{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.19681810726586846,
  "eval_steps": 375,
  "global_step": 1500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00013121207151057897,
      "grad_norm": 721.845947265625,
      "learning_rate": 2e-05,
      "loss": 2.9516,
      "step": 1
    },
    {
      "epoch": 0.00026242414302115794,
      "grad_norm": 587.9494018554688,
      "learning_rate": 4e-05,
      "loss": 3.1565,
      "step": 2
    },
    {
      "epoch": 0.00039363621453173694,
      "grad_norm": 632.3089599609375,
      "learning_rate": 6e-05,
      "loss": 2.4513,
      "step": 3
    },
    {
      "epoch": 0.0005248482860423159,
      "grad_norm": 446.474609375,
      "learning_rate": 8e-05,
      "loss": 2.4105,
      "step": 4
    },
    {
      "epoch": 0.0006560603575528948,
      "grad_norm": 385.1297607421875,
      "learning_rate": 0.0001,
      "loss": 1.8571,
      "step": 5
    },
    {
      "epoch": 0.0007872724290634739,
      "grad_norm": 273.9917907714844,
      "learning_rate": 0.00012,
      "loss": 1.6504,
      "step": 6
    },
    {
      "epoch": 0.0009184845005740528,
      "grad_norm": 268.3256530761719,
      "learning_rate": 0.00014,
      "loss": 1.5392,
      "step": 7
    },
    {
      "epoch": 0.0010496965720846318,
      "grad_norm": 424.15521240234375,
      "learning_rate": 0.00016,
      "loss": 1.6152,
      "step": 8
    },
    {
      "epoch": 0.0011809086435952108,
      "grad_norm": 337.3836669921875,
      "learning_rate": 0.00018,
      "loss": 1.561,
      "step": 9
    },
    {
      "epoch": 0.0013121207151057896,
      "grad_norm": 327.4623718261719,
      "learning_rate": 0.0002,
      "loss": 1.6322,
      "step": 10
    },
    {
      "epoch": 0.0014433327866163687,
      "grad_norm": 252.9476318359375,
      "learning_rate": 0.00019999977772170748,
      "loss": 1.7542,
      "step": 11
    },
    {
      "epoch": 0.0015745448581269477,
      "grad_norm": 275.21063232421875,
      "learning_rate": 0.00019999911088781805,
      "loss": 1.4741,
      "step": 12
    },
    {
      "epoch": 0.0017057569296375266,
      "grad_norm": 226.21009826660156,
      "learning_rate": 0.0001999979995012962,
      "loss": 1.7102,
      "step": 13
    },
    {
      "epoch": 0.0018369690011481056,
      "grad_norm": 238.84994506835938,
      "learning_rate": 0.00019999644356708261,
      "loss": 1.5835,
      "step": 14
    },
    {
      "epoch": 0.0019681810726586845,
      "grad_norm": 245.36904907226562,
      "learning_rate": 0.00019999444309209432,
      "loss": 1.5874,
      "step": 15
    },
    {
      "epoch": 0.0020993931441692635,
      "grad_norm": 194.7904510498047,
      "learning_rate": 0.0001999919980852246,
      "loss": 1.5712,
      "step": 16
    },
    {
      "epoch": 0.0022306052156798426,
      "grad_norm": 188.93511962890625,
      "learning_rate": 0.00019998910855734288,
      "loss": 1.5587,
      "step": 17
    },
    {
      "epoch": 0.0023618172871904216,
      "grad_norm": 170.6442413330078,
      "learning_rate": 0.0001999857745212947,
      "loss": 1.6228,
      "step": 18
    },
    {
      "epoch": 0.0024930293587010007,
      "grad_norm": 186.362060546875,
      "learning_rate": 0.00019998199599190178,
      "loss": 1.6027,
      "step": 19
    },
    {
      "epoch": 0.0026242414302115793,
      "grad_norm": 154.62510681152344,
      "learning_rate": 0.0001999777729859618,
      "loss": 1.5752,
      "step": 20
    },
    {
      "epoch": 0.0027554535017221583,
      "grad_norm": 151.4012908935547,
      "learning_rate": 0.00019997310552224846,
      "loss": 1.4844,
      "step": 21
    },
    {
      "epoch": 0.0028866655732327374,
      "grad_norm": 140.5549774169922,
      "learning_rate": 0.00019996799362151122,
      "loss": 1.389,
      "step": 22
    },
    {
      "epoch": 0.0030178776447433164,
      "grad_norm": 146.2910614013672,
      "learning_rate": 0.00019996243730647538,
      "loss": 1.4649,
      "step": 23
    },
    {
      "epoch": 0.0031490897162538955,
      "grad_norm": 162.38140869140625,
      "learning_rate": 0.00019995643660184191,
      "loss": 1.5736,
      "step": 24
    },
    {
      "epoch": 0.0032803017877644745,
      "grad_norm": 141.8441619873047,
      "learning_rate": 0.00019994999153428737,
      "loss": 1.3662,
      "step": 25
    },
    {
      "epoch": 0.003411513859275053,
      "grad_norm": 139.2669677734375,
      "learning_rate": 0.00019994310213246368,
      "loss": 1.6385,
      "step": 26
    },
    {
      "epoch": 0.003542725930785632,
      "grad_norm": 142.02691650390625,
      "learning_rate": 0.00019993576842699816,
      "loss": 1.3043,
      "step": 27
    },
    {
      "epoch": 0.0036739380022962113,
      "grad_norm": 149.8494873046875,
      "learning_rate": 0.0001999279904504933,
      "loss": 1.6001,
      "step": 28
    },
    {
      "epoch": 0.0038051500738067903,
      "grad_norm": 219.7648468017578,
      "learning_rate": 0.00019991976823752653,
      "loss": 1.2926,
      "step": 29
    },
    {
      "epoch": 0.003936362145317369,
      "grad_norm": 181.8307647705078,
      "learning_rate": 0.00019991110182465032,
      "loss": 1.6259,
      "step": 30
    },
    {
      "epoch": 0.004067574216827948,
      "grad_norm": 154.86196899414062,
      "learning_rate": 0.00019990199125039174,
      "loss": 1.4566,
      "step": 31
    },
    {
      "epoch": 0.004198786288338527,
      "grad_norm": 145.7695770263672,
      "learning_rate": 0.00019989243655525247,
      "loss": 1.5457,
      "step": 32
    },
    {
      "epoch": 0.004329998359849106,
      "grad_norm": 117.9987564086914,
      "learning_rate": 0.00019988243778170853,
      "loss": 1.4211,
      "step": 33
    },
    {
      "epoch": 0.004461210431359685,
      "grad_norm": 147.84857177734375,
      "learning_rate": 0.0001998719949742101,
      "loss": 1.5932,
      "step": 34
    },
    {
      "epoch": 0.004592422502870264,
      "grad_norm": 167.62083435058594,
      "learning_rate": 0.0001998611081791814,
      "loss": 1.3841,
      "step": 35
    },
    {
      "epoch": 0.004723634574380843,
      "grad_norm": 144.39222717285156,
      "learning_rate": 0.00019984977744502038,
      "loss": 1.6815,
      "step": 36
    },
    {
      "epoch": 0.004854846645891422,
      "grad_norm": 160.74594116210938,
      "learning_rate": 0.00019983800282209857,
      "loss": 1.6233,
      "step": 37
    },
    {
      "epoch": 0.004986058717402001,
      "grad_norm": 142.86062622070312,
      "learning_rate": 0.00019982578436276082,
      "loss": 1.5885,
      "step": 38
    },
    {
      "epoch": 0.00511727078891258,
      "grad_norm": 183.543212890625,
      "learning_rate": 0.00019981312212132512,
      "loss": 1.672,
      "step": 39
    },
    {
      "epoch": 0.0052484828604231586,
      "grad_norm": 146.4871826171875,
      "learning_rate": 0.00019980001615408228,
      "loss": 1.5933,
      "step": 40
    },
    {
      "epoch": 0.005379694931933738,
      "grad_norm": 154.18446350097656,
      "learning_rate": 0.00019978646651929572,
      "loss": 1.6446,
      "step": 41
    },
    {
      "epoch": 0.005510907003444317,
      "grad_norm": 131.88648986816406,
      "learning_rate": 0.00019977247327720128,
      "loss": 1.3784,
      "step": 42
    },
    {
      "epoch": 0.005642119074954896,
      "grad_norm": 146.59091186523438,
      "learning_rate": 0.0001997580364900068,
      "loss": 1.4508,
      "step": 43
    },
    {
      "epoch": 0.005773331146465475,
      "grad_norm": 134.70986938476562,
      "learning_rate": 0.000199743156221892,
      "loss": 1.3061,
      "step": 44
    },
    {
      "epoch": 0.005904543217976054,
      "grad_norm": 157.33111572265625,
      "learning_rate": 0.00019972783253900808,
      "loss": 1.581,
      "step": 45
    },
    {
      "epoch": 0.006035755289486633,
      "grad_norm": 171.9995880126953,
      "learning_rate": 0.00019971206550947748,
      "loss": 1.6867,
      "step": 46
    },
    {
      "epoch": 0.006166967360997212,
      "grad_norm": 157.56784057617188,
      "learning_rate": 0.00019969585520339354,
      "loss": 1.4743,
      "step": 47
    },
    {
      "epoch": 0.006298179432507791,
      "grad_norm": 141.30662536621094,
      "learning_rate": 0.0001996792016928203,
      "loss": 1.6981,
      "step": 48
    },
    {
      "epoch": 0.00642939150401837,
      "grad_norm": 178.53582763671875,
      "learning_rate": 0.00019966210505179197,
      "loss": 1.5791,
      "step": 49
    },
    {
      "epoch": 0.006560603575528949,
      "grad_norm": 165.05657958984375,
      "learning_rate": 0.00019964456535631286,
      "loss": 1.3846,
      "step": 50
    },
    {
      "epoch": 0.006691815647039527,
      "grad_norm": 159.24615478515625,
      "learning_rate": 0.0001996265826843568,
      "loss": 1.4014,
      "step": 51
    },
    {
      "epoch": 0.006823027718550106,
      "grad_norm": 151.1456756591797,
      "learning_rate": 0.00019960815711586696,
      "loss": 1.777,
      "step": 52
    },
    {
      "epoch": 0.006954239790060685,
      "grad_norm": 145.77236938476562,
      "learning_rate": 0.00019958928873275539,
      "loss": 1.3307,
      "step": 53
    },
    {
      "epoch": 0.007085451861571264,
      "grad_norm": 134.715087890625,
      "learning_rate": 0.00019956997761890277,
      "loss": 1.3744,
      "step": 54
    },
    {
      "epoch": 0.0072166639330818435,
      "grad_norm": 159.2729949951172,
      "learning_rate": 0.00019955022386015792,
      "loss": 1.682,
      "step": 55
    },
    {
      "epoch": 0.0073478760045924225,
      "grad_norm": 122.81232452392578,
      "learning_rate": 0.00019953002754433743,
      "loss": 1.6228,
      "step": 56
    },
    {
      "epoch": 0.0074790880761030016,
      "grad_norm": 132.26651000976562,
      "learning_rate": 0.00019950938876122542,
      "loss": 1.5912,
      "step": 57
    },
    {
      "epoch": 0.007610300147613581,
      "grad_norm": 118.78931427001953,
      "learning_rate": 0.00019948830760257291,
      "loss": 1.5917,
      "step": 58
    },
    {
      "epoch": 0.00774151221912416,
      "grad_norm": 113.86380767822266,
      "learning_rate": 0.0001994667841620976,
      "loss": 1.477,
      "step": 59
    },
    {
      "epoch": 0.007872724290634738,
      "grad_norm": 110.57164001464844,
      "learning_rate": 0.00019944481853548335,
      "loss": 1.6343,
      "step": 60
    },
    {
      "epoch": 0.008003936362145317,
      "grad_norm": 116.8375015258789,
      "learning_rate": 0.00019942241082037982,
      "loss": 1.5631,
      "step": 61
    },
    {
      "epoch": 0.008135148433655896,
      "grad_norm": 138.9037322998047,
      "learning_rate": 0.00019939956111640197,
      "loss": 1.6706,
      "step": 62
    },
    {
      "epoch": 0.008266360505166475,
      "grad_norm": 119.9926986694336,
      "learning_rate": 0.00019937626952512964,
      "loss": 1.3332,
      "step": 63
    },
    {
      "epoch": 0.008397572576677054,
      "grad_norm": 119.2025375366211,
      "learning_rate": 0.0001993525361501072,
      "loss": 1.5094,
      "step": 64
    },
    {
      "epoch": 0.008528784648187633,
      "grad_norm": 136.4790496826172,
      "learning_rate": 0.00019932836109684286,
      "loss": 1.7717,
      "step": 65
    },
    {
      "epoch": 0.008659996719698212,
      "grad_norm": 111.85336303710938,
      "learning_rate": 0.00019930374447280845,
      "loss": 1.4778,
      "step": 66
    },
    {
      "epoch": 0.008791208791208791,
      "grad_norm": 125.69164276123047,
      "learning_rate": 0.00019927868638743875,
      "loss": 1.7626,
      "step": 67
    },
    {
      "epoch": 0.00892242086271937,
      "grad_norm": 138.92538452148438,
      "learning_rate": 0.0001992531869521312,
      "loss": 1.5218,
      "step": 68
    },
    {
      "epoch": 0.00905363293422995,
      "grad_norm": 119.6041030883789,
      "learning_rate": 0.00019922724628024515,
      "loss": 1.2998,
      "step": 69
    },
    {
      "epoch": 0.009184845005740528,
      "grad_norm": 113.15282440185547,
      "learning_rate": 0.0001992008644871016,
      "loss": 1.1316,
      "step": 70
    },
    {
      "epoch": 0.009316057077251107,
      "grad_norm": 128.54147338867188,
      "learning_rate": 0.00019917404168998256,
      "loss": 1.5693,
      "step": 71
    },
    {
      "epoch": 0.009447269148761686,
      "grad_norm": 125.87454986572266,
      "learning_rate": 0.0001991467780081305,
      "loss": 1.5725,
      "step": 72
    },
    {
      "epoch": 0.009578481220272266,
      "grad_norm": 125.46642303466797,
      "learning_rate": 0.00019911907356274795,
      "loss": 1.4539,
      "step": 73
    },
    {
      "epoch": 0.009709693291782845,
      "grad_norm": 115.56763458251953,
      "learning_rate": 0.00019909092847699683,
      "loss": 1.6424,
      "step": 74
    },
    {
      "epoch": 0.009840905363293424,
      "grad_norm": 119.49039459228516,
      "learning_rate": 0.00019906234287599798,
      "loss": 1.5194,
      "step": 75
    },
    {
      "epoch": 0.009972117434804003,
      "grad_norm": 122.14395141601562,
      "learning_rate": 0.00019903331688683057,
      "loss": 1.5486,
      "step": 76
    },
    {
      "epoch": 0.010103329506314582,
      "grad_norm": 119.73626708984375,
      "learning_rate": 0.00019900385063853154,
      "loss": 1.651,
      "step": 77
    },
    {
      "epoch": 0.01023454157782516,
      "grad_norm": 122.35995483398438,
      "learning_rate": 0.00019897394426209505,
      "loss": 1.5211,
      "step": 78
    },
    {
      "epoch": 0.010365753649335738,
      "grad_norm": 146.387939453125,
      "learning_rate": 0.00019894359789047187,
      "loss": 1.6203,
      "step": 79
    },
    {
      "epoch": 0.010496965720846317,
      "grad_norm": 116.01179504394531,
      "learning_rate": 0.00019891281165856873,
      "loss": 1.367,
      "step": 80
    },
    {
      "epoch": 0.010628177792356896,
      "grad_norm": 142.5491943359375,
      "learning_rate": 0.00019888158570324795,
      "loss": 1.6747,
      "step": 81
    },
    {
      "epoch": 0.010759389863867475,
      "grad_norm": 122.56082153320312,
      "learning_rate": 0.0001988499201633265,
      "loss": 1.5454,
      "step": 82
    },
    {
      "epoch": 0.010890601935378054,
      "grad_norm": 150.46270751953125,
      "learning_rate": 0.00019881781517957562,
      "loss": 1.5616,
      "step": 83
    },
    {
      "epoch": 0.011021814006888633,
      "grad_norm": 105.08324432373047,
      "learning_rate": 0.0001987852708947202,
      "loss": 1.5542,
      "step": 84
    },
    {
      "epoch": 0.011153026078399212,
      "grad_norm": 116.74308776855469,
      "learning_rate": 0.00019875228745343794,
      "loss": 1.4711,
      "step": 85
    },
    {
      "epoch": 0.011284238149909791,
      "grad_norm": 115.30254364013672,
      "learning_rate": 0.0001987188650023589,
      "loss": 1.2222,
      "step": 86
    },
    {
      "epoch": 0.01141545022142037,
      "grad_norm": 122.88642883300781,
      "learning_rate": 0.0001986850036900648,
      "loss": 1.5278,
      "step": 87
    },
    {
      "epoch": 0.01154666229293095,
      "grad_norm": 106.96451568603516,
      "learning_rate": 0.00019865070366708836,
      "loss": 1.3331,
      "step": 88
    },
    {
      "epoch": 0.011677874364441529,
      "grad_norm": 123.54383087158203,
      "learning_rate": 0.00019861596508591255,
      "loss": 1.5431,
      "step": 89
    },
    {
      "epoch": 0.011809086435952108,
      "grad_norm": 108.67639923095703,
      "learning_rate": 0.00019858078810097002,
      "loss": 1.4028,
      "step": 90
    },
    {
      "epoch": 0.011940298507462687,
      "grad_norm": 131.77220153808594,
      "learning_rate": 0.00019854517286864245,
      "loss": 1.5082,
      "step": 91
    },
    {
      "epoch": 0.012071510578973266,
      "grad_norm": 107.17201232910156,
      "learning_rate": 0.0001985091195472596,
      "loss": 1.4177,
      "step": 92
    },
    {
      "epoch": 0.012202722650483845,
      "grad_norm": 119.97637176513672,
      "learning_rate": 0.0001984726282970989,
      "loss": 1.3731,
      "step": 93
    },
    {
      "epoch": 0.012333934721994424,
      "grad_norm": 148.12454223632812,
      "learning_rate": 0.0001984356992803847,
      "loss": 1.7341,
      "step": 94
    },
    {
      "epoch": 0.012465146793505003,
      "grad_norm": 131.28062438964844,
      "learning_rate": 0.00019839833266128724,
      "loss": 1.5652,
      "step": 95
    },
    {
      "epoch": 0.012596358865015582,
      "grad_norm": 121.13946533203125,
      "learning_rate": 0.00019836052860592237,
      "loss": 1.6214,
      "step": 96
    },
    {
      "epoch": 0.012727570936526161,
      "grad_norm": 126.72644805908203,
      "learning_rate": 0.0001983222872823505,
      "loss": 1.5811,
      "step": 97
    },
    {
      "epoch": 0.01285878300803674,
      "grad_norm": 111.71343231201172,
      "learning_rate": 0.00019828360886057594,
      "loss": 1.6433,
      "step": 98
    },
    {
      "epoch": 0.012989995079547319,
      "grad_norm": 111.59823608398438,
      "learning_rate": 0.00019824449351254616,
      "loss": 1.3162,
      "step": 99
    },
    {
      "epoch": 0.013121207151057898,
      "grad_norm": 112.52090454101562,
      "learning_rate": 0.00019820494141215104,
      "loss": 1.3572,
      "step": 100
    },
    {
      "epoch": 0.013252419222568475,
      "grad_norm": 163.2476043701172,
      "learning_rate": 0.000198164952735222,
      "loss": 1.2929,
      "step": 101
    },
    {
      "epoch": 0.013383631294079055,
      "grad_norm": 111.91755676269531,
      "learning_rate": 0.00019812452765953135,
      "loss": 1.3325,
      "step": 102
    },
    {
      "epoch": 0.013514843365589634,
      "grad_norm": 110.2037124633789,
      "learning_rate": 0.00019808366636479147,
      "loss": 1.3612,
      "step": 103
    },
    {
      "epoch": 0.013646055437100213,
      "grad_norm": 162.9724884033203,
      "learning_rate": 0.00019804236903265388,
      "loss": 1.6573,
      "step": 104
    },
    {
      "epoch": 0.013777267508610792,
      "grad_norm": 124.71128845214844,
      "learning_rate": 0.00019800063584670863,
      "loss": 1.5555,
      "step": 105
    },
    {
      "epoch": 0.01390847958012137,
      "grad_norm": 129.85768127441406,
      "learning_rate": 0.00019795846699248332,
      "loss": 1.5338,
      "step": 106
    },
    {
      "epoch": 0.01403969165163195,
      "grad_norm": 109.72705841064453,
      "learning_rate": 0.00019791586265744237,
      "loss": 1.8231,
      "step": 107
    },
    {
      "epoch": 0.014170903723142529,
      "grad_norm": 103.44184875488281,
      "learning_rate": 0.00019787282303098617,
      "loss": 1.6717,
      "step": 108
    },
    {
      "epoch": 0.014302115794653108,
      "grad_norm": 108.34283447265625,
      "learning_rate": 0.0001978293483044502,
      "loss": 1.384,
      "step": 109
    },
    {
      "epoch": 0.014433327866163687,
      "grad_norm": 116.66336822509766,
      "learning_rate": 0.00019778543867110426,
      "loss": 1.7057,
      "step": 110
    },
    {
      "epoch": 0.014564539937674266,
      "grad_norm": 111.7286148071289,
      "learning_rate": 0.00019774109432615147,
      "loss": 1.4811,
      "step": 111
    },
    {
      "epoch": 0.014695752009184845,
      "grad_norm": 112.80461883544922,
      "learning_rate": 0.00019769631546672756,
      "loss": 1.5241,
      "step": 112
    },
    {
      "epoch": 0.014826964080695424,
      "grad_norm": 127.49102020263672,
      "learning_rate": 0.00019765110229189988,
      "loss": 1.4545,
      "step": 113
    },
    {
      "epoch": 0.014958176152206003,
      "grad_norm": 128.5435333251953,
      "learning_rate": 0.00019760545500266657,
      "loss": 1.6771,
      "step": 114
    },
    {
      "epoch": 0.015089388223716582,
      "grad_norm": 114.7703857421875,
      "learning_rate": 0.00019755937380195568,
      "loss": 1.4431,
      "step": 115
    },
    {
      "epoch": 0.015220600295227161,
      "grad_norm": 109.28499603271484,
      "learning_rate": 0.00019751285889462423,
      "loss": 1.5756,
      "step": 116
    },
    {
      "epoch": 0.01535181236673774,
      "grad_norm": 101.58000183105469,
      "learning_rate": 0.0001974659104874573,
      "loss": 1.4165,
      "step": 117
    },
    {
      "epoch": 0.01548302443824832,
      "grad_norm": 103.41732788085938,
      "learning_rate": 0.0001974185287891671,
      "loss": 1.3892,
      "step": 118
    },
    {
      "epoch": 0.015614236509758898,
      "grad_norm": 104.41741180419922,
      "learning_rate": 0.0001973707140103921,
      "loss": 1.3338,
      "step": 119
    },
    {
      "epoch": 0.015745448581269476,
      "grad_norm": 113.5536880493164,
      "learning_rate": 0.00019732246636369605,
      "loss": 1.3046,
      "step": 120
    },
    {
      "epoch": 0.015876660652780056,
      "grad_norm": 131.87432861328125,
      "learning_rate": 0.00019727378606356703,
      "loss": 1.614,
      "step": 121
    },
    {
      "epoch": 0.016007872724290634,
      "grad_norm": 110.47560119628906,
      "learning_rate": 0.00019722467332641656,
      "loss": 1.6724,
      "step": 122
    },
    {
      "epoch": 0.016139084795801215,
      "grad_norm": 106.15983581542969,
      "learning_rate": 0.00019717512837057855,
      "loss": 1.408,
      "step": 123
    },
    {
      "epoch": 0.016270296867311792,
      "grad_norm": 103.1834716796875,
      "learning_rate": 0.0001971251514163083,
      "loss": 1.4777,
      "step": 124
    },
    {
      "epoch": 0.016401508938822373,
      "grad_norm": 113.36947631835938,
      "learning_rate": 0.0001970747426857817,
      "loss": 1.5659,
      "step": 125
    },
    {
      "epoch": 0.01653272101033295,
      "grad_norm": 121.3810806274414,
      "learning_rate": 0.00019702390240309404,
      "loss": 1.4647,
      "step": 126
    },
    {
      "epoch": 0.01666393308184353,
      "grad_norm": 127.23417663574219,
      "learning_rate": 0.0001969726307942592,
      "loss": 1.6911,
      "step": 127
    },
    {
      "epoch": 0.016795145153354108,
      "grad_norm": 140.05999755859375,
      "learning_rate": 0.00019692092808720846,
      "loss": 1.6004,
      "step": 128
    },
    {
      "epoch": 0.01692635722486469,
      "grad_norm": 110.03852081298828,
      "learning_rate": 0.0001968687945117896,
      "loss": 1.4512,
      "step": 129
    },
    {
      "epoch": 0.017057569296375266,
      "grad_norm": 109.4391098022461,
      "learning_rate": 0.00019681623029976588,
      "loss": 1.3854,
      "step": 130
    },
    {
      "epoch": 0.017188781367885847,
      "grad_norm": 108.63151550292969,
      "learning_rate": 0.00019676323568481498,
      "loss": 1.6451,
      "step": 131
    },
    {
      "epoch": 0.017319993439396424,
      "grad_norm": 105.85716247558594,
      "learning_rate": 0.00019670981090252792,
      "loss": 1.4937,
      "step": 132
    },
    {
      "epoch": 0.017451205510907005,
      "grad_norm": 111.54838562011719,
      "learning_rate": 0.00019665595619040808,
      "loss": 1.5797,
      "step": 133
    },
    {
      "epoch": 0.017582417582417582,
      "grad_norm": 98.45042419433594,
      "learning_rate": 0.0001966016717878702,
      "loss": 1.0606,
      "step": 134
    },
    {
      "epoch": 0.01771362965392816,
      "grad_norm": 123.30680847167969,
      "learning_rate": 0.00019654695793623907,
      "loss": 1.5078,
      "step": 135
    },
    {
      "epoch": 0.01784484172543874,
      "grad_norm": 125.98291015625,
      "learning_rate": 0.0001964918148787488,
      "loss": 1.3376,
      "step": 136
    },
    {
      "epoch": 0.017976053796949318,
      "grad_norm": 109.19308471679688,
      "learning_rate": 0.00019643624286054144,
      "loss": 1.3388,
      "step": 137
    },
    {
      "epoch": 0.0181072658684599,
      "grad_norm": 102.06969451904297,
      "learning_rate": 0.00019638024212866606,
      "loss": 1.2907,
      "step": 138
    },
    {
      "epoch": 0.018238477939970476,
      "grad_norm": 108.79220581054688,
      "learning_rate": 0.0001963238129320776,
      "loss": 1.2599,
      "step": 139
    },
    {
      "epoch": 0.018369690011481057,
      "grad_norm": 122.432373046875,
      "learning_rate": 0.00019626695552163578,
      "loss": 1.6109,
      "step": 140
    },
    {
      "epoch": 0.018500902082991634,
      "grad_norm": 114.22309875488281,
      "learning_rate": 0.00019620967015010395,
      "loss": 1.4444,
      "step": 141
    },
    {
      "epoch": 0.018632114154502215,
      "grad_norm": 113.41559600830078,
      "learning_rate": 0.00019615195707214803,
      "loss": 1.295,
      "step": 142
    },
    {
      "epoch": 0.018763326226012792,
      "grad_norm": 127.58735656738281,
      "learning_rate": 0.0001960938165443353,
      "loss": 1.6505,
      "step": 143
    },
    {
      "epoch": 0.018894538297523373,
      "grad_norm": 112.135986328125,
      "learning_rate": 0.00019603524882513327,
      "loss": 1.4577,
      "step": 144
    },
    {
      "epoch": 0.01902575036903395,
      "grad_norm": 113.62690734863281,
      "learning_rate": 0.0001959762541749086,
      "loss": 1.4928,
      "step": 145
    },
    {
      "epoch": 0.01915696244054453,
      "grad_norm": 123.83708953857422,
      "learning_rate": 0.00019591683285592593,
      "loss": 1.682,
      "step": 146
    },
    {
      "epoch": 0.01928817451205511,
      "grad_norm": 112.98189544677734,
      "learning_rate": 0.00019585698513234663,
      "loss": 1.4441,
      "step": 147
    },
    {
      "epoch": 0.01941938658356569,
      "grad_norm": 96.95732116699219,
      "learning_rate": 0.0001957967112702277,
      "loss": 1.3659,
      "step": 148
    },
    {
      "epoch": 0.019550598655076266,
      "grad_norm": 100.92425537109375,
      "learning_rate": 0.00019573601153752052,
      "loss": 1.4038,
      "step": 149
    },
    {
      "epoch": 0.019681810726586847,
      "grad_norm": 89.0321273803711,
      "learning_rate": 0.00019567488620406983,
      "loss": 0.9645,
      "step": 150
    },
    {
      "epoch": 0.019813022798097425,
      "grad_norm": 103.00167846679688,
      "learning_rate": 0.00019561333554161224,
      "loss": 1.6603,
      "step": 151
    },
    {
      "epoch": 0.019944234869608005,
      "grad_norm": 107.33059692382812,
      "learning_rate": 0.0001955513598237753,
      "loss": 1.7484,
      "step": 152
    },
    {
      "epoch": 0.020075446941118583,
      "grad_norm": 99.6077880859375,
      "learning_rate": 0.00019548895932607621,
      "loss": 1.4055,
      "step": 153
    },
    {
      "epoch": 0.020206659012629163,
      "grad_norm": 95.64451599121094,
      "learning_rate": 0.00019542613432592038,
      "loss": 1.3994,
      "step": 154
    },
    {
      "epoch": 0.02033787108413974,
      "grad_norm": 103.3692855834961,
      "learning_rate": 0.00019536288510260056,
      "loss": 1.6396,
      "step": 155
    },
    {
      "epoch": 0.02046908315565032,
      "grad_norm": 97.58636474609375,
      "learning_rate": 0.00019529921193729534,
      "loss": 1.5063,
      "step": 156
    },
    {
      "epoch": 0.0206002952271609,
      "grad_norm": 93.93655395507812,
      "learning_rate": 0.00019523511511306793,
      "loss": 1.5479,
      "step": 157
    },
    {
      "epoch": 0.020731507298671476,
      "grad_norm": 98.70283508300781,
      "learning_rate": 0.000195170594914865,
      "loss": 1.5365,
      "step": 158
    },
    {
      "epoch": 0.020862719370182057,
      "grad_norm": 92.167724609375,
      "learning_rate": 0.00019510565162951537,
      "loss": 1.2108,
      "step": 159
    },
    {
      "epoch": 0.020993931441692634,
      "grad_norm": 102.41527557373047,
      "learning_rate": 0.00019504028554572864,
      "loss": 1.4331,
      "step": 160
    },
    {
      "epoch": 0.021125143513203215,
      "grad_norm": 104.78096008300781,
      "learning_rate": 0.00019497449695409408,
      "loss": 1.679,
      "step": 161
    },
    {
      "epoch": 0.021256355584713792,
      "grad_norm": 103.57251739501953,
      "learning_rate": 0.00019490828614707916,
      "loss": 1.4935,
      "step": 162
    },
    {
      "epoch": 0.021387567656224373,
      "grad_norm": 102.20472717285156,
      "learning_rate": 0.00019484165341902845,
      "loss": 1.522,
      "step": 163
    },
    {
      "epoch": 0.02151877972773495,
      "grad_norm": 99.70207214355469,
      "learning_rate": 0.00019477459906616206,
      "loss": 1.5449,
      "step": 164
    },
    {
      "epoch": 0.02164999179924553,
      "grad_norm": 93.83798217773438,
      "learning_rate": 0.00019470712338657458,
      "loss": 1.4679,
      "step": 165
    },
    {
      "epoch": 0.02178120387075611,
      "grad_norm": 101.6964340209961,
      "learning_rate": 0.0001946392266802336,
      "loss": 1.6945,
      "step": 166
    },
    {
      "epoch": 0.02191241594226669,
      "grad_norm": 94.62825775146484,
      "learning_rate": 0.0001945709092489783,
      "loss": 1.5794,
      "step": 167
    },
    {
      "epoch": 0.022043628013777267,
      "grad_norm": 95.08577728271484,
      "learning_rate": 0.00019450217139651844,
      "loss": 1.593,
      "step": 168
    },
    {
      "epoch": 0.022174840085287847,
      "grad_norm": 90.42913818359375,
      "learning_rate": 0.0001944330134284326,
      "loss": 0.9576,
      "step": 169
    },
    {
      "epoch": 0.022306052156798425,
      "grad_norm": 99.11161804199219,
      "learning_rate": 0.00019436343565216711,
      "loss": 1.2824,
      "step": 170
    },
    {
      "epoch": 0.022437264228309006,
      "grad_norm": 125.21051788330078,
      "learning_rate": 0.00019429343837703455,
      "loss": 1.6119,
      "step": 171
    },
    {
      "epoch": 0.022568476299819583,
      "grad_norm": 133.7545928955078,
      "learning_rate": 0.0001942230219142124,
      "loss": 1.6619,
      "step": 172
    },
    {
      "epoch": 0.022699688371330164,
      "grad_norm": 103.40087890625,
      "learning_rate": 0.0001941521865767417,
      "loss": 1.3251,
      "step": 173
    },
    {
      "epoch": 0.02283090044284074,
      "grad_norm": 100.81037902832031,
      "learning_rate": 0.0001940809326795256,
      "loss": 1.5844,
      "step": 174
    },
    {
      "epoch": 0.022962112514351322,
      "grad_norm": 94.39893341064453,
      "learning_rate": 0.000194009260539328,
      "loss": 1.327,
      "step": 175
    },
    {
      "epoch": 0.0230933245858619,
      "grad_norm": 100.41059112548828,
      "learning_rate": 0.0001939371704747721,
      "loss": 1.6887,
      "step": 176
    },
    {
      "epoch": 0.02322453665737248,
      "grad_norm": 95.26428985595703,
      "learning_rate": 0.00019386466280633906,
      "loss": 1.348,
      "step": 177
    },
    {
      "epoch": 0.023355748728883057,
      "grad_norm": 106.76798248291016,
      "learning_rate": 0.00019379173785636646,
      "loss": 1.4846,
      "step": 178
    },
    {
      "epoch": 0.023486960800393634,
      "grad_norm": 112.457763671875,
      "learning_rate": 0.000193718395949047,
      "loss": 1.2981,
      "step": 179
    },
    {
      "epoch": 0.023618172871904215,
      "grad_norm": 97.22601318359375,
      "learning_rate": 0.00019364463741042694,
      "loss": 1.1909,
      "step": 180
    },
    {
      "epoch": 0.023749384943414793,
      "grad_norm": 133.6383514404297,
      "learning_rate": 0.00019357046256840473,
      "loss": 1.7454,
      "step": 181
    },
    {
      "epoch": 0.023880597014925373,
      "grad_norm": 113.26380157470703,
      "learning_rate": 0.00019349587175272948,
      "loss": 1.537,
      "step": 182
    },
    {
      "epoch": 0.02401180908643595,
      "grad_norm": 109.44064331054688,
      "learning_rate": 0.0001934208652949996,
      "loss": 1.3897,
      "step": 183
    },
    {
      "epoch": 0.02414302115794653,
      "grad_norm": 111.51908874511719,
      "learning_rate": 0.00019334544352866127,
      "loss": 1.5215,
      "step": 184
    },
    {
      "epoch": 0.02427423322945711,
      "grad_norm": 93.65419006347656,
      "learning_rate": 0.00019326960678900688,
      "loss": 1.4067,
      "step": 185
    },
    {
      "epoch": 0.02440544530096769,
      "grad_norm": 94.6824722290039,
      "learning_rate": 0.00019319335541317361,
      "loss": 1.3239,
      "step": 186
    },
    {
      "epoch": 0.024536657372478267,
      "grad_norm": 101.01213073730469,
      "learning_rate": 0.00019311668974014208,
      "loss": 1.4962,
      "step": 187
    },
    {
      "epoch": 0.024667869443988848,
      "grad_norm": 120.41251373291016,
