{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 1623,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0030807147258163892,
      "grad_norm": 1.9506609439849854,
      "learning_rate": 9.756097560975611e-06,
      "loss": 1.0043,
      "step": 5
    },
    {
      "epoch": 0.0061614294516327784,
      "grad_norm": 1.3290749788284302,
      "learning_rate": 2.1951219512195124e-05,
      "loss": 1.0881,
      "step": 10
    },
    {
      "epoch": 0.009242144177449169,
      "grad_norm": 1.0360522270202637,
      "learning_rate": 3.414634146341464e-05,
      "loss": 1.1051,
      "step": 15
    },
    {
      "epoch": 0.012322858903265557,
      "grad_norm": 1.3245832920074463,
      "learning_rate": 4.634146341463415e-05,
      "loss": 0.9605,
      "step": 20
    },
    {
      "epoch": 0.015403573629081947,
      "grad_norm": 3.1614928245544434,
      "learning_rate": 5.853658536585366e-05,
      "loss": 0.8698,
      "step": 25
    },
    {
      "epoch": 0.018484288354898338,
      "grad_norm": 1.4028220176696777,
      "learning_rate": 7.073170731707317e-05,
      "loss": 0.8835,
      "step": 30
    },
    {
      "epoch": 0.021565003080714726,
      "grad_norm": 0.9983139634132385,
      "learning_rate": 8.292682926829268e-05,
      "loss": 0.9123,
      "step": 35
    },
    {
      "epoch": 0.024645717806531114,
      "grad_norm": 1.0538513660430908,
      "learning_rate": 9.51219512195122e-05,
      "loss": 0.885,
      "step": 40
    },
    {
      "epoch": 0.027726432532347505,
      "grad_norm": 1.131333827972412,
      "learning_rate": 0.00010731707317073172,
      "loss": 0.8446,
      "step": 45
    },
    {
      "epoch": 0.030807147258163893,
      "grad_norm": 2.132134199142456,
      "learning_rate": 0.00011951219512195122,
      "loss": 0.7956,
      "step": 50
    },
    {
      "epoch": 0.033887861983980284,
      "grad_norm": 1.0456063747406006,
      "learning_rate": 0.00013170731707317076,
      "loss": 0.8816,
      "step": 55
    },
    {
      "epoch": 0.036968576709796676,
      "grad_norm": 0.8690502643585205,
      "learning_rate": 0.00014390243902439025,
      "loss": 0.8685,
      "step": 60
    },
    {
      "epoch": 0.04004929143561306,
      "grad_norm": 1.0242969989776611,
      "learning_rate": 0.00015609756097560978,
      "loss": 0.8768,
      "step": 65
    },
    {
      "epoch": 0.04313000616142945,
      "grad_norm": 1.328539490699768,
      "learning_rate": 0.00016829268292682927,
      "loss": 0.9351,
      "step": 70
    },
    {
      "epoch": 0.04621072088724584,
      "grad_norm": 1.3852390050888062,
      "learning_rate": 0.0001804878048780488,
      "loss": 0.786,
      "step": 75
    },
    {
      "epoch": 0.04929143561306223,
      "grad_norm": 0.9410791397094727,
      "learning_rate": 0.0001926829268292683,
      "loss": 0.8309,
      "step": 80
    },
    {
      "epoch": 0.05237215033887862,
      "grad_norm": 0.8410763144493103,
      "learning_rate": 0.0001999991687649223,
      "loss": 0.9027,
      "step": 85
    },
    {
      "epoch": 0.05545286506469501,
      "grad_norm": 0.915637195110321,
      "learning_rate": 0.00019998981752900036,
      "loss": 0.9057,
      "step": 90
    },
    {
      "epoch": 0.0585335797905114,
      "grad_norm": 1.2468522787094116,
      "learning_rate": 0.00019997007698817557,
      "loss": 0.9095,
      "step": 95
    },
    {
      "epoch": 0.061614294516327786,
      "grad_norm": 1.3756437301635742,
      "learning_rate": 0.00019993994919356167,
      "loss": 0.8025,
      "step": 100
    },
    {
      "epoch": 0.06469500924214418,
      "grad_norm": 0.8679229021072388,
      "learning_rate": 0.00019989943727554598,
      "loss": 0.8418,
      "step": 105
    },
    {
      "epoch": 0.06777572396796057,
      "grad_norm": 0.8622983694076538,
      "learning_rate": 0.00019984854544346367,
      "loss": 0.8821,
      "step": 110
    },
    {
      "epoch": 0.07085643869377696,
      "grad_norm": 0.8575367331504822,
      "learning_rate": 0.00019978727898516086,
      "loss": 0.9431,
      "step": 115
    },
    {
      "epoch": 0.07393715341959335,
      "grad_norm": 1.2082788944244385,
      "learning_rate": 0.0001997156442664449,
      "loss": 0.8149,
      "step": 120
    },
    {
      "epoch": 0.07701786814540973,
      "grad_norm": 1.507377028465271,
      "learning_rate": 0.00019963364873042298,
      "loss": 0.7926,
      "step": 125
    },
    {
      "epoch": 0.08009858287122612,
      "grad_norm": 0.8779996037483215,
      "learning_rate": 0.0001995413008967289,
      "loss": 0.923,
      "step": 130
    },
    {
      "epoch": 0.08317929759704251,
      "grad_norm": 0.8468291163444519,
      "learning_rate": 0.00019943861036063768,
      "loss": 0.8893,
      "step": 135
    },
    {
      "epoch": 0.0862600123228589,
      "grad_norm": 0.7504271864891052,
      "learning_rate": 0.00019932558779206874,
      "loss": 0.8932,
      "step": 140
    },
    {
      "epoch": 0.0893407270486753,
      "grad_norm": 1.057389736175537,
      "learning_rate": 0.00019920224493447702,
      "loss": 0.8907,
      "step": 145
    },
    {
      "epoch": 0.09242144177449169,
      "grad_norm": 1.2131892442703247,
      "learning_rate": 0.00019906859460363307,
      "loss": 0.7649,
      "step": 150
    },
    {
      "epoch": 0.09550215650030808,
      "grad_norm": 0.7876495718955994,
      "learning_rate": 0.00019892465068629131,
      "loss": 0.8601,
      "step": 155
    },
    {
      "epoch": 0.09858287122612445,
      "grad_norm": 0.7372773885726929,
      "learning_rate": 0.0001987704281387471,
      "loss": 0.9682,
      "step": 160
    },
    {
      "epoch": 0.10166358595194085,
      "grad_norm": 0.9224637150764465,
      "learning_rate": 0.00019860594298528282,
      "loss": 0.8882,
      "step": 165
    },
    {
      "epoch": 0.10474430067775724,
      "grad_norm": 1.195654034614563,
      "learning_rate": 0.0001984312123165028,
      "loss": 0.9287,
      "step": 170
    },
    {
      "epoch": 0.10782501540357363,
      "grad_norm": 1.0784906148910522,
      "learning_rate": 0.0001982462542875576,
      "loss": 0.7491,
      "step": 175
    },
    {
      "epoch": 0.11090573012939002,
      "grad_norm": 0.6587386131286621,
      "learning_rate": 0.00019805108811625773,
      "loss": 0.8577,
      "step": 180
    },
    {
      "epoch": 0.11398644485520641,
      "grad_norm": 0.699715256690979,
      "learning_rate": 0.00019784573408107657,
      "loss": 0.8966,
      "step": 185
    },
    {
      "epoch": 0.1170671595810228,
      "grad_norm": 0.9360162615776062,
      "learning_rate": 0.00019763021351904358,
      "loss": 0.8773,
      "step": 190
    },
    {
      "epoch": 0.12014787430683918,
      "grad_norm": 1.123854160308838,
      "learning_rate": 0.00019740454882352732,
      "loss": 0.8704,
      "step": 195
    },
    {
      "epoch": 0.12322858903265557,
      "grad_norm": 1.2856158018112183,
      "learning_rate": 0.0001971687634419086,
      "loss": 0.7832,
      "step": 200
    },
    {
      "epoch": 0.12630930375847196,
