{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.22180326050792948,
  "eval_steps": 375,
  "global_step": 750,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0002957376806772393,
      "grad_norm": 5.918768405914307,
      "learning_rate": 2e-05,
      "loss": 1.3931,
      "step": 1
    },
    {
      "epoch": 0.0005914753613544786,
      "grad_norm": 6.164484024047852,
      "learning_rate": 4e-05,
      "loss": 1.5398,
      "step": 2
    },
    {
      "epoch": 0.0008872130420317178,
      "grad_norm": 5.500333309173584,
      "learning_rate": 6e-05,
      "loss": 1.4837,
      "step": 3
    },
    {
      "epoch": 0.0011829507227089572,
      "grad_norm": 6.977994441986084,
      "learning_rate": 8e-05,
      "loss": 1.5093,
      "step": 4
    },
    {
      "epoch": 0.0014786884033861965,
      "grad_norm": 10.189373016357422,
      "learning_rate": 0.0001,
      "loss": 1.3296,
      "step": 5
    },
    {
      "epoch": 0.0017744260840634356,
      "grad_norm": 8.993965148925781,
      "learning_rate": 0.00012,
      "loss": 0.9724,
      "step": 6
    },
    {
      "epoch": 0.002070163764740675,
      "grad_norm": 2.4879002571105957,
      "learning_rate": 0.00014,
      "loss": 0.6618,
      "step": 7
    },
    {
      "epoch": 0.0023659014454179145,
      "grad_norm": 3.1456634998321533,
      "learning_rate": 0.00016,
      "loss": 0.5917,
      "step": 8
    },
    {
      "epoch": 0.0026616391260951538,
      "grad_norm": 2.20988392829895,
      "learning_rate": 0.00018,
      "loss": 0.5093,
      "step": 9
    },
    {
      "epoch": 0.002957376806772393,
      "grad_norm": 1.189419150352478,
      "learning_rate": 0.0002,
      "loss": 0.5016,
      "step": 10
    },
    {
      "epoch": 0.003253114487449632,
      "grad_norm": 0.7363479733467102,
      "learning_rate": 0.00019999977772170748,
      "loss": 0.4919,
      "step": 11
    },
    {
      "epoch": 0.0035488521681268713,
      "grad_norm": 0.8657339215278625,
      "learning_rate": 0.00019999911088781805,
      "loss": 0.5377,
      "step": 12
    },
    {
      "epoch": 0.0038445898488041106,
      "grad_norm": 0.5705428123474121,
      "learning_rate": 0.0001999979995012962,
      "loss": 0.5056,
      "step": 13
    },
    {
      "epoch": 0.00414032752948135,
      "grad_norm": 0.6129299402236938,
      "learning_rate": 0.00019999644356708261,
      "loss": 0.4554,
      "step": 14
    },
    {
      "epoch": 0.00443606521015859,
      "grad_norm": 0.4173682928085327,
      "learning_rate": 0.00019999444309209432,
      "loss": 0.3437,
      "step": 15
    },
    {
      "epoch": 0.004731802890835829,
      "grad_norm": 0.39718082547187805,
      "learning_rate": 0.0001999919980852246,
      "loss": 0.5607,
      "step": 16
    },
    {
      "epoch": 0.005027540571513068,
      "grad_norm": 0.2396816462278366,
      "learning_rate": 0.00019998910855734288,
      "loss": 0.4337,
      "step": 17
    },
    {
      "epoch": 0.0053232782521903076,
      "grad_norm": 0.36948394775390625,
      "learning_rate": 0.0001999857745212947,
      "loss": 0.4386,
      "step": 18
    },
    {
      "epoch": 0.005619015932867547,
      "grad_norm": 0.4336908161640167,
      "learning_rate": 0.00019998199599190178,
      "loss": 0.4447,
      "step": 19
    },
    {
      "epoch": 0.005914753613544786,
      "grad_norm": 0.3717949390411377,
      "learning_rate": 0.0001999777729859618,
      "loss": 0.3468,
      "step": 20
    },
    {
      "epoch": 0.0062104912942220255,
      "grad_norm": 0.2210528403520584,
      "learning_rate": 0.00019997310552224846,
      "loss": 0.3254,
      "step": 21
    },
    {
      "epoch": 0.006506228974899264,
      "grad_norm": 0.373976469039917,
      "learning_rate": 0.00019996799362151122,
      "loss": 0.3819,
      "step": 22
    },
    {
      "epoch": 0.006801966655576503,
      "grad_norm": 0.4056873917579651,
      "learning_rate": 0.00019996243730647538,
      "loss": 0.395,
      "step": 23
    },
    {
      "epoch": 0.0070977043362537426,
      "grad_norm": 0.2301822304725647,
      "learning_rate": 0.00019995643660184191,
      "loss": 0.4185,
      "step": 24
    },
    {
      "epoch": 0.007393442016930982,
      "grad_norm": 0.22644546627998352,
      "learning_rate": 0.00019994999153428737,
      "loss": 0.3096,
      "step": 25
    },
    {
      "epoch": 0.007689179697608221,
      "grad_norm": 0.3260557949542999,
      "learning_rate": 0.00019994310213246368,
      "loss": 0.3562,
      "step": 26
    },
    {
      "epoch": 0.007984917378285461,
      "grad_norm": 0.2493218034505844,
      "learning_rate": 0.00019993576842699816,
      "loss": 0.4506,
      "step": 27
    },
    {
      "epoch": 0.0082806550589627,
      "grad_norm": 0.21800749003887177,
      "learning_rate": 0.0001999279904504933,
      "loss": 0.3406,
      "step": 28
    },
    {
      "epoch": 0.00857639273963994,
      "grad_norm": 0.2646995186805725,
      "learning_rate": 0.00019991976823752653,
      "loss": 0.3676,
      "step": 29
    },
    {
      "epoch": 0.00887213042031718,
      "grad_norm": 0.2616923749446869,
      "learning_rate": 0.00019991110182465032,
      "loss": 0.3685,
      "step": 30
    },
    {
      "epoch": 0.009167868100994419,
      "grad_norm": 0.2577691078186035,
      "learning_rate": 0.00019990199125039174,
      "loss": 0.3006,
      "step": 31
    },
    {
      "epoch": 0.009463605781671658,
      "grad_norm": 0.26562878489494324,
      "learning_rate": 0.00019989243655525247,
      "loss": 0.3532,
      "step": 32
    },
    {
      "epoch": 0.009759343462348897,
      "grad_norm": 0.26207393407821655,
      "learning_rate": 0.00019988243778170853,
      "loss": 0.3386,
      "step": 33
    },
    {
      "epoch": 0.010055081143026137,
      "grad_norm": 0.2501125931739807,
      "learning_rate": 0.0001998719949742101,
      "loss": 0.2988,
      "step": 34
    },
    {
      "epoch": 0.010350818823703376,
      "grad_norm": 0.3094285726547241,
      "learning_rate": 0.0001998611081791814,
      "loss": 0.3979,
      "step": 35
    },
    {
      "epoch": 0.010646556504380615,
      "grad_norm": 0.2816741466522217,
      "learning_rate": 0.00019984977744502038,
      "loss": 0.3338,
      "step": 36
    },
    {
      "epoch": 0.010942294185057854,
      "grad_norm": 0.33405667543411255,
      "learning_rate": 0.00019983800282209857,
      "loss": 0.4424,
      "step": 37
    },
    {
      "epoch": 0.011238031865735094,
      "grad_norm": 0.26445043087005615,
      "learning_rate": 0.00019982578436276082,
      "loss": 0.3255,
      "step": 38
    },
    {
      "epoch": 0.011533769546412333,
      "grad_norm": 0.2719035744667053,
      "learning_rate": 0.00019981312212132512,
      "loss": 0.3665,
      "step": 39
    },
    {
      "epoch": 0.011829507227089572,
      "grad_norm": 0.41192904114723206,
      "learning_rate": 0.00019980001615408228,
      "loss": 0.3978,
      "step": 40
    },
    {
      "epoch": 0.012125244907766812,
      "grad_norm": 0.30053630471229553,
      "learning_rate": 0.00019978646651929572,
      "loss": 0.3416,
      "step": 41
    },
    {
      "epoch": 0.012420982588444051,
      "grad_norm": 0.32080647349357605,
      "learning_rate": 0.00019977247327720128,
      "loss": 0.3517,
      "step": 42
    },
    {
      "epoch": 0.012716720269121289,
      "grad_norm": 0.3234342038631439,
      "learning_rate": 0.0001997580364900068,
      "loss": 0.2878,
      "step": 43
    },
    {
      "epoch": 0.013012457949798528,
      "grad_norm": 0.33844128251075745,
      "learning_rate": 0.000199743156221892,
      "loss": 0.2822,
      "step": 44
    },
    {
      "epoch": 0.013308195630475767,
      "grad_norm": 0.36423033475875854,
      "learning_rate": 0.00019972783253900808,
      "loss": 0.3715,
      "step": 45
    },
    {
      "epoch": 0.013603933311153006,
      "grad_norm": 0.34381452202796936,
      "learning_rate": 0.00019971206550947748,
      "loss": 0.3983,
      "step": 46
    },
    {
      "epoch": 0.013899670991830246,
      "grad_norm": 0.3789922595024109,
      "learning_rate": 0.00019969585520339354,
      "loss": 0.2693,
      "step": 47
    },
    {
      "epoch": 0.014195408672507485,
      "grad_norm": 0.4301562011241913,
      "learning_rate": 0.0001996792016928203,
      "loss": 0.3752,
      "step": 48
    },
    {
      "epoch": 0.014491146353184724,
      "grad_norm": 0.4085782766342163,
      "learning_rate": 0.00019966210505179197,
      "loss": 0.3779,
      "step": 49
    },
    {
      "epoch": 0.014786884033861964,
      "grad_norm": 0.7027013897895813,
      "learning_rate": 0.00019964456535631286,
      "loss": 0.4713,
      "step": 50
    },
    {
      "epoch": 0.015082621714539203,
      "grad_norm": 0.12943406403064728,
      "learning_rate": 0.0001996265826843568,
      "loss": 0.1727,
      "step": 51
    },
    {
      "epoch": 0.015378359395216442,
      "grad_norm": 0.20666781067848206,
      "learning_rate": 0.00019960815711586696,
      "loss": 0.2999,
      "step": 52
    },
    {
      "epoch": 0.01567409707589368,
      "grad_norm": 0.18581970036029816,
      "learning_rate": 0.00019958928873275539,
      "loss": 0.3003,
      "step": 53
    },
    {
      "epoch": 0.015969834756570923,
      "grad_norm": 0.1757449209690094,
      "learning_rate": 0.00019956997761890277,
      "loss": 0.2614,
      "step": 54
    },
    {
      "epoch": 0.01626557243724816,
      "grad_norm": 0.1774774044752121,
      "learning_rate": 0.00019955022386015792,
      "loss": 0.3129,
      "step": 55
    },
    {
      "epoch": 0.0165613101179254,
      "grad_norm": 0.17331115901470184,
      "learning_rate": 0.00019953002754433743,
      "loss": 0.3113,
      "step": 56
    },
    {
      "epoch": 0.01685704779860264,
      "grad_norm": 0.18873649835586548,
      "learning_rate": 0.00019950938876122542,
      "loss": 0.2683,
      "step": 57
    },
    {
      "epoch": 0.01715278547927988,
      "grad_norm": 0.15520928800106049,
      "learning_rate": 0.00019948830760257291,
      "loss": 0.2173,
      "step": 58
    },
    {
      "epoch": 0.017448523159957118,
      "grad_norm": 0.14979037642478943,
      "learning_rate": 0.0001994667841620976,
      "loss": 0.23,
      "step": 59
    },
    {
      "epoch": 0.01774426084063436,
      "grad_norm": 0.15720731019973755,
      "learning_rate": 0.00019944481853548335,
      "loss": 0.2606,
      "step": 60
    },
    {
      "epoch": 0.018039998521311596,
      "grad_norm": 0.18836985528469086,
      "learning_rate": 0.00019942241082037982,
      "loss": 0.34,
      "step": 61
    },
    {
      "epoch": 0.018335736201988837,
      "grad_norm": 0.17339695990085602,
      "learning_rate": 0.00019939956111640197,
      "loss": 0.2927,
      "step": 62
    },
    {
      "epoch": 0.018631473882666075,
      "grad_norm": 0.16999205946922302,
      "learning_rate": 0.00019937626952512964,
      "loss": 0.2591,
      "step": 63
    },
    {
      "epoch": 0.018927211563343316,
      "grad_norm": 0.1601104885339737,
      "learning_rate": 0.0001993525361501072,
      "loss": 0.2929,
      "step": 64
    },
    {
      "epoch": 0.019222949244020553,
      "grad_norm": 0.17350880801677704,
      "learning_rate": 0.00019932836109684286,
      "loss": 0.2854,
      "step": 65
    },
    {
      "epoch": 0.019518686924697794,
      "grad_norm": 0.18091264367103577,
      "learning_rate": 0.00019930374447280845,
      "loss": 0.2839,
      "step": 66
    },
    {
      "epoch": 0.019814424605375032,
      "grad_norm": 0.1761292666196823,
      "learning_rate": 0.00019927868638743875,
      "loss": 0.2545,
      "step": 67
    },
    {
      "epoch": 0.020110162286052273,
      "grad_norm": 0.18896149098873138,
      "learning_rate": 0.0001992531869521312,
      "loss": 0.2782,
      "step": 68
    },
    {
      "epoch": 0.02040589996672951,
      "grad_norm": 0.1839849352836609,
      "learning_rate": 0.00019922724628024515,
      "loss": 0.2596,
      "step": 69
    },
    {
      "epoch": 0.02070163764740675,
      "grad_norm": 0.18426676094532013,
      "learning_rate": 0.0001992008644871016,
      "loss": 0.2921,
      "step": 70
    },
    {
      "epoch": 0.02099737532808399,
      "grad_norm": 0.23805294930934906,
      "learning_rate": 0.00019917404168998256,
      "loss": 0.3308,
      "step": 71
    },
    {
      "epoch": 0.02129311300876123,
      "grad_norm": 0.18368969857692719,
      "learning_rate": 0.0001991467780081305,
      "loss": 0.3328,
      "step": 72
    },
    {
      "epoch": 0.021588850689438468,
      "grad_norm": 0.20601886510849,
      "learning_rate": 0.00019911907356274795,
      "loss": 0.3049,
      "step": 73
    },
    {
      "epoch": 0.02188458837011571,
      "grad_norm": 0.21643660962581635,
      "learning_rate": 0.00019909092847699683,
      "loss": 0.3097,
      "step": 74
    },
    {
      "epoch": 0.022180326050792946,
      "grad_norm": 0.1990053504705429,
      "learning_rate": 0.00019906234287599798,
      "loss": 0.3245,
      "step": 75
    },
    {
      "epoch": 0.022476063731470187,
      "grad_norm": 0.18081244826316833,
      "learning_rate": 0.00019903331688683057,
      "loss": 0.2234,
      "step": 76
    },
    {
      "epoch": 0.022771801412147425,
      "grad_norm": 0.19501319527626038,
      "learning_rate": 0.00019900385063853154,
      "loss": 0.2852,
      "step": 77
    },
    {
      "epoch": 0.023067539092824666,
      "grad_norm": 0.19676417112350464,
      "learning_rate": 0.00019897394426209505,
      "loss": 0.1825,
      "step": 78
    },
    {
      "epoch": 0.023363276773501904,
      "grad_norm": 0.22049355506896973,
      "learning_rate": 0.00019894359789047187,
      "loss": 0.2708,
      "step": 79
    },
    {
      "epoch": 0.023659014454179145,
      "grad_norm": 0.24137771129608154,
      "learning_rate": 0.00019891281165856873,
      "loss": 0.317,
      "step": 80
    },
    {
      "epoch": 0.023954752134856382,
      "grad_norm": 0.2464693933725357,
      "learning_rate": 0.00019888158570324795,
      "loss": 0.281,
      "step": 81
    },
    {
      "epoch": 0.024250489815533623,
      "grad_norm": 0.2559899091720581,
      "learning_rate": 0.0001988499201633265,
      "loss": 0.2977,
      "step": 82
    },
    {
      "epoch": 0.02454622749621086,
      "grad_norm": 0.23781578242778778,
      "learning_rate": 0.00019881781517957562,
      "loss": 0.2156,
      "step": 83
    },
    {
      "epoch": 0.024841965176888102,
      "grad_norm": 0.25769996643066406,
      "learning_rate": 0.0001987852708947202,
      "loss": 0.2588,
      "step": 84
    },
    {
      "epoch": 0.02513770285756534,
      "grad_norm": 0.24173754453659058,
      "learning_rate": 0.00019875228745343794,
      "loss": 0.2689,
      "step": 85
    },
    {
      "epoch": 0.025433440538242577,
      "grad_norm": 0.23814626038074493,
      "learning_rate": 0.0001987188650023589,
      "loss": 0.2404,
      "step": 86
    },
    {
      "epoch": 0.025729178218919818,
      "grad_norm": 0.2124136984348297,
      "learning_rate": 0.0001986850036900648,
      "loss": 0.2472,
      "step": 87
    },
    {
      "epoch": 0.026024915899597056,
      "grad_norm": 0.24799956381320953,
      "learning_rate": 0.00019865070366708836,
      "loss": 0.2429,
      "step": 88
    },
    {
      "epoch": 0.026320653580274297,
      "grad_norm": 0.2826499938964844,
      "learning_rate": 0.00019861596508591255,
      "loss": 0.2906,
      "step": 89
    },
    {
      "epoch": 0.026616391260951534,
      "grad_norm": 0.25857439637184143,
      "learning_rate": 0.00019858078810097002,
      "loss": 0.2139,
      "step": 90
    },
    {
      "epoch": 0.026912128941628775,
      "grad_norm": 0.25622865557670593,
      "learning_rate": 0.00019854517286864245,
      "loss": 0.266,
      "step": 91
    },
    {
      "epoch": 0.027207866622306013,
      "grad_norm": 0.2770698368549347,
      "learning_rate": 0.0001985091195472596,
      "loss": 0.254,
      "step": 92
    },
    {
      "epoch": 0.027503604302983254,
      "grad_norm": 0.31867358088493347,
      "learning_rate": 0.0001984726282970989,
      "loss": 0.2851,
      "step": 93
    },
    {
      "epoch": 0.02779934198366049,
      "grad_norm": 0.35994890332221985,
      "learning_rate": 0.0001984356992803847,
      "loss": 0.321,
      "step": 94
    },
    {
      "epoch": 0.028095079664337733,
      "grad_norm": 0.337107390165329,
