{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0223,
  "eval_steps": 500,
  "global_step": 2800,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.001,
      "grad_norm": 0.3872889280319214,
      "learning_rate": 3.8817000000000005e-06,
      "loss": 1.5489984512329102,
      "step": 10
    },
    {
      "epoch": 0.002,
      "grad_norm": 0.3785003423690796,
      "learning_rate": 8.1947e-06,
      "loss": 1.554966926574707,
      "step": 20
    },
    {
      "epoch": 0.003,
      "grad_norm": 0.38793623447418213,
      "learning_rate": 1.25077e-05,
      "loss": 1.546360397338867,
      "step": 30
    },
    {
      "epoch": 0.004,
      "grad_norm": 0.4206889569759369,
      "learning_rate": 1.68207e-05,
      "loss": 1.606406021118164,
      "step": 40
    },
    {
      "epoch": 0.005,
      "grad_norm": 0.38157036900520325,
      "learning_rate": 2.11337e-05,
      "loss": 1.5404697418212892,
      "step": 50
    },
    {
      "epoch": 0.006,
      "grad_norm": 0.39413636922836304,
      "learning_rate": 2.54467e-05,
      "loss": 1.5285799026489257,
      "step": 60
    },
    {
      "epoch": 0.007,
      "grad_norm": 0.40694475173950195,
      "learning_rate": 2.97597e-05,
      "loss": 1.504967498779297,
      "step": 70
    },
    {
      "epoch": 0.008,
      "grad_norm": 0.41990774869918823,
      "learning_rate": 3.40727e-05,
      "loss": 1.5729640960693358,
      "step": 80
    },
    {
      "epoch": 0.009,
      "grad_norm": 0.403705358505249,
      "learning_rate": 3.83857e-05,
      "loss": 1.5551136970520019,
      "step": 90
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.3957897424697876,
      "learning_rate": 4.2698700000000005e-05,
      "loss": 1.5575121879577636,
      "step": 100
    },
    {
      "epoch": 0.011,
      "grad_norm": 0.40792497992515564,
      "learning_rate": 4.70117e-05,
      "loss": 1.518178939819336,
      "step": 110
    },
    {
      "epoch": 0.012,
      "grad_norm": 0.42794761061668396,
      "learning_rate": 5.13247e-05,
      "loss": 1.5593727111816407,
      "step": 120
    },
    {
      "epoch": 0.013,
      "grad_norm": 0.4005564749240875,
      "learning_rate": 5.5637700000000004e-05,
      "loss": 1.562605857849121,
      "step": 130
    },
    {
      "epoch": 0.014,
      "grad_norm": 0.3942921757698059,
      "learning_rate": 5.99507e-05,
      "loss": 1.5700313568115234,
      "step": 140
    },
    {
      "epoch": 0.015,
      "grad_norm": 0.399198055267334,
      "learning_rate": 6.42637e-05,
      "loss": 1.5733401298522949,
      "step": 150
    },
    {
      "epoch": 0.016,
      "grad_norm": 0.39514100551605225,
      "learning_rate": 6.857670000000001e-05,
      "loss": 1.5353406906127929,
      "step": 160
    },
    {
      "epoch": 0.017,
      "grad_norm": 0.3957034945487976,
      "learning_rate": 7.288970000000001e-05,
      "loss": 1.5085016250610352,
      "step": 170
    },
    {
      "epoch": 0.018,
      "grad_norm": 0.4351767897605896,
      "learning_rate": 7.72027e-05,
      "loss": 1.5530328750610352,
      "step": 180
    },
    {
      "epoch": 0.019,
      "grad_norm": 0.418708473443985,
      "learning_rate": 8.15157e-05,
      "loss": 1.5638229370117187,
      "step": 190
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.40256965160369873,
      "learning_rate": 8.58287e-05,
      "loss": 1.5796945571899415,
      "step": 200
    },
    {
      "epoch": 0.021,
      "grad_norm": 0.397319495677948,
      "learning_rate": 8.625982049284077e-05,
      "loss": 1.5816753387451172,
      "step": 210
    },
    {
      "epoch": 0.022,
      "grad_norm": 0.41874557733535767,
      "learning_rate": 8.625919997618398e-05,
      "loss": 1.5102975845336915,
      "step": 220
    },
    {
      "epoch": 0.023,
      "grad_norm": 0.40078213810920715,
      "learning_rate": 8.625813624026713e-05,
      "loss": 1.571668529510498,
      "step": 230
    },
    {
      "epoch": 0.024,
      "grad_norm": 0.43663620948791504,
      "learning_rate": 8.625662929602175e-05,
      "loss": 1.5277430534362793,
      "step": 240
    },
    {
      "epoch": 0.025,
      "grad_norm": 0.4428839385509491,
      "learning_rate": 8.625467915893403e-05,
      "loss": 1.5955425262451173,
      "step": 250
    },
    {
      "epoch": 0.026,
      "grad_norm": 0.4413251578807831,
      "learning_rate": 8.625228584904462e-05,
      "loss": 1.5842302322387696,
      "step": 260
    },
    {
      "epoch": 0.027,
      "grad_norm": 0.42099958658218384,
      "learning_rate": 8.62494493909485e-05,
      "loss": 1.5954334259033203,
      "step": 270
    },
    {
      "epoch": 0.028,
      "grad_norm": 0.43221530318260193,
      "learning_rate": 8.624616981379467e-05,
      "loss": 1.5589290618896485,
      "step": 280
    },
    {
      "epoch": 0.029,
      "grad_norm": 0.42665326595306396,
      "learning_rate": 8.624244715128585e-05,
      "loss": 1.5465147972106934,
      "step": 290
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.42965951561927795,
      "learning_rate": 8.623828144167817e-05,
      "loss": 1.5350761413574219,
      "step": 300
    },
    {
      "epoch": 0.031,
      "grad_norm": 0.42243221402168274,
      "learning_rate": 8.623367272778072e-05,
      "loss": 1.5516321182250976,
      "step": 310
    },
    {
      "epoch": 0.032,
      "grad_norm": 0.4054717421531677,
      "learning_rate": 8.622862105695517e-05,
      "loss": 1.5238536834716796,
      "step": 320
    },
    {
      "epoch": 0.033,
      "grad_norm": 0.43124857544898987,
      "learning_rate": 8.622312648111529e-05,
      "loss": 1.5706765174865722,
      "step": 330
    },
    {
      "epoch": 0.034,
      "grad_norm": 0.4318324625492096,
      "learning_rate": 8.621718905672629e-05,
      "loss": 1.566463565826416,
      "step": 340
    },
    {
      "epoch": 0.035,
      "grad_norm": 0.43307042121887207,
      "learning_rate": 8.621080884480445e-05,
      "loss": 1.530984878540039,
      "step": 350
    },
    {
      "epoch": 0.036,
      "grad_norm": 0.43541616201400757,
