{
  "best_metric": 2.5217931270599365,
  "best_model_checkpoint": "miner_id_24/checkpoint-200",
  "epoch": 0.10509721492380451,
  "eval_steps": 50,
  "global_step": 200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0005254860746190226,
      "grad_norm": 5.3695268630981445,
      "learning_rate": 1e-05,
      "loss": 2.8017,
      "step": 1
    },
    {
      "epoch": 0.0005254860746190226,
      "eval_loss": 4.163503646850586,
      "eval_runtime": 63.816,
      "eval_samples_per_second": 50.223,
      "eval_steps_per_second": 12.567,
      "step": 1
    },
    {
      "epoch": 0.0010509721492380452,
      "grad_norm": 10.700703620910645,
      "learning_rate": 2e-05,
      "loss": 3.0632,
      "step": 2
    },
    {
      "epoch": 0.0015764582238570678,
      "grad_norm": 4.7041497230529785,
      "learning_rate": 3e-05,
      "loss": 2.9208,
      "step": 3
    },
    {
      "epoch": 0.0021019442984760903,
      "grad_norm": 4.3920183181762695,
      "learning_rate": 4e-05,
      "loss": 3.2519,
      "step": 4
    },
    {
      "epoch": 0.002627430373095113,
      "grad_norm": 3.264192819595337,
      "learning_rate": 5e-05,
      "loss": 3.1685,
      "step": 5
    },
    {
      "epoch": 0.0031529164477141357,
      "grad_norm": 2.670915126800537,
      "learning_rate": 6e-05,
      "loss": 3.0917,
      "step": 6
    },
    {
      "epoch": 0.0036784025223331584,
      "grad_norm": 3.014742136001587,
      "learning_rate": 7e-05,
      "loss": 2.9348,
      "step": 7
    },
    {
      "epoch": 0.004203888596952181,
      "grad_norm": 2.923550605773926,
      "learning_rate": 8e-05,
      "loss": 2.9486,
      "step": 8
    },
    {
      "epoch": 0.004729374671571204,
      "grad_norm": 2.537583827972412,
      "learning_rate": 9e-05,
      "loss": 2.9978,
      "step": 9
    },
    {
      "epoch": 0.005254860746190226,
      "grad_norm": 2.5933568477630615,
      "learning_rate": 0.0001,
      "loss": 2.7744,
      "step": 10
    },
    {
      "epoch": 0.005780346820809248,
      "grad_norm": 5.803947925567627,
      "learning_rate": 9.999316524962345e-05,
      "loss": 3.2291,
      "step": 11
    },
    {
      "epoch": 0.006305832895428271,
      "grad_norm": 14.510558128356934,
      "learning_rate": 9.997266286704631e-05,
      "loss": 2.6844,
      "step": 12
    },
    {
      "epoch": 0.006831318970047294,
      "grad_norm": 3.439263105392456,
      "learning_rate": 9.993849845741524e-05,
      "loss": 2.9335,
      "step": 13
    },
    {
      "epoch": 0.007356805044666317,
      "grad_norm": 5.007378101348877,
      "learning_rate": 9.989068136093873e-05,
      "loss": 2.7506,
      "step": 14
    },
    {
      "epoch": 0.007882291119285338,
      "grad_norm": 11.0796480178833,
      "learning_rate": 9.98292246503335e-05,
      "loss": 2.9845,
      "step": 15
    },
    {
      "epoch": 0.008407777193904361,
      "grad_norm": 7.752804279327393,
      "learning_rate": 9.975414512725057e-05,
      "loss": 2.5332,
      "step": 16
    },
    {
      "epoch": 0.008933263268523384,
      "grad_norm": 6.898101329803467,
      "learning_rate": 9.966546331768191e-05,
      "loss": 3.0547,
      "step": 17
    },
    {
      "epoch": 0.009458749343142408,
      "grad_norm": 6.097688674926758,
      "learning_rate": 9.956320346634876e-05,
      "loss": 2.9143,
      "step": 18
    },
    {
      "epoch": 0.009984235417761429,
      "grad_norm": 2.7763419151306152,
      "learning_rate": 9.944739353007344e-05,
      "loss": 2.848,
      "step": 19
    },
    {
      "epoch": 0.010509721492380452,
      "grad_norm": 4.871918201446533,
      "learning_rate": 9.931806517013612e-05,
      "loss": 2.7329,
      "step": 20
    },
    {
      "epoch": 0.011035207566999475,
      "grad_norm": 2.282667875289917,
      "learning_rate": 9.917525374361912e-05,
      "loss": 2.4997,
      "step": 21
    },
    {
      "epoch": 0.011560693641618497,
      "grad_norm": 2.4404683113098145,
      "learning_rate": 9.901899829374047e-05,
      "loss": 2.8231,
      "step": 22
    },
    {
      "epoch": 0.01208617971623752,
      "grad_norm": 2.603766441345215,
      "learning_rate": 9.884934153917997e-05,
      "loss": 2.7033,
      "step": 23
    },
    {
      "epoch": 0.012611665790856543,
      "grad_norm": 3.9431021213531494,
      "learning_rate": 9.86663298624003e-05,
      "loss": 2.6877,
      "step": 24
    },
    {
      "epoch": 0.013137151865475564,
      "grad_norm": 2.8732380867004395,
