{
  "best_metric": 0.8490645885467529,
  "best_model_checkpoint": "miner_id_24/checkpoint-200",
  "epoch": 0.9291521486643438,
  "eval_steps": 50,
  "global_step": 200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.004645760743321719,
      "grad_norm": 1.0351998805999756,
      "learning_rate": 1e-05,
      "loss": 3.7651,
      "step": 1
    },
    {
      "epoch": 0.004645760743321719,
      "eval_loss": 0.9367796778678894,
      "eval_runtime": 24.7892,
      "eval_samples_per_second": 14.643,
      "eval_steps_per_second": 3.671,
      "step": 1
    },
    {
      "epoch": 0.009291521486643438,
      "grad_norm": 0.8076879382133484,
      "learning_rate": 2e-05,
      "loss": 3.6409,
      "step": 2
    },
    {
      "epoch": 0.013937282229965157,
      "grad_norm": 1.0365294218063354,
      "learning_rate": 3e-05,
      "loss": 3.6616,
      "step": 3
    },
    {
      "epoch": 0.018583042973286876,
      "grad_norm": 0.7527905702590942,
      "learning_rate": 4e-05,
      "loss": 3.9612,
      "step": 4
    },
    {
      "epoch": 0.023228803716608595,
      "grad_norm": 0.6469897031784058,
      "learning_rate": 5e-05,
      "loss": 3.484,
      "step": 5
    },
    {
      "epoch": 0.027874564459930314,
      "grad_norm": 0.6465486288070679,
      "learning_rate": 6e-05,
      "loss": 3.4777,
      "step": 6
    },
    {
      "epoch": 0.032520325203252036,
      "grad_norm": 0.5248836278915405,
      "learning_rate": 7e-05,
      "loss": 3.1566,
      "step": 7
    },
    {
      "epoch": 0.03716608594657375,
      "grad_norm": 0.4582987129688263,
      "learning_rate": 8e-05,
      "loss": 3.5161,
      "step": 8
    },
    {
      "epoch": 0.041811846689895474,
      "grad_norm": 0.47173160314559937,
      "learning_rate": 9e-05,
      "loss": 3.5686,
      "step": 9
    },
    {
      "epoch": 0.04645760743321719,
      "grad_norm": 0.48394641280174255,
      "learning_rate": 0.0001,
      "loss": 3.8784,
      "step": 10
    },
    {
      "epoch": 0.05110336817653891,
      "grad_norm": 0.5546756386756897,
      "learning_rate": 9.999316524962345e-05,
      "loss": 3.2326,
      "step": 11
    },
    {
      "epoch": 0.05574912891986063,
      "grad_norm": 0.4756205081939697,
      "learning_rate": 9.997266286704631e-05,
      "loss": 3.6796,
      "step": 12
    },
    {
      "epoch": 0.06039488966318235,
      "grad_norm": 0.575846791267395,
      "learning_rate": 9.993849845741524e-05,
      "loss": 3.433,
      "step": 13
    },
    {
      "epoch": 0.06504065040650407,
      "grad_norm": 0.4966892898082733,
      "learning_rate": 9.989068136093873e-05,
      "loss": 3.5057,
      "step": 14
    },
    {
      "epoch": 0.06968641114982578,
      "grad_norm": 0.47015583515167236,
      "learning_rate": 9.98292246503335e-05,
      "loss": 3.5669,
      "step": 15
    },
    {
      "epoch": 0.0743321718931475,
      "grad_norm": 0.4398474097251892,
      "learning_rate": 9.975414512725057e-05,
      "loss": 3.2473,
      "step": 16
    },
    {
      "epoch": 0.07897793263646923,
      "grad_norm": 0.6636183857917786,
      "learning_rate": 9.966546331768191e-05,
      "loss": 3.433,
      "step": 17
    },
    {
      "epoch": 0.08362369337979095,
      "grad_norm": 0.37653350830078125,
      "learning_rate": 9.956320346634876e-05,
      "loss": 3.5076,
      "step": 18
    },
    {
      "epoch": 0.08826945412311266,
      "grad_norm": 0.447531521320343,
      "learning_rate": 9.944739353007344e-05,
      "loss": 3.9193,
      "step": 19
    },
    {
      "epoch": 0.09291521486643438,
      "grad_norm": 0.457386314868927,
      "learning_rate": 9.931806517013612e-05,
      "loss": 3.1872,
      "step": 20
    },
    {
      "epoch": 0.0975609756097561,
      "grad_norm": 0.34822237491607666,
      "learning_rate": 9.917525374361912e-05,
      "loss": 3.4857,
      "step": 21
    },
    {
      "epoch": 0.10220673635307782,
      "grad_norm": 0.40389588475227356,
      "learning_rate": 9.901899829374047e-05,
      "loss": 3.59,
      "step": 22
    },
    {
      "epoch": 0.10685249709639953,
      "grad_norm": 0.38797664642333984,
      "learning_rate": 9.884934153917997e-05,
      "loss": 3.1609,
      "step": 23
    },
    {
      "epoch": 0.11149825783972125,
      "grad_norm": 0.426193505525589,
      "learning_rate": 9.86663298624003e-05,
      "loss": 2.7746,
      "step": 24
    },
    {
      "epoch": 0.11614401858304298,
      "grad_norm": 0.43234747648239136,
