{
  "best_metric": 2.1184213161468506,
  "best_model_checkpoint": "miner_id_24/checkpoint-200",
  "epoch": 0.7992007992007992,
  "eval_steps": 50,
  "global_step": 200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.003996003996003996,
      "grad_norm": 3.910423994064331,
      "learning_rate": 1e-05,
      "loss": 2.1215,
      "step": 1
    },
    {
      "epoch": 0.003996003996003996,
      "eval_loss": 3.552821159362793,
      "eval_runtime": 7.3726,
      "eval_samples_per_second": 57.239,
      "eval_steps_per_second": 14.378,
      "step": 1
    },
    {
      "epoch": 0.007992007992007992,
      "grad_norm": 4.006191253662109,
      "learning_rate": 2e-05,
      "loss": 2.1664,
      "step": 2
    },
    {
      "epoch": 0.011988011988011988,
      "grad_norm": 4.1190996170043945,
      "learning_rate": 3e-05,
      "loss": 2.3106,
      "step": 3
    },
    {
      "epoch": 0.015984015984015984,
      "grad_norm": 4.885846138000488,
      "learning_rate": 4e-05,
      "loss": 2.2605,
      "step": 4
    },
    {
      "epoch": 0.01998001998001998,
      "grad_norm": 4.7745256423950195,
      "learning_rate": 5e-05,
      "loss": 2.5102,
      "step": 5
    },
    {
      "epoch": 0.023976023976023976,
      "grad_norm": 4.240265369415283,
      "learning_rate": 6e-05,
      "loss": 2.5953,
      "step": 6
    },
    {
      "epoch": 0.027972027972027972,
      "grad_norm": 4.884115219116211,
      "learning_rate": 7e-05,
      "loss": 2.4478,
      "step": 7
    },
    {
      "epoch": 0.03196803196803197,
      "grad_norm": 5.712406635284424,
      "learning_rate": 8e-05,
      "loss": 2.4999,
      "step": 8
    },
    {
      "epoch": 0.03596403596403597,
      "grad_norm": 4.88950252532959,
      "learning_rate": 9e-05,
      "loss": 2.5788,
      "step": 9
    },
    {
      "epoch": 0.03996003996003996,
      "grad_norm": 4.2379841804504395,
      "learning_rate": 0.0001,
      "loss": 2.5255,
      "step": 10
    },
    {
      "epoch": 0.04395604395604396,
      "grad_norm": 3.854949474334717,
      "learning_rate": 9.999316524962345e-05,
      "loss": 2.5106,
      "step": 11
    },
    {
      "epoch": 0.04795204795204795,
      "grad_norm": 4.00169038772583,
      "learning_rate": 9.997266286704631e-05,
      "loss": 2.6816,
      "step": 12
    },
    {
      "epoch": 0.05194805194805195,
      "grad_norm": 4.485476493835449,
      "learning_rate": 9.993849845741524e-05,
      "loss": 2.7688,
      "step": 13
    },
    {
      "epoch": 0.055944055944055944,
      "grad_norm": 4.359032154083252,
      "learning_rate": 9.989068136093873e-05,
      "loss": 2.8721,
      "step": 14
    },
    {
      "epoch": 0.059940059940059943,
      "grad_norm": 4.10183048248291,
      "learning_rate": 9.98292246503335e-05,
      "loss": 2.5593,
      "step": 15
    },
    {
      "epoch": 0.06393606393606394,
      "grad_norm": 4.140375137329102,
      "learning_rate": 9.975414512725057e-05,
      "loss": 2.5027,
      "step": 16
    },
    {
      "epoch": 0.06793206793206794,
      "grad_norm": 4.694578170776367,
      "learning_rate": 9.966546331768191e-05,
      "loss": 2.4229,
      "step": 17
    },
    {
      "epoch": 0.07192807192807193,
      "grad_norm": 4.611745357513428,
      "learning_rate": 9.956320346634876e-05,
      "loss": 2.5744,
      "step": 18
    },
    {
      "epoch": 0.07592407592407592,
      "grad_norm": 4.780086040496826,
      "learning_rate": 9.944739353007344e-05,
      "loss": 2.6549,
      "step": 19
    },
    {
      "epoch": 0.07992007992007992,
      "grad_norm": 4.67626953125,
      "learning_rate": 9.931806517013612e-05,
      "loss": 2.5075,
      "step": 20
    },
    {
      "epoch": 0.08391608391608392,
      "grad_norm": 4.548355579376221,
      "learning_rate": 9.917525374361912e-05,
      "loss": 2.7661,
      "step": 21
    },
    {
      "epoch": 0.08791208791208792,
      "grad_norm": 4.729110240936279,
      "learning_rate": 9.901899829374047e-05,
      "loss": 2.7792,
      "step": 22
    },
    {
      "epoch": 0.0919080919080919,
      "grad_norm": 5.016458988189697,
      "learning_rate": 9.884934153917997e-05,
      "loss": 2.7903,
      "step": 23
    },
    {
      "epoch": 0.0959040959040959,
      "grad_norm": 5.362691879272461,
      "learning_rate": 9.86663298624003e-05,
      "loss": 2.6737,
      "step": 24
    },
    {
      "epoch": 0.0999000999000999,
      "grad_norm": 4.384696006774902,
