{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 7.0,
  "eval_steps": 500,
  "global_step": 231,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.030303030303030304,
      "grad_norm": 6.312150870591461,
      "learning_rate": 8.333333333333333e-07,
      "loss": 1.0115,
      "step": 1
    },
    {
      "epoch": 0.06060606060606061,
      "grad_norm": 6.475694856091368,
      "learning_rate": 1.6666666666666667e-06,
      "loss": 1.0286,
      "step": 2
    },
    {
      "epoch": 0.09090909090909091,
      "grad_norm": 6.510110526436135,
      "learning_rate": 2.5e-06,
      "loss": 1.0389,
      "step": 3
    },
    {
      "epoch": 0.12121212121212122,
      "grad_norm": 5.981791395592498,
      "learning_rate": 3.3333333333333333e-06,
      "loss": 1.0259,
      "step": 4
    },
    {
      "epoch": 0.15151515151515152,
      "grad_norm": 4.53026982334595,
      "learning_rate": 4.166666666666667e-06,
      "loss": 0.9918,
      "step": 5
    },
    {
      "epoch": 0.18181818181818182,
      "grad_norm": 2.8104793581280285,
      "learning_rate": 5e-06,
      "loss": 0.9379,
      "step": 6
    },
    {
      "epoch": 0.21212121212121213,
      "grad_norm": 2.567100887895647,
      "learning_rate": 5.833333333333334e-06,
      "loss": 0.9482,
      "step": 7
    },
    {
      "epoch": 0.24242424242424243,
      "grad_norm": 3.892397384715512,
      "learning_rate": 6.666666666666667e-06,
      "loss": 0.9188,
      "step": 8
    },
    {
      "epoch": 0.2727272727272727,
      "grad_norm": 4.142874339222047,
      "learning_rate": 7.500000000000001e-06,
      "loss": 0.9283,
      "step": 9
    },
    {
      "epoch": 0.30303030303030304,
      "grad_norm": 4.138619036300646,
      "learning_rate": 8.333333333333334e-06,
      "loss": 0.9153,
      "step": 10
    },
    {
      "epoch": 0.3333333333333333,
      "grad_norm": 3.373497058955061,
      "learning_rate": 9.166666666666666e-06,
      "loss": 0.8984,
      "step": 11
    },
    {
      "epoch": 0.36363636363636365,
      "grad_norm": 2.8583835585294537,
      "learning_rate": 1e-05,
      "loss": 0.8745,
      "step": 12
    },
    {
      "epoch": 0.3939393939393939,
      "grad_norm": 2.0867771671497706,
      "learning_rate": 1.0833333333333334e-05,
      "loss": 0.8407,
      "step": 13
    },
    {
      "epoch": 0.42424242424242425,
      "grad_norm": 1.4253099021132698,
      "learning_rate": 1.1666666666666668e-05,
      "loss": 0.8072,
      "step": 14
    },
    {
      "epoch": 0.45454545454545453,
      "grad_norm": 1.5431649525602196,
      "learning_rate": 1.25e-05,
      "loss": 0.8059,
      "step": 15
    },
    {
      "epoch": 0.48484848484848486,
      "grad_norm": 1.2817611290592914,
      "learning_rate": 1.3333333333333333e-05,
      "loss": 0.7992,
      "step": 16
    },
    {
      "epoch": 0.5151515151515151,
      "grad_norm": 1.1356040619726908,
      "learning_rate": 1.416666666666667e-05,
      "loss": 0.781,
      "step": 17
    },
    {
      "epoch": 0.5454545454545454,
      "grad_norm": 1.1421250437266164,
      "learning_rate": 1.5000000000000002e-05,
      "loss": 0.7467,
      "step": 18
    },
    {
      "epoch": 0.5757575757575758,
      "grad_norm": 1.0425455417206984,
      "learning_rate": 1.5833333333333333e-05,
      "loss": 0.7772,
      "step": 19
    },
    {
      "epoch": 0.6060606060606061,
      "grad_norm": 0.9574395482369085,
      "learning_rate": 1.6666666666666667e-05,
      "loss": 0.7322,
      "step": 20
    },
    {
      "epoch": 0.6363636363636364,
      "grad_norm": 1.0470442958944162,
      "learning_rate": 1.7500000000000002e-05,
      "loss": 0.7542,
      "step": 21
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 1.0529632063486736,
      "learning_rate": 1.8333333333333333e-05,
      "loss": 0.7272,
      "step": 22
    },
    {
      "epoch": 0.696969696969697,
      "grad_norm": 1.169220818599545,
      "learning_rate": 1.916666666666667e-05,
      "loss": 0.7264,
      "step": 23
    },
    {
      "epoch": 0.7272727272727273,
      "grad_norm": 0.9407356437242442,
      "learning_rate": 2e-05,
      "loss": 0.7198,
      "step": 24
    },
    {
      "epoch": 0.7575757575757576,
      "grad_norm": 1.0131670420617127,
      "learning_rate": 1.999884834944106e-05,
      "loss": 0.7314,
      "step": 25
    },
    {
      "epoch": 0.7878787878787878,
      "grad_norm": 0.9558934157548457,
      "learning_rate": 1.9995393663024054e-05,
      "loss": 0.7092,
      "step": 26
    },
    {
      "epoch": 0.8181818181818182,
      "grad_norm": 0.7862337023546967,
      "learning_rate": 1.9989636736467278e-05,
      "loss": 0.7022,
      "step": 27
    },
    {
      "epoch": 0.8484848484848485,
      "grad_norm": 0.8015449395501483,
      "learning_rate": 1.9981578895764272e-05,
      "loss": 0.7131,
      "step": 28
    },
    {
      "epoch": 0.8787878787878788,
      "grad_norm": 0.8604817636158697,
      "learning_rate": 1.9971221996878395e-05,
      "loss": 0.7071,
      "step": 29
