{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.9942418426103647,
  "eval_steps": 500,
  "global_step": 585,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.025591810620601407,
      "grad_norm": 0.3465089797973633,
      "learning_rate": 4.999098819177214e-05,
      "loss": 0.8049,
      "step": 5
    },
    {
      "epoch": 0.05118362124120281,
      "grad_norm": 0.288526713848114,
      "learning_rate": 4.9963959264103544e-05,
      "loss": 0.7441,
      "step": 10
    },
    {
      "epoch": 0.07677543186180422,
      "grad_norm": 0.29372313618659973,
      "learning_rate": 4.9918932703355256e-05,
      "loss": 0.6875,
      "step": 15
    },
    {
      "epoch": 0.10236724248240563,
      "grad_norm": 0.28075122833251953,
      "learning_rate": 4.98559409711857e-05,
      "loss": 0.6468,
      "step": 20
    },
    {
      "epoch": 0.12795905310300704,
      "grad_norm": 0.2833946645259857,
      "learning_rate": 4.977502948114772e-05,
      "loss": 0.6117,
      "step": 25
    },
    {
      "epoch": 0.15355086372360843,
      "grad_norm": 0.2722305953502655,
      "learning_rate": 4.967625656594782e-05,
      "loss": 0.5909,
      "step": 30
    },
    {
      "epoch": 0.17914267434420986,
      "grad_norm": 0.25874456763267517,
      "learning_rate": 4.955969343539162e-05,
      "loss": 0.5685,
      "step": 35
    },
    {
      "epoch": 0.20473448496481125,
      "grad_norm": 0.2593054473400116,
      "learning_rate": 4.942542412504543e-05,
      "loss": 0.5595,
      "step": 40
    },
    {
      "epoch": 0.23032629558541268,
      "grad_norm": 0.2561085522174835,
      "learning_rate": 4.92735454356513e-05,
      "loss": 0.5419,
      "step": 45
    },
    {
      "epoch": 0.2559181062060141,
      "grad_norm": 0.2852307856082916,
      "learning_rate": 4.910416686333906e-05,
      "loss": 0.5384,
      "step": 50
    },
    {
      "epoch": 0.28150991682661547,
      "grad_norm": 0.29978978633880615,
      "learning_rate": 4.8917410520685635e-05,
      "loss": 0.5243,
      "step": 55
    },
    {
      "epoch": 0.30710172744721687,
      "grad_norm": 0.30773016810417175,
      "learning_rate": 4.8713411048678635e-05,
      "loss": 0.5229,
      "step": 60
    },
    {
      "epoch": 0.3326935380678183,
      "grad_norm": 0.3103509247303009,
      "learning_rate": 4.849231551964771e-05,
      "loss": 0.5124,
      "step": 65
    },
    {
      "epoch": 0.3582853486884197,
      "grad_norm": 0.32078924775123596,
      "learning_rate": 4.8254283331233464e-05,
      "loss": 0.5118,
      "step": 70
    },
    {
      "epoch": 0.3838771593090211,
      "grad_norm": 0.3069196939468384,
      "learning_rate": 4.799948609147061e-05,
      "loss": 0.501,
      "step": 75
    },
    {
      "epoch": 0.4094689699296225,
      "grad_norm": 0.30792945623397827,
      "learning_rate": 4.77281074950681e-05,
      "loss": 0.5015,
      "step": 80
    },
    {
      "epoch": 0.4350607805502239,
      "grad_norm": 0.3218609094619751,
      "learning_rate": 4.744034319097535e-05,
      "loss": 0.5099,
      "step": 85
    },
    {
      "epoch": 0.46065259117082535,
      "grad_norm": 0.35433730483055115,
      "learning_rate": 4.713640064133025e-05,
      "loss": 0.4912,
      "step": 90
    },
    {
      "epoch": 0.48624440179142675,
      "grad_norm": 0.3382019102573395,
      "learning_rate": 4.681649897189036e-05,
      "loss": 0.4951,
      "step": 95
