{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.5823358136525396,
  "eval_steps": 386,
  "global_step": 450,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0012940795858945326,
      "grad_norm": 1.7405146360397339,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 2.4269,
      "step": 1
    },
    {
      "epoch": 0.0012940795858945326,
      "eval_loss": 2.247628688812256,
      "eval_runtime": 189.8853,
      "eval_samples_per_second": 3.428,
      "eval_steps_per_second": 0.858,
      "step": 1
    },
    {
      "epoch": 0.002588159171789065,
      "grad_norm": 1.6643542051315308,
      "learning_rate": 4.000000000000001e-06,
      "loss": 2.2583,
      "step": 2
    },
    {
      "epoch": 0.0038822387576835974,
      "grad_norm": 1.8690767288208008,
      "learning_rate": 6e-06,
      "loss": 2.2696,
      "step": 3
    },
    {
      "epoch": 0.00517631834357813,
      "grad_norm": 1.828118085861206,
      "learning_rate": 8.000000000000001e-06,
      "loss": 2.3646,
      "step": 4
    },
    {
      "epoch": 0.006470397929472662,
      "grad_norm": 1.9319926500320435,
      "learning_rate": 1e-05,
      "loss": 2.4196,
      "step": 5
    },
    {
      "epoch": 0.007764477515367195,
      "grad_norm": 1.7723782062530518,
      "learning_rate": 1.2e-05,
      "loss": 2.4177,
      "step": 6
    },
    {
      "epoch": 0.009058557101261728,
      "grad_norm": 1.9500815868377686,
      "learning_rate": 1.4000000000000001e-05,
      "loss": 2.3497,
      "step": 7
    },
    {
      "epoch": 0.01035263668715626,
      "grad_norm": 2.3909075260162354,
      "learning_rate": 1.6000000000000003e-05,
      "loss": 2.405,
      "step": 8
    },
    {
      "epoch": 0.011646716273050793,
      "grad_norm": 2.0620856285095215,
      "learning_rate": 1.8e-05,
      "loss": 2.4098,
      "step": 9
    },
    {
      "epoch": 0.012940795858945324,
      "grad_norm": 1.8054910898208618,
      "learning_rate": 2e-05,
      "loss": 2.1233,
      "step": 10
    },
    {
      "epoch": 0.014234875444839857,
      "grad_norm": 2.190964937210083,
      "learning_rate": 2.2000000000000003e-05,
      "loss": 2.3985,
      "step": 11
    },
    {
      "epoch": 0.01552895503073439,
      "grad_norm": 1.9412921667099,
      "learning_rate": 2.4e-05,
      "loss": 2.462,
      "step": 12
    },
    {
      "epoch": 0.016823034616628922,
      "grad_norm": 1.9161555767059326,
      "learning_rate": 2.6000000000000002e-05,
      "loss": 2.2118,
      "step": 13
    },
    {
      "epoch": 0.018117114202523456,
      "grad_norm": 1.7161599397659302,
      "learning_rate": 2.8000000000000003e-05,
      "loss": 2.2175,
      "step": 14
    },
    {
      "epoch": 0.019411193788417987,
      "grad_norm": 2.173877000808716,
      "learning_rate": 3e-05,
      "loss": 2.2521,
      "step": 15
    },
    {
      "epoch": 0.02070527337431252,
      "grad_norm": 2.0000555515289307,
      "learning_rate": 3.2000000000000005e-05,
      "loss": 2.1615,
      "step": 16
    },
    {
      "epoch": 0.021999352960207053,
      "grad_norm": 1.5915080308914185,
      "learning_rate": 3.4000000000000007e-05,
      "loss": 1.9522,
      "step": 17
    },
    {
      "epoch": 0.023293432546101587,
      "grad_norm": 1.6972448825836182,
      "learning_rate": 3.6e-05,
      "loss": 1.7224,
      "step": 18
    },
    {
      "epoch": 0.024587512131996118,
      "grad_norm": 1.7509772777557373,
      "learning_rate": 3.8e-05,
      "loss": 2.0414,
      "step": 19
    },
    {
      "epoch": 0.02588159171789065,
      "grad_norm": 1.697340488433838,
      "learning_rate": 4e-05,
      "loss": 2.0427,
      "step": 20
    },
    {
      "epoch": 0.027175671303785183,
      "grad_norm": 1.8733758926391602,
      "learning_rate": 4.2e-05,
      "loss": 1.6772,
      "step": 21
    },
    {
      "epoch": 0.028469750889679714,
      "grad_norm": 1.6085255146026611,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 1.6527,
      "step": 22
    },
    {
      "epoch": 0.029763830475574248,
      "grad_norm": 1.5792337656021118,
      "learning_rate": 4.600000000000001e-05,
      "loss": 1.6567,
      "step": 23
    },
    {
      "epoch": 0.03105791006146878,
      "grad_norm": 1.4392567873001099,
      "learning_rate": 4.8e-05,
      "loss": 1.508,
      "step": 24
    },
    {
      "epoch": 0.03235198964736331,
      "grad_norm": 1.5222433805465698,
      "learning_rate": 5e-05,
      "loss": 1.4606,
      "step": 25
    },
    {
      "epoch": 0.033646069233257844,
      "grad_norm": 1.5480064153671265,
      "learning_rate": 5.2000000000000004e-05,
      "loss": 1.5027,
      "step": 26
    },
    {
      "epoch": 0.034940148819152375,
      "grad_norm": 1.6736445426940918,
      "learning_rate": 5.4000000000000005e-05,
      "loss": 1.2426,
      "step": 27
    },
    {
      "epoch": 0.03623422840504691,
      "grad_norm": 1.7392551898956299,
      "learning_rate": 5.6000000000000006e-05,
      "loss": 1.4703,
      "step": 28
    },
    {
      "epoch": 0.037528307990941444,
      "grad_norm": 1.6173359155654907,
      "learning_rate": 5.8e-05,
      "loss": 1.4546,
      "step": 29
    },
    {
      "epoch": 0.038822387576835975,
      "grad_norm": 1.3955802917480469,
      "learning_rate": 6e-05,
      "loss": 1.3808,
      "step": 30
    },
    {
      "epoch": 0.040116467162730506,
      "grad_norm": 1.353873372077942,
      "learning_rate": 6.2e-05,
      "loss": 1.229,
      "step": 31
    },
    {
      "epoch": 0.04141054674862504,
      "grad_norm": 1.2547746896743774,
      "learning_rate": 6.400000000000001e-05,
      "loss": 1.1668,
      "step": 32
    },
    {
      "epoch": 0.042704626334519574,
      "grad_norm": 1.3806778192520142,
      "learning_rate": 6.6e-05,
      "loss": 1.0691,
      "step": 33
    },
    {
      "epoch": 0.043998705920414105,
      "grad_norm": 1.2815773487091064,
      "learning_rate": 6.800000000000001e-05,
      "loss": 1.2409,
      "step": 34
    },
    {
      "epoch": 0.045292785506308636,
      "grad_norm": 1.3677266836166382,
      "learning_rate": 7e-05,
      "loss": 0.9668,
      "step": 35
    },
    {
      "epoch": 0.046586865092203174,
      "grad_norm": 1.5457032918930054,
      "learning_rate": 7.2e-05,
      "loss": 1.1385,
      "step": 36
    },
    {
      "epoch": 0.047880944678097705,
      "grad_norm": 1.5587060451507568,
      "learning_rate": 7.4e-05,
      "loss": 1.1707,
      "step": 37
    },
    {
      "epoch": 0.049175024263992236,
      "grad_norm": 1.079053282737732,
      "learning_rate": 7.6e-05,
      "loss": 1.0655,
      "step": 38
    },
    {
      "epoch": 0.050469103849886766,
      "grad_norm": 1.1773897409439087,
      "learning_rate": 7.800000000000001e-05,
      "loss": 1.0465,
      "step": 39
    },
    {
      "epoch": 0.0517631834357813,
      "grad_norm": 1.2437673807144165,
      "learning_rate": 8e-05,
      "loss": 1.2779,
      "step": 40
    },
    {
      "epoch": 0.053057263021675835,
      "grad_norm": 1.254847526550293,
      "learning_rate": 8.2e-05,
      "loss": 1.0898,
      "step": 41
    },
    {
      "epoch": 0.054351342607570366,
      "grad_norm": 1.1771515607833862,
      "learning_rate": 8.4e-05,
      "loss": 1.1827,
      "step": 42
    },
    {
      "epoch": 0.0556454221934649,
      "grad_norm": 1.1400648355484009,
      "learning_rate": 8.6e-05,
      "loss": 1.1066,
      "step": 43
    },
    {
      "epoch": 0.05693950177935943,
      "grad_norm": 1.2047138214111328,
      "learning_rate": 8.800000000000001e-05,
      "loss": 0.8974,
      "step": 44
    },
    {
      "epoch": 0.058233581365253966,
      "grad_norm": 1.1269346475601196,
      "learning_rate": 9e-05,
      "loss": 1.0146,
      "step": 45
    },
    {
      "epoch": 0.059527660951148496,
      "grad_norm": 1.169231653213501,
      "learning_rate": 9.200000000000001e-05,
      "loss": 1.1266,
      "step": 46
    },
    {
      "epoch": 0.06082174053704303,
      "grad_norm": 0.9771779179573059,
      "learning_rate": 9.4e-05,
      "loss": 0.8351,
      "step": 47
    },
    {
      "epoch": 0.06211582012293756,
      "grad_norm": 1.2849314212799072,
      "learning_rate": 9.6e-05,
      "loss": 1.1822,
      "step": 48
    },
    {
      "epoch": 0.0634098997088321,
      "grad_norm": 1.023181676864624,
      "learning_rate": 9.8e-05,
      "loss": 0.9082,
      "step": 49
    },
    {
      "epoch": 0.06470397929472663,
      "grad_norm": 1.135751724243164,
      "learning_rate": 0.0001,
      "loss": 0.9407,
      "step": 50
    },
    {
      "epoch": 0.06599805888062116,
      "grad_norm": 0.9701154828071594,
      "learning_rate": 9.999998300231494e-05,
      "loss": 0.9423,
      "step": 51
    },
    {
      "epoch": 0.06729213846651569,
      "grad_norm": 1.2891143560409546,
      "learning_rate": 9.999993200927133e-05,
      "loss": 0.9757,
      "step": 52
    },
    {
      "epoch": 0.06858621805241022,
      "grad_norm": 1.3360975980758667,
      "learning_rate": 9.999984702090383e-05,
      "loss": 1.0158,
      "step": 53
    },
    {
      "epoch": 0.06988029763830475,
      "grad_norm": 0.977446436882019,
      "learning_rate": 9.999972803727024e-05,
      "loss": 0.8175,
      "step": 54
    },
    {
      "epoch": 0.0711743772241993,
      "grad_norm": 0.9943827390670776,
      "learning_rate": 9.999957505845144e-05,
      "loss": 0.8627,
      "step": 55
    },
    {
      "epoch": 0.07246845681009383,
      "grad_norm": 1.1531224250793457,
      "learning_rate": 9.999938808455145e-05,
      "loss": 1.143,
      "step": 56
    },
    {
      "epoch": 0.07376253639598836,
      "grad_norm": 1.287972092628479,
      "learning_rate": 9.99991671156974e-05,
      "loss": 1.2342,
      "step": 57
