{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 4.0,
  "eval_steps": 500,
  "global_step": 468,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.008547008547008548,
      "grad_norm": 0.8597566485404968,
      "learning_rate": 2e-05,
      "loss": 1.0565,
      "step": 1
    },
    {
      "epoch": 0.017094017094017096,
      "grad_norm": 0.6099331378936768,
      "learning_rate": 4e-05,
      "loss": 0.8421,
      "step": 2
    },
    {
      "epoch": 0.02564102564102564,
      "grad_norm": 0.976015031337738,
      "learning_rate": 6e-05,
      "loss": 1.1244,
      "step": 3
    },
    {
      "epoch": 0.03418803418803419,
      "grad_norm": 1.2120051383972168,
      "learning_rate": 8e-05,
      "loss": 1.4274,
      "step": 4
    },
    {
      "epoch": 0.042735042735042736,
      "grad_norm": 0.7949010133743286,
      "learning_rate": 0.0001,
      "loss": 0.9262,
      "step": 5
    },
    {
      "epoch": 0.05128205128205128,
      "grad_norm": 1.1617588996887207,
      "learning_rate": 0.00012,
      "loss": 1.3036,
      "step": 6
    },
    {
      "epoch": 0.05982905982905983,
      "grad_norm": 0.7767685055732727,
      "learning_rate": 0.00014,
      "loss": 0.8809,
      "step": 7
    },
    {
      "epoch": 0.06837606837606838,
      "grad_norm": 0.4965355098247528,
      "learning_rate": 0.00016,
      "loss": 0.8728,
      "step": 8
    },
    {
      "epoch": 0.07692307692307693,
      "grad_norm": 0.4307010769844055,
      "learning_rate": 0.00018,
      "loss": 0.7268,
      "step": 9
    },
    {
      "epoch": 0.08547008547008547,
      "grad_norm": 0.5121279358863831,
      "learning_rate": 0.0002,
      "loss": 0.9179,
      "step": 10
    },
    {
      "epoch": 0.09401709401709402,
      "grad_norm": 0.5783049464225769,
      "learning_rate": 0.0001999976474595967,
      "loss": 0.8706,
      "step": 11
    },
    {
      "epoch": 0.10256410256410256,
      "grad_norm": 0.407270610332489,
      "learning_rate": 0.00019999058994907564,
      "loss": 0.706,
      "step": 12
    },
    {
      "epoch": 0.1111111111111111,
      "grad_norm": 0.4507271647453308,
      "learning_rate": 0.00019997882780049847,
      "loss": 0.8007,
      "step": 13
    },
    {
      "epoch": 0.11965811965811966,
      "grad_norm": 0.4835984706878662,
      "learning_rate": 0.0001999623615672837,
      "loss": 0.8741,
      "step": 14
    },
    {
      "epoch": 0.1282051282051282,
      "grad_norm": 0.35001033544540405,
      "learning_rate": 0.00019994119202418098,
      "loss": 1.0107,
      "step": 15
    },
    {
      "epoch": 0.13675213675213677,
      "grad_norm": 0.34652167558670044,
      "learning_rate": 0.00019991532016723439,
      "loss": 0.7135,
      "step": 16
    },
    {
      "epoch": 0.1452991452991453,
      "grad_norm": 0.3886638581752777,
      "learning_rate": 0.00019988474721373568,
      "loss": 1.0932,
      "step": 17
    },
    {
      "epoch": 0.15384615384615385,
      "grad_norm": 0.44986358284950256,
      "learning_rate": 0.00019984947460216707,
      "loss": 0.5826,
      "step": 18
    },
    {
      "epoch": 0.1623931623931624,
      "grad_norm": 0.38238662481307983,
      "learning_rate": 0.00019980950399213344,
      "loss": 0.7045,
      "step": 19
    },
    {
      "epoch": 0.17094017094017094,
      "grad_norm": 0.2608371675014496,
      "learning_rate": 0.00019976483726428422,
      "loss": 0.5967,
      "step": 20
    },
    {
      "epoch": 0.1794871794871795,
      "grad_norm": 0.3465232253074646,
      "learning_rate": 0.0001997154765202251,
      "loss": 0.6187,
      "step": 21
    },
    {
      "epoch": 0.18803418803418803,
      "grad_norm": 0.32362625002861023,
      "learning_rate": 0.00019966142408241901,
      "loss": 0.6944,
      "step": 22
    },
    {
      "epoch": 0.19658119658119658,
      "grad_norm": 0.3342239260673523,
      "learning_rate": 0.00019960268249407675,
      "loss": 0.7556,
      "step": 23
    },
    {
      "epoch": 0.20512820512820512,
      "grad_norm": 0.3412131667137146,
      "learning_rate": 0.00019953925451903756,
      "loss": 0.7405,
      "step": 24
    },
    {
      "epoch": 0.21367521367521367,
      "grad_norm": 0.44964078068733215,
      "learning_rate": 0.0001994711431416389,
      "loss": 1.3261,
      "step": 25
    },
    {
      "epoch": 0.2222222222222222,
      "grad_norm": 0.3409959077835083,
      "learning_rate": 0.00019939835156657616,
      "loss": 1.1328,
      "step": 26
    },
    {
      "epoch": 0.23076923076923078,
      "grad_norm": 0.3522396981716156,
      "learning_rate": 0.00019932088321875172,
      "loss": 0.6525,
      "step": 27
    },
    {
      "epoch": 0.23931623931623933,
      "grad_norm": 0.32580652832984924,
      "learning_rate": 0.00019923874174311394,
      "loss": 0.5873,
      "step": 28
    },
    {
      "epoch": 0.24786324786324787,
      "grad_norm": 0.28162461519241333,
      "learning_rate": 0.0001991519310044857,
      "loss": 0.9859,
      "step": 29
    },
    {
      "epoch": 0.2564102564102564,
      "grad_norm": 0.3276708424091339,
      "learning_rate": 0.00019906045508738228,
      "loss": 0.8892,
      "step": 30
    },
    {
      "epoch": 0.26495726495726496,
      "grad_norm": 0.3066442906856537,
      "learning_rate": 0.0001989643182958196,
      "loss": 0.612,
      "step": 31
    },
    {
      "epoch": 0.27350427350427353,
      "grad_norm": 0.3192179203033447,
      "learning_rate": 0.00019886352515311134,
      "loss": 0.6572,
      "step": 32
    },
    {
      "epoch": 0.28205128205128205,
      "grad_norm": 0.3197144567966461,
      "learning_rate": 0.0001987580804016563,
      "loss": 0.8862,
      "step": 33
    },
    {
      "epoch": 0.2905982905982906,
      "grad_norm": 0.3386269807815552,
      "learning_rate": 0.00019864798900271532,
      "loss": 0.7647,
      "step": 34
    },
    {
      "epoch": 0.29914529914529914,
      "grad_norm": 0.2944129705429077,
      "learning_rate": 0.0001985332561361776,
      "loss": 0.6034,
      "step": 35
    },
    {
      "epoch": 0.3076923076923077,
      "grad_norm": 0.2859589457511902,
      "learning_rate": 0.00019841388720031727,
      "loss": 0.5716,
      "step": 36
    },
    {
      "epoch": 0.3162393162393162,
      "grad_norm": 0.33579686284065247,
      "learning_rate": 0.00019828988781153917,
      "loss": 0.6174,
      "step": 37
    },
    {
      "epoch": 0.3247863247863248,
      "grad_norm": 0.3010195195674896,
      "learning_rate": 0.00019816126380411476,
      "loss": 0.6273,
      "step": 38
    },
    {
      "epoch": 0.3333333333333333,
      "grad_norm": 0.28635373711586,
      "learning_rate": 0.00019802802122990758,
      "loss": 0.6698,
      "step": 39
    },
    {
      "epoch": 0.3418803418803419,
      "grad_norm": 0.32364264130592346,
      "learning_rate": 0.00019789016635808837,
      "loss": 0.7271,
      "step": 40
    },
    {
      "epoch": 0.3504273504273504,
      "grad_norm": 0.27588221430778503,
      "learning_rate": 0.00019774770567484022,
      "loss": 0.8396,
      "step": 41
    },
    {
      "epoch": 0.358974358974359,
      "grad_norm": 0.26630881428718567,
      "learning_rate": 0.00019760064588305345,
      "loss": 0.6031,
      "step": 42
    },
    {
      "epoch": 0.36752136752136755,
      "grad_norm": 0.2740603983402252,
      "learning_rate": 0.00019744899390201006,
      "loss": 0.6341,
      "step": 43
    },
    {
      "epoch": 0.37606837606837606,
      "grad_norm": 0.2932704985141754,
      "learning_rate": 0.0001972927568670583,
      "loss": 0.9879,
      "step": 44
    },
    {
      "epoch": 0.38461538461538464,
      "grad_norm": 0.3053448796272278,
      "learning_rate": 0.00019713194212927696,
      "loss": 0.7197,
      "step": 45
    },
    {
      "epoch": 0.39316239316239315,
      "grad_norm": 0.30776503682136536,
      "learning_rate": 0.00019696655725512933,
      "loss": 0.6701,
      "step": 46
    },
    {
      "epoch": 0.4017094017094017,
      "grad_norm": 0.31533047556877136,
      "learning_rate": 0.00019679661002610743,
      "loss": 0.7483,
      "step": 47
    },
    {
      "epoch": 0.41025641025641024,
      "grad_norm": 0.2600354254245758,
      "learning_rate": 0.00019662210843836574,
      "loss": 0.6361,
      "step": 48
    },
    {
      "epoch": 0.4188034188034188,
      "grad_norm": 0.29985320568084717,
      "learning_rate": 0.0001964430607023449,
      "loss": 0.6284,
      "step": 49
    },
    {
      "epoch": 0.42735042735042733,
      "grad_norm": 0.3829474449157715,
      "learning_rate": 0.00019625947524238563,
      "loss": 0.8927,
      "step": 50
    },
    {
      "epoch": 0.4358974358974359,
      "grad_norm": 0.29532748460769653,
      "learning_rate": 0.00019607136069633212,
      "loss": 0.5987,
      "step": 51
    },
    {
      "epoch": 0.4444444444444444,
      "grad_norm": 0.2676311731338501,
      "learning_rate": 0.0001958787259151258,
      "loss": 0.6036,
      "step": 52
    },
    {
      "epoch": 0.452991452991453,
      "grad_norm": 0.34333693981170654,
      "learning_rate": 0.00019568157996238884,
      "loss": 0.5933,
      "step": 53
    },
    {
      "epoch": 0.46153846153846156,
      "grad_norm": 0.2540273368358612,
      "learning_rate": 0.0001954799321139975,
      "loss": 0.8985,
      "step": 54
    },
    {
      "epoch": 0.4700854700854701,
      "grad_norm": 0.24644608795642853,
      "learning_rate": 0.00019527379185764612,
      "loss": 0.5539,
      "step": 55
    },
    {
      "epoch": 0.47863247863247865,
      "grad_norm": 0.37965652346611023,
      "learning_rate": 0.00019506316889240027,
      "loss": 1.0166,
      "step": 56
    },
    {
      "epoch": 0.48717948717948717,
      "grad_norm": 0.2802373766899109,
      "learning_rate": 0.00019484807312824067,
      "loss": 0.5285,
      "step": 57
    },
    {
      "epoch": 0.49572649572649574,
      "grad_norm": 0.3936748802661896,
      "learning_rate": 0.0001946285146855968,
      "loss": 0.6576,
      "step": 58
    },
    {
      "epoch": 0.5042735042735043,
      "grad_norm": 0.29837948083877563,
      "learning_rate": 0.0001944045038948709,