    },
    {
      "epoch": 0.5118362124120281,
      "grad_norm": 0.37824833393096924,
      "learning_rate": 4.6480868814055424e-05,
      "loss": 0.4889,
      "step": 100
    },
    {
      "epoch": 0.5374280230326296,
      "grad_norm": 0.3094428479671478,
      "learning_rate": 4.6129752138594874e-05,
      "loss": 0.4878,
      "step": 105
    },
    {
      "epoch": 0.5630198336532309,
      "grad_norm": 0.39037370681762695,
      "learning_rate": 4.5763402081200294e-05,
      "loss": 0.4913,
      "step": 110
    },
    {
      "epoch": 0.5886116442738324,
      "grad_norm": 0.34269848465919495,
      "learning_rate": 4.538208275998861e-05,
      "loss": 0.4926,
      "step": 115
    },
    {
      "epoch": 0.6142034548944337,
      "grad_norm": 0.3708631098270416,
      "learning_rate": 4.498606908508754e-05,
      "loss": 0.4931,
      "step": 120
    },
    {
      "epoch": 0.6397952655150352,
      "grad_norm": 0.3521125614643097,
      "learning_rate": 4.457564656044056e-05,
      "loss": 0.481,
      "step": 125
    },
    {
      "epoch": 0.6653870761356366,
      "grad_norm": 0.3537481427192688,
      "learning_rate": 4.415111107797445e-05,
      "loss": 0.4842,
      "step": 130
    },
    {
      "epoch": 0.690978886756238,
      "grad_norm": 0.38676100969314575,
      "learning_rate": 4.371276870427753e-05,
      "loss": 0.4765,
      "step": 135
    },
    {
      "epoch": 0.7165706973768394,
      "grad_norm": 0.4416983723640442,
      "learning_rate": 4.3260935459942584e-05,
      "loss": 0.4652,
      "step": 140
    },
    {
      "epoch": 0.7421625079974408,
      "grad_norm": 0.36901989579200745,
      "learning_rate": 4.2795937091733515e-05,
      "loss": 0.4819,
      "step": 145
    },
    {
      "epoch": 0.7677543186180422,
      "grad_norm": 0.3908597528934479,
      "learning_rate": 4.231810883773999e-05,
      "loss": 0.4612,
      "step": 150
    },
    {
      "epoch": 0.7933461292386437,
      "grad_norm": 0.4273853003978729,
      "learning_rate": 4.182779518568926e-05,
      "loss": 0.4713,
      "step": 155
    },
    {
      "epoch": 0.818937939859245,
      "grad_norm": 0.36905694007873535,
      "learning_rate": 4.132534962458962e-05,
      "loss": 0.4734,
      "step": 160
    },
    {
      "epoch": 0.8445297504798465,
      "grad_norm": 0.3763083517551422,
      "learning_rate": 4.0811134389884433e-05,
      "loss": 0.4659,
      "step": 165
    },
    {
      "epoch": 0.8701215611004478,
      "grad_norm": 0.3596407175064087,
      "learning_rate": 4.028552020230031e-05,
      "loss": 0.4631,
      "step": 170
    },
    {
      "epoch": 0.8957133717210493,
      "grad_norm": 0.396517276763916,
      "learning_rate": 3.974888600057808e-05,
      "loss": 0.4634,
      "step": 175
    },
    {
      "epoch": 0.9213051823416507,
      "grad_norm": 0.38290080428123474,
      "learning_rate": 3.920161866827889e-05,
      "loss": 0.4634,
      "step": 180
    },
    {
      "epoch": 0.946896992962252,
      "grad_norm": 0.39791831374168396,
      "learning_rate": 3.8644112754862614e-05,
      "loss": 0.4638,
      "step": 185
    },
    {
      "epoch": 0.9724888035828535,
      "grad_norm": 0.3460437059402466,
      "learning_rate": 3.807677019123944e-05,
      "loss": 0.4647,
      "step": 190
    },
    {
      "epoch": 0.9980806142034548,
      "grad_norm": 0.39563411474227905,
      "learning_rate": 3.7500000000000003e-05,
