{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 1073,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.004659832246039142,
      "grad_norm": 55.73911101293821,
      "learning_rate": 7.763975155279503e-07,
      "loss": 11.0693,
      "step": 5
    },
    {
      "epoch": 0.009319664492078284,
      "grad_norm": 57.1718581665509,
      "learning_rate": 1.5527950310559006e-06,
      "loss": 10.8839,
      "step": 10
    },
    {
      "epoch": 0.013979496738117428,
      "grad_norm": 98.21065880054624,
      "learning_rate": 2.329192546583851e-06,
      "loss": 9.4422,
      "step": 15
    },
    {
      "epoch": 0.01863932898415657,
      "grad_norm": 32.82367751669906,
      "learning_rate": 3.1055900621118013e-06,
      "loss": 3.2146,
      "step": 20
    },
    {
      "epoch": 0.023299161230195712,
      "grad_norm": 3.30697302187253,
      "learning_rate": 3.881987577639752e-06,
      "loss": 1.347,
      "step": 25
    },
    {
      "epoch": 0.027958993476234855,
      "grad_norm": 1.2985174034570703,
      "learning_rate": 4.658385093167702e-06,
      "loss": 1.0061,
      "step": 30
    },
    {
      "epoch": 0.032618825722273995,
      "grad_norm": 0.8623629897688332,
      "learning_rate": 5.4347826086956525e-06,
      "loss": 0.8411,
      "step": 35
    },
    {
      "epoch": 0.03727865796831314,
      "grad_norm": 0.6188703104605264,
      "learning_rate": 6.2111801242236025e-06,
      "loss": 0.7755,
      "step": 40
    },
    {
      "epoch": 0.04193849021435228,
      "grad_norm": 0.461009679090195,
      "learning_rate": 6.9875776397515525e-06,
      "loss": 0.7136,
      "step": 45
    },
    {
      "epoch": 0.046598322460391424,
      "grad_norm": 0.4286257732967241,
      "learning_rate": 7.763975155279503e-06,
      "loss": 0.6831,
      "step": 50
    },
    {
      "epoch": 0.05125815470643057,
      "grad_norm": 0.44241392108033467,
      "learning_rate": 8.540372670807453e-06,
      "loss": 0.646,
      "step": 55
    },
    {
      "epoch": 0.05591798695246971,
      "grad_norm": 0.3111820442035257,
      "learning_rate": 9.316770186335403e-06,
      "loss": 0.6088,
      "step": 60
    },
    {
      "epoch": 0.06057781919850885,
      "grad_norm": 0.30251919432888047,
      "learning_rate": 1.0093167701863353e-05,
      "loss": 0.5932,
      "step": 65
    },
    {
      "epoch": 0.06523765144454799,
      "grad_norm": 0.33324119756936554,
      "learning_rate": 1.0869565217391305e-05,
      "loss": 0.5892,
      "step": 70
    },
    {
      "epoch": 0.06989748369058714,
      "grad_norm": 0.3063487771822568,
      "learning_rate": 1.1645962732919255e-05,
      "loss": 0.5713,
      "step": 75
    },
    {
      "epoch": 0.07455731593662628,
      "grad_norm": 0.3071310049244509,
      "learning_rate": 1.2422360248447205e-05,
      "loss": 0.5618,
      "step": 80
    },
    {
      "epoch": 0.07921714818266543,
      "grad_norm": 0.3358130788082856,
      "learning_rate": 1.3198757763975155e-05,
      "loss": 0.5524,
      "step": 85
    },
    {
      "epoch": 0.08387698042870456,
      "grad_norm": 0.2925977780766404,
      "learning_rate": 1.3975155279503105e-05,
      "loss": 0.5628,
      "step": 90
    },
    {
      "epoch": 0.08853681267474371,
      "grad_norm": 0.25884643717759576,
      "learning_rate": 1.4751552795031057e-05,
      "loss": 0.5442,
      "step": 95
    },
    {
      "epoch": 0.09319664492078285,
      "grad_norm": 0.31707838864665394,
      "learning_rate": 1.5527950310559007e-05,
      "loss": 0.5296,
      "step": 100
    },
    {
      "epoch": 0.097856477166822,
      "grad_norm": 0.3021647811110462,
      "learning_rate": 1.630434782608696e-05,
      "loss": 0.5243,
      "step": 105
    },
    {
      "epoch": 0.10251630941286113,
      "grad_norm": 0.30442757517268515,
      "learning_rate": 1.7080745341614907e-05,
      "loss": 0.526,
      "step": 110
    },
    {
      "epoch": 0.10717614165890028,
      "grad_norm": 0.3036499569102043,
      "learning_rate": 1.785714285714286e-05,
      "loss": 0.5231,
      "step": 115
    },
    {
      "epoch": 0.11183597390493942,
      "grad_norm": 0.37387951427884036,
      "learning_rate": 1.8633540372670807e-05,
      "loss": 0.5098,
      "step": 120
    },
    {
      "epoch": 0.11649580615097857,
      "grad_norm": 0.2871555813435635,
      "learning_rate": 1.940993788819876e-05,
      "loss": 0.5057,
      "step": 125
    },
    {
      "epoch": 0.1211556383970177,
      "grad_norm": 0.3209503297683246,
      "learning_rate": 2.0186335403726707e-05,
      "loss": 0.5113,
      "step": 130
    },
    {
      "epoch": 0.12581547064305684,
      "grad_norm": 0.41553906221283793,