      "learning_rate": 9.847001329696653e-05,
      "loss": 2.4128,
      "step": 25
    },
    {
      "epoch": 0.013662637940094587,
      "grad_norm": 2.627973794937134,
      "learning_rate": 9.826044551386744e-05,
      "loss": 2.5293,
      "step": 26
    },
    {
      "epoch": 0.01418812401471361,
      "grad_norm": 2.4622251987457275,
      "learning_rate": 9.803768380684242e-05,
      "loss": 2.5152,
      "step": 27
    },
    {
      "epoch": 0.014713610089332634,
      "grad_norm": 2.9161486625671387,
      "learning_rate": 9.780178907671789e-05,
      "loss": 2.614,
      "step": 28
    },
    {
      "epoch": 0.015239096163951655,
      "grad_norm": 2.668250799179077,
      "learning_rate": 9.755282581475769e-05,
      "loss": 2.5832,
      "step": 29
    },
    {
      "epoch": 0.015764582238570676,
      "grad_norm": 2.8948163986206055,
      "learning_rate": 9.729086208503174e-05,
      "loss": 2.8288,
      "step": 30
    },
    {
      "epoch": 0.0162900683131897,
      "grad_norm": 3.167529582977295,
      "learning_rate": 9.701596950580806e-05,
      "loss": 3.1371,
      "step": 31
    },
    {
      "epoch": 0.016815554387808723,
      "grad_norm": 5.602165222167969,
      "learning_rate": 9.672822322997305e-05,
      "loss": 2.439,
      "step": 32
    },
    {
      "epoch": 0.017341040462427744,
      "grad_norm": 3.580030679702759,
      "learning_rate": 9.642770192448536e-05,
      "loss": 2.6022,
      "step": 33
    },
    {
      "epoch": 0.01786652653704677,
      "grad_norm": 2.9419171810150146,
      "learning_rate": 9.611448774886924e-05,
      "loss": 2.4721,
      "step": 34
    },
    {
      "epoch": 0.01839201261166579,
      "grad_norm": 3.6252121925354004,
      "learning_rate": 9.578866633275288e-05,
      "loss": 2.6317,
      "step": 35
    },
    {
      "epoch": 0.018917498686284815,
      "grad_norm": 3.8813490867614746,
      "learning_rate": 9.545032675245813e-05,
      "loss": 2.8856,
      "step": 36
    },
    {
      "epoch": 0.019442984760903836,
      "grad_norm": 2.939866781234741,
      "learning_rate": 9.509956150664796e-05,
      "loss": 2.7042,
      "step": 37
    },
    {
      "epoch": 0.019968470835522858,
      "grad_norm": 3.091820240020752,
      "learning_rate": 9.473646649103818e-05,
      "loss": 2.6439,
      "step": 38
    },
    {
      "epoch": 0.020493956910141883,
      "grad_norm": 6.278568267822266,
      "learning_rate": 9.43611409721806e-05,
      "loss": 2.767,
      "step": 39
    },
    {
      "epoch": 0.021019442984760904,
      "grad_norm": 6.981184005737305,
      "learning_rate": 9.397368756032445e-05,
      "loss": 2.4893,
      "step": 40
    },
    {
      "epoch": 0.021544929059379925,
      "grad_norm": 15.421754837036133,
      "learning_rate": 9.357421218136386e-05,
      "loss": 2.6959,
      "step": 41
    },
    {
      "epoch": 0.02207041513399895,
      "grad_norm": 3.8115768432617188,
      "learning_rate": 9.316282404787871e-05,
      "loss": 2.5861,
      "step": 42
    },
    {
      "epoch": 0.02259590120861797,
      "grad_norm": 3.3122410774230957,
      "learning_rate": 9.273963562927695e-05,
      "loss": 2.707,
      "step": 43
    },
    {
      "epoch": 0.023121387283236993,
      "grad_norm": 8.810725212097168,
      "learning_rate": 9.230476262104677e-05,
      "loss": 2.6911,
      "step": 44
    },
    {
      "epoch": 0.023646873357856018,
      "grad_norm": 3.6674933433532715,
      "learning_rate": 9.185832391312644e-05,
      "loss": 2.6285,
      "step": 45
    },
    {
      "epoch": 0.02417235943247504,
      "grad_norm": 3.751573324203491,
      "learning_rate": 9.140044155740101e-05,
      "loss": 2.5582,
      "step": 46
    },
    {
      "epoch": 0.02469784550709406,
      "grad_norm": 3.4803454875946045,
      "learning_rate": 9.093124073433463e-05,
      "loss": 2.6075,
      "step": 47
    },
    {
      "epoch": 0.025223331581713086,
      "grad_norm": 3.6012983322143555,
      "learning_rate": 9.045084971874738e-05,
      "loss": 2.6599,
      "step": 48
    },
    {
      "epoch": 0.025748817656332107,
      "grad_norm": 14.146845817565918,
      "learning_rate": 8.995939984474624e-05,
      "loss": 2.7609,
      "step": 49
    },
    {
      "epoch": 0.02627430373095113,
      "grad_norm": 6.5617570877075195,
      "learning_rate": 8.945702546981969e-05,
      "loss": 2.7783,
      "step": 50
    },
    {
      "epoch": 0.02627430373095113,
