{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 5.0,
  "eval_steps": 500,
  "global_step": 215,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.023255813953488372,
      "grad_norm": 5.434192361984504,
      "learning_rate": 3.6363636363636366e-06,
      "loss": 0.9705,
      "step": 1
    },
    {
      "epoch": 0.046511627906976744,
      "grad_norm": 5.477136338528377,
      "learning_rate": 7.272727272727273e-06,
      "loss": 0.9693,
      "step": 2
    },
    {
      "epoch": 0.06976744186046512,
      "grad_norm": 4.994172293091364,
      "learning_rate": 1.0909090909090909e-05,
      "loss": 0.971,
      "step": 3
    },
    {
      "epoch": 0.09302325581395349,
      "grad_norm": 1.9715768923169905,
      "learning_rate": 1.4545454545454546e-05,
      "loss": 0.882,
      "step": 4
    },
    {
      "epoch": 0.11627906976744186,
      "grad_norm": 4.86558025165536,
      "learning_rate": 1.8181818181818182e-05,
      "loss": 0.9089,
      "step": 5
    },
    {
      "epoch": 0.13953488372093023,
      "grad_norm": 6.162689558609649,
      "learning_rate": 2.1818181818181818e-05,
      "loss": 0.9031,
      "step": 6
    },
    {
      "epoch": 0.16279069767441862,
      "grad_norm": 5.545748731779107,
      "learning_rate": 2.5454545454545457e-05,
      "loss": 0.9058,
      "step": 7
    },
    {
      "epoch": 0.18604651162790697,
      "grad_norm": 3.4785773417590877,
      "learning_rate": 2.9090909090909093e-05,
      "loss": 0.8423,
      "step": 8
    },
    {
      "epoch": 0.20930232558139536,
      "grad_norm": 2.16515163827823,
      "learning_rate": 3.272727272727273e-05,
      "loss": 0.8043,
      "step": 9
    },
    {
      "epoch": 0.23255813953488372,
      "grad_norm": 2.664754416131028,
      "learning_rate": 3.6363636363636364e-05,
      "loss": 0.7799,
      "step": 10
    },
    {
      "epoch": 0.2558139534883721,
      "grad_norm": 1.4902721906661134,
      "learning_rate": 4e-05,
      "loss": 0.7546,
      "step": 11
    },
    {
      "epoch": 0.27906976744186046,
      "grad_norm": 2.015625441498336,
      "learning_rate": 4.3636363636363636e-05,
      "loss": 0.7408,
      "step": 12
    },
    {
      "epoch": 0.3023255813953488,
      "grad_norm": 1.1406136624119514,
      "learning_rate": 4.727272727272728e-05,
      "loss": 0.7216,
      "step": 13
    },
    {
      "epoch": 0.32558139534883723,
      "grad_norm": 1.8195500867443861,
      "learning_rate": 5.0909090909090914e-05,
      "loss": 0.724,
      "step": 14
    },
    {
      "epoch": 0.3488372093023256,
      "grad_norm": 1.595476578575812,
      "learning_rate": 5.4545454545454546e-05,
      "loss": 0.7084,
      "step": 15
    },
    {
      "epoch": 0.37209302325581395,
      "grad_norm": 1.9850762532377662,
      "learning_rate": 5.8181818181818185e-05,
      "loss": 0.6992,
      "step": 16
    },
    {
      "epoch": 0.3953488372093023,
      "grad_norm": 1.3156392269832544,
      "learning_rate": 6.181818181818182e-05,
      "loss": 0.6969,
      "step": 17
    },
    {
      "epoch": 0.4186046511627907,
      "grad_norm": 2.165349958578573,
      "learning_rate": 6.545454545454546e-05,
      "loss": 0.6819,
      "step": 18
    },
    {
      "epoch": 0.4418604651162791,
      "grad_norm": 1.860669140085515,
      "learning_rate": 6.90909090909091e-05,
      "loss": 0.6986,
      "step": 19
    },
    {
      "epoch": 0.46511627906976744,
      "grad_norm": 1.7217756323779048,
      "learning_rate": 7.272727272727273e-05,
      "loss": 0.6801,
      "step": 20
    },
    {
      "epoch": 0.4883720930232558,
      "grad_norm": 1.4025893684782413,
      "learning_rate": 7.636363636363637e-05,
      "loss": 0.6751,
      "step": 21
    },
    {
      "epoch": 0.5116279069767442,
      "grad_norm": 1.7449603840187862,
      "learning_rate": 8e-05,
      "loss": 0.673,
      "step": 22
    },
    {
      "epoch": 0.5348837209302325,
      "grad_norm": 2.482316452405395,
      "learning_rate": 7.999470085828896e-05,
      "loss": 0.6629,
      "step": 23
    },
    {
      "epoch": 0.5581395348837209,
      "grad_norm": 1.2207804383595053,
      "learning_rate": 7.997880483720098e-05,
      "loss": 0.6716,
      "step": 24
    },
    {
      "epoch": 0.5813953488372093,
      "grad_norm": 2.568593544907844,
      "learning_rate": 7.995231614849947e-05,
      "loss": 0.671,
      "step": 25
    },
    {
      "epoch": 0.6046511627906976,
      "grad_norm": 1.502461502116564,
      "learning_rate": 7.991524181055018e-05,
      "loss": 0.6608,
      "step": 26
    },
    {
      "epoch": 0.627906976744186,
      "grad_norm": 2.8212929800172915,
      "learning_rate": 7.986759164646167e-05,