      "learning_rate": 0.00019303961011073447,
      "loss": 1.7658,
      "step": 188
    },
    {
      "epoch": 0.024799081515499425,
      "grad_norm": 114.20089721679688,
      "learning_rate": 0.00019296211686761346,
      "loss": 1.6297,
      "step": 189
    },
    {
      "epoch": 0.024930293587010006,
      "grad_norm": 103.60792541503906,
      "learning_rate": 0.00019288421035528028,
      "loss": 1.4031,
      "step": 190
    },
    {
      "epoch": 0.025061505658520583,
      "grad_norm": 120.98343658447266,
      "learning_rate": 0.00019280589092007352,
      "loss": 1.6236,
      "step": 191
    },
    {
      "epoch": 0.025192717730031164,
      "grad_norm": 116.35912322998047,
      "learning_rate": 0.00019272715891016735,
      "loss": 1.4308,
      "step": 192
    },
    {
      "epoch": 0.02532392980154174,
      "grad_norm": 103.06234741210938,
      "learning_rate": 0.00019264801467557007,
      "loss": 1.4408,
      "step": 193
    },
    {
      "epoch": 0.025455141873052322,
      "grad_norm": 104.24729919433594,
      "learning_rate": 0.00019256845856812266,
      "loss": 1.4133,
      "step": 194
    },
    {
      "epoch": 0.0255863539445629,
      "grad_norm": 100.61005401611328,
      "learning_rate": 0.000192488490941497,
      "loss": 1.5287,
      "step": 195
    },
    {
      "epoch": 0.02571756601607348,
      "grad_norm": 98.92796325683594,
      "learning_rate": 0.00019240811215119448,
      "loss": 1.2894,
      "step": 196
    },
    {
      "epoch": 0.025848778087584057,
      "grad_norm": 105.49788665771484,
      "learning_rate": 0.00019232732255454422,
      "loss": 1.3739,
      "step": 197
    },
    {
      "epoch": 0.025979990159094638,
      "grad_norm": 93.3214340209961,
      "learning_rate": 0.00019224612251070175,
      "loss": 1.0853,
      "step": 198
    },
    {
      "epoch": 0.026111202230605216,
      "grad_norm": 114.79483032226562,
      "learning_rate": 0.0001921645123806472,
      "loss": 1.4507,
      "step": 199
    },
    {
      "epoch": 0.026242414302115796,
      "grad_norm": 95.36935424804688,
      "learning_rate": 0.0001920824925271838,
      "loss": 1.152,
      "step": 200
    },
    {
      "epoch": 0.026373626373626374,
      "grad_norm": 83.26315307617188,
      "learning_rate": 0.0001920000633149362,
      "loss": 1.0678,
      "step": 201
    },
    {
      "epoch": 0.02650483844513695,
      "grad_norm": 121.8676528930664,
      "learning_rate": 0.00019191722511034884,
      "loss": 1.8465,
      "step": 202
    },
    {
      "epoch": 0.02663605051664753,
      "grad_norm": 92.6772232055664,
      "learning_rate": 0.00019183397828168448,
      "loss": 1.3252,
      "step": 203
    },
    {
      "epoch": 0.02676726258815811,
      "grad_norm": 87.77066040039062,
      "learning_rate": 0.00019175032319902234,
      "loss": 1.4254,
      "step": 204
    },
    {
      "epoch": 0.02689847465966869,
      "grad_norm": 94.39461517333984,
      "learning_rate": 0.00019166626023425662,
      "loss": 1.4612,
      "step": 205
    },
    {
      "epoch": 0.027029686731179267,
      "grad_norm": 108.58969116210938,
      "learning_rate": 0.00019158178976109476,
      "loss": 1.4196,
      "step": 206
    },
    {
      "epoch": 0.027160898802689848,
      "grad_norm": 109.79191589355469,
      "learning_rate": 0.0001914969121550558,
      "loss": 1.4102,
      "step": 207
    },
    {
      "epoch": 0.027292110874200425,
      "grad_norm": 113.88539123535156,
      "learning_rate": 0.00019141162779346874,
      "loss": 1.4203,
      "step": 208
    },
    {
      "epoch": 0.027423322945711006,
      "grad_norm": 98.4113540649414,
      "learning_rate": 0.00019132593705547082,
      "loss": 1.2918,
      "step": 209
    },
    {
      "epoch": 0.027554535017221583,
      "grad_norm": 96.7305679321289,
      "learning_rate": 0.00019123984032200586,
      "loss": 1.4801,
      "step": 210
    },
    {
      "epoch": 0.027685747088732164,
      "grad_norm": 108.2973861694336,
      "learning_rate": 0.00019115333797582254,
      "loss": 1.6495,
      "step": 211
    },
    {
      "epoch": 0.02781695916024274,
      "grad_norm": 103.74254608154297,
      "learning_rate": 0.00019106643040147278,
      "loss": 1.5629,
      "step": 212
    },
    {
      "epoch": 0.027948171231753322,
      "grad_norm": 91.67657470703125,
      "learning_rate": 0.00019097911798530987,
      "loss": 1.5293,
      "step": 213
    },
    {
      "epoch": 0.0280793833032639,
      "grad_norm": 94.97773742675781,
      "learning_rate": 0.00019089140111548696,
      "loss": 1.3532,
      "step": 214
    },
    {
      "epoch": 0.02821059537477448,
      "grad_norm": 91.32857513427734,
      "learning_rate": 0.00019080328018195513,
      "loss": 1.3252,
      "step": 215
    },
    {
      "epoch": 0.028341807446285058,
      "grad_norm": 104.70953369140625,
      "learning_rate": 0.0001907147555764618,
      "loss": 1.6138,
      "step": 216
    },
    {
      "epoch": 0.02847301951779564,
      "grad_norm": 120.06690979003906,
      "learning_rate": 0.00019062582769254895,
      "loss": 1.4544,
      "step": 217
    },
    {
      "epoch": 0.028604231589306216,
      "grad_norm": 110.91378021240234,
      "learning_rate": 0.00019053649692555135,
      "loss": 1.5195,
      "step": 218
    },
    {
      "epoch": 0.028735443660816797,
      "grad_norm": 85.34494018554688,
      "learning_rate": 0.00019044676367259476,
      "loss": 1.1306,
      "step": 219
    },
    {
      "epoch": 0.028866655732327374,
      "grad_norm": 100.43746948242188,
      "learning_rate": 0.00019035662833259432,
      "loss": 1.5023,
      "step": 220
    },
    {
      "epoch": 0.028997867803837955,
      "grad_norm": 94.84941101074219,
      "learning_rate": 0.00019026609130625257,
      "loss": 1.4352,
      "step": 221
    },
    {
      "epoch": 0.029129079875348532,
      "grad_norm": 108.02022552490234,
      "learning_rate": 0.00019017515299605788,
      "loss": 1.644,
      "step": 222
    },
    {
      "epoch": 0.029260291946859113,
      "grad_norm": 112.58411407470703,
      "learning_rate": 0.00019008381380628247,
      "loss": 1.5665,
      "step": 223
    },
    {
      "epoch": 0.02939150401836969,
      "grad_norm": 95.44322967529297,
      "learning_rate": 0.00018999207414298067,
      "loss": 1.3208,
      "step": 224
    },
    {
      "epoch": 0.029522716089880267,
      "grad_norm": 115.32520294189453,
      "learning_rate": 0.00018989993441398726,
      "loss": 1.3438,
      "step": 225
    },
    {
      "epoch": 0.029653928161390848,
      "grad_norm": 124.8680419921875,
      "learning_rate": 0.00018980739502891546,
      "loss": 1.6964,
      "step": 226
    },
    {
      "epoch": 0.029785140232901425,
      "grad_norm": 116.53189849853516,
      "learning_rate": 0.0001897144563991552,
      "loss": 1.5132,
      "step": 227
    },
    {
      "epoch": 0.029916352304412006,
      "grad_norm": 115.5568618774414,
      "learning_rate": 0.00018962111893787128,
      "loss": 1.5185,
      "step": 228
    },
    {
      "epoch": 0.030047564375922584,
      "grad_norm": 102.201171875,
      "learning_rate": 0.00018952738306000151,
      "loss": 1.4882,
      "step": 229
    },
    {
      "epoch": 0.030178776447433164,
      "grad_norm": 94.33916473388672,
      "learning_rate": 0.00018943324918225494,
      "loss": 1.1961,
      "step": 230
    },
    {
      "epoch": 0.03030998851894374,
      "grad_norm": 102.50297546386719,
      "learning_rate": 0.0001893387177231099,
      "loss": 1.568,
      "step": 231
    },
    {
      "epoch": 0.030441200590454322,
      "grad_norm": 95.40782928466797,
      "learning_rate": 0.0001892437891028122,
      "loss": 1.5973,
      "step": 232
    },
    {
      "epoch": 0.0305724126619649,
      "grad_norm": 94.1482925415039,
      "learning_rate": 0.0001891484637433733,
      "loss": 1.2577,
      "step": 233
    },
    {
      "epoch": 0.03070362473347548,
      "grad_norm": 102.69287109375,
      "learning_rate": 0.00018905274206856837,
      "loss": 1.2885,
      "step": 234
    },
    {
      "epoch": 0.030834836804986058,
      "grad_norm": 112.06835174560547,
      "learning_rate": 0.00018895662450393438,
      "loss": 1.3743,
      "step": 235
    },
    {
      "epoch": 0.03096604887649664,
      "grad_norm": 85.5040283203125,
      "learning_rate": 0.00018886011147676833,
      "loss": 1.1266,
      "step": 236
    },
    {
      "epoch": 0.031097260948007216,
      "grad_norm": 117.5611801147461,
      "learning_rate": 0.00018876320341612522,
      "loss": 1.5537,
      "step": 237
    },
    {
      "epoch": 0.031228473019517797,
      "grad_norm": 101.97940063476562,
      "learning_rate": 0.00018866590075281624,
      "loss": 1.3272,
      "step": 238
    },
    {
      "epoch": 0.03135968509102838,
      "grad_norm": 93.10289764404297,
      "learning_rate": 0.00018856820391940674,
      "loss": 1.1272,
      "step": 239
    },
    {
      "epoch": 0.03149089716253895,
      "grad_norm": 92.23869323730469,
      "learning_rate": 0.00018847011335021449,
      "loss": 1.3588,
      "step": 240
    },
    {
      "epoch": 0.03162210923404953,
      "grad_norm": 105.07022094726562,
      "learning_rate": 0.00018837162948130752,
      "loss": 1.4837,
      "step": 241
    },
    {
      "epoch": 0.03175332130556011,
      "grad_norm": 98.93236541748047,
      "learning_rate": 0.00018827275275050233,
      "loss": 1.2139,
      "step": 242
    },
    {
      "epoch": 0.031884533377070694,
      "grad_norm": 124.20883178710938,
      "learning_rate": 0.00018817348359736203,
      "loss": 1.4702,
      "step": 243
    },
    {
      "epoch": 0.03201574544858127,
      "grad_norm": 106.99666595458984,
      "learning_rate": 0.00018807382246319412,
      "loss": 1.6555,
      "step": 244
    },
    {
      "epoch": 0.03214695752009185,
      "grad_norm": 118.69978332519531,
      "learning_rate": 0.00018797376979104872,
      "loss": 1.6562,
      "step": 245
    },
    {
      "epoch": 0.03227816959160243,
      "grad_norm": 117.61981201171875,
      "learning_rate": 0.00018787332602571662,
      "loss": 1.5774,
      "step": 246
    },
    {
      "epoch": 0.032409381663113,
      "grad_norm": 136.1822967529297,
      "learning_rate": 0.00018777249161372713,
      "loss": 1.5804,
      "step": 247
    },
    {
      "epoch": 0.032540593734623584,
      "grad_norm": 90.79682922363281,
      "learning_rate": 0.00018767126700334634,
      "loss": 0.9809,
      "step": 248
    },
    {
      "epoch": 0.032671805806134165,
      "grad_norm": 136.71824645996094,
      "learning_rate": 0.0001875696526445749,
      "loss": 1.5187,
      "step": 249
    },
    {
      "epoch": 0.032803017877644745,
      "grad_norm": 122.1123046875,
      "learning_rate": 0.0001874676489891461,
      "loss": 1.3321,
      "step": 250
    },
    {
      "epoch": 0.03293422994915532,
      "grad_norm": 95.92542266845703,
      "learning_rate": 0.00018736525649052394,
      "loss": 1.3213,
      "step": 251
    },
    {
      "epoch": 0.0330654420206659,
      "grad_norm": 108.99664306640625,
      "learning_rate": 0.00018726247560390099,
      "loss": 1.6465,
      "step": 252
    },
    {
      "epoch": 0.03319665409217648,
      "grad_norm": 94.75456237792969,
      "learning_rate": 0.00018715930678619644,
      "loss": 1.1544,
      "step": 253
    },
    {
      "epoch": 0.03332786616368706,
      "grad_norm": 106.3393325805664,
      "learning_rate": 0.00018705575049605413,
      "loss": 1.4198,
      "step": 254
    },
    {
      "epoch": 0.033459078235197635,
      "grad_norm": 96.00926971435547,
      "learning_rate": 0.00018695180719384029,
      "loss": 1.5303,
      "step": 255
    },
    {
      "epoch": 0.033590290306708216,
      "grad_norm": 98.3309326171875,
      "learning_rate": 0.00018684747734164177,
      "loss": 1.5487,
      "step": 256
    },
    {
      "epoch": 0.0337215023782188,
      "grad_norm": 100.9000473022461,
      "learning_rate": 0.00018674276140326376,
      "loss": 1.6227,
      "step": 257
    },
    {
      "epoch": 0.03385271444972938,
      "grad_norm": 108.58922576904297,
      "learning_rate": 0.00018663765984422786,
      "loss": 1.3993,
      "step": 258
    },
    {
      "epoch": 0.03398392652123995,
      "grad_norm": 132.26473999023438,
      "learning_rate": 0.00018653217313177004,
      "loss": 1.5725,
      "step": 259
    },
    {
      "epoch": 0.03411513859275053,
      "grad_norm": 93.20362854003906,
      "learning_rate": 0.00018642630173483832,
      "loss": 1.2196,
      "step": 260
    },
    {
      "epoch": 0.03424635066426111,
      "grad_norm": 105.06778717041016,
      "learning_rate": 0.00018632004612409103,
      "loss": 1.4638,
      "step": 261
    },
    {
      "epoch": 0.034377562735771694,
      "grad_norm": 105.44284057617188,
      "learning_rate": 0.00018621340677189453,
      "loss": 1.4997,
      "step": 262
    },
    {
      "epoch": 0.03450877480728227,
      "grad_norm": 107.55557250976562,
      "learning_rate": 0.00018610638415232097,
      "loss": 1.3686,
      "step": 263
    },
    {
      "epoch": 0.03463998687879285,
      "grad_norm": 108.08594512939453,
      "learning_rate": 0.00018599897874114652,
      "loss": 1.5203,
      "step": 264
    },
    {
      "epoch": 0.03477119895030343,
      "grad_norm": 116.91567993164062,
      "learning_rate": 0.00018589119101584898,
      "loss": 1.5998,
      "step": 265
    },
    {
      "epoch": 0.03490241102181401,
      "grad_norm": 102.79871368408203,
      "learning_rate": 0.00018578302145560584,
      "loss": 1.5781,
      "step": 266
    },
    {
      "epoch": 0.035033623093324584,
      "grad_norm": 94.36803436279297,
      "learning_rate": 0.00018567447054129195,
      "loss": 1.1582,
      "step": 267
    },
    {
      "epoch": 0.035164835164835165,
      "grad_norm": 102.03521728515625,
      "learning_rate": 0.00018556553875547754,
      "loss": 1.446,
      "step": 268
    },
    {
      "epoch": 0.035296047236345746,
      "grad_norm": 111.63703918457031,
      "learning_rate": 0.00018545622658242607,
      "loss": 1.5893,
      "step": 269
    },
    {
      "epoch": 0.03542725930785632,
      "grad_norm": 76.933837890625,
      "learning_rate": 0.00018534653450809197,
      "loss": 1.1533,
      "step": 270
    },
    {
      "epoch": 0.0355584713793669,
      "grad_norm": 108.22994995117188,
      "learning_rate": 0.00018523646302011867,
      "loss": 1.8641,
      "step": 271
    },
    {
      "epoch": 0.03568968345087748,
      "grad_norm": 95.64000701904297,
      "learning_rate": 0.00018512601260783606,
      "loss": 1.4399,
      "step": 272
    },
    {
      "epoch": 0.03582089552238806,
      "grad_norm": 95.00650024414062,
      "learning_rate": 0.00018501518376225887,
      "loss": 1.3788,
      "step": 273
    },
    {
      "epoch": 0.035952107593898636,
      "grad_norm": 97.11001586914062,
      "learning_rate": 0.00018490397697608395,
      "loss": 1.5433,
      "step": 274
    },
    {
      "epoch": 0.036083319665409216,
      "grad_norm": 94.5434341430664,
      "learning_rate": 0.0001847923927436884,
      "loss": 1.4945,
      "step": 275
    },
    {
      "epoch": 0.0362145317369198,
      "grad_norm": 92.63153076171875,
      "learning_rate": 0.00018468043156112728,
      "loss": 1.4194,
      "step": 276
    },
    {
      "epoch": 0.03634574380843038,
      "grad_norm": 99.50650024414062,
      "learning_rate": 0.0001845680939261314,
      "loss": 1.3949,
      "step": 277
    },
    {
      "epoch": 0.03647695587994095,
      "grad_norm": 104.8680419921875,
      "learning_rate": 0.00018445538033810515,
      "loss": 1.4234,
      "step": 278
    },
    {
      "epoch": 0.03660816795145153,
      "grad_norm": 102.6010971069336,
      "learning_rate": 0.00018434229129812418,
      "loss": 1.259,
      "step": 279
    },
    {
      "epoch": 0.03673938002296211,
      "grad_norm": 83.17595672607422,
      "learning_rate": 0.0001842288273089332,
      "loss": 1.1907,
      "step": 280
    },
    {
      "epoch": 0.036870592094472694,
      "grad_norm": 97.24114990234375,
      "learning_rate": 0.00018411498887494396,
      "loss": 0.8478,
      "step": 281
    },
    {
      "epoch": 0.03700180416598327,
      "grad_norm": 112.60641479492188,
      "learning_rate": 0.00018400077650223263,
      "loss": 1.6612,
      "step": 282
    },
    {
      "epoch": 0.03713301623749385,
      "grad_norm": 107.87886810302734,
      "learning_rate": 0.0001838861906985379,
      "loss": 1.5732,
      "step": 283
    },
    {
      "epoch": 0.03726422830900443,
      "grad_norm": 93.29187774658203,
      "learning_rate": 0.00018377123197325842,
      "loss": 1.4329,
      "step": 284
    },
    {
      "epoch": 0.03739544038051501,
      "grad_norm": 89.42298126220703,
      "learning_rate": 0.00018365590083745085,
      "loss": 1.483,
      "step": 285
    },
    {
      "epoch": 0.037526652452025584,
      "grad_norm": 92.40946960449219,
      "learning_rate": 0.00018354019780382735,
      "loss": 1.3736,
      "step": 286
    },
    {
      "epoch": 0.037657864523536165,
      "grad_norm": 84.22126770019531,
      "learning_rate": 0.0001834241233867533,
      "loss": 1.0015,
      "step": 287
    },
    {
      "epoch": 0.037789076595046746,
      "grad_norm": 117.8971939086914,
      "learning_rate": 0.00018330767810224524,
      "loss": 1.45,
      "step": 288
    },
    {
      "epoch": 0.03792028866655733,
      "grad_norm": 108.27301788330078,
      "learning_rate": 0.0001831908624679683,
      "loss": 1.4593,
      "step": 289
    },
    {
      "epoch": 0.0380515007380679,
      "grad_norm": 107.7318115234375,
      "learning_rate": 0.0001830736770032341,
      "loss": 1.3573,
      "step": 290
    },
    {
      "epoch": 0.03818271280957848,
      "grad_norm": 133.5298614501953,
      "learning_rate": 0.0001829561222289984,
      "loss": 1.4224,
      "step": 291
    },
    {
      "epoch": 0.03831392488108906,
      "grad_norm": 101.90618133544922,
      "learning_rate": 0.00018283819866785853,
      "loss": 1.5351,
      "step": 292
    },
    {
      "epoch": 0.038445136952599636,
      "grad_norm": 91.31472778320312,
      "learning_rate": 0.0001827199068440516,
      "loss": 1.4697,
      "step": 293
    },
    {
      "epoch": 0.03857634902411022,
      "grad_norm": 92.80451965332031,
      "learning_rate": 0.00018260124728345162,
      "loss": 1.5675,
      "step": 294
    },
    {
      "epoch": 0.0387075610956208,
      "grad_norm": 105.6081771850586,
      "learning_rate": 0.00018248222051356754,
      "loss": 1.5253,
      "step": 295
    },
    {
      "epoch": 0.03883877316713138,
      "grad_norm": 110.62145233154297,
      "learning_rate": 0.00018236282706354063,
      "loss": 1.3081,
      "step": 296
    },
    {
      "epoch": 0.03896998523864195,
      "grad_norm": 119.20000457763672,
      "learning_rate": 0.00018224306746414238,
      "loss": 1.5794,
      "step": 297
    },
    {
      "epoch": 0.03910119731015253,
      "grad_norm": 162.8583221435547,
      "learning_rate": 0.00018212294224777197,
      "loss": 1.0243,
      "step": 298
    },
    {
      "epoch": 0.039232409381663114,
      "grad_norm": 101.95230865478516,
      "learning_rate": 0.00018200245194845399,
      "loss": 1.3681,
      "step": 299
    },
    {
      "epoch": 0.039363621453173694,
      "grad_norm": 95.07735443115234,
      "learning_rate": 0.00018188159710183594,
      "loss": 0.9938,
      "step": 300
    },
    {
      "epoch": 0.03949483352468427,
      "grad_norm": 100.39617156982422,
      "learning_rate": 0.000181760378245186,
      "loss": 1.3378,
      "step": 301
    },
    {
      "epoch": 0.03962604559619485,
      "grad_norm": 101.49337768554688,
      "learning_rate": 0.00018163879591739067,
      "loss": 1.6379,
      "step": 302
    },
    {
      "epoch": 0.03975725766770543,
      "grad_norm": 77.72740936279297,
      "learning_rate": 0.0001815168506589521,
      "loss": 1.0389,
      "step": 303
    },
    {
      "epoch": 0.03988846973921601,
      "grad_norm": 98.54853057861328,
      "learning_rate": 0.000181394543011986,
      "loss": 1.4934,
      "step": 304
    },
    {
      "epoch": 0.040019681810726584,
      "grad_norm": 87.42035675048828,
      "learning_rate": 0.00018127187352021907,
      "loss": 1.4806,
      "step": 305
    },
    {
      "epoch": 0.040150893882237165,
      "grad_norm": 93.95184326171875,
      "learning_rate": 0.0001811488427289866,
      "loss": 1.3515,
      "step": 306
    },
    {
      "epoch": 0.040282105953747746,
      "grad_norm": 102.16962432861328,
      "learning_rate": 0.00018102545118523007,
      "loss": 1.3886,
      "step": 307
    },
    {
      "epoch": 0.04041331802525833,
      "grad_norm": 99.6520767211914,
      "learning_rate": 0.00018090169943749476,
      "loss": 1.4265,
      "step": 308
    },
    {
      "epoch": 0.0405445300967689,
      "grad_norm": 105.71212768554688,
      "learning_rate": 0.00018077758803592718,
      "loss": 1.1927,
      "step": 309
    },
    {
      "epoch": 0.04067574216827948,
      "grad_norm": 110.70345306396484,
      "learning_rate": 0.00018065311753227273,
      "loss": 1.5832,
      "step": 310
    },
    {
      "epoch": 0.04080695423979006,
      "grad_norm": 110.64224243164062,
      "learning_rate": 0.0001805282884798732,
      "loss": 1.6852,
      "step": 311
    },
    {
      "epoch": 0.04093816631130064,
      "grad_norm": 91.91781616210938,
      "learning_rate": 0.00018040310143366446,
      "loss": 1.4352,
      "step": 312
    },
    {
      "epoch": 0.04106937838281122,
      "grad_norm": 114.888671875,
      "learning_rate": 0.00018027755695017368,
      "loss": 1.5753,
      "step": 313
    },
    {
      "epoch": 0.0412005904543218,
      "grad_norm": 95.87165069580078,
      "learning_rate": 0.00018015165558751717,
      "loss": 1.2975,
      "step": 314
    },
    {
      "epoch": 0.04133180252583238,
      "grad_norm": 110.02605438232422,
      "learning_rate": 0.00018002539790539773,
      "loss": 1.4754,
      "step": 315
    },
    {
      "epoch": 0.04146301459734295,
      "grad_norm": 102.34935760498047,
      "learning_rate": 0.00017989878446510215,
      "loss": 1.109,
      "step": 316
    },
    {
      "epoch": 0.04159422666885353,
      "grad_norm": 100.85880279541016,
      "learning_rate": 0.00017977181582949888,
      "loss": 1.4189,
      "step": 317
    },
    {
      "epoch": 0.041725438740364114,
      "grad_norm": 126.1025161743164,
      "learning_rate": 0.0001796444925630353,
      "loss": 1.6202,
      "step": 318
    },
    {
      "epoch": 0.041856650811874695,
      "grad_norm": 103.51316833496094,
      "learning_rate": 0.00017951681523173542,
      "loss": 1.4254,
      "step": 319
    },
    {
      "epoch": 0.04198786288338527,
      "grad_norm": 89.07659149169922,
      "learning_rate": 0.0001793887844031972,
      "loss": 1.1386,
      "step": 320
    },
    {
      "epoch": 0.04211907495489585,
      "grad_norm": 91.47647094726562,
      "learning_rate": 0.00017926040064659014,
      "loss": 1.5887,
      "step": 321
    },
    {
      "epoch": 0.04225028702640643,
      "grad_norm": 102.70648956298828,
      "learning_rate": 0.0001791316645326526,
      "loss": 1.4146,
      "step": 322
    },
    {
      "epoch": 0.04238149909791701,
      "grad_norm": 90.82884979248047,
      "learning_rate": 0.00017900257663368963,
      "loss": 1.348,
      "step": 323
    },
    {
      "epoch": 0.042512711169427585,
      "grad_norm": 129.57290649414062,
      "learning_rate": 0.0001788731375235698,
      "loss": 1.8249,
      "step": 324
    },
    {
      "epoch": 0.042643923240938165,
      "grad_norm": 112.32279968261719,
      "learning_rate": 0.00017874334777772327,
      "loss": 1.6182,
      "step": 325
    },
    {
      "epoch": 0.042775135312448746,
      "grad_norm": 107.3589859008789,
      "learning_rate": 0.00017861320797313892,
      "loss": 1.4192,
      "step": 326
    },
    {
      "epoch": 0.04290634738395933,
      "grad_norm": 97.4294662475586,
      "learning_rate": 0.0001784827186883618,
      "loss": 1.459,
      "step": 327
    },
    {
      "epoch": 0.0430375594554699,
      "grad_norm": 89.63394927978516,
      "learning_rate": 0.00017835188050349064,
      "loss": 1.342,
      "step": 328
    },
    {
      "epoch": 0.04316877152698048,
      "grad_norm": 104.56636810302734,
      "learning_rate": 0.00017822069400017516,
      "loss": 1.1993,
      "step": 329
    },
    {
      "epoch": 0.04329998359849106,
      "grad_norm": 101.56778717041016,
      "learning_rate": 0.00017808915976161362,
      "loss": 1.6903,
      "step": 330
    },
    {
      "epoch": 0.04343119567000164,
      "grad_norm": 87.88162994384766,
      "learning_rate": 0.00017795727837255015,
      "loss": 1.2707,
      "step": 331
    },
    {
      "epoch": 0.04356240774151222,
      "grad_norm": 109.85099029541016,
      "learning_rate": 0.00017782505041927216,
      "loss": 1.4304,
      "step": 332
    },
    {
      "epoch": 0.0436936198130228,
      "grad_norm": 107.29802703857422,
      "learning_rate": 0.00017769247648960774,
      "loss": 1.4053,
      "step": 333
    },
    {
      "epoch": 0.04382483188453338,
      "grad_norm": 115.34333038330078,
      "learning_rate": 0.00017755955717292296,
      "loss": 1.5417,
      "step": 334
    },
    {
      "epoch": 0.04395604395604396,
      "grad_norm": 93.28298950195312,
      "learning_rate": 0.00017742629306011944,
      "loss": 0.8773,
      "step": 335
    },
    {
      "epoch": 0.04408725602755453,
      "grad_norm": 90.39668273925781,
      "learning_rate": 0.00017729268474363154,
      "loss": 1.2491,
      "step": 336
    },
    {
      "epoch": 0.044218468099065114,
      "grad_norm": 114.6083755493164,
      "learning_rate": 0.0001771587328174239,
      "loss": 1.2966,
      "step": 337
    },
    {
      "epoch": 0.044349680170575695,
      "grad_norm": 115.39427947998047,
      "learning_rate": 0.0001770244378769885,
      "loss": 1.4295,
      "step": 338
    },
    {
      "epoch": 0.04448089224208627,
      "grad_norm": 111.20867919921875,
      "learning_rate": 0.0001768898005193425,
      "loss": 1.3819,
      "step": 339
    },
    {
      "epoch": 0.04461210431359685,
      "grad_norm": 120.4778823852539,
      "learning_rate": 0.000176754821343025,
      "loss": 1.2191,
      "step": 340
    },
    {
      "epoch": 0.04474331638510743,
      "grad_norm": 110.4124755859375,
      "learning_rate": 0.0001766195009480949,
      "loss": 1.4072,
      "step": 341
    },
    {
      "epoch": 0.04487452845661801,
      "grad_norm": 92.29596710205078,
      "learning_rate": 0.0001764838399361279,
      "loss": 1.6048,
      "step": 342
    },
    {
      "epoch": 0.045005740528128585,
      "grad_norm": 157.63006591796875,
      "learning_rate": 0.00017634783891021393,
      "loss": 1.349,
      "step": 343
    },
    {
      "epoch": 0.045136952599639166,
      "grad_norm": 108.58782958984375,
      "learning_rate": 0.00017621149847495458,
      "loss": 1.5907,
      "step": 344
    },
    {
      "epoch": 0.045268164671149747,
      "grad_norm": 99.58897399902344,
      "learning_rate": 0.00017607481923646016,
      "loss": 1.5153,
      "step": 345
    },
    {
      "epoch": 0.04539937674266033,
      "grad_norm": 104.05718231201172,
      "learning_rate": 0.0001759378018023473,
      "loss": 1.3563,
      "step": 346
    },
    {
      "epoch": 0.0455305888141709,
      "grad_norm": 83.61771392822266,
      "learning_rate": 0.00017580044678173592,
      "loss": 1.1481,
      "step": 347
    },
    {
      "epoch": 0.04566180088568148,
      "grad_norm": 128.92269897460938,
      "learning_rate": 0.00017566275478524693,
      "loss": 1.238,
      "step": 348
    },
    {
      "epoch": 0.04579301295719206,
      "grad_norm": 124.21028137207031,
      "learning_rate": 0.0001755247264249991,
      "loss": 1.5199,
      "step": 349
    },
    {
      "epoch": 0.045924225028702644,
      "grad_norm": 107.28356170654297,
      "learning_rate": 0.0001753863623146066,
      "loss": 1.0621,
      "step": 350
    },
    {
      "epoch": 0.04605543710021322,
      "grad_norm": 109.68620300292969,
      "learning_rate": 0.00017524766306917618,
      "loss": 1.0921,
      "step": 351
    },
    {
      "epoch": 0.0461866491717238,
      "grad_norm": 100.94927978515625,
      "learning_rate": 0.0001751086293053045,
      "loss": 1.4859,
      "step": 352
    },
    {
      "epoch": 0.04631786124323438,
      "grad_norm": 125.20081329345703,
      "learning_rate": 0.0001749692616410753,
      "loss": 1.3746,
      "step": 353
    },
    {
      "epoch": 0.04644907331474496,
      "grad_norm": 92.83186340332031,
      "learning_rate": 0.00017482956069605668,
      "loss": 1.4385,
      "step": 354
    },
    {
      "epoch": 0.046580285386255534,
      "grad_norm": 111.00030517578125,
      "learning_rate": 0.00017468952709129846,
      "loss": 1.7226,
      "step": 355
    },
    {
      "epoch": 0.046711497457766114,
      "grad_norm": 89.47738647460938,
      "learning_rate": 0.00017454916144932922,
      "loss": 1.2583,
      "step": 356
    },
    {
      "epoch": 0.046842709529276695,
      "grad_norm": 86.99544525146484,
      "learning_rate": 0.0001744084643941536,
      "loss": 1.4868,
      "step": 357
    },
    {
      "epoch": 0.04697392160078727,
      "grad_norm": 97.1447982788086,
      "learning_rate": 0.00017426743655124974,
      "loss": 1.3634,
      "step": 358
    },
    {
      "epoch": 0.04710513367229785,
      "grad_norm": 105.50357818603516,
      "learning_rate": 0.0001741260785475661,
      "loss": 1.5933,
      "step": 359
    },
    {
      "epoch": 0.04723634574380843,
      "grad_norm": 95.43667602539062,
      "learning_rate": 0.00017398439101151905,
      "loss": 1.3294,
      "step": 360
    },
    {
      "epoch": 0.04736755781531901,
      "grad_norm": 107.23646545410156,
      "learning_rate": 0.00017384237457298987,
      "loss": 1.2716,
      "step": 361
    },
    {
      "epoch": 0.047498769886829585,
      "grad_norm": 93.79784393310547,
      "learning_rate": 0.00017370002986332193,
      "loss": 1.3088,
      "step": 362
    },
    {
      "epoch": 0.047629981958340166,
      "grad_norm": 88.74073791503906,
      "learning_rate": 0.00017355735751531807,
      "loss": 1.3845,
      "step": 363
    },
    {
      "epoch": 0.04776119402985075,
      "grad_norm": 109.1597900390625,
      "learning_rate": 0.00017341435816323756,
      "loss": 1.5577,
      "step": 364
    },
    {
      "epoch": 0.04789240610136133,
      "grad_norm": 90.45240783691406,
      "learning_rate": 0.00017327103244279348,
      "loss": 1.3844,
      "step": 365
    },
    {
      "epoch": 0.0480236181728719,
      "grad_norm": 93.51265716552734,
      "learning_rate": 0.00017312738099114973,
      "loss": 1.2208,
      "step": 366
    },
    {
      "epoch": 0.04815483024438248,
      "grad_norm": 95.80655670166016,
      "learning_rate": 0.00017298340444691835,
      "loss": 1.4481,
      "step": 367
    },
    {
      "epoch": 0.04828604231589306,
      "grad_norm": 100.99742126464844,
      "learning_rate": 0.00017283910345015647,
      "loss": 1.5903,
      "step": 368
    },
    {
      "epoch": 0.048417254387403644,
      "grad_norm": 96.02694702148438,
      "learning_rate": 0.0001726944786423637,
      "loss": 1.5153,
      "step": 369
    },
    {
      "epoch": 0.04854846645891422,
      "grad_norm": 95.6656265258789,
      "learning_rate": 0.00017254953066647913,
      "loss": 1.2167,
      "step": 370
    },
    {
      "epoch": 0.0486796785304248,
      "grad_norm": 99.86067199707031,
      "learning_rate": 0.00017240426016687863,
      "loss": 1.538,
      "step": 371
    },
    {
      "epoch": 0.04881089060193538,
      "grad_norm": 116.48789978027344,
      "learning_rate": 0.00017225866778937165,
      "loss": 1.5232,
      "step": 372
    },
    {
      "epoch": 0.04894210267344596,
      "grad_norm": 116.97354888916016,
      "learning_rate": 0.00017211275418119876,
      "loss": 1.3433,
      "step": 373
    },
    {
      "epoch": 0.049073314744956534,
      "grad_norm": 98.1113052368164,
      "learning_rate": 0.0001719665199910285,
      "loss": 1.3907,
      "step": 374
    },
    {
      "epoch": 0.049204526816467115,
      "grad_norm": 112.6119613647461,
      "learning_rate": 0.00017181996586895454,
      "loss": 1.6813,
      "step": 375