      "learning_rate": 2.096273291925466e-05,
      "loss": 0.5105,
      "step": 135
    },
    {
      "epoch": 0.13047530288909598,
      "grad_norm": 0.447991972215783,
      "learning_rate": 2.173913043478261e-05,
      "loss": 0.4951,
      "step": 140
    },
    {
      "epoch": 0.13513513513513514,
      "grad_norm": 0.36952665164899706,
      "learning_rate": 2.2515527950310562e-05,
      "loss": 0.522,
      "step": 145
    },
    {
      "epoch": 0.13979496738117428,
      "grad_norm": 0.43854865846501767,
      "learning_rate": 2.329192546583851e-05,
      "loss": 0.4958,
      "step": 150
    },
    {
      "epoch": 0.14445479962721341,
      "grad_norm": 0.4551787765980115,
      "learning_rate": 2.4068322981366462e-05,
      "loss": 0.5041,
      "step": 155
    },
    {
      "epoch": 0.14911463187325255,
      "grad_norm": 0.500295346094291,
      "learning_rate": 2.484472049689441e-05,
      "loss": 0.4847,
      "step": 160
    },
    {
      "epoch": 0.15377446411929171,
      "grad_norm": 0.4867357618004153,
      "learning_rate": 2.5621118012422362e-05,
      "loss": 0.4961,
      "step": 165
    },
    {
      "epoch": 0.15843429636533085,
      "grad_norm": 0.40984119146102166,
      "learning_rate": 2.639751552795031e-05,
      "loss": 0.4719,
      "step": 170
    },
    {
      "epoch": 0.16309412861137,
      "grad_norm": 0.4604406630630476,
      "learning_rate": 2.7173913043478262e-05,
      "loss": 0.4852,
      "step": 175
    },
    {
      "epoch": 0.16775396085740912,
      "grad_norm": 0.40948649685519667,
      "learning_rate": 2.795031055900621e-05,
      "loss": 0.478,
      "step": 180
    },
    {
      "epoch": 0.1724137931034483,
      "grad_norm": 0.3653652404773955,
      "learning_rate": 2.8726708074534165e-05,
      "loss": 0.4809,
      "step": 185
    },
    {
      "epoch": 0.17707362534948742,
      "grad_norm": 0.3788769696744858,
      "learning_rate": 2.9503105590062114e-05,
      "loss": 0.4907,
      "step": 190
    },
    {
      "epoch": 0.18173345759552656,
      "grad_norm": 0.4455345320827965,
      "learning_rate": 3.0279503105590062e-05,
      "loss": 0.4802,
      "step": 195
    },
    {
      "epoch": 0.1863932898415657,
      "grad_norm": 0.4872519077868042,
      "learning_rate": 3.1055900621118014e-05,
      "loss": 0.4818,
      "step": 200
    },
    {
      "epoch": 0.19105312208760486,
      "grad_norm": 0.5176809127208468,
      "learning_rate": 3.183229813664597e-05,
      "loss": 0.4707,
      "step": 205
    },
    {
      "epoch": 0.195712954333644,
      "grad_norm": 0.5721945537346641,
      "learning_rate": 3.260869565217392e-05,
      "loss": 0.4848,
      "step": 210
    },
    {
      "epoch": 0.20037278657968313,
      "grad_norm": 0.3770305001602024,
      "learning_rate": 3.3385093167701865e-05,
      "loss": 0.4669,
      "step": 215
    },
    {
      "epoch": 0.20503261882572227,
      "grad_norm": 0.4543945451638326,
      "learning_rate": 3.4161490683229814e-05,
      "loss": 0.4713,
      "step": 220
    },
    {
      "epoch": 0.2096924510717614,
      "grad_norm": 0.49369340193910277,
      "learning_rate": 3.493788819875777e-05,
      "loss": 0.4534,
      "step": 225
    },
    {
      "epoch": 0.21435228331780057,
      "grad_norm": 0.48464327291832504,
      "learning_rate": 3.571428571428572e-05,
      "loss": 0.4668,
      "step": 230
    },
    {
      "epoch": 0.2190121155638397,
      "grad_norm": 0.4111365129541678,
      "learning_rate": 3.6490683229813665e-05,
      "loss": 0.4644,
      "step": 235
    },
    {
      "epoch": 0.22367194780987884,
      "grad_norm": 0.6275047885479089,
      "learning_rate": 3.7267080745341614e-05,
      "loss": 0.4622,
      "step": 240
    },
    {
      "epoch": 0.22833178005591798,
      "grad_norm": 0.556206325520564,
      "learning_rate": 3.804347826086957e-05,
      "loss": 0.4674,
      "step": 245
    },
    {
      "epoch": 0.23299161230195714,
      "grad_norm": 0.4727137299853098,
      "learning_rate": 3.881987577639752e-05,
      "loss": 0.4592,
      "step": 250
    },
    {
      "epoch": 0.23765144454799628,
      "grad_norm": 0.5175385392810632,
      "learning_rate": 3.9596273291925465e-05,
      "loss": 0.4679,
      "step": 255
    },
    {
      "epoch": 0.2423112767940354,
      "grad_norm": 0.5522111161928154,
      "learning_rate": 4.0372670807453414e-05,
      "loss": 0.463,
      "step": 260
    },
    {
      "epoch": 0.24697110904007455,
      "grad_norm": 0.533406707658831,
      "learning_rate": 4.114906832298137e-05,
      "loss": 0.4613,
      "step": 265
    },
    {
      "epoch": 0.2516309412861137,
