{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 6.850632911392405,
  "eval_steps": 500,
  "global_step": 224,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.030379746835443037,
      "grad_norm": 6.724321385441998,
      "learning_rate": 8.695652173913044e-07,
      "loss": 1.0659,
      "step": 1
    },
    {
      "epoch": 0.060759493670886074,
      "grad_norm": 6.673968240891003,
      "learning_rate": 1.7391304347826088e-06,
      "loss": 1.0828,
      "step": 2
    },
    {
      "epoch": 0.09113924050632911,
      "grad_norm": 6.523748536253664,
      "learning_rate": 2.6086956521739132e-06,
      "loss": 1.0319,
      "step": 3
    },
    {
      "epoch": 0.12151898734177215,
      "grad_norm": 6.122873408278022,
      "learning_rate": 3.4782608695652175e-06,
      "loss": 1.0405,
      "step": 4
    },
    {
      "epoch": 0.1518987341772152,
      "grad_norm": 4.801542878629058,
      "learning_rate": 4.347826086956522e-06,
      "loss": 0.9891,
      "step": 5
    },
    {
      "epoch": 0.18227848101265823,
      "grad_norm": 2.689749518380698,
      "learning_rate": 5.2173913043478265e-06,
      "loss": 0.9557,
      "step": 6
    },
    {
      "epoch": 0.21265822784810126,
      "grad_norm": 2.169959433266693,
      "learning_rate": 6.086956521739132e-06,
      "loss": 0.9031,
      "step": 7
    },
    {
      "epoch": 0.2430379746835443,
      "grad_norm": 4.6833611606171885,
      "learning_rate": 6.956521739130435e-06,
      "loss": 0.9259,
      "step": 8
    },
    {
      "epoch": 0.27341772151898736,
      "grad_norm": 4.763333125744323,
      "learning_rate": 7.82608695652174e-06,
      "loss": 0.9703,
      "step": 9
    },
    {
      "epoch": 0.3037974683544304,
      "grad_norm": 4.575166820637519,
      "learning_rate": 8.695652173913044e-06,
      "loss": 0.8889,
      "step": 10
    },
    {
      "epoch": 0.3341772151898734,
      "grad_norm": 4.401938673864446,
      "learning_rate": 9.565217391304349e-06,
      "loss": 0.8995,
      "step": 11
    },
    {
      "epoch": 0.36455696202531646,
      "grad_norm": 4.216432602520236,
      "learning_rate": 1.0434782608695653e-05,
      "loss": 0.8902,
      "step": 12
    },
    {
      "epoch": 0.3949367088607595,
      "grad_norm": 3.3775295795191274,
      "learning_rate": 1.1304347826086957e-05,
      "loss": 0.8755,
      "step": 13
    },
    {
      "epoch": 0.4253164556962025,
      "grad_norm": 2.7502809798077936,
      "learning_rate": 1.2173913043478263e-05,
      "loss": 0.8139,
      "step": 14
    },
    {
      "epoch": 0.45569620253164556,
      "grad_norm": 2.8419087912747134,
      "learning_rate": 1.3043478260869566e-05,
      "loss": 0.798,
      "step": 15
    },
    {
      "epoch": 0.4860759493670886,
      "grad_norm": 2.6688951578471585,
      "learning_rate": 1.391304347826087e-05,
      "loss": 0.825,
      "step": 16
    },
    {
      "epoch": 0.5164556962025316,
      "grad_norm": 2.4029391793919914,
      "learning_rate": 1.4782608695652174e-05,
      "loss": 0.8193,
      "step": 17
    },
    {
      "epoch": 0.5468354430379747,
      "grad_norm": 1.857269035972293,
      "learning_rate": 1.565217391304348e-05,
      "loss": 0.768,
      "step": 18
    },
    {
      "epoch": 0.5772151898734177,
      "grad_norm": 1.6490997914690606,
      "learning_rate": 1.6521739130434785e-05,
      "loss": 0.7852,
      "step": 19
    },
    {
      "epoch": 0.6075949367088608,
      "grad_norm": 1.623783155586043,
      "learning_rate": 1.739130434782609e-05,
      "loss": 0.7524,
      "step": 20
    },
    {
      "epoch": 0.6379746835443038,
      "grad_norm": 1.5106238324794226,
      "learning_rate": 1.8260869565217393e-05,
      "loss": 0.7705,
      "step": 21
    },
    {
      "epoch": 0.6683544303797468,
      "grad_norm": 1.1342268905421653,
      "learning_rate": 1.9130434782608697e-05,
      "loss": 0.7737,
      "step": 22
    },
    {
      "epoch": 0.6987341772151898,
      "grad_norm": 1.1108579601060105,
      "learning_rate": 2e-05,
      "loss": 0.7484,
      "step": 23
    },
    {
      "epoch": 0.7291139240506329,
      "grad_norm": 1.1296419089546688,
      "learning_rate": 1.999877856940653e-05,
      "loss": 0.7728,
      "step": 24
    },
    {
      "epoch": 0.759493670886076,
      "grad_norm": 0.9722164607834988,
      "learning_rate": 1.999511457600466e-05,
      "loss": 0.7574,
      "step": 25
    },
    {
      "epoch": 0.789873417721519,
      "grad_norm": 0.9275590226007782,
      "learning_rate": 1.9989008914857115e-05,
      "loss": 0.7435,
      "step": 26
    },
    {
      "epoch": 0.8202531645569621,
      "grad_norm": 0.9488572495997042,
      "learning_rate": 1.998046307749216e-05,
      "loss": 0.721,
      "step": 27
    },
    {
      "epoch": 0.850632911392405,
      "grad_norm": 0.9249397199574121,
      "learning_rate": 1.9969479151539238e-05,
