{
  "best_metric": 1.3416213989257812,
  "best_model_checkpoint": "miner_id_24/checkpoint-200",
  "epoch": 0.08025682182985554,
  "eval_steps": 50,
  "global_step": 200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0004012841091492777,
      "grad_norm": 31.025239944458008,
      "learning_rate": 1e-05,
      "loss": 5.7267,
      "step": 1
    },
    {
      "epoch": 0.0004012841091492777,
      "eval_loss": 1.5384316444396973,
      "eval_runtime": 76.5735,
      "eval_samples_per_second": 54.823,
      "eval_steps_per_second": 13.712,
      "step": 1
    },
    {
      "epoch": 0.0008025682182985554,
      "grad_norm": 21.58005714416504,
      "learning_rate": 2e-05,
      "loss": 5.4726,
      "step": 2
    },
    {
      "epoch": 0.001203852327447833,
      "grad_norm": 23.27560043334961,
      "learning_rate": 3e-05,
      "loss": 5.7567,
      "step": 3
    },
    {
      "epoch": 0.0016051364365971107,
      "grad_norm": 20.8821964263916,
      "learning_rate": 4e-05,
      "loss": 5.3938,
      "step": 4
    },
    {
      "epoch": 0.0020064205457463883,
      "grad_norm": 23.100244522094727,
      "learning_rate": 5e-05,
      "loss": 5.1156,
      "step": 5
    },
    {
      "epoch": 0.002407704654895666,
      "grad_norm": 24.256513595581055,
      "learning_rate": 6e-05,
      "loss": 5.575,
      "step": 6
    },
    {
      "epoch": 0.0028089887640449437,
      "grad_norm": 27.6983585357666,
      "learning_rate": 7e-05,
      "loss": 5.512,
      "step": 7
    },
    {
      "epoch": 0.0032102728731942215,
      "grad_norm": 28.489892959594727,
      "learning_rate": 8e-05,
      "loss": 5.6409,
      "step": 8
    },
    {
      "epoch": 0.0036115569823434992,
      "grad_norm": 29.85070037841797,
      "learning_rate": 9e-05,
      "loss": 5.1877,
      "step": 9
    },
    {
      "epoch": 0.0040128410914927765,
      "grad_norm": 39.847591400146484,
      "learning_rate": 0.0001,
      "loss": 5.265,
      "step": 10
    },
    {
      "epoch": 0.004414125200642055,
      "grad_norm": 33.815494537353516,
      "learning_rate": 9.999316524962345e-05,
      "loss": 5.6645,
      "step": 11
    },
    {
      "epoch": 0.004815409309791332,
      "grad_norm": 31.305957794189453,
      "learning_rate": 9.997266286704631e-05,
      "loss": 5.7484,
      "step": 12
    },
    {
      "epoch": 0.00521669341894061,
      "grad_norm": 29.45059585571289,
      "learning_rate": 9.993849845741524e-05,
      "loss": 5.7639,
      "step": 13
    },
    {
      "epoch": 0.0056179775280898875,
      "grad_norm": 41.57351303100586,
      "learning_rate": 9.989068136093873e-05,
      "loss": 5.2827,
      "step": 14
    },
    {
      "epoch": 0.006019261637239166,
      "grad_norm": 32.21366500854492,
      "learning_rate": 9.98292246503335e-05,
      "loss": 5.4855,
      "step": 15
    },
    {
      "epoch": 0.006420545746388443,
      "grad_norm": 29.62386703491211,
      "learning_rate": 9.975414512725057e-05,
      "loss": 5.7484,
      "step": 16
    },
    {
      "epoch": 0.006821829855537721,
      "grad_norm": 26.24283218383789,
      "learning_rate": 9.966546331768191e-05,
      "loss": 5.5532,
      "step": 17
    },
    {
      "epoch": 0.0072231139646869984,
      "grad_norm": 26.627737045288086,
      "learning_rate": 9.956320346634876e-05,
      "loss": 5.2349,
      "step": 18
    },
    {
      "epoch": 0.007624398073836276,
      "grad_norm": 27.147537231445312,
      "learning_rate": 9.944739353007344e-05,
      "loss": 5.503,
      "step": 19
    },
    {
      "epoch": 0.008025682182985553,
      "grad_norm": 28.37727165222168,
      "learning_rate": 9.931806517013612e-05,
      "loss": 5.4497,
      "step": 20
    },
    {
      "epoch": 0.008426966292134831,
      "grad_norm": 30.155452728271484,
      "learning_rate": 9.917525374361912e-05,
      "loss": 5.3802,
      "step": 21
    },
    {
      "epoch": 0.00882825040128411,
      "grad_norm": 23.588979721069336,
      "learning_rate": 9.901899829374047e-05,
      "loss": 5.5548,
      "step": 22
    },
    {
      "epoch": 0.009229534510433388,
      "grad_norm": 24.670040130615234,
      "learning_rate": 9.884934153917997e-05,
      "loss": 5.8427,
      "step": 23
    },
    {
      "epoch": 0.009630818619582664,
      "grad_norm": 24.20445442199707,
      "learning_rate": 9.86663298624003e-05,
      "loss": 6.0211,
      "step": 24
    },
    {
      "epoch": 0.010032102728731942,
      "grad_norm": 24.382766723632812,
