{
  "best_metric": 2.652009963989258,
  "best_model_checkpoint": "miner_id_24/checkpoint-200",
  "epoch": 1.4990654205607477,
  "eval_steps": 50,
  "global_step": 200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.007476635514018692,
      "grad_norm": 4.018685817718506,
      "learning_rate": 1e-05,
      "loss": 3.2199,
      "step": 1
    },
    {
      "epoch": 0.007476635514018692,
      "eval_loss": 4.9851226806640625,
      "eval_runtime": 21.1936,
      "eval_samples_per_second": 10.664,
      "eval_steps_per_second": 2.689,
      "step": 1
    },
    {
      "epoch": 0.014953271028037384,
      "grad_norm": 3.7604949474334717,
      "learning_rate": 2e-05,
      "loss": 3.3533,
      "step": 2
    },
    {
      "epoch": 0.022429906542056073,
      "grad_norm": 3.7570433616638184,
      "learning_rate": 3e-05,
      "loss": 3.3395,
      "step": 3
    },
    {
      "epoch": 0.029906542056074768,
      "grad_norm": 3.352917432785034,
      "learning_rate": 4e-05,
      "loss": 3.279,
      "step": 4
    },
    {
      "epoch": 0.037383177570093455,
      "grad_norm": 3.761164903640747,
      "learning_rate": 5e-05,
      "loss": 3.0999,
      "step": 5
    },
    {
      "epoch": 0.044859813084112146,
      "grad_norm": 2.679424524307251,
      "learning_rate": 6e-05,
      "loss": 3.1394,
      "step": 6
    },
    {
      "epoch": 0.052336448598130844,
      "grad_norm": 2.48563551902771,
      "learning_rate": 7e-05,
      "loss": 2.5872,
      "step": 7
    },
    {
      "epoch": 0.059813084112149535,
      "grad_norm": 2.618500232696533,
      "learning_rate": 8e-05,
      "loss": 3.0987,
      "step": 8
    },
    {
      "epoch": 0.06728971962616823,
      "grad_norm": 2.312514305114746,
      "learning_rate": 9e-05,
      "loss": 2.7576,
      "step": 9
    },
    {
      "epoch": 0.07476635514018691,
      "grad_norm": 2.8855197429656982,
      "learning_rate": 0.0001,
      "loss": 3.0275,
      "step": 10
    },
    {
      "epoch": 0.08224299065420561,
      "grad_norm": 2.5778770446777344,
      "learning_rate": 9.999316524962345e-05,
      "loss": 3.1095,
      "step": 11
    },
    {
      "epoch": 0.08971962616822429,
      "grad_norm": 2.2357723712921143,
      "learning_rate": 9.997266286704631e-05,
      "loss": 2.9825,
      "step": 12
    },
    {
      "epoch": 0.09719626168224299,
      "grad_norm": 1.975340485572815,
      "learning_rate": 9.993849845741524e-05,
      "loss": 2.9437,
      "step": 13
    },
    {
      "epoch": 0.10467289719626169,
      "grad_norm": 1.90480637550354,
      "learning_rate": 9.989068136093873e-05,
      "loss": 2.7934,
      "step": 14
    },
    {
      "epoch": 0.11214953271028037,
      "grad_norm": 2.49348783493042,
      "learning_rate": 9.98292246503335e-05,
      "loss": 2.8363,
      "step": 15
    },
    {
      "epoch": 0.11962616822429907,
      "grad_norm": 2.5422096252441406,
      "learning_rate": 9.975414512725057e-05,
      "loss": 3.0426,
      "step": 16
    },
    {
      "epoch": 0.12710280373831775,
      "grad_norm": 2.304914712905884,
      "learning_rate": 9.966546331768191e-05,
      "loss": 3.1102,
      "step": 17
    },
    {
      "epoch": 0.13457943925233645,
      "grad_norm": 2.8444602489471436,
      "learning_rate": 9.956320346634876e-05,
      "loss": 3.0561,
      "step": 18
    },
    {
      "epoch": 0.14205607476635515,
      "grad_norm": 1.9882365465164185,
      "learning_rate": 9.944739353007344e-05,
      "loss": 3.0771,
      "step": 19
    },
    {
      "epoch": 0.14953271028037382,
      "grad_norm": 2.390509605407715,
      "learning_rate": 9.931806517013612e-05,
      "loss": 3.1419,
      "step": 20
    },
    {
      "epoch": 0.15700934579439252,
      "grad_norm": 2.4310107231140137,
      "learning_rate": 9.917525374361912e-05,
      "loss": 3.0508,
      "step": 21
    },
    {
      "epoch": 0.16448598130841122,
      "grad_norm": 2.5127458572387695,
      "learning_rate": 9.901899829374047e-05,
      "loss": 3.0841,
      "step": 22
    },
    {
      "epoch": 0.17196261682242991,
      "grad_norm": 3.741215705871582,
      "learning_rate": 9.884934153917997e-05,
      "loss": 3.0784,
      "step": 23
    },
    {
      "epoch": 0.17943925233644858,
      "grad_norm": 2.4267561435699463,
      "learning_rate": 9.86663298624003e-05,
      "loss": 2.8861,
      "step": 24
    },
    {
      "epoch": 0.18691588785046728,
      "grad_norm": 3.180570125579834,
      "learning_rate": 9.847001329696653e-05,