      "loss": 0.673,
      "step": 27
    },
    {
      "epoch": 0.6511627906976745,
      "grad_norm": 2.5422318457006825,
      "learning_rate": 7.980937828148252e-05,
      "loss": 0.6916,
      "step": 28
    },
    {
      "epoch": 0.6744186046511628,
      "grad_norm": 1.6426727549059879,
      "learning_rate": 7.974061713965625e-05,
      "loss": 0.6487,
      "step": 29
    },
    {
      "epoch": 0.6976744186046512,
      "grad_norm": 1.8002811870915256,
      "learning_rate": 7.966132643973464e-05,
      "loss": 0.6429,
      "step": 30
    },
    {
      "epoch": 0.7209302325581395,
      "grad_norm": 1.4654311466330405,
      "learning_rate": 7.95715271903504e-05,
      "loss": 0.6398,
      "step": 31
    },
    {
      "epoch": 0.7441860465116279,
      "grad_norm": 1.329866741625803,
      "learning_rate": 7.947124318445097e-05,
      "loss": 0.6398,
      "step": 32
    },
    {
      "epoch": 0.7674418604651163,
      "grad_norm": 1.4731638464740446,
      "learning_rate": 7.936050099299425e-05,
      "loss": 0.6243,
      "step": 33
    },
    {
      "epoch": 0.7906976744186046,
      "grad_norm": 1.4894523263564694,
      "learning_rate": 7.923932995790858e-05,
      "loss": 0.6335,
      "step": 34
    },
    {
      "epoch": 0.813953488372093,
      "grad_norm": 1.173972016468102,
      "learning_rate": 7.910776218431821e-05,
      "loss": 0.6224,
      "step": 35
    },
    {
      "epoch": 0.8372093023255814,
      "grad_norm": 1.6637513739112912,
      "learning_rate": 7.896583253203703e-05,
      "loss": 0.6219,
      "step": 36
    },
    {
      "epoch": 0.8604651162790697,
      "grad_norm": 1.122406312089788,
      "learning_rate": 7.881357860633204e-05,
      "loss": 0.6232,
      "step": 37
    },
    {
      "epoch": 0.8837209302325582,
      "grad_norm": 1.4388855947402022,
      "learning_rate": 7.865104074795966e-05,
      "loss": 0.5977,
      "step": 38
    },
    {
      "epoch": 0.9069767441860465,
      "grad_norm": 1.4139997348656486,
      "learning_rate": 7.847826202247715e-05,
      "loss": 0.6193,
      "step": 39
    },
    {
      "epoch": 0.9302325581395349,
      "grad_norm": 1.4041482845121371,
      "learning_rate": 7.829528820883205e-05,
      "loss": 0.604,
      "step": 40
    },
    {
      "epoch": 0.9534883720930233,
      "grad_norm": 1.1795408353423182,
      "learning_rate": 7.810216778723275e-05,
      "loss": 0.5909,
      "step": 41
    },
    {
      "epoch": 0.9767441860465116,
      "grad_norm": 1.0271878945240036,
      "learning_rate": 7.789895192630333e-05,
      "loss": 0.5927,
      "step": 42
    },
    {
      "epoch": 1.0,
      "grad_norm": 1.7552637527608157,
      "learning_rate": 7.768569446952603e-05,
      "loss": 0.5843,
      "step": 43
    },
    {
      "epoch": 1.0232558139534884,
      "grad_norm": 0.9212131281614243,
      "learning_rate": 7.746245192097509e-05,
      "loss": 0.5722,
      "step": 44
    },
    {
      "epoch": 1.0465116279069768,
      "grad_norm": 1.6756296752481579,
      "learning_rate": 7.722928343034551e-05,
      "loss": 0.5773,
      "step": 45
    },
    {
      "epoch": 1.069767441860465,
      "grad_norm": 1.3543712007327793,
      "learning_rate": 7.698625077728103e-05,
      "loss": 0.5672,
      "step": 46
    },
    {
      "epoch": 1.0930232558139534,
      "grad_norm": 1.7410081511463789,
      "learning_rate": 7.67334183550051e-05,
      "loss": 0.5674,
      "step": 47
    },
    {
      "epoch": 1.1162790697674418,
      "grad_norm": 1.0733210547826335,
      "learning_rate": 7.647085315325946e-05,
      "loss": 0.5671,
      "step": 48
    },
    {
      "epoch": 1.1395348837209303,
      "grad_norm": 2.252268000471556,
      "learning_rate": 7.619862474055472e-05,
      "loss": 0.5638,
      "step": 49
    },
    {
      "epoch": 1.1627906976744187,
      "grad_norm": 1.6138509974039352,
      "learning_rate": 7.591680524573773e-05,
      "loss": 0.5625,
      "step": 50
    },
    {
      "epoch": 1.1860465116279069,
      "grad_norm": 2.0237798746000193,
      "learning_rate": 7.56254693388805e-05,
      "loss": 0.5592,
      "step": 51
    },
    {
      "epoch": 1.2093023255813953,
      "grad_norm": 1.4811712482469028,
      "learning_rate": 7.532469421149578e-05,
      "loss": 0.5599,
      "step": 52
    },
    {
      "epoch": 1.2325581395348837,
      "grad_norm": 1.5279703879919382,
      "learning_rate": 7.501455955608479e-05,
      "loss": 0.5559,
      "step": 53
    },
    {
      "epoch": 1.255813953488372,
      "grad_norm": 1.2468524365094296,
      "learning_rate": 7.469514754502191e-05,
      "loss": 0.5549,
      "step": 54
    },
    {
      "epoch": 1.2790697674418605,
