{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.993968636911942,
  "eval_steps": 500,
  "global_step": 620,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.003216726980297547,
      "grad_norm": 81.91886138916016,
      "learning_rate": 1.6129032258064518e-07,
      "loss": 0.3558,
      "step": 1
    },
    {
      "epoch": 0.006433453960595094,
      "grad_norm": 88.99775695800781,
      "learning_rate": 3.2258064516129035e-07,
      "loss": 0.4071,
      "step": 2
    },
    {
      "epoch": 0.009650180940892641,
      "grad_norm": 79.32195281982422,
      "learning_rate": 4.838709677419355e-07,
      "loss": 0.3576,
      "step": 3
    },
    {
      "epoch": 0.012866907921190189,
      "grad_norm": 81.10282897949219,
      "learning_rate": 6.451612903225807e-07,
      "loss": 0.3627,
      "step": 4
    },
    {
      "epoch": 0.016083634901487735,
      "grad_norm": 75.47724914550781,
      "learning_rate": 8.064516129032258e-07,
      "loss": 0.3581,
      "step": 5
    },
    {
      "epoch": 0.019300361881785282,
      "grad_norm": 73.94715118408203,
      "learning_rate": 9.67741935483871e-07,
      "loss": 0.3392,
      "step": 6
    },
    {
      "epoch": 0.02251708886208283,
      "grad_norm": 54.852840423583984,
      "learning_rate": 1.1290322580645162e-06,
      "loss": 0.2786,
      "step": 7
    },
    {
      "epoch": 0.025733815842380377,
      "grad_norm": 51.52576446533203,
      "learning_rate": 1.2903225806451614e-06,
      "loss": 0.27,
      "step": 8
    },
    {
      "epoch": 0.028950542822677925,
      "grad_norm": 17.170745849609375,
      "learning_rate": 1.4516129032258066e-06,
      "loss": 0.1758,
      "step": 9
    },
    {
      "epoch": 0.03216726980297547,
      "grad_norm": 10.435275077819824,
      "learning_rate": 1.6129032258064516e-06,
      "loss": 0.1723,
      "step": 10
    },
    {
      "epoch": 0.03538399678327302,
      "grad_norm": 8.09239387512207,
      "learning_rate": 1.774193548387097e-06,
      "loss": 0.1565,
      "step": 11
    },
    {
      "epoch": 0.038600723763570564,
      "grad_norm": 41.39366149902344,
      "learning_rate": 1.935483870967742e-06,
      "loss": 0.1746,
      "step": 12
    },
    {
      "epoch": 0.04181745074386811,
      "grad_norm": 20.37698745727539,
      "learning_rate": 2.096774193548387e-06,
      "loss": 0.1709,
      "step": 13
    },
    {
      "epoch": 0.04503417772416566,
      "grad_norm": 10.368378639221191,
      "learning_rate": 2.2580645161290324e-06,
      "loss": 0.1567,
      "step": 14
    },
    {
      "epoch": 0.04825090470446321,
      "grad_norm": 14.423890113830566,
      "learning_rate": 2.4193548387096776e-06,
      "loss": 0.1344,
      "step": 15
    },
    {
      "epoch": 0.051467631684760755,
      "grad_norm": 13.449414253234863,
      "learning_rate": 2.580645161290323e-06,
      "loss": 0.1306,
      "step": 16
    },
    {
      "epoch": 0.0546843586650583,
      "grad_norm": 5.699060916900635,
      "learning_rate": 2.7419354838709676e-06,
      "loss": 0.1368,
      "step": 17
    },
    {
      "epoch": 0.05790108564535585,
      "grad_norm": 38.163658142089844,
      "learning_rate": 2.903225806451613e-06,
      "loss": 0.1767,
      "step": 18
    },
    {
      "epoch": 0.0611178126256534,
      "grad_norm": 43.165924072265625,
      "learning_rate": 3.0645161290322584e-06,
      "loss": 0.2103,
      "step": 19
    },
    {
      "epoch": 0.06433453960595094,
      "grad_norm": 34.02187728881836,
      "learning_rate": 3.225806451612903e-06,
      "loss": 0.1798,
      "step": 20
    },
    {
      "epoch": 0.06755126658624849,
      "grad_norm": 15.196747779846191,
      "learning_rate": 3.3870967741935484e-06,
      "loss": 0.1326,
      "step": 21
    },
    {
      "epoch": 0.07076799356654603,
      "grad_norm": 12.960094451904297,
      "learning_rate": 3.548387096774194e-06,
      "loss": 0.132,
      "step": 22
    },
    {
      "epoch": 0.07398472054684359,
      "grad_norm": 18.067081451416016,
      "learning_rate": 3.7096774193548392e-06,
      "loss": 0.1407,
      "step": 23
    },
    {
      "epoch": 0.07720144752714113,
      "grad_norm": 11.63009262084961,
      "learning_rate": 3.870967741935484e-06,
      "loss": 0.1198,
      "step": 24
    },
    {
      "epoch": 0.08041817450743868,
      "grad_norm": 3.6088430881500244,
      "learning_rate": 4.032258064516129e-06,
      "loss": 0.0861,
      "step": 25
    },
    {
      "epoch": 0.08363490148773622,
      "grad_norm": 10.469372749328613,
      "learning_rate": 4.193548387096774e-06,
      "loss": 0.1043,
      "step": 26
    },
    {
      "epoch": 0.08685162846803378,
      "grad_norm": 14.186717987060547,
      "learning_rate": 4.35483870967742e-06,
      "loss": 0.135,
      "step": 27
    },
    {
      "epoch": 0.09006835544833132,
      "grad_norm": 7.1297454833984375,
      "learning_rate": 4.516129032258065e-06,
      "loss": 0.1171,
      "step": 28
    },
    {
      "epoch": 0.09328508242862887,
      "grad_norm": 2.476020574569702,
      "learning_rate": 4.67741935483871e-06,
      "loss": 0.089,
      "step": 29
    },
    {
      "epoch": 0.09650180940892641,
      "grad_norm": 3.7916111946105957,
      "learning_rate": 4.838709677419355e-06,
      "loss": 0.0881,
      "step": 30
    },
    {
      "epoch": 0.09971853638922397,
      "grad_norm": 1.1642123460769653,
      "learning_rate": 5e-06,
      "loss": 0.0965,
      "step": 31
    },
    {
      "epoch": 0.10293526336952151,
      "grad_norm": 3.4518535137176514,
      "learning_rate": 5.161290322580646e-06,
      "loss": 0.0849,
      "step": 32
    },
    {
      "epoch": 0.10615199034981906,
      "grad_norm": 2.2884507179260254,
      "learning_rate": 5.322580645161291e-06,
      "loss": 0.0786,
      "step": 33
    },
    {
      "epoch": 0.1093687173301166,
      "grad_norm": 1.3953402042388916,
      "learning_rate": 5.483870967741935e-06,
      "loss": 0.0675,
      "step": 34
    },
    {
      "epoch": 0.11258544431041416,
      "grad_norm": 5.182158470153809,
      "learning_rate": 5.645161290322582e-06,
      "loss": 0.0977,
      "step": 35
    },
    {
      "epoch": 0.1158021712907117,
      "grad_norm": 2.713275909423828,
      "learning_rate": 5.806451612903226e-06,
      "loss": 0.0648,
      "step": 36
    },
    {
      "epoch": 0.11901889827100925,
      "grad_norm": 2.827744960784912,
      "learning_rate": 5.967741935483872e-06,
      "loss": 0.082,
      "step": 37
    },
    {
      "epoch": 0.1222356252513068,
      "grad_norm": 4.1014723777771,
      "learning_rate": 6.129032258064517e-06,
      "loss": 0.0758,
      "step": 38
    },
    {
      "epoch": 0.12545235223160434,
      "grad_norm": 1.5448578596115112,
      "learning_rate": 6.290322580645162e-06,
      "loss": 0.066,
      "step": 39
    },
    {
      "epoch": 0.12866907921190188,
      "grad_norm": 3.3927855491638184,
      "learning_rate": 6.451612903225806e-06,
      "loss": 0.0709,
      "step": 40
    },
    {
      "epoch": 0.13188580619219945,
      "grad_norm": 2.370821475982666,
      "learning_rate": 6.612903225806452e-06,
      "loss": 0.0852,
      "step": 41
    },
    {
      "epoch": 0.13510253317249699,
      "grad_norm": 3.0705552101135254,
      "learning_rate": 6.774193548387097e-06,
      "loss": 0.0702,
      "step": 42
    },
    {
      "epoch": 0.13831926015279453,
      "grad_norm": 4.215071678161621,
      "learning_rate": 6.935483870967743e-06,
      "loss": 0.0747,
      "step": 43
    },
    {
      "epoch": 0.14153598713309207,
      "grad_norm": 2.137068748474121,
      "learning_rate": 7.096774193548388e-06,
      "loss": 0.0657,
      "step": 44
    },
    {
      "epoch": 0.14475271411338964,
      "grad_norm": 4.441213130950928,
      "learning_rate": 7.258064516129033e-06,
      "loss": 0.0717,
      "step": 45
    },
    {
      "epoch": 0.14796944109368718,
      "grad_norm": 2.896615743637085,
      "learning_rate": 7.4193548387096784e-06,
      "loss": 0.066,
      "step": 46
    },
    {
      "epoch": 0.15118616807398472,
      "grad_norm": 1.594978928565979,
      "learning_rate": 7.580645161290323e-06,
      "loss": 0.0554,
      "step": 47
    },
    {
      "epoch": 0.15440289505428226,
      "grad_norm": 3.3716137409210205,
      "learning_rate": 7.741935483870968e-06,
      "loss": 0.075,
      "step": 48
    },
    {
      "epoch": 0.15761962203457983,
      "grad_norm": 2.1819281578063965,
      "learning_rate": 7.903225806451613e-06,
      "loss": 0.0583,
      "step": 49
    },
    {
      "epoch": 0.16083634901487737,
      "grad_norm": 5.265512943267822,
      "learning_rate": 8.064516129032258e-06,
      "loss": 0.0832,
      "step": 50
    },
    {
      "epoch": 0.1640530759951749,
      "grad_norm": 6.982624053955078,
      "learning_rate": 8.225806451612904e-06,
      "loss": 0.0772,
      "step": 51
    },
    {
      "epoch": 0.16726980297547245,
      "grad_norm": 4.082429885864258,
      "learning_rate": 8.387096774193549e-06,
      "loss": 0.0734,
      "step": 52
    },
    {
      "epoch": 0.17048652995577002,
      "grad_norm": 1.164082646369934,
      "learning_rate": 8.548387096774194e-06,
      "loss": 0.0541,
      "step": 53
    },
    {
      "epoch": 0.17370325693606756,
      "grad_norm": 2.810659646987915,
      "learning_rate": 8.70967741935484e-06,
      "loss": 0.052,
      "step": 54
    },
    {
      "epoch": 0.1769199839163651,
      "grad_norm": 1.5780754089355469,
      "learning_rate": 8.870967741935484e-06,
      "loss": 0.0582,
      "step": 55
    },
    {
      "epoch": 0.18013671089666264,
      "grad_norm": 1.068620204925537,
      "learning_rate": 9.03225806451613e-06,
      "loss": 0.065,
      "step": 56
    },
    {
      "epoch": 0.18335343787696018,
      "grad_norm": 2.523627758026123,
      "learning_rate": 9.193548387096775e-06,
      "loss": 0.0624,
      "step": 57
    },
    {
      "epoch": 0.18657016485725775,
      "grad_norm": 1.244511365890503,
      "learning_rate": 9.35483870967742e-06,
      "loss": 0.0503,
      "step": 58
    },
    {
      "epoch": 0.1897868918375553,
      "grad_norm": 0.9401457905769348,
      "learning_rate": 9.516129032258065e-06,
      "loss": 0.051,
      "step": 59
    },
    {
      "epoch": 0.19300361881785283,
      "grad_norm": 0.9283802509307861,
      "learning_rate": 9.67741935483871e-06,
      "loss": 0.0537,
      "step": 60
    },
    {
      "epoch": 0.19622034579815037,
      "grad_norm": 1.3737329244613647,
      "learning_rate": 9.838709677419356e-06,
      "loss": 0.0528,
      "step": 61
    },
    {
      "epoch": 0.19943707277844794,
      "grad_norm": 0.8466247916221619,
      "learning_rate": 1e-05,
      "loss": 0.0505,
      "step": 62
    },
    {
      "epoch": 0.20265379975874548,
      "grad_norm": 3.419365644454956,
      "learning_rate": 9.999920755303033e-06,
      "loss": 0.0696,
      "step": 63
    },
    {
      "epoch": 0.20587052673904302,
      "grad_norm": 2.1508781909942627,
      "learning_rate": 9.999683023724021e-06,
      "loss": 0.0502,
      "step": 64
    },
    {
      "epoch": 0.20908725371934056,
      "grad_norm": 1.2374366521835327,
      "learning_rate": 9.99928681279855e-06,
      "loss": 0.0536,
      "step": 65
    },
    {
      "epoch": 0.21230398069963813,
      "grad_norm": 2.0750317573547363,
      "learning_rate": 9.998732135085665e-06,
      "loss": 0.0634,
      "step": 66
    },
    {
      "epoch": 0.21552070767993567,
      "grad_norm": 1.1349234580993652,
      "learning_rate": 9.998019008167476e-06,
      "loss": 0.0472,
      "step": 67
    },
    {
      "epoch": 0.2187374346602332,
      "grad_norm": 4.443458080291748,
      "learning_rate": 9.99714745464859e-06,
      "loss": 0.0714,
      "step": 68
    },
    {
      "epoch": 0.22195416164053075,
      "grad_norm": 5.246551036834717,
      "learning_rate": 9.99611750215541e-06,
      "loss": 0.0686,
      "step": 69
    },
    {
      "epoch": 0.22517088862082832,
      "grad_norm": 1.8846714496612549,
      "learning_rate": 9.994929183335237e-06,
      "loss": 0.0707,
      "step": 70
    },
    {
      "epoch": 0.22838761560112586,
      "grad_norm": 2.367842674255371,
      "learning_rate": 9.993582535855265e-06,
      "loss": 0.0596,
      "step": 71
    },
    {
      "epoch": 0.2316043425814234,
      "grad_norm": 3.4915974140167236,
      "learning_rate": 9.992077602401358e-06,
      "loss": 0.0644,
      "step": 72
    },
    {
      "epoch": 0.23482106956172094,
      "grad_norm": 1.2369784116744995,
      "learning_rate": 9.990414430676716e-06,
      "loss": 0.0566,
      "step": 73
    },
    {
      "epoch": 0.2380377965420185,
      "grad_norm": 1.0058597326278687,
      "learning_rate": 9.988593073400354e-06,
      "loss": 0.0518,
      "step": 74
    },
    {
      "epoch": 0.24125452352231605,
      "grad_norm": 1.170558214187622,
      "learning_rate": 9.986613588305435e-06,
      "loss": 0.0481,
      "step": 75
    },
    {
      "epoch": 0.2444712505026136,
      "grad_norm": 0.9504396915435791,
      "learning_rate": 9.984476038137437e-06,
      "loss": 0.0458,
      "step": 76
    },
    {
      "epoch": 0.24768797748291113,
      "grad_norm": 4.625006675720215,
      "learning_rate": 9.982180490652165e-06,
      "loss": 0.0663,
      "step": 77
    },
    {
      "epoch": 0.25090470446320867,
      "grad_norm": 2.3191418647766113,
