{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 4.935779816513762,
  "eval_steps": 500,
  "global_step": 360,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.013761467889908258,
      "grad_norm": 38.02450942993164,
      "learning_rate": 5.0000000000000004e-08,
      "loss": 3.125,
      "step": 1
    },
    {
      "epoch": 0.027522935779816515,
      "grad_norm": 37.864768981933594,
      "learning_rate": 1.0000000000000001e-07,
      "loss": 3.0998,
      "step": 2
    },
    {
      "epoch": 0.04128440366972477,
      "grad_norm": 38.34700012207031,
      "learning_rate": 1.5000000000000002e-07,
      "loss": 3.1533,
      "step": 3
    },
    {
      "epoch": 0.05504587155963303,
      "grad_norm": 38.33641815185547,
      "learning_rate": 2.0000000000000002e-07,
      "loss": 3.1542,
      "step": 4
    },
    {
      "epoch": 0.06880733944954129,
      "grad_norm": 38.064449310302734,
      "learning_rate": 2.5000000000000004e-07,
      "loss": 3.1153,
      "step": 5
    },
    {
      "epoch": 0.08256880733944955,
      "grad_norm": 37.92089080810547,
      "learning_rate": 3.0000000000000004e-07,
      "loss": 3.0867,
      "step": 6
    },
    {
      "epoch": 0.0963302752293578,
      "grad_norm": 38.120323181152344,
      "learning_rate": 3.5000000000000004e-07,
      "loss": 3.093,
      "step": 7
    },
    {
      "epoch": 0.11009174311926606,
      "grad_norm": 38.47222900390625,
      "learning_rate": 4.0000000000000003e-07,
      "loss": 3.1056,
      "step": 8
    },
    {
      "epoch": 0.12385321100917432,
      "grad_norm": 38.013702392578125,
      "learning_rate": 4.5000000000000003e-07,
      "loss": 3.0474,
      "step": 9
    },
    {
      "epoch": 0.13761467889908258,
      "grad_norm": 38.17593002319336,
      "learning_rate": 5.000000000000001e-07,
      "loss": 3.0264,
      "step": 10
    },
    {
      "epoch": 0.15137614678899083,
      "grad_norm": 38.60066604614258,
      "learning_rate": 5.5e-07,
      "loss": 2.9404,
      "step": 11
    },
    {
      "epoch": 0.1651376146788991,
      "grad_norm": 38.83498764038086,
      "learning_rate": 6.000000000000001e-07,
      "loss": 2.9571,
      "step": 12
    },
    {
      "epoch": 0.17889908256880735,
      "grad_norm": 38.942047119140625,
      "learning_rate": 6.5e-07,
      "loss": 2.8849,
      "step": 13
    },
    {
      "epoch": 0.1926605504587156,
      "grad_norm": 38.0286865234375,
      "learning_rate": 7.000000000000001e-07,
      "loss": 2.7486,
      "step": 14
    },
    {
      "epoch": 0.20642201834862386,
      "grad_norm": 38.31856155395508,
      "learning_rate": 7.5e-07,
      "loss": 2.6876,
      "step": 15
    },
    {
      "epoch": 0.22018348623853212,
      "grad_norm": 38.124759674072266,
      "learning_rate": 8.000000000000001e-07,
      "loss": 2.5992,
      "step": 16
    },
    {
      "epoch": 0.23394495412844038,
      "grad_norm": 36.59762191772461,
      "learning_rate": 8.500000000000001e-07,
      "loss": 2.4063,
      "step": 17
    },
    {
      "epoch": 0.24770642201834864,
      "grad_norm": 36.63874435424805,
      "learning_rate": 9.000000000000001e-07,
      "loss": 2.3109,
      "step": 18
    },
    {
      "epoch": 0.26146788990825687,
      "grad_norm": 36.768577575683594,
      "learning_rate": 9.500000000000001e-07,
      "loss": 2.1677,
      "step": 19
    },
    {
      "epoch": 0.27522935779816515,
      "grad_norm": 36.187137603759766,
      "learning_rate": 1.0000000000000002e-06,
      "loss": 1.9551,
      "step": 20
    },
    {
      "epoch": 0.2889908256880734,
      "grad_norm": 35.55617141723633,
      "learning_rate": 1.0500000000000001e-06,
      "loss": 1.8053,
      "step": 21
    },
    {
      "epoch": 0.30275229357798167,
      "grad_norm": 34.60952377319336,
      "learning_rate": 1.1e-06,
      "loss": 1.5654,
      "step": 22
    },
    {
      "epoch": 0.3165137614678899,
      "grad_norm": 33.69620895385742,
      "learning_rate": 1.1500000000000002e-06,
      "loss": 1.3454,
      "step": 23
    },
    {
      "epoch": 0.3302752293577982,
      "grad_norm": 34.33642578125,
      "learning_rate": 1.2000000000000002e-06,
      "loss": 1.2417,
      "step": 24
    },
    {
      "epoch": 0.3440366972477064,
      "grad_norm": 31.23066520690918,
      "learning_rate": 1.25e-06,
      "loss": 0.9839,
      "step": 25
    },
    {
      "epoch": 0.3577981651376147,
      "grad_norm": 25.810237884521484,
      "learning_rate": 1.3e-06,
      "loss": 0.793,
      "step": 26
    },
    {
      "epoch": 0.37155963302752293,
      "grad_norm": 23.06692886352539,
      "learning_rate": 1.3500000000000002e-06,
      "loss": 0.6082,
      "step": 27
    },
    {
      "epoch": 0.3853211009174312,
      "grad_norm": 19.828439712524414,
      "learning_rate": 1.4000000000000001e-06,
      "loss": 0.4845,
      "step": 28
    },
    {
      "epoch": 0.39908256880733944,
      "grad_norm": 14.150300025939941,
      "learning_rate": 1.45e-06,
      "loss": 0.348,
      "step": 29
    },
    {
      "epoch": 0.41284403669724773,
      "grad_norm": 9.044266700744629,
      "learning_rate": 1.5e-06,
      "loss": 0.2516,
      "step": 30
    },
    {
      "epoch": 0.42660550458715596,
      "grad_norm": 5.704404354095459,
      "learning_rate": 1.5500000000000002e-06,
      "loss": 0.177,
      "step": 31
    },
    {
      "epoch": 0.44036697247706424,
      "grad_norm": 3.2953503131866455,
      "learning_rate": 1.6000000000000001e-06,
      "loss": 0.1391,
      "step": 32
    },
    {
      "epoch": 0.4541284403669725,
      "grad_norm": 2.453219413757324,
      "learning_rate": 1.6500000000000003e-06,
      "loss": 0.0982,
      "step": 33
    },
    {
      "epoch": 0.46788990825688076,
      "grad_norm": 2.0325512886047363,
      "learning_rate": 1.7000000000000002e-06,
      "loss": 0.0807,
      "step": 34
    },
    {
      "epoch": 0.481651376146789,
      "grad_norm": 1.6322681903839111,
      "learning_rate": 1.75e-06,
      "loss": 0.0725,
      "step": 35
    },
    {
      "epoch": 0.4954128440366973,
      "grad_norm": 0.9713364839553833,
      "learning_rate": 1.8000000000000001e-06,
      "loss": 0.067,
      "step": 36
    },
    {
      "epoch": 0.5091743119266054,
      "grad_norm": 0.7980225682258606,
      "learning_rate": 1.85e-06,
      "loss": 0.0582,
      "step": 37
    },
    {
      "epoch": 0.5229357798165137,
      "grad_norm": 1.0616590976715088,
      "learning_rate": 1.9000000000000002e-06,
      "loss": 0.0562,
      "step": 38
    },
    {
      "epoch": 0.536697247706422,
      "grad_norm": 1.053462028503418,
      "learning_rate": 1.9500000000000004e-06,
      "loss": 0.0537,
      "step": 39
    },
    {
      "epoch": 0.5504587155963303,
      "grad_norm": 0.9452660083770752,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 0.0602,
      "step": 40
    },
    {
      "epoch": 0.5642201834862385,
      "grad_norm": 0.830368161201477,
      "learning_rate": 2.05e-06,
      "loss": 0.0549,
      "step": 41
    },
    {
      "epoch": 0.5779816513761468,
      "grad_norm": 0.5791187882423401,
      "learning_rate": 2.1000000000000002e-06,
      "loss": 0.0479,
      "step": 42
    },
    {
      "epoch": 0.591743119266055,
      "grad_norm": 0.44175243377685547,
      "learning_rate": 2.15e-06,
      "loss": 0.0461,
      "step": 43
    },
    {
      "epoch": 0.6055045871559633,
      "grad_norm": 0.37655699253082275,
      "learning_rate": 2.2e-06,
      "loss": 0.043,
      "step": 44
    },
    {
      "epoch": 0.6192660550458715,
      "grad_norm": 0.34382495284080505,
      "learning_rate": 2.25e-06,
      "loss": 0.0454,
      "step": 45
    },
    {
      "epoch": 0.6330275229357798,
      "grad_norm": 0.5047216415405273,
