{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 2286,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0004375410194705754,
      "grad_norm": 240.0139923095703,
      "learning_rate": 3.6231884057971015e-07,
      "loss": 11.8567,
      "step": 1
    },
    {
      "epoch": 0.004375410194705754,
      "grad_norm": 92.28858184814453,
      "learning_rate": 3.6231884057971017e-06,
      "loss": 10.1921,
      "step": 10
    },
    {
      "epoch": 0.008750820389411508,
      "grad_norm": 37.24079895019531,
      "learning_rate": 7.246376811594203e-06,
      "loss": 8.7443,
      "step": 20
    },
    {
      "epoch": 0.01312623058411726,
      "grad_norm": 57.094844818115234,
      "learning_rate": 1.0869565217391305e-05,
      "loss": 7.937,
      "step": 30
    },
    {
      "epoch": 0.017501640778823015,
      "grad_norm": 37.80804443359375,
      "learning_rate": 1.4492753623188407e-05,
      "loss": 7.5627,
      "step": 40
    },
    {
      "epoch": 0.021877050973528767,
      "grad_norm": 37.13063049316406,
      "learning_rate": 1.8115942028985507e-05,
      "loss": 7.1986,
      "step": 50
    },
    {
      "epoch": 0.02625246116823452,
      "grad_norm": 85.72801208496094,
      "learning_rate": 2.173913043478261e-05,
      "loss": 6.6604,
      "step": 60
    },
    {
      "epoch": 0.030627871362940276,
      "grad_norm": 61.312721252441406,
      "learning_rate": 2.5362318840579714e-05,
      "loss": 6.1824,
      "step": 70
    },
    {
      "epoch": 0.03500328155764603,
      "grad_norm": 41.373172760009766,
      "learning_rate": 2.8985507246376814e-05,
      "loss": 5.7529,
      "step": 80
    },
    {
      "epoch": 0.039378691752351785,
      "grad_norm": 40.97380828857422,
      "learning_rate": 3.260869565217392e-05,
      "loss": 5.0558,
      "step": 90
    },
    {
      "epoch": 0.04375410194705753,
      "grad_norm": 29.73272132873535,
      "learning_rate": 3.6231884057971014e-05,
      "loss": 4.2438,
      "step": 100
    },
    {
      "epoch": 0.04812951214176329,
      "grad_norm": 18.360260009765625,
      "learning_rate": 3.985507246376812e-05,
      "loss": 3.609,
      "step": 110
    },
    {
      "epoch": 0.05250492233646904,
      "grad_norm": 17.4825439453125,
      "learning_rate": 4.347826086956522e-05,
      "loss": 3.1501,
      "step": 120
    },
    {
      "epoch": 0.0568803325311748,
      "grad_norm": 14.25856876373291,
      "learning_rate": 4.710144927536232e-05,
      "loss": 2.8663,
      "step": 130
    },
    {
      "epoch": 0.06125574272588055,
      "grad_norm": 10.295909881591797,
      "learning_rate": 4.999997487707664e-05,
      "loss": 2.6252,
      "step": 140
    },
    {
      "epoch": 0.0656311529205863,
      "grad_norm": 10.987821578979492,
      "learning_rate": 4.999909558006067e-05,
      "loss": 2.4812,
      "step": 150
    },
    {
      "epoch": 0.07000656311529206,
      "grad_norm": 10.583593368530273,
      "learning_rate": 4.9996960187368994e-05,
      "loss": 2.2509,
      "step": 160
    },
    {
      "epoch": 0.07438197330999781,
      "grad_norm": 7.81511926651001,
      "learning_rate": 4.9993568806295786e-05,
      "loss": 2.2127,
      "step": 170
    },
    {
      "epoch": 0.07875738350470357,
      "grad_norm": 8.283308982849121,
      "learning_rate": 4.998892160724318e-05,
      "loss": 2.1067,
      "step": 180
    },
    {
      "epoch": 0.08313279369940932,
      "grad_norm": 6.635704040527344,
      "learning_rate": 4.998301882371268e-05,
      "loss": 2.062,
      "step": 190
    },
    {
      "epoch": 0.08750820389411507,
      "grad_norm": 5.870543956756592,
      "learning_rate": 4.997586075229346e-05,
      "loss": 1.9649,
      "step": 200
    },
    {
      "epoch": 0.09188361408882083,
      "grad_norm": 5.485969543457031,
      "learning_rate": 4.996744775264743e-05,
      "loss": 1.9267,
      "step": 210
    },
    {
      "epoch": 0.09625902428352658,
      "grad_norm": 5.421557903289795,
      "learning_rate": 4.995778024749118e-05,
      "loss": 1.8854,
      "step": 220
    },
    {
      "epoch": 0.10063443447823234,
      "grad_norm": 7.398969650268555,
      "learning_rate": 4.9946858722574743e-05,
      "loss": 1.8805,
      "step": 230
    },
    {
      "epoch": 0.10500984467293809,
      "grad_norm": 5.875659465789795,
      "learning_rate": 4.993468372665718e-05,
      "loss": 1.8449,
      "step": 240
    },
    {
      "epoch": 0.10938525486764385,
      "grad_norm": 4.7640910148620605,
      "learning_rate": 4.9921255871479e-05,
      "loss": 1.7944,
      "step": 250
    },
    {
      "epoch": 0.1137606650623496,
      "grad_norm": 4.58650016784668,
      "learning_rate": 4.990657583173146e-05,
      "loss": 1.8134,
      "step": 260
    },
    {
      "epoch": 0.11813607525705534,
      "grad_norm": 6.414165496826172,
      "learning_rate": 4.989064434502262e-05,
      "loss": 1.7693,
      "step": 270
    },
    {
      "epoch": 0.1225114854517611,
      "grad_norm": 4.658475399017334,
