{
  "best_metric": 0.49133214354515076,
  "best_model_checkpoint": "miner_id_24/checkpoint-200",
  "epoch": 0.22605255721955356,
  "eval_steps": 50,
  "global_step": 200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0011302627860977678,
      "grad_norm": 801.9556274414062,
      "learning_rate": 1e-05,
      "loss": 11.6478,
      "step": 1
    },
    {
      "epoch": 0.0011302627860977678,
      "eval_loss": 3.2303900718688965,
      "eval_runtime": 104.6049,
      "eval_samples_per_second": 14.244,
      "eval_steps_per_second": 3.566,
      "step": 1
    },
    {
      "epoch": 0.0022605255721955355,
      "grad_norm": 1177.0208740234375,
      "learning_rate": 2e-05,
      "loss": 12.4896,
      "step": 2
    },
    {
      "epoch": 0.003390788358293303,
      "grad_norm": 125.79412841796875,
      "learning_rate": 3e-05,
      "loss": 9.2501,
      "step": 3
    },
    {
      "epoch": 0.004521051144391071,
      "grad_norm": 67.85590362548828,
      "learning_rate": 4e-05,
      "loss": 5.744,
      "step": 4
    },
    {
      "epoch": 0.005651313930488839,
      "grad_norm": 142.0784454345703,
      "learning_rate": 5e-05,
      "loss": 5.0251,
      "step": 5
    },
    {
      "epoch": 0.006781576716586606,
      "grad_norm": 38.53276443481445,
      "learning_rate": 6e-05,
      "loss": 3.5999,
      "step": 6
    },
    {
      "epoch": 0.007911839502684374,
      "grad_norm": 13.22555923461914,
      "learning_rate": 7e-05,
      "loss": 3.0867,
      "step": 7
    },
    {
      "epoch": 0.009042102288782142,
      "grad_norm": 12.12904167175293,
      "learning_rate": 8e-05,
      "loss": 2.9462,
      "step": 8
    },
    {
      "epoch": 0.010172365074879909,
      "grad_norm": 9.223594665527344,
      "learning_rate": 9e-05,
      "loss": 2.7963,
      "step": 9
    },
    {
      "epoch": 0.011302627860977677,
      "grad_norm": 16.560197830200195,
      "learning_rate": 0.0001,
      "loss": 2.7887,
      "step": 10
    },
    {
      "epoch": 0.012432890647075446,
      "grad_norm": 24.951810836791992,
      "learning_rate": 9.999316524962345e-05,
      "loss": 2.6467,
      "step": 11
    },
    {
      "epoch": 0.013563153433173212,
      "grad_norm": 15.974726676940918,
      "learning_rate": 9.997266286704631e-05,
      "loss": 3.1178,
      "step": 12
    },
    {
      "epoch": 0.01469341621927098,
      "grad_norm": 6.611715316772461,
      "learning_rate": 9.993849845741524e-05,
      "loss": 2.6894,
      "step": 13
    },
    {
      "epoch": 0.015823679005368747,
      "grad_norm": 7.117212772369385,
      "learning_rate": 9.989068136093873e-05,
      "loss": 2.7149,
      "step": 14
    },
    {
      "epoch": 0.016953941791466517,
      "grad_norm": 5.9376349449157715,
      "learning_rate": 9.98292246503335e-05,
      "loss": 2.6921,
      "step": 15
    },
    {
      "epoch": 0.018084204577564284,
      "grad_norm": 5.6211018562316895,
      "learning_rate": 9.975414512725057e-05,
      "loss": 2.6859,
      "step": 16
    },
    {
      "epoch": 0.01921446736366205,
      "grad_norm": 5.426237106323242,
      "learning_rate": 9.966546331768191e-05,
      "loss": 2.5345,
      "step": 17
    },
    {
      "epoch": 0.020344730149759818,
      "grad_norm": 4.6002421379089355,
      "learning_rate": 9.956320346634876e-05,
      "loss": 2.5872,
      "step": 18
    },
    {
      "epoch": 0.021474992935857588,
      "grad_norm": 5.197369575500488,
      "learning_rate": 9.944739353007344e-05,
      "loss": 2.7634,
      "step": 19
    },
    {
      "epoch": 0.022605255721955354,
      "grad_norm": 7.112434387207031,
      "learning_rate": 9.931806517013612e-05,
      "loss": 2.861,
      "step": 20
    },
    {
      "epoch": 0.02373551850805312,
      "grad_norm": 4.055138111114502,
      "learning_rate": 9.917525374361912e-05,
      "loss": 2.7568,
      "step": 21
    },
    {
      "epoch": 0.02486578129415089,
      "grad_norm": 3.8503806591033936,
      "learning_rate": 9.901899829374047e-05,
      "loss": 2.5534,
      "step": 22
    },
    {
      "epoch": 0.025996044080248658,
      "grad_norm": 4.7678093910217285,
      "learning_rate": 9.884934153917997e-05,
      "loss": 2.7326,
      "step": 23
    },
    {
      "epoch": 0.027126306866346424,
      "grad_norm": 4.417437553405762,
      "learning_rate": 9.86663298624003e-05,
      "loss": 2.7002,
      "step": 24
    },
    {
      "epoch": 0.028256569652444195,
      "grad_norm": 4.683590888977051,
