{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 135,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.022222222222222223,
      "grad_norm": 0.013690393716568078,
      "learning_rate": 5.7142857142857145e-06,
      "loss": 0.2019,
      "step": 1
    },
    {
      "epoch": 0.044444444444444446,
      "grad_norm": 0.012251154456255115,
      "learning_rate": 1.1428571428571429e-05,
      "loss": 0.1936,
      "step": 2
    },
    {
      "epoch": 0.06666666666666667,
      "grad_norm": 0.012624664264050438,
      "learning_rate": 1.7142857142857142e-05,
      "loss": 0.1939,
      "step": 3
    },
    {
      "epoch": 0.08888888888888889,
      "grad_norm": 0.012082258304141572,
      "learning_rate": 2.2857142857142858e-05,
      "loss": 0.1904,
      "step": 4
    },
    {
      "epoch": 0.1111111111111111,
      "grad_norm": 0.01402134157530267,
      "learning_rate": 2.8571428571428574e-05,
      "loss": 0.2003,
      "step": 5
    },
    {
      "epoch": 0.13333333333333333,
      "grad_norm": 0.01473425246358169,
      "learning_rate": 3.4285714285714284e-05,
      "loss": 0.1982,
      "step": 6
    },
    {
      "epoch": 0.15555555555555556,
      "grad_norm": 0.012501265335035136,
      "learning_rate": 4e-05,
      "loss": 0.1842,
      "step": 7
    },
    {
      "epoch": 0.17777777777777778,
      "grad_norm": 0.0204995837768438,
      "learning_rate": 4.5714285714285716e-05,
      "loss": 0.2085,
      "step": 8
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.014248525223858364,
      "learning_rate": 5.1428571428571436e-05,
      "loss": 0.1861,
      "step": 9
    },
    {
      "epoch": 0.2222222222222222,
      "grad_norm": 0.013991639397531296,
      "learning_rate": 5.714285714285715e-05,
      "loss": 0.1825,
      "step": 10
    },
    {
      "epoch": 0.24444444444444444,
      "grad_norm": 0.015355992445776189,
      "learning_rate": 6.285714285714286e-05,
      "loss": 0.1829,
      "step": 11
    },
    {
      "epoch": 0.26666666666666666,
      "grad_norm": 0.030211594933866753,
      "learning_rate": 6.857142857142857e-05,
      "loss": 0.1758,
      "step": 12
    },
    {
      "epoch": 0.28888888888888886,
      "grad_norm": 0.03300097397844424,
      "learning_rate": 7.42857142857143e-05,
      "loss": 0.1664,
      "step": 13
    },
    {
      "epoch": 0.3111111111111111,
      "grad_norm": 0.033769247278801144,
      "learning_rate": 8e-05,
      "loss": 0.1682,
      "step": 14
    },
    {
      "epoch": 0.3333333333333333,
      "grad_norm": 0.022245093149131256,
      "learning_rate": 7.998651861214051e-05,
      "loss": 0.1639,
      "step": 15
    },
    {
      "epoch": 0.35555555555555557,
      "grad_norm": 0.014555549539416571,
      "learning_rate": 7.994608353595293e-05,
      "loss": 0.1463,
      "step": 16
    },
    {
      "epoch": 0.37777777777777777,
      "grad_norm": 0.013536651810954663,
      "learning_rate": 7.987872202748455e-05,
      "loss": 0.1566,
      "step": 17
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.013535530619433005,
      "learning_rate": 7.978447949306646e-05,
      "loss": 0.1463,
      "step": 18
    },
    {
      "epoch": 0.4222222222222222,
      "grad_norm": 0.011986947697387621,
      "learning_rate": 7.966341945870666e-05,
      "loss": 0.1298,
      "step": 19
    },
    {
      "epoch": 0.4444444444444444,