      "learning_rate": 4.9873462211840296e-05,
      "loss": 1.7147,
      "step": 280
    },
    {
      "epoch": 0.12688689564646685,
      "grad_norm": 5.292407512664795,
      "learning_rate": 4.985503029551184e-05,
      "loss": 1.7935,
      "step": 290
    },
    {
      "epoch": 0.1312623058411726,
      "grad_norm": 5.333080291748047,
      "learning_rate": 4.983534952216079e-05,
      "loss": 1.7151,
      "step": 300
    },
    {
      "epoch": 0.13563771603587838,
      "grad_norm": 4.1388139724731445,
      "learning_rate": 4.981442088066029e-05,
      "loss": 1.7525,
      "step": 310
    },
    {
      "epoch": 0.14001312623058412,
      "grad_norm": 3.429931402206421,
      "learning_rate": 4.9792245422583406e-05,
      "loss": 1.6896,
      "step": 320
    },
    {
      "epoch": 0.14438853642528987,
      "grad_norm": 3.840290069580078,
      "learning_rate": 4.9768824262150344e-05,
      "loss": 1.6801,
      "step": 330
    },
    {
      "epoch": 0.14876394661999562,
      "grad_norm": 5.193508625030518,
      "learning_rate": 4.9744158576172406e-05,
      "loss": 1.664,
      "step": 340
    },
    {
      "epoch": 0.15313935681470137,
      "grad_norm": 3.8273701667785645,
      "learning_rate": 4.971824960399289e-05,
      "loss": 1.7366,
      "step": 350
    },
    {
      "epoch": 0.15751476700940714,
      "grad_norm": 4.011519432067871,
      "learning_rate": 4.9691098647424806e-05,
      "loss": 1.6676,
      "step": 360
    },
    {
      "epoch": 0.1618901772041129,
      "grad_norm": 3.759233236312866,
      "learning_rate": 4.9662707070685476e-05,
      "loss": 1.6592,
      "step": 370
    },
    {
      "epoch": 0.16626558739881864,
      "grad_norm": 4.202290058135986,
      "learning_rate": 4.963307630032797e-05,
      "loss": 1.6202,
      "step": 380
    },
    {
      "epoch": 0.17064099759352439,
      "grad_norm": 3.60914945602417,
      "learning_rate": 4.960220782516946e-05,
      "loss": 1.6365,
      "step": 390
    },
    {
      "epoch": 0.17501640778823013,
      "grad_norm": 3.6276028156280518,
      "learning_rate": 4.957010319621638e-05,
      "loss": 1.578,
      "step": 400
    },
    {
      "epoch": 0.1793918179829359,
      "grad_norm": 3.6587979793548584,
      "learning_rate": 4.95367640265865e-05,
      "loss": 1.6879,
      "step": 410
    },
    {
      "epoch": 0.18376722817764166,
      "grad_norm": 4.733792781829834,
      "learning_rate": 4.950219199142791e-05,
      "loss": 1.6164,
      "step": 420
    },
    {
      "epoch": 0.1881426383723474,
      "grad_norm": 3.628096103668213,
      "learning_rate": 4.9466388827834795e-05,
      "loss": 1.6271,
      "step": 430
    },
    {
      "epoch": 0.19251804856705315,
      "grad_norm": 3.9334936141967773,
      "learning_rate": 4.9429356334760205e-05,
      "loss": 1.6153,
      "step": 440
    },
    {
      "epoch": 0.19689345876175893,
      "grad_norm": 3.697252035140991,
      "learning_rate": 4.9391096372925626e-05,
      "loss": 1.5674,
      "step": 450
    },
    {
      "epoch": 0.20126886895646467,
      "grad_norm": 5.91195011138916,
      "learning_rate": 4.9351610864727504e-05,
      "loss": 1.5977,
      "step": 460
    },
    {
      "epoch": 0.20564427915117042,
      "grad_norm": 4.414886474609375,
      "learning_rate": 4.9310901794140665e-05,
      "loss": 1.6158,
      "step": 470
    },
    {
      "epoch": 0.21001968934587617,
      "grad_norm": 4.002306938171387,
      "learning_rate": 4.92689712066186e-05,
      "loss": 1.6071,
      "step": 480
    },
    {
      "epoch": 0.21439509954058192,
      "grad_norm": 2.9803686141967773,
      "learning_rate": 4.922582120899072e-05,
      "loss": 1.5951,
      "step": 490
    },
    {
      "epoch": 0.2187705097352877,
      "grad_norm": 3.6894145011901855,
      "learning_rate": 4.918145396935648e-05,
      "loss": 1.6099,
      "step": 500
    },
    {
      "epoch": 0.22314591992999344,
      "grad_norm": 3.097684383392334,
      "learning_rate": 4.913587171697643e-05,
      "loss": 1.5629,
      "step": 510
    },
    {
      "epoch": 0.2275213301246992,
      "grad_norm": 3.0725483894348145,
      "learning_rate": 4.908907674216023e-05,
      "loss": 1.5382,
      "step": 520
    },
    {
      "epoch": 0.23189674031940494,
      "grad_norm": 3.557966947555542,
      "learning_rate": 4.9041071396151585e-05,
      "loss": 1.5952,
      "step": 530
    },
    {
      "epoch": 0.23627215051411068,
      "grad_norm": 3.983081817626953,
      "learning_rate": 4.899185809101003e-05,
      "loss": 1.5553,
      "step": 540
    },
    {
      "epoch": 0.24064756070881646,
      "grad_norm": 4.9868974685668945,
      "learning_rate": 4.894143929948982e-05,
      "loss": 1.5193,
      "step": 550
    },
    {
      "epoch": 0.2450229709035222,
      "grad_norm": 2.7370195388793945,
      "learning_rate": 4.8889817554915645e-05,
      "loss": 1.5844,
      "step": 560
    },
    {
      "epoch": 0.24939838109822796,
