{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 4.9914529914529915,
  "eval_steps": 500,
  "global_step": 149,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.03418803418803419,
      "grad_norm": 83.55925960057581,
      "learning_rate": 0.0,
      "loss": 0.1635,
      "step": 1
    },
    {
      "epoch": 0.06837606837606838,
      "grad_norm": 75.5491523818588,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 0.205,
      "step": 2
    },
    {
      "epoch": 0.10256410256410256,
      "grad_norm": 69.68045087594498,
      "learning_rate": 4.000000000000001e-06,
      "loss": 0.2142,
      "step": 3
    },
    {
      "epoch": 0.13675213675213677,
      "grad_norm": 57.086453365632416,
      "learning_rate": 6e-06,
      "loss": 0.1981,
      "step": 4
    },
    {
      "epoch": 0.17094017094017094,
      "grad_norm": 49.239480244872944,
      "learning_rate": 8.000000000000001e-06,
      "loss": 0.3847,
      "step": 5
    },
    {
      "epoch": 0.20512820512820512,
      "grad_norm": 16.322352326647238,
      "learning_rate": 1e-05,
      "loss": 0.1727,
      "step": 6
    },
    {
      "epoch": 0.23931623931623933,
      "grad_norm": 10.054689685370425,
      "learning_rate": 9.99882649009242e-06,
      "loss": 0.1366,
      "step": 7
    },
    {
      "epoch": 0.27350427350427353,
      "grad_norm": 12.472152502380862,
      "learning_rate": 9.995306511219885e-06,
      "loss": 0.2145,
      "step": 8
    },
    {
      "epoch": 0.3076923076923077,
      "grad_norm": 7.146434932900831,
      "learning_rate": 9.989441715674422e-06,
      "loss": 0.175,
      "step": 9
    },
    {
      "epoch": 0.3418803418803419,
      "grad_norm": 10.829219452046477,
      "learning_rate": 9.981234856414306e-06,
      "loss": 0.1697,
      "step": 10
    },
    {
      "epoch": 0.37606837606837606,
      "grad_norm": 3.9257977267573696,
      "learning_rate": 9.970689785771798e-06,
      "loss": 0.1282,
      "step": 11
    },
    {
      "epoch": 0.41025641025641024,
      "grad_norm": 4.881222222368343,
      "learning_rate": 9.957811453644848e-06,
      "loss": 0.1137,
      "step": 12
    },
    {
      "epoch": 0.4444444444444444,
      "grad_norm": 7.191012828431416,
      "learning_rate": 9.942605905173593e-06,
      "loss": 0.0955,
      "step": 13
    },
    {
      "epoch": 0.47863247863247865,
      "grad_norm": 11.172906117826464,
      "learning_rate": 9.925080277902743e-06,
      "loss": 0.1265,
      "step": 14
    },
    {
      "epoch": 0.5128205128205128,
      "grad_norm": 6.787542905973914,
      "learning_rate": 9.905242798431196e-06,
      "loss": 0.1066,
      "step": 15
    },
    {
      "epoch": 0.5470085470085471,
      "grad_norm": 3.7905884982716795,
      "learning_rate": 9.883102778550434e-06,
      "loss": 0.0692,
      "step": 16
    },
    {
      "epoch": 0.5811965811965812,
      "grad_norm": 8.499180844600431,
      "learning_rate": 9.858670610873528e-06,
      "loss": 0.075,
      "step": 17
    },
    {
      "epoch": 0.6153846153846154,
      "grad_norm": 7.628854941819294,
      "learning_rate": 9.831957763956814e-06,
      "loss": 0.0844,
      "step": 18
    },
    {
      "epoch": 0.6495726495726496,
      "grad_norm": 3.4788385946157008,
      "learning_rate": 9.802976776916493e-06,
      "loss": 0.04,
      "step": 19
    },
    {
      "epoch": 0.6837606837606838,
      "grad_norm": 9.925442676615837,
      "learning_rate": 9.771741253542742e-06,
      "loss": 0.1003,
      "step": 20
    },
    {
      "epoch": 0.717948717948718,
      "grad_norm": 8.49069545035768,
      "learning_rate": 9.738265855914014e-06,
      "loss": 0.0564,
      "step": 21
    },
