{
  "best_metric": 1.926206350326538,
  "best_model_checkpoint": "miner_id_24/checkpoint-200",
  "epoch": 0.05737234652897304,
  "eval_steps": 50,
  "global_step": 200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00028686173264486515,
      "grad_norm": 3.0401198863983154,
      "learning_rate": 1e-05,
      "loss": 9.7413,
      "step": 1
    },
    {
      "epoch": 0.00028686173264486515,
      "eval_loss": 2.5668680667877197,
      "eval_runtime": 280.4305,
      "eval_samples_per_second": 20.936,
      "eval_steps_per_second": 5.235,
      "step": 1
    },
    {
      "epoch": 0.0005737234652897303,
      "grad_norm": 3.8561806678771973,
      "learning_rate": 2e-05,
      "loss": 9.9331,
      "step": 2
    },
    {
      "epoch": 0.0008605851979345956,
      "grad_norm": 3.4238851070404053,
      "learning_rate": 3e-05,
      "loss": 9.9092,
      "step": 3
    },
    {
      "epoch": 0.0011474469305794606,
      "grad_norm": 3.9144554138183594,
      "learning_rate": 4e-05,
      "loss": 10.9025,
      "step": 4
    },
    {
      "epoch": 0.001434308663224326,
      "grad_norm": 2.685176372528076,
      "learning_rate": 5e-05,
      "loss": 9.4328,
      "step": 5
    },
    {
      "epoch": 0.0017211703958691911,
      "grad_norm": 1.9112707376480103,
      "learning_rate": 6e-05,
      "loss": 9.8314,
      "step": 6
    },
    {
      "epoch": 0.002008032128514056,
      "grad_norm": 1.4061189889907837,
      "learning_rate": 7e-05,
      "loss": 9.6855,
      "step": 7
    },
    {
      "epoch": 0.002294893861158921,
      "grad_norm": 1.5059212446212769,
      "learning_rate": 8e-05,
      "loss": 9.5399,
      "step": 8
    },
    {
      "epoch": 0.0025817555938037868,
      "grad_norm": 1.5410596132278442,
      "learning_rate": 9e-05,
      "loss": 8.91,
      "step": 9
    },
    {
      "epoch": 0.002868617326448652,
      "grad_norm": 1.7056461572647095,
      "learning_rate": 0.0001,
      "loss": 9.4448,
      "step": 10
    },
    {
      "epoch": 0.003155479059093517,
      "grad_norm": 1.5578986406326294,
      "learning_rate": 9.999316524962345e-05,
      "loss": 8.8125,
      "step": 11
    },
    {
      "epoch": 0.0034423407917383822,
      "grad_norm": 1.3023666143417358,
      "learning_rate": 9.997266286704631e-05,
      "loss": 8.6899,
      "step": 12
    },
    {
      "epoch": 0.0037292025243832474,
      "grad_norm": 1.1935285329818726,
      "learning_rate": 9.993849845741524e-05,
      "loss": 8.8257,
      "step": 13
    },
    {
      "epoch": 0.004016064257028112,
      "grad_norm": 1.308254599571228,
      "learning_rate": 9.989068136093873e-05,
      "loss": 8.8794,
      "step": 14
    },
    {
      "epoch": 0.004302925989672977,
      "grad_norm": 1.273793339729309,
      "learning_rate": 9.98292246503335e-05,
      "loss": 8.6004,
      "step": 15
    },
    {
      "epoch": 0.004589787722317842,
      "grad_norm": 1.2527799606323242,
      "learning_rate": 9.975414512725057e-05,
      "loss": 8.733,
      "step": 16
    },
    {
      "epoch": 0.004876649454962708,
      "grad_norm": 1.179306983947754,
      "learning_rate": 9.966546331768191e-05,
      "loss": 8.7311,
      "step": 17
    },
    {
      "epoch": 0.0051635111876075735,
      "grad_norm": 1.232127070426941,
      "learning_rate": 9.956320346634876e-05,
      "loss": 8.9044,
      "step": 18
    },
    {
      "epoch": 0.005450372920252439,
      "grad_norm": 1.2128486633300781,
      "learning_rate": 9.944739353007344e-05,
      "loss": 8.7441,
      "step": 19
    },
    {
      "epoch": 0.005737234652897304,
      "grad_norm": 1.0975828170776367,
      "learning_rate": 9.931806517013612e-05,
      "loss": 8.8321,
      "step": 20
    },
    {
      "epoch": 0.006024096385542169,
      "grad_norm": 1.1655679941177368,
      "learning_rate": 9.917525374361912e-05,
      "loss": 8.729,
      "step": 21
    },
    {
      "epoch": 0.006310958118187034,
      "grad_norm": 1.1813334226608276,
      "learning_rate": 9.901899829374047e-05,
      "loss": 8.2111,
      "step": 22
    },
    {
      "epoch": 0.006597819850831899,
      "grad_norm": 1.291486382484436,
      "learning_rate": 9.884934153917997e-05,
      "loss": 8.2365,
      "step": 23
    },
    {
      "epoch": 0.0068846815834767644,
      "grad_norm": 1.1758555173873901,
      "learning_rate": 9.86663298624003e-05,
      "loss": 8.1319,
      "step": 24
    },
    {
      "epoch": 0.00717154331612163,
      "grad_norm": 1.3970890045166016,