    },
    {
      "epoch": 0.9090909090909091,
      "grad_norm": 0.7854145700441739,
      "learning_rate": 1.9958568425315316e-05,
      "loss": 0.696,
      "step": 30
    },
    {
      "epoch": 0.9393939393939394,
      "grad_norm": 0.6225530941896312,
      "learning_rate": 1.9943621095573588e-05,
      "loss": 0.6941,
      "step": 31
    },
    {
      "epoch": 0.9696969696969697,
      "grad_norm": 0.7778316880329571,
      "learning_rate": 1.9926383450473344e-05,
      "loss": 0.6933,
      "step": 32
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.5277454875573591,
      "learning_rate": 1.9906859460363307e-05,
      "loss": 0.6915,
      "step": 33
    },
    {
      "epoch": 1.0303030303030303,
      "grad_norm": 0.5805460635972389,
      "learning_rate": 1.9885053622206305e-05,
      "loss": 0.6447,
      "step": 34
    },
    {
      "epoch": 1.0606060606060606,
      "grad_norm": 0.6019817869943789,
      "learning_rate": 1.986097095854347e-05,
      "loss": 0.6683,
      "step": 35
    },
    {
      "epoch": 1.0909090909090908,
      "grad_norm": 0.5088210340957223,
      "learning_rate": 1.9834617016337424e-05,
      "loss": 0.6472,
      "step": 36
    },
    {
      "epoch": 1.121212121212121,
      "grad_norm": 0.5962209257067279,
      "learning_rate": 1.9805997865694616e-05,
      "loss": 0.6632,
      "step": 37
    },
    {
      "epoch": 1.1515151515151516,
      "grad_norm": 0.5831832792689964,
      "learning_rate": 1.9775120098467212e-05,
      "loss": 0.6554,
      "step": 38
    },
    {
      "epoch": 1.1818181818181819,
      "grad_norm": 0.5817683452118627,
      "learning_rate": 1.9741990826734793e-05,
      "loss": 0.6336,
      "step": 39
    },
    {
      "epoch": 1.2121212121212122,
      "grad_norm": 0.5669925733078411,
      "learning_rate": 1.970661768116622e-05,
      "loss": 0.6408,
      "step": 40
    },
    {
      "epoch": 1.2424242424242424,
      "grad_norm": 0.6105774979030593,
      "learning_rate": 1.9669008809262064e-05,
      "loss": 0.6303,
      "step": 41
    },
    {
      "epoch": 1.2727272727272727,
      "grad_norm": 0.6142577901965711,
      "learning_rate": 1.9629172873477995e-05,
      "loss": 0.6398,
      "step": 42
    },
    {
      "epoch": 1.303030303030303,
      "grad_norm": 0.5337799657097911,
      "learning_rate": 1.9587119049229558e-05,
      "loss": 0.6286,
      "step": 43
    },
    {
      "epoch": 1.3333333333333333,
      "grad_norm": 0.5572020819014635,
      "learning_rate": 1.954285702277879e-05,
      "loss": 0.6402,
      "step": 44
    },
    {
      "epoch": 1.3636363636363638,
      "grad_norm": 0.4872837857719787,
      "learning_rate": 1.9496396989003195e-05,
      "loss": 0.6241,
      "step": 45
    },
    {
      "epoch": 1.393939393939394,
      "grad_norm": 0.5181010202053239,
      "learning_rate": 1.944774964904754e-05,
      "loss": 0.6293,
      "step": 46
    },
    {
      "epoch": 1.4242424242424243,
      "grad_norm": 0.5967191523625646,
      "learning_rate": 1.9396926207859085e-05,
      "loss": 0.6316,
      "step": 47
    },
    {
      "epoch": 1.4545454545454546,
      "grad_norm": 0.5703718513429342,
      "learning_rate": 1.9343938371606714e-05,
      "loss": 0.6256,
      "step": 48
    },
    {
      "epoch": 1.4848484848484849,
      "grad_norm": 0.6541547573166652,
      "learning_rate": 1.9288798344984673e-05,
      "loss": 0.6123,
      "step": 49
    },
    {
      "epoch": 1.5151515151515151,
      "grad_norm": 0.6609780061398919,
      "learning_rate": 1.9231518828401458e-05,
      "loss": 0.6221,
      "step": 50
    },
    {
      "epoch": 1.5454545454545454,
      "grad_norm": 0.6894966357223148,
      "learning_rate": 1.917211301505453e-05,
      "loss": 0.6248,
      "step": 51
    },
    {
      "epoch": 1.5757575757575757,
      "grad_norm": 0.5939539044353149,
      "learning_rate": 1.911059458789152e-05,
      "loss": 0.6407,
      "step": 52
    },
    {
      "epoch": 1.606060606060606,
      "grad_norm": 0.7038110533728202,
      "learning_rate": 1.9046977716458627e-05,
      "loss": 0.6184,
      "step": 53
    },
    {
      "epoch": 1.6363636363636362,
      "grad_norm": 0.62732059189463,
      "learning_rate": 1.8981277053636963e-05,
      "loss": 0.6129,
      "step": 54
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 0.5427046281169445,
      "learning_rate": 1.891350773226754e-05,
      "loss": 0.6179,
      "step": 55
    },
    {
      "epoch": 1.696969696969697,
      "grad_norm": 0.6194003921539317,
      "learning_rate": 1.8843685361665724e-05,
      "loss": 0.6113,
      "step": 56
    },
    {
      "epoch": 1.7272727272727273,
      "grad_norm": 0.6410974021298932,
      "learning_rate": 1.8771826024025944e-05,
      "loss": 0.6173,
      "step": 57
    },
    {
      "epoch": 1.7575757575757576,
      "grad_norm": 0.5779009695922482,
      "learning_rate": 1.8697946270717468e-05,
      "loss": 0.617,
      "step": 58
    },
    {
      "epoch": 1.7878787878787878,