      "grad_norm": 1.2825998428965273,
      "learning_rate": 7.436654280878269e-05,
      "loss": 0.546,
      "step": 55
    },
    {
      "epoch": 1.302325581395349,
      "grad_norm": 0.9710691709326068,
      "learning_rate": 7.402883241352037e-05,
      "loss": 0.5492,
      "step": 56
    },
    {
      "epoch": 1.3255813953488373,
      "grad_norm": 1.4012298372305019,
      "learning_rate": 7.368210583799701e-05,
      "loss": 0.5481,
      "step": 57
    },
    {
      "epoch": 1.3488372093023255,
      "grad_norm": 1.2045446990194086,
      "learning_rate": 7.332645494987555e-05,
      "loss": 0.544,
      "step": 58
    },
    {
      "epoch": 1.372093023255814,
      "grad_norm": 1.5478512274120089,
      "learning_rate": 7.296197398137878e-05,
      "loss": 0.5399,
      "step": 59
    },
    {
      "epoch": 1.3953488372093024,
      "grad_norm": 1.0482231123455121,
      "learning_rate": 7.258875950432185e-05,
      "loss": 0.5457,
      "step": 60
    },
    {
      "epoch": 1.4186046511627908,
      "grad_norm": 1.350025832651234,
      "learning_rate": 7.220691040452491e-05,
      "loss": 0.533,
      "step": 61
    },
    {
      "epoch": 1.441860465116279,
      "grad_norm": 1.6379778302822026,
      "learning_rate": 7.181652785561254e-05,
      "loss": 0.5467,
      "step": 62
    },
    {
      "epoch": 1.4651162790697674,
      "grad_norm": 0.7697335505845303,
      "learning_rate": 7.141771529220716e-05,
      "loss": 0.536,
      "step": 63
    },
    {
      "epoch": 1.4883720930232558,
      "grad_norm": 1.019518212307104,
      "learning_rate": 7.101057838252324e-05,
      "loss": 0.5309,
      "step": 64
    },
    {
      "epoch": 1.5116279069767442,
      "grad_norm": 1.1970177274801466,
      "learning_rate": 7.05952250003698e-05,
      "loss": 0.535,
      "step": 65
    },
    {
      "epoch": 1.5348837209302326,
      "grad_norm": 1.0203220373942377,
      "learning_rate": 7.017176519656845e-05,
      "loss": 0.5378,
      "step": 66
    },
    {
      "epoch": 1.558139534883721,
      "grad_norm": 1.0958072090841686,
      "learning_rate": 6.974031116979466e-05,
      "loss": 0.5407,
      "step": 67
    },
    {
      "epoch": 1.5813953488372094,
      "grad_norm": 1.3397310047597761,
      "learning_rate": 6.93009772368499e-05,
      "loss": 0.5375,
      "step": 68
    },
    {
      "epoch": 1.6046511627906976,
      "grad_norm": 0.8782131687864965,
      "learning_rate": 6.885387980237262e-05,
      "loss": 0.5251,
      "step": 69
    },
    {
      "epoch": 1.627906976744186,
      "grad_norm": 1.4357177318604477,
      "learning_rate": 6.839913732799604e-05,
      "loss": 0.5274,
      "step": 70
    },
    {
      "epoch": 1.6511627906976745,
      "grad_norm": 1.215047188865305,
      "learning_rate": 6.793687030096085e-05,
      "loss": 0.5295,
      "step": 71
    },
    {
      "epoch": 1.6744186046511627,
      "grad_norm": 1.2710521920417013,
      "learning_rate": 6.746720120219126e-05,
      "loss": 0.5421,
      "step": 72
    },
    {
      "epoch": 1.697674418604651,
      "grad_norm": 0.7813066526911904,
      "learning_rate": 6.699025447384284e-05,
      "loss": 0.5312,
      "step": 73
    },
    {
      "epoch": 1.7209302325581395,
      "grad_norm": 1.2496179982961972,
      "learning_rate": 6.650615648633075e-05,
      "loss": 0.5246,
      "step": 74
    },
    {
      "epoch": 1.744186046511628,
      "grad_norm": 1.311820641764223,
      "learning_rate": 6.601503550484684e-05,
      "loss": 0.5258,
      "step": 75
    },
    {
      "epoch": 1.7674418604651163,
      "grad_norm": 1.0893635673008315,
      "learning_rate": 6.5517021655375e-05,
      "loss": 0.5271,
      "step": 76
    },
    {
      "epoch": 1.7906976744186047,
      "grad_norm": 0.7546199150125754,
      "learning_rate": 6.501224689021341e-05,
      "loss": 0.5185,
      "step": 77
    },
    {
      "epoch": 1.8139534883720931,
      "grad_norm": 1.1638269494753801,
      "learning_rate": 6.450084495301268e-05,
      "loss": 0.5169,
      "step": 78
    },
    {
      "epoch": 1.8372093023255816,
      "grad_norm": 1.1455758960103528,
      "learning_rate": 6.398295134333962e-05,
      "loss": 0.5197,
      "step": 79
    },
    {
      "epoch": 1.8604651162790697,
      "grad_norm": 0.9787313133874568,
      "learning_rate": 6.345870328077569e-05,
      "loss": 0.5189,
      "step": 80
    },
    {
      "epoch": 1.8837209302325582,
      "grad_norm": 1.2146657230893994,
      "learning_rate": 6.292823966855965e-05,
      "loss": 0.5184,
      "step": 81
    },
    {
      "epoch": 1.9069767441860463,
      "grad_norm": 0.8315295395061865,
      "learning_rate": 6.239170105678419e-05,