      "loss": 0.7169,
      "step": 28
    },
    {
      "epoch": 0.8810126582278481,
      "grad_norm": 0.7523228212322401,
      "learning_rate": 1.9956059820218982e-05,
      "loss": 0.695,
      "step": 29
    },
    {
      "epoch": 0.9113924050632911,
      "grad_norm": 0.9105279081256428,
      "learning_rate": 1.9940208361687762e-05,
      "loss": 0.7423,
      "step": 30
    },
    {
      "epoch": 0.9417721518987342,
      "grad_norm": 0.8016467149598822,
      "learning_rate": 1.9921928648236855e-05,
      "loss": 0.7174,
      "step": 31
    },
    {
      "epoch": 0.9721518987341772,
      "grad_norm": 0.7440010212967557,
      "learning_rate": 1.990122514534651e-05,
      "loss": 0.7159,
      "step": 32
    },
    {
      "epoch": 1.010126582278481,
      "grad_norm": 0.7731147531183545,
      "learning_rate": 1.9878102910595097e-05,
      "loss": 0.6848,
      "step": 33
    },
    {
      "epoch": 1.040506329113924,
      "grad_norm": 0.7564537787045006,
      "learning_rate": 1.985256759242359e-05,
      "loss": 0.6402,
      "step": 34
    },
    {
      "epoch": 1.070886075949367,
      "grad_norm": 0.8281978759652757,
      "learning_rate": 1.982462542875576e-05,
      "loss": 0.6249,
      "step": 35
    },
    {
      "epoch": 1.1012658227848102,
      "grad_norm": 0.8330114235132352,
      "learning_rate": 1.979428324547432e-05,
      "loss": 0.5945,
      "step": 36
    },
    {
      "epoch": 1.1316455696202532,
      "grad_norm": 0.795134793296867,
      "learning_rate": 1.9761548454753455e-05,
      "loss": 0.5984,
      "step": 37
    },
    {
      "epoch": 1.1620253164556962,
      "grad_norm": 0.8928024561000094,
      "learning_rate": 1.972642905324813e-05,
      "loss": 0.5847,
      "step": 38
    },
    {
      "epoch": 1.1924050632911392,
      "grad_norm": 0.810403107459553,
      "learning_rate": 1.9688933620140638e-05,
      "loss": 0.624,
      "step": 39
    },
    {
      "epoch": 1.2227848101265824,
      "grad_norm": 0.9943149256376383,
      "learning_rate": 1.96490713150448e-05,
      "loss": 0.6083,
      "step": 40
    },
    {
      "epoch": 1.2531645569620253,
      "grad_norm": 0.9924643009165688,
      "learning_rate": 1.9606851875768404e-05,
      "loss": 0.6539,
      "step": 41
    },
    {
      "epoch": 1.2835443037974683,
      "grad_norm": 0.7993748151992456,
      "learning_rate": 1.956228561593441e-05,
      "loss": 0.5755,
      "step": 42
    },
    {
      "epoch": 1.3139240506329113,
      "grad_norm": 0.7566228025305142,
      "learning_rate": 1.9515383422461457e-05,
      "loss": 0.6056,
      "step": 43
    },
    {
      "epoch": 1.3443037974683545,
      "grad_norm": 0.8891824813519293,
      "learning_rate": 1.9466156752904344e-05,
      "loss": 0.6296,
      "step": 44
    },
    {
      "epoch": 1.3746835443037975,
      "grad_norm": 0.7209701927038757,
      "learning_rate": 1.9414617632655114e-05,
      "loss": 0.6093,
      "step": 45
    },
    {
      "epoch": 1.4050632911392404,
      "grad_norm": 0.6729020920743402,
      "learning_rate": 1.9360778652005416e-05,
      "loss": 0.5812,
      "step": 46
    },
    {
      "epoch": 1.4354430379746836,
      "grad_norm": 0.7255724729689165,
      "learning_rate": 1.9304652963070868e-05,
      "loss": 0.5793,
      "step": 47
    },
    {
      "epoch": 1.4658227848101266,
      "grad_norm": 0.7307464569892043,
      "learning_rate": 1.9246254276578175e-05,
      "loss": 0.6287,
      "step": 48
    },
    {
      "epoch": 1.4962025316455696,
      "grad_norm": 0.787421338323459,
      "learning_rate": 1.9185596858515797e-05,
      "loss": 0.5866,
      "step": 49
    },
    {
      "epoch": 1.5265822784810128,
      "grad_norm": 0.6637424482404751,
      "learning_rate": 1.9122695526648968e-05,
      "loss": 0.5971,
      "step": 50
    },
    {
      "epoch": 1.5569620253164556,
      "grad_norm": 0.7303549375587066,
      "learning_rate": 1.905756564689991e-05,
      "loss": 0.602,
      "step": 51
    },
    {
      "epoch": 1.5873417721518988,
      "grad_norm": 0.6671718898800668,
      "learning_rate": 1.8990223129594146e-05,
      "loss": 0.5425,
      "step": 52
    },
    {
      "epoch": 1.6177215189873417,
      "grad_norm": 0.7919585705802251,
      "learning_rate": 1.8920684425573865e-05,
      "loss": 0.5738,
      "step": 53
    },
    {
      "epoch": 1.6481012658227847,
      "grad_norm": 0.6990928157475217,
      "learning_rate": 1.884896652217917e-05,
      "loss": 0.5655,
      "step": 54
    },
    {
      "epoch": 1.678481012658228,
      "grad_norm": 0.7430920677636845,
      "learning_rate": 1.877508693909831e-05,
      "loss": 0.5667,
      "step": 55
    },
    {
      "epoch": 1.7088607594936709,
      "grad_norm": 0.7523164073492369,
      "learning_rate": 1.8699063724087905e-05,
      "loss": 0.5926,
      "step": 56
    },
    {
      "epoch": 1.7392405063291139,