      "learning_rate": 9.847001329696653e-05,
      "loss": 2.8746,
      "step": 25
    },
    {
      "epoch": 0.02938683243854196,
      "grad_norm": 4.264058589935303,
      "learning_rate": 9.826044551386744e-05,
      "loss": 2.7484,
      "step": 26
    },
    {
      "epoch": 0.030517095224639728,
      "grad_norm": 4.426316261291504,
      "learning_rate": 9.803768380684242e-05,
      "loss": 2.7455,
      "step": 27
    },
    {
      "epoch": 0.031647358010737495,
      "grad_norm": 4.080742359161377,
      "learning_rate": 9.780178907671789e-05,
      "loss": 2.8513,
      "step": 28
    },
    {
      "epoch": 0.032777620796835265,
      "grad_norm": 4.0141496658325195,
      "learning_rate": 9.755282581475769e-05,
      "loss": 2.7114,
      "step": 29
    },
    {
      "epoch": 0.033907883582933035,
      "grad_norm": 4.260357856750488,
      "learning_rate": 9.729086208503174e-05,
      "loss": 2.8865,
      "step": 30
    },
    {
      "epoch": 0.0350381463690308,
      "grad_norm": 3.850391387939453,
      "learning_rate": 9.701596950580806e-05,
      "loss": 2.6973,
      "step": 31
    },
    {
      "epoch": 0.03616840915512857,
      "grad_norm": 4.164616584777832,
      "learning_rate": 9.672822322997305e-05,
      "loss": 2.7519,
      "step": 32
    },
    {
      "epoch": 0.03729867194122634,
      "grad_norm": 4.058201789855957,
      "learning_rate": 9.642770192448536e-05,
      "loss": 2.8022,
      "step": 33
    },
    {
      "epoch": 0.0384289347273241,
      "grad_norm": 4.295276641845703,
      "learning_rate": 9.611448774886924e-05,
      "loss": 2.8081,
      "step": 34
    },
    {
      "epoch": 0.03955919751342187,
      "grad_norm": 3.63132905960083,
      "learning_rate": 9.578866633275288e-05,
      "loss": 2.7836,
      "step": 35
    },
    {
      "epoch": 0.040689460299519635,
      "grad_norm": 3.809506416320801,
      "learning_rate": 9.545032675245813e-05,
      "loss": 2.752,
      "step": 36
    },
    {
      "epoch": 0.041819723085617405,
      "grad_norm": 4.227634906768799,
      "learning_rate": 9.509956150664796e-05,
      "loss": 2.8855,
      "step": 37
    },
    {
      "epoch": 0.042949985871715175,
      "grad_norm": 3.6339871883392334,
      "learning_rate": 9.473646649103818e-05,
      "loss": 2.7636,
      "step": 38
    },
    {
      "epoch": 0.04408024865781294,
      "grad_norm": 3.7836780548095703,
      "learning_rate": 9.43611409721806e-05,
      "loss": 2.8363,
      "step": 39
    },
    {
      "epoch": 0.04521051144391071,
      "grad_norm": 3.532805919647217,
      "learning_rate": 9.397368756032445e-05,
      "loss": 2.7844,
      "step": 40
    },
    {
      "epoch": 0.04634077423000848,
      "grad_norm": 8.422014236450195,
      "learning_rate": 9.357421218136386e-05,
      "loss": 2.9154,
      "step": 41
    },
    {
      "epoch": 0.04747103701610624,
      "grad_norm": 4.013293743133545,
      "learning_rate": 9.316282404787871e-05,
      "loss": 2.9438,
      "step": 42
    },
    {
      "epoch": 0.04860129980220401,
      "grad_norm": 4.012973308563232,
      "learning_rate": 9.273963562927695e-05,
      "loss": 2.7634,
      "step": 43
    },
    {
      "epoch": 0.04973156258830178,
      "grad_norm": 3.766223430633545,
      "learning_rate": 9.230476262104677e-05,
      "loss": 2.8299,
      "step": 44
    },
    {
      "epoch": 0.050861825374399545,
      "grad_norm": 3.77280855178833,
      "learning_rate": 9.185832391312644e-05,
      "loss": 2.8154,
      "step": 45
    },
    {
      "epoch": 0.051992088160497316,
      "grad_norm": 38.72694396972656,
      "learning_rate": 9.140044155740101e-05,
      "loss": 2.2626,
      "step": 46
    },
    {
      "epoch": 0.053122350946595086,
      "grad_norm": 3.4899163246154785,
      "learning_rate": 9.093124073433463e-05,
      "loss": 1.899,
      "step": 47
    },
    {
      "epoch": 0.05425261373269285,
      "grad_norm": 2.926361083984375,
      "learning_rate": 9.045084971874738e-05,
      "loss": 1.7023,
      "step": 48
    },
    {
      "epoch": 0.05538287651879062,
      "grad_norm": 3.7920968532562256,
      "learning_rate": 8.995939984474624e-05,
      "loss": 2.094,
      "step": 49
    },
    {
      "epoch": 0.05651313930488839,
      "grad_norm": 3.4063291549682617,
      "learning_rate": 8.945702546981969e-05,
      "loss": 1.8472,
      "step": 50
    },
    {
      "epoch": 0.05651313930488839,