      "learning_rate": 0.00019839833266128724,
      "loss": 0.2248,
      "step": 95
    },
    {
      "epoch": 0.02839081734501497,
      "grad_norm": 0.34881696105003357,
      "learning_rate": 0.00019836052860592237,
      "loss": 0.2786,
      "step": 96
    },
    {
      "epoch": 0.02868655502569221,
      "grad_norm": 0.3969687521457672,
      "learning_rate": 0.0001983222872823505,
      "loss": 0.3405,
      "step": 97
    },
    {
      "epoch": 0.02898229270636945,
      "grad_norm": 0.38202965259552,
      "learning_rate": 0.00019828360886057594,
      "loss": 0.275,
      "step": 98
    },
    {
      "epoch": 0.02927803038704669,
      "grad_norm": 0.4236074686050415,
      "learning_rate": 0.00019824449351254616,
      "loss": 0.2973,
      "step": 99
    },
    {
      "epoch": 0.029573768067723927,
      "grad_norm": 0.6627792716026306,
      "learning_rate": 0.00019820494141215104,
      "loss": 0.3571,
      "step": 100
    },
    {
      "epoch": 0.02986950574840117,
      "grad_norm": 0.14578954875469208,
      "learning_rate": 0.000198164952735222,
      "loss": 0.1857,
      "step": 101
    },
    {
      "epoch": 0.030165243429078406,
      "grad_norm": 0.1730443388223648,
      "learning_rate": 0.00019812452765953135,
      "loss": 0.2189,
      "step": 102
    },
    {
      "epoch": 0.030460981109755647,
      "grad_norm": 0.20889230072498322,
      "learning_rate": 0.00019808366636479147,
      "loss": 0.264,
      "step": 103
    },
    {
      "epoch": 0.030756718790432885,
      "grad_norm": 0.19660377502441406,
      "learning_rate": 0.00019804236903265388,
      "loss": 0.3058,
      "step": 104
    },
    {
      "epoch": 0.031052456471110126,
      "grad_norm": 0.16153180599212646,
      "learning_rate": 0.00019800063584670863,
      "loss": 0.2465,
      "step": 105
    },
    {
      "epoch": 0.03134819415178736,
      "grad_norm": 0.17819426953792572,
      "learning_rate": 0.00019795846699248332,
      "loss": 0.2406,
      "step": 106
    },
    {
      "epoch": 0.031643931832464604,
      "grad_norm": 0.16484011709690094,
      "learning_rate": 0.00019791586265744237,
      "loss": 0.211,
      "step": 107
    },
    {
      "epoch": 0.031939669513141845,
      "grad_norm": 0.19172953069210052,
      "learning_rate": 0.00019787282303098617,
      "loss": 0.2686,
      "step": 108
    },
    {
      "epoch": 0.03223540719381908,
      "grad_norm": 0.17775307595729828,
      "learning_rate": 0.0001978293483044502,
      "loss": 0.2464,
      "step": 109
    },
    {
      "epoch": 0.03253114487449632,
      "grad_norm": 0.17548003792762756,
      "learning_rate": 0.00019778543867110426,
      "loss": 0.2489,
      "step": 110
    },
    {
      "epoch": 0.03282688255517356,
      "grad_norm": 0.17837557196617126,
      "learning_rate": 0.00019774109432615147,
      "loss": 0.2305,
      "step": 111
    },
    {
      "epoch": 0.0331226202358508,
      "grad_norm": 0.21005579829216003,
      "learning_rate": 0.00019769631546672756,
      "loss": 0.3109,
      "step": 112
    },
    {
      "epoch": 0.03341835791652804,
      "grad_norm": 0.1606261134147644,
      "learning_rate": 0.00019765110229189988,
      "loss": 0.2297,
      "step": 113
    },
    {
      "epoch": 0.03371409559720528,
      "grad_norm": 0.16127589344978333,
      "learning_rate": 0.00019760545500266657,
      "loss": 0.2121,
      "step": 114
    },
    {
      "epoch": 0.03400983327788252,
      "grad_norm": 0.18138477206230164,
      "learning_rate": 0.00019755937380195568,
      "loss": 0.2556,
      "step": 115
    },
    {
      "epoch": 0.03430557095855976,
      "grad_norm": 0.19160465896129608,
      "learning_rate": 0.00019751285889462423,
      "loss": 0.2534,
      "step": 116
    },
    {
      "epoch": 0.034601308639236994,
      "grad_norm": 0.19009210169315338,
      "learning_rate": 0.0001974659104874573,
      "loss": 0.2289,
      "step": 117
    },
    {
      "epoch": 0.034897046319914235,
      "grad_norm": 0.1938568651676178,
      "learning_rate": 0.0001974185287891671,
      "loss": 0.2483,
      "step": 118
    },
    {
      "epoch": 0.035192784000591476,
      "grad_norm": 0.19120760262012482,
      "learning_rate": 0.0001973707140103921,
      "loss": 0.2522,
      "step": 119
    },
    {
      "epoch": 0.03548852168126872,
      "grad_norm": 0.20597144961357117,
      "learning_rate": 0.00019732246636369605,
      "loss": 0.2776,
      "step": 120
    },
    {
      "epoch": 0.03578425936194595,
      "grad_norm": 0.19169169664382935,
      "learning_rate": 0.00019727378606356703,
      "loss": 0.2821,
      "step": 121
    },
    {
      "epoch": 0.03607999704262319,
      "grad_norm": 0.24905505776405334,
      "learning_rate": 0.00019722467332641656,
      "loss": 0.3392,
      "step": 122
    },
    {
      "epoch": 0.03637573472330043,
      "grad_norm": 0.21315141022205353,
      "learning_rate": 0.00019717512837057855,
      "loss": 0.2514,
      "step": 123
    },
    {
      "epoch": 0.036671472403977674,
      "grad_norm": 0.20183053612709045,
      "learning_rate": 0.0001971251514163083,
      "loss": 0.2522,
      "step": 124
    },
    {
      "epoch": 0.03696721008465491,
      "grad_norm": 0.21825973689556122,
      "learning_rate": 0.0001970747426857817,
      "loss": 0.2249,
      "step": 125
    },
    {
      "epoch": 0.03726294776533215,
      "grad_norm": 0.19961951673030853,
      "learning_rate": 0.00019702390240309404,
      "loss": 0.2081,
      "step": 126
    },
    {
      "epoch": 0.03755868544600939,
      "grad_norm": 0.24127651751041412,
      "learning_rate": 0.0001969726307942592,
      "loss": 0.2543,
      "step": 127
    },
    {
      "epoch": 0.03785442312668663,
      "grad_norm": 0.18847639858722687,
      "learning_rate": 0.00019692092808720846,
      "loss": 0.1804,
      "step": 128
    },
    {
      "epoch": 0.038150160807363866,
      "grad_norm": 0.2245255559682846,
      "learning_rate": 0.0001968687945117896,
      "loss": 0.2615,
      "step": 129
    },
    {
      "epoch": 0.03844589848804111,
      "grad_norm": 0.2330857813358307,
      "learning_rate": 0.00019681623029976588,
      "loss": 0.24,
      "step": 130
    },
    {
      "epoch": 0.03874163616871835,
      "grad_norm": 0.2308180183172226,
      "learning_rate": 0.00019676323568481498,
      "loss": 0.2735,
      "step": 131
    },
    {
      "epoch": 0.03903737384939559,
      "grad_norm": 0.23675177991390228,
      "learning_rate": 0.00019670981090252792,
      "loss": 0.273,
      "step": 132
    },
    {
      "epoch": 0.03933311153007282,
      "grad_norm": 0.2156166285276413,
      "learning_rate": 0.00019665595619040808,
      "loss": 0.2038,
      "step": 133
    },
    {
      "epoch": 0.039628849210750064,
      "grad_norm": 0.22249513864517212,
      "learning_rate": 0.0001966016717878702,
      "loss": 0.2396,
      "step": 134
    },
    {
      "epoch": 0.039924586891427305,
      "grad_norm": 0.25741225481033325,
      "learning_rate": 0.00019654695793623907,
      "loss": 0.2921,
      "step": 135
    },
    {
      "epoch": 0.040220324572104546,
      "grad_norm": 0.25070619583129883,
      "learning_rate": 0.0001964918148787488,
      "loss": 0.2107,
      "step": 136
    },
    {
      "epoch": 0.04051606225278178,
      "grad_norm": 0.2761012017726898,
      "learning_rate": 0.00019643624286054144,
      "loss": 0.2443,
      "step": 137
    },
    {
      "epoch": 0.04081179993345902,
      "grad_norm": 0.22816140949726105,
      "learning_rate": 0.00019638024212866606,
      "loss": 0.1939,
      "step": 138
    },
    {
      "epoch": 0.04110753761413626,
      "grad_norm": 0.2465653121471405,
      "learning_rate": 0.0001963238129320776,
      "loss": 0.2535,
      "step": 139
    },
    {
      "epoch": 0.0414032752948135,
      "grad_norm": 0.38914790749549866,
      "learning_rate": 0.00019626695552163578,
      "loss": 0.3959,
      "step": 140
    },
    {
      "epoch": 0.04169901297549074,
      "grad_norm": 0.2719482481479645,
      "learning_rate": 0.00019620967015010395,
      "loss": 0.273,
      "step": 141
    },
    {
      "epoch": 0.04199475065616798,
      "grad_norm": 0.26919126510620117,
      "learning_rate": 0.00019615195707214803,
      "loss": 0.2197,
      "step": 142
    },
    {
      "epoch": 0.04229048833684522,
      "grad_norm": 0.2746446132659912,
      "learning_rate": 0.0001960938165443353,
      "loss": 0.251,
      "step": 143
    },
    {
      "epoch": 0.04258622601752246,
      "grad_norm": 0.31786006689071655,
      "learning_rate": 0.00019603524882513327,
      "loss": 0.2585,
      "step": 144
    },
    {
      "epoch": 0.042881963698199695,
      "grad_norm": 0.3586949110031128,
      "learning_rate": 0.0001959762541749086,
      "loss": 0.2868,
      "step": 145
    },
    {
      "epoch": 0.043177701378876936,
      "grad_norm": 0.3204839527606964,
      "learning_rate": 0.00019591683285592593,
      "loss": 0.2449,
      "step": 146
    },
    {
      "epoch": 0.04347343905955418,
      "grad_norm": 0.3199440836906433,
      "learning_rate": 0.00019585698513234663,
      "loss": 0.238,
      "step": 147
    },
    {
      "epoch": 0.04376917674023142,
      "grad_norm": 0.460660457611084,
      "learning_rate": 0.0001957967112702277,
      "loss": 0.2707,
      "step": 148
    },
    {
      "epoch": 0.04406491442090865,
      "grad_norm": 0.3981362581253052,
      "learning_rate": 0.00019573601153752052,
      "loss": 0.2259,
      "step": 149
    },
    {
      "epoch": 0.04436065210158589,
      "grad_norm": 0.5862023234367371,
      "learning_rate": 0.00019567488620406983,
      "loss": 0.3031,
      "step": 150
    },
    {
      "epoch": 0.044656389782263134,
      "grad_norm": 0.14805294573307037,
      "learning_rate": 0.00019561333554161224,
      "loss": 0.1515,
      "step": 151
    },
    {
      "epoch": 0.044952127462940375,
      "grad_norm": 0.2211844027042389,
      "learning_rate": 0.0001955513598237753,
      "loss": 0.2199,
      "step": 152
    },
    {
      "epoch": 0.04524786514361761,
      "grad_norm": 0.20431561768054962,
      "learning_rate": 0.00019548895932607621,
      "loss": 0.2077,
      "step": 153
    },
    {
      "epoch": 0.04554360282429485,
      "grad_norm": 0.1585177332162857,
      "learning_rate": 0.00019542613432592038,
      "loss": 0.1702,
      "step": 154
    },
    {
      "epoch": 0.04583934050497209,
      "grad_norm": 0.21959953010082245,
      "learning_rate": 0.00019536288510260056,
      "loss": 0.3018,
      "step": 155
    },
    {
      "epoch": 0.04613507818564933,
      "grad_norm": 0.18285472691059113,
      "learning_rate": 0.00019529921193729534,
      "loss": 0.2621,
      "step": 156
    },
    {
      "epoch": 0.046430815866326566,
      "grad_norm": 0.19817829132080078,
      "learning_rate": 0.00019523511511306793,
      "loss": 0.2542,
      "step": 157
    },
    {
      "epoch": 0.04672655354700381,
      "grad_norm": 0.2109287828207016,
      "learning_rate": 0.000195170594914865,
      "loss": 0.2398,
      "step": 158
    },
    {
      "epoch": 0.04702229122768105,
      "grad_norm": 0.20559227466583252,
      "learning_rate": 0.00019510565162951537,
      "loss": 0.2302,
      "step": 159
    },
    {
      "epoch": 0.04731802890835829,
      "grad_norm": 0.2374628186225891,
      "learning_rate": 0.00019504028554572864,
      "loss": 0.2729,
      "step": 160
    },
    {
      "epoch": 0.047613766589035524,
      "grad_norm": 0.17768986523151398,
      "learning_rate": 0.00019497449695409408,
      "loss": 0.2692,
      "step": 161
    },
    {
      "epoch": 0.047909504269712765,
      "grad_norm": 0.1690852791070938,
      "learning_rate": 0.00019490828614707916,
      "loss": 0.2147,
      "step": 162
    },
    {
      "epoch": 0.048205241950390006,
      "grad_norm": 0.20444796979427338,
      "learning_rate": 0.00019484165341902845,
      "loss": 0.2317,
      "step": 163
    },
    {
      "epoch": 0.04850097963106725,
      "grad_norm": 0.2058713436126709,
      "learning_rate": 0.00019477459906616206,
      "loss": 0.1968,
      "step": 164
    },
    {
      "epoch": 0.04879671731174448,
      "grad_norm": 0.20283836126327515,
      "learning_rate": 0.00019470712338657458,
      "loss": 0.2275,
      "step": 165
    },
    {
      "epoch": 0.04909245499242172,
      "grad_norm": 0.18474236130714417,
      "learning_rate": 0.0001946392266802336,
      "loss": 0.2602,
      "step": 166
    },
    {
      "epoch": 0.04938819267309896,
      "grad_norm": 0.17107436060905457,
      "learning_rate": 0.0001945709092489783,
      "loss": 0.2182,
      "step": 167
    },
    {
      "epoch": 0.049683930353776204,
      "grad_norm": 0.18627148866653442,
      "learning_rate": 0.00019450217139651844,
      "loss": 0.1782,
      "step": 168
    },
    {
      "epoch": 0.04997966803445344,
      "grad_norm": 0.19079387187957764,
      "learning_rate": 0.0001944330134284326,
      "loss": 0.2239,
      "step": 169
    },
    {
      "epoch": 0.05027540571513068,
      "grad_norm": 0.22388561069965363,
      "learning_rate": 0.00019436343565216711,
      "loss": 0.2257,
      "step": 170
    },
    {
      "epoch": 0.05057114339580792,
      "grad_norm": 0.20720115303993225,
      "learning_rate": 0.00019429343837703455,
      "loss": 0.2546,
      "step": 171
    },
    {
      "epoch": 0.050866881076485154,
      "grad_norm": 0.19800806045532227,
      "learning_rate": 0.0001942230219142124,
      "loss": 0.1958,
      "step": 172
    },
    {
      "epoch": 0.051162618757162395,
      "grad_norm": 0.22758133709430695,
      "learning_rate": 0.0001941521865767417,
      "loss": 0.2499,
      "step": 173
    },
    {
      "epoch": 0.051458356437839636,
      "grad_norm": 0.20841172337532043,
      "learning_rate": 0.0001940809326795256,
      "loss": 0.2369,
      "step": 174
    },
    {
      "epoch": 0.05175409411851688,
      "grad_norm": 0.20555779337882996,
      "learning_rate": 0.000194009260539328,
      "loss": 0.2435,
      "step": 175
    },
    {
      "epoch": 0.05204983179919411,
      "grad_norm": 0.2264283448457718,
      "learning_rate": 0.0001939371704747721,
      "loss": 0.2603,
      "step": 176
    },
    {
      "epoch": 0.05234556947987135,
      "grad_norm": 0.22826440632343292,
      "learning_rate": 0.00019386466280633906,
      "loss": 0.2574,
      "step": 177
    },
    {
      "epoch": 0.052641307160548594,
      "grad_norm": 0.2327851504087448,
      "learning_rate": 0.00019379173785636646,
      "loss": 0.2125,
      "step": 178
    },
    {
      "epoch": 0.052937044841225835,
      "grad_norm": 0.25769078731536865,
      "learning_rate": 0.000193718395949047,
      "loss": 0.2594,
      "step": 179
    },
    {
      "epoch": 0.05323278252190307,
      "grad_norm": 0.24143129587173462,
      "learning_rate": 0.00019364463741042694,
      "loss": 0.2285,
      "step": 180
    },
    {
      "epoch": 0.05352852020258031,
      "grad_norm": 0.21483063697814941,
      "learning_rate": 0.00019357046256840473,
      "loss": 0.1883,
      "step": 181
    },
    {
      "epoch": 0.05382425788325755,
      "grad_norm": 0.2387738972902298,
      "learning_rate": 0.00019349587175272948,
      "loss": 0.2697,
      "step": 182
    },
    {
      "epoch": 0.05411999556393479,
      "grad_norm": 0.2653301954269409,
      "learning_rate": 0.0001934208652949996,
      "loss": 0.2403,
      "step": 183
    },
    {
      "epoch": 0.054415733244612026,
      "grad_norm": 0.28791022300720215,
      "learning_rate": 0.00019334544352866127,
      "loss": 0.268,
      "step": 184
    },
    {
      "epoch": 0.05471147092528927,
      "grad_norm": 0.2307739108800888,
      "learning_rate": 0.00019326960678900688,
      "loss": 0.2088,
      "step": 185
    },
    {
      "epoch": 0.05500720860596651,
      "grad_norm": 0.2275734543800354,
      "learning_rate": 0.00019319335541317361,
      "loss": 0.1935,
      "step": 186
    },
    {
      "epoch": 0.05530294628664375,
      "grad_norm": 0.23581242561340332,
      "learning_rate": 0.00019311668974014208,
      "loss": 0.2037,
      "step": 187
    },
    {
      "epoch": 0.05559868396732098,
      "grad_norm": 0.25955161452293396,
      "learning_rate": 0.00019303961011073447,
      "loss": 0.2216,
      "step": 188
    },
    {
      "epoch": 0.055894421647998224,
      "grad_norm": 0.262517511844635,