      "eval_loss": 3.0442087650299072,
      "eval_runtime": 63.8043,
      "eval_samples_per_second": 50.232,
      "eval_steps_per_second": 12.57,
      "step": 50
    },
    {
      "epoch": 0.026799789805570153,
      "grad_norm": 9.739975929260254,
      "learning_rate": 8.894386393810563e-05,
      "loss": 3.1685,
      "step": 51
    },
    {
      "epoch": 0.027325275880189175,
      "grad_norm": 7.816197872161865,
      "learning_rate": 8.842005554284296e-05,
      "loss": 3.0739,
      "step": 52
    },
    {
      "epoch": 0.027850761954808196,
      "grad_norm": 4.996708393096924,
      "learning_rate": 8.788574348801675e-05,
      "loss": 3.0167,
      "step": 53
    },
    {
      "epoch": 0.02837624802942722,
      "grad_norm": 4.00170373916626,
      "learning_rate": 8.73410738492077e-05,
      "loss": 2.7979,
      "step": 54
    },
    {
      "epoch": 0.028901734104046242,
      "grad_norm": 2.1249988079071045,
      "learning_rate": 8.678619553365659e-05,
      "loss": 2.6787,
      "step": 55
    },
    {
      "epoch": 0.029427220178665267,
      "grad_norm": 3.2888216972351074,
      "learning_rate": 8.622126023955446e-05,
      "loss": 2.4514,
      "step": 56
    },
    {
      "epoch": 0.02995270625328429,
      "grad_norm": 1.8928073644638062,
      "learning_rate": 8.564642241456986e-05,
      "loss": 2.5235,
      "step": 57
    },
    {
      "epoch": 0.03047819232790331,
      "grad_norm": 2.292433261871338,
      "learning_rate": 8.506183921362443e-05,
      "loss": 2.6104,
      "step": 58
    },
    {
      "epoch": 0.031003678402522335,
      "grad_norm": 1.8913743495941162,
      "learning_rate": 8.44676704559283e-05,
      "loss": 2.4348,
      "step": 59
    },
    {
      "epoch": 0.03152916447714135,
      "grad_norm": 1.8357709646224976,
      "learning_rate": 8.386407858128706e-05,
      "loss": 2.6611,
      "step": 60
    },
    {
      "epoch": 0.03205465055176038,
      "grad_norm": 1.7449949979782104,
      "learning_rate": 8.32512286056924e-05,
      "loss": 2.6,
      "step": 61
    },
    {
      "epoch": 0.0325801366263794,
      "grad_norm": 1.6825581789016724,
      "learning_rate": 8.262928807620843e-05,
      "loss": 2.3776,
      "step": 62
    },
    {
      "epoch": 0.03310562270099842,
      "grad_norm": 2.3304495811462402,
      "learning_rate": 8.199842702516583e-05,
      "loss": 2.7517,
      "step": 63
    },
    {
      "epoch": 0.033631108775617445,
      "grad_norm": 2.3837075233459473,
      "learning_rate": 8.135881792367686e-05,
      "loss": 2.5219,
      "step": 64
    },
    {
      "epoch": 0.03415659485023647,
      "grad_norm": 1.67093825340271,
      "learning_rate": 8.07106356344834e-05,
      "loss": 2.5317,
      "step": 65
    },
    {
      "epoch": 0.03468208092485549,
      "grad_norm": 1.6712074279785156,
      "learning_rate": 8.005405736415126e-05,
      "loss": 2.5434,
      "step": 66
    },
    {
      "epoch": 0.03520756699947451,
      "grad_norm": 1.8315536975860596,
      "learning_rate": 7.938926261462366e-05,
      "loss": 2.4476,
      "step": 67
    },
    {
      "epoch": 0.03573305307409354,
      "grad_norm": 1.8458882570266724,
      "learning_rate": 7.871643313414718e-05,
      "loss": 2.6224,
      "step": 68
    },
    {
      "epoch": 0.03625853914871256,
      "grad_norm": 1.7561705112457275,
      "learning_rate": 7.803575286758364e-05,
      "loss": 2.5659,
      "step": 69
    },
    {
      "epoch": 0.03678402522333158,
      "grad_norm": 1.8314167261123657,
      "learning_rate": 7.734740790612136e-05,
      "loss": 2.5786,
      "step": 70
    },
    {
      "epoch": 0.037309511297950605,
      "grad_norm": 1.7386704683303833,
      "learning_rate": 7.66515864363997e-05,
      "loss": 2.7321,
      "step": 71
    },
    {
      "epoch": 0.03783499737256963,
      "grad_norm": 1.6980810165405273,
      "learning_rate": 7.594847868906076e-05,
      "loss": 2.5279,
      "step": 72
    },
    {
      "epoch": 0.03836048344718865,
      "grad_norm": 1.7262290716171265,
      "learning_rate": 7.52382768867422e-05,
      "loss": 2.3427,
      "step": 73
    },
    {
      "epoch": 0.03888596952180767,
      "grad_norm": 2.3134450912475586,
      "learning_rate": 7.452117519152542e-05,
      "loss": 2.611,
      "step": 74
    },
    {
      "epoch": 0.0394114555964267,
      "grad_norm": 2.393048048019409,
      "learning_rate": 7.379736965185368e-05,
      "loss": 2.3574,
      "step": 75
    },
    {