      "loss": 0.465,
      "step": 195
    },
    {
      "epoch": 1.0236724248240563,
      "grad_norm": 0.3726441264152527,
      "learning_rate": 3.69142180005327e-05,
      "loss": 0.5229,
      "step": 200
    },
    {
      "epoch": 1.0492642354446577,
      "grad_norm": 0.402972936630249,
      "learning_rate": 3.631984650924094e-05,
      "loss": 0.4445,
      "step": 205
    },
    {
      "epoch": 1.0748560460652592,
      "grad_norm": 0.4153399169445038,
      "learning_rate": 3.5717314035076355e-05,
      "loss": 0.4516,
      "step": 210
    },
    {
      "epoch": 1.1004478566858604,
      "grad_norm": 0.5092347860336304,
      "learning_rate": 3.510705497060762e-05,
      "loss": 0.4528,
      "step": 215
    },
    {
      "epoch": 1.1260396673064619,
      "grad_norm": 0.40579304099082947,
      "learning_rate": 3.4489509278847414e-05,
      "loss": 0.4395,
      "step": 220
    },
    {
      "epoch": 1.1516314779270633,
      "grad_norm": 0.43368446826934814,
      "learning_rate": 3.386512217606339e-05,
      "loss": 0.4484,
      "step": 225
    },
    {
      "epoch": 1.1772232885476648,
      "grad_norm": 0.41571882367134094,
      "learning_rate": 3.323434381080199e-05,
      "loss": 0.4376,
      "step": 230
    },
    {
      "epoch": 1.2028150991682662,
      "grad_norm": 0.44565585255622864,
      "learning_rate": 3.2597628939356175e-05,
      "loss": 0.4578,
      "step": 235
    },
    {
      "epoch": 1.2284069097888675,
      "grad_norm": 0.4231709837913513,
      "learning_rate": 3.195543659791132e-05,
      "loss": 0.439,
      "step": 240
    },
    {
      "epoch": 1.253998720409469,
      "grad_norm": 0.4322074055671692,
      "learning_rate": 3.130822977160554e-05,
      "loss": 0.4429,
      "step": 245
    },
    {
      "epoch": 1.2795905310300704,
      "grad_norm": 0.42521893978118896,
      "learning_rate": 3.065647506074306e-05,
      "loss": 0.4427,
      "step": 250
    },
    {
      "epoch": 1.3051823416506718,
      "grad_norm": 0.416797012090683,
      "learning_rate": 3.0000642344401113e-05,
      "loss": 0.4315,
      "step": 255
    },
    {
      "epoch": 1.3307741522712733,
      "grad_norm": 0.41196638345718384,
      "learning_rate": 2.9341204441673266e-05,
      "loss": 0.4425,
      "step": 260
    },
    {
      "epoch": 1.3563659628918745,
      "grad_norm": 0.4643464982509613,
      "learning_rate": 2.8678636770792906e-05,
      "loss": 0.4488,
      "step": 265
    },
    {
      "epoch": 1.381957773512476,
      "grad_norm": 0.40053775906562805,
      "learning_rate": 2.8013417006383076e-05,
      "loss": 0.4289,
      "step": 270
    },
    {
      "epoch": 1.4075495841330774,
      "grad_norm": 0.4310891032218933,
      "learning_rate": 2.7346024735079486e-05,
      "loss": 0.4348,
      "step": 275
    },
    {
      "epoch": 1.4331413947536789,
      "grad_norm": 0.5526339411735535,
      "learning_rate": 2.667694110977506e-05,
      "loss": 0.4294,
      "step": 280
    },
    {
      "epoch": 1.4587332053742803,
      "grad_norm": 0.473822683095932,
      "learning_rate": 2.600664850273538e-05,
      "loss": 0.4361,
      "step": 285
    },
    {
      "epoch": 1.4843250159948815,
      "grad_norm": 0.4439009726047516,
      "learning_rate": 2.5335630157834937e-05,
      "loss": 0.4413,
      "step": 290
    },
    {
      "epoch": 1.5099168266154832,
      "grad_norm": 0.4225783944129944,
