{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 5.0,
  "eval_steps": 500,
  "global_step": 11575,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "grad_norm": 0.8661230868133192,
      "learning_rate": 1.7271157167530226e-08,
      "loss": 0.1037,
      "step": 1
    },
    {
      "epoch": 0.0,
      "grad_norm": 1.1661544306907603,
      "learning_rate": 8.635578583765113e-08,
      "loss": 0.0948,
      "step": 5
    },
    {
      "epoch": 0.0,
      "grad_norm": 0.7537109348930305,
      "learning_rate": 1.7271157167530226e-07,
      "loss": 0.1015,
      "step": 10
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.2061731506590818,
      "learning_rate": 2.5906735751295336e-07,
      "loss": 0.1073,
      "step": 15
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.0211147540146708,
      "learning_rate": 3.454231433506045e-07,
      "loss": 0.0973,
      "step": 20
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.1617195446495134,
      "learning_rate": 4.317789291882556e-07,
      "loss": 0.1121,
      "step": 25
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.8211862670686246,
      "learning_rate": 5.181347150259067e-07,
      "loss": 0.1026,
      "step": 30
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.9348381607194174,
      "learning_rate": 6.04490500863558e-07,
      "loss": 0.1192,
      "step": 35
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.490417880543222,
      "learning_rate": 6.90846286701209e-07,
      "loss": 0.094,
      "step": 40
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.8388913689714425,
      "learning_rate": 7.772020725388602e-07,
      "loss": 0.0977,
      "step": 45
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.7004297572965578,
      "learning_rate": 8.635578583765112e-07,
      "loss": 0.1032,
      "step": 50
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.6059616583139867,
      "learning_rate": 9.499136442141624e-07,
      "loss": 0.0749,
      "step": 55
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.9955847374719834,
      "learning_rate": 1.0362694300518134e-06,
      "loss": 0.1005,
      "step": 60
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.8320891857428879,
      "learning_rate": 1.1226252158894648e-06,
      "loss": 0.0853,
      "step": 65
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.7954532245692706,
      "learning_rate": 1.208981001727116e-06,
      "loss": 0.0879,
      "step": 70
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.8813868206948325,
      "learning_rate": 1.2953367875647669e-06,
      "loss": 0.0902,
      "step": 75
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.0058746202593967,
      "learning_rate": 1.381692573402418e-06,
      "loss": 0.1145,
      "step": 80
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.7597389247113029,
      "learning_rate": 1.4680483592400694e-06,
      "loss": 0.1021,
      "step": 85
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.46103661687167447,
      "learning_rate": 1.5544041450777204e-06,
      "loss": 0.0806,
      "step": 90
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.737893040063868,
      "learning_rate": 1.6407599309153715e-06,
      "loss": 0.0875,
      "step": 95
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.8310674040774385,
      "learning_rate": 1.7271157167530224e-06,
      "loss": 0.0832,
      "step": 100
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.984904820117129,
      "learning_rate": 1.8134715025906738e-06,
      "loss": 0.1034,
      "step": 105
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.035690398188729,
      "learning_rate": 1.8998272884283248e-06,
      "loss": 0.0915,
      "step": 110
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.8084351137915218,
      "learning_rate": 1.986183074265976e-06,
      "loss": 0.0955,
      "step": 115
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.7974766116289803,
      "learning_rate": 2.072538860103627e-06,
      "loss": 0.0906,
      "step": 120
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.6030316166998918,
      "learning_rate": 2.1588946459412782e-06,
      "loss": 0.1054,
      "step": 125
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.8476703419068902,
      "learning_rate": 2.2452504317789296e-06,
      "loss": 0.0875,
      "step": 130
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.7435355944089723,
      "learning_rate": 2.3316062176165805e-06,
      "loss": 0.0951,
      "step": 135
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.019975713303781,
      "learning_rate": 2.417962003454232e-06,
      "loss": 0.0918,
      "step": 140
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.8408723502054206,
      "learning_rate": 2.504317789291883e-06,
      "loss": 0.0984,
      "step": 145
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.787665840797315,
      "learning_rate": 2.5906735751295338e-06,
      "loss": 0.0905,
      "step": 150
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.7319427568278293,
      "learning_rate": 2.6770293609671847e-06,
      "loss": 0.0802,
      "step": 155
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.7710169679916523,
      "learning_rate": 2.763385146804836e-06,
      "loss": 0.091,
      "step": 160
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.7154619591661221,
      "learning_rate": 2.8497409326424875e-06,
      "loss": 0.0821,
      "step": 165
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.939909195416624,
      "learning_rate": 2.936096718480139e-06,
      "loss": 0.1003,
      "step": 170
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.7738010454826961,
      "learning_rate": 3.0224525043177893e-06,
      "loss": 0.0975,
      "step": 175
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.86507670980491,
      "learning_rate": 3.1088082901554407e-06,
      "loss": 0.0989,
      "step": 180
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.6050718382645259,
      "learning_rate": 3.195164075993092e-06,
      "loss": 0.0843,
      "step": 185
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.8827419506686391,
      "learning_rate": 3.281519861830743e-06,
      "loss": 0.0956,
      "step": 190
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.890758444313074,
      "learning_rate": 3.367875647668394e-06,
      "loss": 0.0937,
      "step": 195
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.0254768465690436,
      "learning_rate": 3.454231433506045e-06,
      "loss": 0.088,
      "step": 200
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.1096548052357276,
      "learning_rate": 3.5405872193436963e-06,
      "loss": 0.1031,
      "step": 205
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.0635828264297262,
      "learning_rate": 3.6269430051813476e-06,
      "loss": 0.0995,
      "step": 210
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.7459111481743179,
      "learning_rate": 3.713298791018999e-06,
      "loss": 0.073,
      "step": 215
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.703283436759907,
      "learning_rate": 3.7996545768566495e-06,
      "loss": 0.074,
      "step": 220
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.0566826153490951,
      "learning_rate": 3.886010362694301e-06,
      "loss": 0.0936,
      "step": 225
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.0016492237862977,
      "learning_rate": 3.972366148531952e-06,
      "loss": 0.1009,
      "step": 230
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.5128173139360626,
      "learning_rate": 4.058721934369604e-06,
      "loss": 0.0904,
      "step": 235
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.856968479324872,
      "learning_rate": 4.145077720207254e-06,
      "loss": 0.088,
      "step": 240
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.7601294161989428,
      "learning_rate": 4.2314335060449055e-06,
      "loss": 0.0895,
      "step": 245
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.0746309115244876,
      "learning_rate": 4.3177892918825564e-06,
      "loss": 0.0831,
      "step": 250
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.7906243315776752,
      "learning_rate": 4.404145077720207e-06,
      "loss": 0.0737,
      "step": 255
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.679401480283055,
      "learning_rate": 4.490500863557859e-06,
      "loss": 0.0848,
      "step": 260
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.374569808090678,
      "learning_rate": 4.576856649395509e-06,
      "loss": 0.0812,
      "step": 265
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.905921167975438,
      "learning_rate": 4.663212435233161e-06,
      "loss": 0.1048,
      "step": 270
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.9148670093529526,
      "learning_rate": 4.749568221070812e-06,
      "loss": 0.1088,
      "step": 275
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.6860981856697029,
      "learning_rate": 4.835924006908464e-06,
      "loss": 0.0953,
      "step": 280
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.7141327749409578,
      "learning_rate": 4.922279792746114e-06,
      "loss": 0.093,
      "step": 285
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.7857764564175557,
      "learning_rate": 5.008635578583766e-06,
      "loss": 0.0865,
      "step": 290
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.4587481775006207,
      "learning_rate": 5.094991364421417e-06,
      "loss": 0.0715,
      "step": 295
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.6672088038539078,
      "learning_rate": 5.1813471502590676e-06,
      "loss": 0.0766,
      "step": 300
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.7695334407032586,
      "learning_rate": 5.267702936096719e-06,
      "loss": 0.0855,
      "step": 305
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.5016272306147599,
      "learning_rate": 5.3540587219343694e-06,
      "loss": 0.0777,
      "step": 310
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.49424188342660924,
      "learning_rate": 5.440414507772021e-06,
      "loss": 0.0878,
      "step": 315
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.2989946292550959,
      "learning_rate": 5.526770293609672e-06,
      "loss": 0.0811,
      "step": 320
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.9640494211712551,
      "learning_rate": 5.613126079447323e-06,
      "loss": 0.0893,
      "step": 325
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.449054653018769,
      "learning_rate": 5.699481865284975e-06,
      "loss": 0.1132,
      "step": 330
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.8085449725247954,
      "learning_rate": 5.785837651122626e-06,
      "loss": 0.0845,
      "step": 335
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.7225319435995665,
      "learning_rate": 5.872193436960278e-06,
      "loss": 0.0904,
      "step": 340
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.2928666812431443,
      "learning_rate": 5.958549222797928e-06,
      "loss": 0.0655,
      "step": 345
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.5315887810709465,
      "learning_rate": 6.044905008635579e-06,
      "loss": 0.1082,
      "step": 350
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.7560372608550086,
      "learning_rate": 6.1312607944732305e-06,
      "loss": 0.0732,
      "step": 355
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.8551844011661183,
      "learning_rate": 6.217616580310881e-06,
      "loss": 0.0957,
      "step": 360
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.5218822357688347,
      "learning_rate": 6.3039723661485315e-06,
      "loss": 0.077,
      "step": 365
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.8145206500610358,
      "learning_rate": 6.390328151986184e-06,
      "loss": 0.0945,
      "step": 370
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.8232997561718624,
      "learning_rate": 6.476683937823834e-06,
      "loss": 0.0858,
      "step": 375
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.6907309189014695,
      "learning_rate": 6.563039723661486e-06,
      "loss": 0.0886,
      "step": 380
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.6155827336032126,
      "learning_rate": 6.649395509499137e-06,
      "loss": 0.0886,
      "step": 385
    },
    {
      "epoch": 0.17,
      "grad_norm": 2.121964258128152,
      "learning_rate": 6.735751295336788e-06,
      "loss": 0.0964,
      "step": 390
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.42147856280146334,
      "learning_rate": 6.82210708117444e-06,
      "loss": 0.0773,
      "step": 395
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.7617970675602044,
      "learning_rate": 6.90846286701209e-06,
      "loss": 0.088,
      "step": 400
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.4880498398490212,
      "learning_rate": 6.994818652849742e-06,
      "loss": 0.0722,
      "step": 405
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.5379171086423975,
      "learning_rate": 7.0811744386873925e-06,
      "loss": 0.0939,
      "step": 410
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.8777252382337768,
      "learning_rate": 7.1675302245250435e-06,
      "loss": 0.1001,
      "step": 415
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.8711901075408542,
      "learning_rate": 7.253886010362695e-06,
      "loss": 0.0829,
      "step": 420
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.725690370014714,
      "learning_rate": 7.340241796200346e-06,
      "loss": 0.0874,
      "step": 425
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.8773286754142784,
      "learning_rate": 7.426597582037998e-06,
      "loss": 0.0955,
      "step": 430
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.636866710021352,
      "learning_rate": 7.512953367875648e-06,
      "loss": 0.0983,
      "step": 435
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.6827849545920833,
      "learning_rate": 7.599309153713299e-06,
      "loss": 0.0848,
      "step": 440
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.7079397816443027,
      "learning_rate": 7.68566493955095e-06,
      "loss": 0.0965,
      "step": 445
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.7181170354131496,
      "learning_rate": 7.772020725388602e-06,
      "loss": 0.0752,
      "step": 450
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.322597394817283,
      "learning_rate": 7.858376511226253e-06,
      "loss": 0.072,
      "step": 455
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.5792459466721516,
      "learning_rate": 7.944732297063904e-06,
      "loss": 0.0821,
      "step": 460
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.8786329648976334,
      "learning_rate": 8.031088082901555e-06,
      "loss": 0.0988,
      "step": 465
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.5945920943627181,
      "learning_rate": 8.117443868739207e-06,
      "loss": 0.0973,
      "step": 470
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.8347979512880902,
      "learning_rate": 8.203799654576856e-06,
      "loss": 0.101,
      "step": 475
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.6469239378245417,
      "learning_rate": 8.290155440414507e-06,
      "loss": 0.087,
      "step": 480
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.890031998434686,
      "learning_rate": 8.37651122625216e-06,
      "loss": 0.0987,
      "step": 485
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.8556583399604323,
      "learning_rate": 8.462867012089811e-06,
      "loss": 0.0763,
      "step": 490
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.5748423587848386,
      "learning_rate": 8.549222797927462e-06,
      "loss": 0.0939,
      "step": 495
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.592030133056585,
      "learning_rate": 8.635578583765113e-06,
      "loss": 0.0776,
      "step": 500
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.8496557430119451,
      "learning_rate": 8.721934369602764e-06,
      "loss": 0.0834,
      "step": 505
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.7433725646125416,
      "learning_rate": 8.808290155440415e-06,
      "loss": 0.086,
      "step": 510
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.8962892707404773,
      "learning_rate": 8.894645941278066e-06,
      "loss": 0.091,
      "step": 515
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.7094137576645999,
      "learning_rate": 8.981001727115718e-06,
      "loss": 0.0873,
      "step": 520
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.004889054863807,
      "learning_rate": 9.06735751295337e-06,
      "loss": 0.1075,
      "step": 525
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.0825605461343522,
      "learning_rate": 9.153713298791019e-06,
      "loss": 0.0917,
      "step": 530
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.6138783402804222,
      "learning_rate": 9.240069084628671e-06,
      "loss": 0.0886,
      "step": 535
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.7381864789865002,
      "learning_rate": 9.326424870466322e-06,
      "loss": 0.0829,
      "step": 540
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.6258130636911013,
      "learning_rate": 9.412780656303973e-06,
      "loss": 0.0851,
      "step": 545
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.4788323221052753,
      "learning_rate": 9.499136442141624e-06,
      "loss": 0.0804,
      "step": 550
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.108997196870818,
      "learning_rate": 9.585492227979275e-06,
      "loss": 0.0875,
      "step": 555
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.5979786366623295,
      "learning_rate": 9.671848013816928e-06,
      "loss": 0.0784,
      "step": 560
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.9964613100797841,
      "learning_rate": 9.758203799654577e-06,
      "loss": 0.1102,
      "step": 565
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.6033999040613355,
      "learning_rate": 9.844559585492228e-06,
      "loss": 0.0763,
      "step": 570
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.8903085892847482,
      "learning_rate": 9.93091537132988e-06,
      "loss": 0.1059,
      "step": 575
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.4824096488930525,
      "learning_rate": 1.0017271157167531e-05,
      "loss": 0.0894,
      "step": 580
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.2433951606130729,
      "learning_rate": 1.0103626943005182e-05,
      "loss": 0.1024,
      "step": 585
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.6778057289429442,
      "learning_rate": 1.0189982728842833e-05,
      "loss": 0.1009,
      "step": 590
    },
    {
      "epoch": 0.26,
      "grad_norm": 448.5872376035829,
      "learning_rate": 1.0276338514680484e-05,
      "loss": 0.136,
      "step": 595
    },
    {
      "epoch": 0.26,
      "grad_norm": 8.21806945206516,
      "learning_rate": 1.0362694300518135e-05,
      "loss": 0.2339,
      "step": 600
    },
    {
      "epoch": 0.26,
      "grad_norm": 32.423915481989496,
      "learning_rate": 1.0449050086355788e-05,
      "loss": 0.1776,
      "step": 605
    },
    {
      "epoch": 0.26,
      "grad_norm": 4.773716201573569,
      "learning_rate": 1.0535405872193439e-05,
      "loss": 0.1198,
      "step": 610
    },
    {
      "epoch": 0.27,
      "grad_norm": 2.2737662019398757,
      "learning_rate": 1.062176165803109e-05,
      "loss": 0.1325,
      "step": 615
    },
    {
      "epoch": 0.27,
      "grad_norm": 2.7445938115491577,
      "learning_rate": 1.0708117443868739e-05,
      "loss": 0.1243,
      "step": 620
    },
    {
      "epoch": 0.27,
      "grad_norm": 4.8485979268084956,
      "learning_rate": 1.079447322970639e-05,
      "loss": 0.1309,
      "step": 625
    },
    {
      "epoch": 0.27,
      "grad_norm": 98.21615534763852,
      "learning_rate": 1.0880829015544042e-05,
      "loss": 0.6446,
      "step": 630
    },
    {
      "epoch": 0.27,
      "grad_norm": 20.369120279227072,
      "learning_rate": 1.0967184801381693e-05,
      "loss": 0.4051,
      "step": 635
    },
    {
      "epoch": 0.28,
      "grad_norm": 9.255717401202803,
      "learning_rate": 1.1053540587219344e-05,
      "loss": 0.2551,
      "step": 640
    },
    {
      "epoch": 0.28,
      "grad_norm": 4.584163246732508,
      "learning_rate": 1.1139896373056995e-05,
      "loss": 0.1466,
      "step": 645
    },
    {
      "epoch": 0.28,
      "grad_norm": 2.6563813074260962,
      "learning_rate": 1.1226252158894646e-05,
      "loss": 0.1524,
      "step": 650
    },
    {
      "epoch": 0.28,
      "grad_norm": 11.743230448649134,
      "learning_rate": 1.1312607944732299e-05,
      "loss": 0.129,
      "step": 655
    },
    {
      "epoch": 0.29,
      "grad_norm": 2.210415579371572,
      "learning_rate": 1.139896373056995e-05,
      "loss": 0.1433,
      "step": 660
    },
    {
      "epoch": 0.29,
      "grad_norm": 2.9933290541845463,
      "learning_rate": 1.14853195164076e-05,
      "loss": 0.133,
      "step": 665
    },
    {
      "epoch": 0.29,
      "grad_norm": 3.9237889199038127,
      "learning_rate": 1.1571675302245252e-05,
      "loss": 0.1038,
      "step": 670
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.135124310946562,
      "learning_rate": 1.1658031088082901e-05,
      "loss": 0.1103,
      "step": 675
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.4743333927949402,
      "learning_rate": 1.1744386873920555e-05,
      "loss": 0.1268,
      "step": 680
    },
    {
      "epoch": 0.3,
      "grad_norm": 4.532155326678469,
      "learning_rate": 1.1830742659758205e-05,
      "loss": 0.1106,
      "step": 685
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.9554908116369246,
      "learning_rate": 1.1917098445595855e-05,
      "loss": 0.0977,
      "step": 690
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.6864808315479922,
      "learning_rate": 1.2003454231433506e-05,
      "loss": 0.103,
      "step": 695
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.949196719147114,
      "learning_rate": 1.2089810017271157e-05,
      "loss": 0.1181,
      "step": 700
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.7696240182489283,
      "learning_rate": 1.217616580310881e-05,
      "loss": 0.1242,
      "step": 705
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.8567236220701305,
      "learning_rate": 1.2262521588946461e-05,
      "loss": 0.096,
      "step": 710
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.8225245866733908,
      "learning_rate": 1.2348877374784112e-05,
      "loss": 0.1088,
      "step": 715
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.9565738061517076,
      "learning_rate": 1.2435233160621763e-05,
      "loss": 0.0966,
      "step": 720
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.6439159172681707,
      "learning_rate": 1.2521588946459414e-05,
      "loss": 0.098,
      "step": 725
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.7777338342869954,
      "learning_rate": 1.2607944732297063e-05,
      "loss": 0.0934,
      "step": 730
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.4576043339962497,
      "learning_rate": 1.2694300518134717e-05,
      "loss": 0.1199,
      "step": 735
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.0279834902840417,
      "learning_rate": 1.2780656303972368e-05,
      "loss": 0.0885,
      "step": 740
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.6026067629298502,
      "learning_rate": 1.2867012089810018e-05,
      "loss": 0.1264,
      "step": 745
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.4783691316537515,
      "learning_rate": 1.2953367875647668e-05,
      "loss": 0.1059,
      "step": 750
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.9486933852319395,
      "learning_rate": 1.303972366148532e-05,
      "loss": 0.0922,
      "step": 755
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.4499456879597254,
      "learning_rate": 1.3126079447322972e-05,
      "loss": 0.1014,
      "step": 760
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.9941662211122442,
      "learning_rate": 1.3212435233160623e-05,
      "loss": 0.1022,
      "step": 765
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.8036557979399938,
      "learning_rate": 1.3298791018998274e-05,
      "loss": 0.0902,
      "step": 770
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.7748669920716171,
      "learning_rate": 1.3385146804835925e-05,
      "loss": 0.0818,
      "step": 775
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.657169437820155,
      "learning_rate": 1.3471502590673576e-05,
      "loss": 0.1055,
      "step": 780
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.46710916166028693,
      "learning_rate": 1.3557858376511228e-05,
      "loss": 0.1182,
      "step": 785
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.8598455308727582,
      "learning_rate": 1.364421416234888e-05,
      "loss": 0.0931,
      "step": 790
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.3357712764845688,
      "learning_rate": 1.373056994818653e-05,
      "loss": 0.0874,
      "step": 795
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.0229109182232703,
      "learning_rate": 1.381692573402418e-05,
      "loss": 0.0952,
      "step": 800
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.5623898387210554,
      "learning_rate": 1.390328151986183e-05,
      "loss": 0.11,
      "step": 805
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.7245600666438339,
      "learning_rate": 1.3989637305699483e-05,
      "loss": 0.1029,
      "step": 810
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.510481458435366,
      "learning_rate": 1.4075993091537134e-05,
      "loss": 0.0986,
      "step": 815
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.6791060587785951,
      "learning_rate": 1.4162348877374785e-05,
      "loss": 0.1105,
      "step": 820
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.41798381472993135,
      "learning_rate": 1.4248704663212436e-05,
      "loss": 0.1066,
      "step": 825
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.9469611913330616,
      "learning_rate": 1.4335060449050087e-05,
      "loss": 0.0805,
      "step": 830
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.7479499626670872,
      "learning_rate": 1.442141623488774e-05,
      "loss": 0.0987,
      "step": 835
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.7514723994820159,
      "learning_rate": 1.450777202072539e-05,
      "loss": 0.1001,
      "step": 840
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.7402662681471279,
      "learning_rate": 1.4594127806563041e-05,
      "loss": 0.0872,
      "step": 845
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.7009014324991375,
      "learning_rate": 1.4680483592400692e-05,
      "loss": 0.0999,
      "step": 850
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.5733120839978919,
      "learning_rate": 1.4766839378238342e-05,
      "loss": 0.0987,
      "step": 855
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.6928308586582662,
      "learning_rate": 1.4853195164075996e-05,
      "loss": 0.0954,
      "step": 860
    },
    {
      "epoch": 0.37,
      "grad_norm": 1.1809120617319104,
      "learning_rate": 1.4939550949913645e-05,
      "loss": 0.0931,
      "step": 865
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.5924468807561128,
      "learning_rate": 1.5025906735751296e-05,
      "loss": 0.079,
      "step": 870
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.6862371179811677,
      "learning_rate": 1.5112262521588947e-05,
      "loss": 0.0963,
      "step": 875
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.8000620112908129,
      "learning_rate": 1.5198618307426598e-05,
      "loss": 0.1036,
      "step": 880
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.2520177997722077,
      "learning_rate": 1.5284974093264252e-05,
      "loss": 0.1247,
      "step": 885
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.8857701186569507,
      "learning_rate": 1.53713298791019e-05,
      "loss": 0.0829,
      "step": 890
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.9391358915324642,
      "learning_rate": 1.545768566493955e-05,
      "loss": 0.1097,
      "step": 895
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.867842560800044,
      "learning_rate": 1.5544041450777204e-05,
      "loss": 0.0946,
      "step": 900
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.9400349073231179,
      "learning_rate": 1.5630397236614853e-05,
      "loss": 0.1056,
      "step": 905
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.6224643324430222,
      "learning_rate": 1.5716753022452505e-05,
      "loss": 0.0884,
      "step": 910
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.6375382173467924,
      "learning_rate": 1.5803108808290158e-05,
      "loss": 0.1019,
      "step": 915
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.5289546723303301,
      "learning_rate": 1.5889464594127807e-05,
      "loss": 0.1053,
      "step": 920
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.7834683566163679,
      "learning_rate": 1.597582037996546e-05,
      "loss": 0.1037,
      "step": 925
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.9867612062335633,
      "learning_rate": 1.606217616580311e-05,
      "loss": 0.1095,
      "step": 930
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.8209415616389953,
      "learning_rate": 1.6148531951640762e-05,
      "loss": 0.0885,
      "step": 935
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.1682789140133532,
      "learning_rate": 1.6234887737478414e-05,
      "loss": 0.0885,
      "step": 940
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.9859818071339542,
      "learning_rate": 1.6321243523316064e-05,
      "loss": 0.0962,
      "step": 945
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.7829183817929026,
      "learning_rate": 1.6407599309153713e-05,
      "loss": 0.0894,
      "step": 950
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.9752553349973404,
      "learning_rate": 1.6493955094991366e-05,
      "loss": 0.0822,
      "step": 955
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.7139804233290242,
      "learning_rate": 1.6580310880829015e-05,
      "loss": 0.1096,
      "step": 960
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.9303149577410212,
      "learning_rate": 1.6666666666666667e-05,
      "loss": 0.0819,
      "step": 965
    },
    {
      "epoch": 0.42,
      "grad_norm": 1.2058202041327308,
      "learning_rate": 1.675302245250432e-05,
      "loss": 0.0982,
      "step": 970
    },
    {
      "epoch": 0.42,
      "grad_norm": 1.9759790480560875,
      "learning_rate": 1.683937823834197e-05,
      "loss": 0.0834,
      "step": 975
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.9091397007217376,
      "learning_rate": 1.6925734024179622e-05,
      "loss": 0.116,
      "step": 980
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.5255731686872832,
      "learning_rate": 1.701208981001727e-05,
      "loss": 0.0884,
      "step": 985
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.9767889473211209,
      "learning_rate": 1.7098445595854924e-05,
      "loss": 0.0974,
      "step": 990
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.4010325380143416,
      "learning_rate": 1.7184801381692577e-05,
      "loss": 0.1152,
      "step": 995
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.8767740796336314,
      "learning_rate": 1.7271157167530226e-05,
      "loss": 0.0906,
      "step": 1000
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.8985824020069753,
      "learning_rate": 1.7357512953367875e-05,
      "loss": 0.1123,
      "step": 1005
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.7505161160084586,
      "learning_rate": 1.7443868739205528e-05,
      "loss": 0.0849,
      "step": 1010
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.6840014025213934,
      "learning_rate": 1.753022452504318e-05,
      "loss": 0.0988,
      "step": 1015
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.7035053998018641,
      "learning_rate": 1.761658031088083e-05,
      "loss": 0.0886,
      "step": 1020
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.8977571896188331,
      "learning_rate": 1.7702936096718482e-05,
      "loss": 0.0838,
      "step": 1025
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.5137751559858482,
      "learning_rate": 1.778929188255613e-05,
      "loss": 0.0803,
      "step": 1030
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.5990089687696084,
      "learning_rate": 1.7875647668393784e-05,
      "loss": 0.0871,
      "step": 1035
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.046378770763183,
      "learning_rate": 1.7962003454231437e-05,
      "loss": 0.0907,
      "step": 1040
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.4817479325028605,
      "learning_rate": 1.8048359240069086e-05,
      "loss": 0.0858,
      "step": 1045
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.9294482820288561,
      "learning_rate": 1.813471502590674e-05,
      "loss": 0.0748,
      "step": 1050
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.5354498293460184,
      "learning_rate": 1.8221070811744388e-05,
      "loss": 0.0939,
      "step": 1055
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.5784525074446918,
      "learning_rate": 1.8307426597582037e-05,
      "loss": 0.0843,
      "step": 1060
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.62801785518792,
      "learning_rate": 1.8393782383419693e-05,
      "loss": 0.0846,
      "step": 1065
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.5139295632612221,
      "learning_rate": 1.8480138169257342e-05,
      "loss": 0.0801,
      "step": 1070
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.9194494828553363,
      "learning_rate": 1.856649395509499e-05,
      "loss": 0.0819,
      "step": 1075
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.9742593365295164,
      "learning_rate": 1.8652849740932644e-05,
      "loss": 0.1078,
      "step": 1080
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.7692295674318097,
      "learning_rate": 1.8739205526770293e-05,
      "loss": 0.1085,
      "step": 1085
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.9845970911106678,
      "learning_rate": 1.8825561312607946e-05,
      "loss": 0.0884,
      "step": 1090
    },
    {
      "epoch": 0.47,
      "grad_norm": 2.082820314979471,
      "learning_rate": 1.89119170984456e-05,
      "loss": 0.12,
      "step": 1095
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.8117837080054929,
      "learning_rate": 1.8998272884283248e-05,
      "loss": 0.0822,
      "step": 1100
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.1460092611090402,
      "learning_rate": 1.90846286701209e-05,
      "loss": 0.1097,
      "step": 1105
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.8899569297894988,
      "learning_rate": 1.917098445595855e-05,
      "loss": 0.0886,
      "step": 1110
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.6410464393227897,
      "learning_rate": 1.9257340241796203e-05,
      "loss": 0.0789,
      "step": 1115
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.4505620588270192,
      "learning_rate": 1.9343696027633855e-05,
      "loss": 0.0909,
      "step": 1120
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.852506053614793,
      "learning_rate": 1.9430051813471504e-05,
      "loss": 0.0974,
      "step": 1125
    },
    {
      "epoch": 0.49,
      "grad_norm": 2.6689573293041553,
      "learning_rate": 1.9516407599309154e-05,
      "loss": 0.0843,
      "step": 1130
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.5230054777855191,
      "learning_rate": 1.9602763385146806e-05,
      "loss": 0.0928,
      "step": 1135
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.6142185715888877,
      "learning_rate": 1.9689119170984456e-05,
      "loss": 0.0945,
      "step": 1140
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.5872875788362734,
      "learning_rate": 1.9775474956822108e-05,
      "loss": 0.0905,
      "step": 1145
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.5007585461555959,
      "learning_rate": 1.986183074265976e-05,
      "loss": 0.0927,
      "step": 1150
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.6808047766488587,
      "learning_rate": 1.994818652849741e-05,
      "loss": 0.0873,
      "step": 1155
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.6689238734320209,
      "learning_rate": 1.9999998180950994e-05,
      "loss": 0.0806,
      "step": 1160
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.7121185292021445,
      "learning_rate": 1.9999977716657265e-05,
      "loss": 0.1034,
      "step": 1165
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.791635702401396,
      "learning_rate": 1.9999934514305234e-05,
      "loss": 0.1118,
      "step": 1170
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.7777895778365356,
      "learning_rate": 1.9999868573993138e-05,
      "loss": 0.1009,
      "step": 1175
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.7114168138902796,
      "learning_rate": 1.999977989587091e-05,
      "loss": 0.0968,
      "step": 1180
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.1214052182705263,
      "learning_rate": 1.999966848014019e-05,
      "loss": 0.0998,
      "step": 1185
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.5835670870079697,
      "learning_rate": 1.999953432705431e-05,
      "loss": 0.0913,
      "step": 1190
    },
    {
      "epoch": 0.52,
      "grad_norm": 1.0152760699918206,
      "learning_rate": 1.9999377436918314e-05,
      "loss": 0.0982,
      "step": 1195
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.3391884623808986,
      "learning_rate": 1.9999197810088943e-05,
      "loss": 0.0781,
      "step": 1200
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.5009493361172995,
      "learning_rate": 1.999899544697463e-05,
      "loss": 0.092,
      "step": 1205
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.4674731980979764,
      "learning_rate": 1.999877034803551e-05,
      "loss": 0.0895,
      "step": 1210
    },
    {
      "epoch": 0.52,
      "grad_norm": 1.9768349974485036,
      "learning_rate": 1.9998522513783415e-05,
      "loss": 0.0954,
      "step": 1215
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.6329573907921768,
      "learning_rate": 1.9998251944781878e-05,
      "loss": 0.0842,
      "step": 1220
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.5513670604045542,
      "learning_rate": 1.9997958641646122e-05,
      "loss": 0.0898,
      "step": 1225
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.2570984021594174,
      "learning_rate": 1.9997642605043056e-05,
      "loss": 0.0966,
      "step": 1230
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.0391042314097374,
      "learning_rate": 1.9997303835691292e-05,
      "loss": 0.0949,
      "step": 1235
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.6263951819312368,
      "learning_rate": 1.9996942334361126e-05,
      "loss": 0.072,
      "step": 1240
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.5995161680677363,
      "learning_rate": 1.9996558101874543e-05,
      "loss": 0.0907,
      "step": 1245
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.5405848608014154,
      "learning_rate": 1.999615113910522e-05,
      "loss": 0.0907,
      "step": 1250
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.7354793370663812,
      "learning_rate": 1.999572144697851e-05,
      "loss": 0.0825,
      "step": 1255
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.8097571767018434,
      "learning_rate": 1.9995269026471447e-05,
      "loss": 0.0873,
      "step": 1260
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.7413997346551716,
      "learning_rate": 1.9994793878612756e-05,
      "loss": 0.0907,
      "step": 1265
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.49108968812165604,
      "learning_rate": 1.9994296004482832e-05,
      "loss": 0.0775,
      "step": 1270
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.1407732391312855,
      "learning_rate": 1.9993775405213748e-05,
      "loss": 0.0883,
      "step": 1275
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.5322078982239706,
      "learning_rate": 1.9993232081989243e-05,
      "loss": 0.1029,
      "step": 1280
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.3707481376397464,
      "learning_rate": 1.9992666036044738e-05,
      "loss": 0.0879,
      "step": 1285
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.43162394310399216,
      "learning_rate": 1.999207726866731e-05,
      "loss": 0.0823,
      "step": 1290
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.39430801592296844,
      "learning_rate": 1.9991465781195706e-05,
      "loss": 0.0896,
      "step": 1295
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.5694159677621585,
      "learning_rate": 1.999083157502033e-05,
      "loss": 0.0816,
      "step": 1300
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.5060620019833312,
      "learning_rate": 1.9990174651583252e-05,
      "loss": 0.0884,
      "step": 1305
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.6293129592641865,
      "learning_rate": 1.998949501237819e-05,
      "loss": 0.0809,
      "step": 1310
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.3046083825730037,
      "learning_rate": 1.998879265895051e-05,
      "loss": 0.0835,
      "step": 1315
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.5537324247444009,
      "learning_rate": 1.998806759289724e-05,
      "loss": 0.0951,
      "step": 1320
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.5478670125082655,
      "learning_rate": 1.9987319815867037e-05,
      "loss": 0.0845,
      "step": 1325
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.7964012064137652,
      "learning_rate": 1.9986549329560204e-05,
      "loss": 0.1053,
      "step": 1330
    },
    {
      "epoch": 0.58,
      "grad_norm": 1.8201537024321701,
      "learning_rate": 1.9985756135728685e-05,
      "loss": 0.0904,
      "step": 1335
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.5932463119726163,
      "learning_rate": 1.998494023617605e-05,
      "loss": 0.0733,
      "step": 1340
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.5212259318017795,
      "learning_rate": 1.99841016327575e-05,
      "loss": 0.0829,
      "step": 1345
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.7552207970663853,
      "learning_rate": 1.9983240327379862e-05,
      "loss": 0.1056,
      "step": 1350
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.5644526446704441,
      "learning_rate": 1.9982356322001584e-05,
      "loss": 0.0815,
      "step": 1355
    },
    {
      "epoch": 0.59,
      "grad_norm": 2.312470171279909,
      "learning_rate": 1.998144961863272e-05,
      "loss": 0.0992,
      "step": 1360
    },
    {
      "epoch": 0.59,
      "grad_norm": 1.1405067091990706,
      "learning_rate": 1.9980520219334947e-05,
      "loss": 0.1161,
      "step": 1365
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.33573877882920905,
      "learning_rate": 1.9979568126221543e-05,
      "loss": 0.0862,
      "step": 1370
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.5219279117942052,
      "learning_rate": 1.997859334145739e-05,
      "loss": 0.1008,
      "step": 1375
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.5106260104942524,
      "learning_rate": 1.997759586725896e-05,
      "loss": 0.0924,
      "step": 1380
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.0248504115962325,
      "learning_rate": 1.997657570589432e-05,
      "loss": 0.1078,
      "step": 1385
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.6094328587436685,
      "learning_rate": 1.9975532859683126e-05,
      "loss": 0.0772,
      "step": 1390
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.1002689291796384,
      "learning_rate": 1.9974467330996616e-05,
      "loss": 0.1,
      "step": 1395
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.5136753011629079,
      "learning_rate": 1.99733791222576e-05,
      "loss": 0.0947,
      "step": 1400
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.7928839770448126,
      "learning_rate": 1.997226823594045e-05,
      "loss": 0.0899,
      "step": 1405
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.3194381533981615,
      "learning_rate": 1.997113467457112e-05,
      "loss": 0.0908,
      "step": 1410
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.4640425843189779,
      "learning_rate": 1.9969978440727116e-05,
      "loss": 0.0975,
      "step": 1415
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.8225929364452668,
      "learning_rate": 1.9968799537037488e-05,
      "loss": 0.0989,
      "step": 1420
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.5521492749721465,
      "learning_rate": 1.9967597966182846e-05,
      "loss": 0.0976,
      "step": 1425
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.5472673648386764,
      "learning_rate": 1.9966373730895333e-05,
      "loss": 0.0908,
      "step": 1430
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.9585374769415203,
      "learning_rate": 1.9965126833958625e-05,
      "loss": 0.0875,
      "step": 1435
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.8371145278722188,
      "learning_rate": 1.9963857278207937e-05,
      "loss": 0.0934,
      "step": 1440
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.7680571773730615,
      "learning_rate": 1.996256506652999e-05,
      "loss": 0.1097,
      "step": 1445
    },
    {
      "epoch": 0.63,
      "grad_norm": 1.0592484287359163,
      "learning_rate": 1.9961250201863037e-05,
      "loss": 0.0967,
      "step": 1450
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.44679514300900086,
      "learning_rate": 1.995991268719683e-05,
      "loss": 0.0884,
      "step": 1455
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.6909332703927928,
