{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.5336179295624333,
  "eval_steps": 500,
  "global_step": 500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "grad_norm": 0.10173338651657104,
      "learning_rate": 5.319148936170213e-07,
      "loss": 1.4985,
      "step": 1
    },
    {
      "epoch": 0.0,
      "grad_norm": 0.10725362598896027,
      "learning_rate": 1.0638297872340427e-06,
      "loss": 1.4086,
      "step": 2
    },
    {
      "epoch": 0.0,
      "grad_norm": 0.1187792494893074,
      "learning_rate": 1.5957446808510639e-06,
      "loss": 1.4706,
      "step": 3
    },
    {
      "epoch": 0.0,
      "grad_norm": 0.10263431817293167,
      "learning_rate": 2.1276595744680853e-06,
      "loss": 1.4262,
      "step": 4
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.10243014246225357,
      "learning_rate": 2.6595744680851065e-06,
      "loss": 1.4314,
      "step": 5
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.11368783563375473,
      "learning_rate": 3.1914893617021277e-06,
      "loss": 1.3997,
      "step": 6
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.11974532902240753,
      "learning_rate": 3.723404255319149e-06,
      "loss": 1.4356,
      "step": 7
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.10068795830011368,
      "learning_rate": 4.255319148936171e-06,
      "loss": 1.4006,
      "step": 8
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.11169079691171646,
      "learning_rate": 4.787234042553191e-06,
      "loss": 1.4415,
      "step": 9
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.10073692351579666,
      "learning_rate": 5.319148936170213e-06,
      "loss": 1.3752,
      "step": 10
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.1180361658334732,
      "learning_rate": 5.851063829787235e-06,
      "loss": 1.425,
      "step": 11
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.09457835555076599,
      "learning_rate": 6.3829787234042555e-06,
      "loss": 1.3732,
      "step": 12
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.10676431655883789,
      "learning_rate": 6.914893617021277e-06,
      "loss": 1.5031,
      "step": 13
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.13407254219055176,
      "learning_rate": 7.446808510638298e-06,
      "loss": 1.4144,
      "step": 14
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.10287946462631226,
      "learning_rate": 7.97872340425532e-06,
      "loss": 1.4369,
      "step": 15
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.10514385998249054,
      "learning_rate": 8.510638297872341e-06,
      "loss": 1.39,
      "step": 16
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.10802456736564636,
      "learning_rate": 9.042553191489362e-06,
      "loss": 1.4308,
      "step": 17
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.11478875577449799,
      "learning_rate": 9.574468085106383e-06,
      "loss": 1.4301,
      "step": 18
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.11263708025217056,
      "learning_rate": 1.0106382978723404e-05,
      "loss": 1.3662,
      "step": 19
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.1158255934715271,
      "learning_rate": 1.0638297872340426e-05,
      "loss": 1.4532,
      "step": 20
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.10566612333059311,
      "learning_rate": 1.1170212765957447e-05,
      "loss": 1.4037,
      "step": 21
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.11653206497430801,
      "learning_rate": 1.170212765957447e-05,
      "loss": 1.5074,
      "step": 22
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.1109832152724266,
      "learning_rate": 1.223404255319149e-05,
      "loss": 1.3657,
      "step": 23
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.11619631946086884,
      "learning_rate": 1.2765957446808511e-05,
      "loss": 1.4038,
      "step": 24
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.1255435049533844,
      "learning_rate": 1.3297872340425532e-05,
      "loss": 1.3373,
      "step": 25
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.11004912108182907,
      "learning_rate": 1.3829787234042554e-05,
      "loss": 1.4355,
      "step": 26
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.103106789290905,
      "learning_rate": 1.4361702127659577e-05,
      "loss": 1.4139,
      "step": 27
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.11609970778226852,
      "learning_rate": 1.4893617021276596e-05,
      "loss": 1.3901,
      "step": 28
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.10627258569002151,
      "learning_rate": 1.5425531914893617e-05,
      "loss": 1.4591,
      "step": 29
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.10686346888542175,
      "learning_rate": 1.595744680851064e-05,
      "loss": 1.3809,
      "step": 30
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.09971443563699722,
      "learning_rate": 1.6489361702127658e-05,
      "loss": 1.3737,
      "step": 31
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.09937144815921783,
      "learning_rate": 1.7021276595744682e-05,
      "loss": 1.3764,
      "step": 32
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.1083306223154068,
      "learning_rate": 1.7553191489361703e-05,
      "loss": 1.3584,
      "step": 33
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.10381340235471725,
      "learning_rate": 1.8085106382978724e-05,
      "loss": 1.4051,
      "step": 34
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.09269995987415314,
      "learning_rate": 1.8617021276595745e-05,
      "loss": 1.3287,
      "step": 35
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.10057959705591202,
      "learning_rate": 1.9148936170212766e-05,
      "loss": 1.4432,
      "step": 36
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.09498997777700424,
      "learning_rate": 1.968085106382979e-05,
      "loss": 1.5223,
      "step": 37
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.09475935995578766,
      "learning_rate": 2.0212765957446807e-05,
      "loss": 1.4697,
      "step": 38
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.09633958339691162,
      "learning_rate": 2.074468085106383e-05,
      "loss": 1.3877,
      "step": 39
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.09594912081956863,
      "learning_rate": 2.1276595744680852e-05,
      "loss": 1.4528,
      "step": 40
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.11214554309844971,
      "learning_rate": 2.1808510638297873e-05,
      "loss": 1.4098,
      "step": 41
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.19427792727947235,
      "learning_rate": 2.2340425531914894e-05,
      "loss": 1.3801,
      "step": 42
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.09848932921886444,
      "learning_rate": 2.2872340425531915e-05,
      "loss": 1.4141,
      "step": 43
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.09963933378458023,
      "learning_rate": 2.340425531914894e-05,
      "loss": 1.4553,
      "step": 44
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.09407307952642441,
      "learning_rate": 2.393617021276596e-05,
      "loss": 1.4168,
      "step": 45
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.10419266670942307,
      "learning_rate": 2.446808510638298e-05,
      "loss": 1.3287,
      "step": 46
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.09864490479230881,
      "learning_rate": 2.5e-05,
      "loss": 1.3317,
      "step": 47
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.11262151598930359,
      "learning_rate": 2.5531914893617022e-05,
      "loss": 1.3246,
      "step": 48
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.10121612995862961,
      "learning_rate": 2.6063829787234046e-05,
      "loss": 1.4149,
      "step": 49
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.0978160873055458,
      "learning_rate": 2.6595744680851064e-05,
      "loss": 1.3421,
      "step": 50
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.08920774608850479,
      "learning_rate": 2.7127659574468084e-05,
      "loss": 1.3861,
      "step": 51
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.09108138084411621,
      "learning_rate": 2.765957446808511e-05,
      "loss": 1.3964,
      "step": 52
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.1604197472333908,
      "learning_rate": 2.819148936170213e-05,
      "loss": 1.3571,
      "step": 53
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.10113083571195602,
      "learning_rate": 2.8723404255319154e-05,
      "loss": 1.3825,
      "step": 54
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.11050890386104584,
      "learning_rate": 2.925531914893617e-05,
      "loss": 1.3951,
      "step": 55
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.0961427092552185,
      "learning_rate": 2.9787234042553192e-05,
      "loss": 1.3434,
      "step": 56
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.1315707117319107,
      "learning_rate": 3.0319148936170216e-05,
      "loss": 1.3636,
      "step": 57
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.10595009475946426,
      "learning_rate": 3.085106382978723e-05,
      "loss": 1.4939,
      "step": 58
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.14528846740722656,
      "learning_rate": 3.1382978723404254e-05,
      "loss": 1.454,
      "step": 59
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.0951959565281868,
      "learning_rate": 3.191489361702128e-05,
      "loss": 1.4043,
      "step": 60
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.10170764476060867,
      "learning_rate": 3.2446808510638296e-05,
      "loss": 1.3428,
      "step": 61
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.1816103458404541,
      "learning_rate": 3.2978723404255317e-05,
      "loss": 1.3395,
      "step": 62
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.1011623665690422,