      "learning_rate": 2.3000000000000004e-06,
      "loss": 0.0437,
      "step": 46
    },
    {
      "epoch": 0.6467889908256881,
      "grad_norm": 0.6318779587745667,
      "learning_rate": 2.35e-06,
      "loss": 0.0468,
      "step": 47
    },
    {
      "epoch": 0.6605504587155964,
      "grad_norm": 0.5135455131530762,
      "learning_rate": 2.4000000000000003e-06,
      "loss": 0.0494,
      "step": 48
    },
    {
      "epoch": 0.6743119266055045,
      "grad_norm": 0.4802612066268921,
      "learning_rate": 2.4500000000000003e-06,
      "loss": 0.0441,
      "step": 49
    },
    {
      "epoch": 0.6880733944954128,
      "grad_norm": 0.6157718300819397,
      "learning_rate": 2.5e-06,
      "loss": 0.0398,
      "step": 50
    },
    {
      "epoch": 0.7018348623853211,
      "grad_norm": 0.4327130913734436,
      "learning_rate": 2.55e-06,
      "loss": 0.0438,
      "step": 51
    },
    {
      "epoch": 0.7155963302752294,
      "grad_norm": 0.46133658289909363,
      "learning_rate": 2.6e-06,
      "loss": 0.041,
      "step": 52
    },
    {
      "epoch": 0.7293577981651376,
      "grad_norm": 0.5729146003723145,
      "learning_rate": 2.6500000000000005e-06,
      "loss": 0.0406,
      "step": 53
    },
    {
      "epoch": 0.7431192660550459,
      "grad_norm": 0.32373812794685364,
      "learning_rate": 2.7000000000000004e-06,
      "loss": 0.0419,
      "step": 54
    },
    {
      "epoch": 0.7568807339449541,
      "grad_norm": 0.29006752371788025,
      "learning_rate": 2.7500000000000004e-06,
      "loss": 0.0415,
      "step": 55
    },
    {
      "epoch": 0.7706422018348624,
      "grad_norm": 0.31038960814476013,
      "learning_rate": 2.8000000000000003e-06,
      "loss": 0.0344,
      "step": 56
    },
    {
      "epoch": 0.7844036697247706,
      "grad_norm": 0.2324836701154709,
      "learning_rate": 2.85e-06,
      "loss": 0.0374,
      "step": 57
    },
    {
      "epoch": 0.7981651376146789,
      "grad_norm": 0.5083625912666321,
      "learning_rate": 2.9e-06,
      "loss": 0.0324,
      "step": 58
    },
    {
      "epoch": 0.8119266055045872,
      "grad_norm": 0.2873130142688751,
      "learning_rate": 2.95e-06,
      "loss": 0.0403,
      "step": 59
    },
    {
      "epoch": 0.8256880733944955,
      "grad_norm": 0.437663197517395,
      "learning_rate": 3e-06,
      "loss": 0.0368,
      "step": 60
    },
    {
      "epoch": 0.8394495412844036,
      "grad_norm": 0.5645247101783752,
      "learning_rate": 3.05e-06,
      "loss": 0.0386,
      "step": 61
    },
    {
      "epoch": 0.8532110091743119,
      "grad_norm": 0.40374210476875305,
      "learning_rate": 3.1000000000000004e-06,
      "loss": 0.0425,
      "step": 62
    },
    {
      "epoch": 0.8669724770642202,
      "grad_norm": 0.46468955278396606,
      "learning_rate": 3.1500000000000003e-06,
      "loss": 0.0323,
      "step": 63
    },
    {
      "epoch": 0.8807339449541285,
      "grad_norm": 0.29952895641326904,
      "learning_rate": 3.2000000000000003e-06,
      "loss": 0.0325,
      "step": 64
    },
    {
      "epoch": 0.8944954128440367,
      "grad_norm": 0.3678436279296875,
      "learning_rate": 3.2500000000000002e-06,
      "loss": 0.036,
      "step": 65
    },
    {
      "epoch": 0.908256880733945,
      "grad_norm": 0.5068934559822083,
      "learning_rate": 3.3000000000000006e-06,
      "loss": 0.0357,
      "step": 66
    },
    {
      "epoch": 0.9220183486238532,
      "grad_norm": 0.2723177671432495,
      "learning_rate": 3.3500000000000005e-06,
      "loss": 0.0333,
      "step": 67
    },
    {
      "epoch": 0.9357798165137615,
      "grad_norm": 0.41696834564208984,
      "learning_rate": 3.4000000000000005e-06,
      "loss": 0.0347,
      "step": 68
    },
    {
      "epoch": 0.9495412844036697,
      "grad_norm": 0.2582981288433075,
      "learning_rate": 3.45e-06,
      "loss": 0.0283,
      "step": 69
    },
    {
      "epoch": 0.963302752293578,
      "grad_norm": 0.40648311376571655,
      "learning_rate": 3.5e-06,
      "loss": 0.0293,
      "step": 70
    },
    {
      "epoch": 0.9770642201834863,
      "grad_norm": 0.4149394631385803,
      "learning_rate": 3.5500000000000003e-06,
      "loss": 0.0311,
      "step": 71
    },
    {
      "epoch": 0.9908256880733946,
      "grad_norm": 0.3800952136516571,
      "learning_rate": 3.6000000000000003e-06,
      "loss": 0.0346,
      "step": 72
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.3800952136516571,
      "learning_rate": 3.65e-06,
      "loss": 0.0291,
      "step": 73
    },
    {
      "epoch": 1.0137614678899083,
      "grad_norm": 0.5016496181488037,
      "learning_rate": 3.7e-06,
      "loss": 0.0286,
      "step": 74
    },
    {
      "epoch": 1.0275229357798166,
      "grad_norm": 0.3533766567707062,
      "learning_rate": 3.7500000000000005e-06,
      "loss": 0.0321,
      "step": 75
    },
    {
      "epoch": 1.0412844036697249,
      "grad_norm": 0.2785470485687256,
      "learning_rate": 3.8000000000000005e-06,
      "loss": 0.0277,
      "step": 76
    },
    {
      "epoch": 1.0550458715596331,
      "grad_norm": 0.4530641436576843,
      "learning_rate": 3.85e-06,
      "loss": 0.0294,
      "step": 77
    },
    {
      "epoch": 1.0688073394495412,
      "grad_norm": 0.3170749843120575,
      "learning_rate": 3.900000000000001e-06,
      "loss": 0.0274,
      "step": 78
    },
    {
      "epoch": 1.0825688073394495,
      "grad_norm": 0.26502758264541626,
      "learning_rate": 3.95e-06,
      "loss": 0.0284,
      "step": 79
    },
    {
      "epoch": 1.0963302752293578,
      "grad_norm": 0.5486436486244202,
      "learning_rate": 4.000000000000001e-06,
      "loss": 0.0285,
      "step": 80
    },
    {
      "epoch": 1.110091743119266,
      "grad_norm": 0.24868083000183105,
      "learning_rate": 4.05e-06,
      "loss": 0.0301,
      "step": 81
    },
    {
      "epoch": 1.1238532110091743,
      "grad_norm": 0.3448987305164337,
      "learning_rate": 4.1e-06,
      "loss": 0.0261,
      "step": 82
    },
    {
      "epoch": 1.1376146788990826,
      "grad_norm": 0.3330553472042084,
      "learning_rate": 4.15e-06,
      "loss": 0.03,
      "step": 83
    },
    {
      "epoch": 1.151376146788991,
      "grad_norm": 0.3379852771759033,
      "learning_rate": 4.2000000000000004e-06,
      "loss": 0.0286,
      "step": 84
    },
    {
      "epoch": 1.165137614678899,
      "grad_norm": 0.23678433895111084,
      "learning_rate": 4.25e-06,
      "loss": 0.0245,
      "step": 85
    },
    {
      "epoch": 1.1788990825688073,
      "grad_norm": 0.24502314627170563,
      "learning_rate": 4.3e-06,
      "loss": 0.0267,
      "step": 86
    },
    {
      "epoch": 1.1926605504587156,
      "grad_norm": 0.34288597106933594,
      "learning_rate": 4.350000000000001e-06,
      "loss": 0.0259,
      "step": 87
    },
    {
      "epoch": 1.2064220183486238,
      "grad_norm": 0.20595045387744904,
      "learning_rate": 4.4e-06,
      "loss": 0.0241,
      "step": 88
    },
    {
      "epoch": 1.2201834862385321,
      "grad_norm": 0.28399360179901123,
      "learning_rate": 4.450000000000001e-06,
      "loss": 0.0277,
      "step": 89
    },
    {
      "epoch": 1.2339449541284404,
      "grad_norm": 0.273929238319397,
      "learning_rate": 4.5e-06,
      "loss": 0.0261,
      "step": 90
    },
    {
      "epoch": 1.2477064220183487,
      "grad_norm": 0.24288330972194672,
      "learning_rate": 4.5500000000000005e-06,
      "loss": 0.0267,
      "step": 91
    },
    {
      "epoch": 1.261467889908257,
      "grad_norm": 0.42502400279045105,
      "learning_rate": 4.600000000000001e-06,
      "loss": 0.0252,
      "step": 92
    },
    {
      "epoch": 1.2752293577981653,