      "eval_loss": 0.6768115758895874,
      "eval_runtime": 106.5638,
      "eval_samples_per_second": 13.982,
      "eval_steps_per_second": 3.5,
      "step": 50
    },
    {
      "epoch": 0.05764340209098615,
      "grad_norm": 16.549467086791992,
      "learning_rate": 8.894386393810563e-05,
      "loss": 2.4248,
      "step": 51
    },
    {
      "epoch": 0.05877366487708392,
      "grad_norm": 5.14683723449707,
      "learning_rate": 8.842005554284296e-05,
      "loss": 2.2162,
      "step": 52
    },
    {
      "epoch": 0.05990392766318169,
      "grad_norm": 3.8495397567749023,
      "learning_rate": 8.788574348801675e-05,
      "loss": 2.2834,
      "step": 53
    },
    {
      "epoch": 0.061034190449279456,
      "grad_norm": 20.76611328125,
      "learning_rate": 8.73410738492077e-05,
      "loss": 2.2471,
      "step": 54
    },
    {
      "epoch": 0.062164453235377226,
      "grad_norm": 3.1339986324310303,
      "learning_rate": 8.678619553365659e-05,
      "loss": 2.2386,
      "step": 55
    },
    {
      "epoch": 0.06329471602147499,
      "grad_norm": 5.5675950050354,
      "learning_rate": 8.622126023955446e-05,
      "loss": 2.3174,
      "step": 56
    },
    {
      "epoch": 0.06442497880757277,
      "grad_norm": 3.2717695236206055,
      "learning_rate": 8.564642241456986e-05,
      "loss": 2.2979,
      "step": 57
    },
    {
      "epoch": 0.06555524159367053,
      "grad_norm": 3.1835787296295166,
      "learning_rate": 8.506183921362443e-05,
      "loss": 2.2476,
      "step": 58
    },
    {
      "epoch": 0.06668550437976829,
      "grad_norm": 3.4862701892852783,
      "learning_rate": 8.44676704559283e-05,
      "loss": 2.4119,
      "step": 59
    },
    {
      "epoch": 0.06781576716586607,
      "grad_norm": 3.279918670654297,
      "learning_rate": 8.386407858128706e-05,
      "loss": 2.3689,
      "step": 60
    },
    {
      "epoch": 0.06894602995196383,
      "grad_norm": 3.209315538406372,
      "learning_rate": 8.32512286056924e-05,
      "loss": 2.2905,
      "step": 61
    },
    {
      "epoch": 0.0700762927380616,
      "grad_norm": 3.4652132987976074,
      "learning_rate": 8.262928807620843e-05,
      "loss": 2.3564,
      "step": 62
    },
    {
      "epoch": 0.07120655552415937,
      "grad_norm": 6.234279155731201,
      "learning_rate": 8.199842702516583e-05,
      "loss": 2.3007,
      "step": 63
    },
    {
      "epoch": 0.07233681831025714,
      "grad_norm": 3.932115077972412,
      "learning_rate": 8.135881792367686e-05,
      "loss": 2.4579,
      "step": 64
    },
    {
      "epoch": 0.0734670810963549,
      "grad_norm": 3.8194704055786133,
      "learning_rate": 8.07106356344834e-05,
      "loss": 2.4234,
      "step": 65
    },
    {
      "epoch": 0.07459734388245268,
      "grad_norm": 3.519242763519287,
      "learning_rate": 8.005405736415126e-05,
      "loss": 2.4634,
      "step": 66
    },
    {
      "epoch": 0.07572760666855044,
      "grad_norm": 3.320915699005127,
      "learning_rate": 7.938926261462366e-05,
      "loss": 2.4572,
      "step": 67
    },
    {
      "epoch": 0.0768578694546482,
      "grad_norm": 3.258530855178833,
      "learning_rate": 7.871643313414718e-05,
      "loss": 2.3929,
      "step": 68
    },
    {
      "epoch": 0.07798813224074597,
      "grad_norm": 3.530489206314087,
      "learning_rate": 7.803575286758364e-05,
      "loss": 2.4867,
      "step": 69
    },
    {
      "epoch": 0.07911839502684374,
      "grad_norm": 3.51053524017334,
      "learning_rate": 7.734740790612136e-05,
      "loss": 2.4485,
      "step": 70
    },
    {
      "epoch": 0.0802486578129415,
      "grad_norm": 3.227036714553833,
      "learning_rate": 7.66515864363997e-05,
      "loss": 2.4643,
      "step": 71
    },
    {
      "epoch": 0.08137892059903927,
      "grad_norm": 3.296581268310547,
      "learning_rate": 7.594847868906076e-05,
      "loss": 2.4749,
      "step": 72
    },
    {
      "epoch": 0.08250918338513705,
      "grad_norm": 3.132627487182617,
      "learning_rate": 7.52382768867422e-05,
      "loss": 2.4998,
      "step": 73
    },
    {
      "epoch": 0.08363944617123481,
      "grad_norm": 3.2647902965545654,
      "learning_rate": 7.452117519152542e-05,
      "loss": 2.5533,
      "step": 74
    },
    {
      "epoch": 0.08476970895733257,
      "grad_norm": 3.1851794719696045,
      "learning_rate": 7.379736965185368e-05,
      "loss": 2.49,
      "step": 75
    },
    {