      "learning_rate": 1.9958552525572615e-05,
      "loss": 0.1118,
      "step": 1460
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.5668123307660959,
      "learning_rate": 1.9957169720083157e-05,
      "loss": 0.089,
      "step": 1465
    },
    {
      "epoch": 0.63,
      "grad_norm": 1.0174208845150305,
      "learning_rate": 1.9955764273872682e-05,
      "loss": 0.0996,
      "step": 1470
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.5429972018826015,
      "learning_rate": 1.995433619013692e-05,
      "loss": 0.0876,
      "step": 1475
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.9880842640326745,
      "learning_rate": 1.9952885472123055e-05,
      "loss": 0.1013,
      "step": 1480
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.7178755350757968,
      "learning_rate": 1.9951412123129744e-05,
      "loss": 0.0979,
      "step": 1485
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.3992707003588622,
      "learning_rate": 1.9949916146507112e-05,
      "loss": 0.1148,
      "step": 1490
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.631851323844705,
      "learning_rate": 1.9948397545656723e-05,
      "loss": 0.0842,
      "step": 1495
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.6374008156020984,
      "learning_rate": 1.9946856324031583e-05,
      "loss": 0.0903,
      "step": 1500
    },
    {
      "epoch": 0.65,
      "grad_norm": 2.6567821637198543,
      "learning_rate": 1.9945292485136147e-05,
      "loss": 0.0823,
      "step": 1505
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.7212264227394842,
      "learning_rate": 1.9943706032526285e-05,
      "loss": 0.0873,
      "step": 1510
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.45023488726069755,
      "learning_rate": 1.994209696980929e-05,
      "loss": 0.0826,
      "step": 1515
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.7119656947598436,
      "learning_rate": 1.9940465300643873e-05,
      "loss": 0.0891,
      "step": 1520
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.5964456186339716,
      "learning_rate": 1.993881102874013e-05,
      "loss": 0.0984,
      "step": 1525
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.6914123199335683,
      "learning_rate": 1.9937134157859564e-05,
      "loss": 0.0758,
      "step": 1530
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.5560481104608749,
      "learning_rate": 1.993543469181507e-05,
      "loss": 0.0668,
      "step": 1535
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.8057082881679304,
      "learning_rate": 1.9933712634470908e-05,
      "loss": 0.0887,
      "step": 1540
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.7710775739750066,
      "learning_rate": 1.993196798974271e-05,
      "loss": 0.0996,
      "step": 1545
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.502429831287691,
      "learning_rate": 1.993020076159747e-05,
      "loss": 0.0803,
      "step": 1550
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.5868894943798167,
      "learning_rate": 1.992841095405353e-05,
      "loss": 0.0864,
      "step": 1555
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.28510438917209935,
      "learning_rate": 1.9926598571180572e-05,
      "loss": 0.0682,
      "step": 1560
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.7590086408954494,
      "learning_rate": 1.9924763617099618e-05,
      "loss": 0.0927,
      "step": 1565
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.5037762336158287,
      "learning_rate": 1.9922906095982997e-05,
      "loss": 0.0869,
      "step": 1570
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.5921698361775859,
      "learning_rate": 1.992102601205437e-05,
      "loss": 0.0819,
      "step": 1575
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.5825079229431567,
      "learning_rate": 1.9919123369588688e-05,
      "loss": 0.0961,
      "step": 1580
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.7608364067381084,
      "learning_rate": 1.9917198172912195e-05,
      "loss": 0.1001,
      "step": 1585
    },
    {
      "epoch": 0.69,
      "grad_norm": 1.1934597870924588,
      "learning_rate": 1.9915250426402436e-05,
      "loss": 0.0933,
      "step": 1590
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.6841623546034824,
      "learning_rate": 1.991328013448821e-05,
      "loss": 0.0828,
      "step": 1595
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.6882329284891476,
      "learning_rate": 1.991128730164959e-05,
      "loss": 0.0808,
      "step": 1600
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.8720753704539425,
      "learning_rate": 1.99092719324179e-05,
      "loss": 0.0837,
      "step": 1605
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.5402610151606684,
      "learning_rate": 1.990723403137571e-05,
      "loss": 0.0916,
      "step": 1610
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.6838808424621337,
      "learning_rate": 1.9905173603156824e-05,
      "loss": 0.0923,
      "step": 1615
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.30202755881155974,
      "learning_rate": 1.990309065244626e-05,
      "loss": 0.0989,
      "step": 1620
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.6086110751842494,
      "learning_rate": 1.9900985183980256e-05,
      "loss": 0.0941,
      "step": 1625
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.6688653967743655,
      "learning_rate": 1.9898857202546255e-05,
      "loss": 0.0959,
      "step": 1630
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.7484858228329726,
      "learning_rate": 1.9896706712982877e-05,
      "loss": 0.0922,
      "step": 1635
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.676025015630956,
      "learning_rate": 1.9894533720179932e-05,
      "loss": 0.0994,
      "step": 1640
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.5491864239928795,
      "learning_rate": 1.9892338229078397e-05,
      "loss": 0.0806,
      "step": 1645
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.725976378150946,
      "learning_rate": 1.98901202446704e-05,
      "loss": 0.1067,
      "step": 1650
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.4913084517466355,
      "learning_rate": 1.9887879771999214e-05,
      "loss": 0.097,
      "step": 1655
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.7471224320617841,
      "learning_rate": 1.988561681615926e-05,
      "loss": 0.0966,
      "step": 1660
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.5014553278723048,
      "learning_rate": 1.9883331382296064e-05,
      "loss": 0.1027,
      "step": 1665
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.6818025963494562,
      "learning_rate": 1.9881023475606277e-05,
      "loss": 0.096,
      "step": 1670
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.960144563702732,
      "learning_rate": 1.9878693101337637e-05,
      "loss": 0.0967,
      "step": 1675
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.5620856263756633,
      "learning_rate": 1.9876340264788975e-05,
      "loss": 0.0749,
      "step": 1680
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.4955492724470792,
      "learning_rate": 1.9873964971310195e-05,
      "loss": 0.0903,
      "step": 1685
    },
    {
      "epoch": 0.73,
      "grad_norm": 1.2069504512746692,
      "learning_rate": 1.987156722630227e-05,
      "loss": 0.104,
      "step": 1690
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.8683018619940261,
      "learning_rate": 1.986914703521722e-05,
      "loss": 0.1089,
      "step": 1695
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.733977762749242,
      "learning_rate": 1.9866704403558092e-05,
      "loss": 0.0784,
      "step": 1700
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.6334701598336082,
      "learning_rate": 1.9864239336878978e-05,
      "loss": 0.089,
      "step": 1705
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.648143567197238,
      "learning_rate": 1.9861751840784967e-05,
      "loss": 0.0903,
      "step": 1710
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.7652710041657063,
      "learning_rate": 1.9859241920932164e-05,
      "loss": 0.0942,
      "step": 1715
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.7305568688315708,
      "learning_rate": 1.9856709583027642e-05,
      "loss": 0.1115,
      "step": 1720
    },
    {
      "epoch": 0.75,
      "grad_norm": 1.7278653299987823,
      "learning_rate": 1.985415483282946e-05,
      "loss": 0.0809,
      "step": 1725
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.6529801862227353,
      "learning_rate": 1.9851577676146644e-05,
      "loss": 0.0755,
      "step": 1730
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.4144692070134487,
      "learning_rate": 1.9848978118839155e-05,
      "loss": 0.0943,
      "step": 1735
    },
    {
      "epoch": 0.75,
      "grad_norm": 1.9734791631021336,
      "learning_rate": 1.9846356166817894e-05,
      "loss": 0.0898,
      "step": 1740
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.5459253898472208,
      "learning_rate": 1.984371182604469e-05,
      "loss": 0.0804,
      "step": 1745
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.5382475774373716,
      "learning_rate": 1.9841045102532263e-05,
      "loss": 0.0851,
      "step": 1750
    },
    {
      "epoch": 0.76,
      "grad_norm": 1.551104458429542,
      "learning_rate": 1.983835600234425e-05,
      "loss": 0.1059,
      "step": 1755
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.7868898688507436,
      "learning_rate": 1.983564453159515e-05,
      "loss": 0.1054,
      "step": 1760
    },
    {
      "epoch": 0.76,
      "grad_norm": 1.020652098910767,
      "learning_rate": 1.983291069645034e-05,
      "loss": 0.0964,
      "step": 1765
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.7107879271932049,
      "learning_rate": 1.9830154503126034e-05,
      "loss": 0.0903,
      "step": 1770
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.6594309834743366,
      "learning_rate": 1.9827375957889303e-05,
      "loss": 0.0846,
      "step": 1775
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.5263342809036547,
      "learning_rate": 1.9824575067058028e-05,
      "loss": 0.0695,
      "step": 1780
    },
    {
      "epoch": 0.77,
      "grad_norm": 1.0348679313164832,
      "learning_rate": 1.9821751837000912e-05,
      "loss": 0.0856,
      "step": 1785
    },
    {
      "epoch": 0.77,
      "grad_norm": 1.117064780483444,
      "learning_rate": 1.981890627413744e-05,
      "loss": 0.1036,
      "step": 1790
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.4180917277551505,
      "learning_rate": 1.981603838493789e-05,
      "loss": 0.0916,
      "step": 1795
    },
    {
      "epoch": 0.78,
      "grad_norm": 2.292965557154143,
      "learning_rate": 1.9813148175923295e-05,
      "loss": 0.0974,
      "step": 1800
    },
    {
      "epoch": 0.78,
      "grad_norm": 1.0931436805632515,
      "learning_rate": 1.981023565366544e-05,
      "loss": 0.0804,
      "step": 1805
    },
    {
      "epoch": 0.78,
      "grad_norm": 1.1197940353971019,
      "learning_rate": 1.980730082478686e-05,
      "loss": 0.0931,
      "step": 1810
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.8241163830027554,
      "learning_rate": 1.980434369596079e-05,
      "loss": 0.0837,
      "step": 1815
    },
    {
      "epoch": 0.79,
      "grad_norm": 1.5779394216611302,
      "learning_rate": 1.980136427391119e-05,
      "loss": 0.0959,
      "step": 1820
    },
    {
      "epoch": 0.79,
      "grad_norm": 1.5568696619522093,
      "learning_rate": 1.97983625654127e-05,
      "loss": 0.1035,
      "step": 1825
    },
    {
      "epoch": 0.79,
      "grad_norm": 5.119739750277216,
      "learning_rate": 1.979533857729064e-05,
      "loss": 0.1234,
      "step": 1830
    },
    {
      "epoch": 0.79,
      "grad_norm": 1.223109478908416,
      "learning_rate": 1.9792292316420977e-05,
      "loss": 0.0979,
      "step": 1835
    },
    {
      "epoch": 0.79,
      "grad_norm": 1.4446616152884881,
      "learning_rate": 1.978922378973034e-05,
      "loss": 0.106,
      "step": 1840
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.214455605272822,
      "learning_rate": 1.9786133004195985e-05,
      "loss": 0.0913,
      "step": 1845
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.5457682654468834,
      "learning_rate": 1.978301996684576e-05,
      "loss": 0.0979,
      "step": 1850
    },
    {
      "epoch": 0.8,
      "grad_norm": 2.3570341758823226,
      "learning_rate": 1.9779884684758134e-05,
      "loss": 0.0886,
      "step": 1855
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.7709673245293045,
      "learning_rate": 1.9776727165062138e-05,
      "loss": 0.0846,
      "step": 1860
    },
    {
      "epoch": 0.81,
      "grad_norm": 1.7004550493706148,
      "learning_rate": 1.9773547414937377e-05,
      "loss": 0.1061,
      "step": 1865
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.8904480634107864,
      "learning_rate": 1.9770345441614006e-05,
      "loss": 0.0898,
      "step": 1870
    },
    {
      "epoch": 0.81,
      "grad_norm": 1.0038580798349137,
      "learning_rate": 1.97671212523727e-05,
      "loss": 0.1083,
      "step": 1875
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.7913132768445936,
      "learning_rate": 1.976387485454466e-05,
      "loss": 0.0821,
      "step": 1880
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.8873095461861253,
      "learning_rate": 1.976060625551158e-05,
      "loss": 0.0971,
      "step": 1885
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.7699631611759321,
      "learning_rate": 1.9757315462705638e-05,
      "loss": 0.1138,
      "step": 1890
    },
    {
      "epoch": 0.82,
      "grad_norm": 1.1036033723581375,
      "learning_rate": 1.975400248360947e-05,
      "loss": 0.1006,
      "step": 1895
    },
    {
      "epoch": 0.82,
      "grad_norm": 2.026065926826909,
      "learning_rate": 1.9750667325756166e-05,
      "loss": 0.114,
      "step": 1900
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.7998712377145,
      "learning_rate": 1.9747309996729247e-05,
      "loss": 0.1032,
      "step": 1905
    },
    {
      "epoch": 0.83,
      "grad_norm": 1.05768089227519,
      "learning_rate": 1.9743930504162638e-05,
      "loss": 0.0856,
      "step": 1910
    },
    {
      "epoch": 0.83,
      "grad_norm": 1.2658434011585944,
      "learning_rate": 1.9740528855740675e-05,
      "loss": 0.0867,
      "step": 1915
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.923144746542567,
      "learning_rate": 1.9737105059198053e-05,
      "loss": 0.1023,
      "step": 1920
    },
    {
      "epoch": 0.83,
      "grad_norm": 12.069080174223988,
      "learning_rate": 1.9733659122319847e-05,
      "loss": 0.0918,
      "step": 1925
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.8369351356561107,
      "learning_rate": 1.9730191052941458e-05,
      "loss": 0.0779,
      "step": 1930
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.8064785728299022,
      "learning_rate": 1.972670085894863e-05,
      "loss": 0.0819,
      "step": 1935
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.700096547854496,
      "learning_rate": 1.972318854827739e-05,
      "loss": 0.0884,
      "step": 1940
    },
    {
      "epoch": 0.84,
      "grad_norm": 1.403023986764824,
      "learning_rate": 1.9719654128914082e-05,
      "loss": 0.0849,
      "step": 1945
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.8097419704135963,
      "learning_rate": 1.97160976088953e-05,
      "loss": 0.0842,
      "step": 1950
    },
    {
      "epoch": 0.84,
      "grad_norm": 2.1665359212528466,
      "learning_rate": 1.9712518996307902e-05,
      "loss": 0.1007,
      "step": 1955
    },
    {
      "epoch": 0.85,
      "grad_norm": 1.4963106562167054,
      "learning_rate": 1.9708918299288968e-05,
      "loss": 0.0959,
      "step": 1960
    },
    {
      "epoch": 0.85,
      "grad_norm": 1.3607731524123874,
      "learning_rate": 1.9705295526025813e-05,
      "loss": 0.1083,
      "step": 1965
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.6737531185490563,
      "learning_rate": 1.970165068475593e-05,
      "loss": 0.1055,
      "step": 1970
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.5919103617849452,
      "learning_rate": 1.9697983783767003e-05,
      "loss": 0.0932,
      "step": 1975
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.6377311219864845,
      "learning_rate": 1.969429483139687e-05,
      "loss": 0.0758,
      "step": 1980
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.9478830645476063,
      "learning_rate": 1.9690583836033514e-05,
      "loss": 0.093,
      "step": 1985
    },
    {
      "epoch": 0.86,
      "grad_norm": 6.8762210911874675,
      "learning_rate": 1.968685080611503e-05,
      "loss": 0.0913,
      "step": 1990
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.7519885869151743,
      "learning_rate": 1.968309575012963e-05,
      "loss": 0.0942,
      "step": 1995
    },
    {
      "epoch": 0.86,
      "grad_norm": 1.3317465601556326,
      "learning_rate": 1.9679318676615596e-05,
      "loss": 0.0933,
      "step": 2000
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.6132759065494825,
      "learning_rate": 1.9675519594161283e-05,
      "loss": 0.082,
      "step": 2005
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.7537614600148012,
      "learning_rate": 1.967169851140508e-05,
      "loss": 0.0951,
      "step": 2010
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.7735155086030167,
      "learning_rate": 1.9667855437035413e-05,
      "loss": 0.0938,
      "step": 2015
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.6692324440774802,
      "learning_rate": 1.9663990379790704e-05,
      "loss": 0.0735,
      "step": 2020
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.9080167579059008,
      "learning_rate": 1.966010334845936e-05,
      "loss": 0.0811,
      "step": 2025
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.4884667463437012,
      "learning_rate": 1.9656194351879757e-05,
      "loss": 0.0866,
      "step": 2030
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.3515686523041155,
      "learning_rate": 1.9652263398940216e-05,
      "loss": 0.0849,
      "step": 2035
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.5009655704966927,
      "learning_rate": 1.9648310498578972e-05,
      "loss": 0.083,
      "step": 2040
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.7143483676761784,
      "learning_rate": 1.9644335659784185e-05,
      "loss": 0.0764,
      "step": 2045
    },
    {
      "epoch": 0.89,
      "grad_norm": 1.8204976491891331,
      "learning_rate": 1.9640338891593883e-05,
      "loss": 0.1264,
      "step": 2050
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.6807437366900461,
      "learning_rate": 1.9636320203095954e-05,
      "loss": 0.1116,
      "step": 2055
    },
    {
      "epoch": 0.89,
      "grad_norm": 1.9341021097885769,
      "learning_rate": 1.9632279603428146e-05,
      "loss": 0.1026,
      "step": 2060
    },
    {
      "epoch": 0.89,
      "grad_norm": 1.179956292587945,
      "learning_rate": 1.9628217101778015e-05,
      "loss": 0.1183,
      "step": 2065
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.8899768937827349,
      "learning_rate": 1.9624132707382918e-05,
      "loss": 0.0977,
      "step": 2070
    },
    {
      "epoch": 0.9,
      "grad_norm": 1.3166443426125516,
      "learning_rate": 1.9620026429529994e-05,
      "loss": 0.1012,
      "step": 2075
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.5339079016472141,
      "learning_rate": 1.961589827755615e-05,
      "loss": 0.1082,
      "step": 2080
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.4596990220083425,
      "learning_rate": 1.961174826084802e-05,
      "loss": 0.0912,
      "step": 2085
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.3915418119448165,
      "learning_rate": 1.9607576388841953e-05,
      "loss": 0.0783,
      "step": 2090
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.6327168049132509,
      "learning_rate": 1.9603382671024e-05,
      "loss": 0.0822,
      "step": 2095
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.6150588801181179,
      "learning_rate": 1.9599167116929886e-05,
      "loss": 0.0871,
      "step": 2100
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.6436422597157241,
      "learning_rate": 1.9594929736144978e-05,
      "loss": 0.093,
      "step": 2105
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.8254699958514967,
      "learning_rate": 1.9590670538304278e-05,
      "loss": 0.096,
      "step": 2110
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.5197543328388274,
      "learning_rate": 1.9586389533092402e-05,
      "loss": 0.0918,
      "step": 2115
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.9557781768242266,
      "learning_rate": 1.9582086730243545e-05,
      "loss": 0.099,
      "step": 2120
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.5427166899498477,
      "learning_rate": 1.9577762139541465e-05,
      "loss": 0.0811,
      "step": 2125
    },
    {
      "epoch": 0.92,
      "grad_norm": 1.8756961308130387,
      "learning_rate": 1.9573415770819463e-05,
      "loss": 0.0945,
      "step": 2130
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.6779111478511379,
      "learning_rate": 1.956904763396036e-05,
      "loss": 0.0986,
      "step": 2135
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.5791588172751216,
      "learning_rate": 1.9564657738896478e-05,
      "loss": 0.0884,
      "step": 2140
    },
    {
      "epoch": 0.93,
      "grad_norm": 1.1065008243272856,
      "learning_rate": 1.95602460956096e-05,
      "loss": 0.1007,
      "step": 2145
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.7131509483548544,
      "learning_rate": 1.9555812714130974e-05,
      "loss": 0.0866,
      "step": 2150
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.4799981904596929,
      "learning_rate": 1.955135760454127e-05,
      "loss": 0.0657,
      "step": 2155
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.5250069540649991,
      "learning_rate": 1.9546880776970566e-05,
      "loss": 0.083,
      "step": 2160
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.7836812084717903,
      "learning_rate": 1.954238224159832e-05,
      "loss": 0.0882,
      "step": 2165
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.560112589183633,
      "learning_rate": 1.9537862008653347e-05,
      "loss": 0.0913,
      "step": 2170
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.6335189752801311,
      "learning_rate": 1.9533320088413807e-05,
      "loss": 0.0812,
      "step": 2175
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.8026969319535522,
      "learning_rate": 1.9528756491207167e-05,
      "loss": 0.1014,
      "step": 2180
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.5958206176680662,
      "learning_rate": 1.9524171227410185e-05,
      "loss": 0.0819,
      "step": 2185
    },
    {
      "epoch": 0.95,
      "grad_norm": 3.8784625039711,
      "learning_rate": 1.951956430744888e-05,
      "loss": 0.0782,
      "step": 2190
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.7051759938172582,
      "learning_rate": 1.951493574179852e-05,
      "loss": 0.1002,
      "step": 2195
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.9979595896719976,
      "learning_rate": 1.9510285540983583e-05,
      "loss": 0.0888,
      "step": 2200
    },
    {
      "epoch": 0.95,
      "grad_norm": 1.8283276654029694,
      "learning_rate": 1.9505613715577754e-05,
      "loss": 0.1212,
      "step": 2205
    },
    {
      "epoch": 0.95,
      "grad_norm": 1.0278569665206916,
      "learning_rate": 1.9500920276203878e-05,
      "loss": 0.1009,
      "step": 2210
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.7532073985765524,
      "learning_rate": 1.9496205233533945e-05,
      "loss": 0.1139,
      "step": 2215
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.6704633875384296,
      "learning_rate": 1.949146859828907e-05,
      "loss": 0.0977,
      "step": 2220
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.6270491613257371,
      "learning_rate": 1.9486710381239472e-05,
      "loss": 0.0874,
      "step": 2225
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.5665665710542613,
      "learning_rate": 1.9481930593204437e-05,
      "loss": 0.1056,
      "step": 2230
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.8350923101831028,
      "learning_rate": 1.947712924505229e-05,
      "loss": 0.0782,
      "step": 2235
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.8029288115147665,
      "learning_rate": 1.9472306347700397e-05,
      "loss": 0.08,
      "step": 2240
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.7177515615744224,
      "learning_rate": 1.946746191211511e-05,
      "loss": 0.0937,
      "step": 2245
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.9012213341405209,
      "learning_rate": 1.946259594931176e-05,
      "loss": 0.0954,
      "step": 2250
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.619578875231581,
      "learning_rate": 1.945770847035463e-05,
      "loss": 0.0877,
      "step": 2255
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.7866371900368929,
      "learning_rate": 1.9452799486356918e-05,
      "loss": 0.094,
      "step": 2260
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.9258671786779609,
      "learning_rate": 1.944786900848073e-05,
      "loss": 0.1004,
      "step": 2265
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.48317941437781525,
      "learning_rate": 1.9442917047937036e-05,
      "loss": 0.0789,
      "step": 2270
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.704690320595154,
      "learning_rate": 1.9437943615985663e-05,
      "loss": 0.1044,
      "step": 2275
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.7643215375132546,
      "learning_rate": 1.943294872393525e-05,
      "loss": 0.0903,
      "step": 2280
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.6086609450905198,
      "learning_rate": 1.9427932383143236e-05,
      "loss": 0.0775,
      "step": 2285
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.5769438697299223,
      "learning_rate": 1.9422894605015838e-05,
      "loss": 0.093,
      "step": 2290
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.5140997819366792,
      "learning_rate": 1.9417835401008004e-05,
      "loss": 0.0825,
      "step": 2295
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.5132532878495304,
      "learning_rate": 1.9412754782623412e-05,
      "loss": 0.0851,
      "step": 2300
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.8932621214196316,
      "learning_rate": 1.9407652761414422e-05,
      "loss": 0.0905,
      "step": 2305
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.6069639942137219,
      "learning_rate": 1.9402529348982072e-05,
      "loss": 0.0877,
      "step": 2310
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.6366735477424228,
      "learning_rate": 1.9397384556976026e-05,
      "loss": 0.0936,
      "step": 2315
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.9145181179046631,
      "eval_runtime": 352.5269,
      "eval_samples_per_second": 21.295,
      "eval_steps_per_second": 0.335,
      "step": 2315
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.5616275078056484,
      "learning_rate": 1.9392218397094575e-05,
      "loss": 0.0657,
      "step": 2320
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.6596828032006968,
      "learning_rate": 1.9387030881084588e-05,
      "loss": 0.0697,
      "step": 2325
    },
    {
      "epoch": 1.01,
      "grad_norm": 0.43755967023151665,
      "learning_rate": 1.938182202074149e-05,
      "loss": 0.0594,
      "step": 2330
    },
    {
      "epoch": 1.01,
      "grad_norm": 0.6234667120691315,
      "learning_rate": 1.9376591827909253e-05,
      "loss": 0.0645,
      "step": 2335
    },
    {
      "epoch": 1.01,
      "grad_norm": 1.3794470677309096,
      "learning_rate": 1.937134031448034e-05,
      "loss": 0.0696,
      "step": 2340
    },
    {
      "epoch": 1.01,
      "grad_norm": 0.5849723108496292,
      "learning_rate": 1.93660674923957e-05,
      "loss": 0.059,
      "step": 2345
    },
    {
      "epoch": 1.02,
      "grad_norm": 0.48370007210601973,
      "learning_rate": 1.9360773373644734e-05,
      "loss": 0.0711,
      "step": 2350
    },
    {
      "epoch": 1.02,
      "grad_norm": 1.1325120023666992,
      "learning_rate": 1.935545797026527e-05,
      "loss": 0.0547,
      "step": 2355
    },
    {
      "epoch": 1.02,
      "grad_norm": 0.28494184406753825,
      "learning_rate": 1.9350121294343526e-05,
      "loss": 0.059,
      "step": 2360
    },
    {
      "epoch": 1.02,
      "grad_norm": 0.6700705334846775,
      "learning_rate": 1.9344763358014095e-05,
      "loss": 0.0755,
      "step": 2365
    },
    {
      "epoch": 1.02,
      "grad_norm": 0.485478426214818,
      "learning_rate": 1.9339384173459913e-05,
      "loss": 0.0697,
      "step": 2370
    },
    {
      "epoch": 1.03,
      "grad_norm": 0.8524933638521246,
      "learning_rate": 1.9333983752912227e-05,
      "loss": 0.0583,
      "step": 2375
    },
    {
      "epoch": 1.03,
      "grad_norm": 0.4269504257394539,
      "learning_rate": 1.9328562108650572e-05,
      "loss": 0.0582,
      "step": 2380
    },
    {
      "epoch": 1.03,
      "grad_norm": 0.6480515153844685,
      "learning_rate": 1.932311925300274e-05,
      "loss": 0.0688,
      "step": 2385
    },
    {
      "epoch": 1.03,
      "grad_norm": 0.34585135913386694,
      "learning_rate": 1.931765519834476e-05,
      "loss": 0.0646,
      "step": 2390
    },
    {
      "epoch": 1.03,
      "grad_norm": 0.6680263661426055,
      "learning_rate": 1.9312169957100857e-05,
      "loss": 0.065,
      "step": 2395
    },
    {
      "epoch": 1.04,
      "grad_norm": 0.9926336730464236,
      "learning_rate": 1.930666354174343e-05,
      "loss": 0.0697,
      "step": 2400
    },
    {
      "epoch": 1.04,
      "grad_norm": 1.0773333832561918,
      "learning_rate": 1.9301135964793024e-05,
      "loss": 0.0701,
      "step": 2405
    },
    {
      "epoch": 1.04,
      "grad_norm": 0.8220974394525326,
      "learning_rate": 1.9295587238818313e-05,
      "loss": 0.0814,
      "step": 2410
    },
    {
      "epoch": 1.04,
      "grad_norm": 0.2662772500245477,
      "learning_rate": 1.9290017376436044e-05,
      "loss": 0.0592,
      "step": 2415
    },
    {
      "epoch": 1.05,
      "grad_norm": 6.222299959444174,
      "learning_rate": 1.9284426390311033e-05,
      "loss": 0.0702,
      "step": 2420
    },
    {
      "epoch": 1.05,
      "grad_norm": 0.6830827737400599,
      "learning_rate": 1.9278814293156122e-05,
      "loss": 0.0626,
      "step": 2425
    },
    {
      "epoch": 1.05,
      "grad_norm": 0.6648648902311897,
      "learning_rate": 1.927318109773216e-05,
      "loss": 0.0596,
      "step": 2430
    },
    {
      "epoch": 1.05,
      "grad_norm": 15.184086990125708,
      "learning_rate": 1.9267526816847972e-05,
      "loss": 0.0944,
      "step": 2435
    },
    {
      "epoch": 1.05,
      "grad_norm": 13.086549805442004,
      "learning_rate": 1.9261851463360323e-05,
      "loss": 1.138,
      "step": 2440
    },
    {
      "epoch": 1.06,
      "grad_norm": 12.903910620618289,
      "learning_rate": 1.9256155050173885e-05,
      "loss": 0.9527,
      "step": 2445
    },
    {
      "epoch": 1.06,
      "grad_norm": 5.01343033739344,
      "learning_rate": 1.9250437590241234e-05,
      "loss": 0.9946,
      "step": 2450
    },
    {
      "epoch": 1.06,
      "grad_norm": 4.334664635242735,
      "learning_rate": 1.9244699096562785e-05,
      "loss": 0.9971,
      "step": 2455
    },
    {
      "epoch": 1.06,
      "grad_norm": 3.744318496627235,
      "learning_rate": 1.9238939582186798e-05,
      "loss": 0.9953,
      "step": 2460
    },
    {
      "epoch": 1.06,
      "grad_norm": 3.0516984634704736,
      "learning_rate": 1.923315906020931e-05,
      "loss": 0.9043,
      "step": 2465
    },
    {
      "epoch": 1.07,
      "grad_norm": 4.056313270501341,
      "learning_rate": 1.9227357543774136e-05,
      "loss": 1.3543,
      "step": 2470
    },
    {
      "epoch": 1.07,
      "grad_norm": 1.9651974681311575,
      "learning_rate": 1.922153504607283e-05,
      "loss": 1.0212,
      "step": 2475
    },
    {
      "epoch": 1.07,
      "grad_norm": 4.101849169363717,
      "learning_rate": 1.9215691580344653e-05,
      "loss": 0.9517,
      "step": 2480
    },
    {
      "epoch": 1.07,
      "grad_norm": 3.1937357498817387,
      "learning_rate": 1.9209827159876538e-05,
      "loss": 1.0283,
      "step": 2485
    },
    {
      "epoch": 1.08,
      "grad_norm": 1.9804787062964042,
      "learning_rate": 1.920394179800306e-05,
      "loss": 1.17,
      "step": 2490
    },
    {
      "epoch": 1.08,
      "grad_norm": 3.9385750059156934,
      "learning_rate": 1.919803550810644e-05,
      "loss": 1.1176,
      "step": 2495
    },
    {
      "epoch": 1.08,
      "grad_norm": 3.3729979129811953,
      "learning_rate": 1.9192108303616443e-05,
      "loss": 1.1851,
      "step": 2500
    },
    {
      "epoch": 1.08,
      "grad_norm": 9.464224848094316,
      "learning_rate": 1.9186160198010424e-05,
      "loss": 1.2127,
      "step": 2505
    },
    {
      "epoch": 1.08,
      "grad_norm": 6.691608094833035,
      "learning_rate": 1.9180191204813243e-05,
      "loss": 0.9996,
      "step": 2510
    },
    {
      "epoch": 1.09,
      "grad_norm": 2.989401318952922,
      "learning_rate": 1.9174201337597266e-05,
      "loss": 1.1232,
      "step": 2515
    },
    {
      "epoch": 1.09,
      "grad_norm": 2.1641301879625052,
      "learning_rate": 1.9168190609982317e-05,
      "loss": 0.9688,
      "step": 2520
    },
    {
      "epoch": 1.09,
      "grad_norm": 2.7269265589356824,
      "learning_rate": 1.9162159035635654e-05,
      "loss": 1.1601,
      "step": 2525
    },
    {
      "epoch": 1.09,
      "grad_norm": 1.7533229665086245,
      "learning_rate": 1.9156106628271938e-05,
      "loss": 1.0813,
      "step": 2530
    },
    {
      "epoch": 1.1,
      "grad_norm": 1.1593285241535218,
      "learning_rate": 1.9150033401653193e-05,
      "loss": 1.1281,
      "step": 2535
    },
    {
      "epoch": 1.1,
      "grad_norm": 1.7658481049626809,
      "learning_rate": 1.9143939369588794e-05,
      "loss": 1.111,
      "step": 2540
    },
    {
      "epoch": 1.1,
      "grad_norm": 3.3407448124691763,
      "learning_rate": 1.913782454593542e-05,
      "loss": 1.3092,
      "step": 2545
    },
    {
      "epoch": 1.1,
      "grad_norm": 2.6070550190431905,
      "learning_rate": 1.9131688944597017e-05,
      "loss": 1.0744,
      "step": 2550
    },
    {
      "epoch": 1.1,
      "grad_norm": 2.7601896393598007,
      "learning_rate": 1.9125532579524783e-05,
      "loss": 1.2218,
      "step": 2555
    },
    {
      "epoch": 1.11,
      "grad_norm": 7.3856240027677815,
      "learning_rate": 1.911935546471713e-05,
      "loss": 1.1591,
      "step": 2560
    },
    {
      "epoch": 1.11,
      "grad_norm": 4.24382439098411,
      "learning_rate": 1.911315761421965e-05,
      "loss": 1.2513,
      "step": 2565
    },
    {
      "epoch": 1.11,
      "grad_norm": 1.8814096184126177,
      "learning_rate": 1.9106939042125085e-05,
      "loss": 1.037,
      "step": 2570
    },
    {
      "epoch": 1.11,
      "grad_norm": 1.8500629691055899,
      "learning_rate": 1.9100699762573282e-05,
      "loss": 1.0205,
      "step": 2575
    },
    {
      "epoch": 1.11,
      "grad_norm": 2.2506993444324785,
      "learning_rate": 1.909443978975119e-05,
      "loss": 1.1285,
      "step": 2580
    },
    {
      "epoch": 1.12,
      "grad_norm": 2.160603867389095,
      "learning_rate": 1.9088159137892806e-05,
      "loss": 1.2168,
      "step": 2585
    },
    {
      "epoch": 1.12,
      "grad_norm": 2.4521010037166753,
      "learning_rate": 1.908185782127914e-05,
      "loss": 0.9641,
      "step": 2590
    },
    {
      "epoch": 1.12,
      "grad_norm": 3.3405574242520215,
      "learning_rate": 1.9075535854238195e-05,
      "loss": 1.1123,
      "step": 2595
    },
    {
      "epoch": 1.12,
      "grad_norm": 1.719202294463501,
      "learning_rate": 1.906919325114493e-05,
      "loss": 1.0601,
      "step": 2600
    },
    {
      "epoch": 1.13,
      "grad_norm": 1.0275191129064691,
      "learning_rate": 1.9062830026421226e-05,
      "loss": 1.067,
      "step": 2605
    },
    {
      "epoch": 1.13,
      "grad_norm": 1.9723618781686103,
      "learning_rate": 1.905644619453585e-05,
      "loss": 1.0872,
      "step": 2610
    },
    {
      "epoch": 1.13,
      "grad_norm": 0.8049997370300347,
      "learning_rate": 1.905004177000443e-05,
      "loss": 1.132,
      "step": 2615
    },
    {
      "epoch": 1.13,
      "grad_norm": 2.8051709497363646,
      "learning_rate": 1.9043616767389416e-05,
      "loss": 1.1224,
      "step": 2620
    },
    {
      "epoch": 1.13,
      "grad_norm": 2.039428292133686,
      "learning_rate": 1.9037171201300045e-05,
      "loss": 0.7828,
      "step": 2625
    },
    {
      "epoch": 1.14,
      "grad_norm": 1.6162927800964706,
      "learning_rate": 1.9030705086392323e-05,
      "loss": 1.1454,
      "step": 2630
    },
    {
      "epoch": 1.14,
      "grad_norm": 1.512216353436176,
      "learning_rate": 1.9024218437368968e-05,
      "loss": 1.0732,
      "step": 2635
    },
    {
      "epoch": 1.14,
      "grad_norm": 1.9741080273381615,
      "learning_rate": 1.9017711268979394e-05,
      "loss": 1.1747,
      "step": 2640
    },
    {
      "epoch": 1.14,
      "grad_norm": 1.745899658973244,
      "learning_rate": 1.9011183596019672e-05,
      "loss": 1.0929,
      "step": 2645
    },
    {
      "epoch": 1.14,
      "grad_norm": 3.1272153267838902,
      "learning_rate": 1.9004635433332498e-05,
      "loss": 1.0864,
      "step": 2650
    },
    {
      "epoch": 1.15,
      "grad_norm": 2.923091983001033,
      "learning_rate": 1.899806679580715e-05,
      "loss": 1.0525,
      "step": 2655
    },
    {
      "epoch": 1.15,
      "grad_norm": 4.984968727888495,
      "learning_rate": 1.899147769837947e-05,
      "loss": 1.0726,
      "step": 2660
    },
    {
      "epoch": 1.15,
      "grad_norm": 1.8217132402018963,
      "learning_rate": 1.898486815603182e-05,
      "loss": 1.1832,
      "step": 2665
    },
    {
      "epoch": 1.15,
      "grad_norm": 1.769890651202349,
      "learning_rate": 1.8978238183793052e-05,
      "loss": 1.0143,
      "step": 2670
    },
    {
      "epoch": 1.16,
      "grad_norm": 2.878970152426382,
      "learning_rate": 1.8971587796738463e-05,
      "loss": 1.3317,
      "step": 2675
    },
    {
      "epoch": 1.16,
      "grad_norm": 3.107994186560129,
      "learning_rate": 1.8964917009989783e-05,
      "loss": 1.1506,
      "step": 2680
    },
    {
      "epoch": 1.16,
      "grad_norm": 2.115211771464286,
      "learning_rate": 1.8958225838715113e-05,
      "loss": 1.0785,
      "step": 2685
    },
    {
      "epoch": 1.16,
      "grad_norm": 1.85498934986892,
      "learning_rate": 1.8951514298128914e-05,
      "loss": 0.9929,
      "step": 2690
    },
    {
      "epoch": 1.16,
      "grad_norm": 1.0586931662100967,
      "learning_rate": 1.8944782403491964e-05,
      "loss": 0.8947,
      "step": 2695
    },
    {
      "epoch": 1.17,
      "grad_norm": 2.9650657595211563,
      "learning_rate": 1.893803017011131e-05,
      "loss": 0.9875,
      "step": 2700
    },
    {
      "epoch": 1.17,
      "grad_norm": 2.278562489218499,
      "learning_rate": 1.8931257613340262e-05,
      "loss": 1.0959,
      "step": 2705
    },
    {
      "epoch": 1.17,
      "grad_norm": 2.9520728930198263,
      "learning_rate": 1.892446474857833e-05,
      "loss": 1.1596,
      "step": 2710
    },
    {
      "epoch": 1.17,
      "grad_norm": 6.671374376765014,
      "learning_rate": 1.89176515912712e-05,
      "loss": 1.179,
      "step": 2715
    },
    {
      "epoch": 1.17,
      "grad_norm": 2.510105971294028,
      "learning_rate": 1.8910818156910707e-05,
      "loss": 1.0341,
      "step": 2720
    },
    {
      "epoch": 1.18,
      "grad_norm": 3.358943297796002,
      "learning_rate": 1.890396446103479e-05,
      "loss": 1.209,
      "step": 2725
    },
    {
      "epoch": 1.18,
      "grad_norm": 0.6924599517955392,
      "learning_rate": 1.8897090519227458e-05,
      "loss": 1.0079,
      "step": 2730
    },
    {
      "epoch": 1.18,
      "grad_norm": 1.2144311754717434,
      "learning_rate": 1.8890196347118747e-05,
      "loss": 0.95,
      "step": 2735
    },
    {
      "epoch": 1.18,
      "grad_norm": 1.628799029535916,
      "learning_rate": 1.888328196038471e-05,
      "loss": 0.999,
      "step": 2740
    },
    {
      "epoch": 1.19,
      "grad_norm": 1.6323476110763768,
      "learning_rate": 1.887634737474735e-05,
      "loss": 1.1144,
      "step": 2745
    },
    {
      "epoch": 1.19,
      "grad_norm": 0.8318865630869238,
      "learning_rate": 1.8869392605974605e-05,
      "loss": 0.8806,
      "step": 2750
    },
    {
      "epoch": 1.19,
      "grad_norm": 2.292543263234408,
      "learning_rate": 1.8862417669880307e-05,
      "loss": 1.0152,
      "step": 2755
    },
    {
      "epoch": 1.19,
      "grad_norm": 1.1718305039079837,
      "learning_rate": 1.8855422582324136e-05,
      "loss": 0.9284,
      "step": 2760
    },
    {
      "epoch": 1.19,
      "grad_norm": 1.8315532803193473,
      "learning_rate": 1.88484073592116e-05,
      "loss": 0.9595,
      "step": 2765
    },
    {
      "epoch": 1.2,
      "grad_norm": 1.0603422444623511,
      "learning_rate": 1.8841372016493987e-05,
      "loss": 1.099,
      "step": 2770
    },
    {
      "epoch": 1.2,
      "grad_norm": 1.1059393746288249,
      "learning_rate": 1.8834316570168344e-05,
      "loss": 0.9097,
      "step": 2775
    },
    {
      "epoch": 1.2,
      "grad_norm": 9.01735421988145,
      "learning_rate": 1.8827241036277415e-05,
      "loss": 1.2295,
      "step": 2780
    },
    {
      "epoch": 1.2,
      "grad_norm": 3.797225020197302,
      "learning_rate": 1.8820145430909627e-05,
      "loss": 0.9622,
      "step": 2785
    },
    {
      "epoch": 1.21,
      "grad_norm": 3.0587273399955777,
      "learning_rate": 1.881302977019905e-05,
      "loss": 1.101,
      "step": 2790
    },
    {
      "epoch": 1.21,
      "grad_norm": 2.6716610674679675,
      "learning_rate": 1.8805894070325343e-05,
      "loss": 1.0955,
      "step": 2795
    },
    {
      "epoch": 1.21,
      "grad_norm": 4.6190166161920905,
      "learning_rate": 1.8798738347513743e-05,
      "loss": 1.1255,
      "step": 2800
    },
    {
      "epoch": 1.21,
      "grad_norm": 1.2516203303283377,
      "learning_rate": 1.879156261803501e-05,
      "loss": 1.2917,
      "step": 2805
    },
    {
      "epoch": 1.21,
      "grad_norm": 1.8669010681781146,
      "learning_rate": 1.8784366898205395e-05,
      "loss": 1.0387,
      "step": 2810
    },
    {
      "epoch": 1.22,
      "grad_norm": 3.3989695616319953,
      "learning_rate": 1.8777151204386604e-05,
      "loss": 1.1687,
      "step": 2815
    },
    {
      "epoch": 1.22,
      "grad_norm": 3.9839501096580054,
      "learning_rate": 1.876991555298576e-05,
      "loss": 1.3672,
      "step": 2820
    },
    {
      "epoch": 1.22,
      "grad_norm": 2.3359598799774464,
      "learning_rate": 1.876265996045537e-05,
      "loss": 1.1501,
      "step": 2825
    },
    {
      "epoch": 1.22,
      "grad_norm": 1.7595264569997804,
      "learning_rate": 1.8755384443293273e-05,
      "loss": 0.9142,
      "step": 2830
    },
    {
      "epoch": 1.22,
      "grad_norm": 1.4577221871495234,
      "learning_rate": 1.8748089018042623e-05,
      "loss": 0.9255,
      "step": 2835
    },
    {
      "epoch": 1.23,
      "grad_norm": 1.5275145635538994,
      "learning_rate": 1.8740773701291836e-05,
      "loss": 0.8818,
      "step": 2840
    },
    {
      "epoch": 1.23,
      "grad_norm": 1.1264301878987903,
      "learning_rate": 1.873343850967456e-05,
      "loss": 0.984,
      "step": 2845
    },
    {
      "epoch": 1.23,
      "grad_norm": 2.863318146720118,
      "learning_rate": 1.8726083459869634e-05,
      "loss": 0.9505,
      "step": 2850
    },
    {
      "epoch": 1.23,
      "grad_norm": 2.8435565960567706,
      "learning_rate": 1.8718708568601047e-05,
      "loss": 1.0112,
      "step": 2855
    },
    {
      "epoch": 1.24,
      "grad_norm": 2.6724494819760025,
      "learning_rate": 1.8711313852637914e-05,
      "loss": 1.194,
      "step": 2860
    },
    {
      "epoch": 1.24,
      "grad_norm": 1.755170687903758,
      "learning_rate": 1.8703899328794412e-05,
      "loss": 1.2048,
      "step": 2865
    },
    {
      "epoch": 1.24,
      "grad_norm": 2.0242885846033074,
      "learning_rate": 1.8696465013929776e-05,
      "loss": 1.0102,
      "step": 2870
    },
    {
      "epoch": 1.24,
      "grad_norm": 2.095574086345825,
      "learning_rate": 1.8689010924948222e-05,
      "loss": 0.9868,
      "step": 2875
    },
    {
      "epoch": 1.24,
      "grad_norm": 2.166749589787262,
      "learning_rate": 1.8681537078798945e-05,
      "loss": 1.0925,
      "step": 2880
    },
    {
      "epoch": 1.25,
      "grad_norm": 3.339195291028359,
      "learning_rate": 1.8674043492476057e-05,
      "loss": 0.8562,
      "step": 2885
    },
    {
      "epoch": 1.25,
      "grad_norm": 4.246668581883667,
      "learning_rate": 1.8666530183018555e-05,
      "loss": 1.0551,
      "step": 2890
    },
    {
      "epoch": 1.25,
      "grad_norm": 1.8753108057969599,
      "learning_rate": 1.8658997167510287e-05,
      "loss": 0.811,
      "step": 2895
    },
    {
      "epoch": 1.25,
      "grad_norm": 2.6262621213201798,
      "learning_rate": 1.8651444463079898e-05,
      "loss": 0.9255,
      "step": 2900
    },
    {
      "epoch": 1.25,
      "grad_norm": 5.271501553303465,
      "learning_rate": 1.864387208690082e-05,
      "loss": 1.1943,
