{
  "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": 2.869056092598831,
      "learning_rate": 8.333333333333333e-07,
      "loss": 0.7829,
      "step": 1
    },
    {
      "epoch": 0.06060606060606061,
      "grad_norm": 2.943265028617576,
      "learning_rate": 1.6666666666666667e-06,
      "loss": 0.7759,
      "step": 2
    },
    {
      "epoch": 0.09090909090909091,
      "grad_norm": 2.9314876817471993,
      "learning_rate": 2.5e-06,
      "loss": 0.7778,
      "step": 3
    },
    {
      "epoch": 0.12121212121212122,
      "grad_norm": 2.7163807523973773,
      "learning_rate": 3.3333333333333333e-06,
      "loss": 0.7609,
      "step": 4
    },
    {
      "epoch": 0.15151515151515152,
      "grad_norm": 2.358009308252273,
      "learning_rate": 4.166666666666667e-06,
      "loss": 0.7442,
      "step": 5
    },
    {
      "epoch": 0.18181818181818182,
      "grad_norm": 1.609545714575771,
      "learning_rate": 5e-06,
      "loss": 0.6855,
      "step": 6
    },
    {
      "epoch": 0.21212121212121213,
      "grad_norm": 1.4216532708220926,
      "learning_rate": 5.833333333333334e-06,
      "loss": 0.679,
      "step": 7
    },
    {
      "epoch": 0.24242424242424243,
      "grad_norm": 1.2672192053469296,
      "learning_rate": 6.666666666666667e-06,
      "loss": 0.6693,
      "step": 8
    },
    {
      "epoch": 0.2727272727272727,
      "grad_norm": 1.1762564557609267,
      "learning_rate": 7.500000000000001e-06,
      "loss": 0.6507,
      "step": 9
    },
    {
      "epoch": 0.30303030303030304,
      "grad_norm": 1.089506316410281,
      "learning_rate": 8.333333333333334e-06,
      "loss": 0.6151,
      "step": 10
    },
    {
      "epoch": 0.3333333333333333,
      "grad_norm": 1.3423589789327068,
      "learning_rate": 9.166666666666666e-06,
      "loss": 0.6213,
      "step": 11
    },
    {
      "epoch": 0.36363636363636365,
      "grad_norm": 1.1356398695966998,
      "learning_rate": 1e-05,
      "loss": 0.5831,
      "step": 12
    },
    {
      "epoch": 0.3939393939393939,
      "grad_norm": 0.83376554972884,
      "learning_rate": 1.0833333333333334e-05,
      "loss": 0.5794,
      "step": 13
    },
    {
      "epoch": 0.42424242424242425,
      "grad_norm": 0.6917831832778619,
      "learning_rate": 1.1666666666666668e-05,
      "loss": 0.5823,
      "step": 14
    },
    {
      "epoch": 0.45454545454545453,
      "grad_norm": 1.191027416515714,
      "learning_rate": 1.25e-05,
      "loss": 0.5558,
      "step": 15
    },
    {
      "epoch": 0.48484848484848486,
      "grad_norm": 0.9585522875430437,
      "learning_rate": 1.3333333333333333e-05,
      "loss": 0.5353,
      "step": 16
    },
    {
      "epoch": 0.5151515151515151,
      "grad_norm": 0.8036676390354438,
      "learning_rate": 1.416666666666667e-05,
      "loss": 0.5625,
      "step": 17
    },
    {
      "epoch": 0.5454545454545454,
      "grad_norm": 0.9533026357903469,
      "learning_rate": 1.5000000000000002e-05,
      "loss": 0.531,
      "step": 18
    },
    {
      "epoch": 0.5757575757575758,
      "grad_norm": 0.6111870818491476,
      "learning_rate": 1.5833333333333333e-05,
      "loss": 0.5369,
      "step": 19
    },
    {
      "epoch": 0.6060606060606061,
      "grad_norm": 0.5552879651238115,
      "learning_rate": 1.6666666666666667e-05,
      "loss": 0.5406,
      "step": 20
    },
    {
      "epoch": 0.6363636363636364,
      "grad_norm": 0.5419242592910598,
      "learning_rate": 1.7500000000000002e-05,
      "loss": 0.4997,
      "step": 21
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 0.47530911796903585,
      "learning_rate": 1.8333333333333333e-05,
      "loss": 0.4809,
      "step": 22
    },
    {
      "epoch": 0.696969696969697,
      "grad_norm": 0.5237872578941504,
      "learning_rate": 1.916666666666667e-05,
      "loss": 0.5204,
      "step": 23
    },
    {
      "epoch": 0.7272727272727273,
      "grad_norm": 0.591785503782615,
      "learning_rate": 2e-05,
      "loss": 0.5191,
      "step": 24
    },
    {
      "epoch": 0.7575757575757576,
      "grad_norm": 0.4773902480822667,
      "learning_rate": 1.999884834944106e-05,
      "loss": 0.5153,
      "step": 25
    },
    {
      "epoch": 0.7878787878787878,
      "grad_norm": 0.4200444462649622,
      "learning_rate": 1.9995393663024054e-05,
      "loss": 0.5021,
      "step": 26
    },
    {
      "epoch": 0.8181818181818182,
      "grad_norm": 0.46007067913695465,
      "learning_rate": 1.9989636736467278e-05,
      "loss": 0.4993,
      "step": 27
    },
    {
      "epoch": 0.8484848484848485,
      "grad_norm": 0.4170021773849236,
      "learning_rate": 1.9981578895764272e-05,
      "loss": 0.48,
      "step": 28
    },
    {
      "epoch": 0.8787878787878788,
      "grad_norm": 0.42560935899010904,
      "learning_rate": 1.9971221996878395e-05,
      "loss": 0.4813,
      "step": 29
    },
    {
      "epoch": 0.9090909090909091,
      "grad_norm": 0.40164950433779595,
