{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 4.87012987012987,
  "eval_steps": 500,
  "global_step": 6000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.008116883116883116,
      "grad_norm": 9.878928184509277,
      "learning_rate": 3e-06,
      "loss": 0.7789,
      "step": 10
    },
    {
      "epoch": 0.016233766233766232,
      "grad_norm": 2.0204501152038574,
      "learning_rate": 6.333333333333334e-06,
      "loss": 0.448,
      "step": 20
    },
    {
      "epoch": 0.024350649350649352,
      "grad_norm": 1.989099383354187,
      "learning_rate": 9.666666666666667e-06,
      "loss": 0.2774,
      "step": 30
    },
    {
      "epoch": 0.032467532467532464,
      "grad_norm": 2.263881206512451,
      "learning_rate": 1.3000000000000001e-05,
      "loss": 0.2339,
      "step": 40
    },
    {
      "epoch": 0.040584415584415584,
      "grad_norm": 1.8205904960632324,
      "learning_rate": 1.6333333333333335e-05,
      "loss": 0.2169,
      "step": 50
    },
    {
      "epoch": 0.048701298701298704,
      "grad_norm": 0.9809542298316956,
      "learning_rate": 1.9666666666666666e-05,
      "loss": 0.1856,
      "step": 60
    },
    {
      "epoch": 0.056818181818181816,
      "grad_norm": 1.3591933250427246,
      "learning_rate": 2.3000000000000003e-05,
      "loss": 0.1674,
      "step": 70
    },
    {
      "epoch": 0.06493506493506493,
      "grad_norm": 1.483484148979187,
      "learning_rate": 2.633333333333333e-05,
      "loss": 0.1764,
      "step": 80
    },
    {
      "epoch": 0.07305194805194805,
      "grad_norm": 0.9977609515190125,
      "learning_rate": 2.9666666666666672e-05,
      "loss": 0.1497,
      "step": 90
    },
    {
      "epoch": 0.08116883116883117,
      "grad_norm": 1.1285325288772583,
      "learning_rate": 3.3e-05,
      "loss": 0.154,
      "step": 100
    },
    {
      "epoch": 0.08928571428571429,
      "grad_norm": 1.2136211395263672,
      "learning_rate": 3.633333333333333e-05,
      "loss": 0.1436,
      "step": 110
    },
    {
      "epoch": 0.09740259740259741,
      "grad_norm": 1.2199478149414062,
      "learning_rate": 3.966666666666667e-05,
      "loss": 0.1325,
      "step": 120
    },
    {
      "epoch": 0.10551948051948051,
      "grad_norm": 0.7286880016326904,
      "learning_rate": 4.3e-05,
      "loss": 0.1254,
      "step": 130
    },
    {
      "epoch": 0.11363636363636363,
      "grad_norm": 1.066806435585022,
      "learning_rate": 4.633333333333333e-05,
      "loss": 0.1234,
      "step": 140
    },
    {
      "epoch": 0.12175324675324675,
      "grad_norm": 1.0203993320465088,
      "learning_rate": 4.966666666666667e-05,
      "loss": 0.1119,
      "step": 150
    },
    {
      "epoch": 0.12987012987012986,
      "grad_norm": 0.610676109790802,
      "learning_rate": 5.300000000000001e-05,
      "loss": 0.1203,
      "step": 160
    },
    {
      "epoch": 0.137987012987013,
      "grad_norm": 1.528229832649231,
      "learning_rate": 5.633333333333334e-05,
      "loss": 0.1168,
      "step": 170
    },
    {
      "epoch": 0.1461038961038961,
      "grad_norm": 1.8649646043777466,
      "learning_rate": 5.966666666666667e-05,
      "loss": 0.1265,
      "step": 180
    },
    {
      "epoch": 0.15422077922077923,
      "grad_norm": 0.7550538182258606,
      "learning_rate": 6.3e-05,
      "loss": 0.114,
      "step": 190
    },
    {
      "epoch": 0.16233766233766234,
      "grad_norm": 0.9938152432441711,
      "learning_rate": 6.633333333333334e-05,
      "loss": 0.1141,
      "step": 200
    },
    {
      "epoch": 0.17045454545454544,
      "grad_norm": 1.0020602941513062,
      "learning_rate": 6.966666666666668e-05,
      "loss": 0.1241,
      "step": 210
    },
    {
      "epoch": 0.17857142857142858,
      "grad_norm": 0.8025875091552734,
      "learning_rate": 7.3e-05,
      "loss": 0.1023,
      "step": 220
    },
    {
      "epoch": 0.18668831168831168,
      "grad_norm": 0.9334425926208496,
      "learning_rate": 7.633333333333334e-05,
      "loss": 0.0927,
      "step": 230
    },
    {
      "epoch": 0.19480519480519481,
      "grad_norm": 0.9842389225959778,
      "learning_rate": 7.966666666666666e-05,
      "loss": 0.1011,
      "step": 240
    },
    {
      "epoch": 0.20292207792207792,
      "grad_norm": 1.6969271898269653,
      "learning_rate": 8.3e-05,
      "loss": 0.0972,
      "step": 250
    },
    {
      "epoch": 0.21103896103896103,
      "grad_norm": 1.097373127937317,
      "learning_rate": 8.633333333333334e-05,
      "loss": 0.0949,
      "step": 260
    },
    {
      "epoch": 0.21915584415584416,
      "grad_norm": 1.2105375528335571,
      "learning_rate": 8.966666666666666e-05,
      "loss": 0.103,
      "step": 270
    },
    {
      "epoch": 0.22727272727272727,
      "grad_norm": 0.764797568321228,
      "learning_rate": 9.300000000000001e-05,
      "loss": 0.0957,
      "step": 280
    },
    {
      "epoch": 0.2353896103896104,
      "grad_norm": 1.3966500759124756,
      "learning_rate": 9.633333333333335e-05,
      "loss": 0.099,
      "step": 290
    },
    {
      "epoch": 0.2435064935064935,
      "grad_norm": 0.8691798448562622,
      "learning_rate": 9.966666666666667e-05,
      "loss": 0.0871,
      "step": 300
    },
    {
      "epoch": 0.25162337662337664,
      "grad_norm": 0.8653181791305542,
      "learning_rate": 9.999938485971279e-05,
      "loss": 0.094,
      "step": 310
    },
    {
      "epoch": 0.2597402597402597,
      "grad_norm": 0.867095410823822,
      "learning_rate": 9.999725846827562e-05,
      "loss": 0.081,
      "step": 320
    },
    {
      "epoch": 0.26785714285714285,
      "grad_norm": 0.7629588842391968,
      "learning_rate": 9.999361329594254e-05,
      "loss": 0.081,
      "step": 330
    },
    {
      "epoch": 0.275974025974026,
      "grad_norm": 1.119461178779602,
      "learning_rate": 9.998844945344405e-05,
      "loss": 0.0746,
      "step": 340
    },
    {
      "epoch": 0.2840909090909091,
      "grad_norm": 0.7320038676261902,
      "learning_rate": 9.99817670976436e-05,
      "loss": 0.0798,
      "step": 350
    },
    {
      "epoch": 0.2922077922077922,
      "grad_norm": 0.8777246475219727,
      "learning_rate": 9.997356643153303e-05,
      "loss": 0.0713,
      "step": 360
    },
    {
      "epoch": 0.3003246753246753,
      "grad_norm": 0.7135207653045654,
      "learning_rate": 9.996384770422629e-05,
      "loss": 0.0754,
      "step": 370
    },
    {
      "epoch": 0.30844155844155846,
      "grad_norm": 0.739890456199646,
      "learning_rate": 9.995261121095194e-05,
      "loss": 0.0725,
      "step": 380
    },
    {
      "epoch": 0.31655844155844154,
      "grad_norm": 0.720205545425415,
      "learning_rate": 9.993985729304408e-05,
      "loss": 0.0734,
      "step": 390
    },
    {
      "epoch": 0.3246753246753247,
      "grad_norm": 0.7397477626800537,
      "learning_rate": 9.992558633793212e-05,
      "loss": 0.0713,
      "step": 400
    },
    {
      "epoch": 0.3327922077922078,
      "grad_norm": 0.46367812156677246,
      "learning_rate": 9.990979877912891e-05,
      "loss": 0.0742,
      "step": 410
    },
    {
      "epoch": 0.3409090909090909,
      "grad_norm": 0.6461692452430725,
      "learning_rate": 9.989249509621759e-05,
      "loss": 0.0634,
      "step": 420
    },
    {
      "epoch": 0.349025974025974,
      "grad_norm": 0.5224289894104004,
      "learning_rate": 9.987367581483705e-05,
      "loss": 0.0647,
      "step": 430
    },
    {
      "epoch": 0.35714285714285715,
      "grad_norm": 0.4992559552192688,
      "learning_rate": 9.985334150666592e-05,
      "loss": 0.0635,
      "step": 440
    },
    {
      "epoch": 0.3652597402597403,
      "grad_norm": 0.728114902973175,
      "learning_rate": 9.983149278940526e-05,
      "loss": 0.0632,
      "step": 450
    },
    {
      "epoch": 0.37337662337662336,
      "grad_norm": 0.6755104660987854,
      "learning_rate": 9.980813032675974e-05,
      "loss": 0.0696,
      "step": 460
    },
    {
      "epoch": 0.3814935064935065,
      "grad_norm": 0.4159291982650757,
      "learning_rate": 9.978325482841753e-05,
      "loss": 0.0676,
      "step": 470
    },
    {
      "epoch": 0.38961038961038963,
      "grad_norm": 0.5511267185211182,
      "learning_rate": 9.975686705002867e-05,
      "loss": 0.0654,
      "step": 480
    },
    {
      "epoch": 0.3977272727272727,
      "grad_norm": 0.6259632110595703,
      "learning_rate": 9.972896779318219e-05,
      "loss": 0.0625,
      "step": 490
    },
    {
      "epoch": 0.40584415584415584,
      "grad_norm": 0.6669071316719055,
      "learning_rate": 9.969955790538175e-05,
      "loss": 0.0609,
      "step": 500
    },
    {
      "epoch": 0.413961038961039,
      "grad_norm": 0.4925576150417328,
      "learning_rate": 9.966863828001982e-05,
      "loss": 0.0618,
      "step": 510
    },
    {
      "epoch": 0.42207792207792205,
      "grad_norm": 0.513395369052887,
      "learning_rate": 9.963620985635065e-05,
      "loss": 0.0606,
      "step": 520
    },
    {
      "epoch": 0.4301948051948052,
      "grad_norm": 0.4852656424045563,
      "learning_rate": 9.960227361946164e-05,
      "loss": 0.0599,
      "step": 530
    },
    {
      "epoch": 0.4383116883116883,
      "grad_norm": 1.118783950805664,
      "learning_rate": 9.95668306002435e-05,
      "loss": 0.06,
      "step": 540
    },
    {
      "epoch": 0.44642857142857145,
      "grad_norm": 0.5577962398529053,
      "learning_rate": 9.952988187535886e-05,
      "loss": 0.056,
      "step": 550
    },
    {
      "epoch": 0.45454545454545453,
      "grad_norm": 0.9623023271560669,
      "learning_rate": 9.949142856720961e-05,
      "loss": 0.0582,
      "step": 560
    },
    {
      "epoch": 0.46266233766233766,
      "grad_norm": 0.5731148719787598,
      "learning_rate": 9.945147184390278e-05,
      "loss": 0.0515,
      "step": 570
    },
    {
      "epoch": 0.4707792207792208,
      "grad_norm": 0.5436367392539978,
      "learning_rate": 9.941001291921512e-05,
      "loss": 0.0562,
      "step": 580
    },
    {
      "epoch": 0.4788961038961039,
      "grad_norm": 0.5028620958328247,
      "learning_rate": 9.936705305255612e-05,
      "loss": 0.0587,
      "step": 590
    },
    {
      "epoch": 0.487012987012987,
      "grad_norm": 0.6722061038017273,
      "learning_rate": 9.932259354892984e-05,
      "loss": 0.0621,
      "step": 600
    },
    {
      "epoch": 0.49512987012987014,
      "grad_norm": 0.4234634339809418,
      "learning_rate": 9.927663575889521e-05,
      "loss": 0.0656,
      "step": 610
    },
    {
      "epoch": 0.5032467532467533,
      "grad_norm": 0.4400225877761841,
      "learning_rate": 9.922918107852504e-05,
      "loss": 0.0575,
      "step": 620
    },
    {
      "epoch": 0.5113636363636364,
      "grad_norm": 0.5401639342308044,
      "learning_rate": 9.918023094936363e-05,
      "loss": 0.0534,
      "step": 630
    },
    {
      "epoch": 0.5194805194805194,
      "grad_norm": 0.471709281206131,
      "learning_rate": 9.912978685838294e-05,
      "loss": 0.0524,
      "step": 640
    },
    {
      "epoch": 0.5275974025974026,
      "grad_norm": 0.6473276615142822,
      "learning_rate": 9.90778503379374e-05,
      "loss": 0.0559,
      "step": 650
    },
    {
      "epoch": 0.5357142857142857,
      "grad_norm": 0.7314488887786865,
      "learning_rate": 9.902442296571743e-05,
      "loss": 0.0582,
      "step": 660
    },
    {
      "epoch": 0.5438311688311688,
      "grad_norm": 0.528820276260376,
      "learning_rate": 9.896950636470147e-05,
      "loss": 0.0624,
      "step": 670
    },
    {
      "epoch": 0.551948051948052,
      "grad_norm": 0.5930554866790771,
      "learning_rate": 9.891310220310666e-05,
      "loss": 0.0617,
      "step": 680
    },
    {
      "epoch": 0.560064935064935,
      "grad_norm": 0.5521166324615479,
      "learning_rate": 9.885521219433823e-05,
      "loss": 0.0552,
      "step": 690
    },
    {
      "epoch": 0.5681818181818182,
      "grad_norm": 0.5110708475112915,
      "learning_rate": 9.879583809693738e-05,
      "loss": 0.0561,
      "step": 700
    },
    {
      "epoch": 0.5762987012987013,
      "grad_norm": 0.5363431572914124,
      "learning_rate": 9.873498171452789e-05,
      "loss": 0.0667,
      "step": 710
    },
    {
      "epoch": 0.5844155844155844,
      "grad_norm": 0.4858008921146393,
      "learning_rate": 9.867264489576135e-05,
      "loss": 0.0526,
      "step": 720
    },
    {
      "epoch": 0.5925324675324676,
      "grad_norm": 0.49196329712867737,
      "learning_rate": 9.860882953426099e-05,
      "loss": 0.0561,
      "step": 730
    },
    {
      "epoch": 0.6006493506493507,
      "grad_norm": 0.42068248987197876,
      "learning_rate": 9.854353756856412e-05,
      "loss": 0.048,
      "step": 740
    },
    {
      "epoch": 0.6087662337662337,
      "grad_norm": 0.6270432472229004,
      "learning_rate": 9.847677098206332e-05,
      "loss": 0.0525,
      "step": 750
    },
    {