      "grad_norm": 0.17670072615146637,
      "learning_rate": 4.65e-06,
      "loss": 0.0231,
      "step": 93
    },
    {
      "epoch": 1.2889908256880733,
      "grad_norm": 0.23585423827171326,
      "learning_rate": 4.7e-06,
      "loss": 0.0213,
      "step": 94
    },
    {
      "epoch": 1.3027522935779816,
      "grad_norm": 0.32558879256248474,
      "learning_rate": 4.75e-06,
      "loss": 0.0226,
      "step": 95
    },
    {
      "epoch": 1.31651376146789,
      "grad_norm": 0.2908780872821808,
      "learning_rate": 4.800000000000001e-06,
      "loss": 0.0274,
      "step": 96
    },
    {
      "epoch": 1.3302752293577982,
      "grad_norm": 0.3725607991218567,
      "learning_rate": 4.85e-06,
      "loss": 0.0241,
      "step": 97
    },
    {
      "epoch": 1.3440366972477065,
      "grad_norm": 0.3833301067352295,
      "learning_rate": 4.9000000000000005e-06,
      "loss": 0.0252,
      "step": 98
    },
    {
      "epoch": 1.3577981651376148,
      "grad_norm": 0.32000771164894104,
      "learning_rate": 4.95e-06,
      "loss": 0.0236,
      "step": 99
    },
    {
      "epoch": 1.3715596330275228,
      "grad_norm": 0.3203510344028473,
      "learning_rate": 5e-06,
      "loss": 0.0235,
      "step": 100
    },
    {
      "epoch": 1.385321100917431,
      "grad_norm": 0.20016217231750488,
      "learning_rate": 4.999888074163108e-06,
      "loss": 0.0218,
      "step": 101
    },
    {
      "epoch": 1.3990825688073394,
      "grad_norm": 0.3012026250362396,
      "learning_rate": 4.999552306674345e-06,
      "loss": 0.0234,
      "step": 102
    },
    {
      "epoch": 1.4128440366972477,
      "grad_norm": 0.22818222641944885,
      "learning_rate": 4.998992727598557e-06,
      "loss": 0.0212,
      "step": 103
    },
    {
      "epoch": 1.426605504587156,
      "grad_norm": 0.24128392338752747,
      "learning_rate": 4.998209387040829e-06,
      "loss": 0.0207,
      "step": 104
    },
    {
      "epoch": 1.4403669724770642,
      "grad_norm": 0.3475555181503296,
      "learning_rate": 4.9972023551419995e-06,
      "loss": 0.0246,
      "step": 105
    },
    {
      "epoch": 1.4541284403669725,
      "grad_norm": 0.21223627030849457,
      "learning_rate": 4.995971722072379e-06,
      "loss": 0.0237,
      "step": 106
    },
    {
      "epoch": 1.4678899082568808,
      "grad_norm": 0.3776336908340454,
      "learning_rate": 4.9945175980236745e-06,
      "loss": 0.0218,
      "step": 107
    },
    {
      "epoch": 1.481651376146789,
      "grad_norm": 0.25027793645858765,
      "learning_rate": 4.992840113199131e-06,
      "loss": 0.0214,
      "step": 108
    },
    {
      "epoch": 1.4954128440366974,
      "grad_norm": 0.2559281885623932,
      "learning_rate": 4.990939417801859e-06,
      "loss": 0.0213,
      "step": 109
    },
    {
      "epoch": 1.5091743119266054,
      "grad_norm": 0.33694687485694885,
      "learning_rate": 4.988815682021398e-06,
      "loss": 0.0182,
      "step": 110
    },
    {
      "epoch": 1.5229357798165137,
      "grad_norm": 0.3175147473812103,
      "learning_rate": 4.986469096018472e-06,
      "loss": 0.0213,
      "step": 111
    },
    {
      "epoch": 1.536697247706422,
      "grad_norm": 0.48777177929878235,
      "learning_rate": 4.983899869907963e-06,
      "loss": 0.0186,
      "step": 112
    },
    {
      "epoch": 1.5504587155963303,
      "grad_norm": 0.34633246064186096,
      "learning_rate": 4.981108233740096e-06,
      "loss": 0.0224,
      "step": 113
    },
    {
      "epoch": 1.5642201834862384,
      "grad_norm": 0.17302758991718292,
      "learning_rate": 4.978094437479843e-06,
      "loss": 0.0189,
      "step": 114
    },
    {
      "epoch": 1.5779816513761467,
      "grad_norm": 0.24110931158065796,
      "learning_rate": 4.97485875098454e-06,
      "loss": 0.0174,
      "step": 115
    },
    {
      "epoch": 1.591743119266055,
      "grad_norm": 0.2965283989906311,
      "learning_rate": 4.971401463979722e-06,
      "loss": 0.0174,
      "step": 116
    },
    {
      "epoch": 1.6055045871559632,
      "grad_norm": 0.34184500575065613,
      "learning_rate": 4.967722886033181e-06,
      "loss": 0.0162,
      "step": 117
    },
    {
      "epoch": 1.6192660550458715,
      "grad_norm": 0.3942873179912567,
      "learning_rate": 4.963823346527249e-06,
      "loss": 0.0176,
      "step": 118
    },
    {
      "epoch": 1.6330275229357798,
      "grad_norm": 0.21840929985046387,
      "learning_rate": 4.959703194629304e-06,
      "loss": 0.018,
      "step": 119
    },
    {
      "epoch": 1.646788990825688,
      "grad_norm": 0.23672759532928467,
      "learning_rate": 4.955362799260507e-06,
      "loss": 0.0204,
      "step": 120
    },
    {
      "epoch": 1.6605504587155964,
      "grad_norm": 0.18776445090770721,
      "learning_rate": 4.950802549062764e-06,
      "loss": 0.0173,
      "step": 121
    },
    {
      "epoch": 1.6743119266055047,
      "grad_norm": 0.279297411441803,
      "learning_rate": 4.946022852363932e-06,
      "loss": 0.0177,
      "step": 122
    },
    {
      "epoch": 1.688073394495413,
      "grad_norm": 0.20893588662147522,
      "learning_rate": 4.9410241371412525e-06,
      "loss": 0.0189,
      "step": 123
    },
    {
      "epoch": 1.7018348623853212,
      "grad_norm": 0.26409876346588135,
      "learning_rate": 4.935806850983034e-06,
      "loss": 0.0195,
      "step": 124
    },
    {
      "epoch": 1.7155963302752295,
      "grad_norm": 0.23863324522972107,
      "learning_rate": 4.9303714610485705e-06,
      "loss": 0.0151,
      "step": 125
    },
    {
      "epoch": 1.7293577981651376,
      "grad_norm": 0.25934213399887085,
      "learning_rate": 4.924718454026318e-06,
      "loss": 0.0157,
      "step": 126
    },
    {
      "epoch": 1.7431192660550459,
      "grad_norm": 0.2923693358898163,
      "learning_rate": 4.918848336090309e-06,
      "loss": 0.0155,
      "step": 127
    },
    {
      "epoch": 1.7568807339449541,
      "grad_norm": 0.16973069310188293,
      "learning_rate": 4.912761632854834e-06,
      "loss": 0.0156,
      "step": 128
    },
    {
      "epoch": 1.7706422018348624,
      "grad_norm": 0.25908610224723816,
      "learning_rate": 4.906458889327375e-06,
      "loss": 0.0159,
      "step": 129
    },
    {
      "epoch": 1.7844036697247705,
      "grad_norm": 0.27444136142730713,
      "learning_rate": 4.899940669859807e-06,
      "loss": 0.0159,
      "step": 130
    },
    {
      "epoch": 1.7981651376146788,
      "grad_norm": 0.25167539715766907,
      "learning_rate": 4.893207558097867e-06,
      "loss": 0.0147,
      "step": 131
    },
    {
      "epoch": 1.811926605504587,
      "grad_norm": 0.25777608156204224,
      "learning_rate": 4.8862601569288885e-06,
      "loss": 0.016,
      "step": 132
    },
    {
      "epoch": 1.8256880733944953,
      "grad_norm": 0.24190428853034973,
      "learning_rate": 4.879099088427824e-06,
      "loss": 0.0128,
      "step": 133
    },
    {
      "epoch": 1.8394495412844036,
      "grad_norm": 0.17028410732746124,
      "learning_rate": 4.871724993801541e-06,
      "loss": 0.0145,
      "step": 134
    },
    {
      "epoch": 1.853211009174312,
      "grad_norm": 0.25141075253486633,
      "learning_rate": 4.864138533331411e-06,
      "loss": 0.0144,
      "step": 135
    },
    {
      "epoch": 1.8669724770642202,
      "grad_norm": 0.6400424838066101,
      "learning_rate": 4.8563403863141825e-06,
      "loss": 0.0128,
      "step": 136
    },
    {
      "epoch": 1.8807339449541285,
      "grad_norm": 0.22450514137744904,
      "learning_rate": 4.84833125100116e-06,
      "loss": 0.0167,
      "step": 137
    },
    {
      "epoch": 1.8944954128440368,
      "grad_norm": 0.19940154254436493,
