{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 804,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.018656716417910446,
      "grad_norm": 3.6549656925784713,
      "learning_rate": 6.0975609756097564e-06,
      "loss": 0.8899,
      "step": 5
    },
    {
      "epoch": 0.03731343283582089,
      "grad_norm": 5.025090574937554,
      "learning_rate": 1.2195121951219513e-05,
      "loss": 0.8747,
      "step": 10
    },
    {
      "epoch": 0.055970149253731345,
      "grad_norm": 2.342131675091634,
      "learning_rate": 1.8292682926829268e-05,
      "loss": 0.8338,
      "step": 15
    },
    {
      "epoch": 0.07462686567164178,
      "grad_norm": 1.1494234292820684,
      "learning_rate": 2.4390243902439026e-05,
      "loss": 0.7846,
      "step": 20
    },
    {
      "epoch": 0.09328358208955224,
      "grad_norm": 0.8632261864875254,
      "learning_rate": 3.048780487804878e-05,
      "loss": 0.7428,
      "step": 25
    },
    {
      "epoch": 0.11194029850746269,
      "grad_norm": 0.657426681372709,
      "learning_rate": 3.6585365853658535e-05,
      "loss": 0.7252,
      "step": 30
    },
    {
      "epoch": 0.13059701492537312,
      "grad_norm": 0.5368416614080481,
      "learning_rate": 4.26829268292683e-05,
      "loss": 0.7007,
      "step": 35
    },
    {
      "epoch": 0.14925373134328357,
      "grad_norm": 0.46353463964140496,
      "learning_rate": 4.878048780487805e-05,
      "loss": 0.6861,
      "step": 40
    },
    {
      "epoch": 0.16791044776119404,
      "grad_norm": 0.44636564035173865,
      "learning_rate": 4.999694850011677e-05,
      "loss": 0.6822,
      "step": 45
    },
    {
      "epoch": 0.1865671641791045,
      "grad_norm": 0.4138434792991525,
      "learning_rate": 4.998455320039942e-05,
      "loss": 0.6609,
      "step": 50
    },
    {
      "epoch": 0.20522388059701493,
      "grad_norm": 0.422728596552499,
      "learning_rate": 4.9962628631365625e-05,
      "loss": 0.6614,
      "step": 55
    },
    {
      "epoch": 0.22388059701492538,
      "grad_norm": 0.38581935884056645,
      "learning_rate": 4.9931184084955565e-05,
      "loss": 0.6551,
      "step": 60
    },
    {
      "epoch": 0.24253731343283583,
      "grad_norm": 0.38792106424667805,
      "learning_rate": 4.989023288780946e-05,
      "loss": 0.644,
      "step": 65
    },
    {
      "epoch": 0.26119402985074625,
      "grad_norm": 0.35124973951258276,
      "learning_rate": 4.9839792395619594e-05,
      "loss": 0.6397,
      "step": 70
    },
    {
      "epoch": 0.2798507462686567,
      "grad_norm": 0.39928220528883474,
      "learning_rate": 4.977988398577472e-05,
      "loss": 0.6374,
      "step": 75
    },
    {
      "epoch": 0.29850746268656714,
      "grad_norm": 0.390200695997598,
      "learning_rate": 4.971053304830001e-05,
      "loss": 0.6361,
      "step": 80
    },
    {
      "epoch": 0.31716417910447764,
      "grad_norm": 0.38998393581603685,
      "learning_rate": 4.96317689750964e-05,
      "loss": 0.6315,
      "step": 85
    },
    {
      "epoch": 0.3358208955223881,
      "grad_norm": 0.3977052204753468,
      "learning_rate": 4.954362514748392e-05,
      "loss": 0.6343,
      "step": 90
    },
    {
      "epoch": 0.35447761194029853,
      "grad_norm": 0.3799021008048012,
      "learning_rate": 4.9446138922054206e-05,
      "loss": 0.6326,
      "step": 95
    },
    {
      "epoch": 0.373134328358209,
      "grad_norm": 0.42417763729903135,
      "learning_rate": 4.933935161483824e-05,
      "loss": 0.628,
