{
  "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": 6.082164551143029,
      "learning_rate": 8.333333333333333e-07,
      "loss": 0.9743,
      "step": 1
    },
    {
      "epoch": 0.06060606060606061,
      "grad_norm": 6.199864643639447,
      "learning_rate": 1.6666666666666667e-06,
      "loss": 0.9824,
      "step": 2
    },
    {
      "epoch": 0.09090909090909091,
      "grad_norm": 5.933068581509626,
      "learning_rate": 2.5e-06,
      "loss": 0.9484,
      "step": 3
    },
    {
      "epoch": 0.12121212121212122,
      "grad_norm": 5.48489512774313,
      "learning_rate": 3.3333333333333333e-06,
      "loss": 0.9593,
      "step": 4
    },
    {
      "epoch": 0.15151515151515152,
      "grad_norm": 4.733804820154826,
      "learning_rate": 4.166666666666667e-06,
      "loss": 0.9806,
      "step": 5
    },
    {
      "epoch": 0.18181818181818182,
      "grad_norm": 2.453251004082805,
      "learning_rate": 5e-06,
      "loss": 0.8577,
      "step": 6
    },
    {
      "epoch": 0.21212121212121213,
      "grad_norm": 2.194692816987401,
      "learning_rate": 5.833333333333334e-06,
      "loss": 0.9069,
      "step": 7
    },
    {
      "epoch": 0.24242424242424243,
      "grad_norm": 4.169693041546917,
      "learning_rate": 6.666666666666667e-06,
      "loss": 0.8973,
      "step": 8
    },
    {
      "epoch": 0.2727272727272727,
      "grad_norm": 4.339203281212194,
      "learning_rate": 7.500000000000001e-06,
      "loss": 0.8624,
      "step": 9
    },
    {
      "epoch": 0.30303030303030304,
      "grad_norm": 4.492627160157046,
      "learning_rate": 8.333333333333334e-06,
      "loss": 0.8693,
      "step": 10
    },
    {
      "epoch": 0.3333333333333333,
      "grad_norm": 3.775232800961184,
      "learning_rate": 9.166666666666666e-06,
      "loss": 0.8402,
      "step": 11
    },
    {
      "epoch": 0.36363636363636365,
      "grad_norm": 3.7624865252490447,
      "learning_rate": 1e-05,
      "loss": 0.8399,
      "step": 12
    },
    {
      "epoch": 0.3939393939393939,
      "grad_norm": 3.030257151028533,
      "learning_rate": 1.0833333333333334e-05,
      "loss": 0.8324,
      "step": 13
    },
    {
      "epoch": 0.42424242424242425,
      "grad_norm": 2.1045530995893955,
      "learning_rate": 1.1666666666666668e-05,
      "loss": 0.7438,
      "step": 14
    },
    {
      "epoch": 0.45454545454545453,
      "grad_norm": 2.529529609047169,
      "learning_rate": 1.25e-05,
      "loss": 0.7935,
      "step": 15
    },
    {
      "epoch": 0.48484848484848486,
      "grad_norm": 2.3960493789409814,
      "learning_rate": 1.3333333333333333e-05,
      "loss": 0.7764,
      "step": 16
    },
    {
      "epoch": 0.5151515151515151,
      "grad_norm": 2.3405459822246595,
      "learning_rate": 1.416666666666667e-05,
      "loss": 0.7612,
      "step": 17
    },
    {
      "epoch": 0.5454545454545454,
      "grad_norm": 1.780242375804455,
      "learning_rate": 1.5000000000000002e-05,
      "loss": 0.7319,
      "step": 18
    },
    {
      "epoch": 0.5757575757575758,
      "grad_norm": 1.8376558126663334,
      "learning_rate": 1.5833333333333333e-05,
      "loss": 0.7525,
      "step": 19
    },
    {
      "epoch": 0.6060606060606061,
      "grad_norm": 1.7841155986438049,
      "learning_rate": 1.6666666666666667e-05,
      "loss": 0.7351,
      "step": 20
    },
    {
      "epoch": 0.6363636363636364,
      "grad_norm": 1.4363844990323147,
      "learning_rate": 1.7500000000000002e-05,
      "loss": 0.7023,
      "step": 21
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 1.157085510392974,
      "learning_rate": 1.8333333333333333e-05,
      "loss": 0.7288,
      "step": 22
    },
    {
      "epoch": 0.696969696969697,
      "grad_norm": 1.2534692241749141,
      "learning_rate": 1.916666666666667e-05,
      "loss": 0.6977,
      "step": 23
    },
    {
      "epoch": 0.7272727272727273,
      "grad_norm": 1.2306556966026905,
      "learning_rate": 2e-05,
      "loss": 0.6841,
      "step": 24
    },
    {
      "epoch": 0.7575757575757576,
      "grad_norm": 0.8716110203881019,
      "learning_rate": 1.999884834944106e-05,
      "loss": 0.708,
      "step": 25
    },
    {
      "epoch": 0.7878787878787878,
      "grad_norm": 1.0877065969299742,
      "learning_rate": 1.9995393663024054e-05,
      "loss": 0.6761,
      "step": 26
    },
    {
      "epoch": 0.8181818181818182,
      "grad_norm": 0.9671933925061544,
      "learning_rate": 1.9989636736467278e-05,
      "loss": 0.6682,
      "step": 27
    },
    {
      "epoch": 0.8484848484848485,
      "grad_norm": 0.9251271064814989,
      "learning_rate": 1.9981578895764272e-05,
      "loss": 0.6939,
      "step": 28
    },
    {
      "epoch": 0.8787878787878788,
      "grad_norm": 0.9021303013268445,
      "learning_rate": 1.9971221996878395e-05,
      "loss": 0.6689,
      "step": 29
    },
    {
      "epoch": 0.9090909090909091,
      "grad_norm": 0.8563239151908998,
