{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 465,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0064516129032258064,
      "grad_norm": 1.8496088327765188,
      "learning_rate": 4.255319148936171e-06,
      "loss": 0.9425,
      "step": 1
    },
    {
      "epoch": 0.012903225806451613,
      "grad_norm": 1.464006720275199,
      "learning_rate": 8.510638297872341e-06,
      "loss": 0.6748,
      "step": 2
    },
    {
      "epoch": 0.01935483870967742,
      "grad_norm": 1.195864584909173,
      "learning_rate": 1.2765957446808511e-05,
      "loss": 0.5742,
      "step": 3
    },
    {
      "epoch": 0.025806451612903226,
      "grad_norm": 1.4795865650630229,
      "learning_rate": 1.7021276595744682e-05,
      "loss": 0.7095,
      "step": 4
    },
    {
      "epoch": 0.03225806451612903,
      "grad_norm": 1.3747220363223371,
      "learning_rate": 2.1276595744680852e-05,
      "loss": 0.6648,
      "step": 5
    },
    {
      "epoch": 0.03870967741935484,
      "grad_norm": 1.301406005018193,
      "learning_rate": 2.5531914893617022e-05,
      "loss": 0.6255,
      "step": 6
    },
    {
      "epoch": 0.04516129032258064,
      "grad_norm": 1.0187483254912535,
      "learning_rate": 2.9787234042553192e-05,
      "loss": 0.4435,
      "step": 7
    },
    {
      "epoch": 0.05161290322580645,
      "grad_norm": 1.0700513081990584,
      "learning_rate": 3.4042553191489365e-05,
      "loss": 0.4263,
      "step": 8
    },
    {
      "epoch": 0.05806451612903226,
      "grad_norm": 0.7712554505163577,
      "learning_rate": 3.829787234042553e-05,
      "loss": 0.2511,
      "step": 9
    },
    {
      "epoch": 0.06451612903225806,
      "grad_norm": 0.7232887438916186,
      "learning_rate": 4.2553191489361704e-05,
      "loss": 0.1781,
      "step": 10
    },
    {
      "epoch": 0.07096774193548387,
      "grad_norm": 0.8652521068380956,
      "learning_rate": 4.680851063829788e-05,
      "loss": 0.1926,
      "step": 11
    },
    {
      "epoch": 0.07741935483870968,
      "grad_norm": 0.5403017280043989,
      "learning_rate": 5.1063829787234044e-05,
      "loss": 0.1125,
      "step": 12
    },
    {
      "epoch": 0.08387096774193549,
      "grad_norm": 0.7070226880901681,
      "learning_rate": 5.531914893617022e-05,
      "loss": 0.101,
      "step": 13
    },
    {
      "epoch": 0.09032258064516129,
      "grad_norm": 0.2650073598784149,
      "learning_rate": 5.9574468085106384e-05,
      "loss": 0.1058,
      "step": 14
    },
    {
      "epoch": 0.0967741935483871,
      "grad_norm": 0.131557187963195,
      "learning_rate": 6.382978723404256e-05,
      "loss": 0.1113,
      "step": 15
    },
    {
      "epoch": 0.1032258064516129,
      "grad_norm": 0.09751111122397599,
      "learning_rate": 6.808510638297873e-05,
      "loss": 0.0834,
      "step": 16
    },
    {
      "epoch": 0.10967741935483871,
      "grad_norm": 0.10433139009644296,
      "learning_rate": 7.23404255319149e-05,
      "loss": 0.0961,
      "step": 17
    },
    {
      "epoch": 0.11612903225806452,
      "grad_norm": 0.1020133297592548,
      "learning_rate": 7.659574468085106e-05,
      "loss": 0.097,
      "step": 18
    },
    {
      "epoch": 0.12258064516129032,
      "grad_norm": 0.07145294843882412,
      "learning_rate": 8.085106382978723e-05,
      "loss": 0.0916,
      "step": 19
    },
    {
      "epoch": 0.12903225806451613,
      "grad_norm": 0.08981924516933591,
      "learning_rate": 8.510638297872341e-05,
      "loss": 0.101,
      "step": 20
    },
    {
      "epoch": 0.13548387096774195,
      "grad_norm": 0.0681524232238508,
      "learning_rate": 8.936170212765958e-05,
      "loss": 0.0736,
      "step": 21
    },
    {
      "epoch": 0.14193548387096774,
      "grad_norm": 0.0981002108730681,
      "learning_rate": 9.361702127659576e-05,
      "loss": 0.1091,
      "step": 22
    },
    {
      "epoch": 0.14838709677419354,
      "grad_norm": 0.11001940230271333,
      "learning_rate": 9.787234042553192e-05,
      "loss": 0.0991,
      "step": 23
    },
    {
      "epoch": 0.15483870967741936,
      "grad_norm": 0.06614356059405047,
      "learning_rate": 0.00010212765957446809,
      "loss": 0.0591,
      "step": 24
    },
    {
      "epoch": 0.16129032258064516,
      "grad_norm": 0.05937801367835722,
      "learning_rate": 0.00010638297872340425,
      "loss": 0.0633,
      "step": 25
    },
    {
      "epoch": 0.16774193548387098,
      "grad_norm": 0.07592110532595596,
      "learning_rate": 0.00011063829787234043,
      "loss": 0.095,
      "step": 26
    },
    {
      "epoch": 0.17419354838709677,
      "grad_norm": 0.07287278501663387,
      "learning_rate": 0.00011489361702127661,
      "loss": 0.09,
      "step": 27
    },
    {
      "epoch": 0.18064516129032257,
      "grad_norm": 0.07270501072864007,
      "learning_rate": 0.00011914893617021277,
      "loss": 0.0764,
      "step": 28
    },
    {
      "epoch": 0.1870967741935484,
      "grad_norm": 0.08202627980248832,
      "learning_rate": 0.00012340425531914893,
      "loss": 0.0841,
      "step": 29
    },
    {
      "epoch": 0.1935483870967742,
      "grad_norm": 0.10309403095606792,
      "learning_rate": 0.00012765957446808513,
      "loss": 0.0989,
      "step": 30
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.08068773698268775,
      "learning_rate": 0.00013191489361702127,
      "loss": 0.0844,
      "step": 31
    },
    {
      "epoch": 0.2064516129032258,
      "grad_norm": 0.06817985516792972,
      "learning_rate": 0.00013617021276595746,
      "loss": 0.0793,
      "step": 32
    },
    {
      "epoch": 0.2129032258064516,
      "grad_norm": 0.06656651609987188,
      "learning_rate": 0.00014042553191489363,
      "loss": 0.0666,
      "step": 33
    },
    {
      "epoch": 0.21935483870967742,
      "grad_norm": 0.09101885574528384,
      "learning_rate": 0.0001446808510638298,
      "loss": 0.0856,
      "step": 34
    },
    {
      "epoch": 0.22580645161290322,
      "grad_norm": 0.11244141382637182,
      "learning_rate": 0.00014893617021276596,
      "loss": 0.1076,
      "step": 35
    },
    {
      "epoch": 0.23225806451612904,
      "grad_norm": 0.07641687845159767,
      "learning_rate": 0.00015319148936170213,
      "loss": 0.0678,
      "step": 36
    },
    {
      "epoch": 0.23870967741935484,
      "grad_norm": 0.08682080497813398,
      "learning_rate": 0.00015744680851063832,
      "loss": 0.0859,
      "step": 37
    },
    {
      "epoch": 0.24516129032258063,
      "grad_norm": 0.07050044811298162,
      "learning_rate": 0.00016170212765957446,
      "loss": 0.0746,
      "step": 38
    },
    {
      "epoch": 0.25161290322580643,
      "grad_norm": 0.10418569203255569,
      "learning_rate": 0.00016595744680851065,
      "loss": 0.1028,
      "step": 39
    },
    {
      "epoch": 0.25806451612903225,
      "grad_norm": 0.14211262039099226,
      "learning_rate": 0.00017021276595744682,
      "loss": 0.101,
      "step": 40
    },
    {
      "epoch": 0.2645161290322581,
      "grad_norm": 0.065192138802457,
      "learning_rate": 0.00017446808510638298,
      "loss": 0.0591,
      "step": 41
    },
    {
      "epoch": 0.2709677419354839,
      "grad_norm": 0.08691739931276934,
      "learning_rate": 0.00017872340425531915,
      "loss": 0.0816,
      "step": 42
    },
    {
      "epoch": 0.27741935483870966,
      "grad_norm": 0.09038372940039405,
      "learning_rate": 0.00018297872340425532,
      "loss": 0.0923,
      "step": 43
    },
    {
      "epoch": 0.2838709677419355,
      "grad_norm": 0.07006837751282566,
      "learning_rate": 0.0001872340425531915,
      "loss": 0.0781,
      "step": 44
    },
    {
      "epoch": 0.2903225806451613,
      "grad_norm": 0.08875720051203487,
      "learning_rate": 0.00019148936170212768,
      "loss": 0.0672,
      "step": 45
    },
    {
      "epoch": 0.2967741935483871,
      "grad_norm": 0.07220322606120361,
      "learning_rate": 0.00019574468085106384,
      "loss": 0.0853,
      "step": 46
    },
    {
      "epoch": 0.3032258064516129,
      "grad_norm": 0.07992244290240197,
      "learning_rate": 0.0002,
      "loss": 0.0846,
      "step": 47
    },
    {
      "epoch": 0.3096774193548387,
      "grad_norm": 0.09917015506018256,
      "learning_rate": 0.0001999971756719333,
      "loss": 0.0934,
      "step": 48
    },
    {
      "epoch": 0.3161290322580645,
      "grad_norm": 0.16671519547672903,
      "learning_rate": 0.00019998870284726968,
      "loss": 0.1228,
      "step": 49
    },
    {
      "epoch": 0.3225806451612903,
      "grad_norm": 0.09107913287594624,
      "learning_rate": 0.00019997458200460993,
      "loss": 0.0847,
      "step": 50
    },
    {
      "epoch": 0.32903225806451614,
      "grad_norm": 0.13795927374096006,
      "learning_rate": 0.00019995481394159188,
      "loss": 0.0904,
      "step": 51
    },
    {
      "epoch": 0.33548387096774196,
      "grad_norm": 0.11415771239086442,
      "learning_rate": 0.0001999293997748454,
      "loss": 0.0782,
      "step": 52
    },
    {
      "epoch": 0.3419354838709677,
      "grad_norm": 0.08413093725717319,
      "learning_rate": 0.00019989834093992945,
      "loss": 0.0818,
      "step": 53
    },
    {
      "epoch": 0.34838709677419355,
      "grad_norm": 0.06942673059881972,
      "learning_rate": 0.00019986163919125075,
      "loss": 0.0746,
      "step": 54
    },
    {
      "epoch": 0.3548387096774194,
      "grad_norm": 0.06779490919317474,
      "learning_rate": 0.00019981929660196492,
      "loss": 0.0704,
      "step": 55
    },
    {
      "epoch": 0.36129032258064514,
      "grad_norm": 0.0705554071643389,
      "learning_rate": 0.0001997713155638592,
      "loss": 0.0718,
      "step": 56
    },
    {
      "epoch": 0.36774193548387096,
      "grad_norm": 0.06094574022360802,
      "learning_rate": 0.00019971769878721743,
      "loss": 0.0615,
      "step": 57
    },
    {
      "epoch": 0.3741935483870968,
      "grad_norm": 0.08174462699234748,
      "learning_rate": 0.000199658449300667,
      "loss": 0.0772,
      "step": 58
    },
    {
      "epoch": 0.38064516129032255,
      "grad_norm": 0.15240897489693175,
