{
  "best_metric": 0.5852221250534058,
  "best_model_checkpoint": "/ephemeral/models/qwen-describe_tasks/checkpoint-200",
  "epoch": 6.0,
  "eval_steps": 200,
  "global_step": 1344,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.004464285714285714,
      "grad_norm": 10.935125350952148,
      "learning_rate": 0.0,
      "loss": 1.3545,
      "step": 1
    },
    {
      "epoch": 0.008928571428571428,
      "grad_norm": 7.386569976806641,
      "learning_rate": 2.626495350371936e-06,
      "loss": 1.0726,
      "step": 2
    },
    {
      "epoch": 0.013392857142857142,
      "grad_norm": 7.35675048828125,
      "learning_rate": 4.162896638657993e-06,
      "loss": 1.0713,
      "step": 3
    },
    {
      "epoch": 0.017857142857142856,
      "grad_norm": 6.851252555847168,
      "learning_rate": 5.252990700743872e-06,
      "loss": 0.9934,
      "step": 4
    },
    {
      "epoch": 0.022321428571428572,
      "grad_norm": 6.426003456115723,
      "learning_rate": 6.098533345119624e-06,
      "loss": 0.7561,
      "step": 5
    },
    {
      "epoch": 0.026785714285714284,
      "grad_norm": 4.757523536682129,
      "learning_rate": 6.7893919890299284e-06,
      "loss": 0.7267,
      "step": 6
    },
    {
      "epoch": 0.03125,
      "grad_norm": 3.9246585369110107,
      "learning_rate": 7.373504649628066e-06,
      "loss": 0.6757,
      "step": 7
    },
    {
      "epoch": 0.03571428571428571,
      "grad_norm": 5.09559965133667,
      "learning_rate": 7.879486051115807e-06,
      "loss": 0.7255,
      "step": 8
    },
    {
      "epoch": 0.04017857142857143,
      "grad_norm": 4.388647079467773,
      "learning_rate": 8.325793277315987e-06,
      "loss": 0.6737,
      "step": 9
    },
    {
      "epoch": 0.044642857142857144,
      "grad_norm": 3.693603754043579,
      "learning_rate": 8.72502869549156e-06,
      "loss": 0.6406,
      "step": 10
    },
    {
      "epoch": 0.049107142857142856,
      "grad_norm": 3.9717981815338135,
      "learning_rate": 9.086181061280522e-06,
      "loss": 0.587,
      "step": 11
    },
    {
      "epoch": 0.05357142857142857,
      "grad_norm": 3.3187167644500732,
      "learning_rate": 9.415887339401865e-06,
      "loss": 0.5724,
      "step": 12
    },
    {
      "epoch": 0.05803571428571429,
      "grad_norm": 3.3204233646392822,
      "learning_rate": 9.719187714029216e-06,
      "loss": 0.6456,
      "step": 13
    },
    {
      "epoch": 0.0625,
      "grad_norm": 3.1636106967926025,
      "learning_rate": 1e-05,
      "loss": 0.573,
      "step": 14
    },
    {
      "epoch": 0.06696428571428571,
      "grad_norm": 3.5804100036621094,
      "learning_rate": 9.999986052613417e-06,
      "loss": 0.5863,
      "step": 15
    },
    {
      "epoch": 0.07142857142857142,
      "grad_norm": 3.142025947570801,
      "learning_rate": 9.99994421053148e-06,
      "loss": 0.6012,
      "step": 16
    },
    {
      "epoch": 0.07589285714285714,
      "grad_norm": 2.6348273754119873,
      "learning_rate": 9.999874473987653e-06,
      "loss": 0.5631,
      "step": 17
    },
    {
      "epoch": 0.08035714285714286,
      "grad_norm": 2.788102149963379,
      "learning_rate": 9.999776843371027e-06,
      "loss": 0.6074,
      "step": 18
    },
    {
      "epoch": 0.08482142857142858,
      "grad_norm": 2.5704562664031982,
      "learning_rate": 9.99965131922634e-06,
      "loss": 0.5218,
      "step": 19
    },
    {
      "epoch": 0.08928571428571429,
      "grad_norm": 2.773559331893921,
      "learning_rate": 9.999497902253949e-06,
      "loss": 0.5921,
      "step": 20
    },
    {
      "epoch": 0.09375,
      "grad_norm": 2.585707664489746,
      "learning_rate": 9.999316593309849e-06,
      "loss": 0.603,
      "step": 21
    },
    {
      "epoch": 0.09821428571428571,
      "grad_norm": 2.896451711654663,
      "learning_rate": 9.999107393405655e-06,
      "loss": 0.588,
      "step": 22
    },
    {
      "epoch": 0.10267857142857142,
      "grad_norm": 2.606309652328491,
      "learning_rate": 9.998870303708601e-06,
      "loss": 0.4943,
      "step": 23
    },
    {
      "epoch": 0.10714285714285714,
      "grad_norm": 2.6549859046936035,
      "learning_rate": 9.998605325541531e-06,
      "loss": 0.5824,
      "step": 24
    },
    {
      "epoch": 0.11160714285714286,
      "grad_norm": 2.7567014694213867,
      "learning_rate": 9.998312460382895e-06,
      "loss": 0.5915,
      "step": 25
    },
    {
      "epoch": 0.11607142857142858,
      "grad_norm": 2.674490451812744,
      "learning_rate": 9.997991709866738e-06,
      "loss": 0.5946,
      "step": 26
    },
    {
      "epoch": 0.12053571428571429,
      "grad_norm": 2.0534379482269287,
      "learning_rate": 9.997643075782691e-06,
      "loss": 0.4713,
      "step": 27
    },
    {
      "epoch": 0.125,
      "grad_norm": 2.5476534366607666,
      "learning_rate": 9.997266560075961e-06,
      "loss": 0.5565,
      "step": 28
    },
    {
      "epoch": 0.12946428571428573,
      "grad_norm": 3.0594515800476074,
      "learning_rate": 9.996862164847323e-06,
      "loss": 0.5195,
      "step": 29
    },
    {
      "epoch": 0.13392857142857142,
      "grad_norm": 2.6829309463500977,
      "learning_rate": 9.996429892353107e-06,
      "loss": 0.4949,
      "step": 30
    },
    {
      "epoch": 0.13839285714285715,
      "grad_norm": 2.6954853534698486,
      "learning_rate": 9.99596974500518e-06,
      "loss": 0.5276,
      "step": 31
    },
    {
      "epoch": 0.14285714285714285,
      "grad_norm": 2.8353872299194336,
      "learning_rate": 9.995481725370941e-06,
      "loss": 0.5713,
      "step": 32
    },
    {
      "epoch": 0.14732142857142858,
      "grad_norm": 2.503706693649292,
      "learning_rate": 9.994965836173303e-06,
      "loss": 0.5171,
      "step": 33
    },
    {
      "epoch": 0.15178571428571427,
      "grad_norm": 2.876476526260376,
      "learning_rate": 9.994422080290675e-06,
      "loss": 0.5046,
      "step": 34
    },
    {
      "epoch": 0.15625,
      "grad_norm": 2.6266987323760986,
      "learning_rate": 9.99385046075695e-06,
      "loss": 0.5366,
      "step": 35
    },
    {
      "epoch": 0.16071428571428573,
      "grad_norm": 2.4998559951782227,
      "learning_rate": 9.993250980761487e-06,
      "loss": 0.5082,
      "step": 36
    },
    {
      "epoch": 0.16517857142857142,
      "grad_norm": 2.6557369232177734,
      "learning_rate": 9.99262364364909e-06,
      "loss": 0.5126,
      "step": 37
    },
    {
      "epoch": 0.16964285714285715,
      "grad_norm": 2.4919795989990234,
      "learning_rate": 9.991968452919999e-06,
      "loss": 0.459,
      "step": 38
    },
    {
      "epoch": 0.17410714285714285,
      "grad_norm": 2.501042604446411,
      "learning_rate": 9.991285412229854e-06,
      "loss": 0.5021,
      "step": 39
    },
    {
      "epoch": 0.17857142857142858,
      "grad_norm": 2.462800979614258,
      "learning_rate": 9.99057452538969e-06,
      "loss": 0.5227,
      "step": 40
    },
    {
      "epoch": 0.18303571428571427,
      "grad_norm": 2.2987539768218994,
      "learning_rate": 9.989835796365911e-06,
      "loss": 0.491,
      "step": 41
    },
    {
      "epoch": 0.1875,
      "grad_norm": 2.5660648345947266,
      "learning_rate": 9.989069229280264e-06,
      "loss": 0.5264,
      "step": 42
    },
    {
      "epoch": 0.19196428571428573,
      "grad_norm": 2.3724122047424316,
      "learning_rate": 9.988274828409821e-06,
      "loss": 0.5362,
      "step": 43
    },
    {
      "epoch": 0.19642857142857142,
      "grad_norm": 2.1337292194366455,
      "learning_rate": 9.987452598186947e-06,
      "loss": 0.4827,
      "step": 44
    },
    {
      "epoch": 0.20089285714285715,
      "grad_norm": 2.4473042488098145,
      "learning_rate": 9.986602543199292e-06,
      "loss": 0.5034,
      "step": 45
    },
    {
      "epoch": 0.20535714285714285,
      "grad_norm": 2.3975181579589844,
      "learning_rate": 9.985724668189744e-06,
      "loss": 0.5155,
      "step": 46
    },
    {
      "epoch": 0.20982142857142858,
      "grad_norm": 2.3558027744293213,
      "learning_rate": 9.98481897805642e-06,
      "loss": 0.5271,
      "step": 47
    },
    {
      "epoch": 0.21428571428571427,
      "grad_norm": 2.3652427196502686,
      "learning_rate": 9.983885477852628e-06,
      "loss": 0.5966,
      "step": 48
    },
    {
      "epoch": 0.21875,
      "grad_norm": 2.551260471343994,
      "learning_rate": 9.982924172786847e-06,
      "loss": 0.5586,
      "step": 49
    },
    {
      "epoch": 0.22321428571428573,
      "grad_norm": 2.4399120807647705,
      "learning_rate": 9.981935068222687e-06,
      "loss": 0.5133,
      "step": 50
    },
    {
      "epoch": 0.22767857142857142,
      "grad_norm": 2.097429037094116,
      "learning_rate": 9.980918169678872e-06,
      "loss": 0.439,
      "step": 51
    },
    {
      "epoch": 0.23214285714285715,
      "grad_norm": 3.4427826404571533,
      "learning_rate": 9.979873482829199e-06,
      "loss": 0.5122,
      "step": 52
    },
    {
      "epoch": 0.23660714285714285,
      "grad_norm": 2.506448268890381,
      "learning_rate": 9.978801013502511e-06,
      "loss": 0.5187,
      "step": 53
    },
    {
      "epoch": 0.24107142857142858,
      "grad_norm": 2.153078556060791,
      "learning_rate": 9.977700767682665e-06,
      "loss": 0.4699,
      "step": 54
    },
    {
      "epoch": 0.24553571428571427,
      "grad_norm": 2.5463476181030273,
      "learning_rate": 9.976572751508497e-06,
      "loss": 0.5425,
      "step": 55
    },
    {
      "epoch": 0.25,
      "grad_norm": 2.3144137859344482,
      "learning_rate": 9.975416971273787e-06,
      "loss": 0.4973,
      "step": 56
    },
    {
      "epoch": 0.2544642857142857,
      "grad_norm": 2.373551607131958,
      "learning_rate": 9.974233433427222e-06,
      "loss": 0.4538,
      "step": 57
    },
    {
      "epoch": 0.25892857142857145,
      "grad_norm": 1.888371467590332,
      "learning_rate": 9.97302214457237e-06,
      "loss": 0.4525,
      "step": 58
    },
    {
      "epoch": 0.26339285714285715,
      "grad_norm": 2.0926015377044678,
      "learning_rate": 9.971783111467635e-06,
      "loss": 0.4793,
      "step": 59
    },
    {
      "epoch": 0.26785714285714285,
      "grad_norm": 2.090359926223755,
      "learning_rate": 9.970516341026211e-06,
      "loss": 0.4818,
      "step": 60
    },
    {
      "epoch": 0.27232142857142855,
      "grad_norm": 2.268264055252075,
      "learning_rate": 9.969221840316066e-06,
      "loss": 0.5479,
      "step": 61
    },
    {
      "epoch": 0.2767857142857143,
      "grad_norm": 2.044856309890747,
      "learning_rate": 9.967899616559879e-06,
      "loss": 0.4726,
      "step": 62
    },
    {
      "epoch": 0.28125,
      "grad_norm": 2.361753225326538,
      "learning_rate": 9.966549677135015e-06,
      "loss": 0.4901,
      "step": 63
    },
    {
      "epoch": 0.2857142857142857,
      "grad_norm": 2.3921494483947754,
      "learning_rate": 9.965172029573479e-06,
      "loss": 0.4856,
      "step": 64
    },
    {
      "epoch": 0.29017857142857145,
      "grad_norm": 2.134155750274658,
      "learning_rate": 9.96376668156187e-06,
      "loss": 0.4621,
      "step": 65
    },
    {
      "epoch": 0.29464285714285715,
      "grad_norm": 2.1714282035827637,
      "learning_rate": 9.962333640941349e-06,
      "loss": 0.5009,
      "step": 66
    },
    {
      "epoch": 0.29910714285714285,
      "grad_norm": 2.107659339904785,
      "learning_rate": 9.960872915707582e-06,
      "loss": 0.5125,
      "step": 67
    },
    {
      "epoch": 0.30357142857142855,
      "grad_norm": 1.874833583831787,
      "learning_rate": 9.959384514010703e-06,
      "loss": 0.4568,
      "step": 68
    },
    {
      "epoch": 0.3080357142857143,
      "grad_norm": 2.18188214302063,
      "learning_rate": 9.95786844415527e-06,
      "loss": 0.5336,
      "step": 69
    },
    {
      "epoch": 0.3125,
      "grad_norm": 1.9995635747909546,
      "learning_rate": 9.956324714600212e-06,
      "loss": 0.507,
      "step": 70
    },
    {
      "epoch": 0.3169642857142857,
      "grad_norm": 2.198190212249756,
      "learning_rate": 9.95475333395879e-06,
      "loss": 0.4801,
      "step": 71
    },
    {
      "epoch": 0.32142857142857145,
      "grad_norm": 2.370695114135742,
      "learning_rate": 9.95315431099854e-06,
      "loss": 0.4944,
      "step": 72
    },
    {
      "epoch": 0.32589285714285715,
      "grad_norm": 2.126605272293091,
      "learning_rate": 9.951527654641231e-06,
      "loss": 0.4874,
      "step": 73
    },
    {
      "epoch": 0.33035714285714285,
      "grad_norm": 2.093745231628418,
      "learning_rate": 9.949873373962814e-06,
      "loss": 0.5515,
      "step": 74
    },
    {
      "epoch": 0.33482142857142855,
      "grad_norm": 1.9955790042877197,
      "learning_rate": 9.948191478193365e-06,
      "loss": 0.4524,
      "step": 75
    },
    {
      "epoch": 0.3392857142857143,
      "grad_norm": 1.8829865455627441,
      "learning_rate": 9.946481976717046e-06,
      "loss": 0.4387,
      "step": 76
    },
    {
      "epoch": 0.34375,
      "grad_norm": 1.8707062005996704,
      "learning_rate": 9.944744879072043e-06,
      "loss": 0.4467,
      "step": 77
    },
    {
      "epoch": 0.3482142857142857,
      "grad_norm": 2.0644657611846924,
      "learning_rate": 9.942980194950511e-06,
      "loss": 0.5096,
      "step": 78
    },
    {
      "epoch": 0.35267857142857145,
      "grad_norm": 2.6129536628723145,
      "learning_rate": 9.941187934198528e-06,
      "loss": 0.5458,
      "step": 79
    },
    {
      "epoch": 0.35714285714285715,
      "grad_norm": 2.0972604751586914,
      "learning_rate": 9.939368106816038e-06,
      "loss": 0.4918,
      "step": 80
    },
    {
      "epoch": 0.36160714285714285,
      "grad_norm": 2.2218029499053955,
      "learning_rate": 9.937520722956789e-06,
      "loss": 0.5811,
      "step": 81
    },
    {
      "epoch": 0.36607142857142855,
      "grad_norm": 2.017685890197754,
      "learning_rate": 9.93564579292828e-06,
      "loss": 0.4626,
      "step": 82
    },
    {
      "epoch": 0.3705357142857143,
      "grad_norm": 2.521484613418579,
      "learning_rate": 9.933743327191711e-06,
      "loss": 0.4818,
      "step": 83
    },
    {
      "epoch": 0.375,
      "grad_norm": 2.1667404174804688,
      "learning_rate": 9.93181333636191e-06,
      "loss": 0.5324,
      "step": 84
    },
    {
      "epoch": 0.3794642857142857,
      "grad_norm": 2.0817134380340576,
      "learning_rate": 9.929855831207288e-06,
      "loss": 0.4595,
      "step": 85
    },
    {
      "epoch": 0.38392857142857145,
      "grad_norm": 1.9740232229232788,
      "learning_rate": 9.92787082264977e-06,
      "loss": 0.4228,
      "step": 86
    },
    {
      "epoch": 0.38839285714285715,
      "grad_norm": 1.985465407371521,
      "learning_rate": 9.925858321764733e-06,
      "loss": 0.4971,
      "step": 87
    },
    {
      "epoch": 0.39285714285714285,
      "grad_norm": 2.02755069732666,
      "learning_rate": 9.923818339780954e-06,
      "loss": 0.5322,
      "step": 88
    },
    {
      "epoch": 0.39732142857142855,
      "grad_norm": 2.2033133506774902,
      "learning_rate": 9.921750888080534e-06,
      "loss": 0.4646,
      "step": 89
    },
    {
      "epoch": 0.4017857142857143,
      "grad_norm": 2.1710963249206543,
      "learning_rate": 9.91965597819885e-06,
      "loss": 0.4665,
      "step": 90
    },
    {
      "epoch": 0.40625,
      "grad_norm": 1.907015085220337,
      "learning_rate": 9.917533621824476e-06,
      "loss": 0.408,
      "step": 91
    },
    {
      "epoch": 0.4107142857142857,
      "grad_norm": 1.8381679058074951,
      "learning_rate": 9.915383830799129e-06,
      "loss": 0.4057,
      "step": 92
    },
    {
      "epoch": 0.41517857142857145,
      "grad_norm": 2.105541229248047,
      "learning_rate": 9.91320661711759e-06,
      "loss": 0.5089,
      "step": 93
    },
    {
      "epoch": 0.41964285714285715,
      "grad_norm": 2.172125816345215,
      "learning_rate": 9.911001992927655e-06,
      "loss": 0.4856,
      "step": 94
    },
    {
      "epoch": 0.42410714285714285,
      "grad_norm": 2.0441055297851562,
      "learning_rate": 9.908769970530049e-06,
      "loss": 0.505,
      "step": 95
    },
    {
      "epoch": 0.42857142857142855,
      "grad_norm": 1.9920746088027954,
      "learning_rate": 9.90651056237837e-06,
      "loss": 0.4962,
      "step": 96
    },
    {
      "epoch": 0.4330357142857143,
      "grad_norm": 2.0216598510742188,
      "learning_rate": 9.904223781079017e-06,
      "loss": 0.5041,
      "step": 97
    },
    {
      "epoch": 0.4375,
      "grad_norm": 2.212711811065674,
      "learning_rate": 9.901909639391111e-06,
      "loss": 0.5078,
      "step": 98
    },
    {
      "epoch": 0.4419642857142857,
      "grad_norm": 2.3241405487060547,
      "learning_rate": 9.899568150226435e-06,
      "loss": 0.4316,
      "step": 99
    },
    {
      "epoch": 0.44642857142857145,
      "grad_norm": 2.092745542526245,
      "learning_rate": 9.897199326649362e-06,
      "loss": 0.4653,
      "step": 100
    },
    {
      "epoch": 0.45089285714285715,
      "grad_norm": 1.8443938493728638,
      "learning_rate": 9.894803181876765e-06,
      "loss": 0.4527,
      "step": 101
    },
    {
      "epoch": 0.45535714285714285,
      "grad_norm": 1.827662467956543,
      "learning_rate": 9.892379729277972e-06,
      "loss": 0.455,
      "step": 102
    },
    {
      "epoch": 0.45982142857142855,
      "grad_norm": 1.8256633281707764,
      "learning_rate": 9.889928982374663e-06,
      "loss": 0.369,
      "step": 103
    },
    {
      "epoch": 0.4642857142857143,
      "grad_norm": 2.305311918258667,
      "learning_rate": 9.887450954840812e-06,
      "loss": 0.4713,
      "step": 104
    },
    {
      "epoch": 0.46875,
      "grad_norm": 2.3441269397735596,
      "learning_rate": 9.884945660502607e-06,
      "loss": 0.4571,
      "step": 105
    },
    {
      "epoch": 0.4732142857142857,
      "grad_norm": 1.9434335231781006,
      "learning_rate": 9.882413113338364e-06,
      "loss": 0.4814,
      "step": 106
    },
    {
      "epoch": 0.47767857142857145,
      "grad_norm": 2.1630849838256836,
      "learning_rate": 9.879853327478465e-06,
      "loss": 0.4856,
      "step": 107
    },
    {
      "epoch": 0.48214285714285715,
      "grad_norm": 2.32119083404541,
      "learning_rate": 9.877266317205268e-06,
      "loss": 0.4614,
      "step": 108
    },
    {
      "epoch": 0.48660714285714285,
      "grad_norm": 2.141681671142578,
      "learning_rate": 9.874652096953028e-06,
      "loss": 0.4633,
      "step": 109
    },
    {
      "epoch": 0.49107142857142855,
      "grad_norm": 2.1267635822296143,
      "learning_rate": 9.872010681307821e-06,
      "loss": 0.5275,
      "step": 110
    },
    {
      "epoch": 0.4955357142857143,
      "grad_norm": 2.1307332515716553,
      "learning_rate": 9.869342085007458e-06,
      "loss": 0.4224,
      "step": 111
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.9681659936904907,
      "learning_rate": 9.866646322941405e-06,
      "loss": 0.4645,
      "step": 112
    },
    {
      "epoch": 0.5044642857142857,
      "grad_norm": 1.8705610036849976,
      "learning_rate": 9.863923410150704e-06,
      "loss": 0.4672,
      "step": 113
    },
    {
      "epoch": 0.5089285714285714,
      "grad_norm": 2.2226390838623047,
      "learning_rate": 9.861173361827876e-06,
      "loss": 0.4418,
      "step": 114
    },
    {
      "epoch": 0.5133928571428571,
      "grad_norm": 1.979615569114685,
      "learning_rate": 9.858396193316853e-06,
      "loss": 0.4622,
      "step": 115
    },
    {
      "epoch": 0.5178571428571429,
      "grad_norm": 2.0155093669891357,
      "learning_rate": 9.855591920112883e-06,
      "loss": 0.4585,
      "step": 116
    },
    {
      "epoch": 0.5223214285714286,
      "grad_norm": 1.7425578832626343,
      "learning_rate": 9.85276055786244e-06,
      "loss": 0.4938,
      "step": 117
    },
    {
      "epoch": 0.5267857142857143,
      "grad_norm": 2.0626533031463623,
      "learning_rate": 9.849902122363148e-06,
      "loss": 0.458,
      "step": 118
    },
    {
      "epoch": 0.53125,
      "grad_norm": 1.8312275409698486,
      "learning_rate": 9.847016629563683e-06,
      "loss": 0.4615,
      "step": 119
    },
    {
      "epoch": 0.5357142857142857,
      "grad_norm": 2.0553085803985596,
      "learning_rate": 9.844104095563689e-06,
      "loss": 0.5286,
      "step": 120
    },
    {
      "epoch": 0.5401785714285714,
      "grad_norm": 1.95139741897583,
      "learning_rate": 9.841164536613685e-06,
      "loss": 0.4863,
      "step": 121
    },
    {
      "epoch": 0.5446428571428571,
      "grad_norm": 1.9694424867630005,
      "learning_rate": 9.83819796911498e-06,
      "loss": 0.5279,
      "step": 122
    },
    {
      "epoch": 0.5491071428571429,
      "grad_norm": 1.9648712873458862,
      "learning_rate": 9.83520440961957e-06,
      "loss": 0.3807,
      "step": 123
    },
    {
      "epoch": 0.5535714285714286,
      "grad_norm": 1.8197687864303589,
      "learning_rate": 9.83218387483006e-06,
      "loss": 0.4943,
      "step": 124
    },
    {
      "epoch": 0.5580357142857143,
      "grad_norm": 1.864090919494629,
      "learning_rate": 9.829136381599563e-06,
      "loss": 0.4804,
      "step": 125
    },
    {
      "epoch": 0.5625,
      "grad_norm": 1.827278971672058,
      "learning_rate": 9.826061946931605e-06,
      "loss": 0.4173,
      "step": 126
    },
    {
      "epoch": 0.5669642857142857,
      "grad_norm": 1.8508820533752441,
      "learning_rate": 9.822960587980034e-06,
      "loss": 0.4831,
      "step": 127
    },
    {
      "epoch": 0.5714285714285714,
      "grad_norm": 1.801448941230774,
      "learning_rate": 9.81983232204892e-06,
      "loss": 0.4823,
      "step": 128
    },
    {
      "epoch": 0.5758928571428571,
      "grad_norm": 1.9353251457214355,
      "learning_rate": 9.816677166592462e-06,
      "loss": 0.4074,
      "step": 129
    },
    {
      "epoch": 0.5803571428571429,
      "grad_norm": 1.9287128448486328,
      "learning_rate": 9.81349513921489e-06,
      "loss": 0.4495,
      "step": 130
    },
    {
      "epoch": 0.5848214285714286,
      "grad_norm": 1.5913102626800537,
      "learning_rate": 9.810286257670365e-06,
      "loss": 0.4161,
      "step": 131
    },
    {
      "epoch": 0.5892857142857143,
      "grad_norm": 1.9423521757125854,
      "learning_rate": 9.807050539862884e-06,
      "loss": 0.4254,
      "step": 132
    },
    {
      "epoch": 0.59375,
      "grad_norm": 1.7807602882385254,
      "learning_rate": 9.803788003846175e-06,
      "loss": 0.4541,
      "step": 133
    },
    {
      "epoch": 0.5982142857142857,
      "grad_norm": 1.9615179300308228,
      "learning_rate": 9.800498667823595e-06,
      "loss": 0.4375,
      "step": 134
    },
    {
      "epoch": 0.6026785714285714,
      "grad_norm": 1.7444758415222168,
      "learning_rate": 9.797182550148039e-06,
      "loss": 0.4568,
      "step": 135
    },
    {
      "epoch": 0.6071428571428571,
      "grad_norm": 1.9967138767242432,
      "learning_rate": 9.793839669321828e-06,
      "loss": 0.4152,
      "step": 136
    },
    {
      "epoch": 0.6116071428571429,
      "grad_norm": 1.8935024738311768,
      "learning_rate": 9.790470043996604e-06,
      "loss": 0.4807,
      "step": 137
    },
    {
      "epoch": 0.6160714285714286,
      "grad_norm": 1.8592950105667114,
      "learning_rate": 9.78707369297324e-06,
      "loss": 0.4849,
      "step": 138
    },
    {
      "epoch": 0.6205357142857143,
      "grad_norm": 1.8688433170318604,
      "learning_rate": 9.783650635201714e-06,
      "loss": 0.5193,
      "step": 139
    },
    {
      "epoch": 0.625,
      "grad_norm": 1.806913137435913,
      "learning_rate": 9.780200889781021e-06,
      "loss": 0.4224,
      "step": 140
    },
    {
      "epoch": 0.6294642857142857,
      "grad_norm": 1.8969337940216064,
      "learning_rate": 9.776724475959061e-06,
      "loss": 0.4986,
      "step": 141
    },
    {
      "epoch": 0.6339285714285714,
      "grad_norm": 1.5755013227462769,
      "learning_rate": 9.773221413132525e-06,
      "loss": 0.4343,
      "step": 142
    },
    {
      "epoch": 0.6383928571428571,
      "grad_norm": 1.70070481300354,
      "learning_rate": 9.769691720846801e-06,
      "loss": 0.4163,
      "step": 143
    },
    {
      "epoch": 0.6428571428571429,
      "grad_norm": 1.9572802782058716,
      "learning_rate": 9.766135418795848e-06,
      "loss": 0.4618,
      "step": 144
    },
    {
      "epoch": 0.6473214285714286,
      "grad_norm": 1.9200794696807861,
      "learning_rate": 9.762552526822098e-06,
      "loss": 0.4258,
      "step": 145
    },
    {
      "epoch": 0.6517857142857143,
      "grad_norm": 1.9915560483932495,
      "learning_rate": 9.758943064916342e-06,
      "loss": 0.4185,
      "step": 146
    },
    {
      "epoch": 0.65625,
      "grad_norm": 1.6791561841964722,
      "learning_rate": 9.755307053217622e-06,
      "loss": 0.4074,
      "step": 147
    },
    {
      "epoch": 0.6607142857142857,
      "grad_norm": 1.984868049621582,
      "learning_rate": 9.751644512013106e-06,
      "loss": 0.461,
      "step": 148
    },
    {
      "epoch": 0.6651785714285714,
      "grad_norm": 1.925027847290039,
      "learning_rate": 9.74795546173799e-06,
      "loss": 0.4601,
      "step": 149
    },
    {
      "epoch": 0.6696428571428571,
      "grad_norm": 2.0234742164611816,
      "learning_rate": 9.744239922975377e-06,
      "loss": 0.4874,
      "step": 150
    },
    {
      "epoch": 0.6741071428571429,
      "grad_norm": 1.8171749114990234,
      "learning_rate": 9.740497916456163e-06,
      "loss": 0.4806,
      "step": 151
    },
    {
      "epoch": 0.6785714285714286,
      "grad_norm": 1.8766131401062012,
      "learning_rate": 9.736729463058921e-06,
      "loss": 0.5144,
      "step": 152
    },
    {
      "epoch": 0.6830357142857143,
      "grad_norm": 1.7700015306472778,
      "learning_rate": 9.732934583809782e-06,
      "loss": 0.4958,
      "step": 153
    },
    {
      "epoch": 0.6875,
      "grad_norm": 1.7756420373916626,
      "learning_rate": 9.729113299882324e-06,
      "loss": 0.4685,
      "step": 154
    },
    {
      "epoch": 0.6919642857142857,
      "grad_norm": 1.7557669878005981,
      "learning_rate": 9.725265632597448e-06,
      "loss": 0.4736,
      "step": 155
    },
    {
      "epoch": 0.6964285714285714,
      "grad_norm": 1.7786169052124023,
      "learning_rate": 9.721391603423263e-06,
      "loss": 0.4582,
      "step": 156
    },
    {
      "epoch": 0.7008928571428571,
      "grad_norm": 1.7477728128433228,
      "learning_rate": 9.717491233974962e-06,
      "loss": 0.4399,
      "step": 157
    },
    {
      "epoch": 0.7053571428571429,
      "grad_norm": 1.8034288883209229,
      "learning_rate": 9.713564546014707e-06,
      "loss": 0.4933,
      "step": 158
    },
    {
      "epoch": 0.7098214285714286,
      "grad_norm": 1.7433615922927856,
      "learning_rate": 9.7096115614515e-06,
      "loss": 0.4367,
      "step": 159
    },
    {
      "epoch": 0.7142857142857143,
      "grad_norm": 1.9310708045959473,
      "learning_rate": 9.705632302341073e-06,
      "loss": 0.4674,
      "step": 160
    },
    {
      "epoch": 0.71875,
      "grad_norm": 1.709022879600525,
      "learning_rate": 9.701626790885749e-06,
      "loss": 0.4361,
      "step": 161
    },
    {
      "epoch": 0.7232142857142857,
      "grad_norm": 1.7623347043991089,
      "learning_rate": 9.69759504943433e-06,
      "loss": 0.4037,
      "step": 162
    },
    {
      "epoch": 0.7276785714285714,
      "grad_norm": 1.6316179037094116,
      "learning_rate": 9.69353710048197e-06,
      "loss": 0.4707,
      "step": 163
    },
    {
      "epoch": 0.7321428571428571,
      "grad_norm": 1.966737151145935,
      "learning_rate": 9.68945296667004e-06,
      "loss": 0.4493,
      "step": 164
    },
    {
      "epoch": 0.7366071428571429,
      "grad_norm": 1.7973566055297852,
      "learning_rate": 9.685342670786025e-06,
      "loss": 0.5024,
      "step": 165
    },
    {
      "epoch": 0.7410714285714286,
      "grad_norm": 2.026155710220337,
      "learning_rate": 9.681206235763367e-06,
      "loss": 0.4879,
      "step": 166
    },
    {
      "epoch": 0.7455357142857143,
      "grad_norm": 1.679463267326355,
      "learning_rate": 9.677043684681358e-06,
      "loss": 0.4495,
      "step": 167
    },
    {
      "epoch": 0.75,
      "grad_norm": 1.7095201015472412,
      "learning_rate": 9.672855040765006e-06,
      "loss": 0.3991,
      "step": 168
    },
    {
      "epoch": 0.7544642857142857,
      "grad_norm": 1.7846328020095825,
      "learning_rate": 9.668640327384899e-06,