      "learning_rate": 1.9958568425315316e-05,
      "loss": 0.4714,
      "step": 30
    },
    {
      "epoch": 0.9393939393939394,
      "grad_norm": 0.3747886605978014,
      "learning_rate": 1.9943621095573588e-05,
      "loss": 0.4715,
      "step": 31
    },
    {
      "epoch": 0.9696969696969697,
      "grad_norm": 0.3856494875066136,
      "learning_rate": 1.9926383450473344e-05,
      "loss": 0.4713,
      "step": 32
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.368995467886064,
      "learning_rate": 1.9906859460363307e-05,
      "loss": 0.4659,
      "step": 33
    },
    {
      "epoch": 1.0303030303030303,
      "grad_norm": 0.37722427170657086,
      "learning_rate": 1.9885053622206305e-05,
      "loss": 0.4227,
      "step": 34
    },
    {
      "epoch": 1.0606060606060606,
      "grad_norm": 0.3721472558361102,
      "learning_rate": 1.986097095854347e-05,
      "loss": 0.4278,
      "step": 35
    },
    {
      "epoch": 1.0909090909090908,
      "grad_norm": 0.3325354640219533,
      "learning_rate": 1.9834617016337424e-05,
      "loss": 0.4222,
      "step": 36
    },
    {
      "epoch": 1.121212121212121,
      "grad_norm": 0.40356081509767827,
      "learning_rate": 1.9805997865694616e-05,
      "loss": 0.4178,
      "step": 37
    },
    {
      "epoch": 1.1515151515151516,
      "grad_norm": 0.38099392189821907,
      "learning_rate": 1.9775120098467212e-05,
      "loss": 0.4131,
      "step": 38
    },
    {
      "epoch": 1.1818181818181819,
      "grad_norm": 0.3251651997566753,
      "learning_rate": 1.9741990826734793e-05,
      "loss": 0.4142,
      "step": 39
    },
    {
      "epoch": 1.2121212121212122,
      "grad_norm": 0.48085377394054535,
      "learning_rate": 1.970661768116622e-05,
      "loss": 0.4337,
      "step": 40
    },
    {
      "epoch": 1.2424242424242424,
      "grad_norm": 0.41903464152333575,
      "learning_rate": 1.9669008809262064e-05,
      "loss": 0.3817,
      "step": 41
    },
    {
      "epoch": 1.2727272727272727,
      "grad_norm": 0.35304367416613097,
      "learning_rate": 1.9629172873477995e-05,
      "loss": 0.4183,
      "step": 42
    },
    {
      "epoch": 1.303030303030303,
      "grad_norm": 0.28435360032609186,
      "learning_rate": 1.9587119049229558e-05,
      "loss": 0.4041,
      "step": 43
    },
    {
      "epoch": 1.3333333333333333,
      "grad_norm": 0.3776549822245415,
      "learning_rate": 1.954285702277879e-05,
      "loss": 0.4041,
      "step": 44
    },
    {
      "epoch": 1.3636363636363638,
      "grad_norm": 0.413353819728181,
      "learning_rate": 1.9496396989003195e-05,
      "loss": 0.3877,
      "step": 45
    },
    {
      "epoch": 1.393939393939394,
      "grad_norm": 0.3121363676940385,
      "learning_rate": 1.944774964904754e-05,
      "loss": 0.3851,
      "step": 46
    },
    {
      "epoch": 1.4242424242424243,
      "grad_norm": 0.317997890277735,
      "learning_rate": 1.9396926207859085e-05,
      "loss": 0.4006,
      "step": 47
    },
    {
      "epoch": 1.4545454545454546,
      "grad_norm": 0.3956555940066266,
      "learning_rate": 1.9343938371606714e-05,
      "loss": 0.4111,
      "step": 48
    },
    {
      "epoch": 1.4848484848484849,
      "grad_norm": 0.35446211830509466,
      "learning_rate": 1.9288798344984673e-05,
      "loss": 0.3822,
      "step": 49
    },
    {
      "epoch": 1.5151515151515151,
      "grad_norm": 0.3172616925095445,
      "learning_rate": 1.9231518828401458e-05,
      "loss": 0.41,
      "step": 50
    },
    {
      "epoch": 1.5454545454545454,
      "grad_norm": 0.3519084739248107,
      "learning_rate": 1.917211301505453e-05,
      "loss": 0.3903,
      "step": 51
    },
    {
      "epoch": 1.5757575757575757,
      "grad_norm": 0.32651728789160184,
      "learning_rate": 1.911059458789152e-05,
      "loss": 0.3657,
      "step": 52
    },
    {
      "epoch": 1.606060606060606,
      "grad_norm": 0.30723386303702965,
      "learning_rate": 1.9046977716458627e-05,
      "loss": 0.3851,
      "step": 53
    },
    {
      "epoch": 1.6363636363636362,
      "grad_norm": 0.304696465821583,
      "learning_rate": 1.8981277053636963e-05,
      "loss": 0.3857,
      "step": 54
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 0.3128850051690239,
      "learning_rate": 1.891350773226754e-05,
      "loss": 0.4044,
      "step": 55
    },
    {
      "epoch": 1.696969696969697,
      "grad_norm": 0.27095466690336234,
      "learning_rate": 1.8843685361665724e-05,
      "loss": 0.409,
      "step": 56
    },
    {
      "epoch": 1.7272727272727273,
      "grad_norm": 0.3224272109058398,
      "learning_rate": 1.8771826024025944e-05,
      "loss": 0.3745,
      "step": 57
    },
    {
      "epoch": 1.7575757575757576,
      "grad_norm": 0.30435817274603993,
      "learning_rate": 1.8697946270717468e-05,
      "loss": 0.3811,
      "step": 58
    },
    {
      "epoch": 1.7878787878787878,
      "grad_norm": 0.28901957409628054,
      "learning_rate": 1.8622063118472135e-05,
