{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.5787781350482315,
  "eval_steps": 500,
  "global_step": 200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.012861736334405145,
      "grad_norm": 1.1440598964691162,
      "learning_rate": 0.0001,
      "loss": 2.6398,
      "step": 1
    },
    {
      "epoch": 0.02572347266881029,
      "grad_norm": 1.078170657157898,
      "learning_rate": 9.949748743718594e-05,
      "loss": 2.6721,
      "step": 2
    },
    {
      "epoch": 0.03858520900321544,
      "grad_norm": 0.9559459686279297,
      "learning_rate": 9.899497487437186e-05,
      "loss": 2.6911,
      "step": 3
    },
    {
      "epoch": 0.05144694533762058,
      "grad_norm": 1.0017387866973877,
      "learning_rate": 9.84924623115578e-05,
      "loss": 2.479,
      "step": 4
    },
    {
      "epoch": 0.06430868167202572,
      "grad_norm": 1.0636959075927734,
      "learning_rate": 9.798994974874372e-05,
      "loss": 2.335,
      "step": 5
    },
    {
      "epoch": 0.07717041800643087,
      "grad_norm": 1.0521775484085083,
      "learning_rate": 9.748743718592965e-05,
      "loss": 2.1896,
      "step": 6
    },
    {
      "epoch": 0.09003215434083602,
      "grad_norm": 1.0728856325149536,
      "learning_rate": 9.698492462311559e-05,
      "loss": 2.0778,
      "step": 7
    },
    {
      "epoch": 0.10289389067524116,
      "grad_norm": 1.031778335571289,
      "learning_rate": 9.64824120603015e-05,
      "loss": 1.9751,
      "step": 8
    },
    {
      "epoch": 0.1157556270096463,
      "grad_norm": 1.076965093612671,
      "learning_rate": 9.597989949748745e-05,
      "loss": 1.7265,
      "step": 9
    },
    {
      "epoch": 0.12861736334405144,
      "grad_norm": 1.214233636856079,
      "learning_rate": 9.547738693467337e-05,
      "loss": 1.729,
      "step": 10
    },
    {
      "epoch": 0.1414790996784566,
      "grad_norm": 0.7060278058052063,
      "learning_rate": 9.49748743718593e-05,
      "loss": 1.9542,
      "step": 11
    },
    {
      "epoch": 0.15434083601286175,
      "grad_norm": 0.827420711517334,
      "learning_rate": 9.447236180904523e-05,
      "loss": 1.4983,
      "step": 12
    },
    {
      "epoch": 0.16720257234726688,
      "grad_norm": 0.6821541786193848,
      "learning_rate": 9.396984924623115e-05,
      "loss": 1.4999,
      "step": 13
    },
    {
      "epoch": 0.18006430868167203,
      "grad_norm": 0.7033599615097046,
      "learning_rate": 9.34673366834171e-05,
      "loss": 1.4026,
      "step": 14
    },
    {
      "epoch": 0.19292604501607716,
      "grad_norm": 0.80780029296875,
      "learning_rate": 9.296482412060302e-05,
      "loss": 1.4809,
      "step": 15
    },
    {
      "epoch": 0.2057877813504823,
      "grad_norm": 0.7556090950965881,
      "learning_rate": 9.246231155778895e-05,
      "loss": 1.4065,
      "step": 16
    },
    {
      "epoch": 0.21864951768488747,
      "grad_norm": 0.6941238641738892,
      "learning_rate": 9.195979899497488e-05,
      "loss": 1.4728,
      "step": 17
    },
    {
      "epoch": 0.2315112540192926,
      "grad_norm": 0.7150428295135498,
      "learning_rate": 9.14572864321608e-05,
      "loss": 1.3528,
      "step": 18
    },
    {
      "epoch": 0.24437299035369775,
      "grad_norm": 0.5909086465835571,
      "learning_rate": 9.095477386934675e-05,
      "loss": 1.3859,
      "step": 19
    },
    {
      "epoch": 0.2572347266881029,
      "grad_norm": 0.7633894681930542,
      "learning_rate": 9.045226130653267e-05,
      "loss": 1.3405,
      "step": 20
    },
    {
      "epoch": 0.27009646302250806,
      "grad_norm": 0.6418241262435913,
      "learning_rate": 8.99497487437186e-05,
      "loss": 1.3162,
      "step": 21
    },
    {
      "epoch": 0.2829581993569132,
      "grad_norm": 0.8278843760490417,
      "learning_rate": 8.944723618090453e-05,
      "loss": 1.3934,
      "step": 22
    },
    {
      "epoch": 0.2958199356913183,
      "grad_norm": 0.7349154353141785,
      "learning_rate": 8.894472361809045e-05,
      "loss": 1.3461,
      "step": 23
    },
    {
      "epoch": 0.3086816720257235,
      "grad_norm": 0.736482560634613,
      "learning_rate": 8.84422110552764e-05,
      "loss": 1.2912,
      "step": 24
    },
    {
      "epoch": 0.3215434083601286,
      "grad_norm": 0.8143345713615417,
