{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 1629,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01841620626151013,
      "grad_norm": 3.54826512795615,
      "learning_rate": 5.521472392638038e-07,
      "loss": 0.4164,
      "step": 10
    },
    {
      "epoch": 0.03683241252302026,
      "grad_norm": 2.150976325911803,
      "learning_rate": 1.165644171779141e-06,
      "loss": 0.3752,
      "step": 20
    },
    {
      "epoch": 0.055248618784530384,
      "grad_norm": 1.301734345190514,
      "learning_rate": 1.7791411042944787e-06,
      "loss": 0.3446,
      "step": 30
    },
    {
      "epoch": 0.07366482504604052,
      "grad_norm": 1.0129957906382288,
      "learning_rate": 2.392638036809816e-06,
      "loss": 0.3248,
      "step": 40
    },
    {
      "epoch": 0.09208103130755065,
      "grad_norm": 1.0713399309035538,
      "learning_rate": 3.0061349693251535e-06,
      "loss": 0.3026,
      "step": 50
    },
    {
      "epoch": 0.11049723756906077,
      "grad_norm": 1.0507425813581446,
      "learning_rate": 3.6196319018404913e-06,
      "loss": 0.304,
      "step": 60
    },
    {
      "epoch": 0.1289134438305709,
      "grad_norm": 0.8381317525552241,
      "learning_rate": 4.233128834355829e-06,
      "loss": 0.291,
      "step": 70
    },
    {
      "epoch": 0.14732965009208104,
      "grad_norm": 0.9207152395423905,
      "learning_rate": 4.846625766871166e-06,
      "loss": 0.2962,
      "step": 80
    },
    {
      "epoch": 0.16574585635359115,
      "grad_norm": 0.9703285578071639,
      "learning_rate": 5.460122699386503e-06,
      "loss": 0.2877,
      "step": 90
    },
    {
      "epoch": 0.1841620626151013,
      "grad_norm": 0.9941438616319775,
      "learning_rate": 6.073619631901841e-06,
      "loss": 0.2894,
      "step": 100
    },
    {
      "epoch": 0.20257826887661143,
      "grad_norm": 1.0580118824781048,
      "learning_rate": 6.687116564417178e-06,
      "loss": 0.2865,
      "step": 110
    },
    {
      "epoch": 0.22099447513812154,
      "grad_norm": 1.0527321061745536,
      "learning_rate": 7.300613496932516e-06,
      "loss": 0.2744,
      "step": 120
    },
    {
      "epoch": 0.23941068139963168,
      "grad_norm": 0.9446407682756467,
      "learning_rate": 7.914110429447854e-06,
      "loss": 0.2844,
      "step": 130
    },
    {
      "epoch": 0.2578268876611418,
      "grad_norm": 1.0199363774215442,
      "learning_rate": 8.527607361963191e-06,
      "loss": 0.2793,
      "step": 140
    },
    {
      "epoch": 0.27624309392265195,
      "grad_norm": 0.9261191382922327,
      "learning_rate": 9.14110429447853e-06,
      "loss": 0.2822,
      "step": 150
    },
    {
      "epoch": 0.2946593001841621,
      "grad_norm": 0.9028861540282759,
      "learning_rate": 9.754601226993867e-06,
      "loss": 0.2758,
      "step": 160
    },
    {
      "epoch": 0.31307550644567217,
      "grad_norm": 0.8298039066694322,
      "learning_rate": 9.999586697215748e-06,
      "loss": 0.2867,
      "step": 170
    },
    {
      "epoch": 0.3314917127071823,
      "grad_norm": 0.9401043712912456,
      "learning_rate": 9.997061205416203e-06,
      "loss": 0.2935,
      "step": 180
    },
    {
      "epoch": 0.34990791896869244,
      "grad_norm": 0.8170942120374912,
      "learning_rate": 9.992240992810445e-06,
      "loss": 0.276,
      "step": 190
    },
    {
      "epoch": 0.3683241252302026,
      "grad_norm": 0.7970942467849401,
      "learning_rate": 9.985128272907917e-06,
      "loss": 0.2797,
      "step": 200
    },
    {
