{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 1656,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.006041839740200891,
      "grad_norm": 8.275440544361343,
      "learning_rate": 5.421686746987952e-07,
      "loss": 0.818,
      "step": 10
    },
    {
      "epoch": 0.012083679480401783,
      "grad_norm": 2.7914609484184627,
      "learning_rate": 1.1445783132530121e-06,
      "loss": 0.6987,
      "step": 20
    },
    {
      "epoch": 0.018125519220602675,
      "grad_norm": 1.3664105415610248,
      "learning_rate": 1.7469879518072292e-06,
      "loss": 0.5802,
      "step": 30
    },
    {
      "epoch": 0.024167358960803565,
      "grad_norm": 0.8594872429754616,
      "learning_rate": 2.349397590361446e-06,
      "loss": 0.5258,
      "step": 40
    },
    {
      "epoch": 0.030209198701004455,
      "grad_norm": 0.7434393729226938,
      "learning_rate": 2.9518072289156627e-06,
      "loss": 0.5145,
      "step": 50
    },
    {
      "epoch": 0.03625103844120535,
      "grad_norm": 0.8872227027342147,
      "learning_rate": 3.5542168674698798e-06,
      "loss": 0.5015,
      "step": 60
    },
    {
      "epoch": 0.042292878181406236,
      "grad_norm": 0.8429713488596823,
      "learning_rate": 4.156626506024097e-06,
      "loss": 0.4901,
      "step": 70
    },
    {
      "epoch": 0.04833471792160713,
      "grad_norm": 0.9065488209295363,
      "learning_rate": 4.759036144578314e-06,
      "loss": 0.4796,
      "step": 80
    },
    {
      "epoch": 0.05437655766180802,
      "grad_norm": 0.9651498125714989,
      "learning_rate": 5.361445783132531e-06,
      "loss": 0.4758,
      "step": 90
    },
    {
      "epoch": 0.06041839740200891,
      "grad_norm": 0.8301785431193227,
      "learning_rate": 5.963855421686747e-06,
      "loss": 0.4829,
      "step": 100
    },
    {
      "epoch": 0.0664602371422098,
      "grad_norm": 0.8886113651753562,
      "learning_rate": 6.566265060240964e-06,
      "loss": 0.4812,
      "step": 110
    },
    {
      "epoch": 0.0725020768824107,
      "grad_norm": 0.9862797318907437,
      "learning_rate": 7.168674698795182e-06,
      "loss": 0.4789,
      "step": 120
    },
    {
      "epoch": 0.07854391662261158,
      "grad_norm": 1.0637568246181046,
      "learning_rate": 7.771084337349398e-06,
      "loss": 0.4767,
      "step": 130
    },
    {
      "epoch": 0.08458575636281247,
      "grad_norm": 1.019587639434787,
      "learning_rate": 8.373493975903614e-06,
      "loss": 0.4748,
      "step": 140
    },
    {
      "epoch": 0.09062759610301337,
      "grad_norm": 1.0494143456183387,
      "learning_rate": 8.975903614457832e-06,
      "loss": 0.471,
      "step": 150
    },
    {
      "epoch": 0.09666943584321426,
      "grad_norm": 1.0167856442156524,
      "learning_rate": 9.57831325301205e-06,
      "loss": 0.482,
      "step": 160
    },
    {
      "epoch": 0.10271127558341515,
      "grad_norm": 0.9927765526681102,
      "learning_rate": 9.99989997506481e-06,
      "loss": 0.4867,
      "step": 170
    },
    {
      "epoch": 0.10875311532361603,
      "grad_norm": 0.8252285160527216,
      "learning_rate": 9.998121865323769e-06,
      "loss": 0.4832,
      "step": 180
    },
    {
      "epoch": 0.11479495506381693,
      "grad_norm": 1.0494152999510933,
      "learning_rate": 9.994121889085426e-06,
      "loss": 0.4836,
      "step": 190
    },
    {
      "epoch": 0.12083679480401782,
      "grad_norm": 0.8683126435696938,
      "learning_rate": 9.987901824500342e-06,
      "loss": 0.4796,
      "step": 200
    },
    {
      "epoch": 0.12687863454421872,
      "grad_norm": 1.0765238837274924,