      "grad_norm": 0.014944887653987982,
      "learning_rate": 7.951562352726902e-05,
      "loss": 0.137,
      "step": 20
    },
    {
      "epoch": 0.4666666666666667,
      "grad_norm": 0.014659729835514146,
      "learning_rate": 7.934119132346731e-05,
      "loss": 0.1259,
      "step": 21
    },
    {
      "epoch": 0.4888888888888889,
      "grad_norm": 0.013998264706758854,
      "learning_rate": 7.914024042671127e-05,
      "loss": 0.1171,
      "step": 22
    },
    {
      "epoch": 0.5111111111111111,
      "grad_norm": 0.013062382810473252,
      "learning_rate": 7.89129062918499e-05,
      "loss": 0.1121,
      "step": 23
    },
    {
      "epoch": 0.5333333333333333,
      "grad_norm": 0.012713743862688602,
      "learning_rate": 7.865934215786548e-05,
      "loss": 0.1126,
      "step": 24
    },
    {
      "epoch": 0.5555555555555556,
      "grad_norm": 0.014541535328817478,
      "learning_rate": 7.837971894457991e-05,
      "loss": 0.1008,
      "step": 25
    },
    {
      "epoch": 0.5777777777777777,
      "grad_norm": 0.013428592114472589,
      "learning_rate": 7.807422513744279e-05,
      "loss": 0.0994,
      "step": 26
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.013722884558950167,
      "learning_rate": 7.774306666047929e-05,
      "loss": 0.0905,
      "step": 27
    },
    {
      "epoch": 0.6222222222222222,
      "grad_norm": 0.01321804661592223,
      "learning_rate": 7.738646673748293e-05,
      "loss": 0.0849,
      "step": 28
    },
    {
      "epoch": 0.6444444444444445,
      "grad_norm": 0.012761871679626948,
      "learning_rate": 7.700466574154735e-05,
      "loss": 0.0837,
      "step": 29
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 0.02177087386269294,
      "learning_rate": 7.659792103303814e-05,
      "loss": 0.0831,
      "step": 30
    },
    {
      "epoch": 0.6888888888888889,
      "grad_norm": 0.00920448676794041,
      "learning_rate": 7.616650678611401e-05,
      "loss": 0.0728,
      "step": 31
    },
    {
      "epoch": 0.7111111111111111,
      "grad_norm": 0.010160647208318184,
      "learning_rate": 7.571071380391455e-05,
      "loss": 0.0721,
      "step": 32
    },
    {
      "epoch": 0.7333333333333333,
      "grad_norm": 0.011503007840350446,
      "learning_rate": 7.523084932253858e-05,
      "loss": 0.0792,
      "step": 33
    },
    {
      "epoch": 0.7555555555555555,
      "grad_norm": 0.009838508262674604,
      "learning_rate": 7.472723680394576e-05,
      "loss": 0.0684,
      "step": 34
    },
    {
      "epoch": 0.7777777777777778,
      "grad_norm": 0.011783911120327735,
      "learning_rate": 7.420021571792081e-05,
      "loss": 0.0761,
      "step": 35
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.010170301881669928,
      "learning_rate": 7.365014131324725e-05,
      "loss": 0.0638,
      "step": 36
    },
    {
      "epoch": 0.8222222222222222,
      "grad_norm": 0.011787395542951468,
      "learning_rate": 7.307738437824515e-05,
      "loss": 0.0648,
      "step": 37
    },
    {
      "epoch": 0.8444444444444444,
      "grad_norm": 0.01201727784549968,
      "learning_rate": 7.248233099083401e-05,
      "loss": 0.0589,
      "step": 38
    },
    {
      "epoch": 0.8666666666666667,
      "grad_norm": 0.013728223860114455,
      "learning_rate": 7.186538225828944e-05,
      "loss": 0.064,
      "step": 39
    },
    {
      "epoch": 0.8888888888888888,