      "grad_norm": 0.5190269056190954,
      "learning_rate": 4.192546583850932e-05,
      "loss": 0.4596,
      "step": 270
    },
    {
      "epoch": 0.25629077353215285,
      "grad_norm": 0.5124788795006116,
      "learning_rate": 4.270186335403727e-05,
      "loss": 0.4654,
      "step": 275
    },
    {
      "epoch": 0.26095060577819196,
      "grad_norm": 0.6274404486819355,
      "learning_rate": 4.347826086956522e-05,
      "loss": 0.4576,
      "step": 280
    },
    {
      "epoch": 0.2656104380242311,
      "grad_norm": 0.4751490435936714,
      "learning_rate": 4.425465838509317e-05,
      "loss": 0.4623,
      "step": 285
    },
    {
      "epoch": 0.2702702702702703,
      "grad_norm": 0.7744692776296426,
      "learning_rate": 4.5031055900621124e-05,
      "loss": 0.4654,
      "step": 290
    },
    {
      "epoch": 0.2749301025163094,
      "grad_norm": 0.7800040335094921,
      "learning_rate": 4.580745341614907e-05,
      "loss": 0.4682,
      "step": 295
    },
    {
      "epoch": 0.27958993476234856,
      "grad_norm": 0.8281886923177356,
      "learning_rate": 4.658385093167702e-05,
      "loss": 0.4703,
      "step": 300
    },
    {
      "epoch": 0.2842497670083877,
      "grad_norm": 0.5858908773046568,
      "learning_rate": 4.736024844720497e-05,
      "loss": 0.4578,
      "step": 305
    },
    {
      "epoch": 0.28890959925442683,
      "grad_norm": 0.45985550866997615,
      "learning_rate": 4.8136645962732924e-05,
      "loss": 0.4497,
      "step": 310
    },
    {
      "epoch": 0.293569431500466,
      "grad_norm": 0.5286612419877652,
      "learning_rate": 4.891304347826087e-05,
      "loss": 0.4575,
      "step": 315
    },
    {
      "epoch": 0.2982292637465051,
      "grad_norm": 0.5552381000432096,
      "learning_rate": 4.968944099378882e-05,
      "loss": 0.4427,
      "step": 320
    },
    {
      "epoch": 0.30288909599254427,
      "grad_norm": 0.3693124262013228,
      "learning_rate": 4.994822229892993e-05,
      "loss": 0.4405,
      "step": 325
    },
    {
      "epoch": 0.30754892823858343,
      "grad_norm": 0.40179673300156293,
      "learning_rate": 4.986192613047981e-05,
      "loss": 0.4512,
      "step": 330
    },
    {
      "epoch": 0.31220876048462254,
      "grad_norm": 0.6909331905128122,
      "learning_rate": 4.977562996202969e-05,
      "loss": 0.4518,
      "step": 335
    },
    {
      "epoch": 0.3168685927306617,
      "grad_norm": 0.36848158859890456,
      "learning_rate": 4.968933379357957e-05,
      "loss": 0.4527,
      "step": 340
    },
    {
      "epoch": 0.32152842497670087,
      "grad_norm": 0.43881556176124087,
      "learning_rate": 4.9603037625129445e-05,
      "loss": 0.4475,
      "step": 345
    },
    {
      "epoch": 0.32618825722274,
      "grad_norm": 0.38991398427916507,
      "learning_rate": 4.951674145667933e-05,
      "loss": 0.461,
      "step": 350
    },
    {
      "epoch": 0.33084808946877914,
      "grad_norm": 0.47212015510204275,
      "learning_rate": 4.94304452882292e-05,
      "loss": 0.4472,
      "step": 355
    },
    {
      "epoch": 0.33550792171481825,
      "grad_norm": 0.44227904867586293,
      "learning_rate": 4.934414911977908e-05,
      "loss": 0.4448,
      "step": 360
    },
    {
      "epoch": 0.3401677539608574,
      "grad_norm": 0.4466830759596917,
      "learning_rate": 4.9257852951328965e-05,
      "loss": 0.4412,
      "step": 365
    },
    {
      "epoch": 0.3448275862068966,
      "grad_norm": 0.4019403755167571,
      "learning_rate": 4.917155678287884e-05,
      "loss": 0.4548,
      "step": 370
    },
    {
      "epoch": 0.3494874184529357,
      "grad_norm": 0.46020164939301034,
      "learning_rate": 4.908526061442872e-05,
      "loss": 0.441,
      "step": 375
    },
    {
      "epoch": 0.35414725069897485,
      "grad_norm": 0.41640820047800664,
      "learning_rate": 4.89989644459786e-05,
      "loss": 0.456,
      "step": 380
    },
    {
      "epoch": 0.35880708294501396,
      "grad_norm": 0.43251075009831547,
      "learning_rate": 4.891266827752848e-05,
      "loss": 0.4575,
      "step": 385
    },
    {
      "epoch": 0.3634669151910531,
      "grad_norm": 0.49587902382827737,
      "learning_rate": 4.882637210907836e-05,
      "loss": 0.4417,
      "step": 390
    },
    {
      "epoch": 0.3681267474370923,
      "grad_norm": 0.45767354851720066,
      "learning_rate": 4.874007594062824e-05,
      "loss": 0.4481,
      "step": 395
    },
    {
      "epoch": 0.3727865796831314,
      "grad_norm": 0.4109939412993998,
      "learning_rate": 4.865377977217811e-05,
      "loss": 0.4328,
      "step": 400
    },
    {
      "epoch": 0.37744641192917056,