      "epoch": 0.3867403314917127,
      "grad_norm": 0.8675972760036051,
      "learning_rate": 9.975726311969664e-06,
      "loss": 0.2782,
      "step": 210
    },
    {
      "epoch": 0.40515653775322286,
      "grad_norm": 0.8765243656942432,
      "learning_rate": 9.964039427508418e-06,
      "loss": 0.2815,
      "step": 220
    },
    {
      "epoch": 0.42357274401473294,
      "grad_norm": 0.8649203720855488,
      "learning_rate": 9.950072986305938e-06,
      "loss": 0.2757,
      "step": 230
    },
    {
      "epoch": 0.4419889502762431,
      "grad_norm": 0.8513089510736624,
      "learning_rate": 9.933833401948514e-06,
      "loss": 0.2728,
      "step": 240
    },
    {
      "epoch": 0.4604051565377532,
      "grad_norm": 0.775184166522136,
      "learning_rate": 9.915328131881745e-06,
      "loss": 0.2741,
      "step": 250
    },
    {
      "epoch": 0.47882136279926335,
      "grad_norm": 0.7978646282069273,
      "learning_rate": 9.894565673985986e-06,
      "loss": 0.2725,
      "step": 260
    },
    {
      "epoch": 0.4972375690607735,
      "grad_norm": 0.6613832505637378,
      "learning_rate": 9.871555562673996e-06,
      "loss": 0.2675,
      "step": 270
    },
    {
      "epoch": 0.5156537753222836,
      "grad_norm": 0.8757001574454977,
      "learning_rate": 9.846308364512607e-06,
      "loss": 0.2661,
      "step": 280
    },
    {
      "epoch": 0.5340699815837937,
      "grad_norm": 0.7547104807299371,
      "learning_rate": 9.8188356733704e-06,
      "loss": 0.2742,
      "step": 290
    },
    {
      "epoch": 0.5524861878453039,
      "grad_norm": 0.7251338284529852,
      "learning_rate": 9.789150105093647e-06,
      "loss": 0.2683,
      "step": 300
    },
    {
      "epoch": 0.570902394106814,
      "grad_norm": 0.7543457855276412,
      "learning_rate": 9.75726529171293e-06,
      "loss": 0.2756,
      "step": 310
    },
    {
      "epoch": 0.5893186003683242,
      "grad_norm": 0.7747270052423587,
      "learning_rate": 9.72319587518312e-06,
      "loss": 0.2687,
      "step": 320
    },
    {
      "epoch": 0.6077348066298343,
      "grad_norm": 0.8708864438161477,
      "learning_rate": 9.68695750065959e-06,
      "loss": 0.2687,
      "step": 330
    },
    {
      "epoch": 0.6261510128913443,
      "grad_norm": 0.7181496119078857,
      "learning_rate": 9.648566809313738e-06,
      "loss": 0.2754,
      "step": 340
    },
    {
      "epoch": 0.6445672191528545,
      "grad_norm": 0.7212363207801823,
      "learning_rate": 9.608041430691126e-06,
      "loss": 0.2699,
      "step": 350
    },
    {
      "epoch": 0.6629834254143646,
      "grad_norm": 0.707979023449836,
      "learning_rate": 9.565399974615744e-06,
      "loss": 0.2688,
      "step": 360
    },
    {
      "epoch": 0.6813996316758748,
      "grad_norm": 0.8334790893435787,
      "learning_rate": 9.52066202264412e-06,
      "loss": 0.2776,
      "step": 370
    },
    {
      "epoch": 0.6998158379373849,
      "grad_norm": 0.7596815946675103,
      "learning_rate": 9.473848119073188e-06,
      "loss": 0.2786,
      "step": 380
    },
    {
      "epoch": 0.7182320441988951,
      "grad_norm": 0.7331861291943189,
      "learning_rate": 9.42497976150607e-06,
      "loss": 0.2622,
      "step": 390
    },
    {
      "epoch": 0.7366482504604052,
      "grad_norm": 0.6706665335188781,
      "learning_rate": 9.374079390980058e-06,
      "loss": 0.2695,
      "step": 400
    },
    {
      "epoch": 0.7550644567219152,
      "grad_norm": 0.7515563648690087,
      "learning_rate": 9.321170381661383e-06,
      "loss": 0.2671,