      "learning_rate": 8.793969849246232e-05,
      "loss": 1.3445,
      "step": 25
    },
    {
      "epoch": 0.33440514469453375,
      "grad_norm": 0.6922888159751892,
      "learning_rate": 8.743718592964825e-05,
      "loss": 1.36,
      "step": 26
    },
    {
      "epoch": 0.34726688102893893,
      "grad_norm": 0.7785637378692627,
      "learning_rate": 8.693467336683418e-05,
      "loss": 1.276,
      "step": 27
    },
    {
      "epoch": 0.36012861736334406,
      "grad_norm": 0.7577537298202515,
      "learning_rate": 8.64321608040201e-05,
      "loss": 1.2323,
      "step": 28
    },
    {
      "epoch": 0.3729903536977492,
      "grad_norm": 0.7919734716415405,
      "learning_rate": 8.592964824120603e-05,
      "loss": 1.4053,
      "step": 29
    },
    {
      "epoch": 0.3858520900321543,
      "grad_norm": 0.821456789970398,
      "learning_rate": 8.542713567839196e-05,
      "loss": 1.251,
      "step": 30
    },
    {
      "epoch": 0.3987138263665595,
      "grad_norm": 0.8000316023826599,
      "learning_rate": 8.49246231155779e-05,
      "loss": 1.3022,
      "step": 31
    },
    {
      "epoch": 0.4115755627009646,
      "grad_norm": 0.7909243702888489,
      "learning_rate": 8.442211055276383e-05,
      "loss": 1.2509,
      "step": 32
    },
    {
      "epoch": 0.42443729903536975,
      "grad_norm": 0.896696150302887,
      "learning_rate": 8.391959798994975e-05,
      "loss": 1.2068,
      "step": 33
    },
    {
      "epoch": 0.43729903536977494,
      "grad_norm": 0.8683031797409058,
      "learning_rate": 8.341708542713568e-05,
      "loss": 1.3849,
      "step": 34
    },
    {
      "epoch": 0.45016077170418006,
      "grad_norm": 0.8988001346588135,
      "learning_rate": 8.291457286432161e-05,
      "loss": 1.1935,
      "step": 35
    },
    {
      "epoch": 0.4630225080385852,
      "grad_norm": 0.8507208824157715,
      "learning_rate": 8.241206030150754e-05,
      "loss": 1.2605,
      "step": 36
    },
    {
      "epoch": 0.4758842443729904,
      "grad_norm": 0.9080713987350464,
      "learning_rate": 8.190954773869348e-05,
      "loss": 1.3304,
      "step": 37
    },
    {
      "epoch": 0.4887459807073955,
      "grad_norm": 0.9472977519035339,
      "learning_rate": 8.14070351758794e-05,
      "loss": 1.3325,
      "step": 38
    },
    {
      "epoch": 0.5016077170418006,
      "grad_norm": 0.9067745804786682,
      "learning_rate": 8.090452261306533e-05,
      "loss": 1.195,
      "step": 39
    },
    {
      "epoch": 0.5144694533762058,
      "grad_norm": 0.7469335198402405,
      "learning_rate": 8.040201005025126e-05,
      "loss": 1.348,
      "step": 40
    },
    {
      "epoch": 0.5273311897106109,
      "grad_norm": 0.915685772895813,
      "learning_rate": 7.989949748743719e-05,
      "loss": 1.2406,
      "step": 41
    },
    {
      "epoch": 0.5401929260450161,
      "grad_norm": 0.6561890244483948,
      "learning_rate": 7.939698492462313e-05,
      "loss": 1.4992,
      "step": 42
    },
    {
      "epoch": 0.5530546623794212,
      "grad_norm": 0.7920067310333252,
      "learning_rate": 7.889447236180904e-05,
      "loss": 1.1802,
      "step": 43
    },
    {
      "epoch": 0.5659163987138264,
      "grad_norm": 0.7714186906814575,
      "learning_rate": 7.839195979899498e-05,
      "loss": 1.2086,
      "step": 44
    },
    {
      "epoch": 0.5787781350482315,
      "grad_norm": 0.7048550844192505,
      "learning_rate": 7.788944723618091e-05,
      "loss": 1.166,
      "step": 45
    },
    {
      "epoch": 0.5916398713826366,
      "grad_norm": 0.5653746128082275,
      "learning_rate": 7.738693467336684e-05,
      "loss": 1.3876,
      "step": 46
    },
    {
      "epoch": 0.6045016077170418,
      "grad_norm": 0.6872064471244812,
      "learning_rate": 7.688442211055277e-05,
      "loss": 1.1946,
      "step": 47
    },
    {
      "epoch": 0.617363344051447,
      "grad_norm": 0.6310857534408569,
      "learning_rate": 7.638190954773869e-05,
      "loss": 1.1509,
      "step": 48
    },
    {
      "epoch": 0.6302250803858521,
      "grad_norm": 0.7433875203132629,
      "learning_rate": 7.587939698492463e-05,
      "loss": 1.1792,
      "step": 49
    },
    {
      "epoch": 0.6430868167202572,
      "grad_norm": 0.7774860262870789,
      "learning_rate": 7.537688442211056e-05,
      "loss": 1.3531,