      "step": 2905
    },
    {
      "epoch": 1.26,
      "grad_norm": 8.306222298118406,
      "learning_rate": 1.8636280056191195e-05,
      "loss": 1.1878,
      "step": 2910
    },
    {
      "epoch": 1.26,
      "grad_norm": 1.1170321181806435,
      "learning_rate": 1.862866838821387e-05,
      "loss": 1.0038,
      "step": 2915
    },
    {
      "epoch": 1.26,
      "grad_norm": 2.2026475395614034,
      "learning_rate": 1.8621037100276342e-05,
      "loss": 1.1979,
      "step": 2920
    },
    {
      "epoch": 1.26,
      "grad_norm": 5.019325813476353,
      "learning_rate": 1.861338620973071e-05,
      "loss": 1.1805,
      "step": 2925
    },
    {
      "epoch": 1.27,
      "grad_norm": 4.975982418870631,
      "learning_rate": 1.8605715733973654e-05,
      "loss": 1.1462,
      "step": 2930
    },
    {
      "epoch": 1.27,
      "grad_norm": 2.1055702133905627,
      "learning_rate": 1.859802569044639e-05,
      "loss": 1.1622,
      "step": 2935
    },
    {
      "epoch": 1.27,
      "grad_norm": 2.2270636877747805,
      "learning_rate": 1.8590316096634615e-05,
      "loss": 1.0749,
      "step": 2940
    },
    {
      "epoch": 1.27,
      "grad_norm": 2.0870628552313883,
      "learning_rate": 1.8582586970068493e-05,
      "loss": 0.882,
      "step": 2945
    },
    {
      "epoch": 1.27,
      "grad_norm": 1.436943410024318,
      "learning_rate": 1.85748383283226e-05,
      "loss": 0.9564,
      "step": 2950
    },
    {
      "epoch": 1.28,
      "grad_norm": 2.8749289552957404,
      "learning_rate": 1.856707018901587e-05,
      "loss": 1.0634,
      "step": 2955
    },
    {
      "epoch": 1.28,
      "grad_norm": 0.9614834633177424,
      "learning_rate": 1.855928256981159e-05,
      "loss": 0.87,
      "step": 2960
    },
    {
      "epoch": 1.28,
      "grad_norm": 1.2686571202532422,
      "learning_rate": 1.8551475488417336e-05,
      "loss": 1.0995,
      "step": 2965
    },
    {
      "epoch": 1.28,
      "grad_norm": 1.5286350406981015,
      "learning_rate": 1.8543648962584932e-05,
      "loss": 1.0547,
      "step": 2970
    },
    {
      "epoch": 1.29,
      "grad_norm": 3.0835167841300257,
      "learning_rate": 1.8535803010110417e-05,
      "loss": 1.2377,
      "step": 2975
    },
    {
      "epoch": 1.29,
      "grad_norm": 2.391017656781517,
      "learning_rate": 1.8527937648834e-05,
      "loss": 1.2353,
      "step": 2980
    },
    {
      "epoch": 1.29,
      "grad_norm": 3.6789551586548352,
      "learning_rate": 1.8520052896640036e-05,
      "loss": 0.9409,
      "step": 2985
    },
    {
      "epoch": 1.29,
      "grad_norm": 1.180581592436138,
      "learning_rate": 1.8512148771456947e-05,
      "loss": 1.0151,
      "step": 2990
    },
    {
      "epoch": 1.29,
      "grad_norm": 2.9530924499547395,
      "learning_rate": 1.850422529125723e-05,
      "loss": 1.1419,
      "step": 2995
    },
    {
      "epoch": 1.3,
      "grad_norm": 2.4122900275395645,
      "learning_rate": 1.8496282474057373e-05,
      "loss": 1.0736,
      "step": 3000
    },
    {
      "epoch": 1.3,
      "grad_norm": 1.1589553877211933,
      "learning_rate": 1.8488320337917843e-05,
      "loss": 1.1795,
      "step": 3005
    },
    {
      "epoch": 1.3,
      "grad_norm": 1.6530700975476218,
      "learning_rate": 1.8480338900943034e-05,
      "loss": 1.304,
      "step": 3010
    },
    {
      "epoch": 1.3,
      "grad_norm": 2.517021072232722,
      "learning_rate": 1.8472338181281222e-05,
      "loss": 0.9507,
      "step": 3015
    },
    {
      "epoch": 1.3,
      "grad_norm": 2.220531755504523,
      "learning_rate": 1.846431819712453e-05,
      "loss": 0.9046,
      "step": 3020
    },
    {
      "epoch": 1.31,
      "grad_norm": 2.764629699311416,
      "learning_rate": 1.845627896670889e-05,
      "loss": 1.009,
      "step": 3025
    },
    {
      "epoch": 1.31,
      "grad_norm": 2.0766834198659314,
      "learning_rate": 1.8448220508313984e-05,
      "loss": 1.1998,
      "step": 3030
    },
    {
      "epoch": 1.31,
      "grad_norm": 2.9669489382279193,
      "learning_rate": 1.844014284026323e-05,
      "loss": 1.307,
      "step": 3035
    },
    {
      "epoch": 1.31,
      "grad_norm": 2.548666722268394,
      "learning_rate": 1.843204598092371e-05,
      "loss": 0.9256,
      "step": 3040
    },
    {
      "epoch": 1.32,
      "grad_norm": 2.0021583660995392,
      "learning_rate": 1.8423929948706162e-05,
      "loss": 1.2201,
      "step": 3045
    },
    {
      "epoch": 1.32,
      "grad_norm": 4.905029779245296,
      "learning_rate": 1.8415794762064898e-05,
      "loss": 0.9246,
      "step": 3050
    },
    {
      "epoch": 1.32,
      "grad_norm": 3.9342941603758885,
      "learning_rate": 1.8407640439497796e-05,
      "loss": 1.1058,
      "step": 3055
    },
    {
      "epoch": 1.32,
      "grad_norm": 3.051956807415952,
      "learning_rate": 1.839946699954625e-05,
      "loss": 0.9203,
      "step": 3060
    },
    {
      "epoch": 1.32,
      "grad_norm": 1.2845686784682366,
      "learning_rate": 1.839127446079511e-05,
      "loss": 1.1921,
      "step": 3065
    },
    {
      "epoch": 1.33,
      "grad_norm": 2.6222870108950374,
      "learning_rate": 1.838306284187266e-05,
      "loss": 0.9654,
      "step": 3070
    },
    {
      "epoch": 1.33,
      "grad_norm": 1.5701543762068435,
      "learning_rate": 1.837483216145057e-05,
      "loss": 1.1834,
      "step": 3075
    },
    {
      "epoch": 1.33,
      "grad_norm": 5.1520937052207545,
      "learning_rate": 1.8366582438243858e-05,
      "loss": 1.0738,
      "step": 3080
    },
    {
      "epoch": 1.33,
      "grad_norm": 4.697802363483906,
      "learning_rate": 1.835831369101082e-05,
      "loss": 0.9288,
      "step": 3085
    },
    {
      "epoch": 1.33,
      "grad_norm": 5.385202446007638,
      "learning_rate": 1.8350025938553035e-05,
      "loss": 0.9979,
      "step": 3090
    },
    {
      "epoch": 1.34,
      "grad_norm": 2.7177333591313473,
      "learning_rate": 1.8341719199715285e-05,
      "loss": 1.2136,
      "step": 3095
    },
    {
      "epoch": 1.34,
      "grad_norm": 3.4538389386304797,
      "learning_rate": 1.8333393493385515e-05,
      "loss": 1.0905,
      "step": 3100
    },
    {
      "epoch": 1.34,
      "grad_norm": 3.3661278616114965,
      "learning_rate": 1.8325048838494814e-05,
      "loss": 1.2895,
      "step": 3105
    },
    {
      "epoch": 1.34,
      "grad_norm": 1.2397004844541852,
      "learning_rate": 1.8316685254017347e-05,
      "loss": 1.0898,
      "step": 3110
    },
    {
      "epoch": 1.35,
      "grad_norm": 1.2760755063636564,
      "learning_rate": 1.830830275897032e-05,
      "loss": 1.0004,
      "step": 3115
    },
    {
      "epoch": 1.35,
      "grad_norm": 1.336182841450121,
      "learning_rate": 1.829990137241395e-05,
      "loss": 0.9718,
      "step": 3120
    },
    {
      "epoch": 1.35,
      "grad_norm": 1.6532051159468562,
      "learning_rate": 1.8291481113451394e-05,
      "loss": 1.2108,
      "step": 3125
    },
    {
      "epoch": 1.35,
      "grad_norm": 1.831500868110592,
      "learning_rate": 1.828304200122873e-05,
      "loss": 1.1943,
      "step": 3130
    },
    {
      "epoch": 1.35,
      "grad_norm": 5.118380579560918,
      "learning_rate": 1.8274584054934904e-05,
      "loss": 1.0709,
      "step": 3135
    },
    {
      "epoch": 1.36,
      "grad_norm": 2.64507819444403,
      "learning_rate": 1.8266107293801686e-05,
      "loss": 0.9887,
      "step": 3140
    },
    {
      "epoch": 1.36,
      "grad_norm": 2.369535703148114,
      "learning_rate": 1.8257611737103632e-05,
      "loss": 1.141,
      "step": 3145
    },
    {
      "epoch": 1.36,
      "grad_norm": 1.105441798485061,
      "learning_rate": 1.8249097404158023e-05,
      "loss": 0.9434,
      "step": 3150
    },
    {
      "epoch": 1.36,
      "grad_norm": 1.1850093992099566,
      "learning_rate": 1.8240564314324848e-05,
      "loss": 1.1414,
      "step": 3155
    },
    {
      "epoch": 1.37,
      "grad_norm": 3.0187844829993997,
      "learning_rate": 1.823201248700674e-05,
      "loss": 1.0408,
      "step": 3160
    },
    {
      "epoch": 1.37,
      "grad_norm": 2.46409735317593,
      "learning_rate": 1.8223441941648934e-05,
      "loss": 1.1907,
      "step": 3165
    },
    {
      "epoch": 1.37,
      "grad_norm": 1.1050394893702846,
      "learning_rate": 1.8214852697739232e-05,
      "loss": 1.1438,
      "step": 3170
    },
    {
      "epoch": 1.37,
      "grad_norm": 1.926128045528065,
      "learning_rate": 1.820624477480795e-05,
      "loss": 1.1118,
      "step": 3175
    },
    {
      "epoch": 1.37,
      "grad_norm": 1.9044939702214854,
      "learning_rate": 1.8197618192427874e-05,
      "loss": 1.2493,
      "step": 3180
    },
    {
      "epoch": 1.38,
      "grad_norm": 6.721570039073311,
      "learning_rate": 1.8188972970214226e-05,
      "loss": 1.1755,
      "step": 3185
    },
    {
      "epoch": 1.38,
      "grad_norm": 2.19822656617475,
      "learning_rate": 1.81803091278246e-05,
      "loss": 1.0238,
      "step": 3190
    },
    {
      "epoch": 1.38,
      "grad_norm": 2.4481511003429963,
      "learning_rate": 1.8171626684958944e-05,
      "loss": 1.0708,
      "step": 3195
    },
    {
      "epoch": 1.38,
      "grad_norm": 1.0407019264410284,
      "learning_rate": 1.8162925661359487e-05,
      "loss": 0.9963,
      "step": 3200
    },
    {
      "epoch": 1.38,
      "grad_norm": 1.0870254854331745,
      "learning_rate": 1.8154206076810706e-05,
      "loss": 1.1444,
      "step": 3205
    },
    {
      "epoch": 1.39,
      "grad_norm": 1.008606553113266,
      "learning_rate": 1.8145467951139294e-05,
      "loss": 1.0423,
      "step": 3210
    },
    {
      "epoch": 1.39,
      "grad_norm": 2.7404259988281,
      "learning_rate": 1.813671130421409e-05,
      "loss": 1.0753,
      "step": 3215
    },
    {
      "epoch": 1.39,
      "grad_norm": 3.822856198318866,
      "learning_rate": 1.8127936155946062e-05,
      "loss": 1.1881,
      "step": 3220
    },
    {
      "epoch": 1.39,
      "grad_norm": 1.763286636104949,
      "learning_rate": 1.8119142526288237e-05,
      "loss": 1.1203,
      "step": 3225
    },
    {
      "epoch": 1.4,
      "grad_norm": 1.3741982296746016,
      "learning_rate": 1.8110330435235656e-05,
      "loss": 1.1767,
      "step": 3230
    },
    {
      "epoch": 1.4,
      "grad_norm": 3.5252778102399747,
      "learning_rate": 1.8101499902825354e-05,
      "loss": 1.0326,
      "step": 3235
    },
    {
      "epoch": 1.4,
      "grad_norm": 0.6361941017976659,
      "learning_rate": 1.8092650949136295e-05,
      "loss": 1.2773,
      "step": 3240
    },
    {
      "epoch": 1.4,
      "grad_norm": 1.0642655331813502,
      "learning_rate": 1.808378359428932e-05,
      "loss": 1.0846,
      "step": 3245
    },
    {
      "epoch": 1.4,
      "grad_norm": 1.300245514729484,
      "learning_rate": 1.8074897858447123e-05,
      "loss": 0.9965,
      "step": 3250
    },
    {
      "epoch": 1.41,
      "grad_norm": 1.8676058682990833,
      "learning_rate": 1.806599376181418e-05,
      "loss": 1.0315,
      "step": 3255
    },
    {
      "epoch": 1.41,
      "grad_norm": 1.8478689368260564,
      "learning_rate": 1.805707132463673e-05,
      "loss": 1.0647,
      "step": 3260
    },
    {
      "epoch": 1.41,
      "grad_norm": 1.5380958946700751,
      "learning_rate": 1.8048130567202707e-05,
      "loss": 1.204,
      "step": 3265
    },
    {
      "epoch": 1.41,
      "grad_norm": 1.1114063704861679,
      "learning_rate": 1.8039171509841697e-05,
      "loss": 1.0207,
      "step": 3270
    },
    {
      "epoch": 1.41,
      "grad_norm": 1.8248625661555662,
      "learning_rate": 1.8030194172924906e-05,
      "loss": 1.3113,
      "step": 3275
    },
    {
      "epoch": 1.42,
      "grad_norm": 2.4034069263852764,
      "learning_rate": 1.80211985768651e-05,
      "loss": 1.1165,
      "step": 3280
    },
    {
      "epoch": 1.42,
      "grad_norm": 1.6590028307791285,
      "learning_rate": 1.8012184742116566e-05,
      "loss": 1.1675,
      "step": 3285
    },
    {
      "epoch": 1.42,
      "grad_norm": 2.4344348218555933,
      "learning_rate": 1.8003152689175055e-05,
      "loss": 0.9994,
      "step": 3290
    },
    {
      "epoch": 1.42,
      "grad_norm": 2.0011634617126592,
      "learning_rate": 1.7994102438577753e-05,
      "loss": 1.1049,
      "step": 3295
    },
    {
      "epoch": 1.43,
      "grad_norm": 1.8236411685650236,
      "learning_rate": 1.7985034010903212e-05,
      "loss": 1.0944,
      "step": 3300
    },
    {
      "epoch": 1.43,
      "grad_norm": 1.4601573929509397,
      "learning_rate": 1.7975947426771327e-05,
      "loss": 1.2179,
      "step": 3305
    },
    {
      "epoch": 1.43,
      "grad_norm": 4.6685454697583015,
      "learning_rate": 1.796684270684327e-05,
      "loss": 1.3198,
      "step": 3310
    },
    {
      "epoch": 1.43,
      "grad_norm": 5.895936325447172,
      "learning_rate": 1.795771987182145e-05,
      "loss": 1.085,
      "step": 3315
    },
    {
      "epoch": 1.43,
      "grad_norm": 3.0349834958019843,
      "learning_rate": 1.794857894244947e-05,
      "loss": 1.243,
      "step": 3320
    },
    {
      "epoch": 1.44,
      "grad_norm": 2.4051043441438447,
      "learning_rate": 1.793941993951208e-05,
      "loss": 1.0862,
      "step": 3325
    },
    {
      "epoch": 1.44,
      "grad_norm": 5.537290266543172,
      "learning_rate": 1.7930242883835114e-05,
      "loss": 0.9081,
      "step": 3330
    },
    {
      "epoch": 1.44,
      "grad_norm": 2.000136679317919,
      "learning_rate": 1.7921047796285464e-05,
      "loss": 0.9509,
      "step": 3335
    },
    {
      "epoch": 1.44,
      "grad_norm": 2.0237555621206975,
      "learning_rate": 1.7911834697771017e-05,
      "loss": 1.0043,
      "step": 3340
    },
    {
      "epoch": 1.44,
      "grad_norm": 2.876113888412484,
      "learning_rate": 1.7902603609240623e-05,
      "loss": 0.9543,
      "step": 3345
    },
    {
      "epoch": 1.45,
      "grad_norm": 6.833172198802181,
      "learning_rate": 1.789335455168403e-05,
      "loss": 1.1886,
      "step": 3350
    },
    {
      "epoch": 1.45,
      "grad_norm": 2.8740987465913173,
      "learning_rate": 1.788408754613184e-05,
      "loss": 1.016,
      "step": 3355
    },
    {
      "epoch": 1.45,
      "grad_norm": 1.9149073164585646,
      "learning_rate": 1.7874802613655478e-05,
      "loss": 0.9419,
      "step": 3360
    },
    {
      "epoch": 1.45,
      "grad_norm": 4.205668899674492,
      "learning_rate": 1.786549977536712e-05,
      "loss": 1.0141,
      "step": 3365
    },
    {
      "epoch": 1.46,
      "grad_norm": 3.7286171999927227,
      "learning_rate": 1.7856179052419667e-05,
      "loss": 1.0132,
      "step": 3370
    },
    {
      "epoch": 1.46,
      "grad_norm": 3.555982799960603,
      "learning_rate": 1.7846840466006675e-05,
      "loss": 0.9953,
      "step": 3375
    },
    {
      "epoch": 1.46,
      "grad_norm": 2.7729455636026015,
      "learning_rate": 1.783748403736232e-05,
      "loss": 0.9252,
      "step": 3380
    },
    {
      "epoch": 1.46,
      "grad_norm": 5.2333848411684905,
      "learning_rate": 1.7828109787761364e-05,
      "loss": 1.03,
      "step": 3385
    },
    {
      "epoch": 1.46,
      "grad_norm": 4.649862211067886,
      "learning_rate": 1.7818717738519064e-05,
      "loss": 1.1632,
      "step": 3390
    },
    {
      "epoch": 1.47,
      "grad_norm": 2.4148563124642726,
      "learning_rate": 1.7809307910991177e-05,
      "loss": 0.8583,
      "step": 3395
    },
    {
      "epoch": 1.47,
      "grad_norm": 3.7308969676385573,
      "learning_rate": 1.779988032657386e-05,
      "loss": 0.921,
      "step": 3400
    },
    {
      "epoch": 1.47,
      "grad_norm": 10.233259023305683,
      "learning_rate": 1.7790435006703663e-05,
      "loss": 1.216,
      "step": 3405
    },
    {
      "epoch": 1.47,
      "grad_norm": 4.750028602895991,
      "learning_rate": 1.7780971972857454e-05,
      "loss": 1.3289,
      "step": 3410
    },
    {
      "epoch": 1.48,
      "grad_norm": 1.605503970424356,
      "learning_rate": 1.7771491246552388e-05,
      "loss": 0.8949,
      "step": 3415
    },
    {
      "epoch": 1.48,
      "grad_norm": 1.7687292828558763,
      "learning_rate": 1.776199284934584e-05,
      "loss": 1.0494,
      "step": 3420
    },
    {
      "epoch": 1.48,
      "grad_norm": 5.172907684395909,
      "learning_rate": 1.775247680283537e-05,
      "loss": 0.8951,
      "step": 3425
    },
    {
      "epoch": 1.48,
      "grad_norm": 1.759596035216593,
      "learning_rate": 1.7742943128658664e-05,
      "loss": 0.9194,
      "step": 3430
    },
    {
      "epoch": 1.48,
      "grad_norm": 2.40491220347799,
      "learning_rate": 1.77333918484935e-05,
      "loss": 0.9912,
      "step": 3435
    },
    {
      "epoch": 1.49,
      "grad_norm": 3.0396454651265197,
      "learning_rate": 1.7723822984057682e-05,
      "loss": 0.9934,
      "step": 3440
    },
    {
      "epoch": 1.49,
      "grad_norm": 5.964381710275646,
      "learning_rate": 1.7714236557108998e-05,
      "loss": 0.9603,
      "step": 3445
    },
    {
      "epoch": 1.49,
      "grad_norm": 3.0976228058993627,
      "learning_rate": 1.7704632589445167e-05,
      "loss": 0.9246,
      "step": 3450
    },
    {
      "epoch": 1.49,
      "grad_norm": 22.29026429234738,
      "learning_rate": 1.7695011102903796e-05,
      "loss": 1.0157,
      "step": 3455
    },
    {
      "epoch": 1.49,
      "grad_norm": 10.692379917639276,
      "learning_rate": 1.768537211936233e-05,
      "loss": 0.9716,
      "step": 3460
    },
    {
      "epoch": 1.5,
      "grad_norm": 3.9612716203667016,
      "learning_rate": 1.767571566073799e-05,
      "loss": 0.8151,
      "step": 3465
    },
    {
      "epoch": 1.5,
      "grad_norm": 1.464022924418229,
      "learning_rate": 1.7666041748987744e-05,
      "loss": 0.9119,
      "step": 3470
    },
    {
      "epoch": 1.5,
      "grad_norm": 1.2077797246084108,
      "learning_rate": 1.7656350406108228e-05,
      "loss": 0.7946,
      "step": 3475
    },
    {
      "epoch": 1.5,
      "grad_norm": 1.296998157972292,
      "learning_rate": 1.764664165413573e-05,
      "loss": 0.905,
      "step": 3480
    },
    {
      "epoch": 1.51,
      "grad_norm": 1.6098965454363554,
      "learning_rate": 1.763691551514611e-05,
      "loss": 0.8778,
      "step": 3485
    },
    {
      "epoch": 1.51,
      "grad_norm": 3.6816917985635995,
      "learning_rate": 1.7627172011254777e-05,
      "loss": 1.0083,
      "step": 3490
    },
    {
      "epoch": 1.51,
      "grad_norm": 3.8188571824489963,
      "learning_rate": 1.7617411164616613e-05,
      "loss": 0.9896,
      "step": 3495
    },
    {
      "epoch": 1.51,
      "grad_norm": 4.920995898630756,
      "learning_rate": 1.7607632997425936e-05,
      "loss": 1.0074,
      "step": 3500
    },
    {
      "epoch": 1.51,
      "grad_norm": 4.921413022519996,
      "learning_rate": 1.7597837531916447e-05,
      "loss": 1.1118,
      "step": 3505
    },
    {
      "epoch": 1.52,
      "grad_norm": 7.1316673838198925,
      "learning_rate": 1.758802479036119e-05,
      "loss": 1.0156,
      "step": 3510
    },
    {
      "epoch": 1.52,
      "grad_norm": 3.939674963752455,
      "learning_rate": 1.7578194795072474e-05,
      "loss": 0.9165,
      "step": 3515
    },
    {
      "epoch": 1.52,
      "grad_norm": 2.5775877484888063,
      "learning_rate": 1.7568347568401857e-05,
      "loss": 1.0156,
      "step": 3520
    },
    {
      "epoch": 1.52,
      "grad_norm": 1.5703752754794065,
      "learning_rate": 1.7558483132740066e-05,
      "loss": 0.9154,
      "step": 3525
    },
    {
      "epoch": 1.52,
      "grad_norm": 1.4765467973691762,
      "learning_rate": 1.7548601510516965e-05,
      "loss": 0.9683,
      "step": 3530
    },
    {
      "epoch": 1.53,
      "grad_norm": 1.4821344651743134,
      "learning_rate": 1.7538702724201492e-05,
      "loss": 0.8375,
      "step": 3535
    },
    {
      "epoch": 1.53,
      "grad_norm": 1.0318729292387396,
      "learning_rate": 1.7528786796301615e-05,
      "loss": 0.8248,
      "step": 3540
    },
    {
      "epoch": 1.53,
      "grad_norm": 2.2167517550488425,
      "learning_rate": 1.7518853749364283e-05,
      "loss": 0.9371,
      "step": 3545
    },
    {
      "epoch": 1.53,
      "grad_norm": 2.876447434863973,
      "learning_rate": 1.750890360597536e-05,
      "loss": 0.9319,
      "step": 3550
    },
    {
      "epoch": 1.54,
      "grad_norm": 4.626600470000748,
      "learning_rate": 1.7498936388759596e-05,
      "loss": 0.8646,
      "step": 3555
    },
    {
      "epoch": 1.54,
      "grad_norm": 3.0238441858852942,
      "learning_rate": 1.7488952120380556e-05,
      "loss": 0.9372,
      "step": 3560
    },
    {
      "epoch": 1.54,
      "grad_norm": 3.3222120957614663,
      "learning_rate": 1.747895082354058e-05,
      "loss": 0.8487,
      "step": 3565
    },
    {
      "epoch": 1.54,
      "grad_norm": 8.28429504970897,
      "learning_rate": 1.7468932520980722e-05,
      "loss": 0.9472,
      "step": 3570
    },
    {
      "epoch": 1.54,
      "grad_norm": 2.807509123378574,
      "learning_rate": 1.745889723548071e-05,
      "loss": 1.0485,
      "step": 3575
    },
    {
      "epoch": 1.55,
      "grad_norm": 5.081306727094978,
      "learning_rate": 1.7448844989858886e-05,
      "loss": 0.7305,
      "step": 3580
    },
    {
      "epoch": 1.55,
      "grad_norm": 3.429895318349554,
      "learning_rate": 1.7438775806972153e-05,
      "loss": 0.8123,
      "step": 3585
    },
    {
      "epoch": 1.55,
      "grad_norm": 3.4459765311838515,
      "learning_rate": 1.7428689709715928e-05,
      "loss": 0.8061,
      "step": 3590
    },
    {
      "epoch": 1.55,
      "grad_norm": 4.417656520258387,
      "learning_rate": 1.7418586721024096e-05,
      "loss": 0.8281,
      "step": 3595
    },
    {
      "epoch": 1.56,
      "grad_norm": 3.502368126389993,
      "learning_rate": 1.7408466863868935e-05,
      "loss": 0.8285,
      "step": 3600
    },
    {
      "epoch": 1.56,
      "grad_norm": 2.1996075307453475,
      "learning_rate": 1.739833016126109e-05,
      "loss": 0.7089,
      "step": 3605
    },
    {
      "epoch": 1.56,
      "grad_norm": 1.5582767433616693,
      "learning_rate": 1.7388176636249508e-05,
      "loss": 0.7903,
      "step": 3610
    },
    {
      "epoch": 1.56,
      "grad_norm": 1.9699367067357896,
      "learning_rate": 1.737800631192138e-05,
      "loss": 0.8573,
      "step": 3615
    },
    {
      "epoch": 1.56,
      "grad_norm": 5.25889815456181,
      "learning_rate": 1.7367819211402102e-05,
      "loss": 0.9748,
      "step": 3620
    },
    {
      "epoch": 1.57,
      "grad_norm": 4.646179346178722,
      "learning_rate": 1.7357615357855214e-05,
      "loss": 0.9594,
      "step": 3625
    },
    {
      "epoch": 1.57,
      "grad_norm": 6.136196120838206,
      "learning_rate": 1.7347394774482352e-05,
      "loss": 0.8544,
      "step": 3630
    },
    {
      "epoch": 1.57,
      "grad_norm": 0.8175188488490657,
      "learning_rate": 1.733715748452319e-05,
      "loss": 0.806,
      "step": 3635
    },
    {
      "epoch": 1.57,
      "grad_norm": 3.460932815967537,
      "learning_rate": 1.7326903511255385e-05,
      "loss": 0.7996,
      "step": 3640
    },
    {
      "epoch": 1.57,
      "grad_norm": 2.878479232664424,
      "learning_rate": 1.7316632877994536e-05,
      "loss": 1.0603,
      "step": 3645
    },
    {
      "epoch": 1.58,
      "grad_norm": 2.8505217948290817,
      "learning_rate": 1.7306345608094118e-05,
      "loss": 0.7371,
      "step": 3650
    },
    {
      "epoch": 1.58,
      "grad_norm": 4.88241799283637,
      "learning_rate": 1.7296041724945444e-05,
      "loss": 0.8434,
      "step": 3655
    },
    {
      "epoch": 1.58,
      "grad_norm": 1.741185669970214,
      "learning_rate": 1.7285721251977587e-05,
      "loss": 1.04,
      "step": 3660
    },
    {
      "epoch": 1.58,
      "grad_norm": 1.3676240280021945,
      "learning_rate": 1.727538421265736e-05,
      "loss": 0.9171,
      "step": 3665
    },
    {
      "epoch": 1.59,
      "grad_norm": 2.8829408528987894,
      "learning_rate": 1.7265030630489225e-05,
      "loss": 0.8408,
      "step": 3670
    },
    {
      "epoch": 1.59,
      "grad_norm": 2.4303674261915633,
      "learning_rate": 1.725466052901528e-05,
      "loss": 0.7715,
      "step": 3675
    },
    {
      "epoch": 1.59,
      "grad_norm": 3.0475458834050135,
      "learning_rate": 1.724427393181517e-05,
      "loss": 0.9401,
      "step": 3680
    },
    {
      "epoch": 1.59,
      "grad_norm": 7.430831583314274,
      "learning_rate": 1.7233870862506056e-05,
      "loss": 0.8857,
      "step": 3685
    },
    {
      "epoch": 1.59,
      "grad_norm": 2.4942060149245893,
      "learning_rate": 1.7223451344742543e-05,
      "loss": 0.8022,
      "step": 3690
    },
    {
      "epoch": 1.6,
      "grad_norm": 3.6192672645199626,
      "learning_rate": 1.721301540221665e-05,
      "loss": 0.7113,
      "step": 3695
    },
    {
      "epoch": 1.6,
      "grad_norm": 5.739159713290128,
      "learning_rate": 1.7202563058657742e-05,
      "loss": 0.8907,
      "step": 3700
    },
    {
      "epoch": 1.6,
      "grad_norm": 3.707239332984955,
      "learning_rate": 1.719209433783246e-05,
      "loss": 0.9802,
      "step": 3705
    },
    {
      "epoch": 1.6,
      "grad_norm": 4.992264989017978,
      "learning_rate": 1.7181609263544706e-05,
      "loss": 0.7683,
      "step": 3710
    },
    {
      "epoch": 1.6,
      "grad_norm": 2.2368251184952026,
      "learning_rate": 1.717110785963555e-05,
      "loss": 0.7963,
      "step": 3715
    },
    {
      "epoch": 1.61,
      "grad_norm": 4.286388500117645,
      "learning_rate": 1.7160590149983195e-05,
      "loss": 0.9671,
      "step": 3720
    },
    {
      "epoch": 1.61,
      "grad_norm": 2.6902602240723628,
      "learning_rate": 1.715005615850293e-05,
      "loss": 0.8457,
      "step": 3725
    },
    {
      "epoch": 1.61,
      "grad_norm": 2.476741017909578,
      "learning_rate": 1.713950590914706e-05,
      "loss": 0.8622,
      "step": 3730
    },
    {
      "epoch": 1.61,
      "grad_norm": 11.302682992388872,
      "learning_rate": 1.7128939425904857e-05,
      "loss": 0.8139,
      "step": 3735
    },
    {
      "epoch": 1.62,
      "grad_norm": 5.760632967816922,
      "learning_rate": 1.7118356732802502e-05,
      "loss": 0.9834,
      "step": 3740
    },
    {
      "epoch": 1.62,
      "grad_norm": 4.72617092213642,
      "learning_rate": 1.710775785390304e-05,
      "loss": 0.8057,
      "step": 3745
    },
    {
      "epoch": 1.62,
      "grad_norm": 2.867135354009946,
      "learning_rate": 1.709714281330632e-05,
      "loss": 0.7903,
      "step": 3750
    },
    {
      "epoch": 1.62,
      "grad_norm": 5.584505123973735,
      "learning_rate": 1.7086511635148926e-05,
      "loss": 1.0098,
      "step": 3755
    },
    {
      "epoch": 1.62,
      "grad_norm": 1.3507238105944468,
      "learning_rate": 1.7075864343604162e-05,
      "loss": 0.6761,
      "step": 3760
    },
    {
      "epoch": 1.63,
      "grad_norm": 5.062844299889445,
      "learning_rate": 1.7065200962881946e-05,
      "loss": 0.9111,
      "step": 3765
    },
    {
      "epoch": 1.63,
      "grad_norm": 1.420206926325427,
      "learning_rate": 1.7054521517228792e-05,
      "loss": 0.9262,
      "step": 3770
    },
    {
      "epoch": 1.63,
      "grad_norm": 3.400362820427134,
      "learning_rate": 1.704382603092774e-05,
      "loss": 0.7034,
      "step": 3775
    },
    {
      "epoch": 1.63,
      "grad_norm": 13.410473444763928,
      "learning_rate": 1.7033114528298304e-05,
      "loss": 0.8797,
      "step": 3780
    },
    {
      "epoch": 1.63,
      "grad_norm": 4.227659804805431,
      "learning_rate": 1.7022387033696413e-05,
      "loss": 0.8266,
      "step": 3785
    },
    {
      "epoch": 1.64,
      "grad_norm": 11.123371302509794,
      "learning_rate": 1.7011643571514365e-05,
      "loss": 0.8825,
      "step": 3790
    },
    {
      "epoch": 1.64,
      "grad_norm": 7.0109121600981466,
      "learning_rate": 1.7000884166180754e-05,
      "loss": 1.0255,
      "step": 3795
    },
    {
      "epoch": 1.64,
      "grad_norm": 3.233226584132477,
      "learning_rate": 1.6990108842160445e-05,
      "loss": 0.8432,
      "step": 3800
    },
    {
      "epoch": 1.64,
      "grad_norm": 2.6243800886514235,
      "learning_rate": 1.6979317623954476e-05,
      "loss": 0.7495,
      "step": 3805
    },
    {
      "epoch": 1.65,
      "grad_norm": 4.804997476595591,
      "learning_rate": 1.6968510536100037e-05,
      "loss": 0.7711,
      "step": 3810
    },
    {
      "epoch": 1.65,
      "grad_norm": 3.084026561599956,
      "learning_rate": 1.6957687603170413e-05,
      "loss": 0.8088,
      "step": 3815
    },
    {
      "epoch": 1.65,
      "grad_norm": 8.949623608113457,
      "learning_rate": 1.6946848849774894e-05,
      "loss": 0.8532,
      "step": 3820
    },
    {
      "epoch": 1.65,
      "grad_norm": 4.778283870889809,
      "learning_rate": 1.693599430055876e-05,
      "loss": 0.9558,
      "step": 3825
    },
    {
      "epoch": 1.65,
      "grad_norm": 4.587210475944901,
      "learning_rate": 1.692512398020321e-05,
      "loss": 0.7157,
      "step": 3830
    },
    {
      "epoch": 1.66,
      "grad_norm": 1.7434908938272295,
      "learning_rate": 1.691423791342529e-05,
      "loss": 0.877,
      "step": 3835
    },
    {
      "epoch": 1.66,
      "grad_norm": 1.6581151680906558,
      "learning_rate": 1.6903336124977858e-05,
      "loss": 0.8128,
      "step": 3840
    },
    {
      "epoch": 1.66,
      "grad_norm": 2.8079967808156883,
      "learning_rate": 1.689241863964952e-05,
      "loss": 0.9081,
      "step": 3845
    },
    {
      "epoch": 1.66,
      "grad_norm": 0.7986015484972644,
      "learning_rate": 1.688148548226457e-05,
      "loss": 0.7597,
      "step": 3850
    },
    {
      "epoch": 1.67,
      "grad_norm": 4.498669451011208,
      "learning_rate": 1.687053667768295e-05,
      "loss": 0.7931,
      "step": 3855
    },
    {
      "epoch": 1.67,
      "grad_norm": 3.124000175896055,
      "learning_rate": 1.685957225080016e-05,
      "loss": 0.7858,
      "step": 3860
    },
    {
      "epoch": 1.67,
      "grad_norm": 4.364340657244522,
      "learning_rate": 1.684859222654724e-05,
      "loss": 0.9062,
      "step": 3865
    },
    {
      "epoch": 1.67,
      "grad_norm": 7.350243693509682,
      "learning_rate": 1.6837596629890683e-05,
      "loss": 1.05,
      "step": 3870
    },
    {
      "epoch": 1.67,
      "grad_norm": 3.2882320910324405,
      "learning_rate": 1.6826585485832408e-05,
      "loss": 0.8192,
      "step": 3875
    },
    {
      "epoch": 1.68,
      "grad_norm": 4.778392575192595,
      "learning_rate": 1.6815558819409662e-05,
      "loss": 0.9379,
      "step": 3880
    },
    {
      "epoch": 1.68,
      "grad_norm": 3.318104844054457,
      "learning_rate": 1.6804516655695003e-05,
      "loss": 0.8725,
      "step": 3885
    },
    {
      "epoch": 1.68,
      "grad_norm": 2.264270858290099,
      "learning_rate": 1.6793459019796226e-05,
      "loss": 0.6963,
      "step": 3890
    },
    {
      "epoch": 1.68,
      "grad_norm": 4.01933074861301,
      "learning_rate": 1.67823859368563e-05,
      "loss": 0.805,
      "step": 3895
    },
    {
      "epoch": 1.68,
      "grad_norm": 2.9667750183356283,
      "learning_rate": 1.6771297432053324e-05,
      "loss": 0.7698,
      "step": 3900
    },
    {
      "epoch": 1.69,
      "grad_norm": 2.8413147859622034,
      "learning_rate": 1.6760193530600463e-05,
      "loss": 0.8624,
      "step": 3905
    },
    {
      "epoch": 1.69,
      "grad_norm": 0.8762946124497302,
      "learning_rate": 1.6749074257745886e-05,
      "loss": 0.7549,
      "step": 3910
    },
    {
      "epoch": 1.69,
      "grad_norm": 1.516106158568189,
      "learning_rate": 1.6737939638772717e-05,
      "loss": 1.0617,
      "step": 3915
    },
    {
      "epoch": 1.69,
      "grad_norm": 2.6633739389265387,
      "learning_rate": 1.6726789698998975e-05,
      "loss": 0.7438,
      "step": 3920
    },
    {
      "epoch": 1.7,
      "grad_norm": 5.437903737579578,
      "learning_rate": 1.6715624463777517e-05,
      "loss": 0.879,
      "step": 3925
    },
    {
      "epoch": 1.7,
      "grad_norm": 1.5278118733471453,
      "learning_rate": 1.6704443958495968e-05,
      "loss": 0.9221,
      "step": 3930
    },
    {
      "epoch": 1.7,
      "grad_norm": 1.463363203352676,
      "learning_rate": 1.6693248208576692e-05,
      "loss": 0.847,
      "step": 3935
    },
    {
      "epoch": 1.7,
      "grad_norm": 1.8308055767789655,
      "learning_rate": 1.6682037239476704e-05,
      "loss": 0.715,
      "step": 3940
    },
    {
      "epoch": 1.7,
      "grad_norm": 3.1489574436454206,
      "learning_rate": 1.6670811076687625e-05,
      "loss": 0.8082,
      "step": 3945
    },
    {
      "epoch": 1.71,
      "grad_norm": 1.7250411190465171,
      "learning_rate": 1.665956974573563e-05,
      "loss": 0.781,
      "step": 3950
    },
    {
      "epoch": 1.71,
      "grad_norm": 1.3382497564513038,
      "learning_rate": 1.6648313272181376e-05,
      "loss": 0.7784,
      "step": 3955
    },
    {
      "epoch": 1.71,
      "grad_norm": 3.455075057480329,
      "learning_rate": 1.6637041681619965e-05,
      "loss": 0.8104,
      "step": 3960
    },
    {
      "epoch": 1.71,
      "grad_norm": 5.418598141306543,
      "learning_rate": 1.6625754999680848e-05,
      "loss": 0.7615,
      "step": 3965
    },
    {
      "epoch": 1.71,
      "grad_norm": 4.440826447631677,
      "learning_rate": 1.6614453252027812e-05,
      "loss": 0.8031,
      "step": 3970
    },
    {
      "epoch": 1.72,
      "grad_norm": 3.70035610129939,
      "learning_rate": 1.6603136464358904e-05,
      "loss": 0.773,
      "step": 3975
    },
    {
      "epoch": 1.72,
      "grad_norm": 5.716173129679144,
      "learning_rate": 1.6591804662406338e-05,
      "loss": 0.7824,
      "step": 3980
    },
    {
      "epoch": 1.72,
      "grad_norm": 1.5729883041830204,
      "learning_rate": 1.6580457871936506e-05,
      "loss": 0.8418,
      "step": 3985
    },
    {
      "epoch": 1.72,
      "grad_norm": 6.022112190463788,
      "learning_rate": 1.6569096118749858e-05,
      "loss": 0.7688,
      "step": 3990
    },
    {
      "epoch": 1.73,
      "grad_norm": 1.3221927220586283,
      "learning_rate": 1.6557719428680872e-05,
      "loss": 0.758,
      "step": 3995
    },
    {
      "epoch": 1.73,
      "grad_norm": 3.799747978987853,
      "learning_rate": 1.654632782759799e-05,
      "loss": 0.8551,
      "step": 4000
    },
    {
      "epoch": 1.73,
      "grad_norm": 1.2942585118978511,
      "learning_rate": 1.653492134140356e-05,
      "loss": 0.7883,
      "step": 4005
    },
    {
      "epoch": 1.73,
      "grad_norm": 2.7212654259408304,
      "learning_rate": 1.652349999603377e-05,
      "loss": 0.8271,
      "step": 4010
    },
    {
      "epoch": 1.73,
      "grad_norm": 2.5262239960585866,
      "learning_rate": 1.6512063817458605e-05,
      "loss": 0.8504,
      "step": 4015
    },
    {
      "epoch": 1.74,
      "grad_norm": 3.1183416250542306,
      "learning_rate": 1.650061283168177e-05,
      "loss": 0.6711,
      "step": 4020
    },
    {
      "epoch": 1.74,
      "grad_norm": 2.397589020478298,
      "learning_rate": 1.6489147064740642e-05,
      "loss": 0.7684,
      "step": 4025
    },
    {
      "epoch": 1.74,
      "grad_norm": 3.656025571054232,
      "learning_rate": 1.6477666542706207e-05,
      "loss": 0.8718,
      "step": 4030
    },
    {
      "epoch": 1.74,
      "grad_norm": 2.168095980113938,
      "learning_rate": 1.6466171291683e-05,
      "loss": 0.821,
      "step": 4035
    },
    {
      "epoch": 1.75,
      "grad_norm": 3.574224637226931,
      "learning_rate": 1.6454661337809055e-05,
      "loss": 0.7481,
      "step": 4040
    },
    {
      "epoch": 1.75,
      "grad_norm": 5.950088811281826,
      "learning_rate": 1.6443136707255818e-05,
      "loss": 0.7502,
      "step": 4045
    },
    {
      "epoch": 1.75,
      "grad_norm": 3.655860963651843,
      "learning_rate": 1.6431597426228126e-05,
      "loss": 0.8137,
      "step": 4050
    },
    {
      "epoch": 1.75,
      "grad_norm": 3.999879119916261,
      "learning_rate": 1.642004352096412e-05,
      "loss": 0.863,
      "step": 4055
    },
    {
      "epoch": 1.75,
      "grad_norm": 2.3109457640682876,
      "learning_rate": 1.6408475017735202e-05,
      "loss": 0.7669,
      "step": 4060
    },
    {
      "epoch": 1.76,
      "grad_norm": 3.0774638339199347,
      "learning_rate": 1.6396891942845954e-05,
      "loss": 0.7902,
      "step": 4065
    },
    {
      "epoch": 1.76,
      "grad_norm": 3.286986353132523,
      "learning_rate": 1.6385294322634098e-05,
      "loss": 0.9422,
      "step": 4070
    },
    {
      "epoch": 1.76,
      "grad_norm": 3.0032940152432426,
      "learning_rate": 1.6373682183470433e-05,
      "loss": 0.8383,
      "step": 4075
    },
    {
      "epoch": 1.76,
      "grad_norm": 4.1268268918520565,
      "learning_rate": 1.6362055551758767e-05,
      "loss": 0.741,
      "step": 4080
    },
    {
      "epoch": 1.76,
      "grad_norm": 3.36178597516033,
      "learning_rate": 1.635041445393586e-05,
      "loss": 0.8094,
      "step": 4085
    },
    {
      "epoch": 1.77,
      "grad_norm": 2.9084683387629076,
      "learning_rate": 1.633875891647137e-05,
      "loss": 0.8543,
      "step": 4090
    },
    {
      "epoch": 1.77,
      "grad_norm": 3.0795495163167526,
      "learning_rate": 1.6327088965867776e-05,
      "loss": 0.868,
      "step": 4095
    },
    {
      "epoch": 1.77,
      "grad_norm": 1.6954478345935486,
      "learning_rate": 1.6315404628660353e-05,
      "loss": 0.8003,
      "step": 4100
    },
    {
      "epoch": 1.77,
      "grad_norm": 2.630623426076288,
      "learning_rate": 1.630370593141706e-05,
      "loss": 0.851,
      "step": 4105
    },
    {
      "epoch": 1.78,
      "grad_norm": 8.6511753159283,
      "learning_rate": 1.6291992900738533e-05,
      "loss": 0.6028,
      "step": 4110
    },
    {
      "epoch": 1.78,
      "grad_norm": 1.8881802149607876,
      "learning_rate": 1.6280265563257985e-05,
      "loss": 0.9412,
      "step": 4115
    },
    {
      "epoch": 1.78,
      "grad_norm": 1.4693485910253388,
      "learning_rate": 1.6268523945641167e-05,
      "loss": 0.6772,
      "step": 4120
    },
    {
      "epoch": 1.78,
      "grad_norm": 2.9003838100116686,
      "learning_rate": 1.625676807458629e-05,
      "loss": 0.8457,
      "step": 4125
    },
    {
      "epoch": 1.78,
      "grad_norm": 2.589692655837404,
      "learning_rate": 1.6244997976823985e-05,
      "loss": 0.6244,
      "step": 4130
    },
    {
      "epoch": 1.79,
      "grad_norm": 2.956900769234369,
      "learning_rate": 1.6233213679117232e-05,
      "loss": 0.6138,
      "step": 4135
    },
    {
      "epoch": 1.79,
      "grad_norm": 4.024754567975696,
      "learning_rate": 1.6221415208261293e-05,
      "loss": 0.7175,
      "step": 4140
    },
    {
      "epoch": 1.79,
      "grad_norm": 2.4529650232289715,
      "learning_rate": 1.6209602591083657e-05,
      "loss": 0.6782,
      "step": 4145
    },
    {
      "epoch": 1.79,
      "grad_norm": 4.58521235942626,
      "learning_rate": 1.6197775854443985e-05,
      "loss": 0.6483,
      "step": 4150
    },
    {
      "epoch": 1.79,
      "grad_norm": 4.733052959198119,
      "learning_rate": 1.6185935025234036e-05,
      "loss": 0.7645,
      "step": 4155
    },
    {
      "epoch": 1.8,
      "grad_norm": 7.068612856829745,
      "learning_rate": 1.6174080130377622e-05,
      "loss": 0.9781,
      "step": 4160
    },
    {
      "epoch": 1.8,
      "grad_norm": 4.417592764702344,
      "learning_rate": 1.6162211196830527e-05,
      "loss": 0.9102,
      "step": 4165
    },
    {
      "epoch": 1.8,
      "grad_norm": 3.6303747370484167,
      "learning_rate": 1.6150328251580468e-05,
      "loss": 0.8287,
      "step": 4170
    },
    {
      "epoch": 1.8,
      "grad_norm": 7.138484240792349,
      "learning_rate": 1.613843132164701e-05,
      "loss": 0.872,
      "step": 4175
    },
    {
      "epoch": 1.81,
      "grad_norm": 3.918036791068465,
      "learning_rate": 1.6126520434081516e-05,
      "loss": 0.7506,
      "step": 4180
    },
    {
      "epoch": 1.81,
      "grad_norm": 5.0901783852976425,
      "learning_rate": 1.6114595615967106e-05,
      "loss": 0.9017,
      "step": 4185
    },
    {
      "epoch": 1.81,
      "grad_norm": 2.045410905898512,
      "learning_rate": 1.6102656894418553e-05,
      "loss": 0.7683,
      "step": 4190
    },
    {
      "epoch": 1.81,
      "grad_norm": 2.6632339628142594,
      "learning_rate": 1.609070429658225e-05,
      "loss": 0.7586,
      "step": 4195
    },
    {
      "epoch": 1.81,
      "grad_norm": 1.3827299405985718,
      "learning_rate": 1.607873784963615e-05,
      "loss": 0.8765,
      "step": 4200
    },
    {
      "epoch": 1.82,
      "grad_norm": 3.0920671391775727,
      "learning_rate": 1.6066757580789688e-05,
      "loss": 0.7257,
      "step": 4205
    },
    {
      "epoch": 1.82,
      "grad_norm": 2.3255694701334635,
      "learning_rate": 1.6054763517283732e-05,
      "loss": 0.6743,
      "step": 4210
    },
    {
      "epoch": 1.82,
      "grad_norm": 2.2854911864107286,
      "learning_rate": 1.604275568639051e-05,
      "loss": 0.6999,
      "step": 4215
    },
    {
      "epoch": 1.82,
      "grad_norm": 2.6973111281931277,
      "learning_rate": 1.603073411541356e-05,
      "loss": 0.6559,
      "step": 4220
    },
    {
      "epoch": 1.83,
      "grad_norm": 1.711774755331081,
      "learning_rate": 1.6018698831687666e-05,
      "loss": 0.7051,
      "step": 4225
    },
    {
      "epoch": 1.83,
      "grad_norm": 1.7553915583220734,
      "learning_rate": 1.6006649862578778e-05,
      "loss": 0.8272,
      "step": 4230
    },
    {
      "epoch": 1.83,
      "grad_norm": 2.6007853775805665,
      "learning_rate": 1.599458723548398e-05,
      "loss": 0.7484,
      "step": 4235
    },
    {
      "epoch": 1.83,
      "grad_norm": 2.5226516379121358,
      "learning_rate": 1.5982510977831408e-05,
      "loss": 0.6877,
      "step": 4240
    },
    {
      "epoch": 1.83,
      "grad_norm": 3.130361730803545,
      "learning_rate": 1.5970421117080177e-05,
      "loss": 0.6707,
      "step": 4245
    },
    {
      "epoch": 1.84,
      "grad_norm": 1.8182316559095062,
      "learning_rate": 1.5958317680720355e-05,
      "loss": 0.6941,
      "step": 4250
    },
    {
      "epoch": 1.84,
      "grad_norm": 2.8174603247256926,
      "learning_rate": 1.5946200696272863e-05,
      "loss": 0.9454,
      "step": 4255
    },
    {
      "epoch": 1.84,
      "grad_norm": 1.5718684552163051,
      "learning_rate": 1.5934070191289433e-05,
      "loss": 0.8102,
      "step": 4260
    },
    {
      "epoch": 1.84,
      "grad_norm": 7.559817790858044,
      "learning_rate": 1.5921926193352538e-05,
      "loss": 0.8674,
      "step": 4265
    },
    {
      "epoch": 1.84,
      "grad_norm": 3.9279347722065308,
      "learning_rate": 1.5909768730075333e-05,
      "loss": 0.8646,
      "step": 4270
    },
    {
      "epoch": 1.85,
      "grad_norm": 7.256637087042067,
      "learning_rate": 1.5897597829101595e-05,
      "loss": 0.7771,
      "step": 4275
    },
    {
      "epoch": 1.85,
      "grad_norm": 6.071624777714126,
      "learning_rate": 1.588541351810564e-05,
      "loss": 0.619,
      "step": 4280
    },
    {
      "epoch": 1.85,
      "grad_norm": 11.2771501825943,
      "learning_rate": 1.58732158247923e-05,
      "loss": 0.7805,
      "step": 4285
    },
    {
      "epoch": 1.85,
      "grad_norm": 13.152048939078416,
      "learning_rate": 1.586100477689682e-05,
      "loss": 0.9266,
      "step": 4290
    },
    {
      "epoch": 1.86,
      "grad_norm": 13.753922000544033,
      "learning_rate": 1.5848780402184808e-05,
      "loss": 0.7702,
      "step": 4295
    },
    {
      "epoch": 1.86,
      "grad_norm": 7.27610899912322,
      "learning_rate": 1.5836542728452185e-05,
      "loss": 0.7946,
      "step": 4300
    },
    {
      "epoch": 1.86,
      "grad_norm": 2.368126946549542,
      "learning_rate": 1.5824291783525106e-05,
      "loss": 0.6348,
      "step": 4305
    },
    {
      "epoch": 1.86,
      "grad_norm": 1.146122972648381,
      "learning_rate": 1.5812027595259906e-05,
      "loss": 0.7269,
      "step": 4310
    },
    {
      "epoch": 1.86,
      "grad_norm": 1.8289636619528542,
      "learning_rate": 1.579975019154302e-05,
      "loss": 0.7495,
      "step": 4315
    },
    {
      "epoch": 1.87,
      "grad_norm": 2.165288148545283,
      "learning_rate": 1.5787459600290954e-05,
      "loss": 0.7305,
      "step": 4320
    },
    {
      "epoch": 1.87,
      "grad_norm": 4.155506709529135,
      "learning_rate": 1.577515584945018e-05,
      "loss": 0.8006,
      "step": 4325
    },
    {
      "epoch": 1.87,
      "grad_norm": 2.6227326930585053,
      "learning_rate": 1.5762838966997107e-05,
      "loss": 0.7165,
      "step": 4330
    },
    {
      "epoch": 1.87,
      "grad_norm": 7.036639513352053,
      "learning_rate": 1.5750508980937993e-05,
      "loss": 0.777,
      "step": 4335
    },
    {
      "epoch": 1.87,
      "grad_norm": 6.766310299818976,
      "learning_rate": 1.5738165919308895e-05,
      "loss": 0.8287,
      "step": 4340
    },
    {
      "epoch": 1.88,
      "grad_norm": 1.2420623522486884,
      "learning_rate": 1.5725809810175602e-05,
      "loss": 0.7338,
      "step": 4345
    },
    {
      "epoch": 1.88,
      "grad_norm": 6.564640372437988,
      "learning_rate": 1.5713440681633566e-05,
      "loss": 0.767,
      "step": 4350
    },
    {
      "epoch": 1.88,
      "grad_norm": 3.0083078851125267,
      "learning_rate": 1.5701058561807853e-05,
      "loss": 0.7465,
      "step": 4355
    },
    {
      "epoch": 1.88,
      "grad_norm": 1.366516155826671,