      "loss": 3.1596,
      "step": 25
    },
    {
      "epoch": 0.19439252336448598,
      "grad_norm": 3.063917636871338,
      "learning_rate": 9.826044551386744e-05,
      "loss": 3.334,
      "step": 26
    },
    {
      "epoch": 0.20186915887850468,
      "grad_norm": 3.1439948081970215,
      "learning_rate": 9.803768380684242e-05,
      "loss": 3.7414,
      "step": 27
    },
    {
      "epoch": 0.20934579439252338,
      "grad_norm": 3.375192403793335,
      "learning_rate": 9.780178907671789e-05,
      "loss": 3.2579,
      "step": 28
    },
    {
      "epoch": 0.21682242990654205,
      "grad_norm": 3.557861089706421,
      "learning_rate": 9.755282581475769e-05,
      "loss": 3.663,
      "step": 29
    },
    {
      "epoch": 0.22429906542056074,
      "grad_norm": 5.07192850112915,
      "learning_rate": 9.729086208503174e-05,
      "loss": 4.0257,
      "step": 30
    },
    {
      "epoch": 0.23177570093457944,
      "grad_norm": 4.732841491699219,
      "learning_rate": 9.701596950580806e-05,
      "loss": 3.7215,
      "step": 31
    },
    {
      "epoch": 0.23925233644859814,
      "grad_norm": 4.269678592681885,
      "learning_rate": 9.672822322997305e-05,
      "loss": 4.0007,
      "step": 32
    },
    {
      "epoch": 0.2467289719626168,
      "grad_norm": 5.582366466522217,
      "learning_rate": 9.642770192448536e-05,
      "loss": 4.1847,
      "step": 33
    },
    {
      "epoch": 0.2542056074766355,
      "grad_norm": 20.683387756347656,
      "learning_rate": 9.611448774886924e-05,
      "loss": 4.7626,
      "step": 34
    },
    {
      "epoch": 0.2616822429906542,
      "grad_norm": 8.848209381103516,
      "learning_rate": 9.578866633275288e-05,
      "loss": 3.5258,
      "step": 35
    },
    {
      "epoch": 0.2691588785046729,
      "grad_norm": 4.423970699310303,
      "learning_rate": 9.545032675245813e-05,
      "loss": 3.1521,
      "step": 36
    },
    {
      "epoch": 0.2766355140186916,
      "grad_norm": 2.0548391342163086,
      "learning_rate": 9.509956150664796e-05,
      "loss": 2.882,
      "step": 37
    },
    {
      "epoch": 0.2841121495327103,
      "grad_norm": 1.6080002784729004,
      "learning_rate": 9.473646649103818e-05,
      "loss": 2.6264,
      "step": 38
    },
    {
      "epoch": 0.29158878504672897,
      "grad_norm": 1.3412339687347412,
      "learning_rate": 9.43611409721806e-05,
      "loss": 2.5175,
      "step": 39
    },
    {
      "epoch": 0.29906542056074764,
      "grad_norm": 1.6938002109527588,
      "learning_rate": 9.397368756032445e-05,
      "loss": 2.6353,
      "step": 40
    },
    {
      "epoch": 0.30654205607476637,
      "grad_norm": 1.3714394569396973,
      "learning_rate": 9.357421218136386e-05,
      "loss": 2.3522,
      "step": 41
    },
    {
      "epoch": 0.31401869158878504,
      "grad_norm": 1.3818203210830688,
      "learning_rate": 9.316282404787871e-05,
      "loss": 2.8121,
      "step": 42
    },
    {
      "epoch": 0.32149532710280376,
      "grad_norm": 1.6380140781402588,
      "learning_rate": 9.273963562927695e-05,
      "loss": 2.5245,
      "step": 43
    },
    {
      "epoch": 0.32897196261682243,
      "grad_norm": 1.567598819732666,
      "learning_rate": 9.230476262104677e-05,
      "loss": 2.5743,
      "step": 44
    },
    {
      "epoch": 0.3364485981308411,
      "grad_norm": 1.3986696004867554,
      "learning_rate": 9.185832391312644e-05,
      "loss": 2.5472,
      "step": 45
    },
    {
      "epoch": 0.34392523364485983,
      "grad_norm": 1.3626127243041992,
      "learning_rate": 9.140044155740101e-05,
      "loss": 2.7207,
      "step": 46
    },
    {
      "epoch": 0.3514018691588785,
      "grad_norm": 1.588208556175232,
      "learning_rate": 9.093124073433463e-05,
      "loss": 2.7109,
      "step": 47
    },
    {
      "epoch": 0.35887850467289717,
      "grad_norm": 1.4778130054473877,
      "learning_rate": 9.045084971874738e-05,
      "loss": 2.6878,
      "step": 48
    },
    {
      "epoch": 0.3663551401869159,
      "grad_norm": 1.4806129932403564,
      "learning_rate": 8.995939984474624e-05,
      "loss": 2.6774,
      "step": 49
    },
    {
      "epoch": 0.37383177570093457,
      "grad_norm": 1.6793596744537354,
      "learning_rate": 8.945702546981969e-05,
      "loss": 2.5917,
      "step": 50
    },
    {
      "epoch": 0.37383177570093457,
      "eval_loss": 3.0735702514648438,
      "eval_runtime": 21.5261,