    },
    {
      "epoch": 0.07505661598188289,
      "grad_norm": 1.1554590463638306,
      "learning_rate": 9.999891215203949e-05,
      "loss": 0.9692,
      "step": 58
    },
    {
      "epoch": 0.07635069556777742,
      "grad_norm": 1.0786008834838867,
      "learning_rate": 9.999862319375113e-05,
      "loss": 1.1254,
      "step": 59
    },
    {
      "epoch": 0.07764477515367195,
      "grad_norm": 1.0764508247375488,
      "learning_rate": 9.999830024102874e-05,
      "loss": 0.9312,
      "step": 60
    },
    {
      "epoch": 0.07893885473956648,
      "grad_norm": 1.1909526586532593,
      "learning_rate": 9.999794329409194e-05,
      "loss": 0.9959,
      "step": 61
    },
    {
      "epoch": 0.08023293432546101,
      "grad_norm": 0.9989166259765625,
      "learning_rate": 9.999755235318337e-05,
      "loss": 0.934,
      "step": 62
    },
    {
      "epoch": 0.08152701391135554,
      "grad_norm": 1.0302046537399292,
      "learning_rate": 9.999712741856889e-05,
      "loss": 1.1017,
      "step": 63
    },
    {
      "epoch": 0.08282109349725009,
      "grad_norm": 0.9583478569984436,
      "learning_rate": 9.999666849053738e-05,
      "loss": 1.1384,
      "step": 64
    },
    {
      "epoch": 0.08411517308314462,
      "grad_norm": 1.001126766204834,
      "learning_rate": 9.999617556940085e-05,
      "loss": 0.9279,
      "step": 65
    },
    {
      "epoch": 0.08540925266903915,
      "grad_norm": 1.0130903720855713,
      "learning_rate": 9.999564865549449e-05,
      "loss": 0.9381,
      "step": 66
    },
    {
      "epoch": 0.08670333225493368,
      "grad_norm": 1.1210829019546509,
      "learning_rate": 9.999508774917652e-05,
      "loss": 0.9607,
      "step": 67
    },
    {
      "epoch": 0.08799741184082821,
      "grad_norm": 1.045749545097351,
      "learning_rate": 9.999449285082831e-05,
      "loss": 1.0037,
      "step": 68
    },
    {
      "epoch": 0.08929149142672274,
      "grad_norm": 1.1308139562606812,
      "learning_rate": 9.999386396085434e-05,
      "loss": 0.9086,
      "step": 69
    },
    {
      "epoch": 0.09058557101261727,
      "grad_norm": 1.1013413667678833,
      "learning_rate": 9.999320107968219e-05,
      "loss": 1.0712,
      "step": 70
    },
    {
      "epoch": 0.0918796505985118,
      "grad_norm": 1.0830566883087158,
      "learning_rate": 9.999250420776258e-05,
      "loss": 1.0326,
      "step": 71
    },
    {
      "epoch": 0.09317373018440635,
      "grad_norm": 1.0673171281814575,
      "learning_rate": 9.999177334556929e-05,
      "loss": 1.0034,
      "step": 72
    },
    {
      "epoch": 0.09446780977030088,
      "grad_norm": 1.1546461582183838,
      "learning_rate": 9.999100849359926e-05,
      "loss": 1.059,
      "step": 73
    },
    {
      "epoch": 0.09576188935619541,
      "grad_norm": 0.9139528870582581,
      "learning_rate": 9.999020965237249e-05,
      "loss": 0.8596,
      "step": 74
    },
    {
      "epoch": 0.09705596894208994,
      "grad_norm": 1.1570812463760376,
      "learning_rate": 9.998937682243215e-05,
      "loss": 1.0456,
      "step": 75
    },
    {
      "epoch": 0.09835004852798447,
      "grad_norm": 1.3232612609863281,
      "learning_rate": 9.998851000434448e-05,
      "loss": 0.9994,
      "step": 76
    },
    {
      "epoch": 0.099644128113879,
      "grad_norm": 1.2017115354537964,
      "learning_rate": 9.998760919869883e-05,
      "loss": 1.2664,
      "step": 77
    },
    {
      "epoch": 0.10093820769977353,
      "grad_norm": 1.0694175958633423,
      "learning_rate": 9.998667440610765e-05,
      "loss": 0.9483,
      "step": 78
    },
    {
      "epoch": 0.10223228728566806,
      "grad_norm": 0.9963059425354004,
      "learning_rate": 9.998570562720654e-05,
      "loss": 0.9577,
      "step": 79
    },
    {
      "epoch": 0.1035263668715626,
      "grad_norm": 0.8873535394668579,
      "learning_rate": 9.998470286265416e-05,
      "loss": 0.8498,
      "step": 80
    },
    {
      "epoch": 0.10482044645745714,
      "grad_norm": 1.1350760459899902,
      "learning_rate": 9.99836661131323e-05,
      "loss": 1.0024,
      "step": 81
    },
    {
      "epoch": 0.10611452604335167,
      "grad_norm": 0.8355389833450317,
      "learning_rate": 9.998259537934586e-05,
      "loss": 0.7399,
      "step": 82
    },
    {
      "epoch": 0.1074086056292462,
      "grad_norm": 0.9935446381568909,
      "learning_rate": 9.998149066202284e-05,
      "loss": 0.9809,
      "step": 83
    },
    {
      "epoch": 0.10870268521514073,
      "grad_norm": 1.0571558475494385,
      "learning_rate": 9.998035196191435e-05,
      "loss": 1.0144,
      "step": 84
    },
    {
      "epoch": 0.10999676480103526,
      "grad_norm": 0.9860286116600037,
      "learning_rate": 9.99791792797946e-05,
      "loss": 1.0467,
      "step": 85
    },
    {
      "epoch": 0.1112908443869298,
      "grad_norm": 1.1422507762908936,
      "learning_rate": 9.997797261646089e-05,
      "loss": 0.9535,
      "step": 86
    },
    {
      "epoch": 0.11258492397282432,
      "grad_norm": 0.8561545014381409,
      "learning_rate": 9.997673197273365e-05,
      "loss": 1.007,
      "step": 87
    },
    {
      "epoch": 0.11387900355871886,
      "grad_norm": 1.0027543306350708,
      "learning_rate": 9.997545734945639e-05,
      "loss": 0.9861,
      "step": 88
    },
    {
      "epoch": 0.1151730831446134,
      "grad_norm": 0.8489773273468018,
      "learning_rate": 9.997414874749575e-05,
      "loss": 0.9672,
      "step": 89
    },
    {
      "epoch": 0.11646716273050793,
      "grad_norm": 1.0517115592956543,
      "learning_rate": 9.997280616774147e-05,
      "loss": 1.1672,
      "step": 90
    },
    {
      "epoch": 0.11776124231640246,
      "grad_norm": 1.0035395622253418,
      "learning_rate": 9.997142961110634e-05,
      "loss": 0.9294,
      "step": 91
    },
    {
      "epoch": 0.11905532190229699,
      "grad_norm": 1.1194915771484375,
      "learning_rate": 9.997001907852635e-05,
      "loss": 1.0857,
      "step": 92
    },
    {
      "epoch": 0.12034940148819152,
      "grad_norm": 1.5234825611114502,
      "learning_rate": 9.996857457096047e-05,
      "loss": 1.027,
      "step": 93
    },
    {
      "epoch": 0.12164348107408605,
      "grad_norm": 0.949878454208374,
      "learning_rate": 9.996709608939088e-05,
      "loss": 0.8173,
      "step": 94
    },
    {
      "epoch": 0.12293756065998059,
      "grad_norm": 0.8736472129821777,
      "learning_rate": 9.996558363482277e-05,
      "loss": 0.855,
      "step": 95
    },
    {
      "epoch": 0.12423164024587512,
      "grad_norm": 0.8604567050933838,
      "learning_rate": 9.996403720828449e-05,
      "loss": 0.9485,
      "step": 96
    },
    {
      "epoch": 0.12552571983176966,
      "grad_norm": 1.020851492881775,
      "learning_rate": 9.996245681082748e-05,
      "loss": 1.0024,
      "step": 97
    },
    {
      "epoch": 0.1268197994176642,
      "grad_norm": 1.0704892873764038,
      "learning_rate": 9.996084244352623e-05,
      "loss": 0.9246,
      "step": 98
    },
    {
      "epoch": 0.12811387900355872,
      "grad_norm": 0.8441987037658691,
      "learning_rate": 9.99591941074784e-05,
      "loss": 1.0343,
      "step": 99
    },
    {
      "epoch": 0.12940795858945325,
      "grad_norm": 1.0280612707138062,
      "learning_rate": 9.995751180380466e-05,
      "loss": 0.9644,
      "step": 100
    },
    {
      "epoch": 0.13070203817534778,
      "grad_norm": 0.9827906489372253,
      "learning_rate": 9.995579553364887e-05,
      "loss": 0.9583,
      "step": 101
    },
    {
      "epoch": 0.13199611776124232,
      "grad_norm": 1.035618543624878,
      "learning_rate": 9.995404529817791e-05,
      "loss": 1.0366,
      "step": 102
    },
    {
      "epoch": 0.13329019734713685,
      "grad_norm": 1.2775524854660034,
      "learning_rate": 9.995226109858178e-05,
      "loss": 0.9353,
      "step": 103
    },
    {
      "epoch": 0.13458427693303138,
      "grad_norm": 1.0101919174194336,
      "learning_rate": 9.995044293607355e-05,
      "loss": 0.9045,
      "step": 104
    },
    {
      "epoch": 0.1358783565189259,
      "grad_norm": 0.8396942019462585,
      "learning_rate": 9.994859081188943e-05,
      "loss": 0.867,
      "step": 105
    },
    {
      "epoch": 0.13717243610482044,
      "grad_norm": 1.04515540599823,
      "learning_rate": 9.99467047272887e-05,
      "loss": 0.9693,
      "step": 106
    },
    {
      "epoch": 0.13846651569071497,
      "grad_norm": 1.099042534828186,
      "learning_rate": 9.994478468355369e-05,
      "loss": 0.8879,
      "step": 107
    },
    {
      "epoch": 0.1397605952766095,
      "grad_norm": 0.8710360527038574,
      "learning_rate": 9.994283068198988e-05,
      "loss": 0.9018,
      "step": 108
    },
    {
      "epoch": 0.14105467486250403,
      "grad_norm": 0.961025059223175,
      "learning_rate": 9.99408427239258e-05,
      "loss": 0.8806,
      "step": 109
    },
    {
      "epoch": 0.1423487544483986,
      "grad_norm": 0.915665328502655,
      "learning_rate": 9.993882081071306e-05,
      "loss": 0.8628,
      "step": 110
    },
    {
      "epoch": 0.14364283403429312,
      "grad_norm": 1.2776648998260498,
      "learning_rate": 9.993676494372642e-05,
      "loss": 0.9742,
      "step": 111
    },
    {
      "epoch": 0.14493691362018765,
      "grad_norm": 1.1270071268081665,
      "learning_rate": 9.993467512436364e-05,
      "loss": 0.9729,
      "step": 112
    },
    {
      "epoch": 0.14623099320608218,
      "grad_norm": 0.8188664317131042,
      "learning_rate": 9.99325513540456e-05,
      "loss": 0.9428,
      "step": 113
    },
    {
      "epoch": 0.1475250727919767,
      "grad_norm": 1.0760393142700195,