      "learning_rate": 9.847001329696653e-05,
      "loss": 5.5022,
      "step": 25
    },
    {
      "epoch": 0.01043338683788122,
      "grad_norm": 25.791048049926758,
      "learning_rate": 9.826044551386744e-05,
      "loss": 5.9042,
      "step": 26
    },
    {
      "epoch": 0.010834670947030497,
      "grad_norm": 26.141822814941406,
      "learning_rate": 9.803768380684242e-05,
      "loss": 5.8266,
      "step": 27
    },
    {
      "epoch": 0.011235955056179775,
      "grad_norm": 29.682100296020508,
      "learning_rate": 9.780178907671789e-05,
      "loss": 5.6217,
      "step": 28
    },
    {
      "epoch": 0.011637239165329053,
      "grad_norm": 28.040651321411133,
      "learning_rate": 9.755282581475769e-05,
      "loss": 6.1496,
      "step": 29
    },
    {
      "epoch": 0.012038523274478331,
      "grad_norm": 34.46293258666992,
      "learning_rate": 9.729086208503174e-05,
      "loss": 5.755,
      "step": 30
    },
    {
      "epoch": 0.012439807383627608,
      "grad_norm": 31.138093948364258,
      "learning_rate": 9.701596950580806e-05,
      "loss": 5.6661,
      "step": 31
    },
    {
      "epoch": 0.012841091492776886,
      "grad_norm": 30.499282836914062,
      "learning_rate": 9.672822322997305e-05,
      "loss": 5.6561,
      "step": 32
    },
    {
      "epoch": 0.013242375601926164,
      "grad_norm": 34.31699752807617,
      "learning_rate": 9.642770192448536e-05,
      "loss": 5.5981,
      "step": 33
    },
    {
      "epoch": 0.013643659711075442,
      "grad_norm": 29.02855110168457,
      "learning_rate": 9.611448774886924e-05,
      "loss": 5.8391,
      "step": 34
    },
    {
      "epoch": 0.014044943820224719,
      "grad_norm": 35.29312515258789,
      "learning_rate": 9.578866633275288e-05,
      "loss": 6.2131,
      "step": 35
    },
    {
      "epoch": 0.014446227929373997,
      "grad_norm": 30.347909927368164,
      "learning_rate": 9.545032675245813e-05,
      "loss": 6.4159,
      "step": 36
    },
    {
      "epoch": 0.014847512038523275,
      "grad_norm": 30.43315887451172,
      "learning_rate": 9.509956150664796e-05,
      "loss": 6.4443,
      "step": 37
    },
    {
      "epoch": 0.015248796147672551,
      "grad_norm": 29.859960556030273,
      "learning_rate": 9.473646649103818e-05,
      "loss": 6.1487,
      "step": 38
    },
    {
      "epoch": 0.01565008025682183,
      "grad_norm": 25.03530502319336,
      "learning_rate": 9.43611409721806e-05,
      "loss": 6.5191,
      "step": 39
    },
    {
      "epoch": 0.016051364365971106,
      "grad_norm": 26.410856246948242,
      "learning_rate": 9.397368756032445e-05,
      "loss": 6.4339,
      "step": 40
    },
    {
      "epoch": 0.016452648475120384,
      "grad_norm": 26.30873680114746,
      "learning_rate": 9.357421218136386e-05,
      "loss": 6.6504,
      "step": 41
    },
    {
      "epoch": 0.016853932584269662,
      "grad_norm": 25.05800437927246,
      "learning_rate": 9.316282404787871e-05,
      "loss": 6.3494,
      "step": 42
    },
    {
      "epoch": 0.01725521669341894,
      "grad_norm": 26.615232467651367,
      "learning_rate": 9.273963562927695e-05,
      "loss": 6.3326,
      "step": 43
    },
    {
      "epoch": 0.01765650080256822,
      "grad_norm": 27.76177978515625,
      "learning_rate": 9.230476262104677e-05,
      "loss": 6.1111,
      "step": 44
    },
    {
      "epoch": 0.018057784911717497,
      "grad_norm": 29.176189422607422,
      "learning_rate": 9.185832391312644e-05,
      "loss": 6.2874,
      "step": 45
    },
    {
      "epoch": 0.018459069020866775,
      "grad_norm": 36.66054916381836,
      "learning_rate": 9.140044155740101e-05,
      "loss": 5.6365,
      "step": 46
    },
    {
      "epoch": 0.01886035313001605,
      "grad_norm": 31.717748641967773,
      "learning_rate": 9.093124073433463e-05,
      "loss": 6.334,
      "step": 47
    },
    {
      "epoch": 0.019261637239165328,
      "grad_norm": 46.143096923828125,
      "learning_rate": 9.045084971874738e-05,
      "loss": 5.9939,
      "step": 48
    },
    {
      "epoch": 0.019662921348314606,
      "grad_norm": 39.511863708496094,
      "learning_rate": 8.995939984474624e-05,
      "loss": 5.9257,
      "step": 49
    },
    {
      "epoch": 0.020064205457463884,
      "grad_norm": 39.54106903076172,
      "learning_rate": 8.945702546981969e-05,
      "loss": 6.1518,
      "step": 50
    },
    {
      "epoch": 0.020064205457463884,
      "eval_loss": 1.4055343866348267,
