{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.0,
  "eval_steps": 200,
  "global_step": 460,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.004347826086956522,
      "grad_norm": 20.069940351259884,
      "learning_rate": 9.999883393595949e-06,
      "loss": 1.0922,
      "step": 1
    },
    {
      "epoch": 0.008695652173913044,
      "grad_norm": 7.542838269922722,
      "learning_rate": 9.999533579822611e-06,
      "loss": 0.9273,
      "step": 2
    },
    {
      "epoch": 0.013043478260869565,
      "grad_norm": 10.605262141069346,
      "learning_rate": 9.9989505749962e-06,
      "loss": 0.799,
      "step": 3
    },
    {
      "epoch": 0.017391304347826087,
      "grad_norm": 7.717415499184863,
      "learning_rate": 9.998134406309555e-06,
      "loss": 0.7944,
      "step": 4
    },
    {
      "epoch": 0.021739130434782608,
      "grad_norm": 6.667883368554055,
      "learning_rate": 9.99708511183087e-06,
      "loss": 0.6172,
      "step": 5
    },
    {
      "epoch": 0.02608695652173913,
      "grad_norm": 6.631085068879644,
      "learning_rate": 9.995802740501933e-06,
      "loss": 0.7764,
      "step": 6
    },
    {
      "epoch": 0.030434782608695653,
      "grad_norm": 6.400521762953455,
      "learning_rate": 9.994287352135826e-06,
      "loss": 0.7722,
      "step": 7
    },
    {
      "epoch": 0.034782608695652174,
      "grad_norm": 7.272975450725693,
      "learning_rate": 9.99253901741414e-06,
      "loss": 0.7411,
      "step": 8
    },
    {
      "epoch": 0.0391304347826087,
      "grad_norm": 6.864310215078276,
      "learning_rate": 9.99055781788369e-06,
      "loss": 0.5343,
      "step": 9
    },
    {
      "epoch": 0.043478260869565216,
      "grad_norm": 7.712777689064331,
      "learning_rate": 9.988343845952697e-06,
      "loss": 0.8379,
      "step": 10
    },
    {
      "epoch": 0.04782608695652174,
      "grad_norm": 5.0645786594764735,
      "learning_rate": 9.985897204886481e-06,
      "loss": 0.6745,
      "step": 11
    },
    {
      "epoch": 0.05217391304347826,
      "grad_norm": 6.72945128964307,
      "learning_rate": 9.983218008802648e-06,
      "loss": 0.7465,
      "step": 12
    },
    {
      "epoch": 0.05652173913043478,
      "grad_norm": 6.6367606607720795,
      "learning_rate": 9.98030638266577e-06,
      "loss": 0.799,
      "step": 13
    },
    {
      "epoch": 0.06086956521739131,
      "grad_norm": 5.528359723258759,
      "learning_rate": 9.977162462281544e-06,
      "loss": 0.7108,
      "step": 14
    },
    {
      "epoch": 0.06521739130434782,
      "grad_norm": 5.264702955307018,
      "learning_rate": 9.973786394290475e-06,
      "loss": 0.7228,
      "step": 15
    },
    {
      "epoch": 0.06956521739130435,
      "grad_norm": 6.771882163796353,
      "learning_rate": 9.970178336161018e-06,
      "loss": 0.6881,
      "step": 16
    },
    {
      "epoch": 0.07391304347826087,
      "grad_norm": 5.879249880250146,
      "learning_rate": 9.96633845618225e-06,
      "loss": 0.8758,
      "step": 17
    },
    {
      "epoch": 0.0782608695652174,
      "grad_norm": 5.307640811008731,
      "learning_rate": 9.962266933456008e-06,
      "loss": 0.5422,
      "step": 18
    },
    {
      "epoch": 0.08260869565217391,
      "grad_norm": 6.966122541129152,
      "learning_rate": 9.957963957888542e-06,
      "loss": 0.8481,
      "step": 19
    },
    {
      "epoch": 0.08695652173913043,
      "grad_norm": 7.249802700623231,
      "learning_rate": 9.953429730181653e-06,
      "loss": 0.7001,
      "step": 20
    },
    {
      "epoch": 0.09130434782608696,
      "grad_norm": 7.0428833711498084,
      "learning_rate": 9.94866446182334e-06,
      "loss": 0.8144,
      "step": 21
    },
    {
      "epoch": 0.09565217391304348,
      "grad_norm": 10.996358543863554,
      "learning_rate": 9.943668375077926e-06,
      "loss": 0.8485,
      "step": 22
    },
    {
      "epoch": 0.1,
      "grad_norm": 7.0517113465910395,
      "learning_rate": 9.938441702975689e-06,
      "loss": 0.9405,
      "step": 23
    },
    {
      "epoch": 0.10434782608695652,
      "grad_norm": 6.695372202707919,
      "learning_rate": 9.932984689302012e-06,
      "loss": 0.9361,
      "step": 24
    },
    {
      "epoch": 0.10869565217391304,
      "grad_norm": 7.686125990917134,
      "learning_rate": 9.927297588585984e-06,
      "loss": 0.7243,
      "step": 25
    },
    {
      "epoch": 0.11304347826086956,
      "grad_norm": 7.875281863853482,
      "learning_rate": 9.921380666088558e-06,
      "loss": 0.6751,
      "step": 26
    },
    {
      "epoch": 0.11739130434782609,
      "grad_norm": 7.1870240888829535,
      "learning_rate": 9.915234197790153e-06,
      "loss": 0.8586,
      "step": 27
    },
    {
      "epoch": 0.12173913043478261,
      "grad_norm": 5.805469924139845,
      "learning_rate": 9.908858470377793e-06,
      "loss": 0.9345,
      "step": 28
    },
    {
      "epoch": 0.12608695652173912,
      "grad_norm": 5.934847009962585,
      "learning_rate": 9.902253781231741e-06,
      "loss": 0.6176,
      "step": 29
    },
    {
      "epoch": 0.13043478260869565,
      "grad_norm": 5.669652079107864,
      "learning_rate": 9.895420438411616e-06,
      "loss": 0.8198,
      "step": 30
    },
    {
      "epoch": 0.13478260869565217,
      "grad_norm": 6.9068296669610865,
      "learning_rate": 9.88835876064203e-06,
      "loss": 0.7229,
      "step": 31
    },
    {
      "epoch": 0.1391304347826087,
      "grad_norm": 5.589461426898666,
      "learning_rate": 9.881069077297724e-06,
      "loss": 0.7851,
      "step": 32
    },
    {
      "epoch": 0.14347826086956522,
      "grad_norm": 5.9925311921594675,
      "learning_rate": 9.873551728388203e-06,
      "loss": 0.7627,
      "step": 33
    },
    {
      "epoch": 0.14782608695652175,
      "grad_norm": 7.235153096975517,
      "learning_rate": 9.865807064541878e-06,
      "loss": 0.7946,
      "step": 34
    },
    {
      "epoch": 0.15217391304347827,
      "grad_norm": 5.736002072478754,
      "learning_rate": 9.857835446989708e-06,
      "loss": 1.0581,
      "step": 35
    },
    {
      "epoch": 0.1565217391304348,
      "grad_norm": 5.761567311753849,
      "learning_rate": 9.849637247548356e-06,
      "loss": 0.506,
      "step": 36
    },
    {
      "epoch": 0.1608695652173913,
      "grad_norm": 6.72869577475054,
      "learning_rate": 9.841212848602848e-06,
      "loss": 0.8714,
      "step": 37
    },
    {
      "epoch": 0.16521739130434782,
      "grad_norm": 5.813625848755491,
      "learning_rate": 9.832562643088724e-06,
      "loss": 0.7487,
      "step": 38
    },
    {
      "epoch": 0.16956521739130434,
      "grad_norm": 5.842266886816961,
      "learning_rate": 9.823687034473734e-06,
      "loss": 0.7917,
      "step": 39
    },
    {
      "epoch": 0.17391304347826086,
      "grad_norm": 6.1778401108169145,
      "learning_rate": 9.814586436738998e-06,
      "loss": 1.0145,
      "step": 40
    },
    {
      "epoch": 0.1782608695652174,
      "grad_norm": 7.862015045061562,
      "learning_rate": 9.805261274359705e-06,
      "loss": 0.7646,
      "step": 41
    },
    {
      "epoch": 0.1826086956521739,
      "grad_norm": 5.602026315245972,
      "learning_rate": 9.795711982285317e-06,
      "loss": 0.8201,
      "step": 42
    },
    {
      "epoch": 0.18695652173913044,
      "grad_norm": 4.929349987457967,
      "learning_rate": 9.785939005919279e-06,
      "loss": 0.7702,
      "step": 43
    },
    {
      "epoch": 0.19130434782608696,
      "grad_norm": 4.871372648031096,
      "learning_rate": 9.775942801098241e-06,
      "loss": 0.6926,
      "step": 44
    },
    {
      "epoch": 0.1956521739130435,
      "grad_norm": 6.719859832428872,
      "learning_rate": 9.765723834070805e-06,
      "loss": 0.8486,
      "step": 45
    },
    {
      "epoch": 0.2,
      "grad_norm": 4.3370504785567165,
      "learning_rate": 9.755282581475769e-06,
      "loss": 0.6513,
      "step": 46
    },
    {
      "epoch": 0.20434782608695654,
      "grad_norm": 7.762057444144393,
      "learning_rate": 9.7446195303199e-06,
      "loss": 0.9406,
      "step": 47
    },
    {
      "epoch": 0.20869565217391303,
      "grad_norm": 4.551538491515239,
      "learning_rate": 9.733735177955219e-06,
      "loss": 0.525,
      "step": 48
    },
    {
      "epoch": 0.21304347826086956,
      "grad_norm": 6.075760774188708,
      "learning_rate": 9.722630032055804e-06,
      "loss": 0.7813,
      "step": 49
    },
    {
      "epoch": 0.21739130434782608,
      "grad_norm": 4.008956324178376,
      "learning_rate": 9.711304610594104e-06,
      "loss": 0.7368,
      "step": 50
    },
    {
      "epoch": 0.2217391304347826,
      "grad_norm": 4.839599477628889,
      "learning_rate": 9.699759441816788e-06,
      "loss": 0.6262,
      "step": 51
    },
    {
      "epoch": 0.22608695652173913,
      "grad_norm": 6.38115691482247,
      "learning_rate": 9.687995064220102e-06,
      "loss": 0.877,
      "step": 52
    },
    {
      "epoch": 0.23043478260869565,
      "grad_norm": 5.197270085744578,
      "learning_rate": 9.676012026524755e-06,
      "loss": 0.6567,
      "step": 53
    },
    {
      "epoch": 0.23478260869565218,
      "grad_norm": 6.133664705996566,
      "learning_rate": 9.66381088765032e-06,
      "loss": 0.8683,
      "step": 54
    },
    {
      "epoch": 0.2391304347826087,
      "grad_norm": 5.767075885504545,
      "learning_rate": 9.651392216689167e-06,
      "loss": 1.0394,
      "step": 55
    },
    {
      "epoch": 0.24347826086956523,
      "grad_norm": 4.1300106739423255,
      "learning_rate": 9.638756592879923e-06,
      "loss": 0.6307,
      "step": 56
    },
    {
      "epoch": 0.24782608695652175,
      "grad_norm": 5.031145143621685,
      "learning_rate": 9.625904605580452e-06,
      "loss": 0.6713,
      "step": 57
    },
    {
      "epoch": 0.25217391304347825,
      "grad_norm": 4.919887444745777,
