{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 2391,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.012547051442910916,
      "grad_norm": 0.716044008731842,
      "learning_rate": 1.9924717691342537e-05,
      "loss": 1.0972,
      "step": 10
    },
    {
      "epoch": 0.025094102885821833,
      "grad_norm": 1.4613193273544312,
      "learning_rate": 1.9841070681723128e-05,
      "loss": 1.088,
      "step": 20
    },
    {
      "epoch": 0.037641154328732745,
      "grad_norm": 0.627024233341217,
      "learning_rate": 1.9757423672103725e-05,
      "loss": 1.0716,
      "step": 30
    },
    {
      "epoch": 0.050188205771643665,
      "grad_norm": 1.224279522895813,
      "learning_rate": 1.9673776662484318e-05,
      "loss": 1.0661,
      "step": 40
    },
    {
      "epoch": 0.06273525721455459,
      "grad_norm": 1.368181586265564,
      "learning_rate": 1.959012965286491e-05,
      "loss": 1.035,
      "step": 50
    },
    {
      "epoch": 0.07528230865746549,
      "grad_norm": 1.0884851217269897,
      "learning_rate": 1.9506482643245505e-05,
      "loss": 1.0288,
      "step": 60
    },
    {
      "epoch": 0.08782936010037641,
      "grad_norm": 1.9844820499420166,
      "learning_rate": 1.9422835633626102e-05,
      "loss": 0.9855,
      "step": 70
    },
    {
      "epoch": 0.10037641154328733,
      "grad_norm": 1.6572856903076172,
      "learning_rate": 1.9339188624006692e-05,
      "loss": 0.991,
      "step": 80
    },
    {
      "epoch": 0.11292346298619825,
      "grad_norm": 2.2602856159210205,
      "learning_rate": 1.925554161438729e-05,
      "loss": 1.0092,
      "step": 90
    },
    {
      "epoch": 0.12547051442910917,
      "grad_norm": 4.111606597900391,
      "learning_rate": 1.9171894604767882e-05,
      "loss": 0.9062,
      "step": 100
    },
    {
      "epoch": 0.13801756587202008,
      "grad_norm": 3.383659601211548,
      "learning_rate": 1.9088247595148476e-05,
      "loss": 0.9749,
      "step": 110
    },
    {
      "epoch": 0.15056461731493098,
      "grad_norm": 1.9777194261550903,
      "learning_rate": 1.900460058552907e-05,
      "loss": 0.9193,
      "step": 120
    },
    {
      "epoch": 0.16311166875784192,
      "grad_norm": 6.869212627410889,
      "learning_rate": 1.8920953575909663e-05,
      "loss": 0.9695,
      "step": 130
    },
    {
      "epoch": 0.17565872020075282,
      "grad_norm": 2.800828695297241,
      "learning_rate": 1.8837306566290257e-05,
      "loss": 0.9375,
      "step": 140
    },
    {
      "epoch": 0.18820577164366373,
      "grad_norm": 3.783500909805298,
      "learning_rate": 1.875365955667085e-05,
      "loss": 0.9294,
      "step": 150
    },
    {
      "epoch": 0.20075282308657466,
      "grad_norm": 2.961916208267212,
      "learning_rate": 1.8670012547051444e-05,
      "loss": 0.9651,
      "step": 160
    },
    {
      "epoch": 0.21329987452948557,
      "grad_norm": 4.599372386932373,
      "learning_rate": 1.8586365537432037e-05,
      "loss": 0.9706,
      "step": 170
    },
    {
      "epoch": 0.2258469259723965,
      "grad_norm": 3.3036141395568848,
      "learning_rate": 1.8502718527812634e-05,
      "loss": 0.9289,
      "step": 180
    },
    {
      "epoch": 0.2383939774153074,
      "grad_norm": 6.022705078125,
      "learning_rate": 1.8419071518193224e-05,
      "loss": 0.9015,
      "step": 190
    },
    {
      "epoch": 0.25094102885821834,
      "grad_norm": 13.482539176940918,
      "learning_rate": 1.833542450857382e-05,
      "loss": 0.9033,
      "step": 200
    },
    {
      "epoch": 0.26348808030112925,
      "grad_norm": 4.047461032867432,
      "learning_rate": 1.8251777498954414e-05,
      "loss": 0.8447,
      "step": 210
    },
    {
      "epoch": 0.27603513174404015,
      "grad_norm": 2.8806040287017822,
      "learning_rate": 1.8168130489335008e-05,
      "loss": 0.8333,
      "step": 220
    },
    {
      "epoch": 0.28858218318695106,
      "grad_norm": 5.517472743988037,
      "learning_rate": 1.80844834797156e-05,
      "loss": 0.8697,
      "step": 230
    },
    {
      "epoch": 0.30112923462986196,
      "grad_norm": 7.646421432495117,
      "learning_rate": 1.8000836470096195e-05,
      "loss": 0.9159,
      "step": 240
    },
    {
      "epoch": 0.3136762860727729,
      "grad_norm": 4.211366176605225,
      "learning_rate": 1.791718946047679e-05,
      "loss": 0.8848,
      "step": 250
    },
    {
      "epoch": 0.32622333751568383,
      "grad_norm": 7.600694179534912,
      "learning_rate": 1.7833542450857385e-05,
      "loss": 0.8085,
      "step": 260
    },
    {
      "epoch": 0.33877038895859474,
      "grad_norm": 4.872680187225342,
      "learning_rate": 1.7749895441237976e-05,
      "loss": 0.8301,
      "step": 270
    },
    {
      "epoch": 0.35131744040150564,
      "grad_norm": 6.380251884460449,
      "learning_rate": 1.7666248431618572e-05,
      "loss": 0.9484,
      "step": 280
    },
    {
      "epoch": 0.36386449184441655,
      "grad_norm": 4.365920066833496,
      "learning_rate": 1.7582601421999166e-05,
      "loss": 0.9457,
      "step": 290
    },
    {
      "epoch": 0.37641154328732745,
      "grad_norm": 3.710908889770508,