      "learning_rate": 9.979464436637771e-06,
      "loss": 0.479,
      "step": 210
    },
    {
      "epoch": 0.1329204742844196,
      "grad_norm": 0.9747782968326728,
      "learning_rate": 9.968813476256483e-06,
      "loss": 0.4774,
      "step": 220
    },
    {
      "epoch": 0.1389623140246205,
      "grad_norm": 1.0422947760995596,
      "learning_rate": 9.955953678137397e-06,
      "loss": 0.4775,
      "step": 230
    },
    {
      "epoch": 0.1450041537648214,
      "grad_norm": 0.8638869125481701,
      "learning_rate": 9.940890758978781e-06,
      "loss": 0.4809,
      "step": 240
    },
    {
      "epoch": 0.1510459935050223,
      "grad_norm": 0.9908759029078118,
      "learning_rate": 9.923631414854946e-06,
      "loss": 0.4823,
      "step": 250
    },
    {
      "epoch": 0.15708783324522316,
      "grad_norm": 0.8594590054265251,
      "learning_rate": 9.904183318239573e-06,
      "loss": 0.4732,
      "step": 260
    },
    {
      "epoch": 0.16312967298542405,
      "grad_norm": 0.8537166815636835,
      "learning_rate": 9.882555114594994e-06,
      "loss": 0.4746,
      "step": 270
    },
    {
      "epoch": 0.16917151272562495,
      "grad_norm": 0.7988607086870596,
      "learning_rate": 9.858756418528928e-06,
      "loss": 0.4771,
      "step": 280
    },
    {
      "epoch": 0.17521335246582584,
      "grad_norm": 0.8128733549061342,
      "learning_rate": 9.832797809520404e-06,
      "loss": 0.4703,
      "step": 290
    },
    {
      "epoch": 0.18125519220602673,
      "grad_norm": 0.8670976233152772,
      "learning_rate": 9.804690827216764e-06,
      "loss": 0.4763,
      "step": 300
    },
    {
      "epoch": 0.18729703194622763,
      "grad_norm": 0.9163703807052254,
      "learning_rate": 9.77444796630381e-06,
      "loss": 0.4665,
      "step": 310
    },
    {
      "epoch": 0.19333887168642852,
      "grad_norm": 1.0505077270930072,
      "learning_rate": 9.742082670951423e-06,
      "loss": 0.4781,
      "step": 320
    },
    {
      "epoch": 0.19938071142662941,
      "grad_norm": 0.9771619804123264,
      "learning_rate": 9.707609328837085e-06,
      "loss": 0.4695,
      "step": 330
    },
    {
      "epoch": 0.2054225511668303,
      "grad_norm": 0.8443160970596196,
      "learning_rate": 9.67104326474998e-06,
      "loss": 0.4709,
      "step": 340
    },
    {
      "epoch": 0.2114643909070312,
      "grad_norm": 0.7405442085507775,
      "learning_rate": 9.632400733778504e-06,
      "loss": 0.4687,
      "step": 350
    },
    {
      "epoch": 0.21750623064723207,
      "grad_norm": 1.06910781919214,
      "learning_rate": 9.591698914084224e-06,
      "loss": 0.4727,
      "step": 360
    },
    {
      "epoch": 0.22354807038743296,
      "grad_norm": 0.9783321165505232,
      "learning_rate": 9.548955899265495e-06,
      "loss": 0.4757,
      "step": 370
    },
    {
      "epoch": 0.22958991012763386,
      "grad_norm": 0.856614778477577,
      "learning_rate": 9.504190690314124e-06,
      "loss": 0.4628,
      "step": 380
    },
    {
      "epoch": 0.23563174986783475,
      "grad_norm": 0.8407994854798428,
      "learning_rate": 9.457423187168667e-06,
      "loss": 0.4751,
      "step": 390
    },
    {
      "epoch": 0.24167358960803564,
      "grad_norm": 0.835330085907184,
      "learning_rate": 9.408674179868101e-06,
      "loss": 0.4755,
      "step": 400
    },
    {
      "epoch": 0.24771542934823654,
      "grad_norm": 0.8184694750958316,
      "learning_rate": 9.357965339309823e-06,
      "loss": 0.4688,
      "step": 410
    },
    {
      "epoch": 0.25375726908843743,
      "grad_norm": 0.8556998098855371,