      "step": 50
    },
    {
      "epoch": 0.6559485530546624,
      "grad_norm": 0.6565365195274353,
      "learning_rate": 7.487437185929649e-05,
      "loss": 1.2448,
      "step": 51
    },
    {
      "epoch": 0.6688102893890675,
      "grad_norm": 0.6934945583343506,
      "learning_rate": 7.437185929648241e-05,
      "loss": 1.3109,
      "step": 52
    },
    {
      "epoch": 0.6816720257234726,
      "grad_norm": 0.662295401096344,
      "learning_rate": 7.386934673366834e-05,
      "loss": 1.3463,
      "step": 53
    },
    {
      "epoch": 0.6945337620578779,
      "grad_norm": 0.6406717300415039,
      "learning_rate": 7.336683417085427e-05,
      "loss": 1.0897,
      "step": 54
    },
    {
      "epoch": 0.707395498392283,
      "grad_norm": 0.7470242977142334,
      "learning_rate": 7.28643216080402e-05,
      "loss": 1.1834,
      "step": 55
    },
    {
      "epoch": 0.7202572347266881,
      "grad_norm": 0.6413742899894714,
      "learning_rate": 7.236180904522614e-05,
      "loss": 1.313,
      "step": 56
    },
    {
      "epoch": 0.7331189710610932,
      "grad_norm": 0.64141845703125,
      "learning_rate": 7.185929648241206e-05,
      "loss": 1.322,
      "step": 57
    },
    {
      "epoch": 0.7459807073954984,
      "grad_norm": 0.6849397420883179,
      "learning_rate": 7.135678391959799e-05,
      "loss": 1.1466,
      "step": 58
    },
    {
      "epoch": 0.7588424437299035,
      "grad_norm": 0.7390387058258057,
      "learning_rate": 7.085427135678392e-05,
      "loss": 1.1692,
      "step": 59
    },
    {
      "epoch": 0.7717041800643086,
      "grad_norm": 0.7512691617012024,
      "learning_rate": 7.035175879396985e-05,
      "loss": 1.2447,
      "step": 60
    },
    {
      "epoch": 0.7845659163987139,
      "grad_norm": 0.7736090421676636,
      "learning_rate": 6.984924623115579e-05,
      "loss": 1.1723,
      "step": 61
    },
    {
      "epoch": 0.797427652733119,
      "grad_norm": 0.7576032280921936,
      "learning_rate": 6.93467336683417e-05,
      "loss": 1.2074,
      "step": 62
    },
    {
      "epoch": 0.8102893890675241,
      "grad_norm": 0.7597335577011108,
      "learning_rate": 6.884422110552764e-05,
      "loss": 1.1425,
      "step": 63
    },
    {
      "epoch": 0.8231511254019293,
      "grad_norm": 0.774594247341156,
      "learning_rate": 6.834170854271357e-05,
      "loss": 1.3168,
      "step": 64
    },
    {
      "epoch": 0.8360128617363344,
      "grad_norm": 0.7835249900817871,
      "learning_rate": 6.78391959798995e-05,
      "loss": 0.9976,
      "step": 65
    },
    {
      "epoch": 0.8488745980707395,
      "grad_norm": 0.915006160736084,
      "learning_rate": 6.733668341708544e-05,
      "loss": 1.2015,
      "step": 66
    },
    {
      "epoch": 0.8617363344051447,
      "grad_norm": 0.7381256818771362,
      "learning_rate": 6.683417085427135e-05,
      "loss": 1.1394,
      "step": 67
    },
    {
      "epoch": 0.8745980707395499,
      "grad_norm": 0.6990655064582825,
      "learning_rate": 6.633165829145729e-05,
      "loss": 1.1879,
      "step": 68
    },
    {
      "epoch": 0.887459807073955,
      "grad_norm": 0.7526246309280396,
      "learning_rate": 6.582914572864322e-05,
      "loss": 1.2836,
      "step": 69
    },
    {
      "epoch": 0.9003215434083601,
      "grad_norm": 0.6938768029212952,
      "learning_rate": 6.532663316582915e-05,
      "loss": 1.1957,
      "step": 70
    },
    {
      "epoch": 0.9131832797427653,
      "grad_norm": 0.7444002032279968,
      "learning_rate": 6.482412060301508e-05,
      "loss": 1.2171,
      "step": 71
    },
    {
      "epoch": 0.9260450160771704,
      "grad_norm": 0.7161276340484619,
      "learning_rate": 6.4321608040201e-05,
      "loss": 1.1213,
      "step": 72
    },
    {
      "epoch": 0.9389067524115756,
      "grad_norm": 0.7644099593162537,
      "learning_rate": 6.381909547738694e-05,
      "loss": 1.0821,
      "step": 73
    },
    {
      "epoch": 0.9517684887459807,
      "grad_norm": 0.6624590754508972,
      "learning_rate": 6.331658291457287e-05,
      "loss": 1.233,
      "step": 74
    },
    {
      "epoch": 0.9646302250803859,
      "grad_norm": 0.7556670308113098,
      "learning_rate": 6.28140703517588e-05,
      "loss": 1.2724,
      "step": 75
    },
    {
      "epoch": 0.977491961414791,