      "learning_rate": 0.00019296211686761346,
      "loss": 0.2143,
      "step": 189
    },
    {
      "epoch": 0.056190159328675465,
      "grad_norm": 0.3077569305896759,
      "learning_rate": 0.00019288421035528028,
      "loss": 0.2356,
      "step": 190
    },
    {
      "epoch": 0.056485897009352706,
      "grad_norm": 0.26791897416114807,
      "learning_rate": 0.00019280589092007352,
      "loss": 0.1884,
      "step": 191
    },
    {
      "epoch": 0.05678163469002994,
      "grad_norm": 0.27822402119636536,
      "learning_rate": 0.00019272715891016735,
      "loss": 0.2118,
      "step": 192
    },
    {
      "epoch": 0.05707737237070718,
      "grad_norm": 0.2592899203300476,
      "learning_rate": 0.00019264801467557007,
      "loss": 0.1402,
      "step": 193
    },
    {
      "epoch": 0.05737311005138442,
      "grad_norm": 0.37037453055381775,
      "learning_rate": 0.00019256845856812266,
      "loss": 0.3472,
      "step": 194
    },
    {
      "epoch": 0.057668847732061664,
      "grad_norm": 0.30506840348243713,
      "learning_rate": 0.000192488490941497,
      "loss": 0.1907,
      "step": 195
    },
    {
      "epoch": 0.0579645854127389,
      "grad_norm": 0.3820033669471741,
      "learning_rate": 0.00019240811215119448,
      "loss": 0.2557,
      "step": 196
    },
    {
      "epoch": 0.05826032309341614,
      "grad_norm": 0.40331634879112244,
      "learning_rate": 0.00019232732255454422,
      "loss": 0.2579,
      "step": 197
    },
    {
      "epoch": 0.05855606077409338,
      "grad_norm": 0.4118432104587555,
      "learning_rate": 0.00019224612251070175,
      "loss": 0.2316,
      "step": 198
    },
    {
      "epoch": 0.05885179845477062,
      "grad_norm": 0.4977587163448334,
      "learning_rate": 0.0001921645123806472,
      "loss": 0.2964,
      "step": 199
    },
    {
      "epoch": 0.059147536135447855,
      "grad_norm": 0.5816993713378906,
      "learning_rate": 0.0001920824925271838,
      "loss": 0.2308,
      "step": 200
    },
    {
      "epoch": 0.059443273816125096,
      "grad_norm": 0.1586022526025772,
      "learning_rate": 0.0001920000633149362,
      "loss": 0.1695,
      "step": 201
    },
    {
      "epoch": 0.05973901149680234,
      "grad_norm": 0.17268988490104675,
      "learning_rate": 0.00019191722511034884,
      "loss": 0.1737,
      "step": 202
    },
    {
      "epoch": 0.06003474917747958,
      "grad_norm": 0.17306619882583618,
      "learning_rate": 0.00019183397828168448,
      "loss": 0.1928,
      "step": 203
    },
    {
      "epoch": 0.06033048685815681,
      "grad_norm": 0.18508224189281464,
      "learning_rate": 0.00019175032319902234,
      "loss": 0.2293,
      "step": 204
    },
    {
      "epoch": 0.06062622453883405,
      "grad_norm": 0.15483330190181732,
      "learning_rate": 0.00019166626023425662,
      "loss": 0.1827,
      "step": 205
    },
    {
      "epoch": 0.060921962219511294,
      "grad_norm": 0.190659299492836,
      "learning_rate": 0.00019158178976109476,
      "loss": 0.2827,
      "step": 206
    },
    {
      "epoch": 0.061217699900188535,
      "grad_norm": 0.16438104212284088,
      "learning_rate": 0.0001914969121550558,
      "loss": 0.2186,
      "step": 207
    },
    {
      "epoch": 0.06151343758086577,
      "grad_norm": 0.17598582804203033,
      "learning_rate": 0.00019141162779346874,
      "loss": 0.1971,
      "step": 208
    },
    {
      "epoch": 0.06180917526154301,
      "grad_norm": 0.18765893578529358,
      "learning_rate": 0.00019132593705547082,
      "loss": 0.1996,
      "step": 209
    },
    {
      "epoch": 0.06210491294222025,
      "grad_norm": 0.21454760432243347,
      "learning_rate": 0.00019123984032200586,
      "loss": 0.2427,
      "step": 210
    },
    {
      "epoch": 0.06240065062289749,
      "grad_norm": 0.21339420974254608,
      "learning_rate": 0.00019115333797582254,
      "loss": 0.2539,
      "step": 211
    },
    {
      "epoch": 0.06269638830357473,
      "grad_norm": 0.18333065509796143,
      "learning_rate": 0.00019106643040147278,
      "loss": 0.2447,
      "step": 212
    },
    {
      "epoch": 0.06299212598425197,
      "grad_norm": 0.19549886882305145,
      "learning_rate": 0.00019097911798530987,
      "loss": 0.2451,
      "step": 213
    },
    {
      "epoch": 0.06328786366492921,
      "grad_norm": 0.19182579219341278,
      "learning_rate": 0.00019089140111548696,
      "loss": 0.2058,
      "step": 214
    },
    {
      "epoch": 0.06358360134560645,
      "grad_norm": 0.19501961767673492,
      "learning_rate": 0.00019080328018195513,
      "loss": 0.2365,
      "step": 215
    },
    {
      "epoch": 0.06387933902628369,
      "grad_norm": 0.19913062453269958,
      "learning_rate": 0.0001907147555764618,
      "loss": 0.2204,
      "step": 216
    },
    {
      "epoch": 0.06417507670696093,
      "grad_norm": 0.1938629001379013,
      "learning_rate": 0.00019062582769254895,
      "loss": 0.2554,
      "step": 217
    },
    {
      "epoch": 0.06447081438763816,
      "grad_norm": 0.20500774681568146,
      "learning_rate": 0.00019053649692555135,
      "loss": 0.2719,
      "step": 218
    },
    {
      "epoch": 0.0647665520683154,
      "grad_norm": 0.19601133465766907,
      "learning_rate": 0.00019044676367259476,
      "loss": 0.2176,
      "step": 219
    },
    {
      "epoch": 0.06506228974899264,
      "grad_norm": 0.21002735197544098,
      "learning_rate": 0.00019035662833259432,
      "loss": 0.2987,
      "step": 220
    },
    {
      "epoch": 0.06535802742966988,
      "grad_norm": 0.19716577231884003,
      "learning_rate": 0.00019026609130625257,
      "loss": 0.2103,
      "step": 221
    },
    {
      "epoch": 0.06565376511034712,
      "grad_norm": 0.20321358740329742,
      "learning_rate": 0.00019017515299605788,
      "loss": 0.2373,
      "step": 222
    },
    {
      "epoch": 0.06594950279102436,
      "grad_norm": 0.21129213273525238,
      "learning_rate": 0.00019008381380628247,
      "loss": 0.2284,
      "step": 223
    },
    {
      "epoch": 0.0662452404717016,
      "grad_norm": 0.20547550916671753,
      "learning_rate": 0.00018999207414298067,
      "loss": 0.2162,
      "step": 224
    },
    {
      "epoch": 0.06654097815237885,
      "grad_norm": 0.19809496402740479,
      "learning_rate": 0.00018989993441398726,
      "loss": 0.2176,
      "step": 225
    },
    {
      "epoch": 0.06683671583305607,
      "grad_norm": 0.23077711462974548,
      "learning_rate": 0.00018980739502891546,
      "loss": 0.2389,
      "step": 226
    },
    {
      "epoch": 0.06713245351373331,
      "grad_norm": 0.24560177326202393,
      "learning_rate": 0.0001897144563991552,
      "loss": 0.2347,
      "step": 227
    },
    {
      "epoch": 0.06742819119441056,
      "grad_norm": 0.24137191474437714,
      "learning_rate": 0.00018962111893787128,
      "loss": 0.2191,
      "step": 228
    },
    {
      "epoch": 0.0677239288750878,
      "grad_norm": 0.23104238510131836,
      "learning_rate": 0.00018952738306000151,
      "loss": 0.2207,
      "step": 229
    },
    {
      "epoch": 0.06801966655576504,
      "grad_norm": 0.22218018770217896,
      "learning_rate": 0.00018943324918225494,
      "loss": 0.21,
      "step": 230
    },
    {
      "epoch": 0.06831540423644228,
      "grad_norm": 0.21070846915245056,
      "learning_rate": 0.0001893387177231099,
      "loss": 0.165,
      "step": 231
    },
    {
      "epoch": 0.06861114191711952,
      "grad_norm": 0.2444663941860199,
      "learning_rate": 0.0001892437891028122,
      "loss": 0.2336,
      "step": 232
    },
    {
      "epoch": 0.06890687959779676,
      "grad_norm": 0.2369096875190735,
      "learning_rate": 0.0001891484637433733,
      "loss": 0.2232,
      "step": 233
    },
    {
      "epoch": 0.06920261727847399,
      "grad_norm": 0.2586706280708313,
      "learning_rate": 0.00018905274206856837,
      "loss": 0.2468,
      "step": 234
    },
    {
      "epoch": 0.06949835495915123,
      "grad_norm": 0.30194905400276184,
      "learning_rate": 0.00018895662450393438,
      "loss": 0.254,
      "step": 235
    },
    {
      "epoch": 0.06979409263982847,
      "grad_norm": 0.2761548161506653,
      "learning_rate": 0.00018886011147676833,
      "loss": 0.2953,
      "step": 236
    },
    {
      "epoch": 0.07008983032050571,
      "grad_norm": 0.23956750333309174,
      "learning_rate": 0.00018876320341612522,
      "loss": 0.2249,
      "step": 237
    },
    {
      "epoch": 0.07038556800118295,
      "grad_norm": 0.2570217549800873,
      "learning_rate": 0.00018866590075281624,
      "loss": 0.2153,
      "step": 238
    },
    {
      "epoch": 0.0706813056818602,
      "grad_norm": 0.2857867181301117,
      "learning_rate": 0.00018856820391940674,
      "loss": 0.3025,
      "step": 239
    },
    {
      "epoch": 0.07097704336253743,
      "grad_norm": 0.25232836604118347,
      "learning_rate": 0.00018847011335021449,
      "loss": 0.201,
      "step": 240
    },
    {
      "epoch": 0.07127278104321468,
      "grad_norm": 0.28288912773132324,
      "learning_rate": 0.00018837162948130752,
      "loss": 0.2385,
      "step": 241
    },
    {
      "epoch": 0.0715685187238919,
      "grad_norm": 0.3228309154510498,
      "learning_rate": 0.00018827275275050233,
      "loss": 0.2761,
      "step": 242
    },
    {
      "epoch": 0.07186425640456914,
      "grad_norm": 0.2894141972064972,
      "learning_rate": 0.00018817348359736203,
      "loss": 0.2615,
      "step": 243
    },
    {
      "epoch": 0.07215999408524638,
      "grad_norm": 0.3046562373638153,
      "learning_rate": 0.00018807382246319412,
      "loss": 0.237,
      "step": 244
    },
    {
      "epoch": 0.07245573176592363,
      "grad_norm": 0.3270159661769867,
      "learning_rate": 0.00018797376979104872,
      "loss": 0.209,
      "step": 245
    },
    {
      "epoch": 0.07275146944660087,
      "grad_norm": 0.2986398935317993,
      "learning_rate": 0.00018787332602571662,
      "loss": 0.2032,
      "step": 246
    },
    {
      "epoch": 0.07304720712727811,
      "grad_norm": 0.3203541040420532,
      "learning_rate": 0.00018777249161372713,
      "loss": 0.2154,
      "step": 247
    },
    {
      "epoch": 0.07334294480795535,
      "grad_norm": 0.3694257438182831,
      "learning_rate": 0.00018767126700334634,
      "loss": 0.1959,
      "step": 248
    },
    {
      "epoch": 0.07363868248863259,
      "grad_norm": 0.6173413991928101,
      "learning_rate": 0.0001875696526445749,
      "loss": 0.1682,
      "step": 249
    },
    {
      "epoch": 0.07393442016930982,
      "grad_norm": 0.5118496417999268,
      "learning_rate": 0.0001874676489891461,
      "loss": 0.276,
      "step": 250
    },
    {
      "epoch": 0.07423015784998706,
      "grad_norm": 0.15977254509925842,
      "learning_rate": 0.00018736525649052394,
      "loss": 0.2124,
      "step": 251
    },
    {
      "epoch": 0.0745258955306643,
      "grad_norm": 0.15199963748455048,
      "learning_rate": 0.00018726247560390099,
      "loss": 0.1718,
      "step": 252
    },
    {
      "epoch": 0.07482163321134154,
      "grad_norm": 0.16801315546035767,
      "learning_rate": 0.00018715930678619644,
      "loss": 0.1961,
      "step": 253
    },
    {
      "epoch": 0.07511737089201878,
      "grad_norm": 0.17173269391059875,
      "learning_rate": 0.00018705575049605413,
      "loss": 0.2053,
      "step": 254
    },
    {
      "epoch": 0.07541310857269602,
      "grad_norm": 0.19244447350502014,
      "learning_rate": 0.00018695180719384029,
      "loss": 0.2841,
      "step": 255
    },
    {
      "epoch": 0.07570884625337326,
      "grad_norm": 0.1731448471546173,
      "learning_rate": 0.00018684747734164177,
      "loss": 0.2246,
      "step": 256
    },
    {
      "epoch": 0.07600458393405049,
      "grad_norm": 0.1740599274635315,
      "learning_rate": 0.00018674276140326376,
      "loss": 0.229,
      "step": 257
    },
    {
      "epoch": 0.07630032161472773,
      "grad_norm": 0.15664459764957428,
      "learning_rate": 0.00018663765984422786,
      "loss": 0.2015,
      "step": 258
    },
    {
      "epoch": 0.07659605929540497,
      "grad_norm": 0.18093277513980865,
      "learning_rate": 0.00018653217313177004,
      "loss": 0.2371,
      "step": 259
    },
    {
      "epoch": 0.07689179697608221,
      "grad_norm": 0.15789927542209625,
      "learning_rate": 0.00018642630173483832,
      "loss": 0.19,
      "step": 260
    },
    {
      "epoch": 0.07718753465675945,
      "grad_norm": 0.1597701758146286,
      "learning_rate": 0.00018632004612409103,
      "loss": 0.2096,
      "step": 261
    },
    {
      "epoch": 0.0774832723374367,
      "grad_norm": 0.17183595895767212,
      "learning_rate": 0.00018621340677189453,
      "loss": 0.2333,
      "step": 262
    },
    {
      "epoch": 0.07777901001811394,
      "grad_norm": 0.17315109074115753,
      "learning_rate": 0.00018610638415232097,
      "loss": 0.2215,
      "step": 263
    },
    {
      "epoch": 0.07807474769879118,
      "grad_norm": 0.17420059442520142,
      "learning_rate": 0.00018599897874114652,
      "loss": 0.2216,
      "step": 264
    },
    {
      "epoch": 0.0783704853794684,
      "grad_norm": 0.19842545688152313,
      "learning_rate": 0.00018589119101584898,
      "loss": 0.212,
      "step": 265
    },
    {
      "epoch": 0.07866622306014565,
      "grad_norm": 0.18231399357318878,
      "learning_rate": 0.00018578302145560584,
      "loss": 0.2284,
      "step": 266
    },
    {
      "epoch": 0.07896196074082289,
      "grad_norm": 0.1868034154176712,
      "learning_rate": 0.00018567447054129195,
      "loss": 0.2458,
      "step": 267
    },
    {
      "epoch": 0.07925769842150013,
      "grad_norm": 0.19086702167987823,
      "learning_rate": 0.00018556553875547754,
      "loss": 0.2335,
      "step": 268
    },
    {
      "epoch": 0.07955343610217737,
      "grad_norm": 0.2202424705028534,
      "learning_rate": 0.00018545622658242607,
      "loss": 0.282,
      "step": 269
    },
    {
      "epoch": 0.07984917378285461,
      "grad_norm": 0.20102711021900177,
      "learning_rate": 0.00018534653450809197,
      "loss": 0.2376,
      "step": 270
    },
    {
      "epoch": 0.08014491146353185,
      "grad_norm": 0.2291051596403122,
      "learning_rate": 0.00018523646302011867,
      "loss": 0.2448,
      "step": 271
    },
    {
      "epoch": 0.08044064914420909,
      "grad_norm": 0.20590786635875702,
      "learning_rate": 0.00018512601260783606,
      "loss": 0.2235,
      "step": 272
    },
    {
      "epoch": 0.08073638682488632,
      "grad_norm": 0.18361449241638184,
      "learning_rate": 0.00018501518376225887,
      "loss": 0.1882,
      "step": 273
    },
    {
      "epoch": 0.08103212450556356,
      "grad_norm": 0.20607051253318787,
      "learning_rate": 0.00018490397697608395,
      "loss": 0.216,
      "step": 274
    },
    {
      "epoch": 0.0813278621862408,
      "grad_norm": 0.20951667428016663,
      "learning_rate": 0.0001847923927436884,
      "loss": 0.2283,
      "step": 275
    },
    {
      "epoch": 0.08162359986691804,
      "grad_norm": 0.23473979532718658,
      "learning_rate": 0.00018468043156112728,
      "loss": 0.2582,
      "step": 276
    },
    {
      "epoch": 0.08191933754759528,
      "grad_norm": 0.1933351308107376,
      "learning_rate": 0.0001845680939261314,
      "loss": 0.1938,
      "step": 277
    },
    {
      "epoch": 0.08221507522827252,
      "grad_norm": 0.23549608886241913,
      "learning_rate": 0.00018445538033810515,
      "loss": 0.2166,
      "step": 278
    },
    {
      "epoch": 0.08251081290894977,
      "grad_norm": 0.2569220960140228,
      "learning_rate": 0.00018434229129812418,
      "loss": 0.2697,
      "step": 279
    },
    {
      "epoch": 0.082806550589627,
      "grad_norm": 0.22212399542331696,
      "learning_rate": 0.0001842288273089332,
      "loss": 0.26,
      "step": 280
    },
    {
      "epoch": 0.08310228827030423,
      "grad_norm": 0.21168506145477295,
      "learning_rate": 0.00018411498887494396,
      "loss": 0.2008,
      "step": 281
    },
    {
      "epoch": 0.08339802595098147,
      "grad_norm": 0.23749881982803345,
      "learning_rate": 0.00018400077650223263,
      "loss": 0.2163,
      "step": 282
    },
    {
      "epoch": 0.08369376363165872,
      "grad_norm": 0.23951789736747742,