      "learning_rate": 9.847001329696653e-05,
      "loss": 2.7849,
      "step": 25
    },
    {
      "epoch": 0.1207897793263647,
      "grad_norm": 0.45846283435821533,
      "learning_rate": 9.826044551386744e-05,
      "loss": 3.5435,
      "step": 26
    },
    {
      "epoch": 0.1254355400696864,
      "grad_norm": 0.4334864020347595,
      "learning_rate": 9.803768380684242e-05,
      "loss": 3.1148,
      "step": 27
    },
    {
      "epoch": 0.13008130081300814,
      "grad_norm": 0.43216222524642944,
      "learning_rate": 9.780178907671789e-05,
      "loss": 3.0365,
      "step": 28
    },
    {
      "epoch": 0.13472706155632985,
      "grad_norm": 0.47313007712364197,
      "learning_rate": 9.755282581475769e-05,
      "loss": 3.4931,
      "step": 29
    },
    {
      "epoch": 0.13937282229965156,
      "grad_norm": 0.46203476190567017,
      "learning_rate": 9.729086208503174e-05,
      "loss": 2.8042,
      "step": 30
    },
    {
      "epoch": 0.1440185830429733,
      "grad_norm": 0.47654610872268677,
      "learning_rate": 9.701596950580806e-05,
      "loss": 3.1323,
      "step": 31
    },
    {
      "epoch": 0.148664343786295,
      "grad_norm": 0.4681512117385864,
      "learning_rate": 9.672822322997305e-05,
      "loss": 3.2426,
      "step": 32
    },
    {
      "epoch": 0.15331010452961671,
      "grad_norm": 0.4921683371067047,
      "learning_rate": 9.642770192448536e-05,
      "loss": 3.4348,
      "step": 33
    },
    {
      "epoch": 0.15795586527293845,
      "grad_norm": 0.551809549331665,
      "learning_rate": 9.611448774886924e-05,
      "loss": 3.2269,
      "step": 34
    },
    {
      "epoch": 0.16260162601626016,
      "grad_norm": 0.5451248288154602,
      "learning_rate": 9.578866633275288e-05,
      "loss": 3.1664,
      "step": 35
    },
    {
      "epoch": 0.1672473867595819,
      "grad_norm": 0.5960503816604614,
      "learning_rate": 9.545032675245813e-05,
      "loss": 3.1891,
      "step": 36
    },
    {
      "epoch": 0.1718931475029036,
      "grad_norm": 0.538750171661377,
      "learning_rate": 9.509956150664796e-05,
      "loss": 3.384,
      "step": 37
    },
    {
      "epoch": 0.1765389082462253,
      "grad_norm": 0.6365413069725037,
      "learning_rate": 9.473646649103818e-05,
      "loss": 2.9003,
      "step": 38
    },
    {
      "epoch": 0.18118466898954705,
      "grad_norm": 0.6132746338844299,
      "learning_rate": 9.43611409721806e-05,
      "loss": 2.7394,
      "step": 39
    },
    {
      "epoch": 0.18583042973286876,
      "grad_norm": 0.672350287437439,
      "learning_rate": 9.397368756032445e-05,
      "loss": 3.5575,
      "step": 40
    },
    {
      "epoch": 0.19047619047619047,
      "grad_norm": 0.6099003553390503,
      "learning_rate": 9.357421218136386e-05,
      "loss": 3.5746,
      "step": 41
    },
    {
      "epoch": 0.1951219512195122,
      "grad_norm": 0.6504359245300293,
      "learning_rate": 9.316282404787871e-05,
      "loss": 3.4364,
      "step": 42
    },
    {
      "epoch": 0.1997677119628339,
      "grad_norm": 0.8467481732368469,
      "learning_rate": 9.273963562927695e-05,
      "loss": 3.3204,
      "step": 43
    },
    {
      "epoch": 0.20441347270615565,
      "grad_norm": 0.7635272741317749,
      "learning_rate": 9.230476262104677e-05,
      "loss": 3.3328,
      "step": 44
    },
    {
      "epoch": 0.20905923344947736,
      "grad_norm": 0.8729445338249207,
      "learning_rate": 9.185832391312644e-05,
      "loss": 3.3263,
      "step": 45
    },
    {
      "epoch": 0.21370499419279906,
      "grad_norm": 0.8443131446838379,
      "learning_rate": 9.140044155740101e-05,
      "loss": 3.4818,
      "step": 46
    },
    {
      "epoch": 0.2183507549361208,
      "grad_norm": 0.8567115068435669,
      "learning_rate": 9.093124073433463e-05,
      "loss": 3.2892,
      "step": 47
    },
    {
      "epoch": 0.2229965156794425,
      "grad_norm": 0.9528262615203857,
      "learning_rate": 9.045084971874738e-05,
      "loss": 3.4194,
      "step": 48
    },
    {
      "epoch": 0.22764227642276422,
      "grad_norm": 1.6198744773864746,
      "learning_rate": 8.995939984474624e-05,
      "loss": 4.4332,
      "step": 49
    },
    {
      "epoch": 0.23228803716608595,
      "grad_norm": 1.9487779140472412,
      "learning_rate": 8.945702546981969e-05,
      "loss": 4.2732,
      "step": 50
    },
    {
      "epoch": 0.23228803716608595,