      "learning_rate": 9.979727018613607e-06,
      "loss": 0.0439,
      "step": 78
    },
    {
      "epoch": 0.25412143144350624,
      "grad_norm": 1.6738172769546509,
      "learning_rate": 9.977115699791622e-06,
      "loss": 0.0668,
      "step": 79
    },
    {
      "epoch": 0.25733815842380375,
      "grad_norm": 2.8376853466033936,
      "learning_rate": 9.974346616959476e-06,
      "loss": 0.0612,
      "step": 80
    },
    {
      "epoch": 0.2605548854041013,
      "grad_norm": 4.625660419464111,
      "learning_rate": 9.971419857891223e-06,
      "loss": 0.0602,
      "step": 81
    },
    {
      "epoch": 0.2637716123843989,
      "grad_norm": 6.061123847961426,
      "learning_rate": 9.968335515358916e-06,
      "loss": 0.0623,
      "step": 82
    },
    {
      "epoch": 0.2669883393646964,
      "grad_norm": 1.5065197944641113,
      "learning_rate": 9.965093687129669e-06,
      "loss": 0.0611,
      "step": 83
    },
    {
      "epoch": 0.27020506634499397,
      "grad_norm": 0.9030428528785706,
      "learning_rate": 9.961694475962562e-06,
      "loss": 0.0623,
      "step": 84
    },
    {
      "epoch": 0.27342179332529154,
      "grad_norm": 5.082237243652344,
      "learning_rate": 9.95813798960538e-06,
      "loss": 0.0758,
      "step": 85
    },
    {
      "epoch": 0.27663852030558905,
      "grad_norm": 5.320428848266602,
      "learning_rate": 9.954424340791195e-06,
      "loss": 0.0675,
      "step": 86
    },
    {
      "epoch": 0.2798552472858866,
      "grad_norm": 6.164220809936523,
      "learning_rate": 9.950553647234798e-06,
      "loss": 0.0713,
      "step": 87
    },
    {
      "epoch": 0.28307197426618413,
      "grad_norm": 3.148167371749878,
      "learning_rate": 9.94652603162896e-06,
      "loss": 0.0577,
      "step": 88
    },
    {
      "epoch": 0.2862887012464817,
      "grad_norm": 1.503680944442749,
      "learning_rate": 9.942341621640558e-06,
      "loss": 0.062,
      "step": 89
    },
    {
      "epoch": 0.28950542822677927,
      "grad_norm": 2.4841108322143555,
      "learning_rate": 9.938000549906509e-06,
      "loss": 0.0594,
      "step": 90
    },
    {
      "epoch": 0.2927221552070768,
      "grad_norm": 4.777047157287598,
      "learning_rate": 9.93350295402958e-06,
      "loss": 0.0601,
      "step": 91
    },
    {
      "epoch": 0.29593888218737435,
      "grad_norm": 4.280494213104248,
      "learning_rate": 9.92884897657402e-06,
      "loss": 0.0654,
      "step": 92
    },
    {
      "epoch": 0.2991556091676719,
      "grad_norm": 3.8119473457336426,
      "learning_rate": 9.924038765061042e-06,
      "loss": 0.0645,
      "step": 93
    },
    {
      "epoch": 0.30237233614796943,
      "grad_norm": 1.2760810852050781,
      "learning_rate": 9.919072471964146e-06,
      "loss": 0.0502,
      "step": 94
    },
    {
      "epoch": 0.305589063128267,
      "grad_norm": 2.0416059494018555,
      "learning_rate": 9.913950254704291e-06,
      "loss": 0.0601,
      "step": 95
    },
    {
      "epoch": 0.3088057901085645,
      "grad_norm": 2.9436042308807373,
      "learning_rate": 9.908672275644898e-06,
      "loss": 0.0552,
      "step": 96
    },
    {
      "epoch": 0.3120225170888621,
      "grad_norm": 2.80222487449646,
      "learning_rate": 9.903238702086707e-06,
      "loss": 0.0534,
      "step": 97
    },
    {
      "epoch": 0.31523924406915965,
      "grad_norm": 2.396406888961792,
      "learning_rate": 9.897649706262474e-06,
      "loss": 0.0516,
      "step": 98
    },
    {
      "epoch": 0.31845597104945716,
      "grad_norm": 1.1176875829696655,
      "learning_rate": 9.89190546533151e-06,
      "loss": 0.0544,
      "step": 99
    },
    {
      "epoch": 0.32167269802975473,
      "grad_norm": 2.592433452606201,
      "learning_rate": 9.88600616137407e-06,
      "loss": 0.0493,
      "step": 100
    },
    {
      "epoch": 0.32488942501005225,
      "grad_norm": 2.838442802429199,
      "learning_rate": 9.879951981385577e-06,
      "loss": 0.0628,
      "step": 101
    },
    {
      "epoch": 0.3281061519903498,
      "grad_norm": 3.090041399002075,
      "learning_rate": 9.873743117270691e-06,
      "loss": 0.0426,
      "step": 102
    },
    {
      "epoch": 0.3313228789706474,
      "grad_norm": 3.3808889389038086,
      "learning_rate": 9.867379765837237e-06,
      "loss": 0.0579,
      "step": 103
    },
    {
      "epoch": 0.3345396059509449,
      "grad_norm": 0.756629228591919,
      "learning_rate": 9.860862128789954e-06,
      "loss": 0.0537,
      "step": 104
    },
    {
      "epoch": 0.33775633293124246,
      "grad_norm": 1.3207106590270996,
      "learning_rate": 9.854190412724114e-06,
      "loss": 0.0406,
      "step": 105
    },
    {
      "epoch": 0.34097305991154003,
      "grad_norm": 2.726410388946533,
      "learning_rate": 9.847364829118963e-06,
      "loss": 0.0603,
      "step": 106
    },
    {
      "epoch": 0.34418978689183755,
      "grad_norm": 1.6553726196289062,
      "learning_rate": 9.840385594331022e-06,
      "loss": 0.0635,
      "step": 107
    },
    {
      "epoch": 0.3474065138721351,
      "grad_norm": 0.6768801808357239,
      "learning_rate": 9.833252929587231e-06,
      "loss": 0.0538,
      "step": 108
    },
    {
      "epoch": 0.3506232408524326,
      "grad_norm": 2.1218836307525635,
      "learning_rate": 9.825967060977933e-06,
      "loss": 0.0503,
      "step": 109
    },
    {
      "epoch": 0.3538399678327302,
      "grad_norm": 2.7606120109558105,
      "learning_rate": 9.818528219449705e-06,
      "loss": 0.05,
      "step": 110
    },
    {
      "epoch": 0.35705669481302776,
      "grad_norm": 1.1629998683929443,
      "learning_rate": 9.810936640798046e-06,
      "loss": 0.0456,
      "step": 111
    },
    {
      "epoch": 0.3602734217933253,
      "grad_norm": 1.027559757232666,
      "learning_rate": 9.803192565659898e-06,
      "loss": 0.0522,
      "step": 112
    },
    {
      "epoch": 0.36349014877362285,
      "grad_norm": 0.8349726796150208,
      "learning_rate": 9.795296239506011e-06,
      "loss": 0.0424,
      "step": 113
    },
    {
      "epoch": 0.36670687575392036,
      "grad_norm": 1.19048273563385,
      "learning_rate": 9.78724791263318e-06,
      "loss": 0.0472,
      "step": 114
    },
    {
      "epoch": 0.3699236027342179,
      "grad_norm": 0.9269819259643555,
      "learning_rate": 9.779047840156288e-06,
      "loss": 0.0528,
      "step": 115
    },
    {
      "epoch": 0.3731403297145155,
      "grad_norm": 0.7386018633842468,
      "learning_rate": 9.770696282000245e-06,
      "loss": 0.0561,
      "step": 116
    },
    {
      "epoch": 0.376357056694813,
      "grad_norm": 0.8170301914215088,
      "learning_rate": 9.762193502891726e-06,
      "loss": 0.0457,
      "step": 117
    },
    {
      "epoch": 0.3795737836751106,
      "grad_norm": 1.0884501934051514,
      "learning_rate": 9.753539772350792e-06,
      "loss": 0.0467,
      "step": 118
    },
    {
      "epoch": 0.38279051065540814,
      "grad_norm": 1.1235942840576172,
      "learning_rate": 9.744735364682347e-06,
      "loss": 0.0406,
      "step": 119
    },
    {
      "epoch": 0.38600723763570566,
      "grad_norm": 0.8838767409324646,
      "learning_rate": 9.735780558967434e-06,
      "loss": 0.0487,
      "step": 120
    },
    {
      "epoch": 0.3892239646160032,
      "grad_norm": 0.9004449248313904,
      "learning_rate": 9.726675639054403e-06,
      "loss": 0.0543,
      "step": 121
    },
    {
      "epoch": 0.39244069159630074,
      "grad_norm": 1.97210693359375,
      "learning_rate": 9.717420893549902e-06,
      "loss": 0.0493,
      "step": 122
    },
    {
      "epoch": 0.3956574185765983,
      "grad_norm": 0.7708803415298462,
      "learning_rate": 9.70801661580973e-06,
      "loss": 0.0437,
      "step": 123
    },
    {
      "epoch": 0.3988741455568959,
      "grad_norm": 1.6689863204956055,
      "learning_rate": 9.698463103929542e-06,
      "loss": 0.0455,
      "step": 124
    },
    {
      "epoch": 0.4020908725371934,
      "grad_norm": 1.5968708992004395,
      "learning_rate": 9.688760660735403e-06,
      "loss": 0.0569,
      "step": 125
    },
    {
      "epoch": 0.40530759951749096,
      "grad_norm": 1.1041440963745117,
      "learning_rate": 9.67890959377418e-06,
      "loss": 0.0576,
      "step": 126
    },
    {
      "epoch": 0.4085243264977885,
      "grad_norm": 2.1897428035736084,
      "learning_rate": 9.668910215303797e-06,
      "loss": 0.0514,
      "step": 127
    },
    {
      "epoch": 0.41174105347808604,
      "grad_norm": 1.594397783279419,
      "learning_rate": 9.658762842283343e-06,
      "loss": 0.0444,
      "step": 128
    },
    {
      "epoch": 0.4149577804583836,
      "grad_norm": 1.404996395111084,
      "learning_rate": 9.648467796363019e-06,
      "loss": 0.0493,
      "step": 129
    },
    {
      "epoch": 0.4181745074386811,
      "grad_norm": 0.889514148235321,
      "learning_rate": 9.638025403873939e-06,
      "loss": 0.0576,
      "step": 130
    },
    {
      "epoch": 0.4213912344189787,
      "grad_norm": 3.748598337173462,
      "learning_rate": 9.627435995817799e-06,
      "loss": 0.0599,
      "step": 131
    },
    {
      "epoch": 0.42460796139927626,
      "grad_norm": 4.13529109954834,
      "learning_rate": 9.616699907856368e-06,
      "loss": 0.0625,
      "step": 132
    },
    {
      "epoch": 0.42782468837957377,
      "grad_norm": 2.6388087272644043,
      "learning_rate": 9.605817480300863e-06,
      "loss": 0.0536,
      "step": 133
    },
    {
      "epoch": 0.43104141535987134,
      "grad_norm": 1.5752240419387817,
      "learning_rate": 9.594789058101154e-06,
      "loss": 0.0636,
      "step": 134
    },
    {
      "epoch": 0.43425814234016885,
      "grad_norm": 1.7247588634490967,
      "learning_rate": 9.58361499083483e-06,
      "loss": 0.0602,
      "step": 135
    },
    {
      "epoch": 0.4374748693204664,
      "grad_norm": 1.2270214557647705,
      "learning_rate": 9.57229563269612e-06,
      "loss": 0.0543,
      "step": 136
    },
    {
      "epoch": 0.440691596300764,
      "grad_norm": 1.9172810316085815,
      "learning_rate": 9.560831342484668e-06,
      "loss": 0.0456,
      "step": 137
    },
    {
      "epoch": 0.4439083232810615,
      "grad_norm": 5.208808898925781,
      "learning_rate": 9.549222483594154e-06,
      "loss": 0.0868,
      "step": 138
    },
    {
      "epoch": 0.44712505026135907,
      "grad_norm": 4.008519172668457,
      "learning_rate": 9.53746942400078e-06,
      "loss": 0.0646,
      "step": 139
    },
    {
      "epoch": 0.45034177724165664,
      "grad_norm": 2.6669585704803467,
      "learning_rate": 9.525572536251608e-06,
      "loss": 0.068,
      "step": 140
    },
    {
      "epoch": 0.45355850422195415,
      "grad_norm": 0.6555745601654053,
      "learning_rate": 9.513532197452737e-06,
      "loss": 0.0539,
      "step": 141
    },
    {
      "epoch": 0.4567752312022517,
      "grad_norm": 3.599552869796753,
      "learning_rate": 9.501348789257373e-06,
      "loss": 0.0515,
      "step": 142
    },
    {
      "epoch": 0.45999195818254923,
      "grad_norm": 5.188879489898682,
      "learning_rate": 9.48902269785371e-06,
      "loss": 0.0754,
      "step": 143
    },
    {
      "epoch": 0.4632086851628468,
      "grad_norm": 3.3630921840667725,
      "learning_rate": 9.476554313952697e-06,
      "loss": 0.0651,
      "step": 144
    },
    {
      "epoch": 0.46642541214314437,
      "grad_norm": 2.4736194610595703,
      "learning_rate": 9.46394403277566e-06,
      "loss": 0.0578,
      "step": 145
    },
    {
      "epoch": 0.4696421391234419,
      "grad_norm": 0.7932046055793762,
      "learning_rate": 9.451192254041759e-06,
      "loss": 0.0466,
      "step": 146
    },
    {
      "epoch": 0.47285886610373945,
      "grad_norm": 0.980038583278656,
      "learning_rate": 9.438299381955333e-06,
      "loss": 0.0452,
      "step": 147
    },
    {
      "epoch": 0.476075593084037,
      "grad_norm": 1.6979252099990845,
      "learning_rate": 9.425265825193077e-06,
      "loss": 0.0445,
      "step": 148
    },
    {
      "epoch": 0.47929232006433453,
      "grad_norm": 2.8460330963134766,
      "learning_rate": 9.412091996891097e-06,
      "loss": 0.051,
      "step": 149
    },
    {
      "epoch": 0.4825090470446321,
      "grad_norm": 1.3673274517059326,
      "learning_rate": 9.398778314631801e-06,
      "loss": 0.0473,
      "step": 150
    },
    {
      "epoch": 0.4857257740249296,
      "grad_norm": 1.6840636730194092,
      "learning_rate": 9.385325200430679e-06,
      "loss": 0.0525,
      "step": 151
    },
    {
      "epoch": 0.4889425010052272,
      "grad_norm": 2.1259257793426514,
      "learning_rate": 9.371733080722911e-06,
      "loss": 0.0548,
      "step": 152
    },
    {
      "epoch": 0.49215922798552475,
      "grad_norm": 0.7495786547660828,
      "learning_rate": 9.358002386349862e-06,
      "loss": 0.0379,
      "step": 153
    },
    {
      "epoch": 0.49537595496582226,
      "grad_norm": 1.248661756515503,
      "learning_rate": 9.34413355254542e-06,
      "loss": 0.0477,
      "step": 154
    },
    {
      "epoch": 0.49859268194611983,
      "grad_norm": 1.250671148300171,
      "learning_rate": 9.330127018922195e-06,
      "loss": 0.0524,
      "step": 155
    },
    {
      "epoch": 0.5018094089264173,
      "grad_norm": 0.5698635578155518,