      "step": 410
    },
    {
      "epoch": 0.7734806629834254,
      "grad_norm": 0.6978960493224456,
      "learning_rate": 9.266277030111474e-06,
      "loss": 0.2721,
      "step": 420
    },
    {
      "epoch": 0.7918968692449355,
      "grad_norm": 0.6317044519904149,
      "learning_rate": 9.209424544129621e-06,
      "loss": 0.2611,
      "step": 430
    },
    {
      "epoch": 0.8103130755064457,
      "grad_norm": 0.6933884011578646,
      "learning_rate": 9.150639031177211e-06,
      "loss": 0.2679,
      "step": 440
    },
    {
      "epoch": 0.8287292817679558,
      "grad_norm": 0.6662516396211534,
      "learning_rate": 9.08994748638881e-06,
      "loss": 0.2718,
      "step": 450
    },
    {
      "epoch": 0.8471454880294659,
      "grad_norm": 0.7537377588910371,
      "learning_rate": 9.02737778017562e-06,
      "loss": 0.2625,
      "step": 460
    },
    {
      "epoch": 0.8655616942909761,
      "grad_norm": 0.6490665535396116,
      "learning_rate": 8.962958645426989e-06,
      "loss": 0.2601,
      "step": 470
    },
    {
      "epoch": 0.8839779005524862,
      "grad_norm": 0.6773727676397064,
      "learning_rate": 8.896719664315866e-06,
      "loss": 0.2651,
      "step": 480
    },
    {
      "epoch": 0.9023941068139963,
      "grad_norm": 0.6798276722239568,
      "learning_rate": 8.828691254714259e-06,
      "loss": 0.2625,
      "step": 490
    },
    {
      "epoch": 0.9208103130755064,
      "grad_norm": 0.6979929906196801,
      "learning_rate": 8.758904656224904e-06,
      "loss": 0.2749,
      "step": 500
    },
    {
      "epoch": 0.9392265193370166,
      "grad_norm": 0.7308459913669413,
      "learning_rate": 8.687391915835617e-06,
      "loss": 0.2666,
      "step": 510
    },
    {
      "epoch": 0.9576427255985267,
      "grad_norm": 0.8055261702848038,
      "learning_rate": 8.614185873202852e-06,
      "loss": 0.2705,
      "step": 520
    },
    {
      "epoch": 0.9760589318600368,
      "grad_norm": 0.7313961658432152,
      "learning_rate": 8.539320145571277e-06,
      "loss": 0.2619,
      "step": 530
    },
    {
      "epoch": 0.994475138121547,
      "grad_norm": 0.6331517103040231,
      "learning_rate": 8.462829112336266e-06,
      "loss": 0.2671,
      "step": 540
    },
    {
      "epoch": 1.0128913443830572,
      "grad_norm": 0.7085468234966358,
      "learning_rate": 8.384747899256386e-06,
      "loss": 0.2388,
      "step": 550
    },
    {
      "epoch": 1.0313075506445673,
      "grad_norm": 0.6748955679522807,
      "learning_rate": 8.30511236232316e-06,
      "loss": 0.2226,
      "step": 560
    },
    {
      "epoch": 1.0497237569060773,
      "grad_norm": 0.7100102091480333,
      "learning_rate": 8.223959071295492e-06,
      "loss": 0.2297,
      "step": 570
    },
    {
      "epoch": 1.0681399631675874,
      "grad_norm": 0.6668769883519982,
      "learning_rate": 8.141325292906325e-06,
      "loss": 0.2263,
      "step": 580
    },
    {
      "epoch": 1.0865561694290977,
      "grad_norm": 0.6720758025127164,
      "learning_rate": 8.057248973749216e-06,
      "loss": 0.2288,
      "step": 590
    },
    {
      "epoch": 1.1049723756906078,
      "grad_norm": 0.6411173465966967,
      "learning_rate": 7.971768722852741e-06,
      "loss": 0.2253,
      "step": 600
    },
    {
      "epoch": 1.1233885819521179,
      "grad_norm": 0.6724293241376955,
      "learning_rate": 7.884923793950684e-06,
      "loss": 0.2326,
      "step": 610
    },
    {
      "epoch": 1.141804788213628,
      "grad_norm": 0.6371467332633137,