      "learning_rate": 9.61283685424036e-06,
      "loss": 0.7606,
      "step": 58
    },
    {
      "epoch": 0.2565217391304348,
      "grad_norm": 6.6285800296846125,
      "learning_rate": 9.599553948373047e-06,
      "loss": 0.6534,
      "step": 59
    },
    {
      "epoch": 0.2608695652173913,
      "grad_norm": 5.774858135760501,
      "learning_rate": 9.586056507527266e-06,
      "loss": 0.7715,
      "step": 60
    },
    {
      "epoch": 0.26521739130434785,
      "grad_norm": 6.723446220459966,
      "learning_rate": 9.572345161258235e-06,
      "loss": 0.762,
      "step": 61
    },
    {
      "epoch": 0.26956521739130435,
      "grad_norm": 5.1899164656824865,
      "learning_rate": 9.558420549098269e-06,
      "loss": 0.7894,
      "step": 62
    },
    {
      "epoch": 0.27391304347826084,
      "grad_norm": 4.6931553923683484,
      "learning_rate": 9.544283320526943e-06,
      "loss": 0.6347,
      "step": 63
    },
    {
      "epoch": 0.2782608695652174,
      "grad_norm": 4.373053450493686,
      "learning_rate": 9.529934134940819e-06,
      "loss": 0.6341,
      "step": 64
    },
    {
      "epoch": 0.2826086956521739,
      "grad_norm": 4.999427694976061,
      "learning_rate": 9.515373661622665e-06,
      "loss": 0.7572,
      "step": 65
    },
    {
      "epoch": 0.28695652173913044,
      "grad_norm": 6.758587468030614,
      "learning_rate": 9.500602579710256e-06,
      "loss": 0.665,
      "step": 66
    },
    {
      "epoch": 0.29130434782608694,
      "grad_norm": 4.891152841404051,
      "learning_rate": 9.48562157816469e-06,
      "loss": 0.6409,
      "step": 67
    },
    {
      "epoch": 0.2956521739130435,
      "grad_norm": 5.097742356761814,
      "learning_rate": 9.470431355738257e-06,
      "loss": 0.7526,
      "step": 68
    },
    {
      "epoch": 0.3,
      "grad_norm": 5.867883989296531,
      "learning_rate": 9.45503262094184e-06,
      "loss": 0.9745,
      "step": 69
    },
    {
      "epoch": 0.30434782608695654,
      "grad_norm": 6.0206884019070115,
      "learning_rate": 9.439426092011877e-06,
      "loss": 0.7403,
      "step": 70
    },
    {
      "epoch": 0.30869565217391304,
      "grad_norm": 7.276309489479921,
      "learning_rate": 9.423612496876856e-06,
      "loss": 0.8695,
      "step": 71
    },
    {
      "epoch": 0.3130434782608696,
      "grad_norm": 5.239327528351938,
      "learning_rate": 9.407592573123359e-06,
      "loss": 0.6404,
      "step": 72
    },
    {
      "epoch": 0.3173913043478261,
      "grad_norm": 5.793898990560427,
      "learning_rate": 9.39136706796167e-06,
      "loss": 0.8614,
      "step": 73
    },
    {
      "epoch": 0.3217391304347826,
      "grad_norm": 6.692327672103591,
      "learning_rate": 9.374936738190913e-06,
      "loss": 1.0608,
      "step": 74
    },
    {
      "epoch": 0.32608695652173914,
      "grad_norm": 5.4432252422315,
      "learning_rate": 9.358302350163758e-06,
      "loss": 0.9542,
      "step": 75
    },
    {
      "epoch": 0.33043478260869563,
      "grad_norm": 6.189724837073554,
      "learning_rate": 9.341464679750669e-06,
      "loss": 0.785,
      "step": 76
    },
    {
      "epoch": 0.3347826086956522,
      "grad_norm": 5.968778873070777,
      "learning_rate": 9.32442451230373e-06,
      "loss": 0.704,
      "step": 77
    },
    {
      "epoch": 0.3391304347826087,
      "grad_norm": 5.739575659080165,
      "learning_rate": 9.307182642620001e-06,
      "loss": 0.8448,
      "step": 78
    },
    {
      "epoch": 0.34347826086956523,
      "grad_norm": 7.1558802535598955,
      "learning_rate": 9.289739874904448e-06,
      "loss": 0.7278,
      "step": 79
    },
    {
      "epoch": 0.34782608695652173,
      "grad_norm": 5.8994399624654505,
      "learning_rate": 9.272097022732444e-06,
      "loss": 0.6515,
      "step": 80
    },
    {
      "epoch": 0.3521739130434783,
      "grad_norm": 5.19589456035667,
      "learning_rate": 9.254254909011805e-06,
      "loss": 0.6578,
      "step": 81
    },
    {
      "epoch": 0.3565217391304348,
      "grad_norm": 6.310044462837421,
      "learning_rate": 9.236214365944418e-06,
      "loss": 0.7226,
      "step": 82
    },
    {
      "epoch": 0.36086956521739133,
      "grad_norm": 7.094343209942033,
      "learning_rate": 9.217976234987429e-06,
      "loss": 0.7561,
      "step": 83
    },
    {
      "epoch": 0.3652173913043478,
      "grad_norm": 5.395238484535205,
      "learning_rate": 9.199541366813984e-06,
      "loss": 0.8546,
      "step": 84
    },
    {
      "epoch": 0.3695652173913043,
      "grad_norm": 6.294601516599205,
      "learning_rate": 9.180910621273555e-06,
      "loss": 0.8843,
      "step": 85
    },
    {
      "epoch": 0.3739130434782609,
      "grad_norm": 6.077039811904942,
      "learning_rate": 9.16208486735184e-06,
      "loss": 0.7714,
      "step": 86
    },
    {
      "epoch": 0.3782608695652174,
      "grad_norm": 6.820954588122634,
      "learning_rate": 9.14306498313023e-06,
      "loss": 0.875,
      "step": 87
    },
    {
      "epoch": 0.3826086956521739,
      "grad_norm": 6.019619698580998,
      "learning_rate": 9.123851855744842e-06,
      "loss": 0.8366,
      "step": 88
    },
    {
      "epoch": 0.3869565217391304,
      "grad_norm": 6.114796321611535,
      "learning_rate": 9.10444638134516e-06,
      "loss": 0.7968,
      "step": 89
    },
    {
      "epoch": 0.391304347826087,
      "grad_norm": 4.636889958853358,
      "learning_rate": 9.08484946505221e-06,
      "loss": 0.427,
      "step": 90
    },
    {
      "epoch": 0.39565217391304347,
      "grad_norm": 5.954435872454598,
      "learning_rate": 9.065062020916376e-06,
      "loss": 1.0072,
      "step": 91
    },
    {
      "epoch": 0.4,
      "grad_norm": 5.724801876279828,
      "learning_rate": 9.045084971874738e-06,
      "loss": 0.7092,
      "step": 92
    },
    {
      "epoch": 0.4043478260869565,
      "grad_norm": 6.8009420865289165,
      "learning_rate": 9.024919249708034e-06,
      "loss": 0.6406,
      "step": 93
    },
    {
      "epoch": 0.40869565217391307,
      "grad_norm": 7.158183866840323,
      "learning_rate": 9.004565794997209e-06,
      "loss": 1.1627,
      "step": 94
    },
    {
      "epoch": 0.41304347826086957,
      "grad_norm": 5.970996901339507,
      "learning_rate": 8.984025557079523e-06,
      "loss": 0.7724,
      "step": 95
    },
    {
      "epoch": 0.41739130434782606,
      "grad_norm": 5.616656204864542,
      "learning_rate": 8.963299494004292e-06,
      "loss": 0.8157,
      "step": 96
    },
    {
      "epoch": 0.4217391304347826,
      "grad_norm": 3.742930995375726,
      "learning_rate": 8.942388572488188e-06,
      "loss": 0.8402,
      "step": 97
    },
    {
      "epoch": 0.4260869565217391,
      "grad_norm": 5.800107420771233,
      "learning_rate": 8.921293767870157e-06,
      "loss": 0.9098,
      "step": 98
    },
    {
      "epoch": 0.43043478260869567,
      "grad_norm": 6.543219958920574,
      "learning_rate": 8.900016064065923e-06,
      "loss": 0.9196,
      "step": 99
    },
    {
      "epoch": 0.43478260869565216,
      "grad_norm": 4.63775639810159,
      "learning_rate": 8.8785564535221e-06,
      "loss": 0.6786,
      "step": 100
    },
    {
      "epoch": 0.4391304347826087,
      "grad_norm": 9.129310067931197,
      "learning_rate": 8.85691593716989e-06,
      "loss": 1.0006,
      "step": 101
    },
    {
      "epoch": 0.4434782608695652,
      "grad_norm": 5.79169101402568,
      "learning_rate": 8.835095524378413e-06,
      "loss": 0.7848,
      "step": 102
    },
    {
      "epoch": 0.44782608695652176,
      "grad_norm": 8.32981156706014,
      "learning_rate": 8.81309623290762e-06,
      "loss": 1.0868,
      "step": 103
    },
    {
      "epoch": 0.45217391304347826,
      "grad_norm": 3.7350886126503804,
      "learning_rate": 8.790919088860815e-06,
      "loss": 0.5445,
      "step": 104
    },
    {
      "epoch": 0.45652173913043476,
      "grad_norm": 6.277328479888874,
      "learning_rate": 8.768565126636806e-06,
      "loss": 1.0354,
      "step": 105
    },
    {
      "epoch": 0.4608695652173913,
      "grad_norm": 5.781414676947912,
      "learning_rate": 8.746035388881655e-06,
      "loss": 0.7979,
      "step": 106
    },
    {
      "epoch": 0.4652173913043478,
      "grad_norm": 5.96663186885216,
      "learning_rate": 8.723330926440045e-06,
      "loss": 0.8951,
      "step": 107
    },
    {
      "epoch": 0.46956521739130436,
      "grad_norm": 4.835300942906349,
      "learning_rate": 8.70045279830626e-06,
      "loss": 0.7859,
      "step": 108
    },
    {
      "epoch": 0.47391304347826085,
      "grad_norm": 4.985942536490541,
      "learning_rate": 8.677402071574806e-06,
      "loss": 0.5022,
      "step": 109
    },
    {
      "epoch": 0.4782608695652174,
      "grad_norm": 3.5391993265706643,
      "learning_rate": 8.65417982139062e-06,
      "loss": 0.5089,
      "step": 110
    },
    {
      "epoch": 0.4826086956521739,
      "grad_norm": 5.27727991244884,
      "learning_rate": 8.630787130898943e-06,
      "loss": 0.611,
      "step": 111
    },
    {
      "epoch": 0.48695652173913045,
      "grad_norm": 4.410902784418352,
      "learning_rate": 8.60722509119478e-06,
      "loss": 0.6555,
      "step": 112
    },
    {
      "epoch": 0.49130434782608695,
      "grad_norm": 5.311875445651678,
      "learning_rate": 8.583494801272018e-06,
      "loss": 0.7914,
      "step": 113
    },
    {
      "epoch": 0.4956521739130435,
      "grad_norm": 5.341460292078737,
      "learning_rate": 8.559597367972168e-06,
      "loss": 0.6386,
      "step": 114
    },
    {
      "epoch": 0.5,
      "grad_norm": 3.8467630962416237,
      "learning_rate": 8.535533905932739e-06,
      "loss": 0.706,
      "step": 115
    },
    {
      "epoch": 0.5043478260869565,
      "grad_norm": 6.272649123388026,
      "learning_rate": 8.511305537535238e-06,
      "loss": 0.8418,
      "step": 116
    },