      "learning_rate": 9.305319207616048e-06,
      "loss": 0.468,
      "step": 420
    },
    {
      "epoch": 0.2597991088286383,
      "grad_norm": 0.7125631433834922,
      "learning_rate": 9.250759188112944e-06,
      "loss": 0.4721,
      "step": 430
    },
    {
      "epoch": 0.2658409485688392,
      "grad_norm": 0.8684064304089488,
      "learning_rate": 9.194309534926896e-06,
      "loss": 0.4719,
      "step": 440
    },
    {
      "epoch": 0.2718827883090401,
      "grad_norm": 0.8581281745281804,
      "learning_rate": 9.13599534220258e-06,
      "loss": 0.4676,
      "step": 450
    },
    {
      "epoch": 0.277924628049241,
      "grad_norm": 0.8073909041822805,
      "learning_rate": 9.075842532947605e-06,
      "loss": 0.4729,
      "step": 460
    },
    {
      "epoch": 0.2839664677894419,
      "grad_norm": 0.7760503268897376,
      "learning_rate": 9.013877847508684e-06,
      "loss": 0.4735,
      "step": 470
    },
    {
      "epoch": 0.2900083075296428,
      "grad_norm": 1.007511777834999,
      "learning_rate": 8.95012883168448e-06,
      "loss": 0.4655,
      "step": 480
    },
    {
      "epoch": 0.2960501472698437,
      "grad_norm": 0.8792795612951879,
      "learning_rate": 8.884623824480388e-06,
      "loss": 0.4604,
      "step": 490
    },
    {
      "epoch": 0.3020919870100446,
      "grad_norm": 0.7649844789740483,
      "learning_rate": 8.817391945510697e-06,
      "loss": 0.4727,
      "step": 500
    },
    {
      "epoch": 0.3081338267502455,
      "grad_norm": 0.7112188721793731,
      "learning_rate": 8.748463082053765e-06,
      "loss": 0.4696,
      "step": 510
    },
    {
      "epoch": 0.3141756664904463,
      "grad_norm": 0.982736127204613,
      "learning_rate": 8.677867875765904e-06,
      "loss": 0.4667,
      "step": 520
    },
    {
      "epoch": 0.3202175062306472,
      "grad_norm": 0.7380835460961964,
      "learning_rate": 8.605637709059937e-06,
      "loss": 0.4665,
      "step": 530
    },
    {
      "epoch": 0.3262593459708481,
      "grad_norm": 0.889341527636709,
      "learning_rate": 8.531804691154454e-06,
      "loss": 0.4596,
      "step": 540
    },
    {
      "epoch": 0.332301185711049,
      "grad_norm": 0.8717590020119719,
      "learning_rate": 8.45640164379996e-06,
      "loss": 0.4635,
      "step": 550
    },
    {
      "epoch": 0.3383430254512499,
      "grad_norm": 0.8593981074116124,
      "learning_rate": 8.379462086688302e-06,
      "loss": 0.4614,
      "step": 560
    },
    {
      "epoch": 0.3443848651914508,
      "grad_norm": 0.7452870255566968,
      "learning_rate": 8.301020222551795e-06,
      "loss": 0.4613,
      "step": 570
    },
    {
      "epoch": 0.3504267049316517,
      "grad_norm": 0.6738237626130682,
      "learning_rate": 8.221110921958748e-06,
      "loss": 0.4625,
      "step": 580
    },
    {
      "epoch": 0.35646854467185257,
      "grad_norm": 0.7475917228131528,
      "learning_rate": 8.139769707812083e-06,
      "loss": 0.4643,
      "step": 590
    },
    {
      "epoch": 0.36251038441205347,
      "grad_norm": 0.8625312431506086,
      "learning_rate": 8.057032739557973e-06,
      "loss": 0.4647,
      "step": 600
    },
    {
      "epoch": 0.36855222415225436,
      "grad_norm": 0.7686541350728158,
      "learning_rate": 7.972936797111515e-06,
      "loss": 0.4664,
      "step": 610
    },
    {
      "epoch": 0.37459406389245525,
      "grad_norm": 0.793893791173724,
      "learning_rate": 7.887519264506577e-06,
      "loss": 0.4687,
      "step": 620
    },
    {
      "epoch": 0.38063590363265615,
      "grad_norm": 0.8889160126310801,
      "learning_rate": 7.800818113277085e-06,
