{
  "best_metric": 2.744900941848755,
  "best_model_checkpoint": "miner_id_24/checkpoint-200",
  "epoch": 0.08714596949891068,
  "eval_steps": 50,
  "global_step": 200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00043572984749455336,
      "grad_norm": 1.8739798069000244,
      "learning_rate": 1e-05,
      "loss": 3.3394,
      "step": 1
    },
    {
      "epoch": 0.00043572984749455336,
      "eval_loss": 6.411465644836426,
      "eval_runtime": 270.3784,
      "eval_samples_per_second": 14.298,
      "eval_steps_per_second": 3.576,
      "step": 1
    },
    {
      "epoch": 0.0008714596949891067,
      "grad_norm": 2.6025726795196533,
      "learning_rate": 2e-05,
      "loss": 4.6002,
      "step": 2
    },
    {
      "epoch": 0.00130718954248366,
      "grad_norm": 2.8490633964538574,
      "learning_rate": 3e-05,
      "loss": 4.8942,
      "step": 3
    },
    {
      "epoch": 0.0017429193899782135,
      "grad_norm": 3.4302775859832764,
      "learning_rate": 4e-05,
      "loss": 4.9611,
      "step": 4
    },
    {
      "epoch": 0.002178649237472767,
      "grad_norm": 2.9345827102661133,
      "learning_rate": 5e-05,
      "loss": 4.8143,
      "step": 5
    },
    {
      "epoch": 0.00261437908496732,
      "grad_norm": 2.4464316368103027,
      "learning_rate": 6e-05,
      "loss": 4.7977,
      "step": 6
    },
    {
      "epoch": 0.0030501089324618735,
      "grad_norm": 1.5789039134979248,
      "learning_rate": 7e-05,
      "loss": 4.4528,
      "step": 7
    },
    {
      "epoch": 0.003485838779956427,
      "grad_norm": 1.471861720085144,
      "learning_rate": 8e-05,
      "loss": 4.4727,
      "step": 8
    },
    {
      "epoch": 0.00392156862745098,
      "grad_norm": 1.4483952522277832,
      "learning_rate": 9e-05,
      "loss": 4.3232,
      "step": 9
    },
    {
      "epoch": 0.004357298474945534,
      "grad_norm": 1.194008708000183,
      "learning_rate": 0.0001,
      "loss": 4.2554,
      "step": 10
    },
    {
      "epoch": 0.004793028322440087,
      "grad_norm": 1.2611711025238037,
      "learning_rate": 9.999316524962345e-05,
      "loss": 4.1843,
      "step": 11
    },
    {
      "epoch": 0.00522875816993464,
      "grad_norm": 1.2563211917877197,
      "learning_rate": 9.997266286704631e-05,
      "loss": 3.8703,
      "step": 12
    },
    {
      "epoch": 0.005664488017429194,
      "grad_norm": 1.665711522102356,
      "learning_rate": 9.993849845741524e-05,
      "loss": 3.8937,
      "step": 13
    },
    {
      "epoch": 0.006100217864923747,
      "grad_norm": 1.3912427425384521,
      "learning_rate": 9.989068136093873e-05,
      "loss": 3.992,
      "step": 14
    },
    {
      "epoch": 0.006535947712418301,
      "grad_norm": 1.6197963953018188,
      "learning_rate": 9.98292246503335e-05,
      "loss": 3.8547,
      "step": 15
    },
    {
      "epoch": 0.006971677559912854,
      "grad_norm": 1.4797712564468384,
      "learning_rate": 9.975414512725057e-05,
      "loss": 3.7176,
      "step": 16
    },
    {
      "epoch": 0.007407407407407408,
      "grad_norm": 2.132702350616455,
      "learning_rate": 9.966546331768191e-05,
      "loss": 3.6034,
      "step": 17
    },
    {
      "epoch": 0.00784313725490196,
      "grad_norm": 1.8697417974472046,
      "learning_rate": 9.956320346634876e-05,
      "loss": 3.7192,
      "step": 18
    },
    {
      "epoch": 0.008278867102396514,
      "grad_norm": 1.627230167388916,
      "learning_rate": 9.944739353007344e-05,
      "loss": 3.4886,
      "step": 19
    },
    {
      "epoch": 0.008714596949891068,
      "grad_norm": 2.071629285812378,
      "learning_rate": 9.931806517013612e-05,
      "loss": 3.4684,
      "step": 20
    },
    {
      "epoch": 0.009150326797385621,
      "grad_norm": 1.7963043451309204,
      "learning_rate": 9.917525374361912e-05,
      "loss": 3.5541,
      "step": 21
    },
    {
      "epoch": 0.009586056644880174,
      "grad_norm": 2.466836929321289,
      "learning_rate": 9.901899829374047e-05,
      "loss": 3.7864,
      "step": 22
    },
    {
      "epoch": 0.010021786492374727,
      "grad_norm": 3.0706896781921387,
      "learning_rate": 9.884934153917997e-05,
      "loss": 3.2833,
      "step": 23
    },
    {
      "epoch": 0.01045751633986928,
      "grad_norm": 2.8026747703552246,
      "learning_rate": 9.86663298624003e-05,
      "loss": 3.5871,
      "step": 24
    },
    {
      "epoch": 0.010893246187363835,
      "grad_norm": 2.7502214908599854,
      "learning_rate": 9.847001329696653e-05,