      "epoch": 0.6168831168831169,
      "grad_norm": 0.6490077376365662,
      "learning_rate": 9.840853180294608e-05,
      "loss": 0.0517,
      "step": 760
    },
    {
      "epoch": 0.625,
      "grad_norm": 0.3505368232727051,
      "learning_rate": 9.833882210413332e-05,
      "loss": 0.0495,
      "step": 770
    },
    {
      "epoch": 0.6331168831168831,
      "grad_norm": 0.4244323968887329,
      "learning_rate": 9.826764400321633e-05,
      "loss": 0.0501,
      "step": 780
    },
    {
      "epoch": 0.6412337662337663,
      "grad_norm": 0.6171508431434631,
      "learning_rate": 9.819499966239243e-05,
      "loss": 0.0521,
      "step": 790
    },
    {
      "epoch": 0.6493506493506493,
      "grad_norm": 0.49943363666534424,
      "learning_rate": 9.812089128839938e-05,
      "loss": 0.0591,
      "step": 800
    },
    {
      "epoch": 0.6574675324675324,
      "grad_norm": 0.3718108534812927,
      "learning_rate": 9.804532113244828e-05,
      "loss": 0.0552,
      "step": 810
    },
    {
      "epoch": 0.6655844155844156,
      "grad_norm": 0.4114764928817749,
      "learning_rate": 9.796829149015517e-05,
      "loss": 0.0519,
      "step": 820
    },
    {
      "epoch": 0.6737012987012987,
      "grad_norm": 0.794797956943512,
      "learning_rate": 9.788980470147132e-05,
      "loss": 0.048,
      "step": 830
    },
    {
      "epoch": 0.6818181818181818,
      "grad_norm": 0.5907167792320251,
      "learning_rate": 9.780986315061218e-05,
      "loss": 0.0468,
      "step": 840
    },
    {
      "epoch": 0.689935064935065,
      "grad_norm": 0.40095576643943787,
      "learning_rate": 9.772846926598491e-05,
      "loss": 0.0544,
      "step": 850
    },
    {
      "epoch": 0.698051948051948,
      "grad_norm": 0.6045254468917847,
      "learning_rate": 9.76456255201146e-05,
      "loss": 0.0527,
      "step": 860
    },
    {
      "epoch": 0.7061688311688312,
      "grad_norm": 0.33633312582969666,
      "learning_rate": 9.756133442956923e-05,
      "loss": 0.0512,
      "step": 870
    },
    {
      "epoch": 0.7142857142857143,
      "grad_norm": 0.5168439149856567,
      "learning_rate": 9.747559855488313e-05,
      "loss": 0.043,
      "step": 880
    },
    {
      "epoch": 0.7224025974025974,
      "grad_norm": 0.5217397212982178,
      "learning_rate": 9.73884205004793e-05,
      "loss": 0.0478,
      "step": 890
    },
    {
      "epoch": 0.7305194805194806,
      "grad_norm": 0.4128414988517761,
      "learning_rate": 9.729980291459019e-05,
      "loss": 0.054,
      "step": 900
    },
    {
      "epoch": 0.7386363636363636,
      "grad_norm": 0.6482959985733032,
      "learning_rate": 9.720974848917735e-05,
      "loss": 0.0456,
      "step": 910
    },
    {
      "epoch": 0.7467532467532467,
      "grad_norm": 0.6063295602798462,
      "learning_rate": 9.711825995984957e-05,
      "loss": 0.0499,
      "step": 920
    },
    {
      "epoch": 0.7548701298701299,
      "grad_norm": 0.6088147163391113,
      "learning_rate": 9.702534010577991e-05,
      "loss": 0.0456,
      "step": 930
    },
    {
      "epoch": 0.762987012987013,
      "grad_norm": 0.6890689730644226,
      "learning_rate": 9.693099174962103e-05,
      "loss": 0.0439,
      "step": 940
    },
    {
      "epoch": 0.7711038961038961,
      "grad_norm": 0.5384606122970581,
      "learning_rate": 9.683521775741977e-05,
      "loss": 0.0548,
      "step": 950
    },
    {
      "epoch": 0.7792207792207793,
      "grad_norm": 0.7219799160957336,
      "learning_rate": 9.673802103852979e-05,
      "loss": 0.0467,
      "step": 960
    },
    {
      "epoch": 0.7873376623376623,
      "grad_norm": 0.5432652831077576,
      "learning_rate": 9.663940454552342e-05,
      "loss": 0.0479,
      "step": 970
    },
    {
      "epoch": 0.7954545454545454,
      "grad_norm": 0.5412060618400574,
      "learning_rate": 9.65393712741018e-05,
      "loss": 0.0552,
      "step": 980
    },
    {
      "epoch": 0.8035714285714286,
      "grad_norm": 0.506218433380127,
      "learning_rate": 9.6437924263004e-05,
      "loss": 0.0517,
      "step": 990
    },
    {
      "epoch": 0.8116883116883117,
      "grad_norm": 0.4060725271701813,
      "learning_rate": 9.63350665939146e-05,
      "loss": 0.0486,
      "step": 1000
    },
    {
      "epoch": 0.8198051948051948,
      "grad_norm": 0.5219970941543579,
      "learning_rate": 9.623080139137023e-05,
      "loss": 0.0441,
      "step": 1010
    },
    {
      "epoch": 0.827922077922078,
      "grad_norm": 0.6814868450164795,
      "learning_rate": 9.612513182266447e-05,
      "loss": 0.0437,
      "step": 1020
    },
    {
      "epoch": 0.836038961038961,
      "grad_norm": 0.5724406242370605,
      "learning_rate": 9.601806109775179e-05,
      "loss": 0.0445,
      "step": 1030
    },
    {
      "epoch": 0.8441558441558441,
      "grad_norm": 0.5323967337608337,
      "learning_rate": 9.590959246914995e-05,
      "loss": 0.0492,
      "step": 1040
    },
    {
      "epoch": 0.8522727272727273,
      "grad_norm": 0.4202612340450287,
      "learning_rate": 9.579972923184122e-05,
      "loss": 0.0404,
      "step": 1050
    },
    {
      "epoch": 0.8603896103896104,
      "grad_norm": 0.5512884259223938,
      "learning_rate": 9.568847472317232e-05,
      "loss": 0.0521,
      "step": 1060
    },
    {
      "epoch": 0.8685064935064936,
      "grad_norm": 0.326740562915802,
      "learning_rate": 9.557583232275303e-05,
      "loss": 0.0513,
      "step": 1070
    },
    {
      "epoch": 0.8766233766233766,
      "grad_norm": 0.49678659439086914,
      "learning_rate": 9.546180545235344e-05,
      "loss": 0.0458,
      "step": 1080
    },
    {
      "epoch": 0.8847402597402597,
      "grad_norm": 0.4488849937915802,
      "learning_rate": 9.534639757580013e-05,
      "loss": 0.0361,
      "step": 1090
    },
    {
      "epoch": 0.8928571428571429,
      "grad_norm": 0.3409590721130371,
      "learning_rate": 9.522961219887092e-05,
      "loss": 0.041,
      "step": 1100
    },
    {
      "epoch": 0.900974025974026,
      "grad_norm": 0.3857138156890869,
      "learning_rate": 9.511145286918828e-05,
      "loss": 0.044,
      "step": 1110
    },
    {
      "epoch": 0.9090909090909091,
      "grad_norm": 0.48027417063713074,
      "learning_rate": 9.499192317611167e-05,
      "loss": 0.0438,
      "step": 1120
    },
    {
      "epoch": 0.9172077922077922,
      "grad_norm": 0.26468542218208313,
      "learning_rate": 9.487102675062851e-05,
      "loss": 0.0495,
      "step": 1130
    },
    {
      "epoch": 0.9253246753246753,
      "grad_norm": 0.5093517899513245,
      "learning_rate": 9.474876726524374e-05,
      "loss": 0.0468,
      "step": 1140
    },
    {
      "epoch": 0.9334415584415584,
      "grad_norm": 0.4471321105957031,
      "learning_rate": 9.462514843386845e-05,
      "loss": 0.0451,
      "step": 1150
    },
    {
      "epoch": 0.9415584415584416,
      "grad_norm": 0.3578501343727112,
      "learning_rate": 9.450017401170689e-05,
      "loss": 0.0404,
      "step": 1160
    },
    {
      "epoch": 0.9496753246753247,
      "grad_norm": 0.46056661009788513,
      "learning_rate": 9.437384779514256e-05,
      "loss": 0.0419,
      "step": 1170
    },
    {
      "epoch": 0.9577922077922078,
      "grad_norm": 0.4485381841659546,
      "learning_rate": 9.424617362162271e-05,
      "loss": 0.0435,
      "step": 1180
    },
    {
      "epoch": 0.9659090909090909,
      "grad_norm": 0.4065086245536804,
      "learning_rate": 9.411715536954196e-05,
      "loss": 0.0422,
      "step": 1190
    },
    {
      "epoch": 0.974025974025974,
      "grad_norm": 0.49561673402786255,
      "learning_rate": 9.39867969581243e-05,
      "loss": 0.0435,
      "step": 1200
    },
    {
      "epoch": 0.9821428571428571,
      "grad_norm": 0.3293704390525818,
      "learning_rate": 9.385510234730415e-05,
      "loss": 0.044,
      "step": 1210
    },
    {
      "epoch": 0.9902597402597403,
      "grad_norm": 0.5016496181488037,
      "learning_rate": 9.372207553760603e-05,
      "loss": 0.0367,
      "step": 1220
    },
    {
      "epoch": 0.9983766233766234,
      "grad_norm": 0.5729497075080872,
      "learning_rate": 9.358772057002312e-05,
      "loss": 0.0478,
      "step": 1230
    },
    {
      "epoch": 1.0064935064935066,
      "grad_norm": 0.6327930092811584,
      "learning_rate": 9.345204152589428e-05,
      "loss": 0.0412,
      "step": 1240
    },
    {
      "epoch": 1.0146103896103895,
      "grad_norm": 0.40651750564575195,
      "learning_rate": 9.331504252678037e-05,
      "loss": 0.0459,
      "step": 1250
    },
    {
      "epoch": 1.0227272727272727,
      "grad_norm": 0.5167247653007507,
      "learning_rate": 9.317672773433876e-05,
      "loss": 0.0454,
      "step": 1260
    },
    {
      "epoch": 1.030844155844156,
      "grad_norm": 0.48275452852249146,
      "learning_rate": 9.30371013501972e-05,
      "loss": 0.0459,
      "step": 1270
    },
    {
      "epoch": 1.0389610389610389,
      "grad_norm": 0.48492079973220825,
      "learning_rate": 9.289616761582587e-05,
      "loss": 0.0391,
      "step": 1280
    },
    {
      "epoch": 1.047077922077922,
      "grad_norm": 0.5485243797302246,
      "learning_rate": 9.275393081240882e-05,
      "loss": 0.0525,
      "step": 1290
    },
    {
      "epoch": 1.0551948051948052,
      "grad_norm": 0.5031293034553528,
      "learning_rate": 9.261039526071374e-05,
      "loss": 0.0462,
      "step": 1300
    },
    {
      "epoch": 1.0633116883116882,
      "grad_norm": 0.5791725516319275,
      "learning_rate": 9.246556532096078e-05,
      "loss": 0.0483,
      "step": 1310
    },
    {
      "epoch": 1.0714285714285714,
      "grad_norm": 0.4918902814388275,
      "learning_rate": 9.231944539269009e-05,
      "loss": 0.0463,
      "step": 1320
    },
    {
      "epoch": 1.0795454545454546,
      "grad_norm": 0.5383262634277344,
      "learning_rate": 9.217203991462815e-05,
      "loss": 0.0408,
      "step": 1330
    },
    {
      "epoch": 1.0876623376623376,
      "grad_norm": 0.6195541620254517,
      "learning_rate": 9.202335336455296e-05,
      "loss": 0.0415,
      "step": 1340
    },
    {
      "epoch": 1.0957792207792207,
      "grad_norm": 0.3492124080657959,
      "learning_rate": 9.187339025915802e-05,
      "loss": 0.0436,
      "step": 1350
    },
    {
      "epoch": 1.103896103896104,
      "grad_norm": 0.3001192510128021,
      "learning_rate": 9.17221551539151e-05,
      "loss": 0.0406,
      "step": 1360
    },
    {
      "epoch": 1.112012987012987,
      "grad_norm": 0.5313782095909119,
      "learning_rate": 9.156965264293586e-05,
      "loss": 0.0306,
      "step": 1370
    },
    {
      "epoch": 1.12012987012987,
      "grad_norm": 0.35381996631622314,
      "learning_rate": 9.141588735883232e-05,
      "loss": 0.0381,
      "step": 1380
    },
    {
      "epoch": 1.1282467532467533,
      "grad_norm": 0.2853216826915741,
      "learning_rate": 9.126086397257612e-05,
      "loss": 0.0366,
      "step": 1390
    },
    {
      "epoch": 1.1363636363636362,
      "grad_norm": 0.2972594201564789,
      "learning_rate": 9.110458719335659e-05,
      "loss": 0.039,
      "step": 1400
    },
    {
      "epoch": 1.1444805194805194,
      "grad_norm": 0.31505510210990906,
      "learning_rate": 9.094706176843777e-05,
      "loss": 0.0353,
      "step": 1410
    },
    {
      "epoch": 1.1525974025974026,
      "grad_norm": 0.41417738795280457,
      "learning_rate": 9.078829248301417e-05,
      "loss": 0.0424,
      "step": 1420
    },
    {
      "epoch": 1.1607142857142858,
      "grad_norm": 0.2667694687843323,
      "learning_rate": 9.062828416006539e-05,
      "loss": 0.0395,
      "step": 1430
    },
    {
      "epoch": 1.1688311688311688,
      "grad_norm": 0.3468317687511444,
      "learning_rate": 9.046704166020961e-05,
      "loss": 0.0395,
      "step": 1440
    },
    {
      "epoch": 1.176948051948052,
      "grad_norm": 0.4519212245941162,
      "learning_rate": 9.030456988155596e-05,
      "loss": 0.0421,
      "step": 1450
    },
    {
      "epoch": 1.1850649350649352,
      "grad_norm": 0.3515416979789734,
      "learning_rate": 9.014087375955573e-05,
      "loss": 0.0356,
      "step": 1460
    },
    {
      "epoch": 1.1931818181818181,
      "grad_norm": 0.42163556814193726,
      "learning_rate": 8.997595826685243e-05,
      "loss": 0.0387,
      "step": 1470
    },
    {
      "epoch": 1.2012987012987013,
      "grad_norm": 0.3507428467273712,
      "learning_rate": 8.980982841313074e-05,
      "loss": 0.0445,
      "step": 1480
    },
    {
      "epoch": 1.2094155844155845,
      "grad_norm": 0.45810309052467346,
      "learning_rate": 8.964248924496435e-05,
      "loss": 0.0336,
      "step": 1490
    },
    {
      "epoch": 1.2175324675324675,
      "grad_norm": 0.2804049253463745,
      "learning_rate": 8.947394584566258e-05,
      "loss": 0.0415,
      "step": 1500
    },
    {
      "epoch": 1.2256493506493507,
      "grad_norm": 0.2912737727165222,