    },
    {
      "epoch": 0.049204526816467115,
      "eval_loss": 1.3800705671310425,
      "eval_runtime": 760.3009,
      "eval_samples_per_second": 16.883,
      "eval_steps_per_second": 8.441,
      "step": 375
    },
    {
      "epoch": 0.049335738887977695,
      "grad_norm": 96.22384643554688,
      "learning_rate": 0.00017167309246649297,
      "loss": 1.5219,
      "step": 376
    },
    {
      "epoch": 0.049466950959488276,
      "grad_norm": 91.31690216064453,
      "learning_rate": 0.0001715259004365791,
      "loss": 1.5053,
      "step": 377
    },
    {
      "epoch": 0.04959816303099885,
      "grad_norm": 102.95918273925781,
      "learning_rate": 0.00017137839043356484,
      "loss": 1.4379,
      "step": 378
    },
    {
      "epoch": 0.04972937510250943,
      "grad_norm": 89.1133804321289,
      "learning_rate": 0.00017123056311321562,
      "loss": 1.178,
      "step": 379
    },
    {
      "epoch": 0.04986058717402001,
      "grad_norm": 95.50212860107422,
      "learning_rate": 0.0001710824191327075,
      "loss": 1.0833,
      "step": 380
    },
    {
      "epoch": 0.049991799245530585,
      "grad_norm": 110.51009368896484,
      "learning_rate": 0.00017093395915062428,
      "loss": 1.6214,
      "step": 381
    },
    {
      "epoch": 0.050123011317041166,
      "grad_norm": 105.03572082519531,
      "learning_rate": 0.00017078518382695465,
      "loss": 1.3617,
      "step": 382
    },
    {
      "epoch": 0.05025422338855175,
      "grad_norm": 102.54866027832031,
      "learning_rate": 0.00017063609382308908,
      "loss": 1.5317,
      "step": 383
    },
    {
      "epoch": 0.05038543546006233,
      "grad_norm": 179.09857177734375,
      "learning_rate": 0.00017048668980181698,
      "loss": 1.3749,
      "step": 384
    },
    {
      "epoch": 0.0505166475315729,
      "grad_norm": 89.51898956298828,
      "learning_rate": 0.00017033697242732377,
      "loss": 1.0856,
      "step": 385
    },
    {
      "epoch": 0.05064785960308348,
      "grad_norm": 94.99905395507812,
      "learning_rate": 0.0001701869423651879,
      "loss": 1.3213,
      "step": 386
    },
    {
      "epoch": 0.05077907167459406,
      "grad_norm": 90.60092163085938,
      "learning_rate": 0.00017003660028237793,
      "loss": 1.3796,
      "step": 387
    },
    {
      "epoch": 0.050910283746104644,
      "grad_norm": 83.35353088378906,
      "learning_rate": 0.00016988594684724947,
      "loss": 1.1707,
      "step": 388
    },
    {
      "epoch": 0.05104149581761522,
      "grad_norm": 91.23401641845703,
      "learning_rate": 0.00016973498272954222,
      "loss": 1.3189,
      "step": 389
    },
    {
      "epoch": 0.0511727078891258,
      "grad_norm": 96.93355560302734,
      "learning_rate": 0.00016958370860037717,
      "loss": 1.3392,
      "step": 390
    },
    {
      "epoch": 0.05130391996063638,
      "grad_norm": 111.30208587646484,
      "learning_rate": 0.00016943212513225345,
      "loss": 1.6205,
      "step": 391
    },
    {
      "epoch": 0.05143513203214696,
      "grad_norm": 98.86795806884766,
      "learning_rate": 0.00016928023299904533,
      "loss": 1.1821,
      "step": 392
    },
    {
      "epoch": 0.051566344103657534,
      "grad_norm": 95.3832015991211,
      "learning_rate": 0.0001691280328759992,
      "loss": 1.393,
      "step": 393
    },
    {
      "epoch": 0.051697556175168115,
      "grad_norm": 94.64776611328125,
      "learning_rate": 0.00016897552543973084,
      "loss": 1.3263,
      "step": 394
    },
    {
      "epoch": 0.051828768246678696,
      "grad_norm": 110.38654327392578,
      "learning_rate": 0.00016882271136822206,
      "loss": 1.4799,
      "step": 395
    },
    {
      "epoch": 0.051959980318189276,
      "grad_norm": 97.55098724365234,
      "learning_rate": 0.0001686695913408179,
      "loss": 1.4768,
      "step": 396
    },
    {
      "epoch": 0.05209119238969985,
      "grad_norm": 109.0049057006836,
      "learning_rate": 0.0001685161660382235,
      "loss": 1.2331,
      "step": 397
    },
    {
      "epoch": 0.05222240446121043,
      "grad_norm": 113.52181243896484,
      "learning_rate": 0.00016836243614250113,
      "loss": 1.2568,
      "step": 398
    },
    {
      "epoch": 0.05235361653272101,
      "grad_norm": 115.9555892944336,
      "learning_rate": 0.00016820840233706719,
      "loss": 1.4199,
      "step": 399
    },
    {
      "epoch": 0.05248482860423159,
      "grad_norm": 105.11410522460938,
      "learning_rate": 0.0001680540653066891,
      "loss": 1.0462,
      "step": 400
    },
    {
      "epoch": 0.052616040675742166,
      "grad_norm": 95.82889556884766,
      "learning_rate": 0.00016789942573748232,
      "loss": 1.2901,
      "step": 401
    },
    {
      "epoch": 0.05274725274725275,
      "grad_norm": 92.86366271972656,
      "learning_rate": 0.0001677444843169072,
      "loss": 1.5465,
      "step": 402
    },
    {
      "epoch": 0.05287846481876333,
      "grad_norm": 85.53275299072266,
      "learning_rate": 0.00016758924173376603,
      "loss": 1.2278,
      "step": 403
    },
    {
      "epoch": 0.0530096768902739,
      "grad_norm": 85.20893859863281,
      "learning_rate": 0.0001674336986781999,
      "loss": 1.2279,
      "step": 404
    },
    {
      "epoch": 0.05314088896178448,
      "grad_norm": 92.3574447631836,
      "learning_rate": 0.00016727785584168581,
      "loss": 1.5276,
      "step": 405
    },
    {
      "epoch": 0.05327210103329506,
      "grad_norm": 94.9033203125,
      "learning_rate": 0.0001671217139170333,
      "loss": 1.4808,
      "step": 406
    },
    {
      "epoch": 0.053403313104805644,
      "grad_norm": 89.22769165039062,
      "learning_rate": 0.00016696527359838154,
      "loss": 1.3736,
      "step": 407
    },
    {
      "epoch": 0.05353452517631622,
      "grad_norm": 101.15306091308594,
      "learning_rate": 0.00016680853558119632,
      "loss": 1.5808,
      "step": 408
    },
    {
      "epoch": 0.0536657372478268,
      "grad_norm": 80.92437744140625,
      "learning_rate": 0.0001666515005622668,
      "loss": 1.0454,
      "step": 409
    },
    {
      "epoch": 0.05379694931933738,
      "grad_norm": 98.29462432861328,
      "learning_rate": 0.0001664941692397025,
      "loss": 1.3269,
      "step": 410
    },
    {
      "epoch": 0.05392816139084796,
      "grad_norm": 95.68324279785156,
      "learning_rate": 0.00016633654231293013,
      "loss": 1.4351,
      "step": 411
    },
    {
      "epoch": 0.054059373462358534,
      "grad_norm": 87.02562713623047,
      "learning_rate": 0.00016617862048269065,
      "loss": 1.2922,
      "step": 412
    },
    {
      "epoch": 0.054190585533869115,
      "grad_norm": 94.31806182861328,
      "learning_rate": 0.00016602040445103588,
      "loss": 1.5084,
      "step": 413
    },
    {
      "epoch": 0.054321797605379696,
      "grad_norm": 104.38877868652344,
      "learning_rate": 0.00016586189492132566,
      "loss": 1.4142,
      "step": 414
    },
    {
      "epoch": 0.05445300967689028,
      "grad_norm": 97.92495727539062,
      "learning_rate": 0.00016570309259822453,
      "loss": 1.3735,
      "step": 415
    },
    {
      "epoch": 0.05458422174840085,
      "grad_norm": 102.59034729003906,
      "learning_rate": 0.0001655439981876987,
      "loss": 1.5029,
      "step": 416
    },
    {
      "epoch": 0.05471543381991143,
      "grad_norm": 87.60053253173828,
      "learning_rate": 0.00016538461239701277,
      "loss": 1.1478,
      "step": 417
    },
    {
      "epoch": 0.05484664589142201,
      "grad_norm": 98.76835632324219,
      "learning_rate": 0.00016522493593472683,
      "loss": 1.308,
      "step": 418
    },
    {
      "epoch": 0.05497785796293259,
      "grad_norm": 100.94078826904297,
      "learning_rate": 0.0001650649695106931,
      "loss": 1.4184,
      "step": 419
    },
    {
      "epoch": 0.05510907003444317,
      "grad_norm": 88.8556900024414,
      "learning_rate": 0.00016490471383605288,
      "loss": 1.2318,
      "step": 420
    },
    {
      "epoch": 0.05524028210595375,
      "grad_norm": 115.70670318603516,
      "learning_rate": 0.00016474416962323325,
      "loss": 1.428,
      "step": 421
    },
    {
      "epoch": 0.05537149417746433,
      "grad_norm": 151.6595001220703,
      "learning_rate": 0.00016458333758594414,
      "loss": 1.4623,
      "step": 422
    },
    {
      "epoch": 0.05550270624897491,
      "grad_norm": 135.5734100341797,
      "learning_rate": 0.00016442221843917496,
      "loss": 1.4891,
      "step": 423
    },
    {
      "epoch": 0.05563391832048548,
      "grad_norm": 97.82564544677734,
      "learning_rate": 0.00016426081289919143,
      "loss": 1.3323,
      "step": 424
    },
    {
      "epoch": 0.055765130391996064,
      "grad_norm": 96.42183685302734,
      "learning_rate": 0.0001640991216835326,
      "loss": 1.4403,
      "step": 425
    },
    {
      "epoch": 0.055896342463506644,
      "grad_norm": 101.82233428955078,
      "learning_rate": 0.00016393714551100734,
      "loss": 1.4059,
      "step": 426
    },
    {
      "epoch": 0.05602755453501722,
      "grad_norm": 121.00505828857422,
      "learning_rate": 0.0001637748851016914,
      "loss": 1.4895,
      "step": 427
    },
    {
      "epoch": 0.0561587666065278,
      "grad_norm": 89.92058563232422,
      "learning_rate": 0.00016361234117692413,
      "loss": 1.3665,
      "step": 428
    },
    {
      "epoch": 0.05628997867803838,
      "grad_norm": 105.40972137451172,
      "learning_rate": 0.00016344951445930526,
      "loss": 1.5343,
      "step": 429
    },
    {
      "epoch": 0.05642119074954896,
      "grad_norm": 109.82108306884766,
      "learning_rate": 0.0001632864056726917,
      "loss": 1.636,
      "step": 430
    },
    {
      "epoch": 0.056552402821059534,
      "grad_norm": 76.5701675415039,
      "learning_rate": 0.00016312301554219426,
      "loss": 0.9097,
      "step": 431
    },
    {
      "epoch": 0.056683614892570115,
      "grad_norm": 93.67756652832031,
      "learning_rate": 0.00016295934479417453,
      "loss": 1.5279,
      "step": 432
    },
    {
      "epoch": 0.056814826964080696,
      "grad_norm": 95.57807922363281,
      "learning_rate": 0.00016279539415624164,
      "loss": 1.5105,
      "step": 433
    },
    {
      "epoch": 0.05694603903559128,
      "grad_norm": 92.7071304321289,
      "learning_rate": 0.0001626311643572489,
      "loss": 1.342,
      "step": 434
    },
    {
      "epoch": 0.05707725110710185,
      "grad_norm": 98.78939819335938,
      "learning_rate": 0.00016246665612729074,
      "loss": 1.5439,
      "step": 435
    },
    {
      "epoch": 0.05720846317861243,
      "grad_norm": 94.4900131225586,
      "learning_rate": 0.00016230187019769928,
      "loss": 1.337,
      "step": 436
    },
    {
      "epoch": 0.05733967525012301,
      "grad_norm": 110.52635955810547,
      "learning_rate": 0.00016213680730104124,
      "loss": 1.2025,
      "step": 437
    },
    {
      "epoch": 0.05747088732163359,
      "grad_norm": 112.14994049072266,
      "learning_rate": 0.0001619714681711146,
      "loss": 1.6216,
      "step": 438
    },
    {
      "epoch": 0.05760209939314417,
      "grad_norm": 106.85348510742188,
      "learning_rate": 0.00016180585354294536,
      "loss": 1.3506,
      "step": 439
    },
    {
      "epoch": 0.05773331146465475,
      "grad_norm": 83.6705551147461,
      "learning_rate": 0.00016163996415278424,
      "loss": 0.9114,
      "step": 440
    },
    {
      "epoch": 0.05786452353616533,
      "grad_norm": 97.01634979248047,
      "learning_rate": 0.00016147380073810346,
      "loss": 1.375,
      "step": 441
    },
    {
      "epoch": 0.05799573560767591,
      "grad_norm": 116.10235595703125,
      "learning_rate": 0.0001613073640375934,
      "loss": 1.3987,
      "step": 442
    },
    {
      "epoch": 0.05812694767918648,
      "grad_norm": 114.28963470458984,
      "learning_rate": 0.00016114065479115946,
      "loss": 1.3638,
      "step": 443
    },
    {
      "epoch": 0.058258159750697064,
      "grad_norm": 88.33671569824219,
      "learning_rate": 0.00016097367373991842,
      "loss": 1.2641,
      "step": 444
    },
    {
      "epoch": 0.058389371822207645,
      "grad_norm": 98.9401626586914,
      "learning_rate": 0.00016080642162619565,
      "loss": 1.3232,
      "step": 445
    },
    {
      "epoch": 0.058520583893718225,
      "grad_norm": 92.24010467529297,
      "learning_rate": 0.0001606388991935214,
      "loss": 1.2703,
      "step": 446
    },
    {
      "epoch": 0.0586517959652288,
      "grad_norm": 112.46167755126953,
      "learning_rate": 0.0001604711071866277,
      "loss": 1.4755,
      "step": 447
    },
    {
      "epoch": 0.05878300803673938,
      "grad_norm": 88.90274810791016,
      "learning_rate": 0.00016030304635144494,
      "loss": 1.1499,
      "step": 448
    },
    {
      "epoch": 0.05891422010824996,
      "grad_norm": 117.99269104003906,
      "learning_rate": 0.00016013471743509862,
      "loss": 1.6385,
      "step": 449
    },
    {
      "epoch": 0.059045432179760535,
      "grad_norm": 91.54945373535156,
      "learning_rate": 0.00015996612118590603,
      "loss": 1.0954,
      "step": 450
    },
    {
      "epoch": 0.059176644251271115,
      "grad_norm": 85.16179656982422,
      "learning_rate": 0.00015979725835337294,
      "loss": 1.3672,
      "step": 451
    },
    {
      "epoch": 0.059307856322781696,
      "grad_norm": 111.78656768798828,
      "learning_rate": 0.00015962812968819016,
      "loss": 1.3957,
      "step": 452
    },
    {
      "epoch": 0.05943906839429228,
      "grad_norm": 109.4968490600586,
      "learning_rate": 0.0001594587359422303,
      "loss": 1.2189,
      "step": 453
    },
    {
      "epoch": 0.05957028046580285,
      "grad_norm": 88.610107421875,
      "learning_rate": 0.0001592890778685444,
      "loss": 1.4252,
      "step": 454
    },
    {
      "epoch": 0.05970149253731343,
      "grad_norm": 97.8349609375,
      "learning_rate": 0.00015911915622135862,
      "loss": 1.4623,
      "step": 455
    },
    {
      "epoch": 0.05983270460882401,
      "grad_norm": 94.22692108154297,
      "learning_rate": 0.00015894897175607086,
      "loss": 1.2172,
      "step": 456
    },
    {
      "epoch": 0.05996391668033459,
      "grad_norm": 88.2801284790039,
      "learning_rate": 0.00015877852522924732,
      "loss": 1.471,
      "step": 457
    },
    {
      "epoch": 0.06009512875184517,
      "grad_norm": 100.88803100585938,
      "learning_rate": 0.00015860781739861928,
      "loss": 1.5061,
      "step": 458
    },
    {
      "epoch": 0.06022634082335575,
      "grad_norm": 94.9809799194336,
      "learning_rate": 0.00015843684902307962,
      "loss": 1.3981,
      "step": 459
    },
    {
      "epoch": 0.06035755289486633,
      "grad_norm": 108.68051147460938,
      "learning_rate": 0.00015826562086267956,
      "loss": 1.2599,
      "step": 460
    },
    {
      "epoch": 0.06048876496637691,
      "grad_norm": 91.68250274658203,
      "learning_rate": 0.00015809413367862512,
      "loss": 1.3566,
      "step": 461
    },
    {
      "epoch": 0.06061997703788748,
      "grad_norm": 101.87088775634766,
      "learning_rate": 0.00015792238823327388,
      "loss": 1.5045,
      "step": 462
    },
    {
      "epoch": 0.060751189109398064,
      "grad_norm": 102.60354614257812,
      "learning_rate": 0.00015775038529013152,
      "loss": 1.4966,
      "step": 463
    },
    {
      "epoch": 0.060882401180908645,
      "grad_norm": 96.80057525634766,
      "learning_rate": 0.0001575781256138485,
      "loss": 1.368,
      "step": 464
    },
    {
      "epoch": 0.061013613252419226,
      "grad_norm": 101.29257202148438,
      "learning_rate": 0.00015740560997021648,
      "loss": 1.4061,
      "step": 465
    },
    {
      "epoch": 0.0611448253239298,
      "grad_norm": 103.30950164794922,
      "learning_rate": 0.00015723283912616513,
      "loss": 1.471,
      "step": 466
    },
    {
      "epoch": 0.06127603739544038,
      "grad_norm": 103.24085998535156,
      "learning_rate": 0.00015705981384975866,
      "loss": 1.4403,
      "step": 467
    },
    {
      "epoch": 0.06140724946695096,
      "grad_norm": 91.03667449951172,
      "learning_rate": 0.0001568865349101923,
      "loss": 1.4934,
      "step": 468
    },
    {
      "epoch": 0.06153846153846154,
      "grad_norm": 94.49728393554688,
      "learning_rate": 0.00015671300307778898,
      "loss": 1.2673,
      "step": 469
    },
    {
      "epoch": 0.061669673609972116,
      "grad_norm": 78.22455596923828,
      "learning_rate": 0.00015653921912399589,
      "loss": 0.9639,
      "step": 470
    },
    {
      "epoch": 0.061800885681482697,
      "grad_norm": 98.11071014404297,
      "learning_rate": 0.00015636518382138107,
      "loss": 1.4333,
      "step": 471
    },
    {
      "epoch": 0.06193209775299328,
      "grad_norm": 109.14305877685547,
      "learning_rate": 0.0001561908979436299,
      "loss": 1.5689,
      "step": 472
    },
    {
      "epoch": 0.06206330982450385,
      "grad_norm": 100.83522033691406,
      "learning_rate": 0.00015601636226554168,
      "loss": 1.4422,
      "step": 473
    },
    {
      "epoch": 0.06219452189601443,
      "grad_norm": 102.82100677490234,
      "learning_rate": 0.00015584157756302634,
      "loss": 1.2241,
      "step": 474
    },
    {
      "epoch": 0.06232573396752501,
      "grad_norm": 109.54855346679688,
      "learning_rate": 0.0001556665446131007,
      "loss": 1.5539,
      "step": 475
    },
    {
      "epoch": 0.062456946039035593,
      "grad_norm": 84.4482421875,
      "learning_rate": 0.00015549126419388536,
      "loss": 1.2777,
      "step": 476
    },
    {
      "epoch": 0.06258815811054617,
      "grad_norm": 92.32514190673828,
      "learning_rate": 0.0001553157370846009,
      "loss": 1.2898,
      "step": 477
    },
    {
      "epoch": 0.06271937018205676,
      "grad_norm": 88.95817565917969,
      "learning_rate": 0.00015513996406556465,
      "loss": 1.17,
      "step": 478
    },
    {
      "epoch": 0.06285058225356732,
      "grad_norm": 95.21998596191406,
      "learning_rate": 0.00015496394591818716,
      "loss": 1.0812,
      "step": 479
    },
    {
      "epoch": 0.0629817943250779,
      "grad_norm": 96.45527648925781,
      "learning_rate": 0.0001547876834249687,
      "loss": 1.339,
      "step": 480
    },
    {
      "epoch": 0.06311300639658848,
      "grad_norm": 100.06550598144531,
      "learning_rate": 0.00015461117736949577,
      "loss": 1.3208,
      "step": 481
    },
    {
      "epoch": 0.06324421846809906,
      "grad_norm": 121.37667846679688,
      "learning_rate": 0.00015443442853643762,
      "loss": 1.5828,
      "step": 482
    },
    {
      "epoch": 0.06337543053960965,
      "grad_norm": 99.197998046875,
      "learning_rate": 0.00015425743771154294,
      "loss": 1.5254,
      "step": 483
    },
    {
      "epoch": 0.06350664261112023,
      "grad_norm": 82.0926742553711,
      "learning_rate": 0.00015408020568163602,
      "loss": 1.2595,
      "step": 484
    },
    {
      "epoch": 0.0636378546826308,
      "grad_norm": 78.20893859863281,
      "learning_rate": 0.00015390273323461352,
      "loss": 1.2667,
      "step": 485
    },
    {
      "epoch": 0.06376906675414139,
      "grad_norm": 82.5278091430664,
      "learning_rate": 0.0001537250211594409,
      "loss": 1.2537,
      "step": 486
    },
    {
      "epoch": 0.06390027882565195,
      "grad_norm": 91.94507598876953,
      "learning_rate": 0.0001535470702461489,
      "loss": 1.4589,
      "step": 487
    },
    {
      "epoch": 0.06403149089716254,
      "grad_norm": 86.41954040527344,
      "learning_rate": 0.00015336888128583,
      "loss": 1.238,
      "step": 488
    },
    {
      "epoch": 0.06416270296867312,
      "grad_norm": 99.39994812011719,
      "learning_rate": 0.000153190455070635,
      "loss": 1.332,
      "step": 489
    },
    {
      "epoch": 0.0642939150401837,
      "grad_norm": 103.71652221679688,
      "learning_rate": 0.00015301179239376938,
      "loss": 1.3798,
      "step": 490
    },
    {
      "epoch": 0.06442512711169428,
      "grad_norm": 89.61128234863281,
      "learning_rate": 0.00015283289404948976,
      "loss": 1.2469,
      "step": 491
    },
    {
      "epoch": 0.06455633918320486,
      "grad_norm": 106.24605560302734,
      "learning_rate": 0.0001526537608331006,
      "loss": 1.2132,
      "step": 492
    },
    {
      "epoch": 0.06468755125471544,
      "grad_norm": 105.75811004638672,
      "learning_rate": 0.00015247439354095041,
      "loss": 1.2991,
      "step": 493
    },
    {
      "epoch": 0.064818763326226,
      "grad_norm": 105.74169158935547,
      "learning_rate": 0.00015229479297042823,
      "loss": 1.4383,
      "step": 494
    },
    {
      "epoch": 0.06494997539773659,
      "grad_norm": 97.50747680664062,
      "learning_rate": 0.00015211495991996027,
      "loss": 1.3649,
      "step": 495
    },
    {
      "epoch": 0.06508118746924717,
      "grad_norm": 158.67910766601562,
      "learning_rate": 0.0001519348951890062,
      "loss": 1.5641,
      "step": 496
    },
    {
      "epoch": 0.06521239954075775,
      "grad_norm": 91.19139862060547,
      "learning_rate": 0.0001517545995780556,
      "loss": 1.4598,
      "step": 497
    },
    {
      "epoch": 0.06534361161226833,
      "grad_norm": 83.49527740478516,
      "learning_rate": 0.00015157407388862452,
      "loss": 1.0082,
      "step": 498
    },
    {
      "epoch": 0.06547482368377891,
      "grad_norm": 93.00768280029297,
      "learning_rate": 0.00015139331892325179,
      "loss": 1.2751,
      "step": 499
    },
    {
      "epoch": 0.06560603575528949,
      "grad_norm": 100.46504974365234,
      "learning_rate": 0.0001512123354854955,
      "loss": 1.1309,
      "step": 500
    },
    {
      "epoch": 0.06573724782680007,
      "grad_norm": 87.32220458984375,
      "learning_rate": 0.0001510311243799295,
      "loss": 1.0387,
      "step": 501
    },
    {
      "epoch": 0.06586845989831064,
      "grad_norm": 85.12271881103516,
      "learning_rate": 0.00015084968641213958,
      "loss": 1.2373,
      "step": 502
    },
    {
      "epoch": 0.06599967196982122,
      "grad_norm": 102.62129211425781,
      "learning_rate": 0.00015066802238872023,
      "loss": 1.3695,
      "step": 503
    },
    {
      "epoch": 0.0661308840413318,
      "grad_norm": 103.07049560546875,
      "learning_rate": 0.0001504861331172709,
      "loss": 1.4536,
      "step": 504
    },
    {
      "epoch": 0.06626209611284238,
      "grad_norm": 107.23957824707031,
      "learning_rate": 0.0001503040194063922,
      "loss": 1.3144,
      "step": 505
    },
    {
      "epoch": 0.06639330818435296,
      "grad_norm": 103.92955017089844,
      "learning_rate": 0.00015012168206568268,
      "loss": 1.3752,
      "step": 506
    },
    {
      "epoch": 0.06652452025586354,
      "grad_norm": 99.0862808227539,
      "learning_rate": 0.00014993912190573505,
      "loss": 1.6014,
      "step": 507
    },
    {
      "epoch": 0.06665573232737412,
      "grad_norm": 107.68649291992188,
      "learning_rate": 0.00014975633973813242,
      "loss": 1.3814,
      "step": 508
    },
    {
      "epoch": 0.0667869443988847,
      "grad_norm": 89.5562973022461,
      "learning_rate": 0.00014957333637544503,
      "loss": 1.3382,
      "step": 509
    },
    {
      "epoch": 0.06691815647039527,
      "grad_norm": 101.75519561767578,
      "learning_rate": 0.00014939011263122634,
      "loss": 1.4943,
      "step": 510
    },
    {
      "epoch": 0.06704936854190585,
      "grad_norm": 89.83662414550781,
      "learning_rate": 0.0001492066693200096,
      "loss": 1.5218,
      "step": 511
    },
    {
      "epoch": 0.06718058061341643,
      "grad_norm": 96.33296966552734,
      "learning_rate": 0.00014902300725730413,
      "loss": 1.3269,
      "step": 512
    },
    {
      "epoch": 0.06731179268492701,
      "grad_norm": 83.73062896728516,
      "learning_rate": 0.00014883912725959167,
      "loss": 1.3324,
      "step": 513
    },
    {
      "epoch": 0.0674430047564376,
      "grad_norm": 95.9453125,
      "learning_rate": 0.00014865503014432292,
      "loss": 1.3229,
      "step": 514
    },
    {
      "epoch": 0.06757421682794817,
      "grad_norm": 91.52754974365234,
      "learning_rate": 0.00014847071672991367,
      "loss": 1.4314,
      "step": 515
    },
    {
      "epoch": 0.06770542889945876,
      "grad_norm": 88.39478302001953,
      "learning_rate": 0.0001482861878357414,
      "loss": 1.2086,
      "step": 516
    },
    {
      "epoch": 0.06783664097096932,
      "grad_norm": 98.7316665649414,
      "learning_rate": 0.00014810144428214144,
      "loss": 1.1989,
      "step": 517
    },
    {
      "epoch": 0.0679678530424799,
      "grad_norm": 97.03109741210938,
      "learning_rate": 0.0001479164868904034,
      "loss": 1.5284,
      "step": 518
    },
    {
      "epoch": 0.06809906511399048,
      "grad_norm": 79.53731536865234,
      "learning_rate": 0.00014773131648276758,
      "loss": 0.8382,
      "step": 519
    },
    {
      "epoch": 0.06823027718550106,
      "grad_norm": 102.88385009765625,
      "learning_rate": 0.00014754593388242117,
      "loss": 1.371,
      "step": 520
    },
    {
      "epoch": 0.06836148925701165,
      "grad_norm": 90.7183837890625,
      "learning_rate": 0.0001473603399134948,
      "loss": 1.243,
      "step": 521
    },
    {
      "epoch": 0.06849270132852223,
      "grad_norm": 95.85237121582031,
      "learning_rate": 0.0001471745354010586,
      "loss": 1.4161,
      "step": 522
    },
    {
      "epoch": 0.06862391340003281,
      "grad_norm": 111.7193374633789,
      "learning_rate": 0.00014698852117111884,
      "loss": 1.3228,
      "step": 523
    },
    {
      "epoch": 0.06875512547154339,
      "grad_norm": 103.87052154541016,
      "learning_rate": 0.000146802298050614,
      "loss": 1.4702,
      "step": 524
    },
    {
      "epoch": 0.06888633754305395,
      "grad_norm": 124.20005798339844,
      "learning_rate": 0.0001466158668674112,
      "loss": 1.4673,
      "step": 525
    },
    {
      "epoch": 0.06901754961456454,
      "grad_norm": 100.72663879394531,
      "learning_rate": 0.00014642922845030257,
      "loss": 1.3444,
      "step": 526
    },
    {
      "epoch": 0.06914876168607512,
      "grad_norm": 94.54740142822266,
      "learning_rate": 0.0001462423836290015,
      "loss": 1.4809,
      "step": 527
    },
    {
      "epoch": 0.0692799737575857,
      "grad_norm": 123.48004150390625,
      "learning_rate": 0.00014605533323413887,
      "loss": 1.5193,
      "step": 528
    },
    {
      "epoch": 0.06941118582909628,
      "grad_norm": 82.52967071533203,
      "learning_rate": 0.00014586807809725962,
      "loss": 1.0382,
      "step": 529
    },
    {
      "epoch": 0.06954239790060686,
      "grad_norm": 110.57383728027344,
      "learning_rate": 0.00014568061905081875,
      "loss": 1.5592,
      "step": 530
    },
    {
      "epoch": 0.06967360997211744,
      "grad_norm": 83.3657455444336,
      "learning_rate": 0.00014549295692817778,
      "loss": 1.1996,
      "step": 531
    },
    {
      "epoch": 0.06980482204362802,
      "grad_norm": 113.93882751464844,
      "learning_rate": 0.00014530509256360102,
      "loss": 1.459,
      "step": 532
    },
    {
      "epoch": 0.06993603411513859,
      "grad_norm": 104.60193634033203,
      "learning_rate": 0.00014511702679225193,
      "loss": 1.1598,
      "step": 533
    },
    {
      "epoch": 0.07006724618664917,
      "grad_norm": 97.72349548339844,
      "learning_rate": 0.0001449287604501893,
      "loss": 1.379,
      "step": 534
    },
    {
      "epoch": 0.07019845825815975,
      "grad_norm": 87.33219146728516,
      "learning_rate": 0.00014474029437436348,
      "loss": 1.0107,
      "step": 535
    },
    {
      "epoch": 0.07032967032967033,
      "grad_norm": 89.23751068115234,
      "learning_rate": 0.00014455162940261285,
      "loss": 1.2164,
      "step": 536
    },
    {
      "epoch": 0.07046088240118091,
      "grad_norm": 123.31934356689453,
      "learning_rate": 0.0001443627663736599,
      "loss": 1.5902,
      "step": 537
    },
    {
      "epoch": 0.07059209447269149,
      "grad_norm": 84.15154266357422,
      "learning_rate": 0.00014417370612710778,
      "loss": 1.1374,
      "step": 538
    },
    {
      "epoch": 0.07072330654420207,
      "grad_norm": 96.65689849853516,
      "learning_rate": 0.00014398444950343623,
      "loss": 1.2253,
      "step": 539
    },
    {
      "epoch": 0.07085451861571264,
      "grad_norm": 92.46644592285156,
      "learning_rate": 0.00014379499734399798,
      "loss": 1.4614,
      "step": 540
    },
    {
      "epoch": 0.07098573068722322,
      "grad_norm": 98.3292465209961,
      "learning_rate": 0.0001436053504910151,
      "loss": 1.4079,
      "step": 541
    },
    {
      "epoch": 0.0711169427587338,
      "grad_norm": 95.9684829711914,
      "learning_rate": 0.0001434155097875752,
      "loss": 1.5408,
      "step": 542
    },
    {
      "epoch": 0.07124815483024438,
      "grad_norm": 99.97079467773438,
      "learning_rate": 0.00014322547607762762,
      "loss": 1.2549,
      "step": 543
    },
    {
      "epoch": 0.07137936690175496,
      "grad_norm": 85.59657287597656,
      "learning_rate": 0.0001430352502059797,
      "loss": 1.1647,
      "step": 544
    },
    {
      "epoch": 0.07151057897326554,
      "grad_norm": 99.05835723876953,
      "learning_rate": 0.0001428448330182931,
      "loss": 1.5245,
      "step": 545
    },
    {
      "epoch": 0.07164179104477612,
      "grad_norm": 100.6355972290039,
      "learning_rate": 0.00014265422536107993,
      "loss": 1.4144,
      "step": 546
    },
    {
      "epoch": 0.0717730031162867,
      "grad_norm": 87.58867645263672,
      "learning_rate": 0.00014246342808169914,
      "loss": 1.2356,
      "step": 547
    },
    {
      "epoch": 0.07190421518779727,
      "grad_norm": 77.95830535888672,
      "learning_rate": 0.00014227244202835257,
      "loss": 1.0183,
      "step": 548
    },
    {
      "epoch": 0.07203542725930785,
      "grad_norm": 117.39501953125,
      "learning_rate": 0.0001420812680500813,
      "loss": 1.5723,
      "step": 549
    },
    {
      "epoch": 0.07216663933081843,
      "grad_norm": 85.6815414428711,
      "learning_rate": 0.00014188990699676184,
      "loss": 1.2084,
      "step": 550
    },
    {
      "epoch": 0.07229785140232901,
      "grad_norm": 93.67658233642578,
      "learning_rate": 0.00014169835971910238,
      "loss": 1.3034,
      "step": 551
    },
    {
      "epoch": 0.0724290634738396,
      "grad_norm": 97.38410949707031,
      "learning_rate": 0.0001415066270686389,
      "loss": 1.4992,
      "step": 552
    },
    {
      "epoch": 0.07256027554535018,
      "grad_norm": 78.78777313232422,
      "learning_rate": 0.00014131470989773158,
      "loss": 1.0619,
      "step": 553
    },
    {
      "epoch": 0.07269148761686076,
      "grad_norm": 103.23997497558594,
      "learning_rate": 0.0001411226090595608,
      "loss": 1.3115,
      "step": 554
    },
    {
      "epoch": 0.07282269968837134,
      "grad_norm": 95.2453842163086,
      "learning_rate": 0.00014093032540812348,
      "loss": 1.3362,
      "step": 555
    },
    {
      "epoch": 0.0729539117598819,
      "grad_norm": 119.82777404785156,
      "learning_rate": 0.0001407378597982293,
      "loss": 1.4418,
      "step": 556
    },
    {
      "epoch": 0.07308512383139248,
      "grad_norm": 100.15574645996094,
      "learning_rate": 0.00014054521308549673,
      "loss": 1.3295,
      "step": 557
    },
    {
      "epoch": 0.07321633590290307,
      "grad_norm": 90.64246368408203,
      "learning_rate": 0.0001403523861263495,
      "loss": 1.1399,
      "step": 558
    },
    {
      "epoch": 0.07334754797441365,
      "grad_norm": 109.21315002441406,
      "learning_rate": 0.00014015937977801256,
      "loss": 1.3947,
      "step": 559
    },
    {
      "epoch": 0.07347876004592423,
      "grad_norm": 84.3509521484375,
      "learning_rate": 0.00013996619489850822,
      "loss": 1.268,
      "step": 560
    },
    {
      "epoch": 0.07360997211743481,
      "grad_norm": 103.73743438720703,
      "learning_rate": 0.00013977283234665273,
      "loss": 1.3775,
      "step": 561
    },
    {
      "epoch": 0.07374118418894539,
      "grad_norm": 85.89759063720703,
      "learning_rate": 0.00013957929298205195,
      "loss": 1.3283,