      "grad_norm": 3.5751953125,
      "learning_rate": 4.8836995451055344e-05,
      "loss": 1.5749,
      "step": 570
    },
    {
      "epoch": 0.2537737912929337,
      "grad_norm": 3.7053322792053223,
      "learning_rate": 4.878297564198959e-05,
      "loss": 1.507,
      "step": 580
    },
    {
      "epoch": 0.2581492014876395,
      "grad_norm": 3.4896960258483887,
      "learning_rate": 4.872776084197851e-05,
      "loss": 1.5189,
      "step": 590
    },
    {
      "epoch": 0.2625246116823452,
      "grad_norm": 3.1286392211914062,
      "learning_rate": 4.867135382532533e-05,
      "loss": 1.5693,
      "step": 600
    },
    {
      "epoch": 0.266900021877051,
      "grad_norm": 2.978731870651245,
      "learning_rate": 4.861375742623697e-05,
      "loss": 1.5304,
      "step": 610
    },
    {
      "epoch": 0.27127543207175675,
      "grad_norm": 7.503092288970947,
      "learning_rate": 4.855497453868164e-05,
      "loss": 1.5307,
      "step": 620
    },
    {
      "epoch": 0.27565084226646247,
      "grad_norm": 3.5043628215789795,
      "learning_rate": 4.849500811624343e-05,
      "loss": 1.5363,
      "step": 630
    },
    {
      "epoch": 0.28002625246116825,
      "grad_norm": 2.7723424434661865,
      "learning_rate": 4.8433861171973896e-05,
      "loss": 1.5473,
      "step": 640
    },
    {
      "epoch": 0.28440166265587397,
      "grad_norm": 2.5967185497283936,
      "learning_rate": 4.837153677824067e-05,
      "loss": 1.5448,
      "step": 650
    },
    {
      "epoch": 0.28877707285057974,
      "grad_norm": 2.6320347785949707,
      "learning_rate": 4.830803806657311e-05,
      "loss": 1.5338,
      "step": 660
    },
    {
      "epoch": 0.2931524830452855,
      "grad_norm": 3.092648506164551,
      "learning_rate": 4.8243368227504924e-05,
      "loss": 1.5539,
      "step": 670
    },
    {
      "epoch": 0.29752789323999124,
      "grad_norm": 3.1972124576568604,
      "learning_rate": 4.817753051041387e-05,
      "loss": 1.5186,
      "step": 680
    },
    {
      "epoch": 0.301903303434697,
      "grad_norm": 3.0039663314819336,
      "learning_rate": 4.811052822335849e-05,
      "loss": 1.558,
      "step": 690
    },
    {
      "epoch": 0.30627871362940273,
      "grad_norm": 2.77423357963562,
      "learning_rate": 4.80423647329119e-05,
      "loss": 1.489,
      "step": 700
    },
    {
      "epoch": 0.3106541238241085,
      "grad_norm": 3.202047348022461,
      "learning_rate": 4.797304346399263e-05,
      "loss": 1.5308,
      "step": 710
    },
    {
      "epoch": 0.3150295340188143,
      "grad_norm": 3.66243052482605,
      "learning_rate": 4.790256789969252e-05,
      "loss": 1.5809,
      "step": 720
    },
    {
      "epoch": 0.31940494421352,
      "grad_norm": 2.8688106536865234,
      "learning_rate": 4.783094158110174e-05,
      "loss": 1.5135,
      "step": 730
    },
    {
      "epoch": 0.3237803544082258,
      "grad_norm": 2.723235607147217,
      "learning_rate": 4.775816810713086e-05,
      "loss": 1.5087,
      "step": 740
    },
    {
      "epoch": 0.3281557646029315,
      "grad_norm": 2.867291212081909,
      "learning_rate": 4.768425113433e-05,
      "loss": 1.5495,
      "step": 750
    },
    {
      "epoch": 0.3325311747976373,
      "grad_norm": 2.7260971069335938,
      "learning_rate": 4.7609194376705105e-05,
      "loss": 1.5182,
      "step": 760
    },
    {
      "epoch": 0.33690658499234305,
      "grad_norm": 2.8537180423736572,
      "learning_rate": 4.753300160553136e-05,
      "loss": 1.5718,
      "step": 770
    },
    {
      "epoch": 0.34128199518704877,
      "grad_norm": 2.657478094100952,
      "learning_rate": 4.745567664916368e-05,
      "loss": 1.5206,
      "step": 780
    },
    {
      "epoch": 0.34565740538175455,
      "grad_norm": 2.704979419708252,
      "learning_rate": 4.737722339284436e-05,
      "loss": 1.484,
      "step": 790
    },
    {
      "epoch": 0.35003281557646027,
      "grad_norm": 2.9366772174835205,
      "learning_rate": 4.7297645778507835e-05,
      "loss": 1.4317,
      "step": 800
    },
    {
      "epoch": 0.35440822577116604,
      "grad_norm": 2.9852333068847656,
      "learning_rate": 4.721694780458266e-05,
      "loss": 1.5432,
      "step": 810
    },
    {
      "epoch": 0.3587836359658718,
      "grad_norm": 2.6309895515441895,
      "learning_rate": 4.7135133525790544e-05,
      "loss": 1.5344,
      "step": 820
    },
    {
      "epoch": 0.36315904616057754,
      "grad_norm": 2.7915596961975098,
      "learning_rate": 4.7052207052942685e-05,
      "loss": 1.4518,
      "step": 830
    },
    {
      "epoch": 0.3675344563552833,
      "grad_norm": 3.265641689300537,
      "learning_rate": 4.6968172552733185e-05,
      "loss": 1.5039,
      "step": 840
    },
    {
      "epoch": 0.3719098665499891,
      "grad_norm": 3.012242078781128,
      "learning_rate": 4.688303424752969e-05,
      "loss": 1.4626,
      "step": 850
    },
    {
      "epoch": 0.3762852767446948,