    {
      "epoch": 0.508695652173913,
      "grad_norm": 4.449361932301055,
      "learning_rate": 8.48691339285283e-06,
      "loss": 0.5468,
      "step": 117
    },
    {
      "epoch": 0.5130434782608696,
      "grad_norm": 7.990771812448774,
      "learning_rate": 8.462358609597629e-06,
      "loss": 0.7669,
      "step": 118
    },
    {
      "epoch": 0.5173913043478261,
      "grad_norm": 5.264147376966948,
      "learning_rate": 8.437642333067626e-06,
      "loss": 0.87,
      "step": 119
    },
    {
      "epoch": 0.5217391304347826,
      "grad_norm": 5.41751513200529,
      "learning_rate": 8.412765716093273e-06,
      "loss": 0.9239,
      "step": 120
    },
    {
      "epoch": 0.5260869565217391,
      "grad_norm": 4.3710732873831635,
      "learning_rate": 8.387729918983706e-06,
      "loss": 0.7447,
      "step": 121
    },
    {
      "epoch": 0.5304347826086957,
      "grad_norm": 4.8604888967385325,
      "learning_rate": 8.362536109472637e-06,
      "loss": 0.8853,
      "step": 122
    },
    {
      "epoch": 0.5347826086956522,
      "grad_norm": 5.885943638280187,
      "learning_rate": 8.33718546266388e-06,
      "loss": 1.0368,
      "step": 123
    },
    {
      "epoch": 0.5391304347826087,
      "grad_norm": 5.215972523152178,
      "learning_rate": 8.31167916097654e-06,
      "loss": 0.6797,
      "step": 124
    },
    {
      "epoch": 0.5434782608695652,
      "grad_norm": 4.5386673376706135,
      "learning_rate": 8.286018394089864e-06,
      "loss": 0.819,
      "step": 125
    },
    {
      "epoch": 0.5478260869565217,
      "grad_norm": 7.394643265606397,
      "learning_rate": 8.260204358887753e-06,
      "loss": 0.8769,
      "step": 126
    },
    {
      "epoch": 0.5521739130434783,
      "grad_norm": 4.309426896828944,
      "learning_rate": 8.234238259402936e-06,
      "loss": 0.6467,
      "step": 127
    },
    {
      "epoch": 0.5565217391304348,
      "grad_norm": 7.0783793651308935,
      "learning_rate": 8.208121306760806e-06,
      "loss": 0.9724,
      "step": 128
    },
    {
      "epoch": 0.5608695652173913,
      "grad_norm": 7.352028331306423,
      "learning_rate": 8.181854719122938e-06,
      "loss": 0.8856,
      "step": 129
    },
    {
      "epoch": 0.5652173913043478,
      "grad_norm": 8.329174594301895,
      "learning_rate": 8.155439721630265e-06,
      "loss": 1.1432,
      "step": 130
    },
    {
      "epoch": 0.5695652173913044,
      "grad_norm": 6.313137339192889,
      "learning_rate": 8.128877546345932e-06,
      "loss": 0.7633,
      "step": 131
    },
    {
      "epoch": 0.5739130434782609,
      "grad_norm": 5.2417272675388675,
      "learning_rate": 8.102169432197842e-06,
      "loss": 0.6172,
      "step": 132
    },
    {
      "epoch": 0.5782608695652174,
      "grad_norm": 6.311371524751384,
      "learning_rate": 8.075316624920848e-06,
      "loss": 0.6881,
      "step": 133
    },
    {
      "epoch": 0.5826086956521739,
      "grad_norm": 4.59066197947503,
      "learning_rate": 8.048320376998675e-06,
      "loss": 0.8151,
      "step": 134
    },
    {
      "epoch": 0.5869565217391305,
      "grad_norm": 5.182510168555989,
      "learning_rate": 8.021181947605474e-06,
      "loss": 0.6908,
      "step": 135
    },
    {
      "epoch": 0.591304347826087,
      "grad_norm": 4.929873250505049,
      "learning_rate": 7.993902602547113e-06,
      "loss": 0.7495,
      "step": 136
    },
    {
      "epoch": 0.5956521739130435,
      "grad_norm": 6.65667604966722,
      "learning_rate": 7.966483614202127e-06,
      "loss": 0.8716,
      "step": 137
    },
    {
      "epoch": 0.6,
      "grad_norm": 4.1724711277921465,
      "learning_rate": 7.938926261462366e-06,
      "loss": 0.5766,
      "step": 138
    },
    {
      "epoch": 0.6043478260869565,
      "grad_norm": 5.2659580658654015,
      "learning_rate": 7.911231829673356e-06,
      "loss": 0.7273,
      "step": 139
    },
    {
      "epoch": 0.6086956521739131,
      "grad_norm": 6.757332780769119,
      "learning_rate": 7.883401610574338e-06,
      "loss": 0.6014,
      "step": 140
    },
    {
      "epoch": 0.6130434782608696,
      "grad_norm": 5.017197202764835,
      "learning_rate": 7.855436902238018e-06,
      "loss": 0.8168,
      "step": 141
    },
    {
      "epoch": 0.6173913043478261,
      "grad_norm": 3.800051418474176,
      "learning_rate": 7.82733900901003e-06,
      "loss": 0.7956,
      "step": 142
    },
    {
      "epoch": 0.6217391304347826,
      "grad_norm": 5.166291160524886,
      "learning_rate": 7.799109241448091e-06,
      "loss": 0.9329,
      "step": 143
    },
    {
      "epoch": 0.6260869565217392,
      "grad_norm": 6.435240811992826,
      "learning_rate": 7.770748916260875e-06,
      "loss": 1.0732,
      "step": 144
    },
    {
      "epoch": 0.6304347826086957,
      "grad_norm": 6.327169651045853,
      "learning_rate": 7.742259356246594e-06,
      "loss": 0.7525,
      "step": 145
    },
    {
      "epoch": 0.6347826086956522,
      "grad_norm": 7.828014215982498,
      "learning_rate": 7.71364189023131e-06,
      "loss": 0.9682,
      "step": 146
    },
    {
      "epoch": 0.6391304347826087,
      "grad_norm": 4.67990423869777,
      "learning_rate": 7.68489785300694e-06,
      "loss": 0.7471,
      "step": 147
    },
    {
      "epoch": 0.6434782608695652,
      "grad_norm": 6.966974153413055,
      "learning_rate": 7.656028585269017e-06,
      "loss": 0.8207,
      "step": 148
    },
    {
      "epoch": 0.6478260869565218,
      "grad_norm": 5.267291157936162,
      "learning_rate": 7.627035433554138e-06,
      "loss": 0.7868,
      "step": 149
    },
    {
      "epoch": 0.6521739130434783,
      "grad_norm": 3.9897462084848985,
      "learning_rate": 7.597919750177168e-06,
      "loss": 0.6855,
      "step": 150
    },
    {
      "epoch": 0.6565217391304348,
      "grad_norm": 6.0251822846734315,
      "learning_rate": 7.5686828931681646e-06,
      "loss": 0.708,
      "step": 151
    },
    {
      "epoch": 0.6608695652173913,
      "grad_norm": 4.6594453655758254,
      "learning_rate": 7.539326226209032e-06,
      "loss": 0.7041,
      "step": 152
    },
    {
      "epoch": 0.6652173913043479,
      "grad_norm": 4.378552460943295,
      "learning_rate": 7.509851118569915e-06,
      "loss": 0.7772,
      "step": 153
    },
    {
      "epoch": 0.6695652173913044,
      "grad_norm": 5.235514230052308,
      "learning_rate": 7.4802589450453415e-06,
      "loss": 0.8908,
      "step": 154
    },
    {
      "epoch": 0.6739130434782609,
      "grad_norm": 6.003166232352742,
      "learning_rate": 7.450551085890087e-06,
      "loss": 0.6599,
      "step": 155
    },
    {
      "epoch": 0.6782608695652174,
      "grad_norm": 6.96693463687554,
      "learning_rate": 7.420728926754803e-06,
      "loss": 0.6983,
      "step": 156
    },
    {
      "epoch": 0.6826086956521739,
      "grad_norm": 6.169751798135285,
      "learning_rate": 7.390793858621386e-06,
      "loss": 0.5701,
      "step": 157
    },
    {
      "epoch": 0.6869565217391305,
      "grad_norm": 4.644138430432932,
      "learning_rate": 7.360747277738094e-06,
      "loss": 0.8904,
      "step": 158
    },
    {
      "epoch": 0.691304347826087,
      "grad_norm": 4.818846895585623,
      "learning_rate": 7.330590585554428e-06,
      "loss": 0.7802,
      "step": 159
    },
    {
      "epoch": 0.6956521739130435,
      "grad_norm": 5.8506930704235485,
      "learning_rate": 7.300325188655762e-06,
      "loss": 0.7679,
      "step": 160
    },
    {
      "epoch": 0.7,
      "grad_norm": 5.1004665200015,
      "learning_rate": 7.269952498697734e-06,
      "loss": 0.9647,
      "step": 161
    },
    {
      "epoch": 0.7043478260869566,
      "grad_norm": 5.183216444935818,
      "learning_rate": 7.2394739323404105e-06,
      "loss": 1.0273,
      "step": 162
    },
    {
      "epoch": 0.7086956521739131,
      "grad_norm": 4.675829262788197,
      "learning_rate": 7.208890911182198e-06,
      "loss": 1.0811,
      "step": 163
    },
    {
      "epoch": 0.7130434782608696,
      "grad_norm": 5.254697834539126,
      "learning_rate": 7.178204861693546e-06,
      "loss": 0.7775,
      "step": 164
    },
    {
      "epoch": 0.717391304347826,
      "grad_norm": 6.4426513627853295,
      "learning_rate": 7.147417215150411e-06,
      "loss": 0.9937,
      "step": 165
    },
    {
      "epoch": 0.7217391304347827,
      "grad_norm": 4.1964116195751116,
      "learning_rate": 7.116529407567489e-06,
      "loss": 0.5633,
      "step": 166
    },
    {
      "epoch": 0.7260869565217392,
      "grad_norm": 9.567832200885338,
      "learning_rate": 7.085542879631253e-06,
      "loss": 0.7036,
      "step": 167
    },
    {
      "epoch": 0.7304347826086957,
      "grad_norm": 5.675320156588058,
      "learning_rate": 7.054459076632742e-06,
      "loss": 0.9462,
      "step": 168
    },
    {
      "epoch": 0.7347826086956522,
      "grad_norm": 6.094382398275439,
      "learning_rate": 7.0232794484001495e-06,
      "loss": 0.6914,
      "step": 169
    },
    {
      "epoch": 0.7391304347826086,
      "grad_norm": 5.727403297838351,
      "learning_rate": 6.9920054492312086e-06,
      "loss": 0.7357,
      "step": 170
    },
    {
      "epoch": 0.7434782608695653,
      "grad_norm": 9.8890170594532,
      "learning_rate": 6.960638537825352e-06,
      "loss": 1.1028,
      "step": 171
    },
    {
      "epoch": 0.7478260869565218,
      "grad_norm": 5.106590647280527,
      "learning_rate": 6.9291801772156775e-06,
      "loss": 0.7062,
      "step": 172
    },
    {
      "epoch": 0.7521739130434782,
      "grad_norm": 4.689025437172321,
      "learning_rate": 6.89763183470071e-06,
      "loss": 0.7428,
      "step": 173
    },
    {
      "epoch": 0.7565217391304347,
      "grad_norm": 4.773319191168081,
      "learning_rate": 6.865994981775958e-06,
      "loss": 0.8531,
      "step": 174
    },
    {
      "epoch": 0.7608695652173914,