      "loss": 0.4653,
      "step": 169
    },
    {
      "epoch": 0.7589285714285714,
      "grad_norm": 1.9439399242401123,
      "learning_rate": 9.664399568057087e-06,
      "loss": 0.4811,
      "step": 170
    },
    {
      "epoch": 0.7633928571428571,
      "grad_norm": 1.8142772912979126,
      "learning_rate": 9.660132786442937e-06,
      "loss": 0.4638,
      "step": 171
    },
    {
      "epoch": 0.7678571428571429,
      "grad_norm": 1.8977925777435303,
      "learning_rate": 9.655840006349014e-06,
      "loss": 0.4698,
      "step": 172
    },
    {
      "epoch": 0.7723214285714286,
      "grad_norm": 1.624567985534668,
      "learning_rate": 9.651521251726936e-06,
      "loss": 0.4651,
      "step": 173
    },
    {
      "epoch": 0.7767857142857143,
      "grad_norm": 1.6974815130233765,
      "learning_rate": 9.64717654667325e-06,
      "loss": 0.5096,
      "step": 174
    },
    {
      "epoch": 0.78125,
      "grad_norm": 1.6671240329742432,
      "learning_rate": 9.642805915429291e-06,
      "loss": 0.4255,
      "step": 175
    },
    {
      "epoch": 0.7857142857142857,
      "grad_norm": 1.703273057937622,
      "learning_rate": 9.638409382381052e-06,
      "loss": 0.4436,
      "step": 176
    },
    {
      "epoch": 0.7901785714285714,
      "grad_norm": 1.924634575843811,
      "learning_rate": 9.633986972059047e-06,
      "loss": 0.4679,
      "step": 177
    },
    {
      "epoch": 0.7946428571428571,
      "grad_norm": 1.6996525526046753,
      "learning_rate": 9.629538709138166e-06,
      "loss": 0.4836,
      "step": 178
    },
    {
      "epoch": 0.7991071428571429,
      "grad_norm": 1.6877120733261108,
      "learning_rate": 9.625064618437549e-06,
      "loss": 0.4473,
      "step": 179
    },
    {
      "epoch": 0.8035714285714286,
      "grad_norm": 1.6341931819915771,
      "learning_rate": 9.620564724920443e-06,
      "loss": 0.4279,
      "step": 180
    },
    {
      "epoch": 0.8080357142857143,
      "grad_norm": 1.5455591678619385,
      "learning_rate": 9.616039053694058e-06,
      "loss": 0.4124,
      "step": 181
    },
    {
      "epoch": 0.8125,
      "grad_norm": 1.6947259902954102,
      "learning_rate": 9.611487630009436e-06,
      "loss": 0.4061,
      "step": 182
    },
    {
      "epoch": 0.8169642857142857,
      "grad_norm": 1.7641195058822632,
      "learning_rate": 9.606910479261301e-06,
      "loss": 0.4839,
      "step": 183
    },
    {
      "epoch": 0.8214285714285714,
      "grad_norm": 1.513904333114624,
      "learning_rate": 9.602307626987925e-06,
      "loss": 0.403,
      "step": 184
    },
    {
      "epoch": 0.8258928571428571,
      "grad_norm": 1.791146993637085,
      "learning_rate": 9.597679098870978e-06,
      "loss": 0.4782,
      "step": 185
    },
    {
      "epoch": 0.8303571428571429,
      "grad_norm": 1.7051180601119995,
      "learning_rate": 9.593024920735393e-06,
      "loss": 0.4754,
      "step": 186
    },
    {
      "epoch": 0.8348214285714286,
      "grad_norm": 1.749476671218872,
      "learning_rate": 9.588345118549214e-06,
      "loss": 0.435,
      "step": 187
    },
    {
      "epoch": 0.8392857142857143,
      "grad_norm": 1.661120057106018,
      "learning_rate": 9.583639718423457e-06,
      "loss": 0.4536,
      "step": 188
    },
    {
      "epoch": 0.84375,
      "grad_norm": 1.4928163290023804,
      "learning_rate": 9.57890874661196e-06,
      "loss": 0.375,
      "step": 189
    },
    {
      "epoch": 0.8482142857142857,
      "grad_norm": 1.6453092098236084,
      "learning_rate": 9.57415222951124e-06,
      "loss": 0.445,
      "step": 190
    },
    {
      "epoch": 0.8526785714285714,
      "grad_norm": 1.6496193408966064,
      "learning_rate": 9.569370193660348e-06,
      "loss": 0.4098,
      "step": 191
    },
    {
      "epoch": 0.8571428571428571,
      "grad_norm": 1.779951810836792,
      "learning_rate": 9.564562665740708e-06,
      "loss": 0.5191,
      "step": 192
    },
    {
      "epoch": 0.8616071428571429,
      "grad_norm": 1.5148218870162964,
      "learning_rate": 9.559729672575985e-06,
      "loss": 0.4001,
      "step": 193
    },
    {
      "epoch": 0.8660714285714286,
      "grad_norm": 1.5799369812011719,
      "learning_rate": 9.554871241131923e-06,
      "loss": 0.4807,
      "step": 194
    },
    {
      "epoch": 0.8705357142857143,
      "grad_norm": 1.6424708366394043,
      "learning_rate": 9.549987398516206e-06,
      "loss": 0.4851,
      "step": 195
    },
    {
      "epoch": 0.875,
      "grad_norm": 1.766772985458374,
      "learning_rate": 9.54507817197829e-06,
      "loss": 0.4295,
      "step": 196
    },
    {
      "epoch": 0.8794642857142857,
      "grad_norm": 1.6645870208740234,
      "learning_rate": 9.540143588909268e-06,
      "loss": 0.426,
      "step": 197
    },
    {
      "epoch": 0.8839285714285714,
      "grad_norm": 1.8015894889831543,
      "learning_rate": 9.535183676841709e-06,
      "loss": 0.4602,
      "step": 198
    },
    {
      "epoch": 0.8883928571428571,
      "grad_norm": 1.8687186241149902,
      "learning_rate": 9.530198463449507e-06,
      "loss": 0.4674,
      "step": 199
    },
    {
      "epoch": 0.8928571428571429,
      "grad_norm": 1.8257641792297363,
      "learning_rate": 9.525187976547718e-06,
      "loss": 0.4584,
      "step": 200
    },
    {
      "epoch": 0.8928571428571429,
      "eval_loss": 0.5852221250534058,
      "eval_runtime": 5.367,
      "eval_samples_per_second": 11.18,
      "eval_steps_per_second": 0.745,
      "step": 200
    },
    {
      "epoch": 0.8973214285714286,
      "grad_norm": 1.5470904111862183,
      "learning_rate": 9.520152244092421e-06,
      "loss": 0.3601,
      "step": 201
    },
    {
      "epoch": 0.9017857142857143,
      "grad_norm": 1.608096957206726,
      "learning_rate": 9.515091294180546e-06,
      "loss": 0.3992,
      "step": 202
    },
    {
      "epoch": 0.90625,
      "grad_norm": 1.8043137788772583,
      "learning_rate": 9.510005155049729e-06,
      "loss": 0.4782,
      "step": 203
    },
    {
      "epoch": 0.9107142857142857,
      "grad_norm": 1.6916441917419434,
      "learning_rate": 9.504893855078144e-06,
      "loss": 0.3765,
      "step": 204
    },
    {
      "epoch": 0.9151785714285714,
      "grad_norm": 1.5904515981674194,
      "learning_rate": 9.499757422784358e-06,
      "loss": 0.4424,
      "step": 205
    },
    {
      "epoch": 0.9196428571428571,
      "grad_norm": 1.6705337762832642,
      "learning_rate": 9.494595886827157e-06,
      "loss": 0.4984,
      "step": 206
    },
    {
      "epoch": 0.9241071428571429,
      "grad_norm": 1.657344937324524,
      "learning_rate": 9.489409276005393e-06,
      "loss": 0.4511,
      "step": 207
    },
    {
      "epoch": 0.9285714285714286,
      "grad_norm": 1.6433309316635132,
      "learning_rate": 9.48419761925783e-06,
      "loss": 0.5035,
      "step": 208
    },
    {
      "epoch": 0.9330357142857143,
      "grad_norm": 1.81150484085083,
      "learning_rate": 9.478960945662974e-06,
      "loss": 0.3958,
      "step": 209
    },
    {
      "epoch": 0.9375,
      "grad_norm": 1.6642935276031494,
      "learning_rate": 9.473699284438908e-06,
      "loss": 0.4011,
      "step": 210
    },
    {
      "epoch": 0.9419642857142857,
      "grad_norm": 1.8117976188659668,
      "learning_rate": 9.468412664943137e-06,
      "loss": 0.5123,
      "step": 211
    },
    {
      "epoch": 0.9464285714285714,
      "grad_norm": 1.5370533466339111,
      "learning_rate": 9.463101116672423e-06,
      "loss": 0.4226,
      "step": 212
    },
    {
      "epoch": 0.9508928571428571,
      "grad_norm": 2.033900499343872,
      "learning_rate": 9.457764669262615e-06,
      "loss": 0.4902,
      "step": 213
    },
    {
      "epoch": 0.9553571428571429,
      "grad_norm": 1.672672986984253,
      "learning_rate": 9.452403352488488e-06,
      "loss": 0.442,
      "step": 214
    },
    {
      "epoch": 0.9598214285714286,
      "grad_norm": 1.9266510009765625,
      "learning_rate": 9.447017196263578e-06,
      "loss": 0.4805,
      "step": 215
    },
    {
      "epoch": 0.9642857142857143,
      "grad_norm": 1.629258155822754,
      "learning_rate": 9.441606230640012e-06,
      "loss": 0.444,
      "step": 216
    },
    {
      "epoch": 0.96875,
      "grad_norm": 1.6848729848861694,
      "learning_rate": 9.436170485808338e-06,
      "loss": 0.4444,
      "step": 217
    },
    {
      "epoch": 0.9732142857142857,
      "grad_norm": 1.693801999092102,
      "learning_rate": 9.430709992097364e-06,
      "loss": 0.4913,
      "step": 218
    },
    {
      "epoch": 0.9776785714285714,
      "grad_norm": 1.5817019939422607,
      "learning_rate": 9.425224779973986e-06,
      "loss": 0.4554,
      "step": 219
    },
    {
      "epoch": 0.9821428571428571,
      "grad_norm": 1.7226394414901733,
      "learning_rate": 9.41971488004301e-06,
      "loss": 0.5002,
      "step": 220
    },
    {
      "epoch": 0.9866071428571429,
      "grad_norm": 1.7014926671981812,
      "learning_rate": 9.414180323046991e-06,
      "loss": 0.4582,
      "step": 221
    },
    {
      "epoch": 0.9910714285714286,
      "grad_norm": 1.5330203771591187,
      "learning_rate": 9.408621139866067e-06,
      "loss": 0.3967,
      "step": 222
    },
    {
      "epoch": 0.9955357142857143,
      "grad_norm": 1.6649432182312012,
      "learning_rate": 9.403037361517762e-06,
      "loss": 0.4442,
      "step": 223
    },
    {
      "epoch": 1.0,
      "grad_norm": 1.5434422492980957,
      "learning_rate": 9.397429019156841e-06,
      "loss": 0.4281,
      "step": 224
    },
    {
      "epoch": 1.0044642857142858,
      "grad_norm": 1.4803626537322998,
      "learning_rate": 9.391796144075123e-06,
      "loss": 0.3408,
      "step": 225
    },
    {
      "epoch": 1.0089285714285714,
      "grad_norm": 1.7202813625335693,
      "learning_rate": 9.386138767701306e-06,
      "loss": 0.3212,
      "step": 226
    },
    {
      "epoch": 1.0133928571428572,
      "grad_norm": 1.5383710861206055,
      "learning_rate": 9.380456921600785e-06,
      "loss": 0.3628,
      "step": 227
    },
    {
      "epoch": 1.0178571428571428,
      "grad_norm": 1.598041296005249,
      "learning_rate": 9.374750637475499e-06,
      "loss": 0.3354,
      "step": 228
    },
    {
      "epoch": 1.0223214285714286,
      "grad_norm": 1.483928918838501,
      "learning_rate": 9.36901994716373e-06,
      "loss": 0.3572,
      "step": 229
    },
    {
      "epoch": 1.0267857142857142,
      "grad_norm": 1.4985754489898682,
      "learning_rate": 9.363264882639936e-06,
      "loss": 0.3117,
      "step": 230
    },
    {
      "epoch": 1.03125,
      "grad_norm": 1.6541988849639893,
      "learning_rate": 9.357485476014573e-06,
      "loss": 0.3235,
      "step": 231
    },
    {
      "epoch": 1.0357142857142858,
      "grad_norm": 1.459572672843933,
      "learning_rate": 9.351681759533914e-06,
      "loss": 0.2768,
      "step": 232
    },
    {
      "epoch": 1.0401785714285714,
      "grad_norm": 1.7430468797683716,
      "learning_rate": 9.345853765579865e-06,
      "loss": 0.2995,
      "step": 233
    },
    {
      "epoch": 1.0446428571428572,
      "grad_norm": 1.4982185363769531,
      "learning_rate": 9.340001526669794e-06,
      "loss": 0.2818,
      "step": 234
    },
    {
      "epoch": 1.0491071428571428,
      "grad_norm": 1.6810327768325806,
      "learning_rate": 9.33412507545634e-06,
      "loss": 0.3382,
      "step": 235
    },
    {
      "epoch": 1.0535714285714286,
      "grad_norm": 1.8060071468353271,
      "learning_rate": 9.32822444472724e-06,
      "loss": 0.343,
      "step": 236
    },
    {
      "epoch": 1.0580357142857142,
      "grad_norm": 1.786448359489441,
      "learning_rate": 9.322299667405134e-06,
      "loss": 0.3411,
      "step": 237
    },
    {
      "epoch": 1.0625,
      "grad_norm": 1.776330590248108,
      "learning_rate": 9.31635077654739e-06,
      "loss": 0.3569,
      "step": 238
    },
    {
      "epoch": 1.0669642857142858,
      "grad_norm": 1.5425933599472046,
      "learning_rate": 9.310377805345926e-06,
      "loss": 0.2962,
      "step": 239
    },
    {
      "epoch": 1.0714285714285714,
      "grad_norm": 1.6801799535751343,
      "learning_rate": 9.304380787127003e-06,
      "loss": 0.3163,
      "step": 240
    },
    {
      "epoch": 1.0758928571428572,
      "grad_norm": 1.6421353816986084,
      "learning_rate": 9.298359755351065e-06,
      "loss": 0.3159,
      "step": 241
    },
    {
      "epoch": 1.0803571428571428,
      "grad_norm": 1.7834713459014893,
      "learning_rate": 9.29231474361253e-06,
      "loss": 0.3795,
      "step": 242
    },
    {
      "epoch": 1.0848214285714286,
      "grad_norm": 1.6435611248016357,
      "learning_rate": 9.28624578563962e-06,
      "loss": 0.3161,
      "step": 243
    },
    {
      "epoch": 1.0892857142857142,
      "grad_norm": 1.579264760017395,
      "learning_rate": 9.280152915294162e-06,
      "loss": 0.3916,
      "step": 244
    },
    {
      "epoch": 1.09375,
      "grad_norm": 1.5122257471084595,
      "learning_rate": 9.274036166571402e-06,
      "loss": 0.3642,
      "step": 245
    },
    {
      "epoch": 1.0982142857142858,
      "grad_norm": 1.4102823734283447,
      "learning_rate": 9.267895573599819e-06,
      "loss": 0.3147,
      "step": 246
    },
    {
      "epoch": 1.1026785714285714,
      "grad_norm": 1.5816137790679932,
      "learning_rate": 9.261731170640923e-06,
      "loss": 0.3346,
      "step": 247
    },
    {
      "epoch": 1.1071428571428572,
      "grad_norm": 1.426952838897705,
      "learning_rate": 9.255542992089086e-06,
      "loss": 0.3033,
      "step": 248
    },
    {
      "epoch": 1.1116071428571428,
      "grad_norm": 1.437669038772583,
      "learning_rate": 9.24933107247132e-06,
      "loss": 0.3195,
      "step": 249
    },
    {
      "epoch": 1.1160714285714286,
      "grad_norm": 1.5356261730194092,
      "learning_rate": 9.243095446447113e-06,
      "loss": 0.306,
      "step": 250
    },
    {
      "epoch": 1.1205357142857142,
      "grad_norm": 1.4559781551361084,
      "learning_rate": 9.23683614880822e-06,
      "loss": 0.3088,
      "step": 251
    },
    {
      "epoch": 1.125,
      "grad_norm": 1.5992493629455566,
      "learning_rate": 9.230553214478469e-06,
      "loss": 0.3614,
      "step": 252
    },
    {
      "epoch": 1.1294642857142858,
      "grad_norm": 1.5381258726119995,
      "learning_rate": 9.224246678513569e-06,
      "loss": 0.322,
      "step": 253
    },
    {
      "epoch": 1.1339285714285714,
      "grad_norm": 1.6599974632263184,
      "learning_rate": 9.217916576100922e-06,
      "loss": 0.3098,
      "step": 254
    },
    {
      "epoch": 1.1383928571428572,
      "grad_norm": 1.5781593322753906,
      "learning_rate": 9.211562942559408e-06,
      "loss": 0.3397,
      "step": 255
    },
    {
      "epoch": 1.1428571428571428,
      "grad_norm": 1.5014089345932007,
      "learning_rate": 9.20518581333921e-06,
      "loss": 0.3195,
      "step": 256
    },
    {
      "epoch": 1.1473214285714286,
      "grad_norm": 1.5395718812942505,
      "learning_rate": 9.1987852240216e-06,
      "loss": 0.3319,
      "step": 257
    },
    {
      "epoch": 1.1517857142857142,
      "grad_norm": 1.4616154432296753,
      "learning_rate": 9.192361210318745e-06,
      "loss": 0.3141,
      "step": 258
    },
    {
      "epoch": 1.15625,
      "grad_norm": 1.5390483140945435,
      "learning_rate": 9.185913808073513e-06,
      "loss": 0.3062,
      "step": 259
    },
    {
      "epoch": 1.1607142857142858,
      "grad_norm": 1.6422654390335083,
      "learning_rate": 9.179443053259263e-06,
      "loss": 0.3551,
      "step": 260
    },
    {
      "epoch": 1.1651785714285714,
      "grad_norm": 1.5900368690490723,
      "learning_rate": 9.172948981979654e-06,
      "loss": 0.3655,
      "step": 261
    },
    {
      "epoch": 1.1696428571428572,
      "grad_norm": 1.582152009010315,
      "learning_rate": 9.166431630468438e-06,
      "loss": 0.3418,
      "step": 262
    },
    {
      "epoch": 1.1741071428571428,
      "grad_norm": 1.4746930599212646,
      "learning_rate": 9.159891035089262e-06,
      "loss": 0.2786,
      "step": 263
    },
    {
      "epoch": 1.1785714285714286,
      "grad_norm": 1.5723954439163208,
      "learning_rate": 9.153327232335455e-06,
      "loss": 0.3347,
      "step": 264
    },
    {
      "epoch": 1.1830357142857142,
      "grad_norm": 1.374165415763855,
      "learning_rate": 9.146740258829844e-06,
      "loss": 0.2811,
      "step": 265
    },
    {
      "epoch": 1.1875,
      "grad_norm": 1.6520895957946777,
      "learning_rate": 9.140130151324526e-06,
      "loss": 0.3786,
      "step": 266
    },
    {
      "epoch": 1.1919642857142858,
      "grad_norm": 1.3879092931747437,
      "learning_rate": 9.13349694670068e-06,
      "loss": 0.3063,
      "step": 267
    },
    {
      "epoch": 1.1964285714285714,
      "grad_norm": 1.3920384645462036,
      "learning_rate": 9.126840681968357e-06,
      "loss": 0.3191,
      "step": 268
    },
    {
      "epoch": 1.2008928571428572,
      "grad_norm": 1.6087769269943237,
      "learning_rate": 9.120161394266266e-06,
      "loss": 0.3101,
      "step": 269
    },
    {
      "epoch": 1.2053571428571428,
      "grad_norm": 1.465286135673523,
      "learning_rate": 9.113459120861579e-06,
      "loss": 0.2833,
      "step": 270
    },
    {
      "epoch": 1.2098214285714286,
      "grad_norm": 1.681369423866272,
      "learning_rate": 9.106733899149715e-06,
      "loss": 0.3389,
      "step": 271
    },
    {
      "epoch": 1.2142857142857142,
      "grad_norm": 1.5745911598205566,
      "learning_rate": 9.099985766654132e-06,
      "loss": 0.3352,
      "step": 272
    },
    {
      "epoch": 1.21875,
      "grad_norm": 1.3523564338684082,
      "learning_rate": 9.093214761026121e-06,
      "loss": 0.2964,
      "step": 273
    },
    {
      "epoch": 1.2232142857142858,
      "grad_norm": 1.2804710865020752,
      "learning_rate": 9.08642092004459e-06,
      "loss": 0.2992,
      "step": 274
    },
    {
      "epoch": 1.2276785714285714,
      "grad_norm": 1.521187424659729,
      "learning_rate": 9.079604281615868e-06,
      "loss": 0.3212,
      "step": 275
    },
    {
      "epoch": 1.2321428571428572,
      "grad_norm": 1.3694134950637817,
      "learning_rate": 9.072764883773464e-06,
      "loss": 0.3195,
      "step": 276
    },
    {
      "epoch": 1.2366071428571428,
      "grad_norm": 1.4687600135803223,
      "learning_rate": 9.065902764677897e-06,
      "loss": 0.2878,
      "step": 277
    },
    {
      "epoch": 1.2410714285714286,
      "grad_norm": 1.5896199941635132,
      "learning_rate": 9.059017962616435e-06,
      "loss": 0.3482,
      "step": 278
    },
    {
      "epoch": 1.2455357142857142,
      "grad_norm": 1.6721631288528442,
      "learning_rate": 9.052110516002925e-06,
      "loss": 0.3629,
      "step": 279
    },
    {
      "epoch": 1.25,
      "grad_norm": 1.703574299812317,
      "learning_rate": 9.04518046337755e-06,
      "loss": 0.3618,
      "step": 280
    },
    {
      "epoch": 1.2544642857142856,
      "grad_norm": 1.704490065574646,
      "learning_rate": 9.038227843406628e-06,
      "loss": 0.3397,
      "step": 281
    },
    {
      "epoch": 1.2589285714285714,
      "grad_norm": 1.508264422416687,
      "learning_rate": 9.031252694882386e-06,
      "loss": 0.3405,
      "step": 282
    },
    {
      "epoch": 1.2633928571428572,
      "grad_norm": 1.4581125974655151,
      "learning_rate": 9.024255056722753e-06,
      "loss": 0.3227,
      "step": 283
    },
    {
      "epoch": 1.2678571428571428,
      "grad_norm": 1.4935333728790283,
      "learning_rate": 9.017234967971143e-06,
      "loss": 0.3248,
      "step": 284
    },
    {
      "epoch": 1.2723214285714286,
      "grad_norm": 1.5179234743118286,
      "learning_rate": 9.010192467796228e-06,
      "loss": 0.3257,
      "step": 285
    },
    {
      "epoch": 1.2767857142857144,
      "grad_norm": 1.5567351579666138,
      "learning_rate": 9.003127595491723e-06,
      "loss": 0.3393,
      "step": 286
    },
    {
      "epoch": 1.28125,
      "grad_norm": 1.6843305826187134,
      "learning_rate": 8.996040390476177e-06,
      "loss": 0.365,
      "step": 287
    },
    {
      "epoch": 1.2857142857142856,
      "grad_norm": 1.558060646057129,
      "learning_rate": 8.988930892292738e-06,
      "loss": 0.3617,
      "step": 288
    },
    {
      "epoch": 1.2901785714285714,
      "grad_norm": 1.5220246315002441,
      "learning_rate": 8.981799140608938e-06,
      "loss": 0.3399,
      "step": 289
    },
    {
      "epoch": 1.2946428571428572,
      "grad_norm": 1.5785568952560425,
      "learning_rate": 8.974645175216478e-06,
      "loss": 0.3267,
      "step": 290
    },
    {
      "epoch": 1.2991071428571428,
      "grad_norm": 1.5936439037322998,
      "learning_rate": 8.967469036030996e-06,
      "loss": 0.3337,
      "step": 291
    },
    {
      "epoch": 1.3035714285714286,
      "grad_norm": 1.7652866840362549,
      "learning_rate": 8.960270763091853e-06,
      "loss": 0.3109,
      "step": 292
    },
    {
      "epoch": 1.3080357142857144,
      "grad_norm": 1.3734923601150513,
      "learning_rate": 8.953050396561904e-06,
      "loss": 0.3026,
      "step": 293
    },
    {
      "epoch": 1.3125,
      "grad_norm": 1.570879578590393,
      "learning_rate": 8.94580797672727e-06,
      "loss": 0.3305,
      "step": 294
    },
    {
      "epoch": 1.3169642857142856,
      "grad_norm": 1.5227439403533936,
      "learning_rate": 8.938543543997129e-06,
      "loss": 0.2904,
      "step": 295
    },
    {
      "epoch": 1.3214285714285714,
      "grad_norm": 1.756388783454895,
      "learning_rate": 8.931257138903474e-06,
      "loss": 0.3017,
      "step": 296
    },
    {
      "epoch": 1.3258928571428572,
      "grad_norm": 1.714189052581787,
      "learning_rate": 8.923948802100891e-06,
      "loss": 0.3486,
      "step": 297
    },
    {
      "epoch": 1.3303571428571428,
      "grad_norm": 1.7315165996551514,
      "learning_rate": 8.916618574366338e-06,
      "loss": 0.3542,
      "step": 298
    },
    {
      "epoch": 1.3348214285714286,
      "grad_norm": 1.5031179189682007,
      "learning_rate": 8.909266496598917e-06,
      "loss": 0.3139,
      "step": 299
    },
    {
      "epoch": 1.3392857142857144,
      "grad_norm": 1.3531993627548218,
      "learning_rate": 8.901892609819632e-06,
      "loss": 0.2723,
      "step": 300
    },
    {
      "epoch": 1.34375,
      "grad_norm": 1.5571861267089844,
      "learning_rate": 8.894496955171182e-06,
      "loss": 0.36,
      "step": 301
    },
    {
      "epoch": 1.3482142857142856,
      "grad_norm": 1.662018060684204,
      "learning_rate": 8.887079573917713e-06,
      "loss": 0.3305,
      "step": 302
    },
    {
      "epoch": 1.3526785714285714,
      "grad_norm": 1.696288824081421,
      "learning_rate": 8.879640507444598e-06,
      "loss": 0.3286,
      "step": 303
    },
    {
      "epoch": 1.3571428571428572,
      "grad_norm": 1.5394744873046875,
      "learning_rate": 8.872179797258202e-06,
      "loss": 0.3549,
      "step": 304
    },
    {
      "epoch": 1.3616071428571428,
      "grad_norm": 1.6615936756134033,
      "learning_rate": 8.86469748498565e-06,
      "loss": 0.3198,
      "step": 305
    },
    {
      "epoch": 1.3660714285714286,
      "grad_norm": 1.4944754838943481,
      "learning_rate": 8.8571936123746e-06,
      "loss": 0.3068,
      "step": 306
    },
    {
      "epoch": 1.3705357142857144,
      "grad_norm": 1.4707030057907104,
      "learning_rate": 8.849668221293e-06,
      "loss": 0.3293,
      "step": 307
    },
    {
      "epoch": 1.375,
      "grad_norm": 1.4484243392944336,
      "learning_rate": 8.842121353728867e-06,
      "loss": 0.3217,
      "step": 308
    },
    {
      "epoch": 1.3794642857142856,
      "grad_norm": 1.544541835784912,
      "learning_rate": 8.834553051790044e-06,
      "loss": 0.3514,
      "step": 309
    },
    {
      "epoch": 1.3839285714285714,
      "grad_norm": 1.3725217580795288,
      "learning_rate": 8.826963357703964e-06,
      "loss": 0.2865,
      "step": 310
    },
    {
      "epoch": 1.3883928571428572,
      "grad_norm": 1.586159110069275,
      "learning_rate": 8.819352313817424e-06,
      "loss": 0.3376,
      "step": 311
    },
    {
      "epoch": 1.3928571428571428,
      "grad_norm": 1.4718273878097534,
      "learning_rate": 8.811719962596338e-06,
      "loss": 0.3243,
      "step": 312
    },
    {
      "epoch": 1.3973214285714286,
      "grad_norm": 1.41029953956604,
      "learning_rate": 8.804066346625506e-06,
      "loss": 0.3142,
      "step": 313
    },
    {
      "epoch": 1.4017857142857144,
      "grad_norm": 1.6441102027893066,
      "learning_rate": 8.796391508608372e-06,
      "loss": 0.362,
      "step": 314
    },
    {
      "epoch": 1.40625,
      "grad_norm": 1.501712441444397,
      "learning_rate": 8.788695491366795e-06,
      "loss": 0.3175,
      "step": 315
    },
    {
      "epoch": 1.4107142857142856,
      "grad_norm": 1.4669523239135742,
      "learning_rate": 8.780978337840796e-06,
      "loss": 0.3059,
      "step": 316
    },
    {
      "epoch": 1.4151785714285714,
      "grad_norm": 1.505470871925354,
      "learning_rate": 8.773240091088335e-06,
      "loss": 0.3163,
      "step": 317
    },
    {
      "epoch": 1.4196428571428572,
      "grad_norm": 1.5221257209777832,
      "learning_rate": 8.765480794285054e-06,
      "loss": 0.3329,
      "step": 318
    },
    {
      "epoch": 1.4241071428571428,
      "grad_norm": 1.4438836574554443,
      "learning_rate": 8.757700490724046e-06,
      "loss": 0.3109,
      "step": 319
    },
    {
      "epoch": 1.4285714285714286,
      "grad_norm": 1.5612000226974487,
      "learning_rate": 8.749899223815618e-06,
      "loss": 0.3512,
      "step": 320
    },
    {
      "epoch": 1.4330357142857144,
      "grad_norm": 1.447957158088684,
      "learning_rate": 8.742077037087032e-06,
      "loss": 0.3244,
      "step": 321
    },
    {
      "epoch": 1.4375,
      "grad_norm": 1.4214884042739868,
      "learning_rate": 8.734233974182276e-06,
      "loss": 0.3149,
      "step": 322
    },
    {
      "epoch": 1.4419642857142856,
      "grad_norm": 1.732344388961792,
      "learning_rate": 8.726370078861825e-06,
      "loss": 0.341,
      "step": 323
    },
    {
      "epoch": 1.4464285714285714,
      "grad_norm": 1.5918134450912476,
      "learning_rate": 8.718485395002377e-06,
      "loss": 0.3636,
      "step": 324
    },
    {
      "epoch": 1.4508928571428572,
      "grad_norm": 1.486258625984192,
      "learning_rate": 8.710579966596625e-06,
      "loss": 0.3428,
      "step": 325
    },
    {
      "epoch": 1.4553571428571428,
      "grad_norm": 1.4833279848098755,
      "learning_rate": 8.702653837753005e-06,
      "loss": 0.3205,
      "step": 326
    },
    {
      "epoch": 1.4598214285714286,
      "grad_norm": 1.620224118232727,
      "learning_rate": 8.694707052695459e-06,
      "loss": 0.3322,
      "step": 327
    },
    {
      "epoch": 1.4642857142857144,
      "grad_norm": 1.5433918237686157,
      "learning_rate": 8.686739655763166e-06,
      "loss": 0.3492,
      "step": 328
    },
    {
      "epoch": 1.46875,
      "grad_norm": 1.4457906484603882,
      "learning_rate": 8.678751691410323e-06,
      "loss": 0.3102,
      "step": 329
    },
    {
      "epoch": 1.4732142857142856,
      "grad_norm": 1.4180353879928589,
      "learning_rate": 8.670743204205875e-06,
      "loss": 0.3432,
      "step": 330
    },
    {
      "epoch": 1.4776785714285714,
      "grad_norm": 1.3930200338363647,
      "learning_rate": 8.662714238833278e-06,
      "loss": 0.294,
      "step": 331
    },
    {
      "epoch": 1.4821428571428572,
      "grad_norm": 1.436591625213623,
      "learning_rate": 8.654664840090247e-06,
      "loss": 0.3666,
      "step": 332
    },
    {
      "epoch": 1.4866071428571428,
      "grad_norm": 1.6145936250686646,
      "learning_rate": 8.6465950528885e-06,
      "loss": 0.323,
      "step": 333
    },
    {
      "epoch": 1.4910714285714286,
      "grad_norm": 1.4172477722167969,
      "learning_rate": 8.638504922253518e-06,
      "loss": 0.3138,
      "step": 334
    },
    {
      "epoch": 1.4955357142857144,
      "grad_norm": 1.4870661497116089,
      "learning_rate": 8.63039449332429e-06,
      "loss": 0.3337,
      "step": 335
    },
    {
      "epoch": 1.5,
      "grad_norm": 1.487033724784851,
      "learning_rate": 8.62226381135305e-06,
      "loss": 0.3157,
      "step": 336
    },
    {
      "epoch": 1.5044642857142856,
      "grad_norm": 1.5234925746917725,
      "learning_rate": 8.614112921705045e-06,
      "loss": 0.3067,
      "step": 337
    },
    {