      "learning_rate": 8.930420333511606e-05,
      "loss": 0.0454,
      "step": 1510
    },
    {
      "epoch": 1.2337662337662338,
      "grad_norm": 0.3981313705444336,
      "learning_rate": 8.913326686964117e-05,
      "loss": 0.0377,
      "step": 1520
    },
    {
      "epoch": 1.2418831168831168,
      "grad_norm": 0.33030176162719727,
      "learning_rate": 8.89611416418234e-05,
      "loss": 0.0449,
      "step": 1530
    },
    {
      "epoch": 1.25,
      "grad_norm": 0.47622916102409363,
      "learning_rate": 8.878783288035957e-05,
      "loss": 0.0462,
      "step": 1540
    },
    {
      "epoch": 1.2581168831168832,
      "grad_norm": 0.3368417024612427,
      "learning_rate": 8.86133458498991e-05,
      "loss": 0.0454,
      "step": 1550
    },
    {
      "epoch": 1.2662337662337662,
      "grad_norm": 0.4337632358074188,
      "learning_rate": 8.843768585088393e-05,
      "loss": 0.0376,
      "step": 1560
    },
    {
      "epoch": 1.2743506493506493,
      "grad_norm": 0.37321558594703674,
      "learning_rate": 8.82608582193877e-05,
      "loss": 0.0357,
      "step": 1570
    },
    {
      "epoch": 1.2824675324675325,
      "grad_norm": 0.4496852457523346,
      "learning_rate": 8.80828683269535e-05,
      "loss": 0.04,
      "step": 1580
    },
    {
      "epoch": 1.2905844155844157,
      "grad_norm": 0.3610592782497406,
      "learning_rate": 8.790372158043074e-05,
      "loss": 0.0394,
      "step": 1590
    },
    {
      "epoch": 1.2987012987012987,
      "grad_norm": 0.4939509928226471,
      "learning_rate": 8.772342342181095e-05,
      "loss": 0.0389,
      "step": 1600
    },
    {
      "epoch": 1.3068181818181819,
      "grad_norm": 0.32088860869407654,
      "learning_rate": 8.75419793280624e-05,
      "loss": 0.0376,
      "step": 1610
    },
    {
      "epoch": 1.314935064935065,
      "grad_norm": 0.24955856800079346,
      "learning_rate": 8.735939481096378e-05,
      "loss": 0.0337,
      "step": 1620
    },
    {
      "epoch": 1.323051948051948,
      "grad_norm": 0.2986941635608673,
      "learning_rate": 8.717567541693673e-05,
      "loss": 0.0356,
      "step": 1630
    },
    {
      "epoch": 1.3311688311688312,
      "grad_norm": 0.6089693903923035,
      "learning_rate": 8.699082672687734e-05,
      "loss": 0.0382,
      "step": 1640
    },
    {
      "epoch": 1.3392857142857144,
      "grad_norm": 0.465394526720047,
      "learning_rate": 8.680485435598673e-05,
      "loss": 0.0462,
      "step": 1650
    },
    {
      "epoch": 1.3474025974025974,
      "grad_norm": 0.2362728714942932,
      "learning_rate": 8.661776395360029e-05,
      "loss": 0.0305,
      "step": 1660
    },
    {
      "epoch": 1.3555194805194806,
      "grad_norm": 0.32173094153404236,
      "learning_rate": 8.642956120301626e-05,
      "loss": 0.0367,
      "step": 1670
    },
    {
      "epoch": 1.3636363636363638,
      "grad_norm": 0.47564932703971863,
      "learning_rate": 8.624025182132292e-05,
      "loss": 0.0421,
      "step": 1680
    },
    {
      "epoch": 1.3717532467532467,
      "grad_norm": 0.2202228456735611,
      "learning_rate": 8.604984155922506e-05,
      "loss": 0.0399,
      "step": 1690
    },
    {
      "epoch": 1.37987012987013,
      "grad_norm": 0.21159934997558594,
      "learning_rate": 8.585833620086918e-05,
      "loss": 0.0326,
      "step": 1700
    },
    {
      "epoch": 1.387987012987013,
      "grad_norm": 0.46340233087539673,
      "learning_rate": 8.566574156366784e-05,
      "loss": 0.0377,
      "step": 1710
    },
    {
      "epoch": 1.396103896103896,
      "grad_norm": 0.44417804479599,
      "learning_rate": 8.547206349812298e-05,
      "loss": 0.0399,
      "step": 1720
    },
    {
      "epoch": 1.4042207792207793,
      "grad_norm": 0.4538224935531616,
      "learning_rate": 8.527730788764805e-05,
      "loss": 0.0356,
      "step": 1730
    },
    {
      "epoch": 1.4123376623376624,
      "grad_norm": 0.45235779881477356,
      "learning_rate": 8.508148064838948e-05,
      "loss": 0.0358,
      "step": 1740
    },
    {
      "epoch": 1.4204545454545454,
      "grad_norm": 0.32103490829467773,
      "learning_rate": 8.488458772904684e-05,
      "loss": 0.0386,
      "step": 1750
    },
    {
      "epoch": 1.4285714285714286,
      "grad_norm": 0.43833354115486145,
      "learning_rate": 8.468663511069217e-05,
      "loss": 0.0429,
      "step": 1760
    },
    {
      "epoch": 1.4366883116883118,
      "grad_norm": 0.4180199205875397,
      "learning_rate": 8.448762880658825e-05,
      "loss": 0.0369,
      "step": 1770
    },
    {
      "epoch": 1.4448051948051948,
      "grad_norm": 0.41014060378074646,
      "learning_rate": 8.428757486200603e-05,
      "loss": 0.0412,
      "step": 1780
    },
    {
      "epoch": 1.452922077922078,
      "grad_norm": 0.3040854334831238,
      "learning_rate": 8.40864793540409e-05,
      "loss": 0.0412,
      "step": 1790
    },
    {
      "epoch": 1.4610389610389611,
      "grad_norm": 0.28535085916519165,
      "learning_rate": 8.388434839142813e-05,
      "loss": 0.0391,
      "step": 1800
    },
    {
      "epoch": 1.469155844155844,
      "grad_norm": 0.5234016180038452,
      "learning_rate": 8.368118811435726e-05,
      "loss": 0.0395,
      "step": 1810
    },
    {
      "epoch": 1.4772727272727273,
      "grad_norm": 0.5190858840942383,
      "learning_rate": 8.347700469428564e-05,
      "loss": 0.0356,
      "step": 1820
    },
    {
      "epoch": 1.4853896103896105,
      "grad_norm": 0.2170562446117401,
      "learning_rate": 8.327180433375091e-05,
      "loss": 0.0373,
      "step": 1830
    },
    {
      "epoch": 1.4935064935064934,
      "grad_norm": 0.3450142741203308,
      "learning_rate": 8.306559326618259e-05,
      "loss": 0.0358,
      "step": 1840
    },
    {
      "epoch": 1.5016233766233766,
      "grad_norm": 0.49485528469085693,
      "learning_rate": 8.285837775571276e-05,
      "loss": 0.0343,
      "step": 1850
    },
    {
      "epoch": 1.5097402597402598,
      "grad_norm": 0.25820446014404297,
      "learning_rate": 8.265016409698573e-05,
      "loss": 0.0391,
      "step": 1860
    },
    {
      "epoch": 1.5178571428571428,
      "grad_norm": 0.37337490916252136,
      "learning_rate": 8.244095861496686e-05,
      "loss": 0.0367,
      "step": 1870
    },
    {
      "epoch": 1.525974025974026,
      "grad_norm": 0.3256986141204834,
      "learning_rate": 8.223076766475035e-05,
      "loss": 0.0328,
      "step": 1880
    },
    {
      "epoch": 1.5340909090909092,
      "grad_norm": 0.3565908968448639,
      "learning_rate": 8.201959763136633e-05,
      "loss": 0.0291,
      "step": 1890
    },
    {
      "epoch": 1.5422077922077921,
      "grad_norm": 0.21973878145217896,
      "learning_rate": 8.180745492958674e-05,
      "loss": 0.039,
      "step": 1900
    },
    {
      "epoch": 1.5503246753246753,
      "grad_norm": 0.6989667415618896,
      "learning_rate": 8.159434600373061e-05,
      "loss": 0.0411,
      "step": 1910
    },
    {
      "epoch": 1.5584415584415585,
      "grad_norm": 0.41696980595588684,
      "learning_rate": 8.138027732746818e-05,
      "loss": 0.0421,
      "step": 1920
    },
    {
      "epoch": 1.5665584415584415,
      "grad_norm": 0.46307626366615295,
      "learning_rate": 8.116525540362434e-05,
      "loss": 0.0376,
      "step": 1930
    },
    {
      "epoch": 1.5746753246753247,
      "grad_norm": 0.4556179940700531,
      "learning_rate": 8.094928676398101e-05,
      "loss": 0.0396,
      "step": 1940
    },
    {
      "epoch": 1.5827922077922079,
      "grad_norm": 0.2597368657588959,
      "learning_rate": 8.073237796907882e-05,
      "loss": 0.0345,
      "step": 1950
    },
    {
      "epoch": 1.5909090909090908,
      "grad_norm": 0.31811609864234924,
      "learning_rate": 8.051453560801772e-05,
      "loss": 0.032,
      "step": 1960
    },
    {
      "epoch": 1.599025974025974,
      "grad_norm": 0.2722020447254181,
      "learning_rate": 8.029576629825687e-05,
      "loss": 0.0388,
      "step": 1970
    },
    {
      "epoch": 1.6071428571428572,
      "grad_norm": 0.3664605915546417,
      "learning_rate": 8.007607668541362e-05,
      "loss": 0.0389,
      "step": 1980
    },
    {
      "epoch": 1.6152597402597402,
      "grad_norm": 0.4249238967895508,
      "learning_rate": 7.985547344306161e-05,
      "loss": 0.0359,
      "step": 1990
    },
    {
      "epoch": 1.6233766233766234,
      "grad_norm": 0.27410855889320374,
      "learning_rate": 7.963396327252812e-05,
      "loss": 0.0337,
      "step": 2000
    },
    {
      "epoch": 1.6314935064935066,
      "grad_norm": 0.5478296875953674,
      "learning_rate": 7.941155290269038e-05,
      "loss": 0.0314,
      "step": 2010
    },
    {
      "epoch": 1.6396103896103895,
      "grad_norm": 0.3749033212661743,
      "learning_rate": 7.918824908977123e-05,
      "loss": 0.0356,
      "step": 2020
    },
    {
      "epoch": 1.6477272727272727,
      "grad_norm": 0.48101192712783813,
      "learning_rate": 7.896405861713394e-05,
      "loss": 0.0407,
      "step": 2030
    },
    {
      "epoch": 1.655844155844156,
      "grad_norm": 0.3538611829280853,
      "learning_rate": 7.873898829507606e-05,
      "loss": 0.0399,
      "step": 2040
    },
    {
      "epoch": 1.6639610389610389,
      "grad_norm": 0.4626486599445343,
      "learning_rate": 7.851304496062254e-05,
      "loss": 0.0346,
      "step": 2050
    },
    {
      "epoch": 1.672077922077922,
      "grad_norm": 0.41147440671920776,
      "learning_rate": 7.828623547731818e-05,
      "loss": 0.0315,
      "step": 2060
    },
    {
      "epoch": 1.6801948051948052,
      "grad_norm": 0.28702419996261597,
      "learning_rate": 7.80585667350189e-05,
      "loss": 0.0299,
      "step": 2070
    },
    {
      "epoch": 1.6883116883116882,
      "grad_norm": 0.3159998059272766,
      "learning_rate": 7.783004564968263e-05,
      "loss": 0.0334,
      "step": 2080
    },
    {
      "epoch": 1.6964285714285714,
      "grad_norm": 0.40764641761779785,
      "learning_rate": 7.760067916315921e-05,
      "loss": 0.0307,
      "step": 2090
    },
    {
      "epoch": 1.7045454545454546,
      "grad_norm": 0.314028263092041,
      "learning_rate": 7.737047424297941e-05,
      "loss": 0.0283,
      "step": 2100
    },
    {
      "epoch": 1.7126623376623376,
      "grad_norm": 0.32912203669548035,
      "learning_rate": 7.713943788214337e-05,
      "loss": 0.0339,
      "step": 2110
    },
    {
      "epoch": 1.7207792207792207,
      "grad_norm": 0.37462761998176575,
      "learning_rate": 7.690757709890812e-05,
      "loss": 0.0285,
      "step": 2120
    },
    {
      "epoch": 1.728896103896104,
      "grad_norm": 0.2835879921913147,
      "learning_rate": 7.66748989365744e-05,
      "loss": 0.0335,
      "step": 2130
    },
    {
      "epoch": 1.737012987012987,
      "grad_norm": 0.23292797803878784,
      "learning_rate": 7.644141046327271e-05,
      "loss": 0.032,
      "step": 2140
    },
    {
      "epoch": 1.74512987012987,
      "grad_norm": 0.31655260920524597,
      "learning_rate": 7.620711877174866e-05,
      "loss": 0.0354,
      "step": 2150
    },
    {
      "epoch": 1.7532467532467533,
      "grad_norm": 0.35904660820961,
      "learning_rate": 7.597203097914732e-05,
      "loss": 0.0295,
      "step": 2160
    },
    {
      "epoch": 1.7613636363636362,
      "grad_norm": 0.5082493424415588,
      "learning_rate": 7.573615422679726e-05,
      "loss": 0.0318,
      "step": 2170
    },
    {
      "epoch": 1.7694805194805194,
      "grad_norm": 0.2941649556159973,
      "learning_rate": 7.549949567999345e-05,
      "loss": 0.0299,
      "step": 2180
    },
    {
      "epoch": 1.7775974025974026,
      "grad_norm": 0.4067605435848236,
      "learning_rate": 7.526206252777968e-05,
      "loss": 0.0323,
      "step": 2190
    },
    {
      "epoch": 1.7857142857142856,
      "grad_norm": 0.4252224564552307,
      "learning_rate": 7.50238619827301e-05,
      "loss": 0.0367,
      "step": 2200
    },
    {
      "epoch": 1.7938311688311688,
      "grad_norm": 0.4501606225967407,
      "learning_rate": 7.478490128073022e-05,
      "loss": 0.0311,
      "step": 2210
    },
    {
      "epoch": 1.801948051948052,
      "grad_norm": 0.4356978237628937,
      "learning_rate": 7.454518768075704e-05,
      "loss": 0.032,
      "step": 2220
    },
    {
      "epoch": 1.810064935064935,
      "grad_norm": 0.2771468460559845,
      "learning_rate": 7.430472846465856e-05,
      "loss": 0.0365,
      "step": 2230
    },
    {
      "epoch": 1.8181818181818183,
      "grad_norm": 0.4239960312843323,
      "learning_rate": 7.406353093693253e-05,
      "loss": 0.0332,
      "step": 2240
    },
    {
      "epoch": 1.8262987012987013,
      "grad_norm": 0.2908805012702942,
      "learning_rate": 7.382160242450469e-05,
      "loss": 0.0355,
      "step": 2250
    },
    {
      "epoch": 1.8344155844155843,
      "grad_norm": 0.36092737317085266,
      "learning_rate": 7.357895027650598e-05,