      "learning_rate": 3.3510638297872344e-05,
      "loss": 1.4016,
      "step": 63
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.09644798189401627,
      "learning_rate": 3.4042553191489365e-05,
      "loss": 1.3991,
      "step": 64
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.11246529221534729,
      "learning_rate": 3.4574468085106386e-05,
      "loss": 1.4112,
      "step": 65
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.0963212326169014,
      "learning_rate": 3.5106382978723407e-05,
      "loss": 1.3987,
      "step": 66
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.10353317856788635,
      "learning_rate": 3.563829787234043e-05,
      "loss": 1.4398,
      "step": 67
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.1050473302602768,
      "learning_rate": 3.617021276595745e-05,
      "loss": 1.4308,
      "step": 68
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.10414024442434311,
      "learning_rate": 3.670212765957447e-05,
      "loss": 1.3739,
      "step": 69
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.1070413663983345,
      "learning_rate": 3.723404255319149e-05,
      "loss": 1.2944,
      "step": 70
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.08770282566547394,
      "learning_rate": 3.776595744680852e-05,
      "loss": 1.2979,
      "step": 71
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.09537198394536972,
      "learning_rate": 3.829787234042553e-05,
      "loss": 1.3194,
      "step": 72
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.168814554810524,
      "learning_rate": 3.882978723404255e-05,
      "loss": 1.3255,
      "step": 73
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.10750961303710938,
      "learning_rate": 3.936170212765958e-05,
      "loss": 1.4541,
      "step": 74
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.09767494350671768,
      "learning_rate": 3.9893617021276594e-05,
      "loss": 1.3108,
      "step": 75
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.09643732011318207,
      "learning_rate": 4.0425531914893614e-05,
      "loss": 1.4451,
      "step": 76
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.10322108119726181,
      "learning_rate": 4.095744680851064e-05,
      "loss": 1.4286,
      "step": 77
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.10831727087497711,
      "learning_rate": 4.148936170212766e-05,
      "loss": 1.415,
      "step": 78
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.11692611873149872,
      "learning_rate": 4.2021276595744684e-05,
      "loss": 1.3944,
      "step": 79
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.10344197601079941,
      "learning_rate": 4.2553191489361704e-05,
      "loss": 1.4005,
      "step": 80
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.10011729598045349,
      "learning_rate": 4.3085106382978725e-05,
      "loss": 1.4402,
      "step": 81
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.22549670934677124,
      "learning_rate": 4.3617021276595746e-05,
      "loss": 1.3343,
      "step": 82
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.11193282157182693,
      "learning_rate": 4.414893617021277e-05,
      "loss": 1.4489,
      "step": 83
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.10785133391618729,
      "learning_rate": 4.468085106382979e-05,
      "loss": 1.402,
      "step": 84
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.10537637770175934,
      "learning_rate": 4.5212765957446815e-05,
      "loss": 1.3552,
      "step": 85
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.1029570996761322,
      "learning_rate": 4.574468085106383e-05,
      "loss": 1.3307,
      "step": 86
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.10009574890136719,
      "learning_rate": 4.627659574468085e-05,
      "loss": 1.3954,
      "step": 87
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.11556259542703629,
      "learning_rate": 4.680851063829788e-05,
      "loss": 1.3959,
      "step": 88
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.10029179602861404,
      "learning_rate": 4.734042553191489e-05,
      "loss": 1.3343,
      "step": 89
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.11009006202220917,
      "learning_rate": 4.787234042553192e-05,
      "loss": 1.5279,
      "step": 90
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.10678206384181976,
      "learning_rate": 4.840425531914894e-05,
      "loss": 1.341,
      "step": 91
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.1142154335975647,
      "learning_rate": 4.893617021276596e-05,
      "loss": 1.4263,
      "step": 92
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.09466603398323059,
      "learning_rate": 4.946808510638298e-05,
      "loss": 1.3956,
      "step": 93
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.1106942892074585,
      "learning_rate": 5e-05,
      "loss": 1.3708,
      "step": 94
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.12606613337993622,
      "learning_rate": 4.9999961062358615e-05,
      "loss": 1.3991,
      "step": 95
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.10292167216539383,
      "learning_rate": 4.999984424955572e-05,
      "loss": 1.3572,
      "step": 96
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.09828190505504608,
      "learning_rate": 4.9999649561955216e-05,
      "loss": 1.31,
      "step": 97
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.109276182949543,
      "learning_rate": 4.9999377000163536e-05,
      "loss": 1.3833,
      "step": 98
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.10516379773616791,
      "learning_rate": 4.999902656502973e-05,
      "loss": 1.3872,
      "step": 99
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.0997815951704979,
      "learning_rate": 4.99985982576454e-05,
      "loss": 1.4308,
      "step": 100
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.1041920930147171,
      "learning_rate": 4.9998092079344723e-05,
      "loss": 1.3832,
      "step": 101
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.09994788467884064,
      "learning_rate": 4.9997508031704456e-05,
      "loss": 1.3408,
      "step": 102
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.13418225944042206,
      "learning_rate": 4.9996846116543915e-05,
      "loss": 1.349,
      "step": 103
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.12500610947608948,
      "learning_rate": 4.999610633592496e-05,
      "loss": 1.308,
      "step": 104
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.15584518015384674,
      "learning_rate": 4.9995288692152044e-05,
      "loss": 1.4055,
      "step": 105
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.10717131942510605,
      "learning_rate": 4.999439318777212e-05,
      "loss": 1.3719,
      "step": 106
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.09883825480937958,
      "learning_rate": 4.999341982557468e-05,
      "loss": 1.3671,
      "step": 107
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.09858126938343048,
      "learning_rate": 4.9992368608591775e-05,
      "loss": 1.4032,
      "step": 108
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.09398377686738968,
      "learning_rate": 4.9991239540097965e-05,
      "loss": 1.2763,
      "step": 109
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.10111579298973083,
      "learning_rate": 4.9990032623610296e-05,
      "loss": 1.3361,
      "step": 110
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.10265881568193436,
      "learning_rate": 4.998874786288833e-05,
      "loss": 1.3745,
      "step": 111
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.10393088310956955,
      "learning_rate": 4.998738526193412e-05,
      "loss": 1.4737,
      "step": 112
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.10454674810171127,
      "learning_rate": 4.9985944824992166e-05,
      "loss": 1.3587,
      "step": 113
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.09539375454187393,
      "learning_rate": 4.9984426556549456e-05,
      "loss": 1.3285,
      "step": 114
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.10025190562009811,
      "learning_rate": 4.998283046133542e-05,
      "loss": 1.3573,
      "step": 115
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.3517231345176697,
      "learning_rate": 4.9981156544321906e-05,
      "loss": 1.3057,
      "step": 116
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.11298110336065292,
      "learning_rate": 4.9979404810723175e-05,
      "loss": 1.376,
      "step": 117
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.0932416245341301,
      "learning_rate": 4.997757526599591e-05,
      "loss": 1.3948,
      "step": 118
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.0937485620379448,
      "learning_rate": 4.997566791583916e-05,
      "loss": 1.2327,
      "step": 119
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.09703949838876724,
      "learning_rate": 4.997368276619435e-05,
      "loss": 1.4109,
      "step": 120
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.10851214081048965,
      "learning_rate": 4.997161982324523e-05,
      "loss": 1.3715,
      "step": 121
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.09600158035755157,
      "learning_rate": 4.9969479093417894e-05,
      "loss": 1.3202,
      "step": 122
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.09330842643976212,
      "learning_rate": 4.996726058338075e-05,
      "loss": 1.3079,
      "step": 123
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.12271397560834885,
      "learning_rate": 4.996496430004446e-05,
      "loss": 1.4488,
      "step": 124
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.09032147377729416,
      "learning_rate": 4.9962590250561996e-05,
      "loss": 1.3593,
      "step": 125
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.0970826968550682,
