{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.22698612862547288,
  "eval_steps": 10,
  "global_step": 90,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "computed_total": 0.45715999603271484,
      "epoch": 0.0025220680958385876,
      "grad_norm": 102.097412109375,
      "kl_loss": 6.5083294309431494e-09,
      "learning_rate": 0.0,
      "loss": 4.3237,
      "sft_loss": 0.45650917291641235,
      "step": 1
    },
    {
      "computed_total": 0.555546760559082,
      "epoch": 0.005044136191677175,
      "grad_norm": 87.49225616455078,
      "kl_loss": 5.4945092919922445e-09,
      "learning_rate": 4.000000000000001e-06,
      "loss": 4.3036,
      "sft_loss": 0.5549973249435425,
      "step": 2
    },
    {
      "computed_total": 0.5587829947471619,
      "epoch": 0.007566204287515763,
      "grad_norm": 64.52155303955078,
      "kl_loss": 3.4872137177899276e-08,
      "learning_rate": 8.000000000000001e-06,
      "loss": 3.5946,
      "sft_loss": 0.5552957653999329,
      "step": 3
    },
    {
      "computed_total": 0.4053139090538025,
      "epoch": 0.01008827238335435,
      "grad_norm": 79.32300567626953,
      "kl_loss": 7.649157964806363e-07,
      "learning_rate": 1.2e-05,
      "loss": 3.3472,
      "sft_loss": 0.3288223147392273,
      "step": 4
    },
    {
      "computed_total": 0.3922015130519867,
      "epoch": 0.012610340479192938,
      "grad_norm": 49.438594818115234,
      "kl_loss": 9.65095068750088e-07,
      "learning_rate": 1.6000000000000003e-05,
      "loss": 3.0682,
      "sft_loss": 0.29569199681282043,
      "step": 5
    },
    {
      "computed_total": 0.3126932382583618,
      "epoch": 0.015132408575031526,
      "grad_norm": 36.26791000366211,
      "kl_loss": 9.634337629904621e-07,
      "learning_rate": 2e-05,
      "loss": 2.7815,
      "sft_loss": 0.21634985506534576,
      "step": 6
    },
    {
      "computed_total": 0.2588472068309784,
      "epoch": 0.017654476670870115,
      "grad_norm": 28.21232795715332,
      "kl_loss": 7.276251494658936e-07,
      "learning_rate": 1.9948979591836737e-05,
      "loss": 2.499,
      "sft_loss": 0.18608468770980835,
      "step": 7
    },
    {
      "computed_total": 0.32910704612731934,
      "epoch": 0.0201765447667087,
      "grad_norm": 24.3712158203125,
      "kl_loss": 8.822735821922834e-07,
      "learning_rate": 1.9897959183673473e-05,
      "loss": 2.4776,
      "sft_loss": 0.24087969958782196,
      "step": 8
    },
    {
      "computed_total": 0.3658309280872345,
      "epoch": 0.02269861286254729,
      "grad_norm": 31.98260498046875,
      "kl_loss": 7.195029070317105e-07,
      "learning_rate": 1.9846938775510205e-05,
      "loss": 2.3718,
      "sft_loss": 0.2938806414604187,
      "step": 9
    },
    {
      "computed_total": 0.22178040444850922,
      "epoch": 0.025220680958385876,
      "grad_norm": 34.0799674987793,
      "kl_loss": 6.166483217384666e-07,
      "learning_rate": 1.979591836734694e-05,
      "loss": 2.4129,
      "sft_loss": 0.1601155698299408,
      "step": 10
    },
    {
      "computed_total": 2.1096696853637695,
      "epoch": 0.025220680958385876,
      "eval_loss": 1.9816868305206299,
      "eval_runtime": 67.6013,
      "eval_samples_per_second": 10.429,
      "eval_steps_per_second": 5.222,
      "kl_loss": 3.9526437944914505e-07,
      "sft_loss": 2.070143222808838,
      "step": 10
    },
    {
      "computed_total": 0.4089103043079376,
      "epoch": 0.027742749054224466,
      "grad_norm": 26.954879760742188,
      "kl_loss": 4.996533107259893e-07,
      "learning_rate": 1.9744897959183677e-05,
      "loss": 2.277,
      "sft_loss": 0.35894498229026794,
      "step": 11
    },
    {
      "computed_total": 0.23354634642601013,
