{
  "best_metric": 3.076195240020752,
  "best_model_checkpoint": "miner_id_24/checkpoint-200",
  "epoch": 0.07432181345224824,
  "eval_steps": 50,
  "global_step": 200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0003716090672612412,
      "grad_norm": 89.91339111328125,
      "learning_rate": 1e-05,
      "loss": 12.8689,
      "step": 1
    },
    {
      "epoch": 0.0003716090672612412,
      "eval_loss": 3.4851932525634766,
      "eval_runtime": 51.7917,
      "eval_samples_per_second": 87.504,
      "eval_steps_per_second": 21.876,
      "step": 1
    },
    {
      "epoch": 0.0007432181345224824,
      "grad_norm": 75.49945831298828,
      "learning_rate": 2e-05,
      "loss": 13.2491,
      "step": 2
    },
    {
      "epoch": 0.0011148272017837235,
      "grad_norm": 100.03617095947266,
      "learning_rate": 3e-05,
      "loss": 12.7671,
      "step": 3
    },
    {
      "epoch": 0.0014864362690449647,
      "grad_norm": 77.02920532226562,
      "learning_rate": 4e-05,
      "loss": 13.5062,
      "step": 4
    },
    {
      "epoch": 0.0018580453363062058,
      "grad_norm": 61.53122329711914,
      "learning_rate": 5e-05,
      "loss": 13.1202,
      "step": 5
    },
    {
      "epoch": 0.002229654403567447,
      "grad_norm": 55.57686996459961,
      "learning_rate": 6e-05,
      "loss": 13.0141,
      "step": 6
    },
    {
      "epoch": 0.002601263470828688,
      "grad_norm": 50.34519577026367,
      "learning_rate": 7e-05,
      "loss": 12.3262,
      "step": 7
    },
    {
      "epoch": 0.0029728725380899295,
      "grad_norm": 49.633968353271484,
      "learning_rate": 8e-05,
      "loss": 12.7377,
      "step": 8
    },
    {
      "epoch": 0.0033444816053511705,
      "grad_norm": 58.044288635253906,
      "learning_rate": 9e-05,
      "loss": 13.0712,
      "step": 9
    },
    {
      "epoch": 0.0037160906726124115,
      "grad_norm": 55.974037170410156,
      "learning_rate": 0.0001,
      "loss": 13.4715,
      "step": 10
    },
    {
      "epoch": 0.004087699739873653,
      "grad_norm": 61.226139068603516,
      "learning_rate": 9.999316524962345e-05,
      "loss": 13.1595,
      "step": 11
    },
    {
      "epoch": 0.004459308807134894,
      "grad_norm": 67.331787109375,
      "learning_rate": 9.997266286704631e-05,
      "loss": 12.8342,
      "step": 12
    },
    {
      "epoch": 0.004830917874396135,
      "grad_norm": 49.407108306884766,
      "learning_rate": 9.993849845741524e-05,
      "loss": 13.2984,
      "step": 13
    },
    {
      "epoch": 0.005202526941657376,
      "grad_norm": 70.76209259033203,
      "learning_rate": 9.989068136093873e-05,
      "loss": 12.9778,
      "step": 14
    },
    {
      "epoch": 0.005574136008918618,
      "grad_norm": 55.556880950927734,
      "learning_rate": 9.98292246503335e-05,
      "loss": 13.2762,
      "step": 15
    },
    {
      "epoch": 0.005945745076179859,
      "grad_norm": 57.44112777709961,
      "learning_rate": 9.975414512725057e-05,
      "loss": 13.4604,
      "step": 16
    },
    {
      "epoch": 0.0063173541434411,
      "grad_norm": 61.65849304199219,
      "learning_rate": 9.966546331768191e-05,
      "loss": 13.5111,
      "step": 17
    },
    {
      "epoch": 0.006688963210702341,
      "grad_norm": 58.80717468261719,
      "learning_rate": 9.956320346634876e-05,
      "loss": 12.933,
      "step": 18
    },
    {
      "epoch": 0.007060572277963582,
      "grad_norm": 57.13853454589844,
      "learning_rate": 9.944739353007344e-05,
      "loss": 13.6066,
      "step": 19
    },
    {
      "epoch": 0.007432181345224823,
      "grad_norm": 58.7703742980957,
      "learning_rate": 9.931806517013612e-05,
      "loss": 13.801,
      "step": 20
    },
    {
      "epoch": 0.007803790412486065,
      "grad_norm": 71.7556381225586,
      "learning_rate": 9.917525374361912e-05,
      "loss": 13.8192,
      "step": 21
    },
    {
      "epoch": 0.008175399479747306,
      "grad_norm": 56.83869171142578,
      "learning_rate": 9.901899829374047e-05,
      "loss": 13.3882,
      "step": 22
    },
    {
      "epoch": 0.008547008547008548,
      "grad_norm": 68.50532531738281,
      "learning_rate": 9.884934153917997e-05,
      "loss": 13.5063,
      "step": 23
    },
    {
      "epoch": 0.008918617614269788,
      "grad_norm": 64.328369140625,
      "learning_rate": 9.86663298624003e-05,
      "loss": 13.4745,
      "step": 24
    },
    {
      "epoch": 0.00929022668153103,
      "grad_norm": 78.07742309570312,