      "learning_rate": 9.31598322945759e-06,
      "loss": 0.0508,
      "step": 156
    },
    {
      "epoch": 0.505026135906715,
      "grad_norm": 0.6476792097091675,
      "learning_rate": 9.301702632479734e-06,
      "loss": 0.0482,
      "step": 157
    },
    {
      "epoch": 0.5082428628870125,
      "grad_norm": 0.6281876564025879,
      "learning_rate": 9.287285680653254e-06,
      "loss": 0.0448,
      "step": 158
    },
    {
      "epoch": 0.51145958986731,
      "grad_norm": 0.8750901222229004,
      "learning_rate": 9.272732830964948e-06,
      "loss": 0.0652,
      "step": 159
    },
    {
      "epoch": 0.5146763168476075,
      "grad_norm": 0.6696067452430725,
      "learning_rate": 9.258044544709276e-06,
      "loss": 0.0432,
      "step": 160
    },
    {
      "epoch": 0.5178930438279051,
      "grad_norm": 0.5400009155273438,
      "learning_rate": 9.243221287473755e-06,
      "loss": 0.0389,
      "step": 161
    },
    {
      "epoch": 0.5211097708082026,
      "grad_norm": 0.59368497133255,
      "learning_rate": 9.228263529124199e-06,
      "loss": 0.0407,
      "step": 162
    },
    {
      "epoch": 0.5243264977885002,
      "grad_norm": 0.8467090725898743,
      "learning_rate": 9.21317174378982e-06,
      "loss": 0.0489,
      "step": 163
    },
    {
      "epoch": 0.5275432247687978,
      "grad_norm": 0.6505163908004761,
      "learning_rate": 9.197946409848196e-06,
      "loss": 0.0448,
      "step": 164
    },
    {
      "epoch": 0.5307599517490953,
      "grad_norm": 1.452300786972046,
      "learning_rate": 9.182588009910119e-06,
      "loss": 0.0487,
      "step": 165
    },
    {
      "epoch": 0.5339766787293928,
      "grad_norm": 1.1281909942626953,
      "learning_rate": 9.167097030804289e-06,
      "loss": 0.0468,
      "step": 166
    },
    {
      "epoch": 0.5371934057096904,
      "grad_norm": 1.3461655378341675,
      "learning_rate": 9.151473963561884e-06,
      "loss": 0.0589,
      "step": 167
    },
    {
      "epoch": 0.5404101326899879,
      "grad_norm": 0.5634208917617798,
      "learning_rate": 9.135719303400995e-06,
      "loss": 0.0405,
      "step": 168
    },
    {
      "epoch": 0.5436268596702855,
      "grad_norm": 1.198769450187683,
      "learning_rate": 9.119833549710927e-06,
      "loss": 0.0635,
      "step": 169
    },
    {
      "epoch": 0.5468435866505831,
      "grad_norm": 1.4971657991409302,
      "learning_rate": 9.103817206036383e-06,
      "loss": 0.05,
      "step": 170
    },
    {
      "epoch": 0.5500603136308806,
      "grad_norm": 0.9563581347465515,
      "learning_rate": 9.087670780061477e-06,
      "loss": 0.0555,
      "step": 171
    },
    {
      "epoch": 0.5532770406111781,
      "grad_norm": 0.7893424034118652,
      "learning_rate": 9.071394783593664e-06,
      "loss": 0.0495,
      "step": 172
    },
    {
      "epoch": 0.5564937675914756,
      "grad_norm": 1.9164668321609497,
      "learning_rate": 9.054989732547507e-06,
      "loss": 0.0519,
      "step": 173
    },
    {
      "epoch": 0.5597104945717732,
      "grad_norm": 0.7731283903121948,
      "learning_rate": 9.038456146928325e-06,
      "loss": 0.047,
      "step": 174
    },
    {
      "epoch": 0.5629272215520708,
      "grad_norm": 1.0580449104309082,
      "learning_rate": 9.021794550815713e-06,
      "loss": 0.052,
      "step": 175
    },
    {
      "epoch": 0.5661439485323683,
      "grad_norm": 0.7994678616523743,
      "learning_rate": 9.005005472346923e-06,
      "loss": 0.0587,
      "step": 176
    },
    {
      "epoch": 0.5693606755126659,
      "grad_norm": 1.5067932605743408,
      "learning_rate": 8.988089443700131e-06,
      "loss": 0.0426,
      "step": 177
    },
    {
      "epoch": 0.5725774024929634,
      "grad_norm": 1.0666264295578003,
      "learning_rate": 8.971047001077561e-06,
      "loss": 0.0432,
      "step": 178
    },
    {
      "epoch": 0.5757941294732609,
      "grad_norm": 2.072291135787964,
      "learning_rate": 8.953878684688492e-06,
      "loss": 0.0467,
      "step": 179
    },
    {
      "epoch": 0.5790108564535585,
      "grad_norm": 0.4865788221359253,
      "learning_rate": 8.936585038732143e-06,
      "loss": 0.0375,
      "step": 180
    },
    {
      "epoch": 0.582227583433856,
      "grad_norm": 2.6773552894592285,
      "learning_rate": 8.919166611380397e-06,
      "loss": 0.065,
      "step": 181
    },
    {
      "epoch": 0.5854443104141536,
      "grad_norm": 3.049079656600952,
      "learning_rate": 8.90162395476046e-06,
      "loss": 0.0652,
      "step": 182
    },
    {
      "epoch": 0.5886610373944512,
      "grad_norm": 1.1698511838912964,
      "learning_rate": 8.883957624937333e-06,
      "loss": 0.0519,
      "step": 183
    },
    {
      "epoch": 0.5918777643747487,
      "grad_norm": 1.1253437995910645,
      "learning_rate": 8.866168181896198e-06,
      "loss": 0.0562,
      "step": 184
    },
    {
      "epoch": 0.5950944913550462,
      "grad_norm": 0.5882744789123535,
      "learning_rate": 8.848256189524661e-06,
      "loss": 0.0472,
      "step": 185
    },
    {
      "epoch": 0.5983112183353438,
      "grad_norm": 0.5303323864936829,
      "learning_rate": 8.83022221559489e-06,
      "loss": 0.0359,
      "step": 186
    },
    {
      "epoch": 0.6015279453156414,
      "grad_norm": 0.8436638712882996,
      "learning_rate": 8.812066831745602e-06,
      "loss": 0.0335,
      "step": 187
    },
    {
      "epoch": 0.6047446722959389,
      "grad_norm": 0.7851281762123108,
      "learning_rate": 8.793790613463956e-06,
      "loss": 0.0459,
      "step": 188
    },
    {
      "epoch": 0.6079613992762364,
      "grad_norm": 0.7391752600669861,
      "learning_rate": 8.775394140067299e-06,
      "loss": 0.0445,
      "step": 189
    },
    {
      "epoch": 0.611178126256534,
      "grad_norm": 0.9071338772773743,
      "learning_rate": 8.756877994684818e-06,
      "loss": 0.0468,
      "step": 190
    },
    {
      "epoch": 0.6143948532368315,
      "grad_norm": 1.225618600845337,
      "learning_rate": 8.738242764239046e-06,
      "loss": 0.0433,
      "step": 191
    },
    {
      "epoch": 0.617611580217129,
      "grad_norm": 1.5072846412658691,
      "learning_rate": 8.719489039427256e-06,
      "loss": 0.0467,
      "step": 192
    },
    {
      "epoch": 0.6208283071974267,
      "grad_norm": 0.7594591975212097,
      "learning_rate": 8.700617414702746e-06,
      "loss": 0.0407,
      "step": 193
    },
    {
      "epoch": 0.6240450341777242,
      "grad_norm": 1.7316999435424805,
      "learning_rate": 8.681628488255986e-06,
      "loss": 0.0524,
      "step": 194
    },
    {
      "epoch": 0.6272617611580217,
      "grad_norm": 0.6949440240859985,
      "learning_rate": 8.66252286199567e-06,
      "loss": 0.0354,
      "step": 195
    },
    {
      "epoch": 0.6304784881383193,
      "grad_norm": 0.6324911117553711,
      "learning_rate": 8.643301141529619e-06,
      "loss": 0.0525,
      "step": 196
    },
    {
      "epoch": 0.6336952151186168,
      "grad_norm": 0.6617627739906311,
      "learning_rate": 8.6239639361456e-06,
      "loss": 0.0471,
      "step": 197
    },
    {
      "epoch": 0.6369119420989143,
      "grad_norm": 1.0735185146331787,
      "learning_rate": 8.604511858792006e-06,
      "loss": 0.0347,
      "step": 198
    },
    {
      "epoch": 0.640128669079212,
      "grad_norm": 0.6503901481628418,
      "learning_rate": 8.584945526058426e-06,
      "loss": 0.0621,
      "step": 199
    },
    {
      "epoch": 0.6433453960595095,
      "grad_norm": 0.5145460367202759,
      "learning_rate": 8.565265558156101e-06,
      "loss": 0.0444,
      "step": 200
    },
    {
      "epoch": 0.646562123039807,
      "grad_norm": 0.5394482016563416,
      "learning_rate": 8.545472578898276e-06,
      "loss": 0.0505,
      "step": 201
    },
    {
      "epoch": 0.6497788500201045,
      "grad_norm": 0.9032294750213623,
      "learning_rate": 8.525567215680397e-06,
      "loss": 0.0522,
      "step": 202
    },
    {
      "epoch": 0.6529955770004021,
      "grad_norm": 0.4706067144870758,
      "learning_rate": 8.505550099460264e-06,
      "loss": 0.0388,
      "step": 203
    },
    {
      "epoch": 0.6562123039806996,
      "grad_norm": 0.602165699005127,
      "learning_rate": 8.485421864737997e-06,
      "loss": 0.0557,
      "step": 204
    },
    {
      "epoch": 0.6594290309609971,
      "grad_norm": 1.2543528079986572,
      "learning_rate": 8.465183149535939e-06,
      "loss": 0.0539,
      "step": 205
    },
    {
      "epoch": 0.6626457579412948,
      "grad_norm": 1.1637508869171143,
      "learning_rate": 8.444834595378434e-06,
      "loss": 0.0526,
      "step": 206
    },
    {
      "epoch": 0.6658624849215923,
      "grad_norm": 2.013043165206909,
      "learning_rate": 8.424376847271483e-06,
      "loss": 0.0523,
      "step": 207
    },
    {
      "epoch": 0.6690792119018898,
      "grad_norm": 1.166269302368164,
      "learning_rate": 8.403810553682307e-06,
      "loss": 0.0482,
      "step": 208
    },
    {
      "epoch": 0.6722959388821874,
      "grad_norm": 1.0098134279251099,
      "learning_rate": 8.383136366518788e-06,
      "loss": 0.0477,
      "step": 209
    },
    {
      "epoch": 0.6755126658624849,
      "grad_norm": 0.7044484615325928,
      "learning_rate": 8.362354941108803e-06,
      "loss": 0.05,
      "step": 210
    },
    {
      "epoch": 0.6787293928427824,
      "grad_norm": 0.5775832533836365,
      "learning_rate": 8.341466936179457e-06,
      "loss": 0.0532,
      "step": 211
    },
    {
      "epoch": 0.6819461198230801,
      "grad_norm": 1.780887246131897,
      "learning_rate": 8.320473013836197e-06,
      "loss": 0.0385,
      "step": 212
    },
    {
      "epoch": 0.6851628468033776,
      "grad_norm": 0.6344701051712036,
      "learning_rate": 8.299373839541829e-06,
      "loss": 0.0345,
      "step": 213
    },
    {
      "epoch": 0.6883795737836751,
      "grad_norm": 0.597938597202301,
      "learning_rate": 8.278170082095422e-06,
      "loss": 0.0549,
      "step": 214
    },
    {
      "epoch": 0.6915963007639726,
      "grad_norm": 0.5144065022468567,
      "learning_rate": 8.256862413611113e-06,
      "loss": 0.0368,
      "step": 215
    },
    {
      "epoch": 0.6948130277442702,
      "grad_norm": 0.5459667444229126,
      "learning_rate": 8.23545150949679e-06,
      "loss": 0.0396,
      "step": 216
    },
    {
      "epoch": 0.6980297547245677,
      "grad_norm": 1.465198278427124,
      "learning_rate": 8.213938048432697e-06,
      "loss": 0.0534,
      "step": 217
    },
    {
      "epoch": 0.7012464817048653,
      "grad_norm": 0.9042430520057678,
      "learning_rate": 8.192322712349917e-06,
      "loss": 0.0463,
      "step": 218
    },
    {
      "epoch": 0.7044632086851629,
      "grad_norm": 1.0576472282409668,
      "learning_rate": 8.170606186408756e-06,
      "loss": 0.0392,
      "step": 219
    },
    {
      "epoch": 0.7076799356654604,
      "grad_norm": 0.7928400039672852,
      "learning_rate": 8.148789158977012e-06,
      "loss": 0.056,
      "step": 220
    },
    {
      "epoch": 0.7108966626457579,
      "grad_norm": 1.0211706161499023,
      "learning_rate": 8.126872321608185e-06,
      "loss": 0.059,
      "step": 221
    },
    {
      "epoch": 0.7141133896260555,
      "grad_norm": 0.5951483845710754,
      "learning_rate": 8.104856369019525e-06,
      "loss": 0.0493,
      "step": 222
    },
    {
      "epoch": 0.717330116606353,
      "grad_norm": 1.3035422563552856,
      "learning_rate": 8.08274199907003e-06,
      "loss": 0.0411,
      "step": 223
    },
    {
      "epoch": 0.7205468435866506,
      "grad_norm": 0.7945671081542969,
      "learning_rate": 8.060529912738316e-06,
      "loss": 0.0419,
      "step": 224
    },
    {
      "epoch": 0.7237635705669482,
      "grad_norm": 0.6303293704986572,
      "learning_rate": 8.038220814100403e-06,
      "loss": 0.0504,
      "step": 225
    },
    {
      "epoch": 0.7269802975472457,
      "grad_norm": 1.6979130506515503,
      "learning_rate": 8.0158154103074e-06,
      "loss": 0.0512,
      "step": 226
    },
    {
      "epoch": 0.7301970245275432,
      "grad_norm": 0.7927567362785339,
      "learning_rate": 7.993314411563075e-06,
      "loss": 0.0559,
      "step": 227
    },
    {
      "epoch": 0.7334137515078407,
      "grad_norm": 0.7262939214706421,
      "learning_rate": 7.970718531101365e-06,
      "loss": 0.0439,
      "step": 228
    },
    {
      "epoch": 0.7366304784881383,
      "grad_norm": 0.7317245006561279,
      "learning_rate": 7.948028485163744e-06,
      "loss": 0.0399,
      "step": 229
    },
    {
      "epoch": 0.7398472054684359,
      "grad_norm": 0.9586815237998962,
      "learning_rate": 7.925244992976538e-06,
      "loss": 0.0573,
      "step": 230
    },
    {
      "epoch": 0.7430639324487334,
      "grad_norm": 1.5452810525894165,
      "learning_rate": 7.902368776728125e-06,
      "loss": 0.0464,
      "step": 231
    },
    {
      "epoch": 0.746280659429031,
      "grad_norm": 0.8795072436332703,
      "learning_rate": 7.879400561546033e-06,
      "loss": 0.0453,
      "step": 232
    },
    {
      "epoch": 0.7494973864093285,
      "grad_norm": 0.648480236530304,
      "learning_rate": 7.856341075473963e-06,
      "loss": 0.0467,
      "step": 233
    },
    {
      "epoch": 0.752714113389626,
      "grad_norm": 0.646035373210907,