      "loss": 0.3823,
      "step": 59
    },
    {
      "epoch": 1.8181818181818183,
      "grad_norm": 0.34222101122768345,
      "learning_rate": 1.8544194045464888e-05,
      "loss": 0.3736,
      "step": 60
    },
    {
      "epoch": 1.8484848484848486,
      "grad_norm": 0.34444671710274855,
      "learning_rate": 1.8464356987288012e-05,
      "loss": 0.3827,
      "step": 61
    },
    {
      "epoch": 1.878787878787879,
      "grad_norm": 0.28985273895052577,
      "learning_rate": 1.8382570332820045e-05,
      "loss": 0.3755,
      "step": 62
    },
    {
      "epoch": 1.9090909090909092,
      "grad_norm": 0.3034716163634918,
      "learning_rate": 1.8298852919990254e-05,
      "loss": 0.3835,
      "step": 63
    },
    {
      "epoch": 1.9393939393939394,
      "grad_norm": 0.2941618478631507,
      "learning_rate": 1.821322403143969e-05,
      "loss": 0.3576,
      "step": 64
    },
    {
      "epoch": 1.9696969696969697,
      "grad_norm": 0.32449993269579835,
      "learning_rate": 1.812570339007983e-05,
      "loss": 0.3919,
      "step": 65
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.32769097208932363,
      "learning_rate": 1.8036311154549783e-05,
      "loss": 0.3981,
      "step": 66
    },
    {
      "epoch": 2.0303030303030303,
      "grad_norm": 0.42740566665401647,
      "learning_rate": 1.7945067914573147e-05,
      "loss": 0.3279,
      "step": 67
    },
    {
      "epoch": 2.0606060606060606,
      "grad_norm": 0.39704291392713154,
      "learning_rate": 1.7851994686215592e-05,
      "loss": 0.3211,
      "step": 68
    },
    {
      "epoch": 2.090909090909091,
      "grad_norm": 0.35163844900027147,
      "learning_rate": 1.77571129070442e-05,
      "loss": 0.3085,
      "step": 69
    },
    {
      "epoch": 2.121212121212121,
      "grad_norm": 0.3967806510402936,
      "learning_rate": 1.766044443118978e-05,
      "loss": 0.3039,
      "step": 70
    },
    {
      "epoch": 2.1515151515151514,
      "grad_norm": 0.34189498714951994,
      "learning_rate": 1.7562011524313187e-05,
      "loss": 0.3196,
      "step": 71
    },
    {
      "epoch": 2.1818181818181817,
      "grad_norm": 0.3547831687071915,
      "learning_rate": 1.7461836858476858e-05,
      "loss": 0.2919,
      "step": 72
    },
    {
      "epoch": 2.212121212121212,
      "grad_norm": 0.32184329282036506,
      "learning_rate": 1.7359943506922775e-05,
      "loss": 0.3076,
      "step": 73
    },
    {
      "epoch": 2.242424242424242,
      "grad_norm": 0.29503920564203745,
      "learning_rate": 1.725635493875799e-05,
      "loss": 0.285,
      "step": 74
    },
    {
      "epoch": 2.2727272727272725,
      "grad_norm": 0.3249522869032218,
      "learning_rate": 1.7151095013548996e-05,
      "loss": 0.3219,
      "step": 75
    },
    {
      "epoch": 2.303030303030303,
      "grad_norm": 0.31739285818308033,
      "learning_rate": 1.7044187975826126e-05,
      "loss": 0.3111,
      "step": 76
    },
    {
      "epoch": 2.3333333333333335,
      "grad_norm": 0.32599170214519907,
      "learning_rate": 1.693565844949933e-05,
      "loss": 0.3268,
      "step": 77
    },
    {
      "epoch": 2.3636363636363638,
      "grad_norm": 0.3376435002388743,
      "learning_rate": 1.6825531432186545e-05,
      "loss": 0.2868,
      "step": 78
    },
    {
      "epoch": 2.393939393939394,
      "grad_norm": 0.2925735136846681,
      "learning_rate": 1.671383228945597e-05,
      "loss": 0.2906,
      "step": 79
    },
    {
      "epoch": 2.4242424242424243,
      "grad_norm": 0.3462868931497353,
      "learning_rate": 1.6600586748983642e-05,
      "loss": 0.2903,
      "step": 80
    },
    {
      "epoch": 2.4545454545454546,
      "grad_norm": 0.3276418329276385,
      "learning_rate": 1.648582089462756e-05,
      "loss": 0.2987,
      "step": 81
    },
    {
      "epoch": 2.484848484848485,
      "grad_norm": 0.3201279533467403,
      "learning_rate": 1.6369561160419783e-05,
      "loss": 0.2949,
      "step": 82
    },
    {
      "epoch": 2.515151515151515,
      "grad_norm": 0.30112516081818,
      "learning_rate": 1.625183432447789e-05,
      "loss": 0.3081,
      "step": 83
    },
    {
      "epoch": 2.5454545454545454,
      "grad_norm": 0.296162508157177,
      "learning_rate": 1.6132667502837164e-05,
      "loss": 0.293,
      "step": 84
    },
    {
      "epoch": 2.5757575757575757,
      "grad_norm": 0.28908697674303724,
      "learning_rate": 1.6012088143204953e-05,
      "loss": 0.3014,
      "step": 85
    },
    {
      "epoch": 2.606060606060606,
      "grad_norm": 0.31650516349569,
      "learning_rate": 1.589012401863864e-05,
      "loss": 0.303,
      "step": 86
    },
    {
      "epoch": 2.6363636363636362,
      "grad_norm": 0.30839635654078507,
      "learning_rate": 1.5766803221148676e-05,
      "loss": 0.2944,
      "step": 87
    },
    {
      "epoch": 2.6666666666666665,
      "grad_norm": 0.29900590705405367,
      "learning_rate": 1.5642154155228124e-05,
      "loss": 0.3019,