      "loss": 0.6493,
      "step": 59
    },
    {
      "epoch": 0.5128205128205128,
      "grad_norm": 0.2878054082393646,
      "learning_rate": 0.00019417605129595157,
      "loss": 0.6055,
      "step": 60
    },
    {
      "epoch": 0.5213675213675214,
      "grad_norm": 0.31146562099456787,
      "learning_rate": 0.0001939431676377183,
      "loss": 0.6076,
      "step": 61
    },
    {
      "epoch": 0.5299145299145299,
      "grad_norm": 0.26937177777290344,
      "learning_rate": 0.0001937058638775353,
      "loss": 0.5893,
      "step": 62
    },
    {
      "epoch": 0.5384615384615384,
      "grad_norm": 0.27157649397850037,
      "learning_rate": 0.00019346415118073632,
      "loss": 1.1555,
      "step": 63
    },
    {
      "epoch": 0.5470085470085471,
      "grad_norm": 0.4730302691459656,
      "learning_rate": 0.00019321804092009906,
      "loss": 0.6345,
      "step": 64
    },
    {
      "epoch": 0.5555555555555556,
      "grad_norm": 0.2813616693019867,
      "learning_rate": 0.00019296754467531014,
      "loss": 0.5973,
      "step": 65
    },
    {
      "epoch": 0.5641025641025641,
      "grad_norm": 0.3468441069126129,
      "learning_rate": 0.00019271267423242024,
      "loss": 0.6335,
      "step": 66
    },
    {
      "epoch": 0.5726495726495726,
      "grad_norm": 0.3037238121032715,
      "learning_rate": 0.00019245344158328972,
      "loss": 0.677,
      "step": 67
    },
    {
      "epoch": 0.5811965811965812,
      "grad_norm": 0.2790398597717285,
      "learning_rate": 0.0001921898589250242,
      "loss": 0.5703,
      "step": 68
    },
    {
      "epoch": 0.5897435897435898,
      "grad_norm": 0.288515567779541,
      "learning_rate": 0.0001919219386594007,
      "loss": 0.6514,
      "step": 69
    },
    {
      "epoch": 0.5982905982905983,
      "grad_norm": 0.2876873016357422,
      "learning_rate": 0.00019164969339228422,
      "loss": 0.6443,
      "step": 70
    },
    {
      "epoch": 0.6068376068376068,
      "grad_norm": 0.32371172308921814,
      "learning_rate": 0.00019137313593303463,
      "loss": 0.7022,
      "step": 71
    },
    {
      "epoch": 0.6153846153846154,
      "grad_norm": 0.2753008008003235,
      "learning_rate": 0.00019109227929390378,
      "loss": 1.2197,
      "step": 72
    },
    {
      "epoch": 0.6239316239316239,
      "grad_norm": 0.26767146587371826,
      "learning_rate": 0.00019080713668942356,
      "loss": 0.5784,
      "step": 73
    },
    {
      "epoch": 0.6324786324786325,
      "grad_norm": 0.2762153148651123,
      "learning_rate": 0.00019051772153578389,
      "loss": 0.5991,
      "step": 74
    },
    {
      "epoch": 0.6410256410256411,
      "grad_norm": 0.28386157751083374,
      "learning_rate": 0.00019022404745020163,
      "loss": 0.6117,
      "step": 75
    },
    {
      "epoch": 0.6495726495726496,
      "grad_norm": 0.24532322585582733,
      "learning_rate": 0.00018992612825027976,
      "loss": 0.6231,
      "step": 76
    },
    {
      "epoch": 0.6581196581196581,
      "grad_norm": 0.32852908968925476,
      "learning_rate": 0.0001896239779533575,
      "loss": 0.7321,
      "step": 77
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 0.24348995089530945,
      "learning_rate": 0.00018931761077585035,
      "loss": 0.589,
      "step": 78
    },
    {
      "epoch": 0.6752136752136753,
      "grad_norm": 0.25331753492355347,
      "learning_rate": 0.00018900704113258165,
      "loss": 0.5817,
      "step": 79
    },
    {
      "epoch": 0.6837606837606838,
      "grad_norm": 0.2514454126358032,
      "learning_rate": 0.00018869228363610404,
      "loss": 0.5693,
      "step": 80
    },
    {
      "epoch": 0.6923076923076923,
      "grad_norm": 0.28637388348579407,
      "learning_rate": 0.00018837335309601213,
      "loss": 0.6363,
      "step": 81
    },
    {
      "epoch": 0.7008547008547008,
      "grad_norm": 0.2514885663986206,
      "learning_rate": 0.00018805026451824546,
      "loss": 1.1033,
      "step": 82
    },
    {
      "epoch": 0.7094017094017094,
      "grad_norm": 0.2492385059595108,
      "learning_rate": 0.00018772303310438275,
      "loss": 1.045,
      "step": 83
    },
    {
      "epoch": 0.717948717948718,
      "grad_norm": 0.287953644990921,
      "learning_rate": 0.00018739167425092644,
      "loss": 1.0097,
      "step": 84
    },
    {
      "epoch": 0.7264957264957265,
      "grad_norm": 0.2525361478328705,
      "learning_rate": 0.00018705620354857833,
      "loss": 0.5465,
      "step": 85
    },
    {
      "epoch": 0.7350427350427351,
      "grad_norm": 0.2608508765697479,
      "learning_rate": 0.00018671663678150607,
      "loss": 0.5291,
      "step": 86
    },
    {
      "epoch": 0.7435897435897436,
      "grad_norm": 0.25817644596099854,
      "learning_rate": 0.0001863729899266004,
      "loss": 0.5783,
      "step": 87
    },
    {
      "epoch": 0.7521367521367521,
      "grad_norm": 0.32855451107025146,
      "learning_rate": 0.0001860252791527236,
      "loss": 0.926,
      "step": 88
    },
    {
      "epoch": 0.7606837606837606,
      "grad_norm": 0.25011545419692993,
      "learning_rate": 0.00018567352081994852,
      "loss": 1.12,
      "step": 89
    },
    {
      "epoch": 0.7692307692307693,
      "grad_norm": 0.24932663142681122,
      "learning_rate": 0.00018531773147878895,
      "loss": 0.9743,
      "step": 90
    },
    {
      "epoch": 0.7777777777777778,
      "grad_norm": 0.2763610780239105,
      "learning_rate": 0.0001849579278694209,
      "loss": 0.6299,
      "step": 91
    },
    {
      "epoch": 0.7863247863247863,
      "grad_norm": 0.27096471190452576,
      "learning_rate": 0.00018459412692089494,
      "loss": 0.5782,
      "step": 92
    },
    {
      "epoch": 0.7948717948717948,
      "grad_norm": 0.246726855635643,
      "learning_rate": 0.0001842263457503397,
      "loss": 0.5432,
      "step": 93
    },
    {
      "epoch": 0.8034188034188035,
      "grad_norm": 0.3436436951160431,
      "learning_rate": 0.00018385460166215638,
      "loss": 0.7304,
      "step": 94
    },
    {
      "epoch": 0.811965811965812,
      "grad_norm": 0.31473803520202637,
      "learning_rate": 0.00018347891214720477,
      "loss": 0.5874,
      "step": 95
    },
    {
      "epoch": 0.8205128205128205,
      "grad_norm": 0.23026835918426514,
      "learning_rate": 0.00018309929488198012,
      "loss": 1.0413,
      "step": 96
    },
    {
      "epoch": 0.8290598290598291,
      "grad_norm": 0.2863934338092804,
      "learning_rate": 0.00018271576772778154,
      "loss": 0.5917,
      "step": 97
    },
    {
      "epoch": 0.8376068376068376,
      "grad_norm": 0.2771088182926178,
      "learning_rate": 0.00018232834872987147,
      "loss": 1.089,
      "step": 98
    },
    {
      "epoch": 0.8461538461538461,
      "grad_norm": 0.33394113183021545,
      "learning_rate": 0.00018193705611662696,
      "loss": 0.6941,
      "step": 99
    },
    {
      "epoch": 0.8547008547008547,
      "grad_norm": 0.3389004170894623,
      "learning_rate": 0.0001815419082986815,
      "loss": 0.6486,
      "step": 100
    },
    {
      "epoch": 0.8632478632478633,
      "grad_norm": 0.2511347830295563,
      "learning_rate": 0.00018114292386805936,
      "loss": 0.952,
      "step": 101
    },
    {
      "epoch": 0.8717948717948718,
      "grad_norm": 0.28700685501098633,
      "learning_rate": 0.00018074012159730032,
      "loss": 0.9952,
      "step": 102
    },
    {
      "epoch": 0.8803418803418803,
      "grad_norm": 0.2681572437286377,
      "learning_rate": 0.00018033352043857675,
      "loss": 0.8697,
      "step": 103
    },
    {
      "epoch": 0.8888888888888888,
      "grad_norm": 0.28524017333984375,
      "learning_rate": 0.00017992313952280172,
      "loss": 0.9502,
      "step": 104
    },
    {
      "epoch": 0.8974358974358975,
      "grad_norm": 0.28100305795669556,
      "learning_rate": 0.00017950899815872892,
      "loss": 1.0808,
      "step": 105
    },
    {
      "epoch": 0.905982905982906,
      "grad_norm": 0.37970542907714844,
      "learning_rate": 0.00017909111583204422,
      "loss": 0.9715,
      "step": 106
    },
    {
      "epoch": 0.9145299145299145,
      "grad_norm": 0.4127101004123688,
      "learning_rate": 0.0001786695122044487,
      "loss": 0.7508,
      "step": 107
    },
    {
      "epoch": 0.9230769230769231,
      "grad_norm": 0.3084346354007721,
      "learning_rate": 0.0001782442071127338,
      "loss": 0.9778,
      "step": 108
    },
    {
      "epoch": 0.9316239316239316,
      "grad_norm": 0.4070857763290405,
      "learning_rate": 0.0001778152205678477,
      "loss": 1.0079,
      "step": 109
    },
    {
      "epoch": 0.9401709401709402,
      "grad_norm": 0.35972508788108826,
      "learning_rate": 0.00017738257275395404,
      "loss": 0.6978,
      "step": 110
    },
    {
      "epoch": 0.9487179487179487,
      "grad_norm": 0.3087528347969055,
      "learning_rate": 0.00017694628402748202,
      "loss": 0.6223,
      "step": 111
    },
    {
      "epoch": 0.9572649572649573,
      "grad_norm": 0.30665698647499084,
      "learning_rate": 0.0001765063749161688,
      "loss": 0.6312,
      "step": 112
    },
    {
      "epoch": 0.9658119658119658,
      "grad_norm": 0.3046605587005615,
      "learning_rate": 0.00017606286611809353,
      "loss": 0.6267,
      "step": 113
    },
    {
      "epoch": 0.9743589743589743,
      "grad_norm": 0.30426326394081116,
      "learning_rate": 0.00017561577850070355,
      "loss": 0.7445,
      "step": 114
    },
    {
      "epoch": 0.9829059829059829,
      "grad_norm": 0.23966537415981293,
      "learning_rate": 0.00017516513309983253,
      "loss": 0.526,
      "step": 115
    },
    {
      "epoch": 0.9914529914529915,
      "grad_norm": 0.3334603011608124,
      "learning_rate": 0.00017471095111871074,
      "loss": 0.6664,
      "step": 116
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.2775772511959076,
      "learning_rate": 0.0001742532539269674,
      "loss": 0.5718,
      "step": 117
    },
    {
      "epoch": 1.0085470085470085,