      "epoch": 0.03026481715006305,
      "grad_norm": 18.464841842651367,
      "kl_loss": 5.376580816118803e-07,
      "learning_rate": 1.969387755102041e-05,
      "loss": 2.2583,
      "sft_loss": 0.17978054285049438,
      "step": 12
    },
    {
      "computed_total": 0.30147990584373474,
      "epoch": 0.03278688524590164,
      "grad_norm": 26.595727920532227,
      "kl_loss": 5.573554062721087e-07,
      "learning_rate": 1.9642857142857145e-05,
      "loss": 2.2624,
      "sft_loss": 0.24574437737464905,
      "step": 13
    },
    {
      "computed_total": 0.24872392416000366,
      "epoch": 0.03530895334174023,
      "grad_norm": 16.613079071044922,
      "kl_loss": 4.973371687810868e-07,
      "learning_rate": 1.9591836734693877e-05,
      "loss": 2.1151,
      "sft_loss": 0.19899021089076996,
      "step": 14
    },
    {
      "computed_total": 0.19776779413223267,
      "epoch": 0.03783102143757881,
      "grad_norm": 15.080662727355957,
      "kl_loss": 4.196848806259368e-07,
      "learning_rate": 1.9540816326530613e-05,
      "loss": 2.0897,
      "sft_loss": 0.15579931437969208,
      "step": 15
    },
    {
      "computed_total": 0.284699946641922,
      "epoch": 0.0403530895334174,
      "grad_norm": 11.973522186279297,
      "kl_loss": 4.3317794506947394e-07,
      "learning_rate": 1.948979591836735e-05,
      "loss": 2.216,
      "sft_loss": 0.2413821518421173,
      "step": 16
    },
    {
      "computed_total": 0.21680015325546265,
      "epoch": 0.04287515762925599,
      "grad_norm": 11.934453964233398,
      "kl_loss": 3.477235850368743e-07,
      "learning_rate": 1.9438775510204085e-05,
      "loss": 1.9235,
      "sft_loss": 0.18202780187129974,
      "step": 17
    },
    {
      "computed_total": 0.18643273413181305,
      "epoch": 0.04539722572509458,
      "grad_norm": 10.074520111083984,
      "kl_loss": 2.590618919384724e-07,
      "learning_rate": 1.9387755102040817e-05,
      "loss": 1.8522,
      "sft_loss": 0.16052654385566711,
      "step": 18
    },
    {
      "computed_total": 0.22932234406471252,
      "epoch": 0.04791929382093316,
      "grad_norm": 15.449592590332031,
      "kl_loss": 2.511018522000086e-07,
      "learning_rate": 1.9336734693877553e-05,
      "loss": 1.7982,
      "sft_loss": 0.20421215891838074,
      "step": 19
    },
    {
      "computed_total": 0.29722726345062256,
      "epoch": 0.05044136191677175,
      "grad_norm": 12.395018577575684,
      "kl_loss": 2.6658420892999857e-07,
      "learning_rate": 1.928571428571429e-05,
      "loss": 1.817,
      "sft_loss": 0.27056884765625,
      "step": 20
    },
    {
      "computed_total": 1.6116889715194702,
      "epoch": 0.05044136191677175,
      "eval_loss": 1.5718247890472412,
      "eval_runtime": 69.8078,
      "eval_samples_per_second": 10.099,
      "eval_steps_per_second": 5.057,
      "kl_loss": 3.4212027344437956e-07,
      "sft_loss": 1.577476978302002,
      "step": 20
    },
    {
      "computed_total": 0.1908552497625351,
      "epoch": 0.05296343001261034,
      "grad_norm": 23.406896591186523,
      "kl_loss": 2.7672922442434356e-07,
      "learning_rate": 1.9234693877551024e-05,
      "loss": 1.7865,
      "sft_loss": 0.163182333111763,
      "step": 21
    },
    {
      "computed_total": 0.2098204642534256,
      "epoch": 0.05548549810844893,
      "grad_norm": 12.387871742248535,
      "kl_loss": 1.9662114425500477e-07,
      "learning_rate": 1.9183673469387756e-05,
      "loss": 1.6599,
      "sft_loss": 0.19015835225582123,
      "step": 22
    },
    {
      "computed_total": 0.27859383821487427,
      "epoch": 0.058007566204287514,
      "grad_norm": 17.064393997192383,
      "kl_loss": 2.6223989380014245e-07,
      "learning_rate": 1.9132653061224492e-05,