      "grad_norm": 5.6351958720310655,
      "learning_rate": 6.834271094065284e-06,
      "loss": 0.8763,
      "step": 175
    },
    {
      "epoch": 0.7652173913043478,
      "grad_norm": 5.450635954315396,
      "learning_rate": 6.802461651252073e-06,
      "loss": 0.7972,
      "step": 176
    },
    {
      "epoch": 0.7695652173913043,
      "grad_norm": 4.476148487838498,
      "learning_rate": 6.770568137010226e-06,
      "loss": 0.8685,
      "step": 177
    },
    {
      "epoch": 0.7739130434782608,
      "grad_norm": 5.862791509115436,
      "learning_rate": 6.738592038934946e-06,
      "loss": 0.805,
      "step": 178
    },
    {
      "epoch": 0.7782608695652173,
      "grad_norm": 7.184841819252107,
      "learning_rate": 6.706534848473353e-06,
      "loss": 0.815,
      "step": 179
    },
    {
      "epoch": 0.782608695652174,
      "grad_norm": 5.671729894985735,
      "learning_rate": 6.674398060854931e-06,
      "loss": 0.7554,
      "step": 180
    },
    {
      "epoch": 0.7869565217391304,
      "grad_norm": 5.64912938658032,
      "learning_rate": 6.642183175021779e-06,
      "loss": 0.6776,
      "step": 181
    },
    {
      "epoch": 0.7913043478260869,
      "grad_norm": 4.264778264903819,
      "learning_rate": 6.609891693558692e-06,
      "loss": 0.6693,
      "step": 182
    },
    {
      "epoch": 0.7956521739130434,
      "grad_norm": 4.313430011704358,
      "learning_rate": 6.5775251226230855e-06,
      "loss": 0.7671,
      "step": 183
    },
    {
      "epoch": 0.8,
      "grad_norm": 4.945139192368261,
      "learning_rate": 6.545084971874738e-06,
      "loss": 0.795,
      "step": 184
    },
    {
      "epoch": 0.8043478260869565,
      "grad_norm": 6.727158152542215,
      "learning_rate": 6.51257275440538e-06,
      "loss": 1.1242,
      "step": 185
    },
    {
      "epoch": 0.808695652173913,
      "grad_norm": 4.78999021690085,
      "learning_rate": 6.479989986668118e-06,
      "loss": 0.6795,
      "step": 186
    },
    {
      "epoch": 0.8130434782608695,
      "grad_norm": 5.14510842920708,
      "learning_rate": 6.447338188406705e-06,
      "loss": 0.8969,
      "step": 187
    },
    {
      "epoch": 0.8173913043478261,
      "grad_norm": 3.749029424520381,
      "learning_rate": 6.41461888258465e-06,
      "loss": 0.6552,
      "step": 188
    },
    {
      "epoch": 0.8217391304347826,
      "grad_norm": 5.15904621321683,
      "learning_rate": 6.3818335953141955e-06,
      "loss": 0.8136,
      "step": 189
    },
    {
      "epoch": 0.8260869565217391,
      "grad_norm": 4.803773818163898,
      "learning_rate": 6.348983855785122e-06,
      "loss": 0.8237,
      "step": 190
    },
    {
      "epoch": 0.8304347826086956,
      "grad_norm": 5.182891820840243,
      "learning_rate": 6.31607119619343e-06,
      "loss": 0.5909,
      "step": 191
    },
    {
      "epoch": 0.8347826086956521,
      "grad_norm": 5.2883602858763235,
      "learning_rate": 6.283097151669869e-06,
      "loss": 0.8036,
      "step": 192
    },
    {
      "epoch": 0.8391304347826087,
      "grad_norm": 5.575481752264669,
      "learning_rate": 6.250063260208345e-06,
      "loss": 0.7226,
      "step": 193
    },
    {
      "epoch": 0.8434782608695652,
      "grad_norm": 4.959328867352831,
      "learning_rate": 6.216971062594179e-06,
      "loss": 0.7519,
      "step": 194
    },
    {
      "epoch": 0.8478260869565217,
      "grad_norm": 7.33419045327728,
      "learning_rate": 6.183822102332234e-06,
      "loss": 0.7705,
      "step": 195
    },
    {
      "epoch": 0.8521739130434782,
      "grad_norm": 4.819663888433419,
      "learning_rate": 6.1506179255749335e-06,
      "loss": 0.875,
      "step": 196
    },
    {
      "epoch": 0.8565217391304348,
      "grad_norm": 4.427401546783955,
      "learning_rate": 6.1173600810501355e-06,
      "loss": 0.6443,
      "step": 197
    },
    {
      "epoch": 0.8608695652173913,
      "grad_norm": 4.106798354300552,
      "learning_rate": 6.084050119988905e-06,
      "loss": 0.8034,
      "step": 198
    },
    {
      "epoch": 0.8652173913043478,
      "grad_norm": 5.960150253865078,
      "learning_rate": 6.050689596053151e-06,
      "loss": 0.8404,
      "step": 199
    },
    {
      "epoch": 0.8695652173913043,
      "grad_norm": 3.929860508285719,
      "learning_rate": 6.0172800652631706e-06,
      "loss": 0.6939,
      "step": 200
    },
    {
      "epoch": 0.8695652173913043,
      "eval_loss": 0.798964262008667,
      "eval_runtime": 2.9749,
      "eval_samples_per_second": 6.387,
      "eval_steps_per_second": 1.681,
      "step": 200
    },
    {
      "epoch": 0.8739130434782608,
      "grad_norm": 4.59188756563693,
      "learning_rate": 5.983823085925059e-06,
      "loss": 0.6078,
      "step": 201
    },
    {
      "epoch": 0.8782608695652174,
      "grad_norm": 5.005739757058599,
      "learning_rate": 5.950320218558037e-06,
      "loss": 0.5929,
      "step": 202
    },
    {
      "epoch": 0.8826086956521739,
      "grad_norm": 5.345651445814748,
      "learning_rate": 5.916773025821662e-06,
      "loss": 0.7753,
      "step": 203
    },
    {
      "epoch": 0.8869565217391304,
      "grad_norm": 4.336577036436432,
      "learning_rate": 5.883183072442938e-06,
      "loss": 0.6188,
      "step": 204
    },
    {
      "epoch": 0.8913043478260869,
      "grad_norm": 6.421225042043299,
      "learning_rate": 5.849551925143334e-06,
      "loss": 0.8011,
      "step": 205
    },
    {
      "epoch": 0.8956521739130435,
      "grad_norm": 4.789728877148738,
      "learning_rate": 5.815881152565712e-06,
      "loss": 0.7802,
      "step": 206
    },
    {
      "epoch": 0.9,
      "grad_norm": 4.587059004368843,
      "learning_rate": 5.782172325201155e-06,
      "loss": 0.953,
      "step": 207
    },
    {
      "epoch": 0.9043478260869565,
      "grad_norm": 5.009838863672234,
      "learning_rate": 5.7484270153157215e-06,
      "loss": 0.6242,
      "step": 208
    },
    {
      "epoch": 0.908695652173913,
      "grad_norm": 6.424271963786697,
      "learning_rate": 5.714646796877108e-06,
      "loss": 0.9939,
      "step": 209
    },
    {
      "epoch": 0.9130434782608695,
      "grad_norm": 3.971263429130473,
      "learning_rate": 5.680833245481234e-06,
      "loss": 0.7929,
      "step": 210
    },
    {
      "epoch": 0.9173913043478261,
      "grad_norm": 6.726704529221084,
      "learning_rate": 5.646987938278753e-06,
      "loss": 0.7622,
      "step": 211
    },
    {
      "epoch": 0.9217391304347826,
      "grad_norm": 5.6341731609562204,
      "learning_rate": 5.613112453901493e-06,
      "loss": 0.815,
      "step": 212
    },
    {
      "epoch": 0.9260869565217391,
      "grad_norm": 5.064362660354317,
      "learning_rate": 5.579208372388822e-06,
      "loss": 0.7597,
      "step": 213
    },
    {
      "epoch": 0.9304347826086956,
      "grad_norm": 5.760921891007834,
      "learning_rate": 5.5452772751139496e-06,
      "loss": 0.7194,
      "step": 214
    },
    {
      "epoch": 0.9347826086956522,
      "grad_norm": 4.666891611911306,
      "learning_rate": 5.511320744710171e-06,
      "loss": 0.651,
      "step": 215
    },
    {
      "epoch": 0.9391304347826087,
      "grad_norm": 6.267747734622622,
      "learning_rate": 5.477340364997051e-06,
      "loss": 0.9107,
      "step": 216
    },
    {
      "epoch": 0.9434782608695652,
      "grad_norm": 5.863468082482575,
      "learning_rate": 5.443337720906542e-06,
      "loss": 0.8784,
      "step": 217
    },
    {
      "epoch": 0.9478260869565217,
      "grad_norm": 4.161066615088509,
      "learning_rate": 5.409314398409067e-06,
      "loss": 0.6682,
      "step": 218
    },
    {
      "epoch": 0.9521739130434783,
      "grad_norm": 5.171402681222928,
      "learning_rate": 5.375271984439541e-06,
      "loss": 0.6462,
      "step": 219
    },
    {
      "epoch": 0.9565217391304348,
      "grad_norm": 5.096714508227432,
      "learning_rate": 5.341212066823356e-06,
      "loss": 0.7665,
      "step": 220
    },
    {
      "epoch": 0.9608695652173913,
      "grad_norm": 4.7801513361154,
      "learning_rate": 5.307136234202318e-06,
      "loss": 0.6241,
      "step": 221
    },
    {
      "epoch": 0.9652173913043478,
      "grad_norm": 5.916836433268943,
      "learning_rate": 5.27304607596055e-06,
      "loss": 1.0114,
      "step": 222
    },
    {
      "epoch": 0.9695652173913043,
      "grad_norm": 5.0695466084844165,
      "learning_rate": 5.238943182150361e-06,
      "loss": 0.6205,
      "step": 223
    },
    {
      "epoch": 0.9739130434782609,
      "grad_norm": 7.607343695393423,
      "learning_rate": 5.204829143418072e-06,
      "loss": 0.7219,
      "step": 224
    },
    {
      "epoch": 0.9782608695652174,
      "grad_norm": 5.071732375655891,
      "learning_rate": 5.17070555092984e-06,
      "loss": 0.6891,
      "step": 225
    },
    {
      "epoch": 0.9826086956521739,
      "grad_norm": 6.457774215325638,
      "learning_rate": 5.136573996297431e-06,
      "loss": 0.9542,
      "step": 226
    },
    {
      "epoch": 0.9869565217391304,
      "grad_norm": 6.582844550791858,
      "learning_rate": 5.102436071503983e-06,
      "loss": 0.8563,
      "step": 227
    },
    {
      "epoch": 0.991304347826087,
      "grad_norm": 2.936968503934489,
      "learning_rate": 5.068293368829755e-06,
      "loss": 0.7973,
      "step": 228
    },
    {
      "epoch": 0.9956521739130435,
      "grad_norm": 6.489247940312402,
      "learning_rate": 5.034147480777867e-06,
      "loss": 0.7773,
      "step": 229
    },
    {
      "epoch": 1.0,
      "grad_norm": 4.047478627971296,
      "learning_rate": 5e-06,
      "loss": 0.8434,
      "step": 230
    },
    {
      "epoch": 1.0043478260869565,
      "grad_norm": 4.837822183603084,
      "learning_rate": 4.965852519222135e-06,
      "loss": 0.5609,
      "step": 231
    },
    {
      "epoch": 1.008695652173913,
      "grad_norm": 4.082899268222064,
