{
  "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": 59.914947522200755,
      "learning_rate": 7.763975155279503e-07,
      "loss": 10.6862,
      "step": 5
    },
    {
      "epoch": 0.009319664492078284,
      "grad_norm": 61.506550472016244,
      "learning_rate": 1.5527950310559006e-06,
      "loss": 10.4784,
      "step": 10
    },
    {
      "epoch": 0.013979496738117428,
      "grad_norm": 100.03132681639305,
      "learning_rate": 2.329192546583851e-06,
      "loss": 8.8593,
      "step": 15
    },
    {
      "epoch": 0.01863932898415657,
      "grad_norm": 28.64812769413447,
      "learning_rate": 3.1055900621118013e-06,
      "loss": 2.9689,
      "step": 20
    },
    {
      "epoch": 0.023299161230195712,
      "grad_norm": 3.1458673773327392,
      "learning_rate": 3.881987577639752e-06,
      "loss": 1.3154,
      "step": 25
    },
    {
      "epoch": 0.027958993476234855,
      "grad_norm": 1.2888476895438312,
      "learning_rate": 4.658385093167702e-06,
      "loss": 0.9997,
      "step": 30
    },
    {
      "epoch": 0.032618825722273995,
      "grad_norm": 0.8962553376010008,
      "learning_rate": 5.4347826086956525e-06,
      "loss": 0.8452,
      "step": 35
    },
    {
      "epoch": 0.03727865796831314,
      "grad_norm": 0.6393961134946683,
      "learning_rate": 6.2111801242236025e-06,
      "loss": 0.7803,
      "step": 40
    },
    {
      "epoch": 0.04193849021435228,
      "grad_norm": 0.513465541203958,
      "learning_rate": 6.9875776397515525e-06,
      "loss": 0.7182,
      "step": 45
    },
    {
      "epoch": 0.046598322460391424,
      "grad_norm": 0.401362860410936,
      "learning_rate": 7.763975155279503e-06,
      "loss": 0.686,
      "step": 50
    },
    {
      "epoch": 0.05125815470643057,
      "grad_norm": 0.4467110756469191,
      "learning_rate": 8.540372670807453e-06,
      "loss": 0.6477,
      "step": 55
    },
    {
      "epoch": 0.05591798695246971,
      "grad_norm": 0.3260347816467283,
      "learning_rate": 9.316770186335403e-06,
      "loss": 0.6096,
      "step": 60
    },
    {
      "epoch": 0.06057781919850885,
      "grad_norm": 0.3077539007822434,
      "learning_rate": 1.0093167701863353e-05,
      "loss": 0.5929,
      "step": 65
    },
    {
      "epoch": 0.06523765144454799,
      "grad_norm": 0.32304663164813313,
      "learning_rate": 1.0869565217391305e-05,
      "loss": 0.5885,
      "step": 70
    },
    {
      "epoch": 0.06989748369058714,
      "grad_norm": 0.29376236300157593,
      "learning_rate": 1.1645962732919255e-05,
      "loss": 0.5703,
      "step": 75
    },
    {
      "epoch": 0.07455731593662628,
      "grad_norm": 0.3348078948651445,
      "learning_rate": 1.2422360248447205e-05,
      "loss": 0.5604,
      "step": 80
    },
    {
      "epoch": 0.07921714818266543,
      "grad_norm": 0.3426802151879899,
      "learning_rate": 1.3198757763975155e-05,
      "loss": 0.5507,
      "step": 85
    },
    {
      "epoch": 0.08387698042870456,
      "grad_norm": 0.28857538776714187,
      "learning_rate": 1.3975155279503105e-05,
      "loss": 0.5607,
      "step": 90
    },
    {
      "epoch": 0.08853681267474371,
      "grad_norm": 0.26144673946425695,
      "learning_rate": 1.4751552795031057e-05,
      "loss": 0.5421,
      "step": 95
    },
    {
      "epoch": 0.09319664492078285,
      "grad_norm": 0.31140980297277515,
      "learning_rate": 1.5527950310559007e-05,
      "loss": 0.5272,
      "step": 100
    },
    {
      "epoch": 0.097856477166822,
      "grad_norm": 0.3266343156750877,
      "learning_rate": 1.630434782608696e-05,
      "loss": 0.5215,
      "step": 105
    },
    {
      "epoch": 0.10251630941286113,
      "grad_norm": 0.29918673347327257,
      "learning_rate": 1.7080745341614907e-05,
      "loss": 0.523,
      "step": 110
    },
    {
      "epoch": 0.10717614165890028,
      "grad_norm": 0.3310333978811669,
      "learning_rate": 1.785714285714286e-05,
      "loss": 0.5202,
      "step": 115
    },
    {
      "epoch": 0.11183597390493942,
      "grad_norm": 0.36321738118381847,
      "learning_rate": 1.8633540372670807e-05,
      "loss": 0.5067,
      "step": 120
    },
    {
      "epoch": 0.11649580615097857,
      "grad_norm": 0.2797666994370987,
      "learning_rate": 1.940993788819876e-05,
      "loss": 0.5026,
      "step": 125
    },
    {
      "epoch": 0.1211556383970177,
      "grad_norm": 0.32298025318616186,
      "learning_rate": 2.0186335403726707e-05,
      "loss": 0.5084,
      "step": 130
    },
    {
      "epoch": 0.12581547064305684,
      "grad_norm": 0.4000752827733501,
      "learning_rate": 2.096273291925466e-05,