      "learning_rate": 0.00019959357045100764,
      "loss": 0.1136,
      "step": 59
    },
    {
      "epoch": 0.3870967741935484,
      "grad_norm": 0.12657819665542305,
      "learning_rate": 0.00019952306590302247,
      "loss": 0.1066,
      "step": 60
    },
    {
      "epoch": 0.3935483870967742,
      "grad_norm": 0.06847012277175818,
      "learning_rate": 0.00019944693963927092,
      "loss": 0.0809,
      "step": 61
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.06995242717454851,
      "learning_rate": 0.00019936519595986394,
      "loss": 0.0732,
      "step": 62
    },
    {
      "epoch": 0.4064516129032258,
      "grad_norm": 0.08400016052467398,
      "learning_rate": 0.00019927783948222084,
      "loss": 0.0838,
      "step": 63
    },
    {
      "epoch": 0.4129032258064516,
      "grad_norm": 0.06773211017815356,
      "learning_rate": 0.00019918487514080865,
      "loss": 0.0786,
      "step": 64
    },
    {
      "epoch": 0.41935483870967744,
      "grad_norm": 0.05982739234697708,
      "learning_rate": 0.00019908630818686338,
      "loss": 0.0643,
      "step": 65
    },
    {
      "epoch": 0.4258064516129032,
      "grad_norm": 0.07075319425828745,
      "learning_rate": 0.0001989821441880933,
      "loss": 0.0765,
      "step": 66
    },
    {
      "epoch": 0.432258064516129,
      "grad_norm": 0.08372038666535586,
      "learning_rate": 0.00019887238902836448,
      "loss": 0.0669,
      "step": 67
    },
    {
      "epoch": 0.43870967741935485,
      "grad_norm": 0.13201060477850413,
      "learning_rate": 0.00019875704890736853,
      "loss": 0.1129,
      "step": 68
    },
    {
      "epoch": 0.44516129032258067,
      "grad_norm": 0.15774619872849757,
      "learning_rate": 0.00019863613034027224,
      "loss": 0.1031,
      "step": 69
    },
    {
      "epoch": 0.45161290322580644,
      "grad_norm": 0.09239020494576224,
      "learning_rate": 0.0001985096401573497,
      "loss": 0.0827,
      "step": 70
    },
    {
      "epoch": 0.45806451612903226,
      "grad_norm": 0.08710163284471478,
      "learning_rate": 0.00019837758550359636,
      "loss": 0.0898,
      "step": 71
    },
    {
      "epoch": 0.4645161290322581,
      "grad_norm": 0.09970425528976597,
      "learning_rate": 0.0001982399738383255,
      "loss": 0.0671,
      "step": 72
    },
    {
      "epoch": 0.47096774193548385,
      "grad_norm": 0.11084567093271626,
      "learning_rate": 0.00019809681293474693,
      "loss": 0.0978,
      "step": 73
    },
    {
      "epoch": 0.4774193548387097,
      "grad_norm": 0.08583289505322891,
      "learning_rate": 0.0001979481108795278,
      "loss": 0.0851,
      "step": 74
    },
    {
      "epoch": 0.4838709677419355,
      "grad_norm": 0.08349769120006256,
      "learning_rate": 0.00019779387607233586,
      "loss": 0.096,
      "step": 75
    },
    {
      "epoch": 0.49032258064516127,
      "grad_norm": 0.08822285487830493,
      "learning_rate": 0.00019763411722536502,
      "loss": 0.0797,
      "step": 76
    },
    {
      "epoch": 0.4967741935483871,
      "grad_norm": 0.08213973307253318,
      "learning_rate": 0.00019746884336284317,
      "loss": 0.0798,
      "step": 77
    },
    {
      "epoch": 0.5032258064516129,
      "grad_norm": 0.13109929672968634,
      "learning_rate": 0.00019729806382052248,
      "loss": 0.1037,
      "step": 78
    },
    {
      "epoch": 0.5096774193548387,
      "grad_norm": 0.09245218089323069,
      "learning_rate": 0.00019712178824515212,
      "loss": 0.106,
      "step": 79
    },
    {
      "epoch": 0.5161290322580645,
      "grad_norm": 0.08163307640075247,
      "learning_rate": 0.00019694002659393305,
      "loss": 0.062,
      "step": 80
    },
    {
      "epoch": 0.5225806451612903,
      "grad_norm": 0.05950648712895881,
      "learning_rate": 0.00019675278913395606,
      "loss": 0.0597,
      "step": 81
    },
    {
      "epoch": 0.5290322580645161,
      "grad_norm": 0.06692605001225288,
      "learning_rate": 0.0001965600864416213,
      "loss": 0.0822,
      "step": 82
    },
    {
      "epoch": 0.535483870967742,
      "grad_norm": 0.08343609115222997,
      "learning_rate": 0.00019636192940204134,
      "loss": 0.0795,
      "step": 83
    },
    {
      "epoch": 0.5419354838709678,
      "grad_norm": 0.06269743932265057,
      "learning_rate": 0.00019615832920842586,
      "loss": 0.0723,
      "step": 84
    },
    {
      "epoch": 0.5483870967741935,
      "grad_norm": 0.08562276972837224,
      "learning_rate": 0.00019594929736144976,
      "loss": 0.0815,
      "step": 85
    },
    {
      "epoch": 0.5548387096774193,
      "grad_norm": 0.15766855113117675,
      "learning_rate": 0.0001957348456686032,
      "loss": 0.116,
      "step": 86
    },
    {
      "epoch": 0.5612903225806452,
      "grad_norm": 0.09597817519413815,
      "learning_rate": 0.00019551498624352496,
      "loss": 0.0748,
      "step": 87
    },
    {
      "epoch": 0.567741935483871,
      "grad_norm": 0.0781406533241935,
      "learning_rate": 0.00019528973150531787,
      "loss": 0.0767,
      "step": 88
    },
    {
      "epoch": 0.5741935483870968,
      "grad_norm": 0.06202378155483603,
      "learning_rate": 0.00019505909417784754,
      "loss": 0.0824,
      "step": 89
    },
    {
      "epoch": 0.5806451612903226,
      "grad_norm": 0.07132758636096014,
      "learning_rate": 0.00019482308728902356,
      "loss": 0.0711,
      "step": 90
    },
    {
      "epoch": 0.5870967741935483,
      "grad_norm": 0.08324703218713819,
      "learning_rate": 0.00019458172417006347,
      "loss": 0.0908,
      "step": 91
    },
    {
      "epoch": 0.5935483870967742,
      "grad_norm": 0.07808035645363601,
      "learning_rate": 0.00019433501845473995,
      "loss": 0.0798,
      "step": 92
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.09922957657013898,
      "learning_rate": 0.00019408298407861042,
      "loss": 0.0938,
      "step": 93
    },
    {
      "epoch": 0.6064516129032258,
      "grad_norm": 0.11239213177839262,
      "learning_rate": 0.00019382563527823026,
      "loss": 0.0958,
      "step": 94
    },
    {
      "epoch": 0.6129032258064516,
      "grad_norm": 0.11717297126046855,
      "learning_rate": 0.00019356298659034817,
      "loss": 0.1095,
      "step": 95
    },
    {
      "epoch": 0.6193548387096774,
      "grad_norm": 0.058447164828301125,
      "learning_rate": 0.00019329505285108542,
      "loss": 0.0656,
      "step": 96
    },
    {
      "epoch": 0.6258064516129033,
      "grad_norm": 0.06781770204369802,
      "learning_rate": 0.00019302184919509755,
      "loss": 0.0731,
      "step": 97
    },
    {
      "epoch": 0.632258064516129,
      "grad_norm": 0.07151161214033872,
      "learning_rate": 0.00019274339105471971,
      "loss": 0.0715,
      "step": 98
    },
    {
      "epoch": 0.6387096774193548,
      "grad_norm": 0.07604008792488949,
      "learning_rate": 0.00019245969415909465,
      "loss": 0.0742,
      "step": 99
    },
    {
      "epoch": 0.6451612903225806,
      "grad_norm": 0.10649337867077462,
      "learning_rate": 0.00019217077453328449,
      "loss": 0.0888,
      "step": 100
    },
    {
      "epoch": 0.6516129032258065,
      "grad_norm": 0.06658172686723186,
      "learning_rate": 0.0001918766484973654,
      "loss": 0.0596,
      "step": 101
    },
    {
      "epoch": 0.6580645161290323,
      "grad_norm": 0.07673513301158036,
      "learning_rate": 0.00019157733266550575,
      "loss": 0.0807,
      "step": 102
    },
    {
      "epoch": 0.6645161290322581,
      "grad_norm": 0.10741827222890765,
      "learning_rate": 0.0001912728439450276,
      "loss": 0.0957,
      "step": 103
    },
    {
      "epoch": 0.6709677419354839,
      "grad_norm": 0.07943997555150896,
      "learning_rate": 0.00019096319953545185,
      "loss": 0.0732,
      "step": 104
    },
    {
      "epoch": 0.6774193548387096,
      "grad_norm": 0.09157580363747424,
      "learning_rate": 0.0001906484169275263,
      "loss": 0.0964,
      "step": 105
    },
    {
      "epoch": 0.6838709677419355,
      "grad_norm": 0.07483334970504105,
      "learning_rate": 0.00019032851390223812,
      "loss": 0.0813,
      "step": 106
    },
    {
      "epoch": 0.6903225806451613,
      "grad_norm": 0.08004302588923752,
      "learning_rate": 0.00019000350852980909,
      "loss": 0.0792,
      "step": 107
    },
    {
      "epoch": 0.6967741935483871,
      "grad_norm": 0.10990654224626453,
      "learning_rate": 0.00018967341916867518,
      "loss": 0.096,
      "step": 108
    },
    {
      "epoch": 0.7032258064516129,
      "grad_norm": 0.16146890814677625,
      "learning_rate": 0.00018933826446444933,
      "loss": 0.1212,
      "step": 109
    },
    {
      "epoch": 0.7096774193548387,
      "grad_norm": 0.0695644520734436,
      "learning_rate": 0.0001889980633488683,
      "loss": 0.072,
      "step": 110
    },
    {
      "epoch": 0.7161290322580646,
      "grad_norm": 0.07364701758879359,
      "learning_rate": 0.00018865283503872324,
      "loss": 0.0831,
      "step": 111
    },
    {
      "epoch": 0.7225806451612903,
      "grad_norm": 0.06683550033086046,
      "learning_rate": 0.00018830259903477426,
      "loss": 0.0801,
      "step": 112
    },
    {
      "epoch": 0.7290322580645161,
      "grad_norm": 0.12012664311858479,
      "learning_rate": 0.0001879473751206489,
      "loss": 0.1029,
      "step": 113
    },
    {
      "epoch": 0.7354838709677419,
      "grad_norm": 0.09146802828659024,
      "learning_rate": 0.0001875871833617246,
      "loss": 0.0894,
      "step": 114
    },
    {
      "epoch": 0.7419354838709677,
      "grad_norm": 0.07579863274704472,
      "learning_rate": 0.0001872220441039952,
      "loss": 0.0713,
      "step": 115
    },
    {
      "epoch": 0.7483870967741936,
      "grad_norm": 0.07086039028499559,
      "learning_rate": 0.0001868519779729218,
      "loss": 0.0731,
      "step": 116
    },
    {
      "epoch": 0.7548387096774194,
      "grad_norm": 0.07163280313926258,
      "learning_rate": 0.0001864770058722676,
      "loss": 0.0727,
      "step": 117
    },
    {
      "epoch": 0.7612903225806451,