      "learning_rate": 1.9958568425315316e-05,
      "loss": 0.6889,
      "step": 30
    },
    {
      "epoch": 0.9393939393939394,
      "grad_norm": 0.8264877357726544,
      "learning_rate": 1.9943621095573588e-05,
      "loss": 0.6756,
      "step": 31
    },
    {
      "epoch": 0.9696969696969697,
      "grad_norm": 0.7549070294670714,
      "learning_rate": 1.9926383450473344e-05,
      "loss": 0.6801,
      "step": 32
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.7528386178645887,
      "learning_rate": 1.9906859460363307e-05,
      "loss": 0.6359,
      "step": 33
    },
    {
      "epoch": 1.0303030303030303,
      "grad_norm": 0.9479487314552061,
      "learning_rate": 1.9885053622206305e-05,
      "loss": 0.664,
      "step": 34
    },
    {
      "epoch": 1.0606060606060606,
      "grad_norm": 0.5579551117644118,
      "learning_rate": 1.986097095854347e-05,
      "loss": 0.5904,
      "step": 35
    },
    {
      "epoch": 1.0909090909090908,
      "grad_norm": 0.8784414158397916,
      "learning_rate": 1.9834617016337424e-05,
      "loss": 0.6009,
      "step": 36
    },
    {
      "epoch": 1.121212121212121,
      "grad_norm": 0.6911146133507795,
      "learning_rate": 1.9805997865694616e-05,
      "loss": 0.5919,
      "step": 37
    },
    {
      "epoch": 1.1515151515151516,
      "grad_norm": 0.6893506149596769,
      "learning_rate": 1.9775120098467212e-05,
      "loss": 0.5933,
      "step": 38
    },
    {
      "epoch": 1.1818181818181819,
      "grad_norm": 0.7255162316422783,
      "learning_rate": 1.9741990826734793e-05,
      "loss": 0.6017,
      "step": 39
    },
    {
      "epoch": 1.2121212121212122,
      "grad_norm": 0.8471072586967952,
      "learning_rate": 1.970661768116622e-05,
      "loss": 0.5851,
      "step": 40
    },
    {
      "epoch": 1.2424242424242424,
      "grad_norm": 0.6529361337239672,
      "learning_rate": 1.9669008809262064e-05,
      "loss": 0.5649,
      "step": 41
    },
    {
      "epoch": 1.2727272727272727,
      "grad_norm": 0.6838430467037275,
      "learning_rate": 1.9629172873477995e-05,
      "loss": 0.5826,
      "step": 42
    },
    {
      "epoch": 1.303030303030303,
      "grad_norm": 0.6558124450536883,
      "learning_rate": 1.9587119049229558e-05,
      "loss": 0.5733,
      "step": 43
    },
    {
      "epoch": 1.3333333333333333,
      "grad_norm": 0.7075844976370234,
      "learning_rate": 1.954285702277879e-05,
      "loss": 0.6049,
      "step": 44
    },
    {
      "epoch": 1.3636363636363638,
      "grad_norm": 0.6741474451437296,
      "learning_rate": 1.9496396989003195e-05,
      "loss": 0.5601,
      "step": 45
    },
    {
      "epoch": 1.393939393939394,
      "grad_norm": 0.7554795068670029,
      "learning_rate": 1.944774964904754e-05,
      "loss": 0.6024,
      "step": 46
    },
    {
      "epoch": 1.4242424242424243,
      "grad_norm": 0.830944144168637,
      "learning_rate": 1.9396926207859085e-05,
      "loss": 0.5467,
      "step": 47
    },
    {
      "epoch": 1.4545454545454546,
      "grad_norm": 0.7147180115547055,
      "learning_rate": 1.9343938371606714e-05,
      "loss": 0.5808,
      "step": 48
    },
    {
      "epoch": 1.4848484848484849,
      "grad_norm": 0.6752142559980884,
      "learning_rate": 1.9288798344984673e-05,
      "loss": 0.5406,
      "step": 49
    },
    {
      "epoch": 1.5151515151515151,
      "grad_norm": 0.806800173109227,
      "learning_rate": 1.9231518828401458e-05,
      "loss": 0.5592,
      "step": 50
    },
    {
      "epoch": 1.5454545454545454,
      "grad_norm": 0.6786507591993622,
      "learning_rate": 1.917211301505453e-05,
      "loss": 0.6109,
      "step": 51
    },
    {
      "epoch": 1.5757575757575757,
      "grad_norm": 0.6341107562740883,
      "learning_rate": 1.911059458789152e-05,
      "loss": 0.5306,
      "step": 52
    },
    {
      "epoch": 1.606060606060606,
      "grad_norm": 0.6568745519323503,
      "learning_rate": 1.9046977716458627e-05,
      "loss": 0.5892,
      "step": 53
    },
    {
      "epoch": 1.6363636363636362,
      "grad_norm": 0.6852239165890293,
      "learning_rate": 1.8981277053636963e-05,
      "loss": 0.5784,
      "step": 54
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 0.700488857549085,
      "learning_rate": 1.891350773226754e-05,
      "loss": 0.5433,
      "step": 55
    },
    {
      "epoch": 1.696969696969697,
      "grad_norm": 0.7945441088005711,
      "learning_rate": 1.8843685361665724e-05,
      "loss": 0.5797,
      "step": 56
    },
    {
      "epoch": 1.7272727272727273,
      "grad_norm": 0.7276040007844824,
      "learning_rate": 1.8771826024025944e-05,
      "loss": 0.5882,
      "step": 57
    },
    {
      "epoch": 1.7575757575757576,
      "grad_norm": 0.8596171576630897,
      "learning_rate": 1.8697946270717468e-05,
      "loss": 0.5888,
      "step": 58
    },
    {
      "epoch": 1.7878787878787878,
      "grad_norm": 0.6610138494716133,
      "learning_rate": 1.8622063118472135e-05,