      "learning_rate": 7.796754067456168e-06,
      "loss": 0.2235,
      "step": 620
    },
    {
      "epoch": 1.160220994475138,
      "grad_norm": 0.6085432732125634,
      "learning_rate": 7.707300032148004e-06,
      "loss": 0.2289,
      "step": 630
    },
    {
      "epoch": 1.1786372007366483,
      "grad_norm": 0.7730530541898509,
      "learning_rate": 7.616602766577683e-06,
      "loss": 0.2263,
      "step": 640
    },
    {
      "epoch": 1.1970534069981584,
      "grad_norm": 0.7420431722106332,
      "learning_rate": 7.524703920205521e-06,
      "loss": 0.2331,
      "step": 650
    },
    {
      "epoch": 1.2154696132596685,
      "grad_norm": 0.6604167477110746,
      "learning_rate": 7.43164569427464e-06,
      "loss": 0.2278,
      "step": 660
    },
    {
      "epoch": 1.2338858195211786,
      "grad_norm": 0.7778434196490822,
      "learning_rate": 7.3374708224315725e-06,
      "loss": 0.2242,
      "step": 670
    },
    {
      "epoch": 1.2523020257826887,
      "grad_norm": 0.7114711250180328,
      "learning_rate": 7.2422225511023555e-06,
      "loss": 0.2261,
      "step": 680
    },
    {
      "epoch": 1.270718232044199,
      "grad_norm": 0.7059769736094257,
      "learning_rate": 7.145944619633176e-06,
      "loss": 0.2218,
      "step": 690
    },
    {
      "epoch": 1.289134438305709,
      "grad_norm": 0.6846868707521909,
      "learning_rate": 7.048681240204641e-06,
      "loss": 0.2305,
      "step": 700
    },
    {
      "epoch": 1.3075506445672191,
      "grad_norm": 0.636969343049607,
      "learning_rate": 6.950477077528927e-06,
      "loss": 0.2242,
      "step": 710
    },
    {
      "epoch": 1.3259668508287292,
      "grad_norm": 0.633812292072214,
      "learning_rate": 6.851377228339106e-06,
      "loss": 0.2302,
      "step": 720
    },
    {
      "epoch": 1.3443830570902393,
      "grad_norm": 0.6630293802277623,
      "learning_rate": 6.751427200680109e-06,
      "loss": 0.2295,
      "step": 730
    },
    {
      "epoch": 1.3627992633517496,
      "grad_norm": 0.6315090136897826,
      "learning_rate": 6.650672893010769e-06,
      "loss": 0.2221,
      "step": 740
    },
    {
      "epoch": 1.3812154696132597,
      "grad_norm": 0.7126623318254198,
      "learning_rate": 6.549160573126623e-06,
      "loss": 0.2293,
      "step": 750
    },
    {
      "epoch": 1.3996316758747698,
      "grad_norm": 0.6653513173488009,
      "learning_rate": 6.4469368569130786e-06,
      "loss": 0.2282,
      "step": 760
    },
    {
      "epoch": 1.4180478821362799,
      "grad_norm": 0.6343921369098496,
      "learning_rate": 6.344048686938745e-06,
      "loss": 0.2273,
      "step": 770
    },
    {
      "epoch": 1.43646408839779,
      "grad_norm": 0.6210167648750854,
      "learning_rate": 6.2405433108987456e-06,
      "loss": 0.2338,
      "step": 780
    },
    {
      "epoch": 1.4548802946593002,
      "grad_norm": 0.6784962640818855,
      "learning_rate": 6.136468259917917e-06,
      "loss": 0.2289,
      "step": 790
    },
    {
      "epoch": 1.4732965009208103,
      "grad_norm": 0.578659502472278,
      "learning_rate": 6.031871326723837e-06,
      "loss": 0.2363,
      "step": 800
    },
    {
      "epoch": 1.4917127071823204,
      "grad_norm": 0.7040713369273666,
      "learning_rate": 5.92680054369974e-06,
      "loss": 0.2233,
      "step": 810
    },
    {
      "epoch": 1.5101289134438307,
      "grad_norm": 0.7100429365248154,
      "learning_rate": 5.821304160827371e-06,
      "loss": 0.2275,
      "step": 820
    },
    {
      "epoch": 1.5285451197053406,