      "loss": 0.4675,
      "step": 630
    },
    {
      "epoch": 0.38667774337285704,
      "grad_norm": 0.73855111981923,
      "learning_rate": 7.712871885577147e-06,
      "loss": 0.4696,
      "step": 640
    },
    {
      "epoch": 0.39271958311305794,
      "grad_norm": 0.9038531350448926,
      "learning_rate": 7.623719677047521e-06,
      "loss": 0.4679,
      "step": 650
    },
    {
      "epoch": 0.39876142285325883,
      "grad_norm": 0.8258158060448149,
      "learning_rate": 7.533401119436012e-06,
      "loss": 0.4636,
      "step": 660
    },
    {
      "epoch": 0.4048032625934597,
      "grad_norm": 0.7984996623335708,
      "learning_rate": 7.441956362979584e-06,
      "loss": 0.4573,
      "step": 670
    },
    {
      "epoch": 0.4108451023336606,
      "grad_norm": 0.9068733833902644,
      "learning_rate": 7.349426058555943e-06,
      "loss": 0.467,
      "step": 680
    },
    {
      "epoch": 0.4168869420738615,
      "grad_norm": 0.7305483298541591,
      "learning_rate": 7.255851339612597e-06,
      "loss": 0.4619,
      "step": 690
    },
    {
      "epoch": 0.4229287818140624,
      "grad_norm": 0.7441065423891078,
      "learning_rate": 7.161273803881381e-06,
      "loss": 0.4491,
      "step": 700
    },
    {
      "epoch": 0.4289706215542633,
      "grad_norm": 0.9276808876681103,
      "learning_rate": 7.0657354948865786e-06,
      "loss": 0.4532,
      "step": 710
    },
    {
      "epoch": 0.43501246129446414,
      "grad_norm": 0.6605412717277142,
      "learning_rate": 6.969278883254896e-06,
      "loss": 0.4646,
      "step": 720
    },
    {
      "epoch": 0.44105430103466503,
      "grad_norm": 0.6593545188739028,
      "learning_rate": 6.871946847835548e-06,
      "loss": 0.4542,
      "step": 730
    },
    {
      "epoch": 0.4470961407748659,
      "grad_norm": 0.7295116693541216,
      "learning_rate": 6.773782656638889e-06,
      "loss": 0.4532,
      "step": 740
    },
    {
      "epoch": 0.4531379805150668,
      "grad_norm": 0.7877222556349484,
      "learning_rate": 6.674829947602034e-06,
      "loss": 0.464,
      "step": 750
    },
    {
      "epoch": 0.4591798202552677,
      "grad_norm": 0.7208466663864777,
      "learning_rate": 6.575132709190041e-06,
      "loss": 0.4604,
      "step": 760
    },
    {
      "epoch": 0.4652216599954686,
      "grad_norm": 0.8296261142467121,
      "learning_rate": 6.474735260841264e-06,
      "loss": 0.4604,
      "step": 770
    },
    {
      "epoch": 0.4712634997356695,
      "grad_norm": 0.7971141568003346,
      "learning_rate": 6.373682233265581e-06,
      "loss": 0.464,
      "step": 780
    },
    {
      "epoch": 0.4773053394758704,
      "grad_norm": 0.8736349174540989,
      "learning_rate": 6.2720185486042405e-06,
      "loss": 0.4679,
      "step": 790
    },
    {
      "epoch": 0.4833471792160713,
      "grad_norm": 0.9309796515999401,
      "learning_rate": 6.169789400460167e-06,
      "loss": 0.4613,
      "step": 800
    },
    {
      "epoch": 0.4893890189562722,
      "grad_norm": 0.7752657579479202,
      "learning_rate": 6.067040233807579e-06,
      "loss": 0.4561,
      "step": 810
    },
    {
      "epoch": 0.4954308586964731,
      "grad_norm": 0.8534120955438902,
      "learning_rate": 5.963816724789868e-06,
      "loss": 0.4488,
      "step": 820
    },
    {
      "epoch": 0.501472698436674,
      "grad_norm": 0.8496125433352629,
      "learning_rate": 5.860164760414715e-06,
      "loss": 0.4618,
      "step": 830
    },
    {
      "epoch": 0.5075145381768749,
      "grad_norm": 0.8251415980778435,
      "learning_rate": 5.7561304181554626e-06,