      "loss": 0.5069,
      "step": 135
    },
    {
      "epoch": 0.13047530288909598,
      "grad_norm": 0.4922924496261864,
      "learning_rate": 2.173913043478261e-05,
      "loss": 0.4919,
      "step": 140
    },
    {
      "epoch": 0.13513513513513514,
      "grad_norm": 0.3528459265317262,
      "learning_rate": 2.2515527950310562e-05,
      "loss": 0.5182,
      "step": 145
    },
    {
      "epoch": 0.13979496738117428,
      "grad_norm": 0.41310907318766044,
      "learning_rate": 2.329192546583851e-05,
      "loss": 0.4924,
      "step": 150
    },
    {
      "epoch": 0.14445479962721341,
      "grad_norm": 0.4269460447807956,
      "learning_rate": 2.4068322981366462e-05,
      "loss": 0.5005,
      "step": 155
    },
    {
      "epoch": 0.14911463187325255,
      "grad_norm": 0.4188942709163961,
      "learning_rate": 2.484472049689441e-05,
      "loss": 0.4809,
      "step": 160
    },
    {
      "epoch": 0.15377446411929171,
      "grad_norm": 0.4224538737371056,
      "learning_rate": 2.5621118012422362e-05,
      "loss": 0.4922,
      "step": 165
    },
    {
      "epoch": 0.15843429636533085,
      "grad_norm": 0.36451695943150036,
      "learning_rate": 2.639751552795031e-05,
      "loss": 0.4682,
      "step": 170
    },
    {
      "epoch": 0.16309412861137,
      "grad_norm": 0.44880938962766675,
      "learning_rate": 2.7173913043478262e-05,
      "loss": 0.4812,
      "step": 175
    },
    {
      "epoch": 0.16775396085740912,
      "grad_norm": 0.399985899809391,
      "learning_rate": 2.795031055900621e-05,
      "loss": 0.4744,
      "step": 180
    },
    {
      "epoch": 0.1724137931034483,
      "grad_norm": 0.3710762382054236,
      "learning_rate": 2.8726708074534165e-05,
      "loss": 0.4773,
      "step": 185
    },
    {
      "epoch": 0.17707362534948742,
      "grad_norm": 0.5067423386101156,
      "learning_rate": 2.9503105590062114e-05,
      "loss": 0.4874,
      "step": 190
    },
    {
      "epoch": 0.18173345759552656,
      "grad_norm": 0.6497058112525416,
      "learning_rate": 3.0279503105590062e-05,
      "loss": 0.4774,
      "step": 195
    },
    {
      "epoch": 0.1863932898415657,
      "grad_norm": 0.5885066139321359,
      "learning_rate": 3.1055900621118014e-05,
      "loss": 0.4796,
      "step": 200
    },
    {
      "epoch": 0.19105312208760486,
      "grad_norm": 1.0003664794306601,
      "learning_rate": 3.183229813664597e-05,
      "loss": 0.4687,
      "step": 205
    },
    {
      "epoch": 0.195712954333644,
      "grad_norm": 0.5038665173622903,
      "learning_rate": 3.260869565217392e-05,
      "loss": 0.4813,
      "step": 210
    },
    {
      "epoch": 0.20037278657968313,
      "grad_norm": 0.39066674417290936,
      "learning_rate": 3.3385093167701865e-05,
      "loss": 0.4632,
      "step": 215
    },
    {
      "epoch": 0.20503261882572227,
      "grad_norm": 0.5461800355437195,
      "learning_rate": 3.4161490683229814e-05,
      "loss": 0.4673,
      "step": 220
    },
    {
      "epoch": 0.2096924510717614,
      "grad_norm": 0.43194458883726633,
      "learning_rate": 3.493788819875777e-05,
      "loss": 0.4496,
      "step": 225
    },
    {
      "epoch": 0.21435228331780057,
      "grad_norm": 0.4440833071381681,
      "learning_rate": 3.571428571428572e-05,
      "loss": 0.4628,
      "step": 230
    },
    {
      "epoch": 0.2190121155638397,
      "grad_norm": 0.41566446947186825,
      "learning_rate": 3.6490683229813665e-05,
      "loss": 0.4604,
      "step": 235
    },
    {
      "epoch": 0.22367194780987884,
      "grad_norm": 0.358310843676591,
      "learning_rate": 3.7267080745341614e-05,
      "loss": 0.4578,
      "step": 240
    },
    {
      "epoch": 0.22833178005591798,
      "grad_norm": 0.44225923549559465,
      "learning_rate": 3.804347826086957e-05,
      "loss": 0.4627,
      "step": 245
    },
    {
      "epoch": 0.23299161230195714,
      "grad_norm": 0.3992256291237462,
      "learning_rate": 3.881987577639752e-05,
      "loss": 0.4552,
      "step": 250
    },
    {
      "epoch": 0.23765144454799628,
      "grad_norm": 0.4694080889691004,
      "learning_rate": 3.9596273291925465e-05,
      "loss": 0.4635,
      "step": 255
    },
    {
      "epoch": 0.2423112767940354,
      "grad_norm": 0.42023599721646526,
      "learning_rate": 4.0372670807453414e-05,
      "loss": 0.4584,
      "step": 260
    },
    {
      "epoch": 0.24697110904007455,
      "grad_norm": 0.46533332747349704,
      "learning_rate": 4.114906832298137e-05,
      "loss": 0.4568,
      "step": 265
    },
    {
      "epoch": 0.2516309412861137,
      "grad_norm": 0.4866315434700336,
