{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 230,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.004347826086956522,
      "grad_norm": 210.10928344726562,
      "learning_rate": 0.0,
      "loss": 5.8188,
      "step": 1
    },
    {
      "epoch": 0.008695652173913044,
      "grad_norm": 216.5006561279297,
      "learning_rate": 2.173913043478261e-06,
      "loss": 5.9259,
      "step": 2
    },
    {
      "epoch": 0.013043478260869565,
      "grad_norm": 144.48963928222656,
      "learning_rate": 4.347826086956522e-06,
      "loss": 5.646,
      "step": 3
    },
    {
      "epoch": 0.017391304347826087,
      "grad_norm": 45.486934661865234,
      "learning_rate": 6.521739130434783e-06,
      "loss": 5.3097,
      "step": 4
    },
    {
      "epoch": 0.021739130434782608,
      "grad_norm": 83.79264831542969,
      "learning_rate": 8.695652173913044e-06,
      "loss": 5.3505,
      "step": 5
    },
    {
      "epoch": 0.02608695652173913,
      "grad_norm": 33.744483947753906,
      "learning_rate": 1.0869565217391305e-05,
      "loss": 5.1314,
      "step": 6
    },
    {
      "epoch": 0.030434782608695653,
      "grad_norm": 22.175418853759766,
      "learning_rate": 1.3043478260869566e-05,
      "loss": 4.8346,
      "step": 7
    },
    {
      "epoch": 0.034782608695652174,
      "grad_norm": 18.40424156188965,
      "learning_rate": 1.5217391304347828e-05,
      "loss": 4.7562,
      "step": 8
    },
    {
      "epoch": 0.0391304347826087,
      "grad_norm": 15.772565841674805,
      "learning_rate": 1.739130434782609e-05,
      "loss": 4.5057,
      "step": 9
    },
    {
      "epoch": 0.043478260869565216,
      "grad_norm": 11.410517692565918,
      "learning_rate": 1.956521739130435e-05,
      "loss": 4.3231,
      "step": 10
    },
    {
      "epoch": 0.04782608695652174,
      "grad_norm": 14.64340877532959,
      "learning_rate": 2.173913043478261e-05,
      "loss": 4.3797,
      "step": 11
    },
    {
      "epoch": 0.05217391304347826,
      "grad_norm": 7.4696946144104,
      "learning_rate": 2.391304347826087e-05,
      "loss": 3.9548,
      "step": 12
    },
    {
      "epoch": 0.05652173913043478,
      "grad_norm": 3.1422557830810547,
      "learning_rate": 2.608695652173913e-05,
      "loss": 3.8226,
      "step": 13
    },
    {
      "epoch": 0.06086956521739131,
      "grad_norm": 2.6594135761260986,
      "learning_rate": 2.826086956521739e-05,
      "loss": 3.8783,
      "step": 14
    },
    {
      "epoch": 0.06521739130434782,
      "grad_norm": 2.0335605144500732,
      "learning_rate": 3.0434782608695656e-05,
      "loss": 3.626,
      "step": 15
    },
    {
      "epoch": 0.06956521739130435,
      "grad_norm": 2.045989513397217,
      "learning_rate": 3.260869565217392e-05,
      "loss": 3.4734,
      "step": 16
    },
    {
      "epoch": 0.07391304347826087,
      "grad_norm": 1.797641396522522,
      "learning_rate": 3.478260869565218e-05,
      "loss": 3.3667,
      "step": 17
    },
    {
      "epoch": 0.0782608695652174,
      "grad_norm": 1.7289575338363647,
      "learning_rate": 3.695652173913043e-05,
      "loss": 3.2171,
      "step": 18
    },
    {
      "epoch": 0.08260869565217391,
      "grad_norm": 1.6280560493469238,
      "learning_rate": 3.91304347826087e-05,
      "loss": 3.0697,
      "step": 19
    },
    {
      "epoch": 0.08695652173913043,
      "grad_norm": 1.5199931859970093,
      "learning_rate": 4.130434782608696e-05,
      "loss": 2.9537,
      "step": 20
    },
    {
      "epoch": 0.09130434782608696,
      "grad_norm": 1.4183111190795898,
      "learning_rate": 4.347826086956522e-05,
      "loss": 2.8091,
      "step": 21
    },
    {
      "epoch": 0.09565217391304348,
      "grad_norm": 1.453029990196228,
      "learning_rate": 4.565217391304348e-05,
      "loss": 2.6457,
      "step": 22
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.135553002357483,
      "learning_rate": 4.782608695652174e-05,
      "loss": 2.4701,
      "step": 23
    },
    {
      "epoch": 0.10434782608695652,
      "grad_norm": 0.9866960644721985,
      "learning_rate": 5e-05,
      "loss": 2.3948,
      "step": 24
    },
    {
      "epoch": 0.10869565217391304,
      "grad_norm": 0.8710840344429016,
      "learning_rate": 5.217391304347826e-05,
      "loss": 2.3239,
      "step": 25
    },
    {
      "epoch": 0.11304347826086956,
      "grad_norm": 0.8170456886291504,
      "learning_rate": 5.4347826086956524e-05,
      "loss": 2.1285,
      "step": 26
    },
    {
      "epoch": 0.11739130434782609,
      "grad_norm": 0.790302038192749,
      "learning_rate": 5.652173913043478e-05,
      "loss": 2.021,
      "step": 27
    },
    {
      "epoch": 0.12173913043478261,
      "grad_norm": 0.7848089933395386,
      "learning_rate": 5.869565217391305e-05,
      "loss": 1.9254,
      "step": 28
    },
    {
      "epoch": 0.12608695652173912,
      "grad_norm": 0.7707406878471375,
      "learning_rate": 6.086956521739131e-05,