      "grad_norm": 0.7385960221290588,
      "learning_rate": 0.0001969228818731442,
      "loss": 0.8582,
      "step": 205
    },
    {
      "epoch": 0.12939001848428835,
      "grad_norm": 0.6473488211631775,
      "learning_rate": 0.00019666692966522145,
      "loss": 0.8792,
      "step": 210
    },
    {
      "epoch": 0.13247073321010475,
      "grad_norm": 0.746126115322113,
      "learning_rate": 0.00019640093341250357,
      "loss": 0.8736,
      "step": 215
    },
    {
      "epoch": 0.13555144793592114,
      "grad_norm": 1.0111432075500488,
      "learning_rate": 0.0001961249207529665,
      "loss": 0.8853,
      "step": 220
    },
    {
      "epoch": 0.13863216266173753,
      "grad_norm": 1.2317752838134766,
      "learning_rate": 0.00019583892036532726,
      "loss": 0.7865,
      "step": 225
    },
    {
      "epoch": 0.14171287738755392,
      "grad_norm": 0.6940191388130188,
      "learning_rate": 0.00019554296196606395,
      "loss": 0.8703,
      "step": 230
    },
    {
      "epoch": 0.1447935921133703,
      "grad_norm": 0.7127991318702698,
      "learning_rate": 0.00019523707630632835,
      "loss": 0.8262,
      "step": 235
    },
    {
      "epoch": 0.1478743068391867,
      "grad_norm": 0.6165663003921509,
      "learning_rate": 0.00019492129516875055,
      "loss": 0.9039,
      "step": 240
    },
    {
      "epoch": 0.15095502156500307,
      "grad_norm": 1.2010482549667358,
      "learning_rate": 0.00019459565136413666,
      "loss": 0.9394,
      "step": 245
    },
    {
      "epoch": 0.15403573629081946,
      "grad_norm": 1.4223132133483887,
      "learning_rate": 0.0001942601787280598,
      "loss": 0.7718,
      "step": 250
    },
    {
      "epoch": 0.15711645101663585,
      "grad_norm": 0.8233282566070557,
      "learning_rate": 0.00019391491211734425,
      "loss": 0.8706,
      "step": 255
    },
    {
      "epoch": 0.16019716574245224,
      "grad_norm": 0.6466130018234253,
      "learning_rate": 0.0001935598874064438,
      "loss": 0.8579,
      "step": 260
    },
    {
      "epoch": 0.16327788046826863,
      "grad_norm": 0.7207822799682617,
      "learning_rate": 0.00019319514148371435,
      "loss": 0.7989,
      "step": 265
    },
    {
      "epoch": 0.16635859519408502,
      "grad_norm": 1.073905110359192,
      "learning_rate": 0.00019282071224758091,
      "loss": 0.8333,
      "step": 270
    },
    {
      "epoch": 0.16943930991990142,
      "grad_norm": 1.2006767988204956,
      "learning_rate": 0.00019243663860259993,
      "loss": 0.7993,
      "step": 275
    },
    {
      "epoch": 0.1725200246457178,
      "grad_norm": 0.6434801816940308,
      "learning_rate": 0.00019204296045541685,
      "loss": 0.8851,
      "step": 280
    },
    {
      "epoch": 0.1756007393715342,
      "grad_norm": 0.6209431290626526,
      "learning_rate": 0.0001916397187106199,
      "loss": 0.8257,
      "step": 285
    },
    {
      "epoch": 0.1786814540973506,
      "grad_norm": 0.5881760120391846,
      "learning_rate": 0.00019122695526648968,
      "loss": 0.8571,
      "step": 290
    },
    {
      "epoch": 0.18176216882316698,
      "grad_norm": 1.2488080263137817,
      "learning_rate": 0.00019080471301064598,
      "loss": 0.895,
      "step": 295
    },
    {
      "epoch": 0.18484288354898337,
      "grad_norm": 1.179758071899414,
      "learning_rate": 0.00019037303581559143,
      "loss": 0.7412,
      "step": 300
    },
    {
      "epoch": 0.18792359827479976,
      "grad_norm": 0.6266173720359802,
      "learning_rate": 0.00018993196853415317,
      "loss": 0.8424,
      "step": 305
    },
    {
      "epoch": 0.19100431300061615,
      "grad_norm": 0.781024694442749,
      "learning_rate": 0.00018948155699482244,
      "loss": 0.818,
      "step": 310
    },
    {
      "epoch": 0.19408502772643252,
      "grad_norm": 0.857995331287384,
      "learning_rate": 0.00018902184799699263,
      "loss": 0.9148,
      "step": 315
    },
    {
      "epoch": 0.1971657424522489,
      "grad_norm": 0.7764838933944702,
      "learning_rate": 0.00018855288930609692,
      "loss": 0.7969,
      "step": 320
    },
    {
      "epoch": 0.2002464571780653,
      "grad_norm": 0.9209476113319397,
      "learning_rate": 0.00018807472964864515,
      "loss": 0.7768,
      "step": 325
    },
    {
      "epoch": 0.2033271719038817,
      "grad_norm": 0.6861506104469299,
      "learning_rate": 0.00018758741870716092,
      "loss": 0.8849,
      "step": 330
    },
    {
      "epoch": 0.20640788662969808,
      "grad_norm": 0.5947197675704956,
      "learning_rate": 0.00018709100711501955,
      "loss": 0.8672,
      "step": 335
    },
    {
      "epoch": 0.20948860135551448,
      "grad_norm": 0.6780321002006531,
      "learning_rate": 0.0001865855464511869,
      "loss": 0.7919,
      "step": 340
    },
    {
      "epoch": 0.21256931608133087,
      "grad_norm": 1.1705702543258667,
      "learning_rate": 0.00018607108923486025,
      "loss": 0.7502,
      "step": 345
    },
    {
      "epoch": 0.21565003080714726,
      "grad_norm": 1.004110336303711,
      "learning_rate": 0.00018554768892001136,
      "loss": 0.7369,
      "step": 350
    },
    {
      "epoch": 0.21873074553296365,
      "grad_norm": 0.6980533003807068,
      "learning_rate": 0.00018501539988983234,
      "loss": 0.7703,
      "step": 355
    },
    {
      "epoch": 0.22181146025878004,
      "grad_norm": 0.6246406435966492,
      "learning_rate": 0.0001844742774510851,
      "loss": 0.8441,
      "step": 360
    },
    {
      "epoch": 0.22489217498459643,
      "grad_norm": 0.5528222322463989,
      "learning_rate": 0.00018392437782835475,
      "loss": 0.8385,
      "step": 365
    },
    {
      "epoch": 0.22797288971041282,
      "grad_norm": 0.9303919076919556,
      "learning_rate": 0.00018336575815820766,
      "loss": 0.8384,
      "step": 370
    },
    {
      "epoch": 0.23105360443622922,
      "grad_norm": 1.3066872358322144,
      "learning_rate": 0.00018279847648325478,
      "loss": 0.7767,
      "step": 375
    },
    {
      "epoch": 0.2341343191620456,
      "grad_norm": 0.7670502662658691,
      "learning_rate": 0.0001822225917461208,
      "loss": 0.8032,
      "step": 380
    },
    {
      "epoch": 0.23721503388786197,
      "grad_norm": 0.7942723631858826,
      "learning_rate": 0.0001816381637833198,
      "loss": 0.8288,
      "step": 385
    },
    {
      "epoch": 0.24029574861367836,
      "grad_norm": 0.6688534617424011,
      "learning_rate": 0.00018104525331903799,
      "loss": 0.8631,
      "step": 390
    },
    {
      "epoch": 0.24337646333949475,
      "grad_norm": 0.9976694583892822,
      "learning_rate": 0.00018044392195882427,
      "loss": 0.9414,
      "step": 395
    },
    {
      "epoch": 0.24645717806531114,
      "grad_norm": 1.142532229423523,
      "learning_rate": 0.00017983423218318918,
      "loss": 0.7797,
      "step": 400
    },
    {
      "epoch": 0.24953789279112754,
      "grad_norm": 0.6933532953262329,
      "learning_rate": 0.00017921624734111292,
      "loss": 0.8074,
      "step": 405
    },
    {