      "learning_rate": 9.847001329696653e-05,
      "loss": 8.4453,
      "step": 25
    },
    {
      "epoch": 0.007458405048766495,
      "grad_norm": 1.2271980047225952,
      "learning_rate": 9.826044551386744e-05,
      "loss": 7.9361,
      "step": 26
    },
    {
      "epoch": 0.00774526678141136,
      "grad_norm": 1.2449077367782593,
      "learning_rate": 9.803768380684242e-05,
      "loss": 8.3274,
      "step": 27
    },
    {
      "epoch": 0.008032128514056224,
      "grad_norm": 1.352083683013916,
      "learning_rate": 9.780178907671789e-05,
      "loss": 8.701,
      "step": 28
    },
    {
      "epoch": 0.00831899024670109,
      "grad_norm": 1.3462893962860107,
      "learning_rate": 9.755282581475769e-05,
      "loss": 7.7078,
      "step": 29
    },
    {
      "epoch": 0.008605851979345954,
      "grad_norm": 1.4451414346694946,
      "learning_rate": 9.729086208503174e-05,
      "loss": 8.1698,
      "step": 30
    },
    {
      "epoch": 0.00889271371199082,
      "grad_norm": 1.4077917337417603,
      "learning_rate": 9.701596950580806e-05,
      "loss": 8.693,
      "step": 31
    },
    {
      "epoch": 0.009179575444635685,
      "grad_norm": 1.3237286806106567,
      "learning_rate": 9.672822322997305e-05,
      "loss": 8.4617,
      "step": 32
    },
    {
      "epoch": 0.00946643717728055,
      "grad_norm": 1.3992210626602173,
      "learning_rate": 9.642770192448536e-05,
      "loss": 8.7108,
      "step": 33
    },
    {
      "epoch": 0.009753298909925417,
      "grad_norm": 1.4127986431121826,
      "learning_rate": 9.611448774886924e-05,
      "loss": 8.7477,
      "step": 34
    },
    {
      "epoch": 0.010040160642570281,
      "grad_norm": 1.378975510597229,
      "learning_rate": 9.578866633275288e-05,
      "loss": 8.6461,
      "step": 35
    },
    {
      "epoch": 0.010327022375215147,
      "grad_norm": 1.3909608125686646,
      "learning_rate": 9.545032675245813e-05,
      "loss": 8.7952,
      "step": 36
    },
    {
      "epoch": 0.010613884107860011,
      "grad_norm": 1.3498965501785278,
      "learning_rate": 9.509956150664796e-05,
      "loss": 8.7529,
      "step": 37
    },
    {
      "epoch": 0.010900745840504877,
      "grad_norm": 1.4007525444030762,
      "learning_rate": 9.473646649103818e-05,
      "loss": 8.2888,
      "step": 38
    },
    {
      "epoch": 0.011187607573149742,
      "grad_norm": 1.3914694786071777,
      "learning_rate": 9.43611409721806e-05,
      "loss": 8.0191,
      "step": 39
    },
    {
      "epoch": 0.011474469305794608,
      "grad_norm": 1.5015257596969604,
      "learning_rate": 9.397368756032445e-05,
      "loss": 8.2274,
      "step": 40
    },
    {
      "epoch": 0.011761331038439472,
      "grad_norm": 1.4857221841812134,
      "learning_rate": 9.357421218136386e-05,
      "loss": 8.9747,
      "step": 41
    },
    {
      "epoch": 0.012048192771084338,
      "grad_norm": 1.5363478660583496,
      "learning_rate": 9.316282404787871e-05,
      "loss": 8.1086,
      "step": 42
    },
    {
      "epoch": 0.012335054503729202,
      "grad_norm": 1.524233102798462,
      "learning_rate": 9.273963562927695e-05,
      "loss": 8.9862,
      "step": 43
    },
    {
      "epoch": 0.012621916236374068,
      "grad_norm": 1.5505753755569458,
      "learning_rate": 9.230476262104677e-05,
      "loss": 7.7927,
      "step": 44
    },
    {
      "epoch": 0.012908777969018933,
      "grad_norm": 1.6508245468139648,
      "learning_rate": 9.185832391312644e-05,
      "loss": 7.6721,
      "step": 45
    },
    {
      "epoch": 0.013195639701663799,
      "grad_norm": 1.5731823444366455,
      "learning_rate": 9.140044155740101e-05,
      "loss": 7.9439,
      "step": 46
    },
    {
      "epoch": 0.013482501434308663,
      "grad_norm": 1.6659985780715942,
      "learning_rate": 9.093124073433463e-05,
      "loss": 8.8125,
      "step": 47
    },
    {
      "epoch": 0.013769363166953529,
      "grad_norm": 1.9959471225738525,
      "learning_rate": 9.045084971874738e-05,
      "loss": 7.2652,
      "step": 48
    },
    {
      "epoch": 0.014056224899598393,
      "grad_norm": 3.823784351348877,
      "learning_rate": 8.995939984474624e-05,
      "loss": 6.558,
      "step": 49
    },
    {
      "epoch": 0.01434308663224326,
      "grad_norm": 6.641433238983154,
      "learning_rate": 8.945702546981969e-05,
      "loss": 7.1197,
      "step": 50
    },
    {
      "epoch": 0.01434308663224326,