      "loss": 1.8048,
      "step": 29
    },
    {
      "epoch": 0.13043478260869565,
      "grad_norm": 0.7862960696220398,
      "learning_rate": 6.304347826086957e-05,
      "loss": 1.6704,
      "step": 30
    },
    {
      "epoch": 0.13478260869565217,
      "grad_norm": 0.8184984922409058,
      "learning_rate": 6.521739130434783e-05,
      "loss": 1.5525,
      "step": 31
    },
    {
      "epoch": 0.1391304347826087,
      "grad_norm": 0.751800537109375,
      "learning_rate": 6.73913043478261e-05,
      "loss": 1.4305,
      "step": 32
    },
    {
      "epoch": 0.14347826086956522,
      "grad_norm": 0.6508727073669434,
      "learning_rate": 6.956521739130436e-05,
      "loss": 1.3082,
      "step": 33
    },
    {
      "epoch": 0.14782608695652175,
      "grad_norm": 0.5927818417549133,
      "learning_rate": 7.17391304347826e-05,
      "loss": 1.2962,
      "step": 34
    },
    {
      "epoch": 0.15217391304347827,
      "grad_norm": 0.48864519596099854,
      "learning_rate": 7.391304347826086e-05,
      "loss": 1.1943,
      "step": 35
    },
    {
      "epoch": 0.1565217391304348,
      "grad_norm": 0.43812891840934753,
      "learning_rate": 7.608695652173914e-05,
      "loss": 1.1367,
      "step": 36
    },
    {
      "epoch": 0.1608695652173913,
      "grad_norm": 0.3985790014266968,
      "learning_rate": 7.82608695652174e-05,
      "loss": 1.0961,
      "step": 37
    },
    {
      "epoch": 0.16521739130434782,
      "grad_norm": 0.3411348760128021,
      "learning_rate": 8.043478260869566e-05,
      "loss": 1.0314,
      "step": 38
    },
    {
      "epoch": 0.16956521739130434,
      "grad_norm": 0.32298171520233154,
      "learning_rate": 8.260869565217392e-05,
      "loss": 0.9771,
      "step": 39
    },
    {
      "epoch": 0.17391304347826086,
      "grad_norm": 0.30958038568496704,
      "learning_rate": 8.478260869565218e-05,
      "loss": 0.9268,
      "step": 40
    },
    {
      "epoch": 0.1782608695652174,
      "grad_norm": 0.2889741063117981,
      "learning_rate": 8.695652173913044e-05,
      "loss": 0.9256,
      "step": 41
    },
    {
      "epoch": 0.1826086956521739,
      "grad_norm": 0.24591656029224396,
      "learning_rate": 8.91304347826087e-05,
      "loss": 0.883,
      "step": 42
    },
    {
      "epoch": 0.18695652173913044,
      "grad_norm": 0.23997186124324799,
      "learning_rate": 9.130434782608696e-05,
      "loss": 0.8786,
      "step": 43
    },
    {
      "epoch": 0.19130434782608696,
      "grad_norm": 0.2006598263978958,
      "learning_rate": 9.347826086956522e-05,
      "loss": 0.8396,
      "step": 44
    },
    {
      "epoch": 0.1956521739130435,
      "grad_norm": 0.18479709327220917,
      "learning_rate": 9.565217391304348e-05,
      "loss": 0.8413,
      "step": 45
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.17641599476337433,
      "learning_rate": 9.782608695652174e-05,
      "loss": 0.8359,
      "step": 46
    },
    {
      "epoch": 0.20434782608695654,
      "grad_norm": 0.15423867106437683,
      "learning_rate": 0.0001,
      "loss": 0.8058,
      "step": 47
    },
    {
      "epoch": 0.20869565217391303,
      "grad_norm": 0.1461988240480423,
      "learning_rate": 9.999856041607731e-05,
      "loss": 0.8029,
      "step": 48
    },
    {
      "epoch": 0.21304347826086956,
      "grad_norm": 0.12839862704277039,
      "learning_rate": 9.999424174720531e-05,
      "loss": 0.7822,
      "step": 49
    },
    {
      "epoch": 0.21739130434782608,
      "grad_norm": 0.12158359587192535,
      "learning_rate": 9.998704424206746e-05,
      "loss": 0.7748,
      "step": 50
    },
    {
      "epoch": 0.2217391304347826,
      "grad_norm": 0.1291743963956833,
      "learning_rate": 9.997696831512027e-05,
      "loss": 0.7661,
      "step": 51
    },
    {
      "epoch": 0.22608695652173913,
      "grad_norm": 0.12144283205270767,
      "learning_rate": 9.99640145465694e-05,
      "loss": 0.7869,
      "step": 52
    },
    {
      "epoch": 0.23043478260869565,
      "grad_norm": 0.1100422814488411,
      "learning_rate": 9.994818368233639e-05,
      "loss": 0.7777,
      "step": 53
    },
    {
      "epoch": 0.23478260869565218,
      "grad_norm": 0.0993693619966507,
      "learning_rate": 9.992947663401548e-05,
      "loss": 0.7473,
      "step": 54
    },
    {
      "epoch": 0.2391304347826087,
      "grad_norm": 0.0941305086016655,
      "learning_rate": 9.990789447882137e-05,
      "loss": 0.7516,
      "step": 55
    },
    {
      "epoch": 0.24347826086956523,
      "grad_norm": 0.09400874376296997,
      "learning_rate": 9.988343845952697e-05,
      "loss": 0.7517,
      "step": 56
    },
    {
      "epoch": 0.24782608695652175,
      "grad_norm": 0.083980493247509,
      "learning_rate": 9.985610998439197e-05,
      "loss": 0.749,
      "step": 57
    },
    {
      "epoch": 0.25217391304347825,
      "grad_norm": 0.08494170755147934,
      "learning_rate": 9.98259106270817e-05,