      "epoch": 0.2526186075169439,
      "grad_norm": 0.7068799138069153,
      "learning_rate": 0.00017859003164346336,
      "loss": 0.9038,
      "step": 410
    },
    {
      "epoch": 0.2556993222427603,
      "grad_norm": 0.7113360166549683,
      "learning_rate": 0.0001779556501563239,
      "loss": 0.8077,
      "step": 415
    },
    {
      "epoch": 0.2587800369685767,
      "grad_norm": 1.0284174680709839,
      "learning_rate": 0.00017731316879423327,
      "loss": 0.9117,
      "step": 420
    },
    {
      "epoch": 0.2618607516943931,
      "grad_norm": 1.0717129707336426,
      "learning_rate": 0.00017666265431333654,
      "loss": 0.8241,
      "step": 425
    },
    {
      "epoch": 0.2649414664202095,
      "grad_norm": 0.6098302006721497,
      "learning_rate": 0.000176004174304449,
      "loss": 0.8526,
      "step": 430
    },
    {
      "epoch": 0.2680221811460259,
      "grad_norm": 0.6730021834373474,
      "learning_rate": 0.00017533779718603313,
      "loss": 0.8473,
      "step": 435
    },
    {
      "epoch": 0.2711028958718423,
      "grad_norm": 0.627232551574707,
      "learning_rate": 0.00017466359219708985,
      "loss": 0.7787,
      "step": 440
    },
    {
      "epoch": 0.27418361059765867,
      "grad_norm": 0.7510082125663757,
      "learning_rate": 0.00017398162938996422,
      "loss": 0.7869,
      "step": 445
    },
    {
      "epoch": 0.27726432532347506,
      "grad_norm": 0.800914466381073,
      "learning_rate": 0.00017329197962306664,
      "loss": 0.7307,
      "step": 450
    },
    {
      "epoch": 0.28034504004929145,
      "grad_norm": 0.688615083694458,
      "learning_rate": 0.00017259471455351072,
      "loss": 0.8078,
      "step": 455
    },
    {
      "epoch": 0.28342575477510784,
      "grad_norm": 0.6369125247001648,
      "learning_rate": 0.0001718899066296675,
      "loss": 0.827,
      "step": 460
    },
    {
      "epoch": 0.28650646950092423,
      "grad_norm": 0.7632527351379395,
      "learning_rate": 0.000171177629083638,
      "loss": 0.8171,
      "step": 465
    },
    {
      "epoch": 0.2895871842267406,
      "grad_norm": 0.8901572227478027,
      "learning_rate": 0.0001704579559236441,
      "loss": 0.8534,
      "step": 470
    },
    {
      "epoch": 0.292667898952557,
      "grad_norm": 1.2739449739456177,
      "learning_rate": 0.00016973096192633884,
      "loss": 0.718,
      "step": 475
    },
    {
      "epoch": 0.2957486136783734,
      "grad_norm": 0.655415952205658,
      "learning_rate": 0.00016899672262903677,
      "loss": 0.7889,
      "step": 480
    },
    {
      "epoch": 0.2988293284041898,
      "grad_norm": 0.6401821970939636,
      "learning_rate": 0.00016825531432186543,
      "loss": 0.9173,
      "step": 485
    },
    {
      "epoch": 0.30191004313000613,
      "grad_norm": 0.5191354155540466,
      "learning_rate": 0.00016750681403983846,
      "loss": 0.931,
      "step": 490
    },
    {
      "epoch": 0.3049907578558225,
      "grad_norm": 0.8743457794189453,
      "learning_rate": 0.00016675129955485152,
      "loss": 0.8111,
      "step": 495
    },
    {
      "epoch": 0.3080714725816389,
      "grad_norm": 1.0666691064834595,
      "learning_rate": 0.00016598884936760131,
      "loss": 0.736,
      "step": 500
    },
    {
      "epoch": 0.3111521873074553,
      "grad_norm": 0.606925368309021,
      "learning_rate": 0.00016521954269942918,
      "loss": 0.8742,
      "step": 505
    },
    {
      "epoch": 0.3142329020332717,
      "grad_norm": 0.5690024495124817,
      "learning_rate": 0.00016444345948408984,
      "loss": 0.8288,
      "step": 510
    },
    {
      "epoch": 0.3173136167590881,
      "grad_norm": 0.7045788168907166,
      "learning_rate": 0.0001636606803594457,
      "loss": 0.8422,
      "step": 515
    },
    {
      "epoch": 0.3203943314849045,
      "grad_norm": 0.9720426201820374,
      "learning_rate": 0.0001628712866590885,
      "loss": 0.8012,
      "step": 520
    },
    {
      "epoch": 0.3234750462107209,
      "grad_norm": 1.1168466806411743,
      "learning_rate": 0.00016207536040388845,
      "loss": 0.7338,
      "step": 525
    },
    {
      "epoch": 0.32655576093653726,
      "grad_norm": 0.6451901197433472,
      "learning_rate": 0.0001612729842934718,
      "loss": 0.8471,
      "step": 530
    },
    {
      "epoch": 0.32963647566235366,
      "grad_norm": 0.7933263778686523,
      "learning_rate": 0.00016046424169762827,
      "loss": 0.8995,
      "step": 535
    },
    {
      "epoch": 0.33271719038817005,
      "grad_norm": 0.6123836636543274,
      "learning_rate": 0.0001596492166476485,
      "loss": 0.8341,
      "step": 540
    },
    {
      "epoch": 0.33579790511398644,
      "grad_norm": 0.8381984233856201,
      "learning_rate": 0.0001588279938275929,
      "loss": 0.8531,
      "step": 545
    },
    {
      "epoch": 0.33887861983980283,
      "grad_norm": 1.155434250831604,
      "learning_rate": 0.00015800065856549269,
      "loss": 0.6915,
      "step": 550
    },
    {
      "epoch": 0.3419593345656192,
      "grad_norm": 0.634437084197998,
      "learning_rate": 0.00015716729682448393,
      "loss": 0.8241,
      "step": 555
    },
    {
      "epoch": 0.3450400492914356,
      "grad_norm": 0.6219022274017334,
      "learning_rate": 0.0001563279951938758,
      "loss": 0.8461,
      "step": 560
    },
    {
      "epoch": 0.348120764017252,
      "grad_norm": 0.4845116138458252,
      "learning_rate": 0.00015548284088015354,
      "loss": 0.8311,
      "step": 565
    },
    {
      "epoch": 0.3512014787430684,
      "grad_norm": 0.9370896816253662,
      "learning_rate": 0.00015463192169791741,
      "loss": 0.8278,
      "step": 570
    },
    {
      "epoch": 0.3542821934688848,
      "grad_norm": 1.1633367538452148,
      "learning_rate": 0.0001537753260607584,
      "loss": 0.7536,
      "step": 575
    },
    {
      "epoch": 0.3573629081947012,
      "grad_norm": 0.5799803733825684,
      "learning_rate": 0.00015291314297207175,
      "loss": 0.7999,
      "step": 580
    },
    {
      "epoch": 0.36044362292051757,
      "grad_norm": 0.47321441769599915,
      "learning_rate": 0.0001520454620158093,
      "loss": 0.8864,
      "step": 585
    },
    {
      "epoch": 0.36352433764633396,
      "grad_norm": 0.5589901804924011,
      "learning_rate": 0.00015117237334717117,
      "loss": 0.7921,
      "step": 590
    },
    {
      "epoch": 0.36660505237215035,
      "grad_norm": 0.9199019074440002,
      "learning_rate": 0.00015029396768323846,
      "loss": 0.8999,
      "step": 595
    },
    {
      "epoch": 0.36968576709796674,
      "grad_norm": 1.062154769897461,
      "learning_rate": 0.00014941033629354734,
      "loss": 0.815,
      "step": 600
    },
    {
      "epoch": 0.37276648182378314,
      "grad_norm": 0.7027643918991089,
      "learning_rate": 0.00014852157099060596,
      "loss": 0.8644,
      "step": 605
    },
    {
      "epoch": 0.3758471965495995,
      "grad_norm": 0.5963114500045776,
      "learning_rate": 0.00014762776412035456,
      "loss": 0.87,
      "step": 610