      "grad_norm": 0.010975012831345344,
      "learning_rate": 7.122695404686912e-05,
      "loss": 0.06,
      "step": 40
    },
    {
      "epoch": 0.9111111111111111,
      "grad_norm": 0.015740657789464877,
      "learning_rate": 7.056747670148997e-05,
      "loss": 0.0519,
      "step": 41
    },
    {
      "epoch": 0.9333333333333333,
      "grad_norm": 0.014720379776587062,
      "learning_rate": 6.988739475564586e-05,
      "loss": 0.0567,
      "step": 42
    },
    {
      "epoch": 0.9555555555555556,
      "grad_norm": 0.01248483370084809,
      "learning_rate": 6.91871666317612e-05,
      "loss": 0.0497,
      "step": 43
    },
    {
      "epoch": 0.9777777777777777,
      "grad_norm": 0.014667442497100976,
      "learning_rate": 6.846726433218241e-05,
      "loss": 0.0476,
      "step": 44
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.01994876001821449,
      "learning_rate": 6.772817312101557e-05,
      "loss": 0.0507,
      "step": 45
    },
    {
      "epoch": 1.0222222222222221,
      "grad_norm": 0.013452054799340921,
      "learning_rate": 6.697039119702473e-05,
      "loss": 0.0529,
      "step": 46
    },
    {
      "epoch": 1.0444444444444445,
      "grad_norm": 0.02140708516746142,
      "learning_rate": 6.619442935781141e-05,
      "loss": 0.048,
      "step": 47
    },
    {
      "epoch": 1.0666666666666667,
      "grad_norm": 0.01957288087879906,
      "learning_rate": 6.540081065550154e-05,
      "loss": 0.0444,
      "step": 48
    },
    {
      "epoch": 1.0888888888888888,
      "grad_norm": 0.012707459777905543,
      "learning_rate": 6.459007004417204e-05,
      "loss": 0.0397,
      "step": 49
    },
    {
      "epoch": 1.1111111111111112,
      "grad_norm": 0.014646209549804124,
      "learning_rate": 6.376275401925466e-05,
      "loss": 0.0388,
      "step": 50
    },
    {
      "epoch": 1.1333333333333333,
      "grad_norm": 0.022176133553592395,
      "learning_rate": 6.29194202491601e-05,
      "loss": 0.0461,
      "step": 51
    },
    {
      "epoch": 1.1555555555555554,
      "grad_norm": 0.016607732134554864,
      "learning_rate": 6.206063719937084e-05,
      "loss": 0.0422,
      "step": 52
    },
    {
      "epoch": 1.1777777777777778,
      "grad_norm": 0.01842741899149879,
      "learning_rate": 6.118698374925596e-05,
      "loss": 0.0472,
      "step": 53
    },
    {
      "epoch": 1.2,
      "grad_norm": 0.012723315232781958,
      "learning_rate": 6.029904880186623e-05,
      "loss": 0.0381,
      "step": 54
    },
    {
      "epoch": 1.2222222222222223,
      "grad_norm": 0.01976204440639505,
      "learning_rate": 5.9397430886972675e-05,
      "loss": 0.039,
      "step": 55
    },
    {
      "epoch": 1.2444444444444445,
      "grad_norm": 0.0150167554079988,
      "learning_rate": 5.848273775761585e-05,
      "loss": 0.0386,
      "step": 56
    },
    {
      "epoch": 1.2666666666666666,
      "grad_norm": 0.009651333070712512,
      "learning_rate": 5.755558598043824e-05,
      "loss": 0.034,
      "step": 57
    },
    {
      "epoch": 1.2888888888888888,
      "grad_norm": 0.024348992618410538,
      "learning_rate": 5.661660052007547e-05,
      "loss": 0.0358,
      "step": 58
    },
    {
      "epoch": 1.3111111111111111,
      "grad_norm": 0.013815164385718691,
      "learning_rate": 5.566641431788682e-05,
      "loss": 0.0462,
      "step": 59