      "grad_norm": 0.7721218219279508,
      "learning_rate": 1.862091544856407e-05,
      "loss": 0.5975,
      "step": 57
    },
    {
      "epoch": 1.769620253164557,
      "grad_norm": 0.7991673790574397,
      "learning_rate": 1.854066120306571e-05,
      "loss": 0.593,
      "step": 58
    },
    {
      "epoch": 1.8,
      "grad_norm": 0.8509470388117943,
      "learning_rate": 1.8458320592590976e-05,
      "loss": 0.613,
      "step": 59
    },
    {
      "epoch": 1.830379746835443,
      "grad_norm": 0.7926546932170354,
      "learning_rate": 1.837391373180801e-05,
      "loss": 0.6109,
      "step": 60
    },
    {
      "epoch": 1.8607594936708862,
      "grad_norm": 0.7373855946290643,
      "learning_rate": 1.8287461240141217e-05,
      "loss": 0.5796,
      "step": 61
    },
    {
      "epoch": 1.891139240506329,
      "grad_norm": 0.804743170937278,
      "learning_rate": 1.8198984236734246e-05,
      "loss": 0.5935,
      "step": 62
    },
    {
      "epoch": 1.9215189873417722,
      "grad_norm": 0.6464493381506916,
      "learning_rate": 1.8108504335290852e-05,
      "loss": 0.5752,
      "step": 63
    },
    {
      "epoch": 1.9518987341772152,
      "grad_norm": 0.897512467411834,
      "learning_rate": 1.8016043638794975e-05,
      "loss": 0.5739,
      "step": 64
    },
    {
      "epoch": 1.9822784810126581,
      "grad_norm": 0.6233712363768099,
      "learning_rate": 1.7921624734111292e-05,
      "loss": 0.552,
      "step": 65
    },
    {
      "epoch": 2.020253164556962,
      "grad_norm": 0.9010607465468213,
      "learning_rate": 1.7825270686467567e-05,
      "loss": 0.5134,
      "step": 66
    },
    {
      "epoch": 2.050632911392405,
      "grad_norm": 0.6231137424036363,
      "learning_rate": 1.7727005033820117e-05,
      "loss": 0.4369,
      "step": 67
    },
    {
      "epoch": 2.081012658227848,
      "grad_norm": 1.131343628324013,
      "learning_rate": 1.762685178110382e-05,
      "loss": 0.4255,
      "step": 68
    },
    {
      "epoch": 2.1113924050632913,
      "grad_norm": 0.6042525745110319,
      "learning_rate": 1.752483539436807e-05,
      "loss": 0.4445,
      "step": 69
    },
    {
      "epoch": 2.141772151898734,
      "grad_norm": 0.7743430859984088,
      "learning_rate": 1.7420980794800013e-05,
      "loss": 0.4385,
      "step": 70
    },
    {
      "epoch": 2.1721518987341772,
      "grad_norm": 0.6983330511726792,
      "learning_rate": 1.731531335263669e-05,
      "loss": 0.4315,
      "step": 71
    },
    {
      "epoch": 2.2025316455696204,
      "grad_norm": 0.6380893104358467,
      "learning_rate": 1.720785888096743e-05,
      "loss": 0.4393,
      "step": 72
    },
    {
      "epoch": 2.232911392405063,
      "grad_norm": 0.6572998772081126,
      "learning_rate": 1.7098643629428035e-05,
      "loss": 0.4231,
      "step": 73
    },
    {
      "epoch": 2.2632911392405064,
      "grad_norm": 0.6177397867260271,
      "learning_rate": 1.698769427778842e-05,
      "loss": 0.4149,
      "step": 74
    },
    {
      "epoch": 2.293670886075949,
      "grad_norm": 0.6282431319340771,
      "learning_rate": 1.687503792943506e-05,
      "loss": 0.457,
      "step": 75
    },
    {
      "epoch": 2.3240506329113924,
      "grad_norm": 0.615512704006609,
      "learning_rate": 1.6760702104750046e-05,
      "loss": 0.4118,
      "step": 76
    },
    {
      "epoch": 2.3544303797468356,
      "grad_norm": 0.64834503427471,
      "learning_rate": 1.664471473438822e-05,
      "loss": 0.4189,
      "step": 77
    },
    {
      "epoch": 2.3848101265822783,
      "grad_norm": 0.604153312916305,
      "learning_rate": 1.6527104152454096e-05,
      "loss": 0.4177,
      "step": 78
    },
    {
      "epoch": 2.4151898734177215,
      "grad_norm": 0.7037011923248295,
      "learning_rate": 1.6407899089580263e-05,
      "loss": 0.4343,
      "step": 79
    },
    {
      "epoch": 2.4455696202531647,
      "grad_norm": 0.6270592698620667,
      "learning_rate": 1.628712866590885e-05,
      "loss": 0.4135,
      "step": 80
    },
    {
      "epoch": 2.4759493670886075,
      "grad_norm": 0.6188407172388372,
      "learning_rate": 1.6164822383977912e-05,
      "loss": 0.399,
      "step": 81
    },
    {
      "epoch": 2.5063291139240507,
      "grad_norm": 0.6146877597480588,
      "learning_rate": 1.604101012151436e-05,
      "loss": 0.4374,
      "step": 82
    },
    {
      "epoch": 2.536708860759494,
      "grad_norm": 0.6352949615373182,
      "learning_rate": 1.5915722124135227e-05,
      "loss": 0.4334,
      "step": 83
    },
    {
      "epoch": 2.5670886075949366,
      "grad_norm": 0.6403150437248757,
      "learning_rate": 1.5788988997959115e-05,
      "loss": 0.4618,
      "step": 84
    },
    {
      "epoch": 2.59746835443038,
      "grad_norm": 0.6402992610460986,
      "learning_rate": 1.5660841702129533e-05,