      "epoch": 0.08589997174343035,
      "grad_norm": 4.002132892608643,
      "learning_rate": 7.30670581489344e-05,
      "loss": 2.5861,
      "step": 76
    },
    {
      "epoch": 0.08703023452952811,
      "grad_norm": 4.411466121673584,
      "learning_rate": 7.233044034264034e-05,
      "loss": 2.5512,
      "step": 77
    },
    {
      "epoch": 0.08816049731562588,
      "grad_norm": 3.805506944656372,
      "learning_rate": 7.158771761692464e-05,
      "loss": 2.5293,
      "step": 78
    },
    {
      "epoch": 0.08929076010172365,
      "grad_norm": 4.370881080627441,
      "learning_rate": 7.083909302476453e-05,
      "loss": 2.6055,
      "step": 79
    },
    {
      "epoch": 0.09042102288782142,
      "grad_norm": 3.3669865131378174,
      "learning_rate": 7.008477123264848e-05,
      "loss": 2.5587,
      "step": 80
    },
    {
      "epoch": 0.09155128567391918,
      "grad_norm": 3.5958588123321533,
      "learning_rate": 6.932495846462261e-05,
      "loss": 2.5615,
      "step": 81
    },
    {
      "epoch": 0.09268154846001696,
      "grad_norm": 3.562636375427246,
      "learning_rate": 6.855986244591104e-05,
      "loss": 2.4759,
      "step": 82
    },
    {
      "epoch": 0.09381181124611472,
      "grad_norm": 3.590280055999756,
      "learning_rate": 6.778969234612584e-05,
      "loss": 2.6407,
      "step": 83
    },
    {
      "epoch": 0.09494207403221248,
      "grad_norm": 3.7248311042785645,
      "learning_rate": 6.701465872208216e-05,
      "loss": 2.5758,
      "step": 84
    },
    {
      "epoch": 0.09607233681831026,
      "grad_norm": 3.400334596633911,
      "learning_rate": 6.623497346023418e-05,
      "loss": 2.6469,
      "step": 85
    },
    {
      "epoch": 0.09720259960440802,
      "grad_norm": 3.411473035812378,
      "learning_rate": 6.545084971874738e-05,
      "loss": 2.64,
      "step": 86
    },
    {
      "epoch": 0.09833286239050579,
      "grad_norm": 3.469587802886963,
      "learning_rate": 6.466250186922325e-05,
      "loss": 2.5562,
      "step": 87
    },
    {
      "epoch": 0.09946312517660356,
      "grad_norm": 3.244494915008545,
      "learning_rate": 6.387014543809223e-05,
      "loss": 2.4905,
      "step": 88
    },
    {
      "epoch": 0.10059338796270133,
      "grad_norm": 26.36758041381836,
      "learning_rate": 6.307399704769099e-05,
      "loss": 2.7957,
      "step": 89
    },
    {
      "epoch": 0.10172365074879909,
      "grad_norm": 3.8822689056396484,
      "learning_rate": 6.227427435703997e-05,
      "loss": 2.5604,
      "step": 90
    },
    {
      "epoch": 0.10285391353489687,
      "grad_norm": 4.2484130859375,
      "learning_rate": 6.147119600233758e-05,
      "loss": 2.6036,
      "step": 91
    },
    {
      "epoch": 0.10398417632099463,
      "grad_norm": 4.178469181060791,
      "learning_rate": 6.066498153718735e-05,
      "loss": 2.554,
      "step": 92
    },
    {
      "epoch": 0.1051144391070924,
      "grad_norm": 5.22417688369751,
      "learning_rate": 5.985585137257401e-05,
      "loss": 2.6619,
      "step": 93
    },
    {
      "epoch": 0.10624470189319017,
      "grad_norm": 5.007574558258057,
      "learning_rate": 5.90440267166055e-05,
      "loss": 2.4233,
      "step": 94
    },
    {
      "epoch": 0.10737496467928793,
      "grad_norm": 3.77988600730896,
      "learning_rate": 5.8229729514036705e-05,
      "loss": 2.2586,
      "step": 95
    },
    {
      "epoch": 0.1085052274653857,
      "grad_norm": 3.4278860092163086,
      "learning_rate": 5.74131823855921e-05,
      "loss": 1.8036,
      "step": 96
    },
    {
      "epoch": 0.10963549025148348,
      "grad_norm": 4.095373630523682,
      "learning_rate": 5.6594608567103456e-05,
      "loss": 1.9984,
      "step": 97
    },
    {
      "epoch": 0.11076575303758124,
      "grad_norm": 4.119393348693848,
      "learning_rate": 5.577423184847932e-05,
      "loss": 2.0249,
      "step": 98
    },
    {
      "epoch": 0.111896015823679,
      "grad_norm": 4.134685516357422,
      "learning_rate": 5.495227651252315e-05,
      "loss": 1.7243,
      "step": 99
    },
    {
      "epoch": 0.11302627860977678,
      "grad_norm": 4.909872055053711,
      "learning_rate": 5.4128967273616625e-05,
      "loss": 1.6684,
      "step": 100
    },
    {
      "epoch": 0.11302627860977678,
      "eval_loss": 0.6366060376167297,
      "eval_runtime": 106.627,
      "eval_samples_per_second": 13.974,
      "eval_steps_per_second": 3.498,