      "grad_norm": 0.09459308338109298,
      "learning_rate": 0.00018609714898291718,
      "loss": 0.074,
      "step": 118
    },
    {
      "epoch": 0.7677419354838709,
      "grad_norm": 0.07703917957047136,
      "learning_rate": 0.00018571242876167996,
      "loss": 0.08,
      "step": 119
    },
    {
      "epoch": 0.7741935483870968,
      "grad_norm": 0.09836589908180506,
      "learning_rate": 0.0001853228669400784,
      "loss": 0.0922,
      "step": 120
    },
    {
      "epoch": 0.7806451612903226,
      "grad_norm": 0.09387479622355775,
      "learning_rate": 0.00018492848552312014,
      "loss": 0.1033,
      "step": 121
    },
    {
      "epoch": 0.7870967741935484,
      "grad_norm": 0.12284239686014672,
      "learning_rate": 0.00018452930678805536,
      "loss": 0.0999,
      "step": 122
    },
    {
      "epoch": 0.7935483870967742,
      "grad_norm": 0.09213360621981392,
      "learning_rate": 0.00018412535328311814,
      "loss": 0.0953,
      "step": 123
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.09837327897834967,
      "learning_rate": 0.00018371664782625287,
      "loss": 0.1015,
      "step": 124
    },
    {
      "epoch": 0.8064516129032258,
      "grad_norm": 0.11241908687558025,
      "learning_rate": 0.00018330321350382544,
      "loss": 0.0871,
      "step": 125
    },
    {
      "epoch": 0.8129032258064516,
      "grad_norm": 0.10016314695009264,
      "learning_rate": 0.00018288507366931905,
      "loss": 0.1017,
      "step": 126
    },
    {
      "epoch": 0.8193548387096774,
      "grad_norm": 0.08602103688908509,
      "learning_rate": 0.00018246225194201517,
      "loss": 0.0812,
      "step": 127
    },
    {
      "epoch": 0.8258064516129032,
      "grad_norm": 0.12818530303124728,
      "learning_rate": 0.00018203477220565912,
      "loss": 0.1021,
      "step": 128
    },
    {
      "epoch": 0.832258064516129,
      "grad_norm": 0.10135096833159363,
      "learning_rate": 0.00018160265860711134,
      "loss": 0.1006,
      "step": 129
    },
    {
      "epoch": 0.8387096774193549,
      "grad_norm": 0.06634065993081388,
      "learning_rate": 0.00018116593555498307,
      "loss": 0.0771,
      "step": 130
    },
    {
      "epoch": 0.8451612903225807,
      "grad_norm": 0.08489626818724719,
      "learning_rate": 0.0001807246277182578,
      "loss": 0.0786,
      "step": 131
    },
    {
      "epoch": 0.8516129032258064,
      "grad_norm": 0.08888467555852377,
      "learning_rate": 0.0001802787600248977,
      "loss": 0.0991,
      "step": 132
    },
    {
      "epoch": 0.8580645161290322,
      "grad_norm": 0.10636042561644678,
      "learning_rate": 0.0001798283576604356,
      "loss": 0.1011,
      "step": 133
    },
    {
      "epoch": 0.864516129032258,
      "grad_norm": 0.07381978120350188,
      "learning_rate": 0.0001793734460665523,
      "loss": 0.0942,
      "step": 134
    },
    {
      "epoch": 0.8709677419354839,
      "grad_norm": 0.09141476973335388,
      "learning_rate": 0.00017891405093963938,
      "loss": 0.1003,
      "step": 135
    },
    {
      "epoch": 0.8774193548387097,
      "grad_norm": 0.08895622256143133,
      "learning_rate": 0.0001784501982293479,
      "loss": 0.0852,
      "step": 136
    },
    {
      "epoch": 0.8838709677419355,
      "grad_norm": 0.08177526392628556,
      "learning_rate": 0.00017798191413712243,
      "loss": 0.084,
      "step": 137
    },
    {
      "epoch": 0.8903225806451613,
      "grad_norm": 0.07208124492736653,
      "learning_rate": 0.0001775092251147211,
      "loss": 0.0765,
      "step": 138
    },
    {
      "epoch": 0.896774193548387,
      "grad_norm": 0.06475169326738762,
      "learning_rate": 0.0001770321578627213,
      "loss": 0.0677,
      "step": 139
    },
    {
      "epoch": 0.9032258064516129,
      "grad_norm": 0.12805303578104438,
      "learning_rate": 0.00017655073932901168,
      "loss": 0.1159,
      "step": 140
    },
    {
      "epoch": 0.9096774193548387,
      "grad_norm": 0.09062553054321057,
      "learning_rate": 0.0001760649967072697,
      "loss": 0.1001,
      "step": 141
    },
    {
      "epoch": 0.9161290322580645,
      "grad_norm": 0.09471211412797043,
      "learning_rate": 0.00017557495743542585,
      "loss": 0.0984,
      "step": 142
    },
    {
      "epoch": 0.9225806451612903,
      "grad_norm": 0.07126884163211589,
      "learning_rate": 0.00017508064919411344,
      "loss": 0.0675,
      "step": 143
    },
    {
      "epoch": 0.9290322580645162,
      "grad_norm": 0.06754737396081824,
      "learning_rate": 0.00017458209990510527,
      "loss": 0.0782,
      "step": 144
    },
    {
      "epoch": 0.9354838709677419,
      "grad_norm": 0.09114175386188283,
      "learning_rate": 0.00017407933772973637,
      "loss": 0.0955,
      "step": 145
    },
    {
      "epoch": 0.9419354838709677,
      "grad_norm": 0.09779279853308924,
      "learning_rate": 0.00017357239106731317,
      "loss": 0.1061,
      "step": 146
    },
    {
      "epoch": 0.9483870967741935,
      "grad_norm": 0.09679111364819645,
      "learning_rate": 0.00017306128855350942,
      "loss": 0.0942,
      "step": 147
    },
    {
      "epoch": 0.9548387096774194,
      "grad_norm": 0.1619352621242093,
      "learning_rate": 0.0001725460590587486,
      "loss": 0.1215,
      "step": 148
    },
    {
      "epoch": 0.9612903225806452,
      "grad_norm": 0.07218712567651438,
      "learning_rate": 0.00017202673168657318,
      "loss": 0.0716,
      "step": 149
    },
    {
      "epoch": 0.967741935483871,
      "grad_norm": 0.0768978967349938,
      "learning_rate": 0.0001715033357720006,
      "loss": 0.0921,
      "step": 150
    },
    {
      "epoch": 0.9741935483870968,
      "grad_norm": 0.0716182862543501,
      "learning_rate": 0.00017097590087986633,
      "loss": 0.069,
      "step": 151
    },
    {
      "epoch": 0.9806451612903225,
      "grad_norm": 0.09641833742826841,
      "learning_rate": 0.00017044445680315372,
      "loss": 0.0975,
      "step": 152
    },
    {
      "epoch": 0.9870967741935484,
      "grad_norm": 0.1015814477912073,
      "learning_rate": 0.00016990903356131124,
      "loss": 0.0935,
      "step": 153
    },
    {
      "epoch": 0.9935483870967742,
      "grad_norm": 0.07050697207500581,
      "learning_rate": 0.00016936966139855663,
      "loss": 0.0728,
      "step": 154
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.12410114043097807,
      "learning_rate": 0.00016882637078216868,
      "loss": 0.0918,
      "step": 155
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.08952951431274414,
      "eval_runtime": 27.6637,
      "eval_samples_per_second": 4.735,
      "eval_steps_per_second": 0.615,
      "step": 155
    },
    {
      "epoch": 1.0064516129032257,
      "grad_norm": 0.05504498458319807,
      "learning_rate": 0.0001682791924007661,
      "loss": 0.0464,
      "step": 156
    },
    {
      "epoch": 1.0129032258064516,
      "grad_norm": 0.06037793847523705,
      "learning_rate": 0.00016772815716257412,
      "loss": 0.0569,
      "step": 157
    },
    {
      "epoch": 1.0193548387096774,
      "grad_norm": 0.06519500890305997,
      "learning_rate": 0.0001671732961936785,
      "loss": 0.0785,
      "step": 158
    },
    {
      "epoch": 1.0258064516129033,
      "grad_norm": 0.06362711429608407,
      "learning_rate": 0.00016661464083626734,
      "loss": 0.0492,
      "step": 159
    },
    {
      "epoch": 1.032258064516129,
      "grad_norm": 0.059790795449393114,
      "learning_rate": 0.00016605222264686086,
      "loss": 0.0511,
      "step": 160
    },
    {
      "epoch": 1.038709677419355,
      "grad_norm": 0.062409758409412354,
      "learning_rate": 0.00016548607339452853,
      "loss": 0.0561,
      "step": 161
    },
    {
      "epoch": 1.0451612903225806,
      "grad_norm": 0.05212609330281915,
      "learning_rate": 0.00016491622505909482,
      "loss": 0.0482,
      "step": 162
    },
    {
      "epoch": 1.0516129032258064,
      "grad_norm": 0.07852785833231216,
      "learning_rate": 0.00016434270982933273,
      "loss": 0.0632,
      "step": 163
    },
    {
      "epoch": 1.0580645161290323,
      "grad_norm": 0.0844199396392266,
      "learning_rate": 0.0001637655601011454,
      "loss": 0.0634,
      "step": 164
    },
    {
      "epoch": 1.064516129032258,
      "grad_norm": 0.07478417013950381,
      "learning_rate": 0.00016318480847573642,
      "loss": 0.0705,
      "step": 165
    },
    {
      "epoch": 1.070967741935484,
      "grad_norm": 0.14677307886341195,
      "learning_rate": 0.00016260048775776804,
      "loss": 0.0707,
      "step": 166
    },
    {
      "epoch": 1.0774193548387097,
      "grad_norm": 0.11159415413369748,
      "learning_rate": 0.00016201263095350833,
      "loss": 0.0552,
      "step": 167
    },
    {
      "epoch": 1.0838709677419356,
      "grad_norm": 0.12431222248668082,
      "learning_rate": 0.0001614212712689668,
      "loss": 0.0521,
      "step": 168
    },
    {
      "epoch": 1.0903225806451613,
      "grad_norm": 0.08991938158139547,
      "learning_rate": 0.00016082644210801844,
      "loss": 0.0454,
      "step": 169
    },
    {
      "epoch": 1.096774193548387,
      "grad_norm": 0.08464763972774585,
      "learning_rate": 0.00016022817707051724,
      "loss": 0.0622,
      "step": 170
    },
    {
      "epoch": 1.103225806451613,
      "grad_norm": 0.12349871251730524,
      "learning_rate": 0.00015962650995039783,
      "loss": 0.0704,
      "step": 171
    },
    {
      "epoch": 1.1096774193548387,
      "grad_norm": 0.1309389973485375,
      "learning_rate": 0.00015902147473376694,
      "loss": 0.0711,
      "step": 172
    },
    {
      "epoch": 1.1161290322580646,
      "grad_norm": 0.06816581304698818,
      "learning_rate": 0.00015841310559698343,
      "loss": 0.0566,
      "step": 173
    },
    {
      "epoch": 1.1225806451612903,
      "grad_norm": 0.0628939350533328,
      "learning_rate": 0.0001578014369047279,
      "loss": 0.0515,
      "step": 174
    },
    {
      "epoch": 1.129032258064516,
      "grad_norm": 0.0859769654461784,
      "learning_rate": 0.00015718650320806142,
      "loss": 0.0666,