      "grad_norm": 0.5147582250315594,
      "learning_rate": 1.8622063118472135e-05,
      "loss": 0.6085,
      "step": 59
    },
    {
      "epoch": 1.8181818181818183,
      "grad_norm": 0.5671710300061014,
      "learning_rate": 1.8544194045464888e-05,
      "loss": 0.6203,
      "step": 60
    },
    {
      "epoch": 1.8484848484848486,
      "grad_norm": 0.7941211783306206,
      "learning_rate": 1.8464356987288012e-05,
      "loss": 0.6146,
      "step": 61
    },
    {
      "epoch": 1.878787878787879,
      "grad_norm": 0.7506842382530591,
      "learning_rate": 1.8382570332820045e-05,
      "loss": 0.6063,
      "step": 62
    },
    {
      "epoch": 1.9090909090909092,
      "grad_norm": 0.5258601890372588,
      "learning_rate": 1.8298852919990254e-05,
      "loss": 0.6301,
      "step": 63
    },
    {
      "epoch": 1.9393939393939394,
      "grad_norm": 0.8268877793831794,
      "learning_rate": 1.821322403143969e-05,
      "loss": 0.6001,
      "step": 64
    },
    {
      "epoch": 1.9696969696969697,
      "grad_norm": 0.5581536666536359,
      "learning_rate": 1.812570339007983e-05,
      "loss": 0.6119,
      "step": 65
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.600234570523279,
      "learning_rate": 1.8036311154549783e-05,
      "loss": 0.6099,
      "step": 66
    },
    {
      "epoch": 2.0303030303030303,
      "grad_norm": 0.7265544201056052,
      "learning_rate": 1.7945067914573147e-05,
      "loss": 0.5596,
      "step": 67
    },
    {
      "epoch": 2.0606060606060606,
      "grad_norm": 0.5248402931161873,
      "learning_rate": 1.7851994686215592e-05,
      "loss": 0.5585,
      "step": 68
    },
    {
      "epoch": 2.090909090909091,
      "grad_norm": 0.585676191853421,
      "learning_rate": 1.77571129070442e-05,
      "loss": 0.56,
      "step": 69
    },
    {
      "epoch": 2.121212121212121,
      "grad_norm": 0.7015489638418316,
      "learning_rate": 1.766044443118978e-05,
      "loss": 0.5631,
      "step": 70
    },
    {
      "epoch": 2.1515151515151514,
      "grad_norm": 0.5842445543540288,
      "learning_rate": 1.7562011524313187e-05,
      "loss": 0.5491,
      "step": 71
    },
    {
      "epoch": 2.1818181818181817,
      "grad_norm": 0.51243844443462,
      "learning_rate": 1.7461836858476858e-05,
      "loss": 0.5528,
      "step": 72
    },
    {
      "epoch": 2.212121212121212,
      "grad_norm": 0.7150352543491989,
      "learning_rate": 1.7359943506922775e-05,
      "loss": 0.545,
      "step": 73
    },
    {
      "epoch": 2.242424242424242,
      "grad_norm": 0.5141968817426051,
      "learning_rate": 1.725635493875799e-05,
      "loss": 0.547,
      "step": 74
    },
    {
      "epoch": 2.2727272727272725,
      "grad_norm": 0.585858728549435,
      "learning_rate": 1.7151095013548996e-05,
      "loss": 0.5488,
      "step": 75
    },
    {
      "epoch": 2.303030303030303,
      "grad_norm": 0.7866792966684663,
      "learning_rate": 1.7044187975826126e-05,
      "loss": 0.5519,
      "step": 76
    },
    {
      "epoch": 2.3333333333333335,
      "grad_norm": 0.5061107670338666,
      "learning_rate": 1.693565844949933e-05,
      "loss": 0.5499,
      "step": 77
    },
    {
      "epoch": 2.3636363636363638,
      "grad_norm": 0.6469169079163084,
      "learning_rate": 1.6825531432186545e-05,
      "loss": 0.5317,
      "step": 78
    },
    {
      "epoch": 2.393939393939394,
      "grad_norm": 0.6060546083978395,
      "learning_rate": 1.671383228945597e-05,
      "loss": 0.5524,
      "step": 79
    },
    {
      "epoch": 2.4242424242424243,
      "grad_norm": 0.4696097237633282,
      "learning_rate": 1.6600586748983642e-05,
      "loss": 0.5426,
      "step": 80
    },
    {
      "epoch": 2.4545454545454546,
      "grad_norm": 0.5399242385302314,
      "learning_rate": 1.648582089462756e-05,
      "loss": 0.5494,
      "step": 81
    },
    {
      "epoch": 2.484848484848485,
      "grad_norm": 0.48245852694053143,
      "learning_rate": 1.6369561160419783e-05,
      "loss": 0.5382,
      "step": 82
    },
    {
      "epoch": 2.515151515151515,
      "grad_norm": 0.47001514713374226,
      "learning_rate": 1.625183432447789e-05,
      "loss": 0.5564,
      "step": 83
    },
    {
      "epoch": 2.5454545454545454,
      "grad_norm": 0.5618528608673509,
      "learning_rate": 1.6132667502837164e-05,
      "loss": 0.5504,
      "step": 84
    },
    {
      "epoch": 2.5757575757575757,
      "grad_norm": 0.525585513155995,
      "learning_rate": 1.6012088143204953e-05,
      "loss": 0.5366,
      "step": 85
    },
    {
      "epoch": 2.606060606060606,
      "grad_norm": 0.47441334531625917,
      "learning_rate": 1.589012401863864e-05,
      "loss": 0.5382,
      "step": 86
    },
    {
      "epoch": 2.6363636363636362,
      "grad_norm": 0.6093467259870757,
      "learning_rate": 1.5766803221148676e-05,
      "loss": 0.5491,
      "step": 87
    },
    {
      "epoch": 2.6666666666666665,
      "grad_norm": 0.561336046190135,
      "learning_rate": 1.5642154155228124e-05,