      "epoch": 0.039936941671045716,
      "grad_norm": 2.068819522857666,
      "learning_rate": 7.30670581489344e-05,
      "loss": 2.4459,
      "step": 76
    },
    {
      "epoch": 0.04046242774566474,
      "grad_norm": 1.764756679534912,
      "learning_rate": 7.233044034264034e-05,
      "loss": 2.4598,
      "step": 77
    },
    {
      "epoch": 0.040987913820283765,
      "grad_norm": 3.356853485107422,
      "learning_rate": 7.158771761692464e-05,
      "loss": 2.3022,
      "step": 78
    },
    {
      "epoch": 0.04151339989490278,
      "grad_norm": 2.490696430206299,
      "learning_rate": 7.083909302476453e-05,
      "loss": 2.5217,
      "step": 79
    },
    {
      "epoch": 0.04203888596952181,
      "grad_norm": 1.7186123132705688,
      "learning_rate": 7.008477123264848e-05,
      "loss": 2.3811,
      "step": 80
    },
    {
      "epoch": 0.04256437204414083,
      "grad_norm": 1.9211206436157227,
      "learning_rate": 6.932495846462261e-05,
      "loss": 2.4413,
      "step": 81
    },
    {
      "epoch": 0.04308985811875985,
      "grad_norm": 2.2291693687438965,
      "learning_rate": 6.855986244591104e-05,
      "loss": 2.6936,
      "step": 82
    },
    {
      "epoch": 0.043615344193378876,
      "grad_norm": 2.4612324237823486,
      "learning_rate": 6.778969234612584e-05,
      "loss": 2.5216,
      "step": 83
    },
    {
      "epoch": 0.0441408302679979,
      "grad_norm": 1.9075177907943726,
      "learning_rate": 6.701465872208216e-05,
      "loss": 2.4772,
      "step": 84
    },
    {
      "epoch": 0.04466631634261692,
      "grad_norm": 3.057823896408081,
      "learning_rate": 6.623497346023418e-05,
      "loss": 2.5391,
      "step": 85
    },
    {
      "epoch": 0.04519180241723594,
      "grad_norm": 3.4438459873199463,
      "learning_rate": 6.545084971874738e-05,
      "loss": 2.3459,
      "step": 86
    },
    {
      "epoch": 0.04571728849185497,
      "grad_norm": 1.949404239654541,
      "learning_rate": 6.466250186922325e-05,
      "loss": 2.3725,
      "step": 87
    },
    {
      "epoch": 0.046242774566473986,
      "grad_norm": 3.3578133583068848,
      "learning_rate": 6.387014543809223e-05,
      "loss": 2.7464,
      "step": 88
    },
    {
      "epoch": 0.04676826064109301,
      "grad_norm": 3.063342571258545,
      "learning_rate": 6.307399704769099e-05,
      "loss": 2.661,
      "step": 89
    },
    {
      "epoch": 0.047293746715712036,
      "grad_norm": 4.749983787536621,
      "learning_rate": 6.227427435703997e-05,
      "loss": 2.6596,
      "step": 90
    },
    {
      "epoch": 0.047819232790331054,
      "grad_norm": 2.3586113452911377,
      "learning_rate": 6.147119600233758e-05,
      "loss": 2.5832,
      "step": 91
    },
    {
      "epoch": 0.04834471886495008,
      "grad_norm": 2.3192555904388428,
      "learning_rate": 6.066498153718735e-05,
      "loss": 2.5755,
      "step": 92
    },
    {
      "epoch": 0.048870204939569103,
      "grad_norm": 4.838554859161377,
      "learning_rate": 5.985585137257401e-05,
      "loss": 2.5979,
      "step": 93
    },
    {
      "epoch": 0.04939569101418812,
      "grad_norm": 2.4375178813934326,
      "learning_rate": 5.90440267166055e-05,
      "loss": 2.8294,
      "step": 94
    },
    {
      "epoch": 0.049921177088807146,
      "grad_norm": 3.2174525260925293,
      "learning_rate": 5.8229729514036705e-05,
      "loss": 2.3626,
      "step": 95
    },
    {
      "epoch": 0.05044666316342617,
      "grad_norm": 2.692962884902954,
      "learning_rate": 5.74131823855921e-05,
      "loss": 2.871,
      "step": 96
    },
    {
      "epoch": 0.05097214923804519,
      "grad_norm": 3.2834389209747314,
      "learning_rate": 5.6594608567103456e-05,
      "loss": 2.8696,
      "step": 97
    },
    {
      "epoch": 0.051497635312664214,
      "grad_norm": 2.701561212539673,
      "learning_rate": 5.577423184847932e-05,
      "loss": 2.5867,
      "step": 98
    },
    {
      "epoch": 0.05202312138728324,
      "grad_norm": 2.957364797592163,
      "learning_rate": 5.495227651252315e-05,
      "loss": 2.4674,
      "step": 99
    },
    {
      "epoch": 0.05254860746190226,
      "grad_norm": 5.590366363525391,
      "learning_rate": 5.4128967273616625e-05,
      "loss": 3.1014,
      "step": 100
    },
    {
      "epoch": 0.05254860746190226,
      "eval_loss": 2.7430338859558105,
      "eval_runtime": 63.8051,
      "eval_samples_per_second": 50.231,
      "eval_steps_per_second": 12.57,