      "learning_rate": 4.840111844535682e-06,
      "loss": 0.0154,
      "step": 138
    },
    {
      "epoch": 1.908256880733945,
      "grad_norm": 0.2946206033229828,
      "learning_rate": 4.8316829028889076e-06,
      "loss": 0.0158,
      "step": 139
    },
    {
      "epoch": 1.9220183486238533,
      "grad_norm": 0.4694023132324219,
      "learning_rate": 4.823045180793914e-06,
      "loss": 0.0153,
      "step": 140
    },
    {
      "epoch": 1.9357798165137616,
      "grad_norm": 0.23655226826667786,
      "learning_rate": 4.8141994516781196e-06,
      "loss": 0.0109,
      "step": 141
    },
    {
      "epoch": 1.9495412844036697,
      "grad_norm": 0.23846553266048431,
      "learning_rate": 4.805146507594034e-06,
      "loss": 0.0129,
      "step": 142
    },
    {
      "epoch": 1.963302752293578,
      "grad_norm": 0.2763686180114746,
      "learning_rate": 4.7958871591483305e-06,
      "loss": 0.0124,
      "step": 143
    },
    {
      "epoch": 1.9770642201834863,
      "grad_norm": 0.23522883653640747,
      "learning_rate": 4.786422235429269e-06,
      "loss": 0.0125,
      "step": 144
    },
    {
      "epoch": 1.9908256880733946,
      "grad_norm": 0.27468597888946533,
      "learning_rate": 4.776752583932455e-06,
      "loss": 0.0137,
      "step": 145
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.26614058017730713,
      "learning_rate": 4.766879070484957e-06,
      "loss": 0.0113,
      "step": 146
    },
    {
      "epoch": 2.0137614678899083,
      "grad_norm": 0.2862620949745178,
      "learning_rate": 4.756802579167781e-06,
      "loss": 0.0104,
      "step": 147
    },
    {
      "epoch": 2.0275229357798166,
      "grad_norm": 0.1770494133234024,
      "learning_rate": 4.746524012236706e-06,
      "loss": 0.0102,
      "step": 148
    },
    {
      "epoch": 2.041284403669725,
      "grad_norm": 0.1850830465555191,
      "learning_rate": 4.736044290041496e-06,
      "loss": 0.0125,
      "step": 149
    },
    {
      "epoch": 2.055045871559633,
      "grad_norm": 0.20552438497543335,
      "learning_rate": 4.725364350943492e-06,
      "loss": 0.0091,
      "step": 150
    },
    {
      "epoch": 2.0688073394495414,
      "grad_norm": 0.12979158759117126,
      "learning_rate": 4.714485151231593e-06,
      "loss": 0.0113,
      "step": 151
    },
    {
      "epoch": 2.0825688073394497,
      "grad_norm": 0.16992178559303284,
      "learning_rate": 4.703407665036622e-06,
      "loss": 0.0075,
      "step": 152
    },
    {
      "epoch": 2.096330275229358,
      "grad_norm": 0.1473352611064911,
      "learning_rate": 4.692132884244113e-06,
      "loss": 0.0091,
      "step": 153
    },
    {
      "epoch": 2.1100917431192663,
      "grad_norm": 0.24198868870735168,
      "learning_rate": 4.680661818405485e-06,
      "loss": 0.0073,
      "step": 154
    },
    {
      "epoch": 2.123853211009174,
      "grad_norm": 0.25488734245300293,
      "learning_rate": 4.668995494647653e-06,
      "loss": 0.0098,
      "step": 155
    },
    {
      "epoch": 2.1376146788990824,
      "grad_norm": 0.17986920475959778,
      "learning_rate": 4.657134957581057e-06,
      "loss": 0.0081,
      "step": 156
    },
    {
      "epoch": 2.1513761467889907,
      "grad_norm": 0.22465726733207703,
      "learning_rate": 4.645081269206128e-06,
      "loss": 0.0074,
      "step": 157
    },
    {
      "epoch": 2.165137614678899,
      "grad_norm": 0.19391046464443207,
      "learning_rate": 4.632835508818192e-06,
      "loss": 0.0077,
      "step": 158
    },
    {
      "epoch": 2.1788990825688073,
      "grad_norm": 0.25997886061668396,
      "learning_rate": 4.620398772910833e-06,
      "loss": 0.0074,
      "step": 159
    },
    {
      "epoch": 2.1926605504587156,
      "grad_norm": 0.2422141134738922,
      "learning_rate": 4.607772175077712e-06,
      "loss": 0.0103,
      "step": 160
    },
    {
      "epoch": 2.206422018348624,
      "grad_norm": 0.2785587012767792,
      "learning_rate": 4.59495684591285e-06,
      "loss": 0.0079,
      "step": 161
    },
    {
      "epoch": 2.220183486238532,
      "grad_norm": 0.19141560792922974,
      "learning_rate": 4.581953932909403e-06,
      "loss": 0.0068,
      "step": 162
    },
    {
      "epoch": 2.2339449541284404,
      "grad_norm": 0.152049720287323,
      "learning_rate": 4.5687646003569055e-06,
      "loss": 0.0063,
      "step": 163
    },
    {
      "epoch": 2.2477064220183487,
      "grad_norm": 0.19313585758209229,
      "learning_rate": 4.555390029237026e-06,
      "loss": 0.0079,
      "step": 164
    },
    {
      "epoch": 2.261467889908257,
      "grad_norm": 0.19979022443294525,
      "learning_rate": 4.541831417117815e-06,
      "loss": 0.0085,
      "step": 165
    },
    {
      "epoch": 2.2752293577981653,
      "grad_norm": 0.2902522683143616,
      "learning_rate": 4.528089978046481e-06,
      "loss": 0.0082,
      "step": 166
    },
    {
      "epoch": 2.2889908256880735,
      "grad_norm": 0.5130491256713867,
      "learning_rate": 4.514166942440679e-06,
      "loss": 0.0067,
      "step": 167
    },
    {
      "epoch": 2.302752293577982,
      "grad_norm": 0.15980036556720734,
      "learning_rate": 4.5000635569783365e-06,
      "loss": 0.0093,
      "step": 168
    },
    {
      "epoch": 2.31651376146789,
      "grad_norm": 0.19247184693813324,
      "learning_rate": 4.4857810844860325e-06,
      "loss": 0.0101,
      "step": 169
    },
    {
      "epoch": 2.330275229357798,
      "grad_norm": 0.19173133373260498,
      "learning_rate": 4.471320803825915e-06,
      "loss": 0.0065,
      "step": 170
    },
    {
      "epoch": 2.3440366972477062,
      "grad_norm": 0.1779325157403946,
      "learning_rate": 4.4566840097811956e-06,
      "loss": 0.0059,
      "step": 171
    },
    {
      "epoch": 2.3577981651376145,
      "grad_norm": 0.1639624983072281,
      "learning_rate": 4.4418720129402145e-06,
      "loss": 0.0052,
      "step": 172
    },
    {
      "epoch": 2.371559633027523,
      "grad_norm": 0.2355070263147354,
      "learning_rate": 4.426886139579083e-06,
      "loss": 0.0046,
      "step": 173
    },
    {
      "epoch": 2.385321100917431,
      "grad_norm": 0.20461603999137878,
      "learning_rate": 4.411727731542937e-06,
      "loss": 0.0071,
      "step": 174
    },
    {
      "epoch": 2.3990825688073394,
      "grad_norm": 0.20251843333244324,
      "learning_rate": 4.39639814612578e-06,
      "loss": 0.0054,
      "step": 175
    },
    {
      "epoch": 2.4128440366972477,
      "grad_norm": 0.17055197060108185,
      "learning_rate": 4.3808987559489536e-06,
      "loss": 0.0052,
      "step": 176
    },
    {
      "epoch": 2.426605504587156,
      "grad_norm": 0.2508833706378937,
      "learning_rate": 4.365230948838232e-06,
      "loss": 0.0071,
      "step": 177
    },
    {
      "epoch": 2.4403669724770642,
      "grad_norm": 0.21865351498126984,
      "learning_rate": 4.349396127699552e-06,
      "loss": 0.0068,
      "step": 178
    },
    {
      "epoch": 2.4541284403669725,
      "grad_norm": 0.225867360830307,
      "learning_rate": 4.3333957103934025e-06,
      "loss": 0.0072,
      "step": 179
    },
    {
      "epoch": 2.467889908256881,
      "grad_norm": 0.48035845160484314,
      "learning_rate": 4.317231129607859e-06,
      "loss": 0.006,
      "step": 180
    },
    {
      "epoch": 2.481651376146789,
      "grad_norm": 0.17027413845062256,
      "learning_rate": 4.30090383273031e-06,
      "loss": 0.0043,
      "step": 181
    },
    {
      "epoch": 2.4954128440366974,
      "grad_norm": 0.2660333216190338,
      "learning_rate": 4.2844152817178476e-06,
      "loss": 0.0082,
      "step": 182
    },
    {
      "epoch": 2.5091743119266052,