      "loss": 3.7186,
      "step": 25
    },
    {
      "epoch": 0.011328976034858388,
      "grad_norm": 3.3825294971466064,
      "learning_rate": 9.826044551386744e-05,
      "loss": 3.6825,
      "step": 26
    },
    {
      "epoch": 0.011764705882352941,
      "grad_norm": 3.1702840328216553,
      "learning_rate": 9.803768380684242e-05,
      "loss": 3.6382,
      "step": 27
    },
    {
      "epoch": 0.012200435729847494,
      "grad_norm": 2.4032654762268066,
      "learning_rate": 9.780178907671789e-05,
      "loss": 3.4767,
      "step": 28
    },
    {
      "epoch": 0.012636165577342049,
      "grad_norm": 2.9371402263641357,
      "learning_rate": 9.755282581475769e-05,
      "loss": 3.1934,
      "step": 29
    },
    {
      "epoch": 0.013071895424836602,
      "grad_norm": 3.0755205154418945,
      "learning_rate": 9.729086208503174e-05,
      "loss": 3.4227,
      "step": 30
    },
    {
      "epoch": 0.013507625272331155,
      "grad_norm": 2.6002798080444336,
      "learning_rate": 9.701596950580806e-05,
      "loss": 3.3,
      "step": 31
    },
    {
      "epoch": 0.013943355119825708,
      "grad_norm": 2.895479679107666,
      "learning_rate": 9.672822322997305e-05,
      "loss": 2.7932,
      "step": 32
    },
    {
      "epoch": 0.01437908496732026,
      "grad_norm": 2.4496476650238037,
      "learning_rate": 9.642770192448536e-05,
      "loss": 2.6877,
      "step": 33
    },
    {
      "epoch": 0.014814814814814815,
      "grad_norm": 3.1120903491973877,
      "learning_rate": 9.611448774886924e-05,
      "loss": 3.0046,
      "step": 34
    },
    {
      "epoch": 0.015250544662309368,
      "grad_norm": 2.330878734588623,
      "learning_rate": 9.578866633275288e-05,
      "loss": 2.9339,
      "step": 35
    },
    {
      "epoch": 0.01568627450980392,
      "grad_norm": 2.8824687004089355,
      "learning_rate": 9.545032675245813e-05,
      "loss": 2.9413,
      "step": 36
    },
    {
      "epoch": 0.016122004357298474,
      "grad_norm": 2.3390908241271973,
      "learning_rate": 9.509956150664796e-05,
      "loss": 2.6987,
      "step": 37
    },
    {
      "epoch": 0.016557734204793027,
      "grad_norm": 2.63877534866333,
      "learning_rate": 9.473646649103818e-05,
      "loss": 3.1289,
      "step": 38
    },
    {
      "epoch": 0.01699346405228758,
      "grad_norm": 2.6740949153900146,
      "learning_rate": 9.43611409721806e-05,
      "loss": 2.9736,
      "step": 39
    },
    {
      "epoch": 0.017429193899782137,
      "grad_norm": 2.445444345474243,
      "learning_rate": 9.397368756032445e-05,
      "loss": 3.051,
      "step": 40
    },
    {
      "epoch": 0.01786492374727669,
      "grad_norm": 2.5592195987701416,
      "learning_rate": 9.357421218136386e-05,
      "loss": 2.5992,
      "step": 41
    },
    {
      "epoch": 0.018300653594771243,
      "grad_norm": 2.5356016159057617,
      "learning_rate": 9.316282404787871e-05,
      "loss": 2.9593,
      "step": 42
    },
    {
      "epoch": 0.018736383442265796,
      "grad_norm": 3.3945510387420654,
      "learning_rate": 9.273963562927695e-05,
      "loss": 3.1721,
      "step": 43
    },
    {
      "epoch": 0.01917211328976035,
      "grad_norm": 2.9297382831573486,
      "learning_rate": 9.230476262104677e-05,
      "loss": 3.0255,
      "step": 44
    },
    {
      "epoch": 0.0196078431372549,
      "grad_norm": 3.0529351234436035,
      "learning_rate": 9.185832391312644e-05,
      "loss": 3.314,
      "step": 45
    },
    {
      "epoch": 0.020043572984749455,
      "grad_norm": 4.365032196044922,
      "learning_rate": 9.140044155740101e-05,
      "loss": 3.314,
      "step": 46
    },
    {
      "epoch": 0.020479302832244008,
      "grad_norm": 2.6716372966766357,
      "learning_rate": 9.093124073433463e-05,
      "loss": 2.7975,
      "step": 47
    },
    {
      "epoch": 0.02091503267973856,
      "grad_norm": 3.0507326126098633,
      "learning_rate": 9.045084971874738e-05,
      "loss": 3.2709,
      "step": 48
    },
    {
      "epoch": 0.021350762527233117,
      "grad_norm": 3.8042397499084473,
      "learning_rate": 8.995939984474624e-05,
      "loss": 3.5221,
      "step": 49
    },
    {
      "epoch": 0.02178649237472767,
      "grad_norm": 11.91222095489502,
      "learning_rate": 8.945702546981969e-05,
      "loss": 3.7155,
      "step": 50
    },
    {
      "epoch": 0.02178649237472767,
      "eval_loss": 5.001107215881348,
      "eval_runtime": 272.3907,
      "eval_samples_per_second": 14.193,