      "learning_rate": 8.620398591091628e-05,
      "loss": 1.5489194869995118,
      "step": 360
    },
    {
      "epoch": 0.037,
      "grad_norm": 0.40485483407974243,
      "learning_rate": 8.619672032517798e-05,
      "loss": 1.576507568359375,
      "step": 370
    },
    {
      "epoch": 0.038,
      "grad_norm": 0.38804423809051514,
      "learning_rate": 8.618901216225471e-05,
      "loss": 1.5632071495056152,
      "step": 380
    },
    {
      "epoch": 0.039,
      "grad_norm": 0.44407907128334045,
      "learning_rate": 8.618086150135976e-05,
      "loss": 1.571460247039795,
      "step": 390
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.4320124387741089,
      "learning_rate": 8.617226842625377e-05,
      "loss": 1.5724874496459962,
      "step": 400
    },
    {
      "epoch": 0.041,
      "grad_norm": 0.41030609607696533,
      "learning_rate": 8.61632330252439e-05,
      "loss": 1.552263069152832,
      "step": 410
    },
    {
      "epoch": 0.042,
      "grad_norm": 0.4280606210231781,
      "learning_rate": 8.615375539118286e-05,
      "loss": 1.5935471534729004,
      "step": 420
    },
    {
      "epoch": 0.043,
      "grad_norm": 0.425942063331604,
      "learning_rate": 8.614383562146802e-05,
      "loss": 1.618431854248047,
      "step": 430
    },
    {
      "epoch": 0.044,
      "grad_norm": 0.4112652540206909,
      "learning_rate": 8.613347381804036e-05,
      "loss": 1.6068254470825196,
      "step": 440
    },
    {
      "epoch": 0.045,
      "grad_norm": 0.43573281168937683,
      "learning_rate": 8.612267008738342e-05,
      "loss": 1.501287078857422,
      "step": 450
    },
    {
      "epoch": 0.046,
      "grad_norm": 0.43779465556144714,
      "learning_rate": 8.611142454052226e-05,
      "loss": 1.5798423767089844,
      "step": 460
    },
    {
      "epoch": 0.047,
      "grad_norm": 0.45562246441841125,
      "learning_rate": 8.609973729302228e-05,
      "loss": 1.5571372985839844,
      "step": 470
    },
    {
      "epoch": 0.048,
      "grad_norm": 0.4224706292152405,
      "learning_rate": 8.608760846498805e-05,
      "loss": 1.5518342971801757,
      "step": 480
    },
    {
      "epoch": 0.049,
      "grad_norm": 0.4402417242527008,
      "learning_rate": 8.6075038181062e-05,
      "loss": 1.5404040336608886,
      "step": 490
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.42017725110054016,
      "learning_rate": 8.606202657042327e-05,
      "loss": 1.5324398040771485,
      "step": 500
    },
    {
      "epoch": 0.051,
      "grad_norm": 0.41633936762809753,
      "learning_rate": 8.604857376678628e-05,
      "loss": 1.6201072692871095,
      "step": 510
    },
    {
      "epoch": 0.052,
      "grad_norm": 0.4232465922832489,
      "learning_rate": 8.603467990839943e-05,
      "loss": 1.579312515258789,
      "step": 520
    },
    {
      "epoch": 0.053,
      "grad_norm": 0.4202162027359009,
      "learning_rate": 8.602034513804359e-05,
      "loss": 1.5241361618041993,
      "step": 530
    },
    {
      "epoch": 0.054,
      "grad_norm": 0.4149995446205139,
      "learning_rate": 8.60055696030307e-05,
      "loss": 1.5884086608886718,
      "step": 540
    },
    {
      "epoch": 0.055,
      "grad_norm": 0.4352942109107971,
      "learning_rate": 8.599035345520225e-05,
      "loss": 1.5795814514160156,
      "step": 550
    },
    {
      "epoch": 0.056,
      "grad_norm": 0.406890332698822,
      "learning_rate": 8.59746968509277e-05,
      "loss": 1.5849136352539062,
      "step": 560
    },
    {
      "epoch": 0.057,
      "grad_norm": 0.4194611608982086,
      "learning_rate": 8.595859995110288e-05,
      "loss": 1.5220673561096192,
      "step": 570
    },
    {
      "epoch": 0.058,
      "grad_norm": 0.42952999472618103,
      "learning_rate": 8.594206292114837e-05,
      "loss": 1.5716930389404298,
      "step": 580
    },
    {
      "epoch": 0.059,
      "grad_norm": 0.4449344575405121,
      "learning_rate": 8.59250859310077e-05,
      "loss": 1.5606157302856445,
      "step": 590
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.4540117383003235,
      "learning_rate": 8.590766915514571e-05,
      "loss": 1.5590102195739746,
      "step": 600
    },
    {
      "epoch": 0.061,
      "grad_norm": 0.4081774652004242,
      "learning_rate": 8.588981277254677e-05,
      "loss": 1.513133430480957,
      "step": 610
    },
    {
      "epoch": 0.062,
      "grad_norm": 0.43133676052093506,
      "learning_rate": 8.587151696671278e-05,
      "loss": 1.5470260620117187,
      "step": 620
    },
    {
      "epoch": 0.063,
      "grad_norm": 0.42611169815063477,
      "learning_rate": 8.585278192566148e-05,
      "loss": 1.5632414817810059,
      "step": 630
    },
    {
      "epoch": 0.064,
      "grad_norm": 0.4463273286819458,
      "learning_rate": 8.583360784192441e-05,
      "loss": 1.6180675506591797,
      "step": 640
    },
    {
      "epoch": 0.065,
      "grad_norm": 0.4165816009044647,
      "learning_rate": 8.581399491254493e-05,
      "loss": 1.5370763778686523,
      "step": 650
    },
    {
      "epoch": 0.066,
      "grad_norm": 0.44205260276794434,
      "learning_rate": 8.579394333907624e-05,
      "loss": 1.558724594116211,
      "step": 660
    },
    {
      "epoch": 0.067,
      "grad_norm": 0.4382662773132324,
      "learning_rate": 8.57734533275793e-05,
      "loss": 1.5468637466430664,
      "step": 670
    },
    {
      "epoch": 0.068,
      "grad_norm": 0.40459638833999634,
      "learning_rate": 8.575252508862066e-05,
      "loss": 1.543410301208496,
      "step": 680
    },
    {
      "epoch": 0.069,
      "grad_norm": 0.42760446667671204,
      "learning_rate": 8.573115883727036e-05,
      "loss": 1.590751838684082,
      "step": 690
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.42787182331085205,
      "learning_rate": 8.570935479309969e-05,
      "loss": 1.591897201538086,
      "step": 700
    },
    {
      "epoch": 0.071,
      "grad_norm": 0.4374030828475952,
      "learning_rate": 8.568711318017893e-05,
      "loss": 1.5722368240356446,