      "learning_rate": 1.749895441237976e-05,
      "loss": 0.8773,
      "step": 300
    },
    {
      "epoch": 0.3889585947302384,
      "grad_norm": 4.279157638549805,
      "learning_rate": 1.7415307402760353e-05,
      "loss": 0.8892,
      "step": 310
    },
    {
      "epoch": 0.4015056461731493,
      "grad_norm": 17.142894744873047,
      "learning_rate": 1.7331660393140947e-05,
      "loss": 0.8723,
      "step": 320
    },
    {
      "epoch": 0.41405269761606023,
      "grad_norm": 3.349973440170288,
      "learning_rate": 1.724801338352154e-05,
      "loss": 0.8092,
      "step": 330
    },
    {
      "epoch": 0.42659974905897113,
      "grad_norm": 7.343029022216797,
      "learning_rate": 1.7164366373902134e-05,
      "loss": 0.9051,
      "step": 340
    },
    {
      "epoch": 0.43914680050188204,
      "grad_norm": 3.02512264251709,
      "learning_rate": 1.7080719364282727e-05,
      "loss": 0.8959,
      "step": 350
    },
    {
      "epoch": 0.451693851944793,
      "grad_norm": 13.072802543640137,
      "learning_rate": 1.699707235466332e-05,
      "loss": 0.9339,
      "step": 360
    },
    {
      "epoch": 0.4642409033877039,
      "grad_norm": 3.5314481258392334,
      "learning_rate": 1.6913425345043917e-05,
      "loss": 0.8883,
      "step": 370
    },
    {
      "epoch": 0.4767879548306148,
      "grad_norm": 3.5592713356018066,
      "learning_rate": 1.6829778335424508e-05,
      "loss": 0.9539,
      "step": 380
    },
    {
      "epoch": 0.4893350062735257,
      "grad_norm": 2.2764039039611816,
      "learning_rate": 1.6746131325805104e-05,
      "loss": 0.9251,
      "step": 390
    },
    {
      "epoch": 0.5018820577164367,
      "grad_norm": 2.953474998474121,
      "learning_rate": 1.6662484316185698e-05,
      "loss": 0.9561,
      "step": 400
    },
    {
      "epoch": 0.5144291091593476,
      "grad_norm": 2.250779628753662,
      "learning_rate": 1.657883730656629e-05,
      "loss": 0.9405,
      "step": 410
    },
    {
      "epoch": 0.5269761606022585,
      "grad_norm": 5.86421537399292,
      "learning_rate": 1.6495190296946885e-05,
      "loss": 0.8936,
      "step": 420
    },
    {
      "epoch": 0.5395232120451694,
      "grad_norm": 4.82138729095459,
      "learning_rate": 1.6411543287327482e-05,
      "loss": 0.8553,
      "step": 430
    },
    {
      "epoch": 0.5520702634880803,
      "grad_norm": 3.090364456176758,
      "learning_rate": 1.6327896277708072e-05,
      "loss": 0.9059,
      "step": 440
    },
    {
      "epoch": 0.5646173149309912,
      "grad_norm": 5.363993167877197,
      "learning_rate": 1.624424926808867e-05,
      "loss": 0.9631,
      "step": 450
    },
    {
      "epoch": 0.5771643663739021,
      "grad_norm": 3.4371540546417236,
      "learning_rate": 1.616060225846926e-05,
      "loss": 0.794,
      "step": 460
    },
    {
      "epoch": 0.589711417816813,
      "grad_norm": 4.080657482147217,
      "learning_rate": 1.6076955248849856e-05,
      "loss": 0.9348,
      "step": 470
    },
    {
      "epoch": 0.6022584692597239,
      "grad_norm": 5.351539134979248,
      "learning_rate": 1.599330823923045e-05,
      "loss": 0.9286,
      "step": 480
    },
    {
      "epoch": 0.6148055207026348,
      "grad_norm": 3.6495888233184814,
      "learning_rate": 1.5909661229611043e-05,
      "loss": 0.7525,
      "step": 490
    },
    {
      "epoch": 0.6273525721455459,
      "grad_norm": 4.849205017089844,
      "learning_rate": 1.5826014219991636e-05,
      "loss": 0.9534,
      "step": 500
    },
    {
      "epoch": 0.6398996235884568,
      "grad_norm": 4.923251152038574,
      "learning_rate": 1.574236721037223e-05,
      "loss": 0.9681,
      "step": 510
    },
    {
      "epoch": 0.6524466750313677,
      "grad_norm": 4.8032073974609375,
      "learning_rate": 1.5658720200752824e-05,
      "loss": 0.8674,
      "step": 520
    },
    {
      "epoch": 0.6649937264742786,
      "grad_norm": 6.583972454071045,
      "learning_rate": 1.5575073191133417e-05,
      "loss": 0.8463,
      "step": 530
    },
    {
      "epoch": 0.6775407779171895,
      "grad_norm": 6.276474475860596,
      "learning_rate": 1.5491426181514014e-05,
      "loss": 0.8481,
      "step": 540
    },
    {
      "epoch": 0.6900878293601004,
      "grad_norm": 6.980823040008545,
      "learning_rate": 1.5407779171894604e-05,
      "loss": 0.8441,
      "step": 550
    },
    {
      "epoch": 0.7026348808030113,
      "grad_norm": 5.919472694396973,
      "learning_rate": 1.53241321622752e-05,
      "loss": 0.7705,
      "step": 560
    },
    {
      "epoch": 0.7151819322459222,
      "grad_norm": 14.738715171813965,
      "learning_rate": 1.5240485152655793e-05,
      "loss": 0.8883,
      "step": 570
    },
    {
      "epoch": 0.7277289836888331,
      "grad_norm": 9.944400787353516,
      "learning_rate": 1.5156838143036388e-05,
      "loss": 0.843,
      "step": 580
    },
    {
      "epoch": 0.740276035131744,
      "grad_norm": 5.619649410247803,
      "learning_rate": 1.5073191133416981e-05,
      "loss": 0.9996,
      "step": 590
    },
    {
      "epoch": 0.7528230865746549,
      "grad_norm": 7.122910499572754,