      "step": 175
    },
    {
      "epoch": 1.135483870967742,
      "grad_norm": 0.09102699765206346,
      "learning_rate": 0.00015656833924247398,
      "loss": 0.0617,
      "step": 176
    },
    {
      "epoch": 1.1419354838709677,
      "grad_norm": 0.0765372107398411,
      "learning_rate": 0.00015594697992592232,
      "loss": 0.0488,
      "step": 177
    },
    {
      "epoch": 1.1483870967741936,
      "grad_norm": 0.06699310113728325,
      "learning_rate": 0.00015532246035685756,
      "loss": 0.0573,
      "step": 178
    },
    {
      "epoch": 1.1548387096774193,
      "grad_norm": 0.08417544962609826,
      "learning_rate": 0.00015469481581224272,
      "loss": 0.0636,
      "step": 179
    },
    {
      "epoch": 1.1612903225806452,
      "grad_norm": 0.1377222230063782,
      "learning_rate": 0.00015406408174555976,
      "loss": 0.0675,
      "step": 180
    },
    {
      "epoch": 1.167741935483871,
      "grad_norm": 0.11567828351658707,
      "learning_rate": 0.0001534302937848073,
      "loss": 0.0694,
      "step": 181
    },
    {
      "epoch": 1.1741935483870969,
      "grad_norm": 0.07435102139494404,
      "learning_rate": 0.00015279348773048786,
      "loss": 0.0619,
      "step": 182
    },
    {
      "epoch": 1.1806451612903226,
      "grad_norm": 0.16550300010492666,
      "learning_rate": 0.00015215369955358566,
      "loss": 0.0863,
      "step": 183
    },
    {
      "epoch": 1.1870967741935483,
      "grad_norm": 0.12805046103524215,
      "learning_rate": 0.0001515109653935348,
      "loss": 0.0787,
      "step": 184
    },
    {
      "epoch": 1.1935483870967742,
      "grad_norm": 0.07302921753172271,
      "learning_rate": 0.00015086532155617784,
      "loss": 0.0616,
      "step": 185
    },
    {
      "epoch": 1.2,
      "grad_norm": 0.06678947449139945,
      "learning_rate": 0.00015021680451171498,
      "loss": 0.049,
      "step": 186
    },
    {
      "epoch": 1.206451612903226,
      "grad_norm": 0.07635737870842979,
      "learning_rate": 0.00014956545089264407,
      "loss": 0.06,
      "step": 187
    },
    {
      "epoch": 1.2129032258064516,
      "grad_norm": 0.09502911856234499,
      "learning_rate": 0.0001489112974916912,
      "loss": 0.0717,
      "step": 188
    },
    {
      "epoch": 1.2193548387096773,
      "grad_norm": 0.10739599373969563,
      "learning_rate": 0.00014825438125973264,
      "loss": 0.0695,
      "step": 189
    },
    {
      "epoch": 1.2258064516129032,
      "grad_norm": 0.07169666556967867,
      "learning_rate": 0.00014759473930370736,
      "loss": 0.0457,
      "step": 190
    },
    {
      "epoch": 1.232258064516129,
      "grad_norm": 0.0938431700551618,
      "learning_rate": 0.0001469324088845212,
      "loss": 0.0512,
      "step": 191
    },
    {
      "epoch": 1.238709677419355,
      "grad_norm": 0.08141491065850508,
      "learning_rate": 0.00014626742741494206,
      "loss": 0.0636,
      "step": 192
    },
    {
      "epoch": 1.2451612903225806,
      "grad_norm": 0.07323985237455122,
      "learning_rate": 0.00014559983245748638,
      "loss": 0.0547,
      "step": 193
    },
    {
      "epoch": 1.2516129032258063,
      "grad_norm": 0.07650175159597551,
      "learning_rate": 0.00014492966172229777,
      "loss": 0.0589,
      "step": 194
    },
    {
      "epoch": 1.2580645161290323,
      "grad_norm": 0.06313573515968769,
      "learning_rate": 0.00014425695306501658,
      "loss": 0.047,
      "step": 195
    },
    {
      "epoch": 1.2645161290322582,
      "grad_norm": 0.08819272769189934,
      "learning_rate": 0.00014358174448464154,
      "loss": 0.0578,
      "step": 196
    },
    {
      "epoch": 1.270967741935484,
      "grad_norm": 0.1258683666392954,
      "learning_rate": 0.00014290407412138366,
      "loss": 0.0694,
      "step": 197
    },
    {
      "epoch": 1.2774193548387096,
      "grad_norm": 0.11414016288586973,
      "learning_rate": 0.00014222398025451135,
      "loss": 0.0604,
      "step": 198
    },
    {
      "epoch": 1.2838709677419355,
      "grad_norm": 0.11137244002349718,
      "learning_rate": 0.00014154150130018866,
      "loss": 0.0756,
      "step": 199
    },
    {
      "epoch": 1.2903225806451613,
      "grad_norm": 0.07153259224988148,
      "learning_rate": 0.0001408566758093048,
      "loss": 0.0538,
      "step": 200
    },
    {
      "epoch": 1.2967741935483872,
      "grad_norm": 0.09526746882665606,
      "learning_rate": 0.00014016954246529696,
      "loss": 0.0494,
      "step": 201
    },
    {
      "epoch": 1.303225806451613,
      "grad_norm": 0.08649485022046847,
      "learning_rate": 0.00013948014008196487,
      "loss": 0.0553,
      "step": 202
    },
    {
      "epoch": 1.3096774193548386,
      "grad_norm": 0.1050606321924827,
      "learning_rate": 0.0001387885076012785,
      "loss": 0.0726,
      "step": 203
    },
    {
      "epoch": 1.3161290322580645,
      "grad_norm": 0.12765438794483042,
      "learning_rate": 0.00013809468409117846,
      "loss": 0.084,
      "step": 204
    },
    {
      "epoch": 1.3225806451612903,
      "grad_norm": 0.0778070748347935,
      "learning_rate": 0.00013739870874336898,
      "loss": 0.0549,
      "step": 205
    },
    {
      "epoch": 1.3290322580645162,
      "grad_norm": 0.08829501946025207,
      "learning_rate": 0.00013670062087110422,
      "loss": 0.054,
      "step": 206
    },
    {
      "epoch": 1.335483870967742,
      "grad_norm": 0.09272743824756946,
      "learning_rate": 0.00013600045990696762,
      "loss": 0.0533,
      "step": 207
    },
    {
      "epoch": 1.3419354838709676,
      "grad_norm": 0.09752395648449025,
      "learning_rate": 0.0001352982654006444,
      "loss": 0.0682,
      "step": 208
    },
    {
      "epoch": 1.3483870967741935,
      "grad_norm": 0.0787282052904063,
      "learning_rate": 0.00013459407701668763,
      "loss": 0.061,
      "step": 209
    },
    {
      "epoch": 1.3548387096774195,
      "grad_norm": 0.09965818076523,
      "learning_rate": 0.00013388793453227767,
      "loss": 0.0708,
      "step": 210
    },
    {
      "epoch": 1.3612903225806452,
      "grad_norm": 0.08510816398839181,
      "learning_rate": 0.0001331798778349752,
      "loss": 0.0612,
      "step": 211
    },
    {
      "epoch": 1.367741935483871,
      "grad_norm": 0.13415837375152884,
      "learning_rate": 0.00013246994692046836,
      "loss": 0.0589,
      "step": 212
    },
    {
      "epoch": 1.3741935483870968,
      "grad_norm": 0.10509841268799006,
      "learning_rate": 0.00013175818189031327,
      "loss": 0.0771,
      "step": 213
    },
    {
      "epoch": 1.3806451612903226,
      "grad_norm": 0.10611783573942027,
      "learning_rate": 0.00013104462294966896,
      "loss": 0.0772,
      "step": 214
    },
    {
      "epoch": 1.3870967741935485,
      "grad_norm": 0.10192686799206796,
      "learning_rate": 0.00013032931040502627,
      "loss": 0.0679,
      "step": 215
    },
    {
      "epoch": 1.3935483870967742,
      "grad_norm": 0.0734691690524852,
      "learning_rate": 0.00012961228466193116,
      "loss": 0.0536,
      "step": 216
    },
    {
      "epoch": 1.4,
      "grad_norm": 0.08238968788193186,
      "learning_rate": 0.00012889358622270223,
      "loss": 0.0469,
      "step": 217
    },
    {
      "epoch": 1.4064516129032258,
      "grad_norm": 0.107308663713204,
      "learning_rate": 0.00012817325568414297,
      "loss": 0.0472,
      "step": 218
    },
    {
      "epoch": 1.4129032258064516,
      "grad_norm": 0.08899570061026768,
      "learning_rate": 0.00012745133373524853,
      "loss": 0.0621,
      "step": 219
    },
    {
      "epoch": 1.4193548387096775,
      "grad_norm": 0.09218740393476688,
      "learning_rate": 0.0001267278611549073,
      "loss": 0.0449,
      "step": 220
    },
    {
      "epoch": 1.4258064516129032,
      "grad_norm": 0.17826386309202488,
      "learning_rate": 0.00012600287880959763,
      "loss": 0.0617,
      "step": 221
    },
    {
      "epoch": 1.432258064516129,
      "grad_norm": 0.10232360588380689,
      "learning_rate": 0.0001252764276510792,
      "loss": 0.0636,
      "step": 222
    },
    {
      "epoch": 1.4387096774193548,
      "grad_norm": 0.07733076855697436,
      "learning_rate": 0.00012454854871407994,
      "loss": 0.0524,
      "step": 223
    },
    {
      "epoch": 1.4451612903225808,
      "grad_norm": 0.1006676835885761,
      "learning_rate": 0.00012381928311397806,
      "loss": 0.057,
      "step": 224
    },
    {
      "epoch": 1.4516129032258065,
      "grad_norm": 0.080575053343473,
      "learning_rate": 0.0001230886720444796,
      "loss": 0.0663,
      "step": 225
    },
    {
      "epoch": 1.4580645161290322,
      "grad_norm": 0.1246560932677115,
      "learning_rate": 0.00012235675677529158,
      "loss": 0.0652,
      "step": 226
    },
    {
      "epoch": 1.4645161290322581,
      "grad_norm": 0.10632779078755829,
      "learning_rate": 0.00012162357864979072,
      "loss": 0.065,
      "step": 227
    },
    {
      "epoch": 1.4709677419354839,
      "grad_norm": 0.095566354892694,
      "learning_rate": 0.00012088917908268821,
      "loss": 0.0704,
      "step": 228
    },
    {
      "epoch": 1.4774193548387098,
      "grad_norm": 0.1101864261825649,
      "learning_rate": 0.00012015359955769021,
      "loss": 0.069,
      "step": 229
    },
    {
      "epoch": 1.4838709677419355,
      "grad_norm": 0.08091933130202467,
      "learning_rate": 0.00011941688162515467,
      "loss": 0.0513,
      "step": 230
    },
    {
      "epoch": 1.4903225806451612,
      "grad_norm": 0.12491882510784613,
      "learning_rate": 0.00011867906689974428,
      "loss": 0.0481,
      "step": 231
    },
    {
      "epoch": 1.4967741935483871,
      "grad_norm": 0.13567923305564322,
      "learning_rate": 0.00011794019705807584,
      "loss": 0.0689,
      "step": 232
    },
    {
      "epoch": 1.5032258064516129,
      "grad_norm": 0.15901158424809853,
      "learning_rate": 0.00011720031383636585,
      "loss": 0.0647,
      "step": 233
    },
    {
      "epoch": 1.5096774193548388,
      "grad_norm": 0.10314424853828444,
      "learning_rate": 0.00011645945902807341,