      "step": 710
    },
    {
      "epoch": 0.072,
      "grad_norm": 0.4523533582687378,
      "learning_rate": 8.566443422707509e-05,
      "loss": 1.5926239013671875,
      "step": 720
    },
    {
      "epoch": 0.073,
      "grad_norm": 0.44648537039756775,
      "learning_rate": 8.564131816684949e-05,
      "loss": 1.5328532218933106,
      "step": 730
    },
    {
      "epoch": 0.074,
      "grad_norm": 0.44178444147109985,
      "learning_rate": 8.561776523705541e-05,
      "loss": 1.5162950515747071,
      "step": 740
    },
    {
      "epoch": 0.075,
      "grad_norm": 0.459213525056839,
      "learning_rate": 8.559377567973564e-05,
      "loss": 1.627875518798828,
      "step": 750
    },
    {
      "epoch": 0.076,
      "grad_norm": 0.4523981511592865,
      "learning_rate": 8.556934974142002e-05,
      "loss": 1.5238592147827148,
      "step": 760
    },
    {
      "epoch": 0.077,
      "grad_norm": 0.4350197911262512,
      "learning_rate": 8.554448767312285e-05,
      "loss": 1.5533188819885253,
      "step": 770
    },
    {
      "epoch": 0.078,
      "grad_norm": 0.461459755897522,
      "learning_rate": 8.551918973034032e-05,
      "loss": 1.5818538665771484,
      "step": 780
    },
    {
      "epoch": 0.079,
      "grad_norm": 0.42179811000823975,
      "learning_rate": 8.549345617304799e-05,
      "loss": 1.5900559425354004,
      "step": 790
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.4687558710575104,
      "learning_rate": 8.546728726569792e-05,
      "loss": 1.5136714935302735,
      "step": 800
    },
    {
      "epoch": 0.081,
      "grad_norm": 0.42299333214759827,
      "learning_rate": 8.544068327721616e-05,
      "loss": 1.6079557418823243,
      "step": 810
    },
    {
      "epoch": 0.082,
      "grad_norm": 0.45741724967956543,
      "learning_rate": 8.541364448099983e-05,
      "loss": 1.6250675201416016,
      "step": 820
    },
    {
      "epoch": 0.083,
      "grad_norm": 0.47489041090011597,
      "learning_rate": 8.538617115491442e-05,
      "loss": 1.5306575775146485,
      "step": 830
    },
    {
      "epoch": 0.084,
      "grad_norm": 0.43424174189567566,
      "learning_rate": 8.535826358129085e-05,
      "loss": 1.5837011337280273,
      "step": 840
    },
    {
      "epoch": 0.085,
      "grad_norm": 0.43130776286125183,
      "learning_rate": 8.53299220469226e-05,
      "loss": 1.560431957244873,
      "step": 850
    },
    {
      "epoch": 0.086,
      "grad_norm": 0.4119592607021332,
      "learning_rate": 8.530114684306282e-05,
      "loss": 1.572993278503418,
      "step": 860
    },
    {
      "epoch": 0.087,
      "grad_norm": 0.4472162425518036,
      "learning_rate": 8.527193826542123e-05,
      "loss": 1.5893125534057617,
      "step": 870
    },
    {
      "epoch": 0.088,
      "grad_norm": 0.43497446179389954,
      "learning_rate": 8.524229661416117e-05,
      "loss": 1.560661506652832,
      "step": 880
    },
    {
      "epoch": 0.089,
      "grad_norm": 0.43729737401008606,
      "learning_rate": 8.521222219389646e-05,
      "loss": 1.5510387420654297,
      "step": 890
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.4537718594074249,
      "learning_rate": 8.518171531368828e-05,
      "loss": 1.5789440155029297,
      "step": 900
    },
    {
      "epoch": 0.091,
      "grad_norm": 0.4269010126590729,
      "learning_rate": 8.515077628704208e-05,
      "loss": 1.5246685028076172,
      "step": 910
    },
    {
      "epoch": 0.092,
      "grad_norm": 0.4279134273529053,
      "learning_rate": 8.511940543190416e-05,
      "loss": 1.5408878326416016,
      "step": 920
    },
    {
      "epoch": 0.093,
      "grad_norm": 0.4487226605415344,
      "learning_rate": 8.508760307065864e-05,
      "loss": 1.5889862060546875,
      "step": 930
    },
    {
      "epoch": 0.094,
      "grad_norm": 0.4346173107624054,
      "learning_rate": 8.505536953012398e-05,
      "loss": 1.4902814865112304,
      "step": 940
    },
    {
      "epoch": 0.095,
      "grad_norm": 0.431169718503952,
      "learning_rate": 8.502270514154968e-05,
      "loss": 1.5507546424865724,
      "step": 950
    },
    {
      "epoch": 0.096,
      "grad_norm": 0.46881142258644104,
      "learning_rate": 8.498961024061287e-05,
      "loss": 1.5701534271240234,
      "step": 960
    },
    {
      "epoch": 0.097,
      "grad_norm": 0.42858099937438965,
      "learning_rate": 8.495608516741486e-05,
      "loss": 1.4937341690063477,
      "step": 970
    },
    {
      "epoch": 0.098,
      "grad_norm": 0.4465695023536682,
      "learning_rate": 8.492213026647764e-05,
      "loss": 1.5935869216918945,
      "step": 980
    },
    {
      "epoch": 0.099,
      "grad_norm": 0.4498782157897949,
      "learning_rate": 8.488774588674037e-05,
      "loss": 1.5635653495788575,
      "step": 990
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.4431057274341583,
      "learning_rate": 8.485293238155573e-05,
      "loss": 1.5423452377319335,
      "step": 1000
    },
    {
      "epoch": 0.101,
      "grad_norm": 0.4575574994087219,
      "learning_rate": 8.481769010868637e-05,
      "loss": 1.5350657463073731,
      "step": 1010
    },
    {
      "epoch": 0.102,
      "grad_norm": 0.4342168867588043,
      "learning_rate": 8.478201943030117e-05,
      "loss": 1.5452421188354493,
      "step": 1020
    },
    {
      "epoch": 0.103,
      "grad_norm": 0.445075124502182,
      "learning_rate": 8.474592071297152e-05,
      "loss": 1.554254150390625,
      "step": 1030
    },
    {
      "epoch": 0.104,
      "grad_norm": 0.4207341969013214,
      "learning_rate": 8.470939432766762e-05,
      "loss": 1.5895505905151368,
      "step": 1040
    },
    {
      "epoch": 0.105,
      "grad_norm": 0.4459976851940155,
      "learning_rate": 8.467244064975459e-05,
      "loss": 1.546200942993164,
      "step": 1050
    },
    {
      "epoch": 0.106,
      "grad_norm": 0.4187023341655731,
      "learning_rate": 8.463506005898863e-05,
      "loss": 1.5332408905029298,
      "step": 1060
    },
    {