      "loss": 0.0335,
      "step": 2260
    },
    {
      "epoch": 1.8425324675324677,
      "grad_norm": 0.3536132276058197,
      "learning_rate": 7.333558186404958e-05,
      "loss": 0.0305,
      "step": 2270
    },
    {
      "epoch": 1.8506493506493507,
      "grad_norm": 0.41365113854408264,
      "learning_rate": 7.309150458000668e-05,
      "loss": 0.0354,
      "step": 2280
    },
    {
      "epoch": 1.8587662337662336,
      "grad_norm": 0.37331002950668335,
      "learning_rate": 7.284672583878219e-05,
      "loss": 0.0368,
      "step": 2290
    },
    {
      "epoch": 1.866883116883117,
      "grad_norm": 0.4106990098953247,
      "learning_rate": 7.260125307608929e-05,
      "loss": 0.0271,
      "step": 2300
    },
    {
      "epoch": 1.875,
      "grad_norm": 0.46040254831314087,
      "learning_rate": 7.235509374872373e-05,
      "loss": 0.037,
      "step": 2310
    },
    {
      "epoch": 1.883116883116883,
      "grad_norm": 0.25721243023872375,
      "learning_rate": 7.210825533433719e-05,
      "loss": 0.0284,
      "step": 2320
    },
    {
      "epoch": 1.8912337662337664,
      "grad_norm": 0.37134552001953125,
      "learning_rate": 7.186074533121013e-05,
      "loss": 0.0396,
      "step": 2330
    },
    {
      "epoch": 1.8993506493506493,
      "grad_norm": 0.36696702241897583,
      "learning_rate": 7.161257125802413e-05,
      "loss": 0.0313,
      "step": 2340
    },
    {
      "epoch": 1.9074675324675323,
      "grad_norm": 0.3280735909938812,
      "learning_rate": 7.136374065363334e-05,
      "loss": 0.0356,
      "step": 2350
    },
    {
      "epoch": 1.9155844155844157,
      "grad_norm": 0.36613646149635315,
      "learning_rate": 7.11142610768356e-05,
      "loss": 0.0329,
      "step": 2360
    },
    {
      "epoch": 1.9237012987012987,
      "grad_norm": 0.5371996164321899,
      "learning_rate": 7.086414010614276e-05,
      "loss": 0.0311,
      "step": 2370
    },
    {
      "epoch": 1.9318181818181817,
      "grad_norm": 0.34445852041244507,
      "learning_rate": 7.061338533955043e-05,
      "loss": 0.0281,
      "step": 2380
    },
    {
      "epoch": 1.939935064935065,
      "grad_norm": 0.4094446003437042,
      "learning_rate": 7.036200439430725e-05,
      "loss": 0.0347,
      "step": 2390
    },
    {
      "epoch": 1.948051948051948,
      "grad_norm": 0.3943575620651245,
      "learning_rate": 7.01100049066835e-05,
      "loss": 0.0339,
      "step": 2400
    },
    {
      "epoch": 1.9561688311688312,
      "grad_norm": 0.427166223526001,
      "learning_rate": 6.985739453173903e-05,
      "loss": 0.0316,
      "step": 2410
    },
    {
      "epoch": 1.9642857142857144,
      "grad_norm": 0.21290753781795502,
      "learning_rate": 6.960418094309085e-05,
      "loss": 0.0351,
      "step": 2420
    },
    {
      "epoch": 1.9724025974025974,
      "grad_norm": 0.32666370272636414,
      "learning_rate": 6.93503718326799e-05,
      "loss": 0.0294,
      "step": 2430
    },
    {
      "epoch": 1.9805194805194806,
      "grad_norm": 0.3321411907672882,
      "learning_rate": 6.909597491053751e-05,
      "loss": 0.0332,
      "step": 2440
    },
    {
      "epoch": 1.9886363636363638,
      "grad_norm": 0.33192306756973267,
      "learning_rate": 6.884099790455113e-05,
      "loss": 0.0301,
      "step": 2450
    },
    {
      "epoch": 1.9967532467532467,
      "grad_norm": 0.5505363941192627,
      "learning_rate": 6.858544856022952e-05,
      "loss": 0.0358,
      "step": 2460
    },
    {
      "epoch": 2.0048701298701297,
      "grad_norm": 0.5219383239746094,
      "learning_rate": 6.83293346404676e-05,
      "loss": 0.0346,
      "step": 2470
    },
    {
      "epoch": 2.012987012987013,
      "grad_norm": 0.34627780318260193,
      "learning_rate": 6.80726639253105e-05,
      "loss": 0.0377,
      "step": 2480
    },
    {
      "epoch": 2.021103896103896,
      "grad_norm": 0.4243893325328827,
      "learning_rate": 6.781544421171732e-05,
      "loss": 0.0301,
      "step": 2490
    },
    {
      "epoch": 2.029220779220779,
      "grad_norm": 0.38806095719337463,
      "learning_rate": 6.755768331332424e-05,
      "loss": 0.0304,
      "step": 2500
    },
    {
      "epoch": 2.0373376623376624,
      "grad_norm": 0.30858781933784485,
      "learning_rate": 6.729938906020713e-05,
      "loss": 0.032,
      "step": 2510
    },
    {
      "epoch": 2.0454545454545454,
      "grad_norm": 0.2511473298072815,
      "learning_rate": 6.704056929864376e-05,
      "loss": 0.0257,
      "step": 2520
    },
    {
      "epoch": 2.0535714285714284,
      "grad_norm": 0.28458356857299805,
      "learning_rate": 6.67812318908754e-05,
      "loss": 0.0294,
      "step": 2530
    },
    {
      "epoch": 2.061688311688312,
      "grad_norm": 0.3707312047481537,
      "learning_rate": 6.6521384714868e-05,
      "loss": 0.0314,
      "step": 2540
    },
    {
      "epoch": 2.0698051948051948,
      "grad_norm": 0.24003523588180542,
      "learning_rate": 6.626103566407295e-05,
      "loss": 0.0338,
      "step": 2550
    },
    {
      "epoch": 2.0779220779220777,
      "grad_norm": 0.3327799439430237,
      "learning_rate": 6.600019264718713e-05,
      "loss": 0.032,
      "step": 2560
    },
    {
      "epoch": 2.086038961038961,
      "grad_norm": 0.3021881878376007,
      "learning_rate": 6.573886358791285e-05,
      "loss": 0.0295,
      "step": 2570
    },
    {
      "epoch": 2.094155844155844,
      "grad_norm": 0.30725860595703125,
      "learning_rate": 6.547705642471703e-05,
      "loss": 0.0287,
      "step": 2580
    },
    {
      "epoch": 2.102272727272727,
      "grad_norm": 0.2985770106315613,
      "learning_rate": 6.521477911059008e-05,
      "loss": 0.0289,
      "step": 2590
    },
    {
      "epoch": 2.1103896103896105,
      "grad_norm": 0.2585412859916687,
      "learning_rate": 6.495203961280434e-05,
      "loss": 0.0264,
      "step": 2600
    },
    {
      "epoch": 2.1185064935064934,
      "grad_norm": 0.3345862627029419,
      "learning_rate": 6.468884591267204e-05,
      "loss": 0.0294,
      "step": 2610
    },
    {
      "epoch": 2.1266233766233764,
      "grad_norm": 0.41437289118766785,
      "learning_rate": 6.44252060053028e-05,
      "loss": 0.027,
      "step": 2620
    },
    {
      "epoch": 2.13474025974026,
      "grad_norm": 0.25786668062210083,
      "learning_rate": 6.416112789936086e-05,
      "loss": 0.0262,
      "step": 2630
    },
    {
      "epoch": 2.142857142857143,
      "grad_norm": 0.2157658040523529,
      "learning_rate": 6.389661961682173e-05,
      "loss": 0.0244,
      "step": 2640
    },
    {
      "epoch": 2.150974025974026,
      "grad_norm": 0.3058822453022003,
      "learning_rate": 6.363168919272846e-05,
      "loss": 0.0258,
      "step": 2650
    },
    {
      "epoch": 2.159090909090909,
      "grad_norm": 0.27886053919792175,
      "learning_rate": 6.336634467494768e-05,
      "loss": 0.0363,
      "step": 2660
    },
    {
      "epoch": 2.167207792207792,
      "grad_norm": 0.3333021402359009,
      "learning_rate": 6.310059412392505e-05,
      "loss": 0.0291,
      "step": 2670
    },
    {
      "epoch": 2.175324675324675,
      "grad_norm": 0.2074045091867447,
      "learning_rate": 6.283444561244042e-05,
      "loss": 0.0315,
      "step": 2680
    },
    {
      "epoch": 2.1834415584415585,
      "grad_norm": 0.24044620990753174,
      "learning_rate": 6.256790722536251e-05,
      "loss": 0.0241,
      "step": 2690
    },
    {
      "epoch": 2.1915584415584415,
      "grad_norm": 0.3287203013896942,
      "learning_rate": 6.230098705940354e-05,
      "loss": 0.0321,
      "step": 2700
    },
    {
      "epoch": 2.199675324675325,
      "grad_norm": 0.43518200516700745,
      "learning_rate": 6.203369322287306e-05,
      "loss": 0.0283,
      "step": 2710
    },
    {
      "epoch": 2.207792207792208,
      "grad_norm": 0.3628130257129669,
      "learning_rate": 6.17660338354317e-05,
      "loss": 0.03,
      "step": 2720
    },
    {
      "epoch": 2.215909090909091,
      "grad_norm": 0.26702725887298584,
      "learning_rate": 6.149801702784456e-05,
      "loss": 0.028,
      "step": 2730
    },
    {
      "epoch": 2.224025974025974,
      "grad_norm": 0.297001451253891,
      "learning_rate": 6.122965094173424e-05,
      "loss": 0.0294,
      "step": 2740
    },
    {
      "epoch": 2.232142857142857,
      "grad_norm": 0.21826975047588348,
      "learning_rate": 6.0960943729333374e-05,
      "loss": 0.0225,
      "step": 2750
    },
    {
      "epoch": 2.24025974025974,
      "grad_norm": 0.2182624638080597,
      "learning_rate": 6.069190355323717e-05,
      "loss": 0.0276,
      "step": 2760
    },
    {
      "epoch": 2.2483766233766236,
      "grad_norm": 0.31292399764060974,
      "learning_rate": 6.042253858615532e-05,
      "loss": 0.0231,
      "step": 2770
    },
    {
      "epoch": 2.2564935064935066,
      "grad_norm": 0.2803086042404175,
      "learning_rate": 6.015285701066382e-05,
      "loss": 0.0255,
      "step": 2780
    },
    {
      "epoch": 2.2646103896103895,
      "grad_norm": 0.3240359425544739,
      "learning_rate": 5.988286701895631e-05,
      "loss": 0.0258,
      "step": 2790
    },
    {
      "epoch": 2.2727272727272725,
      "grad_norm": 0.5147755742073059,
      "learning_rate": 5.961257681259535e-05,
      "loss": 0.0296,
      "step": 2800
    },
    {
      "epoch": 2.280844155844156,
      "grad_norm": 0.3791831135749817,
      "learning_rate": 5.934199460226317e-05,
      "loss": 0.0298,
      "step": 2810
    },
    {
      "epoch": 2.288961038961039,
      "grad_norm": 0.3257302939891815,
      "learning_rate": 5.9071128607512285e-05,
      "loss": 0.0224,
      "step": 2820
    },
    {
      "epoch": 2.2970779220779223,
      "grad_norm": 0.30238914489746094,
      "learning_rate": 5.8799987056515804e-05,
      "loss": 0.0272,
      "step": 2830
    },
    {
      "epoch": 2.3051948051948052,
      "grad_norm": 0.31279265880584717,
      "learning_rate": 5.8528578185817514e-05,
      "loss": 0.0259,
      "step": 2840
    },
    {
      "epoch": 2.313311688311688,
      "grad_norm": 0.23632532358169556,
      "learning_rate": 5.825691024008162e-05,
      "loss": 0.0232,
      "step": 2850
    },
    {
      "epoch": 2.3214285714285716,
      "grad_norm": 0.4876987934112549,
      "learning_rate": 5.798499147184233e-05,
      "loss": 0.0302,
      "step": 2860
    },
    {
      "epoch": 2.3295454545454546,
      "grad_norm": 0.47043561935424805,
      "learning_rate": 5.771283014125317e-05,
      "loss": 0.0245,
      "step": 2870
    },
    {
      "epoch": 2.3376623376623376,
      "grad_norm": 0.2626848518848419,
      "learning_rate": 5.7440434515836064e-05,
      "loss": 0.0259,
      "step": 2880
    },
    {
      "epoch": 2.345779220779221,
      "grad_norm": 0.26216691732406616,
      "learning_rate": 5.7167812870230094e-05,
      "loss": 0.0241,
      "step": 2890
    },
    {
      "epoch": 2.353896103896104,
      "grad_norm": 0.3104308247566223,
      "learning_rate": 5.689497348594035e-05,
      "loss": 0.0274,
      "step": 2900
    },
    {
      "epoch": 2.362012987012987,
      "grad_norm": 0.32353267073631287,
      "learning_rate": 5.662192465108613e-05,
      "loss": 0.0346,
      "step": 2910
    },
    {
      "epoch": 2.3701298701298703,
      "grad_norm": 0.33068498969078064,
      "learning_rate": 5.634867466014932e-05,
      "loss": 0.024,
      "step": 2920
    },
    {
      "epoch": 2.3782467532467533,
      "grad_norm": 0.2560001015663147,
      "learning_rate": 5.607523181372234e-05,
      "loss": 0.0258,
      "step": 2930
    },
    {
      "epoch": 2.3863636363636362,
      "grad_norm": 0.3798504173755646,
      "learning_rate": 5.5801604418256117e-05,
      "loss": 0.0252,
      "step": 2940
    },
    {
      "epoch": 2.3944805194805197,
      "grad_norm": 0.28704068064689636,
      "learning_rate": 5.552780078580756e-05,
      "loss": 0.0249,
      "step": 2950
    },
    {
      "epoch": 2.4025974025974026,
      "grad_norm": 0.22857019305229187,
      "learning_rate": 5.525382923378728e-05,
      "loss": 0.0294,
      "step": 2960
    },
    {
      "epoch": 2.4107142857142856,
      "grad_norm": 0.27925705909729004,
      "learning_rate": 5.49796980847068e-05,
      "loss": 0.0245,
      "step": 2970
    },
    {
      "epoch": 2.418831168831169,
      "grad_norm": 0.259421706199646,
      "learning_rate": 5.470541566592573e-05,
      "loss": 0.0272,
      "step": 2980
    },
    {
      "epoch": 2.426948051948052,
      "grad_norm": 0.33721524477005005,
      "learning_rate": 5.443099030939887e-05,
      "loss": 0.0252,
      "step": 2990
    },
    {
      "epoch": 2.435064935064935,
      "grad_norm": 0.23205623030662537,
      "learning_rate": 5.415643035142309e-05,
      "loss": 0.0254,
      "step": 3000
    },
    {
      "epoch": 2.4431818181818183,
      "grad_norm": 0.21620187163352966,
      "learning_rate": 5.3881744132384104e-05,