      "loss": 0.7332,
      "step": 58
    },
    {
      "epoch": 0.2565217391304348,
      "grad_norm": 0.08159317076206207,
      "learning_rate": 9.979284212657657e-05,
      "loss": 0.7343,
      "step": 59
    },
    {
      "epoch": 0.2608695652173913,
      "grad_norm": 0.08313615620136261,
      "learning_rate": 9.97569063870718e-05,
      "loss": 0.7211,
      "step": 60
    },
    {
      "epoch": 0.26521739130434785,
      "grad_norm": 0.07852096855640411,
      "learning_rate": 9.971810547786793e-05,
      "loss": 0.731,
      "step": 61
    },
    {
      "epoch": 0.26956521739130435,
      "grad_norm": 0.0774468258023262,
      "learning_rate": 9.967644163325156e-05,
      "loss": 0.7198,
      "step": 62
    },
    {
      "epoch": 0.27391304347826084,
      "grad_norm": 0.07157547771930695,
      "learning_rate": 9.963191725236672e-05,
      "loss": 0.6946,
      "step": 63
    },
    {
      "epoch": 0.2782608695652174,
      "grad_norm": 0.07179877161979675,
      "learning_rate": 9.958453489907673e-05,
      "loss": 0.6983,
      "step": 64
    },
    {
      "epoch": 0.2826086956521739,
      "grad_norm": 0.13720852136611938,
      "learning_rate": 9.953429730181653e-05,
      "loss": 0.7209,
      "step": 65
    },
    {
      "epoch": 0.28695652173913044,
      "grad_norm": 0.08586138486862183,
      "learning_rate": 9.948120735343566e-05,
      "loss": 0.7022,
      "step": 66
    },
    {
      "epoch": 0.29130434782608694,
      "grad_norm": 0.06595543771982193,
      "learning_rate": 9.942526811103152e-05,
      "loss": 0.6857,
      "step": 67
    },
    {
      "epoch": 0.2956521739130435,
      "grad_norm": 0.06423239409923553,
      "learning_rate": 9.936648279577349e-05,
      "loss": 0.6924,
      "step": 68
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.07080892473459244,
      "learning_rate": 9.930485479271735e-05,
      "loss": 0.6963,
      "step": 69
    },
    {
      "epoch": 0.30434782608695654,
      "grad_norm": 0.06481339782476425,
      "learning_rate": 9.924038765061042e-05,
      "loss": 0.7055,
      "step": 70
    },
    {
      "epoch": 0.30869565217391304,
      "grad_norm": 0.07143648713827133,
      "learning_rate": 9.91730850816871e-05,
      "loss": 0.6761,
      "step": 71
    },
    {
      "epoch": 0.3130434782608696,
      "grad_norm": 0.06885742396116257,
      "learning_rate": 9.91029509614553e-05,
      "loss": 0.7111,
      "step": 72
    },
    {
      "epoch": 0.3173913043478261,
      "grad_norm": 0.06406974792480469,
      "learning_rate": 9.902998932847307e-05,
      "loss": 0.6971,
      "step": 73
    },
    {
      "epoch": 0.3217391304347826,
      "grad_norm": 0.06285955011844635,
      "learning_rate": 9.895420438411616e-05,
      "loss": 0.681,
      "step": 74
    },
    {
      "epoch": 0.32608695652173914,
      "grad_norm": 0.07179131358861923,
      "learning_rate": 9.887560049233605e-05,
      "loss": 0.7001,
      "step": 75
    },
    {
      "epoch": 0.33043478260869563,
      "grad_norm": 0.06652161478996277,
      "learning_rate": 9.879418217940873e-05,
      "loss": 0.6668,
      "step": 76
    },
    {
      "epoch": 0.3347826086956522,
      "grad_norm": 0.06445639580488205,
      "learning_rate": 9.870995413367397e-05,
      "loss": 0.6981,
      "step": 77
    },
    {
      "epoch": 0.3391304347826087,
      "grad_norm": 0.06834300607442856,
      "learning_rate": 9.862292120526535e-05,
      "loss": 0.6484,
      "step": 78
    },
    {
      "epoch": 0.34347826086956523,
      "grad_norm": 0.06481563299894333,
      "learning_rate": 9.853308840583109e-05,
      "loss": 0.6875,
      "step": 79
    },
    {
      "epoch": 0.34782608695652173,
      "grad_norm": 0.062026482075452805,
      "learning_rate": 9.844046090824533e-05,
      "loss": 0.6889,
      "step": 80
    },
    {
      "epoch": 0.3521739130434783,
      "grad_norm": 0.07275456190109253,
      "learning_rate": 9.834504404631031e-05,
      "loss": 0.6879,
      "step": 81
    },
    {
      "epoch": 0.3565217391304348,
      "grad_norm": 0.06591422110795975,
      "learning_rate": 9.824684331444927e-05,
      "loss": 0.6554,
      "step": 82
    },
    {
      "epoch": 0.36086956521739133,
      "grad_norm": 0.06396066397428513,
      "learning_rate": 9.814586436738998e-05,
      "loss": 0.6925,
      "step": 83
    },
    {
      "epoch": 0.3652173913043478,
      "grad_norm": 0.08825157582759857,
      "learning_rate": 9.804211301983918e-05,
      "loss": 0.6629,
      "step": 84
    },
    {
      "epoch": 0.3695652173913043,
      "grad_norm": 0.06731634587049484,
      "learning_rate": 9.793559524614779e-05,
      "loss": 0.6745,
      "step": 85
    },
    {
      "epoch": 0.3739130434782609,
      "grad_norm": 0.06455274671316147,
      "learning_rate": 9.782631717996675e-05,
      "loss": 0.6851,
      "step": 86
    },
    {
      "epoch": 0.3782608695652174,
      "grad_norm": 0.07710668444633484,