      "grad_norm": 0.2530048191547394,
      "learning_rate": 0.00017379206305962526,
      "loss": 0.4612,
      "step": 118
    },
    {
      "epoch": 1.017094017094017,
      "grad_norm": 0.2800690233707428,
      "learning_rate": 0.00017332740021608722,
      "loss": 0.4748,
      "step": 119
    },
    {
      "epoch": 1.0256410256410255,
      "grad_norm": 0.31294316053390503,
      "learning_rate": 0.00017285928725911562,
      "loss": 0.4615,
      "step": 120
    },
    {
      "epoch": 1.0341880341880343,
      "grad_norm": 0.3340826630592346,
      "learning_rate": 0.00017238774621380337,
      "loss": 0.464,
      "step": 121
    },
    {
      "epoch": 1.0427350427350428,
      "grad_norm": 0.35129979252815247,
      "learning_rate": 0.00017191279926653761,
      "loss": 1.0206,
      "step": 122
    },
    {
      "epoch": 1.0512820512820513,
      "grad_norm": 0.37155213952064514,
      "learning_rate": 0.00017143446876395602,
      "loss": 0.8751,
      "step": 123
    },
    {
      "epoch": 1.0598290598290598,
      "grad_norm": 0.35809117555618286,
      "learning_rate": 0.00017095277721189528,
      "loss": 0.8189,
      "step": 124
    },
    {
      "epoch": 1.0683760683760684,
      "grad_norm": 0.42735549807548523,
      "learning_rate": 0.00017046774727433222,
      "loss": 0.6002,
      "step": 125
    },
    {
      "epoch": 1.0769230769230769,
      "grad_norm": 0.3488568067550659,
      "learning_rate": 0.00016997940177231722,
      "loss": 0.465,
      "step": 126
    },
    {
      "epoch": 1.0854700854700854,
      "grad_norm": 0.3281893730163574,
      "learning_rate": 0.00016948776368290084,
      "loss": 0.9623,
      "step": 127
    },
    {
      "epoch": 1.0940170940170941,
      "grad_norm": 0.3557395935058594,
      "learning_rate": 0.00016899285613805246,
      "loss": 0.4101,
      "step": 128
    },
    {
      "epoch": 1.1025641025641026,
      "grad_norm": 0.4082263112068176,
      "learning_rate": 0.00016849470242357196,
      "loss": 0.4747,
      "step": 129
    },
    {
      "epoch": 1.1111111111111112,
      "grad_norm": 0.3952445089817047,
      "learning_rate": 0.00016799332597799413,
      "loss": 0.4398,
      "step": 130
    },
    {
      "epoch": 1.1196581196581197,
      "grad_norm": 0.43554118275642395,
      "learning_rate": 0.00016748875039148593,
      "loss": 0.6727,
      "step": 131
    },
    {
      "epoch": 1.1282051282051282,
      "grad_norm": 0.43142980337142944,
      "learning_rate": 0.0001669809994047364,
      "loss": 0.7956,
      "step": 132
    },
    {
      "epoch": 1.1367521367521367,
      "grad_norm": 0.39991500973701477,
      "learning_rate": 0.0001664700969078398,
      "loss": 0.4824,
      "step": 133
    },
    {
      "epoch": 1.1452991452991452,
      "grad_norm": 0.3793850243091583,
      "learning_rate": 0.00016595606693917142,
      "loss": 0.7563,
      "step": 134
    },
    {
      "epoch": 1.1538461538461537,
      "grad_norm": 0.3451911211013794,
      "learning_rate": 0.00016543893368425666,
      "loss": 0.4539,
      "step": 135
    },
    {
      "epoch": 1.1623931623931625,
      "grad_norm": 0.3969724178314209,
      "learning_rate": 0.00016491872147463306,
      "loss": 0.4022,
      "step": 136
    },
    {
      "epoch": 1.170940170940171,
      "grad_norm": 0.7453020811080933,
      "learning_rate": 0.00016439545478670543,
      "loss": 0.9909,
      "step": 137
    },
    {
      "epoch": 1.1794871794871795,
      "grad_norm": 0.41800379753112793,
      "learning_rate": 0.00016386915824059427,
      "loss": 0.4215,
      "step": 138
    },
    {
      "epoch": 1.188034188034188,
      "grad_norm": 0.3774307072162628,
      "learning_rate": 0.00016333985659897735,
      "loss": 0.3704,
      "step": 139
    },
    {
      "epoch": 1.1965811965811965,
      "grad_norm": 0.6166864037513733,
      "learning_rate": 0.00016280757476592466,
      "loss": 0.5109,
      "step": 140
    },
    {
      "epoch": 1.205128205128205,
      "grad_norm": 0.4391806423664093,
      "learning_rate": 0.0001622723377857265,
      "loss": 0.4229,
      "step": 141
    },
    {
      "epoch": 1.2136752136752136,
      "grad_norm": 1.190855860710144,
      "learning_rate": 0.00016173417084171536,
      "loss": 0.9702,
      "step": 142
    },
    {
      "epoch": 1.2222222222222223,
      "grad_norm": 0.4401251971721649,
      "learning_rate": 0.00016119309925508078,
      "loss": 0.4004,
      "step": 143
    },
    {
      "epoch": 1.2307692307692308,
      "grad_norm": 0.47323349118232727,
      "learning_rate": 0.0001606491484836782,
      "loss": 0.4234,
      "step": 144
    },
    {
      "epoch": 1.2393162393162394,
      "grad_norm": 0.42880019545555115,
      "learning_rate": 0.00016010234412083086,
      "loss": 0.4897,
      "step": 145
    },
    {
      "epoch": 1.2478632478632479,
      "grad_norm": 0.4045426845550537,
      "learning_rate": 0.00015955271189412598,
      "loss": 0.5023,
      "step": 146
    },
    {
      "epoch": 1.2564102564102564,
      "grad_norm": 0.41800403594970703,
      "learning_rate": 0.00015900027766420393,
      "loss": 0.3848,
      "step": 147
    },
    {
      "epoch": 1.264957264957265,
      "grad_norm": 0.42741239070892334,
      "learning_rate": 0.00015844506742354164,
      "loss": 0.5103,
      "step": 148
    },
    {
      "epoch": 1.2735042735042734,
      "grad_norm": 0.4664075970649719,
      "learning_rate": 0.00015788710729522953,
      "loss": 0.6579,
      "step": 149
    },
    {
      "epoch": 1.282051282051282,
      "grad_norm": 0.5985051393508911,
      "learning_rate": 0.00015732642353174259,
      "loss": 0.717,
      "step": 150
    },
    {
      "epoch": 1.2905982905982907,
      "grad_norm": 0.35360172390937805,
      "learning_rate": 0.0001567630425137049,
      "loss": 0.357,
      "step": 151
    },
    {
      "epoch": 1.2991452991452992,
      "grad_norm": 0.44152647256851196,
      "learning_rate": 0.00015619699074864864,
      "loss": 0.4689,
      "step": 152
    },
    {
      "epoch": 1.3076923076923077,
      "grad_norm": 0.49423593282699585,
      "learning_rate": 0.00015562829486976673,
      "loss": 0.4921,
      "step": 153
    },
    {
      "epoch": 1.3162393162393162,
      "grad_norm": 0.43744710087776184,
      "learning_rate": 0.00015505698163465986,
      "loss": 0.4925,
      "step": 154
    },
    {
      "epoch": 1.3247863247863247,
      "grad_norm": 0.46788179874420166,
      "learning_rate": 0.00015448307792407734,
      "loss": 0.4274,
      "step": 155
    },
    {
      "epoch": 1.3333333333333333,
      "grad_norm": 0.3874301016330719,
      "learning_rate": 0.00015390661074065256,
      "loss": 0.4548,
      "step": 156
    },
    {
      "epoch": 1.341880341880342,
      "grad_norm": 0.4863063097000122,
      "learning_rate": 0.00015332760720763232,
      "loss": 0.8016,
      "step": 157
    },
    {
      "epoch": 1.3504273504273505,
      "grad_norm": 0.5121601819992065,
      "learning_rate": 0.00015274609456760073,
      "loss": 0.5428,
      "step": 158
    },
    {
      "epoch": 1.358974358974359,
      "grad_norm": 0.5354238748550415,
      "learning_rate": 0.00015216210018119733,
      "loss": 0.5371,
      "step": 159
    },
    {
      "epoch": 1.3675213675213675,
      "grad_norm": 0.41091984510421753,
      "learning_rate": 0.00015157565152583002,
      "loss": 0.3991,
      "step": 160
    },
    {
      "epoch": 1.376068376068376,
      "grad_norm": 0.5029374361038208,
      "learning_rate": 0.0001509867761943818,
      "loss": 0.6131,
      "step": 161
    },
    {
      "epoch": 1.3846153846153846,
      "grad_norm": 0.43077659606933594,
      "learning_rate": 0.00015039550189391298,
      "loss": 0.4774,
      "step": 162
    },
    {
      "epoch": 1.393162393162393,
      "grad_norm": 0.6919808387756348,
      "learning_rate": 0.0001498018564443571,
      "loss": 0.6952,
      "step": 163
    },
    {
      "epoch": 1.4017094017094016,
      "grad_norm": 0.6397038102149963,
      "learning_rate": 0.0001492058677772123,
      "loss": 0.5046,
      "step": 164
    },
    {
      "epoch": 1.4102564102564101,
      "grad_norm": 0.4719148278236389,
      "learning_rate": 0.000148607563934227,
      "loss": 0.4855,
      "step": 165
    },
    {
      "epoch": 1.4188034188034189,
      "grad_norm": 0.3356660008430481,
      "learning_rate": 0.00014800697306608044,
      "loss": 0.3561,
      "step": 166
    },
    {
      "epoch": 1.4273504273504274,
      "grad_norm": 0.5359858870506287,
      "learning_rate": 0.00014740412343105828,
      "loss": 0.6659,
      "step": 167
    },
    {
      "epoch": 1.435897435897436,
      "grad_norm": 0.4051946997642517,
      "learning_rate": 0.00014679904339372302,
      "loss": 0.3845,
      "step": 168
    },
    {
      "epoch": 1.4444444444444444,
      "grad_norm": 0.4408167004585266,
      "learning_rate": 0.00014619176142357935,
      "loss": 0.4088,
      "step": 169
    },
    {
      "epoch": 1.452991452991453,
      "grad_norm": 0.44008126854896545,
      "learning_rate": 0.0001455823060937347,
      "loss": 0.4886,
      "step": 170
    },
    {
      "epoch": 1.4615384615384617,
      "grad_norm": 0.440919429063797,
      "learning_rate": 0.00014497070607955476,
      "loss": 0.4249,
      "step": 171
    },
    {
      "epoch": 1.4700854700854702,
      "grad_norm": 0.4290974736213684,
      "learning_rate": 0.00014435699015731448,
      "loss": 0.4312,
      "step": 172
    },
    {
      "epoch": 1.4786324786324787,
      "grad_norm": 0.44406354427337646,
      "learning_rate": 0.00014374118720284388,
      "loss": 0.5196,
      "step": 173
    },
    {
      "epoch": 1.4871794871794872,
      "grad_norm": 0.5584247708320618,
      "learning_rate": 0.00014312332619016965,
      "loss": 0.5115,
      "step": 174
    },
    {
      "epoch": 1.4957264957264957,
      "grad_norm": 0.4923872947692871,
      "learning_rate": 0.0001425034361901516,
      "loss": 0.4886,
      "step": 175
    },
    {
      "epoch": 1.5042735042735043,
      "grad_norm": 0.5677529573440552,
      "learning_rate": 0.00014188154636911524,