      "step": 100
    },
    {
      "epoch": 0.3917910447761194,
      "grad_norm": 0.4175425415930285,
      "learning_rate": 4.922330848379606e-05,
      "loss": 0.6244,
      "step": 105
    },
    {
      "epoch": 0.41044776119402987,
      "grad_norm": 0.3780496454852693,
      "learning_rate": 4.909805870963577e-05,
      "loss": 0.6316,
      "step": 110
    },
    {
      "epoch": 0.4291044776119403,
      "grad_norm": 0.43250766991726497,
      "learning_rate": 4.89636553749701e-05,
      "loss": 0.6219,
      "step": 115
    },
    {
      "epoch": 0.44776119402985076,
      "grad_norm": 0.45900715839677686,
      "learning_rate": 4.882015544181922e-05,
      "loss": 0.6234,
      "step": 120
    },
    {
      "epoch": 0.4664179104477612,
      "grad_norm": 0.3801562720208039,
      "learning_rate": 4.866761972746946e-05,
      "loss": 0.6203,
      "step": 125
    },
    {
      "epoch": 0.48507462686567165,
      "grad_norm": 0.40763368662979566,
      "learning_rate": 4.850611287869809e-05,
      "loss": 0.6192,
      "step": 130
    },
    {
      "epoch": 0.503731343283582,
      "grad_norm": 0.387695878234886,
      "learning_rate": 4.833570334437505e-05,
      "loss": 0.6158,
      "step": 135
    },
    {
      "epoch": 0.5223880597014925,
      "grad_norm": 0.3594282184260463,
      "learning_rate": 4.8156463346453454e-05,
      "loss": 0.6218,
      "step": 140
    },
    {
      "epoch": 0.5410447761194029,
      "grad_norm": 0.38751010489421944,
      "learning_rate": 4.7968468849360844e-05,
      "loss": 0.6137,
      "step": 145
    },
    {
      "epoch": 0.5597014925373134,
      "grad_norm": 0.44064154866322297,
      "learning_rate": 4.777179952780443e-05,
      "loss": 0.6124,
      "step": 150
    },
    {
      "epoch": 0.5783582089552238,
      "grad_norm": 0.3950056095751924,
      "learning_rate": 4.756653873300381e-05,
      "loss": 0.6123,
      "step": 155
    },
    {
      "epoch": 0.5970149253731343,
      "grad_norm": 0.3802169516389078,
      "learning_rate": 4.735277345736555e-05,
      "loss": 0.6101,
      "step": 160
    },
    {
      "epoch": 0.6156716417910447,
      "grad_norm": 0.5190273440198148,
      "learning_rate": 4.713059429761462e-05,
      "loss": 0.6092,
      "step": 165
    },
    {
      "epoch": 0.6343283582089553,
      "grad_norm": 0.47079908923901587,
      "learning_rate": 4.690009541639818e-05,
      "loss": 0.6094,
      "step": 170
    },
    {
      "epoch": 0.6529850746268657,
      "grad_norm": 0.4315284363771648,
      "learning_rate": 4.666137450237816e-05,
      "loss": 0.6095,
      "step": 175
    },
    {
      "epoch": 0.6716417910447762,
      "grad_norm": 0.40236134418201497,
      "learning_rate": 4.641453272882943e-05,
      "loss": 0.6009,
      "step": 180
    },
    {
      "epoch": 0.6902985074626866,
      "grad_norm": 0.43576469418064606,
      "learning_rate": 4.615967471076114e-05,
      "loss": 0.6043,
      "step": 185
    },
    {
      "epoch": 0.7089552238805971,
      "grad_norm": 0.3924375734617119,
      "learning_rate": 4.5896908460579396e-05,
      "loss": 0.6082,
      "step": 190
    },
    {
      "epoch": 0.7276119402985075,
      "grad_norm": 0.37813678352592345,
      "learning_rate": 4.562634534231012e-05,
      "loss": 0.6111,
      "step": 195
    },
    {
      "epoch": 0.746268656716418,
      "grad_norm": 0.4255214995928645,
      "learning_rate": 4.5348100024401387e-05,
      "loss": 0.6053,
      "step": 200
    },
    {
      "epoch": 0.7649253731343284,