      "step": 100
    },
    {
      "epoch": 0.11415654139587454,
      "grad_norm": 16.274993896484375,
      "learning_rate": 5.330452921628497e-05,
      "loss": 2.3684,
      "step": 101
    },
    {
      "epoch": 0.1152868041819723,
      "grad_norm": 4.304095268249512,
      "learning_rate": 5.247918773366112e-05,
      "loss": 2.0797,
      "step": 102
    },
    {
      "epoch": 0.11641706696807008,
      "grad_norm": 3.7952427864074707,
      "learning_rate": 5.165316846586541e-05,
      "loss": 2.1863,
      "step": 103
    },
    {
      "epoch": 0.11754732975416785,
      "grad_norm": 3.1484949588775635,
      "learning_rate": 5.0826697238317935e-05,
      "loss": 2.0027,
      "step": 104
    },
    {
      "epoch": 0.11867759254026561,
      "grad_norm": 3.438645124435425,
      "learning_rate": 5e-05,
      "loss": 2.0559,
      "step": 105
    },
    {
      "epoch": 0.11980785532636339,
      "grad_norm": 2.8130788803100586,
      "learning_rate": 4.917330276168208e-05,
      "loss": 2.0134,
      "step": 106
    },
    {
      "epoch": 0.12093811811246115,
      "grad_norm": 3.237912893295288,
      "learning_rate": 4.834683153413459e-05,
      "loss": 2.1676,
      "step": 107
    },
    {
      "epoch": 0.12206838089855891,
      "grad_norm": 3.323214530944824,
      "learning_rate": 4.7520812266338885e-05,
      "loss": 2.1133,
      "step": 108
    },
    {
      "epoch": 0.12319864368465669,
      "grad_norm": 4.096421718597412,
      "learning_rate": 4.669547078371504e-05,
      "loss": 2.1645,
      "step": 109
    },
    {
      "epoch": 0.12432890647075445,
      "grad_norm": 3.973909378051758,
      "learning_rate": 4.5871032726383386e-05,
      "loss": 2.1192,
      "step": 110
    },
    {
      "epoch": 0.12545916925685222,
      "grad_norm": 3.9636118412017822,
      "learning_rate": 4.504772348747687e-05,
      "loss": 2.2094,
      "step": 111
    },
    {
      "epoch": 0.12658943204294998,
      "grad_norm": 4.142837047576904,
      "learning_rate": 4.4225768151520694e-05,
      "loss": 2.185,
      "step": 112
    },
    {
      "epoch": 0.12771969482904774,
      "grad_norm": 4.369409084320068,
      "learning_rate": 4.3405391432896555e-05,
      "loss": 2.1121,
      "step": 113
    },
    {
      "epoch": 0.12884995761514553,
      "grad_norm": 3.398538827896118,
      "learning_rate": 4.2586817614407895e-05,
      "loss": 2.1533,
      "step": 114
    },
    {
      "epoch": 0.1299802204012433,
      "grad_norm": 3.4510021209716797,
      "learning_rate": 4.17702704859633e-05,
      "loss": 2.2089,
      "step": 115
    },
    {
      "epoch": 0.13111048318734106,
      "grad_norm": 3.939387559890747,
      "learning_rate": 4.095597328339452e-05,
      "loss": 2.17,
      "step": 116
    },
    {
      "epoch": 0.13224074597343882,
      "grad_norm": 3.7233493328094482,
      "learning_rate": 4.0144148627425993e-05,
      "loss": 2.1368,
      "step": 117
    },
    {
      "epoch": 0.13337100875953659,
      "grad_norm": 4.076516628265381,
      "learning_rate": 3.933501846281267e-05,
      "loss": 2.1025,
      "step": 118
    },
    {
      "epoch": 0.13450127154563435,
      "grad_norm": 4.5438761711120605,
      "learning_rate": 3.852880399766243e-05,
      "loss": 2.2489,
      "step": 119
    },
    {
      "epoch": 0.13563153433173214,
      "grad_norm": 3.5553297996520996,
      "learning_rate": 3.772572564296005e-05,
      "loss": 2.1149,
      "step": 120
    },
    {
      "epoch": 0.1367617971178299,
      "grad_norm": 4.277527809143066,
      "learning_rate": 3.6926002952309016e-05,
      "loss": 2.0641,
      "step": 121
    },
    {
      "epoch": 0.13789205990392767,
      "grad_norm": 3.79026198387146,
      "learning_rate": 3.612985456190778e-05,
      "loss": 2.0996,
      "step": 122
    },
    {
      "epoch": 0.13902232269002543,
      "grad_norm": 5.075240135192871,
      "learning_rate": 3.533749813077677e-05,
      "loss": 2.2164,
      "step": 123
    },
    {
      "epoch": 0.1401525854761232,
      "grad_norm": 4.374073028564453,
      "learning_rate": 3.4549150281252636e-05,
      "loss": 2.2146,
      "step": 124
    },
    {
      "epoch": 0.14128284826222096,
      "grad_norm": 4.413872241973877,
      "learning_rate": 3.3765026539765834e-05,
      "loss": 2.4203,
      "step": 125
    },
    {
      "epoch": 0.14241311104831875,
      "grad_norm": 4.2351298332214355,
      "learning_rate": 3.298534127791785e-05,