      "eval_samples_per_second": 10.499,
      "eval_steps_per_second": 2.648,
      "step": 50
    },
    {
      "epoch": 0.3813084112149533,
      "grad_norm": 2.2972981929779053,
      "learning_rate": 8.894386393810563e-05,
      "loss": 2.8403,
      "step": 51
    },
    {
      "epoch": 0.38878504672897196,
      "grad_norm": 1.604483962059021,
      "learning_rate": 8.842005554284296e-05,
      "loss": 2.9061,
      "step": 52
    },
    {
      "epoch": 0.39626168224299063,
      "grad_norm": 2.183002233505249,
      "learning_rate": 8.788574348801675e-05,
      "loss": 2.75,
      "step": 53
    },
    {
      "epoch": 0.40373831775700936,
      "grad_norm": 1.8150001764297485,
      "learning_rate": 8.73410738492077e-05,
      "loss": 2.8232,
      "step": 54
    },
    {
      "epoch": 0.411214953271028,
      "grad_norm": 1.9930814504623413,
      "learning_rate": 8.678619553365659e-05,
      "loss": 2.8307,
      "step": 55
    },
    {
      "epoch": 0.41869158878504675,
      "grad_norm": 2.117896795272827,
      "learning_rate": 8.622126023955446e-05,
      "loss": 2.9888,
      "step": 56
    },
    {
      "epoch": 0.4261682242990654,
      "grad_norm": 2.1339077949523926,
      "learning_rate": 8.564642241456986e-05,
      "loss": 3.1136,
      "step": 57
    },
    {
      "epoch": 0.4336448598130841,
      "grad_norm": 2.300039052963257,
      "learning_rate": 8.506183921362443e-05,
      "loss": 3.1252,
      "step": 58
    },
    {
      "epoch": 0.4411214953271028,
      "grad_norm": 2.5407261848449707,
      "learning_rate": 8.44676704559283e-05,
      "loss": 3.2192,
      "step": 59
    },
    {
      "epoch": 0.4485981308411215,
      "grad_norm": 1.9730626344680786,
      "learning_rate": 8.386407858128706e-05,
      "loss": 2.9494,
      "step": 60
    },
    {
      "epoch": 0.45607476635514016,
      "grad_norm": 2.6272242069244385,
      "learning_rate": 8.32512286056924e-05,
      "loss": 3.2439,
      "step": 61
    },
    {
      "epoch": 0.4635514018691589,
      "grad_norm": 2.9058167934417725,
      "learning_rate": 8.262928807620843e-05,
      "loss": 3.0013,
      "step": 62
    },
    {
      "epoch": 0.47102803738317756,
      "grad_norm": 3.5064244270324707,
      "learning_rate": 8.199842702516583e-05,
      "loss": 3.579,
      "step": 63
    },
    {
      "epoch": 0.4785046728971963,
      "grad_norm": 3.474576234817505,
      "learning_rate": 8.135881792367686e-05,
      "loss": 3.6689,
      "step": 64
    },
    {
      "epoch": 0.48598130841121495,
      "grad_norm": 4.29802942276001,
      "learning_rate": 8.07106356344834e-05,
      "loss": 3.5855,
      "step": 65
    },
    {
      "epoch": 0.4934579439252336,
      "grad_norm": 3.9347784519195557,
      "learning_rate": 8.005405736415126e-05,
      "loss": 4.11,
      "step": 66
    },
    {
      "epoch": 0.5009345794392523,
      "grad_norm": 23.299638748168945,
      "learning_rate": 7.938926261462366e-05,
      "loss": 4.4604,
      "step": 67
    },
    {
      "epoch": 0.508411214953271,
      "grad_norm": 10.690607070922852,
      "learning_rate": 7.871643313414718e-05,
      "loss": 3.4988,
      "step": 68
    },
    {
      "epoch": 0.5158878504672897,
      "grad_norm": 5.5861077308654785,
      "learning_rate": 7.803575286758364e-05,
      "loss": 3.1854,
      "step": 69
    },
    {
      "epoch": 0.5233644859813084,
      "grad_norm": 4.5043487548828125,
      "learning_rate": 7.734740790612136e-05,
      "loss": 2.763,
      "step": 70
    },
    {
      "epoch": 0.5308411214953271,
      "grad_norm": 2.52915620803833,
      "learning_rate": 7.66515864363997e-05,
      "loss": 2.6761,
      "step": 71
    },
    {
      "epoch": 0.5383177570093458,
      "grad_norm": 1.552752137184143,
      "learning_rate": 7.594847868906076e-05,
      "loss": 2.4099,
      "step": 72
    },
    {
      "epoch": 0.5457943925233645,
      "grad_norm": 1.2231338024139404,
      "learning_rate": 7.52382768867422e-05,
      "loss": 2.4981,
      "step": 73
    },
    {
      "epoch": 0.5532710280373832,
      "grad_norm": 1.0463976860046387,
      "learning_rate": 7.452117519152542e-05,
      "loss": 2.5872,
      "step": 74
    },
    {
      "epoch": 0.5607476635514018,
      "grad_norm": 1.1301862001419067,
      "learning_rate": 7.379736965185368e-05,
      "loss": 2.4969,
      "step": 75
    },
    {
      "epoch": 0.5682242990654206,
      "grad_norm": 1.101419448852539,
      "learning_rate": 7.30670581489344e-05,