      "grad_norm": 0.4230323004235886,
      "learning_rate": 4.5062290431125306e-05,
      "loss": 0.6021,
      "step": 205
    },
    {
      "epoch": 0.7835820895522388,
      "grad_norm": 0.3891283415062517,
      "learning_rate": 4.476903769260014e-05,
      "loss": 0.6074,
      "step": 210
    },
    {
      "epoch": 0.8022388059701493,
      "grad_norm": 0.359821193476202,
      "learning_rate": 4.4468466093453555e-05,
      "loss": 0.6011,
      "step": 215
    },
    {
      "epoch": 0.8208955223880597,
      "grad_norm": 0.3922321010105192,
      "learning_rate": 4.416070302014912e-05,
      "loss": 0.6012,
      "step": 220
    },
    {
      "epoch": 0.8395522388059702,
      "grad_norm": 0.4054255182956602,
      "learning_rate": 4.384587890699813e-05,
      "loss": 0.6078,
      "step": 225
    },
    {
      "epoch": 0.8582089552238806,
      "grad_norm": 0.46435941050583623,
      "learning_rate": 4.352412718087967e-05,
      "loss": 0.6036,
      "step": 230
    },
    {
      "epoch": 0.8768656716417911,
      "grad_norm": 0.4515063244889334,
      "learning_rate": 4.31955842046925e-05,
      "loss": 0.6037,
      "step": 235
    },
    {
      "epoch": 0.8955223880597015,
      "grad_norm": 0.3782429863515032,
      "learning_rate": 4.2860389219562457e-05,
      "loss": 0.5992,
      "step": 240
    },
    {
      "epoch": 0.914179104477612,
      "grad_norm": 0.3861236280491249,
      "learning_rate": 4.25186842858302e-05,
      "loss": 0.603,
      "step": 245
    },
    {
      "epoch": 0.9328358208955224,
      "grad_norm": 0.36429692137779823,
      "learning_rate": 4.217061422284397e-05,
      "loss": 0.5994,
      "step": 250
    },
    {
      "epoch": 0.9514925373134329,
      "grad_norm": 0.4118616829934472,
      "learning_rate": 4.181632654758317e-05,
      "loss": 0.5949,
      "step": 255
    },
    {
      "epoch": 0.9701492537313433,
      "grad_norm": 0.41043315907009315,
      "learning_rate": 4.145597141213857e-05,
      "loss": 0.5986,
      "step": 260
    },
    {
      "epoch": 0.9888059701492538,
      "grad_norm": 0.39055993718072285,
      "learning_rate": 4.1089701540075746e-05,
      "loss": 0.6022,
      "step": 265
    },
    {
      "epoch": 1.007462686567164,
      "grad_norm": 0.3847644920200864,
      "learning_rate": 4.07176721617087e-05,
      "loss": 0.5929,
      "step": 270
    },
    {
      "epoch": 1.0261194029850746,
      "grad_norm": 0.41157541834436956,
      "learning_rate": 4.034004094831106e-05,
      "loss": 0.595,
      "step": 275
    },
    {
      "epoch": 1.044776119402985,
      "grad_norm": 0.4187910574304168,
      "learning_rate": 3.995696794529279e-05,
      "loss": 0.5958,
      "step": 280
    },
    {
      "epoch": 1.0634328358208955,
      "grad_norm": 0.3752055081771545,
      "learning_rate": 3.9568615504370675e-05,
      "loss": 0.5961,
      "step": 285
    },
    {
      "epoch": 1.0820895522388059,
      "grad_norm": 0.42027998796322924,
      "learning_rate": 3.9175148214761445e-05,
      "loss": 0.5855,
      "step": 290
    },
    {
      "epoch": 1.1007462686567164,
      "grad_norm": 0.43658130712850085,
      "learning_rate": 3.877673283342647e-05,
      "loss": 0.5938,
      "step": 295
    },
    {
      "epoch": 1.1194029850746268,
      "grad_norm": 0.430457565646321,
      "learning_rate": 3.8373538214397895e-05,
      "loss": 0.5889,
      "step": 300
    },
    {
      "epoch": 1.1380597014925373,
      "grad_norm": 0.369429450742487,
      "learning_rate": 3.796573523721588e-05,