      "learning_rate": 4.192546583850932e-05,
      "loss": 0.4552,
      "step": 270
    },
    {
      "epoch": 0.25629077353215285,
      "grad_norm": 0.4846502580427653,
      "learning_rate": 4.270186335403727e-05,
      "loss": 0.4606,
      "step": 275
    },
    {
      "epoch": 0.26095060577819196,
      "grad_norm": 0.4022476259260207,
      "learning_rate": 4.347826086956522e-05,
      "loss": 0.4528,
      "step": 280
    },
    {
      "epoch": 0.2656104380242311,
      "grad_norm": 0.43388622580915626,
      "learning_rate": 4.425465838509317e-05,
      "loss": 0.4567,
      "step": 285
    },
    {
      "epoch": 0.2702702702702703,
      "grad_norm": 0.4856594556844892,
      "learning_rate": 4.5031055900621124e-05,
      "loss": 0.4585,
      "step": 290
    },
    {
      "epoch": 0.2749301025163094,
      "grad_norm": 0.6213697727877028,
      "learning_rate": 4.580745341614907e-05,
      "loss": 0.4624,
      "step": 295
    },
    {
      "epoch": 0.27958993476234856,
      "grad_norm": 0.660785708911612,
      "learning_rate": 4.658385093167702e-05,
      "loss": 0.4643,
      "step": 300
    },
    {
      "epoch": 0.2842497670083877,
      "grad_norm": 0.5094248475675451,
      "learning_rate": 4.736024844720497e-05,
      "loss": 0.453,
      "step": 305
    },
    {
      "epoch": 0.28890959925442683,
      "grad_norm": 0.5162843441070357,
      "learning_rate": 4.8136645962732924e-05,
      "loss": 0.4454,
      "step": 310
    },
    {
      "epoch": 0.293569431500466,
      "grad_norm": 0.5558477904513442,
      "learning_rate": 4.891304347826087e-05,
      "loss": 0.4533,
      "step": 315
    },
    {
      "epoch": 0.2982292637465051,
      "grad_norm": 0.6726916910854334,
      "learning_rate": 4.968944099378882e-05,
      "loss": 0.4388,
      "step": 320
    },
    {
      "epoch": 0.30288909599254427,
      "grad_norm": 0.5608394756921781,
      "learning_rate": 4.994822229892993e-05,
      "loss": 0.4368,
      "step": 325
    },
    {
      "epoch": 0.30754892823858343,
      "grad_norm": 0.5247959113527437,
      "learning_rate": 4.986192613047981e-05,
      "loss": 0.4468,
      "step": 330
    },
    {
      "epoch": 0.31220876048462254,
      "grad_norm": 0.7466581909085609,
      "learning_rate": 4.977562996202969e-05,
      "loss": 0.4473,
      "step": 335
    },
    {
      "epoch": 0.3168685927306617,
      "grad_norm": 0.40069693827476377,
      "learning_rate": 4.968933379357957e-05,
      "loss": 0.4488,
      "step": 340
    },
    {
      "epoch": 0.32152842497670087,
      "grad_norm": 0.4433073418733753,
      "learning_rate": 4.9603037625129445e-05,
      "loss": 0.4433,
      "step": 345
    },
    {
      "epoch": 0.32618825722274,
      "grad_norm": 0.4402804602848506,
      "learning_rate": 4.951674145667933e-05,
      "loss": 0.4566,
      "step": 350
    },
    {
      "epoch": 0.33084808946877914,
      "grad_norm": 0.5561916690794446,
      "learning_rate": 4.94304452882292e-05,
      "loss": 0.4431,
      "step": 355
    },
    {
      "epoch": 0.33550792171481825,
      "grad_norm": 0.5845890551087048,
      "learning_rate": 4.934414911977908e-05,
      "loss": 0.4405,
      "step": 360
    },
    {
      "epoch": 0.3401677539608574,
      "grad_norm": 0.3866415610222162,
      "learning_rate": 4.9257852951328965e-05,
      "loss": 0.437,
      "step": 365
    },
    {
      "epoch": 0.3448275862068966,
      "grad_norm": 0.4523706035604396,
      "learning_rate": 4.917155678287884e-05,
      "loss": 0.45,
      "step": 370
    },
    {
      "epoch": 0.3494874184529357,
      "grad_norm": 0.45303755908012494,
      "learning_rate": 4.908526061442872e-05,
      "loss": 0.4364,
      "step": 375
    },
    {
      "epoch": 0.35414725069897485,
      "grad_norm": 0.4759810127104163,
      "learning_rate": 4.89989644459786e-05,
      "loss": 0.4508,
      "step": 380
    },
    {
      "epoch": 0.35880708294501396,
      "grad_norm": 0.37302493435804995,
      "learning_rate": 4.891266827752848e-05,
      "loss": 0.4528,
      "step": 385
    },
    {
      "epoch": 0.3634669151910531,
      "grad_norm": 0.39553345753859565,
      "learning_rate": 4.882637210907836e-05,
      "loss": 0.4374,
      "step": 390
    },
    {
      "epoch": 0.3681267474370923,
      "grad_norm": 0.517029746917124,
      "learning_rate": 4.874007594062824e-05,
      "loss": 0.4434,
      "step": 395
    },
    {
      "epoch": 0.3727865796831314,
      "grad_norm": 0.43650791449383186,
      "learning_rate": 4.865377977217811e-05,
      "loss": 0.4286,
      "step": 400
    },
    {
      "epoch": 0.37744641192917056,
      "grad_norm": 0.4287102270886215,