      "learning_rate": 9.771428511389395e-05,
      "loss": 0.6929,
      "step": 87
    },
    {
      "epoch": 0.3826086956521739,
      "grad_norm": 0.0727052241563797,
      "learning_rate": 9.759950549911186e-05,
      "loss": 0.6798,
      "step": 88
    },
    {
      "epoch": 0.3869565217391304,
      "grad_norm": 0.07156208157539368,
      "learning_rate": 9.748198494501597e-05,
      "loss": 0.6807,
      "step": 89
    },
    {
      "epoch": 0.391304347826087,
      "grad_norm": 0.0921456515789032,
      "learning_rate": 9.736173021883432e-05,
      "loss": 0.6435,
      "step": 90
    },
    {
      "epoch": 0.39565217391304347,
      "grad_norm": 0.09094609320163727,
      "learning_rate": 9.723874824523771e-05,
      "loss": 0.6874,
      "step": 91
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.09006571024656296,
      "learning_rate": 9.711304610594104e-05,
      "loss": 0.6778,
      "step": 92
    },
    {
      "epoch": 0.4043478260869565,
      "grad_norm": 0.13732297718524933,
      "learning_rate": 9.698463103929542e-05,
      "loss": 0.6561,
      "step": 93
    },
    {
      "epoch": 0.40869565217391307,
      "grad_norm": 0.09598764777183533,
      "learning_rate": 9.685351043987151e-05,
      "loss": 0.6624,
      "step": 94
    },
    {
      "epoch": 0.41304347826086957,
      "grad_norm": 0.09070798009634018,
      "learning_rate": 9.671969185803356e-05,
      "loss": 0.6684,
      "step": 95
    },
    {
      "epoch": 0.41739130434782606,
      "grad_norm": 0.0911954715847969,
      "learning_rate": 9.658318299950473e-05,
      "loss": 0.6568,
      "step": 96
    },
    {
      "epoch": 0.4217391304347826,
      "grad_norm": 0.08703230321407318,
      "learning_rate": 9.644399172492336e-05,
      "loss": 0.6442,
      "step": 97
    },
    {
      "epoch": 0.4260869565217391,
      "grad_norm": 0.0760849341750145,
      "learning_rate": 9.630212604939026e-05,
      "loss": 0.6551,
      "step": 98
    },
    {
      "epoch": 0.43043478260869567,
      "grad_norm": 0.10621879249811172,
      "learning_rate": 9.615759414200729e-05,
      "loss": 0.6665,
      "step": 99
    },
    {
      "epoch": 0.43478260869565216,
      "grad_norm": 0.08248650282621384,
      "learning_rate": 9.601040432540684e-05,
      "loss": 0.6752,
      "step": 100
    },
    {
      "epoch": 0.4391304347826087,
      "grad_norm": 0.10147503018379211,
      "learning_rate": 9.586056507527266e-05,
      "loss": 0.6602,
      "step": 101
    },
    {
      "epoch": 0.4434782608695652,
      "grad_norm": 0.1442282497882843,
      "learning_rate": 9.570808501985175e-05,
      "loss": 0.6704,
      "step": 102
    },
    {
      "epoch": 0.44782608695652176,
      "grad_norm": 0.11339450627565384,
      "learning_rate": 9.555297293945759e-05,
      "loss": 0.6631,
      "step": 103
    },
    {
      "epoch": 0.45217391304347826,
      "grad_norm": 0.15643437206745148,
      "learning_rate": 9.539523776596445e-05,
      "loss": 0.668,
      "step": 104
    },
    {
      "epoch": 0.45652173913043476,
      "grad_norm": 0.1856074035167694,
      "learning_rate": 9.523488858229313e-05,
      "loss": 0.6413,
      "step": 105
    },
    {
      "epoch": 0.4608695652173913,
      "grad_norm": 0.12280824780464172,
      "learning_rate": 9.507193462188791e-05,
      "loss": 0.6658,
      "step": 106
    },
    {
      "epoch": 0.4652173913043478,
      "grad_norm": 0.18749414384365082,
      "learning_rate": 9.49063852681848e-05,
      "loss": 0.6785,
      "step": 107
    },
    {
      "epoch": 0.46956521739130436,
      "grad_norm": 0.13954943418502808,
      "learning_rate": 9.47382500540714e-05,
      "loss": 0.652,
      "step": 108
    },
    {
      "epoch": 0.47391304347826085,
      "grad_norm": 0.15025292336940765,
      "learning_rate": 9.45675386613377e-05,
      "loss": 0.6622,
      "step": 109
    },
    {
      "epoch": 0.4782608695652174,
      "grad_norm": 0.11263363063335419,
      "learning_rate": 9.439426092011875e-05,
      "loss": 0.6573,
      "step": 110
    },
    {
      "epoch": 0.4826086956521739,
      "grad_norm": 0.12779393792152405,
      "learning_rate": 9.421842680832861e-05,
      "loss": 0.6535,
      "step": 111
    },
    {
      "epoch": 0.48695652173913045,
      "grad_norm": 0.11488567292690277,
      "learning_rate": 9.404004645108568e-05,
      "loss": 0.6438,
      "step": 112
    },
    {
      "epoch": 0.49130434782608695,
      "grad_norm": 0.1706668585538864,
      "learning_rate": 9.385913012012973e-05,
      "loss": 0.6427,
      "step": 113
    },
    {
      "epoch": 0.4956521739130435,
      "grad_norm": 0.13733729720115662,
      "learning_rate": 9.367568823323039e-05,
      "loss": 0.6555,
      "step": 114
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.11061578243970871,
      "learning_rate": 9.348973135358734e-05,
      "loss": 0.6672,
      "step": 115
    },
    {
      "epoch": 0.5043478260869565,
      "grad_norm": 0.18926067650318146,