      "step": 100
    },
    {
      "epoch": 0.05307409353652128,
      "grad_norm": 3.5561394691467285,
      "learning_rate": 5.330452921628497e-05,
      "loss": 2.7885,
      "step": 101
    },
    {
      "epoch": 0.053599579611140306,
      "grad_norm": 4.151655197143555,
      "learning_rate": 5.247918773366112e-05,
      "loss": 2.8183,
      "step": 102
    },
    {
      "epoch": 0.054125065685759324,
      "grad_norm": 3.5874578952789307,
      "learning_rate": 5.165316846586541e-05,
      "loss": 2.7315,
      "step": 103
    },
    {
      "epoch": 0.05465055176037835,
      "grad_norm": 3.2538201808929443,
      "learning_rate": 5.0826697238317935e-05,
      "loss": 2.7938,
      "step": 104
    },
    {
      "epoch": 0.055176037834997374,
      "grad_norm": 2.828874111175537,
      "learning_rate": 5e-05,
      "loss": 2.4741,
      "step": 105
    },
    {
      "epoch": 0.05570152390961639,
      "grad_norm": 2.0678741931915283,
      "learning_rate": 4.917330276168208e-05,
      "loss": 2.5924,
      "step": 106
    },
    {
      "epoch": 0.05622700998423542,
      "grad_norm": 1.8583709001541138,
      "learning_rate": 4.834683153413459e-05,
      "loss": 2.4872,
      "step": 107
    },
    {
      "epoch": 0.05675249605885444,
      "grad_norm": 1.6289417743682861,
      "learning_rate": 4.7520812266338885e-05,
      "loss": 2.5012,
      "step": 108
    },
    {
      "epoch": 0.05727798213347346,
      "grad_norm": 1.6546365022659302,
      "learning_rate": 4.669547078371504e-05,
      "loss": 2.5573,
      "step": 109
    },
    {
      "epoch": 0.057803468208092484,
      "grad_norm": 1.6993639469146729,
      "learning_rate": 4.5871032726383386e-05,
      "loss": 2.7053,
      "step": 110
    },
    {
      "epoch": 0.05832895428271151,
      "grad_norm": 1.5903888940811157,
      "learning_rate": 4.504772348747687e-05,
      "loss": 2.668,
      "step": 111
    },
    {
      "epoch": 0.058854440357330534,
      "grad_norm": 1.5807969570159912,
      "learning_rate": 4.4225768151520694e-05,
      "loss": 2.3191,
      "step": 112
    },
    {
      "epoch": 0.05937992643194955,
      "grad_norm": 1.7308474779129028,
      "learning_rate": 4.3405391432896555e-05,
      "loss": 2.6387,
      "step": 113
    },
    {
      "epoch": 0.05990541250656858,
      "grad_norm": 1.7413314580917358,
      "learning_rate": 4.2586817614407895e-05,
      "loss": 2.4735,
      "step": 114
    },
    {
      "epoch": 0.0604308985811876,
      "grad_norm": 11.982009887695312,
      "learning_rate": 4.17702704859633e-05,
      "loss": 2.6347,
      "step": 115
    },
    {
      "epoch": 0.06095638465580662,
      "grad_norm": 1.9617363214492798,
      "learning_rate": 4.095597328339452e-05,
      "loss": 2.615,
      "step": 116
    },
    {
      "epoch": 0.061481870730425645,
      "grad_norm": 1.60637366771698,
      "learning_rate": 4.0144148627425993e-05,
      "loss": 2.5558,
      "step": 117
    },
    {
      "epoch": 0.06200735680504467,
      "grad_norm": 1.5637458562850952,
      "learning_rate": 3.933501846281267e-05,
      "loss": 2.4934,
      "step": 118
    },
    {
      "epoch": 0.0625328428796637,
      "grad_norm": 1.6383709907531738,
      "learning_rate": 3.852880399766243e-05,
      "loss": 2.3867,
      "step": 119
    },
    {
      "epoch": 0.0630583289542827,
      "grad_norm": 1.4957643747329712,
      "learning_rate": 3.772572564296005e-05,
      "loss": 2.3857,
      "step": 120
    },
    {
      "epoch": 0.06358381502890173,
      "grad_norm": 1.7877415418624878,
      "learning_rate": 3.6926002952309016e-05,
      "loss": 2.2849,
      "step": 121
    },
    {
      "epoch": 0.06410930110352075,
      "grad_norm": 1.5769100189208984,
      "learning_rate": 3.612985456190778e-05,
      "loss": 2.4558,
      "step": 122
    },
    {
      "epoch": 0.06463478717813978,
      "grad_norm": 1.6936702728271484,
      "learning_rate": 3.533749813077677e-05,
      "loss": 2.5472,
      "step": 123
    },
    {
      "epoch": 0.0651602732527588,
      "grad_norm": 1.5664417743682861,
      "learning_rate": 3.4549150281252636e-05,
      "loss": 2.5743,
      "step": 124
    },
    {
      "epoch": 0.06568575932737783,
      "grad_norm": 1.7317452430725098,
      "learning_rate": 3.3765026539765834e-05,
      "loss": 2.3727,
      "step": 125
    },
    {
      "epoch": 0.06621124540199684,
      "grad_norm": 1.6247220039367676,