      "learning_rate": 4.931706631170246e-06,
      "loss": 0.5027,
      "step": 232
    },
    {
      "epoch": 1.0130434782608695,
      "grad_norm": 3.843325733590251,
      "learning_rate": 4.89756392849602e-06,
      "loss": 0.3658,
      "step": 233
    },
    {
      "epoch": 1.017391304347826,
      "grad_norm": 4.836952879734927,
      "learning_rate": 4.863426003702572e-06,
      "loss": 0.5756,
      "step": 234
    },
    {
      "epoch": 1.0217391304347827,
      "grad_norm": 4.103173107463818,
      "learning_rate": 4.829294449070161e-06,
      "loss": 0.4994,
      "step": 235
    },
    {
      "epoch": 1.0260869565217392,
      "grad_norm": 3.8903384128887106,
      "learning_rate": 4.795170856581929e-06,
      "loss": 0.4887,
      "step": 236
    },
    {
      "epoch": 1.0304347826086957,
      "grad_norm": 3.08596959753649,
      "learning_rate": 4.7610568178496405e-06,
      "loss": 0.4403,
      "step": 237
    },
    {
      "epoch": 1.0347826086956522,
      "grad_norm": 3.527272426977584,
      "learning_rate": 4.7269539240394505e-06,
      "loss": 0.3424,
      "step": 238
    },
    {
      "epoch": 1.0391304347826087,
      "grad_norm": 3.2374239776762246,
      "learning_rate": 4.692863765797683e-06,
      "loss": 0.4145,
      "step": 239
    },
    {
      "epoch": 1.0434782608695652,
      "grad_norm": 3.4800072353942224,
      "learning_rate": 4.6587879331766465e-06,
      "loss": 0.5048,
      "step": 240
    },
    {
      "epoch": 1.0478260869565217,
      "grad_norm": 3.192418206458333,
      "learning_rate": 4.624728015560461e-06,
      "loss": 0.4787,
      "step": 241
    },
    {
      "epoch": 1.0521739130434782,
      "grad_norm": 5.143086646430212,
      "learning_rate": 4.5906856015909365e-06,
      "loss": 0.3719,
      "step": 242
    },
    {
      "epoch": 1.0565217391304347,
      "grad_norm": 3.8074108245867997,
      "learning_rate": 4.556662279093461e-06,
      "loss": 0.3785,
      "step": 243
    },
    {
      "epoch": 1.0608695652173914,
      "grad_norm": 4.498314324006594,
      "learning_rate": 4.52265963500295e-06,
      "loss": 0.552,
      "step": 244
    },
    {
      "epoch": 1.065217391304348,
      "grad_norm": 4.163738307816856,
      "learning_rate": 4.488679255289829e-06,
      "loss": 0.4597,
      "step": 245
    },
    {
      "epoch": 1.0695652173913044,
      "grad_norm": 4.350615353682191,
      "learning_rate": 4.454722724886051e-06,
      "loss": 0.3357,
      "step": 246
    },
    {
      "epoch": 1.0739130434782609,
      "grad_norm": 3.943016818527937,
      "learning_rate": 4.4207916276111795e-06,
      "loss": 0.3514,
      "step": 247
    },
    {
      "epoch": 1.0782608695652174,
      "grad_norm": 5.2085391198088535,
      "learning_rate": 4.386887546098509e-06,
      "loss": 0.413,
      "step": 248
    },
    {
      "epoch": 1.0826086956521739,
      "grad_norm": 5.6969414567070045,
      "learning_rate": 4.353012061721249e-06,
      "loss": 0.3608,
      "step": 249
    },
    {
      "epoch": 1.0869565217391304,
      "grad_norm": 4.268384815942112,
      "learning_rate": 4.319166754518768e-06,
      "loss": 0.288,
      "step": 250
    },
    {
      "epoch": 1.0913043478260869,
      "grad_norm": 3.579380382838798,
      "learning_rate": 4.285353203122894e-06,
      "loss": 0.2789,
      "step": 251
    },
    {
      "epoch": 1.0956521739130434,
      "grad_norm": 5.219933048553219,
      "learning_rate": 4.251572984684281e-06,
      "loss": 0.5852,
      "step": 252
    },
    {
      "epoch": 1.1,
      "grad_norm": 5.975508571527501,
      "learning_rate": 4.217827674798845e-06,
      "loss": 0.4421,
      "step": 253
    },
    {
      "epoch": 1.1043478260869566,
      "grad_norm": 3.8240302327268174,
      "learning_rate": 4.18411884743429e-06,
      "loss": 0.3446,
      "step": 254
    },
    {
      "epoch": 1.108695652173913,
      "grad_norm": 3.7445910108565044,
      "learning_rate": 4.150448074856667e-06,
      "loss": 0.3875,
      "step": 255
    },
    {
      "epoch": 1.1130434782608696,
      "grad_norm": 4.3241863758181776,
      "learning_rate": 4.116816927557063e-06,
      "loss": 0.3562,
      "step": 256
    },
    {
      "epoch": 1.117391304347826,
      "grad_norm": 4.547780326995007,
      "learning_rate": 4.083226974178339e-06,
      "loss": 0.2453,
      "step": 257
    },
    {
      "epoch": 1.1217391304347826,
      "grad_norm": 3.687859503920098,
      "learning_rate": 4.0496797814419655e-06,
      "loss": 0.3146,
      "step": 258
    },
    {
      "epoch": 1.126086956521739,
      "grad_norm": 4.946712083509019,
      "learning_rate": 4.016176914074944e-06,
      "loss": 0.3909,
      "step": 259
    },
    {
      "epoch": 1.1304347826086956,
      "grad_norm": 5.24606301014812,
      "learning_rate": 3.982719934736832e-06,
      "loss": 0.3626,
      "step": 260
    },
    {
      "epoch": 1.134782608695652,
      "grad_norm": 3.4719423073847078,
      "learning_rate": 3.949310403946849e-06,
      "loss": 0.3811,
      "step": 261
    },
    {
      "epoch": 1.1391304347826088,
      "grad_norm": 4.29696507978264,
      "learning_rate": 3.915949880011096e-06,
      "loss": 0.332,
      "step": 262
    },
    {
      "epoch": 1.1434782608695653,
      "grad_norm": 3.9771307721259475,
      "learning_rate": 3.882639918949865e-06,
      "loss": 0.4423,
      "step": 263
    },
    {
      "epoch": 1.1478260869565218,
      "grad_norm": 3.744900801031998,
      "learning_rate": 3.849382074425069e-06,
      "loss": 0.3336,
      "step": 264
    },
    {
      "epoch": 1.1521739130434783,
      "grad_norm": 4.938081646469259,
      "learning_rate": 3.816177897667767e-06,
      "loss": 0.6139,
      "step": 265
    },
    {
      "epoch": 1.1565217391304348,
      "grad_norm": 6.709242400647857,
      "learning_rate": 3.7830289374058214e-06,
      "loss": 0.4085,
      "step": 266
    },
    {
      "epoch": 1.1608695652173913,
      "grad_norm": 6.894376674623404,
      "learning_rate": 3.749936739791656e-06,
      "loss": 0.3985,
      "step": 267
    },
    {
      "epoch": 1.1652173913043478,
      "grad_norm": 4.117232458690699,
      "learning_rate": 3.7169028483301333e-06,
      "loss": 0.3541,
      "step": 268
    },
    {
      "epoch": 1.1695652173913043,
      "grad_norm": 3.6495749370260957,
      "learning_rate": 3.6839288038065736e-06,
      "loss": 0.2845,
      "step": 269
    },
    {
      "epoch": 1.1739130434782608,
      "grad_norm": 4.0137488559165355,
      "learning_rate": 3.6510161442148783e-06,
      "loss": 0.3154,
      "step": 270
    },
    {
      "epoch": 1.1782608695652175,
      "grad_norm": 4.8152515686913455,
      "learning_rate": 3.6181664046858045e-06,
      "loss": 0.4594,
      "step": 271
    },
    {
      "epoch": 1.182608695652174,
      "grad_norm": 5.202407554106834,
      "learning_rate": 3.58538111741535e-06,
      "loss": 0.5321,
      "step": 272
    },
    {
      "epoch": 1.1869565217391305,
      "grad_norm": 5.208348292849049,
      "learning_rate": 3.5526618115932974e-06,
      "loss": 0.4076,
      "step": 273
    },
    {
      "epoch": 1.191304347826087,
      "grad_norm": 4.524886701172911,
      "learning_rate": 3.5200100133318836e-06,
      "loss": 0.6158,
      "step": 274
    },
    {
      "epoch": 1.1956521739130435,
      "grad_norm": 5.093651871829818,
      "learning_rate": 3.4874272455946217e-06,
      "loss": 0.2257,
      "step": 275
    },
    {
      "epoch": 1.2,
      "grad_norm": 4.022672833100005,
      "learning_rate": 3.4549150281252635e-06,
      "loss": 0.3768,
      "step": 276
    },
    {
      "epoch": 1.2043478260869565,
      "grad_norm": 2.8604068490644874,
      "learning_rate": 3.4224748773769166e-06,
      "loss": 0.3304,
      "step": 277
    },
    {
      "epoch": 1.208695652173913,
      "grad_norm": 3.563416318864363,
      "learning_rate": 3.39010830644131e-06,
      "loss": 0.4327,
      "step": 278
    },
    {
      "epoch": 1.2130434782608694,
      "grad_norm": 5.462342178891127,
      "learning_rate": 3.357816824978222e-06,
      "loss": 0.3563,
      "step": 279
    },
    {
      "epoch": 1.2173913043478262,
      "grad_norm": 5.776352122429311,
      "learning_rate": 3.3256019391450696e-06,
      "loss": 0.554,
      "step": 280
    },
    {
      "epoch": 1.2217391304347827,
      "grad_norm": 4.280137224720186,
      "learning_rate": 3.2934651515266485e-06,
      "loss": 0.4447,
      "step": 281
    },
    {
      "epoch": 1.2260869565217392,
      "grad_norm": 5.855318929352802,
      "learning_rate": 3.261407961065056e-06,
      "loss": 0.4613,
      "step": 282
    },
    {
      "epoch": 1.2304347826086957,
      "grad_norm": 3.759306773035557,
      "learning_rate": 3.2294318629897746e-06,
      "loss": 0.41,
      "step": 283
    },
    {
      "epoch": 1.2347826086956522,
      "grad_norm": 4.325778018741157,
      "learning_rate": 3.197538348747927e-06,
      "loss": 0.3822,
      "step": 284
    },
    {
      "epoch": 1.2391304347826086,
      "grad_norm": 3.271690860245072,
      "learning_rate": 3.1657289059347184e-06,
      "loss": 0.3265,
      "step": 285
    },
    {
      "epoch": 1.2434782608695651,
      "grad_norm": 8.121167597752617,
      "learning_rate": 3.1340050182240438e-06,
      "loss": 0.3817,
      "step": 286
    },
    {
      "epoch": 1.2478260869565219,
      "grad_norm": 6.009272476981181,
      "learning_rate": 3.1023681652992925e-06,
      "loss": 0.4827,
      "step": 287
    },
    {
      "epoch": 1.2521739130434781,
      "grad_norm": 4.551287991303892,
      "learning_rate": 3.070819822784323e-06,
      "loss": 0.333,
      "step": 288
    },
    {
      "epoch": 1.2565217391304349,
      "grad_norm": 4.40493888329962,
      "learning_rate": 3.03936146217465e-06,
      "loss": 0.2985,
      "step": 289
    },
    {
      "epoch": 1.2608695652173914,
      "grad_norm": 7.457602998331675,