      "eval_loss": 2.0632212162017822,
      "eval_runtime": 282.2747,
      "eval_samples_per_second": 20.799,
      "eval_steps_per_second": 5.201,
      "step": 50
    },
    {
      "epoch": 0.014629948364888123,
      "grad_norm": 1.4122048616409302,
      "learning_rate": 8.894386393810563e-05,
      "loss": 8.9177,
      "step": 51
    },
    {
      "epoch": 0.01491681009753299,
      "grad_norm": 1.3056529760360718,
      "learning_rate": 8.842005554284296e-05,
      "loss": 8.3227,
      "step": 52
    },
    {
      "epoch": 0.015203671830177854,
      "grad_norm": 1.334641695022583,
      "learning_rate": 8.788574348801675e-05,
      "loss": 9.2602,
      "step": 53
    },
    {
      "epoch": 0.01549053356282272,
      "grad_norm": 1.2780232429504395,
      "learning_rate": 8.73410738492077e-05,
      "loss": 8.7697,
      "step": 54
    },
    {
      "epoch": 0.015777395295467584,
      "grad_norm": 1.172772765159607,
      "learning_rate": 8.678619553365659e-05,
      "loss": 8.3807,
      "step": 55
    },
    {
      "epoch": 0.01606425702811245,
      "grad_norm": 1.0277596712112427,
      "learning_rate": 8.622126023955446e-05,
      "loss": 8.4207,
      "step": 56
    },
    {
      "epoch": 0.016351118760757316,
      "grad_norm": 1.2122106552124023,
      "learning_rate": 8.564642241456986e-05,
      "loss": 8.8548,
      "step": 57
    },
    {
      "epoch": 0.01663798049340218,
      "grad_norm": 1.2018460035324097,
      "learning_rate": 8.506183921362443e-05,
      "loss": 8.4117,
      "step": 58
    },
    {
      "epoch": 0.016924842226047045,
      "grad_norm": 1.2418075799942017,
      "learning_rate": 8.44676704559283e-05,
      "loss": 8.663,
      "step": 59
    },
    {
      "epoch": 0.01721170395869191,
      "grad_norm": 1.163619875907898,
      "learning_rate": 8.386407858128706e-05,
      "loss": 8.8763,
      "step": 60
    },
    {
      "epoch": 0.017498565691336777,
      "grad_norm": 1.3117581605911255,
      "learning_rate": 8.32512286056924e-05,
      "loss": 7.7938,
      "step": 61
    },
    {
      "epoch": 0.01778542742398164,
      "grad_norm": 1.201149344444275,
      "learning_rate": 8.262928807620843e-05,
      "loss": 8.1896,
      "step": 62
    },
    {
      "epoch": 0.018072289156626505,
      "grad_norm": 1.252098560333252,
      "learning_rate": 8.199842702516583e-05,
      "loss": 8.0049,
      "step": 63
    },
    {
      "epoch": 0.01835915088927137,
      "grad_norm": 1.183274269104004,
      "learning_rate": 8.135881792367686e-05,
      "loss": 8.2218,
      "step": 64
    },
    {
      "epoch": 0.018646012621916237,
      "grad_norm": 1.4185477495193481,
      "learning_rate": 8.07106356344834e-05,
      "loss": 8.3158,
      "step": 65
    },
    {
      "epoch": 0.0189328743545611,
      "grad_norm": 1.295198678970337,
      "learning_rate": 8.005405736415126e-05,
      "loss": 8.1416,
      "step": 66
    },
    {
      "epoch": 0.019219736087205966,
      "grad_norm": 1.2766518592834473,
      "learning_rate": 7.938926261462366e-05,
      "loss": 7.8335,
      "step": 67
    },
    {
      "epoch": 0.019506597819850834,
      "grad_norm": 1.4730690717697144,
      "learning_rate": 7.871643313414718e-05,
      "loss": 8.3285,
      "step": 68
    },
    {
      "epoch": 0.019793459552495698,
      "grad_norm": 1.3872134685516357,
      "learning_rate": 7.803575286758364e-05,
      "loss": 8.0806,
      "step": 69
    },
    {
      "epoch": 0.020080321285140562,
      "grad_norm": 1.2927069664001465,
      "learning_rate": 7.734740790612136e-05,
      "loss": 7.8835,
      "step": 70
    },
    {
      "epoch": 0.020367183017785426,
      "grad_norm": 1.3059531450271606,
      "learning_rate": 7.66515864363997e-05,
      "loss": 7.5511,
      "step": 71
    },
    {
      "epoch": 0.020654044750430294,
      "grad_norm": 1.565891981124878,
      "learning_rate": 7.594847868906076e-05,
      "loss": 7.6058,
      "step": 72
    },
    {
      "epoch": 0.02094090648307516,
      "grad_norm": 1.339733362197876,
      "learning_rate": 7.52382768867422e-05,
      "loss": 7.6848,
      "step": 73
    },
    {
      "epoch": 0.021227768215720023,
      "grad_norm": 1.474951982498169,
      "learning_rate": 7.452117519152542e-05,
      "loss": 7.7781,
      "step": 74
    },
    {
      "epoch": 0.021514629948364887,
      "grad_norm": 1.404189109802246,
      "learning_rate": 7.379736965185368e-05,
      "loss": 7.8352,
      "step": 75
    },
    {