    },
    {
      "epoch": 1.3333333333333333,
      "grad_norm": 0.012492750529238414,
      "learning_rate": 5.4705667865308805e-05,
      "loss": 0.0458,
      "step": 60
    },
    {
      "epoch": 1.3555555555555556,
      "grad_norm": 0.010908584063330298,
      "learning_rate": 5.373500877211955e-05,
      "loss": 0.0354,
      "step": 61
    },
    {
      "epoch": 1.3777777777777778,
      "grad_norm": 0.010171826825238057,
      "learning_rate": 5.275509132990476e-05,
      "loss": 0.0344,
      "step": 62
    },
    {
      "epoch": 1.4,
      "grad_norm": 0.006559566441352372,
      "learning_rate": 5.176657607101986e-05,
      "loss": 0.0371,
      "step": 63
    },
    {
      "epoch": 1.4222222222222223,
      "grad_norm": 0.00838882340830255,
      "learning_rate": 5.0770129323345385e-05,
      "loss": 0.0354,
      "step": 64
    },
    {
      "epoch": 1.4444444444444444,
      "grad_norm": 0.013397952249981533,
      "learning_rate": 4.976642276113566e-05,
      "loss": 0.0331,
      "step": 65
    },
    {
      "epoch": 1.4666666666666668,
      "grad_norm": 0.006834440077927805,
      "learning_rate": 4.87561329522638e-05,
      "loss": 0.0439,
      "step": 66
    },
    {
      "epoch": 1.488888888888889,
      "grad_norm": 0.008704274818279584,
      "learning_rate": 4.773994090216799e-05,
      "loss": 0.0363,
      "step": 67
    },
    {
      "epoch": 1.511111111111111,
      "grad_norm": 0.008507811927095526,
      "learning_rate": 4.6718531594806595e-05,
      "loss": 0.0387,
      "step": 68
    },
    {
      "epoch": 1.5333333333333332,
      "grad_norm": 0.006945336492276737,
      "learning_rate": 4.5692593530931416e-05,
      "loss": 0.038,
      "step": 69
    },
    {
      "epoch": 1.5555555555555556,
      "grad_norm": 0.0070963814666081945,
      "learning_rate": 4.466281826399037e-05,
      "loss": 0.0388,
      "step": 70
    },
    {
      "epoch": 1.5777777777777777,
      "grad_norm": 0.0067425945763156855,
      "learning_rate": 4.3629899933972576e-05,
      "loss": 0.041,
      "step": 71
    },
    {
      "epoch": 1.6,
      "grad_norm": 0.0054563079803089725,
      "learning_rate": 4.259453479950972e-05,
      "loss": 0.0361,
      "step": 72
    },
    {
      "epoch": 1.6222222222222222,
      "grad_norm": 0.009280396144582837,
      "learning_rate": 4.1557420768549515e-05,
      "loss": 0.0324,
      "step": 73
    },
    {
      "epoch": 1.6444444444444444,
      "grad_norm": 0.010878878607417408,
      "learning_rate": 4.051925692791724e-05,
      "loss": 0.0395,
      "step": 74
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 0.006689006309318525,
      "learning_rate": 3.9480743072082775e-05,
      "loss": 0.0368,
      "step": 75
    },
    {
      "epoch": 1.6888888888888889,
      "grad_norm": 0.008722934065043346,
      "learning_rate": 3.84425792314505e-05,
      "loss": 0.0352,
      "step": 76
    },
    {
      "epoch": 1.7111111111111112,
      "grad_norm": 0.006001571016351971,
      "learning_rate": 3.7405465200490277e-05,
      "loss": 0.0348,
      "step": 77
    },
    {
      "epoch": 1.7333333333333334,
      "grad_norm": 0.006263134409631759,
      "learning_rate": 3.637010006602743e-05,
      "loss": 0.0395,
      "step": 78
    },
    {
      "epoch": 1.7555555555555555,
      "grad_norm": 0.0056495683417859405,
      "learning_rate": 3.533718173600964e-05,