      "loss": 0.4223,
      "step": 85
    },
    {
      "epoch": 2.6278481012658226,
      "grad_norm": 0.6024916966578524,
      "learning_rate": 1.5531311541251995e-05,
      "loss": 0.4165,
      "step": 86
    },
    {
      "epoch": 2.6582278481012658,
      "grad_norm": 0.6459925031270565,
      "learning_rate": 1.540043015774676e-05,
      "loss": 0.3988,
      "step": 87
    },
    {
      "epoch": 2.688607594936709,
      "grad_norm": 0.6594685990251288,
      "learning_rate": 1.5268229524119007e-05,
      "loss": 0.4153,
      "step": 88
    },
    {
      "epoch": 2.7189873417721517,
      "grad_norm": 0.6265585510563371,
      "learning_rate": 1.513474193514842e-05,
      "loss": 0.436,
      "step": 89
    },
    {
      "epoch": 2.749367088607595,
      "grad_norm": 0.709156266918839,
      "learning_rate": 1.5000000000000002e-05,
      "loss": 0.4429,
      "step": 90
    },
    {
      "epoch": 2.779746835443038,
      "grad_norm": 0.581248806563168,
      "learning_rate": 1.4864036634258112e-05,
      "loss": 0.4288,
      "step": 91
    },
    {
      "epoch": 2.810126582278481,
      "grad_norm": 0.5950440856206151,
      "learning_rate": 1.4726885051885654e-05,
      "loss": 0.4349,
      "step": 92
    },
    {
      "epoch": 2.840506329113924,
      "grad_norm": 0.5842118283358446,
      "learning_rate": 1.4588578757110359e-05,
      "loss": 0.423,
      "step": 93
    },
    {
      "epoch": 2.8708860759493673,
      "grad_norm": 0.5564773465959546,
      "learning_rate": 1.4449151536240167e-05,
      "loss": 0.4032,
      "step": 94
    },
    {
      "epoch": 2.90126582278481,
      "grad_norm": 0.5836605885622997,
      "learning_rate": 1.4308637449409705e-05,
      "loss": 0.4258,
      "step": 95
    },
    {
      "epoch": 2.9316455696202532,
      "grad_norm": 0.6015709752970122,
      "learning_rate": 1.4167070822259868e-05,
      "loss": 0.4448,
      "step": 96
    },
    {
      "epoch": 2.962025316455696,
      "grad_norm": 0.5950632127827538,
      "learning_rate": 1.402448623755254e-05,
      "loss": 0.4115,
      "step": 97
    },
    {
      "epoch": 2.992405063291139,
      "grad_norm": 0.5194543478675112,
      "learning_rate": 1.3880918526722497e-05,
      "loss": 0.4009,
      "step": 98
    },
    {
      "epoch": 3.030379746835443,
      "grad_norm": 0.844294428792226,
      "learning_rate": 1.3736402761368597e-05,
      "loss": 0.3378,
      "step": 99
    },
    {
      "epoch": 3.060759493670886,
      "grad_norm": 0.5931310088361406,
      "learning_rate": 1.3590974244686248e-05,
      "loss": 0.3088,
      "step": 100
    },
    {
      "epoch": 3.091139240506329,
      "grad_norm": 0.8535774616922295,
      "learning_rate": 1.344466850284333e-05,
      "loss": 0.318,
      "step": 101
    },
    {
      "epoch": 3.1215189873417724,
      "grad_norm": 0.8630924342324103,
      "learning_rate": 1.3297521276301666e-05,
      "loss": 0.3202,
      "step": 102
    },
    {
      "epoch": 3.151898734177215,
      "grad_norm": 0.5890704904569005,
      "learning_rate": 1.3149568511086104e-05,
      "loss": 0.3274,
      "step": 103
    },
    {
      "epoch": 3.1822784810126583,
      "grad_norm": 0.7805399006223039,
      "learning_rate": 1.300084635000341e-05,
      "loss": 0.3198,
      "step": 104
    },
    {
      "epoch": 3.212658227848101,
      "grad_norm": 0.6327679849640022,
      "learning_rate": 1.2851391123813075e-05,
      "loss": 0.3027,
      "step": 105
    },
    {
      "epoch": 3.2430379746835443,
      "grad_norm": 0.6213264905015883,
      "learning_rate": 1.2701239342352223e-05,
      "loss": 0.3179,
      "step": 106
    },
    {
      "epoch": 3.2734177215189875,
      "grad_norm": 0.623590998028256,
      "learning_rate": 1.2550427685616767e-05,
      "loss": 0.307,
      "step": 107
    },
    {
      "epoch": 3.3037974683544302,
      "grad_norm": 0.6058326286657856,
      "learning_rate": 1.239899299480098e-05,
      "loss": 0.3261,
      "step": 108
    },
    {
      "epoch": 3.3341772151898734,
      "grad_norm": 0.5782122692274975,
      "learning_rate": 1.2246972263297718e-05,
      "loss": 0.2871,
      "step": 109
    },
    {
      "epoch": 3.3645569620253166,
      "grad_norm": 0.5809386465823625,
      "learning_rate": 1.2094402627661447e-05,
      "loss": 0.3246,
      "step": 110
    },
    {
      "epoch": 3.3949367088607594,
      "grad_norm": 0.5702687462780784,
      "learning_rate": 1.1941321358536278e-05,
      "loss": 0.3195,
      "step": 111
    },
    {
      "epoch": 3.4253164556962026,
      "grad_norm": 0.5869217460582679,
      "learning_rate": 1.1787765851551296e-05,
      "loss": 0.2567,
      "step": 112
    },
    {
      "epoch": 3.4556962025316453,
      "grad_norm": 0.5298890621421983,
      "learning_rate": 1.1633773618185302e-05,
      "loss": 0.3243,
      "step": 113
    },
    {
      "epoch": 3.4860759493670885,