      "learning_rate": 3.298534127791785e-05,
      "loss": 2.4799,
      "step": 126
    },
    {
      "epoch": 0.06673673147661587,
      "grad_norm": 1.706653118133545,
      "learning_rate": 3.221030765387417e-05,
      "loss": 2.3447,
      "step": 127
    },
    {
      "epoch": 0.06726221755123489,
      "grad_norm": 1.8045384883880615,
      "learning_rate": 3.144013755408895e-05,
      "loss": 2.569,
      "step": 128
    },
    {
      "epoch": 0.06778770362585392,
      "grad_norm": 1.8199059963226318,
      "learning_rate": 3.0675041535377405e-05,
      "loss": 2.3339,
      "step": 129
    },
    {
      "epoch": 0.06831318970047294,
      "grad_norm": 2.006324529647827,
      "learning_rate": 2.991522876735154e-05,
      "loss": 2.4131,
      "step": 130
    },
    {
      "epoch": 0.06883867577509196,
      "grad_norm": 2.1189417839050293,
      "learning_rate": 2.916090697523549e-05,
      "loss": 2.3111,
      "step": 131
    },
    {
      "epoch": 0.06936416184971098,
      "grad_norm": 1.8146612644195557,
      "learning_rate": 2.8412282383075363e-05,
      "loss": 2.4897,
      "step": 132
    },
    {
      "epoch": 0.06988964792433,
      "grad_norm": 2.3572909832000732,
      "learning_rate": 2.766955965735968e-05,
      "loss": 2.7102,
      "step": 133
    },
    {
      "epoch": 0.07041513399894903,
      "grad_norm": 1.9966086149215698,
      "learning_rate": 2.693294185106562e-05,
      "loss": 2.639,
      "step": 134
    },
    {
      "epoch": 0.07094062007356805,
      "grad_norm": 1.880858302116394,
      "learning_rate": 2.6202630348146324e-05,
      "loss": 2.4948,
      "step": 135
    },
    {
      "epoch": 0.07146610614818708,
      "grad_norm": 1.9570298194885254,
      "learning_rate": 2.547882480847461e-05,
      "loss": 2.5222,
      "step": 136
    },
    {
      "epoch": 0.0719915922228061,
      "grad_norm": 2.5068185329437256,
      "learning_rate": 2.476172311325783e-05,
      "loss": 2.655,
      "step": 137
    },
    {
      "epoch": 0.07251707829742512,
      "grad_norm": 2.5217397212982178,
      "learning_rate": 2.405152131093926e-05,
      "loss": 2.6219,
      "step": 138
    },
    {
      "epoch": 0.07304256437204414,
      "grad_norm": 2.28779673576355,
      "learning_rate": 2.3348413563600325e-05,
      "loss": 2.2438,
      "step": 139
    },
    {
      "epoch": 0.07356805044666316,
      "grad_norm": 2.3459389209747314,
      "learning_rate": 2.2652592093878666e-05,
      "loss": 2.3633,
      "step": 140
    },
    {
      "epoch": 0.07409353652128219,
      "grad_norm": 2.842440366744995,
      "learning_rate": 2.196424713241637e-05,
      "loss": 2.5962,
      "step": 141
    },
    {
      "epoch": 0.07461902259590121,
      "grad_norm": 4.144902229309082,
      "learning_rate": 2.128356686585282e-05,
      "loss": 2.6208,
      "step": 142
    },
    {
      "epoch": 0.07514450867052024,
      "grad_norm": 2.1958553791046143,
      "learning_rate": 2.061073738537635e-05,
      "loss": 2.3283,
      "step": 143
    },
    {
      "epoch": 0.07566999474513926,
      "grad_norm": 3.245858907699585,
      "learning_rate": 1.9945942635848748e-05,
      "loss": 2.7907,
      "step": 144
    },
    {
      "epoch": 0.07619548081975827,
      "grad_norm": 2.396341323852539,
      "learning_rate": 1.928936436551661e-05,
      "loss": 2.5682,
      "step": 145
    },
    {
      "epoch": 0.0767209668943773,
      "grad_norm": 2.905231475830078,
      "learning_rate": 1.8641182076323148e-05,
      "loss": 2.5272,
      "step": 146
    },
    {
      "epoch": 0.07724645296899632,
      "grad_norm": 2.507230758666992,
      "learning_rate": 1.800157297483417e-05,
      "loss": 2.5704,
      "step": 147
    },
    {
      "epoch": 0.07777193904361535,
      "grad_norm": 3.8231539726257324,
      "learning_rate": 1.7370711923791567e-05,
      "loss": 2.7476,
      "step": 148
    },
    {
      "epoch": 0.07829742511823437,
      "grad_norm": 3.1925318241119385,
      "learning_rate": 1.6748771394307585e-05,
      "loss": 2.8047,
      "step": 149
    },
    {
      "epoch": 0.0788229111928534,
      "grad_norm": 13.15278434753418,
      "learning_rate": 1.6135921418712956e-05,
      "loss": 3.2165,
      "step": 150
    },
    {
      "epoch": 0.0788229111928534,
      "eval_loss": 2.5751631259918213,
      "eval_runtime": 63.8172,
      "eval_samples_per_second": 50.222,
      "eval_steps_per_second": 12.567,
      "step": 150
    },
    {
      "epoch": 0.0793483972674724,