      "learning_rate": 2.4664369842165068e-05,
      "loss": 0.4296,
      "step": 295
    },
    {
      "epoch": 1.5355086372360844,
      "grad_norm": 0.497530460357666,
      "learning_rate": 2.399335149726463e-05,
      "loss": 0.4447,
      "step": 300
    },
    {
      "epoch": 1.561100447856686,
      "grad_norm": 0.4586053490638733,
      "learning_rate": 2.3323058890224938e-05,
      "loss": 0.4207,
      "step": 305
    },
    {
      "epoch": 1.5866922584772873,
      "grad_norm": 0.5011757612228394,
      "learning_rate": 2.265397526492052e-05,
      "loss": 0.4398,
      "step": 310
    },
    {
      "epoch": 1.6122840690978886,
      "grad_norm": 0.4521833658218384,
      "learning_rate": 2.1986582993616926e-05,
      "loss": 0.4367,
      "step": 315
    },
    {
      "epoch": 1.6378758797184902,
      "grad_norm": 0.4077029526233673,
      "learning_rate": 2.1321363229207096e-05,
      "loss": 0.4274,
      "step": 320
    },
    {
      "epoch": 1.6634676903390915,
      "grad_norm": 0.4575510621070862,
      "learning_rate": 2.0658795558326743e-05,
      "loss": 0.4412,
      "step": 325
    },
    {
      "epoch": 1.689059500959693,
      "grad_norm": 0.48359596729278564,
      "learning_rate": 1.9999357655598893e-05,
      "loss": 0.4408,
      "step": 330
    },
    {
      "epoch": 1.7146513115802944,
      "grad_norm": 0.4537913203239441,
      "learning_rate": 1.934352493925695e-05,
      "loss": 0.4306,
      "step": 335
    },
    {
      "epoch": 1.7402431222008956,
      "grad_norm": 0.4479089379310608,
      "learning_rate": 1.8691770228394456e-05,
      "loss": 0.4347,
      "step": 340
    },
    {
      "epoch": 1.7658349328214973,
      "grad_norm": 0.432325154542923,
      "learning_rate": 1.8044563402088684e-05,
      "loss": 0.4252,
      "step": 345
    },
    {
      "epoch": 1.7914267434420985,
      "grad_norm": 0.4653799831867218,
      "learning_rate": 1.740237106064383e-05,
      "loss": 0.4306,
      "step": 350
    },
    {
      "epoch": 1.8170185540627,
      "grad_norm": 0.46541592478752136,
      "learning_rate": 1.6765656189198013e-05,
      "loss": 0.4201,
      "step": 355
    },
    {
      "epoch": 1.8426103646833014,
      "grad_norm": 0.5030815005302429,
      "learning_rate": 1.613487782393661e-05,
      "loss": 0.426,
      "step": 360
    },
    {
      "epoch": 1.8682021753039026,
      "grad_norm": 0.5130462646484375,
      "learning_rate": 1.5510490721152592e-05,
      "loss": 0.4276,
      "step": 365
    },
    {
      "epoch": 1.8937939859245043,
      "grad_norm": 0.48607420921325684,
      "learning_rate": 1.489294502939238e-05,
      "loss": 0.418,
      "step": 370
    },
    {
      "epoch": 1.9193857965451055,
      "grad_norm": 0.43401479721069336,
      "learning_rate": 1.4282685964923642e-05,
      "loss": 0.4326,
      "step": 375
    },
    {
      "epoch": 1.944977607165707,
      "grad_norm": 0.4771016538143158,
      "learning_rate": 1.3680153490759073e-05,
      "loss": 0.433,
      "step": 380
    },
    {
      "epoch": 1.9705694177863085,
      "grad_norm": 0.503084123134613,
      "learning_rate": 1.3085781999467303e-05,
      "loss": 0.4278,
      "step": 385
    },
    {
      "epoch": 1.9961612284069097,
      "grad_norm": 0.504097580909729,
      "learning_rate": 1.2500000000000006e-05,
      "loss": 0.4221,
      "step": 390
    },
    {
      "epoch": 2.0217530390275114,