      "loss": 1.7114,
      "sft_loss": 0.25236985087394714,
      "step": 23
    },
    {
      "computed_total": 0.2661365270614624,
      "epoch": 0.0605296343001261,
      "grad_norm": 14.536602973937988,
      "kl_loss": 3.302961033568863e-07,
      "learning_rate": 1.9081632653061225e-05,
      "loss": 1.7381,
      "sft_loss": 0.23310692608356476,
      "step": 24
    },
    {
      "computed_total": 0.21774020791053772,
      "epoch": 0.06305170239596469,
      "grad_norm": 10.612632751464844,
      "kl_loss": 2.2093004758971801e-07,
      "learning_rate": 1.903061224489796e-05,
      "loss": 1.6493,
      "sft_loss": 0.195647194981575,
      "step": 25
    },
    {
      "computed_total": 0.21532197296619415,
      "epoch": 0.06557377049180328,
      "grad_norm": 13.36475944519043,
      "kl_loss": 1.9050941091336426e-07,
      "learning_rate": 1.8979591836734696e-05,
      "loss": 1.65,
      "sft_loss": 0.19627103209495544,
      "step": 26
    },
    {
      "computed_total": 0.21690420806407928,
      "epoch": 0.06809583858764187,
      "grad_norm": 13.064225196838379,
      "kl_loss": 2.1687725393348956e-07,
      "learning_rate": 1.892857142857143e-05,
      "loss": 1.7006,
      "sft_loss": 0.19521647691726685,
      "step": 27
    },
    {
      "computed_total": 0.19565589725971222,
      "epoch": 0.07061790668348046,
      "grad_norm": 12.320195198059082,
      "kl_loss": 2.17976335648018e-07,
      "learning_rate": 1.8877551020408164e-05,
      "loss": 1.6652,
      "sft_loss": 0.17385827004909515,
      "step": 28
    },
    {
      "computed_total": 0.27470341324806213,
      "epoch": 0.07313997477931904,
      "grad_norm": 11.224297523498535,
      "kl_loss": 2.520193618238409e-07,
      "learning_rate": 1.88265306122449e-05,
      "loss": 1.6883,
      "sft_loss": 0.24950148165225983,
      "step": 29
    },
    {
      "computed_total": 0.1886218637228012,
      "epoch": 0.07566204287515763,
      "grad_norm": 14.4179048538208,
      "kl_loss": 3.655400746538362e-07,
      "learning_rate": 1.8775510204081636e-05,
      "loss": 1.7232,
      "sft_loss": 0.1520678550004959,
      "step": 30
    },
    {
      "computed_total": 1.3417655229568481,
      "epoch": 0.07566204287515763,
      "eval_loss": 1.4369149208068848,
      "eval_runtime": 67.0852,
      "eval_samples_per_second": 10.509,
      "eval_steps_per_second": 5.262,
      "kl_loss": 2.1116991888447956e-07,
      "sft_loss": 1.3206485509872437,
      "step": 30
    },
    {
      "computed_total": 0.19783087074756622,
      "epoch": 0.07818411097099622,
      "grad_norm": 14.451069831848145,
      "kl_loss": 1.4020483263266215e-07,
      "learning_rate": 1.8724489795918368e-05,
      "loss": 1.6225,
      "sft_loss": 0.18381038308143616,
      "step": 31
    },
    {
      "computed_total": 0.2664089798927307,
      "epoch": 0.0807061790668348,
      "grad_norm": 10.6675443649292,
      "kl_loss": 2.558473113367654e-07,
      "learning_rate": 1.8673469387755104e-05,
      "loss": 1.6656,
      "sft_loss": 0.24082423746585846,
      "step": 32
    },
    {
      "computed_total": 0.16216596961021423,
      "epoch": 0.0832282471626734,
      "grad_norm": 10.89920711517334,
      "kl_loss": 3.102907157881418e-07,
      "learning_rate": 1.862244897959184e-05,
      "loss": 1.7422,
      "sft_loss": 0.13113689422607422,
      "step": 33
    },
    {
      "computed_total": 0.18052271008491516,
      "epoch": 0.08575031525851198,
      "grad_norm": 10.312054634094238,
      "kl_loss": 2.5258012215090275e-07,
      "learning_rate": 1.8571428571428575e-05,
      "loss": 1.5284,
      "sft_loss": 0.1552647054195404,
      "step": 34
    },
    {
      "computed_total": 0.13403812050819397,
      "epoch": 0.08827238335435057,