      "grad_norm": 0.5875774880678092,
      "learning_rate": 5.71543062352991e-06,
      "loss": 0.225,
      "step": 830
    },
    {
      "epoch": 1.5469613259668509,
      "grad_norm": 0.5895563949605374,
      "learning_rate": 5.609228550425154e-06,
      "loss": 0.2347,
      "step": 840
    },
    {
      "epoch": 1.565377532228361,
      "grad_norm": 0.6923142403050372,
      "learning_rate": 5.50274671099917e-06,
      "loss": 0.2212,
      "step": 850
    },
    {
      "epoch": 1.583793738489871,
      "grad_norm": 0.6402043458583734,
      "learning_rate": 5.3960340032106515e-06,
      "loss": 0.221,
      "step": 860
    },
    {
      "epoch": 1.6022099447513813,
      "grad_norm": 0.6583270982368973,
      "learning_rate": 5.28913943103629e-06,
      "loss": 0.2263,
      "step": 870
    },
    {
      "epoch": 1.6206261510128912,
      "grad_norm": 0.6545061377359965,
      "learning_rate": 5.182112081967467e-06,
      "loss": 0.2221,
      "step": 880
    },
    {
      "epoch": 1.6390423572744015,
      "grad_norm": 0.7057814110328757,
      "learning_rate": 5.075001104468576e-06,
      "loss": 0.2303,
      "step": 890
    },
    {
      "epoch": 1.6574585635359116,
      "grad_norm": 0.6738126191275001,
      "learning_rate": 4.967855685407368e-06,
      "loss": 0.2235,
      "step": 900
    },
    {
      "epoch": 1.6758747697974217,
      "grad_norm": 0.6229975201509437,
      "learning_rate": 4.860725027467641e-06,
      "loss": 0.2276,
      "step": 910
    },
    {
      "epoch": 1.694290976058932,
      "grad_norm": 0.6485619857748801,
      "learning_rate": 4.7536583265546775e-06,
      "loss": 0.2294,
      "step": 920
    },
    {
      "epoch": 1.7127071823204418,
      "grad_norm": 0.6952097611858238,
      "learning_rate": 4.646704749203794e-06,
      "loss": 0.22,
      "step": 930
    },
    {
      "epoch": 1.7311233885819521,
      "grad_norm": 0.6993773129475868,
      "learning_rate": 4.539913410002378e-06,
      "loss": 0.2253,
      "step": 940
    },
    {
      "epoch": 1.7495395948434622,
      "grad_norm": 0.7179073268017699,
      "learning_rate": 4.433333349035773e-06,
      "loss": 0.2262,
      "step": 950
    },
    {
      "epoch": 1.7679558011049723,
      "grad_norm": 0.6202631830964883,
      "learning_rate": 4.327013509367386e-06,
      "loss": 0.2212,
      "step": 960
    },
    {
      "epoch": 1.7863720073664826,
      "grad_norm": 0.6608879257543027,
      "learning_rate": 4.221002714563347e-06,
      "loss": 0.2205,
      "step": 970
    },
    {
      "epoch": 1.8047882136279927,
      "grad_norm": 0.59224317673099,
      "learning_rate": 4.115349646272029e-06,
      "loss": 0.2205,
      "step": 980
    },
    {
      "epoch": 1.8232044198895028,
      "grad_norm": 0.5878582444161091,
      "learning_rate": 4.010102821868762e-06,
      "loss": 0.2229,
      "step": 990
    },
    {
      "epoch": 1.8416206261510129,
      "grad_norm": 0.6461746282587334,
      "learning_rate": 3.90531057217597e-06,
      "loss": 0.2232,
      "step": 1000
    },
    {
      "epoch": 1.860036832412523,
      "grad_norm": 0.6205082651968326,
      "learning_rate": 3.8010210192689688e-06,
      "loss": 0.2277,
      "step": 1010
    },
    {
      "epoch": 1.8784530386740332,
      "grad_norm": 0.7181661264692984,
      "learning_rate": 3.6972820543776404e-06,
      "loss": 0.2261,
      "step": 1020
    },
    {
      "epoch": 1.8968692449355433,
      "grad_norm": 0.6649391919825204,
      "learning_rate": 3.5941413158941086e-06,
      "loss": 0.2259,