      "grad_norm": 0.2485552728176117,
      "learning_rate": 4.267766952966369e-06,
      "loss": 0.0059,
      "step": 183
    },
    {
      "epoch": 2.522935779816514,
      "grad_norm": 0.2048122137784958,
      "learning_rate": 4.2509603371783776e-06,
      "loss": 0.0071,
      "step": 184
    },
    {
      "epoch": 2.536697247706422,
      "grad_norm": 0.1725567877292633,
      "learning_rate": 4.233996939229502e-06,
      "loss": 0.0053,
      "step": 185
    },
    {
      "epoch": 2.5504587155963305,
      "grad_norm": 0.14592835307121277,
      "learning_rate": 4.216878278033753e-06,
      "loss": 0.0044,
      "step": 186
    },
    {
      "epoch": 2.5642201834862384,
      "grad_norm": 0.2868310213088989,
      "learning_rate": 4.199605886407515e-06,
      "loss": 0.0057,
      "step": 187
    },
    {
      "epoch": 2.5779816513761467,
      "grad_norm": 0.5002567172050476,
      "learning_rate": 4.1821813109322975e-06,
      "loss": 0.0072,
      "step": 188
    },
    {
      "epoch": 2.591743119266055,
      "grad_norm": 0.199919655919075,
      "learning_rate": 4.164606111816256e-06,
      "loss": 0.0043,
      "step": 189
    },
    {
      "epoch": 2.6055045871559632,
      "grad_norm": 0.17280228435993195,
      "learning_rate": 4.146881862754485e-06,
      "loss": 0.0048,
      "step": 190
    },
    {
      "epoch": 2.6192660550458715,
      "grad_norm": 0.17505577206611633,
      "learning_rate": 4.129010150788112e-06,
      "loss": 0.0037,
      "step": 191
    },
    {
      "epoch": 2.63302752293578,
      "grad_norm": 0.21955189108848572,
      "learning_rate": 4.110992576162193e-06,
      "loss": 0.0044,
      "step": 192
    },
    {
      "epoch": 2.646788990825688,
      "grad_norm": 0.22081787884235382,
      "learning_rate": 4.092830752182423e-06,
      "loss": 0.0036,
      "step": 193
    },
    {
      "epoch": 2.6605504587155964,
      "grad_norm": 0.26361310482025146,
      "learning_rate": 4.074526305070679e-06,
      "loss": 0.0062,
      "step": 194
    },
    {
      "epoch": 2.6743119266055047,
      "grad_norm": 0.15528841316699982,
      "learning_rate": 4.056080873819412e-06,
      "loss": 0.0029,
      "step": 195
    },
    {
      "epoch": 2.688073394495413,
      "grad_norm": 6.002184867858887,
      "learning_rate": 4.037496110044885e-06,
      "loss": 0.0062,
      "step": 196
    },
    {
      "epoch": 2.7018348623853212,
      "grad_norm": 0.12827160954475403,
      "learning_rate": 4.018773677839289e-06,
      "loss": 0.0025,
      "step": 197
    },
    {
      "epoch": 2.7155963302752295,
      "grad_norm": 0.21857935190200806,
      "learning_rate": 3.999915253621739e-06,
      "loss": 0.004,
      "step": 198
    },
    {
      "epoch": 2.729357798165138,
      "grad_norm": 0.231048583984375,
      "learning_rate": 3.980922525988167e-06,
      "loss": 0.0034,
      "step": 199
    },
    {
      "epoch": 2.7431192660550456,
      "grad_norm": 0.16589054465293884,
      "learning_rate": 3.961797195560118e-06,
      "loss": 0.0045,
      "step": 200
    },
    {
      "epoch": 2.7568807339449544,
      "grad_norm": 0.2721656858921051,
      "learning_rate": 3.942540974832486e-06,
      "loss": 0.0056,
      "step": 201
    },
    {
      "epoch": 2.770642201834862,
      "grad_norm": 0.12930598855018616,
      "learning_rate": 3.9231555880201655e-06,
      "loss": 0.0037,
      "step": 202
    },
    {
      "epoch": 2.7844036697247705,
      "grad_norm": 0.299055814743042,
      "learning_rate": 3.903642770903671e-06,
      "loss": 0.0036,
      "step": 203
    },
    {
      "epoch": 2.7981651376146788,
      "grad_norm": 0.15592966973781586,
      "learning_rate": 3.884004270673711e-06,
      "loss": 0.0023,
      "step": 204
    },
    {
      "epoch": 2.811926605504587,
      "grad_norm": 0.10496195405721664,
      "learning_rate": 3.864241845774746e-06,
      "loss": 0.0029,
      "step": 205
    },
    {
      "epoch": 2.8256880733944953,
      "grad_norm": 0.1604638695716858,
      "learning_rate": 3.844357265747531e-06,
      "loss": 0.0032,
      "step": 206
    },
    {
      "epoch": 2.8394495412844036,
      "grad_norm": 0.1436438113451004,
      "learning_rate": 3.8243523110706736e-06,
      "loss": 0.0035,
      "step": 207
    },
    {
      "epoch": 2.853211009174312,
      "grad_norm": 0.176204651594162,
      "learning_rate": 3.8042287730012117e-06,
      "loss": 0.0022,
      "step": 208
    },
    {
      "epoch": 2.86697247706422,
      "grad_norm": 0.3500923216342926,
      "learning_rate": 3.7839884534142157e-06,
      "loss": 0.0022,
      "step": 209
    },
    {
      "epoch": 2.8807339449541285,
      "grad_norm": 0.2319999784231186,
      "learning_rate": 3.7636331646414524e-06,
      "loss": 0.003,
      "step": 210
    },
    {
      "epoch": 2.8944954128440368,
      "grad_norm": 0.1358226090669632,
      "learning_rate": 3.7431647293091076e-06,
      "loss": 0.0023,
      "step": 211
    },
    {
      "epoch": 2.908256880733945,
      "grad_norm": 0.1323612779378891,
      "learning_rate": 3.7225849801745835e-06,
      "loss": 0.0021,
      "step": 212
    },
    {
      "epoch": 2.9220183486238533,
      "grad_norm": 0.10968377441167831,
      "learning_rate": 3.701895759962397e-06,
      "loss": 0.002,
      "step": 213
    },
    {
      "epoch": 2.9357798165137616,
      "grad_norm": 0.11163649708032608,
      "learning_rate": 3.6810989211991777e-06,
      "loss": 0.0015,
      "step": 214
    },
    {
      "epoch": 2.9495412844036695,
      "grad_norm": 0.6103344559669495,
      "learning_rate": 3.6601963260477923e-06,
      "loss": 0.0051,
      "step": 215
    },
    {
      "epoch": 2.963302752293578,
      "grad_norm": 0.1300484985113144,
      "learning_rate": 3.6391898461406045e-06,
      "loss": 0.0018,
      "step": 216
    },
    {
      "epoch": 2.977064220183486,
      "grad_norm": 0.11599847674369812,
      "learning_rate": 3.6180813624118898e-06,
      "loss": 0.0021,
      "step": 217
    },
    {
      "epoch": 2.9908256880733948,
      "grad_norm": 0.14168186485767365,
      "learning_rate": 3.5968727649294134e-06,
      "loss": 0.0018,
      "step": 218
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.16039852797985077,
      "learning_rate": 3.575565952725193e-06,
      "loss": 0.0014,
      "step": 219
    },
    {
      "epoch": 3.0137614678899083,
      "grad_norm": 0.08175123482942581,
      "learning_rate": 3.55416283362546e-06,
      "loss": 0.0008,
      "step": 220
    },
    {
      "epoch": 3.0275229357798166,
      "grad_norm": 0.07637064158916473,
      "learning_rate": 3.5326653240798283e-06,
      "loss": 0.0007,
      "step": 221
    },
    {
      "epoch": 3.041284403669725,
      "grad_norm": 0.061755988746881485,
      "learning_rate": 3.5110753489896924e-06,
      "loss": 0.0007,
      "step": 222
    },
    {
      "epoch": 3.055045871559633,
      "grad_norm": 0.05573924630880356,
      "learning_rate": 3.4893948415358803e-06,
      "loss": 0.0008,
      "step": 223
    },
    {
      "epoch": 3.0688073394495414,
      "grad_norm": 0.18670693039894104,
      "learning_rate": 3.4676257430055438e-06,
      "loss": 0.0007,
      "step": 224
    },
    {
      "epoch": 3.0825688073394497,
      "grad_norm": 0.05674147605895996,
      "learning_rate": 3.4457700026183378e-06,
      "loss": 0.0008,
      "step": 225
    },
    {
      "epoch": 3.096330275229358,
      "grad_norm": 0.08101407438516617,
      "learning_rate": 3.4238295773518924e-06,
      "loss": 0.0006,
      "step": 226
    },
    {
      "epoch": 3.1100917431192663,
      "grad_norm": 0.09028138220310211,
      "learning_rate": 3.4018064317665745e-06,
      "loss": 0.0011,
      "step": 227
    },
    {