      "learning_rate": 1.4989544123797575e-05,
      "loss": 0.9,
      "step": 600
    },
    {
      "epoch": 0.7653701380175659,
      "grad_norm": 10.766443252563477,
      "learning_rate": 1.490589711417817e-05,
      "loss": 0.8326,
      "step": 610
    },
    {
      "epoch": 0.7779171894604768,
      "grad_norm": 4.103264808654785,
      "learning_rate": 1.4822250104558764e-05,
      "loss": 0.8552,
      "step": 620
    },
    {
      "epoch": 0.7904642409033877,
      "grad_norm": 11.64675521850586,
      "learning_rate": 1.4738603094939357e-05,
      "loss": 0.774,
      "step": 630
    },
    {
      "epoch": 0.8030112923462986,
      "grad_norm": 6.035724639892578,
      "learning_rate": 1.465495608531995e-05,
      "loss": 0.9782,
      "step": 640
    },
    {
      "epoch": 0.8155583437892095,
      "grad_norm": 6.475778102874756,
      "learning_rate": 1.4571309075700546e-05,
      "loss": 0.7997,
      "step": 650
    },
    {
      "epoch": 0.8281053952321205,
      "grad_norm": 5.5077314376831055,
      "learning_rate": 1.4487662066081138e-05,
      "loss": 0.7985,
      "step": 660
    },
    {
      "epoch": 0.8406524466750314,
      "grad_norm": 9.788359642028809,
      "learning_rate": 1.4404015056461733e-05,
      "loss": 0.9237,
      "step": 670
    },
    {
      "epoch": 0.8531994981179423,
      "grad_norm": 5.642941951751709,
      "learning_rate": 1.4320368046842325e-05,
      "loss": 0.7795,
      "step": 680
    },
    {
      "epoch": 0.8657465495608532,
      "grad_norm": 7.9442315101623535,
      "learning_rate": 1.423672103722292e-05,
      "loss": 0.891,
      "step": 690
    },
    {
      "epoch": 0.8782936010037641,
      "grad_norm": 4.023198127746582,
      "learning_rate": 1.4153074027603515e-05,
      "loss": 0.8391,
      "step": 700
    },
    {
      "epoch": 0.890840652446675,
      "grad_norm": 11.08212661743164,
      "learning_rate": 1.4069427017984107e-05,
      "loss": 0.9049,
      "step": 710
    },
    {
      "epoch": 0.903387703889586,
      "grad_norm": 7.001065254211426,
      "learning_rate": 1.3985780008364702e-05,
      "loss": 0.9636,
      "step": 720
    },
    {
      "epoch": 0.9159347553324969,
      "grad_norm": 7.488030433654785,
      "learning_rate": 1.3902132998745297e-05,
      "loss": 0.8491,
      "step": 730
    },
    {
      "epoch": 0.9284818067754078,
      "grad_norm": 5.319072246551514,
      "learning_rate": 1.381848598912589e-05,
      "loss": 0.921,
      "step": 740
    },
    {
      "epoch": 0.9410288582183187,
      "grad_norm": 7.3197126388549805,
      "learning_rate": 1.3734838979506484e-05,
      "loss": 0.9866,
      "step": 750
    },
    {
      "epoch": 0.9535759096612296,
      "grad_norm": 5.280470848083496,
      "learning_rate": 1.3651191969887078e-05,
      "loss": 0.8146,
      "step": 760
    },
    {
      "epoch": 0.9661229611041405,
      "grad_norm": 3.304518938064575,
      "learning_rate": 1.3567544960267671e-05,
      "loss": 0.8724,
      "step": 770
    },
    {
      "epoch": 0.9786700125470514,
      "grad_norm": 3.306471109390259,
      "learning_rate": 1.3483897950648267e-05,
      "loss": 0.8796,
      "step": 780
    },
    {
      "epoch": 0.9912170639899623,
      "grad_norm": 5.108331203460693,
      "learning_rate": 1.3400250941028858e-05,
      "loss": 0.8931,
      "step": 790
    },
    {
      "epoch": 1.0037641154328734,
      "grad_norm": 9.46453857421875,
      "learning_rate": 1.3316603931409454e-05,
      "loss": 0.8834,
      "step": 800
    },
    {
      "epoch": 1.0163111668757843,
      "grad_norm": 8.54848861694336,
      "learning_rate": 1.3232956921790047e-05,
      "loss": 0.7073,
      "step": 810
    },
    {
      "epoch": 1.0288582183186952,
      "grad_norm": 8.689090728759766,
      "learning_rate": 1.314930991217064e-05,
      "loss": 0.8242,
      "step": 820
    },
    {
      "epoch": 1.041405269761606,
      "grad_norm": 8.000664710998535,
      "learning_rate": 1.3065662902551234e-05,
      "loss": 0.8314,
      "step": 830
    },
    {
      "epoch": 1.053952321204517,
      "grad_norm": 11.315167427062988,
      "learning_rate": 1.298201589293183e-05,
      "loss": 0.6794,
      "step": 840
    },
    {
      "epoch": 1.066499372647428,
      "grad_norm": 6.037744522094727,
      "learning_rate": 1.2898368883312423e-05,
      "loss": 0.7796,
      "step": 850
    },
    {
      "epoch": 1.0790464240903388,
      "grad_norm": 9.521272659301758,
      "learning_rate": 1.2814721873693016e-05,
      "loss": 0.92,
      "step": 860
    },
    {
      "epoch": 1.0915934755332497,
      "grad_norm": 18.728450775146484,
      "learning_rate": 1.2731074864073612e-05,
      "loss": 0.8371,
      "step": 870
    },
    {
      "epoch": 1.1041405269761606,
      "grad_norm": 7.143416404724121,
      "learning_rate": 1.2647427854454203e-05,
      "loss": 0.8534,
      "step": 880
    },
    {
      "epoch": 1.1166875784190715,
      "grad_norm": 7.581761837005615,
      "learning_rate": 1.2563780844834799e-05,
      "loss": 0.894,
      "step": 890
    },
    {
      "epoch": 1.1292346298619824,
      "grad_norm": 3.4014008045196533,
      "learning_rate": 1.248013383521539e-05,