      "step": 562
    },
    {
      "epoch": 0.07387239626045596,
      "grad_norm": 94.20722961425781,
      "learning_rate": 0.00013938557766509792,
      "loss": 1.3219,
      "step": 563
    },
    {
      "epoch": 0.07400360833196654,
      "grad_norm": 80.8791732788086,
      "learning_rate": 0.0001391916872569648,
      "loss": 1.1917,
      "step": 564
    },
    {
      "epoch": 0.07413482040347712,
      "grad_norm": 89.1103744506836,
      "learning_rate": 0.00013899762261960518,
      "loss": 1.5825,
      "step": 565
    },
    {
      "epoch": 0.0742660324749877,
      "grad_norm": 85.13253021240234,
      "learning_rate": 0.0001388033846157462,
      "loss": 1.2244,
      "step": 566
    },
    {
      "epoch": 0.07439724454649828,
      "grad_norm": 79.84519958496094,
      "learning_rate": 0.0001386089741088857,
      "loss": 1.2388,
      "step": 567
    },
    {
      "epoch": 0.07452845661800886,
      "grad_norm": 94.13172149658203,
      "learning_rate": 0.00013841439196328836,
      "loss": 1.423,
      "step": 568
    },
    {
      "epoch": 0.07465966868951944,
      "grad_norm": 82.32963562011719,
      "learning_rate": 0.00013821963904398193,
      "loss": 1.1689,
      "step": 569
    },
    {
      "epoch": 0.07479088076103002,
      "grad_norm": 82.66189575195312,
      "learning_rate": 0.00013802471621675338,
      "loss": 1.2138,
      "step": 570
    },
    {
      "epoch": 0.07492209283254059,
      "grad_norm": 94.656494140625,
      "learning_rate": 0.00013782962434814492,
      "loss": 1.5074,
      "step": 571
    },
    {
      "epoch": 0.07505330490405117,
      "grad_norm": 98.77300262451172,
      "learning_rate": 0.00013763436430545034,
      "loss": 1.485,
      "step": 572
    },
    {
      "epoch": 0.07518451697556175,
      "grad_norm": 89.64130401611328,
      "learning_rate": 0.00013743893695671096,
      "loss": 1.3385,
      "step": 573
    },
    {
      "epoch": 0.07531572904707233,
      "grad_norm": 100.80279541015625,
      "learning_rate": 0.00013724334317071198,
      "loss": 1.4083,
      "step": 574
    },
    {
      "epoch": 0.07544694111858291,
      "grad_norm": 91.70159912109375,
      "learning_rate": 0.00013704758381697844,
      "loss": 1.4267,
      "step": 575
    },
    {
      "epoch": 0.07557815319009349,
      "grad_norm": 88.65699768066406,
      "learning_rate": 0.00013685165976577146,
      "loss": 1.3017,
      "step": 576
    },
    {
      "epoch": 0.07570936526160407,
      "grad_norm": 80.58065795898438,
      "learning_rate": 0.0001366555718880843,
      "loss": 1.1337,
      "step": 577
    },
    {
      "epoch": 0.07584057733311465,
      "grad_norm": 101.66384887695312,
      "learning_rate": 0.00013645932105563844,
      "loss": 1.3666,
      "step": 578
    },
    {
      "epoch": 0.07597178940462522,
      "grad_norm": 89.27428436279297,
      "learning_rate": 0.00013626290814088005,
      "loss": 1.3328,
      "step": 579
    },
    {
      "epoch": 0.0761030014761358,
      "grad_norm": 79.54308319091797,
      "learning_rate": 0.00013606633401697557,
      "loss": 1.0088,
      "step": 580
    },
    {
      "epoch": 0.07623421354764638,
      "grad_norm": 92.69046783447266,
      "learning_rate": 0.00013586959955780824,
      "loss": 1.2759,
      "step": 581
    },
    {
      "epoch": 0.07636542561915696,
      "grad_norm": 98.94734954833984,
      "learning_rate": 0.00013567270563797398,
      "loss": 1.2657,
      "step": 582
    },
    {
      "epoch": 0.07649663769066754,
      "grad_norm": 95.38640594482422,
      "learning_rate": 0.00013547565313277776,
      "loss": 1.4724,
      "step": 583
    },
    {
      "epoch": 0.07662784976217812,
      "grad_norm": 101.580078125,
      "learning_rate": 0.00013527844291822948,
      "loss": 1.4412,
      "step": 584
    },
    {
      "epoch": 0.0767590618336887,
      "grad_norm": 103.85546875,
      "learning_rate": 0.0001350810758710401,
      "loss": 1.4579,
      "step": 585
    },
    {
      "epoch": 0.07689027390519927,
      "grad_norm": 85.1213150024414,
      "learning_rate": 0.00013488355286861783,
      "loss": 1.0065,
      "step": 586
    },
    {
      "epoch": 0.07702148597670985,
      "grad_norm": 97.4102554321289,
      "learning_rate": 0.0001346858747890642,
      "loss": 1.3881,
      "step": 587
    },
    {
      "epoch": 0.07715269804822043,
      "grad_norm": 95.22293853759766,
      "learning_rate": 0.00013448804251117003,
      "loss": 1.1938,
      "step": 588
    },
    {
      "epoch": 0.07728391011973101,
      "grad_norm": 110.470703125,
      "learning_rate": 0.0001342900569144119,
      "loss": 1.2816,
      "step": 589
    },
    {
      "epoch": 0.0774151221912416,
      "grad_norm": 97.50569152832031,
      "learning_rate": 0.0001340919188789477,
      "loss": 1.3063,
      "step": 590
    },
    {
      "epoch": 0.07754633426275218,
      "grad_norm": 102.9594955444336,
      "learning_rate": 0.00013389362928561317,
      "loss": 1.2478,
      "step": 591
    },
    {
      "epoch": 0.07767754633426276,
      "grad_norm": 91.47532653808594,
      "learning_rate": 0.00013369518901591772,
      "loss": 1.3588,
      "step": 592
    },
    {
      "epoch": 0.07780875840577334,
      "grad_norm": 109.54425048828125,
      "learning_rate": 0.00013349659895204067,
      "loss": 1.4631,
      "step": 593
    },
    {
      "epoch": 0.0779399704772839,
      "grad_norm": 92.49427795410156,
      "learning_rate": 0.0001332978599768272,
      "loss": 1.368,
      "step": 594
    },
    {
      "epoch": 0.07807118254879448,
      "grad_norm": 101.76089477539062,
      "learning_rate": 0.00013309897297378455,
      "loss": 1.3572,
      "step": 595
    },
    {
      "epoch": 0.07820239462030507,
      "grad_norm": 100.13697052001953,
      "learning_rate": 0.00013289993882707797,
      "loss": 1.4331,
      "step": 596
    },
    {
      "epoch": 0.07833360669181565,
      "grad_norm": 94.05873107910156,
      "learning_rate": 0.00013270075842152678,
      "loss": 1.2936,
      "step": 597
    },
    {
      "epoch": 0.07846481876332623,
      "grad_norm": 92.8931655883789,
      "learning_rate": 0.00013250143264260074,
      "loss": 1.1791,
      "step": 598
    },
    {
      "epoch": 0.07859603083483681,
      "grad_norm": 85.41705322265625,
      "learning_rate": 0.0001323019623764156,
      "loss": 1.3681,
      "step": 599
    },
    {
      "epoch": 0.07872724290634739,
      "grad_norm": 91.26629638671875,
      "learning_rate": 0.00013210234850972964,
      "loss": 1.0341,
      "step": 600
    },
    {
      "epoch": 0.07885845497785797,
      "grad_norm": 98.13883972167969,
      "learning_rate": 0.0001319025919299394,
      "loss": 0.9386,
      "step": 601
    },
    {
      "epoch": 0.07898966704936854,
      "grad_norm": 113.55065155029297,
      "learning_rate": 0.00013170269352507597,
      "loss": 1.5092,
      "step": 602
    },
    {
      "epoch": 0.07912087912087912,
      "grad_norm": 90.46080017089844,
      "learning_rate": 0.0001315026541838008,
      "loss": 1.1674,
      "step": 603
    },
    {
      "epoch": 0.0792520911923897,
      "grad_norm": 90.46232604980469,
      "learning_rate": 0.00013130247479540202,
      "loss": 1.3674,
      "step": 604
    },
    {
      "epoch": 0.07938330326390028,
      "grad_norm": 98.26184844970703,
      "learning_rate": 0.00013110215624979025,
      "loss": 1.4128,
      "step": 605
    },
    {
      "epoch": 0.07951451533541086,
      "grad_norm": 86.61162567138672,
      "learning_rate": 0.00013090169943749476,
      "loss": 1.346,
      "step": 606
    },
    {
      "epoch": 0.07964572740692144,
      "grad_norm": 84.7541732788086,
      "learning_rate": 0.00013070110524965954,
      "loss": 1.3569,
      "step": 607
    },
    {
      "epoch": 0.07977693947843202,
      "grad_norm": 90.7254867553711,
      "learning_rate": 0.00013050037457803924,
      "loss": 1.4229,
      "step": 608
    },
    {
      "epoch": 0.07990815154994259,
      "grad_norm": 80.16986846923828,
      "learning_rate": 0.0001302995083149953,
      "loss": 1.1681,
      "step": 609
    },
    {
      "epoch": 0.08003936362145317,
      "grad_norm": 89.36129760742188,
      "learning_rate": 0.0001300985073534919,
      "loss": 1.4023,
      "step": 610
    },
    {
      "epoch": 0.08017057569296375,
      "grad_norm": 106.40861511230469,
      "learning_rate": 0.00012989737258709203,
      "loss": 1.5128,
      "step": 611
    },
    {
      "epoch": 0.08030178776447433,
      "grad_norm": 98.45272064208984,
      "learning_rate": 0.00012969610490995358,
      "loss": 1.3327,
      "step": 612
    },
    {
      "epoch": 0.08043299983598491,
      "grad_norm": 98.46907806396484,
      "learning_rate": 0.00012949470521682528,
      "loss": 1.1781,
      "step": 613
    },
    {
      "epoch": 0.08056421190749549,
      "grad_norm": 97.92434692382812,
      "learning_rate": 0.0001292931744030427,
      "loss": 1.3189,
      "step": 614
    },
    {
      "epoch": 0.08069542397900607,
      "grad_norm": 87.9881820678711,
      "learning_rate": 0.0001290915133645243,
      "loss": 1.2561,
      "step": 615
    },
    {
      "epoch": 0.08082663605051665,
      "grad_norm": 79.83320617675781,
      "learning_rate": 0.00012888972299776754,
      "loss": 1.1237,
      "step": 616
    },
    {
      "epoch": 0.08095784812202722,
      "grad_norm": 99.0945053100586,
      "learning_rate": 0.00012868780419984482,
      "loss": 1.2593,
      "step": 617
    },
    {
      "epoch": 0.0810890601935378,
      "grad_norm": 96.94082641601562,
      "learning_rate": 0.00012848575786839943,
      "loss": 1.4874,
      "step": 618
    },
    {
      "epoch": 0.08122027226504838,
      "grad_norm": 84.22801208496094,
      "learning_rate": 0.0001282835849016416,
      "loss": 1.0247,
      "step": 619
    },
    {
      "epoch": 0.08135148433655896,
      "grad_norm": 92.93048095703125,
      "learning_rate": 0.00012808128619834461,
      "loss": 1.5083,
      "step": 620
    },
    {
      "epoch": 0.08148269640806954,
      "grad_norm": 89.20956420898438,
      "learning_rate": 0.0001278788626578407,
      "loss": 1.2741,
      "step": 621
    },
    {
      "epoch": 0.08161390847958012,
      "grad_norm": 85.68168640136719,
      "learning_rate": 0.00012767631518001698,
      "loss": 1.3059,
      "step": 622
    },
    {
      "epoch": 0.0817451205510907,
      "grad_norm": 92.48267364501953,
      "learning_rate": 0.00012747364466531163,
      "loss": 1.3309,
      "step": 623
    },
    {
      "epoch": 0.08187633262260129,
      "grad_norm": 86.2343978881836,
      "learning_rate": 0.00012727085201470973,
      "loss": 1.3335,
      "step": 624
    },
    {
      "epoch": 0.08200754469411185,
      "grad_norm": 80.11412048339844,
      "learning_rate": 0.00012706793812973941,
      "loss": 1.2295,
      "step": 625
    },
    {
      "epoch": 0.08213875676562243,
      "grad_norm": 91.04052734375,
      "learning_rate": 0.0001268649039124677,
      "loss": 1.2667,
      "step": 626
    },
    {
      "epoch": 0.08226996883713301,
      "grad_norm": 88.94692993164062,
      "learning_rate": 0.00012666175026549662,
      "loss": 1.2893,
      "step": 627
    },
    {
      "epoch": 0.0824011809086436,
      "grad_norm": 97.49986267089844,
      "learning_rate": 0.000126458478091959,
      "loss": 1.476,
      "step": 628
    },
    {
      "epoch": 0.08253239298015418,
      "grad_norm": 80.16893005371094,
      "learning_rate": 0.00012625508829551473,
      "loss": 1.0821,
      "step": 629
    },
    {
      "epoch": 0.08266360505166476,
      "grad_norm": 94.23833465576172,
      "learning_rate": 0.00012605158178034654,
      "loss": 1.4205,
      "step": 630
    },
    {
      "epoch": 0.08279481712317534,
      "grad_norm": 100.96776580810547,
      "learning_rate": 0.00012584795945115603,
      "loss": 1.4535,
      "step": 631
    },
    {
      "epoch": 0.0829260291946859,
      "grad_norm": 90.95092010498047,
      "learning_rate": 0.0001256442222131597,
      "loss": 1.3493,
      "step": 632
    },
    {
      "epoch": 0.08305724126619649,
      "grad_norm": 111.87983703613281,
      "learning_rate": 0.0001254403709720848,
      "loss": 1.4152,
      "step": 633
    },
    {
      "epoch": 0.08318845333770707,
      "grad_norm": 103.62471008300781,
      "learning_rate": 0.0001252364066341655,
      "loss": 1.2251,
      "step": 634
    },
    {
      "epoch": 0.08331966540921765,
      "grad_norm": 94.95115661621094,
      "learning_rate": 0.00012503233010613865,
      "loss": 1.2112,
      "step": 635
    },
    {
      "epoch": 0.08345087748072823,
      "grad_norm": 88.23070526123047,
      "learning_rate": 0.00012482814229523997,
      "loss": 1.2592,
      "step": 636
    },
    {
      "epoch": 0.08358208955223881,
      "grad_norm": 89.11841583251953,
      "learning_rate": 0.00012462384410919975,
      "loss": 1.3439,
      "step": 637
    },
    {
      "epoch": 0.08371330162374939,
      "grad_norm": 92.19784545898438,
      "learning_rate": 0.00012441943645623903,
      "loss": 1.4071,
      "step": 638
    },
    {
      "epoch": 0.08384451369525997,
      "grad_norm": 88.66581726074219,
      "learning_rate": 0.00012421492024506555,
      "loss": 1.2349,
      "step": 639
    },
    {
      "epoch": 0.08397572576677054,
      "grad_norm": 137.8487091064453,
      "learning_rate": 0.00012401029638486953,
      "loss": 1.3711,
      "step": 640
    },
    {
      "epoch": 0.08410693783828112,
      "grad_norm": 99.01351165771484,
      "learning_rate": 0.0001238055657853198,
      "loss": 1.348,
      "step": 641
    },
    {
      "epoch": 0.0842381499097917,
      "grad_norm": 89.60566711425781,
      "learning_rate": 0.00012360072935655982,
      "loss": 1.3787,
      "step": 642
    },
    {
      "epoch": 0.08436936198130228,
      "grad_norm": 95.86077117919922,
      "learning_rate": 0.00012339578800920332,
      "loss": 1.4273,
      "step": 643
    },
    {
      "epoch": 0.08450057405281286,
      "grad_norm": 94.48822021484375,
      "learning_rate": 0.00012319074265433063,
      "loss": 1.418,
      "step": 644
    },
    {
      "epoch": 0.08463178612432344,
      "grad_norm": 87.32054138183594,
      "learning_rate": 0.00012298559420348437,
      "loss": 1.2337,
      "step": 645
    },
    {
      "epoch": 0.08476299819583402,
      "grad_norm": 86.92813110351562,
      "learning_rate": 0.00012278034356866545,
      "loss": 1.3581,
      "step": 646
    },
    {
      "epoch": 0.0848942102673446,
      "grad_norm": 94.89491271972656,
      "learning_rate": 0.00012257499166232907,
      "loss": 1.3674,
      "step": 647
    },
    {
      "epoch": 0.08502542233885517,
      "grad_norm": 103.08065795898438,
      "learning_rate": 0.0001223695393973807,
      "loss": 1.0211,
      "step": 648
    },
    {
      "epoch": 0.08515663441036575,
      "grad_norm": 94.00289154052734,
      "learning_rate": 0.0001221639876871719,
      "loss": 1.3441,
      "step": 649
    },
    {
      "epoch": 0.08528784648187633,
      "grad_norm": 85.98426055908203,
      "learning_rate": 0.0001219583374454963,
      "loss": 0.9399,
      "step": 650
    },
    {
      "epoch": 0.08541905855338691,
      "grad_norm": 81.55607604980469,
      "learning_rate": 0.00012175258958658564,
      "loss": 1.0386,
      "step": 651
    },
    {
      "epoch": 0.08555027062489749,
      "grad_norm": 97.75609588623047,
      "learning_rate": 0.00012154674502510555,
      "loss": 1.4796,
      "step": 652
    },
    {
      "epoch": 0.08568148269640807,
      "grad_norm": 74.53363800048828,
      "learning_rate": 0.00012134080467615159,
      "loss": 1.1285,
      "step": 653
    },
    {
      "epoch": 0.08581269476791865,
      "grad_norm": 94.61219787597656,
      "learning_rate": 0.00012113476945524513,
      "loss": 1.2521,
      "step": 654
    },
    {
      "epoch": 0.08594390683942922,
      "grad_norm": 99.3115005493164,
      "learning_rate": 0.00012092864027832933,
      "loss": 1.2409,
      "step": 655
    },
    {
      "epoch": 0.0860751189109398,
      "grad_norm": 106.48043060302734,
      "learning_rate": 0.000120722418061765,
      "loss": 1.3513,
      "step": 656
    },
    {
      "epoch": 0.08620633098245038,
      "grad_norm": 90.65664672851562,
      "learning_rate": 0.0001205161037223266,
      "loss": 1.3055,
      "step": 657
    },
    {
      "epoch": 0.08633754305396096,
      "grad_norm": 80.25823211669922,
      "learning_rate": 0.00012030969817719808,
      "loss": 1.3302,
      "step": 658
    },
    {
      "epoch": 0.08646875512547154,
      "grad_norm": 93.70552825927734,
      "learning_rate": 0.00012010320234396894,
      "loss": 1.3059,
      "step": 659
    },
    {
      "epoch": 0.08659996719698212,
      "grad_norm": 95.39839172363281,
      "learning_rate": 0.00011989661714062999,
      "loss": 1.1336,
      "step": 660
    },
    {
      "epoch": 0.0867311792684927,
      "grad_norm": 98.91097259521484,
      "learning_rate": 0.0001196899434855693,
      "loss": 1.3398,
      "step": 661
    },
    {
      "epoch": 0.08686239134000329,
      "grad_norm": 125.88059997558594,
      "learning_rate": 0.00011948318229756827,
      "loss": 1.3293,
      "step": 662
    },
    {
      "epoch": 0.08699360341151385,
      "grad_norm": 101.42369079589844,
      "learning_rate": 0.00011927633449579735,
      "loss": 1.4461,
      "step": 663
    },
    {
      "epoch": 0.08712481548302443,
      "grad_norm": 84.62518310546875,
      "learning_rate": 0.0001190694009998121,
      "loss": 1.1745,
      "step": 664
    },
    {
      "epoch": 0.08725602755453502,
      "grad_norm": 91.96395874023438,
      "learning_rate": 0.00011886238272954897,
      "loss": 1.3255,
      "step": 665
    },
    {
      "epoch": 0.0873872396260456,
      "grad_norm": 86.48896026611328,
      "learning_rate": 0.00011865528060532127,
      "loss": 1.2024,
      "step": 666
    },
    {
      "epoch": 0.08751845169755618,
      "grad_norm": 84.64398956298828,
      "learning_rate": 0.0001184480955478152,
      "loss": 1.2483,
      "step": 667
    },
    {
      "epoch": 0.08764966376906676,
      "grad_norm": 90.9584732055664,
      "learning_rate": 0.00011824082847808558,
      "loss": 1.4197,
      "step": 668
    },
    {
      "epoch": 0.08778087584057734,
      "grad_norm": 80.34623718261719,
      "learning_rate": 0.00011803348031755179,
      "loss": 1.2402,
      "step": 669
    },
    {
      "epoch": 0.08791208791208792,
      "grad_norm": 79.48882293701172,
      "learning_rate": 0.0001178260519879937,
      "loss": 1.0723,
      "step": 670
    },
    {
      "epoch": 0.08804329998359849,
      "grad_norm": 79.95413970947266,
      "learning_rate": 0.00011761854441154767,
      "loss": 1.1993,
      "step": 671
    },
    {
      "epoch": 0.08817451205510907,
      "grad_norm": 98.07939910888672,
      "learning_rate": 0.00011741095851070228,
      "loss": 1.2364,
      "step": 672
    },
    {
      "epoch": 0.08830572412661965,
      "grad_norm": 94.5353012084961,
      "learning_rate": 0.00011720329520829429,
      "loss": 1.3124,
      "step": 673
    },
    {
      "epoch": 0.08843693619813023,
      "grad_norm": 84.25875854492188,
      "learning_rate": 0.0001169955554275046,
      "loss": 1.2399,
      "step": 674
    },
    {
      "epoch": 0.08856814826964081,
      "grad_norm": 101.6513671875,
      "learning_rate": 0.0001167877400918541,
      "loss": 1.4408,
      "step": 675
    },
    {
      "epoch": 0.08869936034115139,
      "grad_norm": 89.99846649169922,
      "learning_rate": 0.00011657985012519952,
      "loss": 1.1339,
      "step": 676
    },
    {
      "epoch": 0.08883057241266197,
      "grad_norm": 96.81571960449219,
      "learning_rate": 0.00011637188645172944,
      "loss": 1.3194,
      "step": 677
    },
    {
      "epoch": 0.08896178448417254,
      "grad_norm": 82.89884185791016,
      "learning_rate": 0.00011616384999596006,
      "loss": 1.3209,
      "step": 678
    },
    {
      "epoch": 0.08909299655568312,
      "grad_norm": 77.22850799560547,
      "learning_rate": 0.00011595574168273111,
      "loss": 1.2296,
      "step": 679
    },
    {
      "epoch": 0.0892242086271937,
      "grad_norm": 103.56705474853516,
      "learning_rate": 0.0001157475624372018,
      "loss": 1.1064,
      "step": 680
    },
    {
      "epoch": 0.08935542069870428,
      "grad_norm": 113.94505310058594,
      "learning_rate": 0.0001155393131848467,
      "loss": 1.3729,
      "step": 681
    },
    {
      "epoch": 0.08948663277021486,
      "grad_norm": 98.26383209228516,
      "learning_rate": 0.00011533099485145155,
      "loss": 1.4269,
      "step": 682
    },
    {
      "epoch": 0.08961784484172544,
      "grad_norm": 94.24766540527344,
      "learning_rate": 0.00011512260836310924,
      "loss": 1.313,
      "step": 683
    },
    {
      "epoch": 0.08974905691323602,
      "grad_norm": 94.0205078125,
      "learning_rate": 0.00011491415464621562,
      "loss": 1.4233,
      "step": 684
    },
    {
      "epoch": 0.0898802689847466,
      "grad_norm": 86.20130920410156,
      "learning_rate": 0.00011470563462746541,
      "loss": 1.2434,
      "step": 685
    },
    {
      "epoch": 0.09001148105625717,
      "grad_norm": 84.50789642333984,
      "learning_rate": 0.00011449704923384812,
      "loss": 1.1149,
      "step": 686
    },
    {
      "epoch": 0.09014269312776775,
      "grad_norm": 89.99745178222656,
      "learning_rate": 0.00011428839939264382,
      "loss": 1.2888,
      "step": 687
    },
    {
      "epoch": 0.09027390519927833,
      "grad_norm": 93.26073455810547,
      "learning_rate": 0.0001140796860314191,
      "loss": 1.2935,
      "step": 688
    },
    {
      "epoch": 0.09040511727078891,
      "grad_norm": 91.77643585205078,
      "learning_rate": 0.00011387091007802297,
      "loss": 1.1229,
      "step": 689
    },
    {
      "epoch": 0.09053632934229949,
      "grad_norm": 96.67001342773438,
      "learning_rate": 0.0001136620724605827,
      "loss": 1.2077,
      "step": 690
    },
    {
      "epoch": 0.09066754141381007,
      "grad_norm": 122.97071075439453,
      "learning_rate": 0.00011345317410749964,
      "loss": 1.559,
      "step": 691
    },
    {
      "epoch": 0.09079875348532065,
      "grad_norm": 92.56646728515625,
      "learning_rate": 0.00011324421594744516,
      "loss": 1.4267,
      "step": 692
    },
    {
      "epoch": 0.09092996555683124,
      "grad_norm": 97.94356536865234,
      "learning_rate": 0.00011303519890935656,
      "loss": 1.3027,
      "step": 693
    },
    {
      "epoch": 0.0910611776283418,
      "grad_norm": 82.18528747558594,
      "learning_rate": 0.00011282612392243286,
      "loss": 1.1712,
      "step": 694
    },
    {
      "epoch": 0.09119238969985238,
      "grad_norm": 92.41801452636719,
      "learning_rate": 0.00011261699191613066,
      "loss": 1.3709,
      "step": 695
    },
    {
      "epoch": 0.09132360177136296,
      "grad_norm": 122.3470687866211,
      "learning_rate": 0.00011240780382016005,
      "loss": 1.7139,
      "step": 696
    },
    {
      "epoch": 0.09145481384287354,
      "grad_norm": 89.3165054321289,
      "learning_rate": 0.00011219856056448051,
      "loss": 1.1107,
      "step": 697
    },
    {
      "epoch": 0.09158602591438413,
      "grad_norm": 69.40094757080078,
      "learning_rate": 0.00011198926307929664,
      "loss": 0.7894,
      "step": 698
    },
    {
      "epoch": 0.0917172379858947,
      "grad_norm": 133.05140686035156,
      "learning_rate": 0.00011177991229505431,
      "loss": 1.5211,
      "step": 699
    },
    {
      "epoch": 0.09184845005740529,
      "grad_norm": 91.17678833007812,
      "learning_rate": 0.00011157050914243614,
      "loss": 1.0069,
      "step": 700
    },
    {
      "epoch": 0.09197966212891585,
      "grad_norm": 83.76356506347656,
      "learning_rate": 0.00011136105455235766,
      "loss": 1.1431,
      "step": 701
    },
    {
      "epoch": 0.09211087420042643,
      "grad_norm": 103.66838836669922,
      "learning_rate": 0.00011115154945596305,
      "loss": 1.5398,
      "step": 702
    },
    {
      "epoch": 0.09224208627193702,
      "grad_norm": 79.48207092285156,
      "learning_rate": 0.00011094199478462095,
      "loss": 1.0131,
      "step": 703
    },
    {
      "epoch": 0.0923732983434476,
      "grad_norm": 96.70210266113281,
      "learning_rate": 0.00011073239146992054,
      "loss": 1.6009,
      "step": 704
    },
    {
      "epoch": 0.09250451041495818,
      "grad_norm": 95.59807586669922,
      "learning_rate": 0.00011052274044366711,
      "loss": 1.3504,
      "step": 705
    },
    {
      "epoch": 0.09263572248646876,
      "grad_norm": 98.45941925048828,
      "learning_rate": 0.00011031304263787812,
      "loss": 1.3459,
      "step": 706
    },
    {
      "epoch": 0.09276693455797934,
      "grad_norm": 91.16413116455078,
      "learning_rate": 0.00011010329898477891,
      "loss": 1.5099,
      "step": 707
    },
    {
      "epoch": 0.09289814662948992,
      "grad_norm": 96.37604522705078,
      "learning_rate": 0.0001098935104167988,
      "loss": 1.3623,
      "step": 708
    },
    {
      "epoch": 0.09302935870100049,
      "grad_norm": 93.7906723022461,
      "learning_rate": 0.00010968367786656663,
      "loss": 1.2122,
      "step": 709
    },
    {
      "epoch": 0.09316057077251107,
      "grad_norm": 310.4654541015625,
      "learning_rate": 0.00010947380226690684,
      "loss": 1.288,
      "step": 710
    },
    {
      "epoch": 0.09329178284402165,
      "grad_norm": 95.25032043457031,
      "learning_rate": 0.00010926388455083522,
      "loss": 1.4071,
      "step": 711
    },
    {
      "epoch": 0.09342299491553223,
      "grad_norm": 91.64644622802734,
      "learning_rate": 0.00010905392565155477,
      "loss": 1.2861,
      "step": 712
    },
    {
      "epoch": 0.09355420698704281,
      "grad_norm": 91.60810089111328,
      "learning_rate": 0.00010884392650245165,
      "loss": 1.4312,
      "step": 713
    },
    {
      "epoch": 0.09368541905855339,
      "grad_norm": 91.65507507324219,
      "learning_rate": 0.00010863388803709089,
      "loss": 1.6009,
      "step": 714
    },
    {
      "epoch": 0.09381663113006397,
      "grad_norm": 89.88408660888672,
      "learning_rate": 0.00010842381118921232,
      "loss": 1.3242,
      "step": 715
    },
    {
      "epoch": 0.09394784320157454,
      "grad_norm": 79.47547149658203,
      "learning_rate": 0.00010821369689272638,
      "loss": 1.108,
      "step": 716
    },
    {
      "epoch": 0.09407905527308512,
      "grad_norm": 85.63246154785156,
      "learning_rate": 0.00010800354608171003,
      "loss": 1.2516,
      "step": 717
    },
    {
      "epoch": 0.0942102673445957,
      "grad_norm": 93.73361206054688,
      "learning_rate": 0.00010779335969040252,
      "loss": 1.6184,
      "step": 718
    },
    {
      "epoch": 0.09434147941610628,
      "grad_norm": 69.55641174316406,
      "learning_rate": 0.00010758313865320134,
      "loss": 0.8027,
      "step": 719
    },
    {
      "epoch": 0.09447269148761686,
      "grad_norm": 79.26919555664062,
      "learning_rate": 0.00010737288390465792,
      "loss": 1.3146,
      "step": 720
    },
    {
      "epoch": 0.09460390355912744,
      "grad_norm": 88.47747802734375,
      "learning_rate": 0.00010716259637947357,
      "loss": 1.4076,
      "step": 721
    },
    {
      "epoch": 0.09473511563063802,
      "grad_norm": 78.3600845336914,
      "learning_rate": 0.00010695227701249537,
      "loss": 1.1143,
      "step": 722
    },
    {
      "epoch": 0.0948663277021486,
      "grad_norm": 96.30194091796875,
      "learning_rate": 0.00010674192673871191,
      "loss": 1.2686,
      "step": 723
    },
    {
      "epoch": 0.09499753977365917,
      "grad_norm": 98.24201202392578,
      "learning_rate": 0.00010653154649324917,
      "loss": 1.2332,
      "step": 724
    },
    {
      "epoch": 0.09512875184516975,
      "grad_norm": 99.90653228759766,
      "learning_rate": 0.00010632113721136636,
      "loss": 1.354,
      "step": 725
    },
    {
      "epoch": 0.09525996391668033,
      "grad_norm": 97.12633514404297,
      "learning_rate": 0.00010611069982845183,
      "loss": 1.2428,
      "step": 726
    },
    {
      "epoch": 0.09539117598819091,
      "grad_norm": 92.02963256835938,
      "learning_rate": 0.00010590023528001884,
      "loss": 1.2524,
      "step": 727
    },
    {
      "epoch": 0.0955223880597015,
      "grad_norm": 83.4082260131836,
      "learning_rate": 0.00010568974450170139,
      "loss": 1.3368,
      "step": 728
    },
    {
      "epoch": 0.09565360013121207,
      "grad_norm": 94.21345520019531,
      "learning_rate": 0.00010547922842925008,
      "loss": 1.4861,
      "step": 729
    },
    {
      "epoch": 0.09578481220272266,
      "grad_norm": 79.10660552978516,
      "learning_rate": 0.00010526868799852796,
      "loss": 1.0983,
      "step": 730
    },
    {
      "epoch": 0.09591602427423324,
      "grad_norm": 82.46759796142578,
      "learning_rate": 0.0001050581241455064,
      "loss": 1.2602,
      "step": 731
    },
    {
      "epoch": 0.0960472363457438,
      "grad_norm": 93.12700653076172,
      "learning_rate": 0.00010484753780626089,
      "loss": 1.2674,
      "step": 732
    },
    {
      "epoch": 0.09617844841725438,
      "grad_norm": 90.5476303100586,
      "learning_rate": 0.00010463692991696685,
      "loss": 1.4807,
      "step": 733
    },
    {
      "epoch": 0.09630966048876496,
      "grad_norm": 94.627685546875,
      "learning_rate": 0.00010442630141389549,
      "loss": 1.3896,
      "step": 734
    },
    {
      "epoch": 0.09644087256027555,
      "grad_norm": 70.74127960205078,
      "learning_rate": 0.00010421565323340971,
      "loss": 0.7906,
      "step": 735
    },
    {
      "epoch": 0.09657208463178613,
      "grad_norm": 77.0815200805664,
      "learning_rate": 0.00010400498631195992,
      "loss": 1.1226,
      "step": 736
    },
    {
      "epoch": 0.0967032967032967,
      "grad_norm": 92.74217987060547,
      "learning_rate": 0.00010379430158607975,
      "loss": 1.36,
      "step": 737
    },
    {
      "epoch": 0.09683450877480729,
      "grad_norm": 85.06411743164062,
      "learning_rate": 0.000103583599992382,
      "loss": 1.0775,
      "step": 738
    },
    {
      "epoch": 0.09696572084631785,
      "grad_norm": 86.8634033203125,
      "learning_rate": 0.0001033728824675545,
      "loss": 1.0086,
      "step": 739
    },
    {
      "epoch": 0.09709693291782844,
      "grad_norm": 107.71949768066406,
      "learning_rate": 0.0001031621499483559,
      "loss": 1.3765,
      "step": 740
    },
    {
      "epoch": 0.09722814498933902,
      "grad_norm": 95.89998626708984,
      "learning_rate": 0.00010295140337161146,
      "loss": 1.1242,
      "step": 741
    },
    {
      "epoch": 0.0973593570608496,
      "grad_norm": 88.23342895507812,
      "learning_rate": 0.00010274064367420897,
      "loss": 1.0392,
      "step": 742
    },
    {
      "epoch": 0.09749056913236018,
      "grad_norm": 92.08667755126953,
      "learning_rate": 0.00010252987179309459,
      "loss": 1.2866,
      "step": 743
    },
    {
      "epoch": 0.09762178120387076,
      "grad_norm": 93.89356231689453,
      "learning_rate": 0.00010231908866526851,
      "loss": 1.0697,
      "step": 744
    },
    {
      "epoch": 0.09775299327538134,
      "grad_norm": 95.6974105834961,
      "learning_rate": 0.00010210829522778111,
      "loss": 1.4399,
      "step": 745
    },
    {
      "epoch": 0.09788420534689192,
      "grad_norm": 91.04817199707031,
      "learning_rate": 0.00010189749241772844,
      "loss": 1.3993,
      "step": 746
    },
    {
      "epoch": 0.09801541741840249,
      "grad_norm": 93.00709533691406,
      "learning_rate": 0.00010168668117224825,
      "loss": 1.1606,
      "step": 747
    },
    {
      "epoch": 0.09814662948991307,
      "grad_norm": 82.47737121582031,
      "learning_rate": 0.00010147586242851585,
      "loss": 1.1779,
      "step": 748
    },
    {
      "epoch": 0.09827784156142365,
      "grad_norm": 84.99987030029297,
      "learning_rate": 0.00010126503712373982,
      "loss": 1.3097,
      "step": 749
    },
    {
      "epoch": 0.09840905363293423,
      "grad_norm": 97.76146697998047,
      "learning_rate": 0.00010105420619515798,
      "loss": 1.231,
      "step": 750
    },
    {
      "epoch": 0.09840905363293423,
      "eval_loss": 1.2570799589157104,