      "grad_norm": 0.7611079812049866,
      "learning_rate": 6.231155778894473e-05,
      "loss": 1.1659,
      "step": 76
    },
    {
      "epoch": 0.9903536977491961,
      "grad_norm": 0.7112507224082947,
      "learning_rate": 6.180904522613065e-05,
      "loss": 1.1363,
      "step": 77
    },
    {
      "epoch": 1.0064308681672025,
      "grad_norm": 1.1092808246612549,
      "learning_rate": 6.130653266331658e-05,
      "loss": 1.6764,
      "step": 78
    },
    {
      "epoch": 1.0192926045016077,
      "grad_norm": 0.5934129357337952,
      "learning_rate": 6.080402010050251e-05,
      "loss": 1.0156,
      "step": 79
    },
    {
      "epoch": 1.0321543408360128,
      "grad_norm": 0.6461953520774841,
      "learning_rate": 6.030150753768844e-05,
      "loss": 1.116,
      "step": 80
    },
    {
      "epoch": 1.045016077170418,
      "grad_norm": 0.6718643307685852,
      "learning_rate": 5.979899497487438e-05,
      "loss": 1.1288,
      "step": 81
    },
    {
      "epoch": 1.0578778135048232,
      "grad_norm": 0.6698076128959656,
      "learning_rate": 5.929648241206031e-05,
      "loss": 1.1071,
      "step": 82
    },
    {
      "epoch": 1.0707395498392283,
      "grad_norm": 0.630225419998169,
      "learning_rate": 5.879396984924623e-05,
      "loss": 1.0615,
      "step": 83
    },
    {
      "epoch": 1.0836012861736335,
      "grad_norm": 0.7957659959793091,
      "learning_rate": 5.829145728643216e-05,
      "loss": 1.1099,
      "step": 84
    },
    {
      "epoch": 1.0964630225080385,
      "grad_norm": 0.7359949946403503,
      "learning_rate": 5.778894472361809e-05,
      "loss": 1.1107,
      "step": 85
    },
    {
      "epoch": 1.1093247588424437,
      "grad_norm": 0.7295353412628174,
      "learning_rate": 5.728643216080403e-05,
      "loss": 1.093,
      "step": 86
    },
    {
      "epoch": 1.122186495176849,
      "grad_norm": 0.6929362416267395,
      "learning_rate": 5.6783919597989955e-05,
      "loss": 1.0855,
      "step": 87
    },
    {
      "epoch": 1.135048231511254,
      "grad_norm": 0.6977247595787048,
      "learning_rate": 5.628140703517588e-05,
      "loss": 1.1056,
      "step": 88
    },
    {
      "epoch": 1.1479099678456592,
      "grad_norm": 0.7273219227790833,
      "learning_rate": 5.577889447236181e-05,
      "loss": 1.1319,
      "step": 89
    },
    {
      "epoch": 1.1607717041800643,
      "grad_norm": 0.7561770081520081,
      "learning_rate": 5.527638190954774e-05,
      "loss": 1.1406,
      "step": 90
    },
    {
      "epoch": 1.1736334405144695,
      "grad_norm": 0.7354375720024109,
      "learning_rate": 5.477386934673368e-05,
      "loss": 1.8666,
      "step": 91
    },
    {
      "epoch": 1.1864951768488745,
      "grad_norm": 0.6309502124786377,
      "learning_rate": 5.4271356783919604e-05,
      "loss": 0.9311,
      "step": 92
    },
    {
      "epoch": 1.1993569131832797,
      "grad_norm": 0.6576774716377258,
      "learning_rate": 5.376884422110553e-05,
      "loss": 0.9488,
      "step": 93
    },
    {
      "epoch": 1.212218649517685,
      "grad_norm": 0.8312212228775024,
      "learning_rate": 5.3266331658291455e-05,
      "loss": 1.2609,
      "step": 94
    },
    {
      "epoch": 1.22508038585209,
      "grad_norm": 0.6623194813728333,
      "learning_rate": 5.276381909547739e-05,
      "loss": 0.9066,
      "step": 95
    },
    {
      "epoch": 1.2379421221864952,
      "grad_norm": 0.8220139741897583,
      "learning_rate": 5.226130653266332e-05,
      "loss": 1.2762,
      "step": 96
    },
    {
      "epoch": 1.2508038585209003,
      "grad_norm": 0.8111358880996704,
      "learning_rate": 5.175879396984925e-05,
      "loss": 1.1121,
      "step": 97
    },
    {
      "epoch": 1.2636655948553055,
      "grad_norm": 0.8110450506210327,
      "learning_rate": 5.125628140703518e-05,
      "loss": 1.1078,
      "step": 98
    },
    {
      "epoch": 1.2765273311897105,
      "grad_norm": 0.7321817874908447,
      "learning_rate": 5.0753768844221104e-05,
      "loss": 1.1252,
      "step": 99
    },
    {
      "epoch": 1.2893890675241158,
      "grad_norm": 0.7316034436225891,
      "learning_rate": 5.0251256281407036e-05,
      "loss": 1.0643,
      "step": 100
    },
    {
      "epoch": 1.302250803858521,
      "grad_norm": 0.7192660570144653,
