{
  "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": 59.78214918326703,
      "learning_rate": 7.763975155279503e-07,
      "loss": 10.6865,
      "step": 5
    },
    {
      "epoch": 0.009319664492078284,
      "grad_norm": 61.61821429517621,
      "learning_rate": 1.5527950310559006e-06,
      "loss": 10.4783,
      "step": 10
    },
    {
      "epoch": 0.013979496738117428,
      "grad_norm": 100.18886340338842,
      "learning_rate": 2.329192546583851e-06,
      "loss": 8.8595,
      "step": 15
    },
    {
      "epoch": 0.01863932898415657,
      "grad_norm": 28.80282471816165,
      "learning_rate": 3.1055900621118013e-06,
      "loss": 2.9699,
      "step": 20
    },
    {
      "epoch": 0.023299161230195712,
      "grad_norm": 2.9469993742141027,
      "learning_rate": 3.881987577639752e-06,
      "loss": 1.3152,
      "step": 25
    },
    {
      "epoch": 0.027958993476234855,
      "grad_norm": 1.273876748177396,
      "learning_rate": 4.658385093167702e-06,
      "loss": 0.9979,
      "step": 30
    },
    {
      "epoch": 0.032618825722273995,
      "grad_norm": 1.1847847745298432,
      "learning_rate": 5.4347826086956525e-06,
      "loss": 0.8462,
      "step": 35
    },
    {
      "epoch": 0.03727865796831314,
      "grad_norm": 0.6184161495875878,
      "learning_rate": 6.2111801242236025e-06,
      "loss": 0.7823,
      "step": 40
    },
    {
      "epoch": 0.04193849021435228,
      "grad_norm": 0.4611063105597401,
      "learning_rate": 6.9875776397515525e-06,
      "loss": 0.7195,
      "step": 45
    },
    {
      "epoch": 0.046598322460391424,
      "grad_norm": 0.3903179820498685,
      "learning_rate": 7.763975155279503e-06,
      "loss": 0.6868,
      "step": 50
    },
    {
      "epoch": 0.05125815470643057,
      "grad_norm": 0.44815192407139737,
      "learning_rate": 8.540372670807453e-06,
      "loss": 0.6483,
      "step": 55
    },
    {
      "epoch": 0.05591798695246971,
      "grad_norm": 0.3363352375310443,
      "learning_rate": 9.316770186335403e-06,
      "loss": 0.61,
      "step": 60
    },
    {
      "epoch": 0.06057781919850885,
      "grad_norm": 0.31214409078055283,
      "learning_rate": 1.0093167701863353e-05,
      "loss": 0.5932,
      "step": 65
    },
    {
      "epoch": 0.06523765144454799,
      "grad_norm": 0.32957450493165075,
      "learning_rate": 1.0869565217391305e-05,
      "loss": 0.5886,
      "step": 70
    },
    {
      "epoch": 0.06989748369058714,
      "grad_norm": 0.3311294918877853,
      "learning_rate": 1.1645962732919255e-05,
      "loss": 0.5704,
      "step": 75
    },
    {
      "epoch": 0.07455731593662628,
      "grad_norm": 0.3181163049876225,
      "learning_rate": 1.2422360248447205e-05,
      "loss": 0.5604,
      "step": 80
    },
    {
      "epoch": 0.07921714818266543,
      "grad_norm": 0.3310250029185392,
      "learning_rate": 1.3198757763975155e-05,
      "loss": 0.5508,
      "step": 85
    },
    {
      "epoch": 0.08387698042870456,
      "grad_norm": 0.2864578039803888,
      "learning_rate": 1.3975155279503105e-05,
      "loss": 0.5606,
      "step": 90
    },
    {
      "epoch": 0.08853681267474371,
      "grad_norm": 0.2527763382663063,
      "learning_rate": 1.4751552795031057e-05,
      "loss": 0.542,
      "step": 95
    },
    {
      "epoch": 0.09319664492078285,
      "grad_norm": 0.30977172064299785,
      "learning_rate": 1.5527950310559007e-05,
      "loss": 0.5272,
      "step": 100
    },
    {
      "epoch": 0.097856477166822,
      "grad_norm": 0.3503262906800675,
      "learning_rate": 1.630434782608696e-05,
      "loss": 0.5214,
      "step": 105
    },
    {
      "epoch": 0.10251630941286113,
      "grad_norm": 0.3032318880335728,
      "learning_rate": 1.7080745341614907e-05,
      "loss": 0.5229,
      "step": 110
    },
    {