      "epoch": 1.5089285714285714,
      "grad_norm": 1.5859482288360596,
      "learning_rate": 8.605941869858265e-06,
      "loss": 0.3364,
      "step": 338
    },
    {
      "epoch": 1.5133928571428572,
      "grad_norm": 1.5042221546173096,
      "learning_rate": 8.597750701403197e-06,
      "loss": 0.3187,
      "step": 339
    },
    {
      "epoch": 1.5178571428571428,
      "grad_norm": 1.379291296005249,
      "learning_rate": 8.589539462042566e-06,
      "loss": 0.3073,
      "step": 340
    },
    {
      "epoch": 1.5223214285714286,
      "grad_norm": 1.4755321741104126,
      "learning_rate": 8.581308197591088e-06,
      "loss": 0.309,
      "step": 341
    },
    {
      "epoch": 1.5267857142857144,
      "grad_norm": 1.5573457479476929,
      "learning_rate": 8.573056953975208e-06,
      "loss": 0.3501,
      "step": 342
    },
    {
      "epoch": 1.53125,
      "grad_norm": 1.4575328826904297,
      "learning_rate": 8.56478577723284e-06,
      "loss": 0.3452,
      "step": 343
    },
    {
      "epoch": 1.5357142857142856,
      "grad_norm": 1.4823976755142212,
      "learning_rate": 8.556494713513123e-06,
      "loss": 0.3331,
      "step": 344
    },
    {
      "epoch": 1.5401785714285714,
      "grad_norm": 1.4269355535507202,
      "learning_rate": 8.548183809076146e-06,
      "loss": 0.3107,
      "step": 345
    },
    {
      "epoch": 1.5446428571428572,
      "grad_norm": 1.4220494031906128,
      "learning_rate": 8.539853110292708e-06,
      "loss": 0.3319,
      "step": 346
    },
    {
      "epoch": 1.5491071428571428,
      "grad_norm": 1.5970823764801025,
      "learning_rate": 8.531502663644046e-06,
      "loss": 0.3548,
      "step": 347
    },
    {
      "epoch": 1.5535714285714286,
      "grad_norm": 1.4499046802520752,
      "learning_rate": 8.523132515721586e-06,
      "loss": 0.3369,
      "step": 348
    },
    {
      "epoch": 1.5580357142857144,
      "grad_norm": 1.3973332643508911,
      "learning_rate": 8.51474271322667e-06,
      "loss": 0.3464,
      "step": 349
    },
    {
      "epoch": 1.5625,
      "grad_norm": 1.7323225736618042,
      "learning_rate": 8.506333302970306e-06,
      "loss": 0.3218,
      "step": 350
    },
    {
      "epoch": 1.5669642857142856,
      "grad_norm": 1.3297864198684692,
      "learning_rate": 8.497904331872909e-06,
      "loss": 0.301,
      "step": 351
    },
    {
      "epoch": 1.5714285714285714,
      "grad_norm": 1.5161700248718262,
      "learning_rate": 8.489455846964027e-06,
      "loss": 0.3468,
      "step": 352
    },
    {
      "epoch": 1.5758928571428572,
      "grad_norm": 1.4331910610198975,
      "learning_rate": 8.480987895382086e-06,
      "loss": 0.3613,
      "step": 353
    },
    {
      "epoch": 1.5803571428571428,
      "grad_norm": 1.5763874053955078,
      "learning_rate": 8.472500524374129e-06,
      "loss": 0.357,
      "step": 354
    },
    {
      "epoch": 1.5848214285714286,
      "grad_norm": 1.50728440284729,
      "learning_rate": 8.463993781295552e-06,
      "loss": 0.3252,
      "step": 355
    },
    {
      "epoch": 1.5892857142857144,
      "grad_norm": 1.4221165180206299,
      "learning_rate": 8.45546771360983e-06,
      "loss": 0.3036,
      "step": 356
    },
    {
      "epoch": 1.59375,
      "grad_norm": 1.430577278137207,
      "learning_rate": 8.44692236888827e-06,
      "loss": 0.3187,
      "step": 357
    },
    {
      "epoch": 1.5982142857142856,
      "grad_norm": 1.316603422164917,
      "learning_rate": 8.43835779480973e-06,
      "loss": 0.3275,
      "step": 358
    },
    {
      "epoch": 1.6026785714285714,
      "grad_norm": 1.6066330671310425,
      "learning_rate": 8.429774039160355e-06,
      "loss": 0.336,
      "step": 359
    },
    {
      "epoch": 1.6071428571428572,
      "grad_norm": 1.6857646703720093,
      "learning_rate": 8.421171149833322e-06,
      "loss": 0.3346,
      "step": 360
    },
    {
      "epoch": 1.6116071428571428,
      "grad_norm": 1.6161949634552002,
      "learning_rate": 8.412549174828558e-06,
      "loss": 0.3806,
      "step": 361
    },
    {
      "epoch": 1.6160714285714286,
      "grad_norm": 1.5804469585418701,
      "learning_rate": 8.403908162252481e-06,
      "loss": 0.3332,
      "step": 362
    },
    {
      "epoch": 1.6205357142857144,
      "grad_norm": 1.568248987197876,
      "learning_rate": 8.395248160317728e-06,
      "loss": 0.3175,
      "step": 363
    },
    {
      "epoch": 1.625,
      "grad_norm": 1.4972864389419556,
      "learning_rate": 8.386569217342893e-06,
      "loss": 0.3233,
      "step": 364
    },
    {
      "epoch": 1.6294642857142856,
      "grad_norm": 1.5257365703582764,
      "learning_rate": 8.377871381752246e-06,
      "loss": 0.3273,
      "step": 365
    },
    {
      "epoch": 1.6339285714285714,
      "grad_norm": 1.4747686386108398,
      "learning_rate": 8.369154702075466e-06,
      "loss": 0.3382,
      "step": 366
    },
    {
      "epoch": 1.6383928571428572,
      "grad_norm": 1.5610405206680298,
      "learning_rate": 8.360419226947383e-06,
      "loss": 0.3203,
      "step": 367
    },
    {
      "epoch": 1.6428571428571428,
      "grad_norm": 1.4387973546981812,
      "learning_rate": 8.351665005107686e-06,
      "loss": 0.3271,
      "step": 368
    },
    {
      "epoch": 1.6473214285714286,
      "grad_norm": 1.5783442258834839,
      "learning_rate": 8.34289208540067e-06,
      "loss": 0.3576,
      "step": 369
    },
    {
      "epoch": 1.6517857142857144,
      "grad_norm": 1.5295450687408447,
      "learning_rate": 8.334100516774946e-06,
      "loss": 0.3488,
      "step": 370
    },
    {
      "epoch": 1.65625,
      "grad_norm": 1.5279836654663086,
      "learning_rate": 8.325290348283186e-06,
      "loss": 0.358,
      "step": 371
    },
    {
      "epoch": 1.6607142857142856,
      "grad_norm": 1.5376560688018799,
      "learning_rate": 8.316461629081833e-06,
      "loss": 0.3025,
      "step": 372
    },
    {
      "epoch": 1.6651785714285714,
      "grad_norm": 1.4480183124542236,
      "learning_rate": 8.307614408430839e-06,
      "loss": 0.3609,
      "step": 373
    },
    {
      "epoch": 1.6696428571428572,
      "grad_norm": 1.5485166311264038,
      "learning_rate": 8.298748735693382e-06,
      "loss": 0.3405,
      "step": 374
    },
    {
      "epoch": 1.6741071428571428,
      "grad_norm": 1.408075213432312,
      "learning_rate": 8.289864660335595e-06,
      "loss": 0.322,
      "step": 375
    },
    {
      "epoch": 1.6785714285714286,
      "grad_norm": 1.5506187677383423,
      "learning_rate": 8.280962231926288e-06,
      "loss": 0.3385,
      "step": 376
    },
    {
      "epoch": 1.6830357142857144,
      "grad_norm": 1.6020299196243286,
      "learning_rate": 8.27204150013667e-06,
      "loss": 0.3353,
      "step": 377
    },
    {
      "epoch": 1.6875,
      "grad_norm": 1.4672306776046753,
      "learning_rate": 8.263102514740082e-06,
      "loss": 0.3086,
      "step": 378
    },
    {
      "epoch": 1.6919642857142856,
      "grad_norm": 1.5637385845184326,
      "learning_rate": 8.2541453256117e-06,
      "loss": 0.3136,
      "step": 379
    },
    {
      "epoch": 1.6964285714285714,
      "grad_norm": 1.446553111076355,
      "learning_rate": 8.245169982728276e-06,
      "loss": 0.3389,
      "step": 380
    },
    {
      "epoch": 1.7008928571428572,
      "grad_norm": 1.4007227420806885,
      "learning_rate": 8.23617653616785e-06,
      "loss": 0.3095,
      "step": 381
    },
    {
      "epoch": 1.7053571428571428,
      "grad_norm": 1.454406976699829,
      "learning_rate": 8.227165036109468e-06,
      "loss": 0.329,
      "step": 382
    },
    {
      "epoch": 1.7098214285714286,
      "grad_norm": 1.7223042249679565,
      "learning_rate": 8.218135532832909e-06,
      "loss": 0.3662,
      "step": 383
    },
    {
      "epoch": 1.7142857142857144,
      "grad_norm": 1.5587764978408813,
      "learning_rate": 8.209088076718398e-06,
      "loss": 0.3348,
      "step": 384
    },
    {
      "epoch": 1.71875,
      "grad_norm": 1.602912187576294,
      "learning_rate": 8.20002271824633e-06,
      "loss": 0.3736,
      "step": 385
    },
    {
      "epoch": 1.7232142857142856,
      "grad_norm": 1.5459792613983154,
      "learning_rate": 8.190939507996992e-06,
      "loss": 0.3559,
      "step": 386
    },
    {
      "epoch": 1.7276785714285714,
      "grad_norm": 1.5487347841262817,
      "learning_rate": 8.181838496650266e-06,
      "loss": 0.3466,
      "step": 387
    },
    {
      "epoch": 1.7321428571428572,
      "grad_norm": 1.5751219987869263,
      "learning_rate": 8.17271973498536e-06,
      "loss": 0.3541,
      "step": 388
    },
    {
      "epoch": 1.7366071428571428,
      "grad_norm": 1.5197848081588745,
      "learning_rate": 8.163583273880519e-06,
      "loss": 0.3276,
      "step": 389
    },
    {
      "epoch": 1.7410714285714286,
      "grad_norm": 1.4693080186843872,
      "learning_rate": 8.154429164312742e-06,
      "loss": 0.3051,
      "step": 390
    },
    {
      "epoch": 1.7455357142857144,
      "grad_norm": 1.4455931186676025,
      "learning_rate": 8.145257457357502e-06,
      "loss": 0.328,
      "step": 391
    },
    {
      "epoch": 1.75,
      "grad_norm": 1.5701878070831299,
      "learning_rate": 8.136068204188448e-06,
      "loss": 0.3301,
      "step": 392
    },
    {
      "epoch": 1.7544642857142856,
      "grad_norm": 1.358162522315979,
      "learning_rate": 8.12686145607714e-06,
      "loss": 0.3074,
      "step": 393
    },
    {
      "epoch": 1.7589285714285714,
      "grad_norm": 1.8741075992584229,
      "learning_rate": 8.11763726439274e-06,
      "loss": 0.3883,
      "step": 394
    },
    {
      "epoch": 1.7633928571428572,
      "grad_norm": 1.5605037212371826,
      "learning_rate": 8.108395680601742e-06,
      "loss": 0.3213,
      "step": 395
    },
    {
      "epoch": 1.7678571428571428,
      "grad_norm": 1.5065466165542603,
      "learning_rate": 8.099136756267682e-06,
      "loss": 0.3213,
      "step": 396
    },
    {
      "epoch": 1.7723214285714286,
      "grad_norm": 1.3987689018249512,
      "learning_rate": 8.089860543050843e-06,
      "loss": 0.324,
      "step": 397
    },
    {
      "epoch": 1.7767857142857144,
      "grad_norm": 1.6072906255722046,
      "learning_rate": 8.080567092707973e-06,
      "loss": 0.344,
      "step": 398
    },
    {
      "epoch": 1.78125,
      "grad_norm": 1.4562393426895142,
      "learning_rate": 8.071256457091994e-06,
      "loss": 0.3173,
      "step": 399
    },
    {
      "epoch": 1.7857142857142856,
      "grad_norm": 1.4061006307601929,
      "learning_rate": 8.06192868815172e-06,
      "loss": 0.321,
      "step": 400
    },
    {
      "epoch": 1.7857142857142856,
      "eval_loss": 0.5984869003295898,
      "eval_runtime": 4.436,
      "eval_samples_per_second": 13.526,
      "eval_steps_per_second": 0.902,
      "step": 400
    },
    {
      "epoch": 1.7901785714285714,
      "grad_norm": 1.521924614906311,
      "learning_rate": 8.05258383793155e-06,
      "loss": 0.3542,
      "step": 401
    },
    {
      "epoch": 1.7946428571428572,
      "grad_norm": 1.4396589994430542,
      "learning_rate": 8.043221958571193e-06,
      "loss": 0.3303,
      "step": 402
    },
    {
      "epoch": 1.7991071428571428,
      "grad_norm": 1.5028449296951294,
      "learning_rate": 8.033843102305376e-06,
      "loss": 0.3436,
      "step": 403
    },
    {
      "epoch": 1.8035714285714286,
      "grad_norm": 1.3993929624557495,
      "learning_rate": 8.024447321463545e-06,
      "loss": 0.3207,
      "step": 404
    },
    {
      "epoch": 1.8080357142857144,
      "grad_norm": 1.4850343465805054,
      "learning_rate": 8.015034668469576e-06,
      "loss": 0.3495,
      "step": 405
    },
    {
      "epoch": 1.8125,
      "grad_norm": 1.5270286798477173,
      "learning_rate": 8.005605195841485e-06,
      "loss": 0.3224,
      "step": 406
    },
    {
      "epoch": 1.8169642857142856,
      "grad_norm": 1.462645411491394,
      "learning_rate": 7.996158956191135e-06,
      "loss": 0.2767,
      "step": 407
    },
    {
      "epoch": 1.8214285714285714,
      "grad_norm": 1.292880654335022,
      "learning_rate": 7.986696002223936e-06,
      "loss": 0.2946,
      "step": 408
    },
    {
      "epoch": 1.8258928571428572,
      "grad_norm": 1.4748491048812866,
      "learning_rate": 7.97721638673856e-06,
      "loss": 0.3342,
      "step": 409
    },
    {
      "epoch": 1.8303571428571428,
      "grad_norm": 1.4022036790847778,
      "learning_rate": 7.967720162626643e-06,
      "loss": 0.333,
      "step": 410
    },
    {
      "epoch": 1.8348214285714286,
      "grad_norm": 1.479123830795288,
      "learning_rate": 7.958207382872486e-06,
      "loss": 0.3424,
      "step": 411
    },
    {
      "epoch": 1.8392857142857144,
      "grad_norm": 1.455062747001648,
      "learning_rate": 7.94867810055276e-06,
      "loss": 0.304,
      "step": 412
    },
    {
      "epoch": 1.84375,
      "grad_norm": 1.506971001625061,
      "learning_rate": 7.93913236883622e-06,
      "loss": 0.3409,
      "step": 413
    },
    {
      "epoch": 1.8482142857142856,
      "grad_norm": 1.5958027839660645,
      "learning_rate": 7.929570240983393e-06,
      "loss": 0.3264,
      "step": 414
    },
    {
      "epoch": 1.8526785714285714,
      "grad_norm": 1.3807499408721924,
      "learning_rate": 7.919991770346295e-06,
      "loss": 0.3436,
      "step": 415
    },
    {
      "epoch": 1.8571428571428572,
      "grad_norm": 1.3814480304718018,
      "learning_rate": 7.910397010368122e-06,
      "loss": 0.3288,
      "step": 416
    },
    {
      "epoch": 1.8616071428571428,
      "grad_norm": 1.412415862083435,
      "learning_rate": 7.900786014582957e-06,
      "loss": 0.353,
      "step": 417
    },
    {
      "epoch": 1.8660714285714286,
      "grad_norm": 1.3772637844085693,
      "learning_rate": 7.891158836615472e-06,
      "loss": 0.2992,
      "step": 418
    },
    {
      "epoch": 1.8705357142857144,
      "grad_norm": 1.425160527229309,
      "learning_rate": 7.881515530180629e-06,
      "loss": 0.3536,
      "step": 419
    },
    {
      "epoch": 1.875,
      "grad_norm": 1.4070899486541748,
      "learning_rate": 7.871856149083377e-06,
      "loss": 0.3252,
      "step": 420
    },
    {
      "epoch": 1.8794642857142856,
      "grad_norm": 1.3919873237609863,
      "learning_rate": 7.862180747218354e-06,
      "loss": 0.3378,
      "step": 421
    },
    {
      "epoch": 1.8839285714285714,
      "grad_norm": 1.4947540760040283,
      "learning_rate": 7.852489378569588e-06,
      "loss": 0.3383,
      "step": 422
    },
    {
      "epoch": 1.8883928571428572,
      "grad_norm": 1.500719428062439,
      "learning_rate": 7.84278209721019e-06,
      "loss": 0.3414,
      "step": 423
    },
    {
      "epoch": 1.8928571428571428,
      "grad_norm": 1.202958583831787,
      "learning_rate": 7.83305895730206e-06,
      "loss": 0.2683,
      "step": 424
    },
    {
      "epoch": 1.8973214285714286,
      "grad_norm": 1.518190860748291,
      "learning_rate": 7.823320013095578e-06,
      "loss": 0.3315,
      "step": 425
    },
    {
      "epoch": 1.9017857142857144,
      "grad_norm": 1.435086965560913,
      "learning_rate": 7.81356531892931e-06,
      "loss": 0.3238,
      "step": 426
    },
    {
      "epoch": 1.90625,
      "grad_norm": 1.4547661542892456,
      "learning_rate": 7.803794929229689e-06,
      "loss": 0.3171,
      "step": 427
    },
    {
      "epoch": 1.9107142857142856,
      "grad_norm": 1.5060296058654785,
      "learning_rate": 7.794008898510731e-06,
      "loss": 0.348,
      "step": 428
    },
    {
      "epoch": 1.9151785714285714,
      "grad_norm": 1.5503387451171875,
      "learning_rate": 7.784207281373716e-06,
      "loss": 0.3573,
      "step": 429
    },
    {
      "epoch": 1.9196428571428572,
      "grad_norm": 1.4911384582519531,
      "learning_rate": 7.774390132506892e-06,
      "loss": 0.3234,
      "step": 430
    },
    {
      "epoch": 1.9241071428571428,
      "grad_norm": 1.5123614072799683,
      "learning_rate": 7.764557506685162e-06,
      "loss": 0.3838,
      "step": 431
    },
    {
      "epoch": 1.9285714285714286,
      "grad_norm": 1.529229760169983,
      "learning_rate": 7.754709458769787e-06,
      "loss": 0.3425,
      "step": 432
    },
    {
      "epoch": 1.9330357142857144,
      "grad_norm": 1.4191179275512695,
      "learning_rate": 7.744846043708076e-06,
      "loss": 0.3175,
      "step": 433
    },
    {
      "epoch": 1.9375,
      "grad_norm": 1.5373347997665405,
      "learning_rate": 7.734967316533074e-06,
      "loss": 0.3379,
      "step": 434
    },
    {
      "epoch": 1.9419642857142856,
      "grad_norm": 1.562549114227295,
      "learning_rate": 7.725073332363265e-06,
      "loss": 0.3521,
      "step": 435
    },
    {
      "epoch": 1.9464285714285714,
      "grad_norm": 1.4550033807754517,
      "learning_rate": 7.715164146402259e-06,
      "loss": 0.3314,
      "step": 436
    },
    {
      "epoch": 1.9508928571428572,
      "grad_norm": 1.5194437503814697,
      "learning_rate": 7.705239813938486e-06,
      "loss": 0.3369,
      "step": 437
    },
    {
      "epoch": 1.9553571428571428,
      "grad_norm": 1.331009030342102,
      "learning_rate": 7.69530039034488e-06,
      "loss": 0.3156,
      "step": 438
    },
    {
      "epoch": 1.9598214285714286,
      "grad_norm": 1.4136159420013428,
      "learning_rate": 7.685345931078579e-06,
      "loss": 0.352,
      "step": 439
    },
    {
      "epoch": 1.9642857142857144,
      "grad_norm": 1.4941082000732422,
      "learning_rate": 7.675376491680617e-06,
      "loss": 0.3487,
      "step": 440
    },
    {
      "epoch": 1.96875,
      "grad_norm": 1.5224647521972656,
      "learning_rate": 7.665392127775605e-06,
      "loss": 0.3472,
      "step": 441
    },
    {
      "epoch": 1.9732142857142856,
      "grad_norm": 1.389153242111206,
      "learning_rate": 7.65539289507143e-06,
      "loss": 0.3188,
      "step": 442
    },
    {
      "epoch": 1.9776785714285714,
      "grad_norm": 1.3954931497573853,
      "learning_rate": 7.645378849358931e-06,
      "loss": 0.2805,
      "step": 443
    },
    {
      "epoch": 1.9821428571428572,
      "grad_norm": 1.4153132438659668,
      "learning_rate": 7.635350046511609e-06,
      "loss": 0.3351,
      "step": 444
    },
    {
      "epoch": 1.9866071428571428,
      "grad_norm": 1.4916757345199585,
      "learning_rate": 7.625306542485289e-06,
      "loss": 0.3388,
      "step": 445
    },
    {
      "epoch": 1.9910714285714286,
      "grad_norm": 1.395013689994812,
      "learning_rate": 7.615248393317833e-06,
      "loss": 0.3422,
      "step": 446
    },
    {
      "epoch": 1.9955357142857144,
      "grad_norm": 1.6745942831039429,
      "learning_rate": 7.605175655128809e-06,
      "loss": 0.3205,
      "step": 447
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.5079350471496582,
      "learning_rate": 7.595088384119186e-06,
      "loss": 0.3489,
      "step": 448
    },
    {
      "epoch": 2.0044642857142856,
      "grad_norm": 1.3217769861221313,
      "learning_rate": 7.58498663657102e-06,
      "loss": 0.2103,
      "step": 449
    },
    {
      "epoch": 2.0089285714285716,
      "grad_norm": 1.3270319700241089,
      "learning_rate": 7.57487046884714e-06,
      "loss": 0.2176,
      "step": 450
    },
    {
      "epoch": 2.013392857142857,
      "grad_norm": 1.1652079820632935,
      "learning_rate": 7.5647399373908296e-06,
      "loss": 0.1737,
      "step": 451
    },
    {
      "epoch": 2.017857142857143,
      "grad_norm": 1.2553536891937256,
      "learning_rate": 7.554595098725515e-06,
      "loss": 0.2092,
      "step": 452
    },
    {
      "epoch": 2.0223214285714284,
      "grad_norm": 1.1758341789245605,
      "learning_rate": 7.544436009454454e-06,
      "loss": 0.1817,
      "step": 453
    },
    {
      "epoch": 2.0267857142857144,
      "grad_norm": 1.359291434288025,
      "learning_rate": 7.534262726260413e-06,
      "loss": 0.2095,
      "step": 454
    },
    {
      "epoch": 2.03125,
      "grad_norm": 1.4187179803848267,
      "learning_rate": 7.524075305905351e-06,
      "loss": 0.199,
      "step": 455
    },
    {
      "epoch": 2.0357142857142856,
      "grad_norm": 1.6442433595657349,
      "learning_rate": 7.513873805230111e-06,
      "loss": 0.2029,
      "step": 456
    },
    {
      "epoch": 2.0401785714285716,
      "grad_norm": 1.8652024269104004,
      "learning_rate": 7.5036582811540935e-06,
      "loss": 0.2194,
      "step": 457
    },
    {
      "epoch": 2.044642857142857,
      "grad_norm": 1.7254689931869507,
      "learning_rate": 7.493428790674943e-06,
      "loss": 0.1686,
      "step": 458
    },
    {
      "epoch": 2.049107142857143,
      "grad_norm": 1.8289271593093872,
      "learning_rate": 7.483185390868232e-06,
      "loss": 0.1941,
      "step": 459
    },
    {
      "epoch": 2.0535714285714284,
      "grad_norm": 1.7519882917404175,
      "learning_rate": 7.472928138887134e-06,
      "loss": 0.1833,
      "step": 460
    },
    {
      "epoch": 2.0580357142857144,
      "grad_norm": 1.9071825742721558,
      "learning_rate": 7.462657091962122e-06,
      "loss": 0.202,
      "step": 461
    },
    {
      "epoch": 2.0625,
      "grad_norm": 1.690094232559204,
      "learning_rate": 7.452372307400626e-06,
      "loss": 0.1903,
      "step": 462
    },
    {
      "epoch": 2.0669642857142856,
      "grad_norm": 1.6332119703292847,
      "learning_rate": 7.442073842586733e-06,
      "loss": 0.1868,
      "step": 463
    },
    {
      "epoch": 2.0714285714285716,
      "grad_norm": 1.5547043085098267,
      "learning_rate": 7.43176175498086e-06,
      "loss": 0.17,
      "step": 464
    },
    {
      "epoch": 2.075892857142857,
      "grad_norm": 1.6427596807479858,
      "learning_rate": 7.421436102119427e-06,
      "loss": 0.2107,
      "step": 465
    },
    {
      "epoch": 2.080357142857143,
      "grad_norm": 1.4404045343399048,
      "learning_rate": 7.411096941614543e-06,
      "loss": 0.1707,
      "step": 466
    },
    {
      "epoch": 2.0848214285714284,
      "grad_norm": 1.3924144506454468,
      "learning_rate": 7.400744331153684e-06,
      "loss": 0.1847,
      "step": 467
    },
    {
      "epoch": 2.0892857142857144,
      "grad_norm": 1.2086669206619263,
      "learning_rate": 7.390378328499372e-06,
      "loss": 0.1564,
      "step": 468
    },
    {
      "epoch": 2.09375,
      "grad_norm": 1.4519152641296387,
      "learning_rate": 7.3799989914888506e-06,
      "loss": 0.1997,
      "step": 469
    },
    {
      "epoch": 2.0982142857142856,
      "grad_norm": 1.478129506111145,
      "learning_rate": 7.3696063780337566e-06,
      "loss": 0.1925,
      "step": 470
    },
    {
      "epoch": 2.1026785714285716,
      "grad_norm": 1.650312900543213,
      "learning_rate": 7.359200546119813e-06,
      "loss": 0.1883,
      "step": 471
    },
    {
      "epoch": 2.107142857142857,
      "grad_norm": 1.4365777969360352,
      "learning_rate": 7.3487815538064865e-06,
      "loss": 0.1868,
      "step": 472
    },
    {
      "epoch": 2.111607142857143,
      "grad_norm": 1.342756748199463,
      "learning_rate": 7.338349459226678e-06,
      "loss": 0.1692,
      "step": 473
    },
    {
      "epoch": 2.1160714285714284,
      "grad_norm": 1.5859627723693848,
      "learning_rate": 7.327904320586387e-06,
      "loss": 0.201,
      "step": 474
    },
    {
      "epoch": 2.1205357142857144,
      "grad_norm": 1.499635100364685,
      "learning_rate": 7.3174461961644e-06,
      "loss": 0.1817,
      "step": 475
    },
    {
      "epoch": 2.125,
      "grad_norm": 2.32686710357666,
      "learning_rate": 7.3069751443119505e-06,
      "loss": 0.1704,
      "step": 476
    },
    {
      "epoch": 2.1294642857142856,
      "grad_norm": 1.412901759147644,
      "learning_rate": 7.296491223452407e-06,
      "loss": 0.1828,
      "step": 477
    },
    {
      "epoch": 2.1339285714285716,
      "grad_norm": 1.5171948671340942,
      "learning_rate": 7.285994492080934e-06,
      "loss": 0.1953,
      "step": 478
    },
    {
      "epoch": 2.138392857142857,
      "grad_norm": 1.5599428415298462,
      "learning_rate": 7.275485008764183e-06,
      "loss": 0.1923,
      "step": 479
    },
    {
      "epoch": 2.142857142857143,
      "grad_norm": 1.5545156002044678,
      "learning_rate": 7.2649628321399415e-06,
      "loss": 0.19,
      "step": 480
    },
    {
      "epoch": 2.1473214285714284,
      "grad_norm": 1.479979157447815,
      "learning_rate": 7.254428020916829e-06,
      "loss": 0.1862,
      "step": 481
    },
    {
      "epoch": 2.1517857142857144,
      "grad_norm": 1.4323021173477173,
      "learning_rate": 7.243880633873957e-06,
      "loss": 0.1968,
      "step": 482
    },
    {
      "epoch": 2.15625,
      "grad_norm": 1.5759376287460327,
      "learning_rate": 7.2333207298606075e-06,
      "loss": 0.1912,
      "step": 483
    },
    {
      "epoch": 2.1607142857142856,
      "grad_norm": 1.4067718982696533,
      "learning_rate": 7.222748367795892e-06,
      "loss": 0.1968,
      "step": 484
    },
    {
      "epoch": 2.1651785714285716,
      "grad_norm": 1.4289413690567017,
      "learning_rate": 7.212163606668442e-06,
      "loss": 0.181,
      "step": 485
    },
    {
      "epoch": 2.169642857142857,
      "grad_norm": 1.386183738708496,
      "learning_rate": 7.201566505536065e-06,
      "loss": 0.1859,
      "step": 486
    },
    {
      "epoch": 2.174107142857143,
      "grad_norm": 1.5737900733947754,
      "learning_rate": 7.190957123525417e-06,
      "loss": 0.1975,
      "step": 487
    },
    {
      "epoch": 2.1785714285714284,
      "grad_norm": 1.4346133470535278,
      "learning_rate": 7.180335519831685e-06,
      "loss": 0.1988,
      "step": 488
    },
    {
      "epoch": 2.1830357142857144,
      "grad_norm": 1.6407471895217896,
      "learning_rate": 7.169701753718232e-06,
      "loss": 0.1942,
      "step": 489
    },
    {
      "epoch": 2.1875,
      "grad_norm": 1.4272671937942505,
      "learning_rate": 7.159055884516297e-06,
      "loss": 0.1709,
      "step": 490
    },
    {
      "epoch": 2.1919642857142856,
      "grad_norm": 1.4495371580123901,
      "learning_rate": 7.148397971624636e-06,
      "loss": 0.1857,
      "step": 491
    },
    {
      "epoch": 2.1964285714285716,
      "grad_norm": 1.474555253982544,
      "learning_rate": 7.137728074509211e-06,
      "loss": 0.1994,
      "step": 492
    },
    {
      "epoch": 2.200892857142857,
      "grad_norm": 1.3996185064315796,
      "learning_rate": 7.127046252702847e-06,
      "loss": 0.1976,
      "step": 493
    },
    {
      "epoch": 2.205357142857143,
      "grad_norm": 1.4525631666183472,
      "learning_rate": 7.116352565804904e-06,
      "loss": 0.1781,
      "step": 494
    },
    {
      "epoch": 2.2098214285714284,
      "grad_norm": 1.3915441036224365,
      "learning_rate": 7.105647073480939e-06,
      "loss": 0.1934,
      "step": 495
    },
    {
      "epoch": 2.2142857142857144,
      "grad_norm": 1.548956274986267,
      "learning_rate": 7.0949298354623855e-06,
      "loss": 0.1956,
      "step": 496
    },
    {
      "epoch": 2.21875,
      "grad_norm": 1.67276930809021,
      "learning_rate": 7.084200911546205e-06,
      "loss": 0.2031,
      "step": 497
    },
    {
      "epoch": 2.2232142857142856,
      "grad_norm": 1.4230951070785522,
      "learning_rate": 7.073460361594565e-06,
      "loss": 0.1747,
      "step": 498
    },
    {
      "epoch": 2.2276785714285716,
      "grad_norm": 1.3589422702789307,
      "learning_rate": 7.0627082455344984e-06,
      "loss": 0.1697,
      "step": 499
    },
    {
      "epoch": 2.232142857142857,
      "grad_norm": 1.5158501863479614,
      "learning_rate": 7.0519446233575715e-06,
      "loss": 0.1931,
      "step": 500
    },
    {
      "epoch": 2.236607142857143,
      "grad_norm": 1.584936499595642,
      "learning_rate": 7.041169555119552e-06,
      "loss": 0.1835,
      "step": 501
    },
    {
      "epoch": 2.2410714285714284,
      "grad_norm": 1.75810968875885,
      "learning_rate": 7.030383100940068e-06,
      "loss": 0.2101,
      "step": 502
    },
    {
      "epoch": 2.2455357142857144,
      "grad_norm": 1.5574495792388916,
      "learning_rate": 7.019585321002276e-06,
      "loss": 0.1967,
      "step": 503
    },
    {
      "epoch": 2.25,
      "grad_norm": 1.4675954580307007,
      "learning_rate": 7.008776275552522e-06,
      "loss": 0.1835,
      "step": 504
    },
    {
      "epoch": 2.2544642857142856,
      "grad_norm": 1.5428667068481445,
      "learning_rate": 6.997956024900014e-06,
      "loss": 0.1758,
      "step": 505
    },
    {
      "epoch": 2.2589285714285716,
      "grad_norm": 1.7386614084243774,