      "epoch": 3.123853211009174,
      "grad_norm": 0.0955042913556099,
      "learning_rate": 3.3797025378295826e-06,
      "loss": 0.0008,
      "step": 228
    },
    {
      "epoch": 3.1376146788990824,
      "grad_norm": 0.09355667978525162,
      "learning_rate": 3.357519874738382e-06,
      "loss": 0.0011,
      "step": 229
    },
    {
      "epoch": 3.1513761467889907,
      "grad_norm": 0.06623287498950958,
      "learning_rate": 3.3352604287434752e-06,
      "loss": 0.0006,
      "step": 230
    },
    {
      "epoch": 3.165137614678899,
      "grad_norm": 0.07880504429340363,
      "learning_rate": 3.31292619297056e-06,
      "loss": 0.0007,
      "step": 231
    },
    {
      "epoch": 3.1788990825688073,
      "grad_norm": 0.06904889643192291,
      "learning_rate": 3.29051916724206e-06,
      "loss": 0.0005,
      "step": 232
    },
    {
      "epoch": 3.1926605504587156,
      "grad_norm": 0.23911181092262268,
      "learning_rate": 3.2680413578980623e-06,
      "loss": 0.0007,
      "step": 233
    },
    {
      "epoch": 3.206422018348624,
      "grad_norm": 0.15154607594013214,
      "learning_rate": 3.245494777616664e-06,
      "loss": 0.0015,
      "step": 234
    },
    {
      "epoch": 3.220183486238532,
      "grad_norm": 0.15627366304397583,
      "learning_rate": 3.2228814452337587e-06,
      "loss": 0.001,
      "step": 235
    },
    {
      "epoch": 3.2339449541284404,
      "grad_norm": 0.0780797079205513,
      "learning_rate": 3.2002033855622683e-06,
      "loss": 0.0005,
      "step": 236
    },
    {
      "epoch": 3.2477064220183487,
      "grad_norm": 0.7883831858634949,
      "learning_rate": 3.177462629210838e-06,
      "loss": 0.0005,
      "step": 237
    },
    {
      "epoch": 3.261467889908257,
      "grad_norm": 0.18375582993030548,
      "learning_rate": 3.154661212402017e-06,
      "loss": 0.0006,
      "step": 238
    },
    {
      "epoch": 3.2752293577981653,
      "grad_norm": 0.09809675812721252,
      "learning_rate": 3.131801176789934e-06,
      "loss": 0.001,
      "step": 239
    },
    {
      "epoch": 3.2889908256880735,
      "grad_norm": 0.04473511874675751,
      "learning_rate": 3.1088845692774798e-06,
      "loss": 0.0002,
      "step": 240
    },
    {
      "epoch": 3.302752293577982,
      "grad_norm": 0.07583656907081604,
      "learning_rate": 3.0859134418330373e-06,
      "loss": 0.0007,
      "step": 241
    },
    {
      "epoch": 3.31651376146789,
      "grad_norm": 0.06534383445978165,
      "learning_rate": 3.0628898513067357e-06,
      "loss": 0.0003,
      "step": 242
    },
    {
      "epoch": 3.330275229357798,
      "grad_norm": 0.09651501476764679,
      "learning_rate": 3.0398158592462847e-06,
      "loss": 0.0012,
      "step": 243
    },
    {
      "epoch": 3.3440366972477062,
      "grad_norm": 0.07052983343601227,
      "learning_rate": 3.0166935317123824e-06,
      "loss": 0.0007,
      "step": 244
    },
    {
      "epoch": 3.3577981651376145,
      "grad_norm": 0.09956210106611252,
      "learning_rate": 2.9935249390937184e-06,
      "loss": 0.0009,
      "step": 245
    },
    {
      "epoch": 3.371559633027523,
      "grad_norm": 0.051604535430669785,
      "learning_rate": 2.970312155921585e-06,
      "loss": 0.0005,
      "step": 246
    },
    {
      "epoch": 3.385321100917431,
      "grad_norm": 0.16182328760623932,
      "learning_rate": 2.9470572606841295e-06,
      "loss": 0.0006,
      "step": 247
    },
    {
      "epoch": 3.3990825688073394,
      "grad_norm": 0.07653603702783585,
      "learning_rate": 2.9237623356402423e-06,
      "loss": 0.0005,
      "step": 248
    },
    {
      "epoch": 3.4128440366972477,
      "grad_norm": 0.0773971900343895,
      "learning_rate": 2.900429466633107e-06,
      "loss": 0.0009,
      "step": 249
    },
    {
      "epoch": 3.426605504587156,
      "grad_norm": 0.1629229336977005,
      "learning_rate": 2.8770607429034352e-06,
      "loss": 0.0013,
      "step": 250
    },
    {
      "epoch": 3.4403669724770642,
      "grad_norm": 0.1734458953142166,
      "learning_rate": 2.8536582569023964e-06,
      "loss": 0.0009,
      "step": 251
    },
    {
      "epoch": 3.4541284403669725,
      "grad_norm": 0.08963964134454727,
      "learning_rate": 2.8302241041042564e-06,
      "loss": 0.0004,
      "step": 252
    },
    {
      "epoch": 3.467889908256881,
      "grad_norm": 0.037656184285879135,
      "learning_rate": 2.8067603828187446e-06,
      "loss": 0.0002,
      "step": 253
    },
    {
      "epoch": 3.481651376146789,
      "grad_norm": 0.07996565848588943,
      "learning_rate": 2.7832691940031755e-06,
      "loss": 0.0006,
      "step": 254
    },
    {
      "epoch": 3.4954128440366974,
      "grad_norm": 0.39782819151878357,
      "learning_rate": 2.759752641074322e-06,
      "loss": 0.0004,
      "step": 255
    },
    {
      "epoch": 3.5091743119266052,
      "grad_norm": 0.056630104780197144,
      "learning_rate": 2.7362128297200784e-06,
      "loss": 0.0004,
      "step": 256
    },
    {
      "epoch": 3.522935779816514,
      "grad_norm": 0.18943996727466583,
      "learning_rate": 2.712651867710914e-06,
      "loss": 0.0017,
      "step": 257
    },
    {
      "epoch": 3.536697247706422,
      "grad_norm": 0.04645173251628876,
      "learning_rate": 2.6890718647111424e-06,
      "loss": 0.0004,
      "step": 258
    },
    {
      "epoch": 3.5504587155963305,
      "grad_norm": 0.07644187659025192,
      "learning_rate": 2.665474932090017e-06,
      "loss": 0.0008,
      "step": 259
    },
    {
      "epoch": 3.5642201834862384,
      "grad_norm": 0.04974009841680527,
      "learning_rate": 2.6418631827326857e-06,
      "loss": 0.0005,
      "step": 260
    },
    {
      "epoch": 3.5779816513761467,
      "grad_norm": 0.06213025003671646,
      "learning_rate": 2.6182387308509927e-06,
      "loss": 0.0005,
      "step": 261
    },
    {
      "epoch": 3.591743119266055,
      "grad_norm": 0.09061244130134583,
      "learning_rate": 2.5946036917941765e-06,
      "loss": 0.0006,
      "step": 262
    },
    {
      "epoch": 3.6055045871559632,
      "grad_norm": 0.0796905905008316,
      "learning_rate": 2.570960181859458e-06,
      "loss": 0.0006,
      "step": 263
    },
    {
      "epoch": 3.6192660550458715,
      "grad_norm": 0.19685125350952148,
      "learning_rate": 2.547310318102548e-06,
      "loss": 0.0006,
      "step": 264
    },
    {
      "epoch": 3.63302752293578,
      "grad_norm": 0.030696067959070206,
      "learning_rate": 2.5236562181480794e-06,
      "loss": 0.0001,
      "step": 265
    },
    {
      "epoch": 3.646788990825688,
      "grad_norm": 0.04516645520925522,
      "learning_rate": 2.5e-06,
      "loss": 0.0004,
      "step": 266
    },
    {
      "epoch": 3.6605504587155964,
      "grad_norm": 0.09353721141815186,
      "learning_rate": 2.4763437818519205e-06,
      "loss": 0.0003,
      "step": 267
    },
    {
      "epoch": 3.6743119266055047,
      "grad_norm": 0.05623761937022209,
      "learning_rate": 2.4526896818974534e-06,
      "loss": 0.0003,
      "step": 268
    },
    {
      "epoch": 3.688073394495413,
      "grad_norm": 0.11353174597024918,
      "learning_rate": 2.429039818140543e-06,
      "loss": 0.0002,
      "step": 269
    },
    {
      "epoch": 3.7018348623853212,
      "grad_norm": 0.03043302521109581,
      "learning_rate": 2.405396308205825e-06,
      "loss": 0.0002,
      "step": 270
    },
    {
      "epoch": 3.7155963302752295,
      "grad_norm": 0.05028878524899483,
      "learning_rate": 2.381761269149009e-06,
      "loss": 0.0002,
      "step": 271
    },
    {
      "epoch": 3.729357798165138,
      "grad_norm": 0.03009975329041481,
      "learning_rate": 2.358136817267315e-06,
      "loss": 0.0002,
