": 7.038825272149799e-05,
      "loss": 0.5875,
      "step": 517
    },
    {
      "epoch": 259.0,
      "grad_norm": 1.5628106594085693,
      "learning_rate": 7.028538580063903e-05,
      "loss": 0.4171,
      "step": 518
    },
    {
      "epoch": 259.5,
      "grad_norm": 1.6990623474121094,
      "learning_rate": 7.018241599895974e-05,
      "loss": 0.7662,
      "step": 519
    },
    {
      "epoch": 260.0,
      "grad_norm": 1.2228469848632812,
      "learning_rate": 7.007934383868917e-05,
      "loss": 0.4534,
      "step": 520
    },
    {
      "epoch": 260.5,
      "grad_norm": 0.6618345975875854,
      "learning_rate": 6.997616984257552e-05,
      "loss": 0.6385,
      "step": 521
    },
    {
      "epoch": 261.0,
      "grad_norm": 1.113016963005066,
      "learning_rate": 6.987289453388342e-05,
      "loss": 0.4592,
      "step": 522
    },
    {
      "epoch": 261.5,
      "grad_norm": 0.7863847017288208,
      "learning_rate": 6.976951843639134e-05,
      "loss": 0.6275,
      "step": 523
    },
    {
      "epoch": 262.0,
      "grad_norm": 1.592556118965149,
      "learning_rate": 6.966604207438894e-05,
      "loss": 0.4762,
      "step": 524
    },
    {
      "epoch": 262.5,
      "grad_norm": 2.564715623855591,
      "learning_rate": 6.956246597267437e-05,
      "loss": 0.8902,
      "step": 525
    },
    {
      "epoch": 263.0,
      "grad_norm": 2.3334994316101074,
      "learning_rate": 6.945879065655165e-05,
      "loss": 0.6336,
      "step": 526
    },
    {
      "epoch": 263.5,
      "grad_norm": 1.7134613990783691,
      "learning_rate": 6.935501665182791e-05,
      "loss": 0.7875,
      "step": 527
    },
    {
      "epoch": 264.0,
      "grad_norm": 1.9639478921890259,
      "learning_rate": 6.92511444848109e-05,
      "loss": 1.0636,
      "step": 528
    },
    {
      "epoch": 264.5,
      "grad_norm": 0.7804299592971802,
      "learning_rate": 6.914717468230614e-05,
      "loss": 0.642,
      "step": 529
    },
    {
      "epoch": 265.0,
      "grad_norm": 1.793656349182129,
      "learning_rate": 6.904310777161438e-05,
      "loss": 0.5575,
      "step": 530
    },
    {
      "epoch": 265.5,
      "grad_norm": 4.0193705558776855,
      "learning_rate": 6.89389442805288e-05,
      "loss": 0.8893,
      "step": 531
    },
    {
      "epoch": 266.0,
      "grad_norm": 1.378307819366455,
      "learning_rate": 6.88346847373325e-05,
      "loss": 0.4658,
      "step": 532
    },
    {
      "epoch": 266.5,
      "grad_norm": 1.1681323051452637,
      "learning_rate": 6.873032967079561e-05,
      "loss": 0.5641,
      "step": 533
    },
    {
      "epoch": 267.0,
      "grad_norm": 1.4974206686019897,
      "learning_rate": 6.862587961017283e-05,
      "loss": 0.4118,
      "step": 534
    },
    {
      "epoch": 267.5,
      "grad_norm": 0.5355740189552307,
      "learning_rate": 6.852133508520057e-05,
      "loss": 0.5712,
      "step": 535
    },
    {
      "epoch": 268.0,
      "grad_norm": 0.9718406796455383,
      "learning_rate": 6.841669662609436e-05,
      "loss": 0.3885,
      "step": 536
    },
    {
      "epoch": 268.5,
      "grad_norm": 0.6695818901062012,
      "learning_rate": 6.831196476354614e-05,
      "loss": 0.5619,
      "step": 537
    },
    {
      "epoch": 269.0,
      "grad_norm": 1.9289757013320923,
      "learning_rate": 6.820714002872153e-05,
      "loss": 0.3731,
      "step": 538
    },
    {
      "epoch": 269.5,
      "grad_norm": 1.2406436204910278,
      "learning_rate": 6.810222295325722e-05,
      "loss": 0.6799,
      "step": 539
    },
    {
      "epoch": 270.0,
      "grad_norm": 1.5236866474151611,
      "learning_rate": 6.799721406925816e-05,
      "loss": 0.6241,
      "step": 540
    },
    {
      "epoch": 270.5,
      "grad_norm": 0.8481728434562683,
      "learning_rate": 6.789211390929497e-05,
      "loss": 0.7699,
      "step": 541
    },
    {
      "epoch": 271.0,
      "grad_norm": 1.0108497142791748,
      "learning_rate": 6.778692300640115e-05,
      "loss": 0.9034,
      "step": 542
    },
    {
      "epoch": 271.5,
      "grad_norm": 0.7054284811019897,
      "learning_rate": 6.768164189407047e-05,
      "loss": 0.577,
      "step": 543
    },
    {
      "epoch": 272.0,
      "grad_norm": 0.961260974407196,
      "learning_rate": 6.757627110625416e-05,
      "loss": 0.3847,
      "step": 544
    },
    {
      "epoch": 272.5,
      "grad_norm": 0.6874154806137085,
      "learning_rate": 6.747081117735828e-05,
      "loss": 0.6187,
      "step": 545
    },
    {
      "epoch": 273.0,
      "grad_norm": 1.1369733810424805,
      "learning_rate": 6.736526264224101e-05,
      "loss": 0.453,
      "step": 546
    },
    {
      "epoch": 273.5,
      "grad_norm": 0.5786665678024292,
      "learning_rate": 6.725962603620988e-05,
      "loss": 0.5516,
      "step": 547
    },
    {
      "epoch": 274.0,
      "grad_norm": 15.180058479309082,
      "learning_rate": 6.715390189501906e-05,
      "loss": 1.6768,
      "step": 548
    },
    {
      "epoch": 274.5,
      "grad_norm": 0.7103780508041382,
      "learning_rate": 6.704809075486674e-05,
      "loss": 0.7192,
      "step": 549
    },
    {
      "epoch": 275.0,
      "grad_norm": 1.5910588502883911,
      "learning_rate": 6.694219315239227e-05,
      "loss": 0.399,
      "step": 550
    },
    {
      "epoch": 275.5,
      "grad_norm": 0.7732727527618408,
      "learning_rate": 6.683620962467357e-05,
      "loss": 0.5537,
      "step": 551
    },
    {
      "epoch": 276.0,
      "grad_norm": 1.4121403694152832,
      "learning_rate": 6.67301407092243e-05,
      "loss": 0.3472,
      "step": 552
    },
    {
      "epoch": 276.5,
      "grad_norm": 0.6016989946365356,
      "learning_rate": 6.662398694399119e-05,
      "loss": 0.6534,
      "step": 553
    },
    {
      "epoch": 277.0,
      "grad_norm": 0.7904322743415833,
      "learning_rate": 6.651774886735131e-05,
      "loss": 0.3585,
      "step": 554
    },
    {
      "epoch": 277.5,
      "grad_norm": 0.4645850956439972,
      "learning_rate": 6.641142701810932e-05,
      "loss": 0.509,
      "step": 555
    },
    {
      "epoch": 278.0,
      "grad_norm": 1.5916553735733032,
      "learning_rate": 6.630502193549474e-05,
      "loss": 0.863,
      "step": 556
    },
    {
      "epoch": 278.5,
      "grad_norm": 0.6173549294471741,
      "learning_rate": 6.619853415915927e-05,
      "loss": 0.5997,
      "step": 557
    },
    {
      "epoch": 279.0,
      "grad_norm": 0.9468961954116821,
      "learning_rate": 6.609196422917394e-05,
      "loss": 0.3834,
      "step": 558
    },
    {
      "epoch": 279.5,
      "grad_norm": 2.065488338470459,
      "learning_rate": 6.598531268602646e-05,
      "loss": 0.5864,
      "step": 559
    },
    {
      "epoch": 280.0,
      "grad_norm": 1.0914835929870605,
      "learning_rate": 6.587858007061847e-05,
      "loss": 0.4273,
      "step": 560
    },
    {
      "epoch": 280.5,
      "grad_norm": 0.7105666995048523,
      "learning_rate": 6.577176692426279e-05,
      "loss": 0.5817,
      "step": 561
    },
    {
      "epoch": 281.0,
      "grad_norm": 1.44219172000885,
      "learning_rate": 6.566487378868061e-05,
      "loss": 0.631,
      "step": 562
    },
    {
      "epoch": 281.5,
      "grad_norm": 1.2486913204193115,
      "learning_rate": 6.555790120599885e-05,
      "loss": 0.6416,
      "step": 563
    },
    {
      "epoch": 282.0,
      "grad_norm": 1.0740281343460083,
      "learning_rate": 6.545084971874738e-05,
      "loss": 0.4218,
      "step": 564
    },
    {
      "epoch": 282.5,
      "grad_norm": 1.6123507022857666,
      "learning_rate": 6.534371986985618e-05,
      "loss": 0.6842,
      "step": 565
    },
    {
      "epoch": 283.0,
      "grad_norm": 1.0170310735702515,
      "learning_rate": 6.523651220265268e-05,
      "loss": 0.4029,
      "step": 566
    },
    {
      "epoch": 283.5,
      "grad_norm": 1.0539488792419434,
      "learning_rate": 6.512922726085904e-05,
      "loss": 0.6037,
      "step": 567
    },
    {
      "epoch": 284.0,
      "grad_norm": 1.4019837379455566,
      "learning_rate": 6.502186558858925e-05,
      "loss": 0.4126,
      "step": 568
    },
    {
      "epoch": 284.5,
      "grad_norm": 0.6338628530502319,
      "learning_rate": 6.491442773034646e-05,
      "loss": 0.5258,
      "step": 569
    },
    {
      "epoch": 285.0,
      "grad_norm": 0.840011715888977,
      "learning_rate": 6.480691423102028e-05,
      "loss": 1.1035,
      "step": 570
    },
    {
      "epoch": 285.5,
      "grad_norm": 1.55778169631958,
      "learning_rate": 6.469932563588386e-05,
      "loss": 0.5671,
      "step": 571
    },
    {
      "epoch": 286.0,
      "grad_norm": 0.9740279912948608,
      "learning_rate": 6.459166249059125e-05,
      "loss": 0.3366,
      "step": 572
    },
    {
      "epoch": 286.5,
      "grad_norm": 1.9214400053024292,
      "learning_rate": 6.448392534117463e-05,
      "loss": 0.7472,
      "step": 573
    },
    {
      "epoch": 287.0,
      "grad_norm": 1.4927865266799927,
      "learning_rate": 6.437611473404142e-05,
      "loss": 0.6364,
      "step": 574
    },
    {
      "epoch": 287.5,
      "grad_norm": 3.6279613971710205,
      "learning_rate": 6.426823121597169e-05,
      "loss": 0.7989,
      "step": 575
    },
    {
      "epoch": 288.0,
      "grad_norm": 1.6166788339614868,
      "learning_rate": 6.416027533411519e-05,
      "loss": 0.8841,
      "step": 576
    },
    {
      "epoch": 288.5,
      "grad_norm": 2.182417869567871,
      "learning_rate": 6.40522476359888e-05,
      "loss": 0.6057,
      "step": 577
    },
    {
      "epoch": 289.0,
      "grad_norm": 10.826478004455566,
      "learning_rate": 6.39441486694735e-05,
      "loss": 1.3541,
      "step": 578
    },
    {
      "epoch": 289.5,
      "grad_norm": 0.6857351660728455,
      "learning_rate": 6.38359789828118e-05,
      "loss": 0.5884,
      "step": 579
    },
    {
      "epoch": 290.0,
      "grad_norm": 1.2314034700393677,
      "learning_rate": 6.372773912460483e-05,
      "loss": 0.3845,
      "step": 580
    },
    {
      "epoch": 290.5,
      "grad_norm": 1.252751350402832,
      "learning_rate": 6.361942964380967e-05,
      "loss": 0.7315,
      "step": 581
    },
    {
      "epoch": 291.0,
      "grad_norm": 0.7485074996948242,
      "learning_rate": 6.351105108973644e-05,
      "loss": 0.3279,
      "step": 582
    },
    {
      "epoch": 291.5,
      "grad_norm": 0.428264856338501,
      "learning_rate": 6.340260401204561e-05,
      "loss": 0.651,
      "step": 583
    },
    {
      "epoch": 292.0,
      "grad_norm": 1.844477653503418,
      "learning_rate": 6.329408896074518e-05,
      "loss": 1.1602,
      "step": 584
    },
    {
      "epoch": 292.5,
      "grad_norm": 0.393625408411026,
      "learning_rate": 6.318550648618786e-05,
      "loss": 0.5349,
      "step": 585
    },
    {
      "epoch": 293.0,
      "grad_norm": 3.897812843322754,
      "learning_rate": 6.307685713906834e-05,
      "loss": 0.7975,
      "step": 586
    },
    {
      "epoch": 293.5,
      "grad_norm": 0.5139439105987549,
      "learning_rate": 6.296814147042047e-05,
      "loss": 0.6555,
      "step": 587
    },
    {
      "epoch": 294.0,
      "grad_norm": 1.1032121181488037,
      "learning_rate": 6.285936003161444e-05,
      "loss": 0.4405,
      "step": 588
    },
    {
      "epoch": 294.5,
      "grad_norm": 0.6604927182197571,
      "learning_rate": 6.275051337435401e-05,
      "loss": 0.5792,
      "step": 589
    },
    {
      "epoch": 295.0,
      "grad_norm": 1.6684532165527344,
      "learning_rate": 6.26416020506737e-05,
      "loss": 0.4126,
      "step": 590
    },
    {
      "epoch": 295.5,
      "grad_norm": 0.6102702021598816,
      "learning_rate": 6.253262661293604e-05,
      "loss": 0.5892,
      "step": 591
    },
    {
      "epoch": 296.0,
      "grad_norm": 1.3736975193023682,
      "learning_rate": 6.242358761382864e-05,
      "loss": 0.3391,
      "step": 592
    },
    {
      "epoch": 296.5,
      "grad_norm": 0.4834466874599457,
      "learning_rate": 6.231448560636156e-05,
      "loss": 0.5209,
      "step": 593
    },
    {
      "epoch": 297.0,
      "grad_norm": 1.2184211015701294,
      "learning_rate": 6.220532114386437e-05,
      "loss": 0.4106,
      "step": 594
    },
    {
      "epoch": 297.5,
      "grad_norm": 0.771481990814209,
      "learning_rate": 6.209609477998338e-05,
      "loss": 0.6153,
      "step": 595
    },
    {
      "epoch": 298.0,
      "grad_norm": 11.0581693649292,
      "learning_rate": 6.198680706867892e-05,
      "loss": 1.3428,
      "step": 596
    },
    {
      "epoch": 298.5,
      "grad_norm": 0.6222519278526306,
      "learning_rate": 6.187745856422235e-05,
      "loss": 0.4776,
      "step": 597
    },
    {
      "epoch": 299.0,
      "grad_norm": 1.2503429651260376,
      "learning_rate": 6.176804982119343e-05,
      "loss": 0.359,
      "step": 598
    },
    {
      "epoch": 299.5,
      "grad_norm": 0.6785809993743896,
      "learning_rate": 6.165858139447739e-05,
      "loss": 0.5179,
      "step": 599
    },
    {
      "epoch": 300.0,
      "grad_norm": 1.532062292098999,
      "learning_rate": 6.154905383926217e-05,
      "loss": 0.356,
      "step": 600
    },
    {
      "epoch": 300.5,
      "grad_norm": 0.5052584409713745,
      "learning_rate": 6.14394677110356e-05,
      "loss": 0.6105,
      "step": 601
    },
    {
      "epoch": 301.0,
      "grad_norm": 1.8669718503952026,
      "learning_rate": 6.132982356558255e-05,
      "loss": 0.7122,
      "step": 602
    },
    {
      "epoch": 301.5,
      "grad_norm": 0.362720787525177,
      "learning_rate": 6.122012195898215e-05,
      "loss": 0.5037,
      "step": 603
    },
    {
      "epoch": 302.0,
      "grad_norm": 1.5966863632202148,
      "learning_rate": 6.111036344760497e-05,
      "loss": 0.674,
      "step": 604
    },
    {
      "epoch": 302.5,
      "grad_norm": 0.4242487847805023,
      "learning_rate": 6.100054858811012e-05,
      "loss": 0.4651,
      "step": 605
    },
    {
      "epoch": 303.0,
      "grad_norm": 1.1737929582595825,
      "learning_rate": 6.0890677937442574e-05,
      "loss": 0.343,
      "step": 606
    },
    {
      "epoch": 303.5,
      "grad_norm": 6.984641075134277,
      "learning_rate": 6.078075205283021e-05,
      "loss": 0.9084,
      "step": 607
    },
    {
      "epoch": 304.0,
      "grad_norm": 1.0746909379959106,
      "learning_rate": 6.0670771491781043e-05,
      "loss": 0.3532,
      "step": 608
    },
    {
      "epoch": 304.5,
      "grad_norm": 1.016340970993042,
      "learning_rate": 6.056073681208038e-05,
      "loss": 0.8443,
      "step": 609
    },
    {
      "epoch": 305.0,
      "grad_norm": 0.7755418419837952,
      "learning_rate": 6.0450648571788004e-05,
      "loss": 0.3082,
      "step": 610
    },
    {
      "epoch": 305.5,
      "grad_norm": 0.7768415212631226,
      "learning_rate": 6.034050732923537e-05,
      "loss": 0.6455,
      "step": 611
    },
    {
      "epoch": 306.0,
      "grad_norm": 1.468597173690796,
      "learning_rate": 6.02303136430227e-05,
      "loss": 0.351,
      "step": 612
    },
    {
      "epoch": 306.5,
      "grad_norm": 0.7404545545578003,
      "learning_rate": 6.0120068072016225e-05,
      "loss": 0.6796,
      "step": 613
    },
    {
      "epoch": 307.0,
      "grad_norm": 1.812406063079834,
      "learning_rate": 6.000977117534526e-05,
      "loss": 0.999,
      "step": 614
    },
    {
      "epoch": 307.5,
      "grad_norm": 0.6493818759918213,
      "learning_rate": 5.9899423512399545e-05,
      "loss": 0.4837,
      "step": 615
    },
    {
      "epoch": 308.0,
      "grad_norm": 1.1872923374176025,
      "learning_rate": 5.978902564282616e-05,
      "loss": 0.8674,
      "step": 616
    },
    {
      "epoch": 308.5,
      "grad_norm": 6.459784984588623,
      "learning_rate": 5.967857812652691e-05,
      "loss": 0.4767,
      "step": 617
    },
    {
      "epoch": 309.0,
      "grad_norm": 4.909338474273682,
      "learning_rate": 5.9568081523655315e-05,
      "loss": 0.8649,
      "step": 618
    },
    {
      "epoch": 309.5,
      "grad_norm": 0.6570463180541992,
      "learning_rate": 5.945753639461393e-05,
      "loss": 0.6933,
      "step": 619
    },
    {
      "epoch": 310.0,
      "grad_norm": 1.4627389907836914,
      "learning_rate": 5.934694330005136e-05,
      "loss": 0.3858,
      "step": 620
    },
    {
      "epoch": 310.5,
      "grad_norm": 0.7542905807495117,
      "learning_rate": 5.923630280085948e-05,
      "loss": 0.5588,
      "step": 621
    },
    {
      "epoch": 311.0,
      "grad_norm": 1.3896498680114746,
      "learning_rate": 5.9125615458170593e-05,
      "loss": 0.648,
      "step": 622
    },
    {
      "epoch": 311.5,
      "grad_norm": 0.8111844062805176,
      "learning_rate": 5.9014881833354596e-05,
      "loss": 0.5504,
      "step": 623
    },
    {
      "epoch": 312.0,
      "grad_norm": 0.5012509822845459,
      "learning_rate": 5.890410248801608e-05,
      "loss": 0.3318,
      "step": 624
    },
    {
      "epoch": 312.5,
      "grad_norm": 0.5347142219543457,
      "learning_rate": 5.879327798399154e-05,
      "loss": 0.5042,
      "step": 625
    },
    {
      "epoch": 313.0,
      "grad_norm": 47.85480499267578,
      "learning_rate": 5.868240888334653e-05,
      "loss": 2.8712,
      "step": 626
    },
    {
      "epoch": 313.5,
      "grad_norm": 0.49037623405456543,
      "learning_rate": 5.8571495748372684e-05,
      "loss": 0.5675,
      "step": 627
    },
    {
      "epoch": 314.0,
      "grad_norm": 0.9824361801147461,
      "learning_rate": 5.846053914158507e-05,
      "loss": 0.3402,