      "learning_rate": 9.993039363421627e-05,
      "loss": 0.9482,
      "step": 114
    },
    {
      "epoch": 0.14881915237787124,
      "grad_norm": 1.019920825958252,
      "learning_rate": 9.992820196634273e-05,
      "loss": 0.9785,
      "step": 115
    },
    {
      "epoch": 0.15011323196376578,
      "grad_norm": 0.8342046737670898,
      "learning_rate": 9.992597635191509e-05,
      "loss": 0.9291,
      "step": 116
    },
    {
      "epoch": 0.1514073115496603,
      "grad_norm": 0.8460632562637329,
      "learning_rate": 9.992371679244658e-05,
      "loss": 0.8797,
      "step": 117
    },
    {
      "epoch": 0.15270139113555484,
      "grad_norm": 0.933060348033905,
      "learning_rate": 9.992142328947345e-05,
      "loss": 0.9657,
      "step": 118
    },
    {
      "epoch": 0.15399547072144937,
      "grad_norm": 0.8822593688964844,
      "learning_rate": 9.991909584455511e-05,
      "loss": 0.8872,
      "step": 119
    },
    {
      "epoch": 0.1552895503073439,
      "grad_norm": 0.9599350094795227,
      "learning_rate": 9.991673445927398e-05,
      "loss": 0.9064,
      "step": 120
    },
    {
      "epoch": 0.15658362989323843,
      "grad_norm": 0.8505874872207642,
      "learning_rate": 9.99143391352356e-05,
      "loss": 0.9966,
      "step": 121
    },
    {
      "epoch": 0.15787770947913296,
      "grad_norm": 1.3977786302566528,
      "learning_rate": 9.991190987406857e-05,
      "loss": 0.9145,
      "step": 122
    },
    {
      "epoch": 0.1591717890650275,
      "grad_norm": 0.8947294354438782,
      "learning_rate": 9.990944667742455e-05,
      "loss": 0.9569,
      "step": 123
    },
    {
      "epoch": 0.16046586865092202,
      "grad_norm": 0.7973839044570923,
      "learning_rate": 9.990694954697828e-05,
      "loss": 0.8853,
      "step": 124
    },
    {
      "epoch": 0.16175994823681655,
      "grad_norm": 0.9481159448623657,
      "learning_rate": 9.99044184844276e-05,
      "loss": 1.04,
      "step": 125
    },
    {
      "epoch": 0.16305402782271108,
      "grad_norm": 1.3568611145019531,
      "learning_rate": 9.990185349149339e-05,
      "loss": 1.1104,
      "step": 126
    },
    {
      "epoch": 0.16434810740860564,
      "grad_norm": 0.900867223739624,
      "learning_rate": 9.98992545699196e-05,
      "loss": 0.8427,
      "step": 127
    },
    {
      "epoch": 0.16564218699450017,
      "grad_norm": 0.9025059938430786,
      "learning_rate": 9.989662172147326e-05,
      "loss": 0.9671,
      "step": 128
    },
    {
      "epoch": 0.1669362665803947,
      "grad_norm": 0.944692850112915,
      "learning_rate": 9.989395494794446e-05,
      "loss": 1.0966,
      "step": 129
    },
    {
      "epoch": 0.16823034616628924,
      "grad_norm": 1.196311116218567,
      "learning_rate": 9.989125425114638e-05,
      "loss": 1.0888,
      "step": 130
    },
    {
      "epoch": 0.16952442575218377,
      "grad_norm": 0.9069584608078003,
      "learning_rate": 9.988851963291522e-05,
      "loss": 0.8579,
      "step": 131
    },
    {
      "epoch": 0.1708185053380783,
      "grad_norm": 0.8150789141654968,
      "learning_rate": 9.988575109511026e-05,
      "loss": 0.7622,
      "step": 132
    },
    {
      "epoch": 0.17211258492397283,
      "grad_norm": 1.0844395160675049,
      "learning_rate": 9.988294863961387e-05,
      "loss": 0.9284,
      "step": 133
    },
    {
      "epoch": 0.17340666450986736,
      "grad_norm": 1.0463049411773682,
      "learning_rate": 9.988011226833146e-05,
      "loss": 0.9185,
      "step": 134
    },
    {
      "epoch": 0.1747007440957619,
      "grad_norm": 0.9481234550476074,
      "learning_rate": 9.987724198319148e-05,
      "loss": 0.8631,
      "step": 135
    },
    {
      "epoch": 0.17599482368165642,
      "grad_norm": 0.882074773311615,
      "learning_rate": 9.987433778614549e-05,
      "loss": 0.8997,
      "step": 136
    },
    {
      "epoch": 0.17728890326755095,
      "grad_norm": 0.9853332042694092,
      "learning_rate": 9.987139967916805e-05,
      "loss": 0.9226,
      "step": 137
    },
    {
      "epoch": 0.17858298285344548,
      "grad_norm": 1.151941180229187,
      "learning_rate": 9.98684276642568e-05,
      "loss": 1.0486,
      "step": 138
    },
    {
      "epoch": 0.17987706243934,
      "grad_norm": 1.0128459930419922,
      "learning_rate": 9.986542174343245e-05,
      "loss": 1.0797,
      "step": 139
    },
    {
      "epoch": 0.18117114202523454,
      "grad_norm": 0.9798718094825745,
      "learning_rate": 9.986238191873874e-05,
      "loss": 0.875,
      "step": 140
    },
    {
      "epoch": 0.18246522161112907,
      "grad_norm": 0.8143295645713806,
      "learning_rate": 9.985930819224247e-05,
      "loss": 0.8454,
      "step": 141
    },
    {
      "epoch": 0.1837593011970236,
      "grad_norm": 0.8755755424499512,
      "learning_rate": 9.985620056603348e-05,
      "loss": 0.8029,
      "step": 142
    },
    {
      "epoch": 0.18505338078291814,
      "grad_norm": 0.899174690246582,
      "learning_rate": 9.985305904222469e-05,
      "loss": 0.9608,
      "step": 143
    },
    {
      "epoch": 0.1863474603688127,
      "grad_norm": 0.920137882232666,
      "learning_rate": 9.984988362295203e-05,
      "loss": 0.9022,
      "step": 144
    },
    {
      "epoch": 0.18764153995470723,
      "grad_norm": 1.1012908220291138,
      "learning_rate": 9.984667431037447e-05,
      "loss": 0.9621,
      "step": 145
    },
    {
      "epoch": 0.18893561954060176,
      "grad_norm": 0.8609358668327332,
      "learning_rate": 9.98434311066741e-05,
      "loss": 0.917,
      "step": 146
    },
    {
      "epoch": 0.1902296991264963,
      "grad_norm": 0.8248727321624756,
      "learning_rate": 9.984015401405594e-05,
      "loss": 0.7864,
      "step": 147
    },
    {
      "epoch": 0.19152377871239082,
      "grad_norm": 0.8680225610733032,
      "learning_rate": 9.983684303474815e-05,
      "loss": 0.9288,
      "step": 148
    },
    {
      "epoch": 0.19281785829828535,
      "grad_norm": 1.0807067155838013,
      "learning_rate": 9.983349817100188e-05,
      "loss": 0.9842,
      "step": 149
    },
    {
      "epoch": 0.19411193788417988,
      "grad_norm": 0.9310898780822754,
      "learning_rate": 9.983011942509131e-05,
      "loss": 1.0568,
      "step": 150
    },
    {
      "epoch": 0.1954060174700744,
      "grad_norm": 0.8052242398262024,
      "learning_rate": 9.98267067993137e-05,
      "loss": 0.8235,
      "step": 151
    },
    {
      "epoch": 0.19670009705596894,
      "grad_norm": 0.9700384140014648,
      "learning_rate": 9.982326029598931e-05,
      "loss": 0.8611,
      "step": 152
    },
    {
      "epoch": 0.19799417664186347,
      "grad_norm": 0.8437764048576355,
      "learning_rate": 9.981977991746142e-05,
      "loss": 0.83,
      "step": 153
    },
    {
      "epoch": 0.199288256227758,
      "grad_norm": 0.930636465549469,
      "learning_rate": 9.98162656660964e-05,
      "loss": 1.0892,
      "step": 154
    },
    {
      "epoch": 0.20058233581365253,
      "grad_norm": 0.9111954569816589,
      "learning_rate": 9.98127175442836e-05,
      "loss": 0.9962,
      "step": 155
    },
    {
      "epoch": 0.20187641539954707,
      "grad_norm": 0.9521974921226501,
      "learning_rate": 9.980913555443541e-05,
      "loss": 0.911,
      "step": 156
    },
    {
      "epoch": 0.2031704949854416,
      "grad_norm": 0.8516745567321777,
      "learning_rate": 9.980551969898727e-05,
      "loss": 0.9009,
      "step": 157
    },
    {
      "epoch": 0.20446457457133613,
      "grad_norm": 0.8302998542785645,
      "learning_rate": 9.98018699803976e-05,
      "loss": 0.8854,
      "step": 158
    },
    {
      "epoch": 0.20575865415723066,
      "grad_norm": 0.814391016960144,
      "learning_rate": 9.979818640114789e-05,
      "loss": 0.9601,
      "step": 159
    },
    {
      "epoch": 0.2070527337431252,
      "grad_norm": 0.8938564658164978,
      "learning_rate": 9.979446896374262e-05,
      "loss": 0.8834,
      "step": 160
    },
    {
      "epoch": 0.20834681332901975,
      "grad_norm": 0.9066985249519348,
      "learning_rate": 9.979071767070932e-05,
      "loss": 0.7427,
      "step": 161
    },
    {
      "epoch": 0.20964089291491428,
      "grad_norm": 0.7866595983505249,
      "learning_rate": 9.978693252459851e-05,
      "loss": 0.8556,
      "step": 162
    },
    {
      "epoch": 0.2109349725008088,
      "grad_norm": 0.9159708023071289,
      "learning_rate": 9.978311352798374e-05,
      "loss": 0.8101,
      "step": 163
    },
    {
      "epoch": 0.21222905208670334,
      "grad_norm": 1.1350793838500977,
      "learning_rate": 9.977926068346157e-05,
      "loss": 0.9374,
      "step": 164
    },
    {
      "epoch": 0.21352313167259787,
      "grad_norm": 1.0535932779312134,
      "learning_rate": 9.977537399365159e-05,
      "loss": 1.0238,
      "step": 165
    },
    {
      "epoch": 0.2148172112584924,
      "grad_norm": 0.8717033267021179,
      "learning_rate": 9.977145346119637e-05,
      "loss": 1.0265,
      "step": 166
    },
    {
      "epoch": 0.21611129084438693,
      "grad_norm": 0.8357003927230835,
      "learning_rate": 9.976749908876152e-05,
      "loss": 0.9016,
      "step": 167
    },
    {
      "epoch": 0.21740537043028146,
      "grad_norm": 0.8369495868682861,
      "learning_rate": 9.976351087903568e-05,
      "loss": 0.8764,
      "step": 168
    },
    {
      "epoch": 0.218699450016176,
      "grad_norm": 0.912352979183197,
      "learning_rate": 9.97594888347304e-05,
      "loss": 0.9078,
      "step": 169
    },
    {
      "epoch": 0.21999352960207053,
      "grad_norm": 0.8475804328918457,
      "learning_rate": 9.975543295858035e-05,
      "loss": 0.8836,
      "step": 170