      "learning_rate": 4.974874371859297e-05,
      "loss": 1.1715,
      "step": 101
    },
    {
      "epoch": 1.315112540192926,
      "grad_norm": 0.7240356206893921,
      "learning_rate": 4.92462311557789e-05,
      "loss": 1.0688,
      "step": 102
    },
    {
      "epoch": 1.3279742765273312,
      "grad_norm": 0.6924866437911987,
      "learning_rate": 4.874371859296483e-05,
      "loss": 0.9077,
      "step": 103
    },
    {
      "epoch": 1.3408360128617363,
      "grad_norm": 0.7501733303070068,
      "learning_rate": 4.824120603015075e-05,
      "loss": 1.5946,
      "step": 104
    },
    {
      "epoch": 1.3536977491961415,
      "grad_norm": 0.807327926158905,
      "learning_rate": 4.7738693467336685e-05,
      "loss": 1.1421,
      "step": 105
    },
    {
      "epoch": 1.3665594855305465,
      "grad_norm": 0.7556862235069275,
      "learning_rate": 4.723618090452262e-05,
      "loss": 0.9923,
      "step": 106
    },
    {
      "epoch": 1.3794212218649518,
      "grad_norm": 0.7407500743865967,
      "learning_rate": 4.673366834170855e-05,
      "loss": 0.9658,
      "step": 107
    },
    {
      "epoch": 1.392282958199357,
      "grad_norm": 0.7646963596343994,
      "learning_rate": 4.6231155778894475e-05,
      "loss": 1.0564,
      "step": 108
    },
    {
      "epoch": 1.405144694533762,
      "grad_norm": 0.782166063785553,
      "learning_rate": 4.57286432160804e-05,
      "loss": 0.9601,
      "step": 109
    },
    {
      "epoch": 1.4180064308681672,
      "grad_norm": 0.7942412495613098,
      "learning_rate": 4.522613065326633e-05,
      "loss": 1.0054,
      "step": 110
    },
    {
      "epoch": 1.4308681672025725,
      "grad_norm": 0.7960532307624817,
      "learning_rate": 4.4723618090452266e-05,
      "loss": 1.0791,
      "step": 111
    },
    {
      "epoch": 1.4437299035369775,
      "grad_norm": 0.7725895643234253,
      "learning_rate": 4.42211055276382e-05,
      "loss": 1.0099,
      "step": 112
    },
    {
      "epoch": 1.4565916398713825,
      "grad_norm": 0.8653061985969543,
      "learning_rate": 4.3718592964824124e-05,
      "loss": 1.168,
      "step": 113
    },
    {
      "epoch": 1.4694533762057878,
      "grad_norm": 0.8689178824424744,
      "learning_rate": 4.321608040201005e-05,
      "loss": 1.1026,
      "step": 114
    },
    {
      "epoch": 1.482315112540193,
      "grad_norm": 0.823765754699707,
      "learning_rate": 4.271356783919598e-05,
      "loss": 1.3272,
      "step": 115
    },
    {
      "epoch": 1.495176848874598,
      "grad_norm": 0.715006947517395,
      "learning_rate": 4.2211055276381914e-05,
      "loss": 0.9572,
      "step": 116
    },
    {
      "epoch": 1.5080385852090032,
      "grad_norm": 0.9381377696990967,
      "learning_rate": 4.170854271356784e-05,
      "loss": 1.0988,
      "step": 117
    },
    {
      "epoch": 1.5209003215434085,
      "grad_norm": 0.7912129759788513,
      "learning_rate": 4.120603015075377e-05,
      "loss": 1.0784,
      "step": 118
    },
    {
      "epoch": 1.5337620578778135,
      "grad_norm": 0.8803995251655579,
      "learning_rate": 4.07035175879397e-05,
      "loss": 1.1845,
      "step": 119
    },
    {
      "epoch": 1.5466237942122185,
      "grad_norm": 0.7034085988998413,
      "learning_rate": 4.020100502512563e-05,
      "loss": 1.016,
      "step": 120
    },
    {
      "epoch": 1.5594855305466238,
      "grad_norm": 0.8203994035720825,
      "learning_rate": 3.969849246231156e-05,
      "loss": 1.063,
      "step": 121
    },
    {
      "epoch": 1.572347266881029,
      "grad_norm": 0.712227463722229,
      "learning_rate": 3.919597989949749e-05,
      "loss": 1.033,
      "step": 122
    },
    {
      "epoch": 1.585209003215434,
      "grad_norm": 0.7563191056251526,
      "learning_rate": 3.869346733668342e-05,
      "loss": 1.0089,
      "step": 123
    },
    {
      "epoch": 1.5980707395498392,
      "grad_norm": 0.8773857355117798,
      "learning_rate": 3.8190954773869346e-05,
      "loss": 1.4975,
      "step": 124
    },
    {
      "epoch": 1.6109324758842445,
      "grad_norm": 0.6820680499076843,
      "learning_rate": 3.768844221105528e-05,
      "loss": 0.8996,
      "step": 125
    },
    {
      "epoch": 1.6237942122186495,
      "grad_norm": 0.7820584774017334,
      "learning_rate": 3.7185929648241204e-05,