    {
      "epoch": 0.7521367521367521,
      "grad_norm": 9.111442462642389,
      "learning_rate": 9.70256629751462e-06,
      "loss": 0.0324,
      "step": 22
    },
    {
      "epoch": 0.7863247863247863,
      "grad_norm": 7.948522736839896,
      "learning_rate": 9.664659335858755e-06,
      "loss": 0.0668,
      "step": 23
    },
    {
      "epoch": 0.8205128205128205,
      "grad_norm": 16.01231870016539,
      "learning_rate": 9.624562764624445e-06,
      "loss": 0.0726,
      "step": 24
    },
    {
      "epoch": 0.8547008547008547,
      "grad_norm": 13.854838853842292,
      "learning_rate": 9.582295405301131e-06,
      "loss": 0.103,
      "step": 25
    },
    {
      "epoch": 0.8888888888888888,
      "grad_norm": 3.4607583628632517,
      "learning_rate": 9.537877098354787e-06,
      "loss": 0.0585,
      "step": 26
    },
    {
      "epoch": 0.9230769230769231,
      "grad_norm": 12.924003397447635,
      "learning_rate": 9.491328693914723e-06,
      "loss": 0.0402,
      "step": 27
    },
    {
      "epoch": 0.9572649572649573,
      "grad_norm": 4.118368098723785,
      "learning_rate": 9.442672041986456e-06,
      "loss": 0.0662,
      "step": 28
    },
    {
      "epoch": 0.9914529914529915,
      "grad_norm": 16.501556013072644,
      "learning_rate": 9.391929982195233e-06,
      "loss": 0.0641,
      "step": 29
    },
    {
      "epoch": 1.0,
      "grad_norm": 16.501556013072644,
      "learning_rate": 9.339126333065008e-06,
      "loss": 0.0179,
      "step": 30
    },
    {
      "epoch": 1.0341880341880343,
      "grad_norm": 10.007135158066854,
      "learning_rate": 9.284285880837947e-06,
      "loss": 0.0725,
      "step": 31
    },
    {
      "epoch": 1.0683760683760684,
      "grad_norm": 4.196446854828196,
      "learning_rate": 9.22743436783966e-06,
      "loss": 0.0495,
      "step": 32
    },
    {
      "epoch": 1.1025641025641026,
      "grad_norm": 4.548096237677968,
      "learning_rate": 9.168598480395653e-06,
      "loss": 0.0403,
      "step": 33
    },
    {
      "epoch": 1.1367521367521367,
      "grad_norm": 10.203571292360358,
      "learning_rate": 9.107805836304658e-06,
      "loss": 0.0786,
      "step": 34
    },
    {
      "epoch": 1.170940170940171,
      "grad_norm": 12.66007143709808,
      "learning_rate": 9.045084971874738e-06,
      "loss": 0.1094,
      "step": 35
    },
    {
      "epoch": 1.205128205128205,
      "grad_norm": 12.611631213450272,
      "learning_rate": 8.98046532852822e-06,
      "loss": 0.0676,
      "step": 36
    },
    {
      "epoch": 1.2393162393162394,
      "grad_norm": 13.35830031699019,
      "learning_rate": 8.91397723898178e-06,
      "loss": 0.0393,
      "step": 37
    },
    {
      "epoch": 1.2735042735042734,
      "grad_norm": 8.690400816793332,
      "learning_rate": 8.845651913008145e-06,
      "loss": 0.065,
      "step": 38
    },
    {
      "epoch": 1.3076923076923077,
      "grad_norm": 14.615618176894753,
      "learning_rate": 8.775521422786104e-06,
      "loss": 0.0212,
      "step": 39
    },
    {
      "epoch": 1.341880341880342,
      "grad_norm": 2.1336317122196418,
      "learning_rate": 8.703618687845697e-06,
      "loss": 0.0288,
      "step": 40
    },
    {
      "epoch": 1.376068376068376,
      "grad_norm": 4.140427071604391,
      "learning_rate": 8.629977459615655e-06,
      "loss": 0.0432,
      "step": 41
    },
    {
      "epoch": 1.4102564102564101,
      "grad_norm": 5.22920788836053,
      "learning_rate": 8.554632305580355e-06,
      "loss": 0.0383,
      "step": 42
    },
    {
      "epoch": 1.4444444444444444,
      "grad_norm": 1.707299197465682,
      "learning_rate": 8.477618593053693e-06,