      "epoch": 0.10717614165890028,
      "grad_norm": 0.32192998759757896,
      "learning_rate": 1.785714285714286e-05,
      "loss": 0.5201,
      "step": 115
    },
    {
      "epoch": 0.11183597390493942,
      "grad_norm": 0.37017049044979194,
      "learning_rate": 1.8633540372670807e-05,
      "loss": 0.5067,
      "step": 120
    },
    {
      "epoch": 0.11649580615097857,
      "grad_norm": 0.27625624649694025,
      "learning_rate": 1.940993788819876e-05,
      "loss": 0.5026,
      "step": 125
    },
    {
      "epoch": 0.1211556383970177,
      "grad_norm": 0.32732522406326287,
      "learning_rate": 2.0186335403726707e-05,
      "loss": 0.5083,
      "step": 130
    },
    {
      "epoch": 0.12581547064305684,
      "grad_norm": 0.3974302327759709,
      "learning_rate": 2.096273291925466e-05,
      "loss": 0.5069,
      "step": 135
    },
    {
      "epoch": 0.13047530288909598,
      "grad_norm": 0.49055099062465235,
      "learning_rate": 2.173913043478261e-05,
      "loss": 0.4918,
      "step": 140
    },
    {
      "epoch": 0.13513513513513514,
      "grad_norm": 0.3509510737287038,
      "learning_rate": 2.2515527950310562e-05,
      "loss": 0.5182,
      "step": 145
    },
    {
      "epoch": 0.13979496738117428,
      "grad_norm": 0.4060738145091598,
      "learning_rate": 2.329192546583851e-05,
      "loss": 0.4924,
      "step": 150
    },
    {
      "epoch": 0.14445479962721341,
      "grad_norm": 0.42238178931670933,
      "learning_rate": 2.4068322981366462e-05,
      "loss": 0.5005,
      "step": 155
    },
    {
      "epoch": 0.14911463187325255,
      "grad_norm": 0.42361270461040995,
      "learning_rate": 2.484472049689441e-05,
      "loss": 0.4809,
      "step": 160
    },
    {
      "epoch": 0.15377446411929171,
      "grad_norm": 0.4419148082648927,
      "learning_rate": 2.5621118012422362e-05,
      "loss": 0.4922,
      "step": 165
    },
    {
      "epoch": 0.15843429636533085,
      "grad_norm": 0.37817457175825797,
      "learning_rate": 2.639751552795031e-05,
      "loss": 0.4682,
      "step": 170
    },
    {
      "epoch": 0.16309412861137,
      "grad_norm": 0.4612740179437555,
      "learning_rate": 2.7173913043478262e-05,
      "loss": 0.4812,
      "step": 175
    },
    {
      "epoch": 0.16775396085740912,
      "grad_norm": 0.4027204736632852,
      "learning_rate": 2.795031055900621e-05,
      "loss": 0.4743,
      "step": 180
    },
    {
      "epoch": 0.1724137931034483,
      "grad_norm": 0.3662916369622068,
      "learning_rate": 2.8726708074534165e-05,
      "loss": 0.4771,
      "step": 185
    },
    {
      "epoch": 0.17707362534948742,
      "grad_norm": 0.44549022951820444,
      "learning_rate": 2.9503105590062114e-05,
      "loss": 0.4872,
      "step": 190
    },
    {
      "epoch": 0.18173345759552656,
      "grad_norm": 0.4421692278535386,
      "learning_rate": 3.0279503105590062e-05,
      "loss": 0.4768,
      "step": 195
    },
    {
      "epoch": 0.1863932898415657,
      "grad_norm": 0.4592171701659634,
      "learning_rate": 3.1055900621118014e-05,
      "loss": 0.4782,
      "step": 200
    },
    {
      "epoch": 0.19105312208760486,
      "grad_norm": 0.5338610618981041,
      "learning_rate": 3.183229813664597e-05,
      "loss": 0.4677,
      "step": 205
    },
    {
      "epoch": 0.195712954333644,
      "grad_norm": 0.679375515050287,
      "learning_rate": 3.260869565217392e-05,
      "loss": 0.4817,
      "step": 210
    },
    {
      "epoch": 0.20037278657968313,
      "grad_norm": 0.5075771202954662,
      "learning_rate": 3.3385093167701865e-05,
      "loss": 0.4632,
      "step": 215
    },
    {
      "epoch": 0.20503261882572227,
      "grad_norm": 0.5271972882853628,
      "learning_rate": 3.4161490683229814e-05,
      "loss": 0.4674,
      "step": 220
    },
    {
      "epoch": 0.2096924510717614,