      "loss": 2.1717,
      "step": 126
    },
    {
      "epoch": 0.1435433738344165,
      "grad_norm": 3.9337973594665527,
      "learning_rate": 3.221030765387417e-05,
      "loss": 2.208,
      "step": 127
    },
    {
      "epoch": 0.14467363662051427,
      "grad_norm": 4.095107555389404,
      "learning_rate": 3.144013755408895e-05,
      "loss": 2.1778,
      "step": 128
    },
    {
      "epoch": 0.14580389940661204,
      "grad_norm": 4.108471870422363,
      "learning_rate": 3.0675041535377405e-05,
      "loss": 2.2488,
      "step": 129
    },
    {
      "epoch": 0.1469341621927098,
      "grad_norm": 4.01255464553833,
      "learning_rate": 2.991522876735154e-05,
      "loss": 2.2685,
      "step": 130
    },
    {
      "epoch": 0.14806442497880756,
      "grad_norm": 4.220801830291748,
      "learning_rate": 2.916090697523549e-05,
      "loss": 2.2855,
      "step": 131
    },
    {
      "epoch": 0.14919468776490535,
      "grad_norm": 4.091708183288574,
      "learning_rate": 2.8412282383075363e-05,
      "loss": 2.3577,
      "step": 132
    },
    {
      "epoch": 0.15032495055100312,
      "grad_norm": 4.274747848510742,
      "learning_rate": 2.766955965735968e-05,
      "loss": 2.2369,
      "step": 133
    },
    {
      "epoch": 0.15145521333710088,
      "grad_norm": 3.8950018882751465,
      "learning_rate": 2.693294185106562e-05,
      "loss": 2.211,
      "step": 134
    },
    {
      "epoch": 0.15258547612319864,
      "grad_norm": 4.154726982116699,
      "learning_rate": 2.6202630348146324e-05,
      "loss": 2.308,
      "step": 135
    },
    {
      "epoch": 0.1537157389092964,
      "grad_norm": 4.183905601501465,
      "learning_rate": 2.547882480847461e-05,
      "loss": 2.2896,
      "step": 136
    },
    {
      "epoch": 0.15484600169539417,
      "grad_norm": 3.8550608158111572,
      "learning_rate": 2.476172311325783e-05,
      "loss": 2.2892,
      "step": 137
    },
    {
      "epoch": 0.15597626448149193,
      "grad_norm": 4.267631530761719,
      "learning_rate": 2.405152131093926e-05,
      "loss": 2.2431,
      "step": 138
    },
    {
      "epoch": 0.15710652726758972,
      "grad_norm": 4.062099456787109,
      "learning_rate": 2.3348413563600325e-05,
      "loss": 2.3272,
      "step": 139
    },
    {
      "epoch": 0.1582367900536875,
      "grad_norm": 4.087035655975342,
      "learning_rate": 2.2652592093878666e-05,
      "loss": 2.2581,
      "step": 140
    },
    {
      "epoch": 0.15936705283978525,
      "grad_norm": 4.308664798736572,
      "learning_rate": 2.196424713241637e-05,
      "loss": 2.3231,
      "step": 141
    },
    {
      "epoch": 0.160497315625883,
      "grad_norm": 4.159082412719727,
      "learning_rate": 2.128356686585282e-05,
      "loss": 2.2854,
      "step": 142
    },
    {
      "epoch": 0.16162757841198078,
      "grad_norm": 4.102118015289307,
      "learning_rate": 2.061073738537635e-05,
      "loss": 2.3272,
      "step": 143
    },
    {
      "epoch": 0.16275784119807854,
      "grad_norm": 4.1677117347717285,
      "learning_rate": 1.9945942635848748e-05,
      "loss": 2.175,
      "step": 144
    },
    {
      "epoch": 0.16388810398417633,
      "grad_norm": 4.319460868835449,
      "learning_rate": 1.928936436551661e-05,
      "loss": 2.1601,
      "step": 145
    },
    {
      "epoch": 0.1650183667702741,
      "grad_norm": 3.8885021209716797,
      "learning_rate": 1.8641182076323148e-05,
      "loss": 1.8595,
      "step": 146
    },
    {
      "epoch": 0.16614862955637186,
      "grad_norm": 4.126462936401367,
      "learning_rate": 1.800157297483417e-05,
      "loss": 1.9056,
      "step": 147
    },
    {
      "epoch": 0.16727889234246962,
      "grad_norm": 3.3327112197875977,
      "learning_rate": 1.7370711923791567e-05,
      "loss": 1.1968,
      "step": 148
    },
    {
      "epoch": 0.16840915512856738,
      "grad_norm": 4.65377140045166,
      "learning_rate": 1.6748771394307585e-05,
      "loss": 1.5292,
      "step": 149
    },
    {
      "epoch": 0.16953941791466515,
      "grad_norm": 4.763574123382568,
      "learning_rate": 1.6135921418712956e-05,
      "loss": 1.6376,
      "step": 150
    },
    {
      "epoch": 0.16953941791466515,
      "eval_loss": 0.5219109058380127,
      "eval_runtime": 106.5495,
      "eval_samples_per_second": 13.984,
      "eval_steps_per_second": 3.501,
      "step": 150
    },
    {
      "epoch": 0.17066968070076294,
