{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 30.0,
  "eval_steps": 500,
  "global_step": 9810,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0030581039755351682,
      "grad_norm": NaN,
      "learning_rate": 0.0,
      "loss": 15.1701,
      "step": 1
    },
    {
      "epoch": 0.03058103975535168,
      "grad_norm": 22.001853942871094,
      "learning_rate": 4.5e-06,
      "loss": 18.0699,
      "step": 10
    },
    {
      "epoch": 0.06116207951070336,
      "grad_norm": 31.736783981323242,
      "learning_rate": 9.5e-06,
      "loss": 16.5865,
      "step": 20
    },
    {
      "epoch": 0.09174311926605505,
      "grad_norm": 39.22868347167969,
      "learning_rate": 1.45e-05,
      "loss": 12.788,
      "step": 30
    },
    {
      "epoch": 0.12232415902140673,
      "grad_norm": 23.849876403808594,
      "learning_rate": 1.9500000000000003e-05,
      "loss": 8.3233,
      "step": 40
    },
    {
      "epoch": 0.1529051987767584,
      "grad_norm": 8.918281555175781,
      "learning_rate": 2.45e-05,
      "loss": 4.9971,
      "step": 50
    },
    {
      "epoch": 0.1834862385321101,
      "grad_norm": 7.4376420974731445,
      "learning_rate": 2.95e-05,
      "loss": 4.2418,
      "step": 60
    },
    {
      "epoch": 0.21406727828746178,
      "grad_norm": 3.3368353843688965,
      "learning_rate": 3.45e-05,
      "loss": 4.0631,
      "step": 70
    },
    {
      "epoch": 0.24464831804281345,
      "grad_norm": 3.5972115993499756,
      "learning_rate": 3.9500000000000005e-05,
      "loss": 3.8988,
      "step": 80
    },
    {
      "epoch": 0.27522935779816515,
      "grad_norm": 1.5172040462493896,
      "learning_rate": 4.4500000000000004e-05,
      "loss": 3.8347,
      "step": 90
    },
    {
      "epoch": 0.3058103975535168,
      "grad_norm": 2.7769057750701904,
      "learning_rate": 4.9500000000000004e-05,
      "loss": 3.7567,
      "step": 100
    },
    {
      "epoch": 0.3363914373088685,
      "grad_norm": 0.8358258605003357,
      "learning_rate": 5.45e-05,
      "loss": 3.6505,
      "step": 110
    },
    {
      "epoch": 0.3669724770642202,
      "grad_norm": 0.9350353479385376,
      "learning_rate": 5.95e-05,
      "loss": 3.5953,
      "step": 120
    },
    {
      "epoch": 0.39755351681957185,
      "grad_norm": 0.7625614404678345,
      "learning_rate": 6.450000000000001e-05,
      "loss": 3.6017,
      "step": 130
    },
    {
      "epoch": 0.42813455657492355,
      "grad_norm": 0.8317046761512756,
      "learning_rate": 6.95e-05,
      "loss": 3.6038,
      "step": 140
    },
    {
      "epoch": 0.45871559633027525,
      "grad_norm": 3.3260114192962646,
      "learning_rate": 7.450000000000001e-05,
      "loss": 3.6195,
      "step": 150
    },
    {
      "epoch": 0.4892966360856269,
      "grad_norm": 3.696671485900879,
      "learning_rate": 7.950000000000001e-05,
      "loss": 3.5211,
      "step": 160
    },
    {
      "epoch": 0.5198776758409785,
      "grad_norm": 1.0185092687606812,
      "learning_rate": 8.450000000000001e-05,
      "loss": 3.5306,
      "step": 170
    },
    {
      "epoch": 0.5504587155963303,
      "grad_norm": 0.7007076740264893,
      "learning_rate": 8.950000000000001e-05,
      "loss": 3.5614,
      "step": 180
    },
    {
      "epoch": 0.581039755351682,
      "grad_norm": 1.0399439334869385,
      "learning_rate": 9.449999999999999e-05,
      "loss": 3.5602,
      "step": 190
    },
    {
      "epoch": 0.6116207951070336,
      "grad_norm": 0.945879340171814,
      "learning_rate": 9.95e-05,
      "loss": 3.5759,
      "step": 200
    },
    {
      "epoch": 0.6422018348623854,
      "grad_norm": 0.6479012966156006,
      "learning_rate": 9.99063475546306e-05,
      "loss": 3.5129,
      "step": 210
    },
    {
      "epoch": 0.672782874617737,
      "grad_norm": 0.9196483492851257,
      "learning_rate": 9.980228928199792e-05,
      "loss": 3.4773,
      "step": 220
    },
    {
      "epoch": 0.7033639143730887,
      "grad_norm": 0.8915090560913086,
      "learning_rate": 9.969823100936525e-05,
      "loss": 3.5257,
      "step": 230
    },
    {
      "epoch": 0.7339449541284404,
      "grad_norm": 0.7033812403678894,
      "learning_rate": 9.959417273673257e-05,
      "loss": 3.5344,
      "step": 240
    },
    {
      "epoch": 0.764525993883792,
      "grad_norm": 2.416579246520996,
      "learning_rate": 9.94901144640999e-05,
      "loss": 3.5397,
      "step": 250
    },
    {
      "epoch": 0.7951070336391437,
      "grad_norm": 0.5419871211051941,
      "learning_rate": 9.938605619146722e-05,
      "loss": 3.4923,
      "step": 260
    },
    {
      "epoch": 0.8256880733944955,
      "grad_norm": 1.1078801155090332,
      "learning_rate": 9.928199791883456e-05,
      "loss": 3.4991,
      "step": 270
    },
    {
      "epoch": 0.8562691131498471,
      "grad_norm": 0.8176569938659668,
      "learning_rate": 9.917793964620187e-05,
      "loss": 3.5006,
      "step": 280
    },
    {
      "epoch": 0.8868501529051988,
      "grad_norm": 1.468253493309021,
      "learning_rate": 9.90738813735692e-05,
      "loss": 3.5245,
      "step": 290
    },
    {
      "epoch": 0.9174311926605505,
      "grad_norm": 1.12344491481781,
      "learning_rate": 9.896982310093654e-05,
      "loss": 3.576,
      "step": 300
    },
    {
      "epoch": 0.9480122324159022,
      "grad_norm": 1.7714366912841797,
      "learning_rate": 9.886576482830385e-05,
      "loss": 3.4923,
      "step": 310
    },
    {
      "epoch": 0.9785932721712538,
      "grad_norm": 0.7636281251907349,
      "learning_rate": 9.876170655567119e-05,
      "loss": 3.4947,
      "step": 320
    },
    {
      "epoch": 1.0091743119266054,
      "grad_norm": 1.032314658164978,
      "learning_rate": 9.865764828303851e-05,
      "loss": 3.4876,
      "step": 330
    },
    {
      "epoch": 1.039755351681957,
      "grad_norm": 1.217951774597168,
      "learning_rate": 9.855359001040582e-05,
      "loss": 3.4614,
      "step": 340
    },
    {
      "epoch": 1.070336391437309,
      "grad_norm": 2.222341299057007,
      "learning_rate": 9.844953173777316e-05,
      "loss": 3.3953,
      "step": 350
    },
    {
      "epoch": 1.1009174311926606,
      "grad_norm": 2.0637242794036865,
      "learning_rate": 9.834547346514049e-05,
      "loss": 3.3575,
      "step": 360
    },
    {
      "epoch": 1.1314984709480123,
      "grad_norm": 2.882171392440796,
      "learning_rate": 9.824141519250781e-05,
      "loss": 3.2256,
      "step": 370
    },
    {
      "epoch": 1.162079510703364,
      "grad_norm": 2.4455676078796387,
      "learning_rate": 9.813735691987514e-05,
      "loss": 3.1562,
      "step": 380
    },
    {
      "epoch": 1.1926605504587156,
      "grad_norm": 2.7748982906341553,
      "learning_rate": 9.803329864724245e-05,
      "loss": 2.9223,
      "step": 390
    },
    {
      "epoch": 1.2232415902140672,
      "grad_norm": 3.552553415298462,
      "learning_rate": 9.792924037460979e-05,
      "loss": 2.8258,
      "step": 400
    },
    {
      "epoch": 1.2538226299694188,
      "grad_norm": 7.971360206604004,
      "learning_rate": 9.782518210197711e-05,
      "loss": 2.711,
      "step": 410
    },
    {
      "epoch": 1.2844036697247707,
      "grad_norm": 4.942215442657471,
      "learning_rate": 9.772112382934444e-05,
      "loss": 2.6485,
      "step": 420
    },
    {
      "epoch": 1.3149847094801224,
      "grad_norm": 2.9292726516723633,
      "learning_rate": 9.761706555671176e-05,
      "loss": 2.6137,
      "step": 430
    },
    {
      "epoch": 1.345565749235474,
      "grad_norm": 3.012336015701294,
      "learning_rate": 9.751300728407909e-05,
      "loss": 2.4069,
      "step": 440
    },
    {
      "epoch": 1.3761467889908257,
      "grad_norm": 3.4564368724823,
      "learning_rate": 9.740894901144641e-05,
      "loss": 2.2903,
      "step": 450
    },
    {
      "epoch": 1.4067278287461773,
      "grad_norm": 5.477275848388672,
      "learning_rate": 9.730489073881374e-05,
      "loss": 2.2817,
      "step": 460
    },
    {
      "epoch": 1.4373088685015292,
      "grad_norm": 4.544290065765381,
      "learning_rate": 9.720083246618106e-05,
      "loss": 2.2782,
      "step": 470
    },
    {
      "epoch": 1.4678899082568808,
      "grad_norm": 3.39713454246521,
      "learning_rate": 9.709677419354839e-05,
      "loss": 2.2134,
      "step": 480
    },
    {
      "epoch": 1.4984709480122325,
      "grad_norm": 11.243364334106445,
      "learning_rate": 9.699271592091573e-05,
      "loss": 2.0284,
      "step": 490
    },
    {
      "epoch": 1.529051987767584,
      "grad_norm": 7.737382411956787,
      "learning_rate": 9.688865764828304e-05,
      "loss": 2.0166,
      "step": 500
    },
    {
      "epoch": 1.529051987767584,
      "eval_loss": 1.865260124206543,
      "eval_runtime": 17.067,
      "eval_samples_per_second": 17.05,
      "eval_steps_per_second": 8.555,
      "step": 500
    },
    {
      "epoch": 1.5596330275229358,
      "grad_norm": 7.2488694190979,
      "learning_rate": 9.678459937565036e-05,
      "loss": 2.0329,
      "step": 510
    },
    {
      "epoch": 1.5902140672782874,
      "grad_norm": 4.852959632873535,
      "learning_rate": 9.66805411030177e-05,
      "loss": 2.015,
      "step": 520
    },
    {
      "epoch": 1.620795107033639,
      "grad_norm": 5.300124645233154,
      "learning_rate": 9.657648283038501e-05,
      "loss": 2.0602,
      "step": 530
    },
    {
      "epoch": 1.6513761467889907,
      "grad_norm": 3.9786946773529053,
      "learning_rate": 9.647242455775235e-05,
      "loss": 1.8694,
      "step": 540
    },
    {
      "epoch": 1.6819571865443423,
      "grad_norm": 10.176203727722168,
      "learning_rate": 9.636836628511968e-05,
      "loss": 1.8715,
      "step": 550
    },
    {
      "epoch": 1.7125382262996942,
      "grad_norm": 4.76723051071167,
      "learning_rate": 9.626430801248699e-05,
      "loss": 1.8218,
      "step": 560
    },
    {
      "epoch": 1.7431192660550459,
      "grad_norm": 7.081244468688965,
      "learning_rate": 9.616024973985433e-05,
      "loss": 1.8937,
      "step": 570
    },
    {
      "epoch": 1.7737003058103975,
      "grad_norm": 5.473796367645264,
      "learning_rate": 9.605619146722165e-05,
      "loss": 1.961,
      "step": 580
    },
    {
      "epoch": 1.8042813455657494,
      "grad_norm": 6.410043239593506,
      "learning_rate": 9.595213319458898e-05,
      "loss": 1.7208,
      "step": 590
    },
    {
      "epoch": 1.834862385321101,
      "grad_norm": 6.149592876434326,
      "learning_rate": 9.58480749219563e-05,
      "loss": 1.7194,
      "step": 600
    },
    {
      "epoch": 1.8654434250764527,
      "grad_norm": 8.549052238464355,
      "learning_rate": 9.574401664932363e-05,
      "loss": 1.7529,
      "step": 610
    },
    {
      "epoch": 1.8960244648318043,
      "grad_norm": 9.397221565246582,
      "learning_rate": 9.563995837669095e-05,
      "loss": 1.8058,
      "step": 620
    },
    {
      "epoch": 1.926605504587156,
      "grad_norm": 5.5874104499816895,
      "learning_rate": 9.553590010405828e-05,
      "loss": 1.7526,
      "step": 630
    },
    {
      "epoch": 1.9571865443425076,
      "grad_norm": 6.131716251373291,
      "learning_rate": 9.54318418314256e-05,
      "loss": 1.6377,
      "step": 640
    },
    {
      "epoch": 1.9877675840978593,
      "grad_norm": 14.399773597717285,
      "learning_rate": 9.532778355879293e-05,
      "loss": 1.7617,
      "step": 650
    },
    {
      "epoch": 2.018348623853211,
      "grad_norm": 3.3374037742614746,
      "learning_rate": 9.522372528616027e-05,
      "loss": 1.6114,
      "step": 660
    },
    {
      "epoch": 2.0489296636085625,
      "grad_norm": 4.952519416809082,
      "learning_rate": 9.511966701352758e-05,
      "loss": 1.4885,
      "step": 670
    },
    {
      "epoch": 2.079510703363914,
      "grad_norm": 13.499002456665039,
      "learning_rate": 9.50156087408949e-05,
      "loss": 1.4875,
      "step": 680
    },
    {
      "epoch": 2.1100917431192663,
      "grad_norm": 5.6005473136901855,
      "learning_rate": 9.491155046826224e-05,
      "loss": 1.5514,
      "step": 690
    },
    {
      "epoch": 2.140672782874618,
      "grad_norm": 7.581764221191406,
      "learning_rate": 9.480749219562955e-05,
      "loss": 1.5322,
      "step": 700
    },
    {
      "epoch": 2.1712538226299696,
      "grad_norm": 4.7952046394348145,
      "learning_rate": 9.470343392299689e-05,
      "loss": 1.5022,
      "step": 710
    },
    {
      "epoch": 2.2018348623853212,
      "grad_norm": 5.151154041290283,
      "learning_rate": 9.45993756503642e-05,
      "loss": 1.3642,
      "step": 720
    },
    {
      "epoch": 2.232415902140673,
      "grad_norm": 4.606308937072754,
      "learning_rate": 9.449531737773153e-05,
      "loss": 1.4213,
      "step": 730
    },
    {
      "epoch": 2.2629969418960245,
      "grad_norm": 6.860902309417725,
      "learning_rate": 9.439125910509887e-05,
      "loss": 1.4115,
      "step": 740
    },
    {
      "epoch": 2.293577981651376,
      "grad_norm": 10.510758399963379,
      "learning_rate": 9.428720083246618e-05,
      "loss": 1.5298,
      "step": 750
    },
    {
      "epoch": 2.324159021406728,
      "grad_norm": 4.704411029815674,
      "learning_rate": 9.418314255983352e-05,
      "loss": 1.4785,
      "step": 760
    },
    {
      "epoch": 2.3547400611620795,
      "grad_norm": 4.206416130065918,
      "learning_rate": 9.407908428720084e-05,
      "loss": 1.3483,
      "step": 770
    },
    {
      "epoch": 2.385321100917431,
      "grad_norm": 8.520705223083496,
      "learning_rate": 9.397502601456815e-05,
      "loss": 1.2905,
      "step": 780
    },
    {
      "epoch": 2.4159021406727827,
      "grad_norm": 7.525352478027344,
      "learning_rate": 9.387096774193549e-05,
      "loss": 1.3142,
      "step": 790
    },
    {
      "epoch": 2.4464831804281344,
      "grad_norm": 8.8917818069458,
      "learning_rate": 9.376690946930282e-05,
      "loss": 1.4135,
      "step": 800
    },
    {
      "epoch": 2.477064220183486,
      "grad_norm": 7.770375728607178,
      "learning_rate": 9.366285119667014e-05,
      "loss": 1.3812,
      "step": 810
    },
    {
      "epoch": 2.5076452599388377,
      "grad_norm": 4.642964839935303,
      "learning_rate": 9.355879292403747e-05,
      "loss": 1.2248,
      "step": 820
    },
    {
      "epoch": 2.5382262996941893,
      "grad_norm": 7.788937091827393,
      "learning_rate": 9.345473465140479e-05,
      "loss": 1.3524,
      "step": 830
    },
    {
      "epoch": 2.5688073394495414,
      "grad_norm": 9.559152603149414,
      "learning_rate": 9.335067637877212e-05,
      "loss": 1.3342,
      "step": 840
    },
    {
      "epoch": 2.599388379204893,
      "grad_norm": 6.956350803375244,
      "learning_rate": 9.324661810613944e-05,
      "loss": 1.3578,
      "step": 850
    },
    {
      "epoch": 2.6299694189602447,
      "grad_norm": 4.262228012084961,
      "learning_rate": 9.314255983350677e-05,
      "loss": 1.3014,
      "step": 860
    },
    {
      "epoch": 2.6605504587155964,
      "grad_norm": 6.106989860534668,
      "learning_rate": 9.303850156087409e-05,
      "loss": 1.1986,
      "step": 870
    },
    {
      "epoch": 2.691131498470948,
      "grad_norm": 9.121493339538574,
      "learning_rate": 9.293444328824143e-05,
      "loss": 1.2134,
      "step": 880
    },
    {
      "epoch": 2.7217125382262997,
      "grad_norm": 6.555141925811768,
      "learning_rate": 9.283038501560874e-05,
      "loss": 1.2174,
      "step": 890
    },
    {
      "epoch": 2.7522935779816513,
      "grad_norm": 14.706107139587402,
      "learning_rate": 9.272632674297607e-05,
      "loss": 1.3036,
      "step": 900
    },
    {
      "epoch": 2.782874617737003,
      "grad_norm": 3.802637815475464,
      "learning_rate": 9.26222684703434e-05,
      "loss": 1.1968,
      "step": 910
    },
    {
      "epoch": 2.8134556574923546,
      "grad_norm": 5.46152925491333,
      "learning_rate": 9.251821019771072e-05,
      "loss": 1.2127,
      "step": 920
    },
    {
      "epoch": 2.8440366972477067,
      "grad_norm": 15.203720092773438,
      "learning_rate": 9.241415192507805e-05,
      "loss": 1.2008,
      "step": 930
    },
    {
      "epoch": 2.8746177370030583,
      "grad_norm": 7.590661525726318,
      "learning_rate": 9.231009365244538e-05,
      "loss": 1.2146,
      "step": 940
    },
    {
      "epoch": 2.90519877675841,
      "grad_norm": 5.586644172668457,
      "learning_rate": 9.220603537981269e-05,
      "loss": 1.2314,
      "step": 950
    },
    {
      "epoch": 2.9357798165137616,
      "grad_norm": 3.6150929927825928,
      "learning_rate": 9.210197710718003e-05,
      "loss": 1.2373,
      "step": 960
    },
    {
      "epoch": 2.9663608562691133,
      "grad_norm": 4.534612655639648,
      "learning_rate": 9.199791883454734e-05,
      "loss": 1.1448,
      "step": 970
    },
    {
      "epoch": 2.996941896024465,
      "grad_norm": 7.266969203948975,
      "learning_rate": 9.189386056191468e-05,
      "loss": 1.22,
      "step": 980
    },
    {
      "epoch": 3.0275229357798166,
      "grad_norm": 4.0980987548828125,
      "learning_rate": 9.1789802289282e-05,
      "loss": 1.0809,
      "step": 990
    },
    {
      "epoch": 3.058103975535168,
      "grad_norm": 7.4049201011657715,
      "learning_rate": 9.168574401664933e-05,
      "loss": 1.0163,
      "step": 1000
    },
    {
      "epoch": 3.058103975535168,
      "eval_loss": 0.9477021098136902,
      "eval_runtime": 15.4641,
      "eval_samples_per_second": 18.818,
      "eval_steps_per_second": 9.441,
      "step": 1000
    },
    {
      "epoch": 3.08868501529052,
      "grad_norm": 6.115686416625977,
      "learning_rate": 9.158168574401665e-05,
      "loss": 1.1167,
      "step": 1010
    },
    {
      "epoch": 3.1192660550458715,
      "grad_norm": 4.0201005935668945,
      "learning_rate": 9.147762747138398e-05,
      "loss": 1.0734,
      "step": 1020
    },
    {
      "epoch": 3.149847094801223,
      "grad_norm": 5.6173176765441895,
      "learning_rate": 9.13735691987513e-05,
      "loss": 1.1944,
      "step": 1030
    },
    {
      "epoch": 3.180428134556575,
      "grad_norm": 6.7310590744018555,
      "learning_rate": 9.126951092611863e-05,
      "loss": 1.0825,
      "step": 1040
    },
    {
      "epoch": 3.2110091743119265,
      "grad_norm": 3.8579742908477783,
      "learning_rate": 9.116545265348595e-05,
      "loss": 0.9264,
      "step": 1050
    },
    {
      "epoch": 3.241590214067278,
      "grad_norm": 6.208363056182861,
      "learning_rate": 9.106139438085328e-05,
      "loss": 1.0291,
      "step": 1060
    },
    {
      "epoch": 3.2721712538226297,
      "grad_norm": 8.63094711303711,
      "learning_rate": 9.09573361082206e-05,
      "loss": 1.0673,
      "step": 1070
    },
    {
      "epoch": 3.302752293577982,
      "grad_norm": 8.783857345581055,
      "learning_rate": 9.085327783558793e-05,
      "loss": 1.1216,
      "step": 1080
    },
    {
      "epoch": 3.3333333333333335,
      "grad_norm": 3.576794147491455,
      "learning_rate": 9.074921956295525e-05,
      "loss": 1.0424,
      "step": 1090
    },
    {
      "epoch": 3.363914373088685,
      "grad_norm": 5.663135528564453,
      "learning_rate": 9.06451612903226e-05,
      "loss": 0.9512,
      "step": 1100
    },
    {
      "epoch": 3.3944954128440368,
      "grad_norm": 5.0192108154296875,
      "learning_rate": 9.05411030176899e-05,
      "loss": 0.9458,
      "step": 1110
    },
    {
      "epoch": 3.4250764525993884,
      "grad_norm": 9.276246070861816,