      "eval_loss": 0.8812066912651062,
      "eval_runtime": 25.1953,
      "eval_samples_per_second": 14.407,
      "eval_steps_per_second": 3.612,
      "step": 50
    },
    {
      "epoch": 0.23693379790940766,
      "grad_norm": 0.5852674841880798,
      "learning_rate": 8.894386393810563e-05,
      "loss": 3.3808,
      "step": 51
    },
    {
      "epoch": 0.2415795586527294,
      "grad_norm": 0.6277382969856262,
      "learning_rate": 8.842005554284296e-05,
      "loss": 3.3238,
      "step": 52
    },
    {
      "epoch": 0.2462253193960511,
      "grad_norm": 0.5941165685653687,
      "learning_rate": 8.788574348801675e-05,
      "loss": 3.238,
      "step": 53
    },
    {
      "epoch": 0.2508710801393728,
      "grad_norm": 0.6185467839241028,
      "learning_rate": 8.73410738492077e-05,
      "loss": 3.4884,
      "step": 54
    },
    {
      "epoch": 0.25551684088269455,
      "grad_norm": 0.5919776558876038,
      "learning_rate": 8.678619553365659e-05,
      "loss": 3.1948,
      "step": 55
    },
    {
      "epoch": 0.2601626016260163,
      "grad_norm": 0.42011991143226624,
      "learning_rate": 8.622126023955446e-05,
      "loss": 3.4183,
      "step": 56
    },
    {
      "epoch": 0.26480836236933797,
      "grad_norm": 0.4252772331237793,
      "learning_rate": 8.564642241456986e-05,
      "loss": 3.0559,
      "step": 57
    },
    {
      "epoch": 0.2694541231126597,
      "grad_norm": 0.3848459720611572,
      "learning_rate": 8.506183921362443e-05,
      "loss": 3.6723,
      "step": 58
    },
    {
      "epoch": 0.27409988385598144,
      "grad_norm": 0.36709821224212646,
      "learning_rate": 8.44676704559283e-05,
      "loss": 3.3125,
      "step": 59
    },
    {
      "epoch": 0.2787456445993031,
      "grad_norm": 0.33688926696777344,
      "learning_rate": 8.386407858128706e-05,
      "loss": 3.4757,
      "step": 60
    },
    {
      "epoch": 0.28339140534262486,
      "grad_norm": 0.3361741006374359,
      "learning_rate": 8.32512286056924e-05,
      "loss": 3.4076,
      "step": 61
    },
    {
      "epoch": 0.2880371660859466,
      "grad_norm": 0.33525484800338745,
      "learning_rate": 8.262928807620843e-05,
      "loss": 3.676,
      "step": 62
    },
    {
      "epoch": 0.2926829268292683,
      "grad_norm": 0.32035547494888306,
      "learning_rate": 8.199842702516583e-05,
      "loss": 3.3031,
      "step": 63
    },
    {
      "epoch": 0.29732868757259,
      "grad_norm": 0.32557880878448486,
      "learning_rate": 8.135881792367686e-05,
      "loss": 3.0611,
      "step": 64
    },
    {
      "epoch": 0.30197444831591175,
      "grad_norm": 0.3346540033817291,
      "learning_rate": 8.07106356344834e-05,
      "loss": 3.4293,
      "step": 65
    },
    {
      "epoch": 0.30662020905923343,
      "grad_norm": 0.34224340319633484,
      "learning_rate": 8.005405736415126e-05,
      "loss": 3.4509,
      "step": 66
    },
    {
      "epoch": 0.31126596980255516,
      "grad_norm": 0.30891191959381104,
      "learning_rate": 7.938926261462366e-05,
      "loss": 3.1424,
      "step": 67
    },
    {
      "epoch": 0.3159117305458769,
      "grad_norm": 0.3381284475326538,
      "learning_rate": 7.871643313414718e-05,
      "loss": 3.4033,
      "step": 68
    },
    {
      "epoch": 0.3205574912891986,
      "grad_norm": 0.2900513708591461,
      "learning_rate": 7.803575286758364e-05,
      "loss": 3.0363,
      "step": 69
    },
    {
      "epoch": 0.3252032520325203,
      "grad_norm": 0.2948324382305145,
      "learning_rate": 7.734740790612136e-05,
      "loss": 3.3241,
      "step": 70
    },
    {
      "epoch": 0.32984901277584205,
      "grad_norm": 0.30632418394088745,
      "learning_rate": 7.66515864363997e-05,
      "loss": 3.5767,
      "step": 71
    },
    {
      "epoch": 0.3344947735191638,
      "grad_norm": 0.3329317569732666,
      "learning_rate": 7.594847868906076e-05,
      "loss": 3.0772,
      "step": 72
    },
    {
      "epoch": 0.33914053426248547,
      "grad_norm": 0.35231274366378784,
      "learning_rate": 7.52382768867422e-05,
      "loss": 2.8675,
      "step": 73
    },
    {
      "epoch": 0.3437862950058072,
      "grad_norm": 0.33026960492134094,
      "learning_rate": 7.452117519152542e-05,
      "loss": 2.8798,
      "step": 74
    },
    {
      "epoch": 0.34843205574912894,
      "grad_norm": 0.4055929481983185,
      "learning_rate": 7.379736965185368e-05,
      "loss": 3.1931,
      "step": 75
    },