      "grad_norm": 1.9700614213943481,
      "learning_rate": 1.553232954407171e-05,
      "loss": 2.6012,
      "step": 151
    },
    {
      "epoch": 0.07987388334209143,
      "grad_norm": 2.375993251800537,
      "learning_rate": 1.4938160786375572e-05,
      "loss": 2.6067,
      "step": 152
    },
    {
      "epoch": 0.08039936941671046,
      "grad_norm": 2.343794345855713,
      "learning_rate": 1.435357758543015e-05,
      "loss": 2.3841,
      "step": 153
    },
    {
      "epoch": 0.08092485549132948,
      "grad_norm": 2.6827147006988525,
      "learning_rate": 1.3778739760445552e-05,
      "loss": 2.3196,
      "step": 154
    },
    {
      "epoch": 0.0814503415659485,
      "grad_norm": 2.185447931289673,
      "learning_rate": 1.3213804466343421e-05,
      "loss": 2.4531,
      "step": 155
    },
    {
      "epoch": 0.08197582764056753,
      "grad_norm": 2.358931064605713,
      "learning_rate": 1.2658926150792322e-05,
      "loss": 2.5774,
      "step": 156
    },
    {
      "epoch": 0.08250131371518654,
      "grad_norm": 1.9362928867340088,
      "learning_rate": 1.2114256511983274e-05,
      "loss": 2.5972,
      "step": 157
    },
    {
      "epoch": 0.08302679978980557,
      "grad_norm": 2.0720343589782715,
      "learning_rate": 1.157994445715706e-05,
      "loss": 2.449,
      "step": 158
    },
    {
      "epoch": 0.08355228586442459,
      "grad_norm": 1.863685965538025,
      "learning_rate": 1.1056136061894384e-05,
      "loss": 2.5223,
      "step": 159
    },
    {
      "epoch": 0.08407777193904362,
      "grad_norm": 2.223606586456299,
      "learning_rate": 1.0542974530180327e-05,
      "loss": 2.5683,
      "step": 160
    },
    {
      "epoch": 0.08460325801366264,
      "grad_norm": 1.8201804161071777,
      "learning_rate": 1.0040600155253765e-05,
      "loss": 2.478,
      "step": 161
    },
    {
      "epoch": 0.08512874408828167,
      "grad_norm": 2.5992581844329834,
      "learning_rate": 9.549150281252633e-06,
      "loss": 2.6768,
      "step": 162
    },
    {
      "epoch": 0.08565423016290068,
      "grad_norm": 1.7471072673797607,
      "learning_rate": 9.068759265665384e-06,
      "loss": 2.5242,
      "step": 163
    },
    {
      "epoch": 0.0861797162375197,
      "grad_norm": 1.6078826189041138,
      "learning_rate": 8.599558442598998e-06,
      "loss": 2.4488,
      "step": 164
    },
    {
      "epoch": 0.08670520231213873,
      "grad_norm": 1.6924360990524292,
      "learning_rate": 8.141676086873572e-06,
      "loss": 2.5203,
      "step": 165
    },
    {
      "epoch": 0.08723068838675775,
      "grad_norm": 1.5796608924865723,
      "learning_rate": 7.695237378953223e-06,
      "loss": 2.4129,
      "step": 166
    },
    {
      "epoch": 0.08775617446137678,
      "grad_norm": 1.7781840562820435,
      "learning_rate": 7.260364370723044e-06,
      "loss": 2.4229,
      "step": 167
    },
    {
      "epoch": 0.0882816605359958,
      "grad_norm": 1.6857123374938965,
      "learning_rate": 6.837175952121306e-06,
      "loss": 2.6196,
      "step": 168
    },
    {
      "epoch": 0.08880714661061481,
      "grad_norm": 1.433090090751648,
      "learning_rate": 6.425787818636131e-06,
      "loss": 2.339,
      "step": 169
    },
    {
      "epoch": 0.08933263268523384,
      "grad_norm": 1.7250750064849854,
      "learning_rate": 6.026312439675552e-06,
      "loss": 2.5213,
      "step": 170
    },
    {
      "epoch": 0.08985811875985286,
      "grad_norm": 1.8775157928466797,
      "learning_rate": 5.6388590278194096e-06,
      "loss": 2.3304,
      "step": 171
    },
    {
      "epoch": 0.09038360483447189,
      "grad_norm": 1.5660293102264404,
      "learning_rate": 5.263533508961827e-06,
      "loss": 2.3937,
      "step": 172
    },
    {
      "epoch": 0.09090909090909091,
      "grad_norm": 1.6316609382629395,
      "learning_rate": 4.900438493352055e-06,
      "loss": 2.2765,
      "step": 173
    },
    {
      "epoch": 0.09143457698370994,
      "grad_norm": 1.8517199754714966,
      "learning_rate": 4.549673247541875e-06,
      "loss": 2.3971,
      "step": 174
    },
    {
      "epoch": 0.09196006305832895,
      "grad_norm": 1.6643317937850952,
      "learning_rate": 4.2113336672471245e-06,
      "loss": 2.6349,
      "step": 175
    },
    {
      "epoch": 0.09248554913294797,
      "grad_norm": 1.6245158910751343,
      "learning_rate": 3.885512251130763e-06,
      "loss": 2.6504,
      "step": 176
