{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.9952,
  "eval_steps": 500,
  "global_step": 312,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0096,
      "grad_norm": 6.292316913604736,
      "learning_rate": 3.125e-07,
      "loss": 1.0104,
      "step": 1
    },
    {
      "epoch": 0.0192,
      "grad_norm": 6.1730875968933105,
      "learning_rate": 6.25e-07,
      "loss": 1.0026,
      "step": 2
    },
    {
      "epoch": 0.0288,
      "grad_norm": 6.434154033660889,
      "learning_rate": 9.375000000000001e-07,
      "loss": 1.0329,
      "step": 3
    },
    {
      "epoch": 0.0384,
      "grad_norm": 6.069917678833008,
      "learning_rate": 1.25e-06,
      "loss": 0.9934,
      "step": 4
    },
    {
      "epoch": 0.048,
      "grad_norm": 5.826951503753662,
      "learning_rate": 1.5625e-06,
      "loss": 0.9941,
      "step": 5
    },
    {
      "epoch": 0.0576,
      "grad_norm": 5.749411582946777,
      "learning_rate": 1.8750000000000003e-06,
      "loss": 0.9974,
      "step": 6
    },
    {
      "epoch": 0.0672,
      "grad_norm": 4.575586318969727,
      "learning_rate": 2.1875000000000002e-06,
      "loss": 0.9291,
      "step": 7
    },
    {
      "epoch": 0.0768,
      "grad_norm": 4.361931800842285,
      "learning_rate": 2.5e-06,
      "loss": 0.9198,
      "step": 8
    },
    {
      "epoch": 0.0864,
      "grad_norm": 2.572038412094116,
      "learning_rate": 2.8125e-06,
      "loss": 0.9106,
      "step": 9
    },
    {
      "epoch": 0.096,
      "grad_norm": 2.4609551429748535,
      "learning_rate": 3.125e-06,
      "loss": 0.9226,
      "step": 10
    },
    {
      "epoch": 0.1056,
      "grad_norm": 2.093930959701538,
      "learning_rate": 3.4375e-06,
      "loss": 0.8899,
      "step": 11
    },
    {
      "epoch": 0.1152,
      "grad_norm": 3.553408622741699,
      "learning_rate": 3.7500000000000005e-06,
      "loss": 0.9156,
      "step": 12
    },
    {
      "epoch": 0.1248,
      "grad_norm": 3.967522621154785,
      "learning_rate": 4.0625000000000005e-06,
      "loss": 0.8831,
      "step": 13
    },
    {
      "epoch": 0.1344,
      "grad_norm": 3.7353246212005615,
      "learning_rate": 4.3750000000000005e-06,
      "loss": 0.8756,
      "step": 14
    },
    {
      "epoch": 0.144,
      "grad_norm": 3.722759962081909,
      "learning_rate": 4.6875000000000004e-06,
      "loss": 0.8935,
      "step": 15
    },
    {
      "epoch": 0.1536,
      "grad_norm": 2.4490113258361816,
      "learning_rate": 5e-06,
      "loss": 0.8094,
      "step": 16
    },
    {
      "epoch": 0.1632,
      "grad_norm": 2.2576043605804443,
      "learning_rate": 5.3125e-06,
      "loss": 0.8232,
      "step": 17
    },
    {
      "epoch": 0.1728,
      "grad_norm": 2.022249460220337,
      "learning_rate": 5.625e-06,
      "loss": 0.8152,
      "step": 18
    },
    {
      "epoch": 0.1824,
      "grad_norm": 1.6432116031646729,
      "learning_rate": 5.9375e-06,
      "loss": 0.8131,
      "step": 19
    },
    {
      "epoch": 0.192,
      "grad_norm": 1.428215742111206,
      "learning_rate": 6.25e-06,
      "loss": 0.7559,
      "step": 20
    },
    {
      "epoch": 0.2016,
      "grad_norm": 1.2886855602264404,
      "learning_rate": 6.5625e-06,
      "loss": 0.7587,
      "step": 21
    },
    {
      "epoch": 0.2112,
      "grad_norm": 1.229178786277771,
      "learning_rate": 6.875e-06,
      "loss": 0.779,
      "step": 22
    },
    {
      "epoch": 0.2208,
      "grad_norm": 1.1774561405181885,
      "learning_rate": 7.1875e-06,
      "loss": 0.7286,
      "step": 23
    },
    {
      "epoch": 0.2304,
      "grad_norm": 1.279067039489746,
      "learning_rate": 7.500000000000001e-06,
      "loss": 0.768,
      "step": 24
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.9670516848564148,
      "learning_rate": 7.8125e-06,
      "loss": 0.7009,
      "step": 25
    },
    {
      "epoch": 0.2496,
      "grad_norm": 1.081884503364563,
      "learning_rate": 8.125000000000001e-06,
      "loss": 0.7441,
      "step": 26
    },
    {
      "epoch": 0.2592,
      "grad_norm": 1.046007752418518,
      "learning_rate": 8.4375e-06,
      "loss": 0.7463,
      "step": 27
    },
    {
      "epoch": 0.2688,
      "grad_norm": 0.8546833395957947,
      "learning_rate": 8.750000000000001e-06,
      "loss": 0.6999,
      "step": 28
    },
    {
      "epoch": 0.2784,
      "grad_norm": 0.8875015377998352,
      "learning_rate": 9.0625e-06,
      "loss": 0.7203,
      "step": 29
    },
    {
      "epoch": 0.288,
      "grad_norm": 0.899333119392395,
      "learning_rate": 9.375000000000001e-06,
      "loss": 0.725,
      "step": 30
    },
    {
      "epoch": 0.2976,
      "grad_norm": 0.9151673316955566,
      "learning_rate": 9.6875e-06,
      "loss": 0.7323,
      "step": 31
    },
    {
      "epoch": 0.3072,
      "grad_norm": 0.7202209830284119,
      "learning_rate": 1e-05,
      "loss": 0.7081,
      "step": 32
    },
    {
      "epoch": 0.3168,
      "grad_norm": 0.8347250819206238,
      "learning_rate": 9.999685283773504e-06,
      "loss": 0.702,
      "step": 33
    },
    {
      "epoch": 0.3264,
      "grad_norm": 0.7798133492469788,
      "learning_rate": 9.998741174712534e-06,
      "loss": 0.679,
      "step": 34
    },
    {
      "epoch": 0.336,
      "grad_norm": 0.8034778833389282,
      "learning_rate": 9.997167791667668e-06,
      "loss": 0.7096,
      "step": 35
    },
    {
      "epoch": 0.3456,
      "grad_norm": 0.7123808860778809,
      "learning_rate": 9.994965332706574e-06,
      "loss": 0.6878,
      "step": 36
    },
    {
      "epoch": 0.3552,
      "grad_norm": 0.8426344990730286,
      "learning_rate": 9.992134075089085e-06,
      "loss": 0.7157,
      "step": 37
    },
    {
      "epoch": 0.3648,
      "grad_norm": 0.6954659819602966,
      "learning_rate": 9.98867437523228e-06,
      "loss": 0.709,
      "step": 38
    },
    {
      "epoch": 0.3744,
      "grad_norm": 0.8144009709358215,
      "learning_rate": 9.984586668665641e-06,
      "loss": 0.6934,
      "step": 39
    },
    {
      "epoch": 0.384,
      "grad_norm": 0.6534774303436279,
      "learning_rate": 9.979871469976197e-06,
      "loss": 0.6694,
      "step": 40
    },
    {
      "epoch": 0.3936,
      "grad_norm": 0.792493999004364,
      "learning_rate": 9.974529372743762e-06,