      "step": 88
    },
    {
      "epoch": 2.6969696969696972,
      "grad_norm": 0.3219033217738759,
      "learning_rate": 1.5516205531310272e-05,
      "loss": 0.3189,
      "step": 89
    },
    {
      "epoch": 2.7272727272727275,
      "grad_norm": 0.28978132649253047,
      "learning_rate": 1.538898635915576e-05,
      "loss": 0.2833,
      "step": 90
    },
    {
      "epoch": 2.757575757575758,
      "grad_norm": 0.28915281358268086,
      "learning_rate": 1.526052594117071e-05,
      "loss": 0.3005,
      "step": 91
    },
    {
      "epoch": 2.787878787878788,
      "grad_norm": 0.317698065513808,
      "learning_rate": 1.513085386565758e-05,
      "loss": 0.2977,
      "step": 92
    },
    {
      "epoch": 2.8181818181818183,
      "grad_norm": 0.2498718163599118,
      "learning_rate": 1.5000000000000002e-05,
      "loss": 0.2985,
      "step": 93
    },
    {
      "epoch": 2.8484848484848486,
      "grad_norm": 0.30799699883272946,
      "learning_rate": 1.4867994483783485e-05,
      "loss": 0.2851,
      "step": 94
    },
    {
      "epoch": 2.878787878787879,
      "grad_norm": 0.306217995774187,
      "learning_rate": 1.4734867721853341e-05,
      "loss": 0.3177,
      "step": 95
    },
    {
      "epoch": 2.909090909090909,
      "grad_norm": 0.2825101074951494,
      "learning_rate": 1.4600650377311523e-05,
      "loss": 0.3062,
      "step": 96
    },
    {
      "epoch": 2.9393939393939394,
      "grad_norm": 0.27641742120479323,
      "learning_rate": 1.4465373364454001e-05,
      "loss": 0.2896,
      "step": 97
    },
    {
      "epoch": 2.9696969696969697,
      "grad_norm": 0.26429984167222165,
      "learning_rate": 1.4329067841650274e-05,
      "loss": 0.2882,
      "step": 98
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.27369017823251496,
      "learning_rate": 1.4191765204166643e-05,
      "loss": 0.2905,
      "step": 99
    },
    {
      "epoch": 3.0303030303030303,
      "grad_norm": 0.4379271665681835,
      "learning_rate": 1.4053497076934948e-05,
      "loss": 0.2651,
      "step": 100
    },
    {
      "epoch": 3.0606060606060606,
      "grad_norm": 0.2576991297801112,
      "learning_rate": 1.3914295307268396e-05,
      "loss": 0.2447,
      "step": 101
    },
    {
      "epoch": 3.090909090909091,
      "grad_norm": 0.4655071493240127,
      "learning_rate": 1.3774191957526144e-05,
      "loss": 0.2428,
      "step": 102
    },
    {
      "epoch": 3.121212121212121,
      "grad_norm": 0.33452872754271457,
      "learning_rate": 1.3633219297728415e-05,
      "loss": 0.2105,
      "step": 103
    },
    {
      "epoch": 3.1515151515151514,
      "grad_norm": 0.35271079395528837,
      "learning_rate": 1.3491409798123687e-05,
      "loss": 0.2452,
      "step": 104
    },
    {
      "epoch": 3.1818181818181817,
      "grad_norm": 0.31256584966425977,
      "learning_rate": 1.3348796121709862e-05,
      "loss": 0.248,
      "step": 105
    },
    {
      "epoch": 3.212121212121212,
      "grad_norm": 0.2672038190973873,
      "learning_rate": 1.3205411116710973e-05,
      "loss": 0.2327,
      "step": 106
    },
    {
      "epoch": 3.242424242424242,
      "grad_norm": 0.3918173558277686,
      "learning_rate": 1.3061287809011243e-05,
      "loss": 0.2367,
      "step": 107
    },
    {
      "epoch": 3.2727272727272725,
      "grad_norm": 0.272384410096658,
      "learning_rate": 1.291645939454825e-05,
      "loss": 0.243,
      "step": 108
    },
    {
      "epoch": 3.303030303030303,
      "grad_norm": 0.2864546238013791,
      "learning_rate": 1.277095923166689e-05,
      "loss": 0.2084,
      "step": 109
    },
    {
      "epoch": 3.3333333333333335,
      "grad_norm": 0.32299757549945335,
      "learning_rate": 1.2624820833435939e-05,
      "loss": 0.2225,
      "step": 110
    },
    {
      "epoch": 3.3636363636363638,
      "grad_norm": 0.2584247076475714,
      "learning_rate": 1.2478077859929e-05,
      "loss": 0.2028,
      "step": 111
    },
    {
      "epoch": 3.393939393939394,
      "grad_norm": 0.30195003802581855,
      "learning_rate": 1.2330764110471567e-05,
      "loss": 0.2464,
      "step": 112
    },
    {
      "epoch": 3.4242424242424243,
      "grad_norm": 0.3157033099217462,
      "learning_rate": 1.2182913515856016e-05,
      "loss": 0.2503,
      "step": 113
    },
    {
      "epoch": 3.4545454545454546,
      "grad_norm": 0.2763772769305203,
      "learning_rate": 1.2034560130526341e-05,
      "loss": 0.2326,
      "step": 114
    },
    {
      "epoch": 3.484848484848485,
      "grad_norm": 0.34154725071250613,
      "learning_rate": 1.1885738124734359e-05,
      "loss": 0.249,
      "step": 115
    },
    {
      "epoch": 3.515151515151515,
      "grad_norm": 0.2779686553812174,
      "learning_rate": 1.1736481776669307e-05,
      "loss": 0.2461,
      "step": 116
    },
    {
      "epoch": 3.5454545454545454,
      "grad_norm": 0.2699920292181013,
      "learning_rate": 1.1586825464562515e-05,
      "loss": 0.217,
      "step": 117