      "loss": 0.5917,
      "step": 305
    },
    {
      "epoch": 1.1567164179104479,
      "grad_norm": 0.35118841190764827,
      "learning_rate": 3.755349673450747e-05,
      "loss": 0.5897,
      "step": 310
    },
    {
      "epoch": 1.1753731343283582,
      "grad_norm": 0.43036161901845593,
      "learning_rate": 3.713699741873769e-05,
      "loss": 0.5899,
      "step": 315
    },
    {
      "epoch": 1.1940298507462686,
      "grad_norm": 0.39048195458912954,
      "learning_rate": 3.6716413808163996e-05,
      "loss": 0.5919,
      "step": 320
    },
    {
      "epoch": 1.212686567164179,
      "grad_norm": 0.38357903651062303,
      "learning_rate": 3.6291924152025287e-05,
      "loss": 0.59,
      "step": 325
    },
    {
      "epoch": 1.2313432835820897,
      "grad_norm": 0.4362556920006053,
      "learning_rate": 3.5863708354997426e-05,
      "loss": 0.5872,
      "step": 330
    },
    {
      "epoch": 1.25,
      "grad_norm": 0.38432808337966534,
      "learning_rate": 3.5431947900947086e-05,
      "loss": 0.5923,
      "step": 335
    },
    {
      "epoch": 1.2686567164179103,
      "grad_norm": 0.3682629714132096,
      "learning_rate": 3.499682577601638e-05,
      "loss": 0.5834,
      "step": 340
    },
    {
      "epoch": 1.287313432835821,
      "grad_norm": 0.4315330247371912,
      "learning_rate": 3.455852639107071e-05,
      "loss": 0.5896,
      "step": 345
    },
    {
      "epoch": 1.3059701492537314,
      "grad_norm": 0.3629622380785368,
      "learning_rate": 3.4117235503542874e-05,
      "loss": 0.5871,
      "step": 350
    },
    {
      "epoch": 1.3246268656716418,
      "grad_norm": 0.3922640004199405,
      "learning_rate": 3.3673140138706474e-05,
      "loss": 0.5914,
      "step": 355
    },
    {
      "epoch": 1.3432835820895521,
      "grad_norm": 0.35817834691642214,
      "learning_rate": 3.322642851041199e-05,
      "loss": 0.5853,
      "step": 360
    },
    {
      "epoch": 1.3619402985074627,
      "grad_norm": 0.3816232424070772,
      "learning_rate": 3.277728994131904e-05,
      "loss": 0.5824,
      "step": 365
    },
    {
      "epoch": 1.3805970149253732,
      "grad_norm": 0.4132410785960827,
      "learning_rate": 3.232591478265887e-05,
      "loss": 0.5934,
      "step": 370
    },
    {
      "epoch": 1.3992537313432836,
      "grad_norm": 0.36404595249873795,
      "learning_rate": 3.187249433356076e-05,
      "loss": 0.5874,
      "step": 375
    },
    {
      "epoch": 1.417910447761194,
      "grad_norm": 0.38024271420241285,
      "learning_rate": 3.141722075997681e-05,
      "loss": 0.5867,
      "step": 380
    },
    {
      "epoch": 1.4365671641791045,
      "grad_norm": 0.3334177836252411,
      "learning_rate": 3.096028701323926e-05,
      "loss": 0.5808,
      "step": 385
    },
    {
      "epoch": 1.455223880597015,
      "grad_norm": 0.35555456743280156,
      "learning_rate": 3.050188674828507e-05,
      "loss": 0.5833,
      "step": 390
    },
    {
      "epoch": 1.4738805970149254,
      "grad_norm": 0.3563293906188617,
      "learning_rate": 3.00422142415822e-05,
      "loss": 0.5794,
      "step": 395
    },
    {
      "epoch": 1.4925373134328357,
      "grad_norm": 0.37347508205822805,
      "learning_rate": 2.958146430879254e-05,
      "loss": 0.5838,
      "step": 400
    },
    {
      "epoch": 1.5111940298507462,
      "grad_norm": 0.3636667109098333,
      "learning_rate": 2.9119832222206262e-05,
      "loss": 0.578,
      "step": 405
    },
    {