      "grad_norm": 0.5920489163165501,
      "learning_rate": 1.14793822766033e-05,
      "loss": 0.3468,
      "step": 114
    },
    {
      "epoch": 3.5164556962025317,
      "grad_norm": 0.5966124077220712,
      "learning_rate": 1.132462954246688e-05,
      "loss": 0.3231,
      "step": 115
    },
    {
      "epoch": 3.546835443037975,
      "grad_norm": 0.5538533170627364,
      "learning_rate": 1.1169553219720828e-05,
      "loss": 0.2997,
      "step": 116
    },
    {
      "epoch": 3.5772151898734177,
      "grad_norm": 0.5592173390657668,
      "learning_rate": 1.1014191191358118e-05,
      "loss": 0.3028,
      "step": 117
    },
    {
      "epoch": 3.607594936708861,
      "grad_norm": 0.5663294244777555,
      "learning_rate": 1.085858141016566e-05,
      "loss": 0.3283,
      "step": 118
    },
    {
      "epoch": 3.6379746835443036,
      "grad_norm": 0.5927983871712702,
      "learning_rate": 1.070276188945293e-05,
      "loss": 0.3063,
      "step": 119
    },
    {
      "epoch": 3.668354430379747,
      "grad_norm": 0.5128681037189097,
      "learning_rate": 1.0546770693765859e-05,
      "loss": 0.2973,
      "step": 120
    },
    {
      "epoch": 3.69873417721519,
      "grad_norm": 0.5013449014586845,
      "learning_rate": 1.0390645929588197e-05,
      "loss": 0.3214,
      "step": 121
    },
    {
      "epoch": 3.729113924050633,
      "grad_norm": 0.5478862358948356,
      "learning_rate": 1.0234425736032607e-05,
      "loss": 0.3189,
      "step": 122
    },
    {
      "epoch": 3.759493670886076,
      "grad_norm": 0.49382091077617946,
      "learning_rate": 1.007814827552384e-05,
      "loss": 0.3157,
      "step": 123
    },
    {
      "epoch": 3.7898734177215188,
      "grad_norm": 0.4692910313144046,
      "learning_rate": 9.92185172447616e-06,
      "loss": 0.2993,
      "step": 124
    },
    {
      "epoch": 3.820253164556962,
      "grad_norm": 0.5063799910727816,
      "learning_rate": 9.765574263967397e-06,
      "loss": 0.3084,
      "step": 125
    },
    {
      "epoch": 3.850632911392405,
      "grad_norm": 0.5080317517382535,
      "learning_rate": 9.609354070411807e-06,
      "loss": 0.3034,
      "step": 126
    },
    {
      "epoch": 3.8810126582278484,
      "grad_norm": 0.5011250158588806,
      "learning_rate": 9.453229306234143e-06,
      "loss": 0.3241,
      "step": 127
    },
    {
      "epoch": 3.911392405063291,
      "grad_norm": 0.5370165374575708,
      "learning_rate": 9.297238110547075e-06,
      "loss": 0.3236,
      "step": 128
    },
    {
      "epoch": 3.9417721518987343,
      "grad_norm": 0.49090697804891814,
      "learning_rate": 9.14141858983434e-06,
      "loss": 0.2996,
      "step": 129
    },
    {
      "epoch": 3.972151898734177,
      "grad_norm": 0.49401526369528825,
      "learning_rate": 8.985808808641883e-06,
      "loss": 0.2551,
      "step": 130
    },
    {
      "epoch": 4.010126582278481,
      "grad_norm": 0.5537769243713863,
      "learning_rate": 8.830446780279175e-06,
      "loss": 0.2737,
      "step": 131
    },
    {
      "epoch": 4.040506329113924,
      "grad_norm": 0.5148535251941122,
      "learning_rate": 8.675370457533122e-06,
      "loss": 0.2398,
      "step": 132
    },
    {
      "epoch": 4.0708860759493675,
      "grad_norm": 0.47821609099353934,
      "learning_rate": 8.520617723396702e-06,
      "loss": 0.2665,
      "step": 133
    },
    {
      "epoch": 4.10126582278481,
      "grad_norm": 0.5405131833634897,
      "learning_rate": 8.366226381814698e-06,
      "loss": 0.2307,
      "step": 134
    },
    {
      "epoch": 4.131645569620253,
      "grad_norm": 0.6044733557667396,
      "learning_rate": 8.212234148448708e-06,
      "loss": 0.2564,
      "step": 135
    },
    {
      "epoch": 4.162025316455696,
      "grad_norm": 0.45030859558024416,
      "learning_rate": 8.058678641463724e-06,
      "loss": 0.2283,
      "step": 136
    },
    {
      "epoch": 4.192405063291139,
      "grad_norm": 0.5072751954046991,
      "learning_rate": 7.905597372338558e-06,
      "loss": 0.2568,
      "step": 137
    },
    {
      "epoch": 4.222784810126583,
      "grad_norm": 0.5644224441115779,
      "learning_rate": 7.753027736702283e-06,
      "loss": 0.2286,
      "step": 138
    },
    {
      "epoch": 4.253164556962025,
      "grad_norm": 0.46193939586692945,
      "learning_rate": 7.601007005199022e-06,
      "loss": 0.2388,
      "step": 139
    },
    {
      "epoch": 4.283544303797468,
      "grad_norm": 0.4426191437097694,
      "learning_rate": 7.449572314383237e-06,
      "loss": 0.2613,
      "step": 140
    },
    {
      "epoch": 4.313924050632911,
      "grad_norm": 0.5116025449694097,
      "learning_rate": 7.298760657647779e-06,
      "loss": 0.2467,
      "step": 141
    },
    {
      "epoch": 4.3443037974683545,
      "grad_norm": 0.5122709380558271,