      "grad_norm": 2.732055425643921,
      "learning_rate": 4.679679641516122e-05,
      "loss": 1.4853,
      "step": 860
    },
    {
      "epoch": 0.3806606869394006,
      "grad_norm": 3.42232346534729,
      "learning_rate": 4.670946338870329e-05,
      "loss": 1.4678,
      "step": 870
    },
    {
      "epoch": 0.3850360971341063,
      "grad_norm": 2.7350080013275146,
      "learning_rate": 4.6621039556260095e-05,
      "loss": 1.4926,
      "step": 880
    },
    {
      "epoch": 0.3894115073288121,
      "grad_norm": 2.6931347846984863,
      "learning_rate": 4.653152936074413e-05,
      "loss": 1.4652,
      "step": 890
    },
    {
      "epoch": 0.39378691752351785,
      "grad_norm": 3.128669023513794,
      "learning_rate": 4.644093729965287e-05,
      "loss": 1.461,
      "step": 900
    },
    {
      "epoch": 0.3981623277182236,
      "grad_norm": 2.5737152099609375,
      "learning_rate": 4.634926792484284e-05,
      "loss": 1.523,
      "step": 910
    },
    {
      "epoch": 0.40253773791292935,
      "grad_norm": 2.422856092453003,
      "learning_rate": 4.625652584230089e-05,
      "loss": 1.5155,
      "step": 920
    },
    {
      "epoch": 0.40691314810763507,
      "grad_norm": 2.1890969276428223,
      "learning_rate": 4.616271571191273e-05,
      "loss": 1.5173,
      "step": 930
    },
    {
      "epoch": 0.41128855830234085,
      "grad_norm": 2.422394037246704,
      "learning_rate": 4.606784224722887e-05,
      "loss": 1.4966,
      "step": 940
    },
    {
      "epoch": 0.4156639684970466,
      "grad_norm": 2.413853406906128,
      "learning_rate": 4.5971910215227684e-05,
      "loss": 1.4246,
      "step": 950
    },
    {
      "epoch": 0.42003937869175234,
      "grad_norm": 2.3929250240325928,
      "learning_rate": 4.5874924436075996e-05,
      "loss": 1.4887,
      "step": 960
    },
    {
      "epoch": 0.4244147888864581,
      "grad_norm": 2.542177200317383,
      "learning_rate": 4.577688978288681e-05,
      "loss": 1.5036,
      "step": 970
    },
    {
      "epoch": 0.42879019908116384,
      "grad_norm": 2.4133691787719727,
      "learning_rate": 4.567781118147448e-05,
      "loss": 1.4935,
      "step": 980
    },
    {
      "epoch": 0.4331656092758696,
      "grad_norm": 2.8484768867492676,
      "learning_rate": 4.557769361010721e-05,
      "loss": 1.4317,
      "step": 990
    },
    {
      "epoch": 0.4375410194705754,
      "grad_norm": 2.7544851303100586,
      "learning_rate": 4.547654209925693e-05,
      "loss": 1.4674,
      "step": 1000
    },
    {
      "epoch": 0.4419164296652811,
      "grad_norm": 2.514936923980713,
      "learning_rate": 4.5374361731346526e-05,
      "loss": 1.4525,
      "step": 1010
    },
    {
      "epoch": 0.4462918398599869,
      "grad_norm": 2.672096014022827,
      "learning_rate": 4.5271157640494456e-05,
      "loss": 1.485,
      "step": 1020
    },
    {
      "epoch": 0.4506672500546926,
      "grad_norm": 2.681631088256836,
      "learning_rate": 4.51669350122568e-05,
      "loss": 1.503,
      "step": 1030
    },
    {
      "epoch": 0.4550426602493984,
      "grad_norm": 2.664799213409424,
      "learning_rate": 4.506169908336673e-05,
      "loss": 1.5043,
      "step": 1040
    },
    {
      "epoch": 0.45941807044410415,
      "grad_norm": 2.5863938331604004,
      "learning_rate": 4.495545514147134e-05,
      "loss": 1.4476,
      "step": 1050
    },
    {
      "epoch": 0.4637934806388099,
      "grad_norm": 2.8567678928375244,
      "learning_rate": 4.484820852486598e-05,
      "loss": 1.4886,
      "step": 1060
    },
    {
      "epoch": 0.46816889083351565,
      "grad_norm": 2.551856279373169,
      "learning_rate": 4.473996462222606e-05,
      "loss": 1.4478,
      "step": 1070
    },
    {
      "epoch": 0.47254430102822137,
      "grad_norm": 2.7946434020996094,
      "learning_rate": 4.463072887233625e-05,
      "loss": 1.5019,
      "step": 1080
    },
    {
      "epoch": 0.47691971122292715,
      "grad_norm": 2.5749003887176514,
      "learning_rate": 4.452050676381725e-05,
      "loss": 1.5127,
      "step": 1090
    },
    {
      "epoch": 0.4812951214176329,
      "grad_norm": 2.6211116313934326,
      "learning_rate": 4.440930383484995e-05,
      "loss": 1.4624,
      "step": 1100
    },
    {
      "epoch": 0.48567053161233864,
      "grad_norm": 3.0213544368743896,
      "learning_rate": 4.429712567289722e-05,
      "loss": 1.4889,
      "step": 1110
    },
    {
      "epoch": 0.4900459418070444,
      "grad_norm": 2.861062526702881,
      "learning_rate": 4.418397791442314e-05,
      "loss": 1.4922,
      "step": 1120
    },
    {
      "epoch": 0.4944213520017502,
      "grad_norm": 2.391467332839966,
      "learning_rate": 4.406986624460979e-05,
      "loss": 1.5009,
      "step": 1130
    },
    {
      "epoch": 0.4987967621964559,
      "grad_norm": 2.5782501697540283,
      "learning_rate": 4.395479639707158e-05,
      "loss": 1.4218,
      "step": 1140
    },
    {