      "learning_rate": 9.847001329696653e-05,
      "loss": 2.4998,
      "step": 25
    },
    {
      "epoch": 0.1038961038961039,
      "grad_norm": 5.082944869995117,
      "learning_rate": 9.826044551386744e-05,
      "loss": 2.5172,
      "step": 26
    },
    {
      "epoch": 0.10789210789210789,
      "grad_norm": 4.218101501464844,
      "learning_rate": 9.803768380684242e-05,
      "loss": 2.4249,
      "step": 27
    },
    {
      "epoch": 0.11188811188811189,
      "grad_norm": 4.740158557891846,
      "learning_rate": 9.780178907671789e-05,
      "loss": 2.6899,
      "step": 28
    },
    {
      "epoch": 0.11588411588411589,
      "grad_norm": 5.189453601837158,
      "learning_rate": 9.755282581475769e-05,
      "loss": 2.2796,
      "step": 29
    },
    {
      "epoch": 0.11988011988011989,
      "grad_norm": 5.367314338684082,
      "learning_rate": 9.729086208503174e-05,
      "loss": 2.5304,
      "step": 30
    },
    {
      "epoch": 0.12387612387612387,
      "grad_norm": 4.755152702331543,
      "learning_rate": 9.701596950580806e-05,
      "loss": 2.287,
      "step": 31
    },
    {
      "epoch": 0.12787212787212787,
      "grad_norm": 4.233535289764404,
      "learning_rate": 9.672822322997305e-05,
      "loss": 2.4822,
      "step": 32
    },
    {
      "epoch": 0.13186813186813187,
      "grad_norm": 5.043883323669434,
      "learning_rate": 9.642770192448536e-05,
      "loss": 2.3557,
      "step": 33
    },
    {
      "epoch": 0.13586413586413587,
      "grad_norm": 4.452463150024414,
      "learning_rate": 9.611448774886924e-05,
      "loss": 2.2276,
      "step": 34
    },
    {
      "epoch": 0.13986013986013987,
      "grad_norm": 4.7310051918029785,
      "learning_rate": 9.578866633275288e-05,
      "loss": 2.3177,
      "step": 35
    },
    {
      "epoch": 0.14385614385614387,
      "grad_norm": 4.885829925537109,
      "learning_rate": 9.545032675245813e-05,
      "loss": 2.4561,
      "step": 36
    },
    {
      "epoch": 0.14785214785214784,
      "grad_norm": 5.9802422523498535,
      "learning_rate": 9.509956150664796e-05,
      "loss": 2.4706,
      "step": 37
    },
    {
      "epoch": 0.15184815184815184,
      "grad_norm": 4.736691951751709,
      "learning_rate": 9.473646649103818e-05,
      "loss": 2.3638,
      "step": 38
    },
    {
      "epoch": 0.15584415584415584,
      "grad_norm": 4.2223734855651855,
      "learning_rate": 9.43611409721806e-05,
      "loss": 2.0792,
      "step": 39
    },
    {
      "epoch": 0.15984015984015984,
      "grad_norm": 4.105318546295166,
      "learning_rate": 9.397368756032445e-05,
      "loss": 1.9058,
      "step": 40
    },
    {
      "epoch": 0.16383616383616384,
      "grad_norm": 3.9662559032440186,
      "learning_rate": 9.357421218136386e-05,
      "loss": 1.9316,
      "step": 41
    },
    {
      "epoch": 0.16783216783216784,
      "grad_norm": 4.318179130554199,
      "learning_rate": 9.316282404787871e-05,
      "loss": 1.7193,
      "step": 42
    },
    {
      "epoch": 0.17182817182817184,
      "grad_norm": 5.287116527557373,
      "learning_rate": 9.273963562927695e-05,
      "loss": 2.1934,
      "step": 43
    },
    {
      "epoch": 0.17582417582417584,
      "grad_norm": 4.280655384063721,
      "learning_rate": 9.230476262104677e-05,
      "loss": 1.8134,
      "step": 44
    },
    {
      "epoch": 0.1798201798201798,
      "grad_norm": 5.311466217041016,
      "learning_rate": 9.185832391312644e-05,
      "loss": 1.6951,
      "step": 45
    },
    {
      "epoch": 0.1838161838161838,
      "grad_norm": 4.899798393249512,
      "learning_rate": 9.140044155740101e-05,
      "loss": 1.9689,
      "step": 46
    },
    {
      "epoch": 0.1878121878121878,
      "grad_norm": 4.340550422668457,
      "learning_rate": 9.093124073433463e-05,
      "loss": 2.018,
      "step": 47
    },
    {
      "epoch": 0.1918081918081918,
      "grad_norm": 4.7391581535339355,
      "learning_rate": 9.045084971874738e-05,
      "loss": 1.7269,
      "step": 48
    },
    {
      "epoch": 0.1958041958041958,
      "grad_norm": 4.794632911682129,
      "learning_rate": 8.995939984474624e-05,
      "loss": 1.8895,
      "step": 49
    },
    {
      "epoch": 0.1998001998001998,
      "grad_norm": 6.271996021270752,
      "learning_rate": 8.945702546981969e-05,
      "loss": 1.8417,
      "step": 50
    },
    {
      "epoch": 0.1998001998001998,
      "eval_loss": 2.408557653427124,