      "learning_rate": 7.148608876186931e-06,
      "loss": 0.2354,
      "step": 142
    },
    {
      "epoch": 4.374683544303798,
      "grad_norm": 0.4512423442981359,
      "learning_rate": 6.999153649996595e-06,
      "loss": 0.2193,
      "step": 143
    },
    {
      "epoch": 4.405063291139241,
      "grad_norm": 0.4593061845189982,
      "learning_rate": 6.8504314889138956e-06,
      "loss": 0.2344,
      "step": 144
    },
    {
      "epoch": 4.435443037974683,
      "grad_norm": 0.4582210472491683,
      "learning_rate": 6.702478723698336e-06,
      "loss": 0.2281,
      "step": 145
    },
    {
      "epoch": 4.465822784810126,
      "grad_norm": 0.5063756184850993,
      "learning_rate": 6.555331497156671e-06,
      "loss": 0.2238,
      "step": 146
    },
    {
      "epoch": 4.49620253164557,
      "grad_norm": 0.4249721125101036,
      "learning_rate": 6.4090257553137566e-06,
      "loss": 0.2533,
      "step": 147
    },
    {
      "epoch": 4.526582278481013,
      "grad_norm": 0.4423013143121516,
      "learning_rate": 6.263597238631405e-06,
      "loss": 0.2379,
      "step": 148
    },
    {
      "epoch": 4.556962025316456,
      "grad_norm": 0.4741331964802554,
      "learning_rate": 6.119081473277502e-06,
      "loss": 0.2633,
      "step": 149
    },
    {
      "epoch": 4.587341772151898,
      "grad_norm": 0.4411782948867419,
      "learning_rate": 5.975513762447465e-06,
      "loss": 0.1939,
      "step": 150
    },
    {
      "epoch": 4.6177215189873415,
      "grad_norm": 0.41894685245950497,
      "learning_rate": 5.832929177740134e-06,
      "loss": 0.2344,
      "step": 151
    },
    {
      "epoch": 4.648101265822785,
      "grad_norm": 0.4332911018848986,
      "learning_rate": 5.6913625505902966e-06,
      "loss": 0.2504,
      "step": 152
    },
    {
      "epoch": 4.678481012658228,
      "grad_norm": 0.4085930701285508,
      "learning_rate": 5.550848463759835e-06,
      "loss": 0.2367,
      "step": 153
    },
    {
      "epoch": 4.708860759493671,
      "grad_norm": 0.429451690556146,
      "learning_rate": 5.411421242889643e-06,
      "loss": 0.2167,
      "step": 154
    },
    {
      "epoch": 4.739240506329114,
      "grad_norm": 0.4254577495687622,
      "learning_rate": 5.273114948114346e-06,
      "loss": 0.2235,
      "step": 155
    },
    {
      "epoch": 4.769620253164557,
      "grad_norm": 0.40479054662232866,
      "learning_rate": 5.135963365741892e-06,
      "loss": 0.2401,
      "step": 156
    },
    {
      "epoch": 4.8,
      "grad_norm": 0.41103272209167985,
      "learning_rate": 5.000000000000003e-06,
      "loss": 0.2295,
      "step": 157
    },
    {
      "epoch": 4.830379746835443,
      "grad_norm": 0.41774063878356243,
      "learning_rate": 4.865258064851579e-06,
      "loss": 0.2239,
      "step": 158
    },
    {
      "epoch": 4.860759493670886,
      "grad_norm": 0.39818053600436154,
      "learning_rate": 4.731770475880995e-06,
      "loss": 0.2106,
      "step": 159
    },
    {
      "epoch": 4.891139240506329,
      "grad_norm": 0.3791825757559602,
      "learning_rate": 4.599569842253244e-06,
      "loss": 0.2104,
      "step": 160
    },
    {
      "epoch": 4.921518987341772,
      "grad_norm": 0.38680889693627346,
      "learning_rate": 4.468688458748006e-06,
      "loss": 0.2112,
      "step": 161
    },
    {
      "epoch": 4.951898734177215,
      "grad_norm": 0.4347540013230296,
      "learning_rate": 4.339158297870469e-06,
      "loss": 0.2011,
      "step": 162
    },
    {
      "epoch": 4.982278481012658,
      "grad_norm": 0.4210347035680862,
      "learning_rate": 4.211011002040885e-06,
      "loss": 0.2351,
      "step": 163
    },
    {
      "epoch": 5.020253164556962,
      "grad_norm": 0.44893426636255773,
      "learning_rate": 4.084277875864776e-06,
      "loss": 0.2037,
      "step": 164
    },
    {
      "epoch": 5.050632911392405,
      "grad_norm": 0.522092934598422,
      "learning_rate": 3.958989878485644e-06,
      "loss": 0.1689,
      "step": 165
    },
    {
      "epoch": 5.0810126582278485,
      "grad_norm": 0.3434404747078812,
      "learning_rate": 3.83517761602209e-06,
      "loss": 0.2016,
      "step": 166
    },
    {
      "epoch": 5.111392405063291,
      "grad_norm": 0.31974471357347284,
      "learning_rate": 3.712871334091154e-06,
      "loss": 0.1856,
      "step": 167
    },
    {
      "epoch": 5.141772151898734,
      "grad_norm": 0.4019447583894554,
      "learning_rate": 3.592100910419738e-06,
      "loss": 0.1953,
      "step": 168
    },
    {
      "epoch": 5.172151898734177,
      "grad_norm": 0.5100811606362374,
      "learning_rate": 3.4728958475459052e-06,
      "loss": 0.2383,
      "step": 169
    },
    {
      "epoch": 5.2025316455696204,
      "grad_norm": 0.48028757812661044,
      "learning_rate": 3.355285265611784e-06,
      "loss": 0.1928,
      "step": 170
    },
    {