      "epoch": 0.021801491681009755,
      "grad_norm": 1.691503643989563,
      "learning_rate": 7.30670581489344e-05,
      "loss": 8.5215,
      "step": 76
    },
    {
      "epoch": 0.02208835341365462,
      "grad_norm": 1.4138275384902954,
      "learning_rate": 7.233044034264034e-05,
      "loss": 7.6912,
      "step": 77
    },
    {
      "epoch": 0.022375215146299483,
      "grad_norm": 1.4546366930007935,
      "learning_rate": 7.158771761692464e-05,
      "loss": 7.9238,
      "step": 78
    },
    {
      "epoch": 0.022662076878944348,
      "grad_norm": 1.4224469661712646,
      "learning_rate": 7.083909302476453e-05,
      "loss": 8.1009,
      "step": 79
    },
    {
      "epoch": 0.022948938611589215,
      "grad_norm": 1.434641718864441,
      "learning_rate": 7.008477123264848e-05,
      "loss": 8.0694,
      "step": 80
    },
    {
      "epoch": 0.02323580034423408,
      "grad_norm": 1.4902080297470093,
      "learning_rate": 6.932495846462261e-05,
      "loss": 8.2426,
      "step": 81
    },
    {
      "epoch": 0.023522662076878944,
      "grad_norm": 1.505750298500061,
      "learning_rate": 6.855986244591104e-05,
      "loss": 8.1906,
      "step": 82
    },
    {
      "epoch": 0.023809523809523808,
      "grad_norm": 1.6639647483825684,
      "learning_rate": 6.778969234612584e-05,
      "loss": 7.6399,
      "step": 83
    },
    {
      "epoch": 0.024096385542168676,
      "grad_norm": 1.533452033996582,
      "learning_rate": 6.701465872208216e-05,
      "loss": 8.0799,
      "step": 84
    },
    {
      "epoch": 0.02438324727481354,
      "grad_norm": 1.6043343544006348,
      "learning_rate": 6.623497346023418e-05,
      "loss": 8.176,
      "step": 85
    },
    {
      "epoch": 0.024670109007458405,
      "grad_norm": 1.53847336769104,
      "learning_rate": 6.545084971874738e-05,
      "loss": 7.1614,
      "step": 86
    },
    {
      "epoch": 0.02495697074010327,
      "grad_norm": 1.7997097969055176,
      "learning_rate": 6.466250186922325e-05,
      "loss": 7.9587,
      "step": 87
    },
    {
      "epoch": 0.025243832472748137,
      "grad_norm": 1.5545953512191772,
      "learning_rate": 6.387014543809223e-05,
      "loss": 7.5016,
      "step": 88
    },
    {
      "epoch": 0.025530694205393,
      "grad_norm": 1.661802887916565,
      "learning_rate": 6.307399704769099e-05,
      "loss": 7.5321,
      "step": 89
    },
    {
      "epoch": 0.025817555938037865,
      "grad_norm": 1.6800179481506348,
      "learning_rate": 6.227427435703997e-05,
      "loss": 7.9831,
      "step": 90
    },
    {
      "epoch": 0.02610441767068273,
      "grad_norm": 1.5589094161987305,
      "learning_rate": 6.147119600233758e-05,
      "loss": 8.0974,
      "step": 91
    },
    {
      "epoch": 0.026391279403327597,
      "grad_norm": 1.8109207153320312,
      "learning_rate": 6.066498153718735e-05,
      "loss": 8.4902,
      "step": 92
    },
    {
      "epoch": 0.02667814113597246,
      "grad_norm": 1.6859471797943115,
      "learning_rate": 5.985585137257401e-05,
      "loss": 8.2357,
      "step": 93
    },
    {
      "epoch": 0.026965002868617326,
      "grad_norm": 1.5752520561218262,
      "learning_rate": 5.90440267166055e-05,
      "loss": 7.745,
      "step": 94
    },
    {
      "epoch": 0.02725186460126219,
      "grad_norm": 1.5889084339141846,
      "learning_rate": 5.8229729514036705e-05,
      "loss": 7.87,
      "step": 95
    },
    {
      "epoch": 0.027538726333907058,
      "grad_norm": 1.6827741861343384,
      "learning_rate": 5.74131823855921e-05,
      "loss": 7.2508,
      "step": 96
    },
    {
      "epoch": 0.027825588066551922,
      "grad_norm": 1.7450107336044312,
      "learning_rate": 5.6594608567103456e-05,
      "loss": 8.1502,
      "step": 97
    },
    {
      "epoch": 0.028112449799196786,
      "grad_norm": 1.688856601715088,
      "learning_rate": 5.577423184847932e-05,
      "loss": 7.798,
      "step": 98
    },
    {
      "epoch": 0.028399311531841654,
      "grad_norm": 3.679314613342285,
      "learning_rate": 5.495227651252315e-05,
      "loss": 5.2979,
      "step": 99
    },
    {
      "epoch": 0.02868617326448652,
      "grad_norm": 5.121074676513672,
      "learning_rate": 5.4128967273616625e-05,
      "loss": 5.7695,
      "step": 100
    },
    {
      "epoch": 0.02868617326448652,
      "eval_loss": 1.9794018268585205,
      "eval_runtime": 282.5194,
      "eval_samples_per_second": 20.781,
      "eval_steps_per_second": 5.196,