      "learning_rate": 7.833191049448706e-06,
      "loss": 0.0377,
      "step": 234
    },
    {
      "epoch": 0.7559308403699236,
      "grad_norm": 1.1588225364685059,
      "learning_rate": 7.809951217276986e-06,
      "loss": 0.0427,
      "step": 235
    },
    {
      "epoch": 0.7591475673502212,
      "grad_norm": 0.6012084484100342,
      "learning_rate": 7.786622315612182e-06,
      "loss": 0.0433,
      "step": 236
    },
    {
      "epoch": 0.7623642943305187,
      "grad_norm": 0.5877156853675842,
      "learning_rate": 7.763205083930995e-06,
      "loss": 0.0394,
      "step": 237
    },
    {
      "epoch": 0.7655810213108163,
      "grad_norm": 0.7161210179328918,
      "learning_rate": 7.739700264509993e-06,
      "loss": 0.0379,
      "step": 238
    },
    {
      "epoch": 0.7687977482911138,
      "grad_norm": 0.819977343082428,
      "learning_rate": 7.716108602402094e-06,
      "loss": 0.0599,
      "step": 239
    },
    {
      "epoch": 0.7720144752714113,
      "grad_norm": 0.6733528971672058,
      "learning_rate": 7.692430845412946e-06,
      "loss": 0.0465,
      "step": 240
    },
    {
      "epoch": 0.7752312022517089,
      "grad_norm": 0.6228499412536621,
      "learning_rate": 7.668667744077215e-06,
      "loss": 0.0478,
      "step": 241
    },
    {
      "epoch": 0.7784479292320065,
      "grad_norm": 0.8519704341888428,
      "learning_rate": 7.644820051634813e-06,
      "loss": 0.0503,
      "step": 242
    },
    {
      "epoch": 0.781664656212304,
      "grad_norm": 0.966986358165741,
      "learning_rate": 7.6208885240069995e-06,
      "loss": 0.0412,
      "step": 243
    },
    {
      "epoch": 0.7848813831926015,
      "grad_norm": 0.8694539070129395,
      "learning_rate": 7.596873919772438e-06,
      "loss": 0.0436,
      "step": 244
    },
    {
      "epoch": 0.7880981101728991,
      "grad_norm": 0.8280279040336609,
      "learning_rate": 7.572777000143145e-06,
      "loss": 0.0347,
      "step": 245
    },
    {
      "epoch": 0.7913148371531966,
      "grad_norm": 1.55968177318573,
      "learning_rate": 7.548598528940354e-06,
      "loss": 0.0478,
      "step": 246
    },
    {
      "epoch": 0.7945315641334941,
      "grad_norm": 0.5930286049842834,
      "learning_rate": 7.524339272570317e-06,
      "loss": 0.0511,
      "step": 247
    },
    {
      "epoch": 0.7977482911137918,
      "grad_norm": 1.2295900583267212,
      "learning_rate": 7.500000000000001e-06,
      "loss": 0.055,
      "step": 248
    },
    {
      "epoch": 0.8009650180940893,
      "grad_norm": 1.3309237957000732,
      "learning_rate": 7.475581482732717e-06,
      "loss": 0.0557,
      "step": 249
    },
    {
      "epoch": 0.8041817450743868,
      "grad_norm": 0.7122506499290466,
      "learning_rate": 7.451084494783668e-06,
      "loss": 0.0469,
      "step": 250
    },
    {
      "epoch": 0.8073984720546844,
      "grad_norm": 1.1560570001602173,
      "learning_rate": 7.4265098126554065e-06,
      "loss": 0.0489,
      "step": 251
    },
    {
      "epoch": 0.8106151990349819,
      "grad_norm": 1.4469757080078125,
      "learning_rate": 7.401858215313228e-06,
      "loss": 0.0501,
      "step": 252
    },
    {
      "epoch": 0.8138319260152794,
      "grad_norm": 0.8810122609138489,
      "learning_rate": 7.3771304841604764e-06,
      "loss": 0.0317,
      "step": 253
    },
    {
      "epoch": 0.817048652995577,
      "grad_norm": 0.5365167260169983,
      "learning_rate": 7.352327403013779e-06,
      "loss": 0.0414,
      "step": 254
    },
    {
      "epoch": 0.8202653799758746,
      "grad_norm": 0.6032179594039917,
      "learning_rate": 7.327449758078194e-06,
      "loss": 0.0413,
      "step": 255
    },
    {
      "epoch": 0.8234821069561721,
      "grad_norm": 0.7471963763237,
      "learning_rate": 7.302498337922293e-06,
      "loss": 0.0406,
      "step": 256
    },
    {
      "epoch": 0.8266988339364696,
      "grad_norm": 0.5315238237380981,
      "learning_rate": 7.27747393345317e-06,
      "loss": 0.0386,
      "step": 257
    },
    {
      "epoch": 0.8299155609167672,
      "grad_norm": 0.7365145683288574,
      "learning_rate": 7.2523773378913655e-06,
      "loss": 0.0554,
      "step": 258
    },
    {
      "epoch": 0.8331322878970647,
      "grad_norm": 1.0224618911743164,
      "learning_rate": 7.2272093467457226e-06,
      "loss": 0.0434,
      "step": 259
    },
    {
      "epoch": 0.8363490148773622,
      "grad_norm": 1.6006726026535034,
      "learning_rate": 7.201970757788172e-06,
      "loss": 0.0486,
      "step": 260
    },
    {
      "epoch": 0.8395657418576599,
      "grad_norm": 1.2492550611495972,
      "learning_rate": 7.17666237102845e-06,
      "loss": 0.0448,
      "step": 261
    },
    {
      "epoch": 0.8427824688379574,
      "grad_norm": 2.2926084995269775,
      "learning_rate": 7.151284988688731e-06,
      "loss": 0.044,
      "step": 262
    },
    {
      "epoch": 0.8459991958182549,
      "grad_norm": 0.7440332174301147,
      "learning_rate": 7.125839415178204e-06,
      "loss": 0.0477,
      "step": 263
    },
    {
      "epoch": 0.8492159227985525,
      "grad_norm": 0.717848002910614,
      "learning_rate": 7.100326457067576e-06,
      "loss": 0.0396,
      "step": 264
    },
    {
      "epoch": 0.85243264977885,
      "grad_norm": 0.6726682782173157,
      "learning_rate": 7.074746923063497e-06,
      "loss": 0.051,
      "step": 265
    },
    {
      "epoch": 0.8556493767591475,
      "grad_norm": 0.8255169987678528,
      "learning_rate": 7.049101623982938e-06,
      "loss": 0.0445,
      "step": 266
    },
    {
      "epoch": 0.8588661037394452,
      "grad_norm": 0.7894156575202942,
      "learning_rate": 7.02339137272748e-06,
      "loss": 0.0378,
      "step": 267
    },
    {
      "epoch": 0.8620828307197427,
      "grad_norm": 0.8157196640968323,
      "learning_rate": 6.9976169842575526e-06,
      "loss": 0.0395,
      "step": 268
    },
    {
      "epoch": 0.8652995577000402,
      "grad_norm": 0.613278329372406,
      "learning_rate": 6.971779275566593e-06,
      "loss": 0.032,
      "step": 269
    },
    {
      "epoch": 0.8685162846803377,
      "grad_norm": 0.6524770855903625,
      "learning_rate": 6.945879065655164e-06,
      "loss": 0.0463,
      "step": 270
    },
    {
      "epoch": 0.8717330116606353,
      "grad_norm": 0.8012259602546692,
      "learning_rate": 6.919917175504978e-06,
      "loss": 0.0556,
      "step": 271
    },
    {
      "epoch": 0.8749497386409328,
      "grad_norm": 0.8766384720802307,
      "learning_rate": 6.893894428052881e-06,
      "loss": 0.0534,
      "step": 272
    },
    {
      "epoch": 0.8781664656212304,
      "grad_norm": 0.6129264831542969,
      "learning_rate": 6.867811648164769e-06,
      "loss": 0.0486,
      "step": 273
    },
    {
      "epoch": 0.881383192601528,
      "grad_norm": 0.7514522075653076,
      "learning_rate": 6.841669662609437e-06,
      "loss": 0.0464,
      "step": 274
    },
    {
      "epoch": 0.8845999195818255,
      "grad_norm": 0.5673539042472839,
      "learning_rate": 6.815469300032374e-06,
      "loss": 0.0338,
      "step": 275
    },
    {
      "epoch": 0.887816646562123,
      "grad_norm": 0.5022168159484863,
      "learning_rate": 6.789211390929497e-06,
      "loss": 0.0283,
      "step": 276
    },
    {
      "epoch": 0.8910333735424206,
      "grad_norm": 1.195987343788147,
      "learning_rate": 6.762896767620827e-06,
      "loss": 0.0488,
      "step": 277
    },
    {
      "epoch": 0.8942501005227181,
      "grad_norm": 1.0466563701629639,
      "learning_rate": 6.736526264224101e-06,
      "loss": 0.0412,
      "step": 278
    },
    {
      "epoch": 0.8974668275030157,
      "grad_norm": 0.6366952061653137,
      "learning_rate": 6.710100716628345e-06,
      "loss": 0.0403,
      "step": 279
    },
    {
      "epoch": 0.9006835544833133,
      "grad_norm": 1.0278160572052002,
      "learning_rate": 6.6836209624673575e-06,
      "loss": 0.0527,
      "step": 280
    },
    {
      "epoch": 0.9039002814636108,
      "grad_norm": 0.8810920715332031,
      "learning_rate": 6.657087841093179e-06,
      "loss": 0.0517,
      "step": 281
    },
    {
      "epoch": 0.9071170084439083,
      "grad_norm": 1.306876301765442,
      "learning_rate": 6.6305021935494755e-06,
      "loss": 0.041,
      "step": 282
    },
    {
      "epoch": 0.9103337354242059,
      "grad_norm": 1.134096622467041,
      "learning_rate": 6.603864862544879e-06,
      "loss": 0.0474,
      "step": 283
    },
    {
      "epoch": 0.9135504624045034,
      "grad_norm": 1.364350438117981,
      "learning_rate": 6.5771766924262795e-06,
      "loss": 0.0418,
      "step": 284
    },
    {
      "epoch": 0.916767189384801,
      "grad_norm": 1.0049413442611694,
      "learning_rate": 6.5504385291520554e-06,
      "loss": 0.0386,
      "step": 285
    },
    {
      "epoch": 0.9199839163650985,
      "grad_norm": 0.5477365255355835,
      "learning_rate": 6.523651220265269e-06,
      "loss": 0.0394,
      "step": 286
    },
    {
      "epoch": 0.9232006433453961,
      "grad_norm": 1.8029471635818481,
      "learning_rate": 6.496815614866792e-06,
      "loss": 0.0486,
      "step": 287
    },
    {
      "epoch": 0.9264173703256936,
      "grad_norm": 1.1982808113098145,
      "learning_rate": 6.469932563588386e-06,
      "loss": 0.0385,
      "step": 288
    },
    {
      "epoch": 0.9296340973059911,
      "grad_norm": 1.8659429550170898,
      "learning_rate": 6.443002918565754e-06,
      "loss": 0.05,
      "step": 289
    },
    {
      "epoch": 0.9328508242862887,
      "grad_norm": 0.9505304098129272,
      "learning_rate": 6.41602753341152e-06,
      "loss": 0.0439,
      "step": 290
    },
    {
      "epoch": 0.9360675512665863,
      "grad_norm": 1.1647857427597046,
      "learning_rate": 6.389007263188176e-06,
      "loss": 0.0472,
      "step": 291
    },
    {
      "epoch": 0.9392842782468838,
      "grad_norm": 1.5960369110107422,
      "learning_rate": 6.361942964380967e-06,
      "loss": 0.0454,
      "step": 292
    },
    {
      "epoch": 0.9425010052271814,
      "grad_norm": 0.6902590990066528,
      "learning_rate": 6.334835494870759e-06,
      "loss": 0.0351,
      "step": 293
    },
    {
      "epoch": 0.9457177322074789,
      "grad_norm": 0.8172485828399658,
      "learning_rate": 6.307685713906835e-06,
      "loss": 0.0344,
      "step": 294
    },
    {
      "epoch": 0.9489344591877764,
      "grad_norm": 0.46139606833457947,
      "learning_rate": 6.2804944820796596e-06,
      "loss": 0.0308,
      "step": 295
    },
    {
      "epoch": 0.952151186168074,
      "grad_norm": 0.7500118613243103,
      "learning_rate": 6.2532626612936035e-06,
      "loss": 0.0541,
      "step": 296
    },
    {
      "epoch": 0.9553679131483716,
      "grad_norm": 0.693679928779602,
      "learning_rate": 6.225991114739622e-06,
      "loss": 0.0351,
      "step": 297
    },
    {
      "epoch": 0.9585846401286691,
      "grad_norm": 1.401061773300171,
      "learning_rate": 6.1986807068678926e-06,
      "loss": 0.0458,
      "step": 298
    },
    {
      "epoch": 0.9618013671089666,
      "grad_norm": 0.5701402425765991,
      "learning_rate": 6.171332303360411e-06,
      "loss": 0.0349,
      "step": 299
    },
    {
      "epoch": 0.9650180940892642,
      "grad_norm": 1.6502150297164917,
      "learning_rate": 6.143946771103561e-06,
      "loss": 0.0553,
      "step": 300
    },
    {
      "epoch": 0.9682348210695617,
      "grad_norm": 0.8593463897705078,
      "learning_rate": 6.11652497816062e-06,
      "loss": 0.0372,
      "step": 301
    },
    {
      "epoch": 0.9714515480498592,
      "grad_norm": 0.8635258674621582,
      "learning_rate": 6.089067793744258e-06,
      "loss": 0.0386,
      "step": 302
    },
    {
      "epoch": 0.9746682750301568,
      "grad_norm": 1.058119535446167,
      "learning_rate": 6.061576088188981e-06,
      "loss": 0.0407,
      "step": 303
    },
    {
      "epoch": 0.9778850020104544,
      "grad_norm": 1.316687822341919,
      "learning_rate": 6.034050732923538e-06,
      "loss": 0.0405,
      "step": 304
    },
    {
      "epoch": 0.9811017289907519,
      "grad_norm": 0.6700626611709595,
      "learning_rate": 6.006492600443301e-06,
      "loss": 0.0456,
      "step": 305
    },
    {
      "epoch": 0.9843184559710495,
      "grad_norm": 1.0319327116012573,
      "learning_rate": 5.978902564282616e-06,
      "loss": 0.0476,
      "step": 306
    },
    {
      "epoch": 0.987535182951347,
      "grad_norm": 0.5660614371299744,
      "learning_rate": 5.951281498987106e-06,
      "loss": 0.0431,
      "step": 307
    },
    {
      "epoch": 0.9907519099316445,
      "grad_norm": 0.5249197483062744,
      "learning_rate": 5.923630280085948e-06,
      "loss": 0.0416,
      "step": 308
    },
    {
      "epoch": 0.9939686369119421,
      "grad_norm": 0.5636727213859558,
      "learning_rate": 5.895949784064126e-06,
      "loss": 0.0316,
      "step": 309
    },
    {
      "epoch": 0.9971853638922397,
      "grad_norm": 0.5312305688858032,
      "learning_rate": 5.8682408883346535e-06,
      "loss": 0.0393,
      "step": 310
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.8194127678871155,
      "learning_rate": 5.840504471210742e-06,
      "loss": 0.0507,
      "step": 311
    },
    {
      "epoch": 1.0032167269802976,