      "learning_rate": 4.9960138442328534e-05,
      "loss": 1.3624,
      "step": 126
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.09610634297132492,
      "learning_rate": 4.995760888298149e-05,
      "loss": 1.2691,
      "step": 127
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.10096735507249832,
      "learning_rate": 4.9955001580400475e-05,
      "loss": 1.4252,
      "step": 128
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.15890567004680634,
      "learning_rate": 4.9952316542707253e-05,
      "loss": 1.3557,
      "step": 129
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.11183907091617584,
      "learning_rate": 4.994955377826577e-05,
      "loss": 1.2537,
      "step": 130
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.09637223184108734,
      "learning_rate": 4.9946713295682034e-05,
      "loss": 1.2973,
      "step": 131
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.08964424580335617,
      "learning_rate": 4.994379510380421e-05,
      "loss": 1.3916,
      "step": 132
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.0993800014257431,
      "learning_rate": 4.9940799211722484e-05,
      "loss": 1.4148,
      "step": 133
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.09357187151908875,
      "learning_rate": 4.9937725628769094e-05,
      "loss": 1.2904,
      "step": 134
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.09439032524824142,
      "learning_rate": 4.9934574364518295e-05,
      "loss": 1.3445,
      "step": 135
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.10771452635526657,
      "learning_rate": 4.993134542878631e-05,
      "loss": 1.3755,
      "step": 136
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.09271609783172607,
      "learning_rate": 4.99280388316313e-05,
      "loss": 1.2766,
      "step": 137
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.1042882427573204,
      "learning_rate": 4.992465458335335e-05,
      "loss": 1.3967,
      "step": 138
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.10192878544330597,
      "learning_rate": 4.9921192694494454e-05,
      "loss": 1.3401,
      "step": 139
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.09331267327070236,
      "learning_rate": 4.9917653175838405e-05,
      "loss": 1.3251,
      "step": 140
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.09516970068216324,
      "learning_rate": 4.991403603841087e-05,
      "loss": 1.3637,
      "step": 141
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.09037785977125168,
      "learning_rate": 4.9910341293479265e-05,
      "loss": 1.417,
      "step": 142
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.10086451470851898,
      "learning_rate": 4.9906568952552754e-05,
      "loss": 1.4657,
      "step": 143
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.09110086411237717,
      "learning_rate": 4.990271902738223e-05,
      "loss": 1.3173,
      "step": 144
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.09085503220558167,
      "learning_rate": 4.989879152996025e-05,
      "loss": 1.3234,
      "step": 145
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.0876404270529747,
      "learning_rate": 4.989478647252101e-05,
      "loss": 1.3367,
      "step": 146
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.09222956746816635,
      "learning_rate": 4.9890703867540314e-05,
      "loss": 1.3408,
      "step": 147
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.09831474721431732,
      "learning_rate": 4.988654372773552e-05,
      "loss": 1.3904,
      "step": 148
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.09171093255281448,
      "learning_rate": 4.988230606606552e-05,
      "loss": 1.3178,
      "step": 149
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.11647266149520874,
      "learning_rate": 4.9877990895730666e-05,
      "loss": 1.3282,
      "step": 150
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.08659853786230087,
      "learning_rate": 4.9873598230172764e-05,
      "loss": 1.3531,
      "step": 151
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.09219101071357727,
      "learning_rate": 4.986912808307502e-05,
      "loss": 1.2975,
      "step": 152
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.0935453400015831,
      "learning_rate": 4.986458046836199e-05,
      "loss": 1.258,
      "step": 153
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.09370335936546326,
      "learning_rate": 4.985995540019955e-05,
      "loss": 1.4368,
      "step": 154
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.08816273510456085,
      "learning_rate": 4.9855252892994844e-05,
      "loss": 1.3168,
      "step": 155
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.10105711221694946,
      "learning_rate": 4.9850472961396215e-05,
      "loss": 1.2625,
      "step": 156
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.09414559602737427,
      "learning_rate": 4.984561562029323e-05,
      "loss": 1.34,
      "step": 157
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.09968777745962143,
      "learning_rate": 4.9840680884816536e-05,
      "loss": 1.4364,
      "step": 158
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.0899839997291565,
      "learning_rate": 4.983566877033791e-05,
      "loss": 1.2951,
      "step": 159
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.08910808712244034,
      "learning_rate": 4.9830579292470136e-05,
      "loss": 1.3533,
      "step": 160
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.09147724509239197,
      "learning_rate": 4.9825412467067e-05,
      "loss": 1.3941,
      "step": 161
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.09149125963449478,
      "learning_rate": 4.9820168310223215e-05,
      "loss": 1.3223,
      "step": 162
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.09467878937721252,
      "learning_rate": 4.981484683827439e-05,
      "loss": 1.2784,
      "step": 163
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.1013043224811554,
      "learning_rate": 4.9809448067796974e-05,
      "loss": 1.3415,
      "step": 164
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.10649001598358154,
      "learning_rate": 4.9803972015608205e-05,
      "loss": 1.3784,
      "step": 165
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.09393008053302765,
      "learning_rate": 4.9798418698766034e-05,
      "loss": 1.3084,
      "step": 166
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.09331285208463669,
      "learning_rate": 4.979278813456911e-05,
      "loss": 1.4026,
      "step": 167
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.10828982293605804,
      "learning_rate": 4.97870803405567e-05,
      "loss": 1.3845,
      "step": 168
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.09335343539714813,
      "learning_rate": 4.978129533450866e-05,
      "loss": 1.3743,
      "step": 169
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.09739939123392105,
      "learning_rate": 4.977543313444534e-05,
      "loss": 1.357,
      "step": 170
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.10757780820131302,
      "learning_rate": 4.976949375862756e-05,
      "loss": 1.4467,
      "step": 171
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.13449633121490479,
      "learning_rate": 4.976347722555654e-05,
      "loss": 1.3811,
      "step": 172
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.10207029432058334,
      "learning_rate": 4.975738355397386e-05,
      "loss": 1.3571,
      "step": 173
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.09364256262779236,
      "learning_rate": 4.975121276286136e-05,
      "loss": 1.298,
      "step": 174
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.10099958628416061,
      "learning_rate": 4.974496487144114e-05,
      "loss": 1.3605,
      "step": 175
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.08877628296613693,
      "learning_rate": 4.973863989917544e-05,
      "loss": 1.3566,
      "step": 176
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.09996306151151657,
      "learning_rate": 4.973223786576663e-05,
      "loss": 1.5077,
      "step": 177
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.08964492380619049,
      "learning_rate": 4.972575879115711e-05,
      "loss": 1.3556,
      "step": 178
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.10891906917095184,
      "learning_rate": 4.971920269552927e-05,
      "loss": 1.3015,
      "step": 179
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.09565063565969467,
      "learning_rate": 4.971256959930541e-05,
      "loss": 1.2955,
      "step": 180
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.09602455049753189,
      "learning_rate": 4.970585952314774e-05,
      "loss": 1.4334,
      "step": 181
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.09518374502658844,
      "learning_rate": 4.969907248795818e-05,
      "loss": 1.3537,
      "step": 182
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.09421328455209732,
      "learning_rate": 4.9692208514878444e-05,
      "loss": 1.4354,
      "step": 183
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.09471164643764496,
      "learning_rate": 4.9685267625289886e-05,
      "loss": 1.278,
      "step": 184
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.10142439603805542,
      "learning_rate": 4.967824984081344e-05,
      "loss": 1.4494,
      "step": 185
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.0980624407529831,
      "learning_rate": 4.9671155183309595e-05,
      "loss": 1.3545,
      "step": 186
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.09215401858091354,
      "learning_rate": 4.9663983674878296e-05,
      "loss": 1.2775,
      "step": 187
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.10461422055959702,
      "learning_rate": 4.965673533785886e-05,
      "loss": 1.3267,