      "epoch": 1.5298507462686568,
      "grad_norm": 0.41991140965552654,
      "learning_rate": 2.8657513627982702e-05,
      "loss": 0.5824,
      "step": 410
    },
    {
      "epoch": 1.5485074626865671,
      "grad_norm": 0.3847464343971541,
      "learning_rate": 2.8194704463232792e-05,
      "loss": 0.588,
      "step": 415
    },
    {
      "epoch": 1.5671641791044775,
      "grad_norm": 0.35032797152530204,
      "learning_rate": 2.7731600872978102e-05,
      "loss": 0.5842,
      "step": 420
    },
    {
      "epoch": 1.585820895522388,
      "grad_norm": 0.3656500421389817,
      "learning_rate": 2.726839912702191e-05,
      "loss": 0.5867,
      "step": 425
    },
    {
      "epoch": 1.6044776119402986,
      "grad_norm": 0.40068784377780137,
      "learning_rate": 2.6805295536767224e-05,
      "loss": 0.5851,
      "step": 430
    },
    {
      "epoch": 1.623134328358209,
      "grad_norm": 0.321173003139559,
      "learning_rate": 2.6342486372017306e-05,
      "loss": 0.5843,
      "step": 435
    },
    {
      "epoch": 1.6417910447761193,
      "grad_norm": 0.32237157232459224,
      "learning_rate": 2.5880167777793746e-05,
      "loss": 0.5846,
      "step": 440
    },
    {
      "epoch": 1.6604477611940298,
      "grad_norm": 0.3520631416157162,
      "learning_rate": 2.5418535691207464e-05,
      "loss": 0.5869,
      "step": 445
    },
    {
      "epoch": 1.6791044776119404,
      "grad_norm": 0.35075232086508196,
      "learning_rate": 2.49577857584178e-05,
      "loss": 0.5783,
      "step": 450
    },
    {
      "epoch": 1.6977611940298507,
      "grad_norm": 0.3292385498979277,
      "learning_rate": 2.4498113251714936e-05,
      "loss": 0.5875,
      "step": 455
    },
    {
      "epoch": 1.716417910447761,
      "grad_norm": 0.3130865482957303,
      "learning_rate": 2.4039712986760755e-05,
      "loss": 0.5878,
      "step": 460
    },
    {
      "epoch": 1.7350746268656716,
      "grad_norm": 0.3478733318118067,
      "learning_rate": 2.35827792400232e-05,
      "loss": 0.5874,
      "step": 465
    },
    {
      "epoch": 1.7537313432835822,
      "grad_norm": 0.3544639262381643,
      "learning_rate": 2.3127505666439243e-05,
      "loss": 0.5854,
      "step": 470
    },
    {
      "epoch": 1.7723880597014925,
      "grad_norm": 0.3587084792300111,
      "learning_rate": 2.267408521734113e-05,
      "loss": 0.5826,
      "step": 475
    },
    {
      "epoch": 1.7910447761194028,
      "grad_norm": 0.3367799313961551,
      "learning_rate": 2.2222710058680963e-05,
      "loss": 0.5796,
      "step": 480
    },
    {
      "epoch": 1.8097014925373134,
      "grad_norm": 0.31508822995257363,
      "learning_rate": 2.1773571489588017e-05,
      "loss": 0.5814,
      "step": 485
    },
    {
      "epoch": 1.828358208955224,
      "grad_norm": 0.31530891070364053,
      "learning_rate": 2.132685986129353e-05,
      "loss": 0.5822,
      "step": 490
    },
    {
      "epoch": 1.8470149253731343,
      "grad_norm": 0.3219435163937165,
      "learning_rate": 2.088276449645714e-05,
      "loss": 0.5886,
      "step": 495
    },
    {
      "epoch": 1.8656716417910446,
      "grad_norm": 0.33551685144976545,
      "learning_rate": 2.0441473608929303e-05,
      "loss": 0.5899,
      "step": 500
    },
    {
      "epoch": 1.8843283582089554,
      "grad_norm": 0.33919972994543235,
      "learning_rate": 2.0003174223983623e-05,
      "loss": 0.5844,
      "step": 505
    },
    {
      "epoch": 1.9029850746268657,
      "grad_norm": 0.334804929939135,