      "loss": 0.0596,
      "step": 234
    },
    {
      "epoch": 1.5161290322580645,
      "grad_norm": 0.08231901481570739,
      "learning_rate": 0.00011571767448153901,
      "loss": 0.0425,
      "step": 235
    },
    {
      "epoch": 1.5225806451612902,
      "grad_norm": 0.24497759631459304,
      "learning_rate": 0.00011497500209762102,
      "loss": 0.0861,
      "step": 236
    },
    {
      "epoch": 1.5290322580645161,
      "grad_norm": 0.08192342540576594,
      "learning_rate": 0.00011423148382732853,
      "loss": 0.037,
      "step": 237
    },
    {
      "epoch": 1.535483870967742,
      "grad_norm": 0.086893577170591,
      "learning_rate": 0.00011348716166945195,
      "loss": 0.053,
      "step": 238
    },
    {
      "epoch": 1.5419354838709678,
      "grad_norm": 0.10993883303276072,
      "learning_rate": 0.0001127420776681905,
      "loss": 0.0584,
      "step": 239
    },
    {
      "epoch": 1.5483870967741935,
      "grad_norm": 0.08488031387161675,
      "learning_rate": 0.00011199627391077732,
      "loss": 0.0618,
      "step": 240
    },
    {
      "epoch": 1.5548387096774192,
      "grad_norm": 0.059796713120191955,
      "learning_rate": 0.00011124979252510208,
      "loss": 0.0468,
      "step": 241
    },
    {
      "epoch": 1.5612903225806452,
      "grad_norm": 0.1333781764848824,
      "learning_rate": 0.0001105026756773314,
      "loss": 0.0435,
      "step": 242
    },
    {
      "epoch": 1.567741935483871,
      "grad_norm": 0.13677532337123358,
      "learning_rate": 0.00010975496556952682,
      "loss": 0.0822,
      "step": 243
    },
    {
      "epoch": 1.5741935483870968,
      "grad_norm": 0.07575070108462173,
      "learning_rate": 0.00010900670443726135,
      "loss": 0.0518,
      "step": 244
    },
    {
      "epoch": 1.5806451612903225,
      "grad_norm": 0.10448261881956887,
      "learning_rate": 0.00010825793454723325,
      "loss": 0.0754,
      "step": 245
    },
    {
      "epoch": 1.5870967741935482,
      "grad_norm": 0.1295950737836306,
      "learning_rate": 0.00010750869819487883,
      "loss": 0.0745,
      "step": 246
    },
    {
      "epoch": 1.5935483870967742,
      "grad_norm": 0.09240489021108271,
      "learning_rate": 0.00010675903770198333,
      "loss": 0.067,
      "step": 247
    },
    {
      "epoch": 1.6,
      "grad_norm": 0.07808190236709202,
      "learning_rate": 0.00010600899541429004,
      "loss": 0.0546,
      "step": 248
    },
    {
      "epoch": 1.6064516129032258,
      "grad_norm": 0.0790414641966529,
      "learning_rate": 0.00010525861369910877,
      "loss": 0.0498,
      "step": 249
    },
    {
      "epoch": 1.6129032258064515,
      "grad_norm": 0.08616829997306262,
      "learning_rate": 0.00010450793494292224,
      "loss": 0.0655,
      "step": 250
    },
    {
      "epoch": 1.6193548387096774,
      "grad_norm": 0.08883827111615517,
      "learning_rate": 0.00010375700154899208,
      "loss": 0.0657,
      "step": 251
    },
    {
      "epoch": 1.6258064516129034,
      "grad_norm": 0.09734167988079584,
      "learning_rate": 0.00010300585593496348,
      "loss": 0.0614,
      "step": 252
    },
    {
      "epoch": 1.632258064516129,
      "grad_norm": 0.08926667751389035,
      "learning_rate": 0.00010225454053046921,
      "loss": 0.0562,
      "step": 253
    },
    {
      "epoch": 1.6387096774193548,
      "grad_norm": 0.13520312388410863,
      "learning_rate": 0.00010150309777473306,
      "loss": 0.0731,
      "step": 254
    },
    {
      "epoch": 1.6451612903225805,
      "grad_norm": 0.08417129846577816,
      "learning_rate": 0.0001007515701141722,
      "loss": 0.0496,
      "step": 255
    },
    {
      "epoch": 1.6516129032258065,
      "grad_norm": 0.15813812964913973,
      "learning_rate": 0.0001,
      "loss": 0.0622,
      "step": 256
    },
    {
      "epoch": 1.6580645161290324,
      "grad_norm": 0.07426217779998688,
      "learning_rate": 9.924842988582782e-05,
      "loss": 0.0487,
      "step": 257
    },
    {
      "epoch": 1.664516129032258,
      "grad_norm": 0.14075828336643512,
      "learning_rate": 9.849690222526698e-05,
      "loss": 0.0623,
      "step": 258
    },
    {
      "epoch": 1.6709677419354838,
      "grad_norm": 0.13317164150901403,
      "learning_rate": 9.77454594695308e-05,
      "loss": 0.0551,
      "step": 259
    },
    {
      "epoch": 1.6774193548387095,
      "grad_norm": 0.08655308767786873,
      "learning_rate": 9.699414406503654e-05,
      "loss": 0.0678,
      "step": 260
    },
    {
      "epoch": 1.6838709677419355,
      "grad_norm": 0.11118369085651197,
      "learning_rate": 9.624299845100795e-05,
      "loss": 0.0622,
      "step": 261
    },
    {
      "epoch": 1.6903225806451614,
      "grad_norm": 0.11077421899130901,
      "learning_rate": 9.549206505707777e-05,
      "loss": 0.0745,
      "step": 262
    },
    {
      "epoch": 1.696774193548387,
      "grad_norm": 0.08190755444484397,
      "learning_rate": 9.474138630089124e-05,
      "loss": 0.0537,
      "step": 263
    },
    {
      "epoch": 1.7032258064516128,
      "grad_norm": 0.1113159750162081,
      "learning_rate": 9.399100458570997e-05,
      "loss": 0.0691,
      "step": 264
    },
    {
      "epoch": 1.7096774193548387,
      "grad_norm": 0.09279661384255962,
      "learning_rate": 9.324096229801674e-05,
      "loss": 0.0583,
      "step": 265
    },
    {
      "epoch": 1.7161290322580647,
      "grad_norm": 0.07523224454997395,
      "learning_rate": 9.249130180512118e-05,
      "loss": 0.0621,
      "step": 266
    },
    {
      "epoch": 1.7225806451612904,
      "grad_norm": 0.09015314442775083,
      "learning_rate": 9.174206545276677e-05,
      "loss": 0.0669,
      "step": 267
    },
    {
      "epoch": 1.729032258064516,
      "grad_norm": 0.10636283133034866,
      "learning_rate": 9.099329556273866e-05,
      "loss": 0.0739,
      "step": 268
    },
    {
      "epoch": 1.7354838709677418,
      "grad_norm": 0.1416802448850876,
      "learning_rate": 9.024503443047319e-05,
      "loss": 0.0684,
      "step": 269
    },
    {
      "epoch": 1.7419354838709677,
      "grad_norm": 0.09949812802402141,
      "learning_rate": 8.949732432266866e-05,
      "loss": 0.0517,
      "step": 270
    },
    {
      "epoch": 1.7483870967741937,
      "grad_norm": 0.09911727857006153,
      "learning_rate": 8.875020747489794e-05,
      "loss": 0.0695,
      "step": 271
    },
    {
      "epoch": 1.7548387096774194,
      "grad_norm": 0.09965173501261688,
      "learning_rate": 8.800372608922271e-05,
      "loss": 0.0601,
      "step": 272
    },
    {
      "epoch": 1.761290322580645,
      "grad_norm": 0.11483628184027991,
      "learning_rate": 8.72579223318095e-05,
      "loss": 0.0747,
      "step": 273
    },
    {
      "epoch": 1.7677419354838708,
      "grad_norm": 0.08398542909094164,
      "learning_rate": 8.651283833054809e-05,
      "loss": 0.0545,
      "step": 274
    },
    {
      "epoch": 1.7741935483870968,
      "grad_norm": 0.09206084337768013,
      "learning_rate": 8.57685161726715e-05,
      "loss": 0.0626,
      "step": 275
    },
    {
      "epoch": 1.7806451612903227,
      "grad_norm": 0.10640419292629341,
      "learning_rate": 8.5024997902379e-05,
      "loss": 0.0694,
      "step": 276
    },
    {
      "epoch": 1.7870967741935484,
      "grad_norm": 0.09655305743665399,
      "learning_rate": 8.428232551846101e-05,
      "loss": 0.063,
      "step": 277
    },
    {
      "epoch": 1.793548387096774,
      "grad_norm": 0.10280172089235683,
      "learning_rate": 8.35405409719266e-05,
      "loss": 0.0628,
      "step": 278
    },
    {
      "epoch": 1.8,
      "grad_norm": 0.0924442773479908,
      "learning_rate": 8.279968616363418e-05,
      "loss": 0.0614,
      "step": 279
    },
    {
      "epoch": 1.8064516129032258,
      "grad_norm": 0.09056943407276584,
      "learning_rate": 8.205980294192421e-05,
      "loss": 0.0587,
      "step": 280
    },
    {
      "epoch": 1.8129032258064517,
      "grad_norm": 0.08536247550150874,
      "learning_rate": 8.132093310025571e-05,
      "loss": 0.0457,
      "step": 281
    },
    {
      "epoch": 1.8193548387096774,
      "grad_norm": 0.10278594457682778,
      "learning_rate": 8.058311837484535e-05,
      "loss": 0.0743,
      "step": 282
    },
    {
      "epoch": 1.8258064516129031,
      "grad_norm": 0.10489767053221469,
      "learning_rate": 7.984640044230983e-05,
      "loss": 0.059,
      "step": 283
    },
    {
      "epoch": 1.832258064516129,
      "grad_norm": 0.07066039733968037,
      "learning_rate": 7.911082091731181e-05,
      "loss": 0.05,
      "step": 284
    },
    {
      "epoch": 1.838709677419355,
      "grad_norm": 0.08869091516941932,
      "learning_rate": 7.837642135020929e-05,
      "loss": 0.0468,
      "step": 285
    },
    {
      "epoch": 1.8451612903225807,
      "grad_norm": 0.07614529156417063,
      "learning_rate": 7.764324322470841e-05,
      "loss": 0.0504,
      "step": 286
    },
    {
      "epoch": 1.8516129032258064,
      "grad_norm": 0.12289914291233693,
      "learning_rate": 7.691132795552043e-05,
      "loss": 0.0654,
      "step": 287
    },
    {
      "epoch": 1.8580645161290321,
      "grad_norm": 0.19187133428553324,
      "learning_rate": 7.618071688602199e-05,
      "loss": 0.0819,
      "step": 288
    },
    {
      "epoch": 1.864516129032258,
      "grad_norm": 0.1132288808580514,
      "learning_rate": 7.54514512859201e-05,
      "loss": 0.068,
      "step": 289
    },
    {
      "epoch": 1.870967741935484,
      "grad_norm": 0.213823719636533,
      "learning_rate": 7.472357234892082e-05,
      "loss": 0.0761,
      "step": 290
    },
    {
      "epoch": 1.8774193548387097,
      "grad_norm": 0.08954961479842181,
      "learning_rate": 7.399712119040238e-05,
      "loss": 0.0599,
      "step": 291
    },
    {
      "epoch": 1.8838709677419354,
      "grad_norm": 0.10264188935178613,
      "learning_rate": 7.327213884509272e-05,
      "loss": 0.0597,
      "step": 292
    },
    {
      "epoch": 1.8903225806451613,
      "grad_norm": 0.12757793483602733,
      "learning_rate": 7.254866626475152e-05,