      "eval_runtime": 7.3755,
      "eval_samples_per_second": 57.217,
      "eval_steps_per_second": 14.372,
      "step": 50
    },
    {
      "epoch": 0.2037962037962038,
      "grad_norm": 6.911890506744385,
      "learning_rate": 8.894386393810563e-05,
      "loss": 2.1049,
      "step": 51
    },
    {
      "epoch": 0.2077922077922078,
      "grad_norm": 5.056367874145508,
      "learning_rate": 8.842005554284296e-05,
      "loss": 1.9969,
      "step": 52
    },
    {
      "epoch": 0.21178821178821178,
      "grad_norm": 3.1770966053009033,
      "learning_rate": 8.788574348801675e-05,
      "loss": 1.9205,
      "step": 53
    },
    {
      "epoch": 0.21578421578421578,
      "grad_norm": 2.8177223205566406,
      "learning_rate": 8.73410738492077e-05,
      "loss": 2.0638,
      "step": 54
    },
    {
      "epoch": 0.21978021978021978,
      "grad_norm": 2.7501065731048584,
      "learning_rate": 8.678619553365659e-05,
      "loss": 2.0561,
      "step": 55
    },
    {
      "epoch": 0.22377622377622378,
      "grad_norm": 2.8661201000213623,
      "learning_rate": 8.622126023955446e-05,
      "loss": 2.2725,
      "step": 56
    },
    {
      "epoch": 0.22777222777222778,
      "grad_norm": 3.2395591735839844,
      "learning_rate": 8.564642241456986e-05,
      "loss": 2.3776,
      "step": 57
    },
    {
      "epoch": 0.23176823176823177,
      "grad_norm": 3.423804759979248,
      "learning_rate": 8.506183921362443e-05,
      "loss": 2.6703,
      "step": 58
    },
    {
      "epoch": 0.23576423576423577,
      "grad_norm": 3.303666353225708,
      "learning_rate": 8.44676704559283e-05,
      "loss": 2.5395,
      "step": 59
    },
    {
      "epoch": 0.23976023976023977,
      "grad_norm": 3.6006393432617188,
      "learning_rate": 8.386407858128706e-05,
      "loss": 2.5096,
      "step": 60
    },
    {
      "epoch": 0.24375624375624375,
      "grad_norm": 3.439183235168457,
      "learning_rate": 8.32512286056924e-05,
      "loss": 2.5177,
      "step": 61
    },
    {
      "epoch": 0.24775224775224775,
      "grad_norm": 3.9183597564697266,
      "learning_rate": 8.262928807620843e-05,
      "loss": 2.5746,
      "step": 62
    },
    {
      "epoch": 0.2517482517482518,
      "grad_norm": 3.8047871589660645,
      "learning_rate": 8.199842702516583e-05,
      "loss": 2.646,
      "step": 63
    },
    {
      "epoch": 0.25574425574425574,
      "grad_norm": 3.4592788219451904,
      "learning_rate": 8.135881792367686e-05,
      "loss": 2.3938,
      "step": 64
    },
    {
      "epoch": 0.2597402597402597,
      "grad_norm": 3.3289968967437744,
      "learning_rate": 8.07106356344834e-05,
      "loss": 2.6414,
      "step": 65
    },
    {
      "epoch": 0.26373626373626374,
      "grad_norm": 3.6790871620178223,
      "learning_rate": 8.005405736415126e-05,
      "loss": 2.6041,
      "step": 66
    },
    {
      "epoch": 0.2677322677322677,
      "grad_norm": 3.4457921981811523,
      "learning_rate": 7.938926261462366e-05,
      "loss": 2.1067,
      "step": 67
    },
    {
      "epoch": 0.27172827172827174,
      "grad_norm": 4.019956588745117,
      "learning_rate": 7.871643313414718e-05,
      "loss": 2.5417,
      "step": 68
    },
    {
      "epoch": 0.2757242757242757,
      "grad_norm": 4.262936592102051,
      "learning_rate": 7.803575286758364e-05,
      "loss": 2.5911,
      "step": 69
    },
    {
      "epoch": 0.27972027972027974,
      "grad_norm": 4.201218605041504,
      "learning_rate": 7.734740790612136e-05,
      "loss": 2.5448,
      "step": 70
    },
    {
      "epoch": 0.2837162837162837,
      "grad_norm": 3.414463996887207,
      "learning_rate": 7.66515864363997e-05,
      "loss": 2.196,
      "step": 71
    },
    {
      "epoch": 0.28771228771228774,
      "grad_norm": 3.717339515686035,
      "learning_rate": 7.594847868906076e-05,
      "loss": 2.5187,
      "step": 72
    },
    {
      "epoch": 0.2917082917082917,
      "grad_norm": 3.4383299350738525,
      "learning_rate": 7.52382768867422e-05,
      "loss": 2.1614,
      "step": 73
    },
    {
      "epoch": 0.2957042957042957,
      "grad_norm": 3.9660472869873047,
      "learning_rate": 7.452117519152542e-05,
      "loss": 2.4798,
      "step": 74
    },
    {
      "epoch": 0.2997002997002997,
      "grad_norm": 3.9959959983825684,
      "learning_rate": 7.379736965185368e-05,
      "loss": 2.6558,
      "step": 75
    },
    {