      "learning_rate": 9.847001329696653e-05,
      "loss": 13.2238,
      "step": 25
    },
    {
      "epoch": 0.00966183574879227,
      "grad_norm": 70.4471664428711,
      "learning_rate": 9.826044551386744e-05,
      "loss": 13.0755,
      "step": 26
    },
    {
      "epoch": 0.010033444816053512,
      "grad_norm": 57.33465576171875,
      "learning_rate": 9.803768380684242e-05,
      "loss": 13.5927,
      "step": 27
    },
    {
      "epoch": 0.010405053883314752,
      "grad_norm": 54.69197463989258,
      "learning_rate": 9.780178907671789e-05,
      "loss": 13.4678,
      "step": 28
    },
    {
      "epoch": 0.010776662950575994,
      "grad_norm": 60.369022369384766,
      "learning_rate": 9.755282581475769e-05,
      "loss": 12.9709,
      "step": 29
    },
    {
      "epoch": 0.011148272017837236,
      "grad_norm": 53.06597137451172,
      "learning_rate": 9.729086208503174e-05,
      "loss": 13.1473,
      "step": 30
    },
    {
      "epoch": 0.011519881085098476,
      "grad_norm": 56.796897888183594,
      "learning_rate": 9.701596950580806e-05,
      "loss": 12.8153,
      "step": 31
    },
    {
      "epoch": 0.011891490152359718,
      "grad_norm": 55.94731903076172,
      "learning_rate": 9.672822322997305e-05,
      "loss": 12.649,
      "step": 32
    },
    {
      "epoch": 0.012263099219620958,
      "grad_norm": 60.808738708496094,
      "learning_rate": 9.642770192448536e-05,
      "loss": 12.5036,
      "step": 33
    },
    {
      "epoch": 0.0126347082868822,
      "grad_norm": 61.68337631225586,
      "learning_rate": 9.611448774886924e-05,
      "loss": 12.7729,
      "step": 34
    },
    {
      "epoch": 0.013006317354143442,
      "grad_norm": 55.44351577758789,
      "learning_rate": 9.578866633275288e-05,
      "loss": 13.3658,
      "step": 35
    },
    {
      "epoch": 0.013377926421404682,
      "grad_norm": 60.463138580322266,
      "learning_rate": 9.545032675245813e-05,
      "loss": 12.8666,
      "step": 36
    },
    {
      "epoch": 0.013749535488665924,
      "grad_norm": 60.136474609375,
      "learning_rate": 9.509956150664796e-05,
      "loss": 13.2752,
      "step": 37
    },
    {
      "epoch": 0.014121144555927164,
      "grad_norm": 61.477603912353516,
      "learning_rate": 9.473646649103818e-05,
      "loss": 13.3831,
      "step": 38
    },
    {
      "epoch": 0.014492753623188406,
      "grad_norm": 65.39891052246094,
      "learning_rate": 9.43611409721806e-05,
      "loss": 12.3385,
      "step": 39
    },
    {
      "epoch": 0.014864362690449646,
      "grad_norm": 48.332157135009766,
      "learning_rate": 9.397368756032445e-05,
      "loss": 12.8809,
      "step": 40
    },
    {
      "epoch": 0.015235971757710888,
      "grad_norm": 64.07525634765625,
      "learning_rate": 9.357421218136386e-05,
      "loss": 13.1503,
      "step": 41
    },
    {
      "epoch": 0.01560758082497213,
      "grad_norm": 61.306514739990234,
      "learning_rate": 9.316282404787871e-05,
      "loss": 13.6669,
      "step": 42
    },
    {
      "epoch": 0.015979189892233372,
      "grad_norm": 57.3792610168457,
      "learning_rate": 9.273963562927695e-05,
      "loss": 14.3407,
      "step": 43
    },
    {
      "epoch": 0.016350798959494612,
      "grad_norm": 49.27093505859375,
      "learning_rate": 9.230476262104677e-05,
      "loss": 13.7153,
      "step": 44
    },
    {
      "epoch": 0.016722408026755852,
      "grad_norm": 44.54463577270508,
      "learning_rate": 9.185832391312644e-05,
      "loss": 13.31,
      "step": 45
    },
    {
      "epoch": 0.017094017094017096,
      "grad_norm": 49.52144241333008,
      "learning_rate": 9.140044155740101e-05,
      "loss": 13.5407,
      "step": 46
    },
    {
      "epoch": 0.017465626161278336,
      "grad_norm": 59.571197509765625,
      "learning_rate": 9.093124073433463e-05,
      "loss": 13.4758,
      "step": 47
    },
    {
      "epoch": 0.017837235228539576,
      "grad_norm": 47.71146774291992,
      "learning_rate": 9.045084971874738e-05,
      "loss": 13.6893,
      "step": 48
    },
    {
      "epoch": 0.018208844295800816,
      "grad_norm": 52.587608337402344,
      "learning_rate": 8.995939984474624e-05,
      "loss": 15.3526,
      "step": 49
    },
    {
      "epoch": 0.01858045336306206,
      "grad_norm": 45.097713470458984,
      "learning_rate": 8.945702546981969e-05,
      "loss": 15.5272,
      "step": 50
    },
    {
      "epoch": 0.01858045336306206,
      "eval_loss": 3.256963014602661,