      "eval_runtime": 76.6741,
      "eval_samples_per_second": 54.751,
      "eval_steps_per_second": 13.694,
      "step": 50
    },
    {
      "epoch": 0.020465489566613163,
      "grad_norm": 43.721588134765625,
      "learning_rate": 8.894386393810563e-05,
      "loss": 4.8454,
      "step": 51
    },
    {
      "epoch": 0.02086677367576244,
      "grad_norm": 42.092079162597656,
      "learning_rate": 8.842005554284296e-05,
      "loss": 4.9015,
      "step": 52
    },
    {
      "epoch": 0.02126805778491172,
      "grad_norm": 40.44691467285156,
      "learning_rate": 8.788574348801675e-05,
      "loss": 5.0475,
      "step": 53
    },
    {
      "epoch": 0.021669341894060994,
      "grad_norm": 40.7920036315918,
      "learning_rate": 8.73410738492077e-05,
      "loss": 5.1365,
      "step": 54
    },
    {
      "epoch": 0.022070626003210272,
      "grad_norm": 38.79573059082031,
      "learning_rate": 8.678619553365659e-05,
      "loss": 4.9113,
      "step": 55
    },
    {
      "epoch": 0.02247191011235955,
      "grad_norm": 37.632667541503906,
      "learning_rate": 8.622126023955446e-05,
      "loss": 5.0678,
      "step": 56
    },
    {
      "epoch": 0.022873194221508828,
      "grad_norm": 35.89990234375,
      "learning_rate": 8.564642241456986e-05,
      "loss": 5.0636,
      "step": 57
    },
    {
      "epoch": 0.023274478330658106,
      "grad_norm": 36.06848907470703,
      "learning_rate": 8.506183921362443e-05,
      "loss": 5.0778,
      "step": 58
    },
    {
      "epoch": 0.023675762439807384,
      "grad_norm": 44.013309478759766,
      "learning_rate": 8.44676704559283e-05,
      "loss": 5.0021,
      "step": 59
    },
    {
      "epoch": 0.024077046548956663,
      "grad_norm": 37.43661117553711,
      "learning_rate": 8.386407858128706e-05,
      "loss": 5.1293,
      "step": 60
    },
    {
      "epoch": 0.024478330658105937,
      "grad_norm": 45.96494674682617,
      "learning_rate": 8.32512286056924e-05,
      "loss": 5.3998,
      "step": 61
    },
    {
      "epoch": 0.024879614767255216,
      "grad_norm": 33.3837890625,
      "learning_rate": 8.262928807620843e-05,
      "loss": 5.2209,
      "step": 62
    },
    {
      "epoch": 0.025280898876404494,
      "grad_norm": 38.15312576293945,
      "learning_rate": 8.199842702516583e-05,
      "loss": 5.39,
      "step": 63
    },
    {
      "epoch": 0.025682182985553772,
      "grad_norm": 43.511070251464844,
      "learning_rate": 8.135881792367686e-05,
      "loss": 5.1686,
      "step": 64
    },
    {
      "epoch": 0.02608346709470305,
      "grad_norm": 39.421226501464844,
      "learning_rate": 8.07106356344834e-05,
      "loss": 5.5708,
      "step": 65
    },
    {
      "epoch": 0.026484751203852328,
      "grad_norm": 67.12967681884766,
      "learning_rate": 8.005405736415126e-05,
      "loss": 5.2274,
      "step": 66
    },
    {
      "epoch": 0.026886035313001606,
      "grad_norm": 42.88607406616211,
      "learning_rate": 7.938926261462366e-05,
      "loss": 5.2944,
      "step": 67
    },
    {
      "epoch": 0.027287319422150885,
      "grad_norm": 51.06602478027344,
      "learning_rate": 7.871643313414718e-05,
      "loss": 5.5377,
      "step": 68
    },
    {
      "epoch": 0.02768860353130016,
      "grad_norm": 42.7625732421875,
      "learning_rate": 7.803575286758364e-05,
      "loss": 5.8507,
      "step": 69
    },
    {
      "epoch": 0.028089887640449437,
      "grad_norm": 52.89863204956055,
      "learning_rate": 7.734740790612136e-05,
      "loss": 5.6088,
      "step": 70
    },
    {
      "epoch": 0.028491171749598716,
      "grad_norm": 43.04888153076172,
      "learning_rate": 7.66515864363997e-05,
      "loss": 5.3938,
      "step": 71
    },
    {
      "epoch": 0.028892455858747994,
      "grad_norm": 35.296104431152344,
      "learning_rate": 7.594847868906076e-05,
      "loss": 5.5896,
      "step": 72
    },
    {
      "epoch": 0.029293739967897272,
      "grad_norm": 34.10617446899414,
      "learning_rate": 7.52382768867422e-05,
      "loss": 5.4214,
      "step": 73
    },
    {
      "epoch": 0.02969502407704655,
      "grad_norm": 37.03001022338867,
      "learning_rate": 7.452117519152542e-05,
      "loss": 5.8333,
      "step": 74
    },
    {
      "epoch": 0.03009630818619583,
      "grad_norm": 39.194862365722656,
      "learning_rate": 7.379736965185368e-05,
      "loss": 5.5827,
      "step": 75
    },
    {
      "epoch": 0.030497592295345103,