      "loss": 0.5415,
      "step": 59
    },
    {
      "epoch": 1.8181818181818183,
      "grad_norm": 0.8106095732614276,
      "learning_rate": 1.8544194045464888e-05,
      "loss": 0.5745,
      "step": 60
    },
    {
      "epoch": 1.8484848484848486,
      "grad_norm": 0.6102717823967825,
      "learning_rate": 1.8464356987288012e-05,
      "loss": 0.5236,
      "step": 61
    },
    {
      "epoch": 1.878787878787879,
      "grad_norm": 0.794138487374143,
      "learning_rate": 1.8382570332820045e-05,
      "loss": 0.5252,
      "step": 62
    },
    {
      "epoch": 1.9090909090909092,
      "grad_norm": 0.5854426107859853,
      "learning_rate": 1.8298852919990254e-05,
      "loss": 0.5607,
      "step": 63
    },
    {
      "epoch": 1.9393939393939394,
      "grad_norm": 0.7441208793477079,
      "learning_rate": 1.821322403143969e-05,
      "loss": 0.5631,
      "step": 64
    },
    {
      "epoch": 1.9696969696969697,
      "grad_norm": 0.6307361361127832,
      "learning_rate": 1.812570339007983e-05,
      "loss": 0.5809,
      "step": 65
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.6386005422389703,
      "learning_rate": 1.8036311154549783e-05,
      "loss": 0.578,
      "step": 66
    },
    {
      "epoch": 2.0303030303030303,
      "grad_norm": 0.7787329060681482,
      "learning_rate": 1.7945067914573147e-05,
      "loss": 0.451,
      "step": 67
    },
    {
      "epoch": 2.0606060606060606,
      "grad_norm": 0.6567460397548995,
      "learning_rate": 1.7851994686215592e-05,
      "loss": 0.4404,
      "step": 68
    },
    {
      "epoch": 2.090909090909091,
      "grad_norm": 0.8241315933972604,
      "learning_rate": 1.77571129070442e-05,
      "loss": 0.4581,
      "step": 69
    },
    {
      "epoch": 2.121212121212121,
      "grad_norm": 0.7534370019747573,
      "learning_rate": 1.766044443118978e-05,
      "loss": 0.4509,
      "step": 70
    },
    {
      "epoch": 2.1515151515151514,
      "grad_norm": 0.6872313509214591,
      "learning_rate": 1.7562011524313187e-05,
      "loss": 0.4417,
      "step": 71
    },
    {
      "epoch": 2.1818181818181817,
      "grad_norm": 0.8003750790780567,
      "learning_rate": 1.7461836858476858e-05,
      "loss": 0.4315,
      "step": 72
    },
    {
      "epoch": 2.212121212121212,
      "grad_norm": 0.6885114529551459,
      "learning_rate": 1.7359943506922775e-05,
      "loss": 0.4357,
      "step": 73
    },
    {
      "epoch": 2.242424242424242,
      "grad_norm": 0.622922312481805,
      "learning_rate": 1.725635493875799e-05,
      "loss": 0.4155,
      "step": 74
    },
    {
      "epoch": 2.2727272727272725,
      "grad_norm": 0.6659684133395284,
      "learning_rate": 1.7151095013548996e-05,
      "loss": 0.4294,
      "step": 75
    },
    {
      "epoch": 2.303030303030303,
      "grad_norm": 0.6355696728345942,
      "learning_rate": 1.7044187975826126e-05,
      "loss": 0.4133,
      "step": 76
    },
    {
      "epoch": 2.3333333333333335,
      "grad_norm": 0.6567767598814616,
      "learning_rate": 1.693565844949933e-05,
      "loss": 0.4255,
      "step": 77
    },
    {
      "epoch": 2.3636363636363638,
      "grad_norm": 0.6933746677111999,
      "learning_rate": 1.6825531432186545e-05,
      "loss": 0.438,
      "step": 78
    },
    {
      "epoch": 2.393939393939394,
      "grad_norm": 0.6056847026397173,
      "learning_rate": 1.671383228945597e-05,
      "loss": 0.398,
      "step": 79
    },
    {
      "epoch": 2.4242424242424243,
      "grad_norm": 0.6597086578872607,
      "learning_rate": 1.6600586748983642e-05,
      "loss": 0.4236,
      "step": 80
    },
    {
      "epoch": 2.4545454545454546,
      "grad_norm": 0.5548480457025179,
      "learning_rate": 1.648582089462756e-05,
      "loss": 0.4369,
      "step": 81
    },
    {
      "epoch": 2.484848484848485,
      "grad_norm": 0.5878138584407185,
      "learning_rate": 1.6369561160419783e-05,
      "loss": 0.4092,
      "step": 82
    },
    {
      "epoch": 2.515151515151515,
      "grad_norm": 0.5865809429691052,
      "learning_rate": 1.625183432447789e-05,
      "loss": 0.4197,
      "step": 83
    },
    {
      "epoch": 2.5454545454545454,
      "grad_norm": 0.6044346276756262,
      "learning_rate": 1.6132667502837164e-05,
      "loss": 0.4242,
      "step": 84
    },
    {
      "epoch": 2.5757575757575757,
      "grad_norm": 0.6368164236990473,
      "learning_rate": 1.6012088143204953e-05,
      "loss": 0.4235,
      "step": 85
    },
    {
      "epoch": 2.606060606060606,
      "grad_norm": 0.5519239137142832,
      "learning_rate": 1.589012401863864e-05,
      "loss": 0.4213,
      "step": 86
    },
    {
      "epoch": 2.6363636363636362,
      "grad_norm": 0.600599957336867,
      "learning_rate": 1.5766803221148676e-05,
      "loss": 0.4057,
      "step": 87
    },
    {
      "epoch": 2.6666666666666665,
      "grad_norm": 0.6039341293280637,
      "learning_rate": 1.5642154155228124e-05,
      "loss": 0.419,