      "loss": 2.4788,
      "step": 76
    },
    {
      "epoch": 0.5757009345794393,
      "grad_norm": 1.1890898942947388,
      "learning_rate": 7.233044034264034e-05,
      "loss": 2.4436,
      "step": 77
    },
    {
      "epoch": 0.5831775700934579,
      "grad_norm": 1.28480863571167,
      "learning_rate": 7.158771761692464e-05,
      "loss": 2.5141,
      "step": 78
    },
    {
      "epoch": 0.5906542056074766,
      "grad_norm": 1.3600590229034424,
      "learning_rate": 7.083909302476453e-05,
      "loss": 2.4151,
      "step": 79
    },
    {
      "epoch": 0.5981308411214953,
      "grad_norm": 1.30199134349823,
      "learning_rate": 7.008477123264848e-05,
      "loss": 2.4949,
      "step": 80
    },
    {
      "epoch": 0.6056074766355141,
      "grad_norm": 1.410233974456787,
      "learning_rate": 6.932495846462261e-05,
      "loss": 2.703,
      "step": 81
    },
    {
      "epoch": 0.6130841121495327,
      "grad_norm": 1.7426823377609253,
      "learning_rate": 6.855986244591104e-05,
      "loss": 2.6741,
      "step": 82
    },
    {
      "epoch": 0.6205607476635514,
      "grad_norm": 1.559410572052002,
      "learning_rate": 6.778969234612584e-05,
      "loss": 2.8253,
      "step": 83
    },
    {
      "epoch": 0.6280373831775701,
      "grad_norm": 1.5492641925811768,
      "learning_rate": 6.701465872208216e-05,
      "loss": 2.6632,
      "step": 84
    },
    {
      "epoch": 0.6355140186915887,
      "grad_norm": 1.700284481048584,
      "learning_rate": 6.623497346023418e-05,
      "loss": 2.7667,
      "step": 85
    },
    {
      "epoch": 0.6429906542056075,
      "grad_norm": 1.6237927675247192,
      "learning_rate": 6.545084971874738e-05,
      "loss": 2.672,
      "step": 86
    },
    {
      "epoch": 0.6504672897196262,
      "grad_norm": 2.156932830810547,
      "learning_rate": 6.466250186922325e-05,
      "loss": 2.6041,
      "step": 87
    },
    {
      "epoch": 0.6579439252336449,
      "grad_norm": 2.0370500087738037,
      "learning_rate": 6.387014543809223e-05,
      "loss": 2.8062,
      "step": 88
    },
    {
      "epoch": 0.6654205607476635,
      "grad_norm": 2.302056312561035,
      "learning_rate": 6.307399704769099e-05,
      "loss": 3.1741,
      "step": 89
    },
    {
      "epoch": 0.6728971962616822,
      "grad_norm": 2.2182841300964355,
      "learning_rate": 6.227427435703997e-05,
      "loss": 3.11,
      "step": 90
    },
    {
      "epoch": 0.680373831775701,
      "grad_norm": 2.420154094696045,
      "learning_rate": 6.147119600233758e-05,
      "loss": 3.1094,
      "step": 91
    },
    {
      "epoch": 0.6878504672897197,
      "grad_norm": 2.4488508701324463,
      "learning_rate": 6.066498153718735e-05,
      "loss": 3.0456,
      "step": 92
    },
    {
      "epoch": 0.6953271028037383,
      "grad_norm": 2.7070367336273193,
      "learning_rate": 5.985585137257401e-05,
      "loss": 3.2094,
      "step": 93
    },
    {
      "epoch": 0.702803738317757,
      "grad_norm": 2.7324931621551514,
      "learning_rate": 5.90440267166055e-05,
      "loss": 3.426,
      "step": 94
    },
    {
      "epoch": 0.7102803738317757,
      "grad_norm": 3.2201993465423584,
      "learning_rate": 5.8229729514036705e-05,
      "loss": 3.8129,
      "step": 95
    },
    {
      "epoch": 0.7177570093457943,
      "grad_norm": 3.0422415733337402,
      "learning_rate": 5.74131823855921e-05,
      "loss": 3.558,
      "step": 96
    },
    {
      "epoch": 0.7252336448598131,
      "grad_norm": 3.3264083862304688,
      "learning_rate": 5.6594608567103456e-05,
      "loss": 3.3661,
      "step": 97
    },
    {
      "epoch": 0.7327102803738318,
      "grad_norm": 3.449193239212036,
      "learning_rate": 5.577423184847932e-05,
      "loss": 3.4449,
      "step": 98
    },
    {
      "epoch": 0.7401869158878505,
      "grad_norm": 6.108767032623291,
      "learning_rate": 5.495227651252315e-05,
      "loss": 4.2715,
      "step": 99
    },
    {
      "epoch": 0.7476635514018691,
      "grad_norm": 24.272310256958008,
      "learning_rate": 5.4128967273616625e-05,
      "loss": 4.2324,
      "step": 100
    },
    {
      "epoch": 0.7476635514018691,
      "eval_loss": 3.636786699295044,
      "eval_runtime": 21.5204,
      "eval_samples_per_second": 10.502,
      "eval_steps_per_second": 2.649,
      "step": 100
    },
    {
      "epoch": 0.7551401869158878,
      "grad_norm": 21.568483352661133,
      "learning_rate": 5.330452921628497e-05,