      "learning_rate": 4.8567483603728e-05,
      "loss": 0.4468,
      "step": 405
    },
    {
      "epoch": 0.3821062441752097,
      "grad_norm": 0.43073044657348636,
      "learning_rate": 4.8481187435277875e-05,
      "loss": 0.4426,
      "step": 410
    },
    {
      "epoch": 0.38676607642124883,
      "grad_norm": 0.5387025600608393,
      "learning_rate": 4.839489126682776e-05,
      "loss": 0.4296,
      "step": 415
    },
    {
      "epoch": 0.391425908667288,
      "grad_norm": 0.5153252513836688,
      "learning_rate": 4.830859509837763e-05,
      "loss": 0.4475,
      "step": 420
    },
    {
      "epoch": 0.3960857409133271,
      "grad_norm": 0.3664353088914084,
      "learning_rate": 4.822229892992751e-05,
      "loss": 0.445,
      "step": 425
    },
    {
      "epoch": 0.40074557315936626,
      "grad_norm": 0.4646069378801538,
      "learning_rate": 4.8136002761477395e-05,
      "loss": 0.4321,
      "step": 430
    },
    {
      "epoch": 0.40540540540540543,
      "grad_norm": 0.5337046566267092,
      "learning_rate": 4.804970659302727e-05,
      "loss": 0.4452,
      "step": 435
    },
    {
      "epoch": 0.41006523765144454,
      "grad_norm": 0.5277173274418405,
      "learning_rate": 4.796341042457715e-05,
      "loss": 0.4393,
      "step": 440
    },
    {
      "epoch": 0.4147250698974837,
      "grad_norm": 0.5328906770207152,
      "learning_rate": 4.787711425612703e-05,
      "loss": 0.4339,
      "step": 445
    },
    {
      "epoch": 0.4193849021435228,
      "grad_norm": 0.4906064823628884,
      "learning_rate": 4.779081808767691e-05,
      "loss": 0.4246,
      "step": 450
    },
    {
      "epoch": 0.424044734389562,
      "grad_norm": 0.4871704896808195,
      "learning_rate": 4.770452191922679e-05,
      "loss": 0.4316,
      "step": 455
    },
    {
      "epoch": 0.42870456663560114,
      "grad_norm": 0.3872613851854932,
      "learning_rate": 4.761822575077667e-05,
      "loss": 0.4283,
      "step": 460
    },
    {
      "epoch": 0.43336439888164024,
      "grad_norm": 0.48560447593847655,
      "learning_rate": 4.753192958232654e-05,
      "loss": 0.4255,
      "step": 465
    },
    {
      "epoch": 0.4380242311276794,
      "grad_norm": 0.43287841878863453,
      "learning_rate": 4.744563341387643e-05,
      "loss": 0.4377,
      "step": 470
    },
    {
      "epoch": 0.4426840633737186,
      "grad_norm": 0.46424167359241597,
      "learning_rate": 4.7359337245426306e-05,
      "loss": 0.4277,
      "step": 475
    },
    {
      "epoch": 0.4473438956197577,
      "grad_norm": 0.39322918496511977,
      "learning_rate": 4.7273041076976184e-05,
      "loss": 0.4393,
      "step": 480
    },
    {
      "epoch": 0.45200372786579684,
      "grad_norm": 0.38542293023219015,
      "learning_rate": 4.718674490852606e-05,
      "loss": 0.432,
      "step": 485
    },
    {
      "epoch": 0.45666356011183595,
      "grad_norm": 0.41246395275559605,
      "learning_rate": 4.710044874007594e-05,
      "loss": 0.4263,
      "step": 490
    },
    {
      "epoch": 0.4613233923578751,
      "grad_norm": 0.3975264680297866,
      "learning_rate": 4.7014152571625826e-05,
      "loss": 0.4323,
      "step": 495
    },
    {
      "epoch": 0.4659832246039143,
      "grad_norm": 0.45318918462749547,
      "learning_rate": 4.6927856403175704e-05,
      "loss": 0.4348,
      "step": 500
    },
    {
      "epoch": 0.4706430568499534,
      "grad_norm": 0.36573583820356864,
      "learning_rate": 4.684156023472558e-05,
      "loss": 0.4219,
      "step": 505
    },
    {
      "epoch": 0.47530288909599255,
      "grad_norm": 0.44294897087649004,
      "learning_rate": 4.675526406627546e-05,
      "loss": 0.4301,
      "step": 510
    },
    {
      "epoch": 0.47996272134203166,
      "grad_norm": 0.6261618649064988,
      "learning_rate": 4.666896789782534e-05,
      "loss": 0.4281,
      "step": 515
    },
    {
      "epoch": 0.4846225535880708,
      "grad_norm": 0.5618813957954623,
      "learning_rate": 4.658267172937522e-05,
      "loss": 0.432,
      "step": 520
    },
    {
      "epoch": 0.48928238583411,
      "grad_norm": 0.4221390092567116,
      "learning_rate": 4.64963755609251e-05,
      "loss": 0.4309,
      "step": 525
    },
    {
      "epoch": 0.4939422180801491,
      "grad_norm": 0.5054553118583148,
      "learning_rate": 4.641007939247497e-05,
      "loss": 0.428,
      "step": 530
    },
    {
      "epoch": 0.49860205032618826,
      "grad_norm": 0.3674283025881192,
      "learning_rate": 4.632378322402486e-05,
      "loss": 0.4282,
      "step": 535
    },
    {
      "epoch": 0.5032618825722274,