      "learning_rate": 1.5688663478853058e-05,
      "loss": 0.7986,
      "step": 4360
    },
    {
      "epoch": 1.89,
      "grad_norm": 1.5525097970638504,
      "learning_rate": 1.5676255460953254e-05,
      "loss": 0.8143,
      "step": 4365
    },
    {
      "epoch": 1.89,
      "grad_norm": 2.4683161535080473,
      "learning_rate": 1.566383453632193e-05,
      "loss": 0.7335,
      "step": 4370
    },
    {
      "epoch": 1.89,
      "grad_norm": 1.852877416756737,
      "learning_rate": 1.565140073320192e-05,
      "loss": 0.7425,
      "step": 4375
    },
    {
      "epoch": 1.89,
      "grad_norm": 1.4269521394757336,
      "learning_rate": 1.5638954079865342e-05,
      "loss": 0.7322,
      "step": 4380
    },
    {
      "epoch": 1.89,
      "grad_norm": 2.338580529714504,
      "learning_rate": 1.5626494604613528e-05,
      "loss": 0.9325,
      "step": 4385
    },
    {
      "epoch": 1.9,
      "grad_norm": 2.708983569930317,
      "learning_rate": 1.561402233577697e-05,
      "loss": 0.8415,
      "step": 4390
    },
    {
      "epoch": 1.9,
      "grad_norm": 2.8397511266914304,
      "learning_rate": 1.5601537301715247e-05,
      "loss": 0.7672,
      "step": 4395
    },
    {
      "epoch": 1.9,
      "grad_norm": 1.4728688538265324,
      "learning_rate": 1.558903953081697e-05,
      "loss": 0.6873,
      "step": 4400
    },
    {
      "epoch": 1.9,
      "grad_norm": 2.4185636195682747,
      "learning_rate": 1.5576529051499707e-05,
      "loss": 0.7151,
      "step": 4405
    },
    {
      "epoch": 1.9,
      "grad_norm": 3.931984808710276,
      "learning_rate": 1.5564005892209916e-05,
      "loss": 0.7446,
      "step": 4410
    },
    {
      "epoch": 1.91,
      "grad_norm": 3.9161564917715075,
      "learning_rate": 1.55514700814229e-05,
      "loss": 0.8391,
      "step": 4415
    },
    {
      "epoch": 1.91,
      "grad_norm": 1.3389597797447057,
      "learning_rate": 1.5538921647642718e-05,
      "loss": 0.7746,
      "step": 4420
    },
    {
      "epoch": 1.91,
      "grad_norm": 2.1550095910375564,
      "learning_rate": 1.5526360619402138e-05,
      "loss": 0.6443,
      "step": 4425
    },
    {
      "epoch": 1.91,
      "grad_norm": 5.377982371644991,
      "learning_rate": 1.5513787025262556e-05,
      "loss": 0.7172,
      "step": 4430
    },
    {
      "epoch": 1.92,
      "grad_norm": 3.4585542128723183,
      "learning_rate": 1.550120089381396e-05,
      "loss": 0.8732,
      "step": 4435
    },
    {
      "epoch": 1.92,
      "grad_norm": 15.135963545385831,
      "learning_rate": 1.548860225367482e-05,
      "loss": 0.8467,
      "step": 4440
    },
    {
      "epoch": 1.92,
      "grad_norm": 4.608444491060904,
      "learning_rate": 1.5475991133492068e-05,
      "loss": 0.6725,
      "step": 4445
    },
    {
      "epoch": 1.92,
      "grad_norm": 15.522224892151614,
      "learning_rate": 1.5463367561941008e-05,
      "loss": 0.6184,
      "step": 4450
    },
    {
      "epoch": 1.92,
      "grad_norm": 13.838682827729304,
      "learning_rate": 1.5450731567725248e-05,
      "loss": 0.8473,
      "step": 4455
    },
    {
      "epoch": 1.93,
      "grad_norm": 10.190353817674348,
      "learning_rate": 1.5438083179576654e-05,
      "loss": 0.8967,
      "step": 4460
    },
    {
      "epoch": 1.93,
      "grad_norm": 4.046631043939276,
      "learning_rate": 1.542542242625527e-05,
      "loss": 0.8575,
      "step": 4465
    },
    {
      "epoch": 1.93,
      "grad_norm": 9.244099006352315,
      "learning_rate": 1.541274933654925e-05,
      "loss": 0.783,
      "step": 4470
    },
    {
      "epoch": 1.93,
      "grad_norm": 5.994363890660864,
      "learning_rate": 1.540006393927481e-05,
      "loss": 0.8754,
      "step": 4475
    },
    {
      "epoch": 1.94,
      "grad_norm": 16.806360328638917,
      "learning_rate": 1.538736626327614e-05,
      "loss": 0.7075,
      "step": 4480
    },
    {
      "epoch": 1.94,
      "grad_norm": 3.7191893623583687,
      "learning_rate": 1.5374656337425356e-05,
      "loss": 0.6038,
      "step": 4485
    },
    {
      "epoch": 1.94,
      "grad_norm": 2.6278813426794176,
      "learning_rate": 1.536193419062242e-05,
      "loss": 0.8328,
      "step": 4490
    },
    {
      "epoch": 1.94,
      "grad_norm": 3.1733988195669207,
      "learning_rate": 1.5349199851795095e-05,
      "loss": 0.8071,
      "step": 4495
    },
    {
      "epoch": 1.94,
      "grad_norm": 4.189600230957689,
      "learning_rate": 1.533645334989886e-05,
      "loss": 0.8508,
      "step": 4500
    },
    {
      "epoch": 1.95,
      "grad_norm": 10.22818859017594,
      "learning_rate": 1.5323694713916845e-05,
      "loss": 0.7205,
      "step": 4505
    },
    {
      "epoch": 1.95,
      "grad_norm": 1.9802329865191948,
      "learning_rate": 1.531092397285978e-05,
      "loss": 0.7696,
      "step": 4510
    },
    {
      "epoch": 1.95,
      "grad_norm": 6.172727439763833,
      "learning_rate": 1.5298141155765907e-05,
      "loss": 0.7767,
      "step": 4515
    },
    {
      "epoch": 1.95,
      "grad_norm": 3.088104692698883,
      "learning_rate": 1.5285346291700945e-05,
      "loss": 0.7553,
      "step": 4520
    },
    {
      "epoch": 1.95,
      "grad_norm": 1.83864815532713,
      "learning_rate": 1.5272539409757992e-05,
      "loss": 0.6233,
      "step": 4525
    },
    {
      "epoch": 1.96,
      "grad_norm": 3.214823490347084,
      "learning_rate": 1.5259720539057474e-05,
      "loss": 0.6767,
      "step": 4530
    },
    {
      "epoch": 1.96,
      "grad_norm": 3.844831063170245,
      "learning_rate": 1.524688970874708e-05,
      "loss": 0.7351,
      "step": 4535
    },
    {
      "epoch": 1.96,
      "grad_norm": 11.119244612077713,
      "learning_rate": 1.5234046948001688e-05,
      "loss": 0.7056,
      "step": 4540
    },
    {
      "epoch": 1.96,
      "grad_norm": 2.0176072517741206,
      "learning_rate": 1.5221192286023311e-05,
      "loss": 0.6993,
      "step": 4545
    },
    {
      "epoch": 1.97,
      "grad_norm": 2.6533601248298404,
      "learning_rate": 1.5208325752041023e-05,
      "loss": 0.7078,
      "step": 4550
    },
    {
      "epoch": 1.97,
      "grad_norm": 4.065798443051646,
      "learning_rate": 1.5195447375310884e-05,
      "loss": 0.6986,
      "step": 4555
    },
    {
      "epoch": 1.97,
      "grad_norm": 3.744211055171865,
      "learning_rate": 1.518255718511589e-05,
      "loss": 0.6592,
      "step": 4560
    },
    {
      "epoch": 1.97,
      "grad_norm": 3.9609503029691746,
      "learning_rate": 1.516965521076589e-05,
      "loss": 0.8221,
      "step": 4565
    },
    {
      "epoch": 1.97,
      "grad_norm": 1.6094015126459182,
      "learning_rate": 1.515674148159754e-05,
      "loss": 0.6243,
      "step": 4570
    },
    {
      "epoch": 1.98,
      "grad_norm": 3.4156081882227625,
      "learning_rate": 1.5143816026974221e-05,
      "loss": 0.7139,
      "step": 4575
    },
    {
      "epoch": 1.98,
      "grad_norm": 1.9510998473531977,
      "learning_rate": 1.5130878876285965e-05,
      "loss": 0.7618,
      "step": 4580
    },
    {
      "epoch": 1.98,
      "grad_norm": 1.668314185972532,
      "learning_rate": 1.511793005894941e-05,
      "loss": 0.7907,
      "step": 4585
    },
    {
      "epoch": 1.98,
      "grad_norm": 1.6058117505109508,
      "learning_rate": 1.5104969604407716e-05,
      "loss": 0.659,
      "step": 4590
    },
    {
      "epoch": 1.98,
      "grad_norm": 3.5132785483692635,
      "learning_rate": 1.5091997542130506e-05,
      "loss": 0.786,
      "step": 4595
    },
    {
      "epoch": 1.99,
      "grad_norm": 6.732887645085844,
      "learning_rate": 1.5079013901613802e-05,
      "loss": 0.7016,
      "step": 4600
    },
    {
      "epoch": 1.99,
      "grad_norm": 1.423445894358691,
      "learning_rate": 1.5066018712379939e-05,
      "loss": 0.7545,
      "step": 4605
    },
    {
      "epoch": 1.99,
      "grad_norm": 1.5043065712268024,
      "learning_rate": 1.5053012003977527e-05,
      "loss": 0.8126,
      "step": 4610
    },
    {
      "epoch": 1.99,
      "grad_norm": 3.0586231775629615,
      "learning_rate": 1.5039993805981357e-05,
      "loss": 0.7461,
      "step": 4615
    },
    {
      "epoch": 2.0,
      "grad_norm": 6.052595972323311,
      "learning_rate": 1.5026964147992348e-05,
      "loss": 0.7427,
      "step": 4620
    },
    {
      "epoch": 2.0,
      "grad_norm": 2.0733693886967393,
      "learning_rate": 1.5013923059637486e-05,
      "loss": 0.6624,
      "step": 4625
    },
    {
      "epoch": 2.0,
      "grad_norm": 4.687342403320873,
      "learning_rate": 1.500087057056973e-05,
      "loss": 0.7793,
      "step": 4630
    },
    {
      "epoch": 2.0,
      "eval_loss": 5.413473606109619,
      "eval_runtime": 351.2973,
      "eval_samples_per_second": 21.369,
      "eval_steps_per_second": 0.336,
      "step": 4630
    },
    {
      "epoch": 2.0,
      "grad_norm": 3.6681478113794004,
      "learning_rate": 1.498780671046798e-05,
      "loss": 0.734,
      "step": 4635
    },
    {
      "epoch": 2.0,
      "grad_norm": 7.802830523737377,
      "learning_rate": 1.497473150903698e-05,
      "loss": 0.7815,
      "step": 4640
    },
    {
      "epoch": 2.01,
      "grad_norm": 8.109033409317203,
      "learning_rate": 1.496164499600726e-05,
      "loss": 0.829,
      "step": 4645
    },
    {
      "epoch": 2.01,
      "grad_norm": 8.074131689515943,
      "learning_rate": 1.4948547201135088e-05,
      "loss": 0.6535,
      "step": 4650
    },
    {
      "epoch": 2.01,
      "grad_norm": 6.596520551200253,
      "learning_rate": 1.4935438154202363e-05,
      "loss": 0.7058,
      "step": 4655
    },
    {
      "epoch": 2.01,
      "grad_norm": 1.6424859881344187,
      "learning_rate": 1.4922317885016584e-05,
      "loss": 0.643,
      "step": 4660
    },
    {
      "epoch": 2.02,
      "grad_norm": 4.977709318530804,
      "learning_rate": 1.490918642341076e-05,
      "loss": 0.731,
      "step": 4665
    },
    {
      "epoch": 2.02,
      "grad_norm": 2.795003746798207,
      "learning_rate": 1.4896043799243349e-05,
      "loss": 0.7434,
      "step": 4670
    },
    {
      "epoch": 2.02,
      "grad_norm": 2.106674558037682,
      "learning_rate": 1.4882890042398197e-05,
      "loss": 0.6729,
      "step": 4675
    },
    {
      "epoch": 2.02,
      "grad_norm": 3.4589623366999875,
      "learning_rate": 1.4869725182784458e-05,
      "loss": 0.8524,
      "step": 4680
    },
    {
      "epoch": 2.02,
      "grad_norm": 2.702534222166854,
      "learning_rate": 1.4856549250336537e-05,
      "loss": 0.6632,
      "step": 4685
    },
    {
      "epoch": 2.03,
      "grad_norm": 2.8843345731755354,
      "learning_rate": 1.4843362275014004e-05,
      "loss": 0.8314,
      "step": 4690
    },
    {
      "epoch": 2.03,
      "grad_norm": 2.2974327677286053,
      "learning_rate": 1.4830164286801552e-05,
      "loss": 0.6472,
      "step": 4695
    },
    {
      "epoch": 2.03,
      "grad_norm": 2.6153862975813094,
      "learning_rate": 1.4816955315708916e-05,
      "loss": 0.6175,
      "step": 4700
    },
    {
      "epoch": 2.03,
      "grad_norm": 1.812689881534979,
      "learning_rate": 1.4803735391770788e-05,
      "loss": 0.6033,
      "step": 4705
    },
    {
      "epoch": 2.03,
      "grad_norm": 8.657508696336997,
      "learning_rate": 1.4790504545046785e-05,
      "loss": 0.8608,
      "step": 4710
    },
    {
      "epoch": 2.04,
      "grad_norm": 10.64182267161339,
      "learning_rate": 1.4777262805621341e-05,
      "loss": 0.686,
      "step": 4715
    },
    {
      "epoch": 2.04,
      "grad_norm": 10.917597635768443,
      "learning_rate": 1.4764010203603673e-05,
      "loss": 0.6776,
      "step": 4720
    },
    {
      "epoch": 2.04,
      "grad_norm": 4.894550754609609,
      "learning_rate": 1.4750746769127695e-05,
      "loss": 0.6753,
      "step": 4725
    },
    {
      "epoch": 2.04,
      "grad_norm": 7.90131501061075,
      "learning_rate": 1.4737472532351941e-05,
      "loss": 0.6693,
      "step": 4730
    },
    {
      "epoch": 2.05,
      "grad_norm": 7.234315155347728,
      "learning_rate": 1.4724187523459524e-05,
      "loss": 0.8296,
      "step": 4735
    },
    {
      "epoch": 2.05,
      "grad_norm": 2.621839934063872,
      "learning_rate": 1.4710891772658034e-05,
      "loss": 0.8339,
      "step": 4740
    },
    {
      "epoch": 2.05,
      "grad_norm": 4.132144800302082,
      "learning_rate": 1.4697585310179495e-05,
      "loss": 0.7128,
      "step": 4745
    },
    {
      "epoch": 2.05,
      "grad_norm": 1.392978638244078,
      "learning_rate": 1.4684268166280293e-05,
      "loss": 0.5983,
      "step": 4750
    },
    {
      "epoch": 2.05,
      "grad_norm": 2.346021627010998,
      "learning_rate": 1.4670940371241086e-05,
      "loss": 0.6504,
      "step": 4755
    },
    {
      "epoch": 2.06,
      "grad_norm": 1.6569506181595035,
      "learning_rate": 1.4657601955366767e-05,
      "loss": 0.693,
      "step": 4760
    },
    {
      "epoch": 2.06,
      "grad_norm": 2.2965091368781776,
      "learning_rate": 1.4644252948986368e-05,
      "loss": 0.6942,
      "step": 4765
    },
    {
      "epoch": 2.06,
      "grad_norm": 1.6884518445524517,
      "learning_rate": 1.4630893382453004e-05,
      "loss": 0.7618,
      "step": 4770
    },
    {
      "epoch": 2.06,
      "grad_norm": 3.0291222685763852,
      "learning_rate": 1.4617523286143806e-05,
      "loss": 0.7276,
      "step": 4775
    },
    {
      "epoch": 2.06,
      "grad_norm": 2.752424933714392,
      "learning_rate": 1.4604142690459843e-05,
      "loss": 0.6047,
      "step": 4780
    },
    {
      "epoch": 2.07,
      "grad_norm": 2.913663764157102,
      "learning_rate": 1.459075162582606e-05,
      "loss": 0.7256,
      "step": 4785
    },
    {
      "epoch": 2.07,
      "grad_norm": 2.2389568862285567,
      "learning_rate": 1.4577350122691205e-05,
      "loss": 0.7098,
      "step": 4790
    },
    {
      "epoch": 2.07,
      "grad_norm": 3.8588853245465953,
      "learning_rate": 1.4563938211527761e-05,
      "loss": 0.8396,
      "step": 4795
    },
    {
      "epoch": 2.07,
      "grad_norm": 1.8326704667975264,
      "learning_rate": 1.4550515922831881e-05,
      "loss": 0.6518,
      "step": 4800
    },
    {
      "epoch": 2.08,
      "grad_norm": 2.01601811375239,
      "learning_rate": 1.4537083287123309e-05,
      "loss": 0.7681,
      "step": 4805
    },
    {
      "epoch": 2.08,
      "grad_norm": 3.4581655103666966,
      "learning_rate": 1.4523640334945317e-05,
      "loss": 0.6857,
      "step": 4810
    },
    {
      "epoch": 2.08,
      "grad_norm": 2.5903440925338317,
      "learning_rate": 1.4510187096864638e-05,
      "loss": 0.6613,
      "step": 4815
    },
    {
      "epoch": 2.08,
      "grad_norm": 1.8100171699807792,
      "learning_rate": 1.449672360347139e-05,
      "loss": 0.6798,
      "step": 4820
    },
    {
      "epoch": 2.08,
      "grad_norm": 2.1781013189873946,
      "learning_rate": 1.4483249885379014e-05,
      "loss": 0.6552,
      "step": 4825
    },
    {
      "epoch": 2.09,
      "grad_norm": 6.251914333658113,
      "learning_rate": 1.4469765973224194e-05,
      "loss": 0.7461,
      "step": 4830
    },
    {
      "epoch": 2.09,
      "grad_norm": 5.817878468854903,
      "learning_rate": 1.4456271897666798e-05,
      "loss": 0.6437,
      "step": 4835
    },
    {
      "epoch": 2.09,
      "grad_norm": 5.662860653064379,
      "learning_rate": 1.44427676893898e-05,
      "loss": 0.7011,
      "step": 4840
    },
    {
      "epoch": 2.09,
      "grad_norm": 4.510980150786451,
      "learning_rate": 1.442925337909922e-05,
      "loss": 0.5923,
      "step": 4845
    },
    {
      "epoch": 2.1,
      "grad_norm": 1.9814598174187572,
      "learning_rate": 1.441572899752404e-05,
      "loss": 0.7218,
      "step": 4850
    },
    {
      "epoch": 2.1,
      "grad_norm": 1.8871473249191737,
      "learning_rate": 1.4402194575416148e-05,
      "loss": 0.7841,
      "step": 4855
    },
    {
      "epoch": 2.1,
      "grad_norm": 2.729798404906885,
      "learning_rate": 1.438865014355026e-05,
      "loss": 0.5294,
      "step": 4860
    },
    {
      "epoch": 2.1,
      "grad_norm": 3.1101404087816444,
      "learning_rate": 1.4375095732723852e-05,
      "loss": 0.6459,
      "step": 4865
    },
    {
      "epoch": 2.1,
      "grad_norm": 2.5343679096514107,
      "learning_rate": 1.4361531373757092e-05,
      "loss": 0.5827,
      "step": 4870
    },
    {
      "epoch": 2.11,
      "grad_norm": 2.4084831856604536,
      "learning_rate": 1.434795709749277e-05,
      "loss": 0.6763,
      "step": 4875
    },
    {
      "epoch": 2.11,
      "grad_norm": 2.6059335460748896,
      "learning_rate": 1.4334372934796218e-05,
      "loss": 0.5522,
      "step": 4880
    },
    {
      "epoch": 2.11,
      "grad_norm": 4.261437272764679,
      "learning_rate": 1.4320778916555255e-05,
      "loss": 0.6038,
      "step": 4885
    },
    {
      "epoch": 2.11,
      "grad_norm": 1.2672251951057814,
      "learning_rate": 1.430717507368011e-05,
      "loss": 0.688,
      "step": 4890
    },
    {
      "epoch": 2.11,
      "grad_norm": 2.1511319225380094,
      "learning_rate": 1.4293561437103348e-05,
      "loss": 0.7069,
      "step": 4895
    },
    {
      "epoch": 2.12,
      "grad_norm": 4.212127220928629,
      "learning_rate": 1.4279938037779801e-05,
      "loss": 0.6505,
      "step": 4900
    },
    {
      "epoch": 2.12,
      "grad_norm": 2.1893645392849868,
      "learning_rate": 1.4266304906686508e-05,
      "loss": 0.5228,
      "step": 4905
    },
    {
      "epoch": 2.12,
      "grad_norm": 3.537460380045355,
      "learning_rate": 1.425266207482263e-05,
      "loss": 0.718,
      "step": 4910
    },
    {
      "epoch": 2.12,
      "grad_norm": 3.404220405341098,
      "learning_rate": 1.4239009573209381e-05,
      "loss": 0.7759,
      "step": 4915
    },
    {
      "epoch": 2.13,
      "grad_norm": 3.159580780643254,
      "learning_rate": 1.4225347432889978e-05,
      "loss": 0.6648,
      "step": 4920
    },
    {
      "epoch": 2.13,
      "grad_norm": 3.037253480076717,
      "learning_rate": 1.4211675684929533e-05,
      "loss": 0.6586,
      "step": 4925
    },
    {
      "epoch": 2.13,
      "grad_norm": 3.869361730397183,
      "learning_rate": 1.4197994360415023e-05,
      "loss": 0.6963,
      "step": 4930
    },
    {
      "epoch": 2.13,
      "grad_norm": 2.2272003374579468,
      "learning_rate": 1.4184303490455194e-05,
      "loss": 0.8277,
      "step": 4935
    },
    {
      "epoch": 2.13,
      "grad_norm": 3.034166829374501,
      "learning_rate": 1.4170603106180489e-05,
      "loss": 0.7326,
      "step": 4940
    },
    {
      "epoch": 2.14,
      "grad_norm": 3.0837813829439136,
      "learning_rate": 1.4156893238743e-05,
      "loss": 0.6207,
      "step": 4945
    },
    {
      "epoch": 2.14,
      "grad_norm": 4.059850226368533,
      "learning_rate": 1.414317391931636e-05,
      "loss": 0.7122,
      "step": 4950
    },
    {
      "epoch": 2.14,
      "grad_norm": 3.4181474527577977,
      "learning_rate": 1.4129445179095718e-05,
      "loss": 0.6949,
      "step": 4955
    },
    {
      "epoch": 2.14,
      "grad_norm": 2.331066437043077,
      "learning_rate": 1.4115707049297631e-05,
      "loss": 0.5983,
      "step": 4960
    },
    {
      "epoch": 2.14,
      "grad_norm": 3.0238435031005064,
      "learning_rate": 1.4101959561160003e-05,
      "loss": 0.6895,
      "step": 4965
    },
    {
      "epoch": 2.15,
      "grad_norm": 2.761852205828301,
      "learning_rate": 1.408820274594203e-05,
      "loss": 0.7393,
      "step": 4970
    },
    {
      "epoch": 2.15,
      "grad_norm": 2.544915824724106,
      "learning_rate": 1.4074436634924102e-05,
      "loss": 0.6316,
      "step": 4975
    },
    {
      "epoch": 2.15,
      "grad_norm": 1.3244544600527366,
      "learning_rate": 1.4060661259407758e-05,
      "loss": 0.5658,
      "step": 4980
    },
    {
      "epoch": 2.15,
      "grad_norm": 2.772971136679998,
      "learning_rate": 1.4046876650715591e-05,
      "loss": 0.6778,
      "step": 4985
    },
    {
      "epoch": 2.16,
      "grad_norm": 2.5012677940913757,
      "learning_rate": 1.4033082840191196e-05,
      "loss": 0.7587,
      "step": 4990
    },
    {
      "epoch": 2.16,
      "grad_norm": 3.64540541690908,
      "learning_rate": 1.4019279859199096e-05,
      "loss": 0.7228,
      "step": 4995
    },
    {
      "epoch": 2.16,
      "grad_norm": 8.366734997146574,
      "learning_rate": 1.400546773912465e-05,
      "loss": 0.7256,
      "step": 5000
    },
    {
      "epoch": 2.16,
      "grad_norm": 2.142213618690858,
      "learning_rate": 1.3991646511374011e-05,
      "loss": 0.7008,
      "step": 5005
    },
    {
      "epoch": 2.16,
      "grad_norm": 4.019248315380808,
      "learning_rate": 1.3977816207374039e-05,
      "loss": 0.7018,
      "step": 5010
    },
    {
      "epoch": 2.17,
      "grad_norm": 5.554098532914148,
      "learning_rate": 1.396397685857223e-05,
      "loss": 0.7446,
      "step": 5015
    },
    {
      "epoch": 2.17,
      "grad_norm": 3.359959733126847,
      "learning_rate": 1.3950128496436646e-05,
      "loss": 0.601,
      "step": 5020
    },
    {
      "epoch": 2.17,
      "grad_norm": 2.569653753156473,
      "learning_rate": 1.393627115245584e-05,
      "loss": 0.6059,
      "step": 5025
    },
    {
      "epoch": 2.17,
      "grad_norm": 2.9822193273204873,
      "learning_rate": 1.3922404858138794e-05,
      "loss": 0.6179,
      "step": 5030
    },
    {
      "epoch": 2.17,
      "grad_norm": 2.0489495833913782,
      "learning_rate": 1.390852964501484e-05,
      "loss": 0.6472,
      "step": 5035
    },
    {
      "epoch": 2.18,
      "grad_norm": 1.4261825333330067,
      "learning_rate": 1.389464554463359e-05,
      "loss": 0.5776,
      "step": 5040
    },
    {
      "epoch": 2.18,
      "grad_norm": 2.400830227229986,
      "learning_rate": 1.3880752588564861e-05,
      "loss": 0.7632,
      "step": 5045
    },
    {
      "epoch": 2.18,
      "grad_norm": 2.9766721821975106,
      "learning_rate": 1.3866850808398607e-05,
      "loss": 0.6854,
      "step": 5050
    },
    {
      "epoch": 2.18,
      "grad_norm": 5.204975829961181,
      "learning_rate": 1.3852940235744848e-05,
      "loss": 0.6408,
      "step": 5055
    },
    {
      "epoch": 2.19,
      "grad_norm": 12.80490051664339,
      "learning_rate": 1.3839020902233595e-05,
      "loss": 0.6775,
      "step": 5060
    },
    {
      "epoch": 2.19,
      "grad_norm": 2.8672289829708397,
      "learning_rate": 1.3825092839514782e-05,
      "loss": 0.6438,
      "step": 5065
    },
    {
      "epoch": 2.19,
      "grad_norm": 4.051389466862681,
      "learning_rate": 1.381115607925819e-05,
      "loss": 0.6045,
      "step": 5070
    },
    {
      "epoch": 2.19,
      "grad_norm": 6.952904080521865,
      "learning_rate": 1.3797210653153372e-05,
      "loss": 0.7735,
      "step": 5075
    },
    {
      "epoch": 2.19,
      "grad_norm": 2.94230699295151,
      "learning_rate": 1.3783256592909596e-05,
      "loss": 0.6059,
      "step": 5080
    },
    {
      "epoch": 2.2,
      "grad_norm": 2.9411457261542724,
      "learning_rate": 1.3769293930255752e-05,
      "loss": 0.6946,
      "step": 5085
    },
    {
      "epoch": 2.2,
      "grad_norm": 3.4557539077990724,
      "learning_rate": 1.3755322696940293e-05,
      "loss": 0.5876,
      "step": 5090
    },
    {
      "epoch": 2.2,
      "grad_norm": 1.856220373216668,
      "learning_rate": 1.3741342924731171e-05,
      "loss": 0.8175,
      "step": 5095
    },
    {
      "epoch": 2.2,
      "grad_norm": 5.806548304083525,
      "learning_rate": 1.3727354645415734e-05,
      "loss": 0.6969,
      "step": 5100
    },
    {
      "epoch": 2.21,
      "grad_norm": 1.9514309269566108,
      "learning_rate": 1.3713357890800685e-05,
      "loss": 0.585,
      "step": 5105
    },
    {
      "epoch": 2.21,
      "grad_norm": 1.495350677319604,
      "learning_rate": 1.3699352692712004e-05,
      "loss": 0.5571,
      "step": 5110
    },
    {
      "epoch": 2.21,
      "grad_norm": 2.6614436176053577,
      "learning_rate": 1.3685339082994856e-05,
      "loss": 0.6419,
      "step": 5115
    },
    {
      "epoch": 2.21,
      "grad_norm": 2.707576709000901,
      "learning_rate": 1.3671317093513546e-05,
      "loss": 0.7347,
      "step": 5120
    },
    {
      "epoch": 2.21,
      "grad_norm": 2.2162068090827214,
      "learning_rate": 1.3657286756151418e-05,
      "loss": 0.6805,
      "step": 5125
    },
    {
      "epoch": 2.22,
      "grad_norm": 2.995075044694853,
      "learning_rate": 1.3643248102810812e-05,
      "loss": 0.6363,
      "step": 5130
    },
    {
      "epoch": 2.22,
      "grad_norm": 2.685871623205705,
      "learning_rate": 1.362920116541297e-05,
      "loss": 0.6295,
      "step": 5135
    },
    {
      "epoch": 2.22,
      "grad_norm": 3.174472117863071,
      "learning_rate": 1.3615145975897973e-05,
      "loss": 0.7393,
      "step": 5140
    },
    {
      "epoch": 2.22,
      "grad_norm": 2.0902726071931075,
      "learning_rate": 1.360108256622466e-05,
      "loss": 0.6473,
      "step": 5145
    },
    {
      "epoch": 2.22,
      "grad_norm": 1.9372938517889469,
      "learning_rate": 1.3587010968370569e-05,
      "loss": 0.6455,
      "step": 5150
    },
    {
      "epoch": 2.23,
      "grad_norm": 2.3024147029017534,
      "learning_rate": 1.3572931214331848e-05,
      "loss": 0.6314,
      "step": 5155
    },
    {
      "epoch": 2.23,
      "grad_norm": 4.320602900604779,
      "learning_rate": 1.3558843336123201e-05,
      "loss": 0.7002,
      "step": 5160
    },
    {
      "epoch": 2.23,
      "grad_norm": 1.6210451039283706,
      "learning_rate": 1.3544747365777795e-05,
      "loss": 0.7742,
      "step": 5165
    },
    {
      "epoch": 2.23,
      "grad_norm": 1.3075788822230048,
      "learning_rate": 1.3530643335347203e-05,
      "loss": 0.5962,
      "step": 5170
    },
    {
      "epoch": 2.24,
      "grad_norm": 1.7855687906302533,
      "learning_rate": 1.351653127690132e-05,
      "loss": 0.5874,
      "step": 5175
    },
    {
      "epoch": 2.24,
      "grad_norm": 2.3951063501074583,
      "learning_rate": 1.35024112225283e-05,
      "loss": 0.5974,
      "step": 5180
    },
    {
      "epoch": 2.24,
      "grad_norm": 2.1622174835372188,
      "learning_rate": 1.3488283204334479e-05,
      "loss": 0.6406,
      "step": 5185
    },
    {
      "epoch": 2.24,
      "grad_norm": 2.132938791139148,
      "learning_rate": 1.3474147254444293e-05,
      "loss": 0.6773,
      "step": 5190
    },
    {
      "epoch": 2.24,
      "grad_norm": 4.669387578737708,
      "learning_rate": 1.3460003405000226e-05,
      "loss": 0.58,
      "step": 5195
    },
    {
      "epoch": 2.25,
      "grad_norm": 4.131229331144581,
      "learning_rate": 1.3445851688162713e-05,
      "loss": 0.5259,
      "step": 5200
    },
    {
      "epoch": 2.25,
      "grad_norm": 2.101898168571244,
      "learning_rate": 1.3431692136110079e-05,
      "loss": 0.5822,
      "step": 5205
    },
    {
      "epoch": 2.25,
      "grad_norm": 2.9631752802509537,
      "learning_rate": 1.3417524781038473e-05,
      "loss": 0.7066,
      "step": 5210
    },
    {
      "epoch": 2.25,
      "grad_norm": 2.8623277308886537,
      "learning_rate": 1.3403349655161782e-05,
      "loss": 0.6715,
      "step": 5215
    },
    {
      "epoch": 2.25,
      "grad_norm": 3.3350327819908356,
      "learning_rate": 1.3389166790711556e-05,
      "loss": 0.7183,
      "step": 5220
    },
    {
      "epoch": 2.26,
      "grad_norm": 3.295562949758625,
      "learning_rate": 1.3374976219936952e-05,
      "loss": 0.7453,
      "step": 5225
    },
    {
      "epoch": 2.26,
      "grad_norm": 1.5495019986962577,
      "learning_rate": 1.336077797510464e-05,
      "loss": 0.6477,
      "step": 5230
    },
    {
      "epoch": 2.26,
      "grad_norm": 3.244632762508368,
      "learning_rate": 1.3346572088498746e-05,
      "loss": 0.7053,
      "step": 5235
    },
    {
      "epoch": 2.26,
      "grad_norm": 10.458889203896769,
      "learning_rate": 1.333235859242077e-05,
      "loss": 0.6152,
      "step": 5240
    },
    {
      "epoch": 2.27,
      "grad_norm": 2.817571457186845,
      "learning_rate": 1.3318137519189518e-05,
      "loss": 0.5173,
      "step": 5245
    },
    {
      "epoch": 2.27,
      "grad_norm": 2.2804314138434525,
      "learning_rate": 1.3303908901141016e-05,
      "loss": 0.6432,
      "step": 5250
    },
    {
      "epoch": 2.27,
      "grad_norm": 2.6539794325977764,
      "learning_rate": 1.3289672770628449e-05,
      "loss": 0.727,
      "step": 5255
    },
    {
      "epoch": 2.27,
      "grad_norm": 3.0261414160956615,
      "learning_rate": 1.327542916002209e-05,
      "loss": 0.64,
      "step": 5260
    },
    {
      "epoch": 2.27,
      "grad_norm": 2.6157623856222334,
      "learning_rate": 1.3261178101709215e-05,
      "loss": 0.4689,
      "step": 5265
    },
    {
      "epoch": 2.28,
      "grad_norm": 2.2558817062570116,
      "learning_rate": 1.3246919628094037e-05,
      "loss": 0.7602,
      "step": 5270
    },
    {
      "epoch": 2.28,
      "grad_norm": 3.70735926592346,
      "learning_rate": 1.3232653771597627e-05,
      "loss": 0.6818,
      "step": 5275
    },
    {
      "epoch": 2.28,
      "grad_norm": 3.63267852687161,
      "learning_rate": 1.3218380564657842e-05,
      "loss": 0.7698,
      "step": 5280
    },
    {
      "epoch": 2.28,
      "grad_norm": 3.5578281169494503,
      "learning_rate": 1.3204100039729263e-05,
      "loss": 0.6531,
      "step": 5285
    },
    {
      "epoch": 2.29,
      "grad_norm": 1.5663803805551082,
      "learning_rate": 1.3189812229283096e-05,
      "loss": 0.6545,
      "step": 5290
    },
    {
      "epoch": 2.29,
      "grad_norm": 3.6169189509060873,
      "learning_rate": 1.3175517165807126e-05,
      "loss": 0.671,
      "step": 5295
    },
    {
      "epoch": 2.29,
      "grad_norm": 3.164169667424972,
      "learning_rate": 1.3161214881805622e-05,
      "loss": 0.6619,
      "step": 5300
    },
    {
      "epoch": 2.29,
      "grad_norm": 2.7729255955211984,
      "learning_rate": 1.3146905409799266e-05,
      "loss": 0.6725,
      "step": 5305
    },
    {
      "epoch": 2.29,
      "grad_norm": 2.0512913530098693,
      "learning_rate": 1.3132588782325103e-05,
      "loss": 0.658,
      "step": 5310
    },
    {
      "epoch": 2.3,
      "grad_norm": 1.8793852520626917,
      "learning_rate": 1.3118265031936427e-05,
      "loss": 0.6442,
      "step": 5315
    },
    {
      "epoch": 2.3,
      "grad_norm": 4.26316521105751,
      "learning_rate": 1.3103934191202746e-05,
      "loss": 0.651,
      "step": 5320
    },
    {
      "epoch": 2.3,
      "grad_norm": 2.02079446252766,
      "learning_rate": 1.3089596292709677e-05,
      "loss": 0.6715,
      "step": 5325
    },
    {
      "epoch": 2.3,
      "grad_norm": 1.3143490394654598,
      "learning_rate": 1.307525136905889e-05,
      "loss": 0.5558,
      "step": 5330
    },
    {
      "epoch": 2.3,
      "grad_norm": 3.9706238434807273,
      "learning_rate": 1.3060899452868028e-05,
      "loss": 0.6195,
      "step": 5335
    },
    {
      "epoch": 2.31,
      "grad_norm": 4.653747012803783,
      "learning_rate": 1.3046540576770635e-05,
      "loss": 0.7183,
      "step": 5340
    },
    {
      "epoch": 2.31,
      "grad_norm": 6.238182045888077,
      "learning_rate": 1.3032174773416085e-05,
      "loss": 0.6659,
      "step": 5345
    },
    {
      "epoch": 2.31,
      "grad_norm": 4.559310932480753,
      "learning_rate": 1.3017802075469492e-05,
      "loss": 0.6809,
      "step": 5350
    },
    {
      "epoch": 2.31,
      "grad_norm": 1.467398670493438,
      "learning_rate": 1.3003422515611654e-05,
      "loss": 0.6352,
      "step": 5355
    },
    {
      "epoch": 2.32,
      "grad_norm": 4.642235893269442,
      "learning_rate": 1.2989036126538973e-05,
      "loss": 0.5664,
      "step": 5360
    },
    {
      "epoch": 2.32,
      "grad_norm": 3.938217402793309,
      "learning_rate": 1.2974642940963373e-05,
      "loss": 0.6056,
      "step": 5365
    },
    {
      "epoch": 2.32,
      "grad_norm": 6.885204245370707,
      "learning_rate": 1.2960242991612244e-05,
      "loss": 0.6308,
      "step": 5370
    },
    {
      "epoch": 2.32,
      "grad_norm": 4.5212483786095765,
      "learning_rate": 1.2945836311228342e-05,
      "loss": 0.5929,
      "step": 5375
    },
    {
      "epoch": 2.32,
      "grad_norm": 4.641571097379382,
      "learning_rate": 1.2931422932569732e-05,
      "loss": 0.5578,
      "step": 5380
    },
    {
      "epoch": 2.33,
      "grad_norm": 2.8833078407731842,
      "learning_rate": 1.2917002888409718e-05,
      "loss": 0.5554,
      "step": 5385
    },
    {
      "epoch": 2.33,
      "grad_norm": 1.173901145308014,
      "learning_rate": 1.2902576211536748e-05,
      "loss": 0.5722,
      "step": 5390
    },
    {
      "epoch": 2.33,
      "grad_norm": 0.9358893996361893,
      "learning_rate": 1.288814293475436e-05,
      "loss": 0.7776,
      "step": 5395
    },
    {
      "epoch": 2.33,
      "grad_norm": 1.4558280571808577,
      "learning_rate": 1.2873703090881096e-05,
      "loss": 0.4994,
      "step": 5400
    },
    {
      "epoch": 2.33,
      "grad_norm": 4.079557487260838,
      "learning_rate": 1.2859256712750428e-05,
      "loss": 0.6432,
      "step": 5405
    },
    {
      "epoch": 2.34,
      "grad_norm": 2.551436632509885,
      "learning_rate": 1.284480383321069e-05,
      "loss": 0.539,
      "step": 5410
    },
    {
      "epoch": 2.34,
      "grad_norm": 4.073997211595631,
      "learning_rate": 1.2830344485124995e-05,
      "loss": 0.6014,
      "step": 5415
    },
    {
      "epoch": 2.34,
      "grad_norm": 1.547482223872806,
      "learning_rate": 1.2815878701371172e-05,
      "loss": 0.6883,
      "step": 5420
    },
    {
      "epoch": 2.34,
      "grad_norm": 7.302871222039995,
      "learning_rate": 1.280140651484167e-05,
      "loss": 0.6608,
      "step": 5425
    },
    {
      "epoch": 2.35,
      "grad_norm": 4.71763130175225,
      "learning_rate": 1.278692795844351e-05,
      "loss": 0.6053,
      "step": 5430
    },
    {
      "epoch": 2.35,
      "grad_norm": 2.376185014669388,
      "learning_rate": 1.2772443065098186e-05,
      "loss": 0.5509,
      "step": 5435
    },
    {
      "epoch": 2.35,
      "grad_norm": 2.8825193085694707,
      "learning_rate": 1.275795186774161e-05,
      "loss": 0.8276,
      "step": 5440
    },
    {
      "epoch": 2.35,
      "grad_norm": 3.2274170276467036,
      "learning_rate": 1.2743454399324027e-05,
      "loss": 0.6181,
      "step": 5445
    },
    {
      "epoch": 2.35,
      "grad_norm": 2.543424038374083,
      "learning_rate": 1.2728950692809936e-05,
      "loss": 0.5948,
      "step": 5450
    },
    {
      "epoch": 2.36,
      "grad_norm": 3.9232792043148637,
      "learning_rate": 1.2714440781178019e-05,
      "loss": 0.6813,
      "step": 5455
    },
    {
      "epoch": 2.36,
      "grad_norm": 4.311356998187637,
      "learning_rate": 1.2699924697421075e-05,
      "loss": 0.599,
      "step": 5460
    },
    {
      "epoch": 2.36,
      "grad_norm": 2.394891057498207,
      "learning_rate": 1.2685402474545934e-05,
      "loss": 0.5836,
      "step": 5465
    },
    {
      "epoch": 2.36,
      "grad_norm": 4.232798939575906,
      "learning_rate": 1.2670874145573384e-05,
      "loss": 0.7191,
      "step": 5470
    },
    {
      "epoch": 2.37,
      "grad_norm": 2.546588455428622,
      "learning_rate": 1.2656339743538099e-05,
      "loss": 0.6068,
      "step": 5475
    },
    {
      "epoch": 2.37,
      "grad_norm": 2.397110488213129,
      "learning_rate": 1.2641799301488557e-05,
      "loss": 0.6745,
      "step": 5480
    },
    {
      "epoch": 2.37,
      "grad_norm": 2.3290923862939388,
      "learning_rate": 1.2627252852486978e-05,
      "loss": 0.5364,
      "step": 5485
    },
    {
      "epoch": 2.37,
      "grad_norm": 2.320282415146021,
      "learning_rate": 1.2612700429609231e-05,
      "loss": 0.5121,
      "step": 5490
    },
    {
      "epoch": 2.37,
      "grad_norm": 2.865501228371246,
      "learning_rate": 1.259814206594478e-05,
      "loss": 0.656,
      "step": 5495
    },
    {
      "epoch": 2.38,
      "grad_norm": 2.929289607937329,
      "learning_rate": 1.2583577794596588e-05,
      "loss": 0.6746,
      "step": 5500
    },
    {
      "epoch": 2.38,
      "grad_norm": 1.0123923944249917,
      "learning_rate": 1.256900764868105e-05,
      "loss": 0.7035,
      "step": 5505
    },
    {
      "epoch": 2.38,
      "grad_norm": 2.5671713727059498,
      "learning_rate": 1.2554431661327929e-05,
      "loss": 0.5534,
      "step": 5510
    },
    {
      "epoch": 2.38,
      "grad_norm": 2.713465275447893,
      "learning_rate": 1.253984986568026e-05,
      "loss": 0.6707,
      "step": 5515
    },
    {
      "epoch": 2.38,
      "grad_norm": 3.2388981875859666,
      "learning_rate": 1.2525262294894291e-05,
      "loss": 0.5429,
      "step": 5520
    },
    {
      "epoch": 2.39,
      "grad_norm": 4.429034301281724,
      "learning_rate": 1.25106689821394e-05,
      "loss": 0.634,
      "step": 5525
    },
    {
      "epoch": 2.39,
      "grad_norm": 3.527254640727882,
      "learning_rate": 1.2496069960598015e-05,
      "loss": 0.5531,
      "step": 5530
    },
    {
      "epoch": 2.39,
      "grad_norm": 3.9778021787359807,
      "learning_rate": 1.2481465263465556e-05,
      "loss": 0.5832,
      "step": 5535
    },
    {
      "epoch": 2.39,
      "grad_norm": 5.7570845677942994,
      "learning_rate": 1.2466854923950337e-05,
      "loss": 0.6605,
      "step": 5540
    },
    {
      "epoch": 2.4,
      "grad_norm": 2.5136660861714533,
      "learning_rate": 1.2452238975273517e-05,
      "loss": 0.6726,
      "step": 5545
    },
    {
      "epoch": 2.4,
      "grad_norm": 4.778955739958259,
      "learning_rate": 1.2437617450668991e-05,
      "loss": 0.6528,
      "step": 5550
    },
    {
      "epoch": 2.4,
      "grad_norm": 2.822118632646264,
      "learning_rate": 1.2422990383383342e-05,
      "loss": 0.7309,
      "step": 5555
    },
    {
      "epoch": 2.4,
      "grad_norm": 3.101702662145458,
      "learning_rate": 1.2408357806675755e-05,
      "loss": 0.6144,
      "step": 5560
    },
    {
      "epoch": 2.4,
      "grad_norm": 2.7283728602937067,
      "learning_rate": 1.2393719753817945e-05,
      "loss": 0.6402,
      "step": 5565
    },
    {
      "epoch": 2.41,
      "grad_norm": 3.941793534144611,
      "learning_rate": 1.2379076258094076e-05,
      "loss": 0.5565,
      "step": 5570
    },
    {
      "epoch": 2.41,
      "grad_norm": 2.0507521443057612,
      "learning_rate": 1.2364427352800688e-05,
      "loss": 0.5296,
      "step": 5575
    },
    {
      "epoch": 2.41,
      "grad_norm": 2.2791441959795167,
      "learning_rate": 1.2349773071246615e-05,
      "loss": 0.6071,
      "step": 5580
    },
    {
      "epoch": 2.41,
      "grad_norm": 1.4181544900587406,
      "learning_rate": 1.233511344675293e-05,
      "loss": 0.5631,
      "step": 5585
    },
    {
      "epoch": 2.41,
      "grad_norm": 3.0642106256181467,
      "learning_rate": 1.2320448512652845e-05,
      "loss": 0.5379,
      "step": 5590
    },
    {
      "epoch": 2.42,
      "grad_norm": 3.0825089074185787,
      "learning_rate": 1.2305778302291647e-05,
      "loss": 0.6778,
      "step": 5595
    },
    {
      "epoch": 2.42,
      "grad_norm": 2.3649728516693553,
      "learning_rate": 1.229110284902662e-05,
      "loss": 0.7069,
      "step": 5600
    },
    {
      "epoch": 2.42,
      "grad_norm": 4.418260582924241,
      "learning_rate": 1.2276422186226965e-05,
      "loss": 0.6248,
      "step": 5605
    },
    {
      "epoch": 2.42,
      "grad_norm": 3.393194032854879,
      "learning_rate": 1.2261736347273739e-05,
      "loss": 0.5673,
      "step": 5610
    },
    {
      "epoch": 2.43,
      "grad_norm": 3.256965708075714,
      "learning_rate": 1.2247045365559758e-05,
      "loss": 0.6426,
      "step": 5615
    },
    {
      "epoch": 2.43,
      "grad_norm": 1.9498668554848662,
      "learning_rate": 1.2232349274489543e-05,
      "loss": 0.6806,
      "step": 5620
    },
    {
      "epoch": 2.43,
      "grad_norm": 3.8443153243697936,
      "learning_rate": 1.2217648107479222e-05,
      "loss": 0.6157,
      "step": 5625
    },
    {
      "epoch": 2.43,
      "grad_norm": 2.4099977548359326,
      "learning_rate": 1.2202941897956468e-05,
      "loss": 0.5046,
      "step": 5630
    },
    {
      "epoch": 2.43,
      "grad_norm": 3.164543987009733,
      "learning_rate": 1.2188230679360417e-05,
      "loss": 0.6559,
      "step": 5635
    },
    {
      "epoch": 2.44,
      "grad_norm": 1.3937539752455785,
      "learning_rate": 1.2173514485141606e-05,
      "loss": 0.524,
      "step": 5640
    },
    {
      "epoch": 2.44,
      "grad_norm": 1.0846844450463282,
      "learning_rate": 1.2158793348761878e-05,
      "loss": 0.494,
      "step": 5645
    },
    {
      "epoch": 2.44,
      "grad_norm": 2.2885642706729197,
      "learning_rate": 1.2144067303694304e-05,
      "loss": 0.5825,
      "step": 5650
    },
    {
      "epoch": 2.44,
      "grad_norm": 1.3565551486811531,
      "learning_rate": 1.212933638342313e-05,
      "loss": 0.6171,
      "step": 5655
    },
    {
      "epoch": 2.44,
      "grad_norm": 1.328702556248931,
      "learning_rate": 1.211460062144369e-05,
      "loss": 0.5537,
      "step": 5660
    },
    {
      "epoch": 2.45,
      "grad_norm": 2.2535405004738798,
      "learning_rate": 1.2099860051262308e-05,
      "loss": 0.5915,
      "step": 5665
    },
    {
      "epoch": 2.45,
      "grad_norm": 2.2859547603183534,
      "learning_rate": 1.2085114706396264e-05,
      "loss": 0.5191,
      "step": 5670
    },
    {
      "epoch": 2.45,
      "grad_norm": 2.386753821740332,
      "learning_rate": 1.207036462037368e-05,
      "loss": 0.5119,
      "step": 5675
    },
    {
      "epoch": 2.45,
      "grad_norm": 0.8861234521427775,
      "learning_rate": 1.2055609826733459e-05,
      "loss": 0.5889,
      "step": 5680
    },
    {
      "epoch": 2.46,
      "grad_norm": 5.447653168410484,
      "learning_rate": 1.2040850359025216e-05,
      "loss": 0.656,
      "step": 5685
    },
    {
      "epoch": 2.46,
      "grad_norm": 2.874911271848961,
      "learning_rate": 1.2026086250809181e-05,
      "loss": 0.6001,
      "step": 5690
    },
    {
      "epoch": 2.46,
      "grad_norm": 3.116810605679269,
      "learning_rate": 1.201131753565616e-05,
      "loss": 0.6166,
      "step": 5695
    },
    {
      "epoch": 2.46,
      "grad_norm": 2.638413391139646,
      "learning_rate": 1.1996544247147405e-05,
      "loss": 0.5621,
      "step": 5700
    },
    {
      "epoch": 2.46,
      "grad_norm": 1.6687401902988719,
      "learning_rate": 1.1981766418874585e-05,
      "loss": 0.6026,
      "step": 5705
    },
    {
      "epoch": 2.47,
      "grad_norm": 2.3214904478487557,
      "learning_rate": 1.1966984084439688e-05,
      "loss": 0.6314,
      "step": 5710
    },
    {
      "epoch": 2.47,
      "grad_norm": 1.9982906558239586,
      "learning_rate": 1.1952197277454943e-05,
      "loss": 0.5246,
      "step": 5715
    },
    {
      "epoch": 2.47,
      "grad_norm": 1.2589374703878966,
      "learning_rate": 1.1937406031542758e-05,
      "loss": 0.4738,
      "step": 5720
    },
    {
      "epoch": 2.47,
      "grad_norm": 1.6703887329525249,
      "learning_rate": 1.1922610380335626e-05,
      "loss": 0.6611,
      "step": 5725
    },
    {
      "epoch": 2.48,
      "grad_norm": 1.9141396640048154,
      "learning_rate": 1.190781035747606e-05,
      "loss": 0.4246,
      "step": 5730
    },
    {
      "epoch": 2.48,
      "grad_norm": 1.7078230496055034,
      "learning_rate": 1.1893005996616516e-05,
      "loss": 0.637,
      "step": 5735
    },
    {
      "epoch": 2.48,
      "grad_norm": 1.8474307055974233,
      "learning_rate": 1.1878197331419306e-05,
      "loss": 0.6485,
      "step": 5740
    },
    {
      "epoch": 2.48,
      "grad_norm": 2.153158902463925,
      "learning_rate": 1.1863384395556538e-05,
      "loss": 0.5999,
      "step": 5745
    },
    {
      "epoch": 2.48,
      "grad_norm": 1.803005771677984,
      "learning_rate": 1.1848567222710027e-05,
      "loss": 0.5679,
      "step": 5750
    },
    {
      "epoch": 2.49,
      "grad_norm": 3.335477853835864,
      "learning_rate": 1.1833745846571222e-05,
      "loss": 0.6034,
      "step": 5755
    },
    {
      "epoch": 2.49,
      "grad_norm": 1.878366420821759,
      "learning_rate": 1.181892030084113e-05,
      "loss": 0.5744,
      "step": 5760
    },
    {
      "epoch": 2.49,
      "grad_norm": 2.0165345116465745,
      "learning_rate": 1.1804090619230235e-05,
      "loss": 0.4466,
      "step": 5765
    },
    {
      "epoch": 2.49,
      "grad_norm": 3.5504172356816466,
      "learning_rate": 1.1789256835458429e-05,
      "loss": 0.608,
      "step": 5770
    },
    {
      "epoch": 2.49,
      "grad_norm": 2.245792656102741,
      "learning_rate": 1.1774418983254939e-05,
      "loss": 0.5836,
      "step": 5775
    },
    {
      "epoch": 2.5,
      "grad_norm": 2.5643770754859148,
      "learning_rate": 1.1759577096358223e-05,
      "loss": 0.5998,
      "step": 5780
    },
    {
      "epoch": 2.5,
      "grad_norm": 3.401996030820411,
      "learning_rate": 1.1744731208515936e-05,
      "loss": 0.5699,
      "step": 5785
    },
    {
      "epoch": 2.5,
      "grad_norm": 2.1812346079873346,
      "learning_rate": 1.1729881353484813e-05,
      "loss": 0.6498,
      "step": 5790
    },
    {
      "epoch": 2.5,
      "grad_norm": 2.899854876371358,
      "learning_rate": 1.1715027565030618e-05,
      "loss": 0.5407,
      "step": 5795
    },
    {
      "epoch": 2.51,
      "grad_norm": 5.374546420936496,
      "learning_rate": 1.1700169876928057e-05,
      "loss": 0.7177,
      "step": 5800
    },
    {
      "epoch": 2.51,
      "grad_norm": 2.4212010959195664,
      "learning_rate": 1.1685308322960705e-05,