      "loss": 0.0264,
      "step": 43
    },
    {
      "epoch": 1.4786324786324787,
      "grad_norm": 5.940268523708764,
      "learning_rate": 8.39897247257754e-06,
      "loss": 0.0236,
      "step": 44
    },
    {
      "epoch": 1.5128205128205128,
      "grad_norm": 3.603359880690343,
      "learning_rate": 8.318730860952523e-06,
      "loss": 0.045,
      "step": 45
    },
    {
      "epoch": 1.547008547008547,
      "grad_norm": 5.184468932618174,
      "learning_rate": 8.23693142390914e-06,
      "loss": 0.0181,
      "step": 46
    },
    {
      "epoch": 1.5811965811965814,
      "grad_norm": 4.729640403035601,
      "learning_rate": 8.153612558427311e-06,
      "loss": 0.0245,
      "step": 47
    },
    {
      "epoch": 1.6153846153846154,
      "grad_norm": 1.4004395210335507,
      "learning_rate": 8.068813374712689e-06,
      "loss": 0.032,
      "step": 48
    },
    {
      "epoch": 1.6495726495726495,
      "grad_norm": 5.483564379255356,
      "learning_rate": 7.982573677838172e-06,
      "loss": 0.0391,
      "step": 49
    },
    {
      "epoch": 1.6837606837606838,
      "grad_norm": 12.48720699386242,
      "learning_rate": 7.894933949059245e-06,
      "loss": 0.0305,
      "step": 50
    },
    {
      "epoch": 1.717948717948718,
      "grad_norm": 11.67481825168841,
      "learning_rate": 7.805935326811913e-06,
      "loss": 0.0643,
      "step": 51
    },
    {
      "epoch": 1.7521367521367521,
      "grad_norm": 8.31052833795307,
      "learning_rate": 7.715619587402165e-06,
      "loss": 0.0325,
      "step": 52
    },
    {
      "epoch": 1.7863247863247862,
      "grad_norm": 3.45860780579628,
      "learning_rate": 7.624029125396004e-06,
      "loss": 0.0213,
      "step": 53
    },
    {
      "epoch": 1.8205128205128205,
      "grad_norm": 3.3328870728402125,
      "learning_rate": 7.53120693371927e-06,
      "loss": 0.003,
      "step": 54
    },
    {
      "epoch": 1.8547008547008548,
      "grad_norm": 1.0551381306202823,
      "learning_rate": 7.437196583476597e-06,
      "loss": 0.0316,
      "step": 55
    },
    {
      "epoch": 1.8888888888888888,
      "grad_norm": 4.940796522342355,
      "learning_rate": 7.342042203498952e-06,
      "loss": 0.0283,
      "step": 56
    },
    {
      "epoch": 1.9230769230769231,
      "grad_norm": 1.3838725683764157,
      "learning_rate": 7.245788459629397e-06,
      "loss": 0.006,
      "step": 57
    },
    {
      "epoch": 1.9572649572649574,
      "grad_norm": 1.105772730767765,
      "learning_rate": 7.148480533756759e-06,
      "loss": 0.0196,
      "step": 58
    },
    {
      "epoch": 1.9914529914529915,
      "grad_norm": 3.4472527456227082,
      "learning_rate": 7.050164102607081e-06,
      "loss": 0.0189,
      "step": 59
    },
    {
      "epoch": 2.0,
      "grad_norm": 3.4472527456227082,
      "learning_rate": 6.950885316302773e-06,
      "loss": 0.0017,
      "step": 60
    },
    {
      "epoch": 2.034188034188034,
      "grad_norm": 2.5594058854187103,
      "learning_rate": 6.850690776699574e-06,
      "loss": 0.0294,
      "step": 61
    },
    {
      "epoch": 2.0683760683760686,
      "grad_norm": 2.38316198568798,
      "learning_rate": 6.749627515511443e-06,
      "loss": 0.0197,
      "step": 62
    },
    {
      "epoch": 2.1025641025641026,
      "grad_norm": 3.9379010709797893,
      "learning_rate": 6.647742972233703e-06,
      "loss": 0.0296,
      "step": 63
    },
    {
      "epoch": 2.1367521367521367,
      "grad_norm": 3.4756056396321586,
      "learning_rate": 6.545084971874738e-06,
      "loss": 0.0167,
      "step": 64
    },
    {
      "epoch": 2.1709401709401708,
      "grad_norm": 25.199221603893488,