      "eval_steps_per_second": 3.55,
      "step": 50
    },
    {
      "epoch": 0.022222222222222223,
      "grad_norm": 15.804675102233887,
      "learning_rate": 8.894386393810563e-05,
      "loss": 4.3576,
      "step": 51
    },
    {
      "epoch": 0.022657952069716776,
      "grad_norm": 14.754591941833496,
      "learning_rate": 8.842005554284296e-05,
      "loss": 4.8768,
      "step": 52
    },
    {
      "epoch": 0.02309368191721133,
      "grad_norm": 7.184947967529297,
      "learning_rate": 8.788574348801675e-05,
      "loss": 4.442,
      "step": 53
    },
    {
      "epoch": 0.023529411764705882,
      "grad_norm": 2.8941123485565186,
      "learning_rate": 8.73410738492077e-05,
      "loss": 3.909,
      "step": 54
    },
    {
      "epoch": 0.023965141612200435,
      "grad_norm": 2.292079448699951,
      "learning_rate": 8.678619553365659e-05,
      "loss": 3.9243,
      "step": 55
    },
    {
      "epoch": 0.024400871459694988,
      "grad_norm": 1.7536393404006958,
      "learning_rate": 8.622126023955446e-05,
      "loss": 3.603,
      "step": 56
    },
    {
      "epoch": 0.02483660130718954,
      "grad_norm": 1.7443733215332031,
      "learning_rate": 8.564642241456986e-05,
      "loss": 3.661,
      "step": 57
    },
    {
      "epoch": 0.025272331154684097,
      "grad_norm": 1.4649640321731567,
      "learning_rate": 8.506183921362443e-05,
      "loss": 3.5779,
      "step": 58
    },
    {
      "epoch": 0.02570806100217865,
      "grad_norm": 1.2020493745803833,
      "learning_rate": 8.44676704559283e-05,
      "loss": 3.3405,
      "step": 59
    },
    {
      "epoch": 0.026143790849673203,
      "grad_norm": 1.1484709978103638,
      "learning_rate": 8.386407858128706e-05,
      "loss": 3.3672,
      "step": 60
    },
    {
      "epoch": 0.026579520697167756,
      "grad_norm": 1.375651240348816,
      "learning_rate": 8.32512286056924e-05,
      "loss": 3.4851,
      "step": 61
    },
    {
      "epoch": 0.02701525054466231,
      "grad_norm": 1.6574229001998901,
      "learning_rate": 8.262928807620843e-05,
      "loss": 3.3412,
      "step": 62
    },
    {
      "epoch": 0.027450980392156862,
      "grad_norm": 1.3422226905822754,
      "learning_rate": 8.199842702516583e-05,
      "loss": 3.5403,
      "step": 63
    },
    {
      "epoch": 0.027886710239651415,
      "grad_norm": 1.2499260902404785,
      "learning_rate": 8.135881792367686e-05,
      "loss": 3.442,
      "step": 64
    },
    {
      "epoch": 0.02832244008714597,
      "grad_norm": 1.185653805732727,
      "learning_rate": 8.07106356344834e-05,
      "loss": 3.1791,
      "step": 65
    },
    {
      "epoch": 0.02875816993464052,
      "grad_norm": 1.2521653175354004,
      "learning_rate": 8.005405736415126e-05,
      "loss": 3.2464,
      "step": 66
    },
    {
      "epoch": 0.029193899782135078,
      "grad_norm": 1.4586526155471802,
      "learning_rate": 7.938926261462366e-05,
      "loss": 3.1141,
      "step": 67
    },
    {
      "epoch": 0.02962962962962963,
      "grad_norm": 2.601705312728882,
      "learning_rate": 7.871643313414718e-05,
      "loss": 2.3987,
      "step": 68
    },
    {
      "epoch": 0.030065359477124184,
      "grad_norm": 1.4250242710113525,
      "learning_rate": 7.803575286758364e-05,
      "loss": 3.0466,
      "step": 69
    },
    {
      "epoch": 0.030501089324618737,
      "grad_norm": 1.6503407955169678,
      "learning_rate": 7.734740790612136e-05,
      "loss": 2.9567,
      "step": 70
    },
    {
      "epoch": 0.03093681917211329,
      "grad_norm": 2.354557991027832,
      "learning_rate": 7.66515864363997e-05,
      "loss": 2.5175,
      "step": 71
    },
    {
      "epoch": 0.03137254901960784,
      "grad_norm": 1.949331521987915,
      "learning_rate": 7.594847868906076e-05,
      "loss": 2.9373,
      "step": 72
    },
    {
      "epoch": 0.0318082788671024,
      "grad_norm": 2.3972008228302,
      "learning_rate": 7.52382768867422e-05,
      "loss": 3.2467,
      "step": 73
    },
    {
      "epoch": 0.03224400871459695,
      "grad_norm": 2.647298574447632,
      "learning_rate": 7.452117519152542e-05,
      "loss": 3.2091,
      "step": 74
    },
    {
      "epoch": 0.032679738562091505,
      "grad_norm": 3.0107853412628174,
      "learning_rate": 7.379736965185368e-05,
      "loss": 3.4211,
      "step": 75
    },
    {
      "epoch": 0.033115468409586055,
      "grad_norm": 3.170661211013794,
      "learning_rate": 7.30670581489344e-05,
      "loss": 3.358,