      "loss": 0.0248,
      "step": 3010
    },
    {
      "epoch": 2.4512987012987013,
      "grad_norm": 0.3049747347831726,
      "learning_rate": 5.360693999650303e-05,
      "loss": 0.0301,
      "step": 3020
    },
    {
      "epoch": 2.4594155844155843,
      "grad_norm": 0.22451385855674744,
      "learning_rate": 5.3332026291583016e-05,
      "loss": 0.0234,
      "step": 3030
    },
    {
      "epoch": 2.4675324675324677,
      "grad_norm": 0.28510794043540955,
      "learning_rate": 5.305701136875566e-05,
      "loss": 0.024,
      "step": 3040
    },
    {
      "epoch": 2.4756493506493507,
      "grad_norm": 0.3264691233634949,
      "learning_rate": 5.278190358222721e-05,
      "loss": 0.0227,
      "step": 3050
    },
    {
      "epoch": 2.4837662337662336,
      "grad_norm": 0.3188452422618866,
      "learning_rate": 5.25067112890249e-05,
      "loss": 0.0305,
      "step": 3060
    },
    {
      "epoch": 2.491883116883117,
      "grad_norm": 0.33883917331695557,
      "learning_rate": 5.2231442848743064e-05,
      "loss": 0.0275,
      "step": 3070
    },
    {
      "epoch": 2.5,
      "grad_norm": 0.3571816384792328,
      "learning_rate": 5.1956106623289145e-05,
      "loss": 0.0259,
      "step": 3080
    },
    {
      "epoch": 2.508116883116883,
      "grad_norm": 0.28985506296157837,
      "learning_rate": 5.168071097662972e-05,
      "loss": 0.0297,
      "step": 3090
    },
    {
      "epoch": 2.5162337662337664,
      "grad_norm": 0.3391208052635193,
      "learning_rate": 5.1405264274536445e-05,
      "loss": 0.0273,
      "step": 3100
    },
    {
      "epoch": 2.5243506493506493,
      "grad_norm": 0.32588908076286316,
      "learning_rate": 5.112977488433188e-05,
      "loss": 0.0308,
      "step": 3110
    },
    {
      "epoch": 2.5324675324675323,
      "grad_norm": 0.22015531361103058,
      "learning_rate": 5.085425117463533e-05,
      "loss": 0.0252,
      "step": 3120
    },
    {
      "epoch": 2.5405844155844157,
      "grad_norm": 0.32106679677963257,
      "learning_rate": 5.057870151510864e-05,
      "loss": 0.0275,
      "step": 3130
    },
    {
      "epoch": 2.5487012987012987,
      "grad_norm": 0.27428922057151794,
      "learning_rate": 5.030313427620197e-05,
      "loss": 0.0269,
      "step": 3140
    },
    {
      "epoch": 2.5568181818181817,
      "grad_norm": 0.2819305956363678,
      "learning_rate": 5.0027557828899426e-05,
      "loss": 0.0266,
      "step": 3150
    },
    {
      "epoch": 2.564935064935065,
      "grad_norm": 0.48730844259262085,
      "learning_rate": 4.975198054446492e-05,
      "loss": 0.0259,
      "step": 3160
    },
    {
      "epoch": 2.573051948051948,
      "grad_norm": 0.2712593078613281,
      "learning_rate": 4.947641079418773e-05,
      "loss": 0.0271,
      "step": 3170
    },
    {
      "epoch": 2.5811688311688314,
      "grad_norm": 0.26792702078819275,
      "learning_rate": 4.920085694912828e-05,
      "loss": 0.0256,
      "step": 3180
    },
    {
      "epoch": 2.5892857142857144,
      "grad_norm": 0.2116631269454956,
      "learning_rate": 4.892532737986387e-05,
      "loss": 0.0241,
      "step": 3190
    },
    {
      "epoch": 2.5974025974025974,
      "grad_norm": 0.20508752763271332,
      "learning_rate": 4.864983045623434e-05,
      "loss": 0.029,
      "step": 3200
    },
    {
      "epoch": 2.6055194805194803,
      "grad_norm": 0.3375774323940277,
      "learning_rate": 4.837437454708784e-05,
      "loss": 0.0257,
      "step": 3210
    },
    {
      "epoch": 2.6136363636363638,
      "grad_norm": 0.275234580039978,
      "learning_rate": 4.809896802002662e-05,
      "loss": 0.0228,
      "step": 3220
    },
    {
      "epoch": 2.6217532467532467,
      "grad_norm": 0.33019763231277466,
      "learning_rate": 4.7823619241152854e-05,
      "loss": 0.0243,
      "step": 3230
    },
    {
      "epoch": 2.62987012987013,
      "grad_norm": 0.28387683629989624,
      "learning_rate": 4.754833657481445e-05,
      "loss": 0.02,
      "step": 3240
    },
    {
      "epoch": 2.637987012987013,
      "grad_norm": 0.29118308424949646,
      "learning_rate": 4.7273128383351015e-05,
      "loss": 0.0213,
      "step": 3250
    },
    {
      "epoch": 2.646103896103896,
      "grad_norm": 0.2830529808998108,
      "learning_rate": 4.699800302683981e-05,
      "loss": 0.03,
      "step": 3260
    },
    {
      "epoch": 2.654220779220779,
      "grad_norm": 0.30629491806030273,
      "learning_rate": 4.6722968862841806e-05,
      "loss": 0.023,
      "step": 3270
    },
    {
      "epoch": 2.6623376623376624,
      "grad_norm": 0.22736108303070068,
      "learning_rate": 4.6448034246147754e-05,
      "loss": 0.0242,
      "step": 3280
    },
    {
      "epoch": 2.6704545454545454,
      "grad_norm": 0.30054470896720886,
      "learning_rate": 4.6173207528524476e-05,
      "loss": 0.0266,
      "step": 3290
    },
    {
      "epoch": 2.678571428571429,
      "grad_norm": 0.1740064173936844,
      "learning_rate": 4.58984970584611e-05,
      "loss": 0.0195,
      "step": 3300
    },
    {
      "epoch": 2.686688311688312,
      "grad_norm": 0.26478099822998047,
      "learning_rate": 4.562391118091544e-05,
      "loss": 0.0238,
      "step": 3310
    },
    {
      "epoch": 2.6948051948051948,
      "grad_norm": 0.25808119773864746,
      "learning_rate": 4.534945823706056e-05,
      "loss": 0.0214,
      "step": 3320
    },
    {
      "epoch": 2.7029220779220777,
      "grad_norm": 0.2761845588684082,
      "learning_rate": 4.507514656403137e-05,
      "loss": 0.0245,
      "step": 3330
    },
    {
      "epoch": 2.711038961038961,
      "grad_norm": 0.27152833342552185,
      "learning_rate": 4.480098449467132e-05,
      "loss": 0.0227,
      "step": 3340
    },
    {
      "epoch": 2.719155844155844,
      "grad_norm": 0.1654919534921646,
      "learning_rate": 4.452698035727929e-05,
      "loss": 0.023,
      "step": 3350
    },
    {
      "epoch": 2.7272727272727275,
      "grad_norm": 0.28704938292503357,
      "learning_rate": 4.425314247535668e-05,
      "loss": 0.0246,
      "step": 3360
    },
    {
      "epoch": 2.7353896103896105,
      "grad_norm": 0.2609517276287079,
      "learning_rate": 4.3979479167354477e-05,
      "loss": 0.0254,
      "step": 3370
    },
    {
      "epoch": 2.7435064935064934,
      "grad_norm": 0.16841895878314972,
      "learning_rate": 4.370599874642055e-05,
      "loss": 0.0225,
      "step": 3380
    },
    {
      "epoch": 2.7516233766233764,
      "grad_norm": 0.2550669014453888,
      "learning_rate": 4.3432709520147205e-05,
      "loss": 0.0227,
      "step": 3390
    },
    {
      "epoch": 2.75974025974026,
      "grad_norm": 0.29165688157081604,
      "learning_rate": 4.315961979031875e-05,
      "loss": 0.0276,
      "step": 3400
    },
    {
      "epoch": 2.767857142857143,
      "grad_norm": 0.2834565043449402,
      "learning_rate": 4.2886737852659325e-05,
      "loss": 0.0226,
      "step": 3410
    },
    {
      "epoch": 2.775974025974026,
      "grad_norm": 0.3727666139602661,
      "learning_rate": 4.261407199658093e-05,
      "loss": 0.0272,
      "step": 3420
    },
    {
      "epoch": 2.784090909090909,
      "grad_norm": 0.34494835138320923,
      "learning_rate": 4.234163050493158e-05,
      "loss": 0.0205,
      "step": 3430
    },
    {
      "epoch": 2.792207792207792,
      "grad_norm": 0.2529587745666504,
      "learning_rate": 4.2069421653743706e-05,
      "loss": 0.0208,
      "step": 3440
    },
    {
      "epoch": 2.800324675324675,
      "grad_norm": 0.3394658863544464,
      "learning_rate": 4.179745371198276e-05,
      "loss": 0.0262,
      "step": 3450
    },
    {
      "epoch": 2.8084415584415585,
      "grad_norm": 0.22549283504486084,
      "learning_rate": 4.1525734941296026e-05,
      "loss": 0.0261,
      "step": 3460
    },
    {
      "epoch": 2.8165584415584415,
      "grad_norm": 0.2669041156768799,
      "learning_rate": 4.125427359576162e-05,
      "loss": 0.0212,
      "step": 3470
    },
    {
      "epoch": 2.824675324675325,
      "grad_norm": 0.26632052659988403,
      "learning_rate": 4.0983077921637815e-05,
      "loss": 0.0228,
      "step": 3480
    },
    {
      "epoch": 2.832792207792208,
      "grad_norm": 0.20987533032894135,
      "learning_rate": 4.07121561571125e-05,
      "loss": 0.0262,
      "step": 3490
    },
    {
      "epoch": 2.840909090909091,
      "grad_norm": 0.42086461186408997,
      "learning_rate": 4.044151653205292e-05,
      "loss": 0.0265,
      "step": 3500
    },
    {
      "epoch": 2.849025974025974,
      "grad_norm": 0.23293422162532806,
      "learning_rate": 4.0171167267755696e-05,
      "loss": 0.0273,
      "step": 3510
    },
    {
      "epoch": 2.857142857142857,
      "grad_norm": 0.21170319616794586,
      "learning_rate": 3.9901116576697083e-05,
      "loss": 0.0225,
      "step": 3520
    },
    {
      "epoch": 2.86525974025974,
      "grad_norm": 0.26338985562324524,
      "learning_rate": 3.963137266228349e-05,
      "loss": 0.0188,
      "step": 3530
    },
    {
      "epoch": 2.8733766233766236,
      "grad_norm": 0.2753160297870636,
      "learning_rate": 3.93619437186023e-05,
      "loss": 0.0214,
      "step": 3540
    },
    {
      "epoch": 2.8814935064935066,
      "grad_norm": 0.2192564606666565,
      "learning_rate": 3.9092837930172884e-05,
      "loss": 0.0258,
      "step": 3550
    },
    {
      "epoch": 2.8896103896103895,
      "grad_norm": 0.1939588338136673,
      "learning_rate": 3.8824063471698105e-05,
      "loss": 0.0215,
      "step": 3560
    },
    {
      "epoch": 2.8977272727272725,
      "grad_norm": 0.3316612243652344,
      "learning_rate": 3.855562850781589e-05,
      "loss": 0.028,
      "step": 3570
    },
    {
      "epoch": 2.905844155844156,
      "grad_norm": 0.35597196221351624,
      "learning_rate": 3.828754119285123e-05,
      "loss": 0.0247,
      "step": 3580
    },
    {
      "epoch": 2.913961038961039,
      "grad_norm": 0.17957353591918945,
      "learning_rate": 3.801980967056851e-05,
      "loss": 0.0243,
      "step": 3590
    },
    {
      "epoch": 2.9220779220779223,
      "grad_norm": 0.2684866189956665,
      "learning_rate": 3.77524420739241e-05,
      "loss": 0.024,
      "step": 3600
    },
    {
      "epoch": 2.9301948051948052,
      "grad_norm": 0.27119967341423035,
      "learning_rate": 3.748544652481927e-05,
      "loss": 0.0223,
      "step": 3610
    },
    {
      "epoch": 2.938311688311688,
      "grad_norm": 0.2573845684528351,
      "learning_rate": 3.721883113385353e-05,
      "loss": 0.0224,
      "step": 3620
    },
    {
      "epoch": 2.946428571428571,
      "grad_norm": 0.22625279426574707,
      "learning_rate": 3.695260400007819e-05,
      "loss": 0.025,
      "step": 3630
    },
    {
      "epoch": 2.9545454545454546,
      "grad_norm": 0.254618376493454,
      "learning_rate": 3.6686773210750385e-05,
      "loss": 0.0242,
      "step": 3640
    },
    {
      "epoch": 2.9626623376623376,
      "grad_norm": 0.2085765153169632,
      "learning_rate": 3.642134684108737e-05,
      "loss": 0.0216,
      "step": 3650
    },
    {
      "epoch": 2.970779220779221,
      "grad_norm": 0.27070721983909607,
      "learning_rate": 3.615633295402123e-05,
      "loss": 0.0215,
      "step": 3660
    },
    {
      "epoch": 2.978896103896104,
      "grad_norm": 0.43438518047332764,
      "learning_rate": 3.5891739599953945e-05,
      "loss": 0.0232,
      "step": 3670
    },
    {
      "epoch": 2.987012987012987,
      "grad_norm": 0.2724141776561737,
      "learning_rate": 3.5627574816512846e-05,
      "loss": 0.022,
      "step": 3680
    },
    {
      "epoch": 2.99512987012987,
      "grad_norm": 0.2144225537776947,
      "learning_rate": 3.536384662830648e-05,
      "loss": 0.0222,
      "step": 3690
    },
    {
      "epoch": 3.0032467532467533,
      "grad_norm": 0.3140086531639099,
      "learning_rate": 3.5100563046680764e-05,
      "loss": 0.0274,
      "step": 3700
    },
    {
      "epoch": 3.0113636363636362,
      "grad_norm": 0.23359528183937073,
      "learning_rate": 3.483773206947572e-05,
      "loss": 0.0267,
      "step": 3710
    },
    {
      "epoch": 3.0194805194805197,
      "grad_norm": 0.26310062408447266,
      "learning_rate": 3.457536168078247e-05,
      "loss": 0.0266,
      "step": 3720
    },
    {
      "epoch": 3.0275974025974026,
      "grad_norm": 0.29639682173728943,
      "learning_rate": 3.431345985070067e-05,
      "loss": 0.0204,
      "step": 3730
    },
    {
      "epoch": 3.0357142857142856,
      "grad_norm": 0.2499479353427887,
      "learning_rate": 3.40520345350965e-05,
      "loss": 0.0226,
      "step": 3740
    },
    {
      "epoch": 3.043831168831169,
      "grad_norm": 0.2562311887741089,
      "learning_rate": 3.379109367536089e-05,
      "loss": 0.0217,
      "step": 3750
    },
    {
      "epoch": 3.051948051948052,
      "grad_norm": 0.27222245931625366,
      "learning_rate": 3.3530645198168295e-05,
      "loss": 0.0233,