      "grad_norm": 0.4939586818218231,
      "learning_rate": 5.8127414118779825e-06,
      "loss": 0.0331,
      "step": 312
    },
    {
      "epoch": 1.006433453960595,
      "grad_norm": 0.9499027729034424,
      "learning_rate": 5.7849525903664636e-06,
      "loss": 0.0234,
      "step": 313
    },
    {
      "epoch": 1.0096501809408926,
      "grad_norm": 1.0846751928329468,
      "learning_rate": 5.757138887522884e-06,
      "loss": 0.0244,
      "step": 314
    },
    {
      "epoch": 1.0128669079211903,
      "grad_norm": 0.39767810702323914,
      "learning_rate": 5.729301184982622e-06,
      "loss": 0.0334,
      "step": 315
    },
    {
      "epoch": 1.0160836349014877,
      "grad_norm": 0.38344645500183105,
      "learning_rate": 5.701440365141799e-06,
      "loss": 0.0301,
      "step": 316
    },
    {
      "epoch": 1.0193003618817853,
      "grad_norm": 0.7626691460609436,
      "learning_rate": 5.673557311129306e-06,
      "loss": 0.0322,
      "step": 317
    },
    {
      "epoch": 1.022517088862083,
      "grad_norm": 0.5533730387687683,
      "learning_rate": 5.645652906778808e-06,
      "loss": 0.0304,
      "step": 318
    },
    {
      "epoch": 1.0257338158423803,
      "grad_norm": 0.7238726615905762,
      "learning_rate": 5.617728036600734e-06,
      "loss": 0.0219,
      "step": 319
    },
    {
      "epoch": 1.028950542822678,
      "grad_norm": 0.6678081750869751,
      "learning_rate": 5.5897835857542315e-06,
      "loss": 0.0271,
      "step": 320
    },
    {
      "epoch": 1.0321672698029756,
      "grad_norm": 1.125412940979004,
      "learning_rate": 5.561820440019117e-06,
      "loss": 0.0281,
      "step": 321
    },
    {
      "epoch": 1.035383996783273,
      "grad_norm": 0.9026764631271362,
      "learning_rate": 5.533839485767795e-06,
      "loss": 0.0181,
      "step": 322
    },
    {
      "epoch": 1.0386007237635706,
      "grad_norm": 0.6368532776832581,
      "learning_rate": 5.505841609937162e-06,
      "loss": 0.0261,
      "step": 323
    },
    {
      "epoch": 1.041817450743868,
      "grad_norm": 0.48171374201774597,
      "learning_rate": 5.477827700000492e-06,
      "loss": 0.0168,
      "step": 324
    },
    {
      "epoch": 1.0450341777241656,
      "grad_norm": 0.7617553472518921,
      "learning_rate": 5.449798643939305e-06,
      "loss": 0.0294,
      "step": 325
    },
    {
      "epoch": 1.0482509047044632,
      "grad_norm": 0.7659939527511597,
      "learning_rate": 5.421755330215223e-06,
      "loss": 0.0233,
      "step": 326
    },
    {
      "epoch": 1.0514676316847607,
      "grad_norm": 1.1631779670715332,
      "learning_rate": 5.39369864774181e-06,
      "loss": 0.031,
      "step": 327
    },
    {
      "epoch": 1.0546843586650583,
      "grad_norm": 0.5692664384841919,
      "learning_rate": 5.365629485856381e-06,
      "loss": 0.017,
      "step": 328
    },
    {
      "epoch": 1.057901085645356,
      "grad_norm": 0.8848438858985901,
      "learning_rate": 5.337548734291827e-06,
      "loss": 0.0372,
      "step": 329
    },
    {
      "epoch": 1.0611178126256533,
      "grad_norm": 0.6400728225708008,
      "learning_rate": 5.30945728314841e-06,
      "loss": 0.0213,
      "step": 330
    },
    {
      "epoch": 1.064334539605951,
      "grad_norm": 0.9902735352516174,
      "learning_rate": 5.281356022865542e-06,
      "loss": 0.0262,
      "step": 331
    },
    {
      "epoch": 1.0675512665862485,
      "grad_norm": 0.5511563420295715,
      "learning_rate": 5.253245844193564e-06,
      "loss": 0.0212,
      "step": 332
    },
    {
      "epoch": 1.070767993566546,
      "grad_norm": 0.7277644276618958,
      "learning_rate": 5.225127638165514e-06,
      "loss": 0.0164,
      "step": 333
    },
    {
      "epoch": 1.0739847205468436,
      "grad_norm": 0.5540727376937866,
      "learning_rate": 5.197002296068878e-06,
      "loss": 0.024,
      "step": 334
    },
    {
      "epoch": 1.0772014475271412,
      "grad_norm": 0.55019611120224,
      "learning_rate": 5.168870709417342e-06,
      "loss": 0.0203,
      "step": 335
    },
    {
      "epoch": 1.0804181745074386,
      "grad_norm": 0.5772176384925842,
      "learning_rate": 5.140733769922525e-06,
      "loss": 0.021,
      "step": 336
    },
    {
      "epoch": 1.0836349014877362,
      "grad_norm": 0.7441113591194153,
      "learning_rate": 5.112592369465731e-06,
      "loss": 0.0372,
      "step": 337
    },
    {
      "epoch": 1.0868516284680338,
      "grad_norm": 0.5592713952064514,
      "learning_rate": 5.084447400069656e-06,
      "loss": 0.02,
      "step": 338
    },
    {
      "epoch": 1.0900683554483312,
      "grad_norm": 0.6708924770355225,
      "learning_rate": 5.0562997538701295e-06,
      "loss": 0.0207,
      "step": 339
    },
    {
      "epoch": 1.0932850824286289,
      "grad_norm": 0.6388556361198425,
      "learning_rate": 5.0281503230878304e-06,
      "loss": 0.0234,
      "step": 340
    },
    {
      "epoch": 1.0965018094089265,
      "grad_norm": 0.5663427114486694,
      "learning_rate": 5e-06,
      "loss": 0.02,
      "step": 341
    },
    {
      "epoch": 1.099718536389224,
      "grad_norm": 1.1264530420303345,
      "learning_rate": 4.971849676912172e-06,
      "loss": 0.0243,
      "step": 342
    },
    {
      "epoch": 1.1029352633695215,
      "grad_norm": 1.2352030277252197,
      "learning_rate": 4.943700246129871e-06,
      "loss": 0.0297,
      "step": 343
    },
    {
      "epoch": 1.1061519903498191,
      "grad_norm": 1.2657285928726196,
      "learning_rate": 4.915552599930345e-06,
      "loss": 0.0204,
      "step": 344
    },
    {
      "epoch": 1.1093687173301165,
      "grad_norm": 0.7982221245765686,
      "learning_rate": 4.887407630534271e-06,
      "loss": 0.0306,
      "step": 345
    },
    {
      "epoch": 1.1125854443104142,
      "grad_norm": 0.7376164197921753,
      "learning_rate": 4.859266230077474e-06,
      "loss": 0.0299,
      "step": 346
    },
    {
      "epoch": 1.1158021712907118,
      "grad_norm": 0.7835693359375,
      "learning_rate": 4.83112929058266e-06,
      "loss": 0.0231,
      "step": 347
    },
    {
      "epoch": 1.1190188982710092,
      "grad_norm": 1.1944645643234253,
      "learning_rate": 4.802997703931124e-06,
      "loss": 0.0277,
      "step": 348
    },
    {
      "epoch": 1.1222356252513068,
      "grad_norm": 0.6810235977172852,
      "learning_rate": 4.7748723618344865e-06,
      "loss": 0.0261,
      "step": 349
    },
    {
      "epoch": 1.1254523522316044,
      "grad_norm": 0.6911215782165527,
      "learning_rate": 4.746754155806437e-06,
      "loss": 0.0213,
      "step": 350
    },
    {
      "epoch": 1.1286690792119018,
      "grad_norm": 0.8558758497238159,
      "learning_rate": 4.71864397713446e-06,
      "loss": 0.0238,
      "step": 351
    },
    {
      "epoch": 1.1318858061921995,
      "grad_norm": 0.9453890919685364,
      "learning_rate": 4.6905427168515914e-06,
      "loss": 0.0303,
      "step": 352
    },
    {
      "epoch": 1.135102533172497,
      "grad_norm": 0.666305661201477,
      "learning_rate": 4.662451265708174e-06,
      "loss": 0.0188,
      "step": 353
    },
    {
      "epoch": 1.1383192601527945,
      "grad_norm": 0.7106337547302246,
      "learning_rate": 4.63437051414362e-06,
      "loss": 0.0305,
      "step": 354
    },
    {
      "epoch": 1.1415359871330921,
      "grad_norm": 0.5249149203300476,
      "learning_rate": 4.606301352258192e-06,
      "loss": 0.0186,
      "step": 355
    },
    {
      "epoch": 1.1447527141133897,
      "grad_norm": 0.6891714334487915,
      "learning_rate": 4.5782446697847775e-06,
      "loss": 0.0193,
      "step": 356
    },
    {
      "epoch": 1.1479694410936871,
      "grad_norm": 1.2555615901947021,
      "learning_rate": 4.550201356060695e-06,
      "loss": 0.0274,
      "step": 357
    },
    {
      "epoch": 1.1511861680739848,
      "grad_norm": 1.113802433013916,
      "learning_rate": 4.52217229999951e-06,
      "loss": 0.0256,
      "step": 358
    },
    {
      "epoch": 1.1544028950542822,
      "grad_norm": 0.6190158724784851,
      "learning_rate": 4.49415839006284e-06,
      "loss": 0.0265,
      "step": 359
    },
    {
      "epoch": 1.1576196220345798,
      "grad_norm": 0.8673491477966309,
      "learning_rate": 4.466160514232206e-06,
      "loss": 0.0341,
      "step": 360
    },
    {
      "epoch": 1.1608363490148774,
      "grad_norm": 0.4235200583934784,
      "learning_rate": 4.438179559980885e-06,
      "loss": 0.0163,
      "step": 361
    },
    {
      "epoch": 1.1640530759951748,
      "grad_norm": 0.9669196605682373,
      "learning_rate": 4.410216414245771e-06,
      "loss": 0.0357,
      "step": 362
    },
    {
      "epoch": 1.1672698029754724,
      "grad_norm": 0.8336294293403625,
      "learning_rate": 4.382271963399268e-06,
      "loss": 0.0161,
      "step": 363
    },
    {
      "epoch": 1.17048652995577,
      "grad_norm": 0.8520212769508362,
      "learning_rate": 4.354347093221194e-06,
      "loss": 0.0206,
      "step": 364
    },
    {
      "epoch": 1.1737032569360675,
      "grad_norm": 0.9719181060791016,
      "learning_rate": 4.326442688870697e-06,
      "loss": 0.0207,
      "step": 365
    },
    {
      "epoch": 1.176919983916365,
      "grad_norm": 0.6609673500061035,
      "learning_rate": 4.298559634858202e-06,
      "loss": 0.0288,
      "step": 366
    },
    {
      "epoch": 1.1801367108966627,
      "grad_norm": 0.9857617616653442,
      "learning_rate": 4.270698815017379e-06,
      "loss": 0.0279,
      "step": 367
    },
    {
      "epoch": 1.1833534378769601,
      "grad_norm": 0.788272500038147,
      "learning_rate": 4.2428611124771184e-06,
      "loss": 0.0144,
      "step": 368
    },
    {
      "epoch": 1.1865701648572577,
      "grad_norm": 0.8060084581375122,
      "learning_rate": 4.2150474096335356e-06,
      "loss": 0.023,
      "step": 369
    },
    {
      "epoch": 1.1897868918375554,
      "grad_norm": 1.2030549049377441,
      "learning_rate": 4.187258588122019e-06,
      "loss": 0.0355,
      "step": 370
    },
    {
      "epoch": 1.1930036188178528,
      "grad_norm": 0.6166083812713623,
      "learning_rate": 4.15949552878926e-06,
      "loss": 0.0165,
      "step": 371
    },
    {
      "epoch": 1.1962203457981504,
      "grad_norm": 1.5844203233718872,
      "learning_rate": 4.131759111665349e-06,
      "loss": 0.0323,
      "step": 372
    },
    {
      "epoch": 1.199437072778448,
      "grad_norm": 0.9516481757164001,
      "learning_rate": 4.104050215935875e-06,
      "loss": 0.0306,
      "step": 373
    },
    {
      "epoch": 1.2026537997587454,
      "grad_norm": 1.0186715126037598,
      "learning_rate": 4.076369719914055e-06,
      "loss": 0.0252,
      "step": 374
    },
    {
      "epoch": 1.205870526739043,
      "grad_norm": 0.891091525554657,
      "learning_rate": 4.048718501012895e-06,
      "loss": 0.0168,
      "step": 375
    },
    {
      "epoch": 1.2090872537193404,
      "grad_norm": 0.6966618299484253,
      "learning_rate": 4.021097435717386e-06,
      "loss": 0.0297,
      "step": 376
    },
    {
      "epoch": 1.212303980699638,
      "grad_norm": 0.9956583380699158,
      "learning_rate": 3.993507399556699e-06,
      "loss": 0.0189,
      "step": 377
    },
    {
      "epoch": 1.2155207076799357,
      "grad_norm": 0.6981390714645386,
      "learning_rate": 3.965949267076465e-06,
      "loss": 0.0175,
      "step": 378
    },
    {
      "epoch": 1.218737434660233,
      "grad_norm": 0.9271097779273987,
      "learning_rate": 3.938423911811021e-06,
      "loss": 0.0215,
      "step": 379
    },
    {
      "epoch": 1.2219541616405307,
      "grad_norm": 1.0739299058914185,
      "learning_rate": 3.910932206255742e-06,
      "loss": 0.0304,
      "step": 380
    },
    {
      "epoch": 1.2251708886208283,
      "grad_norm": 0.7238112092018127,
      "learning_rate": 3.883475021839382e-06,
      "loss": 0.0269,
      "step": 381
    },
    {
      "epoch": 1.2283876156011257,
      "grad_norm": 0.8449429273605347,
      "learning_rate": 3.856053228896442e-06,
      "loss": 0.0253,
      "step": 382
    },
    {
      "epoch": 1.2316043425814234,
      "grad_norm": 1.0511232614517212,
      "learning_rate": 3.8286676966395895e-06,
      "loss": 0.0286,
      "step": 383
    },
    {
      "epoch": 1.234821069561721,
      "grad_norm": 0.6601850390434265,
      "learning_rate": 3.8013192931321095e-06,
      "loss": 0.0161,
      "step": 384
    },
    {
      "epoch": 1.2380377965420184,
      "grad_norm": 1.7054966688156128,
      "learning_rate": 3.77400888526038e-06,
      "loss": 0.0382,
      "step": 385
    },
    {
      "epoch": 1.241254523522316,
      "grad_norm": 0.5929433703422546,
      "learning_rate": 3.7467373387063973e-06,
      "loss": 0.027,
      "step": 386
    },
    {
      "epoch": 1.2444712505026136,
      "grad_norm": 0.6725859642028809,
      "learning_rate": 3.719505517920342e-06,
      "loss": 0.0283,
      "step": 387
    },
    {
      "epoch": 1.247687977482911,
      "grad_norm": 0.6051467657089233,
      "learning_rate": 3.692314286093167e-06,
      "loss": 0.024,
      "step": 388
    },
    {
      "epoch": 1.2509047044632087,
      "grad_norm": 0.7530882358551025,
      "learning_rate": 3.6651645051292415e-06,
      "loss": 0.0226,
      "step": 389
    },
    {
      "epoch": 1.2541214314435063,