      "step": 76
    },
    {
      "epoch": 0.03355119825708061,
      "grad_norm": 2.9121828079223633,
      "learning_rate": 7.233044034264034e-05,
      "loss": 3.2367,
      "step": 77
    },
    {
      "epoch": 0.03398692810457516,
      "grad_norm": 2.53003191947937,
      "learning_rate": 7.158771761692464e-05,
      "loss": 2.9897,
      "step": 78
    },
    {
      "epoch": 0.03442265795206972,
      "grad_norm": 2.4336366653442383,
      "learning_rate": 7.083909302476453e-05,
      "loss": 2.4128,
      "step": 79
    },
    {
      "epoch": 0.034858387799564274,
      "grad_norm": 2.6205925941467285,
      "learning_rate": 7.008477123264848e-05,
      "loss": 2.7036,
      "step": 80
    },
    {
      "epoch": 0.03529411764705882,
      "grad_norm": 2.623055934906006,
      "learning_rate": 6.932495846462261e-05,
      "loss": 2.5179,
      "step": 81
    },
    {
      "epoch": 0.03572984749455338,
      "grad_norm": 2.927945375442505,
      "learning_rate": 6.855986244591104e-05,
      "loss": 2.5236,
      "step": 82
    },
    {
      "epoch": 0.03616557734204793,
      "grad_norm": 2.3569791316986084,
      "learning_rate": 6.778969234612584e-05,
      "loss": 2.4889,
      "step": 83
    },
    {
      "epoch": 0.036601307189542485,
      "grad_norm": 2.9156198501586914,
      "learning_rate": 6.701465872208216e-05,
      "loss": 2.8055,
      "step": 84
    },
    {
      "epoch": 0.037037037037037035,
      "grad_norm": 3.061814785003662,
      "learning_rate": 6.623497346023418e-05,
      "loss": 2.7943,
      "step": 85
    },
    {
      "epoch": 0.03747276688453159,
      "grad_norm": 2.637924909591675,
      "learning_rate": 6.545084971874738e-05,
      "loss": 2.6648,
      "step": 86
    },
    {
      "epoch": 0.03790849673202614,
      "grad_norm": 2.399538040161133,
      "learning_rate": 6.466250186922325e-05,
      "loss": 2.5321,
      "step": 87
    },
    {
      "epoch": 0.0383442265795207,
      "grad_norm": 2.7483348846435547,
      "learning_rate": 6.387014543809223e-05,
      "loss": 2.6865,
      "step": 88
    },
    {
      "epoch": 0.038779956427015254,
      "grad_norm": 2.459527015686035,
      "learning_rate": 6.307399704769099e-05,
      "loss": 2.4833,
      "step": 89
    },
    {
      "epoch": 0.0392156862745098,
      "grad_norm": 2.7004268169403076,
      "learning_rate": 6.227427435703997e-05,
      "loss": 2.6851,
      "step": 90
    },
    {
      "epoch": 0.03965141612200436,
      "grad_norm": 2.567392349243164,
      "learning_rate": 6.147119600233758e-05,
      "loss": 2.7779,
      "step": 91
    },
    {
      "epoch": 0.04008714596949891,
      "grad_norm": 3.3482472896575928,
      "learning_rate": 6.066498153718735e-05,
      "loss": 3.1234,
      "step": 92
    },
    {
      "epoch": 0.040522875816993466,
      "grad_norm": 2.5612170696258545,
      "learning_rate": 5.985585137257401e-05,
      "loss": 2.9383,
      "step": 93
    },
    {
      "epoch": 0.040958605664488015,
      "grad_norm": 3.3072261810302734,
      "learning_rate": 5.90440267166055e-05,
      "loss": 2.7391,
      "step": 94
    },
    {
      "epoch": 0.04139433551198257,
      "grad_norm": 3.518063545227051,
      "learning_rate": 5.8229729514036705e-05,
      "loss": 2.4007,
      "step": 95
    },
    {
      "epoch": 0.04183006535947712,
      "grad_norm": 3.3459455966949463,
      "learning_rate": 5.74131823855921e-05,
      "loss": 3.2812,
      "step": 96
    },
    {
      "epoch": 0.04226579520697168,
      "grad_norm": 2.8966217041015625,
      "learning_rate": 5.6594608567103456e-05,
      "loss": 2.6464,
      "step": 97
    },
    {
      "epoch": 0.042701525054466234,
      "grad_norm": 4.4983367919921875,
      "learning_rate": 5.577423184847932e-05,
      "loss": 2.7673,
      "step": 98
    },
    {
      "epoch": 0.043137254901960784,
      "grad_norm": 3.090879440307617,
      "learning_rate": 5.495227651252315e-05,
      "loss": 2.9085,
      "step": 99
    },
    {
      "epoch": 0.04357298474945534,
      "grad_norm": 8.59165096282959,
      "learning_rate": 5.4128967273616625e-05,
      "loss": 3.1583,
      "step": 100
    },
    {
      "epoch": 0.04357298474945534,
      "eval_loss": 3.9073877334594727,
      "eval_runtime": 272.4178,
      "eval_samples_per_second": 14.191,
      "eval_steps_per_second": 3.55,
      "step": 100
    },
    {
      "epoch": 0.04400871459694989,
      "grad_norm": 4.844549655914307,
      "learning_rate": 5.330452921628497e-05,
      "loss": 3.0795,