      "learning_rate": 9.330127018922194e-05,
      "loss": 0.6573,
      "step": 116
    },
    {
      "epoch": 0.508695652173913,
      "grad_norm": 0.15428727865219116,
      "learning_rate": 9.311031559236067e-05,
      "loss": 0.6708,
      "step": 117
    },
    {
      "epoch": 0.5130434782608696,
      "grad_norm": 0.16264328360557556,
      "learning_rate": 9.291687855881026e-05,
      "loss": 0.6446,
      "step": 118
    },
    {
      "epoch": 0.5173913043478261,
      "grad_norm": 0.11342114955186844,
      "learning_rate": 9.272097022732443e-05,
      "loss": 0.6571,
      "step": 119
    },
    {
      "epoch": 0.5217391304347826,
      "grad_norm": 0.15034589171409607,
      "learning_rate": 9.252260187896256e-05,
      "loss": 0.6408,
      "step": 120
    },
    {
      "epoch": 0.5260869565217391,
      "grad_norm": 0.21747715771198273,
      "learning_rate": 9.232178493644006e-05,
      "loss": 0.6346,
      "step": 121
    },
    {
      "epoch": 0.5304347826086957,
      "grad_norm": 0.27781569957733154,
      "learning_rate": 9.211853096347058e-05,
      "loss": 0.6541,
      "step": 122
    },
    {
      "epoch": 0.5347826086956522,
      "grad_norm": 0.2587333023548126,
      "learning_rate": 9.191285166410022e-05,
      "loss": 0.6516,
      "step": 123
    },
    {
      "epoch": 0.5391304347826087,
      "grad_norm": 0.16397182643413544,
      "learning_rate": 9.170475888203347e-05,
      "loss": 0.6716,
      "step": 124
    },
    {
      "epoch": 0.5434782608695652,
      "grad_norm": 0.12862510979175568,
      "learning_rate": 9.149426459995126e-05,
      "loss": 0.6596,
      "step": 125
    },
    {
      "epoch": 0.5478260869565217,
      "grad_norm": 0.15427789092063904,
      "learning_rate": 9.128138093882098e-05,
      "loss": 0.6588,
      "step": 126
    },
    {
      "epoch": 0.5521739130434783,
      "grad_norm": 0.22064033150672913,
      "learning_rate": 9.106612015719845e-05,
      "loss": 0.6314,
      "step": 127
    },
    {
      "epoch": 0.5565217391304348,
      "grad_norm": 0.1941988468170166,
      "learning_rate": 9.08484946505221e-05,
      "loss": 0.648,
      "step": 128
    },
    {
      "epoch": 0.5608695652173913,
      "grad_norm": 0.18163767457008362,
      "learning_rate": 9.062851695039915e-05,
      "loss": 0.6738,
      "step": 129
    },
    {
      "epoch": 0.5652173913043478,
      "grad_norm": 0.16294820606708527,
      "learning_rate": 9.040619972388403e-05,
      "loss": 0.6534,
      "step": 130
    },
    {
      "epoch": 0.5695652173913044,
      "grad_norm": 0.23330819606781006,
      "learning_rate": 9.018155577274892e-05,
      "loss": 0.6478,
      "step": 131
    },
    {
      "epoch": 0.5739130434782609,
      "grad_norm": 0.3880465030670166,
      "learning_rate": 8.995459803274664e-05,
      "loss": 0.6566,
      "step": 132
    },
    {
      "epoch": 0.5782608695652174,
      "grad_norm": 0.6047540903091431,
      "learning_rate": 8.972533957286573e-05,
      "loss": 0.6321,
      "step": 133
    },
    {
      "epoch": 0.5826086956521739,
      "grad_norm": 0.526760995388031,
      "learning_rate": 8.949379359457793e-05,
      "loss": 0.6501,
      "step": 134
    },
    {
      "epoch": 0.5869565217391305,
      "grad_norm": 0.26121070981025696,
      "learning_rate": 8.925997343107795e-05,
      "loss": 0.6462,
      "step": 135
    },
    {
      "epoch": 0.591304347826087,
      "grad_norm": 0.3640858232975006,
      "learning_rate": 8.902389254651569e-05,
      "loss": 0.6378,
      "step": 136
    },
    {
      "epoch": 0.5956521739130435,
      "grad_norm": 0.3413775861263275,
      "learning_rate": 8.8785564535221e-05,
      "loss": 0.6549,
      "step": 137
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.25738435983657837,
      "learning_rate": 8.854500312092081e-05,
      "loss": 0.6266,
      "step": 138
    },
    {
      "epoch": 0.6043478260869565,
      "grad_norm": 0.3974941670894623,
      "learning_rate": 8.83022221559489e-05,
      "loss": 0.6207,
      "step": 139
    },
    {
      "epoch": 0.6086956521739131,
      "grad_norm": 0.3541712462902069,
      "learning_rate": 8.805723562044824e-05,
      "loss": 0.6623,
      "step": 140
    },
    {
      "epoch": 0.6130434782608696,
      "grad_norm": 0.29466933012008667,
      "learning_rate": 8.781005762156593e-05,
      "loss": 0.6753,
      "step": 141
    },
    {
      "epoch": 0.6173913043478261,
      "grad_norm": 0.429376482963562,
      "learning_rate": 8.75607023926409e-05,
      "loss": 0.6351,
      "step": 142
    },
    {
      "epoch": 0.6217391304347826,
      "grad_norm": 0.28085529804229736,
      "learning_rate": 8.730918429238428e-05,
      "loss": 0.6584,
      "step": 143
    },
    {
      "epoch": 0.6260869565217392,
      "grad_norm": 0.34451988339424133,
      "learning_rate": 8.705551780405263e-05,
      "loss": 0.6619,
      "step": 144
    },
    {
      "epoch": 0.6304347826086957,