      "loss": 0.9921,
      "step": 176
    },
    {
      "epoch": 1.5128205128205128,
      "grad_norm": 0.4640873968601227,
      "learning_rate": 0.0001412576859874791,
      "loss": 0.4434,
      "step": 177
    },
    {
      "epoch": 1.5213675213675213,
      "grad_norm": 0.41292521357536316,
      "learning_rate": 0.00014063188439837832,
      "loss": 0.3768,
      "step": 178
    },
    {
      "epoch": 1.5299145299145298,
      "grad_norm": 0.42541882395744324,
      "learning_rate": 0.0001400041710462833,
      "loss": 0.3756,
      "step": 179
    },
    {
      "epoch": 1.5384615384615383,
      "grad_norm": 0.3998405635356903,
      "learning_rate": 0.0001393745754656146,
      "loss": 0.4429,
      "step": 180
    },
    {
      "epoch": 1.547008547008547,
      "grad_norm": 0.3835245966911316,
      "learning_rate": 0.00013874312727935292,
      "loss": 0.3823,
      "step": 181
    },
    {
      "epoch": 1.5555555555555556,
      "grad_norm": 0.6474157571792603,
      "learning_rate": 0.00013810985619764572,
      "loss": 0.6064,
      "step": 182
    },
    {
      "epoch": 1.564102564102564,
      "grad_norm": 0.39667806029319763,
      "learning_rate": 0.00013747479201640914,
      "loss": 0.4091,
      "step": 183
    },
    {
      "epoch": 1.5726495726495726,
      "grad_norm": 0.523284912109375,
      "learning_rate": 0.00013683796461592604,
      "loss": 0.4822,
      "step": 184
    },
    {
      "epoch": 1.5811965811965814,
      "grad_norm": 0.42217087745666504,
      "learning_rate": 0.00013619940395944027,
      "loss": 0.4485,
      "step": 185
    },
    {
      "epoch": 1.5897435897435899,
      "grad_norm": 0.4510805308818817,
      "learning_rate": 0.00013555914009174663,
      "loss": 0.4146,
      "step": 186
    },
    {
      "epoch": 1.5982905982905984,
      "grad_norm": 0.6063178777694702,
      "learning_rate": 0.00013491720313777756,
      "loss": 0.6233,
      "step": 187
    },
    {
      "epoch": 1.606837606837607,
      "grad_norm": 0.4440964162349701,
      "learning_rate": 0.00013427362330118543,
      "loss": 0.4274,
      "step": 188
    },
    {
      "epoch": 1.6153846153846154,
      "grad_norm": 0.46828019618988037,
      "learning_rate": 0.0001336284308629216,
      "loss": 0.5501,
      "step": 189
    },
    {
      "epoch": 1.623931623931624,
      "grad_norm": 0.5643227100372314,
      "learning_rate": 0.00013298165617981172,
      "loss": 0.434,
      "step": 190
    },
    {
      "epoch": 1.6324786324786325,
      "grad_norm": 0.5508803129196167,
      "learning_rate": 0.00013233332968312715,
      "loss": 0.5851,
      "step": 191
    },
    {
      "epoch": 1.641025641025641,
      "grad_norm": 0.5036980509757996,
      "learning_rate": 0.0001316834818771535,
      "loss": 0.4734,
      "step": 192
    },
    {
      "epoch": 1.6495726495726495,
      "grad_norm": 0.5561577677726746,
      "learning_rate": 0.00013103214333775521,
      "loss": 0.4829,
      "step": 193
    },
    {
      "epoch": 1.658119658119658,
      "grad_norm": 0.4673156142234802,
      "learning_rate": 0.00013037934471093682,
      "loss": 0.4453,
      "step": 194
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 0.4580596685409546,
      "learning_rate": 0.00012972511671140125,
      "loss": 0.3879,
      "step": 195
    },
    {
      "epoch": 1.6752136752136753,
      "grad_norm": 1.055134892463684,
      "learning_rate": 0.00012906949012110456,
      "loss": 0.6662,
      "step": 196
    },
    {
      "epoch": 1.6837606837606838,
      "grad_norm": 0.512134850025177,
      "learning_rate": 0.00012841249578780757,
      "loss": 0.579,
      "step": 197
    },
    {
      "epoch": 1.6923076923076923,
      "grad_norm": 0.3832627534866333,
      "learning_rate": 0.00012775416462362457,
      "loss": 0.4728,
      "step": 198
    },
    {
      "epoch": 1.7008547008547008,
      "grad_norm": 0.49170422554016113,
      "learning_rate": 0.00012709452760356884,
      "loss": 1.0981,
      "step": 199
    },
    {
      "epoch": 1.7094017094017095,
      "grad_norm": 0.3917134702205658,
      "learning_rate": 0.00012643361576409516,
      "loss": 0.4617,
      "step": 200
    },
    {
      "epoch": 1.717948717948718,
      "grad_norm": 0.49506673216819763,
      "learning_rate": 0.00012577146020163968,
      "loss": 0.4588,
      "step": 201
    },
    {
      "epoch": 1.7264957264957266,
      "grad_norm": 0.5297417640686035,
      "learning_rate": 0.00012510809207115666,
      "loss": 0.4749,
      "step": 202
    },
    {
      "epoch": 1.735042735042735,
      "grad_norm": 0.40945613384246826,
      "learning_rate": 0.00012444354258465268,
      "loss": 0.422,
      "step": 203
    },
    {
      "epoch": 1.7435897435897436,
      "grad_norm": 0.6515336632728577,
      "learning_rate": 0.00012377784300971807,
      "loss": 0.5492,
      "step": 204
    },
    {
      "epoch": 1.7521367521367521,
      "grad_norm": 0.7161153554916382,
      "learning_rate": 0.0001231110246680558,
      "loss": 0.7015,
      "step": 205
    },
    {
      "epoch": 1.7606837606837606,
      "grad_norm": 0.5005588531494141,
      "learning_rate": 0.00012244311893400763,
      "loss": 0.4901,
      "step": 206
    },
    {
      "epoch": 1.7692307692307692,
      "grad_norm": 0.5528268814086914,
      "learning_rate": 0.00012177415723307808,
      "loss": 0.4627,
      "step": 207
    },
    {
      "epoch": 1.7777777777777777,
      "grad_norm": 0.41800305247306824,
      "learning_rate": 0.00012110417104045575,
      "loss": 0.4003,
      "step": 208
    },
    {
      "epoch": 1.7863247863247862,
      "grad_norm": 0.42275306582450867,
      "learning_rate": 0.00012043319187953241,
      "loss": 0.5079,
      "step": 209
    },
    {
      "epoch": 1.7948717948717947,
      "grad_norm": 0.5063667893409729,
      "learning_rate": 0.00011976125132041974,
      "loss": 0.4823,
      "step": 210
    },
    {
      "epoch": 1.8034188034188035,
      "grad_norm": 0.5957720875740051,
      "learning_rate": 0.00011908838097846404,
      "loss": 0.584,
      "step": 211
    },
    {
      "epoch": 1.811965811965812,
      "grad_norm": 0.7528586983680725,
      "learning_rate": 0.00011841461251275867,
      "loss": 0.5715,
      "step": 212
    },
    {
      "epoch": 1.8205128205128205,
      "grad_norm": 0.572335958480835,
      "learning_rate": 0.00011773997762465429,
      "loss": 0.4524,
      "step": 213
    },
    {
      "epoch": 1.8290598290598292,
      "grad_norm": 0.47848159074783325,
      "learning_rate": 0.0001170645080562676,
      "loss": 0.5314,
      "step": 214
    },
    {
      "epoch": 1.8376068376068377,
      "grad_norm": 0.40428680181503296,
      "learning_rate": 0.00011638823558898762,
      "loss": 0.3909,
      "step": 215
    },
    {
      "epoch": 1.8461538461538463,
      "grad_norm": 0.44216388463974,
      "learning_rate": 0.00011571119204198037,
      "loss": 0.446,
      "step": 216
    },
    {
      "epoch": 1.8547008547008548,
      "grad_norm": 0.43098968267440796,
      "learning_rate": 0.00011503340927069189,
      "loss": 0.5271,
      "step": 217
    },
    {
      "epoch": 1.8632478632478633,
      "grad_norm": 0.39306917786598206,
      "learning_rate": 0.00011435491916534919,
      "loss": 0.4278,
      "step": 218
    },
    {
      "epoch": 1.8717948717948718,
      "grad_norm": 0.41507989168167114,
      "learning_rate": 0.00011367575364946006,
      "loss": 0.4277,
      "step": 219
    },
    {
      "epoch": 1.8803418803418803,
      "grad_norm": 0.4549039602279663,
      "learning_rate": 0.00011299594467831078,
      "loss": 0.4413,
      "step": 220
    },
    {
      "epoch": 1.8888888888888888,
      "grad_norm": 0.5150819420814514,
      "learning_rate": 0.00011231552423746283,
      "loss": 0.5057,
      "step": 221
    },
    {
      "epoch": 1.8974358974358974,
      "grad_norm": 0.4306662678718567,
      "learning_rate": 0.00011163452434124773,
      "loss": 0.4069,
      "step": 222
    },
    {
      "epoch": 1.9059829059829059,
      "grad_norm": 0.4371308386325836,
      "learning_rate": 0.00011095297703126093,
      "loss": 0.4142,
      "step": 223
    },
    {
      "epoch": 1.9145299145299144,
      "grad_norm": 0.4569212794303894,
      "learning_rate": 0.00011027091437485404,
      "loss": 0.4749,
      "step": 224
    },
    {
      "epoch": 1.9230769230769231,
      "grad_norm": 0.6140167117118835,
      "learning_rate": 0.00010958836846362621,
      "loss": 0.3719,
      "step": 225
    },
    {
      "epoch": 1.9316239316239316,
      "grad_norm": 0.44062310457229614,
      "learning_rate": 0.00010890537141191417,
      "loss": 0.4199,
      "step": 226
    },
    {
      "epoch": 1.9401709401709402,
      "grad_norm": 0.8398754596710205,
      "learning_rate": 0.00010822195535528106,
      "loss": 0.4809,
      "step": 227
    },
    {
      "epoch": 1.9487179487179487,
      "grad_norm": 0.41064441204071045,
      "learning_rate": 0.00010753815244900458,
      "loss": 0.4311,
      "step": 228
    },
    {
      "epoch": 1.9572649572649574,
      "grad_norm": 0.4826710820198059,
      "learning_rate": 0.00010685399486656406,
      "loss": 0.3993,
      "step": 229
    },
    {
      "epoch": 1.965811965811966,
      "grad_norm": 0.4640951454639435,
      "learning_rate": 0.00010616951479812658,
      "loss": 0.4229,
      "step": 230
    },
    {
      "epoch": 1.9743589743589745,
      "grad_norm": 0.5007233619689941,
      "learning_rate": 0.00010548474444903247,
      "loss": 0.3777,
      "step": 231
    },
    {
      "epoch": 1.982905982905983,
      "grad_norm": 0.4510847330093384,
      "learning_rate": 0.00010479971603828,
      "loss": 0.4129,
      "step": 232
    },
    {
      "epoch": 1.9914529914529915,
      "grad_norm": 0.388022243976593,
      "learning_rate": 0.00010411446179700943,
      "loss": 0.4013,
      "step": 233
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.6255751848220825,
      "learning_rate": 0.00010342901396698659,
      "loss": 0.3998,
      "step": 234
    },
    {
      "epoch": 2.0085470085470085,