      "learning_rate": 0.0001838861906985379,
      "loss": 0.2235,
      "step": 283
    },
    {
      "epoch": 0.08398950131233596,
      "grad_norm": 0.23467004299163818,
      "learning_rate": 0.00018377123197325842,
      "loss": 0.2119,
      "step": 284
    },
    {
      "epoch": 0.0842852389930132,
      "grad_norm": 0.25748157501220703,
      "learning_rate": 0.00018365590083745085,
      "loss": 0.2498,
      "step": 285
    },
    {
      "epoch": 0.08458097667369044,
      "grad_norm": 0.2640005946159363,
      "learning_rate": 0.00018354019780382735,
      "loss": 0.2297,
      "step": 286
    },
    {
      "epoch": 0.08487671435436768,
      "grad_norm": 0.22059418261051178,
      "learning_rate": 0.0001834241233867533,
      "loss": 0.1856,
      "step": 287
    },
    {
      "epoch": 0.08517245203504492,
      "grad_norm": 0.2318236380815506,
      "learning_rate": 0.00018330767810224524,
      "loss": 0.186,
      "step": 288
    },
    {
      "epoch": 0.08546818971572215,
      "grad_norm": 0.22884918749332428,
      "learning_rate": 0.0001831908624679683,
      "loss": 0.2019,
      "step": 289
    },
    {
      "epoch": 0.08576392739639939,
      "grad_norm": 0.26514625549316406,
      "learning_rate": 0.0001830736770032341,
      "loss": 0.2338,
      "step": 290
    },
    {
      "epoch": 0.08605966507707663,
      "grad_norm": 0.2545234262943268,
      "learning_rate": 0.0001829561222289984,
      "loss": 0.2051,
      "step": 291
    },
    {
      "epoch": 0.08635540275775387,
      "grad_norm": 0.31720927357673645,
      "learning_rate": 0.00018283819866785853,
      "loss": 0.2127,
      "step": 292
    },
    {
      "epoch": 0.08665114043843111,
      "grad_norm": 0.24507926404476166,
      "learning_rate": 0.0001827199068440516,
      "loss": 0.1481,
      "step": 293
    },
    {
      "epoch": 0.08694687811910835,
      "grad_norm": 0.2604232728481293,
      "learning_rate": 0.00018260124728345162,
      "loss": 0.1804,
      "step": 294
    },
    {
      "epoch": 0.0872426157997856,
      "grad_norm": 0.2785932123661041,
      "learning_rate": 0.00018248222051356754,
      "loss": 0.2188,
      "step": 295
    },
    {
      "epoch": 0.08753835348046284,
      "grad_norm": 0.3473379611968994,
      "learning_rate": 0.00018236282706354063,
      "loss": 0.2223,
      "step": 296
    },
    {
      "epoch": 0.08783409116114006,
      "grad_norm": 0.35401713848114014,
      "learning_rate": 0.00018224306746414238,
      "loss": 0.2602,
      "step": 297
    },
    {
      "epoch": 0.0881298288418173,
      "grad_norm": 0.4033154249191284,
      "learning_rate": 0.00018212294224777197,
      "loss": 0.2324,
      "step": 298
    },
    {
      "epoch": 0.08842556652249454,
      "grad_norm": 0.5243091583251953,
      "learning_rate": 0.00018200245194845399,
      "loss": 0.3447,
      "step": 299
    },
    {
      "epoch": 0.08872130420317179,
      "grad_norm": 0.5621427297592163,
      "learning_rate": 0.00018188159710183594,
      "loss": 0.2627,
      "step": 300
    },
    {
      "epoch": 0.08901704188384903,
      "grad_norm": 0.1358180195093155,
      "learning_rate": 0.000181760378245186,
      "loss": 0.1579,
      "step": 301
    },
    {
      "epoch": 0.08931277956452627,
      "grad_norm": 0.16979822516441345,
      "learning_rate": 0.00018163879591739067,
      "loss": 0.2157,
      "step": 302
    },
    {
      "epoch": 0.08960851724520351,
      "grad_norm": 0.18489453196525574,
      "learning_rate": 0.0001815168506589521,
      "loss": 0.2149,
      "step": 303
    },
    {
      "epoch": 0.08990425492588075,
      "grad_norm": 0.19990918040275574,
      "learning_rate": 0.000181394543011986,
      "loss": 0.254,
      "step": 304
    },
    {
      "epoch": 0.09019999260655798,
      "grad_norm": 0.17698073387145996,
      "learning_rate": 0.00018127187352021907,
      "loss": 0.2131,
      "step": 305
    },
    {
      "epoch": 0.09049573028723522,
      "grad_norm": 0.1820111721754074,
      "learning_rate": 0.0001811488427289866,
      "loss": 0.2372,
      "step": 306
    },
    {
      "epoch": 0.09079146796791246,
      "grad_norm": 0.1698734015226364,
      "learning_rate": 0.00018102545118523007,
      "loss": 0.2199,
      "step": 307
    },
    {
      "epoch": 0.0910872056485897,
      "grad_norm": 0.17691490054130554,
      "learning_rate": 0.00018090169943749476,
      "loss": 0.2474,
      "step": 308
    },
    {
      "epoch": 0.09138294332926694,
      "grad_norm": 0.1919829100370407,
      "learning_rate": 0.00018077758803592718,
      "loss": 0.2447,
      "step": 309
    },
    {
      "epoch": 0.09167868100994418,
      "grad_norm": 0.4307519495487213,
      "learning_rate": 0.00018065311753227273,
      "loss": 0.2342,
      "step": 310
    },
    {
      "epoch": 0.09197441869062142,
      "grad_norm": 0.18838980793952942,
      "learning_rate": 0.0001805282884798732,
      "loss": 0.1829,
      "step": 311
    },
    {
      "epoch": 0.09227015637129866,
      "grad_norm": 0.1960785984992981,
      "learning_rate": 0.00018040310143366446,
      "loss": 0.2279,
      "step": 312
    },
    {
      "epoch": 0.09256589405197589,
      "grad_norm": 0.1780913770198822,
      "learning_rate": 0.00018027755695017368,
      "loss": 0.222,
      "step": 313
    },
    {
      "epoch": 0.09286163173265313,
      "grad_norm": 0.1910991370677948,
      "learning_rate": 0.00018015165558751717,
      "loss": 0.2378,
      "step": 314
    },
    {
      "epoch": 0.09315736941333037,
      "grad_norm": 0.17035876214504242,
      "learning_rate": 0.00018002539790539773,
      "loss": 0.1834,
      "step": 315
    },
    {
      "epoch": 0.09345310709400761,
      "grad_norm": 0.18461062014102936,
      "learning_rate": 0.00017989878446510215,
      "loss": 0.2255,
      "step": 316
    },
    {
      "epoch": 0.09374884477468486,
      "grad_norm": 0.18789273500442505,
      "learning_rate": 0.00017977181582949888,
      "loss": 0.2236,
      "step": 317
    },
    {
      "epoch": 0.0940445824553621,
      "grad_norm": 0.18466150760650635,
      "learning_rate": 0.0001796444925630353,
      "loss": 0.1987,
      "step": 318
    },
    {
      "epoch": 0.09434032013603934,
      "grad_norm": 0.2048148661851883,
      "learning_rate": 0.00017951681523173542,
      "loss": 0.2059,
      "step": 319
    },
    {
      "epoch": 0.09463605781671658,
      "grad_norm": 0.2097977101802826,
      "learning_rate": 0.0001793887844031972,
      "loss": 0.231,
      "step": 320
    },
    {
      "epoch": 0.0949317954973938,
      "grad_norm": 0.20636926591396332,
      "learning_rate": 0.00017926040064659014,
      "loss": 0.2326,
      "step": 321
    },
    {
      "epoch": 0.09522753317807105,
      "grad_norm": 0.2094830870628357,
      "learning_rate": 0.0001791316645326526,
      "loss": 0.2059,
      "step": 322
    },
    {
      "epoch": 0.09552327085874829,
      "grad_norm": 0.22407394647598267,
      "learning_rate": 0.00017900257663368963,
      "loss": 0.226,
      "step": 323
    },
    {
      "epoch": 0.09581900853942553,
      "grad_norm": 0.19603627920150757,
      "learning_rate": 0.0001788731375235698,
      "loss": 0.2188,
      "step": 324
    },
    {
      "epoch": 0.09611474622010277,
      "grad_norm": 0.23884060978889465,
      "learning_rate": 0.00017874334777772327,
      "loss": 0.2372,
      "step": 325
    },
    {
      "epoch": 0.09641048390078001,
      "grad_norm": 0.22886501252651215,
      "learning_rate": 0.00017861320797313892,
      "loss": 0.2336,
      "step": 326
    },
    {
      "epoch": 0.09670622158145725,
      "grad_norm": 0.22485503554344177,
      "learning_rate": 0.0001784827186883618,
      "loss": 0.2339,
      "step": 327
    },
    {
      "epoch": 0.0970019592621345,
      "grad_norm": 0.21763861179351807,
      "learning_rate": 0.00017835188050349064,
      "loss": 0.2002,
      "step": 328
    },
    {
      "epoch": 0.09729769694281172,
      "grad_norm": 0.2242451012134552,
      "learning_rate": 0.00017822069400017516,
      "loss": 0.2347,
      "step": 329
    },
    {
      "epoch": 0.09759343462348896,
      "grad_norm": 0.24562625586986542,
      "learning_rate": 0.00017808915976161362,
      "loss": 0.2339,
      "step": 330
    },
    {
      "epoch": 0.0978891723041662,
      "grad_norm": 0.2314700335264206,
      "learning_rate": 0.00017795727837255015,
      "loss": 0.2708,
      "step": 331
    },
    {
      "epoch": 0.09818490998484344,
      "grad_norm": 0.26339125633239746,
      "learning_rate": 0.00017782505041927216,
      "loss": 0.2557,
      "step": 332
    },
    {
      "epoch": 0.09848064766552068,
      "grad_norm": 0.22803835570812225,
      "learning_rate": 0.00017769247648960774,
      "loss": 0.196,
      "step": 333
    },
    {
      "epoch": 0.09877638534619793,
      "grad_norm": 0.22814923524856567,
      "learning_rate": 0.00017755955717292296,
      "loss": 0.1687,
      "step": 334
    },
    {
      "epoch": 0.09907212302687517,
      "grad_norm": 0.25240692496299744,
      "learning_rate": 0.00017742629306011944,
      "loss": 0.2064,
      "step": 335
    },
    {
      "epoch": 0.09936786070755241,
      "grad_norm": 0.23173846304416656,
      "learning_rate": 0.00017729268474363154,
      "loss": 0.1797,
      "step": 336
    },
    {
      "epoch": 0.09966359838822964,
      "grad_norm": 0.2367689609527588,
      "learning_rate": 0.0001771587328174239,
      "loss": 0.1796,
      "step": 337
    },
    {
      "epoch": 0.09995933606890688,
      "grad_norm": 0.2612314522266388,
      "learning_rate": 0.0001770244378769885,
      "loss": 0.2132,
      "step": 338
    },
    {
      "epoch": 0.10025507374958412,
      "grad_norm": 0.2403004765510559,
      "learning_rate": 0.0001768898005193425,
      "loss": 0.1919,
      "step": 339
    },
    {
      "epoch": 0.10055081143026136,
      "grad_norm": 0.26991987228393555,
      "learning_rate": 0.000176754821343025,
      "loss": 0.2262,
      "step": 340
    },
    {
      "epoch": 0.1008465491109386,
      "grad_norm": 0.28544318675994873,
      "learning_rate": 0.0001766195009480949,
      "loss": 0.2229,
      "step": 341
    },
    {
      "epoch": 0.10114228679161584,
      "grad_norm": 0.2717716693878174,
      "learning_rate": 0.0001764838399361279,
      "loss": 0.2146,
      "step": 342
    },
    {
      "epoch": 0.10143802447229308,
      "grad_norm": 0.2891225218772888,
      "learning_rate": 0.00017634783891021393,
      "loss": 0.2462,
      "step": 343
    },
    {
      "epoch": 0.10173376215297031,
      "grad_norm": 0.3248477578163147,
      "learning_rate": 0.00017621149847495458,
      "loss": 0.2475,
      "step": 344
    },
    {
      "epoch": 0.10202949983364755,
      "grad_norm": 0.3657064735889435,
      "learning_rate": 0.00017607481923646016,
      "loss": 0.28,
      "step": 345
    },
    {
      "epoch": 0.10232523751432479,
      "grad_norm": 0.3362722396850586,
      "learning_rate": 0.0001759378018023473,
      "loss": 0.2462,
      "step": 346
    },
    {
      "epoch": 0.10262097519500203,
      "grad_norm": 0.3996289074420929,
      "learning_rate": 0.00017580044678173592,
      "loss": 0.2258,
      "step": 347
    },
    {
      "epoch": 0.10291671287567927,
      "grad_norm": 0.36670979857444763,
      "learning_rate": 0.00017566275478524693,
      "loss": 0.2466,
      "step": 348
    },
    {
      "epoch": 0.10321245055635651,
      "grad_norm": 0.48458173871040344,
      "learning_rate": 0.0001755247264249991,
      "loss": 0.3022,
      "step": 349
    },
    {
      "epoch": 0.10350818823703375,
      "grad_norm": 0.49845030903816223,
      "learning_rate": 0.0001753863623146066,
      "loss": 0.2424,
      "step": 350
    },
    {
      "epoch": 0.103803925917711,
      "grad_norm": 0.11847740411758423,
      "learning_rate": 0.00017524766306917618,
      "loss": 0.1449,
      "step": 351
    },
    {
      "epoch": 0.10409966359838822,
      "grad_norm": 0.14606598019599915,
      "learning_rate": 0.0001751086293053045,
      "loss": 0.1713,
      "step": 352
    },
    {
      "epoch": 0.10439540127906546,
      "grad_norm": 0.1666148602962494,
      "learning_rate": 0.0001749692616410753,
      "loss": 0.2063,
      "step": 353
    },
    {
      "epoch": 0.1046911389597427,
      "grad_norm": 0.1671825349330902,
      "learning_rate": 0.00017482956069605668,
      "loss": 0.2283,
      "step": 354
    },
    {
      "epoch": 0.10498687664041995,
      "grad_norm": 0.17825308442115784,
      "learning_rate": 0.00017468952709129846,
      "loss": 0.2379,
      "step": 355
    },
    {
      "epoch": 0.10528261432109719,
      "grad_norm": 0.1678950935602188,
      "learning_rate": 0.00017454916144932922,
      "loss": 0.2227,
      "step": 356
    },
    {
      "epoch": 0.10557835200177443,
      "grad_norm": 0.16522592306137085,
      "learning_rate": 0.0001744084643941536,
      "loss": 0.2099,
      "step": 357
    },
    {
      "epoch": 0.10587408968245167,
      "grad_norm": 0.17215241491794586,
      "learning_rate": 0.00017426743655124974,
      "loss": 0.2612,
      "step": 358
    },
    {
      "epoch": 0.10616982736312891,
      "grad_norm": 0.17087267339229584,
      "learning_rate": 0.0001741260785475661,
      "loss": 0.2301,
      "step": 359
    },
    {
      "epoch": 0.10646556504380614,
      "grad_norm": 0.1661638766527176,
      "learning_rate": 0.00017398439101151905,
      "loss": 0.2174,
      "step": 360
    },
    {
      "epoch": 0.10676130272448338,
      "grad_norm": 0.17075875401496887,
      "learning_rate": 0.00017384237457298987,
      "loss": 0.1701,
      "step": 361
    },
    {
      "epoch": 0.10705704040516062,
      "grad_norm": 0.1797911673784256,
      "learning_rate": 0.00017370002986332193,
      "loss": 0.2192,
      "step": 362
    },
    {
      "epoch": 0.10735277808583786,
      "grad_norm": 0.17848677933216095,
      "learning_rate": 0.00017355735751531807,
      "loss": 0.2193,
      "step": 363
    },
    {
      "epoch": 0.1076485157665151,
      "grad_norm": 0.17944815754890442,
      "learning_rate": 0.00017341435816323756,
      "loss": 0.2008,
      "step": 364
    },
    {
      "epoch": 0.10794425344719234,
      "grad_norm": 0.17322401702404022,
      "learning_rate": 0.00017327103244279348,
      "loss": 0.2084,
      "step": 365
    },
    {
      "epoch": 0.10823999112786958,
      "grad_norm": 0.19021335244178772,
      "learning_rate": 0.00017312738099114973,
      "loss": 0.2398,
      "step": 366
    },
    {
      "epoch": 0.10853572880854682,
      "grad_norm": 0.18541230261325836,
      "learning_rate": 0.00017298340444691835,
      "loss": 0.2618,
      "step": 367
    },
    {
      "epoch": 0.10883146648922405,
      "grad_norm": 0.18366391956806183,
      "learning_rate": 0.00017283910345015647,
      "loss": 0.2356,
      "step": 368
    },
    {
      "epoch": 0.10912720416990129,
      "grad_norm": 0.1941191852092743,
      "learning_rate": 0.0001726944786423637,
      "loss": 0.2228,
      "step": 369
    },
    {
      "epoch": 0.10942294185057853,
      "grad_norm": 0.16953422129154205,
      "learning_rate": 0.00017254953066647913,
      "loss": 0.1555,
      "step": 370
    },
    {
      "epoch": 0.10971867953125577,
      "grad_norm": 0.20326456427574158,
      "learning_rate": 0.00017240426016687863,
      "loss": 0.2514,
      "step": 371
    },
    {
      "epoch": 0.11001441721193302,
      "grad_norm": 0.1932348906993866,
      "learning_rate": 0.00017225866778937165,
      "loss": 0.2068,
      "step": 372
    },
    {
      "epoch": 0.11031015489261026,
      "grad_norm": 0.21190403401851654,
      "learning_rate": 0.00017211275418119876,
      "loss": 0.2196,
      "step": 373
    },
    {
      "epoch": 0.1106058925732875,
      "grad_norm": 0.20619221031665802,
      "learning_rate": 0.0001719665199910285,
      "loss": 0.2368,
      "step": 374
    },
    {
      "epoch": 0.11090163025396474,
      "grad_norm": 0.1878051608800888,
      "learning_rate": 0.00017181996586895454,
      "loss": 0.1801,
      "step": 375
    },
    {
      "epoch": 0.11090163025396474,
      "eval_loss": 0.21186766028404236,
      "eval_runtime": 323.3922,
      "eval_samples_per_second": 17.61,
      "eval_steps_per_second": 8.807,
      "step": 375
    },
    {
      "epoch": 0.11119736793464197,
      "grad_norm": 0.22650651633739471,