      "loss": 0.5151,
      "step": 82
    },
    {
      "epoch": 1.9302325581395348,
      "grad_norm": 0.43739239970652544,
      "learning_rate": 6.184922960515616e-05,
      "loss": 0.5088,
      "step": 83
    },
    {
      "epoch": 1.9534883720930232,
      "grad_norm": 0.8529923746757689,
      "learning_rate": 6.13009690453304e-05,
      "loss": 0.5212,
      "step": 84
    },
    {
      "epoch": 1.9767441860465116,
      "grad_norm": 0.7915534358179345,
      "learning_rate": 6.074706464282695e-05,
      "loss": 0.5081,
      "step": 85
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.838942938146,
      "learning_rate": 6.018766315854196e-05,
      "loss": 0.4843,
      "step": 86
    },
    {
      "epoch": 2.0232558139534884,
      "grad_norm": 1.0664271997362778,
      "learning_rate": 5.9622912809862384e-05,
      "loss": 0.4787,
      "step": 87
    },
    {
      "epoch": 2.046511627906977,
      "grad_norm": 1.4645846831681002,
      "learning_rate": 5.905296323139467e-05,
      "loss": 0.4779,
      "step": 88
    },
    {
      "epoch": 2.0697674418604652,
      "grad_norm": 1.0222504468766858,
      "learning_rate": 5.847796543531801e-05,
      "loss": 0.4708,
      "step": 89
    },
    {
      "epoch": 2.0930232558139537,
      "grad_norm": 1.446291157107826,
      "learning_rate": 5.789807177137269e-05,
      "loss": 0.4747,
      "step": 90
    },
    {
      "epoch": 2.116279069767442,
      "grad_norm": 0.894564054089064,
      "learning_rate": 5.731343588649382e-05,
      "loss": 0.4707,
      "step": 91
    },
    {
      "epoch": 2.13953488372093,
      "grad_norm": 1.1812623112632408,
      "learning_rate": 5.6724212684101567e-05,
      "loss": 0.4724,
      "step": 92
    },
    {
      "epoch": 2.1627906976744184,
      "grad_norm": 1.1141414953554294,
      "learning_rate": 5.613055828305838e-05,
      "loss": 0.4659,
      "step": 93
    },
    {
      "epoch": 2.186046511627907,
      "grad_norm": 0.7770582052783432,
      "learning_rate": 5.553262997630419e-05,
      "loss": 0.4592,
      "step": 94
    },
    {
      "epoch": 2.2093023255813953,
      "grad_norm": 1.0201596899389407,
      "learning_rate": 5.493058618918051e-05,
      "loss": 0.464,
      "step": 95
    },
    {
      "epoch": 2.2325581395348837,
      "grad_norm": 0.99732039808284,
      "learning_rate": 5.432458643745456e-05,
      "loss": 0.4728,
      "step": 96
    },
    {
      "epoch": 2.255813953488372,
      "grad_norm": 0.9946765474703833,
      "learning_rate": 5.37147912850544e-05,
      "loss": 0.4719,
      "step": 97
    },
    {
      "epoch": 2.2790697674418605,
      "grad_norm": 0.67376712726853,
      "learning_rate": 5.31013623015264e-05,
      "loss": 0.4635,
      "step": 98
    },
    {
      "epoch": 2.302325581395349,
      "grad_norm": 0.5331380825906256,
      "learning_rate": 5.2484462019226216e-05,
      "loss": 0.4636,
      "step": 99
    },
    {
      "epoch": 2.3255813953488373,
      "grad_norm": 0.8114012150052442,
      "learning_rate": 5.1864253890254744e-05,
      "loss": 0.4669,
      "step": 100
    },
    {
      "epoch": 2.3488372093023258,
      "grad_norm": 0.7817511193460437,
      "learning_rate": 5.1240902243150245e-05,
      "loss": 0.4643,
      "step": 101
    },
    {
      "epoch": 2.3720930232558137,
      "grad_norm": 0.5742798063922762,
      "learning_rate": 5.0614572239348436e-05,
      "loss": 0.468,
      "step": 102
    },
    {
      "epoch": 2.395348837209302,
      "grad_norm": 0.7545379109683643,
      "learning_rate": 4.998542982942169e-05,
      "loss": 0.4646,
      "step": 103
    },
    {
      "epoch": 2.4186046511627906,
      "grad_norm": 0.6867300007616275,
      "learning_rate": 4.9353641709109366e-05,
      "loss": 0.466,
      "step": 104
    },
    {
      "epoch": 2.441860465116279,
      "grad_norm": 0.5994645048448306,
      "learning_rate": 4.87193752751505e-05,
      "loss": 0.4605,
      "step": 105
    },
    {
      "epoch": 2.4651162790697674,
      "grad_norm": 0.5648107826549134,
      "learning_rate": 4.808279858093088e-05,
      "loss": 0.4633,
      "step": 106
    },
    {
      "epoch": 2.488372093023256,
      "grad_norm": 0.41656800662182036,
      "learning_rate": 4.744408029195616e-05,
      "loss": 0.458,
      "step": 107
    },
    {
      "epoch": 2.511627906976744,
      "grad_norm": 0.560471313591134,
      "learning_rate": 4.680338964116267e-05,
      "loss": 0.4694,
      "step": 108
    },
    {
      "epoch": 2.5348837209302326,
      "grad_norm": 0.3737750854348531,
      "learning_rate": 4.616089638407798e-05,
      "loss": 0.4572,
      "step": 109
    },
    {