      "learning_rate": 9.043704474505723e-05,
      "loss": 0.9878,
      "step": 1120
    },
    {
      "epoch": 3.45565749235474,
      "grad_norm": 6.982539653778076,
      "learning_rate": 9.033298647242457e-05,
      "loss": 1.0918,
      "step": 1130
    },
    {
      "epoch": 3.4862385321100917,
      "grad_norm": 4.335603713989258,
      "learning_rate": 9.022892819979188e-05,
      "loss": 0.9489,
      "step": 1140
    },
    {
      "epoch": 3.5168195718654434,
      "grad_norm": 4.5788092613220215,
      "learning_rate": 9.012486992715922e-05,
      "loss": 0.9649,
      "step": 1150
    },
    {
      "epoch": 3.547400611620795,
      "grad_norm": 4.774909496307373,
      "learning_rate": 9.002081165452654e-05,
      "loss": 0.9678,
      "step": 1160
    },
    {
      "epoch": 3.5779816513761467,
      "grad_norm": 9.079809188842773,
      "learning_rate": 8.991675338189387e-05,
      "loss": 1.049,
      "step": 1170
    },
    {
      "epoch": 3.6085626911314987,
      "grad_norm": 5.256271839141846,
      "learning_rate": 8.98126951092612e-05,
      "loss": 1.039,
      "step": 1180
    },
    {
      "epoch": 3.6391437308868504,
      "grad_norm": 5.269153118133545,
      "learning_rate": 8.970863683662852e-05,
      "loss": 0.9424,
      "step": 1190
    },
    {
      "epoch": 3.669724770642202,
      "grad_norm": 10.608845710754395,
      "learning_rate": 8.960457856399584e-05,
      "loss": 0.9202,
      "step": 1200
    },
    {
      "epoch": 3.7003058103975537,
      "grad_norm": 5.39443302154541,
      "learning_rate": 8.950052029136317e-05,
      "loss": 0.9395,
      "step": 1210
    },
    {
      "epoch": 3.7308868501529053,
      "grad_norm": 10.286017417907715,
      "learning_rate": 8.93964620187305e-05,
      "loss": 0.9538,
      "step": 1220
    },
    {
      "epoch": 3.761467889908257,
      "grad_norm": 7.594948768615723,
      "learning_rate": 8.929240374609782e-05,
      "loss": 1.0294,
      "step": 1230
    },
    {
      "epoch": 3.7920489296636086,
      "grad_norm": 6.064659595489502,
      "learning_rate": 8.918834547346514e-05,
      "loss": 0.9486,
      "step": 1240
    },
    {
      "epoch": 3.8226299694189603,
      "grad_norm": 5.366261959075928,
      "learning_rate": 8.908428720083247e-05,
      "loss": 0.9198,
      "step": 1250
    },
    {
      "epoch": 3.853211009174312,
      "grad_norm": 5.648370265960693,
      "learning_rate": 8.89802289281998e-05,
      "loss": 0.9617,
      "step": 1260
    },
    {
      "epoch": 3.8837920489296636,
      "grad_norm": 10.285215377807617,
      "learning_rate": 8.887617065556713e-05,
      "loss": 0.9628,
      "step": 1270
    },
    {
      "epoch": 3.914373088685015,
      "grad_norm": 5.895651340484619,
      "learning_rate": 8.877211238293444e-05,
      "loss": 0.9326,
      "step": 1280
    },
    {
      "epoch": 3.944954128440367,
      "grad_norm": 3.7789382934570312,
      "learning_rate": 8.866805411030177e-05,
      "loss": 0.8489,
      "step": 1290
    },
    {
      "epoch": 3.9755351681957185,
      "grad_norm": 8.380213737487793,
      "learning_rate": 8.85639958376691e-05,
      "loss": 0.8593,
      "step": 1300
    },
    {
      "epoch": 4.00611620795107,
      "grad_norm": 4.148671627044678,
      "learning_rate": 8.845993756503642e-05,
      "loss": 0.9416,
      "step": 1310
    },
    {
      "epoch": 4.036697247706422,
      "grad_norm": 4.745501518249512,
      "learning_rate": 8.835587929240376e-05,
      "loss": 0.7869,
      "step": 1320
    },
    {
      "epoch": 4.0672782874617734,
      "grad_norm": 6.28198766708374,
      "learning_rate": 8.825182101977107e-05,
      "loss": 0.8129,
      "step": 1330
    },
    {
      "epoch": 4.097859327217125,
      "grad_norm": 7.487410545349121,
      "learning_rate": 8.814776274713841e-05,
      "loss": 0.8523,
      "step": 1340
    },
    {
      "epoch": 4.128440366972477,
      "grad_norm": 7.346683979034424,
      "learning_rate": 8.804370447450573e-05,
      "loss": 0.8112,
      "step": 1350
    },
    {
      "epoch": 4.159021406727828,
      "grad_norm": 5.072343349456787,
      "learning_rate": 8.793964620187304e-05,
      "loss": 0.9141,
      "step": 1360
    },
    {
      "epoch": 4.18960244648318,
      "grad_norm": 6.268205165863037,
      "learning_rate": 8.783558792924038e-05,
      "loss": 0.7041,
      "step": 1370
    },
    {
      "epoch": 4.220183486238533,
      "grad_norm": 6.782501697540283,
      "learning_rate": 8.773152965660771e-05,
      "loss": 0.7478,
      "step": 1380
    },
    {
      "epoch": 4.250764525993883,
      "grad_norm": 6.050166606903076,
      "learning_rate": 8.762747138397503e-05,
      "loss": 0.8574,
      "step": 1390
    },
    {
      "epoch": 4.281345565749236,
      "grad_norm": 5.380760669708252,
      "learning_rate": 8.752341311134236e-05,
      "loss": 0.7921,
      "step": 1400
    },
    {
      "epoch": 4.3119266055045875,
      "grad_norm": 4.861249923706055,
      "learning_rate": 8.741935483870968e-05,
      "loss": 0.9141,
      "step": 1410
    },
    {
      "epoch": 4.342507645259939,
      "grad_norm": 7.817598819732666,
      "learning_rate": 8.731529656607701e-05,
      "loss": 0.7957,
      "step": 1420
    },
    {
      "epoch": 4.373088685015291,
      "grad_norm": 4.377925395965576,
      "learning_rate": 8.721123829344433e-05,
      "loss": 0.7507,
      "step": 1430
    },
    {
      "epoch": 4.4036697247706424,
      "grad_norm": 5.141635417938232,
      "learning_rate": 8.710718002081166e-05,
      "loss": 0.7613,
      "step": 1440
    },
    {
      "epoch": 4.434250764525994,
      "grad_norm": 4.599380970001221,
      "learning_rate": 8.700312174817898e-05,
      "loss": 0.8004,
      "step": 1450
    },
    {
      "epoch": 4.464831804281346,
      "grad_norm": 4.748518466949463,
      "learning_rate": 8.689906347554631e-05,
      "loss": 0.902,
      "step": 1460
    },
    {
      "epoch": 4.495412844036697,
      "grad_norm": 5.193148612976074,
      "learning_rate": 8.679500520291363e-05,
      "loss": 0.7295,
      "step": 1470
    },
    {
      "epoch": 4.525993883792049,
      "grad_norm": 8.4327392578125,
      "learning_rate": 8.669094693028096e-05,
      "loss": 0.7065,
      "step": 1480
    },
    {
      "epoch": 4.556574923547401,
      "grad_norm": 7.619236469268799,
      "learning_rate": 8.65868886576483e-05,
      "loss": 0.8132,
      "step": 1490
    },
    {
      "epoch": 4.587155963302752,
      "grad_norm": 5.090475559234619,
      "learning_rate": 8.648283038501561e-05,
      "loss": 0.8103,
      "step": 1500
    },
    {
      "epoch": 4.587155963302752,
      "eval_loss": 0.6886358857154846,
      "eval_runtime": 17.1364,
      "eval_samples_per_second": 16.981,
      "eval_steps_per_second": 8.52,
      "step": 1500
    },
    {
      "epoch": 4.617737003058104,
      "grad_norm": 4.3185133934021,
      "learning_rate": 8.637877211238295e-05,
      "loss": 0.8655,
      "step": 1510
    },
    {
      "epoch": 4.648318042813456,
      "grad_norm": 3.376910448074341,
      "learning_rate": 8.627471383975027e-05,
      "loss": 0.7238,
      "step": 1520
    },
    {
      "epoch": 4.678899082568807,
      "grad_norm": 3.3280067443847656,
      "learning_rate": 8.617065556711758e-05,
      "loss": 0.7003,
      "step": 1530
    },
    {
      "epoch": 4.709480122324159,
      "grad_norm": 9.140504837036133,
      "learning_rate": 8.606659729448492e-05,
      "loss": 0.7757,
      "step": 1540
    },
    {
      "epoch": 4.740061162079511,
      "grad_norm": 4.642270088195801,
      "learning_rate": 8.596253902185225e-05,
      "loss": 0.7487,
      "step": 1550
    },
    {
      "epoch": 4.770642201834862,
      "grad_norm": 4.702116966247559,
      "learning_rate": 8.585848074921957e-05,
      "loss": 0.8316,
      "step": 1560
    },
    {
      "epoch": 4.801223241590214,
      "grad_norm": 7.388566970825195,
      "learning_rate": 8.57544224765869e-05,
      "loss": 0.6916,
      "step": 1570
    },
    {
      "epoch": 4.8318042813455655,
      "grad_norm": 8.01981258392334,
      "learning_rate": 8.565036420395421e-05,
      "loss": 0.7335,
      "step": 1580
    },
    {
      "epoch": 4.862385321100917,
      "grad_norm": 16.20261573791504,
      "learning_rate": 8.554630593132155e-05,
      "loss": 0.75,
      "step": 1590
    },
    {
      "epoch": 4.892966360856269,
      "grad_norm": 10.875531196594238,
      "learning_rate": 8.544224765868887e-05,
      "loss": 0.7442,
      "step": 1600
    },
    {
      "epoch": 4.92354740061162,
      "grad_norm": 3.217120409011841,
      "learning_rate": 8.53381893860562e-05,
      "loss": 0.8253,
      "step": 1610
    },
    {
      "epoch": 4.954128440366972,
      "grad_norm": 5.443752765655518,
      "learning_rate": 8.523413111342352e-05,
      "loss": 0.7353,
      "step": 1620
    },
    {
      "epoch": 4.984709480122325,
      "grad_norm": 6.295218467712402,
      "learning_rate": 8.513007284079085e-05,
      "loss": 0.7714,
      "step": 1630
    },
    {
      "epoch": 5.015290519877676,
      "grad_norm": 4.707128524780273,
      "learning_rate": 8.502601456815817e-05,
      "loss": 0.7602,
      "step": 1640
    },
    {
      "epoch": 5.045871559633028,
      "grad_norm": 6.996250152587891,
      "learning_rate": 8.49219562955255e-05,
      "loss": 0.6049,
      "step": 1650
    },
    {
      "epoch": 5.07645259938838,
      "grad_norm": 5.075336933135986,
      "learning_rate": 8.481789802289282e-05,
      "loss": 0.6293,
      "step": 1660
    },
    {
      "epoch": 5.107033639143731,
      "grad_norm": 5.5144476890563965,
      "learning_rate": 8.471383975026015e-05,
      "loss": 0.6873,
      "step": 1670
    },
    {
      "epoch": 5.137614678899083,
      "grad_norm": 6.9491167068481445,
      "learning_rate": 8.460978147762747e-05,
      "loss": 0.6847,
      "step": 1680
    },
    {
      "epoch": 5.1681957186544345,
      "grad_norm": 4.417248249053955,
      "learning_rate": 8.45057232049948e-05,
      "loss": 0.7481,
      "step": 1690
    },
    {
      "epoch": 5.198776758409786,
      "grad_norm": 4.493064880371094,
      "learning_rate": 8.440166493236212e-05,
      "loss": 0.5935,
      "step": 1700
    },
    {
      "epoch": 5.229357798165138,
      "grad_norm": 5.201015472412109,
      "learning_rate": 8.429760665972946e-05,
      "loss": 0.6575,
      "step": 1710
    },
    {
      "epoch": 5.259938837920489,
      "grad_norm": 6.402657985687256,
      "learning_rate": 8.419354838709677e-05,
      "loss": 0.6748,
      "step": 1720
    },
    {
      "epoch": 5.290519877675841,
      "grad_norm": 12.927905082702637,
      "learning_rate": 8.408949011446411e-05,
      "loss": 0.7321,
      "step": 1730
    },
    {
      "epoch": 5.321100917431193,
      "grad_norm": 3.7228381633758545,
      "learning_rate": 8.398543184183144e-05,
      "loss": 0.7307,
      "step": 1740
    },
    {
      "epoch": 5.351681957186544,
      "grad_norm": 5.624021530151367,
      "learning_rate": 8.388137356919875e-05,
      "loss": 0.5725,
      "step": 1750
    },
    {
      "epoch": 5.382262996941896,
      "grad_norm": 7.779863357543945,
      "learning_rate": 8.377731529656609e-05,
      "loss": 0.6315,
      "step": 1760
    },
    {
      "epoch": 5.412844036697248,
      "grad_norm": 10.124401092529297,
      "learning_rate": 8.367325702393341e-05,
      "loss": 0.6838,
      "step": 1770
    },
    {
      "epoch": 5.443425076452599,
      "grad_norm": 6.292506217956543,
      "learning_rate": 8.356919875130074e-05,
      "loss": 0.6162,
      "step": 1780
    },
    {
      "epoch": 5.474006116207951,
      "grad_norm": 5.369190692901611,
      "learning_rate": 8.346514047866806e-05,
      "loss": 0.6983,
      "step": 1790
    },
    {
      "epoch": 5.504587155963303,
      "grad_norm": 4.705302715301514,
      "learning_rate": 8.336108220603539e-05,
      "loss": 0.5728,
      "step": 1800
    },
    {
      "epoch": 5.535168195718654,
      "grad_norm": 4.299398422241211,
      "learning_rate": 8.325702393340271e-05,
      "loss": 0.626,
      "step": 1810
    },
    {
      "epoch": 5.565749235474006,
      "grad_norm": 6.148344993591309,
      "learning_rate": 8.315296566077004e-05,
      "loss": 0.6327,
      "step": 1820
    },
    {
      "epoch": 5.5963302752293576,
      "grad_norm": 5.602970600128174,
      "learning_rate": 8.304890738813736e-05,
      "loss": 0.6552,
      "step": 1830
    },
    {
      "epoch": 5.626911314984709,
      "grad_norm": 3.46807861328125,
      "learning_rate": 8.294484911550469e-05,
      "loss": 0.7196,
      "step": 1840
    },
    {
      "epoch": 5.657492354740061,
      "grad_norm": 3.249087333679199,
      "learning_rate": 8.284079084287201e-05,
      "loss": 0.6087,
      "step": 1850
    },
    {
      "epoch": 5.6880733944954125,
      "grad_norm": 7.348158836364746,
      "learning_rate": 8.273673257023934e-05,
      "loss": 0.5612,
      "step": 1860
    },
    {
      "epoch": 5.718654434250764,
      "grad_norm": 5.426358222961426,
      "learning_rate": 8.263267429760666e-05,
      "loss": 0.6494,
      "step": 1870
    },
    {
      "epoch": 5.749235474006117,
      "grad_norm": 7.073860168457031,
      "learning_rate": 8.2528616024974e-05,
      "loss": 0.6407,
      "step": 1880
    },
    {
      "epoch": 5.779816513761467,
      "grad_norm": 4.780755519866943,
      "learning_rate": 8.242455775234131e-05,
      "loss": 0.6997,
      "step": 1890
    },
    {
      "epoch": 5.81039755351682,
      "grad_norm": 5.361114978790283,
      "learning_rate": 8.232049947970865e-05,
      "loss": 0.6212,
      "step": 1900
    },
    {
      "epoch": 5.840978593272172,
      "grad_norm": 5.250582695007324,
      "learning_rate": 8.221644120707596e-05,
      "loss": 0.6705,
      "step": 1910
    },
    {
      "epoch": 5.871559633027523,
      "grad_norm": 4.780821323394775,
      "learning_rate": 8.211238293444329e-05,
      "loss": 0.6421,
      "step": 1920
    },
    {
      "epoch": 5.902140672782875,
      "grad_norm": 6.975560188293457,
      "learning_rate": 8.200832466181062e-05,
      "loss": 0.7433,
      "step": 1930
    },
    {
      "epoch": 5.9327217125382266,
      "grad_norm": 3.3447937965393066,
      "learning_rate": 8.190426638917794e-05,
      "loss": 0.6109,
      "step": 1940
    },
    {
      "epoch": 5.963302752293578,
      "grad_norm": 9.269634246826172,
      "learning_rate": 8.180020811654527e-05,
      "loss": 0.5586,
      "step": 1950
    },
    {
      "epoch": 5.99388379204893,
      "grad_norm": 7.32548713684082,
      "learning_rate": 8.16961498439126e-05,
      "loss": 0.6575,
      "step": 1960
    },
    {
      "epoch": 6.0244648318042815,
      "grad_norm": 4.461253643035889,
      "learning_rate": 8.159209157127991e-05,
      "loss": 0.5831,
      "step": 1970
    },
    {
      "epoch": 6.055045871559633,
      "grad_norm": 4.997203350067139,
      "learning_rate": 8.148803329864725e-05,
      "loss": 0.5412,
      "step": 1980
    },
    {
      "epoch": 6.085626911314985,
      "grad_norm": 5.2825751304626465,
      "learning_rate": 8.138397502601457e-05,
      "loss": 0.5012,
      "step": 1990
    },
    {
      "epoch": 6.116207951070336,
      "grad_norm": 5.710301399230957,
      "learning_rate": 8.12799167533819e-05,
      "loss": 0.5548,
      "step": 2000
    },
    {
      "epoch": 6.116207951070336,
      "eval_loss": 0.5635126233100891,
      "eval_runtime": 17.154,
      "eval_samples_per_second": 16.964,
      "eval_steps_per_second": 8.511,
      "step": 2000
    },
    {
      "epoch": 6.146788990825688,
      "grad_norm": 6.4876179695129395,
      "learning_rate": 8.117585848074922e-05,
      "loss": 0.6589,
      "step": 2010
    },
    {
      "epoch": 6.17737003058104,
      "grad_norm": 3.520118474960327,
      "learning_rate": 8.107180020811655e-05,
      "loss": 0.5431,
      "step": 2020
    },
    {
      "epoch": 6.207951070336391,
      "grad_norm": 4.1320648193359375,
      "learning_rate": 8.096774193548387e-05,
      "loss": 0.5069,
      "step": 2030
    },
    {
      "epoch": 6.238532110091743,
      "grad_norm": 6.541991710662842,
      "learning_rate": 8.08636836628512e-05,
      "loss": 0.5477,
      "step": 2040
    },
    {
      "epoch": 6.269113149847095,
      "grad_norm": 6.10244607925415,
      "learning_rate": 8.075962539021852e-05,
      "loss": 0.5639,
      "step": 2050
    },
    {
      "epoch": 6.299694189602446,
      "grad_norm": 6.1276469230651855,
      "learning_rate": 8.065556711758585e-05,
      "loss": 0.5773,
      "step": 2060
    },
    {
      "epoch": 6.330275229357798,
      "grad_norm": 3.827613353729248,
      "learning_rate": 8.055150884495319e-05,
      "loss": 0.5283,
      "step": 2070
    },
    {
      "epoch": 6.36085626911315,
      "grad_norm": 5.0163421630859375,
      "learning_rate": 8.04474505723205e-05,
      "loss": 0.5174,
      "step": 2080
    },
    {
      "epoch": 6.391437308868501,
      "grad_norm": 5.943496227264404,
      "learning_rate": 8.034339229968782e-05,
      "loss": 0.5364,
      "step": 2090
    },
    {
      "epoch": 6.422018348623853,
      "grad_norm": 7.260120391845703,
      "learning_rate": 8.023933402705516e-05,
      "loss": 0.6289,
      "step": 2100
    },
    {
      "epoch": 6.4525993883792045,
      "grad_norm": 12.01782512664795,
      "learning_rate": 8.013527575442247e-05,
      "loss": 0.617,
      "step": 2110
    },
    {
      "epoch": 6.483180428134556,
      "grad_norm": 6.774069309234619,
      "learning_rate": 8.003121748178981e-05,
      "loss": 0.579,
      "step": 2120
    },
    {
      "epoch": 6.513761467889909,
      "grad_norm": 7.0621442794799805,
      "learning_rate": 7.992715920915714e-05,
      "loss": 0.4988,
      "step": 2130
    },
    {
      "epoch": 6.5443425076452595,
      "grad_norm": 9.868996620178223,
      "learning_rate": 7.982310093652445e-05,
      "loss": 0.53,
      "step": 2140
    },
    {
      "epoch": 6.574923547400612,
      "grad_norm": 20.389284133911133,
      "learning_rate": 7.971904266389179e-05,
      "loss": 0.5561,
      "step": 2150
    },
    {
      "epoch": 6.605504587155964,
      "grad_norm": 7.879746913909912,
      "learning_rate": 7.961498439125911e-05,
      "loss": 0.582,
      "step": 2160
    },
    {
      "epoch": 6.636085626911315,
      "grad_norm": 3.271695137023926,
      "learning_rate": 7.951092611862644e-05,
      "loss": 0.564,
      "step": 2170
    },
    {
      "epoch": 6.666666666666667,
      "grad_norm": 4.457621097564697,
      "learning_rate": 7.940686784599376e-05,
      "loss": 0.4886,
      "step": 2180
    },
    {
      "epoch": 6.697247706422019,
      "grad_norm": 4.690130710601807,
      "learning_rate": 7.930280957336107e-05,
      "loss": 0.4988,
      "step": 2190
    },
    {
      "epoch": 6.72782874617737,
      "grad_norm": 17.970327377319336,
      "learning_rate": 7.919875130072841e-05,
      "loss": 0.5308,
      "step": 2200
    },
    {
      "epoch": 6.758409785932722,
      "grad_norm": 8.012701034545898,
      "learning_rate": 7.909469302809574e-05,
      "loss": 0.6291,
      "step": 2210
    },
    {
      "epoch": 6.7889908256880735,
      "grad_norm": 4.772938251495361,
      "learning_rate": 7.899063475546306e-05,
      "loss": 0.58,
      "step": 2220
    },
    {
      "epoch": 6.819571865443425,
      "grad_norm": 3.9855034351348877,
      "learning_rate": 7.888657648283039e-05,
      "loss": 0.4787,
      "step": 2230
    },
    {
      "epoch": 6.850152905198777,
      "grad_norm": 10.868754386901855,