      "learning_rate": 3.007994550768793e-06,
      "loss": 0.2688,
      "step": 290
    },
    {
      "epoch": 1.2652173913043478,
      "grad_norm": 3.5245316191041773,
      "learning_rate": 2.976720551599852e-06,
      "loss": 0.3413,
      "step": 291
    },
    {
      "epoch": 1.2695652173913043,
      "grad_norm": 5.458404782644148,
      "learning_rate": 2.9455409233672594e-06,
      "loss": 0.5688,
      "step": 292
    },
    {
      "epoch": 1.2739130434782608,
      "grad_norm": 5.849014777064085,
      "learning_rate": 2.914457120368748e-06,
      "loss": 0.5278,
      "step": 293
    },
    {
      "epoch": 1.2782608695652173,
      "grad_norm": 4.524775615715545,
      "learning_rate": 2.883470592432512e-06,
      "loss": 0.3695,
      "step": 294
    },
    {
      "epoch": 1.2826086956521738,
      "grad_norm": 3.0639485751220197,
      "learning_rate": 2.8525827848495912e-06,
      "loss": 0.3673,
      "step": 295
    },
    {
      "epoch": 1.2869565217391306,
      "grad_norm": 4.854031752617082,
      "learning_rate": 2.8217951383064546e-06,
      "loss": 0.3701,
      "step": 296
    },
    {
      "epoch": 1.2913043478260868,
      "grad_norm": 4.277779062420338,
      "learning_rate": 2.7911090888178033e-06,
      "loss": 0.4181,
      "step": 297
    },
    {
      "epoch": 1.2956521739130435,
      "grad_norm": 4.347824665030716,
      "learning_rate": 2.760526067659591e-06,
      "loss": 0.4506,
      "step": 298
    },
    {
      "epoch": 1.3,
      "grad_norm": 4.68297313331208,
      "learning_rate": 2.7300475013022666e-06,
      "loss": 0.45,
      "step": 299
    },
    {
      "epoch": 1.3043478260869565,
      "grad_norm": 3.0283941903778895,
      "learning_rate": 2.6996748113442397e-06,
      "loss": 0.2597,
      "step": 300
    },
    {
      "epoch": 1.308695652173913,
      "grad_norm": 4.491981965834742,
      "learning_rate": 2.669409414445574e-06,
      "loss": 0.382,
      "step": 301
    },
    {
      "epoch": 1.3130434782608695,
      "grad_norm": 3.990717626507738,
      "learning_rate": 2.6392527222619078e-06,
      "loss": 0.3516,
      "step": 302
    },
    {
      "epoch": 1.317391304347826,
      "grad_norm": 4.136276302818288,
      "learning_rate": 2.6092061413786158e-06,
      "loss": 0.3829,
      "step": 303
    },
    {
      "epoch": 1.3217391304347825,
      "grad_norm": 6.106967961277055,
      "learning_rate": 2.5792710732452e-06,
      "loss": 0.4728,
      "step": 304
    },
    {
      "epoch": 1.3260869565217392,
      "grad_norm": 3.5028685987260175,
      "learning_rate": 2.5494489141099155e-06,
      "loss": 0.2598,
      "step": 305
    },
    {
      "epoch": 1.3304347826086955,
      "grad_norm": 4.559833842039468,
      "learning_rate": 2.5197410549546598e-06,
      "loss": 0.253,
      "step": 306
    },
    {
      "epoch": 1.3347826086956522,
      "grad_norm": 4.5448409327910415,
      "learning_rate": 2.4901488814300855e-06,
      "loss": 0.4016,
      "step": 307
    },
    {
      "epoch": 1.3391304347826087,
      "grad_norm": 3.0979476281469265,
      "learning_rate": 2.4606737737909696e-06,
      "loss": 0.2321,
      "step": 308
    },
    {
      "epoch": 1.3434782608695652,
      "grad_norm": 5.981685141898299,
      "learning_rate": 2.431317106831836e-06,
      "loss": 0.5274,
      "step": 309
    },
    {
      "epoch": 1.3478260869565217,
      "grad_norm": 2.98685529750352,
      "learning_rate": 2.4020802498228333e-06,
      "loss": 0.396,
      "step": 310
    },
    {
      "epoch": 1.3521739130434782,
      "grad_norm": 5.125431488555637,
      "learning_rate": 2.3729645664458637e-06,
      "loss": 0.505,
      "step": 311
    },
    {
      "epoch": 1.3565217391304347,
      "grad_norm": 5.094394216531229,
      "learning_rate": 2.3439714147309845e-06,
      "loss": 0.4012,
      "step": 312
    },
    {
      "epoch": 1.3608695652173912,
      "grad_norm": 4.289708162043688,
      "learning_rate": 2.315102146993061e-06,
      "loss": 0.2589,
      "step": 313
    },
    {
      "epoch": 1.365217391304348,
      "grad_norm": 4.3432839576954745,
      "learning_rate": 2.286358109768693e-06,
      "loss": 0.3326,
      "step": 314
    },
    {
      "epoch": 1.3695652173913042,
      "grad_norm": 4.063212467164958,
      "learning_rate": 2.2577406437534055e-06,
      "loss": 0.2893,
      "step": 315
    },
    {
      "epoch": 1.373913043478261,
      "grad_norm": 3.1836830488379033,
      "learning_rate": 2.229251083739127e-06,
      "loss": 0.2955,
      "step": 316
    },
    {
      "epoch": 1.3782608695652174,
      "grad_norm": 4.9246519028776685,
      "learning_rate": 2.2008907585519094e-06,
      "loss": 0.4412,
      "step": 317
    },
    {
      "epoch": 1.382608695652174,
      "grad_norm": 5.147968019536769,
      "learning_rate": 2.172660990989971e-06,
      "loss": 0.3728,
      "step": 318
    },
    {
      "epoch": 1.3869565217391304,
      "grad_norm": 5.0886629945359365,
      "learning_rate": 2.144563097761984e-06,
      "loss": 0.4549,
      "step": 319
    },
    {
      "epoch": 1.391304347826087,
      "grad_norm": 4.110328785074501,
      "learning_rate": 2.1165983894256647e-06,
      "loss": 0.399,
      "step": 320
    },
    {
      "epoch": 1.3956521739130434,
      "grad_norm": 5.678018080542645,
      "learning_rate": 2.0887681703266453e-06,
      "loss": 0.4922,
      "step": 321
    },
    {
      "epoch": 1.4,
      "grad_norm": 3.522299014746889,
      "learning_rate": 2.061073738537635e-06,
      "loss": 0.2851,
      "step": 322
    },
    {
      "epoch": 1.4043478260869566,
      "grad_norm": 6.210071220589272,
      "learning_rate": 2.0335163857978747e-06,
      "loss": 0.4381,
      "step": 323
    },
    {
      "epoch": 1.4086956521739131,
      "grad_norm": 4.677651092740623,
      "learning_rate": 2.0060973974528873e-06,
      "loss": 0.3304,
      "step": 324
    },
    {
      "epoch": 1.4130434782608696,
      "grad_norm": 3.7070115905602585,
      "learning_rate": 1.978818052394528e-06,
      "loss": 0.2646,
      "step": 325
    },
    {
      "epoch": 1.4173913043478261,
      "grad_norm": 4.375456577369518,
      "learning_rate": 1.9516796230013275e-06,
      "loss": 0.2989,
      "step": 326
    },
    {
      "epoch": 1.4217391304347826,
      "grad_norm": 5.570245887780959,
      "learning_rate": 1.9246833750791526e-06,
      "loss": 0.4208,
      "step": 327
    },
    {
      "epoch": 1.4260869565217391,
      "grad_norm": 4.886938750299713,
      "learning_rate": 1.8978305678021598e-06,
      "loss": 0.4118,
      "step": 328
    },
    {
      "epoch": 1.4304347826086956,
      "grad_norm": 4.620746234726746,
      "learning_rate": 1.8711224536540678e-06,
      "loss": 0.2954,
      "step": 329
    },
    {
      "epoch": 1.434782608695652,
      "grad_norm": 4.633932338754762,
      "learning_rate": 1.8445602783697375e-06,
      "loss": 0.4206,
      "step": 330
    },
    {
      "epoch": 1.4391304347826086,
      "grad_norm": 3.999563113243534,
      "learning_rate": 1.8181452808770638e-06,
      "loss": 0.3652,
      "step": 331
    },
    {
      "epoch": 1.4434782608695653,
      "grad_norm": 4.232490038953125,
      "learning_rate": 1.7918786932391945e-06,
      "loss": 0.4048,
      "step": 332
    },
    {
      "epoch": 1.4478260869565218,
      "grad_norm": 4.376778688265914,
      "learning_rate": 1.765761740597065e-06,
      "loss": 0.5538,
      "step": 333
    },
    {
      "epoch": 1.4521739130434783,
      "grad_norm": 3.6752993411272086,
      "learning_rate": 1.739795641112248e-06,
      "loss": 0.4687,
      "step": 334
    },
    {
      "epoch": 1.4565217391304348,
      "grad_norm": 4.55460302537388,
      "learning_rate": 1.7139816059101372e-06,
      "loss": 0.7519,
      "step": 335
    },
    {
      "epoch": 1.4608695652173913,
      "grad_norm": 5.114759853944812,
      "learning_rate": 1.688320839023463e-06,
      "loss": 0.3998,
      "step": 336
    },
    {
      "epoch": 1.4652173913043478,
      "grad_norm": 3.1716648657362128,
      "learning_rate": 1.662814537336122e-06,
      "loss": 0.3869,
      "step": 337
    },
    {
      "epoch": 1.4695652173913043,
      "grad_norm": 3.7587234242982905,
      "learning_rate": 1.6374638905273643e-06,
      "loss": 0.3841,
      "step": 338
    },
    {
      "epoch": 1.4739130434782608,
      "grad_norm": 3.63611176729071,
      "learning_rate": 1.6122700810162967e-06,
      "loss": 0.3801,
      "step": 339
    },
    {
      "epoch": 1.4782608695652173,
      "grad_norm": 4.796431057792193,
      "learning_rate": 1.5872342839067305e-06,
      "loss": 0.2504,
      "step": 340
    },
    {
      "epoch": 1.482608695652174,
      "grad_norm": 4.964948958311623,
      "learning_rate": 1.5623576669323743e-06,
      "loss": 0.4862,
      "step": 341
    },
    {
      "epoch": 1.4869565217391305,
      "grad_norm": 3.9303479836058464,
      "learning_rate": 1.5376413904023723e-06,
      "loss": 0.3774,
      "step": 342
    },
    {
      "epoch": 1.491304347826087,
      "grad_norm": 4.11021434418603,
      "learning_rate": 1.5130866071471717e-06,
      "loss": 0.3495,
      "step": 343
    },
    {
      "epoch": 1.4956521739130435,
      "grad_norm": 4.0020770636229575,
      "learning_rate": 1.4886944624647647e-06,
      "loss": 0.2751,
      "step": 344
    },
    {
      "epoch": 1.5,
      "grad_norm": 5.903091442855636,
      "learning_rate": 1.4644660940672628e-06,
      "loss": 0.4938,
      "step": 345
    },
    {
      "epoch": 1.5043478260869565,
      "grad_norm": 4.598078158755451,
      "learning_rate": 1.4404026320278318e-06,
      "loss": 0.2931,
      "step": 346
    },
    {
      "epoch": 1.508695652173913,
      "grad_norm": 6.359179166789268,
      "learning_rate": 1.4165051987279832e-06,
      "loss": 0.4609,
      "step": 347
    },
    {
      "epoch": 1.5130434782608697,
      "grad_norm": 3.7748104229756305,