      "grad_norm": 0.35581013560295105,
      "learning_rate": 0.00010274340479908568,
      "loss": 0.225,
      "step": 235
    },
    {
      "epoch": 2.017094017094017,
      "grad_norm": 0.38416168093681335,
      "learning_rate": 0.00010205766655177215,
      "loss": 0.2054,
      "step": 236
    },
    {
      "epoch": 2.0256410256410255,
      "grad_norm": 0.40108367800712585,
      "learning_rate": 0.00010137183148958463,
      "loss": 0.3509,
      "step": 237
    },
    {
      "epoch": 2.034188034188034,
      "grad_norm": 0.42894577980041504,
      "learning_rate": 0.00010068593188161697,
      "loss": 0.2074,
      "step": 238
    },
    {
      "epoch": 2.0427350427350426,
      "grad_norm": 0.529785692691803,
      "learning_rate": 0.0001,
      "loss": 0.2683,
      "step": 239
    },
    {
      "epoch": 2.051282051282051,
      "grad_norm": 0.49012964963912964,
      "learning_rate": 9.931406811838308e-05,
      "loss": 0.1633,
      "step": 240
    },
    {
      "epoch": 2.0598290598290596,
      "grad_norm": 0.578618049621582,
      "learning_rate": 9.862816851041541e-05,
      "loss": 0.1707,
      "step": 241
    },
    {
      "epoch": 2.0683760683760686,
      "grad_norm": 0.5279390215873718,
      "learning_rate": 9.79423334482279e-05,
      "loss": 0.2751,
      "step": 242
    },
    {
      "epoch": 2.076923076923077,
      "grad_norm": 0.6561891436576843,
      "learning_rate": 9.725659520091433e-05,
      "loss": 0.2322,
      "step": 243
    },
    {
      "epoch": 2.0854700854700856,
      "grad_norm": 0.6351572275161743,
      "learning_rate": 9.657098603301346e-05,
      "loss": 0.2151,
      "step": 244
    },
    {
      "epoch": 2.094017094017094,
      "grad_norm": 0.5479522347450256,
      "learning_rate": 9.588553820299056e-05,
      "loss": 0.2322,
      "step": 245
    },
    {
      "epoch": 2.1025641025641026,
      "grad_norm": 0.6824434995651245,
      "learning_rate": 9.520028396172003e-05,
      "loss": 0.2547,
      "step": 246
    },
    {
      "epoch": 2.111111111111111,
      "grad_norm": 0.5938841700553894,
      "learning_rate": 9.451525555096753e-05,
      "loss": 0.21,
      "step": 247
    },
    {
      "epoch": 2.1196581196581197,
      "grad_norm": 0.4802488386631012,
      "learning_rate": 9.383048520187344e-05,
      "loss": 0.1805,
      "step": 248
    },
    {
      "epoch": 2.128205128205128,
      "grad_norm": 0.5632796883583069,
      "learning_rate": 9.314600513343595e-05,
      "loss": 0.2037,
      "step": 249
    },
    {
      "epoch": 2.1367521367521367,
      "grad_norm": 0.7644915580749512,
      "learning_rate": 9.246184755099545e-05,
      "loss": 0.2885,
      "step": 250
    },
    {
      "epoch": 2.1452991452991452,
      "grad_norm": 0.6049590110778809,
      "learning_rate": 9.177804464471898e-05,
      "loss": 0.3792,
      "step": 251
    },
    {
      "epoch": 2.1538461538461537,
      "grad_norm": 0.6754825115203857,
      "learning_rate": 9.109462858808586e-05,
      "loss": 0.3549,
      "step": 252
    },
    {
      "epoch": 2.1623931623931623,
      "grad_norm": 0.5245388746261597,
      "learning_rate": 9.041163153637381e-05,
      "loss": 0.2115,
      "step": 253
    },
    {
      "epoch": 2.1709401709401708,
      "grad_norm": 0.6746419668197632,
      "learning_rate": 8.972908562514598e-05,
      "loss": 0.2281,
      "step": 254
    },
    {
      "epoch": 2.1794871794871793,
      "grad_norm": 0.7673494815826416,
      "learning_rate": 8.904702296873912e-05,
      "loss": 0.2436,
      "step": 255
    },
    {
      "epoch": 2.1880341880341883,
      "grad_norm": 0.6082167625427246,
      "learning_rate": 8.836547565875227e-05,
      "loss": 0.2141,
      "step": 256
    },
    {
      "epoch": 2.1965811965811968,
      "grad_norm": 0.611365795135498,
      "learning_rate": 8.76844757625372e-05,
      "loss": 0.75,
      "step": 257
    },
    {
      "epoch": 2.2051282051282053,
      "grad_norm": 0.5843011736869812,
      "learning_rate": 8.70040553216892e-05,
      "loss": 0.2264,
      "step": 258
    },
    {
      "epoch": 2.213675213675214,
      "grad_norm": 0.7557019591331482,
      "learning_rate": 8.632424635053997e-05,
      "loss": 0.3125,
      "step": 259
    },
    {
      "epoch": 2.2222222222222223,
      "grad_norm": 0.5172224640846252,
      "learning_rate": 8.564508083465079e-05,
      "loss": 0.1868,
      "step": 260
    },
    {
      "epoch": 2.230769230769231,
      "grad_norm": 0.5741866827011108,
      "learning_rate": 8.496659072930813e-05,
      "loss": 0.2006,
      "step": 261
    },
    {
      "epoch": 2.2393162393162394,
      "grad_norm": 0.5482658743858337,
      "learning_rate": 8.428880795801965e-05,
      "loss": 0.1925,
      "step": 262
    },
    {
      "epoch": 2.247863247863248,
      "grad_norm": 0.6574217677116394,
      "learning_rate": 8.36117644110124e-05,
      "loss": 0.1934,
      "step": 263
    },
    {
      "epoch": 2.2564102564102564,
      "grad_norm": 0.556007444858551,
      "learning_rate": 8.293549194373243e-05,
      "loss": 0.1876,
      "step": 264
    },
    {
      "epoch": 2.264957264957265,
      "grad_norm": 0.49290239810943604,
      "learning_rate": 8.226002237534572e-05,
      "loss": 0.1896,
      "step": 265
    },
    {
      "epoch": 2.2735042735042734,
      "grad_norm": 1.305418848991394,
      "learning_rate": 8.158538748724139e-05,
      "loss": 0.2409,
      "step": 266
    },
    {
      "epoch": 2.282051282051282,
      "grad_norm": 0.7417373657226562,
      "learning_rate": 8.091161902153595e-05,
      "loss": 0.2225,
      "step": 267
    },
    {
      "epoch": 2.2905982905982905,
      "grad_norm": 0.6486262083053589,
      "learning_rate": 8.023874867958027e-05,
      "loss": 0.3485,
      "step": 268
    },
    {
      "epoch": 2.299145299145299,
      "grad_norm": 0.5938575267791748,
      "learning_rate": 7.95668081204676e-05,
      "loss": 0.1438,
      "step": 269
    },
    {
      "epoch": 2.3076923076923075,
      "grad_norm": 0.5281603932380676,
      "learning_rate": 7.889582895954427e-05,
      "loss": 0.2291,
      "step": 270
    },
    {
      "epoch": 2.316239316239316,
      "grad_norm": 0.5182002782821655,
      "learning_rate": 7.822584276692191e-05,
      "loss": 0.2025,
      "step": 271
    },
    {
      "epoch": 2.324786324786325,
      "grad_norm": 0.5637621879577637,
      "learning_rate": 7.755688106599241e-05,
      "loss": 0.2628,
      "step": 272
    },
    {
      "epoch": 2.3333333333333335,
      "grad_norm": 0.6143607497215271,
      "learning_rate": 7.688897533194424e-05,
      "loss": 0.2666,
      "step": 273
    },
    {
      "epoch": 2.341880341880342,
      "grad_norm": 0.6833069920539856,
      "learning_rate": 7.622215699028196e-05,
      "loss": 0.1836,
      "step": 274
    },
    {
      "epoch": 2.3504273504273505,
      "grad_norm": 0.5094839334487915,
      "learning_rate": 7.555645741534736e-05,
      "loss": 0.2325,
      "step": 275
    },
    {
      "epoch": 2.358974358974359,
      "grad_norm": 1.1273527145385742,
      "learning_rate": 7.489190792884338e-05,
      "loss": 0.2334,
      "step": 276
    },
    {
      "epoch": 2.3675213675213675,
      "grad_norm": 1.1840169429779053,
      "learning_rate": 7.422853979836034e-05,
      "loss": 0.2739,
      "step": 277
    },
    {
      "epoch": 2.376068376068376,
      "grad_norm": 0.5058795213699341,
      "learning_rate": 7.356638423590485e-05,
      "loss": 0.191,
      "step": 278
    },
    {
      "epoch": 2.3846153846153846,
      "grad_norm": 0.5227871537208557,
      "learning_rate": 7.290547239643117e-05,
      "loss": 0.2161,
      "step": 279
    },
    {
      "epoch": 2.393162393162393,
      "grad_norm": 0.4699568748474121,
      "learning_rate": 7.224583537637544e-05,
      "loss": 0.2204,
      "step": 280
    },
    {
      "epoch": 2.4017094017094016,
      "grad_norm": 1.2125422954559326,
      "learning_rate": 7.158750421219244e-05,
      "loss": 0.2986,
      "step": 281
    },
    {
      "epoch": 2.41025641025641,
      "grad_norm": 0.5665399432182312,
      "learning_rate": 7.093050987889547e-05,
      "loss": 0.2013,
      "step": 282
    },
    {
      "epoch": 2.4188034188034186,
      "grad_norm": 0.5025288462638855,
      "learning_rate": 7.027488328859876e-05,
      "loss": 0.1908,
      "step": 283
    },
    {
      "epoch": 2.427350427350427,
      "grad_norm": 0.5425287485122681,
      "learning_rate": 6.96206552890632e-05,
      "loss": 0.2184,
      "step": 284
    },
    {
      "epoch": 2.435897435897436,
      "grad_norm": 0.640358567237854,
      "learning_rate": 6.896785666224481e-05,
      "loss": 0.2142,
      "step": 285
    },
    {
      "epoch": 2.4444444444444446,
      "grad_norm": 0.7539050579071045,
      "learning_rate": 6.831651812284652e-05,
      "loss": 0.1929,
      "step": 286
    },
    {
      "epoch": 2.452991452991453,
      "grad_norm": 0.5728548765182495,
      "learning_rate": 6.766667031687286e-05,
      "loss": 0.2363,
      "step": 287
    },
    {
      "epoch": 2.4615384615384617,
      "grad_norm": 0.5997483730316162,
      "learning_rate": 6.701834382018832e-05,
      "loss": 0.1565,
      "step": 288
    },
    {
      "epoch": 2.47008547008547,
      "grad_norm": 0.5362933278083801,
      "learning_rate": 6.637156913707839e-05,
      "loss": 0.189,
      "step": 289
    },
    {
      "epoch": 2.4786324786324787,
      "grad_norm": 0.5454270243644714,
      "learning_rate": 6.572637669881458e-05,
      "loss": 0.2281,
      "step": 290
    },
    {
      "epoch": 2.4871794871794872,
      "grad_norm": 1.2282865047454834,
      "learning_rate": 6.508279686222243e-05,
      "loss": 0.3461,
      "step": 291
    },
    {
      "epoch": 2.4957264957264957,
      "grad_norm": 0.5579683184623718,
      "learning_rate": 6.444085990825338e-05,
      "loss": 0.1748,
      "step": 292
    },
    {
      "epoch": 2.5042735042735043,
      "grad_norm": 0.4940085709095001,
      "learning_rate": 6.380059604055974e-05,
      "loss": 0.216,
      "step": 293
    },
    {