      "loss": 0.4593,
      "step": 840
    },
    {
      "epoch": 0.5135563779170758,
      "grad_norm": 0.7139950510608393,
      "learning_rate": 5.651759945467829e-06,
      "loss": 0.4502,
      "step": 850
    },
    {
      "epoch": 0.5195982176572767,
      "grad_norm": 0.601943553224525,
      "learning_rate": 5.5470997392310475e-06,
      "loss": 0.4619,
      "step": 860
    },
    {
      "epoch": 0.5256400573974775,
      "grad_norm": 0.8153866464407951,
      "learning_rate": 5.442196325122583e-06,
      "loss": 0.4588,
      "step": 870
    },
    {
      "epoch": 0.5316818971376784,
      "grad_norm": 0.8344346217932002,
      "learning_rate": 5.337096336935596e-06,
      "loss": 0.4606,
      "step": 880
    },
    {
      "epoch": 0.5377237368778793,
      "grad_norm": 0.7492780666259381,
      "learning_rate": 5.231846495848343e-06,
      "loss": 0.4408,
      "step": 890
    },
    {
      "epoch": 0.5437655766180802,
      "grad_norm": 0.7261206428098568,
      "learning_rate": 5.12649358965473e-06,
      "loss": 0.4538,
      "step": 900
    },
    {
      "epoch": 0.5498074163582811,
      "grad_norm": 0.7279393955754347,
      "learning_rate": 5.021084451965257e-06,
      "loss": 0.4532,
      "step": 910
    },
    {
      "epoch": 0.555849256098482,
      "grad_norm": 0.825743775909579,
      "learning_rate": 4.915665941387589e-06,
      "loss": 0.4496,
      "step": 920
    },
    {
      "epoch": 0.5618910958386829,
      "grad_norm": 0.7524696024763261,
      "learning_rate": 4.8102849206960144e-06,
      "loss": 0.4521,
      "step": 930
    },
    {
      "epoch": 0.5679329355788838,
      "grad_norm": 0.6103762891076722,
      "learning_rate": 4.704988235999059e-06,
      "loss": 0.4537,
      "step": 940
    },
    {
      "epoch": 0.5739747753190847,
      "grad_norm": 0.7615515295808474,
      "learning_rate": 4.5998226959145e-06,
      "loss": 0.4437,
      "step": 950
    },
    {
      "epoch": 0.5800166150592856,
      "grad_norm": 0.8148357524689337,
      "learning_rate": 4.494835050761055e-06,
      "loss": 0.4506,
      "step": 960
    },
    {
      "epoch": 0.5860584547994865,
      "grad_norm": 0.746845262451762,
      "learning_rate": 4.390071971775977e-06,
      "loss": 0.4496,
      "step": 970
    },
    {
      "epoch": 0.5921002945396874,
      "grad_norm": 0.7650124837023387,
      "learning_rate": 4.285580030367812e-06,
      "loss": 0.4442,
      "step": 980
    },
    {
      "epoch": 0.5981421342798883,
      "grad_norm": 0.7044872850784959,
      "learning_rate": 4.18140567741353e-06,
      "loss": 0.4515,
      "step": 990
    },
    {
      "epoch": 0.6041839740200892,
      "grad_norm": 0.6835203586178121,
      "learning_rate": 4.07759522260924e-06,
      "loss": 0.4484,
      "step": 1000
    },
    {
      "epoch": 0.6102258137602901,
      "grad_norm": 0.7875522158264618,
      "learning_rate": 3.974194813883672e-06,
      "loss": 0.4487,
      "step": 1010
    },
    {
      "epoch": 0.616267653500491,
      "grad_norm": 1.0731871419616035,
      "learning_rate": 3.871250416883547e-06,
      "loss": 0.4487,
      "step": 1020
    },
    {
      "epoch": 0.6223094932406917,
      "grad_norm": 0.7645842030603645,
      "learning_rate": 3.7688077945400135e-06,
      "loss": 0.4476,
      "step": 1030
    },
    {
      "epoch": 0.6283513329808926,
      "grad_norm": 0.8262016423595615,
      "learning_rate": 3.6669124867251705e-06,
      "loss": 0.4419,
      "step": 1040
    },
    {
      "epoch": 0.6343931727210935,
      "grad_norm": 0.8630082766433129,
      "learning_rate": 3.56560979000776e-06,
      "loss": 0.4433,