      "loss": 0.5387,
      "step": 88
    },
    {
      "epoch": 2.6969696969696972,
      "grad_norm": 0.49872044831217777,
      "learning_rate": 1.5516205531310272e-05,
      "loss": 0.5374,
      "step": 89
    },
    {
      "epoch": 2.7272727272727275,
      "grad_norm": 0.45268268638401327,
      "learning_rate": 1.538898635915576e-05,
      "loss": 0.539,
      "step": 90
    },
    {
      "epoch": 2.757575757575758,
      "grad_norm": 0.49399347923936493,
      "learning_rate": 1.526052594117071e-05,
      "loss": 0.5483,
      "step": 91
    },
    {
      "epoch": 2.787878787878788,
      "grad_norm": 0.5051110919057764,
      "learning_rate": 1.513085386565758e-05,
      "loss": 0.524,
      "step": 92
    },
    {
      "epoch": 2.8181818181818183,
      "grad_norm": 0.5378366080363427,
      "learning_rate": 1.5000000000000002e-05,
      "loss": 0.5512,
      "step": 93
    },
    {
      "epoch": 2.8484848484848486,
      "grad_norm": 0.46256558272521914,
      "learning_rate": 1.4867994483783485e-05,
      "loss": 0.5448,
      "step": 94
    },
    {
      "epoch": 2.878787878787879,
      "grad_norm": 0.5939458305415402,
      "learning_rate": 1.4734867721853341e-05,
      "loss": 0.541,
      "step": 95
    },
    {
      "epoch": 2.909090909090909,
      "grad_norm": 0.48938531706905364,
      "learning_rate": 1.4600650377311523e-05,
      "loss": 0.5473,
      "step": 96
    },
    {
      "epoch": 2.9393939393939394,
      "grad_norm": 0.5012717641433172,
      "learning_rate": 1.4465373364454001e-05,
      "loss": 0.5615,
      "step": 97
    },
    {
      "epoch": 2.9696969696969697,
      "grad_norm": 0.534292141394524,
      "learning_rate": 1.4329067841650274e-05,
      "loss": 0.535,
      "step": 98
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.534314937899243,
      "learning_rate": 1.4191765204166643e-05,
      "loss": 0.5424,
      "step": 99
    },
    {
      "epoch": 3.0303030303030303,
      "grad_norm": 0.5582906692438339,
      "learning_rate": 1.4053497076934948e-05,
      "loss": 0.4838,
      "step": 100
    },
    {
      "epoch": 3.0606060606060606,
      "grad_norm": 0.6022338728928069,
      "learning_rate": 1.3914295307268396e-05,
      "loss": 0.4822,
      "step": 101
    },
    {
      "epoch": 3.090909090909091,
      "grad_norm": 0.6062975278971561,
      "learning_rate": 1.3774191957526144e-05,
      "loss": 0.4866,
      "step": 102
    },
    {
      "epoch": 3.121212121212121,
      "grad_norm": 0.6965168962135356,
      "learning_rate": 1.3633219297728415e-05,
      "loss": 0.4875,
      "step": 103
    },
    {
      "epoch": 3.1515151515151514,
      "grad_norm": 0.5740971337746343,
      "learning_rate": 1.3491409798123687e-05,
      "loss": 0.4973,
      "step": 104
    },
    {
      "epoch": 3.1818181818181817,
      "grad_norm": 0.6037151795891081,
      "learning_rate": 1.3348796121709862e-05,
      "loss": 0.4838,
      "step": 105
    },
    {
      "epoch": 3.212121212121212,
      "grad_norm": 0.5209861578836725,
      "learning_rate": 1.3205411116710973e-05,
      "loss": 0.4742,
      "step": 106
    },
    {
      "epoch": 3.242424242424242,
      "grad_norm": 0.5456291719112359,
      "learning_rate": 1.3061287809011243e-05,
      "loss": 0.4888,
      "step": 107
    },
    {
      "epoch": 3.2727272727272725,
      "grad_norm": 0.48032197533105836,
      "learning_rate": 1.291645939454825e-05,
      "loss": 0.491,
      "step": 108
    },
    {
      "epoch": 3.303030303030303,
      "grad_norm": 0.4580756881783694,
      "learning_rate": 1.277095923166689e-05,
      "loss": 0.4843,
      "step": 109
    },
    {
      "epoch": 3.3333333333333335,
      "grad_norm": 0.46235698412068515,
      "learning_rate": 1.2624820833435939e-05,
      "loss": 0.4916,
      "step": 110
    },
    {
      "epoch": 3.3636363636363638,
      "grad_norm": 0.4489151537798375,
      "learning_rate": 1.2478077859929e-05,
      "loss": 0.4855,
      "step": 111
    },
    {
      "epoch": 3.393939393939394,
      "grad_norm": 0.4383101762947854,
      "learning_rate": 1.2330764110471567e-05,
      "loss": 0.4858,
      "step": 112
    },
    {
      "epoch": 3.4242424242424243,
      "grad_norm": 0.45451185564173385,
      "learning_rate": 1.2182913515856016e-05,
      "loss": 0.4755,
      "step": 113
    },
    {
      "epoch": 3.4545454545454546,
      "grad_norm": 0.42050469812801256,
      "learning_rate": 1.2034560130526341e-05,
      "loss": 0.4803,
      "step": 114
    },
    {
      "epoch": 3.484848484848485,
      "grad_norm": 0.45831672585375605,
      "learning_rate": 1.1885738124734359e-05,
      "loss": 0.4859,
      "step": 115
    },
    {
      "epoch": 3.515151515151515,
      "grad_norm": 0.4798493173213465,
      "learning_rate": 1.1736481776669307e-05,
      "loss": 0.4758,
      "step": 116
    },
    {
      "epoch": 3.5454545454545454,
      "grad_norm": 0.5047184645263595,
      "learning_rate": 1.1586825464562515e-05,