      "step": 101
    },
    {
      "epoch": 0.044444444444444446,
      "grad_norm": 4.860598087310791,
      "learning_rate": 5.247918773366112e-05,
      "loss": 3.8187,
      "step": 102
    },
    {
      "epoch": 0.044880174291938996,
      "grad_norm": 3.8574204444885254,
      "learning_rate": 5.165316846586541e-05,
      "loss": 3.5216,
      "step": 103
    },
    {
      "epoch": 0.04531590413943355,
      "grad_norm": 2.277156352996826,
      "learning_rate": 5.0826697238317935e-05,
      "loss": 3.6533,
      "step": 104
    },
    {
      "epoch": 0.0457516339869281,
      "grad_norm": 1.8573322296142578,
      "learning_rate": 5e-05,
      "loss": 3.5722,
      "step": 105
    },
    {
      "epoch": 0.04618736383442266,
      "grad_norm": 1.4282306432724,
      "learning_rate": 4.917330276168208e-05,
      "loss": 3.1226,
      "step": 106
    },
    {
      "epoch": 0.046623093681917215,
      "grad_norm": 1.3700124025344849,
      "learning_rate": 4.834683153413459e-05,
      "loss": 3.272,
      "step": 107
    },
    {
      "epoch": 0.047058823529411764,
      "grad_norm": 1.1703954935073853,
      "learning_rate": 4.7520812266338885e-05,
      "loss": 3.3815,
      "step": 108
    },
    {
      "epoch": 0.04749455337690632,
      "grad_norm": 1.1776833534240723,
      "learning_rate": 4.669547078371504e-05,
      "loss": 3.1371,
      "step": 109
    },
    {
      "epoch": 0.04793028322440087,
      "grad_norm": 1.345227837562561,
      "learning_rate": 4.5871032726383386e-05,
      "loss": 3.192,
      "step": 110
    },
    {
      "epoch": 0.048366013071895426,
      "grad_norm": 1.2462388277053833,
      "learning_rate": 4.504772348747687e-05,
      "loss": 2.9725,
      "step": 111
    },
    {
      "epoch": 0.048801742919389976,
      "grad_norm": 1.332241415977478,
      "learning_rate": 4.4225768151520694e-05,
      "loss": 3.2054,
      "step": 112
    },
    {
      "epoch": 0.04923747276688453,
      "grad_norm": 1.3160040378570557,
      "learning_rate": 4.3405391432896555e-05,
      "loss": 2.9311,
      "step": 113
    },
    {
      "epoch": 0.04967320261437908,
      "grad_norm": 1.4543054103851318,
      "learning_rate": 4.2586817614407895e-05,
      "loss": 3.26,
      "step": 114
    },
    {
      "epoch": 0.05010893246187364,
      "grad_norm": 1.3486131429672241,
      "learning_rate": 4.17702704859633e-05,
      "loss": 2.974,
      "step": 115
    },
    {
      "epoch": 0.050544662309368195,
      "grad_norm": 1.4071096181869507,
      "learning_rate": 4.095597328339452e-05,
      "loss": 2.6225,
      "step": 116
    },
    {
      "epoch": 0.050980392156862744,
      "grad_norm": 1.479636788368225,
      "learning_rate": 4.0144148627425993e-05,
      "loss": 2.4074,
      "step": 117
    },
    {
      "epoch": 0.0514161220043573,
      "grad_norm": 1.4064018726348877,
      "learning_rate": 3.933501846281267e-05,
      "loss": 2.1019,
      "step": 118
    },
    {
      "epoch": 0.05185185185185185,
      "grad_norm": 1.546591877937317,
      "learning_rate": 3.852880399766243e-05,
      "loss": 2.8507,
      "step": 119
    },
    {
      "epoch": 0.05228758169934641,
      "grad_norm": 1.6136044263839722,
      "learning_rate": 3.772572564296005e-05,
      "loss": 2.8392,
      "step": 120
    },
    {
      "epoch": 0.052723311546840956,
      "grad_norm": 1.658595085144043,
      "learning_rate": 3.6926002952309016e-05,
      "loss": 2.3807,
      "step": 121
    },
    {
      "epoch": 0.05315904139433551,
      "grad_norm": 2.2032907009124756,
      "learning_rate": 3.612985456190778e-05,
      "loss": 3.2297,
      "step": 122
    },
    {
      "epoch": 0.05359477124183006,
      "grad_norm": 2.3491697311401367,
      "learning_rate": 3.533749813077677e-05,
      "loss": 3.2603,
      "step": 123
    },
    {
      "epoch": 0.05403050108932462,
      "grad_norm": 3.0305137634277344,
      "learning_rate": 3.4549150281252636e-05,
      "loss": 2.9607,
      "step": 124
    },
    {
      "epoch": 0.054466230936819175,
      "grad_norm": 2.5042619705200195,
      "learning_rate": 3.3765026539765834e-05,
      "loss": 2.9974,
      "step": 125
    },
    {
      "epoch": 0.054901960784313725,
      "grad_norm": 2.7703745365142822,
      "learning_rate": 3.298534127791785e-05,
      "loss": 2.9027,
      "step": 126
    },
    {
      "epoch": 0.05533769063180828,
      "grad_norm": 2.762169361114502,
      "learning_rate": 3.221030765387417e-05,
      "loss": 2.4001,
      "step": 127