      "epoch": 0.5031721723911616,
      "grad_norm": 2.4004220962524414,
      "learning_rate": 4.3838774153567195e-05,
      "loss": 1.4559,
      "step": 1150
    },
    {
      "epoch": 0.5075475825858674,
      "grad_norm": 2.3091509342193604,
      "learning_rate": 4.372180534370908e-05,
      "loss": 1.4379,
      "step": 1160
    },
    {
      "epoch": 0.5119229927805732,
      "grad_norm": 2.7878313064575195,
      "learning_rate": 4.3603895844670495e-05,
      "loss": 1.4394,
      "step": 1170
    },
    {
      "epoch": 0.516298402975279,
      "grad_norm": 2.5346367359161377,
      "learning_rate": 4.3485051580890236e-05,
      "loss": 1.4965,
      "step": 1180
    },
    {
      "epoch": 0.5206738131699847,
      "grad_norm": 2.4604122638702393,
      "learning_rate": 4.336527852377497e-05,
      "loss": 1.4792,
      "step": 1190
    },
    {
      "epoch": 0.5250492233646904,
      "grad_norm": 2.473877191543579,
      "learning_rate": 4.324458269139916e-05,
      "loss": 1.4478,
      "step": 1200
    },
    {
      "epoch": 0.5294246335593962,
      "grad_norm": 2.3778798580169678,
      "learning_rate": 4.3122970148202744e-05,
      "loss": 1.4573,
      "step": 1210
    },
    {
      "epoch": 0.533800043754102,
      "grad_norm": 2.4219954013824463,
      "learning_rate": 4.300044700468635e-05,
      "loss": 1.4806,
      "step": 1220
    },
    {
      "epoch": 0.5381754539488077,
      "grad_norm": 2.640268564224243,
      "learning_rate": 4.2877019417104334e-05,
      "loss": 1.4323,
      "step": 1230
    },
    {
      "epoch": 0.5425508641435135,
      "grad_norm": 2.3779873847961426,
      "learning_rate": 4.27526935871554e-05,
      "loss": 1.4614,
      "step": 1240
    },
    {
      "epoch": 0.5469262743382192,
      "grad_norm": 2.5593316555023193,
      "learning_rate": 4.262747576167106e-05,
      "loss": 1.4291,
      "step": 1250
    },
    {
      "epoch": 0.5513016845329249,
      "grad_norm": 2.619673252105713,
      "learning_rate": 4.250137223230165e-05,
      "loss": 1.4378,
      "step": 1260
    },
    {
      "epoch": 0.5556770947276307,
      "grad_norm": 2.7467172145843506,
      "learning_rate": 4.2374389335200335e-05,
      "loss": 1.4629,
      "step": 1270
    },
    {
      "epoch": 0.5600525049223365,
      "grad_norm": 2.7677481174468994,
      "learning_rate": 4.224653345070466e-05,
      "loss": 1.3885,
      "step": 1280
    },
    {
      "epoch": 0.5644279151170423,
      "grad_norm": 2.4805145263671875,
      "learning_rate": 4.211781100301596e-05,
      "loss": 1.4683,
      "step": 1290
    },
    {
      "epoch": 0.5688033253117479,
      "grad_norm": 2.3270130157470703,
      "learning_rate": 4.198822845987661e-05,
      "loss": 1.4484,
      "step": 1300
    },
    {
      "epoch": 0.5731787355064537,
      "grad_norm": 2.714475393295288,
      "learning_rate": 4.185779233224506e-05,
      "loss": 1.3896,
      "step": 1310
    },
    {
      "epoch": 0.5775541457011595,
      "grad_norm": 2.5880696773529053,
      "learning_rate": 4.172650917396863e-05,
      "loss": 1.5171,
      "step": 1320
    },
    {
      "epoch": 0.5819295558958653,
      "grad_norm": 2.7589426040649414,
      "learning_rate": 4.1594385581454245e-05,
      "loss": 1.5071,
      "step": 1330
    },
    {
      "epoch": 0.586304966090571,
      "grad_norm": 2.3339285850524902,
      "learning_rate": 4.146142819333702e-05,
      "loss": 1.446,
      "step": 1340
    },
    {
      "epoch": 0.5906803762852767,
      "grad_norm": 2.597846746444702,
      "learning_rate": 4.132764369014662e-05,
      "loss": 1.5085,
      "step": 1350
    },
    {
      "epoch": 0.5950557864799825,
      "grad_norm": 2.3371880054473877,
      "learning_rate": 4.119303879397169e-05,
      "loss": 1.4464,
      "step": 1360
    },
    {
      "epoch": 0.5994311966746882,
      "grad_norm": 2.333690881729126,
      "learning_rate": 4.1057620268122004e-05,
      "loss": 1.4367,
      "step": 1370
    },
    {
      "epoch": 0.603806606869394,
      "grad_norm": 2.8487133979797363,
      "learning_rate": 4.092139491678873e-05,
      "loss": 1.4151,
      "step": 1380
    },
    {
      "epoch": 0.6081820170640998,
      "grad_norm": 2.409637212753296,
      "learning_rate": 4.078436958470244e-05,
      "loss": 1.4162,
      "step": 1390
    },
    {
      "epoch": 0.6125574272588055,
      "grad_norm": 2.4450595378875732,
      "learning_rate": 4.064655115678933e-05,
      "loss": 1.3824,
      "step": 1400
    },
    {
      "epoch": 0.6169328374535112,
      "grad_norm": 2.278193950653076,
      "learning_rate": 4.050794655782515e-05,
      "loss": 1.4173,
      "step": 1410
    },
    {
      "epoch": 0.621308247648217,
      "grad_norm": 2.50050950050354,
      "learning_rate": 4.036856275208736e-05,
      "loss": 1.3888,
      "step": 1420
    },
    {
      "epoch": 0.6256836578429228,
      "grad_norm": 2.2079274654388428,
      "learning_rate": 4.022840674300512e-05,
      "loss": 1.4386,