      "grad_norm": 0.45927485782401883,
      "learning_rate": 3.493788819875777e-05,
      "loss": 0.4496,
      "step": 225
    },
    {
      "epoch": 0.21435228331780057,
      "grad_norm": 0.3875430276374643,
      "learning_rate": 3.571428571428572e-05,
      "loss": 0.4628,
      "step": 230
    },
    {
      "epoch": 0.2190121155638397,
      "grad_norm": 0.43592470909651004,
      "learning_rate": 3.6490683229813665e-05,
      "loss": 0.4604,
      "step": 235
    },
    {
      "epoch": 0.22367194780987884,
      "grad_norm": 0.45541861707423287,
      "learning_rate": 3.7267080745341614e-05,
      "loss": 0.4578,
      "step": 240
    },
    {
      "epoch": 0.22833178005591798,
      "grad_norm": 0.5415802082513514,
      "learning_rate": 3.804347826086957e-05,
      "loss": 0.4628,
      "step": 245
    },
    {
      "epoch": 0.23299161230195714,
      "grad_norm": 0.48756789908879733,
      "learning_rate": 3.881987577639752e-05,
      "loss": 0.4551,
      "step": 250
    },
    {
      "epoch": 0.23765144454799628,
      "grad_norm": 0.48641029082339876,
      "learning_rate": 3.9596273291925465e-05,
      "loss": 0.4636,
      "step": 255
    },
    {
      "epoch": 0.2423112767940354,
      "grad_norm": 0.44167537560377024,
      "learning_rate": 4.0372670807453414e-05,
      "loss": 0.4584,
      "step": 260
    },
    {
      "epoch": 0.24697110904007455,
      "grad_norm": 0.49523500628083084,
      "learning_rate": 4.114906832298137e-05,
      "loss": 0.457,
      "step": 265
    },
    {
      "epoch": 0.2516309412861137,
      "grad_norm": 0.46362590765498735,
      "learning_rate": 4.192546583850932e-05,
      "loss": 0.4553,
      "step": 270
    },
    {
      "epoch": 0.25629077353215285,
      "grad_norm": 0.41387594745783063,
      "learning_rate": 4.270186335403727e-05,
      "loss": 0.4606,
      "step": 275
    },
    {
      "epoch": 0.26095060577819196,
      "grad_norm": 0.4327868844195844,
      "learning_rate": 4.347826086956522e-05,
      "loss": 0.4529,
      "step": 280
    },
    {
      "epoch": 0.2656104380242311,
      "grad_norm": 0.425878793618421,
      "learning_rate": 4.425465838509317e-05,
      "loss": 0.457,
      "step": 285
    },
    {
      "epoch": 0.2702702702702703,
      "grad_norm": 0.49167640847055744,
      "learning_rate": 4.5031055900621124e-05,
      "loss": 0.4586,
      "step": 290
    },
    {
      "epoch": 0.2749301025163094,
      "grad_norm": 0.6223762427590684,
      "learning_rate": 4.580745341614907e-05,
      "loss": 0.4627,
      "step": 295
    },
    {
      "epoch": 0.27958993476234856,
      "grad_norm": 0.6138447122463503,
      "learning_rate": 4.658385093167702e-05,
      "loss": 0.4643,
      "step": 300
    },
    {
      "epoch": 0.2842497670083877,
      "grad_norm": 0.5410077879825527,
      "learning_rate": 4.736024844720497e-05,
      "loss": 0.4531,
      "step": 305
    },
    {
      "epoch": 0.28890959925442683,
      "grad_norm": 0.5248106064020833,
      "learning_rate": 4.8136645962732924e-05,
      "loss": 0.4453,
      "step": 310
    },
    {
      "epoch": 0.293569431500466,
      "grad_norm": 0.4576312238038398,
      "learning_rate": 4.891304347826087e-05,
      "loss": 0.4531,
      "step": 315
    },
    {
      "epoch": 0.2982292637465051,
      "grad_norm": 0.7616375740291316,
      "learning_rate": 4.968944099378882e-05,
      "loss": 0.4387,
      "step": 320
    },
    {
      "epoch": 0.30288909599254427,
      "grad_norm": 0.5230246761762287,
      "learning_rate": 4.994822229892993e-05,
      "loss": 0.4367,
      "step": 325
    },
    {
      "epoch": 0.30754892823858343,
      "grad_norm": 0.4333888474147187,
      "learning_rate": 4.986192613047981e-05,
      "loss": 0.4468,
      "step": 330
    },
    {
      "epoch": 0.31220876048462254,