    },
    {
      "epoch": 0.22128760918796506,
      "grad_norm": 0.8391397595405579,
      "learning_rate": 9.97513432533431e-05,
      "loss": 0.9003,
      "step": 171
    },
    {
      "epoch": 0.2225816887738596,
      "grad_norm": 0.9666821360588074,
      "learning_rate": 9.974721972179931e-05,
      "loss": 0.9528,
      "step": 172
    },
    {
      "epoch": 0.22387576835975412,
      "grad_norm": 0.9321691393852234,
      "learning_rate": 9.974306236675259e-05,
      "loss": 0.9575,
      "step": 173
    },
    {
      "epoch": 0.22516984794564865,
      "grad_norm": 0.8022271990776062,
      "learning_rate": 9.973887119102957e-05,
      "loss": 0.8731,
      "step": 174
    },
    {
      "epoch": 0.22646392753154318,
      "grad_norm": 1.1056872606277466,
      "learning_rate": 9.973464619747983e-05,
      "loss": 0.9925,
      "step": 175
    },
    {
      "epoch": 0.2277580071174377,
      "grad_norm": 0.810420036315918,
      "learning_rate": 9.9730387388976e-05,
      "loss": 1.0073,
      "step": 176
    },
    {
      "epoch": 0.22905208670333224,
      "grad_norm": 0.9536454677581787,
      "learning_rate": 9.972609476841367e-05,
      "loss": 0.9595,
      "step": 177
    },
    {
      "epoch": 0.2303461662892268,
      "grad_norm": 0.8205066919326782,
      "learning_rate": 9.972176833871142e-05,
      "loss": 0.8146,
      "step": 178
    },
    {
      "epoch": 0.23164024587512133,
      "grad_norm": 0.9716495275497437,
      "learning_rate": 9.971740810281083e-05,
      "loss": 1.0377,
      "step": 179
    },
    {
      "epoch": 0.23293432546101586,
      "grad_norm": 0.828642725944519,
      "learning_rate": 9.971301406367644e-05,
      "loss": 0.8619,
      "step": 180
    },
    {
      "epoch": 0.2342284050469104,
      "grad_norm": 0.6980477571487427,
      "learning_rate": 9.970858622429579e-05,
      "loss": 0.8271,
      "step": 181
    },
    {
      "epoch": 0.23552248463280492,
      "grad_norm": 0.954387903213501,
      "learning_rate": 9.970412458767943e-05,
      "loss": 0.8465,
      "step": 182
    },
    {
      "epoch": 0.23681656421869945,
      "grad_norm": 0.8425692915916443,
      "learning_rate": 9.969962915686083e-05,
      "loss": 0.8893,
      "step": 183
    },
    {
      "epoch": 0.23811064380459399,
      "grad_norm": 0.8565071225166321,
      "learning_rate": 9.969509993489647e-05,
      "loss": 0.939,
      "step": 184
    },
    {
      "epoch": 0.23940472339048852,
      "grad_norm": 0.8831691145896912,
      "learning_rate": 9.969053692486583e-05,
      "loss": 0.8907,
      "step": 185
    },
    {
      "epoch": 0.24069880297638305,
      "grad_norm": 0.9661678075790405,
      "learning_rate": 9.96859401298713e-05,
      "loss": 0.9191,
      "step": 186
    },
    {
      "epoch": 0.24199288256227758,
      "grad_norm": 0.8784729838371277,
      "learning_rate": 9.968130955303828e-05,
      "loss": 1.0393,
      "step": 187
    },
    {
      "epoch": 0.2432869621481721,
      "grad_norm": 0.8830071091651917,
      "learning_rate": 9.967664519751515e-05,
      "loss": 0.9837,
      "step": 188
    },
    {
      "epoch": 0.24458104173406664,
      "grad_norm": 0.862108588218689,
      "learning_rate": 9.967194706647322e-05,
      "loss": 0.7871,
      "step": 189
    },
    {
      "epoch": 0.24587512131996117,
      "grad_norm": 1.0068063735961914,
      "learning_rate": 9.966721516310682e-05,
      "loss": 0.9526,
      "step": 190
    },
    {
      "epoch": 0.2471692009058557,
      "grad_norm": 0.9828710556030273,
      "learning_rate": 9.966244949063316e-05,
      "loss": 0.8923,
      "step": 191
    },
    {
      "epoch": 0.24846328049175023,
      "grad_norm": 1.0729883909225464,
      "learning_rate": 9.965765005229248e-05,
      "loss": 1.0115,
      "step": 192
    },
    {
      "epoch": 0.24975736007764476,
      "grad_norm": 0.9844326972961426,
      "learning_rate": 9.965281685134796e-05,
      "loss": 0.9855,
      "step": 193
    },
    {
      "epoch": 0.2510514396635393,
      "grad_norm": 1.1593172550201416,
      "learning_rate": 9.96479498910857e-05,
      "loss": 1.0912,
      "step": 194
    },
    {
      "epoch": 0.2523455192494338,
      "grad_norm": 0.8835370540618896,
      "learning_rate": 9.964304917481482e-05,
      "loss": 0.9951,
      "step": 195
    },
    {
      "epoch": 0.2536395988353284,
      "grad_norm": 0.9553850889205933,
      "learning_rate": 9.963811470586733e-05,
      "loss": 0.9335,
      "step": 196
    },
    {
      "epoch": 0.2549336784212229,
      "grad_norm": 0.863814115524292,
      "learning_rate": 9.963314648759823e-05,
      "loss": 1.0203,
      "step": 197
    },
    {
      "epoch": 0.25622775800711745,
      "grad_norm": 0.9639378786087036,
      "learning_rate": 9.962814452338542e-05,
      "loss": 1.0357,
      "step": 198
    },
    {
      "epoch": 0.25752183759301195,
      "grad_norm": 0.880519688129425,
      "learning_rate": 9.96231088166298e-05,
      "loss": 0.9964,
      "step": 199
    },
    {
      "epoch": 0.2588159171789065,
      "grad_norm": 0.8445360064506531,
      "learning_rate": 9.961803937075516e-05,
      "loss": 0.9766,
      "step": 200
    },
    {
      "epoch": 0.260109996764801,
      "grad_norm": 0.8204835057258606,
      "learning_rate": 9.961293618920826e-05,
      "loss": 0.8864,
      "step": 201
    },
    {
      "epoch": 0.26140407635069557,
      "grad_norm": 0.9315406084060669,
      "learning_rate": 9.960779927545883e-05,
      "loss": 1.0388,
      "step": 202
    },
    {
      "epoch": 0.2626981559365901,
      "grad_norm": 0.9286762475967407,
      "learning_rate": 9.960262863299943e-05,
      "loss": 0.9653,
      "step": 203
    },
    {
      "epoch": 0.26399223552248463,
      "grad_norm": 0.8037816882133484,
      "learning_rate": 9.959742426534566e-05,
      "loss": 0.765,
      "step": 204
    },
    {
      "epoch": 0.2652863151083792,
      "grad_norm": 0.9435904622077942,
      "learning_rate": 9.9592186176036e-05,
      "loss": 0.8889,
      "step": 205
    },
    {
      "epoch": 0.2665803946942737,
      "grad_norm": 1.0072762966156006,
      "learning_rate": 9.958691436863188e-05,
      "loss": 0.8358,
      "step": 206
    },
    {
      "epoch": 0.26787447428016825,
      "grad_norm": 0.9463568329811096,
      "learning_rate": 9.958160884671761e-05,
      "loss": 0.8815,
      "step": 207
    },
    {
      "epoch": 0.26916855386606275,
      "grad_norm": 0.9203188419342041,
      "learning_rate": 9.957626961390047e-05,
      "loss": 0.9312,
      "step": 208
    },
    {
      "epoch": 0.2704626334519573,
      "grad_norm": 1.0614677667617798,
      "learning_rate": 9.957089667381064e-05,
      "loss": 0.9822,
      "step": 209
    },
    {
      "epoch": 0.2717567130378518,
      "grad_norm": 0.8971818089485168,
      "learning_rate": 9.956549003010123e-05,
      "loss": 0.9421,
      "step": 210
    },
    {
      "epoch": 0.2730507926237464,
      "grad_norm": 0.9978768825531006,
      "learning_rate": 9.956004968644825e-05,
      "loss": 0.9539,
      "step": 211
    },
    {
      "epoch": 0.2743448722096409,
      "grad_norm": 0.7017274498939514,
      "learning_rate": 9.955457564655064e-05,
      "loss": 0.665,
      "step": 212
    },
    {
      "epoch": 0.27563895179553544,
      "grad_norm": 0.8292055726051331,
      "learning_rate": 9.954906791413023e-05,
      "loss": 0.922,
      "step": 213
    },
    {
      "epoch": 0.27693303138142994,
      "grad_norm": 0.978084146976471,
      "learning_rate": 9.954352649293178e-05,
      "loss": 0.9465,
      "step": 214
    },
    {
      "epoch": 0.2782271109673245,
      "grad_norm": 1.0260313749313354,
      "learning_rate": 9.953795138672291e-05,
      "loss": 0.9093,
      "step": 215
    },
    {
      "epoch": 0.279521190553219,
      "grad_norm": 1.162850022315979,
      "learning_rate": 9.95323425992942e-05,
      "loss": 1.0372,
      "step": 216
    },
    {
      "epoch": 0.28081527013911356,
      "grad_norm": 0.9785279631614685,
      "learning_rate": 9.952670013445913e-05,
      "loss": 0.8818,
      "step": 217
    },
    {
      "epoch": 0.28210934972500806,
      "grad_norm": 0.9386499524116516,
      "learning_rate": 9.9521023996054e-05,
      "loss": 0.8711,
      "step": 218
    },
    {
      "epoch": 0.2834034293109026,
      "grad_norm": 0.8620506525039673,
      "learning_rate": 9.951531418793812e-05,
      "loss": 1.011,
      "step": 219
    },
    {
      "epoch": 0.2846975088967972,
      "grad_norm": 0.9523435831069946,
      "learning_rate": 9.950957071399357e-05,
      "loss": 0.8541,
      "step": 220
    },
    {
      "epoch": 0.2859915884826917,
      "grad_norm": 0.8993477821350098,
      "learning_rate": 9.950379357812543e-05,
      "loss": 1.0253,
      "step": 221
    },
    {
      "epoch": 0.28728566806858624,
      "grad_norm": 1.073880910873413,
      "learning_rate": 9.949798278426158e-05,
      "loss": 1.115,
      "step": 222
    },
    {
      "epoch": 0.28857974765448074,
      "grad_norm": 0.7941976189613342,
      "learning_rate": 9.949213833635285e-05,
      "loss": 0.9398,
      "step": 223
    },
    {
      "epoch": 0.2898738272403753,
      "grad_norm": 0.798089325428009,
      "learning_rate": 9.948626023837291e-05,
      "loss": 0.8523,
      "step": 224
    },
    {
      "epoch": 0.2911679068262698,
      "grad_norm": 1.0251280069351196,
      "learning_rate": 9.948034849431831e-05,
      "loss": 0.939,
      "step": 225
    },
    {
      "epoch": 0.29246198641216437,
      "grad_norm": 0.9793195724487305,
      "learning_rate": 9.947440310820852e-05,
      "loss": 1.0998,
      "step": 226
    },
    {
      "epoch": 0.29375606599805887,
      "grad_norm": 0.8190125823020935,