      "eval_runtime": 51.9824,
      "eval_samples_per_second": 87.183,
      "eval_steps_per_second": 21.796,
      "step": 50
    },
    {
      "epoch": 0.0189520624303233,
      "grad_norm": 20.3822021484375,
      "learning_rate": 8.894386393810563e-05,
      "loss": 12.8168,
      "step": 51
    },
    {
      "epoch": 0.01932367149758454,
      "grad_norm": 22.23453712463379,
      "learning_rate": 8.842005554284296e-05,
      "loss": 12.9706,
      "step": 52
    },
    {
      "epoch": 0.019695280564845784,
      "grad_norm": 21.059741973876953,
      "learning_rate": 8.788574348801675e-05,
      "loss": 12.7409,
      "step": 53
    },
    {
      "epoch": 0.020066889632107024,
      "grad_norm": 20.219327926635742,
      "learning_rate": 8.73410738492077e-05,
      "loss": 13.0011,
      "step": 54
    },
    {
      "epoch": 0.020438498699368264,
      "grad_norm": 23.87364959716797,
      "learning_rate": 8.678619553365659e-05,
      "loss": 12.6852,
      "step": 55
    },
    {
      "epoch": 0.020810107766629504,
      "grad_norm": 20.999147415161133,
      "learning_rate": 8.622126023955446e-05,
      "loss": 12.8306,
      "step": 56
    },
    {
      "epoch": 0.021181716833890748,
      "grad_norm": 21.016027450561523,
      "learning_rate": 8.564642241456986e-05,
      "loss": 12.928,
      "step": 57
    },
    {
      "epoch": 0.021553325901151988,
      "grad_norm": 21.58234405517578,
      "learning_rate": 8.506183921362443e-05,
      "loss": 13.0229,
      "step": 58
    },
    {
      "epoch": 0.021924934968413228,
      "grad_norm": 22.88654136657715,
      "learning_rate": 8.44676704559283e-05,
      "loss": 12.0128,
      "step": 59
    },
    {
      "epoch": 0.022296544035674472,
      "grad_norm": 20.62936019897461,
      "learning_rate": 8.386407858128706e-05,
      "loss": 12.3145,
      "step": 60
    },
    {
      "epoch": 0.022668153102935712,
      "grad_norm": 22.425981521606445,
      "learning_rate": 8.32512286056924e-05,
      "loss": 12.3533,
      "step": 61
    },
    {
      "epoch": 0.023039762170196952,
      "grad_norm": 21.995954513549805,
      "learning_rate": 8.262928807620843e-05,
      "loss": 12.8239,
      "step": 62
    },
    {
      "epoch": 0.023411371237458192,
      "grad_norm": 23.067846298217773,
      "learning_rate": 8.199842702516583e-05,
      "loss": 13.1718,
      "step": 63
    },
    {
      "epoch": 0.023782980304719436,
      "grad_norm": 27.200054168701172,
      "learning_rate": 8.135881792367686e-05,
      "loss": 12.663,
      "step": 64
    },
    {
      "epoch": 0.024154589371980676,
      "grad_norm": 23.297252655029297,
      "learning_rate": 8.07106356344834e-05,
      "loss": 12.2497,
      "step": 65
    },
    {
      "epoch": 0.024526198439241916,
      "grad_norm": 25.43965721130371,
      "learning_rate": 8.005405736415126e-05,
      "loss": 12.2888,
      "step": 66
    },
    {
      "epoch": 0.02489780750650316,
      "grad_norm": 24.800180435180664,
      "learning_rate": 7.938926261462366e-05,
      "loss": 11.9027,
      "step": 67
    },
    {
      "epoch": 0.0252694165737644,
      "grad_norm": 26.041688919067383,
      "learning_rate": 7.871643313414718e-05,
      "loss": 12.3735,
      "step": 68
    },
    {
      "epoch": 0.02564102564102564,
      "grad_norm": 24.485660552978516,
      "learning_rate": 7.803575286758364e-05,
      "loss": 12.9005,
      "step": 69
    },
    {
      "epoch": 0.026012634708286884,
      "grad_norm": 27.68210220336914,
      "learning_rate": 7.734740790612136e-05,
      "loss": 12.2957,
      "step": 70
    },
    {
      "epoch": 0.026384243775548124,
      "grad_norm": 27.342363357543945,
      "learning_rate": 7.66515864363997e-05,
      "loss": 12.4048,
      "step": 71
    },
    {
      "epoch": 0.026755852842809364,
      "grad_norm": 26.84868812561035,
      "learning_rate": 7.594847868906076e-05,
      "loss": 12.5084,
      "step": 72
    },
    {
      "epoch": 0.027127461910070604,
      "grad_norm": 28.721134185791016,
      "learning_rate": 7.52382768867422e-05,
      "loss": 12.3547,
      "step": 73
    },
    {
      "epoch": 0.027499070977331848,
      "grad_norm": 30.7074031829834,
      "learning_rate": 7.452117519152542e-05,
      "loss": 12.5016,
      "step": 74
    },
    {
      "epoch": 0.027870680044593088,
      "grad_norm": 30.42992401123047,
      "learning_rate": 7.379736965185368e-05,
      "loss": 12.6369,
      "step": 75
    },
    {