      "eval_runtime": 761.1535,
      "eval_samples_per_second": 16.864,
      "eval_steps_per_second": 8.432,
      "step": 750
    },
    {
      "epoch": 0.09854026570444481,
      "grad_norm": 97.69029998779297,
      "learning_rate": 0.00010084337058003303,
      "loss": 1.1091,
      "step": 751
    },
    {
      "epoch": 0.09867147777595539,
      "grad_norm": 93.81666564941406,
      "learning_rate": 0.00010063253121564868,
      "loss": 1.4851,
      "step": 752
    },
    {
      "epoch": 0.09880268984746597,
      "grad_norm": 61.353031158447266,
      "learning_rate": 0.00010042168903930514,
      "loss": 0.7647,
      "step": 753
    },
    {
      "epoch": 0.09893390191897655,
      "grad_norm": 91.40865325927734,
      "learning_rate": 0.00010021084498831522,
      "loss": 1.3089,
      "step": 754
    },
    {
      "epoch": 0.09906511399048712,
      "grad_norm": 101.84923553466797,
      "learning_rate": 0.0001,
      "loss": 1.4239,
      "step": 755
    },
    {
      "epoch": 0.0991963260619977,
      "grad_norm": 97.03103637695312,
      "learning_rate": 9.97891550116848e-05,
      "loss": 1.3613,
      "step": 756
    },
    {
      "epoch": 0.09932753813350828,
      "grad_norm": 89.84929656982422,
      "learning_rate": 9.957831096069488e-05,
      "loss": 1.3314,
      "step": 757
    },
    {
      "epoch": 0.09945875020501886,
      "grad_norm": 74.20384979248047,
      "learning_rate": 9.936746878435136e-05,
      "loss": 1.1078,
      "step": 758
    },
    {
      "epoch": 0.09958996227652944,
      "grad_norm": 85.77796936035156,
      "learning_rate": 9.915662941996699e-05,
      "loss": 1.3821,
      "step": 759
    },
    {
      "epoch": 0.09972117434804002,
      "grad_norm": 80.18834686279297,
      "learning_rate": 9.894579380484204e-05,
      "loss": 1.4241,
      "step": 760
    },
    {
      "epoch": 0.0998523864195506,
      "grad_norm": 105.5184097290039,
      "learning_rate": 9.873496287626019e-05,
      "loss": 1.4285,
      "step": 761
    },
    {
      "epoch": 0.09998359849106117,
      "grad_norm": 80.73155975341797,
      "learning_rate": 9.852413757148417e-05,
      "loss": 1.3009,
      "step": 762
    },
    {
      "epoch": 0.10011481056257175,
      "grad_norm": 77.9684829711914,
      "learning_rate": 9.831331882775178e-05,
      "loss": 1.2135,
      "step": 763
    },
    {
      "epoch": 0.10024602263408233,
      "grad_norm": 90.05245971679688,
      "learning_rate": 9.81025075822716e-05,
      "loss": 1.3076,
      "step": 764
    },
    {
      "epoch": 0.10037723470559291,
      "grad_norm": 90.34455108642578,
      "learning_rate": 9.789170477221891e-05,
      "loss": 1.4653,
      "step": 765
    },
    {
      "epoch": 0.1005084467771035,
      "grad_norm": 78.91496276855469,
      "learning_rate": 9.76809113347315e-05,
      "loss": 1.2576,
      "step": 766
    },
    {
      "epoch": 0.10063965884861407,
      "grad_norm": 89.83158111572266,
      "learning_rate": 9.747012820690543e-05,
      "loss": 1.2744,
      "step": 767
    },
    {
      "epoch": 0.10077087092012466,
      "grad_norm": 101.65666961669922,
      "learning_rate": 9.725935632579104e-05,
      "loss": 1.2676,
      "step": 768
    },
    {
      "epoch": 0.10090208299163524,
      "grad_norm": 84.8772964477539,
      "learning_rate": 9.704859662838855e-05,
      "loss": 1.0624,
      "step": 769
    },
    {
      "epoch": 0.1010332950631458,
      "grad_norm": 86.49850463867188,
      "learning_rate": 9.683785005164411e-05,
      "loss": 1.4636,
      "step": 770
    },
    {
      "epoch": 0.10116450713465638,
      "grad_norm": 96.36685943603516,
      "learning_rate": 9.662711753244551e-05,
      "loss": 1.2227,
      "step": 771
    },
    {
      "epoch": 0.10129571920616696,
      "grad_norm": 90.46879577636719,
      "learning_rate": 9.641640000761802e-05,
      "loss": 1.2893,
      "step": 772
    },
    {
      "epoch": 0.10142693127767755,
      "grad_norm": 87.37088012695312,
      "learning_rate": 9.620569841392029e-05,
      "loss": 1.301,
      "step": 773
    },
    {
      "epoch": 0.10155814334918813,
      "grad_norm": 82.70668029785156,
      "learning_rate": 9.59950136880401e-05,
      "loss": 1.2097,
      "step": 774
    },
    {
      "epoch": 0.10168935542069871,
      "grad_norm": 80.88251495361328,
      "learning_rate": 9.57843467665903e-05,
      "loss": 1.2546,
      "step": 775
    },
    {
      "epoch": 0.10182056749220929,
      "grad_norm": 87.06242370605469,
      "learning_rate": 9.557369858610453e-05,
      "loss": 1.2342,
      "step": 776
    },
    {
      "epoch": 0.10195177956371987,
      "grad_norm": 95.01030731201172,
      "learning_rate": 9.53630700830332e-05,
      "loss": 1.3376,
      "step": 777
    },
    {
      "epoch": 0.10208299163523044,
      "grad_norm": 87.33165740966797,
      "learning_rate": 9.51524621937391e-05,
      "loss": 1.331,
      "step": 778
    },
    {
      "epoch": 0.10221420370674102,
      "grad_norm": 61.26624298095703,
      "learning_rate": 9.494187585449358e-05,
      "loss": 0.659,
      "step": 779
    },
    {
      "epoch": 0.1023454157782516,
      "grad_norm": 92.8788833618164,
      "learning_rate": 9.473131200147205e-05,
      "loss": 1.395,
      "step": 780
    },
    {
      "epoch": 0.10247662784976218,
      "grad_norm": 73.27294158935547,
      "learning_rate": 9.452077157074994e-05,
      "loss": 1.0632,
      "step": 781
    },
    {
      "epoch": 0.10260783992127276,
      "grad_norm": 91.13346862792969,
      "learning_rate": 9.431025549829862e-05,
      "loss": 1.3306,
      "step": 782
    },
    {
      "epoch": 0.10273905199278334,
      "grad_norm": 95.94681549072266,
      "learning_rate": 9.409976471998118e-05,
      "loss": 1.1825,
      "step": 783
    },
    {
      "epoch": 0.10287026406429392,
      "grad_norm": 97.73973083496094,
      "learning_rate": 9.388930017154819e-05,
      "loss": 1.3928,
      "step": 784
    },
    {
      "epoch": 0.10300147613580449,
      "grad_norm": 96.57628631591797,
      "learning_rate": 9.367886278863366e-05,
      "loss": 1.3381,
      "step": 785
    },
    {
      "epoch": 0.10313268820731507,
      "grad_norm": 76.6468276977539,
      "learning_rate": 9.346845350675088e-05,
      "loss": 0.8431,
      "step": 786
    },
    {
      "epoch": 0.10326390027882565,
      "grad_norm": 91.9226303100586,
      "learning_rate": 9.325807326128814e-05,
      "loss": 1.3495,
      "step": 787
    },
    {
      "epoch": 0.10339511235033623,
      "grad_norm": 83.96055603027344,
      "learning_rate": 9.304772298750463e-05,
      "loss": 1.1399,
      "step": 788
    },
    {
      "epoch": 0.10352632442184681,
      "grad_norm": 110.01771545410156,
      "learning_rate": 9.283740362052642e-05,
      "loss": 1.2838,
      "step": 789
    },
    {
      "epoch": 0.10365753649335739,
      "grad_norm": 85.91791534423828,
      "learning_rate": 9.26271160953421e-05,
      "loss": 1.1938,
      "step": 790
    },
    {
      "epoch": 0.10378874856486797,
      "grad_norm": 84.15980529785156,
      "learning_rate": 9.241686134679867e-05,
      "loss": 1.3805,
      "step": 791
    },
    {
      "epoch": 0.10391996063637855,
      "grad_norm": 101.84619903564453,
      "learning_rate": 9.220664030959749e-05,
      "loss": 1.4642,
      "step": 792
    },
    {
      "epoch": 0.10405117270788912,
      "grad_norm": 82.17317199707031,
      "learning_rate": 9.199645391828999e-05,
      "loss": 1.1267,
      "step": 793
    },
    {
      "epoch": 0.1041823847793997,
      "grad_norm": 86.27032470703125,
      "learning_rate": 9.178630310727365e-05,
      "loss": 1.3437,
      "step": 794
    },
    {
      "epoch": 0.10431359685091028,
      "grad_norm": 79.38684844970703,
      "learning_rate": 9.157618881078772e-05,
      "loss": 1.1762,
      "step": 795
    },
    {
      "epoch": 0.10444480892242086,
      "grad_norm": 89.62006378173828,
      "learning_rate": 9.136611196290915e-05,
      "loss": 1.4621,
      "step": 796
    },
    {
      "epoch": 0.10457602099393144,
      "grad_norm": 95.24529266357422,
      "learning_rate": 9.115607349754834e-05,
      "loss": 1.199,
      "step": 797
    },
    {
      "epoch": 0.10470723306544202,
      "grad_norm": 84.98068237304688,
      "learning_rate": 9.094607434844523e-05,
      "loss": 1.1134,
      "step": 798
    },
    {
      "epoch": 0.1048384451369526,
      "grad_norm": 103.12364959716797,
      "learning_rate": 9.07361154491648e-05,
      "loss": 1.352,
      "step": 799
    },
    {
      "epoch": 0.10496965720846319,
      "grad_norm": 86.50543975830078,
      "learning_rate": 9.052619773309317e-05,
      "loss": 0.9736,
      "step": 800
    },
    {
      "epoch": 0.10510086927997375,
      "grad_norm": 91.45311737060547,
      "learning_rate": 9.031632213343339e-05,
      "loss": 1.3331,
      "step": 801
    },
    {
      "epoch": 0.10523208135148433,
      "grad_norm": 85.41616821289062,
      "learning_rate": 9.01064895832012e-05,
      "loss": 1.3502,
      "step": 802
    },
    {
      "epoch": 0.10536329342299491,
      "grad_norm": 74.28988647460938,
      "learning_rate": 8.98967010152211e-05,
      "loss": 0.9898,
      "step": 803
    },
    {
      "epoch": 0.1054945054945055,
      "grad_norm": 89.07300567626953,
      "learning_rate": 8.968695736212193e-05,
      "loss": 1.4575,
      "step": 804
    },
    {
      "epoch": 0.10562571756601608,
      "grad_norm": 93.72269439697266,
      "learning_rate": 8.947725955633294e-05,
      "loss": 1.6465,
      "step": 805
    },
    {
      "epoch": 0.10575692963752666,
      "grad_norm": 88.29570770263672,
      "learning_rate": 8.926760853007946e-05,
      "loss": 1.424,
      "step": 806
    },
    {
      "epoch": 0.10588814170903724,
      "grad_norm": 73.67832946777344,
      "learning_rate": 8.905800521537905e-05,
      "loss": 1.1864,
      "step": 807
    },
    {
      "epoch": 0.1060193537805478,
      "grad_norm": 80.58191680908203,
      "learning_rate": 8.884845054403699e-05,
      "loss": 1.2277,
      "step": 808
    },
    {
      "epoch": 0.10615056585205838,
      "grad_norm": 75.96244812011719,
      "learning_rate": 8.863894544764236e-05,
      "loss": 1.3263,
      "step": 809
    },
    {
      "epoch": 0.10628177792356897,
      "grad_norm": 79.91869354248047,
      "learning_rate": 8.84294908575639e-05,
      "loss": 1.236,
      "step": 810
    },
    {
      "epoch": 0.10641298999507955,
      "grad_norm": 83.43290710449219,
      "learning_rate": 8.822008770494572e-05,
      "loss": 1.3687,
      "step": 811
    },
    {
      "epoch": 0.10654420206659013,
      "grad_norm": 78.77812957763672,
      "learning_rate": 8.801073692070337e-05,
      "loss": 1.2445,
      "step": 812
    },
    {
      "epoch": 0.10667541413810071,
      "grad_norm": 80.77157592773438,
      "learning_rate": 8.780143943551954e-05,
      "loss": 1.2807,
      "step": 813
    },
    {
      "epoch": 0.10680662620961129,
      "grad_norm": 82.76246643066406,
      "learning_rate": 8.759219617983999e-05,
      "loss": 1.4222,
      "step": 814
    },
    {
      "epoch": 0.10693783828112187,
      "grad_norm": 82.53105163574219,
      "learning_rate": 8.738300808386935e-05,
      "loss": 1.1988,
      "step": 815
    },
    {
      "epoch": 0.10706905035263244,
      "grad_norm": 85.68115997314453,
      "learning_rate": 8.717387607756713e-05,
      "loss": 1.1988,
      "step": 816
    },
    {
      "epoch": 0.10720026242414302,
      "grad_norm": 94.20657348632812,
      "learning_rate": 8.696480109064342e-05,
      "loss": 1.4334,
      "step": 817
    },
    {
      "epoch": 0.1073314744956536,
      "grad_norm": 89.79777526855469,
      "learning_rate": 8.675578405255485e-05,
      "loss": 1.4274,
      "step": 818
    },
    {
      "epoch": 0.10746268656716418,
      "grad_norm": 74.69121551513672,
      "learning_rate": 8.654682589250038e-05,
      "loss": 0.9087,
      "step": 819
    },
    {
      "epoch": 0.10759389863867476,
      "grad_norm": 85.3201675415039,
      "learning_rate": 8.633792753941733e-05,
      "loss": 1.3313,
      "step": 820
    },
    {
      "epoch": 0.10772511071018534,
      "grad_norm": 71.2101058959961,
      "learning_rate": 8.612908992197705e-05,
      "loss": 1.132,
      "step": 821
    },
    {
      "epoch": 0.10785632278169592,
      "grad_norm": 74.04344940185547,
      "learning_rate": 8.592031396858093e-05,
      "loss": 1.2489,
      "step": 822
    },
    {
      "epoch": 0.1079875348532065,
      "grad_norm": 85.58378601074219,
      "learning_rate": 8.571160060735624e-05,
      "loss": 1.4977,
      "step": 823
    },
    {
      "epoch": 0.10811874692471707,
      "grad_norm": 84.59700775146484,
      "learning_rate": 8.550295076615188e-05,
      "loss": 1.3275,
      "step": 824
    },
    {
      "epoch": 0.10824995899622765,
      "grad_norm": 86.85831451416016,
      "learning_rate": 8.529436537253458e-05,
      "loss": 1.4293,
      "step": 825
    },
    {
      "epoch": 0.10838117106773823,
      "grad_norm": 95.97257995605469,
      "learning_rate": 8.508584535378439e-05,
      "loss": 1.2916,
      "step": 826
    },
    {
      "epoch": 0.10851238313924881,
      "grad_norm": 89.05290985107422,
      "learning_rate": 8.487739163689079e-05,
      "loss": 1.2153,
      "step": 827
    },
    {
      "epoch": 0.10864359521075939,
      "grad_norm": 81.8623046875,
      "learning_rate": 8.466900514854847e-05,
      "loss": 1.0889,
      "step": 828
    },
    {
      "epoch": 0.10877480728226997,
      "grad_norm": 90.15696716308594,
      "learning_rate": 8.446068681515334e-05,
      "loss": 1.223,
      "step": 829
    },
    {
      "epoch": 0.10890601935378055,
      "grad_norm": 79.51602172851562,
      "learning_rate": 8.425243756279824e-05,
      "loss": 0.9173,
      "step": 830
    },
    {
      "epoch": 0.10903723142529112,
      "grad_norm": 86.3885726928711,
      "learning_rate": 8.404425831726894e-05,
      "loss": 1.3716,
      "step": 831
    },
    {
      "epoch": 0.1091684434968017,
      "grad_norm": 99.4853286743164,
      "learning_rate": 8.383615000404e-05,
      "loss": 1.4296,
      "step": 832
    },
    {
      "epoch": 0.10929965556831228,
      "grad_norm": 84.57745361328125,
      "learning_rate": 8.362811354827059e-05,
      "loss": 1.2572,
      "step": 833
    },
    {
      "epoch": 0.10943086763982286,
      "grad_norm": 81.55428314208984,
      "learning_rate": 8.342014987480047e-05,
      "loss": 1.1257,
      "step": 834
    },
    {
      "epoch": 0.10956207971133344,
      "grad_norm": 86.83169555664062,
      "learning_rate": 8.321225990814591e-05,
      "loss": 1.3076,
      "step": 835
    },
    {
      "epoch": 0.10969329178284402,
      "grad_norm": 85.37054443359375,
      "learning_rate": 8.300444457249543e-05,
      "loss": 0.985,
      "step": 836
    },
    {
      "epoch": 0.1098245038543546,
      "grad_norm": 84.12065887451172,
      "learning_rate": 8.279670479170573e-05,
      "loss": 1.2913,
      "step": 837
    },
    {
      "epoch": 0.10995571592586519,
      "grad_norm": 94.91079711914062,
      "learning_rate": 8.258904148929775e-05,
      "loss": 1.2579,
      "step": 838
    },
    {
      "epoch": 0.11008692799737575,
      "grad_norm": 74.27262878417969,
      "learning_rate": 8.238145558845235e-05,
      "loss": 1.1638,
      "step": 839
    },
    {
      "epoch": 0.11021814006888633,
      "grad_norm": 75.5075454711914,
      "learning_rate": 8.217394801200631e-05,
      "loss": 1.2136,
      "step": 840
    },
    {
      "epoch": 0.11034935214039691,
      "grad_norm": 91.46548461914062,
      "learning_rate": 8.196651968244826e-05,
      "loss": 1.3125,
      "step": 841
    },
    {
      "epoch": 0.1104805642119075,
      "grad_norm": 110.3213119506836,
      "learning_rate": 8.175917152191447e-05,
      "loss": 1.1892,
      "step": 842
    },
    {
      "epoch": 0.11061177628341808,
      "grad_norm": 90.02698516845703,
      "learning_rate": 8.15519044521848e-05,
      "loss": 1.2794,
      "step": 843
    },
    {
      "epoch": 0.11074298835492866,
      "grad_norm": 78.54035186767578,
      "learning_rate": 8.134471939467874e-05,
      "loss": 1.2092,
      "step": 844
    },
    {
      "epoch": 0.11087420042643924,
      "grad_norm": 87.6242904663086,
      "learning_rate": 8.113761727045105e-05,
      "loss": 1.3359,
      "step": 845
    },
    {
      "epoch": 0.11100541249794982,
      "grad_norm": 97.61687469482422,
      "learning_rate": 8.093059900018792e-05,
      "loss": 1.5146,
      "step": 846
    },
    {
      "epoch": 0.11113662456946038,
      "grad_norm": 77.97388458251953,
      "learning_rate": 8.072366550420266e-05,
      "loss": 1.184,
      "step": 847
    },
    {
      "epoch": 0.11126783664097097,
      "grad_norm": 106.9185791015625,
      "learning_rate": 8.051681770243175e-05,
      "loss": 1.2764,
      "step": 848
    },
    {
      "epoch": 0.11139904871248155,
      "grad_norm": 104.53775024414062,
      "learning_rate": 8.031005651443073e-05,
      "loss": 1.3336,
      "step": 849
    },
    {
      "epoch": 0.11153026078399213,
      "grad_norm": 70.69134521484375,
      "learning_rate": 8.010338285937006e-05,
      "loss": 0.8172,
      "step": 850
    },
    {
      "epoch": 0.11166147285550271,
      "grad_norm": 72.92957305908203,
      "learning_rate": 7.989679765603108e-05,
      "loss": 1.1501,
      "step": 851
    },
    {
      "epoch": 0.11179268492701329,
      "grad_norm": 93.121337890625,
      "learning_rate": 7.969030182280192e-05,
      "loss": 1.3874,
      "step": 852
    },
    {
      "epoch": 0.11192389699852387,
      "grad_norm": 90.40266418457031,
      "learning_rate": 7.948389627767343e-05,
      "loss": 1.2266,
      "step": 853
    },
    {
      "epoch": 0.11205510907003444,
      "grad_norm": 80.4605484008789,
      "learning_rate": 7.927758193823501e-05,
      "loss": 1.2867,
      "step": 854
    },
    {
      "epoch": 0.11218632114154502,
      "grad_norm": 76.1406478881836,
      "learning_rate": 7.907135972167069e-05,
      "loss": 1.1436,
      "step": 855
    },
    {
      "epoch": 0.1123175332130556,
      "grad_norm": 117.57950592041016,
      "learning_rate": 7.88652305447549e-05,
      "loss": 1.4234,
      "step": 856
    },
    {
      "epoch": 0.11244874528456618,
      "grad_norm": 83.21184539794922,
      "learning_rate": 7.865919532384844e-05,
      "loss": 1.406,
      "step": 857
    },
    {
      "epoch": 0.11257995735607676,
      "grad_norm": 81.59228515625,
      "learning_rate": 7.845325497489449e-05,
      "loss": 1.1652,
      "step": 858
    },
    {
      "epoch": 0.11271116942758734,
      "grad_norm": 78.01521301269531,
      "learning_rate": 7.82474104134144e-05,
      "loss": 1.1525,
      "step": 859
    },
    {
      "epoch": 0.11284238149909792,
      "grad_norm": 79.91583251953125,
      "learning_rate": 7.804166255450373e-05,
      "loss": 1.1489,
      "step": 860
    },
    {
      "epoch": 0.1129735935706085,
      "grad_norm": 93.8932876586914,
      "learning_rate": 7.783601231282812e-05,
      "loss": 1.5593,
      "step": 861
    },
    {
      "epoch": 0.11310480564211907,
      "grad_norm": 81.15459442138672,
      "learning_rate": 7.763046060261932e-05,
      "loss": 1.1166,
      "step": 862
    },
    {
      "epoch": 0.11323601771362965,
      "grad_norm": 81.07710266113281,
      "learning_rate": 7.742500833767094e-05,
      "loss": 1.259,
      "step": 863
    },
    {
      "epoch": 0.11336722978514023,
      "grad_norm": 86.97444152832031,
      "learning_rate": 7.721965643133458e-05,
      "loss": 1.375,
      "step": 864
    },
    {
      "epoch": 0.11349844185665081,
      "grad_norm": 86.7616195678711,
      "learning_rate": 7.701440579651566e-05,
      "loss": 1.2574,
      "step": 865
    },
    {
      "epoch": 0.11362965392816139,
      "grad_norm": 72.26502227783203,
      "learning_rate": 7.680925734566937e-05,
      "loss": 0.9285,
      "step": 866
    },
    {
      "epoch": 0.11376086599967197,
      "grad_norm": 85.68944549560547,
      "learning_rate": 7.660421199079669e-05,
      "loss": 1.3314,
      "step": 867
    },
    {
      "epoch": 0.11389207807118255,
      "grad_norm": 105.01937103271484,
      "learning_rate": 7.639927064344022e-05,
      "loss": 1.4191,
      "step": 868
    },
    {
      "epoch": 0.11402329014269313,
      "grad_norm": 76.75748443603516,
      "learning_rate": 7.619443421468021e-05,
      "loss": 1.2067,
      "step": 869
    },
    {
      "epoch": 0.1141545022142037,
      "grad_norm": 71.16960144042969,
      "learning_rate": 7.598970361513051e-05,
      "loss": 1.1203,
      "step": 870
    },
    {
      "epoch": 0.11428571428571428,
      "grad_norm": 91.62995147705078,
      "learning_rate": 7.578507975493448e-05,
      "loss": 1.4054,
      "step": 871
    },
    {
      "epoch": 0.11441692635722486,
      "grad_norm": 93.81131744384766,
      "learning_rate": 7.558056354376098e-05,
      "loss": 1.4383,
      "step": 872
    },
    {
      "epoch": 0.11454813842873544,
      "grad_norm": 90.25621032714844,
      "learning_rate": 7.537615589080027e-05,
      "loss": 1.3775,
      "step": 873
    },
    {
      "epoch": 0.11467935050024602,
      "grad_norm": 91.32057189941406,
      "learning_rate": 7.517185770476006e-05,
      "loss": 1.3503,
      "step": 874
    },
    {
      "epoch": 0.1148105625717566,
      "grad_norm": 84.20240020751953,
      "learning_rate": 7.496766989386136e-05,
      "loss": 1.2805,
      "step": 875
    },
    {
      "epoch": 0.11494177464326719,
      "grad_norm": 89.70222473144531,
      "learning_rate": 7.476359336583454e-05,
      "loss": 1.0532,
      "step": 876
    },
    {
      "epoch": 0.11507298671477775,
      "grad_norm": 84.67049407958984,
      "learning_rate": 7.455962902791522e-05,
      "loss": 1.2905,
      "step": 877
    },
    {
      "epoch": 0.11520419878628833,
      "grad_norm": 87.71664428710938,
      "learning_rate": 7.435577778684033e-05,
      "loss": 1.2024,
      "step": 878
    },
    {
      "epoch": 0.11533541085779891,
      "grad_norm": 79.67853546142578,
      "learning_rate": 7.415204054884399e-05,
      "loss": 1.1998,
      "step": 879
    },
    {
      "epoch": 0.1154666229293095,
      "grad_norm": 73.01593780517578,
      "learning_rate": 7.394841821965345e-05,
      "loss": 0.8399,
      "step": 880
    },
    {
      "epoch": 0.11559783500082008,
      "grad_norm": 89.91947174072266,
      "learning_rate": 7.374491170448525e-05,
      "loss": 1.1487,
      "step": 881
    },
    {
      "epoch": 0.11572904707233066,
      "grad_norm": 84.5108642578125,
      "learning_rate": 7.3541521908041e-05,
      "loss": 1.2394,
      "step": 882
    },
    {
      "epoch": 0.11586025914384124,
      "grad_norm": 76.39949035644531,
      "learning_rate": 7.33382497345034e-05,
      "loss": 1.1252,
      "step": 883
    },
    {
      "epoch": 0.11599147121535182,
      "grad_norm": 97.69725799560547,
      "learning_rate": 7.313509608753231e-05,
      "loss": 1.3952,
      "step": 884
    },
    {
      "epoch": 0.11612268328686239,
      "grad_norm": 81.33671569824219,
      "learning_rate": 7.293206187026061e-05,
      "loss": 1.2021,
      "step": 885
    },
    {
      "epoch": 0.11625389535837297,
      "grad_norm": 69.44561767578125,
      "learning_rate": 7.27291479852903e-05,
      "loss": 1.0298,
      "step": 886
    },
    {
      "epoch": 0.11638510742988355,
      "grad_norm": 69.80852508544922,
      "learning_rate": 7.252635533468843e-05,
      "loss": 1.1035,
      "step": 887
    },
    {
      "epoch": 0.11651631950139413,
      "grad_norm": 88.15672302246094,
      "learning_rate": 7.232368481998309e-05,
      "loss": 1.1554,
      "step": 888
    },
    {
      "epoch": 0.11664753157290471,
      "grad_norm": 79.34785461425781,
      "learning_rate": 7.212113734215932e-05,
      "loss": 1.2661,
      "step": 889
    },
    {
      "epoch": 0.11677874364441529,
      "grad_norm": 76.99122619628906,
      "learning_rate": 7.191871380165538e-05,
      "loss": 1.1134,
      "step": 890
    },
    {
      "epoch": 0.11690995571592587,
      "grad_norm": 83.0638656616211,
      "learning_rate": 7.17164150983584e-05,
      "loss": 1.2442,
      "step": 891
    },
    {
      "epoch": 0.11704116778743645,
      "grad_norm": 82.86811065673828,
      "learning_rate": 7.151424213160061e-05,
      "loss": 1.316,
      "step": 892
    },
    {
      "epoch": 0.11717237985894702,
      "grad_norm": 80.93962097167969,
      "learning_rate": 7.131219580015521e-05,
      "loss": 1.2626,
      "step": 893
    },
    {
      "epoch": 0.1173035919304576,
      "grad_norm": 96.833984375,
      "learning_rate": 7.11102770022325e-05,
      "loss": 1.4285,
      "step": 894
    },
    {
      "epoch": 0.11743480400196818,
      "grad_norm": 77.8813705444336,
      "learning_rate": 7.090848663547574e-05,
      "loss": 1.2855,
      "step": 895
    },
    {
      "epoch": 0.11756601607347876,
      "grad_norm": 102.7423095703125,
      "learning_rate": 7.070682559695736e-05,
      "loss": 1.5953,
      "step": 896
    },
    {
      "epoch": 0.11769722814498934,
      "grad_norm": 77.7080078125,
      "learning_rate": 7.050529478317476e-05,
      "loss": 1.282,
      "step": 897
    },
    {
      "epoch": 0.11782844021649992,
      "grad_norm": 87.76219940185547,
      "learning_rate": 7.03038950900464e-05,
      "loss": 1.1632,
      "step": 898
    },
    {
      "epoch": 0.1179596522880105,
      "grad_norm": 78.38594818115234,
      "learning_rate": 7.010262741290798e-05,
      "loss": 1.0336,
      "step": 899
    },
    {
      "epoch": 0.11809086435952107,
      "grad_norm": 73.42670440673828,
      "learning_rate": 6.990149264650814e-05,
      "loss": 0.9046,
      "step": 900
    },
    {
      "epoch": 0.11822207643103165,
      "grad_norm": 76.01429748535156,
      "learning_rate": 6.970049168500474e-05,
      "loss": 1.0919,
      "step": 901
    },
    {
      "epoch": 0.11835328850254223,
      "grad_norm": 102.96312713623047,
      "learning_rate": 6.94996254219608e-05,
      "loss": 1.1038,
      "step": 902
    },
    {
      "epoch": 0.11848450057405281,
      "grad_norm": 95.3603744506836,
      "learning_rate": 6.929889475034048e-05,
      "loss": 1.427,
      "step": 903
    },
    {
      "epoch": 0.11861571264556339,
      "grad_norm": 95.59170532226562,
      "learning_rate": 6.909830056250527e-05,
      "loss": 1.4869,
      "step": 904
    },
    {
      "epoch": 0.11874692471707397,
      "grad_norm": 80.96112823486328,
      "learning_rate": 6.889784375020978e-05,
      "loss": 0.9966,
      "step": 905
    },
    {
      "epoch": 0.11887813678858455,
      "grad_norm": 86.87613677978516,
      "learning_rate": 6.869752520459803e-05,
      "loss": 1.4059,
      "step": 906
    },
    {
      "epoch": 0.11900934886009513,
      "grad_norm": 80.7220687866211,
      "learning_rate": 6.849734581619918e-05,
      "loss": 1.2773,
      "step": 907
    },
    {
      "epoch": 0.1191405609316057,
      "grad_norm": 74.66370391845703,
      "learning_rate": 6.829730647492404e-05,
      "loss": 0.9948,
      "step": 908
    },
    {
      "epoch": 0.11927177300311628,
      "grad_norm": 80.79701232910156,
      "learning_rate": 6.80974080700606e-05,
      "loss": 1.3181,
      "step": 909
    },
    {
      "epoch": 0.11940298507462686,
      "grad_norm": 83.53121948242188,
      "learning_rate": 6.789765149027039e-05,
      "loss": 1.1188,
      "step": 910
    },
    {
      "epoch": 0.11953419714613744,
      "grad_norm": 79.6371078491211,
      "learning_rate": 6.769803762358443e-05,
      "loss": 1.0999,
      "step": 911
    },
    {
      "epoch": 0.11966540921764802,
      "grad_norm": 78.49540710449219,
      "learning_rate": 6.749856735739928e-05,
      "loss": 1.1543,
      "step": 912
    },
    {
      "epoch": 0.1197966212891586,
      "grad_norm": 90.95901489257812,
      "learning_rate": 6.729924157847323e-05,
      "loss": 1.077,
      "step": 913
    },
    {
      "epoch": 0.11992783336066919,
      "grad_norm": 89.46601104736328,
      "learning_rate": 6.710006117292209e-05,
      "loss": 1.2309,
      "step": 914
    },
    {
      "epoch": 0.12005904543217977,
      "grad_norm": 81.42534637451172,
      "learning_rate": 6.690102702621548e-05,
      "loss": 0.8973,
      "step": 915
    },
    {
      "epoch": 0.12019025750369033,
      "grad_norm": 86.31146240234375,
      "learning_rate": 6.670214002317278e-05,
      "loss": 1.1906,
      "step": 916
    },
    {
      "epoch": 0.12032146957520092,
      "grad_norm": 77.64242553710938,
      "learning_rate": 6.650340104795932e-05,
      "loss": 1.1253,
      "step": 917
    },
    {
      "epoch": 0.1204526816467115,
      "grad_norm": 96.30457305908203,
      "learning_rate": 6.630481098408228e-05,
      "loss": 1.0787,
      "step": 918
    },
    {
      "epoch": 0.12058389371822208,
      "grad_norm": 86.64209747314453,
      "learning_rate": 6.610637071438686e-05,
      "loss": 1.0846,
      "step": 919
    },
    {
      "epoch": 0.12071510578973266,
      "grad_norm": 120.56439208984375,
      "learning_rate": 6.590808112105232e-05,
      "loss": 1.406,
      "step": 920
    },
    {
      "epoch": 0.12084631786124324,
      "grad_norm": 82.7140121459961,
      "learning_rate": 6.570994308558812e-05,
      "loss": 1.2397,
      "step": 921
    },
    {
      "epoch": 0.12097752993275382,
      "grad_norm": 100.1551742553711,
      "learning_rate": 6.551195748882997e-05,
      "loss": 1.3067,
      "step": 922
    },
    {
      "epoch": 0.12110874200426439,
      "grad_norm": 98.75762939453125,
      "learning_rate": 6.531412521093586e-05,
      "loss": 1.4248,
      "step": 923
    },
    {
      "epoch": 0.12123995407577497,
      "grad_norm": 84.06700134277344,
      "learning_rate": 6.51164471313822e-05,
      "loss": 1.2859,
      "step": 924
    },
    {
      "epoch": 0.12137116614728555,
      "grad_norm": 101.31942749023438,
      "learning_rate": 6.491892412895995e-05,
      "loss": 1.339,
      "step": 925
    },
    {
      "epoch": 0.12150237821879613,
      "grad_norm": 80.5140380859375,
      "learning_rate": 6.472155708177052e-05,
      "loss": 1.1143,
      "step": 926
    },
    {
      "epoch": 0.12163359029030671,
      "grad_norm": 94.0484848022461,
      "learning_rate": 6.452434686722224e-05,
      "loss": 1.2961,
      "step": 927
    },
    {
      "epoch": 0.12176480236181729,
      "grad_norm": 79.46549987792969,
      "learning_rate": 6.432729436202604e-05,
      "loss": 1.1237,
      "step": 928
    },
    {
      "epoch": 0.12189601443332787,
      "grad_norm": 71.63682556152344,
      "learning_rate": 6.41304004421918e-05,
      "loss": 1.0096,
      "step": 929
    },
    {
      "epoch": 0.12202722650483845,
      "grad_norm": 89.28536987304688,
      "learning_rate": 6.393366598302446e-05,
      "loss": 1.3502,
      "step": 930
    },
    {
      "epoch": 0.12215843857634902,
      "grad_norm": 84.19918823242188,
      "learning_rate": 6.373709185911998e-05,
      "loss": 1.2199,
      "step": 931
    },
    {
      "epoch": 0.1222896506478596,
      "grad_norm": 80.94573211669922,
      "learning_rate": 6.354067894436155e-05,
      "loss": 1.3371,
      "step": 932
    },
    {
      "epoch": 0.12242086271937018,
      "grad_norm": 79.59129333496094,
      "learning_rate": 6.334442811191576e-05,
      "loss": 1.2141,
      "step": 933
    },
    {
      "epoch": 0.12255207479088076,
      "grad_norm": 88.73218536376953,
      "learning_rate": 6.314834023422858e-05,
      "loss": 1.2853,
      "step": 934
    },
    {
      "epoch": 0.12268328686239134,
      "grad_norm": 66.05457305908203,
      "learning_rate": 6.295241618302156e-05,
      "loss": 0.8253,
      "step": 935
    },
    {
      "epoch": 0.12281449893390192,
      "grad_norm": 100.30980682373047,
      "learning_rate": 6.275665682928803e-05,
      "loss": 1.3808,
      "step": 936
    },
    {
      "epoch": 0.1229457110054125,
      "grad_norm": 81.35166931152344,
      "learning_rate": 6.256106304328905e-05,
      "loss": 1.1028,
      "step": 937
    },
    {
      "epoch": 0.12307692307692308,
      "grad_norm": 90.19493103027344,
      "learning_rate": 6.23656356945497e-05,
      "loss": 1.3201,
      "step": 938
    },
    {
      "epoch": 0.12320813514843365,
      "grad_norm": 79.76177215576172,
      "learning_rate": 6.21703756518551e-05,