      "loss": 0.6651,
      "step": 41
    },
    {
      "epoch": 0.4032,
      "grad_norm": 0.7658901810646057,
      "learning_rate": 9.968561049466214e-06,
      "loss": 0.6625,
      "step": 42
    },
    {
      "epoch": 0.4128,
      "grad_norm": 0.5957349538803101,
      "learning_rate": 9.961967251474823e-06,
      "loss": 0.6556,
      "step": 43
    },
    {
      "epoch": 0.4224,
      "grad_norm": 0.6661067605018616,
      "learning_rate": 9.954748808839675e-06,
      "loss": 0.6638,
      "step": 44
    },
    {
      "epoch": 0.432,
      "grad_norm": 0.6937732100486755,
      "learning_rate": 9.946906630265184e-06,
      "loss": 0.6636,
      "step": 45
    },
    {
      "epoch": 0.4416,
      "grad_norm": 0.6557593941688538,
      "learning_rate": 9.938441702975689e-06,
      "loss": 0.6641,
      "step": 46
    },
    {
      "epoch": 0.4512,
      "grad_norm": 0.7512499094009399,
      "learning_rate": 9.92935509259118e-06,
      "loss": 0.6812,
      "step": 47
    },
    {
      "epoch": 0.4608,
      "grad_norm": 0.626158595085144,
      "learning_rate": 9.91964794299315e-06,
      "loss": 0.6607,
      "step": 48
    },
    {
      "epoch": 0.4704,
      "grad_norm": 0.5737528204917908,
      "learning_rate": 9.909321476180594e-06,
      "loss": 0.695,
      "step": 49
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.6223921775817871,
      "learning_rate": 9.898376992116179e-06,
      "loss": 0.6679,
      "step": 50
    },
    {
      "epoch": 0.4896,
      "grad_norm": 0.6310274600982666,
      "learning_rate": 9.886815868562596e-06,
      "loss": 0.6666,
      "step": 51
    },
    {
      "epoch": 0.4992,
      "grad_norm": 0.7005355954170227,
      "learning_rate": 9.874639560909118e-06,
      "loss": 0.6745,
      "step": 52
    },
    {
      "epoch": 0.5088,
      "grad_norm": 0.6608271598815918,
      "learning_rate": 9.861849601988384e-06,
      "loss": 0.6894,
      "step": 53
    },
    {
      "epoch": 0.5184,
      "grad_norm": 0.6414551734924316,
      "learning_rate": 9.848447601883436e-06,
      "loss": 0.6771,
      "step": 54
    },
    {
      "epoch": 0.528,
      "grad_norm": 0.7467500567436218,
      "learning_rate": 9.834435247725032e-06,
      "loss": 0.6468,
      "step": 55
    },
    {
      "epoch": 0.5376,
      "grad_norm": 0.5815238952636719,
      "learning_rate": 9.819814303479268e-06,
      "loss": 0.6622,
      "step": 56
    },
    {
      "epoch": 0.5472,
      "grad_norm": 0.6850916147232056,
      "learning_rate": 9.804586609725499e-06,
      "loss": 0.623,
      "step": 57
    },
    {
      "epoch": 0.5568,
      "grad_norm": 0.67253178358078,
      "learning_rate": 9.788754083424654e-06,
      "loss": 0.6524,
      "step": 58
    },
    {
      "epoch": 0.5664,
      "grad_norm": 0.6569348573684692,
      "learning_rate": 9.772318717677905e-06,
      "loss": 0.6649,
      "step": 59
    },
    {
      "epoch": 0.576,
      "grad_norm": 0.7703596949577332,
      "learning_rate": 9.755282581475769e-06,
      "loss": 0.6653,
      "step": 60
    },
    {
      "epoch": 0.5856,
      "grad_norm": 0.5406990647315979,
      "learning_rate": 9.737647819437645e-06,
      "loss": 0.664,
      "step": 61
    },
    {
      "epoch": 0.5952,
      "grad_norm": 0.6827347278594971,
      "learning_rate": 9.719416651541839e-06,
      "loss": 0.6854,
      "step": 62
    },
    {
      "epoch": 0.6048,
      "grad_norm": 0.6011564135551453,
      "learning_rate": 9.700591372846096e-06,
      "loss": 0.6595,
      "step": 63
    },
    {
      "epoch": 0.6144,
      "grad_norm": 0.6541777849197388,
      "learning_rate": 9.681174353198687e-06,
      "loss": 0.6764,
      "step": 64
    },
    {
      "epoch": 0.624,
      "grad_norm": 0.6818787455558777,
      "learning_rate": 9.661168036940071e-06,
      "loss": 0.6201,
      "step": 65
    },
    {
      "epoch": 0.6336,
      "grad_norm": 0.6184643507003784,
      "learning_rate": 9.640574942595195e-06,
      "loss": 0.6488,
      "step": 66
    },
    {
      "epoch": 0.6432,
      "grad_norm": 0.6250948309898376,
      "learning_rate": 9.619397662556434e-06,
      "loss": 0.6462,
      "step": 67
    },
    {
      "epoch": 0.6528,
      "grad_norm": 0.6833012104034424,
      "learning_rate": 9.597638862757255e-06,
      "loss": 0.6527,
      "step": 68
    },
    {
      "epoch": 0.6624,
      "grad_norm": 0.5511516332626343,
      "learning_rate": 9.5753012823366e-06,
      "loss": 0.6502,
      "step": 69
    },
    {
      "epoch": 0.672,
      "grad_norm": 0.7079241275787354,
      "learning_rate": 9.552387733294081e-06,
      "loss": 0.6747,
      "step": 70
    },
    {
      "epoch": 0.6816,
      "grad_norm": 0.637936532497406,
      "learning_rate": 9.528901100135971e-06,
      "loss": 0.669,
      "step": 71
    },
    {
      "epoch": 0.6912,
      "grad_norm": 0.6084673404693604,
      "learning_rate": 9.504844339512096e-06,
      "loss": 0.6397,
      "step": 72
    },
    {
      "epoch": 0.7008,
      "grad_norm": 0.6382216811180115,
      "learning_rate": 9.480220479843627e-06,
      "loss": 0.6775,
      "step": 73
    },
    {
      "epoch": 0.7104,
      "grad_norm": 0.5169996023178101,
      "learning_rate": 9.45503262094184e-06,
      "loss": 0.6424,
      "step": 74
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.6096466779708862,
      "learning_rate": 9.4292839336179e-06,
      "loss": 0.642,
      "step": 75
    },
    {
      "epoch": 0.7296,
      "grad_norm": 0.5510257482528687,
      "learning_rate": 9.40297765928369e-06,
      "loss": 0.6335,
      "step": 76
    },
    {
      "epoch": 0.7392,
      "grad_norm": 0.5844236612319946,
      "learning_rate": 9.376117109543769e-06,
      "loss": 0.6247,
      "step": 77
    },
    {
      "epoch": 0.7488,
      "grad_norm": 0.5517457127571106,
      "learning_rate": 9.348705665778479e-06,
      "loss": 0.6216,
      "step": 78
    },
    {
      "epoch": 0.7584,
      "grad_norm": 0.5716204047203064,
      "learning_rate": 9.320746778718274e-06,
      "loss": 0.6496,
      "step": 79
    },
    {
      "epoch": 0.768,
      "grad_norm": 0.5770816206932068,
      "learning_rate": 9.292243968009332e-06,
      "loss": 0.6431,
      "step": 80
    },
    {
      "epoch": 0.7776,
      "grad_norm": 0.5499193668365479,
      "learning_rate": 9.263200821770462e-06,