      "loss": 0.4733,
      "step": 117
    },
    {
      "epoch": 3.5757575757575757,
      "grad_norm": 0.43050751564999873,
      "learning_rate": 1.1436803658769082e-05,
      "loss": 0.4843,
      "step": 118
    },
    {
      "epoch": 3.606060606060606,
      "grad_norm": 0.507474061745988,
      "learning_rate": 1.1286450913828313e-05,
      "loss": 0.4752,
      "step": 119
    },
    {
      "epoch": 3.6363636363636362,
      "grad_norm": 0.5027468266837049,
      "learning_rate": 1.113580186050475e-05,
      "loss": 0.4642,
      "step": 120
    },
    {
      "epoch": 3.6666666666666665,
      "grad_norm": 0.5002534033302283,
      "learning_rate": 1.0984891197811686e-05,
      "loss": 0.475,
      "step": 121
    },
    {
      "epoch": 3.6969696969696972,
      "grad_norm": 0.5099369195818414,
      "learning_rate": 1.0833753685018935e-05,
      "loss": 0.487,
      "step": 122
    },
    {
      "epoch": 3.7272727272727275,
      "grad_norm": 0.48144394820768416,
      "learning_rate": 1.0682424133646712e-05,
      "loss": 0.4728,
      "step": 123
    },
    {
      "epoch": 3.757575757575758,
      "grad_norm": 0.48509841581231067,
      "learning_rate": 1.0530937399447496e-05,
      "loss": 0.4822,
      "step": 124
    },
    {
      "epoch": 3.787878787878788,
      "grad_norm": 0.5232355102786402,
      "learning_rate": 1.0379328374377715e-05,
      "loss": 0.4814,
      "step": 125
    },
    {
      "epoch": 3.8181818181818183,
      "grad_norm": 0.4522597876855734,
      "learning_rate": 1.0227631978561057e-05,
      "loss": 0.4752,
      "step": 126
    },
    {
      "epoch": 3.8484848484848486,
      "grad_norm": 0.48608519066110173,
      "learning_rate": 1.0075883152245334e-05,
      "loss": 0.4589,
      "step": 127
    },
    {
      "epoch": 3.878787878787879,
      "grad_norm": 0.4175574940160927,
      "learning_rate": 9.92411684775467e-06,
      "loss": 0.4797,
      "step": 128
    },
    {
      "epoch": 3.909090909090909,
      "grad_norm": 0.4081235310325298,
      "learning_rate": 9.772368021438943e-06,
      "loss": 0.4855,
      "step": 129
    },
    {
      "epoch": 3.9393939393939394,
      "grad_norm": 0.4672381986062692,
      "learning_rate": 9.620671625622287e-06,
      "loss": 0.4701,
      "step": 130
    },
    {
      "epoch": 3.9696969696969697,
      "grad_norm": 0.4013104550104878,
      "learning_rate": 9.469062600552509e-06,
      "loss": 0.4669,
      "step": 131
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.40311810148872756,
      "learning_rate": 9.317575866353293e-06,
      "loss": 0.4694,
      "step": 132
    },
    {
      "epoch": 4.03030303030303,
      "grad_norm": 0.5253103516174084,
      "learning_rate": 9.166246314981066e-06,
      "loss": 0.4278,
      "step": 133
    },
    {
      "epoch": 4.0606060606060606,
      "grad_norm": 0.4229995957542638,
      "learning_rate": 9.015108802188314e-06,
      "loss": 0.4199,
      "step": 134
    },
    {
      "epoch": 4.090909090909091,
      "grad_norm": 0.4834993887824011,
      "learning_rate": 8.86419813949525e-06,
      "loss": 0.4233,
      "step": 135
    },
    {
      "epoch": 4.121212121212121,
      "grad_norm": 0.5083836778846227,
      "learning_rate": 8.71354908617169e-06,
      "loss": 0.4134,
      "step": 136
    },
    {
      "epoch": 4.151515151515151,
      "grad_norm": 0.4629826983464886,
      "learning_rate": 8.56319634123092e-06,
      "loss": 0.4153,
      "step": 137
    },
    {
      "epoch": 4.181818181818182,
      "grad_norm": 0.4262442369411264,
      "learning_rate": 8.413174535437486e-06,
      "loss": 0.4245,
      "step": 138
    },
    {
      "epoch": 4.212121212121212,
      "grad_norm": 0.48584922213655163,
      "learning_rate": 8.263518223330698e-06,
      "loss": 0.4163,
      "step": 139
    },
    {
      "epoch": 4.242424242424242,
      "grad_norm": 0.43099027346452917,
      "learning_rate": 8.114261875265643e-06,
      "loss": 0.4219,
      "step": 140
    },
    {
      "epoch": 4.2727272727272725,
      "grad_norm": 0.45792891092025656,
      "learning_rate": 7.965439869473664e-06,
      "loss": 0.4252,
      "step": 141
    },
    {
      "epoch": 4.303030303030303,
      "grad_norm": 0.45821297586471343,
      "learning_rate": 7.817086484143987e-06,
      "loss": 0.4275,
      "step": 142
    },
    {
      "epoch": 4.333333333333333,
      "grad_norm": 0.4440069719961142,
      "learning_rate": 7.669235889528436e-06,
      "loss": 0.4293,
      "step": 143
    },
    {
      "epoch": 4.363636363636363,
      "grad_norm": 0.4061400289189544,
      "learning_rate": 7.521922140071003e-06,
      "loss": 0.4331,
      "step": 144
    },
    {
      "epoch": 4.393939393939394,
      "grad_norm": 0.37815756277205015,
      "learning_rate": 7.375179166564062e-06,
      "loss": 0.408,
      "step": 145
    },
    {
      "epoch": 4.424242424242424,
      "grad_norm": 0.4742561168827842,
      "learning_rate": 7.2290407683331154e-06,
      "loss": 0.4241,
      "step": 146