      "step": 188
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.0982186496257782,
      "learning_rate": 4.9649410194829945e-05,
      "loss": 1.3111,
      "step": 189
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.09877320379018784,
      "learning_rate": 4.9642008268609454e-05,
      "loss": 1.3306,
      "step": 190
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.10117904841899872,
      "learning_rate": 4.9634529582254466e-05,
      "loss": 1.3162,
      "step": 191
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.10001292824745178,
      "learning_rate": 4.962697415906118e-05,
      "loss": 1.3379,
      "step": 192
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.09294575452804565,
      "learning_rate": 4.961934202256482e-05,
      "loss": 1.3189,
      "step": 193
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.09697602689266205,
      "learning_rate": 4.9611633196539584e-05,
      "loss": 1.3325,
      "step": 194
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.09828122705221176,
      "learning_rate": 4.960384770499855e-05,
      "loss": 1.4084,
      "step": 195
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.0948810800909996,
      "learning_rate": 4.95959855721936e-05,
      "loss": 1.2597,
      "step": 196
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.10147445648908615,
      "learning_rate": 4.958804682261539e-05,
      "loss": 1.2529,
      "step": 197
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.14694844186306,
      "learning_rate": 4.95800314809932e-05,
      "loss": 1.3571,
      "step": 198
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.21705757081508636,
      "learning_rate": 4.957193957229491e-05,
      "loss": 1.4377,
      "step": 199
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.12128579616546631,
      "learning_rate": 4.956377112172691e-05,
      "loss": 1.3725,
      "step": 200
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.09640059620141983,
      "learning_rate": 4.955552615473401e-05,
      "loss": 1.3913,
      "step": 201
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.19373305141925812,
      "learning_rate": 4.954720469699938e-05,
      "loss": 1.4382,
      "step": 202
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.0959777757525444,
      "learning_rate": 4.953880677444446e-05,
      "loss": 1.3134,
      "step": 203
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.10545915365219116,
      "learning_rate": 4.953033241322886e-05,
      "loss": 1.3788,
      "step": 204
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.1017264649271965,
      "learning_rate": 4.952178163975033e-05,
      "loss": 1.3932,
      "step": 205
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.1435861438512802,
      "learning_rate": 4.951315448064462e-05,
      "loss": 1.2377,
      "step": 206
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.10849795490503311,
      "learning_rate": 4.950445096278541e-05,
      "loss": 1.2753,
      "step": 207
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.09382300078868866,
      "learning_rate": 4.949567111328428e-05,
      "loss": 1.2816,
      "step": 208
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.09901299327611923,
      "learning_rate": 4.9486814959490544e-05,
      "loss": 1.3306,
      "step": 209
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.09280100464820862,
      "learning_rate": 4.947788252899124e-05,
      "loss": 1.2148,
      "step": 210
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.10457581281661987,
      "learning_rate": 4.946887384961097e-05,
      "loss": 1.4392,
      "step": 211
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.09088419377803802,
      "learning_rate": 4.9459788949411886e-05,
      "loss": 1.3393,
      "step": 212
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.09645824134349823,
      "learning_rate": 4.9450627856693545e-05,
      "loss": 1.4496,
      "step": 213
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.09751391410827637,
      "learning_rate": 4.9441390599992864e-05,
      "loss": 1.3765,
      "step": 214
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.09599944204092026,
      "learning_rate": 4.943207720808399e-05,
      "loss": 1.3751,
      "step": 215
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.16549742221832275,
      "learning_rate": 4.9422687709978254e-05,
      "loss": 1.3439,
      "step": 216
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.09610256552696228,
      "learning_rate": 4.941322213492405e-05,
      "loss": 1.2716,
      "step": 217
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.09371083229780197,
      "learning_rate": 4.940368051240675e-05,
      "loss": 1.2652,
      "step": 218
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.09995076060295105,
      "learning_rate": 4.9394062872148604e-05,
      "loss": 1.2524,
      "step": 219
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.09488219022750854,
      "learning_rate": 4.9384369244108685e-05,
      "loss": 1.2505,
      "step": 220
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.1144493967294693,
      "learning_rate": 4.937459965848275e-05,
      "loss": 1.3485,
      "step": 221
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.09584493935108185,
      "learning_rate": 4.9364754145703165e-05,
      "loss": 1.2722,
      "step": 222
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.12084224820137024,
      "learning_rate": 4.935483273643882e-05,
      "loss": 1.4047,
      "step": 223
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.10730363428592682,
      "learning_rate": 4.9344835461595014e-05,
      "loss": 1.2585,
      "step": 224
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.08327766507863998,
      "learning_rate": 4.933476235231337e-05,
      "loss": 1.205,
      "step": 225
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.09607108682394028,
      "learning_rate": 4.9324613439971736e-05,
      "loss": 1.3635,
      "step": 226
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.09505125880241394,
      "learning_rate": 4.931438875618408e-05,
      "loss": 1.3224,
      "step": 227
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.09247023612260818,
      "learning_rate": 4.930408833280043e-05,
      "loss": 1.3292,
      "step": 228
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.09089571982622147,
      "learning_rate": 4.929371220190671e-05,
      "loss": 1.3271,
      "step": 229
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.0896812379360199,
      "learning_rate": 4.928326039582468e-05,
      "loss": 1.281,
      "step": 230
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.10084407031536102,
      "learning_rate": 4.927273294711184e-05,
      "loss": 1.354,
      "step": 231
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.09202654659748077,
      "learning_rate": 4.9262129888561316e-05,
      "loss": 1.3373,
      "step": 232
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.09410908818244934,
      "learning_rate": 4.925145125320175e-05,
      "loss": 1.4251,
      "step": 233
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.08918359875679016,
      "learning_rate": 4.9240697074297206e-05,
      "loss": 1.2926,
      "step": 234
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.09744445979595184,
      "learning_rate": 4.9229867385347086e-05,
      "loss": 1.2261,
      "step": 235
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.09546810388565063,
      "learning_rate": 4.921896222008598e-05,
      "loss": 1.3187,
      "step": 236
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.09971386194229126,
      "learning_rate": 4.920798161248361e-05,
      "loss": 1.3097,
      "step": 237
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.09423212707042694,
      "learning_rate": 4.9196925596744684e-05,
      "loss": 1.3812,
      "step": 238
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.09651963412761688,
      "learning_rate": 4.918579420730883e-05,
      "loss": 1.2818,
      "step": 239
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.10617819428443909,
      "learning_rate": 4.9174587478850445e-05,
      "loss": 1.3433,
      "step": 240
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.09576583653688431,
      "learning_rate": 4.916330544627861e-05,
      "loss": 1.3034,
      "step": 241
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.09386958181858063,
      "learning_rate": 4.915194814473699e-05,
      "loss": 1.2635,
      "step": 242
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.1020546555519104,
      "learning_rate": 4.914051560960371e-05,
      "loss": 1.4071,
      "step": 243
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.09851020574569702,
      "learning_rate": 4.912900787649124e-05,
      "loss": 1.3625,
      "step": 244
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.09171383827924728,
      "learning_rate": 4.9117424981246304e-05,
      "loss": 1.3349,
      "step": 245
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.09428778290748596,
      "learning_rate": 4.910576695994975e-05,
      "loss": 1.3469,
      "step": 246
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.09907223284244537,
      "learning_rate": 4.909403384891644e-05,
      "loss": 1.3375,
      "step": 247
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.09412197768688202,
      "learning_rate": 4.908222568469516e-05,
      "loss": 1.3386,
      "step": 248
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.13928060233592987,
      "learning_rate": 4.907034250406846e-05,
      "loss": 1.3403,
      "step": 249
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.09467838704586029,
      "learning_rate": 4.9058384344052587e-05,
      "loss": 1.3549,
      "step": 250
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.08943888545036316,