      "step": 628
    },
    {
      "epoch": 314.5,
      "grad_norm": 0.5542188286781311,
      "learning_rate": 5.8349539625719184e-05,
      "loss": 0.6333,
      "step": 629
    },
    {
      "epoch": 315.0,
      "grad_norm": 0.8394861221313477,
      "learning_rate": 5.823849776372814e-05,
      "loss": 0.322,
      "step": 630
    },
    {
      "epoch": 315.5,
      "grad_norm": 1.123844861984253,
      "learning_rate": 5.812741411877982e-05,
      "loss": 0.7907,
      "step": 631
    },
    {
      "epoch": 316.0,
      "grad_norm": 1.0530474185943604,
      "learning_rate": 5.801628925425403e-05,
      "loss": 0.3698,
      "step": 632
    },
    {
      "epoch": 316.5,
      "grad_norm": 0.5948439240455627,
      "learning_rate": 5.790512373373962e-05,
      "loss": 0.5343,
      "step": 633
    },
    {
      "epoch": 317.0,
      "grad_norm": 1.4498809576034546,
      "learning_rate": 5.779391812103162e-05,
      "loss": 0.4547,
      "step": 634
    },
    {
      "epoch": 317.5,
      "grad_norm": 3.35131573677063,
      "learning_rate": 5.7682672980128406e-05,
      "loss": 0.6478,
      "step": 635
    },
    {
      "epoch": 318.0,
      "grad_norm": 1.3095659017562866,
      "learning_rate": 5.757138887522884e-05,
      "loss": 0.4043,
      "step": 636
    },
    {
      "epoch": 318.5,
      "grad_norm": 0.49627795815467834,
      "learning_rate": 5.746006637072938e-05,
      "loss": 0.5414,
      "step": 637
    },
    {
      "epoch": 319.0,
      "grad_norm": 0.7660489082336426,
      "learning_rate": 5.7348706031221247e-05,
      "loss": 1.0857,
      "step": 638
    },
    {
      "epoch": 319.5,
      "grad_norm": 1.3646880388259888,
      "learning_rate": 5.723730842148752e-05,
      "loss": 0.6487,
      "step": 639
    },
    {
      "epoch": 320.0,
      "grad_norm": 3.388394594192505,
      "learning_rate": 5.712587410650034e-05,
      "loss": 0.5239,
      "step": 640
    },
    {
      "epoch": 320.5,
      "grad_norm": 1.2308599948883057,
      "learning_rate": 5.701440365141799e-05,
      "loss": 0.7771,
      "step": 641
    },
    {
      "epoch": 321.0,
      "grad_norm": 1.5851600170135498,
      "learning_rate": 5.690289762158203e-05,
      "loss": 0.4139,
      "step": 642
    },
    {
      "epoch": 321.5,
      "grad_norm": 1.0507495403289795,
      "learning_rate": 5.6791356582514444e-05,
      "loss": 0.5564,
      "step": 643
    },
    {
      "epoch": 322.0,
      "grad_norm": 1.7389483451843262,
      "learning_rate": 5.667978109991481e-05,
      "loss": 1.0529,
      "step": 644
    },
    {
      "epoch": 322.5,
      "grad_norm": 2.415520191192627,
      "learning_rate": 5.656817173965733e-05,
      "loss": 0.6332,
      "step": 645
    },
    {
      "epoch": 323.0,
      "grad_norm": 2.2303061485290527,
      "learning_rate": 5.645652906778808e-05,
      "loss": 0.3998,
      "step": 646
    },
    {
      "epoch": 323.5,
      "grad_norm": 1.617475152015686,
      "learning_rate": 5.6344853650522043e-05,
      "loss": 0.7289,
      "step": 647
    },
    {
      "epoch": 324.0,
      "grad_norm": 0.7127124667167664,
      "learning_rate": 5.6233146054240305e-05,
      "loss": 1.0924,
      "step": 648
    },
    {
      "epoch": 324.5,
      "grad_norm": 0.5544408559799194,
      "learning_rate": 5.612140684548712e-05,
      "loss": 0.5652,
      "step": 649
    },
    {
      "epoch": 325.0,
      "grad_norm": 1.1636828184127808,
      "learning_rate": 5.6009636590967106e-05,
      "loss": 1.0842,
      "step": 650
    },
    {
      "epoch": 325.5,
      "grad_norm": 1.549724817276001,
      "learning_rate": 5.5897835857542317e-05,
      "loss": 0.6118,
      "step": 651
    },
    {
      "epoch": 326.0,
      "grad_norm": 1.182888388633728,
      "learning_rate": 5.578600521222936e-05,
      "loss": 0.356,
      "step": 652
    },
    {
      "epoch": 326.5,
      "grad_norm": 0.5167127251625061,
      "learning_rate": 5.5674145222196596e-05,
      "loss": 0.4882,
      "step": 653
    },
    {
      "epoch": 327.0,
      "grad_norm": 1.7193734645843506,
      "learning_rate": 5.556225645476118e-05,
      "loss": 0.407,
      "step": 654
    },
    {
      "epoch": 327.5,
      "grad_norm": 2.6632983684539795,
      "learning_rate": 5.545033947738624e-05,
      "loss": 0.8979,
      "step": 655
    },
    {
      "epoch": 328.0,
      "grad_norm": 3.879645347595215,
      "learning_rate": 5.5338394857677945e-05,
      "loss": 0.7762,
      "step": 656
    },
    {
      "epoch": 328.5,
      "grad_norm": 1.3633205890655518,
      "learning_rate": 5.522642316338268e-05,
      "loss": 0.6186,
      "step": 657
    },
    {
      "epoch": 329.0,
      "grad_norm": 1.3031944036483765,
      "learning_rate": 5.5114424962384124e-05,
      "loss": 0.3738,
      "step": 658
    },
    {
      "epoch": 329.5,
      "grad_norm": 0.8823007345199585,
      "learning_rate": 5.5002400822700415e-05,
      "loss": 0.4081,
      "step": 659
    },
    {
      "epoch": 330.0,
      "grad_norm": 1.5013563632965088,
      "learning_rate": 5.489035131248123e-05,
      "loss": 0.597,
      "step": 660
    },
    {
      "epoch": 330.5,
      "grad_norm": 2.054853677749634,
      "learning_rate": 5.477827700000492e-05,
      "loss": 0.6673,
      "step": 661
    },
    {
      "epoch": 331.0,
      "grad_norm": 1.4972254037857056,
      "learning_rate": 5.46661784536756e-05,
      "loss": 0.7105,
      "step": 662
    },
    {
      "epoch": 331.5,
      "grad_norm": 0.5177363157272339,
      "learning_rate": 5.4554056242020325e-05,
      "loss": 0.4252,
      "step": 663
    },
    {
      "epoch": 332.0,
      "grad_norm": 2.8590264320373535,
      "learning_rate": 5.444191093368616e-05,
      "loss": 0.7269,
      "step": 664
    },
    {
      "epoch": 332.5,
      "grad_norm": 2.1307737827301025,
      "learning_rate": 5.432974309743731e-05,
      "loss": 0.6633,
      "step": 665
    },
    {
      "epoch": 333.0,
      "grad_norm": 2.0082290172576904,
      "learning_rate": 5.4217553302152237e-05,
      "loss": 0.4807,
      "step": 666
    },
    {
      "epoch": 333.5,
      "grad_norm": 1.205055832862854,
      "learning_rate": 5.410534211682078e-05,
      "loss": 0.5974,
      "step": 667
    },
    {
      "epoch": 334.0,
      "grad_norm": 2.481665849685669,
      "learning_rate": 5.3993110110541226e-05,
      "loss": 0.5644,
      "step": 668
    },
    {
      "epoch": 334.5,
      "grad_norm": 0.6096961498260498,
      "learning_rate": 5.3880857852517496e-05,
      "loss": 0.4628,
      "step": 669
    },
    {
      "epoch": 335.0,
      "grad_norm": 1.256682276725769,
      "learning_rate": 5.3768585912056215e-05,
      "loss": 0.3198,
      "step": 670
    },
    {
      "epoch": 335.5,
      "grad_norm": 0.51405930519104,
      "learning_rate": 5.3656294858563803e-05,
      "loss": 0.5181,
      "step": 671
    },
    {
      "epoch": 336.0,
      "grad_norm": 0.7644615173339844,
      "learning_rate": 5.354398526154365e-05,
      "loss": 0.322,
      "step": 672
    },
    {
      "epoch": 336.5,
      "grad_norm": 3.858107566833496,
      "learning_rate": 5.3431657690593187e-05,
      "loss": 0.472,
      "step": 673
    },
    {
      "epoch": 337.0,
      "grad_norm": 1.5808908939361572,
      "learning_rate": 5.331931271540096e-05,
      "loss": 0.6868,
      "step": 674
    },
    {
      "epoch": 337.5,
      "grad_norm": 0.8845391869544983,
      "learning_rate": 5.320695090574386e-05,
      "loss": 0.4245,
      "step": 675
    },
    {
      "epoch": 338.0,
      "grad_norm": 2.135876417160034,
      "learning_rate": 5.30945728314841e-05,
      "loss": 0.3822,
      "step": 676
    },
    {
      "epoch": 338.5,
      "grad_norm": 0.745249330997467,
      "learning_rate": 5.2982179062566385e-05,
      "loss": 0.5694,
      "step": 677
    },
    {
      "epoch": 339.0,
      "grad_norm": 1.2785981893539429,
      "learning_rate": 5.2869770169015034e-05,
      "loss": 0.5969,
      "step": 678
    },
    {
      "epoch": 339.5,
      "grad_norm": 1.5919989347457886,
      "learning_rate": 5.275734672093109e-05,
      "loss": 0.5957,
      "step": 679
    },
    {
      "epoch": 340.0,
      "grad_norm": 1.1622087955474854,
      "learning_rate": 5.2644909288489365e-05,
      "loss": 0.3329,
      "step": 680
    },
    {
      "epoch": 340.5,
      "grad_norm": 1.0013360977172852,
      "learning_rate": 5.2532458441935636e-05,
      "loss": 0.5313,
      "step": 681
    },
    {
      "epoch": 341.0,
      "grad_norm": 1.1525441408157349,
      "learning_rate": 5.24199947515837e-05,
      "loss": 0.4165,
      "step": 682
    },
    {
      "epoch": 341.5,
      "grad_norm": 1.0278726816177368,
      "learning_rate": 5.230751878781247e-05,
      "loss": 0.6566,
      "step": 683
    },
    {
      "epoch": 342.0,
      "grad_norm": 1.1601136922836304,
      "learning_rate": 5.219503112106314e-05,
      "loss": 0.3285,
      "step": 684
    },
    {
      "epoch": 342.5,
      "grad_norm": 1.7861090898513794,
      "learning_rate": 5.2082532321836256e-05,
      "loss": 0.7569,
      "step": 685
    },
    {
      "epoch": 343.0,
      "grad_norm": 0.9920994639396667,
      "learning_rate": 5.197002296068878e-05,
      "loss": 0.31,
      "step": 686
    },
    {
      "epoch": 343.5,
      "grad_norm": 0.8111425042152405,
      "learning_rate": 5.1857503608231304e-05,
      "loss": 0.5063,
      "step": 687
    },
    {
      "epoch": 344.0,
      "grad_norm": 2.2094032764434814,
      "learning_rate": 5.174497483512506e-05,
      "loss": 0.4633,
      "step": 688
    },
    {
      "epoch": 344.5,
      "grad_norm": 0.6735397577285767,
      "learning_rate": 5.163243721207903e-05,
      "loss": 0.5271,
      "step": 689
    },
    {
      "epoch": 345.0,
      "grad_norm": 1.682518482208252,
      "learning_rate": 5.151989130984715e-05,
      "loss": 0.3251,
      "step": 690
    },
    {
      "epoch": 345.5,
      "grad_norm": 0.5200764536857605,
      "learning_rate": 5.140733769922526e-05,
      "loss": 0.573,
      "step": 691
    },
    {
      "epoch": 346.0,
      "grad_norm": 1.7628211975097656,
      "learning_rate": 5.1294776951048386e-05,
      "loss": 0.5762,
      "step": 692
    },
    {
      "epoch": 346.5,
      "grad_norm": 0.6817848682403564,
      "learning_rate": 5.1182209636187674e-05,
      "loss": 0.4942,
      "step": 693
    },
    {
      "epoch": 347.0,
      "grad_norm": 0.7215993404388428,
      "learning_rate": 5.106963632554762e-05,
      "loss": 1.1083,
      "step": 694
    },
    {
      "epoch": 347.5,
      "grad_norm": 0.4209991991519928,
      "learning_rate": 5.09570575900631e-05,
      "loss": 0.5077,
      "step": 695
    },
    {
      "epoch": 348.0,
      "grad_norm": 1.8245824575424194,
      "learning_rate": 5.084447400069655e-05,
      "loss": 0.4267,
      "step": 696
    },
    {
      "epoch": 348.5,
      "grad_norm": 4.792304039001465,
      "learning_rate": 5.073188612843498e-05,
      "loss": 0.6107,
      "step": 697
    },
    {
      "epoch": 349.0,
      "grad_norm": 0.6443562507629395,
      "learning_rate": 5.0619294544287124e-05,
      "loss": 0.3326,
      "step": 698
    },
    {
      "epoch": 349.5,
      "grad_norm": 2.2389540672302246,
      "learning_rate": 5.0506699819280556e-05,
      "loss": 0.5906,
      "step": 699
    },
    {
      "epoch": 350.0,
      "grad_norm": 0.563458263874054,
      "learning_rate": 5.039410252445878e-05,
      "loss": 0.276,
      "step": 700
    },
    {
      "epoch": 350.5,
      "grad_norm": 1.1636115312576294,
      "learning_rate": 5.02815032308783e-05,
      "loss": 0.6247,
      "step": 701
    },
    {
      "epoch": 351.0,
      "grad_norm": 1.2161306142807007,
      "learning_rate": 5.016890250960582e-05,
      "loss": 1.0409,
      "step": 702
    },
    {
      "epoch": 351.5,
      "grad_norm": 0.47416549921035767,
      "learning_rate": 5.0056300931715226e-05,
      "loss": 0.527,
      "step": 703
    },
    {
      "epoch": 352.0,
      "grad_norm": 0.7766896486282349,
      "learning_rate": 4.9943699068284786e-05,
      "loss": 0.3566,
      "step": 704
    },
    {
      "epoch": 352.5,
      "grad_norm": 1.030655026435852,
      "learning_rate": 4.9831097490394194e-05,
      "loss": 0.4807,
      "step": 705
    },
    {
      "epoch": 353.0,
      "grad_norm": 1.4332993030548096,
      "learning_rate": 4.971849676912172e-05,
      "loss": 0.5652,
      "step": 706
    },
    {
      "epoch": 353.5,
      "grad_norm": 0.436355322599411,
      "learning_rate": 4.960589747554124e-05,
      "loss": 0.4786,
      "step": 707
    },
    {
      "epoch": 354.0,
      "grad_norm": 0.8764670491218567,
      "learning_rate": 4.949330018071946e-05,
      "loss": 0.3349,
      "step": 708
    },
    {
      "epoch": 354.5,
      "grad_norm": 0.47430935502052307,
      "learning_rate": 4.938070545571289e-05,
      "loss": 0.5645,
      "step": 709
    },
    {
      "epoch": 355.0,
      "grad_norm": 1.080704927444458,
      "learning_rate": 4.926811387156502e-05,
      "loss": 0.4866,
      "step": 710
    },
    {
      "epoch": 355.5,
      "grad_norm": 0.43620017170906067,
      "learning_rate": 4.915552599930345e-05,
      "loss": 0.4525,
      "step": 711
    },
    {
      "epoch": 356.0,
      "grad_norm": 0.8112178444862366,
      "learning_rate": 4.9042942409936895e-05,
      "loss": 0.4016,
      "step": 712
    },
    {
      "epoch": 356.5,
      "grad_norm": 0.8193455338478088,
      "learning_rate": 4.893036367445239e-05,
      "loss": 0.4762,
      "step": 713
    },
    {
      "epoch": 357.0,
      "grad_norm": 2.1425654888153076,
      "learning_rate": 4.881779036381234e-05,
      "loss": 0.8261,
      "step": 714
    },
    {
      "epoch": 357.5,
      "grad_norm": 1.8430966138839722,
      "learning_rate": 4.870522304895163e-05,
      "loss": 0.5222,
      "step": 715
    },
    {
      "epoch": 358.0,
      "grad_norm": 1.4282432794570923,
      "learning_rate": 4.859266230077474e-05,
      "loss": 0.3847,
      "step": 716
    },
    {
      "epoch": 358.5,
      "grad_norm": 0.5180120468139648,
      "learning_rate": 4.8480108690152876e-05,
      "loss": 0.4603,
      "step": 717
    },
    {
      "epoch": 359.0,
      "grad_norm": 0.5665010809898376,
      "learning_rate": 4.8367562787920975e-05,
      "loss": 0.3111,
      "step": 718
    },
    {
      "epoch": 359.5,
      "grad_norm": 0.6947895288467407,
      "learning_rate": 4.825502516487497e-05,
      "loss": 0.5334,
      "step": 719
    },
    {
      "epoch": 360.0,
      "grad_norm": 0.6546884775161743,
      "learning_rate": 4.81424963917687e-05,
      "loss": 0.3005,
      "step": 720
    },
    {
      "epoch": 360.5,
      "grad_norm": 0.36885982751846313,
      "learning_rate": 4.8029977039311236e-05,
      "loss": 0.6127,
      "step": 721
    },
    {
      "epoch": 361.0,
      "grad_norm": 0.9827126860618591,
      "learning_rate": 4.791746767816376e-05,
      "loss": 0.3043,
      "step": 722
    },
    {
      "epoch": 361.5,
      "grad_norm": 0.5609752535820007,
      "learning_rate": 4.780496887893686e-05,
      "loss": 0.4179,
      "step": 723
    },
    {
      "epoch": 362.0,
      "grad_norm": 1.5731890201568604,
      "learning_rate": 4.769248121218753e-05,
      "loss": 0.5522,
      "step": 724
    },
    {
      "epoch": 362.5,
      "grad_norm": 0.499330073595047,
      "learning_rate": 4.7580005248416314e-05,
      "loss": 0.4708,
      "step": 725
    },
    {
      "epoch": 363.0,
      "grad_norm": 1.821414589881897,
      "learning_rate": 4.746754155806437e-05,
      "loss": 0.7208,
      "step": 726
    },
    {
      "epoch": 363.5,
      "grad_norm": 0.4948866069316864,
      "learning_rate": 4.735509071151063e-05,
      "loss": 0.5543,
      "step": 727
    },
    {
      "epoch": 364.0,
      "grad_norm": 1.543725848197937,
      "learning_rate": 4.724265327906892e-05,
      "loss": 0.9646,
      "step": 728
    },
    {
      "epoch": 364.5,
      "grad_norm": 1.1979113817214966,
      "learning_rate": 4.7130229830984964e-05,
      "loss": 0.7539,
      "step": 729
    },
    {
      "epoch": 365.0,
      "grad_norm": 1.1699392795562744,
      "learning_rate": 4.701782093743363e-05,
      "loss": 1.044,
      "step": 730
    },
    {
      "epoch": 365.5,
      "grad_norm": 0.6509461998939514,
      "learning_rate": 4.690542716851591e-05,
      "loss": 0.5045,
      "step": 731
    },
    {
      "epoch": 366.0,
      "grad_norm": 27.287565231323242,
      "learning_rate": 4.6793049094256144e-05,
      "loss": 3.3403,
      "step": 732
    },
    {
      "epoch": 366.5,
      "grad_norm": 0.9202408790588379,
      "learning_rate": 4.668068728459905e-05,
      "loss": 0.484,
      "step": 733
    },
    {
      "epoch": 367.0,
      "grad_norm": 1.6131242513656616,
      "learning_rate": 4.656834230940684e-05,
      "loss": 0.6677,
      "step": 734
    },
    {
      "epoch": 367.5,
      "grad_norm": 0.9662175178527832,
      "learning_rate": 4.6456014738456356e-05,
      "loss": 0.5157,
      "step": 735
    },
    {
      "epoch": 368.0,
      "grad_norm": 0.7148228883743286,
      "learning_rate": 4.6343705141436195e-05,
      "loss": 0.35,
      "step": 736
    },
    {
      "epoch": 368.5,
      "grad_norm": 0.6274757981300354,
      "learning_rate": 4.6231414087943796e-05,
      "loss": 0.5181,
      "step": 737
    },
    {
      "epoch": 369.0,
      "grad_norm": 0.6102350950241089,
      "learning_rate": 4.6119142147482495e-05,
      "loss": 0.3305,
      "step": 738
    },
    {
      "epoch": 369.5,
      "grad_norm": 0.4765862226486206,
      "learning_rate": 4.600688988945878e-05,
      "loss": 0.5495,
      "step": 739