      "loss": 0.8155,
      "step": 900
    },
    {
      "epoch": 1.1417816813048933,
      "grad_norm": 13.978635787963867,
      "learning_rate": 1.2396486825595986e-05,
      "loss": 0.754,
      "step": 910
    },
    {
      "epoch": 1.1543287327478042,
      "grad_norm": 5.18642520904541,
      "learning_rate": 1.2312839815976581e-05,
      "loss": 0.8493,
      "step": 920
    },
    {
      "epoch": 1.1668757841907151,
      "grad_norm": 4.153924942016602,
      "learning_rate": 1.2229192806357173e-05,
      "loss": 0.813,
      "step": 930
    },
    {
      "epoch": 1.179422835633626,
      "grad_norm": 9.386292457580566,
      "learning_rate": 1.2145545796737768e-05,
      "loss": 0.7526,
      "step": 940
    },
    {
      "epoch": 1.191969887076537,
      "grad_norm": 5.997525215148926,
      "learning_rate": 1.2061898787118363e-05,
      "loss": 0.8602,
      "step": 950
    },
    {
      "epoch": 1.2045169385194479,
      "grad_norm": 7.475282669067383,
      "learning_rate": 1.1978251777498955e-05,
      "loss": 0.6822,
      "step": 960
    },
    {
      "epoch": 1.2170639899623588,
      "grad_norm": 11.12500286102295,
      "learning_rate": 1.189460476787955e-05,
      "loss": 0.7505,
      "step": 970
    },
    {
      "epoch": 1.2296110414052697,
      "grad_norm": 10.952397346496582,
      "learning_rate": 1.1810957758260144e-05,
      "loss": 0.8621,
      "step": 980
    },
    {
      "epoch": 1.2421580928481806,
      "grad_norm": 16.098909378051758,
      "learning_rate": 1.1727310748640737e-05,
      "loss": 0.8263,
      "step": 990
    },
    {
      "epoch": 1.2547051442910915,
      "grad_norm": 7.630945682525635,
      "learning_rate": 1.164366373902133e-05,
      "loss": 0.7283,
      "step": 1000
    },
    {
      "epoch": 1.2672521957340024,
      "grad_norm": 5.213379859924316,
      "learning_rate": 1.1560016729401924e-05,
      "loss": 0.7849,
      "step": 1010
    },
    {
      "epoch": 1.2797992471769133,
      "grad_norm": 9.576950073242188,
      "learning_rate": 1.147636971978252e-05,
      "loss": 0.7845,
      "step": 1020
    },
    {
      "epoch": 1.2923462986198244,
      "grad_norm": 7.526945114135742,
      "learning_rate": 1.1392722710163113e-05,
      "loss": 0.8226,
      "step": 1030
    },
    {
      "epoch": 1.3048933500627353,
      "grad_norm": 8.773786544799805,
      "learning_rate": 1.1309075700543706e-05,
      "loss": 0.7912,
      "step": 1040
    },
    {
      "epoch": 1.3174404015056462,
      "grad_norm": 5.114284038543701,
      "learning_rate": 1.12254286909243e-05,
      "loss": 0.7414,
      "step": 1050
    },
    {
      "epoch": 1.3299874529485571,
      "grad_norm": 6.488368988037109,
      "learning_rate": 1.1141781681304895e-05,
      "loss": 0.8588,
      "step": 1060
    },
    {
      "epoch": 1.342534504391468,
      "grad_norm": 6.658433437347412,
      "learning_rate": 1.1058134671685487e-05,
      "loss": 0.9214,
      "step": 1070
    },
    {
      "epoch": 1.355081555834379,
      "grad_norm": 10.264144897460938,
      "learning_rate": 1.0974487662066082e-05,
      "loss": 0.7492,
      "step": 1080
    },
    {
      "epoch": 1.3676286072772899,
      "grad_norm": 7.86182165145874,
      "learning_rate": 1.0890840652446677e-05,
      "loss": 0.7919,
      "step": 1090
    },
    {
      "epoch": 1.3801756587202008,
      "grad_norm": 7.3965535163879395,
      "learning_rate": 1.080719364282727e-05,
      "loss": 0.8695,
      "step": 1100
    },
    {
      "epoch": 1.3927227101631117,
      "grad_norm": 9.31979751586914,
      "learning_rate": 1.0723546633207864e-05,
      "loss": 0.7495,
      "step": 1110
    },
    {
      "epoch": 1.4052697616060226,
      "grad_norm": 9.471373558044434,
      "learning_rate": 1.0639899623588456e-05,
      "loss": 0.8729,
      "step": 1120
    },
    {
      "epoch": 1.4178168130489335,
      "grad_norm": 13.156733512878418,
      "learning_rate": 1.0556252613969051e-05,
      "loss": 0.8688,
      "step": 1130
    },
    {
      "epoch": 1.4303638644918444,
      "grad_norm": 13.249898910522461,
      "learning_rate": 1.0472605604349647e-05,
      "loss": 0.8097,
      "step": 1140
    },
    {
      "epoch": 1.4429109159347553,
      "grad_norm": 6.98345947265625,
      "learning_rate": 1.0388958594730238e-05,
      "loss": 0.7635,
      "step": 1150
    },
    {
      "epoch": 1.4554579673776662,
      "grad_norm": 18.78076171875,
      "learning_rate": 1.0305311585110834e-05,
      "loss": 0.957,
      "step": 1160
    },
    {
      "epoch": 1.468005018820577,
      "grad_norm": 25.165569305419922,
      "learning_rate": 1.0221664575491427e-05,
      "loss": 0.7616,
      "step": 1170
    },
    {
      "epoch": 1.480552070263488,
      "grad_norm": 8.38501262664795,
      "learning_rate": 1.013801756587202e-05,
      "loss": 0.7768,
      "step": 1180
    },
    {
      "epoch": 1.4930991217063991,
      "grad_norm": 10.194948196411133,
      "learning_rate": 1.0054370556252616e-05,
      "loss": 0.8924,
      "step": 1190
    },
    {
      "epoch": 1.50564617314931,
      "grad_norm": 10.704676628112793,
      "learning_rate": 9.97072354663321e-06,
      "loss": 0.9066,