      "grad_norm": 0.4267479303093156,
      "learning_rate": 4.8567483603728e-05,
      "loss": 0.4512,
      "step": 405
    },
    {
      "epoch": 0.3821062441752097,
      "grad_norm": 0.46022700390396754,
      "learning_rate": 4.8481187435277875e-05,
      "loss": 0.4473,
      "step": 410
    },
    {
      "epoch": 0.38676607642124883,
      "grad_norm": 0.5818788628854474,
      "learning_rate": 4.839489126682776e-05,
      "loss": 0.4338,
      "step": 415
    },
    {
      "epoch": 0.391425908667288,
      "grad_norm": 0.6348777601747888,
      "learning_rate": 4.830859509837763e-05,
      "loss": 0.4521,
      "step": 420
    },
    {
      "epoch": 0.3960857409133271,
      "grad_norm": 0.3406273073070538,
      "learning_rate": 4.822229892992751e-05,
      "loss": 0.4495,
      "step": 425
    },
    {
      "epoch": 0.40074557315936626,
      "grad_norm": 0.39490973933479806,
      "learning_rate": 4.8136002761477395e-05,
      "loss": 0.4364,
      "step": 430
    },
    {
      "epoch": 0.40540540540540543,
      "grad_norm": 0.5621814721849863,
      "learning_rate": 4.804970659302727e-05,
      "loss": 0.4497,
      "step": 435
    },
    {
      "epoch": 0.41006523765144454,
      "grad_norm": 0.5250828265047137,
      "learning_rate": 4.796341042457715e-05,
      "loss": 0.4438,
      "step": 440
    },
    {
      "epoch": 0.4147250698974837,
      "grad_norm": 0.49605694640709724,
      "learning_rate": 4.787711425612703e-05,
      "loss": 0.4384,
      "step": 445
    },
    {
      "epoch": 0.4193849021435228,
      "grad_norm": 0.49091569267750296,
      "learning_rate": 4.779081808767691e-05,
      "loss": 0.4292,
      "step": 450
    },
    {
      "epoch": 0.424044734389562,
      "grad_norm": 0.5449853039218947,
      "learning_rate": 4.770452191922679e-05,
      "loss": 0.4359,
      "step": 455
    },
    {
      "epoch": 0.42870456663560114,
      "grad_norm": 0.40595214751293,
      "learning_rate": 4.761822575077667e-05,
      "loss": 0.4327,
      "step": 460
    },
    {
      "epoch": 0.43336439888164024,
      "grad_norm": 0.4959979164579244,
      "learning_rate": 4.753192958232654e-05,
      "loss": 0.4297,
      "step": 465
    },
    {
      "epoch": 0.4380242311276794,
      "grad_norm": 0.42025573002718436,
      "learning_rate": 4.744563341387643e-05,
      "loss": 0.4426,
      "step": 470
    },
    {
      "epoch": 0.4426840633737186,
      "grad_norm": 0.4082690281031086,
      "learning_rate": 4.7359337245426306e-05,
      "loss": 0.432,
      "step": 475
    },
    {
      "epoch": 0.4473438956197577,
      "grad_norm": 0.46608619482702135,
      "learning_rate": 4.7273041076976184e-05,
      "loss": 0.4441,
      "step": 480
    },
    {
      "epoch": 0.45200372786579684,
      "grad_norm": 0.41782415569185494,
      "learning_rate": 4.718674490852606e-05,
      "loss": 0.4367,
      "step": 485
    },
    {
      "epoch": 0.45666356011183595,
      "grad_norm": 0.4159441336334524,
      "learning_rate": 4.710044874007594e-05,
      "loss": 0.431,
      "step": 490
    },
    {
      "epoch": 0.4613233923578751,
      "grad_norm": 0.4267635887639151,
      "learning_rate": 4.7014152571625826e-05,
      "loss": 0.437,
      "step": 495
    },
    {
      "epoch": 0.4659832246039143,
      "grad_norm": 0.536368733946403,
      "learning_rate": 4.6927856403175704e-05,
      "loss": 0.4398,
      "step": 500
    },
    {
      "epoch": 0.4706430568499534,
      "grad_norm": 0.38365822324922083,
      "learning_rate": 4.684156023472558e-05,
      "loss": 0.4264,
      "step": 505
    },
    {
      "epoch": 0.47530288909599255,
      "grad_norm": 0.49638181546901267,
      "learning_rate": 4.675526406627546e-05,
      "loss": 0.4349,
      "step": 510
    },
    {
      "epoch": 0.47996272134203166,
      "grad_norm": 0.6039980202672868,
      "learning_rate": 4.666896789782534e-05,
      "loss": 0.4329,
      "step": 515
    },
    {
      "epoch": 0.4846225535880708,
      "grad_norm": 0.531071639674577,
      "learning_rate": 4.658267172937522e-05,
      "loss": 0.4367,
      "step": 520
    },
    {
      "epoch": 0.48928238583411,
      "grad_norm": 0.41927620753314093,
      "learning_rate": 4.64963755609251e-05,
      "loss": 0.4356,
      "step": 525
    },
    {
      "epoch": 0.4939422180801491,
      "grad_norm": 0.49167935938181634,
      "learning_rate": 4.641007939247497e-05,
      "loss": 0.4324,
      "step": 530
    },
    {
      "epoch": 0.49860205032618826,
      "grad_norm": 0.35194515788154196,
      "learning_rate": 4.632378322402486e-05,
      "loss": 0.4327,
      "step": 535
    },
    {