      "epoch": 2.558139534883721,
      "grad_norm": 0.4130815840573508,
      "learning_rate": 4.551677075384296e-05,
      "loss": 0.4642,
      "step": 110
    },
    {
      "epoch": 2.5813953488372094,
      "grad_norm": 0.45613348607122495,
      "learning_rate": 4.487118341610736e-05,
      "loss": 0.4558,
      "step": 111
    },
    {
      "epoch": 2.604651162790698,
      "grad_norm": 0.3913887000750076,
      "learning_rate": 4.422430542381064e-05,
      "loss": 0.4646,
      "step": 112
    },
    {
      "epoch": 2.6279069767441863,
      "grad_norm": 0.30374911049832554,
      "learning_rate": 4.3576308171860336e-05,
      "loss": 0.4624,
      "step": 113
    },
    {
      "epoch": 2.6511627906976747,
      "grad_norm": 0.3917025778134578,
      "learning_rate": 4.292736335171978e-05,
      "loss": 0.4566,
      "step": 114
    },
    {
      "epoch": 2.6744186046511627,
      "grad_norm": 0.3717043272468876,
      "learning_rate": 4.2277642905917205e-05,
      "loss": 0.4566,
      "step": 115
    },
    {
      "epoch": 2.697674418604651,
      "grad_norm": 0.30603425852066757,
      "learning_rate": 4.162731898248834e-05,
      "loss": 0.4599,
      "step": 116
    },
    {
      "epoch": 2.7209302325581395,
      "grad_norm": 0.3251855900379686,
      "learning_rate": 4.097656388936462e-05,
      "loss": 0.4623,
      "step": 117
    },
    {
      "epoch": 2.744186046511628,
      "grad_norm": 0.33442925355720576,
      "learning_rate": 4.032555004871892e-05,
      "loss": 0.4587,
      "step": 118
    },
    {
      "epoch": 2.7674418604651163,
      "grad_norm": 0.33761110069481104,
      "learning_rate": 3.9674449951281095e-05,
      "loss": 0.4576,
      "step": 119
    },
    {
      "epoch": 2.7906976744186047,
      "grad_norm": 0.26743941116764636,
      "learning_rate": 3.902343611063539e-05,
      "loss": 0.4556,
      "step": 120
    },
    {
      "epoch": 2.813953488372093,
      "grad_norm": 0.2128179360983353,
      "learning_rate": 3.837268101751166e-05,
      "loss": 0.4558,
      "step": 121
    },
    {
      "epoch": 2.8372093023255816,
      "grad_norm": 0.2896189154702043,
      "learning_rate": 3.7722357094082815e-05,
      "loss": 0.4618,
      "step": 122
    },
    {
      "epoch": 2.8604651162790695,
      "grad_norm": 0.21296934932760417,
      "learning_rate": 3.707263664828023e-05,
      "loss": 0.4591,
      "step": 123
    },
    {
      "epoch": 2.883720930232558,
      "grad_norm": 0.2791061519264039,
      "learning_rate": 3.6423691828139685e-05,
      "loss": 0.4524,
      "step": 124
    },
    {
      "epoch": 2.9069767441860463,
      "grad_norm": 0.2631221143960164,
      "learning_rate": 3.577569457618937e-05,
      "loss": 0.4606,
      "step": 125
    },
    {
      "epoch": 2.9302325581395348,
      "grad_norm": 0.20684807194617985,
      "learning_rate": 3.512881658389265e-05,
      "loss": 0.4578,
      "step": 126
    },
    {
      "epoch": 2.953488372093023,
      "grad_norm": 0.2697328366910901,
      "learning_rate": 3.448322924615703e-05,
      "loss": 0.4544,
      "step": 127
    },
    {
      "epoch": 2.9767441860465116,
      "grad_norm": 0.22910217651885814,
      "learning_rate": 3.383910361592204e-05,
      "loss": 0.458,
      "step": 128
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.23244831316179806,
      "learning_rate": 3.319661035883734e-05,
      "loss": 0.427,
      "step": 129
    },
    {
      "epoch": 3.0232558139534884,
      "grad_norm": 0.20986791210333858,
      "learning_rate": 3.2555919708043856e-05,
      "loss": 0.4232,
      "step": 130
    },
    {
      "epoch": 3.046511627906977,
      "grad_norm": 0.2570519418980638,
      "learning_rate": 3.1917201419069126e-05,
      "loss": 0.4149,
      "step": 131
    },
    {
      "epoch": 3.0697674418604652,
      "grad_norm": 0.24795179030963221,
      "learning_rate": 3.128062472484952e-05,
      "loss": 0.4193,
      "step": 132
    },
    {
      "epoch": 3.0930232558139537,
      "grad_norm": 0.2576172200398066,
      "learning_rate": 3.064635829089064e-05,
      "loss": 0.4113,
      "step": 133
    },
    {
      "epoch": 3.116279069767442,
      "grad_norm": 0.22687653576900754,
      "learning_rate": 3.0014570170578324e-05,
      "loss": 0.417,
      "step": 134
    },
    {
      "epoch": 3.13953488372093,
      "grad_norm": 0.26047808565271,
      "learning_rate": 2.938542776065158e-05,
      "loss": 0.4137,
      "step": 135
    },
    {
      "epoch": 3.1627906976744184,
      "grad_norm": 0.22319212273415287,
      "learning_rate": 2.875909775684977e-05,
      "loss": 0.4211,
      "step": 136
    },
    {
      "epoch": 3.186046511627907,