      "loss": 0.0318,
      "step": 79
    },
    {
      "epoch": 1.7777777777777777,
      "grad_norm": 0.006299687163765627,
      "learning_rate": 3.4307406469068604e-05,
      "loss": 0.0333,
      "step": 80
    },
    {
      "epoch": 1.8,
      "grad_norm": 0.007777720989947799,
      "learning_rate": 3.3281468405193405e-05,
      "loss": 0.0411,
      "step": 81
    },
    {
      "epoch": 1.8222222222222222,
      "grad_norm": 0.007253962148324405,
      "learning_rate": 3.2260059097832015e-05,
      "loss": 0.0399,
      "step": 82
    },
    {
      "epoch": 1.8444444444444446,
      "grad_norm": 0.007278848381773977,
      "learning_rate": 3.124386704773621e-05,
      "loss": 0.036,
      "step": 83
    },
    {
      "epoch": 1.8666666666666667,
      "grad_norm": 0.005846464796215279,
      "learning_rate": 3.0233577238864353e-05,
      "loss": 0.0452,
      "step": 84
    },
    {
      "epoch": 1.8888888888888888,
      "grad_norm": 0.006161431929118998,
      "learning_rate": 2.922987067665462e-05,
      "loss": 0.0377,
      "step": 85
    },
    {
      "epoch": 1.911111111111111,
      "grad_norm": 0.006454018505853082,
      "learning_rate": 2.8233423928980152e-05,
      "loss": 0.0441,
      "step": 86
    },
    {
      "epoch": 1.9333333333333333,
      "grad_norm": 0.009478193725520544,
      "learning_rate": 2.7244908670095262e-05,
      "loss": 0.0305,
      "step": 87
    },
    {
      "epoch": 1.9555555555555557,
      "grad_norm": 0.008826780614828972,
      "learning_rate": 2.6264991227880474e-05,
      "loss": 0.0397,
      "step": 88
    },
    {
      "epoch": 1.9777777777777779,
      "grad_norm": 0.0054777007788867965,
      "learning_rate": 2.5294332134691195e-05,
      "loss": 0.0376,
      "step": 89
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.005815916801456319,
      "learning_rate": 2.4333585682113192e-05,
      "loss": 0.0329,
      "step": 90
    },
    {
      "epoch": 2.022222222222222,
      "grad_norm": 0.005028733765259896,
      "learning_rate": 2.338339947992455e-05,
      "loss": 0.0404,
      "step": 91
    },
    {
      "epoch": 2.0444444444444443,
      "grad_norm": 0.00498376987246792,
      "learning_rate": 2.2444414019561775e-05,
      "loss": 0.0333,
      "step": 92
    },
    {
      "epoch": 2.066666666666667,
      "grad_norm": 0.005497192345444023,
      "learning_rate": 2.151726224238415e-05,
      "loss": 0.0388,
      "step": 93
    },
    {
      "epoch": 2.088888888888889,
      "grad_norm": 0.005313566098249345,
      "learning_rate": 2.0602569113027338e-05,
      "loss": 0.0372,
      "step": 94
    },
    {
      "epoch": 2.111111111111111,
      "grad_norm": 0.005174482058644561,
      "learning_rate": 1.970095119813378e-05,
      "loss": 0.0343,
      "step": 95
    },
    {
      "epoch": 2.1333333333333333,
      "grad_norm": 0.005802915599904005,
      "learning_rate": 1.8813016250744068e-05,
      "loss": 0.0445,
      "step": 96
    },
    {
      "epoch": 2.1555555555555554,
      "grad_norm": 0.005626340514047014,
      "learning_rate": 1.7939362800629178e-05,
      "loss": 0.033,
      "step": 97
    },
    {
      "epoch": 2.1777777777777776,
      "grad_norm": 0.006068854158275442,
      "learning_rate": 1.7080579750839925e-05,
      "loss": 0.039,
      "step": 98
    },
    {
      "epoch": 2.2,
      "grad_norm": 0.009173229404682466,