      "loss": 0.9505,
      "step": 126
    },
    {
      "epoch": 1.6366559485530545,
      "grad_norm": 0.7456291913986206,
      "learning_rate": 3.668341708542714e-05,
      "loss": 0.9537,
      "step": 127
    },
    {
      "epoch": 1.6495176848874598,
      "grad_norm": 0.9636368155479431,
      "learning_rate": 3.618090452261307e-05,
      "loss": 1.1609,
      "step": 128
    },
    {
      "epoch": 1.662379421221865,
      "grad_norm": 0.83909672498703,
      "learning_rate": 3.5678391959798995e-05,
      "loss": 1.0519,
      "step": 129
    },
    {
      "epoch": 1.67524115755627,
      "grad_norm": 0.9138084053993225,
      "learning_rate": 3.517587939698493e-05,
      "loss": 0.977,
      "step": 130
    },
    {
      "epoch": 1.6881028938906752,
      "grad_norm": 0.8310115337371826,
      "learning_rate": 3.467336683417085e-05,
      "loss": 1.0701,
      "step": 131
    },
    {
      "epoch": 1.7009646302250805,
      "grad_norm": 0.8289808630943298,
      "learning_rate": 3.4170854271356785e-05,
      "loss": 1.0607,
      "step": 132
    },
    {
      "epoch": 1.7138263665594855,
      "grad_norm": 0.9004020094871521,
      "learning_rate": 3.366834170854272e-05,
      "loss": 1.3209,
      "step": 133
    },
    {
      "epoch": 1.7266881028938905,
      "grad_norm": 0.7358340620994568,
      "learning_rate": 3.3165829145728643e-05,
      "loss": 0.9646,
      "step": 134
    },
    {
      "epoch": 1.739549839228296,
      "grad_norm": 0.8724610805511475,
      "learning_rate": 3.2663316582914576e-05,
      "loss": 0.9834,
      "step": 135
    },
    {
      "epoch": 1.752411575562701,
      "grad_norm": 0.9433433413505554,
      "learning_rate": 3.21608040201005e-05,
      "loss": 0.9332,
      "step": 136
    },
    {
      "epoch": 1.765273311897106,
      "grad_norm": 0.9258175492286682,
      "learning_rate": 3.1658291457286434e-05,
      "loss": 1.0992,
      "step": 137
    },
    {
      "epoch": 1.7781350482315113,
      "grad_norm": 0.8666226267814636,
      "learning_rate": 3.1155778894472366e-05,
      "loss": 1.0032,
      "step": 138
    },
    {
      "epoch": 1.7909967845659165,
      "grad_norm": 0.9296693205833435,
      "learning_rate": 3.065326633165829e-05,
      "loss": 1.082,
      "step": 139
    },
    {
      "epoch": 1.8038585209003215,
      "grad_norm": 0.8250362873077393,
      "learning_rate": 3.015075376884422e-05,
      "loss": 0.8665,
      "step": 140
    },
    {
      "epoch": 1.8167202572347267,
      "grad_norm": 0.9356509447097778,
      "learning_rate": 2.9648241206030153e-05,
      "loss": 1.1933,
      "step": 141
    },
    {
      "epoch": 1.829581993569132,
      "grad_norm": 0.935892641544342,
      "learning_rate": 2.914572864321608e-05,
      "loss": 1.0981,
      "step": 142
    },
    {
      "epoch": 1.842443729903537,
      "grad_norm": 0.8205680251121521,
      "learning_rate": 2.8643216080402015e-05,
      "loss": 1.0501,
      "step": 143
    },
    {
      "epoch": 1.855305466237942,
      "grad_norm": 0.9360036253929138,
      "learning_rate": 2.814070351758794e-05,
      "loss": 1.1749,
      "step": 144
    },
    {
      "epoch": 1.8681672025723473,
      "grad_norm": 0.9020069241523743,
      "learning_rate": 2.763819095477387e-05,
      "loss": 1.0402,
      "step": 145
    },
    {
      "epoch": 1.8810289389067525,
      "grad_norm": 0.7428980469703674,
      "learning_rate": 2.7135678391959802e-05,
      "loss": 0.9236,
      "step": 146
    },
    {
      "epoch": 1.8938906752411575,
      "grad_norm": 0.8405928611755371,
      "learning_rate": 2.6633165829145728e-05,
      "loss": 1.2336,
      "step": 147
    },
    {
      "epoch": 1.9067524115755627,
      "grad_norm": 0.954319417476654,
      "learning_rate": 2.613065326633166e-05,
      "loss": 1.0338,
      "step": 148
    },
    {
      "epoch": 1.919614147909968,
      "grad_norm": 0.836933970451355,
      "learning_rate": 2.562814070351759e-05,
      "loss": 0.9513,
      "step": 149
    },
    {
      "epoch": 1.932475884244373,
      "grad_norm": 0.8751674294471741,
      "learning_rate": 2.5125628140703518e-05,
      "loss": 1.1418,
      "step": 150
    },
    {
      "epoch": 1.945337620578778,
      "grad_norm": 0.7700965404510498,
      "learning_rate": 2.462311557788945e-05,
      "loss": 1.0421,