      "grad_norm": 0.22428976031004036,
      "learning_rate": 2.8135746109745273e-05,
      "loss": 0.4138,
      "step": 137
    },
    {
      "epoch": 3.2093023255813953,
      "grad_norm": 0.22247453156579614,
      "learning_rate": 2.751553798077379e-05,
      "loss": 0.4153,
      "step": 138
    },
    {
      "epoch": 3.2325581395348837,
      "grad_norm": 0.20007992804768981,
      "learning_rate": 2.6898637698473605e-05,
      "loss": 0.4135,
      "step": 139
    },
    {
      "epoch": 3.255813953488372,
      "grad_norm": 0.2451508001851349,
      "learning_rate": 2.6285208714945615e-05,
      "loss": 0.407,
      "step": 140
    },
    {
      "epoch": 3.2790697674418605,
      "grad_norm": 0.18902377319628208,
      "learning_rate": 2.5675413562545453e-05,
      "loss": 0.4044,
      "step": 141
    },
    {
      "epoch": 3.302325581395349,
      "grad_norm": 0.24194497338579668,
      "learning_rate": 2.5069413810819503e-05,
      "loss": 0.4073,
      "step": 142
    },
    {
      "epoch": 3.3255813953488373,
      "grad_norm": 0.1989749287423866,
      "learning_rate": 2.446737002369582e-05,
      "loss": 0.4169,
      "step": 143
    },
    {
      "epoch": 3.3488372093023258,
      "grad_norm": 0.2095582088688536,
      "learning_rate": 2.3869441716941627e-05,
      "loss": 0.4204,
      "step": 144
    },
    {
      "epoch": 3.3720930232558137,
      "grad_norm": 0.162353572736833,
      "learning_rate": 2.327578731589843e-05,
      "loss": 0.4065,
      "step": 145
    },
    {
      "epoch": 3.395348837209302,
      "grad_norm": 0.17279337899190159,
      "learning_rate": 2.2686564113506188e-05,
      "loss": 0.412,
      "step": 146
    },
    {
      "epoch": 3.4186046511627906,
      "grad_norm": 0.16979965281268794,
      "learning_rate": 2.210192822862733e-05,
      "loss": 0.4097,
      "step": 147
    },
    {
      "epoch": 3.441860465116279,
      "grad_norm": 0.1394513868982289,
      "learning_rate": 2.1522034564681997e-05,
      "loss": 0.4136,
      "step": 148
    },
    {
      "epoch": 3.4651162790697674,
      "grad_norm": 0.15491336187962532,
      "learning_rate": 2.0947036768605343e-05,
      "loss": 0.4138,
      "step": 149
    },
    {
      "epoch": 3.488372093023256,
      "grad_norm": 0.15331640718626022,
      "learning_rate": 2.0377087190137615e-05,
      "loss": 0.4121,
      "step": 150
    },
    {
      "epoch": 3.511627906976744,
      "grad_norm": 0.14333365303471923,
      "learning_rate": 1.9812336841458032e-05,
      "loss": 0.412,
      "step": 151
    },
    {
      "epoch": 3.5348837209302326,
      "grad_norm": 0.15342863178771485,
      "learning_rate": 1.925293535717306e-05,
      "loss": 0.4128,
      "step": 152
    },
    {
      "epoch": 3.558139534883721,
      "grad_norm": 0.1515728462605629,
      "learning_rate": 1.869903095466962e-05,
      "loss": 0.4092,
      "step": 153
    },
    {
      "epoch": 3.5813953488372094,
      "grad_norm": 0.16216900903578507,
      "learning_rate": 1.815077039484385e-05,
      "loss": 0.418,
      "step": 154
    },
    {
      "epoch": 3.604651162790698,
      "grad_norm": 0.14923319773401145,
      "learning_rate": 1.760829894321582e-05,
      "loss": 0.4124,
      "step": 155
    },
    {
      "epoch": 3.6279069767441863,
      "grad_norm": 0.1388357623921274,
      "learning_rate": 1.7071760331440352e-05,
      "loss": 0.4134,
      "step": 156
    },
    {
      "epoch": 3.6511627906976747,
      "grad_norm": 0.14463436978493738,
      "learning_rate": 1.6541296719224315e-05,
      "loss": 0.4152,
      "step": 157
    },
    {
      "epoch": 3.6744186046511627,
      "grad_norm": 0.13948667157444467,
      "learning_rate": 1.6017048656660392e-05,
      "loss": 0.4127,
      "step": 158
    },
    {
      "epoch": 3.697674418604651,
      "grad_norm": 0.12962305417153203,
      "learning_rate": 1.5499155046987343e-05,
      "loss": 0.4102,
      "step": 159
    },
    {
      "epoch": 3.7209302325581395,
      "grad_norm": 0.13488967836960994,
      "learning_rate": 1.49877531097866e-05,
      "loss": 0.412,
      "step": 160
    },
    {
      "epoch": 3.744186046511628,
      "grad_norm": 0.14161978366263686,
      "learning_rate": 1.4482978344624999e-05,
      "loss": 0.4112,
      "step": 161
    },
    {
      "epoch": 3.7674418604651163,
      "grad_norm": 0.12104463455571417,
      "learning_rate": 1.3984964495153178e-05,
      "loss": 0.4107,
      "step": 162
    },
    {
      "epoch": 3.7906976744186047,
      "grad_norm": 0.14778280010357478,
      "learning_rate": 1.349384351366926e-05,
      "loss": 0.4103,
      "step": 163
    },
    {
      "epoch": 3.813953488372093,