      "grad_norm": 0.3307543098926544,
      "learning_rate": 8.679971753461387e-05,
      "loss": 0.6448,
      "step": 145
    },
    {
      "epoch": 0.6347826086956522,
      "grad_norm": 0.2655896842479706,
      "learning_rate": 8.654179821390621e-05,
      "loss": 0.6442,
      "step": 146
    },
    {
      "epoch": 0.6391304347826087,
      "grad_norm": 0.4360576868057251,
      "learning_rate": 8.628177469378995e-05,
      "loss": 0.6487,
      "step": 147
    },
    {
      "epoch": 0.6434782608695652,
      "grad_norm": 0.35094520449638367,
      "learning_rate": 8.601966194729227e-05,
      "loss": 0.6359,
      "step": 148
    },
    {
      "epoch": 0.6478260869565218,
      "grad_norm": 0.4109646677970886,
      "learning_rate": 8.575547506774497e-05,
      "loss": 0.6519,
      "step": 149
    },
    {
      "epoch": 0.6521739130434783,
      "grad_norm": 0.3401927053928375,
      "learning_rate": 8.548922926791545e-05,
      "loss": 0.6375,
      "step": 150
    },
    {
      "epoch": 0.6565217391304348,
      "grad_norm": 0.22073158621788025,
      "learning_rate": 8.522093987913062e-05,
      "loss": 0.6462,
      "step": 151
    },
    {
      "epoch": 0.6608695652173913,
      "grad_norm": 0.43310844898223877,
      "learning_rate": 8.495062235039411e-05,
      "loss": 0.6697,
      "step": 152
    },
    {
      "epoch": 0.6652173913043479,
      "grad_norm": 0.42843684554100037,
      "learning_rate": 8.467829224749665e-05,
      "loss": 0.6169,
      "step": 153
    },
    {
      "epoch": 0.6695652173913044,
      "grad_norm": 0.4057531952857971,
      "learning_rate": 8.440396525211975e-05,
      "loss": 0.6625,
      "step": 154
    },
    {
      "epoch": 0.6739130434782609,
      "grad_norm": 0.23454974591732025,
      "learning_rate": 8.412765716093272e-05,
      "loss": 0.616,
      "step": 155
    },
    {
      "epoch": 0.6782608695652174,
      "grad_norm": 0.3723919987678528,
      "learning_rate": 8.384938388468296e-05,
      "loss": 0.6576,
      "step": 156
    },
    {
      "epoch": 0.6826086956521739,
      "grad_norm": 0.44731444120407104,
      "learning_rate": 8.356916144727985e-05,
      "loss": 0.6408,
      "step": 157
    },
    {
      "epoch": 0.6869565217391305,
      "grad_norm": 0.4032682180404663,
      "learning_rate": 8.328700598487203e-05,
      "loss": 0.6541,
      "step": 158
    },
    {
      "epoch": 0.691304347826087,
      "grad_norm": 0.34927839040756226,
      "learning_rate": 8.300293374491821e-05,
      "loss": 0.641,
      "step": 159
    },
    {
      "epoch": 0.6956521739130435,
      "grad_norm": 0.37337374687194824,
      "learning_rate": 8.271696108525157e-05,
      "loss": 0.6409,
      "step": 160
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.33849138021469116,
      "learning_rate": 8.24291044731378e-05,
      "loss": 0.6571,
      "step": 161
    },
    {
      "epoch": 0.7043478260869566,
      "grad_norm": 0.26402008533477783,
      "learning_rate": 8.213938048432697e-05,
      "loss": 0.6467,
      "step": 162
    },
    {
      "epoch": 0.7086956521739131,
      "grad_norm": 0.31730157136917114,
      "learning_rate": 8.184780580209892e-05,
      "loss": 0.6519,
      "step": 163
    },
    {
      "epoch": 0.7130434782608696,
      "grad_norm": 0.47295334935188293,
      "learning_rate": 8.155439721630264e-05,
      "loss": 0.629,
      "step": 164
    },
    {
      "epoch": 0.717391304347826,
      "grad_norm": 0.3847337067127228,
      "learning_rate": 8.125917162238945e-05,
      "loss": 0.6404,
      "step": 165
    },
    {
      "epoch": 0.7217391304347827,
      "grad_norm": 0.30035194754600525,
      "learning_rate": 8.09621460204401e-05,
      "loss": 0.6697,
      "step": 166
    },
    {
      "epoch": 0.7260869565217392,
      "grad_norm": 0.4391736686229706,
      "learning_rate": 8.066333751418583e-05,
      "loss": 0.6399,
      "step": 167
    },
    {
      "epoch": 0.7304347826086957,
      "grad_norm": 0.6057283878326416,
      "learning_rate": 8.036276331002348e-05,
      "loss": 0.6341,
      "step": 168
    },
    {
      "epoch": 0.7347826086956522,
      "grad_norm": 0.644005537033081,
      "learning_rate": 8.006044071602477e-05,
      "loss": 0.662,
      "step": 169
    },
    {
      "epoch": 0.7391304347826086,
      "grad_norm": 0.6034097671508789,
      "learning_rate": 7.975638714093949e-05,
      "loss": 0.638,
      "step": 170
    },
    {
      "epoch": 0.7434782608695653,
      "grad_norm": 0.49918678402900696,
      "learning_rate": 7.945062009319319e-05,
      "loss": 0.6322,
      "step": 171
    },
    {
      "epoch": 0.7478260869565218,
      "grad_norm": 0.4087945520877838,
      "learning_rate": 7.914315717987892e-05,
      "loss": 0.6419,
      "step": 172
    },
    {
      "epoch": 0.7521739130434782,
      "grad_norm": 0.3623512089252472,
      "learning_rate": 7.883401610574336e-05,
      "loss": 0.6618,
      "step": 173
    },
    {
      "epoch": 0.7565217391304347,