      "learning_rate": 6.441701702506755e-06,
      "loss": 0.0614,
      "step": 65
    },
    {
      "epoch": 2.2051282051282053,
      "grad_norm": 6.642877984563196,
      "learning_rate": 6.337641692646106e-06,
      "loss": 0.0214,
      "step": 66
    },
    {
      "epoch": 2.2393162393162394,
      "grad_norm": 2.797157842559016,
      "learning_rate": 6.2329537884738115e-06,
      "loss": 0.0104,
      "step": 67
    },
    {
      "epoch": 2.2735042735042734,
      "grad_norm": 4.3025299601872256,
      "learning_rate": 6.127687130906972e-06,
      "loss": 0.0232,
      "step": 68
    },
    {
      "epoch": 2.3076923076923075,
      "grad_norm": 11.127364863419961,
      "learning_rate": 6.021891132531825e-06,
      "loss": 0.0129,
      "step": 69
    },
    {
      "epoch": 2.341880341880342,
      "grad_norm": 2.6327078429367723,
      "learning_rate": 5.915615454409281e-06,
      "loss": 0.02,
      "step": 70
    },
    {
      "epoch": 2.376068376068376,
      "grad_norm": 6.642980085247562,
      "learning_rate": 5.808909982763825e-06,
      "loss": 0.001,
      "step": 71
    },
    {
      "epoch": 2.41025641025641,
      "grad_norm": 64.72672342279493,
      "learning_rate": 5.701824805566722e-06,
      "loss": 0.0466,
      "step": 72
    },
    {
      "epoch": 2.4444444444444446,
      "grad_norm": 7.531014554064458,
      "learning_rate": 5.594410189024533e-06,
      "loss": 0.0019,
      "step": 73
    },
    {
      "epoch": 2.4786324786324787,
      "grad_norm": 0.9212587754890356,
      "learning_rate": 5.4867165539839505e-06,
      "loss": 0.01,
      "step": 74
    },
    {
      "epoch": 2.5128205128205128,
      "grad_norm": 5.3676628920085125,
      "learning_rate": 5.378794452264053e-06,
      "loss": 0.0702,
      "step": 75
    },
    {
      "epoch": 2.547008547008547,
      "grad_norm": 8.604201249574777,
      "learning_rate": 5.270694542927089e-06,
      "loss": 0.0078,
      "step": 76
    },
    {
      "epoch": 2.5811965811965814,
      "grad_norm": 3.442392132103759,
      "learning_rate": 5.1624675684989035e-06,
      "loss": 0.0099,
      "step": 77
    },
    {
      "epoch": 2.6153846153846154,
      "grad_norm": 1.226217039772073,
      "learning_rate": 5.054164331150199e-06,
      "loss": 0.0037,
      "step": 78
    },
    {
      "epoch": 2.6495726495726495,
      "grad_norm": 8.349284917924178,
      "learning_rate": 4.945835668849801e-06,
      "loss": 0.0486,
      "step": 79
    },
    {
      "epoch": 2.683760683760684,
      "grad_norm": 13.983929022139195,
      "learning_rate": 4.837532431501098e-06,
      "loss": 0.08,
      "step": 80
    },
    {
      "epoch": 2.717948717948718,
      "grad_norm": 7.290271590657317,
      "learning_rate": 4.729305457072913e-06,
      "loss": 0.0113,
      "step": 81
    },
    {
      "epoch": 2.752136752136752,
      "grad_norm": 5.948542359802803,
      "learning_rate": 4.621205547735949e-06,
      "loss": 0.0308,
      "step": 82
    },
    {
      "epoch": 2.786324786324786,
      "grad_norm": 1.7311032900556051,
      "learning_rate": 4.513283446016052e-06,
      "loss": 0.0176,
      "step": 83
    },
    {
      "epoch": 2.8205128205128203,
      "grad_norm": 2.919516222473814,
      "learning_rate": 4.4055898109754684e-06,
      "loss": 0.0049,
      "step": 84
    },
    {
      "epoch": 2.8547008547008548,
      "grad_norm": 1.9073912741858117,
      "learning_rate": 4.298175194433279e-06,
      "loss": 0.009,
      "step": 85
    },
    {
      "epoch": 2.888888888888889,
      "grad_norm": 3.474022268526306,
      "learning_rate": 4.191090017236177e-06,
      "loss": 0.0154,
      "step": 86