      "learning_rate": 1.3927749088052218e-06,
      "loss": 0.4253,
      "step": 348
    },
    {
      "epoch": 1.517391304347826,
      "grad_norm": 4.449981154160408,
      "learning_rate": 1.3692128691010592e-06,
      "loss": 0.3325,
      "step": 349
    },
    {
      "epoch": 1.5217391304347827,
      "grad_norm": 5.0666604676427305,
      "learning_rate": 1.3458201786093795e-06,
      "loss": 0.4088,
      "step": 350
    },
    {
      "epoch": 1.526086956521739,
      "grad_norm": 4.630295107740943,
      "learning_rate": 1.3225979284251955e-06,
      "loss": 0.2542,
      "step": 351
    },
    {
      "epoch": 1.5304347826086957,
      "grad_norm": 4.7269256098025085,
      "learning_rate": 1.2995472016937405e-06,
      "loss": 0.3752,
      "step": 352
    },
    {
      "epoch": 1.5347826086956522,
      "grad_norm": 4.340635258567047,
      "learning_rate": 1.2766690735599569e-06,
      "loss": 0.408,
      "step": 353
    },
    {
      "epoch": 1.5391304347826087,
      "grad_norm": 4.770896513431023,
      "learning_rate": 1.2539646111183452e-06,
      "loss": 0.29,
      "step": 354
    },
    {
      "epoch": 1.5434782608695652,
      "grad_norm": 4.228926300066491,
      "learning_rate": 1.2314348733631958e-06,
      "loss": 0.3813,
      "step": 355
    },
    {
      "epoch": 1.5478260869565217,
      "grad_norm": 2.7862114049851052,
      "learning_rate": 1.209080911139187e-06,
      "loss": 0.2217,
      "step": 356
    },
    {
      "epoch": 1.5521739130434784,
      "grad_norm": 4.834499117010849,
      "learning_rate": 1.1869037670923817e-06,
      "loss": 0.3568,
      "step": 357
    },
    {
      "epoch": 1.5565217391304347,
      "grad_norm": 4.436098328818527,
      "learning_rate": 1.1649044756215872e-06,
      "loss": 0.4675,
      "step": 358
    },
    {
      "epoch": 1.5608695652173914,
      "grad_norm": 3.505364253769214,
      "learning_rate": 1.1430840628301093e-06,
      "loss": 0.4031,
      "step": 359
    },
    {
      "epoch": 1.5652173913043477,
      "grad_norm": 3.5337346862290686,
      "learning_rate": 1.1214435464779006e-06,
      "loss": 0.4156,
      "step": 360
    },
    {
      "epoch": 1.5695652173913044,
      "grad_norm": 6.840098625250913,
      "learning_rate": 1.099983935934077e-06,
      "loss": 0.4857,
      "step": 361
    },
    {
      "epoch": 1.5739130434782609,
      "grad_norm": 6.213293552584588,
      "learning_rate": 1.0787062321298441e-06,
      "loss": 0.4374,
      "step": 362
    },
    {
      "epoch": 1.5782608695652174,
      "grad_norm": 5.404535179527454,
      "learning_rate": 1.0576114275118132e-06,
      "loss": 0.3358,
      "step": 363
    },
    {
      "epoch": 1.5826086956521739,
      "grad_norm": 3.3759662275329965,
      "learning_rate": 1.0367005059957097e-06,
      "loss": 0.3992,
      "step": 364
    },
    {
      "epoch": 1.5869565217391304,
      "grad_norm": 4.33324237420986,
      "learning_rate": 1.0159744429204776e-06,
      "loss": 0.3414,
      "step": 365
    },
    {
      "epoch": 1.591304347826087,
      "grad_norm": 8.014303432550552,
      "learning_rate": 9.954342050027922e-07,
      "loss": 0.5959,
      "step": 366
    },
    {
      "epoch": 1.5956521739130434,
      "grad_norm": 5.063352120229355,
      "learning_rate": 9.75080750291965e-07,
      "loss": 0.4683,
      "step": 367
    },
    {
      "epoch": 1.6,
      "grad_norm": 2.6582958542181774,
      "learning_rate": 9.549150281252633e-07,
      "loss": 0.3139,
      "step": 368
    },
    {
      "epoch": 1.6043478260869564,
      "grad_norm": 3.534876787372464,
      "learning_rate": 9.349379790836243e-07,
      "loss": 0.3933,
      "step": 369
    },
    {
      "epoch": 1.608695652173913,
      "grad_norm": 4.091568890168747,
      "learning_rate": 9.151505349477901e-07,
      "loss": 0.4115,
      "step": 370
    },
    {
      "epoch": 1.6130434782608696,
      "grad_norm": 4.587879328576348,
      "learning_rate": 8.955536186548425e-07,
      "loss": 0.4574,
      "step": 371
    },
    {
      "epoch": 1.617391304347826,
      "grad_norm": 4.11512437136083,
      "learning_rate": 8.761481442551573e-07,
      "loss": 0.4054,
      "step": 372
    },
    {
      "epoch": 1.6217391304347826,
      "grad_norm": 4.948336357859377,
      "learning_rate": 8.569350168697705e-07,
      "loss": 0.4593,
      "step": 373
    },
    {
      "epoch": 1.626086956521739,
      "grad_norm": 4.571512826347378,
      "learning_rate": 8.379151326481588e-07,
      "loss": 0.3444,
      "step": 374
    },
    {
      "epoch": 1.6304347826086958,
      "grad_norm": 4.328104628253105,
      "learning_rate": 8.19089378726447e-07,
      "loss": 0.4923,
      "step": 375
    },
    {
      "epoch": 1.634782608695652,
      "grad_norm": 3.760449059509458,
      "learning_rate": 8.004586331860176e-07,
      "loss": 0.2563,
      "step": 376
    },
    {
      "epoch": 1.6391304347826088,
      "grad_norm": 4.322342061262116,
      "learning_rate": 7.820237650125711e-07,
      "loss": 0.2826,
      "step": 377
    },
    {
      "epoch": 1.643478260869565,
      "grad_norm": 5.664753766424785,
      "learning_rate": 7.637856340555822e-07,
      "loss": 0.5416,
      "step": 378
    },
    {
      "epoch": 1.6478260869565218,
      "grad_norm": 4.168173552178201,
      "learning_rate": 7.457450909881969e-07,
      "loss": 0.4541,
      "step": 379
    },
    {
      "epoch": 1.6521739130434783,
      "grad_norm": 3.746985781310187,
      "learning_rate": 7.279029772675572e-07,
      "loss": 0.3148,
      "step": 380
    },
    {
      "epoch": 1.6565217391304348,
      "grad_norm": 4.0890307040702,
      "learning_rate": 7.102601250955526e-07,
      "loss": 0.2536,
      "step": 381
    },
    {
      "epoch": 1.6608695652173913,
      "grad_norm": 4.147467566269263,
      "learning_rate": 6.928173573800007e-07,
      "loss": 0.5819,
      "step": 382
    },
    {
      "epoch": 1.6652173913043478,
      "grad_norm": 3.470213468989912,
      "learning_rate": 6.755754876962711e-07,
      "loss": 0.3264,
      "step": 383
    },
    {
      "epoch": 1.6695652173913045,
      "grad_norm": 3.790885229353616,
      "learning_rate": 6.585353202493322e-07,
      "loss": 0.4863,
      "step": 384
    },
    {
      "epoch": 1.6739130434782608,
      "grad_norm": 3.2098499133289873,
      "learning_rate": 6.416976498362432e-07,
      "loss": 0.4062,
      "step": 385
    },
    {
      "epoch": 1.6782608695652175,
      "grad_norm": 3.052567811452033,
      "learning_rate": 6.250632618090868e-07,
      "loss": 0.3766,
      "step": 386
    },
    {
      "epoch": 1.6826086956521737,
      "grad_norm": 3.251506558261061,
      "learning_rate": 6.086329320383311e-07,
      "loss": 0.2694,
      "step": 387
    },
    {
      "epoch": 1.6869565217391305,
      "grad_norm": 4.429111727950372,
      "learning_rate": 5.924074268766422e-07,
      "loss": 0.3576,
      "step": 388
    },
    {
      "epoch": 1.691304347826087,
      "grad_norm": 5.864544691259148,
      "learning_rate": 5.763875031231464e-07,
      "loss": 0.4093,
      "step": 389
    },
    {
      "epoch": 1.6956521739130435,
      "grad_norm": 3.999472642222693,
      "learning_rate": 5.60573907988124e-07,
      "loss": 0.2376,
      "step": 390
    },
    {
      "epoch": 1.7,
      "grad_norm": 4.508948892750066,
      "learning_rate": 5.449673790581611e-07,
      "loss": 0.3957,
      "step": 391
    },
    {
      "epoch": 1.7043478260869565,
      "grad_norm": 4.096578412910045,
      "learning_rate": 5.295686442617442e-07,
      "loss": 0.3033,
      "step": 392
    },
    {
      "epoch": 1.7086956521739132,
      "grad_norm": 3.958895722894547,
      "learning_rate": 5.143784218353104e-07,
      "loss": 0.59,
      "step": 393
    },
    {
      "epoch": 1.7130434782608694,
      "grad_norm": 3.6190572925004485,
      "learning_rate": 4.993974202897456e-07,
      "loss": 0.473,
      "step": 394
    },
    {
      "epoch": 1.7173913043478262,
      "grad_norm": 5.161364474058307,
      "learning_rate": 4.846263383773364e-07,
      "loss": 0.3472,
      "step": 395
    },
    {
      "epoch": 1.7217391304347827,
      "grad_norm": 4.226485622410246,
      "learning_rate": 4.7006586505918273e-07,
      "loss": 0.3042,
      "step": 396
    },
    {
      "epoch": 1.7260869565217392,
      "grad_norm": 3.9403605213034347,
      "learning_rate": 4.557166794730572e-07,
      "loss": 0.3942,
      "step": 397
    },
    {
      "epoch": 1.7304347826086957,
      "grad_norm": 3.464344850791239,
      "learning_rate": 4.4157945090173294e-07,
      "loss": 0.3491,
      "step": 398
    },
    {
      "epoch": 1.7347826086956522,
      "grad_norm": 3.6832590658291253,
      "learning_rate": 4.276548387417656e-07,
      "loss": 0.4446,
      "step": 399
    },
    {
      "epoch": 1.7391304347826086,
      "grad_norm": 3.9257122718747173,
      "learning_rate": 4.139434924727359e-07,
      "loss": 0.1904,
      "step": 400
    },
    {
      "epoch": 1.7391304347826086,
      "eval_loss": 0.84710693359375,
      "eval_runtime": 2.9637,
      "eval_samples_per_second": 6.411,
      "eval_steps_per_second": 1.687,
      "step": 400
    },
    {
      "epoch": 1.7434782608695651,
      "grad_norm": 4.2238949061823625,
      "learning_rate": 4.004460516269554e-07,
      "loss": 0.2747,
      "step": 401
    },
    {
      "epoch": 1.7478260869565219,
      "grad_norm": 4.415499878703038,
      "learning_rate": 3.8716314575964197e-07,
      "loss": 0.434,
      "step": 402
    },
    {
      "epoch": 1.7521739130434781,
      "grad_norm": 4.413123149635052,
      "learning_rate": 3.740953944195497e-07,
      "loss": 0.3275,
      "step": 403
    },
    {
      "epoch": 1.7565217391304349,
      "grad_norm": 5.005472770791753,
      "learning_rate": 3.612434071200771e-07,
      "loss": 0.4393,
      "step": 404
    },
    {
      "epoch": 1.7608695652173914,
      "grad_norm": 4.77916502918349,