      "grad_norm": 35.79296112060547,
      "learning_rate": 7.30670581489344e-05,
      "loss": 5.6704,
      "step": 76
    },
    {
      "epoch": 0.03089887640449438,
      "grad_norm": 40.25273513793945,
      "learning_rate": 7.233044034264034e-05,
      "loss": 5.6064,
      "step": 77
    },
    {
      "epoch": 0.03130016051364366,
      "grad_norm": 38.52794647216797,
      "learning_rate": 7.158771761692464e-05,
      "loss": 5.7733,
      "step": 78
    },
    {
      "epoch": 0.03170144462279294,
      "grad_norm": 38.72677230834961,
      "learning_rate": 7.083909302476453e-05,
      "loss": 5.631,
      "step": 79
    },
    {
      "epoch": 0.03210272873194221,
      "grad_norm": 37.08574295043945,
      "learning_rate": 7.008477123264848e-05,
      "loss": 5.5958,
      "step": 80
    },
    {
      "epoch": 0.032504012841091494,
      "grad_norm": 38.327022552490234,
      "learning_rate": 6.932495846462261e-05,
      "loss": 5.7647,
      "step": 81
    },
    {
      "epoch": 0.03290529695024077,
      "grad_norm": 41.12373352050781,
      "learning_rate": 6.855986244591104e-05,
      "loss": 5.7133,
      "step": 82
    },
    {
      "epoch": 0.03330658105939005,
      "grad_norm": 36.328975677490234,
      "learning_rate": 6.778969234612584e-05,
      "loss": 5.766,
      "step": 83
    },
    {
      "epoch": 0.033707865168539325,
      "grad_norm": 35.401458740234375,
      "learning_rate": 6.701465872208216e-05,
      "loss": 5.8813,
      "step": 84
    },
    {
      "epoch": 0.03410914927768861,
      "grad_norm": 32.8640022277832,
      "learning_rate": 6.623497346023418e-05,
      "loss": 5.9751,
      "step": 85
    },
    {
      "epoch": 0.03451043338683788,
      "grad_norm": 35.46437454223633,
      "learning_rate": 6.545084971874738e-05,
      "loss": 6.6593,
      "step": 86
    },
    {
      "epoch": 0.034911717495987156,
      "grad_norm": 41.714378356933594,
      "learning_rate": 6.466250186922325e-05,
      "loss": 5.9974,
      "step": 87
    },
    {
      "epoch": 0.03531300160513644,
      "grad_norm": 42.388187408447266,
      "learning_rate": 6.387014543809223e-05,
      "loss": 6.3697,
      "step": 88
    },
    {
      "epoch": 0.03571428571428571,
      "grad_norm": 36.15264892578125,
      "learning_rate": 6.307399704769099e-05,
      "loss": 6.1781,
      "step": 89
    },
    {
      "epoch": 0.036115569823434994,
      "grad_norm": 35.308414459228516,
      "learning_rate": 6.227427435703997e-05,
      "loss": 6.2428,
      "step": 90
    },
    {
      "epoch": 0.03651685393258427,
      "grad_norm": 38.426918029785156,
      "learning_rate": 6.147119600233758e-05,
      "loss": 6.528,
      "step": 91
    },
    {
      "epoch": 0.03691813804173355,
      "grad_norm": 43.396968841552734,
      "learning_rate": 6.066498153718735e-05,
      "loss": 6.3805,
      "step": 92
    },
    {
      "epoch": 0.037319422150882825,
      "grad_norm": 42.94819259643555,
      "learning_rate": 5.985585137257401e-05,
      "loss": 5.8774,
      "step": 93
    },
    {
      "epoch": 0.0377207062600321,
      "grad_norm": 74.18196105957031,
      "learning_rate": 5.90440267166055e-05,
      "loss": 6.0958,
      "step": 94
    },
    {
      "epoch": 0.03812199036918138,
      "grad_norm": 37.53769302368164,
      "learning_rate": 5.8229729514036705e-05,
      "loss": 5.9197,
      "step": 95
    },
    {
      "epoch": 0.038523274478330656,
      "grad_norm": 44.90113830566406,
      "learning_rate": 5.74131823855921e-05,
      "loss": 6.2594,
      "step": 96
    },
    {
      "epoch": 0.03892455858747994,
      "grad_norm": 40.159690856933594,
      "learning_rate": 5.6594608567103456e-05,
      "loss": 6.1026,
      "step": 97
    },
    {
      "epoch": 0.03932584269662921,
      "grad_norm": 37.56284713745117,
      "learning_rate": 5.577423184847932e-05,
      "loss": 6.1867,
      "step": 98
    },
    {
      "epoch": 0.039727126805778494,
      "grad_norm": 43.6816520690918,
      "learning_rate": 5.495227651252315e-05,
      "loss": 6.0633,
      "step": 99
    },
    {
      "epoch": 0.04012841091492777,
      "grad_norm": 45.78097152709961,
      "learning_rate": 5.4128967273616625e-05,
      "loss": 5.6578,
      "step": 100
    },
    {
      "epoch": 0.04012841091492777,
      "eval_loss": 1.3837623596191406,
      "eval_runtime": 76.8022,
      "eval_samples_per_second": 54.66,
      "eval_steps_per_second": 13.671,
      "step": 100
    },
    {