      "epoch": 0.3036963036963037,
      "grad_norm": 3.653013229370117,
      "learning_rate": 7.30670581489344e-05,
      "loss": 1.9736,
      "step": 76
    },
    {
      "epoch": 0.3076923076923077,
      "grad_norm": 3.6085028648376465,
      "learning_rate": 7.233044034264034e-05,
      "loss": 2.2469,
      "step": 77
    },
    {
      "epoch": 0.3116883116883117,
      "grad_norm": 4.391664505004883,
      "learning_rate": 7.158771761692464e-05,
      "loss": 2.4239,
      "step": 78
    },
    {
      "epoch": 0.3156843156843157,
      "grad_norm": 4.454457759857178,
      "learning_rate": 7.083909302476453e-05,
      "loss": 2.5525,
      "step": 79
    },
    {
      "epoch": 0.3196803196803197,
      "grad_norm": 3.9900012016296387,
      "learning_rate": 7.008477123264848e-05,
      "loss": 2.4509,
      "step": 80
    },
    {
      "epoch": 0.32367632367632365,
      "grad_norm": 4.066675186157227,
      "learning_rate": 6.932495846462261e-05,
      "loss": 2.3202,
      "step": 81
    },
    {
      "epoch": 0.3276723276723277,
      "grad_norm": 3.738349676132202,
      "learning_rate": 6.855986244591104e-05,
      "loss": 2.1657,
      "step": 82
    },
    {
      "epoch": 0.33166833166833165,
      "grad_norm": 3.7364532947540283,
      "learning_rate": 6.778969234612584e-05,
      "loss": 2.3194,
      "step": 83
    },
    {
      "epoch": 0.3356643356643357,
      "grad_norm": 4.640803813934326,
      "learning_rate": 6.701465872208216e-05,
      "loss": 2.293,
      "step": 84
    },
    {
      "epoch": 0.33966033966033965,
      "grad_norm": 5.231298923492432,
      "learning_rate": 6.623497346023418e-05,
      "loss": 2.5187,
      "step": 85
    },
    {
      "epoch": 0.3436563436563437,
      "grad_norm": 3.566009759902954,
      "learning_rate": 6.545084971874738e-05,
      "loss": 2.2727,
      "step": 86
    },
    {
      "epoch": 0.34765234765234765,
      "grad_norm": 3.6806423664093018,
      "learning_rate": 6.466250186922325e-05,
      "loss": 2.0119,
      "step": 87
    },
    {
      "epoch": 0.3516483516483517,
      "grad_norm": 3.7033236026763916,
      "learning_rate": 6.387014543809223e-05,
      "loss": 2.0647,
      "step": 88
    },
    {
      "epoch": 0.35564435564435565,
      "grad_norm": 3.734884023666382,
      "learning_rate": 6.307399704769099e-05,
      "loss": 1.9621,
      "step": 89
    },
    {
      "epoch": 0.3596403596403596,
      "grad_norm": 4.0137434005737305,
      "learning_rate": 6.227427435703997e-05,
      "loss": 1.9407,
      "step": 90
    },
    {
      "epoch": 0.36363636363636365,
      "grad_norm": 3.5453217029571533,
      "learning_rate": 6.147119600233758e-05,
      "loss": 1.8273,
      "step": 91
    },
    {
      "epoch": 0.3676323676323676,
      "grad_norm": 4.36460018157959,
      "learning_rate": 6.066498153718735e-05,
      "loss": 2.0767,
      "step": 92
    },
    {
      "epoch": 0.37162837162837165,
      "grad_norm": 3.9199297428131104,
      "learning_rate": 5.985585137257401e-05,
      "loss": 1.755,
      "step": 93
    },
    {
      "epoch": 0.3756243756243756,
      "grad_norm": 3.9695377349853516,
      "learning_rate": 5.90440267166055e-05,
      "loss": 1.7596,
      "step": 94
    },
    {
      "epoch": 0.37962037962037964,
      "grad_norm": 3.833652973175049,
      "learning_rate": 5.8229729514036705e-05,
      "loss": 1.8189,
      "step": 95
    },
    {
      "epoch": 0.3836163836163836,
      "grad_norm": 3.3542890548706055,
      "learning_rate": 5.74131823855921e-05,
      "loss": 1.7539,
      "step": 96
    },
    {
      "epoch": 0.3876123876123876,
      "grad_norm": 4.119942665100098,
      "learning_rate": 5.6594608567103456e-05,
      "loss": 1.8562,
      "step": 97
    },
    {
      "epoch": 0.3916083916083916,
      "grad_norm": 3.4092016220092773,
      "learning_rate": 5.577423184847932e-05,
      "loss": 1.5566,
      "step": 98
    },
    {
      "epoch": 0.3956043956043956,
      "grad_norm": 3.7112979888916016,
      "learning_rate": 5.495227651252315e-05,
      "loss": 1.7212,
      "step": 99
    },
    {
      "epoch": 0.3996003996003996,
      "grad_norm": 4.721656799316406,
      "learning_rate": 5.4128967273616625e-05,
      "loss": 1.7987,
      "step": 100
    },
    {
      "epoch": 0.3996003996003996,
      "eval_loss": 2.1971940994262695,
      "eval_runtime": 7.3736,
      "eval_samples_per_second": 57.231,
      "eval_steps_per_second": 14.376,