      "loss": 0.6548,
      "step": 81
    },
    {
      "epoch": 0.7872,
      "grad_norm": 0.6924810409545898,
      "learning_rate": 9.233620996141421e-06,
      "loss": 0.6227,
      "step": 82
    },
    {
      "epoch": 0.7968,
      "grad_norm": 0.5241541862487793,
      "learning_rate": 9.203508214822652e-06,
      "loss": 0.6568,
      "step": 83
    },
    {
      "epoch": 0.8064,
      "grad_norm": 0.6041975617408752,
      "learning_rate": 9.172866268606514e-06,
      "loss": 0.6451,
      "step": 84
    },
    {
      "epoch": 0.816,
      "grad_norm": 0.6413403749465942,
      "learning_rate": 9.141699014900084e-06,
      "loss": 0.6489,
      "step": 85
    },
    {
      "epoch": 0.8256,
      "grad_norm": 0.5075194835662842,
      "learning_rate": 9.110010377239552e-06,
      "loss": 0.6278,
      "step": 86
    },
    {
      "epoch": 0.8352,
      "grad_norm": 0.6670998334884644,
      "learning_rate": 9.077804344796302e-06,
      "loss": 0.6798,
      "step": 87
    },
    {
      "epoch": 0.8448,
      "grad_norm": 0.5765168070793152,
      "learning_rate": 9.045084971874738e-06,
      "loss": 0.6451,
      "step": 88
    },
    {
      "epoch": 0.8544,
      "grad_norm": 0.555912435054779,
      "learning_rate": 9.011856377401891e-06,
      "loss": 0.6401,
      "step": 89
    },
    {
      "epoch": 0.864,
      "grad_norm": 0.5889183878898621,
      "learning_rate": 8.978122744408905e-06,
      "loss": 0.6513,
      "step": 90
    },
    {
      "epoch": 0.8736,
      "grad_norm": 0.558150053024292,
      "learning_rate": 8.943888319504456e-06,
      "loss": 0.6286,
      "step": 91
    },
    {
      "epoch": 0.8832,
      "grad_norm": 0.6536504626274109,
      "learning_rate": 8.90915741234015e-06,
      "loss": 0.6396,
      "step": 92
    },
    {
      "epoch": 0.8928,
      "grad_norm": 0.6340594291687012,
      "learning_rate": 8.873934395068006e-06,
      "loss": 0.6436,
      "step": 93
    },
    {
      "epoch": 0.9024,
      "grad_norm": 0.7267080545425415,
      "learning_rate": 8.838223701790057e-06,
      "loss": 0.6352,
      "step": 94
    },
    {
      "epoch": 0.912,
      "grad_norm": 0.5870280265808105,
      "learning_rate": 8.802029828000157e-06,
      "loss": 0.6036,
      "step": 95
    },
    {
      "epoch": 0.9216,
      "grad_norm": 0.6412261128425598,
      "learning_rate": 8.765357330018056e-06,
      "loss": 0.6242,
      "step": 96
    },
    {
      "epoch": 0.9312,
      "grad_norm": 0.6203837394714355,
      "learning_rate": 8.728210824415829e-06,
      "loss": 0.6363,
      "step": 97
    },
    {
      "epoch": 0.9408,
      "grad_norm": 0.6868276000022888,
      "learning_rate": 8.690594987436705e-06,
      "loss": 0.6171,
      "step": 98
    },
    {
      "epoch": 0.9504,
      "grad_norm": 0.5638834238052368,
      "learning_rate": 8.652514554406388e-06,
      "loss": 0.635,
      "step": 99
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.69553542137146,
      "learning_rate": 8.613974319136959e-06,
      "loss": 0.6527,
      "step": 100
    },
    {
      "epoch": 0.9696,
      "grad_norm": 0.5303117036819458,
      "learning_rate": 8.574979133323378e-06,
      "loss": 0.6403,
      "step": 101
    },
    {
      "epoch": 0.9792,
      "grad_norm": 0.6957817673683167,
      "learning_rate": 8.535533905932739e-06,
      "loss": 0.6295,
      "step": 102
    },
    {
      "epoch": 0.9888,
      "grad_norm": 0.6576924324035645,
      "learning_rate": 8.495643602586287e-06,
      "loss": 0.6362,
      "step": 103
    },
    {
      "epoch": 0.9984,
      "grad_norm": 0.6190880537033081,
      "learning_rate": 8.455313244934324e-06,
      "loss": 0.6343,
      "step": 104
    },
    {
      "epoch": 1.008,
      "grad_norm": 1.2567451000213623,
      "learning_rate": 8.414547910024035e-06,
      "loss": 1.0639,
      "step": 105
    },
    {
      "epoch": 1.0176,
      "grad_norm": 0.5477228164672852,
      "learning_rate": 8.373352729660373e-06,
      "loss": 0.546,
      "step": 106
    },
    {
      "epoch": 1.0272,
      "grad_norm": 0.569487452507019,
      "learning_rate": 8.331732889760021e-06,
      "loss": 0.5735,
      "step": 107
    },
    {
      "epoch": 1.0368,
      "grad_norm": 0.6835495233535767,
      "learning_rate": 8.289693629698564e-06,
      "loss": 0.6376,
      "step": 108
    },
    {
      "epoch": 1.0464,
      "grad_norm": 0.5983431339263916,
      "learning_rate": 8.247240241650918e-06,
      "loss": 0.6078,
      "step": 109
    },
    {
      "epoch": 1.056,
      "grad_norm": 0.6319621801376343,
      "learning_rate": 8.204378069925121e-06,
      "loss": 0.591,
      "step": 110
    },
    {
      "epoch": 1.0656,
      "grad_norm": 0.6427493691444397,
      "learning_rate": 8.16111251028955e-06,
      "loss": 0.5695,
      "step": 111
    },
    {
      "epoch": 1.0752,
      "grad_norm": 0.5300183892250061,
      "learning_rate": 8.117449009293668e-06,
      "loss": 0.5742,
      "step": 112
    },
    {
      "epoch": 1.0848,
      "grad_norm": 0.5667701959609985,
      "learning_rate": 8.073393063582386e-06,
      "loss": 0.5429,
      "step": 113
    },
    {
      "epoch": 1.0944,
      "grad_norm": 0.6527385711669922,
      "learning_rate": 8.0289502192041e-06,
      "loss": 0.5633,
      "step": 114
    },
    {
      "epoch": 1.104,
      "grad_norm": 0.6647496819496155,
      "learning_rate": 7.984126070912519e-06,
      "loss": 0.6611,
      "step": 115
    },
    {
      "epoch": 1.1136,
      "grad_norm": 0.47675877809524536,
      "learning_rate": 7.938926261462366e-06,
      "loss": 0.5287,
      "step": 116
    },
    {
      "epoch": 1.1232,
      "grad_norm": 0.738347589969635,
      "learning_rate": 7.89335648089903e-06,
      "loss": 0.6157,
      "step": 117
    },
    {
      "epoch": 1.1328,
      "grad_norm": 0.57480788230896,
      "learning_rate": 7.84742246584226e-06,
      "loss": 0.5565,
      "step": 118
    },
    {
      "epoch": 1.1424,
      "grad_norm": 0.5129834413528442,
      "learning_rate": 7.801129998764014e-06,
      "loss": 0.5546,
      "step": 119
    },
    {
      "epoch": 1.152,
      "grad_norm": 0.7716547250747681,
      "learning_rate": 7.754484907260513e-06,
      "loss": 0.6147,
      "step": 120
    },
    {
      "epoch": 1.1616,
      "grad_norm": 0.545978307723999,