      "learning_rate": 3.486077833108342e-07,
      "loss": 0.3882,
      "step": 405
    },
    {
      "epoch": 1.7652173913043478,
      "grad_norm": 4.888332155893412,
      "learning_rate": 3.361891123496824e-07,
      "loss": 0.2972,
      "step": 406
    },
    {
      "epoch": 1.7695652173913043,
      "grad_norm": 8.555798239045828,
      "learning_rate": 3.2398797347524656e-07,
      "loss": 0.416,
      "step": 407
    },
    {
      "epoch": 1.7739130434782608,
      "grad_norm": 4.311854645200206,
      "learning_rate": 3.1200493577989875e-07,
      "loss": 0.4497,
      "step": 408
    },
    {
      "epoch": 1.7782608695652173,
      "grad_norm": 4.916122775051447,
      "learning_rate": 3.002405581832135e-07,
      "loss": 0.639,
      "step": 409
    },
    {
      "epoch": 1.7826086956521738,
      "grad_norm": 5.578493709235784,
      "learning_rate": 2.88695389405898e-07,
      "loss": 0.5367,
      "step": 410
    },
    {
      "epoch": 1.7869565217391306,
      "grad_norm": 4.301048656163112,
      "learning_rate": 2.7736996794419767e-07,
      "loss": 0.4093,
      "step": 411
    },
    {
      "epoch": 1.7913043478260868,
      "grad_norm": 4.646134977327319,
      "learning_rate": 2.662648220447811e-07,
      "loss": 0.4533,
      "step": 412
    },
    {
      "epoch": 1.7956521739130435,
      "grad_norm": 4.061060226299757,
      "learning_rate": 2.5538046968010097e-07,
      "loss": 0.417,
      "step": 413
    },
    {
      "epoch": 1.8,
      "grad_norm": 4.907143921361194,
      "learning_rate": 2.447174185242324e-07,
      "loss": 0.4717,
      "step": 414
    },
    {
      "epoch": 1.8043478260869565,
      "grad_norm": 4.6576414692745445,
      "learning_rate": 2.3427616592919587e-07,
      "loss": 0.1985,
      "step": 415
    },
    {
      "epoch": 1.808695652173913,
      "grad_norm": 7.551034297687061,
      "learning_rate": 2.240571989017598e-07,
      "loss": 0.2973,
      "step": 416
    },
    {
      "epoch": 1.8130434782608695,
      "grad_norm": 4.711055462866289,
      "learning_rate": 2.1406099408072256e-07,
      "loss": 0.332,
      "step": 417
    },
    {
      "epoch": 1.8173913043478263,
      "grad_norm": 5.451809254942853,
      "learning_rate": 2.0428801771468388e-07,
      "loss": 0.4386,
      "step": 418
    },
    {
      "epoch": 1.8217391304347825,
      "grad_norm": 4.973253351932573,
      "learning_rate": 1.947387256402966e-07,
      "loss": 0.4752,
      "step": 419
    },
    {
      "epoch": 1.8260869565217392,
      "grad_norm": 5.230761253263795,
      "learning_rate": 1.8541356326100436e-07,
      "loss": 0.4217,
      "step": 420
    },
    {
      "epoch": 1.8304347826086955,
      "grad_norm": 5.234189528741744,
      "learning_rate": 1.7631296552626687e-07,
      "loss": 0.4596,
      "step": 421
    },
    {
      "epoch": 1.8347826086956522,
      "grad_norm": 8.698679258939933,
      "learning_rate": 1.6743735691127639e-07,
      "loss": 0.3867,
      "step": 422
    },
    {
      "epoch": 1.8391304347826087,
      "grad_norm": 4.0581036301103826,
      "learning_rate": 1.5878715139715395e-07,
      "loss": 0.3501,
      "step": 423
    },
    {
      "epoch": 1.8434782608695652,
      "grad_norm": 3.3295020409277134,
      "learning_rate": 1.5036275245164377e-07,
      "loss": 0.333,
      "step": 424
    },
    {
      "epoch": 1.8478260869565217,
      "grad_norm": 5.625873750329853,
      "learning_rate": 1.4216455301029274e-07,
      "loss": 0.2544,
      "step": 425
    },
    {
      "epoch": 1.8521739130434782,
      "grad_norm": 21.31563926936729,
      "learning_rate": 1.341929354581234e-07,
      "loss": 0.3744,
      "step": 426
    },
    {
      "epoch": 1.856521739130435,
      "grad_norm": 3.6645244784937554,
      "learning_rate": 1.2644827161179763e-07,
      "loss": 0.3328,
      "step": 427
    },
    {
      "epoch": 1.8608695652173912,
      "grad_norm": 4.035545215811824,
      "learning_rate": 1.1893092270227724e-07,
      "loss": 0.3116,
      "step": 428
    },
    {
      "epoch": 1.865217391304348,
      "grad_norm": 4.778912796903033,
      "learning_rate": 1.1164123935797189e-07,
      "loss": 0.4285,
      "step": 429
    },
    {
      "epoch": 1.8695652173913042,
      "grad_norm": 4.365193705408143,
      "learning_rate": 1.0457956158838545e-07,
      "loss": 0.3649,
      "step": 430
    },
    {
      "epoch": 1.873913043478261,
      "grad_norm": 4.931094956899494,
      "learning_rate": 9.774621876825985e-08,
      "loss": 0.569,
      "step": 431
    },
    {
      "epoch": 1.8782608695652174,
      "grad_norm": 4.916680811083485,
      "learning_rate": 9.114152962220734e-08,
      "loss": 0.4065,
      "step": 432
    },
    {
      "epoch": 1.882608695652174,
      "grad_norm": 5.327664249364534,
      "learning_rate": 8.476580220984854e-08,
      "loss": 0.4887,
      "step": 433
    },
    {
      "epoch": 1.8869565217391304,
      "grad_norm": 5.630899155916819,
      "learning_rate": 7.861933391144272e-08,
      "loss": 0.3431,
      "step": 434
    },
    {
      "epoch": 1.891304347826087,
      "grad_norm": 4.192157519887303,
      "learning_rate": 7.270241141401568e-08,
      "loss": 0.4561,
      "step": 435
    },
    {
      "epoch": 1.8956521739130436,
      "grad_norm": 3.3402242500527346,
      "learning_rate": 6.701531069799039e-08,
      "loss": 0.2535,
      "step": 436
    },
    {
      "epoch": 1.9,
      "grad_norm": 5.05505766202891,
      "learning_rate": 6.15582970243117e-08,
      "loss": 0.4571,
      "step": 437
    },
    {
      "epoch": 1.9043478260869566,
      "grad_norm": 4.482518001328448,
      "learning_rate": 5.633162492207633e-08,
      "loss": 0.4594,
      "step": 438
    },
    {
      "epoch": 1.908695652173913,
      "grad_norm": 4.747721812858791,
      "learning_rate": 5.133553817665948e-08,
      "loss": 0.4966,
      "step": 439
    },
    {
      "epoch": 1.9130434782608696,
      "grad_norm": 5.406074579801078,
      "learning_rate": 4.657026981834623e-08,
      "loss": 0.4339,
      "step": 440
    },
    {
      "epoch": 1.9173913043478261,
      "grad_norm": 3.305337867337301,
      "learning_rate": 4.203604211145851e-08,
      "loss": 0.2366,
      "step": 441
    },
    {
      "epoch": 1.9217391304347826,
      "grad_norm": 3.4638544230453303,
      "learning_rate": 3.773306654399234e-08,
      "loss": 0.3431,
      "step": 442
    },
    {
      "epoch": 1.9260869565217391,
      "grad_norm": 4.732049116559757,
      "learning_rate": 3.366154381775011e-08,
      "loss": 0.4252,
      "step": 443
    },
    {
      "epoch": 1.9304347826086956,
      "grad_norm": 6.208168501352479,
      "learning_rate": 2.9821663838981994e-08,
      "loss": 0.4341,
      "step": 444
    },
    {
      "epoch": 1.9347826086956523,
      "grad_norm": 3.631999073133573,
      "learning_rate": 2.6213605709525803e-08,
      "loss": 0.2918,
      "step": 445
    },
    {
      "epoch": 1.9391304347826086,
      "grad_norm": 3.4071911648402047,
      "learning_rate": 2.283753771845587e-08,
      "loss": 0.2576,
      "step": 446
    },
    {
      "epoch": 1.9434782608695653,
      "grad_norm": 4.341970244912221,
      "learning_rate": 1.969361733423103e-08,
      "loss": 0.4952,
      "step": 447
    },
    {
      "epoch": 1.9478260869565216,
      "grad_norm": 5.0988876739070665,
      "learning_rate": 1.6781991197352133e-08,
      "loss": 0.3971,
      "step": 448
    },
    {
      "epoch": 1.9521739130434783,
      "grad_norm": 5.344651969376342,
      "learning_rate": 1.4102795113520307e-08,
      "loss": 0.3179,
      "step": 449
    },
    {
      "epoch": 1.9565217391304348,
      "grad_norm": 3.1321296957399816,
      "learning_rate": 1.1656154047303691e-08,
      "loss": 0.3592,
      "step": 450
    },
    {
      "epoch": 1.9608695652173913,
      "grad_norm": 5.289764246110828,
      "learning_rate": 9.442182116309872e-09,
      "loss": 0.5675,
      "step": 451
    },
    {
      "epoch": 1.9652173913043478,
      "grad_norm": 3.539029704298015,
      "learning_rate": 7.460982585860144e-09,
      "loss": 0.2401,
      "step": 452
    },
    {
      "epoch": 1.9695652173913043,
      "grad_norm": 3.3170116889868084,
      "learning_rate": 5.712647864176135e-09,
      "loss": 0.4096,
      "step": 453
    },
    {
      "epoch": 1.973913043478261,
      "grad_norm": 4.591533334025833,
      "learning_rate": 4.197259498067707e-09,
      "loss": 0.3984,
      "step": 454
    },
    {
      "epoch": 1.9782608695652173,
      "grad_norm": 3.485313689550101,
      "learning_rate": 2.9148881691298812e-09,
      "loss": 0.3549,
      "step": 455
    },
    {
      "epoch": 1.982608695652174,
      "grad_norm": 5.597770599873441,
      "learning_rate": 1.865593690446588e-09,
      "loss": 0.3956,
      "step": 456
    },
    {
      "epoch": 1.9869565217391303,
      "grad_norm": 3.9083218733863356,
      "learning_rate": 1.0494250038006747e-09,
      "loss": 0.3444,
      "step": 457
    },
    {
      "epoch": 1.991304347826087,
      "grad_norm": 5.853251874420562,
      "learning_rate": 4.664201773896259e-10,
      "loss": 0.373,
      "step": 458
    },
    {
      "epoch": 1.9956521739130435,
      "grad_norm": 4.471869039132135,
      "learning_rate": 1.1660640405308787e-10,
      "loss": 0.4927,
      "step": 459
    },
    {
      "epoch": 2.0,
      "grad_norm": 4.073032878536336,
      "learning_rate": 0.0,
      "loss": 0.3028,
      "step": 460
    },
    {
      "epoch": 2.0,
      "step": 460,
      "total_flos": 3525630345216.0,
      "train_loss": 0.5884321572663991,
      "train_runtime": 1069.0797,
      "train_samples_per_second": 3.442,
      "train_steps_per_second": 0.43
    }
  ],
  "logging_steps": 1,
  "max_steps": 460,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 5000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 3525630345216.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}