      "loss": 1.1902,
      "step": 939
    },
    {
      "epoch": 0.12333934721994423,
      "grad_norm": 86.34595489501953,
      "learning_rate": 6.197528378324665e-05,
      "loss": 1.2645,
      "step": 940
    },
    {
      "epoch": 0.12347055929145481,
      "grad_norm": 102.84590911865234,
      "learning_rate": 6.17803609560181e-05,
      "loss": 1.4919,
      "step": 941
    },
    {
      "epoch": 0.12360177136296539,
      "grad_norm": 93.21085357666016,
      "learning_rate": 6.158560803671168e-05,
      "loss": 1.315,
      "step": 942
    },
    {
      "epoch": 0.12373298343447597,
      "grad_norm": 80.23776245117188,
      "learning_rate": 6.139102589111435e-05,
      "loss": 1.1273,
      "step": 943
    },
    {
      "epoch": 0.12386419550598655,
      "grad_norm": 77.1685791015625,
      "learning_rate": 6.119661538425381e-05,
      "loss": 1.2854,
      "step": 944
    },
    {
      "epoch": 0.12399540757749714,
      "grad_norm": 83.83993530273438,
      "learning_rate": 6.100237738039484e-05,
      "loss": 1.1223,
      "step": 945
    },
    {
      "epoch": 0.1241266196490077,
      "grad_norm": 83.0691146850586,
      "learning_rate": 6.0808312743035236e-05,
      "loss": 1.1178,
      "step": 946
    },
    {
      "epoch": 0.12425783172051828,
      "grad_norm": 98.6527328491211,
      "learning_rate": 6.061442233490211e-05,
      "loss": 1.2167,
      "step": 947
    },
    {
      "epoch": 0.12438904379202886,
      "grad_norm": 94.2676773071289,
      "learning_rate": 6.042070701794806e-05,
      "loss": 1.0235,
      "step": 948
    },
    {
      "epoch": 0.12452025586353944,
      "grad_norm": 79.8330307006836,
      "learning_rate": 6.0227167653347305e-05,
      "loss": 1.0503,
      "step": 949
    },
    {
      "epoch": 0.12465146793505003,
      "grad_norm": 80.43983459472656,
      "learning_rate": 6.0033805101491794e-05,
      "loss": 1.0692,
      "step": 950
    },
    {
      "epoch": 0.1247826800065606,
      "grad_norm": 75.39767456054688,
      "learning_rate": 5.98406202219875e-05,
      "loss": 1.1895,
      "step": 951
    },
    {
      "epoch": 0.12491389207807119,
      "grad_norm": 84.26925659179688,
      "learning_rate": 5.964761387365052e-05,
      "loss": 1.3174,
      "step": 952
    },
    {
      "epoch": 0.12504510414958175,
      "grad_norm": 83.67276000976562,
      "learning_rate": 5.9454786914503255e-05,
      "loss": 1.1032,
      "step": 953
    },
    {
      "epoch": 0.12517631622109235,
      "grad_norm": 76.90235900878906,
      "learning_rate": 5.926214020177074e-05,
      "loss": 1.1672,
      "step": 954
    },
    {
      "epoch": 0.12530752829260292,
      "grad_norm": 80.15973663330078,
      "learning_rate": 5.9069674591876534e-05,
      "loss": 1.3362,
      "step": 955
    },
    {
      "epoch": 0.1254387403641135,
      "grad_norm": 77.25790405273438,
      "learning_rate": 5.887739094043923e-05,
      "loss": 1.2179,
      "step": 956
    },
    {
      "epoch": 0.12556995243562408,
      "grad_norm": 78.84754943847656,
      "learning_rate": 5.868529010226845e-05,
      "loss": 1.2128,
      "step": 957
    },
    {
      "epoch": 0.12570116450713464,
      "grad_norm": 73.49491882324219,
      "learning_rate": 5.849337293136112e-05,
      "loss": 1.1622,
      "step": 958
    },
    {
      "epoch": 0.12583237657864524,
      "grad_norm": 67.60201263427734,
      "learning_rate": 5.830164028089766e-05,
      "loss": 0.8057,
      "step": 959
    },
    {
      "epoch": 0.1259635886501558,
      "grad_norm": 89.39290618896484,
      "learning_rate": 5.811009300323818e-05,
      "loss": 1.2938,
      "step": 960
    },
    {
      "epoch": 0.1260948007216664,
      "grad_norm": 74.048095703125,
      "learning_rate": 5.791873194991872e-05,
      "loss": 1.1232,
      "step": 961
    },
    {
      "epoch": 0.12622601279317697,
      "grad_norm": 66.7199478149414,
      "learning_rate": 5.7727557971647427e-05,
      "loss": 0.93,
      "step": 962
    },
    {
      "epoch": 0.12635722486468756,
      "grad_norm": 81.19363403320312,
      "learning_rate": 5.7536571918300864e-05,
      "loss": 1.1275,
      "step": 963
    },
    {
      "epoch": 0.12648843693619813,
      "grad_norm": 83.35269927978516,
      "learning_rate": 5.734577463892008e-05,
      "loss": 1.4104,
      "step": 964
    },
    {
      "epoch": 0.1266196490077087,
      "grad_norm": 73.7481460571289,
      "learning_rate": 5.7155166981706956e-05,
      "loss": 1.1557,
      "step": 965
    },
    {
      "epoch": 0.1267508610792193,
      "grad_norm": 74.92057037353516,
      "learning_rate": 5.6964749794020354e-05,
      "loss": 1.1487,
      "step": 966
    },
    {
      "epoch": 0.12688207315072986,
      "grad_norm": 76.392578125,
      "learning_rate": 5.6774523922372394e-05,
      "loss": 1.2103,
      "step": 967
    },
    {
      "epoch": 0.12701328522224045,
      "grad_norm": 95.00236511230469,
      "learning_rate": 5.6584490212424804e-05,
      "loss": 1.3699,
      "step": 968
    },
    {
      "epoch": 0.12714449729375102,
      "grad_norm": 91.27318572998047,
      "learning_rate": 5.639464950898491e-05,
      "loss": 1.273,
      "step": 969
    },
    {
      "epoch": 0.1272757093652616,
      "grad_norm": 69.59209442138672,
      "learning_rate": 5.620500265600206e-05,
      "loss": 0.9777,
      "step": 970
    },
    {
      "epoch": 0.12740692143677218,
      "grad_norm": 79.14124298095703,
      "learning_rate": 5.601555049656382e-05,
      "loss": 1.2307,
      "step": 971
    },
    {
      "epoch": 0.12753813350828277,
      "grad_norm": 78.90977478027344,
      "learning_rate": 5.58262938728922e-05,
      "loss": 1.168,
      "step": 972
    },
    {
      "epoch": 0.12766934557979334,
      "grad_norm": 73.35558319091797,
      "learning_rate": 5.563723362634008e-05,
      "loss": 1.1331,
      "step": 973
    },
    {
      "epoch": 0.1278005576513039,
      "grad_norm": 68.26853942871094,
      "learning_rate": 5.544837059738719e-05,
      "loss": 1.0879,
      "step": 974
    },
    {
      "epoch": 0.1279317697228145,
      "grad_norm": 88.05345916748047,
      "learning_rate": 5.525970562563656e-05,
      "loss": 1.1776,
      "step": 975
    },
    {
      "epoch": 0.12806298179432507,
      "grad_norm": 89.8759994506836,
      "learning_rate": 5.507123954981073e-05,
      "loss": 1.3996,
      "step": 976
    },
    {
      "epoch": 0.12819419386583567,
      "grad_norm": 81.99223327636719,
      "learning_rate": 5.488297320774807e-05,
      "loss": 1.2289,
      "step": 977
    },
    {
      "epoch": 0.12832540593734623,
      "grad_norm": 89.96368408203125,
      "learning_rate": 5.4694907436399e-05,
      "loss": 1.3267,
      "step": 978
    },
    {
      "epoch": 0.12845661800885683,
      "grad_norm": 65.02106475830078,
      "learning_rate": 5.4507043071822284e-05,
      "loss": 0.8666,
      "step": 979
    },
    {
      "epoch": 0.1285878300803674,
      "grad_norm": 72.8196792602539,
      "learning_rate": 5.431938094918132e-05,
      "loss": 1.0557,
      "step": 980
    },
    {
      "epoch": 0.12871904215187796,
      "grad_norm": 80.00843048095703,
      "learning_rate": 5.41319219027404e-05,
      "loss": 1.1253,
      "step": 981
    },
    {
      "epoch": 0.12885025422338856,
      "grad_norm": 87.73186492919922,
      "learning_rate": 5.394466676586114e-05,
      "loss": 1.3141,
      "step": 982
    },
    {
      "epoch": 0.12898146629489912,
      "grad_norm": 82.19802856445312,
      "learning_rate": 5.375761637099854e-05,
      "loss": 1.2266,
      "step": 983
    },
    {
      "epoch": 0.12911267836640972,
      "grad_norm": 81.80357360839844,
      "learning_rate": 5.357077154969742e-05,
      "loss": 1.1676,
      "step": 984
    },
    {
      "epoch": 0.12924389043792028,
      "grad_norm": 64.08479309082031,
      "learning_rate": 5.3384133132588784e-05,
      "loss": 0.9184,
      "step": 985
    },
    {
      "epoch": 0.12937510250943088,
      "grad_norm": 85.98280334472656,
      "learning_rate": 5.3197701949386e-05,
      "loss": 1.3069,
      "step": 986
    },
    {
      "epoch": 0.12950631458094145,
      "grad_norm": 75.3571548461914,
      "learning_rate": 5.301147882888116e-05,
      "loss": 1.0578,
      "step": 987
    },
    {
      "epoch": 0.129637526652452,
      "grad_norm": 69.68834686279297,
      "learning_rate": 5.28254645989414e-05,
      "loss": 0.9666,
      "step": 988
    },
    {
      "epoch": 0.1297687387239626,
      "grad_norm": 70.9851303100586,
      "learning_rate": 5.2639660086505226e-05,
      "loss": 0.9498,
      "step": 989
    },
    {
      "epoch": 0.12989995079547317,
      "grad_norm": 88.52960968017578,
      "learning_rate": 5.2454066117578815e-05,
      "loss": 1.2682,
      "step": 990
    },
    {
      "epoch": 0.13003116286698377,
      "grad_norm": 88.15095520019531,
      "learning_rate": 5.226868351723244e-05,
      "loss": 1.2018,
      "step": 991
    },
    {
      "epoch": 0.13016237493849434,
      "grad_norm": 95.7833480834961,
      "learning_rate": 5.2083513109596616e-05,
      "loss": 1.0795,
      "step": 992
    },
    {
      "epoch": 0.13029358701000493,
      "grad_norm": 81.77001953125,
      "learning_rate": 5.189855571785859e-05,
      "loss": 1.2438,
      "step": 993
    },
    {
      "epoch": 0.1304247990815155,
      "grad_norm": 78.49679565429688,
      "learning_rate": 5.171381216425863e-05,
      "loss": 1.0536,
      "step": 994
    },
    {
      "epoch": 0.1305560111530261,
      "grad_norm": 85.6697998046875,
      "learning_rate": 5.152928327008635e-05,
      "loss": 1.0529,
      "step": 995
    },
    {
      "epoch": 0.13068722322453666,
      "grad_norm": 81.8764419555664,
      "learning_rate": 5.134496985567714e-05,
      "loss": 1.1992,
      "step": 996
    },
    {
      "epoch": 0.13081843529604723,
      "grad_norm": 79.9547119140625,
      "learning_rate": 5.116087274040837e-05,
      "loss": 1.2193,
      "step": 997
    },
    {
      "epoch": 0.13094964736755782,
      "grad_norm": 85.14453125,
      "learning_rate": 5.0976992742695925e-05,
      "loss": 1.0081,
      "step": 998
    },
    {
      "epoch": 0.1310808594390684,
      "grad_norm": 83.98960876464844,
      "learning_rate": 5.07933306799904e-05,
      "loss": 1.1891,
      "step": 999
    },
    {
      "epoch": 0.13121207151057898,
      "grad_norm": 85.7392349243164,
      "learning_rate": 5.060988736877366e-05,
      "loss": 1.2281,
      "step": 1000
    },
    {
      "epoch": 0.13134328358208955,
      "grad_norm": 78.68650817871094,
      "learning_rate": 5.042666362455498e-05,
      "loss": 0.9805,
      "step": 1001
    },
    {
      "epoch": 0.13147449565360014,
      "grad_norm": 87.75737762451172,
      "learning_rate": 5.024366026186755e-05,
      "loss": 1.4402,
      "step": 1002
    },
    {
      "epoch": 0.1316057077251107,
      "grad_norm": 74.76168823242188,
      "learning_rate": 5.006087809426496e-05,
      "loss": 0.9178,
      "step": 1003
    },
    {
      "epoch": 0.13173691979662128,
      "grad_norm": 87.87181854248047,
      "learning_rate": 4.987831793431731e-05,
      "loss": 1.3922,
      "step": 1004
    },
    {
      "epoch": 0.13186813186813187,
      "grad_norm": 80.9386215209961,
      "learning_rate": 4.9695980593607817e-05,
      "loss": 1.1955,
      "step": 1005
    },
    {
      "epoch": 0.13199934393964244,
      "grad_norm": 75.18540954589844,
      "learning_rate": 4.9513866882729146e-05,
      "loss": 1.224,
      "step": 1006
    },
    {
      "epoch": 0.13213055601115303,
      "grad_norm": 82.27153778076172,
      "learning_rate": 4.9331977611279777e-05,
      "loss": 1.2444,
      "step": 1007
    },
    {
      "epoch": 0.1322617680826636,
      "grad_norm": 89.88761138916016,
      "learning_rate": 4.9150313587860433e-05,
      "loss": 1.3091,
      "step": 1008
    },
    {
      "epoch": 0.1323929801541742,
      "grad_norm": 62.554744720458984,
      "learning_rate": 4.896887562007054e-05,
      "loss": 0.8667,
      "step": 1009
    },
    {
      "epoch": 0.13252419222568476,
      "grad_norm": 72.94403839111328,
      "learning_rate": 4.8787664514504504e-05,
      "loss": 1.1776,
      "step": 1010
    },
    {
      "epoch": 0.13265540429719533,
      "grad_norm": 79.98250579833984,
      "learning_rate": 4.860668107674823e-05,
      "loss": 1.2097,
      "step": 1011
    },
    {
      "epoch": 0.13278661636870592,
      "grad_norm": 81.76577758789062,
      "learning_rate": 4.8425926111375506e-05,
      "loss": 1.2133,
      "step": 1012
    },
    {
      "epoch": 0.1329178284402165,
      "grad_norm": 81.15084075927734,
      "learning_rate": 4.824540042194443e-05,
      "loss": 1.1422,
      "step": 1013
    },
    {
      "epoch": 0.13304904051172708,
      "grad_norm": 73.56816101074219,
      "learning_rate": 4.8065104810993856e-05,
      "loss": 0.9783,
      "step": 1014
    },
    {
      "epoch": 0.13318025258323765,
      "grad_norm": 92.46258544921875,
      "learning_rate": 4.788504008003978e-05,
      "loss": 1.1844,
      "step": 1015
    },
    {
      "epoch": 0.13331146465474825,
      "grad_norm": 74.20399475097656,
      "learning_rate": 4.770520702957182e-05,
      "loss": 1.1262,
      "step": 1016
    },
    {
      "epoch": 0.1334426767262588,
      "grad_norm": 77.79273986816406,
      "learning_rate": 4.752560645904962e-05,
      "loss": 1.138,
      "step": 1017
    },
    {
      "epoch": 0.1335738887977694,
      "grad_norm": 80.13951873779297,
      "learning_rate": 4.734623916689941e-05,
      "loss": 1.1467,
      "step": 1018
    },
    {
      "epoch": 0.13370510086927997,
      "grad_norm": 81.62757110595703,
      "learning_rate": 4.716710595051022e-05,
      "loss": 1.159,
      "step": 1019
    },
    {
      "epoch": 0.13383631294079054,
      "grad_norm": 66.25305938720703,
      "learning_rate": 4.698820760623064e-05,
      "loss": 0.8899,
      "step": 1020
    },
    {
      "epoch": 0.13396752501230114,
      "grad_norm": 88.61091613769531,
      "learning_rate": 4.6809544929365004e-05,
      "loss": 1.3789,
      "step": 1021
    },
    {
      "epoch": 0.1340987370838117,
      "grad_norm": 68.76612854003906,
      "learning_rate": 4.663111871417e-05,
      "loss": 0.8624,
      "step": 1022
    },
    {
      "epoch": 0.1342299491553223,
      "grad_norm": 83.45545196533203,
      "learning_rate": 4.645292975385111e-05,
      "loss": 1.1052,
      "step": 1023
    },
    {
      "epoch": 0.13436116122683286,
      "grad_norm": 83.41942596435547,
      "learning_rate": 4.627497884055912e-05,
      "loss": 1.2137,
      "step": 1024
    },
    {
      "epoch": 0.13449237329834346,
      "grad_norm": 87.13421630859375,
      "learning_rate": 4.609726676538652e-05,
      "loss": 1.3902,
      "step": 1025
    },
    {
      "epoch": 0.13462358536985403,
      "grad_norm": 72.57111358642578,
      "learning_rate": 4.591979431836402e-05,
      "loss": 1.0551,
      "step": 1026
    },
    {
      "epoch": 0.1347547974413646,
      "grad_norm": 77.98977661132812,
      "learning_rate": 4.574256228845706e-05,
      "loss": 1.2052,
      "step": 1027
    },
    {
      "epoch": 0.1348860095128752,
      "grad_norm": 89.48870086669922,
      "learning_rate": 4.5565571463562365e-05,
      "loss": 1.3083,
      "step": 1028
    },
    {
      "epoch": 0.13501722158438575,
      "grad_norm": 76.56085205078125,
      "learning_rate": 4.5388822630504256e-05,
      "loss": 1.1178,
      "step": 1029
    },
    {
      "epoch": 0.13514843365589635,
      "grad_norm": 72.62358093261719,
      "learning_rate": 4.521231657503132e-05,
      "loss": 0.9959,
      "step": 1030
    },
    {
      "epoch": 0.13527964572740692,
      "grad_norm": 74.80716705322266,
      "learning_rate": 4.503605408181286e-05,
      "loss": 1.1967,
      "step": 1031
    },
    {
      "epoch": 0.1354108577989175,
      "grad_norm": 70.13179016113281,
      "learning_rate": 4.486003593443537e-05,
      "loss": 1.0661,
      "step": 1032
    },
    {
      "epoch": 0.13554206987042808,
      "grad_norm": 83.18326568603516,
      "learning_rate": 4.468426291539914e-05,
      "loss": 1.1548,
      "step": 1033
    },
    {
      "epoch": 0.13567328194193864,
      "grad_norm": 71.34968566894531,
      "learning_rate": 4.4508735806114654e-05,
      "loss": 1.0385,
      "step": 1034
    },
    {
      "epoch": 0.13580449401344924,
      "grad_norm": 83.74834442138672,
      "learning_rate": 4.433345538689929e-05,
      "loss": 1.2305,
      "step": 1035
    },
    {
      "epoch": 0.1359357060849598,
      "grad_norm": 78.68016815185547,
      "learning_rate": 4.415842243697369e-05,
      "loss": 1.1928,
      "step": 1036
    },
    {
      "epoch": 0.1360669181564704,
      "grad_norm": 72.68682098388672,
      "learning_rate": 4.39836377344583e-05,
      "loss": 0.891,
      "step": 1037
    },
    {
      "epoch": 0.13619813022798097,
      "grad_norm": 81.73300170898438,
      "learning_rate": 4.380910205637012e-05,
      "loss": 1.1436,
      "step": 1038
    },
    {
      "epoch": 0.13632934229949156,
      "grad_norm": 84.24224090576172,
      "learning_rate": 4.363481617861893e-05,
      "loss": 1.1992,
      "step": 1039
    },
    {
      "epoch": 0.13646055437100213,
      "grad_norm": 76.39823150634766,
      "learning_rate": 4.346078087600412e-05,
      "loss": 1.1263,
      "step": 1040
    },
    {
      "epoch": 0.13659176644251272,
      "grad_norm": 81.22613525390625,
      "learning_rate": 4.3286996922211034e-05,
      "loss": 1.2476,
      "step": 1041
    },
    {
      "epoch": 0.1367229785140233,
      "grad_norm": 84.17645263671875,
      "learning_rate": 4.311346508980772e-05,
      "loss": 1.1626,
      "step": 1042
    },
    {
      "epoch": 0.13685419058553386,
      "grad_norm": 77.53297424316406,
      "learning_rate": 4.2940186150241365e-05,
      "loss": 1.1552,
      "step": 1043
    },
    {
      "epoch": 0.13698540265704445,
      "grad_norm": 78.96730041503906,
      "learning_rate": 4.27671608738349e-05,
      "loss": 1.1313,
      "step": 1044
    },
    {
      "epoch": 0.13711661472855502,
      "grad_norm": 74.94930267333984,
      "learning_rate": 4.2594390029783534e-05,
      "loss": 1.0609,
      "step": 1045
    },
    {
      "epoch": 0.13724782680006561,
      "grad_norm": 77.02103424072266,
      "learning_rate": 4.242187438615153e-05,
      "loss": 1.0584,
      "step": 1046
    },
    {
      "epoch": 0.13737903887157618,
      "grad_norm": 82.13597869873047,
      "learning_rate": 4.224961470986849e-05,
      "loss": 1.2085,
      "step": 1047
    },
    {
      "epoch": 0.13751025094308678,
      "grad_norm": 78.32071685791016,
      "learning_rate": 4.207761176672614e-05,
      "loss": 1.0407,
      "step": 1048
    },
    {
      "epoch": 0.13764146301459734,
      "grad_norm": 91.04803466796875,
      "learning_rate": 4.190586632137491e-05,
      "loss": 1.1072,
      "step": 1049
    },
    {
      "epoch": 0.1377726750861079,
      "grad_norm": 95.67418670654297,
      "learning_rate": 4.173437913732048e-05,
      "loss": 1.2645,
      "step": 1050
    },
    {
      "epoch": 0.1379038871576185,
      "grad_norm": 82.18598175048828,
      "learning_rate": 4.156315097692037e-05,
      "loss": 1.1896,
      "step": 1051
    },
    {
      "epoch": 0.13803509922912907,
      "grad_norm": 69.84553527832031,
      "learning_rate": 4.139218260138074e-05,
      "loss": 0.9984,
      "step": 1052
    },
    {
      "epoch": 0.13816631130063967,
      "grad_norm": 77.92560577392578,
      "learning_rate": 4.12214747707527e-05,
      "loss": 1.2164,
      "step": 1053
    },
    {
      "epoch": 0.13829752337215023,
      "grad_norm": 78.45890808105469,
      "learning_rate": 4.1051028243929125e-05,
      "loss": 1.3791,
      "step": 1054
    },
    {
      "epoch": 0.13842873544366083,
      "grad_norm": 84.03868103027344,
      "learning_rate": 4.088084377864135e-05,
      "loss": 1.0555,
      "step": 1055
    },
    {
      "epoch": 0.1385599475151714,
      "grad_norm": 80.94784545898438,
      "learning_rate": 4.07109221314556e-05,
      "loss": 1.34,
      "step": 1056
    },
    {
      "epoch": 0.13869115958668196,
      "grad_norm": 81.8576431274414,
      "learning_rate": 4.054126405776971e-05,
      "loss": 1.1598,
      "step": 1057
    },
    {
      "epoch": 0.13882237165819256,
      "grad_norm": 76.01251220703125,
      "learning_rate": 4.037187031180985e-05,
      "loss": 1.118,
      "step": 1058
    },
    {
      "epoch": 0.13895358372970312,
      "grad_norm": 71.07621765136719,
      "learning_rate": 4.020274164662707e-05,
      "loss": 1.1594,
      "step": 1059
    },
    {
      "epoch": 0.13908479580121372,
      "grad_norm": 80.01371002197266,
      "learning_rate": 4.003387881409397e-05,
      "loss": 1.0442,
      "step": 1060
    },
    {
      "epoch": 0.13921600787272428,
      "grad_norm": 76.29126739501953,
      "learning_rate": 3.986528256490141e-05,
      "loss": 1.2139,
      "step": 1061
    },
    {
      "epoch": 0.13934721994423488,
      "grad_norm": 78.54533386230469,
      "learning_rate": 3.969695364855511e-05,
      "loss": 1.1682,
      "step": 1062
    },
    {
      "epoch": 0.13947843201574545,
      "grad_norm": 84.36077117919922,
      "learning_rate": 3.952889281337235e-05,
      "loss": 1.2345,
      "step": 1063
    },
    {
      "epoch": 0.13960964408725604,
      "grad_norm": 75.85221862792969,
      "learning_rate": 3.93611008064786e-05,
      "loss": 1.0906,
      "step": 1064
    },
    {
      "epoch": 0.1397408561587666,
      "grad_norm": 76.5043716430664,
      "learning_rate": 3.9193578373804364e-05,
      "loss": 1.0302,
      "step": 1065
    },
    {
      "epoch": 0.13987206823027717,
      "grad_norm": 72.82099151611328,
      "learning_rate": 3.90263262600816e-05,
      "loss": 0.9824,
      "step": 1066
    },
    {
      "epoch": 0.14000328030178777,
      "grad_norm": 73.16133880615234,
      "learning_rate": 3.88593452088406e-05,
      "loss": 1.0664,
      "step": 1067
    },
    {
      "epoch": 0.14013449237329834,
      "grad_norm": 69.83660125732422,
      "learning_rate": 3.869263596240661e-05,
      "loss": 1.1104,
      "step": 1068
    },
    {
      "epoch": 0.14026570444480893,
      "grad_norm": 59.826416015625,
      "learning_rate": 3.8526199261896544e-05,
      "loss": 0.7265,
      "step": 1069
    },
    {
      "epoch": 0.1403969165163195,
      "grad_norm": 75.51627349853516,
      "learning_rate": 3.836003584721577e-05,
      "loss": 1.1188,
      "step": 1070
    },
    {
      "epoch": 0.1405281285878301,
      "grad_norm": 76.36466979980469,
      "learning_rate": 3.8194146457054655e-05,
      "loss": 1.0615,
      "step": 1071
    },
    {
      "epoch": 0.14065934065934066,
      "grad_norm": 72.96395111083984,
      "learning_rate": 3.802853182888543e-05,
      "loss": 1.1025,
      "step": 1072
    },
    {
      "epoch": 0.14079055273085123,
      "grad_norm": 89.4021224975586,
      "learning_rate": 3.786319269895877e-05,
      "loss": 1.2447,
      "step": 1073
    },
    {
      "epoch": 0.14092176480236182,
      "grad_norm": 90.27672576904297,
      "learning_rate": 3.769812980230074e-05,
      "loss": 1.0668,
      "step": 1074
    },
    {
      "epoch": 0.1410529768738724,
      "grad_norm": 77.8091812133789,
      "learning_rate": 3.7533343872709294e-05,
      "loss": 1.1986,
      "step": 1075
    },
    {
      "epoch": 0.14118418894538298,
      "grad_norm": 71.89775085449219,
      "learning_rate": 3.736883564275112e-05,
      "loss": 1.0691,
      "step": 1076
    },
    {
      "epoch": 0.14131540101689355,
      "grad_norm": 77.78672790527344,
      "learning_rate": 3.7204605843758386e-05,
      "loss": 1.1897,
      "step": 1077
    },
    {
      "epoch": 0.14144661308840414,
      "grad_norm": 80.20154571533203,
      "learning_rate": 3.704065520582549e-05,
      "loss": 1.271,
      "step": 1078
    },
    {
      "epoch": 0.1415778251599147,
      "grad_norm": 63.987327575683594,
      "learning_rate": 3.6876984457805786e-05,
      "loss": 0.7241,
      "step": 1079
    },
    {
      "epoch": 0.14170903723142528,
      "grad_norm": 79.87057495117188,
      "learning_rate": 3.671359432730834e-05,
      "loss": 1.2956,
      "step": 1080
    },
    {
      "epoch": 0.14184024930293587,
      "grad_norm": 80.74537658691406,
      "learning_rate": 3.655048554069478e-05,
      "loss": 1.2117,
      "step": 1081
    },
    {
      "epoch": 0.14197146137444644,
      "grad_norm": 81.21459197998047,
      "learning_rate": 3.638765882307589e-05,
      "loss": 1.2179,
      "step": 1082
    },
    {
      "epoch": 0.14210267344595703,
      "grad_norm": 92.52425384521484,
      "learning_rate": 3.6225114898308634e-05,
      "loss": 1.3275,
      "step": 1083
    },
    {
      "epoch": 0.1422338855174676,
      "grad_norm": 74.58345031738281,
      "learning_rate": 3.6062854488992714e-05,
      "loss": 0.9602,
      "step": 1084
    },
    {
      "epoch": 0.1423650975889782,
      "grad_norm": 73.6881332397461,
      "learning_rate": 3.5900878316467454e-05,
      "loss": 1.0582,
      "step": 1085
    },
    {
      "epoch": 0.14249630966048876,
      "grad_norm": 83.31182098388672,
      "learning_rate": 3.573918710080857e-05,
      "loss": 1.3372,
      "step": 1086
    },
    {
      "epoch": 0.14262752173199936,
      "grad_norm": 77.0564956665039,
      "learning_rate": 3.5577781560825066e-05,
      "loss": 1.0955,
      "step": 1087
    },
    {
      "epoch": 0.14275873380350992,
      "grad_norm": 79.33744049072266,
      "learning_rate": 3.541666241405588e-05,
      "loss": 1.277,
      "step": 1088
    },
    {
      "epoch": 0.1428899458750205,
      "grad_norm": 70.41979217529297,
      "learning_rate": 3.5255830376766764e-05,
      "loss": 0.9681,
      "step": 1089
    },
    {
      "epoch": 0.14302115794653109,
      "grad_norm": 107.60590362548828,
      "learning_rate": 3.509528616394716e-05,
      "loss": 1.1465,
      "step": 1090
    },
    {
      "epoch": 0.14315237001804165,
      "grad_norm": 81.1191177368164,
      "learning_rate": 3.4935030489306883e-05,
      "loss": 1.1125,
      "step": 1091
    },
    {
      "epoch": 0.14328358208955225,
      "grad_norm": 74.24160766601562,
      "learning_rate": 3.4775064065273165e-05,
      "loss": 1.2363,
      "step": 1092
    },
    {
      "epoch": 0.14341479416106281,
      "grad_norm": 93.1820297241211,
      "learning_rate": 3.4615387602987236e-05,
      "loss": 1.4472,
      "step": 1093
    },
    {
      "epoch": 0.1435460062325734,
      "grad_norm": 71.58454895019531,
      "learning_rate": 3.445600181230134e-05,
      "loss": 1.1634,
      "step": 1094
    },
    {
      "epoch": 0.14367721830408398,
      "grad_norm": 82.26412963867188,
      "learning_rate": 3.429690740177549e-05,
      "loss": 1.1711,
      "step": 1095
    },
    {
      "epoch": 0.14380843037559454,
      "grad_norm": 77.81709289550781,
      "learning_rate": 3.413810507867436e-05,
      "loss": 1.1686,
      "step": 1096
    },
    {
      "epoch": 0.14393964244710514,
      "grad_norm": 90.8698959350586,
      "learning_rate": 3.397959554896415e-05,
      "loss": 1.2168,
      "step": 1097
    },
    {
      "epoch": 0.1440708545186157,
      "grad_norm": 73.94993591308594,
      "learning_rate": 3.3821379517309405e-05,
      "loss": 1.1268,
      "step": 1098
    },
    {
      "epoch": 0.1442020665901263,
      "grad_norm": 100.52758026123047,
      "learning_rate": 3.3663457687069924e-05,
      "loss": 1.2303,
      "step": 1099
    },
    {
      "epoch": 0.14433327866163687,
      "grad_norm": 102.02374267578125,
      "learning_rate": 3.350583076029754e-05,
      "loss": 1.2834,
      "step": 1100
    },
    {
      "epoch": 0.14446449073314746,
      "grad_norm": 71.72270965576172,
      "learning_rate": 3.334849943773323e-05,
      "loss": 1.0353,
      "step": 1101
    },
    {
      "epoch": 0.14459570280465803,
      "grad_norm": 67.39696502685547,
      "learning_rate": 3.319146441880371e-05,
      "loss": 1.0956,
      "step": 1102
    },
    {
      "epoch": 0.1447269148761686,
      "grad_norm": 72.23811340332031,
      "learning_rate": 3.3034726401618444e-05,
      "loss": 1.1044,
      "step": 1103
    },
    {
      "epoch": 0.1448581269476792,
      "grad_norm": 88.7195816040039,
      "learning_rate": 3.28782860829667e-05,
      "loss": 1.0921,
      "step": 1104
    },
    {
      "epoch": 0.14498933901918976,
      "grad_norm": 76.1098861694336,
      "learning_rate": 3.272214415831418e-05,
      "loss": 1.3162,
      "step": 1105
    },
    {
      "epoch": 0.14512055109070035,
      "grad_norm": 81.88814544677734,
      "learning_rate": 3.2566301321800085e-05,
      "loss": 1.2916,
      "step": 1106
    },
    {
      "epoch": 0.14525176316221092,
      "grad_norm": 74.16931915283203,
      "learning_rate": 3.241075826623401e-05,
      "loss": 1.2259,
      "step": 1107
    },
    {
      "epoch": 0.1453829752337215,
      "grad_norm": 70.73887634277344,
      "learning_rate": 3.225551568309284e-05,
      "loss": 1.0428,
      "step": 1108
    },
    {
      "epoch": 0.14551418730523208,
      "grad_norm": 79.63719940185547,
      "learning_rate": 3.210057426251773e-05,
      "loss": 1.0832,
      "step": 1109
    },
    {
      "epoch": 0.14564539937674267,
      "grad_norm": 80.26901245117188,
      "learning_rate": 3.1945934693310896e-05,
      "loss": 1.2107,
      "step": 1110
    },
    {
      "epoch": 0.14577661144825324,
      "grad_norm": 73.19075012207031,
      "learning_rate": 3.179159766293282e-05,
      "loss": 1.2271,
      "step": 1111
    },
    {
      "epoch": 0.1459078235197638,
      "grad_norm": 77.24995422363281,
      "learning_rate": 3.163756385749889e-05,
      "loss": 1.0926,
      "step": 1112
    },
    {
      "epoch": 0.1460390355912744,
      "grad_norm": 78.86018371582031,
      "learning_rate": 3.148383396177653e-05,
      "loss": 1.0718,
      "step": 1113
    },
    {
      "epoch": 0.14617024766278497,
      "grad_norm": 74.18000793457031,
      "learning_rate": 3.133040865918213e-05,
      "loss": 1.1134,
      "step": 1114
    },
    {
      "epoch": 0.14630145973429556,
      "grad_norm": 71.132080078125,
      "learning_rate": 3.117728863177796e-05,
      "loss": 1.1511,
      "step": 1115
    },
    {
      "epoch": 0.14643267180580613,
      "grad_norm": 73.233642578125,
      "learning_rate": 3.102447456026919e-05,
      "loss": 1.0333,
      "step": 1116
    },
    {
      "epoch": 0.14656388387731673,
      "grad_norm": 68.40826416015625,
      "learning_rate": 3.0871967124000834e-05,
      "loss": 1.0732,
      "step": 1117
    },
    {
      "epoch": 0.1466950959488273,
      "grad_norm": 73.65232849121094,
      "learning_rate": 3.0719767000954714e-05,
      "loss": 1.1053,
      "step": 1118
    },
    {
      "epoch": 0.14682630802033786,
      "grad_norm": 100.50982666015625,
      "learning_rate": 3.056787486774656e-05,
      "loss": 1.3524,
      "step": 1119
    },
    {
      "epoch": 0.14695752009184845,
      "grad_norm": 64.12255096435547,
      "learning_rate": 3.041629139962283e-05,
      "loss": 0.8443,
      "step": 1120
    },
    {
      "epoch": 0.14708873216335902,
      "grad_norm": 79.35578155517578,
      "learning_rate": 3.0265017270457775e-05,
      "loss": 1.2219,
      "step": 1121
    },
    {
      "epoch": 0.14721994423486962,
      "grad_norm": 73.58233642578125,
      "learning_rate": 3.0114053152750556e-05,
      "loss": 1.0997,
      "step": 1122
    },
    {
      "epoch": 0.14735115630638018,
      "grad_norm": 81.35496520996094,
      "learning_rate": 2.9963399717622077e-05,
      "loss": 1.1356,
      "step": 1123
    },
    {
      "epoch": 0.14748236837789078,
      "grad_norm": 75.18378448486328,
      "learning_rate": 2.98130576348121e-05,
      "loss": 1.0517,
      "step": 1124
    },
    {
      "epoch": 0.14761358044940134,
      "grad_norm": 77.94686126708984,
      "learning_rate": 2.966302757267625e-05,
      "loss": 0.9647,
      "step": 1125
    },
    {
      "epoch": 0.14761358044940134,
      "eval_loss": 1.1374354362487793,
      "eval_runtime": 760.6543,
      "eval_samples_per_second": 16.875,
      "eval_steps_per_second": 8.437,
      "step": 1125
    },
    {
      "epoch": 0.1477447925209119,
      "grad_norm": 80.55128479003906,
      "learning_rate": 2.9513310198183065e-05,
      "loss": 1.2022,
      "step": 1126
    },
    {