      "grad_norm": 0.7307239174842834,
      "learning_rate": 7.85232146721673e-05,
      "loss": 0.6572,
      "step": 174
    },
    {
      "epoch": 0.7608695652173914,
      "grad_norm": 0.8763480186462402,
      "learning_rate": 7.821077077614061e-05,
      "loss": 0.6434,
      "step": 175
    },
    {
      "epoch": 0.7652173913043478,
      "grad_norm": 0.5741376280784607,
      "learning_rate": 7.789670240923168e-05,
      "loss": 0.6539,
      "step": 176
    },
    {
      "epoch": 0.7695652173913043,
      "grad_norm": 0.4742548167705536,
      "learning_rate": 7.758102765655137e-05,
      "loss": 0.6435,
      "step": 177
    },
    {
      "epoch": 0.7739130434782608,
      "grad_norm": 0.6679338216781616,
      "learning_rate": 7.726376469571164e-05,
      "loss": 0.6654,
      "step": 178
    },
    {
      "epoch": 0.7782608695652173,
      "grad_norm": 0.7236630320549011,
      "learning_rate": 7.694493179577879e-05,
      "loss": 0.655,
      "step": 179
    },
    {
      "epoch": 0.782608695652174,
      "grad_norm": 0.3954794704914093,
      "learning_rate": 7.662454731622148e-05,
      "loss": 0.6733,
      "step": 180
    },
    {
      "epoch": 0.7869565217391304,
      "grad_norm": 0.41423317790031433,
      "learning_rate": 7.630262970585356e-05,
      "loss": 0.6466,
      "step": 181
    },
    {
      "epoch": 0.7913043478260869,
      "grad_norm": 0.5248022675514221,
      "learning_rate": 7.597919750177168e-05,
      "loss": 0.6343,
      "step": 182
    },
    {
      "epoch": 0.7956521739130434,
      "grad_norm": 0.4523037075996399,
      "learning_rate": 7.56542693282879e-05,
      "loss": 0.6494,
      "step": 183
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.3984985947608948,
      "learning_rate": 7.532786389585716e-05,
      "loss": 0.6511,
      "step": 184
    },
    {
      "epoch": 0.8043478260869565,
      "grad_norm": 0.3854583501815796,
      "learning_rate": 7.500000000000001e-05,
      "loss": 0.6628,
      "step": 185
    },
    {
      "epoch": 0.808695652173913,
      "grad_norm": 0.43506574630737305,
      "learning_rate": 7.467069652022016e-05,
      "loss": 0.6603,
      "step": 186
    },
    {
      "epoch": 0.8130434782608695,
      "grad_norm": 0.41759249567985535,
      "learning_rate": 7.433997241891742e-05,
      "loss": 0.6674,
      "step": 187
    },
    {
      "epoch": 0.8173913043478261,
      "grad_norm": 0.2531141936779022,
      "learning_rate": 7.400784674029578e-05,
      "loss": 0.6395,
      "step": 188
    },
    {
      "epoch": 0.8217391304347826,
      "grad_norm": 0.4464227259159088,
      "learning_rate": 7.36743386092667e-05,
      "loss": 0.6576,
      "step": 189
    },
    {
      "epoch": 0.8260869565217391,
      "grad_norm": 0.47379711270332336,
      "learning_rate": 7.333946723034794e-05,
      "loss": 0.6423,
      "step": 190
    },
    {
      "epoch": 0.8304347826086956,
      "grad_norm": 0.3901284635066986,
      "learning_rate": 7.300325188655761e-05,
      "loss": 0.6594,
      "step": 191
    },
    {
      "epoch": 0.8347826086956521,
      "grad_norm": 0.3132023811340332,
      "learning_rate": 7.266571193830387e-05,
      "loss": 0.6611,
      "step": 192
    },
    {
      "epoch": 0.8391304347826087,
      "grad_norm": 0.4356115460395813,
      "learning_rate": 7.232686682227001e-05,
      "loss": 0.6376,
      "step": 193
    },
    {
      "epoch": 0.8434782608695652,
      "grad_norm": 0.5321224331855774,
      "learning_rate": 7.198673605029528e-05,
      "loss": 0.6643,
      "step": 194
    },
    {
      "epoch": 0.8478260869565217,
      "grad_norm": 0.3640391230583191,
      "learning_rate": 7.164533920825137e-05,
      "loss": 0.6476,
      "step": 195
    },
    {
      "epoch": 0.8521739130434782,
      "grad_norm": 0.2873951494693756,
      "learning_rate": 7.130269595491443e-05,
      "loss": 0.649,
      "step": 196
    },
    {
      "epoch": 0.8565217391304348,
      "grad_norm": 0.45937976241111755,
      "learning_rate": 7.095882602083322e-05,
      "loss": 0.648,
      "step": 197
    },
    {
      "epoch": 0.8608695652173913,
      "grad_norm": 0.5308820009231567,
      "learning_rate": 7.061374920719288e-05,
      "loss": 0.6458,
      "step": 198
    },
    {
      "epoch": 0.8652173913043478,
      "grad_norm": 0.43750235438346863,
      "learning_rate": 7.026748538467474e-05,
      "loss": 0.6457,
      "step": 199
    },
    {
      "epoch": 0.8695652173913043,
      "grad_norm": 0.27052804827690125,
      "learning_rate": 6.992005449231208e-05,
      "loss": 0.6452,
      "step": 200
    },
    {
      "epoch": 0.8739130434782608,
      "grad_norm": 0.37570297718048096,
      "learning_rate": 6.957147653634198e-05,
      "loss": 0.6566,
      "step": 201
    },
    {
      "epoch": 0.8782608695652174,
      "grad_norm": 0.32025307416915894,
      "learning_rate": 6.922177158905325e-05,
      "loss": 0.6655,
      "step": 202
    },
    {
      "epoch": 0.8826086956521739,