      "learning_rate": 6.9871246294164775e-06,
      "loss": 0.2004,
      "step": 506
    },
    {
      "epoch": 2.263392857142857,
      "grad_norm": 1.6915313005447388,
      "learning_rate": 6.9762821495358194e-06,
      "loss": 0.2123,
      "step": 507
    },
    {
      "epoch": 2.267857142857143,
      "grad_norm": 1.512141227722168,
      "learning_rate": 6.965428645753792e-06,
      "loss": 0.1849,
      "step": 508
    },
    {
      "epoch": 2.2723214285714284,
      "grad_norm": 1.3899809122085571,
      "learning_rate": 6.954564178627655e-06,
      "loss": 0.1755,
      "step": 509
    },
    {
      "epoch": 2.2767857142857144,
      "grad_norm": 1.446179747581482,
      "learning_rate": 6.943688808775843e-06,
      "loss": 0.2002,
      "step": 510
    },
    {
      "epoch": 2.28125,
      "grad_norm": 1.4694817066192627,
      "learning_rate": 6.9328025968776155e-06,
      "loss": 0.1733,
      "step": 511
    },
    {
      "epoch": 2.2857142857142856,
      "grad_norm": 1.4220669269561768,
      "learning_rate": 6.921905603672733e-06,
      "loss": 0.1865,
      "step": 512
    },
    {
      "epoch": 2.2901785714285716,
      "grad_norm": 1.5075308084487915,
      "learning_rate": 6.910997889961098e-06,
      "loss": 0.2068,
      "step": 513
    },
    {
      "epoch": 2.294642857142857,
      "grad_norm": 1.433173418045044,
      "learning_rate": 6.900079516602445e-06,
      "loss": 0.1646,
      "step": 514
    },
    {
      "epoch": 2.299107142857143,
      "grad_norm": 1.5466556549072266,
      "learning_rate": 6.889150544515972e-06,
      "loss": 0.1876,
      "step": 515
    },
    {
      "epoch": 2.3035714285714284,
      "grad_norm": 1.5720268487930298,
      "learning_rate": 6.8782110346800155e-06,
      "loss": 0.1895,
      "step": 516
    },
    {
      "epoch": 2.3080357142857144,
      "grad_norm": 1.5228114128112793,
      "learning_rate": 6.867261048131712e-06,
      "loss": 0.1775,
      "step": 517
    },
    {
      "epoch": 2.3125,
      "grad_norm": 1.5427687168121338,
      "learning_rate": 6.856300645966645e-06,
      "loss": 0.1949,
      "step": 518
    },
    {
      "epoch": 2.3169642857142856,
      "grad_norm": 1.4784200191497803,
      "learning_rate": 6.845329889338519e-06,
      "loss": 0.1972,
      "step": 519
    },
    {
      "epoch": 2.3214285714285716,
      "grad_norm": 1.5751314163208008,
      "learning_rate": 6.834348839458806e-06,
      "loss": 0.192,
      "step": 520
    },
    {
      "epoch": 2.325892857142857,
      "grad_norm": 1.4170998334884644,
      "learning_rate": 6.823357557596416e-06,
      "loss": 0.1739,
      "step": 521
    },
    {
      "epoch": 2.330357142857143,
      "grad_norm": 1.5813453197479248,
      "learning_rate": 6.81235610507734e-06,
      "loss": 0.1973,
      "step": 522
    },
    {
      "epoch": 2.3348214285714284,
      "grad_norm": 1.4130178689956665,
      "learning_rate": 6.801344543284324e-06,
      "loss": 0.1784,
      "step": 523
    },
    {
      "epoch": 2.3392857142857144,
      "grad_norm": 1.503487229347229,
      "learning_rate": 6.790322933656515e-06,
      "loss": 0.1758,
      "step": 524
    },
    {
      "epoch": 2.34375,
      "grad_norm": 1.5043838024139404,
      "learning_rate": 6.779291337689122e-06,
      "loss": 0.1931,
      "step": 525
    },
    {
      "epoch": 2.3482142857142856,
      "grad_norm": 1.5936574935913086,
      "learning_rate": 6.768249816933074e-06,
      "loss": 0.2102,
      "step": 526
    },
    {
      "epoch": 2.3526785714285716,
      "grad_norm": 1.39701247215271,
      "learning_rate": 6.757198432994674e-06,
      "loss": 0.1888,
      "step": 527
    },
    {
      "epoch": 2.357142857142857,
      "grad_norm": 1.4147759675979614,
      "learning_rate": 6.7461372475352585e-06,
      "loss": 0.1728,
      "step": 528
    },
    {
      "epoch": 2.361607142857143,
      "grad_norm": 1.5155245065689087,
      "learning_rate": 6.73506632227085e-06,
      "loss": 0.2029,
      "step": 529
    },
    {
      "epoch": 2.3660714285714284,
      "grad_norm": 1.3971562385559082,
      "learning_rate": 6.723985718971818e-06,
      "loss": 0.1712,
      "step": 530
    },
    {
      "epoch": 2.3705357142857144,
      "grad_norm": 1.4848732948303223,
      "learning_rate": 6.712895499462527e-06,
      "loss": 0.1778,
      "step": 531
    },
    {
      "epoch": 2.375,
      "grad_norm": 1.3537763357162476,
      "learning_rate": 6.701795725620995e-06,
      "loss": 0.1807,
      "step": 532
    },
    {
      "epoch": 2.3794642857142856,
      "grad_norm": 1.278257131576538,
      "learning_rate": 6.69068645937855e-06,
      "loss": 0.1739,
      "step": 533
    },
    {
      "epoch": 2.3839285714285716,
      "grad_norm": 1.4302947521209717,
      "learning_rate": 6.6795677627194835e-06,
      "loss": 0.1757,
      "step": 534
    },
    {
      "epoch": 2.388392857142857,
      "grad_norm": 1.5175752639770508,
      "learning_rate": 6.668439697680704e-06,
      "loss": 0.1744,
      "step": 535
    },
    {
      "epoch": 2.392857142857143,
      "grad_norm": 1.582108974456787,
      "learning_rate": 6.65730232635139e-06,
      "loss": 0.2061,
      "step": 536
    },
    {
      "epoch": 2.3973214285714284,
      "grad_norm": 1.3695740699768066,
      "learning_rate": 6.6461557108726435e-06,
      "loss": 0.1861,
      "step": 537
    },
    {
      "epoch": 2.4017857142857144,
      "grad_norm": 1.4187901020050049,
      "learning_rate": 6.634999913437148e-06,
      "loss": 0.1646,
      "step": 538
    },
    {
      "epoch": 2.40625,
      "grad_norm": 1.5142133235931396,
      "learning_rate": 6.623834996288816e-06,
      "loss": 0.1878,
      "step": 539
    },
    {
      "epoch": 2.4107142857142856,
      "grad_norm": 1.4513888359069824,
      "learning_rate": 6.6126610217224405e-06,
      "loss": 0.2083,
      "step": 540
    },
    {
      "epoch": 2.4151785714285716,
      "grad_norm": 1.6881784200668335,
      "learning_rate": 6.601478052083356e-06,
      "loss": 0.215,
      "step": 541
    },
    {
      "epoch": 2.419642857142857,
      "grad_norm": 1.5719410181045532,
      "learning_rate": 6.59028614976708e-06,
      "loss": 0.2007,
      "step": 542
    },
    {
      "epoch": 2.424107142857143,
      "grad_norm": 1.5767160654067993,
      "learning_rate": 6.579085377218973e-06,
      "loss": 0.1993,
      "step": 543
    },
    {
      "epoch": 2.4285714285714284,
      "grad_norm": 1.4855055809020996,
      "learning_rate": 6.567875796933888e-06,
      "loss": 0.1962,
      "step": 544
    },
    {
      "epoch": 2.4330357142857144,
      "grad_norm": 1.4299588203430176,
      "learning_rate": 6.556657471455817e-06,
      "loss": 0.1904,
      "step": 545
    },
    {
      "epoch": 2.4375,
      "grad_norm": 1.4422001838684082,
      "learning_rate": 6.54543046337755e-06,
      "loss": 0.211,
      "step": 546
    },
    {
      "epoch": 2.4419642857142856,
      "grad_norm": 1.6163188219070435,
      "learning_rate": 6.534194835340321e-06,
      "loss": 0.2123,
      "step": 547
    },
    {
      "epoch": 2.4464285714285716,
      "grad_norm": 1.485174298286438,
      "learning_rate": 6.522950650033454e-06,
      "loss": 0.1834,
      "step": 548
    },
    {
      "epoch": 2.450892857142857,
      "grad_norm": 1.5538384914398193,
      "learning_rate": 6.511697970194024e-06,
      "loss": 0.1982,
      "step": 549
    },
    {
      "epoch": 2.455357142857143,
      "grad_norm": 1.3437016010284424,
      "learning_rate": 6.500436858606501e-06,
      "loss": 0.1809,
      "step": 550
    },
    {
      "epoch": 2.4598214285714284,
      "grad_norm": 1.5390422344207764,
      "learning_rate": 6.4891673781023975e-06,
      "loss": 0.1861,
      "step": 551
    },
    {
      "epoch": 2.4642857142857144,
      "grad_norm": 1.5996623039245605,
      "learning_rate": 6.477889591559926e-06,
      "loss": 0.2132,
      "step": 552
    },
    {
      "epoch": 2.46875,
      "grad_norm": 1.3555296659469604,
      "learning_rate": 6.466603561903633e-06,
      "loss": 0.1709,
      "step": 553
    },
    {
      "epoch": 2.4732142857142856,
      "grad_norm": 1.4055936336517334,
      "learning_rate": 6.455309352104065e-06,
      "loss": 0.1931,
      "step": 554
    },
    {
      "epoch": 2.4776785714285716,
      "grad_norm": 1.490340232849121,
      "learning_rate": 6.444007025177407e-06,
      "loss": 0.1923,
      "step": 555
    },
    {
      "epoch": 2.482142857142857,
      "grad_norm": 1.451197862625122,
      "learning_rate": 6.4326966441851355e-06,
      "loss": 0.1822,
      "step": 556
    },
    {
      "epoch": 2.486607142857143,
      "grad_norm": 1.33744215965271,
      "learning_rate": 6.4213782722336625e-06,
      "loss": 0.1557,
      "step": 557
    },
    {
      "epoch": 2.4910714285714284,
      "grad_norm": 1.7766042947769165,
      "learning_rate": 6.4100519724739875e-06,
      "loss": 0.2113,
      "step": 558
    },
    {
      "epoch": 2.4955357142857144,
      "grad_norm": 1.5189027786254883,
      "learning_rate": 6.3987178081013446e-06,
      "loss": 0.1917,
      "step": 559
    },
    {
      "epoch": 2.5,
      "grad_norm": 1.4922457933425903,
      "learning_rate": 6.387375842354843e-06,
      "loss": 0.1928,
      "step": 560
    },
    {
      "epoch": 2.5044642857142856,
      "grad_norm": 1.5204309225082397,
      "learning_rate": 6.376026138517125e-06,
      "loss": 0.1852,
      "step": 561
    },
    {
      "epoch": 2.508928571428571,
      "grad_norm": 1.5281089544296265,
      "learning_rate": 6.364668759914005e-06,
      "loss": 0.1955,
      "step": 562
    },
    {
      "epoch": 2.513392857142857,
      "grad_norm": 1.4325025081634521,
      "learning_rate": 6.353303769914121e-06,
      "loss": 0.2076,
      "step": 563
    },
    {
      "epoch": 2.517857142857143,
      "grad_norm": 1.4212669134140015,
      "learning_rate": 6.341931231928577e-06,
      "loss": 0.1674,
      "step": 564
    },
    {
      "epoch": 2.522321428571429,
      "grad_norm": 1.6153303384780884,
      "learning_rate": 6.330551209410593e-06,
      "loss": 0.1878,
      "step": 565
    },
    {
      "epoch": 2.5267857142857144,
      "grad_norm": 1.956329345703125,
      "learning_rate": 6.319163765855146e-06,
      "loss": 0.2096,
      "step": 566
    },
    {
      "epoch": 2.53125,
      "grad_norm": 1.5983258485794067,
      "learning_rate": 6.307768964798623e-06,
      "loss": 0.1859,
      "step": 567
    },
    {
      "epoch": 2.5357142857142856,
      "grad_norm": 1.6349304914474487,
      "learning_rate": 6.296366869818458e-06,
      "loss": 0.2049,
      "step": 568
    },
    {
      "epoch": 2.540178571428571,
      "grad_norm": 1.6352767944335938,
      "learning_rate": 6.284957544532783e-06,
      "loss": 0.2139,
      "step": 569
    },
    {
      "epoch": 2.544642857142857,
      "grad_norm": 1.6097818613052368,
      "learning_rate": 6.273541052600074e-06,
      "loss": 0.2227,
      "step": 570
    },
    {
      "epoch": 2.549107142857143,
      "grad_norm": 1.4204213619232178,
      "learning_rate": 6.2621174577187895e-06,
      "loss": 0.1626,
      "step": 571
    },
    {
      "epoch": 2.553571428571429,
      "grad_norm": 1.5108147859573364,
      "learning_rate": 6.250686823627022e-06,
      "loss": 0.1906,
      "step": 572
    },
    {
      "epoch": 2.5580357142857144,
      "grad_norm": 1.318894863128662,
      "learning_rate": 6.239249214102139e-06,
      "loss": 0.1644,
      "step": 573
    },
    {
      "epoch": 2.5625,
      "grad_norm": 1.4823791980743408,
      "learning_rate": 6.2278046929604265e-06,
      "loss": 0.1953,
      "step": 574
    },
    {
      "epoch": 2.5669642857142856,
      "grad_norm": 1.440727949142456,
      "learning_rate": 6.216353324056732e-06,
      "loss": 0.182,
      "step": 575
    },
    {
      "epoch": 2.571428571428571,
      "grad_norm": 1.3238186836242676,
      "learning_rate": 6.204895171284115e-06,
      "loss": 0.1706,
      "step": 576
    },
    {
      "epoch": 2.575892857142857,
      "grad_norm": 1.4335119724273682,
      "learning_rate": 6.193430298573481e-06,
      "loss": 0.2052,
      "step": 577
    },
    {
      "epoch": 2.580357142857143,
      "grad_norm": 1.4563027620315552,
      "learning_rate": 6.181958769893234e-06,
      "loss": 0.2076,
      "step": 578
    },
    {
      "epoch": 2.584821428571429,
      "grad_norm": 1.4558268785476685,
      "learning_rate": 6.17048064924891e-06,
      "loss": 0.2074,
      "step": 579
    },
    {
      "epoch": 2.5892857142857144,
      "grad_norm": 1.4916455745697021,
      "learning_rate": 6.158996000682829e-06,
      "loss": 0.1868,
      "step": 580
    },
    {
      "epoch": 2.59375,
      "grad_norm": 1.5160123109817505,
      "learning_rate": 6.147504888273736e-06,
      "loss": 0.1952,
      "step": 581
    },
    {
      "epoch": 2.5982142857142856,
      "grad_norm": 1.518829107284546,
      "learning_rate": 6.136007376136429e-06,
      "loss": 0.2029,
      "step": 582
    },
    {
      "epoch": 2.602678571428571,
      "grad_norm": 1.2285252809524536,
      "learning_rate": 6.124503528421429e-06,
      "loss": 0.1513,
      "step": 583
    },
    {
      "epoch": 2.607142857142857,
      "grad_norm": 1.465366244316101,
      "learning_rate": 6.112993409314594e-06,
      "loss": 0.183,
      "step": 584
    },
    {
      "epoch": 2.611607142857143,
      "grad_norm": 1.3672759532928467,
      "learning_rate": 6.101477083036783e-06,
      "loss": 0.1763,
      "step": 585
    },
    {
      "epoch": 2.616071428571429,
      "grad_norm": 1.5228221416473389,
      "learning_rate": 6.0899546138434785e-06,
      "loss": 0.2012,
      "step": 586
    },
    {
      "epoch": 2.6205357142857144,
      "grad_norm": 1.5724462270736694,
      "learning_rate": 6.0784260660244475e-06,
      "loss": 0.1689,
      "step": 587
    },
    {
      "epoch": 2.625,
      "grad_norm": 1.5473616123199463,
      "learning_rate": 6.066891503903363e-06,
      "loss": 0.1887,
      "step": 588
    },
    {
      "epoch": 2.6294642857142856,
      "grad_norm": 1.4898332357406616,
      "learning_rate": 6.0553509918374635e-06,
      "loss": 0.1752,
      "step": 589
    },
    {
      "epoch": 2.633928571428571,
      "grad_norm": 1.443292498588562,
      "learning_rate": 6.0438045942171775e-06,
      "loss": 0.1846,
      "step": 590
    },
    {
      "epoch": 2.638392857142857,
      "grad_norm": 1.4720582962036133,
      "learning_rate": 6.032252375465778e-06,
      "loss": 0.1795,
      "step": 591
    },
    {
      "epoch": 2.642857142857143,
      "grad_norm": 1.4213638305664062,
      "learning_rate": 6.020694400039017e-06,
      "loss": 0.1745,
      "step": 592
    },
    {
      "epoch": 2.647321428571429,
      "grad_norm": 1.5590949058532715,
      "learning_rate": 6.009130732424758e-06,
      "loss": 0.1866,
      "step": 593
    },
    {
      "epoch": 2.6517857142857144,
      "grad_norm": 1.6477134227752686,
      "learning_rate": 5.997561437142636e-06,
      "loss": 0.2065,
      "step": 594
    },
    {
      "epoch": 2.65625,
      "grad_norm": 12.8923921585083,
      "learning_rate": 5.985986578743676e-06,
      "loss": 0.1789,
      "step": 595
    },
    {
      "epoch": 2.6607142857142856,
      "grad_norm": 1.4159936904907227,
      "learning_rate": 5.974406221809949e-06,
      "loss": 0.1875,
      "step": 596
    },
    {
      "epoch": 2.665178571428571,
      "grad_norm": 1.6385328769683838,
      "learning_rate": 5.962820430954198e-06,
      "loss": 0.2118,
      "step": 597
    },
    {
      "epoch": 2.669642857142857,
      "grad_norm": 1.4674583673477173,
      "learning_rate": 5.951229270819494e-06,
      "loss": 0.1992,
      "step": 598
    },
    {
      "epoch": 2.674107142857143,
      "grad_norm": 1.3399938344955444,
      "learning_rate": 5.9396328060788576e-06,
      "loss": 0.1903,
      "step": 599
    },
    {
      "epoch": 2.678571428571429,
      "grad_norm": 1.516242265701294,
      "learning_rate": 5.928031101434908e-06,
      "loss": 0.1835,
      "step": 600
    },
    {
      "epoch": 2.678571428571429,
      "eval_loss": 0.6832678914070129,
      "eval_runtime": 4.5972,
      "eval_samples_per_second": 13.051,
      "eval_steps_per_second": 0.87,
      "step": 600
    },
    {
      "epoch": 2.6830357142857144,
      "grad_norm": 1.591635823249817,
      "learning_rate": 5.916424221619507e-06,
      "loss": 0.2111,
      "step": 601
    },
    {
      "epoch": 2.6875,
      "grad_norm": 1.3842867612838745,
      "learning_rate": 5.904812231393383e-06,
      "loss": 0.1948,
      "step": 602
    },
    {
      "epoch": 2.6919642857142856,
      "grad_norm": 1.2690483331680298,
      "learning_rate": 5.893195195545784e-06,
      "loss": 0.1747,
      "step": 603
    },
    {
      "epoch": 2.696428571428571,
      "grad_norm": 1.5160295963287354,
      "learning_rate": 5.8815731788941064e-06,
      "loss": 0.185,
      "step": 604
    },
    {
      "epoch": 2.700892857142857,
      "grad_norm": 1.5125572681427002,
      "learning_rate": 5.86994624628354e-06,
      "loss": 0.1879,
      "step": 605
    },
    {
      "epoch": 2.705357142857143,
      "grad_norm": 1.4671415090560913,
      "learning_rate": 5.858314462586697e-06,
      "loss": 0.1826,
      "step": 606
    },
    {
      "epoch": 2.709821428571429,
      "grad_norm": 1.5814452171325684,
      "learning_rate": 5.846677892703268e-06,
      "loss": 0.2095,
      "step": 607
    },
    {
      "epoch": 2.7142857142857144,
      "grad_norm": 1.3826875686645508,
      "learning_rate": 5.835036601559634e-06,
      "loss": 0.1791,
      "step": 608
    },
    {
      "epoch": 2.71875,
      "grad_norm": 1.4754812717437744,
      "learning_rate": 5.82339065410853e-06,
      "loss": 0.1866,
      "step": 609
    },
    {
      "epoch": 2.7232142857142856,
      "grad_norm": 1.541463017463684,
      "learning_rate": 5.811740115328665e-06,
      "loss": 0.1812,
      "step": 610
    },
    {
      "epoch": 2.727678571428571,
      "grad_norm": 1.5529158115386963,
      "learning_rate": 5.800085050224367e-06,
      "loss": 0.198,
      "step": 611
    },
    {
      "epoch": 2.732142857142857,
      "grad_norm": 1.5426830053329468,
      "learning_rate": 5.7884255238252175e-06,
      "loss": 0.1981,
      "step": 612
    },
    {
      "epoch": 2.736607142857143,
      "grad_norm": 1.610823154449463,
      "learning_rate": 5.776761601185692e-06,
      "loss": 0.1886,
      "step": 613
    },
    {
      "epoch": 2.741071428571429,
      "grad_norm": 1.5132354497909546,
      "learning_rate": 5.765093347384793e-06,
      "loss": 0.1957,
      "step": 614
    },
    {
      "epoch": 2.7455357142857144,
      "grad_norm": 1.4942512512207031,
      "learning_rate": 5.753420827525691e-06,
      "loss": 0.2006,
      "step": 615
    },
    {
      "epoch": 2.75,
      "grad_norm": 1.5573617219924927,
      "learning_rate": 5.741744106735354e-06,
      "loss": 0.1947,
      "step": 616
    },
    {
      "epoch": 2.7544642857142856,
      "grad_norm": 1.4934524297714233,
      "learning_rate": 5.730063250164196e-06,
      "loss": 0.177,
      "step": 617
    },
    {
      "epoch": 2.758928571428571,
      "grad_norm": 1.6269088983535767,
      "learning_rate": 5.718378322985702e-06,
      "loss": 0.1986,
      "step": 618
    },
    {
      "epoch": 2.763392857142857,
      "grad_norm": 1.3660807609558105,
      "learning_rate": 5.70668939039607e-06,
      "loss": 0.1755,
      "step": 619
    },
    {
      "epoch": 2.767857142857143,
      "grad_norm": 1.5547926425933838,
      "learning_rate": 5.694996517613847e-06,
      "loss": 0.2123,
      "step": 620
    },
    {
      "epoch": 2.772321428571429,
      "grad_norm": 1.579438328742981,
      "learning_rate": 5.683299769879562e-06,
      "loss": 0.1965,
      "step": 621
    },
    {
      "epoch": 2.7767857142857144,
      "grad_norm": 1.6027774810791016,
      "learning_rate": 5.6715992124553685e-06,
      "loss": 0.1928,
      "step": 622
    },
    {
      "epoch": 2.78125,
      "grad_norm": 1.556585669517517,
      "learning_rate": 5.6598949106246734e-06,
      "loss": 0.203,
      "step": 623
    },
    {
      "epoch": 2.7857142857142856,
      "grad_norm": 1.7546970844268799,
      "learning_rate": 5.648186929691776e-06,
      "loss": 0.2128,
      "step": 624
    },
    {
      "epoch": 2.790178571428571,
      "grad_norm": 1.5303380489349365,
      "learning_rate": 5.6364753349815035e-06,
      "loss": 0.1956,
      "step": 625
    },
    {
      "epoch": 2.794642857142857,
      "grad_norm": 1.4398573637008667,
      "learning_rate": 5.624760191838845e-06,
      "loss": 0.1818,
      "step": 626
    },
    {
      "epoch": 2.799107142857143,
      "grad_norm": 1.4977809190750122,
      "learning_rate": 5.61304156562859e-06,
      "loss": 0.2104,
      "step": 627
    },
    {
      "epoch": 2.803571428571429,
      "grad_norm": 1.465651512145996,
      "learning_rate": 5.60131952173496e-06,
      "loss": 0.1905,
      "step": 628
    },
    {
      "epoch": 2.8080357142857144,
      "grad_norm": 1.5513821840286255,
      "learning_rate": 5.589594125561246e-06,
      "loss": 0.201,
      "step": 629
    },
    {
      "epoch": 2.8125,
      "grad_norm": 1.4582335948944092,
      "learning_rate": 5.577865442529447e-06,
      "loss": 0.2003,
      "step": 630
    },
    {
      "epoch": 2.8169642857142856,
      "grad_norm": 1.513297438621521,
      "learning_rate": 5.566133538079893e-06,
      "loss": 0.1771,
      "step": 631
    },
    {
      "epoch": 2.821428571428571,
      "grad_norm": 1.5092986822128296,
      "learning_rate": 5.554398477670895e-06,
      "loss": 0.185,
      "step": 632
    },
    {
      "epoch": 2.825892857142857,
      "grad_norm": 1.3965249061584473,
      "learning_rate": 5.54266032677837e-06,
      "loss": 0.1931,
      "step": 633
    },
    {
      "epoch": 2.830357142857143,
      "grad_norm": 1.5133389234542847,
      "learning_rate": 5.53091915089548e-06,
      "loss": 0.1805,
      "step": 634
    },
    {
      "epoch": 2.834821428571429,
      "grad_norm": 1.5035574436187744,
      "learning_rate": 5.5191750155322595e-06,
      "loss": 0.1936,
      "step": 635
    },
    {
      "epoch": 2.8392857142857144,
      "grad_norm": 1.4027115106582642,
      "learning_rate": 5.507427986215265e-06,
      "loss": 0.188,
      "step": 636
    },
    {
      "epoch": 2.84375,
      "grad_norm": 1.6537582874298096,
      "learning_rate": 5.49567812848719e-06,
      "loss": 0.1989,
      "step": 637
    },
    {
      "epoch": 2.8482142857142856,
      "grad_norm": 1.6560035943984985,
      "learning_rate": 5.483925507906514e-06,
      "loss": 0.2003,
      "step": 638
    },
    {
      "epoch": 2.852678571428571,
      "grad_norm": 1.5221431255340576,
      "learning_rate": 5.4721701900471335e-06,
      "loss": 0.1898,
      "step": 639
    },
    {
      "epoch": 2.857142857142857,
      "grad_norm": 1.5042264461517334,
      "learning_rate": 5.460412240497993e-06,
      "loss": 0.1964,
      "step": 640
    },
    {
      "epoch": 2.861607142857143,
      "grad_norm": 1.5381810665130615,
      "learning_rate": 5.448651724862716e-06,
      "loss": 0.185,
      "step": 641
    },
    {
      "epoch": 2.866071428571429,
      "grad_norm": 1.4257649183273315,
      "learning_rate": 5.436888708759253e-06,
      "loss": 0.189,
      "step": 642
    },
    {
      "epoch": 2.8705357142857144,
      "grad_norm": 1.5205702781677246,
      "learning_rate": 5.425123257819494e-06,
      "loss": 0.2022,
      "step": 643
    },
    {
      "epoch": 2.875,
      "grad_norm": 1.4278076887130737,
      "learning_rate": 5.413355437688926e-06,
      "loss": 0.1724,
      "step": 644
    },
    {
      "epoch": 2.8794642857142856,
      "grad_norm": 1.529142141342163,
      "learning_rate": 5.401585314026248e-06,
      "loss": 0.1946,
      "step": 645
    },
    {
      "epoch": 2.883928571428571,
      "grad_norm": 1.5609850883483887,
      "learning_rate": 5.3898129525030105e-06,
      "loss": 0.2121,
      "step": 646
    },
    {
      "epoch": 2.888392857142857,
      "grad_norm": 1.690317153930664,
      "learning_rate": 5.378038418803256e-06,
      "loss": 0.2023,
      "step": 647
    },
    {
      "epoch": 2.892857142857143,
      "grad_norm": 1.5702646970748901,
      "learning_rate": 5.366261778623143e-06,
      "loss": 0.2036,
      "step": 648
    },
    {
      "epoch": 2.897321428571429,
      "grad_norm": 1.6084643602371216,
      "learning_rate": 5.354483097670584e-06,
      "loss": 0.2009,
      "step": 649
    },
    {
      "epoch": 2.9017857142857144,
      "grad_norm": 1.7028254270553589,
      "learning_rate": 5.342702441664875e-06,
      "loss": 0.2093,
      "step": 650
    },
    {
      "epoch": 2.90625,
      "grad_norm": 1.5291582345962524,
      "learning_rate": 5.3309198763363345e-06,
      "loss": 0.2031,
      "step": 651
    },
    {
      "epoch": 2.9107142857142856,
      "grad_norm": 1.5348578691482544,
      "learning_rate": 5.319135467425937e-06,
      "loss": 0.199,
      "step": 652
    },
    {
      "epoch": 2.915178571428571,
      "grad_norm": 1.3724780082702637,
      "learning_rate": 5.3073492806849405e-06,
      "loss": 0.1765,
      "step": 653
    },
    {
      "epoch": 2.919642857142857,
      "grad_norm": 1.4268510341644287,
      "learning_rate": 5.295561381874518e-06,
      "loss": 0.17,
      "step": 654
    },
    {
      "epoch": 2.924107142857143,
      "grad_norm": 1.462509036064148,
      "learning_rate": 5.2837718367654036e-06,
      "loss": 0.1968,
      "step": 655
    },
    {
      "epoch": 2.928571428571429,
      "grad_norm": 1.5478267669677734,
      "learning_rate": 5.2719807111375096e-06,
      "loss": 0.2011,
      "step": 656
    },
    {
      "epoch": 2.9330357142857144,
      "grad_norm": 1.6338170766830444,
      "learning_rate": 5.260188070779573e-06,
      "loss": 0.1974,
      "step": 657
    },
    {
      "epoch": 2.9375,
      "grad_norm": 1.4555277824401855,
      "learning_rate": 5.248393981488777e-06,
      "loss": 0.1859,
      "step": 658
    },
    {
      "epoch": 2.9419642857142856,
      "grad_norm": 1.3970043659210205,
      "learning_rate": 5.236598509070389e-06,
      "loss": 0.1768,
      "step": 659
    },
    {
      "epoch": 2.946428571428571,
      "grad_norm": 1.533681869506836,
      "learning_rate": 5.2248017193374e-06,
      "loss": 0.2143,
      "step": 660
    },
    {
      "epoch": 2.950892857142857,
      "grad_norm": 1.6132566928863525,
      "learning_rate": 5.2130036781101455e-06,
      "loss": 0.1895,
      "step": 661
    },
    {
      "epoch": 2.955357142857143,
      "grad_norm": 1.5554327964782715,
      "learning_rate": 5.201204451215945e-06,
      "loss": 0.1883,
      "step": 662
    },
    {
      "epoch": 2.959821428571429,
      "grad_norm": 1.3858691453933716,
      "learning_rate": 5.18940410448873e-06,
      "loss": 0.1775,
      "step": 663
    },
    {
      "epoch": 2.9642857142857144,
      "grad_norm": 1.49596107006073,
      "learning_rate": 5.1776027037686895e-06,
      "loss": 0.1754,
      "step": 664
    },
    {
      "epoch": 2.96875,
      "grad_norm": 1.5804927349090576,
      "learning_rate": 5.165800314901883e-06,
      "loss": 0.1922,
      "step": 665
    },
    {
      "epoch": 2.9732142857142856,
      "grad_norm": 1.5328700542449951,
      "learning_rate": 5.15399700373989e-06,
      "loss": 0.1994,
      "step": 666
    },
    {
      "epoch": 2.977678571428571,
      "grad_norm": 1.593167781829834,
      "learning_rate": 5.142192836139432e-06,
      "loss": 0.1943,
      "step": 667
    },
    {
      "epoch": 2.982142857142857,
      "grad_norm": 1.4839582443237305,
      "learning_rate": 5.130387877962012e-06,
      "loss": 0.1697,
      "step": 668
    },
    {
      "epoch": 2.986607142857143,
      "grad_norm": 1.5925782918930054,
      "learning_rate": 5.118582195073542e-06,
      "loss": 0.1986,
      "step": 669
    },
    {
      "epoch": 2.991071428571429,
      "grad_norm": 1.4743068218231201,
      "learning_rate": 5.1067758533439804e-06,
      "loss": 0.1959,
      "step": 670
    },
    {
      "epoch": 2.9955357142857144,
      "grad_norm": 1.3711611032485962,
      "learning_rate": 5.094968918646954e-06,
      "loss": 0.1737,
      "step": 671
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.3573774099349976,
      "learning_rate": 5.0831614568594105e-06,
      "loss": 0.1787,
      "step": 672
    },
    {
      "epoch": 3.0044642857142856,
      "grad_norm": 1.1417394876480103,
      "learning_rate": 5.071353533861225e-06,
      "loss": 0.114,
      "step": 673
    },
    {
      "epoch": 3.0089285714285716,
      "grad_norm": 1.105907678604126,
      "learning_rate": 5.059545215534859e-06,
      "loss": 0.1086,