      "step": 88
    },
    {
      "epoch": 2.6969696969696972,
      "grad_norm": 0.5580092804048937,
      "learning_rate": 1.5516205531310272e-05,
      "loss": 0.4093,
      "step": 89
    },
    {
      "epoch": 2.7272727272727275,
      "grad_norm": 0.6116420692274311,
      "learning_rate": 1.538898635915576e-05,
      "loss": 0.4106,
      "step": 90
    },
    {
      "epoch": 2.757575757575758,
      "grad_norm": 0.48991663926802204,
      "learning_rate": 1.526052594117071e-05,
      "loss": 0.4227,
      "step": 91
    },
    {
      "epoch": 2.787878787878788,
      "grad_norm": 0.6560670019285071,
      "learning_rate": 1.513085386565758e-05,
      "loss": 0.4253,
      "step": 92
    },
    {
      "epoch": 2.8181818181818183,
      "grad_norm": 0.5300396641790873,
      "learning_rate": 1.5000000000000002e-05,
      "loss": 0.4025,
      "step": 93
    },
    {
      "epoch": 2.8484848484848486,
      "grad_norm": 0.5434375934212474,
      "learning_rate": 1.4867994483783485e-05,
      "loss": 0.4332,
      "step": 94
    },
    {
      "epoch": 2.878787878787879,
      "grad_norm": 0.5828297559054213,
      "learning_rate": 1.4734867721853341e-05,
      "loss": 0.4296,
      "step": 95
    },
    {
      "epoch": 2.909090909090909,
      "grad_norm": 0.539521181319008,
      "learning_rate": 1.4600650377311523e-05,
      "loss": 0.4095,
      "step": 96
    },
    {
      "epoch": 2.9393939393939394,
      "grad_norm": 0.5408066308418242,
      "learning_rate": 1.4465373364454001e-05,
      "loss": 0.4375,
      "step": 97
    },
    {
      "epoch": 2.9696969696969697,
      "grad_norm": 0.5153917595894786,
      "learning_rate": 1.4329067841650274e-05,
      "loss": 0.4412,
      "step": 98
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.5086632643785411,
      "learning_rate": 1.4191765204166643e-05,
      "loss": 0.4179,
      "step": 99
    },
    {
      "epoch": 3.0303030303030303,
      "grad_norm": 0.8397857310579694,
      "learning_rate": 1.4053497076934948e-05,
      "loss": 0.3316,
      "step": 100
    },
    {
      "epoch": 3.0606060606060606,
      "grad_norm": 0.6169887350063111,
      "learning_rate": 1.3914295307268396e-05,
      "loss": 0.2961,
      "step": 101
    },
    {
      "epoch": 3.090909090909091,
      "grad_norm": 0.7373719556030066,
      "learning_rate": 1.3774191957526144e-05,
      "loss": 0.3223,
      "step": 102
    },
    {
      "epoch": 3.121212121212121,
      "grad_norm": 0.9689178088933306,
      "learning_rate": 1.3633219297728415e-05,
      "loss": 0.3014,
      "step": 103
    },
    {
      "epoch": 3.1515151515151514,
      "grad_norm": 0.65450824052187,
      "learning_rate": 1.3491409798123687e-05,
      "loss": 0.3173,
      "step": 104
    },
    {
      "epoch": 3.1818181818181817,
      "grad_norm": 0.6379328128026068,
      "learning_rate": 1.3348796121709862e-05,
      "loss": 0.3068,
      "step": 105
    },
    {
      "epoch": 3.212121212121212,
      "grad_norm": 0.6941873237026441,
      "learning_rate": 1.3205411116710973e-05,
      "loss": 0.3075,
      "step": 106
    },
    {
      "epoch": 3.242424242424242,
      "grad_norm": 0.6163483028866464,
      "learning_rate": 1.3061287809011243e-05,
      "loss": 0.3061,
      "step": 107
    },
    {
      "epoch": 3.2727272727272725,
      "grad_norm": 0.6393176916923681,
      "learning_rate": 1.291645939454825e-05,
      "loss": 0.3259,
      "step": 108
    },
    {
      "epoch": 3.303030303030303,
      "grad_norm": 0.5507965203981176,
      "learning_rate": 1.277095923166689e-05,
      "loss": 0.305,
      "step": 109
    },
    {
      "epoch": 3.3333333333333335,
      "grad_norm": 0.617267947967507,
      "learning_rate": 1.2624820833435939e-05,
      "loss": 0.2969,
      "step": 110
    },
    {
      "epoch": 3.3636363636363638,
      "grad_norm": 0.6909455689002703,
      "learning_rate": 1.2478077859929e-05,
      "loss": 0.2983,
      "step": 111
    },
    {
      "epoch": 3.393939393939394,
      "grad_norm": 0.5328958418698717,
      "learning_rate": 1.2330764110471567e-05,
      "loss": 0.3005,
      "step": 112
    },
    {
      "epoch": 3.4242424242424243,
      "grad_norm": 0.7520834074743395,
      "learning_rate": 1.2182913515856016e-05,
      "loss": 0.3187,
      "step": 113
    },
    {
      "epoch": 3.4545454545454546,
      "grad_norm": 0.6416314170295463,
      "learning_rate": 1.2034560130526341e-05,
      "loss": 0.296,
      "step": 114
    },
    {
      "epoch": 3.484848484848485,
      "grad_norm": 0.5565966816209027,
      "learning_rate": 1.1885738124734359e-05,
      "loss": 0.3096,
      "step": 115
    },
    {
      "epoch": 3.515151515151515,
      "grad_norm": 0.6505365256952365,
      "learning_rate": 1.1736481776669307e-05,
      "loss": 0.2866,
      "step": 116
    },
    {
      "epoch": 3.5454545454545454,
      "grad_norm": 0.5028288108042958,
      "learning_rate": 1.1586825464562515e-05,
      "loss": 0.2986,