      "epoch": 0.5032618825722274,
      "grad_norm": 0.3880164900857034,
      "learning_rate": 4.6237487055574736e-05,
      "loss": 0.4227,
      "step": 540
    },
    {
      "epoch": 0.5079217148182665,
      "grad_norm": 0.4647026652109177,
      "learning_rate": 4.6151190887124615e-05,
      "loss": 0.4289,
      "step": 545
    },
    {
      "epoch": 0.5125815470643057,
      "grad_norm": 0.4369224822745491,
      "learning_rate": 4.606489471867449e-05,
      "loss": 0.4292,
      "step": 550
    },
    {
      "epoch": 0.5172413793103449,
      "grad_norm": 0.3516777481993619,
      "learning_rate": 4.597859855022437e-05,
      "loss": 0.4236,
      "step": 555
    },
    {
      "epoch": 0.5219012115563839,
      "grad_norm": 0.4189980745198798,
      "learning_rate": 4.589230238177425e-05,
      "loss": 0.4442,
      "step": 560
    },
    {
      "epoch": 0.5265610438024231,
      "grad_norm": 0.4546880509441404,
      "learning_rate": 4.5806006213324134e-05,
      "loss": 0.4219,
      "step": 565
    },
    {
      "epoch": 0.5312208760484622,
      "grad_norm": 0.5415119247861921,
      "learning_rate": 4.5719710044874006e-05,
      "loss": 0.4218,
      "step": 570
    },
    {
      "epoch": 0.5358807082945014,
      "grad_norm": 0.4163996631718338,
      "learning_rate": 4.563341387642389e-05,
      "loss": 0.428,
      "step": 575
    },
    {
      "epoch": 0.5405405405405406,
      "grad_norm": 0.3619146303574906,
      "learning_rate": 4.554711770797377e-05,
      "loss": 0.428,
      "step": 580
    },
    {
      "epoch": 0.5452003727865797,
      "grad_norm": 0.4528885219717091,
      "learning_rate": 4.546082153952365e-05,
      "loss": 0.4248,
      "step": 585
    },
    {
      "epoch": 0.5498602050326188,
      "grad_norm": 0.42455500708425176,
      "learning_rate": 4.5374525371073526e-05,
      "loss": 0.4212,
      "step": 590
    },
    {
      "epoch": 0.554520037278658,
      "grad_norm": 0.39895122252871523,
      "learning_rate": 4.5288229202623404e-05,
      "loss": 0.4233,
      "step": 595
    },
    {
      "epoch": 0.5591798695246971,
      "grad_norm": 0.3990992719666481,
      "learning_rate": 4.520193303417328e-05,
      "loss": 0.4236,
      "step": 600
    },
    {
      "epoch": 0.5638397017707363,
      "grad_norm": 0.3553585945932701,
      "learning_rate": 4.511563686572317e-05,
      "loss": 0.4223,
      "step": 605
    },
    {
      "epoch": 0.5684995340167754,
      "grad_norm": 0.42046831338605023,
      "learning_rate": 4.5029340697273045e-05,
      "loss": 0.4334,
      "step": 610
    },
    {
      "epoch": 0.5731593662628145,
      "grad_norm": 0.34391259032310467,
      "learning_rate": 4.4943044528822923e-05,
      "loss": 0.42,
      "step": 615
    },
    {
      "epoch": 0.5778191985088537,
      "grad_norm": 0.4087272926784166,
      "learning_rate": 4.48567483603728e-05,
      "loss": 0.417,
      "step": 620
    },
    {
      "epoch": 0.5824790307548928,
      "grad_norm": 0.3589614143705352,
      "learning_rate": 4.477045219192268e-05,
      "loss": 0.4387,
      "step": 625
    },
    {
      "epoch": 0.587138863000932,
      "grad_norm": 0.3629824785143128,
      "learning_rate": 4.4684156023472565e-05,
      "loss": 0.4294,
      "step": 630
    },
    {
      "epoch": 0.5917986952469712,
      "grad_norm": 0.47917310854378126,
      "learning_rate": 4.4597859855022436e-05,
      "loss": 0.4339,
      "step": 635
    },
    {
      "epoch": 0.5964585274930102,
      "grad_norm": 0.4782662682845663,
      "learning_rate": 4.4511563686572315e-05,
      "loss": 0.4247,
      "step": 640
    },
    {
      "epoch": 0.6011183597390494,
      "grad_norm": 0.471784972910924,
      "learning_rate": 4.44252675181222e-05,
      "loss": 0.4243,
      "step": 645
    },
    {
      "epoch": 0.6057781919850885,
      "grad_norm": 0.42122505647880576,
      "learning_rate": 4.433897134967208e-05,
      "loss": 0.433,
      "step": 650
    },
    {
      "epoch": 0.6104380242311277,
      "grad_norm": 0.4025673330141692,
      "learning_rate": 4.4252675181221956e-05,
      "loss": 0.4306,
      "step": 655
    },
    {
      "epoch": 0.6150978564771669,
      "grad_norm": 0.33637741802274024,
      "learning_rate": 4.4166379012771834e-05,
      "loss": 0.4124,
      "step": 660
    },
    {
      "epoch": 0.6197576887232059,
      "grad_norm": 0.46827398668452175,
      "learning_rate": 4.408008284432171e-05,
      "loss": 0.4132,
      "step": 665
    },
    {
      "epoch": 0.6244175209692451,
      "grad_norm": 0.4372278275475368,
      "learning_rate": 4.39937866758716e-05,
      "loss": 0.4118,
      "step": 670
    },
    {