      "loss": 0.0736,
      "step": 293
    },
    {
      "epoch": 1.896774193548387,
      "grad_norm": 0.1424378202330991,
      "learning_rate": 7.182674431585704e-05,
      "loss": 0.0758,
      "step": 294
    },
    {
      "epoch": 1.903225806451613,
      "grad_norm": 0.09267218373713473,
      "learning_rate": 7.110641377729778e-05,
      "loss": 0.0602,
      "step": 295
    },
    {
      "epoch": 1.9096774193548387,
      "grad_norm": 0.07573069220027309,
      "learning_rate": 7.038771533806884e-05,
      "loss": 0.0479,
      "step": 296
    },
    {
      "epoch": 1.9161290322580644,
      "grad_norm": 0.11850380653093652,
      "learning_rate": 6.967068959497376e-05,
      "loss": 0.061,
      "step": 297
    },
    {
      "epoch": 1.9225806451612903,
      "grad_norm": 0.0952951197418935,
      "learning_rate": 6.895537705033108e-05,
      "loss": 0.0719,
      "step": 298
    },
    {
      "epoch": 1.9290322580645163,
      "grad_norm": 0.0869206540087956,
      "learning_rate": 6.824181810968675e-05,
      "loss": 0.0614,
      "step": 299
    },
    {
      "epoch": 1.935483870967742,
      "grad_norm": 0.11514530265371259,
      "learning_rate": 6.753005307953167e-05,
      "loss": 0.0631,
      "step": 300
    },
    {
      "epoch": 1.9419354838709677,
      "grad_norm": 0.113758246927865,
      "learning_rate": 6.682012216502484e-05,
      "loss": 0.0737,
      "step": 301
    },
    {
      "epoch": 1.9483870967741934,
      "grad_norm": 0.1202740564887612,
      "learning_rate": 6.611206546772237e-05,
      "loss": 0.0665,
      "step": 302
    },
    {
      "epoch": 1.9548387096774194,
      "grad_norm": 0.12119681869029483,
      "learning_rate": 6.54059229833124e-05,
      "loss": 0.0617,
      "step": 303
    },
    {
      "epoch": 1.9612903225806453,
      "grad_norm": 0.11167959831458651,
      "learning_rate": 6.47017345993556e-05,
      "loss": 0.0636,
      "step": 304
    },
    {
      "epoch": 1.967741935483871,
      "grad_norm": 0.10236314721384843,
      "learning_rate": 6.39995400930324e-05,
      "loss": 0.0691,
      "step": 305
    },
    {
      "epoch": 1.9741935483870967,
      "grad_norm": 0.08978452227249356,
      "learning_rate": 6.329937912889582e-05,
      "loss": 0.0486,
      "step": 306
    },
    {
      "epoch": 1.9806451612903224,
      "grad_norm": 0.09402215683973497,
      "learning_rate": 6.260129125663106e-05,
      "loss": 0.0575,
      "step": 307
    },
    {
      "epoch": 1.9870967741935484,
      "grad_norm": 0.09176207813711797,
      "learning_rate": 6.190531590882159e-05,
      "loss": 0.0616,
      "step": 308
    },
    {
      "epoch": 1.9935483870967743,
      "grad_norm": 0.1461148677177572,
      "learning_rate": 6.121149239872151e-05,
      "loss": 0.084,
      "step": 309
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.0936208966101867,
      "learning_rate": 6.051985991803517e-05,
      "loss": 0.0463,
      "step": 310
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.09596683084964752,
      "eval_runtime": 25.5767,
      "eval_samples_per_second": 5.122,
      "eval_steps_per_second": 0.665,
      "step": 310
    },
    {
      "epoch": 2.0064516129032257,
      "grad_norm": 0.06773450997938296,
      "learning_rate": 5.983045753470308e-05,
      "loss": 0.0453,
      "step": 311
    },
    {
      "epoch": 2.0129032258064514,
      "grad_norm": 0.05341229626959274,
      "learning_rate": 5.9143324190695196e-05,
      "loss": 0.036,
      "step": 312
    },
    {
      "epoch": 2.0193548387096776,
      "grad_norm": 0.06711247984899421,
      "learning_rate": 5.845849869981137e-05,
      "loss": 0.046,
      "step": 313
    },
    {
      "epoch": 2.0258064516129033,
      "grad_norm": 0.061859413521882546,
      "learning_rate": 5.777601974548866e-05,
      "loss": 0.0352,
      "step": 314
    },
    {
      "epoch": 2.032258064516129,
      "grad_norm": 0.05446833363095474,
      "learning_rate": 5.709592587861637e-05,
      "loss": 0.0369,
      "step": 315
    },
    {
      "epoch": 2.0387096774193547,
      "grad_norm": 0.0782453373649888,
      "learning_rate": 5.6418255515358486e-05,
      "loss": 0.0349,
      "step": 316
    },
    {
      "epoch": 2.0451612903225804,
      "grad_norm": 0.04790490679204087,
      "learning_rate": 5.574304693498346e-05,
      "loss": 0.0308,
      "step": 317
    },
    {
      "epoch": 2.0516129032258066,
      "grad_norm": 0.062494284774837545,
      "learning_rate": 5.507033827770225e-05,
      "loss": 0.0353,
      "step": 318
    },
    {
      "epoch": 2.0580645161290323,
      "grad_norm": 0.07037383356374984,
      "learning_rate": 5.4400167542513636e-05,
      "loss": 0.0412,
      "step": 319
    },
    {
      "epoch": 2.064516129032258,
      "grad_norm": 0.08297196040492472,
      "learning_rate": 5.3732572585057974e-05,
      "loss": 0.045,
      "step": 320
    },
    {
      "epoch": 2.0709677419354837,
      "grad_norm": 0.05677712601652311,
      "learning_rate": 5.306759111547881e-05,
      "loss": 0.0314,
      "step": 321
    },
    {
      "epoch": 2.07741935483871,
      "grad_norm": 0.10970235904680202,
      "learning_rate": 5.240526069629265e-05,
      "loss": 0.0426,
      "step": 322
    },
    {
      "epoch": 2.0838709677419356,
      "grad_norm": 0.1074904876869991,
      "learning_rate": 5.174561874026741e-05,
      "loss": 0.0453,
      "step": 323
    },
    {
      "epoch": 2.0903225806451613,
      "grad_norm": 0.05559023402895479,
      "learning_rate": 5.108870250830882e-05,
      "loss": 0.0295,
      "step": 324
    },
    {
      "epoch": 2.096774193548387,
      "grad_norm": 0.09545856608388173,
      "learning_rate": 5.0434549107355944e-05,
      "loss": 0.0441,
      "step": 325
    },
    {
      "epoch": 2.1032258064516127,
      "grad_norm": 0.09664814857644259,
      "learning_rate": 4.978319548828504e-05,
      "loss": 0.0335,
      "step": 326
    },
    {
      "epoch": 2.109677419354839,
      "grad_norm": 0.09467665510676392,
      "learning_rate": 4.9134678443822166e-05,
      "loss": 0.0251,
      "step": 327
    },
    {
      "epoch": 2.1161290322580646,
      "grad_norm": 0.1384628187897895,
      "learning_rate": 4.8489034606465225e-05,
      "loss": 0.0369,
      "step": 328
    },
    {
      "epoch": 2.1225806451612903,
      "grad_norm": 0.12578896351994917,
      "learning_rate": 4.784630044641435e-05,
      "loss": 0.0392,
      "step": 329
    },
    {
      "epoch": 2.129032258064516,
      "grad_norm": 0.16521279278081405,
      "learning_rate": 4.7206512269512124e-05,
      "loss": 0.0463,
      "step": 330
    },
    {
      "epoch": 2.135483870967742,
      "grad_norm": 0.11984973498381633,
      "learning_rate": 4.65697062151927e-05,
      "loss": 0.0394,
      "step": 331
    },
    {
      "epoch": 2.141935483870968,
      "grad_norm": 0.09544223502654232,
      "learning_rate": 4.593591825444028e-05,
      "loss": 0.037,
      "step": 332
    },
    {
      "epoch": 2.1483870967741936,
      "grad_norm": 0.13562773315198307,
      "learning_rate": 4.530518418775733e-05,
      "loss": 0.039,
      "step": 333
    },
    {
      "epoch": 2.1548387096774193,
      "grad_norm": 0.08782158354204284,
      "learning_rate": 4.4677539643142454e-05,
      "loss": 0.0392,
      "step": 334
    },
    {
      "epoch": 2.161290322580645,
      "grad_norm": 0.08174883141208888,
      "learning_rate": 4.40530200740777e-05,
      "loss": 0.0313,
      "step": 335
    },
    {
      "epoch": 2.167741935483871,
      "grad_norm": 0.09796897925277187,
      "learning_rate": 4.343166075752605e-05,
      "loss": 0.0337,
      "step": 336
    },
    {
      "epoch": 2.174193548387097,
      "grad_norm": 0.11851301692643428,
      "learning_rate": 4.281349679193861e-05,
      "loss": 0.039,
      "step": 337
    },
    {
      "epoch": 2.1806451612903226,
      "grad_norm": 0.10807728159748675,
      "learning_rate": 4.2198563095272116e-05,
      "loss": 0.0301,
      "step": 338
    },
    {
      "epoch": 2.1870967741935483,
      "grad_norm": 0.2141192740082474,
      "learning_rate": 4.158689440301657e-05,
      "loss": 0.0354,
      "step": 339
    },
    {
      "epoch": 2.193548387096774,
      "grad_norm": 0.07257549552594048,
      "learning_rate": 4.097852526623307e-05,
      "loss": 0.0339,
      "step": 340
    },
    {
      "epoch": 2.2,
      "grad_norm": 0.19631930583890622,
      "learning_rate": 4.0373490049602204e-05,
      "loss": 0.0497,
      "step": 341
    },
    {
      "epoch": 2.206451612903226,
      "grad_norm": 0.08740964228587876,
      "learning_rate": 3.977182292948283e-05,
      "loss": 0.0478,
      "step": 342
    },
    {
      "epoch": 2.2129032258064516,
      "grad_norm": 0.08272419362723894,
      "learning_rate": 3.9173557891981573e-05,
      "loss": 0.0379,
      "step": 343
    },
    {
      "epoch": 2.2193548387096773,
      "grad_norm": 0.08040985475184476,
      "learning_rate": 3.857872873103322e-05,
      "loss": 0.0447,
      "step": 344
    },
    {
      "epoch": 2.225806451612903,
      "grad_norm": 0.09245277473267594,
      "learning_rate": 3.7987369046491684e-05,
      "loss": 0.0391,
      "step": 345
    },
    {
      "epoch": 2.232258064516129,
      "grad_norm": 0.12614672525793685,
      "learning_rate": 3.7399512242231995e-05,
      "loss": 0.0342,
      "step": 346
    },
    {
      "epoch": 2.238709677419355,
      "grad_norm": 0.09266235955840761,
      "learning_rate": 3.6815191524263624e-05,
      "loss": 0.0354,
      "step": 347
    },
    {
      "epoch": 2.2451612903225806,
      "grad_norm": 0.06080492053598232,
      "learning_rate": 3.623443989885462e-05,
      "loss": 0.0335,
      "step": 348
    },
    {
      "epoch": 2.2516129032258063,
      "grad_norm": 0.07807865557506005,
      "learning_rate": 3.565729017066729e-05,
      "loss": 0.0347,
      "step": 349
    },
    {
      "epoch": 2.258064516129032,
      "grad_norm": 0.08087459186315774,
      "learning_rate": 3.508377494090521e-05,
      "loss": 0.036,
      "step": 350
    },
    {
      "epoch": 2.264516129032258,
      "grad_norm": 0.08899211474214212,
