{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.5125815470643057,
  "eval_steps": 500,
  "global_step": 550,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.004659832246039142,
      "grad_norm": 55.709484937361424,
      "learning_rate": 7.763975155279503e-07,
      "loss": 11.0694,
      "step": 5
    },
    {
      "epoch": 0.009319664492078284,
      "grad_norm": 57.1398137451701,
      "learning_rate": 1.5527950310559006e-06,
      "loss": 10.8846,
      "step": 10
    },
    {
      "epoch": 0.013979496738117428,
      "grad_norm": 97.98047478379912,
      "learning_rate": 2.329192546583851e-06,
      "loss": 9.4413,
      "step": 15
    },
    {
      "epoch": 0.01863932898415657,
      "grad_norm": 32.76887701215898,
      "learning_rate": 3.1055900621118013e-06,
      "loss": 3.214,
      "step": 20
    },
    {
      "epoch": 0.023299161230195712,
      "grad_norm": 3.315394944010228,
      "learning_rate": 3.881987577639752e-06,
      "loss": 1.346,
      "step": 25
    },
    {
      "epoch": 0.027958993476234855,
      "grad_norm": 1.29231664681427,
      "learning_rate": 4.658385093167702e-06,
      "loss": 1.006,
      "step": 30
    },
    {
      "epoch": 0.032618825722273995,
      "grad_norm": 0.8629586527731651,
      "learning_rate": 5.4347826086956525e-06,
      "loss": 0.8411,
      "step": 35
    },
    {
      "epoch": 0.03727865796831314,
      "grad_norm": 0.6212648141101664,
      "learning_rate": 6.2111801242236025e-06,
      "loss": 0.7754,
      "step": 40
    },
    {
      "epoch": 0.04193849021435228,
      "grad_norm": 0.46040470834395936,
      "learning_rate": 6.9875776397515525e-06,
      "loss": 0.7136,
      "step": 45
    },
    {
      "epoch": 0.046598322460391424,
      "grad_norm": 0.42693871288027446,
      "learning_rate": 7.763975155279503e-06,
      "loss": 0.6831,
      "step": 50
    },
    {
      "epoch": 0.05125815470643057,
      "grad_norm": 0.44192720551532577,
      "learning_rate": 8.540372670807453e-06,
      "loss": 0.646,
      "step": 55
    },
    {
      "epoch": 0.05591798695246971,
      "grad_norm": 0.31222059058105195,
      "learning_rate": 9.316770186335403e-06,
      "loss": 0.6088,
      "step": 60
    },
    {
      "epoch": 0.06057781919850885,
      "grad_norm": 0.30027349299916717,
      "learning_rate": 1.0093167701863353e-05,
      "loss": 0.5932,
      "step": 65
    },
    {
      "epoch": 0.06523765144454799,
      "grad_norm": 0.3320480978786744,
      "learning_rate": 1.0869565217391305e-05,
      "loss": 0.5892,
      "step": 70
    },
    {
      "epoch": 0.06989748369058714,
      "grad_norm": 0.3001101299512147,
      "learning_rate": 1.1645962732919255e-05,
      "loss": 0.5713,
      "step": 75
    },
    {
      "epoch": 0.07455731593662628,
      "grad_norm": 0.30892715626333545,
      "learning_rate": 1.2422360248447205e-05,
      "loss": 0.5619,
      "step": 80
    },
    {
      "epoch": 0.07921714818266543,
      "grad_norm": 0.3319072685834915,
      "learning_rate": 1.3198757763975155e-05,
      "loss": 0.5525,
      "step": 85
    },
    {
      "epoch": 0.08387698042870456,
      "grad_norm": 0.29047402255141175,
      "learning_rate": 1.3975155279503105e-05,
      "loss": 0.5628,
      "step": 90
    },
    {
      "epoch": 0.08853681267474371,
      "grad_norm": 0.25922715271689156,
      "learning_rate": 1.4751552795031057e-05,
      "loss": 0.5443,
      "step": 95
    },
    {
      "epoch": 0.09319664492078285,
      "grad_norm": 0.31531185304643106,
      "learning_rate": 1.5527950310559007e-05,
      "loss": 0.5296,
      "step": 100
    },
    {
      "epoch": 0.097856477166822,
      "grad_norm": 0.3013647908850556,
      "learning_rate": 1.630434782608696e-05,
      "loss": 0.5243,
      "step": 105
    },
    {
      "epoch": 0.10251630941286113,
      "grad_norm": 0.3047350591065195,