      "epoch": 5.232911392405064,
      "grad_norm": 0.3425405707645625,
      "learning_rate": 3.2392978952499553e-06,
      "loss": 0.1602,
      "step": 171
    },
    {
      "epoch": 5.263291139240506,
      "grad_norm": 0.35077982550672243,
      "learning_rate": 3.1249620705649417e-06,
      "loss": 0.1991,
      "step": 172
    },
    {
      "epoch": 5.293670886075949,
      "grad_norm": 0.40015354653414825,
      "learning_rate": 3.0123057222115835e-06,
      "loss": 0.1865,
      "step": 173
    },
    {
      "epoch": 5.324050632911392,
      "grad_norm": 0.3744320304586383,
      "learning_rate": 2.9013563705719673e-06,
      "loss": 0.2009,
      "step": 174
    },
    {
      "epoch": 5.3544303797468356,
      "grad_norm": 0.36873467362269485,
      "learning_rate": 2.7921411190325753e-06,
      "loss": 0.1906,
      "step": 175
    },
    {
      "epoch": 5.384810126582279,
      "grad_norm": 0.35930799527816304,
      "learning_rate": 2.6846866473633126e-06,
      "loss": 0.19,
      "step": 176
    },
    {
      "epoch": 5.415189873417722,
      "grad_norm": 0.31030408724986336,
      "learning_rate": 2.579019205199992e-06,
      "loss": 0.175,
      "step": 177
    },
    {
      "epoch": 5.445569620253164,
      "grad_norm": 0.34039179147748905,
      "learning_rate": 2.4751646056319334e-06,
      "loss": 0.1959,
      "step": 178
    },
    {
      "epoch": 5.4759493670886075,
      "grad_norm": 0.35650223026553185,
      "learning_rate": 2.373148218896182e-06,
      "loss": 0.191,
      "step": 179
    },
    {
      "epoch": 5.506329113924051,
      "grad_norm": 0.33442141425419425,
      "learning_rate": 2.2729949661798876e-06,
      "loss": 0.1615,
      "step": 180
    },
    {
      "epoch": 5.536708860759494,
      "grad_norm": 0.34825977633892563,
      "learning_rate": 2.174729313532433e-06,
      "loss": 0.2029,
      "step": 181
    },
    {
      "epoch": 5.567088607594937,
      "grad_norm": 0.3159822910248115,
      "learning_rate": 2.078375265888707e-06,
      "loss": 0.1808,
      "step": 182
    },
    {
      "epoch": 5.597468354430379,
      "grad_norm": 0.31768645413132673,
      "learning_rate": 1.9839563612050273e-06,
      "loss": 0.181,
      "step": 183
    },
    {
      "epoch": 5.627848101265823,
      "grad_norm": 0.32572428592834557,
      "learning_rate": 1.8914956647091497e-06,
      "loss": 0.1823,
      "step": 184
    },
    {
      "epoch": 5.658227848101266,
      "grad_norm": 0.39224101741095874,
      "learning_rate": 1.8010157632657544e-06,
      "loss": 0.1862,
      "step": 185
    },
    {
      "epoch": 5.688607594936709,
      "grad_norm": 0.32631604828850613,
      "learning_rate": 1.7125387598587862e-06,
      "loss": 0.1853,
      "step": 186
    },
    {
      "epoch": 5.718987341772152,
      "grad_norm": 0.33319360264226067,
      "learning_rate": 1.6260862681919965e-06,
      "loss": 0.1958,
      "step": 187
    },
    {
      "epoch": 5.749367088607595,
      "grad_norm": 0.3281321636883402,
      "learning_rate": 1.5416794074090258e-06,
      "loss": 0.1912,
      "step": 188
    },
    {
      "epoch": 5.779746835443038,
      "grad_norm": 0.31016456100887957,
      "learning_rate": 1.459338796934293e-06,
      "loss": 0.1801,
      "step": 189
    },
    {
      "epoch": 5.810126582278481,
      "grad_norm": 0.30350412661300336,
      "learning_rate": 1.3790845514359363e-06,
      "loss": 0.1859,
      "step": 190
    },
    {
      "epoch": 5.840506329113924,
      "grad_norm": 0.32167848034180074,
      "learning_rate": 1.300936275912098e-06,
      "loss": 0.1819,
      "step": 191
    },
    {
      "epoch": 5.870886075949367,
      "grad_norm": 0.3359060630149123,
      "learning_rate": 1.224913060901688e-06,
      "loss": 0.2042,
      "step": 192
    },
    {
      "epoch": 5.9012658227848105,
      "grad_norm": 0.32357420028237616,
      "learning_rate": 1.1510334778208332e-06,
      "loss": 0.1859,
      "step": 193
    },
    {
      "epoch": 5.931645569620253,
      "grad_norm": 0.3285673148430449,
      "learning_rate": 1.0793155744261352e-06,
      "loss": 0.1797,
      "step": 194
    },
    {
      "epoch": 5.962025316455696,
      "grad_norm": 0.32457902216346124,
      "learning_rate": 1.0097768704058542e-06,
      "loss": 0.1815,
      "step": 195
    },
    {
      "epoch": 5.992405063291139,
      "grad_norm": 0.32110370640958585,
      "learning_rate": 9.424343531000968e-07,
      "loss": 0.1817,
      "step": 196
    },
    {
      "epoch": 6.030379746835443,
      "grad_norm": 0.31726626079789855,
      "learning_rate": 8.773044733510338e-07,
      "loss": 0.1987,
      "step": 197
    },
    {
      "epoch": 6.060759493670886,
      "grad_norm": 0.30930728491074694,
      "learning_rate": 8.144031414842012e-07,
      "loss": 0.1793,
      "step": 198
    },
    {
      "epoch": 6.091139240506329,