    },
    {
      "epoch": 370.0,
      "grad_norm": 1.4294583797454834,
      "learning_rate": 4.589465788317922e-05,
      "loss": 0.331,
      "step": 740
    },
    {
      "epoch": 370.5,
      "grad_norm": 0.3986988067626953,
      "learning_rate": 4.578244669784777e-05,
      "loss": 0.4844,
      "step": 741
    },
    {
      "epoch": 371.0,
      "grad_norm": 1.7126579284667969,
      "learning_rate": 4.5670256902562694e-05,
      "loss": 0.7701,
      "step": 742
    },
    {
      "epoch": 371.5,
      "grad_norm": 0.7031809687614441,
      "learning_rate": 4.5558089066313854e-05,
      "loss": 0.3879,
      "step": 743
    },
    {
      "epoch": 372.0,
      "grad_norm": 0.8251434564590454,
      "learning_rate": 4.544594375797968e-05,
      "loss": 0.3357,
      "step": 744
    },
    {
      "epoch": 372.5,
      "grad_norm": 1.3418869972229004,
      "learning_rate": 4.533382154632442e-05,
      "loss": 0.5457,
      "step": 745
    },
    {
      "epoch": 373.0,
      "grad_norm": 1.2738980054855347,
      "learning_rate": 4.52217229999951e-05,
      "loss": 0.5005,
      "step": 746
    },
    {
      "epoch": 373.5,
      "grad_norm": 2.683478593826294,
      "learning_rate": 4.5109648687518785e-05,
      "loss": 0.6966,
      "step": 747
    },
    {
      "epoch": 374.0,
      "grad_norm": 1.8470619916915894,
      "learning_rate": 4.499759917729959e-05,
      "loss": 0.5483,
      "step": 748
    },
    {
      "epoch": 374.5,
      "grad_norm": 0.5200942754745483,
      "learning_rate": 4.4885575037615874e-05,
      "loss": 0.5303,
      "step": 749
    },
    {
      "epoch": 375.0,
      "grad_norm": 1.564513921737671,
      "learning_rate": 4.477357683661734e-05,
      "loss": 0.3621,
      "step": 750
    },
    {
      "epoch": 375.5,
      "grad_norm": 1.4996962547302246,
      "learning_rate": 4.466160514232206e-05,
      "loss": 0.721,
      "step": 751
    },
    {
      "epoch": 376.0,
      "grad_norm": 0.959261417388916,
      "learning_rate": 4.454966052261378e-05,
      "loss": 0.2735,
      "step": 752
    },
    {
      "epoch": 376.5,
      "grad_norm": 0.6466561555862427,
      "learning_rate": 4.443774354523882e-05,
      "loss": 0.5933,
      "step": 753
    },
    {
      "epoch": 377.0,
      "grad_norm": 0.6928865313529968,
      "learning_rate": 4.4325854777803416e-05,
      "loss": 0.3201,
      "step": 754
    },
    {
      "epoch": 377.5,
      "grad_norm": 1.5327074527740479,
      "learning_rate": 4.4213994787770644e-05,
      "loss": 0.4991,
      "step": 755
    },
    {
      "epoch": 378.0,
      "grad_norm": 0.8794979453086853,
      "learning_rate": 4.410216414245771e-05,
      "loss": 0.3057,
      "step": 756
    },
    {
      "epoch": 378.5,
      "grad_norm": 0.948381781578064,
      "learning_rate": 4.39903634090329e-05,
      "loss": 0.3979,
      "step": 757
    },
    {
      "epoch": 379.0,
      "grad_norm": 0.695127546787262,
      "learning_rate": 4.3878593154512895e-05,
      "loss": 0.3253,
      "step": 758
    },
    {
      "epoch": 379.5,
      "grad_norm": 0.5920587778091431,
      "learning_rate": 4.376685394575971e-05,
      "loss": 0.429,
      "step": 759
    },
    {
      "epoch": 380.0,
      "grad_norm": 2.077747106552124,
      "learning_rate": 4.3655146349477975e-05,
      "loss": 0.5833,
      "step": 760
    },
    {
      "epoch": 380.5,
      "grad_norm": 0.5286732912063599,
      "learning_rate": 4.354347093221194e-05,
      "loss": 0.4387,
      "step": 761
    },
    {
      "epoch": 381.0,
      "grad_norm": 2.106220006942749,
      "learning_rate": 4.3431828260342674e-05,
      "loss": 0.5925,
      "step": 762
    },
    {
      "epoch": 381.5,
      "grad_norm": 0.671046793460846,
      "learning_rate": 4.3320218900085204e-05,
      "loss": 0.4904,
      "step": 763
    },
    {
      "epoch": 382.0,
      "grad_norm": 1.2230089902877808,
      "learning_rate": 4.320864341748556e-05,
      "loss": 0.4323,
      "step": 764
    },
    {
      "epoch": 382.5,
      "grad_norm": 0.9491264820098877,
      "learning_rate": 4.309710237841798e-05,
      "loss": 0.6952,
      "step": 765
    },
    {
      "epoch": 383.0,
      "grad_norm": 0.7745018601417542,
      "learning_rate": 4.2985596348582016e-05,
      "loss": 0.3556,
      "step": 766
    },
    {
      "epoch": 383.5,
      "grad_norm": 1.336092233657837,
      "learning_rate": 4.2874125893499676e-05,
      "loss": 0.5156,
      "step": 767
    },
    {
      "epoch": 384.0,
      "grad_norm": 0.7443627715110779,
      "learning_rate": 4.276269157851249e-05,
      "loss": 0.2786,
      "step": 768
    },
    {
      "epoch": 384.5,
      "grad_norm": 0.6343594789505005,
      "learning_rate": 4.265129396877878e-05,
      "loss": 0.4851,
      "step": 769
    },
    {
      "epoch": 385.0,
      "grad_norm": 1.935417890548706,
      "learning_rate": 4.253993362927063e-05,
      "loss": 0.5012,
      "step": 770
    },
    {
      "epoch": 385.5,
      "grad_norm": 1.0597429275512695,
      "learning_rate": 4.2428611124771184e-05,
      "loss": 0.5131,
      "step": 771
    },
    {
      "epoch": 386.0,
      "grad_norm": 1.3614938259124756,
      "learning_rate": 4.231732701987161e-05,
      "loss": 0.4643,
      "step": 772
    },
    {
      "epoch": 386.5,
      "grad_norm": 1.5061779022216797,
      "learning_rate": 4.2206081878968404e-05,
      "loss": 0.703,
      "step": 773
    },
    {
      "epoch": 387.0,
      "grad_norm": 0.8944714665412903,
      "learning_rate": 4.209487626626039e-05,
      "loss": 0.3434,
      "step": 774
    },
    {
      "epoch": 387.5,
      "grad_norm": 0.5271845459938049,
      "learning_rate": 4.1983710745745964e-05,
      "loss": 0.4324,
      "step": 775
    },
    {
      "epoch": 388.0,
      "grad_norm": 1.3463325500488281,
      "learning_rate": 4.187258588122019e-05,
      "loss": 0.5492,
      "step": 776
    },
    {
      "epoch": 388.5,
      "grad_norm": 1.2681238651275635,
      "learning_rate": 4.176150223627186e-05,
      "loss": 0.5158,
      "step": 777
    },
    {
      "epoch": 389.0,
      "grad_norm": 0.9463877081871033,
      "learning_rate": 4.165046037428083e-05,
      "loss": 0.4036,
      "step": 778
    },
    {
      "epoch": 389.5,
      "grad_norm": 0.4172000288963318,
      "learning_rate": 4.1539460858414924e-05,
      "loss": 0.5094,
      "step": 779
    },
    {
      "epoch": 390.0,
      "grad_norm": 0.9338898658752441,
      "learning_rate": 4.1428504251627335e-05,
      "loss": 0.3072,
      "step": 780
    },
    {
      "epoch": 390.5,
      "grad_norm": 2.133432149887085,
      "learning_rate": 4.131759111665349e-05,
      "loss": 0.6543,
      "step": 781
    },
    {
      "epoch": 391.0,
      "grad_norm": 1.6524803638458252,
      "learning_rate": 4.120672201600846e-05,
      "loss": 0.3051,
      "step": 782
    },
    {
      "epoch": 391.5,
      "grad_norm": 0.9547257423400879,
      "learning_rate": 4.1095897511983934e-05,
      "loss": 0.4665,
      "step": 783
    },
    {
      "epoch": 392.0,
      "grad_norm": 0.6648340225219727,
      "learning_rate": 4.098511816664542e-05,
      "loss": 0.2947,
      "step": 784
    },
    {
      "epoch": 392.5,
      "grad_norm": 1.129219651222229,
      "learning_rate": 4.087438454182942e-05,
      "loss": 0.49,
      "step": 785
    },
    {
      "epoch": 393.0,
      "grad_norm": 0.805383563041687,
      "learning_rate": 4.076369719914055e-05,
      "loss": 0.3174,
      "step": 786
    },
    {
      "epoch": 393.5,
      "grad_norm": 0.39121708273887634,
      "learning_rate": 4.0653056699948656e-05,
      "loss": 0.4699,
      "step": 787
    },
    {
      "epoch": 394.0,
      "grad_norm": 3.5439791679382324,
      "learning_rate": 4.054246360538607e-05,
      "loss": 0.5919,
      "step": 788
    },
    {
      "epoch": 394.5,
      "grad_norm": 1.417879343032837,
      "learning_rate": 4.043191847634469e-05,
      "loss": 0.4958,
      "step": 789
    },
    {
      "epoch": 395.0,
      "grad_norm": 0.8236833810806274,
      "learning_rate": 4.0321421873473095e-05,
      "loss": 0.2945,
      "step": 790
    },
    {
      "epoch": 395.5,
      "grad_norm": 0.9563655257225037,
      "learning_rate": 4.021097435717385e-05,
      "loss": 0.4518,
      "step": 791
    },
    {
      "epoch": 396.0,
      "grad_norm": 0.9118824005126953,
      "learning_rate": 4.010057648760047e-05,
      "loss": 0.3141,
      "step": 792
    },
    {
      "epoch": 396.5,
      "grad_norm": 0.5774078965187073,
      "learning_rate": 3.999022882465474e-05,
      "loss": 0.5431,
      "step": 793
    },
    {
      "epoch": 397.0,
      "grad_norm": 8.662175178527832,
      "learning_rate": 3.9879931927983793e-05,
      "loss": 0.9706,
      "step": 794
    },
    {
      "epoch": 397.5,
      "grad_norm": 0.6393212676048279,
      "learning_rate": 3.976968635697732e-05,
      "loss": 0.5056,
      "step": 795
    },
    {
      "epoch": 398.0,
      "grad_norm": 1.1805341243743896,
      "learning_rate": 3.965949267076464e-05,
      "loss": 0.3166,
      "step": 796
    },
    {
      "epoch": 398.5,
      "grad_norm": 1.7247583866119385,
      "learning_rate": 3.9549351428212015e-05,
      "loss": 0.5934,
      "step": 797
    },
    {
      "epoch": 399.0,
      "grad_norm": 1.286820411682129,
      "learning_rate": 3.943926318791963e-05,
      "loss": 1.0755,
      "step": 798
    },
    {
      "epoch": 399.5,
      "grad_norm": 0.8619692921638489,
      "learning_rate": 3.932922850821898e-05,
      "loss": 0.3948,
      "step": 799
    },
    {
      "epoch": 400.0,
      "grad_norm": 1.0435948371887207,
      "learning_rate": 3.92192479471698e-05,
      "loss": 0.3603,
      "step": 800
    },
    {
      "epoch": 400.5,
      "grad_norm": 0.5307633280754089,
      "learning_rate": 3.9109322062557424e-05,
      "loss": 0.4845,
      "step": 801
    },
    {
      "epoch": 401.0,
      "grad_norm": 1.0714030265808105,
      "learning_rate": 3.8999451411889894e-05,
      "loss": 0.2792,
      "step": 802
    },
    {
      "epoch": 401.5,
      "grad_norm": 0.5888768434524536,
      "learning_rate": 3.888963655239505e-05,
      "loss": 0.5502,
      "step": 803
    },
    {
      "epoch": 402.0,
      "grad_norm": 1.147822618484497,
      "learning_rate": 3.877987804101786e-05,
      "loss": 0.3137,
      "step": 804
    },
    {
      "epoch": 402.5,
      "grad_norm": 3.8641574382781982,
      "learning_rate": 3.867017643441745e-05,
      "loss": 0.5796,
      "step": 805
    },
    {
      "epoch": 403.0,
      "grad_norm": 0.8890193104743958,
      "learning_rate": 3.856053228896442e-05,
      "loss": 0.3258,
      "step": 806
    },
    {
      "epoch": 403.5,
      "grad_norm": 0.6431386470794678,
      "learning_rate": 3.845094616073783e-05,
      "loss": 0.4491,
      "step": 807
    },
    {
      "epoch": 404.0,
      "grad_norm": 1.1839590072631836,
      "learning_rate": 3.8341418605522625e-05,
      "loss": 0.8286,
      "step": 808
    },
    {
      "epoch": 404.5,
      "grad_norm": 1.9728081226348877,
      "learning_rate": 3.823195017880658e-05,
      "loss": 0.5337,
      "step": 809
    },
    {
      "epoch": 405.0,
      "grad_norm": 1.244590163230896,
      "learning_rate": 3.812254143577767e-05,
      "loss": 0.9212,
      "step": 810
    },
    {
      "epoch": 405.5,
      "grad_norm": 0.4246683716773987,
      "learning_rate": 3.8013192931321096e-05,
      "loss": 0.4703,
      "step": 811
    },
    {
      "epoch": 406.0,
      "grad_norm": 1.8208593130111694,
      "learning_rate": 3.790390522001662e-05,
      "loss": 0.3155,
      "step": 812
    },
    {
      "epoch": 406.5,
      "grad_norm": 1.1932415962219238,
      "learning_rate": 3.7794678856135643e-05,
      "loss": 0.5103,
      "step": 813
    },
    {
      "epoch": 407.0,
      "grad_norm": 5.409270763397217,
      "learning_rate": 3.768551439363845e-05,
      "loss": 0.4284,
      "step": 814
    },
    {
      "epoch": 407.5,
      "grad_norm": 1.1557362079620361,
      "learning_rate": 3.757641238617137e-05,
      "loss": 0.5291,
      "step": 815
    },
    {
      "epoch": 408.0,
      "grad_norm": 1.3448107242584229,
      "learning_rate": 3.746737338706397e-05,
      "loss": 0.3839,
      "step": 816
    },
    {
      "epoch": 408.5,
      "grad_norm": 0.7283592820167542,
      "learning_rate": 3.735839794932631e-05,
      "loss": 0.4393,
      "step": 817
    },
    {
      "epoch": 409.0,
      "grad_norm": 0.9953801035881042,
      "learning_rate": 3.724948662564599e-05,
      "loss": 0.3655,
      "step": 818
    },
    {
      "epoch": 409.5,
      "grad_norm": 4.571325302124023,
      "learning_rate": 3.714063996838558e-05,
      "loss": 0.6867,
      "step": 819
    },
    {
      "epoch": 410.0,
      "grad_norm": 1.234880805015564,
      "learning_rate": 3.703185852957953e-05,
      "loss": 0.3767,
      "step": 820
    },
    {
      "epoch": 410.5,
      "grad_norm": 0.7834144830703735,
      "learning_rate": 3.692314286093167e-05,
      "loss": 0.3751,
      "step": 821
    },
    {
      "epoch": 411.0,
      "grad_norm": 0.9509165287017822,
      "learning_rate": 3.681449351381216e-05,
      "loss": 0.3874,
      "step": 822
    },
    {
      "epoch": 411.5,
      "grad_norm": 0.6251256465911865,
      "learning_rate": 3.670591103925485e-05,
      "loss": 0.4081,
      "step": 823
    },
    {
      "epoch": 412.0,
      "grad_norm": 1.1587973833084106,
      "learning_rate": 3.65973959879544e-05,
      "loss": 0.2894,
      "step": 824
    },
    {
      "epoch": 412.5,
      "grad_norm": 1.3861202001571655,
      "learning_rate": 3.6488948910263584e-05,
      "loss": 0.552,
      "step": 825
    },
    {
      "epoch": 413.0,
      "grad_norm": 1.6265616416931152,
      "learning_rate": 3.6380570356190344e-05,
      "loss": 0.419,
      "step": 826
    },
    {
      "epoch": 413.5,
      "grad_norm": 0.6835845708847046,
      "learning_rate": 3.627226087539517e-05,
      "loss": 0.4641,
      "step": 827
    },
    {
      "epoch": 414.0,
      "grad_norm": 2.4812233448028564,
      "learning_rate": 3.616402101718822e-05,
      "loss": 0.3631,
      "step": 828
    },
    {
      "epoch": 414.5,
      "grad_norm": 1.2474125623703003,
      "learning_rate": 3.60558513305265e-05,
      "loss": 0.5602,
      "step": 829
    },
    {
      "epoch": 415.0,
      "grad_norm": 1.2531253099441528,
      "learning_rate": 3.594775236401122e-05,
      "loss": 0.3266,
      "step": 830
    },
    {
      "epoch": 415.5,
      "grad_norm": 0.6381431221961975,
      "learning_rate": 3.58397246658848e-05,
      "loss": 0.4453,
      "step": 831
    },
    {
      "epoch": 416.0,
      "grad_norm": 0.965083658695221,
      "learning_rate": 3.573176878402833e-05,
      "loss": 0.413,
      "step": 832
    },
    {
      "epoch": 416.5,
      "grad_norm": 0.39097273349761963,
      "learning_rate": 3.5623885265958576e-05,
      "loss": 0.4282,
      "step": 833
    },
    {
      "epoch": 417.0,
      "grad_norm": 0.7393054366111755,
      "learning_rate": 3.55160746588254e-05,
      "loss": 0.3271,
      "step": 834
    },
    {
      "epoch": 417.5,
      "grad_norm": 0.41721704602241516,
      "learning_rate": 3.5408337509408764e-05,
      "loss": 0.3984,
      "step": 835
    },
    {
      "epoch": 418.0,
      "grad_norm": 0.9765669703483582,
      "learning_rate": 3.530067436411617e-05,
      "loss": 0.3203,
      "step": 836
    },
    {
      "epoch": 418.5,
      "grad_norm": 0.40932708978652954,
      "learning_rate": 3.519308576897974e-05,
      "loss": 0.5239,
      "step": 837
    },
    {
      "epoch": 419.0,
      "grad_norm": 1.1959574222564697,
      "learning_rate": 3.5085572269653533e-05,
      "loss": 0.3844,
      "step": 838
    },
    {
      "epoch": 419.5,
      "grad_norm": 1.3366146087646484,
      "learning_rate": 3.497813441141077e-05,
      "loss": 0.4664,
      "step": 839
    },
    {
      "epoch": 420.0,
      "grad_norm": 1.4755361080169678,
      "learning_rate": 3.4870772739140955e-05,
      "loss": 0.3329,
      "step": 840
    },
    {
      "epoch": 420.5,
      "grad_norm": 1.525937557220459,
      "learning_rate": 3.476348779734732e-05,
      "loss": 0.6967,
      "step": 841
    },
    {
      "epoch": 421.0,
      "grad_norm": 0.6713773012161255,
      "learning_rate": 3.4656280130143835e-05,
      "loss": 0.2996,
      "step": 842
    },
    {
      "epoch": 421.5,
      "grad_norm": 1.4801234006881714,
      "learning_rate": 3.4549150281252636e-05,
      "loss": 0.5395,
      "step": 843
    },
    {
      "epoch": 422.0,
      "grad_norm": 1.2286485433578491,
      "learning_rate": 3.4442098794001144e-05,
      "loss": 0.7762,
      "step": 844
    },
    {
      "epoch": 422.5,
      "grad_norm": 0.5565068125724792,
      "learning_rate": 3.433512621131941e-05,
      "loss": 0.4198,
      "step": 845
    },
    {
      "epoch": 423.0,
      "grad_norm": 0.713882327079773,
      "learning_rate": 3.422823307573722e-05,
      "loss": 0.3578,
      "step": 846
    },
    {
      "epoch": 423.5,
      "grad_norm": 0.9831527471542358,
      "learning_rate": 3.412141992938155e-05,
      "loss": 0.5304,
      "step": 847
    },
    {
      "epoch": 424.0,
      "grad_norm": 3.615861177444458,
      "learning_rate": 3.401468731397355e-05,
      "loss": 0.6533,
      "step": 848
    },
    {
      "epoch": 424.5,
      "grad_norm": 0.7093788385391235,
      "learning_rate": 3.390803577082609e-05,
      "loss": 0.5724,
      "step": 849
    },
    {
      "epoch": 425.0,
      "grad_norm": 0.7051779627799988,
      "learning_rate": 3.380146584084074e-05,
      "loss": 0.3081,
      "step": 850