      "loss": 0.5105,
      "step": 5805
    },
    {
      "epoch": 2.51,
      "grad_norm": 0.9555119992317347,
      "learning_rate": 1.1670442936920923e-05,
      "loss": 0.5637,
      "step": 5810
    },
    {
      "epoch": 2.51,
      "grad_norm": 2.7380695984695365,
      "learning_rate": 1.1655573752609789e-05,
      "loss": 0.5801,
      "step": 5815
    },
    {
      "epoch": 2.51,
      "grad_norm": 2.5075837508881804,
      "learning_rate": 1.164070080383701e-05,
      "loss": 0.533,
      "step": 5820
    },
    {
      "epoch": 2.52,
      "grad_norm": 5.6592126518142765,
      "learning_rate": 1.1625824124420873e-05,
      "loss": 0.6799,
      "step": 5825
    },
    {
      "epoch": 2.52,
      "grad_norm": 2.6552120212255543,
      "learning_rate": 1.161094374818812e-05,
      "loss": 0.5297,
      "step": 5830
    },
    {
      "epoch": 2.52,
      "grad_norm": 2.2459110733361736,
      "learning_rate": 1.159605970897392e-05,
      "loss": 0.6522,
      "step": 5835
    },
    {
      "epoch": 2.52,
      "grad_norm": 2.446676539882337,
      "learning_rate": 1.1581172040621759e-05,
      "loss": 0.5082,
      "step": 5840
    },
    {
      "epoch": 2.52,
      "grad_norm": 2.9109242525611707,
      "learning_rate": 1.1566280776983385e-05,
      "loss": 0.5508,
      "step": 5845
    },
    {
      "epoch": 2.53,
      "grad_norm": 1.65087084427715,
      "learning_rate": 1.1551385951918714e-05,
      "loss": 0.5428,
      "step": 5850
    },
    {
      "epoch": 2.53,
      "grad_norm": 2.6075258277791344,
      "learning_rate": 1.1536487599295758e-05,
      "loss": 0.5509,
      "step": 5855
    },
    {
      "epoch": 2.53,
      "grad_norm": 3.7449485712262245,
      "learning_rate": 1.1521585752990556e-05,
      "loss": 0.5438,
      "step": 5860
    },
    {
      "epoch": 2.53,
      "grad_norm": 2.412438766415541,
      "learning_rate": 1.1506680446887088e-05,
      "loss": 0.5831,
      "step": 5865
    },
    {
      "epoch": 2.54,
      "grad_norm": 7.446052091029832,
      "learning_rate": 1.14917717148772e-05,
      "loss": 0.6561,
      "step": 5870
    },
    {
      "epoch": 2.54,
      "grad_norm": 4.023471716380351,
      "learning_rate": 1.1476859590860534e-05,
      "loss": 0.5676,
      "step": 5875
    },
    {
      "epoch": 2.54,
      "grad_norm": 1.1134623043453562,
      "learning_rate": 1.1461944108744436e-05,
      "loss": 0.6136,
      "step": 5880
    },
    {
      "epoch": 2.54,
      "grad_norm": 2.956812319516977,
      "learning_rate": 1.1447025302443896e-05,
      "loss": 0.5827,
      "step": 5885
    },
    {
      "epoch": 2.54,
      "grad_norm": 1.9327663678979348,
      "learning_rate": 1.1432103205881451e-05,
      "loss": 0.5183,
      "step": 5890
    },
    {
      "epoch": 2.55,
      "grad_norm": 1.8084939065396937,
      "learning_rate": 1.1417177852987132e-05,
      "loss": 0.4991,
      "step": 5895
    },
    {
      "epoch": 2.55,
      "grad_norm": 1.9430549423028185,
      "learning_rate": 1.1402249277698367e-05,
      "loss": 0.4841,
      "step": 5900
    },
    {
      "epoch": 2.55,
      "grad_norm": 2.8708341115461233,
      "learning_rate": 1.1387317513959914e-05,
      "loss": 0.6687,
      "step": 5905
    },
    {
      "epoch": 2.55,
      "grad_norm": 2.529911695150623,
      "learning_rate": 1.1372382595723783e-05,
      "loss": 0.5151,
      "step": 5910
    },
    {
      "epoch": 2.56,
      "grad_norm": 2.4366426081266983,
      "learning_rate": 1.1357444556949151e-05,
      "loss": 0.5137,
      "step": 5915
    },
    {
      "epoch": 2.56,
      "grad_norm": 3.2681063789494766,
      "learning_rate": 1.134250343160229e-05,
      "loss": 0.5975,
      "step": 5920
    },
    {
      "epoch": 2.56,
      "grad_norm": 4.32348306671456,
      "learning_rate": 1.1327559253656502e-05,
      "loss": 0.561,
      "step": 5925
    },
    {
      "epoch": 2.56,
      "grad_norm": 3.027049309126319,
      "learning_rate": 1.1312612057092011e-05,
      "loss": 0.5233,
      "step": 5930
    },
    {
      "epoch": 2.56,
      "grad_norm": 3.013990620653381,
      "learning_rate": 1.1297661875895927e-05,
      "loss": 0.5997,
      "step": 5935
    },
    {
      "epoch": 2.57,
      "grad_norm": 3.0287001852233413,
      "learning_rate": 1.1282708744062125e-05,
      "loss": 0.509,
      "step": 5940
    },
    {
      "epoch": 2.57,
      "grad_norm": 1.79953582506735,
      "learning_rate": 1.1267752695591203e-05,
      "loss": 0.622,
      "step": 5945
    },
    {
      "epoch": 2.57,
      "grad_norm": 2.5864541268100023,
      "learning_rate": 1.125279376449039e-05,
      "loss": 0.4962,
      "step": 5950
    },
    {
      "epoch": 2.57,
      "grad_norm": 2.8771122683468864,
      "learning_rate": 1.1237831984773462e-05,
      "loss": 0.5531,
      "step": 5955
    },
    {
      "epoch": 2.57,
      "grad_norm": 2.0367422844037546,
      "learning_rate": 1.1222867390460681e-05,
      "loss": 0.5465,
      "step": 5960
    },
    {
      "epoch": 2.58,
      "grad_norm": 1.267993602452498,
      "learning_rate": 1.12079000155787e-05,
      "loss": 0.4367,
      "step": 5965
    },
    {
      "epoch": 2.58,
      "grad_norm": 2.135027107867463,
      "learning_rate": 1.11929298941605e-05,
      "loss": 0.4705,
      "step": 5970
    },
    {
      "epoch": 2.58,
      "grad_norm": 1.7457748728023093,
      "learning_rate": 1.1177957060245311e-05,
      "loss": 0.5437,
      "step": 5975
    },
    {
      "epoch": 2.58,
      "grad_norm": 1.912347529458289,
      "learning_rate": 1.1162981547878518e-05,
      "loss": 0.5349,
      "step": 5980
    },
    {
      "epoch": 2.59,
      "grad_norm": 4.50751108797099,
      "learning_rate": 1.1148003391111611e-05,
      "loss": 0.6474,
      "step": 5985
    },
    {
      "epoch": 2.59,
      "grad_norm": 1.5938058439774518,
      "learning_rate": 1.1133022624002083e-05,
      "loss": 0.6542,
      "step": 5990
    },
    {
      "epoch": 2.59,
      "grad_norm": 4.147373275718441,
      "learning_rate": 1.1118039280613365e-05,
      "loss": 0.5725,
      "step": 5995
    },
    {
      "epoch": 2.59,
      "grad_norm": 2.6356352190755605,
      "learning_rate": 1.1103053395014749e-05,
      "loss": 0.553,
      "step": 6000
    },
    {
      "epoch": 2.59,
      "grad_norm": 2.498630347261751,
      "learning_rate": 1.1088065001281302e-05,
      "loss": 0.5169,
      "step": 6005
    },
    {
      "epoch": 2.6,
      "grad_norm": 6.7469795499587555,
      "learning_rate": 1.1073074133493802e-05,
      "loss": 0.6078,
      "step": 6010
    },
    {
      "epoch": 2.6,
      "grad_norm": 5.1539900000529775,
      "learning_rate": 1.1058080825738643e-05,
      "loss": 0.5672,
      "step": 6015
    },
    {
      "epoch": 2.6,
      "grad_norm": 2.8156989728257864,
      "learning_rate": 1.104308511210777e-05,
      "loss": 0.4806,
      "step": 6020
    },
    {
      "epoch": 2.6,
      "grad_norm": 3.3213185297521712,
      "learning_rate": 1.1028087026698604e-05,
      "loss": 0.4861,
      "step": 6025
    },
    {
      "epoch": 2.6,
      "grad_norm": 3.5440873009748146,
      "learning_rate": 1.1013086603613953e-05,
      "loss": 0.4882,
      "step": 6030
    },
    {
      "epoch": 2.61,
      "grad_norm": 2.6503633600511955,
      "learning_rate": 1.0998083876961948e-05,
      "loss": 0.5048,
      "step": 6035
    },
    {
      "epoch": 2.61,
      "grad_norm": 1.159740466993807,
      "learning_rate": 1.0983078880855942e-05,
      "loss": 0.4544,
      "step": 6040
    },
    {
      "epoch": 2.61,
      "grad_norm": 3.28865816392566,
      "learning_rate": 1.0968071649414465e-05,
      "loss": 0.509,
      "step": 6045
    },
    {
      "epoch": 2.61,
      "grad_norm": 2.0646591256328293,
      "learning_rate": 1.0953062216761123e-05,
      "loss": 0.5019,
      "step": 6050
    },
    {
      "epoch": 2.62,
      "grad_norm": 2.352642135677423,
      "learning_rate": 1.0938050617024528e-05,
      "loss": 0.5987,
      "step": 6055
    },
    {
      "epoch": 2.62,
      "grad_norm": 2.457968428431282,
      "learning_rate": 1.0923036884338218e-05,
      "loss": 0.5434,
      "step": 6060
    },
    {
      "epoch": 2.62,
      "grad_norm": 4.850215654720314,
      "learning_rate": 1.0908021052840582e-05,
      "loss": 0.5991,
      "step": 6065
    },
    {
      "epoch": 2.62,
      "grad_norm": 1.6414624184793067,
      "learning_rate": 1.089300315667478e-05,
      "loss": 0.467,
      "step": 6070
    },
    {
      "epoch": 2.62,
      "grad_norm": 1.8565846208125867,
      "learning_rate": 1.0877983229988671e-05,
      "loss": 0.4879,
      "step": 6075
    },
    {
      "epoch": 2.63,
      "grad_norm": 2.57669455439607,
      "learning_rate": 1.0862961306934727e-05,
      "loss": 0.5034,
      "step": 6080
    },
    {
      "epoch": 2.63,
      "grad_norm": 2.1781249624833516,
      "learning_rate": 1.0847937421669958e-05,
      "loss": 0.5489,
      "step": 6085
    },
    {
      "epoch": 2.63,
      "grad_norm": 3.2642079291697113,
      "learning_rate": 1.083291160835584e-05,
      "loss": 0.5428,
      "step": 6090
    },
    {
      "epoch": 2.63,
      "grad_norm": 1.520362141386093,
      "learning_rate": 1.0817883901158233e-05,
      "loss": 0.527,
      "step": 6095
    },
    {
      "epoch": 2.63,
      "grad_norm": 1.7772580788569998,
      "learning_rate": 1.0802854334247297e-05,
      "loss": 0.5083,
      "step": 6100
    },
    {
      "epoch": 2.64,
      "grad_norm": 1.996720547533441,
      "learning_rate": 1.0787822941797429e-05,
      "loss": 0.5448,
      "step": 6105
    },
    {
      "epoch": 2.64,
      "grad_norm": 1.523628857374489,
      "learning_rate": 1.0772789757987168e-05,
      "loss": 0.5466,
      "step": 6110
    },
    {
      "epoch": 2.64,
      "grad_norm": 1.242041010323239,
      "learning_rate": 1.0757754816999136e-05,
      "loss": 0.5819,
      "step": 6115
    },
    {
      "epoch": 2.64,
      "grad_norm": 2.1655405402416563,
      "learning_rate": 1.0742718153019945e-05,
      "loss": 0.453,
      "step": 6120
    },
    {
      "epoch": 2.65,
      "grad_norm": 2.6841517683072853,
      "learning_rate": 1.0727679800240123e-05,
      "loss": 0.434,
      "step": 6125
    },
    {
      "epoch": 2.65,
      "grad_norm": 2.7477547553589803,
      "learning_rate": 1.071263979285404e-05,
      "loss": 0.5315,
      "step": 6130
    },
    {
      "epoch": 2.65,
      "grad_norm": 2.485348126111469,
      "learning_rate": 1.0697598165059833e-05,
      "loss": 0.5192,
      "step": 6135
    },
    {
      "epoch": 2.65,
      "grad_norm": 3.066169512052378,
      "learning_rate": 1.0682554951059316e-05,
      "loss": 0.4982,
      "step": 6140
    },
    {
      "epoch": 2.65,
      "grad_norm": 1.4248750822788414,
      "learning_rate": 1.0667510185057915e-05,
      "loss": 0.5432,
      "step": 6145
    },
    {
      "epoch": 2.66,
      "grad_norm": 2.0714398287890092,
      "learning_rate": 1.0652463901264578e-05,
      "loss": 0.6539,
      "step": 6150
    },
    {
      "epoch": 2.66,
      "grad_norm": 5.174378433465227,
      "learning_rate": 1.0637416133891714e-05,
      "loss": 0.6034,
      "step": 6155
    },
    {
      "epoch": 2.66,
      "grad_norm": 3.170144848595313,
      "learning_rate": 1.06223669171551e-05,
      "loss": 0.4693,
      "step": 6160
    },
    {
      "epoch": 2.66,
      "grad_norm": 2.301582029494036,
      "learning_rate": 1.0607316285273808e-05,
      "loss": 0.5155,
      "step": 6165
    },
    {
      "epoch": 2.67,
      "grad_norm": 2.455215341523393,
      "learning_rate": 1.0592264272470128e-05,
      "loss": 0.5141,
      "step": 6170
    },
    {
      "epoch": 2.67,
      "grad_norm": 3.0805989635867594,
      "learning_rate": 1.0577210912969492e-05,
      "loss": 0.4961,
      "step": 6175
    },
    {
      "epoch": 2.67,
      "grad_norm": 4.048856790949565,
      "learning_rate": 1.0562156241000389e-05,
      "loss": 0.6041,
      "step": 6180
    },
    {
      "epoch": 2.67,
      "grad_norm": 3.2536278909993346,
      "learning_rate": 1.0547100290794303e-05,
      "loss": 0.4917,
      "step": 6185
    },
    {
      "epoch": 2.67,
      "grad_norm": 5.798717038594271,
      "learning_rate": 1.0532043096585613e-05,
      "loss": 0.5321,
      "step": 6190
    },
    {
      "epoch": 2.68,
      "grad_norm": 4.3837410681270805,
      "learning_rate": 1.0516984692611533e-05,
      "loss": 0.6015,
      "step": 6195
    },
    {
      "epoch": 2.68,
      "grad_norm": 4.576176396461095,
      "learning_rate": 1.0501925113112024e-05,
      "loss": 0.6082,
      "step": 6200
    },
    {
      "epoch": 2.68,
      "grad_norm": 1.707471946131464,
      "learning_rate": 1.0486864392329723e-05,
      "loss": 0.508,
      "step": 6205
    },
    {
      "epoch": 2.68,
      "grad_norm": 2.2217634890861824,
      "learning_rate": 1.0471802564509865e-05,
      "loss": 0.5729,
      "step": 6210
    },
    {
      "epoch": 2.68,
      "grad_norm": 1.6203039931412289,
      "learning_rate": 1.045673966390019e-05,
      "loss": 0.5472,
      "step": 6215
    },
    {
      "epoch": 2.69,
      "grad_norm": 2.4729542822145607,
      "learning_rate": 1.0441675724750895e-05,
      "loss": 0.533,
      "step": 6220
    },
    {
      "epoch": 2.69,
      "grad_norm": 1.841780535568983,
      "learning_rate": 1.042661078131452e-05,
      "loss": 0.468,
      "step": 6225
    },
    {
      "epoch": 2.69,
      "grad_norm": 0.9034137425899711,
      "learning_rate": 1.0411544867845904e-05,
      "loss": 0.4318,
      "step": 6230
    },
    {
      "epoch": 2.69,
      "grad_norm": 2.2121584496901514,
      "learning_rate": 1.0396478018602083e-05,
      "loss": 0.4904,
      "step": 6235
    },
    {
      "epoch": 2.7,
      "grad_norm": 2.0715237369990356,
      "learning_rate": 1.038141026784222e-05,
      "loss": 0.4411,
      "step": 6240
    },
    {
      "epoch": 2.7,
      "grad_norm": 3.8613319036660907,
      "learning_rate": 1.0366341649827538e-05,
      "loss": 0.549,
      "step": 6245
    },
    {
      "epoch": 2.7,
      "grad_norm": 4.138638245292264,
      "learning_rate": 1.0351272198821218e-05,
      "loss": 0.5471,
      "step": 6250
    },
    {
      "epoch": 2.7,
      "grad_norm": 3.4322114787684606,
      "learning_rate": 1.0336201949088344e-05,
      "loss": 0.5714,
      "step": 6255
    },
    {
      "epoch": 2.7,
      "grad_norm": 3.7460770229320297,
      "learning_rate": 1.0321130934895818e-05,
      "loss": 0.5299,
      "step": 6260
    },
    {
      "epoch": 2.71,
      "grad_norm": 1.0187985411942515,
      "learning_rate": 1.030605919051227e-05,
      "loss": 0.4599,
      "step": 6265
    },
    {
      "epoch": 2.71,
      "grad_norm": 3.060934636461662,
      "learning_rate": 1.0290986750208001e-05,
      "loss": 0.4518,
      "step": 6270
    },
    {
      "epoch": 2.71,
      "grad_norm": 2.6057030776915115,
      "learning_rate": 1.0275913648254887e-05,
      "loss": 0.4628,
      "step": 6275
    },
    {
      "epoch": 2.71,
      "grad_norm": 1.6893046080930116,
      "learning_rate": 1.0260839918926314e-05,
      "loss": 0.4309,
      "step": 6280
    },
    {
      "epoch": 2.71,
      "grad_norm": 1.4400368288913405,
      "learning_rate": 1.024576559649709e-05,
      "loss": 0.5131,
      "step": 6285
    },
    {
      "epoch": 2.72,
      "grad_norm": 1.070692376269189,
      "learning_rate": 1.0230690715243375e-05,
      "loss": 0.4188,
      "step": 6290
    },
    {
      "epoch": 2.72,
      "grad_norm": 2.058634828733364,
      "learning_rate": 1.02156153094426e-05,
      "loss": 0.5291,
      "step": 6295
    },
    {
      "epoch": 2.72,
      "grad_norm": 3.4252175800130167,
      "learning_rate": 1.0200539413373381e-05,
      "loss": 0.4725,
      "step": 6300
    },
    {
      "epoch": 2.72,
      "grad_norm": 3.9280466206448046,
      "learning_rate": 1.018546306131546e-05,
      "loss": 0.4743,
      "step": 6305
    },
    {
      "epoch": 2.73,
      "grad_norm": 2.151132653512246,
      "learning_rate": 1.0170386287549612e-05,
      "loss": 0.4581,
      "step": 6310
    },
    {
      "epoch": 2.73,
      "grad_norm": 1.2839090665270678,
      "learning_rate": 1.0155309126357562e-05,
      "loss": 0.5197,
      "step": 6315
    },
    {
      "epoch": 2.73,
      "grad_norm": 1.075875412375835,
      "learning_rate": 1.0140231612021931e-05,
      "loss": 0.4433,
      "step": 6320
    },
    {
      "epoch": 2.73,
      "grad_norm": 2.632088454474315,
      "learning_rate": 1.0125153778826133e-05,
      "loss": 0.4793,
      "step": 6325
    },
    {
      "epoch": 2.73,
      "grad_norm": 2.4920741381531326,
      "learning_rate": 1.0110075661054306e-05,
      "loss": 0.5394,
      "step": 6330
    },
    {
      "epoch": 2.74,
      "grad_norm": 1.018665224091436,
      "learning_rate": 1.0094997292991244e-05,
      "loss": 0.3782,
      "step": 6335
    },
    {
      "epoch": 2.74,
      "grad_norm": 2.5508307612111056,
      "learning_rate": 1.0079918708922303e-05,
      "loss": 0.4532,
      "step": 6340
    },
    {
      "epoch": 2.74,
      "grad_norm": 1.655282795993782,
      "learning_rate": 1.0064839943133328e-05,
      "loss": 0.4194,
      "step": 6345
    },
    {
      "epoch": 2.74,
      "grad_norm": 2.361478377401841,
      "learning_rate": 1.0049761029910585e-05,
      "loss": 0.5437,
      "step": 6350
    },
    {
      "epoch": 2.75,
      "grad_norm": 3.649049573919179,
      "learning_rate": 1.003468200354067e-05,
      "loss": 0.543,
      "step": 6355
    },
    {
      "epoch": 2.75,
      "grad_norm": 1.2283646653492042,
      "learning_rate": 1.0019602898310438e-05,
      "loss": 0.502,
      "step": 6360
    },
    {
      "epoch": 2.75,
      "grad_norm": 3.1541246866307904,
      "learning_rate": 1.000452374850692e-05,
      "loss": 0.4247,
      "step": 6365
    },
    {
      "epoch": 2.75,
      "grad_norm": 4.11299828550103,
      "learning_rate": 9.989444588417254e-06,
      "loss": 0.5446,
      "step": 6370
    },
    {
      "epoch": 2.75,
      "grad_norm": 4.914138694275413,
      "learning_rate": 9.9743654523286e-06,
      "loss": 0.4956,
      "step": 6375
    },
    {
      "epoch": 2.76,
      "grad_norm": 2.9470768143393022,
      "learning_rate": 9.959286374528059e-06,
      "loss": 0.4672,
      "step": 6380
    },
    {
      "epoch": 2.76,
      "grad_norm": 2.1677074867145225,
      "learning_rate": 9.944207389302603e-06,
      "loss": 0.5284,
      "step": 6385
    },
    {
      "epoch": 2.76,
      "grad_norm": 1.9429131304932807,
      "learning_rate": 9.929128530938996e-06,
      "loss": 0.4724,
      "step": 6390
    },
    {
      "epoch": 2.76,
      "grad_norm": 1.918155595325713,
      "learning_rate": 9.9140498337237e-06,
      "loss": 0.4078,
      "step": 6395
    },
    {
      "epoch": 2.76,
      "grad_norm": 1.2710780443623437,
      "learning_rate": 9.898971331942836e-06,
      "loss": 0.4892,
      "step": 6400
    },
    {
      "epoch": 2.77,
      "grad_norm": 2.0197316051827365,
      "learning_rate": 9.883893059882054e-06,
      "loss": 0.4773,
      "step": 6405
    },
    {
      "epoch": 2.77,
      "grad_norm": 1.9979343758482182,
      "learning_rate": 9.8688150518265e-06,
      "loss": 0.4322,
      "step": 6410
    },
    {
      "epoch": 2.77,
      "grad_norm": 2.9015111491639023,
      "learning_rate": 9.853737342060711e-06,
      "loss": 0.4626,
      "step": 6415
    },
    {
      "epoch": 2.77,
      "grad_norm": 1.9284808560910882,
      "learning_rate": 9.838659964868545e-06,
      "loss": 0.4723,
      "step": 6420
    },
    {
      "epoch": 2.78,
      "grad_norm": 1.696924508811764,
      "learning_rate": 9.82358295453311e-06,
      "loss": 0.4322,
      "step": 6425
    },
    {
      "epoch": 2.78,
      "grad_norm": 1.5827504845958853,
      "learning_rate": 9.808506345336675e-06,
      "loss": 0.4958,
      "step": 6430
    },
    {
      "epoch": 2.78,
      "grad_norm": 2.933433261057401,
      "learning_rate": 9.793430171560602e-06,
      "loss": 0.4268,
      "step": 6435
    },
    {
      "epoch": 2.78,
      "grad_norm": 2.4259662609504917,
      "learning_rate": 9.778354467485255e-06,
      "loss": 0.4229,
      "step": 6440
    },
    {
      "epoch": 2.78,
      "grad_norm": 2.5082682873508784,
      "learning_rate": 9.763279267389929e-06,
      "loss": 0.4694,
      "step": 6445
    },
    {
      "epoch": 2.79,
      "grad_norm": 2.25023974453689,
      "learning_rate": 9.74820460555279e-06,
      "loss": 0.4411,
      "step": 6450
    },
    {
      "epoch": 2.79,
      "grad_norm": 1.4082748459739975,
      "learning_rate": 9.733130516250761e-06,
      "loss": 0.355,
      "step": 6455
    },
    {
      "epoch": 2.79,
      "grad_norm": 1.4067199247341755,
      "learning_rate": 9.718057033759474e-06,
      "loss": 0.4116,
      "step": 6460
    },
    {
      "epoch": 2.79,
      "grad_norm": 1.7519299418804521,
      "learning_rate": 9.702984192353176e-06,
      "loss": 0.4818,
      "step": 6465
    },
    {
      "epoch": 2.79,
      "grad_norm": 3.4175019738202104,
      "learning_rate": 9.687912026304653e-06,
      "loss": 0.5346,
      "step": 6470
    },
    {
      "epoch": 2.8,
      "grad_norm": 1.9864314826474614,
      "learning_rate": 9.67284056988517e-06,
      "loss": 0.4581,
      "step": 6475
    },
    {
      "epoch": 2.8,
      "grad_norm": 3.138077655049651,
      "learning_rate": 9.657769857364363e-06,
      "loss": 0.518,
      "step": 6480
    },
    {
      "epoch": 2.8,
      "grad_norm": 2.502870249731197,
      "learning_rate": 9.642699923010185e-06,
      "loss": 0.4508,
      "step": 6485
    },
    {
      "epoch": 2.8,
      "grad_norm": 1.7930765574987866,
      "learning_rate": 9.627630801088816e-06,
      "loss": 0.4101,
      "step": 6490
    },
    {
      "epoch": 2.81,
      "grad_norm": 2.544947230994809,
      "learning_rate": 9.612562525864584e-06,
      "loss": 0.4858,
      "step": 6495
    },
    {
      "epoch": 2.81,
      "grad_norm": 4.137243268762429,
      "learning_rate": 9.597495131599906e-06,
      "loss": 0.4585,
      "step": 6500
    },
    {
      "epoch": 2.81,
      "grad_norm": 3.9778120532643086,
      "learning_rate": 9.582428652555185e-06,
      "loss": 0.4436,
      "step": 6505
    },
    {
      "epoch": 2.81,
      "grad_norm": 2.2441182670565736,
      "learning_rate": 9.567363122988741e-06,
      "loss": 0.4107,
      "step": 6510
    },
    {
      "epoch": 2.81,
      "grad_norm": 1.141091245400905,
      "learning_rate": 9.552298577156747e-06,
      "loss": 0.4256,
      "step": 6515
    },
    {
      "epoch": 2.82,
      "grad_norm": 1.7129366841254983,
      "learning_rate": 9.537235049313121e-06,
      "loss": 0.3952,
      "step": 6520
    },
    {
      "epoch": 2.82,
      "grad_norm": 1.236988405519333,
      "learning_rate": 9.522172573709488e-06,
      "loss": 0.4265,
      "step": 6525
    },
    {
      "epoch": 2.82,
      "grad_norm": 1.576104076344442,
      "learning_rate": 9.507111184595062e-06,
      "loss": 0.4207,
      "step": 6530
    },
    {
      "epoch": 2.82,
      "grad_norm": 1.3526871132184435,
      "learning_rate": 9.492050916216593e-06,
      "loss": 0.5159,
      "step": 6535
    },
    {
      "epoch": 2.83,
      "grad_norm": 2.2800239749148594,
      "learning_rate": 9.47699180281829e-06,
      "loss": 0.4127,
      "step": 6540
    },
    {
      "epoch": 2.83,
      "grad_norm": 1.2561916542114218,
      "learning_rate": 9.461933878641718e-06,
      "loss": 0.3991,
      "step": 6545
    },
    {
      "epoch": 2.83,
      "grad_norm": 2.2628973811496476,
      "learning_rate": 9.446877177925764e-06,
      "loss": 0.4333,
      "step": 6550
    },
    {
      "epoch": 2.83,
      "grad_norm": 1.4247293460628585,
      "learning_rate": 9.431821734906506e-06,
      "loss": 0.4714,
      "step": 6555
    },
    {
      "epoch": 2.83,
      "grad_norm": 1.619837488784781,
      "learning_rate": 9.416767583817178e-06,
      "loss": 0.3538,
      "step": 6560
    },
    {
      "epoch": 2.84,
      "grad_norm": 1.598948161513362,
      "learning_rate": 9.401714758888076e-06,
      "loss": 0.4258,
      "step": 6565
    },
    {
      "epoch": 2.84,
      "grad_norm": 1.0926972438062792,
      "learning_rate": 9.386663294346472e-06,
      "loss": 0.4779,
      "step": 6570
    },
    {
      "epoch": 2.84,
      "grad_norm": 2.508482094322588,
      "learning_rate": 9.371613224416554e-06,
      "loss": 0.4806,
      "step": 6575
    },
    {
      "epoch": 2.84,
      "grad_norm": 1.7024240871896414,
      "learning_rate": 9.356564583319335e-06,
      "loss": 0.4665,
      "step": 6580
    },
    {
      "epoch": 2.84,
      "grad_norm": 2.0503223331113056,
      "learning_rate": 9.341517405272575e-06,
      "loss": 0.4305,
      "step": 6585
    },
    {
      "epoch": 2.85,
      "grad_norm": 1.443289846371451,
      "learning_rate": 9.326471724490717e-06,
      "loss": 0.4521,
      "step": 6590
    },
    {
      "epoch": 2.85,
      "grad_norm": 2.2239936478869726,
      "learning_rate": 9.311427575184788e-06,
      "loss": 0.4195,
      "step": 6595
    },
    {
      "epoch": 2.85,
      "grad_norm": 1.5756273116741237,
      "learning_rate": 9.296384991562346e-06,
      "loss": 0.4334,
      "step": 6600
    },
    {
      "epoch": 2.85,
      "grad_norm": 1.6584753210048995,
      "learning_rate": 9.281344007827376e-06,
      "loss": 0.4295,
      "step": 6605
    },
    {
      "epoch": 2.86,
      "grad_norm": 2.8103529367397115,
      "learning_rate": 9.26630465818023e-06,
      "loss": 0.4584,
      "step": 6610
    },
    {
      "epoch": 2.86,
      "grad_norm": 2.691772042232581,
      "learning_rate": 9.25126697681755e-06,
      "loss": 0.4621,
      "step": 6615
    },
    {
      "epoch": 2.86,
      "grad_norm": 2.7735396415701468,
      "learning_rate": 9.236230997932174e-06,
      "loss": 0.4358,
      "step": 6620
    },
    {
      "epoch": 2.86,
      "grad_norm": 2.8240980703269316,
      "learning_rate": 9.221196755713082e-06,
      "loss": 0.4101,
      "step": 6625
    },
    {
      "epoch": 2.86,
      "grad_norm": 5.0412248180564605,
      "learning_rate": 9.206164284345291e-06,
      "loss": 0.4649,
      "step": 6630
    },
    {
      "epoch": 2.87,
      "grad_norm": 1.1581478268127954,
      "learning_rate": 9.1911336180098e-06,
      "loss": 0.413,
      "step": 6635
    },
    {
      "epoch": 2.87,
      "grad_norm": 4.453405657732692,
      "learning_rate": 9.176104790883504e-06,
      "loss": 0.4488,
      "step": 6640
    },
    {
      "epoch": 2.87,
      "grad_norm": 3.012774695677809,
      "learning_rate": 9.16107783713911e-06,
      "loss": 0.4058,
      "step": 6645
    },
    {
      "epoch": 2.87,
      "grad_norm": 1.2290176178612868,
      "learning_rate": 9.146052790945072e-06,
      "loss": 0.3824,
      "step": 6650
    },
    {
      "epoch": 2.87,
      "grad_norm": 2.232066797254159,
      "learning_rate": 9.131029686465506e-06,
      "loss": 0.506,
      "step": 6655
    },
    {
      "epoch": 2.88,
      "grad_norm": 4.16063650508915,
      "learning_rate": 9.116008557860106e-06,
      "loss": 0.4308,
      "step": 6660
    },
    {
      "epoch": 2.88,
      "grad_norm": 1.8369268794794247,
      "learning_rate": 9.100989439284081e-06,
      "loss": 0.4339,
      "step": 6665
    },
    {
      "epoch": 2.88,
      "grad_norm": 1.4077468997120663,
      "learning_rate": 9.08597236488806e-06,
      "loss": 0.413,
      "step": 6670
    },
    {
      "epoch": 2.88,
      "grad_norm": 1.2057626155911438,
      "learning_rate": 9.070957368818043e-06,
      "loss": 0.4537,
      "step": 6675
    },
    {
      "epoch": 2.89,
      "grad_norm": 1.5566729642358135,
      "learning_rate": 9.055944485215283e-06,
      "loss": 0.3966,
      "step": 6680
    },
    {
      "epoch": 2.89,
      "grad_norm": 2.600064929447201,
      "learning_rate": 9.04093374821624e-06,
      "loss": 0.4803,
      "step": 6685
    },
    {
      "epoch": 2.89,
      "grad_norm": 1.6680538318410998,
      "learning_rate": 9.02592519195249e-06,
      "loss": 0.4409,
      "step": 6690
    },
    {
      "epoch": 2.89,
      "grad_norm": 2.1954047183681937,
      "learning_rate": 9.01091885055065e-06,
      "loss": 0.4342,
      "step": 6695
    },
    {
      "epoch": 2.89,
      "grad_norm": 4.49678925419594,
      "learning_rate": 8.99591475813231e-06,
      "loss": 0.4316,
      "step": 6700
    },
    {
      "epoch": 2.9,
      "grad_norm": 2.2027231960939346,
      "learning_rate": 8.980912948813932e-06,
      "loss": 0.3865,
      "step": 6705
    },
    {
      "epoch": 2.9,
      "grad_norm": 2.581265968046604,
      "learning_rate": 8.965913456706796e-06,
      "loss": 0.4989,
      "step": 6710
    },
    {
      "epoch": 2.9,
      "grad_norm": 1.6644182265975438,
      "learning_rate": 8.950916315916912e-06,
      "loss": 0.3984,
      "step": 6715
    },
    {
      "epoch": 2.9,
      "grad_norm": 1.4570520063725072,
      "learning_rate": 8.935921560544938e-06,
      "loss": 0.437,
      "step": 6720
    },
    {
      "epoch": 2.9,
      "grad_norm": 1.1355513140288211,
      "learning_rate": 8.920929224686117e-06,
      "loss": 0.3342,
      "step": 6725
    },
    {
      "epoch": 2.91,
      "grad_norm": 1.7876848271437962,
      "learning_rate": 8.905939342430185e-06,
      "loss": 0.4636,
      "step": 6730
    },
    {
      "epoch": 2.91,
      "grad_norm": 2.4404383696926386,
      "learning_rate": 8.890951947861296e-06,
      "loss": 0.44,
      "step": 6735
    },
    {
      "epoch": 2.91,
      "grad_norm": 1.0600647901504878,
      "learning_rate": 8.875967075057953e-06,
      "loss": 0.4511,
      "step": 6740
    },
    {
      "epoch": 2.91,
      "grad_norm": 1.9431591387196776,
      "learning_rate": 8.86098475809292e-06,
      "loss": 0.3206,
      "step": 6745
    },
    {
      "epoch": 2.92,
      "grad_norm": 1.727807582573479,
      "learning_rate": 8.84600503103316e-06,
      "loss": 0.414,
      "step": 6750
    },
    {
      "epoch": 2.92,
      "grad_norm": 1.923727161874462,
      "learning_rate": 8.831027927939736e-06,
      "loss": 0.415,
      "step": 6755
    },
    {
      "epoch": 2.92,
      "grad_norm": 3.373896731208882,
      "learning_rate": 8.816053482867745e-06,
      "loss": 0.527,
      "step": 6760
    },
    {
      "epoch": 2.92,
      "grad_norm": 1.5335921617244541,
      "learning_rate": 8.80108172986625e-06,
      "loss": 0.4036,
      "step": 6765
    },
    {
      "epoch": 2.92,
      "grad_norm": 1.9471891820925509,
      "learning_rate": 8.786112702978176e-06,
      "loss": 0.5486,
      "step": 6770
    },
    {
      "epoch": 2.93,
      "grad_norm": 2.2347916473025022,
      "learning_rate": 8.77114643624027e-06,
      "loss": 0.4457,
      "step": 6775
    },
    {
      "epoch": 2.93,
      "grad_norm": 2.3678639583636265,
      "learning_rate": 8.756182963682987e-06,
      "loss": 0.4018,
      "step": 6780
    },
    {
      "epoch": 2.93,
      "grad_norm": 2.6214495125461847,
      "learning_rate": 8.741222319330434e-06,
      "loss": 0.4425,
      "step": 6785
    },
    {
      "epoch": 2.93,
      "grad_norm": 2.5343906167412498,
      "learning_rate": 8.72626453720029e-06,
      "loss": 0.4468,
      "step": 6790
    },
    {
      "epoch": 2.94,
      "grad_norm": 1.9361513781363973,
      "learning_rate": 8.711309651303718e-06,
      "loss": 0.378,
      "step": 6795
    },
    {
      "epoch": 2.94,
      "grad_norm": 4.69711181302915,
      "learning_rate": 8.696357695645308e-06,
      "loss": 0.3949,
      "step": 6800
    },
    {
      "epoch": 2.94,
      "grad_norm": 5.346661659511524,
      "learning_rate": 8.681408704222974e-06,
      "loss": 0.4211,
      "step": 6805
    },
    {
      "epoch": 2.94,
      "grad_norm": 2.449125693644456,
      "learning_rate": 8.666462711027894e-06,
      "loss": 0.3589,
      "step": 6810
    },
    {
      "epoch": 2.94,
      "grad_norm": 1.6096907460963736,
      "learning_rate": 8.651519750044438e-06,
      "loss": 0.3983,
      "step": 6815
    },
    {
      "epoch": 2.95,
      "grad_norm": 1.3264924866914174,
      "learning_rate": 8.636579855250062e-06,
      "loss": 0.394,
      "step": 6820
    },
    {
      "epoch": 2.95,
      "grad_norm": 4.37270376441934,
      "learning_rate": 8.621643060615272e-06,
      "loss": 0.4499,
      "step": 6825
    },
    {
      "epoch": 2.95,
      "grad_norm": 1.6547788544441633,
      "learning_rate": 8.606709400103512e-06,
      "loss": 0.4202,
      "step": 6830
    },
    {
      "epoch": 2.95,
      "grad_norm": 2.0890274265714206,
      "learning_rate": 8.591778907671099e-06,
      "loss": 0.3356,
      "step": 6835
    },
    {
      "epoch": 2.95,
      "grad_norm": 4.268006056243279,
      "learning_rate": 8.576851617267151e-06,
      "loss": 0.3998,
      "step": 6840
    },
    {
      "epoch": 2.96,
      "grad_norm": 2.863642651618915,
      "learning_rate": 8.5619275628335e-06,
      "loss": 0.3956,
      "step": 6845
    },
    {
      "epoch": 2.96,
      "grad_norm": 1.1601259335650023,
      "learning_rate": 8.547006778304632e-06,
      "loss": 0.3632,
      "step": 6850
    },
    {
      "epoch": 2.96,
      "grad_norm": 2.0777569420111854,
      "learning_rate": 8.532089297607585e-06,
      "loss": 0.3995,
      "step": 6855
    },
    {
      "epoch": 2.96,
      "grad_norm": 2.2903854384703433,
      "learning_rate": 8.517175154661887e-06,
      "loss": 0.4314,
      "step": 6860
    },
    {
      "epoch": 2.97,
      "grad_norm": 1.734580558138607,
      "learning_rate": 8.50226438337948e-06,
      "loss": 0.3829,
      "step": 6865
    },
    {
      "epoch": 2.97,
      "grad_norm": 3.0950107309613637,
      "learning_rate": 8.487357017664637e-06,
      "loss": 0.4173,
      "step": 6870
    },
    {
      "epoch": 2.97,
      "grad_norm": 1.2769399862835231,
      "learning_rate": 8.472453091413895e-06,
      "loss": 0.3269,
      "step": 6875
    },
    {
      "epoch": 2.97,
      "grad_norm": 2.874297218160978,
      "learning_rate": 8.457552638515957e-06,
      "loss": 0.5393,
      "step": 6880
    },
    {
      "epoch": 2.97,
      "grad_norm": 1.0257328536932597,
      "learning_rate": 8.44265569285164e-06,
      "loss": 0.3151,
      "step": 6885
    },
    {
      "epoch": 2.98,
      "grad_norm": 2.3351313764160255,
      "learning_rate": 8.427762288293774e-06,
      "loss": 0.4007,
      "step": 6890
    },
    {
      "epoch": 2.98,
      "grad_norm": 1.5703654283551154,
      "learning_rate": 8.412872458707149e-06,
      "loss": 0.3595,
      "step": 6895
    },
    {
      "epoch": 2.98,
      "grad_norm": 1.8729325717727523,
      "learning_rate": 8.397986237948426e-06,
      "loss": 0.3091,
      "step": 6900
    },
    {
      "epoch": 2.98,
      "grad_norm": 1.6325566546085524,
      "learning_rate": 8.383103659866051e-06,
      "loss": 0.3869,
      "step": 6905
    },
    {
      "epoch": 2.98,
      "grad_norm": 1.1253158005213604,
      "learning_rate": 8.368224758300196e-06,
      "loss": 0.3839,
      "step": 6910
    },
    {
      "epoch": 2.99,
      "grad_norm": 1.7683227163639756,
      "learning_rate": 8.353349567082664e-06,
      "loss": 0.377,
      "step": 6915
    },
    {
      "epoch": 2.99,
      "grad_norm": 2.5880147569316163,
      "learning_rate": 8.338478120036827e-06,
      "loss": 0.4638,
      "step": 6920
    },
    {
      "epoch": 2.99,
      "grad_norm": 2.3260828538915503,
      "learning_rate": 8.32361045097755e-06,
      "loss": 0.5224,
      "step": 6925
    },
    {
      "epoch": 2.99,
      "grad_norm": 1.2921962675871295,
      "learning_rate": 8.308746593711097e-06,
      "loss": 0.3786,
      "step": 6930
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.373713282307955,
      "learning_rate": 8.293886582035068e-06,
      "loss": 0.4048,
      "step": 6935
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.1792303077370565,
      "learning_rate": 8.27903044973832e-06,
      "loss": 0.3728,
      "step": 6940
    },
    {
      "epoch": 3.0,
      "grad_norm": 2.478527290979682,
      "learning_rate": 8.264178230600886e-06,
      "loss": 0.3835,
      "step": 6945
    },
    {
      "epoch": 3.0,
      "eval_loss": 3.1219844818115234,
      "eval_runtime": 352.5163,
      "eval_samples_per_second": 21.295,
      "eval_steps_per_second": 0.335,
      "step": 6945
    },
    {
      "epoch": 3.0,
      "grad_norm": 2.4438409801730288,
      "learning_rate": 8.24932995839391e-06,
      "loss": 0.3604,
      "step": 6950
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.9829599516746514,
      "learning_rate": 8.234485666879552e-06,
      "loss": 0.3694,
      "step": 6955
    },
    {
      "epoch": 3.01,
      "grad_norm": 1.6859068677210045,
      "learning_rate": 8.219645389810923e-06,
      "loss": 0.4267,
      "step": 6960
    },
    {
      "epoch": 3.01,
      "grad_norm": 3.0400939723252676,
      "learning_rate": 8.204809160932004e-06,
      "loss": 0.382,
      "step": 6965
    },
    {
      "epoch": 3.01,
      "grad_norm": 2.365210178859166,
      "learning_rate": 8.189977013977575e-06,
      "loss": 0.404,
      "step": 6970
    },
    {
      "epoch": 3.01,
      "grad_norm": 1.6313288974490237,
      "learning_rate": 8.175148982673141e-06,
      "loss": 0.3676,
      "step": 6975
    },
    {
      "epoch": 3.02,
      "grad_norm": 1.5199651747372227,
      "learning_rate": 8.160325100734835e-06,
      "loss": 0.4012,
      "step": 6980
    },
    {
      "epoch": 3.02,
      "grad_norm": 1.7919489483520143,
      "learning_rate": 8.14550540186936e-06,
      "loss": 0.3834,
      "step": 6985
    },
    {
      "epoch": 3.02,
      "grad_norm": 1.376372130365141,
      "learning_rate": 8.13068991977391e-06,
      "loss": 0.3562,
      "step": 6990
    },
    {
      "epoch": 3.02,
      "grad_norm": 1.0713239095717266,
      "learning_rate": 8.115878688136088e-06,
      "loss": 0.3052,
      "step": 6995
    },
    {
      "epoch": 3.02,
      "grad_norm": 1.3659799887935307,
      "learning_rate": 8.10107174063384e-06,
      "loss": 0.3978,
      "step": 7000
    },
    {
      "epoch": 3.03,
      "grad_norm": 1.4462952426077653,
      "learning_rate": 8.08626911093536e-06,
      "loss": 0.3724,
      "step": 7005
    },
    {
      "epoch": 3.03,
      "grad_norm": 1.4157895388816184,
      "learning_rate": 8.07147083269903e-06,
      "loss": 0.3455,
      "step": 7010
    },
    {
      "epoch": 3.03,
      "grad_norm": 0.8782933167988847,
      "learning_rate": 8.056676939573334e-06,
      "loss": 0.3594,
      "step": 7015
    },
    {
      "epoch": 3.03,
      "grad_norm": 1.6277189620907764,
      "learning_rate": 8.041887465196788e-06,
      "loss": 0.4029,
      "step": 7020
    },
    {
      "epoch": 3.03,
      "grad_norm": 2.6546278749619274,
      "learning_rate": 8.02710244319786e-06,
      "loss": 0.4223,
      "step": 7025
    },
    {
      "epoch": 3.04,
      "grad_norm": 1.211076145212302,
      "learning_rate": 8.0123219071949e-06,
      "loss": 0.3598,
      "step": 7030
    },
    {
      "epoch": 3.04,
      "grad_norm": 1.343844774689017,
      "learning_rate": 7.997545890796045e-06,
      "loss": 0.3523,
      "step": 7035
    },
    {
      "epoch": 3.04,
      "grad_norm": 1.6274284116880298,
      "learning_rate": 7.982774427599162e-06,
      "loss": 0.3976,
      "step": 7040
    },
    {
      "epoch": 3.04,
      "grad_norm": 2.8144432271137214,
      "learning_rate": 7.968007551191767e-06,
      "loss": 0.3435,
      "step": 7045
    },
    {
      "epoch": 3.05,
      "grad_norm": 3.2774547224149004,
      "learning_rate": 7.953245295150945e-06,
      "loss": 0.4223,
      "step": 7050
    },
    {
      "epoch": 3.05,
      "grad_norm": 1.7535061060351183,
      "learning_rate": 7.938487693043277e-06,
      "loss": 0.3808,
      "step": 7055
    },
    {
      "epoch": 3.05,
      "grad_norm": 2.185881923891815,
      "learning_rate": 7.923734778424755e-06,
      "loss": 0.3872,
      "step": 7060
    },
    {
      "epoch": 3.05,
      "grad_norm": 2.5655611819438864,
      "learning_rate": 7.908986584840719e-06,
      "loss": 0.3469,
      "step": 7065
    },
    {
      "epoch": 3.05,
      "grad_norm": 2.0458124980828414,
      "learning_rate": 7.894243145825769e-06,
      "loss": 0.3517,
      "step": 7070
    },
    {
      "epoch": 3.06,
      "grad_norm": 2.09865517921951,
      "learning_rate": 7.879504494903703e-06,
      "loss": 0.3747,
      "step": 7075
    },
    {
      "epoch": 3.06,
      "grad_norm": 1.836960027260886,
      "learning_rate": 7.864770665587424e-06,
      "loss": 0.3784,
      "step": 7080
    },
    {
      "epoch": 3.06,
      "grad_norm": 2.161059273404664,
      "learning_rate": 7.850041691378873e-06,
      "loss": 0.3529,
      "step": 7085
    },
    {
      "epoch": 3.06,
      "grad_norm": 2.36114844324576,
      "learning_rate": 7.835317605768949e-06,
      "loss": 0.3875,
      "step": 7090
    },
    {
      "epoch": 3.06,
      "grad_norm": 1.7898708810429749,
      "learning_rate": 7.820598442237444e-06,
      "loss": 0.3854,
      "step": 7095
    },
    {
      "epoch": 3.07,
      "grad_norm": 1.9649168757553455,
      "learning_rate": 7.805884234252947e-06,
      "loss": 0.3459,
      "step": 7100
    },
    {
      "epoch": 3.07,
      "grad_norm": 1.9301743683275345,
      "learning_rate": 7.791175015272791e-06,
      "loss": 0.3507,
      "step": 7105
    },
    {
      "epoch": 3.07,
      "grad_norm": 1.3558476831220407,
      "learning_rate": 7.776470818742954e-06,
      "loss": 0.3579,
      "step": 7110
    },
    {
      "epoch": 3.07,
      "grad_norm": 1.5222520880851103,
      "learning_rate": 7.761771678097997e-06,
      "loss": 0.3348,
      "step": 7115
    },
    {
      "epoch": 3.08,
      "grad_norm": 2.3225581037412923,
      "learning_rate": 7.747077626760989e-06,
      "loss": 0.365,
      "step": 7120
    },
    {
      "epoch": 3.08,
      "grad_norm": 1.4854703386339154,
      "learning_rate": 7.732388698143421e-06,
      "loss": 0.3497,
      "step": 7125
    },
    {
      "epoch": 3.08,
      "grad_norm": 2.4825826951759966,
      "learning_rate": 7.717704925645145e-06,
      "loss": 0.4039,
      "step": 7130
    },
    {
      "epoch": 3.08,
      "grad_norm": 1.1508037345497122,
      "learning_rate": 7.703026342654279e-06,
      "loss": 0.3357,
      "step": 7135
    },
    {
      "epoch": 3.08,
      "grad_norm": 1.401657005259486,
      "learning_rate": 7.68835298254714e-06,
      "loss": 0.346,
      "step": 7140
    },
    {
      "epoch": 3.09,
      "grad_norm": 0.616882217177935,
      "learning_rate": 7.673684878688183e-06,
      "loss": 0.3683,
      "step": 7145
    },
    {
      "epoch": 3.09,
      "grad_norm": 1.4440225390107253,
      "learning_rate": 7.659022064429897e-06,
      "loss": 0.4064,
      "step": 7150
    },
    {
      "epoch": 3.09,
      "grad_norm": 1.3782845348525206,
      "learning_rate": 7.644364573112752e-06,
      "loss": 0.3057,
      "step": 7155
    },
    {
      "epoch": 3.09,
      "grad_norm": 1.1185076699186876,
      "learning_rate": 7.629712438065112e-06,
      "loss": 0.3627,
      "step": 7160
    },
    {
      "epoch": 3.1,
      "grad_norm": 0.9216251938487858,
      "learning_rate": 7.615065692603154e-06,
      "loss": 0.2782,
      "step": 7165
    },
    {
      "epoch": 3.1,
      "grad_norm": 1.0953121936604426,
      "learning_rate": 7.6004243700308136e-06,
      "loss": 0.3553,
      "step": 7170
    },
    {
      "epoch": 3.1,
      "grad_norm": 1.1345399633196933,
      "learning_rate": 7.585788503639688e-06,
      "loss": 0.3723,
      "step": 7175
    },
    {
      "epoch": 3.1,
      "grad_norm": 1.3955329601427968,
      "learning_rate": 7.571158126708974e-06,
      "loss": 0.3782,
      "step": 7180
    },
    {
      "epoch": 3.1,
      "grad_norm": 1.9115571763477264,
      "learning_rate": 7.556533272505377e-06,
      "loss": 0.3813,
      "step": 7185
    },
    {
      "epoch": 3.11,
      "grad_norm": 1.7498110352468872,
      "learning_rate": 7.5419139742830494e-06,
      "loss": 0.3144,
      "step": 7190
    },
    {
      "epoch": 3.11,
      "grad_norm": 1.0163021017255334,
      "learning_rate": 7.5273002652835124e-06,
      "loss": 0.3512,
      "step": 7195
    },
    {
      "epoch": 3.11,
      "grad_norm": 1.536549393457774,
      "learning_rate": 7.512692178735578e-06,
      "loss": 0.3413,
      "step": 7200
    },
    {
      "epoch": 3.11,
      "grad_norm": 1.8775277627792926,
      "learning_rate": 7.4980897478552725e-06,
      "loss": 0.406,
      "step": 7205
    },
    {
      "epoch": 3.11,
      "grad_norm": 1.9644401322054297,
      "learning_rate": 7.483493005845762e-06,
      "loss": 0.4718,
      "step": 7210
    },
    {
      "epoch": 3.12,
      "grad_norm": 2.625406118579211,
      "learning_rate": 7.468901985897277e-06,
      "loss": 0.344,
      "step": 7215
    },
    {
      "epoch": 3.12,
      "grad_norm": 2.163595461889686,
      "learning_rate": 7.454316721187034e-06,
      "loss": 0.3996,
      "step": 7220
    },
    {
      "epoch": 3.12,
      "grad_norm": 2.228158971734925,
      "learning_rate": 7.43973724487917e-06,
      "loss": 0.3664,
      "step": 7225
    },
    {
      "epoch": 3.12,
      "grad_norm": 2.524343595169013,
      "learning_rate": 7.425163590124662e-06,
      "loss": 0.3957,
      "step": 7230
    },
    {
      "epoch": 3.13,
      "grad_norm": 0.675888461514967,
      "learning_rate": 7.4105957900612405e-06,
      "loss": 0.3292,
      "step": 7235
    },
    {
      "epoch": 3.13,
      "grad_norm": 1.6605683075642659,
      "learning_rate": 7.396033877813327e-06,
      "loss": 0.3192,
      "step": 7240
    },
    {
      "epoch": 3.13,
      "grad_norm": 1.569991195286367,
      "learning_rate": 7.381477886491953e-06,
      "loss": 0.4003,
      "step": 7245
    },
    {
      "epoch": 3.13,
      "grad_norm": 2.632148940238971,
      "learning_rate": 7.366927849194698e-06,
      "loss": 0.3839,
      "step": 7250
    },
    {
      "epoch": 3.13,
      "grad_norm": 1.352149468997016,
      "learning_rate": 7.352383799005593e-06,