      "grad_norm": 0.2932673990726471,
      "learning_rate": 6.887095978763072e-05,
      "loss": 0.6749,
      "step": 203
    },
    {
      "epoch": 0.8869565217391304,
      "grad_norm": 0.23213867843151093,
      "learning_rate": 6.851906133299557e-05,
      "loss": 0.6631,
      "step": 204
    },
    {
      "epoch": 0.8913043478260869,
      "grad_norm": 0.3275505602359772,
      "learning_rate": 6.816609648864208e-05,
      "loss": 0.6758,
      "step": 205
    },
    {
      "epoch": 0.8956521739130435,
      "grad_norm": 0.32032299041748047,
      "learning_rate": 6.781208557947086e-05,
      "loss": 0.662,
      "step": 206
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.26808756589889526,
      "learning_rate": 6.745704899061843e-05,
      "loss": 0.6464,
      "step": 207
    },
    {
      "epoch": 0.9043478260869565,
      "grad_norm": 0.25998106598854065,
      "learning_rate": 6.710100716628344e-05,
      "loss": 0.6556,
      "step": 208
    },
    {
      "epoch": 0.908695652173913,
      "grad_norm": 0.36953797936439514,
      "learning_rate": 6.674398060854931e-05,
      "loss": 0.6761,
      "step": 209
    },
    {
      "epoch": 0.9130434782608695,
      "grad_norm": 0.43774327635765076,
      "learning_rate": 6.638598987620375e-05,
      "loss": 0.6481,
      "step": 210
    },
    {
      "epoch": 0.9173913043478261,
      "grad_norm": 0.24901102483272552,
      "learning_rate": 6.602705558355486e-05,
      "loss": 0.675,
      "step": 211
    },
    {
      "epoch": 0.9217391304347826,
      "grad_norm": 0.3668375313282013,
      "learning_rate": 6.566719839924412e-05,
      "loss": 0.6619,
      "step": 212
    },
    {
      "epoch": 0.9260869565217391,
      "grad_norm": 0.5943741202354431,
      "learning_rate": 6.530643904505621e-05,
      "loss": 0.6561,
      "step": 213
    },
    {
      "epoch": 0.9304347826086956,
      "grad_norm": 0.6538096070289612,
      "learning_rate": 6.49447982947258e-05,
      "loss": 0.6297,
      "step": 214
    },
    {
      "epoch": 0.9347826086956522,
      "grad_norm": 0.5622021555900574,
      "learning_rate": 6.458229697274125e-05,
      "loss": 0.6602,
      "step": 215
    },
    {
      "epoch": 0.9391304347826087,
      "grad_norm": 0.45731329917907715,
      "learning_rate": 6.42189559531456e-05,
      "loss": 0.6686,
      "step": 216
    },
    {
      "epoch": 0.9434782608695652,
      "grad_norm": 0.26856303215026855,
      "learning_rate": 6.385479615833445e-05,
      "loss": 0.6358,
      "step": 217
    },
    {
      "epoch": 0.9478260869565217,
      "grad_norm": 0.24898113310337067,
      "learning_rate": 6.348983855785121e-05,
      "loss": 0.6579,
      "step": 218
    },
    {
      "epoch": 0.9521739130434783,
      "grad_norm": 0.3039465844631195,
      "learning_rate": 6.312410416717968e-05,
      "loss": 0.6493,
      "step": 219
    },
    {
      "epoch": 0.9565217391304348,
      "grad_norm": 0.36120837926864624,
      "learning_rate": 6.27576140465338e-05,
      "loss": 0.6524,
      "step": 220
    },
    {
      "epoch": 0.9608695652173913,
      "grad_norm": 0.35759392380714417,
      "learning_rate": 6.2390389299645e-05,
      "loss": 0.6247,
      "step": 221
    },
    {
      "epoch": 0.9652173913043478,
      "grad_norm": 0.29583072662353516,
      "learning_rate": 6.202245107254693e-05,
      "loss": 0.642,
      "step": 222
    },
    {
      "epoch": 0.9695652173913043,
      "grad_norm": 0.24323242902755737,
      "learning_rate": 6.165382055235783e-05,
      "loss": 0.6683,
      "step": 223
    },
    {
      "epoch": 0.9739130434782609,
      "grad_norm": 0.4522090256214142,
      "learning_rate": 6.128451896606053e-05,
      "loss": 0.639,
      "step": 224
    },
    {
      "epoch": 0.9782608695652174,
      "grad_norm": 0.6692441701889038,
      "learning_rate": 6.091456757928008e-05,
      "loss": 0.6628,
      "step": 225
    },
    {
      "epoch": 0.9826086956521739,
      "grad_norm": 0.7985122203826904,
      "learning_rate": 6.054398769505924e-05,
      "loss": 0.6585,
      "step": 226
    },
    {
      "epoch": 0.9869565217391304,
      "grad_norm": 0.6464029550552368,
      "learning_rate": 6.01728006526317e-05,
      "loss": 0.6563,
      "step": 227
    },
    {
      "epoch": 0.991304347826087,
      "grad_norm": 0.36494386196136475,
      "learning_rate": 5.980102782619342e-05,
      "loss": 0.648,
      "step": 228
    },
    {
      "epoch": 0.9956521739130435,
      "grad_norm": 0.40735068917274475,
      "learning_rate": 5.942869062367179e-05,
      "loss": 0.6502,
      "step": 229
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.6993163228034973,
      "learning_rate": 5.905581048549279e-05,
      "loss": 0.6682,
      "step": 230
    }
  ],
  "logging_steps": 1,
  "max_steps": 460,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 8.970864260913562e+18,
  "train_batch_size": 24,
  "trial_name": null,
  "trial_params": null
}