      "epoch": 0.04052969502407704,
      "grad_norm": 38.71050262451172,
      "learning_rate": 5.330452921628497e-05,
      "loss": 4.3026,
      "step": 101
    },
    {
      "epoch": 0.040930979133226325,
      "grad_norm": 35.11087417602539,
      "learning_rate": 5.247918773366112e-05,
      "loss": 4.8271,
      "step": 102
    },
    {
      "epoch": 0.0413322632423756,
      "grad_norm": 31.431264877319336,
      "learning_rate": 5.165316846586541e-05,
      "loss": 4.8693,
      "step": 103
    },
    {
      "epoch": 0.04173354735152488,
      "grad_norm": 27.17133140563965,
      "learning_rate": 5.0826697238317935e-05,
      "loss": 4.5891,
      "step": 104
    },
    {
      "epoch": 0.042134831460674156,
      "grad_norm": 28.50444984436035,
      "learning_rate": 5e-05,
      "loss": 4.6927,
      "step": 105
    },
    {
      "epoch": 0.04253611556982344,
      "grad_norm": 30.044137954711914,
      "learning_rate": 4.917330276168208e-05,
      "loss": 4.9476,
      "step": 106
    },
    {
      "epoch": 0.04293739967897271,
      "grad_norm": 29.493112564086914,
      "learning_rate": 4.834683153413459e-05,
      "loss": 4.9597,
      "step": 107
    },
    {
      "epoch": 0.04333868378812199,
      "grad_norm": 31.95983123779297,
      "learning_rate": 4.7520812266338885e-05,
      "loss": 4.8486,
      "step": 108
    },
    {
      "epoch": 0.04373996789727127,
      "grad_norm": 29.99119758605957,
      "learning_rate": 4.669547078371504e-05,
      "loss": 5.1202,
      "step": 109
    },
    {
      "epoch": 0.044141252006420544,
      "grad_norm": 30.794391632080078,
      "learning_rate": 4.5871032726383386e-05,
      "loss": 5.0138,
      "step": 110
    },
    {
      "epoch": 0.044542536115569825,
      "grad_norm": 30.934053421020508,
      "learning_rate": 4.504772348747687e-05,
      "loss": 5.0704,
      "step": 111
    },
    {
      "epoch": 0.0449438202247191,
      "grad_norm": 25.249662399291992,
      "learning_rate": 4.4225768151520694e-05,
      "loss": 5.4297,
      "step": 112
    },
    {
      "epoch": 0.04534510433386838,
      "grad_norm": 34.12894058227539,
      "learning_rate": 4.3405391432896555e-05,
      "loss": 5.3021,
      "step": 113
    },
    {
      "epoch": 0.045746388443017656,
      "grad_norm": 30.923583984375,
      "learning_rate": 4.2586817614407895e-05,
      "loss": 5.4493,
      "step": 114
    },
    {
      "epoch": 0.04614767255216693,
      "grad_norm": 27.576171875,
      "learning_rate": 4.17702704859633e-05,
      "loss": 5.4067,
      "step": 115
    },
    {
      "epoch": 0.04654895666131621,
      "grad_norm": 33.756290435791016,
      "learning_rate": 4.095597328339452e-05,
      "loss": 5.3337,
      "step": 116
    },
    {
      "epoch": 0.04695024077046549,
      "grad_norm": 39.03615951538086,
      "learning_rate": 4.0144148627425993e-05,
      "loss": 5.2089,
      "step": 117
    },
    {
      "epoch": 0.04735152487961477,
      "grad_norm": 38.68584060668945,
      "learning_rate": 3.933501846281267e-05,
      "loss": 5.1072,
      "step": 118
    },
    {
      "epoch": 0.047752808988764044,
      "grad_norm": 39.365177154541016,
      "learning_rate": 3.852880399766243e-05,
      "loss": 5.1986,
      "step": 119
    },
    {
      "epoch": 0.048154093097913325,
      "grad_norm": 39.52337646484375,
      "learning_rate": 3.772572564296005e-05,
      "loss": 5.3352,
      "step": 120
    },
    {
      "epoch": 0.0485553772070626,
      "grad_norm": 73.31407165527344,
      "learning_rate": 3.6926002952309016e-05,
      "loss": 5.2127,
      "step": 121
    },
    {
      "epoch": 0.048956661316211875,
      "grad_norm": 39.27695846557617,
      "learning_rate": 3.612985456190778e-05,
      "loss": 5.4688,
      "step": 122
    },
    {
      "epoch": 0.049357945425361156,
      "grad_norm": 31.94614028930664,
      "learning_rate": 3.533749813077677e-05,
      "loss": 5.1711,
      "step": 123
    },
    {
      "epoch": 0.04975922953451043,
      "grad_norm": 50.28776931762695,
      "learning_rate": 3.4549150281252636e-05,
      "loss": 5.6199,
      "step": 124
    },
    {
      "epoch": 0.05016051364365971,
      "grad_norm": 30.917068481445312,
      "learning_rate": 3.3765026539765834e-05,
      "loss": 5.4583,
      "step": 125
    },
    {
      "epoch": 0.05056179775280899,
      "grad_norm": 47.983707427978516,
      "learning_rate": 3.298534127791785e-05,
      "loss": 5.4741,
      "step": 126
    },
    {