      "grad_norm": 9.937405586242676,
      "kl_loss": 1.9699928088812158e-07,
      "learning_rate": 1.8520408163265307e-05,
      "loss": 1.6276,
      "sft_loss": 0.11433819681406021,
      "step": 35
    },
    {
      "computed_total": 0.21145765483379364,
      "epoch": 0.09079445145018916,
      "grad_norm": 10.810340881347656,
      "kl_loss": 2.5251029001083225e-07,
      "learning_rate": 1.8469387755102043e-05,
      "loss": 1.5674,
      "sft_loss": 0.1862066239118576,
      "step": 36
    },
    {
      "computed_total": 0.28667497634887695,
      "epoch": 0.09331651954602774,
      "grad_norm": 13.322245597839355,
      "kl_loss": 2.524957380956039e-07,
      "learning_rate": 1.8418367346938776e-05,
      "loss": 1.6924,
      "sft_loss": 0.2614254057407379,
      "step": 37
    },
    {
      "computed_total": 0.17698872089385986,
      "epoch": 0.09583858764186633,
      "grad_norm": 12.67447566986084,
      "kl_loss": 1.8715130067903374e-07,
      "learning_rate": 1.836734693877551e-05,
      "loss": 1.6326,
      "sft_loss": 0.1582735925912857,
      "step": 38
    },
    {
      "computed_total": 0.218657985329628,
      "epoch": 0.09836065573770492,
      "grad_norm": 10.07801342010498,
      "kl_loss": 2.1152843032723467e-07,
      "learning_rate": 1.8316326530612247e-05,
      "loss": 1.5745,
      "sft_loss": 0.1975051462650299,
      "step": 39
    },
    {
      "computed_total": 0.1585373878479004,
      "epoch": 0.1008827238335435,
      "grad_norm": 10.709808349609375,
      "kl_loss": 1.5567046318665234e-07,
      "learning_rate": 1.826530612244898e-05,
      "loss": 1.4899,
      "sft_loss": 0.14297033846378326,
      "step": 40
    },
    {
      "computed_total": 1.3272018432617188,
      "epoch": 0.1008827238335435,
      "eval_loss": 1.3992165327072144,
      "eval_runtime": 67.3288,
      "eval_samples_per_second": 10.471,
      "eval_steps_per_second": 5.243,
      "kl_loss": 1.587673352787533e-07,
      "sft_loss": 1.3113250732421875,
      "step": 40
    },
    {
      "computed_total": 0.16187940537929535,
      "epoch": 0.1034047919293821,
      "grad_norm": 12.045276641845703,
      "kl_loss": 1.6272514358206536e-07,
      "learning_rate": 1.8214285714285715e-05,
      "loss": 1.4241,
      "sft_loss": 0.1456068903207779,
      "step": 41
    },
    {
      "computed_total": 0.16236107051372528,
      "epoch": 0.10592686002522068,
      "grad_norm": 9.8399658203125,
      "kl_loss": 1.7566540577718115e-07,
      "learning_rate": 1.816326530612245e-05,
      "loss": 1.557,
      "sft_loss": 0.1447945237159729,
      "step": 42
    },
    {
      "computed_total": 0.15767036378383636,
      "epoch": 0.10844892812105927,
      "grad_norm": 12.114518165588379,
      "kl_loss": 1.7348739334011043e-07,
      "learning_rate": 1.8112244897959187e-05,
      "loss": 1.583,
      "sft_loss": 0.14032162725925446,
      "step": 43
    },
    {
      "computed_total": 0.09624896198511124,
      "epoch": 0.11097099621689786,
      "grad_norm": 9.963354110717773,
      "kl_loss": 1.7239732130747143e-07,
      "learning_rate": 1.806122448979592e-05,
      "loss": 1.4214,
      "sft_loss": 0.07900922745466232,
      "step": 44
    },
    {
      "computed_total": 0.2713112533092499,
      "epoch": 0.11349306431273644,
      "grad_norm": 10.126675605773926,
      "kl_loss": 1.4448913532305596e-07,
      "learning_rate": 1.8010204081632655e-05,
      "loss": 1.5783,
      "sft_loss": 0.2568623423576355,
      "step": 45
    },
    {
      "computed_total": 0.1786537766456604,
      "epoch": 0.11601513240857503,
      "grad_norm": 9.506120681762695,
      "kl_loss": 1.1718055503706637e-07,
      "learning_rate": 1.795918367346939e-05,
      "loss": 1.5218,