      "step": 1200
    },
    {
      "epoch": 1.518193224592221,
      "grad_norm": 21.507707595825195,
      "learning_rate": 9.887076537013803e-06,
      "loss": 0.9013,
      "step": 1210
    },
    {
      "epoch": 1.5307402760351319,
      "grad_norm": 8.45915699005127,
      "learning_rate": 9.803429527394396e-06,
      "loss": 1.0474,
      "step": 1220
    },
    {
      "epoch": 1.5432873274780428,
      "grad_norm": 5.320367336273193,
      "learning_rate": 9.71978251777499e-06,
      "loss": 0.8313,
      "step": 1230
    },
    {
      "epoch": 1.5558343789209537,
      "grad_norm": 10.026762962341309,
      "learning_rate": 9.636135508155583e-06,
      "loss": 0.8881,
      "step": 1240
    },
    {
      "epoch": 1.5683814303638646,
      "grad_norm": 9.318647384643555,
      "learning_rate": 9.552488498536177e-06,
      "loss": 0.8274,
      "step": 1250
    },
    {
      "epoch": 1.5809284818067755,
      "grad_norm": 9.438924789428711,
      "learning_rate": 9.468841488916772e-06,
      "loss": 0.9072,
      "step": 1260
    },
    {
      "epoch": 1.5934755332496864,
      "grad_norm": 11.8627290725708,
      "learning_rate": 9.385194479297366e-06,
      "loss": 0.746,
      "step": 1270
    },
    {
      "epoch": 1.6060225846925973,
      "grad_norm": 5.647990703582764,
      "learning_rate": 9.30154746967796e-06,
      "loss": 0.8526,
      "step": 1280
    },
    {
      "epoch": 1.6185696361355082,
      "grad_norm": 17.41633415222168,
      "learning_rate": 9.217900460058554e-06,
      "loss": 0.7216,
      "step": 1290
    },
    {
      "epoch": 1.631116687578419,
      "grad_norm": 10.73770523071289,
      "learning_rate": 9.134253450439148e-06,
      "loss": 0.7405,
      "step": 1300
    },
    {
      "epoch": 1.64366373902133,
      "grad_norm": 9.316720962524414,
      "learning_rate": 9.050606440819741e-06,
      "loss": 0.727,
      "step": 1310
    },
    {
      "epoch": 1.656210790464241,
      "grad_norm": 8.067049980163574,
      "learning_rate": 8.966959431200335e-06,
      "loss": 0.7992,
      "step": 1320
    },
    {
      "epoch": 1.6687578419071518,
      "grad_norm": 23.564279556274414,
      "learning_rate": 8.88331242158093e-06,
      "loss": 0.7617,
      "step": 1330
    },
    {
      "epoch": 1.6813048933500627,
      "grad_norm": 10.805792808532715,
      "learning_rate": 8.799665411961524e-06,
      "loss": 0.9413,
      "step": 1340
    },
    {
      "epoch": 1.6938519447929736,
      "grad_norm": 8.173852920532227,
      "learning_rate": 8.716018402342117e-06,
      "loss": 0.8204,
      "step": 1350
    },
    {
      "epoch": 1.7063989962358845,
      "grad_norm": 16.444934844970703,
      "learning_rate": 8.63237139272271e-06,
      "loss": 0.7124,
      "step": 1360
    },
    {
      "epoch": 1.7189460476787954,
      "grad_norm": 19.09043312072754,
      "learning_rate": 8.548724383103306e-06,
      "loss": 0.8573,
      "step": 1370
    },
    {
      "epoch": 1.7314930991217063,
      "grad_norm": 15.454293251037598,
      "learning_rate": 8.4650773734839e-06,
      "loss": 0.7964,
      "step": 1380
    },
    {
      "epoch": 1.7440401505646173,
      "grad_norm": 9.219696044921875,
      "learning_rate": 8.381430363864493e-06,
      "loss": 0.7494,
      "step": 1390
    },
    {
      "epoch": 1.7565872020075282,
      "grad_norm": 12.434812545776367,
      "learning_rate": 8.297783354245086e-06,
      "loss": 0.7066,
      "step": 1400
    },
    {
      "epoch": 1.769134253450439,
      "grad_norm": 8.64736270904541,
      "learning_rate": 8.21413634462568e-06,
      "loss": 0.8049,
      "step": 1410
    },
    {
      "epoch": 1.78168130489335,
      "grad_norm": 8.343708038330078,
      "learning_rate": 8.130489335006273e-06,
      "loss": 0.7935,
      "step": 1420
    },
    {
      "epoch": 1.7942283563362609,
      "grad_norm": 6.48344612121582,
      "learning_rate": 8.046842325386867e-06,
      "loss": 0.7431,
      "step": 1430
    },
    {
      "epoch": 1.8067754077791718,
      "grad_norm": 7.362094402313232,
      "learning_rate": 7.963195315767462e-06,
      "loss": 0.6778,
      "step": 1440
    },
    {
      "epoch": 1.8193224592220827,
      "grad_norm": 10.41412353515625,
      "learning_rate": 7.879548306148056e-06,
      "loss": 0.8869,
      "step": 1450
    },
    {
      "epoch": 1.8318695106649936,
      "grad_norm": 11.075960159301758,
      "learning_rate": 7.79590129652865e-06,
      "loss": 0.8913,
      "step": 1460
    },
    {
      "epoch": 1.8444165621079045,
      "grad_norm": 3.257545232772827,
      "learning_rate": 7.712254286909243e-06,
      "loss": 0.8214,
      "step": 1470
    },
    {
      "epoch": 1.8569636135508154,
      "grad_norm": 2.363823175430298,
      "learning_rate": 7.628607277289838e-06,
      "loss": 0.8268,
      "step": 1480
    },
    {
      "epoch": 1.8695106649937263,
      "grad_norm": 16.772573471069336,
      "learning_rate": 7.544960267670431e-06,
      "loss": 0.9585,
      "step": 1490
    },
    {
      "epoch": 1.8820577164366374,
      "grad_norm": 23.401941299438477,
      "learning_rate": 7.461313258051025e-06,
      "loss": 0.8636,
      "step": 1500