      "learning_rate": 7.878251821019771e-05,
      "loss": 0.5189,
      "step": 2240
    },
    {
      "epoch": 6.8807339449541285,
      "grad_norm": 4.059515476226807,
      "learning_rate": 7.867845993756504e-05,
      "loss": 0.5256,
      "step": 2250
    },
    {
      "epoch": 6.91131498470948,
      "grad_norm": 4.996955871582031,
      "learning_rate": 7.857440166493236e-05,
      "loss": 0.6207,
      "step": 2260
    },
    {
      "epoch": 6.941896024464832,
      "grad_norm": 3.1101465225219727,
      "learning_rate": 7.847034339229969e-05,
      "loss": 0.4958,
      "step": 2270
    },
    {
      "epoch": 6.972477064220183,
      "grad_norm": 8.754892349243164,
      "learning_rate": 7.836628511966701e-05,
      "loss": 0.5257,
      "step": 2280
    },
    {
      "epoch": 7.003058103975535,
      "grad_norm": 3.744485378265381,
      "learning_rate": 7.826222684703435e-05,
      "loss": 0.6034,
      "step": 2290
    },
    {
      "epoch": 7.033639143730887,
      "grad_norm": 4.851626396179199,
      "learning_rate": 7.815816857440166e-05,
      "loss": 0.3741,
      "step": 2300
    },
    {
      "epoch": 7.064220183486238,
      "grad_norm": 7.789270877838135,
      "learning_rate": 7.805411030176899e-05,
      "loss": 0.4436,
      "step": 2310
    },
    {
      "epoch": 7.09480122324159,
      "grad_norm": 7.35498571395874,
      "learning_rate": 7.795005202913633e-05,
      "loss": 0.4353,
      "step": 2320
    },
    {
      "epoch": 7.125382262996942,
      "grad_norm": 6.431797981262207,
      "learning_rate": 7.784599375650364e-05,
      "loss": 0.4976,
      "step": 2330
    },
    {
      "epoch": 7.155963302752293,
      "grad_norm": 4.586643218994141,
      "learning_rate": 7.774193548387098e-05,
      "loss": 0.6474,
      "step": 2340
    },
    {
      "epoch": 7.186544342507645,
      "grad_norm": 5.96403169631958,
      "learning_rate": 7.76378772112383e-05,
      "loss": 0.4536,
      "step": 2350
    },
    {
      "epoch": 7.217125382262997,
      "grad_norm": 4.443822860717773,
      "learning_rate": 7.753381893860561e-05,
      "loss": 0.4222,
      "step": 2360
    },
    {
      "epoch": 7.247706422018348,
      "grad_norm": 5.37199592590332,
      "learning_rate": 7.742976066597295e-05,
      "loss": 0.4821,
      "step": 2370
    },
    {
      "epoch": 7.2782874617737,
      "grad_norm": 5.589402675628662,
      "learning_rate": 7.732570239334028e-05,
      "loss": 0.4723,
      "step": 2380
    },
    {
      "epoch": 7.3088685015290515,
      "grad_norm": 2.493727207183838,
      "learning_rate": 7.72216441207076e-05,
      "loss": 0.5624,
      "step": 2390
    },
    {
      "epoch": 7.339449541284404,
      "grad_norm": 4.967494010925293,
      "learning_rate": 7.711758584807493e-05,
      "loss": 0.4189,
      "step": 2400
    },
    {
      "epoch": 7.370030581039756,
      "grad_norm": 4.298621654510498,
      "learning_rate": 7.701352757544225e-05,
      "loss": 0.4328,
      "step": 2410
    },
    {
      "epoch": 7.400611620795107,
      "grad_norm": 6.013545989990234,
      "learning_rate": 7.690946930280958e-05,
      "loss": 0.4803,
      "step": 2420
    },
    {
      "epoch": 7.431192660550459,
      "grad_norm": 5.232725620269775,
      "learning_rate": 7.68054110301769e-05,
      "loss": 0.4758,
      "step": 2430
    },
    {
      "epoch": 7.461773700305811,
      "grad_norm": 5.801144123077393,
      "learning_rate": 7.670135275754423e-05,
      "loss": 0.495,
      "step": 2440
    },
    {
      "epoch": 7.492354740061162,
      "grad_norm": 2.754412889480591,
      "learning_rate": 7.659729448491155e-05,
      "loss": 0.4522,
      "step": 2450
    },
    {
      "epoch": 7.522935779816514,
      "grad_norm": 6.403210163116455,
      "learning_rate": 7.649323621227889e-05,
      "loss": 0.4281,
      "step": 2460
    },
    {
      "epoch": 7.553516819571866,
      "grad_norm": 5.0712056159973145,
      "learning_rate": 7.63891779396462e-05,
      "loss": 0.4274,
      "step": 2470
    },
    {
      "epoch": 7.584097859327217,
      "grad_norm": 6.147752285003662,
      "learning_rate": 7.628511966701353e-05,
      "loss": 0.4729,
      "step": 2480
    },
    {
      "epoch": 7.614678899082569,
      "grad_norm": 3.0379574298858643,
      "learning_rate": 7.618106139438087e-05,
      "loss": 0.5294,
      "step": 2490
    },
    {
      "epoch": 7.6452599388379205,
      "grad_norm": 3.286498785018921,
      "learning_rate": 7.607700312174818e-05,
      "loss": 0.4543,
      "step": 2500
    },
    {
      "epoch": 7.6452599388379205,
      "eval_loss": 0.48799338936805725,
      "eval_runtime": 16.6393,
      "eval_samples_per_second": 17.489,
      "eval_steps_per_second": 8.774,
      "step": 2500
    },
    {
      "epoch": 7.675840978593272,
      "grad_norm": 9.105060577392578,
      "learning_rate": 7.597294484911552e-05,
      "loss": 0.428,
      "step": 2510
    },
    {
      "epoch": 7.706422018348624,
      "grad_norm": 4.746338844299316,
      "learning_rate": 7.586888657648283e-05,
      "loss": 0.4439,
      "step": 2520
    },
    {
      "epoch": 7.7370030581039755,
      "grad_norm": 4.770055770874023,
      "learning_rate": 7.576482830385015e-05,
      "loss": 0.4902,
      "step": 2530
    },
    {
      "epoch": 7.767584097859327,
      "grad_norm": 5.230956554412842,
      "learning_rate": 7.566077003121749e-05,
      "loss": 0.4818,
      "step": 2540
    },
    {
      "epoch": 7.798165137614679,
      "grad_norm": 5.195344924926758,
      "learning_rate": 7.55567117585848e-05,
      "loss": 0.4143,
      "step": 2550
    },
    {
      "epoch": 7.82874617737003,
      "grad_norm": 3.9105961322784424,
      "learning_rate": 7.545265348595214e-05,
      "loss": 0.4193,
      "step": 2560
    },
    {
      "epoch": 7.859327217125382,
      "grad_norm": 9.038330078125,
      "learning_rate": 7.534859521331947e-05,
      "loss": 0.4578,
      "step": 2570
    },
    {
      "epoch": 7.889908256880734,
      "grad_norm": 8.360684394836426,
      "learning_rate": 7.524453694068678e-05,
      "loss": 0.4504,
      "step": 2580
    },
    {
      "epoch": 7.920489296636085,
      "grad_norm": 4.143953800201416,
      "learning_rate": 7.514047866805412e-05,
      "loss": 0.5283,
      "step": 2590
    },
    {
      "epoch": 7.951070336391437,
      "grad_norm": 3.5100886821746826,
      "learning_rate": 7.503642039542144e-05,
      "loss": 0.45,
      "step": 2600
    },
    {
      "epoch": 7.981651376146789,
      "grad_norm": 4.23678731918335,
      "learning_rate": 7.493236212278877e-05,
      "loss": 0.4842,
      "step": 2610
    },
    {
      "epoch": 8.01223241590214,
      "grad_norm": 3.595038414001465,
      "learning_rate": 7.482830385015609e-05,
      "loss": 0.4797,
      "step": 2620
    },
    {
      "epoch": 8.042813455657493,
      "grad_norm": 4.2236504554748535,
      "learning_rate": 7.472424557752342e-05,
      "loss": 0.3455,
      "step": 2630
    },
    {
      "epoch": 8.073394495412844,
      "grad_norm": 3.604245901107788,
      "learning_rate": 7.462018730489074e-05,
      "loss": 0.3759,
      "step": 2640
    },
    {
      "epoch": 8.103975535168196,
      "grad_norm": 5.744846820831299,
      "learning_rate": 7.451612903225807e-05,
      "loss": 0.4431,
      "step": 2650
    },
    {
      "epoch": 8.134556574923547,
      "grad_norm": 6.683863639831543,
      "learning_rate": 7.441207075962539e-05,
      "loss": 0.5102,
      "step": 2660
    },
    {
      "epoch": 8.1651376146789,
      "grad_norm": 3.7429959774017334,
      "learning_rate": 7.430801248699272e-05,
      "loss": 0.5118,
      "step": 2670
    },
    {
      "epoch": 8.19571865443425,
      "grad_norm": 4.308508396148682,
      "learning_rate": 7.420395421436005e-05,
      "loss": 0.3776,
      "step": 2680
    },
    {
      "epoch": 8.226299694189603,
      "grad_norm": 4.801083564758301,
      "learning_rate": 7.409989594172737e-05,
      "loss": 0.378,
      "step": 2690
    },
    {
      "epoch": 8.256880733944953,
      "grad_norm": 5.005183219909668,
      "learning_rate": 7.399583766909469e-05,
      "loss": 0.4783,
      "step": 2700
    },
    {
      "epoch": 8.287461773700306,
      "grad_norm": 6.771042346954346,
      "learning_rate": 7.389177939646203e-05,
      "loss": 0.4222,
      "step": 2710
    },
    {
      "epoch": 8.318042813455657,
      "grad_norm": 3.7550055980682373,
      "learning_rate": 7.378772112382934e-05,
      "loss": 0.4088,
      "step": 2720
    },
    {
      "epoch": 8.34862385321101,
      "grad_norm": 3.91751766204834,
      "learning_rate": 7.368366285119668e-05,
      "loss": 0.3798,
      "step": 2730
    },
    {
      "epoch": 8.37920489296636,
      "grad_norm": 5.044673919677734,
      "learning_rate": 7.3579604578564e-05,
      "loss": 0.3783,
      "step": 2740
    },
    {
      "epoch": 8.409785932721713,
      "grad_norm": 10.467144012451172,
      "learning_rate": 7.347554630593132e-05,
      "loss": 0.4106,
      "step": 2750
    },
    {
      "epoch": 8.440366972477065,
      "grad_norm": 8.879876136779785,
      "learning_rate": 7.337148803329866e-05,
      "loss": 0.4369,
      "step": 2760
    },
    {
      "epoch": 8.470948012232416,
      "grad_norm": 2.9480531215667725,
      "learning_rate": 7.326742976066597e-05,
      "loss": 0.4218,
      "step": 2770
    },
    {
      "epoch": 8.501529051987767,
      "grad_norm": 5.187920093536377,
      "learning_rate": 7.31633714880333e-05,
      "loss": 0.3972,
      "step": 2780
    },
    {
      "epoch": 8.53211009174312,
      "grad_norm": 9.206551551818848,
      "learning_rate": 7.305931321540063e-05,
      "loss": 0.4019,
      "step": 2790
    },
    {
      "epoch": 8.562691131498472,
      "grad_norm": 15.84662914276123,
      "learning_rate": 7.295525494276796e-05,
      "loss": 0.3883,
      "step": 2800
    },
    {
      "epoch": 8.593272171253822,
      "grad_norm": 8.379307746887207,
      "learning_rate": 7.285119667013528e-05,
      "loss": 0.4943,
      "step": 2810
    },
    {
      "epoch": 8.623853211009175,
      "grad_norm": 4.381606578826904,
      "learning_rate": 7.27471383975026e-05,
      "loss": 0.4541,
      "step": 2820
    },
    {
      "epoch": 8.654434250764526,
      "grad_norm": 6.217578411102295,
      "learning_rate": 7.264308012486993e-05,
      "loss": 0.4034,
      "step": 2830
    },
    {
      "epoch": 8.685015290519878,
      "grad_norm": 4.155258655548096,
      "learning_rate": 7.253902185223726e-05,
      "loss": 0.3874,
      "step": 2840
    },
    {
      "epoch": 8.715596330275229,
      "grad_norm": 5.202000617980957,
      "learning_rate": 7.243496357960458e-05,
      "loss": 0.4272,
      "step": 2850
    },
    {
      "epoch": 8.746177370030582,
      "grad_norm": 7.137246131896973,
      "learning_rate": 7.23309053069719e-05,
      "loss": 0.4251,
      "step": 2860
    },
    {
      "epoch": 8.776758409785932,
      "grad_norm": 3.7905514240264893,
      "learning_rate": 7.222684703433923e-05,
      "loss": 0.5237,
      "step": 2870
    },
    {
      "epoch": 8.807339449541285,
      "grad_norm": 4.002864837646484,
      "learning_rate": 7.212278876170656e-05,
      "loss": 0.3235,
      "step": 2880
    },
    {
      "epoch": 8.837920489296636,
      "grad_norm": 9.481928825378418,
      "learning_rate": 7.201873048907388e-05,
      "loss": 0.4215,
      "step": 2890
    },
    {
      "epoch": 8.868501529051988,
      "grad_norm": 3.8038759231567383,
      "learning_rate": 7.191467221644122e-05,
      "loss": 0.3849,
      "step": 2900
    },
    {
      "epoch": 8.899082568807339,
      "grad_norm": 15.837183952331543,
      "learning_rate": 7.181061394380853e-05,
      "loss": 0.4646,
      "step": 2910
    },
    {
      "epoch": 8.929663608562691,
      "grad_norm": 5.814904689788818,
      "learning_rate": 7.170655567117586e-05,
      "loss": 0.4505,
      "step": 2920
    },
    {
      "epoch": 8.960244648318042,
      "grad_norm": 5.111677169799805,
      "learning_rate": 7.16024973985432e-05,
      "loss": 0.3565,
      "step": 2930
    },
    {
      "epoch": 8.990825688073395,
      "grad_norm": 20.767641067504883,
      "learning_rate": 7.14984391259105e-05,
      "loss": 0.4058,
      "step": 2940
    },
    {
      "epoch": 9.021406727828746,
      "grad_norm": 5.076265811920166,
      "learning_rate": 7.139438085327784e-05,
      "loss": 0.3846,
      "step": 2950
    },
    {
      "epoch": 9.051987767584098,
      "grad_norm": 3.0534474849700928,
      "learning_rate": 7.129032258064517e-05,
      "loss": 0.2875,
      "step": 2960
    },
    {
      "epoch": 9.082568807339449,
      "grad_norm": 3.4252254962921143,
      "learning_rate": 7.11862643080125e-05,
      "loss": 0.3941,
      "step": 2970
    },
    {
      "epoch": 9.113149847094801,
      "grad_norm": 7.061174392700195,
      "learning_rate": 7.108220603537982e-05,
      "loss": 0.439,
      "step": 2980
    },
    {
      "epoch": 9.143730886850152,
      "grad_norm": 5.309439659118652,
      "learning_rate": 7.097814776274714e-05,
      "loss": 0.3839,
      "step": 2990
    },
    {
      "epoch": 9.174311926605505,
      "grad_norm": 2.809702157974243,
      "learning_rate": 7.087408949011447e-05,
      "loss": 0.3664,
      "step": 3000
    },
    {
      "epoch": 9.174311926605505,
      "eval_loss": 0.48507484793663025,
      "eval_runtime": 17.7814,
      "eval_samples_per_second": 16.365,
      "eval_steps_per_second": 8.211,
      "step": 3000
    },
    {
      "epoch": 9.204892966360855,
      "grad_norm": 4.209444522857666,
      "learning_rate": 7.07700312174818e-05,
      "loss": 0.3277,
      "step": 3010
    },
    {
      "epoch": 9.235474006116208,
      "grad_norm": 4.111099720001221,
      "learning_rate": 7.066597294484912e-05,
      "loss": 0.3804,
      "step": 3020
    },
    {
      "epoch": 9.26605504587156,
      "grad_norm": 7.38637638092041,
      "learning_rate": 7.056191467221644e-05,
      "loss": 0.3965,
      "step": 3030
    },
    {
      "epoch": 9.296636085626911,
      "grad_norm": 4.757504463195801,
      "learning_rate": 7.045785639958377e-05,
      "loss": 0.3981,
      "step": 3040
    },
    {
      "epoch": 9.327217125382264,
      "grad_norm": 3.749709367752075,
      "learning_rate": 7.03537981269511e-05,
      "loss": 0.405,
      "step": 3050
    },
    {
      "epoch": 9.357798165137615,
      "grad_norm": 3.167997121810913,
      "learning_rate": 7.024973985431842e-05,
      "loss": 0.292,
      "step": 3060
    },
    {
      "epoch": 9.388379204892967,
      "grad_norm": 4.194080352783203,
      "learning_rate": 7.014568158168576e-05,
      "loss": 0.3458,
      "step": 3070
    },
    {
      "epoch": 9.418960244648318,
      "grad_norm": 4.001060962677002,
      "learning_rate": 7.004162330905307e-05,
      "loss": 0.3434,
      "step": 3080
    },
    {
      "epoch": 9.44954128440367,
      "grad_norm": 7.514442443847656,
      "learning_rate": 6.99375650364204e-05,
      "loss": 0.3827,
      "step": 3090
    },
    {
      "epoch": 9.480122324159021,
      "grad_norm": 3.734842300415039,
      "learning_rate": 6.983350676378772e-05,
      "loss": 0.3885,
      "step": 3100
    },
    {
      "epoch": 9.510703363914374,
      "grad_norm": 5.834413051605225,
      "learning_rate": 6.972944849115504e-05,
      "loss": 0.3644,
      "step": 3110
    },
    {
      "epoch": 9.541284403669724,
      "grad_norm": 13.83302116394043,
      "learning_rate": 6.962539021852238e-05,
      "loss": 0.3304,
      "step": 3120
    },
    {
      "epoch": 9.571865443425077,
      "grad_norm": 4.693499565124512,
      "learning_rate": 6.95213319458897e-05,
      "loss": 0.3597,
      "step": 3130
    },
    {
      "epoch": 9.602446483180428,
      "grad_norm": 8.545695304870605,
      "learning_rate": 6.941727367325703e-05,
      "loss": 0.4419,
      "step": 3140
    },
    {
      "epoch": 9.63302752293578,
      "grad_norm": 3.0196399688720703,
      "learning_rate": 6.931321540062436e-05,
      "loss": 0.3519,
      "step": 3150
    },
    {
      "epoch": 9.663608562691131,
      "grad_norm": 4.620004177093506,
      "learning_rate": 6.920915712799167e-05,
      "loss": 0.3464,
      "step": 3160
    },
    {
      "epoch": 9.694189602446484,
      "grad_norm": 3.939121961593628,
      "learning_rate": 6.910509885535901e-05,
      "loss": 0.3331,
      "step": 3170
    },
    {
      "epoch": 9.724770642201834,
      "grad_norm": 6.618518352508545,
      "learning_rate": 6.900104058272633e-05,
      "loss": 0.3621,
      "step": 3180
    },
    {
      "epoch": 9.755351681957187,
      "grad_norm": 6.519993782043457,
      "learning_rate": 6.889698231009366e-05,
      "loss": 0.3965,
      "step": 3190
    },
    {
      "epoch": 9.785932721712538,
      "grad_norm": 4.361577033996582,
      "learning_rate": 6.879292403746098e-05,
      "loss": 0.4112,
      "step": 3200
    },
    {
      "epoch": 9.81651376146789,
      "grad_norm": 4.898428916931152,
      "learning_rate": 6.868886576482831e-05,
      "loss": 0.3391,
      "step": 3210
    },
    {
      "epoch": 9.84709480122324,
      "grad_norm": 7.772880554199219,
      "learning_rate": 6.858480749219563e-05,
      "loss": 0.35,
      "step": 3220
    },
    {
      "epoch": 9.877675840978593,
      "grad_norm": 7.611955642700195,
      "learning_rate": 6.848074921956296e-05,
      "loss": 0.4289,
      "step": 3230
    },
    {
      "epoch": 9.908256880733944,
      "grad_norm": 4.031200885772705,
      "learning_rate": 6.837669094693028e-05,
      "loss": 0.4108,
      "step": 3240
    },
    {
      "epoch": 9.938837920489297,
      "grad_norm": 3.0720908641815186,
      "learning_rate": 6.827263267429761e-05,
      "loss": 0.371,
      "step": 3250
    },
    {
      "epoch": 9.96941896024465,
      "grad_norm": 9.898725509643555,
      "learning_rate": 6.816857440166493e-05,
      "loss": 0.3057,
      "step": 3260
    },
    {
      "epoch": 10.0,
      "grad_norm": 10.916779518127441,
      "learning_rate": 6.806451612903226e-05,
      "loss": 0.3364,
      "step": 3270
    },
    {
      "epoch": 10.030581039755353,
      "grad_norm": 7.887121200561523,
      "learning_rate": 6.796045785639958e-05,
      "loss": 0.257,
      "step": 3280
    },
    {
      "epoch": 10.061162079510703,
      "grad_norm": 5.914815425872803,
      "learning_rate": 6.785639958376692e-05,
      "loss": 0.3088,
      "step": 3290
    },
    {
      "epoch": 10.091743119266056,
      "grad_norm": 4.333220481872559,
      "learning_rate": 6.775234131113423e-05,
      "loss": 0.3107,
      "step": 3300
    },
    {
      "epoch": 10.122324159021407,
      "grad_norm": 3.6494174003601074,
      "learning_rate": 6.764828303850156e-05,
      "loss": 0.3632,
      "step": 3310
    },
    {
      "epoch": 10.15290519877676,
      "grad_norm": 10.518165588378906,
      "learning_rate": 6.75442247658689e-05,
      "loss": 0.4544,
      "step": 3320
    },
    {
      "epoch": 10.18348623853211,
      "grad_norm": 8.803187370300293,
      "learning_rate": 6.744016649323621e-05,
      "loss": 0.3239,
      "step": 3330
    },
    {
      "epoch": 10.214067278287462,
      "grad_norm": 4.997848033905029,
      "learning_rate": 6.733610822060355e-05,
      "loss": 0.2882,
      "step": 3340
    },
    {
      "epoch": 10.244648318042813,
      "grad_norm": 4.853428363800049,
      "learning_rate": 6.723204994797087e-05,
      "loss": 0.3277,
      "step": 3350
    },
    {
      "epoch": 10.275229357798166,
      "grad_norm": 8.042211532592773,
      "learning_rate": 6.71279916753382e-05,
      "loss": 0.3872,
      "step": 3360