      "epoch": 0.6290773532152842,
      "grad_norm": 0.41420214852536175,
      "learning_rate": 4.3907490507421476e-05,
      "loss": 0.4204,
      "step": 675
    },
    {
      "epoch": 0.6337371854613234,
      "grad_norm": 0.4227796669483225,
      "learning_rate": 4.382119433897135e-05,
      "loss": 0.4233,
      "step": 680
    },
    {
      "epoch": 0.6383970177073626,
      "grad_norm": 0.42862178932562206,
      "learning_rate": 4.373489817052123e-05,
      "loss": 0.4329,
      "step": 685
    },
    {
      "epoch": 0.6430568499534017,
      "grad_norm": 0.4382533539684448,
      "learning_rate": 4.364860200207111e-05,
      "loss": 0.4151,
      "step": 690
    },
    {
      "epoch": 0.6477166821994408,
      "grad_norm": 0.35948786609504135,
      "learning_rate": 4.356230583362099e-05,
      "loss": 0.4202,
      "step": 695
    },
    {
      "epoch": 0.65237651444548,
      "grad_norm": 0.35194142984278265,
      "learning_rate": 4.347600966517087e-05,
      "loss": 0.4288,
      "step": 700
    },
    {
      "epoch": 0.6570363466915191,
      "grad_norm": 0.4108441539595895,
      "learning_rate": 4.3389713496720745e-05,
      "loss": 0.4223,
      "step": 705
    },
    {
      "epoch": 0.6616961789375583,
      "grad_norm": 0.48807209838383736,
      "learning_rate": 4.330341732827063e-05,
      "loss": 0.4229,
      "step": 710
    },
    {
      "epoch": 0.6663560111835974,
      "grad_norm": 0.4177025446331477,
      "learning_rate": 4.321712115982051e-05,
      "loss": 0.4114,
      "step": 715
    },
    {
      "epoch": 0.6710158434296365,
      "grad_norm": 0.3755176399576921,
      "learning_rate": 4.3130824991370387e-05,
      "loss": 0.4217,
      "step": 720
    },
    {
      "epoch": 0.6756756756756757,
      "grad_norm": 0.36461380313474906,
      "learning_rate": 4.3044528822920265e-05,
      "loss": 0.4205,
      "step": 725
    },
    {
      "epoch": 0.6803355079217148,
      "grad_norm": 0.4251420641738401,
      "learning_rate": 4.295823265447014e-05,
      "loss": 0.4153,
      "step": 730
    },
    {
      "epoch": 0.684995340167754,
      "grad_norm": 0.4003093585866478,
      "learning_rate": 4.287193648602002e-05,
      "loss": 0.4193,
      "step": 735
    },
    {
      "epoch": 0.6896551724137931,
      "grad_norm": 0.34876369977589905,
      "learning_rate": 4.27856403175699e-05,
      "loss": 0.4129,
      "step": 740
    },
    {
      "epoch": 0.6943150046598322,
      "grad_norm": 0.3707569691409881,
      "learning_rate": 4.269934414911978e-05,
      "loss": 0.4087,
      "step": 745
    },
    {
      "epoch": 0.6989748369058714,
      "grad_norm": 0.4316626941569482,
      "learning_rate": 4.261304798066966e-05,
      "loss": 0.4135,
      "step": 750
    },
    {
      "epoch": 0.7036346691519105,
      "grad_norm": 0.39303618105943594,
      "learning_rate": 4.252675181221954e-05,
      "loss": 0.4159,
      "step": 755
    },
    {
      "epoch": 0.7082945013979497,
      "grad_norm": 0.4267487540956833,
      "learning_rate": 4.244045564376942e-05,
      "loss": 0.419,
      "step": 760
    },
    {
      "epoch": 0.7129543336439889,
      "grad_norm": 0.3664167792787496,
      "learning_rate": 4.23541594753193e-05,
      "loss": 0.403,
      "step": 765
    },
    {
      "epoch": 0.7176141658900279,
      "grad_norm": 0.3577500899585447,
      "learning_rate": 4.2267863306869176e-05,
      "loss": 0.4122,
      "step": 770
    },
    {
      "epoch": 0.7222739981360671,
      "grad_norm": 0.39507959012374977,
      "learning_rate": 4.2181567138419054e-05,
      "loss": 0.4249,
      "step": 775
    },
    {
      "epoch": 0.7269338303821062,
      "grad_norm": 0.3688889563306457,
      "learning_rate": 4.209527096996894e-05,
      "loss": 0.4094,
      "step": 780
    },
    {
      "epoch": 0.7315936626281454,
      "grad_norm": 0.3846331698540081,
      "learning_rate": 4.200897480151881e-05,
      "loss": 0.4167,
      "step": 785
    },
    {
      "epoch": 0.7362534948741846,
      "grad_norm": 0.43990146629985694,
      "learning_rate": 4.1922678633068695e-05,
      "loss": 0.42,
      "step": 790
    },
    {
      "epoch": 0.7409133271202236,
      "grad_norm": 0.4264573932403777,
      "learning_rate": 4.1836382464618573e-05,
      "loss": 0.4142,
      "step": 795
    },
    {
      "epoch": 0.7455731593662628,
      "grad_norm": 0.43741863957364285,
      "learning_rate": 4.175008629616845e-05,
      "loss": 0.4124,
      "step": 800
    },
    {
      "epoch": 0.750232991612302,
      "grad_norm": 0.38911299384649556,
      "learning_rate": 4.166379012771833e-05,
      "loss": 0.4122,
      "step": 805
    },
    {