      "learning_rate": 7.70749306331863e-06,
      "loss": 0.5565,
      "step": 121
    },
    {
      "epoch": 1.1712,
      "grad_norm": 0.5517440438270569,
      "learning_rate": 7.660160382576683e-06,
      "loss": 0.608,
      "step": 122
    },
    {
      "epoch": 1.1808,
      "grad_norm": 0.6543371081352234,
      "learning_rate": 7.612492823579744e-06,
      "loss": 0.5766,
      "step": 123
    },
    {
      "epoch": 1.1904,
      "grad_norm": 0.5610912442207336,
      "learning_rate": 7.564496387029532e-06,
      "loss": 0.5889,
      "step": 124
    },
    {
      "epoch": 1.2,
      "grad_norm": 0.725592315196991,
      "learning_rate": 7.516177115029002e-06,
      "loss": 0.576,
      "step": 125
    },
    {
      "epoch": 1.2096,
      "grad_norm": 0.5460718870162964,
      "learning_rate": 7.467541090321735e-06,
      "loss": 0.5926,
      "step": 126
    },
    {
      "epoch": 1.2192,
      "grad_norm": 0.587978720664978,
      "learning_rate": 7.4185944355261996e-06,
      "loss": 0.544,
      "step": 127
    },
    {
      "epoch": 1.2288000000000001,
      "grad_norm": 0.5988969802856445,
      "learning_rate": 7.369343312364994e-06,
      "loss": 0.5996,
      "step": 128
    },
    {
      "epoch": 1.2384,
      "grad_norm": 0.5384567379951477,
      "learning_rate": 7.319793920889171e-06,
      "loss": 0.5328,
      "step": 129
    },
    {
      "epoch": 1.248,
      "grad_norm": 0.6369662880897522,
      "learning_rate": 7.269952498697734e-06,
      "loss": 0.5878,
      "step": 130
    },
    {
      "epoch": 1.2576,
      "grad_norm": 0.5659148097038269,
      "learning_rate": 7.219825320152411e-06,
      "loss": 0.6203,
      "step": 131
    },
    {
      "epoch": 1.2671999999999999,
      "grad_norm": 0.5150180459022522,
      "learning_rate": 7.169418695587791e-06,
      "loss": 0.4977,
      "step": 132
    },
    {
      "epoch": 1.2768,
      "grad_norm": 0.6585394740104675,
      "learning_rate": 7.118738970516944e-06,
      "loss": 0.6233,
      "step": 133
    },
    {
      "epoch": 1.2864,
      "grad_norm": 0.49442875385284424,
      "learning_rate": 7.067792524832604e-06,
      "loss": 0.5462,
      "step": 134
    },
    {
      "epoch": 1.296,
      "grad_norm": 0.5812093019485474,
      "learning_rate": 7.016585772004026e-06,
      "loss": 0.6037,
      "step": 135
    },
    {
      "epoch": 1.3056,
      "grad_norm": 0.6433523297309875,
      "learning_rate": 6.965125158269619e-06,
      "loss": 0.6334,
      "step": 136
    },
    {
      "epoch": 1.3152,
      "grad_norm": 0.5826403498649597,
      "learning_rate": 6.913417161825449e-06,
      "loss": 0.5644,
      "step": 137
    },
    {
      "epoch": 1.3248,
      "grad_norm": 0.492260605096817,
      "learning_rate": 6.8614682920097265e-06,
      "loss": 0.5566,
      "step": 138
    },
    {
      "epoch": 1.3344,
      "grad_norm": 0.5302029848098755,
      "learning_rate": 6.809285088483361e-06,
      "loss": 0.576,
      "step": 139
    },
    {
      "epoch": 1.3439999999999999,
      "grad_norm": 0.5525418519973755,
      "learning_rate": 6.7568741204067145e-06,
      "loss": 0.5355,
      "step": 140
    },
    {
      "epoch": 1.3536000000000001,
      "grad_norm": 0.5225656628608704,
      "learning_rate": 6.704241985612625e-06,
      "loss": 0.5751,
      "step": 141
    },
    {
      "epoch": 1.3632,
      "grad_norm": 0.4587990939617157,
      "learning_rate": 6.651395309775837e-06,
      "loss": 0.5381,
      "step": 142
    },
    {
      "epoch": 1.3728,
      "grad_norm": 0.579556941986084,
      "learning_rate": 6.598340745578908e-06,
      "loss": 0.6322,
      "step": 143
    },
    {
      "epoch": 1.3824,
      "grad_norm": 0.5754554271697998,
      "learning_rate": 6.545084971874738e-06,
      "loss": 0.575,
      "step": 144
    },
    {
      "epoch": 1.392,
      "grad_norm": 0.5633704662322998,
      "learning_rate": 6.491634692845781e-06,
      "loss": 0.6123,
      "step": 145
    },
    {
      "epoch": 1.4016,
      "grad_norm": 0.4322211742401123,
      "learning_rate": 6.437996637160086e-06,
      "loss": 0.5378,
      "step": 146
    },
    {
      "epoch": 1.4112,
      "grad_norm": 0.5670461058616638,
      "learning_rate": 6.384177557124247e-06,
      "loss": 0.6098,
      "step": 147
    },
    {
      "epoch": 1.4208,
      "grad_norm": 0.501027524471283,
      "learning_rate": 6.330184227833376e-06,
      "loss": 0.5874,
      "step": 148
    },
    {
      "epoch": 1.4304000000000001,
      "grad_norm": 0.44575896859169006,
      "learning_rate": 6.276023446318214e-06,
      "loss": 0.5889,
      "step": 149
    },
    {
      "epoch": 1.44,
      "grad_norm": 0.48979705572128296,
      "learning_rate": 6.2217020306894705e-06,
      "loss": 0.5575,
      "step": 150
    },
    {
      "epoch": 1.4496,
      "grad_norm": 0.5052200555801392,
      "learning_rate": 6.1672268192795285e-06,
      "loss": 0.551,
      "step": 151
    },
    {
      "epoch": 1.4592,
      "grad_norm": 0.4936927855014801,
      "learning_rate": 6.112604669781572e-06,
      "loss": 0.5616,
      "step": 152
    },
    {
      "epoch": 1.4687999999999999,
      "grad_norm": 0.4863929748535156,
      "learning_rate": 6.057842458386315e-06,
      "loss": 0.6105,
      "step": 153
    },
    {
      "epoch": 1.4784,
      "grad_norm": 0.4865013659000397,
      "learning_rate": 6.002947078916365e-06,
      "loss": 0.5389,
      "step": 154
    },
    {
      "epoch": 1.488,
      "grad_norm": 0.5264416933059692,
      "learning_rate": 5.947925441958393e-06,
      "loss": 0.5851,
      "step": 155
    },
    {
      "epoch": 1.4976,
      "grad_norm": 0.4394872784614563,
      "learning_rate": 5.892784473993184e-06,
      "loss": 0.5623,
      "step": 156
    },
    {
      "epoch": 1.5072,
      "grad_norm": 0.5211396813392639,
      "learning_rate": 5.837531116523683e-06,
      "loss": 0.6225,
      "step": 157
    },
    {
      "epoch": 1.5168,
      "grad_norm": 0.4977400302886963,
      "learning_rate": 5.782172325201155e-06,
      "loss": 0.5915,
      "step": 158
    },
    {
      "epoch": 1.5264,
      "grad_norm": 0.4418800175189972,
      "learning_rate": 5.726715068949564e-06,
      "loss": 0.4918,
      "step": 159
    },
    {
      "epoch": 1.536,
      "grad_norm": 0.5426927208900452,