    },
    {
      "epoch": 10.305810397553516,
      "grad_norm": 6.132227897644043,
      "learning_rate": 6.702393340270552e-05,
      "loss": 0.3562,
      "step": 3370
    },
    {
      "epoch": 10.336391437308869,
      "grad_norm": 3.8088772296905518,
      "learning_rate": 6.691987513007283e-05,
      "loss": 0.3054,
      "step": 3380
    },
    {
      "epoch": 10.36697247706422,
      "grad_norm": 6.178166389465332,
      "learning_rate": 6.681581685744017e-05,
      "loss": 0.3423,
      "step": 3390
    },
    {
      "epoch": 10.397553516819572,
      "grad_norm": 6.065817832946777,
      "learning_rate": 6.67117585848075e-05,
      "loss": 0.3252,
      "step": 3400
    },
    {
      "epoch": 10.428134556574923,
      "grad_norm": 5.649379730224609,
      "learning_rate": 6.660770031217482e-05,
      "loss": 0.3581,
      "step": 3410
    },
    {
      "epoch": 10.458715596330276,
      "grad_norm": 6.1458563804626465,
      "learning_rate": 6.650364203954215e-05,
      "loss": 0.453,
      "step": 3420
    },
    {
      "epoch": 10.489296636085626,
      "grad_norm": 3.4970128536224365,
      "learning_rate": 6.639958376690947e-05,
      "loss": 0.3148,
      "step": 3430
    },
    {
      "epoch": 10.519877675840979,
      "grad_norm": 2.8173370361328125,
      "learning_rate": 6.62955254942768e-05,
      "loss": 0.2592,
      "step": 3440
    },
    {
      "epoch": 10.55045871559633,
      "grad_norm": 8.554411888122559,
      "learning_rate": 6.619146722164412e-05,
      "loss": 0.3032,
      "step": 3450
    },
    {
      "epoch": 10.581039755351682,
      "grad_norm": 7.391510009765625,
      "learning_rate": 6.608740894901145e-05,
      "loss": 0.3607,
      "step": 3460
    },
    {
      "epoch": 10.611620795107033,
      "grad_norm": 7.969365119934082,
      "learning_rate": 6.598335067637877e-05,
      "loss": 0.3567,
      "step": 3470
    },
    {
      "epoch": 10.642201834862385,
      "grad_norm": 4.969130039215088,
      "learning_rate": 6.58792924037461e-05,
      "loss": 0.3078,
      "step": 3480
    },
    {
      "epoch": 10.672782874617736,
      "grad_norm": 4.509185314178467,
      "learning_rate": 6.577523413111342e-05,
      "loss": 0.3299,
      "step": 3490
    },
    {
      "epoch": 10.703363914373089,
      "grad_norm": 4.516345024108887,
      "learning_rate": 6.567117585848075e-05,
      "loss": 0.3525,
      "step": 3500
    },
    {
      "epoch": 10.703363914373089,
      "eval_loss": 0.45856353640556335,
      "eval_runtime": 17.4638,
      "eval_samples_per_second": 16.663,
      "eval_steps_per_second": 8.36,
      "step": 3500
    },
    {
      "epoch": 10.73394495412844,
      "grad_norm": 6.109631538391113,
      "learning_rate": 6.556711758584809e-05,
      "loss": 0.35,
      "step": 3510
    },
    {
      "epoch": 10.764525993883792,
      "grad_norm": 5.233398914337158,
      "learning_rate": 6.54630593132154e-05,
      "loss": 0.3717,
      "step": 3520
    },
    {
      "epoch": 10.795107033639145,
      "grad_norm": 3.2442190647125244,
      "learning_rate": 6.535900104058274e-05,
      "loss": 0.2845,
      "step": 3530
    },
    {
      "epoch": 10.825688073394495,
      "grad_norm": 4.037608623504639,
      "learning_rate": 6.525494276795006e-05,
      "loss": 0.2987,
      "step": 3540
    },
    {
      "epoch": 10.856269113149848,
      "grad_norm": 3.441417932510376,
      "learning_rate": 6.515088449531737e-05,
      "loss": 0.34,
      "step": 3550
    },
    {
      "epoch": 10.886850152905199,
      "grad_norm": 7.213795185089111,
      "learning_rate": 6.504682622268471e-05,
      "loss": 0.3387,
      "step": 3560
    },
    {
      "epoch": 10.917431192660551,
      "grad_norm": 5.970945358276367,
      "learning_rate": 6.494276795005204e-05,
      "loss": 0.4408,
      "step": 3570
    },
    {
      "epoch": 10.948012232415902,
      "grad_norm": 6.310761451721191,
      "learning_rate": 6.483870967741936e-05,
      "loss": 0.2882,
      "step": 3580
    },
    {
      "epoch": 10.978593272171254,
      "grad_norm": 12.499755859375,
      "learning_rate": 6.473465140478669e-05,
      "loss": 0.2837,
      "step": 3590
    },
    {
      "epoch": 11.009174311926605,
      "grad_norm": 4.7203369140625,
      "learning_rate": 6.463059313215401e-05,
      "loss": 0.3552,
      "step": 3600
    },
    {
      "epoch": 11.039755351681958,
      "grad_norm": 5.568735599517822,
      "learning_rate": 6.452653485952134e-05,
      "loss": 0.2643,
      "step": 3610
    },
    {
      "epoch": 11.070336391437309,
      "grad_norm": 5.771465301513672,
      "learning_rate": 6.442247658688866e-05,
      "loss": 0.2558,
      "step": 3620
    },
    {
      "epoch": 11.100917431192661,
      "grad_norm": 3.8241806030273438,
      "learning_rate": 6.431841831425599e-05,
      "loss": 0.2871,
      "step": 3630
    },
    {
      "epoch": 11.131498470948012,
      "grad_norm": 4.5878777503967285,
      "learning_rate": 6.421436004162331e-05,
      "loss": 0.3779,
      "step": 3640
    },
    {
      "epoch": 11.162079510703364,
      "grad_norm": 4.807687759399414,
      "learning_rate": 6.411030176899064e-05,
      "loss": 0.3694,
      "step": 3650
    },
    {
      "epoch": 11.192660550458715,
      "grad_norm": 4.102357387542725,
      "learning_rate": 6.400624349635796e-05,
      "loss": 0.2575,
      "step": 3660
    },
    {
      "epoch": 11.223241590214068,
      "grad_norm": 4.665652275085449,
      "learning_rate": 6.390218522372529e-05,
      "loss": 0.3022,
      "step": 3670
    },
    {
      "epoch": 11.253822629969418,
      "grad_norm": 4.404171943664551,
      "learning_rate": 6.379812695109262e-05,
      "loss": 0.2951,
      "step": 3680
    },
    {
      "epoch": 11.284403669724771,
      "grad_norm": 6.702642917633057,
      "learning_rate": 6.369406867845994e-05,
      "loss": 0.2925,
      "step": 3690
    },
    {
      "epoch": 11.314984709480122,
      "grad_norm": 3.5163631439208984,
      "learning_rate": 6.359001040582727e-05,
      "loss": 0.3496,
      "step": 3700
    },
    {
      "epoch": 11.345565749235474,
      "grad_norm": 4.392800331115723,
      "learning_rate": 6.348595213319459e-05,
      "loss": 0.2605,
      "step": 3710
    },
    {
      "epoch": 11.376146788990825,
      "grad_norm": 5.423502445220947,
      "learning_rate": 6.338189386056191e-05,
      "loss": 0.2948,
      "step": 3720
    },
    {
      "epoch": 11.406727828746178,
      "grad_norm": 4.449337959289551,
      "learning_rate": 6.327783558792925e-05,
      "loss": 0.2971,
      "step": 3730
    },
    {
      "epoch": 11.437308868501528,
      "grad_norm": 6.444003582000732,
      "learning_rate": 6.317377731529656e-05,
      "loss": 0.3255,
      "step": 3740
    },
    {
      "epoch": 11.46788990825688,
      "grad_norm": 2.682093858718872,
      "learning_rate": 6.30697190426639e-05,
      "loss": 0.3969,
      "step": 3750
    },
    {
      "epoch": 11.498470948012232,
      "grad_norm": 4.033663272857666,
      "learning_rate": 6.296566077003122e-05,
      "loss": 0.2669,
      "step": 3760
    },
    {
      "epoch": 11.529051987767584,
      "grad_norm": 3.125478506088257,
      "learning_rate": 6.286160249739854e-05,
      "loss": 0.3093,
      "step": 3770
    },
    {
      "epoch": 11.559633027522935,
      "grad_norm": 5.048392295837402,
      "learning_rate": 6.275754422476587e-05,
      "loss": 0.2829,
      "step": 3780
    },
    {
      "epoch": 11.590214067278287,
      "grad_norm": 3.8091633319854736,
      "learning_rate": 6.26534859521332e-05,
      "loss": 0.3207,
      "step": 3790
    },
    {
      "epoch": 11.62079510703364,
      "grad_norm": 5.138482093811035,
      "learning_rate": 6.254942767950052e-05,
      "loss": 0.3514,
      "step": 3800
    },
    {
      "epoch": 11.65137614678899,
      "grad_norm": 3.7958028316497803,
      "learning_rate": 6.244536940686785e-05,
      "loss": 0.2416,
      "step": 3810
    },
    {
      "epoch": 11.681957186544343,
      "grad_norm": 5.143332481384277,
      "learning_rate": 6.234131113423518e-05,
      "loss": 0.2647,
      "step": 3820
    },
    {
      "epoch": 11.712538226299694,
      "grad_norm": 6.322818756103516,
      "learning_rate": 6.22372528616025e-05,
      "loss": 0.323,
      "step": 3830
    },
    {
      "epoch": 11.743119266055047,
      "grad_norm": 5.7363996505737305,
      "learning_rate": 6.213319458896983e-05,
      "loss": 0.3049,
      "step": 3840
    },
    {
      "epoch": 11.773700305810397,
      "grad_norm": 4.893774032592773,
      "learning_rate": 6.202913631633715e-05,
      "loss": 0.3502,
      "step": 3850
    },
    {
      "epoch": 11.80428134556575,
      "grad_norm": 3.848017930984497,
      "learning_rate": 6.192507804370448e-05,
      "loss": 0.2771,
      "step": 3860
    },
    {
      "epoch": 11.8348623853211,
      "grad_norm": 5.943398952484131,
      "learning_rate": 6.182101977107181e-05,
      "loss": 0.301,
      "step": 3870
    },
    {
      "epoch": 11.865443425076453,
      "grad_norm": 3.520704984664917,
      "learning_rate": 6.171696149843913e-05,
      "loss": 0.2926,
      "step": 3880
    },
    {
      "epoch": 11.896024464831804,
      "grad_norm": 4.850738048553467,
      "learning_rate": 6.161290322580645e-05,
      "loss": 0.2805,
      "step": 3890
    },
    {
      "epoch": 11.926605504587156,
      "grad_norm": 2.6821794509887695,
      "learning_rate": 6.150884495317379e-05,
      "loss": 0.3268,
      "step": 3900
    },
    {
      "epoch": 11.957186544342507,
      "grad_norm": 4.901425838470459,
      "learning_rate": 6.14047866805411e-05,
      "loss": 0.2954,
      "step": 3910
    },
    {
      "epoch": 11.98776758409786,
      "grad_norm": 4.856095314025879,
      "learning_rate": 6.130072840790844e-05,
      "loss": 0.3739,
      "step": 3920
    },
    {
      "epoch": 12.01834862385321,
      "grad_norm": 5.4404616355896,
      "learning_rate": 6.119667013527576e-05,
      "loss": 0.28,
      "step": 3930
    },
    {
      "epoch": 12.048929663608563,
      "grad_norm": 4.431248188018799,
      "learning_rate": 6.109261186264308e-05,
      "loss": 0.2479,
      "step": 3940
    },
    {
      "epoch": 12.079510703363914,
      "grad_norm": 7.073538780212402,
      "learning_rate": 6.0988553590010414e-05,
      "loss": 0.266,
      "step": 3950
    },
    {
      "epoch": 12.110091743119266,
      "grad_norm": 6.206909656524658,
      "learning_rate": 6.088449531737773e-05,
      "loss": 0.3466,
      "step": 3960
    },
    {
      "epoch": 12.140672782874617,
      "grad_norm": 6.207326412200928,
      "learning_rate": 6.0780437044745064e-05,
      "loss": 0.3262,
      "step": 3970
    },
    {
      "epoch": 12.17125382262997,
      "grad_norm": 3.017399787902832,
      "learning_rate": 6.067637877211239e-05,
      "loss": 0.2778,
      "step": 3980
    },
    {
      "epoch": 12.20183486238532,
      "grad_norm": 3.116441488265991,
      "learning_rate": 6.057232049947971e-05,
      "loss": 0.2127,
      "step": 3990
    },
    {
      "epoch": 12.232415902140673,
      "grad_norm": 5.957989692687988,
      "learning_rate": 6.046826222684704e-05,
      "loss": 0.2901,
      "step": 4000
    },
    {
      "epoch": 12.232415902140673,
      "eval_loss": 0.4600982069969177,
      "eval_runtime": 16.8374,
      "eval_samples_per_second": 17.283,
      "eval_steps_per_second": 8.671,
      "step": 4000
    },
    {
      "epoch": 12.262996941896024,
      "grad_norm": 3.1144094467163086,
      "learning_rate": 6.036420395421436e-05,
      "loss": 0.2878,
      "step": 4010
    },
    {
      "epoch": 12.293577981651376,
      "grad_norm": 5.286678314208984,
      "learning_rate": 6.026014568158169e-05,
      "loss": 0.298,
      "step": 4020
    },
    {
      "epoch": 12.324159021406729,
      "grad_norm": 3.7269203662872314,
      "learning_rate": 6.0156087408949014e-05,
      "loss": 0.2888,
      "step": 4030
    },
    {
      "epoch": 12.35474006116208,
      "grad_norm": 3.3287158012390137,
      "learning_rate": 6.005202913631633e-05,
      "loss": 0.2063,
      "step": 4040
    },
    {
      "epoch": 12.385321100917432,
      "grad_norm": 4.839067459106445,
      "learning_rate": 5.9947970863683664e-05,
      "loss": 0.2779,
      "step": 4050
    },
    {
      "epoch": 12.415902140672783,
      "grad_norm": 5.016460418701172,
      "learning_rate": 5.984391259105099e-05,
      "loss": 0.2527,
      "step": 4060
    },
    {
      "epoch": 12.446483180428135,
      "grad_norm": 5.951940536499023,
      "learning_rate": 5.973985431841832e-05,
      "loss": 0.3517,
      "step": 4070
    },
    {
      "epoch": 12.477064220183486,
      "grad_norm": 3.960165023803711,
      "learning_rate": 5.963579604578564e-05,
      "loss": 0.317,
      "step": 4080
    },
    {
      "epoch": 12.507645259938839,
      "grad_norm": 2.830717086791992,
      "learning_rate": 5.953173777315297e-05,
      "loss": 0.3019,
      "step": 4090
    },
    {
      "epoch": 12.53822629969419,
      "grad_norm": 2.6632418632507324,
      "learning_rate": 5.9427679500520296e-05,
      "loss": 0.2186,
      "step": 4100
    },
    {
      "epoch": 12.568807339449542,
      "grad_norm": 5.966037273406982,
      "learning_rate": 5.9323621227887614e-05,
      "loss": 0.2752,
      "step": 4110
    },
    {
      "epoch": 12.599388379204893,
      "grad_norm": 7.089763164520264,
      "learning_rate": 5.9219562955254946e-05,
      "loss": 0.3096,
      "step": 4120
    },
    {
      "epoch": 12.629969418960245,
      "grad_norm": 2.9808754920959473,
      "learning_rate": 5.911550468262227e-05,
      "loss": 0.262,
      "step": 4130
    },
    {
      "epoch": 12.660550458715596,
      "grad_norm": 3.30267596244812,
      "learning_rate": 5.90114464099896e-05,
      "loss": 0.2317,
      "step": 4140
    },
    {
      "epoch": 12.691131498470948,
      "grad_norm": 3.521730422973633,
      "learning_rate": 5.890738813735692e-05,
      "loss": 0.2719,
      "step": 4150
    },
    {
      "epoch": 12.7217125382263,
      "grad_norm": 4.6262078285217285,
      "learning_rate": 5.8803329864724246e-05,
      "loss": 0.2598,
      "step": 4160
    },
    {
      "epoch": 12.752293577981652,
      "grad_norm": 8.410408973693848,
      "learning_rate": 5.869927159209158e-05,
      "loss": 0.3356,
      "step": 4170
    },
    {
      "epoch": 12.782874617737003,
      "grad_norm": 3.9074435234069824,
      "learning_rate": 5.8595213319458896e-05,
      "loss": 0.2509,
      "step": 4180
    },
    {
      "epoch": 12.813455657492355,
      "grad_norm": 3.5229334831237793,
      "learning_rate": 5.849115504682623e-05,
      "loss": 0.2223,
      "step": 4190
    },
    {
      "epoch": 12.844036697247706,
      "grad_norm": 6.239946365356445,
      "learning_rate": 5.838709677419355e-05,
      "loss": 0.2412,
      "step": 4200
    },
    {
      "epoch": 12.874617737003058,
      "grad_norm": 6.538459300994873,
      "learning_rate": 5.828303850156087e-05,
      "loss": 0.259,
      "step": 4210
    },
    {
      "epoch": 12.905198776758409,
      "grad_norm": 5.794137954711914,
      "learning_rate": 5.81789802289282e-05,
      "loss": 0.2934,
      "step": 4220
    },
    {
      "epoch": 12.935779816513762,
      "grad_norm": 4.787842273712158,
      "learning_rate": 5.807492195629553e-05,
      "loss": 0.3066,
      "step": 4230
    },
    {
      "epoch": 12.966360856269112,
      "grad_norm": 2.687736988067627,
      "learning_rate": 5.797086368366286e-05,
      "loss": 0.2459,
      "step": 4240
    },
    {
      "epoch": 12.996941896024465,
      "grad_norm": 4.587925910949707,
      "learning_rate": 5.786680541103018e-05,
      "loss": 0.2548,
      "step": 4250
    },
    {
      "epoch": 13.027522935779816,
      "grad_norm": 3.2574503421783447,
      "learning_rate": 5.776274713839751e-05,
      "loss": 0.2463,
      "step": 4260
    },
    {
      "epoch": 13.058103975535168,
      "grad_norm": 4.4264397621154785,
      "learning_rate": 5.7658688865764835e-05,
      "loss": 0.2214,
      "step": 4270
    },
    {
      "epoch": 13.08868501529052,
      "grad_norm": 3.6344454288482666,
      "learning_rate": 5.755463059313215e-05,
      "loss": 0.2442,
      "step": 4280
    },
    {
      "epoch": 13.119266055045872,
      "grad_norm": 5.120426654815674,
      "learning_rate": 5.7450572320499485e-05,
      "loss": 0.2977,
      "step": 4290
    },
    {
      "epoch": 13.149847094801224,
      "grad_norm": 5.738251686096191,
      "learning_rate": 5.73465140478668e-05,
      "loss": 0.3104,
      "step": 4300
    },
    {
      "epoch": 13.180428134556575,
      "grad_norm": 3.5654373168945312,
      "learning_rate": 5.724245577523414e-05,
      "loss": 0.2432,
      "step": 4310
    },
    {
      "epoch": 13.211009174311927,
      "grad_norm": 4.370543956756592,
      "learning_rate": 5.713839750260146e-05,
      "loss": 0.2373,
      "step": 4320
    },
    {
      "epoch": 13.241590214067278,
      "grad_norm": 6.025704383850098,
      "learning_rate": 5.703433922996878e-05,
      "loss": 0.2658,
      "step": 4330
    },
    {
      "epoch": 13.27217125382263,
      "grad_norm": 4.021401882171631,
      "learning_rate": 5.693028095733611e-05,
      "loss": 0.2805,
      "step": 4340
    },
    {
      "epoch": 13.302752293577981,
      "grad_norm": 8.499091148376465,
      "learning_rate": 5.6826222684703435e-05,
      "loss": 0.3026,
      "step": 4350
    },
    {
      "epoch": 13.333333333333334,
      "grad_norm": 2.7453525066375732,
      "learning_rate": 5.672216441207077e-05,
      "loss": 0.2622,
      "step": 4360
    },
    {
      "epoch": 13.363914373088685,
      "grad_norm": 5.434438228607178,
      "learning_rate": 5.6618106139438085e-05,
      "loss": 0.2392,
      "step": 4370
    },
    {
      "epoch": 13.394495412844037,
      "grad_norm": 6.497437000274658,
      "learning_rate": 5.651404786680541e-05,
      "loss": 0.243,
      "step": 4380
    },
    {
      "epoch": 13.425076452599388,
      "grad_norm": 5.23800802230835,
      "learning_rate": 5.640998959417274e-05,
      "loss": 0.3129,
      "step": 4390
    },
    {
      "epoch": 13.45565749235474,
      "grad_norm": 4.006523609161377,
      "learning_rate": 5.630593132154006e-05,
      "loss": 0.2886,
      "step": 4400
    },
    {
      "epoch": 13.486238532110091,
      "grad_norm": 9.827424049377441,
      "learning_rate": 5.620187304890739e-05,
      "loss": 0.2281,
      "step": 4410
    },
    {
      "epoch": 13.516819571865444,
      "grad_norm": 4.516007900238037,
      "learning_rate": 5.609781477627472e-05,
      "loss": 0.2079,
      "step": 4420
    },
    {
      "epoch": 13.547400611620795,
      "grad_norm": 4.366119384765625,
      "learning_rate": 5.599375650364205e-05,
      "loss": 0.2361,
      "step": 4430
    },
    {
      "epoch": 13.577981651376147,
      "grad_norm": 5.7396111488342285,
      "learning_rate": 5.588969823100937e-05,
      "loss": 0.2765,
      "step": 4440
    },
    {
      "epoch": 13.608562691131498,
      "grad_norm": 4.654730796813965,
      "learning_rate": 5.578563995837669e-05,
      "loss": 0.3094,
      "step": 4450
    },
    {
      "epoch": 13.63914373088685,
      "grad_norm": 2.6884396076202393,
      "learning_rate": 5.5681581685744024e-05,
      "loss": 0.2124,
      "step": 4460
    },
    {
      "epoch": 13.669724770642201,
      "grad_norm": 6.302408695220947,
      "learning_rate": 5.557752341311134e-05,
      "loss": 0.2303,
      "step": 4470
    },
    {
      "epoch": 13.700305810397554,
      "grad_norm": 2.9143168926239014,