    {
      "epoch": 0.3530778164924506,
      "grad_norm": 0.3513950705528259,
      "learning_rate": 7.30670581489344e-05,
      "loss": 3.1714,
      "step": 76
    },
    {
      "epoch": 0.35772357723577236,
      "grad_norm": 0.3544432520866394,
      "learning_rate": 7.233044034264034e-05,
      "loss": 3.0544,
      "step": 77
    },
    {
      "epoch": 0.3623693379790941,
      "grad_norm": 0.39695414900779724,
      "learning_rate": 7.158771761692464e-05,
      "loss": 3.347,
      "step": 78
    },
    {
      "epoch": 0.3670150987224158,
      "grad_norm": 0.390946626663208,
      "learning_rate": 7.083909302476453e-05,
      "loss": 2.9558,
      "step": 79
    },
    {
      "epoch": 0.3716608594657375,
      "grad_norm": 0.522554337978363,
      "learning_rate": 7.008477123264848e-05,
      "loss": 2.8602,
      "step": 80
    },
    {
      "epoch": 0.37630662020905925,
      "grad_norm": 0.4002372920513153,
      "learning_rate": 6.932495846462261e-05,
      "loss": 3.2478,
      "step": 81
    },
    {
      "epoch": 0.38095238095238093,
      "grad_norm": 0.422383189201355,
      "learning_rate": 6.855986244591104e-05,
      "loss": 3.0952,
      "step": 82
    },
    {
      "epoch": 0.38559814169570267,
      "grad_norm": 0.4236571490764618,
      "learning_rate": 6.778969234612584e-05,
      "loss": 3.0962,
      "step": 83
    },
    {
      "epoch": 0.3902439024390244,
      "grad_norm": 0.43771952390670776,
      "learning_rate": 6.701465872208216e-05,
      "loss": 3.2654,
      "step": 84
    },
    {
      "epoch": 0.3948896631823461,
      "grad_norm": 0.5260652899742126,
      "learning_rate": 6.623497346023418e-05,
      "loss": 3.5664,
      "step": 85
    },
    {
      "epoch": 0.3995354239256678,
      "grad_norm": 0.46263155341148376,
      "learning_rate": 6.545084971874738e-05,
      "loss": 3.5696,
      "step": 86
    },
    {
      "epoch": 0.40418118466898956,
      "grad_norm": 0.47570663690567017,
      "learning_rate": 6.466250186922325e-05,
      "loss": 2.8717,
      "step": 87
    },
    {
      "epoch": 0.4088269454123113,
      "grad_norm": 0.5219430327415466,
      "learning_rate": 6.387014543809223e-05,
      "loss": 3.4943,
      "step": 88
    },
    {
      "epoch": 0.413472706155633,
      "grad_norm": 0.5462395548820496,
      "learning_rate": 6.307399704769099e-05,
      "loss": 3.3059,
      "step": 89
    },
    {
      "epoch": 0.4181184668989547,
      "grad_norm": 0.612158477306366,
      "learning_rate": 6.227427435703997e-05,
      "loss": 3.6598,
      "step": 90
    },
    {
      "epoch": 0.42276422764227645,
      "grad_norm": 0.62015300989151,
      "learning_rate": 6.147119600233758e-05,
      "loss": 3.6712,
      "step": 91
    },
    {
      "epoch": 0.4274099883855981,
      "grad_norm": 0.5732129812240601,
      "learning_rate": 6.066498153718735e-05,
      "loss": 3.029,
      "step": 92
    },
    {
      "epoch": 0.43205574912891986,
      "grad_norm": 0.7043212652206421,
      "learning_rate": 5.985585137257401e-05,
      "loss": 3.5952,
      "step": 93
    },
    {
      "epoch": 0.4367015098722416,
      "grad_norm": 0.65862637758255,
      "learning_rate": 5.90440267166055e-05,
      "loss": 3.4988,
      "step": 94
    },
    {
      "epoch": 0.4413472706155633,
      "grad_norm": 0.6996191143989563,
      "learning_rate": 5.8229729514036705e-05,
      "loss": 3.3605,
      "step": 95
    },
    {
      "epoch": 0.445993031358885,
      "grad_norm": 0.9373538494110107,
      "learning_rate": 5.74131823855921e-05,
      "loss": 3.9063,
      "step": 96
    },
    {
      "epoch": 0.45063879210220675,
      "grad_norm": 0.8076989650726318,
      "learning_rate": 5.6594608567103456e-05,
      "loss": 3.6397,
      "step": 97
    },
    {
      "epoch": 0.45528455284552843,
      "grad_norm": 1.1284147500991821,
      "learning_rate": 5.577423184847932e-05,
      "loss": 3.5403,
      "step": 98
    },
    {
      "epoch": 0.45993031358885017,
      "grad_norm": 0.948340117931366,
      "learning_rate": 5.495227651252315e-05,
      "loss": 3.5053,
      "step": 99
    },
    {
      "epoch": 0.4645760743321719,
      "grad_norm": 2.0170774459838867,
      "learning_rate": 5.4128967273616625e-05,
      "loss": 3.9853,
      "step": 100
    },
    {
      "epoch": 0.4645760743321719,
      "eval_loss": 0.8647764325141907,
      "eval_runtime": 25.34,
      "eval_samples_per_second": 14.325,
      "eval_steps_per_second": 3.591,