    },
    {
      "epoch": 0.05577342047930283,
      "grad_norm": 3.0867695808410645,
      "learning_rate": 3.144013755408895e-05,
      "loss": 2.4678,
      "step": 128
    },
    {
      "epoch": 0.05620915032679739,
      "grad_norm": 3.359591484069824,
      "learning_rate": 3.0675041535377405e-05,
      "loss": 2.6955,
      "step": 129
    },
    {
      "epoch": 0.05664488017429194,
      "grad_norm": 3.013075351715088,
      "learning_rate": 2.991522876735154e-05,
      "loss": 2.6053,
      "step": 130
    },
    {
      "epoch": 0.05708061002178649,
      "grad_norm": 3.158442497253418,
      "learning_rate": 2.916090697523549e-05,
      "loss": 2.9712,
      "step": 131
    },
    {
      "epoch": 0.05751633986928104,
      "grad_norm": 2.680692195892334,
      "learning_rate": 2.8412282383075363e-05,
      "loss": 2.5613,
      "step": 132
    },
    {
      "epoch": 0.0579520697167756,
      "grad_norm": 2.756537437438965,
      "learning_rate": 2.766955965735968e-05,
      "loss": 2.7349,
      "step": 133
    },
    {
      "epoch": 0.058387799564270156,
      "grad_norm": 2.7178499698638916,
      "learning_rate": 2.693294185106562e-05,
      "loss": 2.5581,
      "step": 134
    },
    {
      "epoch": 0.058823529411764705,
      "grad_norm": 2.7326412200927734,
      "learning_rate": 2.6202630348146324e-05,
      "loss": 2.5431,
      "step": 135
    },
    {
      "epoch": 0.05925925925925926,
      "grad_norm": 2.7082595825195312,
      "learning_rate": 2.547882480847461e-05,
      "loss": 2.7033,
      "step": 136
    },
    {
      "epoch": 0.05969498910675381,
      "grad_norm": 2.597834825515747,
      "learning_rate": 2.476172311325783e-05,
      "loss": 2.4039,
      "step": 137
    },
    {
      "epoch": 0.06013071895424837,
      "grad_norm": 2.6988589763641357,
      "learning_rate": 2.405152131093926e-05,
      "loss": 2.3628,
      "step": 138
    },
    {
      "epoch": 0.06056644880174292,
      "grad_norm": 2.561375856399536,
      "learning_rate": 2.3348413563600325e-05,
      "loss": 2.626,
      "step": 139
    },
    {
      "epoch": 0.06100217864923747,
      "grad_norm": 2.8920652866363525,
      "learning_rate": 2.2652592093878666e-05,
      "loss": 2.4874,
      "step": 140
    },
    {
      "epoch": 0.06143790849673202,
      "grad_norm": 2.745953321456909,
      "learning_rate": 2.196424713241637e-05,
      "loss": 2.6994,
      "step": 141
    },
    {
      "epoch": 0.06187363834422658,
      "grad_norm": 2.8938546180725098,
      "learning_rate": 2.128356686585282e-05,
      "loss": 2.7683,
      "step": 142
    },
    {
      "epoch": 0.062309368191721136,
      "grad_norm": 3.5593948364257812,
      "learning_rate": 2.061073738537635e-05,
      "loss": 3.0153,
      "step": 143
    },
    {
      "epoch": 0.06274509803921569,
      "grad_norm": 2.6813762187957764,
      "learning_rate": 1.9945942635848748e-05,
      "loss": 2.7665,
      "step": 144
    },
    {
      "epoch": 0.06318082788671024,
      "grad_norm": 4.160965442657471,
      "learning_rate": 1.928936436551661e-05,
      "loss": 2.6576,
      "step": 145
    },
    {
      "epoch": 0.0636165577342048,
      "grad_norm": 3.5698304176330566,
      "learning_rate": 1.8641182076323148e-05,
      "loss": 2.6612,
      "step": 146
    },
    {
      "epoch": 0.06405228758169934,
      "grad_norm": 3.7868640422821045,
      "learning_rate": 1.800157297483417e-05,
      "loss": 3.03,
      "step": 147
    },
    {
      "epoch": 0.0644880174291939,
      "grad_norm": 3.7534987926483154,
      "learning_rate": 1.7370711923791567e-05,
      "loss": 2.9499,
      "step": 148
    },
    {
      "epoch": 0.06492374727668845,
      "grad_norm": 4.871495246887207,
      "learning_rate": 1.6748771394307585e-05,
      "loss": 3.3347,
      "step": 149
    },
    {
      "epoch": 0.06535947712418301,
      "grad_norm": 6.4089250564575195,
      "learning_rate": 1.6135921418712956e-05,
      "loss": 3.6969,
      "step": 150
    },
    {
      "epoch": 0.06535947712418301,
      "eval_loss": 2.9679818153381348,
      "eval_runtime": 272.5597,
      "eval_samples_per_second": 14.184,
      "eval_steps_per_second": 3.548,
      "step": 150
    },
    {
      "epoch": 0.06579520697167757,
      "grad_norm": 1.6404249668121338,
      "learning_rate": 1.553232954407171e-05,
      "loss": 2.6682,
      "step": 151
    },
    {
      "epoch": 0.06623093681917211,
      "grad_norm": 2.034848928451538,
      "learning_rate": 1.4938160786375572e-05,