    },
    {
      "epoch": 0.3789279112754159,
      "grad_norm": 0.5938383340835571,
      "learning_rate": 0.00014672900855257056,
      "loss": 0.8137,
      "step": 615
    },
    {
      "epoch": 0.3820086260012323,
      "grad_norm": 0.9398343563079834,
      "learning_rate": 0.00014582539767121904,
      "loss": 0.8821,
      "step": 620
    },
    {
      "epoch": 0.3850893407270487,
      "grad_norm": 0.99173504114151,
      "learning_rate": 0.0001449170253647498,
      "loss": 0.6784,
      "step": 625
    },
    {
      "epoch": 0.38817005545286504,
      "grad_norm": 0.7030093669891357,
      "learning_rate": 0.0001440039860163419,
      "loss": 0.832,
      "step": 630
    },
    {
      "epoch": 0.39125077017868143,
      "grad_norm": 0.6641373038291931,
      "learning_rate": 0.00014308637449409706,
      "loss": 0.826,
      "step": 635
    },
    {
      "epoch": 0.3943314849044978,
      "grad_norm": 0.6074934005737305,
      "learning_rate": 0.00014216428614118243,
      "loss": 0.8566,
      "step": 640
    },
    {
      "epoch": 0.3974121996303142,
      "grad_norm": 0.8580813407897949,
      "learning_rate": 0.00014123781676592418,
      "loss": 0.8423,
      "step": 645
    },
    {
      "epoch": 0.4004929143561306,
      "grad_norm": 1.0865176916122437,
      "learning_rate": 0.00014030706263185247,
      "loss": 0.769,
      "step": 650
    },
    {
      "epoch": 0.403573629081947,
      "grad_norm": 0.6008116602897644,
      "learning_rate": 0.00013937212044769955,
      "loss": 0.8003,
      "step": 655
    },
    {
      "epoch": 0.4066543438077634,
      "grad_norm": 0.7620061635971069,
      "learning_rate": 0.0001384330873573513,
      "loss": 0.863,
      "step": 660
    },
    {
      "epoch": 0.4097350585335798,
      "grad_norm": 0.5859779119491577,
      "learning_rate": 0.00013749006092975347,
      "loss": 0.8224,
      "step": 665
    },
    {
      "epoch": 0.41281577325939617,
      "grad_norm": 0.8230900764465332,
      "learning_rate": 0.00013654313914877414,
      "loss": 0.8245,
      "step": 670
    },
    {
      "epoch": 0.41589648798521256,
      "grad_norm": 0.9619393348693848,
      "learning_rate": 0.00013559242040302272,
      "loss": 0.7234,
      "step": 675
    },
    {
      "epoch": 0.41897720271102895,
      "grad_norm": 0.687427282333374,
      "learning_rate": 0.00013463800347562706,
      "loss": 0.8489,
      "step": 680
    },
    {
      "epoch": 0.42205791743684534,
      "grad_norm": 0.5357842445373535,
      "learning_rate": 0.00013367998753396944,
      "loss": 0.755,
      "step": 685
    },
    {
      "epoch": 0.42513863216266173,
      "grad_norm": 0.5615360736846924,
      "learning_rate": 0.00013271847211938285,
      "loss": 0.8116,
      "step": 690
    },
    {
      "epoch": 0.4282193468884781,
      "grad_norm": 1.0073713064193726,
      "learning_rate": 0.0001317535571368082,
      "loss": 0.8125,
      "step": 695
    },
    {
      "epoch": 0.4313000616142945,
      "grad_norm": 0.9638437032699585,
      "learning_rate": 0.00013078534284441382,
      "loss": 0.7871,
      "step": 700
    },
    {
      "epoch": 0.4343807763401109,
      "grad_norm": 0.6262618899345398,
      "learning_rate": 0.00012981392984317834,
      "loss": 0.7716,
      "step": 705
    },
    {
      "epoch": 0.4374614910659273,
      "grad_norm": 0.6202210187911987,
      "learning_rate": 0.00012883941906643786,
      "loss": 0.8464,
      "step": 710
    },
    {
      "epoch": 0.4405422057917437,
      "grad_norm": 0.636193037033081,
      "learning_rate": 0.00012786191176939848,
      "loss": 0.7936,
      "step": 715
    },
    {
      "epoch": 0.4436229205175601,
      "grad_norm": 0.7928021550178528,
      "learning_rate": 0.00012688150951861582,
      "loss": 0.7535,
      "step": 720
    },
    {
      "epoch": 0.4467036352433765,
      "grad_norm": 1.1193562746047974,
      "learning_rate": 0.00012589831418144154,
      "loss": 0.7378,
      "step": 725
    },
    {
      "epoch": 0.44978434996919286,
      "grad_norm": 0.5501682162284851,
      "learning_rate": 0.00012491242791543922,
      "loss": 0.8251,
      "step": 730
    },
    {
      "epoch": 0.45286506469500926,
      "grad_norm": 0.5638311505317688,
      "learning_rate": 0.00012392395315776963,
      "loss": 0.8488,
      "step": 735
    },
    {
      "epoch": 0.45594577942082565,
      "grad_norm": 0.5718980431556702,
      "learning_rate": 0.00012293299261454725,
      "loss": 0.8058,
      "step": 740
    },
    {
      "epoch": 0.45902649414664204,
      "grad_norm": 0.9007247090339661,
      "learning_rate": 0.00012193964925016872,
      "loss": 0.7745,
      "step": 745
    },
    {
      "epoch": 0.46210720887245843,
      "grad_norm": 1.2628989219665527,
      "learning_rate": 0.00012094402627661447,
      "loss": 0.7156,
      "step": 750
    },
    {
      "epoch": 0.4651879235982748,
      "grad_norm": 0.6456303596496582,
      "learning_rate": 0.00011994622714272448,
      "loss": 0.8305,
      "step": 755
    },
    {
      "epoch": 0.4682686383240912,
      "grad_norm": 0.7610649466514587,
      "learning_rate": 0.00011894635552344975,
      "loss": 0.8419,
      "step": 760
    },
    {
      "epoch": 0.4713493530499076,
      "grad_norm": 0.415209025144577,
      "learning_rate": 0.00011794451530908011,
      "loss": 0.7674,
      "step": 765
    },
    {
      "epoch": 0.47443006777572394,
      "grad_norm": 0.8516692519187927,
      "learning_rate": 0.00011694081059444946,
      "loss": 0.8148,
      "step": 770
    },
    {
      "epoch": 0.47751078250154033,
      "grad_norm": 1.0319640636444092,
      "learning_rate": 0.0001159353456681201,
      "loss": 0.7435,
      "step": 775
    },
    {
      "epoch": 0.4805914972273567,
      "grad_norm": 0.5645594000816345,
      "learning_rate": 0.00011492822500154667,
      "loss": 0.7572,
      "step": 780
    },
    {
      "epoch": 0.4836722119531731,
      "grad_norm": 0.5694653987884521,
      "learning_rate": 0.00011391955323822126,
      "loss": 0.7624,
      "step": 785
    },
    {
      "epoch": 0.4867529266789895,
      "grad_norm": 0.5023643374443054,
      "learning_rate": 0.00011290943518280057,
      "loss": 0.8524,
      "step": 790
    },
    {
      "epoch": 0.4898336414048059,
      "grad_norm": 0.9564359784126282,
      "learning_rate": 0.0001118979757902162,
      "loss": 0.7575,
      "step": 795
    },
    {
      "epoch": 0.4929143561306223,
      "grad_norm": 1.1121801137924194,
      "learning_rate": 0.00011088528015476964,
      "loss": 0.7799,
      "step": 800
    },
    {
      "epoch": 0.4959950708564387,
      "grad_norm": 0.5354447364807129,
      "learning_rate": 0.00010987145349921251,
      "loss": 0.7722,
      "step": 805
    },
    {
      "epoch": 0.49907578558225507,
      "grad_norm": 0.5652374625205994,
      "learning_rate": 0.0001088566011638134,
      "loss": 0.7795,
      "step": 810
    },
    {
      "epoch": 0.5021565003080715,
      "grad_norm": 0.5385531783103943,
      "learning_rate": 0.00010784082859541292,
      "loss": 0.7919,