      "learning_rate": 4.904635124189736e-05,
      "loss": 1.2899,
      "step": 251
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.08892600983381271,
      "learning_rate": 4.903424323508601e-05,
      "loss": 1.2404,
      "step": 252
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.08917702734470367,
      "learning_rate": 4.902206036133512e-05,
      "loss": 1.368,
      "step": 253
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.10863058269023895,
      "learning_rate": 4.900980265859448e-05,
      "loss": 1.4147,
      "step": 254
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.09652639180421829,
      "learning_rate": 4.8997470165046976e-05,
      "loss": 1.2795,
      "step": 255
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.09574037045240402,
      "learning_rate": 4.8985062919108474e-05,
      "loss": 1.3202,
      "step": 256
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.10350500792264938,
      "learning_rate": 4.897258095942766e-05,
      "loss": 1.3874,
      "step": 257
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.0920305848121643,
      "learning_rate": 4.896002432488599e-05,
      "loss": 1.3145,
      "step": 258
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.12242481857538223,
      "learning_rate": 4.8947393054597534e-05,
      "loss": 1.3162,
      "step": 259
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.08995360136032104,
      "learning_rate": 4.8934687187908834e-05,
      "loss": 1.3715,
      "step": 260
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.09082052856683731,
      "learning_rate": 4.8921906764398805e-05,
      "loss": 1.2569,
      "step": 261
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.09512560814619064,
      "learning_rate": 4.890905182387861e-05,
      "loss": 1.4848,
      "step": 262
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.08932676166296005,
      "learning_rate": 4.8896122406391556e-05,
      "loss": 1.314,
      "step": 263
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.0890311747789383,
      "learning_rate": 4.888311855221289e-05,
      "loss": 1.3381,
      "step": 264
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.08900325000286102,
      "learning_rate": 4.887004030184979e-05,
      "loss": 1.3001,
      "step": 265
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.08832228183746338,
      "learning_rate": 4.885688769604114e-05,
      "loss": 1.3714,
      "step": 266
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.09763582050800323,
      "learning_rate": 4.884366077575747e-05,
      "loss": 1.4229,
      "step": 267
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.1476629674434662,
      "learning_rate": 4.883035958220077e-05,
      "loss": 1.5083,
      "step": 268
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.10069846361875534,
      "learning_rate": 4.881698415680442e-05,
      "loss": 1.3742,
      "step": 269
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.0915488749742508,
      "learning_rate": 4.8803534541233014e-05,
      "loss": 1.3738,
      "step": 270
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.09633270651102066,
      "learning_rate": 4.879001077738227e-05,
      "loss": 1.3044,
      "step": 271
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.09047040343284607,
      "learning_rate": 4.877641290737884e-05,
      "loss": 1.2579,
      "step": 272
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.09998361766338348,
      "learning_rate": 4.876274097358027e-05,
      "loss": 1.3221,
      "step": 273
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.10295653343200684,
      "learning_rate": 4.874899501857477e-05,
      "loss": 1.3666,
      "step": 274
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.09751718491315842,
      "learning_rate": 4.873517508518116e-05,
      "loss": 1.3144,
      "step": 275
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.09435424208641052,
      "learning_rate": 4.872128121644868e-05,
      "loss": 1.3125,
      "step": 276
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.08840420097112656,
      "learning_rate": 4.870731345565689e-05,
      "loss": 1.3244,
      "step": 277
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.09967703372240067,
      "learning_rate": 4.8693271846315515e-05,
      "loss": 1.3218,
      "step": 278
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.09583844244480133,
      "learning_rate": 4.867915643216434e-05,
      "loss": 1.3866,
      "step": 279
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.09268955886363983,
      "learning_rate": 4.866496725717303e-05,
      "loss": 1.2995,
      "step": 280
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.09287814050912857,
      "learning_rate": 4.8650704365541035e-05,
      "loss": 1.3388,
      "step": 281
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.09162658452987671,
      "learning_rate": 4.863636780169742e-05,
      "loss": 1.2241,
      "step": 282
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.10253284871578217,
      "learning_rate": 4.8621957610300736e-05,
      "loss": 1.3496,
      "step": 283
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.09558489173650742,
      "learning_rate": 4.860747383623889e-05,
      "loss": 1.3608,
      "step": 284
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.11398740112781525,
      "learning_rate": 4.859291652462903e-05,
      "loss": 1.3924,
      "step": 285
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.10008247196674347,
      "learning_rate": 4.8578285720817314e-05,
      "loss": 1.2955,
      "step": 286
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.09404585510492325,
      "learning_rate": 4.8563581470378875e-05,
      "loss": 1.4093,
      "step": 287
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.09330190718173981,
      "learning_rate": 4.8548803819117614e-05,
      "loss": 1.413,
      "step": 288
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.09668520838022232,
      "learning_rate": 4.8533952813066094e-05,
      "loss": 1.3729,
      "step": 289
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.09101221710443497,
      "learning_rate": 4.851902849848536e-05,
      "loss": 1.3243,
      "step": 290
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.09626659750938416,
      "learning_rate": 4.8504030921864816e-05,
      "loss": 1.2227,
      "step": 291
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.10459615290164948,
      "learning_rate": 4.848896012992208e-05,
      "loss": 1.4128,
      "step": 292
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.09928146004676819,
      "learning_rate": 4.847381616960286e-05,
      "loss": 1.4355,
      "step": 293
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.09046845883131027,
      "learning_rate": 4.8458599088080735e-05,
      "loss": 1.3679,
      "step": 294
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.10265221446752548,
      "learning_rate": 4.844330893275711e-05,
      "loss": 1.329,
      "step": 295
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.09289583563804626,
      "learning_rate": 4.8427945751260986e-05,
      "loss": 1.3476,
      "step": 296
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.09216302633285522,
      "learning_rate": 4.8412509591448835e-05,
      "loss": 1.2461,
      "step": 297
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.09431559592485428,
      "learning_rate": 4.839700050140448e-05,
      "loss": 1.3014,
      "step": 298
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.09718261659145355,
      "learning_rate": 4.838141852943891e-05,
      "loss": 1.4044,
      "step": 299
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.09670992195606232,
      "learning_rate": 4.836576372409015e-05,
      "loss": 1.3345,
      "step": 300
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.09435348957777023,
      "learning_rate": 4.835003613412308e-05,
      "loss": 1.254,
      "step": 301
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.09258078038692474,
      "learning_rate": 4.8334235808529335e-05,
      "loss": 1.3898,
      "step": 302
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.0906069427728653,
      "learning_rate": 4.83183627965271e-05,
      "loss": 1.3753,
      "step": 303
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.159384086728096,
      "learning_rate": 4.830241714756099e-05,
      "loss": 1.2721,
      "step": 304
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.0922141820192337,
      "learning_rate": 4.828639891130189e-05,
      "loss": 1.2806,
      "step": 305
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.09928394109010696,
      "learning_rate": 4.827030813764677e-05,
      "loss": 1.3605,
      "step": 306
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.09432511776685715,
      "learning_rate": 4.825414487671859e-05,
      "loss": 1.4276,
      "step": 307
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.09470757842063904,
      "learning_rate": 4.823790917886607e-05,
      "loss": 1.2807,
      "step": 308
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.09105392545461655,
      "learning_rate": 4.822160109466361e-05,
      "loss": 1.3163,
      "step": 309
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.09101832658052444,
      "learning_rate": 4.8205220674911074e-05,
      "loss": 1.2988,
      "step": 310
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.09433624148368835,
      "learning_rate": 4.8188767970633647e-05,
      "loss": 1.3743,
      "step": 311
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.1052064523100853,
      "learning_rate": 4.8172243033081695e-05,
      "loss": 1.4047,
      "step": 312
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.09831178188323975,
      "learning_rate": 4.8155645913730585e-05,
      "loss": 1.4189,
      "step": 313