      "learning_rate": 1.9568052099052912e-05,
      "loss": 0.5793,
      "step": 510
    },
    {
      "epoch": 1.921641791044776,
      "grad_norm": 0.3678088109878758,
      "learning_rate": 1.913629164500258e-05,
      "loss": 0.5876,
      "step": 515
    },
    {
      "epoch": 1.9402985074626866,
      "grad_norm": 0.3556925565126293,
      "learning_rate": 1.8708075847974722e-05,
      "loss": 0.5873,
      "step": 520
    },
    {
      "epoch": 1.9589552238805972,
      "grad_norm": 0.32600987065289666,
      "learning_rate": 1.8283586191836006e-05,
      "loss": 0.5817,
      "step": 525
    },
    {
      "epoch": 1.9776119402985075,
      "grad_norm": 0.3449350706208757,
      "learning_rate": 1.786300258126231e-05,
      "loss": 0.5875,
      "step": 530
    },
    {
      "epoch": 1.9962686567164178,
      "grad_norm": 0.31606697268698664,
      "learning_rate": 1.744650326549254e-05,
      "loss": 0.5771,
      "step": 535
    },
    {
      "epoch": 2.014925373134328,
      "grad_norm": 0.3226406048094545,
      "learning_rate": 1.703426476278413e-05,
      "loss": 0.5804,
      "step": 540
    },
    {
      "epoch": 2.033582089552239,
      "grad_norm": 0.3192747041987855,
      "learning_rate": 1.6626461785602114e-05,
      "loss": 0.5779,
      "step": 545
    },
    {
      "epoch": 2.0522388059701493,
      "grad_norm": 0.32942244104358903,
      "learning_rate": 1.622326716657353e-05,
      "loss": 0.5784,
      "step": 550
    },
    {
      "epoch": 2.0708955223880596,
      "grad_norm": 0.3208641410944032,
      "learning_rate": 1.582485178523856e-05,
      "loss": 0.5752,
      "step": 555
    },
    {
      "epoch": 2.08955223880597,
      "grad_norm": 0.32990050316542857,
      "learning_rate": 1.5431384495629337e-05,
      "loss": 0.5776,
      "step": 560
    },
    {
      "epoch": 2.1082089552238807,
      "grad_norm": 0.32151659170031266,
      "learning_rate": 1.504303205470723e-05,
      "loss": 0.5791,
      "step": 565
    },
    {
      "epoch": 2.126865671641791,
      "grad_norm": 0.3190957196858403,
      "learning_rate": 1.4659959051688944e-05,
      "loss": 0.5823,
      "step": 570
    },
    {
      "epoch": 2.1455223880597014,
      "grad_norm": 0.46789518766409627,
      "learning_rate": 1.4282327838291304e-05,
      "loss": 0.5738,
      "step": 575
    },
    {
      "epoch": 2.1641791044776117,
      "grad_norm": 0.3289296714854712,
      "learning_rate": 1.391029845992426e-05,
      "loss": 0.5784,
      "step": 580
    },
    {
      "epoch": 2.1828358208955225,
      "grad_norm": 0.33039007431665324,
      "learning_rate": 1.3544028587861441e-05,
      "loss": 0.5783,
      "step": 585
    },
    {
      "epoch": 2.201492537313433,
      "grad_norm": 0.3211020903500206,
      "learning_rate": 1.3183673452416833e-05,
      "loss": 0.5741,
      "step": 590
    },
    {
      "epoch": 2.220149253731343,
      "grad_norm": 0.33763161409176584,
      "learning_rate": 1.2829385777156036e-05,
      "loss": 0.5729,
      "step": 595
    },
    {
      "epoch": 2.2388059701492535,
      "grad_norm": 0.33060117221059376,
      "learning_rate": 1.2481315714169812e-05,
      "loss": 0.5781,
      "step": 600
    },
    {
      "epoch": 2.2574626865671643,
      "grad_norm": 0.33725376949413927,
      "learning_rate": 1.2139610780437552e-05,
      "loss": 0.582,
      "step": 605
    },
    {
      "epoch": 2.2761194029850746,
      "grad_norm": 0.2995179260780104,
      "learning_rate": 1.1804415795307511e-05,
      "loss": 0.5769,