      "step": 117
    },
    {
      "epoch": 3.5757575757575757,
      "grad_norm": 0.5560879755249306,
      "learning_rate": 1.1436803658769082e-05,
      "loss": 0.2984,
      "step": 118
    },
    {
      "epoch": 3.606060606060606,
      "grad_norm": 0.561674570380799,
      "learning_rate": 1.1286450913828313e-05,
      "loss": 0.2898,
      "step": 119
    },
    {
      "epoch": 3.6363636363636362,
      "grad_norm": 0.47347423026956065,
      "learning_rate": 1.113580186050475e-05,
      "loss": 0.2976,
      "step": 120
    },
    {
      "epoch": 3.6666666666666665,
      "grad_norm": 0.5127687105759521,
      "learning_rate": 1.0984891197811686e-05,
      "loss": 0.3019,
      "step": 121
    },
    {
      "epoch": 3.6969696969696972,
      "grad_norm": 0.5236365734999248,
      "learning_rate": 1.0833753685018935e-05,
      "loss": 0.287,
      "step": 122
    },
    {
      "epoch": 3.7272727272727275,
      "grad_norm": 0.4811598328453427,
      "learning_rate": 1.0682424133646712e-05,
      "loss": 0.301,
      "step": 123
    },
    {
      "epoch": 3.757575757575758,
      "grad_norm": 0.4964461036072366,
      "learning_rate": 1.0530937399447496e-05,
      "loss": 0.3142,
      "step": 124
    },
    {
      "epoch": 3.787878787878788,
      "grad_norm": 0.542572482795018,
      "learning_rate": 1.0379328374377715e-05,
      "loss": 0.2854,
      "step": 125
    },
    {
      "epoch": 3.8181818181818183,
      "grad_norm": 0.49529270879602094,
      "learning_rate": 1.0227631978561057e-05,
      "loss": 0.2826,
      "step": 126
    },
    {
      "epoch": 3.8484848484848486,
      "grad_norm": 0.4825419630192013,
      "learning_rate": 1.0075883152245334e-05,
      "loss": 0.2737,
      "step": 127
    },
    {
      "epoch": 3.878787878787879,
      "grad_norm": 0.5369360705017022,
      "learning_rate": 9.92411684775467e-06,
      "loss": 0.3052,
      "step": 128
    },
    {
      "epoch": 3.909090909090909,
      "grad_norm": 0.49405796903698684,
      "learning_rate": 9.772368021438943e-06,
      "loss": 0.3027,
      "step": 129
    },
    {
      "epoch": 3.9393939393939394,
      "grad_norm": 0.48724290865126735,
      "learning_rate": 9.620671625622287e-06,
      "loss": 0.2916,
      "step": 130
    },
    {
      "epoch": 3.9696969696969697,
      "grad_norm": 0.5459544698031724,
      "learning_rate": 9.469062600552509e-06,
      "loss": 0.3132,
      "step": 131
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.5039711265231206,
      "learning_rate": 9.317575866353293e-06,
      "loss": 0.2724,
      "step": 132
    },
    {
      "epoch": 4.03030303030303,
      "grad_norm": 0.7277166782918293,
      "learning_rate": 9.166246314981066e-06,
      "loss": 0.2173,
      "step": 133
    },
    {
      "epoch": 4.0606060606060606,
      "grad_norm": 0.5882472305139107,
      "learning_rate": 9.015108802188314e-06,
      "loss": 0.2279,
      "step": 134
    },
    {
      "epoch": 4.090909090909091,
      "grad_norm": 0.5298145162290456,
      "learning_rate": 8.86419813949525e-06,
      "loss": 0.2097,
      "step": 135
    },
    {
      "epoch": 4.121212121212121,
      "grad_norm": 0.7622376401983684,
      "learning_rate": 8.71354908617169e-06,
      "loss": 0.2114,
      "step": 136
    },
    {
      "epoch": 4.151515151515151,
      "grad_norm": 0.8091106259564409,
      "learning_rate": 8.56319634123092e-06,
      "loss": 0.2211,
      "step": 137
    },
    {
      "epoch": 4.181818181818182,
      "grad_norm": 0.552735676904124,
      "learning_rate": 8.413174535437486e-06,
      "loss": 0.2046,
      "step": 138
    },
    {
      "epoch": 4.212121212121212,
      "grad_norm": 0.5619974706460401,
      "learning_rate": 8.263518223330698e-06,
      "loss": 0.2059,
      "step": 139
    },
    {
      "epoch": 4.242424242424242,
      "grad_norm": 0.6022642886123147,
      "learning_rate": 8.114261875265643e-06,
      "loss": 0.1994,
      "step": 140
    },
    {
      "epoch": 4.2727272727272725,
      "grad_norm": 0.5702540230632025,
      "learning_rate": 7.965439869473664e-06,
      "loss": 0.2242,
      "step": 141
    },
    {
      "epoch": 4.303030303030303,
      "grad_norm": 0.5722229410713374,
      "learning_rate": 7.817086484143987e-06,
      "loss": 0.198,
      "step": 142
    },
    {
      "epoch": 4.333333333333333,
      "grad_norm": 0.5411390812679117,
      "learning_rate": 7.669235889528436e-06,
      "loss": 0.2016,
      "step": 143
    },
    {
      "epoch": 4.363636363636363,
      "grad_norm": 0.5898502091121667,
      "learning_rate": 7.521922140071003e-06,
      "loss": 0.227,
      "step": 144
    },
    {
      "epoch": 4.393939393939394,
      "grad_norm": 0.5419999988005103,
      "learning_rate": 7.375179166564062e-06,
      "loss": 0.2037,
      "step": 145
    },
    {
      "epoch": 4.424242424242424,
      "grad_norm": 0.5255083099684936,
      "learning_rate": 7.2290407683331154e-06,
      "loss": 0.1998,
      "step": 146