    },
    {
      "epoch": 4.454545454545454,
      "grad_norm": 0.3850909440657431,
      "learning_rate": 7.0835406054517505e-06,
      "loss": 0.4199,
      "step": 147
    },
    {
      "epoch": 4.484848484848484,
      "grad_norm": 0.3850307914357772,
      "learning_rate": 6.93871219098876e-06,
      "loss": 0.414,
      "step": 148
    },
    {
      "epoch": 4.515151515151516,
      "grad_norm": 0.37628212538187894,
      "learning_rate": 6.79458888328903e-06,
      "loss": 0.4142,
      "step": 149
    },
    {
      "epoch": 4.545454545454545,
      "grad_norm": 0.3697164917421783,
      "learning_rate": 6.651203878290139e-06,
      "loss": 0.4199,
      "step": 150
    },
    {
      "epoch": 4.575757575757576,
      "grad_norm": 0.3936732478606088,
      "learning_rate": 6.508590201876317e-06,
      "loss": 0.4392,
      "step": 151
    },
    {
      "epoch": 4.606060606060606,
      "grad_norm": 0.36979958098093274,
      "learning_rate": 6.366780702271589e-06,
      "loss": 0.429,
      "step": 152
    },
    {
      "epoch": 4.636363636363637,
      "grad_norm": 0.4120222499856573,
      "learning_rate": 6.225808042473857e-06,
      "loss": 0.4245,
      "step": 153
    },
    {
      "epoch": 4.666666666666667,
      "grad_norm": 0.37901191700407694,
      "learning_rate": 6.085704692731609e-06,
      "loss": 0.4339,
      "step": 154
    },
    {
      "epoch": 4.696969696969697,
      "grad_norm": 0.34436938180267346,
      "learning_rate": 5.946502923065054e-06,
      "loss": 0.4195,
      "step": 155
    },
    {
      "epoch": 4.7272727272727275,
      "grad_norm": 0.3954985824584074,
      "learning_rate": 5.8082347958333625e-06,
      "loss": 0.4257,
      "step": 156
    },
    {
      "epoch": 4.757575757575758,
      "grad_norm": 0.3847038911028595,
      "learning_rate": 5.670932158349732e-06,
      "loss": 0.4105,
      "step": 157
    },
    {
      "epoch": 4.787878787878788,
      "grad_norm": 0.35182897006844766,
      "learning_rate": 5.534626635546e-06,
      "loss": 0.4161,
      "step": 158
    },
    {
      "epoch": 4.818181818181818,
      "grad_norm": 0.3395768604052704,
      "learning_rate": 5.399349622688479e-06,
      "loss": 0.4274,
      "step": 159
    },
    {
      "epoch": 4.848484848484849,
      "grad_norm": 0.37610765942404767,
      "learning_rate": 5.2651322781466606e-06,
      "loss": 0.4093,
      "step": 160
    },
    {
      "epoch": 4.878787878787879,
      "grad_norm": 0.34369090240001643,
      "learning_rate": 5.132005516216512e-06,
      "loss": 0.4117,
      "step": 161
    },
    {
      "epoch": 4.909090909090909,
      "grad_norm": 0.3193574933254639,
      "learning_rate": 5.000000000000003e-06,
      "loss": 0.4244,
      "step": 162
    },
    {
      "epoch": 4.9393939393939394,
      "grad_norm": 0.3223949271599354,
      "learning_rate": 4.869146134342426e-06,
      "loss": 0.4103,
      "step": 163
    },
    {
      "epoch": 4.96969696969697,
      "grad_norm": 0.3244127700665561,
      "learning_rate": 4.739474058829288e-06,
      "loss": 0.413,
      "step": 164
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.32409787537740004,
      "learning_rate": 4.611013640844245e-06,
      "loss": 0.4165,
      "step": 165
    },
    {
      "epoch": 5.03030303030303,
      "grad_norm": 0.4575500171903933,
      "learning_rate": 4.483794468689728e-06,
      "loss": 0.3721,
      "step": 166
    },
    {
      "epoch": 5.0606060606060606,
      "grad_norm": 0.36883989819661467,
      "learning_rate": 4.357845844771881e-06,
      "loss": 0.3834,
      "step": 167
    },
    {
      "epoch": 5.090909090909091,
      "grad_norm": 0.33445890390189614,
      "learning_rate": 4.2331967788513295e-06,
      "loss": 0.3911,
      "step": 168
    },
    {
      "epoch": 5.121212121212121,
      "grad_norm": 0.41535879066134074,
      "learning_rate": 4.109875981361363e-06,
      "loss": 0.379,
      "step": 169
    },
    {
      "epoch": 5.151515151515151,
      "grad_norm": 0.4967611991379499,
      "learning_rate": 3.987911856795047e-06,
      "loss": 0.3838,
      "step": 170
    },
    {
      "epoch": 5.181818181818182,
      "grad_norm": 0.4117803449507636,
      "learning_rate": 3.867332497162836e-06,
      "loss": 0.3806,
      "step": 171
    },
    {
      "epoch": 5.212121212121212,
      "grad_norm": 0.3805214746037457,
      "learning_rate": 3.748165675522113e-06,
      "loss": 0.3716,
      "step": 172
    },
    {
      "epoch": 5.242424242424242,
      "grad_norm": 0.3538575739009125,
      "learning_rate": 3.630438839580217e-06,
      "loss": 0.3894,
      "step": 173
    },
    {
      "epoch": 5.2727272727272725,
      "grad_norm": 0.4055868044908981,
      "learning_rate": 3.5141791053724405e-06,
      "loss": 0.3814,
      "step": 174
    },
    {
      "epoch": 5.303030303030303,
      "grad_norm": 0.4047775213330113,
      "learning_rate": 3.399413251016359e-06,
      "loss": 0.3742,
      "step": 175
    },
    {
      "epoch": 5.333333333333333,