      "step": 100
    },
    {
      "epoch": 0.028973034997131383,
      "grad_norm": 1.6453404426574707,
      "learning_rate": 5.330452921628497e-05,
      "loss": 8.4342,
      "step": 101
    },
    {
      "epoch": 0.029259896729776247,
      "grad_norm": 1.5428962707519531,
      "learning_rate": 5.247918773366112e-05,
      "loss": 8.7705,
      "step": 102
    },
    {
      "epoch": 0.029546758462421115,
      "grad_norm": 1.420920968055725,
      "learning_rate": 5.165316846586541e-05,
      "loss": 8.2905,
      "step": 103
    },
    {
      "epoch": 0.02983362019506598,
      "grad_norm": 1.3663030862808228,
      "learning_rate": 5.0826697238317935e-05,
      "loss": 7.8795,
      "step": 104
    },
    {
      "epoch": 0.030120481927710843,
      "grad_norm": 1.268221378326416,
      "learning_rate": 5e-05,
      "loss": 8.6724,
      "step": 105
    },
    {
      "epoch": 0.030407343660355707,
      "grad_norm": 1.284090518951416,
      "learning_rate": 4.917330276168208e-05,
      "loss": 8.2105,
      "step": 106
    },
    {
      "epoch": 0.030694205393000575,
      "grad_norm": 1.2277450561523438,
      "learning_rate": 4.834683153413459e-05,
      "loss": 8.0819,
      "step": 107
    },
    {
      "epoch": 0.03098106712564544,
      "grad_norm": 1.3332046270370483,
      "learning_rate": 4.7520812266338885e-05,
      "loss": 8.3792,
      "step": 108
    },
    {
      "epoch": 0.03126792885829031,
      "grad_norm": 1.3728914260864258,
      "learning_rate": 4.669547078371504e-05,
      "loss": 7.7785,
      "step": 109
    },
    {
      "epoch": 0.03155479059093517,
      "grad_norm": 1.2537291049957275,
      "learning_rate": 4.5871032726383386e-05,
      "loss": 7.9977,
      "step": 110
    },
    {
      "epoch": 0.031841652323580036,
      "grad_norm": 1.3703155517578125,
      "learning_rate": 4.504772348747687e-05,
      "loss": 8.0916,
      "step": 111
    },
    {
      "epoch": 0.0321285140562249,
      "grad_norm": 1.5226688385009766,
      "learning_rate": 4.4225768151520694e-05,
      "loss": 8.1762,
      "step": 112
    },
    {
      "epoch": 0.032415375788869764,
      "grad_norm": 1.7661747932434082,
      "learning_rate": 4.3405391432896555e-05,
      "loss": 7.6323,
      "step": 113
    },
    {
      "epoch": 0.03270223752151463,
      "grad_norm": 1.8368374109268188,
      "learning_rate": 4.2586817614407895e-05,
      "loss": 7.4977,
      "step": 114
    },
    {
      "epoch": 0.03298909925415949,
      "grad_norm": 1.5700205564498901,
      "learning_rate": 4.17702704859633e-05,
      "loss": 7.6413,
      "step": 115
    },
    {
      "epoch": 0.03327596098680436,
      "grad_norm": 1.5076812505722046,
      "learning_rate": 4.095597328339452e-05,
      "loss": 7.8666,
      "step": 116
    },
    {
      "epoch": 0.03356282271944923,
      "grad_norm": 1.3722950220108032,
      "learning_rate": 4.0144148627425993e-05,
      "loss": 7.7111,
      "step": 117
    },
    {
      "epoch": 0.03384968445209409,
      "grad_norm": 1.434427261352539,
      "learning_rate": 3.933501846281267e-05,
      "loss": 7.9437,
      "step": 118
    },
    {
      "epoch": 0.03413654618473896,
      "grad_norm": 1.5034595727920532,
      "learning_rate": 3.852880399766243e-05,
      "loss": 7.8544,
      "step": 119
    },
    {
      "epoch": 0.03442340791738382,
      "grad_norm": 1.5724352598190308,
      "learning_rate": 3.772572564296005e-05,
      "loss": 7.7647,
      "step": 120
    },
    {
      "epoch": 0.034710269650028686,
      "grad_norm": 1.4990090131759644,
      "learning_rate": 3.6926002952309016e-05,
      "loss": 8.2473,
      "step": 121
    },
    {
      "epoch": 0.03499713138267355,
      "grad_norm": 1.6667665243148804,
      "learning_rate": 3.612985456190778e-05,
      "loss": 8.0882,
      "step": 122
    },
    {
      "epoch": 0.035283993115318414,
      "grad_norm": 1.458492636680603,
      "learning_rate": 3.533749813077677e-05,
      "loss": 7.0904,
      "step": 123
    },
    {
      "epoch": 0.03557085484796328,
      "grad_norm": 1.465876579284668,
      "learning_rate": 3.4549150281252636e-05,
      "loss": 7.9857,
      "step": 124
    },
    {
      "epoch": 0.03585771658060815,
      "grad_norm": 1.5374915599822998,
      "learning_rate": 3.3765026539765834e-05,
      "loss": 7.3472,
      "step": 125
    },
    {
      "epoch": 0.03614457831325301,
      "grad_norm": 1.7170230150222778,
      "learning_rate": 3.298534127791785e-05,