      "step": 100
    },
    {
      "epoch": 0.4692218350754936,
      "grad_norm": 0.5611105561256409,
      "learning_rate": 5.330452921628497e-05,
      "loss": 3.3805,
      "step": 101
    },
    {
      "epoch": 0.4738675958188153,
      "grad_norm": 0.5419340133666992,
      "learning_rate": 5.247918773366112e-05,
      "loss": 3.2216,
      "step": 102
    },
    {
      "epoch": 0.47851335656213706,
      "grad_norm": 0.5218324661254883,
      "learning_rate": 5.165316846586541e-05,
      "loss": 3.2722,
      "step": 103
    },
    {
      "epoch": 0.4831591173054588,
      "grad_norm": 0.5150585174560547,
      "learning_rate": 5.0826697238317935e-05,
      "loss": 3.6204,
      "step": 104
    },
    {
      "epoch": 0.4878048780487805,
      "grad_norm": 0.4715979993343353,
      "learning_rate": 5e-05,
      "loss": 3.5924,
      "step": 105
    },
    {
      "epoch": 0.4924506387921022,
      "grad_norm": 5.683842658996582,
      "learning_rate": 4.917330276168208e-05,
      "loss": 3.0733,
      "step": 106
    },
    {
      "epoch": 0.49709639953542395,
      "grad_norm": 0.3685954213142395,
      "learning_rate": 4.834683153413459e-05,
      "loss": 3.4506,
      "step": 107
    },
    {
      "epoch": 0.5017421602787456,
      "grad_norm": 0.38257554173469543,
      "learning_rate": 4.7520812266338885e-05,
      "loss": 3.0914,
      "step": 108
    },
    {
      "epoch": 0.5063879210220673,
      "grad_norm": 0.3234102129936218,
      "learning_rate": 4.669547078371504e-05,
      "loss": 3.3799,
      "step": 109
    },
    {
      "epoch": 0.5110336817653891,
      "grad_norm": 0.3754776120185852,
      "learning_rate": 4.5871032726383386e-05,
      "loss": 3.5261,
      "step": 110
    },
    {
      "epoch": 0.5156794425087108,
      "grad_norm": 0.3709120452404022,
      "learning_rate": 4.504772348747687e-05,
      "loss": 3.8127,
      "step": 111
    },
    {
      "epoch": 0.5203252032520326,
      "grad_norm": 0.3062532842159271,
      "learning_rate": 4.4225768151520694e-05,
      "loss": 3.4371,
      "step": 112
    },
    {
      "epoch": 0.5249709639953543,
      "grad_norm": 0.32583022117614746,
      "learning_rate": 4.3405391432896555e-05,
      "loss": 3.2271,
      "step": 113
    },
    {
      "epoch": 0.5296167247386759,
      "grad_norm": 0.30351176857948303,
      "learning_rate": 4.2586817614407895e-05,
      "loss": 3.2774,
      "step": 114
    },
    {
      "epoch": 0.5342624854819977,
      "grad_norm": 0.29279470443725586,
      "learning_rate": 4.17702704859633e-05,
      "loss": 3.5312,
      "step": 115
    },
    {
      "epoch": 0.5389082462253194,
      "grad_norm": 0.2936018109321594,
      "learning_rate": 4.095597328339452e-05,
      "loss": 3.1139,
      "step": 116
    },
    {
      "epoch": 0.5435540069686411,
      "grad_norm": 0.4334779679775238,
      "learning_rate": 4.0144148627425993e-05,
      "loss": 3.3485,
      "step": 117
    },
    {
      "epoch": 0.5481997677119629,
      "grad_norm": 0.35199621319770813,
      "learning_rate": 3.933501846281267e-05,
      "loss": 3.7469,
      "step": 118
    },
    {
      "epoch": 0.5528455284552846,
      "grad_norm": 0.32896292209625244,
      "learning_rate": 3.852880399766243e-05,
      "loss": 3.6798,
      "step": 119
    },
    {
      "epoch": 0.5574912891986062,
      "grad_norm": 0.3263486623764038,
      "learning_rate": 3.772572564296005e-05,
      "loss": 3.091,
      "step": 120
    },
    {
      "epoch": 0.562137049941928,
      "grad_norm": 0.2836499810218811,
      "learning_rate": 3.6926002952309016e-05,
      "loss": 3.274,
      "step": 121
    },
    {
      "epoch": 0.5667828106852497,
      "grad_norm": 0.32200539112091064,
      "learning_rate": 3.612985456190778e-05,
      "loss": 3.6974,
      "step": 122
    },
    {
      "epoch": 0.5714285714285714,
      "grad_norm": 0.32614466547966003,
      "learning_rate": 3.533749813077677e-05,
      "loss": 2.945,
      "step": 123
    },
    {
      "epoch": 0.5760743321718932,
      "grad_norm": 0.36022284626960754,
      "learning_rate": 3.4549150281252636e-05,
      "loss": 3.1623,
      "step": 124
    },
    {
      "epoch": 0.5807200929152149,
      "grad_norm": 0.35150450468063354,
      "learning_rate": 3.3765026539765834e-05,
      "loss": 2.9815,
      "step": 125
    },
    {
      "epoch": 0.5853658536585366,
      "grad_norm": 0.39378654956817627,
      "learning_rate": 3.298534127791785e-05,