      "learning_rate": 5.5473465140478674e-05,
      "loss": 0.2771,
      "step": 4480
    },
    {
      "epoch": 13.730886850152904,
      "grad_norm": 4.365865707397461,
      "learning_rate": 5.5369406867846e-05,
      "loss": 0.2519,
      "step": 4490
    },
    {
      "epoch": 13.761467889908257,
      "grad_norm": 5.209056854248047,
      "learning_rate": 5.526534859521332e-05,
      "loss": 0.3578,
      "step": 4500
    },
    {
      "epoch": 13.761467889908257,
      "eval_loss": 0.44830259680747986,
      "eval_runtime": 17.1439,
      "eval_samples_per_second": 16.974,
      "eval_steps_per_second": 8.516,
      "step": 4500
    },
    {
      "epoch": 13.792048929663608,
      "grad_norm": 3.6942572593688965,
      "learning_rate": 5.516129032258065e-05,
      "loss": 0.2407,
      "step": 4510
    },
    {
      "epoch": 13.82262996941896,
      "grad_norm": 3.4600281715393066,
      "learning_rate": 5.5057232049947974e-05,
      "loss": 0.2362,
      "step": 4520
    },
    {
      "epoch": 13.853211009174313,
      "grad_norm": 5.58060359954834,
      "learning_rate": 5.4953173777315306e-05,
      "loss": 0.2432,
      "step": 4530
    },
    {
      "epoch": 13.883792048929664,
      "grad_norm": 6.685091018676758,
      "learning_rate": 5.4849115504682624e-05,
      "loss": 0.2267,
      "step": 4540
    },
    {
      "epoch": 13.914373088685016,
      "grad_norm": 5.045437335968018,
      "learning_rate": 5.474505723204994e-05,
      "loss": 0.3498,
      "step": 4550
    },
    {
      "epoch": 13.944954128440367,
      "grad_norm": 3.360764741897583,
      "learning_rate": 5.464099895941728e-05,
      "loss": 0.2323,
      "step": 4560
    },
    {
      "epoch": 13.97553516819572,
      "grad_norm": 8.18934440612793,
      "learning_rate": 5.45369406867846e-05,
      "loss": 0.2403,
      "step": 4570
    },
    {
      "epoch": 14.00611620795107,
      "grad_norm": 2.808126926422119,
      "learning_rate": 5.443288241415193e-05,
      "loss": 0.3151,
      "step": 4580
    },
    {
      "epoch": 14.036697247706423,
      "grad_norm": 4.768661975860596,
      "learning_rate": 5.432882414151925e-05,
      "loss": 0.2096,
      "step": 4590
    },
    {
      "epoch": 14.067278287461773,
      "grad_norm": 3.9848787784576416,
      "learning_rate": 5.422476586888659e-05,
      "loss": 0.187,
      "step": 4600
    },
    {
      "epoch": 14.097859327217126,
      "grad_norm": 8.465381622314453,
      "learning_rate": 5.4120707596253906e-05,
      "loss": 0.2029,
      "step": 4610
    },
    {
      "epoch": 14.128440366972477,
      "grad_norm": 5.414556503295898,
      "learning_rate": 5.4016649323621224e-05,
      "loss": 0.2401,
      "step": 4620
    },
    {
      "epoch": 14.15902140672783,
      "grad_norm": 3.5335428714752197,
      "learning_rate": 5.3912591050988556e-05,
      "loss": 0.3059,
      "step": 4630
    },
    {
      "epoch": 14.18960244648318,
      "grad_norm": 3.0602972507476807,
      "learning_rate": 5.380853277835588e-05,
      "loss": 0.2089,
      "step": 4640
    },
    {
      "epoch": 14.220183486238533,
      "grad_norm": 4.8753557205200195,
      "learning_rate": 5.370447450572321e-05,
      "loss": 0.2213,
      "step": 4650
    },
    {
      "epoch": 14.250764525993883,
      "grad_norm": 4.4842987060546875,
      "learning_rate": 5.360041623309053e-05,
      "loss": 0.2155,
      "step": 4660
    },
    {
      "epoch": 14.281345565749236,
      "grad_norm": 4.742305278778076,
      "learning_rate": 5.3496357960457856e-05,
      "loss": 0.2515,
      "step": 4670
    },
    {
      "epoch": 14.311926605504587,
      "grad_norm": 2.4470579624176025,
      "learning_rate": 5.339229968782519e-05,
      "loss": 0.3141,
      "step": 4680
    },
    {
      "epoch": 14.34250764525994,
      "grad_norm": 2.6575069427490234,
      "learning_rate": 5.3288241415192506e-05,
      "loss": 0.1942,
      "step": 4690
    },
    {
      "epoch": 14.37308868501529,
      "grad_norm": 6.382633686065674,
      "learning_rate": 5.318418314255984e-05,
      "loss": 0.24,
      "step": 4700
    },
    {
      "epoch": 14.403669724770642,
      "grad_norm": 4.684356212615967,
      "learning_rate": 5.308012486992716e-05,
      "loss": 0.2164,
      "step": 4710
    },
    {
      "epoch": 14.434250764525993,
      "grad_norm": 6.916840076446533,
      "learning_rate": 5.297606659729448e-05,
      "loss": 0.2048,
      "step": 4720
    },
    {
      "epoch": 14.464831804281346,
      "grad_norm": 3.9860119819641113,
      "learning_rate": 5.287200832466181e-05,
      "loss": 0.2869,
      "step": 4730
    },
    {
      "epoch": 14.495412844036696,
      "grad_norm": 4.020848274230957,
      "learning_rate": 5.276795005202914e-05,
      "loss": 0.204,
      "step": 4740
    },
    {
      "epoch": 14.525993883792049,
      "grad_norm": 4.997827053070068,
      "learning_rate": 5.266389177939647e-05,
      "loss": 0.2295,
      "step": 4750
    },
    {
      "epoch": 14.5565749235474,
      "grad_norm": 4.601790904998779,
      "learning_rate": 5.255983350676379e-05,
      "loss": 0.2526,
      "step": 4760
    },
    {
      "epoch": 14.587155963302752,
      "grad_norm": 5.125568389892578,
      "learning_rate": 5.245577523413111e-05,
      "loss": 0.2559,
      "step": 4770
    },
    {
      "epoch": 14.617737003058103,
      "grad_norm": 4.057760715484619,
      "learning_rate": 5.2351716961498445e-05,
      "loss": 0.3112,
      "step": 4780
    },
    {
      "epoch": 14.648318042813456,
      "grad_norm": 3.295330047607422,
      "learning_rate": 5.224765868886576e-05,
      "loss": 0.2071,
      "step": 4790
    },
    {
      "epoch": 14.678899082568808,
      "grad_norm": 3.893869638442993,
      "learning_rate": 5.2143600416233095e-05,
      "loss": 0.1763,
      "step": 4800
    },
    {
      "epoch": 14.709480122324159,
      "grad_norm": 8.921279907226562,
      "learning_rate": 5.203954214360042e-05,
      "loss": 0.2331,
      "step": 4810
    },
    {
      "epoch": 14.740061162079511,
      "grad_norm": 3.6646783351898193,
      "learning_rate": 5.193548387096775e-05,
      "loss": 0.246,
      "step": 4820
    },
    {
      "epoch": 14.770642201834862,
      "grad_norm": 2.544281482696533,
      "learning_rate": 5.183142559833507e-05,
      "loss": 0.3041,
      "step": 4830
    },
    {
      "epoch": 14.801223241590215,
      "grad_norm": 2.496971607208252,
      "learning_rate": 5.1727367325702395e-05,
      "loss": 0.1993,
      "step": 4840
    },
    {
      "epoch": 14.831804281345565,
      "grad_norm": 5.756928443908691,
      "learning_rate": 5.162330905306973e-05,
      "loss": 0.2215,
      "step": 4850
    },
    {
      "epoch": 14.862385321100918,
      "grad_norm": 4.231186866760254,
      "learning_rate": 5.1519250780437045e-05,
      "loss": 0.1927,
      "step": 4860
    },
    {
      "epoch": 14.892966360856269,
      "grad_norm": 9.099519729614258,
      "learning_rate": 5.141519250780438e-05,
      "loss": 0.2556,
      "step": 4870
    },
    {
      "epoch": 14.923547400611621,
      "grad_norm": 3.435333728790283,
      "learning_rate": 5.13111342351717e-05,
      "loss": 0.2672,
      "step": 4880
    },
    {
      "epoch": 14.954128440366972,
      "grad_norm": 4.523071765899658,
      "learning_rate": 5.120707596253902e-05,
      "loss": 0.2166,
      "step": 4890
    },
    {
      "epoch": 14.984709480122325,
      "grad_norm": 7.479111671447754,
      "learning_rate": 5.110301768990635e-05,
      "loss": 0.2489,
      "step": 4900
    },
    {
      "epoch": 15.015290519877675,
      "grad_norm": 6.366982936859131,
      "learning_rate": 5.099895941727367e-05,
      "loss": 0.2488,
      "step": 4910
    },
    {
      "epoch": 15.045871559633028,
      "grad_norm": 3.5568156242370605,
      "learning_rate": 5.0894901144641e-05,
      "loss": 0.1866,
      "step": 4920
    },
    {
      "epoch": 15.076452599388379,
      "grad_norm": 4.744721412658691,
      "learning_rate": 5.079084287200833e-05,
      "loss": 0.2186,
      "step": 4930
    },
    {
      "epoch": 15.107033639143731,
      "grad_norm": 4.74827241897583,
      "learning_rate": 5.0686784599375645e-05,
      "loss": 0.2372,
      "step": 4940
    },
    {
      "epoch": 15.137614678899082,
      "grad_norm": 14.134720802307129,
      "learning_rate": 5.058272632674298e-05,
      "loss": 0.235,
      "step": 4950
    },
    {
      "epoch": 15.168195718654435,
      "grad_norm": 5.064000606536865,
      "learning_rate": 5.04786680541103e-05,
      "loss": 0.2712,
      "step": 4960
    },
    {
      "epoch": 15.198776758409785,
      "grad_norm": 3.46882700920105,
      "learning_rate": 5.0374609781477634e-05,
      "loss": 0.1872,
      "step": 4970
    },
    {
      "epoch": 15.229357798165138,
      "grad_norm": 6.42874002456665,
      "learning_rate": 5.027055150884495e-05,
      "loss": 0.1819,
      "step": 4980
    },
    {
      "epoch": 15.259938837920489,
      "grad_norm": 8.690680503845215,
      "learning_rate": 5.0166493236212284e-05,
      "loss": 0.2267,
      "step": 4990
    },
    {
      "epoch": 15.290519877675841,
      "grad_norm": 5.257503032684326,
      "learning_rate": 5.006243496357961e-05,
      "loss": 0.218,
      "step": 5000
    },
    {
      "epoch": 15.290519877675841,
      "eval_loss": 0.45569944381713867,
      "eval_runtime": 17.4347,
      "eval_samples_per_second": 16.691,
      "eval_steps_per_second": 8.374,
      "step": 5000
    },
    {
      "epoch": 15.321100917431192,
      "grad_norm": 3.1005799770355225,
      "learning_rate": 4.9958376690946934e-05,
      "loss": 0.2364,
      "step": 5010
    },
    {
      "epoch": 15.351681957186544,
      "grad_norm": 6.173751354217529,
      "learning_rate": 4.985431841831426e-05,
      "loss": 0.1919,
      "step": 5020
    },
    {
      "epoch": 15.382262996941897,
      "grad_norm": 3.07297682762146,
      "learning_rate": 4.9750260145681584e-05,
      "loss": 0.154,
      "step": 5030
    },
    {
      "epoch": 15.412844036697248,
      "grad_norm": 5.775254249572754,
      "learning_rate": 4.964620187304891e-05,
      "loss": 0.2291,
      "step": 5040
    },
    {
      "epoch": 15.4434250764526,
      "grad_norm": 3.3316097259521484,
      "learning_rate": 4.9542143600416234e-05,
      "loss": 0.2434,
      "step": 5050
    },
    {
      "epoch": 15.474006116207951,
      "grad_norm": 3.749220848083496,
      "learning_rate": 4.943808532778356e-05,
      "loss": 0.2795,
      "step": 5060
    },
    {
      "epoch": 15.504587155963304,
      "grad_norm": 2.5141384601593018,
      "learning_rate": 4.933402705515089e-05,
      "loss": 0.1833,
      "step": 5070
    },
    {
      "epoch": 15.535168195718654,
      "grad_norm": 3.9967784881591797,
      "learning_rate": 4.922996878251821e-05,
      "loss": 0.1838,
      "step": 5080
    },
    {
      "epoch": 15.565749235474007,
      "grad_norm": 3.8655612468719482,
      "learning_rate": 4.9125910509885534e-05,
      "loss": 0.2571,
      "step": 5090
    },
    {
      "epoch": 15.596330275229358,
      "grad_norm": 8.397068977355957,
      "learning_rate": 4.9021852237252866e-05,
      "loss": 0.2467,
      "step": 5100
    },
    {
      "epoch": 15.62691131498471,
      "grad_norm": 4.210080623626709,
      "learning_rate": 4.891779396462019e-05,
      "loss": 0.2353,
      "step": 5110
    },
    {
      "epoch": 15.65749235474006,
      "grad_norm": 3.7114758491516113,
      "learning_rate": 4.8813735691987516e-05,
      "loss": 0.1764,
      "step": 5120
    },
    {
      "epoch": 15.688073394495413,
      "grad_norm": 4.410521507263184,
      "learning_rate": 4.870967741935484e-05,
      "loss": 0.2373,
      "step": 5130
    },
    {
      "epoch": 15.718654434250764,
      "grad_norm": 4.798542022705078,
      "learning_rate": 4.8605619146722166e-05,
      "loss": 0.2254,
      "step": 5140
    },
    {
      "epoch": 15.749235474006117,
      "grad_norm": 5.082320690155029,
      "learning_rate": 4.850156087408949e-05,
      "loss": 0.2278,
      "step": 5150
    },
    {
      "epoch": 15.779816513761467,
      "grad_norm": 3.4736669063568115,
      "learning_rate": 4.8397502601456816e-05,
      "loss": 0.2479,
      "step": 5160
    },
    {
      "epoch": 15.81039755351682,
      "grad_norm": 8.030004501342773,
      "learning_rate": 4.829344432882415e-05,
      "loss": 0.1571,
      "step": 5170
    },
    {
      "epoch": 15.84097859327217,
      "grad_norm": 4.511882305145264,
      "learning_rate": 4.818938605619147e-05,
      "loss": 0.2471,
      "step": 5180
    },
    {
      "epoch": 15.871559633027523,
      "grad_norm": 4.218798637390137,
      "learning_rate": 4.80853277835588e-05,
      "loss": 0.2226,
      "step": 5190
    },
    {
      "epoch": 15.902140672782874,
      "grad_norm": 9.834012985229492,
      "learning_rate": 4.7981269510926116e-05,
      "loss": 0.2012,
      "step": 5200
    },
    {
      "epoch": 15.932721712538227,
      "grad_norm": 3.036139965057373,
      "learning_rate": 4.787721123829345e-05,
      "loss": 0.2287,
      "step": 5210
    },
    {
      "epoch": 15.963302752293577,
      "grad_norm": 4.578927516937256,
      "learning_rate": 4.777315296566077e-05,
      "loss": 0.1971,
      "step": 5220
    },
    {
      "epoch": 15.99388379204893,
      "grad_norm": 26.36978530883789,
      "learning_rate": 4.76690946930281e-05,
      "loss": 0.2227,
      "step": 5230
    },
    {
      "epoch": 16.02446483180428,
      "grad_norm": 3.6557741165161133,
      "learning_rate": 4.756503642039542e-05,
      "loss": 0.1704,
      "step": 5240
    },
    {
      "epoch": 16.05504587155963,
      "grad_norm": 4.2049031257629395,
      "learning_rate": 4.746097814776275e-05,
      "loss": 0.1698,
      "step": 5250
    },
    {
      "epoch": 16.085626911314986,
      "grad_norm": 6.779061794281006,
      "learning_rate": 4.735691987513007e-05,
      "loss": 0.2246,
      "step": 5260
    },
    {
      "epoch": 16.116207951070336,
      "grad_norm": 4.245777130126953,
      "learning_rate": 4.72528616024974e-05,
      "loss": 0.2008,
      "step": 5270
    },
    {
      "epoch": 16.146788990825687,
      "grad_norm": 10.013638496398926,
      "learning_rate": 4.714880332986473e-05,
      "loss": 0.2578,
      "step": 5280
    },
    {
      "epoch": 16.17737003058104,
      "grad_norm": 2.074145555496216,
      "learning_rate": 4.7044745057232055e-05,
      "loss": 0.1921,
      "step": 5290
    },
    {
      "epoch": 16.207951070336392,
      "grad_norm": 4.0752434730529785,
      "learning_rate": 4.694068678459938e-05,
      "loss": 0.1822,
      "step": 5300
    },
    {
      "epoch": 16.238532110091743,
      "grad_norm": 4.347853660583496,
      "learning_rate": 4.6836628511966705e-05,
      "loss": 0.1856,
      "step": 5310
    },
    {
      "epoch": 16.269113149847094,
      "grad_norm": 4.117323398590088,
      "learning_rate": 4.673257023933403e-05,
      "loss": 0.1841,
      "step": 5320
    },
    {
      "epoch": 16.299694189602448,
      "grad_norm": 4.033961772918701,
      "learning_rate": 4.6628511966701355e-05,
      "loss": 0.2984,
      "step": 5330
    },
    {
      "epoch": 16.3302752293578,
      "grad_norm": 2.2842674255371094,
      "learning_rate": 4.652445369406868e-05,
      "loss": 0.1907,
      "step": 5340
    },
    {
      "epoch": 16.36085626911315,
      "grad_norm": 3.3764960765838623,
      "learning_rate": 4.6420395421436005e-05,
      "loss": 0.1756,
      "step": 5350
    },
    {
      "epoch": 16.3914373088685,
      "grad_norm": 6.444164276123047,
      "learning_rate": 4.6316337148803337e-05,
      "loss": 0.2153,
      "step": 5360
    },
    {
      "epoch": 16.422018348623855,
      "grad_norm": 8.567184448242188,
      "learning_rate": 4.6212278876170655e-05,
      "loss": 0.2273,
      "step": 5370
    },
    {
      "epoch": 16.452599388379205,
      "grad_norm": 5.575711250305176,
      "learning_rate": 4.610822060353798e-05,
      "loss": 0.2563,
      "step": 5380
    },
    {
      "epoch": 16.483180428134556,
      "grad_norm": 3.3823914527893066,
      "learning_rate": 4.600416233090531e-05,
      "loss": 0.2264,
      "step": 5390
    },
    {
      "epoch": 16.513761467889907,
      "grad_norm": 3.6987643241882324,
      "learning_rate": 4.590010405827264e-05,
      "loss": 0.1638,
      "step": 5400
    },
    {
      "epoch": 16.54434250764526,
      "grad_norm": 2.492995262145996,
      "learning_rate": 4.579604578563996e-05,
      "loss": 0.2032,
      "step": 5410
    },
    {
      "epoch": 16.574923547400612,
      "grad_norm": 4.453232288360596,
      "learning_rate": 4.569198751300729e-05,
      "loss": 0.2184,
      "step": 5420
    },
    {
      "epoch": 16.605504587155963,
      "grad_norm": 7.290363788604736,
      "learning_rate": 4.558792924037461e-05,
      "loss": 0.2135,
      "step": 5430
    },
    {
      "epoch": 16.636085626911314,
      "grad_norm": 4.6387810707092285,
      "learning_rate": 4.548387096774194e-05,
      "loss": 0.2016,
      "step": 5440
    },
    {
      "epoch": 16.666666666666668,
      "grad_norm": 5.628077983856201,
      "learning_rate": 4.537981269510926e-05,
      "loss": 0.1585,
      "step": 5450
    },
    {
      "epoch": 16.69724770642202,
      "grad_norm": 3.2144572734832764,
      "learning_rate": 4.5275754422476594e-05,
      "loss": 0.2066,
      "step": 5460
    },
    {
      "epoch": 16.72782874617737,
      "grad_norm": 16.819278717041016,
      "learning_rate": 4.517169614984392e-05,
      "loss": 0.2294,
      "step": 5470
    },
    {
      "epoch": 16.75840978593272,
      "grad_norm": 3.323765516281128,
      "learning_rate": 4.506763787721124e-05,
      "loss": 0.2446,
      "step": 5480
    },
    {
      "epoch": 16.788990825688074,
      "grad_norm": 3.800917625427246,
      "learning_rate": 4.496357960457856e-05,
      "loss": 0.2067,
      "step": 5490
    },
    {
      "epoch": 16.819571865443425,
      "grad_norm": 4.269155025482178,
      "learning_rate": 4.4859521331945894e-05,
      "loss": 0.1493,
      "step": 5500
    },
    {
      "epoch": 16.819571865443425,
      "eval_loss": 0.45923885703086853,
      "eval_runtime": 17.6763,
      "eval_samples_per_second": 16.463,
      "eval_steps_per_second": 8.26,
      "step": 5500
    },
    {
      "epoch": 16.850152905198776,
      "grad_norm": 4.534053802490234,
      "learning_rate": 4.475546305931322e-05,
      "loss": 0.1816,
      "step": 5510
    },
    {
      "epoch": 16.88073394495413,
      "grad_norm": 3.5566651821136475,
      "learning_rate": 4.4651404786680544e-05,
      "loss": 0.1771,
      "step": 5520
    },
    {
      "epoch": 16.91131498470948,
      "grad_norm": 7.620478630065918,
      "learning_rate": 4.454734651404787e-05,
      "loss": 0.2447,
      "step": 5530
    },
    {
      "epoch": 16.941896024464832,
      "grad_norm": 1.7376987934112549,
      "learning_rate": 4.4443288241415194e-05,
      "loss": 0.1855,
      "step": 5540
    },
    {
      "epoch": 16.972477064220183,
      "grad_norm": 4.00967264175415,
      "learning_rate": 4.433922996878252e-05,
      "loss": 0.2166,
      "step": 5550
    },
    {
      "epoch": 17.003058103975537,
      "grad_norm": 2.8900110721588135,
      "learning_rate": 4.4235171696149844e-05,
      "loss": 0.2166,
      "step": 5560
    },
    {
      "epoch": 17.033639143730888,
      "grad_norm": 3.8202905654907227,
      "learning_rate": 4.4131113423517176e-05,
      "loss": 0.174,
      "step": 5570
    },
    {
      "epoch": 17.06422018348624,
      "grad_norm": 3.6928577423095703,
      "learning_rate": 4.40270551508845e-05,