      "epoch": 2.5128205128205128,
      "grad_norm": 0.7586284279823303,
      "learning_rate": 6.316203538407397e-05,
      "loss": 0.3407,
      "step": 294
    },
    {
      "epoch": 2.5213675213675213,
      "grad_norm": 0.5672574639320374,
      "learning_rate": 6.252520798359092e-05,
      "loss": 0.2409,
      "step": 295
    },
    {
      "epoch": 2.52991452991453,
      "grad_norm": 0.5276070833206177,
      "learning_rate": 6.18901438023543e-05,
      "loss": 0.1942,
      "step": 296
    },
    {
      "epoch": 2.5384615384615383,
      "grad_norm": 0.533193051815033,
      "learning_rate": 6.125687272064713e-05,
      "loss": 0.2027,
      "step": 297
    },
    {
      "epoch": 2.547008547008547,
      "grad_norm": 0.4329485595226288,
      "learning_rate": 6.0625424534385425e-05,
      "loss": 0.1332,
      "step": 298
    },
    {
      "epoch": 2.5555555555555554,
      "grad_norm": 0.6938366293907166,
      "learning_rate": 5.9995828953716695e-05,
      "loss": 0.2901,
      "step": 299
    },
    {
      "epoch": 2.564102564102564,
      "grad_norm": 0.5417672991752625,
      "learning_rate": 5.936811560162169e-05,
      "loss": 0.2002,
      "step": 300
    },
    {
      "epoch": 2.5726495726495724,
      "grad_norm": 0.4884963035583496,
      "learning_rate": 5.87423140125209e-05,
      "loss": 0.1767,
      "step": 301
    },
    {
      "epoch": 2.5811965811965814,
      "grad_norm": 0.48789867758750916,
      "learning_rate": 5.811845363088477e-05,
      "loss": 0.1532,
      "step": 302
    },
    {
      "epoch": 2.58974358974359,
      "grad_norm": 0.6069095134735107,
      "learning_rate": 5.749656380984844e-05,
      "loss": 0.2281,
      "step": 303
    },
    {
      "epoch": 2.5982905982905984,
      "grad_norm": 0.7241275310516357,
      "learning_rate": 5.687667380983037e-05,
      "loss": 0.2334,
      "step": 304
    },
    {
      "epoch": 2.606837606837607,
      "grad_norm": 0.5266017317771912,
      "learning_rate": 5.625881279715615e-05,
      "loss": 0.2092,
      "step": 305
    },
    {
      "epoch": 2.6153846153846154,
      "grad_norm": 0.8533578515052795,
      "learning_rate": 5.5643009842685554e-05,
      "loss": 0.3899,
      "step": 306
    },
    {
      "epoch": 2.623931623931624,
      "grad_norm": 0.545475423336029,
      "learning_rate": 5.502929392044528e-05,
      "loss": 0.2184,
      "step": 307
    },
    {
      "epoch": 2.6324786324786325,
      "grad_norm": 0.560219943523407,
      "learning_rate": 5.4417693906265365e-05,
      "loss": 0.1736,
      "step": 308
    },
    {
      "epoch": 2.641025641025641,
      "grad_norm": 0.6965181827545166,
      "learning_rate": 5.380823857642069e-05,
      "loss": 0.2107,
      "step": 309
    },
    {
      "epoch": 2.6495726495726495,
      "grad_norm": 0.6126929521560669,
      "learning_rate": 5.3200956606277006e-05,
      "loss": 0.2235,
      "step": 310
    },
    {
      "epoch": 2.658119658119658,
      "grad_norm": 0.5365376472473145,
      "learning_rate": 5.259587656894174e-05,
      "loss": 0.197,
      "step": 311
    },
    {
      "epoch": 2.6666666666666665,
      "grad_norm": 0.7516108751296997,
      "learning_rate": 5.199302693391959e-05,
      "loss": 0.1798,
      "step": 312
    },
    {
      "epoch": 2.6752136752136755,
      "grad_norm": 0.9562693238258362,
      "learning_rate": 5.139243606577302e-05,
      "loss": 0.239,
      "step": 313
    },
    {
      "epoch": 2.683760683760684,
      "grad_norm": 0.68837970495224,
      "learning_rate": 5.0794132222787707e-05,
      "loss": 0.1942,
      "step": 314
    },
    {
      "epoch": 2.6923076923076925,
      "grad_norm": 0.5753955841064453,
      "learning_rate": 5.019814355564292e-05,
      "loss": 0.2182,
      "step": 315
    },
    {
      "epoch": 2.700854700854701,
      "grad_norm": 0.5992863774299622,
      "learning_rate": 4.960449810608705e-05,
      "loss": 0.2762,
      "step": 316
    },
    {
      "epoch": 2.7094017094017095,
      "grad_norm": 0.5354930758476257,
      "learning_rate": 4.90132238056182e-05,
      "loss": 0.2099,
      "step": 317
    },
    {
      "epoch": 2.717948717948718,
      "grad_norm": 1.1121389865875244,
      "learning_rate": 4.8424348474170014e-05,
      "loss": 0.4311,
      "step": 318
    },
    {
      "epoch": 2.7264957264957266,
      "grad_norm": 0.6793778538703918,
      "learning_rate": 4.783789981880267e-05,
      "loss": 0.1744,
      "step": 319
    },
    {
      "epoch": 2.735042735042735,
      "grad_norm": 0.4686272144317627,
      "learning_rate": 4.725390543239929e-05,
      "loss": 0.124,
      "step": 320
    },
    {
      "epoch": 2.7435897435897436,
      "grad_norm": 0.6915261745452881,
      "learning_rate": 4.667239279236768e-05,
      "loss": 0.2213,
      "step": 321
    },
    {
      "epoch": 2.752136752136752,
      "grad_norm": 0.6020219326019287,
      "learning_rate": 4.609338925934743e-05,
      "loss": 0.2431,
      "step": 322
    },
    {
      "epoch": 2.7606837606837606,
      "grad_norm": 1.5203578472137451,
      "learning_rate": 4.551692207592265e-05,
      "loss": 0.2611,
      "step": 323
    },
    {
      "epoch": 2.769230769230769,
      "grad_norm": 0.5536499619483948,
      "learning_rate": 4.494301836534016e-05,
      "loss": 0.1738,
      "step": 324
    },
    {
      "epoch": 2.7777777777777777,
      "grad_norm": 0.6723442077636719,
      "learning_rate": 4.4371705130233275e-05,
      "loss": 0.2588,
      "step": 325
    },
    {
      "epoch": 2.786324786324786,
      "grad_norm": 0.5099270939826965,
      "learning_rate": 4.380300925135138e-05,
      "loss": 0.1586,
      "step": 326
    },
    {
      "epoch": 2.7948717948717947,
      "grad_norm": 0.6832568645477295,
      "learning_rate": 4.3236957486295115e-05,
      "loss": 0.2723,
      "step": 327
    },
    {
      "epoch": 2.8034188034188032,
      "grad_norm": 0.5315645337104797,
      "learning_rate": 4.267357646825746e-05,
      "loss": 0.119,
      "step": 328
    },
    {
      "epoch": 2.8119658119658117,
      "grad_norm": 0.5762079954147339,
      "learning_rate": 4.211289270477047e-05,
      "loss": 0.1864,
      "step": 329
    },
    {
      "epoch": 2.8205128205128203,
      "grad_norm": 0.5557078123092651,
      "learning_rate": 4.1554932576458415e-05,
      "loss": 0.2014,
      "step": 330
    },
    {
      "epoch": 2.8290598290598292,
      "grad_norm": 0.664307713508606,
      "learning_rate": 4.0999722335796075e-05,
      "loss": 0.2065,
      "step": 331
    },
    {
      "epoch": 2.8376068376068377,
      "grad_norm": 0.6074508428573608,
      "learning_rate": 4.044728810587406e-05,
      "loss": 0.1941,
      "step": 332
    },
    {
      "epoch": 2.8461538461538463,
      "grad_norm": 0.680601954460144,
      "learning_rate": 3.989765587916914e-05,
      "loss": 0.1957,
      "step": 333
    },
    {
      "epoch": 2.8547008547008548,
      "grad_norm": 0.6200893521308899,
      "learning_rate": 3.935085151632185e-05,
      "loss": 0.2238,
      "step": 334
    },
    {
      "epoch": 2.8632478632478633,
      "grad_norm": 0.5177062153816223,
      "learning_rate": 3.8806900744919205e-05,
      "loss": 0.186,
      "step": 335
    },
    {
      "epoch": 2.871794871794872,
      "grad_norm": 0.6182951331138611,
      "learning_rate": 3.826582915828468e-05,
      "loss": 0.2267,
      "step": 336
    },
    {
      "epoch": 2.8803418803418803,
      "grad_norm": 0.6131262183189392,
      "learning_rate": 3.7727662214273495e-05,
      "loss": 0.1839,
      "step": 337
    },
    {
      "epoch": 2.888888888888889,
      "grad_norm": 0.44177737832069397,
      "learning_rate": 3.719242523407539e-05,
      "loss": 0.1495,
      "step": 338
    },
    {
      "epoch": 2.8974358974358974,
      "grad_norm": 0.643297553062439,
      "learning_rate": 3.666014340102268e-05,
      "loss": 0.1797,
      "step": 339
    },
    {
      "epoch": 2.905982905982906,
      "grad_norm": 0.8198469877243042,
      "learning_rate": 3.613084175940578e-05,
      "loss": 0.2215,
      "step": 340
    },
    {
      "epoch": 2.9145299145299144,
      "grad_norm": 0.8104227781295776,
      "learning_rate": 3.5604545213294616e-05,
      "loss": 0.2336,
      "step": 341
    },
    {
      "epoch": 2.9230769230769234,
      "grad_norm": 0.6274427175521851,
      "learning_rate": 3.508127852536698e-05,
      "loss": 0.193,
      "step": 342
    },
    {
      "epoch": 2.931623931623932,
      "grad_norm": 0.6716679930686951,
      "learning_rate": 3.456106631574336e-05,
      "loss": 0.2574,
      "step": 343
    },
    {
      "epoch": 2.9401709401709404,
      "grad_norm": 0.5092660784721375,
      "learning_rate": 3.4043933060828605e-05,
      "loss": 0.2021,
      "step": 344
    },
    {
      "epoch": 2.948717948717949,
      "grad_norm": 0.6550085544586182,
      "learning_rate": 3.352990309216022e-05,
      "loss": 0.1957,
      "step": 345
    },
    {
      "epoch": 2.9572649572649574,
      "grad_norm": 0.6786168217658997,
      "learning_rate": 3.3019000595263574e-05,
      "loss": 0.1915,
      "step": 346
    },
    {
      "epoch": 2.965811965811966,
      "grad_norm": 0.6020535826683044,
      "learning_rate": 3.251124960851408e-05,
      "loss": 0.1923,
      "step": 347
    },
    {
      "epoch": 2.9743589743589745,
      "grad_norm": 0.7327200770378113,
      "learning_rate": 3.200667402200586e-05,
      "loss": 0.2175,
      "step": 348
    },
    {
      "epoch": 2.982905982905983,
      "grad_norm": 0.6219268441200256,
      "learning_rate": 3.1505297576428075e-05,
      "loss": 0.2009,
      "step": 349
    },
    {
      "epoch": 2.9914529914529915,
      "grad_norm": 0.5827919840812683,
      "learning_rate": 3.100714386194757e-05,
      "loss": 0.2109,
      "step": 350
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.543898344039917,
      "learning_rate": 3.0512236317099175e-05,
      "loss": 0.1786,
      "step": 351
    },
    {
      "epoch": 3.0085470085470085,
      "grad_norm": 0.4246410131454468,
      "learning_rate": 3.0020598227682795e-05,
      "loss": 0.1215,