      "learning_rate": 1.6237245980745352e-05,
      "loss": 0.043,
      "step": 99
    },
    {
      "epoch": 2.2222222222222223,
      "grad_norm": 0.005635702088268802,
      "learning_rate": 1.540992995582797e-05,
      "loss": 0.0317,
      "step": 100
    },
    {
      "epoch": 2.2444444444444445,
      "grad_norm": 0.005725252790395505,
      "learning_rate": 1.4599189344498461e-05,
      "loss": 0.0431,
      "step": 101
    },
    {
      "epoch": 2.2666666666666666,
      "grad_norm": 0.00853576875952399,
      "learning_rate": 1.3805570642188602e-05,
      "loss": 0.0373,
      "step": 102
    },
    {
      "epoch": 2.2888888888888888,
      "grad_norm": 0.0060708375689423905,
      "learning_rate": 1.3029608802975284e-05,
      "loss": 0.0319,
      "step": 103
    },
    {
      "epoch": 2.311111111111111,
      "grad_norm": 0.005989456547387701,
      "learning_rate": 1.2271826878984454e-05,
      "loss": 0.0337,
      "step": 104
    },
    {
      "epoch": 2.3333333333333335,
      "grad_norm": 0.006188821285807428,
      "learning_rate": 1.1532735667817599e-05,
      "loss": 0.0342,
      "step": 105
    },
    {
      "epoch": 2.3555555555555556,
      "grad_norm": 0.006413950741567938,
      "learning_rate": 1.08128333682388e-05,
      "loss": 0.031,
      "step": 106
    },
    {
      "epoch": 2.3777777777777778,
      "grad_norm": 0.006225347087198183,
      "learning_rate": 1.0112605244354147e-05,
      "loss": 0.0309,
      "step": 107
    },
    {
      "epoch": 2.4,
      "grad_norm": 0.008751130751586886,
      "learning_rate": 9.432523298510041e-06,
      "loss": 0.0322,
      "step": 108
    },
    {
      "epoch": 2.422222222222222,
      "grad_norm": 0.0057667405967167075,
      "learning_rate": 8.773045953130878e-06,
      "loss": 0.0349,
      "step": 109
    },
    {
      "epoch": 2.4444444444444446,
      "grad_norm": 0.005861465816534307,
      "learning_rate": 8.13461774171056e-06,
      "loss": 0.0365,
      "step": 110
    },
    {
      "epoch": 2.466666666666667,
      "grad_norm": 0.0059333326199326126,
      "learning_rate": 7.517669009166009e-06,
      "loss": 0.0312,
      "step": 111
    },
    {
      "epoch": 2.488888888888889,
      "grad_norm": 0.006205164659404803,
      "learning_rate": 6.922615621754856e-06,
      "loss": 0.0363,
      "step": 112
    },
    {
      "epoch": 2.511111111111111,
      "grad_norm": 0.0059685274212395,
      "learning_rate": 6.349858686752748e-06,
      "loss": 0.0335,
      "step": 113
    },
    {
      "epoch": 2.533333333333333,
      "grad_norm": 0.006057414196908381,
      "learning_rate": 5.7997842820792e-06,
      "loss": 0.0365,
      "step": 114
    },
    {
      "epoch": 2.5555555555555554,
      "grad_norm": 0.00570488547636213,
      "learning_rate": 5.2727631960542495e-06,
      "loss": 0.0357,
      "step": 115
    },
    {
      "epoch": 2.5777777777777775,
      "grad_norm": 0.007899468870847144,
      "learning_rate": 4.769150677461434e-06,
      "loss": 0.0335,
      "step": 116
    },
    {
      "epoch": 2.6,
      "grad_norm": 0.006434287558194806,
      "learning_rate": 4.289286196085454e-06,
      "loss": 0.0388,
      "step": 117
    },
    {
      "epoch": 2.6222222222222222,
      "grad_norm": 0.006550099151422101,
      "learning_rate": 3.833493213885993e-06,
      "loss": 0.0349,
      "step": 118
    },
    {