      "epoch": 0.107,
      "grad_norm": 0.4594968557357788,
      "learning_rate": 8.459725293951314e-05,
      "loss": 1.5771045684814453,
      "step": 1070
    },
    {
      "epoch": 0.108,
      "grad_norm": 0.4230520725250244,
      "learning_rate": 8.455901967985479e-05,
      "loss": 1.5950983047485352,
      "step": 1080
    },
    {
      "epoch": 0.109,
      "grad_norm": 0.43380364775657654,
      "learning_rate": 8.452036067291948e-05,
      "loss": 1.6060211181640625,
      "step": 1090
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.4305785596370697,
      "learning_rate": 8.448127631598831e-05,
      "loss": 1.5464977264404296,
      "step": 1100
    },
    {
      "epoch": 0.111,
      "grad_norm": 0.46058452129364014,
      "learning_rate": 8.44417670107135e-05,
      "loss": 1.5613780975341798,
      "step": 1110
    },
    {
      "epoch": 0.112,
      "grad_norm": 0.42492201924324036,
      "learning_rate": 8.440183316311433e-05,
      "loss": 1.5371337890625,
      "step": 1120
    },
    {
      "epoch": 0.113,
      "grad_norm": 0.45703360438346863,
      "learning_rate": 8.436147518357284e-05,
      "loss": 1.5895401000976563,
      "step": 1130
    },
    {
      "epoch": 0.114,
      "grad_norm": 0.4336182773113251,
      "learning_rate": 8.432069348682973e-05,
      "loss": 1.5472881317138671,
      "step": 1140
    },
    {
      "epoch": 0.115,
      "grad_norm": 0.44153451919555664,
      "learning_rate": 8.427948849198002e-05,
      "loss": 1.5661738395690918,
      "step": 1150
    },
    {
      "epoch": 0.116,
      "grad_norm": 0.43905603885650635,
      "learning_rate": 8.42378606224688e-05,
      "loss": 1.5574228286743164,
      "step": 1160
    },
    {
      "epoch": 0.117,
      "grad_norm": 0.43091803789138794,
      "learning_rate": 8.419581030608683e-05,
      "loss": 1.5603097915649413,
      "step": 1170
    },
    {
      "epoch": 0.118,
      "grad_norm": 0.4270052909851074,
      "learning_rate": 8.415333797496616e-05,
      "loss": 1.552748680114746,
      "step": 1180
    },
    {
      "epoch": 0.119,
      "grad_norm": 0.4377681016921997,
      "learning_rate": 8.411044406557571e-05,
      "loss": 1.5209136962890626,
      "step": 1190
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.4468444585800171,
      "learning_rate": 8.406712901871679e-05,
      "loss": 1.5935579299926759,
      "step": 1200
    },
    {
      "epoch": 0.121,
      "grad_norm": 0.40383124351501465,
      "learning_rate": 8.402339327951851e-05,
      "loss": 1.5492631912231445,
      "step": 1210
    },
    {
      "epoch": 0.122,
      "grad_norm": 0.453901469707489,
      "learning_rate": 8.397923729743324e-05,
      "loss": 1.5286017417907716,
      "step": 1220
    },
    {
      "epoch": 0.123,
      "grad_norm": 0.45355984568595886,
      "learning_rate": 8.393466152623203e-05,
      "loss": 1.534010887145996,
      "step": 1230
    },
    {
      "epoch": 0.124,
      "grad_norm": 0.41849273443222046,
      "learning_rate": 8.388966642399991e-05,
      "loss": 1.5546277999877929,
      "step": 1240
    },
    {
      "epoch": 0.125,
      "grad_norm": 0.42275470495224,
      "learning_rate": 8.384425245313115e-05,
      "loss": 1.5799736976623535,
      "step": 1250
    },
    {
      "epoch": 0.126,
      "grad_norm": 0.44112563133239746,
      "learning_rate": 8.379842008032457e-05,
      "loss": 1.5725579261779785,
      "step": 1260
    },
    {
      "epoch": 0.127,
      "grad_norm": 0.42606085538864136,
      "learning_rate": 8.375216977657873e-05,
      "loss": 1.534956932067871,
      "step": 1270
    },
    {
      "epoch": 0.128,
      "grad_norm": 0.44213414192199707,
      "learning_rate": 8.370550201718704e-05,
      "loss": 1.496871566772461,
      "step": 1280
    },
    {
      "epoch": 0.129,
      "grad_norm": 0.47176358103752136,
      "learning_rate": 8.365841728173296e-05,
      "loss": 1.5814940452575683,
      "step": 1290
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.4457390606403351,
      "learning_rate": 8.361091605408497e-05,
      "loss": 1.599702739715576,
      "step": 1300
    },
    {
      "epoch": 0.131,
      "grad_norm": 0.43326595425605774,
      "learning_rate": 8.356299882239168e-05,
      "loss": 1.5610076904296875,
      "step": 1310
    },
    {
      "epoch": 0.132,
      "grad_norm": 0.43634170293807983,
      "learning_rate": 8.35146660790768e-05,
      "loss": 1.5391773223876952,
      "step": 1320
    },
    {
      "epoch": 0.133,
      "grad_norm": 0.4215056002140045,
      "learning_rate": 8.346591832083405e-05,
      "loss": 1.538853359222412,
      "step": 1330
    },
    {
      "epoch": 0.134,
      "grad_norm": 0.4359005093574524,
      "learning_rate": 8.341675604862204e-05,
      "loss": 1.5893823623657226,
      "step": 1340
    },
    {
      "epoch": 0.135,
      "grad_norm": 0.4620271623134613,
      "learning_rate": 8.336717976765922e-05,
      "loss": 1.544410514831543,
      "step": 1350
    },
    {
      "epoch": 0.136,
      "grad_norm": 0.45624276995658875,
      "learning_rate": 8.331718998741857e-05,
      "loss": 1.5727739334106445,
      "step": 1360
    },
    {
      "epoch": 0.137,
      "grad_norm": 0.47488826513290405,
      "learning_rate": 8.32667872216224e-05,
      "loss": 1.5734460830688477,
      "step": 1370
    },
    {
      "epoch": 0.138,
      "grad_norm": 0.4481564462184906,
      "learning_rate": 8.321597198823715e-05,
      "loss": 1.5132453918457032,
      "step": 1380
    },
    {
      "epoch": 0.139,
      "grad_norm": 0.44024789333343506,
      "learning_rate": 8.316474480946796e-05,
      "loss": 1.5181768417358399,
      "step": 1390
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.43633797764778137,
      "learning_rate": 8.311310621175332e-05,
      "loss": 1.5812137603759766,
      "step": 1400
    },
    {
      "epoch": 0.141,
      "grad_norm": 0.4629945158958435,
      "learning_rate": 8.306105672575975e-05,
      "loss": 1.5248623847961427,
      "step": 1410
    },
    {
      "epoch": 0.142,