      "step": 1030
    },
    {
      "epoch": 1.9152854511970534,
      "grad_norm": 0.6044692798894635,
      "learning_rate": 3.4916461674965074e-06,
      "loss": 0.222,
      "step": 1040
    },
    {
      "epoch": 1.9337016574585635,
      "grad_norm": 0.615450434298784,
      "learning_rate": 3.389843676398925e-06,
      "loss": 0.2254,
      "step": 1050
    },
    {
      "epoch": 1.9521178637200736,
      "grad_norm": 0.6691163050124782,
      "learning_rate": 3.2887805917374736e-06,
      "loss": 0.2189,
      "step": 1060
    },
    {
      "epoch": 1.9705340699815839,
      "grad_norm": 0.6022414370557024,
      "learning_rate": 3.1885033231024253e-06,
      "loss": 0.2227,
      "step": 1070
    },
    {
      "epoch": 1.988950276243094,
      "grad_norm": 0.6942520950174368,
      "learning_rate": 3.089057919226277e-06,
      "loss": 0.2254,
      "step": 1080
    },
    {
      "epoch": 2.007366482504604,
      "grad_norm": 0.5986817795885931,
      "learning_rate": 2.9904900468375298e-06,
      "loss": 0.2091,
      "step": 1090
    },
    {
      "epoch": 2.0257826887661143,
      "grad_norm": 0.5655950429797426,
      "learning_rate": 2.892844969689876e-06,
      "loss": 0.184,
      "step": 1100
    },
    {
      "epoch": 2.044198895027624,
      "grad_norm": 0.6492187365821179,
      "learning_rate": 2.7961675277764498e-06,
      "loss": 0.1815,
      "step": 1110
    },
    {
      "epoch": 2.0626151012891345,
      "grad_norm": 0.5494541259912535,
      "learning_rate": 2.7005021167386804e-06,
      "loss": 0.1782,
      "step": 1120
    },
    {
      "epoch": 2.0810313075506444,
      "grad_norm": 0.6018040555497874,
      "learning_rate": 2.605892667479173e-06,
      "loss": 0.1806,
      "step": 1130
    },
    {
      "epoch": 2.0994475138121547,
      "grad_norm": 0.609141353634607,
      "learning_rate": 2.5123826259880324e-06,
      "loss": 0.1827,
      "step": 1140
    },
    {
      "epoch": 2.117863720073665,
      "grad_norm": 0.6522701323065019,
      "learning_rate": 2.420014933391849e-06,
      "loss": 0.1834,
      "step": 1150
    },
    {
      "epoch": 2.136279926335175,
      "grad_norm": 0.6209038557527796,
      "learning_rate": 2.3288320062345276e-06,
      "loss": 0.1832,
      "step": 1160
    },
    {
      "epoch": 2.154696132596685,
      "grad_norm": 0.671734526610132,
      "learning_rate": 2.238875716999019e-06,
      "loss": 0.181,
      "step": 1170
    },
    {
      "epoch": 2.1731123388581954,
      "grad_norm": 0.6211775861295186,
      "learning_rate": 2.1501873748788804e-06,
      "loss": 0.1859,
      "step": 1180
    },
    {
      "epoch": 2.1915285451197053,
      "grad_norm": 0.6161892750894669,
      "learning_rate": 2.0628077068085173e-06,
      "loss": 0.1833,
      "step": 1190
    },
    {
      "epoch": 2.2099447513812156,
      "grad_norm": 0.6171060565609299,
      "learning_rate": 1.976776838760801e-06,
      "loss": 0.1823,
      "step": 1200
    },
    {
      "epoch": 2.2283609576427255,
      "grad_norm": 0.6627457414621876,
      "learning_rate": 1.8921342773206553e-06,
      "loss": 0.1802,
      "step": 1210
    },
    {
      "epoch": 2.2467771639042358,
      "grad_norm": 0.5766657805895978,
      "learning_rate": 1.8089188915430794e-06,
      "loss": 0.1801,
      "step": 1220
    },
    {
      "epoch": 2.265193370165746,
      "grad_norm": 0.6528068390986206,
      "learning_rate": 1.7271688951039312e-06,
      "loss": 0.1817,
      "step": 1230
    },
    {
      "epoch": 2.283609576427256,
      "grad_norm": 0.6108145113882844,