      "epoch": 0.028242289111854328,
      "grad_norm": 30.07449722290039,
      "learning_rate": 7.30670581489344e-05,
      "loss": 12.439,
      "step": 76
    },
    {
      "epoch": 0.028613898179115572,
      "grad_norm": 27.86932373046875,
      "learning_rate": 7.233044034264034e-05,
      "loss": 13.5454,
      "step": 77
    },
    {
      "epoch": 0.028985507246376812,
      "grad_norm": 24.85785484313965,
      "learning_rate": 7.158771761692464e-05,
      "loss": 12.5962,
      "step": 78
    },
    {
      "epoch": 0.029357116313638052,
      "grad_norm": 27.810264587402344,
      "learning_rate": 7.083909302476453e-05,
      "loss": 12.9454,
      "step": 79
    },
    {
      "epoch": 0.029728725380899292,
      "grad_norm": 30.700931549072266,
      "learning_rate": 7.008477123264848e-05,
      "loss": 11.9123,
      "step": 80
    },
    {
      "epoch": 0.030100334448160536,
      "grad_norm": 30.442121505737305,
      "learning_rate": 6.932495846462261e-05,
      "loss": 11.9333,
      "step": 81
    },
    {
      "epoch": 0.030471943515421776,
      "grad_norm": 35.74739074707031,
      "learning_rate": 6.855986244591104e-05,
      "loss": 11.866,
      "step": 82
    },
    {
      "epoch": 0.030843552582683016,
      "grad_norm": 33.87689208984375,
      "learning_rate": 6.778969234612584e-05,
      "loss": 12.5872,
      "step": 83
    },
    {
      "epoch": 0.03121516164994426,
      "grad_norm": 39.369693756103516,
      "learning_rate": 6.701465872208216e-05,
      "loss": 12.3437,
      "step": 84
    },
    {
      "epoch": 0.031586770717205497,
      "grad_norm": 37.18006896972656,
      "learning_rate": 6.623497346023418e-05,
      "loss": 12.5386,
      "step": 85
    },
    {
      "epoch": 0.031958379784466744,
      "grad_norm": 43.344722747802734,
      "learning_rate": 6.545084971874738e-05,
      "loss": 11.9742,
      "step": 86
    },
    {
      "epoch": 0.032329988851727984,
      "grad_norm": 38.42266845703125,
      "learning_rate": 6.466250186922325e-05,
      "loss": 12.0524,
      "step": 87
    },
    {
      "epoch": 0.032701597918989224,
      "grad_norm": 36.284507751464844,
      "learning_rate": 6.387014543809223e-05,
      "loss": 12.3979,
      "step": 88
    },
    {
      "epoch": 0.033073206986250464,
      "grad_norm": 42.46220016479492,
      "learning_rate": 6.307399704769099e-05,
      "loss": 11.8786,
      "step": 89
    },
    {
      "epoch": 0.033444816053511704,
      "grad_norm": 39.43769836425781,
      "learning_rate": 6.227427435703997e-05,
      "loss": 11.391,
      "step": 90
    },
    {
      "epoch": 0.033816425120772944,
      "grad_norm": 39.12789535522461,
      "learning_rate": 6.147119600233758e-05,
      "loss": 11.8605,
      "step": 91
    },
    {
      "epoch": 0.03418803418803419,
      "grad_norm": 41.41144561767578,
      "learning_rate": 6.066498153718735e-05,
      "loss": 12.5651,
      "step": 92
    },
    {
      "epoch": 0.03455964325529543,
      "grad_norm": 42.6218376159668,
      "learning_rate": 5.985585137257401e-05,
      "loss": 12.6488,
      "step": 93
    },
    {
      "epoch": 0.03493125232255667,
      "grad_norm": 51.30817413330078,
      "learning_rate": 5.90440267166055e-05,
      "loss": 12.5156,
      "step": 94
    },
    {
      "epoch": 0.03530286138981791,
      "grad_norm": 48.467193603515625,
      "learning_rate": 5.8229729514036705e-05,
      "loss": 12.2167,
      "step": 95
    },
    {
      "epoch": 0.03567447045707915,
      "grad_norm": 51.6811408996582,
      "learning_rate": 5.74131823855921e-05,
      "loss": 12.9945,
      "step": 96
    },
    {
      "epoch": 0.03604607952434039,
      "grad_norm": 48.028831481933594,
      "learning_rate": 5.6594608567103456e-05,
      "loss": 13.3435,
      "step": 97
    },
    {
      "epoch": 0.03641768859160163,
      "grad_norm": 50.69572448730469,
      "learning_rate": 5.577423184847932e-05,
      "loss": 13.6653,
      "step": 98
    },
    {
      "epoch": 0.03678929765886288,
      "grad_norm": 44.864540100097656,
      "learning_rate": 5.495227651252315e-05,
      "loss": 14.7863,
      "step": 99
    },
    {
      "epoch": 0.03716090672612412,
      "grad_norm": 47.65779495239258,
      "learning_rate": 5.4128967273616625e-05,
      "loss": 14.949,
      "step": 100
    },
    {
      "epoch": 0.03716090672612412,
      "eval_loss": 3.154662609100342,
      "eval_runtime": 52.0726,
      "eval_samples_per_second": 87.032,
      "eval_steps_per_second": 21.758,