    },
    {
      "epoch": 2.9230769230769234,
      "grad_norm": 5.209234490321998,
      "learning_rate": 4.0843845455907195e-06,
      "loss": 0.0179,
      "step": 87
    },
    {
      "epoch": 2.9572649572649574,
      "grad_norm": 5.172963112227645,
      "learning_rate": 3.9781088674681764e-06,
      "loss": 0.0155,
      "step": 88
    },
    {
      "epoch": 2.9914529914529915,
      "grad_norm": 3.346150099815718,
      "learning_rate": 3.87231286909303e-06,
      "loss": 0.0194,
      "step": 89
    },
    {
      "epoch": 3.0,
      "grad_norm": 3.346150099815718,
      "learning_rate": 3.767046211526191e-06,
      "loss": 0.0016,
      "step": 90
    },
    {
      "epoch": 3.034188034188034,
      "grad_norm": 3.628053542950287,
      "learning_rate": 3.662358307353897e-06,
      "loss": 0.0115,
      "step": 91
    },
    {
      "epoch": 3.0683760683760686,
      "grad_norm": 2.4035577316004533,
      "learning_rate": 3.5582982974932467e-06,
      "loss": 0.0122,
      "step": 92
    },
    {
      "epoch": 3.1025641025641026,
      "grad_norm": 2.0115150221716145,
      "learning_rate": 3.4549150281252635e-06,
      "loss": 0.0065,
      "step": 93
    },
    {
      "epoch": 3.1367521367521367,
      "grad_norm": 1.7423751005899801,
      "learning_rate": 3.3522570277662986e-06,
      "loss": 0.0053,
      "step": 94
    },
    {
      "epoch": 3.1709401709401708,
      "grad_norm": 1.2120823372902632,
      "learning_rate": 3.250372484488558e-06,
      "loss": 0.0325,
      "step": 95
    },
    {
      "epoch": 3.2051282051282053,
      "grad_norm": 4.5742221694331535,
      "learning_rate": 3.149309223300428e-06,
      "loss": 0.0107,
      "step": 96
    },
    {
      "epoch": 3.2393162393162394,
      "grad_norm": 5.680204394389736,
      "learning_rate": 3.0491146836972273e-06,
      "loss": 0.0129,
      "step": 97
    },
    {
      "epoch": 3.2735042735042734,
      "grad_norm": 2.7770967720336697,
      "learning_rate": 2.9498358973929197e-06,
      "loss": 0.0051,
      "step": 98
    },
    {
      "epoch": 3.3076923076923075,
      "grad_norm": 2.8490979218142036,
      "learning_rate": 2.8515194662432423e-06,
      "loss": 0.0163,
      "step": 99
    },
    {
      "epoch": 3.341880341880342,
      "grad_norm": 2.131593020337306,
      "learning_rate": 2.7542115403706067e-06,
      "loss": 0.0193,
      "step": 100
    },
    {
      "epoch": 3.376068376068376,
      "grad_norm": 3.0496672724782465,
      "learning_rate": 2.65795779650105e-06,
      "loss": 0.007,
      "step": 101
    },
    {
      "epoch": 3.41025641025641,
      "grad_norm": 2.4627336519789607,
      "learning_rate": 2.562803416523405e-06,
      "loss": 0.0564,
      "step": 102
    },
    {
      "epoch": 3.4444444444444446,
      "grad_norm": 4.227140605397066,
      "learning_rate": 2.46879306628073e-06,
      "loss": 0.029,
      "step": 103
    },
    {
      "epoch": 3.4786324786324787,
      "grad_norm": 12.74938404068687,
      "learning_rate": 2.375970874603998e-06,
      "loss": 0.004,
      "step": 104
    },
    {
      "epoch": 3.5128205128205128,
      "grad_norm": 1.2426442297293803,
      "learning_rate": 2.2843804125978356e-06,
      "loss": 0.0018,
      "step": 105
    },
    {
      "epoch": 3.547008547008547,
      "grad_norm": 0.49342076095930326,
      "learning_rate": 2.1940646731880887e-06,
      "loss": 0.0039,
      "step": 106
    },
    {
      "epoch": 3.5811965811965814,
      "grad_norm": 2.69930114004212,
      "learning_rate": 2.105066050940758e-06,
      "loss": 0.0052,
      "step": 107
    },
    {
      "epoch": 3.6153846153846154,
      "grad_norm": 0.6214273697117633,