      "learning_rate": 3.45139266054715e-05,
      "loss": 0.0343,
      "step": 351
    },
    {
      "epoch": 2.270967741935484,
      "grad_norm": 0.07056377080769806,
      "learning_rate": 3.394777735313919e-05,
      "loss": 0.033,
      "step": 352
    },
    {
      "epoch": 2.2774193548387096,
      "grad_norm": 0.0786360395215486,
      "learning_rate": 3.338535916373266e-05,
      "loss": 0.0319,
      "step": 353
    },
    {
      "epoch": 2.2838709677419353,
      "grad_norm": 0.0870351921736199,
      "learning_rate": 3.2826703806321525e-05,
      "loss": 0.0408,
      "step": 354
    },
    {
      "epoch": 2.2903225806451615,
      "grad_norm": 0.07340127696415631,
      "learning_rate": 3.227184283742591e-05,
      "loss": 0.0348,
      "step": 355
    },
    {
      "epoch": 2.296774193548387,
      "grad_norm": 0.08727491345423612,
      "learning_rate": 3.17208075992339e-05,
      "loss": 0.0349,
      "step": 356
    },
    {
      "epoch": 2.303225806451613,
      "grad_norm": 0.08641599877739083,
      "learning_rate": 3.117362921783134e-05,
      "loss": 0.0435,
      "step": 357
    },
    {
      "epoch": 2.3096774193548386,
      "grad_norm": 0.06573525447698249,
      "learning_rate": 3.063033860144339e-05,
      "loss": 0.0316,
      "step": 358
    },
    {
      "epoch": 2.3161290322580643,
      "grad_norm": 0.07221179151878702,
      "learning_rate": 3.0090966438688772e-05,
      "loss": 0.0324,
      "step": 359
    },
    {
      "epoch": 2.3225806451612905,
      "grad_norm": 0.18424602936724097,
      "learning_rate": 2.9555543196846292e-05,
      "loss": 0.0396,
      "step": 360
    },
    {
      "epoch": 2.329032258064516,
      "grad_norm": 0.08124593683663159,
      "learning_rate": 2.9024099120133673e-05,
      "loss": 0.0365,
      "step": 361
    },
    {
      "epoch": 2.335483870967742,
      "grad_norm": 0.10243519568473496,
      "learning_rate": 2.8496664227999415e-05,
      "loss": 0.0431,
      "step": 362
    },
    {
      "epoch": 2.3419354838709676,
      "grad_norm": 0.08863145799938825,
      "learning_rate": 2.7973268313426837e-05,
      "loss": 0.0431,
      "step": 363
    },
    {
      "epoch": 2.3483870967741938,
      "grad_norm": 0.1157770288973089,
      "learning_rate": 2.745394094125141e-05,
      "loss": 0.0384,
      "step": 364
    },
    {
      "epoch": 2.3548387096774195,
      "grad_norm": 0.06937293824933606,
      "learning_rate": 2.6938711446490606e-05,
      "loss": 0.0326,
      "step": 365
    },
    {
      "epoch": 2.361290322580645,
      "grad_norm": 0.11070613688051265,
      "learning_rate": 2.6427608932686843e-05,
      "loss": 0.0453,
      "step": 366
    },
    {
      "epoch": 2.367741935483871,
      "grad_norm": 0.07304602035921624,
      "learning_rate": 2.5920662270263653e-05,
      "loss": 0.0349,
      "step": 367
    },
    {
      "epoch": 2.3741935483870966,
      "grad_norm": 0.0957880185255825,
      "learning_rate": 2.5417900094894744e-05,
      "loss": 0.0384,
      "step": 368
    },
    {
      "epoch": 2.3806451612903228,
      "grad_norm": 0.06844023851775795,
      "learning_rate": 2.4919350805886577e-05,
      "loss": 0.0312,
      "step": 369
    },
    {
      "epoch": 2.3870967741935485,
      "grad_norm": 0.10510477036054879,
      "learning_rate": 2.4425042564574184e-05,
      "loss": 0.0337,
      "step": 370
    },
    {
      "epoch": 2.393548387096774,
      "grad_norm": 0.0969617018143466,
      "learning_rate": 2.3935003292730296e-05,
      "loss": 0.035,
      "step": 371
    },
    {
      "epoch": 2.4,
      "grad_norm": 0.07967820297616653,
      "learning_rate": 2.344926067098836e-05,
      "loss": 0.037,
      "step": 372
    },
    {
      "epoch": 2.4064516129032256,
      "grad_norm": 0.07497176613411485,
      "learning_rate": 2.2967842137278706e-05,
      "loss": 0.0328,
      "step": 373
    },
    {
      "epoch": 2.412903225806452,
      "grad_norm": 0.09754092780042643,
      "learning_rate": 2.2490774885278908e-05,
      "loss": 0.0313,
      "step": 374
    },
    {
      "epoch": 2.4193548387096775,
      "grad_norm": 0.08185743462021658,
      "learning_rate": 2.201808586287757e-05,
      "loss": 0.031,
      "step": 375
    },
    {
      "epoch": 2.425806451612903,
      "grad_norm": 0.08264089376813998,
      "learning_rate": 2.15498017706521e-05,
      "loss": 0.0385,
      "step": 376
    },
    {
      "epoch": 2.432258064516129,
      "grad_norm": 0.0823740031947979,
      "learning_rate": 2.1085949060360654e-05,
      "loss": 0.039,
      "step": 377
    },
    {
      "epoch": 2.4387096774193546,
      "grad_norm": 0.07247982019193418,
      "learning_rate": 2.0626553933447734e-05,
      "loss": 0.0264,
      "step": 378
    },
    {
      "epoch": 2.445161290322581,
      "grad_norm": 0.08404435825498824,
      "learning_rate": 2.01716423395644e-05,
      "loss": 0.0479,
      "step": 379
    },
    {
      "epoch": 2.4516129032258065,
      "grad_norm": 0.08666218883884351,
      "learning_rate": 1.9721239975102313e-05,
      "loss": 0.0377,
      "step": 380
    },
    {
      "epoch": 2.458064516129032,
      "grad_norm": 0.0813934369010793,
      "learning_rate": 1.9275372281742242e-05,
      "loss": 0.0287,
      "step": 381
    },
    {
      "epoch": 2.464516129032258,
      "grad_norm": 0.08490400932239225,
      "learning_rate": 1.8834064445016953e-05,
      "loss": 0.0287,
      "step": 382
    },
    {
      "epoch": 2.4709677419354836,
      "grad_norm": 0.0745619544814596,
      "learning_rate": 1.839734139288868e-05,
      "loss": 0.0313,
      "step": 383
    },
    {
      "epoch": 2.47741935483871,
      "grad_norm": 0.07603999026318416,
      "learning_rate": 1.7965227794340877e-05,
      "loss": 0.0282,
      "step": 384
    },
    {
      "epoch": 2.4838709677419355,
      "grad_norm": 0.07482170653624513,
      "learning_rate": 1.753774805798486e-05,
      "loss": 0.0338,
      "step": 385
    },
    {
      "epoch": 2.490322580645161,
      "grad_norm": 0.11239247582773948,
      "learning_rate": 1.7114926330680957e-05,
      "loss": 0.0332,
      "step": 386
    },
    {
      "epoch": 2.496774193548387,
      "grad_norm": 0.08250289864329428,
      "learning_rate": 1.6696786496174578e-05,
      "loss": 0.0402,
      "step": 387
    },
    {
      "epoch": 2.5032258064516126,
      "grad_norm": 0.10452752517237242,
      "learning_rate": 1.6283352173747145e-05,
      "loss": 0.0333,
      "step": 388
    },
    {
      "epoch": 2.509677419354839,
      "grad_norm": 0.09847614894022073,
      "learning_rate": 1.587464671688187e-05,
      "loss": 0.0349,
      "step": 389
    },
    {
      "epoch": 2.5161290322580645,
      "grad_norm": 0.08854153674623017,
      "learning_rate": 1.5470693211944643e-05,
      "loss": 0.0294,
      "step": 390
    },
    {
      "epoch": 2.52258064516129,
      "grad_norm": 0.07568785710192437,
      "learning_rate": 1.5071514476879878e-05,
      "loss": 0.0304,
      "step": 391
    },
    {
      "epoch": 2.5290322580645164,
      "grad_norm": 0.10747782045229992,
      "learning_rate": 1.4677133059921632e-05,
      "loss": 0.034,
      "step": 392
    },
    {
      "epoch": 2.535483870967742,
      "grad_norm": 0.07944602809272536,
      "learning_rate": 1.4287571238320053e-05,
      "loss": 0.0362,
      "step": 393
    },
    {
      "epoch": 2.541935483870968,
      "grad_norm": 0.09637365953273927,
      "learning_rate": 1.3902851017082864e-05,
      "loss": 0.0351,
      "step": 394
    },
    {
      "epoch": 2.5483870967741935,
      "grad_norm": 0.07895627852796791,
      "learning_rate": 1.3522994127732414e-05,
      "loss": 0.0348,
      "step": 395
    },
    {
      "epoch": 2.554838709677419,
      "grad_norm": 0.11668198814666138,
      "learning_rate": 1.3148022027078222e-05,
      "loss": 0.0377,
      "step": 396
    },
    {
      "epoch": 2.5612903225806454,
      "grad_norm": 0.08859887668842747,
      "learning_rate": 1.2777955896004812e-05,
      "loss": 0.0339,
      "step": 397
    },
    {
      "epoch": 2.567741935483871,
      "grad_norm": 0.0955648903311151,
      "learning_rate": 1.2412816638275404e-05,
      "loss": 0.0429,
      "step": 398
    },
    {
      "epoch": 2.574193548387097,
      "grad_norm": 0.08953857592938266,
      "learning_rate": 1.2052624879351104e-05,
      "loss": 0.0354,
      "step": 399
    },
    {
      "epoch": 2.5806451612903225,
      "grad_norm": 0.12572795234550732,
      "learning_rate": 1.1697400965225747e-05,
      "loss": 0.0403,
      "step": 400
    },
    {
      "epoch": 2.587096774193548,
      "grad_norm": 0.10694439782055375,
      "learning_rate": 1.134716496127679e-05,
      "loss": 0.0336,
      "step": 401
    },
    {
      "epoch": 2.5935483870967744,
      "grad_norm": 0.06513844231274085,
      "learning_rate": 1.1001936651131717e-05,
      "loss": 0.0329,
      "step": 402
    },
    {
      "epoch": 2.6,
      "grad_norm": 0.09974875770813588,
      "learning_rate": 1.0661735535550666e-05,
      "loss": 0.0365,
      "step": 403
    },
    {
      "epoch": 2.606451612903226,
      "grad_norm": 0.06817268977991534,
      "learning_rate": 1.0326580831324817e-05,
      "loss": 0.0299,
      "step": 404
    },
    {
      "epoch": 2.6129032258064515,
      "grad_norm": 0.06220845407702803,
      "learning_rate": 9.996491470190917e-06,
      "loss": 0.027,
      "step": 405
    },
    {
      "epoch": 2.6193548387096772,
      "grad_norm": 0.08453636414206647,
      "learning_rate": 9.671486097761917e-06,
      "loss": 0.0418,
      "step": 406
    },
    {
      "epoch": 2.6258064516129034,
      "grad_norm": 0.10895972341564872,
      "learning_rate": 9.351583072473713e-06,
      "loss": 0.0411,
      "step": 407
    },
    {
      "epoch": 2.632258064516129,
      "grad_norm": 0.08350751437942873,
      "learning_rate": 9.036800464548157e-06,
      "loss": 0.0434,
      "step": 408
    },
    {
      "epoch": 2.638709677419355,
      "grad_norm": 0.10366684902670063,
      "learning_rate": 8.727156054972374e-06,
      "loss": 0.039,
      "step": 409
    },
    {
      "epoch": 2.6451612903225805,
      "grad_norm": 0.09956778774869587,