      "step": 3760
    },
    {
      "epoch": 3.060064935064935,
      "grad_norm": 0.37355130910873413,
      "learning_rate": 3.327069701523595e-05,
      "loss": 0.0299,
      "step": 3770
    },
    {
      "epoch": 3.0681818181818183,
      "grad_norm": 0.2403850555419922,
      "learning_rate": 3.301125702308353e-05,
      "loss": 0.0222,
      "step": 3780
    },
    {
      "epoch": 3.0762987012987013,
      "grad_norm": 0.1968202441930771,
      "learning_rate": 3.275233310279321e-05,
      "loss": 0.0203,
      "step": 3790
    },
    {
      "epoch": 3.0844155844155843,
      "grad_norm": 0.23642730712890625,
      "learning_rate": 3.249393311977037e-05,
      "loss": 0.0229,
      "step": 3800
    },
    {
      "epoch": 3.0925324675324677,
      "grad_norm": 0.19872798025608063,
      "learning_rate": 3.223606492350451e-05,
      "loss": 0.0208,
      "step": 3810
    },
    {
      "epoch": 3.1006493506493507,
      "grad_norm": 0.20454058051109314,
      "learning_rate": 3.197873634733096e-05,
      "loss": 0.0242,
      "step": 3820
    },
    {
      "epoch": 3.1087662337662336,
      "grad_norm": 0.31078484654426575,
      "learning_rate": 3.172195520819285e-05,
      "loss": 0.0236,
      "step": 3830
    },
    {
      "epoch": 3.116883116883117,
      "grad_norm": 0.3581089377403259,
      "learning_rate": 3.146572930640362e-05,
      "loss": 0.0216,
      "step": 3840
    },
    {
      "epoch": 3.125,
      "grad_norm": 0.2517269551753998,
      "learning_rate": 3.121006642541014e-05,
      "loss": 0.0183,
      "step": 3850
    },
    {
      "epoch": 3.133116883116883,
      "grad_norm": 0.26109376549720764,
      "learning_rate": 3.095497433155626e-05,
      "loss": 0.0242,
      "step": 3860
    },
    {
      "epoch": 3.1412337662337664,
      "grad_norm": 0.20480413734912872,
      "learning_rate": 3.070046077384682e-05,
      "loss": 0.0202,
      "step": 3870
    },
    {
      "epoch": 3.1493506493506493,
      "grad_norm": 0.2771162986755371,
      "learning_rate": 3.0446533483712304e-05,
      "loss": 0.0231,
      "step": 3880
    },
    {
      "epoch": 3.1574675324675323,
      "grad_norm": 0.16268500685691833,
      "learning_rate": 3.0193200174774038e-05,
      "loss": 0.0173,
      "step": 3890
    },
    {
      "epoch": 3.1655844155844157,
      "grad_norm": 0.2741657793521881,
      "learning_rate": 2.994046854260974e-05,
      "loss": 0.0241,
      "step": 3900
    },
    {
      "epoch": 3.1737012987012987,
      "grad_norm": 0.38963595032691956,
      "learning_rate": 2.9688346264519866e-05,
      "loss": 0.019,
      "step": 3910
    },
    {
      "epoch": 3.1818181818181817,
      "grad_norm": 0.36379268765449524,
      "learning_rate": 2.943684099929436e-05,
      "loss": 0.0218,
      "step": 3920
    },
    {
      "epoch": 3.189935064935065,
      "grad_norm": 0.34124332666397095,
      "learning_rate": 2.918596038697995e-05,
      "loss": 0.0185,
      "step": 3930
    },
    {
      "epoch": 3.198051948051948,
      "grad_norm": 0.25402340292930603,
      "learning_rate": 2.8935712048648112e-05,
      "loss": 0.0204,
      "step": 3940
    },
    {
      "epoch": 3.206168831168831,
      "grad_norm": 0.2678219676017761,
      "learning_rate": 2.8686103586163626e-05,
      "loss": 0.0212,
      "step": 3950
    },
    {
      "epoch": 3.2142857142857144,
      "grad_norm": 0.2375226765871048,
      "learning_rate": 2.843714258195346e-05,
      "loss": 0.0211,
      "step": 3960
    },
    {
      "epoch": 3.2224025974025974,
      "grad_norm": 0.22339288890361786,
      "learning_rate": 2.8188836598776662e-05,
      "loss": 0.0194,
      "step": 3970
    },
    {
      "epoch": 3.2305194805194803,
      "grad_norm": 0.25037524104118347,
      "learning_rate": 2.7941193179494484e-05,
      "loss": 0.0205,
      "step": 3980
    },
    {
      "epoch": 3.2386363636363638,
      "grad_norm": 0.19681259989738464,
      "learning_rate": 2.7694219846841262e-05,
      "loss": 0.0197,
      "step": 3990
    },
    {
      "epoch": 3.2467532467532467,
      "grad_norm": 0.2873547375202179,
      "learning_rate": 2.7447924103195976e-05,
      "loss": 0.0212,
      "step": 4000
    },
    {
      "epoch": 3.2548701298701297,
      "grad_norm": 0.33498382568359375,
      "learning_rate": 2.7202313430354253e-05,
      "loss": 0.0202,
      "step": 4010
    },
    {
      "epoch": 3.262987012987013,
      "grad_norm": 0.2162805050611496,
      "learning_rate": 2.695739528930111e-05,
      "loss": 0.0179,
      "step": 4020
    },
    {
      "epoch": 3.271103896103896,
      "grad_norm": 0.25468140840530396,
      "learning_rate": 2.67131771199844e-05,
      "loss": 0.02,
      "step": 4030
    },
    {
      "epoch": 3.279220779220779,
      "grad_norm": 0.21498627960681915,
      "learning_rate": 2.6469666341088677e-05,
      "loss": 0.0243,
      "step": 4040
    },
    {
      "epoch": 3.2873376623376624,
      "grad_norm": 0.3046325445175171,
      "learning_rate": 2.6226870349809885e-05,
      "loss": 0.0197,
      "step": 4050
    },
    {
      "epoch": 3.2954545454545454,
      "grad_norm": 0.2851937413215637,
      "learning_rate": 2.5984796521630737e-05,
      "loss": 0.0233,
      "step": 4060
    },
    {
      "epoch": 3.3035714285714284,
      "grad_norm": 0.2599593997001648,
      "learning_rate": 2.574345221009653e-05,
      "loss": 0.0207,
      "step": 4070
    },
    {
      "epoch": 3.311688311688312,
      "grad_norm": 0.19292496144771576,
      "learning_rate": 2.5502844746591804e-05,
      "loss": 0.0184,
      "step": 4080
    },
    {
      "epoch": 3.3198051948051948,
      "grad_norm": 0.2586632966995239,
      "learning_rate": 2.526298144011775e-05,
      "loss": 0.0184,
      "step": 4090
    },
    {
      "epoch": 3.3279220779220777,
      "grad_norm": 0.2998685836791992,
      "learning_rate": 2.5023869577070013e-05,
      "loss": 0.0225,
      "step": 4100
    },
    {
      "epoch": 3.336038961038961,
      "grad_norm": 0.3807864189147949,
      "learning_rate": 2.478551642101743e-05,
      "loss": 0.0193,
      "step": 4110
    },
    {
      "epoch": 3.344155844155844,
      "grad_norm": 0.2566285729408264,
      "learning_rate": 2.4547929212481435e-05,
      "loss": 0.0191,
      "step": 4120
    },
    {
      "epoch": 3.3522727272727275,
      "grad_norm": 0.17539237439632416,
      "learning_rate": 2.4311115168716013e-05,
      "loss": 0.0208,
      "step": 4130
    },
    {
      "epoch": 3.3603896103896105,
      "grad_norm": 0.18529631197452545,
      "learning_rate": 2.4075081483488494e-05,
      "loss": 0.0191,
      "step": 4140
    },
    {
      "epoch": 3.3685064935064934,
      "grad_norm": 0.19923549890518188,
      "learning_rate": 2.3839835326861104e-05,
      "loss": 0.0213,
      "step": 4150
    },
    {
      "epoch": 3.3766233766233764,
      "grad_norm": 0.2135431319475174,
      "learning_rate": 2.3605383844972966e-05,
      "loss": 0.0129,
      "step": 4160
    },
    {
      "epoch": 3.38474025974026,
      "grad_norm": 0.3140435814857483,
      "learning_rate": 2.3371734159823284e-05,
      "loss": 0.0233,
      "step": 4170
    },
    {
      "epoch": 3.392857142857143,
      "grad_norm": 0.18021227419376373,
      "learning_rate": 2.3138893369054766e-05,
      "loss": 0.0186,
      "step": 4180
    },
    {
      "epoch": 3.400974025974026,
      "grad_norm": 0.2429431974887848,
      "learning_rate": 2.2906868545738102e-05,
      "loss": 0.0241,
      "step": 4190
    },
    {
      "epoch": 3.409090909090909,
      "grad_norm": 0.15611667931079865,
      "learning_rate": 2.2675666738157186e-05,
      "loss": 0.0177,
      "step": 4200
    },
    {
      "epoch": 3.417207792207792,
      "grad_norm": 0.30634909868240356,
      "learning_rate": 2.2445294969594844e-05,
      "loss": 0.0154,
      "step": 4210
    },
    {
      "epoch": 3.425324675324675,
      "grad_norm": 0.23167884349822998,
      "learning_rate": 2.22157602381196e-05,
      "loss": 0.015,
      "step": 4220
    },
    {
      "epoch": 3.4334415584415585,
      "grad_norm": 0.325353741645813,
      "learning_rate": 2.1987069516373098e-05,
      "loss": 0.0144,
      "step": 4230
    },
    {
      "epoch": 3.4415584415584415,
      "grad_norm": 0.21330159902572632,
      "learning_rate": 2.1759229751358217e-05,
      "loss": 0.0183,
      "step": 4240
    },
    {
      "epoch": 3.449675324675325,
      "grad_norm": 0.3478533625602722,
      "learning_rate": 2.1532247864228084e-05,
      "loss": 0.02,
      "step": 4250
    },
    {
      "epoch": 3.457792207792208,
      "grad_norm": 0.1727161556482315,
      "learning_rate": 2.1306130750075865e-05,
      "loss": 0.0249,
      "step": 4260
    },
    {
      "epoch": 3.465909090909091,
      "grad_norm": 0.19407041370868683,
      "learning_rate": 2.1080885277725236e-05,
      "loss": 0.0174,
      "step": 4270
    },
    {
      "epoch": 3.474025974025974,
      "grad_norm": 0.35378298163414,
      "learning_rate": 2.085651828952175e-05,
      "loss": 0.0193,
      "step": 4280
    },
    {
      "epoch": 3.482142857142857,
      "grad_norm": 0.2549465000629425,
      "learning_rate": 2.063303660112506e-05,
      "loss": 0.0163,
      "step": 4290
    },
    {
      "epoch": 3.49025974025974,
      "grad_norm": 0.2552514970302582,
      "learning_rate": 2.0410447001301753e-05,
      "loss": 0.0171,
      "step": 4300
    },
    {
      "epoch": 3.4983766233766236,
      "grad_norm": 0.2298896312713623,
      "learning_rate": 2.0188756251719203e-05,
      "loss": 0.0167,
      "step": 4310
    },
    {
      "epoch": 3.5064935064935066,
      "grad_norm": 0.23686964809894562,
      "learning_rate": 1.9967971086740195e-05,
      "loss": 0.0193,
      "step": 4320
    },
    {
      "epoch": 3.5146103896103895,
      "grad_norm": 0.2347656786441803,
      "learning_rate": 1.974809821321827e-05,
      "loss": 0.016,
      "step": 4330
    },
    {
      "epoch": 3.5227272727272725,
      "grad_norm": 0.2654639780521393,
      "learning_rate": 1.9529144310294023e-05,
      "loss": 0.0152,
      "step": 4340
    },
    {
      "epoch": 3.530844155844156,
      "grad_norm": 0.1813756227493286,
      "learning_rate": 1.9311116029192278e-05,
      "loss": 0.0171,
      "step": 4350
    },
    {
      "epoch": 3.538961038961039,
      "grad_norm": 0.2936214804649353,
      "learning_rate": 1.909401999301993e-05,
      "loss": 0.0162,
      "step": 4360
    },
    {
      "epoch": 3.5470779220779223,
      "grad_norm": 0.21046514809131622,
      "learning_rate": 1.887786279656482e-05,
      "loss": 0.0166,
      "step": 4370
    },
    {
      "epoch": 3.5551948051948052,
      "grad_norm": 0.3618803918361664,
      "learning_rate": 1.8662651006095387e-05,
      "loss": 0.0156,
      "step": 4380
    },
    {
      "epoch": 3.563311688311688,
      "grad_norm": 0.23544606566429138,
      "learning_rate": 1.8448391159161204e-05,
      "loss": 0.0183,
      "step": 4390
    },
    {
      "epoch": 3.571428571428571,
      "grad_norm": 0.28176313638687134,
      "learning_rate": 1.8235089764394408e-05,
      "loss": 0.023,
      "step": 4400
    },
    {
      "epoch": 3.5795454545454546,
      "grad_norm": 0.26535889506340027,
      "learning_rate": 1.8022753301311935e-05,
      "loss": 0.0213,
      "step": 4410
    },
    {
      "epoch": 3.5876623376623376,
      "grad_norm": 0.25943446159362793,
      "learning_rate": 1.7811388220118707e-05,
      "loss": 0.0175,
      "step": 4420
    },
    {
      "epoch": 3.595779220779221,
      "grad_norm": 0.15602117776870728,
      "learning_rate": 1.7601000941511757e-05,
      "loss": 0.0173,
      "step": 4430
    },
    {
      "epoch": 3.603896103896104,
      "grad_norm": 0.2504715919494629,
      "learning_rate": 1.7391597856485083e-05,
      "loss": 0.023,
      "step": 4440
    },
    {
      "epoch": 3.612012987012987,
      "grad_norm": 0.1567242443561554,
      "learning_rate": 1.7183185326135543e-05,
      "loss": 0.0167,
      "step": 4450
    },
    {
      "epoch": 3.62012987012987,
      "grad_norm": 0.268656462430954,
      "learning_rate": 1.6975769681469705e-05,
      "loss": 0.018,
      "step": 4460
    },
    {
      "epoch": 3.6282467532467533,
      "grad_norm": 0.37558749318122864,
      "learning_rate": 1.676935722321139e-05,
      "loss": 0.0174,
      "step": 4470
    },
    {
      "epoch": 3.6363636363636362,
      "grad_norm": 0.1570769101381302,
      "learning_rate": 1.6563954221610355e-05,
      "loss": 0.0175,
      "step": 4480
    },
    {
      "epoch": 3.6444805194805197,
      "grad_norm": 0.19073347747325897,
      "learning_rate": 1.6359566916251845e-05,
      "loss": 0.018,
      "step": 4490
    },
    {
      "epoch": 3.6525974025974026,
      "grad_norm": 0.17310558259487152,
      "learning_rate": 1.615620151586697e-05,
      "loss": 0.0168,
      "step": 4500
    },
    {
      "epoch": 3.6607142857142856,
      "grad_norm": 0.31775960326194763,
      "learning_rate": 1.5953864198144135e-05,