      "loss": 2.8921,
      "step": 126
    },
    {
      "epoch": 0.5900116144018583,
      "grad_norm": 0.37585243582725525,
      "learning_rate": 3.221030765387417e-05,
      "loss": 3.3454,
      "step": 127
    },
    {
      "epoch": 0.59465737514518,
      "grad_norm": 0.37854430079460144,
      "learning_rate": 3.144013755408895e-05,
      "loss": 3.2877,
      "step": 128
    },
    {
      "epoch": 0.5993031358885017,
      "grad_norm": 0.4166063666343689,
      "learning_rate": 3.0675041535377405e-05,
      "loss": 2.9788,
      "step": 129
    },
    {
      "epoch": 0.6039488966318235,
      "grad_norm": 0.3931642770767212,
      "learning_rate": 2.991522876735154e-05,
      "loss": 3.0149,
      "step": 130
    },
    {
      "epoch": 0.6085946573751452,
      "grad_norm": 0.40521040558815,
      "learning_rate": 2.916090697523549e-05,
      "loss": 3.0367,
      "step": 131
    },
    {
      "epoch": 0.6132404181184669,
      "grad_norm": 0.4211152195930481,
      "learning_rate": 2.8412282383075363e-05,
      "loss": 2.7199,
      "step": 132
    },
    {
      "epoch": 0.6178861788617886,
      "grad_norm": 0.41897135972976685,
      "learning_rate": 2.766955965735968e-05,
      "loss": 3.334,
      "step": 133
    },
    {
      "epoch": 0.6225319396051103,
      "grad_norm": 0.4201313555240631,
      "learning_rate": 2.693294185106562e-05,
      "loss": 3.2249,
      "step": 134
    },
    {
      "epoch": 0.627177700348432,
      "grad_norm": 0.4595812261104584,
      "learning_rate": 2.6202630348146324e-05,
      "loss": 3.4358,
      "step": 135
    },
    {
      "epoch": 0.6318234610917538,
      "grad_norm": 0.4839074909687042,
      "learning_rate": 2.547882480847461e-05,
      "loss": 3.1214,
      "step": 136
    },
    {
      "epoch": 0.6364692218350755,
      "grad_norm": 0.5086528062820435,
      "learning_rate": 2.476172311325783e-05,
      "loss": 3.3255,
      "step": 137
    },
    {
      "epoch": 0.6411149825783972,
      "grad_norm": 0.527389407157898,
      "learning_rate": 2.405152131093926e-05,
      "loss": 3.2876,
      "step": 138
    },
    {
      "epoch": 0.645760743321719,
      "grad_norm": 0.5561457276344299,
      "learning_rate": 2.3348413563600325e-05,
      "loss": 3.0794,
      "step": 139
    },
    {
      "epoch": 0.6504065040650406,
      "grad_norm": 0.5368041396141052,
      "learning_rate": 2.2652592093878666e-05,
      "loss": 2.9482,
      "step": 140
    },
    {
      "epoch": 0.6550522648083623,
      "grad_norm": 0.5672673583030701,
      "learning_rate": 2.196424713241637e-05,
      "loss": 3.2003,
      "step": 141
    },
    {
      "epoch": 0.6596980255516841,
      "grad_norm": 0.6208717226982117,
      "learning_rate": 2.128356686585282e-05,
      "loss": 3.4714,
      "step": 142
    },
    {
      "epoch": 0.6643437862950058,
      "grad_norm": 0.6111561059951782,
      "learning_rate": 2.061073738537635e-05,
      "loss": 3.2983,
      "step": 143
    },
    {
      "epoch": 0.6689895470383276,
      "grad_norm": 0.6501817107200623,
      "learning_rate": 1.9945942635848748e-05,
      "loss": 3.3733,
      "step": 144
    },
    {
      "epoch": 0.6736353077816493,
      "grad_norm": 0.5993291139602661,
      "learning_rate": 1.928936436551661e-05,
      "loss": 2.8484,
      "step": 145
    },
    {
      "epoch": 0.6782810685249709,
      "grad_norm": 0.7727030515670776,
      "learning_rate": 1.8641182076323148e-05,
      "loss": 3.8404,
      "step": 146
    },
    {
      "epoch": 0.6829268292682927,
      "grad_norm": 0.7875126600265503,
      "learning_rate": 1.800157297483417e-05,
      "loss": 3.4292,
      "step": 147
    },
    {
      "epoch": 0.6875725900116144,
      "grad_norm": 1.5537347793579102,
      "learning_rate": 1.7370711923791567e-05,
      "loss": 3.9599,
      "step": 148
    },
    {
      "epoch": 0.6922183507549361,
      "grad_norm": 1.4659509658813477,
      "learning_rate": 1.6748771394307585e-05,
      "loss": 3.1774,
      "step": 149
    },
    {
      "epoch": 0.6968641114982579,
      "grad_norm": 2.1345794200897217,
      "learning_rate": 1.6135921418712956e-05,
      "loss": 4.5479,
      "step": 150
    },
    {
      "epoch": 0.6968641114982579,
      "eval_loss": 0.8539987206459045,
      "eval_runtime": 25.2942,
      "eval_samples_per_second": 14.351,
      "eval_steps_per_second": 3.598,
      "step": 150
    },
    {
      "epoch": 0.7015098722415796,