      "grad_norm": 0.6657713651657104,
      "learning_rate": 3.6380570356190346e-06,
      "loss": 0.0197,
      "step": 390
    },
    {
      "epoch": 1.2573381584238037,
      "grad_norm": 1.1519615650177002,
      "learning_rate": 3.610992736811827e-06,
      "loss": 0.0188,
      "step": 391
    },
    {
      "epoch": 1.2605548854041013,
      "grad_norm": 0.8955692648887634,
      "learning_rate": 3.58397246658848e-06,
      "loss": 0.0193,
      "step": 392
    },
    {
      "epoch": 1.263771612384399,
      "grad_norm": 1.1424133777618408,
      "learning_rate": 3.556997081434248e-06,
      "loss": 0.025,
      "step": 393
    },
    {
      "epoch": 1.2669883393646963,
      "grad_norm": 0.634136974811554,
      "learning_rate": 3.5300674364116173e-06,
      "loss": 0.0237,
      "step": 394
    },
    {
      "epoch": 1.270205066344994,
      "grad_norm": 0.9286861419677734,
      "learning_rate": 3.5031843851332105e-06,
      "loss": 0.0231,
      "step": 395
    },
    {
      "epoch": 1.2734217933252916,
      "grad_norm": 0.7624120712280273,
      "learning_rate": 3.476348779734732e-06,
      "loss": 0.0246,
      "step": 396
    },
    {
      "epoch": 1.276638520305589,
      "grad_norm": 0.8140252828598022,
      "learning_rate": 3.449561470847947e-06,
      "loss": 0.0262,
      "step": 397
    },
    {
      "epoch": 1.2798552472858866,
      "grad_norm": 1.130080223083496,
      "learning_rate": 3.4228233075737225e-06,
      "loss": 0.0294,
      "step": 398
    },
    {
      "epoch": 1.2830719742661842,
      "grad_norm": 0.8101189136505127,
      "learning_rate": 3.3961351374551234e-06,
      "loss": 0.0279,
      "step": 399
    },
    {
      "epoch": 1.2862887012464816,
      "grad_norm": 0.5413293838500977,
      "learning_rate": 3.3694978064505258e-06,
      "loss": 0.0104,
      "step": 400
    },
    {
      "epoch": 1.2895054282267793,
      "grad_norm": 0.940980851650238,
      "learning_rate": 3.3429121589068213e-06,
      "loss": 0.023,
      "step": 401
    },
    {
      "epoch": 1.292722155207077,
      "grad_norm": 1.4094127416610718,
      "learning_rate": 3.316379037532644e-06,
      "loss": 0.0238,
      "step": 402
    },
    {
      "epoch": 1.2959388821873743,
      "grad_norm": 0.6321690082550049,
      "learning_rate": 3.289899283371657e-06,
      "loss": 0.0233,
      "step": 403
    },
    {
      "epoch": 1.299155609167672,
      "grad_norm": 0.7912178635597229,
      "learning_rate": 3.2634737357758994e-06,
      "loss": 0.0247,
      "step": 404
    },
    {
      "epoch": 1.3023723361479695,
      "grad_norm": 0.808509111404419,
      "learning_rate": 3.2371032323791757e-06,
      "loss": 0.0209,
      "step": 405
    },
    {
      "epoch": 1.305589063128267,
      "grad_norm": 0.5239757895469666,
      "learning_rate": 3.2107886090705035e-06,
      "loss": 0.0188,
      "step": 406
    },
    {
      "epoch": 1.3088057901085646,
      "grad_norm": 0.7358213067054749,
      "learning_rate": 3.1845306999676274e-06,
      "loss": 0.0328,
      "step": 407
    },
    {
      "epoch": 1.3120225170888622,
      "grad_norm": 0.8183391690254211,
      "learning_rate": 3.158330337390565e-06,
      "loss": 0.0263,
      "step": 408
    },
    {
      "epoch": 1.3152392440691596,
      "grad_norm": 0.7185299396514893,
      "learning_rate": 3.132188351835232e-06,
      "loss": 0.0316,
      "step": 409
    },
    {
      "epoch": 1.3184559710494572,
      "grad_norm": 0.7911690473556519,
      "learning_rate": 3.10610557194712e-06,
      "loss": 0.0229,
      "step": 410
    },
    {
      "epoch": 1.3216726980297548,
      "grad_norm": 0.5570195317268372,
      "learning_rate": 3.080082824495024e-06,
      "loss": 0.0276,
      "step": 411
    },
    {
      "epoch": 1.3248894250100522,
      "grad_norm": 0.760032057762146,
      "learning_rate": 3.0541209343448373e-06,
      "loss": 0.0228,
      "step": 412
    },
    {
      "epoch": 1.3281061519903499,
      "grad_norm": 0.6067129373550415,
      "learning_rate": 3.0282207244334084e-06,
      "loss": 0.0211,
      "step": 413
    },
    {
      "epoch": 1.3313228789706475,
      "grad_norm": 0.8982436656951904,
      "learning_rate": 3.0023830157424504e-06,
      "loss": 0.0269,
      "step": 414
    },
    {
      "epoch": 1.334539605950945,
      "grad_norm": 0.9359907507896423,
      "learning_rate": 2.97660862727252e-06,
      "loss": 0.0226,
      "step": 415
    },
    {
      "epoch": 1.3377563329312425,
      "grad_norm": 0.5233786106109619,
      "learning_rate": 2.950898376017064e-06,
      "loss": 0.0226,
      "step": 416
    },
    {
      "epoch": 1.3409730599115401,
      "grad_norm": 0.5477368235588074,
      "learning_rate": 2.9252530769365053e-06,
      "loss": 0.0177,
      "step": 417
    },
    {
      "epoch": 1.3441897868918375,
      "grad_norm": 0.7857730984687805,
      "learning_rate": 2.8996735429324256e-06,
      "loss": 0.0208,
      "step": 418
    },
    {
      "epoch": 1.3474065138721352,
      "grad_norm": 0.6837164163589478,
      "learning_rate": 2.874160584821798e-06,
      "loss": 0.0204,
      "step": 419
    },
    {
      "epoch": 1.3506232408524326,
      "grad_norm": 0.5833801627159119,
      "learning_rate": 2.848715011311271e-06,
      "loss": 0.0179,
      "step": 420
    },
    {
      "epoch": 1.3538399678327302,
      "grad_norm": 0.7275317311286926,
      "learning_rate": 2.823337628971551e-06,
      "loss": 0.0176,
      "step": 421
    },
    {
      "epoch": 1.3570566948130278,
      "grad_norm": 1.1249991655349731,
      "learning_rate": 2.7980292422118282e-06,
      "loss": 0.0264,
      "step": 422
    },
    {
      "epoch": 1.3602734217933252,
      "grad_norm": 0.8892255425453186,
      "learning_rate": 2.7727906532542783e-06,
      "loss": 0.021,
      "step": 423
    },
    {
      "epoch": 1.3634901487736228,
      "grad_norm": 0.8986218571662903,
      "learning_rate": 2.7476226621086354e-06,
      "loss": 0.0209,
      "step": 424
    },
    {
      "epoch": 1.3667068757539202,
      "grad_norm": 0.8153395056724548,
      "learning_rate": 2.72252606654683e-06,
      "loss": 0.0209,
      "step": 425
    },
    {
      "epoch": 1.3699236027342179,
      "grad_norm": 0.8225369453430176,
      "learning_rate": 2.697501662077707e-06,
      "loss": 0.0299,
      "step": 426
    },
    {
      "epoch": 1.3731403297145155,
      "grad_norm": 0.8588507175445557,
      "learning_rate": 2.6725502419218084e-06,
      "loss": 0.0346,
      "step": 427
    },
    {
      "epoch": 1.376357056694813,
      "grad_norm": 1.9007744789123535,
      "learning_rate": 2.6476725969862227e-06,
      "loss": 0.0418,
      "step": 428
    },
    {
      "epoch": 1.3795737836751105,
      "grad_norm": 1.121382474899292,
      "learning_rate": 2.622869515839524e-06,
      "loss": 0.0179,
      "step": 429
    },
    {
      "epoch": 1.3827905106554081,
      "grad_norm": 0.7519135475158691,
      "learning_rate": 2.5981417846867753e-06,
      "loss": 0.0242,
      "step": 430
    },
    {
      "epoch": 1.3860072376357055,
      "grad_norm": 0.7129932641983032,
      "learning_rate": 2.573490187344596e-06,
      "loss": 0.0137,
      "step": 431
    },
    {
      "epoch": 1.3892239646160032,
      "grad_norm": 0.7935872673988342,
      "learning_rate": 2.548915505216333e-06,
      "loss": 0.0287,
      "step": 432
    },
    {
      "epoch": 1.3924406915963008,
      "grad_norm": 1.06389582157135,
      "learning_rate": 2.524418517267283e-06,
      "loss": 0.0251,
      "step": 433
    },
    {
      "epoch": 1.3956574185765982,
      "grad_norm": 0.8240230679512024,
      "learning_rate": 2.5000000000000015e-06,
      "loss": 0.0297,
      "step": 434
    },
    {
      "epoch": 1.3988741455568958,
      "grad_norm": 0.9097860455513,
      "learning_rate": 2.4756607274296844e-06,
      "loss": 0.0236,
      "step": 435
    },
    {
      "epoch": 1.4020908725371934,
      "grad_norm": 0.634899377822876,
      "learning_rate": 2.4514014710596467e-06,
      "loss": 0.0205,
      "step": 436
    },
    {
      "epoch": 1.4053075995174908,
      "grad_norm": 1.0875910520553589,
      "learning_rate": 2.4272229998568576e-06,
      "loss": 0.0367,
      "step": 437
    },
    {
      "epoch": 1.4085243264977885,
      "grad_norm": 1.6455551385879517,
      "learning_rate": 2.4031260802275623e-06,
      "loss": 0.0271,
      "step": 438
    },
    {
      "epoch": 1.411741053478086,
      "grad_norm": 0.7204722166061401,
      "learning_rate": 2.3791114759930013e-06,
      "loss": 0.0188,
      "step": 439
    },
    {
      "epoch": 1.4149577804583835,
      "grad_norm": 1.0633268356323242,
      "learning_rate": 2.3551799483651894e-06,
      "loss": 0.0264,
      "step": 440
    },
    {
      "epoch": 1.4181745074386811,
      "grad_norm": 0.6467992663383484,
      "learning_rate": 2.331332255922784e-06,
      "loss": 0.0184,
      "step": 441
    },
    {
      "epoch": 1.4213912344189787,
      "grad_norm": 0.5554575324058533,
      "learning_rate": 2.307569154587056e-06,
      "loss": 0.0184,
      "step": 442
    },
    {
      "epoch": 1.4246079613992761,
      "grad_norm": 0.6711392998695374,
      "learning_rate": 2.283891397597908e-06,
      "loss": 0.0206,
      "step": 443
    },
    {
      "epoch": 1.4278246883795738,
      "grad_norm": 0.7576943635940552,
      "learning_rate": 2.2602997354900075e-06,
      "loss": 0.031,
      "step": 444
    },
    {
      "epoch": 1.4310414153598714,
      "grad_norm": 0.6453703045845032,
      "learning_rate": 2.236794916069007e-06,
      "loss": 0.0171,
      "step": 445
    },
    {
      "epoch": 1.4342581423401688,
      "grad_norm": 0.5706139802932739,
      "learning_rate": 2.2133776843878185e-06,
      "loss": 0.0191,
      "step": 446
    },
    {
      "epoch": 1.4374748693204664,
      "grad_norm": 0.7611203789710999,
      "learning_rate": 2.190048782723015e-06,
      "loss": 0.0225,
      "step": 447
    },
    {
      "epoch": 1.440691596300764,
      "grad_norm": 0.6822800040245056,
      "learning_rate": 2.166808950551296e-06,
      "loss": 0.0256,
      "step": 448
    },
    {
      "epoch": 1.4439083232810614,
      "grad_norm": 0.798964262008667,
      "learning_rate": 2.1436589245260375e-06,
      "loss": 0.0236,
      "step": 449
    },
    {
      "epoch": 1.447125050261359,
      "grad_norm": 0.8539571166038513,
      "learning_rate": 2.120599438453968e-06,
      "loss": 0.0241,
      "step": 450
    },
    {
      "epoch": 1.4503417772416567,
      "grad_norm": 0.8464952707290649,
      "learning_rate": 2.0976312232718763e-06,
      "loss": 0.0282,
      "step": 451
    },
    {
      "epoch": 1.453558504221954,
      "grad_norm": 0.7654972076416016,
      "learning_rate": 2.074755007023461e-06,
      "loss": 0.0278,
      "step": 452
    },
    {
      "epoch": 1.4567752312022517,
      "grad_norm": 0.9112337231636047,
      "learning_rate": 2.0519715148362585e-06,
      "loss": 0.0176,
      "step": 453
    },
    {
      "epoch": 1.4599919581825493,
      "grad_norm": 0.6856303215026855,
      "learning_rate": 2.0292814688986375e-06,
      "loss": 0.0224,
      "step": 454
    },
    {
      "epoch": 1.4632086851628467,
      "grad_norm": 0.9206861257553101,
      "learning_rate": 2.0066855884369246e-06,
      "loss": 0.0312,
      "step": 455
    },
    {
      "epoch": 1.4664254121431444,
      "grad_norm": 0.9237130880355835,
      "learning_rate": 1.9841845896926022e-06,
      "loss": 0.0256,
      "step": 456
    },
    {
      "epoch": 1.469642139123442,
      "grad_norm": 0.5882164835929871,
      "learning_rate": 1.961779185899597e-06,
      "loss": 0.0224,
      "step": 457
    },
    {
      "epoch": 1.4728588661037394,
      "grad_norm": 0.5949448347091675,
      "learning_rate": 1.9394700872616856e-06,
      "loss": 0.0197,
      "step": 458
    },
    {
      "epoch": 1.476075593084037,
      "grad_norm": 0.5350119471549988,
      "learning_rate": 1.9172580009299735e-06,
      "loss": 0.0164,
      "step": 459
    },
    {
      "epoch": 1.4792923200643346,
      "grad_norm": 1.140408992767334,
      "learning_rate": 1.8951436309804766e-06,
      "loss": 0.0344,
      "step": 460
    },
    {
      "epoch": 1.482509047044632,
      "grad_norm": 0.7622132301330566,
      "learning_rate": 1.8731276783918162e-06,
      "loss": 0.0195,
      "step": 461
    },
    {
      "epoch": 1.4857257740249297,
      "grad_norm": 1.5069661140441895,
      "learning_rate": 1.8512108410229878e-06,
      "loss": 0.0306,
      "step": 462
    },
    {
      "epoch": 1.4889425010052273,
      "grad_norm": 0.7332004904747009,
      "learning_rate": 1.8293938135912475e-06,
      "loss": 0.0217,
      "step": 463
    },
    {
      "epoch": 1.4921592279855247,
      "grad_norm": 0.7239487767219543,
      "learning_rate": 1.8076772876500831e-06,
      "loss": 0.0298,
      "step": 464
    },
    {
      "epoch": 1.4953759549658223,
      "grad_norm": 0.8309620022773743,
      "learning_rate": 1.7860619515673034e-06,
      "loss": 0.0316,
      "step": 465
    },
    {
      "epoch": 1.49859268194612,
      "grad_norm": 0.6418971419334412,
      "learning_rate": 1.7645484905032129e-06,
      "loss": 0.0182,
      "step": 466
    },
    {
      "epoch": 1.5018094089264173,
      "grad_norm": 0.5187109112739563,
      "learning_rate": 1.74313758638889e-06,
      "loss": 0.025,
      "step": 467
    },