      "loss": 0.02,
      "step": 4510
    },
    {
      "epoch": 3.6688311688311686,
      "grad_norm": 0.1850634217262268,
      "learning_rate": 1.5752561109541447e-05,
      "loss": 0.0195,
      "step": 4520
    },
    {
      "epoch": 3.676948051948052,
      "grad_norm": 0.21364572644233704,
      "learning_rate": 1.5552298365099882e-05,
      "loss": 0.017,
      "step": 4530
    },
    {
      "epoch": 3.685064935064935,
      "grad_norm": 0.2515982687473297,
      "learning_rate": 1.5353082048257596e-05,
      "loss": 0.0148,
      "step": 4540
    },
    {
      "epoch": 3.6931818181818183,
      "grad_norm": 0.21852737665176392,
      "learning_rate": 1.5154918210665148e-05,
      "loss": 0.0166,
      "step": 4550
    },
    {
      "epoch": 3.7012987012987013,
      "grad_norm": 0.1786801964044571,
      "learning_rate": 1.4957812872001614e-05,
      "loss": 0.0138,
      "step": 4560
    },
    {
      "epoch": 3.7094155844155843,
      "grad_norm": 0.14969338476657867,
      "learning_rate": 1.4761772019791748e-05,
      "loss": 0.0181,
      "step": 4570
    },
    {
      "epoch": 3.7175324675324677,
      "grad_norm": 0.21216048300266266,
      "learning_rate": 1.4566801609224096e-05,
      "loss": 0.017,
      "step": 4580
    },
    {
      "epoch": 3.7256493506493507,
      "grad_norm": 0.16176529228687286,
      "learning_rate": 1.4372907562970079e-05,
      "loss": 0.016,
      "step": 4590
    },
    {
      "epoch": 3.7337662337662336,
      "grad_norm": 0.25220224261283875,
      "learning_rate": 1.4180095771004154e-05,
      "loss": 0.0148,
      "step": 4600
    },
    {
      "epoch": 3.741883116883117,
      "grad_norm": 0.1625714898109436,
      "learning_rate": 1.3988372090424773e-05,
      "loss": 0.0192,
      "step": 4610
    },
    {
      "epoch": 3.75,
      "grad_norm": 0.26604965329170227,
      "learning_rate": 1.3797742345276521e-05,
      "loss": 0.017,
      "step": 4620
    },
    {
      "epoch": 3.758116883116883,
      "grad_norm": 0.27260708808898926,
      "learning_rate": 1.3608212326373249e-05,
      "loss": 0.0176,
      "step": 4630
    },
    {
      "epoch": 3.7662337662337664,
      "grad_norm": 0.15837593376636505,
      "learning_rate": 1.3419787791122062e-05,
      "loss": 0.0142,
      "step": 4640
    },
    {
      "epoch": 3.7743506493506493,
      "grad_norm": 0.18480677902698517,
      "learning_rate": 1.323247446334847e-05,
      "loss": 0.019,
      "step": 4650
    },
    {
      "epoch": 3.7824675324675323,
      "grad_norm": 0.2447320967912674,
      "learning_rate": 1.3046278033122577e-05,
      "loss": 0.0244,
      "step": 4660
    },
    {
      "epoch": 3.7905844155844157,
      "grad_norm": 0.22955277562141418,
      "learning_rate": 1.286120415658611e-05,
      "loss": 0.0196,
      "step": 4670
    },
    {
      "epoch": 3.7987012987012987,
      "grad_norm": 0.15361738204956055,
      "learning_rate": 1.2677258455780683e-05,
      "loss": 0.0152,
      "step": 4680
    },
    {
      "epoch": 3.8068181818181817,
      "grad_norm": 0.2417902648448944,
      "learning_rate": 1.2494446518477022e-05,
      "loss": 0.0164,
      "step": 4690
    },
    {
      "epoch": 3.814935064935065,
      "grad_norm": 0.18717902898788452,
      "learning_rate": 1.2312773898005175e-05,
      "loss": 0.0178,
      "step": 4700
    },
    {
      "epoch": 3.823051948051948,
      "grad_norm": 0.20804652571678162,
      "learning_rate": 1.2132246113085822e-05,
      "loss": 0.0171,
      "step": 4710
    },
    {
      "epoch": 3.8311688311688314,
      "grad_norm": 0.2959846258163452,
      "learning_rate": 1.1952868647662696e-05,
      "loss": 0.0161,
      "step": 4720
    },
    {
      "epoch": 3.8392857142857144,
      "grad_norm": 0.15556474030017853,
      "learning_rate": 1.1774646950735913e-05,
      "loss": 0.0164,
      "step": 4730
    },
    {
      "epoch": 3.8474025974025974,
      "grad_norm": 0.24023498594760895,
      "learning_rate": 1.1597586436196473e-05,
      "loss": 0.0136,
      "step": 4740
    },
    {
      "epoch": 3.8555194805194803,
      "grad_norm": 0.2286766916513443,
      "learning_rate": 1.1421692482661856e-05,
      "loss": 0.015,
      "step": 4750
    },
    {
      "epoch": 3.8636363636363638,
      "grad_norm": 0.191964790225029,
      "learning_rate": 1.124697043331256e-05,
      "loss": 0.0157,
      "step": 4760
    },
    {
      "epoch": 3.8717532467532467,
      "grad_norm": 0.17993806302547455,
      "learning_rate": 1.107342559572977e-05,
      "loss": 0.0187,
      "step": 4770
    },
    {
      "epoch": 3.87987012987013,
      "grad_norm": 0.20845425128936768,
      "learning_rate": 1.090106324173426e-05,
      "loss": 0.0194,
      "step": 4780
    },
    {
      "epoch": 3.887987012987013,
      "grad_norm": 0.20088578760623932,
      "learning_rate": 1.0729888607226113e-05,
      "loss": 0.0146,
      "step": 4790
    },
    {
      "epoch": 3.896103896103896,
      "grad_norm": 0.24833039939403534,
      "learning_rate": 1.0559906892025745e-05,
      "loss": 0.0177,
      "step": 4800
    },
    {
      "epoch": 3.904220779220779,
      "grad_norm": 0.18850843608379364,
      "learning_rate": 1.0391123259715906e-05,
      "loss": 0.0142,
      "step": 4810
    },
    {
      "epoch": 3.9123376623376624,
      "grad_norm": 0.1293126493692398,
      "learning_rate": 1.0223542837484839e-05,
      "loss": 0.012,
      "step": 4820
    },
    {
      "epoch": 3.9204545454545454,
      "grad_norm": 0.18274357914924622,
      "learning_rate": 1.0057170715970559e-05,
      "loss": 0.0162,
      "step": 4830
    },
    {
      "epoch": 3.928571428571429,
      "grad_norm": 0.1782849133014679,
      "learning_rate": 9.892011949106172e-06,
      "loss": 0.0295,
      "step": 4840
    },
    {
      "epoch": 3.936688311688312,
      "grad_norm": 0.40481868386268616,
      "learning_rate": 9.728071553966339e-06,
      "loss": 0.0174,
      "step": 4850
    },
    {
      "epoch": 3.9448051948051948,
      "grad_norm": 0.20304250717163086,
      "learning_rate": 9.56535451061496e-06,
      "loss": 0.0138,
      "step": 4860
    },
    {
      "epoch": 3.9529220779220777,
      "grad_norm": 0.22350141406059265,
      "learning_rate": 9.403865761953779e-06,
      "loss": 0.0212,
      "step": 4870
    },
    {
      "epoch": 3.961038961038961,
      "grad_norm": 0.28146418929100037,
      "learning_rate": 9.243610213572285e-06,
      "loss": 0.0194,
      "step": 4880
    },
    {
      "epoch": 3.969155844155844,
      "grad_norm": 0.2405545562505722,
      "learning_rate": 9.084592733598735e-06,
      "loss": 0.0129,
      "step": 4890
    },
    {
      "epoch": 3.9772727272727275,
      "grad_norm": 0.27352192997932434,
      "learning_rate": 8.92681815255219e-06,
      "loss": 0.0174,
      "step": 4900
    },
    {
      "epoch": 3.9853896103896105,
      "grad_norm": 0.1327618509531021,
      "learning_rate": 8.770291263195819e-06,
      "loss": 0.0176,
      "step": 4910
    },
    {
      "epoch": 3.9935064935064934,
      "grad_norm": 0.1761268973350525,
      "learning_rate": 8.615016820391342e-06,
      "loss": 0.0119,
      "step": 4920
    },
    {
      "epoch": 4.001623376623376,
      "grad_norm": 0.35601338744163513,
      "learning_rate": 8.460999540954517e-06,
      "loss": 0.0158,
      "step": 4930
    },
    {
      "epoch": 4.009740259740259,
      "grad_norm": 0.17832595109939575,
      "learning_rate": 8.308244103511909e-06,
      "loss": 0.015,
      "step": 4940
    },
    {
      "epoch": 4.017857142857143,
      "grad_norm": 0.15967492759227753,
      "learning_rate": 8.156755148358764e-06,
      "loss": 0.0156,
      "step": 4950
    },
    {
      "epoch": 4.025974025974026,
      "grad_norm": 0.17017978429794312,
      "learning_rate": 8.00653727731801e-06,
      "loss": 0.0147,
      "step": 4960
    },
    {
      "epoch": 4.034090909090909,
      "grad_norm": 0.1537715643644333,
      "learning_rate": 7.857595053600513e-06,
      "loss": 0.0185,
      "step": 4970
    },
    {
      "epoch": 4.042207792207792,
      "grad_norm": 0.2602587938308716,
      "learning_rate": 7.709933001666431e-06,
      "loss": 0.021,
      "step": 4980
    },
    {
      "epoch": 4.050324675324675,
      "grad_norm": 0.2224591225385666,
      "learning_rate": 7.56355560708778e-06,
      "loss": 0.0131,
      "step": 4990
    },
    {
      "epoch": 4.058441558441558,
      "grad_norm": 0.3965739905834198,
      "learning_rate": 7.418467316412158e-06,
      "loss": 0.0173,
      "step": 5000
    },
    {
      "epoch": 4.066558441558442,
      "grad_norm": 0.20591789484024048,
      "learning_rate": 7.2746725370277435e-06,
      "loss": 0.0154,
      "step": 5010
    },
    {
      "epoch": 4.074675324675325,
      "grad_norm": 0.19456496834754944,
      "learning_rate": 7.132175637029293e-06,
      "loss": 0.017,
      "step": 5020
    },
    {
      "epoch": 4.082792207792208,
      "grad_norm": 0.226767435669899,
      "learning_rate": 6.9909809450855345e-06,
      "loss": 0.0169,
      "step": 5030
    },
    {
      "epoch": 4.090909090909091,
      "grad_norm": 0.2061874121427536,
      "learning_rate": 6.851092750307686e-06,
      "loss": 0.0153,
      "step": 5040
    },
    {
      "epoch": 4.099025974025974,
      "grad_norm": 0.16970008611679077,
      "learning_rate": 6.712515302119077e-06,
      "loss": 0.0149,
      "step": 5050
    },
    {
      "epoch": 4.107142857142857,
      "grad_norm": 0.16201592981815338,
      "learning_rate": 6.575252810126143e-06,
      "loss": 0.0125,
      "step": 5060
    },
    {
      "epoch": 4.115259740259741,
      "grad_norm": 0.2043023258447647,
      "learning_rate": 6.439309443990532e-06,
      "loss": 0.0145,
      "step": 5070
    },
    {
      "epoch": 4.123376623376624,
      "grad_norm": 0.19296449422836304,
      "learning_rate": 6.304689333302416e-06,
      "loss": 0.0146,
      "step": 5080
    },
    {
      "epoch": 4.1314935064935066,
      "grad_norm": 0.22057557106018066,
      "learning_rate": 6.171396567455051e-06,
      "loss": 0.0159,
      "step": 5090
    },
    {
      "epoch": 4.1396103896103895,
      "grad_norm": 0.15243513882160187,
      "learning_rate": 6.039435195520604e-06,
      "loss": 0.0179,
      "step": 5100
    },
    {
      "epoch": 4.1477272727272725,
      "grad_norm": 0.11597718298435211,
      "learning_rate": 5.908809226127054e-06,
      "loss": 0.0135,
      "step": 5110
    },
    {
      "epoch": 4.1558441558441555,
      "grad_norm": 0.35636165738105774,
      "learning_rate": 5.779522627336537e-06,
      "loss": 0.0182,
      "step": 5120
    },
    {
      "epoch": 4.163961038961039,
      "grad_norm": 0.2736373841762543,
      "learning_rate": 5.651579326524709e-06,
      "loss": 0.0139,
      "step": 5130
    },
    {
      "epoch": 4.172077922077922,
      "grad_norm": 0.15613017976284027,
      "learning_rate": 5.524983210261481e-06,
      "loss": 0.0139,
      "step": 5140
    },
    {
      "epoch": 4.180194805194805,
      "grad_norm": 0.26912152767181396,
      "learning_rate": 5.399738124192988e-06,
      "loss": 0.0146,
      "step": 5150
    },
    {
      "epoch": 4.188311688311688,
      "grad_norm": 0.22851672768592834,
      "learning_rate": 5.2758478729247164e-06,
      "loss": 0.0146,
      "step": 5160
    },
    {
      "epoch": 4.196428571428571,
      "grad_norm": 0.1734984815120697,
      "learning_rate": 5.153316219905946e-06,
      "loss": 0.0119,
      "step": 5170
    },
    {
      "epoch": 4.204545454545454,
      "grad_norm": 0.12929925322532654,
      "learning_rate": 5.032146887315448e-06,
      "loss": 0.0117,
      "step": 5180
    },
    {
      "epoch": 4.212662337662338,
      "grad_norm": 0.25070181488990784,
      "learning_rate": 4.91234355594839e-06,
      "loss": 0.016,
      "step": 5190
    },
    {
      "epoch": 4.220779220779221,
      "grad_norm": 0.12695631384849548,
      "learning_rate": 4.7939098651045235e-06,
      "loss": 0.0177,
      "step": 5200
    },
    {
      "epoch": 4.228896103896104,
      "grad_norm": 0.18711066246032715,
      "learning_rate": 4.67684941247768e-06,
      "loss": 0.0148,
      "step": 5210
    },
    {
      "epoch": 4.237012987012987,
      "grad_norm": 0.1875857561826706,
      "learning_rate": 4.5611657540464036e-06,
      "loss": 0.0153,
      "step": 5220
    },
    {
      "epoch": 4.24512987012987,
      "grad_norm": 0.18307802081108093,
      "learning_rate": 4.446862403965984e-06,
      "loss": 0.0147,
      "step": 5230
    },
    {
      "epoch": 4.253246753246753,
      "grad_norm": 0.29160308837890625,
      "learning_rate": 4.333942834461702e-06,
      "loss": 0.0135,
      "step": 5240
    },
    {
      "epoch": 4.261363636363637,
      "grad_norm": 0.26900240778923035,
      "learning_rate": 4.222410475723326e-06,
      "loss": 0.0128,
      "step": 5250
    },
    {
      "epoch": 4.26948051948052,
      "grad_norm": 0.22926048934459686,
      "learning_rate": 4.112268715800943e-06,
      "loss": 0.0141,