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.09389661997556686,
      "learning_rate": 4.8138976664280536e-05,
      "loss": 1.3911,
      "step": 314
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.0961923897266388,
      "learning_rate": 4.812223533665643e-05,
      "loss": 1.3879,
      "step": 315
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.09563787281513214,
      "learning_rate": 4.8105421983007715e-05,
      "loss": 1.3871,
      "step": 316
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.0972226932644844,
      "learning_rate": 4.808853665570816e-05,
      "loss": 1.4138,
      "step": 317
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.09669040143489838,
      "learning_rate": 4.807157940735577e-05,
      "loss": 1.3376,
      "step": 318
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.10110729932785034,
      "learning_rate": 4.805455029077255e-05,
      "loss": 1.3442,
      "step": 319
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.10544568300247192,
      "learning_rate": 4.803744935900439e-05,
      "loss": 1.259,
      "step": 320
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.09788301587104797,
      "learning_rate": 4.802027666532089e-05,
      "loss": 1.3767,
      "step": 321
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.09752319008111954,
      "learning_rate": 4.8003032263215185e-05,
      "loss": 1.4368,
      "step": 322
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.1186918169260025,
      "learning_rate": 4.798571620640378e-05,
      "loss": 1.2999,
      "step": 323
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.09278569370508194,
      "learning_rate": 4.79683285488264e-05,
      "loss": 1.3786,
      "step": 324
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.11016564816236496,
      "learning_rate": 4.795086934464579e-05,
      "loss": 1.3475,
      "step": 325
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.11838585883378983,
      "learning_rate": 4.7933338648247563e-05,
      "loss": 1.3264,
      "step": 326
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.09825286269187927,
      "learning_rate": 4.791573651424003e-05,
      "loss": 1.314,
      "step": 327
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.12940770387649536,
      "learning_rate": 4.789806299745405e-05,
      "loss": 1.2689,
      "step": 328
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.09449135512113571,
      "learning_rate": 4.7880318152942816e-05,
      "loss": 1.3096,
      "step": 329
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.10377182811498642,
      "learning_rate": 4.786250203598174e-05,
      "loss": 1.3784,
      "step": 330
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.09918166697025299,
      "learning_rate": 4.78446147020682e-05,
      "loss": 1.44,
      "step": 331
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.09931398928165436,
      "learning_rate": 4.782665620692147e-05,
      "loss": 1.3034,
      "step": 332
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.20402568578720093,
      "learning_rate": 4.780862660648244e-05,
      "loss": 1.3637,
      "step": 333
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.10577716678380966,
      "learning_rate": 4.779052595691355e-05,
      "loss": 1.2544,
      "step": 334
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.1048274114727974,
      "learning_rate": 4.77723543145985e-05,
      "loss": 1.4564,
      "step": 335
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.09773129224777222,
      "learning_rate": 4.775411173614218e-05,
      "loss": 1.3061,
      "step": 336
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.10128381103277206,
      "learning_rate": 4.773579827837041e-05,
      "loss": 1.3197,
      "step": 337
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.087751604616642,
      "learning_rate": 4.7717413998329844e-05,
      "loss": 1.2628,
      "step": 338
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.09778417646884918,
      "learning_rate": 4.769895895328769e-05,
      "loss": 1.4227,
      "step": 339
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.10379809141159058,
      "learning_rate": 4.768043320073165e-05,
      "loss": 1.347,
      "step": 340
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.09904753416776657,
      "learning_rate": 4.766183679836964e-05,
      "loss": 1.3168,
      "step": 341
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.2791866958141327,
      "learning_rate": 4.7643169804129665e-05,
      "loss": 1.2482,
      "step": 342
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.24763400852680206,
      "learning_rate": 4.762443227615963e-05,
      "loss": 1.4319,
      "step": 343
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.0879005715250969,
      "learning_rate": 4.7605624272827126e-05,
      "loss": 1.2904,
      "step": 344
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.09311103075742722,
      "learning_rate": 4.7586745852719316e-05,
      "loss": 1.3017,
      "step": 345
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.09575958549976349,
      "learning_rate": 4.756779707464269e-05,
      "loss": 1.354,
      "step": 346
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.09556791186332703,
      "learning_rate": 4.754877799762291e-05,
      "loss": 1.3307,
      "step": 347
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.09889087080955505,
      "learning_rate": 4.752968868090459e-05,
      "loss": 1.3329,
      "step": 348
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.0879189744591713,
      "learning_rate": 4.75105291839512e-05,
      "loss": 1.3677,
      "step": 349
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.09398270398378372,
      "learning_rate": 4.749129956644477e-05,
      "loss": 1.3096,
      "step": 350
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.0942855104804039,
      "learning_rate": 4.747199988828579e-05,
      "loss": 1.4615,
      "step": 351
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.09009843319654465,
      "learning_rate": 4.7452630209592966e-05,
      "loss": 1.3422,
      "step": 352
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.11878379434347153,
      "learning_rate": 4.7433190590703055e-05,
      "loss": 1.3167,
      "step": 353
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.10389739274978638,
      "learning_rate": 4.7413681092170715e-05,
      "loss": 1.3042,
      "step": 354
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.09575341641902924,
      "learning_rate": 4.739410177476824e-05,
      "loss": 1.3589,
      "step": 355
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.0868607833981514,
      "learning_rate": 4.7374452699485426e-05,
      "loss": 1.3461,
      "step": 356
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.09217596054077148,
      "learning_rate": 4.7354733927529374e-05,
      "loss": 1.3186,
      "step": 357
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.08772280067205429,
      "learning_rate": 4.7334945520324265e-05,
      "loss": 1.3425,
      "step": 358
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.10162526369094849,
      "learning_rate": 4.731508753951122e-05,
      "loss": 1.4125,
      "step": 359
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.09132665395736694,
      "learning_rate": 4.729516004694808e-05,
      "loss": 1.3316,
      "step": 360
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.0943852961063385,
      "learning_rate": 4.72751631047092e-05,
      "loss": 1.2758,
      "step": 361
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.08872327208518982,
      "learning_rate": 4.725509677508528e-05,
      "loss": 1.3476,
      "step": 362
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.09403718262910843,
      "learning_rate": 4.7234961120583174e-05,
      "loss": 1.2631,
      "step": 363
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.08866386115550995,
      "learning_rate": 4.7214756203925676e-05,
      "loss": 1.3324,
      "step": 364
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.10122651606798172,
      "learning_rate": 4.719448208805132e-05,
      "loss": 1.3209,
      "step": 365
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.09378830343484879,
      "learning_rate": 4.71741388361142e-05,
      "loss": 1.3856,
      "step": 366
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.09723149240016937,
      "learning_rate": 4.7153726511483786e-05,
      "loss": 1.3676,
      "step": 367
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.08986017853021622,
      "learning_rate": 4.713324517774471e-05,
      "loss": 1.324,
      "step": 368
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.08899199962615967,
      "learning_rate": 4.711269489869653e-05,
      "loss": 1.213,
      "step": 369
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.0900987833738327,
      "learning_rate": 4.709207573835363e-05,
      "loss": 1.3337,
      "step": 370
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.0994141548871994,
      "learning_rate": 4.70713877609449e-05,
      "loss": 1.458,
      "step": 371
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.10680098831653595,
      "learning_rate": 4.7050631030913644e-05,
      "loss": 1.3494,
      "step": 372
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.17104795575141907,
      "learning_rate": 4.70298056129173e-05,
      "loss": 1.2644,
      "step": 373
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.08228933811187744,
      "learning_rate": 4.700891157182729e-05,
      "loss": 1.2638,
      "step": 374
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.08857174217700958,
      "learning_rate": 4.698794897272877e-05,
      "loss": 1.3473,
      "step": 375
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.08300795406103134,
      "learning_rate": 4.696691788092049e-05,
      "loss": 1.3295,