    },
    {
      "epoch": 4.454545454545454,
      "grad_norm": 0.4997836198066876,
      "learning_rate": 7.0835406054517505e-06,
      "loss": 0.1984,
      "step": 147
    },
    {
      "epoch": 4.484848484848484,
      "grad_norm": 0.5153790148750838,
      "learning_rate": 6.93871219098876e-06,
      "loss": 0.1884,
      "step": 148
    },
    {
      "epoch": 4.515151515151516,
      "grad_norm": 0.48931878967443426,
      "learning_rate": 6.79458888328903e-06,
      "loss": 0.2031,
      "step": 149
    },
    {
      "epoch": 4.545454545454545,
      "grad_norm": 0.5277305274168856,
      "learning_rate": 6.651203878290139e-06,
      "loss": 0.1959,
      "step": 150
    },
    {
      "epoch": 4.575757575757576,
      "grad_norm": 0.46438939206102664,
      "learning_rate": 6.508590201876317e-06,
      "loss": 0.2005,
      "step": 151
    },
    {
      "epoch": 4.606060606060606,
      "grad_norm": 0.46881918533467243,
      "learning_rate": 6.366780702271589e-06,
      "loss": 0.1928,
      "step": 152
    },
    {
      "epoch": 4.636363636363637,
      "grad_norm": 0.4764736017188973,
      "learning_rate": 6.225808042473857e-06,
      "loss": 0.1977,
      "step": 153
    },
    {
      "epoch": 4.666666666666667,
      "grad_norm": 0.47046144151353025,
      "learning_rate": 6.085704692731609e-06,
      "loss": 0.1988,
      "step": 154
    },
    {
      "epoch": 4.696969696969697,
      "grad_norm": 0.4855529970086327,
      "learning_rate": 5.946502923065054e-06,
      "loss": 0.2107,
      "step": 155
    },
    {
      "epoch": 4.7272727272727275,
      "grad_norm": 0.46000018432335865,
      "learning_rate": 5.8082347958333625e-06,
      "loss": 0.2251,
      "step": 156
    },
    {
      "epoch": 4.757575757575758,
      "grad_norm": 0.46395250190385995,
      "learning_rate": 5.670932158349732e-06,
      "loss": 0.1868,
      "step": 157
    },
    {
      "epoch": 4.787878787878788,
      "grad_norm": 0.5087778457525394,
      "learning_rate": 5.534626635546e-06,
      "loss": 0.1788,
      "step": 158
    },
    {
      "epoch": 4.818181818181818,
      "grad_norm": 0.4795841560481975,
      "learning_rate": 5.399349622688479e-06,
      "loss": 0.1982,
      "step": 159
    },
    {
      "epoch": 4.848484848484849,
      "grad_norm": 0.4370991242352397,
      "learning_rate": 5.2651322781466606e-06,
      "loss": 0.2075,
      "step": 160
    },
    {
      "epoch": 4.878787878787879,
      "grad_norm": 0.4293123093795792,
      "learning_rate": 5.132005516216512e-06,
      "loss": 0.1988,
      "step": 161
    },
    {
      "epoch": 4.909090909090909,
      "grad_norm": 0.5787382975004897,
      "learning_rate": 5.000000000000003e-06,
      "loss": 0.1958,
      "step": 162
    },
    {
      "epoch": 4.9393939393939394,
      "grad_norm": 0.4574130412997006,
      "learning_rate": 4.869146134342426e-06,
      "loss": 0.1899,
      "step": 163
    },
    {
      "epoch": 4.96969696969697,
      "grad_norm": 0.4380726185982083,
      "learning_rate": 4.739474058829288e-06,
      "loss": 0.1931,
      "step": 164
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.4539513661465005,
      "learning_rate": 4.611013640844245e-06,
      "loss": 0.1958,
      "step": 165
    },
    {
      "epoch": 5.03030303030303,
      "grad_norm": 0.5975706051263661,
      "learning_rate": 4.483794468689728e-06,
      "loss": 0.1634,
      "step": 166
    },
    {
      "epoch": 5.0606060606060606,
      "grad_norm": 0.5308345825385193,
      "learning_rate": 4.357845844771881e-06,
      "loss": 0.1484,
      "step": 167
    },
    {
      "epoch": 5.090909090909091,
      "grad_norm": 0.4306605521264556,
      "learning_rate": 4.2331967788513295e-06,
      "loss": 0.1422,
      "step": 168
    },
    {
      "epoch": 5.121212121212121,
      "grad_norm": 0.44799985389855046,
      "learning_rate": 4.109875981361363e-06,
      "loss": 0.1326,
      "step": 169
    },
    {
      "epoch": 5.151515151515151,
      "grad_norm": 0.5396573031305413,
      "learning_rate": 3.987911856795047e-06,
      "loss": 0.1453,
      "step": 170
    },
    {
      "epoch": 5.181818181818182,
      "grad_norm": 0.6258082214119879,
      "learning_rate": 3.867332497162836e-06,
      "loss": 0.1398,
      "step": 171
    },
    {
      "epoch": 5.212121212121212,
      "grad_norm": 0.6066507081440136,
      "learning_rate": 3.748165675522113e-06,
      "loss": 0.1392,
      "step": 172
    },
    {
      "epoch": 5.242424242424242,
      "grad_norm": 0.5232848536220339,
      "learning_rate": 3.630438839580217e-06,
      "loss": 0.1397,
      "step": 173
    },
    {
      "epoch": 5.2727272727272725,
      "grad_norm": 0.4720769063490613,
      "learning_rate": 3.5141791053724405e-06,
      "loss": 0.1414,
      "step": 174
    },
    {
      "epoch": 5.303030303030303,
      "grad_norm": 0.4592538846411448,
      "learning_rate": 3.399413251016359e-06,
      "loss": 0.1363,
      "step": 175
    },
    {