      "loss": 4.0683,
      "step": 101
    },
    {
      "epoch": 0.7626168224299066,
      "grad_norm": 13.84886646270752,
      "learning_rate": 5.247918773366112e-05,
      "loss": 3.6184,
      "step": 102
    },
    {
      "epoch": 0.7700934579439253,
      "grad_norm": 7.929118633270264,
      "learning_rate": 5.165316846586541e-05,
      "loss": 2.8035,
      "step": 103
    },
    {
      "epoch": 0.7775700934579439,
      "grad_norm": 5.159682273864746,
      "learning_rate": 5.0826697238317935e-05,
      "loss": 2.6174,
      "step": 104
    },
    {
      "epoch": 0.7850467289719626,
      "grad_norm": 4.41166877746582,
      "learning_rate": 5e-05,
      "loss": 2.5002,
      "step": 105
    },
    {
      "epoch": 0.7925233644859813,
      "grad_norm": 3.252392292022705,
      "learning_rate": 4.917330276168208e-05,
      "loss": 2.4566,
      "step": 106
    },
    {
      "epoch": 0.8,
      "grad_norm": 3.977404832839966,
      "learning_rate": 4.834683153413459e-05,
      "loss": 2.7188,
      "step": 107
    },
    {
      "epoch": 0.8074766355140187,
      "grad_norm": 3.309093475341797,
      "learning_rate": 4.7520812266338885e-05,
      "loss": 2.8443,
      "step": 108
    },
    {
      "epoch": 0.8149532710280374,
      "grad_norm": 1.8352887630462646,
      "learning_rate": 4.669547078371504e-05,
      "loss": 2.8052,
      "step": 109
    },
    {
      "epoch": 0.822429906542056,
      "grad_norm": 1.6216166019439697,
      "learning_rate": 4.5871032726383386e-05,
      "loss": 2.6602,
      "step": 110
    },
    {
      "epoch": 0.8299065420560747,
      "grad_norm": 1.2183775901794434,
      "learning_rate": 4.504772348747687e-05,
      "loss": 2.5022,
      "step": 111
    },
    {
      "epoch": 0.8373831775700935,
      "grad_norm": 1.2618286609649658,
      "learning_rate": 4.4225768151520694e-05,
      "loss": 2.4272,
      "step": 112
    },
    {
      "epoch": 0.8448598130841122,
      "grad_norm": 1.3538342714309692,
      "learning_rate": 4.3405391432896555e-05,
      "loss": 2.7235,
      "step": 113
    },
    {
      "epoch": 0.8523364485981308,
      "grad_norm": 1.4312119483947754,
      "learning_rate": 4.2586817614407895e-05,
      "loss": 2.6716,
      "step": 114
    },
    {
      "epoch": 0.8598130841121495,
      "grad_norm": 1.2790806293487549,
      "learning_rate": 4.17702704859633e-05,
      "loss": 2.6267,
      "step": 115
    },
    {
      "epoch": 0.8672897196261682,
      "grad_norm": 1.5556808710098267,
      "learning_rate": 4.095597328339452e-05,
      "loss": 2.7905,
      "step": 116
    },
    {
      "epoch": 0.874766355140187,
      "grad_norm": 1.5428481101989746,
      "learning_rate": 4.0144148627425993e-05,
      "loss": 2.8523,
      "step": 117
    },
    {
      "epoch": 0.8822429906542056,
      "grad_norm": 1.4691976308822632,
      "learning_rate": 3.933501846281267e-05,
      "loss": 2.7614,
      "step": 118
    },
    {
      "epoch": 0.8897196261682243,
      "grad_norm": 1.5811524391174316,
      "learning_rate": 3.852880399766243e-05,
      "loss": 3.0149,
      "step": 119
    },
    {
      "epoch": 0.897196261682243,
      "grad_norm": 1.6296933889389038,
      "learning_rate": 3.772572564296005e-05,
      "loss": 2.7567,
      "step": 120
    },
    {
      "epoch": 0.9046728971962616,
      "grad_norm": 1.7295292615890503,
      "learning_rate": 3.6926002952309016e-05,
      "loss": 2.9281,
      "step": 121
    },
    {
      "epoch": 0.9121495327102803,
      "grad_norm": 3.255584955215454,
      "learning_rate": 3.612985456190778e-05,
      "loss": 2.8509,
      "step": 122
    },
    {
      "epoch": 0.9196261682242991,
      "grad_norm": 2.0937676429748535,
      "learning_rate": 3.533749813077677e-05,
      "loss": 2.7884,
      "step": 123
    },
    {
      "epoch": 0.9271028037383178,
      "grad_norm": 2.0866222381591797,
      "learning_rate": 3.4549150281252636e-05,
      "loss": 2.781,
      "step": 124
    },
    {
      "epoch": 0.9345794392523364,
      "grad_norm": 2.126271963119507,
      "learning_rate": 3.3765026539765834e-05,
      "loss": 2.9911,
      "step": 125
    },
    {
      "epoch": 0.9420560747663551,
      "grad_norm": 2.351102590560913,
      "learning_rate": 3.298534127791785e-05,
      "loss": 3.1141,
      "step": 126
    },
    {
      "epoch": 0.9495327102803738,
      "grad_norm": 3.0210423469543457,
      "learning_rate": 3.221030765387417e-05,
      "loss": 3.3582,
      "step": 127