      "grad_norm": 0.46889811754226685,
      "learning_rate": 1.1923229808760564e-05,
      "loss": 0.4834,
      "step": 395
    },
    {
      "epoch": 2.0473448496481126,
      "grad_norm": 0.45117413997650146,
      "learning_rate": 1.1355887245137383e-05,
      "loss": 0.4055,
      "step": 400
    },
    {
      "epoch": 2.072936660268714,
      "grad_norm": 0.47571200132369995,
      "learning_rate": 1.0798381331721109e-05,
      "loss": 0.4143,
      "step": 405
    },
    {
      "epoch": 2.0985284708893155,
      "grad_norm": 0.5209171175956726,
      "learning_rate": 1.0251113999421935e-05,
      "loss": 0.4189,
      "step": 410
    },
    {
      "epoch": 2.1241202815099167,
      "grad_norm": 0.49102234840393066,
      "learning_rate": 9.714479797699694e-06,
      "loss": 0.4051,
      "step": 415
    },
    {
      "epoch": 2.1497120921305184,
      "grad_norm": 0.5026312470436096,
      "learning_rate": 9.18886561011557e-06,
      "loss": 0.4181,
      "step": 420
    },
    {
      "epoch": 2.1753039027511196,
      "grad_norm": 0.4433274269104004,
      "learning_rate": 8.67465037541038e-06,
      "loss": 0.4141,
      "step": 425
    },
    {
      "epoch": 2.200895713371721,
      "grad_norm": 0.4839063286781311,
      "learning_rate": 8.172204814310742e-06,
      "loss": 0.4148,
      "step": 430
    },
    {
      "epoch": 2.2264875239923225,
      "grad_norm": 0.46064066886901855,
      "learning_rate": 7.681891162260015e-06,
      "loss": 0.4093,
      "step": 435
    },
    {
      "epoch": 2.2520793346129238,
      "grad_norm": 0.47170594334602356,
      "learning_rate": 7.20406290826649e-06,
      "loss": 0.4152,
      "step": 440
    },
    {
      "epoch": 2.2776711452335254,
      "grad_norm": 0.4596922695636749,
      "learning_rate": 6.739064540057424e-06,
      "loss": 0.4219,
      "step": 445
    },
    {
      "epoch": 2.3032629558541267,
      "grad_norm": 0.45822423696517944,
      "learning_rate": 6.28723129572247e-06,
      "loss": 0.4169,
      "step": 450
    },
    {
      "epoch": 2.328854766474728,
      "grad_norm": 0.46449440717697144,
      "learning_rate": 5.848888922025553e-06,
      "loss": 0.4033,
      "step": 455
    },
    {
      "epoch": 2.3544465770953296,
      "grad_norm": 0.4573959708213806,
      "learning_rate": 5.424353439559446e-06,
      "loss": 0.4139,
      "step": 460
    },
    {
      "epoch": 2.380038387715931,
      "grad_norm": 0.48029977083206177,
      "learning_rate": 5.013930914912476e-06,
      "loss": 0.4184,
      "step": 465
    },
    {
      "epoch": 2.4056301983365325,
      "grad_norm": 0.5101778507232666,
      "learning_rate": 4.617917240011394e-06,
      "loss": 0.4142,
      "step": 470
    },
    {
      "epoch": 2.4312220089571337,
      "grad_norm": 0.5021511912345886,
      "learning_rate": 4.236597918799709e-06,
      "loss": 0.4105,
      "step": 475
    },
    {
      "epoch": 2.456813819577735,
      "grad_norm": 0.46041542291641235,
      "learning_rate": 3.8702478614051355e-06,
      "loss": 0.4033,
      "step": 480
    },
    {
      "epoch": 2.4824056301983366,
      "grad_norm": 0.4977727234363556,
      "learning_rate": 3.5191311859445796e-06,
      "loss": 0.4136,
      "step": 485
    },
    {
      "epoch": 2.507997440818938,
      "grad_norm": 0.4449191093444824,
      "learning_rate": 3.183501028109642e-06,
      "loss": 0.4026,
      "step": 490
    },
    {