      "step": 151
    },
    {
      "epoch": 1.9581993569131833,
      "grad_norm": 0.8978580832481384,
      "learning_rate": 2.4120603015075376e-05,
      "loss": 1.124,
      "step": 152
    },
    {
      "epoch": 1.9710610932475885,
      "grad_norm": 0.8578283786773682,
      "learning_rate": 2.361809045226131e-05,
      "loss": 1.1622,
      "step": 153
    },
    {
      "epoch": 1.9839228295819935,
      "grad_norm": 0.8631901741027832,
      "learning_rate": 2.3115577889447238e-05,
      "loss": 1.0854,
      "step": 154
    },
    {
      "epoch": 1.9967845659163987,
      "grad_norm": 1.506197452545166,
      "learning_rate": 2.2613065326633167e-05,
      "loss": 1.5602,
      "step": 155
    },
    {
      "epoch": 2.012861736334405,
      "grad_norm": 0.8579190969467163,
      "learning_rate": 2.21105527638191e-05,
      "loss": 1.0135,
      "step": 156
    },
    {
      "epoch": 2.0257234726688105,
      "grad_norm": 0.791259229183197,
      "learning_rate": 2.1608040201005025e-05,
      "loss": 1.063,
      "step": 157
    },
    {
      "epoch": 2.0385852090032155,
      "grad_norm": 0.7824810743331909,
      "learning_rate": 2.1105527638190957e-05,
      "loss": 0.9797,
      "step": 158
    },
    {
      "epoch": 2.0514469453376205,
      "grad_norm": 0.7896823287010193,
      "learning_rate": 2.0603015075376886e-05,
      "loss": 0.9882,
      "step": 159
    },
    {
      "epoch": 2.0643086816720255,
      "grad_norm": 0.8175792098045349,
      "learning_rate": 2.0100502512562815e-05,
      "loss": 0.9608,
      "step": 160
    },
    {
      "epoch": 2.077170418006431,
      "grad_norm": 0.819749116897583,
      "learning_rate": 1.9597989949748744e-05,
      "loss": 1.0491,
      "step": 161
    },
    {
      "epoch": 2.090032154340836,
      "grad_norm": 0.8855689764022827,
      "learning_rate": 1.9095477386934673e-05,
      "loss": 1.0228,
      "step": 162
    },
    {
      "epoch": 2.102893890675241,
      "grad_norm": 0.8208020925521851,
      "learning_rate": 1.8592964824120602e-05,
      "loss": 0.9609,
      "step": 163
    },
    {
      "epoch": 2.1157556270096465,
      "grad_norm": 0.7887352108955383,
      "learning_rate": 1.8090452261306535e-05,
      "loss": 0.9935,
      "step": 164
    },
    {
      "epoch": 2.1286173633440515,
      "grad_norm": 0.7535598874092102,
      "learning_rate": 1.7587939698492464e-05,
      "loss": 1.0082,
      "step": 165
    },
    {
      "epoch": 2.1414790996784565,
      "grad_norm": 0.8124526739120483,
      "learning_rate": 1.7085427135678393e-05,
      "loss": 1.0349,
      "step": 166
    },
    {
      "epoch": 2.154340836012862,
      "grad_norm": 0.8419150114059448,
      "learning_rate": 1.6582914572864322e-05,
      "loss": 0.9734,
      "step": 167
    },
    {
      "epoch": 2.167202572347267,
      "grad_norm": 0.7929818034172058,
      "learning_rate": 1.608040201005025e-05,
      "loss": 0.9844,
      "step": 168
    },
    {
      "epoch": 2.180064308681672,
      "grad_norm": 0.8568313121795654,
      "learning_rate": 1.5577889447236183e-05,
      "loss": 0.9512,
      "step": 169
    },
    {
      "epoch": 2.192926045016077,
      "grad_norm": 0.7779914736747742,
      "learning_rate": 1.507537688442211e-05,
      "loss": 1.4218,
      "step": 170
    },
    {
      "epoch": 2.2057877813504825,
      "grad_norm": 0.8760497570037842,
      "learning_rate": 1.457286432160804e-05,
      "loss": 0.9962,
      "step": 171
    },
    {
      "epoch": 2.2186495176848875,
      "grad_norm": 0.9114797711372375,
      "learning_rate": 1.407035175879397e-05,
      "loss": 0.9851,
      "step": 172
    },
    {
      "epoch": 2.2315112540192925,
      "grad_norm": 0.9769560694694519,
      "learning_rate": 1.3567839195979901e-05,
      "loss": 0.9286,
      "step": 173
    },
    {
      "epoch": 2.244372990353698,
      "grad_norm": 0.8264016509056091,
      "learning_rate": 1.306532663316583e-05,
      "loss": 1.105,
      "step": 174
    },
    {
      "epoch": 2.257234726688103,
      "grad_norm": 0.845151960849762,
      "learning_rate": 1.2562814070351759e-05,
      "loss": 1.0246,
      "step": 175
    },
    {
      "epoch": 2.270096463022508,
      "grad_norm": 0.8815337419509888,
      "learning_rate": 1.2060301507537688e-05,
      "loss": 1.123,
      "step": 176
    },
    {