    {
      "epoch": 1.505026135906715,
      "grad_norm": 0.9927024245262146,
      "learning_rate": 1.7218299179045789e-06,
      "loss": 0.0256,
      "step": 468
    },
    {
      "epoch": 1.5082428628870126,
      "grad_norm": 0.6816557049751282,
      "learning_rate": 1.7006261604581725e-06,
      "loss": 0.0144,
      "step": 469
    },
    {
      "epoch": 1.51145958986731,
      "grad_norm": 0.6767432689666748,
      "learning_rate": 1.6795269861638041e-06,
      "loss": 0.0256,
      "step": 470
    },
    {
      "epoch": 1.5146763168476074,
      "grad_norm": 0.8080036044120789,
      "learning_rate": 1.6585330638205454e-06,
      "loss": 0.0283,
      "step": 471
    },
    {
      "epoch": 1.5178930438279052,
      "grad_norm": 0.7607430219650269,
      "learning_rate": 1.6376450588911985e-06,
      "loss": 0.0223,
      "step": 472
    },
    {
      "epoch": 1.5211097708082026,
      "grad_norm": 0.4729120433330536,
      "learning_rate": 1.6168636334812126e-06,
      "loss": 0.0142,
      "step": 473
    },
    {
      "epoch": 1.5243264977885,
      "grad_norm": 0.5791823267936707,
      "learning_rate": 1.5961894463176942e-06,
      "loss": 0.0258,
      "step": 474
    },
    {
      "epoch": 1.527543224768798,
      "grad_norm": 1.1532747745513916,
      "learning_rate": 1.5756231527285181e-06,
      "loss": 0.027,
      "step": 475
    },
    {
      "epoch": 1.5307599517490953,
      "grad_norm": 0.7709406018257141,
      "learning_rate": 1.555165404621567e-06,
      "loss": 0.0273,
      "step": 476
    },
    {
      "epoch": 1.5339766787293927,
      "grad_norm": 1.0912410020828247,
      "learning_rate": 1.5348168504640631e-06,
      "loss": 0.0282,
      "step": 477
    },
    {
      "epoch": 1.5371934057096905,
      "grad_norm": 0.7054423689842224,
      "learning_rate": 1.5145781352620054e-06,
      "loss": 0.0273,
      "step": 478
    },
    {
      "epoch": 1.540410132689988,
      "grad_norm": 0.46668004989624023,
      "learning_rate": 1.4944499005397372e-06,
      "loss": 0.0164,
      "step": 479
    },
    {
      "epoch": 1.5436268596702853,
      "grad_norm": 0.7803506851196289,
      "learning_rate": 1.4744327843196043e-06,
      "loss": 0.0283,
      "step": 480
    },
    {
      "epoch": 1.5468435866505832,
      "grad_norm": 0.7836430668830872,
      "learning_rate": 1.4545274211017264e-06,
      "loss": 0.0278,
      "step": 481
    },
    {
      "epoch": 1.5500603136308806,
      "grad_norm": 1.152942419052124,
      "learning_rate": 1.434734441843899e-06,
      "loss": 0.0273,
      "step": 482
    },
    {
      "epoch": 1.553277040611178,
      "grad_norm": 0.7189183831214905,
      "learning_rate": 1.4150544739415755e-06,
      "loss": 0.0238,
      "step": 483
    },
    {
      "epoch": 1.5564937675914756,
      "grad_norm": 1.2334411144256592,
      "learning_rate": 1.3954881412079945e-06,
      "loss": 0.0296,
      "step": 484
    },
    {
      "epoch": 1.5597104945717732,
      "grad_norm": 1.2712557315826416,
      "learning_rate": 1.3760360638544012e-06,
      "loss": 0.0258,
      "step": 485
    },
    {
      "epoch": 1.5629272215520706,
      "grad_norm": 0.6514741778373718,
      "learning_rate": 1.3566988584703817e-06,
      "loss": 0.0256,
      "step": 486
    },
    {
      "epoch": 1.5661439485323683,
      "grad_norm": 0.7159017324447632,
      "learning_rate": 1.3374771380043306e-06,
      "loss": 0.0221,
      "step": 487
    },
    {
      "epoch": 1.569360675512666,
      "grad_norm": 0.7216795086860657,
      "learning_rate": 1.3183715117440143e-06,
      "loss": 0.0293,
      "step": 488
    },
    {
      "epoch": 1.5725774024929633,
      "grad_norm": 0.5431943535804749,
      "learning_rate": 1.2993825852972559e-06,
      "loss": 0.0152,
      "step": 489
    },
    {
      "epoch": 1.575794129473261,
      "grad_norm": 1.1157512664794922,
      "learning_rate": 1.280510960572745e-06,
      "loss": 0.0297,
      "step": 490
    },
    {
      "epoch": 1.5790108564535585,
      "grad_norm": 0.8432955741882324,
      "learning_rate": 1.2617572357609565e-06,
      "loss": 0.023,
      "step": 491
    },
    {
      "epoch": 1.582227583433856,
      "grad_norm": 0.8817299008369446,
      "learning_rate": 1.2431220053151832e-06,
      "loss": 0.0276,
      "step": 492
    },
    {
      "epoch": 1.5854443104141536,
      "grad_norm": 0.6438712477684021,
      "learning_rate": 1.2246058599327021e-06,
      "loss": 0.0197,
      "step": 493
    },
    {
      "epoch": 1.5886610373944512,
      "grad_norm": 0.947399377822876,
      "learning_rate": 1.2062093865360458e-06,
      "loss": 0.0266,
      "step": 494
    },
    {
      "epoch": 1.5918777643747486,
      "grad_norm": 0.8782487511634827,
      "learning_rate": 1.1879331682543972e-06,
      "loss": 0.0146,
      "step": 495
    },
    {
      "epoch": 1.5950944913550462,
      "grad_norm": 0.676629364490509,
      "learning_rate": 1.1697777844051105e-06,
      "loss": 0.0266,
      "step": 496
    },
    {
      "epoch": 1.5983112183353438,
      "grad_norm": 1.0170975923538208,
      "learning_rate": 1.1517438104753386e-06,
      "loss": 0.0278,
      "step": 497
    },
    {
      "epoch": 1.6015279453156412,
      "grad_norm": 0.8391772508621216,
      "learning_rate": 1.1338318181038037e-06,
      "loss": 0.0264,
      "step": 498
    },
    {
      "epoch": 1.6047446722959389,
      "grad_norm": 0.6024744510650635,
      "learning_rate": 1.1160423750626693e-06,
      "loss": 0.0297,
      "step": 499
    },
    {
      "epoch": 1.6079613992762365,
      "grad_norm": 0.5469645261764526,
      "learning_rate": 1.0983760452395415e-06,
      "loss": 0.0181,
      "step": 500
    },
    {
      "epoch": 1.611178126256534,
      "grad_norm": 0.6368651986122131,
      "learning_rate": 1.0808333886196038e-06,
      "loss": 0.0203,
      "step": 501
    },
    {
      "epoch": 1.6143948532368315,
      "grad_norm": 0.6552625894546509,
      "learning_rate": 1.063414961267859e-06,
      "loss": 0.0243,
      "step": 502
    },
    {
      "epoch": 1.6176115802171291,
      "grad_norm": 0.5399875640869141,
      "learning_rate": 1.046121315311508e-06,
      "loss": 0.0201,
      "step": 503
    },
    {
      "epoch": 1.6208283071974265,
      "grad_norm": 0.7226431369781494,
      "learning_rate": 1.02895299892244e-06,
      "loss": 0.0269,
      "step": 504
    },
    {
      "epoch": 1.6240450341777242,
      "grad_norm": 0.7085054516792297,
      "learning_rate": 1.01191055629987e-06,
      "loss": 0.0253,
      "step": 505
    },
    {
      "epoch": 1.6272617611580218,
      "grad_norm": 0.7027333974838257,
      "learning_rate": 9.949945276530782e-07,
      "loss": 0.0202,
      "step": 506
    },
    {
      "epoch": 1.6304784881383192,
      "grad_norm": 0.669688880443573,
      "learning_rate": 9.782054491842879e-07,
      "loss": 0.0184,
      "step": 507
    },
    {
      "epoch": 1.6336952151186168,
      "grad_norm": 0.7410572171211243,
      "learning_rate": 9.615438530716753e-07,
      "loss": 0.0144,
      "step": 508
    },
    {
      "epoch": 1.6369119420989144,
      "grad_norm": 0.8965449333190918,
      "learning_rate": 9.450102674524952e-07,
      "loss": 0.0206,
      "step": 509
    },
    {
      "epoch": 1.6401286690792118,
      "grad_norm": 0.8675963878631592,
      "learning_rate": 9.286052164063369e-07,
      "loss": 0.0267,
      "step": 510
    },
    {
      "epoch": 1.6433453960595095,
      "grad_norm": 0.9799548387527466,
      "learning_rate": 9.123292199385247e-07,
      "loss": 0.0334,
      "step": 511
    },
    {
      "epoch": 1.646562123039807,
      "grad_norm": 0.5996445417404175,
      "learning_rate": 8.961827939636198e-07,
      "loss": 0.023,
      "step": 512
    },
    {
      "epoch": 1.6497788500201045,
      "grad_norm": 0.6122687458992004,
      "learning_rate": 8.801664502890722e-07,
      "loss": 0.0214,
      "step": 513
    },
    {
      "epoch": 1.6529955770004021,
      "grad_norm": 0.6234684586524963,
      "learning_rate": 8.64280696599008e-07,
      "loss": 0.0215,
      "step": 514
    },
    {
      "epoch": 1.6562123039806997,
      "grad_norm": 0.5246209502220154,
      "learning_rate": 8.485260364381187e-07,
      "loss": 0.0186,
      "step": 515
    },
    {
      "epoch": 1.6594290309609971,
      "grad_norm": 1.306154489517212,
      "learning_rate": 8.329029691957124e-07,
      "loss": 0.0257,
      "step": 516
    },
    {
      "epoch": 1.6626457579412948,
      "grad_norm": 0.8695035576820374,
      "learning_rate": 8.17411990089883e-07,
      "loss": 0.0389,
      "step": 517
    },
    {
      "epoch": 1.6658624849215924,
      "grad_norm": 0.5970683693885803,
      "learning_rate": 8.02053590151805e-07,
      "loss": 0.0281,
      "step": 518
    },
    {
      "epoch": 1.6690792119018898,
      "grad_norm": 0.7056528329849243,
      "learning_rate": 7.868282562101819e-07,
      "loss": 0.0212,
      "step": 519
    },
    {
      "epoch": 1.6722959388821874,
      "grad_norm": 0.8744556307792664,
      "learning_rate": 7.717364708758024e-07,
      "loss": 0.0265,
      "step": 520
    },
    {
      "epoch": 1.675512665862485,
      "grad_norm": 0.8215451240539551,
      "learning_rate": 7.567787125262449e-07,
      "loss": 0.0197,
      "step": 521
    },
    {
      "epoch": 1.6787293928427824,
      "grad_norm": 1.1487066745758057,
      "learning_rate": 7.41955455290726e-07,
      "loss": 0.03,
      "step": 522
    },
    {
      "epoch": 1.68194611982308,
      "grad_norm": 0.7303391098976135,
      "learning_rate": 7.27267169035053e-07,
      "loss": 0.0253,
      "step": 523
    },
    {
      "epoch": 1.6851628468033777,
      "grad_norm": 0.5748582482337952,
      "learning_rate": 7.127143193467445e-07,
      "loss": 0.0209,
      "step": 524
    },
    {
      "epoch": 1.688379573783675,
      "grad_norm": 1.0102810859680176,
      "learning_rate": 6.982973675202676e-07,
      "loss": 0.0263,
      "step": 525
    },
    {
      "epoch": 1.6915963007639725,
      "grad_norm": 0.7919561862945557,
      "learning_rate": 6.840167705424106e-07,
      "loss": 0.0268,
      "step": 526
    },
    {
      "epoch": 1.6948130277442703,
      "grad_norm": 0.792404294013977,
      "learning_rate": 6.698729810778065e-07,
      "loss": 0.0158,
      "step": 527
    },
    {
      "epoch": 1.6980297547245677,
      "grad_norm": 1.081179141998291,
      "learning_rate": 6.558664474545817e-07,
      "loss": 0.0312,
      "step": 528
    },
    {
      "epoch": 1.7012464817048651,
      "grad_norm": 0.6410866975784302,
      "learning_rate": 6.419976136501377e-07,
      "loss": 0.0188,
      "step": 529
    },
    {
      "epoch": 1.704463208685163,
      "grad_norm": 0.5669673085212708,
      "learning_rate": 6.282669192770896e-07,
      "loss": 0.0164,
      "step": 530
    },
    {
      "epoch": 1.7076799356654604,
      "grad_norm": 0.6292052268981934,
      "learning_rate": 6.146747995693225e-07,
      "loss": 0.0251,
      "step": 531
    },
    {
      "epoch": 1.7108966626457578,
      "grad_norm": 0.8831068277359009,
      "learning_rate": 6.012216853682001e-07,
      "loss": 0.0271,
      "step": 532
    },
    {
      "epoch": 1.7141133896260556,
      "grad_norm": 0.8742364048957825,
      "learning_rate": 5.879080031089047e-07,
      "loss": 0.0297,
      "step": 533
    },
    {
      "epoch": 1.717330116606353,
      "grad_norm": 0.6837791800498962,
      "learning_rate": 5.747341748069229e-07,
      "loss": 0.0184,
      "step": 534
    },
    {
      "epoch": 1.7205468435866504,
      "grad_norm": 0.7482956647872925,
      "learning_rate": 5.617006180446688e-07,
      "loss": 0.0237,
      "step": 535
    },
    {
      "epoch": 1.7237635705669483,
      "grad_norm": 0.5687646269798279,
      "learning_rate": 5.488077459582425e-07,
      "loss": 0.0219,
      "step": 536
    },
    {
      "epoch": 1.7269802975472457,
      "grad_norm": 0.5424864292144775,
      "learning_rate": 5.360559672243421e-07,
      "loss": 0.0228,
      "step": 537
    },
    {
      "epoch": 1.730197024527543,
      "grad_norm": 0.6900858879089355,
      "learning_rate": 5.234456860473042e-07,
      "loss": 0.0227,
      "step": 538
    },
    {
      "epoch": 1.7334137515078407,
      "grad_norm": 0.5527268648147583,
      "learning_rate": 5.109773021462921e-07,
      "loss": 0.0156,
      "step": 539
    },
    {
      "epoch": 1.7366304784881383,
      "grad_norm": 0.7617378234863281,
      "learning_rate": 4.986512107426283e-07,
      "loss": 0.0308,
      "step": 540
    },
    {
      "epoch": 1.7398472054684357,
      "grad_norm": 0.8059920072555542,
      "learning_rate": 4.864678025472635e-07,
      "loss": 0.0119,
      "step": 541
    },
    {
      "epoch": 1.7430639324487334,
      "grad_norm": 0.8831843137741089,
      "learning_rate": 4.7442746374839363e-07,
      "loss": 0.0267,
      "step": 542
    },
    {
      "epoch": 1.746280659429031,
      "grad_norm": 0.5986456274986267,
      "learning_rate": 4.625305759992205e-07,
      "loss": 0.0185,
      "step": 543
    },
    {
      "epoch": 1.7494973864093284,
      "grad_norm": 0.9453976154327393,
      "learning_rate": 4.50777516405847e-07,
      "loss": 0.0244,
      "step": 544
    },
    {
      "epoch": 1.752714113389626,
      "grad_norm": 0.7596896290779114,
      "learning_rate": 4.3916865751533313e-07,