      "learning_rate": 1.6469218287516664e-06,
      "loss": 0.1824,
      "step": 1240
    },
    {
      "epoch": 2.3020257826887662,
      "grad_norm": 0.6223360388944672,
      "learning_rate": 1.568214543068103e-06,
      "loss": 0.1817,
      "step": 1250
    },
    {
      "epoch": 2.320441988950276,
      "grad_norm": 0.5730293037393086,
      "learning_rate": 1.4910831815461125e-06,
      "loss": 0.1783,
      "step": 1260
    },
    {
      "epoch": 2.3388581952117864,
      "grad_norm": 0.6804328273268474,
      "learning_rate": 1.4155631639920208e-06,
      "loss": 0.1793,
      "step": 1270
    },
    {
      "epoch": 2.3572744014732967,
      "grad_norm": 0.5936583186703833,
      "learning_rate": 1.3416891702603357e-06,
      "loss": 0.1772,
      "step": 1280
    },
    {
      "epoch": 2.3756906077348066,
      "grad_norm": 0.6646979614560136,
      "learning_rate": 1.2694951243282682e-06,
      "loss": 0.1805,
      "step": 1290
    },
    {
      "epoch": 2.394106813996317,
      "grad_norm": 0.5867258571902689,
      "learning_rate": 1.1990141787173648e-06,
      "loss": 0.1793,
      "step": 1300
    },
    {
      "epoch": 2.4125230202578267,
      "grad_norm": 0.6356400483556068,
      "learning_rate": 1.1302786992694049e-06,
      "loss": 0.1793,
      "step": 1310
    },
    {
      "epoch": 2.430939226519337,
      "grad_norm": 0.6071714234787918,
      "learning_rate": 1.0633202502835494e-06,
      "loss": 0.1831,
      "step": 1320
    },
    {
      "epoch": 2.4493554327808473,
      "grad_norm": 0.6031318900784148,
      "learning_rate": 9.981695800215701e-07,
      "loss": 0.1795,
      "step": 1330
    },
    {
      "epoch": 2.467771639042357,
      "grad_norm": 0.6195067631818659,
      "learning_rate": 9.348566065878218e-07,
      "loss": 0.1807,
      "step": 1340
    },
    {
      "epoch": 2.4861878453038675,
      "grad_norm": 0.640824325612606,
      "learning_rate": 8.734104041904129e-07,
      "loss": 0.1835,
      "step": 1350
    },
    {
      "epoch": 2.5046040515653774,
      "grad_norm": 0.5801881807297086,
      "learning_rate": 8.138591897899345e-07,
      "loss": 0.1792,
      "step": 1360
    },
    {
      "epoch": 2.5230202578268877,
      "grad_norm": 0.6192805732625175,
      "learning_rate": 7.56230310141835e-07,
      "loss": 0.179,
      "step": 1370
    },
    {
      "epoch": 2.541436464088398,
      "grad_norm": 0.6016304803765787,
      "learning_rate": 7.005502292383898e-07,
      "loss": 0.1799,
      "step": 1380
    },
    {
      "epoch": 2.559852670349908,
      "grad_norm": 0.6115554388128484,
      "learning_rate": 6.46844516156081e-07,
      "loss": 0.1758,
      "step": 1390
    },
    {
      "epoch": 2.578268876611418,
      "grad_norm": 0.5927866192371272,
      "learning_rate": 5.951378333139118e-07,
      "loss": 0.1796,
      "step": 1400
    },
    {
      "epoch": 2.596685082872928,
      "grad_norm": 0.6811377078596266,
      "learning_rate": 5.454539251480739e-07,
      "loss": 0.1808,
      "step": 1410
    },
    {
      "epoch": 2.6151012891344383,
      "grad_norm": 0.6077558733405245,
      "learning_rate": 4.978156072081669e-07,
      "loss": 0.1784,
      "step": 1420
    },
    {
      "epoch": 2.6335174953959486,
      "grad_norm": 0.5643531059218968,
      "learning_rate": 4.522447556799875e-07,
      "loss": 0.1795,
      "step": 1430
    },
    {
      "epoch": 2.6519337016574585,
      "grad_norm": 0.5619239404266387,
      "learning_rate": 4.0876229733966655e-07,
      "loss": 0.1759,