      "grad_norm": 0.351123670955561,
      "learning_rate": 3.2861677105440335e-06,
      "loss": 0.3725,
      "step": 176
    },
    {
      "epoch": 5.363636363636363,
      "grad_norm": 0.3587210333393083,
      "learning_rate": 3.174468567813461e-06,
      "loss": 0.376,
      "step": 177
    },
    {
      "epoch": 5.393939393939394,
      "grad_norm": 0.38060043226029316,
      "learning_rate": 3.0643415505006733e-06,
      "loss": 0.3776,
      "step": 178
    },
    {
      "epoch": 5.424242424242424,
      "grad_norm": 0.33805978166739914,
      "learning_rate": 2.9558120241738786e-06,
      "loss": 0.3796,
      "step": 179
    },
    {
      "epoch": 5.454545454545454,
      "grad_norm": 0.3221363257994335,
      "learning_rate": 2.8489049864510053e-06,
      "loss": 0.3724,
      "step": 180
    },
    {
      "epoch": 5.484848484848484,
      "grad_norm": 0.3248118221029093,
      "learning_rate": 2.7436450612420098e-06,
      "loss": 0.3768,
      "step": 181
    },
    {
      "epoch": 5.515151515151516,
      "grad_norm": 0.3571193688032367,
      "learning_rate": 2.640056493077231e-06,
      "loss": 0.3718,
      "step": 182
    },
    {
      "epoch": 5.545454545454545,
      "grad_norm": 0.325622710888302,
      "learning_rate": 2.5381631415231455e-06,
      "loss": 0.374,
      "step": 183
    },
    {
      "epoch": 5.575757575757576,
      "grad_norm": 0.30970994937371304,
      "learning_rate": 2.4379884756868167e-06,
      "loss": 0.365,
      "step": 184
    },
    {
      "epoch": 5.606060606060606,
      "grad_norm": 0.2939098855993905,
      "learning_rate": 2.339555568810221e-06,
      "loss": 0.3728,
      "step": 185
    },
    {
      "epoch": 5.636363636363637,
      "grad_norm": 0.3327725237141841,
      "learning_rate": 2.2428870929558012e-06,
      "loss": 0.3696,
      "step": 186
    },
    {
      "epoch": 5.666666666666667,
      "grad_norm": 0.33733530820889834,
      "learning_rate": 2.1480053137844115e-06,
      "loss": 0.3688,
      "step": 187
    },
    {
      "epoch": 5.696969696969697,
      "grad_norm": 0.3267723664611921,
      "learning_rate": 2.054932085426856e-06,
      "loss": 0.3743,
      "step": 188
    },
    {
      "epoch": 5.7272727272727275,
      "grad_norm": 0.30477416794739226,
      "learning_rate": 1.963688845450218e-06,
      "loss": 0.3771,
      "step": 189
    },
    {
      "epoch": 5.757575757575758,
      "grad_norm": 0.31237110797322104,
      "learning_rate": 1.8742966099201699e-06,
      "loss": 0.3713,
      "step": 190
    },
    {
      "epoch": 5.787878787878788,
      "grad_norm": 0.3270216603267311,
      "learning_rate": 1.7867759685603115e-06,
      "loss": 0.3735,
      "step": 191
    },
    {
      "epoch": 5.818181818181818,
      "grad_norm": 0.3086374044328664,
      "learning_rate": 1.7011470800097496e-06,
      "loss": 0.3792,
      "step": 192
    },
    {
      "epoch": 5.848484848484849,
      "grad_norm": 0.3154329043876322,
      "learning_rate": 1.6174296671799571e-06,
      "loss": 0.3746,
      "step": 193
    },
    {
      "epoch": 5.878787878787879,
      "grad_norm": 0.3186084501919431,
      "learning_rate": 1.5356430127119915e-06,
      "loss": 0.386,
      "step": 194
    },
    {
      "epoch": 5.909090909090909,
      "grad_norm": 0.3129459155979679,
      "learning_rate": 1.4558059545351144e-06,
      "loss": 0.3787,
      "step": 195
    },
    {
      "epoch": 5.9393939393939394,
      "grad_norm": 0.2849812085273699,
      "learning_rate": 1.3779368815278648e-06,
      "loss": 0.3801,
      "step": 196
    },
    {
      "epoch": 5.96969696969697,
      "grad_norm": 0.29770863297104755,
      "learning_rate": 1.302053729282533e-06,
      "loss": 0.3907,
      "step": 197
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.29769942695520835,
      "learning_rate": 1.2281739759740575e-06,
      "loss": 0.3755,
      "step": 198
    },
    {
      "epoch": 6.03030303030303,
      "grad_norm": 0.37028292991289424,
      "learning_rate": 1.156314638334277e-06,
      "loss": 0.3658,
      "step": 199
    },
    {
      "epoch": 6.0606060606060606,
      "grad_norm": 0.3640687593367054,
      "learning_rate": 1.086492267732462e-06,
      "loss": 0.3603,
      "step": 200
    },
    {
      "epoch": 6.090909090909091,
      "grad_norm": 0.34956891428811765,
      "learning_rate": 1.01872294636304e-06,
      "loss": 0.3529,
      "step": 201
    },
    {
      "epoch": 6.121212121212121,
      "grad_norm": 0.31208168864820524,
      "learning_rate": 9.530222835413739e-07,
      "loss": 0.359,
      "step": 202
    },
    {
      "epoch": 6.151515151515151,
      "grad_norm": 0.2985599079997785,
      "learning_rate": 8.894054121084839e-07,
      "loss": 0.3641,
      "step": 203
    },
    {
      "epoch": 6.181818181818182,
      "grad_norm": 0.3045681297282551,
      "learning_rate": 8.278869849454718e-07,
      "loss": 0.3479,
      "step": 204
    },
    {
      "epoch": 6.212121212121212,
      "grad_norm": 0.3337677361693306,