      "step": 100
    },
    {
      "epoch": 0.4035964035964036,
      "grad_norm": 2.137390375137329,
      "learning_rate": 5.330452921628497e-05,
      "loss": 1.7543,
      "step": 101
    },
    {
      "epoch": 0.4075924075924076,
      "grad_norm": 2.097003698348999,
      "learning_rate": 5.247918773366112e-05,
      "loss": 1.6215,
      "step": 102
    },
    {
      "epoch": 0.4115884115884116,
      "grad_norm": 2.1690244674682617,
      "learning_rate": 5.165316846586541e-05,
      "loss": 1.6602,
      "step": 103
    },
    {
      "epoch": 0.4155844155844156,
      "grad_norm": 2.294987678527832,
      "learning_rate": 5.0826697238317935e-05,
      "loss": 1.8769,
      "step": 104
    },
    {
      "epoch": 0.4195804195804196,
      "grad_norm": 2.4727396965026855,
      "learning_rate": 5e-05,
      "loss": 1.9594,
      "step": 105
    },
    {
      "epoch": 0.42357642357642356,
      "grad_norm": 2.6860365867614746,
      "learning_rate": 4.917330276168208e-05,
      "loss": 2.1873,
      "step": 106
    },
    {
      "epoch": 0.4275724275724276,
      "grad_norm": 2.7127106189727783,
      "learning_rate": 4.834683153413459e-05,
      "loss": 2.3708,
      "step": 107
    },
    {
      "epoch": 0.43156843156843155,
      "grad_norm": 2.6260106563568115,
      "learning_rate": 4.7520812266338885e-05,
      "loss": 2.2876,
      "step": 108
    },
    {
      "epoch": 0.4355644355644356,
      "grad_norm": 2.768350601196289,
      "learning_rate": 4.669547078371504e-05,
      "loss": 2.3566,
      "step": 109
    },
    {
      "epoch": 0.43956043956043955,
      "grad_norm": 3.178553342819214,
      "learning_rate": 4.5871032726383386e-05,
      "loss": 2.6334,
      "step": 110
    },
    {
      "epoch": 0.4435564435564436,
      "grad_norm": 3.3227720260620117,
      "learning_rate": 4.504772348747687e-05,
      "loss": 2.5637,
      "step": 111
    },
    {
      "epoch": 0.44755244755244755,
      "grad_norm": 3.1499240398406982,
      "learning_rate": 4.4225768151520694e-05,
      "loss": 2.4859,
      "step": 112
    },
    {
      "epoch": 0.4515484515484515,
      "grad_norm": 2.7684566974639893,
      "learning_rate": 4.3405391432896555e-05,
      "loss": 2.1494,
      "step": 113
    },
    {
      "epoch": 0.45554445554445555,
      "grad_norm": 3.4616615772247314,
      "learning_rate": 4.2586817614407895e-05,
      "loss": 2.1193,
      "step": 114
    },
    {
      "epoch": 0.4595404595404595,
      "grad_norm": 3.2123327255249023,
      "learning_rate": 4.17702704859633e-05,
      "loss": 2.5909,
      "step": 115
    },
    {
      "epoch": 0.46353646353646355,
      "grad_norm": 3.047816753387451,
      "learning_rate": 4.095597328339452e-05,
      "loss": 2.0229,
      "step": 116
    },
    {
      "epoch": 0.4675324675324675,
      "grad_norm": 3.488494396209717,
      "learning_rate": 4.0144148627425993e-05,
      "loss": 2.2963,
      "step": 117
    },
    {
      "epoch": 0.47152847152847155,
      "grad_norm": 3.587735891342163,
      "learning_rate": 3.933501846281267e-05,
      "loss": 2.2049,
      "step": 118
    },
    {
      "epoch": 0.4755244755244755,
      "grad_norm": 3.056833505630493,
      "learning_rate": 3.852880399766243e-05,
      "loss": 1.8591,
      "step": 119
    },
    {
      "epoch": 0.47952047952047955,
      "grad_norm": 3.218198776245117,
      "learning_rate": 3.772572564296005e-05,
      "loss": 2.2895,
      "step": 120
    },
    {
      "epoch": 0.4835164835164835,
      "grad_norm": 3.371720314025879,
      "learning_rate": 3.6926002952309016e-05,
      "loss": 2.4071,
      "step": 121
    },
    {
      "epoch": 0.4875124875124875,
      "grad_norm": 3.7063543796539307,
      "learning_rate": 3.612985456190778e-05,
      "loss": 2.2914,
      "step": 122
    },
    {
      "epoch": 0.4915084915084915,
      "grad_norm": 3.269237518310547,
      "learning_rate": 3.533749813077677e-05,
      "loss": 2.2477,
      "step": 123
    },
    {
      "epoch": 0.4955044955044955,
      "grad_norm": 3.816117525100708,
      "learning_rate": 3.4549150281252636e-05,
      "loss": 2.3856,
      "step": 124
    },
    {
      "epoch": 0.4995004995004995,
      "grad_norm": 3.4660842418670654,
      "learning_rate": 3.3765026539765834e-05,
      "loss": 2.0488,
      "step": 125
    },
    {
      "epoch": 0.5034965034965035,
      "grad_norm": 3.094273328781128,
      "learning_rate": 3.298534127791785e-05,