    },
    {
      "epoch": 1.8946047678795483,
      "grad_norm": 8.009302139282227,
      "learning_rate": 7.37766624843162e-06,
      "loss": 0.6728,
      "step": 1510
    },
    {
      "epoch": 1.9071518193224593,
      "grad_norm": 19.989112854003906,
      "learning_rate": 7.294019238812214e-06,
      "loss": 0.8526,
      "step": 1520
    },
    {
      "epoch": 1.9196988707653702,
      "grad_norm": 11.849247932434082,
      "learning_rate": 7.210372229192807e-06,
      "loss": 0.8825,
      "step": 1530
    },
    {
      "epoch": 1.932245922208281,
      "grad_norm": 8.389686584472656,
      "learning_rate": 7.126725219573401e-06,
      "loss": 0.8261,
      "step": 1540
    },
    {
      "epoch": 1.944792973651192,
      "grad_norm": 11.961833000183105,
      "learning_rate": 7.043078209953995e-06,
      "loss": 0.6564,
      "step": 1550
    },
    {
      "epoch": 1.9573400250941029,
      "grad_norm": 15.514620780944824,
      "learning_rate": 6.9594312003345885e-06,
      "loss": 0.7943,
      "step": 1560
    },
    {
      "epoch": 1.9698870765370138,
      "grad_norm": 10.584662437438965,
      "learning_rate": 6.875784190715182e-06,
      "loss": 0.7593,
      "step": 1570
    },
    {
      "epoch": 1.9824341279799247,
      "grad_norm": 15.600502014160156,
      "learning_rate": 6.7921371810957755e-06,
      "loss": 0.8643,
      "step": 1580
    },
    {
      "epoch": 1.9949811794228356,
      "grad_norm": 11.883516311645508,
      "learning_rate": 6.708490171476371e-06,
      "loss": 0.8381,
      "step": 1590
    },
    {
      "epoch": 2.0075282308657467,
      "grad_norm": 6.613313674926758,
      "learning_rate": 6.624843161856964e-06,
      "loss": 0.8074,
      "step": 1600
    },
    {
      "epoch": 2.0200752823086576,
      "grad_norm": 10.316112518310547,
      "learning_rate": 6.541196152237558e-06,
      "loss": 0.7161,
      "step": 1610
    },
    {
      "epoch": 2.0326223337515685,
      "grad_norm": 34.29435348510742,
      "learning_rate": 6.457549142618152e-06,
      "loss": 0.7553,
      "step": 1620
    },
    {
      "epoch": 2.0451693851944794,
      "grad_norm": 13.848133087158203,
      "learning_rate": 6.373902132998746e-06,
      "loss": 0.7486,
      "step": 1630
    },
    {
      "epoch": 2.0577164366373903,
      "grad_norm": 6.787230014801025,
      "learning_rate": 6.29025512337934e-06,
      "loss": 0.7867,
      "step": 1640
    },
    {
      "epoch": 2.0702634880803013,
      "grad_norm": 12.09325122833252,
      "learning_rate": 6.2066081137599335e-06,
      "loss": 0.7999,
      "step": 1650
    },
    {
      "epoch": 2.082810539523212,
      "grad_norm": 9.453817367553711,
      "learning_rate": 6.122961104140528e-06,
      "loss": 0.7668,
      "step": 1660
    },
    {
      "epoch": 2.095357590966123,
      "grad_norm": 10.155051231384277,
      "learning_rate": 6.039314094521121e-06,
      "loss": 0.7595,
      "step": 1670
    },
    {
      "epoch": 2.107904642409034,
      "grad_norm": 26.19032859802246,
      "learning_rate": 5.955667084901715e-06,
      "loss": 0.9065,
      "step": 1680
    },
    {
      "epoch": 2.120451693851945,
      "grad_norm": 9.538314819335938,
      "learning_rate": 5.872020075282308e-06,
      "loss": 0.7759,
      "step": 1690
    },
    {
      "epoch": 2.132998745294856,
      "grad_norm": 6.306014060974121,
      "learning_rate": 5.788373065662904e-06,
      "loss": 0.7431,
      "step": 1700
    },
    {
      "epoch": 2.1455457967377667,
      "grad_norm": 8.156487464904785,
      "learning_rate": 5.704726056043497e-06,
      "loss": 0.7029,
      "step": 1710
    },
    {
      "epoch": 2.1580928481806776,
      "grad_norm": 14.053692817687988,
      "learning_rate": 5.621079046424091e-06,
      "loss": 0.8357,
      "step": 1720
    },
    {
      "epoch": 2.1706398996235885,
      "grad_norm": 10.932027816772461,
      "learning_rate": 5.537432036804685e-06,
      "loss": 0.723,
      "step": 1730
    },
    {
      "epoch": 2.1831869510664994,
      "grad_norm": 23.12853240966797,
      "learning_rate": 5.4537850271852785e-06,
      "loss": 0.819,
      "step": 1740
    },
    {
      "epoch": 2.1957340025094103,
      "grad_norm": 10.468832015991211,
      "learning_rate": 5.370138017565872e-06,
      "loss": 0.7771,
      "step": 1750
    },
    {
      "epoch": 2.208281053952321,
      "grad_norm": 16.84822654724121,
      "learning_rate": 5.286491007946466e-06,
      "loss": 0.7895,
      "step": 1760
    },
    {
      "epoch": 2.220828105395232,
      "grad_norm": 17.586076736450195,
      "learning_rate": 5.202843998327061e-06,
      "loss": 0.8417,
      "step": 1770
    },
    {
      "epoch": 2.233375156838143,
      "grad_norm": 5.2970662117004395,
      "learning_rate": 5.119196988707654e-06,
      "loss": 0.7451,
      "step": 1780
    },
    {
      "epoch": 2.245922208281054,
      "grad_norm": 11.92138385772705,
      "learning_rate": 5.035549979088248e-06,
      "loss": 0.8368,
      "step": 1790
    },
    {
      "epoch": 2.258469259723965,
      "grad_norm": 22.60080909729004,
      "learning_rate": 4.951902969468842e-06,
      "loss": 0.7133,
      "step": 1800
    },
    {