      "step": 1050
    },
    {
      "epoch": 0.6404350124612944,
      "grad_norm": 0.6739630234814128,
      "learning_rate": 3.4649447375170243e-06,
      "loss": 0.434,
      "step": 1060
    },
    {
      "epoch": 0.6464768522014953,
      "grad_norm": 0.8005468876181695,
      "learning_rate": 3.3649620789236613e-06,
      "loss": 0.4449,
      "step": 1070
    },
    {
      "epoch": 0.6525186919416962,
      "grad_norm": 0.6927017507913021,
      "learning_rate": 3.265706260546793e-06,
      "loss": 0.4487,
      "step": 1080
    },
    {
      "epoch": 0.6585605316818971,
      "grad_norm": 0.7470116822150206,
      "learning_rate": 3.1672214055957885e-06,
      "loss": 0.4492,
      "step": 1090
    },
    {
      "epoch": 0.664602371422098,
      "grad_norm": 0.7165981129271416,
      "learning_rate": 3.0695512945557175e-06,
      "loss": 0.4435,
      "step": 1100
    },
    {
      "epoch": 0.6706442111622989,
      "grad_norm": 0.6713758708690007,
      "learning_rate": 2.972739345725163e-06,
      "loss": 0.445,
      "step": 1110
    },
    {
      "epoch": 0.6766860509024998,
      "grad_norm": 0.7528881198034107,
      "learning_rate": 2.876828595915043e-06,
      "loss": 0.4458,
      "step": 1120
    },
    {
      "epoch": 0.6827278906427007,
      "grad_norm": 1.3727779517766678,
      "learning_rate": 2.781861681317004e-06,
      "loss": 0.4362,
      "step": 1130
    },
    {
      "epoch": 0.6887697303829016,
      "grad_norm": 0.8146352120444986,
      "learning_rate": 2.687880818549927e-06,
      "loss": 0.4396,
      "step": 1140
    },
    {
      "epoch": 0.6948115701231025,
      "grad_norm": 0.6965825401455522,
      "learning_rate": 2.5949277858929297e-06,
      "loss": 0.4362,
      "step": 1150
    },
    {
      "epoch": 0.7008534098633034,
      "grad_norm": 0.6495955670121708,
      "learning_rate": 2.5030439047132484e-06,
      "loss": 0.4453,
      "step": 1160
    },
    {
      "epoch": 0.7068952496035042,
      "grad_norm": 0.6114269361750394,
      "learning_rate": 2.4122700210972218e-06,
      "loss": 0.4436,
      "step": 1170
    },
    {
      "epoch": 0.7129370893437051,
      "grad_norm": 0.851795747172313,
      "learning_rate": 2.322646487692556e-06,
      "loss": 0.4437,
      "step": 1180
    },
    {
      "epoch": 0.718978929083906,
      "grad_norm": 0.775135762171182,
      "learning_rate": 2.2342131457699575e-06,
      "loss": 0.4495,
      "step": 1190
    },
    {
      "epoch": 0.7250207688241069,
      "grad_norm": 0.6479257056864712,
      "learning_rate": 2.1470093075120686e-06,
      "loss": 0.4383,
      "step": 1200
    },
    {
      "epoch": 0.7310626085643078,
      "grad_norm": 0.8263286588325557,
      "learning_rate": 2.061073738537635e-06,
      "loss": 0.4463,
      "step": 1210
    },
    {
      "epoch": 0.7371044483045087,
      "grad_norm": 0.6982438518266665,
      "learning_rate": 1.9764446406686177e-06,
      "loss": 0.4436,
      "step": 1220
    },
    {
      "epoch": 0.7431462880447096,
      "grad_norm": 0.6700592747869402,
      "learning_rate": 1.8931596349479385e-06,
      "loss": 0.4432,
      "step": 1230
    },
    {
      "epoch": 0.7491881277849105,
      "grad_norm": 0.5967953856950272,
      "learning_rate": 1.8112557449154316e-06,
      "loss": 0.4338,
      "step": 1240
    },
    {
      "epoch": 0.7552299675251114,
      "grad_norm": 0.6985853030295077,
      "learning_rate": 1.7307693801493619e-06,
      "loss": 0.4407,
      "step": 1250
    },
    {
      "epoch": 0.7612718072653123,
      "grad_norm": 0.7016362891096635,
      "learning_rate": 1.6517363200809222e-06,
      "loss": 0.4488,