      "step": 352
    },
    {
      "epoch": 3.017094017094017,
      "grad_norm": 0.46924176812171936,
      "learning_rate": 2.953225272566782e-05,
      "loss": 0.0977,
      "step": 353
    },
    {
      "epoch": 3.0256410256410255,
      "grad_norm": 0.4527635872364044,
      "learning_rate": 2.904722278810471e-05,
      "loss": 0.0965,
      "step": 354
    },
    {
      "epoch": 3.034188034188034,
      "grad_norm": 0.46280795335769653,
      "learning_rate": 2.8565531236043997e-05,
      "loss": 0.1156,
      "step": 355
    },
    {
      "epoch": 3.0427350427350426,
      "grad_norm": 0.4730513095855713,
      "learning_rate": 2.8087200733462425e-05,
      "loss": 0.1271,
      "step": 356
    },
    {
      "epoch": 3.051282051282051,
      "grad_norm": 0.45704519748687744,
      "learning_rate": 2.7612253786196664e-05,
      "loss": 0.0993,
      "step": 357
    },
    {
      "epoch": 3.0598290598290596,
      "grad_norm": 0.42013949155807495,
      "learning_rate": 2.7140712740884376e-05,
      "loss": 0.0626,
      "step": 358
    },
    {
      "epoch": 3.0683760683760686,
      "grad_norm": 0.5590519905090332,
      "learning_rate": 2.667259978391281e-05,
      "loss": 0.1003,
      "step": 359
    },
    {
      "epoch": 3.076923076923077,
      "grad_norm": 0.49460965394973755,
      "learning_rate": 2.6207936940374767e-05,
      "loss": 0.1081,
      "step": 360
    },
    {
      "epoch": 3.0854700854700856,
      "grad_norm": 0.556666910648346,
      "learning_rate": 2.5746746073032625e-05,
      "loss": 0.0839,
      "step": 361
    },
    {
      "epoch": 3.094017094017094,
      "grad_norm": 0.37445566058158875,
      "learning_rate": 2.5289048881289256e-05,
      "loss": 0.074,
      "step": 362
    },
    {
      "epoch": 3.1025641025641026,
      "grad_norm": 0.6359053254127502,
      "learning_rate": 2.4834866900167475e-05,
      "loss": 0.0835,
      "step": 363
    },
    {
      "epoch": 3.111111111111111,
      "grad_norm": 0.4775371551513672,
      "learning_rate": 2.4384221499296466e-05,
      "loss": 0.0871,
      "step": 364
    },
    {
      "epoch": 3.1196581196581197,
      "grad_norm": 0.4734942615032196,
      "learning_rate": 2.393713388190648e-05,
      "loss": 0.0736,
      "step": 365
    },
    {
      "epoch": 3.128205128205128,
      "grad_norm": 0.5424053072929382,
      "learning_rate": 2.3493625083831217e-05,
      "loss": 0.1266,
      "step": 366
    },
    {
      "epoch": 3.1367521367521367,
      "grad_norm": 0.5089436173439026,
      "learning_rate": 2.3053715972518e-05,
      "loss": 0.1056,
      "step": 367
    },
    {
      "epoch": 3.1452991452991452,
      "grad_norm": 0.9008356332778931,
      "learning_rate": 2.2617427246045973e-05,
      "loss": 0.1107,
      "step": 368
    },
    {
      "epoch": 3.1538461538461537,
      "grad_norm": 0.8132445812225342,
      "learning_rate": 2.218477943215229e-05,
      "loss": 0.1306,
      "step": 369
    },
    {
      "epoch": 3.1623931623931623,
      "grad_norm": 0.7188067436218262,
      "learning_rate": 2.1755792887266234e-05,
      "loss": 0.1387,
      "step": 370
    },
    {
      "epoch": 3.1709401709401708,
      "grad_norm": 0.610883891582489,
      "learning_rate": 2.133048779555129e-05,
      "loss": 0.0733,
      "step": 371
    },
    {
      "epoch": 3.1794871794871793,
      "grad_norm": 0.7963797450065613,
      "learning_rate": 2.0908884167955824e-05,
      "loss": 0.0862,
      "step": 372
    },
    {
      "epoch": 3.1880341880341883,
      "grad_norm": 0.6727620363235474,
      "learning_rate": 2.0491001841271074e-05,
      "loss": 0.1219,
      "step": 373
    },
    {
      "epoch": 3.1965811965811968,
      "grad_norm": 0.6130483150482178,
      "learning_rate": 2.0076860477198313e-05,
      "loss": 0.0784,
      "step": 374
    },
    {
      "epoch": 3.2051282051282053,
      "grad_norm": 0.6352599859237671,
      "learning_rate": 1.9666479561423244e-05,
      "loss": 0.1099,
      "step": 375
    },
    {
      "epoch": 3.213675213675214,
      "grad_norm": 0.5709249973297119,
      "learning_rate": 1.9259878402699705e-05,
      "loss": 0.0724,
      "step": 376
    },
    {
      "epoch": 3.2222222222222223,
      "grad_norm": 0.782482385635376,
      "learning_rate": 1.8857076131940642e-05,
      "loss": 0.0999,
      "step": 377
    },
    {
      "epoch": 3.230769230769231,
      "grad_norm": 0.586031973361969,
      "learning_rate": 1.8458091701318504e-05,
      "loss": 0.0677,
      "step": 378
    },
    {
      "epoch": 3.2393162393162394,
      "grad_norm": 0.5336410403251648,
      "learning_rate": 1.806294388337305e-05,
      "loss": 0.082,
      "step": 379
    },
    {
      "epoch": 3.247863247863248,
      "grad_norm": 0.6292747259140015,
      "learning_rate": 1.7671651270128532e-05,
      "loss": 0.0945,
      "step": 380
    },
    {
      "epoch": 3.2564102564102564,
      "grad_norm": 0.5497437119483948,
      "learning_rate": 1.7284232272218504e-05,
      "loss": 0.0798,
      "step": 381
    },
    {
      "epoch": 3.264957264957265,
      "grad_norm": 0.4750556945800781,
      "learning_rate": 1.69007051180199e-05,
      "loss": 0.1399,
      "step": 382
    },
    {
      "epoch": 3.2735042735042734,
      "grad_norm": 0.6771523356437683,
      "learning_rate": 1.652108785279526e-05,
      "loss": 0.0975,
      "step": 383
    },
    {
      "epoch": 3.282051282051282,
      "grad_norm": 0.5702567100524902,
      "learning_rate": 1.6145398337843652e-05,
      "loss": 0.0612,
      "step": 384
    },
    {
      "epoch": 3.2905982905982905,
      "grad_norm": 0.6195294857025146,
      "learning_rate": 1.577365424966034e-05,
      "loss": 0.0898,
      "step": 385
    },
    {
      "epoch": 3.299145299145299,
      "grad_norm": 0.5133700370788574,
      "learning_rate": 1.540587307910508e-05,
      "loss": 0.0637,
      "step": 386
    },
    {
      "epoch": 3.3076923076923075,
      "grad_norm": 0.5438370704650879,
      "learning_rate": 1.504207213057912e-05,
      "loss": 0.0556,
      "step": 387
    },
    {
      "epoch": 3.316239316239316,
      "grad_norm": 0.7359026670455933,
      "learning_rate": 1.4682268521211073e-05,
      "loss": 0.531,
      "step": 388
    },
    {
      "epoch": 3.324786324786325,
      "grad_norm": 0.4169284999370575,
      "learning_rate": 1.43264791800515e-05,
      "loss": 0.07,
      "step": 389
    },
    {
      "epoch": 3.3333333333333335,
      "grad_norm": 0.37228280305862427,
      "learning_rate": 1.3974720847276412e-05,
      "loss": 0.0422,
      "step": 390
    },
    {
      "epoch": 3.341880341880342,
      "grad_norm": 0.4790070056915283,
      "learning_rate": 1.3627010073399604e-05,
      "loss": 0.0848,
      "step": 391
    },
    {
      "epoch": 3.3504273504273505,
      "grad_norm": 0.5548518896102905,
      "learning_rate": 1.328336321849396e-05,
      "loss": 0.1011,
      "step": 392
    },
    {
      "epoch": 3.358974358974359,
      "grad_norm": 0.43558043241500854,
      "learning_rate": 1.2943796451421686e-05,
      "loss": 0.0887,
      "step": 393
    },
    {
      "epoch": 3.3675213675213675,
      "grad_norm": 0.6515349745750427,
      "learning_rate": 1.2608325749073591e-05,
      "loss": 0.0965,
      "step": 394
    },
    {
      "epoch": 3.376068376068376,
      "grad_norm": 0.5635669827461243,
      "learning_rate": 1.227696689561727e-05,
      "loss": 0.0963,
      "step": 395
    },
    {
      "epoch": 3.3846153846153846,
      "grad_norm": 0.5640111565589905,
      "learning_rate": 1.1949735481754565e-05,
      "loss": 0.0746,
      "step": 396
    },
    {
      "epoch": 3.393162393162393,
      "grad_norm": 0.5030996203422546,
      "learning_rate": 1.1626646903987904e-05,
      "loss": 0.0653,
      "step": 397
    },
    {
      "epoch": 3.4017094017094016,
      "grad_norm": 0.8200929164886475,
      "learning_rate": 1.130771636389596e-05,
      "loss": 0.1202,
      "step": 398
    },
    {
      "epoch": 3.41025641025641,
      "grad_norm": 0.61598801612854,
      "learning_rate": 1.0992958867418357e-05,
      "loss": 0.0539,
      "step": 399
    },
    {
      "epoch": 3.4188034188034186,
      "grad_norm": 0.6611541509628296,
      "learning_rate": 1.0682389224149647e-05,
      "loss": 0.1072,
      "step": 400
    },
    {
      "epoch": 3.427350427350427,
      "grad_norm": 0.5062597393989563,
      "learning_rate": 1.037602204664252e-05,
      "loss": 0.1026,
      "step": 401
    },
    {
      "epoch": 3.435897435897436,
      "grad_norm": 0.5125107765197754,
      "learning_rate": 1.0073871749720221e-05,
      "loss": 0.0938,
      "step": 402
    },
    {
      "epoch": 3.4444444444444446,
      "grad_norm": 0.47890546917915344,
      "learning_rate": 9.775952549798406e-06,
      "loss": 0.0642,
      "step": 403
    },
    {
      "epoch": 3.452991452991453,
      "grad_norm": 0.49492985010147095,
      "learning_rate": 9.482278464216121e-06,
      "loss": 0.0762,
      "step": 404
    },
    {
      "epoch": 3.4615384615384617,
      "grad_norm": 0.7054003477096558,
      "learning_rate": 9.192863310576472e-06,
      "loss": 0.1039,
      "step": 405
    },
    {
      "epoch": 3.47008547008547,
      "grad_norm": 0.7291774749755859,
      "learning_rate": 8.907720706096224e-06,
      "loss": 0.0794,
      "step": 406
    },
    {
      "epoch": 3.4786324786324787,
      "grad_norm": 0.5116931796073914,
      "learning_rate": 8.626864066965402e-06,
      "loss": 0.0521,
      "step": 407
    },
    {
      "epoch": 3.4871794871794872,
      "grad_norm": 0.5735267996788025,
      "learning_rate": 8.350306607715774e-06,
      "loss": 0.0644,
      "step": 408
    },
    {
      "epoch": 3.4957264957264957,
      "grad_norm": 0.5400829911231995,
      "learning_rate": 8.07806134059933e-06,
      "loss": 0.0656,
      "step": 409
    },
    {
      "epoch": 3.5042735042735043,
      "grad_norm": 0.6491343975067139,
      "learning_rate": 7.810141074975818e-06,
      "loss": 0.0807,
      "step": 410
    },
    {
      "epoch": 3.5128205128205128,
      "grad_norm": 0.4385392367839813,