      "loss": 3.193,
      "step": 152
    },
    {
      "epoch": 0.06666666666666667,
      "grad_norm": 2.01973295211792,
      "learning_rate": 1.435357758543015e-05,
      "loss": 3.1358,
      "step": 153
    },
    {
      "epoch": 0.06710239651416122,
      "grad_norm": 2.0093135833740234,
      "learning_rate": 1.3778739760445552e-05,
      "loss": 3.0531,
      "step": 154
    },
    {
      "epoch": 0.06753812636165578,
      "grad_norm": 1.9192451238632202,
      "learning_rate": 1.3213804466343421e-05,
      "loss": 3.2221,
      "step": 155
    },
    {
      "epoch": 0.06797385620915032,
      "grad_norm": 1.72286057472229,
      "learning_rate": 1.2658926150792322e-05,
      "loss": 2.9439,
      "step": 156
    },
    {
      "epoch": 0.06840958605664488,
      "grad_norm": 1.5966286659240723,
      "learning_rate": 1.2114256511983274e-05,
      "loss": 2.9585,
      "step": 157
    },
    {
      "epoch": 0.06884531590413943,
      "grad_norm": 1.5907948017120361,
      "learning_rate": 1.157994445715706e-05,
      "loss": 2.9148,
      "step": 158
    },
    {
      "epoch": 0.06928104575163399,
      "grad_norm": 1.539564609527588,
      "learning_rate": 1.1056136061894384e-05,
      "loss": 2.8848,
      "step": 159
    },
    {
      "epoch": 0.06971677559912855,
      "grad_norm": 1.3045552968978882,
      "learning_rate": 1.0542974530180327e-05,
      "loss": 3.1617,
      "step": 160
    },
    {
      "epoch": 0.07015250544662309,
      "grad_norm": 1.3007259368896484,
      "learning_rate": 1.0040600155253765e-05,
      "loss": 2.9992,
      "step": 161
    },
    {
      "epoch": 0.07058823529411765,
      "grad_norm": 1.2513185739517212,
      "learning_rate": 9.549150281252633e-06,
      "loss": 2.9753,
      "step": 162
    },
    {
      "epoch": 0.0710239651416122,
      "grad_norm": 1.300350308418274,
      "learning_rate": 9.068759265665384e-06,
      "loss": 2.696,
      "step": 163
    },
    {
      "epoch": 0.07145969498910676,
      "grad_norm": 1.2691736221313477,
      "learning_rate": 8.599558442598998e-06,
      "loss": 3.0073,
      "step": 164
    },
    {
      "epoch": 0.0718954248366013,
      "grad_norm": 1.2600065469741821,
      "learning_rate": 8.141676086873572e-06,
      "loss": 2.9148,
      "step": 165
    },
    {
      "epoch": 0.07233115468409586,
      "grad_norm": 1.5369460582733154,
      "learning_rate": 7.695237378953223e-06,
      "loss": 2.864,
      "step": 166
    },
    {
      "epoch": 0.07276688453159041,
      "grad_norm": 1.4069645404815674,
      "learning_rate": 7.260364370723044e-06,
      "loss": 2.9255,
      "step": 167
    },
    {
      "epoch": 0.07320261437908497,
      "grad_norm": 1.4547737836837769,
      "learning_rate": 6.837175952121306e-06,
      "loss": 2.4449,
      "step": 168
    },
    {
      "epoch": 0.07363834422657953,
      "grad_norm": 1.6118872165679932,
      "learning_rate": 6.425787818636131e-06,
      "loss": 2.6838,
      "step": 169
    },
    {
      "epoch": 0.07407407407407407,
      "grad_norm": 1.8233840465545654,
      "learning_rate": 6.026312439675552e-06,
      "loss": 3.0111,
      "step": 170
    },
    {
      "epoch": 0.07450980392156863,
      "grad_norm": 1.7893010377883911,
      "learning_rate": 5.6388590278194096e-06,
      "loss": 2.3569,
      "step": 171
    },
    {
      "epoch": 0.07494553376906318,
      "grad_norm": 2.110873222351074,
      "learning_rate": 5.263533508961827e-06,
      "loss": 2.9776,
      "step": 172
    },
    {
      "epoch": 0.07538126361655774,
      "grad_norm": 3.9048421382904053,
      "learning_rate": 4.900438493352055e-06,
      "loss": 3.0479,
      "step": 173
    },
    {
      "epoch": 0.07581699346405228,
      "grad_norm": 5.902091026306152,
      "learning_rate": 4.549673247541875e-06,
      "loss": 3.5068,
      "step": 174
    },
    {
      "epoch": 0.07625272331154684,
      "grad_norm": 6.569426536560059,
      "learning_rate": 4.2113336672471245e-06,
      "loss": 3.2988,
      "step": 175
    },
    {
      "epoch": 0.0766884531590414,
      "grad_norm": 5.962541580200195,
      "learning_rate": 3.885512251130763e-06,
      "loss": 2.862,
      "step": 176
    },
    {
      "epoch": 0.07712418300653595,
      "grad_norm": 5.360730171203613,
      "learning_rate": 3.5722980755146517e-06,
      "loss": 2.8814,
      "step": 177
    },
    {
      "epoch": 0.07755991285403051,
      "grad_norm": 4.786664962768555,
      "learning_rate": 3.271776770026963e-06,
      "loss": 2.7255,