      "step": 272
    },
    {
      "epoch": 3.7431192660550456,
      "grad_norm": 0.08547350764274597,
      "learning_rate": 2.334525067909983e-06,
      "loss": 0.0005,
      "step": 273
    },
    {
      "epoch": 3.7568807339449544,
      "grad_norm": 0.03611677512526512,
      "learning_rate": 2.3109281352888593e-06,
      "loss": 0.0002,
      "step": 274
    },
    {
      "epoch": 3.770642201834862,
      "grad_norm": 0.03364509344100952,
      "learning_rate": 2.2873481322890866e-06,
      "loss": 0.0001,
      "step": 275
    },
    {
      "epoch": 3.7844036697247705,
      "grad_norm": 0.04283633828163147,
      "learning_rate": 2.263787170279922e-06,
      "loss": 0.0003,
      "step": 276
    },
    {
      "epoch": 3.7981651376146788,
      "grad_norm": 0.057849906384944916,
      "learning_rate": 2.2402473589256793e-06,
      "loss": 0.0002,
      "step": 277
    },
    {
      "epoch": 3.811926605504587,
      "grad_norm": 0.12497337907552719,
      "learning_rate": 2.2167308059968258e-06,
      "loss": 0.0002,
      "step": 278
    },
    {
      "epoch": 3.8256880733944953,
      "grad_norm": 0.044824711978435516,
      "learning_rate": 2.193239617181256e-06,
      "loss": 0.0003,
      "step": 279
    },
    {
      "epoch": 3.8394495412844036,
      "grad_norm": 0.06536471843719482,
      "learning_rate": 2.169775895895745e-06,
      "loss": 0.0002,
      "step": 280
    },
    {
      "epoch": 3.853211009174312,
      "grad_norm": 0.07215467095375061,
      "learning_rate": 2.146341743097604e-06,
      "loss": 0.0002,
      "step": 281
    },
    {
      "epoch": 3.86697247706422,
      "grad_norm": 0.01871681585907936,
      "learning_rate": 2.1229392570965656e-06,
      "loss": 0.0001,
      "step": 282
    },
    {
      "epoch": 3.8807339449541285,
      "grad_norm": 0.043813955038785934,
      "learning_rate": 2.0995705333668948e-06,
      "loss": 0.0002,
      "step": 283
    },
    {
      "epoch": 3.8944954128440368,
      "grad_norm": 0.046582091599702835,
      "learning_rate": 2.0762376643597586e-06,
      "loss": 0.0003,
      "step": 284
    },
    {
      "epoch": 3.908256880733945,
      "grad_norm": 0.06702767312526703,
      "learning_rate": 2.0529427393158704e-06,
      "loss": 0.0004,
      "step": 285
    },
    {
      "epoch": 3.9220183486238533,
      "grad_norm": 0.053124528378248215,
      "learning_rate": 2.0296878440784164e-06,
      "loss": 0.0005,
      "step": 286
    },
    {
      "epoch": 3.9357798165137616,
      "grad_norm": 0.029215684160590172,
      "learning_rate": 2.006475060906283e-06,
      "loss": 0.0002,
      "step": 287
    },
    {
      "epoch": 3.9495412844036695,
      "grad_norm": 0.03736970201134682,
      "learning_rate": 1.9833064682876175e-06,
      "loss": 0.0001,
      "step": 288
    },
    {
      "epoch": 3.963302752293578,
      "grad_norm": 0.0413820743560791,
      "learning_rate": 1.9601841407537157e-06,
      "loss": 0.0002,
      "step": 289
    },
    {
      "epoch": 3.977064220183486,
      "grad_norm": 0.07081856578588486,
      "learning_rate": 1.937110148693265e-06,
      "loss": 0.0004,
      "step": 290
    },
    {
      "epoch": 3.9908256880733948,
      "grad_norm": 0.0220099538564682,
      "learning_rate": 1.9140865581669627e-06,
      "loss": 0.0001,
      "step": 291
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.0220099538564682,
      "learning_rate": 1.8911154307225204e-06,
      "loss": 0.0001,
      "step": 292
    },
    {
      "epoch": 4.013761467889908,
      "grad_norm": 0.0368737168610096,
      "learning_rate": 1.8681988232100674e-06,
      "loss": 0.0002,
      "step": 293
    },
    {
      "epoch": 4.027522935779817,
      "grad_norm": 0.024728944525122643,
      "learning_rate": 1.8453387875979834e-06,
      "loss": 0.0001,
      "step": 294
    },
    {
      "epoch": 4.041284403669724,
      "grad_norm": 0.027091216295957565,
      "learning_rate": 1.822537370789163e-06,
      "loss": 0.0002,
      "step": 295
    },
    {
      "epoch": 4.055045871559633,
      "grad_norm": 0.05742163583636284,
      "learning_rate": 1.7997966144377328e-06,
      "loss": 0.0001,
      "step": 296
    },
    {
      "epoch": 4.068807339449541,
      "grad_norm": 0.011909076012670994,
      "learning_rate": 1.7771185547662417e-06,
      "loss": 0.0,
      "step": 297
    },
    {
      "epoch": 4.08256880733945,
      "grad_norm": 0.005773312412202358,
      "learning_rate": 1.754505222383337e-06,
      "loss": 0.0,
      "step": 298
    },
    {
      "epoch": 4.0963302752293576,
      "grad_norm": 0.012064033187925816,
      "learning_rate": 1.7319586421019383e-06,
      "loss": 0.0,
      "step": 299
    },
    {
      "epoch": 4.110091743119266,
      "grad_norm": 0.00871270801872015,
      "learning_rate": 1.7094808327579401e-06,
      "loss": 0.0001,
      "step": 300
    },
    {
      "epoch": 4.123853211009174,
      "grad_norm": 0.015500242821872234,
      "learning_rate": 1.6870738070294412e-06,
      "loss": 0.0001,
      "step": 301
    },
    {
      "epoch": 4.137614678899083,
      "grad_norm": 0.010978137142956257,
      "learning_rate": 1.6647395712565256e-06,
      "loss": 0.0,
      "step": 302
    },
    {
      "epoch": 4.151376146788991,
      "grad_norm": 0.011058066971600056,
      "learning_rate": 1.6424801252616186e-06,
      "loss": 0.0001,
      "step": 303
    },
    {
      "epoch": 4.165137614678899,
      "grad_norm": 0.029183728620409966,
      "learning_rate": 1.6202974621704176e-06,
      "loss": 0.0001,
      "step": 304
    },
    {
      "epoch": 4.178899082568807,
      "grad_norm": 0.016758723184466362,
      "learning_rate": 1.5981935682334266e-06,
      "loss": 0.0002,
      "step": 305
    },
    {
      "epoch": 4.192660550458716,
      "grad_norm": 0.02776522748172283,
      "learning_rate": 1.5761704226481078e-06,
      "loss": 0.0001,
      "step": 306
    },
    {
      "epoch": 4.206422018348624,
      "grad_norm": 0.008353229612112045,
      "learning_rate": 1.5542299973816626e-06,
      "loss": 0.0001,
      "step": 307
    },
    {
      "epoch": 4.220183486238533,
      "grad_norm": 0.019200339913368225,
      "learning_rate": 1.5323742569944573e-06,
      "loss": 0.0001,
      "step": 308
    },
    {
      "epoch": 4.23394495412844,
      "grad_norm": 0.023347314447164536,
      "learning_rate": 1.5106051584641208e-06,
      "loss": 0.0001,
      "step": 309
    },
    {
      "epoch": 4.247706422018348,
      "grad_norm": 0.035208187997341156,
      "learning_rate": 1.4889246510103078e-06,
      "loss": 0.0001,
      "step": 310
    },
    {
      "epoch": 4.261467889908257,
      "grad_norm": 0.0710497498512268,
      "learning_rate": 1.4673346759201728e-06,
      "loss": 0.0,
      "step": 311
    },
    {
      "epoch": 4.275229357798165,
      "grad_norm": 0.018748018890619278,
      "learning_rate": 1.44583716637454e-06,
      "loss": 0.0001,
      "step": 312
    },
    {
      "epoch": 4.2889908256880735,
      "grad_norm": 0.054301682859659195,
      "learning_rate": 1.4244340472748076e-06,
      "loss": 0.0001,
      "step": 313
    },
    {
      "epoch": 4.302752293577981,
      "grad_norm": 0.020265033468604088,
      "learning_rate": 1.403127235070587e-06,
      "loss": 0.0001,
      "step": 314
    },
    {
      "epoch": 4.31651376146789,
      "grad_norm": 0.008297888562083244,
      "learning_rate": 1.381918637588112e-06,
      "loss": 0.0,
      "step": 315
    },
    {
      "epoch": 4.330275229357798,
      "grad_norm": 0.018647175282239914,
      "learning_rate": 1.3608101538593965e-06,
      "loss": 0.0001,
      "step": 316
    },
    {
      "epoch": 4.344036697247707,
      "grad_norm": 0.02466970682144165,