      "grad_norm": 0.44547826051712036,
      "learning_rate": 8.300859688637624e-05,
      "loss": 1.603738021850586,
      "step": 1420
    },
    {
      "epoch": 0.143,
      "grad_norm": 0.4610430896282196,
      "learning_rate": 8.295572723270881e-05,
      "loss": 1.5599722862243652,
      "step": 1430
    },
    {
      "epoch": 0.144,
      "grad_norm": 0.4239096939563751,
      "learning_rate": 8.290244830807493e-05,
      "loss": 1.5240245819091798,
      "step": 1440
    },
    {
      "epoch": 0.145,
      "grad_norm": 0.4612870216369629,
      "learning_rate": 8.284876065999803e-05,
      "loss": 1.5548505783081055,
      "step": 1450
    },
    {
      "epoch": 0.146,
      "grad_norm": 0.437900185585022,
      "learning_rate": 8.279466484020172e-05,
      "loss": 1.5739557266235351,
      "step": 1460
    },
    {
      "epoch": 0.147,
      "grad_norm": 0.459146648645401,
      "learning_rate": 8.274016140460427e-05,
      "loss": 1.5370497703552246,
      "step": 1470
    },
    {
      "epoch": 0.148,
      "grad_norm": 0.4295809268951416,
      "learning_rate": 8.268525091331279e-05,
      "loss": 1.5780242919921874,
      "step": 1480
    },
    {
      "epoch": 0.149,
      "grad_norm": 0.43193328380584717,
      "learning_rate": 8.262993393061756e-05,
      "loss": 1.5061514854431153,
      "step": 1490
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.44761526584625244,
      "learning_rate": 8.257421102498613e-05,
      "loss": 1.5490300178527832,
      "step": 1500
    },
    {
      "epoch": 0.151,
      "grad_norm": 0.4909959137439728,
      "learning_rate": 8.251808276905761e-05,
      "loss": 1.5348563194274902,
      "step": 1510
    },
    {
      "epoch": 0.152,
      "grad_norm": 0.45388928055763245,
      "learning_rate": 8.246154973963663e-05,
      "loss": 1.5422224044799804,
      "step": 1520
    },
    {
      "epoch": 0.153,
      "grad_norm": 0.44226154685020447,
      "learning_rate": 8.240461251768759e-05,
      "loss": 1.5589332580566406,
      "step": 1530
    },
    {
      "epoch": 0.154,
      "grad_norm": 0.45566076040267944,
      "learning_rate": 8.23472716883285e-05,
      "loss": 1.55558500289917,
      "step": 1540
    },
    {
      "epoch": 0.155,
      "grad_norm": 0.45849281549453735,
      "learning_rate": 8.228952784082511e-05,
      "loss": 1.60096435546875,
      "step": 1550
    },
    {
      "epoch": 0.156,
      "grad_norm": 0.4728466272354126,
      "learning_rate": 8.22313815685848e-05,
      "loss": 1.594934844970703,
      "step": 1560
    },
    {
      "epoch": 0.157,
      "grad_norm": 0.4469038248062134,
      "learning_rate": 8.217283346915048e-05,
      "loss": 1.553289031982422,
      "step": 1570
    },
    {
      "epoch": 0.158,
      "grad_norm": 0.48276886343955994,
      "learning_rate": 8.211388414419445e-05,
      "loss": 1.503675937652588,
      "step": 1580
    },
    {
      "epoch": 0.159,
      "grad_norm": 0.4695449769496918,
      "learning_rate": 8.205453419951223e-05,
      "loss": 1.5649538040161133,
      "step": 1590
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.43924853205680847,
      "learning_rate": 8.199478424501635e-05,
      "loss": 1.527259635925293,
      "step": 1600
    },
    {
      "epoch": 0.161,
      "grad_norm": 0.4189310073852539,
      "learning_rate": 8.193463489473004e-05,
      "loss": 1.55230655670166,
      "step": 1610
    },
    {
      "epoch": 0.162,
      "grad_norm": 0.44848498702049255,
      "learning_rate": 8.187408676678092e-05,
      "loss": 1.5449299812316895,
      "step": 1620
    },
    {
      "epoch": 0.163,
      "grad_norm": 0.4382420480251312,
      "learning_rate": 8.181314048339468e-05,
      "loss": 1.541883659362793,
      "step": 1630
    },
    {
      "epoch": 0.164,
      "grad_norm": 0.42705485224723816,
      "learning_rate": 8.175179667088869e-05,
      "loss": 1.5482391357421874,
      "step": 1640
    },
    {
      "epoch": 0.165,
      "grad_norm": 0.4264799654483795,
      "learning_rate": 8.169005595966554e-05,
      "loss": 1.562957191467285,
      "step": 1650
    },
    {
      "epoch": 0.166,
      "grad_norm": 0.46075376868247986,
      "learning_rate": 8.162791898420655e-05,
      "loss": 1.533614730834961,
      "step": 1660
    },
    {
      "epoch": 0.167,
      "grad_norm": 0.46955326199531555,
      "learning_rate": 8.156538638306531e-05,
      "loss": 1.5150230407714844,
      "step": 1670
    },
    {
      "epoch": 0.168,
      "grad_norm": 0.4578472077846527,
      "learning_rate": 8.150245879886102e-05,
      "loss": 1.579137420654297,
      "step": 1680
    },
    {
      "epoch": 0.169,
      "grad_norm": 0.42824026942253113,
      "learning_rate": 8.143913687827199e-05,
      "loss": 1.5843921661376954,
      "step": 1690
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.43932676315307617,
      "learning_rate": 8.137542127202888e-05,
      "loss": 1.605410385131836,
      "step": 1700
    },
    {
      "epoch": 0.171,
      "grad_norm": 0.4417635500431061,
      "learning_rate": 8.131131263490822e-05,
      "loss": 1.5659252166748048,
      "step": 1710
    },
    {
      "epoch": 0.172,
      "grad_norm": 0.448393315076828,
      "learning_rate": 8.124681162572536e-05,
      "loss": 1.5661985397338867,
      "step": 1720
    },
    {
      "epoch": 0.173,
      "grad_norm": 0.4424435496330261,
      "learning_rate": 8.118191890732799e-05,
      "loss": 1.5341227531433106,
      "step": 1730
    },
    {
      "epoch": 0.174,
      "grad_norm": 0.4247022867202759,
      "learning_rate": 8.111663514658923e-05,
      "loss": 1.5798539161682128,
      "step": 1740
    },
    {
      "epoch": 0.175,
      "grad_norm": 0.4436150789260864,
      "learning_rate": 8.105096101440067e-05,
      "loss": 1.5780141830444336,
      "step": 1750
    },
    {
      "epoch": 0.176,
      "grad_norm": 0.4493919312953949,
      "learning_rate": 8.098489718566567e-05,
      "loss": 1.5496795654296875,
      "step": 1760
    },
    {
      "epoch": 0.177,