      "grad_norm": 0.37646280701753526,
      "learning_rate": 4.6237487055574736e-05,
      "loss": 0.4176,
      "step": 540
    },
    {
      "epoch": 0.5079217148182665,
      "grad_norm": 0.4667754579895071,
      "learning_rate": 4.6151190887124615e-05,
      "loss": 0.4243,
      "step": 545
    },
    {
      "epoch": 0.5125815470643057,
      "grad_norm": 0.43689496394898064,
      "learning_rate": 4.606489471867449e-05,
      "loss": 0.4245,
      "step": 550
    },
    {
      "epoch": 0.5172413793103449,
      "grad_norm": 0.35100380853775026,
      "learning_rate": 4.597859855022437e-05,
      "loss": 0.4192,
      "step": 555
    },
    {
      "epoch": 0.5219012115563839,
      "grad_norm": 0.4310432615306107,
      "learning_rate": 4.589230238177425e-05,
      "loss": 0.4395,
      "step": 560
    },
    {
      "epoch": 0.5265610438024231,
      "grad_norm": 0.4828668909643852,
      "learning_rate": 4.5806006213324134e-05,
      "loss": 0.4174,
      "step": 565
    },
    {
      "epoch": 0.5312208760484622,
      "grad_norm": 0.5359345219410544,
      "learning_rate": 4.5719710044874006e-05,
      "loss": 0.4172,
      "step": 570
    },
    {
      "epoch": 0.5358807082945014,
      "grad_norm": 0.44233747611422647,
      "learning_rate": 4.563341387642389e-05,
      "loss": 0.423,
      "step": 575
    },
    {
      "epoch": 0.5405405405405406,
      "grad_norm": 0.35787489087133517,
      "learning_rate": 4.554711770797377e-05,
      "loss": 0.4235,
      "step": 580
    },
    {
      "epoch": 0.5452003727865797,
      "grad_norm": 0.4480426020707225,
      "learning_rate": 4.546082153952365e-05,
      "loss": 0.4206,
      "step": 585
    },
    {
      "epoch": 0.5498602050326188,
      "grad_norm": 0.40460727804169483,
      "learning_rate": 4.5374525371073526e-05,
      "loss": 0.4168,
      "step": 590
    },
    {
      "epoch": 0.554520037278658,
      "grad_norm": 0.37469384681867557,
      "learning_rate": 4.5288229202623404e-05,
      "loss": 0.4188,
      "step": 595
    },
    {
      "epoch": 0.5591798695246971,
      "grad_norm": 0.38124595301547787,
      "learning_rate": 4.520193303417328e-05,
      "loss": 0.4192,
      "step": 600
    },
    {
      "epoch": 0.5638397017707363,
      "grad_norm": 0.3602652491312147,
      "learning_rate": 4.511563686572317e-05,
      "loss": 0.4179,
      "step": 605
    },
    {
      "epoch": 0.5684995340167754,
      "grad_norm": 0.39572056728413657,
      "learning_rate": 4.5029340697273045e-05,
      "loss": 0.4287,
      "step": 610
    },
    {
      "epoch": 0.5731593662628145,
      "grad_norm": 0.33403742184402074,
      "learning_rate": 4.4943044528822923e-05,
      "loss": 0.4155,
      "step": 615
    },
    {
      "epoch": 0.5778191985088537,
      "grad_norm": 0.40306442434828443,
      "learning_rate": 4.48567483603728e-05,
      "loss": 0.4125,
      "step": 620
    },
    {
      "epoch": 0.5824790307548928,
      "grad_norm": 0.3774250530609767,
      "learning_rate": 4.477045219192268e-05,
      "loss": 0.4341,
      "step": 625
    },
    {
      "epoch": 0.587138863000932,
      "grad_norm": 0.39959005575592826,
      "learning_rate": 4.4684156023472565e-05,
      "loss": 0.4245,
      "step": 630
    },
    {
      "epoch": 0.5917986952469712,
      "grad_norm": 0.47668132423182985,
      "learning_rate": 4.4597859855022436e-05,
      "loss": 0.4293,
      "step": 635
    },
    {
      "epoch": 0.5964585274930102,
      "grad_norm": 0.5031461055880807,
      "learning_rate": 4.4511563686572315e-05,
      "loss": 0.42,
      "step": 640
    },
    {
      "epoch": 0.6011183597390494,
      "grad_norm": 0.41239093059640153,
      "learning_rate": 4.44252675181222e-05,
      "loss": 0.4196,
      "step": 645
    },
    {
      "epoch": 0.6057781919850885,
      "grad_norm": 0.40854847004740863,
      "learning_rate": 4.433897134967208e-05,
      "loss": 0.428,
      "step": 650
    },
    {
      "epoch": 0.6104380242311277,
      "grad_norm": 0.41242124461222524,
      "learning_rate": 4.4252675181221956e-05,
      "loss": 0.426,
      "step": 655
    },
    {
      "epoch": 0.6150978564771669,
      "grad_norm": 0.3573616803868084,
      "learning_rate": 4.4166379012771834e-05,
      "loss": 0.4078,
      "step": 660
    },
    {
      "epoch": 0.6197576887232059,
      "grad_norm": 0.4985584033933677,
      "learning_rate": 4.408008284432171e-05,
      "loss": 0.4089,
      "step": 665
    },
    {
      "epoch": 0.6244175209692451,
      "grad_norm": 0.41754039223029704,
      "learning_rate": 4.39937866758716e-05,
      "loss": 0.4075,
      "step": 670
    },
    {