      "learning_rate": 1.3398036739522088e-06,
      "loss": 0.0002,
      "step": 317
    },
    {
      "epoch": 4.3577981651376145,
      "grad_norm": 0.04142339527606964,
      "learning_rate": 1.3189010788008234e-06,
      "loss": 0.0001,
      "step": 318
    },
    {
      "epoch": 4.371559633027523,
      "grad_norm": 0.025369996204972267,
      "learning_rate": 1.2981042400376032e-06,
      "loss": 0.0001,
      "step": 319
    },
    {
      "epoch": 4.385321100917431,
      "grad_norm": 0.009671038947999477,
      "learning_rate": 1.277415019825417e-06,
      "loss": 0.0,
      "step": 320
    },
    {
      "epoch": 4.39908256880734,
      "grad_norm": 0.012685295194387436,
      "learning_rate": 1.2568352706908937e-06,
      "loss": 0.0001,
      "step": 321
    },
    {
      "epoch": 4.412844036697248,
      "grad_norm": 0.05089607089757919,
      "learning_rate": 1.2363668353585486e-06,
      "loss": 0.0001,
      "step": 322
    },
    {
      "epoch": 4.426605504587156,
      "grad_norm": 0.024581842124462128,
      "learning_rate": 1.216011546585785e-06,
      "loss": 0.0001,
      "step": 323
    },
    {
      "epoch": 4.440366972477064,
      "grad_norm": 0.012231200002133846,
      "learning_rate": 1.195771226998789e-06,
      "loss": 0.0001,
      "step": 324
    },
    {
      "epoch": 4.454128440366972,
      "grad_norm": 0.02831755019724369,
      "learning_rate": 1.1756476889293269e-06,
      "loss": 0.0001,
      "step": 325
    },
    {
      "epoch": 4.467889908256881,
      "grad_norm": 0.05837830901145935,
      "learning_rate": 1.1556427342524698e-06,
      "loss": 0.0001,
      "step": 326
    },
    {
      "epoch": 4.481651376146789,
      "grad_norm": 0.015133843757212162,
      "learning_rate": 1.1357581542252555e-06,
      "loss": 0.0,
      "step": 327
    },
    {
      "epoch": 4.495412844036697,
      "grad_norm": 0.004880247637629509,
      "learning_rate": 1.1159957293262888e-06,
      "loss": 0.0,
      "step": 328
    },
    {
      "epoch": 4.509174311926605,
      "grad_norm": 0.017852261662483215,
      "learning_rate": 1.0963572290963298e-06,
      "loss": 0.0001,
      "step": 329
    },
    {
      "epoch": 4.522935779816514,
      "grad_norm": 0.007527775596827269,
      "learning_rate": 1.0768444119798357e-06,
      "loss": 0.0001,
      "step": 330
    },
    {
      "epoch": 4.536697247706422,
      "grad_norm": 0.00978136993944645,
      "learning_rate": 1.0574590251675145e-06,
      "loss": 0.0,
      "step": 331
    },
    {
      "epoch": 4.5504587155963305,
      "grad_norm": 0.011405237950384617,
      "learning_rate": 1.0382028044398823e-06,
      "loss": 0.0,
      "step": 332
    },
    {
      "epoch": 4.564220183486238,
      "grad_norm": 0.004018036648631096,
      "learning_rate": 1.0190774740118343e-06,
      "loss": 0.0,
      "step": 333
    },
    {
      "epoch": 4.577981651376147,
      "grad_norm": 0.0168945100158453,
      "learning_rate": 1.0000847463782615e-06,
      "loss": 0.0001,
      "step": 334
    },
    {
      "epoch": 4.591743119266055,
      "grad_norm": 0.008481111377477646,
      "learning_rate": 9.812263221607114e-07,
      "loss": 0.0,
      "step": 335
    },
    {
      "epoch": 4.605504587155964,
      "grad_norm": 0.014810923486948013,
      "learning_rate": 9.625038899551162e-07,
      "loss": 0.0,
      "step": 336
    },
    {
      "epoch": 4.6192660550458715,
      "grad_norm": 0.03142401948571205,
      "learning_rate": 9.439191261805894e-07,
      "loss": 0.0001,
      "step": 337
    },
    {
      "epoch": 4.63302752293578,
      "grad_norm": 0.025308527052402496,
      "learning_rate": 9.254736949293216e-07,
      "loss": 0.0,
      "step": 338
    },
    {
      "epoch": 4.646788990825688,
      "grad_norm": 0.021518364548683167,
      "learning_rate": 9.07169247817579e-07,
      "loss": 0.0001,
      "step": 339
    },
    {
      "epoch": 4.660550458715596,
      "grad_norm": 0.007280074991285801,
      "learning_rate": 8.890074238378074e-07,
      "loss": 0.0,
      "step": 340
    },
    {
      "epoch": 4.674311926605505,
      "grad_norm": 0.007103382609784603,
      "learning_rate": 8.709898492118885e-07,
      "loss": 0.0,
      "step": 341
    },
    {
      "epoch": 4.6880733944954125,
      "grad_norm": 0.0127399992197752,
      "learning_rate": 8.531181372455161e-07,
      "loss": 0.0001,
      "step": 342
    },
    {
      "epoch": 4.701834862385321,
      "grad_norm": 0.01236600149422884,
      "learning_rate": 8.353938881837445e-07,
      "loss": 0.0001,
      "step": 343
    },
    {
      "epoch": 4.715596330275229,
      "grad_norm": 0.01739378273487091,
      "learning_rate": 8.178186890677029e-07,
      "loss": 0.0001,
      "step": 344
    },
    {
      "epoch": 4.729357798165138,
      "grad_norm": 0.02507130056619644,
      "learning_rate": 8.003941135924859e-07,
      "loss": 0.0001,
      "step": 345
    },
    {
      "epoch": 4.743119266055046,
      "grad_norm": 0.04195310175418854,
      "learning_rate": 7.83121721966248e-07,
      "loss": 0.0,
      "step": 346
    },
    {
      "epoch": 4.756880733944954,
      "grad_norm": 0.04957769811153412,
      "learning_rate": 7.66003060770498e-07,
      "loss": 0.0002,
      "step": 347
    },
    {
      "epoch": 4.770642201834862,
      "grad_norm": 0.015677858144044876,
      "learning_rate": 7.490396628216237e-07,
      "loss": 0.0,
      "step": 348
    },
    {
      "epoch": 4.784403669724771,
      "grad_norm": 0.009202621877193451,
      "learning_rate": 7.322330470336314e-07,
      "loss": 0.0001,
      "step": 349
    },
    {
      "epoch": 4.798165137614679,
      "grad_norm": 0.0277025755494833,
      "learning_rate": 7.155847182821524e-07,
      "loss": 0.0001,
      "step": 350
    },
    {
      "epoch": 4.8119266055045875,
      "grad_norm": 0.022610262036323547,
      "learning_rate": 6.990961672696908e-07,
      "loss": 0.0001,
      "step": 351
    },
    {
      "epoch": 4.825688073394495,
      "grad_norm": 0.008024164475500584,
      "learning_rate": 6.827688703921407e-07,
      "loss": 0.0001,
      "step": 352
    },
    {
      "epoch": 4.839449541284404,
      "grad_norm": 0.025318168103694916,
      "learning_rate": 6.666042896065983e-07,
      "loss": 0.0001,
      "step": 353
    },
    {
      "epoch": 4.853211009174312,
      "grad_norm": 0.022787343710660934,
      "learning_rate": 6.506038723004484e-07,
      "loss": 0.0001,
      "step": 354
    },
    {
      "epoch": 4.86697247706422,
      "grad_norm": 0.04321616515517235,
      "learning_rate": 6.347690511617693e-07,
      "loss": 0.0002,
      "step": 355
    },
    {
      "epoch": 4.8807339449541285,
      "grad_norm": 0.03710443153977394,
      "learning_rate": 6.191012440510469e-07,
      "loss": 0.0,
      "step": 356
    },
    {
      "epoch": 4.894495412844036,
      "grad_norm": 0.012855797074735165,
      "learning_rate": 6.036018538742208e-07,
      "loss": 0.0,
      "step": 357
    },
    {
      "epoch": 4.908256880733945,
      "grad_norm": 0.013949613086879253,
      "learning_rate": 5.882722684570638e-07,
      "loss": 0.0001,
      "step": 358
    },
    {
      "epoch": 4.922018348623853,
      "grad_norm": 0.03625642880797386,
      "learning_rate": 5.731138604209169e-07,
      "loss": 0.0,
      "step": 359
    },
    {
      "epoch": 4.935779816513762,
      "grad_norm": 0.006596633233129978,
      "learning_rate": 5.581279870597866e-07,
      "loss": 0.0,
      "step": 360
    }
  ],
  "logging_steps": 1,
  "max_steps": 432,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 6,
  "save_steps": 72,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 2.3061175005735813e+19,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}