      "step": 178
    },
    {
      "epoch": 0.07799564270152505,
      "grad_norm": 4.172339916229248,
      "learning_rate": 2.9840304941919415e-06,
      "loss": 2.5797,
      "step": 179
    },
    {
      "epoch": 0.0784313725490196,
      "grad_norm": 3.4753611087799072,
      "learning_rate": 2.7091379149682685e-06,
      "loss": 2.6019,
      "step": 180
    },
    {
      "epoch": 0.07886710239651416,
      "grad_norm": 3.5787813663482666,
      "learning_rate": 2.4471741852423237e-06,
      "loss": 2.4009,
      "step": 181
    },
    {
      "epoch": 0.07930283224400872,
      "grad_norm": 3.9149110317230225,
      "learning_rate": 2.1982109232821178e-06,
      "loss": 2.8912,
      "step": 182
    },
    {
      "epoch": 0.07973856209150326,
      "grad_norm": 3.64512038230896,
      "learning_rate": 1.962316193157593e-06,
      "loss": 2.8082,
      "step": 183
    },
    {
      "epoch": 0.08017429193899782,
      "grad_norm": 3.0708370208740234,
      "learning_rate": 1.7395544861325718e-06,
      "loss": 2.4552,
      "step": 184
    },
    {
      "epoch": 0.08061002178649238,
      "grad_norm": 3.097843647003174,
      "learning_rate": 1.5299867030334814e-06,
      "loss": 2.2263,
      "step": 185
    },
    {
      "epoch": 0.08104575163398693,
      "grad_norm": 3.018770694732666,
      "learning_rate": 1.333670137599713e-06,
      "loss": 2.4401,
      "step": 186
    },
    {
      "epoch": 0.08148148148148149,
      "grad_norm": 2.7417657375335693,
      "learning_rate": 1.1506584608200367e-06,
      "loss": 2.363,
      "step": 187
    },
    {
      "epoch": 0.08191721132897603,
      "grad_norm": 3.513536214828491,
      "learning_rate": 9.810017062595322e-07,
      "loss": 2.7238,
      "step": 188
    },
    {
      "epoch": 0.08235294117647059,
      "grad_norm": 2.9783589839935303,
      "learning_rate": 8.247462563808817e-07,
      "loss": 2.4749,
      "step": 189
    },
    {
      "epoch": 0.08278867102396514,
      "grad_norm": 2.5987086296081543,
      "learning_rate": 6.819348298638839e-07,
      "loss": 2.354,
      "step": 190
    },
    {
      "epoch": 0.0832244008714597,
      "grad_norm": 3.0120317935943604,
      "learning_rate": 5.526064699265753e-07,
      "loss": 2.3191,
      "step": 191
    },
    {
      "epoch": 0.08366013071895424,
      "grad_norm": 3.8709325790405273,
      "learning_rate": 4.367965336512403e-07,
      "loss": 2.7396,
      "step": 192
    },
    {
      "epoch": 0.0840958605664488,
      "grad_norm": 2.7988052368164062,
      "learning_rate": 3.3453668231809286e-07,
      "loss": 2.5555,
      "step": 193
    },
    {
      "epoch": 0.08453159041394336,
      "grad_norm": 2.9350998401641846,
      "learning_rate": 2.458548727494292e-07,
      "loss": 2.6935,
      "step": 194
    },
    {
      "epoch": 0.08496732026143791,
      "grad_norm": 3.2310903072357178,
      "learning_rate": 1.7077534966650766e-07,
      "loss": 2.2639,
      "step": 195
    },
    {
      "epoch": 0.08540305010893247,
      "grad_norm": 3.172008991241455,
      "learning_rate": 1.0931863906127327e-07,
      "loss": 2.8087,
      "step": 196
    },
    {
      "epoch": 0.08583877995642701,
      "grad_norm": 3.7203173637390137,
      "learning_rate": 6.150154258476315e-08,
      "loss": 2.7701,
      "step": 197
    },
    {
      "epoch": 0.08627450980392157,
      "grad_norm": 3.748854637145996,
      "learning_rate": 2.7337132953697554e-08,
      "loss": 2.8688,
      "step": 198
    },
    {
      "epoch": 0.08671023965141612,
      "grad_norm": 4.906322479248047,
      "learning_rate": 6.834750376549792e-09,
      "loss": 3.3322,
      "step": 199
    },
    {
      "epoch": 0.08714596949891068,
      "grad_norm": 7.119653701782227,
      "learning_rate": 0.0,
      "loss": 3.3285,
      "step": 200
    },
    {
      "epoch": 0.08714596949891068,
      "eval_loss": 2.744900941848755,
      "eval_runtime": 272.4101,
      "eval_samples_per_second": 14.192,
      "eval_steps_per_second": 3.55,
      "step": 200
    }
  ],
  "logging_steps": 1,
  "max_steps": 200,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 50,
  "stateful_callbacks": {
    "EarlyStoppingCallback": {
      "args": {
        "early_stopping_patience": 5,
        "early_stopping_threshold": 0.0
      },
      "attributes": {
        "early_stopping_patience_counter": 0
      }
    },
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2.6676698616220877e+17,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}