      "epoch": 0.1478760045924225,
      "grad_norm": 77.05827331542969,
      "learning_rate": 2.936390617691097e-05,
      "loss": 1.277,
      "step": 1127
    },
    {
      "epoch": 0.14800721666393307,
      "grad_norm": 71.95764923095703,
      "learning_rate": 2.9214816173045356e-05,
      "loss": 1.1832,
      "step": 1128
    },
    {
      "epoch": 0.14813842873544367,
      "grad_norm": 67.0936050415039,
      "learning_rate": 2.906604084937572e-05,
      "loss": 0.9739,
      "step": 1129
    },
    {
      "epoch": 0.14826964080695423,
      "grad_norm": 62.5437126159668,
      "learning_rate": 2.8917580867292526e-05,
      "loss": 0.8689,
      "step": 1130
    },
    {
      "epoch": 0.14840085287846483,
      "grad_norm": 73.13948822021484,
      "learning_rate": 2.8769436886784408e-05,
      "loss": 1.1177,
      "step": 1131
    },
    {
      "epoch": 0.1485320649499754,
      "grad_norm": 83.93120574951172,
      "learning_rate": 2.862160956643517e-05,
      "loss": 1.3056,
      "step": 1132
    },
    {
      "epoch": 0.148663277021486,
      "grad_norm": 74.29683685302734,
      "learning_rate": 2.847409956342092e-05,
      "loss": 1.0937,
      "step": 1133
    },
    {
      "epoch": 0.14879448909299656,
      "grad_norm": 77.46924591064453,
      "learning_rate": 2.8326907533507074e-05,
      "loss": 1.0821,
      "step": 1134
    },
    {
      "epoch": 0.14892570116450712,
      "grad_norm": 73.16531372070312,
      "learning_rate": 2.8180034131045464e-05,
      "loss": 1.0129,
      "step": 1135
    },
    {
      "epoch": 0.14905691323601772,
      "grad_norm": 79.64446258544922,
      "learning_rate": 2.8033480008971546e-05,
      "loss": 1.1054,
      "step": 1136
    },
    {
      "epoch": 0.14918812530752829,
      "grad_norm": 75.18922424316406,
      "learning_rate": 2.7887245818801277e-05,
      "loss": 1.101,
      "step": 1137
    },
    {
      "epoch": 0.14931933737903888,
      "grad_norm": 88.12415313720703,
      "learning_rate": 2.7741332210628345e-05,
      "loss": 1.1752,
      "step": 1138
    },
    {
      "epoch": 0.14945054945054945,
      "grad_norm": 69.41979217529297,
      "learning_rate": 2.759573983312138e-05,
      "loss": 0.9464,
      "step": 1139
    },
    {
      "epoch": 0.14958176152206004,
      "grad_norm": 80.63298797607422,
      "learning_rate": 2.7450469333520855e-05,
      "loss": 1.0859,
      "step": 1140
    },
    {
      "epoch": 0.1497129735935706,
      "grad_norm": 83.3411636352539,
      "learning_rate": 2.730552135763632e-05,
      "loss": 1.2127,
      "step": 1141
    },
    {
      "epoch": 0.14984418566508118,
      "grad_norm": 78.10100555419922,
      "learning_rate": 2.7160896549843562e-05,
      "loss": 1.1401,
      "step": 1142
    },
    {
      "epoch": 0.14997539773659177,
      "grad_norm": 77.15515899658203,
      "learning_rate": 2.701659555308169e-05,
      "loss": 1.0748,
      "step": 1143
    },
    {
      "epoch": 0.15010660980810234,
      "grad_norm": 83.33522033691406,
      "learning_rate": 2.6872619008850274e-05,
      "loss": 1.4414,
      "step": 1144
    },
    {
      "epoch": 0.15023782187961293,
      "grad_norm": 88.73904418945312,
      "learning_rate": 2.672896755720654e-05,
      "loss": 1.4055,
      "step": 1145
    },
    {
      "epoch": 0.1503690339511235,
      "grad_norm": 81.29682159423828,
      "learning_rate": 2.6585641836762433e-05,
      "loss": 1.1386,
      "step": 1146
    },
    {
      "epoch": 0.1505002460226341,
      "grad_norm": 69.88158416748047,
      "learning_rate": 2.6442642484681944e-05,
      "loss": 1.0921,
      "step": 1147
    },
    {
      "epoch": 0.15063145809414466,
      "grad_norm": 83.93527221679688,
      "learning_rate": 2.6299970136678077e-05,
      "loss": 1.2659,
      "step": 1148
    },
    {
      "epoch": 0.15076267016565523,
      "grad_norm": 79.46372985839844,
      "learning_rate": 2.6157625427010156e-05,
      "loss": 0.9937,
      "step": 1149
    },
    {
      "epoch": 0.15089388223716582,
      "grad_norm": 63.39399337768555,
      "learning_rate": 2.6015608988480955e-05,
      "loss": 0.5788,
      "step": 1150
    },
    {
      "epoch": 0.1510250943086764,
      "grad_norm": 71.88567352294922,
      "learning_rate": 2.5873921452433915e-05,
      "loss": 1.0378,
      "step": 1151
    },
    {
      "epoch": 0.15115630638018698,
      "grad_norm": 76.33042907714844,
      "learning_rate": 2.57325634487503e-05,
      "loss": 1.1971,
      "step": 1152
    },
    {
      "epoch": 0.15128751845169755,
      "grad_norm": 80.09848022460938,
      "learning_rate": 2.5591535605846383e-05,
      "loss": 0.8381,
      "step": 1153
    },
    {
      "epoch": 0.15141873052320814,
      "grad_norm": 81.7151107788086,
      "learning_rate": 2.5450838550670808e-05,
      "loss": 1.1233,
      "step": 1154
    },
    {
      "epoch": 0.1515499425947187,
      "grad_norm": 71.63130187988281,
      "learning_rate": 2.5310472908701555e-05,
      "loss": 1.2958,
      "step": 1155
    },
    {
      "epoch": 0.1516811546662293,
      "grad_norm": 74.99315643310547,
      "learning_rate": 2.5170439303943294e-05,
      "loss": 1.1438,
      "step": 1156
    },
    {
      "epoch": 0.15181236673773987,
      "grad_norm": 73.38867950439453,
      "learning_rate": 2.503073835892471e-05,
      "loss": 0.9832,
      "step": 1157
    },
    {
      "epoch": 0.15194357880925044,
      "grad_norm": 82.92842864990234,
      "learning_rate": 2.4891370694695517e-05,
      "loss": 1.3461,
      "step": 1158
    },
    {
      "epoch": 0.15207479088076103,
      "grad_norm": 66.21943664550781,
      "learning_rate": 2.4752336930823837e-05,
      "loss": 0.911,
      "step": 1159
    },
    {
      "epoch": 0.1522060029522716,
      "grad_norm": 79.16444396972656,
      "learning_rate": 2.4613637685393432e-05,
      "loss": 1.2775,
      "step": 1160
    },
    {
      "epoch": 0.1523372150237822,
      "grad_norm": 73.5926284790039,
      "learning_rate": 2.4475273575000936e-05,
      "loss": 1.1897,
      "step": 1161
    },
    {
      "epoch": 0.15246842709529276,
      "grad_norm": 82.16194915771484,
      "learning_rate": 2.4337245214753103e-05,
      "loss": 1.1723,
      "step": 1162
    },
    {
      "epoch": 0.15259963916680336,
      "grad_norm": 74.63170623779297,
      "learning_rate": 2.4199553218264093e-05,
      "loss": 1.255,
      "step": 1163
    },
    {
      "epoch": 0.15273085123831392,
      "grad_norm": 72.17273712158203,
      "learning_rate": 2.4062198197652752e-05,
      "loss": 1.0489,
      "step": 1164
    },
    {
      "epoch": 0.1528620633098245,
      "grad_norm": 82.76921844482422,
      "learning_rate": 2.3925180763539844e-05,
      "loss": 1.2001,
      "step": 1165
    },
    {
      "epoch": 0.1529932753813351,
      "grad_norm": 71.44467163085938,
      "learning_rate": 2.3788501525045438e-05,
      "loss": 0.994,
      "step": 1166
    },
    {
      "epoch": 0.15312448745284565,
      "grad_norm": 80.5171127319336,
      "learning_rate": 2.3652161089786086e-05,
      "loss": 1.1954,
      "step": 1167
    },
    {
      "epoch": 0.15325569952435625,
      "grad_norm": 77.30850219726562,
      "learning_rate": 2.351616006387214e-05,
      "loss": 1.197,
      "step": 1168
    },
    {
      "epoch": 0.15338691159586682,
      "grad_norm": 74.52938079833984,
      "learning_rate": 2.3380499051905137e-05,
      "loss": 1.0782,
      "step": 1169
    },
    {
      "epoch": 0.1535181236673774,
      "grad_norm": 111.87449645996094,
      "learning_rate": 2.324517865697501e-05,
      "loss": 1.1419,
      "step": 1170
    },
    {
      "epoch": 0.15364933573888798,
      "grad_norm": 79.39338684082031,
      "learning_rate": 2.3110199480657525e-05,
      "loss": 1.3398,
      "step": 1171
    },
    {
      "epoch": 0.15378054781039854,
      "grad_norm": 76.5979995727539,
      "learning_rate": 2.2975562123011495e-05,
      "loss": 1.1692,
      "step": 1172
    },
    {
      "epoch": 0.15391175988190914,
      "grad_norm": 67.45117950439453,
      "learning_rate": 2.2841267182576143e-05,
      "loss": 1.0275,
      "step": 1173
    },
    {
      "epoch": 0.1540429719534197,
      "grad_norm": 82.69200897216797,
      "learning_rate": 2.2707315256368433e-05,
      "loss": 1.3477,
      "step": 1174
    },
    {
      "epoch": 0.1541741840249303,
      "grad_norm": 93.0841064453125,
      "learning_rate": 2.2573706939880555e-05,
      "loss": 1.2809,
      "step": 1175
    },
    {
      "epoch": 0.15430539609644087,
      "grad_norm": 82.80020141601562,
      "learning_rate": 2.2440442827077045e-05,
      "loss": 1.2777,
      "step": 1176
    },
    {
      "epoch": 0.15443660816795146,
      "grad_norm": 70.98883056640625,
      "learning_rate": 2.230752351039228e-05,
      "loss": 1.0981,
      "step": 1177
    },
    {
      "epoch": 0.15456782023946203,
      "grad_norm": 81.68639373779297,
      "learning_rate": 2.2174949580727832e-05,
      "loss": 1.1855,
      "step": 1178
    },
    {
      "epoch": 0.15469903231097262,
      "grad_norm": 70.6236343383789,
      "learning_rate": 2.2042721627449846e-05,
      "loss": 0.8975,
      "step": 1179
    },
    {
      "epoch": 0.1548302443824832,
      "grad_norm": 69.52759552001953,
      "learning_rate": 2.1910840238386398e-05,
      "loss": 0.9286,
      "step": 1180
    },
    {
      "epoch": 0.15496145645399376,
      "grad_norm": 71.5075454711914,
      "learning_rate": 2.1779305999824884e-05,
      "loss": 1.0774,
      "step": 1181
    },
    {
      "epoch": 0.15509266852550435,
      "grad_norm": 82.03826141357422,
      "learning_rate": 2.164811949650942e-05,
      "loss": 1.0548,
      "step": 1182
    },
    {
      "epoch": 0.15522388059701492,
      "grad_norm": 71.9784164428711,
      "learning_rate": 2.1517281311638217e-05,
      "loss": 1.0132,
      "step": 1183
    },
    {
      "epoch": 0.1553550926685255,
      "grad_norm": 73.0723876953125,
      "learning_rate": 2.1386792026861103e-05,
      "loss": 1.1327,
      "step": 1184
    },
    {
      "epoch": 0.15548630474003608,
      "grad_norm": 82.465576171875,
      "learning_rate": 2.125665222227675e-05,
      "loss": 1.1278,
      "step": 1185
    },
    {
      "epoch": 0.15561751681154667,
      "grad_norm": 80.635009765625,
      "learning_rate": 2.112686247643024e-05,
      "loss": 1.1343,
      "step": 1186
    },
    {
      "epoch": 0.15574872888305724,
      "grad_norm": 82.60955810546875,
      "learning_rate": 2.09974233663104e-05,
      "loss": 1.3297,
      "step": 1187
    },
    {
      "epoch": 0.1558799409545678,
      "grad_norm": 87.00865936279297,
      "learning_rate": 2.0868335467347366e-05,
      "loss": 1.2736,
      "step": 1188
    },
    {
      "epoch": 0.1560111530260784,
      "grad_norm": 79.19508361816406,
      "learning_rate": 2.073959935340988e-05,
      "loss": 1.1201,
      "step": 1189
    },
    {
      "epoch": 0.15614236509758897,
      "grad_norm": 74.8446273803711,
      "learning_rate": 2.06112155968028e-05,
      "loss": 1.0599,
      "step": 1190
    },
    {
      "epoch": 0.15627357716909956,
      "grad_norm": 88.28968048095703,
      "learning_rate": 2.0483184768264596e-05,
      "loss": 1.2803,
      "step": 1191
    },
    {
      "epoch": 0.15640478924061013,
      "grad_norm": 80.58595275878906,
      "learning_rate": 2.035550743696468e-05,
      "loss": 1.2338,
      "step": 1192
    },
    {
      "epoch": 0.15653600131212073,
      "grad_norm": 90.64918518066406,
      "learning_rate": 2.022818417050113e-05,
      "loss": 1.3544,
      "step": 1193
    },
    {
      "epoch": 0.1566672133836313,
      "grad_norm": 79.92119598388672,
      "learning_rate": 2.0101215534897855e-05,
      "loss": 1.3407,
      "step": 1194
    },
    {
      "epoch": 0.15679842545514186,
      "grad_norm": 73.43045806884766,
      "learning_rate": 1.99746020946023e-05,
      "loss": 1.0047,
      "step": 1195
    },
    {
      "epoch": 0.15692963752665245,
      "grad_norm": 74.78981018066406,
      "learning_rate": 1.9848344412482854e-05,
      "loss": 1.0394,
      "step": 1196
    },
    {
      "epoch": 0.15706084959816302,
      "grad_norm": 82.09642028808594,
      "learning_rate": 1.9722443049826344e-05,
      "loss": 1.1259,
      "step": 1197
    },
    {
      "epoch": 0.15719206166967362,
      "grad_norm": 79.5477523803711,
      "learning_rate": 1.9596898566335576e-05,
      "loss": 0.9226,
      "step": 1198
    },
    {
      "epoch": 0.15732327374118418,
      "grad_norm": 83.33988189697266,
      "learning_rate": 1.9471711520126824e-05,
      "loss": 1.2415,
      "step": 1199
    },
    {
      "epoch": 0.15745448581269478,
      "grad_norm": 71.45040130615234,
      "learning_rate": 1.9346882467727325e-05,
      "loss": 0.7793,
      "step": 1200
    },
    {
      "epoch": 0.15758569788420534,
      "grad_norm": 56.93532943725586,
      "learning_rate": 1.9222411964072884e-05,
      "loss": 0.8564,
      "step": 1201
    },
    {
      "epoch": 0.15771690995571594,
      "grad_norm": 85.16472625732422,
      "learning_rate": 1.9098300562505266e-05,
      "loss": 1.051,
      "step": 1202
    },
    {
      "epoch": 0.1578481220272265,
      "grad_norm": 61.01642608642578,
      "learning_rate": 1.8974548814769944e-05,
      "loss": 0.8961,
      "step": 1203
    },
    {
      "epoch": 0.15797933409873707,
      "grad_norm": 80.30467224121094,
      "learning_rate": 1.8851157271013442e-05,
      "loss": 1.2335,
      "step": 1204
    },
    {
      "epoch": 0.15811054617024767,
      "grad_norm": 74.60105895996094,
      "learning_rate": 1.872812647978095e-05,
      "loss": 1.3467,
      "step": 1205
    },
    {
      "epoch": 0.15824175824175823,
      "grad_norm": 75.2890396118164,
      "learning_rate": 1.8605456988014015e-05,
      "loss": 1.0769,
      "step": 1206
    },
    {
      "epoch": 0.15837297031326883,
      "grad_norm": 71.32281494140625,
      "learning_rate": 1.8483149341047923e-05,
      "loss": 1.1516,
      "step": 1207
    },
    {
      "epoch": 0.1585041823847794,
      "grad_norm": 79.65695190429688,
      "learning_rate": 1.8361204082609352e-05,
      "loss": 1.3148,
      "step": 1208
    },
    {
      "epoch": 0.15863539445629,
      "grad_norm": 70.46035766601562,
      "learning_rate": 1.8239621754813995e-05,
      "loss": 1.0424,
      "step": 1209
    },
    {
      "epoch": 0.15876660652780056,
      "grad_norm": 70.63594818115234,
      "learning_rate": 1.811840289816409e-05,
      "loss": 0.968,
      "step": 1210
    },
    {
      "epoch": 0.15889781859931112,
      "grad_norm": 74.76544189453125,
      "learning_rate": 1.799754805154603e-05,
      "loss": 1.1901,
      "step": 1211
    },
    {
      "epoch": 0.15902903067082172,
      "grad_norm": 72.52140045166016,
      "learning_rate": 1.787705775222802e-05,
      "loss": 1.0694,
      "step": 1212
    },
    {
      "epoch": 0.1591602427423323,
      "grad_norm": 74.32793426513672,
      "learning_rate": 1.775693253585763e-05,
      "loss": 1.0596,
      "step": 1213
    },
    {
      "epoch": 0.15929145481384288,
      "grad_norm": 77.31381225585938,
      "learning_rate": 1.763717293645939e-05,
      "loss": 1.1495,
      "step": 1214
    },
    {
      "epoch": 0.15942266688535345,
      "grad_norm": 73.98658752441406,
      "learning_rate": 1.7517779486432495e-05,
      "loss": 1.3738,
      "step": 1215
    },
    {
      "epoch": 0.15955387895686404,
      "grad_norm": 83.4875259399414,
      "learning_rate": 1.7398752716548395e-05,
      "loss": 1.2768,
      "step": 1216
    },
    {
      "epoch": 0.1596850910283746,
      "grad_norm": 79.35005187988281,
      "learning_rate": 1.728009315594843e-05,
      "loss": 1.0543,
      "step": 1217
    },
    {
      "epoch": 0.15981630309988518,
      "grad_norm": 79.31051635742188,
      "learning_rate": 1.716180133214149e-05,
      "loss": 1.106,
      "step": 1218
    },
    {
      "epoch": 0.15994751517139577,
      "grad_norm": 72.3387680053711,
      "learning_rate": 1.704387777100165e-05,
      "loss": 1.1441,
      "step": 1219
    },
    {
      "epoch": 0.16007872724290634,
      "grad_norm": 69.92068481445312,
      "learning_rate": 1.6926322996765897e-05,
      "loss": 0.9798,
      "step": 1220
    },
    {
      "epoch": 0.16020993931441693,
      "grad_norm": 74.88939666748047,
      "learning_rate": 1.6809137532031704e-05,
      "loss": 1.2322,
      "step": 1221
    },
    {
      "epoch": 0.1603411513859275,
      "grad_norm": 67.44734954833984,
      "learning_rate": 1.6692321897754758e-05,
      "loss": 0.9565,
      "step": 1222
    },
    {
      "epoch": 0.1604723634574381,
      "grad_norm": 73.39762878417969,
      "learning_rate": 1.65758766132467e-05,
      "loss": 1.1751,
      "step": 1223
    },
    {
      "epoch": 0.16060357552894866,
      "grad_norm": 75.21737670898438,
      "learning_rate": 1.6459802196172668e-05,
      "loss": 1.2423,
      "step": 1224
    },
    {
      "epoch": 0.16073478760045926,
      "grad_norm": 78.3393325805664,
      "learning_rate": 1.634409916254914e-05,
      "loss": 1.0497,
      "step": 1225
    },
    {
      "epoch": 0.16086599967196982,
      "grad_norm": 69.83857727050781,
      "learning_rate": 1.622876802674158e-05,
      "loss": 1.1109,
      "step": 1226
    },
    {
      "epoch": 0.1609972117434804,
      "grad_norm": 69.7317886352539,
      "learning_rate": 1.6113809301462125e-05,
      "loss": 1.1341,
      "step": 1227
    },
    {
      "epoch": 0.16112842381499098,
      "grad_norm": 74.22737884521484,
      "learning_rate": 1.599922349776738e-05,
      "loss": 1.2194,
      "step": 1228
    },
    {
      "epoch": 0.16125963588650155,
      "grad_norm": 69.4048080444336,
      "learning_rate": 1.5885011125056047e-05,
      "loss": 1.1333,
      "step": 1229
    },
    {
      "epoch": 0.16139084795801215,
      "grad_norm": 75.17791748046875,
      "learning_rate": 1.5771172691066794e-05,
      "loss": 0.7728,
      "step": 1230
    },
    {
      "epoch": 0.1615220600295227,
      "grad_norm": 69.05467987060547,
      "learning_rate": 1.565770870187585e-05,
      "loss": 1.0621,
      "step": 1231
    },
    {
      "epoch": 0.1616532721010333,
      "grad_norm": 82.75867462158203,
      "learning_rate": 1.5544619661894864e-05,
      "loss": 1.0526,
      "step": 1232
    },
    {
      "epoch": 0.16178448417254387,
      "grad_norm": 86.98299407958984,
      "learning_rate": 1.543190607386861e-05,
      "loss": 1.2254,
      "step": 1233
    },
    {
      "epoch": 0.16191569624405444,
      "grad_norm": 80.13260650634766,
      "learning_rate": 1.5319568438872745e-05,
      "loss": 1.2042,
      "step": 1234
    },
    {
      "epoch": 0.16204690831556504,
      "grad_norm": 78.50768280029297,
      "learning_rate": 1.520760725631164e-05,
      "loss": 1.0739,
      "step": 1235
    },
    {
      "epoch": 0.1621781203870756,
      "grad_norm": 79.86534118652344,
      "learning_rate": 1.5096023023916094e-05,
      "loss": 1.0542,
      "step": 1236
    },
    {
      "epoch": 0.1623093324585862,
      "grad_norm": 80.50312042236328,
      "learning_rate": 1.498481623774115e-05,
      "loss": 1.2894,
      "step": 1237
    },
    {
      "epoch": 0.16244054453009676,
      "grad_norm": 69.0424575805664,
      "learning_rate": 1.4873987392163947e-05,
      "loss": 1.0506,
      "step": 1238
    },
    {
      "epoch": 0.16257175660160736,
      "grad_norm": 82.29652404785156,
      "learning_rate": 1.4763536979881354e-05,
      "loss": 1.0455,
      "step": 1239
    },
    {
      "epoch": 0.16270296867311793,
      "grad_norm": 69.99261474609375,
      "learning_rate": 1.4653465491908003e-05,
      "loss": 0.9798,
      "step": 1240
    },
    {
      "epoch": 0.1628341807446285,
      "grad_norm": 94.4976577758789,
      "learning_rate": 1.4543773417573925e-05,
      "loss": 1.1891,
      "step": 1241
    },
    {
      "epoch": 0.1629653928161391,
      "grad_norm": 75.3335952758789,
      "learning_rate": 1.4434461244522458e-05,
      "loss": 1.1469,
      "step": 1242
    },
    {
      "epoch": 0.16309660488764965,
      "grad_norm": 89.48926544189453,
      "learning_rate": 1.4325529458708065e-05,
      "loss": 1.2977,
      "step": 1243
    },
    {
      "epoch": 0.16322781695916025,
      "grad_norm": 86.60089111328125,
      "learning_rate": 1.4216978544394177e-05,
      "loss": 1.229,
      "step": 1244
    },
    {
      "epoch": 0.16335902903067082,
      "grad_norm": 74.6124038696289,
      "learning_rate": 1.4108808984151023e-05,
      "loss": 1.0453,
      "step": 1245
    },
    {
      "epoch": 0.1634902411021814,
      "grad_norm": 78.98786926269531,
      "learning_rate": 1.4001021258853509e-05,
      "loss": 1.1977,
      "step": 1246
    },
    {
      "epoch": 0.16362145317369198,
      "grad_norm": 78.58567810058594,
      "learning_rate": 1.3893615847679065e-05,
      "loss": 1.0623,
      "step": 1247
    },
    {
      "epoch": 0.16375266524520257,
      "grad_norm": 63.26145935058594,
      "learning_rate": 1.3786593228105494e-05,
      "loss": 0.777,
      "step": 1248
    },
    {
      "epoch": 0.16388387731671314,
      "grad_norm": 80.01986694335938,
      "learning_rate": 1.3679953875908957e-05,
      "loss": 1.1423,
      "step": 1249
    },
    {
      "epoch": 0.1640150893882237,
      "grad_norm": 71.79955291748047,
      "learning_rate": 1.3573698265161683e-05,
      "loss": 0.8927,
      "step": 1250
    },
    {
      "epoch": 0.1641463014597343,
      "grad_norm": 75.02603149414062,
      "learning_rate": 1.3467826868229994e-05,
      "loss": 1.0336,
      "step": 1251
    },
    {
      "epoch": 0.16427751353124487,
      "grad_norm": 65.329345703125,
      "learning_rate": 1.3362340155772146e-05,
      "loss": 0.9182,
      "step": 1252
    },
    {
      "epoch": 0.16440872560275546,
      "grad_norm": 76.2947006225586,
      "learning_rate": 1.3257238596736266e-05,
      "loss": 0.9962,
      "step": 1253
    },
    {
      "epoch": 0.16453993767426603,
      "grad_norm": 73.18352508544922,
      "learning_rate": 1.3152522658358245e-05,
      "loss": 1.169,
      "step": 1254
    },
    {
      "epoch": 0.16467114974577662,
      "grad_norm": 75.478515625,
      "learning_rate": 1.3048192806159721e-05,
      "loss": 1.2466,
      "step": 1255
    },
    {
      "epoch": 0.1648023618172872,
      "grad_norm": 70.7972640991211,
      "learning_rate": 1.2944249503945894e-05,
      "loss": 1.0656,
      "step": 1256
    },
    {
      "epoch": 0.16493357388879776,
      "grad_norm": 72.05496215820312,
      "learning_rate": 1.2840693213803545e-05,
      "loss": 1.2562,
      "step": 1257
    },
    {
      "epoch": 0.16506478596030835,
      "grad_norm": 63.51637268066406,
      "learning_rate": 1.2737524396099032e-05,
      "loss": 0.9396,
      "step": 1258
    },
    {
      "epoch": 0.16519599803181892,
      "grad_norm": 82.81568908691406,
      "learning_rate": 1.2634743509476088e-05,
      "loss": 1.2662,
      "step": 1259
    },
    {
      "epoch": 0.1653272101033295,
      "grad_norm": 80.03089904785156,
      "learning_rate": 1.2532351010853916e-05,
      "loss": 1.2708,
      "step": 1260
    },
    {
      "epoch": 0.16545842217484008,
      "grad_norm": 87.27790832519531,
      "learning_rate": 1.243034735542512e-05,
      "loss": 1.2648,
      "step": 1261
    },
    {
      "epoch": 0.16558963424635068,
      "grad_norm": 74.34496307373047,
      "learning_rate": 1.2328732996653669e-05,
      "loss": 1.1532,
      "step": 1262
    },
    {
      "epoch": 0.16572084631786124,
      "grad_norm": 71.83038330078125,
      "learning_rate": 1.2227508386272878e-05,
      "loss": 1.1601,
      "step": 1263
    },
    {
      "epoch": 0.1658520583893718,
      "grad_norm": 67.8835220336914,
      "learning_rate": 1.212667397428342e-05,
      "loss": 1.0528,
      "step": 1264
    },
    {
      "epoch": 0.1659832704608824,
      "grad_norm": 68.12181854248047,
      "learning_rate": 1.2026230208951306e-05,
      "loss": 1.0721,
      "step": 1265
    },
    {
      "epoch": 0.16611448253239297,
      "grad_norm": 72.7912368774414,
      "learning_rate": 1.1926177536805905e-05,
      "loss": 1.1382,
      "step": 1266
    },
    {
      "epoch": 0.16624569460390357,
      "grad_norm": 76.56819152832031,
      "learning_rate": 1.1826516402637989e-05,
      "loss": 1.3199,
      "step": 1267
    },
    {
      "epoch": 0.16637690667541413,
      "grad_norm": 67.23606872558594,
      "learning_rate": 1.1727247249497685e-05,
      "loss": 1.0491,
      "step": 1268
    },
    {
      "epoch": 0.16650811874692473,
      "grad_norm": 65.64242553710938,
      "learning_rate": 1.1628370518692533e-05,
      "loss": 0.9767,
      "step": 1269
    },
    {
      "epoch": 0.1666393308184353,
      "grad_norm": 73.8215103149414,
      "learning_rate": 1.152988664978556e-05,
      "loss": 1.0467,
      "step": 1270
    },
    {
      "epoch": 0.1667705428899459,
      "grad_norm": 77.5945816040039,
      "learning_rate": 1.1431796080593283e-05,
      "loss": 1.2715,
      "step": 1271
    },
    {
      "epoch": 0.16690175496145646,
      "grad_norm": 86.18306732177734,
      "learning_rate": 1.1334099247183783e-05,
      "loss": 1.2757,
      "step": 1272
    },
    {
      "epoch": 0.16703296703296702,
      "grad_norm": 67.52281188964844,
      "learning_rate": 1.1236796583874787e-05,
      "loss": 1.0308,
      "step": 1273
    },
    {
      "epoch": 0.16716417910447762,
      "grad_norm": 70.18402862548828,
      "learning_rate": 1.1139888523231678e-05,
      "loss": 1.1145,
      "step": 1274
    },
    {
      "epoch": 0.16729539117598818,
      "grad_norm": 69.39935302734375,
      "learning_rate": 1.1043375496065611e-05,
      "loss": 0.9124,
      "step": 1275
    },
    {
      "epoch": 0.16742660324749878,
      "grad_norm": 75.80091857910156,
      "learning_rate": 1.0947257931431642e-05,
      "loss": 1.1383,
      "step": 1276
    },
    {
      "epoch": 0.16755781531900935,
      "grad_norm": 74.49039459228516,
      "learning_rate": 1.0851536256626705e-05,
      "loss": 1.0753,
      "step": 1277
    },
    {
      "epoch": 0.16768902739051994,
      "grad_norm": 65.25264739990234,
      "learning_rate": 1.0756210897187812e-05,
      "loss": 1.0133,
      "step": 1278
    },
    {
      "epoch": 0.1678202394620305,
      "grad_norm": 71.50276947021484,
      "learning_rate": 1.0661282276890127e-05,
      "loss": 1.0614,
      "step": 1279
    },
    {
      "epoch": 0.16795145153354107,
      "grad_norm": 83.18603515625,
      "learning_rate": 1.0566750817745074e-05,
      "loss": 1.2532,
      "step": 1280
    },
    {
      "epoch": 0.16808266360505167,
      "grad_norm": 84.468017578125,
      "learning_rate": 1.0472616939998492e-05,
      "loss": 1.3327,
      "step": 1281
    },
    {
      "epoch": 0.16821387567656224,
      "grad_norm": 75.32342529296875,
      "learning_rate": 1.0378881062128731e-05,
      "loss": 1.0962,
      "step": 1282
    },
    {
      "epoch": 0.16834508774807283,
      "grad_norm": 84.1277847290039,
      "learning_rate": 1.0285543600844804e-05,
      "loss": 1.1601,
      "step": 1283
    },
    {
      "epoch": 0.1684762998195834,
      "grad_norm": 76.08843994140625,
      "learning_rate": 1.019260497108453e-05,
      "loss": 1.0772,
      "step": 1284
    },
    {
      "epoch": 0.168607511891094,
      "grad_norm": 55.647377014160156,
      "learning_rate": 1.010006558601274e-05,
      "loss": 0.6075,
      "step": 1285
    },
    {
      "epoch": 0.16873872396260456,
      "grad_norm": 73.25740814208984,
      "learning_rate": 1.000792585701934e-05,
      "loss": 0.9784,
      "step": 1286
    },
    {
      "epoch": 0.16886993603411513,
      "grad_norm": 72.43972778320312,
      "learning_rate": 9.91618619371757e-06,
      "loss": 1.0008,
      "step": 1287
    },
    {
      "epoch": 0.16900114810562572,
      "grad_norm": 84.33741760253906,
      "learning_rate": 9.82484700394215e-06,
      "loss": 1.1662,
      "step": 1288
    },
    {
      "epoch": 0.1691323601771363,
      "grad_norm": 66.02348327636719,
      "learning_rate": 9.73390869374743e-06,
      "loss": 0.7879,
      "step": 1289
    },
    {
      "epoch": 0.16926357224864688,
      "grad_norm": 93.93775939941406,
      "learning_rate": 9.643371667405698e-06,
      "loss": 1.1358,
      "step": 1290
    },
    {
      "epoch": 0.16939478432015745,
      "grad_norm": 78.03810119628906,
      "learning_rate": 9.553236327405246e-06,
      "loss": 1.0244,
      "step": 1291
    },
    {
      "epoch": 0.16952599639166804,
      "grad_norm": 75.52627563476562,
      "learning_rate": 9.463503074448677e-06,
      "loss": 1.0874,
      "step": 1292
    },
    {
      "epoch": 0.1696572084631786,
      "grad_norm": 86.1750717163086,
      "learning_rate": 9.374172307451068e-06,
      "loss": 1.015,
      "step": 1293
    },
    {
      "epoch": 0.1697884205346892,
      "grad_norm": 72.38187408447266,
      "learning_rate": 9.285244423538197e-06,
      "loss": 0.9809,
      "step": 1294
    },
    {
      "epoch": 0.16991963260619977,
      "grad_norm": 75.3546371459961,
      "learning_rate": 9.196719818044886e-06,
      "loss": 0.9504,
      "step": 1295
    },
    {
      "epoch": 0.17005084467771034,
      "grad_norm": 84.32721710205078,
      "learning_rate": 9.108598884513053e-06,
      "loss": 1.128,
      "step": 1296
    },
    {
      "epoch": 0.17018205674922093,
      "grad_norm": 78.09004211425781,
      "learning_rate": 9.020882014690136e-06,
      "loss": 1.1886,
      "step": 1297
    },
    {
      "epoch": 0.1703132688207315,
      "grad_norm": 70.17216491699219,
      "learning_rate": 8.933569598527247e-06,
      "loss": 0.8075,
      "step": 1298
    },
    {
      "epoch": 0.1704444808922421,
      "grad_norm": 68.59973907470703,
      "learning_rate": 8.846662024177477e-06,
      "loss": 0.9946,
      "step": 1299
    },
    {
      "epoch": 0.17057569296375266,
      "grad_norm": 80.15505981445312,
      "learning_rate": 8.760159677994172e-06,
      "loss": 1.047,
      "step": 1300
    },
    {
      "epoch": 0.17070690503526326,
      "grad_norm": 83.08203125,
      "learning_rate": 8.674062944529216e-06,
      "loss": 1.2655,
      "step": 1301
    },
    {
      "epoch": 0.17083811710677382,
      "grad_norm": 79.71021270751953,
      "learning_rate": 8.588372206531292e-06,
      "loss": 1.2491,
      "step": 1302
    },
    {
      "epoch": 0.1709693291782844,
      "grad_norm": 63.94439697265625,
      "learning_rate": 8.503087844944213e-06,
      "loss": 0.9798,
      "step": 1303
    },
    {
      "epoch": 0.17110054124979499,
      "grad_norm": 78.41773223876953,
      "learning_rate": 8.418210238905256e-06,
      "loss": 0.9977,
      "step": 1304
    },
    {
      "epoch": 0.17123175332130555,
      "grad_norm": 79.95426177978516,
      "learning_rate": 8.333739765743398e-06,
      "loss": 1.2061,
      "step": 1305
    },
    {
      "epoch": 0.17136296539281615,
      "grad_norm": 74.92925262451172,
      "learning_rate": 8.249676800977658e-06,
      "loss": 1.1326,
      "step": 1306
    },
    {
      "epoch": 0.1714941774643267,
      "grad_norm": 60.196258544921875,
      "learning_rate": 8.16602171831553e-06,
      "loss": 0.8887,
      "step": 1307
    },
    {
      "epoch": 0.1716253895358373,
      "grad_norm": 83.44329071044922,
      "learning_rate": 8.082774889651168e-06,
      "loss": 1.3348,
      "step": 1308
    },
    {
      "epoch": 0.17175660160734788,
      "grad_norm": 83.41243743896484,
      "learning_rate": 7.999936685063835e-06,
      "loss": 1.2002,
      "step": 1309
    },
    {
      "epoch": 0.17188781367885844,
      "grad_norm": 74.7232437133789,
      "learning_rate": 7.91750747281621e-06,
      "loss": 1.1554,
      "step": 1310
    },
    {
      "epoch": 0.17201902575036904,
      "grad_norm": 83.086181640625,
      "learning_rate": 7.835487619352811e-06,
      "loss": 1.1803,
      "step": 1311
    },
    {
      "epoch": 0.1721502378218796,
      "grad_norm": 77.87909698486328,
      "learning_rate": 7.753877489298244e-06,
      "loss": 1.1366,
      "step": 1312
    },
    {
      "epoch": 0.1722814498933902,
      "grad_norm": 81.8636245727539,
      "learning_rate": 7.67267744545579e-06,
      "loss": 1.1778,
      "step": 1313
    },
    {
      "epoch": 0.17241266196490077,
      "grad_norm": 71.06631469726562,
      "learning_rate": 7.591887848805545e-06,
      "loss": 1.1026,
      "step": 1314
    },
    {