      "grad_norm": 0.13974610415305352,
      "learning_rate": 1.3009745526157165e-05,
      "loss": 0.4089,
      "step": 164
    },
    {
      "epoch": 3.8372093023255816,
      "grad_norm": 0.12567459622985308,
      "learning_rate": 1.2532798797808767e-05,
      "loss": 0.4082,
      "step": 165
    },
    {
      "epoch": 3.8604651162790695,
      "grad_norm": 0.13111313669138286,
      "learning_rate": 1.2063129699039169e-05,
      "loss": 0.4072,
      "step": 166
    },
    {
      "epoch": 3.883720930232558,
      "grad_norm": 0.11859583351508307,
      "learning_rate": 1.1600862672003964e-05,
      "loss": 0.408,
      "step": 167
    },
    {
      "epoch": 3.9069767441860463,
      "grad_norm": 0.12426128460062202,
      "learning_rate": 1.1146120197627375e-05,
      "loss": 0.41,
      "step": 168
    },
    {
      "epoch": 3.9302325581395348,
      "grad_norm": 0.12183542106771499,
      "learning_rate": 1.0699022763150104e-05,
      "loss": 0.4115,
      "step": 169
    },
    {
      "epoch": 3.953488372093023,
      "grad_norm": 0.1104533940594946,
      "learning_rate": 1.0259688830205348e-05,
      "loss": 0.4055,
      "step": 170
    },
    {
      "epoch": 3.9767441860465116,
      "grad_norm": 0.11121249489433677,
      "learning_rate": 9.828234803431559e-06,
      "loss": 0.411,
      "step": 171
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.15736411698947347,
      "learning_rate": 9.404774999630204e-06,
      "loss": 0.3931,
      "step": 172
    },
    {
      "epoch": 4.023255813953488,
      "grad_norm": 0.14076644290125107,
      "learning_rate": 8.989421617476766e-06,
      "loss": 0.3903,
      "step": 173
    },
    {
      "epoch": 4.046511627906977,
      "grad_norm": 0.11712681895575112,
      "learning_rate": 8.582284707792845e-06,
      "loss": 0.387,
      "step": 174
    },
    {
      "epoch": 4.069767441860465,
      "grad_norm": 0.12595405435756382,
      "learning_rate": 8.183472144387456e-06,
      "loss": 0.3866,
      "step": 175
    },
    {
      "epoch": 4.093023255813954,
      "grad_norm": 0.13565427155194576,
      "learning_rate": 7.793089595475094e-06,
      "loss": 0.3913,
      "step": 176
    },
    {
      "epoch": 4.116279069767442,
      "grad_norm": 0.1412644950291927,
      "learning_rate": 7.411240495678158e-06,
      "loss": 0.3876,
      "step": 177
    },
    {
      "epoch": 4.1395348837209305,
      "grad_norm": 0.13949397861852225,
      "learning_rate": 7.038026018621234e-06,
      "loss": 0.3839,
      "step": 178
    },
    {
      "epoch": 4.162790697674419,
      "grad_norm": 0.12339126854002234,
      "learning_rate": 6.673545050124466e-06,
      "loss": 0.3894,
      "step": 179
    },
    {
      "epoch": 4.186046511627907,
      "grad_norm": 0.12932008059607125,
      "learning_rate": 6.317894162003004e-06,
      "loss": 0.3863,
      "step": 180
    },
    {
      "epoch": 4.209302325581396,
      "grad_norm": 0.12611084979552667,
      "learning_rate": 5.971167586479638e-06,
      "loss": 0.3878,
      "step": 181
    },
    {
      "epoch": 4.232558139534884,
      "grad_norm": 0.12145402717998689,
      "learning_rate": 5.633457191217311e-06,
      "loss": 0.3806,
      "step": 182
    },
    {
      "epoch": 4.2558139534883725,
      "grad_norm": 0.12362886923322695,
      "learning_rate": 5.3048524549781066e-06,
      "loss": 0.3833,
      "step": 183
    },
    {
      "epoch": 4.27906976744186,
      "grad_norm": 0.1114983032825903,
      "learning_rate": 4.985440443915223e-06,
      "loss": 0.3873,
      "step": 184
    },
    {
      "epoch": 4.3023255813953485,
      "grad_norm": 0.11722293815194844,
      "learning_rate": 4.6753057885042186e-06,
      "loss": 0.3827,
      "step": 185
    },
    {
      "epoch": 4.325581395348837,
      "grad_norm": 0.1250948440370564,
      "learning_rate": 4.374530661119512e-06,
      "loss": 0.3855,
      "step": 186
    },
    {
      "epoch": 4.348837209302325,
      "grad_norm": 0.1106181593146437,
      "learning_rate": 4.0831947542622655e-06,
      "loss": 0.3855,
      "step": 187
    },
    {
      "epoch": 4.372093023255814,
      "grad_norm": 0.10428284018843,
      "learning_rate": 3.8013752594452834e-06,
      "loss": 0.3862,
      "step": 188
    },
    {
      "epoch": 4.395348837209302,
      "grad_norm": 0.1049647850048849,
      "learning_rate": 3.529146846740559e-06,
      "loss": 0.3853,
      "step": 189
    },
    {
      "epoch": 4.4186046511627906,
      "grad_norm": 0.10140234914167741,
      "learning_rate": 3.266581644994915e-06,
      "loss": 0.3832,
      "step": 190
    },
    {
      "epoch": 4.441860465116279,
      "grad_norm": 0.1084178720824487,