      "step": 100
    },
    {
      "epoch": 0.03753251579338536,
      "grad_norm": 28.16967010498047,
      "learning_rate": 5.330452921628497e-05,
      "loss": 12.9067,
      "step": 101
    },
    {
      "epoch": 0.0379041248606466,
      "grad_norm": 30.37758445739746,
      "learning_rate": 5.247918773366112e-05,
      "loss": 12.7569,
      "step": 102
    },
    {
      "epoch": 0.03827573392790784,
      "grad_norm": 24.201208114624023,
      "learning_rate": 5.165316846586541e-05,
      "loss": 12.8335,
      "step": 103
    },
    {
      "epoch": 0.03864734299516908,
      "grad_norm": 23.62079429626465,
      "learning_rate": 5.0826697238317935e-05,
      "loss": 12.2592,
      "step": 104
    },
    {
      "epoch": 0.03901895206243032,
      "grad_norm": 22.814987182617188,
      "learning_rate": 5e-05,
      "loss": 12.4559,
      "step": 105
    },
    {
      "epoch": 0.03939056112969157,
      "grad_norm": 21.530044555664062,
      "learning_rate": 4.917330276168208e-05,
      "loss": 12.5896,
      "step": 106
    },
    {
      "epoch": 0.03976217019695281,
      "grad_norm": 23.73486328125,
      "learning_rate": 4.834683153413459e-05,
      "loss": 12.3527,
      "step": 107
    },
    {
      "epoch": 0.04013377926421405,
      "grad_norm": 21.59313201904297,
      "learning_rate": 4.7520812266338885e-05,
      "loss": 12.1843,
      "step": 108
    },
    {
      "epoch": 0.04050538833147529,
      "grad_norm": 23.252063751220703,
      "learning_rate": 4.669547078371504e-05,
      "loss": 12.442,
      "step": 109
    },
    {
      "epoch": 0.04087699739873653,
      "grad_norm": 23.04511833190918,
      "learning_rate": 4.5871032726383386e-05,
      "loss": 12.0954,
      "step": 110
    },
    {
      "epoch": 0.04124860646599777,
      "grad_norm": 21.15141487121582,
      "learning_rate": 4.504772348747687e-05,
      "loss": 12.3417,
      "step": 111
    },
    {
      "epoch": 0.04162021553325901,
      "grad_norm": 22.450578689575195,
      "learning_rate": 4.4225768151520694e-05,
      "loss": 12.1134,
      "step": 112
    },
    {
      "epoch": 0.041991824600520256,
      "grad_norm": 23.66915512084961,
      "learning_rate": 4.3405391432896555e-05,
      "loss": 12.4642,
      "step": 113
    },
    {
      "epoch": 0.042363433667781496,
      "grad_norm": 24.237220764160156,
      "learning_rate": 4.2586817614407895e-05,
      "loss": 12.2302,
      "step": 114
    },
    {
      "epoch": 0.042735042735042736,
      "grad_norm": 22.29314422607422,
      "learning_rate": 4.17702704859633e-05,
      "loss": 12.2421,
      "step": 115
    },
    {
      "epoch": 0.043106651802303976,
      "grad_norm": 25.172727584838867,
      "learning_rate": 4.095597328339452e-05,
      "loss": 12.1878,
      "step": 116
    },
    {
      "epoch": 0.043478260869565216,
      "grad_norm": 26.602157592773438,
      "learning_rate": 4.0144148627425993e-05,
      "loss": 12.2879,
      "step": 117
    },
    {
      "epoch": 0.043849869936826456,
      "grad_norm": 24.090085983276367,
      "learning_rate": 3.933501846281267e-05,
      "loss": 12.4924,
      "step": 118
    },
    {
      "epoch": 0.044221479004087696,
      "grad_norm": 26.16728973388672,
      "learning_rate": 3.852880399766243e-05,
      "loss": 12.145,
      "step": 119
    },
    {
      "epoch": 0.044593088071348944,
      "grad_norm": 27.40218734741211,
      "learning_rate": 3.772572564296005e-05,
      "loss": 11.8146,
      "step": 120
    },
    {
      "epoch": 0.044964697138610184,
      "grad_norm": 24.844179153442383,
      "learning_rate": 3.6926002952309016e-05,
      "loss": 12.1576,
      "step": 121
    },
    {
      "epoch": 0.045336306205871424,
      "grad_norm": 25.813631057739258,
      "learning_rate": 3.612985456190778e-05,
      "loss": 12.2498,
      "step": 122
    },
    {
      "epoch": 0.045707915273132664,
      "grad_norm": 27.253564834594727,
      "learning_rate": 3.533749813077677e-05,
      "loss": 13.0846,
      "step": 123
    },
    {
      "epoch": 0.046079524340393904,
      "grad_norm": 30.898393630981445,
      "learning_rate": 3.4549150281252636e-05,
      "loss": 12.5038,
      "step": 124
    },
    {
      "epoch": 0.046451133407655144,
      "grad_norm": 28.49885368347168,
      "learning_rate": 3.3765026539765834e-05,
      "loss": 12.0529,
      "step": 125
    },
    {
      "epoch": 0.046822742474916385,
      "grad_norm": 33.739097595214844,
      "learning_rate": 3.298534127791785e-05,