      "grad_norm": 0.44591131806373596,
      "learning_rate": 8.091844433929229e-05,
      "loss": 1.5309679985046387,
      "step": 1770
    },
    {
      "epoch": 0.178,
      "grad_norm": 0.44779840111732483,
      "learning_rate": 8.085160315818632e-05,
      "loss": 1.5156972885131836,
      "step": 1780
    },
    {
      "epoch": 0.179,
      "grad_norm": 0.44151243567466736,
      "learning_rate": 8.078437432924434e-05,
      "loss": 1.5683908462524414,
      "step": 1790
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.43576961755752563,
      "learning_rate": 8.071675854334655e-05,
      "loss": 1.479705810546875,
      "step": 1800
    },
    {
      "epoch": 0.181,
      "grad_norm": 0.4513467848300934,
      "learning_rate": 8.06487564953498e-05,
      "loss": 1.553599739074707,
      "step": 1810
    },
    {
      "epoch": 0.182,
      "grad_norm": 0.4427294433116913,
      "learning_rate": 8.058036888408034e-05,
      "loss": 1.5128995895385742,
      "step": 1820
    },
    {
      "epoch": 0.183,
      "grad_norm": 0.458186537027359,
      "learning_rate": 8.051159641232666e-05,
      "loss": 1.5448317527770996,
      "step": 1830
    },
    {
      "epoch": 0.184,
      "grad_norm": 0.4836799204349518,
      "learning_rate": 8.044243978683234e-05,
      "loss": 1.5993682861328125,
      "step": 1840
    },
    {
      "epoch": 0.185,
      "grad_norm": 0.44361555576324463,
      "learning_rate": 8.03728997182887e-05,
      "loss": 1.5255595207214356,
      "step": 1850
    },
    {
      "epoch": 0.186,
      "grad_norm": 0.41420966386795044,
      "learning_rate": 8.030297692132756e-05,
      "loss": 1.582503890991211,
      "step": 1860
    },
    {
      "epoch": 0.187,
      "grad_norm": 0.44590333104133606,
      "learning_rate": 8.023267211451381e-05,
      "loss": 1.570404624938965,
      "step": 1870
    },
    {
      "epoch": 0.188,
      "grad_norm": 0.43962764739990234,
      "learning_rate": 8.016198602033813e-05,
      "loss": 1.5464605331420898,
      "step": 1880
    },
    {
      "epoch": 0.189,
      "grad_norm": 0.4597555994987488,
      "learning_rate": 8.009091936520949e-05,
      "loss": 1.4956825256347657,
      "step": 1890
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.4748958349227905,
      "learning_rate": 8.001947287944775e-05,
      "loss": 1.5619684219360352,
      "step": 1900
    },
    {
      "epoch": 0.191,
      "grad_norm": 0.44536739587783813,
      "learning_rate": 7.99476472972761e-05,
      "loss": 1.5296592712402344,
      "step": 1910
    },
    {
      "epoch": 0.192,
      "grad_norm": 0.43877002596855164,
      "learning_rate": 7.98754433568135e-05,
      "loss": 1.531735324859619,
      "step": 1920
    },
    {
      "epoch": 0.193,
      "grad_norm": 0.4474166929721832,
      "learning_rate": 7.980286180006717e-05,
      "loss": 1.5319225311279296,
      "step": 1930
    },
    {
      "epoch": 0.194,
      "grad_norm": 0.4478398859500885,
      "learning_rate": 7.972990337292492e-05,
      "loss": 1.598667812347412,
      "step": 1940
    },
    {
      "epoch": 0.195,
      "grad_norm": 0.4543236196041107,
      "learning_rate": 7.965656882514742e-05,
      "loss": 1.5022565841674804,
      "step": 1950
    },
    {
      "epoch": 0.196,
      "grad_norm": 0.4551485776901245,
      "learning_rate": 7.958285891036068e-05,
      "loss": 1.4970108032226563,
      "step": 1960
    },
    {
      "epoch": 0.197,
      "grad_norm": 0.4381597638130188,
      "learning_rate": 7.950877438604807e-05,
      "loss": 1.5610149383544922,
      "step": 1970
    },
    {
      "epoch": 0.198,
      "grad_norm": 0.47578102350234985,
      "learning_rate": 7.943431601354271e-05,
      "loss": 1.4768410682678224,
      "step": 1980
    },
    {
      "epoch": 0.199,
      "grad_norm": 0.43383491039276123,
      "learning_rate": 7.93594845580196e-05,
      "loss": 1.615484619140625,
      "step": 1990
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.4749927520751953,
      "learning_rate": 7.928428078848771e-05,
      "loss": 1.5539310455322266,
      "step": 2000
    },
    {
      "epoch": 0.201,
      "grad_norm": 0.4608052372932434,
      "learning_rate": 7.920870547778215e-05,
      "loss": 1.5318676948547363,
      "step": 2010
    },
    {
      "epoch": 0.202,
      "grad_norm": 0.45553329586982727,
      "learning_rate": 7.913275940255613e-05,
      "loss": 1.5542593002319336,
      "step": 2020
    },
    {
      "epoch": 0.203,
      "grad_norm": 0.4951193630695343,
      "learning_rate": 7.905644334327311e-05,
      "loss": 1.5068115234375,
      "step": 2030
    },
    {
      "epoch": 0.204,
      "grad_norm": 0.4648756980895996,
      "learning_rate": 7.897975808419868e-05,
      "loss": 1.5019064903259278,
      "step": 2040
    },
    {
      "epoch": 0.205,
      "grad_norm": 0.43299543857574463,
      "learning_rate": 7.890270441339251e-05,
      "loss": 1.536239242553711,
      "step": 2050
    },
    {
      "epoch": 0.206,
      "grad_norm": 0.4560064971446991,
      "learning_rate": 7.882528312270028e-05,
      "loss": 1.5249292373657226,
      "step": 2060
    },
    {
      "epoch": 0.207,
      "grad_norm": 0.4477216601371765,
      "learning_rate": 7.874749500774555e-05,
      "loss": 1.515114974975586,
      "step": 2070
    },
    {
      "epoch": 0.208,
      "grad_norm": 0.4781554341316223,
      "learning_rate": 7.866934086792156e-05,
      "loss": 1.5463275909423828,
      "step": 2080
    },
    {
      "epoch": 0.209,
      "grad_norm": 0.45784494280815125,
      "learning_rate": 7.859082150638303e-05,
      "loss": 1.5834733963012695,
      "step": 2090
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.41224566102027893,
      "learning_rate": 7.851193773003785e-05,
      "loss": 1.5228580474853515,
      "step": 2100
    },
    {
      "epoch": 0.211,
      "grad_norm": 0.4573986530303955,
      "learning_rate": 7.84326903495389e-05,
      "loss": 1.4756481170654296,
      "step": 2110
    },
    {
      "epoch": 0.212,
      "grad_norm": 0.477107971906662,