      "grad_norm": 4.40399169921875,
      "learning_rate": 1.553232954407171e-05,
      "loss": 1.7138,
      "step": 151
    },
    {
      "epoch": 0.1717999434868607,
      "grad_norm": 12.577033042907715,
      "learning_rate": 1.4938160786375572e-05,
      "loss": 1.8675,
      "step": 152
    },
    {
      "epoch": 0.17293020627295846,
      "grad_norm": 3.7592997550964355,
      "learning_rate": 1.435357758543015e-05,
      "loss": 1.8429,
      "step": 153
    },
    {
      "epoch": 0.17406046905905623,
      "grad_norm": 3.5341546535491943,
      "learning_rate": 1.3778739760445552e-05,
      "loss": 1.848,
      "step": 154
    },
    {
      "epoch": 0.175190731845154,
      "grad_norm": 3.8451528549194336,
      "learning_rate": 1.3213804466343421e-05,
      "loss": 1.8224,
      "step": 155
    },
    {
      "epoch": 0.17632099463125175,
      "grad_norm": 3.349514961242676,
      "learning_rate": 1.2658926150792322e-05,
      "loss": 1.8962,
      "step": 156
    },
    {
      "epoch": 0.17745125741734954,
      "grad_norm": 3.187197685241699,
      "learning_rate": 1.2114256511983274e-05,
      "loss": 1.8941,
      "step": 157
    },
    {
      "epoch": 0.1785815202034473,
      "grad_norm": 3.436824083328247,
      "learning_rate": 1.157994445715706e-05,
      "loss": 1.9308,
      "step": 158
    },
    {
      "epoch": 0.17971178298954507,
      "grad_norm": 3.1081900596618652,
      "learning_rate": 1.1056136061894384e-05,
      "loss": 1.9113,
      "step": 159
    },
    {
      "epoch": 0.18084204577564283,
      "grad_norm": 3.790692090988159,
      "learning_rate": 1.0542974530180327e-05,
      "loss": 1.9905,
      "step": 160
    },
    {
      "epoch": 0.1819723085617406,
      "grad_norm": 3.599813461303711,
      "learning_rate": 1.0040600155253765e-05,
      "loss": 1.9209,
      "step": 161
    },
    {
      "epoch": 0.18310257134783836,
      "grad_norm": 3.283684730529785,
      "learning_rate": 9.549150281252633e-06,
      "loss": 1.9674,
      "step": 162
    },
    {
      "epoch": 0.18423283413393615,
      "grad_norm": 3.6617507934570312,
      "learning_rate": 9.068759265665384e-06,
      "loss": 1.9887,
      "step": 163
    },
    {
      "epoch": 0.18536309692003392,
      "grad_norm": 3.457580089569092,
      "learning_rate": 8.599558442598998e-06,
      "loss": 1.9888,
      "step": 164
    },
    {
      "epoch": 0.18649335970613168,
      "grad_norm": 3.2257258892059326,
      "learning_rate": 8.141676086873572e-06,
      "loss": 1.9991,
      "step": 165
    },
    {
      "epoch": 0.18762362249222944,
      "grad_norm": 3.4789462089538574,
      "learning_rate": 7.695237378953223e-06,
      "loss": 1.9922,
      "step": 166
    },
    {
      "epoch": 0.1887538852783272,
      "grad_norm": 3.293551445007324,
      "learning_rate": 7.260364370723044e-06,
      "loss": 2.0187,
      "step": 167
    },
    {
      "epoch": 0.18988414806442497,
      "grad_norm": 3.351811408996582,
      "learning_rate": 6.837175952121306e-06,
      "loss": 1.9505,
      "step": 168
    },
    {
      "epoch": 0.19101441085052276,
      "grad_norm": 3.2482376098632812,
      "learning_rate": 6.425787818636131e-06,
      "loss": 1.9509,
      "step": 169
    },
    {
      "epoch": 0.19214467363662052,
      "grad_norm": 3.266819715499878,
      "learning_rate": 6.026312439675552e-06,
      "loss": 2.0122,
      "step": 170
    },
    {
      "epoch": 0.19327493642271829,
      "grad_norm": 3.631998300552368,
      "learning_rate": 5.6388590278194096e-06,
      "loss": 1.9606,
      "step": 171
    },
    {
      "epoch": 0.19440519920881605,
      "grad_norm": 3.7456886768341064,
      "learning_rate": 5.263533508961827e-06,
      "loss": 2.0287,
      "step": 172
    },
    {
      "epoch": 0.1955354619949138,
      "grad_norm": 3.2755744457244873,
      "learning_rate": 4.900438493352055e-06,
      "loss": 1.9983,
      "step": 173
    },
    {
      "epoch": 0.19666572478101157,
      "grad_norm": 3.4375557899475098,
      "learning_rate": 4.549673247541875e-06,
      "loss": 2.0436,
      "step": 174
    },
    {
      "epoch": 0.19779598756710937,
      "grad_norm": 3.4069182872772217,
      "learning_rate": 4.2113336672471245e-06,
      "loss": 2.0593,
      "step": 175
    },
    {
      "epoch": 0.19892625035320713,
      "grad_norm": 3.283660650253296,
      "learning_rate": 3.885512251130763e-06,
      "loss": 2.0617,
      "step": 176
    },
    {