    },
    {
      "epoch": 3.5757575757575757,
      "grad_norm": 0.2663980204395775,
      "learning_rate": 1.1436803658769082e-05,
      "loss": 0.2596,
      "step": 118
    },
    {
      "epoch": 3.606060606060606,
      "grad_norm": 0.27551935059361116,
      "learning_rate": 1.1286450913828313e-05,
      "loss": 0.2398,
      "step": 119
    },
    {
      "epoch": 3.6363636363636362,
      "grad_norm": 0.2878469322295445,
      "learning_rate": 1.113580186050475e-05,
      "loss": 0.2248,
      "step": 120
    },
    {
      "epoch": 3.6666666666666665,
      "grad_norm": 0.2319224266307929,
      "learning_rate": 1.0984891197811686e-05,
      "loss": 0.1987,
      "step": 121
    },
    {
      "epoch": 3.6969696969696972,
      "grad_norm": 0.22772317212547327,
      "learning_rate": 1.0833753685018935e-05,
      "loss": 0.2252,
      "step": 122
    },
    {
      "epoch": 3.7272727272727275,
      "grad_norm": 0.2697799612191607,
      "learning_rate": 1.0682424133646712e-05,
      "loss": 0.2114,
      "step": 123
    },
    {
      "epoch": 3.757575757575758,
      "grad_norm": 0.21882269739846996,
      "learning_rate": 1.0530937399447496e-05,
      "loss": 0.2387,
      "step": 124
    },
    {
      "epoch": 3.787878787878788,
      "grad_norm": 0.23342957026997507,
      "learning_rate": 1.0379328374377715e-05,
      "loss": 0.2216,
      "step": 125
    },
    {
      "epoch": 3.8181818181818183,
      "grad_norm": 0.24027126602054363,
      "learning_rate": 1.0227631978561057e-05,
      "loss": 0.2395,
      "step": 126
    },
    {
      "epoch": 3.8484848484848486,
      "grad_norm": 0.2391444739791165,
      "learning_rate": 1.0075883152245334e-05,
      "loss": 0.2174,
      "step": 127
    },
    {
      "epoch": 3.878787878787879,
      "grad_norm": 0.2333196145254371,
      "learning_rate": 9.92411684775467e-06,
      "loss": 0.2105,
      "step": 128
    },
    {
      "epoch": 3.909090909090909,
      "grad_norm": 0.24482124526909385,
      "learning_rate": 9.772368021438943e-06,
      "loss": 0.2204,
      "step": 129
    },
    {
      "epoch": 3.9393939393939394,
      "grad_norm": 0.2272373276105048,
      "learning_rate": 9.620671625622287e-06,
      "loss": 0.228,
      "step": 130
    },
    {
      "epoch": 3.9696969696969697,
      "grad_norm": 0.24386679078044413,
      "learning_rate": 9.469062600552509e-06,
      "loss": 0.2332,
      "step": 131
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.25792001157745154,
      "learning_rate": 9.317575866353293e-06,
      "loss": 0.2136,
      "step": 132
    },
    {
      "epoch": 4.03030303030303,
      "grad_norm": 0.4497958640812333,
      "learning_rate": 9.166246314981066e-06,
      "loss": 0.1768,
      "step": 133
    },
    {
      "epoch": 4.0606060606060606,
      "grad_norm": 0.2495000974300817,
      "learning_rate": 9.015108802188314e-06,
      "loss": 0.1768,
      "step": 134
    },
    {
      "epoch": 4.090909090909091,
      "grad_norm": 0.37634461388937573,
      "learning_rate": 8.86419813949525e-06,
      "loss": 0.1843,
      "step": 135
    },
    {
      "epoch": 4.121212121212121,
      "grad_norm": 0.49240312132092817,
      "learning_rate": 8.71354908617169e-06,
      "loss": 0.1848,
      "step": 136
    },
    {
      "epoch": 4.151515151515151,
      "grad_norm": 0.26158831464133503,
      "learning_rate": 8.56319634123092e-06,
      "loss": 0.173,
      "step": 137
    },
    {
      "epoch": 4.181818181818182,
      "grad_norm": 0.2871673803376742,
      "learning_rate": 8.413174535437486e-06,
      "loss": 0.1967,
      "step": 138
    },
    {
      "epoch": 4.212121212121212,
      "grad_norm": 0.32788733992070856,
      "learning_rate": 8.263518223330698e-06,
      "loss": 0.1508,
      "step": 139
    },
    {
      "epoch": 4.242424242424242,
      "grad_norm": 0.2824673197031457,
      "learning_rate": 8.114261875265643e-06,
      "loss": 0.1709,
      "step": 140
    },
    {
      "epoch": 4.2727272727272725,
      "grad_norm": 0.24692664537214748,
      "learning_rate": 7.965439869473664e-06,
      "loss": 0.1573,
      "step": 141
    },
    {
      "epoch": 4.303030303030303,
      "grad_norm": 0.3223408935785085,
      "learning_rate": 7.817086484143987e-06,
      "loss": 0.1676,
      "step": 142
    },
    {
      "epoch": 4.333333333333333,
      "grad_norm": 0.2543042663079055,
      "learning_rate": 7.669235889528436e-06,
      "loss": 0.1753,
      "step": 143
    },
    {
      "epoch": 4.363636363636363,
      "grad_norm": 0.25847402114728113,
      "learning_rate": 7.521922140071003e-06,
      "loss": 0.1693,
      "step": 144
    },
    {
      "epoch": 4.393939393939394,
      "grad_norm": 0.3032943296847799,
      "learning_rate": 7.375179166564062e-06,
      "loss": 0.1853,
      "step": 145
    },
    {
      "epoch": 4.424242424242424,
      "grad_norm": 0.22023320582819936,
      "learning_rate": 7.2290407683331154e-06,
      "loss": 0.1687,
      "step": 146