      "learning_rate": 1.7080745341614907e-05,
      "loss": 0.5261,
      "step": 110
    },
    {
      "epoch": 0.10717614165890028,
      "grad_norm": 0.30780196150291267,
      "learning_rate": 1.785714285714286e-05,
      "loss": 0.5231,
      "step": 115
    },
    {
      "epoch": 0.11183597390493942,
      "grad_norm": 0.3804666896095633,
      "learning_rate": 1.8633540372670807e-05,
      "loss": 0.5098,
      "step": 120
    },
    {
      "epoch": 0.11649580615097857,
      "grad_norm": 0.2868238329098672,
      "learning_rate": 1.940993788819876e-05,
      "loss": 0.5057,
      "step": 125
    },
    {
      "epoch": 0.1211556383970177,
      "grad_norm": 0.32341698454060924,
      "learning_rate": 2.0186335403726707e-05,
      "loss": 0.5114,
      "step": 130
    },
    {
      "epoch": 0.12581547064305684,
      "grad_norm": 0.4221000244611152,
      "learning_rate": 2.096273291925466e-05,
      "loss": 0.5106,
      "step": 135
    },
    {
      "epoch": 0.13047530288909598,
      "grad_norm": 0.47197386491626697,
      "learning_rate": 2.173913043478261e-05,
      "loss": 0.4951,
      "step": 140
    },
    {
      "epoch": 0.13513513513513514,
      "grad_norm": 0.3721540533945164,
      "learning_rate": 2.2515527950310562e-05,
      "loss": 0.522,
      "step": 145
    },
    {
      "epoch": 0.13979496738117428,
      "grad_norm": 0.4388944569421825,
      "learning_rate": 2.329192546583851e-05,
      "loss": 0.4959,
      "step": 150
    },
    {
      "epoch": 0.14445479962721341,
      "grad_norm": 0.4622898334441083,
      "learning_rate": 2.4068322981366462e-05,
      "loss": 0.5041,
      "step": 155
    },
    {
      "epoch": 0.14911463187325255,
      "grad_norm": 0.49630920093444675,
      "learning_rate": 2.484472049689441e-05,
      "loss": 0.4849,
      "step": 160
    },
    {
      "epoch": 0.15377446411929171,
      "grad_norm": 0.46849210325540636,
      "learning_rate": 2.5621118012422362e-05,
      "loss": 0.4961,
      "step": 165
    },
    {
      "epoch": 0.15843429636533085,
      "grad_norm": 0.4082563546888238,
      "learning_rate": 2.639751552795031e-05,
      "loss": 0.4719,
      "step": 170
    },
    {
      "epoch": 0.16309412861137,
      "grad_norm": 0.4549138940471945,
      "learning_rate": 2.7173913043478262e-05,
      "loss": 0.4852,
      "step": 175
    },
    {
      "epoch": 0.16775396085740912,
      "grad_norm": 0.41032309538662626,
      "learning_rate": 2.795031055900621e-05,
      "loss": 0.478,
      "step": 180
    },
    {
      "epoch": 0.1724137931034483,
      "grad_norm": 0.36866874054662224,
      "learning_rate": 2.8726708074534165e-05,
      "loss": 0.481,
      "step": 185
    },
    {
      "epoch": 0.17707362534948742,
      "grad_norm": 0.36280659725390446,
      "learning_rate": 2.9503105590062114e-05,
      "loss": 0.4908,
      "step": 190
    },
    {
      "epoch": 0.18173345759552656,
      "grad_norm": 0.4446282244198985,
      "learning_rate": 3.0279503105590062e-05,
      "loss": 0.4802,
      "step": 195
    },
    {
      "epoch": 0.1863932898415657,
      "grad_norm": 0.4818264381472692,
      "learning_rate": 3.1055900621118014e-05,
      "loss": 0.4818,
      "step": 200
    },
    {
      "epoch": 0.19105312208760486,
      "grad_norm": 0.5191372230429591,
      "learning_rate": 3.183229813664597e-05,
      "loss": 0.4708,
      "step": 205
    },
    {
      "epoch": 0.195712954333644,
      "grad_norm": 0.5690145277815086,
      "learning_rate": 3.260869565217392e-05,
      "loss": 0.4848,
      "step": 210
    },
    {
      "epoch": 0.20037278657968313,
      "grad_norm": 0.41006685540485915,
      "learning_rate": 3.3385093167701865e-05,
      "loss": 0.467,
      "step": 215
    },
    {
      "epoch": 0.20503261882572227,
      "grad_norm": 0.47272355547942274,
      "learning_rate": 3.4161490683229814e-05,
      "loss": 0.4716,
      "step": 220