      "grad_norm": 0.31153226594751643,
      "learning_rate": 7.537457234218271e-07,
      "loss": 0.1939,
      "step": 199
    },
    {
      "epoch": 6.121518987341772,
      "grad_norm": 0.29120716501386895,
      "learning_rate": 6.953470369291349e-07,
      "loss": 0.193,
      "step": 200
    },
    {
      "epoch": 6.151898734177215,
      "grad_norm": 0.2758476986724978,
      "learning_rate": 6.392213479945852e-07,
      "loss": 0.1532,
      "step": 201
    },
    {
      "epoch": 6.182278481012658,
      "grad_norm": 0.2826804584168648,
      "learning_rate": 5.853823673448877e-07,
      "loss": 0.1601,
      "step": 202
    },
    {
      "epoch": 6.2126582278481015,
      "grad_norm": 0.2988206443232864,
      "learning_rate": 5.33843247095659e-07,
      "loss": 0.1641,
      "step": 203
    },
    {
      "epoch": 6.243037974683545,
      "grad_norm": 0.27842473789042943,
      "learning_rate": 4.846165775385459e-07,
      "loss": 0.1749,
      "step": 204
    },
    {
      "epoch": 6.273417721518987,
      "grad_norm": 0.27330864248275616,
      "learning_rate": 4.3771438406559173e-07,
      "loss": 0.172,
      "step": 205
    },
    {
      "epoch": 6.30379746835443,
      "grad_norm": 0.2646478326584866,
      "learning_rate": 3.931481242315993e-07,
      "loss": 0.1614,
      "step": 206
    },
    {
      "epoch": 6.334177215189873,
      "grad_norm": 0.273413308078989,
      "learning_rate": 3.5092868495520294e-07,
      "loss": 0.1701,
      "step": 207
    },
    {
      "epoch": 6.364556962025317,
      "grad_norm": 0.3056544530454989,
      "learning_rate": 3.110663798593616e-07,
      "loss": 0.1987,
      "step": 208
    },
    {
      "epoch": 6.39493670886076,
      "grad_norm": 0.2862189315220373,
      "learning_rate": 2.735709467518699e-07,
      "loss": 0.17,
      "step": 209
    },
    {
      "epoch": 6.425316455696202,
      "grad_norm": 0.30465513350706486,
      "learning_rate": 2.384515452465475e-07,
      "loss": 0.2012,
      "step": 210
    },
    {
      "epoch": 6.455696202531645,
      "grad_norm": 0.26859066396400905,
      "learning_rate": 2.0571675452567997e-07,
      "loss": 0.16,
      "step": 211
    },
    {
      "epoch": 6.4860759493670885,
      "grad_norm": 0.2756726057256473,
      "learning_rate": 1.7537457124423896e-07,
      "loss": 0.1639,
      "step": 212
    },
    {
      "epoch": 6.516455696202532,
      "grad_norm": 0.3055101341245335,
      "learning_rate": 1.474324075764111e-07,
      "loss": 0.1742,
      "step": 213
    },
    {
      "epoch": 6.546835443037975,
      "grad_norm": 0.29664410167328803,
      "learning_rate": 1.2189708940490653e-07,
      "loss": 0.1978,
      "step": 214
    },
    {
      "epoch": 6.577215189873417,
      "grad_norm": 0.30358315554916937,
      "learning_rate": 9.877485465349057e-08,
      "loss": 0.1938,
      "step": 215
    },
    {
      "epoch": 6.6075949367088604,
      "grad_norm": 0.26783707540570906,
      "learning_rate": 7.807135176314707e-08,
      "loss": 0.1565,
      "step": 216
    },
    {
      "epoch": 6.637974683544304,
      "grad_norm": 0.2467456266971825,
      "learning_rate": 5.979163831223988e-08,
      "loss": 0.1383,
      "step": 217
    },
    {
      "epoch": 6.668354430379747,
      "grad_norm": 0.25532915377077153,
      "learning_rate": 4.394017978101905e-08,
      "loss": 0.1435,
      "step": 218
    },
    {
      "epoch": 6.69873417721519,
      "grad_norm": 0.27934539566822314,
      "learning_rate": 3.0520848460765525e-08,
      "loss": 0.1709,
      "step": 219
    },
    {
      "epoch": 6.729113924050633,
      "grad_norm": 0.279947104862462,
      "learning_rate": 1.9536922507841227e-08,
      "loss": 0.1625,
      "step": 220
    },
    {
      "epoch": 6.759493670886076,
      "grad_norm": 0.3026493096332586,
      "learning_rate": 1.099108514288627e-08,
      "loss": 0.181,
      "step": 221
    },
    {
      "epoch": 6.789873417721519,
      "grad_norm": 0.28800583807923036,
      "learning_rate": 4.885423995341088e-09,
      "loss": 0.1472,
      "step": 222
    },
    {
      "epoch": 6.820253164556962,
      "grad_norm": 0.2857308690133591,
      "learning_rate": 1.2214305934699078e-09,
      "loss": 0.157,
      "step": 223
    },
    {
      "epoch": 6.850632911392405,
      "grad_norm": 0.299525443636244,
      "learning_rate": 0.0,
      "loss": 0.214,
      "step": 224
    },
    {
      "epoch": 6.850632911392405,
      "step": 224,
      "total_flos": 3.6328761054776525e+17,
      "train_loss": 0.4001694982871413,
      "train_runtime": 15845.3324,
      "train_samples_per_second": 1.396,
      "train_steps_per_second": 0.014
    }
  ],
  "logging_steps": 1,
  "max_steps": 224,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 7,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 3.6328761054776525e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}