      "step": 1260
    },
    {
      "epoch": 0.7673136470055132,
      "grad_norm": 0.8402985241208492,
      "learning_rate": 1.5741916980888267e-06,
      "loss": 0.435,
      "step": 1270
    },
    {
      "epoch": 0.7733554867457141,
      "grad_norm": 0.8240304279049092,
      "learning_rate": 1.4981699858811038e-06,
      "loss": 0.4436,
      "step": 1280
    },
    {
      "epoch": 0.779397326485915,
      "grad_norm": 0.6660945380738031,
      "learning_rate": 1.423704978171046e-06,
      "loss": 0.4407,
      "step": 1290
    },
    {
      "epoch": 0.7854391662261159,
      "grad_norm": 0.5874782271939601,
      "learning_rate": 1.3508297776540845e-06,
      "loss": 0.4498,
      "step": 1300
    },
    {
      "epoch": 0.7914810059663168,
      "grad_norm": 0.5796653208549384,
      "learning_rate": 1.2795767802923192e-06,
      "loss": 0.4352,
      "step": 1310
    },
    {
      "epoch": 0.7975228457065177,
      "grad_norm": 0.9463374855160357,
      "learning_rate": 1.2099776609132048e-06,
      "loss": 0.4384,
      "step": 1320
    },
    {
      "epoch": 0.8035646854467186,
      "grad_norm": 0.6012369149099516,
      "learning_rate": 1.1420633591288072e-06,
      "loss": 0.441,
      "step": 1330
    },
    {
      "epoch": 0.8096065251869194,
      "grad_norm": 0.7550415373921882,
      "learning_rate": 1.0758640655819107e-06,
      "loss": 0.4375,
      "step": 1340
    },
    {
      "epoch": 0.8156483649271203,
      "grad_norm": 0.7006495875182843,
      "learning_rate": 1.0114092085250566e-06,
      "loss": 0.4439,
      "step": 1350
    },
    {
      "epoch": 0.8216902046673212,
      "grad_norm": 0.6206161322638641,
      "learning_rate": 9.487274407384972e-07,
      "loss": 0.4359,
      "step": 1360
    },
    {
      "epoch": 0.8277320444075221,
      "grad_norm": 0.6290155968597686,
      "learning_rate": 8.878466267928814e-07,
      "loss": 0.4366,
      "step": 1370
    },
    {
      "epoch": 0.833773884147723,
      "grad_norm": 0.6195315234512612,
      "learning_rate": 8.287938306623349e-07,
      "loss": 0.4398,
      "step": 1380
    },
    {
      "epoch": 0.8398157238879239,
      "grad_norm": 0.8102446928428507,
      "learning_rate": 7.715953036934304e-07,
      "loss": 0.4409,
      "step": 1390
    },
    {
      "epoch": 0.8458575636281248,
      "grad_norm": 0.789309132604837,
      "learning_rate": 7.162764729354033e-07,
      "loss": 0.4407,
      "step": 1400
    },
    {
      "epoch": 0.8518994033683257,
      "grad_norm": 0.6558463034047459,
      "learning_rate": 6.628619298368133e-07,
      "loss": 0.4297,
      "step": 1410
    },
    {
      "epoch": 0.8579412431085266,
      "grad_norm": 0.6589202868449741,
      "learning_rate": 6.11375419313644e-07,
      "loss": 0.4397,
      "step": 1420
    },
    {
      "epoch": 0.8639830828487275,
      "grad_norm": 0.6171089586498126,
      "learning_rate": 5.618398291937393e-07,
      "loss": 0.4473,
      "step": 1430
    },
    {
      "epoch": 0.8700249225889283,
      "grad_norm": 0.8171417567736048,
      "learning_rate": 5.142771800422403e-07,
      "loss": 0.4424,
      "step": 1440
    },
    {
      "epoch": 0.8760667623291292,
      "grad_norm": 0.813136783771392,
      "learning_rate": 4.687086153725534e-07,
      "loss": 0.4412,
      "step": 1450
    },
    {
      "epoch": 0.8821086020693301,
      "grad_norm": 0.6976453824184702,
      "learning_rate": 4.2515439224721066e-07,
      "loss": 0.4332,
      "step": 1460
    },
    {
      "epoch": 0.888150441809531,
      "grad_norm": 0.5748251232369562,
      "learning_rate": 3.8363387227278947e-07,