      "sft_loss": 0.1669357270002365,
      "step": 46
    },
    {
      "computed_total": 0.15938213467597961,
      "epoch": 0.11853720050441362,
      "grad_norm": 11.758268356323242,
      "kl_loss": 1.7643755256813165e-07,
      "learning_rate": 1.7908163265306123e-05,
      "loss": 1.479,
      "sft_loss": 0.1417383849620819,
      "step": 47
    },
    {
      "computed_total": 0.15835891664028168,
      "epoch": 0.1210592686002522,
      "grad_norm": 10.227970123291016,
      "kl_loss": 1.3144477861715131e-07,
      "learning_rate": 1.785714285714286e-05,
      "loss": 1.5222,
      "sft_loss": 0.14521443843841553,
      "step": 48
    },
    {
      "computed_total": 0.1753772348165512,
      "epoch": 0.1235813366960908,
      "grad_norm": 9.49471378326416,
      "kl_loss": 1.2467818066852487e-07,
      "learning_rate": 1.780612244897959e-05,
      "loss": 1.4884,
      "sft_loss": 0.16290941834449768,
      "step": 49
    },
    {
      "computed_total": 0.14545822143554688,
      "epoch": 0.12610340479192939,
      "grad_norm": 9.987504005432129,
      "kl_loss": 1.3850477387222782e-07,
      "learning_rate": 1.7755102040816327e-05,
      "loss": 1.5382,
      "sft_loss": 0.13160774111747742,
      "step": 50
    },
    {
      "computed_total": 1.2876274585723877,
      "epoch": 0.12610340479192939,
      "eval_loss": 1.3761128187179565,
      "eval_runtime": 65.8814,
      "eval_samples_per_second": 10.701,
      "eval_steps_per_second": 5.358,
      "kl_loss": 1.1736652538729686e-07,
      "sft_loss": 1.275890827178955,
      "step": 50
    },
    {
      "computed_total": 0.13572491705417633,
      "epoch": 0.12862547288776796,
      "grad_norm": 8.282198905944824,
      "kl_loss": 1.2732888876598736e-07,
      "learning_rate": 1.7704081632653062e-05,
      "loss": 1.4672,
      "sft_loss": 0.12299202382564545,
      "step": 51
    },
    {
      "computed_total": 0.2022746205329895,
      "epoch": 0.13114754098360656,
      "grad_norm": 8.850266456604004,
      "kl_loss": 9.734989703247265e-08,
      "learning_rate": 1.7653061224489798e-05,
      "loss": 1.4807,
      "sft_loss": 0.19253963232040405,
      "step": 52
    },
    {
      "computed_total": 0.2428017556667328,
      "epoch": 0.13366960907944514,
      "grad_norm": 7.4850568771362305,
      "kl_loss": 1.472172925787163e-07,
      "learning_rate": 1.760204081632653e-05,
      "loss": 1.4415,
      "sft_loss": 0.22808003425598145,
      "step": 53
    },
    {
      "computed_total": 0.22275952994823456,
      "epoch": 0.13619167717528374,
      "grad_norm": 9.871981620788574,
      "kl_loss": 1.480846094636945e-07,
      "learning_rate": 1.7551020408163266e-05,
      "loss": 1.6587,
      "sft_loss": 0.20795106887817383,
      "step": 54
    },
    {
      "computed_total": 0.1835995763540268,
      "epoch": 0.13871374527112232,
      "grad_norm": 8.844654083251953,
      "kl_loss": 1.2977692165350163e-07,
      "learning_rate": 1.7500000000000002e-05,
      "loss": 1.5753,
      "sft_loss": 0.17062188684940338,
      "step": 55
    },
    {
      "computed_total": 0.19904766976833344,
      "epoch": 0.14123581336696092,
      "grad_norm": 7.038422584533691,
      "kl_loss": 1.1099445629270122e-07,
      "learning_rate": 1.7448979591836738e-05,
      "loss": 1.4909,
      "sft_loss": 0.18794822692871094,
      "step": 56
    },
    {
      "computed_total": 0.14884871244430542,
      "epoch": 0.1437578814627995,
      "grad_norm": 7.760448455810547,
      "kl_loss": 1.0769576164193495e-07,
      "learning_rate": 1.7397959183673473e-05,
      "loss": 1.5988,
      "sft_loss": 0.13807913661003113,
      "step": 57
    },
    {
      "computed_total": 0.22467643022537231,
      "epoch": 0.14627994955863807,