    },
    {
      "epoch": 0.2096924510717614,
      "grad_norm": 0.5135260711958035,
      "learning_rate": 3.493788819875777e-05,
      "loss": 0.4536,
      "step": 225
    },
    {
      "epoch": 0.21435228331780057,
      "grad_norm": 0.5124408748792798,
      "learning_rate": 3.571428571428572e-05,
      "loss": 0.467,
      "step": 230
    },
    {
      "epoch": 0.2190121155638397,
      "grad_norm": 0.4205466417850958,
      "learning_rate": 3.6490683229813665e-05,
      "loss": 0.4647,
      "step": 235
    },
    {
      "epoch": 0.22367194780987884,
      "grad_norm": 0.6251611138936444,
      "learning_rate": 3.7267080745341614e-05,
      "loss": 0.4624,
      "step": 240
    },
    {
      "epoch": 0.22833178005591798,
      "grad_norm": 0.5610747747609784,
      "learning_rate": 3.804347826086957e-05,
      "loss": 0.4676,
      "step": 245
    },
    {
      "epoch": 0.23299161230195714,
      "grad_norm": 0.4692752738735764,
      "learning_rate": 3.881987577639752e-05,
      "loss": 0.4592,
      "step": 250
    },
    {
      "epoch": 0.23765144454799628,
      "grad_norm": 0.5435654364057863,
      "learning_rate": 3.9596273291925465e-05,
      "loss": 0.4679,
      "step": 255
    },
    {
      "epoch": 0.2423112767940354,
      "grad_norm": 0.5264727485653637,
      "learning_rate": 4.0372670807453414e-05,
      "loss": 0.463,
      "step": 260
    },
    {
      "epoch": 0.24697110904007455,
      "grad_norm": 0.5096976732033014,
      "learning_rate": 4.114906832298137e-05,
      "loss": 0.4613,
      "step": 265
    },
    {
      "epoch": 0.2516309412861137,
      "grad_norm": 0.4649125930738299,
      "learning_rate": 4.192546583850932e-05,
      "loss": 0.4594,
      "step": 270
    },
    {
      "epoch": 0.25629077353215285,
      "grad_norm": 0.5173502447936299,
      "learning_rate": 4.270186335403727e-05,
      "loss": 0.4652,
      "step": 275
    },
    {
      "epoch": 0.26095060577819196,
      "grad_norm": 0.3868002746815184,
      "learning_rate": 4.347826086956522e-05,
      "loss": 0.4572,
      "step": 280
    },
    {
      "epoch": 0.2656104380242311,
      "grad_norm": 0.5001940885551612,
      "learning_rate": 4.425465838509317e-05,
      "loss": 0.4613,
      "step": 285
    },
    {
      "epoch": 0.2702702702702703,
      "grad_norm": 0.6213239402228385,
      "learning_rate": 4.5031055900621124e-05,
      "loss": 0.4633,
      "step": 290
    },
    {
      "epoch": 0.2749301025163094,
      "grad_norm": 0.6125652570862165,
      "learning_rate": 4.580745341614907e-05,
      "loss": 0.4673,
      "step": 295
    },
    {
      "epoch": 0.27958993476234856,
      "grad_norm": 0.7479197761570839,
      "learning_rate": 4.658385093167702e-05,
      "loss": 0.4693,
      "step": 300
    },
    {
      "epoch": 0.2842497670083877,
      "grad_norm": 0.5483826450591869,
      "learning_rate": 4.736024844720497e-05,
      "loss": 0.4579,
      "step": 305
    },
    {
      "epoch": 0.28890959925442683,
      "grad_norm": 0.6644586513259841,
      "learning_rate": 4.8136645962732924e-05,
      "loss": 0.45,
      "step": 310
    },
    {
      "epoch": 0.293569431500466,
      "grad_norm": 0.8188490252541196,
      "learning_rate": 4.891304347826087e-05,
      "loss": 0.459,
      "step": 315
    },
    {
      "epoch": 0.2982292637465051,
      "grad_norm": 1.034709523419619,
      "learning_rate": 4.968944099378882e-05,
      "loss": 0.4441,
      "step": 320
    },
    {
      "epoch": 0.30288909599254427,
      "grad_norm": 0.759919110757548,
      "learning_rate": 4.994822229892993e-05,
      "loss": 0.4415,
      "step": 325
    },
    {
      "epoch": 0.30754892823858343,
      "grad_norm": 0.6470711913841568,
      "learning_rate": 4.986192613047981e-05,
      "loss": 0.4519,
      "step": 330
    },
    {
      "epoch": 0.31220876048462254,
      "grad_norm": 0.8153464642106912,