      "epoch": 2.282958199356913,
      "grad_norm": 0.8236774802207947,
      "learning_rate": 1.1557788944723619e-05,
      "loss": 1.0557,
      "step": 177
    },
    {
      "epoch": 2.2958199356913185,
      "grad_norm": 0.9845472574234009,
      "learning_rate": 1.105527638190955e-05,
      "loss": 1.0297,
      "step": 178
    },
    {
      "epoch": 2.3086816720257235,
      "grad_norm": 0.9996894001960754,
      "learning_rate": 1.0552763819095479e-05,
      "loss": 0.9584,
      "step": 179
    },
    {
      "epoch": 2.3215434083601285,
      "grad_norm": 0.873965859413147,
      "learning_rate": 1.0050251256281408e-05,
      "loss": 0.9794,
      "step": 180
    },
    {
      "epoch": 2.334405144694534,
      "grad_norm": 0.8431399464607239,
      "learning_rate": 9.547738693467337e-06,
      "loss": 0.982,
      "step": 181
    },
    {
      "epoch": 2.347266881028939,
      "grad_norm": 0.8434001207351685,
      "learning_rate": 9.045226130653267e-06,
      "loss": 0.9264,
      "step": 182
    },
    {
      "epoch": 2.360128617363344,
      "grad_norm": 0.9110084176063538,
      "learning_rate": 8.542713567839196e-06,
      "loss": 0.9953,
      "step": 183
    },
    {
      "epoch": 2.372990353697749,
      "grad_norm": 0.7942981719970703,
      "learning_rate": 8.040201005025125e-06,
      "loss": 1.0177,
      "step": 184
    },
    {
      "epoch": 2.3858520900321545,
      "grad_norm": 0.8465241193771362,
      "learning_rate": 7.537688442211055e-06,
      "loss": 0.9383,
      "step": 185
    },
    {
      "epoch": 2.3987138263665595,
      "grad_norm": 0.9114559888839722,
      "learning_rate": 7.035175879396985e-06,
      "loss": 1.0373,
      "step": 186
    },
    {
      "epoch": 2.4115755627009645,
      "grad_norm": 0.9119040369987488,
      "learning_rate": 6.532663316582915e-06,
      "loss": 0.9853,
      "step": 187
    },
    {
      "epoch": 2.42443729903537,
      "grad_norm": 0.8535617589950562,
      "learning_rate": 6.030150753768844e-06,
      "loss": 1.0043,
      "step": 188
    },
    {
      "epoch": 2.437299035369775,
      "grad_norm": 0.8427352905273438,
      "learning_rate": 5.527638190954775e-06,
      "loss": 0.9835,
      "step": 189
    },
    {
      "epoch": 2.45016077170418,
      "grad_norm": 0.8029307723045349,
      "learning_rate": 5.025125628140704e-06,
      "loss": 1.0006,
      "step": 190
    },
    {
      "epoch": 2.463022508038585,
      "grad_norm": 0.7396455407142639,
      "learning_rate": 4.522613065326634e-06,
      "loss": 1.2503,
      "step": 191
    },
    {
      "epoch": 2.4758842443729905,
      "grad_norm": 0.9325175881385803,
      "learning_rate": 4.020100502512563e-06,
      "loss": 1.014,
      "step": 192
    },
    {
      "epoch": 2.4887459807073955,
      "grad_norm": 0.8711210489273071,
      "learning_rate": 3.5175879396984926e-06,
      "loss": 0.8779,
      "step": 193
    },
    {
      "epoch": 2.5016077170418005,
      "grad_norm": 0.9022113680839539,
      "learning_rate": 3.015075376884422e-06,
      "loss": 1.0297,
      "step": 194
    },
    {
      "epoch": 2.514469453376206,
      "grad_norm": 0.8506854176521301,
      "learning_rate": 2.512562814070352e-06,
      "loss": 0.8196,
      "step": 195
    },
    {
      "epoch": 2.527331189710611,
      "grad_norm": 0.9505279064178467,
      "learning_rate": 2.0100502512562813e-06,
      "loss": 0.9124,
      "step": 196
    },
    {
      "epoch": 2.540192926045016,
      "grad_norm": 0.8015897870063782,
      "learning_rate": 1.507537688442211e-06,
      "loss": 0.9492,
      "step": 197
    },
    {
      "epoch": 2.553054662379421,
      "grad_norm": 0.8406469821929932,
      "learning_rate": 1.0050251256281407e-06,
      "loss": 0.9474,
      "step": 198
    },
    {
      "epoch": 2.5659163987138265,
      "grad_norm": 0.8296630382537842,
      "learning_rate": 5.025125628140703e-07,
      "loss": 0.9759,
      "step": 199
    },
    {
      "epoch": 2.5787781350482315,
      "grad_norm": 0.8576996922492981,
      "learning_rate": 0.0,
      "loss": 0.849,
      "step": 200
    }
  ],
  "logging_steps": 1,
  "max_steps": 200,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "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": 2.879413794353971e+16,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}