      "step": 1430
    },
    {
      "epoch": 0.6300590680376286,
      "grad_norm": 2.575258255004883,
      "learning_rate": 4.0087485572807485e-05,
      "loss": 1.387,
      "step": 1440
    },
    {
      "epoch": 0.6344344782323342,
      "grad_norm": 2.4015772342681885,
      "learning_rate": 3.994580632216952e-05,
      "loss": 1.4638,
      "step": 1450
    },
    {
      "epoch": 0.63880988842704,
      "grad_norm": 2.506242275238037,
      "learning_rate": 3.980337610985651e-05,
      "loss": 1.4322,
      "step": 1460
    },
    {
      "epoch": 0.6431852986217458,
      "grad_norm": 2.6933438777923584,
      "learning_rate": 3.9660202092366316e-05,
      "loss": 1.4494,
      "step": 1470
    },
    {
      "epoch": 0.6475607088164516,
      "grad_norm": 2.4342904090881348,
      "learning_rate": 3.951629146356975e-05,
      "loss": 1.4017,
      "step": 1480
    },
    {
      "epoch": 0.6519361190111573,
      "grad_norm": 2.187755823135376,
      "learning_rate": 3.937165145434914e-05,
      "loss": 1.4485,
      "step": 1490
    },
    {
      "epoch": 0.656311529205863,
      "grad_norm": 2.433896541595459,
      "learning_rate": 3.922628933223502e-05,
      "loss": 1.4724,
      "step": 1500
    },
    {
      "epoch": 0.6606869394005688,
      "grad_norm": 2.2796449661254883,
      "learning_rate": 3.9080212401040914e-05,
      "loss": 1.3787,
      "step": 1510
    },
    {
      "epoch": 0.6650623495952745,
      "grad_norm": 2.273935079574585,
      "learning_rate": 3.8933428000496455e-05,
      "loss": 1.458,
      "step": 1520
    },
    {
      "epoch": 0.6694377597899803,
      "grad_norm": 2.4992547035217285,
      "learning_rate": 3.878594350587844e-05,
      "loss": 1.3719,
      "step": 1530
    },
    {
      "epoch": 0.6738131699846861,
      "grad_norm": 2.31856369972229,
      "learning_rate": 3.863776632764043e-05,
      "loss": 1.4759,
      "step": 1540
    },
    {
      "epoch": 0.6781885801793918,
      "grad_norm": 2.4243216514587402,
      "learning_rate": 3.848890391104028e-05,
      "loss": 1.4131,
      "step": 1550
    },
    {
      "epoch": 0.6825639903740975,
      "grad_norm": 2.269465208053589,
      "learning_rate": 3.833936373576606e-05,
      "loss": 1.4198,
      "step": 1560
    },
    {
      "epoch": 0.6869394005688033,
      "grad_norm": 2.356196403503418,
      "learning_rate": 3.8189153315560324e-05,
      "loss": 1.4296,
      "step": 1570
    },
    {
      "epoch": 0.6913148107635091,
      "grad_norm": 2.0902884006500244,
      "learning_rate": 3.803828019784248e-05,
      "loss": 1.4082,
      "step": 1580
    },
    {
      "epoch": 0.6956902209582149,
      "grad_norm": 2.3475914001464844,
      "learning_rate": 3.788675196332959e-05,
      "loss": 1.4851,
      "step": 1590
    },
    {
      "epoch": 0.7000656311529205,
      "grad_norm": 2.5512330532073975,
      "learning_rate": 3.7734576225655504e-05,
      "loss": 1.487,
      "step": 1600
    },
    {
      "epoch": 0.7044410413476263,
      "grad_norm": 2.4078125953674316,
      "learning_rate": 3.758176063098829e-05,
      "loss": 1.394,
      "step": 1610
    },
    {
      "epoch": 0.7088164515423321,
      "grad_norm": 2.2228713035583496,
      "learning_rate": 3.742831285764602e-05,
      "loss": 1.391,
      "step": 1620
    },
    {
      "epoch": 0.7131918617370379,
      "grad_norm": 2.348409414291382,
      "learning_rate": 3.7274240615710986e-05,
      "loss": 1.3963,
      "step": 1630
    },
    {
      "epoch": 0.7175672719317436,
      "grad_norm": 2.439955711364746,
      "learning_rate": 3.711955164664235e-05,
      "loss": 1.4384,
      "step": 1640
    },
    {
      "epoch": 0.7219426821264494,
      "grad_norm": 2.3827335834503174,
      "learning_rate": 3.6964253722887075e-05,
      "loss": 1.415,
      "step": 1650
    },
    {
      "epoch": 0.7263180923211551,
      "grad_norm": 2.362487316131592,
      "learning_rate": 3.680835464748949e-05,
      "loss": 1.3842,
      "step": 1660
    },
    {
      "epoch": 0.7306935025158608,
      "grad_norm": 2.2841691970825195,
      "learning_rate": 3.6651862253699144e-05,
      "loss": 1.3775,
      "step": 1670
    },
    {
      "epoch": 0.7350689127105666,
      "grad_norm": 2.589276075363159,
      "learning_rate": 3.649478440457726e-05,
      "loss": 1.4037,
      "step": 1680
    },
    {
      "epoch": 0.7394443229052724,
      "grad_norm": 2.2496328353881836,
      "learning_rate": 3.633712899260166e-05,
      "loss": 1.382,
      "step": 1690
    },
    {
      "epoch": 0.7438197330999782,
      "grad_norm": 2.38676381111145,
      "learning_rate": 3.617890393927015e-05,
      "loss": 1.4889,
      "step": 1700
    },
    {
      "epoch": 0.7481951432946838,
      "grad_norm": 2.377511739730835,
      "learning_rate": 3.6020117194702555e-05,
      "loss": 1.381,
      "step": 1710
    },
    {
      "epoch": 0.7525705534893896,
      "grad_norm": 2.30507755279541,
      "learning_rate": 3.586077673724124e-05,