      "epoch": 0.05096308186195827,
      "grad_norm": 37.81007385253906,
      "learning_rate": 3.221030765387417e-05,
      "loss": 5.3603,
      "step": 127
    },
    {
      "epoch": 0.051364365971107544,
      "grad_norm": 38.755558013916016,
      "learning_rate": 3.144013755408895e-05,
      "loss": 5.2369,
      "step": 128
    },
    {
      "epoch": 0.05176565008025682,
      "grad_norm": 33.99113464355469,
      "learning_rate": 3.0675041535377405e-05,
      "loss": 5.7339,
      "step": 129
    },
    {
      "epoch": 0.0521669341894061,
      "grad_norm": 28.21030044555664,
      "learning_rate": 2.991522876735154e-05,
      "loss": 5.8732,
      "step": 130
    },
    {
      "epoch": 0.052568218298555375,
      "grad_norm": 34.20122146606445,
      "learning_rate": 2.916090697523549e-05,
      "loss": 5.6373,
      "step": 131
    },
    {
      "epoch": 0.052969502407704656,
      "grad_norm": 23.36191177368164,
      "learning_rate": 2.8412282383075363e-05,
      "loss": 5.5812,
      "step": 132
    },
    {
      "epoch": 0.05337078651685393,
      "grad_norm": 22.0201473236084,
      "learning_rate": 2.766955965735968e-05,
      "loss": 6.0476,
      "step": 133
    },
    {
      "epoch": 0.05377207062600321,
      "grad_norm": 33.709163665771484,
      "learning_rate": 2.693294185106562e-05,
      "loss": 5.8909,
      "step": 134
    },
    {
      "epoch": 0.05417335473515249,
      "grad_norm": 29.467906951904297,
      "learning_rate": 2.6202630348146324e-05,
      "loss": 5.4756,
      "step": 135
    },
    {
      "epoch": 0.05457463884430177,
      "grad_norm": 40.85588836669922,
      "learning_rate": 2.547882480847461e-05,
      "loss": 5.993,
      "step": 136
    },
    {
      "epoch": 0.054975922953451044,
      "grad_norm": 24.37715721130371,
      "learning_rate": 2.476172311325783e-05,
      "loss": 6.2587,
      "step": 137
    },
    {
      "epoch": 0.05537720706260032,
      "grad_norm": 22.168224334716797,
      "learning_rate": 2.405152131093926e-05,
      "loss": 6.3656,
      "step": 138
    },
    {
      "epoch": 0.0557784911717496,
      "grad_norm": 21.486522674560547,
      "learning_rate": 2.3348413563600325e-05,
      "loss": 6.3176,
      "step": 139
    },
    {
      "epoch": 0.056179775280898875,
      "grad_norm": 25.41644287109375,
      "learning_rate": 2.2652592093878666e-05,
      "loss": 6.5618,
      "step": 140
    },
    {
      "epoch": 0.056581059390048156,
      "grad_norm": 22.89642906188965,
      "learning_rate": 2.196424713241637e-05,
      "loss": 6.7259,
      "step": 141
    },
    {
      "epoch": 0.05698234349919743,
      "grad_norm": 23.43470573425293,
      "learning_rate": 2.128356686585282e-05,
      "loss": 5.9541,
      "step": 142
    },
    {
      "epoch": 0.05738362760834671,
      "grad_norm": 30.06966209411621,
      "learning_rate": 2.061073738537635e-05,
      "loss": 5.8297,
      "step": 143
    },
    {
      "epoch": 0.05778491171749599,
      "grad_norm": 29.469423294067383,
      "learning_rate": 1.9945942635848748e-05,
      "loss": 5.5952,
      "step": 144
    },
    {
      "epoch": 0.05818619582664526,
      "grad_norm": 28.6286678314209,
      "learning_rate": 1.928936436551661e-05,
      "loss": 5.9743,
      "step": 145
    },
    {
      "epoch": 0.058587479935794544,
      "grad_norm": 24.507835388183594,
      "learning_rate": 1.8641182076323148e-05,
      "loss": 5.534,
      "step": 146
    },
    {
      "epoch": 0.05898876404494382,
      "grad_norm": 28.478206634521484,
      "learning_rate": 1.800157297483417e-05,
      "loss": 6.1821,
      "step": 147
    },
    {
      "epoch": 0.0593900481540931,
      "grad_norm": 26.507993698120117,
      "learning_rate": 1.7370711923791567e-05,
      "loss": 5.6764,
      "step": 148
    },
    {
      "epoch": 0.059791332263242375,
      "grad_norm": 45.15052032470703,
      "learning_rate": 1.6748771394307585e-05,
      "loss": 5.7563,
      "step": 149
    },
    {
      "epoch": 0.06019261637239166,
      "grad_norm": 32.099510192871094,
      "learning_rate": 1.6135921418712956e-05,
      "loss": 5.6664,
      "step": 150
    },
    {
      "epoch": 0.06019261637239166,
      "eval_loss": 1.3496063947677612,
      "eval_runtime": 76.8392,
      "eval_samples_per_second": 54.634,
      "eval_steps_per_second": 13.665,
      "step": 150
    },
    {
      "epoch": 0.06059390048154093,
      "grad_norm": 28.771425247192383,
      "learning_rate": 1.553232954407171e-05,