      "step": 610
    },
    {
      "epoch": 2.294776119402985,
      "grad_norm": 0.31943181761280764,
      "learning_rate": 1.1475872819120328e-05,
      "loss": 0.5777,
      "step": 615
    },
    {
      "epoch": 2.3134328358208958,
      "grad_norm": 0.29904241071720106,
      "learning_rate": 1.1154121093001874e-05,
      "loss": 0.5751,
      "step": 620
    },
    {
      "epoch": 2.332089552238806,
      "grad_norm": 0.29530206657628216,
      "learning_rate": 1.083929697985089e-05,
      "loss": 0.5641,
      "step": 625
    },
    {
      "epoch": 2.3507462686567164,
      "grad_norm": 0.32449591504588926,
      "learning_rate": 1.0531533906546454e-05,
      "loss": 0.5783,
      "step": 630
    },
    {
      "epoch": 2.3694029850746268,
      "grad_norm": 0.335784339109831,
      "learning_rate": 1.023096230739987e-05,
      "loss": 0.5723,
      "step": 635
    },
    {
      "epoch": 2.388059701492537,
      "grad_norm": 0.31020627295629266,
      "learning_rate": 9.937709568874698e-06,
      "loss": 0.581,
      "step": 640
    },
    {
      "epoch": 2.406716417910448,
      "grad_norm": 0.3012216543355107,
      "learning_rate": 9.651899975598627e-06,
      "loss": 0.5769,
      "step": 645
    },
    {
      "epoch": 2.425373134328358,
      "grad_norm": 0.32450430571066397,
      "learning_rate": 9.373654657689884e-06,
      "loss": 0.5762,
      "step": 650
    },
    {
      "epoch": 2.4440298507462686,
      "grad_norm": 0.3012930563631571,
      "learning_rate": 9.103091539420603e-06,
      "loss": 0.5747,
      "step": 655
    },
    {
      "epoch": 2.4626865671641793,
      "grad_norm": 0.3100034944540072,
      "learning_rate": 8.840325289238862e-06,
      "loss": 0.5751,
      "step": 660
    },
    {
      "epoch": 2.4813432835820897,
      "grad_norm": 0.29337231270061,
      "learning_rate": 8.585467271170572e-06,
      "loss": 0.5778,
      "step": 665
    },
    {
      "epoch": 2.5,
      "grad_norm": 0.29327922763808,
      "learning_rate": 8.338625497621846e-06,
      "loss": 0.5751,
      "step": 670
    },
    {
      "epoch": 2.5186567164179103,
      "grad_norm": 0.3044981666505116,
      "learning_rate": 8.099904583601826e-06,
      "loss": 0.5777,
      "step": 675
    },
    {
      "epoch": 2.5373134328358207,
      "grad_norm": 0.30077068543810176,
      "learning_rate": 7.869405702385388e-06,
      "loss": 0.5733,
      "step": 680
    },
    {
      "epoch": 2.5559701492537314,
      "grad_norm": 0.31187553828615966,
      "learning_rate": 7.647226542634454e-06,
      "loss": 0.5786,
      "step": 685
    },
    {
      "epoch": 2.574626865671642,
      "grad_norm": 0.2928003883848623,
      "learning_rate": 7.433461266996197e-06,
      "loss": 0.5746,
      "step": 690
    },
    {
      "epoch": 2.593283582089552,
      "grad_norm": 0.289384782079962,
      "learning_rate": 7.228200472195573e-06,
      "loss": 0.5749,
      "step": 695
    },
    {
      "epoch": 2.611940298507463,
      "grad_norm": 0.2966613143366587,
      "learning_rate": 7.031531150639156e-06,
      "loss": 0.5769,
      "step": 700
    },
    {
      "epoch": 2.6305970149253732,
      "grad_norm": 0.2955583658511631,
      "learning_rate": 6.843536653546554e-06,
      "loss": 0.5724,
      "step": 705
    },
    {
      "epoch": 2.6492537313432836,
      "grad_norm": 0.2839683212080737,
      "learning_rate": 6.664296655624957e-06,
      "loss": 0.5767,
      "step": 710
    },
    {
      "epoch": 2.667910447761194,
      "grad_norm": 0.3111148068064659,