    },
    {
      "epoch": 425.5,
      "grad_norm": 0.45248836278915405,
      "learning_rate": 3.3694978064505254e-05,
      "loss": 0.4441,
      "step": 851
    },
    {
      "epoch": 426.0,
      "grad_norm": 0.6368857622146606,
      "learning_rate": 3.3588572981890685e-05,
      "loss": 0.284,
      "step": 852
    },
    {
      "epoch": 426.5,
      "grad_norm": 0.5291391015052795,
      "learning_rate": 3.34822511326487e-05,
      "loss": 0.4573,
      "step": 853
    },
    {
      "epoch": 427.0,
      "grad_norm": 1.032855749130249,
      "learning_rate": 3.337601305600882e-05,
      "loss": 0.3181,
      "step": 854
    },
    {
      "epoch": 427.5,
      "grad_norm": 1.5788564682006836,
      "learning_rate": 3.3269859290775704e-05,
      "loss": 0.6879,
      "step": 855
    },
    {
      "epoch": 428.0,
      "grad_norm": 1.0379509925842285,
      "learning_rate": 3.316379037532644e-05,
      "loss": 0.3628,
      "step": 856
    },
    {
      "epoch": 428.5,
      "grad_norm": 0.6462805271148682,
      "learning_rate": 3.305780684760773e-05,
      "loss": 0.4379,
      "step": 857
    },
    {
      "epoch": 429.0,
      "grad_norm": 0.83830726146698,
      "learning_rate": 3.2951909245133276e-05,
      "loss": 0.3294,
      "step": 858
    },
    {
      "epoch": 429.5,
      "grad_norm": 0.6177251935005188,
      "learning_rate": 3.2846098104980937e-05,
      "loss": 0.5267,
      "step": 859
    },
    {
      "epoch": 430.0,
      "grad_norm": 0.8889451026916504,
      "learning_rate": 3.2740373963790146e-05,
      "loss": 0.3189,
      "step": 860
    },
    {
      "epoch": 430.5,
      "grad_norm": 0.46218279004096985,
      "learning_rate": 3.263473735775899e-05,
      "loss": 0.4275,
      "step": 861
    },
    {
      "epoch": 431.0,
      "grad_norm": 1.3931822776794434,
      "learning_rate": 3.252918882264172e-05,
      "loss": 1.0849,
      "step": 862
    },
    {
      "epoch": 431.5,
      "grad_norm": 0.7624893188476562,
      "learning_rate": 3.2423728893745855e-05,
      "loss": 0.3929,
      "step": 863
    },
    {
      "epoch": 432.0,
      "grad_norm": 0.8794317245483398,
      "learning_rate": 3.231835810592954e-05,
      "loss": 0.4069,
      "step": 864
    },
    {
      "epoch": 432.5,
      "grad_norm": 0.74793541431427,
      "learning_rate": 3.221307699359886e-05,
      "loss": 0.5508,
      "step": 865
    },
    {
      "epoch": 433.0,
      "grad_norm": 30.461246490478516,
      "learning_rate": 3.210788609070504e-05,
      "loss": 1.7061,
      "step": 866
    },
    {
      "epoch": 433.5,
      "grad_norm": 0.9695721864700317,
      "learning_rate": 3.200278593074185e-05,
      "loss": 0.6087,
      "step": 867
    },
    {
      "epoch": 434.0,
      "grad_norm": 0.8778077363967896,
      "learning_rate": 3.189777704674278e-05,
      "loss": 0.3003,
      "step": 868
    },
    {
      "epoch": 434.5,
      "grad_norm": 0.6212223172187805,
      "learning_rate": 3.179285997127848e-05,
      "loss": 0.4942,
      "step": 869
    },
    {
      "epoch": 435.0,
      "grad_norm": 0.6060783267021179,
      "learning_rate": 3.168803523645387e-05,
      "loss": 0.2747,
      "step": 870
    },
    {
      "epoch": 435.5,
      "grad_norm": 0.44244325160980225,
      "learning_rate": 3.158330337390565e-05,
      "loss": 0.397,
      "step": 871
    },
    {
      "epoch": 436.0,
      "grad_norm": 0.7622948288917542,
      "learning_rate": 3.147866491479945e-05,
      "loss": 0.3164,
      "step": 872
    },
    {
      "epoch": 436.5,
      "grad_norm": 0.4214974343776703,
      "learning_rate": 3.137412038982719e-05,
      "loss": 0.4769,
      "step": 873
    },
    {
      "epoch": 437.0,
      "grad_norm": 27.40167808532715,
      "learning_rate": 3.12696703292044e-05,
      "loss": 2.2768,
      "step": 874
    },
    {
      "epoch": 437.5,
      "grad_norm": 0.6034371256828308,
      "learning_rate": 3.116531526266753e-05,
      "loss": 0.5137,
      "step": 875
    },
    {
      "epoch": 438.0,
      "grad_norm": 0.7559489607810974,
      "learning_rate": 3.10610557194712e-05,
      "loss": 0.3348,
      "step": 876
    },
    {
      "epoch": 438.5,
      "grad_norm": 0.39580613374710083,
      "learning_rate": 3.095689222838563e-05,
      "loss": 0.3713,
      "step": 877
    },
    {
      "epoch": 439.0,
      "grad_norm": 0.7628496289253235,
      "learning_rate": 3.085282531769387e-05,
      "loss": 0.3016,
      "step": 878
    },
    {
      "epoch": 439.5,
      "grad_norm": 0.5997427105903625,
      "learning_rate": 3.0748855515189104e-05,
      "loss": 0.5073,
      "step": 879
    },
    {
      "epoch": 440.0,
      "grad_norm": 0.9870160222053528,
      "learning_rate": 3.064498334817211e-05,
      "loss": 0.286,
      "step": 880
    },
    {
      "epoch": 440.5,
      "grad_norm": 0.32235556840896606,
      "learning_rate": 3.0541209343448376e-05,
      "loss": 0.4198,
      "step": 881
    },
    {
      "epoch": 441.0,
      "grad_norm": 0.4676821529865265,
      "learning_rate": 3.0437534027325633e-05,
      "loss": 0.271,
      "step": 882
    },
    {
      "epoch": 441.5,
      "grad_norm": 0.35034623742103577,
      "learning_rate": 3.0333957925611056e-05,
      "loss": 0.4694,
      "step": 883
    },
    {
      "epoch": 442.0,
      "grad_norm": 1.3105062246322632,
      "learning_rate": 3.0230481563608672e-05,
      "loss": 0.9939,
      "step": 884
    },
    {
      "epoch": 442.5,
      "grad_norm": 0.5963773131370544,
      "learning_rate": 3.0127105466116588e-05,
      "loss": 0.3962,
      "step": 885
    },
    {
      "epoch": 443.0,
      "grad_norm": 6.630970478057861,
      "learning_rate": 3.0023830157424503e-05,
      "loss": 0.8818,
      "step": 886
    },
    {
      "epoch": 443.5,
      "grad_norm": 1.374496579170227,
      "learning_rate": 2.9920656161310834e-05,
      "loss": 0.6273,
      "step": 887
    },
    {
      "epoch": 444.0,
      "grad_norm": 1.0595529079437256,
      "learning_rate": 2.9817584001040282e-05,
      "loss": 0.3133,
      "step": 888
    },
    {
      "epoch": 444.5,
      "grad_norm": 0.28964123129844666,
      "learning_rate": 2.971461419936099e-05,
      "loss": 0.3956,
      "step": 889
    },
    {
      "epoch": 445.0,
      "grad_norm": 1.0725104808807373,
      "learning_rate": 2.9611747278502012e-05,
      "loss": 0.3964,
      "step": 890
    },
    {
      "epoch": 445.5,
      "grad_norm": 0.48483845591545105,
      "learning_rate": 2.950898376017064e-05,
      "loss": 0.3572,
      "step": 891
    },
    {
      "epoch": 446.0,
      "grad_norm": 1.017894983291626,
      "learning_rate": 2.940632416554967e-05,
      "loss": 0.3405,
      "step": 892
    },
    {
      "epoch": 446.5,
      "grad_norm": 0.4015076160430908,
      "learning_rate": 2.9303769015294926e-05,
      "loss": 0.4297,
      "step": 893
    },
    {
      "epoch": 447.0,
      "grad_norm": 1.1023564338684082,
      "learning_rate": 2.9201318829532447e-05,
      "loss": 0.2936,
      "step": 894
    },
    {
      "epoch": 447.5,
      "grad_norm": 0.33282431960105896,
      "learning_rate": 2.9098974127856e-05,
      "loss": 0.3725,
      "step": 895
    },
    {
      "epoch": 448.0,
      "grad_norm": 1.274607539176941,
      "learning_rate": 2.8996735429324255e-05,
      "loss": 0.522,
      "step": 896
    },
    {
      "epoch": 448.5,
      "grad_norm": 0.3814760744571686,
      "learning_rate": 2.8894603252458408e-05,
      "loss": 0.4108,
      "step": 897
    },
    {
      "epoch": 449.0,
      "grad_norm": 1.072039008140564,
      "learning_rate": 2.8792578115239323e-05,
      "loss": 0.3243,
      "step": 898
    },
    {
      "epoch": 449.5,
      "grad_norm": 2.360821008682251,
      "learning_rate": 2.8690660535105008e-05,
      "loss": 0.5375,
      "step": 899
    },
    {
      "epoch": 450.0,
      "grad_norm": 1.2944382429122925,
      "learning_rate": 2.8588851028948004e-05,
      "loss": 0.303,
      "step": 900
    },
    {
      "epoch": 450.5,
      "grad_norm": 0.7298043966293335,
      "learning_rate": 2.848715011311271e-05,
      "loss": 0.524,
      "step": 901
    },
    {
      "epoch": 451.0,
      "grad_norm": 0.7443506717681885,
      "learning_rate": 2.8385558303392795e-05,
      "loss": 0.2819,
      "step": 902
    },
    {
      "epoch": 451.5,
      "grad_norm": 0.5754256844520569,
      "learning_rate": 2.828407611502857e-05,
      "loss": 0.3595,
      "step": 903
    },
    {
      "epoch": 452.0,
      "grad_norm": 1.281030535697937,
      "learning_rate": 2.818270406270444e-05,
      "loss": 1.0711,
      "step": 904
    },
    {
      "epoch": 452.5,
      "grad_norm": 0.42849820852279663,
      "learning_rate": 2.8081442660546125e-05,
      "loss": 0.445,
      "step": 905
    },
    {
      "epoch": 453.0,
      "grad_norm": 0.9029936790466309,
      "learning_rate": 2.798029242211828e-05,
      "loss": 0.3381,
      "step": 906
    },
    {
      "epoch": 453.5,
      "grad_norm": 0.4585476219654083,
      "learning_rate": 2.787925386042171e-05,
      "loss": 0.4402,
      "step": 907
    },
    {
      "epoch": 454.0,
      "grad_norm": 0.9678264260292053,
      "learning_rate": 2.7778327487890837e-05,
      "loss": 0.3294,
      "step": 908
    },
    {
      "epoch": 454.5,
      "grad_norm": 0.5579683780670166,
      "learning_rate": 2.767751381639111e-05,
      "loss": 0.3938,
      "step": 909
    },
    {
      "epoch": 455.0,
      "grad_norm": 0.9896472692489624,
      "learning_rate": 2.757681335721639e-05,
      "loss": 0.3663,
      "step": 910
    },
    {
      "epoch": 455.5,
      "grad_norm": 0.56807541847229,
      "learning_rate": 2.7476226621086354e-05,
      "loss": 0.3673,
      "step": 911
    },
    {
      "epoch": 456.0,
      "grad_norm": 1.372495174407959,
      "learning_rate": 2.7375754118143932e-05,
      "loss": 1.0935,
      "step": 912
    },
    {
      "epoch": 456.5,
      "grad_norm": 3.563377857208252,
      "learning_rate": 2.727539635795268e-05,
      "loss": 0.5135,
      "step": 913
    },
    {
      "epoch": 457.0,
      "grad_norm": 0.9328904151916504,
      "learning_rate": 2.717515384949424e-05,
      "loss": 0.2898,
      "step": 914
    },
    {
      "epoch": 457.5,
      "grad_norm": 0.5711988210678101,
      "learning_rate": 2.7075027101165707e-05,
      "loss": 0.4798,
      "step": 915
    },
    {
      "epoch": 458.0,
      "grad_norm": 7.771111011505127,
      "learning_rate": 2.6975016620777073e-05,
      "loss": 0.7028,
      "step": 916
    },
    {
      "epoch": 458.5,
      "grad_norm": 0.6978394985198975,
      "learning_rate": 2.6875122915548724e-05,
      "loss": 0.3836,
      "step": 917
    },
    {
      "epoch": 459.0,
      "grad_norm": 0.7085824608802795,
      "learning_rate": 2.6775346492108734e-05,
      "loss": 0.3223,
      "step": 918
    },
    {
      "epoch": 459.5,
      "grad_norm": 0.9997246861457825,
      "learning_rate": 2.6675687856490382e-05,
      "loss": 0.5242,
      "step": 919
    },
    {
      "epoch": 460.0,
      "grad_norm": 5.7084479331970215,
      "learning_rate": 2.6576147514129558e-05,
      "loss": 0.5673,
      "step": 920
    },
    {
      "epoch": 460.5,
      "grad_norm": 0.5244137048721313,
      "learning_rate": 2.6476725969862227e-05,
      "loss": 0.5355,
      "step": 921
    },
    {
      "epoch": 461.0,
      "grad_norm": 0.723539412021637,
      "learning_rate": 2.6377423727921825e-05,
      "loss": 0.2838,
      "step": 922
    },
    {
      "epoch": 461.5,
      "grad_norm": 0.50766521692276,
      "learning_rate": 2.6278241291936743e-05,
      "loss": 0.4366,
      "step": 923
    },
    {
      "epoch": 462.0,
      "grad_norm": 1.0225409269332886,
      "learning_rate": 2.6179179164927758e-05,
      "loss": 0.3396,
      "step": 924
    },
    {
      "epoch": 462.5,
      "grad_norm": 0.35129672288894653,
      "learning_rate": 2.6080237849305468e-05,
      "loss": 0.3804,
      "step": 925
    },
    {
      "epoch": 463.0,
      "grad_norm": 0.6580302715301514,
      "learning_rate": 2.598141784686775e-05,
      "loss": 0.3155,
      "step": 926
    },
    {
      "epoch": 463.5,
      "grad_norm": 0.5437390804290771,
      "learning_rate": 2.5882719658797233e-05,
      "loss": 0.4589,
      "step": 927
    },
    {
      "epoch": 464.0,
      "grad_norm": 1.3478631973266602,
      "learning_rate": 2.5784143785658733e-05,
      "loss": 0.2943,
      "step": 928
    },
    {
      "epoch": 464.5,
      "grad_norm": 1.1417429447174072,
      "learning_rate": 2.568569072739675e-05,
      "loss": 0.4489,
      "step": 929
    },
    {
      "epoch": 465.0,
      "grad_norm": 1.2299023866653442,
      "learning_rate": 2.5587360983332887e-05,
      "loss": 0.3255,
      "step": 930
    },
    {
      "epoch": 465.5,
      "grad_norm": 0.3698990046977997,
      "learning_rate": 2.548915505216333e-05,
      "loss": 0.3591,
      "step": 931
    },
    {
      "epoch": 466.0,
      "grad_norm": 0.6597965359687805,
      "learning_rate": 2.5391073431956348e-05,
      "loss": 0.3044,
      "step": 932
    },
    {
      "epoch": 466.5,
      "grad_norm": 1.27484929561615,
      "learning_rate": 2.529311662014972e-05,
      "loss": 0.4845,
      "step": 933
    },
    {
      "epoch": 467.0,
      "grad_norm": 1.4219127893447876,
      "learning_rate": 2.519528511354826e-05,
      "loss": 0.3169,
      "step": 934
    },
    {
      "epoch": 467.5,
      "grad_norm": 1.2331948280334473,
      "learning_rate": 2.5097579408321266e-05,
      "loss": 0.5113,
      "step": 935
    },
    {
      "epoch": 468.0,
      "grad_norm": 1.2256078720092773,
      "learning_rate": 2.500000000000001e-05,
      "loss": 0.3409,
      "step": 936
    },
    {
      "epoch": 468.5,
      "grad_norm": 0.36103686690330505,
      "learning_rate": 2.4902547383475216e-05,
      "loss": 0.3645,
      "step": 937
    },
    {
      "epoch": 469.0,
      "grad_norm": 1.0091758966445923,
      "learning_rate": 2.4805222052994638e-05,
      "loss": 0.2789,
      "step": 938
    },
    {
      "epoch": 469.5,
      "grad_norm": 0.9824321269989014,
      "learning_rate": 2.4708024502160325e-05,
      "loss": 0.5688,
      "step": 939
    },
    {
      "epoch": 470.0,
      "grad_norm": 1.1693477630615234,
      "learning_rate": 2.4610955223926457e-05,
      "loss": 0.577,
      "step": 940
    },
    {
      "epoch": 470.5,
      "grad_norm": 4.739299297332764,
      "learning_rate": 2.4514014710596466e-05,
      "loss": 0.6932,
      "step": 941
    },
    {
      "epoch": 471.0,
      "grad_norm": 0.48150163888931274,
      "learning_rate": 2.4417203453820892e-05,
      "loss": 0.2786,
      "step": 942
    },
    {
      "epoch": 471.5,
      "grad_norm": 1.6477049589157104,
      "learning_rate": 2.4320521944594642e-05,
      "loss": 0.562,
      "step": 943
    },
    {
      "epoch": 472.0,
      "grad_norm": 0.6504332423210144,
      "learning_rate": 2.4223970673254602e-05,
      "loss": 0.2849,
      "step": 944
    },
    {
      "epoch": 472.5,
      "grad_norm": 0.8466479182243347,
      "learning_rate": 2.4127550129477144e-05,
      "loss": 0.4568,
      "step": 945
    },
    {
      "epoch": 473.0,
      "grad_norm": 1.4963505268096924,
      "learning_rate": 2.4031260802275623e-05,
      "loss": 0.5689,
      "step": 946
    },
    {
      "epoch": 473.5,
      "grad_norm": 0.45644015073776245,
      "learning_rate": 2.393510317999791e-05,
      "loss": 0.3971,
      "step": 947
    },
    {
      "epoch": 474.0,
      "grad_norm": 0.8477082848548889,
      "learning_rate": 2.383907775032389e-05,
      "loss": 0.3494,
      "step": 948
    },
    {
      "epoch": 474.5,
      "grad_norm": 1.7000072002410889,
      "learning_rate": 2.3743185000263097e-05,
      "loss": 0.4902,
      "step": 949
    },
    {
      "epoch": 475.0,
      "grad_norm": 1.326416254043579,
      "learning_rate": 2.364742541615201e-05,
      "loss": 0.3054,
      "step": 950
    },
    {
      "epoch": 475.5,
      "grad_norm": 0.3694341480731964,
      "learning_rate": 2.3551799483651894e-05,
      "loss": 0.3988,
      "step": 951
    },
    {
      "epoch": 476.0,
      "grad_norm": 0.7845640778541565,
      "learning_rate": 2.3456307687746027e-05,
      "loss": 0.2931,
      "step": 952
    },
    {
      "epoch": 476.5,
      "grad_norm": 1.0786240100860596,
      "learning_rate": 2.336095051273754e-05,
      "loss": 0.4742,
      "step": 953
    },
    {
      "epoch": 477.0,
      "grad_norm": 2.012897491455078,
      "learning_rate": 2.326572844224665e-05,
      "loss": 0.6121,
      "step": 954
    },
    {
      "epoch": 477.5,
      "grad_norm": 0.34849363565444946,
      "learning_rate": 2.3170641959208516e-05,
      "loss": 0.4423,
      "step": 955
    },
    {
      "epoch": 478.0,
      "grad_norm": 1.6290795803070068,
      "learning_rate": 2.3075691545870558e-05,
      "loss": 0.4713,
      "step": 956
    },
    {
      "epoch": 478.5,
      "grad_norm": 0.5413875579833984,
      "learning_rate": 2.298087768379012e-05,
      "loss": 0.3606,
      "step": 957
    },
    {
      "epoch": 479.0,
      "grad_norm": 0.9787770509719849,
      "learning_rate": 2.288620085383199e-05,
      "loss": 0.3942,
      "step": 958
    },
    {
      "epoch": 479.5,
      "grad_norm": 2.884401559829712,
      "learning_rate": 2.2791661536165964e-05,
      "loss": 0.4498,
      "step": 959
    },
    {
      "epoch": 480.0,
      "grad_norm": 0.9656736254692078,
      "learning_rate": 2.2697260210264505e-05,
      "loss": 0.2764,
      "step": 960
    },
    {
      "epoch": 480.5,
      "grad_norm": 0.7164297103881836,
      "learning_rate": 2.2602997354900075e-05,
      "loss": 0.3474,