      "loss": 2.1264,
      "step": 126
    },
    {
      "epoch": 0.5074925074925075,
      "grad_norm": 3.5829083919525146,
      "learning_rate": 3.221030765387417e-05,
      "loss": 2.1837,
      "step": 127
    },
    {
      "epoch": 0.5114885114885115,
      "grad_norm": 4.308327674865723,
      "learning_rate": 3.144013755408895e-05,
      "loss": 2.1031,
      "step": 128
    },
    {
      "epoch": 0.5154845154845155,
      "grad_norm": 3.1789207458496094,
      "learning_rate": 3.0675041535377405e-05,
      "loss": 2.2333,
      "step": 129
    },
    {
      "epoch": 0.5194805194805194,
      "grad_norm": 3.8885791301727295,
      "learning_rate": 2.991522876735154e-05,
      "loss": 2.266,
      "step": 130
    },
    {
      "epoch": 0.5234765234765235,
      "grad_norm": 4.62883186340332,
      "learning_rate": 2.916090697523549e-05,
      "loss": 2.3179,
      "step": 131
    },
    {
      "epoch": 0.5274725274725275,
      "grad_norm": 4.235198497772217,
      "learning_rate": 2.8412282383075363e-05,
      "loss": 2.2452,
      "step": 132
    },
    {
      "epoch": 0.5314685314685315,
      "grad_norm": 3.6604795455932617,
      "learning_rate": 2.766955965735968e-05,
      "loss": 2.2782,
      "step": 133
    },
    {
      "epoch": 0.5354645354645354,
      "grad_norm": 3.711357831954956,
      "learning_rate": 2.693294185106562e-05,
      "loss": 2.3947,
      "step": 134
    },
    {
      "epoch": 0.5394605394605395,
      "grad_norm": 3.2384161949157715,
      "learning_rate": 2.6202630348146324e-05,
      "loss": 1.8946,
      "step": 135
    },
    {
      "epoch": 0.5434565434565435,
      "grad_norm": 3.569737672805786,
      "learning_rate": 2.547882480847461e-05,
      "loss": 2.1051,
      "step": 136
    },
    {
      "epoch": 0.5474525474525475,
      "grad_norm": 4.256958961486816,
      "learning_rate": 2.476172311325783e-05,
      "loss": 1.8519,
      "step": 137
    },
    {
      "epoch": 0.5514485514485514,
      "grad_norm": 3.9488539695739746,
      "learning_rate": 2.405152131093926e-05,
      "loss": 2.3631,
      "step": 138
    },
    {
      "epoch": 0.5554445554445554,
      "grad_norm": 3.606555223464966,
      "learning_rate": 2.3348413563600325e-05,
      "loss": 1.8756,
      "step": 139
    },
    {
      "epoch": 0.5594405594405595,
      "grad_norm": 3.8353779315948486,
      "learning_rate": 2.2652592093878666e-05,
      "loss": 2.2032,
      "step": 140
    },
    {
      "epoch": 0.5634365634365635,
      "grad_norm": 2.9946072101593018,
      "learning_rate": 2.196424713241637e-05,
      "loss": 1.5247,
      "step": 141
    },
    {
      "epoch": 0.5674325674325674,
      "grad_norm": 4.481592655181885,
      "learning_rate": 2.128356686585282e-05,
      "loss": 2.3981,
      "step": 142
    },
    {
      "epoch": 0.5714285714285714,
      "grad_norm": 3.6957621574401855,
      "learning_rate": 2.061073738537635e-05,
      "loss": 2.1031,
      "step": 143
    },
    {
      "epoch": 0.5754245754245755,
      "grad_norm": 4.094795227050781,
      "learning_rate": 1.9945942635848748e-05,
      "loss": 1.5181,
      "step": 144
    },
    {
      "epoch": 0.5794205794205795,
      "grad_norm": 4.133942127227783,
      "learning_rate": 1.928936436551661e-05,
      "loss": 1.9686,
      "step": 145
    },
    {
      "epoch": 0.5834165834165834,
      "grad_norm": 3.659259557723999,
      "learning_rate": 1.8641182076323148e-05,
      "loss": 1.7523,
      "step": 146
    },
    {
      "epoch": 0.5874125874125874,
      "grad_norm": 3.222001791000366,
      "learning_rate": 1.800157297483417e-05,
      "loss": 1.6258,
      "step": 147
    },
    {
      "epoch": 0.5914085914085914,
      "grad_norm": 3.470677614212036,
      "learning_rate": 1.7370711923791567e-05,
      "loss": 1.5365,
      "step": 148
    },
    {
      "epoch": 0.5954045954045954,
      "grad_norm": 4.083232402801514,
      "learning_rate": 1.6748771394307585e-05,
      "loss": 2.005,
      "step": 149
    },
    {
      "epoch": 0.5994005994005994,
      "grad_norm": 4.134530544281006,
      "learning_rate": 1.6135921418712956e-05,
      "loss": 1.5766,
      "step": 150
    },
    {
      "epoch": 0.5994005994005994,
      "eval_loss": 2.1309282779693604,
      "eval_runtime": 7.3681,
      "eval_samples_per_second": 57.274,
      "eval_steps_per_second": 14.386,
      "step": 150
    },
    {
      "epoch": 0.6033966033966034,