      "learning_rate": 4.977562996202969e-05,
      "loss": 0.4527,
      "step": 335
    },
    {
      "epoch": 0.3168685927306617,
      "grad_norm": 0.4698984383111023,
      "learning_rate": 4.968933379357957e-05,
      "loss": 0.4532,
      "step": 340
    },
    {
      "epoch": 0.32152842497670087,
      "grad_norm": 0.4756319068347418,
      "learning_rate": 4.9603037625129445e-05,
      "loss": 0.4479,
      "step": 345
    },
    {
      "epoch": 0.32618825722274,
      "grad_norm": 0.35969300915800667,
      "learning_rate": 4.951674145667933e-05,
      "loss": 0.4613,
      "step": 350
    },
    {
      "epoch": 0.33084808946877914,
      "grad_norm": 0.5175115868112082,
      "learning_rate": 4.94304452882292e-05,
      "loss": 0.4473,
      "step": 355
    },
    {
      "epoch": 0.33550792171481825,
      "grad_norm": 0.46094013566363407,
      "learning_rate": 4.934414911977908e-05,
      "loss": 0.445,
      "step": 360
    },
    {
      "epoch": 0.3401677539608574,
      "grad_norm": 0.4104764392290741,
      "learning_rate": 4.9257852951328965e-05,
      "loss": 0.4415,
      "step": 365
    },
    {
      "epoch": 0.3448275862068966,
      "grad_norm": 0.40096213441032663,
      "learning_rate": 4.917155678287884e-05,
      "loss": 0.4546,
      "step": 370
    },
    {
      "epoch": 0.3494874184529357,
      "grad_norm": 0.4645867033422015,
      "learning_rate": 4.908526061442872e-05,
      "loss": 0.4408,
      "step": 375
    },
    {
      "epoch": 0.35414725069897485,
      "grad_norm": 0.4400737008116359,
      "learning_rate": 4.89989644459786e-05,
      "loss": 0.4557,
      "step": 380
    },
    {
      "epoch": 0.35880708294501396,
      "grad_norm": 0.4260635316927096,
      "learning_rate": 4.891266827752848e-05,
      "loss": 0.4575,
      "step": 385
    },
    {
      "epoch": 0.3634669151910531,
      "grad_norm": 0.5323727061907297,
      "learning_rate": 4.882637210907836e-05,
      "loss": 0.4417,
      "step": 390
    },
    {
      "epoch": 0.3681267474370923,
      "grad_norm": 0.5836242210222355,
      "learning_rate": 4.874007594062824e-05,
      "loss": 0.4481,
      "step": 395
    },
    {
      "epoch": 0.3727865796831314,
      "grad_norm": 0.4273563357499149,
      "learning_rate": 4.865377977217811e-05,
      "loss": 0.4327,
      "step": 400
    },
    {
      "epoch": 0.37744641192917056,
      "grad_norm": 0.3953667888190435,
      "learning_rate": 4.8567483603728e-05,
      "loss": 0.4512,
      "step": 405
    },
    {
      "epoch": 0.3821062441752097,
      "grad_norm": 0.41871105603942904,
      "learning_rate": 4.8481187435277875e-05,
      "loss": 0.4471,
      "step": 410
    },
    {
      "epoch": 0.38676607642124883,
      "grad_norm": 0.5437922609735028,
      "learning_rate": 4.839489126682776e-05,
      "loss": 0.4338,
      "step": 415
    },
    {
      "epoch": 0.391425908667288,
      "grad_norm": 0.5564237695130677,
      "learning_rate": 4.830859509837763e-05,
      "loss": 0.4521,
      "step": 420
    },
    {
      "epoch": 0.3960857409133271,
      "grad_norm": 0.34206255911684263,
      "learning_rate": 4.822229892992751e-05,
      "loss": 0.4495,
      "step": 425
    },
    {
      "epoch": 0.40074557315936626,
      "grad_norm": 0.4019378807561449,
      "learning_rate": 4.8136002761477395e-05,
      "loss": 0.4365,
      "step": 430
    },
    {
      "epoch": 0.40540540540540543,
      "grad_norm": 0.5440015382583567,
      "learning_rate": 4.804970659302727e-05,
      "loss": 0.4499,
      "step": 435
    },
    {
      "epoch": 0.41006523765144454,
      "grad_norm": 0.5537615079643711,
      "learning_rate": 4.796341042457715e-05,
      "loss": 0.4438,
      "step": 440
    },
    {
      "epoch": 0.4147250698974837,
      "grad_norm": 0.586960808161654,
      "learning_rate": 4.787711425612703e-05,
      "loss": 0.4386,