      "learning_rate": 0.00017167309246649297,
      "loss": 0.2119,
      "step": 376
    },
    {
      "epoch": 0.11149310561531921,
      "grad_norm": 0.21036145091056824,
      "learning_rate": 0.0001715259004365791,
      "loss": 0.2579,
      "step": 377
    },
    {
      "epoch": 0.11178884329599645,
      "grad_norm": 0.2401810735464096,
      "learning_rate": 0.00017137839043356484,
      "loss": 0.2931,
      "step": 378
    },
    {
      "epoch": 0.11208458097667369,
      "grad_norm": 0.22352933883666992,
      "learning_rate": 0.00017123056311321562,
      "loss": 0.213,
      "step": 379
    },
    {
      "epoch": 0.11238031865735093,
      "grad_norm": 0.24567389488220215,
      "learning_rate": 0.0001710824191327075,
      "loss": 0.2738,
      "step": 380
    },
    {
      "epoch": 0.11267605633802817,
      "grad_norm": 0.24588455259799957,
      "learning_rate": 0.00017093395915062428,
      "loss": 0.2252,
      "step": 381
    },
    {
      "epoch": 0.11297179401870541,
      "grad_norm": 0.21331793069839478,
      "learning_rate": 0.00017078518382695465,
      "loss": 0.165,
      "step": 382
    },
    {
      "epoch": 0.11326753169938265,
      "grad_norm": 0.2419130951166153,
      "learning_rate": 0.00017063609382308908,
      "loss": 0.2355,
      "step": 383
    },
    {
      "epoch": 0.11356326938005988,
      "grad_norm": 0.25924235582351685,
      "learning_rate": 0.00017048668980181698,
      "loss": 0.233,
      "step": 384
    },
    {
      "epoch": 0.11385900706073712,
      "grad_norm": 0.26302260160446167,
      "learning_rate": 0.00017033697242732377,
      "loss": 0.2379,
      "step": 385
    },
    {
      "epoch": 0.11415474474141436,
      "grad_norm": 0.24101543426513672,
      "learning_rate": 0.0001701869423651879,
      "loss": 0.1956,
      "step": 386
    },
    {
      "epoch": 0.1144504824220916,
      "grad_norm": 0.26949170231819153,
      "learning_rate": 0.00017003660028237793,
      "loss": 0.2365,
      "step": 387
    },
    {
      "epoch": 0.11474622010276884,
      "grad_norm": 0.22656166553497314,
      "learning_rate": 0.00016988594684724947,
      "loss": 0.1488,
      "step": 388
    },
    {
      "epoch": 0.11504195778344609,
      "grad_norm": 0.25537586212158203,
      "learning_rate": 0.00016973498272954222,
      "loss": 0.2277,
      "step": 389
    },
    {
      "epoch": 0.11533769546412333,
      "grad_norm": 0.2855624556541443,
      "learning_rate": 0.00016958370860037717,
      "loss": 0.2155,
      "step": 390
    },
    {
      "epoch": 0.11563343314480057,
      "grad_norm": 0.2733490467071533,
      "learning_rate": 0.00016943212513225345,
      "loss": 0.2189,
      "step": 391
    },
    {
      "epoch": 0.1159291708254778,
      "grad_norm": 0.2860839366912842,
      "learning_rate": 0.00016928023299904533,
      "loss": 0.276,
      "step": 392
    },
    {
      "epoch": 0.11622490850615504,
      "grad_norm": 0.29845502972602844,
      "learning_rate": 0.0001691280328759992,
      "loss": 0.2542,
      "step": 393
    },
    {
      "epoch": 0.11652064618683228,
      "grad_norm": 0.3229489326477051,
      "learning_rate": 0.00016897552543973084,
      "loss": 0.239,
      "step": 394
    },
    {
      "epoch": 0.11681638386750952,
      "grad_norm": 0.31158486008644104,
      "learning_rate": 0.00016882271136822206,
      "loss": 0.2078,
      "step": 395
    },
    {
      "epoch": 0.11711212154818676,
      "grad_norm": 0.33740320801734924,
      "learning_rate": 0.0001686695913408179,
      "loss": 0.2092,
      "step": 396
    },
    {
      "epoch": 0.117407859228864,
      "grad_norm": 0.42870861291885376,
      "learning_rate": 0.0001685161660382235,
      "loss": 0.3145,
      "step": 397
    },
    {
      "epoch": 0.11770359690954124,
      "grad_norm": 0.41335102915763855,
      "learning_rate": 0.00016836243614250113,
      "loss": 0.2852,
      "step": 398
    },
    {
      "epoch": 0.11799933459021848,
      "grad_norm": 0.3501723110675812,
      "learning_rate": 0.00016820840233706719,
      "loss": 0.2022,
      "step": 399
    },
    {
      "epoch": 0.11829507227089571,
      "grad_norm": 0.511377215385437,
      "learning_rate": 0.0001680540653066891,
      "loss": 0.3176,
      "step": 400
    },
    {
      "epoch": 0.11859080995157295,
      "grad_norm": 0.13668859004974365,
      "learning_rate": 0.00016789942573748232,
      "loss": 0.1905,
      "step": 401
    },
    {
      "epoch": 0.11888654763225019,
      "grad_norm": 0.15192629396915436,
      "learning_rate": 0.0001677444843169072,
      "loss": 0.2082,
      "step": 402
    },
    {
      "epoch": 0.11918228531292743,
      "grad_norm": 0.16673530638217926,
      "learning_rate": 0.00016758924173376603,
      "loss": 0.2217,
      "step": 403
    },
    {
      "epoch": 0.11947802299360467,
      "grad_norm": 0.14643211662769318,
      "learning_rate": 0.0001674336986781999,
      "loss": 0.1797,
      "step": 404
    },
    {
      "epoch": 0.11977376067428191,
      "grad_norm": 0.1829608678817749,
      "learning_rate": 0.00016727785584168581,
      "loss": 0.2738,
      "step": 405
    },
    {
      "epoch": 0.12006949835495916,
      "grad_norm": 0.1723003387451172,
      "learning_rate": 0.0001671217139170333,
      "loss": 0.2193,
      "step": 406
    },
    {
      "epoch": 0.1203652360356364,
      "grad_norm": 0.175453320145607,
      "learning_rate": 0.00016696527359838154,
      "loss": 0.2252,
      "step": 407
    },
    {
      "epoch": 0.12066097371631362,
      "grad_norm": 0.18448546528816223,
      "learning_rate": 0.00016680853558119632,
      "loss": 0.2511,
      "step": 408
    },
    {
      "epoch": 0.12095671139699087,
      "grad_norm": 0.18752612173557281,
      "learning_rate": 0.0001666515005622668,
      "loss": 0.2624,
      "step": 409
    },
    {
      "epoch": 0.1212524490776681,
      "grad_norm": 0.15819008648395538,
      "learning_rate": 0.0001664941692397025,
      "loss": 0.1884,
      "step": 410
    },
    {
      "epoch": 0.12154818675834535,
      "grad_norm": 0.17365534603595734,
      "learning_rate": 0.00016633654231293013,
      "loss": 0.2065,
      "step": 411
    },
    {
      "epoch": 0.12184392443902259,
      "grad_norm": 0.1719895303249359,
      "learning_rate": 0.00016617862048269065,
      "loss": 0.2239,
      "step": 412
    },
    {
      "epoch": 0.12213966211969983,
      "grad_norm": 0.1667676866054535,
      "learning_rate": 0.00016602040445103588,
      "loss": 0.1775,
      "step": 413
    },
    {
      "epoch": 0.12243539980037707,
      "grad_norm": 0.17780280113220215,
      "learning_rate": 0.00016586189492132566,
      "loss": 0.2271,
      "step": 414
    },
    {
      "epoch": 0.12273113748105431,
      "grad_norm": 0.16043958067893982,
      "learning_rate": 0.00016570309259822453,
      "loss": 0.19,
      "step": 415
    },
    {
      "epoch": 0.12302687516173154,
      "grad_norm": 0.2024742215871811,
      "learning_rate": 0.0001655439981876987,
      "loss": 0.207,
      "step": 416
    },
    {
      "epoch": 0.12332261284240878,
      "grad_norm": 0.1686304360628128,
      "learning_rate": 0.00016538461239701277,
      "loss": 0.1874,
      "step": 417
    },
    {
      "epoch": 0.12361835052308602,
      "grad_norm": 0.19449354708194733,
      "learning_rate": 0.00016522493593472683,
      "loss": 0.2119,
      "step": 418
    },
    {
      "epoch": 0.12391408820376326,
      "grad_norm": 0.16912192106246948,
      "learning_rate": 0.0001650649695106931,
      "loss": 0.1543,
      "step": 419
    },
    {
      "epoch": 0.1242098258844405,
      "grad_norm": 0.20457497239112854,
      "learning_rate": 0.00016490471383605288,
      "loss": 0.206,
      "step": 420
    },
    {
      "epoch": 0.12450556356511774,
      "grad_norm": 0.16567295789718628,
      "learning_rate": 0.00016474416962323325,
      "loss": 0.1518,
      "step": 421
    },
    {
      "epoch": 0.12480130124579498,
      "grad_norm": 0.19134259223937988,
      "learning_rate": 0.00016458333758594414,
      "loss": 0.207,
      "step": 422
    },
    {
      "epoch": 0.12509703892647223,
      "grad_norm": 0.22228240966796875,
      "learning_rate": 0.00016442221843917496,
      "loss": 0.2208,
      "step": 423
    },
    {
      "epoch": 0.12539277660714945,
      "grad_norm": 0.23446987569332123,
      "learning_rate": 0.00016426081289919143,
      "loss": 0.2229,
      "step": 424
    },
    {
      "epoch": 0.1256885142878267,
      "grad_norm": 0.20911334455013275,
      "learning_rate": 0.0001640991216835326,
      "loss": 0.1946,
      "step": 425
    },
    {
      "epoch": 0.12598425196850394,
      "grad_norm": 0.21853110194206238,
      "learning_rate": 0.00016393714551100734,
      "loss": 0.2157,
      "step": 426
    },
    {
      "epoch": 0.12627998964918116,
      "grad_norm": 0.23242445290088654,
      "learning_rate": 0.0001637748851016914,
      "loss": 0.2161,
      "step": 427
    },
    {
      "epoch": 0.12657572732985842,
      "grad_norm": 0.24394045770168304,
      "learning_rate": 0.00016361234117692413,
      "loss": 0.232,
      "step": 428
    },
    {
      "epoch": 0.12687146501053564,
      "grad_norm": 0.21284179389476776,
      "learning_rate": 0.00016344951445930526,
      "loss": 0.162,
      "step": 429
    },
    {
      "epoch": 0.1271672026912129,
      "grad_norm": 0.24403221905231476,
      "learning_rate": 0.0001632864056726917,
      "loss": 0.1994,
      "step": 430
    },
    {
      "epoch": 0.12746294037189013,
      "grad_norm": 0.2400125414133072,
      "learning_rate": 0.00016312301554219426,
      "loss": 0.2004,
      "step": 431
    },
    {
      "epoch": 0.12775867805256738,
      "grad_norm": 0.2092212736606598,
      "learning_rate": 0.00016295934479417453,
      "loss": 0.1618,
      "step": 432
    },
    {
      "epoch": 0.1280544157332446,
      "grad_norm": 0.21888083219528198,
      "learning_rate": 0.00016279539415624164,
      "loss": 0.181,
      "step": 433
    },
    {
      "epoch": 0.12835015341392186,
      "grad_norm": 0.23019088804721832,
      "learning_rate": 0.0001626311643572489,
      "loss": 0.1816,
      "step": 434
    },
    {
      "epoch": 0.1286458910945991,
      "grad_norm": 0.23748983442783356,
      "learning_rate": 0.00016246665612729074,
      "loss": 0.1869,
      "step": 435
    },
    {
      "epoch": 0.12894162877527632,
      "grad_norm": 0.23818852007389069,
      "learning_rate": 0.00016230187019769928,
      "loss": 0.1959,
      "step": 436
    },
    {
      "epoch": 0.12923736645595357,
      "grad_norm": 0.27259862422943115,
      "learning_rate": 0.00016213680730104124,
      "loss": 0.2183,
      "step": 437
    },
    {
      "epoch": 0.1295331041366308,
      "grad_norm": 0.2296111136674881,
      "learning_rate": 0.0001619714681711146,
      "loss": 0.1615,
      "step": 438
    },
    {
      "epoch": 0.12982884181730805,
      "grad_norm": 0.2418239265680313,
      "learning_rate": 0.00016180585354294536,
      "loss": 0.2222,
      "step": 439
    },
    {
      "epoch": 0.13012457949798528,
      "grad_norm": 0.2754133641719818,
      "learning_rate": 0.00016163996415278424,
      "loss": 0.2433,
      "step": 440
    },
    {
      "epoch": 0.13042031717866254,
      "grad_norm": 0.31593406200408936,
      "learning_rate": 0.00016147380073810346,
      "loss": 0.2492,
      "step": 441
    },
    {
      "epoch": 0.13071605485933976,
      "grad_norm": 0.30033108592033386,
      "learning_rate": 0.0001613073640375934,
      "loss": 0.2539,
      "step": 442
    },
    {
      "epoch": 0.131011792540017,
      "grad_norm": 0.2534523606300354,
      "learning_rate": 0.00016114065479115946,
      "loss": 0.1382,
      "step": 443
    },
    {
      "epoch": 0.13130753022069425,
      "grad_norm": 0.31533321738243103,
      "learning_rate": 0.00016097367373991842,
      "loss": 0.2456,
      "step": 444
    },
    {
      "epoch": 0.13160326790137147,
      "grad_norm": 0.3007992208003998,
      "learning_rate": 0.00016080642162619565,
      "loss": 0.2018,
      "step": 445
    },
    {
      "epoch": 0.13189900558204873,
      "grad_norm": 0.3328956961631775,
      "learning_rate": 0.0001606388991935214,
      "loss": 0.2181,
      "step": 446
    },
    {
      "epoch": 0.13219474326272596,
      "grad_norm": 0.34528401494026184,
      "learning_rate": 0.0001604711071866277,
      "loss": 0.2354,
      "step": 447
    },
    {
      "epoch": 0.1324904809434032,
      "grad_norm": 0.3941013216972351,
      "learning_rate": 0.00016030304635144494,
      "loss": 0.2796,
      "step": 448
    },
    {
      "epoch": 0.13278621862408044,
      "grad_norm": 0.36703160405158997,
      "learning_rate": 0.00016013471743509862,
      "loss": 0.2258,
      "step": 449
    },
    {
      "epoch": 0.1330819563047577,
      "grad_norm": 0.44615501165390015,
      "learning_rate": 0.00015996612118590603,
      "loss": 0.1753,
      "step": 450
    },
    {
      "epoch": 0.13337769398543492,
      "grad_norm": 0.09873225539922714,
      "learning_rate": 0.00015979725835337294,
      "loss": 0.1385,
      "step": 451
    },
    {
      "epoch": 0.13367343166611215,
      "grad_norm": 0.13112150132656097,
      "learning_rate": 0.00015962812968819016,
      "loss": 0.2112,
      "step": 452
    },
    {
      "epoch": 0.1339691693467894,
      "grad_norm": 0.1437048316001892,
      "learning_rate": 0.0001594587359422303,
      "loss": 0.1976,
      "step": 453
    },
    {
      "epoch": 0.13426490702746663,
      "grad_norm": 0.14785504341125488,
      "learning_rate": 0.0001592890778685444,
      "loss": 0.1953,
      "step": 454
    },
    {
      "epoch": 0.13456064470814388,
      "grad_norm": 0.15337800979614258,
      "learning_rate": 0.00015911915622135862,
      "loss": 0.1938,
      "step": 455
    },
    {
      "epoch": 0.1348563823888211,
      "grad_norm": 0.16650818288326263,
      "learning_rate": 0.00015894897175607086,
      "loss": 0.1915,
      "step": 456
    },
    {
      "epoch": 0.13515212006949837,
      "grad_norm": 0.15892527997493744,
      "learning_rate": 0.00015877852522924732,
      "loss": 0.1972,
      "step": 457
    },
    {
      "epoch": 0.1354478577501756,
      "grad_norm": 0.16916155815124512,
      "learning_rate": 0.00015860781739861928,
      "loss": 0.2193,
      "step": 458
    },
    {
      "epoch": 0.13574359543085282,
      "grad_norm": 0.1601080596446991,
      "learning_rate": 0.00015843684902307962,
      "loss": 0.1921,
      "step": 459
    },
    {
      "epoch": 0.13603933311153008,
      "grad_norm": 0.16756726801395416,
      "learning_rate": 0.00015826562086267956,
      "loss": 0.241,
      "step": 460
    },
    {
      "epoch": 0.1363350707922073,
      "grad_norm": 0.16822628676891327,
      "learning_rate": 0.00015809413367862512,
      "loss": 0.2028,
      "step": 461
    },
    {
      "epoch": 0.13663080847288456,
      "grad_norm": 0.15600813925266266,
      "learning_rate": 0.00015792238823327388,
      "loss": 0.1639,
      "step": 462
    },
    {
      "epoch": 0.13692654615356178,
      "grad_norm": 0.16066166758537292,
      "learning_rate": 0.00015775038529013152,
      "loss": 0.197,
      "step": 463
    },
    {
      "epoch": 0.13722228383423904,
      "grad_norm": 0.181250661611557,
      "learning_rate": 0.0001575781256138485,
      "loss": 0.2464,
      "step": 464
    },
    {
      "epoch": 0.13751802151491627,
      "grad_norm": 0.19740843772888184,
      "learning_rate": 0.00015740560997021648,
      "loss": 0.2688,
      "step": 465
    },
    {
      "epoch": 0.13781375919559352,
      "grad_norm": 0.178134024143219,
      "learning_rate": 0.00015723283912616513,
      "loss": 0.2315,
      "step": 466
    },
    {
      "epoch": 0.13810949687627075,
      "grad_norm": 0.16021274030208588,
      "learning_rate": 0.00015705981384975866,
      "loss": 0.1819,
      "step": 467
    },
    {
      "epoch": 0.13840523455694798,
      "grad_norm": 0.20027248561382294,
      "learning_rate": 0.0001568865349101923,
      "loss": 0.2789,
      "step": 468
    },
    {
      "epoch": 0.13870097223762523,
      "grad_norm": 0.1927415281534195,
      "learning_rate": 0.00015671300307778898,
      "loss": 0.2,
      "step": 469
    },
    {
      "epoch": 0.13899670991830246,
      "grad_norm": 0.1823057234287262,
      "learning_rate": 0.00015653921912399589,