      "learning_rate": 7.835308017927563e-05,
      "loss": 1.563486671447754,
      "step": 2120
    },
    {
      "epoch": 0.213,
      "grad_norm": 0.4521276652812958,
      "learning_rate": 7.82731080373657e-05,
      "loss": 1.5458997726440429,
      "step": 2130
    },
    {
      "epoch": 0.214,
      "grad_norm": 0.4387427568435669,
      "learning_rate": 7.819277474564661e-05,
      "loss": 1.5429878234863281,
      "step": 2140
    },
    {
      "epoch": 0.215,
      "grad_norm": 0.45496708154678345,
      "learning_rate": 7.811208112966722e-05,
      "loss": 1.5660066604614258,
      "step": 2150
    },
    {
      "epoch": 0.216,
      "grad_norm": 0.5154712796211243,
      "learning_rate": 7.803102801867928e-05,
      "loss": 1.5109113693237304,
      "step": 2160
    },
    {
      "epoch": 0.217,
      "grad_norm": 0.4356984496116638,
      "learning_rate": 7.794961624562892e-05,
      "loss": 1.5474672317504883,
      "step": 2170
    },
    {
      "epoch": 0.218,
      "grad_norm": 0.4773666560649872,
      "learning_rate": 7.786784664714808e-05,
      "loss": 1.52636775970459,
      "step": 2180
    },
    {
      "epoch": 0.219,
      "grad_norm": 0.4520673453807831,
      "learning_rate": 7.77857200635459e-05,
      "loss": 1.4999986648559571,
      "step": 2190
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.45495954155921936,
      "learning_rate": 7.770323733880007e-05,
      "loss": 1.5395520210266114,
      "step": 2200
    },
    {
      "epoch": 0.221,
      "grad_norm": 0.46883487701416016,
      "learning_rate": 7.762039932054823e-05,
      "loss": 1.553999900817871,
      "step": 2210
    },
    {
      "epoch": 0.222,
      "grad_norm": 0.4731461703777313,
      "learning_rate": 7.753720686007916e-05,
      "loss": 1.5087655067443848,
      "step": 2220
    },
    {
      "epoch": 0.223,
      "grad_norm": 0.43786755204200745,
      "learning_rate": 7.745366081232414e-05,
      "loss": 1.5776697158813477,
      "step": 2230
    },
    {
      "epoch": 0.224,
      "grad_norm": 0.44203466176986694,
      "learning_rate": 7.736976203584805e-05,
      "loss": 1.5327131271362304,
      "step": 2240
    },
    {
      "epoch": 0.225,
      "grad_norm": 0.4511767625808716,
      "learning_rate": 7.728551139284063e-05,
      "loss": 1.5300714492797851,
      "step": 2250
    },
    {
      "epoch": 0.226,
      "grad_norm": 0.4910801649093628,
      "learning_rate": 7.720090974910761e-05,
      "loss": 1.4961822509765625,
      "step": 2260
    },
    {
      "epoch": 0.227,
      "grad_norm": 0.46052250266075134,
      "learning_rate": 7.711595797406177e-05,
      "loss": 1.5509103775024413,
      "step": 2270
    },
    {
      "epoch": 0.228,
      "grad_norm": 0.45214343070983887,
      "learning_rate": 7.703065694071401e-05,
      "loss": 1.528483772277832,
      "step": 2280
    },
    {
      "epoch": 0.229,
      "grad_norm": 0.46709343791007996,
      "learning_rate": 7.694500752566447e-05,
      "loss": 1.5288267135620117,
      "step": 2290
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.45310071110725403,
      "learning_rate": 7.685901060909339e-05,
      "loss": 1.5159454345703125,
      "step": 2300
    },
    {
      "epoch": 0.231,
      "grad_norm": 0.4652255177497864,
      "learning_rate": 7.677266707475213e-05,
      "loss": 1.5301589012145995,
      "step": 2310
    },
    {
      "epoch": 0.232,
      "grad_norm": 0.45159515738487244,
      "learning_rate": 7.66859778099541e-05,
      "loss": 1.5472338676452637,
      "step": 2320
    },
    {
      "epoch": 0.233,
      "grad_norm": 0.4177556335926056,
      "learning_rate": 7.659894370556565e-05,
      "loss": 1.519093132019043,
      "step": 2330
    },
    {
      "epoch": 0.234,
      "grad_norm": 0.429817795753479,
      "learning_rate": 7.651156565599683e-05,
      "loss": 1.497427272796631,
      "step": 2340
    },
    {
      "epoch": 0.235,
      "grad_norm": 0.43000954389572144,
      "learning_rate": 7.642384455919233e-05,
      "loss": 1.4777793884277344,
      "step": 2350
    },
    {
      "epoch": 0.236,
      "grad_norm": 0.43708905577659607,
      "learning_rate": 7.63357813166221e-05,
      "loss": 1.5647817611694337,
      "step": 2360
    },
    {
      "epoch": 0.237,
      "grad_norm": 0.45687180757522583,
      "learning_rate": 7.624737683327225e-05,
      "loss": 1.4947479248046875,
      "step": 2370
    },
    {
      "epoch": 0.238,
      "grad_norm": 0.4629661440849304,
      "learning_rate": 7.615863201763559e-05,
      "loss": 1.5427149772644042,
      "step": 2380
    },
    {
      "epoch": 0.239,
      "grad_norm": 0.46899276971817017,
      "learning_rate": 7.606954778170244e-05,
      "loss": 1.5839698791503907,
      "step": 2390
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.4606948494911194,
      "learning_rate": 7.598012504095115e-05,
      "loss": 1.541771697998047,
      "step": 2400
    },
    {
      "epoch": 0.241,
      "grad_norm": 0.45212459564208984,
      "learning_rate": 7.58903647143387e-05,
      "loss": 1.5472356796264648,
      "step": 2410
    },
    {
      "epoch": 0.242,
      "grad_norm": 0.45408064126968384,
      "learning_rate": 7.58002677242914e-05,
      "loss": 1.5932246208190919,
      "step": 2420
    },
    {
      "epoch": 0.243,
      "grad_norm": 0.48358407616615295,
      "learning_rate": 7.570983499669515e-05,
      "loss": 1.5044713973999024,
      "step": 2430
    },
    {
      "epoch": 0.244,
      "grad_norm": 0.44570690393447876,
      "learning_rate": 7.561906746088618e-05,
      "loss": 1.5048210144042968,
      "step": 2440
    },
    {
      "epoch": 0.245,
      "grad_norm": 0.4519187808036804,
      "learning_rate": 7.552796604964136e-05,
      "loss": 1.523080539703369,
      "step": 2450
    },
    {
      "epoch": 0.246,
      "grad_norm": 0.44542405009269714,
      "learning_rate": 7.543653169916864e-05,
      "loss": 1.5862494468688966,
      "step": 2460
    },
    {
      "epoch": 0.247,
      "grad_norm": 0.43162643909454346,