      "grad_norm": 1.708056926727295,
      "learning_rate": 1.553232954407171e-05,
      "loss": 1.5096,
      "step": 151
    },
    {
      "epoch": 0.6073926073926074,
      "grad_norm": 2.054734945297241,
      "learning_rate": 1.4938160786375572e-05,
      "loss": 1.7285,
      "step": 152
    },
    {
      "epoch": 0.6113886113886113,
      "grad_norm": 2.0054144859313965,
      "learning_rate": 1.435357758543015e-05,
      "loss": 1.8064,
      "step": 153
    },
    {
      "epoch": 0.6153846153846154,
      "grad_norm": 1.9428958892822266,
      "learning_rate": 1.3778739760445552e-05,
      "loss": 1.6102,
      "step": 154
    },
    {
      "epoch": 0.6193806193806194,
      "grad_norm": 2.3465700149536133,
      "learning_rate": 1.3213804466343421e-05,
      "loss": 2.0891,
      "step": 155
    },
    {
      "epoch": 0.6233766233766234,
      "grad_norm": 2.355849027633667,
      "learning_rate": 1.2658926150792322e-05,
      "loss": 1.8678,
      "step": 156
    },
    {
      "epoch": 0.6273726273726273,
      "grad_norm": 2.2271692752838135,
      "learning_rate": 1.2114256511983274e-05,
      "loss": 2.2338,
      "step": 157
    },
    {
      "epoch": 0.6313686313686314,
      "grad_norm": 2.71498966217041,
      "learning_rate": 1.157994445715706e-05,
      "loss": 2.2785,
      "step": 158
    },
    {
      "epoch": 0.6353646353646354,
      "grad_norm": 2.537884473800659,
      "learning_rate": 1.1056136061894384e-05,
      "loss": 2.3177,
      "step": 159
    },
    {
      "epoch": 0.6393606393606394,
      "grad_norm": 2.685389280319214,
      "learning_rate": 1.0542974530180327e-05,
      "loss": 2.4615,
      "step": 160
    },
    {
      "epoch": 0.6433566433566433,
      "grad_norm": 3.180253744125366,
      "learning_rate": 1.0040600155253765e-05,
      "loss": 2.6329,
      "step": 161
    },
    {
      "epoch": 0.6473526473526473,
      "grad_norm": 3.1242849826812744,
      "learning_rate": 9.549150281252633e-06,
      "loss": 2.3837,
      "step": 162
    },
    {
      "epoch": 0.6513486513486514,
      "grad_norm": 3.144672393798828,
      "learning_rate": 9.068759265665384e-06,
      "loss": 2.2722,
      "step": 163
    },
    {
      "epoch": 0.6553446553446554,
      "grad_norm": 3.5322771072387695,
      "learning_rate": 8.599558442598998e-06,
      "loss": 2.3846,
      "step": 164
    },
    {
      "epoch": 0.6593406593406593,
      "grad_norm": 3.1850361824035645,
      "learning_rate": 8.141676086873572e-06,
      "loss": 2.1708,
      "step": 165
    },
    {
      "epoch": 0.6633366633366633,
      "grad_norm": 4.227579593658447,
      "learning_rate": 7.695237378953223e-06,
      "loss": 2.4368,
      "step": 166
    },
    {
      "epoch": 0.6673326673326674,
      "grad_norm": 3.3327395915985107,
      "learning_rate": 7.260364370723044e-06,
      "loss": 2.4646,
      "step": 167
    },
    {
      "epoch": 0.6713286713286714,
      "grad_norm": 3.2608273029327393,
      "learning_rate": 6.837175952121306e-06,
      "loss": 2.1706,
      "step": 168
    },
    {
      "epoch": 0.6753246753246753,
      "grad_norm": 3.7514514923095703,
      "learning_rate": 6.425787818636131e-06,
      "loss": 2.2712,
      "step": 169
    },
    {
      "epoch": 0.6793206793206793,
      "grad_norm": 3.3286333084106445,
      "learning_rate": 6.026312439675552e-06,
      "loss": 2.2386,
      "step": 170
    },
    {
      "epoch": 0.6833166833166833,
      "grad_norm": 3.551285982131958,
      "learning_rate": 5.6388590278194096e-06,
      "loss": 2.245,
      "step": 171
    },
    {
      "epoch": 0.6873126873126874,
      "grad_norm": 3.367011785507202,
      "learning_rate": 5.263533508961827e-06,
      "loss": 2.15,
      "step": 172
    },
    {
      "epoch": 0.6913086913086913,
      "grad_norm": 3.3601531982421875,
      "learning_rate": 4.900438493352055e-06,
      "loss": 2.317,
      "step": 173
    },
    {
      "epoch": 0.6953046953046953,
      "grad_norm": 3.221449851989746,
      "learning_rate": 4.549673247541875e-06,
      "loss": 2.1853,
      "step": 174
    },
    {
      "epoch": 0.6993006993006993,
      "grad_norm": 3.9633541107177734,
      "learning_rate": 4.2113336672471245e-06,
      "loss": 2.4192,
      "step": 175
    },
    {
      "epoch": 0.7032967032967034,
      "grad_norm": 3.4577462673187256,
      "learning_rate": 3.885512251130763e-06,
      "loss": 2.0783,
      "step": 176
    },
    {