      "learning_rate": 2.0174263221618307e-06,
      "loss": 0.0042,
      "step": 108
    },
    {
      "epoch": 3.6495726495726495,
      "grad_norm": 2.182802172826988,
      "learning_rate": 1.931186625287313e-06,
      "loss": 0.0034,
      "step": 109
    },
    {
      "epoch": 3.683760683760684,
      "grad_norm": 1.2502779472369923,
      "learning_rate": 1.8463874415726918e-06,
      "loss": 0.0057,
      "step": 110
    },
    {
      "epoch": 3.717948717948718,
      "grad_norm": 1.6497554269207777,
      "learning_rate": 1.7630685760908623e-06,
      "loss": 0.0043,
      "step": 111
    },
    {
      "epoch": 3.752136752136752,
      "grad_norm": 1.004233298802305,
      "learning_rate": 1.6812691390474788e-06,
      "loss": 0.0008,
      "step": 112
    },
    {
      "epoch": 3.786324786324786,
      "grad_norm": 3.151655137691081,
      "learning_rate": 1.6010275274224607e-06,
      "loss": 0.0146,
      "step": 113
    },
    {
      "epoch": 3.8205128205128203,
      "grad_norm": 2.663608462515639,
      "learning_rate": 1.5223814069463077e-06,
      "loss": 0.0006,
      "step": 114
    },
    {
      "epoch": 3.8547008547008548,
      "grad_norm": 2.4166544278661175,
      "learning_rate": 1.4453676944196477e-06,
      "loss": 0.0062,
      "step": 115
    },
    {
      "epoch": 3.888888888888889,
      "grad_norm": 0.42828282447090554,
      "learning_rate": 1.370022540384347e-06,
      "loss": 0.002,
      "step": 116
    },
    {
      "epoch": 3.9230769230769234,
      "grad_norm": 0.5035353786016956,
      "learning_rate": 1.296381312154305e-06,
      "loss": 0.0016,
      "step": 117
    },
    {
      "epoch": 3.9572649572649574,
      "grad_norm": 0.8145640782367408,
      "learning_rate": 1.2244785772138972e-06,
      "loss": 0.0019,
      "step": 118
    },
    {
      "epoch": 3.9914529914529915,
      "grad_norm": 0.6555442820439389,
      "learning_rate": 1.1543480869918555e-06,
      "loss": 0.0025,
      "step": 119
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.1211572517989323,
      "learning_rate": 1.0860227610182222e-06,
      "loss": 0.0016,
      "step": 120
    },
    {
      "epoch": 4.034188034188034,
      "grad_norm": 2.837751985589557,
      "learning_rate": 1.0195346714717813e-06,
      "loss": 0.0059,
      "step": 121
    },
    {
      "epoch": 4.068376068376068,
      "grad_norm": 9.72793521179094,
      "learning_rate": 9.549150281252633e-07,
      "loss": 0.0521,
      "step": 122
    },
    {
      "epoch": 4.102564102564102,
      "grad_norm": 0.40180119914672424,
      "learning_rate": 8.921941636953435e-07,
      "loss": 0.0008,
      "step": 123
    },
    {
      "epoch": 4.136752136752137,
      "grad_norm": 0.38580788543923267,
      "learning_rate": 8.314015196043501e-07,
      "loss": 0.0009,
      "step": 124
    },
    {
      "epoch": 4.170940170940171,
      "grad_norm": 4.7944314874118374,
      "learning_rate": 7.725656321603414e-07,
      "loss": 0.0047,
      "step": 125
    },
    {
      "epoch": 4.205128205128205,
      "grad_norm": 1.380709153215471,
      "learning_rate": 7.157141191620548e-07,
      "loss": 0.0022,
      "step": 126
    },
    {
      "epoch": 4.239316239316239,
      "grad_norm": 0.09055290150130697,
      "learning_rate": 6.60873666934993e-07,
      "loss": 0.0004,
      "step": 127
    },
    {
      "epoch": 4.273504273504273,
      "grad_norm": 0.26721903473765835,
      "learning_rate": 6.080700178047688e-07,
      "loss": 0.001,
      "step": 128
    },
    {
      "epoch": 4.3076923076923075,
      "grad_norm": 4.427205308427123,
      "learning_rate": 5.573279580135438e-07,
      "loss": 0.0307,
      "step": 129