    },
    {
      "epoch": 0.9570093457943926,
      "grad_norm": 3.0929412841796875,
      "learning_rate": 3.144013755408895e-05,
      "loss": 3.3628,
      "step": 128
    },
    {
      "epoch": 0.9644859813084112,
      "grad_norm": 2.898117780685425,
      "learning_rate": 3.0675041535377405e-05,
      "loss": 3.3368,
      "step": 129
    },
    {
      "epoch": 0.9719626168224299,
      "grad_norm": 3.4047045707702637,
      "learning_rate": 2.991522876735154e-05,
      "loss": 3.5536,
      "step": 130
    },
    {
      "epoch": 0.9794392523364486,
      "grad_norm": 3.896332263946533,
      "learning_rate": 2.916090697523549e-05,
      "loss": 3.6717,
      "step": 131
    },
    {
      "epoch": 0.9869158878504672,
      "grad_norm": 8.250482559204102,
      "learning_rate": 2.8412282383075363e-05,
      "loss": 4.393,
      "step": 132
    },
    {
      "epoch": 0.994392523364486,
      "grad_norm": 5.29136848449707,
      "learning_rate": 2.766955965735968e-05,
      "loss": 2.5479,
      "step": 133
    },
    {
      "epoch": 1.005607476635514,
      "grad_norm": 8.564682006835938,
      "learning_rate": 2.693294185106562e-05,
      "loss": 4.9411,
      "step": 134
    },
    {
      "epoch": 1.0130841121495326,
      "grad_norm": 5.876297950744629,
      "learning_rate": 2.6202630348146324e-05,
      "loss": 2.4473,
      "step": 135
    },
    {
      "epoch": 1.0205607476635514,
      "grad_norm": 4.5796732902526855,
      "learning_rate": 2.547882480847461e-05,
      "loss": 2.6543,
      "step": 136
    },
    {
      "epoch": 1.02803738317757,
      "grad_norm": 4.264084815979004,
      "learning_rate": 2.476172311325783e-05,
      "loss": 2.5728,
      "step": 137
    },
    {
      "epoch": 1.0355140186915888,
      "grad_norm": 3.245776891708374,
      "learning_rate": 2.405152131093926e-05,
      "loss": 2.1302,
      "step": 138
    },
    {
      "epoch": 1.0429906542056075,
      "grad_norm": 2.914738178253174,
      "learning_rate": 2.3348413563600325e-05,
      "loss": 2.0389,
      "step": 139
    },
    {
      "epoch": 1.050467289719626,
      "grad_norm": 2.3504302501678467,
      "learning_rate": 2.2652592093878666e-05,
      "loss": 2.1766,
      "step": 140
    },
    {
      "epoch": 1.0579439252336449,
      "grad_norm": 2.0428102016448975,
      "learning_rate": 2.196424713241637e-05,
      "loss": 1.8922,
      "step": 141
    },
    {
      "epoch": 1.0654205607476634,
      "grad_norm": 1.764786958694458,
      "learning_rate": 2.128356686585282e-05,
      "loss": 2.2562,
      "step": 142
    },
    {
      "epoch": 1.0728971962616822,
      "grad_norm": 1.4716768264770508,
      "learning_rate": 2.061073738537635e-05,
      "loss": 2.1486,
      "step": 143
    },
    {
      "epoch": 1.080373831775701,
      "grad_norm": 1.3650274276733398,
      "learning_rate": 1.9945942635848748e-05,
      "loss": 2.2323,
      "step": 144
    },
    {
      "epoch": 1.0878504672897196,
      "grad_norm": 1.2428630590438843,
      "learning_rate": 1.928936436551661e-05,
      "loss": 2.3782,
      "step": 145
    },
    {
      "epoch": 1.0953271028037384,
      "grad_norm": 1.2568516731262207,
      "learning_rate": 1.8641182076323148e-05,
      "loss": 2.1139,
      "step": 146
    },
    {
      "epoch": 1.102803738317757,
      "grad_norm": 1.243582844734192,
      "learning_rate": 1.800157297483417e-05,
      "loss": 2.0919,
      "step": 147
    },
    {
      "epoch": 1.1102803738317757,
      "grad_norm": 1.1808128356933594,
      "learning_rate": 1.7370711923791567e-05,
      "loss": 2.1847,
      "step": 148
    },
    {
      "epoch": 1.1177570093457945,
      "grad_norm": 1.260305404663086,
      "learning_rate": 1.6748771394307585e-05,
      "loss": 2.2889,
      "step": 149
    },
    {
      "epoch": 1.125233644859813,
      "grad_norm": 1.7860231399536133,
      "learning_rate": 1.6135921418712956e-05,
      "loss": 2.3896,
      "step": 150
    },
    {
      "epoch": 1.125233644859813,
      "eval_loss": 2.7041280269622803,
      "eval_runtime": 21.4376,
      "eval_samples_per_second": 10.542,
      "eval_steps_per_second": 2.659,
      "step": 150
    },
    {
      "epoch": 1.1327102803738318,
      "grad_norm": 1.4049072265625,
      "learning_rate": 1.553232954407171e-05,
      "loss": 2.3502,
      "step": 151
    },
    {
      "epoch": 1.1401869158878504,
      "grad_norm": 1.9700613021850586,
      "learning_rate": 1.4938160786375572e-05,