      "grad_norm": 8.861936569213867,
      "kl_loss": 1.3167736767627503e-07,
      "learning_rate": 1.7346938775510206e-05,
      "loss": 1.4759,
      "sft_loss": 0.21150869131088257,
      "step": 58
    },
    {
      "computed_total": 0.1402980089187622,
      "epoch": 0.14880201765447668,
      "grad_norm": 8.389931678771973,
      "kl_loss": 1.2574119523378613e-07,
      "learning_rate": 1.729591836734694e-05,
      "loss": 1.4478,
      "sft_loss": 0.12772388756275177,
      "step": 59
    },
    {
      "computed_total": 0.2214464396238327,
      "epoch": 0.15132408575031525,
      "grad_norm": 11.03455924987793,
      "kl_loss": 1.229836925631389e-07,
      "learning_rate": 1.7244897959183674e-05,
      "loss": 1.542,
      "sft_loss": 0.20914806425571442,
      "step": 60
    },
    {
      "computed_total": 1.2934421300888062,
      "epoch": 0.15132408575031525,
      "eval_loss": 1.3563388586044312,
      "eval_runtime": 66.808,
      "eval_samples_per_second": 10.553,
      "eval_steps_per_second": 5.284,
      "kl_loss": 1.0278420603526683e-07,
      "sft_loss": 1.2831636667251587,
      "step": 60
    },
    {
      "computed_total": 0.19582681357860565,
      "epoch": 0.15384615384615385,
      "grad_norm": 8.09454345703125,
      "kl_loss": 1.0761530688796483e-07,
      "learning_rate": 1.719387755102041e-05,
      "loss": 1.4979,
      "sft_loss": 0.18506528437137604,
      "step": 61
    },
    {
      "computed_total": 0.15016137063503265,
      "epoch": 0.15636822194199243,
      "grad_norm": 6.767925262451172,
      "kl_loss": 9.817436819048453e-08,
      "learning_rate": 1.7142857142857142e-05,
      "loss": 1.3421,
      "sft_loss": 0.14034393429756165,
      "step": 62
    },
    {
      "computed_total": 0.23630164563655853,
      "epoch": 0.15889029003783103,
      "grad_norm": 6.9385552406311035,
      "kl_loss": 1.2005209271137574e-07,
      "learning_rate": 1.7091836734693878e-05,
      "loss": 1.4917,
      "sft_loss": 0.224296435713768,
      "step": 63
    },
    {
      "computed_total": 0.23200978338718414,
      "epoch": 0.1614123581336696,
      "grad_norm": 7.448540210723877,
      "kl_loss": 8.879172952447334e-08,
      "learning_rate": 1.7040816326530613e-05,
      "loss": 1.4776,
      "sft_loss": 0.22313061356544495,
      "step": 64
    },
    {
      "computed_total": 0.16305182874202728,
      "epoch": 0.16393442622950818,
      "grad_norm": 7.1528401374816895,
      "kl_loss": 9.930875677355289e-08,
      "learning_rate": 1.698979591836735e-05,
      "loss": 1.3943,
      "sft_loss": 0.1531209498643875,
      "step": 65
    },
    {
      "computed_total": 0.14623767137527466,
      "epoch": 0.1664564943253468,
      "grad_norm": 7.3966474533081055,
      "kl_loss": 7.404720037129664e-08,
      "learning_rate": 1.6938775510204085e-05,
      "loss": 1.5426,
      "sft_loss": 0.1388329565525055,
      "step": 66
    },
    {
      "computed_total": 0.14071843028068542,
      "epoch": 0.16897856242118536,
      "grad_norm": 7.82870626449585,
      "kl_loss": 9.07349146928027e-08,
      "learning_rate": 1.6887755102040817e-05,
      "loss": 1.4386,
      "sft_loss": 0.13164493441581726,
      "step": 67
    },
    {
      "computed_total": 0.14310967922210693,
      "epoch": 0.17150063051702397,
      "grad_norm": 6.757770538330078,
      "kl_loss": 7.598376328132872e-08,
      "learning_rate": 1.6836734693877553e-05,
      "loss": 1.3732,
      "sft_loss": 0.13551130890846252,
      "step": 68
    },
    {
      "computed_total": 0.14132973551750183,
      "epoch": 0.17402269861286254,
      "grad_norm": 8.344841957092285,
      "kl_loss": 1.0905234404390285e-07,
      "learning_rate": 1.678571428571429e-05,
      "loss": 1.4286,