      "grad_norm": 0.3620260953903198,
      "learning_rate": 1.553232954407171e-05,
      "loss": 3.1435,
      "step": 151
    },
    {
      "epoch": 0.7061556329849012,
      "grad_norm": 0.3619486391544342,
      "learning_rate": 1.4938160786375572e-05,
      "loss": 3.3398,
      "step": 152
    },
    {
      "epoch": 0.710801393728223,
      "grad_norm": 0.4846741855144501,
      "learning_rate": 1.435357758543015e-05,
      "loss": 3.3573,
      "step": 153
    },
    {
      "epoch": 0.7154471544715447,
      "grad_norm": 0.3455192744731903,
      "learning_rate": 1.3778739760445552e-05,
      "loss": 3.0987,
      "step": 154
    },
    {
      "epoch": 0.7200929152148664,
      "grad_norm": 0.3357820510864258,
      "learning_rate": 1.3213804466343421e-05,
      "loss": 3.3989,
      "step": 155
    },
    {
      "epoch": 0.7247386759581882,
      "grad_norm": 0.36104816198349,
      "learning_rate": 1.2658926150792322e-05,
      "loss": 3.2489,
      "step": 156
    },
    {
      "epoch": 0.7293844367015099,
      "grad_norm": 0.36337921023368835,
      "learning_rate": 1.2114256511983274e-05,
      "loss": 3.6525,
      "step": 157
    },
    {
      "epoch": 0.7340301974448316,
      "grad_norm": 0.35237929224967957,
      "learning_rate": 1.157994445715706e-05,
      "loss": 3.1483,
      "step": 158
    },
    {
      "epoch": 0.7386759581881533,
      "grad_norm": 0.37043413519859314,
      "learning_rate": 1.1056136061894384e-05,
      "loss": 3.5891,
      "step": 159
    },
    {
      "epoch": 0.743321718931475,
      "grad_norm": 0.31886643171310425,
      "learning_rate": 1.0542974530180327e-05,
      "loss": 3.6873,
      "step": 160
    },
    {
      "epoch": 0.7479674796747967,
      "grad_norm": 0.3651953339576721,
      "learning_rate": 1.0040600155253765e-05,
      "loss": 3.0335,
      "step": 161
    },
    {
      "epoch": 0.7526132404181185,
      "grad_norm": 0.3227655589580536,
      "learning_rate": 9.549150281252633e-06,
      "loss": 3.4653,
      "step": 162
    },
    {
      "epoch": 0.7572590011614402,
      "grad_norm": 0.3414734899997711,
      "learning_rate": 9.068759265665384e-06,
      "loss": 3.5061,
      "step": 163
    },
    {
      "epoch": 0.7619047619047619,
      "grad_norm": 0.3186526894569397,
      "learning_rate": 8.599558442598998e-06,
      "loss": 3.4749,
      "step": 164
    },
    {
      "epoch": 0.7665505226480837,
      "grad_norm": 0.3220815658569336,
      "learning_rate": 8.141676086873572e-06,
      "loss": 3.801,
      "step": 165
    },
    {
      "epoch": 0.7711962833914053,
      "grad_norm": 0.3330041170120239,
      "learning_rate": 7.695237378953223e-06,
      "loss": 3.4381,
      "step": 166
    },
    {
      "epoch": 0.775842044134727,
      "grad_norm": 0.33168530464172363,
      "learning_rate": 7.260364370723044e-06,
      "loss": 3.0619,
      "step": 167
    },
    {
      "epoch": 0.7804878048780488,
      "grad_norm": 0.3295885920524597,
      "learning_rate": 6.837175952121306e-06,
      "loss": 3.4097,
      "step": 168
    },
    {
      "epoch": 0.7851335656213705,
      "grad_norm": 0.34456390142440796,
      "learning_rate": 6.425787818636131e-06,
      "loss": 3.404,
      "step": 169
    },
    {
      "epoch": 0.7897793263646922,
      "grad_norm": 0.3085865378379822,
      "learning_rate": 6.026312439675552e-06,
      "loss": 3.5768,
      "step": 170
    },
    {
      "epoch": 0.794425087108014,
      "grad_norm": 0.3196917176246643,
      "learning_rate": 5.6388590278194096e-06,
      "loss": 3.4992,
      "step": 171
    },
    {
      "epoch": 0.7990708478513356,
      "grad_norm": 0.37706947326660156,
      "learning_rate": 5.263533508961827e-06,
      "loss": 3.3834,
      "step": 172
    },
    {
      "epoch": 0.8037166085946573,
      "grad_norm": 0.30736809968948364,
      "learning_rate": 4.900438493352055e-06,
      "loss": 2.775,
      "step": 173
    },
    {
      "epoch": 0.8083623693379791,
      "grad_norm": 0.3423340320587158,
      "learning_rate": 4.549673247541875e-06,
      "loss": 2.6582,
      "step": 174
    },
    {
      "epoch": 0.8130081300813008,
      "grad_norm": 0.358732670545578,
      "learning_rate": 4.2113336672471245e-06,
      "loss": 2.8197,
      "step": 175
    },
    {
      "epoch": 0.8176538908246226,
      "grad_norm": 0.36769434809684753,
      "learning_rate": 3.885512251130763e-06,
      "loss": 3.1628,
      "step": 176
    },