      "loss": 0.3815,
      "step": 7255
    },
    {
      "epoch": 3.14,
      "grad_norm": 1.662409127769855,
      "learning_rate": 7.3378457689950565e-06,
      "loss": 0.3485,
      "step": 7260
    },
    {
      "epoch": 3.14,
      "grad_norm": 1.7956039040207492,
      "learning_rate": 7.323313792219819e-06,
      "loss": 0.367,
      "step": 7265
    },
    {
      "epoch": 3.14,
      "grad_norm": 1.8540078766409693,
      "learning_rate": 7.308787901722847e-06,
      "loss": 0.3722,
      "step": 7270
    },
    {
      "epoch": 3.14,
      "grad_norm": 0.7085122569624023,
      "learning_rate": 7.294268130533268e-06,
      "loss": 0.3218,
      "step": 7275
    },
    {
      "epoch": 3.14,
      "grad_norm": 2.497501765694081,
      "learning_rate": 7.279754511666304e-06,
      "loss": 0.4098,
      "step": 7280
    },
    {
      "epoch": 3.15,
      "grad_norm": 1.285506504028356,
      "learning_rate": 7.265247078123171e-06,
      "loss": 0.3316,
      "step": 7285
    },
    {
      "epoch": 3.15,
      "grad_norm": 2.0715015673160715,
      "learning_rate": 7.250745862891033e-06,
      "loss": 0.316,
      "step": 7290
    },
    {
      "epoch": 3.15,
      "grad_norm": 1.6351432229152634,
      "learning_rate": 7.236250898942906e-06,
      "loss": 0.342,
      "step": 7295
    },
    {
      "epoch": 3.15,
      "grad_norm": 1.8714186683779126,
      "learning_rate": 7.221762219237603e-06,
      "loss": 0.4127,
      "step": 7300
    },
    {
      "epoch": 3.16,
      "grad_norm": 1.2075723689154334,
      "learning_rate": 7.207279856719642e-06,
      "loss": 0.39,
      "step": 7305
    },
    {
      "epoch": 3.16,
      "grad_norm": 1.842860731387192,
      "learning_rate": 7.192803844319176e-06,
      "loss": 0.3206,
      "step": 7310
    },
    {
      "epoch": 3.16,
      "grad_norm": 0.9966979046850045,
      "learning_rate": 7.178334214951919e-06,
      "loss": 0.3021,
      "step": 7315
    },
    {
      "epoch": 3.16,
      "grad_norm": 1.645332076913864,
      "learning_rate": 7.163871001519065e-06,
      "loss": 0.3523,
      "step": 7320
    },
    {
      "epoch": 3.16,
      "grad_norm": 2.5864286290966634,
      "learning_rate": 7.149414236907238e-06,
      "loss": 0.4284,
      "step": 7325
    },
    {
      "epoch": 3.17,
      "grad_norm": 2.1920020064756747,
      "learning_rate": 7.134963953988381e-06,
      "loss": 0.3299,
      "step": 7330
    },
    {
      "epoch": 3.17,
      "grad_norm": 0.7475437303317399,
      "learning_rate": 7.120520185619706e-06,
      "loss": 0.3112,
      "step": 7335
    },
    {
      "epoch": 3.17,
      "grad_norm": 2.14902808663687,
      "learning_rate": 7.1060829646436104e-06,
      "loss": 0.3632,
      "step": 7340
    },
    {
      "epoch": 3.17,
      "grad_norm": 1.5457835835946556,
      "learning_rate": 7.091652323887601e-06,
      "loss": 0.3356,
      "step": 7345
    },
    {
      "epoch": 3.17,
      "grad_norm": 1.5894112066172816,
      "learning_rate": 7.077228296164231e-06,
      "loss": 0.357,
      "step": 7350
    },
    {
      "epoch": 3.18,
      "grad_norm": 1.2040205720783954,
      "learning_rate": 7.062810914271012e-06,
      "loss": 0.3663,
      "step": 7355
    },
    {
      "epoch": 3.18,
      "grad_norm": 1.33985630492986,
      "learning_rate": 7.048400210990344e-06,
      "loss": 0.3366,
      "step": 7360
    },
    {
      "epoch": 3.18,
      "grad_norm": 1.127291173301345,
      "learning_rate": 7.033996219089438e-06,
      "loss": 0.3272,
      "step": 7365
    },
    {
      "epoch": 3.18,
      "grad_norm": 1.7185161121647898,
      "learning_rate": 7.0195989713202465e-06,
      "loss": 0.3234,
      "step": 7370
    },
    {
      "epoch": 3.19,
      "grad_norm": 1.1498105845305382,
      "learning_rate": 7.005208500419393e-06,
      "loss": 0.2983,
      "step": 7375
    },
    {
      "epoch": 3.19,
      "grad_norm": 2.2371882954147235,
      "learning_rate": 6.990824839108086e-06,
      "loss": 0.3753,
      "step": 7380
    },
    {
      "epoch": 3.19,
      "grad_norm": 0.7884854218345562,
      "learning_rate": 6.976448020092051e-06,
      "loss": 0.3763,
      "step": 7385
    },
    {
      "epoch": 3.19,
      "grad_norm": 2.09921574448507,
      "learning_rate": 6.962078076061456e-06,
      "loss": 0.3753,
      "step": 7390
    },
    {
      "epoch": 3.19,
      "grad_norm": 2.482230635386038,
      "learning_rate": 6.947715039690831e-06,
      "loss": 0.4298,
      "step": 7395
    },
    {
      "epoch": 3.2,
      "grad_norm": 2.108335121499453,
      "learning_rate": 6.933358943639014e-06,
      "loss": 0.3505,
      "step": 7400
    },
    {
      "epoch": 3.2,
      "grad_norm": 1.5449594462702159,
      "learning_rate": 6.919009820549045e-06,
      "loss": 0.394,
      "step": 7405
    },
    {
      "epoch": 3.2,
      "grad_norm": 1.2438744387949128,
      "learning_rate": 6.904667703048121e-06,
      "loss": 0.3462,
      "step": 7410
    },
    {
      "epoch": 3.2,
      "grad_norm": 1.289234931861233,
      "learning_rate": 6.890332623747504e-06,
      "loss": 0.3629,
      "step": 7415
    },
    {
      "epoch": 3.21,
      "grad_norm": 0.8977406690803236,
      "learning_rate": 6.876004615242449e-06,
      "loss": 0.3823,
      "step": 7420
    },
    {
      "epoch": 3.21,
      "grad_norm": 2.1847093235275623,
      "learning_rate": 6.861683710112146e-06,
      "loss": 0.397,
      "step": 7425
    },
    {
      "epoch": 3.21,
      "grad_norm": 3.0155479918164647,
      "learning_rate": 6.847369940919618e-06,
      "loss": 0.3929,
      "step": 7430
    },
    {
      "epoch": 3.21,
      "grad_norm": 4.104350300963877,
      "learning_rate": 6.833063340211674e-06,
      "loss": 0.4007,
      "step": 7435
    },
    {
      "epoch": 3.21,
      "grad_norm": 1.8049485982992373,
      "learning_rate": 6.818763940518818e-06,
      "loss": 0.354,
      "step": 7440
    },
    {
      "epoch": 3.22,
      "grad_norm": 1.2427853854370055,
      "learning_rate": 6.804471774355175e-06,
      "loss": 0.3642,
      "step": 7445
    },
    {
      "epoch": 3.22,
      "grad_norm": 2.331588538403581,
      "learning_rate": 6.790186874218437e-06,
      "loss": 0.3201,
      "step": 7450
    },
    {
      "epoch": 3.22,
      "grad_norm": 1.7805164540285714,
      "learning_rate": 6.77590927258976e-06,
      "loss": 0.3839,
      "step": 7455
    },
    {
      "epoch": 3.22,
      "grad_norm": 1.341701157263398,
      "learning_rate": 6.761639001933713e-06,
      "loss": 0.3224,
      "step": 7460
    },
    {
      "epoch": 3.22,
      "grad_norm": 1.2404201279129343,
      "learning_rate": 6.747376094698191e-06,
      "loss": 0.39,
      "step": 7465
    },
    {
      "epoch": 3.23,
      "grad_norm": 2.5663217984821713,
      "learning_rate": 6.7331205833143456e-06,
      "loss": 0.3605,
      "step": 7470
    },
    {
      "epoch": 3.23,
      "grad_norm": 1.2525035749489488,
      "learning_rate": 6.718872500196521e-06,
      "loss": 0.3395,
      "step": 7475
    },
    {
      "epoch": 3.23,
      "grad_norm": 1.5167515096191115,
      "learning_rate": 6.704631877742156e-06,
      "loss": 0.3075,
      "step": 7480
    },
    {
      "epoch": 3.23,
      "grad_norm": 1.662607642634134,
      "learning_rate": 6.690398748331739e-06,
      "loss": 0.3361,
      "step": 7485
    },
    {
      "epoch": 3.24,
      "grad_norm": 1.0759956437678158,
      "learning_rate": 6.676173144328713e-06,
      "loss": 0.3006,
      "step": 7490
    },
    {
      "epoch": 3.24,
      "grad_norm": 1.045165095111159,
      "learning_rate": 6.661955098079407e-06,
      "loss": 0.3062,
      "step": 7495
    },
    {
      "epoch": 3.24,
      "grad_norm": 1.028477080146992,
      "learning_rate": 6.6477446419129765e-06,
      "loss": 0.2854,
      "step": 7500
    },
    {
      "epoch": 3.24,
      "grad_norm": 1.839431683459406,
      "learning_rate": 6.633541808141307e-06,
      "loss": 0.2772,
      "step": 7505
    },
    {
      "epoch": 3.24,
      "grad_norm": 1.2499750070114224,
      "learning_rate": 6.61934662905896e-06,
      "loss": 0.3068,
      "step": 7510
    },
    {
      "epoch": 3.25,
      "grad_norm": 1.4318397803172023,
      "learning_rate": 6.605159136943086e-06,
      "loss": 0.3998,
      "step": 7515
    },
    {
      "epoch": 3.25,
      "grad_norm": 1.6896153505949632,
      "learning_rate": 6.590979364053358e-06,
      "loss": 0.336,
      "step": 7520
    },
    {
      "epoch": 3.25,
      "grad_norm": 1.8291472755252598,
      "learning_rate": 6.576807342631901e-06,
      "loss": 0.3195,
      "step": 7525
    },
    {
      "epoch": 3.25,
      "grad_norm": 1.086754478155717,
      "learning_rate": 6.56264310490321e-06,
      "loss": 0.3124,
      "step": 7530
    },
    {
      "epoch": 3.25,
      "grad_norm": 0.8549620256435886,
      "learning_rate": 6.548486683074086e-06,
      "loss": 0.3528,
      "step": 7535
    },
    {
      "epoch": 3.26,
      "grad_norm": 0.858456604542488,
      "learning_rate": 6.534338109333551e-06,
      "loss": 0.3175,
      "step": 7540
    },
    {
      "epoch": 3.26,
      "grad_norm": 1.870253816789839,
      "learning_rate": 6.5201974158527855e-06,
      "loss": 0.3451,
      "step": 7545
    },
    {
      "epoch": 3.26,
      "grad_norm": 1.7701261568886084,
      "learning_rate": 6.506064634785059e-06,
      "loss": 0.3468,
      "step": 7550
    },
    {
      "epoch": 3.26,
      "grad_norm": 1.3093267757249363,
      "learning_rate": 6.4919397982656365e-06,
      "loss": 0.3102,
      "step": 7555
    },
    {
      "epoch": 3.27,
      "grad_norm": 1.4187325171777991,
      "learning_rate": 6.47782293841173e-06,
      "loss": 0.3231,
      "step": 7560
    },
    {
      "epoch": 3.27,
      "grad_norm": 1.3886715755141823,
      "learning_rate": 6.463714087322406e-06,
      "loss": 0.3683,
      "step": 7565
    },
    {
      "epoch": 3.27,
      "grad_norm": 2.306803637558559,
      "learning_rate": 6.4496132770785215e-06,
      "loss": 0.4046,
      "step": 7570
    },
    {
      "epoch": 3.27,
      "grad_norm": 1.4954651335990372,
      "learning_rate": 6.43552053974266e-06,
      "loss": 0.2898,
      "step": 7575
    },
    {
      "epoch": 3.27,
      "grad_norm": 1.1220262633443565,
      "learning_rate": 6.421435907359035e-06,
      "loss": 0.3033,
      "step": 7580
    },
    {
      "epoch": 3.28,
      "grad_norm": 1.8256112203819537,
      "learning_rate": 6.407359411953441e-06,
      "loss": 0.2982,
      "step": 7585
    },
    {
      "epoch": 3.28,
      "grad_norm": 1.7431511555290342,
      "learning_rate": 6.393291085533166e-06,
      "loss": 0.2926,
      "step": 7590
    },
    {
      "epoch": 3.28,
      "grad_norm": 0.9850643228063368,
      "learning_rate": 6.379230960086916e-06,
      "loss": 0.2945,
      "step": 7595
    },
    {
      "epoch": 3.28,
      "grad_norm": 1.7655045331686694,
      "learning_rate": 6.36517906758477e-06,
      "loss": 0.4019,
      "step": 7600
    },
    {
      "epoch": 3.29,
      "grad_norm": 1.3966168616487404,
      "learning_rate": 6.351135439978067e-06,
      "loss": 0.2689,
      "step": 7605
    },
    {
      "epoch": 3.29,
      "grad_norm": 0.9472185721991497,
      "learning_rate": 6.337100109199362e-06,
      "loss": 0.312,
      "step": 7610
    },
    {
      "epoch": 3.29,
      "grad_norm": 1.7266621380307388,
      "learning_rate": 6.3230731071623445e-06,
      "loss": 0.3638,
      "step": 7615
    },
    {
      "epoch": 3.29,
      "grad_norm": 2.091724927308154,
      "learning_rate": 6.309054465761758e-06,
      "loss": 0.2872,
      "step": 7620
    },
    {
      "epoch": 3.29,
      "grad_norm": 1.073271495200662,
      "learning_rate": 6.295044216873351e-06,
      "loss": 0.3007,
      "step": 7625
    },
    {
      "epoch": 3.3,
      "grad_norm": 1.117360265840462,
      "learning_rate": 6.281042392353775e-06,
      "loss": 0.3308,
      "step": 7630
    },
    {
      "epoch": 3.3,
      "grad_norm": 1.5440549578000846,
      "learning_rate": 6.267049024040535e-06,
      "loss": 0.2729,
      "step": 7635
    },
    {
      "epoch": 3.3,
      "grad_norm": 1.5481051987579446,
      "learning_rate": 6.253064143751901e-06,
      "loss": 0.3386,
      "step": 7640
    },
    {
      "epoch": 3.3,
      "grad_norm": 3.786049943590816,
      "learning_rate": 6.239087783286842e-06,
      "loss": 0.3006,
      "step": 7645
    },
    {
      "epoch": 3.3,
      "grad_norm": 1.7723356489577555,
      "learning_rate": 6.225119974424969e-06,
      "loss": 0.3207,
      "step": 7650
    },
    {
      "epoch": 3.31,
      "grad_norm": 1.319330771697122,
      "learning_rate": 6.211160748926431e-06,
      "loss": 0.3429,
      "step": 7655
    },
    {
      "epoch": 3.31,
      "grad_norm": 0.9383400641879289,
      "learning_rate": 6.19721013853187e-06,
      "loss": 0.285,
      "step": 7660
    },
    {
      "epoch": 3.31,
      "grad_norm": 1.5060664851409171,
      "learning_rate": 6.183268174962333e-06,
      "loss": 0.3419,
      "step": 7665
    },
    {
      "epoch": 3.31,
      "grad_norm": 1.2514169013060776,
      "learning_rate": 6.16933488991921e-06,
      "loss": 0.3034,
      "step": 7670
    },
    {
      "epoch": 3.32,
      "grad_norm": 1.3352849856980331,
      "learning_rate": 6.1554103150841585e-06,
      "loss": 0.3616,
      "step": 7675
    },
    {
      "epoch": 3.32,
      "grad_norm": 2.519664197804798,
      "learning_rate": 6.141494482119024e-06,
      "loss": 0.2999,
      "step": 7680
    },
    {
      "epoch": 3.32,
      "grad_norm": 1.645506931854804,
      "learning_rate": 6.127587422665785e-06,
      "loss": 0.2752,
      "step": 7685
    },
    {
      "epoch": 3.32,
      "grad_norm": 1.26311160539326,
      "learning_rate": 6.113689168346462e-06,
      "loss": 0.2893,
      "step": 7690
    },
    {
      "epoch": 3.32,
      "grad_norm": 1.1737096926315287,
      "learning_rate": 6.099799750763051e-06,
      "loss": 0.3313,
      "step": 7695
    },
    {
      "epoch": 3.33,
      "grad_norm": 1.3998936159924413,
      "learning_rate": 6.085919201497472e-06,
      "loss": 0.3007,
      "step": 7700
    },
    {
      "epoch": 3.33,
      "grad_norm": 1.4549429673890482,
      "learning_rate": 6.072047552111462e-06,
      "loss": 0.2869,
      "step": 7705
    },
    {
      "epoch": 3.33,
      "grad_norm": 1.746910600245538,
      "learning_rate": 6.058184834146532e-06,
      "loss": 0.3488,
      "step": 7710
    },
    {
      "epoch": 3.33,
      "grad_norm": 1.0407842846096143,
      "learning_rate": 6.044331079123882e-06,
      "loss": 0.3,
      "step": 7715
    },
    {
      "epoch": 3.33,
      "grad_norm": 1.6015467888098907,
      "learning_rate": 6.030486318544322e-06,
      "loss": 0.3575,
      "step": 7720
    },
    {
      "epoch": 3.34,
      "grad_norm": 1.4990017412694239,
      "learning_rate": 6.016650583888231e-06,
      "loss": 0.3011,
      "step": 7725
    },
    {
      "epoch": 3.34,
      "grad_norm": 1.0912555050972461,
      "learning_rate": 6.0028239066154495e-06,
      "loss": 0.2803,
      "step": 7730
    },
    {
      "epoch": 3.34,
      "grad_norm": 1.8084455742572358,
      "learning_rate": 5.989006318165224e-06,
      "loss": 0.301,
      "step": 7735
    },
    {
      "epoch": 3.34,
      "grad_norm": 0.5940513087308738,
      "learning_rate": 5.97519784995614e-06,
      "loss": 0.3093,
      "step": 7740
    },
    {
      "epoch": 3.35,
      "grad_norm": 1.8390181703969186,
      "learning_rate": 5.96139853338604e-06,
      "loss": 0.3046,
      "step": 7745
    },
    {
      "epoch": 3.35,
      "grad_norm": 1.2236133534368583,
      "learning_rate": 5.947608399831968e-06,
      "loss": 0.3721,
      "step": 7750
    },
    {
      "epoch": 3.35,
      "grad_norm": 1.3411315105622825,
      "learning_rate": 5.933827480650073e-06,
      "loss": 0.3068,
      "step": 7755
    },
    {
      "epoch": 3.35,
      "grad_norm": 1.683704960625591,
      "learning_rate": 5.92005580717556e-06,
      "loss": 0.4067,
      "step": 7760
    },
    {
      "epoch": 3.35,
      "grad_norm": 1.4060882167908326,
      "learning_rate": 5.906293410722613e-06,
      "loss": 0.3623,
      "step": 7765
    },
    {
      "epoch": 3.36,
      "grad_norm": 1.727295083424792,
      "learning_rate": 5.892540322584315e-06,
      "loss": 0.3564,
      "step": 7770
    },
    {
      "epoch": 3.36,
      "grad_norm": 0.8805570616777468,
      "learning_rate": 5.878796574032592e-06,
      "loss": 0.3287,
      "step": 7775
    },
    {
      "epoch": 3.36,
      "grad_norm": 0.7273911720571473,
      "learning_rate": 5.865062196318128e-06,
      "loss": 0.273,
      "step": 7780
    },
    {
      "epoch": 3.36,
      "grad_norm": 1.3971767306685285,
      "learning_rate": 5.851337220670299e-06,
      "loss": 0.3476,
      "step": 7785
    },
    {
      "epoch": 3.37,
      "grad_norm": 1.5662729786773861,
      "learning_rate": 5.837621678297104e-06,
      "loss": 0.3471,
      "step": 7790
    },
    {
      "epoch": 3.37,
      "grad_norm": 1.6410213733545815,
      "learning_rate": 5.823915600385095e-06,
      "loss": 0.3553,
      "step": 7795
    },
    {
      "epoch": 3.37,
      "grad_norm": 1.4199761399560424,
      "learning_rate": 5.810219018099299e-06,
      "loss": 0.378,
      "step": 7800
    },
    {
      "epoch": 3.37,
      "grad_norm": 1.6803309581824162,
      "learning_rate": 5.796531962583161e-06,
      "loss": 0.2952,
      "step": 7805
    },
    {
      "epoch": 3.37,
      "grad_norm": 0.8482833840019645,
      "learning_rate": 5.782854464958445e-06,
      "loss": 0.322,
      "step": 7810
    },
    {
      "epoch": 3.38,
      "grad_norm": 2.011908070466235,
      "learning_rate": 5.769186556325206e-06,
      "loss": 0.2833,
      "step": 7815
    },
    {
      "epoch": 3.38,
      "grad_norm": 0.992761062940804,
      "learning_rate": 5.755528267761669e-06,
      "loss": 0.3253,
      "step": 7820
    },
    {
      "epoch": 3.38,
      "grad_norm": 1.295611381712756,
      "learning_rate": 5.741879630324214e-06,
      "loss": 0.2747,
      "step": 7825
    },
    {
      "epoch": 3.38,
      "grad_norm": 1.1938601977260708,
      "learning_rate": 5.728240675047259e-06,
      "loss": 0.3383,
      "step": 7830
    },
    {
      "epoch": 3.38,
      "grad_norm": 0.8344509260934919,
      "learning_rate": 5.714611432943202e-06,
      "loss": 0.2881,
      "step": 7835
    },
    {
      "epoch": 3.39,
      "grad_norm": 1.3994896482762227,
      "learning_rate": 5.700991935002371e-06,
      "loss": 0.3118,
      "step": 7840
    },
    {
      "epoch": 3.39,
      "grad_norm": 0.7711648002211058,
      "learning_rate": 5.687382212192915e-06,
      "loss": 0.3013,
      "step": 7845
    },
    {
      "epoch": 3.39,
      "grad_norm": 1.0473806175451943,
      "learning_rate": 5.673782295460784e-06,
      "loss": 0.2497,
      "step": 7850
    },
    {
      "epoch": 3.39,
      "grad_norm": 1.4164383045109232,
      "learning_rate": 5.660192215729615e-06,
      "loss": 0.3133,
      "step": 7855
    },
    {
      "epoch": 3.4,
      "grad_norm": 0.9574061018200604,
      "learning_rate": 5.646612003900674e-06,
      "loss": 0.3171,
      "step": 7860
    },
    {
      "epoch": 3.4,
      "grad_norm": 1.708395064070684,
      "learning_rate": 5.633041690852798e-06,
      "loss": 0.3316,
      "step": 7865
    },
    {
      "epoch": 3.4,
      "grad_norm": 1.2330174450753875,
      "learning_rate": 5.619481307442306e-06,
      "loss": 0.3241,
      "step": 7870
    },
    {
      "epoch": 3.4,
      "grad_norm": 1.2419549464618695,
      "learning_rate": 5.605930884502954e-06,
      "loss": 0.3131,
      "step": 7875
    },
    {
      "epoch": 3.4,
      "grad_norm": 1.0772138558401552,
      "learning_rate": 5.592390452845844e-06,
      "loss": 0.3444,
      "step": 7880
    },
    {
      "epoch": 3.41,
      "grad_norm": 1.4128274027698726,
      "learning_rate": 5.578860043259346e-06,
      "loss": 0.2587,
      "step": 7885
    },
    {
      "epoch": 3.41,
      "grad_norm": 0.9309254408786461,
      "learning_rate": 5.5653396865090635e-06,
      "loss": 0.3206,
      "step": 7890
    },
    {
      "epoch": 3.41,
      "grad_norm": 1.233915081617126,
      "learning_rate": 5.5518294133377195e-06,
      "loss": 0.2833,
      "step": 7895
    },
    {
      "epoch": 3.41,
      "grad_norm": 1.0971867385885703,
      "learning_rate": 5.538329254465128e-06,
      "loss": 0.2947,
      "step": 7900
    },
    {
      "epoch": 3.41,
      "grad_norm": 2.4423525002869053,
      "learning_rate": 5.5248392405881e-06,
      "loss": 0.386,
      "step": 7905
    },
    {
      "epoch": 3.42,
      "grad_norm": 1.6676084708192005,
      "learning_rate": 5.51135940238037e-06,
      "loss": 0.2742,
      "step": 7910
    },
    {
      "epoch": 3.42,
      "grad_norm": 0.9193407813258758,
      "learning_rate": 5.497889770492546e-06,
      "loss": 0.2955,
      "step": 7915
    },
    {
      "epoch": 3.42,
      "grad_norm": 1.6840232895053835,
      "learning_rate": 5.484430375552015e-06,
      "loss": 0.2896,
      "step": 7920
    },
    {
      "epoch": 3.42,
      "grad_norm": 2.089338229394022,
      "learning_rate": 5.4709812481629035e-06,
      "loss": 0.3333,
      "step": 7925
    },
    {
      "epoch": 3.43,
      "grad_norm": 1.2417429731256493,
      "learning_rate": 5.457542418905986e-06,
      "loss": 0.3443,
      "step": 7930
    },
    {
      "epoch": 3.43,
      "grad_norm": 1.187138692055809,
      "learning_rate": 5.4441139183386116e-06,
      "loss": 0.3059,
      "step": 7935
    },
    {
      "epoch": 3.43,
      "grad_norm": 1.380993547712161,
      "learning_rate": 5.430695776994659e-06,
      "loss": 0.35,
      "step": 7940
    },
    {
      "epoch": 3.43,
      "grad_norm": 1.1629205213157978,
      "learning_rate": 5.4172880253844325e-06,
      "loss": 0.3202,
      "step": 7945
    },
    {
      "epoch": 3.43,
      "grad_norm": 1.4779553786860646,
      "learning_rate": 5.403890693994635e-06,
      "loss": 0.3535,
      "step": 7950
    },
    {
      "epoch": 3.44,
      "grad_norm": 1.0785096047218297,
      "learning_rate": 5.3905038132882644e-06,
      "loss": 0.2486,
      "step": 7955
    },
    {
      "epoch": 3.44,
      "grad_norm": 1.6003168513345345,
      "learning_rate": 5.377127413704548e-06,
      "loss": 0.3267,
      "step": 7960
    },
    {
      "epoch": 3.44,
      "grad_norm": 0.8661954065325067,
      "learning_rate": 5.3637615256589e-06,
      "loss": 0.2747,
      "step": 7965
    },
    {
      "epoch": 3.44,
      "grad_norm": 1.0632816286685356,
      "learning_rate": 5.350406179542804e-06,
      "loss": 0.2608,
      "step": 7970
    },
    {
      "epoch": 3.44,
      "grad_norm": 1.4309598280517628,
      "learning_rate": 5.337061405723813e-06,
      "loss": 0.2653,
      "step": 7975
    },
    {
      "epoch": 3.45,
      "grad_norm": 1.4823225940472557,
      "learning_rate": 5.323727234545406e-06,
      "loss": 0.3331,
      "step": 7980
    },
    {
      "epoch": 3.45,
      "grad_norm": 1.1371470643568795,
      "learning_rate": 5.310403696326969e-06,
      "loss": 0.3151,
      "step": 7985
    },
    {
      "epoch": 3.45,
      "grad_norm": 1.0474163097333449,
      "learning_rate": 5.297090821363713e-06,
      "loss": 0.2394,
      "step": 7990
    },
    {
      "epoch": 3.45,
      "grad_norm": 2.0260247807879663,
      "learning_rate": 5.2837886399265855e-06,
      "loss": 0.3131,
      "step": 7995
    },
    {
      "epoch": 3.46,
      "grad_norm": 1.257262975782416,
      "learning_rate": 5.270497182262251e-06,
      "loss": 0.2739,
      "step": 8000
    },
    {
      "epoch": 3.46,
      "grad_norm": 1.3679796310400674,
      "learning_rate": 5.257216478592956e-06,
      "loss": 0.2799,
      "step": 8005
    },
    {
      "epoch": 3.46,
      "grad_norm": 1.2880772135115273,
      "learning_rate": 5.243946559116513e-06,
      "loss": 0.2591,
      "step": 8010
    },
    {
      "epoch": 3.46,
      "grad_norm": 1.3722257951161014,
      "learning_rate": 5.230687454006213e-06,
      "loss": 0.2561,
      "step": 8015
    },
    {
      "epoch": 3.46,
      "grad_norm": 1.4788204117150598,
      "learning_rate": 5.2174391934107445e-06,
      "loss": 0.3161,
      "step": 8020
    },
    {
      "epoch": 3.47,
      "grad_norm": 1.1950353284608026,
      "learning_rate": 5.204201807454161e-06,
      "loss": 0.2837,
      "step": 8025
    },
    {
      "epoch": 3.47,
      "grad_norm": 1.47777539000232,
      "learning_rate": 5.190975326235764e-06,
      "loss": 0.3246,
      "step": 8030
    },
    {
      "epoch": 3.47,
      "grad_norm": 1.800356278890997,
      "learning_rate": 5.177759779830072e-06,
      "loss": 0.3721,
      "step": 8035
    },
    {
      "epoch": 3.47,
      "grad_norm": 0.879211002173503,
      "learning_rate": 5.164555198286745e-06,
      "loss": 0.2545,
      "step": 8040
    },
    {
      "epoch": 3.48,
      "grad_norm": 1.6605980255623658,
      "learning_rate": 5.151361611630491e-06,
      "loss": 0.3551,
      "step": 8045
    },
    {
      "epoch": 3.48,
      "grad_norm": 0.9258465918526644,
      "learning_rate": 5.138179049861047e-06,
      "loss": 0.2511,
      "step": 8050
    },
    {
      "epoch": 3.48,
      "grad_norm": 1.362803220020721,
      "learning_rate": 5.125007542953056e-06,
      "loss": 0.2996,
      "step": 8055
    },
    {
      "epoch": 3.48,
      "grad_norm": 1.4846626929901958,
      "learning_rate": 5.111847120856035e-06,
      "loss": 0.2861,
      "step": 8060
    },
    {
      "epoch": 3.48,
      "grad_norm": 1.3120726030735523,
      "learning_rate": 5.0986978134943e-06,
      "loss": 0.3216,
      "step": 8065
    },
    {
      "epoch": 3.49,
      "grad_norm": 2.3701644636458017,
      "learning_rate": 5.085559650766878e-06,
      "loss": 0.3636,
      "step": 8070
    },
    {
      "epoch": 3.49,
      "grad_norm": 2.2927283455933685,
      "learning_rate": 5.072432662547483e-06,
      "loss": 0.2699,
      "step": 8075
    },
    {
      "epoch": 3.49,
      "grad_norm": 1.4104584634968191,
      "learning_rate": 5.05931687868439e-06,
      "loss": 0.2846,
      "step": 8080
    },
    {
      "epoch": 3.49,
      "grad_norm": 1.6350220111620895,
      "learning_rate": 5.046212329000417e-06,
      "loss": 0.3417,
      "step": 8085
    },
    {
      "epoch": 3.49,
      "grad_norm": 0.9603072152402977,
      "learning_rate": 5.033119043292835e-06,
      "loss": 0.3143,
      "step": 8090
    },
    {
      "epoch": 3.5,
      "grad_norm": 1.5098697111760186,
      "learning_rate": 5.02003705133329e-06,
      "loss": 0.3101,
      "step": 8095
    },
    {
      "epoch": 3.5,
      "grad_norm": 1.241925543830471,
      "learning_rate": 5.006966382867771e-06,
      "loss": 0.2992,
      "step": 8100
    },
    {
      "epoch": 3.5,
      "grad_norm": 1.5474911953971806,
      "learning_rate": 4.9939070676164935e-06,
      "loss": 0.3282,
      "step": 8105
    },
    {
      "epoch": 3.5,
      "grad_norm": 4.243167744270766,
      "learning_rate": 4.980859135273878e-06,
      "loss": 0.3118,
      "step": 8110
    },
    {
      "epoch": 3.51,
      "grad_norm": 2.5000761058036507,
      "learning_rate": 4.967822615508457e-06,
      "loss": 0.2975,
      "step": 8115
    },
    {
      "epoch": 3.51,
      "grad_norm": 3.5097390106620856,
      "learning_rate": 4.9547975379628e-06,
      "loss": 0.2945,
      "step": 8120
    },
    {
      "epoch": 3.51,
      "grad_norm": 2.209763598381993,
      "learning_rate": 4.941783932253487e-06,
      "loss": 0.3477,
      "step": 8125
    },
    {
      "epoch": 3.51,
      "grad_norm": 1.4044042169605864,
      "learning_rate": 4.928781827970983e-06,
      "loss": 0.2261,
      "step": 8130
    },
    {
      "epoch": 3.51,
      "grad_norm": 1.3885448858555944,
      "learning_rate": 4.9157912546796205e-06,
      "loss": 0.3017,
      "step": 8135
    },
    {
      "epoch": 3.52,
      "grad_norm": 1.2411471192003973,
      "learning_rate": 4.9028122419175075e-06,
      "loss": 0.2904,
      "step": 8140
    },
    {
      "epoch": 3.52,
      "grad_norm": 1.2413220137474514,
      "learning_rate": 4.889844819196455e-06,
      "loss": 0.2782,
      "step": 8145
    },
    {
      "epoch": 3.52,
      "grad_norm": 1.250835967591782,
      "learning_rate": 4.876889016001946e-06,
      "loss": 0.3283,
      "step": 8150
    },
    {
      "epoch": 3.52,
      "grad_norm": 0.9716840825910151,
      "learning_rate": 4.863944861793014e-06,
      "loss": 0.3037,
      "step": 8155
    },
    {
      "epoch": 3.52,
      "grad_norm": 2.260877892021157,
      "learning_rate": 4.851012386002223e-06,
      "loss": 0.3047,
      "step": 8160
    },
    {
      "epoch": 3.53,
      "grad_norm": 1.681828011210631,
      "learning_rate": 4.83809161803558e-06,
      "loss": 0.2672,
      "step": 8165
    },
    {
      "epoch": 3.53,
      "grad_norm": 1.495162454290298,
      "learning_rate": 4.825182587272458e-06,
      "loss": 0.2552,
      "step": 8170
    },
    {
      "epoch": 3.53,
      "grad_norm": 1.292898157049884,
      "learning_rate": 4.812285323065565e-06,
      "loss": 0.2828,
      "step": 8175
    },
    {
      "epoch": 3.53,
      "grad_norm": 0.952818561352293,
      "learning_rate": 4.799399854740832e-06,
      "loss": 0.2739,
      "step": 8180
    },
    {
      "epoch": 3.54,
      "grad_norm": 0.723324136997605,
      "learning_rate": 4.786526211597378e-06,
      "loss": 0.2376,
      "step": 8185
    },
    {
      "epoch": 3.54,
      "grad_norm": 1.2593200948705463,
      "learning_rate": 4.7736644229074396e-06,
      "loss": 0.2956,
      "step": 8190
    },
    {
      "epoch": 3.54,
      "grad_norm": 1.3296718387734723,
      "learning_rate": 4.760814517916279e-06,
      "loss": 0.2842,
      "step": 8195
    },
    {
      "epoch": 3.54,
      "grad_norm": 0.8986198037137462,
      "learning_rate": 4.747976525842167e-06,
      "loss": 0.2481,
      "step": 8200
    },
    {
      "epoch": 3.54,
      "grad_norm": 1.7112417432680016,
      "learning_rate": 4.7351504758762565e-06,
      "loss": 0.301,
      "step": 8205
    },
    {
      "epoch": 3.55,
      "grad_norm": 1.3088144678719051,
      "learning_rate": 4.7223363971825655e-06,
      "loss": 0.268,
      "step": 8210
    },
    {
      "epoch": 3.55,
      "grad_norm": 1.0416234935318807,
      "learning_rate": 4.7095343188978885e-06,
      "loss": 0.2657,
      "step": 8215
    },
    {
      "epoch": 3.55,
      "grad_norm": 1.050594486786159,
      "learning_rate": 4.696744270131719e-06,
      "loss": 0.2606,
      "step": 8220
    },
    {
      "epoch": 3.55,
      "grad_norm": 1.0367524661999572,
      "learning_rate": 4.683966279966225e-06,
      "loss": 0.2551,
      "step": 8225
    },
    {
      "epoch": 3.56,
      "grad_norm": 1.1194180754236926,
      "learning_rate": 4.6712003774561275e-06,
      "loss": 0.2901,
      "step": 8230
    },
    {
      "epoch": 3.56,
      "grad_norm": 0.8141812221195328,
      "learning_rate": 4.65844659162868e-06,
      "loss": 0.2653,
      "step": 8235
    },
    {
      "epoch": 3.56,
      "grad_norm": 0.8263363451947964,
      "learning_rate": 4.645704951483583e-06,
      "loss": 0.2908,
      "step": 8240
    },
    {
      "epoch": 3.56,
      "grad_norm": 0.6953214011687743,
      "learning_rate": 4.632975485992903e-06,
      "loss": 0.2778,
      "step": 8245
    },
    {
      "epoch": 3.56,
      "grad_norm": 1.2865699217071096,
      "learning_rate": 4.620258224101057e-06,
      "loss": 0.2438,
      "step": 8250
    },
    {
      "epoch": 3.57,
      "grad_norm": 0.6117621448414158,
      "learning_rate": 4.607553194724677e-06,
      "loss": 0.2361,
      "step": 8255
    },
    {
      "epoch": 3.57,
      "grad_norm": 0.8914164995414522,
      "learning_rate": 4.594860426752607e-06,
      "loss": 0.2586,
      "step": 8260
    },
    {
      "epoch": 3.57,
      "grad_norm": 1.6881361423247938,
      "learning_rate": 4.5821799490458e-06,
      "loss": 0.287,
      "step": 8265
    },
    {
      "epoch": 3.57,
      "grad_norm": 1.2899428254311105,
      "learning_rate": 4.569511790437254e-06,
      "loss": 0.3003,
      "step": 8270
    },
    {
      "epoch": 3.57,
      "grad_norm": 1.6864476846021144,
      "learning_rate": 4.556855979731983e-06,
      "loss": 0.2636,
      "step": 8275
    },
    {
      "epoch": 3.58,
      "grad_norm": 0.9578718355240123,
      "learning_rate": 4.544212545706895e-06,
      "loss": 0.2438,
      "step": 8280
    },
    {
      "epoch": 3.58,
      "grad_norm": 1.2250154123094863,
      "learning_rate": 4.531581517110771e-06,
      "loss": 0.2761,
      "step": 8285
    },
    {
      "epoch": 3.58,
      "grad_norm": 0.5692185705890477,
      "learning_rate": 4.5189629226641876e-06,
      "loss": 0.2421,
      "step": 8290
    },
    {
      "epoch": 3.58,
      "grad_norm": 1.4743329422756897,
      "learning_rate": 4.506356791059427e-06,
      "loss": 0.2361,
      "step": 8295
    },
    {
      "epoch": 3.59,
      "grad_norm": 1.4258334907405563,
      "learning_rate": 4.493763150960468e-06,
      "loss": 0.2737,
      "step": 8300
    },
    {
      "epoch": 3.59,
      "grad_norm": 1.0513338499870875,
      "learning_rate": 4.481182031002853e-06,
      "loss": 0.2387,
      "step": 8305
    },
    {
      "epoch": 3.59,
      "grad_norm": 1.5143591042082543,
      "learning_rate": 4.468613459793675e-06,
      "loss": 0.2581,
      "step": 8310
    },
    {
      "epoch": 3.59,
      "grad_norm": 0.8425164174407974,
      "learning_rate": 4.456057465911489e-06,
      "loss": 0.2787,
      "step": 8315
    },
    {
      "epoch": 3.59,
      "grad_norm": 1.1182810076099785,
      "learning_rate": 4.443514077906242e-06,
      "loss": 0.2749,
      "step": 8320
    },
    {
      "epoch": 3.6,
      "grad_norm": 0.9606133791212749,
      "learning_rate": 4.430983324299242e-06,
      "loss": 0.2766,
      "step": 8325
    },
    {
      "epoch": 3.6,
      "grad_norm": 2.106283493636646,
      "learning_rate": 4.418465233583039e-06,
      "loss": 0.3489,
      "step": 8330
    },
    {
      "epoch": 3.6,
      "grad_norm": 0.8807348145669678,
      "learning_rate": 4.405959834221411e-06,
      "loss": 0.2673,
      "step": 8335
    },
    {
      "epoch": 3.6,
      "grad_norm": 1.2524322039073619,
      "learning_rate": 4.3934671546492715e-06,
      "loss": 0.247,
      "step": 8340
    },
    {
      "epoch": 3.6,
      "grad_norm": 1.0394890463062174,
      "learning_rate": 4.380987223272602e-06,
      "loss": 0.256,
      "step": 8345
    },
    {
      "epoch": 3.61,
      "grad_norm": 1.00137341927747,
      "learning_rate": 4.368520068468423e-06,
      "loss": 0.2457,
      "step": 8350
    },
    {
      "epoch": 3.61,
      "grad_norm": 1.7009249189017777,
      "learning_rate": 4.356065718584673e-06,
      "loss": 0.3016,
      "step": 8355
    },
    {
      "epoch": 3.61,
      "grad_norm": 1.5553016418878596,
      "learning_rate": 4.343624201940195e-06,
      "loss": 0.3043,
      "step": 8360
    },
    {
      "epoch": 3.61,
      "grad_norm": 1.197565393045384,
      "learning_rate": 4.331195546824646e-06,
      "loss": 0.2368,
      "step": 8365
    },
    {
      "epoch": 3.62,
      "grad_norm": 1.4071865316430026,
      "learning_rate": 4.318779781498429e-06,
      "loss": 0.3015,
      "step": 8370
    },
    {
      "epoch": 3.62,
      "grad_norm": 1.3096415330679698,
      "learning_rate": 4.306376934192659e-06,
      "loss": 0.2764,
      "step": 8375
    },
    {
      "epoch": 3.62,
      "grad_norm": 1.209092329560919,
      "learning_rate": 4.293987033109055e-06,
      "loss": 0.2275,
      "step": 8380
    },
    {
      "epoch": 3.62,
      "grad_norm": 1.7613370849719026,
      "learning_rate": 4.281610106419913e-06,
      "loss": 0.2764,
      "step": 8385
    },
    {
      "epoch": 3.62,
      "grad_norm": 1.295214792210852,
      "learning_rate": 4.269246182268027e-06,
      "loss": 0.2701,
      "step": 8390
    },
    {
      "epoch": 3.63,
      "grad_norm": 1.50791699725049,
      "learning_rate": 4.256895288766614e-06,
      "loss": 0.258,
      "step": 8395
    },
    {
      "epoch": 3.63,
      "grad_norm": 1.654373380876267,
      "learning_rate": 4.244557453999282e-06,
      "loss": 0.2752,
      "step": 8400
    },
    {
      "epoch": 3.63,
      "grad_norm": 1.9897713100742098,
      "learning_rate": 4.232232706019923e-06,
      "loss": 0.325,
      "step": 8405
    },
    {
      "epoch": 3.63,
      "grad_norm": 1.02148414818951,
      "learning_rate": 4.219921072852689e-06,
      "loss": 0.2122,
      "step": 8410
    },
    {
      "epoch": 3.63,
      "grad_norm": 1.1892877259874424,
      "learning_rate": 4.207622582491907e-06,
      "loss": 0.2297,
      "step": 8415
    },
    {
      "epoch": 3.64,
      "grad_norm": 1.237889295378514,
      "learning_rate": 4.195337262902008e-06,
      "loss": 0.3163,
      "step": 8420
    },
    {
      "epoch": 3.64,
      "grad_norm": 1.0589613077003643,
      "learning_rate": 4.183065142017503e-06,
      "loss": 0.2808,
      "step": 8425
    },
    {
      "epoch": 3.64,
      "grad_norm": 0.7728753942869604,
      "learning_rate": 4.17080624774286e-06,
      "loss": 0.2444,
      "step": 8430
    },
    {
      "epoch": 3.64,
      "grad_norm": 1.6520086613499878,
      "learning_rate": 4.158560607952491e-06,
      "loss": 0.2676,
      "step": 8435
    },
    {
      "epoch": 3.65,
      "grad_norm": 1.3800566942565582,
      "learning_rate": 4.146328250490669e-06,
      "loss": 0.2525,
      "step": 8440
    },
    {
      "epoch": 3.65,
      "grad_norm": 1.0603690959383418,
      "learning_rate": 4.13410920317145e-06,
      "loss": 0.2182,
      "step": 8445
    },
    {
      "epoch": 3.65,
      "grad_norm": 1.1364069940034944,
      "learning_rate": 4.121903493778653e-06,
      "loss": 0.2353,
      "step": 8450
    },
    {
      "epoch": 3.65,
      "grad_norm": 1.2550604556115696,
      "learning_rate": 4.1097111500657406e-06,
      "loss": 0.2702,
      "step": 8455
    },
    {
      "epoch": 3.65,
      "grad_norm": 1.0855724510468503,
      "learning_rate": 4.097532199755801e-06,
      "loss": 0.3182,
      "step": 8460
    },
    {
      "epoch": 3.66,
      "grad_norm": 0.8061230337692806,
      "learning_rate": 4.085366670541469e-06,
      "loss": 0.274,
      "step": 8465
    },
    {
      "epoch": 3.66,
      "grad_norm": 0.8247683248127725,
      "learning_rate": 4.073214590084846e-06,
      "loss": 0.2667,
      "step": 8470
    },
    {
      "epoch": 3.66,
      "grad_norm": 1.1839230068133735,
      "learning_rate": 4.061075986017477e-06,
      "loss": 0.2364,
      "step": 8475
    },
    {
      "epoch": 3.66,
      "grad_norm": 1.620173384449278,
      "learning_rate": 4.048950885940244e-06,
      "loss": 0.212,
      "step": 8480
    },
    {
      "epoch": 3.67,
      "grad_norm": 1.193378324765039,
      "learning_rate": 4.036839317423335e-06,
      "loss": 0.1993,
      "step": 8485
    },
    {
      "epoch": 3.67,
      "grad_norm": 0.8733667456843335,
      "learning_rate": 4.024741308006168e-06,
      "loss": 0.2398,
      "step": 8490
    },
    {
      "epoch": 3.67,
      "grad_norm": 1.102519578294471,
      "learning_rate": 4.012656885197321e-06,
      "loss": 0.2587,
      "step": 8495
    },
    {
      "epoch": 3.67,
      "grad_norm": 2.102655622053936,
      "learning_rate": 4.0005860764744996e-06,
      "loss": 0.3024,
      "step": 8500
    },
    {
      "epoch": 3.67,
      "grad_norm": 1.2982264667539742,
      "learning_rate": 3.988528909284428e-06,
      "loss": 0.2769,
      "step": 8505
    },
    {
      "epoch": 3.68,
      "grad_norm": 0.9207503337068603,
      "learning_rate": 3.97648541104283e-06,
      "loss": 0.2297,
      "step": 8510
    },
    {
      "epoch": 3.68,
      "grad_norm": 1.2932708481162427,
      "learning_rate": 3.964455609134345e-06,
      "loss": 0.2235,
      "step": 8515
    },
    {
      "epoch": 3.68,
      "grad_norm": 0.7143624184996928,
      "learning_rate": 3.952439530912457e-06,
      "loss": 0.2468,
      "step": 8520
    },
    {
      "epoch": 3.68,
      "grad_norm": 1.1683102857430236,
      "learning_rate": 3.94043720369947e-06,
      "loss": 0.2602,
      "step": 8525
    },
    {
      "epoch": 3.68,
      "grad_norm": 0.8805470000188161,
      "learning_rate": 3.928448654786397e-06,
      "loss": 0.2471,
      "step": 8530
    },
    {
      "epoch": 3.69,
      "grad_norm": 1.5405448678459484,
      "learning_rate": 3.916473911432934e-06,
      "loss": 0.2497,
      "step": 8535
    },
    {
      "epoch": 3.69,
      "grad_norm": 3.2238368617167614,
      "learning_rate": 3.904513000867386e-06,
      "loss": 0.2465,
      "step": 8540
    },
    {
      "epoch": 3.69,
      "grad_norm": 1.4638528980101133,
      "learning_rate": 3.892565950286592e-06,
      "loss": 0.2678,
      "step": 8545
    },
    {
      "epoch": 3.69,
      "grad_norm": 1.5184496363579658,
      "learning_rate": 3.880632786855899e-06,
      "loss": 0.2024,
      "step": 8550
    },
    {
      "epoch": 3.7,
      "grad_norm": 1.8252758600534205,
      "learning_rate": 3.868713537709055e-06,
      "loss": 0.3266,
      "step": 8555
    },
    {
      "epoch": 3.7,
      "grad_norm": 0.64264179568571,
      "learning_rate": 3.856808229948179e-06,
      "loss": 0.1998,
      "step": 8560
    },
    {
      "epoch": 3.7,
      "grad_norm": 1.1336128812569761,
      "learning_rate": 3.844916890643696e-06,
      "loss": 0.2246,
      "step": 8565
    },
    {
      "epoch": 3.7,
      "grad_norm": 0.941977941979213,
      "learning_rate": 3.833039546834252e-06,
      "loss": 0.2437,
      "step": 8570
    },
    {
      "epoch": 3.7,
      "grad_norm": 0.9723728987072218,
      "learning_rate": 3.821176225526694e-06,
      "loss": 0.2441,
      "step": 8575
    },
    {
      "epoch": 3.71,
      "grad_norm": 1.886020160147755,
      "learning_rate": 3.80932695369596e-06,
      "loss": 0.2605,
      "step": 8580
    },
    {
      "epoch": 3.71,
      "grad_norm": 0.8056642057647267,
      "learning_rate": 3.797491758285059e-06,
      "loss": 0.2323,
      "step": 8585
    },
    {
      "epoch": 3.71,
      "grad_norm": 1.2590539891249422,
      "learning_rate": 3.7856706662049893e-06,
      "loss": 0.2743,
      "step": 8590
    },
    {
      "epoch": 3.71,
      "grad_norm": 1.4292301576478457,
      "learning_rate": 3.7738637043346683e-06,
      "loss": 0.2592,
      "step": 8595
    },
    {
      "epoch": 3.71,
      "grad_norm": 0.9180460073700387,
      "learning_rate": 3.7620708995209087e-06,
      "loss": 0.2075,
      "step": 8600
    },
    {
      "epoch": 3.72,
      "grad_norm": 0.6673357074467728,
      "learning_rate": 3.7502922785783093e-06,
      "loss": 0.1997,
      "step": 8605
    },
    {
      "epoch": 3.72,
      "grad_norm": 0.762136138706124,
      "learning_rate": 3.738527868289229e-06,
      "loss": 0.2719,
      "step": 8610
    },
    {
      "epoch": 3.72,
      "grad_norm": 1.1211722000299258,
      "learning_rate": 3.726777695403716e-06,
      "loss": 0.2276,
      "step": 8615
    },
    {
      "epoch": 3.72,
      "grad_norm": 1.5634256260632298,
      "learning_rate": 3.7150417866394304e-06,
      "loss": 0.2348,
      "step": 8620
    },
    {
      "epoch": 3.73,
      "grad_norm": 1.6256231637279743,
      "learning_rate": 3.7033201686816257e-06,
      "loss": 0.2406,
      "step": 8625
    },
    {
      "epoch": 3.73,
      "grad_norm": 1.2187105096731732,
      "learning_rate": 3.691612868183031e-06,
      "loss": 0.2668,
      "step": 8630
    },
    {
      "epoch": 3.73,
      "grad_norm": 1.2149726719763858,
      "learning_rate": 3.6799199117638383e-06,
      "loss": 0.2592,
      "step": 8635
    },
    {
      "epoch": 3.73,
      "grad_norm": 0.764253856485244,
      "learning_rate": 3.6682413260116235e-06,
      "loss": 0.2218,
      "step": 8640
    },
    {
      "epoch": 3.73,
      "grad_norm": 0.9006111227777452,
      "learning_rate": 3.656577137481271e-06,
      "loss": 0.2331,
      "step": 8645
    },
    {
      "epoch": 3.74,
      "grad_norm": 0.9146889853788844,
      "learning_rate": 3.644927372694954e-06,
      "loss": 0.2657,
      "step": 8650
    },
    {
      "epoch": 3.74,
      "grad_norm": 1.1118503466233536,
      "learning_rate": 3.633292058142024e-06,
      "loss": 0.2393,
      "step": 8655
    },
    {
      "epoch": 3.74,
      "grad_norm": 1.0922268807899702,
      "learning_rate": 3.621671220278993e-06,
      "loss": 0.2595,
      "step": 8660
    },
    {
      "epoch": 3.74,
      "grad_norm": 1.425843520802282,
      "learning_rate": 3.6100648855294417e-06,
      "loss": 0.2259,
      "step": 8665
    },
    {
      "epoch": 3.75,
      "grad_norm": 1.0010134323284015,
      "learning_rate": 3.598473080283982e-06,
      "loss": 0.2282,
      "step": 8670
    },
    {
      "epoch": 3.75,
      "grad_norm": 1.6319837617789361,
      "learning_rate": 3.5868958309001942e-06,
      "loss": 0.2152,
      "step": 8675
    },
    {
      "epoch": 3.75,
      "grad_norm": 1.1531835012647713,
      "learning_rate": 3.575333163702547e-06,
      "loss": 0.1998,
      "step": 8680
    },
    {
      "epoch": 3.75,
      "grad_norm": 0.7140806404064783,
      "learning_rate": 3.5637851049823645e-06,
      "loss": 0.1986,
      "step": 8685
    },
    {
      "epoch": 3.75,
      "grad_norm": 1.396396649138274,
      "learning_rate": 3.5522516809977416e-06,
      "loss": 0.2749,
      "step": 8690
    },
    {
      "epoch": 3.76,
      "grad_norm": 1.790703192319098,
      "learning_rate": 3.5407329179735027e-06,
      "loss": 0.2027,
      "step": 8695
    },
    {
      "epoch": 3.76,
      "grad_norm": 1.1619765530746924,
      "learning_rate": 3.529228842101149e-06,
      "loss": 0.2178,
      "step": 8700
    },
    {
      "epoch": 3.76,
      "grad_norm": 1.1226391583686264,
      "learning_rate": 3.517739479538762e-06,
      "loss": 0.2347,
      "step": 8705
    },
    {
      "epoch": 3.76,
      "grad_norm": 1.5086143605668283,
      "learning_rate": 3.5062648564109857e-06,
      "loss": 0.2359,