      "epoch": 5.333333333333333,
      "grad_norm": 0.4314115482649703,
      "learning_rate": 3.2861677105440335e-06,
      "loss": 0.1379,
      "step": 176
    },
    {
      "epoch": 5.363636363636363,
      "grad_norm": 0.4183800342884364,
      "learning_rate": 3.174468567813461e-06,
      "loss": 0.1326,
      "step": 177
    },
    {
      "epoch": 5.393939393939394,
      "grad_norm": 0.43272870332998015,
      "learning_rate": 3.0643415505006733e-06,
      "loss": 0.1295,
      "step": 178
    },
    {
      "epoch": 5.424242424242424,
      "grad_norm": 0.4318367293901279,
      "learning_rate": 2.9558120241738786e-06,
      "loss": 0.1283,
      "step": 179
    },
    {
      "epoch": 5.454545454545454,
      "grad_norm": 0.44965374722824286,
      "learning_rate": 2.8489049864510053e-06,
      "loss": 0.1466,
      "step": 180
    },
    {
      "epoch": 5.484848484848484,
      "grad_norm": 0.4069977439485942,
      "learning_rate": 2.7436450612420098e-06,
      "loss": 0.1517,
      "step": 181
    },
    {
      "epoch": 5.515151515151516,
      "grad_norm": 0.39337556446353444,
      "learning_rate": 2.640056493077231e-06,
      "loss": 0.1292,
      "step": 182
    },
    {
      "epoch": 5.545454545454545,
      "grad_norm": 0.4207085912362768,
      "learning_rate": 2.5381631415231455e-06,
      "loss": 0.1299,
      "step": 183
    },
    {
      "epoch": 5.575757575757576,
      "grad_norm": 0.4592367809245842,
      "learning_rate": 2.4379884756868167e-06,
      "loss": 0.136,
      "step": 184
    },
    {
      "epoch": 5.606060606060606,
      "grad_norm": 0.4292783238689395,
      "learning_rate": 2.339555568810221e-06,
      "loss": 0.1297,
      "step": 185
    },
    {
      "epoch": 5.636363636363637,
      "grad_norm": 0.4369252482459367,
      "learning_rate": 2.2428870929558012e-06,
      "loss": 0.1311,
      "step": 186
    },
    {
      "epoch": 5.666666666666667,
      "grad_norm": 0.4298607160404583,
      "learning_rate": 2.1480053137844115e-06,
      "loss": 0.1414,
      "step": 187
    },
    {
      "epoch": 5.696969696969697,
      "grad_norm": 0.44145537137502966,
      "learning_rate": 2.054932085426856e-06,
      "loss": 0.1407,
      "step": 188
    },
    {
      "epoch": 5.7272727272727275,
      "grad_norm": 0.3968739541513693,
      "learning_rate": 1.963688845450218e-06,
      "loss": 0.1422,
      "step": 189
    },
    {
      "epoch": 5.757575757575758,
      "grad_norm": 0.41925298626523944,
      "learning_rate": 1.8742966099201699e-06,
      "loss": 0.136,
      "step": 190
    },
    {
      "epoch": 5.787878787878788,
      "grad_norm": 0.4084034837164202,
      "learning_rate": 1.7867759685603115e-06,
      "loss": 0.1571,
      "step": 191
    },
    {
      "epoch": 5.818181818181818,
      "grad_norm": 0.37618501943941296,
      "learning_rate": 1.7011470800097496e-06,
      "loss": 0.1258,
      "step": 192
    },
    {
      "epoch": 5.848484848484849,
      "grad_norm": 0.39252498544081876,
      "learning_rate": 1.6174296671799571e-06,
      "loss": 0.1233,
      "step": 193
    },
    {
      "epoch": 5.878787878787879,
      "grad_norm": 0.40389203534972906,
      "learning_rate": 1.5356430127119915e-06,
      "loss": 0.1328,
      "step": 194
    },
    {
      "epoch": 5.909090909090909,
      "grad_norm": 0.40549008704684514,
      "learning_rate": 1.4558059545351144e-06,
      "loss": 0.1373,
      "step": 195
    },
    {
      "epoch": 5.9393939393939394,
      "grad_norm": 0.3999848448814429,
      "learning_rate": 1.3779368815278648e-06,
      "loss": 0.1378,
      "step": 196
    },
    {
      "epoch": 5.96969696969697,
      "grad_norm": 0.4161486440331444,
      "learning_rate": 1.302053729282533e-06,
      "loss": 0.1366,
      "step": 197
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.4114995630693783,
      "learning_rate": 1.2281739759740575e-06,
      "loss": 0.1365,
      "step": 198
    },
    {
      "epoch": 6.03030303030303,
      "grad_norm": 0.432622913410847,
      "learning_rate": 1.156314638334277e-06,
      "loss": 0.1209,
      "step": 199
    },
    {
      "epoch": 6.0606060606060606,
      "grad_norm": 0.44266928378283915,
      "learning_rate": 1.086492267732462e-06,
      "loss": 0.1189,
      "step": 200
    },
    {
      "epoch": 6.090909090909091,
      "grad_norm": 0.40020191270495875,
      "learning_rate": 1.01872294636304e-06,
      "loss": 0.108,
      "step": 201
    },
    {
      "epoch": 6.121212121212121,
      "grad_norm": 0.40940444298781203,
      "learning_rate": 9.530222835413739e-07,
      "loss": 0.1067,
      "step": 202
    },
    {
      "epoch": 6.151515151515151,
      "grad_norm": 0.3694176417338478,
      "learning_rate": 8.894054121084839e-07,
      "loss": 0.1154,
      "step": 203
    },
    {
      "epoch": 6.181818181818182,
      "grad_norm": 0.3512238164978757,
      "learning_rate": 8.278869849454718e-07,
      "loss": 0.1106,
      "step": 204
    },
    {