      "step": 376
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.11059122532606125,
      "learning_rate": 4.694581836191454e-05,
      "loss": 1.2383,
      "step": 377
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.09859713912010193,
      "learning_rate": 4.692465048143615e-05,
      "loss": 1.3404,
      "step": 378
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.08896754682064056,
      "learning_rate": 4.6903414305423507e-05,
      "loss": 1.3865,
      "step": 379
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.091724693775177,
      "learning_rate": 4.688210990002755e-05,
      "loss": 1.3131,
      "step": 380
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.09057147055864334,
      "learning_rate": 4.686073733161173e-05,
      "loss": 1.2792,
      "step": 381
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.09829523414373398,
      "learning_rate": 4.683929666675185e-05,
      "loss": 1.331,
      "step": 382
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.09669854491949081,
      "learning_rate": 4.681778797223582e-05,
      "loss": 1.3238,
      "step": 383
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.0903388261795044,
      "learning_rate": 4.679621131506347e-05,
      "loss": 1.295,
      "step": 384
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.09009499102830887,
      "learning_rate": 4.6774566762446324e-05,
      "loss": 1.3714,
      "step": 385
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.09612104296684265,
      "learning_rate": 4.675285438180741e-05,
      "loss": 1.3032,
      "step": 386
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.09140989184379578,
      "learning_rate": 4.673107424078105e-05,
      "loss": 1.4714,
      "step": 387
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.0838828906416893,
      "learning_rate": 4.670922640721261e-05,
      "loss": 1.2755,
      "step": 388
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.09146388620138168,
      "learning_rate": 4.6687310949158344e-05,
      "loss": 1.3965,
      "step": 389
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.09024867415428162,
      "learning_rate": 4.6665327934885174e-05,
      "loss": 1.3613,
      "step": 390
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.08649884164333344,
      "learning_rate": 4.664327743287041e-05,
      "loss": 1.2359,
      "step": 391
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.0954437181353569,
      "learning_rate": 4.6621159511801635e-05,
      "loss": 1.315,
      "step": 392
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.09183992445468903,
      "learning_rate": 4.6598974240576406e-05,
      "loss": 1.3338,
      "step": 393
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.09896053373813629,
      "learning_rate": 4.6576721688302105e-05,
      "loss": 1.3156,
      "step": 394
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.0989176332950592,
      "learning_rate": 4.6554401924295686e-05,
      "loss": 1.3154,
      "step": 395
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.08492989093065262,
      "learning_rate": 4.653201501808346e-05,
      "loss": 1.3708,
      "step": 396
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.09674088656902313,
      "learning_rate": 4.650956103940089e-05,
      "loss": 1.3821,
      "step": 397
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.09763908386230469,
      "learning_rate": 4.648704005819238e-05,
      "loss": 1.4568,
      "step": 398
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.09475667029619217,
      "learning_rate": 4.6464452144611046e-05,
      "loss": 1.3513,
      "step": 399
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.09377523511648178,
      "learning_rate": 4.644179736901848e-05,
      "loss": 1.3554,
      "step": 400
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.08964923769235611,
      "learning_rate": 4.641907580198458e-05,
      "loss": 1.3095,
      "step": 401
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.08872678130865097,
      "learning_rate": 4.6396287514287275e-05,
      "loss": 1.3737,
      "step": 402
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.0877537652850151,
      "learning_rate": 4.637343257691234e-05,
      "loss": 1.3265,
      "step": 403
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.08949404209852219,
      "learning_rate": 4.635051106105316e-05,
      "loss": 1.3761,
      "step": 404
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.1385626494884491,
      "learning_rate": 4.632752303811053e-05,
      "loss": 1.4187,
      "step": 405
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.09133272618055344,
      "learning_rate": 4.6304468579692384e-05,
      "loss": 1.1747,
      "step": 406
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.09598804265260696,
      "learning_rate": 4.6281347757613626e-05,
      "loss": 1.3986,
      "step": 407
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.09424811601638794,
      "learning_rate": 4.625816064389589e-05,
      "loss": 1.3617,
      "step": 408
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.10442469269037247,
      "learning_rate": 4.623490731076728e-05,
      "loss": 1.2631,
      "step": 409
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.08816258609294891,
      "learning_rate": 4.62115878306622e-05,
      "loss": 1.3181,
      "step": 410
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.09299376606941223,
      "learning_rate": 4.61882022762211e-05,
      "loss": 1.3832,
      "step": 411
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.09412654489278793,
      "learning_rate": 4.616475072029024e-05,
      "loss": 1.2522,
      "step": 412
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.16249871253967285,
      "learning_rate": 4.614123323592148e-05,
      "loss": 1.4142,
      "step": 413
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.08951661735773087,
      "learning_rate": 4.611764989637205e-05,
      "loss": 1.26,
      "step": 414
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.09598527103662491,
      "learning_rate": 4.609400077510433e-05,
      "loss": 1.3028,
      "step": 415
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.09367496520280838,
      "learning_rate": 4.607028594578559e-05,
      "loss": 1.2739,
      "step": 416
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.09410126507282257,
      "learning_rate": 4.6046505482287794e-05,
      "loss": 1.4011,
      "step": 417
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.0993422120809555,
      "learning_rate": 4.602265945868735e-05,
      "loss": 1.2641,
      "step": 418
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.08851709961891174,
      "learning_rate": 4.59987479492649e-05,
      "loss": 1.3415,
      "step": 419
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.10108134895563126,
      "learning_rate": 4.5974771028505064e-05,
      "loss": 1.3679,
      "step": 420
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.08684688806533813,
      "learning_rate": 4.595072877109622e-05,
      "loss": 1.3404,
      "step": 421
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.102836474776268,
      "learning_rate": 4.592662125193027e-05,
      "loss": 1.2668,
      "step": 422
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.08989983052015305,
      "learning_rate": 4.59024485461024e-05,
      "loss": 1.3408,
      "step": 423
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.11563542485237122,
      "learning_rate": 4.5878210728910894e-05,
      "loss": 1.4273,
      "step": 424
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.09047998487949371,
      "learning_rate": 4.585390787585679e-05,
      "loss": 1.3598,
      "step": 425
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.17246578633785248,
      "learning_rate": 4.5829540062643773e-05,
      "loss": 1.4129,
      "step": 426
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.09310886263847351,
      "learning_rate": 4.5805107365177844e-05,
      "loss": 1.3474,
      "step": 427
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.12132611870765686,
      "learning_rate": 4.5780609859567136e-05,
      "loss": 1.2837,
      "step": 428
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.09810657054185867,
      "learning_rate": 4.575604762212167e-05,
      "loss": 1.2571,
      "step": 429
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.1014370396733284,
      "learning_rate": 4.573142072935307e-05,
      "loss": 1.3435,
      "step": 430
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.09363662451505661,
      "learning_rate": 4.570672925797439e-05,
      "loss": 1.3725,
      "step": 431
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.11186376214027405,
      "learning_rate": 4.568197328489986e-05,
      "loss": 1.3095,
      "step": 432
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.09949055314064026,
      "learning_rate": 4.5657152887244606e-05,
      "loss": 1.2912,
      "step": 433
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.10397903621196747,
      "learning_rate": 4.563226814232444e-05,
      "loss": 1.3221,
      "step": 434
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.09242431074380875,
      "learning_rate": 4.5607319127655636e-05,
      "loss": 1.3132,
      "step": 435
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.09308359771966934,
      "learning_rate": 4.5582305920954643e-05,
      "loss": 1.3784,
      "step": 436
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.10053109377622604,
      "learning_rate": 4.555722860013789e-05,
      "loss": 1.3767,
      "step": 437
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.0856412872672081,
      "learning_rate": 4.553208724332153e-05,
      "loss": 1.3518,
      "step": 438
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.09756068885326385,
      "learning_rate": 4.550688192882115e-05,