      "loss": 12.4644,
      "step": 126
    },
    {
      "epoch": 0.04719435154217763,
      "grad_norm": 27.413183212280273,
      "learning_rate": 3.221030765387417e-05,
      "loss": 11.8988,
      "step": 127
    },
    {
      "epoch": 0.04756596060943887,
      "grad_norm": 28.41815757751465,
      "learning_rate": 3.144013755408895e-05,
      "loss": 12.3175,
      "step": 128
    },
    {
      "epoch": 0.04793756967670011,
      "grad_norm": 24.673856735229492,
      "learning_rate": 3.0675041535377405e-05,
      "loss": 11.0224,
      "step": 129
    },
    {
      "epoch": 0.04830917874396135,
      "grad_norm": 29.299766540527344,
      "learning_rate": 2.991522876735154e-05,
      "loss": 12.2135,
      "step": 130
    },
    {
      "epoch": 0.04868078781122259,
      "grad_norm": 26.782230377197266,
      "learning_rate": 2.916090697523549e-05,
      "loss": 12.197,
      "step": 131
    },
    {
      "epoch": 0.04905239687848383,
      "grad_norm": 40.575321197509766,
      "learning_rate": 2.8412282383075363e-05,
      "loss": 12.0718,
      "step": 132
    },
    {
      "epoch": 0.04942400594574508,
      "grad_norm": 24.29773712158203,
      "learning_rate": 2.766955965735968e-05,
      "loss": 12.2244,
      "step": 133
    },
    {
      "epoch": 0.04979561501300632,
      "grad_norm": 25.015186309814453,
      "learning_rate": 2.693294185106562e-05,
      "loss": 12.3849,
      "step": 134
    },
    {
      "epoch": 0.05016722408026756,
      "grad_norm": 36.70491409301758,
      "learning_rate": 2.6202630348146324e-05,
      "loss": 11.7532,
      "step": 135
    },
    {
      "epoch": 0.0505388331475288,
      "grad_norm": 29.158506393432617,
      "learning_rate": 2.547882480847461e-05,
      "loss": 12.1616,
      "step": 136
    },
    {
      "epoch": 0.05091044221479004,
      "grad_norm": 25.96609115600586,
      "learning_rate": 2.476172311325783e-05,
      "loss": 12.0101,
      "step": 137
    },
    {
      "epoch": 0.05128205128205128,
      "grad_norm": 32.55073547363281,
      "learning_rate": 2.405152131093926e-05,
      "loss": 12.3175,
      "step": 138
    },
    {
      "epoch": 0.05165366034931252,
      "grad_norm": 28.77853775024414,
      "learning_rate": 2.3348413563600325e-05,
      "loss": 12.2389,
      "step": 139
    },
    {
      "epoch": 0.05202526941657377,
      "grad_norm": 29.920581817626953,
      "learning_rate": 2.2652592093878666e-05,
      "loss": 11.7101,
      "step": 140
    },
    {
      "epoch": 0.05239687848383501,
      "grad_norm": 30.31206512451172,
      "learning_rate": 2.196424713241637e-05,
      "loss": 12.373,
      "step": 141
    },
    {
      "epoch": 0.05276848755109625,
      "grad_norm": 34.891883850097656,
      "learning_rate": 2.128356686585282e-05,
      "loss": 12.0662,
      "step": 142
    },
    {
      "epoch": 0.05314009661835749,
      "grad_norm": 40.02173614501953,
      "learning_rate": 2.061073738537635e-05,
      "loss": 12.6474,
      "step": 143
    },
    {
      "epoch": 0.05351170568561873,
      "grad_norm": 36.588050842285156,
      "learning_rate": 1.9945942635848748e-05,
      "loss": 13.1617,
      "step": 144
    },
    {
      "epoch": 0.05388331475287997,
      "grad_norm": 34.155696868896484,
      "learning_rate": 1.928936436551661e-05,
      "loss": 12.7202,
      "step": 145
    },
    {
      "epoch": 0.05425492382014121,
      "grad_norm": 33.35672378540039,
      "learning_rate": 1.8641182076323148e-05,
      "loss": 11.7438,
      "step": 146
    },
    {
      "epoch": 0.054626532887402456,
      "grad_norm": 34.478858947753906,
      "learning_rate": 1.800157297483417e-05,
      "loss": 13.309,
      "step": 147
    },
    {
      "epoch": 0.054998141954663696,
      "grad_norm": 42.10382843017578,
      "learning_rate": 1.7370711923791567e-05,
      "loss": 13.2585,
      "step": 148
    },
    {
      "epoch": 0.055369751021924936,
      "grad_norm": 32.687721252441406,
      "learning_rate": 1.6748771394307585e-05,
      "loss": 15.1013,
      "step": 149
    },
    {
      "epoch": 0.055741360089186176,
      "grad_norm": 35.498931884765625,
      "learning_rate": 1.6135921418712956e-05,
      "loss": 15.1974,
      "step": 150
    },
    {
      "epoch": 0.055741360089186176,
      "eval_loss": 3.0939109325408936,
      "eval_runtime": 52.034,
      "eval_samples_per_second": 87.097,
      "eval_steps_per_second": 21.774,
      "step": 150
    },
    {
      "epoch": 0.056112969156447416,