      "loss": 0.1549,
      "step": 470
    },
    {
      "epoch": 0.1392924475989797,
      "grad_norm": 0.19187289476394653,
      "learning_rate": 0.00015636518382138107,
      "loss": 0.1975,
      "step": 471
    },
    {
      "epoch": 0.13958818527965694,
      "grad_norm": 0.21980108320713043,
      "learning_rate": 0.0001561908979436299,
      "loss": 0.2242,
      "step": 472
    },
    {
      "epoch": 0.1398839229603342,
      "grad_norm": 0.21112766861915588,
      "learning_rate": 0.00015601636226554168,
      "loss": 0.2254,
      "step": 473
    },
    {
      "epoch": 0.14017966064101142,
      "grad_norm": 0.1739622801542282,
      "learning_rate": 0.00015584157756302634,
      "loss": 0.1647,
      "step": 474
    },
    {
      "epoch": 0.14047539832168865,
      "grad_norm": 0.2054637223482132,
      "learning_rate": 0.0001556665446131007,
      "loss": 0.1974,
      "step": 475
    },
    {
      "epoch": 0.1407711360023659,
      "grad_norm": 0.2365804761648178,
      "learning_rate": 0.00015549126419388536,
      "loss": 0.2244,
      "step": 476
    },
    {
      "epoch": 0.14106687368304313,
      "grad_norm": 0.20265153050422668,
      "learning_rate": 0.0001553157370846009,
      "loss": 0.1727,
      "step": 477
    },
    {
      "epoch": 0.1413626113637204,
      "grad_norm": 0.1899023801088333,
      "learning_rate": 0.00015513996406556465,
      "loss": 0.1573,
      "step": 478
    },
    {
      "epoch": 0.1416583490443976,
      "grad_norm": 0.25980237126350403,
      "learning_rate": 0.00015496394591818716,
      "loss": 0.2644,
      "step": 479
    },
    {
      "epoch": 0.14195408672507487,
      "grad_norm": 0.20405927300453186,
      "learning_rate": 0.0001547876834249687,
      "loss": 0.1915,
      "step": 480
    },
    {
      "epoch": 0.1422498244057521,
      "grad_norm": 0.22452029585838318,
      "learning_rate": 0.00015461117736949577,
      "loss": 0.2066,
      "step": 481
    },
    {
      "epoch": 0.14254556208642935,
      "grad_norm": 0.25543588399887085,
      "learning_rate": 0.00015443442853643762,
      "loss": 0.2462,
      "step": 482
    },
    {
      "epoch": 0.14284129976710658,
      "grad_norm": 0.26937565207481384,
      "learning_rate": 0.00015425743771154294,
      "loss": 0.2779,
      "step": 483
    },
    {
      "epoch": 0.1431370374477838,
      "grad_norm": 0.24261003732681274,
      "learning_rate": 0.00015408020568163602,
      "loss": 0.1922,
      "step": 484
    },
    {
      "epoch": 0.14343277512846106,
      "grad_norm": 0.23764987289905548,
      "learning_rate": 0.00015390273323461352,
      "loss": 0.1766,
      "step": 485
    },
    {
      "epoch": 0.1437285128091383,
      "grad_norm": 0.26802945137023926,
      "learning_rate": 0.0001537250211594409,
      "loss": 0.2005,
      "step": 486
    },
    {
      "epoch": 0.14402425048981554,
      "grad_norm": 0.2354622781276703,
      "learning_rate": 0.0001535470702461489,
      "loss": 0.1849,
      "step": 487
    },
    {
      "epoch": 0.14431998817049277,
      "grad_norm": 0.2462967336177826,
      "learning_rate": 0.00015336888128583,
      "loss": 0.1995,
      "step": 488
    },
    {
      "epoch": 0.14461572585117002,
      "grad_norm": 0.2529776394367218,
      "learning_rate": 0.000153190455070635,
      "loss": 0.1994,
      "step": 489
    },
    {
      "epoch": 0.14491146353184725,
      "grad_norm": 0.27993232011795044,
      "learning_rate": 0.00015301179239376938,
      "loss": 0.2212,
      "step": 490
    },
    {
      "epoch": 0.14520720121252448,
      "grad_norm": 0.2503523528575897,
      "learning_rate": 0.00015283289404948976,
      "loss": 0.1711,
      "step": 491
    },
    {
      "epoch": 0.14550293889320173,
      "grad_norm": 0.2650829255580902,
      "learning_rate": 0.0001526537608331006,
      "loss": 0.1991,
      "step": 492
    },
    {
      "epoch": 0.14579867657387896,
      "grad_norm": 0.2573229670524597,
      "learning_rate": 0.00015247439354095041,
      "loss": 0.1629,
      "step": 493
    },
    {
      "epoch": 0.14609441425455622,
      "grad_norm": 0.2993515431880951,
      "learning_rate": 0.00015229479297042823,
      "loss": 0.2513,
      "step": 494
    },
    {
      "epoch": 0.14639015193523344,
      "grad_norm": 0.2918243706226349,
      "learning_rate": 0.00015211495991996027,
      "loss": 0.1799,
      "step": 495
    },
    {
      "epoch": 0.1466858896159107,
      "grad_norm": 0.32243257761001587,
      "learning_rate": 0.0001519348951890062,
      "loss": 0.2384,
      "step": 496
    },
    {
      "epoch": 0.14698162729658792,
      "grad_norm": 0.32650378346443176,
      "learning_rate": 0.0001517545995780556,
      "loss": 0.2419,
      "step": 497
    },
    {
      "epoch": 0.14727736497726518,
      "grad_norm": 0.28698667883872986,
      "learning_rate": 0.00015157407388862452,
      "loss": 0.1575,
      "step": 498
    },
    {
      "epoch": 0.1475731026579424,
      "grad_norm": 0.5042946934700012,
      "learning_rate": 0.00015139331892325179,
      "loss": 0.3989,
      "step": 499
    },
    {
      "epoch": 0.14786884033861963,
      "grad_norm": 0.6159984469413757,
      "learning_rate": 0.0001512123354854955,
      "loss": 0.3383,
      "step": 500
    },
    {
      "epoch": 0.1481645780192969,
      "grad_norm": 0.12425012141466141,
      "learning_rate": 0.0001510311243799295,
      "loss": 0.1775,
      "step": 501
    },
    {
      "epoch": 0.14846031569997412,
      "grad_norm": 0.12108102440834045,
      "learning_rate": 0.00015084968641213958,
      "loss": 0.1404,
      "step": 502
    },
    {
      "epoch": 0.14875605338065137,
      "grad_norm": 0.12233590334653854,
      "learning_rate": 0.00015066802238872023,
      "loss": 0.1668,
      "step": 503
    },
    {
      "epoch": 0.1490517910613286,
      "grad_norm": 0.1354806274175644,
      "learning_rate": 0.0001504861331172709,
      "loss": 0.1827,
      "step": 504
    },
    {
      "epoch": 0.14934752874200585,
      "grad_norm": 0.1641916036605835,
      "learning_rate": 0.0001503040194063922,
      "loss": 0.1819,
      "step": 505
    },
    {
      "epoch": 0.14964326642268308,
      "grad_norm": 0.1794111579656601,
      "learning_rate": 0.00015012168206568268,
      "loss": 0.2923,
      "step": 506
    },
    {
      "epoch": 0.1499390041033603,
      "grad_norm": 0.1664586067199707,
      "learning_rate": 0.00014993912190573505,
      "loss": 0.222,
      "step": 507
    },
    {
      "epoch": 0.15023474178403756,
      "grad_norm": 0.16165000200271606,
      "learning_rate": 0.00014975633973813242,
      "loss": 0.2264,
      "step": 508
    },
    {
      "epoch": 0.1505304794647148,
      "grad_norm": 0.17169785499572754,
      "learning_rate": 0.00014957333637544503,
      "loss": 0.2151,
      "step": 509
    },
    {
      "epoch": 0.15082621714539204,
      "grad_norm": 0.1526673138141632,
      "learning_rate": 0.00014939011263122634,
      "loss": 0.1745,
      "step": 510
    },
    {
      "epoch": 0.15112195482606927,
      "grad_norm": 0.15362168848514557,
      "learning_rate": 0.0001492066693200096,
      "loss": 0.1649,
      "step": 511
    },
    {
      "epoch": 0.15141769250674653,
      "grad_norm": 0.16641461849212646,
      "learning_rate": 0.00014902300725730413,
      "loss": 0.2128,
      "step": 512
    },
    {
      "epoch": 0.15171343018742375,
      "grad_norm": 0.15893928706645966,
      "learning_rate": 0.00014883912725959167,
      "loss": 0.1949,
      "step": 513
    },
    {
      "epoch": 0.15200916786810098,
      "grad_norm": 0.17531882226467133,
      "learning_rate": 0.00014865503014432292,
      "loss": 0.2354,
      "step": 514
    },
    {
      "epoch": 0.15230490554877824,
      "grad_norm": 0.20174053311347961,
      "learning_rate": 0.00014847071672991367,
      "loss": 0.2515,
      "step": 515
    },
    {
      "epoch": 0.15260064322945546,
      "grad_norm": 0.17568667232990265,
      "learning_rate": 0.0001482861878357414,
      "loss": 0.2216,
      "step": 516
    },
    {
      "epoch": 0.15289638091013272,
      "grad_norm": 0.17869949340820312,
      "learning_rate": 0.00014810144428214144,
      "loss": 0.2044,
      "step": 517
    },
    {
      "epoch": 0.15319211859080994,
      "grad_norm": 0.18138998746871948,
      "learning_rate": 0.0001479164868904034,
      "loss": 0.1876,
      "step": 518
    },
    {
      "epoch": 0.1534878562714872,
      "grad_norm": 0.17516851425170898,
      "learning_rate": 0.00014773131648276758,
      "loss": 0.1834,
      "step": 519
    },
    {
      "epoch": 0.15378359395216443,
      "grad_norm": 0.1887725293636322,
      "learning_rate": 0.00014754593388242117,
      "loss": 0.1885,
      "step": 520
    },
    {
      "epoch": 0.15407933163284168,
      "grad_norm": 0.18247954547405243,
      "learning_rate": 0.0001473603399134948,
      "loss": 0.1621,
      "step": 521
    },
    {
      "epoch": 0.1543750693135189,
      "grad_norm": 0.1879930943250656,
      "learning_rate": 0.0001471745354010586,
      "loss": 0.2155,
      "step": 522
    },
    {
      "epoch": 0.15467080699419614,
      "grad_norm": 0.16984637081623077,
      "learning_rate": 0.00014698852117111884,
      "loss": 0.1724,
      "step": 523
    },
    {
      "epoch": 0.1549665446748734,
      "grad_norm": 0.1831916868686676,
      "learning_rate": 0.000146802298050614,
      "loss": 0.157,
      "step": 524
    },
    {
      "epoch": 0.15526228235555062,
      "grad_norm": 0.20019155740737915,
      "learning_rate": 0.0001466158668674112,
      "loss": 0.1544,
      "step": 525
    },
    {
      "epoch": 0.15555802003622787,
      "grad_norm": 0.1930214762687683,
      "learning_rate": 0.00014642922845030257,
      "loss": 0.1891,
      "step": 526
    },
    {
      "epoch": 0.1558537577169051,
      "grad_norm": 0.26559117436408997,
      "learning_rate": 0.0001462423836290015,
      "loss": 0.289,
      "step": 527
    },
    {
      "epoch": 0.15614949539758236,
      "grad_norm": 0.2079695165157318,
      "learning_rate": 0.00014605533323413887,
      "loss": 0.1474,
      "step": 528
    },
    {
      "epoch": 0.15644523307825958,
      "grad_norm": 0.23225319385528564,
      "learning_rate": 0.00014586807809725962,
      "loss": 0.177,
      "step": 529
    },
    {
      "epoch": 0.1567409707589368,
      "grad_norm": 0.2178119421005249,
      "learning_rate": 0.00014568061905081875,
      "loss": 0.2066,
      "step": 530
    },
    {
      "epoch": 0.15703670843961406,
      "grad_norm": 0.2345496565103531,
      "learning_rate": 0.00014549295692817778,
      "loss": 0.2144,
      "step": 531
    },
    {
      "epoch": 0.1573324461202913,
      "grad_norm": 0.24489574134349823,
      "learning_rate": 0.00014530509256360102,
      "loss": 0.2046,
      "step": 532
    },
    {
      "epoch": 0.15762818380096855,
      "grad_norm": 0.23358239233493805,
      "learning_rate": 0.00014511702679225193,
      "loss": 0.2251,
      "step": 533
    },
    {
      "epoch": 0.15792392148164577,
      "grad_norm": 0.26981320977211,
      "learning_rate": 0.0001449287604501893,
      "loss": 0.2222,
      "step": 534
    },
    {
      "epoch": 0.15821965916232303,
      "grad_norm": 0.24023444950580597,
      "learning_rate": 0.00014474029437436348,
      "loss": 0.2155,
      "step": 535
    },
    {
      "epoch": 0.15851539684300026,
      "grad_norm": 0.19078750908374786,
      "learning_rate": 0.00014455162940261285,
      "loss": 0.1276,
      "step": 536
    },
    {
      "epoch": 0.1588111345236775,
      "grad_norm": 0.22319287061691284,
      "learning_rate": 0.0001443627663736599,
      "loss": 0.2082,
      "step": 537
    },
    {
      "epoch": 0.15910687220435474,
      "grad_norm": 0.24529913067817688,
      "learning_rate": 0.00014417370612710778,
      "loss": 0.177,
      "step": 538
    },
    {
      "epoch": 0.15940260988503197,
      "grad_norm": 0.26183465123176575,
      "learning_rate": 0.00014398444950343623,
      "loss": 0.1725,
      "step": 539
    },
    {
      "epoch": 0.15969834756570922,
      "grad_norm": 0.2594473361968994,
      "learning_rate": 0.00014379499734399798,
      "loss": 0.2477,
      "step": 540
    },
    {
      "epoch": 0.15999408524638645,
      "grad_norm": 0.2427864521741867,
      "learning_rate": 0.0001436053504910151,
      "loss": 0.1737,
      "step": 541
    },
    {
      "epoch": 0.1602898229270637,
      "grad_norm": 0.30450040102005005,
      "learning_rate": 0.0001434155097875752,
      "loss": 0.2318,
      "step": 542
    },
    {
      "epoch": 0.16058556060774093,
      "grad_norm": 0.3049463927745819,
      "learning_rate": 0.00014322547607762762,
      "loss": 0.2495,
      "step": 543
    },
    {
      "epoch": 0.16088129828841818,
      "grad_norm": 0.27216774225234985,
      "learning_rate": 0.0001430352502059797,
      "loss": 0.1733,
      "step": 544
    },
    {
      "epoch": 0.1611770359690954,
      "grad_norm": 0.3504975140094757,
      "learning_rate": 0.0001428448330182931,
      "loss": 0.209,
      "step": 545
    },
    {
      "epoch": 0.16147277364977264,
      "grad_norm": 0.3599260151386261,
      "learning_rate": 0.00014265422536107993,
      "loss": 0.2185,
      "step": 546
    },
    {
      "epoch": 0.1617685113304499,
      "grad_norm": 0.31020569801330566,
      "learning_rate": 0.00014246342808169914,
      "loss": 0.2292,
      "step": 547
    },
    {
      "epoch": 0.16206424901112712,
      "grad_norm": 0.3614603281021118,
      "learning_rate": 0.00014227244202835257,
      "loss": 0.1979,
      "step": 548
    },
    {
      "epoch": 0.16235998669180438,
      "grad_norm": 0.44161731004714966,
      "learning_rate": 0.0001420812680500813,
      "loss": 0.1933,
      "step": 549
    },
    {
      "epoch": 0.1626557243724816,
      "grad_norm": 0.6861674189567566,
      "learning_rate": 0.00014188990699676184,
      "loss": 0.3917,
      "step": 550
    },
    {
      "epoch": 0.16295146205315886,
      "grad_norm": 0.10810007899999619,
      "learning_rate": 0.00014169835971910238,
      "loss": 0.1231,
      "step": 551
    },
    {
      "epoch": 0.16324719973383608,
      "grad_norm": 0.1477546989917755,
      "learning_rate": 0.0001415066270686389,
      "loss": 0.1923,
      "step": 552
    },
    {
      "epoch": 0.16354293741451334,
      "grad_norm": 0.14868442714214325,
      "learning_rate": 0.00014131470989773158,
      "loss": 0.1956,
      "step": 553
    },
    {
      "epoch": 0.16383867509519057,
      "grad_norm": 0.1274333894252777,
      "learning_rate": 0.0001411226090595608,
      "loss": 0.1461,
      "step": 554
    },
    {
      "epoch": 0.1641344127758678,
      "grad_norm": 0.1790267527103424,
      "learning_rate": 0.00014093032540812348,
      "loss": 0.2404,
      "step": 555
    },
    {
      "epoch": 0.16443015045654505,
      "grad_norm": 0.1788930892944336,
      "learning_rate": 0.0001407378597982293,
      "loss": 0.2194,
      "step": 556
    },
    {
      "epoch": 0.16472588813722228,
      "grad_norm": 0.20187751948833466,
      "learning_rate": 0.00014054521308549673,
      "loss": 0.2107,
      "step": 557
    },
    {
      "epoch": 0.16502162581789953,
      "grad_norm": 0.16323895752429962,
      "learning_rate": 0.0001403523861263495,
      "loss": 0.1895,
      "step": 558
    },
    {
      "epoch": 0.16531736349857676,
      "grad_norm": 0.14459684491157532,
      "learning_rate": 0.00014015937977801256,
      "loss": 0.1856,
      "step": 559
    },
    {
      "epoch": 0.165613101179254,
      "grad_norm": 0.16795897483825684,
      "learning_rate": 0.00013996619489850822,
      "loss": 0.2049,
      "step": 560
    },
    {
      "epoch": 0.16590883885993124,
      "grad_norm": 0.16099309921264648,
      "learning_rate": 0.00013977283234665273,
      "loss": 0.2114,
      "step": 561
    },
    {
      "epoch": 0.16620457654060847,
      "grad_norm": 0.1711655557155609,
      "learning_rate": 0.00013957929298205195,
      "loss": 0.1889,
      "step": 562
    },
    {
      "epoch": 0.16650031422128572,
      "grad_norm": 0.20089977979660034,
      "learning_rate": 0.00013938557766509792,
      "loss": 0.2932,
      "step": 563
    },
    {
      "epoch": 0.16679605190196295,
      "grad_norm": 0.18003849685192108,
      "learning_rate": 0.0001391916872569648,
      "loss": 0.1676,
      "step": 564
    },
    {