      "epoch": 2.2710163111668757,
      "grad_norm": 12.277490615844727,
      "learning_rate": 4.868255959849436e-06,
      "loss": 0.8418,
      "step": 1810
    },
    {
      "epoch": 2.2835633626097867,
      "grad_norm": 6.862631320953369,
      "learning_rate": 4.78460895023003e-06,
      "loss": 0.7564,
      "step": 1820
    },
    {
      "epoch": 2.2961104140526976,
      "grad_norm": 7.9091267585754395,
      "learning_rate": 4.7009619406106235e-06,
      "loss": 0.7582,
      "step": 1830
    },
    {
      "epoch": 2.3086574654956085,
      "grad_norm": 8.295835494995117,
      "learning_rate": 4.617314930991217e-06,
      "loss": 0.8465,
      "step": 1840
    },
    {
      "epoch": 2.3212045169385194,
      "grad_norm": 11.34477710723877,
      "learning_rate": 4.533667921371811e-06,
      "loss": 0.7846,
      "step": 1850
    },
    {
      "epoch": 2.3337515683814303,
      "grad_norm": 32.95912170410156,
      "learning_rate": 4.450020911752405e-06,
      "loss": 0.6656,
      "step": 1860
    },
    {
      "epoch": 2.346298619824341,
      "grad_norm": 17.355579376220703,
      "learning_rate": 4.366373902132999e-06,
      "loss": 0.8187,
      "step": 1870
    },
    {
      "epoch": 2.358845671267252,
      "grad_norm": 8.741580963134766,
      "learning_rate": 4.282726892513593e-06,
      "loss": 0.744,
      "step": 1880
    },
    {
      "epoch": 2.371392722710163,
      "grad_norm": 12.800196647644043,
      "learning_rate": 4.199079882894187e-06,
      "loss": 0.6891,
      "step": 1890
    },
    {
      "epoch": 2.383939774153074,
      "grad_norm": 6.387624740600586,
      "learning_rate": 4.115432873274781e-06,
      "loss": 0.6968,
      "step": 1900
    },
    {
      "epoch": 2.396486825595985,
      "grad_norm": 9.609091758728027,
      "learning_rate": 4.031785863655375e-06,
      "loss": 0.7412,
      "step": 1910
    },
    {
      "epoch": 2.4090338770388957,
      "grad_norm": 16.369497299194336,
      "learning_rate": 3.9481388540359685e-06,
      "loss": 0.7136,
      "step": 1920
    },
    {
      "epoch": 2.4215809284818066,
      "grad_norm": 7.245341777801514,
      "learning_rate": 3.864491844416563e-06,
      "loss": 0.6953,
      "step": 1930
    },
    {
      "epoch": 2.4341279799247175,
      "grad_norm": 21.487659454345703,
      "learning_rate": 3.7808448347971568e-06,
      "loss": 0.776,
      "step": 1940
    },
    {
      "epoch": 2.4466750313676284,
      "grad_norm": 13.91763973236084,
      "learning_rate": 3.6971978251777503e-06,
      "loss": 0.9644,
      "step": 1950
    },
    {
      "epoch": 2.4592220828105393,
      "grad_norm": 16.034120559692383,
      "learning_rate": 3.6135508155583442e-06,
      "loss": 0.7451,
      "step": 1960
    },
    {
      "epoch": 2.4717691342534502,
      "grad_norm": 15.478806495666504,
      "learning_rate": 3.5299038059389377e-06,
      "loss": 0.6669,
      "step": 1970
    },
    {
      "epoch": 2.484316185696361,
      "grad_norm": 14.39084529876709,
      "learning_rate": 3.446256796319532e-06,
      "loss": 0.7545,
      "step": 1980
    },
    {
      "epoch": 2.496863237139272,
      "grad_norm": 27.011598587036133,
      "learning_rate": 3.3626097867001256e-06,
      "loss": 0.8599,
      "step": 1990
    },
    {
      "epoch": 2.509410288582183,
      "grad_norm": 15.559931755065918,
      "learning_rate": 3.27896277708072e-06,
      "loss": 0.8266,
      "step": 2000
    },
    {
      "epoch": 2.521957340025094,
      "grad_norm": 8.093729019165039,
      "learning_rate": 3.1953157674613135e-06,
      "loss": 0.7978,
      "step": 2010
    },
    {
      "epoch": 2.5345043914680048,
      "grad_norm": 13.65913200378418,
      "learning_rate": 3.1116687578419074e-06,
      "loss": 0.8145,
      "step": 2020
    },
    {
      "epoch": 2.5470514429109157,
      "grad_norm": 10.85305404663086,
      "learning_rate": 3.028021748222501e-06,
      "loss": 0.7976,
      "step": 2030
    },
    {
      "epoch": 2.5595984943538266,
      "grad_norm": 21.572189331054688,
      "learning_rate": 2.9443747386030953e-06,
      "loss": 0.7435,
      "step": 2040
    },
    {
      "epoch": 2.572145545796738,
      "grad_norm": 14.962433815002441,
      "learning_rate": 2.860727728983689e-06,
      "loss": 0.7856,
      "step": 2050
    },
    {
      "epoch": 2.584692597239649,
      "grad_norm": 8.81922435760498,
      "learning_rate": 2.7770807193642827e-06,
      "loss": 0.6729,
      "step": 2060
    },
    {
      "epoch": 2.5972396486825597,
      "grad_norm": 36.92323684692383,
      "learning_rate": 2.693433709744877e-06,
      "loss": 0.7374,
      "step": 2070
    },
    {
      "epoch": 2.6097867001254706,
      "grad_norm": 19.693674087524414,
      "learning_rate": 2.6097867001254706e-06,
      "loss": 0.6577,
      "step": 2080
    },
    {
      "epoch": 2.6223337515683816,
      "grad_norm": 6.363992691040039,
      "learning_rate": 2.526139690506065e-06,
      "loss": 0.7902,
      "step": 2090
    },
    {
      "epoch": 2.6348808030112925,
      "grad_norm": 18.565412521362305,
      "learning_rate": 2.4424926808866585e-06,
      "loss": 0.638,
      "step": 2100
    },
    {