    },
    {
      "epoch": 4.454545454545454,
      "grad_norm": 0.2278457431924142,
      "learning_rate": 7.0835406054517505e-06,
      "loss": 0.1759,
      "step": 147
    },
    {
      "epoch": 4.484848484848484,
      "grad_norm": 0.25096265671310064,
      "learning_rate": 6.93871219098876e-06,
      "loss": 0.1695,
      "step": 148
    },
    {
      "epoch": 4.515151515151516,
      "grad_norm": 0.24920165807552327,
      "learning_rate": 6.79458888328903e-06,
      "loss": 0.1897,
      "step": 149
    },
    {
      "epoch": 4.545454545454545,
      "grad_norm": 0.20960258984474384,
      "learning_rate": 6.651203878290139e-06,
      "loss": 0.1941,
      "step": 150
    },
    {
      "epoch": 4.575757575757576,
      "grad_norm": 0.23033932076241462,
      "learning_rate": 6.508590201876317e-06,
      "loss": 0.1942,
      "step": 151
    },
    {
      "epoch": 4.606060606060606,
      "grad_norm": 0.23131764454005913,
      "learning_rate": 6.366780702271589e-06,
      "loss": 0.1906,
      "step": 152
    },
    {
      "epoch": 4.636363636363637,
      "grad_norm": 0.20899100229523568,
      "learning_rate": 6.225808042473857e-06,
      "loss": 0.1857,
      "step": 153
    },
    {
      "epoch": 4.666666666666667,
      "grad_norm": 0.21305875064424762,
      "learning_rate": 6.085704692731609e-06,
      "loss": 0.1902,
      "step": 154
    },
    {
      "epoch": 4.696969696969697,
      "grad_norm": 0.23257289947800558,
      "learning_rate": 5.946502923065054e-06,
      "loss": 0.1816,
      "step": 155
    },
    {
      "epoch": 4.7272727272727275,
      "grad_norm": 0.21456406928205674,
      "learning_rate": 5.8082347958333625e-06,
      "loss": 0.1739,
      "step": 156
    },
    {
      "epoch": 4.757575757575758,
      "grad_norm": 0.21975605738343512,
      "learning_rate": 5.670932158349732e-06,
      "loss": 0.1689,
      "step": 157
    },
    {
      "epoch": 4.787878787878788,
      "grad_norm": 0.20273833751689185,
      "learning_rate": 5.534626635546e-06,
      "loss": 0.2137,
      "step": 158
    },
    {
      "epoch": 4.818181818181818,
      "grad_norm": 0.2241675411103492,
      "learning_rate": 5.399349622688479e-06,
      "loss": 0.1819,
      "step": 159
    },
    {
      "epoch": 4.848484848484849,
      "grad_norm": 0.2117126616447277,
      "learning_rate": 5.2651322781466606e-06,
      "loss": 0.1682,
      "step": 160
    },
    {
      "epoch": 4.878787878787879,
      "grad_norm": 0.23009636595672053,
      "learning_rate": 5.132005516216512e-06,
      "loss": 0.1841,
      "step": 161
    },
    {
      "epoch": 4.909090909090909,
      "grad_norm": 0.20039506801416299,
      "learning_rate": 5.000000000000003e-06,
      "loss": 0.1755,
      "step": 162
    },
    {
      "epoch": 4.9393939393939394,
      "grad_norm": 0.3522687784544819,
      "learning_rate": 4.869146134342426e-06,
      "loss": 0.1563,
      "step": 163
    },
    {
      "epoch": 4.96969696969697,
      "grad_norm": 0.18977326916591455,
      "learning_rate": 4.739474058829288e-06,
      "loss": 0.179,
      "step": 164
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.19888889302278237,
      "learning_rate": 4.611013640844245e-06,
      "loss": 0.1688,
      "step": 165
    },
    {
      "epoch": 5.03030303030303,
      "grad_norm": 0.32964305091003154,
      "learning_rate": 4.483794468689728e-06,
      "loss": 0.174,
      "step": 166
    },
    {
      "epoch": 5.0606060606060606,
      "grad_norm": 0.2560513783200906,
      "learning_rate": 4.357845844771881e-06,
      "loss": 0.169,
      "step": 167
    },
    {
      "epoch": 5.090909090909091,
      "grad_norm": 0.16313677453301711,
      "learning_rate": 4.2331967788513295e-06,
      "loss": 0.1439,
      "step": 168
    },
    {
      "epoch": 5.121212121212121,
      "grad_norm": 0.20091492477532555,
      "learning_rate": 4.109875981361363e-06,
      "loss": 0.1564,
      "step": 169
    },
    {
      "epoch": 5.151515151515151,
      "grad_norm": 0.29151486179330227,
      "learning_rate": 3.987911856795047e-06,
      "loss": 0.1452,
      "step": 170
    },
    {
      "epoch": 5.181818181818182,
      "grad_norm": 0.2717084479543799,
      "learning_rate": 3.867332497162836e-06,
      "loss": 0.1563,
      "step": 171
    },
    {
      "epoch": 5.212121212121212,
      "grad_norm": 0.19361601408502685,
      "learning_rate": 3.748165675522113e-06,
      "loss": 0.1393,
      "step": 172
    },
    {
      "epoch": 5.242424242424242,
      "grad_norm": 0.1776164713562721,
      "learning_rate": 3.630438839580217e-06,
      "loss": 0.1386,
      "step": 173
    },
    {
      "epoch": 5.2727272727272725,
      "grad_norm": 0.21149360588780272,
      "learning_rate": 3.5141791053724405e-06,
      "loss": 0.1278,
      "step": 174
    },
    {
      "epoch": 5.303030303030303,
      "grad_norm": 0.1812641490516299,
      "learning_rate": 3.399413251016359e-06,
      "loss": 0.127,
      "step": 175
    },
    {