      "epoch": 0.1670917895826402,
      "grad_norm": 0.19262616336345673,
      "learning_rate": 0.00013899762261960518,
      "loss": 0.1983,
      "step": 565
    },
    {
      "epoch": 0.16738752726331743,
      "grad_norm": 0.17830273509025574,
      "learning_rate": 0.0001388033846157462,
      "loss": 0.1977,
      "step": 566
    },
    {
      "epoch": 0.1676832649439947,
      "grad_norm": 0.18847207725048065,
      "learning_rate": 0.0001386089741088857,
      "loss": 0.2294,
      "step": 567
    },
    {
      "epoch": 0.1679790026246719,
      "grad_norm": 0.18042877316474915,
      "learning_rate": 0.00013841439196328836,
      "loss": 0.1961,
      "step": 568
    },
    {
      "epoch": 0.16827474030534917,
      "grad_norm": 0.18396486341953278,
      "learning_rate": 0.00013821963904398193,
      "loss": 0.1872,
      "step": 569
    },
    {
      "epoch": 0.1685704779860264,
      "grad_norm": 0.2012127786874771,
      "learning_rate": 0.00013802471621675338,
      "loss": 0.194,
      "step": 570
    },
    {
      "epoch": 0.16886621566670362,
      "grad_norm": 0.18555906414985657,
      "learning_rate": 0.00013782962434814492,
      "loss": 0.1971,
      "step": 571
    },
    {
      "epoch": 0.16916195334738088,
      "grad_norm": 0.1829305738210678,
      "learning_rate": 0.00013763436430545034,
      "loss": 0.1687,
      "step": 572
    },
    {
      "epoch": 0.1694576910280581,
      "grad_norm": 0.18491117656230927,
      "learning_rate": 0.00013743893695671096,
      "loss": 0.1691,
      "step": 573
    },
    {
      "epoch": 0.16975342870873536,
      "grad_norm": 0.1929929107427597,
      "learning_rate": 0.00013724334317071198,
      "loss": 0.2112,
      "step": 574
    },
    {
      "epoch": 0.1700491663894126,
      "grad_norm": 0.21051141619682312,
      "learning_rate": 0.00013704758381697844,
      "loss": 0.1932,
      "step": 575
    },
    {
      "epoch": 0.17034490407008984,
      "grad_norm": 0.23053379356861115,
      "learning_rate": 0.00013685165976577146,
      "loss": 0.2552,
      "step": 576
    },
    {
      "epoch": 0.17064064175076707,
      "grad_norm": 0.2063068449497223,
      "learning_rate": 0.0001366555718880843,
      "loss": 0.1479,
      "step": 577
    },
    {
      "epoch": 0.1709363794314443,
      "grad_norm": 0.21455182135105133,
      "learning_rate": 0.00013645932105563844,
      "loss": 0.204,
      "step": 578
    },
    {
      "epoch": 0.17123211711212155,
      "grad_norm": 0.23406703770160675,
      "learning_rate": 0.00013626290814088005,
      "loss": 0.2021,
      "step": 579
    },
    {
      "epoch": 0.17152785479279878,
      "grad_norm": 0.2362387478351593,
      "learning_rate": 0.00013606633401697557,
      "loss": 0.2392,
      "step": 580
    },
    {
      "epoch": 0.17182359247347603,
      "grad_norm": 0.19675102829933167,
      "learning_rate": 0.00013586959955780824,
      "loss": 0.1491,
      "step": 581
    },
    {
      "epoch": 0.17211933015415326,
      "grad_norm": 0.25349652767181396,
      "learning_rate": 0.00013567270563797398,
      "loss": 0.2077,
      "step": 582
    },
    {
      "epoch": 0.17241506783483052,
      "grad_norm": 0.24108415842056274,
      "learning_rate": 0.00013547565313277776,
      "loss": 0.2037,
      "step": 583
    },
    {
      "epoch": 0.17271080551550774,
      "grad_norm": 0.2342023402452469,
      "learning_rate": 0.00013527844291822948,
      "loss": 0.1752,
      "step": 584
    },
    {
      "epoch": 0.173006543196185,
      "grad_norm": 0.21363839507102966,
      "learning_rate": 0.0001350810758710401,
      "loss": 0.1707,
      "step": 585
    },
    {
      "epoch": 0.17330228087686222,
      "grad_norm": 0.25588366389274597,
      "learning_rate": 0.00013488355286861783,
      "loss": 0.2348,
      "step": 586
    },
    {
      "epoch": 0.17359801855753945,
      "grad_norm": 0.26161813735961914,
      "learning_rate": 0.0001346858747890642,
      "loss": 0.2326,
      "step": 587
    },
    {
      "epoch": 0.1738937562382167,
      "grad_norm": 0.26976850628852844,
      "learning_rate": 0.00013448804251117003,
      "loss": 0.2392,
      "step": 588
    },
    {
      "epoch": 0.17418949391889393,
      "grad_norm": 0.25815528631210327,
      "learning_rate": 0.0001342900569144119,
      "loss": 0.2083,
      "step": 589
    },
    {
      "epoch": 0.1744852315995712,
      "grad_norm": 0.2357477843761444,
      "learning_rate": 0.0001340919188789477,
      "loss": 0.1978,
      "step": 590
    },
    {
      "epoch": 0.17478096928024842,
      "grad_norm": 0.23640482127666473,
      "learning_rate": 0.00013389362928561317,
      "loss": 0.1737,
      "step": 591
    },
    {
      "epoch": 0.17507670696092567,
      "grad_norm": 0.2584919035434723,
      "learning_rate": 0.00013369518901591772,
      "loss": 0.18,
      "step": 592
    },
    {
      "epoch": 0.1753724446416029,
      "grad_norm": 0.2640099823474884,
      "learning_rate": 0.00013349659895204067,
      "loss": 0.1936,
      "step": 593
    },
    {
      "epoch": 0.17566818232228013,
      "grad_norm": 0.30272793769836426,
      "learning_rate": 0.0001332978599768272,
      "loss": 0.2426,
      "step": 594
    },
    {
      "epoch": 0.17596392000295738,
      "grad_norm": 0.29123789072036743,
      "learning_rate": 0.00013309897297378455,
      "loss": 0.1799,
      "step": 595
    },
    {
      "epoch": 0.1762596576836346,
      "grad_norm": 0.31216293573379517,
      "learning_rate": 0.00013289993882707797,
      "loss": 0.2074,
      "step": 596
    },
    {
      "epoch": 0.17655539536431186,
      "grad_norm": 0.32396575808525085,
      "learning_rate": 0.00013270075842152678,
      "loss": 0.2328,
      "step": 597
    },
    {
      "epoch": 0.1768511330449891,
      "grad_norm": 0.3656167984008789,
      "learning_rate": 0.00013250143264260074,
      "loss": 0.2458,
      "step": 598
    },
    {
      "epoch": 0.17714687072566634,
      "grad_norm": 0.4183952212333679,
      "learning_rate": 0.0001323019623764156,
      "loss": 0.2968,
      "step": 599
    },
    {
      "epoch": 0.17744260840634357,
      "grad_norm": 0.563667356967926,
      "learning_rate": 0.00013210234850972964,
      "loss": 0.3082,
      "step": 600
    },
    {
      "epoch": 0.1777383460870208,
      "grad_norm": 0.11673900485038757,
      "learning_rate": 0.0001319025919299394,
      "loss": 0.1225,
      "step": 601
    },
    {
      "epoch": 0.17803408376769805,
      "grad_norm": 0.15502867102622986,
      "learning_rate": 0.00013170269352507597,
      "loss": 0.1869,
      "step": 602
    },
    {
      "epoch": 0.17832982144837528,
      "grad_norm": 0.1288733333349228,
      "learning_rate": 0.0001315026541838008,
      "loss": 0.1496,
      "step": 603
    },
    {
      "epoch": 0.17862555912905254,
      "grad_norm": 0.16827955842018127,
      "learning_rate": 0.00013130247479540202,
      "loss": 0.2337,
      "step": 604
    },
    {
      "epoch": 0.17892129680972976,
      "grad_norm": 0.1500786542892456,
      "learning_rate": 0.00013110215624979025,
      "loss": 0.1962,
      "step": 605
    },
    {
      "epoch": 0.17921703449040702,
      "grad_norm": 0.16338172554969788,
      "learning_rate": 0.00013090169943749476,
      "loss": 0.2002,
      "step": 606
    },
    {
      "epoch": 0.17951277217108424,
      "grad_norm": 0.17155061662197113,
      "learning_rate": 0.00013070110524965954,
      "loss": 0.2127,
      "step": 607
    },
    {
      "epoch": 0.1798085098517615,
      "grad_norm": 0.1854557991027832,
      "learning_rate": 0.00013050037457803924,
      "loss": 0.2304,
      "step": 608
    },
    {
      "epoch": 0.18010424753243873,
      "grad_norm": 0.16282689571380615,
      "learning_rate": 0.0001302995083149953,
      "loss": 0.1854,
      "step": 609
    },
    {
      "epoch": 0.18039998521311595,
      "grad_norm": 0.1648624688386917,
      "learning_rate": 0.0001300985073534919,
      "loss": 0.2144,
      "step": 610
    },
    {
      "epoch": 0.1806957228937932,
      "grad_norm": 0.15584151446819305,
      "learning_rate": 0.00012989737258709203,
      "loss": 0.1721,
      "step": 611
    },
    {
      "epoch": 0.18099146057447044,
      "grad_norm": 0.16049699485301971,
      "learning_rate": 0.00012969610490995358,
      "loss": 0.165,
      "step": 612
    },
    {
      "epoch": 0.1812871982551477,
      "grad_norm": 0.17223642766475677,
      "learning_rate": 0.00012949470521682528,
      "loss": 0.1902,
      "step": 613
    },
    {
      "epoch": 0.18158293593582492,
      "grad_norm": 0.18705599009990692,
      "learning_rate": 0.0001292931744030427,
      "loss": 0.2414,
      "step": 614
    },
    {
      "epoch": 0.18187867361650217,
      "grad_norm": 0.21196113526821136,
      "learning_rate": 0.0001290915133645243,
      "loss": 0.2609,
      "step": 615
    },
    {
      "epoch": 0.1821744112971794,
      "grad_norm": 0.1778162270784378,
      "learning_rate": 0.00012888972299776754,
      "loss": 0.1849,
      "step": 616
    },
    {
      "epoch": 0.18247014897785663,
      "grad_norm": 0.17998971045017242,
      "learning_rate": 0.00012868780419984482,
      "loss": 0.1704,
      "step": 617
    },
    {
      "epoch": 0.18276588665853388,
      "grad_norm": 0.19640791416168213,
      "learning_rate": 0.00012848575786839943,
      "loss": 0.2088,
      "step": 618
    },
    {
      "epoch": 0.1830616243392111,
      "grad_norm": 0.19208578765392303,
      "learning_rate": 0.0001282835849016416,
      "loss": 0.1754,
      "step": 619
    },
    {
      "epoch": 0.18335736201988836,
      "grad_norm": 0.1946994960308075,
      "learning_rate": 0.00012808128619834461,
      "loss": 0.1852,
      "step": 620
    },
    {
      "epoch": 0.1836530997005656,
      "grad_norm": 0.1998986154794693,
      "learning_rate": 0.0001278788626578407,
      "loss": 0.2074,
      "step": 621
    },
    {
      "epoch": 0.18394883738124285,
      "grad_norm": 0.1936517357826233,
      "learning_rate": 0.00012767631518001698,
      "loss": 0.1938,
      "step": 622
    },
    {
      "epoch": 0.18424457506192007,
      "grad_norm": 0.2083950638771057,
      "learning_rate": 0.00012747364466531163,
      "loss": 0.1993,
      "step": 623
    },
    {
      "epoch": 0.18454031274259733,
      "grad_norm": 0.20452117919921875,
      "learning_rate": 0.00012727085201470973,
      "loss": 0.1721,
      "step": 624
    },
    {
      "epoch": 0.18483605042327456,
      "grad_norm": 0.23634770512580872,
      "learning_rate": 0.00012706793812973941,
      "loss": 0.2541,
      "step": 625
    },
    {
      "epoch": 0.18513178810395178,
      "grad_norm": 0.24558795988559723,
      "learning_rate": 0.0001268649039124677,
      "loss": 0.2825,
      "step": 626
    },
    {
      "epoch": 0.18542752578462904,
      "grad_norm": 0.2423345446586609,
      "learning_rate": 0.00012666175026549662,
      "loss": 0.2449,
      "step": 627
    },
    {
      "epoch": 0.18572326346530627,
      "grad_norm": 0.20482169091701508,
      "learning_rate": 0.000126458478091959,
      "loss": 0.2034,
      "step": 628
    },
    {
      "epoch": 0.18601900114598352,
      "grad_norm": 0.2271885871887207,
      "learning_rate": 0.00012625508829551473,
      "loss": 0.2243,
      "step": 629
    },
    {
      "epoch": 0.18631473882666075,
      "grad_norm": 0.2068728655576706,
      "learning_rate": 0.00012605158178034654,
      "loss": 0.1797,
      "step": 630
    },
    {
      "epoch": 0.186610476507338,
      "grad_norm": 0.2342994064092636,
      "learning_rate": 0.00012584795945115603,
      "loss": 0.2005,
      "step": 631
    },
    {
      "epoch": 0.18690621418801523,
      "grad_norm": 0.22803792357444763,
      "learning_rate": 0.0001256442222131597,
      "loss": 0.2167,
      "step": 632
    },
    {
      "epoch": 0.18720195186869246,
      "grad_norm": 0.23853746056556702,
      "learning_rate": 0.0001254403709720848,
      "loss": 0.2077,
      "step": 633
    },
    {
      "epoch": 0.1874976895493697,
      "grad_norm": 0.2190757393836975,
      "learning_rate": 0.0001252364066341655,
      "loss": 0.1949,
      "step": 634
    },
    {
      "epoch": 0.18779342723004694,
      "grad_norm": 0.22064319252967834,
      "learning_rate": 0.00012503233010613865,
      "loss": 0.1563,
      "step": 635
    },
    {
      "epoch": 0.1880891649107242,
      "grad_norm": 0.22475862503051758,
      "learning_rate": 0.00012482814229523997,
      "loss": 0.164,
      "step": 636
    },
    {
      "epoch": 0.18838490259140142,
      "grad_norm": 0.23672211170196533,
      "learning_rate": 0.00012462384410919975,
      "loss": 0.1975,
      "step": 637
    },
    {
      "epoch": 0.18868064027207868,
      "grad_norm": 0.22147880494594574,
      "learning_rate": 0.00012441943645623903,
      "loss": 0.1799,
      "step": 638
    },
    {
      "epoch": 0.1889763779527559,
      "grad_norm": 0.23822712898254395,
      "learning_rate": 0.00012421492024506555,
      "loss": 0.197,
      "step": 639
    },
    {
      "epoch": 0.18927211563343316,
      "grad_norm": 0.2708362638950348,
      "learning_rate": 0.00012401029638486953,
      "loss": 0.2175,
      "step": 640
    },
    {
      "epoch": 0.18956785331411038,
      "grad_norm": 0.23913449048995972,
      "learning_rate": 0.0001238055657853198,
      "loss": 0.1567,
      "step": 641
    },
    {
      "epoch": 0.1898635909947876,
      "grad_norm": 0.2158363163471222,
      "learning_rate": 0.00012360072935655982,
      "loss": 0.1398,
      "step": 642
    },
    {
      "epoch": 0.19015932867546487,
      "grad_norm": 0.2530500888824463,
      "learning_rate": 0.00012339578800920332,
      "loss": 0.1711,
      "step": 643
    },
    {
      "epoch": 0.1904550663561421,
      "grad_norm": 0.2804064154624939,
      "learning_rate": 0.00012319074265433063,
      "loss": 0.1629,
      "step": 644
    },
    {
      "epoch": 0.19075080403681935,
      "grad_norm": 0.38679811358451843,
      "learning_rate": 0.00012298559420348437,
      "loss": 0.2072,
      "step": 645
    },
    {
      "epoch": 0.19104654171749658,
      "grad_norm": 0.25860467553138733,
      "learning_rate": 0.00012278034356866545,
      "loss": 0.1584,
      "step": 646
    },
    {
      "epoch": 0.19134227939817383,
      "grad_norm": 0.42005908489227295,
      "learning_rate": 0.00012257499166232907,
      "loss": 0.2906,
      "step": 647
    },
    {
      "epoch": 0.19163801707885106,
      "grad_norm": 0.3620118498802185,
      "learning_rate": 0.0001223695393973807,
      "loss": 0.2341,
      "step": 648
    },
    {
      "epoch": 0.19193375475952829,
      "grad_norm": 0.43819329142570496,
      "learning_rate": 0.0001221639876871719,
      "loss": 0.237,
      "step": 649
    },
    {
      "epoch": 0.19222949244020554,
      "grad_norm": 0.6452730894088745,
      "learning_rate": 0.0001219583374454963,
      "loss": 0.2783,
      "step": 650
    },
    {
      "epoch": 0.19252523012088277,
      "grad_norm": 0.13952232897281647,
      "learning_rate": 0.00012175258958658564,
      "loss": 0.1729,
      "step": 651
    },
    {
      "epoch": 0.19282096780156002,
      "grad_norm": 0.16192123293876648,
      "learning_rate": 0.00012154674502510555,
      "loss": 0.2102,
      "step": 652
    },
    {
      "epoch": 0.19311670548223725,
      "grad_norm": 0.1605851650238037,
      "learning_rate": 0.00012134080467615159,
      "loss": 0.1727,
      "step": 653
    },
    {
      "epoch": 0.1934124431629145,
      "grad_norm": 0.1601710319519043,
      "learning_rate": 0.00012113476945524513,
      "loss": 0.1528,
      "step": 654
    },
    {
      "epoch": 0.19370818084359173,
      "grad_norm": 0.1721370667219162,
      "learning_rate": 0.00012092864027832933,
      "loss": 0.1819,
      "step": 655
    },
    {
      "epoch": 0.194003918524269,
      "grad_norm": 0.16613177955150604,
      "learning_rate": 0.000120722418061765,
      "loss": 0.2218,
      "step": 656
    },
    {
      "epoch": 0.19429965620494621,
      "grad_norm": 0.17850197851657867,
      "learning_rate": 0.0001205161037223266,
      "loss": 0.2397,
      "step": 657
    },
    {
      "epoch": 0.19459539388562344,
      "grad_norm": 0.18693998456001282,
      "learning_rate": 0.00012030969817719808,
      "loss": 0.238,
      "step": 658
    },
    {
      "epoch": 0.1948911315663007,
      "grad_norm": 0.17822137475013733,