      "step": 961
    },
    {
      "epoch": 481.0,
      "grad_norm": 1.9617085456848145,
      "learning_rate": 2.2508873448143025e-05,
      "loss": 0.9071,
      "step": 962
    },
    {
      "epoch": 481.5,
      "grad_norm": 0.6157837510108948,
      "learning_rate": 2.2414888967358844e-05,
      "loss": 0.3981,
      "step": 963
    },
    {
      "epoch": 482.0,
      "grad_norm": 1.7687170505523682,
      "learning_rate": 2.232104438920604e-05,
      "loss": 0.4634,
      "step": 964
    },
    {
      "epoch": 482.5,
      "grad_norm": 0.5850176215171814,
      "learning_rate": 2.2227340189633505e-05,
      "loss": 0.4553,
      "step": 965
    },
    {
      "epoch": 483.0,
      "grad_norm": 1.6233370304107666,
      "learning_rate": 2.2133776843878186e-05,
      "loss": 0.9948,
      "step": 966
    },
    {
      "epoch": 483.5,
      "grad_norm": 0.41611436009407043,
      "learning_rate": 2.2040354826462668e-05,
      "loss": 0.4906,
      "step": 967
    },
    {
      "epoch": 484.0,
      "grad_norm": 1.4037021398544312,
      "learning_rate": 2.1947074611192748e-05,
      "loss": 0.9941,
      "step": 968
    },
    {
      "epoch": 484.5,
      "grad_norm": 1.5858484506607056,
      "learning_rate": 2.1853936671155127e-05,
      "loss": 0.5178,
      "step": 969
    },
    {
      "epoch": 485.0,
      "grad_norm": 1.8030040264129639,
      "learning_rate": 2.1760941478714808e-05,
      "loss": 0.4306,
      "step": 970
    },
    {
      "epoch": 485.5,
      "grad_norm": 0.7251501679420471,
      "learning_rate": 2.1668089505512956e-05,
      "loss": 0.4579,
      "step": 971
    },
    {
      "epoch": 486.0,
      "grad_norm": 0.5055922865867615,
      "learning_rate": 2.1575381222464237e-05,
      "loss": 0.2633,
      "step": 972
    },
    {
      "epoch": 486.5,
      "grad_norm": 2.275453567504883,
      "learning_rate": 2.1482817099754715e-05,
      "loss": 0.5743,
      "step": 973
    },
    {
      "epoch": 487.0,
      "grad_norm": 1.7829012870788574,
      "learning_rate": 2.139039760683918e-05,
      "loss": 0.754,
      "step": 974
    },
    {
      "epoch": 487.5,
      "grad_norm": 0.5397523045539856,
      "learning_rate": 2.1298123212439026e-05,
      "loss": 0.3586,
      "step": 975
    },
    {
      "epoch": 488.0,
      "grad_norm": 1.3121229410171509,
      "learning_rate": 2.120599438453968e-05,
      "loss": 0.3468,
      "step": 976
    },
    {
      "epoch": 488.5,
      "grad_norm": 0.44475266337394714,
      "learning_rate": 2.1114011590388322e-05,
      "loss": 0.3799,
      "step": 977
    },
    {
      "epoch": 489.0,
      "grad_norm": 2.2388317584991455,
      "learning_rate": 2.1022175296491515e-05,
      "loss": 0.9219,
      "step": 978
    },
    {
      "epoch": 489.5,
      "grad_norm": 0.46189919114112854,
      "learning_rate": 2.0930485968612775e-05,
      "loss": 0.3733,
      "step": 979
    },
    {
      "epoch": 490.0,
      "grad_norm": 1.6083147525787354,
      "learning_rate": 2.0838944071770343e-05,
      "loss": 0.4656,
      "step": 980
    },
    {
      "epoch": 490.5,
      "grad_norm": 3.0061259269714355,
      "learning_rate": 2.074755007023461e-05,
      "loss": 0.5864,
      "step": 981
    },
    {
      "epoch": 491.0,
      "grad_norm": 0.48393088579177856,
      "learning_rate": 2.065630442752604e-05,
      "loss": 0.2721,
      "step": 982
    },
    {
      "epoch": 491.5,
      "grad_norm": 0.5466325283050537,
      "learning_rate": 2.0565207606412513e-05,
      "loss": 0.5295,
      "step": 983
    },
    {
      "epoch": 492.0,
      "grad_norm": 1.1526198387145996,
      "learning_rate": 2.047426006890728e-05,
      "loss": 0.3666,
      "step": 984
    },
    {
      "epoch": 492.5,
      "grad_norm": 0.6325269937515259,
      "learning_rate": 2.0383462276266347e-05,
      "loss": 0.4363,
      "step": 985
    },
    {
      "epoch": 493.0,
      "grad_norm": 1.1646724939346313,
      "learning_rate": 2.0292814688986372e-05,
      "loss": 0.2818,
      "step": 986
    },
    {
      "epoch": 493.5,
      "grad_norm": 0.34880033135414124,
      "learning_rate": 2.0202317766802154e-05,
      "loss": 0.3418,
      "step": 987
    },
    {
      "epoch": 494.0,
      "grad_norm": 1.619282841682434,
      "learning_rate": 2.011197196868439e-05,
      "loss": 0.8805,
      "step": 988
    },
    {
      "epoch": 494.5,
      "grad_norm": 0.3844090700149536,
      "learning_rate": 2.0021777752837316e-05,
      "loss": 0.4108,
      "step": 989
    },
    {
      "epoch": 495.0,
      "grad_norm": 1.332435965538025,
      "learning_rate": 1.9931735576696402e-05,
      "loss": 0.4071,
      "step": 990
    },
    {
      "epoch": 495.5,
      "grad_norm": 0.3266851305961609,
      "learning_rate": 1.984184589692602e-05,
      "loss": 0.4192,
      "step": 991
    },
    {
      "epoch": 496.0,
      "grad_norm": 1.6106630563735962,
      "learning_rate": 1.975210916941711e-05,
      "loss": 0.6745,
      "step": 992
    },
    {
      "epoch": 496.5,
      "grad_norm": 0.39592084288597107,
      "learning_rate": 1.9662525849284962e-05,
      "loss": 0.3322,
      "step": 993
    },
    {
      "epoch": 497.0,
      "grad_norm": 38.08456802368164,
      "learning_rate": 1.9573096390866707e-05,
      "loss": 2.3484,
      "step": 994
    },
    {
      "epoch": 497.5,
      "grad_norm": 0.54253751039505,
      "learning_rate": 1.948382124771927e-05,
      "loss": 0.4635,
      "step": 995
    },
    {
      "epoch": 498.0,
      "grad_norm": 0.9106139540672302,
      "learning_rate": 1.9394700872616855e-05,
      "loss": 0.3165,
      "step": 996
    },
    {
      "epoch": 498.5,
      "grad_norm": 1.6156233549118042,
      "learning_rate": 1.9305735717548772e-05,
      "loss": 0.5144,
      "step": 997
    },
    {
      "epoch": 499.0,
      "grad_norm": 1.9692583084106445,
      "learning_rate": 1.9216926233717085e-05,
      "loss": 1.1089,
      "step": 998
    },
    {
      "epoch": 499.5,
      "grad_norm": 0.7056724429130554,
      "learning_rate": 1.9128272871534362e-05,
      "loss": 0.4645,
      "step": 999
    },
    {
      "epoch": 500.0,
      "grad_norm": 2.410404682159424,
      "learning_rate": 1.9039776080621363e-05,
      "loss": 0.3709,
      "step": 1000
    },
    {
      "epoch": 500.5,
      "grad_norm": 0.3767830431461334,
      "learning_rate": 1.8951436309804765e-05,
      "loss": 0.4989,
      "step": 1001
    },
    {
      "epoch": 501.0,
      "grad_norm": 1.0018634796142578,
      "learning_rate": 1.886325400711491e-05,
      "loss": 0.3223,
      "step": 1002
    },
    {
      "epoch": 501.5,
      "grad_norm": 3.7476444244384766,
      "learning_rate": 1.877522961978349e-05,
      "loss": 0.537,
      "step": 1003
    },
    {
      "epoch": 502.0,
      "grad_norm": 2.227095603942871,
      "learning_rate": 1.868736359424132e-05,
      "loss": 0.6825,
      "step": 1004
    },
    {
      "epoch": 502.5,
      "grad_norm": 0.39756157994270325,
      "learning_rate": 1.8599656376116026e-05,
      "loss": 0.3634,
      "step": 1005
    },
    {
      "epoch": 503.0,
      "grad_norm": 1.1212667226791382,
      "learning_rate": 1.8512108410229878e-05,
      "loss": 0.3991,
      "step": 1006
    },
    {
      "epoch": 503.5,
      "grad_norm": 0.7371616363525391,
      "learning_rate": 1.8424720140597408e-05,
      "loss": 0.3955,
      "step": 1007
    },
    {
      "epoch": 504.0,
      "grad_norm": 1.3234190940856934,
      "learning_rate": 1.8337492010423252e-05,
      "loss": 0.4309,
      "step": 1008
    },
    {
      "epoch": 504.5,
      "grad_norm": 0.8846190571784973,
      "learning_rate": 1.825042446209985e-05,
      "loss": 0.3815,
      "step": 1009
    },
    {
      "epoch": 505.0,
      "grad_norm": 1.1892130374908447,
      "learning_rate": 1.8163517937205256e-05,
      "loss": 0.3009,
      "step": 1010
    },
    {
      "epoch": 505.5,
      "grad_norm": 0.35917583107948303,
      "learning_rate": 1.807677287650083e-05,
      "loss": 0.3997,
      "step": 1011
    },
    {
      "epoch": 506.0,
      "grad_norm": 15.531428337097168,
      "learning_rate": 1.7990189719929063e-05,
      "loss": 3.3518,
      "step": 1012
    },
    {
      "epoch": 506.5,
      "grad_norm": 1.147783637046814,
      "learning_rate": 1.7903768906611312e-05,
      "loss": 0.4782,
      "step": 1013
    },
    {
      "epoch": 507.0,
      "grad_norm": 1.2089029550552368,
      "learning_rate": 1.7817510874845583e-05,
      "loss": 0.3992,
      "step": 1014
    },
    {
      "epoch": 507.5,
      "grad_norm": 1.1861144304275513,
      "learning_rate": 1.773141606210431e-05,
      "loss": 0.6336,
      "step": 1015
    },
    {
      "epoch": 508.0,
      "grad_norm": 0.8112418055534363,
      "learning_rate": 1.7645484905032128e-05,
      "loss": 0.3005,
      "step": 1016
    },
    {
      "epoch": 508.5,
      "grad_norm": 0.6656642556190491,
      "learning_rate": 1.755971783944366e-05,
      "loss": 0.4219,
      "step": 1017
    },
    {
      "epoch": 509.0,
      "grad_norm": 0.7651844024658203,
      "learning_rate": 1.7474115300321342e-05,
      "loss": 0.2799,
      "step": 1018
    },
    {
      "epoch": 509.5,
      "grad_norm": 0.5986937880516052,
      "learning_rate": 1.738867772181316e-05,
      "loss": 0.3718,
      "step": 1019
    },
    {
      "epoch": 510.0,
      "grad_norm": 0.6151016354560852,
      "learning_rate": 1.7303405537230455e-05,
      "loss": 0.2591,
      "step": 1020
    },
    {
      "epoch": 510.5,
      "grad_norm": 0.6035138368606567,
      "learning_rate": 1.721829917904579e-05,
      "loss": 0.3612,
      "step": 1021
    },
    {
      "epoch": 511.0,
      "grad_norm": 1.8672927618026733,
      "learning_rate": 1.7133359078890665e-05,
      "loss": 0.3346,
      "step": 1022
    },
    {
      "epoch": 511.5,
      "grad_norm": 0.9273606538772583,
      "learning_rate": 1.7048585667553412e-05,
      "loss": 0.5253,
      "step": 1023
    },
    {
      "epoch": 512.0,
      "grad_norm": 1.692618727684021,
      "learning_rate": 1.6963979374976947e-05,
      "loss": 0.526,
      "step": 1024
    },
    {
      "epoch": 512.5,
      "grad_norm": 1.4525573253631592,
      "learning_rate": 1.687954063025663e-05,
      "loss": 0.4485,
      "step": 1025
    },
    {
      "epoch": 513.0,
      "grad_norm": 1.9588305950164795,
      "learning_rate": 1.6795269861638042e-05,
      "loss": 0.4237,
      "step": 1026
    },
    {
      "epoch": 513.5,
      "grad_norm": 0.30284547805786133,
      "learning_rate": 1.6711167496514925e-05,
      "loss": 0.3776,
      "step": 1027
    },
    {
      "epoch": 514.0,
      "grad_norm": 1.181299090385437,
      "learning_rate": 1.662723396142681e-05,
      "loss": 0.9873,
      "step": 1028
    },
    {
      "epoch": 514.5,
      "grad_norm": 0.5849606990814209,
      "learning_rate": 1.6543469682057106e-05,
      "loss": 0.3747,
      "step": 1029
    },
    {
      "epoch": 515.0,
      "grad_norm": 0.8990150094032288,
      "learning_rate": 1.645987508323069e-05,
      "loss": 0.3591,
      "step": 1030
    },
    {
      "epoch": 515.5,
      "grad_norm": 1.2334336042404175,
      "learning_rate": 1.6376450588911984e-05,
      "loss": 0.5033,
      "step": 1031
    },
    {
      "epoch": 516.0,
      "grad_norm": 1.7191448211669922,
      "learning_rate": 1.6293196622202635e-05,
      "loss": 0.5267,
      "step": 1032
    },
    {
      "epoch": 516.5,
      "grad_norm": 3.416567802429199,
      "learning_rate": 1.621011360533944e-05,
      "loss": 0.4534,
      "step": 1033
    },
    {
      "epoch": 517.0,
      "grad_norm": 1.7406679391860962,
      "learning_rate": 1.6127201959692195e-05,
      "loss": 0.5335,
      "step": 1034
    },
    {
      "epoch": 517.5,
      "grad_norm": 0.743309497833252,
      "learning_rate": 1.6044462105761572e-05,
      "loss": 0.4143,
      "step": 1035
    },
    {
      "epoch": 518.0,
      "grad_norm": 0.7291414737701416,
      "learning_rate": 1.5961894463176942e-05,
      "loss": 0.2879,
      "step": 1036
    },
    {
      "epoch": 518.5,
      "grad_norm": 0.4149529039859772,
      "learning_rate": 1.587949945069429e-05,
      "loss": 0.4032,
      "step": 1037
    },
    {
      "epoch": 519.0,
      "grad_norm": 0.5417309999465942,
      "learning_rate": 1.5797277486194133e-05,
      "loss": 0.2724,
      "step": 1038
    },
    {
      "epoch": 519.5,
      "grad_norm": 4.0986809730529785,
      "learning_rate": 1.571522898667922e-05,
      "loss": 0.5884,
      "step": 1039
    },
    {
      "epoch": 520.0,
      "grad_norm": 3.0832550525665283,
      "learning_rate": 1.563335436827269e-05,
      "loss": 0.8514,
      "step": 1040
    },
    {
      "epoch": 520.5,
      "grad_norm": 0.47477272152900696,
      "learning_rate": 1.555165404621567e-05,
      "loss": 0.4769,
      "step": 1041
    },
    {
      "epoch": 521.0,
      "grad_norm": 0.6663293242454529,
      "learning_rate": 1.5470128434865454e-05,
      "loss": 0.2425,
      "step": 1042
    },
    {
      "epoch": 521.5,
      "grad_norm": 0.9349833726882935,
      "learning_rate": 1.5388777947693128e-05,
      "loss": 0.4928,
      "step": 1043
    },
    {
      "epoch": 522.0,
      "grad_norm": 0.525383710861206,
      "learning_rate": 1.5307602997281706e-05,
      "loss": 0.2636,
      "step": 1044
    },
    {
      "epoch": 522.5,
      "grad_norm": 1.125815510749817,
      "learning_rate": 1.5226603995323895e-05,
      "loss": 0.4082,
      "step": 1045
    },
    {
      "epoch": 523.0,
      "grad_norm": 6.597041130065918,
      "learning_rate": 1.5145781352620053e-05,
      "loss": 0.7882,
      "step": 1046
    },
    {
      "epoch": 523.5,
      "grad_norm": 1.4680581092834473,
      "learning_rate": 1.5065135479076099e-05,
      "loss": 0.3411,
      "step": 1047
    },
    {
      "epoch": 524.0,
      "grad_norm": 1.492038607597351,
      "learning_rate": 1.4984666783701418e-05,
      "loss": 0.3677,
      "step": 1048
    },
    {
      "epoch": 524.5,
      "grad_norm": 0.30255770683288574,
      "learning_rate": 1.4904375674606886e-05,
      "loss": 0.4146,
      "step": 1049
    },
    {
      "epoch": 525.0,
      "grad_norm": 0.7858338952064514,
      "learning_rate": 1.4824262559002594e-05,
      "loss": 0.3208,
      "step": 1050
    },
    {
      "epoch": 525.5,
      "grad_norm": 3.7921156883239746,
      "learning_rate": 1.4744327843196043e-05,
      "loss": 0.7807,
      "step": 1051
    },
    {
      "epoch": 526.0,
      "grad_norm": 1.0449076890945435,
      "learning_rate": 1.4664571932589815e-05,
      "loss": 0.3059,
      "step": 1052
    },
    {
      "epoch": 526.5,
      "grad_norm": 0.5799958109855652,
      "learning_rate": 1.4584995231679777e-05,
      "loss": 0.4463,
      "step": 1053
    },
    {
      "epoch": 527.0,
      "grad_norm": 1.3100086450576782,
      "learning_rate": 1.450559814405279e-05,
      "loss": 0.6078,
      "step": 1054
    },
    {
      "epoch": 527.5,
      "grad_norm": 0.3562907874584198,
      "learning_rate": 1.4426381072384865e-05,
      "loss": 0.4386,
      "step": 1055
    },
    {
      "epoch": 528.0,
      "grad_norm": 0.9038794636726379,
      "learning_rate": 1.434734441843899e-05,
      "loss": 0.2846,
      "step": 1056
    },
    {
      "epoch": 528.5,
      "grad_norm": 0.35892513394355774,
      "learning_rate": 1.4268488583063116e-05,
      "loss": 0.3681,
      "step": 1057
    },
    {
      "epoch": 529.0,
      "grad_norm": 1.8041563034057617,
      "learning_rate": 1.4189813966188214e-05,
      "loss": 0.5023,
      "step": 1058
    },
    {
      "epoch": 529.5,
      "grad_norm": 0.4089813232421875,
      "learning_rate": 1.4111320966826059e-05,
      "loss": 0.3706,
      "step": 1059
    },
    {
      "epoch": 530.0,
      "grad_norm": 0.7910075783729553,
      "learning_rate": 1.4033009983067452e-05,
      "loss": 0.3134,
      "step": 1060
    },
    {
      "epoch": 530.5,
      "grad_norm": 1.7190093994140625,
      "learning_rate": 1.3954881412079945e-05,
      "loss": 0.5805,
      "step": 1061
    },
    {
      "epoch": 531.0,
      "grad_norm": 0.8881721496582031,
      "learning_rate": 1.3876935650106071e-05,
      "loss": 0.2823,
      "step": 1062
    },
    {
      "epoch": 531.5,
      "grad_norm": 0.2901274561882019,
      "learning_rate": 1.3799173092461105e-05,
      "loss": 0.3872,
      "step": 1063
    },
    {
      "epoch": 532.0,
      "grad_norm": 1.0518895387649536,
      "learning_rate": 1.3721594133531252e-05,
      "loss": 0.3506,
      "step": 1064
    },
    {
      "epoch": 532.5,
      "grad_norm": 0.40705567598342896,
      "learning_rate": 1.3644199166771532e-05,
      "loss": 0.4508,
      "step": 1065
    },
    {
      "epoch": 533.0,
      "grad_norm": 1.7172421216964722,
      "learning_rate": 1.3566988584703816e-05,
      "loss": 0.8024,
      "step": 1066
    },
    {
      "epoch": 533.5,
      "grad_norm": 0.3861737847328186,
      "learning_rate": 1.3489962778914845e-05,
      "loss": 0.3341,
      "step": 1067
    },
    {
      "epoch": 534.0,
      "grad_norm": 1.28693425655365,
      "learning_rate": 1.3413122140054218e-05,
      "loss": 0.4927,
      "step": 1068
    },
    {
      "epoch": 534.5,
      "grad_norm": 0.7647294402122498,
      "learning_rate": 1.3336467057832458e-05,
      "loss": 0.4685,
      "step": 1069
    },
    {
      "epoch": 535.0,
      "grad_norm": 0.6108217835426331,
      "learning_rate": 1.3259997921018963e-05,
      "loss": 0.2599,
      "step": 1070
    },
    {
      "epoch": 535.5,
      "grad_norm": 0.43750202655792236,
      "learning_rate": 1.3183715117440142e-05,
      "loss": 0.4396,
      "step": 1071