      "loss": 1.4403,
      "step": 1720
    },
    {
      "epoch": 0.7569459636840954,
      "grad_norm": 2.377375602722168,
      "learning_rate": 3.570089057305022e-05,
      "loss": 1.4609,
      "step": 1730
    },
    {
      "epoch": 0.7613213738788012,
      "grad_norm": 2.38913631439209,
      "learning_rate": 3.5540466735712906e-05,
      "loss": 1.4117,
      "step": 1740
    },
    {
      "epoch": 0.7656967840735069,
      "grad_norm": 2.3428103923797607,
      "learning_rate": 3.5379513285828433e-05,
      "loss": 1.4044,
      "step": 1750
    },
    {
      "epoch": 0.7700721942682126,
      "grad_norm": 2.3192853927612305,
      "learning_rate": 3.521803831060668e-05,
      "loss": 1.3802,
      "step": 1760
    },
    {
      "epoch": 0.7744476044629184,
      "grad_norm": 2.675574779510498,
      "learning_rate": 3.505604992346187e-05,
      "loss": 1.4073,
      "step": 1770
    },
    {
      "epoch": 0.7788230146576242,
      "grad_norm": 2.3397696018218994,
      "learning_rate": 3.489355626360497e-05,
      "loss": 1.4628,
      "step": 1780
    },
    {
      "epoch": 0.7831984248523299,
      "grad_norm": 2.750725746154785,
      "learning_rate": 3.473056549563469e-05,
      "loss": 1.4195,
      "step": 1790
    },
    {
      "epoch": 0.7875738350470357,
      "grad_norm": 2.485304832458496,
      "learning_rate": 3.456708580912725e-05,
      "loss": 1.4301,
      "step": 1800
    },
    {
      "epoch": 0.7919492452417414,
      "grad_norm": 2.172199010848999,
      "learning_rate": 3.44031254182249e-05,
      "loss": 1.378,
      "step": 1810
    },
    {
      "epoch": 0.7963246554364471,
      "grad_norm": 2.1977224349975586,
      "learning_rate": 3.42386925612232e-05,
      "loss": 1.3745,
      "step": 1820
    },
    {
      "epoch": 0.8007000656311529,
      "grad_norm": 2.402909517288208,
      "learning_rate": 3.407379550015707e-05,
      "loss": 1.4497,
      "step": 1830
    },
    {
      "epoch": 0.8050754758258587,
      "grad_norm": 2.286860466003418,
      "learning_rate": 3.390844252038565e-05,
      "loss": 1.3943,
      "step": 1840
    },
    {
      "epoch": 0.8094508860205645,
      "grad_norm": 2.354567527770996,
      "learning_rate": 3.3742641930176045e-05,
      "loss": 1.3246,
      "step": 1850
    },
    {
      "epoch": 0.8138262962152701,
      "grad_norm": 2.4535744190216064,
      "learning_rate": 3.3576402060285786e-05,
      "loss": 1.412,
      "step": 1860
    },
    {
      "epoch": 0.8182017064099759,
      "grad_norm": 2.4066927433013916,
      "learning_rate": 3.3409731263544345e-05,
      "loss": 1.3651,
      "step": 1870
    },
    {
      "epoch": 0.8225771166046817,
      "grad_norm": 2.4528560638427734,
      "learning_rate": 3.324263791443333e-05,
      "loss": 1.3898,
      "step": 1880
    },
    {
      "epoch": 0.8269525267993875,
      "grad_norm": 2.25467586517334,
      "learning_rate": 3.307513040866584e-05,
      "loss": 1.3769,
      "step": 1890
    },
    {
      "epoch": 0.8313279369940932,
      "grad_norm": 2.2125041484832764,
      "learning_rate": 3.290721716276448e-05,
      "loss": 1.3883,
      "step": 1900
    },
    {
      "epoch": 0.8357033471887989,
      "grad_norm": 2.3503026962280273,
      "learning_rate": 3.273890661363855e-05,
      "loss": 1.4702,
      "step": 1910
    },
    {
      "epoch": 0.8400787573835047,
      "grad_norm": 2.447185754776001,
      "learning_rate": 3.257020721816013e-05,
      "loss": 1.4266,
      "step": 1920
    },
    {
      "epoch": 0.8444541675782105,
      "grad_norm": 2.5357518196105957,
      "learning_rate": 3.240112745273911e-05,
      "loss": 1.42,
      "step": 1930
    },
    {
      "epoch": 0.8488295777729162,
      "grad_norm": 2.4189131259918213,
      "learning_rate": 3.223167581289729e-05,
      "loss": 1.4111,
      "step": 1940
    },
    {
      "epoch": 0.853204987967622,
      "grad_norm": 2.338618516921997,
      "learning_rate": 3.206186081284161e-05,
      "loss": 1.3689,
      "step": 1950
    },
    {
      "epoch": 0.8575803981623277,
      "grad_norm": 2.4587111473083496,
      "learning_rate": 3.1891690985036215e-05,
      "loss": 1.4042,
      "step": 1960
    },
    {
      "epoch": 0.8619558083570334,
      "grad_norm": 2.3514132499694824,
      "learning_rate": 3.1721174879773826e-05,
      "loss": 1.378,
      "step": 1970
    },
    {
      "epoch": 0.8663312185517392,
      "grad_norm": 2.7501258850097656,
      "learning_rate": 3.1550321064746086e-05,
      "loss": 1.3892,
      "step": 1980
    },
    {
      "epoch": 0.870706628746445,
      "grad_norm": 2.2368621826171875,
      "learning_rate": 3.1379138124613105e-05,
      "loss": 1.3823,
      "step": 1990
    },
    {
      "epoch": 0.8750820389411508,
      "grad_norm": 2.383025884628296,
      "learning_rate": 3.120763466057206e-05,
      "loss": 1.3619,
      "step": 2000
    },
    {
      "epoch": 0.8794574491358564,
      "grad_norm": 2.347883939743042,