      "epoch": 5.333333333333333,
      "grad_norm": 0.20316650540360767,
      "learning_rate": 3.2861677105440335e-06,
      "loss": 0.1556,
      "step": 176
    },
    {
      "epoch": 5.363636363636363,
      "grad_norm": 0.20743132556817231,
      "learning_rate": 3.174468567813461e-06,
      "loss": 0.1416,
      "step": 177
    },
    {
      "epoch": 5.393939393939394,
      "grad_norm": 0.15166134187287506,
      "learning_rate": 3.0643415505006733e-06,
      "loss": 0.1595,
      "step": 178
    },
    {
      "epoch": 5.424242424242424,
      "grad_norm": 0.19878635072355103,
      "learning_rate": 2.9558120241738786e-06,
      "loss": 0.1385,
      "step": 179
    },
    {
      "epoch": 5.454545454545454,
      "grad_norm": 0.20472542514886816,
      "learning_rate": 2.8489049864510053e-06,
      "loss": 0.1398,
      "step": 180
    },
    {
      "epoch": 5.484848484848484,
      "grad_norm": 0.18038410104890473,
      "learning_rate": 2.7436450612420098e-06,
      "loss": 0.1345,
      "step": 181
    },
    {
      "epoch": 5.515151515151516,
      "grad_norm": 0.16870782003870408,
      "learning_rate": 2.640056493077231e-06,
      "loss": 0.1557,
      "step": 182
    },
    {
      "epoch": 5.545454545454545,
      "grad_norm": 0.17267230249611165,
      "learning_rate": 2.5381631415231455e-06,
      "loss": 0.1492,
      "step": 183
    },
    {
      "epoch": 5.575757575757576,
      "grad_norm": 0.16322539444562315,
      "learning_rate": 2.4379884756868167e-06,
      "loss": 0.1562,
      "step": 184
    },
    {
      "epoch": 5.606060606060606,
      "grad_norm": 0.1798535094635071,
      "learning_rate": 2.339555568810221e-06,
      "loss": 0.1509,
      "step": 185
    },
    {
      "epoch": 5.636363636363637,
      "grad_norm": 0.1817266499609535,
      "learning_rate": 2.2428870929558012e-06,
      "loss": 0.1461,
      "step": 186
    },
    {
      "epoch": 5.666666666666667,
      "grad_norm": 0.175952260181304,
      "learning_rate": 2.1480053137844115e-06,
      "loss": 0.1375,
      "step": 187
    },
    {
      "epoch": 5.696969696969697,
      "grad_norm": 0.176625403397863,
      "learning_rate": 2.054932085426856e-06,
      "loss": 0.1487,
      "step": 188
    },
    {
      "epoch": 5.7272727272727275,
      "grad_norm": 0.15366517772742452,
      "learning_rate": 1.963688845450218e-06,
      "loss": 0.1401,
      "step": 189
    },
    {
      "epoch": 5.757575757575758,
      "grad_norm": 0.1748429855704524,
      "learning_rate": 1.8742966099201699e-06,
      "loss": 0.1414,
      "step": 190
    },
    {
      "epoch": 5.787878787878788,
      "grad_norm": 0.17411294450742076,
      "learning_rate": 1.7867759685603115e-06,
      "loss": 0.1283,
      "step": 191
    },
    {
      "epoch": 5.818181818181818,
      "grad_norm": 0.16742314943694317,
      "learning_rate": 1.7011470800097496e-06,
      "loss": 0.133,
      "step": 192
    },
    {
      "epoch": 5.848484848484849,
      "grad_norm": 0.17309315906666728,
      "learning_rate": 1.6174296671799571e-06,
      "loss": 0.143,
      "step": 193
    },
    {
      "epoch": 5.878787878787879,
      "grad_norm": 0.17504293024016182,
      "learning_rate": 1.5356430127119915e-06,
      "loss": 0.1342,
      "step": 194
    },
    {
      "epoch": 5.909090909090909,
      "grad_norm": 0.16468434385732947,
      "learning_rate": 1.4558059545351144e-06,
      "loss": 0.1393,
      "step": 195
    },
    {
      "epoch": 5.9393939393939394,
      "grad_norm": 0.15478117129103133,
      "learning_rate": 1.3779368815278648e-06,
      "loss": 0.152,
      "step": 196
    },
    {
      "epoch": 5.96969696969697,
      "grad_norm": 0.15582554266386373,
      "learning_rate": 1.302053729282533e-06,
      "loss": 0.1484,
      "step": 197
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.16478648898818704,
      "learning_rate": 1.2281739759740575e-06,
      "loss": 0.1227,
      "step": 198
    },
    {
      "epoch": 6.03030303030303,
      "grad_norm": 0.24905591211291422,
      "learning_rate": 1.156314638334277e-06,
      "loss": 0.1269,
      "step": 199
    },
    {
      "epoch": 6.0606060606060606,
      "grad_norm": 0.1827020945064971,
      "learning_rate": 1.086492267732462e-06,
      "loss": 0.1196,
      "step": 200
    },
    {
      "epoch": 6.090909090909091,
      "grad_norm": 0.1995680558560439,
      "learning_rate": 1.01872294636304e-06,
      "loss": 0.1188,
      "step": 201
    },
    {
      "epoch": 6.121212121212121,
      "grad_norm": 0.15288343845707877,
      "learning_rate": 9.530222835413739e-07,
      "loss": 0.1002,
      "step": 202
    },
    {
      "epoch": 6.151515151515151,
      "grad_norm": 0.1495322751013784,
      "learning_rate": 8.894054121084839e-07,
      "loss": 0.1471,
      "step": 203
    },
    {
      "epoch": 6.181818181818182,
      "grad_norm": 0.13239874942254765,
      "learning_rate": 8.278869849454718e-07,
      "loss": 0.1095,
      "step": 204
    },
    {