    },
    {
      "epoch": 0.093011035207567,
      "grad_norm": 1.7814983129501343,
      "learning_rate": 3.5722980755146517e-06,
      "loss": 2.5802,
      "step": 177
    },
    {
      "epoch": 0.09353652128218602,
      "grad_norm": 3.3906099796295166,
      "learning_rate": 3.271776770026963e-06,
      "loss": 2.5609,
      "step": 178
    },
    {
      "epoch": 0.09406200735680505,
      "grad_norm": 2.0150277614593506,
      "learning_rate": 2.9840304941919415e-06,
      "loss": 2.6728,
      "step": 179
    },
    {
      "epoch": 0.09458749343142407,
      "grad_norm": 1.8740047216415405,
      "learning_rate": 2.7091379149682685e-06,
      "loss": 2.5437,
      "step": 180
    },
    {
      "epoch": 0.0951129795060431,
      "grad_norm": 2.12048602104187,
      "learning_rate": 2.4471741852423237e-06,
      "loss": 2.4097,
      "step": 181
    },
    {
      "epoch": 0.09563846558066211,
      "grad_norm": 2.16339111328125,
      "learning_rate": 2.1982109232821178e-06,
      "loss": 2.5361,
      "step": 182
    },
    {
      "epoch": 0.09616395165528113,
      "grad_norm": 4.183311939239502,
      "learning_rate": 1.962316193157593e-06,
      "loss": 2.316,
      "step": 183
    },
    {
      "epoch": 0.09668943772990016,
      "grad_norm": 2.0543057918548584,
      "learning_rate": 1.7395544861325718e-06,
      "loss": 2.4998,
      "step": 184
    },
    {
      "epoch": 0.09721492380451918,
      "grad_norm": 1.9849950075149536,
      "learning_rate": 1.5299867030334814e-06,
      "loss": 2.56,
      "step": 185
    },
    {
      "epoch": 0.09774040987913821,
      "grad_norm": 1.90382719039917,
      "learning_rate": 1.333670137599713e-06,
      "loss": 2.4199,
      "step": 186
    },
    {
      "epoch": 0.09826589595375723,
      "grad_norm": 3.6999776363372803,
      "learning_rate": 1.1506584608200367e-06,
      "loss": 2.6445,
      "step": 187
    },
    {
      "epoch": 0.09879138202837624,
      "grad_norm": 1.9546655416488647,
      "learning_rate": 9.810017062595322e-07,
      "loss": 2.4109,
      "step": 188
    },
    {
      "epoch": 0.09931686810299527,
      "grad_norm": 9.042537689208984,
      "learning_rate": 8.247462563808817e-07,
      "loss": 2.7544,
      "step": 189
    },
    {
      "epoch": 0.09984235417761429,
      "grad_norm": 5.109813690185547,
      "learning_rate": 6.819348298638839e-07,
      "loss": 2.5811,
      "step": 190
    },
    {
      "epoch": 0.10036784025223332,
      "grad_norm": 3.159303665161133,
      "learning_rate": 5.526064699265753e-07,
      "loss": 2.5114,
      "step": 191
    },
    {
      "epoch": 0.10089332632685234,
      "grad_norm": 2.221721649169922,
      "learning_rate": 4.367965336512403e-07,
      "loss": 2.5756,
      "step": 192
    },
    {
      "epoch": 0.10141881240147137,
      "grad_norm": 2.153102397918701,
      "learning_rate": 3.3453668231809286e-07,
      "loss": 2.651,
      "step": 193
    },
    {
      "epoch": 0.10194429847609038,
      "grad_norm": 2.269869089126587,
      "learning_rate": 2.458548727494292e-07,
      "loss": 2.6683,
      "step": 194
    },
    {
      "epoch": 0.1024697845507094,
      "grad_norm": 2.653329372406006,
      "learning_rate": 1.7077534966650766e-07,
      "loss": 2.5345,
      "step": 195
    },
    {
      "epoch": 0.10299527062532843,
      "grad_norm": 2.67691969871521,
      "learning_rate": 1.0931863906127327e-07,
      "loss": 2.4389,
      "step": 196
    },
    {
      "epoch": 0.10352075669994745,
      "grad_norm": 3.1515066623687744,
      "learning_rate": 6.150154258476315e-08,
      "loss": 3.1161,
      "step": 197
    },
    {
      "epoch": 0.10404624277456648,
      "grad_norm": 3.2619149684906006,
      "learning_rate": 2.7337132953697554e-08,
      "loss": 2.8707,
      "step": 198
    },
    {
      "epoch": 0.1045717288491855,
      "grad_norm": 3.538806915283203,
      "learning_rate": 6.834750376549792e-09,
      "loss": 2.8636,
      "step": 199
    },
    {
      "epoch": 0.10509721492380451,
      "grad_norm": 5.6363701820373535,
      "learning_rate": 0.0,
      "loss": 3.2003,
      "step": 200
    },
    {
      "epoch": 0.10509721492380451,
      "eval_loss": 2.5217931270599365,
      "eval_runtime": 63.9234,
      "eval_samples_per_second": 50.138,
      "eval_steps_per_second": 12.546,
      "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": 4.26629570494464e+16,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}