      "step": 815
    },
    {
      "epoch": 0.5052372150338879,
      "grad_norm": 0.7922360897064209,
      "learning_rate": 0.0001068242413364671,
      "loss": 0.8262,
      "step": 820
    },
    {
      "epoch": 0.5083179297597042,
      "grad_norm": 1.1030114889144897,
      "learning_rate": 0.00010580694501408138,
      "loss": 0.7565,
      "step": 825
    },
    {
      "epoch": 0.5113986444855206,
      "grad_norm": 0.5422970056533813,
      "learning_rate": 0.00010478904532903535,
      "loss": 0.7479,
      "step": 830
    },
    {
      "epoch": 0.514479359211337,
      "grad_norm": 0.5856009125709534,
      "learning_rate": 0.00010377064804480025,
      "loss": 0.8519,
      "step": 835
    },
    {
      "epoch": 0.5175600739371534,
      "grad_norm": 0.5326692461967468,
      "learning_rate": 0.00010275185897654971,
      "loss": 0.7604,
      "step": 840
    },
    {
      "epoch": 0.5206407886629698,
      "grad_norm": 0.9409604072570801,
      "learning_rate": 0.00010173278398016501,
      "loss": 0.7729,
      "step": 845
    },
    {
      "epoch": 0.5237215033887862,
      "grad_norm": 0.8584169149398804,
      "learning_rate": 0.00010071352894123654,
      "loss": 0.7233,
      "step": 850
    },
    {
      "epoch": 0.5268022181146026,
      "grad_norm": 0.5935032963752747,
      "learning_rate": 9.969419976406165e-05,
      "loss": 0.7798,
      "step": 855
    },
    {
      "epoch": 0.529882932840419,
      "grad_norm": 0.6444724202156067,
      "learning_rate": 9.867490236064108e-05,
      "loss": 0.7783,
      "step": 860
    },
    {
      "epoch": 0.5329636475662354,
      "grad_norm": 0.5421064496040344,
      "learning_rate": 9.765574263967396e-05,
      "loss": 0.8457,
      "step": 865
    },
    {
      "epoch": 0.5360443622920518,
      "grad_norm": 0.7332625985145569,
      "learning_rate": 9.66368264955539e-05,
      "loss": 0.7861,
      "step": 870
    },
    {
      "epoch": 0.5391250770178682,
      "grad_norm": 0.9407272338867188,
      "learning_rate": 9.56182597973658e-05,
      "loss": 0.738,
      "step": 875
    },
    {
      "epoch": 0.5422057917436846,
      "grad_norm": 0.5053693652153015,
      "learning_rate": 9.460014837788605e-05,
      "loss": 0.7868,
      "step": 880
    },
    {
      "epoch": 0.5452865064695009,
      "grad_norm": 0.5454622507095337,
      "learning_rate": 9.358259802258581e-05,
      "loss": 0.8042,
      "step": 885
    },
    {
      "epoch": 0.5483672211953173,
      "grad_norm": 0.5724400281906128,
      "learning_rate": 9.256571445863972e-05,
      "loss": 0.7704,
      "step": 890
    },
    {
      "epoch": 0.5514479359211337,
      "grad_norm": 0.843951404094696,
      "learning_rate": 9.154960334394027e-05,
      "loss": 0.8044,
      "step": 895
    },
    {
      "epoch": 0.5545286506469501,
      "grad_norm": 1.0994093418121338,
      "learning_rate": 9.053437025611973e-05,
      "loss": 0.7098,
      "step": 900
    },
    {
      "epoch": 0.5576093653727665,
      "grad_norm": 0.6525376439094543,
      "learning_rate": 8.952012068158027e-05,
      "loss": 0.8139,
      "step": 905
    },
    {
      "epoch": 0.5606900800985829,
      "grad_norm": 0.49181467294692993,
      "learning_rate": 8.850696000453326e-05,
      "loss": 0.8226,
      "step": 910
    },
    {
      "epoch": 0.5637707948243993,
      "grad_norm": 0.5223445296287537,
      "learning_rate": 8.749499349604993e-05,
      "loss": 0.7899,
      "step": 915
    },
    {
      "epoch": 0.5668515095502157,
      "grad_norm": 0.8057864308357239,
      "learning_rate": 8.64843263031228e-05,
      "loss": 0.7831,
      "step": 920
    },
    {
      "epoch": 0.5699322242760321,
      "grad_norm": 1.0692858695983887,
      "learning_rate": 8.547506343774097e-05,
      "loss": 0.6825,
      "step": 925
    },
    {
      "epoch": 0.5730129390018485,
      "grad_norm": 0.6106774806976318,
      "learning_rate": 8.446730976597878e-05,
      "loss": 0.8087,
      "step": 930
    },
    {
      "epoch": 0.5760936537276649,
      "grad_norm": 0.657316267490387,
      "learning_rate": 8.346116999709975e-05,
      "loss": 0.7957,
      "step": 935
    },
    {
      "epoch": 0.5791743684534812,
      "grad_norm": 0.4877719581127167,
      "learning_rate": 8.245674867267724e-05,
      "loss": 0.7564,
      "step": 940
    },
    {
      "epoch": 0.5822550831792976,
      "grad_norm": 0.8313435912132263,
      "learning_rate": 8.145415015573183e-05,
      "loss": 0.8325,
      "step": 945
    },
    {
      "epoch": 0.585335797905114,
      "grad_norm": 0.9311845302581787,
      "learning_rate": 8.045347861988789e-05,
      "loss": 0.6653,
      "step": 950
    },
    {
      "epoch": 0.5884165126309304,
      "grad_norm": 0.6919590830802917,
      "learning_rate": 7.945483803854936e-05,
      "loss": 0.7861,
      "step": 955
    },
    {
      "epoch": 0.5914972273567468,
      "grad_norm": 0.5540125370025635,
      "learning_rate": 7.845833217409675e-05,
      "loss": 0.8685,
      "step": 960
    },
    {
      "epoch": 0.5945779420825632,
      "grad_norm": 0.5159959197044373,
      "learning_rate": 7.746406456710564e-05,
      "loss": 0.7727,
      "step": 965
    },
    {
      "epoch": 0.5976586568083796,
      "grad_norm": 0.8416092395782471,
      "learning_rate": 7.64721385255886e-05,
      "loss": 0.7798,
      "step": 970
    },
    {
      "epoch": 0.600739371534196,
      "grad_norm": 1.0320320129394531,
      "learning_rate": 7.548265711426104e-05,
      "loss": 0.6814,
      "step": 975
    },
    {
      "epoch": 0.6038200862600123,
      "grad_norm": 0.626734733581543,
      "learning_rate": 7.449572314383237e-05,
      "loss": 0.8495,
      "step": 980
    },
    {
      "epoch": 0.6069008009858287,
      "grad_norm": 0.6575467586517334,
      "learning_rate": 7.351143916032374e-05,
      "loss": 0.8553,
      "step": 985
    },
    {
      "epoch": 0.609981515711645,
      "grad_norm": 0.5149129629135132,
      "learning_rate": 7.252990743441293e-05,
      "loss": 0.7646,
      "step": 990
    },
    {
      "epoch": 0.6130622304374614,
      "grad_norm": 0.8680890798568726,
      "learning_rate": 7.155122995080827e-05,
      "loss": 0.8524,
      "step": 995
    },
    {
      "epoch": 0.6161429451632778,
      "grad_norm": 1.129459023475647,
      "learning_rate": 7.057550839765188e-05,
      "loss": 0.7465,
      "step": 1000
    },
    {
      "epoch": 0.6192236598890942,
      "grad_norm": 0.7005685567855835,
      "learning_rate": 6.960284415595407e-05,
      "loss": 0.7945,
      "step": 1005
    },
    {
      "epoch": 0.6223043746149106,
      "grad_norm": 0.5934234261512756,
      "learning_rate": 6.863333828905929e-05,
      "loss": 0.9171,
      "step": 1010
    },
    {
      "epoch": 0.625385089340727,
      "grad_norm": 0.5575773119926453,
      "learning_rate": 6.766709153214542e-05,
      "loss": 0.8228,
      "step": 1015
    },
    {
      "epoch": 0.6284658040665434,
      "grad_norm": 0.7771855592727661,
      "learning_rate": 6.670420428175705e-05,
      "loss": 0.8146,