      "learning_rate": 5.671166329088278e-06,
      "loss": 0.6375,
      "step": 160
    },
    {
      "epoch": 1.5455999999999999,
      "grad_norm": 0.4744737446308136,
      "learning_rate": 5.615533098453215e-06,
      "loss": 0.6084,
      "step": 161
    },
    {
      "epoch": 1.5552000000000001,
      "grad_norm": 0.4746991991996765,
      "learning_rate": 5.559822380516539e-06,
      "loss": 0.5142,
      "step": 162
    },
    {
      "epoch": 1.5648,
      "grad_norm": 0.4827795624732971,
      "learning_rate": 5.504041188505022e-06,
      "loss": 0.5769,
      "step": 163
    },
    {
      "epoch": 1.5744,
      "grad_norm": 0.468890905380249,
      "learning_rate": 5.448196544517168e-06,
      "loss": 0.5877,
      "step": 164
    },
    {
      "epoch": 1.584,
      "grad_norm": 0.48452773690223694,
      "learning_rate": 5.392295478639226e-06,
      "loss": 0.5833,
      "step": 165
    },
    {
      "epoch": 1.5936,
      "grad_norm": 0.5015302896499634,
      "learning_rate": 5.336345028060199e-06,
      "loss": 0.6116,
      "step": 166
    },
    {
      "epoch": 1.6032,
      "grad_norm": 0.43964874744415283,
      "learning_rate": 5.2803522361859596e-06,
      "loss": 0.5582,
      "step": 167
    },
    {
      "epoch": 1.6128,
      "grad_norm": 0.4486831724643707,
      "learning_rate": 5.224324151752575e-06,
      "loss": 0.5489,
      "step": 168
    },
    {
      "epoch": 1.6223999999999998,
      "grad_norm": 0.47891250252723694,
      "learning_rate": 5.168267827938971e-06,
      "loss": 0.5648,
      "step": 169
    },
    {
      "epoch": 1.6320000000000001,
      "grad_norm": 0.4622023105621338,
      "learning_rate": 5.112190321479026e-06,
      "loss": 0.5918,
      "step": 170
    },
    {
      "epoch": 1.6416,
      "grad_norm": 0.5129140615463257,
      "learning_rate": 5.05609869177323e-06,
      "loss": 0.6332,
      "step": 171
    },
    {
      "epoch": 1.6512,
      "grad_norm": 0.4973113536834717,
      "learning_rate": 5e-06,
      "loss": 0.5124,
      "step": 172
    },
    {
      "epoch": 1.6608,
      "grad_norm": 0.4771605134010315,
      "learning_rate": 4.943901308226771e-06,
      "loss": 0.6114,
      "step": 173
    },
    {
      "epoch": 1.6703999999999999,
      "grad_norm": 0.45923885703086853,
      "learning_rate": 4.887809678520976e-06,
      "loss": 0.5534,
      "step": 174
    },
    {
      "epoch": 1.6800000000000002,
      "grad_norm": 0.5138639211654663,
      "learning_rate": 4.831732172061032e-06,
      "loss": 0.5472,
      "step": 175
    },
    {
      "epoch": 1.6896,
      "grad_norm": 0.49369630217552185,
      "learning_rate": 4.775675848247427e-06,
      "loss": 0.6085,
      "step": 176
    },
    {
      "epoch": 1.6992,
      "grad_norm": 0.48870953917503357,
      "learning_rate": 4.719647763814041e-06,
      "loss": 0.5798,
      "step": 177
    },
    {
      "epoch": 1.7088,
      "grad_norm": 0.4958065152168274,
      "learning_rate": 4.663654971939802e-06,
      "loss": 0.5883,
      "step": 178
    },
    {
      "epoch": 1.7184,
      "grad_norm": 0.49560844898223877,
      "learning_rate": 4.6077045213607765e-06,
      "loss": 0.5757,
      "step": 179
    },
    {
      "epoch": 1.728,
      "grad_norm": 0.4331684112548828,
      "learning_rate": 4.551803455482833e-06,
      "loss": 0.5657,
      "step": 180
    },
    {
      "epoch": 1.7376,
      "grad_norm": 0.5511804223060608,
      "learning_rate": 4.4959588114949785e-06,
      "loss": 0.6329,
      "step": 181
    },
    {
      "epoch": 1.7471999999999999,
      "grad_norm": 0.48789289593696594,
      "learning_rate": 4.4401776194834615e-06,
      "loss": 0.5785,
      "step": 182
    },
    {
      "epoch": 1.7568000000000001,
      "grad_norm": 0.4831961691379547,
      "learning_rate": 4.384466901546786e-06,
      "loss": 0.581,
      "step": 183
    },
    {
      "epoch": 1.7664,
      "grad_norm": 0.45741593837738037,
      "learning_rate": 4.3288336709117246e-06,
      "loss": 0.5404,
      "step": 184
    },
    {
      "epoch": 1.776,
      "grad_norm": 0.4758993685245514,
      "learning_rate": 4.273284931050438e-06,
      "loss": 0.5718,
      "step": 185
    },
    {
      "epoch": 1.7856,
      "grad_norm": 0.4522536098957062,
      "learning_rate": 4.217827674798845e-06,
      "loss": 0.5726,
      "step": 186
    },
    {
      "epoch": 1.7952,
      "grad_norm": 0.5672460794448853,
      "learning_rate": 4.162468883476319e-06,
      "loss": 0.6464,
      "step": 187
    },
    {
      "epoch": 1.8048,
      "grad_norm": 0.4288261830806732,
      "learning_rate": 4.107215526006818e-06,
      "loss": 0.5412,
      "step": 188
    },
    {
      "epoch": 1.8144,
      "grad_norm": 0.4769589900970459,
      "learning_rate": 4.052074558041608e-06,
      "loss": 0.6142,
      "step": 189
    },
    {
      "epoch": 1.8239999999999998,
      "grad_norm": 0.5145401954650879,
      "learning_rate": 3.997052921083637e-06,
      "loss": 0.5725,
      "step": 190
    },
    {
      "epoch": 1.8336000000000001,
      "grad_norm": 0.46486806869506836,
      "learning_rate": 3.9421575416136866e-06,
      "loss": 0.621,
      "step": 191
    },
    {
      "epoch": 1.8432,
      "grad_norm": 0.42144760489463806,
      "learning_rate": 3.887395330218429e-06,
      "loss": 0.4962,
      "step": 192
    },
    {
      "epoch": 1.8528,
      "grad_norm": 0.5083242654800415,
      "learning_rate": 3.832773180720475e-06,
      "loss": 0.6506,
      "step": 193
    },
    {
      "epoch": 1.8624,
      "grad_norm": 0.4573202431201935,
      "learning_rate": 3.778297969310529e-06,
      "loss": 0.5586,
      "step": 194
    },
    {
      "epoch": 1.8719999999999999,
      "grad_norm": 0.4502522647380829,
      "learning_rate": 3.723976553681787e-06,
      "loss": 0.5363,
      "step": 195
    },
    {
      "epoch": 1.8816000000000002,
      "grad_norm": 0.42463284730911255,
      "learning_rate": 3.669815772166625e-06,
      "loss": 0.6078,
      "step": 196
    },
    {
      "epoch": 1.8912,
      "grad_norm": 0.4534298777580261,
      "learning_rate": 3.6158224428757538e-06,
      "loss": 0.5645,
      "step": 197
    },
    {
      "epoch": 1.9008,
      "grad_norm": 0.4633239507675171,
      "learning_rate": 3.562003362839914e-06,
      "loss": 0.5657,
      "step": 198
    },