      "epoch": 0.6290773532152842,
      "grad_norm": 0.42762641741838325,
      "learning_rate": 4.3907490507421476e-05,
      "loss": 0.4157,
      "step": 675
    },
    {
      "epoch": 0.6337371854613234,
      "grad_norm": 0.4381716045704423,
      "learning_rate": 4.382119433897135e-05,
      "loss": 0.4188,
      "step": 680
    },
    {
      "epoch": 0.6383970177073626,
      "grad_norm": 0.42615839208636525,
      "learning_rate": 4.373489817052123e-05,
      "loss": 0.4283,
      "step": 685
    },
    {
      "epoch": 0.6430568499534017,
      "grad_norm": 0.41547768813271807,
      "learning_rate": 4.364860200207111e-05,
      "loss": 0.4107,
      "step": 690
    },
    {
      "epoch": 0.6477166821994408,
      "grad_norm": 0.4257514042629888,
      "learning_rate": 4.356230583362099e-05,
      "loss": 0.4152,
      "step": 695
    },
    {
      "epoch": 0.65237651444548,
      "grad_norm": 0.3718564529022634,
      "learning_rate": 4.347600966517087e-05,
      "loss": 0.4245,
      "step": 700
    },
    {
      "epoch": 0.6570363466915191,
      "grad_norm": 0.4097816553154767,
      "learning_rate": 4.3389713496720745e-05,
      "loss": 0.4178,
      "step": 705
    },
    {
      "epoch": 0.6616961789375583,
      "grad_norm": 0.5185129505497016,
      "learning_rate": 4.330341732827063e-05,
      "loss": 0.4184,
      "step": 710
    },
    {
      "epoch": 0.6663560111835974,
      "grad_norm": 0.38356789237987543,
      "learning_rate": 4.321712115982051e-05,
      "loss": 0.4066,
      "step": 715
    },
    {
      "epoch": 0.6710158434296365,
      "grad_norm": 0.39820661867372364,
      "learning_rate": 4.3130824991370387e-05,
      "loss": 0.4169,
      "step": 720
    },
    {
      "epoch": 0.6756756756756757,
      "grad_norm": 0.3468550490700257,
      "learning_rate": 4.3044528822920265e-05,
      "loss": 0.4159,
      "step": 725
    },
    {
      "epoch": 0.6803355079217148,
      "grad_norm": 0.42714803174042604,
      "learning_rate": 4.295823265447014e-05,
      "loss": 0.4107,
      "step": 730
    },
    {
      "epoch": 0.684995340167754,
      "grad_norm": 0.35975622295120474,
      "learning_rate": 4.287193648602002e-05,
      "loss": 0.4148,
      "step": 735
    },
    {
      "epoch": 0.6896551724137931,
      "grad_norm": 0.3398454105039195,
      "learning_rate": 4.27856403175699e-05,
      "loss": 0.4085,
      "step": 740
    },
    {
      "epoch": 0.6943150046598322,
      "grad_norm": 0.402262044573312,
      "learning_rate": 4.269934414911978e-05,
      "loss": 0.4044,
      "step": 745
    },
    {
      "epoch": 0.6989748369058714,
      "grad_norm": 0.42023295565874197,
      "learning_rate": 4.261304798066966e-05,
      "loss": 0.409,
      "step": 750
    },
    {
      "epoch": 0.7036346691519105,
      "grad_norm": 0.3886074714281147,
      "learning_rate": 4.252675181221954e-05,
      "loss": 0.4113,
      "step": 755
    },
    {
      "epoch": 0.7082945013979497,
      "grad_norm": 0.42499664467971804,
      "learning_rate": 4.244045564376942e-05,
      "loss": 0.4146,
      "step": 760
    },
    {
      "epoch": 0.7129543336439889,
      "grad_norm": 0.3605554838945289,
      "learning_rate": 4.23541594753193e-05,
      "loss": 0.399,
      "step": 765
    },
    {
      "epoch": 0.7176141658900279,
      "grad_norm": 0.35331727829864007,
      "learning_rate": 4.2267863306869176e-05,
      "loss": 0.4076,
      "step": 770
    },
    {
      "epoch": 0.7222739981360671,
      "grad_norm": 0.4047141050647308,
      "learning_rate": 4.2181567138419054e-05,
      "loss": 0.4203,
      "step": 775
    },
    {
      "epoch": 0.7269338303821062,
      "grad_norm": 0.370214336816459,
      "learning_rate": 4.209527096996894e-05,
      "loss": 0.4047,
      "step": 780
    },
    {
      "epoch": 0.7315936626281454,
      "grad_norm": 0.39480883644980663,
      "learning_rate": 4.200897480151881e-05,
      "loss": 0.4121,
      "step": 785
    },
    {
      "epoch": 0.7362534948741846,
      "grad_norm": 0.5146939285368787,
      "learning_rate": 4.1922678633068695e-05,
      "loss": 0.4151,
      "step": 790
    },
    {
      "epoch": 0.7409133271202236,
      "grad_norm": 0.43348578379113595,
      "learning_rate": 4.1836382464618573e-05,
      "loss": 0.4096,
      "step": 795
    },
    {
      "epoch": 0.7455731593662628,
      "grad_norm": 0.3922926303913187,
      "learning_rate": 4.175008629616845e-05,
      "loss": 0.4078,
      "step": 800
    },
    {
      "epoch": 0.750232991612302,
      "grad_norm": 0.3640607241449352,
      "learning_rate": 4.166379012771833e-05,
      "loss": 0.4078,
      "step": 805