      "epoch": 0.17254387403641136,
      "grad_norm": 81.228515625,
      "learning_rate": 7.5115090585029966e-06,
      "loss": 1.2549,
      "step": 1315
    },
    {
      "epoch": 0.17267508610792193,
      "grad_norm": 79.7537841796875,
      "learning_rate": 7.431541431877342e-06,
      "loss": 1.2129,
      "step": 1316
    },
    {
      "epoch": 0.17280629817943252,
      "grad_norm": 65.91927337646484,
      "learning_rate": 7.351985324429933e-06,
      "loss": 0.9899,
      "step": 1317
    },
    {
      "epoch": 0.1729375102509431,
      "grad_norm": 72.38845825195312,
      "learning_rate": 7.272841089832694e-06,
      "loss": 1.1895,
      "step": 1318
    },
    {
      "epoch": 0.17306872232245366,
      "grad_norm": 61.024662017822266,
      "learning_rate": 7.194109079926514e-06,
      "loss": 0.8673,
      "step": 1319
    },
    {
      "epoch": 0.17319993439396425,
      "grad_norm": 78.17933654785156,
      "learning_rate": 7.115789644719728e-06,
      "loss": 1.1239,
      "step": 1320
    },
    {
      "epoch": 0.17333114646547482,
      "grad_norm": 84.16743469238281,
      "learning_rate": 7.037883132386547e-06,
      "loss": 1.178,
      "step": 1321
    },
    {
      "epoch": 0.1734623585369854,
      "grad_norm": 77.6754379272461,
      "learning_rate": 6.960389889265517e-06,
      "loss": 1.2245,
      "step": 1322
    },
    {
      "epoch": 0.17359357060849598,
      "grad_norm": 73.6377182006836,
      "learning_rate": 6.883310259857944e-06,
      "loss": 1.1687,
      "step": 1323
    },
    {
      "epoch": 0.17372478268000657,
      "grad_norm": 83.02607727050781,
      "learning_rate": 6.806644586826383e-06,
      "loss": 1.2379,
      "step": 1324
    },
    {
      "epoch": 0.17385599475151714,
      "grad_norm": 82.90789031982422,
      "learning_rate": 6.730393210993147e-06,
      "loss": 1.2754,
      "step": 1325
    },
    {
      "epoch": 0.1739872068230277,
      "grad_norm": 102.3957290649414,
      "learning_rate": 6.654556471338746e-06,
      "loss": 1.2487,
      "step": 1326
    },
    {
      "epoch": 0.1741184188945383,
      "grad_norm": 74.44820404052734,
      "learning_rate": 6.579134705000412e-06,
      "loss": 0.9587,
      "step": 1327
    },
    {
      "epoch": 0.17424963096604887,
      "grad_norm": 74.1882095336914,
      "learning_rate": 6.504128247270546e-06,
      "loss": 1.1237,
      "step": 1328
    },
    {
      "epoch": 0.17438084303755946,
      "grad_norm": 68.57334899902344,
      "learning_rate": 6.429537431595312e-06,
      "loss": 0.9795,
      "step": 1329
    },
    {
      "epoch": 0.17451205510907003,
      "grad_norm": 69.1958999633789,
      "learning_rate": 6.355362589573077e-06,
      "loss": 0.8111,
      "step": 1330
    },
    {
      "epoch": 0.17464326718058062,
      "grad_norm": 67.0915298461914,
      "learning_rate": 6.2816040509530165e-06,
      "loss": 1.134,
      "step": 1331
    },
    {
      "epoch": 0.1747744792520912,
      "grad_norm": 79.82743835449219,
      "learning_rate": 6.2082621436335475e-06,
      "loss": 1.2666,
      "step": 1332
    },
    {
      "epoch": 0.17490569132360176,
      "grad_norm": 76.10675811767578,
      "learning_rate": 6.135337193660962e-06,
      "loss": 1.0995,
      "step": 1333
    },
    {
      "epoch": 0.17503690339511235,
      "grad_norm": 82.753662109375,
      "learning_rate": 6.062829525227909e-06,
      "loss": 1.1632,
      "step": 1334
    },
    {
      "epoch": 0.17516811546662292,
      "grad_norm": 73.97330474853516,
      "learning_rate": 5.990739460672024e-06,
      "loss": 0.9425,
      "step": 1335
    },
    {
      "epoch": 0.17529932753813351,
      "grad_norm": 62.40740203857422,
      "learning_rate": 5.9190673204744255e-06,
      "loss": 0.9758,
      "step": 1336
    },
    {
      "epoch": 0.17543053960964408,
      "grad_norm": 75.70748138427734,
      "learning_rate": 5.84781342325833e-06,
      "loss": 1.1283,
      "step": 1337
    },
    {
      "epoch": 0.17556175168115468,
      "grad_norm": 75.87496185302734,
      "learning_rate": 5.77697808578761e-06,
      "loss": 1.0084,
      "step": 1338
    },
    {
      "epoch": 0.17569296375266524,
      "grad_norm": 72.41253662109375,
      "learning_rate": 5.706561622965467e-06,
      "loss": 1.0456,
      "step": 1339
    },
    {
      "epoch": 0.17582417582417584,
      "grad_norm": 89.17688751220703,
      "learning_rate": 5.636564347832907e-06,
      "loss": 0.9937,
      "step": 1340
    },
    {
      "epoch": 0.1759553878956864,
      "grad_norm": 82.14151000976562,
      "learning_rate": 5.566986571567401e-06,
      "loss": 1.0701,
      "step": 1341
    },
    {
      "epoch": 0.17608659996719697,
      "grad_norm": 83.62397003173828,
      "learning_rate": 5.497828603481569e-06,
      "loss": 1.2314,
      "step": 1342
    },
    {
      "epoch": 0.17621781203870757,
      "grad_norm": 77.08482360839844,
      "learning_rate": 5.429090751021704e-06,
      "loss": 1.1964,
      "step": 1343
    },
    {
      "epoch": 0.17634902411021813,
      "grad_norm": 79.92993927001953,
      "learning_rate": 5.3607733197664436e-06,
      "loss": 1.1152,
      "step": 1344
    },
    {
      "epoch": 0.17648023618172873,
      "grad_norm": 87.93970489501953,
      "learning_rate": 5.2928766134254345e-06,
      "loss": 1.0511,
      "step": 1345
    },
    {
      "epoch": 0.1766114482532393,
      "grad_norm": 86.56250762939453,
      "learning_rate": 5.225400933837954e-06,
      "loss": 1.1245,
      "step": 1346
    },
    {
      "epoch": 0.1767426603247499,
      "grad_norm": 82.0915298461914,
      "learning_rate": 5.158346580971573e-06,
      "loss": 1.1714,
      "step": 1347
    },
    {
      "epoch": 0.17687387239626046,
      "grad_norm": 76.38935852050781,
      "learning_rate": 5.091713852920854e-06,
      "loss": 0.9775,
      "step": 1348
    },
    {
      "epoch": 0.17700508446777102,
      "grad_norm": 82.80403137207031,
      "learning_rate": 5.025503045905933e-06,
      "loss": 1.0522,
      "step": 1349
    },
    {
      "epoch": 0.17713629653928162,
      "grad_norm": 74.62669372558594,
      "learning_rate": 4.959714454271369e-06,
      "loss": 0.8874,
      "step": 1350
    },
    {
      "epoch": 0.17726750861079218,
      "grad_norm": 68.98562622070312,
      "learning_rate": 4.8943483704846475e-06,
      "loss": 1.0053,
      "step": 1351
    },
    {
      "epoch": 0.17739872068230278,
      "grad_norm": 82.97789764404297,
      "learning_rate": 4.829405085134997e-06,
      "loss": 1.2898,
      "step": 1352
    },
    {
      "epoch": 0.17752993275381335,
      "grad_norm": 73.07791137695312,
      "learning_rate": 4.764884886932086e-06,
      "loss": 1.0344,
      "step": 1353
    },
    {
      "epoch": 0.17766114482532394,
      "grad_norm": 77.70623016357422,
      "learning_rate": 4.700788062704687e-06,
      "loss": 1.0887,
      "step": 1354
    },
    {
      "epoch": 0.1777923568968345,
      "grad_norm": 82.22172546386719,
      "learning_rate": 4.6371148973994525e-06,
      "loss": 1.151,
      "step": 1355
    },
    {
      "epoch": 0.17792356896834507,
      "grad_norm": 66.64952087402344,
      "learning_rate": 4.573865674079625e-06,
      "loss": 1.0883,
      "step": 1356
    },
    {
      "epoch": 0.17805478103985567,
      "grad_norm": 97.41364288330078,
      "learning_rate": 4.511040673923828e-06,
      "loss": 1.2314,
      "step": 1357
    },
    {
      "epoch": 0.17818599311136624,
      "grad_norm": 85.74223327636719,
      "learning_rate": 4.448640176224694e-06,
      "loss": 1.2398,
      "step": 1358
    },
    {
      "epoch": 0.17831720518287683,
      "grad_norm": 82.7259292602539,
      "learning_rate": 4.386664458387779e-06,
      "loss": 1.2842,
      "step": 1359
    },
    {
      "epoch": 0.1784484172543874,
      "grad_norm": 63.9185905456543,
      "learning_rate": 4.325113795930203e-06,
      "loss": 0.9728,
      "step": 1360
    },
    {
      "epoch": 0.178579629325898,
      "grad_norm": 78.6162338256836,
      "learning_rate": 4.263988462479484e-06,
      "loss": 1.1413,
      "step": 1361
    },
    {
      "epoch": 0.17871084139740856,
      "grad_norm": 74.86878204345703,
      "learning_rate": 4.203288729772326e-06,
      "loss": 1.2287,
      "step": 1362
    },
    {
      "epoch": 0.17884205346891915,
      "grad_norm": 75.49349212646484,
      "learning_rate": 4.143014867653383e-06,
      "loss": 1.1461,
      "step": 1363
    },
    {
      "epoch": 0.17897326554042972,
      "grad_norm": 96.10244750976562,
      "learning_rate": 4.083167144074073e-06,
      "loss": 1.3352,
      "step": 1364
    },
    {
      "epoch": 0.1791044776119403,
      "grad_norm": 87.82105255126953,
      "learning_rate": 4.023745825091407e-06,
      "loss": 1.3263,
      "step": 1365
    },
    {
      "epoch": 0.17923568968345088,
      "grad_norm": 68.80326080322266,
      "learning_rate": 3.964751174866765e-06,
      "loss": 0.9917,
      "step": 1366
    },
    {
      "epoch": 0.17936690175496145,
      "grad_norm": 77.65306854248047,
      "learning_rate": 3.906183455664725e-06,
      "loss": 1.3275,
      "step": 1367
    },
    {
      "epoch": 0.17949811382647204,
      "grad_norm": 80.83008575439453,
      "learning_rate": 3.84804292785198e-06,
      "loss": 1.3739,
      "step": 1368
    },
    {
      "epoch": 0.1796293258979826,
      "grad_norm": 72.2458267211914,
      "learning_rate": 3.7903298498960572e-06,
      "loss": 0.9313,
      "step": 1369
    },
    {
      "epoch": 0.1797605379694932,
      "grad_norm": 72.01419067382812,
      "learning_rate": 3.7330444783642338e-06,
      "loss": 1.0628,
      "step": 1370
    },
    {
      "epoch": 0.17989175004100377,
      "grad_norm": 80.39382934570312,
      "learning_rate": 3.676187067922421e-06,
      "loss": 1.0405,
      "step": 1371
    },
    {
      "epoch": 0.18002296211251434,
      "grad_norm": 83.1916732788086,
      "learning_rate": 3.619757871333973e-06,
      "loss": 1.2964,
      "step": 1372
    },
    {
      "epoch": 0.18015417418402493,
      "grad_norm": 70.99177551269531,
      "learning_rate": 3.563757139458579e-06,
      "loss": 1.0668,
      "step": 1373
    },
    {
      "epoch": 0.1802853862555355,
      "grad_norm": 73.41712188720703,
      "learning_rate": 3.5081851212512175e-06,
      "loss": 1.0965,
      "step": 1374
    },
    {
      "epoch": 0.1804165983270461,
      "grad_norm": 68.72308349609375,
      "learning_rate": 3.4530420637609363e-06,
      "loss": 1.0938,
      "step": 1375
    },
    {
      "epoch": 0.18054781039855666,
      "grad_norm": 79.68484497070312,
      "learning_rate": 3.3983282121298086e-06,
      "loss": 1.157,
      "step": 1376
    },
    {
      "epoch": 0.18067902247006726,
      "grad_norm": 77.23564147949219,
      "learning_rate": 3.3440438095919126e-06,
      "loss": 1.1229,
      "step": 1377
    },
    {
      "epoch": 0.18081023454157782,
      "grad_norm": 81.44544982910156,
      "learning_rate": 3.290189097472096e-06,
      "loss": 1.2141,
      "step": 1378
    },
    {
      "epoch": 0.1809414466130884,
      "grad_norm": 79.25465393066406,
      "learning_rate": 3.236764315185037e-06,
      "loss": 1.1874,
      "step": 1379
    },
    {
      "epoch": 0.18107265868459899,
      "grad_norm": 66.08275604248047,
      "learning_rate": 3.1837697002341293e-06,
      "loss": 0.8021,
      "step": 1380
    },
    {
      "epoch": 0.18120387075610955,
      "grad_norm": 63.78189468383789,
      "learning_rate": 3.131205488210409e-06,
      "loss": 0.9785,
      "step": 1381
    },
    {
      "epoch": 0.18133508282762015,
      "grad_norm": 87.28047180175781,
      "learning_rate": 3.0790719127915646e-06,
      "loss": 1.149,
      "step": 1382
    },
    {
      "epoch": 0.18146629489913071,
      "grad_norm": 77.17674255371094,
      "learning_rate": 3.0273692057408265e-06,
      "loss": 1.0645,
      "step": 1383
    },
    {
      "epoch": 0.1815975069706413,
      "grad_norm": 75.72791290283203,
      "learning_rate": 2.976097596905969e-06,
      "loss": 1.1052,
      "step": 1384
    },
    {
      "epoch": 0.18172871904215188,
      "grad_norm": 75.91800689697266,
      "learning_rate": 2.9252573142183326e-06,
      "loss": 1.1013,
      "step": 1385
    },
    {
      "epoch": 0.18185993111366247,
      "grad_norm": 75.53617095947266,
      "learning_rate": 2.874848583691714e-06,
      "loss": 1.0824,
      "step": 1386
    },
    {
      "epoch": 0.18199114318517304,
      "grad_norm": 77.76245880126953,
      "learning_rate": 2.8248716294214774e-06,
      "loss": 1.0884,
      "step": 1387
    },
    {
      "epoch": 0.1821223552566836,
      "grad_norm": 77.03966522216797,
      "learning_rate": 2.7753266735834338e-06,
      "loss": 1.2191,
      "step": 1388
    },
    {
      "epoch": 0.1822535673281942,
      "grad_norm": 74.19985961914062,
      "learning_rate": 2.7262139364329643e-06,
      "loss": 1.0083,
      "step": 1389
    },
    {
      "epoch": 0.18238477939970477,
      "grad_norm": 70.45758819580078,
      "learning_rate": 2.677533636303964e-06,
      "loss": 1.1458,
      "step": 1390
    },
    {
      "epoch": 0.18251599147121536,
      "grad_norm": 78.09001159667969,
      "learning_rate": 2.6292859896079213e-06,
      "loss": 0.9875,
      "step": 1391
    },
    {
      "epoch": 0.18264720354272593,
      "grad_norm": 86.18848419189453,
      "learning_rate": 2.581471210832931e-06,
      "loss": 1.369,
      "step": 1392
    },
    {
      "epoch": 0.18277841561423652,
      "grad_norm": 75.87664031982422,
      "learning_rate": 2.5340895125427364e-06,
      "loss": 1.1006,
      "step": 1393
    },
    {
      "epoch": 0.1829096276857471,
      "grad_norm": 77.39629364013672,
      "learning_rate": 2.4871411053757898e-06,
      "loss": 1.1746,
      "step": 1394
    },
    {
      "epoch": 0.18304083975725766,
      "grad_norm": 81.56670379638672,
      "learning_rate": 2.440626198044327e-06,
      "loss": 1.2028,
      "step": 1395
    },
    {
      "epoch": 0.18317205182876825,
      "grad_norm": 78.2774658203125,
      "learning_rate": 2.394544997333437e-06,
      "loss": 1.065,
      "step": 1396
    },
    {
      "epoch": 0.18330326390027882,
      "grad_norm": 76.36490631103516,
      "learning_rate": 2.3488977081001394e-06,
      "loss": 0.9461,
      "step": 1397
    },
    {
      "epoch": 0.1834344759717894,
      "grad_norm": 67.82685089111328,
      "learning_rate": 2.3036845332724543e-06,
      "loss": 0.9506,
      "step": 1398
    },
    {
      "epoch": 0.18356568804329998,
      "grad_norm": 71.72674560546875,
      "learning_rate": 2.2589056738485324e-06,
      "loss": 0.9184,
      "step": 1399
    },
    {
      "epoch": 0.18369690011481057,
      "grad_norm": 73.05416107177734,
      "learning_rate": 2.2145613288957478e-06,
      "loss": 1.0358,
      "step": 1400
    },
    {
      "epoch": 0.18382811218632114,
      "grad_norm": 78.15758514404297,
      "learning_rate": 2.170651695549786e-06,
      "loss": 0.9577,
      "step": 1401
    },
    {
      "epoch": 0.1839593242578317,
      "grad_norm": 75.2143783569336,
      "learning_rate": 2.1271769690138332e-06,
      "loss": 1.1971,
      "step": 1402
    },
    {
      "epoch": 0.1840905363293423,
      "grad_norm": 69.4281234741211,
      "learning_rate": 2.084137342557646e-06,
      "loss": 0.9872,
      "step": 1403
    },
    {
      "epoch": 0.18422174840085287,
      "grad_norm": 71.05724334716797,
      "learning_rate": 2.0415330075166937e-06,
      "loss": 0.9218,
      "step": 1404
    },
    {
      "epoch": 0.18435296047236346,
      "grad_norm": 80.1824722290039,
      "learning_rate": 1.9993641532913833e-06,
      "loss": 1.4059,
      "step": 1405
    },
    {
      "epoch": 0.18448417254387403,
      "grad_norm": 77.80947875976562,
      "learning_rate": 1.9576309673461357e-06,
      "loss": 1.0619,
      "step": 1406
    },
    {
      "epoch": 0.18461538461538463,
      "grad_norm": 75.29508209228516,
      "learning_rate": 1.916333635208556e-06,
      "loss": 1.1667,
      "step": 1407
    },
    {
      "epoch": 0.1847465966868952,
      "grad_norm": 76.84593200683594,
      "learning_rate": 1.8754723404686425e-06,
      "loss": 1.2245,
      "step": 1408
    },
    {
      "epoch": 0.1848778087584058,
      "grad_norm": 76.54517364501953,
      "learning_rate": 1.8350472647780116e-06,
      "loss": 1.064,
      "step": 1409
    },
    {
      "epoch": 0.18500902082991635,
      "grad_norm": 74.9830093383789,
      "learning_rate": 1.7950585878489856e-06,
      "loss": 1.1909,
      "step": 1410
    },
    {
      "epoch": 0.18514023290142692,
      "grad_norm": 77.46250915527344,
      "learning_rate": 1.7555064874538397e-06,
      "loss": 1.2028,
      "step": 1411
    },
    {
      "epoch": 0.18527144497293752,
      "grad_norm": 77.85829162597656,
      "learning_rate": 1.7163911394240672e-06,
      "loss": 1.1558,
      "step": 1412
    },
    {
      "epoch": 0.18540265704444808,
      "grad_norm": 76.5018310546875,
      "learning_rate": 1.6777127176495043e-06,
      "loss": 0.965,
      "step": 1413
    },
    {
      "epoch": 0.18553386911595868,
      "grad_norm": 71.07376861572266,
      "learning_rate": 1.6394713940776296e-06,
      "loss": 1.2065,
      "step": 1414
    },
    {
      "epoch": 0.18566508118746924,
      "grad_norm": 70.07061004638672,
      "learning_rate": 1.6016673387127646e-06,
      "loss": 1.0297,
      "step": 1415
    },
    {
      "epoch": 0.18579629325897984,
      "grad_norm": 77.03162384033203,
      "learning_rate": 1.5643007196153302e-06,
      "loss": 1.1695,
      "step": 1416
    },
    {
      "epoch": 0.1859275053304904,
      "grad_norm": 66.60127258300781,
      "learning_rate": 1.5273717029010925e-06,
      "loss": 0.7601,
      "step": 1417
    },
    {
      "epoch": 0.18605871740200097,
      "grad_norm": 69.15326690673828,
      "learning_rate": 1.4908804527404286e-06,
      "loss": 1.0758,
      "step": 1418
    },
    {
      "epoch": 0.18618992947351157,
      "grad_norm": 79.64305114746094,
      "learning_rate": 1.4548271313575835e-06,
      "loss": 1.1285,
      "step": 1419
    },
    {
      "epoch": 0.18632114154502213,
      "grad_norm": 66.04415893554688,
      "learning_rate": 1.4192118990299707e-06,
      "loss": 1.0119,
      "step": 1420
    },
    {
      "epoch": 0.18645235361653273,
      "grad_norm": 84.27361297607422,
      "learning_rate": 1.3840349140874619e-06,
      "loss": 1.1431,
      "step": 1421
    },
    {
      "epoch": 0.1865835656880433,
      "grad_norm": 74.79593658447266,
      "learning_rate": 1.3492963329116537e-06,
      "loss": 1.1957,
      "step": 1422
    },
    {
      "epoch": 0.1867147777595539,
      "grad_norm": 75.47400665283203,
      "learning_rate": 1.3149963099352014e-06,
      "loss": 1.2316,
      "step": 1423
    },
    {
      "epoch": 0.18684598983106446,
      "grad_norm": 84.40567016601562,
      "learning_rate": 1.2811349976411202e-06,
      "loss": 1.1573,
      "step": 1424
    },
    {
      "epoch": 0.18697720190257502,
      "grad_norm": 79.46822357177734,
      "learning_rate": 1.2477125465620853e-06,
      "loss": 1.2147,
      "step": 1425
    },
    {
      "epoch": 0.18710841397408562,
      "grad_norm": 82.89403533935547,
      "learning_rate": 1.2147291052798216e-06,
      "loss": 1.1681,
      "step": 1426
    },
    {
      "epoch": 0.18723962604559619,
      "grad_norm": 79.21101379394531,
      "learning_rate": 1.1821848204243814e-06,
      "loss": 1.1436,
      "step": 1427
    },
    {
      "epoch": 0.18737083811710678,
      "grad_norm": 76.48783874511719,
      "learning_rate": 1.1500798366735233e-06,
      "loss": 1.0509,
      "step": 1428
    },
    {
      "epoch": 0.18750205018861735,
      "grad_norm": 70.59283447265625,
      "learning_rate": 1.1184142967520794e-06,
      "loss": 1.0912,
      "step": 1429
    },
    {
      "epoch": 0.18763326226012794,
      "grad_norm": 80.98322296142578,
      "learning_rate": 1.0871883414312777e-06,
      "loss": 0.9694,
      "step": 1430
    },
    {
      "epoch": 0.1877644743316385,
      "grad_norm": 82.34637451171875,
      "learning_rate": 1.0564021095281652e-06,
      "loss": 1.2436,
      "step": 1431
    },
    {
      "epoch": 0.18789568640314908,
      "grad_norm": 84.2276611328125,
      "learning_rate": 1.0260557379049519e-06,
      "loss": 1.2572,
      "step": 1432
    },
    {
      "epoch": 0.18802689847465967,
      "grad_norm": 76.35618591308594,
      "learning_rate": 9.96149361468457e-07,
      "loss": 1.1525,
      "step": 1433
    },
    {
      "epoch": 0.18815811054617024,
      "grad_norm": 75.3700942993164,
      "learning_rate": 9.66683113169431e-07,
      "loss": 0.995,
      "step": 1434
    },
    {
      "epoch": 0.18828932261768083,
      "grad_norm": 77.96818542480469,
      "learning_rate": 9.376571240020227e-07,
      "loss": 1.0751,
      "step": 1435
    },
    {
      "epoch": 0.1884205346891914,
      "grad_norm": 66.5455093383789,
      "learning_rate": 9.090715230031688e-07,
      "loss": 0.9993,
      "step": 1436
    },
    {
      "epoch": 0.188551746760702,
      "grad_norm": 80.74208068847656,
      "learning_rate": 8.809264372520609e-07,
      "loss": 1.2364,
      "step": 1437
    },
    {
      "epoch": 0.18868295883221256,
      "grad_norm": 94.85200500488281,
      "learning_rate": 8.532219918695128e-07,
      "loss": 1.3317,
      "step": 1438
    },
    {
      "epoch": 0.18881417090372316,
      "grad_norm": 64.76235961914062,
      "learning_rate": 8.259583100174606e-07,
      "loss": 0.8093,
      "step": 1439
    },
    {
      "epoch": 0.18894538297523372,
      "grad_norm": 80.01709747314453,
      "learning_rate": 7.991355128984079e-07,
      "loss": 1.1863,
      "step": 1440
    },
    {
      "epoch": 0.1890765950467443,
      "grad_norm": 84.49312591552734,
      "learning_rate": 7.727537197548707e-07,
      "loss": 1.1203,
      "step": 1441
    },
    {
      "epoch": 0.18920780711825488,
      "grad_norm": 81.1505126953125,
      "learning_rate": 7.468130478688218e-07,
      "loss": 1.0597,
      "step": 1442
    },
    {
      "epoch": 0.18933901918976545,
      "grad_norm": 83.70401000976562,
      "learning_rate": 7.213136125612586e-07,
      "loss": 1.1287,
      "step": 1443
    },
    {
      "epoch": 0.18947023126127605,
      "grad_norm": 84.43779754638672,
      "learning_rate": 6.962555271915805e-07,
      "loss": 1.1866,
      "step": 1444
    },
    {
      "epoch": 0.1896014433327866,
      "grad_norm": 80.268310546875,
      "learning_rate": 6.716389031571568e-07,
      "loss": 1.1742,
      "step": 1445
    },
    {
      "epoch": 0.1897326554042972,
      "grad_norm": 84.96894836425781,
      "learning_rate": 6.474638498928265e-07,
      "loss": 1.2499,
      "step": 1446
    },
    {
      "epoch": 0.18986386747580777,
      "grad_norm": 78.9056167602539,
      "learning_rate": 6.237304748703543e-07,
      "loss": 1.1802,
      "step": 1447
    },
    {
      "epoch": 0.18999507954731834,
      "grad_norm": 76.67760467529297,
      "learning_rate": 6.004388835980423e-07,
      "loss": 0.8561,
      "step": 1448
    },
    {
      "epoch": 0.19012629161882894,
      "grad_norm": 77.4666748046875,
      "learning_rate": 5.77589179620186e-07,
      "loss": 1.0832,
      "step": 1449
    },
    {
      "epoch": 0.1902575036903395,
      "grad_norm": 76.02294921875,
      "learning_rate": 5.55181464516652e-07,
      "loss": 0.9128,
      "step": 1450
    },
    {
      "epoch": 0.1903887157618501,
      "grad_norm": 85.0171890258789,
      "learning_rate": 5.332158379024122e-07,
      "loss": 1.0897,
      "step": 1451
    },
    {
      "epoch": 0.19051992783336066,
      "grad_norm": 78.6381607055664,
      "learning_rate": 5.116923974270993e-07,
      "loss": 1.1214,
      "step": 1452
    },
    {
      "epoch": 0.19065113990487126,
      "grad_norm": 76.16878509521484,
      "learning_rate": 4.906112387745965e-07,
      "loss": 1.0595,
      "step": 1453
    },
    {
      "epoch": 0.19078235197638183,
      "grad_norm": 78.52881622314453,
      "learning_rate": 4.6997245566257064e-07,
      "loss": 1.1227,
      "step": 1454
    },
    {
      "epoch": 0.1909135640478924,
      "grad_norm": 79.7121353149414,
      "learning_rate": 4.497761398421063e-07,
      "loss": 1.142,
      "step": 1455
    },
    {
      "epoch": 0.191044776119403,
      "grad_norm": 81.0895004272461,
      "learning_rate": 4.3002238109723927e-07,
      "loss": 1.1348,
      "step": 1456
    },
    {
      "epoch": 0.19117598819091355,
      "grad_norm": 78.26695251464844,
      "learning_rate": 4.107112672446123e-07,
      "loss": 1.2305,
      "step": 1457
    },
    {
      "epoch": 0.19130720026242415,
      "grad_norm": 77.22274780273438,
      "learning_rate": 3.9184288413306456e-07,
      "loss": 1.086,
      "step": 1458
    },
    {
      "epoch": 0.19143841233393472,
      "grad_norm": 74.83328247070312,
      "learning_rate": 3.734173156432208e-07,
      "loss": 1.1381,
      "step": 1459
    },
    {
      "epoch": 0.1915696244054453,
      "grad_norm": 84.7967529296875,
      "learning_rate": 3.554346436871581e-07,
      "loss": 1.1587,
      "step": 1460
    },
    {
      "epoch": 0.19170083647695588,
      "grad_norm": 67.37308502197266,
      "learning_rate": 3.3789494820803957e-07,
      "loss": 0.8724,
      "step": 1461
    },
    {
      "epoch": 0.19183204854846647,
      "grad_norm": 71.20195007324219,
      "learning_rate": 3.2079830717972606e-07,
      "loss": 1.0631,
      "step": 1462
    },
    {
      "epoch": 0.19196326061997704,
      "grad_norm": 76.33344268798828,
      "learning_rate": 3.041447966064648e-07,
      "loss": 1.2219,
      "step": 1463
    },
    {
      "epoch": 0.1920944726914876,
      "grad_norm": 67.13754272460938,
      "learning_rate": 2.8793449052254563e-07,
      "loss": 0.8903,
      "step": 1464
    },
    {
      "epoch": 0.1922256847629982,
      "grad_norm": 67.1708984375,
      "learning_rate": 2.721674609919345e-07,
      "loss": 1.0607,
      "step": 1465
    },
    {
      "epoch": 0.19235689683450877,
      "grad_norm": 79.79737854003906,
      "learning_rate": 2.568437781080069e-07,
      "loss": 1.0538,
      "step": 1466
    },
    {
      "epoch": 0.19248810890601936,
      "grad_norm": 79.96922302246094,
      "learning_rate": 2.4196350999320384e-07,
      "loss": 1.1157,
      "step": 1467
    },
    {
      "epoch": 0.19261932097752993,
      "grad_norm": 74.80127716064453,
      "learning_rate": 2.275267227987321e-07,
      "loss": 1.0899,
      "step": 1468
    },
    {
      "epoch": 0.19275053304904052,
      "grad_norm": 76.32405090332031,
      "learning_rate": 2.135334807042866e-07,
      "loss": 1.0197,
      "step": 1469
    },
    {
      "epoch": 0.1928817451205511,
      "grad_norm": 80.53022003173828,
      "learning_rate": 1.9998384591773944e-07,
      "loss": 1.283,
      "step": 1470
    },
    {
      "epoch": 0.19301295719206166,
      "grad_norm": 76.90985107421875,
      "learning_rate": 1.8687787867489592e-07,
      "loss": 1.0975,
      "step": 1471
    },
    {
      "epoch": 0.19314416926357225,
      "grad_norm": 71.95731353759766,
      "learning_rate": 1.7421563723919454e-07,
      "loss": 1.1401,
      "step": 1472
    },
    {
      "epoch": 0.19327538133508282,
      "grad_norm": 84.62464904785156,
      "learning_rate": 1.6199717790145174e-07,
      "loss": 1.1067,
      "step": 1473
    },
    {
      "epoch": 0.1934065934065934,
      "grad_norm": 88.183837890625,
      "learning_rate": 1.5022255497962879e-07,
      "loss": 1.0753,
      "step": 1474
    },
    {
      "epoch": 0.19353780547810398,
      "grad_norm": 66.31878662109375,
      "learning_rate": 1.3889182081860962e-07,
      "loss": 1.0303,
      "step": 1475
    },
    {
      "epoch": 0.19366901754961457,
      "grad_norm": 79.98758697509766,
      "learning_rate": 1.2800502578991235e-07,
      "loss": 1.0929,
      "step": 1476
    },
    {
      "epoch": 0.19380022962112514,
      "grad_norm": 80.35233306884766,
      "learning_rate": 1.1756221829148928e-07,
      "loss": 1.0508,
      "step": 1477
    },
    {
      "epoch": 0.1939314416926357,
      "grad_norm": 80.60159301757812,
      "learning_rate": 1.0756344474753821e-07,
      "loss": 1.1465,
      "step": 1478
    },
    {
      "epoch": 0.1940626537641463,
      "grad_norm": 67.90547180175781,
      "learning_rate": 9.800874960826933e-08,
      "loss": 0.9428,
      "step": 1479
    },
    {
      "epoch": 0.19419386583565687,
      "grad_norm": 71.87849426269531,
      "learning_rate": 8.889817534969425e-08,
      "loss": 1.1096,
      "step": 1480
    },
    {
      "epoch": 0.19432507790716746,
      "grad_norm": 67.80226135253906,
      "learning_rate": 8.023176247348163e-08,
      "loss": 0.9938,
      "step": 1481
    },
    {
      "epoch": 0.19445628997867803,
      "grad_norm": 72.56398010253906,
      "learning_rate": 7.200954950673522e-08,
      "loss": 1.1022,
      "step": 1482
    },
    {
      "epoch": 0.19458750205018863,
      "grad_norm": 76.51422882080078,
      "learning_rate": 6.423157300184946e-08,
      "loss": 1.152,
      "step": 1483
    },
    {
      "epoch": 0.1947187141216992,
      "grad_norm": 74.26021575927734,
      "learning_rate": 5.6897867536331864e-08,
      "loss": 1.1147,
      "step": 1484
    },
    {
      "epoch": 0.1948499261932098,
      "grad_norm": 75.1959228515625,
      "learning_rate": 5.000846571264761e-08,
      "loss": 1.1507,
      "step": 1485
    },
    {
      "epoch": 0.19498113826472036,
      "grad_norm": 64.08539581298828,
      "learning_rate": 4.35633981580974e-08,
      "loss": 0.7566,
      "step": 1486
    },
    {
      "epoch": 0.19511235033623092,
      "grad_norm": 115.13754272460938,
      "learning_rate": 3.756269352462871e-08,
      "loss": 1.2823,
      "step": 1487
    },
    {
      "epoch": 0.19524356240774152,
      "grad_norm": 77.61669921875,
      "learning_rate": 3.20063784888025e-08,
      "loss": 1.0631,
      "step": 1488
    },
    {
      "epoch": 0.19537477447925208,
      "grad_norm": 71.88060760498047,
      "learning_rate": 2.6894477751548964e-08,
      "loss": 0.9842,
      "step": 1489
    },
    {
      "epoch": 0.19550598655076268,
      "grad_norm": 81.05320739746094,
      "learning_rate": 2.222701403818972e-08,
      "loss": 1.0942,
      "step": 1490
    },
    {
      "epoch": 0.19563719862227325,
      "grad_norm": 81.63275146484375,
      "learning_rate": 1.8004008098226887e-08,
      "loss": 1.071,
      "step": 1491
    },
    {
      "epoch": 0.19576841069378384,
      "grad_norm": 76.47649383544922,
      "learning_rate": 1.4225478705309769e-08,
      "loss": 1.1566,
      "step": 1492
    },
    {
      "epoch": 0.1958996227652944,
      "grad_norm": 80.46495056152344,
      "learning_rate": 1.0891442657134932e-08,
      "loss": 1.1781,
      "step": 1493
    },
    {
      "epoch": 0.19603083483680497,
      "grad_norm": 85.7220458984375,
      "learning_rate": 8.001914775401798e-09,
      "loss": 1.2146,
      "step": 1494
    },
    {
      "epoch": 0.19616204690831557,
      "grad_norm": 73.49262237548828,
      "learning_rate": 5.5569079056794206e-09,
      "loss": 1.006,
      "step": 1495
    },
    {
      "epoch": 0.19629325897982614,
      "grad_norm": 80.52070617675781,
      "learning_rate": 3.5564329174064824e-09,
      "loss": 1.254,
      "step": 1496
    },
    {
      "epoch": 0.19642447105133673,
      "grad_norm": 85.13038635253906,
      "learning_rate": 2.0004987038246824e-09,
      "loss": 1.174,
      "step": 1497
    },
    {
      "epoch": 0.1965556831228473,
      "grad_norm": 70.47352600097656,
      "learning_rate": 8.891121819565306e-10,
      "loss": 0.8638,
      "step": 1498
    },
    {
      "epoch": 0.1966868951943579,
      "grad_norm": 90.5329818725586,
      "learning_rate": 2.2227829252763344e-10,
      "loss": 1.2457,
      "step": 1499
    },
    {
      "epoch": 0.19681810726586846,
      "grad_norm": 84.50530242919922,
      "learning_rate": 0.0,
      "loss": 0.9787,
      "step": 1500
    },
    {
      "epoch": 0.19681810726586846,
      "eval_loss": 1.102668046951294,
      "eval_runtime": 760.3463,
      "eval_samples_per_second": 16.882,
      "eval_steps_per_second": 8.441,
      "step": 1500
    }
  ],
  "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": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2.1456139615280824e+18,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}