      "learning_rate": 9.946842408408583e-05,
      "loss": 0.9606,
      "step": 227
    },
    {
      "epoch": 0.2950501455839534,
      "grad_norm": 0.8229602575302124,
      "learning_rate": 9.946241142601543e-05,
      "loss": 0.7944,
      "step": 228
    },
    {
      "epoch": 0.29634422516984793,
      "grad_norm": 0.8640865683555603,
      "learning_rate": 9.945636513808537e-05,
      "loss": 1.112,
      "step": 229
    },
    {
      "epoch": 0.2976383047557425,
      "grad_norm": 0.774501621723175,
      "learning_rate": 9.945028522440653e-05,
      "loss": 0.8986,
      "step": 230
    },
    {
      "epoch": 0.298932384341637,
      "grad_norm": 0.9039688110351562,
      "learning_rate": 9.944417168911275e-05,
      "loss": 1.0461,
      "step": 231
    },
    {
      "epoch": 0.30022646392753155,
      "grad_norm": 0.8048250675201416,
      "learning_rate": 9.943802453636065e-05,
      "loss": 0.89,
      "step": 232
    },
    {
      "epoch": 0.30152054351342605,
      "grad_norm": 0.8166521787643433,
      "learning_rate": 9.94318437703297e-05,
      "loss": 0.9149,
      "step": 233
    },
    {
      "epoch": 0.3028146230993206,
      "grad_norm": 0.7571333646774292,
      "learning_rate": 9.942562939522228e-05,
      "loss": 0.9655,
      "step": 234
    },
    {
      "epoch": 0.3041087026852151,
      "grad_norm": 0.6913223266601562,
      "learning_rate": 9.941938141526354e-05,
      "loss": 0.869,
      "step": 235
    },
    {
      "epoch": 0.3054027822711097,
      "grad_norm": 0.8476676344871521,
      "learning_rate": 9.94130998347016e-05,
      "loss": 0.8849,
      "step": 236
    },
    {
      "epoch": 0.30669686185700423,
      "grad_norm": 0.8454031944274902,
      "learning_rate": 9.940678465780728e-05,
      "loss": 0.9102,
      "step": 237
    },
    {
      "epoch": 0.30799094144289874,
      "grad_norm": 0.8514583706855774,
      "learning_rate": 9.940043588887438e-05,
      "loss": 0.9723,
      "step": 238
    },
    {
      "epoch": 0.3092850210287933,
      "grad_norm": 0.7330415844917297,
      "learning_rate": 9.939405353221942e-05,
      "loss": 0.9537,
      "step": 239
    },
    {
      "epoch": 0.3105791006146878,
      "grad_norm": 0.9652897715568542,
      "learning_rate": 9.938763759218185e-05,
      "loss": 0.9736,
      "step": 240
    },
    {
      "epoch": 0.31187318020058236,
      "grad_norm": 0.7517886161804199,
      "learning_rate": 9.93811880731239e-05,
      "loss": 0.8436,
      "step": 241
    },
    {
      "epoch": 0.31316725978647686,
      "grad_norm": 0.8159210681915283,
      "learning_rate": 9.937470497943064e-05,
      "loss": 0.7521,
      "step": 242
    },
    {
      "epoch": 0.3144613393723714,
      "grad_norm": 0.9554911851882935,
      "learning_rate": 9.936818831550998e-05,
      "loss": 1.1076,
      "step": 243
    },
    {
      "epoch": 0.3157554189582659,
      "grad_norm": 0.8745877742767334,
      "learning_rate": 9.936163808579266e-05,
      "loss": 0.8908,
      "step": 244
    },
    {
      "epoch": 0.3170494985441605,
      "grad_norm": 0.8050674200057983,
      "learning_rate": 9.93550542947322e-05,
      "loss": 1.0134,
      "step": 245
    },
    {
      "epoch": 0.318343578130055,
      "grad_norm": 1.009790062904358,
      "learning_rate": 9.9348436946805e-05,
      "loss": 1.0264,
      "step": 246
    },
    {
      "epoch": 0.31963765771594954,
      "grad_norm": 0.8702448606491089,
      "learning_rate": 9.934178604651023e-05,
      "loss": 1.0067,
      "step": 247
    },
    {
      "epoch": 0.32093173730184404,
      "grad_norm": 0.8105303049087524,
      "learning_rate": 9.933510159836989e-05,
      "loss": 0.8121,
      "step": 248
    },
    {
      "epoch": 0.3222258168877386,
      "grad_norm": 0.7680085897445679,
      "learning_rate": 9.932838360692878e-05,
      "loss": 0.8951,
      "step": 249
    },
    {
      "epoch": 0.3235198964736331,
      "grad_norm": 0.8338052034378052,
      "learning_rate": 9.93216320767545e-05,
      "loss": 0.8878,
      "step": 250
    },
    {
      "epoch": 0.32481397605952766,
      "grad_norm": 0.8709661960601807,
      "learning_rate": 9.93148470124375e-05,
      "loss": 0.8786,
      "step": 251
    },
    {
      "epoch": 0.32610805564542217,
      "grad_norm": 0.9593453407287598,
      "learning_rate": 9.930802841859095e-05,
      "loss": 1.0659,
      "step": 252
    },
    {
      "epoch": 0.3274021352313167,
      "grad_norm": 1.0130974054336548,
      "learning_rate": 9.93011762998509e-05,
      "loss": 0.9626,
      "step": 253
    },
    {
      "epoch": 0.3286962148172113,
      "grad_norm": 0.9949910640716553,
      "learning_rate": 9.929429066087616e-05,
      "loss": 1.0499,
      "step": 254
    },
    {
      "epoch": 0.3299902944031058,
      "grad_norm": 0.927542507648468,
      "learning_rate": 9.92873715063483e-05,
      "loss": 0.9571,
      "step": 255
    },
    {
      "epoch": 0.33128437398900035,
      "grad_norm": 1.0638468265533447,
      "learning_rate": 9.92804188409717e-05,
      "loss": 1.0293,
      "step": 256
    },
    {
      "epoch": 0.33257845357489485,
      "grad_norm": 0.7083877325057983,
      "learning_rate": 9.927343266947356e-05,
      "loss": 0.875,
      "step": 257
    },
    {
      "epoch": 0.3338725331607894,
      "grad_norm": 0.7915517091751099,
      "learning_rate": 9.92664129966038e-05,
      "loss": 0.8848,
      "step": 258
    },
    {
      "epoch": 0.3351666127466839,
      "grad_norm": 0.8054295182228088,
      "learning_rate": 9.925935982713518e-05,
      "loss": 0.8981,
      "step": 259
    },
    {
      "epoch": 0.33646069233257847,
      "grad_norm": 1.012574553489685,
      "learning_rate": 9.925227316586316e-05,
      "loss": 0.8119,
      "step": 260
    },
    {
      "epoch": 0.337754771918473,
      "grad_norm": 0.8329979181289673,
      "learning_rate": 9.924515301760606e-05,
      "loss": 0.8467,
      "step": 261
    },
    {
      "epoch": 0.33904885150436753,
      "grad_norm": 0.801017701625824,
      "learning_rate": 9.923799938720488e-05,
      "loss": 0.8333,
      "step": 262
    },
    {
      "epoch": 0.34034293109026204,
      "grad_norm": 0.9083892703056335,
      "learning_rate": 9.923081227952347e-05,
      "loss": 0.8727,
      "step": 263
    },
    {
      "epoch": 0.3416370106761566,
      "grad_norm": 0.7917154431343079,
      "learning_rate": 9.922359169944834e-05,
      "loss": 1.0341,
      "step": 264
    },
    {
      "epoch": 0.3429310902620511,
      "grad_norm": 0.6865798234939575,
      "learning_rate": 9.921633765188886e-05,
      "loss": 0.9117,
      "step": 265
    },
    {
      "epoch": 0.34422516984794566,
      "grad_norm": 0.9448872208595276,
      "learning_rate": 9.92090501417771e-05,
      "loss": 1.1399,
      "step": 266
    },
    {
      "epoch": 0.34551924943384016,
      "grad_norm": 0.8711137175559998,
      "learning_rate": 9.920172917406789e-05,
      "loss": 0.8158,
      "step": 267
    },
    {
      "epoch": 0.3468133290197347,
      "grad_norm": 0.7883780598640442,
      "learning_rate": 9.919437475373882e-05,
      "loss": 1.0259,
      "step": 268
    },
    {
      "epoch": 0.3481074086056292,
      "grad_norm": 0.7776859402656555,
      "learning_rate": 9.91869868857902e-05,
      "loss": 0.9145,
      "step": 269
    },
    {
      "epoch": 0.3494014881915238,
      "grad_norm": 0.8383511900901794,
      "learning_rate": 9.91795655752451e-05,
      "loss": 0.8769,
      "step": 270
    },
    {
      "epoch": 0.35069556777741834,
      "grad_norm": 0.8562968373298645,
      "learning_rate": 9.917211082714933e-05,
      "loss": 1.0078,
      "step": 271
    },
    {
      "epoch": 0.35198964736331284,
      "grad_norm": 0.810517430305481,
      "learning_rate": 9.916462264657142e-05,
      "loss": 0.8847,
      "step": 272
    },
    {
      "epoch": 0.3532837269492074,
      "grad_norm": 0.8934103846549988,
      "learning_rate": 9.915710103860263e-05,
      "loss": 0.8633,
      "step": 273
    },
    {
      "epoch": 0.3545778065351019,
      "grad_norm": 0.8963167667388916,
      "learning_rate": 9.914954600835699e-05,
      "loss": 0.8411,
      "step": 274
    },
    {
      "epoch": 0.35587188612099646,
      "grad_norm": 0.774557888507843,
      "learning_rate": 9.91419575609712e-05,
      "loss": 0.958,
      "step": 275
    },
    {
      "epoch": 0.35716596570689096,
      "grad_norm": 0.8483523726463318,
      "learning_rate": 9.913433570160469e-05,
      "loss": 0.905,
      "step": 276
    },
    {
      "epoch": 0.3584600452927855,
      "grad_norm": 0.8082010746002197,
      "learning_rate": 9.912668043543964e-05,
      "loss": 0.8633,
      "step": 277
    },
    {
      "epoch": 0.35975412487868,
      "grad_norm": 0.8654133677482605,
      "learning_rate": 9.911899176768091e-05,
      "loss": 0.8465,
      "step": 278
    },
    {
      "epoch": 0.3610482044645746,
      "grad_norm": 0.8709694743156433,
      "learning_rate": 9.911126970355609e-05,
      "loss": 0.9286,
      "step": 279
    },
    {
      "epoch": 0.3623422840504691,
      "grad_norm": 0.8783992528915405,
      "learning_rate": 9.910351424831546e-05,
      "loss": 0.9349,
      "step": 280
    },
    {
      "epoch": 0.36363636363636365,
      "grad_norm": 0.8846459984779358,
      "learning_rate": 9.909572540723202e-05,
      "loss": 0.7986,
      "step": 281
    },
    {
      "epoch": 0.36493044322225815,
      "grad_norm": 0.8271780610084534,
      "learning_rate": 9.908790318560146e-05,
      "loss": 0.8179,
      "step": 282
    },
    {
      "epoch": 0.3662245228081527,
      "grad_norm": 0.8694506883621216,
      "learning_rate": 9.908004758874216e-05,
      "loss": 0.8453,
      "step": 283