      "loss": 2.4684,
      "step": 152
    },
    {
      "epoch": 1.1476635514018692,
      "grad_norm": 1.6345374584197998,
      "learning_rate": 1.435357758543015e-05,
      "loss": 2.3376,
      "step": 153
    },
    {
      "epoch": 1.155140186915888,
      "grad_norm": 1.9372590780258179,
      "learning_rate": 1.3778739760445552e-05,
      "loss": 2.4287,
      "step": 154
    },
    {
      "epoch": 1.1626168224299065,
      "grad_norm": 1.86532461643219,
      "learning_rate": 1.3213804466343421e-05,
      "loss": 2.3123,
      "step": 155
    },
    {
      "epoch": 1.1700934579439253,
      "grad_norm": 2.1371116638183594,
      "learning_rate": 1.2658926150792322e-05,
      "loss": 2.5992,
      "step": 156
    },
    {
      "epoch": 1.1775700934579438,
      "grad_norm": 2.4250404834747314,
      "learning_rate": 1.2114256511983274e-05,
      "loss": 2.5268,
      "step": 157
    },
    {
      "epoch": 1.1850467289719626,
      "grad_norm": 2.1564767360687256,
      "learning_rate": 1.157994445715706e-05,
      "loss": 2.2943,
      "step": 158
    },
    {
      "epoch": 1.1925233644859814,
      "grad_norm": 2.2312123775482178,
      "learning_rate": 1.1056136061894384e-05,
      "loss": 2.2666,
      "step": 159
    },
    {
      "epoch": 1.2,
      "grad_norm": 2.733131170272827,
      "learning_rate": 1.0542974530180327e-05,
      "loss": 2.8414,
      "step": 160
    },
    {
      "epoch": 1.2074766355140187,
      "grad_norm": 2.6753244400024414,
      "learning_rate": 1.0040600155253765e-05,
      "loss": 2.5758,
      "step": 161
    },
    {
      "epoch": 1.2149532710280373,
      "grad_norm": 2.994265079498291,
      "learning_rate": 9.549150281252633e-06,
      "loss": 2.9366,
      "step": 162
    },
    {
      "epoch": 1.222429906542056,
      "grad_norm": 3.397216796875,
      "learning_rate": 9.068759265665384e-06,
      "loss": 3.1434,
      "step": 163
    },
    {
      "epoch": 1.2299065420560749,
      "grad_norm": 3.5160207748413086,
      "learning_rate": 8.599558442598998e-06,
      "loss": 3.0147,
      "step": 164
    },
    {
      "epoch": 1.2373831775700934,
      "grad_norm": 3.6264634132385254,
      "learning_rate": 8.141676086873572e-06,
      "loss": 3.3812,
      "step": 165
    },
    {
      "epoch": 1.2448598130841122,
      "grad_norm": 5.191136360168457,
      "learning_rate": 7.695237378953223e-06,
      "loss": 3.7827,
      "step": 166
    },
    {
      "epoch": 1.2523364485981308,
      "grad_norm": 2.7781307697296143,
      "learning_rate": 7.260364370723044e-06,
      "loss": 3.1838,
      "step": 167
    },
    {
      "epoch": 1.2598130841121495,
      "grad_norm": 1.6204148530960083,
      "learning_rate": 6.837175952121306e-06,
      "loss": 2.1856,
      "step": 168
    },
    {
      "epoch": 1.2672897196261683,
      "grad_norm": 1.6356444358825684,
      "learning_rate": 6.425787818636131e-06,
      "loss": 2.2609,
      "step": 169
    },
    {
      "epoch": 1.2747663551401869,
      "grad_norm": 1.655259370803833,
      "learning_rate": 6.026312439675552e-06,
      "loss": 2.1581,
      "step": 170
    },
    {
      "epoch": 1.2822429906542057,
      "grad_norm": 1.5593515634536743,
      "learning_rate": 5.6388590278194096e-06,
      "loss": 1.968,
      "step": 171
    },
    {
      "epoch": 1.2897196261682242,
      "grad_norm": 1.6624542474746704,
      "learning_rate": 5.263533508961827e-06,
      "loss": 2.171,
      "step": 172
    },
    {
      "epoch": 1.297196261682243,
      "grad_norm": 1.402922511100769,
      "learning_rate": 4.900438493352055e-06,
      "loss": 1.9894,
      "step": 173
    },
    {
      "epoch": 1.3046728971962618,
      "grad_norm": 1.5571993589401245,
      "learning_rate": 4.549673247541875e-06,
      "loss": 2.1963,
      "step": 174
    },
    {
      "epoch": 1.3121495327102803,
      "grad_norm": 1.5523626804351807,
      "learning_rate": 4.2113336672471245e-06,
      "loss": 2.2046,
      "step": 175
    },
    {
      "epoch": 1.3196261682242991,
      "grad_norm": 1.363818883895874,
      "learning_rate": 3.885512251130763e-06,
      "loss": 2.0201,
      "step": 176
    },
    {
      "epoch": 1.3271028037383177,
      "grad_norm": 1.460727334022522,
      "learning_rate": 3.5722980755146517e-06,
      "loss": 2.321,
      "step": 177
    },
    {
      "epoch": 1.3345794392523365,
      "grad_norm": 1.5263203382492065,
      "learning_rate": 3.271776770026963e-06,
      "loss": 2.1787,