      "loss": 7.2393,
      "step": 126
    },
    {
      "epoch": 0.03643144004589788,
      "grad_norm": 1.6013517379760742,
      "learning_rate": 3.221030765387417e-05,
      "loss": 8.2777,
      "step": 127
    },
    {
      "epoch": 0.03671830177854274,
      "grad_norm": 1.6421613693237305,
      "learning_rate": 3.144013755408895e-05,
      "loss": 8.2855,
      "step": 128
    },
    {
      "epoch": 0.03700516351118761,
      "grad_norm": 1.7048721313476562,
      "learning_rate": 3.0675041535377405e-05,
      "loss": 7.9614,
      "step": 129
    },
    {
      "epoch": 0.037292025243832475,
      "grad_norm": 1.7380433082580566,
      "learning_rate": 2.991522876735154e-05,
      "loss": 7.7988,
      "step": 130
    },
    {
      "epoch": 0.037578886976477335,
      "grad_norm": 1.5603336095809937,
      "learning_rate": 2.916090697523549e-05,
      "loss": 7.1458,
      "step": 131
    },
    {
      "epoch": 0.0378657487091222,
      "grad_norm": 1.633914589881897,
      "learning_rate": 2.8412282383075363e-05,
      "loss": 7.6048,
      "step": 132
    },
    {
      "epoch": 0.03815261044176707,
      "grad_norm": 1.5596357583999634,
      "learning_rate": 2.766955965735968e-05,
      "loss": 8.247,
      "step": 133
    },
    {
      "epoch": 0.03843947217441193,
      "grad_norm": 1.6335374116897583,
      "learning_rate": 2.693294185106562e-05,
      "loss": 7.8567,
      "step": 134
    },
    {
      "epoch": 0.0387263339070568,
      "grad_norm": 1.94757080078125,
      "learning_rate": 2.6202630348146324e-05,
      "loss": 7.4744,
      "step": 135
    },
    {
      "epoch": 0.03901319563970167,
      "grad_norm": 1.5955264568328857,
      "learning_rate": 2.547882480847461e-05,
      "loss": 7.905,
      "step": 136
    },
    {
      "epoch": 0.03930005737234653,
      "grad_norm": 1.644698143005371,
      "learning_rate": 2.476172311325783e-05,
      "loss": 7.8786,
      "step": 137
    },
    {
      "epoch": 0.039586919104991396,
      "grad_norm": 1.6589207649230957,
      "learning_rate": 2.405152131093926e-05,
      "loss": 7.8557,
      "step": 138
    },
    {
      "epoch": 0.039873780837636257,
      "grad_norm": 1.7729710340499878,
      "learning_rate": 2.3348413563600325e-05,
      "loss": 7.8488,
      "step": 139
    },
    {
      "epoch": 0.040160642570281124,
      "grad_norm": 1.8386496305465698,
      "learning_rate": 2.2652592093878666e-05,
      "loss": 7.7263,
      "step": 140
    },
    {
      "epoch": 0.04044750430292599,
      "grad_norm": 1.8583810329437256,
      "learning_rate": 2.196424713241637e-05,
      "loss": 7.7595,
      "step": 141
    },
    {
      "epoch": 0.04073436603557085,
      "grad_norm": 1.8354265689849854,
      "learning_rate": 2.128356686585282e-05,
      "loss": 7.7039,
      "step": 142
    },
    {
      "epoch": 0.04102122776821572,
      "grad_norm": 1.7606815099716187,
      "learning_rate": 2.061073738537635e-05,
      "loss": 7.5557,
      "step": 143
    },
    {
      "epoch": 0.04130808950086059,
      "grad_norm": 1.671601414680481,
      "learning_rate": 1.9945942635848748e-05,
      "loss": 7.8577,
      "step": 144
    },
    {
      "epoch": 0.04159495123350545,
      "grad_norm": 1.8720660209655762,
      "learning_rate": 1.928936436551661e-05,
      "loss": 7.8286,
      "step": 145
    },
    {
      "epoch": 0.04188181296615032,
      "grad_norm": 1.9522260427474976,
      "learning_rate": 1.8641182076323148e-05,
      "loss": 7.4915,
      "step": 146
    },
    {
      "epoch": 0.04216867469879518,
      "grad_norm": 1.9963221549987793,
      "learning_rate": 1.800157297483417e-05,
      "loss": 7.5029,
      "step": 147
    },
    {
      "epoch": 0.042455536431440045,
      "grad_norm": 1.8929063081741333,
      "learning_rate": 1.7370711923791567e-05,
      "loss": 6.9962,
      "step": 148
    },
    {
      "epoch": 0.04274239816408491,
      "grad_norm": 2.757354259490967,
      "learning_rate": 1.6748771394307585e-05,
      "loss": 4.6136,
      "step": 149
    },
    {
      "epoch": 0.043029259896729774,
      "grad_norm": 4.945831775665283,
      "learning_rate": 1.6135921418712956e-05,
      "loss": 4.7154,
      "step": 150
    },
    {
      "epoch": 0.043029259896729774,
      "eval_loss": 1.9377771615982056,
      "eval_runtime": 282.5484,
      "eval_samples_per_second": 20.779,
      "eval_steps_per_second": 5.196,
      "step": 150
    },
    {
      "epoch": 0.04331612162937464,