      "loss": 0.1657,
      "step": 5580
    },
    {
      "epoch": 17.09480122324159,
      "grad_norm": 5.6383957862854,
      "learning_rate": 4.392299687825182e-05,
      "loss": 0.1918,
      "step": 5590
    },
    {
      "epoch": 17.125382262996943,
      "grad_norm": 4.009942054748535,
      "learning_rate": 4.381893860561915e-05,
      "loss": 0.1959,
      "step": 5600
    },
    {
      "epoch": 17.155963302752294,
      "grad_norm": 2.3905227184295654,
      "learning_rate": 4.3714880332986476e-05,
      "loss": 0.2115,
      "step": 5610
    },
    {
      "epoch": 17.186544342507645,
      "grad_norm": 4.9664740562438965,
      "learning_rate": 4.36108220603538e-05,
      "loss": 0.171,
      "step": 5620
    },
    {
      "epoch": 17.217125382262996,
      "grad_norm": 4.175055980682373,
      "learning_rate": 4.3506763787721126e-05,
      "loss": 0.1552,
      "step": 5630
    },
    {
      "epoch": 17.24770642201835,
      "grad_norm": 5.902973651885986,
      "learning_rate": 4.340270551508846e-05,
      "loss": 0.1792,
      "step": 5640
    },
    {
      "epoch": 17.2782874617737,
      "grad_norm": 4.277851104736328,
      "learning_rate": 4.3298647242455776e-05,
      "loss": 0.225,
      "step": 5650
    },
    {
      "epoch": 17.30886850152905,
      "grad_norm": 3.0731680393218994,
      "learning_rate": 4.31945889698231e-05,
      "loss": 0.2421,
      "step": 5660
    },
    {
      "epoch": 17.339449541284402,
      "grad_norm": 3.9331276416778564,
      "learning_rate": 4.3090530697190426e-05,
      "loss": 0.1605,
      "step": 5670
    },
    {
      "epoch": 17.370030581039757,
      "grad_norm": 5.0467424392700195,
      "learning_rate": 4.298647242455776e-05,
      "loss": 0.1699,
      "step": 5680
    },
    {
      "epoch": 17.400611620795107,
      "grad_norm": 4.776144504547119,
      "learning_rate": 4.288241415192508e-05,
      "loss": 0.167,
      "step": 5690
    },
    {
      "epoch": 17.431192660550458,
      "grad_norm": 23.452909469604492,
      "learning_rate": 4.27783558792924e-05,
      "loss": 0.1966,
      "step": 5700
    },
    {
      "epoch": 17.46177370030581,
      "grad_norm": 3.5143637657165527,
      "learning_rate": 4.267429760665973e-05,
      "loss": 0.2255,
      "step": 5710
    },
    {
      "epoch": 17.492354740061163,
      "grad_norm": 3.8650062084198,
      "learning_rate": 4.257023933402706e-05,
      "loss": 0.1401,
      "step": 5720
    },
    {
      "epoch": 17.522935779816514,
      "grad_norm": 4.191610336303711,
      "learning_rate": 4.246618106139438e-05,
      "loss": 0.192,
      "step": 5730
    },
    {
      "epoch": 17.553516819571865,
      "grad_norm": 5.632226943969727,
      "learning_rate": 4.236212278876171e-05,
      "loss": 0.1822,
      "step": 5740
    },
    {
      "epoch": 17.584097859327215,
      "grad_norm": 6.700896739959717,
      "learning_rate": 4.225806451612904e-05,
      "loss": 0.2147,
      "step": 5750
    },
    {
      "epoch": 17.61467889908257,
      "grad_norm": 2.0719738006591797,
      "learning_rate": 4.215400624349636e-05,
      "loss": 0.2758,
      "step": 5760
    },
    {
      "epoch": 17.64525993883792,
      "grad_norm": 3.314340829849243,
      "learning_rate": 4.204994797086368e-05,
      "loss": 0.1544,
      "step": 5770
    },
    {
      "epoch": 17.67584097859327,
      "grad_norm": 3.5887811183929443,
      "learning_rate": 4.1945889698231015e-05,
      "loss": 0.1905,
      "step": 5780
    },
    {
      "epoch": 17.706422018348626,
      "grad_norm": 4.045060634613037,
      "learning_rate": 4.184183142559834e-05,
      "loss": 0.1851,
      "step": 5790
    },
    {
      "epoch": 17.737003058103976,
      "grad_norm": 5.480162620544434,
      "learning_rate": 4.1737773152965665e-05,
      "loss": 0.1968,
      "step": 5800
    },
    {
      "epoch": 17.767584097859327,
      "grad_norm": 2.7789804935455322,
      "learning_rate": 4.163371488033298e-05,
      "loss": 0.2081,
      "step": 5810
    },
    {
      "epoch": 17.798165137614678,
      "grad_norm": 3.0406036376953125,
      "learning_rate": 4.1529656607700315e-05,
      "loss": 0.16,
      "step": 5820
    },
    {
      "epoch": 17.828746177370032,
      "grad_norm": 4.462724208831787,
      "learning_rate": 4.142559833506764e-05,
      "loss": 0.1612,
      "step": 5830
    },
    {
      "epoch": 17.859327217125383,
      "grad_norm": 6.905117511749268,
      "learning_rate": 4.1321540062434965e-05,
      "loss": 0.2294,
      "step": 5840
    },
    {
      "epoch": 17.889908256880734,
      "grad_norm": 5.394728660583496,
      "learning_rate": 4.121748178980229e-05,
      "loss": 0.172,
      "step": 5850
    },
    {
      "epoch": 17.920489296636084,
      "grad_norm": 4.424241065979004,
      "learning_rate": 4.111342351716962e-05,
      "loss": 0.2024,
      "step": 5860
    },
    {
      "epoch": 17.95107033639144,
      "grad_norm": 4.529147148132324,
      "learning_rate": 4.100936524453694e-05,
      "loss": 0.1698,
      "step": 5870
    },
    {
      "epoch": 17.98165137614679,
      "grad_norm": 4.008469581604004,
      "learning_rate": 4.0905306971904265e-05,
      "loss": 0.2036,
      "step": 5880
    },
    {
      "epoch": 18.01223241590214,
      "grad_norm": 2.6121578216552734,
      "learning_rate": 4.0801248699271597e-05,
      "loss": 0.2187,
      "step": 5890
    },
    {
      "epoch": 18.04281345565749,
      "grad_norm": 2.463407278060913,
      "learning_rate": 4.069719042663892e-05,
      "loss": 0.1473,
      "step": 5900
    },
    {
      "epoch": 18.073394495412845,
      "grad_norm": 4.79568338394165,
      "learning_rate": 4.0593132154006247e-05,
      "loss": 0.1553,
      "step": 5910
    },
    {
      "epoch": 18.103975535168196,
      "grad_norm": 3.8196969032287598,
      "learning_rate": 4.048907388137357e-05,
      "loss": 0.1382,
      "step": 5920
    },
    {
      "epoch": 18.134556574923547,
      "grad_norm": 2.5826215744018555,
      "learning_rate": 4.03850156087409e-05,
      "loss": 0.1984,
      "step": 5930
    },
    {
      "epoch": 18.165137614678898,
      "grad_norm": 5.074633598327637,
      "learning_rate": 4.028095733610822e-05,
      "loss": 0.2118,
      "step": 5940
    },
    {
      "epoch": 18.195718654434252,
      "grad_norm": 2.8331735134124756,
      "learning_rate": 4.017689906347555e-05,
      "loss": 0.1504,
      "step": 5950
    },
    {
      "epoch": 18.226299694189603,
      "grad_norm": 6.5332207679748535,
      "learning_rate": 4.007284079084287e-05,
      "loss": 0.1873,
      "step": 5960
    },
    {
      "epoch": 18.256880733944953,
      "grad_norm": 5.0922651290893555,
      "learning_rate": 3.9968782518210203e-05,
      "loss": 0.1896,
      "step": 5970
    },
    {
      "epoch": 18.287461773700304,
      "grad_norm": 6.440834045410156,
      "learning_rate": 3.986472424557752e-05,
      "loss": 0.2065,
      "step": 5980
    },
    {
      "epoch": 18.31804281345566,
      "grad_norm": 3.152738332748413,
      "learning_rate": 3.976066597294485e-05,
      "loss": 0.2464,
      "step": 5990
    },
    {
      "epoch": 18.34862385321101,
      "grad_norm": 2.717283248901367,
      "learning_rate": 3.965660770031218e-05,
      "loss": 0.1428,
      "step": 6000
    },
    {
      "epoch": 18.34862385321101,
      "eval_loss": 0.44961076974868774,
      "eval_runtime": 17.3868,
      "eval_samples_per_second": 16.737,
      "eval_steps_per_second": 8.397,
      "step": 6000
    },
    {
      "epoch": 18.37920489296636,
      "grad_norm": 5.773160457611084,
      "learning_rate": 3.9552549427679504e-05,
      "loss": 0.204,
      "step": 6010
    },
    {
      "epoch": 18.40978593272171,
      "grad_norm": 7.26622200012207,
      "learning_rate": 3.944849115504683e-05,
      "loss": 0.1919,
      "step": 6020
    },
    {
      "epoch": 18.440366972477065,
      "grad_norm": 4.916703224182129,
      "learning_rate": 3.9344432882414154e-05,
      "loss": 0.1785,
      "step": 6030
    },
    {
      "epoch": 18.470948012232416,
      "grad_norm": 2.300110101699829,
      "learning_rate": 3.924037460978148e-05,
      "loss": 0.1954,
      "step": 6040
    },
    {
      "epoch": 18.501529051987767,
      "grad_norm": 3.844594955444336,
      "learning_rate": 3.9136316337148804e-05,
      "loss": 0.1527,
      "step": 6050
    },
    {
      "epoch": 18.53211009174312,
      "grad_norm": 2.7383902072906494,
      "learning_rate": 3.903225806451613e-05,
      "loss": 0.1831,
      "step": 6060
    },
    {
      "epoch": 18.56269113149847,
      "grad_norm": 22.698638916015625,
      "learning_rate": 3.892819979188346e-05,
      "loss": 0.2091,
      "step": 6070
    },
    {
      "epoch": 18.593272171253822,
      "grad_norm": 8.911148071289062,
      "learning_rate": 3.8824141519250785e-05,
      "loss": 0.2182,
      "step": 6080
    },
    {
      "epoch": 18.623853211009173,
      "grad_norm": 2.808556079864502,
      "learning_rate": 3.872008324661811e-05,
      "loss": 0.2172,
      "step": 6090
    },
    {
      "epoch": 18.654434250764528,
      "grad_norm": 2.2882444858551025,
      "learning_rate": 3.861602497398543e-05,
      "loss": 0.1442,
      "step": 6100
    },
    {
      "epoch": 18.68501529051988,
      "grad_norm": 2.9601521492004395,
      "learning_rate": 3.851196670135276e-05,
      "loss": 0.1747,
      "step": 6110
    },
    {
      "epoch": 18.71559633027523,
      "grad_norm": 3.576287269592285,
      "learning_rate": 3.8407908428720086e-05,
      "loss": 0.1767,
      "step": 6120
    },
    {
      "epoch": 18.74617737003058,
      "grad_norm": 4.117494583129883,
      "learning_rate": 3.830385015608741e-05,
      "loss": 0.179,
      "step": 6130
    },
    {
      "epoch": 18.776758409785934,
      "grad_norm": 3.8094818592071533,
      "learning_rate": 3.8199791883454736e-05,
      "loss": 0.2504,
      "step": 6140
    },
    {
      "epoch": 18.807339449541285,
      "grad_norm": 2.3974802494049072,
      "learning_rate": 3.809573361082206e-05,
      "loss": 0.1616,
      "step": 6150
    },
    {
      "epoch": 18.837920489296636,
      "grad_norm": 3.671895742416382,
      "learning_rate": 3.7991675338189386e-05,
      "loss": 0.1324,
      "step": 6160
    },
    {
      "epoch": 18.868501529051986,
      "grad_norm": 3.833237648010254,
      "learning_rate": 3.788761706555671e-05,
      "loss": 0.1597,
      "step": 6170
    },
    {
      "epoch": 18.89908256880734,
      "grad_norm": 5.964929580688477,
      "learning_rate": 3.778355879292404e-05,
      "loss": 0.2068,
      "step": 6180
    },
    {
      "epoch": 18.92966360856269,
      "grad_norm": 3.1389944553375244,
      "learning_rate": 3.767950052029137e-05,
      "loss": 0.2317,
      "step": 6190
    },
    {
      "epoch": 18.960244648318042,
      "grad_norm": 4.046453952789307,
      "learning_rate": 3.757544224765869e-05,
      "loss": 0.1787,
      "step": 6200
    },
    {
      "epoch": 18.990825688073393,
      "grad_norm": 3.8653130531311035,
      "learning_rate": 3.747138397502602e-05,
      "loss": 0.1851,
      "step": 6210
    },
    {
      "epoch": 19.021406727828747,
      "grad_norm": 2.5173943042755127,
      "learning_rate": 3.736732570239334e-05,
      "loss": 0.1761,
      "step": 6220
    },
    {
      "epoch": 19.051987767584098,
      "grad_norm": 2.8177404403686523,
      "learning_rate": 3.726326742976067e-05,
      "loss": 0.1356,
      "step": 6230
    },
    {
      "epoch": 19.08256880733945,
      "grad_norm": 4.992142200469971,
      "learning_rate": 3.715920915712799e-05,
      "loss": 0.1783,
      "step": 6240
    },
    {
      "epoch": 19.1131498470948,
      "grad_norm": 3.925229072570801,
      "learning_rate": 3.7055150884495324e-05,
      "loss": 0.1689,
      "step": 6250
    },
    {
      "epoch": 19.143730886850154,
      "grad_norm": 4.234835624694824,
      "learning_rate": 3.695109261186264e-05,
      "loss": 0.1858,
      "step": 6260
    },
    {
      "epoch": 19.174311926605505,
      "grad_norm": 6.015228748321533,
      "learning_rate": 3.684703433922997e-05,
      "loss": 0.1984,
      "step": 6270
    },
    {
      "epoch": 19.204892966360855,
      "grad_norm": 3.5733656883239746,
      "learning_rate": 3.674297606659729e-05,
      "loss": 0.1424,
      "step": 6280
    },
    {
      "epoch": 19.235474006116206,
      "grad_norm": 2.8804519176483154,
      "learning_rate": 3.6638917793964624e-05,
      "loss": 0.1838,
      "step": 6290
    },
    {
      "epoch": 19.26605504587156,
      "grad_norm": 6.07488489151001,
      "learning_rate": 3.653485952133195e-05,
      "loss": 0.1532,
      "step": 6300
    },
    {
      "epoch": 19.29663608562691,
      "grad_norm": 4.1511054039001465,
      "learning_rate": 3.6430801248699275e-05,
      "loss": 0.2108,
      "step": 6310
    },
    {
      "epoch": 19.327217125382262,
      "grad_norm": 3.896618366241455,
      "learning_rate": 3.63267429760666e-05,
      "loss": 0.1359,
      "step": 6320
    },
    {
      "epoch": 19.357798165137616,
      "grad_norm": 3.186441659927368,
      "learning_rate": 3.6222684703433925e-05,
      "loss": 0.1386,
      "step": 6330
    },
    {
      "epoch": 19.388379204892967,
      "grad_norm": 5.969933986663818,
      "learning_rate": 3.611862643080125e-05,
      "loss": 0.1675,
      "step": 6340
    },
    {
      "epoch": 19.418960244648318,
      "grad_norm": 5.708403587341309,
      "learning_rate": 3.6014568158168575e-05,
      "loss": 0.1671,
      "step": 6350
    },
    {
      "epoch": 19.44954128440367,
      "grad_norm": 6.588597297668457,
      "learning_rate": 3.5910509885535906e-05,
      "loss": 0.1699,
      "step": 6360
    },
    {
      "epoch": 19.480122324159023,
      "grad_norm": 3.4485206604003906,
      "learning_rate": 3.580645161290323e-05,
      "loss": 0.1553,
      "step": 6370
    },
    {
      "epoch": 19.510703363914374,
      "grad_norm": 2.0754480361938477,
      "learning_rate": 3.570239334027055e-05,
      "loss": 0.1325,
      "step": 6380
    },
    {
      "epoch": 19.541284403669724,
      "grad_norm": 5.2881693840026855,
      "learning_rate": 3.5598335067637875e-05,
      "loss": 0.1833,
      "step": 6390
    },
    {
      "epoch": 19.571865443425075,
      "grad_norm": 7.182536602020264,
      "learning_rate": 3.5494276795005206e-05,
      "loss": 0.1694,
      "step": 6400
    },
    {
      "epoch": 19.60244648318043,
      "grad_norm": 6.095035076141357,
      "learning_rate": 3.539021852237253e-05,
      "loss": 0.2415,
      "step": 6410
    },
    {
      "epoch": 19.63302752293578,
      "grad_norm": 3.0211124420166016,
      "learning_rate": 3.5286160249739857e-05,
      "loss": 0.198,
      "step": 6420
    },
    {
      "epoch": 19.66360856269113,
      "grad_norm": 3.4166922569274902,
      "learning_rate": 3.518210197710718e-05,
      "loss": 0.1311,
      "step": 6430
    },
    {
      "epoch": 19.69418960244648,
      "grad_norm": 2.852652072906494,
      "learning_rate": 3.5078043704474507e-05,
      "loss": 0.1563,
      "step": 6440
    },
    {
      "epoch": 19.724770642201836,
      "grad_norm": 7.064545154571533,
      "learning_rate": 3.497398543184183e-05,
      "loss": 0.1824,
      "step": 6450
    },
    {
      "epoch": 19.755351681957187,
      "grad_norm": 6.215057849884033,
      "learning_rate": 3.4869927159209157e-05,
      "loss": 0.1797,
      "step": 6460
    },
    {
      "epoch": 19.785932721712538,
      "grad_norm": 3.3205864429473877,
      "learning_rate": 3.476586888657649e-05,
      "loss": 0.1543,
      "step": 6470
    },
    {
      "epoch": 19.81651376146789,
      "grad_norm": 4.110857963562012,
      "learning_rate": 3.4661810613943813e-05,
      "loss": 0.1745,
      "step": 6480
    },
    {
      "epoch": 19.847094801223243,
      "grad_norm": 3.430475950241089,
      "learning_rate": 3.455775234131113e-05,
      "loss": 0.1741,
      "step": 6490
    },
    {
      "epoch": 19.877675840978593,
      "grad_norm": 4.670803546905518,
      "learning_rate": 3.4453694068678463e-05,
      "loss": 0.1856,
      "step": 6500
    },
    {
      "epoch": 19.877675840978593,
      "eval_loss": 0.44353702664375305,
      "eval_runtime": 16.8697,
      "eval_samples_per_second": 17.25,
      "eval_steps_per_second": 8.655,
      "step": 6500
    },
    {
      "epoch": 19.908256880733944,
      "grad_norm": 5.290159702301025,
      "learning_rate": 3.434963579604579e-05,
      "loss": 0.2133,
      "step": 6510
    },
    {
      "epoch": 19.938837920489295,
      "grad_norm": 3.216883420944214,
      "learning_rate": 3.4245577523413113e-05,
      "loss": 0.1675,
      "step": 6520
    },
    {
      "epoch": 19.96941896024465,
      "grad_norm": 3.890711784362793,
      "learning_rate": 3.414151925078044e-05,
      "loss": 0.1533,
      "step": 6530
    },
    {
      "epoch": 20.0,
      "grad_norm": 12.24163818359375,
      "learning_rate": 3.403746097814777e-05,
      "loss": 0.2362,
      "step": 6540
    },
    {
      "epoch": 20.03058103975535,
      "grad_norm": 4.171586990356445,
      "learning_rate": 3.393340270551509e-05,
      "loss": 0.1078,
      "step": 6550
    },
    {
      "epoch": 20.061162079510705,
      "grad_norm": 4.116422176361084,
      "learning_rate": 3.3829344432882414e-05,
      "loss": 0.1466,
      "step": 6560
    },
    {
      "epoch": 20.091743119266056,
      "grad_norm": 4.5232625007629395,
      "learning_rate": 3.372528616024974e-05,
      "loss": 0.1598,
      "step": 6570
    },
    {
      "epoch": 20.122324159021407,
      "grad_norm": 5.76244592666626,
      "learning_rate": 3.362122788761707e-05,
      "loss": 0.1706,
      "step": 6580
    },
    {
      "epoch": 20.152905198776757,
      "grad_norm": 6.140753746032715,
      "learning_rate": 3.3517169614984395e-05,
      "loss": 0.2467,
      "step": 6590
    },
    {
      "epoch": 20.18348623853211,
      "grad_norm": 3.3037118911743164,
      "learning_rate": 3.3413111342351714e-05,
      "loss": 0.1593,
      "step": 6600
    },
    {
      "epoch": 20.214067278287462,
      "grad_norm": 7.814075469970703,
      "learning_rate": 3.3309053069719045e-05,
      "loss": 0.1354,
      "step": 6610
    },
    {
      "epoch": 20.244648318042813,
      "grad_norm": 4.812346458435059,
      "learning_rate": 3.320499479708637e-05,
      "loss": 0.1493,
      "step": 6620
    },
    {
      "epoch": 20.275229357798164,
      "grad_norm": 4.187191009521484,
      "learning_rate": 3.3100936524453695e-05,
      "loss": 0.1723,
      "step": 6630
    },
    {
      "epoch": 20.30581039755352,
      "grad_norm": 8.240633010864258,
      "learning_rate": 3.299687825182102e-05,
      "loss": 0.249,
      "step": 6640
    },
    {
      "epoch": 20.33639143730887,
      "grad_norm": 2.668346881866455,
      "learning_rate": 3.289281997918835e-05,
      "loss": 0.1191,
      "step": 6650
    },
    {
      "epoch": 20.36697247706422,
      "grad_norm": 3.798555850982666,
      "learning_rate": 3.278876170655567e-05,
      "loss": 0.1465,
      "step": 6660
    },
    {
      "epoch": 20.39755351681957,
      "grad_norm": 5.0725417137146,
      "learning_rate": 3.2684703433922996e-05,
      "loss": 0.161,
      "step": 6670
    },
    {
      "epoch": 20.428134556574925,
      "grad_norm": 2.8009026050567627,
      "learning_rate": 3.258064516129033e-05,
      "loss": 0.2017,
      "step": 6680
    },
    {
      "epoch": 20.458715596330276,
      "grad_norm": 4.939574718475342,
      "learning_rate": 3.247658688865765e-05,
      "loss": 0.1874,
      "step": 6690
    },
    {
      "epoch": 20.489296636085626,