      "loss": 4.5297,
      "step": 151
    },
    {
      "epoch": 0.060995184590690206,
      "grad_norm": 24.234621047973633,
      "learning_rate": 1.4938160786375572e-05,
      "loss": 4.8205,
      "step": 152
    },
    {
      "epoch": 0.06139646869983949,
      "grad_norm": 31.92659568786621,
      "learning_rate": 1.435357758543015e-05,
      "loss": 5.2605,
      "step": 153
    },
    {
      "epoch": 0.06179775280898876,
      "grad_norm": 31.136581420898438,
      "learning_rate": 1.3778739760445552e-05,
      "loss": 4.8474,
      "step": 154
    },
    {
      "epoch": 0.062199036918138044,
      "grad_norm": 26.611351013183594,
      "learning_rate": 1.3213804466343421e-05,
      "loss": 4.8461,
      "step": 155
    },
    {
      "epoch": 0.06260032102728733,
      "grad_norm": 25.387786865234375,
      "learning_rate": 1.2658926150792322e-05,
      "loss": 4.7304,
      "step": 156
    },
    {
      "epoch": 0.0630016051364366,
      "grad_norm": 31.357276916503906,
      "learning_rate": 1.2114256511983274e-05,
      "loss": 5.0579,
      "step": 157
    },
    {
      "epoch": 0.06340288924558588,
      "grad_norm": 24.882827758789062,
      "learning_rate": 1.157994445715706e-05,
      "loss": 4.6774,
      "step": 158
    },
    {
      "epoch": 0.06380417335473515,
      "grad_norm": 29.332609176635742,
      "learning_rate": 1.1056136061894384e-05,
      "loss": 5.0525,
      "step": 159
    },
    {
      "epoch": 0.06420545746388442,
      "grad_norm": 24.775238037109375,
      "learning_rate": 1.0542974530180327e-05,
      "loss": 4.9378,
      "step": 160
    },
    {
      "epoch": 0.06460674157303371,
      "grad_norm": 26.499711990356445,
      "learning_rate": 1.0040600155253765e-05,
      "loss": 4.7736,
      "step": 161
    },
    {
      "epoch": 0.06500802568218299,
      "grad_norm": 28.938749313354492,
      "learning_rate": 9.549150281252633e-06,
      "loss": 5.2366,
      "step": 162
    },
    {
      "epoch": 0.06540930979133226,
      "grad_norm": 24.08362579345703,
      "learning_rate": 9.068759265665384e-06,
      "loss": 5.1181,
      "step": 163
    },
    {
      "epoch": 0.06581059390048154,
      "grad_norm": 23.927825927734375,
      "learning_rate": 8.599558442598998e-06,
      "loss": 5.0338,
      "step": 164
    },
    {
      "epoch": 0.06621187800963081,
      "grad_norm": 25.204330444335938,
      "learning_rate": 8.141676086873572e-06,
      "loss": 5.4492,
      "step": 165
    },
    {
      "epoch": 0.0666131621187801,
      "grad_norm": 26.78458023071289,
      "learning_rate": 7.695237378953223e-06,
      "loss": 5.1372,
      "step": 166
    },
    {
      "epoch": 0.06701444622792938,
      "grad_norm": 25.849794387817383,
      "learning_rate": 7.260364370723044e-06,
      "loss": 5.0265,
      "step": 167
    },
    {
      "epoch": 0.06741573033707865,
      "grad_norm": 22.05133056640625,
      "learning_rate": 6.837175952121306e-06,
      "loss": 5.0657,
      "step": 168
    },
    {
      "epoch": 0.06781701444622792,
      "grad_norm": 30.659290313720703,
      "learning_rate": 6.425787818636131e-06,
      "loss": 5.4726,
      "step": 169
    },
    {
      "epoch": 0.06821829855537721,
      "grad_norm": 33.32087326049805,
      "learning_rate": 6.026312439675552e-06,
      "loss": 4.9849,
      "step": 170
    },
    {
      "epoch": 0.06861958266452649,
      "grad_norm": 26.374847412109375,
      "learning_rate": 5.6388590278194096e-06,
      "loss": 5.6218,
      "step": 171
    },
    {
      "epoch": 0.06902086677367576,
      "grad_norm": 24.329620361328125,
      "learning_rate": 5.263533508961827e-06,
      "loss": 5.2738,
      "step": 172
    },
    {
      "epoch": 0.06942215088282504,
      "grad_norm": 28.822603225708008,
      "learning_rate": 4.900438493352055e-06,
      "loss": 5.2804,
      "step": 173
    },
    {
      "epoch": 0.06982343499197431,
      "grad_norm": 31.650981903076172,
      "learning_rate": 4.549673247541875e-06,
      "loss": 5.6559,
      "step": 174
    },
    {
      "epoch": 0.0702247191011236,
      "grad_norm": 24.61418914794922,
      "learning_rate": 4.2113336672471245e-06,
      "loss": 5.3051,
      "step": 175
    },
    {
      "epoch": 0.07062600321027288,
      "grad_norm": 40.52798080444336,
      "learning_rate": 3.885512251130763e-06,
      "loss": 5.6262,
      "step": 176
    },
    {
      "epoch": 0.07102728731942215,
      "grad_norm": 29.653003692626953,