      "grad_norm": 18.009008407592773,
      "learning_rate": 1.553232954407171e-05,
      "loss": 12.5073,
      "step": 151
    },
    {
      "epoch": 0.056484578223708656,
      "grad_norm": 19.124990463256836,
      "learning_rate": 1.4938160786375572e-05,
      "loss": 12.8545,
      "step": 152
    },
    {
      "epoch": 0.056856187290969896,
      "grad_norm": 19.760417938232422,
      "learning_rate": 1.435357758543015e-05,
      "loss": 12.2158,
      "step": 153
    },
    {
      "epoch": 0.057227796358231144,
      "grad_norm": 19.974246978759766,
      "learning_rate": 1.3778739760445552e-05,
      "loss": 12.0157,
      "step": 154
    },
    {
      "epoch": 0.057599405425492384,
      "grad_norm": 19.707813262939453,
      "learning_rate": 1.3213804466343421e-05,
      "loss": 12.3223,
      "step": 155
    },
    {
      "epoch": 0.057971014492753624,
      "grad_norm": 19.958818435668945,
      "learning_rate": 1.2658926150792322e-05,
      "loss": 12.4417,
      "step": 156
    },
    {
      "epoch": 0.058342623560014864,
      "grad_norm": 24.03920555114746,
      "learning_rate": 1.2114256511983274e-05,
      "loss": 12.352,
      "step": 157
    },
    {
      "epoch": 0.058714232627276104,
      "grad_norm": 20.127941131591797,
      "learning_rate": 1.157994445715706e-05,
      "loss": 12.2601,
      "step": 158
    },
    {
      "epoch": 0.059085841694537344,
      "grad_norm": 21.126739501953125,
      "learning_rate": 1.1056136061894384e-05,
      "loss": 12.4955,
      "step": 159
    },
    {
      "epoch": 0.059457450761798585,
      "grad_norm": 21.151691436767578,
      "learning_rate": 1.0542974530180327e-05,
      "loss": 11.9164,
      "step": 160
    },
    {
      "epoch": 0.05982905982905983,
      "grad_norm": 20.782501220703125,
      "learning_rate": 1.0040600155253765e-05,
      "loss": 12.0629,
      "step": 161
    },
    {
      "epoch": 0.06020066889632107,
      "grad_norm": 21.038799285888672,
      "learning_rate": 9.549150281252633e-06,
      "loss": 12.1151,
      "step": 162
    },
    {
      "epoch": 0.06057227796358231,
      "grad_norm": 21.63349151611328,
      "learning_rate": 9.068759265665384e-06,
      "loss": 12.1234,
      "step": 163
    },
    {
      "epoch": 0.06094388703084355,
      "grad_norm": 20.985212326049805,
      "learning_rate": 8.599558442598998e-06,
      "loss": 12.1586,
      "step": 164
    },
    {
      "epoch": 0.06131549609810479,
      "grad_norm": 20.65945816040039,
      "learning_rate": 8.141676086873572e-06,
      "loss": 11.9846,
      "step": 165
    },
    {
      "epoch": 0.06168710516536603,
      "grad_norm": 19.89647674560547,
      "learning_rate": 7.695237378953223e-06,
      "loss": 12.38,
      "step": 166
    },
    {
      "epoch": 0.06205871423262728,
      "grad_norm": 18.34578514099121,
      "learning_rate": 7.260364370723044e-06,
      "loss": 11.904,
      "step": 167
    },
    {
      "epoch": 0.06243032329988852,
      "grad_norm": 20.550670623779297,
      "learning_rate": 6.837175952121306e-06,
      "loss": 11.9591,
      "step": 168
    },
    {
      "epoch": 0.06280193236714976,
      "grad_norm": 22.544872283935547,
      "learning_rate": 6.425787818636131e-06,
      "loss": 11.61,
      "step": 169
    },
    {
      "epoch": 0.06317354143441099,
      "grad_norm": 21.69040298461914,
      "learning_rate": 6.026312439675552e-06,
      "loss": 12.3752,
      "step": 170
    },
    {
      "epoch": 0.06354515050167224,
      "grad_norm": 23.17534828186035,
      "learning_rate": 5.6388590278194096e-06,
      "loss": 11.8754,
      "step": 171
    },
    {
      "epoch": 0.06391675956893349,
      "grad_norm": 24.032636642456055,
      "learning_rate": 5.263533508961827e-06,
      "loss": 12.1612,
      "step": 172
    },
    {
      "epoch": 0.06428836863619472,
      "grad_norm": 25.65900421142578,
      "learning_rate": 4.900438493352055e-06,
      "loss": 12.5446,
      "step": 173
    },
    {
      "epoch": 0.06465997770345597,
      "grad_norm": 22.62616539001465,
      "learning_rate": 4.549673247541875e-06,
      "loss": 12.5365,
      "step": 174
    },
    {
      "epoch": 0.0650315867707172,
      "grad_norm": 22.821971893310547,
      "learning_rate": 4.2113336672471245e-06,
      "loss": 11.96,
      "step": 175
    },
    {
      "epoch": 0.06540319583797845,
      "grad_norm": 25.016874313354492,
      "learning_rate": 3.885512251130763e-06,
      "loss": 12.2364,
      "step": 176
    },