    },
    {
      "epoch": 536.0,
      "grad_norm": 1.2389155626296997,
      "learning_rate": 1.3107619033977297e-05,
      "loss": 0.428,
      "step": 1072
    },
    {
      "epoch": 536.5,
      "grad_norm": 0.44225823879241943,
      "learning_rate": 1.3031710056564845e-05,
      "loss": 0.4248,
      "step": 1073
    },
    {
      "epoch": 537.0,
      "grad_norm": 2.291903018951416,
      "learning_rate": 1.2955988570188155e-05,
      "loss": 0.5914,
      "step": 1074
    },
    {
      "epoch": 537.5,
      "grad_norm": 1.7432060241699219,
      "learning_rate": 1.2880454958881794e-05,
      "loss": 0.4318,
      "step": 1075
    },
    {
      "epoch": 538.0,
      "grad_norm": 0.5825529098510742,
      "learning_rate": 1.280510960572745e-05,
      "loss": 0.2735,
      "step": 1076
    },
    {
      "epoch": 538.5,
      "grad_norm": 1.065626859664917,
      "learning_rate": 1.2729952892852021e-05,
      "loss": 0.4796,
      "step": 1077
    },
    {
      "epoch": 539.0,
      "grad_norm": 0.616766095161438,
      "learning_rate": 1.2654985201425701e-05,
      "loss": 0.2607,
      "step": 1078
    },
    {
      "epoch": 539.5,
      "grad_norm": 2.7967593669891357,
      "learning_rate": 1.2580206911660014e-05,
      "loss": 0.6349,
      "step": 1079
    },
    {
      "epoch": 540.0,
      "grad_norm": 1.9492617845535278,
      "learning_rate": 1.2505618402805907e-05,
      "loss": 0.7395,
      "step": 1080
    },
    {
      "epoch": 540.5,
      "grad_norm": 0.38567250967025757,
      "learning_rate": 1.2431220053151833e-05,
      "loss": 0.4176,
      "step": 1081
    },
    {
      "epoch": 541.0,
      "grad_norm": 1.254096508026123,
      "learning_rate": 1.2357012240021793e-05,
      "loss": 0.3162,
      "step": 1082
    },
    {
      "epoch": 541.5,
      "grad_norm": 1.4545925855636597,
      "learning_rate": 1.2282995339773457e-05,
      "loss": 0.4286,
      "step": 1083
    },
    {
      "epoch": 542.0,
      "grad_norm": 0.615583598613739,
      "learning_rate": 1.2209169727796289e-05,
      "loss": 0.2498,
      "step": 1084
    },
    {
      "epoch": 542.5,
      "grad_norm": 2.906977653503418,
      "learning_rate": 1.2135535778509544e-05,
      "loss": 0.4825,
      "step": 1085
    },
    {
      "epoch": 543.0,
      "grad_norm": 0.9604778289794922,
      "learning_rate": 1.2062093865360458e-05,
      "loss": 0.2636,
      "step": 1086
    },
    {
      "epoch": 543.5,
      "grad_norm": 0.44585615396499634,
      "learning_rate": 1.198884436082231e-05,
      "loss": 0.411,
      "step": 1087
    },
    {
      "epoch": 544.0,
      "grad_norm": 1.1484570503234863,
      "learning_rate": 1.191578763639255e-05,
      "loss": 0.353,
      "step": 1088
    },
    {
      "epoch": 544.5,
      "grad_norm": 0.5033761858940125,
      "learning_rate": 1.184292406259092e-05,
      "loss": 0.417,
      "step": 1089
    },
    {
      "epoch": 545.0,
      "grad_norm": 0.7544223070144653,
      "learning_rate": 1.1770254008957549e-05,
      "loss": 0.2978,
      "step": 1090
    },
    {
      "epoch": 545.5,
      "grad_norm": 0.48072242736816406,
      "learning_rate": 1.1697777844051105e-05,
      "loss": 0.495,
      "step": 1091
    },
    {
      "epoch": 546.0,
      "grad_norm": 0.6462698578834534,
      "learning_rate": 1.1625495935446918e-05,
      "loss": 0.2956,
      "step": 1092
    },
    {
      "epoch": 546.5,
      "grad_norm": 0.3545149564743042,
      "learning_rate": 1.1553408649735098e-05,
      "loss": 0.3855,
      "step": 1093
    },
    {
      "epoch": 547.0,
      "grad_norm": 2.0209903717041016,
      "learning_rate": 1.1481516352518701e-05,
      "loss": 0.404,
      "step": 1094
    },
    {
      "epoch": 547.5,
      "grad_norm": 2.841571092605591,
      "learning_rate": 1.1409819408411898e-05,
      "loss": 0.4851,
      "step": 1095
    },
    {
      "epoch": 548.0,
      "grad_norm": 0.505448579788208,
      "learning_rate": 1.1338318181038038e-05,
      "loss": 0.2947,
      "step": 1096
    },
    {
      "epoch": 548.5,
      "grad_norm": 1.485593318939209,
      "learning_rate": 1.1267013033027884e-05,
      "loss": 0.5424,
      "step": 1097
    },
    {
      "epoch": 549.0,
      "grad_norm": 0.8760380148887634,
      "learning_rate": 1.1195904326017759e-05,
      "loss": 0.2931,
      "step": 1098
    },
    {
      "epoch": 549.5,
      "grad_norm": 2.215996742248535,
      "learning_rate": 1.1124992420647684e-05,
      "loss": 0.5809,
      "step": 1099
    },
    {
      "epoch": 550.0,
      "grad_norm": 0.8580252528190613,
      "learning_rate": 1.1054277676559576e-05,
      "loss": 0.2665,
      "step": 1100
    },
    {
      "epoch": 550.5,
      "grad_norm": 0.36559003591537476,
      "learning_rate": 1.0983760452395413e-05,
      "loss": 0.3444,
      "step": 1101
    },
    {
      "epoch": 551.0,
      "grad_norm": 0.5747761130332947,
      "learning_rate": 1.0913441105795418e-05,
      "loss": 0.2831,
      "step": 1102
    },
    {
      "epoch": 551.5,
      "grad_norm": 0.24126912653446198,
      "learning_rate": 1.0843319993396234e-05,
      "loss": 0.3561,
      "step": 1103
    },
    {
      "epoch": 552.0,
      "grad_norm": 0.6384169459342957,
      "learning_rate": 1.0773397470829145e-05,
      "loss": 0.2913,
      "step": 1104
    },
    {
      "epoch": 552.5,
      "grad_norm": 1.3589190244674683,
      "learning_rate": 1.070367389271823e-05,
      "loss": 0.3985,
      "step": 1105
    },
    {
      "epoch": 553.0,
      "grad_norm": 0.7840595245361328,
      "learning_rate": 1.063414961267859e-05,
      "loss": 0.2608,
      "step": 1106
    },
    {
      "epoch": 553.5,
      "grad_norm": 0.5262243151664734,
      "learning_rate": 1.0564824983314592e-05,
      "loss": 0.3449,
      "step": 1107
    },
    {
      "epoch": 554.0,
      "grad_norm": 1.770022988319397,
      "learning_rate": 1.0495700356217985e-05,
      "loss": 0.7868,
      "step": 1108
    },
    {
      "epoch": 554.5,
      "grad_norm": 0.4141281247138977,
      "learning_rate": 1.04267760819662e-05,
      "loss": 0.3923,
      "step": 1109
    },
    {
      "epoch": 555.0,
      "grad_norm": 0.6363136172294617,
      "learning_rate": 1.0358052510120536e-05,
      "loss": 0.2811,
      "step": 1110
    },
    {
      "epoch": 555.5,
      "grad_norm": 0.40090975165367126,
      "learning_rate": 1.02895299892244e-05,
      "loss": 0.3789,
      "step": 1111
    },
    {
      "epoch": 556.0,
      "grad_norm": 0.5406911969184875,
      "learning_rate": 1.0221208866801536e-05,
      "loss": 0.3066,
      "step": 1112
    },
    {
      "epoch": 556.5,
      "grad_norm": 0.31015992164611816,
      "learning_rate": 1.0153089489354257e-05,
      "loss": 0.3686,
      "step": 1113
    },
    {
      "epoch": 557.0,
      "grad_norm": 0.5437638759613037,
      "learning_rate": 1.0085172202361687e-05,
      "loss": 0.2694,
      "step": 1114
    },
    {
      "epoch": 557.5,
      "grad_norm": 0.9308782815933228,
      "learning_rate": 1.001745735027801e-05,
      "loss": 0.4513,
      "step": 1115
    },
    {
      "epoch": 558.0,
      "grad_norm": 1.5071659088134766,
      "learning_rate": 9.949945276530781e-06,
      "loss": 0.5128,
      "step": 1116
    },
    {
      "epoch": 558.5,
      "grad_norm": 0.8416696786880493,
      "learning_rate": 9.882636323519018e-06,
      "loss": 0.4141,
      "step": 1117
    },
    {
      "epoch": 559.0,
      "grad_norm": 1.2935975790023804,
      "learning_rate": 9.815530832611708e-06,
      "loss": 0.9914,
      "step": 1118
    },
    {
      "epoch": 559.5,
      "grad_norm": 0.7350620627403259,
      "learning_rate": 9.748629144145827e-06,
      "loss": 0.4467,
      "step": 1119
    },
    {
      "epoch": 560.0,
      "grad_norm": 2.991042375564575,
      "learning_rate": 9.68193159742482e-06,
      "loss": 0.5512,
      "step": 1120
    },
    {
      "epoch": 560.5,
      "grad_norm": 0.3175847828388214,
      "learning_rate": 9.615438530716753e-06,
      "loss": 0.3306,
      "step": 1121
    },
    {
      "epoch": 561.0,
      "grad_norm": 1.2846187353134155,
      "learning_rate": 9.549150281252633e-06,
      "loss": 0.315,
      "step": 1122
    },
    {
      "epoch": 561.5,
      "grad_norm": 1.775314211845398,
      "learning_rate": 9.483067185224708e-06,
      "loss": 0.4889,
      "step": 1123
    },
    {
      "epoch": 562.0,
      "grad_norm": 0.41615575551986694,
      "learning_rate": 9.417189577784768e-06,
      "loss": 0.2416,
      "step": 1124
    },
    {
      "epoch": 562.5,
      "grad_norm": 1.3527556657791138,
      "learning_rate": 9.351517793042408e-06,
      "loss": 0.5783,
      "step": 1125
    },
    {
      "epoch": 563.0,
      "grad_norm": 0.5528387427330017,
      "learning_rate": 9.28605216406337e-06,
      "loss": 0.2812,
      "step": 1126
    },
    {
      "epoch": 563.5,
      "grad_norm": 0.40397512912750244,
      "learning_rate": 9.22079302286787e-06,
      "loss": 0.4187,
      "step": 1127
    },
    {
      "epoch": 564.0,
      "grad_norm": 0.7269508838653564,
      "learning_rate": 9.15574070042881e-06,
      "loss": 0.3451,
      "step": 1128
    },
    {
      "epoch": 564.5,
      "grad_norm": 1.8183125257492065,
      "learning_rate": 9.090895526670278e-06,
      "loss": 0.4054,
      "step": 1129
    },
    {
      "epoch": 565.0,
      "grad_norm": 1.5190259218215942,
      "learning_rate": 9.026257830465661e-06,
      "loss": 1.0072,
      "step": 1130
    },
    {
      "epoch": 565.5,
      "grad_norm": 0.3671301007270813,
      "learning_rate": 8.961827939636196e-06,
      "loss": 0.3681,
      "step": 1131
    },
    {
      "epoch": 566.0,
      "grad_norm": 1.7227364778518677,
      "learning_rate": 8.897606180949098e-06,
      "loss": 0.8687,
      "step": 1132
    },
    {
      "epoch": 566.5,
      "grad_norm": 1.5525699853897095,
      "learning_rate": 8.833592880116094e-06,
      "loss": 0.6911,
      "step": 1133
    },
    {
      "epoch": 567.0,
      "grad_norm": 1.8225774765014648,
      "learning_rate": 8.76978836179162e-06,
      "loss": 0.433,
      "step": 1134
    },
    {
      "epoch": 567.5,
      "grad_norm": 3.1174187660217285,
      "learning_rate": 8.706192949571262e-06,
      "loss": 0.5234,
      "step": 1135
    },
    {
      "epoch": 568.0,
      "grad_norm": 0.831122875213623,
      "learning_rate": 8.64280696599008e-06,
      "loss": 0.2715,
      "step": 1136
    },
    {
      "epoch": 568.5,
      "grad_norm": 0.30995166301727295,
      "learning_rate": 8.579630732520976e-06,
      "loss": 0.4308,
      "step": 1137
    },
    {
      "epoch": 569.0,
      "grad_norm": 0.46403107047080994,
      "learning_rate": 8.516664569573107e-06,
      "loss": 0.2637,
      "step": 1138
    },
    {
      "epoch": 569.5,
      "grad_norm": 0.41073691844940186,
      "learning_rate": 8.45390879649014e-06,
      "loss": 0.3747,
      "step": 1139
    },
    {
      "epoch": 570.0,
      "grad_norm": 0.6022565960884094,
      "learning_rate": 8.391363731548812e-06,
      "loss": 0.2797,
      "step": 1140
    },
    {
      "epoch": 570.5,
      "grad_norm": 0.29865241050720215,
      "learning_rate": 8.329029691957124e-06,
      "loss": 0.3187,
      "step": 1141
    },
    {
      "epoch": 571.0,
      "grad_norm": 1.3946223258972168,
      "learning_rate": 8.266906993852919e-06,
      "loss": 0.5184,
      "step": 1142
    },
    {
      "epoch": 571.5,
      "grad_norm": 0.2808627784252167,
      "learning_rate": 8.204995952302087e-06,
      "loss": 0.397,
      "step": 1143
    },
    {
      "epoch": 572.0,
      "grad_norm": 1.6276495456695557,
      "learning_rate": 8.143296881297147e-06,
      "loss": 1.0006,
      "step": 1144
    },
    {
      "epoch": 572.5,
      "grad_norm": 0.8327716588973999,
      "learning_rate": 8.081810093755538e-06,
      "loss": 0.3979,
      "step": 1145
    },
    {
      "epoch": 573.0,
      "grad_norm": 4.5547075271606445,
      "learning_rate": 8.02053590151805e-06,
      "loss": 0.8208,
      "step": 1146
    },
    {
      "epoch": 573.5,
      "grad_norm": 3.6351969242095947,
      "learning_rate": 7.959474615347318e-06,
      "loss": 0.4755,
      "step": 1147
    },
    {
      "epoch": 574.0,
      "grad_norm": 0.8956000208854675,
      "learning_rate": 7.898626544926102e-06,
      "loss": 0.2987,
      "step": 1148
    },
    {
      "epoch": 574.5,
      "grad_norm": 0.3933062255382538,
      "learning_rate": 7.837991998855897e-06,
      "loss": 0.4121,
      "step": 1149
    },
    {
      "epoch": 575.0,
      "grad_norm": 0.5727701187133789,
      "learning_rate": 7.777571284655177e-06,
      "loss": 0.2658,
      "step": 1150
    },
    {
      "epoch": 575.5,
      "grad_norm": 0.26141640543937683,
      "learning_rate": 7.717364708758024e-06,
      "loss": 0.4022,
      "step": 1151
    },
    {
      "epoch": 576.0,
      "grad_norm": 0.4763989746570587,
      "learning_rate": 7.657372576512383e-06,
      "loss": 0.2493,
      "step": 1152
    },
    {
      "epoch": 576.5,
      "grad_norm": 0.3561572730541229,
      "learning_rate": 7.597595192178702e-06,
      "loss": 0.3351,
      "step": 1153
    },
    {
      "epoch": 577.0,
      "grad_norm": 0.46987786889076233,
      "learning_rate": 7.538032858928251e-06,
      "loss": 0.2706,
      "step": 1154
    },
    {
      "epoch": 577.5,
      "grad_norm": 0.7194994688034058,
      "learning_rate": 7.47868587884163e-06,
      "loss": 0.3806,
      "step": 1155
    },
    {
      "epoch": 578.0,
      "grad_norm": 0.5587533116340637,
      "learning_rate": 7.41955455290726e-06,
      "loss": 0.2646,
      "step": 1156
    },
    {
      "epoch": 578.5,
      "grad_norm": 0.7956721186637878,
      "learning_rate": 7.360639181019813e-06,
      "loss": 0.496,
      "step": 1157
    },
    {
      "epoch": 579.0,
      "grad_norm": 1.1659671068191528,
      "learning_rate": 7.3019400619787236e-06,
      "loss": 0.3703,
      "step": 1158
    },
    {
      "epoch": 579.5,
      "grad_norm": 1.0656088590621948,
      "learning_rate": 7.24345749348666e-06,
      "loss": 0.5279,
      "step": 1159
    },
    {
      "epoch": 580.0,
      "grad_norm": 1.574374794960022,
      "learning_rate": 7.185191772148042e-06,
      "loss": 0.4469,
      "step": 1160
    },
    {
      "epoch": 580.5,
      "grad_norm": 1.7413619756698608,
      "learning_rate": 7.127143193467445e-06,
      "loss": 0.5501,
      "step": 1161
    },
    {
      "epoch": 581.0,
      "grad_norm": 1.1038923263549805,
      "learning_rate": 7.0693120518482645e-06,
      "loss": 0.4766,
      "step": 1162
    },
    {
      "epoch": 581.5,
      "grad_norm": 0.5392162799835205,
      "learning_rate": 7.011698640591024e-06,
      "loss": 0.4108,
      "step": 1163
    },
    {
      "epoch": 582.0,
      "grad_norm": 1.399778127670288,
      "learning_rate": 6.954303251892069e-06,
      "loss": 0.3837,
      "step": 1164
    },
    {
      "epoch": 582.5,
      "grad_norm": 0.3778377175331116,
      "learning_rate": 6.897126176841978e-06,
      "loss": 0.3313,
      "step": 1165
    },
    {
      "epoch": 583.0,
      "grad_norm": 0.49324139952659607,
      "learning_rate": 6.840167705424105e-06,
      "loss": 0.2814,
      "step": 1166
    },
    {
      "epoch": 583.5,
      "grad_norm": 0.6666603684425354,
      "learning_rate": 6.783428126513125e-06,
      "loss": 0.362,
      "step": 1167
    },
    {
      "epoch": 584.0,
      "grad_norm": 0.5295634865760803,
      "learning_rate": 6.726907727873572e-06,
      "loss": 0.2739,
      "step": 1168
    },
    {
      "epoch": 584.5,
      "grad_norm": 0.37020954489707947,
      "learning_rate": 6.670606796158352e-06,
      "loss": 0.3629,
      "step": 1169
    },
    {
      "epoch": 585.0,
      "grad_norm": 0.528693437576294,
      "learning_rate": 6.6145256169073185e-06,
      "loss": 0.2599,
      "step": 1170
    },
    {
      "epoch": 585.5,
      "grad_norm": 0.7501449584960938,
      "learning_rate": 6.558664474545817e-06,
      "loss": 0.3316,
      "step": 1171
    },
    {
      "epoch": 586.0,
      "grad_norm": 0.5042018890380859,
      "learning_rate": 6.503023652383211e-06,
      "loss": 0.2958,
      "step": 1172
    },
    {
      "epoch": 586.5,
      "grad_norm": 0.5048187971115112,
      "learning_rate": 6.447603432611532e-06,
      "loss": 0.3468,
      "step": 1173
    },
    {
      "epoch": 587.0,
      "grad_norm": 0.6499981880187988,
      "learning_rate": 6.392404096303906e-06,
      "loss": 0.2805,
      "step": 1174
    },
    {
      "epoch": 587.5,
      "grad_norm": 0.28584927320480347,
      "learning_rate": 6.337425923413276e-06,
      "loss": 0.2895,
      "step": 1175
    },
    {
      "epoch": 588.0,
      "grad_norm": 1.0394264459609985,
      "learning_rate": 6.282669192770896e-06,
      "loss": 0.3297,
      "step": 1176
    },
    {
      "epoch": 588.5,
      "grad_norm": 0.32219651341438293,
      "learning_rate": 6.228134182084927e-06,
      "loss": 0.4279,
      "step": 1177
    },
    {
      "epoch": 589.0,
      "grad_norm": 0.46984800696372986,
      "learning_rate": 6.173821167939053e-06,
      "loss": 0.2534,
      "step": 1178
    },
    {
      "epoch": 589.5,
      "grad_norm": 5.781923770904541,
      "learning_rate": 6.119730425791059e-06,
      "loss": 0.354,
      "step": 1179
    },
    {
      "epoch": 590.0,
      "grad_norm": 1.857688546180725,
      "learning_rate": 6.065862229971437e-06,
      "loss": 0.4509,
      "step": 1180
    },
    {
      "epoch": 590.5,
      "grad_norm": 0.2846425175666809,
      "learning_rate": 6.012216853682001e-06,
      "loss": 0.3414,
      "step": 1181
    },
    {
      "epoch": 591.0,