      "learning_rate": 8.422667334494249e-06,
      "loss": 0.0408,
      "step": 410
    },
    {
      "epoch": 2.6516129032258062,
      "grad_norm": 0.08758251356693646,
      "learning_rate": 8.123351502634625e-06,
      "loss": 0.0348,
      "step": 411
    },
    {
      "epoch": 2.6580645161290324,
      "grad_norm": 0.10844348207137618,
      "learning_rate": 7.82922546671555e-06,
      "loss": 0.0408,
      "step": 412
    },
    {
      "epoch": 2.664516129032258,
      "grad_norm": 0.10636911692675921,
      "learning_rate": 7.54030584090537e-06,
      "loss": 0.0429,
      "step": 413
    },
    {
      "epoch": 2.670967741935484,
      "grad_norm": 0.09542792766451057,
      "learning_rate": 7.256608945280319e-06,
      "loss": 0.0447,
      "step": 414
    },
    {
      "epoch": 2.6774193548387095,
      "grad_norm": 0.07697171392030314,
      "learning_rate": 6.97815080490245e-06,
      "loss": 0.0367,
      "step": 415
    },
    {
      "epoch": 2.6838709677419352,
      "grad_norm": 0.072266260251159,
      "learning_rate": 6.704947148914609e-06,
      "loss": 0.0325,
      "step": 416
    },
    {
      "epoch": 2.6903225806451614,
      "grad_norm": 0.0945534978466976,
      "learning_rate": 6.437013409651849e-06,
      "loss": 0.0296,
      "step": 417
    },
    {
      "epoch": 2.696774193548387,
      "grad_norm": 0.10612283288506409,
      "learning_rate": 6.174364721769743e-06,
      "loss": 0.0492,
      "step": 418
    },
    {
      "epoch": 2.703225806451613,
      "grad_norm": 0.08110795988519982,
      "learning_rate": 5.917015921389568e-06,
      "loss": 0.0313,
      "step": 419
    },
    {
      "epoch": 2.709677419354839,
      "grad_norm": 0.0662490047037332,
      "learning_rate": 5.664981545260073e-06,
      "loss": 0.028,
      "step": 420
    },
    {
      "epoch": 2.7161290322580647,
      "grad_norm": 0.11866120174733467,
      "learning_rate": 5.418275829936537e-06,
      "loss": 0.0335,
      "step": 421
    },
    {
      "epoch": 2.7225806451612904,
      "grad_norm": 0.09690307384076605,
      "learning_rate": 5.176912710976467e-06,
      "loss": 0.0328,
      "step": 422
    },
    {
      "epoch": 2.729032258064516,
      "grad_norm": 0.08332254956064335,
      "learning_rate": 4.940905822152453e-06,
      "loss": 0.0376,
      "step": 423
    },
    {
      "epoch": 2.735483870967742,
      "grad_norm": 0.10534480078797799,
      "learning_rate": 4.710268494682146e-06,
      "loss": 0.0427,
      "step": 424
    },
    {
      "epoch": 2.741935483870968,
      "grad_norm": 0.07751933249718136,
      "learning_rate": 4.485013756475076e-06,
      "loss": 0.0351,
      "step": 425
    },
    {
      "epoch": 2.7483870967741937,
      "grad_norm": 0.08421370184302716,
      "learning_rate": 4.2651543313968145e-06,
      "loss": 0.0349,
      "step": 426
    },
    {
      "epoch": 2.7548387096774194,
      "grad_norm": 0.09663498717781645,
      "learning_rate": 4.050702638550275e-06,
      "loss": 0.0375,
      "step": 427
    },
    {
      "epoch": 2.761290322580645,
      "grad_norm": 0.10109504165697188,
      "learning_rate": 3.841670791574137e-06,
      "loss": 0.0389,
      "step": 428
    },
    {
      "epoch": 2.767741935483871,
      "grad_norm": 0.07242313321239978,
      "learning_rate": 3.638070597958665e-06,
      "loss": 0.032,
      "step": 429
    },
    {
      "epoch": 2.774193548387097,
      "grad_norm": 0.07339732439560019,
      "learning_rate": 3.4399135583787043e-06,
      "loss": 0.0354,
      "step": 430
    },
    {
      "epoch": 2.7806451612903227,
      "grad_norm": 0.10075765784942739,
      "learning_rate": 3.2472108660439706e-06,
      "loss": 0.0492,
      "step": 431
    },
    {
      "epoch": 2.7870967741935484,
      "grad_norm": 0.07589971575113699,
      "learning_rate": 3.059973406066963e-06,
      "loss": 0.031,
      "step": 432
    },
    {
      "epoch": 2.793548387096774,
      "grad_norm": 0.08019217757997281,
      "learning_rate": 2.878211754847926e-06,
      "loss": 0.0421,
      "step": 433
    },
    {
      "epoch": 2.8,
      "grad_norm": 0.12861410711276888,
      "learning_rate": 2.7019361794775156e-06,
      "loss": 0.0365,
      "step": 434
    },
    {
      "epoch": 2.806451612903226,
      "grad_norm": 0.09493420573436114,
      "learning_rate": 2.5311566371568507e-06,
      "loss": 0.0326,
      "step": 435
    },
    {
      "epoch": 2.8129032258064517,
      "grad_norm": 0.07274924242928238,
      "learning_rate": 2.365882774634998e-06,
      "loss": 0.0398,
      "step": 436
    },
    {
      "epoch": 2.8193548387096774,
      "grad_norm": 0.08169132644155476,
      "learning_rate": 2.206123927664161e-06,
      "loss": 0.0379,
      "step": 437
    },
    {
      "epoch": 2.825806451612903,
      "grad_norm": 0.07933207328170928,
      "learning_rate": 2.0518891204722168e-06,
      "loss": 0.0255,
      "step": 438
    },
    {
      "epoch": 2.832258064516129,
      "grad_norm": 0.09957580250269343,
      "learning_rate": 1.903187065253076e-06,
      "loss": 0.0445,
      "step": 439
    },
    {
      "epoch": 2.838709677419355,
      "grad_norm": 0.08137735740771984,
      "learning_rate": 1.7600261616745106e-06,
      "loss": 0.0422,
      "step": 440
    },
    {
      "epoch": 2.8451612903225807,
      "grad_norm": 0.10587692098340466,
      "learning_rate": 1.6224144964036681e-06,
      "loss": 0.0364,
      "step": 441
    },
    {
      "epoch": 2.8516129032258064,
      "grad_norm": 0.10343920804144822,
      "learning_rate": 1.4903598426503241e-06,
      "loss": 0.0327,
      "step": 442
    },
    {
      "epoch": 2.858064516129032,
      "grad_norm": 0.07946725527009221,
      "learning_rate": 1.3638696597277679e-06,
      "loss": 0.0341,
      "step": 443
    },
    {
      "epoch": 2.864516129032258,
      "grad_norm": 0.09586325289816539,
      "learning_rate": 1.2429510926314836e-06,
      "loss": 0.0267,
      "step": 444
    },
    {
      "epoch": 2.870967741935484,
      "grad_norm": 0.0827004341912287,
      "learning_rate": 1.1276109716355287e-06,
      "loss": 0.0416,
      "step": 445
    },
    {
      "epoch": 2.8774193548387097,
      "grad_norm": 0.07005026862609978,
      "learning_rate": 1.0178558119067315e-06,
      "loss": 0.0344,
      "step": 446
    },
    {
      "epoch": 2.8838709677419354,
      "grad_norm": 0.09331217539366578,
      "learning_rate": 9.136918131366412e-07,
      "loss": 0.0353,
      "step": 447
    },
    {
      "epoch": 2.8903225806451616,
      "grad_norm": 0.08846943376776958,
      "learning_rate": 8.151248591913518e-07,
      "loss": 0.0386,
      "step": 448
    },
    {
      "epoch": 2.896774193548387,
      "grad_norm": 0.1051100476797595,
      "learning_rate": 7.221605177791691e-07,
      "loss": 0.0411,
      "step": 449
    },
    {
      "epoch": 2.903225806451613,
      "grad_norm": 0.10646364294182206,
      "learning_rate": 6.348040401360833e-07,
      "loss": 0.0346,
      "step": 450
    },
    {
      "epoch": 2.9096774193548387,
      "grad_norm": 0.14514647433288821,
      "learning_rate": 5.530603607290851e-07,
      "loss": 0.0432,
      "step": 451
    },
    {
      "epoch": 2.9161290322580644,
      "grad_norm": 0.08548221741715605,
      "learning_rate": 4.76934096977566e-07,
      "loss": 0.0318,
      "step": 452
    },
    {
      "epoch": 2.9225806451612906,
      "grad_norm": 0.07886089519019772,
      "learning_rate": 4.0642954899238197e-07,
      "loss": 0.0387,
      "step": 453
    },
    {
      "epoch": 2.9290322580645163,
      "grad_norm": 0.10187728036061035,
      "learning_rate": 3.415506993330153e-07,
      "loss": 0.0385,
      "step": 454
    },
    {
      "epoch": 2.935483870967742,
      "grad_norm": 0.06877182838764333,
      "learning_rate": 2.8230121278257637e-07,
      "loss": 0.0268,
      "step": 455
    },
    {
      "epoch": 2.9419354838709677,
      "grad_norm": 0.07135072782410999,
      "learning_rate": 2.2868443614082469e-07,
      "loss": 0.0348,
      "step": 456
    },
    {
      "epoch": 2.9483870967741934,
      "grad_norm": 0.0784846474201083,
      "learning_rate": 1.8070339803509807e-07,
      "loss": 0.0401,
      "step": 457
    },
    {
      "epoch": 2.9548387096774196,
      "grad_norm": 0.07055206711925587,
      "learning_rate": 1.3836080874926049e-07,
      "loss": 0.0341,
      "step": 458
    },
    {
      "epoch": 2.9612903225806453,
      "grad_norm": 0.12997432093488728,
      "learning_rate": 1.0165906007056914e-07,
      "loss": 0.0397,
      "step": 459
    },
    {
      "epoch": 2.967741935483871,
      "grad_norm": 0.07787008340791496,
      "learning_rate": 7.060022515460451e-08,
      "loss": 0.0327,
      "step": 460
    },
    {
      "epoch": 2.9741935483870967,
      "grad_norm": 0.08207645810528069,
      "learning_rate": 4.518605840815315e-08,
      "loss": 0.0433,
      "step": 461
    },
    {
      "epoch": 2.9806451612903224,
      "grad_norm": 0.0969168591567496,
      "learning_rate": 2.5417995390086824e-08,
      "loss": 0.0358,
      "step": 462
    },
    {
      "epoch": 2.9870967741935486,
      "grad_norm": 0.08709768527509966,
      "learning_rate": 1.129715273033849e-08,
      "loss": 0.0363,
      "step": 463
    },
    {
      "epoch": 2.9935483870967743,
      "grad_norm": 0.09987493142651711,
      "learning_rate": 2.824328066730608e-09,
      "loss": 0.0362,
      "step": 464
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.06314894199443474,
      "learning_rate": 0.0,
      "loss": 0.0271,
      "step": 465
    },
    {
      "epoch": 3.0,
      "eval_loss": 0.11768443137407303,
      "eval_runtime": 25.378,
      "eval_samples_per_second": 5.162,
      "eval_steps_per_second": 0.67,
      "step": 465
    },
    {
      "epoch": 3.0,
      "step": 465,
      "total_flos": 261374226563072.0,
      "train_loss": 0.07172503977693537,
      "train_runtime": 3552.9052,
      "train_samples_per_second": 2.091,
      "train_steps_per_second": 0.131
    }
  ],
  "logging_steps": 1,
  "max_steps": 465,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 261374226563072.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}