      "epoch": 0.7072927072927073,
      "grad_norm": 3.5073444843292236,
      "learning_rate": 3.5722980755146517e-06,
      "loss": 2.2715,
      "step": 177
    },
    {
      "epoch": 0.7112887112887113,
      "grad_norm": 3.881265163421631,
      "learning_rate": 3.271776770026963e-06,
      "loss": 2.2345,
      "step": 178
    },
    {
      "epoch": 0.7152847152847153,
      "grad_norm": 3.6585028171539307,
      "learning_rate": 2.9840304941919415e-06,
      "loss": 2.1253,
      "step": 179
    },
    {
      "epoch": 0.7192807192807192,
      "grad_norm": 3.2855727672576904,
      "learning_rate": 2.7091379149682685e-06,
      "loss": 1.9831,
      "step": 180
    },
    {
      "epoch": 0.7232767232767233,
      "grad_norm": 3.994173765182495,
      "learning_rate": 2.4471741852423237e-06,
      "loss": 2.381,
      "step": 181
    },
    {
      "epoch": 0.7272727272727273,
      "grad_norm": 3.5176312923431396,
      "learning_rate": 2.1982109232821178e-06,
      "loss": 1.9802,
      "step": 182
    },
    {
      "epoch": 0.7312687312687313,
      "grad_norm": 3.536353826522827,
      "learning_rate": 1.962316193157593e-06,
      "loss": 2.1217,
      "step": 183
    },
    {
      "epoch": 0.7352647352647352,
      "grad_norm": 3.537053108215332,
      "learning_rate": 1.7395544861325718e-06,
      "loss": 2.0782,
      "step": 184
    },
    {
      "epoch": 0.7392607392607392,
      "grad_norm": 3.4762892723083496,
      "learning_rate": 1.5299867030334814e-06,
      "loss": 1.9867,
      "step": 185
    },
    {
      "epoch": 0.7432567432567433,
      "grad_norm": 4.145040988922119,
      "learning_rate": 1.333670137599713e-06,
      "loss": 2.2047,
      "step": 186
    },
    {
      "epoch": 0.7472527472527473,
      "grad_norm": 4.042585849761963,
      "learning_rate": 1.1506584608200367e-06,
      "loss": 2.1904,
      "step": 187
    },
    {
      "epoch": 0.7512487512487512,
      "grad_norm": 3.6518189907073975,
      "learning_rate": 9.810017062595322e-07,
      "loss": 2.0323,
      "step": 188
    },
    {
      "epoch": 0.7552447552447552,
      "grad_norm": 3.288264513015747,
      "learning_rate": 8.247462563808817e-07,
      "loss": 1.5844,
      "step": 189
    },
    {
      "epoch": 0.7592407592407593,
      "grad_norm": 4.0721845626831055,
      "learning_rate": 6.819348298638839e-07,
      "loss": 2.0555,
      "step": 190
    },
    {
      "epoch": 0.7632367632367633,
      "grad_norm": 3.880002737045288,
      "learning_rate": 5.526064699265753e-07,
      "loss": 2.0766,
      "step": 191
    },
    {
      "epoch": 0.7672327672327672,
      "grad_norm": 3.861696720123291,
      "learning_rate": 4.367965336512403e-07,
      "loss": 1.9422,
      "step": 192
    },
    {
      "epoch": 0.7712287712287712,
      "grad_norm": 3.4892125129699707,
      "learning_rate": 3.3453668231809286e-07,
      "loss": 2.1398,
      "step": 193
    },
    {
      "epoch": 0.7752247752247752,
      "grad_norm": 3.707634210586548,
      "learning_rate": 2.458548727494292e-07,
      "loss": 1.7149,
      "step": 194
    },
    {
      "epoch": 0.7792207792207793,
      "grad_norm": 3.2461206912994385,
      "learning_rate": 1.7077534966650766e-07,
      "loss": 1.6082,
      "step": 195
    },
    {
      "epoch": 0.7832167832167832,
      "grad_norm": 3.7491872310638428,
      "learning_rate": 1.0931863906127327e-07,
      "loss": 1.8066,
      "step": 196
    },
    {
      "epoch": 0.7872127872127872,
      "grad_norm": 4.023907661437988,
      "learning_rate": 6.150154258476315e-08,
      "loss": 2.0895,
      "step": 197
    },
    {
      "epoch": 0.7912087912087912,
      "grad_norm": 3.4938385486602783,
      "learning_rate": 2.7337132953697554e-08,
      "loss": 1.4863,
      "step": 198
    },
    {
      "epoch": 0.7952047952047953,
      "grad_norm": 3.258133888244629,
      "learning_rate": 6.834750376549792e-09,
      "loss": 1.5181,
      "step": 199
    },
    {
      "epoch": 0.7992007992007992,
      "grad_norm": 4.141680717468262,
      "learning_rate": 0.0,
      "loss": 1.7035,
      "step": 200
    },
    {
      "epoch": 0.7992007992007992,
      "eval_loss": 2.1184213161468506,
      "eval_runtime": 7.3672,
      "eval_samples_per_second": 57.281,
      "eval_steps_per_second": 14.388,
      "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": 1.54569648635904e+16,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}