      "step": 674
    },
    {
      "epoch": 3.013392857142857,
      "grad_norm": 1.189517617225647,
      "learning_rate": 5.047736567764967e-06,
      "loss": 0.1138,
      "step": 675
    },
    {
      "epoch": 3.017857142857143,
      "grad_norm": 1.0811058282852173,
      "learning_rate": 5.0359276564380514e-06,
      "loss": 0.1067,
      "step": 676
    },
    {
      "epoch": 3.0223214285714284,
      "grad_norm": 1.2044225931167603,
      "learning_rate": 5.024118547442083e-06,
      "loss": 0.0996,
      "step": 677
    },
    {
      "epoch": 3.0267857142857144,
      "grad_norm": 1.1959599256515503,
      "learning_rate": 5.012309306666129e-06,
      "loss": 0.0986,
      "step": 678
    },
    {
      "epoch": 3.03125,
      "grad_norm": 1.1417187452316284,
      "learning_rate": 5.000500000000001e-06,
      "loss": 0.0816,
      "step": 679
    },
    {
      "epoch": 3.0357142857142856,
      "grad_norm": 1.2518037557601929,
      "learning_rate": 4.988690693333873e-06,
      "loss": 0.0855,
      "step": 680
    },
    {
      "epoch": 3.0401785714285716,
      "grad_norm": 1.4060600996017456,
      "learning_rate": 4.97688145255792e-06,
      "loss": 0.0997,
      "step": 681
    },
    {
      "epoch": 3.044642857142857,
      "grad_norm": 1.3264249563217163,
      "learning_rate": 4.965072343561948e-06,
      "loss": 0.0907,
      "step": 682
    },
    {
      "epoch": 3.049107142857143,
      "grad_norm": 1.5176544189453125,
      "learning_rate": 4.953263432235034e-06,
      "loss": 0.0882,
      "step": 683
    },
    {
      "epoch": 3.0535714285714284,
      "grad_norm": 1.6541138887405396,
      "learning_rate": 4.941454784465144e-06,
      "loss": 0.0876,
      "step": 684
    },
    {
      "epoch": 3.0580357142857144,
      "grad_norm": 1.5412651300430298,
      "learning_rate": 4.929646466138777e-06,
      "loss": 0.0923,
      "step": 685
    },
    {
      "epoch": 3.0625,
      "grad_norm": 1.4274966716766357,
      "learning_rate": 4.917838543140591e-06,
      "loss": 0.0732,
      "step": 686
    },
    {
      "epoch": 3.0669642857142856,
      "grad_norm": 1.5422570705413818,
      "learning_rate": 4.906031081353047e-06,
      "loss": 0.0882,
      "step": 687
    },
    {
      "epoch": 3.0714285714285716,
      "grad_norm": 1.7544273138046265,
      "learning_rate": 4.8942241466560226e-06,
      "loss": 0.0925,
      "step": 688
    },
    {
      "epoch": 3.075892857142857,
      "grad_norm": 1.6329741477966309,
      "learning_rate": 4.882417804926457e-06,
      "loss": 0.0908,
      "step": 689
    },
    {
      "epoch": 3.080357142857143,
      "grad_norm": 1.8030346632003784,
      "learning_rate": 4.870612122037989e-06,
      "loss": 0.1051,
      "step": 690
    },
    {
      "epoch": 3.0848214285714284,
      "grad_norm": 1.8434675931930542,
      "learning_rate": 4.858807163860569e-06,
      "loss": 0.1039,
      "step": 691
    },
    {
      "epoch": 3.0892857142857144,
      "grad_norm": 1.7246928215026855,
      "learning_rate": 4.847002996260113e-06,
      "loss": 0.0913,
      "step": 692
    },
    {
      "epoch": 3.09375,
      "grad_norm": 1.6196149587631226,
      "learning_rate": 4.835199685098117e-06,
      "loss": 0.1038,
      "step": 693
    },
    {
      "epoch": 3.0982142857142856,
      "grad_norm": 1.6847678422927856,
      "learning_rate": 4.823397296231313e-06,
      "loss": 0.0855,
      "step": 694
    },
    {
      "epoch": 3.1026785714285716,
      "grad_norm": 1.7434442043304443,
      "learning_rate": 4.8115958955112715e-06,
      "loss": 0.094,
      "step": 695
    },
    {
      "epoch": 3.107142857142857,
      "grad_norm": 1.6175901889801025,
      "learning_rate": 4.799795548784058e-06,
      "loss": 0.0908,
      "step": 696
    },
    {
      "epoch": 3.111607142857143,
      "grad_norm": 1.6020538806915283,
      "learning_rate": 4.787996321889856e-06,
      "loss": 0.0943,
      "step": 697
    },
    {
      "epoch": 3.1160714285714284,
      "grad_norm": 1.4542748928070068,
      "learning_rate": 4.7761982806626015e-06,
      "loss": 0.098,
      "step": 698
    },
    {
      "epoch": 3.1205357142857144,
      "grad_norm": 1.4568634033203125,
      "learning_rate": 4.764401490929613e-06,
      "loss": 0.0884,
      "step": 699
    },
    {
      "epoch": 3.125,
      "grad_norm": 1.4934730529785156,
      "learning_rate": 4.752606018511225e-06,
      "loss": 0.0941,
      "step": 700
    },
    {
      "epoch": 3.1294642857142856,
      "grad_norm": 1.4770476818084717,
      "learning_rate": 4.740811929220429e-06,
      "loss": 0.1118,
      "step": 701
    },
    {
      "epoch": 3.1339285714285716,
      "grad_norm": 1.416353464126587,
      "learning_rate": 4.729019288862492e-06,
      "loss": 0.0996,
      "step": 702
    },
    {
      "epoch": 3.138392857142857,
      "grad_norm": 1.4401273727416992,
      "learning_rate": 4.717228163234599e-06,
      "loss": 0.092,
      "step": 703
    },
    {
      "epoch": 3.142857142857143,
      "grad_norm": 1.1891121864318848,
      "learning_rate": 4.705438618125482e-06,
      "loss": 0.0741,
      "step": 704
    },
    {
      "epoch": 3.1473214285714284,
      "grad_norm": 1.386759877204895,
      "learning_rate": 4.693650719315062e-06,
      "loss": 0.0998,
      "step": 705
    },
    {
      "epoch": 3.1517857142857144,
      "grad_norm": 1.3926823139190674,
      "learning_rate": 4.681864532574064e-06,
      "loss": 0.1015,
      "step": 706
    },
    {
      "epoch": 3.15625,
      "grad_norm": 1.3910696506500244,
      "learning_rate": 4.670080123663668e-06,
      "loss": 0.1002,
      "step": 707
    },
    {
      "epoch": 3.1607142857142856,
      "grad_norm": 1.2177248001098633,
      "learning_rate": 4.658297558335127e-06,
      "loss": 0.0903,
      "step": 708
    },
    {
      "epoch": 3.1651785714285716,
      "grad_norm": 1.434577465057373,
      "learning_rate": 4.64651690232942e-06,
      "loss": 0.0956,
      "step": 709
    },
    {
      "epoch": 3.169642857142857,
      "grad_norm": 1.2175168991088867,
      "learning_rate": 4.634738221376858e-06,
      "loss": 0.0805,
      "step": 710
    },
    {
      "epoch": 3.174107142857143,
      "grad_norm": 1.2530947923660278,
      "learning_rate": 4.622961581196743e-06,
      "loss": 0.0804,
      "step": 711
    },
    {
      "epoch": 3.1785714285714284,
      "grad_norm": 1.273105502128601,
      "learning_rate": 4.611187047496989e-06,
      "loss": 0.0912,
      "step": 712
    },
    {
      "epoch": 3.1830357142857144,
      "grad_norm": 1.1457451581954956,
      "learning_rate": 4.599414685973754e-06,
      "loss": 0.0794,
      "step": 713
    },
    {
      "epoch": 3.1875,
      "grad_norm": 1.20325767993927,
      "learning_rate": 4.587644562311077e-06,
      "loss": 0.0809,
      "step": 714
    },
    {
      "epoch": 3.1919642857142856,
      "grad_norm": 1.5332928895950317,
      "learning_rate": 4.575876742180506e-06,
      "loss": 0.0959,
      "step": 715
    },
    {
      "epoch": 3.1964285714285716,
      "grad_norm": 1.1909514665603638,
      "learning_rate": 4.56411129124075e-06,
      "loss": 0.0843,
      "step": 716
    },
    {
      "epoch": 3.200892857142857,
      "grad_norm": 1.472267508506775,
      "learning_rate": 4.552348275137285e-06,
      "loss": 0.1156,
      "step": 717
    },
    {
      "epoch": 3.205357142857143,
      "grad_norm": 1.532912254333496,
      "learning_rate": 4.54058775950201e-06,
      "loss": 0.0935,
      "step": 718
    },
    {
      "epoch": 3.2098214285714284,
      "grad_norm": 1.4942517280578613,
      "learning_rate": 4.528829809952867e-06,
      "loss": 0.0945,
      "step": 719
    },
    {
      "epoch": 3.2142857142857144,
      "grad_norm": 1.5307639837265015,
      "learning_rate": 4.517074492093487e-06,
      "loss": 0.1106,
      "step": 720
    },
    {
      "epoch": 3.21875,
      "grad_norm": 1.477367639541626,
      "learning_rate": 4.505321871512813e-06,
      "loss": 0.1024,
      "step": 721
    },
    {
      "epoch": 3.2232142857142856,
      "grad_norm": 1.6967626810073853,
      "learning_rate": 4.493572013784737e-06,
      "loss": 0.1039,
      "step": 722
    },
    {
      "epoch": 3.2276785714285716,
      "grad_norm": 1.4869396686553955,
      "learning_rate": 4.481824984467742e-06,
      "loss": 0.0929,
      "step": 723
    },
    {
      "epoch": 3.232142857142857,
      "grad_norm": 1.466221570968628,
      "learning_rate": 4.470080849104521e-06,
      "loss": 0.1008,
      "step": 724
    },
    {
      "epoch": 3.236607142857143,
      "grad_norm": 1.7747437953948975,
      "learning_rate": 4.458339673221631e-06,
      "loss": 0.1099,
      "step": 725
    },
    {
      "epoch": 3.2410714285714284,
      "grad_norm": 1.5948566198349,
      "learning_rate": 4.446601522329105e-06,
      "loss": 0.0876,
      "step": 726
    },
    {
      "epoch": 3.2455357142857144,
      "grad_norm": 1.3861619234085083,
      "learning_rate": 4.434866461920108e-06,
      "loss": 0.093,
      "step": 727
    },
    {
      "epoch": 3.25,
      "grad_norm": 1.7178670167922974,
      "learning_rate": 4.4231345574705555e-06,
      "loss": 0.1074,
      "step": 728
    },
    {
      "epoch": 3.2544642857142856,
      "grad_norm": 1.6227957010269165,
      "learning_rate": 4.4114058744387535e-06,
      "loss": 0.0984,
      "step": 729
    },
    {
      "epoch": 3.2589285714285716,
      "grad_norm": 1.5968222618103027,
      "learning_rate": 4.399680478265042e-06,
      "loss": 0.1015,
      "step": 730
    },
    {
      "epoch": 3.263392857142857,
      "grad_norm": 1.6373909711837769,
      "learning_rate": 4.387958434371413e-06,
      "loss": 0.099,
      "step": 731
    },
    {
      "epoch": 3.267857142857143,
      "grad_norm": 1.3308444023132324,
      "learning_rate": 4.376239808161157e-06,
      "loss": 0.0819,
      "step": 732
    },
    {
      "epoch": 3.2723214285714284,
      "grad_norm": 1.4825462102890015,
      "learning_rate": 4.364524665018496e-06,
      "loss": 0.0872,
      "step": 733
    },
    {
      "epoch": 3.2767857142857144,
      "grad_norm": 1.5847879648208618,
      "learning_rate": 4.3528130703082245e-06,
      "loss": 0.0873,
      "step": 734
    },
    {
      "epoch": 3.28125,
      "grad_norm": 1.3668209314346313,
      "learning_rate": 4.341105089375328e-06,
      "loss": 0.0843,
      "step": 735
    },
    {
      "epoch": 3.2857142857142856,
      "grad_norm": 1.4361255168914795,
      "learning_rate": 4.329400787544633e-06,
      "loss": 0.093,
      "step": 736
    },
    {
      "epoch": 3.2901785714285716,
      "grad_norm": 1.2126491069793701,
      "learning_rate": 4.317700230120438e-06,
      "loss": 0.0825,
      "step": 737
    },
    {
      "epoch": 3.294642857142857,
      "grad_norm": 1.4545817375183105,
      "learning_rate": 4.306003482386156e-06,
      "loss": 0.1061,
      "step": 738
    },
    {
      "epoch": 3.299107142857143,
      "grad_norm": 1.3296757936477661,
      "learning_rate": 4.2943106096039315e-06,
      "loss": 0.0888,
      "step": 739
    },
    {
      "epoch": 3.3035714285714284,
      "grad_norm": 1.4436862468719482,
      "learning_rate": 4.282621677014299e-06,
      "loss": 0.0955,
      "step": 740
    },
    {
      "epoch": 3.3080357142857144,
      "grad_norm": 1.313293218612671,
      "learning_rate": 4.270936749835805e-06,
      "loss": 0.0979,
      "step": 741
    },
    {
      "epoch": 3.3125,
      "grad_norm": 1.2750566005706787,
      "learning_rate": 4.259255893264647e-06,
      "loss": 0.0907,
      "step": 742
    },
    {
      "epoch": 3.3169642857142856,
      "grad_norm": 1.6833667755126953,
      "learning_rate": 4.247579172474312e-06,
      "loss": 0.1025,
      "step": 743
    },
    {
      "epoch": 3.3214285714285716,
      "grad_norm": 1.4311575889587402,
      "learning_rate": 4.235906652615207e-06,
      "loss": 0.1006,
      "step": 744
    },
    {
      "epoch": 3.325892857142857,
      "grad_norm": 1.5903573036193848,
      "learning_rate": 4.224238398814309e-06,
      "loss": 0.0953,
      "step": 745
    },
    {
      "epoch": 3.330357142857143,
      "grad_norm": 1.3978747129440308,
      "learning_rate": 4.212574476174784e-06,
      "loss": 0.095,
      "step": 746
    },
    {
      "epoch": 3.3348214285714284,
      "grad_norm": 1.4329924583435059,
      "learning_rate": 4.2009149497756355e-06,
      "loss": 0.1007,
      "step": 747
    },
    {
      "epoch": 3.3392857142857144,
      "grad_norm": 1.2830747365951538,
      "learning_rate": 4.189259884671336e-06,
      "loss": 0.0862,
      "step": 748
    },
    {
      "epoch": 3.34375,
      "grad_norm": 1.3609569072723389,
      "learning_rate": 4.177609345891472e-06,
      "loss": 0.0875,
      "step": 749
    },
    {
      "epoch": 3.3482142857142856,
      "grad_norm": 1.3331035375595093,
      "learning_rate": 4.165963398440368e-06,
      "loss": 0.1016,
      "step": 750
    },
    {
      "epoch": 3.3526785714285716,
      "grad_norm": 1.4151555299758911,
      "learning_rate": 4.1543221072967334e-06,
      "loss": 0.0975,
      "step": 751
    },
    {
      "epoch": 3.357142857142857,
      "grad_norm": 1.490128755569458,
      "learning_rate": 4.142685537413303e-06,
      "loss": 0.0915,
      "step": 752
    },
    {
      "epoch": 3.361607142857143,
      "grad_norm": 1.3559775352478027,
      "learning_rate": 4.1310537537164615e-06,
      "loss": 0.0979,
      "step": 753
    },
    {
      "epoch": 3.3660714285714284,
      "grad_norm": 1.4377248287200928,
      "learning_rate": 4.119426821105895e-06,
      "loss": 0.0935,
      "step": 754
    },
    {
      "epoch": 3.3705357142857144,
      "grad_norm": 1.4320459365844727,
      "learning_rate": 4.107804804454215e-06,
      "loss": 0.0976,
      "step": 755
    },
    {
      "epoch": 3.375,
      "grad_norm": 1.473273754119873,
      "learning_rate": 4.096187768606617e-06,
      "loss": 0.0949,
      "step": 756
    },
    {
      "epoch": 3.3794642857142856,
      "grad_norm": 1.240604043006897,
      "learning_rate": 4.084575778380495e-06,
      "loss": 0.077,
      "step": 757
    },
    {
      "epoch": 3.3839285714285716,
      "grad_norm": 1.6010382175445557,
      "learning_rate": 4.072968898565094e-06,
      "loss": 0.0987,
      "step": 758
    },
    {
      "epoch": 3.388392857142857,
      "grad_norm": 1.5959616899490356,
      "learning_rate": 4.061367193921145e-06,
      "loss": 0.1125,
      "step": 759
    },
    {
      "epoch": 3.392857142857143,
      "grad_norm": 1.5466870069503784,
      "learning_rate": 4.049770729180508e-06,
      "loss": 0.0985,
      "step": 760
    },
    {
      "epoch": 3.3973214285714284,
      "grad_norm": 1.485848307609558,
      "learning_rate": 4.038179569045803e-06,
      "loss": 0.0942,
      "step": 761
    },
    {
      "epoch": 3.4017857142857144,
      "grad_norm": 1.4044053554534912,
      "learning_rate": 4.026593778190052e-06,
      "loss": 0.096,
      "step": 762
    },
    {
      "epoch": 3.40625,
      "grad_norm": 1.2948501110076904,
      "learning_rate": 4.015013421256324e-06,
      "loss": 0.0852,
      "step": 763
    },
    {
      "epoch": 3.4107142857142856,
      "grad_norm": 1.2491077184677124,
      "learning_rate": 4.0034385628573655e-06,
      "loss": 0.0814,
      "step": 764
    },
    {
      "epoch": 3.4151785714285716,
      "grad_norm": 1.4824570417404175,
      "learning_rate": 3.991869267575243e-06,
      "loss": 0.0848,
      "step": 765
    },
    {
      "epoch": 3.419642857142857,
      "grad_norm": 1.6389641761779785,
      "learning_rate": 3.9803055999609855e-06,
      "loss": 0.0929,
      "step": 766
    },
    {
      "epoch": 3.424107142857143,
      "grad_norm": 1.5796787738800049,
      "learning_rate": 3.9687476245342234e-06,
      "loss": 0.1006,
      "step": 767
    },
    {
      "epoch": 3.4285714285714284,
      "grad_norm": 1.4374067783355713,
      "learning_rate": 3.957195405782824e-06,
      "loss": 0.0938,
      "step": 768
    },
    {
      "epoch": 3.4330357142857144,
      "grad_norm": 1.5482842922210693,
      "learning_rate": 3.9456490081625396e-06,
      "loss": 0.1003,
      "step": 769
    },
    {
      "epoch": 3.4375,
      "grad_norm": 1.366696834564209,
      "learning_rate": 3.934108496096638e-06,
      "loss": 0.0896,
      "step": 770
    },
    {
      "epoch": 3.4419642857142856,
      "grad_norm": 1.4534419775009155,
      "learning_rate": 3.922573933975555e-06,
      "loss": 0.0954,
      "step": 771
    },
    {
      "epoch": 3.4464285714285716,
      "grad_norm": 1.3738148212432861,
      "learning_rate": 3.911045386156523e-06,
      "loss": 0.0986,
      "step": 772
    },
    {
      "epoch": 3.450892857142857,
      "grad_norm": 1.4549438953399658,
      "learning_rate": 3.899522916963219e-06,
      "loss": 0.0937,
      "step": 773
    },
    {
      "epoch": 3.455357142857143,
      "grad_norm": 1.5725458860397339,
      "learning_rate": 3.888006590685407e-06,
      "loss": 0.0884,
      "step": 774
    },
    {
      "epoch": 3.4598214285714284,
      "grad_norm": 1.409229040145874,
      "learning_rate": 3.876496471578572e-06,
      "loss": 0.0985,
      "step": 775
    },
    {
      "epoch": 3.4642857142857144,
      "grad_norm": 1.4481651782989502,
      "learning_rate": 3.864992623863572e-06,
      "loss": 0.0993,
      "step": 776
    },
    {
      "epoch": 3.46875,
      "grad_norm": 1.3321104049682617,
      "learning_rate": 3.853495111726265e-06,
      "loss": 0.0938,
      "step": 777
    },
    {
      "epoch": 3.4732142857142856,
      "grad_norm": 1.501019835472107,
      "learning_rate": 3.84200399931717e-06,
      "loss": 0.0955,
      "step": 778
    },
    {
      "epoch": 3.4776785714285716,
      "grad_norm": 1.3454251289367676,
      "learning_rate": 3.8305193507510905e-06,
      "loss": 0.0955,
      "step": 779
    },
    {
      "epoch": 3.482142857142857,
      "grad_norm": 1.5015486478805542,
      "learning_rate": 3.819041230106768e-06,
      "loss": 0.1041,
      "step": 780
    },
    {
      "epoch": 3.486607142857143,
      "grad_norm": 1.1266005039215088,
      "learning_rate": 3.807569701426519e-06,
      "loss": 0.0815,
      "step": 781
    },
    {
      "epoch": 3.4910714285714284,
      "grad_norm": 1.322966456413269,
      "learning_rate": 3.7961048287158865e-06,
      "loss": 0.088,
      "step": 782
    },
    {
      "epoch": 3.4955357142857144,
      "grad_norm": 1.31028151512146,
      "learning_rate": 3.784646675943269e-06,
      "loss": 0.0905,
      "step": 783
    },
    {
      "epoch": 3.5,
      "grad_norm": 1.4570220708847046,
      "learning_rate": 3.773195307039574e-06,
      "loss": 0.0966,
      "step": 784
    },
    {
      "epoch": 3.5044642857142856,
      "grad_norm": 1.4365588426589966,
      "learning_rate": 3.7617507858978615e-06,
      "loss": 0.0924,
      "step": 785
    },
    {
      "epoch": 3.508928571428571,
      "grad_norm": 1.2722663879394531,
      "learning_rate": 3.7503131763729785e-06,
      "loss": 0.0861,
      "step": 786
    },
    {
      "epoch": 3.513392857142857,
      "grad_norm": 1.4443727731704712,
      "learning_rate": 3.738882542281212e-06,
      "loss": 0.0943,
      "step": 787
    },
    {
      "epoch": 3.517857142857143,
      "grad_norm": 1.4455920457839966,
      "learning_rate": 3.727458947399927e-06,
      "loss": 0.0924,
      "step": 788
    },
    {
      "epoch": 3.522321428571429,
      "grad_norm": 1.2739661931991577,
      "learning_rate": 3.7160424554672187e-06,
      "loss": 0.0938,
      "step": 789
    },
    {
      "epoch": 3.5267857142857144,
      "grad_norm": 1.3610420227050781,
      "learning_rate": 3.7046331301815435e-06,
      "loss": 0.0928,
      "step": 790
    },
    {
      "epoch": 3.53125,
      "grad_norm": 1.360996127128601,
      "learning_rate": 3.6932310352013796e-06,
      "loss": 0.0848,
      "step": 791
    },
    {
      "epoch": 3.5357142857142856,
      "grad_norm": 1.3822091817855835,
      "learning_rate": 3.6818362341448545e-06,
      "loss": 0.0937,
      "step": 792
    },
    {
      "epoch": 3.540178571428571,
      "grad_norm": 1.5511473417282104,
      "learning_rate": 3.670448790589408e-06,
      "loss": 0.0964,
      "step": 793
    },
    {
      "epoch": 3.544642857142857,
      "grad_norm": 1.322980284690857,
      "learning_rate": 3.659068768071425e-06,
      "loss": 0.0828,
      "step": 794
    },
    {
      "epoch": 3.549107142857143,
      "grad_norm": 1.4569833278656006,
      "learning_rate": 3.6476962300858793e-06,
      "loss": 0.0982,
      "step": 795
    },
    {
      "epoch": 3.553571428571429,
      "grad_norm": 1.2515519857406616,
      "learning_rate": 3.6363312400859963e-06,
      "loss": 0.0874,
      "step": 796
    },
    {
      "epoch": 3.5580357142857144,
      "grad_norm": 1.5618796348571777,
      "learning_rate": 3.6249738614828765e-06,
      "loss": 0.0951,
      "step": 797
    },
    {
      "epoch": 3.5625,
      "grad_norm": 1.530705213546753,
      "learning_rate": 3.613624157645159e-06,
      "loss": 0.0983,
      "step": 798
    },
    {
      "epoch": 3.5669642857142856,
      "grad_norm": 1.6114871501922607,
      "learning_rate": 3.6022821918986563e-06,
      "loss": 0.0946,
      "step": 799
    },
    {
      "epoch": 3.571428571428571,
      "grad_norm": 1.43599271774292,
      "learning_rate": 3.590948027526012e-06,
      "loss": 0.0989,
      "step": 800
    },
    {
      "epoch": 3.571428571428571,
      "eval_loss": 0.843691885471344,
      "eval_runtime": 4.4852,
      "eval_samples_per_second": 13.377,
      "eval_steps_per_second": 0.892,
      "step": 800
    },
    {
      "epoch": 3.575892857142857,
      "grad_norm": 1.2312313318252563,
      "learning_rate": 3.579621727766339e-06,
      "loss": 0.079,
      "step": 801
    },
    {
      "epoch": 3.580357142857143,
      "grad_norm": 1.3309111595153809,
      "learning_rate": 3.568303355814867e-06,
      "loss": 0.0926,
      "step": 802
    },
    {
      "epoch": 3.584821428571429,
      "grad_norm": 1.293211817741394,
      "learning_rate": 3.5569929748225945e-06,
      "loss": 0.0879,
      "step": 803
    },
    {
      "epoch": 3.5892857142857144,
      "grad_norm": 1.532378911972046,
      "learning_rate": 3.5456906478959367e-06,
      "loss": 0.1051,
      "step": 804
    },
    {
      "epoch": 3.59375,
      "grad_norm": 1.4125474691390991,
      "learning_rate": 3.534396438096369e-06,
      "loss": 0.0953,
      "step": 805
    },
    {
      "epoch": 3.5982142857142856,
      "grad_norm": 1.3551851511001587,
      "learning_rate": 3.5231104084400745e-06,
      "loss": 0.0889,
      "step": 806
    },
    {
      "epoch": 3.602678571428571,
      "grad_norm": 1.4057469367980957,
      "learning_rate": 3.5118326218976013e-06,
      "loss": 0.0955,
      "step": 807
    },
    {
      "epoch": 3.607142857142857,
      "grad_norm": 1.5146610736846924,
      "learning_rate": 3.5005631413935006e-06,
      "loss": 0.0932,
      "step": 808
    },
    {
      "epoch": 3.611607142857143,
      "grad_norm": 1.3674921989440918,
      "learning_rate": 3.4893020298059784e-06,
      "loss": 0.0831,
      "step": 809
    },
    {
      "epoch": 3.616071428571429,
      "grad_norm": 1.6528156995773315,
      "learning_rate": 3.4780493499665478e-06,
      "loss": 0.103,
      "step": 810
    },
    {
      "epoch": 3.6205357142857144,
      "grad_norm": 1.5097229480743408,
      "learning_rate": 3.4668051646596825e-06,
      "loss": 0.1007,
      "step": 811
    },
    {
      "epoch": 3.625,
      "grad_norm": 1.6443347930908203,
      "learning_rate": 3.455569536622451e-06,
      "loss": 0.0968,
      "step": 812
    },
    {
      "epoch": 3.6294642857142856,
      "grad_norm": 1.417252779006958,
      "learning_rate": 3.4443425285441847e-06,
      "loss": 0.0944,
      "step": 813
    },
    {
      "epoch": 3.633928571428571,
      "grad_norm": 1.4208124876022339,
      "learning_rate": 3.433124203066113e-06,
      "loss": 0.0938,
      "step": 814
    },
    {
      "epoch": 3.638392857142857,
      "grad_norm": 1.4724245071411133,
      "learning_rate": 3.421914622781028e-06,
      "loss": 0.1021,
      "step": 815
    },
    {
      "epoch": 3.642857142857143,
      "grad_norm": 1.5548927783966064,
      "learning_rate": 3.4107138502329225e-06,
      "loss": 0.1037,
      "step": 816
    },
    {
      "epoch": 3.647321428571429,
      "grad_norm": 1.5089542865753174,
      "learning_rate": 3.399521947916646e-06,
      "loss": 0.1108,
      "step": 817
    },
    {
      "epoch": 3.6517857142857144,
      "grad_norm": 1.24465012550354,
      "learning_rate": 3.3883389782775604e-06,
      "loss": 0.0848,
      "step": 818
    },
    {
      "epoch": 3.65625,
      "grad_norm": 1.3363871574401855,
      "learning_rate": 3.377165003711185e-06,
      "loss": 0.0811,
      "step": 819
    },
    {
      "epoch": 3.6607142857142856,
      "grad_norm": 1.3324394226074219,
      "learning_rate": 3.3660000865628523e-06,
      "loss": 0.0767,
      "step": 820
    },
    {
      "epoch": 3.665178571428571,
      "grad_norm": 1.3909485340118408,
      "learning_rate": 3.3548442891273553e-06,
      "loss": 0.0924,
      "step": 821
    },
    {
      "epoch": 3.669642857142857,
      "grad_norm": 1.3440436124801636,
      "learning_rate": 3.343697673648611e-06,
      "loss": 0.0909,
      "step": 822
    },
    {
      "epoch": 3.674107142857143,
      "grad_norm": 1.4336508512496948,
      "learning_rate": 3.332560302319297e-06,
      "loss": 0.1151,
      "step": 823
    },
    {
      "epoch": 3.678571428571429,
      "grad_norm": 1.3243669271469116,
      "learning_rate": 3.321432237280518e-06,
      "loss": 0.0892,
      "step": 824
    },
    {
      "epoch": 3.6830357142857144,
      "grad_norm": 1.288888692855835,
      "learning_rate": 3.3103135406214506e-06,
      "loss": 0.0858,
      "step": 825
    },
    {
      "epoch": 3.6875,
      "grad_norm": 1.386723279953003,
      "learning_rate": 3.2992042743790055e-06,
      "loss": 0.0881,
      "step": 826
    },
    {
      "epoch": 3.6919642857142856,
      "grad_norm": 1.3573461771011353,
      "learning_rate": 3.2881045005374747e-06,
      "loss": 0.087,
      "step": 827
    },
    {
      "epoch": 3.696428571428571,
      "grad_norm": 1.379010558128357,
      "learning_rate": 3.277014281028181e-06,
      "loss": 0.0835,
      "step": 828
    },
    {
      "epoch": 3.700892857142857,
      "grad_norm": 1.4972076416015625,
      "learning_rate": 3.2659336777291497e-06,
      "loss": 0.1027,
      "step": 829
    },
    {
      "epoch": 3.705357142857143,
      "grad_norm": 1.3756992816925049,
      "learning_rate": 3.254862752464743e-06,
      "loss": 0.0912,
      "step": 830
    },
    {
      "epoch": 3.709821428571429,
      "grad_norm": 1.473894476890564,
      "learning_rate": 3.243801567005329e-06,
      "loss": 0.1016,
      "step": 831
    },
    {
      "epoch": 3.7142857142857144,
      "grad_norm": 1.4770474433898926,
      "learning_rate": 3.232750183066928e-06,
      "loss": 0.0974,
      "step": 832
    },
    {
      "epoch": 3.71875,
      "grad_norm": 1.5533777475357056,
      "learning_rate": 3.2217086623108796e-06,
      "loss": 0.0994,
      "step": 833
    },
    {
      "epoch": 3.7232142857142856,
      "grad_norm": 1.3398988246917725,
      "learning_rate": 3.2106770663434867e-06,
      "loss": 0.0838,
      "step": 834
    },
    {
      "epoch": 3.727678571428571,
      "grad_norm": 1.4815514087677002,
      "learning_rate": 3.1996554567156774e-06,
      "loss": 0.0977,
      "step": 835
    },
    {
      "epoch": 3.732142857142857,
      "grad_norm": 1.1235636472702026,
      "learning_rate": 3.18864389492266e-06,
      "loss": 0.0828,
      "step": 836
    },
    {
      "epoch": 3.736607142857143,
      "grad_norm": 1.524712085723877,
      "learning_rate": 3.1776424424035857e-06,
      "loss": 0.1056,
      "step": 837
    },
    {
      "epoch": 3.741071428571429,
      "grad_norm": 1.3527683019638062,
      "learning_rate": 3.1666511605411947e-06,
      "loss": 0.0901,
      "step": 838
    },
    {
      "epoch": 3.7455357142857144,
      "grad_norm": 1.340248703956604,
      "learning_rate": 3.155670110661482e-06,
      "loss": 0.0885,
      "step": 839
    },
    {
      "epoch": 3.75,
      "grad_norm": 1.3833794593811035,
      "learning_rate": 3.144699354033356e-06,
      "loss": 0.086,
      "step": 840
    },
    {
      "epoch": 3.7544642857142856,
      "grad_norm": 1.4830819368362427,
      "learning_rate": 3.1337389518682894e-06,
      "loss": 0.0985,
      "step": 841
    },
    {
      "epoch": 3.758928571428571,
      "grad_norm": 1.5664781332015991,
      "learning_rate": 3.122788965319985e-06,
      "loss": 0.0789,
      "step": 842
    },