    {
      "epoch": 0.06577480490523968,
      "grad_norm": 21.750635147094727,
      "learning_rate": 3.5722980755146517e-06,
      "loss": 12.28,
      "step": 177
    },
    {
      "epoch": 0.06614641397250093,
      "grad_norm": 23.131282806396484,
      "learning_rate": 3.271776770026963e-06,
      "loss": 12.5989,
      "step": 178
    },
    {
      "epoch": 0.06651802303976218,
      "grad_norm": 30.947845458984375,
      "learning_rate": 2.9840304941919415e-06,
      "loss": 12.5971,
      "step": 179
    },
    {
      "epoch": 0.06688963210702341,
      "grad_norm": 24.951862335205078,
      "learning_rate": 2.7091379149682685e-06,
      "loss": 11.8475,
      "step": 180
    },
    {
      "epoch": 0.06726124117428466,
      "grad_norm": 26.0683536529541,
      "learning_rate": 2.4471741852423237e-06,
      "loss": 12.5458,
      "step": 181
    },
    {
      "epoch": 0.06763285024154589,
      "grad_norm": 23.516315460205078,
      "learning_rate": 2.1982109232821178e-06,
      "loss": 11.5941,
      "step": 182
    },
    {
      "epoch": 0.06800445930880714,
      "grad_norm": 27.001205444335938,
      "learning_rate": 1.962316193157593e-06,
      "loss": 12.2687,
      "step": 183
    },
    {
      "epoch": 0.06837606837606838,
      "grad_norm": 27.13043212890625,
      "learning_rate": 1.7395544861325718e-06,
      "loss": 11.9019,
      "step": 184
    },
    {
      "epoch": 0.06874767744332962,
      "grad_norm": 24.126502990722656,
      "learning_rate": 1.5299867030334814e-06,
      "loss": 11.2569,
      "step": 185
    },
    {
      "epoch": 0.06911928651059086,
      "grad_norm": 27.644119262695312,
      "learning_rate": 1.333670137599713e-06,
      "loss": 12.1752,
      "step": 186
    },
    {
      "epoch": 0.0694908955778521,
      "grad_norm": 28.741046905517578,
      "learning_rate": 1.1506584608200367e-06,
      "loss": 12.3401,
      "step": 187
    },
    {
      "epoch": 0.06986250464511334,
      "grad_norm": 29.9339542388916,
      "learning_rate": 9.810017062595322e-07,
      "loss": 12.243,
      "step": 188
    },
    {
      "epoch": 0.07023411371237458,
      "grad_norm": 28.82630157470703,
      "learning_rate": 8.247462563808817e-07,
      "loss": 12.8568,
      "step": 189
    },
    {
      "epoch": 0.07060572277963582,
      "grad_norm": 36.50189208984375,
      "learning_rate": 6.819348298638839e-07,
      "loss": 12.9596,
      "step": 190
    },
    {
      "epoch": 0.07097733184689707,
      "grad_norm": 37.89533233642578,
      "learning_rate": 5.526064699265753e-07,
      "loss": 13.0177,
      "step": 191
    },
    {
      "epoch": 0.0713489409141583,
      "grad_norm": 29.354293823242188,
      "learning_rate": 4.367965336512403e-07,
      "loss": 11.5076,
      "step": 192
    },
    {
      "epoch": 0.07172054998141955,
      "grad_norm": 45.329193115234375,
      "learning_rate": 3.3453668231809286e-07,
      "loss": 11.8622,
      "step": 193
    },
    {
      "epoch": 0.07209215904868078,
      "grad_norm": 34.94525909423828,
      "learning_rate": 2.458548727494292e-07,
      "loss": 12.3988,
      "step": 194
    },
    {
      "epoch": 0.07246376811594203,
      "grad_norm": 29.820608139038086,
      "learning_rate": 1.7077534966650766e-07,
      "loss": 12.1844,
      "step": 195
    },
    {
      "epoch": 0.07283537718320326,
      "grad_norm": 30.504676818847656,
      "learning_rate": 1.0931863906127327e-07,
      "loss": 12.6609,
      "step": 196
    },
    {
      "epoch": 0.07320698625046451,
      "grad_norm": 31.953720092773438,
      "learning_rate": 6.150154258476315e-08,
      "loss": 12.272,
      "step": 197
    },
    {
      "epoch": 0.07357859531772576,
      "grad_norm": 41.56123733520508,
      "learning_rate": 2.7337132953697554e-08,
      "loss": 13.9901,
      "step": 198
    },
    {
      "epoch": 0.07395020438498699,
      "grad_norm": 36.15668869018555,
      "learning_rate": 6.834750376549792e-09,
      "loss": 13.9894,
      "step": 199
    },
    {
      "epoch": 0.07432181345224824,
      "grad_norm": 40.40428161621094,
      "learning_rate": 0.0,
      "loss": 15.0873,
      "step": 200
    },
    {
      "epoch": 0.07432181345224824,
      "eval_loss": 3.076195240020752,
      "eval_runtime": 51.9862,
      "eval_samples_per_second": 87.177,
      "eval_steps_per_second": 21.794,
      "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": 1.4939692021579776e+16,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}