    },
    {
      "epoch": 0.7548928238583411,
      "grad_norm": 0.349690850963141,
      "learning_rate": 4.157749395926821e-05,
      "loss": 0.4256,
      "step": 810
    },
    {
      "epoch": 0.7595526561043803,
      "grad_norm": 0.36285524732493346,
      "learning_rate": 4.1491197790818086e-05,
      "loss": 0.4004,
      "step": 815
    },
    {
      "epoch": 0.7642124883504194,
      "grad_norm": 0.5210490846235861,
      "learning_rate": 4.140490162236797e-05,
      "loss": 0.4151,
      "step": 820
    },
    {
      "epoch": 0.7688723205964585,
      "grad_norm": 0.41942137907605526,
      "learning_rate": 4.131860545391785e-05,
      "loss": 0.4077,
      "step": 825
    },
    {
      "epoch": 0.7735321528424977,
      "grad_norm": 0.3901545056499728,
      "learning_rate": 4.123230928546773e-05,
      "loss": 0.4024,
      "step": 830
    },
    {
      "epoch": 0.7781919850885368,
      "grad_norm": 0.4038340272785345,
      "learning_rate": 4.1146013117017606e-05,
      "loss": 0.4095,
      "step": 835
    },
    {
      "epoch": 0.782851817334576,
      "grad_norm": 0.4532227365804893,
      "learning_rate": 4.1059716948567484e-05,
      "loss": 0.4127,
      "step": 840
    },
    {
      "epoch": 0.7875116495806151,
      "grad_norm": 0.4770900996270386,
      "learning_rate": 4.097342078011737e-05,
      "loss": 0.4204,
      "step": 845
    },
    {
      "epoch": 0.7921714818266542,
      "grad_norm": 0.37619746602800846,
      "learning_rate": 4.088712461166724e-05,
      "loss": 0.4105,
      "step": 850
    },
    {
      "epoch": 0.7968313140726934,
      "grad_norm": 0.5274698191091436,
      "learning_rate": 4.080082844321712e-05,
      "loss": 0.4121,
      "step": 855
    },
    {
      "epoch": 0.8014911463187325,
      "grad_norm": 0.41457417749149855,
      "learning_rate": 4.0714532274767004e-05,
      "loss": 0.419,
      "step": 860
    },
    {
      "epoch": 0.8061509785647717,
      "grad_norm": 0.3476134180441984,
      "learning_rate": 4.062823610631688e-05,
      "loss": 0.4001,
      "step": 865
    },
    {
      "epoch": 0.8108108108108109,
      "grad_norm": 0.3018646760101771,
      "learning_rate": 4.054193993786676e-05,
      "loss": 0.4144,
      "step": 870
    },
    {
      "epoch": 0.8154706430568499,
      "grad_norm": 0.3505985836303181,
      "learning_rate": 4.045564376941664e-05,
      "loss": 0.4099,
      "step": 875
    },
    {
      "epoch": 0.8201304753028891,
      "grad_norm": 0.37454036950977276,
      "learning_rate": 4.036934760096652e-05,
      "loss": 0.4015,
      "step": 880
    },
    {
      "epoch": 0.8247903075489282,
      "grad_norm": 0.412814499179022,
      "learning_rate": 4.02830514325164e-05,
      "loss": 0.4152,
      "step": 885
    },
    {
      "epoch": 0.8294501397949674,
      "grad_norm": 0.43525980740205655,
      "learning_rate": 4.019675526406628e-05,
      "loss": 0.402,
      "step": 890
    },
    {
      "epoch": 0.8341099720410066,
      "grad_norm": 0.3504008123184863,
      "learning_rate": 4.011045909561615e-05,
      "loss": 0.4211,
      "step": 895
    },
    {
      "epoch": 0.8387698042870456,
      "grad_norm": 0.3451666606741711,
      "learning_rate": 4.0024162927166037e-05,
      "loss": 0.4042,
      "step": 900
    },
    {
      "epoch": 0.8434296365330848,
      "grad_norm": 0.3412474402734993,
      "learning_rate": 3.9937866758715915e-05,
      "loss": 0.4057,
      "step": 905
    },
    {
      "epoch": 0.848089468779124,
      "grad_norm": 0.47725878975243735,
      "learning_rate": 3.98515705902658e-05,
      "loss": 0.4116,
      "step": 910
    },
    {
      "epoch": 0.8527493010251631,
      "grad_norm": 0.6035020610905382,
      "learning_rate": 3.976527442181567e-05,
      "loss": 0.4046,
      "step": 915
    },
    {
      "epoch": 0.8574091332712023,
      "grad_norm": 0.45947885014374223,
      "learning_rate": 3.967897825336555e-05,
      "loss": 0.4109,
      "step": 920
    },
    {
      "epoch": 0.8620689655172413,
      "grad_norm": 0.3346849536418383,
      "learning_rate": 3.9592682084915434e-05,
      "loss": 0.3976,
      "step": 925
    },
    {
      "epoch": 0.8667287977632805,
      "grad_norm": 0.47015004763550144,
      "learning_rate": 3.950638591646531e-05,
      "loss": 0.3977,
      "step": 930
    },
    {
      "epoch": 0.8713886300093197,
      "grad_norm": 0.3751739393328928,
      "learning_rate": 3.942008974801519e-05,
      "loss": 0.4088,
      "step": 935
    },
    {
      "epoch": 0.8760484622553588,
      "grad_norm": 0.31035212858665634,
      "learning_rate": 3.933379357956507e-05,
      "loss": 0.4049,