    {
      "epoch": 3.763392857142857,
      "grad_norm": 1.232924222946167,
      "learning_rate": 3.1118494554840284e-06,
      "loss": 0.0832,
      "step": 843
    },
    {
      "epoch": 3.767857142857143,
      "grad_norm": 1.2607110738754272,
      "learning_rate": 3.100920483397556e-06,
      "loss": 0.0773,
      "step": 844
    },
    {
      "epoch": 3.772321428571429,
      "grad_norm": 1.5223504304885864,
      "learning_rate": 3.090002110038903e-06,
      "loss": 0.0961,
      "step": 845
    },
    {
      "epoch": 3.7767857142857144,
      "grad_norm": 1.514897108078003,
      "learning_rate": 3.0790943963272697e-06,
      "loss": 0.0978,
      "step": 846
    },
    {
      "epoch": 3.78125,
      "grad_norm": 1.4400715827941895,
      "learning_rate": 3.0681974031223854e-06,
      "loss": 0.0971,
      "step": 847
    },
    {
      "epoch": 3.7857142857142856,
      "grad_norm": 1.479879379272461,
      "learning_rate": 3.0573111912241575e-06,
      "loss": 0.0862,
      "step": 848
    },
    {
      "epoch": 3.790178571428571,
      "grad_norm": 1.3477786779403687,
      "learning_rate": 3.0464358213723455e-06,
      "loss": 0.0891,
      "step": 849
    },
    {
      "epoch": 3.794642857142857,
      "grad_norm": 1.3734490871429443,
      "learning_rate": 3.0355713542462086e-06,
      "loss": 0.0995,
      "step": 850
    },
    {
      "epoch": 3.799107142857143,
      "grad_norm": 1.3643180131912231,
      "learning_rate": 3.024717850464181e-06,
      "loss": 0.0954,
      "step": 851
    },
    {
      "epoch": 3.803571428571429,
      "grad_norm": 1.619181752204895,
      "learning_rate": 3.0138753705835234e-06,
      "loss": 0.1008,
      "step": 852
    },
    {
      "epoch": 3.8080357142857144,
      "grad_norm": 1.4963918924331665,
      "learning_rate": 3.003043975099988e-06,
      "loss": 0.091,
      "step": 853
    },
    {
      "epoch": 3.8125,
      "grad_norm": 1.2831984758377075,
      "learning_rate": 2.99222372444748e-06,
      "loss": 0.0788,
      "step": 854
    },
    {
      "epoch": 3.8169642857142856,
      "grad_norm": 1.3784717321395874,
      "learning_rate": 2.9814146789977278e-06,
      "loss": 0.09,
      "step": 855
    },
    {
      "epoch": 3.821428571428571,
      "grad_norm": 1.6730880737304688,
      "learning_rate": 2.970616899059934e-06,
      "loss": 0.0965,
      "step": 856
    },
    {
      "epoch": 3.825892857142857,
      "grad_norm": 1.5248148441314697,
      "learning_rate": 2.959830444880447e-06,
      "loss": 0.0909,
      "step": 857
    },
    {
      "epoch": 3.830357142857143,
      "grad_norm": 1.5161914825439453,
      "learning_rate": 2.949055376642428e-06,
      "loss": 0.0933,
      "step": 858
    },
    {
      "epoch": 3.834821428571429,
      "grad_norm": 1.4900877475738525,
      "learning_rate": 2.9382917544655025e-06,
      "loss": 0.0886,
      "step": 859
    },
    {
      "epoch": 3.8392857142857144,
      "grad_norm": 1.3055517673492432,
      "learning_rate": 2.9275396384054373e-06,
      "loss": 0.083,
      "step": 860
    },
    {
      "epoch": 3.84375,
      "grad_norm": 1.5641956329345703,
      "learning_rate": 2.9167990884537943e-06,
      "loss": 0.097,
      "step": 861
    },
    {
      "epoch": 3.8482142857142856,
      "grad_norm": 1.521967887878418,
      "learning_rate": 2.906070164537616e-06,
      "loss": 0.1076,
      "step": 862
    },
    {
      "epoch": 3.852678571428571,
      "grad_norm": 1.445181131362915,
      "learning_rate": 2.8953529265190618e-06,
      "loss": 0.0904,
      "step": 863
    },
    {
      "epoch": 3.857142857142857,
      "grad_norm": 1.4985555410385132,
      "learning_rate": 2.884647434195099e-06,
      "loss": 0.0992,
      "step": 864
    },
    {
      "epoch": 3.861607142857143,
      "grad_norm": 1.4231829643249512,
      "learning_rate": 2.873953747297153e-06,
      "loss": 0.0799,
      "step": 865
    },
    {
      "epoch": 3.866071428571429,
      "grad_norm": 1.472952961921692,
      "learning_rate": 2.863271925490791e-06,
      "loss": 0.1004,
      "step": 866
    },
    {
      "epoch": 3.8705357142857144,
      "grad_norm": 1.5234057903289795,
      "learning_rate": 2.8526020283753658e-06,
      "loss": 0.0958,
      "step": 867
    },
    {
      "epoch": 3.875,
      "grad_norm": 1.4517230987548828,
      "learning_rate": 2.841944115483703e-06,
      "loss": 0.0918,
      "step": 868
    },
    {
      "epoch": 3.8794642857142856,
      "grad_norm": 1.4124717712402344,
      "learning_rate": 2.8312982462817686e-06,
      "loss": 0.0946,
      "step": 869
    },
    {
      "epoch": 3.883928571428571,
      "grad_norm": 1.6698354482650757,
      "learning_rate": 2.820664480168317e-06,
      "loss": 0.1049,
      "step": 870
    },
    {
      "epoch": 3.888392857142857,
      "grad_norm": 1.4935716390609741,
      "learning_rate": 2.810042876474584e-06,
      "loss": 0.1022,
      "step": 871
    },
    {
      "epoch": 3.892857142857143,
      "grad_norm": 1.5294708013534546,
      "learning_rate": 2.799433494463935e-06,
      "loss": 0.1049,
      "step": 872
    },
    {
      "epoch": 3.897321428571429,
      "grad_norm": 1.5460635423660278,
      "learning_rate": 2.7888363933315593e-06,
      "loss": 0.1035,
      "step": 873
    },
    {
      "epoch": 3.9017857142857144,
      "grad_norm": 1.3909894227981567,
      "learning_rate": 2.7782516322041087e-06,
      "loss": 0.084,
      "step": 874
    },
    {
      "epoch": 3.90625,
      "grad_norm": 1.3862639665603638,
      "learning_rate": 2.767679270139394e-06,
      "loss": 0.0936,
      "step": 875
    },
    {
      "epoch": 3.9107142857142856,
      "grad_norm": 1.602386713027954,
      "learning_rate": 2.7571193661260427e-06,
      "loss": 0.1008,
      "step": 876
    },
    {
      "epoch": 3.915178571428571,
      "grad_norm": 1.4371018409729004,
      "learning_rate": 2.746571979083172e-06,
      "loss": 0.0975,
      "step": 877
    },
    {
      "epoch": 3.919642857142857,
      "grad_norm": 1.3875517845153809,
      "learning_rate": 2.736037167860061e-06,
      "loss": 0.0999,
      "step": 878
    },
    {
      "epoch": 3.924107142857143,
      "grad_norm": 1.4412161111831665,
      "learning_rate": 2.725514991235818e-06,
      "loss": 0.099,
      "step": 879
    },
    {
      "epoch": 3.928571428571429,
      "grad_norm": 1.3014588356018066,
      "learning_rate": 2.7150055079190663e-06,
      "loss": 0.0833,
      "step": 880
    },
    {
      "epoch": 3.9330357142857144,
      "grad_norm": 1.4020464420318604,
      "learning_rate": 2.704508776547595e-06,
      "loss": 0.093,
      "step": 881
    },
    {
      "epoch": 3.9375,
      "grad_norm": 1.3537229299545288,
      "learning_rate": 2.6940248556880512e-06,
      "loss": 0.0858,
      "step": 882
    },
    {
      "epoch": 3.9419642857142856,
      "grad_norm": 1.2978018522262573,
      "learning_rate": 2.6835538038356017e-06,
      "loss": 0.0779,
      "step": 883
    },
    {
      "epoch": 3.946428571428571,
      "grad_norm": 1.2745367288589478,
      "learning_rate": 2.6730956794136138e-06,
      "loss": 0.0867,
      "step": 884
    },
    {
      "epoch": 3.950892857142857,
      "grad_norm": 1.3864595890045166,
      "learning_rate": 2.6626505407733255e-06,
      "loss": 0.0926,
      "step": 885
    },
    {
      "epoch": 3.955357142857143,
      "grad_norm": 1.4206799268722534,
      "learning_rate": 2.6522184461935153e-06,
      "loss": 0.0893,
      "step": 886
    },
    {
      "epoch": 3.959821428571429,
      "grad_norm": 1.637922763824463,
      "learning_rate": 2.6417994538801882e-06,
      "loss": 0.1107,
      "step": 887
    },
    {
      "epoch": 3.9642857142857144,
      "grad_norm": 1.5130659341812134,
      "learning_rate": 2.6313936219662435e-06,
      "loss": 0.0949,
      "step": 888
    },
    {
      "epoch": 3.96875,
      "grad_norm": 1.4699238538742065,
      "learning_rate": 2.6210010085111507e-06,
      "loss": 0.0965,
      "step": 889
    },
    {
      "epoch": 3.9732142857142856,
      "grad_norm": 1.3603074550628662,
      "learning_rate": 2.6106216715006282e-06,
      "loss": 0.0914,
      "step": 890
    },
    {
      "epoch": 3.977678571428571,
      "grad_norm": 1.3753093481063843,
      "learning_rate": 2.600255668846316e-06,
      "loss": 0.0817,
      "step": 891
    },
    {
      "epoch": 3.982142857142857,
      "grad_norm": 1.5180922746658325,
      "learning_rate": 2.58990305838546e-06,
      "loss": 0.1041,
      "step": 892
    },
    {
      "epoch": 3.986607142857143,
      "grad_norm": 1.5158816576004028,
      "learning_rate": 2.5795638978805755e-06,
      "loss": 0.0873,
      "step": 893
    },
    {
      "epoch": 3.991071428571429,
      "grad_norm": 1.4908344745635986,
      "learning_rate": 2.5692382450191404e-06,
      "loss": 0.103,
      "step": 894
    },
    {
      "epoch": 3.9955357142857144,
      "grad_norm": 1.3467048406600952,
      "learning_rate": 2.558926157413266e-06,
      "loss": 0.1036,
      "step": 895
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.4331597089767456,
      "learning_rate": 2.5486276925993746e-06,
      "loss": 0.0932,
      "step": 896
    },
    {
      "epoch": 4.004464285714286,
      "grad_norm": 0.806198239326477,
      "learning_rate": 2.5383429080378807e-06,
      "loss": 0.0455,
      "step": 897
    },
    {
      "epoch": 4.008928571428571,
      "grad_norm": 0.8375071287155151,
      "learning_rate": 2.5280718611128657e-06,
      "loss": 0.0485,
      "step": 898
    },
    {
      "epoch": 4.013392857142857,
      "grad_norm": 0.7727735042572021,
      "learning_rate": 2.517814609131772e-06,
      "loss": 0.0439,
      "step": 899
    },
    {
      "epoch": 4.017857142857143,
      "grad_norm": 0.8304307460784912,
      "learning_rate": 2.507571209325058e-06,
      "loss": 0.0482,
      "step": 900
    },
    {
      "epoch": 4.022321428571429,
      "grad_norm": 0.9408035278320312,
      "learning_rate": 2.4973417188459074e-06,
      "loss": 0.048,
      "step": 901
    },
    {
      "epoch": 4.026785714285714,
      "grad_norm": 0.8631353378295898,
      "learning_rate": 2.4871261947698892e-06,
      "loss": 0.0499,
      "step": 902
    },
    {
      "epoch": 4.03125,
      "grad_norm": 1.2712801694869995,
      "learning_rate": 2.4769246940946487e-06,
      "loss": 0.0539,
      "step": 903
    },
    {
      "epoch": 4.035714285714286,
      "grad_norm": 0.8662086129188538,
      "learning_rate": 2.4667372737395894e-06,
      "loss": 0.046,
      "step": 904
    },
    {
      "epoch": 4.040178571428571,
      "grad_norm": 0.874553918838501,
      "learning_rate": 2.4565639905455455e-06,
      "loss": 0.0431,
      "step": 905
    },
    {
      "epoch": 4.044642857142857,
      "grad_norm": 0.9109020829200745,
      "learning_rate": 2.446404901274486e-06,
      "loss": 0.0451,
      "step": 906
    },
    {
      "epoch": 4.049107142857143,
      "grad_norm": 0.915429413318634,
      "learning_rate": 2.436260062609173e-06,
      "loss": 0.0453,
      "step": 907
    },
    {
      "epoch": 4.053571428571429,
      "grad_norm": 0.9552186131477356,
      "learning_rate": 2.4261295311528632e-06,
      "loss": 0.0376,
      "step": 908
    },
    {
      "epoch": 4.058035714285714,
      "grad_norm": 1.0016251802444458,
      "learning_rate": 2.4160133634289804e-06,
      "loss": 0.0448,
      "step": 909
    },
    {
      "epoch": 4.0625,
      "grad_norm": 1.0590428113937378,
      "learning_rate": 2.4059116158808147e-06,
      "loss": 0.0487,
      "step": 910
    },
    {
      "epoch": 4.066964285714286,
      "grad_norm": 1.1106352806091309,
      "learning_rate": 2.395824344871193e-06,
      "loss": 0.0439,
      "step": 911
    },
    {
      "epoch": 4.071428571428571,
      "grad_norm": 1.1301071643829346,
      "learning_rate": 2.385751606682167e-06,
      "loss": 0.047,
      "step": 912
    },
    {
      "epoch": 4.075892857142857,
      "grad_norm": 1.0470985174179077,
      "learning_rate": 2.3756934575147117e-06,
      "loss": 0.04,
      "step": 913
    },
    {
      "epoch": 4.080357142857143,
      "grad_norm": 1.318620204925537,
      "learning_rate": 2.365649953488393e-06,
      "loss": 0.0402,
      "step": 914
    },
    {
      "epoch": 4.084821428571429,
      "grad_norm": 1.3197895288467407,
      "learning_rate": 2.3556211506410708e-06,
      "loss": 0.0381,
      "step": 915
    },
    {
      "epoch": 4.089285714285714,
      "grad_norm": 1.20299232006073,
      "learning_rate": 2.3456071049285717e-06,
      "loss": 0.0475,
      "step": 916
    },
    {
      "epoch": 4.09375,
      "grad_norm": 1.209742784500122,
      "learning_rate": 2.3356078722243963e-06,
      "loss": 0.043,
      "step": 917
    },
    {
      "epoch": 4.098214285714286,
      "grad_norm": 1.1857240200042725,
      "learning_rate": 2.325623508319385e-06,
      "loss": 0.0479,
      "step": 918
    },
    {
      "epoch": 4.102678571428571,
      "grad_norm": 0.9122520089149475,
      "learning_rate": 2.3156540689214227e-06,
      "loss": 0.0338,
      "step": 919
    },
    {
      "epoch": 4.107142857142857,
      "grad_norm": 1.1389027833938599,
      "learning_rate": 2.3056996096551228e-06,
      "loss": 0.0403,
      "step": 920
    },
    {
      "epoch": 4.111607142857143,
      "grad_norm": 1.3362747430801392,
      "learning_rate": 2.2957601860615152e-06,
      "loss": 0.0476,
      "step": 921
    },
    {
      "epoch": 4.116071428571429,
      "grad_norm": 1.198205828666687,
      "learning_rate": 2.285835853597742e-06,
      "loss": 0.0427,
      "step": 922
    },
    {
      "epoch": 4.120535714285714,
      "grad_norm": 1.4842236042022705,
      "learning_rate": 2.2759266676367345e-06,
      "loss": 0.0544,
      "step": 923
    },
    {
      "epoch": 4.125,
      "grad_norm": 1.2623423337936401,
      "learning_rate": 2.266032683466928e-06,
      "loss": 0.04,
      "step": 924
    },
    {
      "epoch": 4.129464285714286,
      "grad_norm": 1.2799335718154907,
      "learning_rate": 2.2561539562919265e-06,
      "loss": 0.0392,
      "step": 925
    },
    {
      "epoch": 4.133928571428571,
      "grad_norm": 1.0505305528640747,
      "learning_rate": 2.246290541230214e-06,
      "loss": 0.044,
      "step": 926
    },
    {
      "epoch": 4.138392857142857,
      "grad_norm": 1.0607540607452393,
      "learning_rate": 2.236442493314839e-06,
      "loss": 0.0408,
      "step": 927
    },
    {
      "epoch": 4.142857142857143,
      "grad_norm": 1.1474037170410156,
      "learning_rate": 2.2266098674931094e-06,
      "loss": 0.0391,
      "step": 928
    },
    {
      "epoch": 4.147321428571429,
      "grad_norm": 1.168620228767395,
      "learning_rate": 2.216792718626286e-06,
      "loss": 0.0479,
      "step": 929
    },
    {
      "epoch": 4.151785714285714,
      "grad_norm": 1.028488278388977,
      "learning_rate": 2.2069911014892712e-06,
      "loss": 0.0389,
      "step": 930
    },
    {
      "epoch": 4.15625,
      "grad_norm": 1.0126864910125732,
      "learning_rate": 2.197205070770313e-06,
      "loss": 0.0322,
      "step": 931
    },
    {
      "epoch": 4.160714285714286,
      "grad_norm": 1.1278817653656006,
      "learning_rate": 2.1874346810706925e-06,
      "loss": 0.0411,
      "step": 932
    },
    {
      "epoch": 4.165178571428571,
      "grad_norm": 0.9957524538040161,
      "learning_rate": 2.177679986904422e-06,
      "loss": 0.0381,
      "step": 933
    },
    {
      "epoch": 4.169642857142857,
      "grad_norm": 1.0937297344207764,
      "learning_rate": 2.1679410426979412e-06,
      "loss": 0.0517,
      "step": 934
    },
    {
      "epoch": 4.174107142857143,
      "grad_norm": 1.3177183866500854,
      "learning_rate": 2.1582179027898102e-06,
      "loss": 0.0497,
      "step": 935
    },
    {
      "epoch": 4.178571428571429,
      "grad_norm": 1.0554062128067017,
      "learning_rate": 2.148510621430414e-06,
      "loss": 0.0414,
      "step": 936
    },
    {
      "epoch": 4.183035714285714,
      "grad_norm": 0.9176511764526367,
      "learning_rate": 2.1388192527816472e-06,
      "loss": 0.0335,
      "step": 937
    },
    {
      "epoch": 4.1875,
      "grad_norm": 1.0542479753494263,
      "learning_rate": 2.1291438509166236e-06,
      "loss": 0.0473,
      "step": 938
    },
    {
      "epoch": 4.191964285714286,
      "grad_norm": 1.0238994359970093,
      "learning_rate": 2.119484469819371e-06,
      "loss": 0.0466,
      "step": 939
    },
    {
      "epoch": 4.196428571428571,
      "grad_norm": 1.262686014175415,
      "learning_rate": 2.109841163384528e-06,
      "loss": 0.0456,
      "step": 940
    },
    {
      "epoch": 4.200892857142857,
      "grad_norm": 1.0453062057495117,
      "learning_rate": 2.100213985417045e-06,
      "loss": 0.04,
      "step": 941
    },
    {
      "epoch": 4.205357142857143,
      "grad_norm": 1.3999793529510498,
      "learning_rate": 2.090602989631878e-06,
      "loss": 0.0408,
      "step": 942
    },
    {
      "epoch": 4.209821428571429,
      "grad_norm": 1.0657180547714233,
      "learning_rate": 2.081008229653706e-06,
      "loss": 0.0416,
      "step": 943
    },
    {
      "epoch": 4.214285714285714,
      "grad_norm": 1.0057283639907837,
      "learning_rate": 2.071429759016607e-06,
      "loss": 0.0419,
      "step": 944
    },
    {
      "epoch": 4.21875,
      "grad_norm": 1.26461923122406,
      "learning_rate": 2.061867631163781e-06,
      "loss": 0.0497,
      "step": 945
    },
    {
      "epoch": 4.223214285714286,
      "grad_norm": 0.9643348455429077,
      "learning_rate": 2.0523218994472408e-06,
      "loss": 0.0354,
      "step": 946
    },
    {
      "epoch": 4.227678571428571,
      "grad_norm": 1.0940420627593994,
      "learning_rate": 2.0427926171275157e-06,
      "loss": 0.0415,
      "step": 947
    },
    {
      "epoch": 4.232142857142857,
      "grad_norm": 1.087691068649292,
      "learning_rate": 2.033279837373359e-06,
      "loss": 0.0457,
      "step": 948
    },
    {
      "epoch": 4.236607142857143,
      "grad_norm": 0.987881064414978,
      "learning_rate": 2.023783613261439e-06,
      "loss": 0.038,
      "step": 949
    },
    {
      "epoch": 4.241071428571429,
      "grad_norm": 1.2403086423873901,
      "learning_rate": 2.0143039977760663e-06,
      "loss": 0.0516,
      "step": 950
    },
    {
      "epoch": 4.245535714285714,
      "grad_norm": 1.2339435815811157,
      "learning_rate": 2.0048410438088675e-06,
      "loss": 0.0547,
      "step": 951
    },
    {
      "epoch": 4.25,
      "grad_norm": 1.1114577054977417,
      "learning_rate": 1.995394804158516e-06,
      "loss": 0.0481,
      "step": 952
    },
    {
      "epoch": 4.254464285714286,
      "grad_norm": 0.9100790619850159,
      "learning_rate": 1.9859653315304254e-06,
      "loss": 0.0408,
      "step": 953
    },
    {
      "epoch": 4.258928571428571,
      "grad_norm": 1.1362229585647583,
      "learning_rate": 1.976552678536456e-06,
      "loss": 0.0463,
      "step": 954
    },
    {
      "epoch": 4.263392857142857,
      "grad_norm": 1.3644853830337524,
      "learning_rate": 1.9671568976946257e-06,
      "loss": 0.0584,
      "step": 955
    },
    {
      "epoch": 4.267857142857143,
      "grad_norm": 1.1988905668258667,
      "learning_rate": 1.9577780414288066e-06,
      "loss": 0.0482,
      "step": 956
    },
    {
      "epoch": 4.272321428571429,
      "grad_norm": 1.1114509105682373,
      "learning_rate": 1.9484161620684524e-06,
      "loss": 0.0503,
      "step": 957
    },
    {
      "epoch": 4.276785714285714,
      "grad_norm": 1.1508238315582275,
      "learning_rate": 1.939071311848282e-06,
      "loss": 0.0467,
      "step": 958
    },
    {
      "epoch": 4.28125,
      "grad_norm": 1.444758415222168,
      "learning_rate": 1.9297435429080076e-06,
      "loss": 0.0572,
      "step": 959
    },
    {
      "epoch": 4.285714285714286,
      "grad_norm": 1.0676677227020264,
      "learning_rate": 1.9204329072920285e-06,
      "loss": 0.0368,
      "step": 960
    },
    {
      "epoch": 4.290178571428571,
      "grad_norm": 1.0957095623016357,
      "learning_rate": 1.911139456949158e-06,
      "loss": 0.0472,
      "step": 961
    },
    {
      "epoch": 4.294642857142857,
      "grad_norm": 1.1863818168640137,
      "learning_rate": 1.9018632437323198e-06,
      "loss": 0.0518,
      "step": 962
    },
    {
      "epoch": 4.299107142857143,
      "grad_norm": 1.166928768157959,
      "learning_rate": 1.892604319398259e-06,
      "loss": 0.051,
      "step": 963
    },
    {
      "epoch": 4.303571428571429,
      "grad_norm": 1.0898560285568237,
      "learning_rate": 1.883362735607262e-06,
      "loss": 0.0412,
      "step": 964
    },
    {
      "epoch": 4.308035714285714,
      "grad_norm": 1.1429767608642578,
      "learning_rate": 1.8741385439228616e-06,
      "loss": 0.0495,
      "step": 965
    },
    {
      "epoch": 4.3125,
      "grad_norm": 1.1606385707855225,
      "learning_rate": 1.8649317958115534e-06,
      "loss": 0.0481,
      "step": 966
    },
    {
      "epoch": 4.316964285714286,
      "grad_norm": 1.2927366495132446,
      "learning_rate": 1.8557425426424989e-06,
      "loss": 0.0535,
      "step": 967
    },
    {
      "epoch": 4.321428571428571,
      "grad_norm": 1.053138017654419,
      "learning_rate": 1.8465708356872592e-06,
      "loss": 0.0379,
      "step": 968
    },
    {
      "epoch": 4.325892857142857,
      "grad_norm": 1.1609596014022827,
      "learning_rate": 1.8374167261194826e-06,
      "loss": 0.0478,
      "step": 969
    },
    {
      "epoch": 4.330357142857143,
      "grad_norm": 1.1894792318344116,
      "learning_rate": 1.8282802650146408e-06,
      "loss": 0.0437,
      "step": 970
    },
    {
      "epoch": 4.334821428571429,
      "grad_norm": 1.1033045053482056,
      "learning_rate": 1.8191615033497345e-06,
      "loss": 0.0417,
      "step": 971
    },
    {
      "epoch": 4.339285714285714,
      "grad_norm": 1.0112273693084717,
      "learning_rate": 1.810060492003008e-06,
      "loss": 0.0431,
      "step": 972
    },
    {
      "epoch": 4.34375,
      "grad_norm": 1.02801513671875,
      "learning_rate": 1.800977281753671e-06,
      "loss": 0.0439,
      "step": 973
    },
    {
      "epoch": 4.348214285714286,
      "grad_norm": 1.085830569267273,
      "learning_rate": 1.7919119232816049e-06,
      "loss": 0.0426,
      "step": 974
    },
    {
      "epoch": 4.352678571428571,
      "grad_norm": 1.0054593086242676,
      "learning_rate": 1.7828644671670943e-06,
      "loss": 0.0403,
      "step": 975
    },
    {
      "epoch": 4.357142857142857,
      "grad_norm": 1.0667251348495483,
      "learning_rate": 1.773834963890534e-06,
      "loss": 0.0441,
      "step": 976
    },
    {
      "epoch": 4.361607142857143,
      "grad_norm": 1.149455189704895,
      "learning_rate": 1.764823463832151e-06,
      "loss": 0.041,
      "step": 977
    },
    {
      "epoch": 4.366071428571429,
      "grad_norm": 1.0217986106872559,
      "learning_rate": 1.7558300172717234e-06,
      "loss": 0.0354,
      "step": 978
    },
    {
      "epoch": 4.370535714285714,
      "grad_norm": 1.0308144092559814,
      "learning_rate": 1.7468546743882997e-06,
      "loss": 0.0432,
      "step": 979
    },
    {
      "epoch": 4.375,
      "grad_norm": 1.0886852741241455,
      "learning_rate": 1.7378974852599203e-06,
      "loss": 0.0379,
      "step": 980
    },
    {
      "epoch": 4.379464285714286,
      "grad_norm": 1.1773141622543335,
      "learning_rate": 1.7289584998633307e-06,
      "loss": 0.0476,
      "step": 981
    },
    {
      "epoch": 4.383928571428571,
      "grad_norm": 1.2766141891479492,
      "learning_rate": 1.7200377680737148e-06,
      "loss": 0.0503,
      "step": 982
    },
    {
      "epoch": 4.388392857142857,
      "grad_norm": 1.2046881914138794,
      "learning_rate": 1.7111353396644071e-06,
      "loss": 0.0448,
      "step": 983
    },
    {
      "epoch": 4.392857142857143,
      "grad_norm": 1.4284559488296509,
      "learning_rate": 1.7022512643066196e-06,
      "loss": 0.0487,
      "step": 984
    },
    {
      "epoch": 4.397321428571429,
      "grad_norm": 1.4075138568878174,
      "learning_rate": 1.6933855915691622e-06,
      "loss": 0.0494,
      "step": 985
    },
    {
      "epoch": 4.401785714285714,
      "grad_norm": 1.065433144569397,
      "learning_rate": 1.6845383709181676e-06,
      "loss": 0.0383,
      "step": 986
    },
    {
      "epoch": 4.40625,
      "grad_norm": 1.1404274702072144,
      "learning_rate": 1.675709651716817e-06,
      "loss": 0.0416,
      "step": 987
    },
    {
      "epoch": 4.410714285714286,
      "grad_norm": 1.093176245689392,
      "learning_rate": 1.6668994832250556e-06,
      "loss": 0.0437,
      "step": 988
    },
    {
      "epoch": 4.415178571428571,
      "grad_norm": 1.237988829612732,
      "learning_rate": 1.6581079145993323e-06,
      "loss": 0.0475,
      "step": 989
    },
    {
      "epoch": 4.419642857142857,
      "grad_norm": 1.1673163175582886,
      "learning_rate": 1.649334994892314e-06,
      "loss": 0.0512,
      "step": 990
    },
    {
      "epoch": 4.424107142857143,
      "grad_norm": 1.1633106470108032,
      "learning_rate": 1.640580773052618e-06,
      "loss": 0.0458,
      "step": 991
    },
    {
      "epoch": 4.428571428571429,
      "grad_norm": 1.2017102241516113,
      "learning_rate": 1.6318452979245355e-06,
      "loss": 0.048,
      "step": 992
    },
    {
      "epoch": 4.433035714285714,
      "grad_norm": 1.0068191289901733,
      "learning_rate": 1.6231286182477555e-06,
      "loss": 0.0446,
      "step": 993
    },
    {
      "epoch": 4.4375,
      "grad_norm": 1.1095032691955566,
      "learning_rate": 1.6144307826571085e-06,
      "loss": 0.0489,
      "step": 994
    },
    {
      "epoch": 4.441964285714286,
      "grad_norm": 1.0608917474746704,
      "learning_rate": 1.6057518396822724e-06,
      "loss": 0.0508,
      "step": 995
    },
    {
      "epoch": 4.446428571428571,
      "grad_norm": 1.0646238327026367,
      "learning_rate": 1.5970918377475208e-06,
      "loss": 0.0497,
      "step": 996
    },
    {
      "epoch": 4.450892857142857,
      "grad_norm": 1.0634431838989258,
      "learning_rate": 1.5884508251714436e-06,
      "loss": 0.0402,
      "step": 997
    },
    {
      "epoch": 4.455357142857143,
      "grad_norm": 0.9448213577270508,
      "learning_rate": 1.5798288501666793e-06,
      "loss": 0.0413,
      "step": 998
    },
    {
      "epoch": 4.459821428571429,
      "grad_norm": 1.16658353805542,
      "learning_rate": 1.5712259608396462e-06,
      "loss": 0.0425,
      "step": 999
    },
    {
      "epoch": 4.464285714285714,
      "grad_norm": 1.0001455545425415,
      "learning_rate": 1.5626422051902709e-06,
      "loss": 0.0403,
      "step": 1000
    },
    {
      "epoch": 4.464285714285714,
      "eval_loss": 1.0235848426818848,
      "eval_runtime": 4.4483,
      "eval_samples_per_second": 13.488,
      "eval_steps_per_second": 0.899,
      "step": 1000
    },
    {
      "epoch": 4.46875,
      "grad_norm": 1.3267871141433716,
      "learning_rate": 1.5540776311117304e-06,
      "loss": 0.0441,
      "step": 1001
    },
    {
      "epoch": 4.473214285714286,
      "grad_norm": 1.2346901893615723,
      "learning_rate": 1.5455322863901704e-06,
      "loss": 0.0457,
      "step": 1002
    },
    {
      "epoch": 4.477678571428571,
      "grad_norm": 1.1277903318405151,
      "learning_rate": 1.5370062187044502e-06,
      "loss": 0.0493,
      "step": 1003
    },
    {
      "epoch": 4.482142857142857,
      "grad_norm": 1.0025169849395752,
      "learning_rate": 1.5284994756258718e-06,
      "loss": 0.0436,
      "step": 1004
    },
    {
      "epoch": 4.486607142857143,
      "grad_norm": 1.0304396152496338,
      "learning_rate": 1.5200121046179151e-06,
      "loss": 0.0436,
      "step": 1005
    },
    {
      "epoch": 4.491071428571429,
      "grad_norm": 0.9576307535171509,
      "learning_rate": 1.511544153035975e-06,
      "loss": 0.0403,
      "step": 1006
    },
    {
      "epoch": 4.495535714285714,
      "grad_norm": 0.9636532068252563,
      "learning_rate": 1.5030956681270903e-06,
      "loss": 0.039,
      "step": 1007
    },
    {
      "epoch": 4.5,
      "grad_norm": 1.277954339981079,
      "learning_rate": 1.4946666970296932e-06,
      "loss": 0.0375,
      "step": 1008
    },
    {
      "epoch": 4.504464285714286,
      "grad_norm": 0.9996533393859863,
      "learning_rate": 1.486257286773331e-06,
      "loss": 0.0448,
      "step": 1009
    },
    {
      "epoch": 4.508928571428571,
      "grad_norm": 1.0826925039291382,
      "learning_rate": 1.4778674842784168e-06,
      "loss": 0.0477,
      "step": 1010
    },
    {
      "epoch": 4.513392857142857,