      "grad_norm": 3.652327060699463,
      "learning_rate": 3.237252861602498e-05,
      "loss": 0.1294,
      "step": 6700
    },
    {
      "epoch": 20.519877675840977,
      "grad_norm": 4.777719020843506,
      "learning_rate": 3.2268470343392296e-05,
      "loss": 0.1049,
      "step": 6710
    },
    {
      "epoch": 20.55045871559633,
      "grad_norm": 4.19663667678833,
      "learning_rate": 3.216441207075963e-05,
      "loss": 0.1251,
      "step": 6720
    },
    {
      "epoch": 20.581039755351682,
      "grad_norm": 7.416243553161621,
      "learning_rate": 3.206035379812695e-05,
      "loss": 0.186,
      "step": 6730
    },
    {
      "epoch": 20.611620795107033,
      "grad_norm": 16.07038688659668,
      "learning_rate": 3.195629552549428e-05,
      "loss": 0.201,
      "step": 6740
    },
    {
      "epoch": 20.642201834862384,
      "grad_norm": 5.279475212097168,
      "learning_rate": 3.18522372528616e-05,
      "loss": 0.1347,
      "step": 6750
    },
    {
      "epoch": 20.672782874617738,
      "grad_norm": 4.603892803192139,
      "learning_rate": 3.1748178980228934e-05,
      "loss": 0.1489,
      "step": 6760
    },
    {
      "epoch": 20.70336391437309,
      "grad_norm": 6.145246982574463,
      "learning_rate": 3.164412070759625e-05,
      "loss": 0.1138,
      "step": 6770
    },
    {
      "epoch": 20.73394495412844,
      "grad_norm": 5.94775915145874,
      "learning_rate": 3.154006243496358e-05,
      "loss": 0.1769,
      "step": 6780
    },
    {
      "epoch": 20.764525993883794,
      "grad_norm": 6.3351826667785645,
      "learning_rate": 3.143600416233091e-05,
      "loss": 0.1765,
      "step": 6790
    },
    {
      "epoch": 20.795107033639145,
      "grad_norm": 1.8415037393569946,
      "learning_rate": 3.1331945889698234e-05,
      "loss": 0.131,
      "step": 6800
    },
    {
      "epoch": 20.825688073394495,
      "grad_norm": 2.168361186981201,
      "learning_rate": 3.122788761706556e-05,
      "loss": 0.1462,
      "step": 6810
    },
    {
      "epoch": 20.856269113149846,
      "grad_norm": 7.718116760253906,
      "learning_rate": 3.1123829344432884e-05,
      "loss": 0.1667,
      "step": 6820
    },
    {
      "epoch": 20.8868501529052,
      "grad_norm": 4.486772060394287,
      "learning_rate": 3.101977107180021e-05,
      "loss": 0.1617,
      "step": 6830
    },
    {
      "epoch": 20.91743119266055,
      "grad_norm": 9.251279830932617,
      "learning_rate": 3.0915712799167534e-05,
      "loss": 0.2185,
      "step": 6840
    },
    {
      "epoch": 20.948012232415902,
      "grad_norm": 2.624577522277832,
      "learning_rate": 3.081165452653486e-05,
      "loss": 0.1468,
      "step": 6850
    },
    {
      "epoch": 20.978593272171253,
      "grad_norm": 4.017003059387207,
      "learning_rate": 3.0707596253902185e-05,
      "loss": 0.1703,
      "step": 6860
    },
    {
      "epoch": 21.009174311926607,
      "grad_norm": 5.159716606140137,
      "learning_rate": 3.0603537981269516e-05,
      "loss": 0.1809,
      "step": 6870
    },
    {
      "epoch": 21.039755351681958,
      "grad_norm": 3.0058741569519043,
      "learning_rate": 3.0499479708636835e-05,
      "loss": 0.1118,
      "step": 6880
    },
    {
      "epoch": 21.07033639143731,
      "grad_norm": 2.598550796508789,
      "learning_rate": 3.0395421436004163e-05,
      "loss": 0.1684,
      "step": 6890
    },
    {
      "epoch": 21.10091743119266,
      "grad_norm": 4.881763935089111,
      "learning_rate": 3.0291363163371488e-05,
      "loss": 0.1274,
      "step": 6900
    },
    {
      "epoch": 21.131498470948014,
      "grad_norm": 6.203042984008789,
      "learning_rate": 3.0187304890738816e-05,
      "loss": 0.1916,
      "step": 6910
    },
    {
      "epoch": 21.162079510703364,
      "grad_norm": 2.4897990226745605,
      "learning_rate": 3.008324661810614e-05,
      "loss": 0.2015,
      "step": 6920
    },
    {
      "epoch": 21.192660550458715,
      "grad_norm": 2.7538557052612305,
      "learning_rate": 2.997918834547347e-05,
      "loss": 0.1044,
      "step": 6930
    },
    {
      "epoch": 21.223241590214066,
      "grad_norm": 7.675657272338867,
      "learning_rate": 2.987513007284079e-05,
      "loss": 0.1387,
      "step": 6940
    },
    {
      "epoch": 21.25382262996942,
      "grad_norm": 3.915666341781616,
      "learning_rate": 2.9771071800208116e-05,
      "loss": 0.1505,
      "step": 6950
    },
    {
      "epoch": 21.28440366972477,
      "grad_norm": 6.621923446655273,
      "learning_rate": 2.9667013527575445e-05,
      "loss": 0.2056,
      "step": 6960
    },
    {
      "epoch": 21.31498470948012,
      "grad_norm": 2.253448486328125,
      "learning_rate": 2.956295525494277e-05,
      "loss": 0.1935,
      "step": 6970
    },
    {
      "epoch": 21.345565749235472,
      "grad_norm": 3.531188726425171,
      "learning_rate": 2.94588969823101e-05,
      "loss": 0.1359,
      "step": 6980
    },
    {
      "epoch": 21.376146788990827,
      "grad_norm": 4.113672256469727,
      "learning_rate": 2.9354838709677417e-05,
      "loss": 0.1521,
      "step": 6990
    },
    {
      "epoch": 21.406727828746178,
      "grad_norm": 4.267278671264648,
      "learning_rate": 2.9250780437044745e-05,
      "loss": 0.1129,
      "step": 7000
    },
    {
      "epoch": 21.406727828746178,
      "eval_loss": 0.4731081426143646,
      "eval_runtime": 16.9507,
      "eval_samples_per_second": 17.167,
      "eval_steps_per_second": 8.613,
      "step": 7000
    },
    {
      "epoch": 21.43730886850153,
      "grad_norm": 5.572481155395508,
      "learning_rate": 2.914672216441207e-05,
      "loss": 0.1351,
      "step": 7010
    },
    {
      "epoch": 21.46788990825688,
      "grad_norm": 2.6719911098480225,
      "learning_rate": 2.90426638917794e-05,
      "loss": 0.1672,
      "step": 7020
    },
    {
      "epoch": 21.498470948012233,
      "grad_norm": 6.055787086486816,
      "learning_rate": 2.8938605619146723e-05,
      "loss": 0.1013,
      "step": 7030
    },
    {
      "epoch": 21.529051987767584,
      "grad_norm": 4.567263603210449,
      "learning_rate": 2.8834547346514052e-05,
      "loss": 0.1349,
      "step": 7040
    },
    {
      "epoch": 21.559633027522935,
      "grad_norm": 5.775519847869873,
      "learning_rate": 2.8730489073881373e-05,
      "loss": 0.1357,
      "step": 7050
    },
    {
      "epoch": 21.59021406727829,
      "grad_norm": 7.519472122192383,
      "learning_rate": 2.86264308012487e-05,
      "loss": 0.1715,
      "step": 7060
    },
    {
      "epoch": 21.62079510703364,
      "grad_norm": 2.7148330211639404,
      "learning_rate": 2.8522372528616027e-05,
      "loss": 0.1807,
      "step": 7070
    },
    {
      "epoch": 21.65137614678899,
      "grad_norm": 3.665200710296631,
      "learning_rate": 2.8418314255983352e-05,
      "loss": 0.1299,
      "step": 7080
    },
    {
      "epoch": 21.68195718654434,
      "grad_norm": 3.740471601486206,
      "learning_rate": 2.831425598335068e-05,
      "loss": 0.1203,
      "step": 7090
    },
    {
      "epoch": 21.712538226299696,
      "grad_norm": 6.551999092102051,
      "learning_rate": 2.8210197710718005e-05,
      "loss": 0.1833,
      "step": 7100
    },
    {
      "epoch": 21.743119266055047,
      "grad_norm": 9.063637733459473,
      "learning_rate": 2.8106139438085327e-05,
      "loss": 0.1705,
      "step": 7110
    },
    {
      "epoch": 21.773700305810397,
      "grad_norm": 4.627596855163574,
      "learning_rate": 2.8002081165452655e-05,
      "loss": 0.1727,
      "step": 7120
    },
    {
      "epoch": 21.804281345565748,
      "grad_norm": 2.3520357608795166,
      "learning_rate": 2.789802289281998e-05,
      "loss": 0.125,
      "step": 7130
    },
    {
      "epoch": 21.834862385321102,
      "grad_norm": 5.15336275100708,
      "learning_rate": 2.779396462018731e-05,
      "loss": 0.1442,
      "step": 7140
    },
    {
      "epoch": 21.865443425076453,
      "grad_norm": 9.608550071716309,
      "learning_rate": 2.7689906347554634e-05,
      "loss": 0.1595,
      "step": 7150
    },
    {
      "epoch": 21.896024464831804,
      "grad_norm": 6.136108875274658,
      "learning_rate": 2.7585848074921955e-05,
      "loss": 0.1819,
      "step": 7160
    },
    {
      "epoch": 21.926605504587155,
      "grad_norm": 2.7484302520751953,
      "learning_rate": 2.748178980228928e-05,
      "loss": 0.1706,
      "step": 7170
    },
    {
      "epoch": 21.95718654434251,
      "grad_norm": 4.393002510070801,
      "learning_rate": 2.737773152965661e-05,
      "loss": 0.1237,
      "step": 7180
    },
    {
      "epoch": 21.98776758409786,
      "grad_norm": 2.5787153244018555,
      "learning_rate": 2.7273673257023934e-05,
      "loss": 0.1456,
      "step": 7190
    },
    {
      "epoch": 22.01834862385321,
      "grad_norm": 5.6595377922058105,
      "learning_rate": 2.7169614984391262e-05,
      "loss": 0.1583,
      "step": 7200
    },
    {
      "epoch": 22.04892966360856,
      "grad_norm": 2.430314779281616,
      "learning_rate": 2.7065556711758587e-05,
      "loss": 0.1023,
      "step": 7210
    },
    {
      "epoch": 22.079510703363916,
      "grad_norm": 4.03518533706665,
      "learning_rate": 2.696149843912591e-05,
      "loss": 0.1829,
      "step": 7220
    },
    {
      "epoch": 22.110091743119266,
      "grad_norm": 5.839175701141357,
      "learning_rate": 2.6857440166493237e-05,
      "loss": 0.1378,
      "step": 7230
    },
    {
      "epoch": 22.140672782874617,
      "grad_norm": 3.737353563308716,
      "learning_rate": 2.6753381893860562e-05,
      "loss": 0.1801,
      "step": 7240
    },
    {
      "epoch": 22.171253822629968,
      "grad_norm": 1.6595263481140137,
      "learning_rate": 2.664932362122789e-05,
      "loss": 0.1422,
      "step": 7250
    },
    {
      "epoch": 22.201834862385322,
      "grad_norm": 2.8589601516723633,
      "learning_rate": 2.6545265348595216e-05,
      "loss": 0.1056,
      "step": 7260
    },
    {
      "epoch": 22.232415902140673,
      "grad_norm": 2.1064066886901855,
      "learning_rate": 2.6441207075962544e-05,
      "loss": 0.1541,
      "step": 7270
    },
    {
      "epoch": 22.262996941896024,
      "grad_norm": 3.4547598361968994,
      "learning_rate": 2.6337148803329862e-05,
      "loss": 0.1684,
      "step": 7280
    },
    {
      "epoch": 22.293577981651374,
      "grad_norm": 5.208943843841553,
      "learning_rate": 2.623309053069719e-05,
      "loss": 0.1583,
      "step": 7290
    },
    {
      "epoch": 22.32415902140673,
      "grad_norm": 2.59063458442688,
      "learning_rate": 2.6129032258064516e-05,
      "loss": 0.147,
      "step": 7300
    },
    {
      "epoch": 22.35474006116208,
      "grad_norm": 2.2470364570617676,
      "learning_rate": 2.6024973985431844e-05,
      "loss": 0.1587,
      "step": 7310
    },
    {
      "epoch": 22.38532110091743,
      "grad_norm": 3.6315572261810303,
      "learning_rate": 2.592091571279917e-05,
      "loss": 0.1373,
      "step": 7320
    },
    {
      "epoch": 22.415902140672785,
      "grad_norm": 6.476180076599121,
      "learning_rate": 2.581685744016649e-05,
      "loss": 0.194,
      "step": 7330
    },
    {
      "epoch": 22.446483180428135,
      "grad_norm": 5.465244293212891,
      "learning_rate": 2.571279916753382e-05,
      "loss": 0.1674,
      "step": 7340
    },
    {
      "epoch": 22.477064220183486,
      "grad_norm": 3.601835250854492,
      "learning_rate": 2.5608740894901144e-05,
      "loss": 0.1335,
      "step": 7350
    },
    {
      "epoch": 22.507645259938837,
      "grad_norm": 2.8011441230773926,
      "learning_rate": 2.5504682622268473e-05,
      "loss": 0.0966,
      "step": 7360
    },
    {
      "epoch": 22.53822629969419,
      "grad_norm": 3.8074159622192383,
      "learning_rate": 2.5400624349635798e-05,
      "loss": 0.1206,
      "step": 7370
    },
    {
      "epoch": 22.568807339449542,
      "grad_norm": 2.789634943008423,
      "learning_rate": 2.5296566077003126e-05,
      "loss": 0.1234,
      "step": 7380
    },
    {
      "epoch": 22.599388379204893,
      "grad_norm": 3.4443519115448,
      "learning_rate": 2.5192507804370448e-05,
      "loss": 0.1578,
      "step": 7390
    },
    {
      "epoch": 22.629969418960243,
      "grad_norm": 3.9452195167541504,
      "learning_rate": 2.5088449531737773e-05,
      "loss": 0.1737,
      "step": 7400
    },
    {
      "epoch": 22.660550458715598,
      "grad_norm": 2.4158213138580322,
      "learning_rate": 2.49843912591051e-05,
      "loss": 0.1387,
      "step": 7410
    },
    {
      "epoch": 22.69113149847095,
      "grad_norm": 6.3911356925964355,
      "learning_rate": 2.4880332986472426e-05,
      "loss": 0.1519,
      "step": 7420
    },
    {
      "epoch": 22.7217125382263,
      "grad_norm": 2.70169734954834,
      "learning_rate": 2.477627471383975e-05,
      "loss": 0.16,
      "step": 7430
    },
    {
      "epoch": 22.75229357798165,
      "grad_norm": 4.864673614501953,
      "learning_rate": 2.4672216441207076e-05,
      "loss": 0.133,
      "step": 7440
    },
    {
      "epoch": 22.782874617737004,
      "grad_norm": 1.9068208932876587,
      "learning_rate": 2.4568158168574405e-05,
      "loss": 0.1434,
      "step": 7450
    },
    {
      "epoch": 22.813455657492355,
      "grad_norm": 1.8495513200759888,
      "learning_rate": 2.4464099895941726e-05,
      "loss": 0.1397,
      "step": 7460
    },
    {
      "epoch": 22.844036697247706,
      "grad_norm": 4.008700370788574,
      "learning_rate": 2.4360041623309055e-05,
      "loss": 0.1333,
      "step": 7470
    },
    {
      "epoch": 22.874617737003057,
      "grad_norm": 3.1677236557006836,
      "learning_rate": 2.425598335067638e-05,
      "loss": 0.1479,
      "step": 7480
    },
    {
      "epoch": 22.90519877675841,
      "grad_norm": 3.1551225185394287,
      "learning_rate": 2.4151925078043705e-05,
      "loss": 0.1471,
      "step": 7490
    },
    {
      "epoch": 22.93577981651376,
      "grad_norm": 3.30476450920105,
      "learning_rate": 2.4047866805411033e-05,
      "loss": 0.1454,
      "step": 7500
    },
    {
      "epoch": 22.93577981651376,
      "eval_loss": 0.459464430809021,
      "eval_runtime": 16.9937,
      "eval_samples_per_second": 17.124,
      "eval_steps_per_second": 8.591,
      "step": 7500
    },
    {
      "epoch": 22.966360856269112,
      "grad_norm": 2.7632486820220947,
      "learning_rate": 2.3943808532778355e-05,
      "loss": 0.1611,
      "step": 7510
    },
    {
      "epoch": 22.996941896024463,
      "grad_norm": 6.361327171325684,
      "learning_rate": 2.3839750260145683e-05,
      "loss": 0.1678,
      "step": 7520
    },
    {
      "epoch": 23.027522935779817,
      "grad_norm": 3.0553674697875977,
      "learning_rate": 2.373569198751301e-05,
      "loss": 0.1319,
      "step": 7530
    },
    {
      "epoch": 23.058103975535168,
      "grad_norm": 5.1710357666015625,
      "learning_rate": 2.3631633714880333e-05,
      "loss": 0.1498,
      "step": 7540
    },
    {
      "epoch": 23.08868501529052,
      "grad_norm": 5.715560436248779,
      "learning_rate": 2.352757544224766e-05,
      "loss": 0.1304,
      "step": 7550
    },
    {
      "epoch": 23.119266055045873,
      "grad_norm": 6.18096923828125,
      "learning_rate": 2.3423517169614987e-05,
      "loss": 0.1642,
      "step": 7560
    },
    {
      "epoch": 23.149847094801224,
      "grad_norm": 18.369741439819336,
      "learning_rate": 2.3319458896982312e-05,
      "loss": 0.1894,
      "step": 7570
    },
    {
      "epoch": 23.180428134556575,
      "grad_norm": 2.6054799556732178,
      "learning_rate": 2.3215400624349637e-05,
      "loss": 0.0919,
      "step": 7580
    },
    {
      "epoch": 23.211009174311926,
      "grad_norm": 5.282476425170898,
      "learning_rate": 2.3111342351716965e-05,
      "loss": 0.0982,
      "step": 7590
    },
    {
      "epoch": 23.24159021406728,
      "grad_norm": 5.387862205505371,
      "learning_rate": 2.3007284079084287e-05,
      "loss": 0.1567,
      "step": 7600
    },
    {
      "epoch": 23.27217125382263,
      "grad_norm": 3.5458264350891113,
      "learning_rate": 2.2903225806451615e-05,
      "loss": 0.1508,
      "step": 7610
    },
    {
      "epoch": 23.30275229357798,
      "grad_norm": 5.379571914672852,
      "learning_rate": 2.279916753381894e-05,
      "loss": 0.2044,
      "step": 7620
    },
    {
      "epoch": 23.333333333333332,
      "grad_norm": 3.6042699813842773,
      "learning_rate": 2.2695109261186265e-05,
      "loss": 0.119,
      "step": 7630
    },
    {
      "epoch": 23.363914373088686,
      "grad_norm": 2.713719606399536,
      "learning_rate": 2.259105098855359e-05,
      "loss": 0.1356,
      "step": 7640
    },
    {
      "epoch": 23.394495412844037,
      "grad_norm": 4.7480597496032715,
      "learning_rate": 2.248699271592092e-05,
      "loss": 0.1428,
      "step": 7650
    },
    {
      "epoch": 23.425076452599388,
      "grad_norm": 5.461562156677246,
      "learning_rate": 2.2382934443288244e-05,
      "loss": 0.1555,
      "step": 7660
    },
    {
      "epoch": 23.45565749235474,
      "grad_norm": 8.009157180786133,
      "learning_rate": 2.227887617065557e-05,
      "loss": 0.1491,
      "step": 7670
    },
    {
      "epoch": 23.486238532110093,
      "grad_norm": 2.9787027835845947,
      "learning_rate": 2.2174817898022894e-05,
      "loss": 0.1062,
      "step": 7680
    },
    {
      "epoch": 23.516819571865444,
      "grad_norm": 3.0121078491210938,
      "learning_rate": 2.207075962539022e-05,
      "loss": 0.1247,
      "step": 7690
    },
    {
      "epoch": 23.547400611620795,
      "grad_norm": 6.196150302886963,
      "learning_rate": 2.1966701352757547e-05,
      "loss": 0.1115,
      "step": 7700
    },
    {
      "epoch": 23.577981651376145,
      "grad_norm": 3.6438441276550293,
      "learning_rate": 2.186264308012487e-05,
      "loss": 0.1245,
      "step": 7710
    },
    {
      "epoch": 23.6085626911315,
      "grad_norm": 9.425365447998047,
      "learning_rate": 2.1758584807492197e-05,
      "loss": 0.1827,
      "step": 7720
    },
    {
      "epoch": 23.63914373088685,
      "grad_norm": 3.6670544147491455,
      "learning_rate": 2.1654526534859522e-05,
      "loss": 0.1519,
      "step": 7730
    },
    {
      "epoch": 23.6697247706422,
      "grad_norm": 3.302945137023926,
      "learning_rate": 2.1550468262226847e-05,
      "loss": 0.1221,
      "step": 7740
    },
    {
      "epoch": 23.700305810397552,
      "grad_norm": 3.1169800758361816,
      "learning_rate": 2.1446409989594172e-05,
      "loss": 0.1274,
      "step": 7750
    },
    {
      "epoch": 23.730886850152906,
      "grad_norm": 3.494316577911377,
      "learning_rate": 2.13423517169615e-05,
      "loss": 0.2026,
      "step": 7760
    },
    {
      "epoch": 23.761467889908257,
      "grad_norm": 4.792800426483154,
      "learning_rate": 2.1238293444328826e-05,
      "loss": 0.1336,
      "step": 7770
    },
    {
      "epoch": 23.792048929663608,
      "grad_norm": 1.7731387615203857,
      "learning_rate": 2.113423517169615e-05,
      "loss": 0.0959,
      "step": 7780
    },
    {
      "epoch": 23.822629969418962,
      "grad_norm": 4.196662425994873,
      "learning_rate": 2.103017689906348e-05,
      "loss": 0.1162,
      "step": 7790
    },
    {
      "epoch": 23.853211009174313,
      "grad_norm": 5.490009784698486,
      "learning_rate": 2.09261186264308e-05,
      "loss": 0.1544,
      "step": 7800
    },
    {
      "epoch": 23.883792048929664,
      "grad_norm": 2.1621978282928467,
      "learning_rate": 2.082206035379813e-05,