    {
      "epoch": 1.9104,
      "grad_norm": 0.4833671450614929,
      "learning_rate": 3.50836530715422e-06,
      "loss": 0.5732,
      "step": 199
    },
    {
      "epoch": 1.92,
      "grad_norm": 0.511324942111969,
      "learning_rate": 3.4549150281252635e-06,
      "loss": 0.6541,
      "step": 200
    },
    {
      "epoch": 1.9296,
      "grad_norm": 0.41902709007263184,
      "learning_rate": 3.4016592544210937e-06,
      "loss": 0.5181,
      "step": 201
    },
    {
      "epoch": 1.9392,
      "grad_norm": 0.49513956904411316,
      "learning_rate": 3.3486046902241663e-06,
      "loss": 0.594,
      "step": 202
    },
    {
      "epoch": 1.9487999999999999,
      "grad_norm": 0.4521034359931946,
      "learning_rate": 3.295758014387375e-06,
      "loss": 0.5428,
      "step": 203
    },
    {
      "epoch": 1.9584000000000001,
      "grad_norm": 0.48090317845344543,
      "learning_rate": 3.2431258795932863e-06,
      "loss": 0.5857,
      "step": 204
    },
    {
      "epoch": 1.968,
      "grad_norm": 0.42190292477607727,
      "learning_rate": 3.1907149115166403e-06,
      "loss": 0.5711,
      "step": 205
    },
    {
      "epoch": 1.9776,
      "grad_norm": 0.5230044722557068,
      "learning_rate": 3.1385317079902743e-06,
      "loss": 0.5769,
      "step": 206
    },
    {
      "epoch": 1.9872,
      "grad_norm": 0.47211894392967224,
      "learning_rate": 3.0865828381745515e-06,
      "loss": 0.596,
      "step": 207
    },
    {
      "epoch": 1.9968,
      "grad_norm": 0.4404360353946686,
      "learning_rate": 3.0348748417303826e-06,
      "loss": 0.5447,
      "step": 208
    },
    {
      "epoch": 2.0064,
      "grad_norm": 1.2754789590835571,
      "learning_rate": 2.9834142279959754e-06,
      "loss": 1.0076,
      "step": 209
    },
    {
      "epoch": 2.016,
      "grad_norm": 0.445046067237854,
      "learning_rate": 2.932207475167398e-06,
      "loss": 0.5325,
      "step": 210
    },
    {
      "epoch": 2.0256,
      "grad_norm": 0.5322118401527405,
      "learning_rate": 2.8812610294830568e-06,
      "loss": 0.5812,
      "step": 211
    },
    {
      "epoch": 2.0352,
      "grad_norm": 0.5101974606513977,
      "learning_rate": 2.83058130441221e-06,
      "loss": 0.5126,
      "step": 212
    },
    {
      "epoch": 2.0448,
      "grad_norm": 0.469745397567749,
      "learning_rate": 2.7801746798475905e-06,
      "loss": 0.5893,
      "step": 213
    },
    {
      "epoch": 2.0544,
      "grad_norm": 0.47024786472320557,
      "learning_rate": 2.7300475013022666e-06,
      "loss": 0.5568,
      "step": 214
    },
    {
      "epoch": 2.064,
      "grad_norm": 0.49212148785591125,
      "learning_rate": 2.6802060791108304e-06,
      "loss": 0.5738,
      "step": 215
    },
    {
      "epoch": 2.0736,
      "grad_norm": 0.4421893060207367,
      "learning_rate": 2.6306566876350072e-06,
      "loss": 0.4556,
      "step": 216
    },
    {
      "epoch": 2.0832,
      "grad_norm": 0.47675591707229614,
      "learning_rate": 2.5814055644738013e-06,
      "loss": 0.5503,
      "step": 217
    },
    {
      "epoch": 2.0928,
      "grad_norm": 0.40364202857017517,
      "learning_rate": 2.532458909678266e-06,
      "loss": 0.507,
      "step": 218
    },
    {
      "epoch": 2.1024,
      "grad_norm": 0.49315252900123596,
      "learning_rate": 2.483822884971e-06,
      "loss": 0.5551,
      "step": 219
    },
    {
      "epoch": 2.112,
      "grad_norm": 0.4763312339782715,
      "learning_rate": 2.43550361297047e-06,
      "loss": 0.5125,
      "step": 220
    },
    {
      "epoch": 2.1216,
      "grad_norm": 0.4296324551105499,
      "learning_rate": 2.387507176420256e-06,
      "loss": 0.5489,
      "step": 221
    },
    {
      "epoch": 2.1312,
      "grad_norm": 0.4918951690196991,
      "learning_rate": 2.339839617423318e-06,
      "loss": 0.5195,
      "step": 222
    },
    {
      "epoch": 2.1408,
      "grad_norm": 0.4372878670692444,
      "learning_rate": 2.2925069366813718e-06,
      "loss": 0.5181,
      "step": 223
    },
    {
      "epoch": 2.1504,
      "grad_norm": 0.4088146686553955,
      "learning_rate": 2.245515092739488e-06,
      "loss": 0.5357,
      "step": 224
    },
    {
      "epoch": 2.16,
      "grad_norm": 0.43566828966140747,
      "learning_rate": 2.1988700012359865e-06,
      "loss": 0.5389,
      "step": 225
    },
    {
      "epoch": 2.1696,
      "grad_norm": 0.4104842245578766,
      "learning_rate": 2.1525775341577404e-06,
      "loss": 0.4976,
      "step": 226
    },
    {
      "epoch": 2.1792,
      "grad_norm": 0.41833817958831787,
      "learning_rate": 2.1066435191009717e-06,
      "loss": 0.5311,
      "step": 227
    },
    {
      "epoch": 2.1888,
      "grad_norm": 0.40475812554359436,
      "learning_rate": 2.061073738537635e-06,
      "loss": 0.5153,
      "step": 228
    },
    {
      "epoch": 2.1984,
      "grad_norm": 0.4294889271259308,
      "learning_rate": 2.0158739290874822e-06,
      "loss": 0.5571,
      "step": 229
    },
    {
      "epoch": 2.208,
      "grad_norm": 0.49284547567367554,
      "learning_rate": 1.971049780795901e-06,
      "loss": 0.5989,
      "step": 230
    },
    {
      "epoch": 2.2176,
      "grad_norm": 0.3920767307281494,
      "learning_rate": 1.9266069364176144e-06,
      "loss": 0.4818,
      "step": 231
    },
    {
      "epoch": 2.2272,
      "grad_norm": 0.4156140089035034,
      "learning_rate": 1.8825509907063328e-06,
      "loss": 0.5287,
      "step": 232
    },
    {
      "epoch": 2.2368,
      "grad_norm": 0.4189988672733307,
      "learning_rate": 1.838887489710452e-06,
      "loss": 0.5527,
      "step": 233
    },
    {
      "epoch": 2.2464,
      "grad_norm": 0.4066447913646698,
      "learning_rate": 1.7956219300748796e-06,
      "loss": 0.4751,
      "step": 234
    },
    {
      "epoch": 2.2560000000000002,
      "grad_norm": 0.394909143447876,
      "learning_rate": 1.7527597583490825e-06,
      "loss": 0.5778,
      "step": 235
    },
    {
      "epoch": 2.2656,
      "grad_norm": 0.4090768098831177,
      "learning_rate": 1.7103063703014372e-06,
      "loss": 0.5591,
      "step": 236
    },
    {
      "epoch": 2.2752,
      "grad_norm": 0.6263449788093567,
      "learning_rate": 1.6682671102399806e-06,
      "loss": 0.5066,
      "step": 237
    },
    {