      "loss": 0.024,
      "step": 545
    },
    {
      "epoch": 1.7559308403699236,
      "grad_norm": 0.9080878496170044,
      "learning_rate": 4.2770436730388166e-07,
      "loss": 0.0281,
      "step": 546
    },
    {
      "epoch": 1.759147567350221,
      "grad_norm": 1.2852977514266968,
      "learning_rate": 4.163850091651717e-07,
      "loss": 0.0345,
      "step": 547
    },
    {
      "epoch": 1.7623642943305187,
      "grad_norm": 0.9868035912513733,
      "learning_rate": 4.05210941898847e-07,
      "loss": 0.0203,
      "step": 548
    },
    {
      "epoch": 1.7655810213108163,
      "grad_norm": 0.749136745929718,
      "learning_rate": 3.941825196991378e-07,
      "loss": 0.0249,
      "step": 549
    },
    {
      "epoch": 1.7687977482911137,
      "grad_norm": 0.9028546810150146,
      "learning_rate": 3.8330009214363197e-07,
      "loss": 0.0218,
      "step": 550
    },
    {
      "epoch": 1.7720144752714113,
      "grad_norm": 0.8566029667854309,
      "learning_rate": 3.725640041822026e-07,
      "loss": 0.0251,
      "step": 551
    },
    {
      "epoch": 1.775231202251709,
      "grad_norm": 0.6764525771141052,
      "learning_rate": 3.619745961260623e-07,
      "loss": 0.0255,
      "step": 552
    },
    {
      "epoch": 1.7784479292320063,
      "grad_norm": 0.8676577210426331,
      "learning_rate": 3.5153220363698225e-07,
      "loss": 0.0233,
      "step": 553
    },
    {
      "epoch": 1.781664656212304,
      "grad_norm": 0.7067648768424988,
      "learning_rate": 3.4123715771665786e-07,
      "loss": 0.0211,
      "step": 554
    },
    {
      "epoch": 1.7848813831926016,
      "grad_norm": 0.6299604773521423,
      "learning_rate": 3.310897846962041e-07,
      "loss": 0.0158,
      "step": 555
    },
    {
      "epoch": 1.788098110172899,
      "grad_norm": 0.5761677026748657,
      "learning_rate": 3.2109040622582186e-07,
      "loss": 0.0129,
      "step": 556
    },
    {
      "epoch": 1.7913148371531966,
      "grad_norm": 0.7623018622398376,
      "learning_rate": 3.112393392645985e-07,
      "loss": 0.0241,
      "step": 557
    },
    {
      "epoch": 1.7945315641334942,
      "grad_norm": 0.8402174711227417,
      "learning_rate": 3.015368960704584e-07,
      "loss": 0.026,
      "step": 558
    },
    {
      "epoch": 1.7977482911137916,
      "grad_norm": 0.7599536776542664,
      "learning_rate": 2.919833841902714e-07,
      "loss": 0.0229,
      "step": 559
    },
    {
      "epoch": 1.8009650180940893,
      "grad_norm": 0.5555005073547363,
      "learning_rate": 2.8257910645009935e-07,
      "loss": 0.025,
      "step": 560
    },
    {
      "epoch": 1.8041817450743869,
      "grad_norm": 0.9722334742546082,
      "learning_rate": 2.733243609455971e-07,
      "loss": 0.0248,
      "step": 561
    },
    {
      "epoch": 1.8073984720546843,
      "grad_norm": 0.47031423449516296,
      "learning_rate": 2.6421944103256657e-07,
      "loss": 0.0118,
      "step": 562
    },
    {
      "epoch": 1.810615199034982,
      "grad_norm": 0.6891758441925049,
      "learning_rate": 2.5526463531765467e-07,
      "loss": 0.0272,
      "step": 563
    },
    {
      "epoch": 1.8138319260152795,
      "grad_norm": 0.5683274269104004,
      "learning_rate": 2.4646022764920843e-07,
      "loss": 0.0181,
      "step": 564
    },
    {
      "epoch": 1.817048652995577,
      "grad_norm": 0.5917408466339111,
      "learning_rate": 2.3780649710827552e-07,
      "loss": 0.0223,
      "step": 565
    },
    {
      "epoch": 1.8202653799758746,
      "grad_norm": 0.6025788187980652,
      "learning_rate": 2.2930371799975593e-07,
      "loss": 0.0209,
      "step": 566
    },
    {
      "epoch": 1.8234821069561722,
      "grad_norm": 0.6595741510391235,
      "learning_rate": 2.20952159843712e-07,
      "loss": 0.0197,
      "step": 567
    },
    {
      "epoch": 1.8266988339364696,
      "grad_norm": 0.5931892395019531,
      "learning_rate": 2.1275208736682262e-07,
      "loss": 0.0149,
      "step": 568
    },
    {
      "epoch": 1.8299155609167672,
      "grad_norm": 0.45539724826812744,
      "learning_rate": 2.0470376049398944e-07,
      "loss": 0.0132,
      "step": 569
    },
    {
      "epoch": 1.8331322878970648,
      "grad_norm": 0.7427828311920166,
      "learning_rate": 1.9680743434010385e-07,
      "loss": 0.0201,
      "step": 570
    },
    {
      "epoch": 1.8363490148773622,
      "grad_norm": 0.7957307696342468,
      "learning_rate": 1.8906335920195418e-07,
      "loss": 0.0172,
      "step": 571
    },
    {
      "epoch": 1.8395657418576599,
      "grad_norm": 0.6128239035606384,
      "learning_rate": 1.814717805502958e-07,
      "loss": 0.023,
      "step": 572
    },
    {
      "epoch": 1.8427824688379575,
      "grad_norm": 0.6324180364608765,
      "learning_rate": 1.7403293902206851e-07,
      "loss": 0.0231,
      "step": 573
    },
    {
      "epoch": 1.845999195818255,
      "grad_norm": 0.781023383140564,
      "learning_rate": 1.667470704127694e-07,
      "loss": 0.0323,
      "step": 574
    },
    {
      "epoch": 1.8492159227985525,
      "grad_norm": 0.5555590391159058,
      "learning_rate": 1.5961440566897913e-07,
      "loss": 0.0237,
      "step": 575
    },
    {
      "epoch": 1.8524326497788501,
      "grad_norm": 0.8592231869697571,
      "learning_rate": 1.5263517088103862e-07,
      "loss": 0.0232,
      "step": 576
    },
    {
      "epoch": 1.8556493767591475,
      "grad_norm": 0.6279895901679993,
      "learning_rate": 1.4580958727588746e-07,
      "loss": 0.0154,
      "step": 577
    },
    {
      "epoch": 1.8588661037394452,
      "grad_norm": 0.7780999541282654,
      "learning_rate": 1.3913787121004717e-07,
      "loss": 0.0255,
      "step": 578
    },
    {
      "epoch": 1.8620828307197428,
      "grad_norm": 0.9826112985610962,
      "learning_rate": 1.3262023416276414e-07,
      "loss": 0.0293,
      "step": 579
    },
    {
      "epoch": 1.8652995577000402,
      "grad_norm": 0.7337117791175842,
      "learning_rate": 1.2625688272930925e-07,
      "loss": 0.0244,
      "step": 580
    },
    {
      "epoch": 1.8685162846803376,
      "grad_norm": 0.7108997106552124,
      "learning_rate": 1.2004801861442373e-07,
      "loss": 0.0227,
      "step": 581
    },
    {
      "epoch": 1.8717330116606354,
      "grad_norm": 0.6001144647598267,
      "learning_rate": 1.1399383862592928e-07,
      "loss": 0.0242,
      "step": 582
    },
    {
      "epoch": 1.8749497386409328,
      "grad_norm": 1.1188668012619019,
      "learning_rate": 1.0809453466849029e-07,
      "loss": 0.0277,
      "step": 583
    },
    {
      "epoch": 1.8781664656212302,
      "grad_norm": 0.7146681547164917,
      "learning_rate": 1.0235029373752758e-07,
      "loss": 0.0218,
      "step": 584
    },
    {
      "epoch": 1.881383192601528,
      "grad_norm": 0.6120503544807434,
      "learning_rate": 9.676129791329481e-08,
      "loss": 0.0232,
      "step": 585
    },
    {
      "epoch": 1.8845999195818255,
      "grad_norm": 0.69977205991745,
      "learning_rate": 9.132772435510362e-08,
      "loss": 0.0255,
      "step": 586
    },
    {
      "epoch": 1.887816646562123,
      "grad_norm": 0.9345167279243469,
      "learning_rate": 8.604974529571042e-08,
      "loss": 0.0368,
      "step": 587
    },
    {
      "epoch": 1.8910333735424207,
      "grad_norm": 0.7977281212806702,
      "learning_rate": 8.092752803585513e-08,
      "loss": 0.0234,
      "step": 588
    },
    {
      "epoch": 1.8942501005227181,
      "grad_norm": 0.7188998460769653,
      "learning_rate": 7.59612349389599e-08,
      "loss": 0.0222,
      "step": 589
    },
    {
      "epoch": 1.8974668275030155,
      "grad_norm": 1.3486685752868652,
      "learning_rate": 7.115102342598101e-08,
      "loss": 0.0347,
      "step": 590
    },
    {
      "epoch": 1.9006835544833134,
      "grad_norm": 0.7500903010368347,
      "learning_rate": 6.649704597042061e-08,
      "loss": 0.022,
      "step": 591
    },
    {
      "epoch": 1.9039002814636108,
      "grad_norm": 0.5467425584793091,
      "learning_rate": 6.199945009349173e-08,
      "loss": 0.0195,
      "step": 592
    },
    {
      "epoch": 1.9071170084439082,
      "grad_norm": 0.7178412079811096,
      "learning_rate": 5.7658378359443104e-08,
      "loss": 0.0307,
      "step": 593
    },
    {
      "epoch": 1.910333735424206,
      "grad_norm": 0.8701561093330383,
      "learning_rate": 5.3473968371040575e-08,
      "loss": 0.0206,
      "step": 594
    },
    {
      "epoch": 1.9135504624045034,
      "grad_norm": 0.7027232050895691,
      "learning_rate": 4.944635276520393e-08,
      "loss": 0.0231,
      "step": 595
    },
    {
      "epoch": 1.9167671893848008,
      "grad_norm": 0.711676299571991,
      "learning_rate": 4.55756592088058e-08,
      "loss": 0.0219,
      "step": 596
    },
    {
      "epoch": 1.9199839163650985,
      "grad_norm": 0.8163571357727051,
      "learning_rate": 4.186201039462046e-08,
      "loss": 0.0243,
      "step": 597
    },
    {
      "epoch": 1.923200643345396,
      "grad_norm": 0.781928539276123,
      "learning_rate": 3.8305524037438035e-08,
      "loss": 0.0269,
      "step": 598
    },
    {
      "epoch": 1.9264173703256935,
      "grad_norm": 0.7710465788841248,
      "learning_rate": 3.4906312870331973e-08,
      "loss": 0.0172,
      "step": 599
    },
    {
      "epoch": 1.9296340973059911,
      "grad_norm": 0.9015802145004272,
      "learning_rate": 3.166448464108629e-08,
      "loss": 0.0302,
      "step": 600
    },
    {
      "epoch": 1.9328508242862887,
      "grad_norm": 0.857291042804718,
      "learning_rate": 2.8580142108778354e-08,
      "loss": 0.0182,
      "step": 601
    },
    {
      "epoch": 1.9360675512665861,
      "grad_norm": 0.8512669801712036,
      "learning_rate": 2.5653383040524228e-08,
      "loss": 0.0299,
      "step": 602
    },
    {
      "epoch": 1.9392842782468838,
      "grad_norm": 0.8255121111869812,
      "learning_rate": 2.2884300208378395e-08,
      "loss": 0.0278,
      "step": 603
    },
    {
      "epoch": 1.9425010052271814,
      "grad_norm": 0.5912429690361023,
      "learning_rate": 2.0272981386393332e-08,
      "loss": 0.0207,
      "step": 604
    },
    {
      "epoch": 1.9457177322074788,
      "grad_norm": 0.6523793935775757,
      "learning_rate": 1.781950934783505e-08,
      "loss": 0.0177,
      "step": 605
    },
    {
      "epoch": 1.9489344591877764,
      "grad_norm": 0.7320718169212341,
      "learning_rate": 1.552396186256411e-08,
      "loss": 0.0255,
      "step": 606
    },
    {
      "epoch": 1.952151186168074,
      "grad_norm": 0.6994187235832214,
      "learning_rate": 1.3386411694565894e-08,
      "loss": 0.0278,
      "step": 607
    },
    {
      "epoch": 1.9553679131483714,
      "grad_norm": 0.5923306941986084,
      "learning_rate": 1.1406926599646373e-08,
      "loss": 0.0192,
      "step": 608
    },
    {
      "epoch": 1.958584640128669,
      "grad_norm": 0.7130267024040222,
      "learning_rate": 9.585569323284915e-09,
      "loss": 0.0265,
      "step": 609
    },
    {
      "epoch": 1.9618013671089667,
      "grad_norm": 0.9426466226577759,
      "learning_rate": 7.922397598642551e-09,
      "loss": 0.0357,
      "step": 610
    },
    {
      "epoch": 1.965018094089264,
      "grad_norm": 0.6841176748275757,
      "learning_rate": 6.417464144736208e-09,
      "loss": 0.0232,
      "step": 611
    },
    {
      "epoch": 1.9682348210695617,
      "grad_norm": 0.6261034607887268,
      "learning_rate": 5.0708166647628345e-09,
      "loss": 0.0205,
      "step": 612
    },
    {
      "epoch": 1.9714515480498593,
      "grad_norm": 0.7374976873397827,
      "learning_rate": 3.88249784459227e-09,
      "loss": 0.0193,
      "step": 613
    },
    {
      "epoch": 1.9746682750301567,
      "grad_norm": 0.676289975643158,
      "learning_rate": 2.8525453514099966e-09,
      "loss": 0.0177,
      "step": 614
    },
    {
      "epoch": 1.9778850020104544,
      "grad_norm": 0.8624617457389832,
      "learning_rate": 1.980991832524759e-09,
      "loss": 0.0385,
      "step": 615
    },
    {
      "epoch": 1.981101728990752,
      "grad_norm": 0.6072701811790466,
      "learning_rate": 1.2678649143349485e-09,
      "loss": 0.0219,
      "step": 616
    },
    {
      "epoch": 1.9843184559710494,
      "grad_norm": 0.620798647403717,
      "learning_rate": 7.131872014509711e-10,
      "loss": 0.0211,
      "step": 617
    },
    {
      "epoch": 1.987535182951347,
      "grad_norm": 0.5879101753234863,
      "learning_rate": 3.1697627597970794e-10,
      "loss": 0.0169,
      "step": 618
    },
    {
      "epoch": 1.9907519099316446,
      "grad_norm": 1.4460906982421875,
      "learning_rate": 7.924469696718451e-11,
      "loss": 0.0259,
      "step": 619
    },
    {
      "epoch": 1.993968636911942,
      "grad_norm": 0.9200376868247986,
      "learning_rate": 0.0,
      "loss": 0.0252,
      "step": 620
    },
    {
      "epoch": 1.993968636911942,
      "step": 620,
      "total_flos": 7.073469349725471e+17,
      "train_loss": 0.04489339354387935,
      "train_runtime": 2537.7411,
      "train_samples_per_second": 125.415,
      "train_steps_per_second": 0.244
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 620,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 7.073469349725471e+17,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}