      "grad_norm": 0.6558069118015527,
      "learning_rate": 4.977562996202969e-05,
      "loss": 0.447,
      "step": 335
    },
    {
      "epoch": 0.3168685927306617,
      "grad_norm": 0.4781469048268466,
      "learning_rate": 4.968933379357957e-05,
      "loss": 0.4487,
      "step": 340
    },
    {
      "epoch": 0.32152842497670087,
      "grad_norm": 0.4582486158349481,
      "learning_rate": 4.9603037625129445e-05,
      "loss": 0.4433,
      "step": 345
    },
    {
      "epoch": 0.32618825722274,
      "grad_norm": 0.4234664780262906,
      "learning_rate": 4.951674145667933e-05,
      "loss": 0.4568,
      "step": 350
    },
    {
      "epoch": 0.33084808946877914,
      "grad_norm": 0.4766441062632173,
      "learning_rate": 4.94304452882292e-05,
      "loss": 0.4431,
      "step": 355
    },
    {
      "epoch": 0.33550792171481825,
      "grad_norm": 0.5294140666427822,
      "learning_rate": 4.934414911977908e-05,
      "loss": 0.4405,
      "step": 360
    },
    {
      "epoch": 0.3401677539608574,
      "grad_norm": 0.3838225157756671,
      "learning_rate": 4.9257852951328965e-05,
      "loss": 0.437,
      "step": 365
    },
    {
      "epoch": 0.3448275862068966,
      "grad_norm": 0.4289011151413574,
      "learning_rate": 4.917155678287884e-05,
      "loss": 0.4498,
      "step": 370
    },
    {
      "epoch": 0.3494874184529357,
      "grad_norm": 0.4739139367556648,
      "learning_rate": 4.908526061442872e-05,
      "loss": 0.4364,
      "step": 375
    },
    {
      "epoch": 0.35414725069897485,
      "grad_norm": 0.6531685734098429,
      "learning_rate": 4.89989644459786e-05,
      "loss": 0.4507,
      "step": 380
    },
    {
      "epoch": 0.35880708294501396,
      "grad_norm": 0.38736077721702633,
      "learning_rate": 4.891266827752848e-05,
      "loss": 0.4526,
      "step": 385
    },
    {
      "epoch": 0.3634669151910531,
      "grad_norm": 0.39074723361967706,
      "learning_rate": 4.882637210907836e-05,
      "loss": 0.4372,
      "step": 390
    },
    {
      "epoch": 0.3681267474370923,
      "grad_norm": 0.4998720672786241,
      "learning_rate": 4.874007594062824e-05,
      "loss": 0.4432,
      "step": 395
    },
    {
      "epoch": 0.3727865796831314,
      "grad_norm": 0.41449696680501236,
      "learning_rate": 4.865377977217811e-05,
      "loss": 0.4284,
      "step": 400
    },
    {
      "epoch": 0.37744641192917056,
      "grad_norm": 0.42798981510455325,
      "learning_rate": 4.8567483603728e-05,
      "loss": 0.4466,
      "step": 405
    },
    {
      "epoch": 0.3821062441752097,
      "grad_norm": 0.453510571603012,
      "learning_rate": 4.8481187435277875e-05,
      "loss": 0.4425,
      "step": 410
    },
    {
      "epoch": 0.38676607642124883,
      "grad_norm": 0.5613852564226437,
      "learning_rate": 4.839489126682776e-05,
      "loss": 0.4296,
      "step": 415
    },
    {
      "epoch": 0.391425908667288,
      "grad_norm": 0.4991311823287778,
      "learning_rate": 4.830859509837763e-05,
      "loss": 0.4477,
      "step": 420
    },
    {
      "epoch": 0.3960857409133271,
      "grad_norm": 0.41710093216407557,
      "learning_rate": 4.822229892992751e-05,
      "loss": 0.4451,
      "step": 425
    },
    {
      "epoch": 0.40074557315936626,
      "grad_norm": 0.44852195468081424,
      "learning_rate": 4.8136002761477395e-05,
      "loss": 0.4322,
      "step": 430
    },
    {
      "epoch": 0.40540540540540543,
      "grad_norm": 0.5873258297489329,
      "learning_rate": 4.804970659302727e-05,
      "loss": 0.445,
      "step": 435
    },
    {
      "epoch": 0.41006523765144454,
      "grad_norm": 0.5301184440251494,
      "learning_rate": 4.796341042457715e-05,
      "loss": 0.439,
      "step": 440
    },
    {
      "epoch": 0.4147250698974837,
      "grad_norm": 0.554152067322795,