      "learning_rate": 0.00012010320234396894,
      "loss": 0.1988,
      "step": 659
    },
    {
      "epoch": 0.19518686924697792,
      "grad_norm": 0.23123416304588318,
      "learning_rate": 0.00011989661714062999,
      "loss": 0.2258,
      "step": 660
    },
    {
      "epoch": 0.19548260692765518,
      "grad_norm": 0.1795627623796463,
      "learning_rate": 0.0001196899434855693,
      "loss": 0.1911,
      "step": 661
    },
    {
      "epoch": 0.1957783446083324,
      "grad_norm": 0.1633031815290451,
      "learning_rate": 0.00011948318229756827,
      "loss": 0.1737,
      "step": 662
    },
    {
      "epoch": 0.19607408228900966,
      "grad_norm": 0.1819830685853958,
      "learning_rate": 0.00011927633449579735,
      "loss": 0.179,
      "step": 663
    },
    {
      "epoch": 0.1963698199696869,
      "grad_norm": 0.18557539582252502,
      "learning_rate": 0.0001190694009998121,
      "loss": 0.2491,
      "step": 664
    },
    {
      "epoch": 0.19666555765036411,
      "grad_norm": 0.19799882173538208,
      "learning_rate": 0.00011886238272954897,
      "loss": 0.2407,
      "step": 665
    },
    {
      "epoch": 0.19696129533104137,
      "grad_norm": 0.19097891449928284,
      "learning_rate": 0.00011865528060532127,
      "loss": 0.2252,
      "step": 666
    },
    {
      "epoch": 0.1972570330117186,
      "grad_norm": 0.170382559299469,
      "learning_rate": 0.0001184480955478152,
      "loss": 0.1763,
      "step": 667
    },
    {
      "epoch": 0.19755277069239585,
      "grad_norm": 0.18122434616088867,
      "learning_rate": 0.00011824082847808558,
      "loss": 0.1891,
      "step": 668
    },
    {
      "epoch": 0.19784850837307308,
      "grad_norm": 0.20141829550266266,
      "learning_rate": 0.00011803348031755179,
      "loss": 0.1939,
      "step": 669
    },
    {
      "epoch": 0.19814424605375033,
      "grad_norm": 0.19852298498153687,
      "learning_rate": 0.0001178260519879937,
      "loss": 0.2078,
      "step": 670
    },
    {
      "epoch": 0.19843998373442756,
      "grad_norm": 0.18741367757320404,
      "learning_rate": 0.00011761854441154767,
      "loss": 0.1668,
      "step": 671
    },
    {
      "epoch": 0.19873572141510482,
      "grad_norm": 0.18189528584480286,
      "learning_rate": 0.00011741095851070228,
      "loss": 0.1611,
      "step": 672
    },
    {
      "epoch": 0.19903145909578204,
      "grad_norm": 0.1981499046087265,
      "learning_rate": 0.00011720329520829429,
      "loss": 0.2201,
      "step": 673
    },
    {
      "epoch": 0.19932719677645927,
      "grad_norm": 0.19393107295036316,
      "learning_rate": 0.0001169955554275046,
      "loss": 0.193,
      "step": 674
    },
    {
      "epoch": 0.19962293445713652,
      "grad_norm": 0.19245851039886475,
      "learning_rate": 0.0001167877400918541,
      "loss": 0.1546,
      "step": 675
    },
    {
      "epoch": 0.19991867213781375,
      "grad_norm": 0.2009516805410385,
      "learning_rate": 0.00011657985012519952,
      "loss": 0.1963,
      "step": 676
    },
    {
      "epoch": 0.200214409818491,
      "grad_norm": 0.19694218039512634,
      "learning_rate": 0.00011637188645172944,
      "loss": 0.1624,
      "step": 677
    },
    {
      "epoch": 0.20051014749916823,
      "grad_norm": 0.21188542246818542,
      "learning_rate": 0.00011616384999596006,
      "loss": 0.1885,
      "step": 678
    },
    {
      "epoch": 0.2008058851798455,
      "grad_norm": 0.18305402994155884,
      "learning_rate": 0.00011595574168273111,
      "loss": 0.1353,
      "step": 679
    },
    {
      "epoch": 0.20110162286052272,
      "grad_norm": 0.23467271029949188,
      "learning_rate": 0.0001157475624372018,
      "loss": 0.2274,
      "step": 680
    },
    {
      "epoch": 0.20139736054119994,
      "grad_norm": 0.21421073377132416,
      "learning_rate": 0.0001155393131848467,
      "loss": 0.1823,
      "step": 681
    },
    {
      "epoch": 0.2016930982218772,
      "grad_norm": 0.23332397639751434,
      "learning_rate": 0.00011533099485145155,
      "loss": 0.2598,
      "step": 682
    },
    {
      "epoch": 0.20198883590255443,
      "grad_norm": 0.21139463782310486,
      "learning_rate": 0.00011512260836310924,
      "loss": 0.1666,
      "step": 683
    },
    {
      "epoch": 0.20228457358323168,
      "grad_norm": 0.25425291061401367,
      "learning_rate": 0.00011491415464621562,
      "loss": 0.2299,
      "step": 684
    },
    {
      "epoch": 0.2025803112639089,
      "grad_norm": 0.26843997836112976,
      "learning_rate": 0.00011470563462746541,
      "loss": 0.2056,
      "step": 685
    },
    {
      "epoch": 0.20287604894458616,
      "grad_norm": 0.26034682989120483,
      "learning_rate": 0.00011449704923384812,
      "loss": 0.2092,
      "step": 686
    },
    {
      "epoch": 0.2031717866252634,
      "grad_norm": 0.25974202156066895,
      "learning_rate": 0.00011428839939264382,
      "loss": 0.1912,
      "step": 687
    },
    {
      "epoch": 0.20346752430594062,
      "grad_norm": 0.2530285716056824,
      "learning_rate": 0.0001140796860314191,
      "loss": 0.1965,
      "step": 688
    },
    {
      "epoch": 0.20376326198661787,
      "grad_norm": 0.2534104883670807,
      "learning_rate": 0.00011387091007802297,
      "loss": 0.1936,
      "step": 689
    },
    {
      "epoch": 0.2040589996672951,
      "grad_norm": 0.24163153767585754,
      "learning_rate": 0.0001136620724605827,
      "loss": 0.1617,
      "step": 690
    },
    {
      "epoch": 0.20435473734797235,
      "grad_norm": 0.24182729423046112,
      "learning_rate": 0.00011345317410749964,
      "loss": 0.1716,
      "step": 691
    },
    {
      "epoch": 0.20465047502864958,
      "grad_norm": 0.26009559631347656,
      "learning_rate": 0.00011324421594744516,
      "loss": 0.1832,
      "step": 692
    },
    {
      "epoch": 0.20494621270932684,
      "grad_norm": 0.24706415832042694,
      "learning_rate": 0.00011303519890935656,
      "loss": 0.1604,
      "step": 693
    },
    {
      "epoch": 0.20524195039000406,
      "grad_norm": 0.3163825571537018,
      "learning_rate": 0.00011282612392243286,
      "loss": 0.2144,
      "step": 694
    },
    {
      "epoch": 0.20553768807068132,
      "grad_norm": 0.25135427713394165,
      "learning_rate": 0.00011261699191613066,
      "loss": 0.1179,
      "step": 695
    },
    {
      "epoch": 0.20583342575135855,
      "grad_norm": 0.31126582622528076,
      "learning_rate": 0.00011240780382016005,
      "loss": 0.1907,
      "step": 696
    },
    {
      "epoch": 0.20612916343203577,
      "grad_norm": 0.34292182326316833,
      "learning_rate": 0.00011219856056448051,
      "loss": 0.2255,
      "step": 697
    },
    {
      "epoch": 0.20642490111271303,
      "grad_norm": 0.3281469941139221,
      "learning_rate": 0.00011198926307929664,
      "loss": 0.1755,
      "step": 698
    },
    {
      "epoch": 0.20672063879339025,
      "grad_norm": 0.4065954387187958,
      "learning_rate": 0.00011177991229505431,
      "loss": 0.2255,
      "step": 699
    },
    {
      "epoch": 0.2070163764740675,
      "grad_norm": 0.5669332146644592,
      "learning_rate": 0.00011157050914243614,
      "loss": 0.3493,
      "step": 700
    },
    {
      "epoch": 0.20731211415474474,
      "grad_norm": 0.10655932128429413,
      "learning_rate": 0.00011136105455235766,
      "loss": 0.1189,
      "step": 701
    },
    {
      "epoch": 0.207607851835422,
      "grad_norm": 0.12350799888372421,
      "learning_rate": 0.00011115154945596305,
      "loss": 0.1384,
      "step": 702
    },
    {
      "epoch": 0.20790358951609922,
      "grad_norm": 0.16289597749710083,
      "learning_rate": 0.00011094199478462095,
      "loss": 0.1981,
      "step": 703
    },
    {
      "epoch": 0.20819932719677645,
      "grad_norm": 0.16098204255104065,
      "learning_rate": 0.00011073239146992054,
      "loss": 0.1988,
      "step": 704
    },
    {
      "epoch": 0.2084950648774537,
      "grad_norm": 0.17264235019683838,
      "learning_rate": 0.00011052274044366711,
      "loss": 0.1934,
      "step": 705
    },
    {
      "epoch": 0.20879080255813093,
      "grad_norm": 0.18324457108974457,
      "learning_rate": 0.00011031304263787812,
      "loss": 0.2202,
      "step": 706
    },
    {
      "epoch": 0.20908654023880818,
      "grad_norm": 0.1545696258544922,
      "learning_rate": 0.00011010329898477891,
      "loss": 0.1787,
      "step": 707
    },
    {
      "epoch": 0.2093822779194854,
      "grad_norm": 0.1769752949476242,
      "learning_rate": 0.0001098935104167988,
      "loss": 0.1949,
      "step": 708
    },
    {
      "epoch": 0.20967801560016266,
      "grad_norm": 0.1559659242630005,
      "learning_rate": 0.00010968367786656663,
      "loss": 0.1562,
      "step": 709
    },
    {
      "epoch": 0.2099737532808399,
      "grad_norm": 0.1613316386938095,
      "learning_rate": 0.00010947380226690684,
      "loss": 0.1836,
      "step": 710
    },
    {
      "epoch": 0.21026949096151715,
      "grad_norm": 0.18986111879348755,
      "learning_rate": 0.00010926388455083522,
      "loss": 0.241,
      "step": 711
    },
    {
      "epoch": 0.21056522864219437,
      "grad_norm": 0.17520417273044586,
      "learning_rate": 0.00010905392565155477,
      "loss": 0.181,
      "step": 712
    },
    {
      "epoch": 0.2108609663228716,
      "grad_norm": 0.17421351373195648,
      "learning_rate": 0.00010884392650245165,
      "loss": 0.2037,
      "step": 713
    },
    {
      "epoch": 0.21115670400354886,
      "grad_norm": 0.17320917546749115,
      "learning_rate": 0.00010863388803709089,
      "loss": 0.18,
      "step": 714
    },
    {
      "epoch": 0.21145244168422608,
      "grad_norm": 0.19977597892284393,
      "learning_rate": 0.00010842381118921232,
      "loss": 0.2179,
      "step": 715
    },
    {
      "epoch": 0.21174817936490334,
      "grad_norm": 0.19101393222808838,
      "learning_rate": 0.00010821369689272638,
      "loss": 0.2022,
      "step": 716
    },
    {
      "epoch": 0.21204391704558057,
      "grad_norm": 0.22305834293365479,
      "learning_rate": 0.00010800354608171003,
      "loss": 0.2826,
      "step": 717
    },
    {
      "epoch": 0.21233965472625782,
      "grad_norm": 0.1698966771364212,
      "learning_rate": 0.00010779335969040252,
      "loss": 0.1619,
      "step": 718
    },
    {
      "epoch": 0.21263539240693505,
      "grad_norm": 0.1868032068014145,
      "learning_rate": 0.00010758313865320134,
      "loss": 0.1907,
      "step": 719
    },
    {
      "epoch": 0.21293113008761227,
      "grad_norm": 0.1948738545179367,
      "learning_rate": 0.00010737288390465792,
      "loss": 0.2101,
      "step": 720
    },
    {
      "epoch": 0.21322686776828953,
      "grad_norm": 0.18316994607448578,
      "learning_rate": 0.00010716259637947357,
      "loss": 0.1733,
      "step": 721
    },
    {
      "epoch": 0.21352260544896676,
      "grad_norm": 0.18635033071041107,
      "learning_rate": 0.00010695227701249537,
      "loss": 0.177,
      "step": 722
    },
    {
      "epoch": 0.213818343129644,
      "grad_norm": 0.22876830399036407,
      "learning_rate": 0.00010674192673871191,
      "loss": 0.1832,
      "step": 723
    },
    {
      "epoch": 0.21411408081032124,
      "grad_norm": 0.17134256660938263,
      "learning_rate": 0.00010653154649324917,
      "loss": 0.1529,
      "step": 724
    },
    {
      "epoch": 0.2144098184909985,
      "grad_norm": 0.2593710422515869,
      "learning_rate": 0.00010632113721136636,
      "loss": 0.2832,
      "step": 725
    },
    {
      "epoch": 0.21470555617167572,
      "grad_norm": 0.24183620512485504,
      "learning_rate": 0.00010611069982845183,
      "loss": 0.2632,
      "step": 726
    },
    {
      "epoch": 0.21500129385235298,
      "grad_norm": 0.23827098309993744,
      "learning_rate": 0.00010590023528001884,
      "loss": 0.227,
      "step": 727
    },
    {
      "epoch": 0.2152970315330302,
      "grad_norm": 0.20484110713005066,
      "learning_rate": 0.00010568974450170139,
      "loss": 0.2038,
      "step": 728
    },
    {
      "epoch": 0.21559276921370743,
      "grad_norm": 0.22389110922813416,
      "learning_rate": 0.00010547922842925008,
      "loss": 0.2181,
      "step": 729
    },
    {
      "epoch": 0.21588850689438469,
      "grad_norm": 0.22689417004585266,
      "learning_rate": 0.00010526868799852796,
      "loss": 0.2301,
      "step": 730
    },
    {
      "epoch": 0.2161842445750619,
      "grad_norm": 0.21914827823638916,
      "learning_rate": 0.0001050581241455064,
      "loss": 0.1857,
      "step": 731
    },
    {
      "epoch": 0.21647998225573917,
      "grad_norm": 0.2330751121044159,
      "learning_rate": 0.00010484753780626089,
      "loss": 0.1797,
      "step": 732
    },
    {
      "epoch": 0.2167757199364164,
      "grad_norm": 0.22893856465816498,
      "learning_rate": 0.00010463692991696685,
      "loss": 0.2063,
      "step": 733
    },
    {
      "epoch": 0.21707145761709365,
      "grad_norm": 0.24380573630332947,
      "learning_rate": 0.00010442630141389549,
      "loss": 0.1874,
      "step": 734
    },
    {
      "epoch": 0.21736719529777088,
      "grad_norm": 0.24543069303035736,
      "learning_rate": 0.00010421565323340971,
      "loss": 0.1931,
      "step": 735
    },
    {
      "epoch": 0.2176629329784481,
      "grad_norm": 0.22604742646217346,
      "learning_rate": 0.00010400498631195992,
      "loss": 0.143,
      "step": 736
    },
    {
      "epoch": 0.21795867065912536,
      "grad_norm": 0.2889672815799713,
      "learning_rate": 0.00010379430158607975,
      "loss": 0.2563,
      "step": 737
    },
    {
      "epoch": 0.21825440833980259,
      "grad_norm": 0.24721504747867584,
      "learning_rate": 0.000103583599992382,
      "loss": 0.1586,
      "step": 738
    },
    {
      "epoch": 0.21855014602047984,
      "grad_norm": 0.23907063901424408,
      "learning_rate": 0.0001033728824675545,
      "loss": 0.1921,
      "step": 739
    },
    {
      "epoch": 0.21884588370115707,
      "grad_norm": 0.22747616469860077,
      "learning_rate": 0.0001031621499483559,
      "loss": 0.1351,
      "step": 740
    },
    {
      "epoch": 0.21914162138183432,
      "grad_norm": 0.3032531142234802,
      "learning_rate": 0.00010295140337161146,
      "loss": 0.2032,
      "step": 741
    },
    {
      "epoch": 0.21943735906251155,
      "grad_norm": 0.24150900542736053,
      "learning_rate": 0.00010274064367420897,
      "loss": 0.1577,
      "step": 742
    },
    {
      "epoch": 0.2197330967431888,
      "grad_norm": 0.29281744360923767,
      "learning_rate": 0.00010252987179309459,
      "loss": 0.2267,
      "step": 743
    },
    {
      "epoch": 0.22002883442386603,
      "grad_norm": 0.2816901206970215,
      "learning_rate": 0.00010231908866526851,
      "loss": 0.2086,
      "step": 744
    },
    {
      "epoch": 0.22032457210454326,
      "grad_norm": 0.343010276556015,
      "learning_rate": 0.00010210829522778111,
      "loss": 0.224,
      "step": 745
    },
    {
      "epoch": 0.22062030978522051,
      "grad_norm": 0.31681838631629944,
      "learning_rate": 0.00010189749241772844,
      "loss": 0.2536,
      "step": 746
    },
    {
      "epoch": 0.22091604746589774,
      "grad_norm": 0.2629532516002655,
      "learning_rate": 0.00010168668117224825,
      "loss": 0.1457,
      "step": 747
    },
    {
      "epoch": 0.221211785146575,
      "grad_norm": 0.33355703949928284,
      "learning_rate": 0.00010147586242851585,
      "loss": 0.2115,
      "step": 748
    },
    {
      "epoch": 0.22150752282725222,
      "grad_norm": 0.3667524755001068,
      "learning_rate": 0.00010126503712373982,
      "loss": 0.2341,
      "step": 749
    },
    {
      "epoch": 0.22180326050792948,
      "grad_norm": 0.5130096673965454,
      "learning_rate": 0.00010105420619515798,
      "loss": 0.2986,
      "step": 750
    },
    {
      "epoch": 0.22180326050792948,
      "eval_loss": 0.194038987159729,
      "eval_runtime": 323.8656,
      "eval_samples_per_second": 17.584,
      "eval_steps_per_second": 8.794,
      "step": 750
    }
  ],
  "logging_steps": 1,
  "max_steps": 1500,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 375,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 9.919782432528138e+17,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}