      "grad_norm": 1.0207204818725586,
      "learning_rate": 1.4694973363559539e-06,
      "loss": 0.0358,
      "step": 1011
    },
    {
      "epoch": 4.517857142857143,
      "grad_norm": 0.9995800256729126,
      "learning_rate": 1.4611468897072933e-06,
      "loss": 0.0394,
      "step": 1012
    },
    {
      "epoch": 4.522321428571429,
      "grad_norm": 1.090086579322815,
      "learning_rate": 1.4528161909238569e-06,
      "loss": 0.0356,
      "step": 1013
    },
    {
      "epoch": 4.526785714285714,
      "grad_norm": 1.2063642740249634,
      "learning_rate": 1.44450528648688e-06,
      "loss": 0.0521,
      "step": 1014
    },
    {
      "epoch": 4.53125,
      "grad_norm": 1.2835434675216675,
      "learning_rate": 1.4362142227671607e-06,
      "loss": 0.0491,
      "step": 1015
    },
    {
      "epoch": 4.535714285714286,
      "grad_norm": 1.219610571861267,
      "learning_rate": 1.427943046024793e-06,
      "loss": 0.0428,
      "step": 1016
    },
    {
      "epoch": 4.540178571428571,
      "grad_norm": 1.305721402168274,
      "learning_rate": 1.4196918024089133e-06,
      "loss": 0.0505,
      "step": 1017
    },
    {
      "epoch": 4.544642857142857,
      "grad_norm": 1.181258201599121,
      "learning_rate": 1.4114605379574342e-06,
      "loss": 0.0449,
      "step": 1018
    },
    {
      "epoch": 4.549107142857143,
      "grad_norm": 1.1678040027618408,
      "learning_rate": 1.4032492985968057e-06,
      "loss": 0.0507,
      "step": 1019
    },
    {
      "epoch": 4.553571428571429,
      "grad_norm": 1.034266710281372,
      "learning_rate": 1.3950581301417365e-06,
      "loss": 0.0413,
      "step": 1020
    },
    {
      "epoch": 4.558035714285714,
      "grad_norm": 1.1394208669662476,
      "learning_rate": 1.3868870782949565e-06,
      "loss": 0.0484,
      "step": 1021
    },
    {
      "epoch": 4.5625,
      "grad_norm": 1.031049370765686,
      "learning_rate": 1.3787361886469509e-06,
      "loss": 0.0441,
      "step": 1022
    },
    {
      "epoch": 4.566964285714286,
      "grad_norm": 1.0913993120193481,
      "learning_rate": 1.3706055066757116e-06,
      "loss": 0.0497,
      "step": 1023
    },
    {
      "epoch": 4.571428571428571,
      "grad_norm": 1.1953216791152954,
      "learning_rate": 1.3624950777464828e-06,
      "loss": 0.0466,
      "step": 1024
    },
    {
      "epoch": 4.575892857142857,
      "grad_norm": 1.2053303718566895,
      "learning_rate": 1.3544049471115017e-06,
      "loss": 0.0413,
      "step": 1025
    },
    {
      "epoch": 4.580357142857143,
      "grad_norm": 1.085877537727356,
      "learning_rate": 1.3463351599097552e-06,
      "loss": 0.0429,
      "step": 1026
    },
    {
      "epoch": 4.584821428571429,
      "grad_norm": 1.0058659315109253,
      "learning_rate": 1.3382857611667233e-06,
      "loss": 0.0396,
      "step": 1027
    },
    {
      "epoch": 4.589285714285714,
      "grad_norm": 1.1939631700515747,
      "learning_rate": 1.3302567957941265e-06,
      "loss": 0.0467,
      "step": 1028
    },
    {
      "epoch": 4.59375,
      "grad_norm": 1.085060954093933,
      "learning_rate": 1.3222483085896786e-06,
      "loss": 0.0412,
      "step": 1029
    },
    {
      "epoch": 4.598214285714286,
      "grad_norm": 1.1042296886444092,
      "learning_rate": 1.3142603442368346e-06,
      "loss": 0.0484,
      "step": 1030
    },
    {
      "epoch": 4.602678571428571,
      "grad_norm": 1.2152049541473389,
      "learning_rate": 1.3062929473045442e-06,
      "loss": 0.0445,
      "step": 1031
    },
    {
      "epoch": 4.607142857142857,
      "grad_norm": 1.1404128074645996,
      "learning_rate": 1.2983461622469953e-06,
      "loss": 0.047,
      "step": 1032
    },
    {
      "epoch": 4.611607142857143,
      "grad_norm": 1.044474482536316,
      "learning_rate": 1.290420033403377e-06,
      "loss": 0.0417,
      "step": 1033
    },
    {
      "epoch": 4.616071428571429,
      "grad_norm": 1.0207549333572388,
      "learning_rate": 1.2825146049976244e-06,
      "loss": 0.0408,
      "step": 1034
    },
    {
      "epoch": 4.620535714285714,
      "grad_norm": 1.1905019283294678,
      "learning_rate": 1.2746299211381755e-06,
      "loss": 0.0459,
      "step": 1035
    },
    {
      "epoch": 4.625,
      "grad_norm": 0.9946685433387756,
      "learning_rate": 1.2667660258177241e-06,
      "loss": 0.0416,
      "step": 1036
    },
    {
      "epoch": 4.629464285714286,
      "grad_norm": 0.8535659909248352,
      "learning_rate": 1.258922962912969e-06,
      "loss": 0.0366,
      "step": 1037
    },
    {
      "epoch": 4.633928571428571,
      "grad_norm": 1.1927975416183472,
      "learning_rate": 1.2511007761843839e-06,
      "loss": 0.048,
      "step": 1038
    },
    {
      "epoch": 4.638392857142857,
      "grad_norm": 1.1664050817489624,
      "learning_rate": 1.2432995092759538e-06,
      "loss": 0.0462,
      "step": 1039
    },
    {
      "epoch": 4.642857142857143,
      "grad_norm": 1.0218449831008911,
      "learning_rate": 1.2355192057149477e-06,
      "loss": 0.0455,
      "step": 1040
    },
    {
      "epoch": 4.647321428571429,
      "grad_norm": 1.0675235986709595,
      "learning_rate": 1.2277599089116662e-06,
      "loss": 0.0428,
      "step": 1041
    },
    {
      "epoch": 4.651785714285714,
      "grad_norm": 1.136800765991211,
      "learning_rate": 1.220021662159204e-06,
      "loss": 0.0536,
      "step": 1042
    },
    {
      "epoch": 4.65625,
      "grad_norm": 0.9919513463973999,
      "learning_rate": 1.2123045086332076e-06,
      "loss": 0.0417,
      "step": 1043
    },
    {
      "epoch": 4.660714285714286,
      "grad_norm": 0.9957274198532104,
      "learning_rate": 1.204608491391628e-06,
      "loss": 0.038,
      "step": 1044
    },
    {
      "epoch": 4.665178571428571,
      "grad_norm": 1.0251184701919556,
      "learning_rate": 1.1969336533744965e-06,
      "loss": 0.04,
      "step": 1045
    },
    {
      "epoch": 4.669642857142857,
      "grad_norm": 1.3310158252716064,
      "learning_rate": 1.1892800374036632e-06,
      "loss": 0.0517,
      "step": 1046
    },
    {
      "epoch": 4.674107142857143,
      "grad_norm": 1.0729650259017944,
      "learning_rate": 1.181647686182576e-06,
      "loss": 0.0465,
      "step": 1047
    },
    {
      "epoch": 4.678571428571429,
      "grad_norm": 1.2038968801498413,
      "learning_rate": 1.174036642296036e-06,
      "loss": 0.041,
      "step": 1048
    },
    {
      "epoch": 4.683035714285714,
      "grad_norm": 1.0825115442276,
      "learning_rate": 1.166446948209957e-06,
      "loss": 0.0459,
      "step": 1049
    },
    {
      "epoch": 4.6875,
      "grad_norm": 1.114462971687317,
      "learning_rate": 1.1588786462711347e-06,
      "loss": 0.0471,
      "step": 1050
    },
    {
      "epoch": 4.691964285714286,
      "grad_norm": 0.9582335352897644,
      "learning_rate": 1.1513317787070002e-06,
      "loss": 0.0332,
      "step": 1051
    },
    {
      "epoch": 4.696428571428571,
      "grad_norm": 1.0500428676605225,
      "learning_rate": 1.1438063876254025e-06,
      "loss": 0.0417,
      "step": 1052
    },
    {
      "epoch": 4.700892857142857,
      "grad_norm": 1.2004417181015015,
      "learning_rate": 1.1363025150143508e-06,
      "loss": 0.0448,
      "step": 1053
    },
    {
      "epoch": 4.705357142857143,
      "grad_norm": 1.3093647956848145,
      "learning_rate": 1.1288202027417996e-06,
      "loss": 0.0477,
      "step": 1054
    },
    {
      "epoch": 4.709821428571429,
      "grad_norm": 1.1933882236480713,
      "learning_rate": 1.1213594925554027e-06,
      "loss": 0.0509,
      "step": 1055
    },
    {
      "epoch": 4.714285714285714,
      "grad_norm": 1.1147135496139526,
      "learning_rate": 1.1139204260822874e-06,
      "loss": 0.0512,
      "step": 1056
    },
    {
      "epoch": 4.71875,
      "grad_norm": 1.097452163696289,
      "learning_rate": 1.1065030448288196e-06,
      "loss": 0.047,
      "step": 1057
    },
    {
      "epoch": 4.723214285714286,
      "grad_norm": 1.1414300203323364,
      "learning_rate": 1.0991073901803692e-06,
      "loss": 0.049,
      "step": 1058
    },
    {
      "epoch": 4.727678571428571,
      "grad_norm": 1.1289511919021606,
      "learning_rate": 1.091733503401085e-06,
      "loss": 0.0458,
      "step": 1059
    },
    {
      "epoch": 4.732142857142857,
      "grad_norm": 1.0961730480194092,
      "learning_rate": 1.0843814256336622e-06,
      "loss": 0.0435,
      "step": 1060
    },
    {
      "epoch": 4.736607142857143,
      "grad_norm": 1.0257600545883179,
      "learning_rate": 1.0770511978991116e-06,
      "loss": 0.0424,
      "step": 1061
    },
    {
      "epoch": 4.741071428571429,
      "grad_norm": 1.095285415649414,
      "learning_rate": 1.0697428610965275e-06,
      "loss": 0.0457,
      "step": 1062
    },
    {
      "epoch": 4.745535714285714,
      "grad_norm": 1.035879135131836,
      "learning_rate": 1.0624564560028723e-06,
      "loss": 0.0435,
      "step": 1063
    },
    {
      "epoch": 4.75,
      "grad_norm": 1.1119691133499146,
      "learning_rate": 1.0551920232727309e-06,
      "loss": 0.0406,
      "step": 1064
    },
    {
      "epoch": 4.754464285714286,
      "grad_norm": 1.1761224269866943,
      "learning_rate": 1.0479496034380988e-06,
      "loss": 0.0529,
      "step": 1065
    },
    {
      "epoch": 4.758928571428571,
      "grad_norm": 1.2393746376037598,
      "learning_rate": 1.0407292369081479e-06,
      "loss": 0.049,
      "step": 1066
    },
    {
      "epoch": 4.763392857142857,
      "grad_norm": 1.089432716369629,
      "learning_rate": 1.0335309639690041e-06,
      "loss": 0.0469,
      "step": 1067
    },
    {
      "epoch": 4.767857142857143,
      "grad_norm": 1.0081895589828491,
      "learning_rate": 1.0263548247835246e-06,
      "loss": 0.0438,
      "step": 1068
    },
    {
      "epoch": 4.772321428571429,
      "grad_norm": 0.9703455567359924,
      "learning_rate": 1.0192008593910643e-06,
      "loss": 0.048,
      "step": 1069
    },
    {
      "epoch": 4.776785714285714,
      "grad_norm": 1.0442850589752197,
      "learning_rate": 1.0120691077072643e-06,
      "loss": 0.0432,
      "step": 1070
    },
    {
      "epoch": 4.78125,
      "grad_norm": 1.0315920114517212,
      "learning_rate": 1.004959609523824e-06,
      "loss": 0.0485,
      "step": 1071
    },
    {
      "epoch": 4.785714285714286,
      "grad_norm": 1.118602991104126,
      "learning_rate": 9.978724045082772e-07,
      "loss": 0.0393,
      "step": 1072
    },
    {
      "epoch": 4.790178571428571,
      "grad_norm": 0.991577684879303,
      "learning_rate": 9.908075322037738e-07,
      "loss": 0.0418,
      "step": 1073
    },
    {
      "epoch": 4.794642857142857,
      "grad_norm": 0.9192765951156616,
      "learning_rate": 9.83765032028858e-07,
      "loss": 0.0427,
      "step": 1074
    },
    {
      "epoch": 4.799107142857143,
      "grad_norm": 1.1283397674560547,
      "learning_rate": 9.767449432772485e-07,
      "loss": 0.0403,
      "step": 1075
    },
    {
      "epoch": 4.803571428571429,
      "grad_norm": 1.3338288068771362,
      "learning_rate": 9.697473051176173e-07,
      "loss": 0.0537,
      "step": 1076
    },
    {
      "epoch": 4.808035714285714,
      "grad_norm": 1.03440523147583,
      "learning_rate": 9.627721565933749e-07,
      "loss": 0.0419,
      "step": 1077
    },
    {
      "epoch": 4.8125,
      "grad_norm": 1.070723295211792,
      "learning_rate": 9.558195366224508e-07,
      "loss": 0.0444,
      "step": 1078
    },
    {
      "epoch": 4.816964285714286,
      "grad_norm": 1.0093774795532227,
      "learning_rate": 9.488894839970758e-07,
      "loss": 0.0408,
      "step": 1079
    },
    {
      "epoch": 4.821428571428571,
      "grad_norm": 1.1505217552185059,
      "learning_rate": 9.419820373835668e-07,
      "loss": 0.0429,
      "step": 1080
    },
    {
      "epoch": 4.825892857142857,
      "grad_norm": 1.3309438228607178,
      "learning_rate": 9.350972353221052e-07,
      "loss": 0.0432,
      "step": 1081
    },
    {
      "epoch": 4.830357142857143,
      "grad_norm": 1.3498495817184448,
      "learning_rate": 9.282351162265363e-07,
      "loss": 0.0491,
      "step": 1082
    },
    {
      "epoch": 4.834821428571429,
      "grad_norm": 1.1168887615203857,
      "learning_rate": 9.213957183841355e-07,
      "loss": 0.0436,
      "step": 1083
    },
    {
      "epoch": 4.839285714285714,
      "grad_norm": 1.1466056108474731,
      "learning_rate": 9.145790799554101e-07,
      "loss": 0.045,
      "step": 1084
    },
    {
      "epoch": 4.84375,
      "grad_norm": 0.865831196308136,
      "learning_rate": 9.077852389738817e-07,
      "loss": 0.0386,
      "step": 1085
    },
    {
      "epoch": 4.848214285714286,
      "grad_norm": 1.1055983304977417,
      "learning_rate": 9.010142333458698e-07,
      "loss": 0.038,
      "step": 1086
    },
    {
      "epoch": 4.852678571428571,
      "grad_norm": 1.069543480873108,
      "learning_rate": 8.942661008502875e-07,
      "loss": 0.0436,
      "step": 1087
    },
    {
      "epoch": 4.857142857142857,
      "grad_norm": 1.2057677507400513,
      "learning_rate": 8.87540879138421e-07,
      "loss": 0.0481,
      "step": 1088
    },
    {
      "epoch": 4.861607142857143,
      "grad_norm": 1.28882896900177,
      "learning_rate": 8.808386057337353e-07,
      "loss": 0.0495,
      "step": 1089
    },
    {
      "epoch": 4.866071428571429,
      "grad_norm": 1.4611623287200928,
      "learning_rate": 8.741593180316439e-07,
      "loss": 0.0496,
      "step": 1090
    },
    {
      "epoch": 4.870535714285714,
      "grad_norm": 1.2761002779006958,
      "learning_rate": 8.675030532993193e-07,
      "loss": 0.0443,
      "step": 1091
    },
    {
      "epoch": 4.875,
      "grad_norm": 0.9505082964897156,
      "learning_rate": 8.608698486754739e-07,
      "loss": 0.0426,
      "step": 1092
    },
    {
      "epoch": 4.879464285714286,
      "grad_norm": 1.2187812328338623,
      "learning_rate": 8.542597411701563e-07,
      "loss": 0.045,
      "step": 1093
    },
    {
      "epoch": 4.883928571428571,
      "grad_norm": 0.8786472082138062,
      "learning_rate": 8.476727676645453e-07,
      "loss": 0.0393,
      "step": 1094
    },
    {
      "epoch": 4.888392857142857,
      "grad_norm": 1.0366308689117432,
      "learning_rate": 8.411089649107396e-07,
      "loss": 0.0389,
      "step": 1095
    },
    {
      "epoch": 4.892857142857143,
      "grad_norm": 1.0801434516906738,
      "learning_rate": 8.345683695315633e-07,
      "loss": 0.0406,
      "step": 1096
    },
    {
      "epoch": 4.897321428571429,
      "grad_norm": 1.2877177000045776,
      "learning_rate": 8.280510180203476e-07,
      "loss": 0.0495,
      "step": 1097
    },
    {
      "epoch": 4.901785714285714,
      "grad_norm": 1.1729180812835693,
      "learning_rate": 8.215569467407386e-07,
      "loss": 0.0476,
      "step": 1098
    },
    {
      "epoch": 4.90625,
      "grad_norm": 1.261544942855835,
      "learning_rate": 8.150861919264887e-07,
      "loss": 0.0464,
      "step": 1099
    },
    {
      "epoch": 4.910714285714286,
      "grad_norm": 1.185400128364563,
      "learning_rate": 8.086387896812546e-07,
      "loss": 0.0468,
      "step": 1100
    },
    {
      "epoch": 4.915178571428571,
      "grad_norm": 1.1032450199127197,
      "learning_rate": 8.022147759784016e-07,
      "loss": 0.0435,
      "step": 1101
    },
    {
      "epoch": 4.919642857142857,
      "grad_norm": 1.142734408378601,
      "learning_rate": 7.958141866607897e-07,
      "loss": 0.0418,
      "step": 1102
    },
    {
      "epoch": 4.924107142857143,
      "grad_norm": 1.157031774520874,
      "learning_rate": 7.894370574405928e-07,
      "loss": 0.0509,
      "step": 1103
    },
    {
      "epoch": 4.928571428571429,
      "grad_norm": 1.0427000522613525,
      "learning_rate": 7.830834238990803e-07,
      "loss": 0.0427,
      "step": 1104
    },
    {
      "epoch": 4.933035714285714,
      "grad_norm": 1.059389591217041,
      "learning_rate": 7.767533214864331e-07,
      "loss": 0.0464,
      "step": 1105
    },
    {
      "epoch": 4.9375,
      "grad_norm": 0.9791935086250305,
      "learning_rate": 7.70446785521533e-07,
      "loss": 0.039,
      "step": 1106
    },
    {
      "epoch": 4.941964285714286,
      "grad_norm": 1.1133091449737549,
      "learning_rate": 7.641638511917806e-07,
      "loss": 0.043,
      "step": 1107
    },
    {
      "epoch": 4.946428571428571,
      "grad_norm": 1.1858481168746948,
      "learning_rate": 7.579045535528878e-07,
      "loss": 0.0409,
      "step": 1108
    },
    {
      "epoch": 4.950892857142857,
      "grad_norm": 0.9668167233467102,
      "learning_rate": 7.516689275286813e-07,
      "loss": 0.0392,
      "step": 1109
    },
    {
      "epoch": 4.955357142857143,
      "grad_norm": 1.0199722051620483,
      "learning_rate": 7.454570079109164e-07,
      "loss": 0.0414,
      "step": 1110
    },
    {
      "epoch": 4.959821428571429,
      "grad_norm": 0.9346474409103394,
      "learning_rate": 7.392688293590767e-07,
      "loss": 0.0379,
      "step": 1111
    },
    {
      "epoch": 4.964285714285714,
      "grad_norm": 1.0724751949310303,
      "learning_rate": 7.331044264001836e-07,
      "loss": 0.0368,
      "step": 1112
    },
    {
      "epoch": 4.96875,
      "grad_norm": 1.1779240369796753,
      "learning_rate": 7.269638334285973e-07,
      "loss": 0.0391,
      "step": 1113
    },
    {
      "epoch": 4.973214285714286,
      "grad_norm": 1.2583742141723633,
      "learning_rate": 7.208470847058387e-07,
      "loss": 0.0558,
      "step": 1114
    },
    {
      "epoch": 4.977678571428571,
      "grad_norm": 0.9401357769966125,
      "learning_rate": 7.147542143603806e-07,
      "loss": 0.037,
      "step": 1115
    },
    {
      "epoch": 4.982142857142857,
      "grad_norm": 1.1139758825302124,
      "learning_rate": 7.086852563874705e-07,
      "loss": 0.0452,
      "step": 1116
    },
    {
      "epoch": 4.986607142857143,
      "grad_norm": 1.3145822286605835,
      "learning_rate": 7.026402446489368e-07,
      "loss": 0.0417,
      "step": 1117
    },
    {
      "epoch": 4.991071428571429,
      "grad_norm": 1.0069618225097656,
      "learning_rate": 6.966192128729969e-07,
      "loss": 0.0385,
      "step": 1118
    },
    {
      "epoch": 4.995535714285714,
      "grad_norm": 0.9981904029846191,
      "learning_rate": 6.906221946540765e-07,
      "loss": 0.0386,
      "step": 1119
    },
    {
      "epoch": 5.0,
      "grad_norm": 1.0711932182312012,
      "learning_rate": 6.846492234526105e-07,
      "loss": 0.0469,
      "step": 1120
    },
    {
      "epoch": 5.004464285714286,
      "grad_norm": 0.7480913996696472,
      "learning_rate": 6.787003325948681e-07,
      "loss": 0.0244,
      "step": 1121
    },
    {
      "epoch": 5.008928571428571,
      "grad_norm": 0.4924456775188446,
      "learning_rate": 6.727755552727618e-07,
      "loss": 0.021,
      "step": 1122
    },
    {
      "epoch": 5.013392857142857,
      "grad_norm": 0.608626127243042,
      "learning_rate": 6.668749245436603e-07,
      "loss": 0.0251,
      "step": 1123
    },
    {
      "epoch": 5.017857142857143,
      "grad_norm": 0.5998596549034119,
      "learning_rate": 6.609984733302069e-07,
      "loss": 0.0279,
      "step": 1124
    },
    {
      "epoch": 5.022321428571429,
      "grad_norm": 0.7042911648750305,
      "learning_rate": 6.551462344201356e-07,
      "loss": 0.0283,
      "step": 1125
    },
    {
      "epoch": 5.026785714285714,
      "grad_norm": 0.6839467287063599,
      "learning_rate": 6.493182404660884e-07,
      "loss": 0.029,
      "step": 1126
    },
    {
      "epoch": 5.03125,
      "grad_norm": 0.60224449634552,
      "learning_rate": 6.435145239854279e-07,
      "loss": 0.0263,
      "step": 1127
    },
    {
      "epoch": 5.035714285714286,
      "grad_norm": 0.6765567064285278,
      "learning_rate": 6.377351173600649e-07,
      "loss": 0.0286,
      "step": 1128
    },
    {
      "epoch": 5.040178571428571,
      "grad_norm": 0.6019296646118164,
      "learning_rate": 6.319800528362713e-07,
      "loss": 0.0213,
      "step": 1129
    },
    {
      "epoch": 5.044642857142857,
      "grad_norm": 0.661238968372345,
      "learning_rate": 6.262493625245017e-07,
      "loss": 0.0224,
      "step": 1130
    },
    {
      "epoch": 5.049107142857143,
      "grad_norm": 0.8073251247406006,
      "learning_rate": 6.205430783992163e-07,
      "loss": 0.0335,
      "step": 1131
    },
    {
      "epoch": 5.053571428571429,
      "grad_norm": 0.5287904739379883,
      "learning_rate": 6.148612322986963e-07,
      "loss": 0.0215,
      "step": 1132
    },
    {
      "epoch": 5.058035714285714,
      "grad_norm": 0.7667572498321533,
      "learning_rate": 6.092038559248772e-07,
      "loss": 0.026,
      "step": 1133
    },
    {
      "epoch": 5.0625,
      "grad_norm": 0.6493704319000244,
      "learning_rate": 6.035709808431585e-07,
      "loss": 0.0211,
      "step": 1134
    },
    {
      "epoch": 5.066964285714286,
      "grad_norm": 0.6587660312652588,
      "learning_rate": 5.979626384822384e-07,
      "loss": 0.0259,
      "step": 1135
    },
    {
      "epoch": 5.071428571428571,
      "grad_norm": 0.5626607537269592,
      "learning_rate": 5.923788601339348e-07,
      "loss": 0.02,
      "step": 1136
    },
    {
      "epoch": 5.075892857142857,
      "grad_norm": 0.6060606837272644,
      "learning_rate": 5.868196769530085e-07,
      "loss": 0.021,
      "step": 1137
    },
    {
      "epoch": 5.080357142857143,
      "grad_norm": 0.6305381059646606,
      "learning_rate": 5.81285119956993e-07,
      "loss": 0.0206,
      "step": 1138
    },
    {
      "epoch": 5.084821428571429,
      "grad_norm": 0.6716199517250061,
      "learning_rate": 5.757752200260156e-07,
      "loss": 0.0247,
      "step": 1139
    },
    {
      "epoch": 5.089285714285714,
      "grad_norm": 0.5531538128852844,
      "learning_rate": 5.702900079026365e-07,
      "loss": 0.0214,
      "step": 1140
    },
    {
      "epoch": 5.09375,
      "grad_norm": 0.5417418479919434,
      "learning_rate": 5.648295141916629e-07,
      "loss": 0.0207,
      "step": 1141
    },
    {
      "epoch": 5.098214285714286,
      "grad_norm": 0.7253334522247314,
      "learning_rate": 5.593937693599892e-07,
      "loss": 0.0235,
      "step": 1142
    },
    {
      "epoch": 5.102678571428571,
      "grad_norm": 0.6683683395385742,
      "learning_rate": 5.539828037364222e-07,
      "loss": 0.0243,
      "step": 1143
    },
    {
      "epoch": 5.107142857142857,
      "grad_norm": 0.6907817721366882,
      "learning_rate": 5.485966475115118e-07,
      "loss": 0.0261,
      "step": 1144
    },
    {
      "epoch": 5.111607142857143,
      "grad_norm": 0.6566014289855957,
      "learning_rate": 5.432353307373865e-07,
      "loss": 0.025,
      "step": 1145
    },
    {
      "epoch": 5.116071428571429,
      "grad_norm": 0.8714102506637573,
      "learning_rate": 5.378988833275772e-07,
      "loss": 0.0321,
      "step": 1146
    },
    {
      "epoch": 5.120535714285714,
      "grad_norm": 0.7149608731269836,
      "learning_rate": 5.325873350568641e-07,
      "loss": 0.0248,
      "step": 1147
    },
    {
      "epoch": 5.125,
      "grad_norm": 0.7687610983848572,
      "learning_rate": 5.27300715561093e-07,
      "loss": 0.0284,
      "step": 1148
    },
    {
      "epoch": 5.129464285714286,
      "grad_norm": 0.7787485122680664,
      "learning_rate": 5.220390543370269e-07,
      "loss": 0.0267,
      "step": 1149
    },
    {
      "epoch": 5.133928571428571,
      "grad_norm": 0.6577015519142151,
      "learning_rate": 5.16802380742169e-07,
      "loss": 0.0231,
      "step": 1150
    },
    {
      "epoch": 5.138392857142857,
      "grad_norm": 0.6675891280174255,
      "learning_rate": 5.115907239946071e-07,
      "loss": 0.0213,
      "step": 1151
    },
    {
      "epoch": 5.142857142857143,
      "grad_norm": 0.7006237506866455,
      "learning_rate": 5.064041131728456e-07,
      "loss": 0.0224,
      "step": 1152
    },
    {
      "epoch": 5.147321428571429,
      "grad_norm": 0.7089455127716064,
      "learning_rate": 5.012425772156433e-07,
      "loss": 0.0247,
      "step": 1153
    },
    {
      "epoch": 5.151785714285714,
      "grad_norm": 0.7653855085372925,
      "learning_rate": 4.961061449218561e-07,
      "loss": 0.0211,
      "step": 1154
    },
    {
      "epoch": 5.15625,
      "grad_norm": 0.7013386487960815,
      "learning_rate": 4.90994844950272e-07,
      "loss": 0.0242,
      "step": 1155
    },
    {
      "epoch": 5.160714285714286,
      "grad_norm": 0.743937611579895,
      "learning_rate": 4.859087058194547e-07,
      "loss": 0.024,
      "step": 1156
    },
    {
      "epoch": 5.165178571428571,
      "grad_norm": 0.8444202542304993,
      "learning_rate": 4.808477559075791e-07,
      "loss": 0.0253,
      "step": 1157
    },
    {
      "epoch": 5.169642857142857,
      "grad_norm": 0.8176196217536926,
      "learning_rate": 4.758120234522819e-07,
      "loss": 0.0266,
      "step": 1158
    },
    {
      "epoch": 5.174107142857143,
      "grad_norm": 0.8155895471572876,
      "learning_rate": 4.708015365504947e-07,
      "loss": 0.0277,
      "step": 1159
    },
    {
      "epoch": 5.178571428571429,
      "grad_norm": 0.6969717741012573,
      "learning_rate": 4.658163231582916e-07,
      "loss": 0.0218,
      "step": 1160
    },
    {
      "epoch": 5.183035714285714,
      "grad_norm": 0.776943564414978,
      "learning_rate": 4.6085641109073313e-07,
      "loss": 0.0272,
      "step": 1161
    },
    {
      "epoch": 5.1875,
      "grad_norm": 0.8029648661613464,
      "learning_rate": 4.559218280217121e-07,
      "loss": 0.0265,
      "step": 1162
    },
    {
      "epoch": 5.191964285714286,
      "grad_norm": 0.7203864455223083,
      "learning_rate": 4.5101260148379735e-07,
      "loss": 0.0221,
      "step": 1163
    },
    {
      "epoch": 5.196428571428571,
      "grad_norm": 0.747048556804657,
      "learning_rate": 4.461287588680783e-07,
      "loss": 0.0301,
      "step": 1164
    },
    {
      "epoch": 5.200892857142857,
      "grad_norm": 0.7206503748893738,
      "learning_rate": 4.4127032742401697e-07,
      "loss": 0.0258,
      "step": 1165
    },
    {
      "epoch": 5.205357142857143,
      "grad_norm": 0.7037203311920166,
      "learning_rate": 4.364373342592935e-07,
      "loss": 0.0247,
      "step": 1166
    },
    {
      "epoch": 5.209821428571429,
      "grad_norm": 0.9221356511116028,
      "learning_rate": 4.316298063396534e-07,
      "loss": 0.0226,
      "step": 1167
    },
    {
      "epoch": 5.214285714285714,
      "grad_norm": 0.7772340774536133,
      "learning_rate": 4.2684777048875913e-07,
      "loss": 0.028,
      "step": 1168
    },
    {
      "epoch": 5.21875,
      "grad_norm": 0.9022529721260071,
      "learning_rate": 4.2209125338804007e-07,
      "loss": 0.0297,
      "step": 1169
    },
    {
      "epoch": 5.223214285714286,
      "grad_norm": 0.7532088756561279,
      "learning_rate": 4.173602815765447e-07,
      "loss": 0.025,
      "step": 1170
    },
    {
      "epoch": 5.227678571428571,
      "grad_norm": 0.7878860831260681,
      "learning_rate": 4.126548814507876e-07,
      "loss": 0.0214,
      "step": 1171
    },
    {
      "epoch": 5.232142857142857,
      "grad_norm": 0.7220774292945862,
      "learning_rate": 4.079750792646085e-07,
      "loss": 0.0228,
      "step": 1172
    },
    {
      "epoch": 5.236607142857143,
      "grad_norm": 0.8477165699005127,
      "learning_rate": 4.0332090112902294e-07,
      "loss": 0.0291,
      "step": 1173
    },
    {
      "epoch": 5.241071428571429,
      "grad_norm": 0.7177145481109619,
      "learning_rate": 3.98692373012076e-07,
      "loss": 0.0246,
      "step": 1174
    },
    {
      "epoch": 5.245535714285714,
      "grad_norm": 0.7874996662139893,
      "learning_rate": 3.940895207387007e-07,
      "loss": 0.0221,
      "step": 1175
    },
    {
      "epoch": 5.25,
      "grad_norm": 0.7333866357803345,
      "learning_rate": 3.89512369990565e-07,
      "loss": 0.0248,
      "step": 1176
    },
    {
      "epoch": 5.254464285714286,
      "grad_norm": 0.6481684446334839,
      "learning_rate": 3.849609463059437e-07,
      "loss": 0.0211,
      "step": 1177
    },
    {
      "epoch": 5.258928571428571,
      "grad_norm": 0.7252910733222961,
      "learning_rate": 3.8043527507955926e-07,
      "loss": 0.0242,
      "step": 1178
    },
    {
      "epoch": 5.263392857142857,
      "grad_norm": 1.1857779026031494,
      "learning_rate": 3.759353815624526e-07,
      "loss": 0.0276,
      "step": 1179
    },
    {