      "epoch": 6.212121212121212,
      "grad_norm": 0.3572165347480285,
      "learning_rate": 7.684811715985429e-07,
      "loss": 0.1059,
      "step": 205
    },
    {
      "epoch": 6.242424242424242,
      "grad_norm": 0.34746945972713617,
      "learning_rate": 7.1120165501533e-07,
      "loss": 0.1035,
      "step": 206
    },
    {
      "epoch": 6.2727272727272725,
      "grad_norm": 0.331349142326655,
      "learning_rate": 6.560616283932897e-07,
      "loss": 0.1095,
      "step": 207
    },
    {
      "epoch": 6.303030303030303,
      "grad_norm": 0.41627031545081433,
      "learning_rate": 6.030737921409169e-07,
      "loss": 0.1,
      "step": 208
    },
    {
      "epoch": 6.333333333333333,
      "grad_norm": 0.3703896686565648,
      "learning_rate": 5.522503509524591e-07,
      "loss": 0.1064,
      "step": 209
    },
    {
      "epoch": 6.363636363636363,
      "grad_norm": 0.39260491910249645,
      "learning_rate": 5.036030109968082e-07,
      "loss": 0.1172,
      "step": 210
    },
    {
      "epoch": 6.393939393939394,
      "grad_norm": 0.37440287555335755,
      "learning_rate": 4.5714297722121105e-07,
      "loss": 0.1123,
      "step": 211
    },
    {
      "epoch": 6.424242424242424,
      "grad_norm": 0.4282474761319765,
      "learning_rate": 4.128809507704445e-07,
      "loss": 0.1197,
      "step": 212
    },
    {
      "epoch": 6.454545454545454,
      "grad_norm": 0.3633684114874979,
      "learning_rate": 3.708271265220087e-07,
      "loss": 0.1144,
      "step": 213
    },
    {
      "epoch": 6.484848484848484,
      "grad_norm": 0.38411665635234393,
      "learning_rate": 3.309911907379393e-07,
      "loss": 0.1255,
      "step": 214
    },
    {
      "epoch": 6.515151515151516,
      "grad_norm": 0.374091633669975,
      "learning_rate": 2.9338231883378365e-07,
      "loss": 0.1204,
      "step": 215
    },
    {
      "epoch": 6.545454545454545,
      "grad_norm": 0.37109340802594226,
      "learning_rate": 2.5800917326521013e-07,
      "loss": 0.1042,
      "step": 216
    },
    {
      "epoch": 6.575757575757576,
      "grad_norm": 0.35478388580467535,
      "learning_rate": 2.248799015327907e-07,
      "loss": 0.1104,
      "step": 217
    },
    {
      "epoch": 6.606060606060606,
      "grad_norm": 0.3885722436279484,
      "learning_rate": 1.9400213430538773e-07,
      "loss": 0.1165,
      "step": 218
    },
    {
      "epoch": 6.636363636363637,
      "grad_norm": 0.37371970260395854,
      "learning_rate": 1.6538298366257975e-07,
      "loss": 0.1262,
      "step": 219
    },
    {
      "epoch": 6.666666666666667,
      "grad_norm": 0.3903538774663792,
      "learning_rate": 1.3902904145653094e-07,
      "loss": 0.1175,
      "step": 220
    },
    {
      "epoch": 6.696969696969697,
      "grad_norm": 0.3610511715229448,
      "learning_rate": 1.1494637779369766e-07,
      "loss": 0.0973,
      "step": 221
    },
    {
      "epoch": 6.7272727272727275,
      "grad_norm": 0.3628989645199989,
      "learning_rate": 9.314053963669245e-08,
      "loss": 0.1053,
      "step": 222
    },
    {
      "epoch": 6.757575757575758,
      "grad_norm": 0.34183166951991567,
      "learning_rate": 7.361654952665608e-08,
      "loss": 0.1035,
      "step": 223
    },
    {
      "epoch": 6.787878787878788,
      "grad_norm": 0.34550763956865527,
      "learning_rate": 5.637890442641403e-08,
      "loss": 0.1115,
      "step": 224
    },
    {
      "epoch": 6.818181818181818,
      "grad_norm": 0.3560200973058206,
      "learning_rate": 4.143157468468717e-08,
      "loss": 0.1065,
      "step": 225
    },
    {
      "epoch": 6.848484848484849,
      "grad_norm": 0.3700351136531262,
      "learning_rate": 2.8778003121607834e-08,
      "loss": 0.119,
      "step": 226
    },
    {
      "epoch": 6.878787878787879,
      "grad_norm": 0.38171282360266545,
      "learning_rate": 1.8421104235727406e-08,
      "loss": 0.1187,
      "step": 227
    },
    {
      "epoch": 6.909090909090909,
      "grad_norm": 0.3571610695253175,
      "learning_rate": 1.0363263532724433e-08,
      "loss": 0.1111,
      "step": 228
    },
    {
      "epoch": 6.9393939393939394,
      "grad_norm": 0.36033729473264947,
      "learning_rate": 4.606336975948589e-09,
      "loss": 0.103,
      "step": 229
    },
    {
      "epoch": 6.96969696969697,
      "grad_norm": 0.3348421864242949,
      "learning_rate": 1.1516505589381777e-09,
      "loss": 0.1124,
      "step": 230
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.32938291762287747,
      "learning_rate": 0.0,
      "loss": 0.1079,
      "step": 231
    },
    {
      "epoch": 7.0,
      "step": 231,
      "total_flos": 2.7774205071405875e+17,
      "train_loss": 0.363200804404108,
      "train_runtime": 5432.6588,
      "train_samples_per_second": 4.072,
      "train_steps_per_second": 0.043
    }
  ],
  "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": 2.7774205071405875e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}