      "step": 445
    },
    {
      "epoch": 0.4193849021435228,
      "grad_norm": 0.5072343488418793,
      "learning_rate": 4.779081808767691e-05,
      "loss": 0.4292,
      "step": 450
    },
    {
      "epoch": 0.424044734389562,
      "grad_norm": 0.48278977249280486,
      "learning_rate": 4.770452191922679e-05,
      "loss": 0.4362,
      "step": 455
    },
    {
      "epoch": 0.42870456663560114,
      "grad_norm": 0.39818179591075503,
      "learning_rate": 4.761822575077667e-05,
      "loss": 0.4329,
      "step": 460
    },
    {
      "epoch": 0.43336439888164024,
      "grad_norm": 0.48993775783018667,
      "learning_rate": 4.753192958232654e-05,
      "loss": 0.4299,
      "step": 465
    },
    {
      "epoch": 0.4380242311276794,
      "grad_norm": 0.41171920657337485,
      "learning_rate": 4.744563341387643e-05,
      "loss": 0.4428,
      "step": 470
    },
    {
      "epoch": 0.4426840633737186,
      "grad_norm": 0.4400559670696649,
      "learning_rate": 4.7359337245426306e-05,
      "loss": 0.432,
      "step": 475
    },
    {
      "epoch": 0.4473438956197577,
      "grad_norm": 0.4686484619433322,
      "learning_rate": 4.7273041076976184e-05,
      "loss": 0.4441,
      "step": 480
    },
    {
      "epoch": 0.45200372786579684,
      "grad_norm": 0.4455986109270811,
      "learning_rate": 4.718674490852606e-05,
      "loss": 0.4368,
      "step": 485
    },
    {
      "epoch": 0.45666356011183595,
      "grad_norm": 0.4366809315888367,
      "learning_rate": 4.710044874007594e-05,
      "loss": 0.4312,
      "step": 490
    },
    {
      "epoch": 0.4613233923578751,
      "grad_norm": 0.4587131719176713,
      "learning_rate": 4.7014152571625826e-05,
      "loss": 0.4369,
      "step": 495
    },
    {
      "epoch": 0.4659832246039143,
      "grad_norm": 0.5269402211397143,
      "learning_rate": 4.6927856403175704e-05,
      "loss": 0.44,
      "step": 500
    },
    {
      "epoch": 0.4706430568499534,
      "grad_norm": 0.387320399512531,
      "learning_rate": 4.684156023472558e-05,
      "loss": 0.4266,
      "step": 505
    },
    {
      "epoch": 0.47530288909599255,
      "grad_norm": 0.4446811703899567,
      "learning_rate": 4.675526406627546e-05,
      "loss": 0.435,
      "step": 510
    },
    {
      "epoch": 0.47996272134203166,
      "grad_norm": 0.6228000644001809,
      "learning_rate": 4.666896789782534e-05,
      "loss": 0.433,
      "step": 515
    },
    {
      "epoch": 0.4846225535880708,
      "grad_norm": 0.5372584652001571,
      "learning_rate": 4.658267172937522e-05,
      "loss": 0.4368,
      "step": 520
    },
    {
      "epoch": 0.48928238583411,
      "grad_norm": 0.4279339582606534,
      "learning_rate": 4.64963755609251e-05,
      "loss": 0.4358,
      "step": 525
    },
    {
      "epoch": 0.4939422180801491,
      "grad_norm": 0.5212302132415493,
      "learning_rate": 4.641007939247497e-05,
      "loss": 0.4327,
      "step": 530
    },
    {
      "epoch": 0.49860205032618826,
      "grad_norm": 0.38993326302525777,
      "learning_rate": 4.632378322402486e-05,
      "loss": 0.4329,
      "step": 535
    },
    {
      "epoch": 0.5032618825722274,
      "grad_norm": 0.3927986683558084,
      "learning_rate": 4.6237487055574736e-05,
      "loss": 0.4228,
      "step": 540
    },
    {
      "epoch": 0.5079217148182665,
      "grad_norm": 0.48245624377171165,
      "learning_rate": 4.6151190887124615e-05,
      "loss": 0.4292,
      "step": 545
    },
    {
      "epoch": 0.5125815470643057,
      "grad_norm": 0.4417482867433448,
      "learning_rate": 4.606489471867449e-05,
      "loss": 0.4293,
      "step": 550
    }
  ],
  "logging_steps": 5,
  "max_steps": 3219,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 550,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 4.7079688949989376e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}