      "sft_loss": 0.13042449951171875,
      "step": 69
    },
    {
      "computed_total": 0.13716241717338562,
      "epoch": 0.17654476670870115,
      "grad_norm": 7.953212738037109,
      "kl_loss": 7.554955061550572e-08,
      "learning_rate": 1.673469387755102e-05,
      "loss": 1.3187,
      "sft_loss": 0.12960746884346008,
      "step": 70
    },
    {
      "computed_total": 1.2660552263259888,
      "epoch": 0.17654476670870115,
      "eval_loss": 1.348509669303894,
      "eval_runtime": 66.6809,
      "eval_samples_per_second": 10.573,
      "eval_steps_per_second": 5.294,
      "kl_loss": 5.885167198016461e-08,
      "sft_loss": 1.2601701021194458,
      "step": 70
    },
    {
      "computed_total": 0.21419619023799896,
      "epoch": 0.17906683480453972,
      "grad_norm": 7.022055625915527,
      "kl_loss": 5.5535913645599067e-08,
      "learning_rate": 1.6683673469387757e-05,
      "loss": 1.3962,
      "sft_loss": 0.2086426019668579,
      "step": 71
    },
    {
      "computed_total": 0.1429305225610733,
      "epoch": 0.18158890290037832,
      "grad_norm": 6.416485786437988,
      "kl_loss": 5.6032408934925115e-08,
      "learning_rate": 1.6632653061224492e-05,
      "loss": 1.4386,
      "sft_loss": 0.13732728362083435,
      "step": 72
    },
    {
      "computed_total": 0.1595781445503235,
      "epoch": 0.1841109709962169,
      "grad_norm": 6.244477272033691,
      "kl_loss": 1.0954471463264781e-07,
      "learning_rate": 1.6581632653061225e-05,
      "loss": 1.4295,
      "sft_loss": 0.14862367510795593,
      "step": 73
    },
    {
      "computed_total": 0.1830446571111679,
      "epoch": 0.18663303909205547,
      "grad_norm": 6.70966100692749,
      "kl_loss": 1.011455665889116e-07,
      "learning_rate": 1.653061224489796e-05,
      "loss": 1.5022,
      "sft_loss": 0.17293010652065277,
      "step": 74
    },
    {
      "computed_total": 0.13835778832435608,
      "epoch": 0.18915510718789408,
      "grad_norm": 7.2905378341674805,
      "kl_loss": 1.1240518205113403e-07,
      "learning_rate": 1.6479591836734696e-05,
      "loss": 1.4917,
      "sft_loss": 0.12711727619171143,
      "step": 75
    },
    {
      "computed_total": 0.17067813873291016,
      "epoch": 0.19167717528373265,
      "grad_norm": 6.712996482849121,
      "kl_loss": 7.433935422795912e-08,
      "learning_rate": 1.642857142857143e-05,
      "loss": 1.3446,
      "sft_loss": 0.16324420273303986,
      "step": 76
    },
    {
      "computed_total": 0.1600562483072281,
      "epoch": 0.19419924337957126,
      "grad_norm": 6.522639274597168,
      "kl_loss": 6.566633459215154e-08,
      "learning_rate": 1.6377551020408164e-05,
      "loss": 1.2858,
      "sft_loss": 0.1534896194934845,
      "step": 77
    },
    {
      "computed_total": 0.15458209812641144,
      "epoch": 0.19672131147540983,
      "grad_norm": 6.488280773162842,
      "kl_loss": 8.56703010754245e-08,
      "learning_rate": 1.63265306122449e-05,
      "loss": 1.3759,
      "sft_loss": 0.14601506292819977,
      "step": 78
    },
    {
      "computed_total": 0.14832951128482819,
      "epoch": 0.19924337957124844,
      "grad_norm": 7.430500507354736,
      "kl_loss": 1.0060924182653253e-07,
      "learning_rate": 1.6275510204081636e-05,
      "loss": 1.1762,
      "sft_loss": 0.1382685899734497,
      "step": 79
    },
    {
      "computed_total": 0.1753160059452057,
      "epoch": 0.201765447667087,
      "grad_norm": 8.094470024108887,
      "kl_loss": 8.996708089625827e-08,
      "learning_rate": 1.6224489795918368e-05,
      "loss": 1.5172,
      "sft_loss": 0.16631929576396942,
      "step": 80
    },
    {
      "computed_total": 1.2994036674499512,
      "epoch": 0.201765447667087,
      "eval_loss": 1.349037766456604,