      "epoch": 2.6474278544542034,
      "grad_norm": 23.991640090942383,
      "learning_rate": 2.3588456712672524e-06,
      "loss": 0.8057,
      "step": 2110
    },
    {
      "epoch": 2.6599749058971143,
      "grad_norm": 11.251315116882324,
      "learning_rate": 2.2751986616478463e-06,
      "loss": 0.7766,
      "step": 2120
    },
    {
      "epoch": 2.672521957340025,
      "grad_norm": 13.534173011779785,
      "learning_rate": 2.1915516520284403e-06,
      "loss": 0.7792,
      "step": 2130
    },
    {
      "epoch": 2.685069008782936,
      "grad_norm": 28.147571563720703,
      "learning_rate": 2.107904642409034e-06,
      "loss": 0.7725,
      "step": 2140
    },
    {
      "epoch": 2.697616060225847,
      "grad_norm": 25.395660400390625,
      "learning_rate": 2.024257632789628e-06,
      "loss": 0.7114,
      "step": 2150
    },
    {
      "epoch": 2.710163111668758,
      "grad_norm": 11.658987998962402,
      "learning_rate": 1.9406106231702216e-06,
      "loss": 0.7515,
      "step": 2160
    },
    {
      "epoch": 2.722710163111669,
      "grad_norm": 7.770451068878174,
      "learning_rate": 1.8569636135508156e-06,
      "loss": 0.7042,
      "step": 2170
    },
    {
      "epoch": 2.7352572145545797,
      "grad_norm": 15.549407005310059,
      "learning_rate": 1.7733166039314095e-06,
      "loss": 0.8029,
      "step": 2180
    },
    {
      "epoch": 2.7478042659974906,
      "grad_norm": 9.49413013458252,
      "learning_rate": 1.6896695943120034e-06,
      "loss": 0.642,
      "step": 2190
    },
    {
      "epoch": 2.7603513174404015,
      "grad_norm": 4.831557750701904,
      "learning_rate": 1.6060225846925972e-06,
      "loss": 0.6564,
      "step": 2200
    },
    {
      "epoch": 2.7728983688833124,
      "grad_norm": 12.240572929382324,
      "learning_rate": 1.5223755750731913e-06,
      "loss": 0.7423,
      "step": 2210
    },
    {
      "epoch": 2.7854454203262233,
      "grad_norm": 5.629863739013672,
      "learning_rate": 1.4387285654537853e-06,
      "loss": 0.8664,
      "step": 2220
    },
    {
      "epoch": 2.7979924717691342,
      "grad_norm": 14.666424751281738,
      "learning_rate": 1.3550815558343792e-06,
      "loss": 0.8401,
      "step": 2230
    },
    {
      "epoch": 2.810539523212045,
      "grad_norm": 21.368179321289062,
      "learning_rate": 1.271434546214973e-06,
      "loss": 0.8129,
      "step": 2240
    },
    {
      "epoch": 2.823086574654956,
      "grad_norm": 16.279918670654297,
      "learning_rate": 1.1877875365955668e-06,
      "loss": 0.676,
      "step": 2250
    },
    {
      "epoch": 2.835633626097867,
      "grad_norm": 9.378284454345703,
      "learning_rate": 1.1041405269761608e-06,
      "loss": 0.6352,
      "step": 2260
    },
    {
      "epoch": 2.848180677540778,
      "grad_norm": 21.204431533813477,
      "learning_rate": 1.0204935173567545e-06,
      "loss": 0.7627,
      "step": 2270
    },
    {
      "epoch": 2.8607277289836888,
      "grad_norm": 23.616268157958984,
      "learning_rate": 9.368465077373484e-07,
      "loss": 0.6356,
      "step": 2280
    },
    {
      "epoch": 2.8732747804265997,
      "grad_norm": 25.798891067504883,
      "learning_rate": 8.531994981179424e-07,
      "loss": 0.7385,
      "step": 2290
    },
    {
      "epoch": 2.8858218318695106,
      "grad_norm": 11.757165908813477,
      "learning_rate": 7.695524884985363e-07,
      "loss": 0.6858,
      "step": 2300
    },
    {
      "epoch": 2.8983688833124215,
      "grad_norm": 9.811064720153809,
      "learning_rate": 6.859054788791301e-07,
      "loss": 0.6791,
      "step": 2310
    },
    {
      "epoch": 2.9109159347553324,
      "grad_norm": 14.021916389465332,
      "learning_rate": 6.02258469259724e-07,
      "loss": 0.8106,
      "step": 2320
    },
    {
      "epoch": 2.9234629861982433,
      "grad_norm": 16.58710479736328,
      "learning_rate": 5.186114596403179e-07,
      "loss": 0.8049,
      "step": 2330
    },
    {
      "epoch": 2.936010037641154,
      "grad_norm": 8.84688663482666,
      "learning_rate": 4.349644500209118e-07,
      "loss": 0.6731,
      "step": 2340
    },
    {
      "epoch": 2.948557089084065,
      "grad_norm": 9.54117488861084,
      "learning_rate": 3.5131744040150566e-07,
      "loss": 0.7293,
      "step": 2350
    },
    {
      "epoch": 2.961104140526976,
      "grad_norm": 17.779142379760742,
      "learning_rate": 2.6767043078209955e-07,
      "loss": 0.7803,
      "step": 2360
    },
    {
      "epoch": 2.973651191969887,
      "grad_norm": 7.330904006958008,
      "learning_rate": 1.8402342116269345e-07,
      "loss": 0.7281,
      "step": 2370
    },
    {
      "epoch": 2.9861982434127983,
      "grad_norm": 31.703744888305664,
      "learning_rate": 1.0037641154328732e-07,
      "loss": 0.8839,
      "step": 2380
    },
    {
      "epoch": 2.998745294855709,
      "grad_norm": 13.579190254211426,
      "learning_rate": 1.672940192388122e-08,
      "loss": 0.8382,
      "step": 2390
    }
  ],
  "logging_steps": 10,
  "max_steps": 2391,
  "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": 274122632540160.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}