      "epoch": 0.7548928238583411,
      "grad_norm": 0.34798977131970454,
      "learning_rate": 4.157749395926821e-05,
      "loss": 0.4304,
      "step": 810
    },
    {
      "epoch": 0.7595526561043803,
      "grad_norm": 0.360973732965553,
      "learning_rate": 4.1491197790818086e-05,
      "loss": 0.4048,
      "step": 815
    },
    {
      "epoch": 0.7642124883504194,
      "grad_norm": 0.5342748114831669,
      "learning_rate": 4.140490162236797e-05,
      "loss": 0.4202,
      "step": 820
    },
    {
      "epoch": 0.7688723205964585,
      "grad_norm": 0.4164036109796091,
      "learning_rate": 4.131860545391785e-05,
      "loss": 0.4124,
      "step": 825
    },
    {
      "epoch": 0.7735321528424977,
      "grad_norm": 0.3661344541633312,
      "learning_rate": 4.123230928546773e-05,
      "loss": 0.4071,
      "step": 830
    },
    {
      "epoch": 0.7781919850885368,
      "grad_norm": 0.3754811730264276,
      "learning_rate": 4.1146013117017606e-05,
      "loss": 0.4142,
      "step": 835
    },
    {
      "epoch": 0.782851817334576,
      "grad_norm": 0.419200409043472,
      "learning_rate": 4.1059716948567484e-05,
      "loss": 0.4173,
      "step": 840
    },
    {
      "epoch": 0.7875116495806151,
      "grad_norm": 0.43893714450759036,
      "learning_rate": 4.097342078011737e-05,
      "loss": 0.4253,
      "step": 845
    },
    {
      "epoch": 0.7921714818266542,
      "grad_norm": 0.41322614598162755,
      "learning_rate": 4.088712461166724e-05,
      "loss": 0.4151,
      "step": 850
    },
    {
      "epoch": 0.7968313140726934,
      "grad_norm": 0.5117391056019034,
      "learning_rate": 4.080082844321712e-05,
      "loss": 0.417,
      "step": 855
    },
    {
      "epoch": 0.8014911463187325,
      "grad_norm": 0.39482750255468374,
      "learning_rate": 4.0714532274767004e-05,
      "loss": 0.4239,
      "step": 860
    },
    {
      "epoch": 0.8061509785647717,
      "grad_norm": 0.35016978679527583,
      "learning_rate": 4.062823610631688e-05,
      "loss": 0.405,
      "step": 865
    },
    {
      "epoch": 0.8108108108108109,
      "grad_norm": 0.32094903787924467,
      "learning_rate": 4.054193993786676e-05,
      "loss": 0.4194,
      "step": 870
    },
    {
      "epoch": 0.8154706430568499,
      "grad_norm": 0.36889819669973173,
      "learning_rate": 4.045564376941664e-05,
      "loss": 0.4149,
      "step": 875
    },
    {
      "epoch": 0.8201304753028891,
      "grad_norm": 0.3880605400422551,
      "learning_rate": 4.036934760096652e-05,
      "loss": 0.4059,
      "step": 880
    },
    {
      "epoch": 0.8247903075489282,
      "grad_norm": 0.377773246265045,
      "learning_rate": 4.02830514325164e-05,
      "loss": 0.42,
      "step": 885
    },
    {
      "epoch": 0.8294501397949674,
      "grad_norm": 0.38824868030619064,
      "learning_rate": 4.019675526406628e-05,
      "loss": 0.4067,
      "step": 890
    },
    {
      "epoch": 0.8341099720410066,
      "grad_norm": 0.34547566635591975,
      "learning_rate": 4.011045909561615e-05,
      "loss": 0.4261,
      "step": 895
    },
    {
      "epoch": 0.8387698042870456,
      "grad_norm": 0.3630734939739242,
      "learning_rate": 4.0024162927166037e-05,
      "loss": 0.409,
      "step": 900
    },
    {
      "epoch": 0.8434296365330848,
      "grad_norm": 0.3215149282810052,
      "learning_rate": 3.9937866758715915e-05,
      "loss": 0.4105,
      "step": 905
    },
    {
      "epoch": 0.848089468779124,
      "grad_norm": 0.5156466838457244,
      "learning_rate": 3.98515705902658e-05,
      "loss": 0.4167,
      "step": 910
    },
    {
      "epoch": 0.8527493010251631,
      "grad_norm": 0.58719622036636,
      "learning_rate": 3.976527442181567e-05,
      "loss": 0.4094,
      "step": 915
    },
    {
      "epoch": 0.8574091332712023,
      "grad_norm": 0.44663910672199086,
      "learning_rate": 3.967897825336555e-05,
      "loss": 0.4158,
      "step": 920
    },
    {
      "epoch": 0.8620689655172413,
      "grad_norm": 0.3593752942361397,
      "learning_rate": 3.9592682084915434e-05,
      "loss": 0.4024,
      "step": 925
    },
    {
      "epoch": 0.8667287977632805,
      "grad_norm": 0.5114184723138864,
      "learning_rate": 3.950638591646531e-05,
      "loss": 0.402,
      "step": 930
    },
    {
      "epoch": 0.8713886300093197,
      "grad_norm": 0.3926412761807418,
      "learning_rate": 3.942008974801519e-05,
      "loss": 0.4139,
      "step": 935
    },
    {
      "epoch": 0.8760484622553588,
      "grad_norm": 0.3111102224238461,
      "learning_rate": 3.933379357956507e-05,
      "loss": 0.4095,
      "step": 940
    },
    {