    },
    {
      "epoch": 0.3675186023940472,
      "grad_norm": 0.8625207543373108,
      "learning_rate": 9.90721586219952e-05,
      "loss": 0.7825,
      "step": 284
    },
    {
      "epoch": 0.36881268197994177,
      "grad_norm": 0.8270084261894226,
      "learning_rate": 9.906423629072434e-05,
      "loss": 1.0889,
      "step": 285
    },
    {
      "epoch": 0.3701067615658363,
      "grad_norm": 0.8891452550888062,
      "learning_rate": 9.905628060031605e-05,
      "loss": 0.8847,
      "step": 286
    },
    {
      "epoch": 0.37140084115173083,
      "grad_norm": 0.8682214021682739,
      "learning_rate": 9.904829155617945e-05,
      "loss": 0.9311,
      "step": 287
    },
    {
      "epoch": 0.3726949207376254,
      "grad_norm": 1.2218021154403687,
      "learning_rate": 9.904026916374636e-05,
      "loss": 0.92,
      "step": 288
    },
    {
      "epoch": 0.3739890003235199,
      "grad_norm": 1.0069034099578857,
      "learning_rate": 9.903221342847125e-05,
      "loss": 1.0061,
      "step": 289
    },
    {
      "epoch": 0.37528307990941445,
      "grad_norm": 0.7723405361175537,
      "learning_rate": 9.902412435583128e-05,
      "loss": 0.8627,
      "step": 290
    },
    {
      "epoch": 0.37657715949530896,
      "grad_norm": 0.8336161971092224,
      "learning_rate": 9.901600195132627e-05,
      "loss": 0.9815,
      "step": 291
    },
    {
      "epoch": 0.3778712390812035,
      "grad_norm": 0.8170490264892578,
      "learning_rate": 9.90078462204787e-05,
      "loss": 0.8305,
      "step": 292
    },
    {
      "epoch": 0.379165318667098,
      "grad_norm": 0.7367318868637085,
      "learning_rate": 9.899965716883372e-05,
      "loss": 0.8314,
      "step": 293
    },
    {
      "epoch": 0.3804593982529926,
      "grad_norm": 0.8122355937957764,
      "learning_rate": 9.899143480195913e-05,
      "loss": 0.8369,
      "step": 294
    },
    {
      "epoch": 0.3817534778388871,
      "grad_norm": 0.7415926456451416,
      "learning_rate": 9.898317912544536e-05,
      "loss": 0.8985,
      "step": 295
    },
    {
      "epoch": 0.38304755742478164,
      "grad_norm": 0.8654418587684631,
      "learning_rate": 9.897489014490553e-05,
      "loss": 0.8752,
      "step": 296
    },
    {
      "epoch": 0.38434163701067614,
      "grad_norm": 0.7863161563873291,
      "learning_rate": 9.896656786597535e-05,
      "loss": 1.0138,
      "step": 297
    },
    {
      "epoch": 0.3856357165965707,
      "grad_norm": 0.8346667885780334,
      "learning_rate": 9.895821229431323e-05,
      "loss": 0.94,
      "step": 298
    },
    {
      "epoch": 0.3869297961824652,
      "grad_norm": 0.9577547907829285,
      "learning_rate": 9.894982343560016e-05,
      "loss": 0.9967,
      "step": 299
    },
    {
      "epoch": 0.38822387576835976,
      "grad_norm": 0.7633039951324463,
      "learning_rate": 9.894140129553981e-05,
      "loss": 0.8469,
      "step": 300
    },
    {
      "epoch": 0.38951795535425426,
      "grad_norm": 0.8275448083877563,
      "learning_rate": 9.893294587985843e-05,
      "loss": 0.8295,
      "step": 301
    },
    {
      "epoch": 0.3908120349401488,
      "grad_norm": 0.9197372198104858,
      "learning_rate": 9.892445719430493e-05,
      "loss": 0.9363,
      "step": 302
    },
    {
      "epoch": 0.3921061145260433,
      "grad_norm": 0.7137726545333862,
      "learning_rate": 9.891593524465083e-05,
      "loss": 0.8207,
      "step": 303
    },
    {
      "epoch": 0.3934001941119379,
      "grad_norm": 0.9908462762832642,
      "learning_rate": 9.890738003669029e-05,
      "loss": 0.9287,
      "step": 304
    },
    {
      "epoch": 0.39469427369783244,
      "grad_norm": 1.0070343017578125,
      "learning_rate": 9.889879157624002e-05,
      "loss": 1.2092,
      "step": 305
    },
    {
      "epoch": 0.39598835328372695,
      "grad_norm": 0.7812051177024841,
      "learning_rate": 9.889016986913941e-05,
      "loss": 0.8149,
      "step": 306
    },
    {
      "epoch": 0.3972824328696215,
      "grad_norm": 0.8907694816589355,
      "learning_rate": 9.888151492125039e-05,
      "loss": 0.9008,
      "step": 307
    },
    {
      "epoch": 0.398576512455516,
      "grad_norm": 0.960164487361908,
      "learning_rate": 9.887282673845754e-05,
      "loss": 0.9437,
      "step": 308
    },
    {
      "epoch": 0.39987059204141057,
      "grad_norm": 0.8578314185142517,
      "learning_rate": 9.886410532666805e-05,
      "loss": 0.8422,
      "step": 309
    },
    {
      "epoch": 0.40116467162730507,
      "grad_norm": 0.8671669960021973,
      "learning_rate": 9.885535069181162e-05,
      "loss": 0.924,
      "step": 310
    },
    {
      "epoch": 0.40245875121319963,
      "grad_norm": 1.037150263786316,
      "learning_rate": 9.884656283984062e-05,
      "loss": 1.0074,
      "step": 311
    },
    {
      "epoch": 0.40375283079909413,
      "grad_norm": 0.7935649156570435,
      "learning_rate": 9.883774177672998e-05,
      "loss": 0.9804,
      "step": 312
    },
    {
      "epoch": 0.4050469103849887,
      "grad_norm": 0.8249583840370178,
      "learning_rate": 9.882888750847717e-05,
      "loss": 0.9627,
      "step": 313
    },
    {
      "epoch": 0.4063409899708832,
      "grad_norm": 0.7363491058349609,
      "learning_rate": 9.882000004110233e-05,
      "loss": 0.8309,
      "step": 314
    },
    {
      "epoch": 0.40763506955677775,
      "grad_norm": 0.8029589653015137,
      "learning_rate": 9.881107938064806e-05,
      "loss": 0.8762,
      "step": 315
    },
    {
      "epoch": 0.40892914914267225,
      "grad_norm": 0.7131720781326294,
      "learning_rate": 9.880212553317963e-05,
      "loss": 0.882,
      "step": 316
    },
    {
      "epoch": 0.4102232287285668,
      "grad_norm": 0.7951234579086304,
      "learning_rate": 9.879313850478478e-05,
      "loss": 1.0095,
      "step": 317
    },
    {
      "epoch": 0.4115173083144613,
      "grad_norm": 0.7714753150939941,
      "learning_rate": 9.87841183015739e-05,
      "loss": 0.8688,
      "step": 318
    },
    {
      "epoch": 0.4128113879003559,
      "grad_norm": 0.7753300666809082,
      "learning_rate": 9.877506492967987e-05,
      "loss": 0.9669,
      "step": 319
    },
    {
      "epoch": 0.4141054674862504,
      "grad_norm": 0.9767395853996277,
      "learning_rate": 9.876597839525814e-05,
      "loss": 1.1169,
      "step": 320
    },
    {
      "epoch": 0.41539954707214494,
      "grad_norm": 0.7923420667648315,
      "learning_rate": 9.875685870448672e-05,
      "loss": 0.9942,
      "step": 321
    },
    {
      "epoch": 0.4166936266580395,
      "grad_norm": 0.7265552282333374,
      "learning_rate": 9.874770586356616e-05,
      "loss": 1.0377,
      "step": 322
    },
    {
      "epoch": 0.417987706243934,
      "grad_norm": 0.7586270570755005,
      "learning_rate": 9.873851987871954e-05,
      "loss": 0.9172,
      "step": 323
    },
    {
      "epoch": 0.41928178582982856,
      "grad_norm": 0.782192587852478,
      "learning_rate": 9.872930075619249e-05,
      "loss": 0.9219,
      "step": 324
    },
    {
      "epoch": 0.42057586541572306,
      "grad_norm": 0.8508116602897644,
      "learning_rate": 9.872004850225313e-05,
      "loss": 0.939,
      "step": 325
    },
    {
      "epoch": 0.4218699450016176,
      "grad_norm": 0.8639410138130188,
      "learning_rate": 9.871076312319218e-05,
      "loss": 1.0854,
      "step": 326
    },
    {
      "epoch": 0.4231640245875121,
      "grad_norm": 0.809862494468689,
      "learning_rate": 9.870144462532281e-05,
      "loss": 0.8283,
      "step": 327
    },
    {
      "epoch": 0.4244581041734067,
      "grad_norm": 0.7755741477012634,
      "learning_rate": 9.869209301498072e-05,
      "loss": 0.7645,
      "step": 328
    },
    {
      "epoch": 0.4257521837593012,
      "grad_norm": 0.8238282203674316,
      "learning_rate": 9.868270829852416e-05,
      "loss": 0.8858,
      "step": 329
    },
    {
      "epoch": 0.42704626334519574,
      "grad_norm": 0.810691773891449,
      "learning_rate": 9.867329048233387e-05,
      "loss": 0.7913,
      "step": 330
    },
    {
      "epoch": 0.42834034293109025,
      "grad_norm": 0.8913648128509521,
      "learning_rate": 9.866383957281309e-05,
      "loss": 0.7885,
      "step": 331
    },
    {
      "epoch": 0.4296344225169848,
      "grad_norm": 0.9160618782043457,
      "learning_rate": 9.865435557638757e-05,
      "loss": 0.7296,
      "step": 332
    },
    {
      "epoch": 0.4309285021028793,
      "grad_norm": 0.7498170137405396,
      "learning_rate": 9.864483849950553e-05,
      "loss": 0.9655,
      "step": 333
    },
    {
      "epoch": 0.43222258168877387,
      "grad_norm": 0.7315449714660645,
      "learning_rate": 9.863528834863773e-05,
      "loss": 0.6886,
      "step": 334
    },
    {
      "epoch": 0.43351666127466837,
      "grad_norm": 0.8278869390487671,
      "learning_rate": 9.862570513027735e-05,
      "loss": 0.9637,
      "step": 335
    },
    {
      "epoch": 0.4348107408605629,
      "grad_norm": 0.8372804522514343,
      "learning_rate": 9.861608885094012e-05,
      "loss": 0.8609,
      "step": 336
    },
    {
      "epoch": 0.43610482044645743,
      "grad_norm": 0.8712325096130371,
      "learning_rate": 9.860643951716421e-05,
      "loss": 0.9718,
      "step": 337
    },
    {
      "epoch": 0.437398900032352,
      "grad_norm": 0.9869045615196228,
      "learning_rate": 9.859675713551028e-05,
      "loss": 0.887,
      "step": 338
    },
    {
      "epoch": 0.43869297961824655,
      "grad_norm": 0.9166460037231445,
      "learning_rate": 9.858704171256145e-05,
      "loss": 1.0751,
      "step": 339
    },
    {
      "epoch": 0.43998705920414105,
      "grad_norm": 1.1965091228485107,