      "step": 1020
    },
    {
      "epoch": 0.6315465187923598,
      "grad_norm": 0.8821945190429688,
      "learning_rate": 6.574477658537375e-05,
      "loss": 0.7083,
      "step": 1025
    },
    {
      "epoch": 0.6346272335181762,
      "grad_norm": 0.6677148342132568,
      "learning_rate": 6.4788908131015e-05,
      "loss": 0.7567,
      "step": 1030
    },
    {
      "epoch": 0.6377079482439926,
      "grad_norm": 0.5348629951477051,
      "learning_rate": 6.38366982368819e-05,
      "loss": 0.7256,
      "step": 1035
    },
    {
      "epoch": 0.640788662969809,
      "grad_norm": 0.539318323135376,
      "learning_rate": 6.288824584103816e-05,
      "loss": 0.7838,
      "step": 1040
    },
    {
      "epoch": 0.6438693776956254,
      "grad_norm": 0.8003283739089966,
      "learning_rate": 6.194364949112953e-05,
      "loss": 0.7608,
      "step": 1045
    },
    {
      "epoch": 0.6469500924214417,
      "grad_norm": 0.9320285320281982,
      "learning_rate": 6.100300733414474e-05,
      "loss": 0.6823,
      "step": 1050
    },
    {
      "epoch": 0.6500308071472581,
      "grad_norm": 0.6213887333869934,
      "learning_rate": 6.0066417106217455e-05,
      "loss": 0.7781,
      "step": 1055
    },
    {
      "epoch": 0.6531115218730745,
      "grad_norm": 0.5485357046127319,
      "learning_rate": 5.9133976122471214e-05,
      "loss": 0.8137,
      "step": 1060
    },
    {
      "epoch": 0.6561922365988909,
      "grad_norm": 0.5012507438659668,
      "learning_rate": 5.82057812669081e-05,
      "loss": 0.7741,
      "step": 1065
    },
    {
      "epoch": 0.6592729513247073,
      "grad_norm": 0.6728748083114624,
      "learning_rate": 5.728192898234195e-05,
      "loss": 0.7326,
      "step": 1070
    },
    {
      "epoch": 0.6623536660505237,
      "grad_norm": 1.107875943183899,
      "learning_rate": 5.6362515260377835e-05,
      "loss": 0.676,
      "step": 1075
    },
    {
      "epoch": 0.6654343807763401,
      "grad_norm": 0.6420050263404846,
      "learning_rate": 5.544763563143793e-05,
      "loss": 0.8215,
      "step": 1080
    },
    {
      "epoch": 0.6685150955021565,
      "grad_norm": 0.5939176082611084,
      "learning_rate": 5.4537385154835864e-05,
      "loss": 0.7669,
      "step": 1085
    },
    {
      "epoch": 0.6715958102279729,
      "grad_norm": 0.4939870536327362,
      "learning_rate": 5.363185840889935e-05,
      "loss": 0.7439,
      "step": 1090
    },
    {
      "epoch": 0.6746765249537893,
      "grad_norm": 0.7583256363868713,
      "learning_rate": 5.273114948114346e-05,
      "loss": 0.81,
      "step": 1095
    },
    {
      "epoch": 0.6777572396796057,
      "grad_norm": 1.1161051988601685,
      "learning_rate": 5.1835351958494515e-05,
      "loss": 0.683,
      "step": 1100
    },
    {
      "epoch": 0.680837954405422,
      "grad_norm": 0.5778560042381287,
      "learning_rate": 5.094455891756587e-05,
      "loss": 0.7839,
      "step": 1105
    },
    {
      "epoch": 0.6839186691312384,
      "grad_norm": 0.5644716024398804,
      "learning_rate": 5.00588629149872e-05,
      "loss": 0.8279,
      "step": 1110
    },
    {
      "epoch": 0.6869993838570548,
      "grad_norm": 0.525404691696167,
      "learning_rate": 4.91783559777873e-05,
      "loss": 0.7721,
      "step": 1115
    },
    {
      "epoch": 0.6900800985828712,
      "grad_norm": 0.9260256290435791,
      "learning_rate": 4.830312959383238e-05,
      "loss": 0.7555,
      "step": 1120
    },
    {
      "epoch": 0.6931608133086876,
      "grad_norm": 0.8119556903839111,
      "learning_rate": 4.7433274702319815e-05,
      "loss": 0.6236,
      "step": 1125
    },
    {
      "epoch": 0.696241528034504,
      "grad_norm": 0.6038579344749451,
      "learning_rate": 4.656888168432962e-05,
      "loss": 0.7604,
      "step": 1130
    },
    {
      "epoch": 0.6993222427603204,
      "grad_norm": 0.6174635887145996,
      "learning_rate": 4.571004035343315e-05,
      "loss": 0.8142,
      "step": 1135
    },
    {
      "epoch": 0.7024029574861368,
      "grad_norm": 0.470345139503479,
      "learning_rate": 4.485683994636144e-05,
      "loss": 0.7306,
      "step": 1140
    },
    {
      "epoch": 0.7054836722119532,
      "grad_norm": 0.8356139659881592,
      "learning_rate": 4.400936911373308e-05,
      "loss": 0.8464,
      "step": 1145
    },
    {
      "epoch": 0.7085643869377696,
      "grad_norm": 0.8929637670516968,
      "learning_rate": 4.3167715910842966e-05,
      "loss": 0.6958,
      "step": 1150
    },
    {
      "epoch": 0.711645101663586,
      "grad_norm": 0.5765171051025391,
      "learning_rate": 4.2331967788513295e-05,
      "loss": 0.7714,
      "step": 1155
    },
    {
      "epoch": 0.7147258163894024,
      "grad_norm": 0.5743905901908875,
      "learning_rate": 4.1502211584006836e-05,
      "loss": 0.7777,
      "step": 1160
    },
    {
      "epoch": 0.7178065311152187,
      "grad_norm": 0.47613048553466797,
      "learning_rate": 4.067853351200446e-05,
      "loss": 0.8229,
      "step": 1165
    },
    {
      "epoch": 0.7208872458410351,
      "grad_norm": 0.6945850849151611,
      "learning_rate": 3.986101915564695e-05,
      "loss": 0.7934,
      "step": 1170
    },
    {
      "epoch": 0.7239679605668515,
      "grad_norm": 1.2221349477767944,
      "learning_rate": 3.904975345764262e-05,
      "loss": 0.7063,
      "step": 1175
    },
    {
      "epoch": 0.7270486752926679,
      "grad_norm": 0.6141952872276306,
      "learning_rate": 3.824482071144163e-05,
      "loss": 0.7582,
      "step": 1180
    },
    {
      "epoch": 0.7301293900184843,
      "grad_norm": 0.6697849035263062,
      "learning_rate": 3.744630455247739e-05,
      "loss": 0.7688,
      "step": 1185
    },
    {
      "epoch": 0.7332101047443007,
      "grad_norm": 0.45467403531074524,
      "learning_rate": 3.6654287949476626e-05,
      "loss": 0.7579,
      "step": 1190
    },
    {
      "epoch": 0.7362908194701171,
      "grad_norm": 0.7754799127578735,
      "learning_rate": 3.586885319583858e-05,
      "loss": 0.7618,
      "step": 1195
    },
    {
      "epoch": 0.7393715341959335,
      "grad_norm": 1.0018911361694336,
      "learning_rate": 3.5090081901084525e-05,
      "loss": 0.6976,
      "step": 1200
    },
    {
      "epoch": 0.7424522489217499,
      "grad_norm": 0.6909040808677673,
      "learning_rate": 3.431805498237808e-05,
      "loss": 0.8187,
      "step": 1205
    },
    {
      "epoch": 0.7455329636475663,
      "grad_norm": 0.5961218476295471,
      "learning_rate": 3.355285265611784e-05,
      "loss": 0.7922,
      "step": 1210
    },
    {
      "epoch": 0.7486136783733827,
      "grad_norm": 0.4505811333656311,
      "learning_rate": 3.279455442960238e-05,
      "loss": 0.7868,
      "step": 1215
    },
    {
      "epoch": 0.751694393099199,
      "grad_norm": 0.7794224619865417,
      "learning_rate": 3.204323909276924e-05,
      "loss": 0.813,
      "step": 1220
    },
    {
      "epoch": 0.7547751078250154,
      "grad_norm": 0.8165008425712585,
      "learning_rate": 3.1298984710008484e-05,