      "learning_rate": 3.0137492227189803e-06,
      "loss": 0.3874,
      "step": 191
    },
    {
      "epoch": 4.465116279069767,
      "grad_norm": 0.1035454061342504,
      "learning_rate": 2.7707165696545035e-06,
      "loss": 0.3846,
      "step": 192
    },
    {
      "epoch": 4.488372093023256,
      "grad_norm": 0.10465143616293164,
      "learning_rate": 2.537548079024923e-06,
      "loss": 0.3875,
      "step": 193
    },
    {
      "epoch": 4.511627906976744,
      "grad_norm": 0.10287343108965315,
      "learning_rate": 2.3143055304739725e-06,
      "loss": 0.3878,
      "step": 194
    },
    {
      "epoch": 4.534883720930233,
      "grad_norm": 0.09899078580945446,
      "learning_rate": 2.1010480736966788e-06,
      "loss": 0.3844,
      "step": 195
    },
    {
      "epoch": 4.558139534883721,
      "grad_norm": 0.0947795487504229,
      "learning_rate": 1.8978322127672522e-06,
      "loss": 0.3864,
      "step": 196
    },
    {
      "epoch": 4.5813953488372094,
      "grad_norm": 0.09351592903837026,
      "learning_rate": 1.7047117911679567e-06,
      "loss": 0.3764,
      "step": 197
    },
    {
      "epoch": 4.604651162790698,
      "grad_norm": 0.09075701229280372,
      "learning_rate": 1.5217379775228503e-06,
      "loss": 0.3804,
      "step": 198
    },
    {
      "epoch": 4.627906976744186,
      "grad_norm": 0.09563361674368318,
      "learning_rate": 1.3489592520403405e-06,
      "loss": 0.3852,
      "step": 199
    },
    {
      "epoch": 4.651162790697675,
      "grad_norm": 0.09593704632301235,
      "learning_rate": 1.186421393667967e-06,
      "loss": 0.3802,
      "step": 200
    },
    {
      "epoch": 4.674418604651163,
      "grad_norm": 0.09257390445253719,
      "learning_rate": 1.034167467962983e-06,
      "loss": 0.3867,
      "step": 201
    },
    {
      "epoch": 4.6976744186046515,
      "grad_norm": 0.09289347063379441,
      "learning_rate": 8.922378156817957e-07,
      "loss": 0.3829,
      "step": 202
    },
    {
      "epoch": 4.720930232558139,
      "grad_norm": 0.09257169253912409,
      "learning_rate": 7.606700420914381e-07,
      "loss": 0.3821,
      "step": 203
    },
    {
      "epoch": 4.7441860465116275,
      "grad_norm": 0.09310627098479367,
      "learning_rate": 6.39499007005746e-07,
      "loss": 0.3903,
      "step": 204
    },
    {
      "epoch": 4.767441860465116,
      "grad_norm": 0.08944939258724487,
      "learning_rate": 5.287568155490341e-07,
      "loss": 0.3885,
      "step": 205
    },
    {
      "epoch": 4.790697674418604,
      "grad_norm": 0.09592002472628403,
      "learning_rate": 4.284728096496027e-07,
      "loss": 0.3875,
      "step": 206
    },
    {
      "epoch": 4.813953488372093,
      "grad_norm": 0.09180161205563094,
      "learning_rate": 3.3867356026537457e-07,
      "loss": 0.3855,
      "step": 207
    },
    {
      "epoch": 4.837209302325581,
      "grad_norm": 0.08528621258416194,
      "learning_rate": 2.593828603437487e-07,
      "loss": 0.3846,
      "step": 208
    },
    {
      "epoch": 4.8604651162790695,
      "grad_norm": 0.08815943836016799,
      "learning_rate": 1.906217185174919e-07,
      "loss": 0.3839,
      "step": 209
    },
    {
      "epoch": 4.883720930232558,
      "grad_norm": 0.08604030139119473,
      "learning_rate": 1.3240835353834247e-07,
      "loss": 0.3892,
      "step": 210
    },
    {
      "epoch": 4.906976744186046,
      "grad_norm": 0.0865890927390752,
      "learning_rate": 8.475818944982728e-08,
      "loss": 0.3835,
      "step": 211
    },
    {
      "epoch": 4.930232558139535,
      "grad_norm": 0.09058754572938608,
      "learning_rate": 4.7683851500548437e-08,
      "loss": 0.3788,
      "step": 212
    },
    {
      "epoch": 4.953488372093023,
      "grad_norm": 0.08787858276738178,
      "learning_rate": 2.1195162799032554e-08,
      "loss": 0.3903,
      "step": 213
    },
    {
      "epoch": 4.976744186046512,
      "grad_norm": 0.08813542305445697,
      "learning_rate": 5.299141711043732e-09,
      "loss": 0.3865,
      "step": 214
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.09709093222975328,
      "learning_rate": 0.0,
      "loss": 0.379,
      "step": 215
    },
    {
      "epoch": 5.0,
      "step": 215,
      "total_flos": 3606870585507840.0,
      "train_loss": 0.5023871626964835,
      "train_runtime": 3261.3525,
      "train_samples_per_second": 33.201,
      "train_steps_per_second": 0.066
    }
  ],
  "logging_steps": 1,
  "max_steps": 215,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "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": 3606870585507840.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}