      "epoch": 6.212121212121212,
      "grad_norm": 0.14965056372026952,
      "learning_rate": 7.684811715985429e-07,
      "loss": 0.1207,
      "step": 205
    },
    {
      "epoch": 6.242424242424242,
      "grad_norm": 0.1555591619487068,
      "learning_rate": 7.1120165501533e-07,
      "loss": 0.1164,
      "step": 206
    },
    {
      "epoch": 6.2727272727272725,
      "grad_norm": 0.22050786978749573,
      "learning_rate": 6.560616283932897e-07,
      "loss": 0.1354,
      "step": 207
    },
    {
      "epoch": 6.303030303030303,
      "grad_norm": 0.15847848132858028,
      "learning_rate": 6.030737921409169e-07,
      "loss": 0.1299,
      "step": 208
    },
    {
      "epoch": 6.333333333333333,
      "grad_norm": 0.17273454607795033,
      "learning_rate": 5.522503509524591e-07,
      "loss": 0.1546,
      "step": 209
    },
    {
      "epoch": 6.363636363636363,
      "grad_norm": 0.20176869735673397,
      "learning_rate": 5.036030109968082e-07,
      "loss": 0.1274,
      "step": 210
    },
    {
      "epoch": 6.393939393939394,
      "grad_norm": 0.13648820775005097,
      "learning_rate": 4.5714297722121105e-07,
      "loss": 0.1239,
      "step": 211
    },
    {
      "epoch": 6.424242424242424,
      "grad_norm": 0.18568154870934608,
      "learning_rate": 4.128809507704445e-07,
      "loss": 0.1332,
      "step": 212
    },
    {
      "epoch": 6.454545454545454,
      "grad_norm": 0.16013659759492688,
      "learning_rate": 3.708271265220087e-07,
      "loss": 0.173,
      "step": 213
    },
    {
      "epoch": 6.484848484848484,
      "grad_norm": 0.1272351001677736,
      "learning_rate": 3.309911907379393e-07,
      "loss": 0.116,
      "step": 214
    },
    {
      "epoch": 6.515151515151516,
      "grad_norm": 0.13748206637327287,
      "learning_rate": 2.9338231883378365e-07,
      "loss": 0.1305,
      "step": 215
    },
    {
      "epoch": 6.545454545454545,
      "grad_norm": 0.1389062314037023,
      "learning_rate": 2.5800917326521013e-07,
      "loss": 0.1359,
      "step": 216
    },
    {
      "epoch": 6.575757575757576,
      "grad_norm": 0.13540280025821855,
      "learning_rate": 2.248799015327907e-07,
      "loss": 0.1468,
      "step": 217
    },
    {
      "epoch": 6.606060606060606,
      "grad_norm": 0.1317157448820014,
      "learning_rate": 1.9400213430538773e-07,
      "loss": 0.1329,
      "step": 218
    },
    {
      "epoch": 6.636363636363637,
      "grad_norm": 0.12995697988304855,
      "learning_rate": 1.6538298366257975e-07,
      "loss": 0.1254,
      "step": 219
    },
    {
      "epoch": 6.666666666666667,
      "grad_norm": 0.1466780133162069,
      "learning_rate": 1.3902904145653094e-07,
      "loss": 0.1271,
      "step": 220
    },
    {
      "epoch": 6.696969696969697,
      "grad_norm": 0.12014356166279773,
      "learning_rate": 1.1494637779369766e-07,
      "loss": 0.1148,
      "step": 221
    },
    {
      "epoch": 6.7272727272727275,
      "grad_norm": 0.12244331989103567,
      "learning_rate": 9.314053963669245e-08,
      "loss": 0.1315,
      "step": 222
    },
    {
      "epoch": 6.757575757575758,
      "grad_norm": 0.12359131379931636,
      "learning_rate": 7.361654952665608e-08,
      "loss": 0.1348,
      "step": 223
    },
    {
      "epoch": 6.787878787878788,
      "grad_norm": 0.13143701512591202,
      "learning_rate": 5.637890442641403e-08,
      "loss": 0.1252,
      "step": 224
    },
    {
      "epoch": 6.818181818181818,
      "grad_norm": 0.12993742599930214,
      "learning_rate": 4.143157468468717e-08,
      "loss": 0.1242,
      "step": 225
    },
    {
      "epoch": 6.848484848484849,
      "grad_norm": 0.137924427525877,
      "learning_rate": 2.8778003121607834e-08,
      "loss": 0.1384,
      "step": 226
    },
    {
      "epoch": 6.878787878787879,
      "grad_norm": 0.14052459669579978,
      "learning_rate": 1.8421104235727406e-08,
      "loss": 0.1267,
      "step": 227
    },
    {
      "epoch": 6.909090909090909,
      "grad_norm": 0.13787931148251978,
      "learning_rate": 1.0363263532724433e-08,
      "loss": 0.112,
      "step": 228
    },
    {
      "epoch": 6.9393939393939394,
      "grad_norm": 0.14649122214098168,
      "learning_rate": 4.606336975948589e-09,
      "loss": 0.1518,
      "step": 229
    },
    {
      "epoch": 6.96969696969697,
      "grad_norm": 0.14336516467399005,
      "learning_rate": 1.1516505589381777e-09,
      "loss": 0.1522,
      "step": 230
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.1365687268742612,
      "learning_rate": 0.0,
      "loss": 0.1402,
      "step": 231
    },
    {
      "epoch": 7.0,
      "step": 231,
      "total_flos": 3.4960602118383206e+17,
      "train_loss": 0.28039634318062756,
      "train_runtime": 11689.2231,
      "train_samples_per_second": 1.892,
      "train_steps_per_second": 0.02
    }
  ],
  "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": 3.4960602118383206e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}