      "epoch": 2.2848,
      "grad_norm": 0.42953482270240784,
      "learning_rate": 1.6266472703396286e-06,
      "loss": 0.5288,
      "step": 238
    },
    {
      "epoch": 2.2944,
      "grad_norm": 0.3976917266845703,
      "learning_rate": 1.5854520899759656e-06,
      "loss": 0.5317,
      "step": 239
    },
    {
      "epoch": 2.304,
      "grad_norm": 0.41823217272758484,
      "learning_rate": 1.544686755065677e-06,
      "loss": 0.5226,
      "step": 240
    },
    {
      "epoch": 2.3136,
      "grad_norm": 0.42223042249679565,
      "learning_rate": 1.5043563974137132e-06,
      "loss": 0.5536,
      "step": 241
    },
    {
      "epoch": 2.3232,
      "grad_norm": 0.42981192469596863,
      "learning_rate": 1.4644660940672628e-06,
      "loss": 0.5563,
      "step": 242
    },
    {
      "epoch": 2.3327999999999998,
      "grad_norm": 0.35931915044784546,
      "learning_rate": 1.4250208666766235e-06,
      "loss": 0.4857,
      "step": 243
    },
    {
      "epoch": 2.3424,
      "grad_norm": 0.38428106904029846,
      "learning_rate": 1.3860256808630429e-06,
      "loss": 0.4783,
      "step": 244
    },
    {
      "epoch": 2.352,
      "grad_norm": 0.4292716085910797,
      "learning_rate": 1.3474854455936126e-06,
      "loss": 0.5553,
      "step": 245
    },
    {
      "epoch": 2.3616,
      "grad_norm": 0.4062524437904358,
      "learning_rate": 1.3094050125632973e-06,
      "loss": 0.5389,
      "step": 246
    },
    {
      "epoch": 2.3712,
      "grad_norm": 0.3885515630245209,
      "learning_rate": 1.2717891755841722e-06,
      "loss": 0.4804,
      "step": 247
    },
    {
      "epoch": 2.3808,
      "grad_norm": 0.4065932035446167,
      "learning_rate": 1.234642669981946e-06,
      "loss": 0.5717,
      "step": 248
    },
    {
      "epoch": 2.3904,
      "grad_norm": 0.38877761363983154,
      "learning_rate": 1.1979701719998454e-06,
      "loss": 0.5226,
      "step": 249
    },
    {
      "epoch": 2.4,
      "grad_norm": 0.40351080894470215,
      "learning_rate": 1.1617762982099446e-06,
      "loss": 0.5395,
      "step": 250
    },
    {
      "epoch": 2.4096,
      "grad_norm": 0.37985244393348694,
      "learning_rate": 1.1260656049319957e-06,
      "loss": 0.509,
      "step": 251
    },
    {
      "epoch": 2.4192,
      "grad_norm": 0.43029841780662537,
      "learning_rate": 1.0908425876598512e-06,
      "loss": 0.5682,
      "step": 252
    },
    {
      "epoch": 2.4288,
      "grad_norm": 0.40593045949935913,
      "learning_rate": 1.0561116804955451e-06,
      "loss": 0.5389,
      "step": 253
    },
    {
      "epoch": 2.4384,
      "grad_norm": 0.36474233865737915,
      "learning_rate": 1.0218772555910955e-06,
      "loss": 0.4982,
      "step": 254
    },
    {
      "epoch": 2.448,
      "grad_norm": 0.4089374840259552,
      "learning_rate": 9.881436225981107e-07,
      "loss": 0.5787,
      "step": 255
    },
    {
      "epoch": 2.4576000000000002,
      "grad_norm": 0.36549311876296997,
      "learning_rate": 9.549150281252633e-07,
      "loss": 0.4473,
      "step": 256
    },
    {
      "epoch": 2.4672,
      "grad_norm": 0.41079387068748474,
      "learning_rate": 9.221956552036992e-07,
      "loss": 0.5655,
      "step": 257
    },
    {
      "epoch": 2.4768,
      "grad_norm": 0.4164627194404602,
      "learning_rate": 8.899896227604509e-07,
      "loss": 0.5304,
      "step": 258
    },
    {
      "epoch": 2.4864,
      "grad_norm": 0.42988118529319763,
      "learning_rate": 8.58300985099918e-07,
      "loss": 0.5282,
      "step": 259
    },
    {
      "epoch": 2.496,
      "grad_norm": 0.4137679934501648,
      "learning_rate": 8.271337313934869e-07,
      "loss": 0.5628,
      "step": 260
    },
    {
      "epoch": 2.5056000000000003,
      "grad_norm": 0.3814125657081604,
      "learning_rate": 7.964917851773496e-07,
      "loss": 0.5245,
      "step": 261
    },
    {
      "epoch": 2.5152,
      "grad_norm": 0.38825955986976624,
      "learning_rate": 7.663790038585794e-07,
      "loss": 0.4994,
      "step": 262
    },
    {
      "epoch": 2.5248,
      "grad_norm": 0.4235158860683441,
      "learning_rate": 7.367991782295392e-07,
      "loss": 0.5233,
      "step": 263
    },
    {
      "epoch": 2.5343999999999998,
      "grad_norm": 0.4122079908847809,
      "learning_rate": 7.077560319906696e-07,
      "loss": 0.5305,
      "step": 264
    },
    {
      "epoch": 2.544,
      "grad_norm": 0.38463661074638367,
      "learning_rate": 6.792532212817271e-07,
      "loss": 0.5075,
      "step": 265
    },
    {
      "epoch": 2.5536,
      "grad_norm": 0.46616417169570923,
      "learning_rate": 6.512943342215234e-07,
      "loss": 0.5858,
      "step": 266
    },
    {
      "epoch": 2.5632,
      "grad_norm": 0.3797391653060913,
      "learning_rate": 6.238828904562316e-07,
      "loss": 0.5053,
      "step": 267
    },
    {
      "epoch": 2.5728,
      "grad_norm": 0.40550941228866577,
      "learning_rate": 5.9702234071631e-07,
      "loss": 0.5295,
      "step": 268
    },
    {
      "epoch": 2.5824,
      "grad_norm": 0.3795751929283142,
      "learning_rate": 5.707160663821009e-07,
      "loss": 0.5053,
      "step": 269
    },
    {
      "epoch": 2.592,
      "grad_norm": 0.38254594802856445,
      "learning_rate": 5.449673790581611e-07,
      "loss": 0.5196,
      "step": 270
    },
    {
      "epoch": 2.6016,
      "grad_norm": 0.4230276644229889,
      "learning_rate": 5.197795201563744e-07,
      "loss": 0.5181,
      "step": 271
    },
    {
      "epoch": 2.6112,
      "grad_norm": 0.39369747042655945,
      "learning_rate": 4.951556604879049e-07,
      "loss": 0.5888,
      "step": 272
    },
    {
      "epoch": 2.6208,
      "grad_norm": 0.3621313273906708,
      "learning_rate": 4.710988998640298e-07,
      "loss": 0.5424,
      "step": 273
    },
    {
      "epoch": 2.6304,
      "grad_norm": 0.3762126564979553,
      "learning_rate": 4.4761226670592074e-07,
      "loss": 0.5004,
      "step": 274
    },
    {
      "epoch": 2.64,
      "grad_norm": 0.3839513957500458,
      "learning_rate": 4.2469871766340096e-07,
      "loss": 0.5305,
      "step": 275
    },
    {
      "epoch": 2.6496,
      "grad_norm": 0.40465694665908813,
      "learning_rate": 4.0236113724274716e-07,
      "loss": 0.5355,
      "step": 276
    },
    {