      "loss": 0.666,
      "step": 1225
    },
    {
      "epoch": 0.7578558225508318,
      "grad_norm": 0.72403484582901,
      "learning_rate": 3.056186861205136e-05,
      "loss": 0.7433,
      "step": 1230
    },
    {
      "epoch": 0.7609365372766482,
      "grad_norm": 0.5372538566589355,
      "learning_rate": 2.9831967387935467e-05,
      "loss": 0.8582,
      "step": 1235
    },
    {
      "epoch": 0.7640172520024646,
      "grad_norm": 0.5056626796722412,
      "learning_rate": 2.9109356877046712e-05,
      "loss": 0.7281,
      "step": 1240
    },
    {
      "epoch": 0.767097966728281,
      "grad_norm": 0.8713842630386353,
      "learning_rate": 2.8394112161239605e-05,
      "loss": 0.794,
      "step": 1245
    },
    {
      "epoch": 0.7701786814540974,
      "grad_norm": 1.0365170240402222,
      "learning_rate": 2.7686307557035685e-05,
      "loss": 0.6647,
      "step": 1250
    },
    {
      "epoch": 0.7732593961799138,
      "grad_norm": 0.534180760383606,
      "learning_rate": 2.6986016607901908e-05,
      "loss": 0.7198,
      "step": 1255
    },
    {
      "epoch": 0.7763401109057301,
      "grad_norm": 0.5429548025131226,
      "learning_rate": 2.629331207660931e-05,
      "loss": 0.7513,
      "step": 1260
    },
    {
      "epoch": 0.7794208256315465,
      "grad_norm": 0.49868354201316833,
      "learning_rate": 2.5608265937672436e-05,
      "loss": 0.7611,
      "step": 1265
    },
    {
      "epoch": 0.7825015403573629,
      "grad_norm": 0.7774792909622192,
      "learning_rate": 2.4930949369871203e-05,
      "loss": 0.7762,
      "step": 1270
    },
    {
      "epoch": 0.7855822550831792,
      "grad_norm": 1.040366291999817,
      "learning_rate": 2.426143274885493e-05,
      "loss": 0.635,
      "step": 1275
    },
    {
      "epoch": 0.7886629698089956,
      "grad_norm": 0.6801935434341431,
      "learning_rate": 2.359978563983022e-05,
      "loss": 0.7553,
      "step": 1280
    },
    {
      "epoch": 0.791743684534812,
      "grad_norm": 0.4836885929107666,
      "learning_rate": 2.2946076790332827e-05,
      "loss": 0.7228,
      "step": 1285
    },
    {
      "epoch": 0.7948243992606284,
      "grad_norm": 0.5009395480155945,
      "learning_rate": 2.2300374123084522e-05,
      "loss": 0.706,
      "step": 1290
    },
    {
      "epoch": 0.7979051139864448,
      "grad_norm": 0.8395462036132812,
      "learning_rate": 2.166274472893567e-05,
      "loss": 0.7395,
      "step": 1295
    },
    {
      "epoch": 0.8009858287122612,
      "grad_norm": 0.9128504395484924,
      "learning_rate": 2.1033254859894226e-05,
      "loss": 0.6511,
      "step": 1300
    },
    {
      "epoch": 0.8040665434380776,
      "grad_norm": 0.4820072650909424,
      "learning_rate": 2.041196992224206e-05,
      "loss": 0.7469,
      "step": 1305
    },
    {
      "epoch": 0.807147258163894,
      "grad_norm": 0.5351672768592834,
      "learning_rate": 1.9798954469738762e-05,
      "loss": 0.7357,
      "step": 1310
    },
    {
      "epoch": 0.8102279728897104,
      "grad_norm": 0.46705517172813416,
      "learning_rate": 1.919427219691453e-05,
      "loss": 0.7738,
      "step": 1315
    },
    {
      "epoch": 0.8133086876155268,
      "grad_norm": 0.767042875289917,
      "learning_rate": 1.8597985932451856e-05,
      "loss": 0.7527,
      "step": 1320
    },
    {
      "epoch": 0.8163894023413432,
      "grad_norm": 1.0822641849517822,
      "learning_rate": 1.8010157632657543e-05,
      "loss": 0.6507,
      "step": 1325
    },
    {
      "epoch": 0.8194701170671596,
      "grad_norm": 0.6491265892982483,
      "learning_rate": 1.7430848375025176e-05,
      "loss": 0.7882,
      "step": 1330
    },
    {
      "epoch": 0.822550831792976,
      "grad_norm": 0.6176543831825256,
      "learning_rate": 1.686011835188891e-05,
      "loss": 0.7986,
      "step": 1335
    },
    {
      "epoch": 0.8256315465187923,
      "grad_norm": 0.41295069456100464,
      "learning_rate": 1.6298026864169335e-05,
      "loss": 0.7427,
      "step": 1340
    },
    {
      "epoch": 0.8287122612446087,
      "grad_norm": 0.6847618818283081,
      "learning_rate": 1.5744632315211815e-05,
      "loss": 0.7551,
      "step": 1345
    },
    {
      "epoch": 0.8317929759704251,
      "grad_norm": 0.9406694769859314,
      "learning_rate": 1.5199992204718294e-05,
      "loss": 0.6678,
      "step": 1350
    },
    {
      "epoch": 0.8348736906962415,
      "grad_norm": 0.609660804271698,
      "learning_rate": 1.4664163122772689e-05,
      "loss": 0.7761,
      "step": 1355
    },
    {
      "epoch": 0.8379544054220579,
      "grad_norm": 0.6481778621673584,
      "learning_rate": 1.4137200743961188e-05,
      "loss": 0.7838,
      "step": 1360
    },
    {
      "epoch": 0.8410351201478743,
      "grad_norm": 0.4781896770000458,
      "learning_rate": 1.3619159821587235e-05,
      "loss": 0.7483,
      "step": 1365
    },
    {
      "epoch": 0.8441158348736907,
      "grad_norm": 0.6182886958122253,
      "learning_rate": 1.3110094181982657e-05,
      "loss": 0.7589,
      "step": 1370
    },
    {
      "epoch": 0.8471965495995071,
      "grad_norm": 1.1230385303497314,
      "learning_rate": 1.261005671891482e-05,
      "loss": 0.6927,
      "step": 1375
    },
    {
      "epoch": 0.8502772643253235,
      "grad_norm": 0.6041083335876465,
      "learning_rate": 1.2119099388090716e-05,
      "loss": 0.7507,
      "step": 1380
    },
    {
      "epoch": 0.8533579790511399,
      "grad_norm": 0.5692098140716553,
      "learning_rate": 1.1637273201758748e-05,
      "loss": 0.7854,
      "step": 1385
    },
    {
      "epoch": 0.8564386937769563,
      "grad_norm": 0.4730149209499359,
      "learning_rate": 1.1164628223408168e-05,
      "loss": 0.7449,
      "step": 1390
    },
    {
      "epoch": 0.8595194085027726,
      "grad_norm": 0.868976354598999,
      "learning_rate": 1.0701213562567492e-05,
      "loss": 0.7069,
      "step": 1395
    },
    {
      "epoch": 0.862600123228589,
      "grad_norm": 1.2989338636398315,
      "learning_rate": 1.0247077369701653e-05,
      "loss": 0.6703,
      "step": 1400
    },
    {
      "epoch": 0.8656808379544054,
      "grad_norm": 0.5826016068458557,
      "learning_rate": 9.802266831209206e-06,
      "loss": 0.7252,
      "step": 1405
    },
    {
      "epoch": 0.8687615526802218,
      "grad_norm": 0.5001908540725708,
      "learning_rate": 9.366828164519258e-06,
      "loss": 0.7289,
      "step": 1410
    },
    {
      "epoch": 0.8718422674060382,
      "grad_norm": 0.49539172649383545,
      "learning_rate": 8.940806613289498e-06,
      "loss": 0.7756,
      "step": 1415
    },
    {
      "epoch": 0.8749229821318546,
      "grad_norm": 0.8087944984436035,
      "learning_rate": 8.524246442705153e-06,
      "loss": 0.7602,
      "step": 1420
    },
    {
      "epoch": 0.878003696857671,
      "grad_norm": 1.1548421382904053,
      "learning_rate": 8.117190934879593e-06,
      "loss": 0.659,
      "step": 1425
    },
    {
      "epoch": 0.8810844115834874,
      "grad_norm": 0.6721958518028259,