      "learning_rate": 3.10358192899251e-05,
      "loss": 1.4178,
      "step": 2010
    },
    {
      "epoch": 0.8838328593305622,
      "grad_norm": 2.2251715660095215,
      "learning_rate": 3.086370064564629e-05,
      "loss": 1.3897,
      "step": 2020
    },
    {
      "epoch": 0.888208269525268,
      "grad_norm": 2.6489715576171875,
      "learning_rate": 3.069128737594789e-05,
      "loss": 1.4098,
      "step": 2030
    },
    {
      "epoch": 0.8925836797199738,
      "grad_norm": 2.17360520362854,
      "learning_rate": 3.051858814384581e-05,
      "loss": 1.3731,
      "step": 2040
    },
    {
      "epoch": 0.8969590899146795,
      "grad_norm": 2.4053566455841064,
      "learning_rate": 3.0345611626724285e-05,
      "loss": 1.4091,
      "step": 2050
    },
    {
      "epoch": 0.9013345001093852,
      "grad_norm": 2.2361433506011963,
      "learning_rate": 3.0172366515899982e-05,
      "loss": 1.3995,
      "step": 2060
    },
    {
      "epoch": 0.905709910304091,
      "grad_norm": 2.22719669342041,
      "learning_rate": 2.9998861516185173e-05,
      "loss": 1.3895,
      "step": 2070
    },
    {
      "epoch": 0.9100853204987968,
      "grad_norm": 2.1987080574035645,
      "learning_rate": 2.982510534545043e-05,
      "loss": 1.3554,
      "step": 2080
    },
    {
      "epoch": 0.9144607306935025,
      "grad_norm": 2.434157609939575,
      "learning_rate": 2.9651106734186618e-05,
      "loss": 1.3782,
      "step": 2090
    },
    {
      "epoch": 0.9188361408882083,
      "grad_norm": 2.3075335025787354,
      "learning_rate": 2.9476874425066105e-05,
      "loss": 1.4163,
      "step": 2100
    },
    {
      "epoch": 0.923211551082914,
      "grad_norm": 2.301095485687256,
      "learning_rate": 2.9302417172503622e-05,
      "loss": 1.4531,
      "step": 2110
    },
    {
      "epoch": 0.9275869612776197,
      "grad_norm": 2.3101842403411865,
      "learning_rate": 2.9127743742216303e-05,
      "loss": 1.3829,
      "step": 2120
    },
    {
      "epoch": 0.9319623714723255,
      "grad_norm": 2.3680262565612793,
      "learning_rate": 2.8952862910783267e-05,
      "loss": 1.3996,
      "step": 2130
    },
    {
      "epoch": 0.9363377816670313,
      "grad_norm": 2.216562271118164,
      "learning_rate": 2.877778346520466e-05,
      "loss": 1.358,
      "step": 2140
    },
    {
      "epoch": 0.9407131918617371,
      "grad_norm": 2.2377235889434814,
      "learning_rate": 2.860251420246009e-05,
      "loss": 1.4145,
      "step": 2150
    },
    {
      "epoch": 0.9450886020564427,
      "grad_norm": 2.5587806701660156,
      "learning_rate": 2.842706392906669e-05,
      "loss": 1.4431,
      "step": 2160
    },
    {
      "epoch": 0.9494640122511485,
      "grad_norm": 2.181572198867798,
      "learning_rate": 2.825144146063654e-05,
      "loss": 1.4452,
      "step": 2170
    },
    {
      "epoch": 0.9538394224458543,
      "grad_norm": 2.3660833835601807,
      "learning_rate": 2.8075655621433826e-05,
      "loss": 1.3885,
      "step": 2180
    },
    {
      "epoch": 0.9582148326405601,
      "grad_norm": 2.268667459487915,
      "learning_rate": 2.7899715243931375e-05,
      "loss": 1.3624,
      "step": 2190
    },
    {
      "epoch": 0.9625902428352658,
      "grad_norm": 2.330254316329956,
      "learning_rate": 2.7723629168366883e-05,
      "loss": 1.4075,
      "step": 2200
    },
    {
      "epoch": 0.9669656530299715,
      "grad_norm": 2.351877450942993,
      "learning_rate": 2.754740624229875e-05,
      "loss": 1.3761,
      "step": 2210
    },
    {
      "epoch": 0.9713410632246773,
      "grad_norm": 2.4575061798095703,
      "learning_rate": 2.737105532016152e-05,
      "loss": 1.3712,
      "step": 2220
    },
    {
      "epoch": 0.9757164734193831,
      "grad_norm": 2.4166197776794434,
      "learning_rate": 2.719458526282099e-05,
      "loss": 1.3884,
      "step": 2230
    },
    {
      "epoch": 0.9800918836140888,
      "grad_norm": 2.2318737506866455,
      "learning_rate": 2.7018004937128937e-05,
      "loss": 1.3351,
      "step": 2240
    },
    {
      "epoch": 0.9844672938087946,
      "grad_norm": 2.2672455310821533,
      "learning_rate": 2.6841323215477714e-05,
      "loss": 1.3915,
      "step": 2250
    },
    {
      "epoch": 0.9888427040035004,
      "grad_norm": 2.065308094024658,
      "learning_rate": 2.6664548975354332e-05,
      "loss": 1.3697,
      "step": 2260
    },
    {
      "epoch": 0.993218114198206,
      "grad_norm": 2.2431817054748535,
      "learning_rate": 2.6487691098894447e-05,
      "loss": 1.3566,
      "step": 2270
    },
    {
      "epoch": 0.9975935243929118,
      "grad_norm": 2.384570837020874,
      "learning_rate": 2.6310758472436103e-05,
      "loss": 1.3354,
      "step": 2280
    }
  ],
  "logging_steps": 10,
  "max_steps": 4570,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 500.0,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 9.493158922149888e+17,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}