      "learning_rate": 6.49388712130192e-06,
      "loss": 0.5812,
      "step": 715
    },
    {
      "epoch": 2.6865671641791042,
      "grad_norm": 0.29595259288458947,
      "learning_rate": 6.332380272530536e-06,
      "loss": 0.5787,
      "step": 720
    },
    {
      "epoch": 2.705223880597015,
      "grad_norm": 0.30049652644929575,
      "learning_rate": 6.17984455818078e-06,
      "loss": 0.569,
      "step": 725
    },
    {
      "epoch": 2.7238805970149254,
      "grad_norm": 0.28596569042503084,
      "learning_rate": 6.036344625029903e-06,
      "loss": 0.573,
      "step": 730
    },
    {
      "epoch": 2.7425373134328357,
      "grad_norm": 0.3137116309508735,
      "learning_rate": 5.901941290364234e-06,
      "loss": 0.5736,
      "step": 735
    },
    {
      "epoch": 2.7611940298507465,
      "grad_norm": 0.3009948081748677,
      "learning_rate": 5.776691516203942e-06,
      "loss": 0.5787,
      "step": 740
    },
    {
      "epoch": 2.779850746268657,
      "grad_norm": 0.3040010450151502,
      "learning_rate": 5.660648385161759e-06,
      "loss": 0.5767,
      "step": 745
    },
    {
      "epoch": 2.798507462686567,
      "grad_norm": 0.29325707371427784,
      "learning_rate": 5.5538610779457975e-06,
      "loss": 0.5754,
      "step": 750
    },
    {
      "epoch": 2.8171641791044775,
      "grad_norm": 0.2828494761102343,
      "learning_rate": 5.456374852516083e-06,
      "loss": 0.5767,
      "step": 755
    },
    {
      "epoch": 2.835820895522388,
      "grad_norm": 0.2919715511195579,
      "learning_rate": 5.368231024903606e-06,
      "loss": 0.5731,
      "step": 760
    },
    {
      "epoch": 2.8544776119402986,
      "grad_norm": 0.2984091496111364,
      "learning_rate": 5.289466951699997e-06,
      "loss": 0.5766,
      "step": 765
    },
    {
      "epoch": 2.873134328358209,
      "grad_norm": 0.28179592978223916,
      "learning_rate": 5.2201160142252795e-06,
      "loss": 0.5777,
      "step": 770
    },
    {
      "epoch": 2.8917910447761193,
      "grad_norm": 0.29010245215139613,
      "learning_rate": 5.1602076043804036e-06,
      "loss": 0.5717,
      "step": 775
    },
    {
      "epoch": 2.91044776119403,
      "grad_norm": 0.28595377774074365,
      "learning_rate": 5.1097671121905425e-06,
      "loss": 0.5778,
      "step": 780
    },
    {
      "epoch": 2.9291044776119404,
      "grad_norm": 0.29043591574669186,
      "learning_rate": 5.0688159150444395e-06,
      "loss": 0.5703,
      "step": 785
    },
    {
      "epoch": 2.9477611940298507,
      "grad_norm": 0.28909438769255513,
      "learning_rate": 5.0373713686343774e-06,
      "loss": 0.5809,
      "step": 790
    },
    {
      "epoch": 2.966417910447761,
      "grad_norm": 0.300234068745217,
      "learning_rate": 5.015446799600588e-06,
      "loss": 0.5728,
      "step": 795
    },
    {
      "epoch": 2.9850746268656714,
      "grad_norm": 0.2900267072345629,
      "learning_rate": 5.003051499883236e-06,
      "loss": 0.5827,
      "step": 800
    },
    {
      "epoch": 3.0,
      "step": 804,
      "total_flos": 1.3049081114918912e+16,
      "train_loss": 0.6017516222759266,
      "train_runtime": 25832.0855,
      "train_samples_per_second": 3.984,
      "train_steps_per_second": 0.031
    }
  ],
  "logging_steps": 5,
  "max_steps": 804,
  "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": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.3049081114918912e+16,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}