      "loss": 0.4348,
      "step": 1470
    },
    {
      "epoch": 0.8941922815497318,
      "grad_norm": 0.720005515762108,
      "learning_rate": 3.441655129928972e-07,
      "loss": 0.4442,
      "step": 1480
    },
    {
      "epoch": 0.9002341212899327,
      "grad_norm": 0.7252408165512142,
      "learning_rate": 3.067668596830481e-07,
      "loss": 0.4395,
      "step": 1490
    },
    {
      "epoch": 0.9062759610301336,
      "grad_norm": 0.7144631857492548,
      "learning_rate": 2.714545375510852e-07,
      "loss": 0.4366,
      "step": 1500
    },
    {
      "epoch": 0.9123178007703345,
      "grad_norm": 0.6635911944336934,
      "learning_rate": 2.382442443466043e-07,
      "loss": 0.4321,
      "step": 1510
    },
    {
      "epoch": 0.9183596405105354,
      "grad_norm": 0.7758826608638922,
      "learning_rate": 2.0715074338266915e-07,
      "loss": 0.4312,
      "step": 1520
    },
    {
      "epoch": 0.9244014802507363,
      "grad_norm": 0.7230232505488107,
      "learning_rate": 1.7818785697292895e-07,
      "loss": 0.4388,
      "step": 1530
    },
    {
      "epoch": 0.9304433199909372,
      "grad_norm": 0.6580184350979571,
      "learning_rate": 1.5136846028704132e-07,
      "loss": 0.4491,
      "step": 1540
    },
    {
      "epoch": 0.9364851597311381,
      "grad_norm": 0.5586825913101796,
      "learning_rate": 1.2670447562713684e-07,
      "loss": 0.4322,
      "step": 1550
    },
    {
      "epoch": 0.942526999471339,
      "grad_norm": 0.647195128183336,
      "learning_rate": 1.042068671278823e-07,
      "loss": 0.4354,
      "step": 1560
    },
    {
      "epoch": 0.9485688392115399,
      "grad_norm": 0.6035271594962103,
      "learning_rate": 8.388563588247523e-08,
      "loss": 0.4316,
      "step": 1570
    },
    {
      "epoch": 0.9546106789517408,
      "grad_norm": 0.6113112433338288,
      "learning_rate": 6.574981549676007e-08,
      "loss": 0.4354,
      "step": 1580
    },
    {
      "epoch": 0.9606525186919417,
      "grad_norm": 0.7146951417286433,
      "learning_rate": 4.980746807342285e-08,
      "loss": 0.4331,
      "step": 1590
    },
    {
      "epoch": 0.9666943584321426,
      "grad_norm": 0.6521596489801758,
      "learning_rate": 3.6065680628062924e-08,
      "loss": 0.4397,
      "step": 1600
    },
    {
      "epoch": 0.9727361981723435,
      "grad_norm": 0.8208153994914311,
      "learning_rate": 2.4530561938729825e-08,
      "loss": 0.4266,
      "step": 1610
    },
    {
      "epoch": 0.9787780379125444,
      "grad_norm": 0.6500288352676025,
      "learning_rate": 1.520723983032324e-08,
      "loss": 0.4335,
      "step": 1620
    },
    {
      "epoch": 0.9848198776527453,
      "grad_norm": 0.6538530895079608,
      "learning_rate": 8.099858895072587e-09,
      "loss": 0.4373,
      "step": 1630
    },
    {
      "epoch": 0.9908617173929462,
      "grad_norm": 0.8433082091701005,
      "learning_rate": 3.2115786500924728e-09,
      "loss": 0.4371,
      "step": 1640
    },
    {
      "epoch": 0.996903557133147,
      "grad_norm": 0.6909298344072548,
      "learning_rate": 5.445721328567466e-10,
      "loss": 0.4329,
      "step": 1650
    },
    {
      "epoch": 1.0,
      "step": 1656,
      "total_flos": 2042833294327808.0,
      "train_loss": 0.4611289887036678,
      "train_runtime": 68178.0817,
      "train_samples_per_second": 1.554,
      "train_steps_per_second": 0.024
    }
  ],
  "logging_steps": 10,
  "max_steps": 1656,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "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": 2042833294327808.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}