      "learning_rate": 7.719682384357308e-06,
      "loss": 0.7557,
      "step": 1430
    },
    {
      "epoch": 0.8841651263093038,
      "grad_norm": 0.5357010364532471,
      "learning_rate": 7.33176209371923e-06,
      "loss": 0.7661,
      "step": 1435
    },
    {
      "epoch": 0.8872458410351202,
      "grad_norm": 0.5159677863121033,
      "learning_rate": 6.953470369291348e-06,
      "loss": 0.743,
      "step": 1440
    },
    {
      "epoch": 0.8903265557609366,
      "grad_norm": 0.8002091646194458,
      "learning_rate": 6.5848465169566e-06,
      "loss": 0.7694,
      "step": 1445
    },
    {
      "epoch": 0.893407270486753,
      "grad_norm": 0.9943484663963318,
      "learning_rate": 6.225928838071016e-06,
      "loss": 0.6581,
      "step": 1450
    },
    {
      "epoch": 0.8964879852125693,
      "grad_norm": 0.7004392743110657,
      "learning_rate": 5.876754625483904e-06,
      "loss": 0.6769,
      "step": 1455
    },
    {
      "epoch": 0.8995686999383857,
      "grad_norm": 0.5294577479362488,
      "learning_rate": 5.537360159663108e-06,
      "loss": 0.7258,
      "step": 1460
    },
    {
      "epoch": 0.9026494146642021,
      "grad_norm": 0.5086947679519653,
      "learning_rate": 5.207780704925314e-06,
      "loss": 0.7234,
      "step": 1465
    },
    {
      "epoch": 0.9057301293900185,
      "grad_norm": 0.6993657350540161,
      "learning_rate": 4.888050505771868e-06,
      "loss": 0.7496,
      "step": 1470
    },
    {
      "epoch": 0.9088108441158349,
      "grad_norm": 1.0011433362960815,
      "learning_rate": 4.578202783330799e-06,
      "loss": 0.6805,
      "step": 1475
    },
    {
      "epoch": 0.9118915588416513,
      "grad_norm": 0.5664235353469849,
      "learning_rate": 4.2782697319048605e-06,
      "loss": 0.7404,
      "step": 1480
    },
    {
      "epoch": 0.9149722735674677,
      "grad_norm": 0.5276147723197937,
      "learning_rate": 3.988282515626585e-06,
      "loss": 0.732,
      "step": 1485
    },
    {
      "epoch": 0.9180529882932841,
      "grad_norm": 0.512634813785553,
      "learning_rate": 3.7082712652200867e-06,
      "loss": 0.7366,
      "step": 1490
    },
    {
      "epoch": 0.9211337030191005,
      "grad_norm": 0.7594392895698547,
      "learning_rate": 3.438265074870417e-06,
      "loss": 0.7826,
      "step": 1495
    },
    {
      "epoch": 0.9242144177449169,
      "grad_norm": 1.0028218030929565,
      "learning_rate": 3.1782919992006333e-06,
      "loss": 0.665,
      "step": 1500
    },
    {
      "epoch": 0.9272951324707333,
      "grad_norm": 0.5730507969856262,
      "learning_rate": 2.9283790503567222e-06,
      "loss": 0.7497,
      "step": 1505
    },
    {
      "epoch": 0.9303758471965496,
      "grad_norm": 0.5541625618934631,
      "learning_rate": 2.6885521952010105e-06,
      "loss": 0.7605,
      "step": 1510
    },
    {
      "epoch": 0.933456561922366,
      "grad_norm": 0.5235902070999146,
      "learning_rate": 2.458836352614069e-06,
      "loss": 0.7322,
      "step": 1515
    },
    {
      "epoch": 0.9365372766481824,
      "grad_norm": 0.787257730960846,
      "learning_rate": 2.239255390905581e-06,
      "loss": 0.7623,
      "step": 1520
    },
    {
      "epoch": 0.9396179913739988,
      "grad_norm": 1.2801662683486938,
      "learning_rate": 2.029832125334319e-06,
      "loss": 0.6707,
      "step": 1525
    },
    {
      "epoch": 0.9426987060998152,
      "grad_norm": 0.604186475276947,
      "learning_rate": 1.8305883157375804e-06,
      "loss": 0.733,
      "step": 1530
    },
    {
      "epoch": 0.9457794208256316,
      "grad_norm": 0.5822204351425171,
      "learning_rate": 1.6415446642702337e-06,
      "loss": 0.7948,
      "step": 1535
    },
    {
      "epoch": 0.9488601355514479,
      "grad_norm": 0.49571382999420166,
      "learning_rate": 1.462720813253682e-06,
      "loss": 0.743,
      "step": 1540
    },
    {
      "epoch": 0.9519408502772643,
      "grad_norm": 0.8096848726272583,
      "learning_rate": 1.2941353431350056e-06,
      "loss": 0.7614,
      "step": 1545
    },
    {
      "epoch": 0.9550215650030807,
      "grad_norm": 1.2853180170059204,
      "learning_rate": 1.135805770556364e-06,
      "loss": 0.7004,
      "step": 1550
    },
    {
      "epoch": 0.958102279728897,
      "grad_norm": 0.5894697308540344,
      "learning_rate": 9.877485465349058e-07,
      "loss": 0.8042,
      "step": 1555
    },
    {
      "epoch": 0.9611829944547134,
      "grad_norm": 0.5564848780632019,
      "learning_rate": 8.499790547535025e-07,
      "loss": 0.742,
      "step": 1560
    },
    {
      "epoch": 0.9642637091805298,
      "grad_norm": 0.4396416246891022,
      "learning_rate": 7.225116099623286e-07,
      "loss": 0.6936,
      "step": 1565
    },
    {
      "epoch": 0.9673444239063462,
      "grad_norm": 0.8373792171478271,
      "learning_rate": 6.053594564914611e-07,
      "loss": 0.7837,
      "step": 1570
    },
    {
      "epoch": 0.9704251386321626,
      "grad_norm": 1.2192115783691406,
      "learning_rate": 4.985347668747809e-07,
      "loss": 0.6541,
      "step": 1575
    },
    {
      "epoch": 0.973505853357979,
      "grad_norm": 0.6488357782363892,
      "learning_rate": 4.0204864058522864e-07,
      "loss": 0.8073,
      "step": 1580
    },
    {
      "epoch": 0.9765865680837954,
      "grad_norm": 0.456813782453537,
      "learning_rate": 3.15911102881461e-07,
      "loss": 0.837,
      "step": 1585
    },
    {
      "epoch": 0.9796672828096118,
      "grad_norm": 0.514924168586731,
      "learning_rate": 2.40131103766239e-07,
      "loss": 0.7338,
      "step": 1590
    },
    {
      "epoch": 0.9827479975354282,
      "grad_norm": 0.7612470984458923,
      "learning_rate": 1.747165170564724e-07,
      "loss": 0.7998,
      "step": 1595
    },
    {
      "epoch": 0.9858287122612446,
      "grad_norm": 0.8824509978294373,
      "learning_rate": 1.1967413956510686e-07,
      "loss": 0.6986,
      "step": 1600
    },
    {
      "epoch": 0.988909426987061,
      "grad_norm": 0.6480757594108582,
      "learning_rate": 7.500969039491157e-08,
      "loss": 0.7331,
      "step": 1605
    },
    {
      "epoch": 0.9919901417128774,
      "grad_norm": 0.6705480217933655,
      "learning_rate": 4.0727810344254325e-08,
      "loss": 0.8027,
      "step": 1610
    },
    {
      "epoch": 0.9950708564386938,
      "grad_norm": 0.4745963215827942,
      "learning_rate": 1.6832061424865153e-08,
      "loss": 0.7123,
      "step": 1615
    },
    {
      "epoch": 0.9981515711645101,
      "grad_norm": 0.5832483172416687,
      "learning_rate": 3.3249264917878387e-09,
      "loss": 0.6986,
      "step": 1620
    },
    {
      "epoch": 1.0,
      "step": 1623,
      "total_flos": 123288491655168.0,
      "train_loss": 0.7956359339436115,
      "train_runtime": 17588.8577,
      "train_samples_per_second": 2.952,
      "train_steps_per_second": 0.092
    }
  ],
  "logging_steps": 5,
  "max_steps": 1623,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 123288491655168.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}