      "epoch": 2.6444444444444444,
      "grad_norm": 0.00604738519186005,
      "learning_rate": 3.4020789669618747e-06,
      "loss": 0.0304,
      "step": 119
    },
    {
      "epoch": 2.6666666666666665,
      "grad_norm": 0.006884084873030972,
      "learning_rate": 2.9953342584526513e-06,
      "loss": 0.033,
      "step": 120
    },
    {
      "epoch": 2.688888888888889,
      "grad_norm": 0.0068910182456209034,
      "learning_rate": 2.613533262517072e-06,
      "loss": 0.0303,
      "step": 121
    },
    {
      "epoch": 2.7111111111111112,
      "grad_norm": 0.006383781371596263,
      "learning_rate": 2.256933339520724e-06,
      "loss": 0.0389,
      "step": 122
    },
    {
      "epoch": 2.7333333333333334,
      "grad_norm": 0.0061729370851098305,
      "learning_rate": 1.9257748625572193e-06,
      "loss": 0.0361,
      "step": 123
    },
    {
      "epoch": 2.7555555555555555,
      "grad_norm": 0.00550594243445581,
      "learning_rate": 1.6202810554201099e-06,
      "loss": 0.0322,
      "step": 124
    },
    {
      "epoch": 2.7777777777777777,
      "grad_norm": 0.006027669834456689,
      "learning_rate": 1.3406578421345206e-06,
      "loss": 0.0327,
      "step": 125
    },
    {
      "epoch": 2.8,
      "grad_norm": 0.005933443655863191,
      "learning_rate": 1.0870937081501088e-06,
      "loss": 0.0348,
      "step": 126
    },
    {
      "epoch": 2.822222222222222,
      "grad_norm": 0.005541514565731296,
      "learning_rate": 8.597595732887387e-07,
      "loss": 0.0291,
      "step": 127
    },
    {
      "epoch": 2.8444444444444446,
      "grad_norm": 0.007890681000633296,
      "learning_rate": 6.588086765327051e-07,
      "loss": 0.0392,
      "step": 128
    },
    {
      "epoch": 2.8666666666666667,
      "grad_norm": 0.006674624304626657,
      "learning_rate": 4.843764727309941e-07,
      "loss": 0.0313,
      "step": 129
    },
    {
      "epoch": 2.888888888888889,
      "grad_norm": 0.006175178690692166,
      "learning_rate": 3.365805412933476e-07,
      "loss": 0.0444,
      "step": 130
    },
    {
      "epoch": 2.911111111111111,
      "grad_norm": 0.005861225194458993,
      "learning_rate": 2.1552050693354643e-07,
      "loss": 0.0298,
      "step": 131
    },
    {
      "epoch": 2.9333333333333336,
      "grad_norm": 0.006560297564688616,
      "learning_rate": 1.212779725154656e-07,
      "loss": 0.0354,
      "step": 132
    },
    {
      "epoch": 2.9555555555555557,
      "grad_norm": 0.0059990590847810565,
      "learning_rate": 5.3916464047079064e-08,
      "loss": 0.0378,
      "step": 133
    },
    {
      "epoch": 2.977777777777778,
      "grad_norm": 0.006280301065210583,
      "learning_rate": 1.3481387859504147e-08,
      "loss": 0.0361,
      "step": 134
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.006155812375320129,
      "learning_rate": 0.0,
      "loss": 0.0339,
      "step": 135
    },
    {
      "epoch": 3.0,
      "step": 135,
      "total_flos": 75470737768448.0,
      "train_loss": 0.06517314999191849,
      "train_runtime": 812.7994,
      "train_samples_per_second": 1.329,
      "train_steps_per_second": 0.166
    }
  ],
  "logging_steps": 1,
  "max_steps": 135,
  "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": false,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 75470737768448.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}