      "epoch": 2.6592000000000002,
      "grad_norm": 0.41210877895355225,
      "learning_rate": 3.8060233744356634e-07,
      "loss": 0.5671,
      "step": 277
    },
    {
      "epoch": 2.6688,
      "grad_norm": 0.3980127274990082,
      "learning_rate": 3.5942505740480583e-07,
      "loss": 0.5108,
      "step": 278
    },
    {
      "epoch": 2.6784,
      "grad_norm": 0.4046895503997803,
      "learning_rate": 3.3883196305992906e-07,
      "loss": 0.5229,
      "step": 279
    },
    {
      "epoch": 2.6879999999999997,
      "grad_norm": 0.38082319498062134,
      "learning_rate": 3.18825646801314e-07,
      "loss": 0.5244,
      "step": 280
    },
    {
      "epoch": 2.6976,
      "grad_norm": 0.4220086336135864,
      "learning_rate": 2.9940862715390483e-07,
      "loss": 0.5123,
      "step": 281
    },
    {
      "epoch": 2.7072000000000003,
      "grad_norm": 0.4062383472919464,
      "learning_rate": 2.8058334845816214e-07,
      "loss": 0.5092,
      "step": 282
    },
    {
      "epoch": 2.7168,
      "grad_norm": 0.386997252702713,
      "learning_rate": 2.6235218056235633e-07,
      "loss": 0.5442,
      "step": 283
    },
    {
      "epoch": 2.7264,
      "grad_norm": 0.39148667454719543,
      "learning_rate": 2.447174185242324e-07,
      "loss": 0.5711,
      "step": 284
    },
    {
      "epoch": 2.7359999999999998,
      "grad_norm": 0.37360548973083496,
      "learning_rate": 2.276812823220964e-07,
      "loss": 0.5,
      "step": 285
    },
    {
      "epoch": 2.7456,
      "grad_norm": 0.39551421999931335,
      "learning_rate": 2.1124591657534776e-07,
      "loss": 0.5256,
      "step": 286
    },
    {
      "epoch": 2.7552,
      "grad_norm": 0.3795803487300873,
      "learning_rate": 1.9541339027450256e-07,
      "loss": 0.5371,
      "step": 287
    },
    {
      "epoch": 2.7648,
      "grad_norm": 0.3761843740940094,
      "learning_rate": 1.801856965207338e-07,
      "loss": 0.5725,
      "step": 288
    },
    {
      "epoch": 2.7744,
      "grad_norm": 0.3336462676525116,
      "learning_rate": 1.6556475227496816e-07,
      "loss": 0.4762,
      "step": 289
    },
    {
      "epoch": 2.784,
      "grad_norm": 0.3762846887111664,
      "learning_rate": 1.5155239811656562e-07,
      "loss": 0.5404,
      "step": 290
    },
    {
      "epoch": 2.7936,
      "grad_norm": 0.3734802305698395,
      "learning_rate": 1.3815039801161723e-07,
      "loss": 0.5515,
      "step": 291
    },
    {
      "epoch": 2.8032,
      "grad_norm": 0.3967645466327667,
      "learning_rate": 1.253604390908819e-07,
      "loss": 0.5259,
      "step": 292
    },
    {
      "epoch": 2.8128,
      "grad_norm": 0.3874190151691437,
      "learning_rate": 1.1318413143740436e-07,
      "loss": 0.5366,
      "step": 293
    },
    {
      "epoch": 2.8224,
      "grad_norm": 0.36908280849456787,
      "learning_rate": 1.0162300788382263e-07,
      "loss": 0.5674,
      "step": 294
    },
    {
      "epoch": 2.832,
      "grad_norm": 0.34430015087127686,
      "learning_rate": 9.0678523819408e-08,
      "loss": 0.4563,
      "step": 295
    },
    {
      "epoch": 2.8416,
      "grad_norm": 0.4032260775566101,
      "learning_rate": 8.035205700685167e-08,
      "loss": 0.5201,
      "step": 296
    },
    {
      "epoch": 2.8512,
      "grad_norm": 0.385893315076828,
      "learning_rate": 7.064490740882057e-08,
      "loss": 0.574,
      "step": 297
    },
    {
      "epoch": 2.8608000000000002,
      "grad_norm": 0.3859463930130005,
      "learning_rate": 6.15582970243117e-08,
      "loss": 0.5587,
      "step": 298
    },
    {
      "epoch": 2.8704,
      "grad_norm": 0.35366320610046387,
      "learning_rate": 5.3093369734816824e-08,
      "loss": 0.4771,
      "step": 299
    },
    {
      "epoch": 2.88,
      "grad_norm": 0.35460808873176575,
      "learning_rate": 4.52511911603265e-08,
      "loss": 0.5471,
      "step": 300
    },
    {
      "epoch": 2.8895999999999997,
      "grad_norm": 0.36865419149398804,
      "learning_rate": 3.8032748525179684e-08,
      "loss": 0.5279,
      "step": 301
    },
    {
      "epoch": 2.8992,
      "grad_norm": 0.39643099904060364,
      "learning_rate": 3.143895053378698e-08,
      "loss": 0.5488,
      "step": 302
    },
    {
      "epoch": 2.9088000000000003,
      "grad_norm": 0.3732302486896515,
      "learning_rate": 2.547062725623828e-08,
      "loss": 0.5458,
      "step": 303
    },
    {
      "epoch": 2.9184,
      "grad_norm": 0.34452953934669495,
      "learning_rate": 2.012853002380466e-08,
      "loss": 0.4718,
      "step": 304
    },
    {
      "epoch": 2.928,
      "grad_norm": 0.3710247278213501,
      "learning_rate": 1.541333133436018e-08,
      "loss": 0.552,
      "step": 305
    },
    {
      "epoch": 2.9375999999999998,
      "grad_norm": 0.3867627680301666,
      "learning_rate": 1.132562476771959e-08,
      "loss": 0.5287,
      "step": 306
    },
    {
      "epoch": 2.9472,
      "grad_norm": 0.36105963587760925,
      "learning_rate": 7.865924910916977e-09,
      "loss": 0.4764,
      "step": 307
    },
    {
      "epoch": 2.9568,
      "grad_norm": 0.42076411843299866,
      "learning_rate": 5.034667293427053e-09,
      "loss": 0.5559,
      "step": 308
    },
    {
      "epoch": 2.9664,
      "grad_norm": 0.34399110078811646,
      "learning_rate": 2.8322083323334417e-09,
      "loss": 0.4897,
      "step": 309
    },
    {
      "epoch": 2.976,
      "grad_norm": 0.36756882071495056,
      "learning_rate": 1.2588252874673469e-09,
      "loss": 0.5773,
      "step": 310
    },
    {
      "epoch": 2.9856,
      "grad_norm": 0.3748956024646759,
      "learning_rate": 3.147162264971471e-10,
      "loss": 0.5365,
      "step": 311
    },
    {
      "epoch": 2.9952,
      "grad_norm": 0.3594142496585846,
      "learning_rate": 0.0,
      "loss": 0.5758,
      "step": 312
    },
    {
      "epoch": 2.9952,
      "step": 312,
      "total_flos": 312970832904192.0,
      "train_loss": 0.610789420035405,
      "train_runtime": 29538.0563,
      "train_samples_per_second": 1.016,
      "train_steps_per_second": 0.011
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 312,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "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": 312970832904192.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}