      "learning_rate": 7.534476534909743e-05,
      "loss": 1.553419303894043,
      "step": 2470
    },
    {
      "epoch": 0.248,
      "grad_norm": 0.4649028778076172,
      "learning_rate": 7.525266794246896e-05,
      "loss": 1.549135398864746,
      "step": 2480
    },
    {
      "epoch": 0.249,
      "grad_norm": 0.454800009727478,
      "learning_rate": 7.51602404257266e-05,
      "loss": 1.5687236785888672,
      "step": 2490
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.4387585520744324,
      "learning_rate": 7.506748374870607e-05,
      "loss": 1.4585626602172852,
      "step": 2500
    },
    {
      "epoch": 0.251,
      "grad_norm": 0.44236883521080017,
      "learning_rate": 7.497439886462573e-05,
      "loss": 1.5203914642333984,
      "step": 2510
    },
    {
      "epoch": 0.252,
      "grad_norm": 0.4657570719718933,
      "learning_rate": 7.488098673007681e-05,
      "loss": 1.5481887817382813,
      "step": 2520
    },
    {
      "epoch": 0.253,
      "grad_norm": 0.4656415283679962,
      "learning_rate": 7.478724830501353e-05,
      "loss": 1.5086076736450196,
      "step": 2530
    },
    {
      "epoch": 0.254,
      "grad_norm": 0.4566405713558197,
      "learning_rate": 7.469318455274322e-05,
      "loss": 1.497421932220459,
      "step": 2540
    },
    {
      "epoch": 0.255,
      "grad_norm": 0.4292711019515991,
      "learning_rate": 7.459879643991647e-05,
      "loss": 1.5872754096984862,
      "step": 2550
    },
    {
      "epoch": 0.256,
      "grad_norm": 0.4766315519809723,
      "learning_rate": 7.450408493651717e-05,
      "loss": 1.4764188766479491,
      "step": 2560
    },
    {
      "epoch": 0.257,
      "grad_norm": 0.4654238820075989,
      "learning_rate": 7.440905101585258e-05,
      "loss": 1.5425111770629882,
      "step": 2570
    },
    {
      "epoch": 1.0003,
      "grad_norm": 0.42733854055404663,
      "learning_rate": 7.431369565454328e-05,
      "loss": 1.7178598403930665,
      "step": 2580
    },
    {
      "epoch": 1.0013,
      "grad_norm": 0.4512520730495453,
      "learning_rate": 7.421801983251313e-05,
      "loss": 1.434187126159668,
      "step": 2590
    },
    {
      "epoch": 1.0023,
      "grad_norm": 0.45792531967163086,
      "learning_rate": 7.412202453297922e-05,
      "loss": 1.5031817436218262,
      "step": 2600
    },
    {
      "epoch": 1.0033,
      "grad_norm": 0.4429328739643097,
      "learning_rate": 7.402571074244179e-05,
      "loss": 1.4843154907226563,
      "step": 2610
    },
    {
      "epoch": 1.0043,
      "grad_norm": 0.43615370988845825,
      "learning_rate": 7.392907945067407e-05,
      "loss": 1.4864605903625487,
      "step": 2620
    },
    {
      "epoch": 1.0053,
      "grad_norm": 0.469474196434021,
      "learning_rate": 7.383213165071205e-05,
      "loss": 1.4709732055664062,
      "step": 2630
    },
    {
      "epoch": 1.0063,
      "grad_norm": 0.45088157057762146,
      "learning_rate": 7.373486833884443e-05,
      "loss": 1.4104438781738282,
      "step": 2640
    },
    {
      "epoch": 1.0073,
      "grad_norm": 0.458600252866745,
      "learning_rate": 7.363729051460219e-05,
      "loss": 1.4738828659057617,
      "step": 2650
    },
    {
      "epoch": 1.0083,
      "grad_norm": 0.46691879630088806,
      "learning_rate": 7.353939918074846e-05,
      "loss": 1.4707263946533202,
      "step": 2660
    },
    {
      "epoch": 1.0093,
      "grad_norm": 0.46489620208740234,
      "learning_rate": 7.344119534326815e-05,
      "loss": 1.3846729278564454,
      "step": 2670
    },
    {
      "epoch": 1.0103,
      "grad_norm": 0.44752174615859985,
      "learning_rate": 7.334268001135764e-05,
      "loss": 1.4604171752929687,
      "step": 2680
    },
    {
      "epoch": 1.0113,
      "grad_norm": 0.4594590365886688,
      "learning_rate": 7.324385419741438e-05,
      "loss": 1.4276948928833009,
      "step": 2690
    },
    {
      "epoch": 1.0123,
      "grad_norm": 0.4594690501689911,
      "learning_rate": 7.314471891702651e-05,
      "loss": 1.3739115715026855,
      "step": 2700
    },
    {
      "epoch": 1.0133,
      "grad_norm": 0.4549918472766876,
      "learning_rate": 7.304527518896245e-05,
      "loss": 1.4289373397827148,
      "step": 2710
    },
    {
      "epoch": 1.0143,
      "grad_norm": 0.46599602699279785,
      "learning_rate": 7.294552403516034e-05,
      "loss": 1.447694969177246,
      "step": 2720
    },
    {
      "epoch": 1.0153,
      "grad_norm": 0.4508320987224579,
      "learning_rate": 7.284546648071763e-05,
      "loss": 1.4010438919067383,
      "step": 2730
    },
    {
      "epoch": 1.0163,
      "grad_norm": 0.4645399749279022,
      "learning_rate": 7.274510355388049e-05,
      "loss": 1.3724050521850586,
      "step": 2740
    },
    {
      "epoch": 1.0173,
      "grad_norm": 0.5009335875511169,
      "learning_rate": 7.26444362860333e-05,
      "loss": 1.390232467651367,
      "step": 2750
    },
    {
      "epoch": 1.0183,
      "grad_norm": 0.4698573648929596,
      "learning_rate": 7.254346571168798e-05,
      "loss": 1.3995256423950195,
      "step": 2760
    },
    {
      "epoch": 1.0193,
      "grad_norm": 0.46238020062446594,
      "learning_rate": 7.24421928684734e-05,
      "loss": 1.3738612174987792,
      "step": 2770
    },
    {
      "epoch": 1.0203,
      "grad_norm": 0.45451462268829346,
      "learning_rate": 7.234061879712472e-05,
      "loss": 1.3291207313537599,
      "step": 2780
    },
    {
      "epoch": 1.0213,
      "grad_norm": 0.47129520773887634,
      "learning_rate": 7.223874454147269e-05,
      "loss": 1.356761646270752,
      "step": 2790
    },
    {
      "epoch": 1.0223,
      "grad_norm": 0.4773348569869995,
      "learning_rate": 7.213657114843287e-05,
      "loss": 1.3456995010375976,
      "step": 2800
    }
  ],
  "logging_steps": 10,
  "max_steps": 10000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9223372036854775807,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 3.7514887942478234e+17,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}