      "learning_rate": 3.5722980755146517e-06,
      "loss": 5.4403,
      "step": 177
    },
    {
      "epoch": 0.07142857142857142,
      "grad_norm": 30.37826919555664,
      "learning_rate": 3.271776770026963e-06,
      "loss": 5.4099,
      "step": 178
    },
    {
      "epoch": 0.07182985553772071,
      "grad_norm": 35.208473205566406,
      "learning_rate": 2.9840304941919415e-06,
      "loss": 5.3701,
      "step": 179
    },
    {
      "epoch": 0.07223113964686999,
      "grad_norm": 32.27305221557617,
      "learning_rate": 2.7091379149682685e-06,
      "loss": 5.5316,
      "step": 180
    },
    {
      "epoch": 0.07263242375601926,
      "grad_norm": 50.74858856201172,
      "learning_rate": 2.4471741852423237e-06,
      "loss": 5.6867,
      "step": 181
    },
    {
      "epoch": 0.07303370786516854,
      "grad_norm": 24.291109085083008,
      "learning_rate": 2.1982109232821178e-06,
      "loss": 5.4605,
      "step": 182
    },
    {
      "epoch": 0.07343499197431781,
      "grad_norm": 42.08782196044922,
      "learning_rate": 1.962316193157593e-06,
      "loss": 5.7305,
      "step": 183
    },
    {
      "epoch": 0.0738362760834671,
      "grad_norm": 29.564800262451172,
      "learning_rate": 1.7395544861325718e-06,
      "loss": 5.6542,
      "step": 184
    },
    {
      "epoch": 0.07423756019261638,
      "grad_norm": 25.039794921875,
      "learning_rate": 1.5299867030334814e-06,
      "loss": 5.7802,
      "step": 185
    },
    {
      "epoch": 0.07463884430176565,
      "grad_norm": 23.65619659423828,
      "learning_rate": 1.333670137599713e-06,
      "loss": 5.6438,
      "step": 186
    },
    {
      "epoch": 0.07504012841091492,
      "grad_norm": 24.868261337280273,
      "learning_rate": 1.1506584608200367e-06,
      "loss": 5.8165,
      "step": 187
    },
    {
      "epoch": 0.0754414125200642,
      "grad_norm": 20.64823341369629,
      "learning_rate": 9.810017062595322e-07,
      "loss": 6.2111,
      "step": 188
    },
    {
      "epoch": 0.07584269662921349,
      "grad_norm": 24.644554138183594,
      "learning_rate": 8.247462563808817e-07,
      "loss": 6.5188,
      "step": 189
    },
    {
      "epoch": 0.07624398073836276,
      "grad_norm": 35.02384567260742,
      "learning_rate": 6.819348298638839e-07,
      "loss": 6.5091,
      "step": 190
    },
    {
      "epoch": 0.07664526484751204,
      "grad_norm": 23.783903121948242,
      "learning_rate": 5.526064699265753e-07,
      "loss": 6.3317,
      "step": 191
    },
    {
      "epoch": 0.07704654895666131,
      "grad_norm": 26.307065963745117,
      "learning_rate": 4.367965336512403e-07,
      "loss": 5.9314,
      "step": 192
    },
    {
      "epoch": 0.0774478330658106,
      "grad_norm": 28.82275390625,
      "learning_rate": 3.3453668231809286e-07,
      "loss": 5.5426,
      "step": 193
    },
    {
      "epoch": 0.07784911717495988,
      "grad_norm": 31.975770950317383,
      "learning_rate": 2.458548727494292e-07,
      "loss": 5.5197,
      "step": 194
    },
    {
      "epoch": 0.07825040128410915,
      "grad_norm": 27.233463287353516,
      "learning_rate": 1.7077534966650766e-07,
      "loss": 5.8527,
      "step": 195
    },
    {
      "epoch": 0.07865168539325842,
      "grad_norm": 26.5628719329834,
      "learning_rate": 1.0931863906127327e-07,
      "loss": 5.8266,
      "step": 196
    },
    {
      "epoch": 0.0790529695024077,
      "grad_norm": 31.205305099487305,
      "learning_rate": 6.150154258476315e-08,
      "loss": 6.1969,
      "step": 197
    },
    {
      "epoch": 0.07945425361155699,
      "grad_norm": 34.60139083862305,
      "learning_rate": 2.7337132953697554e-08,
      "loss": 6.0804,
      "step": 198
    },
    {
      "epoch": 0.07985553772070626,
      "grad_norm": 23.974050521850586,
      "learning_rate": 6.834750376549792e-09,
      "loss": 5.548,
      "step": 199
    },
    {
      "epoch": 0.08025682182985554,
      "grad_norm": 30.602638244628906,
      "learning_rate": 0.0,
      "loss": 5.8415,
      "step": 200
    },
    {
      "epoch": 0.08025682182985554,
      "eval_loss": 1.3416213989257812,
      "eval_runtime": 76.8073,
      "eval_samples_per_second": 54.656,
      "eval_steps_per_second": 13.671,
      "step": 200
    }
  ],
  "logging_steps": 1,
  "max_steps": 200,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "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": 2.7439110159335424e+16,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}