      "epoch": 2.5335892514395395,
      "grad_norm": 0.4880577027797699,
      "learning_rate": 2.8635993586697553e-06,
      "loss": 0.4144,
      "step": 495
    },
    {
      "epoch": 2.5591810620601407,
      "grad_norm": 0.4731748402118683,
      "learning_rate": 2.5596568090246548e-06,
      "loss": 0.4156,
      "step": 500
    },
    {
      "epoch": 2.584772872680742,
      "grad_norm": 0.4792173206806183,
      "learning_rate": 2.271892504931905e-06,
      "loss": 0.4129,
      "step": 505
    },
    {
      "epoch": 2.6103646833013436,
      "grad_norm": 0.4491451680660248,
      "learning_rate": 2.0005139085293945e-06,
      "loss": 0.3978,
      "step": 510
    },
    {
      "epoch": 2.635956493921945,
      "grad_norm": 0.5214707851409912,
      "learning_rate": 1.7457166687665449e-06,
      "loss": 0.4162,
      "step": 515
    },
    {
      "epoch": 2.6615483045425465,
      "grad_norm": 0.44630053639411926,
      "learning_rate": 1.5076844803522922e-06,
      "loss": 0.4048,
      "step": 520
    },
    {
      "epoch": 2.6871401151631478,
      "grad_norm": 0.45181167125701904,
      "learning_rate": 1.286588951321363e-06,
      "loss": 0.4159,
      "step": 525
    },
    {
      "epoch": 2.712731925783749,
      "grad_norm": 0.4882276952266693,
      "learning_rate": 1.0825894793143721e-06,
      "loss": 0.4171,
      "step": 530
    },
    {
      "epoch": 2.7383237364043507,
      "grad_norm": 0.470037043094635,
      "learning_rate": 8.958331366609423e-07,
      "loss": 0.4049,
      "step": 535
    },
    {
      "epoch": 2.763915547024952,
      "grad_norm": 0.4617624580860138,
      "learning_rate": 7.264545643486997e-07,
      "loss": 0.4062,
      "step": 540
    },
    {
      "epoch": 2.7895073576455536,
      "grad_norm": 0.47536271810531616,
      "learning_rate": 5.745758749545749e-07,
      "loss": 0.4153,
      "step": 545
    },
    {
      "epoch": 2.815099168266155,
      "grad_norm": 0.4715920388698578,
      "learning_rate": 4.403065646083809e-07,
      "loss": 0.4208,
      "step": 550
    },
    {
      "epoch": 2.840690978886756,
      "grad_norm": 0.4840976297855377,
      "learning_rate": 3.237434340521789e-07,
      "loss": 0.4083,
      "step": 555
    },
    {
      "epoch": 2.8662827895073577,
      "grad_norm": 0.4537844657897949,
      "learning_rate": 2.2497051885228827e-07,
      "loss": 0.4085,
      "step": 560
    },
    {
      "epoch": 2.891874600127959,
      "grad_norm": 0.4958924651145935,
      "learning_rate": 1.4405902881430288e-07,
      "loss": 0.417,
      "step": 565
    },
    {
      "epoch": 2.9174664107485606,
      "grad_norm": 0.48199719190597534,
      "learning_rate": 8.106729664475176e-08,
      "loss": 0.415,
      "step": 570
    },
    {
      "epoch": 2.943058221369162,
      "grad_norm": 0.45770469307899475,
      "learning_rate": 3.604073589645596e-08,
      "loss": 0.4043,
      "step": 575
    },
    {
      "epoch": 2.968650031989763,
      "grad_norm": 0.5054221153259277,
      "learning_rate": 9.011808227865625e-09,
      "loss": 0.4034,
      "step": 580
    },
    {
      "epoch": 2.9942418426103647,
      "grad_norm": 0.4703996777534485,
      "learning_rate": 0.0,
      "loss": 0.4106,
      "step": 585
    }
  ],
  "logging_steps": 5,
  "max_steps": 585,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.5149343615798477e+18,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}