      "learning_rate": 7.684811715985429e-07,
      "loss": 0.3661,
      "step": 205
    },
    {
      "epoch": 6.242424242424242,
      "grad_norm": 0.330091275503482,
      "learning_rate": 7.1120165501533e-07,
      "loss": 0.3638,
      "step": 206
    },
    {
      "epoch": 6.2727272727272725,
      "grad_norm": 0.33236530374834883,
      "learning_rate": 6.560616283932897e-07,
      "loss": 0.3585,
      "step": 207
    },
    {
      "epoch": 6.303030303030303,
      "grad_norm": 0.33403382077537863,
      "learning_rate": 6.030737921409169e-07,
      "loss": 0.3523,
      "step": 208
    },
    {
      "epoch": 6.333333333333333,
      "grad_norm": 0.3076896715478002,
      "learning_rate": 5.522503509524591e-07,
      "loss": 0.3562,
      "step": 209
    },
    {
      "epoch": 6.363636363636363,
      "grad_norm": 0.2892632966187589,
      "learning_rate": 5.036030109968082e-07,
      "loss": 0.3558,
      "step": 210
    },
    {
      "epoch": 6.393939393939394,
      "grad_norm": 0.3006521646154084,
      "learning_rate": 4.5714297722121105e-07,
      "loss": 0.3598,
      "step": 211
    },
    {
      "epoch": 6.424242424242424,
      "grad_norm": 0.29904963807062096,
      "learning_rate": 4.128809507704445e-07,
      "loss": 0.3521,
      "step": 212
    },
    {
      "epoch": 6.454545454545454,
      "grad_norm": 0.29793493646756863,
      "learning_rate": 3.708271265220087e-07,
      "loss": 0.349,
      "step": 213
    },
    {
      "epoch": 6.484848484848484,
      "grad_norm": 0.2935354245570347,
      "learning_rate": 3.309911907379393e-07,
      "loss": 0.3649,
      "step": 214
    },
    {
      "epoch": 6.515151515151516,
      "grad_norm": 0.27964737258039646,
      "learning_rate": 2.9338231883378365e-07,
      "loss": 0.3576,
      "step": 215
    },
    {
      "epoch": 6.545454545454545,
      "grad_norm": 0.301582739448306,
      "learning_rate": 2.5800917326521013e-07,
      "loss": 0.3602,
      "step": 216
    },
    {
      "epoch": 6.575757575757576,
      "grad_norm": 0.2936742547927523,
      "learning_rate": 2.248799015327907e-07,
      "loss": 0.3573,
      "step": 217
    },
    {
      "epoch": 6.606060606060606,
      "grad_norm": 0.29954291916250736,
      "learning_rate": 1.9400213430538773e-07,
      "loss": 0.3546,
      "step": 218
    },
    {
      "epoch": 6.636363636363637,
      "grad_norm": 0.27606871733537713,
      "learning_rate": 1.6538298366257975e-07,
      "loss": 0.3578,
      "step": 219
    },
    {
      "epoch": 6.666666666666667,
      "grad_norm": 0.30802643125453266,
      "learning_rate": 1.3902904145653094e-07,
      "loss": 0.3449,
      "step": 220
    },
    {
      "epoch": 6.696969696969697,
      "grad_norm": 0.2962737721620968,
      "learning_rate": 1.1494637779369766e-07,
      "loss": 0.3513,
      "step": 221
    },
    {
      "epoch": 6.7272727272727275,
      "grad_norm": 0.2747413575856551,
      "learning_rate": 9.314053963669245e-08,
      "loss": 0.3638,
      "step": 222
    },
    {
      "epoch": 6.757575757575758,
      "grad_norm": 0.2854609348497825,
      "learning_rate": 7.361654952665608e-08,
      "loss": 0.3538,
      "step": 223
    },
    {
      "epoch": 6.787878787878788,
      "grad_norm": 0.28835820053227823,
      "learning_rate": 5.637890442641403e-08,
      "loss": 0.3525,
      "step": 224
    },
    {
      "epoch": 6.818181818181818,
      "grad_norm": 0.2912380841058537,
      "learning_rate": 4.143157468468717e-08,
      "loss": 0.352,
      "step": 225
    },
    {
      "epoch": 6.848484848484849,
      "grad_norm": 0.2752077854667188,
      "learning_rate": 2.8778003121607834e-08,
      "loss": 0.3573,
      "step": 226
    },
    {
      "epoch": 6.878787878787879,
      "grad_norm": 0.2879200632137298,
      "learning_rate": 1.8421104235727406e-08,
      "loss": 0.3513,
      "step": 227
    },
    {
      "epoch": 6.909090909090909,
      "grad_norm": 0.2906879499875626,
      "learning_rate": 1.0363263532724433e-08,
      "loss": 0.3506,
      "step": 228
    },
    {
      "epoch": 6.9393939393939394,
      "grad_norm": 0.2787965911550042,
      "learning_rate": 4.606336975948589e-09,
      "loss": 0.3495,
      "step": 229
    },
    {
      "epoch": 6.96969696969697,
      "grad_norm": 0.27756524800483817,
      "learning_rate": 1.1516505589381777e-09,
      "loss": 0.3507,
      "step": 230
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.2770002873936819,
      "learning_rate": 0.0,
      "loss": 0.357,
      "step": 231
    },
    {
      "epoch": 7.0,
      "step": 231,
      "total_flos": 6.474382962841354e+17,
      "train_loss": 0.5183511296392003,
      "train_runtime": 8671.5363,
      "train_samples_per_second": 2.551,
      "train_steps_per_second": 0.027
    }
  ],
  "logging_steps": 1,
  "max_steps": 231,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 7,
  "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": 6.474382962841354e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}