      "step": 178
    },
    {
      "epoch": 1.3420560747663552,
      "grad_norm": 1.4303926229476929,
      "learning_rate": 2.9840304941919415e-06,
      "loss": 2.1308,
      "step": 179
    },
    {
      "epoch": 1.3495327102803738,
      "grad_norm": 1.602799415588379,
      "learning_rate": 2.7091379149682685e-06,
      "loss": 2.0897,
      "step": 180
    },
    {
      "epoch": 1.3570093457943926,
      "grad_norm": 1.4464118480682373,
      "learning_rate": 2.4471741852423237e-06,
      "loss": 2.1445,
      "step": 181
    },
    {
      "epoch": 1.3644859813084111,
      "grad_norm": 1.4721251726150513,
      "learning_rate": 2.1982109232821178e-06,
      "loss": 2.227,
      "step": 182
    },
    {
      "epoch": 1.37196261682243,
      "grad_norm": 1.4104818105697632,
      "learning_rate": 1.962316193157593e-06,
      "loss": 2.5595,
      "step": 183
    },
    {
      "epoch": 1.3794392523364487,
      "grad_norm": 1.5188437700271606,
      "learning_rate": 1.7395544861325718e-06,
      "loss": 2.4144,
      "step": 184
    },
    {
      "epoch": 1.3869158878504673,
      "grad_norm": 1.4749935865402222,
      "learning_rate": 1.5299867030334814e-06,
      "loss": 2.2084,
      "step": 185
    },
    {
      "epoch": 1.394392523364486,
      "grad_norm": 1.587631106376648,
      "learning_rate": 1.333670137599713e-06,
      "loss": 2.3531,
      "step": 186
    },
    {
      "epoch": 1.4018691588785046,
      "grad_norm": 1.6825957298278809,
      "learning_rate": 1.1506584608200367e-06,
      "loss": 2.282,
      "step": 187
    },
    {
      "epoch": 1.4093457943925234,
      "grad_norm": 1.9881898164749146,
      "learning_rate": 9.810017062595322e-07,
      "loss": 2.4404,
      "step": 188
    },
    {
      "epoch": 1.4168224299065422,
      "grad_norm": 1.9033918380737305,
      "learning_rate": 8.247462563808817e-07,
      "loss": 2.683,
      "step": 189
    },
    {
      "epoch": 1.4242990654205607,
      "grad_norm": 1.6983411312103271,
      "learning_rate": 6.819348298638839e-07,
      "loss": 2.2103,
      "step": 190
    },
    {
      "epoch": 1.4317757009345795,
      "grad_norm": 2.0023140907287598,
      "learning_rate": 5.526064699265753e-07,
      "loss": 2.4267,
      "step": 191
    },
    {
      "epoch": 1.439252336448598,
      "grad_norm": 2.0997490882873535,
      "learning_rate": 4.367965336512403e-07,
      "loss": 2.3739,
      "step": 192
    },
    {
      "epoch": 1.4467289719626168,
      "grad_norm": 2.378817558288574,
      "learning_rate": 3.3453668231809286e-07,
      "loss": 2.3076,
      "step": 193
    },
    {
      "epoch": 1.4542056074766356,
      "grad_norm": 2.561694622039795,
      "learning_rate": 2.458548727494292e-07,
      "loss": 2.7308,
      "step": 194
    },
    {
      "epoch": 1.4616822429906542,
      "grad_norm": 2.7574121952056885,
      "learning_rate": 1.7077534966650766e-07,
      "loss": 3.0258,
      "step": 195
    },
    {
      "epoch": 1.469158878504673,
      "grad_norm": 2.83764386177063,
      "learning_rate": 1.0931863906127327e-07,
      "loss": 2.5581,
      "step": 196
    },
    {
      "epoch": 1.4766355140186915,
      "grad_norm": 3.6413447856903076,
      "learning_rate": 6.150154258476315e-08,
      "loss": 3.0831,
      "step": 197
    },
    {
      "epoch": 1.4841121495327103,
      "grad_norm": 4.1161627769470215,
      "learning_rate": 2.7337132953697554e-08,
      "loss": 3.2026,
      "step": 198
    },
    {
      "epoch": 1.491588785046729,
      "grad_norm": 6.31775426864624,
      "learning_rate": 6.834750376549792e-09,
      "loss": 3.9549,
      "step": 199
    },
    {
      "epoch": 1.4990654205607477,
      "grad_norm": 2.250831365585327,
      "learning_rate": 0.0,
      "loss": 2.6104,
      "step": 200
    },
    {
      "epoch": 1.4990654205607477,
      "eval_loss": 2.652009963989258,
      "eval_runtime": 21.5142,
      "eval_samples_per_second": 10.505,
      "eval_steps_per_second": 2.649,
      "step": 200
    }
  ],
  "logging_steps": 1,
  "max_steps": 200,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 50,
  "stateful_callbacks": {
    "EarlyStoppingCallback": {
      "args": {
        "early_stopping_patience": 5,
        "early_stopping_threshold": 0.0
      },
      "attributes": {
        "early_stopping_patience_counter": 0
      }
    },
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 3.454443806785536e+17,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}