      "grad_norm": 1.7315561771392822,
      "learning_rate": 5.9587945689945025e-06,
      "loss": 0.5269,
      "step": 1182
    },
    {
      "epoch": 591.5,
      "grad_norm": 0.30941149592399597,
      "learning_rate": 5.905595646849227e-06,
      "loss": 0.3838,
      "step": 1183
    },
    {
      "epoch": 592.0,
      "grad_norm": 0.8480091691017151,
      "learning_rate": 5.852620357053651e-06,
      "loss": 0.266,
      "step": 1184
    },
    {
      "epoch": 592.5,
      "grad_norm": 0.26204994320869446,
      "learning_rate": 5.7998689682810746e-06,
      "loss": 0.4148,
      "step": 1185
    },
    {
      "epoch": 593.0,
      "grad_norm": 0.5316544771194458,
      "learning_rate": 5.747341748069229e-06,
      "loss": 0.2899,
      "step": 1186
    },
    {
      "epoch": 593.5,
      "grad_norm": 0.5450106859207153,
      "learning_rate": 5.695038962818933e-06,
      "loss": 0.383,
      "step": 1187
    },
    {
      "epoch": 594.0,
      "grad_norm": 0.8020029067993164,
      "learning_rate": 5.642960877792752e-06,
      "loss": 0.2932,
      "step": 1188
    },
    {
      "epoch": 594.5,
      "grad_norm": 0.2605772912502289,
      "learning_rate": 5.591107757113645e-06,
      "loss": 0.3488,
      "step": 1189
    },
    {
      "epoch": 595.0,
      "grad_norm": 0.868437647819519,
      "learning_rate": 5.539479863763624e-06,
      "loss": 0.3417,
      "step": 1190
    },
    {
      "epoch": 595.5,
      "grad_norm": 6.5039238929748535,
      "learning_rate": 5.488077459582425e-06,
      "loss": 0.3398,
      "step": 1191
    },
    {
      "epoch": 596.0,
      "grad_norm": 0.6516555547714233,
      "learning_rate": 5.436900805266171e-06,
      "loss": 0.3156,
      "step": 1192
    },
    {
      "epoch": 596.5,
      "grad_norm": 0.27916818857192993,
      "learning_rate": 5.385950160366066e-06,
      "loss": 0.3285,
      "step": 1193
    },
    {
      "epoch": 597.0,
      "grad_norm": 0.4588022232055664,
      "learning_rate": 5.335225783287051e-06,
      "loss": 0.2575,
      "step": 1194
    },
    {
      "epoch": 597.5,
      "grad_norm": 0.538681149482727,
      "learning_rate": 5.284727931286526e-06,
      "loss": 0.3606,
      "step": 1195
    },
    {
      "epoch": 598.0,
      "grad_norm": 2.8163933753967285,
      "learning_rate": 5.234456860473042e-06,
      "loss": 0.4859,
      "step": 1196
    },
    {
      "epoch": 598.5,
      "grad_norm": 0.37342265248298645,
      "learning_rate": 5.1844128258049465e-06,
      "loss": 0.4177,
      "step": 1197
    },
    {
      "epoch": 599.0,
      "grad_norm": 0.6619336605072021,
      "learning_rate": 5.134596081089183e-06,
      "loss": 0.2779,
      "step": 1198
    },
    {
      "epoch": 599.5,
      "grad_norm": 0.30484306812286377,
      "learning_rate": 5.085006878979926e-06,
      "loss": 0.4449,
      "step": 1199
    },
    {
      "epoch": 600.0,
      "grad_norm": 1.492018222808838,
      "learning_rate": 5.03564547097734e-06,
      "loss": 0.6049,
      "step": 1200
    },
    {
      "epoch": 600.5,
      "grad_norm": 0.9707176685333252,
      "learning_rate": 4.986512107426283e-06,
      "loss": 0.3695,
      "step": 1201
    },
    {
      "epoch": 601.0,
      "grad_norm": 0.5508577823638916,
      "learning_rate": 4.937607037515057e-06,
      "loss": 0.2743,
      "step": 1202
    },
    {
      "epoch": 601.5,
      "grad_norm": 0.2802048325538635,
      "learning_rate": 4.888930509274126e-06,
      "loss": 0.3575,
      "step": 1203
    },
    {
      "epoch": 602.0,
      "grad_norm": 2.079087972640991,
      "learning_rate": 4.840482769574861e-06,
      "loss": 0.7976,
      "step": 1204
    },
    {
      "epoch": 602.5,
      "grad_norm": 0.23147372901439667,
      "learning_rate": 4.792264064128327e-06,
      "loss": 0.3581,
      "step": 1205
    },
    {
      "epoch": 603.0,
      "grad_norm": 1.0730218887329102,
      "learning_rate": 4.744274637483936e-06,
      "loss": 0.3138,
      "step": 1206
    },
    {
      "epoch": 603.5,
      "grad_norm": 0.8075739145278931,
      "learning_rate": 4.696514733028357e-06,
      "loss": 0.473,
      "step": 1207
    },
    {
      "epoch": 604.0,
      "grad_norm": 0.7302966713905334,
      "learning_rate": 4.648984592984112e-06,
      "loss": 0.2429,
      "step": 1208
    },
    {
      "epoch": 604.5,
      "grad_norm": 0.27118468284606934,
      "learning_rate": 4.601684458408506e-06,
      "loss": 0.4207,
      "step": 1209
    },
    {
      "epoch": 605.0,
      "grad_norm": 1.341869592666626,
      "learning_rate": 4.5546145691922725e-06,
      "loss": 0.5257,
      "step": 1210
    },
    {
      "epoch": 605.5,
      "grad_norm": 7.416314125061035,
      "learning_rate": 4.507775164058469e-06,
      "loss": 0.6631,
      "step": 1211
    },
    {
      "epoch": 606.0,
      "grad_norm": 0.4223253130912781,
      "learning_rate": 4.46116648056118e-06,
      "loss": 0.2613,
      "step": 1212
    },
    {
      "epoch": 606.5,
      "grad_norm": 0.4507334530353546,
      "learning_rate": 4.414788755084348e-06,
      "loss": 0.4246,
      "step": 1213
    },
    {
      "epoch": 607.0,
      "grad_norm": 0.7952383160591125,
      "learning_rate": 4.368642222840585e-06,
      "loss": 0.316,
      "step": 1214
    },
    {
      "epoch": 607.5,
      "grad_norm": 0.2478264570236206,
      "learning_rate": 4.322727117869951e-06,
      "loss": 0.3359,
      "step": 1215
    },
    {
      "epoch": 608.0,
      "grad_norm": 0.7629574537277222,
      "learning_rate": 4.277043673038816e-06,
      "loss": 0.2851,
      "step": 1216
    },
    {
      "epoch": 608.5,
      "grad_norm": 0.31850215792655945,
      "learning_rate": 4.231592120038585e-06,
      "loss": 0.3685,
      "step": 1217
    },
    {
      "epoch": 609.0,
      "grad_norm": 1.3444807529449463,
      "learning_rate": 4.186372689384654e-06,
      "loss": 0.4876,
      "step": 1218
    },
    {
      "epoch": 609.5,
      "grad_norm": 0.2763107120990753,
      "learning_rate": 4.141385610415105e-06,
      "loss": 0.3674,
      "step": 1219
    },
    {
      "epoch": 610.0,
      "grad_norm": 0.5751998424530029,
      "learning_rate": 4.096631111289678e-06,
      "loss": 0.3032,
      "step": 1220
    },
    {
      "epoch": 610.5,
      "grad_norm": 0.2718237638473511,
      "learning_rate": 4.05210941898847e-06,
      "loss": 0.4133,
      "step": 1221
    },
    {
      "epoch": 611.0,
      "grad_norm": 1.536370038986206,
      "learning_rate": 4.007820759310921e-06,
      "loss": 0.4649,
      "step": 1222
    },
    {
      "epoch": 611.5,
      "grad_norm": 0.2352786511182785,
      "learning_rate": 3.963765356874566e-06,
      "loss": 0.3518,
      "step": 1223
    },
    {
      "epoch": 612.0,
      "grad_norm": 0.7556079626083374,
      "learning_rate": 3.919943435113954e-06,
      "loss": 0.3113,
      "step": 1224
    },
    {
      "epoch": 612.5,
      "grad_norm": 0.30604568123817444,
      "learning_rate": 3.876355216279498e-06,
      "loss": 0.2886,
      "step": 1225
    },
    {
      "epoch": 613.0,
      "grad_norm": 0.5049639940261841,
      "learning_rate": 3.833000921436319e-06,
      "loss": 0.2671,
      "step": 1226
    },
    {
      "epoch": 613.5,
      "grad_norm": 1.8689523935317993,
      "learning_rate": 3.7898807704632167e-06,
      "loss": 0.5063,
      "step": 1227
    },
    {
      "epoch": 614.0,
      "grad_norm": 1.5286250114440918,
      "learning_rate": 3.7469949820513993e-06,
      "loss": 0.6401,
      "step": 1228
    },
    {
      "epoch": 614.5,
      "grad_norm": 1.3831666707992554,
      "learning_rate": 3.7043437737035503e-06,
      "loss": 0.5887,
      "step": 1229
    },
    {
      "epoch": 615.0,
      "grad_norm": 1.5179754495620728,
      "learning_rate": 3.66192736173257e-06,
      "loss": 0.8891,
      "step": 1230
    },
    {
      "epoch": 615.5,
      "grad_norm": 0.2437417060136795,
      "learning_rate": 3.619745961260623e-06,
      "loss": 0.3524,
      "step": 1231
    },
    {
      "epoch": 616.0,
      "grad_norm": 0.5126246809959412,
      "learning_rate": 3.5777997862178837e-06,
      "loss": 0.2674,
      "step": 1232
    },
    {
      "epoch": 616.5,
      "grad_norm": 0.24882115423679352,
      "learning_rate": 3.536089049341623e-06,
      "loss": 0.3614,
      "step": 1233
    },
    {
      "epoch": 617.0,
      "grad_norm": 0.6144912838935852,
      "learning_rate": 3.4946139621749974e-06,
      "loss": 0.2714,
      "step": 1234
    },
    {
      "epoch": 617.5,
      "grad_norm": 0.5393758416175842,
      "learning_rate": 3.453374735066034e-06,
      "loss": 0.3857,
      "step": 1235
    },
    {
      "epoch": 618.0,
      "grad_norm": 1.9871344566345215,
      "learning_rate": 3.4123715771665786e-06,
      "loss": 0.5247,
      "step": 1236
    },
    {
      "epoch": 618.5,
      "grad_norm": 2.2139732837677,
      "learning_rate": 3.371604696431174e-06,
      "loss": 0.4481,
      "step": 1237
    },
    {
      "epoch": 619.0,
      "grad_norm": 0.6167898774147034,
      "learning_rate": 3.3310742996160824e-06,
      "loss": 0.2579,
      "step": 1238
    },
    {
      "epoch": 619.5,
      "grad_norm": 1.643285870552063,
      "learning_rate": 3.2907805922781476e-06,
      "loss": 0.6078,
      "step": 1239
    },
    {
      "epoch": 620.0,
      "grad_norm": 0.47234612703323364,
      "learning_rate": 3.2507237787738676e-06,
      "loss": 0.2534,
      "step": 1240
    },
    {
      "epoch": 620.5,
      "grad_norm": 0.3556380271911621,
      "learning_rate": 3.2109040622582186e-06,
      "loss": 0.4078,
      "step": 1241
    },
    {
      "epoch": 621.0,
      "grad_norm": 0.490627259016037,
      "learning_rate": 3.171321644683761e-06,
      "loss": 0.2422,
      "step": 1242
    },
    {
      "epoch": 621.5,
      "grad_norm": 0.3128645718097687,
      "learning_rate": 3.1319767267995236e-06,
      "loss": 0.3562,
      "step": 1243
    },
    {
      "epoch": 622.0,
      "grad_norm": 16.325393676757812,
      "learning_rate": 3.09286950815002e-06,
      "loss": 1.71,
      "step": 1244
    },
    {
      "epoch": 622.5,
      "grad_norm": 0.4415295124053955,
      "learning_rate": 3.054000187074224e-06,
      "loss": 0.4385,
      "step": 1245
    },
    {
      "epoch": 623.0,
      "grad_norm": 1.0011006593704224,
      "learning_rate": 3.0153689607045845e-06,
      "loss": 0.3607,
      "step": 1246
    },
    {
      "epoch": 623.5,
      "grad_norm": 0.461182564496994,
      "learning_rate": 2.9769760249659995e-06,
      "loss": 0.4444,
      "step": 1247
    },
    {
      "epoch": 624.0,
      "grad_norm": 0.4573282301425934,
      "learning_rate": 2.938821574574835e-06,
      "loss": 0.2705,
      "step": 1248
    },
    {
      "epoch": 624.5,
      "grad_norm": 0.3119376599788666,
      "learning_rate": 2.900905803037968e-06,
      "loss": 0.3822,
      "step": 1249
    },
    {
      "epoch": 625.0,
      "grad_norm": 0.46923312544822693,
      "learning_rate": 2.863228902651721e-06,
      "loss": 0.2717,
      "step": 1250
    },
    {
      "epoch": 625.5,
      "grad_norm": 1.7543494701385498,
      "learning_rate": 2.825791064500993e-06,
      "loss": 0.4766,
      "step": 1251
    },
    {
      "epoch": 626.0,
      "grad_norm": 1.4127765893936157,
      "learning_rate": 2.7885924784581905e-06,
      "loss": 0.3946,
      "step": 1252
    },
    {
      "epoch": 626.5,
      "grad_norm": 0.6524680256843567,
      "learning_rate": 2.7516333331823575e-06,
      "loss": 0.3899,
      "step": 1253
    },
    {
      "epoch": 627.0,
      "grad_norm": 0.9214707016944885,
      "learning_rate": 2.714913816118142e-06,
      "loss": 0.347,
      "step": 1254
    },
    {
      "epoch": 627.5,
      "grad_norm": 0.2354903370141983,
      "learning_rate": 2.678434113494882e-06,
      "loss": 0.3143,
      "step": 1255
    },
    {
      "epoch": 628.0,
      "grad_norm": 0.7629495859146118,
      "learning_rate": 2.6421944103256657e-06,
      "loss": 0.3531,
      "step": 1256
    },
    {
      "epoch": 628.5,
      "grad_norm": 0.33466026186943054,
      "learning_rate": 2.606194890406366e-06,
      "loss": 0.3377,
      "step": 1257
    },
    {
      "epoch": 629.0,
      "grad_norm": 0.9416338801383972,
      "learning_rate": 2.570435736314747e-06,
      "loss": 0.333,
      "step": 1258
    },
    {
      "epoch": 629.5,
      "grad_norm": 1.765978217124939,
      "learning_rate": 2.5349171294094975e-06,
      "loss": 0.4167,
      "step": 1259
    },
    {
      "epoch": 630.0,
      "grad_norm": 1.5732077360153198,
      "learning_rate": 2.499639249829333e-06,
      "loss": 1.0772,
      "step": 1260
    },
    {
      "epoch": 630.5,
      "grad_norm": 0.1897096484899521,
      "learning_rate": 2.464602276492084e-06,
      "loss": 0.2933,
      "step": 1261
    },
    {
      "epoch": 631.0,
      "grad_norm": 0.5945242643356323,
      "learning_rate": 2.4298063870938025e-06,
      "loss": 0.2484,
      "step": 1262
    },
    {
      "epoch": 631.5,
      "grad_norm": 0.2862352728843689,
      "learning_rate": 2.395251758107786e-06,
      "loss": 0.4434,
      "step": 1263
    },
    {
      "epoch": 632.0,
      "grad_norm": 0.6925152540206909,
      "learning_rate": 2.360938564783799e-06,
      "loss": 0.2707,
      "step": 1264
    },
    {
      "epoch": 632.5,
      "grad_norm": 0.8513469099998474,
      "learning_rate": 2.326866981147091e-06,
      "loss": 0.3542,
      "step": 1265
    },
    {
      "epoch": 633.0,
      "grad_norm": 0.5723398327827454,
      "learning_rate": 2.2930371799975594e-06,
      "loss": 0.2755,
      "step": 1266
    },
    {
      "epoch": 633.5,
      "grad_norm": 2.3658180236816406,
      "learning_rate": 2.2594493329088474e-06,
      "loss": 0.7395,
      "step": 1267
    },
    {
      "epoch": 634.0,
      "grad_norm": 0.6319108009338379,
      "learning_rate": 2.2261036102274977e-06,
      "loss": 0.2619,
      "step": 1268
    },
    {
      "epoch": 634.5,
      "grad_norm": 0.24553509056568146,
      "learning_rate": 2.193000181072069e-06,
      "loss": 0.443,
      "step": 1269
    },
    {
      "epoch": 635.0,
      "grad_norm": 1.732064962387085,
      "learning_rate": 2.1601392133322962e-06,
      "loss": 0.8924,
      "step": 1270
    },
    {
      "epoch": 635.5,
      "grad_norm": 0.2352485954761505,
      "learning_rate": 2.127520873668226e-06,
      "loss": 0.3984,
      "step": 1271
    },
    {
      "epoch": 636.0,
      "grad_norm": 39.863914489746094,
      "learning_rate": 2.095145327509368e-06,
      "loss": 1.9335,
      "step": 1272
    },
    {
      "epoch": 636.5,
      "grad_norm": 0.37388208508491516,
      "learning_rate": 2.0630127390538655e-06,
      "loss": 0.344,
      "step": 1273
    },
    {
      "epoch": 637.0,
      "grad_norm": 0.5663019418716431,
      "learning_rate": 2.0311232712676765e-06,
      "loss": 0.2539,
      "step": 1274
    },
    {
      "epoch": 637.5,
      "grad_norm": 0.3063144087791443,
      "learning_rate": 1.999477085883711e-06,
      "loss": 0.4231,
      "step": 1275
    },
    {
      "epoch": 638.0,
      "grad_norm": 0.4968956708908081,
      "learning_rate": 1.9680743434010384e-06,
      "loss": 0.2408,
      "step": 1276
    },
    {
      "epoch": 638.5,
      "grad_norm": 0.5581031441688538,
      "learning_rate": 1.9369152030840556e-06,
      "loss": 0.4181,
      "step": 1277
    },
    {
      "epoch": 639.0,
      "grad_norm": 21.25269317626953,
      "learning_rate": 1.905999822961707e-06,
      "loss": 1.1821,
      "step": 1278
    },
    {
      "epoch": 639.5,
      "grad_norm": 0.8949183821678162,
      "learning_rate": 1.8753283598266558e-06,
      "loss": 0.4662,
      "step": 1279
    },
    {
      "epoch": 640.0,
      "grad_norm": 0.5083120465278625,
      "learning_rate": 1.844900969234492e-06,
      "loss": 0.2637,
      "step": 1280
    },
    {
      "epoch": 640.5,
      "grad_norm": 0.30995485186576843,
      "learning_rate": 1.8147178055029579e-06,
      "loss": 0.3297,
      "step": 1281
    },
    {
      "epoch": 641.0,
      "grad_norm": 0.45998522639274597,
      "learning_rate": 1.7847790217111636e-06,
      "loss": 0.2463,
      "step": 1282
    },
    {
      "epoch": 641.5,
      "grad_norm": 0.28498056530952454,
      "learning_rate": 1.7550847696987838e-06,
      "loss": 0.3782,
      "step": 1283
    },
    {
      "epoch": 642.0,
      "grad_norm": 1.854949951171875,
      "learning_rate": 1.725635200065323e-06,
      "loss": 0.7288,
      "step": 1284
    },
    {
      "epoch": 642.5,
      "grad_norm": 0.8054349422454834,
      "learning_rate": 1.6964304621693517e-06,
      "loss": 0.5488,
      "step": 1285
    },
    {
      "epoch": 643.0,
      "grad_norm": 0.6534746885299683,
      "learning_rate": 1.6674707041276938e-06,
      "loss": 0.2798,
      "step": 1286
    },
    {
      "epoch": 643.5,
      "grad_norm": 0.30258724093437195,
      "learning_rate": 1.638756072814751e-06,
      "loss": 0.3603,
      "step": 1287
    },
    {
      "epoch": 644.0,
      "grad_norm": 0.7822498083114624,
      "learning_rate": 1.610286713861714e-06,
      "loss": 0.3271,
      "step": 1288
    },
    {
      "epoch": 644.5,
      "grad_norm": 0.8773853182792664,
      "learning_rate": 1.5820627716558233e-06,
      "loss": 0.362,
      "step": 1289
    },
    {
      "epoch": 645.0,
      "grad_norm": 1.2205160856246948,
      "learning_rate": 1.554084389339655e-06,
      "loss": 0.4103,
      "step": 1290
    },
    {
      "epoch": 645.5,
      "grad_norm": 0.3368886113166809,
      "learning_rate": 1.5263517088103863e-06,
      "loss": 0.3514,
      "step": 1291
    },
    {
      "epoch": 646.0,
      "grad_norm": 1.4454164505004883,