      "eval_runtime": 64.527,
      "eval_samples_per_second": 10.926,
      "eval_steps_per_second": 5.471,
      "kl_loss": 7.120447520492235e-08,
      "sft_loss": 1.2922831773757935,
      "step": 80
    },
    {
      "computed_total": 0.1462080031633377,
      "epoch": 0.2042875157629256,
      "grad_norm": 7.84616756439209,
      "kl_loss": 9.43423970056756e-08,
      "learning_rate": 1.6173469387755104e-05,
      "loss": 1.3986,
      "sft_loss": 0.13677376508712769,
      "step": 81
    },
    {
      "computed_total": 0.20616410672664642,
      "epoch": 0.2068095838587642,
      "grad_norm": 7.06512975692749,
      "kl_loss": 7.940719370935767e-08,
      "learning_rate": 1.612244897959184e-05,
      "loss": 1.5103,
      "sft_loss": 0.19822338223457336,
      "step": 82
    },
    {
      "computed_total": 0.2241806834936142,
      "epoch": 0.20933165195460277,
      "grad_norm": 6.878040313720703,
      "kl_loss": 1.1082394024697351e-07,
      "learning_rate": 1.6071428571428572e-05,
      "loss": 1.4422,
      "sft_loss": 0.21309828758239746,
      "step": 83
    },
    {
      "computed_total": 0.19030708074569702,
      "epoch": 0.21185372005044137,
      "grad_norm": 6.8972344398498535,
      "kl_loss": 7.487692244012578e-08,
      "learning_rate": 1.6020408163265308e-05,
      "loss": 1.4802,
      "sft_loss": 0.18281938135623932,
      "step": 84
    },
    {
      "computed_total": 0.18855753540992737,
      "epoch": 0.21437578814627994,
      "grad_norm": 6.072068214416504,
      "kl_loss": 6.224269810672922e-08,
      "learning_rate": 1.596938775510204e-05,
      "loss": 1.3092,
      "sft_loss": 0.18233326077461243,
      "step": 85
    },
    {
      "computed_total": 0.17298777401447296,
      "epoch": 0.21689785624211855,
      "grad_norm": 6.168684005737305,
      "kl_loss": 7.623211928375895e-08,
      "learning_rate": 1.5918367346938776e-05,
      "loss": 1.4069,
      "sft_loss": 0.1653645634651184,
      "step": 86
    },
    {
      "computed_total": 0.19384431838989258,
      "epoch": 0.21941992433795712,
      "grad_norm": 6.2668352127075195,
      "kl_loss": 6.887150760803706e-08,
      "learning_rate": 1.586734693877551e-05,
      "loss": 1.4861,
      "sft_loss": 0.18695716559886932,
      "step": 87
    },
    {
      "computed_total": 0.23912306129932404,
      "epoch": 0.22194199243379573,
      "grad_norm": 5.836297035217285,
      "kl_loss": 7.468336349347737e-08,
      "learning_rate": 1.5816326530612247e-05,
      "loss": 1.4451,
      "sft_loss": 0.23165471851825714,
      "step": 88
    },
    {
      "computed_total": 0.1360023021697998,
      "epoch": 0.2244640605296343,
      "grad_norm": 7.015989303588867,
      "kl_loss": 6.417322140350734e-08,
      "learning_rate": 1.576530612244898e-05,
      "loss": 1.3913,
      "sft_loss": 0.12958498299121857,
      "step": 89
    },
    {
      "computed_total": 0.12024354189634323,
      "epoch": 0.22698612862547288,
      "grad_norm": 6.264023780822754,
      "kl_loss": 7.851058825281143e-08,
      "learning_rate": 1.5714285714285715e-05,
      "loss": 1.3727,
      "sft_loss": 0.11239248514175415,
      "step": 90
    },
    {
      "computed_total": 1.277982473373413,
      "epoch": 0.22698612862547288,
      "eval_loss": 1.3299487829208374,
      "eval_runtime": 63.6992,
      "eval_samples_per_second": 11.068,
      "eval_steps_per_second": 5.542,
      "kl_loss": 4.547840504187661e-08,
      "sft_loss": 1.2734346389770508,
      "step": 90
    }
  ],
  "logging_steps": 1,
  "max_steps": 397,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 10,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.651854088636416e+16,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}