      "step": 8710
    },
    {
      "epoch": 3.76,
      "grad_norm": 1.6945930578072637,
      "learning_rate": 3.4948049988089373e-06,
      "loss": 0.2149,
      "step": 8715
    },
    {
      "epoch": 3.77,
      "grad_norm": 1.0255293861017403,
      "learning_rate": 3.4833599327901656e-06,
      "loss": 0.2112,
      "step": 8720
    },
    {
      "epoch": 3.77,
      "grad_norm": 1.7693180895647955,
      "learning_rate": 3.4719296843785977e-06,
      "loss": 0.2457,
      "step": 8725
    },
    {
      "epoch": 3.77,
      "grad_norm": 1.4966986753679365,
      "learning_rate": 3.4605142795644454e-06,
      "loss": 0.2843,
      "step": 8730
    },
    {
      "epoch": 3.77,
      "grad_norm": 0.670891091445466,
      "learning_rate": 3.4491137443041865e-06,
      "loss": 0.2146,
      "step": 8735
    },
    {
      "epoch": 3.78,
      "grad_norm": 1.8052307720833396,
      "learning_rate": 3.437728104520478e-06,
      "loss": 0.2104,
      "step": 8740
    },
    {
      "epoch": 3.78,
      "grad_norm": 1.1813095312113593,
      "learning_rate": 3.4263573861021104e-06,
      "loss": 0.2118,
      "step": 8745
    },
    {
      "epoch": 3.78,
      "grad_norm": 1.4173325757700486,
      "learning_rate": 3.415001614903958e-06,
      "loss": 0.2096,
      "step": 8750
    },
    {
      "epoch": 3.78,
      "grad_norm": 2.4218252609777213,
      "learning_rate": 3.4036608167468856e-06,
      "loss": 0.2576,
      "step": 8755
    },
    {
      "epoch": 3.78,
      "grad_norm": 1.461680496985334,
      "learning_rate": 3.3923350174177304e-06,
      "loss": 0.2608,
      "step": 8760
    },
    {
      "epoch": 3.79,
      "grad_norm": 1.2997679357447258,
      "learning_rate": 3.381024242669213e-06,
      "loss": 0.2357,
      "step": 8765
    },
    {
      "epoch": 3.79,
      "grad_norm": 0.8294127235093294,
      "learning_rate": 3.3697285182198957e-06,
      "loss": 0.2536,
      "step": 8770
    },
    {
      "epoch": 3.79,
      "grad_norm": 1.0896228556623861,
      "learning_rate": 3.35844786975413e-06,
      "loss": 0.2489,
      "step": 8775
    },
    {
      "epoch": 3.79,
      "grad_norm": 2.3166544955166914,
      "learning_rate": 3.347182322921967e-06,
      "loss": 0.1742,
      "step": 8780
    },
    {
      "epoch": 3.79,
      "grad_norm": 1.2716000940595575,
      "learning_rate": 3.3359319033391366e-06,
      "loss": 0.2001,
      "step": 8785
    },
    {
      "epoch": 3.8,
      "grad_norm": 1.3878114041564946,
      "learning_rate": 3.3246966365869593e-06,
      "loss": 0.1999,
      "step": 8790
    },
    {
      "epoch": 3.8,
      "grad_norm": 1.4073458518658912,
      "learning_rate": 3.3134765482123056e-06,
      "loss": 0.2048,
      "step": 8795
    },
    {
      "epoch": 3.8,
      "grad_norm": 2.257280307648163,
      "learning_rate": 3.3022716637275466e-06,
      "loss": 0.2383,
      "step": 8800
    },
    {
      "epoch": 3.8,
      "grad_norm": 1.4865011423989034,
      "learning_rate": 3.291082008610459e-06,
      "loss": 0.2236,
      "step": 8805
    },
    {
      "epoch": 3.81,
      "grad_norm": 1.6028396692657991,
      "learning_rate": 3.2799076083042104e-06,
      "loss": 0.2184,
      "step": 8810
    },
    {
      "epoch": 3.81,
      "grad_norm": 0.7425058030584089,
      "learning_rate": 3.268748488217267e-06,
      "loss": 0.1825,
      "step": 8815
    },
    {
      "epoch": 3.81,
      "grad_norm": 1.8488035722222276,
      "learning_rate": 3.2576046737233567e-06,
      "loss": 0.2211,
      "step": 8820
    },
    {
      "epoch": 3.81,
      "grad_norm": 1.3816883588335565,
      "learning_rate": 3.246476190161418e-06,
      "loss": 0.2256,
      "step": 8825
    },
    {
      "epoch": 3.81,
      "grad_norm": 2.071282818859703,
      "learning_rate": 3.235363062835507e-06,
      "loss": 0.2327,
      "step": 8830
    },
    {
      "epoch": 3.82,
      "grad_norm": 1.4979771549526812,
      "learning_rate": 3.224265317014781e-06,
      "loss": 0.213,
      "step": 8835
    },
    {
      "epoch": 3.82,
      "grad_norm": 1.056089605035267,
      "learning_rate": 3.213182977933408e-06,
      "loss": 0.2149,
      "step": 8840
    },
    {
      "epoch": 3.82,
      "grad_norm": 1.352835390586533,
      "learning_rate": 3.2021160707905353e-06,
      "loss": 0.1822,
      "step": 8845
    },
    {
      "epoch": 3.82,
      "grad_norm": 1.010725156923398,
      "learning_rate": 3.1910646207502173e-06,
      "loss": 0.1812,
      "step": 8850
    },
    {
      "epoch": 3.83,
      "grad_norm": 1.9017812068754216,
      "learning_rate": 3.1800286529413605e-06,
      "loss": 0.2013,
      "step": 8855
    },
    {
      "epoch": 3.83,
      "grad_norm": 1.6218418481134442,
      "learning_rate": 3.169008192457672e-06,
      "loss": 0.1713,
      "step": 8860
    },
    {
      "epoch": 3.83,
      "grad_norm": 1.8330711024738942,
      "learning_rate": 3.15800326435759e-06,
      "loss": 0.2366,
      "step": 8865
    },
    {
      "epoch": 3.83,
      "grad_norm": 1.325482478087709,
      "learning_rate": 3.1470138936642415e-06,
      "loss": 0.1753,
      "step": 8870
    },
    {
      "epoch": 3.83,
      "grad_norm": 1.2317388715406854,
      "learning_rate": 3.1360401053653776e-06,
      "loss": 0.1867,
      "step": 8875
    },
    {
      "epoch": 3.84,
      "grad_norm": 1.383078361027162,
      "learning_rate": 3.125081924413318e-06,
      "loss": 0.1834,
      "step": 8880
    },
    {
      "epoch": 3.84,
      "grad_norm": 1.6364075116336452,
      "learning_rate": 3.1141393757248973e-06,
      "loss": 0.1965,
      "step": 8885
    },
    {
      "epoch": 3.84,
      "grad_norm": 1.5408796381251828,
      "learning_rate": 3.103212484181396e-06,
      "loss": 0.1789,
      "step": 8890
    },
    {
      "epoch": 3.84,
      "grad_norm": 0.7313558930891003,
      "learning_rate": 3.092301274628502e-06,
      "loss": 0.221,
      "step": 8895
    },
    {
      "epoch": 3.84,
      "grad_norm": 1.7236990433362605,
      "learning_rate": 3.0814057718762413e-06,
      "loss": 0.1794,
      "step": 8900
    },
    {
      "epoch": 3.85,
      "grad_norm": 1.309025789799166,
      "learning_rate": 3.070526000698928e-06,
      "loss": 0.1904,
      "step": 8905
    },
    {
      "epoch": 3.85,
      "grad_norm": 1.3057919563887057,
      "learning_rate": 3.0596619858351073e-06,
      "loss": 0.1305,
      "step": 8910
    },
    {
      "epoch": 3.85,
      "grad_norm": 1.2826997903059916,
      "learning_rate": 3.0488137519874896e-06,
      "loss": 0.1854,
      "step": 8915
    },
    {
      "epoch": 3.85,
      "grad_norm": 2.0736399556689205,
      "learning_rate": 3.0379813238229074e-06,
      "loss": 0.1735,
      "step": 8920
    },
    {
      "epoch": 3.86,
      "grad_norm": 0.9527611277386501,
      "learning_rate": 3.027164725972256e-06,
      "loss": 0.1645,
      "step": 8925
    },
    {
      "epoch": 3.86,
      "grad_norm": 1.1335358478265696,
      "learning_rate": 3.016363983030434e-06,
      "loss": 0.1621,
      "step": 8930
    },
    {
      "epoch": 3.86,
      "grad_norm": 2.521984786858169,
      "learning_rate": 3.0055791195562887e-06,
      "loss": 0.2026,
      "step": 8935
    },
    {
      "epoch": 3.86,
      "grad_norm": 1.836220144705608,
      "learning_rate": 2.9948101600725565e-06,
      "loss": 0.1724,
      "step": 8940
    },
    {
      "epoch": 3.86,
      "grad_norm": 1.136504620875479,
      "learning_rate": 2.9840571290658184e-06,
      "loss": 0.1774,
      "step": 8945
    },
    {
      "epoch": 3.87,
      "grad_norm": 1.4816584731960452,
      "learning_rate": 2.9733200509864314e-06,
      "loss": 0.178,
      "step": 8950
    },
    {
      "epoch": 3.87,
      "grad_norm": 1.1642803297715238,
      "learning_rate": 2.962598950248481e-06,
      "loss": 0.154,
      "step": 8955
    },
    {
      "epoch": 3.87,
      "grad_norm": 2.7690606796997552,
      "learning_rate": 2.9518938512297267e-06,
      "loss": 0.1782,
      "step": 8960
    },
    {
      "epoch": 3.87,
      "grad_norm": 1.6223686343268287,
      "learning_rate": 2.9412047782715334e-06,
      "loss": 0.1524,
      "step": 8965
    },
    {
      "epoch": 3.87,
      "grad_norm": 1.3358911181846544,
      "learning_rate": 2.930531755678835e-06,
      "loss": 0.1703,
      "step": 8970
    },
    {
      "epoch": 3.88,
      "grad_norm": 0.7845302281471236,
      "learning_rate": 2.9198748077200644e-06,
      "loss": 0.1376,
      "step": 8975
    },
    {
      "epoch": 3.88,
      "grad_norm": 3.196234583882958,
      "learning_rate": 2.909233958627108e-06,
      "loss": 0.1851,
      "step": 8980
    },
    {
      "epoch": 3.88,
      "grad_norm": 1.5247189113614013,
      "learning_rate": 2.8986092325952474e-06,
      "loss": 0.1915,
      "step": 8985
    },
    {
      "epoch": 3.88,
      "grad_norm": 1.7392497600102992,
      "learning_rate": 2.888000653783094e-06,
      "loss": 0.1903,
      "step": 8990
    },
    {
      "epoch": 3.89,
      "grad_norm": 1.2139519723864711,
      "learning_rate": 2.877408246312552e-06,
      "loss": 0.1696,
      "step": 8995
    },
    {
      "epoch": 3.89,
      "grad_norm": 2.3036921230601677,
      "learning_rate": 2.866832034268754e-06,
      "loss": 0.1861,
      "step": 9000
    },
    {
      "epoch": 3.89,
      "grad_norm": 1.998159592775058,
      "learning_rate": 2.8562720417000056e-06,
      "loss": 0.1493,
      "step": 9005
    },
    {
      "epoch": 3.89,
      "grad_norm": 2.0385570109143663,
      "learning_rate": 2.8457282926177376e-06,
      "loss": 0.1725,
      "step": 9010
    },
    {
      "epoch": 3.89,
      "grad_norm": 1.038409279355451,
      "learning_rate": 2.8352008109964336e-06,
      "loss": 0.143,
      "step": 9015
    },
    {
      "epoch": 3.9,
      "grad_norm": 2.8111977365091607,
      "learning_rate": 2.8246896207735997e-06,
      "loss": 0.1763,
      "step": 9020
    },
    {
      "epoch": 3.9,
      "grad_norm": 1.115676301969164,
      "learning_rate": 2.8141947458496943e-06,
      "loss": 0.1369,
      "step": 9025
    },
    {
      "epoch": 3.9,
      "grad_norm": 2.2547605054805384,
      "learning_rate": 2.803716210088079e-06,
      "loss": 0.1525,
      "step": 9030
    },
    {
      "epoch": 3.9,
      "grad_norm": 2.655789824220769,
      "learning_rate": 2.7932540373149654e-06,
      "loss": 0.1598,
      "step": 9035
    },
    {
      "epoch": 3.9,
      "grad_norm": 1.366174287885307,
      "learning_rate": 2.78280825131935e-06,
      "loss": 0.1455,
      "step": 9040
    },
    {
      "epoch": 3.91,
      "grad_norm": 1.5512552019581032,
      "learning_rate": 2.772378875852978e-06,
      "loss": 0.1598,
      "step": 9045
    },
    {
      "epoch": 3.91,
      "grad_norm": 1.7769531325546062,
      "learning_rate": 2.7619659346302775e-06,
      "loss": 0.1343,
      "step": 9050
    },
    {
      "epoch": 3.91,
      "grad_norm": 1.7409746059130171,
      "learning_rate": 2.751569451328305e-06,
      "loss": 0.1497,
      "step": 9055
    },
    {
      "epoch": 3.91,
      "grad_norm": 3.2371623664471656,
      "learning_rate": 2.741189449586704e-06,
      "loss": 0.1452,
      "step": 9060
    },
    {
      "epoch": 3.92,
      "grad_norm": 1.756639522059405,
      "learning_rate": 2.7308259530076286e-06,
      "loss": 0.127,
      "step": 9065
    },
    {
      "epoch": 3.92,
      "grad_norm": 1.4592380310799502,
      "learning_rate": 2.720478985155712e-06,
      "loss": 0.1181,
      "step": 9070
    },
    {
      "epoch": 3.92,
      "grad_norm": 0.5540812705657977,
      "learning_rate": 2.7101485695580033e-06,
      "loss": 0.1276,
      "step": 9075
    },
    {
      "epoch": 3.92,
      "grad_norm": 0.730628334741152,
      "learning_rate": 2.699834729703914e-06,
      "loss": 0.1395,
      "step": 9080
    },
    {
      "epoch": 3.92,
      "grad_norm": 1.2814134539945647,
      "learning_rate": 2.6895374890451686e-06,
      "loss": 0.1406,
      "step": 9085
    },
    {
      "epoch": 3.93,
      "grad_norm": 1.6017173096861919,
      "learning_rate": 2.679256870995739e-06,
      "loss": 0.1497,
      "step": 9090
    },
    {
      "epoch": 3.93,
      "grad_norm": 4.377136176957885,
      "learning_rate": 2.668992898931809e-06,
      "loss": 0.1313,
      "step": 9095
    },
    {
      "epoch": 3.93,
      "grad_norm": 1.1466406266728912,
      "learning_rate": 2.658745596191712e-06,
      "loss": 0.1178,
      "step": 9100
    },
    {
      "epoch": 3.93,
      "grad_norm": 1.4412542325562125,
      "learning_rate": 2.648514986075874e-06,
      "loss": 0.1213,
      "step": 9105
    },
    {
      "epoch": 3.94,
      "grad_norm": 1.0094886930705556,
      "learning_rate": 2.638301091846772e-06,
      "loss": 0.1282,
      "step": 9110
    },
    {
      "epoch": 3.94,
      "grad_norm": 1.9523149478373765,
      "learning_rate": 2.6281039367288653e-06,
      "loss": 0.1332,
      "step": 9115
    },
    {
      "epoch": 3.94,
      "grad_norm": 1.6163037295208362,
      "learning_rate": 2.6179235439085584e-06,
      "loss": 0.1318,
      "step": 9120
    },
    {
      "epoch": 3.94,
      "grad_norm": 1.6749344686293903,
      "learning_rate": 2.607759936534138e-06,
      "loss": 0.1138,
      "step": 9125
    },
    {
      "epoch": 3.94,
      "grad_norm": 3.742751088082727,
      "learning_rate": 2.5976131377157264e-06,
      "loss": 0.1408,
      "step": 9130
    },
    {
      "epoch": 3.95,
      "grad_norm": 1.2402729828056847,
      "learning_rate": 2.587483170525229e-06,
      "loss": 0.1155,
      "step": 9135
    },
    {
      "epoch": 3.95,
      "grad_norm": 1.5290565815887704,
      "learning_rate": 2.577370057996268e-06,
      "loss": 0.1458,
      "step": 9140
    },
    {
      "epoch": 3.95,
      "grad_norm": 1.728302719729036,
      "learning_rate": 2.5672738231241547e-06,
      "loss": 0.1222,
      "step": 9145
    },
    {
      "epoch": 3.95,
      "grad_norm": 1.6795308054327185,
      "learning_rate": 2.557194488865814e-06,
      "loss": 0.1015,
      "step": 9150
    },
    {
      "epoch": 3.95,
      "grad_norm": 1.1889232094471702,
      "learning_rate": 2.5471320781397493e-06,
      "loss": 0.1276,
      "step": 9155
    },
    {
      "epoch": 3.96,
      "grad_norm": 1.677934768849866,
      "learning_rate": 2.537086613825981e-06,
      "loss": 0.1133,
      "step": 9160
    },
    {
      "epoch": 3.96,
      "grad_norm": 0.8772658619002762,
      "learning_rate": 2.5270581187659903e-06,
      "loss": 0.1193,
      "step": 9165
    },
    {
      "epoch": 3.96,
      "grad_norm": 1.4849426603993638,
      "learning_rate": 2.5170466157626803e-06,
      "loss": 0.1348,
      "step": 9170
    },
    {
      "epoch": 3.96,
      "grad_norm": 1.2635413147574273,
      "learning_rate": 2.5070521275803173e-06,
      "loss": 0.1225,
      "step": 9175
    },
    {
      "epoch": 3.97,
      "grad_norm": 1.921588573285143,
      "learning_rate": 2.4970746769444763e-06,
      "loss": 0.1412,
      "step": 9180
    },
    {
      "epoch": 3.97,
      "grad_norm": 2.084574262761302,
      "learning_rate": 2.487114286541997e-06,
      "loss": 0.1192,
      "step": 9185
    },
    {
      "epoch": 3.97,
      "grad_norm": 4.839684727447521,
      "learning_rate": 2.4771709790209163e-06,
      "loss": 0.0956,
      "step": 9190
    },
    {
      "epoch": 3.97,
      "grad_norm": 2.0640268839481233,
      "learning_rate": 2.4672447769904395e-06,
      "loss": 0.1107,
      "step": 9195
    },
    {
      "epoch": 3.97,
      "grad_norm": 0.8056949030488845,
      "learning_rate": 2.4573357030208724e-06,
      "loss": 0.1021,
      "step": 9200
    },
    {
      "epoch": 3.98,
      "grad_norm": 1.4232707574434897,
      "learning_rate": 2.447443779643576e-06,
      "loss": 0.0994,
      "step": 9205
    },
    {
      "epoch": 3.98,
      "grad_norm": 1.2627211187763787,
      "learning_rate": 2.4375690293509147e-06,
      "loss": 0.1016,
      "step": 9210
    },
    {
      "epoch": 3.98,
      "grad_norm": 2.7302827184383833,
      "learning_rate": 2.4277114745961994e-06,
      "loss": 0.1271,
      "step": 9215
    },
    {
      "epoch": 3.98,
      "grad_norm": 1.3872210610935318,
      "learning_rate": 2.4178711377936483e-06,
      "loss": 0.1084,
      "step": 9220
    },
    {
      "epoch": 3.98,
      "grad_norm": 2.743388377176648,
      "learning_rate": 2.408048041318325e-06,
      "loss": 0.1177,
      "step": 9225
    },
    {
      "epoch": 3.99,
      "grad_norm": 1.617957502406765,
      "learning_rate": 2.3982422075060942e-06,
      "loss": 0.1141,
      "step": 9230
    },
    {
      "epoch": 3.99,
      "grad_norm": 0.9313918034039486,
      "learning_rate": 2.38845365865357e-06,
      "loss": 0.0999,
      "step": 9235
    },
    {
      "epoch": 3.99,
      "grad_norm": 0.961739039698364,
      "learning_rate": 2.3786824170180577e-06,
      "loss": 0.1015,
      "step": 9240
    },
    {
      "epoch": 3.99,
      "grad_norm": 0.9267930669073516,
      "learning_rate": 2.3689285048175125e-06,
      "loss": 0.1093,
      "step": 9245
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.2497965354801768,
      "learning_rate": 2.35919194423049e-06,
      "loss": 0.1134,
      "step": 9250
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.7334404670095594,
      "learning_rate": 2.349472757396084e-06,
      "loss": 0.0949,
      "step": 9255
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.5321023821904252,
      "learning_rate": 2.3397709664138933e-06,
      "loss": 0.0959,
      "step": 9260
    },
    {
      "epoch": 4.0,
      "eval_loss": 1.1934491395950317,
      "eval_runtime": 352.8997,
      "eval_samples_per_second": 21.272,
      "eval_steps_per_second": 0.334,
      "step": 9260
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.0653990301674712,
      "learning_rate": 2.33008659334395e-06,
      "loss": 0.0982,
      "step": 9265
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.3878497410391784,
      "learning_rate": 2.3204196602066875e-06,
      "loss": 0.1019,
      "step": 9270
    },
    {
      "epoch": 4.01,
      "grad_norm": 1.3837070784620797,
      "learning_rate": 2.310770188982886e-06,
      "loss": 0.1095,
      "step": 9275
    },
    {
      "epoch": 4.01,
      "grad_norm": 1.3288988915285944,
      "learning_rate": 2.3011382016136176e-06,
      "loss": 0.0963,
      "step": 9280
    },
    {
      "epoch": 4.01,
      "grad_norm": 0.5189759633457958,
      "learning_rate": 2.2915237200002017e-06,
      "loss": 0.079,
      "step": 9285
    },
    {
      "epoch": 4.01,
      "grad_norm": 0.4571507747282879,
      "learning_rate": 2.2819267660041488e-06,
      "loss": 0.1004,
      "step": 9290
    },
    {
      "epoch": 4.02,
      "grad_norm": 1.0405793385567381,
      "learning_rate": 2.272347361447117e-06,
      "loss": 0.0998,
      "step": 9295
    },
    {
      "epoch": 4.02,
      "grad_norm": 1.047472346373859,
      "learning_rate": 2.2627855281108637e-06,
      "loss": 0.0876,
      "step": 9300
    },
    {
      "epoch": 4.02,
      "grad_norm": 0.8366712173763914,
      "learning_rate": 2.2532412877371866e-06,
      "loss": 0.0993,
      "step": 9305
    },
    {
      "epoch": 4.02,
      "grad_norm": 1.2776670445610616,
      "learning_rate": 2.2437146620278892e-06,
      "loss": 0.0909,
      "step": 9310
    },
    {
      "epoch": 4.02,
      "grad_norm": 0.8102516110822815,
      "learning_rate": 2.23420567264471e-06,
      "loss": 0.0914,
      "step": 9315
    },
    {
      "epoch": 4.03,
      "grad_norm": 1.025504447258419,
      "learning_rate": 2.2247143412092954e-06,
      "loss": 0.095,
      "step": 9320
    },
    {
      "epoch": 4.03,
      "grad_norm": 1.0695140092879165,
      "learning_rate": 2.2152406893031375e-06,
      "loss": 0.0822,
      "step": 9325
    },
    {
      "epoch": 4.03,
      "grad_norm": 0.9489084492072272,
      "learning_rate": 2.2057847384675313e-06,
      "loss": 0.0793,
      "step": 9330
    },
    {
      "epoch": 4.03,
      "grad_norm": 0.411369707442681,
      "learning_rate": 2.196346510203522e-06,
      "loss": 0.0785,
      "step": 9335
    },
    {
      "epoch": 4.03,
      "grad_norm": 0.9137657417624943,
      "learning_rate": 2.1869260259718496e-06,
      "loss": 0.094,
      "step": 9340
    },
    {
      "epoch": 4.04,
      "grad_norm": 1.0826618792252996,
      "learning_rate": 2.1775233071929192e-06,
      "loss": 0.0841,
      "step": 9345
    },
    {
      "epoch": 4.04,
      "grad_norm": 1.0704564255497284,
      "learning_rate": 2.1681383752467322e-06,
      "loss": 0.1025,
      "step": 9350
    },
    {
      "epoch": 4.04,
      "grad_norm": 1.0041128774767623,
      "learning_rate": 2.158771251472852e-06,
      "loss": 0.0908,
      "step": 9355
    },
    {
      "epoch": 4.04,
      "grad_norm": 0.8948045733048677,
      "learning_rate": 2.1494219571703477e-06,
      "loss": 0.0813,
      "step": 9360
    },
    {
      "epoch": 4.05,
      "grad_norm": 1.582811230887437,
      "learning_rate": 2.1400905135977413e-06,
      "loss": 0.096,
      "step": 9365
    },
    {
      "epoch": 4.05,
      "grad_norm": 1.4642103391429275,
      "learning_rate": 2.130776941972974e-06,
      "loss": 0.0998,
      "step": 9370
    },
    {
      "epoch": 4.05,
      "grad_norm": 0.7389464692871076,
      "learning_rate": 2.121481263473345e-06,
      "loss": 0.1021,
      "step": 9375
    },
    {
      "epoch": 4.05,
      "grad_norm": 0.9025632399425323,
      "learning_rate": 2.112203499235472e-06,
      "loss": 0.091,
      "step": 9380
    },
    {
      "epoch": 4.05,
      "grad_norm": 0.9300066681199844,
      "learning_rate": 2.1029436703552376e-06,
      "loss": 0.0855,
      "step": 9385
    },
    {
      "epoch": 4.06,
      "grad_norm": 0.8485203002577559,
      "learning_rate": 2.0937017978877363e-06,
      "loss": 0.076,
      "step": 9390
    },
    {
      "epoch": 4.06,
      "grad_norm": 0.905058540351386,
      "learning_rate": 2.0844779028472407e-06,
      "loss": 0.0905,
      "step": 9395
    },
    {
      "epoch": 4.06,
      "grad_norm": 1.1361180090897958,
      "learning_rate": 2.075272006207144e-06,
      "loss": 0.0845,
      "step": 9400
    },
    {
      "epoch": 4.06,
      "grad_norm": 0.9367653552399894,
      "learning_rate": 2.0660841288999158e-06,
      "loss": 0.0797,
      "step": 9405
    },
    {
      "epoch": 4.06,
      "grad_norm": 0.918363753389822,
      "learning_rate": 2.0569142918170514e-06,
      "loss": 0.0822,
      "step": 9410
    },
    {
      "epoch": 4.07,
      "grad_norm": 1.4120591611331728,
      "learning_rate": 2.0477625158090242e-06,
      "loss": 0.0776,
      "step": 9415
    },
    {
      "epoch": 4.07,
      "grad_norm": 0.8011915390179812,
      "learning_rate": 2.038628821685242e-06,
      "loss": 0.0819,
      "step": 9420
    },
    {
      "epoch": 4.07,
      "grad_norm": 1.2772706148258295,
      "learning_rate": 2.029513230214001e-06,
      "loss": 0.0807,
      "step": 9425
    },
    {
      "epoch": 4.07,
      "grad_norm": 0.9856245196630447,
      "learning_rate": 2.0204157621224297e-06,
      "loss": 0.0789,
      "step": 9430
    },
    {
      "epoch": 4.08,
      "grad_norm": 1.0849195766702093,
      "learning_rate": 2.011336438096454e-06,
      "loss": 0.0847,
      "step": 9435
    },
    {
      "epoch": 4.08,
      "grad_norm": 0.9174908838938806,
      "learning_rate": 2.0022752787807354e-06,
      "loss": 0.0801,
      "step": 9440
    },
    {
      "epoch": 4.08,
      "grad_norm": 0.8817523981249608,
      "learning_rate": 1.9932323047786373e-06,
      "loss": 0.0814,
      "step": 9445
    },
    {
      "epoch": 4.08,
      "grad_norm": 0.7006834202212598,
      "learning_rate": 1.984207536652174e-06,
      "loss": 0.0724,
      "step": 9450
    },
    {
      "epoch": 4.08,
      "grad_norm": 0.9119370518619543,
      "learning_rate": 1.975200994921961e-06,
      "loss": 0.0753,
      "step": 9455
    },
    {
      "epoch": 4.09,
      "grad_norm": 0.6143322482587131,
      "learning_rate": 1.9662127000671738e-06,
      "loss": 0.0905,
      "step": 9460
    },
    {
      "epoch": 4.09,
      "grad_norm": 1.1511015120231334,
      "learning_rate": 1.9572426725254902e-06,
      "loss": 0.1014,
      "step": 9465
    },
    {
      "epoch": 4.09,
      "grad_norm": 0.6188993889727594,
      "learning_rate": 1.9482909326930586e-06,
      "loss": 0.0747,
      "step": 9470
    },
    {
      "epoch": 4.09,
      "grad_norm": 1.3750071264768102,
      "learning_rate": 1.9393575009244437e-06,
      "loss": 0.0899,
      "step": 9475
    },
    {
      "epoch": 4.1,
      "grad_norm": 1.2496684123687336,
      "learning_rate": 1.930442397532578e-06,
      "loss": 0.0851,
      "step": 9480
    },
    {
      "epoch": 4.1,
      "grad_norm": 1.3788685782832,
      "learning_rate": 1.9215456427887257e-06,
      "loss": 0.0787,
      "step": 9485
    },
    {
      "epoch": 4.1,
      "grad_norm": 0.792384806701021,
      "learning_rate": 1.912667256922418e-06,
      "loss": 0.0767,
      "step": 9490
    },
    {
      "epoch": 4.1,
      "grad_norm": 0.6332410126137582,
      "learning_rate": 1.9038072601214285e-06,
      "loss": 0.0763,
      "step": 9495
    },
    {
      "epoch": 4.1,
      "grad_norm": 0.7226336577944539,
      "learning_rate": 1.8949656725317144e-06,
      "loss": 0.0907,
      "step": 9500
    },
    {
      "epoch": 4.11,
      "grad_norm": 0.8517719067216786,
      "learning_rate": 1.8861425142573742e-06,
      "loss": 0.0793,
      "step": 9505
    },
    {
      "epoch": 4.11,
      "grad_norm": 0.7481972063779955,
      "learning_rate": 1.877337805360604e-06,
      "loss": 0.0822,
      "step": 9510
    },
    {
      "epoch": 4.11,
      "grad_norm": 1.2484833698888498,
      "learning_rate": 1.8685515658616404e-06,
      "loss": 0.0834,
      "step": 9515
    },
    {
      "epoch": 4.11,
      "grad_norm": 0.5698463781012858,
      "learning_rate": 1.8597838157387339e-06,
      "loss": 0.0817,
      "step": 9520
    },
    {
      "epoch": 4.11,
      "grad_norm": 0.9409277359440581,
      "learning_rate": 1.851034574928091e-06,
      "loss": 0.0747,
      "step": 9525
    },
    {
      "epoch": 4.12,
      "grad_norm": 0.590114308975083,
      "learning_rate": 1.8423038633238322e-06,
      "loss": 0.0787,
      "step": 9530
    },
    {
      "epoch": 4.12,
      "grad_norm": 1.0433688110927457,
      "learning_rate": 1.8335917007779403e-06,
      "loss": 0.0789,
      "step": 9535
    },
    {
      "epoch": 4.12,
      "grad_norm": 1.5691971033584857,
      "learning_rate": 1.824898107100227e-06,
      "loss": 0.0957,
      "step": 9540
    },
    {
      "epoch": 4.12,
      "grad_norm": 0.6829735197372712,
      "learning_rate": 1.816223102058282e-06,
      "loss": 0.0688,
      "step": 9545
    },
    {
      "epoch": 4.13,
      "grad_norm": 1.1699567655757332,
      "learning_rate": 1.8075667053774248e-06,
      "loss": 0.0761,
      "step": 9550
    },
    {
      "epoch": 4.13,
      "grad_norm": 0.8808444999201729,
      "learning_rate": 1.7989289367406682e-06,
      "loss": 0.0857,
      "step": 9555
    },
    {
      "epoch": 4.13,
      "grad_norm": 0.7162116245902845,
      "learning_rate": 1.790309815788659e-06,
      "loss": 0.0695,
      "step": 9560
    },
    {
      "epoch": 4.13,
      "grad_norm": 0.754845419443789,
      "learning_rate": 1.78170936211965e-06,
      "loss": 0.0741,
      "step": 9565
    },
    {
      "epoch": 4.13,
      "grad_norm": 0.5516335707135073,
      "learning_rate": 1.7731275952894488e-06,
      "loss": 0.0727,
      "step": 9570
    },
    {
      "epoch": 4.14,
      "grad_norm": 0.523203245900932,
      "learning_rate": 1.7645645348113683e-06,
      "loss": 0.0729,
      "step": 9575
    },
    {
      "epoch": 4.14,
      "grad_norm": 0.7940801541327239,
      "learning_rate": 1.756020200156191e-06,
      "loss": 0.0738,
      "step": 9580
    },
    {
      "epoch": 4.14,
      "grad_norm": 0.4578960285809017,
      "learning_rate": 1.7474946107521152e-06,
      "loss": 0.0746,
      "step": 9585
    },
    {
      "epoch": 4.14,
      "grad_norm": 0.8322294793091983,
      "learning_rate": 1.7389877859847171e-06,
      "loss": 0.0787,
      "step": 9590
    },
    {
      "epoch": 4.14,
      "grad_norm": 0.5550793793378821,
      "learning_rate": 1.7304997451969153e-06,
      "loss": 0.0727,
      "step": 9595
    },
    {
      "epoch": 4.15,
      "grad_norm": 0.7568868583227343,
      "learning_rate": 1.722030507688902e-06,
      "loss": 0.0783,
      "step": 9600
    },
    {
      "epoch": 4.15,
      "grad_norm": 0.7148668063423238,
      "learning_rate": 1.713580092718129e-06,
      "loss": 0.0751,
      "step": 9605
    },
    {
      "epoch": 4.15,
      "grad_norm": 0.6017299148369583,
      "learning_rate": 1.7051485194992335e-06,
      "loss": 0.0675,
      "step": 9610
    },
    {
      "epoch": 4.15,
      "grad_norm": 0.7850003228031432,
      "learning_rate": 1.6967358072040208e-06,
      "loss": 0.0701,
      "step": 9615
    },
    {
      "epoch": 4.16,
      "grad_norm": 0.6014246832157257,
      "learning_rate": 1.6883419749614137e-06,
      "loss": 0.0716,
      "step": 9620
    },
    {
      "epoch": 4.16,
      "grad_norm": 0.7364660205675014,
      "learning_rate": 1.6799670418573932e-06,
      "loss": 0.0715,
      "step": 9625
    },
    {
      "epoch": 4.16,
      "grad_norm": 0.7778637332406131,
      "learning_rate": 1.6716110269349782e-06,
      "loss": 0.0728,
      "step": 9630
    },
    {
      "epoch": 4.16,
      "grad_norm": 0.5496082761498826,
      "learning_rate": 1.6632739491941608e-06,
      "loss": 0.0851,
      "step": 9635
    },
    {
      "epoch": 4.16,
      "grad_norm": 1.039006198166973,
      "learning_rate": 1.654955827591881e-06,
      "loss": 0.0663,
      "step": 9640
    },
    {
      "epoch": 4.17,
      "grad_norm": 1.1424991869818257,
      "learning_rate": 1.6466566810419793e-06,
      "loss": 0.0778,
      "step": 9645
    },
    {
      "epoch": 4.17,
      "grad_norm": 0.6614813628662064,
      "learning_rate": 1.6383765284151398e-06,
      "loss": 0.0807,
      "step": 9650
    },
    {
      "epoch": 4.17,
      "grad_norm": 1.0752557823535098,
      "learning_rate": 1.6301153885388688e-06,
      "loss": 0.0752,
      "step": 9655
    },
    {
      "epoch": 4.17,
      "grad_norm": 1.158097069047512,
      "learning_rate": 1.6218732801974302e-06,
      "loss": 0.0829,
      "step": 9660
    },
    {
      "epoch": 4.17,
      "grad_norm": 0.81209934506329,
      "learning_rate": 1.6136502221318195e-06,
      "loss": 0.077,
      "step": 9665
    },
    {
      "epoch": 4.18,
      "grad_norm": 0.6199113304485403,
      "learning_rate": 1.6054462330397259e-06,
      "loss": 0.0724,
      "step": 9670
    },
    {
      "epoch": 4.18,
      "grad_norm": 0.6901257372883581,
      "learning_rate": 1.5972613315754583e-06,
      "loss": 0.0708,
      "step": 9675
    },
    {
      "epoch": 4.18,
      "grad_norm": 0.5410878140359482,
      "learning_rate": 1.5890955363499416e-06,
      "loss": 0.0643,
      "step": 9680
    },
    {
      "epoch": 4.18,
      "grad_norm": 0.5222408854828781,
      "learning_rate": 1.580948865930645e-06,
      "loss": 0.0689,
      "step": 9685
    },
    {
      "epoch": 4.19,
      "grad_norm": 1.17027325579557,
      "learning_rate": 1.572821338841558e-06,
      "loss": 0.0836,
      "step": 9690
    },
    {
      "epoch": 4.19,
      "grad_norm": 1.206300905777825,
      "learning_rate": 1.5647129735631416e-06,
      "loss": 0.071,
      "step": 9695
    },
    {
      "epoch": 4.19,
      "grad_norm": 0.6861332362617869,
      "learning_rate": 1.5566237885322843e-06,
      "loss": 0.0816,
      "step": 9700
    },
    {
      "epoch": 4.19,
      "grad_norm": 0.9367655165575982,
      "learning_rate": 1.5485538021422653e-06,
      "loss": 0.0861,
      "step": 9705
    },
    {
      "epoch": 4.19,
      "grad_norm": 0.6259566627418062,
      "learning_rate": 1.540503032742703e-06,
      "loss": 0.0645,
      "step": 9710
    },
    {
      "epoch": 4.2,
      "grad_norm": 0.8621509722069886,
      "learning_rate": 1.5324714986395284e-06,
      "loss": 0.0761,
      "step": 9715
    },
    {
      "epoch": 4.2,
      "grad_norm": 0.9362839555402197,
      "learning_rate": 1.5244592180949292e-06,
      "loss": 0.0759,
      "step": 9720
    },
    {
      "epoch": 4.2,
      "grad_norm": 0.8468366297743869,
      "learning_rate": 1.5164662093273174e-06,
      "loss": 0.0691,
      "step": 9725
    },
    {
      "epoch": 4.2,
      "grad_norm": 1.5679881546372574,
      "learning_rate": 1.5084924905112853e-06,
      "loss": 0.0762,
      "step": 9730
    },
    {
      "epoch": 4.21,
      "grad_norm": 0.8035954051894941,
      "learning_rate": 1.500538079777557e-06,
      "loss": 0.0709,
      "step": 9735
    },
    {
      "epoch": 4.21,
      "grad_norm": 0.7669725757041786,
      "learning_rate": 1.4926029952129607e-06,
      "loss": 0.0815,
      "step": 9740
    },
    {
      "epoch": 4.21,
      "grad_norm": 0.7855723793535851,
      "learning_rate": 1.4846872548603787e-06,
      "loss": 0.086,
      "step": 9745
    },
    {
      "epoch": 4.21,
      "grad_norm": 0.2646335220272136,
      "learning_rate": 1.4767908767187067e-06,
      "loss": 0.0582,
      "step": 9750
    },
    {
      "epoch": 4.21,
      "grad_norm": 0.8482657133461957,
      "learning_rate": 1.468913878742818e-06,
      "loss": 0.0664,
      "step": 9755
    },
    {
      "epoch": 4.22,
      "grad_norm": 0.7945340610515387,
      "learning_rate": 1.4610562788435123e-06,
      "loss": 0.067,
      "step": 9760
    },
    {
      "epoch": 4.22,
      "grad_norm": 0.6109194241490044,
      "learning_rate": 1.4532180948874874e-06,
      "loss": 0.0693,
      "step": 9765
    },
    {
      "epoch": 4.22,
      "grad_norm": 0.9556555429097761,
      "learning_rate": 1.4453993446972925e-06,
      "loss": 0.0762,
      "step": 9770
    },
    {
      "epoch": 4.22,
      "grad_norm": 0.5329908247970897,
      "learning_rate": 1.437600046051285e-06,
      "loss": 0.0714,
      "step": 9775
    },
    {
      "epoch": 4.22,
      "grad_norm": 0.8007351984555957,
      "learning_rate": 1.4298202166836006e-06,
      "loss": 0.0677,
      "step": 9780
    },
    {
      "epoch": 4.23,
      "grad_norm": 0.9599502718080456,
      "learning_rate": 1.422059874284092e-06,
      "loss": 0.0798,
      "step": 9785
    },
    {
      "epoch": 4.23,
      "grad_norm": 0.8461958157094835,
      "learning_rate": 1.414319036498315e-06,
      "loss": 0.0728,
      "step": 9790
    },
    {
      "epoch": 4.23,
      "grad_norm": 0.8684313938243562,
      "learning_rate": 1.4065977209274706e-06,
      "loss": 0.0694,
      "step": 9795
    },
    {
      "epoch": 4.23,
      "grad_norm": 0.45658884011612155,
      "learning_rate": 1.398895945128369e-06,
      "loss": 0.0736,
      "step": 9800
    },
    {
      "epoch": 4.24,
      "grad_norm": 0.4987698551679359,
      "learning_rate": 1.3912137266133962e-06,
      "loss": 0.0709,
      "step": 9805
    },
    {
      "epoch": 4.24,
      "grad_norm": 0.7190456258252003,
      "learning_rate": 1.3835510828504573e-06,
      "loss": 0.0712,
      "step": 9810
    },
    {
      "epoch": 4.24,
      "grad_norm": 0.7903660619505879,
      "learning_rate": 1.3759080312629602e-06,
      "loss": 0.0735,
      "step": 9815
    },
    {
      "epoch": 4.24,
      "grad_norm": 0.4239086367212096,
      "learning_rate": 1.3682845892297568e-06,
      "loss": 0.0645,
      "step": 9820
    },
    {
      "epoch": 4.24,
      "grad_norm": 0.7296580218816523,
      "learning_rate": 1.3606807740851136e-06,
      "loss": 0.0686,
      "step": 9825
    },
    {
      "epoch": 4.25,
      "grad_norm": 1.0482753668393678,
      "learning_rate": 1.3530966031186698e-06,
      "loss": 0.0774,
      "step": 9830
    },
    {
      "epoch": 4.25,
      "grad_norm": 3.23798164205354,
      "learning_rate": 1.3455320935753924e-06,
      "loss": 0.0751,
      "step": 9835
    },
    {
      "epoch": 4.25,
      "grad_norm": 0.7752045430553508,
      "learning_rate": 1.337987262655548e-06,
      "loss": 0.0716,
      "step": 9840
    },
    {
      "epoch": 4.25,
      "grad_norm": 1.0027667237986686,
      "learning_rate": 1.3304621275146546e-06,
      "loss": 0.0724,
      "step": 9845
    },
    {
      "epoch": 4.25,
      "grad_norm": 0.5926695037422461,
      "learning_rate": 1.3229567052634461e-06,
      "loss": 0.0601,
      "step": 9850
    },
    {
      "epoch": 4.26,
      "grad_norm": 0.6538303894841858,
      "learning_rate": 1.3154710129678372e-06,
      "loss": 0.0612,
      "step": 9855
    },
    {
      "epoch": 4.26,
      "grad_norm": 0.4649486420264303,
      "learning_rate": 1.308005067648872e-06,
      "loss": 0.0684,
      "step": 9860
    },
    {
      "epoch": 4.26,
      "grad_norm": 0.8426888212283112,
      "learning_rate": 1.300558886282699e-06,
      "loss": 0.0833,
      "step": 9865
    },
    {
      "epoch": 4.26,
      "grad_norm": 0.6257879916376212,
      "learning_rate": 1.2931324858005268e-06,
      "loss": 0.0634,
      "step": 9870
    },
    {
      "epoch": 4.27,
      "grad_norm": 0.7619274786441721,
      "learning_rate": 1.2857258830885865e-06,
      "loss": 0.0739,
      "step": 9875
    },
    {
      "epoch": 4.27,
      "grad_norm": 0.714597826232532,
      "learning_rate": 1.2783390949880947e-06,
      "loss": 0.0616,
      "step": 9880
    },
    {
      "epoch": 4.27,
      "grad_norm": 0.7202242150486695,
      "learning_rate": 1.2709721382952055e-06,
      "loss": 0.0626,
      "step": 9885
    },
    {
      "epoch": 4.27,
      "grad_norm": 0.41782547564555067,
      "learning_rate": 1.2636250297609876e-06,
      "loss": 0.0671,
      "step": 9890
    },
    {
      "epoch": 4.27,
      "grad_norm": 0.46816271815456095,
      "learning_rate": 1.2562977860913762e-06,
      "loss": 0.0689,
      "step": 9895
    },
    {
      "epoch": 4.28,
      "grad_norm": 1.0028729164346737,
      "learning_rate": 1.2489904239471384e-06,
      "loss": 0.0683,
      "step": 9900
    },
    {
      "epoch": 4.28,
      "grad_norm": 0.8422597030333709,
      "learning_rate": 1.2417029599438356e-06,
      "loss": 0.0718,
      "step": 9905
    },
    {
      "epoch": 4.28,
      "grad_norm": 1.1068122147961652,
      "learning_rate": 1.2344354106517786e-06,
      "loss": 0.0738,
      "step": 9910
    },
    {
      "epoch": 4.28,
      "grad_norm": 0.5512640293205736,
      "learning_rate": 1.2271877925960019e-06,
      "loss": 0.0708,
      "step": 9915
    },
    {
      "epoch": 4.29,
      "grad_norm": 0.6105033804061096,
      "learning_rate": 1.2199601222562196e-06,
      "loss": 0.0619,
      "step": 9920
    },
    {
      "epoch": 4.29,
      "grad_norm": 0.4856188866921057,
      "learning_rate": 1.2127524160667869e-06,
      "loss": 0.0685,
      "step": 9925
    },
    {
      "epoch": 4.29,
      "grad_norm": 0.9256462061093675,
      "learning_rate": 1.2055646904166652e-06,
      "loss": 0.0604,
      "step": 9930
    },
    {
      "epoch": 4.29,
      "grad_norm": 0.7952559019318669,
      "learning_rate": 1.1983969616493818e-06,
      "loss": 0.0677,
      "step": 9935
    },
    {
      "epoch": 4.29,
      "grad_norm": 0.5933272207096911,
      "learning_rate": 1.1912492460629964e-06,
      "loss": 0.0586,
      "step": 9940
    },
    {
      "epoch": 4.3,
      "grad_norm": 0.7335180054141032,
      "learning_rate": 1.1841215599100643e-06,
      "loss": 0.0896,
      "step": 9945
    },
    {
      "epoch": 4.3,
      "grad_norm": 0.5807708318876802,
      "learning_rate": 1.1770139193975938e-06,
      "loss": 0.0669,
      "step": 9950
    },
    {
      "epoch": 4.3,
      "grad_norm": 3.2465728542332126,
      "learning_rate": 1.1699263406870198e-06,
      "loss": 0.0747,
      "step": 9955
    },
    {
      "epoch": 4.3,
      "grad_norm": 0.9376095352267071,
      "learning_rate": 1.1628588398941498e-06,
      "loss": 0.0709,
      "step": 9960
    },
    {
      "epoch": 4.3,
      "grad_norm": 0.8747907603190435,
      "learning_rate": 1.155811433089148e-06,
      "loss": 0.0723,
      "step": 9965
    },
    {
      "epoch": 4.31,
      "grad_norm": 0.720029499713049,
      "learning_rate": 1.1487841362964835e-06,
      "loss": 0.0683,
      "step": 9970
    },
    {
      "epoch": 4.31,
      "grad_norm": 0.7103885214539689,
      "learning_rate": 1.1417769654949006e-06,
      "loss": 0.0711,
      "step": 9975
    },
    {
      "epoch": 4.31,
      "grad_norm": 0.5180032364147855,
      "learning_rate": 1.1347899366173841e-06,
      "loss": 0.0582,
      "step": 9980
    },
    {
      "epoch": 4.31,
      "grad_norm": 0.8132409849677699,
      "learning_rate": 1.1278230655511103e-06,
      "loss": 0.0675,
      "step": 9985
    },
    {
      "epoch": 4.32,
      "grad_norm": 0.24139115913417963,
      "learning_rate": 1.1208763681374292e-06,
      "loss": 0.0637,
      "step": 9990
    },
    {
      "epoch": 4.32,
      "grad_norm": 1.0353499843369223,
      "learning_rate": 1.1139498601718169e-06,
      "loss": 0.0717,
      "step": 9995
    },
    {
      "epoch": 4.32,
      "grad_norm": 1.2256119940521908,
      "learning_rate": 1.107043557403843e-06,
      "loss": 0.0805,
      "step": 10000
    },
    {
      "epoch": 4.32,
      "grad_norm": 0.5239300408769509,
      "learning_rate": 1.1001574755371359e-06,
      "loss": 0.0775,
      "step": 10005
    },
    {
      "epoch": 4.32,
      "grad_norm": 0.7855124071489255,
      "learning_rate": 1.0932916302293406e-06,
      "loss": 0.065,
      "step": 10010
    },
    {
      "epoch": 4.33,
      "grad_norm": 0.704823659963824,
      "learning_rate": 1.0864460370920904e-06,
      "loss": 0.0712,
      "step": 10015
    },
    {
      "epoch": 4.33,
      "grad_norm": 0.5511957029973897,
      "learning_rate": 1.079620711690973e-06,
      "loss": 0.0667,
      "step": 10020
    },
    {
      "epoch": 4.33,
      "grad_norm": 0.4287620497702917,
      "learning_rate": 1.072815669545485e-06,
      "loss": 0.0638,
      "step": 10025
    },
    {
      "epoch": 4.33,
      "grad_norm": 0.8820143962824251,
      "learning_rate": 1.066030926129008e-06,
      "loss": 0.0702,
      "step": 10030
    },
    {
      "epoch": 4.33,
      "grad_norm": 0.3713151760478781,
      "learning_rate": 1.0592664968687627e-06,
      "loss": 0.0737,
      "step": 10035
    },
    {
      "epoch": 4.34,
      "grad_norm": 0.9403215381824361,
      "learning_rate": 1.0525223971457821e-06,
      "loss": 0.0643,
      "step": 10040
    },
    {
      "epoch": 4.34,
      "grad_norm": 0.7591698484596258,
      "learning_rate": 1.0457986422948752e-06,
      "loss": 0.0671,
      "step": 10045
    },
    {
      "epoch": 4.34,
      "grad_norm": 0.6323614044936294,
      "learning_rate": 1.039095247604589e-06,
      "loss": 0.0681,
      "step": 10050
    },
    {
      "epoch": 4.34,
      "grad_norm": 0.8848119028119328,
      "learning_rate": 1.0324122283171767e-06,
      "loss": 0.0765,
      "step": 10055
    },
    {
      "epoch": 4.35,
      "grad_norm": 0.8298278075654341,
      "learning_rate": 1.025749599628556e-06,
      "loss": 0.0641,
      "step": 10060
    },
    {
      "epoch": 4.35,
      "grad_norm": 0.5541866046706723,
      "learning_rate": 1.0191073766882875e-06,
      "loss": 0.0566,
      "step": 10065
    },
    {
      "epoch": 4.35,
      "grad_norm": 0.6386188598418999,
      "learning_rate": 1.012485574599531e-06,
      "loss": 0.0654,
      "step": 10070
    },
    {
      "epoch": 4.35,
      "grad_norm": 0.4624781345723369,
      "learning_rate": 1.005884208419009e-06,
      "loss": 0.069,
      "step": 10075
    },
    {
      "epoch": 4.35,
      "grad_norm": 0.8121132787817534,
      "learning_rate": 9.993032931569858e-07,
      "loss": 0.0652,
      "step": 10080
    },
    {
      "epoch": 4.36,
      "grad_norm": 0.9607080496359577,
      "learning_rate": 9.927428437772113e-07,
      "loss": 0.0673,
      "step": 10085
    },
    {
      "epoch": 4.36,
      "grad_norm": 0.627392336343592,
      "learning_rate": 9.862028751969099e-07,
      "loss": 0.073,
      "step": 10090
    },
    {
      "epoch": 4.36,
      "grad_norm": 0.7200512938130613,
      "learning_rate": 9.79683402286734e-07,
      "loss": 0.0661,
      "step": 10095
    },
    {
      "epoch": 4.36,
      "grad_norm": 0.5809004286224504,
      "learning_rate": 9.731844398707313e-07,
      "loss": 0.0645,
      "step": 10100
    },
    {
      "epoch": 4.37,
      "grad_norm": 0.4701499769785015,
      "learning_rate": 9.667060027263152e-07,
      "loss": 0.0624,
      "step": 10105
    },
    {
      "epoch": 4.37,
      "grad_norm": 1.0412408044015375,
      "learning_rate": 9.602481055842228e-07,
      "loss": 0.0706,
      "step": 10110
    },
    {
      "epoch": 4.37,
      "grad_norm": 0.7109713416597109,
      "learning_rate": 9.538107631284943e-07,
      "loss": 0.0654,
      "step": 10115
    },
    {
      "epoch": 4.37,
      "grad_norm": 0.4148240501136039,
      "learning_rate": 9.47393989996428e-07,
      "loss": 0.0673,
      "step": 10120
    },
    {
      "epoch": 4.37,
      "grad_norm": 0.49130418610183135,
      "learning_rate": 9.409978007785514e-07,
      "loss": 0.0661,
      "step": 10125
    },
    {
      "epoch": 4.38,
      "grad_norm": 0.4060476502578717,
      "learning_rate": 9.346222100185931e-07,
      "loss": 0.0564,
      "step": 10130
    },
    {
      "epoch": 4.38,
      "grad_norm": 0.7474394992207792,
      "learning_rate": 9.282672322134367e-07,
      "loss": 0.0599,
      "step": 10135
    },
    {
      "epoch": 4.38,
      "grad_norm": 0.42235346298794035,
      "learning_rate": 9.219328818131013e-07,
      "loss": 0.0652,
      "step": 10140
    },
    {
      "epoch": 4.38,
      "grad_norm": 0.6853745391337568,
      "learning_rate": 9.15619173220702e-07,
      "loss": 0.0674,
      "step": 10145
    },
    {
      "epoch": 4.38,
      "grad_norm": 0.4667409504567176,
      "learning_rate": 9.093261207924175e-07,
      "loss": 0.0695,
      "step": 10150
    },
    {
      "epoch": 4.39,
      "grad_norm": 0.5579367649666931,