      "learning_rate": 7.546558416710292e-06,
      "loss": 0.0672,
      "step": 411
    },
    {
      "epoch": 3.5213675213675213,
      "grad_norm": 0.9656842350959778,
      "learning_rate": 7.287325767579756e-06,
      "loss": 0.1266,
      "step": 412
    },
    {
      "epoch": 3.52991452991453,
      "grad_norm": 0.7359745502471924,
      "learning_rate": 7.032455324689902e-06,
      "loss": 0.1098,
      "step": 413
    },
    {
      "epoch": 3.5384615384615383,
      "grad_norm": 0.36696016788482666,
      "learning_rate": 6.781959079900957e-06,
      "loss": 0.0543,
      "step": 414
    },
    {
      "epoch": 3.547008547008547,
      "grad_norm": 0.4409100413322449,
      "learning_rate": 6.535848819263679e-06,
      "loss": 0.0671,
      "step": 415
    },
    {
      "epoch": 3.5555555555555554,
      "grad_norm": 0.7295477390289307,
      "learning_rate": 6.2941361224647e-06,
      "loss": 0.1577,
      "step": 416
    },
    {
      "epoch": 3.564102564102564,
      "grad_norm": 0.6044880747795105,
      "learning_rate": 6.056832362281728e-06,
      "loss": 0.0997,
      "step": 417
    },
    {
      "epoch": 3.5726495726495724,
      "grad_norm": 0.49388715624809265,
      "learning_rate": 5.823948704048443e-06,
      "loss": 0.0605,
      "step": 418
    },
    {
      "epoch": 3.5811965811965814,
      "grad_norm": 0.5975767374038696,
      "learning_rate": 5.5954961051291384e-06,
      "loss": 0.085,
      "step": 419
    },
    {
      "epoch": 3.58974358974359,
      "grad_norm": 0.5086919665336609,
      "learning_rate": 5.371485314403202e-06,
      "loss": 0.0819,
      "step": 420
    },
    {
      "epoch": 3.5982905982905984,
      "grad_norm": 0.6395756602287292,
      "learning_rate": 5.151926871759349e-06,
      "loss": 0.1753,
      "step": 421
    },
    {
      "epoch": 3.606837606837607,
      "grad_norm": 0.6909961700439453,
      "learning_rate": 4.936831107599749e-06,
      "loss": 0.0692,
      "step": 422
    },
    {
      "epoch": 3.6153846153846154,
      "grad_norm": 0.5048989057540894,
      "learning_rate": 4.7262081423538716e-06,
      "loss": 0.0767,
      "step": 423
    },
    {
      "epoch": 3.623931623931624,
      "grad_norm": 0.4353529214859009,
      "learning_rate": 4.5200678860024885e-06,
      "loss": 0.0605,
      "step": 424
    },
    {
      "epoch": 3.6324786324786325,
      "grad_norm": 0.5182363986968994,
      "learning_rate": 4.3184200376111815e-06,
      "loss": 0.0897,
      "step": 425
    },
    {
      "epoch": 3.641025641025641,
      "grad_norm": 0.5179640650749207,
      "learning_rate": 4.121274084874194e-06,
      "loss": 0.0913,
      "step": 426
    },
    {
      "epoch": 3.6495726495726495,
      "grad_norm": 0.5857887268066406,
      "learning_rate": 3.928639303667891e-06,
      "loss": 0.0812,
      "step": 427
    },
    {
      "epoch": 3.658119658119658,
      "grad_norm": 0.5451228022575378,
      "learning_rate": 3.7405247576144054e-06,
      "loss": 0.0793,
      "step": 428
    },
    {
      "epoch": 3.6666666666666665,
      "grad_norm": 0.4748663306236267,
      "learning_rate": 3.556939297655115e-06,
      "loss": 0.0584,
      "step": 429
    },
    {
      "epoch": 3.6752136752136755,
      "grad_norm": 0.6166920065879822,
      "learning_rate": 3.3778915616342943e-06,
      "loss": 0.0841,
      "step": 430
    },
    {
      "epoch": 3.683760683760684,
      "grad_norm": 0.49632978439331055,
      "learning_rate": 3.203389973892579e-06,
      "loss": 0.0723,
      "step": 431
    },
    {
      "epoch": 3.6923076923076925,
      "grad_norm": 0.5901334881782532,
      "learning_rate": 3.0334427448706847e-06,
      "loss": 0.0846,
      "step": 432
    },
    {
      "epoch": 3.700854700854701,
      "grad_norm": 0.5666071772575378,
      "learning_rate": 2.868057870723073e-06,
      "loss": 0.0611,
      "step": 433
    },
    {
      "epoch": 3.7094017094017095,
      "grad_norm": 0.5421086549758911,
      "learning_rate": 2.707243132941717e-06,
      "loss": 0.0918,
      "step": 434
    },
    {
      "epoch": 3.717948717948718,
      "grad_norm": 0.551996648311615,
      "learning_rate": 2.5510060979899607e-06,
      "loss": 0.0677,
      "step": 435
    },
    {
      "epoch": 3.7264957264957266,
      "grad_norm": 0.4528348445892334,
      "learning_rate": 2.3993541169465837e-06,
      "loss": 0.0557,
      "step": 436
    },
    {
      "epoch": 3.735042735042735,
      "grad_norm": 0.602454662322998,
      "learning_rate": 2.2522943251597873e-06,
      "loss": 0.0964,
      "step": 437
    },
    {
      "epoch": 3.7435897435897436,
      "grad_norm": 0.5131720900535583,
      "learning_rate": 2.1098336419116625e-06,
      "loss": 0.0983,
      "step": 438
    },
    {
      "epoch": 3.752136752136752,
      "grad_norm": 0.630336344242096,
      "learning_rate": 1.971978770092431e-06,
      "loss": 0.0866,
      "step": 439
    },
    {
      "epoch": 3.7606837606837606,
      "grad_norm": 0.7707011103630066,
      "learning_rate": 1.838736195885238e-06,
      "loss": 0.1065,
      "step": 440
    },
    {
      "epoch": 3.769230769230769,
      "grad_norm": 0.402682363986969,
      "learning_rate": 1.710112188460844e-06,
      "loss": 0.0416,
      "step": 441
    },
    {
      "epoch": 3.7777777777777777,
      "grad_norm": 0.6648341417312622,
      "learning_rate": 1.5861127996827597e-06,
      "loss": 0.0921,
      "step": 442
    },
    {
      "epoch": 3.786324786324786,
      "grad_norm": 0.5792677998542786,
      "learning_rate": 1.4667438638224062e-06,
      "loss": 0.0822,
      "step": 443
    },
    {
      "epoch": 3.7948717948717947,
      "grad_norm": 0.4458604156970978,
      "learning_rate": 1.3520109972846917e-06,
      "loss": 0.0481,
      "step": 444
    },
    {
      "epoch": 3.8034188034188032,
      "grad_norm": 0.5295193791389465,
      "learning_rate": 1.2419195983436881e-06,
      "loss": 0.0961,
      "step": 445
    },
    {
      "epoch": 3.8119658119658117,
      "grad_norm": 0.6242722272872925,
      "learning_rate": 1.1364748468886687e-06,
      "loss": 0.1165,
      "step": 446
    },
    {
      "epoch": 3.8205128205128203,
      "grad_norm": 0.525872528553009,
      "learning_rate": 1.0356817041804246e-06,
      "loss": 0.0876,
      "step": 447
    },
    {
      "epoch": 3.8290598290598292,
      "grad_norm": 0.5753360390663147,
      "learning_rate": 9.395449126177291e-07,
      "loss": 0.0768,
      "step": 448
    },
    {
      "epoch": 3.8376068376068377,
      "grad_norm": 0.4595930278301239,
      "learning_rate": 8.480689955143395e-07,
      "loss": 0.0567,
      "step": 449
    },
    {
      "epoch": 3.8461538461538463,
      "grad_norm": 0.6583009362220764,
      "learning_rate": 7.612582568860549e-07,
      "loss": 0.1289,
      "step": 450
    },
    {
      "epoch": 3.8547008547008548,
      "grad_norm": 0.4355236887931824,
      "learning_rate": 6.791167812483012e-07,
      "loss": 0.0742,
      "step": 451
    },
    {
      "epoch": 3.8632478632478633,
      "grad_norm": 0.5870593190193176,
      "learning_rate": 6.016484334238515e-07,
      "loss": 0.1046,
      "step": 452
    },
    {
      "epoch": 3.871794871794872,
      "grad_norm": 0.5397793054580688,
      "learning_rate": 5.288568583610931e-07,
      "loss": 0.0989,
      "step": 453
    },
    {
      "epoch": 3.8803418803418803,
      "grad_norm": 0.4530383348464966,
      "learning_rate": 4.607454809624434e-07,
      "loss": 0.0757,
      "step": 454
    },
    {
      "epoch": 3.888888888888889,
      "grad_norm": 0.5205908417701721,
      "learning_rate": 3.9731750592325587e-07,
      "loss": 0.0908,
      "step": 455
    },
    {
      "epoch": 3.8974358974358974,
      "grad_norm": 0.5037741661071777,
      "learning_rate": 3.385759175809966e-07,
      "loss": 0.0862,
      "step": 456
    },
    {
      "epoch": 3.905982905982906,
      "grad_norm": 0.5087939500808716,
      "learning_rate": 2.845234797748897e-07,
      "loss": 0.091,
      "step": 457
    },
    {
      "epoch": 3.9145299145299144,
      "grad_norm": 0.5621718764305115,
      "learning_rate": 2.3516273571577708e-07,
      "loss": 0.0813,
      "step": 458
    },
    {
      "epoch": 3.9230769230769234,
      "grad_norm": 0.5326392650604248,
      "learning_rate": 1.9049600786658073e-07,
      "loss": 0.0834,
      "step": 459
    },
    {
      "epoch": 3.931623931623932,
      "grad_norm": 0.4818204939365387,
      "learning_rate": 1.505253978329235e-07,
      "loss": 0.0851,
      "step": 460
    },
    {
      "epoch": 3.9401709401709404,
      "grad_norm": 0.581664502620697,
      "learning_rate": 1.1525278626431934e-07,
      "loss": 0.0763,
      "step": 461
    },
    {
      "epoch": 3.948717948717949,
      "grad_norm": 0.3236628770828247,
      "learning_rate": 8.467983276563284e-08,
      "loss": 0.0613,
      "step": 462
    },
    {
      "epoch": 3.9572649572649574,
      "grad_norm": 0.5696295499801636,
      "learning_rate": 5.880797581904185e-08,
      "loss": 0.0902,
      "step": 463
    },
    {
      "epoch": 3.965811965811966,
      "grad_norm": 0.6869117021560669,
      "learning_rate": 3.763843271631373e-08,
      "loss": 0.1285,
      "step": 464
    },
    {
      "epoch": 3.9743589743589745,
      "grad_norm": 0.6082459688186646,
      "learning_rate": 2.1172199501573455e-08,
      "loss": 0.0887,
      "step": 465
    },
    {
      "epoch": 3.982905982905983,
      "grad_norm": 0.6671838164329529,
      "learning_rate": 9.410050924374415e-09,
      "loss": 0.1125,
      "step": 466
    },
    {
      "epoch": 3.9914529914529915,
      "grad_norm": 0.661057710647583,
      "learning_rate": 2.3525404033275523e-09,
      "loss": 0.1273,
      "step": 467
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.5786474943161011,
      "learning_rate": 0.0,
      "loss": 0.0884,
      "step": 468
    }
  ],
  "logging_steps": 1,
  "max_steps": 468,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 117,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.7404232262588826e+18,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}