      "step": 5260
    },
    {
      "epoch": 4.277597402597403,
      "grad_norm": 0.21062804758548737,
      "learning_rate": 4.003520900502028e-06,
      "loss": 0.0177,
      "step": 5270
    },
    {
      "epoch": 4.285714285714286,
      "grad_norm": 0.16253219544887543,
      "learning_rate": 3.8961703332898e-06,
      "loss": 0.0144,
      "step": 5280
    },
    {
      "epoch": 4.2938311688311686,
      "grad_norm": 0.29484206438064575,
      "learning_rate": 3.790220275182854e-06,
      "loss": 0.0135,
      "step": 5290
    },
    {
      "epoch": 4.301948051948052,
      "grad_norm": 0.20472891628742218,
      "learning_rate": 3.685673944656176e-06,
      "loss": 0.0137,
      "step": 5300
    },
    {
      "epoch": 4.310064935064935,
      "grad_norm": 0.20979756116867065,
      "learning_rate": 3.582534517543268e-06,
      "loss": 0.015,
      "step": 5310
    },
    {
      "epoch": 4.318181818181818,
      "grad_norm": 0.19138938188552856,
      "learning_rate": 3.4808051269397512e-06,
      "loss": 0.0142,
      "step": 5320
    },
    {
      "epoch": 4.326298701298701,
      "grad_norm": 0.1692737489938736,
      "learning_rate": 3.380488863108183e-06,
      "loss": 0.014,
      "step": 5330
    },
    {
      "epoch": 4.334415584415584,
      "grad_norm": 0.22084228694438934,
      "learning_rate": 3.2815887733841365e-06,
      "loss": 0.0143,
      "step": 5340
    },
    {
      "epoch": 4.342532467532467,
      "grad_norm": 0.12135329097509384,
      "learning_rate": 3.1841078620836683e-06,
      "loss": 0.0126,
      "step": 5350
    },
    {
      "epoch": 4.35064935064935,
      "grad_norm": 0.2081880420446396,
      "learning_rate": 3.0880490904120874e-06,
      "loss": 0.0148,
      "step": 5360
    },
    {
      "epoch": 4.358766233766234,
      "grad_norm": 0.21467359364032745,
      "learning_rate": 2.9934153763739205e-06,
      "loss": 0.0175,
      "step": 5370
    },
    {
      "epoch": 4.366883116883117,
      "grad_norm": 0.14720773696899414,
      "learning_rate": 2.9002095946843277e-06,
      "loss": 0.014,
      "step": 5380
    },
    {
      "epoch": 4.375,
      "grad_norm": 0.18871958553791046,
      "learning_rate": 2.8084345766817676e-06,
      "loss": 0.0123,
      "step": 5390
    },
    {
      "epoch": 4.383116883116883,
      "grad_norm": 0.19331897795200348,
      "learning_rate": 2.718093110241976e-06,
      "loss": 0.0125,
      "step": 5400
    },
    {
      "epoch": 4.391233766233766,
      "grad_norm": 0.23791873455047607,
      "learning_rate": 2.6291879396933004e-06,
      "loss": 0.0154,
      "step": 5410
    },
    {
      "epoch": 4.39935064935065,
      "grad_norm": 0.16598720848560333,
      "learning_rate": 2.541721765733318e-06,
      "loss": 0.0117,
      "step": 5420
    },
    {
      "epoch": 4.407467532467533,
      "grad_norm": 0.14102214574813843,
      "learning_rate": 2.455697245346783e-06,
      "loss": 0.0137,
      "step": 5430
    },
    {
      "epoch": 4.415584415584416,
      "grad_norm": 0.12769535183906555,
      "learning_rate": 2.371116991724953e-06,
      "loss": 0.0157,
      "step": 5440
    },
    {
      "epoch": 4.423701298701299,
      "grad_norm": 0.12857921421527863,
      "learning_rate": 2.2879835741861586e-06,
      "loss": 0.0196,
      "step": 5450
    },
    {
      "epoch": 4.431818181818182,
      "grad_norm": 0.30980998277664185,
      "learning_rate": 2.206299518097804e-06,
      "loss": 0.0117,
      "step": 5460
    },
    {
      "epoch": 4.439935064935065,
      "grad_norm": 0.2178056240081787,
      "learning_rate": 2.1260673047996227e-06,
      "loss": 0.0173,
      "step": 5470
    },
    {
      "epoch": 4.448051948051948,
      "grad_norm": 0.15620489418506622,
      "learning_rate": 2.047289371528299e-06,
      "loss": 0.0104,
      "step": 5480
    },
    {
      "epoch": 4.4561688311688314,
      "grad_norm": 0.1251637190580368,
      "learning_rate": 1.96996811134344e-06,
      "loss": 0.0136,
      "step": 5490
    },
    {
      "epoch": 4.464285714285714,
      "grad_norm": 0.10176528990268707,
      "learning_rate": 1.8941058730549132e-06,
      "loss": 0.0115,
      "step": 5500
    },
    {
      "epoch": 4.472402597402597,
      "grad_norm": 0.19561536610126495,
      "learning_rate": 1.8197049611514194e-06,
      "loss": 0.0139,
      "step": 5510
    },
    {
      "epoch": 4.48051948051948,
      "grad_norm": 0.18119896948337555,
      "learning_rate": 1.7467676357305561e-06,
      "loss": 0.0123,
      "step": 5520
    },
    {
      "epoch": 4.488636363636363,
      "grad_norm": 0.2555669844150543,
      "learning_rate": 1.6752961124301415e-06,
      "loss": 0.0154,
      "step": 5530
    },
    {
      "epoch": 4.496753246753247,
      "grad_norm": 0.21792449057102203,
      "learning_rate": 1.6052925623609049e-06,
      "loss": 0.0189,
      "step": 5540
    },
    {
      "epoch": 4.50487012987013,
      "grad_norm": 0.1742662638425827,
      "learning_rate": 1.5367591120405256e-06,
      "loss": 0.0112,
      "step": 5550
    },
    {
      "epoch": 4.512987012987013,
      "grad_norm": 0.14939716458320618,
      "learning_rate": 1.4696978433290653e-06,
      "loss": 0.0137,
      "step": 5560
    },
    {
      "epoch": 4.521103896103896,
      "grad_norm": 0.14489960670471191,
      "learning_rate": 1.4041107933656928e-06,
      "loss": 0.0148,
      "step": 5570
    },
    {
      "epoch": 4.529220779220779,
      "grad_norm": 0.2126537412405014,
      "learning_rate": 1.339999954506821e-06,
      "loss": 0.0133,
      "step": 5580
    },
    {
      "epoch": 4.537337662337662,
      "grad_norm": 0.2052183747291565,
      "learning_rate": 1.2773672742655784e-06,
      "loss": 0.0143,
      "step": 5590
    },
    {
      "epoch": 4.545454545454545,
      "grad_norm": 0.16533197462558746,
      "learning_rate": 1.2162146552526399e-06,
      "loss": 0.0145,
      "step": 5600
    },
    {
      "epoch": 4.553571428571429,
      "grad_norm": 0.1644553244113922,
      "learning_rate": 1.1565439551184664e-06,
      "loss": 0.019,
      "step": 5610
    },
    {
      "epoch": 4.561688311688312,
      "grad_norm": 0.11637627333402634,
      "learning_rate": 1.0983569864968346e-06,
      "loss": 0.013,
      "step": 5620
    },
    {
      "epoch": 4.569805194805195,
      "grad_norm": 0.12196806818246841,
      "learning_rate": 1.0416555169497688e-06,
      "loss": 0.0124,
      "step": 5630
    },
    {
      "epoch": 4.577922077922078,
      "grad_norm": 0.19040703773498535,
      "learning_rate": 9.864412689139123e-07,
      "loss": 0.0145,
      "step": 5640
    },
    {
      "epoch": 4.586038961038961,
      "grad_norm": 0.1617046594619751,
      "learning_rate": 9.327159196481138e-07,
      "loss": 0.0142,
      "step": 5650
    },
    {
      "epoch": 4.5941558441558445,
      "grad_norm": 0.14342036843299866,
      "learning_rate": 8.804811011825398e-07,
      "loss": 0.0102,
      "step": 5660
    },
    {
      "epoch": 4.6022727272727275,
      "grad_norm": 0.17284558713436127,
      "learning_rate": 8.297384002690866e-07,
      "loss": 0.0147,
      "step": 5670
    },
    {
      "epoch": 4.6103896103896105,
      "grad_norm": 0.2236797958612442,
      "learning_rate": 7.804893583331696e-07,
      "loss": 0.0127,
      "step": 5680
    },
    {
      "epoch": 4.6185064935064934,
      "grad_norm": 0.17601999640464783,
      "learning_rate": 7.32735471426893e-07,
      "loss": 0.017,
      "step": 5690
    },
    {
      "epoch": 4.626623376623376,
      "grad_norm": 0.14382389187812805,
      "learning_rate": 6.864781901836259e-07,
      "loss": 0.0143,
      "step": 5700
    },
    {
      "epoch": 4.634740259740259,
      "grad_norm": 0.13551688194274902,
      "learning_rate": 6.417189197739093e-07,
      "loss": 0.0156,
      "step": 5710
    },
    {
      "epoch": 4.642857142857143,
      "grad_norm": 0.14876428246498108,
      "learning_rate": 5.984590198627849e-07,
      "loss": 0.0171,
      "step": 5720
    },
    {
      "epoch": 4.650974025974026,
      "grad_norm": 0.1803891956806183,
      "learning_rate": 5.566998045685112e-07,
      "loss": 0.0161,
      "step": 5730
    },
    {
      "epoch": 4.659090909090909,
      "grad_norm": 0.2185366451740265,
      "learning_rate": 5.164425424226016e-07,
      "loss": 0.0119,
      "step": 5740
    },
    {
      "epoch": 4.667207792207792,
      "grad_norm": 0.1409493088722229,
      "learning_rate": 4.776884563313266e-07,
      "loss": 0.0127,
      "step": 5750
    },
    {
      "epoch": 4.675324675324675,
      "grad_norm": 0.12052121758460999,
      "learning_rate": 4.404387235385443e-07,
      "loss": 0.0128,
      "step": 5760
    },
    {
      "epoch": 4.683441558441558,
      "grad_norm": 0.14814621210098267,
      "learning_rate": 4.0469447558995065e-07,
      "loss": 0.0157,
      "step": 5770
    },
    {
      "epoch": 4.691558441558442,
      "grad_norm": 0.2510521113872528,
      "learning_rate": 3.7045679829870175e-07,
      "loss": 0.0152,
      "step": 5780
    },
    {
      "epoch": 4.699675324675325,
      "grad_norm": 0.18528954684734344,
      "learning_rate": 3.377267317124233e-07,
      "loss": 0.0153,
      "step": 5790
    },
    {
      "epoch": 4.707792207792208,
      "grad_norm": 0.3942748010158539,
      "learning_rate": 3.0650527008162513e-07,
      "loss": 0.0141,
      "step": 5800
    },
    {
      "epoch": 4.715909090909091,
      "grad_norm": 0.10147012770175934,
      "learning_rate": 2.767933618295082e-07,
      "loss": 0.0122,
      "step": 5810
    },
    {
      "epoch": 4.724025974025974,
      "grad_norm": 0.19354303181171417,
      "learning_rate": 2.485919095231326e-07,
      "loss": 0.014,
      "step": 5820
    },
    {
      "epoch": 4.732142857142857,
      "grad_norm": 0.1840514838695526,
      "learning_rate": 2.219017698460002e-07,
      "loss": 0.0164,
      "step": 5830
    },
    {
      "epoch": 4.740259740259741,
      "grad_norm": 0.3346545994281769,
      "learning_rate": 1.9672375357206452e-07,
      "loss": 0.0134,
      "step": 5840
    },
    {
      "epoch": 4.748376623376624,
      "grad_norm": 0.24521546065807343,
      "learning_rate": 1.73058625541056e-07,
      "loss": 0.0135,
      "step": 5850
    },
    {
      "epoch": 4.7564935064935066,
      "grad_norm": 0.184413880109787,
      "learning_rate": 1.5090710463527836e-07,
      "loss": 0.0173,
      "step": 5860
    },
    {
      "epoch": 4.7646103896103895,
      "grad_norm": 0.15935760736465454,
      "learning_rate": 1.3026986375776485e-07,
      "loss": 0.014,
      "step": 5870
    },
    {
      "epoch": 4.7727272727272725,
      "grad_norm": 0.24353346228599548,
      "learning_rate": 1.1114752981183917e-07,
      "loss": 0.0159,
      "step": 5880
    },
    {
      "epoch": 4.7808441558441555,
      "grad_norm": 0.17594245076179504,
      "learning_rate": 9.354068368204739e-08,
      "loss": 0.0159,
      "step": 5890
    },
    {
      "epoch": 4.788961038961039,
      "grad_norm": 0.15030580759048462,
      "learning_rate": 7.744986021656076e-08,
      "loss": 0.0138,
      "step": 5900
    },
    {
      "epoch": 4.797077922077922,
      "grad_norm": 0.20552325248718262,
      "learning_rate": 6.287554821087783e-08,
      "loss": 0.0141,
      "step": 5910
    },
    {
      "epoch": 4.805194805194805,
      "grad_norm": 0.16335158050060272,
      "learning_rate": 4.981819039300284e-08,
      "loss": 0.0156,
      "step": 5920
    },
    {
      "epoch": 4.813311688311688,
      "grad_norm": 0.29560235142707825,
      "learning_rate": 3.827818341000655e-08,
      "loss": 0.0161,
      "step": 5930
    },
    {
      "epoch": 4.821428571428571,
      "grad_norm": 0.41275930404663086,
      "learning_rate": 2.8255877815946963e-08,
      "loss": 0.0145,
      "step": 5940
    },
    {
      "epoch": 4.829545454545455,
      "grad_norm": 0.21795588731765747,
      "learning_rate": 1.9751578061244504e-08,
      "loss": 0.0118,
      "step": 5950
    },
    {
      "epoch": 4.837662337662338,
      "grad_norm": 0.16346313059329987,
      "learning_rate": 1.2765542483417214e-08,
      "loss": 0.0106,
      "step": 5960
    },
    {
      "epoch": 4.845779220779221,
      "grad_norm": 0.2356872707605362,
      "learning_rate": 7.2979832992592365e-09,
      "loss": 0.0133,
      "step": 5970
    },
    {
      "epoch": 4.853896103896104,
      "grad_norm": 0.14218561351299286,
      "learning_rate": 3.349066598362649e-09,
      "loss": 0.013,
      "step": 5980
    },
    {
      "epoch": 4.862012987012987,
      "grad_norm": 0.173988476395607,
      "learning_rate": 9.189123380826114e-10,
      "loss": 0.0128,
      "step": 5990
    },
    {
      "epoch": 4.87012987012987,
      "grad_norm": 0.30807921290397644,
      "learning_rate": 7.594339912486703e-12,
      "loss": 0.0155,
      "step": 6000
    }
  ],
  "logging_steps": 10,
  "max_steps": 6000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 2000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}