      "loss": 1.5058,
      "step": 439
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.10069140046834946,
      "learning_rate": 4.54816127351516e-05,
      "loss": 1.3025,
      "step": 440
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.0943472757935524,
      "learning_rate": 4.545627974102671e-05,
      "loss": 1.3474,
      "step": 441
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.094652459025383,
      "learning_rate": 4.543088302535903e-05,
      "loss": 1.3433,
      "step": 442
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.09766332805156708,
      "learning_rate": 4.540542266725963e-05,
      "loss": 1.2912,
      "step": 443
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.0971674770116806,
      "learning_rate": 4.5379898746037804e-05,
      "loss": 1.269,
      "step": 444
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.09349120408296585,
      "learning_rate": 4.535431134120086e-05,
      "loss": 1.2813,
      "step": 445
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.15912680327892303,
      "learning_rate": 4.532866053245385e-05,
      "loss": 1.323,
      "step": 446
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.1172761395573616,
      "learning_rate": 4.530294639969934e-05,
      "loss": 1.2852,
      "step": 447
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.09645481407642365,
      "learning_rate": 4.527716902303714e-05,
      "loss": 1.3793,
      "step": 448
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.09305194765329361,
      "learning_rate": 4.525132848276406e-05,
      "loss": 1.4287,
      "step": 449
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.09805100411176682,
      "learning_rate": 4.522542485937369e-05,
      "loss": 1.4003,
      "step": 450
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.09443584084510803,
      "learning_rate": 4.51994582335561e-05,
      "loss": 1.4108,
      "step": 451
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.09847624599933624,
      "learning_rate": 4.517342868619764e-05,
      "loss": 1.3774,
      "step": 452
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.09076975286006927,
      "learning_rate": 4.514733629838063e-05,
      "loss": 1.3411,
      "step": 453
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.09129495918750763,
      "learning_rate": 4.5121181151383143e-05,
      "loss": 1.2343,
      "step": 454
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.09538775682449341,
      "learning_rate": 4.509496332667878e-05,
      "loss": 1.309,
      "step": 455
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.0932200700044632,
      "learning_rate": 4.506868290593635e-05,
      "loss": 1.4046,
      "step": 456
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.09325648844242096,
      "learning_rate": 4.5042339971019666e-05,
      "loss": 1.2966,
      "step": 457
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.10446179658174515,
      "learning_rate": 4.501593460398726e-05,
      "loss": 1.3569,
      "step": 458
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.09409455209970474,
      "learning_rate": 4.498946688709216e-05,
      "loss": 1.3351,
      "step": 459
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.09080103039741516,
      "learning_rate": 4.4962936902781594e-05,
      "loss": 1.2639,
      "step": 460
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.09168045967817307,
      "learning_rate": 4.493634473369677e-05,
      "loss": 1.3146,
      "step": 461
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.08616971224546432,
      "learning_rate": 4.4909690462672585e-05,
      "loss": 1.3693,
      "step": 462
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.09884229302406311,
      "learning_rate": 4.488297417273741e-05,
      "loss": 1.3904,
      "step": 463
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.09650994092226028,
      "learning_rate": 4.4856195947112776e-05,
      "loss": 1.4442,
      "step": 464
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.09592213481664658,
      "learning_rate": 4.482935586921316e-05,
      "loss": 1.3532,
      "step": 465
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.1017376109957695,
      "learning_rate": 4.480245402264572e-05,
      "loss": 1.3927,
      "step": 466
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.08732669800519943,
      "learning_rate": 4.4775490491210016e-05,
      "loss": 1.3695,
      "step": 467
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.09283038228750229,
      "learning_rate": 4.474846535889773e-05,
      "loss": 1.3465,
      "step": 468
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.09399215877056122,
      "learning_rate": 4.472137870989247e-05,
      "loss": 1.3504,
      "step": 469
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.09396019577980042,
      "learning_rate": 4.4694230628569454e-05,
      "loss": 1.328,
      "step": 470
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.09818755835294724,
      "learning_rate": 4.466702119949526e-05,
      "loss": 1.3597,
      "step": 471
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.10373959690332413,
      "learning_rate": 4.463975050742757e-05,
      "loss": 1.2681,
      "step": 472
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.0993281677365303,
      "learning_rate": 4.461241863731489e-05,
      "loss": 1.3057,
      "step": 473
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.09096087515354156,
      "learning_rate": 4.4585025674296315e-05,
      "loss": 1.3061,
      "step": 474
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.09843237698078156,
      "learning_rate": 4.4557571703701226e-05,
      "loss": 1.2969,
      "step": 475
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.09445469826459885,
      "learning_rate": 4.453005681104906e-05,
      "loss": 1.2051,
      "step": 476
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.10724281519651413,
      "learning_rate": 4.4502481082049016e-05,
      "loss": 1.3629,
      "step": 477
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.08827083557844162,
      "learning_rate": 4.4474844602599795e-05,
      "loss": 1.342,
      "step": 478
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.09455974400043488,
      "learning_rate": 4.444714745878936e-05,
      "loss": 1.2408,
      "step": 479
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.0963449627161026,
      "learning_rate": 4.44193897368946e-05,
      "loss": 1.3956,
      "step": 480
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.1244681105017662,
      "learning_rate": 4.439157152338116e-05,
      "loss": 1.3086,
      "step": 481
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.10625208914279938,
      "learning_rate": 4.436369290490307e-05,
      "loss": 1.3869,
      "step": 482
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.10267725586891174,
      "learning_rate": 4.433575396830256e-05,
      "loss": 1.3136,
      "step": 483
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.09765701740980148,
      "learning_rate": 4.4307754800609725e-05,
      "loss": 1.2566,
      "step": 484
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.08589779585599899,
      "learning_rate": 4.427969548904228e-05,
      "loss": 1.3314,
      "step": 485
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.08467262983322144,
      "learning_rate": 4.4251576121005314e-05,
      "loss": 1.228,
      "step": 486
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.09379052370786667,
      "learning_rate": 4.422339678409096e-05,
      "loss": 1.3658,
      "step": 487
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.10410299897193909,
      "learning_rate": 4.4195157566078186e-05,
      "loss": 1.2542,
      "step": 488
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.09087986499071121,
      "learning_rate": 4.416685855493247e-05,
      "loss": 1.2092,
      "step": 489
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.09085581451654434,
      "learning_rate": 4.413849983880554e-05,
      "loss": 1.2962,
      "step": 490
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.09474420547485352,
      "learning_rate": 4.411008150603514e-05,
      "loss": 1.3953,
      "step": 491
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.09279070049524307,
      "learning_rate": 4.408160364514468e-05,
      "loss": 1.3184,
      "step": 492
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.09756158292293549,
      "learning_rate": 4.405306634484303e-05,
      "loss": 1.3197,
      "step": 493
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.09180128574371338,
      "learning_rate": 4.40244696940242e-05,
      "loss": 1.2856,
      "step": 494
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.09457723051309586,
      "learning_rate": 4.399581378176707e-05,
      "loss": 1.3081,
      "step": 495
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.09794706106185913,
      "learning_rate": 4.396709869733515e-05,
      "loss": 1.3562,
      "step": 496
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.0874537006020546,
      "learning_rate": 4.3938324530176236e-05,
      "loss": 1.3235,
      "step": 497
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.0974416732788086,
      "learning_rate": 4.39094913699222e-05,
      "loss": 1.4397,
      "step": 498
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.09341360628604889,
      "learning_rate": 4.388059930638865e-05,
      "loss": 1.3448,
      "step": 499
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.08748738467693329,
      "learning_rate": 4.385164842957469e-05,
      "loss": 1.2659,
      "step": 500
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 1874,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 50,
  "total_flos": 1.3535591043956736e+19,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}