      "epoch": 5.267857142857143,
      "grad_norm": 0.8104342222213745,
      "learning_rate": 3.7146129086183547e-07,
      "loss": 0.0235,
      "step": 1180
    },
    {
      "epoch": 5.272321428571429,
      "grad_norm": 0.6407288908958435,
      "learning_rate": 3.6701302794095416e-07,
      "loss": 0.0202,
      "step": 1181
    },
    {
      "epoch": 5.276785714285714,
      "grad_norm": 0.8074800968170166,
      "learning_rate": 3.625906176189484e-07,
      "loss": 0.0269,
      "step": 1182
    },
    {
      "epoch": 5.28125,
      "grad_norm": 0.8074230551719666,
      "learning_rate": 3.5819408457070893e-07,
      "loss": 0.024,
      "step": 1183
    },
    {
      "epoch": 5.285714285714286,
      "grad_norm": 0.7819578051567078,
      "learning_rate": 3.5382345332675154e-07,
      "loss": 0.0246,
      "step": 1184
    },
    {
      "epoch": 5.290178571428571,
      "grad_norm": 0.7379055619239807,
      "learning_rate": 3.494787482730647e-07,
      "loss": 0.0213,
      "step": 1185
    },
    {
      "epoch": 5.294642857142857,
      "grad_norm": 0.8835303783416748,
      "learning_rate": 3.4515999365098667e-07,
      "loss": 0.0215,
      "step": 1186
    },
    {
      "epoch": 5.299107142857143,
      "grad_norm": 0.7571390867233276,
      "learning_rate": 3.4086721355706303e-07,
      "loss": 0.0243,
      "step": 1187
    },
    {
      "epoch": 5.303571428571429,
      "grad_norm": 0.9748966693878174,
      "learning_rate": 3.366004319429139e-07,
      "loss": 0.0285,
      "step": 1188
    },
    {
      "epoch": 5.308035714285714,
      "grad_norm": 0.8343793153762817,
      "learning_rate": 3.323596726151021e-07,
      "loss": 0.0257,
      "step": 1189
    },
    {
      "epoch": 5.3125,
      "grad_norm": 0.7045021057128906,
      "learning_rate": 3.2814495923499496e-07,
      "loss": 0.0185,
      "step": 1190
    },
    {
      "epoch": 5.316964285714286,
      "grad_norm": 0.788077712059021,
      "learning_rate": 3.239563153186429e-07,
      "loss": 0.0214,
      "step": 1191
    },
    {
      "epoch": 5.321428571428571,
      "grad_norm": 0.8326147198677063,
      "learning_rate": 3.197937642366339e-07,
      "loss": 0.0285,
      "step": 1192
    },
    {
      "epoch": 5.325892857142857,
      "grad_norm": 0.7847570180892944,
      "learning_rate": 3.1565732921397583e-07,
      "loss": 0.026,
      "step": 1193
    },
    {
      "epoch": 5.330357142857143,
      "grad_norm": 0.8788312673568726,
      "learning_rate": 3.115470333299593e-07,
      "loss": 0.0242,
      "step": 1194
    },
    {
      "epoch": 5.334821428571429,
      "grad_norm": 0.7230253219604492,
      "learning_rate": 3.0746289951803197e-07,
      "loss": 0.0227,
      "step": 1195
    },
    {
      "epoch": 5.339285714285714,
      "grad_norm": 0.8304932117462158,
      "learning_rate": 3.03404950565671e-07,
      "loss": 0.0264,
      "step": 1196
    },
    {
      "epoch": 5.34375,
      "grad_norm": 0.8365828990936279,
      "learning_rate": 2.9937320911425226e-07,
      "loss": 0.0245,
      "step": 1197
    },
    {
      "epoch": 5.348214285714286,
      "grad_norm": 0.7228520512580872,
      "learning_rate": 2.953676976589278e-07,
      "loss": 0.0201,
      "step": 1198
    },
    {
      "epoch": 5.352678571428571,
      "grad_norm": 0.6695645451545715,
      "learning_rate": 2.9138843854849964e-07,
      "loss": 0.0212,
      "step": 1199
    },
    {
      "epoch": 5.357142857142857,
      "grad_norm": 0.8191896677017212,
      "learning_rate": 2.8743545398529436e-07,
      "loss": 0.0206,
      "step": 1200
    },
    {
      "epoch": 5.357142857142857,
      "eval_loss": 1.1710518598556519,
      "eval_runtime": 4.4446,
      "eval_samples_per_second": 13.5,
      "eval_steps_per_second": 0.9,
      "step": 1200
    },
    {
      "epoch": 5.361607142857143,
      "grad_norm": 0.6906821131706238,
      "learning_rate": 2.8350876602503893e-07,
      "loss": 0.0234,
      "step": 1201
    },
    {
      "epoch": 5.366071428571429,
      "grad_norm": 0.742746889591217,
      "learning_rate": 2.7960839657673844e-07,
      "loss": 0.0249,
      "step": 1202
    },
    {
      "epoch": 5.370535714285714,
      "grad_norm": 0.7856932282447815,
      "learning_rate": 2.7573436740255337e-07,
      "loss": 0.0276,
      "step": 1203
    },
    {
      "epoch": 5.375,
      "grad_norm": 0.8468236327171326,
      "learning_rate": 2.7188670011767715e-07,
      "loss": 0.0295,
      "step": 1204
    },
    {
      "epoch": 5.379464285714286,
      "grad_norm": 0.8368995785713196,
      "learning_rate": 2.680654161902189e-07,
      "loss": 0.0236,
      "step": 1205
    },
    {
      "epoch": 5.383928571428571,
      "grad_norm": 0.6481898427009583,
      "learning_rate": 2.6427053694107966e-07,
      "loss": 0.0171,
      "step": 1206
    },
    {
      "epoch": 5.388392857142857,
      "grad_norm": 0.6581536531448364,
      "learning_rate": 2.605020835438375e-07,
      "loss": 0.0208,
      "step": 1207
    },
    {
      "epoch": 5.392857142857143,
      "grad_norm": 1.2420670986175537,
      "learning_rate": 2.567600770246237e-07,
      "loss": 0.0228,
      "step": 1208
    },
    {
      "epoch": 5.397321428571429,
      "grad_norm": 0.6781761050224304,
      "learning_rate": 2.5304453826201084e-07,
      "loss": 0.0196,
      "step": 1209
    },
    {
      "epoch": 5.401785714285714,
      "grad_norm": 0.8960991501808167,
      "learning_rate": 2.493554879868958e-07,
      "loss": 0.023,
      "step": 1210
    },
    {
      "epoch": 5.40625,
      "grad_norm": 0.815574049949646,
      "learning_rate": 2.4569294678237995e-07,
      "loss": 0.0243,
      "step": 1211
    },
    {
      "epoch": 5.410714285714286,
      "grad_norm": 0.6534021496772766,
      "learning_rate": 2.42056935083658e-07,
      "loss": 0.0209,
      "step": 1212
    },
    {
      "epoch": 5.415178571428571,
      "grad_norm": 0.7151815891265869,
      "learning_rate": 2.38447473177903e-07,
      "loss": 0.0195,
      "step": 1213
    },
    {
      "epoch": 5.419642857142857,
      "grad_norm": 0.6617888808250427,
      "learning_rate": 2.3486458120415415e-07,
      "loss": 0.0215,
      "step": 1214
    },
    {
      "epoch": 5.424107142857143,
      "grad_norm": 0.6826675534248352,
      "learning_rate": 2.3130827915320015e-07,
      "loss": 0.0202,
      "step": 1215
    },
    {
      "epoch": 5.428571428571429,
      "grad_norm": 0.7722153067588806,
      "learning_rate": 2.2777858686747495e-07,
      "loss": 0.0269,
      "step": 1216
    },
    {
      "epoch": 5.433035714285714,
      "grad_norm": 0.651387095451355,
      "learning_rate": 2.242755240409399e-07,
      "loss": 0.0185,
      "step": 1217
    },
    {
      "epoch": 5.4375,
      "grad_norm": 0.7878575921058655,
      "learning_rate": 2.20799110218979e-07,
      "loss": 0.0264,
      "step": 1218
    },
    {
      "epoch": 5.441964285714286,
      "grad_norm": 0.920578122138977,
      "learning_rate": 2.173493647982873e-07,
      "loss": 0.0242,
      "step": 1219
    },
    {
      "epoch": 5.446428571428571,
      "grad_norm": 0.8522472381591797,
      "learning_rate": 2.139263070267605e-07,
      "loss": 0.0261,
      "step": 1220
    },
    {
      "epoch": 5.450892857142857,
      "grad_norm": 0.7734021544456482,
      "learning_rate": 2.105299560033954e-07,
      "loss": 0.0231,
      "step": 1221
    },
    {
      "epoch": 5.455357142857143,
      "grad_norm": 0.6533644795417786,
      "learning_rate": 2.0716033067817308e-07,
      "loss": 0.0235,
      "step": 1222
    },
    {
      "epoch": 5.459821428571429,
      "grad_norm": 0.7817345261573792,
      "learning_rate": 2.03817449851962e-07,
      "loss": 0.0214,
      "step": 1223
    },
    {
      "epoch": 5.464285714285714,
      "grad_norm": 0.671183168888092,
      "learning_rate": 2.0050133217640672e-07,
      "loss": 0.0229,
      "step": 1224
    },
    {
      "epoch": 5.46875,
      "grad_norm": 0.725639283657074,
      "learning_rate": 1.972119961538277e-07,
      "loss": 0.0236,
      "step": 1225
    },
    {
      "epoch": 5.473214285714286,
      "grad_norm": 0.7955847382545471,
      "learning_rate": 1.9394946013711787e-07,
      "loss": 0.0311,
      "step": 1226
    },
    {
      "epoch": 5.477678571428571,
      "grad_norm": 0.8474836349487305,
      "learning_rate": 1.9071374232963564e-07,
      "loss": 0.0263,
      "step": 1227
    },
    {
      "epoch": 5.482142857142857,
      "grad_norm": 0.8439944982528687,
      "learning_rate": 1.8750486078511206e-07,
      "loss": 0.0297,
      "step": 1228
    },
    {
      "epoch": 5.486607142857143,
      "grad_norm": 0.8103631734848022,
      "learning_rate": 1.8432283340754e-07,
      "loss": 0.0256,
      "step": 1229
    },
    {
      "epoch": 5.491071428571429,
      "grad_norm": 0.8478034734725952,
      "learning_rate": 1.8116767795108175e-07,
      "loss": 0.0265,
      "step": 1230
    },
    {
      "epoch": 5.495535714285714,
      "grad_norm": 0.7951886057853699,
      "learning_rate": 1.780394120199672e-07,
      "loss": 0.0232,
      "step": 1231
    },
    {
      "epoch": 5.5,
      "grad_norm": 0.7319163084030151,
      "learning_rate": 1.7493805306839532e-07,
      "loss": 0.029,
      "step": 1232
    },
    {
      "epoch": 5.504464285714286,
      "grad_norm": 0.8299695253372192,
      "learning_rate": 1.718636184004378e-07,
      "loss": 0.0256,
      "step": 1233
    },
    {
      "epoch": 5.508928571428571,
      "grad_norm": 0.7443589568138123,
      "learning_rate": 1.688161251699405e-07,
      "loss": 0.0202,
      "step": 1234
    },
    {
      "epoch": 5.513392857142857,
      "grad_norm": 0.7679203152656555,
      "learning_rate": 1.6579559038043186e-07,
      "loss": 0.0251,
      "step": 1235
    },
    {
      "epoch": 5.517857142857143,
      "grad_norm": 0.6225228309631348,
      "learning_rate": 1.6280203088502275e-07,
      "loss": 0.0222,
      "step": 1236
    },
    {
      "epoch": 5.522321428571429,
      "grad_norm": 0.6747012734413147,
      "learning_rate": 1.5983546338631578e-07,
      "loss": 0.0203,
      "step": 1237
    },
    {
      "epoch": 5.526785714285714,
      "grad_norm": 0.8515854477882385,
      "learning_rate": 1.5689590443631178e-07,
      "loss": 0.0243,
      "step": 1238
    },
    {
      "epoch": 5.53125,
      "grad_norm": 0.7704442143440247,
      "learning_rate": 1.5398337043631723e-07,
      "loss": 0.0222,
      "step": 1239
    },
    {
      "epoch": 5.535714285714286,
      "grad_norm": 0.715097963809967,
      "learning_rate": 1.5109787763685323e-07,
      "loss": 0.024,
      "step": 1240
    },
    {
      "epoch": 5.540178571428571,
      "grad_norm": 0.70245760679245,
      "learning_rate": 1.4823944213756056e-07,
      "loss": 0.0226,
      "step": 1241
    },
    {
      "epoch": 5.544642857142857,
      "grad_norm": 0.8160264492034912,
      "learning_rate": 1.4540807988711857e-07,
      "loss": 0.0249,
      "step": 1242
    },
    {
      "epoch": 5.549107142857143,
      "grad_norm": 0.7433526515960693,
      "learning_rate": 1.4260380668314764e-07,
      "loss": 0.0222,
      "step": 1243
    },
    {
      "epoch": 5.553571428571429,
      "grad_norm": 0.7798489332199097,
      "learning_rate": 1.3982663817212475e-07,
      "loss": 0.0268,
      "step": 1244
    },
    {
      "epoch": 5.558035714285714,
      "grad_norm": 0.7786112427711487,
      "learning_rate": 1.3707658984929738e-07,
      "loss": 0.0252,
      "step": 1245
    },
    {
      "epoch": 5.5625,
      "grad_norm": 0.8275200128555298,
      "learning_rate": 1.3435367705859475e-07,
      "loss": 0.0245,
      "step": 1246
    },
    {
      "epoch": 5.566964285714286,
      "grad_norm": 0.770211935043335,
      "learning_rate": 1.3165791499254294e-07,
      "loss": 0.0254,
      "step": 1247
    },
    {
      "epoch": 5.571428571428571,
      "grad_norm": 0.9335598945617676,
      "learning_rate": 1.2898931869218046e-07,
      "loss": 0.0264,
      "step": 1248
    },
    {
      "epoch": 5.575892857142857,
      "grad_norm": 0.7596639394760132,
      "learning_rate": 1.2634790304697283e-07,
      "loss": 0.0241,
      "step": 1249
    },
    {
      "epoch": 5.580357142857143,
      "grad_norm": 0.7580435276031494,
      "learning_rate": 1.2373368279473364e-07,
      "loss": 0.0217,
      "step": 1250
    },
    {
      "epoch": 5.584821428571429,
      "grad_norm": 0.8928846120834351,
      "learning_rate": 1.2114667252153644e-07,
      "loss": 0.0288,
      "step": 1251
    },
    {
      "epoch": 5.589285714285714,
      "grad_norm": 0.7398735880851746,
      "learning_rate": 1.1858688666163752e-07,
      "loss": 0.0185,
      "step": 1252
    },
    {
      "epoch": 5.59375,
      "grad_norm": 0.8171510100364685,
      "learning_rate": 1.1605433949739546e-07,
      "loss": 0.0287,
      "step": 1253
    },
    {
      "epoch": 5.598214285714286,
      "grad_norm": 0.71816086769104,
      "learning_rate": 1.1354904515918834e-07,
      "loss": 0.0248,
      "step": 1254
    },
    {
      "epoch": 5.602678571428571,
      "grad_norm": 0.7628265023231506,
      "learning_rate": 1.1107101762533725e-07,
      "loss": 0.0242,
      "step": 1255
    },
    {
      "epoch": 5.607142857142857,
      "grad_norm": 0.620901882648468,
      "learning_rate": 1.0862027072202796e-07,
      "loss": 0.0199,
      "step": 1256
    },
    {
      "epoch": 5.611607142857143,
      "grad_norm": 0.9000399112701416,
      "learning_rate": 1.0619681812323437e-07,
      "loss": 0.0303,
      "step": 1257
    },
    {
      "epoch": 5.616071428571429,
      "grad_norm": 0.7469123005867004,
      "learning_rate": 1.0380067335064019e-07,
      "loss": 0.0252,
      "step": 1258
    },
    {
      "epoch": 5.620535714285714,
      "grad_norm": 0.7491857409477234,
      "learning_rate": 1.0143184977356513e-07,
      "loss": 0.0224,
      "step": 1259
    },
    {
      "epoch": 5.625,
      "grad_norm": 0.7108142971992493,
      "learning_rate": 9.909036060889063e-08,
      "loss": 0.0245,
      "step": 1260
    },
    {
      "epoch": 5.629464285714286,
      "grad_norm": 0.6871334910392761,
      "learning_rate": 9.677621892098471e-08,
      "loss": 0.0216,
      "step": 1261
    },
    {
      "epoch": 5.633928571428571,
      "grad_norm": 0.7799824476242065,
      "learning_rate": 9.448943762163063e-08,
      "loss": 0.0258,
      "step": 1262
    },
    {
      "epoch": 5.638392857142857,
      "grad_norm": 0.6933593153953552,
      "learning_rate": 9.223002946995223e-08,
      "loss": 0.0218,
      "step": 1263
    },
    {
      "epoch": 5.642857142857143,
      "grad_norm": 0.7434856295585632,
      "learning_rate": 8.999800707234651e-08,
      "loss": 0.0216,
      "step": 1264
    },
    {
      "epoch": 5.647321428571429,
      "grad_norm": 0.679129421710968,
      "learning_rate": 8.77933828824112e-08,
      "loss": 0.0178,
      "step": 1265
    },
    {
      "epoch": 5.651785714285714,
      "grad_norm": 0.7621864080429077,
      "learning_rate": 8.561616920087338e-08,
      "loss": 0.0188,
      "step": 1266
    },
    {
      "epoch": 5.65625,
      "grad_norm": 0.8270425796508789,
      "learning_rate": 8.346637817552435e-08,
      "loss": 0.0285,
      "step": 1267
    },
    {
      "epoch": 5.660714285714286,
      "grad_norm": 0.718245267868042,
      "learning_rate": 8.134402180115097e-08,
      "loss": 0.0262,
      "step": 1268
    },
    {
      "epoch": 5.665178571428571,
      "grad_norm": 0.7689396739006042,
      "learning_rate": 7.924911191946728e-08,
      "loss": 0.031,
      "step": 1269
    },
    {
      "epoch": 5.669642857142857,
      "grad_norm": 0.7054774165153503,
      "learning_rate": 7.718166021904903e-08,
      "loss": 0.0245,
      "step": 1270
    },
    {
      "epoch": 5.674107142857143,
      "grad_norm": 0.7122982740402222,
      "learning_rate": 7.514167823526817e-08,
      "loss": 0.0266,
      "step": 1271
    },
    {
      "epoch": 5.678571428571429,
      "grad_norm": 0.8461933732032776,
      "learning_rate": 7.31291773502313e-08,
      "loss": 0.0215,
      "step": 1272
    },
    {
      "epoch": 5.683035714285714,
      "grad_norm": 0.8166012763977051,
      "learning_rate": 7.114416879271188e-08,
      "loss": 0.0226,
      "step": 1273
    },
    {
      "epoch": 5.6875,
      "grad_norm": 0.6390535831451416,
      "learning_rate": 6.918666363808976e-08,
      "loss": 0.0203,
      "step": 1274
    },
    {
      "epoch": 5.691964285714286,
      "grad_norm": 0.8909913301467896,
      "learning_rate": 6.725667280828959e-08,
      "loss": 0.026,
      "step": 1275
    },
    {
      "epoch": 5.696428571428571,
      "grad_norm": 0.6632590889930725,
      "learning_rate": 6.535420707172025e-08,
      "loss": 0.0237,
      "step": 1276
    },
    {
      "epoch": 5.700892857142857,
      "grad_norm": 0.575396716594696,
      "learning_rate": 6.347927704321335e-08,
      "loss": 0.018,
      "step": 1277
    },
    {
      "epoch": 5.705357142857143,
      "grad_norm": 0.7212481498718262,
      "learning_rate": 6.163189318396261e-08,
      "loss": 0.0223,
      "step": 1278
    },
    {
      "epoch": 5.709821428571429,
      "grad_norm": 0.8443297147750854,
      "learning_rate": 5.981206580147232e-08,
      "loss": 0.021,
      "step": 1279
    },
    {
      "epoch": 5.714285714285714,
      "grad_norm": 0.8731318116188049,
      "learning_rate": 5.8019805049490143e-08,
      "loss": 0.0276,
      "step": 1280
    },
    {
      "epoch": 5.71875,
      "grad_norm": 1.091413140296936,
      "learning_rate": 5.625512092795828e-08,
      "loss": 0.02,
      "step": 1281
    },
    {
      "epoch": 5.723214285714286,
      "grad_norm": 0.6113536357879639,
      "learning_rate": 5.451802328295408e-08,
      "loss": 0.0197,
      "step": 1282
    },
    {
      "epoch": 5.727678571428571,
      "grad_norm": 0.897653341293335,
      "learning_rate": 5.2808521806635646e-08,
      "loss": 0.0318,
      "step": 1283
    },
    {
      "epoch": 5.732142857142857,
      "grad_norm": 0.7698624134063721,
      "learning_rate": 5.1126626037188537e-08,
      "loss": 0.0226,
      "step": 1284
    },
    {
      "epoch": 5.736607142857143,
      "grad_norm": 0.6883206963539124,
      "learning_rate": 4.9472345358769714e-08,
      "loss": 0.0205,
      "step": 1285
    },
    {
      "epoch": 5.741071428571429,
      "grad_norm": 0.6476832628250122,
      "learning_rate": 4.784568900146095e-08,
      "loss": 0.0199,
      "step": 1286
    },
    {
      "epoch": 5.745535714285714,
      "grad_norm": 0.671054482460022,
      "learning_rate": 4.624666604121047e-08,
      "loss": 0.0202,
      "step": 1287
    },
    {
      "epoch": 5.75,
      "grad_norm": 0.7022548317909241,
      "learning_rate": 4.4675285399787523e-08,
      "loss": 0.021,
      "step": 1288
    },
    {
      "epoch": 5.754464285714286,
      "grad_norm": 0.801397442817688,
      "learning_rate": 4.3131555844730135e-08,
      "loss": 0.023,
      "step": 1289
    },
    {
      "epoch": 5.758928571428571,
      "grad_norm": 1.183430790901184,
      "learning_rate": 4.161548598929689e-08,
      "loss": 0.0221,
      "step": 1290
    },
    {
      "epoch": 5.763392857142857,
      "grad_norm": 1.0177810192108154,
      "learning_rate": 4.01270842924191e-08,
      "loss": 0.0236,
      "step": 1291
    },
    {
      "epoch": 5.767857142857143,
      "grad_norm": 0.7828777432441711,
      "learning_rate": 3.8666359058652064e-08,
      "loss": 0.0213,
      "step": 1292
    },
    {
      "epoch": 5.772321428571429,
      "grad_norm": 0.7310546040534973,
      "learning_rate": 3.7233318438130064e-08,
      "loss": 0.0224,
      "step": 1293
    },
    {
      "epoch": 5.776785714285714,
      "grad_norm": 0.7637031078338623,
      "learning_rate": 3.582797042652248e-08,
      "loss": 0.0254,
      "step": 1294
    },
    {
      "epoch": 5.78125,
      "grad_norm": 0.9804436564445496,
      "learning_rate": 3.4450322864986106e-08,
      "loss": 0.0278,
      "step": 1295
    },
    {
      "epoch": 5.785714285714286,
      "grad_norm": 0.8916503190994263,
      "learning_rate": 3.310038344012184e-08,
      "loss": 0.0256,
      "step": 1296
    },
    {
      "epoch": 5.790178571428571,
      "grad_norm": 0.8148120641708374,
      "learning_rate": 3.17781596839347e-08,
      "loss": 0.0209,
      "step": 1297
    },
    {
      "epoch": 5.794642857142857,
      "grad_norm": 0.8668297529220581,
      "learning_rate": 3.0483658973788894e-08,
      "loss": 0.0228,
      "step": 1298
    },
    {
      "epoch": 5.799107142857143,
      "grad_norm": 0.6615080833435059,
      "learning_rate": 2.92168885323662e-08,
      "loss": 0.0184,
      "step": 1299
    },
    {
      "epoch": 5.803571428571429,
      "grad_norm": 0.7689619064331055,
      "learning_rate": 2.797785542762927e-08,
      "loss": 0.0235,
      "step": 1300
    },
    {
      "epoch": 5.808035714285714,
      "grad_norm": 0.9581748843193054,
      "learning_rate": 2.67665665727784e-08,
      "loss": 0.0265,
      "step": 1301
    },
    {
      "epoch": 5.8125,
      "grad_norm": 0.7389383316040039,
      "learning_rate": 2.5583028726215427e-08,
      "loss": 0.021,
      "step": 1302
    },
    {
      "epoch": 5.816964285714286,
      "grad_norm": 0.752805233001709,
      "learning_rate": 2.442724849150487e-08,
      "loss": 0.024,
      "step": 1303
    },
    {
      "epoch": 5.821428571428571,
      "grad_norm": 0.8107995986938477,
      "learning_rate": 2.3299232317335643e-08,
      "loss": 0.0256,
      "step": 1304
    },
    {
      "epoch": 5.825892857142857,
      "grad_norm": 0.9850029349327087,
      "learning_rate": 2.2198986497489963e-08,
      "loss": 0.0234,
      "step": 1305
    },
    {
      "epoch": 5.830357142857143,
      "grad_norm": 0.6757587194442749,
      "learning_rate": 2.112651717080228e-08,
      "loss": 0.0188,
      "step": 1306
    },
    {
      "epoch": 5.834821428571429,
      "grad_norm": 0.849671483039856,
      "learning_rate": 2.0081830321129298e-08,
      "loss": 0.0245,
      "step": 1307
    },
    {
      "epoch": 5.839285714285714,
      "grad_norm": 0.7331900000572205,
      "learning_rate": 1.906493177731391e-08,
      "loss": 0.0203,
      "step": 1308
    },
    {
      "epoch": 5.84375,
      "grad_norm": 0.7909244894981384,
      "learning_rate": 1.80758272131541e-08,
      "loss": 0.0239,
      "step": 1309
    },
    {
      "epoch": 5.848214285714286,
      "grad_norm": 0.6593401432037354,
      "learning_rate": 1.711452214737187e-08,
      "loss": 0.0192,
      "step": 1310
    },
    {
      "epoch": 5.852678571428571,
      "grad_norm": 0.8781512975692749,
      "learning_rate": 1.6181021943580477e-08,
      "loss": 0.0226,
      "step": 1311
    },
    {
      "epoch": 5.857142857142857,
      "grad_norm": 0.7089523077011108,
      "learning_rate": 1.5275331810256708e-08,
      "loss": 0.0216,
      "step": 1312
    },
    {
      "epoch": 5.861607142857143,
      "grad_norm": 0.9708623886108398,
      "learning_rate": 1.439745680070921e-08,
      "loss": 0.0273,
      "step": 1313
    },
    {
      "epoch": 5.866071428571429,
      "grad_norm": 0.7651482224464417,
      "learning_rate": 1.3547401813053533e-08,
      "loss": 0.0234,
      "step": 1314
    },
    {
      "epoch": 5.870535714285714,
      "grad_norm": 0.721098780632019,
      "learning_rate": 1.2725171590181043e-08,
      "loss": 0.0199,
      "step": 1315
    },
    {
      "epoch": 5.875,
      "grad_norm": 0.8044819831848145,
      "learning_rate": 1.1930770719736715e-08,
      "loss": 0.0258,
      "step": 1316
    },
    {
      "epoch": 5.879464285714286,
      "grad_norm": 0.6233481168746948,
      "learning_rate": 1.1164203634089718e-08,
      "loss": 0.0193,
      "step": 1317
    },
    {
      "epoch": 5.883928571428571,
      "grad_norm": 0.881504237651825,
      "learning_rate": 1.0425474610310654e-08,
      "loss": 0.028,
      "step": 1318
    },
    {
      "epoch": 5.888392857142857,
      "grad_norm": 0.6835102438926697,
      "learning_rate": 9.714587770147148e-09,
      "loss": 0.0201,
      "step": 1319
    },
    {
      "epoch": 5.892857142857143,
      "grad_norm": 0.6504446268081665,
      "learning_rate": 9.031547080002185e-09,
      "loss": 0.0197,
      "step": 1320
    },
    {
      "epoch": 5.897321428571429,
      "grad_norm": 0.6710385084152222,
      "learning_rate": 8.376356350909694e-09,
      "loss": 0.0208,
      "step": 1321
    },
    {
      "epoch": 5.901785714285714,
      "grad_norm": 0.8464110493659973,
      "learning_rate": 7.749019238513461e-09,
      "loss": 0.0312,
      "step": 1322
    },
    {
      "epoch": 5.90625,
      "grad_norm": 0.8968429565429688,
      "learning_rate": 7.149539243050468e-09,
      "loss": 0.0272,
      "step": 1323
    },
    {
      "epoch": 5.910714285714286,
      "grad_norm": 1.0725476741790771,
      "learning_rate": 6.577919709325367e-09,
      "loss": 0.0222,
      "step": 1324
    },
    {
      "epoch": 5.915178571428571,
      "grad_norm": 0.6525650024414062,
      "learning_rate": 6.034163826697711e-09,
      "loss": 0.0204,
      "step": 1325
    },
    {
      "epoch": 5.919642857142857,
      "grad_norm": 0.8145922422409058,
      "learning_rate": 5.518274629059199e-09,
      "loss": 0.023,
      "step": 1326
    },
    {
      "epoch": 5.924107142857143,
      "grad_norm": 0.7454310059547424,
      "learning_rate": 5.030254994820907e-09,
      "loss": 0.0241,
      "step": 1327
    },
    {
      "epoch": 5.928571428571429,
      "grad_norm": 0.7932132482528687,
      "learning_rate": 4.570107646894414e-09,
      "loss": 0.0249,
      "step": 1328
    },
    {
      "epoch": 5.933035714285714,
      "grad_norm": 0.8640964031219482,
      "learning_rate": 4.137835152677938e-09,
      "loss": 0.0223,
      "step": 1329
    },
    {
      "epoch": 5.9375,
      "grad_norm": 0.904984176158905,
      "learning_rate": 3.7334399240402185e-09,
      "loss": 0.0231,
      "step": 1330
    },
    {
      "epoch": 5.941964285714286,
      "grad_norm": 0.849513590335846,
      "learning_rate": 3.356924217310546e-09,
      "loss": 0.0276,
      "step": 1331
    },
    {
      "epoch": 5.946428571428571,
      "grad_norm": 0.7117705345153809,
      "learning_rate": 3.008290133262653e-09,
      "loss": 0.019,
      "step": 1332
    },
    {
      "epoch": 5.950892857142857,
      "grad_norm": 0.7012656927108765,
      "learning_rate": 2.687539617105282e-09,
      "loss": 0.0193,
      "step": 1333
    },
    {
      "epoch": 5.955357142857143,
      "grad_norm": 0.7271556854248047,
      "learning_rate": 2.3946744584688626e-09,
      "loss": 0.0241,
      "step": 1334
    },
    {
      "epoch": 5.959821428571429,
      "grad_norm": 0.6695514917373657,
      "learning_rate": 2.1296962913994105e-09,
      "loss": 0.0197,
      "step": 1335
    },
    {
      "epoch": 5.964285714285714,
      "grad_norm": 0.7411221265792847,
      "learning_rate": 1.892606594345199e-09,
      "loss": 0.0201,
      "step": 1336
    },
    {
      "epoch": 5.96875,
      "grad_norm": 0.8356890678405762,
      "learning_rate": 1.6834066901512136e-09,
      "loss": 0.0249,
      "step": 1337
    },
    {
      "epoch": 5.973214285714286,
      "grad_norm": 0.7158775329589844,
      "learning_rate": 1.5020977460513809e-09,
      "loss": 0.0232,
      "step": 1338
    },
    {
      "epoch": 5.977678571428571,
      "grad_norm": 0.7861185669898987,
      "learning_rate": 1.3486807736613498e-09,
      "loss": 0.0214,
      "step": 1339
    },
    {
      "epoch": 5.982142857142857,
      "grad_norm": 0.770765483379364,
      "learning_rate": 1.2231566289723888e-09,
      "loss": 0.0168,
      "step": 1340
    },
    {
      "epoch": 5.986607142857143,
      "grad_norm": 0.6375583410263062,
      "learning_rate": 1.1255260123486095e-09,
      "loss": 0.0216,
      "step": 1341
    },
    {
      "epoch": 5.991071428571429,
      "grad_norm": 0.795700192451477,
      "learning_rate": 1.0557894685208617e-09,
      "loss": 0.0239,
      "step": 1342
    },
    {
      "epoch": 5.995535714285714,
      "grad_norm": 0.7411868572235107,
      "learning_rate": 1.013947386585067e-09,
      "loss": 0.0249,
      "step": 1343
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.7516708970069885,
      "learning_rate": 1e-09,
      "loss": 0.02,
      "step": 1344
    },
    {
      "epoch": 6.0,
      "step": 1344,
      "total_flos": 2.429040165805621e+18,
      "train_loss": 0.1964568888388818,
      "train_runtime": 10784.1142,
      "train_samples_per_second": 1.993,
      "train_steps_per_second": 0.125
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 1344,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 6,
  "save_steps": 200,
  "total_flos": 2.429040165805621e+18,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}