      "epoch": 0.2000565131393049,
      "grad_norm": 3.3320672512054443,
      "learning_rate": 3.5722980755146517e-06,
      "loss": 2.0627,
      "step": 177
    },
    {
      "epoch": 0.20118677592540266,
      "grad_norm": 3.507112741470337,
      "learning_rate": 3.271776770026963e-06,
      "loss": 2.0202,
      "step": 178
    },
    {
      "epoch": 0.20231703871150042,
      "grad_norm": 3.4486937522888184,
      "learning_rate": 2.9840304941919415e-06,
      "loss": 2.0541,
      "step": 179
    },
    {
      "epoch": 0.20344730149759818,
      "grad_norm": 3.7197585105895996,
      "learning_rate": 2.7091379149682685e-06,
      "loss": 2.1123,
      "step": 180
    },
    {
      "epoch": 0.20457756428369597,
      "grad_norm": 3.5073843002319336,
      "learning_rate": 2.4471741852423237e-06,
      "loss": 2.1078,
      "step": 181
    },
    {
      "epoch": 0.20570782706979374,
      "grad_norm": 3.6182987689971924,
      "learning_rate": 2.1982109232821178e-06,
      "loss": 2.156,
      "step": 182
    },
    {
      "epoch": 0.2068380898558915,
      "grad_norm": 3.3913204669952393,
      "learning_rate": 1.962316193157593e-06,
      "loss": 2.0867,
      "step": 183
    },
    {
      "epoch": 0.20796835264198926,
      "grad_norm": 3.729597568511963,
      "learning_rate": 1.7395544861325718e-06,
      "loss": 2.1191,
      "step": 184
    },
    {
      "epoch": 0.20909861542808703,
      "grad_norm": 4.039198875427246,
      "learning_rate": 1.5299867030334814e-06,
      "loss": 2.1348,
      "step": 185
    },
    {
      "epoch": 0.2102288782141848,
      "grad_norm": 3.6647441387176514,
      "learning_rate": 1.333670137599713e-06,
      "loss": 2.1544,
      "step": 186
    },
    {
      "epoch": 0.21135914100028255,
      "grad_norm": 3.687973976135254,
      "learning_rate": 1.1506584608200367e-06,
      "loss": 2.1856,
      "step": 187
    },
    {
      "epoch": 0.21248940378638034,
      "grad_norm": 4.044662952423096,
      "learning_rate": 9.810017062595322e-07,
      "loss": 2.1615,
      "step": 188
    },
    {
      "epoch": 0.2136196665724781,
      "grad_norm": 3.7878575325012207,
      "learning_rate": 8.247462563808817e-07,
      "loss": 2.1719,
      "step": 189
    },
    {
      "epoch": 0.21474992935857587,
      "grad_norm": 3.7030394077301025,
      "learning_rate": 6.819348298638839e-07,
      "loss": 2.2528,
      "step": 190
    },
    {
      "epoch": 0.21588019214467363,
      "grad_norm": 4.257616996765137,
      "learning_rate": 5.526064699265753e-07,
      "loss": 2.1805,
      "step": 191
    },
    {
      "epoch": 0.2170104549307714,
      "grad_norm": 3.8384816646575928,
      "learning_rate": 4.367965336512403e-07,
      "loss": 2.2072,
      "step": 192
    },
    {
      "epoch": 0.21814071771686916,
      "grad_norm": 4.030503273010254,
      "learning_rate": 3.3453668231809286e-07,
      "loss": 2.1861,
      "step": 193
    },
    {
      "epoch": 0.21927098050296695,
      "grad_norm": 3.915443181991577,
      "learning_rate": 2.458548727494292e-07,
      "loss": 2.1201,
      "step": 194
    },
    {
      "epoch": 0.2204012432890647,
      "grad_norm": 3.867893695831299,
      "learning_rate": 1.7077534966650766e-07,
      "loss": 2.0177,
      "step": 195
    },
    {
      "epoch": 0.22153150607516248,
      "grad_norm": 3.743412733078003,
      "learning_rate": 1.0931863906127327e-07,
      "loss": 1.679,
      "step": 196
    },
    {
      "epoch": 0.22266176886126024,
      "grad_norm": 4.369097709655762,
      "learning_rate": 6.150154258476315e-08,
      "loss": 2.0535,
      "step": 197
    },
    {
      "epoch": 0.223792031647358,
      "grad_norm": 3.6869876384735107,
      "learning_rate": 2.7337132953697554e-08,
      "loss": 1.532,
      "step": 198
    },
    {
      "epoch": 0.22492229443345577,
      "grad_norm": 4.66172981262207,
      "learning_rate": 6.834750376549792e-09,
      "loss": 1.6097,
      "step": 199
    },
    {
      "epoch": 0.22605255721955356,
      "grad_norm": 4.779449939727783,
      "learning_rate": 0.0,
      "loss": 1.6369,
      "step": 200
    },
    {
      "epoch": 0.22605255721955356,
      "eval_loss": 0.49133214354515076,
      "eval_runtime": 106.6213,
      "eval_samples_per_second": 13.975,
      "eval_steps_per_second": 3.498,
      "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": 2.863232968556544e+17,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}