      "step": 940
    },
    {
      "epoch": 0.880708294501398,
      "grad_norm": 0.37862709572624254,
      "learning_rate": 3.924749741111495e-05,
      "loss": 0.407,
      "step": 945
    },
    {
      "epoch": 0.8853681267474371,
      "grad_norm": 0.4375453312898591,
      "learning_rate": 3.916120124266483e-05,
      "loss": 0.4146,
      "step": 950
    },
    {
      "epoch": 0.8900279589934762,
      "grad_norm": 0.41556861405255147,
      "learning_rate": 3.9074905074214704e-05,
      "loss": 0.4135,
      "step": 955
    },
    {
      "epoch": 0.8946877912395154,
      "grad_norm": 0.34721066251213883,
      "learning_rate": 3.898860890576458e-05,
      "loss": 0.4091,
      "step": 960
    },
    {
      "epoch": 0.8993476234855545,
      "grad_norm": 0.4189318356066771,
      "learning_rate": 3.890231273731447e-05,
      "loss": 0.4007,
      "step": 965
    },
    {
      "epoch": 0.9040074557315937,
      "grad_norm": 0.3129065312499714,
      "learning_rate": 3.8816016568864345e-05,
      "loss": 0.3985,
      "step": 970
    },
    {
      "epoch": 0.9086672879776329,
      "grad_norm": 0.339604217707666,
      "learning_rate": 3.8729720400414224e-05,
      "loss": 0.4169,
      "step": 975
    },
    {
      "epoch": 0.9133271202236719,
      "grad_norm": 0.425269526417587,
      "learning_rate": 3.86434242319641e-05,
      "loss": 0.404,
      "step": 980
    },
    {
      "epoch": 0.9179869524697111,
      "grad_norm": 0.45528213892828173,
      "learning_rate": 3.855712806351398e-05,
      "loss": 0.4058,
      "step": 985
    },
    {
      "epoch": 0.9226467847157502,
      "grad_norm": 0.43623028648882733,
      "learning_rate": 3.8470831895063865e-05,
      "loss": 0.4032,
      "step": 990
    },
    {
      "epoch": 0.9273066169617894,
      "grad_norm": 0.3755484829385926,
      "learning_rate": 3.838453572661374e-05,
      "loss": 0.399,
      "step": 995
    },
    {
      "epoch": 0.9319664492078286,
      "grad_norm": 0.4208532884531691,
      "learning_rate": 3.8298239558163615e-05,
      "loss": 0.402,
      "step": 1000
    },
    {
      "epoch": 0.9366262814538676,
      "grad_norm": 0.4006384178165892,
      "learning_rate": 3.82119433897135e-05,
      "loss": 0.4072,
      "step": 1005
    },
    {
      "epoch": 0.9412861136999068,
      "grad_norm": 0.3959495097559784,
      "learning_rate": 3.812564722126338e-05,
      "loss": 0.3996,
      "step": 1010
    },
    {
      "epoch": 0.9459459459459459,
      "grad_norm": 0.412986020520787,
      "learning_rate": 3.8039351052813256e-05,
      "loss": 0.408,
      "step": 1015
    },
    {
      "epoch": 0.9506057781919851,
      "grad_norm": 0.39141710105069416,
      "learning_rate": 3.7953054884363134e-05,
      "loss": 0.4019,
      "step": 1020
    },
    {
      "epoch": 0.9552656104380243,
      "grad_norm": 0.3681512788462775,
      "learning_rate": 3.786675871591301e-05,
      "loss": 0.4053,
      "step": 1025
    },
    {
      "epoch": 0.9599254426840633,
      "grad_norm": 0.2849981991126552,
      "learning_rate": 3.77804625474629e-05,
      "loss": 0.4103,
      "step": 1030
    },
    {
      "epoch": 0.9645852749301025,
      "grad_norm": 0.4064379280880787,
      "learning_rate": 3.7694166379012776e-05,
      "loss": 0.4027,
      "step": 1035
    },
    {
      "epoch": 0.9692451071761417,
      "grad_norm": 0.38383331397218684,
      "learning_rate": 3.7607870210562654e-05,
      "loss": 0.4002,
      "step": 1040
    },
    {
      "epoch": 0.9739049394221808,
      "grad_norm": 0.27852308304758455,
      "learning_rate": 3.752157404211253e-05,
      "loss": 0.3932,
      "step": 1045
    },
    {
      "epoch": 0.97856477166822,
      "grad_norm": 0.38819077991615814,
      "learning_rate": 3.743527787366241e-05,
      "loss": 0.4221,
      "step": 1050
    },
    {
      "epoch": 0.983224603914259,
      "grad_norm": 0.3660717452223031,
      "learning_rate": 3.734898170521229e-05,
      "loss": 0.4002,
      "step": 1055
    },
    {
      "epoch": 0.9878844361602982,
      "grad_norm": 0.3950869687400394,
      "learning_rate": 3.7262685536762174e-05,
      "loss": 0.3992,
      "step": 1060
    },
    {
      "epoch": 0.9925442684063374,
      "grad_norm": 0.37399341432624805,
      "learning_rate": 3.7176389368312045e-05,
      "loss": 0.4035,
      "step": 1065
    },
    {
      "epoch": 0.9972041006523765,
      "grad_norm": 0.4579382708170151,
      "learning_rate": 3.709009319986193e-05,
      "loss": 0.4037,
      "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": 9.188014991511388e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}