    },
    {
      "epoch": 4.3418803418803416,
      "grad_norm": 3.7731219517331533,
      "learning_rate": 5.086713060852788e-07,
      "loss": 0.0196,
      "step": 130
    },
    {
      "epoch": 4.3760683760683765,
      "grad_norm": 0.3206150214839682,
      "learning_rate": 4.6212290164521554e-07,
      "loss": 0.0014,
      "step": 131
    },
    {
      "epoch": 4.410256410256411,
      "grad_norm": 4.3532485442057745,
      "learning_rate": 4.1770459469887003e-07,
      "loss": 0.0309,
      "step": 132
    },
    {
      "epoch": 4.444444444444445,
      "grad_norm": 0.5802295362163845,
      "learning_rate": 3.754372353755559e-07,
      "loss": 0.0023,
      "step": 133
    },
    {
      "epoch": 4.478632478632479,
      "grad_norm": 2.568497803610388,
      "learning_rate": 3.35340664141246e-07,
      "loss": 0.0168,
      "step": 134
    },
    {
      "epoch": 4.512820512820513,
      "grad_norm": 1.1779499885511535,
      "learning_rate": 2.974337024853802e-07,
      "loss": 0.0022,
      "step": 135
    },
    {
      "epoch": 4.547008547008547,
      "grad_norm": 1.3941782756718661,
      "learning_rate": 2.617341440859883e-07,
      "loss": 0.0026,
      "step": 136
    },
    {
      "epoch": 4.581196581196581,
      "grad_norm": 1.6601906514570437,
      "learning_rate": 2.2825874645725942e-07,
      "loss": 0.0017,
      "step": 137
    },
    {
      "epoch": 4.615384615384615,
      "grad_norm": 0.383326235850294,
      "learning_rate": 1.9702322308350675e-07,
      "loss": 0.0015,
      "step": 138
    },
    {
      "epoch": 4.64957264957265,
      "grad_norm": 2.401959029947897,
      "learning_rate": 1.6804223604318825e-07,
      "loss": 0.0067,
      "step": 139
    },
    {
      "epoch": 4.683760683760684,
      "grad_norm": 0.9855732491355061,
      "learning_rate": 1.413293891264722e-07,
      "loss": 0.0036,
      "step": 140
    },
    {
      "epoch": 4.717948717948718,
      "grad_norm": 0.12231214645102874,
      "learning_rate": 1.1689722144956672e-07,
      "loss": 0.0005,
      "step": 141
    },
    {
      "epoch": 4.752136752136752,
      "grad_norm": 2.5878205380346957,
      "learning_rate": 9.475720156880419e-08,
      "loss": 0.0107,
      "step": 142
    },
    {
      "epoch": 4.786324786324786,
      "grad_norm": 0.5782321688283493,
      "learning_rate": 7.491972209725807e-08,
      "loss": 0.0023,
      "step": 143
    },
    {
      "epoch": 4.82051282051282,
      "grad_norm": 0.15708765591840856,
      "learning_rate": 5.739409482640956e-08,
      "loss": 0.0006,
      "step": 144
    },
    {
      "epoch": 4.854700854700854,
      "grad_norm": 0.883976824477256,
      "learning_rate": 4.2188546355153016e-08,
      "loss": 0.0018,
      "step": 145
    },
    {
      "epoch": 4.888888888888889,
      "grad_norm": 5.372037807057783,
      "learning_rate": 2.9310214228202016e-08,
      "loss": 0.0142,
      "step": 146
    },
    {
      "epoch": 4.923076923076923,
      "grad_norm": 1.1060803400734966,
      "learning_rate": 1.8765143585693924e-08,
      "loss": 0.0025,
      "step": 147
    },
    {
      "epoch": 4.957264957264957,
      "grad_norm": 0.40249495145373326,
      "learning_rate": 1.0558284325578038e-08,
      "loss": 0.0013,
      "step": 148
    },
    {
      "epoch": 4.9914529914529915,
      "grad_norm": 0.10203486906721038,
      "learning_rate": 4.69348878011644e-09,
      "loss": 0.0005,
      "step": 149
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 150,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 149,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 29626003253248.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}