      "learning_rate": 9.857729325492329e-05,
      "loss": 1.0093,
      "step": 340
    },
    {
      "epoch": 0.4412811387900356,
      "grad_norm": 0.8646867275238037,
      "learning_rate": 9.856751176922388e-05,
      "loss": 0.9235,
      "step": 341
    },
    {
      "epoch": 0.4425752183759301,
      "grad_norm": 0.7576479315757751,
      "learning_rate": 9.85576972621137e-05,
      "loss": 0.8323,
      "step": 342
    },
    {
      "epoch": 0.44386929796182467,
      "grad_norm": 0.8257366418838501,
      "learning_rate": 9.854784974026572e-05,
      "loss": 0.8478,
      "step": 343
    },
    {
      "epoch": 0.4451633775477192,
      "grad_norm": 0.8963577747344971,
      "learning_rate": 9.853796921037534e-05,
      "loss": 0.8943,
      "step": 344
    },
    {
      "epoch": 0.44645745713361373,
      "grad_norm": 0.8696343898773193,
      "learning_rate": 9.85280556791604e-05,
      "loss": 0.991,
      "step": 345
    },
    {
      "epoch": 0.44775153671950824,
      "grad_norm": 0.8321841359138489,
      "learning_rate": 9.851810915336119e-05,
      "loss": 1.0462,
      "step": 346
    },
    {
      "epoch": 0.4490456163054028,
      "grad_norm": 0.8194176554679871,
      "learning_rate": 9.850812963974042e-05,
      "loss": 0.9484,
      "step": 347
    },
    {
      "epoch": 0.4503396958912973,
      "grad_norm": 0.9131197333335876,
      "learning_rate": 9.849811714508323e-05,
      "loss": 0.9687,
      "step": 348
    },
    {
      "epoch": 0.45163377547719186,
      "grad_norm": 0.891231894493103,
      "learning_rate": 9.848807167619721e-05,
      "loss": 0.9605,
      "step": 349
    },
    {
      "epoch": 0.45292785506308636,
      "grad_norm": 0.9574115872383118,
      "learning_rate": 9.847799323991234e-05,
      "loss": 1.077,
      "step": 350
    },
    {
      "epoch": 0.4542219346489809,
      "grad_norm": 0.9179444909095764,
      "learning_rate": 9.8467881843081e-05,
      "loss": 0.8353,
      "step": 351
    },
    {
      "epoch": 0.4555160142348754,
      "grad_norm": 0.855802595615387,
      "learning_rate": 9.845773749257804e-05,
      "loss": 0.9764,
      "step": 352
    },
    {
      "epoch": 0.45681009382077,
      "grad_norm": 0.9086332321166992,
      "learning_rate": 9.844756019530066e-05,
      "loss": 1.0526,
      "step": 353
    },
    {
      "epoch": 0.4581041734066645,
      "grad_norm": 0.890271782875061,
      "learning_rate": 9.843734995816848e-05,
      "loss": 0.8905,
      "step": 354
    },
    {
      "epoch": 0.45939825299255904,
      "grad_norm": 0.7878096699714661,
      "learning_rate": 9.842710678812351e-05,
      "loss": 0.8706,
      "step": 355
    },
    {
      "epoch": 0.4606923325784536,
      "grad_norm": 0.9886014461517334,
      "learning_rate": 9.841683069213017e-05,
      "loss": 0.9579,
      "step": 356
    },
    {
      "epoch": 0.4619864121643481,
      "grad_norm": 0.8265432119369507,
      "learning_rate": 9.840652167717526e-05,
      "loss": 0.8528,
      "step": 357
    },
    {
      "epoch": 0.46328049175024266,
      "grad_norm": 0.7354372143745422,
      "learning_rate": 9.839617975026793e-05,
      "loss": 0.775,
      "step": 358
    },
    {
      "epoch": 0.46457457133613717,
      "grad_norm": 0.8311409950256348,
      "learning_rate": 9.838580491843976e-05,
      "loss": 1.0374,
      "step": 359
    },
    {
      "epoch": 0.4658686509220317,
      "grad_norm": 0.8180521130561829,
      "learning_rate": 9.837539718874464e-05,
      "loss": 0.818,
      "step": 360
    },
    {
      "epoch": 0.4671627305079262,
      "grad_norm": 0.7231691479682922,
      "learning_rate": 9.83649565682589e-05,
      "loss": 0.9141,
      "step": 361
    },
    {
      "epoch": 0.4684568100938208,
      "grad_norm": 0.7644848227500916,
      "learning_rate": 9.835448306408118e-05,
      "loss": 0.9565,
      "step": 362
    },
    {
      "epoch": 0.4697508896797153,
      "grad_norm": 0.7103495001792908,
      "learning_rate": 9.83439766833325e-05,
      "loss": 0.7706,
      "step": 363
    },
    {
      "epoch": 0.47104496926560985,
      "grad_norm": 0.8206077218055725,
      "learning_rate": 9.83334374331562e-05,
      "loss": 1.0923,
      "step": 364
    },
    {
      "epoch": 0.47233904885150435,
      "grad_norm": 0.7409754991531372,
      "learning_rate": 9.832286532071802e-05,
      "loss": 0.8534,
      "step": 365
    },
    {
      "epoch": 0.4736331284373989,
      "grad_norm": 0.8880231380462646,
      "learning_rate": 9.831226035320602e-05,
      "loss": 0.9025,
      "step": 366
    },
    {
      "epoch": 0.4749272080232934,
      "grad_norm": 0.9805209636688232,
      "learning_rate": 9.830162253783058e-05,
      "loss": 0.9939,
      "step": 367
    },
    {
      "epoch": 0.47622128760918797,
      "grad_norm": 0.9529579281806946,
      "learning_rate": 9.829095188182442e-05,
      "loss": 0.9645,
      "step": 368
    },
    {
      "epoch": 0.4775153671950825,
      "grad_norm": 0.8122926354408264,
      "learning_rate": 9.828024839244263e-05,
      "loss": 0.9269,
      "step": 369
    },
    {
      "epoch": 0.47880944678097703,
      "grad_norm": 0.8557335734367371,
      "learning_rate": 9.826951207696258e-05,
      "loss": 0.8781,
      "step": 370
    },
    {
      "epoch": 0.48010352636687154,
      "grad_norm": 0.959077775478363,
      "learning_rate": 9.825874294268396e-05,
      "loss": 0.9448,
      "step": 371
    },
    {
      "epoch": 0.4813976059527661,
      "grad_norm": 0.7479959726333618,
      "learning_rate": 9.824794099692878e-05,
      "loss": 0.8608,
      "step": 372
    },
    {
      "epoch": 0.48269168553866065,
      "grad_norm": 0.8064723014831543,
      "learning_rate": 9.823710624704137e-05,
      "loss": 1.1222,
      "step": 373
    },
    {
      "epoch": 0.48398576512455516,
      "grad_norm": 0.8007084131240845,
      "learning_rate": 9.822623870038838e-05,
      "loss": 0.8967,
      "step": 374
    },
    {
      "epoch": 0.4852798447104497,
      "grad_norm": 0.988571286201477,
      "learning_rate": 9.82153383643587e-05,
      "loss": 0.9574,
      "step": 375
    },
    {
      "epoch": 0.4865739242963442,
      "grad_norm": 0.9317258596420288,
      "learning_rate": 9.820440524636356e-05,
      "loss": 0.929,
      "step": 376
    },
    {
      "epoch": 0.4878680038822388,
      "grad_norm": 0.8651334643363953,
      "learning_rate": 9.819343935383649e-05,
      "loss": 0.9483,
      "step": 377
    },
    {
      "epoch": 0.4891620834681333,
      "grad_norm": 0.791306734085083,
      "learning_rate": 9.818244069423325e-05,
      "loss": 0.9277,
      "step": 378
    },
    {
      "epoch": 0.49045616305402784,
      "grad_norm": 0.6706424355506897,
      "learning_rate": 9.817140927503192e-05,
      "loss": 0.9013,
      "step": 379
    },
    {
      "epoch": 0.49175024263992234,
      "grad_norm": 0.8574838042259216,
      "learning_rate": 9.816034510373286e-05,
      "loss": 0.9889,
      "step": 380
    },
    {
      "epoch": 0.4930443222258169,
      "grad_norm": 0.829309344291687,
      "learning_rate": 9.814924818785865e-05,
      "loss": 0.8139,
      "step": 381
    },
    {
      "epoch": 0.4943384018117114,
      "grad_norm": 0.83943110704422,
      "learning_rate": 9.81381185349542e-05,
      "loss": 0.9202,
      "step": 382
    },
    {
      "epoch": 0.49563248139760596,
      "grad_norm": 0.7981933355331421,
      "learning_rate": 9.812695615258662e-05,
      "loss": 0.9131,
      "step": 383
    },
    {
      "epoch": 0.49692656098350046,
      "grad_norm": 0.7930905818939209,
      "learning_rate": 9.81157610483453e-05,
      "loss": 0.769,
      "step": 384
    },
    {
      "epoch": 0.498220640569395,
      "grad_norm": 0.8699679970741272,
      "learning_rate": 9.81045332298419e-05,
      "loss": 0.9468,
      "step": 385
    },
    {
      "epoch": 0.4995147201552895,
      "grad_norm": 0.7733138799667358,
      "learning_rate": 9.809327270471025e-05,
      "loss": 0.8982,
      "step": 386
    },
    {
      "epoch": 0.4995147201552895,
      "eval_loss": 0.8872498869895935,
      "eval_runtime": 189.8035,
      "eval_samples_per_second": 3.43,
      "eval_steps_per_second": 0.859,
      "step": 386
    },
    {
      "epoch": 0.5008087997411841,
      "grad_norm": 0.9202268123626709,
      "learning_rate": 9.808197948060651e-05,
      "loss": 0.9664,
      "step": 387
    },
    {
      "epoch": 0.5021028793270786,
      "grad_norm": 0.7772640585899353,
      "learning_rate": 9.807065356520899e-05,
      "loss": 0.8825,
      "step": 388
    },
    {
      "epoch": 0.5033969589129732,
      "grad_norm": 0.8117741346359253,
      "learning_rate": 9.805929496621828e-05,
      "loss": 0.8166,
      "step": 389
    },
    {
      "epoch": 0.5046910384988677,
      "grad_norm": 0.7308422327041626,
      "learning_rate": 9.804790369135718e-05,
      "loss": 0.8657,
      "step": 390
    },
    {
      "epoch": 0.5059851180847622,
      "grad_norm": 0.8373358845710754,
      "learning_rate": 9.80364797483707e-05,
      "loss": 0.9175,
      "step": 391
    },
    {
      "epoch": 0.5072791976706568,
      "grad_norm": 0.8288902640342712,
      "learning_rate": 9.802502314502607e-05,
      "loss": 0.7463,
      "step": 392
    },
    {
      "epoch": 0.5085732772565513,
      "grad_norm": 0.6780114769935608,
      "learning_rate": 9.801353388911269e-05,
      "loss": 0.7973,
      "step": 393
    },
    {
      "epoch": 0.5098673568424458,
      "grad_norm": 0.9328367710113525,
      "learning_rate": 9.800201198844221e-05,
      "loss": 1.0405,
      "step": 394
    },
    {
      "epoch": 0.5111614364283403,
      "grad_norm": 0.9010327458381653,
      "learning_rate": 9.799045745084847e-05,