      "grad_norm": 1.4190233945846558,
      "learning_rate": 1.553232954407171e-05,
      "loss": 8.5484,
      "step": 151
    },
    {
      "epoch": 0.04360298336201951,
      "grad_norm": 1.4434151649475098,
      "learning_rate": 1.4938160786375572e-05,
      "loss": 8.409,
      "step": 152
    },
    {
      "epoch": 0.04388984509466437,
      "grad_norm": 1.424311637878418,
      "learning_rate": 1.435357758543015e-05,
      "loss": 7.9482,
      "step": 153
    },
    {
      "epoch": 0.04417670682730924,
      "grad_norm": 1.3701916933059692,
      "learning_rate": 1.3778739760445552e-05,
      "loss": 7.8839,
      "step": 154
    },
    {
      "epoch": 0.0444635685599541,
      "grad_norm": 1.6575655937194824,
      "learning_rate": 1.3213804466343421e-05,
      "loss": 8.2305,
      "step": 155
    },
    {
      "epoch": 0.04475043029259897,
      "grad_norm": 1.5214037895202637,
      "learning_rate": 1.2658926150792322e-05,
      "loss": 9.1491,
      "step": 156
    },
    {
      "epoch": 0.045037292025243834,
      "grad_norm": 1.3671691417694092,
      "learning_rate": 1.2114256511983274e-05,
      "loss": 7.6137,
      "step": 157
    },
    {
      "epoch": 0.045324153757888695,
      "grad_norm": 1.5913439989089966,
      "learning_rate": 1.157994445715706e-05,
      "loss": 7.8928,
      "step": 158
    },
    {
      "epoch": 0.04561101549053356,
      "grad_norm": 1.4422390460968018,
      "learning_rate": 1.1056136061894384e-05,
      "loss": 8.291,
      "step": 159
    },
    {
      "epoch": 0.04589787722317843,
      "grad_norm": 1.4854753017425537,
      "learning_rate": 1.0542974530180327e-05,
      "loss": 7.656,
      "step": 160
    },
    {
      "epoch": 0.04618473895582329,
      "grad_norm": 1.5909982919692993,
      "learning_rate": 1.0040600155253765e-05,
      "loss": 6.7902,
      "step": 161
    },
    {
      "epoch": 0.04647160068846816,
      "grad_norm": 1.481354832649231,
      "learning_rate": 9.549150281252633e-06,
      "loss": 7.0523,
      "step": 162
    },
    {
      "epoch": 0.04675846242111302,
      "grad_norm": 1.4266046285629272,
      "learning_rate": 9.068759265665384e-06,
      "loss": 7.3054,
      "step": 163
    },
    {
      "epoch": 0.04704532415375789,
      "grad_norm": 1.4152878522872925,
      "learning_rate": 8.599558442598998e-06,
      "loss": 7.2685,
      "step": 164
    },
    {
      "epoch": 0.047332185886402756,
      "grad_norm": 1.4369451999664307,
      "learning_rate": 8.141676086873572e-06,
      "loss": 7.7174,
      "step": 165
    },
    {
      "epoch": 0.047619047619047616,
      "grad_norm": 1.4816739559173584,
      "learning_rate": 7.695237378953223e-06,
      "loss": 7.3185,
      "step": 166
    },
    {
      "epoch": 0.047905909351692484,
      "grad_norm": 1.4862289428710938,
      "learning_rate": 7.260364370723044e-06,
      "loss": 7.788,
      "step": 167
    },
    {
      "epoch": 0.04819277108433735,
      "grad_norm": 1.4083917140960693,
      "learning_rate": 6.837175952121306e-06,
      "loss": 8.0159,
      "step": 168
    },
    {
      "epoch": 0.04847963281698221,
      "grad_norm": 1.6308103799819946,
      "learning_rate": 6.425787818636131e-06,
      "loss": 7.3942,
      "step": 169
    },
    {
      "epoch": 0.04876649454962708,
      "grad_norm": 1.5693858861923218,
      "learning_rate": 6.026312439675552e-06,
      "loss": 8.0154,
      "step": 170
    },
    {
      "epoch": 0.04905335628227195,
      "grad_norm": 1.5414382219314575,
      "learning_rate": 5.6388590278194096e-06,
      "loss": 7.5487,
      "step": 171
    },
    {
      "epoch": 0.04934021801491681,
      "grad_norm": 1.5920078754425049,
      "learning_rate": 5.263533508961827e-06,
      "loss": 7.6236,
      "step": 172
    },
    {
      "epoch": 0.04962707974756168,
      "grad_norm": 1.5594935417175293,
      "learning_rate": 4.900438493352055e-06,
      "loss": 8.1749,
      "step": 173
    },
    {
      "epoch": 0.04991394148020654,
      "grad_norm": 1.8867764472961426,
      "learning_rate": 4.549673247541875e-06,
      "loss": 7.8641,
      "step": 174
    },
    {
      "epoch": 0.050200803212851405,
      "grad_norm": 1.6867051124572754,
      "learning_rate": 4.2113336672471245e-06,
      "loss": 7.9475,
      "step": 175
    },
    {
      "epoch": 0.05048766494549627,
      "grad_norm": 1.6259570121765137,
      "learning_rate": 3.885512251130763e-06,
      "loss": 7.2952,
      "step": 176
    },