      "step": 1440
    },
    {
      "epoch": 2.6703499079189688,
      "grad_norm": 0.6328209657239623,
      "learning_rate": 3.6738819994379946e-07,
      "loss": 0.1808,
      "step": 1450
    },
    {
      "epoch": 2.6887661141804786,
      "grad_norm": 0.6086714396313364,
      "learning_rate": 3.281414630599811e-07,
      "loss": 0.1772,
      "step": 1460
    },
    {
      "epoch": 2.707182320441989,
      "grad_norm": 0.6266202928122402,
      "learning_rate": 2.9104010934192795e-07,
      "loss": 0.1798,
      "step": 1470
    },
    {
      "epoch": 2.7255985267034992,
      "grad_norm": 0.61076576372029,
      "learning_rate": 2.561011762532212e-07,
      "loss": 0.1749,
      "step": 1480
    },
    {
      "epoch": 2.744014732965009,
      "grad_norm": 0.6615569148493705,
      "learning_rate": 2.2334070824347243e-07,
      "loss": 0.1805,
      "step": 1490
    },
    {
      "epoch": 2.7624309392265194,
      "grad_norm": 0.6278844879654595,
      "learning_rate": 1.9277374938047989e-07,
      "loss": 0.1808,
      "step": 1500
    },
    {
      "epoch": 2.7808471454880292,
      "grad_norm": 0.57717673931454,
      "learning_rate": 1.644143364417794e-07,
      "loss": 0.1775,
      "step": 1510
    },
    {
      "epoch": 2.7992633517495396,
      "grad_norm": 0.5988782652066212,
      "learning_rate": 1.3827549246876627e-07,
      "loss": 0.1841,
      "step": 1520
    },
    {
      "epoch": 2.81767955801105,
      "grad_norm": 0.6220611373904551,
      "learning_rate": 1.1436922078632395e-07,
      "loss": 0.1773,
      "step": 1530
    },
    {
      "epoch": 2.8360957642725597,
      "grad_norm": 0.6549658009335396,
      "learning_rate": 9.270649949073229e-08,
      "loss": 0.1796,
      "step": 1540
    },
    {
      "epoch": 2.85451197053407,
      "grad_norm": 0.5897064017884592,
      "learning_rate": 7.329727640837059e-08,
      "loss": 0.1781,
      "step": 1550
    },
    {
      "epoch": 2.87292817679558,
      "grad_norm": 0.6006548076063878,
      "learning_rate": 5.615046452753403e-08,
      "loss": 0.1789,
      "step": 1560
    },
    {
      "epoch": 2.89134438305709,
      "grad_norm": 0.5940577802255843,
      "learning_rate": 4.127393790546719e-08,
      "loss": 0.1801,
      "step": 1570
    },
    {
      "epoch": 2.9097605893186005,
      "grad_norm": 0.6226132514184528,
      "learning_rate": 2.8674528052484162e-08,
      "loss": 0.1774,
      "step": 1580
    },
    {
      "epoch": 2.9281767955801103,
      "grad_norm": 0.6274949409362496,
      "learning_rate": 1.8358020794843056e-08,
      "loss": 0.1827,
      "step": 1590
    },
    {
      "epoch": 2.9465930018416207,
      "grad_norm": 0.6038116585141188,
      "learning_rate": 1.0329153617812948e-08,
      "loss": 0.1781,
      "step": 1600
    },
    {
      "epoch": 2.9650092081031305,
      "grad_norm": 0.5802100431657281,
      "learning_rate": 4.5916134901552445e-09,
      "loss": 0.1825,
      "step": 1610
    },
    {
      "epoch": 2.983425414364641,
      "grad_norm": 0.6005447860892711,
      "learning_rate": 1.148035171014139e-09,
      "loss": 0.1839,
      "step": 1620
    },
    {
      "epoch": 3.0,
      "step": 1629,
      "total_flos": 2800162131935232.0,
      "train_loss": 0.22954999712656582,
      "train_runtime": 49788.0682,
      "train_samples_per_second": 2.094,
      "train_steps_per_second": 0.033
    }
  ],
  "logging_steps": 10,
  "max_steps": 1629,
  "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": 2800162131935232.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}