    {
      "epoch": 0.8222996515679443,
      "grad_norm": 0.3900729715824127,
      "learning_rate": 3.5722980755146517e-06,
      "loss": 2.7941,
      "step": 177
    },
    {
      "epoch": 0.826945412311266,
      "grad_norm": 0.416951447725296,
      "learning_rate": 3.271776770026963e-06,
      "loss": 2.9058,
      "step": 178
    },
    {
      "epoch": 0.8315911730545877,
      "grad_norm": 0.37644949555397034,
      "learning_rate": 2.9840304941919415e-06,
      "loss": 2.9874,
      "step": 179
    },
    {
      "epoch": 0.8362369337979094,
      "grad_norm": 0.42537301778793335,
      "learning_rate": 2.7091379149682685e-06,
      "loss": 2.9036,
      "step": 180
    },
    {
      "epoch": 0.8408826945412311,
      "grad_norm": 0.46386778354644775,
      "learning_rate": 2.4471741852423237e-06,
      "loss": 3.2971,
      "step": 181
    },
    {
      "epoch": 0.8455284552845529,
      "grad_norm": 0.45652905106544495,
      "learning_rate": 2.1982109232821178e-06,
      "loss": 3.4172,
      "step": 182
    },
    {
      "epoch": 0.8501742160278746,
      "grad_norm": 0.4876401424407959,
      "learning_rate": 1.962316193157593e-06,
      "loss": 3.2619,
      "step": 183
    },
    {
      "epoch": 0.8548199767711963,
      "grad_norm": 0.47403448820114136,
      "learning_rate": 1.7395544861325718e-06,
      "loss": 3.1696,
      "step": 184
    },
    {
      "epoch": 0.859465737514518,
      "grad_norm": 0.49599969387054443,
      "learning_rate": 1.5299867030334814e-06,
      "loss": 3.2579,
      "step": 185
    },
    {
      "epoch": 0.8641114982578397,
      "grad_norm": 0.5189352631568909,
      "learning_rate": 1.333670137599713e-06,
      "loss": 2.9723,
      "step": 186
    },
    {
      "epoch": 0.8687572590011614,
      "grad_norm": 0.5410155057907104,
      "learning_rate": 1.1506584608200367e-06,
      "loss": 3.497,
      "step": 187
    },
    {
      "epoch": 0.8734030197444832,
      "grad_norm": 0.497802734375,
      "learning_rate": 9.810017062595322e-07,
      "loss": 3.2479,
      "step": 188
    },
    {
      "epoch": 0.8780487804878049,
      "grad_norm": 0.5857788920402527,
      "learning_rate": 8.247462563808817e-07,
      "loss": 3.5006,
      "step": 189
    },
    {
      "epoch": 0.8826945412311266,
      "grad_norm": 0.6114856004714966,
      "learning_rate": 6.819348298638839e-07,
      "loss": 3.5122,
      "step": 190
    },
    {
      "epoch": 0.8873403019744484,
      "grad_norm": 0.6082403063774109,
      "learning_rate": 5.526064699265753e-07,
      "loss": 3.5667,
      "step": 191
    },
    {
      "epoch": 0.89198606271777,
      "grad_norm": 0.6140199899673462,
      "learning_rate": 4.367965336512403e-07,
      "loss": 3.3399,
      "step": 192
    },
    {
      "epoch": 0.8966318234610917,
      "grad_norm": 0.6315708756446838,
      "learning_rate": 3.3453668231809286e-07,
      "loss": 3.877,
      "step": 193
    },
    {
      "epoch": 0.9012775842044135,
      "grad_norm": 0.6813123226165771,
      "learning_rate": 2.458548727494292e-07,
      "loss": 3.3962,
      "step": 194
    },
    {
      "epoch": 0.9059233449477352,
      "grad_norm": 0.5829943418502808,
      "learning_rate": 1.7077534966650766e-07,
      "loss": 3.2863,
      "step": 195
    },
    {
      "epoch": 0.9105691056910569,
      "grad_norm": 0.9277510046958923,
      "learning_rate": 1.0931863906127327e-07,
      "loss": 3.7485,
      "step": 196
    },
    {
      "epoch": 0.9152148664343787,
      "grad_norm": 1.1095796823501587,
      "learning_rate": 6.150154258476315e-08,
      "loss": 3.5525,
      "step": 197
    },
    {
      "epoch": 0.9198606271777003,
      "grad_norm": 1.0247440338134766,
      "learning_rate": 2.7337132953697554e-08,
      "loss": 3.5524,
      "step": 198
    },
    {
      "epoch": 0.924506387921022,
      "grad_norm": 0.9513870477676392,
      "learning_rate": 6.834750376549792e-09,
      "loss": 3.6051,
      "step": 199
    },
    {
      "epoch": 0.9291521486643438,
      "grad_norm": 2.0042521953582764,
      "learning_rate": 0.0,
      "loss": 4.5657,
      "step": 200
    },
    {
      "epoch": 0.9291521486643438,
      "eval_loss": 0.8490645885467529,
      "eval_runtime": 25.3502,
      "eval_samples_per_second": 14.319,
      "eval_steps_per_second": 3.59,
      "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": 2.1774141599042765e+17,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}