      "loss": 1.1194,
      "step": 395
    },
    {
      "epoch": 0.5124555160142349,
      "grad_norm": 0.7900609374046326,
      "learning_rate": 9.797887028418746e-05,
      "loss": 0.9747,
      "step": 396
    },
    {
      "epoch": 0.5137495956001294,
      "grad_norm": 0.8245125412940979,
      "learning_rate": 9.796725049633741e-05,
      "loss": 0.8594,
      "step": 397
    },
    {
      "epoch": 0.5150436751860239,
      "grad_norm": 0.7617928385734558,
      "learning_rate": 9.795559809519866e-05,
      "loss": 0.8176,
      "step": 398
    },
    {
      "epoch": 0.5163377547719185,
      "grad_norm": 0.9844444990158081,
      "learning_rate": 9.79439130886938e-05,
      "loss": 1.0116,
      "step": 399
    },
    {
      "epoch": 0.517631834357813,
      "grad_norm": 0.8849684000015259,
      "learning_rate": 9.793219548476753e-05,
      "loss": 0.8706,
      "step": 400
    },
    {
      "epoch": 0.5189259139437076,
      "grad_norm": 0.893902063369751,
      "learning_rate": 9.792044529138674e-05,
      "loss": 0.8217,
      "step": 401
    },
    {
      "epoch": 0.520219993529602,
      "grad_norm": 0.7412934899330139,
      "learning_rate": 9.79086625165405e-05,
      "loss": 0.868,
      "step": 402
    },
    {
      "epoch": 0.5215140731154966,
      "grad_norm": 0.796435534954071,
      "learning_rate": 9.789684716823995e-05,
      "loss": 0.8691,
      "step": 403
    },
    {
      "epoch": 0.5228081527013911,
      "grad_norm": 1.063193440437317,
      "learning_rate": 9.788499925451849e-05,
      "loss": 1.0085,
      "step": 404
    },
    {
      "epoch": 0.5241022322872857,
      "grad_norm": 0.952882707118988,
      "learning_rate": 9.787311878343157e-05,
      "loss": 0.8378,
      "step": 405
    },
    {
      "epoch": 0.5253963118731803,
      "grad_norm": 0.7899916768074036,
      "learning_rate": 9.786120576305682e-05,
      "loss": 0.8917,
      "step": 406
    },
    {
      "epoch": 0.5266903914590747,
      "grad_norm": 0.8242781758308411,
      "learning_rate": 9.784926020149398e-05,
      "loss": 0.9778,
      "step": 407
    },
    {
      "epoch": 0.5279844710449693,
      "grad_norm": 0.9736928343772888,
      "learning_rate": 9.783728210686496e-05,
      "loss": 1.0145,
      "step": 408
    },
    {
      "epoch": 0.5292785506308638,
      "grad_norm": 0.8070263862609863,
      "learning_rate": 9.782527148731372e-05,
      "loss": 0.9923,
      "step": 409
    },
    {
      "epoch": 0.5305726302167584,
      "grad_norm": 0.861262857913971,
      "learning_rate": 9.781322835100638e-05,
      "loss": 1.0261,
      "step": 410
    },
    {
      "epoch": 0.5318667098026528,
      "grad_norm": 0.8204033970832825,
      "learning_rate": 9.780115270613115e-05,
      "loss": 0.8865,
      "step": 411
    },
    {
      "epoch": 0.5331607893885474,
      "grad_norm": 0.8053429126739502,
      "learning_rate": 9.778904456089838e-05,
      "loss": 0.8606,
      "step": 412
    },
    {
      "epoch": 0.5344548689744419,
      "grad_norm": 0.7944838404655457,
      "learning_rate": 9.777690392354045e-05,
      "loss": 0.8925,
      "step": 413
    },
    {
      "epoch": 0.5357489485603365,
      "grad_norm": 0.9882616400718689,
      "learning_rate": 9.77647308023119e-05,
      "loss": 0.8577,
      "step": 414
    },
    {
      "epoch": 0.537043028146231,
      "grad_norm": 0.6937118768692017,
      "learning_rate": 9.77525252054893e-05,
      "loss": 0.6972,
      "step": 415
    },
    {
      "epoch": 0.5383371077321255,
      "grad_norm": 1.0071479082107544,
      "learning_rate": 9.774028714137133e-05,
      "loss": 0.9233,
      "step": 416
    },
    {
      "epoch": 0.5396311873180201,
      "grad_norm": 0.916771411895752,
      "learning_rate": 9.772801661827874e-05,
      "loss": 1.0072,
      "step": 417
    },
    {
      "epoch": 0.5409252669039146,
      "grad_norm": 0.8663278818130493,
      "learning_rate": 9.771571364455439e-05,
      "loss": 1.1011,
      "step": 418
    },
    {
      "epoch": 0.5422193464898091,
      "grad_norm": 0.7842355370521545,
      "learning_rate": 9.77033782285631e-05,
      "loss": 1.0382,
      "step": 419
    },
    {
      "epoch": 0.5435134260757036,
      "grad_norm": 0.8407487273216248,
      "learning_rate": 9.769101037869187e-05,
      "loss": 0.9612,
      "step": 420
    },
    {
      "epoch": 0.5448075056615982,
      "grad_norm": 0.7557950615882874,
      "learning_rate": 9.767861010334962e-05,
      "loss": 0.9965,
      "step": 421
    },
    {
      "epoch": 0.5461015852474927,
      "grad_norm": 0.812099814414978,
      "learning_rate": 9.766617741096746e-05,
      "loss": 0.8824,
      "step": 422
    },
    {
      "epoch": 0.5473956648333873,
      "grad_norm": 0.8438544273376465,
      "learning_rate": 9.765371230999843e-05,
      "loss": 0.8852,
      "step": 423
    },
    {
      "epoch": 0.5486897444192818,
      "grad_norm": 0.7006101608276367,
      "learning_rate": 9.764121480891765e-05,
      "loss": 0.808,
      "step": 424
    },
    {
      "epoch": 0.5499838240051763,
      "grad_norm": 0.7312132716178894,
      "learning_rate": 9.76286849162223e-05,
      "loss": 0.7964,
      "step": 425
    },
    {
      "epoch": 0.5512779035910709,
      "grad_norm": 0.939939558506012,
      "learning_rate": 9.76161226404315e-05,
      "loss": 0.9434,
      "step": 426
    },
    {
      "epoch": 0.5525719831769654,
      "grad_norm": 0.7815152406692505,
      "learning_rate": 9.760352799008643e-05,
      "loss": 1.0487,
      "step": 427
    },
    {
      "epoch": 0.5538660627628599,
      "grad_norm": 0.8934217691421509,
      "learning_rate": 9.759090097375032e-05,
      "loss": 1.0036,
      "step": 428
    },
    {
      "epoch": 0.5551601423487544,
      "grad_norm": 0.7264319658279419,
      "learning_rate": 9.757824160000837e-05,
      "loss": 0.7866,
      "step": 429
    },
    {
      "epoch": 0.556454221934649,
      "grad_norm": 0.8536062836647034,
      "learning_rate": 9.756554987746776e-05,
      "loss": 0.9949,
      "step": 430
    },
    {
      "epoch": 0.5577483015205436,
      "grad_norm": 0.8253523111343384,
      "learning_rate": 9.755282581475769e-05,
      "loss": 0.9801,
      "step": 431
    },
    {
      "epoch": 0.559042381106438,
      "grad_norm": 0.6686350703239441,
      "learning_rate": 9.754006942052936e-05,
      "loss": 0.8388,
      "step": 432
    },
    {
      "epoch": 0.5603364606923326,
      "grad_norm": 0.8560882210731506,
      "learning_rate": 9.752728070345591e-05,
      "loss": 1.0167,
      "step": 433
    },
    {
      "epoch": 0.5616305402782271,
      "grad_norm": 0.765852153301239,
      "learning_rate": 9.751445967223252e-05,
      "loss": 0.7865,
      "step": 434
    },
    {
      "epoch": 0.5629246198641217,
      "grad_norm": 0.8869863152503967,
      "learning_rate": 9.750160633557627e-05,
      "loss": 0.9424,
      "step": 435
    },
    {
      "epoch": 0.5642186994500161,
      "grad_norm": 0.8522469401359558,
      "learning_rate": 9.748872070222625e-05,
      "loss": 0.8151,
      "step": 436
    },
    {
      "epoch": 0.5655127790359107,
      "grad_norm": 0.862091064453125,
      "learning_rate": 9.747580278094352e-05,
      "loss": 1.0131,
      "step": 437
    },
    {
      "epoch": 0.5668068586218052,
      "grad_norm": 0.8809049725532532,
      "learning_rate": 9.746285258051104e-05,
      "loss": 0.9092,
      "step": 438
    },
    {
      "epoch": 0.5681009382076998,
      "grad_norm": 0.8506961464881897,
      "learning_rate": 9.744987010973377e-05,
      "loss": 0.8756,
      "step": 439
    },
    {
      "epoch": 0.5693950177935944,
      "grad_norm": 0.7440844774246216,
      "learning_rate": 9.743685537743856e-05,
      "loss": 0.7948,
      "step": 440
    },
    {
      "epoch": 0.5706890973794888,
      "grad_norm": 0.8117014169692993,
      "learning_rate": 9.742380839247425e-05,
      "loss": 0.9382,
      "step": 441
    },
    {
      "epoch": 0.5719831769653834,
      "grad_norm": 0.8515580296516418,
      "learning_rate": 9.741072916371157e-05,
      "loss": 0.9202,
      "step": 442
    },
    {
      "epoch": 0.5732772565512779,
      "grad_norm": 0.8774899244308472,
      "learning_rate": 9.739761770004318e-05,
      "loss": 0.9129,
      "step": 443
    },
    {
      "epoch": 0.5745713361371725,
      "grad_norm": 0.7189667820930481,
      "learning_rate": 9.738447401038367e-05,
      "loss": 0.7392,
      "step": 444
    },
    {
      "epoch": 0.5758654157230669,
      "grad_norm": 0.7225956916809082,
      "learning_rate": 9.737129810366952e-05,
      "loss": 0.8703,
      "step": 445
    },
    {
      "epoch": 0.5771594953089615,
      "grad_norm": 0.8351484537124634,
      "learning_rate": 9.735808998885915e-05,
      "loss": 0.9626,
      "step": 446
    },
    {
      "epoch": 0.578453574894856,
      "grad_norm": 0.860819399356842,
      "learning_rate": 9.734484967493282e-05,
      "loss": 0.9061,
      "step": 447
    },
    {
      "epoch": 0.5797476544807506,
      "grad_norm": 0.7928848266601562,
      "learning_rate": 9.733157717089277e-05,
      "loss": 0.9141,
      "step": 448
    },
    {
      "epoch": 0.581041734066645,
      "grad_norm": 0.9132540822029114,
      "learning_rate": 9.7318272485763e-05,
      "loss": 0.7613,
      "step": 449
    },
    {
      "epoch": 0.5823358136525396,
      "grad_norm": 0.7894258499145508,
      "learning_rate": 9.730493562858953e-05,
      "loss": 0.8234,
      "step": 450
    }
  ],
  "logging_steps": 1,
  "max_steps": 3860,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 5,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 5.030717272621056e+17,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}