      "learning_rate": 4.787711425612703e-05,
      "loss": 0.4337,
      "step": 445
    },
    {
      "epoch": 0.4193849021435228,
      "grad_norm": 0.4875794890348032,
      "learning_rate": 4.779081808767691e-05,
      "loss": 0.4245,
      "step": 450
    },
    {
      "epoch": 0.424044734389562,
      "grad_norm": 0.4690783572871423,
      "learning_rate": 4.770452191922679e-05,
      "loss": 0.4314,
      "step": 455
    },
    {
      "epoch": 0.42870456663560114,
      "grad_norm": 0.3725289513240759,
      "learning_rate": 4.761822575077667e-05,
      "loss": 0.4283,
      "step": 460
    },
    {
      "epoch": 0.43336439888164024,
      "grad_norm": 0.4830268598668616,
      "learning_rate": 4.753192958232654e-05,
      "loss": 0.4255,
      "step": 465
    },
    {
      "epoch": 0.4380242311276794,
      "grad_norm": 0.43173494250112954,
      "learning_rate": 4.744563341387643e-05,
      "loss": 0.4378,
      "step": 470
    },
    {
      "epoch": 0.4426840633737186,
      "grad_norm": 0.43237002431737065,
      "learning_rate": 4.7359337245426306e-05,
      "loss": 0.4277,
      "step": 475
    },
    {
      "epoch": 0.4473438956197577,
      "grad_norm": 0.41385681702794824,
      "learning_rate": 4.7273041076976184e-05,
      "loss": 0.4394,
      "step": 480
    },
    {
      "epoch": 0.45200372786579684,
      "grad_norm": 0.40157124060011173,
      "learning_rate": 4.718674490852606e-05,
      "loss": 0.432,
      "step": 485
    },
    {
      "epoch": 0.45666356011183595,
      "grad_norm": 0.39938983254093463,
      "learning_rate": 4.710044874007594e-05,
      "loss": 0.4264,
      "step": 490
    },
    {
      "epoch": 0.4613233923578751,
      "grad_norm": 0.39732323279012777,
      "learning_rate": 4.7014152571625826e-05,
      "loss": 0.4321,
      "step": 495
    },
    {
      "epoch": 0.4659832246039143,
      "grad_norm": 0.4747358464791143,
      "learning_rate": 4.6927856403175704e-05,
      "loss": 0.435,
      "step": 500
    },
    {
      "epoch": 0.4706430568499534,
      "grad_norm": 0.3698718174123855,
      "learning_rate": 4.684156023472558e-05,
      "loss": 0.4221,
      "step": 505
    },
    {
      "epoch": 0.47530288909599255,
      "grad_norm": 0.4305572996344627,
      "learning_rate": 4.675526406627546e-05,
      "loss": 0.4303,
      "step": 510
    },
    {
      "epoch": 0.47996272134203166,
      "grad_norm": 0.6085259797324423,
      "learning_rate": 4.666896789782534e-05,
      "loss": 0.4281,
      "step": 515
    },
    {
      "epoch": 0.4846225535880708,
      "grad_norm": 0.5730318489213171,
      "learning_rate": 4.658267172937522e-05,
      "loss": 0.4321,
      "step": 520
    },
    {
      "epoch": 0.48928238583411,
      "grad_norm": 0.4332245949035479,
      "learning_rate": 4.64963755609251e-05,
      "loss": 0.4309,
      "step": 525
    },
    {
      "epoch": 0.4939422180801491,
      "grad_norm": 0.508102013567185,
      "learning_rate": 4.641007939247497e-05,
      "loss": 0.428,
      "step": 530
    },
    {
      "epoch": 0.49860205032618826,
      "grad_norm": 0.34669842614662666,
      "learning_rate": 4.632378322402486e-05,
      "loss": 0.4283,
      "step": 535
    },
    {
      "epoch": 0.5032618825722274,
      "grad_norm": 0.3889254150420956,
      "learning_rate": 4.6237487055574736e-05,
      "loss": 0.4178,
      "step": 540
    },
    {
      "epoch": 0.5079217148182665,
      "grad_norm": 0.49239466237923585,
      "learning_rate": 4.6151190887124615e-05,
      "loss": 0.4244,
      "step": 545
    },
    {
      "epoch": 0.5125815470643057,
      "grad_norm": 0.4397316581317278,
      "learning_rate": 4.606489471867449e-05,
      "loss": 0.4245,
      "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": 4.708780893462856e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}