    {
      "epoch": 0.050774526678141134,
      "grad_norm": 1.492007851600647,
      "learning_rate": 3.5722980755146517e-06,
      "loss": 7.8935,
      "step": 177
    },
    {
      "epoch": 0.051061388410786,
      "grad_norm": 1.7875683307647705,
      "learning_rate": 3.271776770026963e-06,
      "loss": 8.1946,
      "step": 178
    },
    {
      "epoch": 0.05134825014343087,
      "grad_norm": 1.941951036453247,
      "learning_rate": 2.9840304941919415e-06,
      "loss": 7.8806,
      "step": 179
    },
    {
      "epoch": 0.05163511187607573,
      "grad_norm": 1.7543801069259644,
      "learning_rate": 2.7091379149682685e-06,
      "loss": 8.0804,
      "step": 180
    },
    {
      "epoch": 0.0519219736087206,
      "grad_norm": 1.8822728395462036,
      "learning_rate": 2.4471741852423237e-06,
      "loss": 7.4285,
      "step": 181
    },
    {
      "epoch": 0.05220883534136546,
      "grad_norm": 1.6893922090530396,
      "learning_rate": 2.1982109232821178e-06,
      "loss": 7.2991,
      "step": 182
    },
    {
      "epoch": 0.05249569707401033,
      "grad_norm": 1.7529574632644653,
      "learning_rate": 1.962316193157593e-06,
      "loss": 7.2662,
      "step": 183
    },
    {
      "epoch": 0.052782558806655194,
      "grad_norm": 1.6813589334487915,
      "learning_rate": 1.7395544861325718e-06,
      "loss": 8.4343,
      "step": 184
    },
    {
      "epoch": 0.053069420539300055,
      "grad_norm": 1.794797658920288,
      "learning_rate": 1.5299867030334814e-06,
      "loss": 7.3421,
      "step": 185
    },
    {
      "epoch": 0.05335628227194492,
      "grad_norm": 1.813694715499878,
      "learning_rate": 1.333670137599713e-06,
      "loss": 7.4303,
      "step": 186
    },
    {
      "epoch": 0.05364314400458979,
      "grad_norm": 1.7764848470687866,
      "learning_rate": 1.1506584608200367e-06,
      "loss": 7.2409,
      "step": 187
    },
    {
      "epoch": 0.05393000573723465,
      "grad_norm": 1.6146047115325928,
      "learning_rate": 9.810017062595322e-07,
      "loss": 7.5112,
      "step": 188
    },
    {
      "epoch": 0.05421686746987952,
      "grad_norm": 1.809037208557129,
      "learning_rate": 8.247462563808817e-07,
      "loss": 7.6657,
      "step": 189
    },
    {
      "epoch": 0.05450372920252438,
      "grad_norm": 1.8529231548309326,
      "learning_rate": 6.819348298638839e-07,
      "loss": 7.7016,
      "step": 190
    },
    {
      "epoch": 0.05479059093516925,
      "grad_norm": 1.789148211479187,
      "learning_rate": 5.526064699265753e-07,
      "loss": 8.2768,
      "step": 191
    },
    {
      "epoch": 0.055077452667814115,
      "grad_norm": 1.7630912065505981,
      "learning_rate": 4.367965336512403e-07,
      "loss": 8.1753,
      "step": 192
    },
    {
      "epoch": 0.055364314400458976,
      "grad_norm": 1.8875055313110352,
      "learning_rate": 3.3453668231809286e-07,
      "loss": 7.4745,
      "step": 193
    },
    {
      "epoch": 0.055651176133103844,
      "grad_norm": 1.9013909101486206,
      "learning_rate": 2.458548727494292e-07,
      "loss": 7.5211,
      "step": 194
    },
    {
      "epoch": 0.05593803786574871,
      "grad_norm": 1.8378793001174927,
      "learning_rate": 1.7077534966650766e-07,
      "loss": 7.4938,
      "step": 195
    },
    {
      "epoch": 0.05622489959839357,
      "grad_norm": 2.173513174057007,
      "learning_rate": 1.0931863906127327e-07,
      "loss": 7.4864,
      "step": 196
    },
    {
      "epoch": 0.05651176133103844,
      "grad_norm": 1.9688032865524292,
      "learning_rate": 6.150154258476315e-08,
      "loss": 7.3977,
      "step": 197
    },
    {
      "epoch": 0.05679862306368331,
      "grad_norm": 2.012406349182129,
      "learning_rate": 2.7337132953697554e-08,
      "loss": 7.163,
      "step": 198
    },
    {
      "epoch": 0.05708548479632817,
      "grad_norm": 2.227052927017212,
      "learning_rate": 6.834750376549792e-09,
      "loss": 4.878,
      "step": 199
    },
    {
      "epoch": 0.05737234652897304,
      "grad_norm": 4.585555076599121,
      "learning_rate": 0.0,
      "loss": 4.5421,
      "step": 200
    },
    {
      "epoch": 0.05737234652897304,
      "eval_loss": 1.926206350326538,
      "eval_runtime": 282.6339,
      "eval_samples_per_second": 20.772,
      "eval_steps_per_second": 5.194,
      "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": 1.5108768207470592e+17,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}