      "epoch": 0.880708294501398,
      "grad_norm": 0.42810427582045707,
      "learning_rate": 3.924749741111495e-05,
      "loss": 0.4117,
      "step": 945
    },
    {
      "epoch": 0.8853681267474371,
      "grad_norm": 0.48569011616159047,
      "learning_rate": 3.916120124266483e-05,
      "loss": 0.4195,
      "step": 950
    },
    {
      "epoch": 0.8900279589934762,
      "grad_norm": 0.4190117006397043,
      "learning_rate": 3.9074905074214704e-05,
      "loss": 0.4184,
      "step": 955
    },
    {
      "epoch": 0.8946877912395154,
      "grad_norm": 0.33914591782346504,
      "learning_rate": 3.898860890576458e-05,
      "loss": 0.4137,
      "step": 960
    },
    {
      "epoch": 0.8993476234855545,
      "grad_norm": 0.43035587029435407,
      "learning_rate": 3.890231273731447e-05,
      "loss": 0.4053,
      "step": 965
    },
    {
      "epoch": 0.9040074557315937,
      "grad_norm": 0.3101483588886324,
      "learning_rate": 3.8816016568864345e-05,
      "loss": 0.4031,
      "step": 970
    },
    {
      "epoch": 0.9086672879776329,
      "grad_norm": 0.3322568146796224,
      "learning_rate": 3.8729720400414224e-05,
      "loss": 0.422,
      "step": 975
    },
    {
      "epoch": 0.9133271202236719,
      "grad_norm": 0.3917813978291434,
      "learning_rate": 3.86434242319641e-05,
      "loss": 0.4091,
      "step": 980
    },
    {
      "epoch": 0.9179869524697111,
      "grad_norm": 0.4742078695343531,
      "learning_rate": 3.855712806351398e-05,
      "loss": 0.4103,
      "step": 985
    },
    {
      "epoch": 0.9226467847157502,
      "grad_norm": 0.45678911380863857,
      "learning_rate": 3.8470831895063865e-05,
      "loss": 0.4079,
      "step": 990
    },
    {
      "epoch": 0.9273066169617894,
      "grad_norm": 0.4015797894562155,
      "learning_rate": 3.838453572661374e-05,
      "loss": 0.4036,
      "step": 995
    },
    {
      "epoch": 0.9319664492078286,
      "grad_norm": 0.4393074039612925,
      "learning_rate": 3.8298239558163615e-05,
      "loss": 0.4066,
      "step": 1000
    },
    {
      "epoch": 0.9366262814538676,
      "grad_norm": 0.401959095434801,
      "learning_rate": 3.82119433897135e-05,
      "loss": 0.4113,
      "step": 1005
    },
    {
      "epoch": 0.9412861136999068,
      "grad_norm": 0.38718845936575524,
      "learning_rate": 3.812564722126338e-05,
      "loss": 0.4042,
      "step": 1010
    },
    {
      "epoch": 0.9459459459459459,
      "grad_norm": 0.40104704083420933,
      "learning_rate": 3.8039351052813256e-05,
      "loss": 0.4126,
      "step": 1015
    },
    {
      "epoch": 0.9506057781919851,
      "grad_norm": 0.4415733445950018,
      "learning_rate": 3.7953054884363134e-05,
      "loss": 0.4063,
      "step": 1020
    },
    {
      "epoch": 0.9552656104380243,
      "grad_norm": 0.37032743287501774,
      "learning_rate": 3.786675871591301e-05,
      "loss": 0.41,
      "step": 1025
    },
    {
      "epoch": 0.9599254426840633,
      "grad_norm": 0.3024002245710326,
      "learning_rate": 3.77804625474629e-05,
      "loss": 0.4151,
      "step": 1030
    },
    {
      "epoch": 0.9645852749301025,
      "grad_norm": 0.43466697925061876,
      "learning_rate": 3.7694166379012776e-05,
      "loss": 0.4074,
      "step": 1035
    },
    {
      "epoch": 0.9692451071761417,
      "grad_norm": 0.4045497692524809,
      "learning_rate": 3.7607870210562654e-05,
      "loss": 0.4051,
      "step": 1040
    },
    {
      "epoch": 0.9739049394221808,
      "grad_norm": 0.2870249098276766,
      "learning_rate": 3.752157404211253e-05,
      "loss": 0.3975,
      "step": 1045
    },
    {
      "epoch": 0.97856477166822,
      "grad_norm": 0.44417765948538723,
      "learning_rate": 3.743527787366241e-05,
      "loss": 0.4273,
      "step": 1050
    },
    {
      "epoch": 0.983224603914259,
      "grad_norm": 0.3866227301554364,
      "learning_rate": 3.734898170521229e-05,
      "loss": 0.4049,
      "step": 1055
    },
    {
      "epoch": 0.9878844361602982,
      "grad_norm": 0.3834591341930073,
      "learning_rate": 3.7262685536762174e-05,
      "loss": 0.4041,
      "step": 1060
    },
    {
      "epoch": 0.9925442684063374,
      "grad_norm": 0.3608189448212277,
      "learning_rate": 3.7176389368312045e-05,
      "loss": 0.4082,
      "step": 1065
    },
    {
      "epoch": 0.9972041006523765,
      "grad_norm": 0.4677480050799797,
      "learning_rate": 3.709009319986193e-05,
      "loss": 0.4086,
      "step": 1070
    }
  ],
  "logging_steps": 5,
  "max_steps": 3219,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 9.186429923093381e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}