      "learning_rate": 1.4988648707190634e-06,
      "loss": 0.4095,
      "step": 1292
    },
    {
      "epoch": 646.5,
      "grad_norm": 0.1670636087656021,
      "learning_rate": 1.471624014469919e-06,
      "loss": 0.284,
      "step": 1293
    },
    {
      "epoch": 647.0,
      "grad_norm": 0.655076265335083,
      "learning_rate": 1.444629278219628e-06,
      "loss": 0.3017,
      "step": 1294
    },
    {
      "epoch": 647.5,
      "grad_norm": 0.2972862124443054,
      "learning_rate": 1.4178807988766419e-06,
      "loss": 0.3647,
      "step": 1295
    },
    {
      "epoch": 648.0,
      "grad_norm": 1.9007000923156738,
      "learning_rate": 1.3913787121004719e-06,
      "loss": 0.7543,
      "step": 1296
    },
    {
      "epoch": 648.5,
      "grad_norm": 0.33977222442626953,
      "learning_rate": 1.3651231523009955e-06,
      "loss": 0.3384,
      "step": 1297
    },
    {
      "epoch": 649.0,
      "grad_norm": 0.5207720398902893,
      "learning_rate": 1.3391142526378131e-06,
      "loss": 0.2946,
      "step": 1298
    },
    {
      "epoch": 649.5,
      "grad_norm": 0.5146612524986267,
      "learning_rate": 1.3133521450195086e-06,
      "loss": 0.3655,
      "step": 1299
    },
    {
      "epoch": 650.0,
      "grad_norm": 1.0224529504776,
      "learning_rate": 1.287836960103056e-06,
      "loss": 0.3197,
      "step": 1300
    },
    {
      "epoch": 650.5,
      "grad_norm": 0.31942233443260193,
      "learning_rate": 1.2625688272930926e-06,
      "loss": 0.4195,
      "step": 1301
    },
    {
      "epoch": 651.0,
      "grad_norm": 0.6055216193199158,
      "learning_rate": 1.2375478747413017e-06,
      "loss": 0.2808,
      "step": 1302
    },
    {
      "epoch": 651.5,
      "grad_norm": 0.8240653276443481,
      "learning_rate": 1.2127742293457312e-06,
      "loss": 0.5575,
      "step": 1303
    },
    {
      "epoch": 652.0,
      "grad_norm": 0.6095658540725708,
      "learning_rate": 1.188248016750193e-06,
      "loss": 0.264,
      "step": 1304
    },
    {
      "epoch": 652.5,
      "grad_norm": 1.572292685508728,
      "learning_rate": 1.163969361343592e-06,
      "loss": 0.386,
      "step": 1305
    },
    {
      "epoch": 653.0,
      "grad_norm": 0.5676502585411072,
      "learning_rate": 1.1399383862592927e-06,
      "loss": 0.2509,
      "step": 1306
    },
    {
      "epoch": 653.5,
      "grad_norm": 0.2616370618343353,
      "learning_rate": 1.1161552133745201e-06,
      "loss": 0.3769,
      "step": 1307
    },
    {
      "epoch": 654.0,
      "grad_norm": 0.5516897439956665,
      "learning_rate": 1.0926199633097157e-06,
      "loss": 0.2769,
      "step": 1308
    },
    {
      "epoch": 654.5,
      "grad_norm": 0.2853024899959564,
      "learning_rate": 1.0693327554279598e-06,
      "loss": 0.3763,
      "step": 1309
    },
    {
      "epoch": 655.0,
      "grad_norm": 1.1394164562225342,
      "learning_rate": 1.0462937078343116e-06,
      "loss": 0.3303,
      "step": 1310
    },
    {
      "epoch": 655.5,
      "grad_norm": 0.26393836736679077,
      "learning_rate": 1.0235029373752758e-06,
      "loss": 0.3776,
      "step": 1311
    },
    {
      "epoch": 656.0,
      "grad_norm": 22.04922866821289,
      "learning_rate": 1.0009605596381534e-06,
      "loss": 1.6696,
      "step": 1312
    },
    {
      "epoch": 656.5,
      "grad_norm": 1.0814285278320312,
      "learning_rate": 9.786666889504859e-07,
      "loss": 0.4845,
      "step": 1313
    },
    {
      "epoch": 657.0,
      "grad_norm": 0.7056344747543335,
      "learning_rate": 9.566214383794736e-07,
      "loss": 0.2862,
      "step": 1314
    },
    {
      "epoch": 657.5,
      "grad_norm": 2.1829237937927246,
      "learning_rate": 9.348249197313919e-07,
      "loss": 0.4779,
      "step": 1315
    },
    {
      "epoch": 658.0,
      "grad_norm": 0.4273681938648224,
      "learning_rate": 9.132772435510362e-07,
      "loss": 0.2495,
      "step": 1316
    },
    {
      "epoch": 658.5,
      "grad_norm": 2.198198080062866,
      "learning_rate": 8.919785191211394e-07,
      "loss": 0.4886,
      "step": 1317
    },
    {
      "epoch": 659.0,
      "grad_norm": 1.2046173810958862,
      "learning_rate": 8.709288544618555e-07,
      "loss": 0.4499,
      "step": 1318
    },
    {
      "epoch": 659.5,
      "grad_norm": 0.5892188549041748,
      "learning_rate": 8.501283563301599e-07,
      "loss": 0.3737,
      "step": 1319
    },
    {
      "epoch": 660.0,
      "grad_norm": 2.018260955810547,
      "learning_rate": 8.295771302193722e-07,
      "loss": 0.4783,
      "step": 1320
    },
    {
      "epoch": 660.5,
      "grad_norm": 0.3263697028160095,
      "learning_rate": 8.092752803585513e-07,
      "loss": 0.3715,
      "step": 1321
    },
    {
      "epoch": 661.0,
      "grad_norm": 1.3145312070846558,
      "learning_rate": 7.89222909712023e-07,
      "loss": 0.6621,
      "step": 1322
    },
    {
      "epoch": 661.5,
      "grad_norm": 0.24664939939975739,
      "learning_rate": 7.69420119978842e-07,
      "loss": 0.3539,
      "step": 1323
    },
    {
      "epoch": 662.0,
      "grad_norm": 1.21776282787323,
      "learning_rate": 7.498670115922424e-07,
      "loss": 0.3488,
      "step": 1324
    },
    {
      "epoch": 662.5,
      "grad_norm": 0.37903350591659546,
      "learning_rate": 7.305636837191876e-07,
      "loss": 0.3421,
      "step": 1325
    },
    {
      "epoch": 663.0,
      "grad_norm": 0.5102131962776184,
      "learning_rate": 7.115102342598101e-07,
      "loss": 0.2507,
      "step": 1326
    },
    {
      "epoch": 663.5,
      "grad_norm": 1.527514100074768,
      "learning_rate": 6.927067598469672e-07,
      "loss": 0.5747,
      "step": 1327
    },
    {
      "epoch": 664.0,
      "grad_norm": 1.472798466682434,
      "learning_rate": 6.741533558456969e-07,
      "loss": 0.5742,
      "step": 1328
    },
    {
      "epoch": 664.5,
      "grad_norm": 1.0608159303665161,
      "learning_rate": 6.558501163527964e-07,
      "loss": 0.4761,
      "step": 1329
    },
    {
      "epoch": 665.0,
      "grad_norm": 0.7804026007652283,
      "learning_rate": 6.377971341962607e-07,
      "loss": 0.3191,
      "step": 1330
    },
    {
      "epoch": 665.5,
      "grad_norm": 0.483767569065094,
      "learning_rate": 6.199945009349173e-07,
      "loss": 0.3136,
      "step": 1331
    },
    {
      "epoch": 666.0,
      "grad_norm": 0.8419970273971558,
      "learning_rate": 6.0244230685787e-07,
      "loss": 0.3487,
      "step": 1332
    },
    {
      "epoch": 666.5,
      "grad_norm": 0.2199861705303192,
      "learning_rate": 5.851406409840888e-07,
      "loss": 0.335,
      "step": 1333
    },
    {
      "epoch": 667.0,
      "grad_norm": 1.8586952686309814,
      "learning_rate": 5.680895910619433e-07,
      "loss": 0.3682,
      "step": 1334
    },
    {
      "epoch": 667.5,
      "grad_norm": 0.29940807819366455,
      "learning_rate": 5.512892435687645e-07,
      "loss": 0.391,
      "step": 1335
    },
    {
      "epoch": 668.0,
      "grad_norm": 0.49157649278640747,
      "learning_rate": 5.347396837104057e-07,
      "loss": 0.256,
      "step": 1336
    },
    {
      "epoch": 668.5,
      "grad_norm": 0.21423356235027313,
      "learning_rate": 5.184409954207991e-07,
      "loss": 0.341,
      "step": 1337
    },
    {
      "epoch": 669.0,
      "grad_norm": 1.649492859840393,
      "learning_rate": 5.023932613615445e-07,
      "loss": 0.7191,
      "step": 1338
    },
    {
      "epoch": 669.5,
      "grad_norm": 0.21800589561462402,
      "learning_rate": 4.865965629214819e-07,
      "loss": 0.3285,
      "step": 1339
    },
    {
      "epoch": 670.0,
      "grad_norm": 0.911331057548523,
      "learning_rate": 4.710509802162921e-07,
      "loss": 0.3469,
      "step": 1340
    },
    {
      "epoch": 670.5,
      "grad_norm": 0.35991451144218445,
      "learning_rate": 4.55756592088058e-07,
      "loss": 0.3943,
      "step": 1341
    },
    {
      "epoch": 671.0,
      "grad_norm": 0.5479896664619446,
      "learning_rate": 4.407134761048981e-07,
      "loss": 0.2726,
      "step": 1342
    },
    {
      "epoch": 671.5,
      "grad_norm": 0.2348278909921646,
      "learning_rate": 4.2592170856056714e-07,
      "loss": 0.3434,
      "step": 1343
    },
    {
      "epoch": 672.0,
      "grad_norm": 0.4538048803806305,
      "learning_rate": 4.1138136447405606e-07,
      "loss": 0.2342,
      "step": 1344
    },
    {
      "epoch": 672.5,
      "grad_norm": 0.33911028504371643,
      "learning_rate": 3.970925175892093e-07,
      "loss": 0.3688,
      "step": 1345
    },
    {
      "epoch": 673.0,
      "grad_norm": 0.7749882340431213,
      "learning_rate": 3.8305524037438034e-07,
      "loss": 0.2564,
      "step": 1346
    },
    {
      "epoch": 673.5,
      "grad_norm": 0.18411526083946228,
      "learning_rate": 3.692696040220267e-07,
      "loss": 0.3228,
      "step": 1347
    },
    {
      "epoch": 674.0,
      "grad_norm": 0.4457007944583893,
      "learning_rate": 3.557356784483712e-07,
      "loss": 0.248,
      "step": 1348
    },
    {
      "epoch": 674.5,
      "grad_norm": 0.24026712775230408,
      "learning_rate": 3.424535322930522e-07,
      "loss": 0.3356,
      "step": 1349
    },
    {
      "epoch": 675.0,
      "grad_norm": 0.3915645182132721,
      "learning_rate": 3.2942323291874076e-07,
      "loss": 0.2488,
      "step": 1350
    },
    {
      "epoch": 675.5,
      "grad_norm": 0.8330556154251099,
      "learning_rate": 3.1664484641086287e-07,
      "loss": 0.4177,
      "step": 1351
    },
    {
      "epoch": 676.0,
      "grad_norm": 0.6467349529266357,
      "learning_rate": 3.041184375771777e-07,
      "loss": 0.2896,
      "step": 1352
    },
    {
      "epoch": 676.5,
      "grad_norm": 0.18735870718955994,
      "learning_rate": 2.9184406994753334e-07,
      "loss": 0.3114,
      "step": 1353
    },
    {
      "epoch": 677.0,
      "grad_norm": 2.297031879425049,
      "learning_rate": 2.798218057734947e-07,
      "loss": 0.9512,
      "step": 1354
    },
    {
      "epoch": 677.5,
      "grad_norm": 0.2191716879606247,
      "learning_rate": 2.68051706028033e-07,
      "loss": 0.3562,
      "step": 1355
    },
    {
      "epoch": 678.0,
      "grad_norm": 0.6837000846862793,
      "learning_rate": 2.5653383040524227e-07,
      "loss": 0.2828,
      "step": 1356
    },
    {
      "epoch": 678.5,
      "grad_norm": 1.4173405170440674,
      "learning_rate": 2.4526823732000104e-07,
      "loss": 0.4227,
      "step": 1357
    },
    {
      "epoch": 679.0,
      "grad_norm": 0.5415307283401489,
      "learning_rate": 2.3425498390770573e-07,
      "loss": 0.2509,
      "step": 1358
    },
    {
      "epoch": 679.5,
      "grad_norm": 0.2828369140625,
      "learning_rate": 2.234941260239598e-07,
      "loss": 0.5046,
      "step": 1359
    },
    {
      "epoch": 680.0,
      "grad_norm": 0.4111093282699585,
      "learning_rate": 2.129857182443018e-07,
      "loss": 0.24,
      "step": 1360
    },
    {
      "epoch": 680.5,
      "grad_norm": 0.2846771776676178,
      "learning_rate": 2.027298138639333e-07,
      "loss": 0.3662,
      "step": 1361
    },
    {
      "epoch": 681.0,
      "grad_norm": 0.4653342664241791,
      "learning_rate": 1.9272646489743029e-07,
      "loss": 0.2735,
      "step": 1362
    },
    {
      "epoch": 681.5,
      "grad_norm": 0.3511224091053009,
      "learning_rate": 1.829757220784989e-07,
      "loss": 0.4753,
      "step": 1363
    },
    {
      "epoch": 682.0,
      "grad_norm": 1.3656400442123413,
      "learning_rate": 1.734776348597089e-07,
      "loss": 0.6139,
      "step": 1364
    },
    {
      "epoch": 682.5,
      "grad_norm": 2.3750061988830566,
      "learning_rate": 1.6423225141223852e-07,
      "loss": 0.3289,
      "step": 1365
    },
    {
      "epoch": 683.0,
      "grad_norm": 0.726895809173584,
      "learning_rate": 1.5523961862564106e-07,
      "loss": 0.2698,
      "step": 1366
    },
    {
      "epoch": 683.5,
      "grad_norm": 0.22589653730392456,
      "learning_rate": 1.4649978210759528e-07,
      "loss": 0.3594,
      "step": 1367
    },
    {
      "epoch": 684.0,
      "grad_norm": 0.48099565505981445,
      "learning_rate": 1.380127861836944e-07,
      "loss": 0.2561,
      "step": 1368
    },
    {
      "epoch": 684.5,
      "grad_norm": 0.14806623756885529,
      "learning_rate": 1.2977867389718512e-07,
      "loss": 0.2767,
      "step": 1369
    },
    {
      "epoch": 685.0,
      "grad_norm": 0.5614773631095886,
      "learning_rate": 1.2179748700879012e-07,
      "loss": 0.2568,
      "step": 1370
    },
    {
      "epoch": 685.5,
      "grad_norm": 0.1871563196182251,
      "learning_rate": 1.1406926599646373e-07,
      "loss": 0.3211,
      "step": 1371
    },
    {
      "epoch": 686.0,
      "grad_norm": 0.43379056453704834,
      "learning_rate": 1.0659405005521428e-07,
      "loss": 0.2493,
      "step": 1372
    },
    {
      "epoch": 686.5,
      "grad_norm": 0.23022551834583282,
      "learning_rate": 9.937187709688212e-08,
      "loss": 0.3495,
      "step": 1373
    },
    {
      "epoch": 687.0,
      "grad_norm": 0.8524991869926453,
      "learning_rate": 9.240278374995637e-08,
      "loss": 0.2775,
      "step": 1374
    },
    {
      "epoch": 687.5,
      "grad_norm": 0.5772730112075806,
      "learning_rate": 8.568680535939178e-08,
      "loss": 0.3658,
      "step": 1375
    },
    {
      "epoch": 688.0,
      "grad_norm": 0.7931545972824097,
      "learning_rate": 7.92239759864255e-08,
      "loss": 0.3094,
      "step": 1376
    },
    {
      "epoch": 688.5,
      "grad_norm": 0.28452348709106445,
      "learning_rate": 7.30143284084106e-08,
      "loss": 0.3466,
      "step": 1377
    },
    {
      "epoch": 689.0,
      "grad_norm": 2.6673531532287598,
      "learning_rate": 6.705789411864393e-08,
      "loss": 0.5088,
      "step": 1378
    },
    {
      "epoch": 689.5,
      "grad_norm": 1.891019582748413,
      "learning_rate": 6.135470332620519e-08,
      "loss": 0.471,
      "step": 1379
    },
    {
      "epoch": 690.0,
      "grad_norm": 0.6604340076446533,
      "learning_rate": 5.590478495580143e-08,
      "loss": 0.2691,
      "step": 1380
    },
    {
      "epoch": 690.5,
      "grad_norm": 0.2169937938451767,
      "learning_rate": 5.070816664762834e-08,
      "loss": 0.3441,
      "step": 1381
    },
    {
      "epoch": 691.0,
      "grad_norm": 0.5572425127029419,
      "learning_rate": 4.576487475724256e-08,
      "loss": 0.2692,
      "step": 1382
    },
    {
      "epoch": 691.5,
      "grad_norm": 0.25903815031051636,
      "learning_rate": 4.107493435538401e-08,
      "loss": 0.3773,
      "step": 1383
    },
    {
      "epoch": 692.0,
      "grad_norm": 5.266674995422363,
      "learning_rate": 3.663836922789821e-08,
      "loss": 0.6638,
      "step": 1384
    },
    {
      "epoch": 692.5,
      "grad_norm": 0.22817941009998322,
      "learning_rate": 3.245520187558637e-08,
      "loss": 0.3557,
      "step": 1385
    },
    {
      "epoch": 693.0,
      "grad_norm": 0.5723909735679626,
      "learning_rate": 2.8525453514099965e-08,
      "loss": 0.2797,
      "step": 1386
    },
    {
      "epoch": 693.5,
      "grad_norm": 0.4235692024230957,
      "learning_rate": 2.4849144073835206e-08,
      "loss": 0.4595,
      "step": 1387
    },
    {
      "epoch": 694.0,
      "grad_norm": 0.8978646993637085,
      "learning_rate": 2.1426292199822063e-08,
      "loss": 0.3392,
      "step": 1388
    },
    {
      "epoch": 694.5,
      "grad_norm": 0.3319687843322754,
      "learning_rate": 1.8256915251646522e-08,
      "loss": 0.4234,
      "step": 1389
    },
    {
      "epoch": 695.0,
      "grad_norm": 0.754235565662384,
      "learning_rate": 1.5341029303350686e-08,
      "loss": 0.3131,
      "step": 1390
    },
    {
      "epoch": 695.5,
      "grad_norm": 3.311337471008301,
      "learning_rate": 1.2678649143349486e-08,
      "loss": 0.558,
      "step": 1391
    },
    {
      "epoch": 696.0,
      "grad_norm": 1.2778728008270264,
      "learning_rate": 1.0269788274358538e-08,
      "loss": 1.071,
      "step": 1392
    },
    {
      "epoch": 696.5,
      "grad_norm": 0.3060607612133026,
      "learning_rate": 8.114458913333068e-09,
      "loss": 0.3505,
      "step": 1393
    },
    {
      "epoch": 697.0,
      "grad_norm": 2.1305930614471436,
      "learning_rate": 6.212671991390195e-09,
      "loss": 0.4714,
      "step": 1394
    },
    {
      "epoch": 697.5,
      "grad_norm": 1.2966057062149048,
      "learning_rate": 4.564437153775636e-09,
      "loss": 0.4233,
      "step": 1395
    },
    {
      "epoch": 698.0,
      "grad_norm": 0.4793960750102997,
      "learning_rate": 3.1697627597970793e-09,
      "loss": 0.252,
      "step": 1396
    },
    {
      "epoch": 698.5,
      "grad_norm": 0.31018561124801636,
      "learning_rate": 2.028655882779784e-09,
      "loss": 0.3243,
      "step": 1397
    },
    {
      "epoch": 699.0,
      "grad_norm": 0.5495659112930298,
      "learning_rate": 1.1411223100443735e-09,
      "loss": 0.2615,
      "step": 1398
    },
    {
      "epoch": 699.5,
      "grad_norm": 0.22182947397232056,
      "learning_rate": 5.071665428735272e-10,
      "loss": 0.3121,
      "step": 1399
    },
    {
      "epoch": 700.0,
      "grad_norm": 0.411379873752594,
      "learning_rate": 1.2679179647867578e-10,
      "loss": 0.2432,
      "step": 1400
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 1400,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 700,
  "save_steps": 1000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 6.725003617553613e+16,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}