      "learning_rate": 9.030537388374627e-07,
      "loss": 0.0659,
      "step": 10155
    },
    {
      "epoch": 4.39,
      "grad_norm": 0.6908953456786018,
      "learning_rate": 8.968020416180434e-07,
      "loss": 0.0659,
      "step": 10160
    },
    {
      "epoch": 4.39,
      "grad_norm": 0.9183434314862593,
      "learning_rate": 8.905710433493375e-07,
      "loss": 0.0778,
      "step": 10165
    },
    {
      "epoch": 4.39,
      "grad_norm": 0.4109391359251177,
      "learning_rate": 8.843607581994585e-07,
      "loss": 0.0722,
      "step": 10170
    },
    {
      "epoch": 4.4,
      "grad_norm": 1.4316872445801794,
      "learning_rate": 8.781712002894205e-07,
      "loss": 0.0696,
      "step": 10175
    },
    {
      "epoch": 4.4,
      "grad_norm": 0.5923901753548628,
      "learning_rate": 8.720023836931079e-07,
      "loss": 0.067,
      "step": 10180
    },
    {
      "epoch": 4.4,
      "grad_norm": 0.5945036394746386,
      "learning_rate": 8.658543224372407e-07,
      "loss": 0.0605,
      "step": 10185
    },
    {
      "epoch": 4.4,
      "grad_norm": 0.7776265459380092,
      "learning_rate": 8.597270305013483e-07,
      "loss": 0.0739,
      "step": 10190
    },
    {
      "epoch": 4.4,
      "grad_norm": 0.5753102906059375,
      "learning_rate": 8.536205218177351e-07,
      "loss": 0.0649,
      "step": 10195
    },
    {
      "epoch": 4.41,
      "grad_norm": 0.7139209337155082,
      "learning_rate": 8.475348102714464e-07,
      "loss": 0.0743,
      "step": 10200
    },
    {
      "epoch": 4.41,
      "grad_norm": 0.3196259522914314,
      "learning_rate": 8.414699097002422e-07,
      "loss": 0.0727,
      "step": 10205
    },
    {
      "epoch": 4.41,
      "grad_norm": 0.41683002866071117,
      "learning_rate": 8.354258338945542e-07,
      "loss": 0.0588,
      "step": 10210
    },
    {
      "epoch": 4.41,
      "grad_norm": 0.850604799529318,
      "learning_rate": 8.294025965974717e-07,
      "loss": 0.0652,
      "step": 10215
    },
    {
      "epoch": 4.41,
      "grad_norm": 0.5300497819554433,
      "learning_rate": 8.234002115046957e-07,
      "loss": 0.0647,
      "step": 10220
    },
    {
      "epoch": 4.42,
      "grad_norm": 0.5035451796576669,
      "learning_rate": 8.174186922645155e-07,
      "loss": 0.0696,
      "step": 10225
    },
    {
      "epoch": 4.42,
      "grad_norm": 0.380831280980267,
      "learning_rate": 8.11458052477776e-07,
      "loss": 0.0647,
      "step": 10230
    },
    {
      "epoch": 4.42,
      "grad_norm": 0.7292244058285478,
      "learning_rate": 8.055183056978421e-07,
      "loss": 0.0664,
      "step": 10235
    },
    {
      "epoch": 4.42,
      "grad_norm": 0.8310509794310577,
      "learning_rate": 7.99599465430575e-07,
      "loss": 0.0673,
      "step": 10240
    },
    {
      "epoch": 4.43,
      "grad_norm": 1.0816929263590291,
      "learning_rate": 7.937015451342999e-07,
      "loss": 0.0766,
      "step": 10245
    },
    {
      "epoch": 4.43,
      "grad_norm": 0.683497006037796,
      "learning_rate": 7.878245582197718e-07,
      "loss": 0.0615,
      "step": 10250
    },
    {
      "epoch": 4.43,
      "grad_norm": 0.6580740484054255,
      "learning_rate": 7.819685180501502e-07,
      "loss": 0.0583,
      "step": 10255
    },
    {
      "epoch": 4.43,
      "grad_norm": 0.6176177788270545,
      "learning_rate": 7.761334379409569e-07,
      "loss": 0.0602,
      "step": 10260
    },
    {
      "epoch": 4.43,
      "grad_norm": 0.5662803463718628,
      "learning_rate": 7.703193311600654e-07,
      "loss": 0.0712,
      "step": 10265
    },
    {
      "epoch": 4.44,
      "grad_norm": 0.6550414709715258,
      "learning_rate": 7.64526210927653e-07,
      "loss": 0.0621,
      "step": 10270
    },
    {
      "epoch": 4.44,
      "grad_norm": 0.342545878900588,
      "learning_rate": 7.587540904161794e-07,
      "loss": 0.0657,
      "step": 10275
    },
    {
      "epoch": 4.44,
      "grad_norm": 0.5255994175005089,
      "learning_rate": 7.530029827503593e-07,
      "loss": 0.0672,
      "step": 10280
    },
    {
      "epoch": 4.44,
      "grad_norm": 0.4457783191130157,
      "learning_rate": 7.472729010071178e-07,
      "loss": 0.061,
      "step": 10285
    },
    {
      "epoch": 4.44,
      "grad_norm": 0.41693577159351586,
      "learning_rate": 7.415638582155804e-07,
      "loss": 0.0616,
      "step": 10290
    },
    {
      "epoch": 4.45,
      "grad_norm": 0.5062588412348282,
      "learning_rate": 7.35875867357031e-07,
      "loss": 0.07,
      "step": 10295
    },
    {
      "epoch": 4.45,
      "grad_norm": 0.7012064353284458,
      "learning_rate": 7.302089413648838e-07,
      "loss": 0.0662,
      "step": 10300
    },
    {
      "epoch": 4.45,
      "grad_norm": 0.49602914879558746,
      "learning_rate": 7.245630931246606e-07,
      "loss": 0.0709,
      "step": 10305
    },
    {
      "epoch": 4.45,
      "grad_norm": 0.877101152761563,
      "learning_rate": 7.189383354739476e-07,
      "loss": 0.0746,
      "step": 10310
    },
    {
      "epoch": 4.46,
      "grad_norm": 0.6109331747310377,
      "learning_rate": 7.133346812023823e-07,
      "loss": 0.0619,
      "step": 10315
    },
    {
      "epoch": 4.46,
      "grad_norm": 0.6710843159924726,
      "learning_rate": 7.077521430516143e-07,
      "loss": 0.0635,
      "step": 10320
    },
    {
      "epoch": 4.46,
      "grad_norm": 1.4984514944289056,
      "learning_rate": 7.0219073371528e-07,
      "loss": 0.0677,
      "step": 10325
    },
    {
      "epoch": 4.46,
      "grad_norm": 0.65040299798008,
      "learning_rate": 6.96650465838975e-07,
      "loss": 0.0699,
      "step": 10330
    },
    {
      "epoch": 4.46,
      "grad_norm": 0.6274345261918733,
      "learning_rate": 6.911313520202167e-07,
      "loss": 0.0735,
      "step": 10335
    },
    {
      "epoch": 4.47,
      "grad_norm": 1.0263488571321568,
      "learning_rate": 6.85633404808429e-07,
      "loss": 0.0717,
      "step": 10340
    },
    {
      "epoch": 4.47,
      "grad_norm": 0.5816865823316484,
      "learning_rate": 6.801566367049018e-07,
      "loss": 0.0673,
      "step": 10345
    },
    {
      "epoch": 4.47,
      "grad_norm": 0.48045338154681266,
      "learning_rate": 6.747010601627724e-07,
      "loss": 0.0658,
      "step": 10350
    },
    {
      "epoch": 4.47,
      "grad_norm": 0.6100745958999338,
      "learning_rate": 6.692666875869902e-07,
      "loss": 0.0762,
      "step": 10355
    },
    {
      "epoch": 4.48,
      "grad_norm": 0.6287186987071508,
      "learning_rate": 6.638535313342886e-07,
      "loss": 0.0631,
      "step": 10360
    },
    {
      "epoch": 4.48,
      "grad_norm": 0.5641903490443426,
      "learning_rate": 6.584616037131631e-07,
      "loss": 0.0686,
      "step": 10365
    },
    {
      "epoch": 4.48,
      "grad_norm": 0.8508675496237825,
      "learning_rate": 6.530909169838362e-07,
      "loss": 0.0599,
      "step": 10370
    },
    {
      "epoch": 4.48,
      "grad_norm": 0.8371094642421063,
      "learning_rate": 6.477414833582363e-07,
      "loss": 0.0668,
      "step": 10375
    },
    {
      "epoch": 4.48,
      "grad_norm": 0.6556091987202345,
      "learning_rate": 6.424133149999611e-07,
      "loss": 0.0662,
      "step": 10380
    },
    {
      "epoch": 4.49,
      "grad_norm": 0.5788970639761051,
      "learning_rate": 6.371064240242586e-07,
      "loss": 0.0737,
      "step": 10385
    },
    {
      "epoch": 4.49,
      "grad_norm": 0.9576674573709517,
      "learning_rate": 6.31820822497996e-07,
      "loss": 0.0725,
      "step": 10390
    },
    {
      "epoch": 4.49,
      "grad_norm": 1.2339994955128544,
      "learning_rate": 6.265565224396319e-07,
      "loss": 0.0633,
      "step": 10395
    },
    {
      "epoch": 4.49,
      "grad_norm": 0.7727346443142231,
      "learning_rate": 6.21313535819189e-07,
      "loss": 0.0599,
      "step": 10400
    },
    {
      "epoch": 4.49,
      "grad_norm": 0.5883563416250307,
      "learning_rate": 6.160918745582268e-07,
      "loss": 0.0587,
      "step": 10405
    },
    {
      "epoch": 4.5,
      "grad_norm": 0.6543057962328555,
      "learning_rate": 6.108915505298141e-07,
      "loss": 0.065,
      "step": 10410
    },
    {
      "epoch": 4.5,
      "grad_norm": 0.45099270153584864,
      "learning_rate": 6.057125755585069e-07,
      "loss": 0.0617,
      "step": 10415
    },
    {
      "epoch": 4.5,
      "grad_norm": 0.6813172291762168,
      "learning_rate": 6.005549614203144e-07,
      "loss": 0.0711,
      "step": 10420
    },
    {
      "epoch": 4.5,
      "grad_norm": 0.42037421073670783,
      "learning_rate": 5.954187198426764e-07,
      "loss": 0.0664,
      "step": 10425
    },
    {
      "epoch": 4.51,
      "grad_norm": 0.7336243689992468,
      "learning_rate": 5.903038625044332e-07,
      "loss": 0.0708,
      "step": 10430
    },
    {
      "epoch": 4.51,
      "grad_norm": 0.8346702406824394,
      "learning_rate": 5.852104010358039e-07,
      "loss": 0.0615,
      "step": 10435
    },
    {
      "epoch": 4.51,
      "grad_norm": 0.8369876118704074,
      "learning_rate": 5.801383470183564e-07,
      "loss": 0.0607,
      "step": 10440
    },
    {
      "epoch": 4.51,
      "grad_norm": 0.7884787540617048,
      "learning_rate": 5.75087711984983e-07,
      "loss": 0.0612,
      "step": 10445
    },
    {
      "epoch": 4.51,
      "grad_norm": 0.9952443752024916,
      "learning_rate": 5.700585074198739e-07,
      "loss": 0.0622,
      "step": 10450
    },
    {
      "epoch": 4.52,
      "grad_norm": 0.7486367420029569,
      "learning_rate": 5.650507447584863e-07,
      "loss": 0.0682,
      "step": 10455
    },
    {
      "epoch": 4.52,
      "grad_norm": 0.5399364925124402,
      "learning_rate": 5.600644353875273e-07,
      "loss": 0.0704,
      "step": 10460
    },
    {
      "epoch": 4.52,
      "grad_norm": 0.7213995163485463,
      "learning_rate": 5.550995906449208e-07,
      "loss": 0.064,
      "step": 10465
    },
    {
      "epoch": 4.52,
      "grad_norm": 0.5658492332641553,
      "learning_rate": 5.501562218197853e-07,
      "loss": 0.0629,
      "step": 10470
    },
    {
      "epoch": 4.52,
      "grad_norm": 0.3714861481274943,
      "learning_rate": 5.452343401524085e-07,
      "loss": 0.0627,
      "step": 10475
    },
    {
      "epoch": 4.53,
      "grad_norm": 1.0177022722601001,
      "learning_rate": 5.403339568342158e-07,
      "loss": 0.0729,
      "step": 10480
    },
    {
      "epoch": 4.53,
      "grad_norm": 0.6677804981240922,
      "learning_rate": 5.354550830077521e-07,
      "loss": 0.0543,
      "step": 10485
    },
    {
      "epoch": 4.53,
      "grad_norm": 0.3341148617290627,
      "learning_rate": 5.305977297666553e-07,
      "loss": 0.0636,
      "step": 10490
    },
    {
      "epoch": 4.53,
      "grad_norm": 0.5413680488310173,
      "learning_rate": 5.257619081556276e-07,
      "loss": 0.0613,
      "step": 10495
    },
    {
      "epoch": 4.54,
      "grad_norm": 0.6914227311788763,
      "learning_rate": 5.209476291704152e-07,
      "loss": 0.0593,
      "step": 10500
    },
    {
      "epoch": 4.54,
      "grad_norm": 0.6314441804068395,
      "learning_rate": 5.16154903757774e-07,
      "loss": 0.0682,
      "step": 10505
    },
    {
      "epoch": 4.54,
      "grad_norm": 0.5451852735147008,
      "learning_rate": 5.113837428154567e-07,
      "loss": 0.0585,
      "step": 10510
    },
    {
      "epoch": 4.54,
      "grad_norm": 0.759423987615597,
      "learning_rate": 5.066341571921818e-07,
      "loss": 0.0654,
      "step": 10515
    },
    {
      "epoch": 4.54,
      "grad_norm": 0.3161641862772605,
      "learning_rate": 5.019061576876094e-07,
      "loss": 0.0651,
      "step": 10520
    },
    {
      "epoch": 4.55,
      "grad_norm": 0.7204327028723595,
      "learning_rate": 4.971997550523156e-07,
      "loss": 0.0604,
      "step": 10525
    },
    {
      "epoch": 4.55,
      "grad_norm": 0.6399016153777878,
      "learning_rate": 4.925149599877699e-07,
      "loss": 0.0653,
      "step": 10530
    },
    {
      "epoch": 4.55,
      "grad_norm": 0.5888152115518005,
      "learning_rate": 4.878517831463092e-07,
      "loss": 0.0692,
      "step": 10535
    },
    {
      "epoch": 4.55,
      "grad_norm": 1.1000748717903832,
      "learning_rate": 4.832102351311175e-07,
      "loss": 0.0775,
      "step": 10540
    },
    {
      "epoch": 4.56,
      "grad_norm": 0.833611260996263,
      "learning_rate": 4.785903264961967e-07,
      "loss": 0.0654,
      "step": 10545
    },
    {
      "epoch": 4.56,
      "grad_norm": 0.41303077294718554,
      "learning_rate": 4.739920677463461e-07,
      "loss": 0.0688,
      "step": 10550
    },
    {
      "epoch": 4.56,
      "grad_norm": 0.53144152081788,
      "learning_rate": 4.6941546933713553e-07,
      "loss": 0.0535,
      "step": 10555
    },
    {
      "epoch": 4.56,
      "grad_norm": 0.3593735462606911,
      "learning_rate": 4.6486054167488414e-07,
      "loss": 0.0688,
      "step": 10560
    },
    {
      "epoch": 4.56,
      "grad_norm": 0.5132813504324446,
      "learning_rate": 4.603272951166371e-07,
      "loss": 0.0598,
      "step": 10565
    },
    {
      "epoch": 4.57,
      "grad_norm": 0.37892051369945495,
      "learning_rate": 4.558157399701379e-07,
      "loss": 0.0631,
      "step": 10570
    },
    {
      "epoch": 4.57,
      "grad_norm": 0.7363528376552387,
      "learning_rate": 4.5132588649381413e-07,
      "loss": 0.0606,
      "step": 10575
    },
    {
      "epoch": 4.57,
      "grad_norm": 0.628861659000855,
      "learning_rate": 4.4685774489673926e-07,
      "loss": 0.0652,
      "step": 10580
    },
    {
      "epoch": 4.57,
      "grad_norm": 0.4642134473139565,
      "learning_rate": 4.4241132533862306e-07,
      "loss": 0.0599,
      "step": 10585
    },
    {
      "epoch": 4.57,
      "grad_norm": 0.4240911790644064,
      "learning_rate": 4.3798663792978367e-07,
      "loss": 0.0653,
      "step": 10590
    },
    {
      "epoch": 4.58,
      "grad_norm": 0.5919110283678517,
      "learning_rate": 4.335836927311232e-07,
      "loss": 0.0676,
      "step": 10595
    },
    {
      "epoch": 4.58,
      "grad_norm": 0.6943140171165241,
      "learning_rate": 4.292024997541078e-07,
      "loss": 0.0603,
      "step": 10600
    },
    {
      "epoch": 4.58,
      "grad_norm": 0.6557567667966683,
      "learning_rate": 4.248430689607397e-07,
      "loss": 0.0611,
      "step": 10605
    },
    {
      "epoch": 4.58,
      "grad_norm": 0.9732091473705298,
      "learning_rate": 4.205054102635409e-07,
      "loss": 0.067,
      "step": 10610
    },
    {
      "epoch": 4.59,
      "grad_norm": 0.6632492498420535,
      "learning_rate": 4.1618953352552616e-07,
      "loss": 0.0581,
      "step": 10615
    },
    {
      "epoch": 4.59,
      "grad_norm": 0.5304776663139763,
      "learning_rate": 4.118954485601845e-07,
      "loss": 0.059,
      "step": 10620
    },
    {
      "epoch": 4.59,
      "grad_norm": 0.5358707936464782,
      "learning_rate": 4.0762316513145436e-07,
      "loss": 0.0514,
      "step": 10625
    },
    {
      "epoch": 4.59,
      "grad_norm": 0.5362570016989209,
      "learning_rate": 4.033726929536974e-07,
      "loss": 0.0593,
      "step": 10630
    },
    {
      "epoch": 4.59,
      "grad_norm": 0.5513726440823109,
      "learning_rate": 3.991440416916836e-07,
      "loss": 0.0649,
      "step": 10635
    },
    {
      "epoch": 4.6,
      "grad_norm": 0.6461092576242113,
      "learning_rate": 3.9493722096056844e-07,
      "loss": 0.0601,
      "step": 10640
    },
    {
      "epoch": 4.6,
      "grad_norm": 0.38614596195664247,
      "learning_rate": 3.907522403258657e-07,
      "loss": 0.0544,
      "step": 10645
    },
    {
      "epoch": 4.6,
      "grad_norm": 0.9317175798093095,
      "learning_rate": 3.865891093034313e-07,
      "loss": 0.0807,
      "step": 10650
    },
    {
      "epoch": 4.6,
      "grad_norm": 0.4708803679522015,
      "learning_rate": 3.824478373594342e-07,
      "loss": 0.0625,
      "step": 10655
    },
    {
      "epoch": 4.6,
      "grad_norm": 0.5846023860363823,
      "learning_rate": 3.7832843391034766e-07,
      "loss": 0.0617,
      "step": 10660
    },
    {
      "epoch": 4.61,
      "grad_norm": 0.8074441691925042,
      "learning_rate": 3.742309083229134e-07,
      "loss": 0.0756,
      "step": 10665
    },
    {
      "epoch": 4.61,
      "grad_norm": 0.7419611525652908,
      "learning_rate": 3.7015526991413195e-07,
      "loss": 0.0681,
      "step": 10670
    },
    {
      "epoch": 4.61,
      "grad_norm": 0.49521738757182165,
      "learning_rate": 3.661015279512337e-07,
      "loss": 0.059,
      "step": 10675
    },
    {
      "epoch": 4.61,
      "grad_norm": 0.6363196159249647,
      "learning_rate": 3.620696916516597e-07,
      "loss": 0.0675,
      "step": 10680
    },
    {
      "epoch": 4.62,
      "grad_norm": 0.648756517846174,
      "learning_rate": 3.580597701830446e-07,
      "loss": 0.0564,
      "step": 10685
    },
    {
      "epoch": 4.62,
      "grad_norm": 0.22145343101134532,
      "learning_rate": 3.540717726631915e-07,
      "loss": 0.0657,
      "step": 10690
    },
    {
      "epoch": 4.62,
      "grad_norm": 0.745216261651182,
      "learning_rate": 3.5010570816005227e-07,
      "loss": 0.0617,
      "step": 10695
    },
    {
      "epoch": 4.62,
      "grad_norm": 0.5619150318693038,
      "learning_rate": 3.4616158569170887e-07,
      "loss": 0.0639,
      "step": 10700
    },
    {
      "epoch": 4.62,
      "grad_norm": 0.9913195001224414,
      "learning_rate": 3.4223941422634856e-07,
      "loss": 0.0707,
      "step": 10705
    },
    {
      "epoch": 4.63,
      "grad_norm": 0.9018805551119227,
      "learning_rate": 3.383392026822463e-07,
      "loss": 0.0768,
      "step": 10710
    },
    {
      "epoch": 4.63,
      "grad_norm": 0.5561922448862705,
      "learning_rate": 3.344609599277482e-07,
      "loss": 0.0529,
      "step": 10715
    },
    {
      "epoch": 4.63,
      "grad_norm": 0.6365875808087165,
      "learning_rate": 3.306046947812447e-07,
      "loss": 0.0689,
      "step": 10720
    },
    {
      "epoch": 4.63,
      "grad_norm": 0.7609070256401496,
      "learning_rate": 3.267704160111529e-07,
      "loss": 0.0623,
      "step": 10725
    },
    {
      "epoch": 4.63,
      "grad_norm": 0.9504410302089583,
      "learning_rate": 3.2295813233589544e-07,
      "loss": 0.0588,
      "step": 10730
    },
    {
      "epoch": 4.64,
      "grad_norm": 0.7288968981026365,
      "learning_rate": 3.191678524238884e-07,
      "loss": 0.0659,
      "step": 10735
    },
    {
      "epoch": 4.64,
      "grad_norm": 0.7518592308885353,
      "learning_rate": 3.1539958489350875e-07,
      "loss": 0.0751,
      "step": 10740
    },
    {
      "epoch": 4.64,
      "grad_norm": 0.24201120306356882,
      "learning_rate": 3.1165333831308487e-07,
      "loss": 0.0592,
      "step": 10745
    },
    {
      "epoch": 4.64,
      "grad_norm": 0.5676494509868415,
      "learning_rate": 3.0792912120087394e-07,
      "loss": 0.0596,
      "step": 10750
    },
    {
      "epoch": 4.65,
      "grad_norm": 0.47676863842198597,
      "learning_rate": 3.04226942025041e-07,
      "loss": 0.0588,
      "step": 10755
    },
    {
      "epoch": 4.65,
      "grad_norm": 0.48666045227889837,
      "learning_rate": 3.005468092036401e-07,
      "loss": 0.0661,
      "step": 10760
    },
    {
      "epoch": 4.65,
      "grad_norm": 0.5972850094581258,
      "learning_rate": 2.968887311045998e-07,
      "loss": 0.0547,
      "step": 10765
    },
    {
      "epoch": 4.65,
      "grad_norm": 0.9857922947921498,
      "learning_rate": 2.932527160456955e-07,
      "loss": 0.0756,
      "step": 10770
    },
    {
      "epoch": 4.65,
      "grad_norm": 0.6562993291105274,
      "learning_rate": 2.896387722945404e-07,
      "loss": 0.0561,
      "step": 10775
    },
    {
      "epoch": 4.66,
      "grad_norm": 0.8811705396317993,
      "learning_rate": 2.8604690806855684e-07,
      "loss": 0.0677,
      "step": 10780
    },
    {
      "epoch": 4.66,
      "grad_norm": 0.7505002155595076,
      "learning_rate": 2.8247713153496504e-07,
      "loss": 0.062,
      "step": 10785
    },
    {
      "epoch": 4.66,
      "grad_norm": 0.5670638816746271,
      "learning_rate": 2.789294508107632e-07,
      "loss": 0.0603,
      "step": 10790
    },
    {
      "epoch": 4.66,
      "grad_norm": 0.937199394901662,
      "learning_rate": 2.754038739627052e-07,
      "loss": 0.0651,
      "step": 10795
    },
    {
      "epoch": 4.67,
      "grad_norm": 0.49139293261777545,
      "learning_rate": 2.7190040900728855e-07,
      "loss": 0.0632,
      "step": 10800
    },
    {
      "epoch": 4.67,
      "grad_norm": 0.48692569853606976,
      "learning_rate": 2.684190639107265e-07,
      "loss": 0.0651,
      "step": 10805
    },
    {
      "epoch": 4.67,
      "grad_norm": 0.49372578994517236,
      "learning_rate": 2.6495984658894024e-07,
      "loss": 0.0546,
      "step": 10810
    },
    {
      "epoch": 4.67,
      "grad_norm": 0.6689499796929367,
      "learning_rate": 2.6152276490753804e-07,
      "loss": 0.0565,
      "step": 10815
    },
    {
      "epoch": 4.67,
      "grad_norm": 0.5439861457356306,
      "learning_rate": 2.5810782668179156e-07,
      "loss": 0.0683,
      "step": 10820
    },
    {
      "epoch": 4.68,
      "grad_norm": 0.7221669120831035,
      "learning_rate": 2.547150396766263e-07,
      "loss": 0.0616,
      "step": 10825
    },
    {
      "epoch": 4.68,
      "grad_norm": 0.5912466008937799,
      "learning_rate": 2.513444116065966e-07,
      "loss": 0.066,
      "step": 10830
    },
    {
      "epoch": 4.68,
      "grad_norm": 0.41578702847781396,
      "learning_rate": 2.4799595013587305e-07,
      "loss": 0.052,
      "step": 10835
    },
    {
      "epoch": 4.68,
      "grad_norm": 0.4394587515378354,
      "learning_rate": 2.446696628782252e-07,
      "loss": 0.0572,
      "step": 10840
    },
    {
      "epoch": 4.68,
      "grad_norm": 0.5604383766879553,
      "learning_rate": 2.4136555739700193e-07,
      "loss": 0.0663,
      "step": 10845
    },
    {
      "epoch": 4.69,
      "grad_norm": 0.9935450578535743,
      "learning_rate": 2.3808364120511241e-07,
      "loss": 0.064,
      "step": 10850
    },
    {
      "epoch": 4.69,
      "grad_norm": 0.44874457386774114,
      "learning_rate": 2.348239217650139e-07,
      "loss": 0.0602,
      "step": 10855
    },
    {
      "epoch": 4.69,
      "grad_norm": 0.23246095201744477,
      "learning_rate": 2.3158640648869302e-07,
      "loss": 0.06,
      "step": 10860
    },
    {
      "epoch": 4.69,
      "grad_norm": 0.6255865619079833,
      "learning_rate": 2.2837110273764563e-07,
      "loss": 0.061,
      "step": 10865
    },
    {
      "epoch": 4.7,
      "grad_norm": 0.6191570836290404,
      "learning_rate": 2.2517801782286574e-07,
      "loss": 0.0584,
      "step": 10870
    },
    {
      "epoch": 4.7,
      "grad_norm": 0.46531391044936,
      "learning_rate": 2.220071590048256e-07,
      "loss": 0.0518,
      "step": 10875
    },
    {
      "epoch": 4.7,
      "grad_norm": 0.6023954576902755,
      "learning_rate": 2.1885853349345565e-07,
      "loss": 0.0611,
      "step": 10880
    },
    {
      "epoch": 4.7,
      "grad_norm": 0.6359453628805575,
      "learning_rate": 2.1573214844813673e-07,
      "loss": 0.065,
      "step": 10885
    },
    {
      "epoch": 4.7,
      "grad_norm": 0.6893301879412579,
      "learning_rate": 2.1262801097767682e-07,
      "loss": 0.0551,
      "step": 10890
    },
    {
      "epoch": 4.71,
      "grad_norm": 0.4997543153017668,
      "learning_rate": 2.095461281402966e-07,
      "loss": 0.0653,
      "step": 10895
    },
    {
      "epoch": 4.71,
      "grad_norm": 0.7737567709025983,
      "learning_rate": 2.064865069436184e-07,
      "loss": 0.0745,
      "step": 10900
    },
    {
      "epoch": 4.71,
      "grad_norm": 0.8763437454871517,
      "learning_rate": 2.0344915434463819e-07,
      "loss": 0.0631,
      "step": 10905
    },
    {
      "epoch": 4.71,
      "grad_norm": 0.8875366806583803,
      "learning_rate": 2.0043407724972265e-07,
      "loss": 0.0596,
      "step": 10910
    },
    {
      "epoch": 4.71,
      "grad_norm": 0.7156184277959455,
      "learning_rate": 1.9744128251458773e-07,
      "loss": 0.0627,
      "step": 10915
    },
    {
      "epoch": 4.72,
      "grad_norm": 0.5289051350934006,
      "learning_rate": 1.9447077694428107e-07,
      "loss": 0.0512,
      "step": 10920
    },
    {
      "epoch": 4.72,
      "grad_norm": 0.522043354481508,
      "learning_rate": 1.915225672931742e-07,
      "loss": 0.0575,
      "step": 10925
    },
    {
      "epoch": 4.72,
      "grad_norm": 0.5634897346612658,
      "learning_rate": 1.8859666026493363e-07,
      "loss": 0.0638,
      "step": 10930
    },
    {
      "epoch": 4.72,
      "grad_norm": 0.5283634108884022,
      "learning_rate": 1.8569306251252196e-07,
      "loss": 0.0602,
      "step": 10935
    },
    {
      "epoch": 4.73,
      "grad_norm": 0.6740764935995843,
      "learning_rate": 1.8281178063817017e-07,
      "loss": 0.0595,
      "step": 10940
    },
    {
      "epoch": 4.73,
      "grad_norm": 0.6823376637996936,
      "learning_rate": 1.7995282119336765e-07,
      "loss": 0.0722,
      "step": 10945
    },
    {
      "epoch": 4.73,
      "grad_norm": 0.6746547854494919,
      "learning_rate": 1.7711619067884988e-07,
      "loss": 0.0688,
      "step": 10950
    },
    {
      "epoch": 4.73,
      "grad_norm": 0.5892333526896758,
      "learning_rate": 1.7430189554457634e-07,
      "loss": 0.0638,
      "step": 10955
    },
    {
      "epoch": 4.73,
      "grad_norm": 0.5044180039297371,
      "learning_rate": 1.715099421897215e-07,
      "loss": 0.0606,
      "step": 10960
    },
    {
      "epoch": 4.74,
      "grad_norm": 0.49999321097375804,
      "learning_rate": 1.6874033696266056e-07,
      "loss": 0.0635,
      "step": 10965
    },
    {
      "epoch": 4.74,
      "grad_norm": 0.6194582830877933,
      "learning_rate": 1.6599308616095266e-07,
      "loss": 0.0579,
      "step": 10970
    },
    {
      "epoch": 4.74,
      "grad_norm": 0.7205882487405464,
      "learning_rate": 1.6326819603132537e-07,
      "loss": 0.0665,
      "step": 10975
    },
    {
      "epoch": 4.74,
      "grad_norm": 0.7802188871645003,
      "learning_rate": 1.6056567276966362e-07,
      "loss": 0.0778,
      "step": 10980
    },
    {
      "epoch": 4.75,
      "grad_norm": 0.6375719460013147,
      "learning_rate": 1.5788552252099409e-07,
      "loss": 0.0675,
      "step": 10985
    },
    {
      "epoch": 4.75,
      "grad_norm": 0.7194773791511878,
      "learning_rate": 1.552277513794731e-07,
      "loss": 0.0645,
      "step": 10990
    },
    {
      "epoch": 4.75,
      "grad_norm": 0.8278357416000771,
      "learning_rate": 1.5259236538836765e-07,
      "loss": 0.0628,
      "step": 10995
    },
    {
      "epoch": 4.75,
      "grad_norm": 0.45801396543779305,
      "learning_rate": 1.4997937054004874e-07,
      "loss": 0.0517,
      "step": 11000
    },
    {
      "epoch": 4.75,
      "grad_norm": 0.6376220924448733,
      "learning_rate": 1.4738877277597153e-07,
      "loss": 0.0593,
      "step": 11005
    },
    {
      "epoch": 4.76,
      "grad_norm": 0.608559541624421,
      "learning_rate": 1.4482057798666516e-07,
      "loss": 0.0554,
      "step": 11010
    },
    {
      "epoch": 4.76,
      "grad_norm": 0.5119755335379622,
      "learning_rate": 1.4227479201172067e-07,
      "loss": 0.0672,
      "step": 11015
    },
    {
      "epoch": 4.76,
      "grad_norm": 0.4490665941016147,
      "learning_rate": 1.397514206397721e-07,
      "loss": 0.0667,
      "step": 11020
    },
    {
      "epoch": 4.76,
      "grad_norm": 0.6251201055280632,
      "learning_rate": 1.3725046960848975e-07,
      "loss": 0.0625,
      "step": 11025
    },
    {
      "epoch": 4.76,
      "grad_norm": 0.6564636586318714,
      "learning_rate": 1.3477194460456256e-07,
      "loss": 0.0684,
      "step": 11030
    },
    {
      "epoch": 4.77,
      "grad_norm": 0.4745320704645207,
      "learning_rate": 1.3231585126368905e-07,
      "loss": 0.0609,
      "step": 11035
    },
    {
      "epoch": 4.77,
      "grad_norm": 0.5746810367692561,
      "learning_rate": 1.298821951705598e-07,
      "loss": 0.0595,
      "step": 11040
    },
    {
      "epoch": 4.77,
      "grad_norm": 0.5623447071438609,
      "learning_rate": 1.274709818588482e-07,
      "loss": 0.0682,
      "step": 11045
    },
    {
      "epoch": 4.77,
      "grad_norm": 0.6650477950040313,
      "learning_rate": 1.2508221681119869e-07,
      "loss": 0.0659,
      "step": 11050
    },
    {
      "epoch": 4.78,
      "grad_norm": 0.7413436883098828,
      "learning_rate": 1.2271590545921086e-07,
      "loss": 0.0726,
      "step": 11055
    },
    {
      "epoch": 4.78,
      "grad_norm": 0.8125027194289178,
      "learning_rate": 1.2037205318342959e-07,
      "loss": 0.0627,
      "step": 11060
    },
    {
      "epoch": 4.78,
      "grad_norm": 0.5557374623847701,
      "learning_rate": 1.1805066531332954e-07,
      "loss": 0.0609,
      "step": 11065
    },
    {
      "epoch": 4.78,
      "grad_norm": 0.38154583494463035,
      "learning_rate": 1.1575174712731063e-07,
      "loss": 0.0628,
      "step": 11070
    },
    {
      "epoch": 4.78,
      "grad_norm": 0.6962083608684523,
      "learning_rate": 1.1347530385267702e-07,
      "loss": 0.0603,
      "step": 11075
    },
    {
      "epoch": 4.79,
      "grad_norm": 0.5587167971298492,
      "learning_rate": 1.1122134066562929e-07,
      "loss": 0.0648,
      "step": 11080
    },
    {
      "epoch": 4.79,
      "grad_norm": 0.3344941093274932,
      "learning_rate": 1.0898986269125444e-07,
      "loss": 0.0553,
      "step": 11085
    },
    {
      "epoch": 4.79,
      "grad_norm": 0.5161742767786995,
      "learning_rate": 1.0678087500351041e-07,
      "loss": 0.0593,
      "step": 11090
    },
    {
      "epoch": 4.79,
      "grad_norm": 1.0950079374259003,
      "learning_rate": 1.0459438262521937e-07,
      "loss": 0.0719,
      "step": 11095
    },
    {
      "epoch": 4.79,
      "grad_norm": 0.6120823620469348,
      "learning_rate": 1.0243039052804993e-07,
      "loss": 0.068,
      "step": 11100
    },
    {
      "epoch": 4.8,
      "grad_norm": 0.8295518627255871,
      "learning_rate": 1.0028890363251165e-07,
      "loss": 0.0681,
      "step": 11105
    },
    {
      "epoch": 4.8,
      "grad_norm": 0.6779051073720008,
      "learning_rate": 9.816992680794057e-08,
      "loss": 0.0603,
      "step": 11110
    },
    {
      "epoch": 4.8,
      "grad_norm": 0.5656436041883544,
      "learning_rate": 9.607346487248926e-08,
      "loss": 0.0622,
      "step": 11115
    },
    {
      "epoch": 4.8,
      "grad_norm": 0.5831695643818923,
      "learning_rate": 9.399952259311451e-08,
      "loss": 0.0544,
      "step": 11120
    },
    {
      "epoch": 4.81,
      "grad_norm": 0.7699048409977948,
      "learning_rate": 9.194810468557191e-08,
      "loss": 0.0616,
      "step": 11125
    },
    {
      "epoch": 4.81,
      "grad_norm": 0.5542117866160443,
      "learning_rate": 8.991921581439466e-08,
      "loss": 0.0656,
      "step": 11130
    },
    {
      "epoch": 4.81,
      "grad_norm": 0.523700705648689,
      "learning_rate": 8.791286059289472e-08,
      "loss": 0.0536,
      "step": 11135
    },
    {
      "epoch": 4.81,
      "grad_norm": 0.42022329976259326,
      "learning_rate": 8.59290435831428e-08,
      "loss": 0.0685,
      "step": 11140
    },
    {
      "epoch": 4.81,
      "grad_norm": 0.6101833141730744,
      "learning_rate": 8.396776929596395e-08,
      "loss": 0.0637,
      "step": 11145
    },
    {
      "epoch": 4.82,
      "grad_norm": 0.5631499331780958,
      "learning_rate": 8.202904219092645e-08,
      "loss": 0.0633,
      "step": 11150
    },
    {
      "epoch": 4.82,
      "grad_norm": 0.5029707683863562,
      "learning_rate": 8.011286667632734e-08,
      "loss": 0.0554,
      "step": 11155
    },
    {
      "epoch": 4.82,
      "grad_norm": 0.2800645526623471,
      "learning_rate": 7.821924710918804e-08,
      "loss": 0.0641,
      "step": 11160
    },
    {
      "epoch": 4.82,
      "grad_norm": 0.6921504299490085,
      "learning_rate": 7.6348187795241e-08,
      "loss": 0.0663,
      "step": 11165
    },
    {
      "epoch": 4.83,
      "grad_norm": 0.5401882659743539,
      "learning_rate": 7.44996929889219e-08,
      "loss": 0.0613,
      "step": 11170
    },
    {
      "epoch": 4.83,
      "grad_norm": 0.7010012725517893,
      "learning_rate": 7.267376689335859e-08,
      "loss": 0.0614,
      "step": 11175
    },
    {
      "epoch": 4.83,
      "grad_norm": 1.170334350237117,
      "learning_rate": 7.087041366036107e-08,
      "loss": 0.0691,
      "step": 11180
    },
    {
      "epoch": 4.83,
      "grad_norm": 0.4312270382192399,
      "learning_rate": 6.908963739041375e-08,
      "loss": 0.0555,
      "step": 11185
    },
    {
      "epoch": 4.83,
      "grad_norm": 0.8744098840264921,
      "learning_rate": 6.73314421326643e-08,
      "loss": 0.0699,
      "step": 11190
    },
    {
      "epoch": 4.84,
      "grad_norm": 0.654485383566514,
      "learning_rate": 6.559583188491813e-08,
      "loss": 0.0654,
      "step": 11195
    },
    {
      "epoch": 4.84,
      "grad_norm": 0.9216834484238109,
      "learning_rate": 6.3882810593624e-08,
      "loss": 0.0615,
      "step": 11200
    },
    {
      "epoch": 4.84,
      "grad_norm": 0.38282636888860044,
      "learning_rate": 6.219238215386724e-08,
      "loss": 0.0614,
      "step": 11205
    },
    {
      "epoch": 4.84,
      "grad_norm": 0.8347631702339198,
      "learning_rate": 6.052455040936323e-08,
      "loss": 0.0654,
      "step": 11210
    },
    {
      "epoch": 4.84,
      "grad_norm": 0.5972877912821046,
      "learning_rate": 5.887931915244616e-08,
      "loss": 0.0576,
      "step": 11215
    },
    {
      "epoch": 4.85,
      "grad_norm": 0.3402597401613388,
      "learning_rate": 5.7256692124060266e-08,
      "loss": 0.0666,
      "step": 11220
    },
    {
      "epoch": 4.85,
      "grad_norm": 0.5181058757005162,
      "learning_rate": 5.565667301375421e-08,
      "loss": 0.0574,
      "step": 11225
    },
    {
      "epoch": 4.85,
      "grad_norm": 0.7291902925580718,
      "learning_rate": 5.407926545966669e-08,
      "loss": 0.0671,
      "step": 11230
    },
    {
      "epoch": 4.85,
      "grad_norm": 0.6238851287111219,
      "learning_rate": 5.252447304852637e-08,
      "loss": 0.0612,
      "step": 11235
    },
    {
      "epoch": 4.86,
      "grad_norm": 0.47992124625711946,
      "learning_rate": 5.0992299315634205e-08,
      "loss": 0.0624,
      "step": 11240
    },
    {
      "epoch": 4.86,
      "grad_norm": 0.903082591577723,
      "learning_rate": 4.948274774486672e-08,
      "loss": 0.0593,
      "step": 11245
    },
    {
      "epoch": 4.86,
      "grad_norm": 0.5865369565493873,
      "learning_rate": 4.7995821768656023e-08,
      "loss": 0.0582,
      "step": 11250
    },
    {
      "epoch": 4.86,
      "grad_norm": 0.4105720861995551,
      "learning_rate": 4.653152476799205e-08,
      "loss": 0.0679,
      "step": 11255
    },
    {
      "epoch": 4.86,
      "grad_norm": 0.5013484723911407,
      "learning_rate": 4.508986007241034e-08,
      "loss": 0.052,
      "step": 11260
    },
    {
      "epoch": 4.87,
      "grad_norm": 0.46155962018502744,
      "learning_rate": 4.3670830959980926e-08,
      "loss": 0.0592,
      "step": 11265
    },
    {
      "epoch": 4.87,
      "grad_norm": 0.3917564639266349,
      "learning_rate": 4.227444065730946e-08,
      "loss": 0.0717,
      "step": 11270
    },
    {
      "epoch": 4.87,
      "grad_norm": 0.7102335770665541,
      "learning_rate": 4.090069233952387e-08,
      "loss": 0.0592,
      "step": 11275
    },
    {
      "epoch": 4.87,
      "grad_norm": 0.9293214472850908,
      "learning_rate": 3.954958913026774e-08,
      "loss": 0.0692,
      "step": 11280
    },
    {
      "epoch": 4.87,
      "grad_norm": 0.35019330798760945,
      "learning_rate": 3.822113410169359e-08,
      "loss": 0.0517,
      "step": 11285
    },
    {
      "epoch": 4.88,
      "grad_norm": 0.6879669929612432,
      "learning_rate": 3.691533027445626e-08,
      "loss": 0.0632,
      "step": 11290
    },
    {
      "epoch": 4.88,
      "grad_norm": 0.6725965830420647,
      "learning_rate": 3.5632180617708454e-08,
      "loss": 0.0558,
      "step": 11295
    },
    {
      "epoch": 4.88,
      "grad_norm": 0.48268769982003784,
      "learning_rate": 3.4371688049089634e-08,
      "loss": 0.0612,
      "step": 11300
    },
    {
      "epoch": 4.88,
      "grad_norm": 0.5835298871058775,
      "learning_rate": 3.313385543472048e-08,
      "loss": 0.0539,
      "step": 11305
    },
    {
      "epoch": 4.89,
      "grad_norm": 0.5209353031141057,
      "learning_rate": 3.1918685589199524e-08,
      "loss": 0.0527,
      "step": 11310
    },
    {
      "epoch": 4.89,
      "grad_norm": 0.8854807664918086,
      "learning_rate": 3.0726181275592123e-08,
      "loss": 0.0646,
      "step": 11315
    },
    {
      "epoch": 4.89,
      "grad_norm": 0.6794898404276607,
      "learning_rate": 2.9556345205428162e-08,
      "loss": 0.066,
      "step": 11320
    },
    {
      "epoch": 4.89,
      "grad_norm": 0.6889118722336371,
      "learning_rate": 2.8409180038693197e-08,
      "loss": 0.0656,
      "step": 11325
    },
    {
      "epoch": 4.89,
      "grad_norm": 0.8580280186583628,
      "learning_rate": 2.7284688383824032e-08,
      "loss": 0.0717,
      "step": 11330
    },
    {
      "epoch": 4.9,
      "grad_norm": 0.44207148094124704,
      "learning_rate": 2.6182872797702042e-08,
      "loss": 0.0583,
      "step": 11335
    },
    {
      "epoch": 4.9,
      "grad_norm": 0.6360109435337133,
      "learning_rate": 2.5103735785647622e-08,
      "loss": 0.0706,
      "step": 11340
    },
    {
      "epoch": 4.9,
      "grad_norm": 0.7462959795590766,
      "learning_rate": 2.4047279801414634e-08,
      "loss": 0.0622,
      "step": 11345
    },
    {
      "epoch": 4.9,
      "grad_norm": 0.6410660993010558,
      "learning_rate": 2.3013507247183764e-08,
      "loss": 0.0549,
      "step": 11350
    },
    {
      "epoch": 4.9,
      "grad_norm": 0.5495295224336869,
      "learning_rate": 2.200242047355805e-08,
      "loss": 0.0641,
      "step": 11355
    },
    {
      "epoch": 4.91,
      "grad_norm": 0.7576999245837946,
      "learning_rate": 2.101402177955847e-08,
      "loss": 0.0597,
      "step": 11360
    },
    {
      "epoch": 4.91,
      "grad_norm": 0.3694481911504727,
      "learning_rate": 2.0048313412616148e-08,
      "loss": 0.0594,
      "step": 11365
    },
    {
      "epoch": 4.91,
      "grad_norm": 0.8524024130383686,
      "learning_rate": 1.910529756856905e-08,
      "loss": 0.0687,
      "step": 11370
    },
    {
      "epoch": 4.91,
      "grad_norm": 0.6884949477612426,
      "learning_rate": 1.8184976391657506e-08,
      "loss": 0.0626,
      "step": 11375
    },
    {
      "epoch": 4.92,
      "grad_norm": 0.4649978643711475,
      "learning_rate": 1.7287351974517587e-08,
      "loss": 0.061,
      "step": 11380
    },
    {
      "epoch": 4.92,
      "grad_norm": 0.5122065337906697,
      "learning_rate": 1.6412426358177747e-08,
      "loss": 0.0691,
      "step": 11385
    },
    {
      "epoch": 4.92,
      "grad_norm": 0.7026091205226528,
      "learning_rate": 1.5560201532052178e-08,
      "loss": 0.0581,
      "step": 11390
    },
    {
      "epoch": 4.92,
      "grad_norm": 0.554757648960058,
      "learning_rate": 1.4730679433940797e-08,
      "loss": 0.0613,
      "step": 11395
    },
    {
      "epoch": 4.92,
      "grad_norm": 0.6596515020568042,
      "learning_rate": 1.3923861950019268e-08,
      "loss": 0.0624,
      "step": 11400
    },
    {
      "epoch": 4.93,
      "grad_norm": 0.7074006847005723,
      "learning_rate": 1.3139750914836769e-08,
      "loss": 0.0737,
      "step": 11405
    },
    {
      "epoch": 4.93,
      "grad_norm": 0.3975859392319982,
      "learning_rate": 1.2378348111314886e-08,
      "loss": 0.0594,
      "step": 11410
    },
    {
      "epoch": 4.93,
      "grad_norm": 0.46001198290010065,
      "learning_rate": 1.1639655270738737e-08,
      "loss": 0.0635,
      "step": 11415
    },
    {
      "epoch": 4.93,
      "grad_norm": 1.0500438224677553,
      "learning_rate": 1.0923674072758073e-08,
      "loss": 0.0647,
      "step": 11420
    },
    {
      "epoch": 4.94,
      "grad_norm": 1.0586761531771032,
      "learning_rate": 1.0230406145376182e-08,
      "loss": 0.0646,
      "step": 11425
    },
    {
      "epoch": 4.94,
      "grad_norm": 0.49977826374240264,
      "learning_rate": 9.559853064954328e-09,
      "loss": 0.0654,
      "step": 11430
    },
    {
      "epoch": 4.94,
      "grad_norm": 0.3655032988191377,
      "learning_rate": 8.912016356203979e-09,
      "loss": 0.0531,
      "step": 11435
    },
    {
      "epoch": 4.94,
      "grad_norm": 0.5741494921527955,
      "learning_rate": 8.286897492182366e-09,
      "loss": 0.0593,
      "step": 11440
    },
    {
      "epoch": 4.94,
      "grad_norm": 0.31099743151940534,
      "learning_rate": 7.684497894292486e-09,
      "loss": 0.0557,
      "step": 11445
    },
    {
      "epoch": 4.95,
      "grad_norm": 0.6795734502665878,
      "learning_rate": 7.104818932276436e-09,
      "loss": 0.0611,
      "step": 11450
    },
    {
      "epoch": 4.95,
      "grad_norm": 0.5347254342737182,
      "learning_rate": 6.5478619242143045e-09,
      "loss": 0.0529,
      "step": 11455
    },
    {
      "epoch": 4.95,
      "grad_norm": 0.8047733415516746,
      "learning_rate": 6.013628136521954e-09,
      "loss": 0.0665,
      "step": 11460
    },
    {
      "epoch": 4.95,
      "grad_norm": 0.6337459372389843,
      "learning_rate": 5.502118783945465e-09,
      "loss": 0.0605,
      "step": 11465
    },
    {
      "epoch": 4.95,
      "grad_norm": 0.9441016264698104,
      "learning_rate": 5.013335029560029e-09,
      "loss": 0.0728,
      "step": 11470
    },
    {
      "epoch": 4.96,
      "grad_norm": 1.0466550453370893,
      "learning_rate": 4.547277984767728e-09,
      "loss": 0.0632,
      "step": 11475
    },
    {
      "epoch": 4.96,
      "grad_norm": 0.6740902732659063,
      "learning_rate": 4.103948709295313e-09,
      "loss": 0.068,
      "step": 11480
    },
    {
      "epoch": 4.96,
      "grad_norm": 0.7021120177061997,
      "learning_rate": 3.683348211187543e-09,
      "loss": 0.0636,
      "step": 11485
    },
    {
      "epoch": 4.96,
      "grad_norm": 0.8641288363790272,
      "learning_rate": 3.2854774468116246e-09,
      "loss": 0.0668,
      "step": 11490
    },
    {
      "epoch": 4.97,
      "grad_norm": 0.6076936871599368,
      "learning_rate": 2.9103373208505536e-09,
      "loss": 0.0645,
      "step": 11495
    },
    {
      "epoch": 4.97,
      "grad_norm": 0.7107727964656887,
      "learning_rate": 2.5579286863020025e-09,
      "loss": 0.0598,
      "step": 11500
    },
    {
      "epoch": 4.97,
      "grad_norm": 0.694761980405919,
      "learning_rate": 2.2282523444761006e-09,
      "loss": 0.0633,
      "step": 11505
    },
    {
      "epoch": 4.97,
      "grad_norm": 0.7006893637526823,
      "learning_rate": 1.921309044995434e-09,
      "loss": 0.0677,
      "step": 11510
    },
    {
      "epoch": 4.97,
      "grad_norm": 0.40600992482089776,
      "learning_rate": 1.637099485789495e-09,
      "loss": 0.0542,
      "step": 11515
    },
    {
      "epoch": 4.98,
      "grad_norm": 0.582747199794571,
      "learning_rate": 1.375624313099122e-09,
      "loss": 0.0648,
      "step": 11520
    },
    {
      "epoch": 4.98,
      "grad_norm": 1.068294118794195,
      "learning_rate": 1.1368841214676184e-09,
      "loss": 0.0733,
      "step": 11525
    },
    {
      "epoch": 4.98,
      "grad_norm": 0.9704193069259763,
      "learning_rate": 9.208794537451938e-10,
      "loss": 0.0677,
      "step": 11530
    },
    {
      "epoch": 4.98,
      "grad_norm": 0.3587463908186778,
      "learning_rate": 7.276108010867422e-10,
      "loss": 0.056,
      "step": 11535
    },
    {
      "epoch": 4.98,
      "grad_norm": 0.5892073955445215,
      "learning_rate": 5.570786029474029e-10,
      "loss": 0.06,
      "step": 11540
    },
    {
      "epoch": 4.99,
      "grad_norm": 1.2197020031797046,
      "learning_rate": 4.0928324708588986e-10,
      "loss": 0.0729,
      "step": 11545
    },
    {
      "epoch": 4.99,
      "grad_norm": 0.6267489883478871,
      "learning_rate": 2.8422506956116147e-10,
      "loss": 0.0588,
      "step": 11550
    },
    {
      "epoch": 4.99,
      "grad_norm": 0.5631817849121783,
      "learning_rate": 1.819043547313104e-10,
      "loss": 0.0566,
      "step": 11555
    },
    {
      "epoch": 4.99,
      "grad_norm": 0.5248536348350673,
      "learning_rate": 1.0232133525467369e-10,
      "loss": 0.0602,
      "step": 11560
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.44706874544550185,
      "learning_rate": 4.5476192086502115e-11,
      "loss": 0.0615,
      "step": 11565
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.7352828065983817,
      "learning_rate": 1.136905448451131e-11,
      "loss": 0.0618,
      "step": 11570
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.6400220635512455,
      "learning_rate": 0.0,
      "loss": 0.0725,
      "step": 11575
    },
    {
      "epoch": 5.0,
      "eval_loss": 0.9067042469978333,
      "eval_runtime": 352.1066,
      "eval_samples_per_second": 21.32,
      "eval_steps_per_second": 0.335,
      "step": 11575
    },
    {
      "epoch": 5.0,
      "step": 11575,
      "total_flos": 2423569514496000.0,
      "train_loss": 0.37871731976260375,
      "train_runtime": 72163.6976,
      "train_samples_per_second": 5.131,
      "train_steps_per_second": 0.16
    }
  ],
  "logging_steps": 5,
  "max_steps": 11575,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "total_flos": 2423569514496000.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}