      "loss": 0.1212,
      "step": 7810
    },
    {
      "epoch": 23.914373088685014,
      "grad_norm": 4.357117176055908,
      "learning_rate": 2.071800208116545e-05,
      "loss": 0.165,
      "step": 7820
    },
    {
      "epoch": 23.94495412844037,
      "grad_norm": 4.021772861480713,
      "learning_rate": 2.061394380853278e-05,
      "loss": 0.1099,
      "step": 7830
    },
    {
      "epoch": 23.97553516819572,
      "grad_norm": 2.0271944999694824,
      "learning_rate": 2.0509885535900104e-05,
      "loss": 0.1267,
      "step": 7840
    },
    {
      "epoch": 24.00611620795107,
      "grad_norm": 3.9538087844848633,
      "learning_rate": 2.040582726326743e-05,
      "loss": 0.1773,
      "step": 7850
    },
    {
      "epoch": 24.03669724770642,
      "grad_norm": 4.695821285247803,
      "learning_rate": 2.0301768990634758e-05,
      "loss": 0.0898,
      "step": 7860
    },
    {
      "epoch": 24.067278287461775,
      "grad_norm": 3.817899227142334,
      "learning_rate": 2.0197710718002083e-05,
      "loss": 0.1105,
      "step": 7870
    },
    {
      "epoch": 24.097859327217126,
      "grad_norm": 6.2242536544799805,
      "learning_rate": 2.0093652445369408e-05,
      "loss": 0.1275,
      "step": 7880
    },
    {
      "epoch": 24.128440366972477,
      "grad_norm": 6.821825981140137,
      "learning_rate": 1.9989594172736733e-05,
      "loss": 0.1637,
      "step": 7890
    },
    {
      "epoch": 24.159021406727827,
      "grad_norm": 1.6077946424484253,
      "learning_rate": 1.988553590010406e-05,
      "loss": 0.1687,
      "step": 7900
    },
    {
      "epoch": 24.189602446483182,
      "grad_norm": 3.4850125312805176,
      "learning_rate": 1.9781477627471383e-05,
      "loss": 0.1124,
      "step": 7910
    },
    {
      "epoch": 24.220183486238533,
      "grad_norm": 7.470705509185791,
      "learning_rate": 1.967741935483871e-05,
      "loss": 0.106,
      "step": 7920
    },
    {
      "epoch": 24.250764525993883,
      "grad_norm": 7.171663761138916,
      "learning_rate": 1.9573361082206036e-05,
      "loss": 0.1532,
      "step": 7930
    },
    {
      "epoch": 24.281345565749234,
      "grad_norm": 1.5503133535385132,
      "learning_rate": 1.946930280957336e-05,
      "loss": 0.1559,
      "step": 7940
    },
    {
      "epoch": 24.31192660550459,
      "grad_norm": 2.006579875946045,
      "learning_rate": 1.936524453694069e-05,
      "loss": 0.1435,
      "step": 7950
    },
    {
      "epoch": 24.34250764525994,
      "grad_norm": 5.5806756019592285,
      "learning_rate": 1.926118626430801e-05,
      "loss": 0.1127,
      "step": 7960
    },
    {
      "epoch": 24.37308868501529,
      "grad_norm": 4.723907470703125,
      "learning_rate": 1.915712799167534e-05,
      "loss": 0.1293,
      "step": 7970
    },
    {
      "epoch": 24.40366972477064,
      "grad_norm": 4.086251258850098,
      "learning_rate": 1.9053069719042665e-05,
      "loss": 0.1407,
      "step": 7980
    },
    {
      "epoch": 24.434250764525995,
      "grad_norm": 4.814991474151611,
      "learning_rate": 1.894901144640999e-05,
      "loss": 0.1398,
      "step": 7990
    },
    {
      "epoch": 24.464831804281346,
      "grad_norm": 2.7642455101013184,
      "learning_rate": 1.8844953173777315e-05,
      "loss": 0.1645,
      "step": 8000
    },
    {
      "epoch": 24.464831804281346,
      "eval_loss": 0.4339035153388977,
      "eval_runtime": 17.7992,
      "eval_samples_per_second": 16.349,
      "eval_steps_per_second": 8.203,
      "step": 8000
    },
    {
      "epoch": 24.495412844036696,
      "grad_norm": 2.2885043621063232,
      "learning_rate": 1.8740894901144643e-05,
      "loss": 0.1168,
      "step": 8010
    },
    {
      "epoch": 24.525993883792047,
      "grad_norm": 4.910865783691406,
      "learning_rate": 1.8636836628511968e-05,
      "loss": 0.1254,
      "step": 8020
    },
    {
      "epoch": 24.5565749235474,
      "grad_norm": 3.811713457107544,
      "learning_rate": 1.8532778355879293e-05,
      "loss": 0.1535,
      "step": 8030
    },
    {
      "epoch": 24.587155963302752,
      "grad_norm": 3.527937650680542,
      "learning_rate": 1.842872008324662e-05,
      "loss": 0.1472,
      "step": 8040
    },
    {
      "epoch": 24.617737003058103,
      "grad_norm": 1.450607180595398,
      "learning_rate": 1.8324661810613943e-05,
      "loss": 0.2001,
      "step": 8050
    },
    {
      "epoch": 24.648318042813457,
      "grad_norm": 3.5444912910461426,
      "learning_rate": 1.822060353798127e-05,
      "loss": 0.1147,
      "step": 8060
    },
    {
      "epoch": 24.678899082568808,
      "grad_norm": 6.098001480102539,
      "learning_rate": 1.8116545265348597e-05,
      "loss": 0.1144,
      "step": 8070
    },
    {
      "epoch": 24.70948012232416,
      "grad_norm": 2.527642011642456,
      "learning_rate": 1.8012486992715922e-05,
      "loss": 0.1105,
      "step": 8080
    },
    {
      "epoch": 24.74006116207951,
      "grad_norm": 3.0823781490325928,
      "learning_rate": 1.7908428720083247e-05,
      "loss": 0.1126,
      "step": 8090
    },
    {
      "epoch": 24.770642201834864,
      "grad_norm": 2.1876089572906494,
      "learning_rate": 1.7804370447450572e-05,
      "loss": 0.1586,
      "step": 8100
    },
    {
      "epoch": 24.801223241590215,
      "grad_norm": 2.728623390197754,
      "learning_rate": 1.77003121748179e-05,
      "loss": 0.1086,
      "step": 8110
    },
    {
      "epoch": 24.831804281345565,
      "grad_norm": 1.5861458778381348,
      "learning_rate": 1.7596253902185225e-05,
      "loss": 0.0836,
      "step": 8120
    },
    {
      "epoch": 24.862385321100916,
      "grad_norm": 2.1710517406463623,
      "learning_rate": 1.749219562955255e-05,
      "loss": 0.1129,
      "step": 8130
    },
    {
      "epoch": 24.89296636085627,
      "grad_norm": 4.549278736114502,
      "learning_rate": 1.7388137356919875e-05,
      "loss": 0.0956,
      "step": 8140
    },
    {
      "epoch": 24.92354740061162,
      "grad_norm": 2.860056161880493,
      "learning_rate": 1.7284079084287204e-05,
      "loss": 0.1406,
      "step": 8150
    },
    {
      "epoch": 24.954128440366972,
      "grad_norm": 2.902945041656494,
      "learning_rate": 1.7180020811654525e-05,
      "loss": 0.1217,
      "step": 8160
    },
    {
      "epoch": 24.984709480122323,
      "grad_norm": 6.512890815734863,
      "learning_rate": 1.7075962539021854e-05,
      "loss": 0.1534,
      "step": 8170
    },
    {
      "epoch": 25.015290519877677,
      "grad_norm": 2.204719066619873,
      "learning_rate": 1.697190426638918e-05,
      "loss": 0.1133,
      "step": 8180
    },
    {
      "epoch": 25.045871559633028,
      "grad_norm": 4.600652694702148,
      "learning_rate": 1.6867845993756504e-05,
      "loss": 0.1018,
      "step": 8190
    },
    {
      "epoch": 25.07645259938838,
      "grad_norm": 3.88714861869812,
      "learning_rate": 1.676378772112383e-05,
      "loss": 0.1141,
      "step": 8200
    },
    {
      "epoch": 25.10703363914373,
      "grad_norm": 2.694763660430908,
      "learning_rate": 1.6659729448491157e-05,
      "loss": 0.105,
      "step": 8210
    },
    {
      "epoch": 25.137614678899084,
      "grad_norm": 5.237843990325928,
      "learning_rate": 1.6555671175858482e-05,
      "loss": 0.134,
      "step": 8220
    },
    {
      "epoch": 25.168195718654435,
      "grad_norm": 2.7006750106811523,
      "learning_rate": 1.6451612903225807e-05,
      "loss": 0.1189,
      "step": 8230
    },
    {
      "epoch": 25.198776758409785,
      "grad_norm": 5.1693115234375,
      "learning_rate": 1.6347554630593136e-05,
      "loss": 0.1248,
      "step": 8240
    },
    {
      "epoch": 25.229357798165136,
      "grad_norm": 4.986877918243408,
      "learning_rate": 1.6243496357960457e-05,
      "loss": 0.1507,
      "step": 8250
    },
    {
      "epoch": 25.25993883792049,
      "grad_norm": 1.4471940994262695,
      "learning_rate": 1.6139438085327786e-05,
      "loss": 0.1265,
      "step": 8260
    },
    {
      "epoch": 25.29051987767584,
      "grad_norm": 5.540067672729492,
      "learning_rate": 1.6035379812695107e-05,
      "loss": 0.1422,
      "step": 8270
    },
    {
      "epoch": 25.321100917431192,
      "grad_norm": 4.367918014526367,
      "learning_rate": 1.5931321540062436e-05,
      "loss": 0.1546,
      "step": 8280
    },
    {
      "epoch": 25.351681957186543,
      "grad_norm": 2.7890803813934326,
      "learning_rate": 1.582726326742976e-05,
      "loss": 0.0769,
      "step": 8290
    },
    {
      "epoch": 25.382262996941897,
      "grad_norm": 1.6187372207641602,
      "learning_rate": 1.5723204994797086e-05,
      "loss": 0.1068,
      "step": 8300
    },
    {
      "epoch": 25.412844036697248,
      "grad_norm": 3.0006723403930664,
      "learning_rate": 1.5619146722164414e-05,
      "loss": 0.1262,
      "step": 8310
    },
    {
      "epoch": 25.4434250764526,
      "grad_norm": 4.67561674118042,
      "learning_rate": 1.551508844953174e-05,
      "loss": 0.1522,
      "step": 8320
    },
    {
      "epoch": 25.474006116207953,
      "grad_norm": 1.59788179397583,
      "learning_rate": 1.5411030176899064e-05,
      "loss": 0.1875,
      "step": 8330
    },
    {
      "epoch": 25.504587155963304,
      "grad_norm": 2.5928938388824463,
      "learning_rate": 1.530697190426639e-05,
      "loss": 0.0929,
      "step": 8340
    },
    {
      "epoch": 25.535168195718654,
      "grad_norm": 3.5686800479888916,
      "learning_rate": 1.5202913631633716e-05,
      "loss": 0.1185,
      "step": 8350
    },
    {
      "epoch": 25.565749235474005,
      "grad_norm": 5.571007251739502,
      "learning_rate": 1.5098855359001041e-05,
      "loss": 0.1214,
      "step": 8360
    },
    {
      "epoch": 25.59633027522936,
      "grad_norm": 9.126275062561035,
      "learning_rate": 1.4994797086368368e-05,
      "loss": 0.1516,
      "step": 8370
    },
    {
      "epoch": 25.62691131498471,
      "grad_norm": 2.9007301330566406,
      "learning_rate": 1.4890738813735694e-05,
      "loss": 0.1372,
      "step": 8380
    },
    {
      "epoch": 25.65749235474006,
      "grad_norm": 3.1054575443267822,
      "learning_rate": 1.4786680541103018e-05,
      "loss": 0.1055,
      "step": 8390
    },
    {
      "epoch": 25.68807339449541,
      "grad_norm": 4.602766990661621,
      "learning_rate": 1.4682622268470344e-05,
      "loss": 0.1056,
      "step": 8400
    },
    {
      "epoch": 25.718654434250766,
      "grad_norm": 2.9134981632232666,
      "learning_rate": 1.4578563995837668e-05,
      "loss": 0.1074,
      "step": 8410
    },
    {
      "epoch": 25.749235474006117,
      "grad_norm": 4.738123893737793,
      "learning_rate": 1.4474505723204994e-05,
      "loss": 0.1139,
      "step": 8420
    },
    {
      "epoch": 25.779816513761467,
      "grad_norm": 5.146528720855713,
      "learning_rate": 1.4370447450572321e-05,
      "loss": 0.1333,
      "step": 8430
    },
    {
      "epoch": 25.810397553516818,
      "grad_norm": 1.8651458024978638,
      "learning_rate": 1.4266389177939646e-05,
      "loss": 0.1101,
      "step": 8440
    },
    {
      "epoch": 25.840978593272173,
      "grad_norm": 3.372619152069092,
      "learning_rate": 1.4162330905306973e-05,
      "loss": 0.1285,
      "step": 8450
    },
    {
      "epoch": 25.871559633027523,
      "grad_norm": 4.9253363609313965,
      "learning_rate": 1.40582726326743e-05,
      "loss": 0.127,
      "step": 8460
    },
    {
      "epoch": 25.902140672782874,
      "grad_norm": 4.081017971038818,
      "learning_rate": 1.3954214360041623e-05,
      "loss": 0.1061,
      "step": 8470
    },
    {
      "epoch": 25.932721712538225,
      "grad_norm": 1.752782940864563,
      "learning_rate": 1.385015608740895e-05,
      "loss": 0.138,
      "step": 8480
    },
    {
      "epoch": 25.96330275229358,
      "grad_norm": 1.8544362783432007,
      "learning_rate": 1.3746097814776276e-05,
      "loss": 0.0969,
      "step": 8490
    },
    {
      "epoch": 25.99388379204893,
      "grad_norm": 6.175636291503906,
      "learning_rate": 1.36420395421436e-05,
      "loss": 0.1628,
      "step": 8500
    },
    {
      "epoch": 25.99388379204893,
      "eval_loss": 0.46008336544036865,
      "eval_runtime": 17.473,
      "eval_samples_per_second": 16.654,
      "eval_steps_per_second": 8.356,
      "step": 8500
    },
    {
      "epoch": 26.02446483180428,
      "grad_norm": 4.042583465576172,
      "learning_rate": 1.3537981269510926e-05,
      "loss": 0.116,
      "step": 8510
    },
    {
      "epoch": 26.05504587155963,
      "grad_norm": 3.8415026664733887,
      "learning_rate": 1.3433922996878253e-05,
      "loss": 0.1106,
      "step": 8520
    },
    {
      "epoch": 26.085626911314986,
      "grad_norm": 3.7799582481384277,
      "learning_rate": 1.3329864724245578e-05,
      "loss": 0.0886,
      "step": 8530
    },
    {
      "epoch": 26.116207951070336,
      "grad_norm": 6.632818698883057,
      "learning_rate": 1.3225806451612905e-05,
      "loss": 0.1219,
      "step": 8540
    },
    {
      "epoch": 26.146788990825687,
      "grad_norm": 6.225608825683594,
      "learning_rate": 1.3121748178980228e-05,
      "loss": 0.1348,
      "step": 8550
    },
    {
      "epoch": 26.17737003058104,
      "grad_norm": 5.72769832611084,
      "learning_rate": 1.3017689906347555e-05,
      "loss": 0.1163,
      "step": 8560
    },
    {
      "epoch": 26.207951070336392,
      "grad_norm": 3.572054147720337,
      "learning_rate": 1.2913631633714882e-05,
      "loss": 0.0977,
      "step": 8570
    },
    {
      "epoch": 26.238532110091743,
      "grad_norm": 5.419766426086426,
      "learning_rate": 1.2809573361082205e-05,
      "loss": 0.1244,
      "step": 8580
    },
    {
      "epoch": 26.269113149847094,
      "grad_norm": 4.254158973693848,
      "learning_rate": 1.2705515088449532e-05,
      "loss": 0.1376,
      "step": 8590
    },
    {
      "epoch": 26.299694189602448,
      "grad_norm": 7.157454013824463,
      "learning_rate": 1.2601456815816858e-05,
      "loss": 0.182,
      "step": 8600
    },
    {
      "epoch": 26.3302752293578,
      "grad_norm": 1.7647346258163452,
      "learning_rate": 1.2497398543184183e-05,
      "loss": 0.1129,
      "step": 8610
    },
    {
      "epoch": 26.36085626911315,
      "grad_norm": 3.3649981021881104,
      "learning_rate": 1.239334027055151e-05,
      "loss": 0.0961,
      "step": 8620
    },
    {
      "epoch": 26.3914373088685,
      "grad_norm": 7.596597194671631,
      "learning_rate": 1.2289281997918835e-05,
      "loss": 0.1253,
      "step": 8630
    },
    {
      "epoch": 26.422018348623855,
      "grad_norm": 4.112176418304443,
      "learning_rate": 1.2185223725286162e-05,
      "loss": 0.129,
      "step": 8640
    },
    {
      "epoch": 26.452599388379205,
      "grad_norm": 2.766486644744873,
      "learning_rate": 1.2081165452653487e-05,
      "loss": 0.1485,
      "step": 8650
    },
    {
      "epoch": 26.483180428134556,
      "grad_norm": 3.5616629123687744,
      "learning_rate": 1.1977107180020812e-05,
      "loss": 0.1298,
      "step": 8660
    },
    {
      "epoch": 26.513761467889907,
      "grad_norm": 2.5189101696014404,
      "learning_rate": 1.1873048907388137e-05,
      "loss": 0.1013,
      "step": 8670
    },
    {
      "epoch": 26.54434250764526,
      "grad_norm": 2.5541913509368896,
      "learning_rate": 1.1768990634755464e-05,
      "loss": 0.1185,
      "step": 8680
    },
    {
      "epoch": 26.574923547400612,
      "grad_norm": 11.571924209594727,
      "learning_rate": 1.1664932362122789e-05,
      "loss": 0.1364,
      "step": 8690
    },
    {
      "epoch": 26.605504587155963,
      "grad_norm": 6.36802339553833,
      "learning_rate": 1.1560874089490115e-05,
      "loss": 0.1866,
      "step": 8700
    },
    {
      "epoch": 26.636085626911314,
      "grad_norm": 3.861549139022827,
      "learning_rate": 1.1456815816857442e-05,
      "loss": 0.0996,
      "step": 8710
    },
    {
      "epoch": 26.666666666666668,
      "grad_norm": 1.4031004905700684,
      "learning_rate": 1.1352757544224767e-05,
      "loss": 0.0686,
      "step": 8720
    },
    {
      "epoch": 26.69724770642202,
      "grad_norm": 3.146599054336548,
      "learning_rate": 1.1248699271592092e-05,
      "loss": 0.1375,
      "step": 8730
    },
    {
      "epoch": 26.72782874617737,
      "grad_norm": 3.636478900909424,
      "learning_rate": 1.1144640998959417e-05,
      "loss": 0.1234,
      "step": 8740
    },
    {
      "epoch": 26.75840978593272,
      "grad_norm": 2.630683660507202,
      "learning_rate": 1.1040582726326744e-05,
      "loss": 0.1324,
      "step": 8750
    },
    {
      "epoch": 26.788990825688074,
      "grad_norm": 2.7682950496673584,
      "learning_rate": 1.0936524453694069e-05,
      "loss": 0.08,
      "step": 8760
    },
    {
      "epoch": 26.819571865443425,
      "grad_norm": 2.6922128200531006,
      "learning_rate": 1.0832466181061394e-05,
      "loss": 0.1352,
      "step": 8770
    },
    {
      "epoch": 26.850152905198776,
      "grad_norm": 6.487355709075928,
      "learning_rate": 1.072840790842872e-05,
      "loss": 0.1123,
      "step": 8780
    },
    {
      "epoch": 26.88073394495413,
      "grad_norm": 2.877392530441284,
      "learning_rate": 1.0624349635796047e-05,
      "loss": 0.1286,
      "step": 8790
    },
    {
      "epoch": 26.91131498470948,
      "grad_norm": 4.019444465637207,
      "learning_rate": 1.0520291363163372e-05,
      "loss": 0.1137,
      "step": 8800
    },
    {
      "epoch": 26.941896024464832,
      "grad_norm": 2.5582876205444336,
      "learning_rate": 1.0416233090530697e-05,
      "loss": 0.1026,
      "step": 8810
    },
    {
      "epoch": 26.972477064220183,
      "grad_norm": 2.6786487102508545,
      "learning_rate": 1.0312174817898024e-05,
      "loss": 0.1093,
      "step": 8820
    },
    {
      "epoch": 27.003058103975537,
      "grad_norm": 1.792412281036377,
      "learning_rate": 1.0208116545265349e-05,
      "loss": 0.1588,
      "step": 8830
    },
    {
      "epoch": 27.033639143730888,
      "grad_norm": 1.3316484689712524,
      "learning_rate": 1.0104058272632674e-05,
      "loss": 0.0828,
      "step": 8840
    },
    {
      "epoch": 27.06422018348624,
      "grad_norm": 2.7321407794952393,
      "learning_rate": 1e-05,
      "loss": 0.0917,
      "step": 8850
    },
    {
      "epoch": 27.09480122324159,
      "grad_norm": 1.614942193031311,
      "learning_rate": 9.895941727367326e-06,
      "loss": 0.1273,
      "step": 8860
    },
    {
      "epoch": 27.125382262996943,
      "grad_norm": 6.456067085266113,
      "learning_rate": 9.791883454734651e-06,
      "loss": 0.1361,
      "step": 8870
    },
    {
      "epoch": 27.155963302752294,
      "grad_norm": 1.9082986116409302,
      "learning_rate": 9.687825182101978e-06,
      "loss": 0.1529,
      "step": 8880
    },
    {
      "epoch": 27.186544342507645,
      "grad_norm": 3.1089670658111572,
      "learning_rate": 9.583766909469304e-06,
      "loss": 0.0918,
      "step": 8890
    },
    {
      "epoch": 27.217125382262996,
      "grad_norm": 2.7883081436157227,
      "learning_rate": 9.47970863683663e-06,
      "loss": 0.1191,
      "step": 8900
    },
    {
      "epoch": 27.24770642201835,
      "grad_norm": 3.825026035308838,
      "learning_rate": 9.375650364203954e-06,
      "loss": 0.113,
      "step": 8910
    },
    {
      "epoch": 27.2782874617737,
      "grad_norm": 9.698514938354492,
      "learning_rate": 9.271592091571281e-06,
      "loss": 0.0969,
      "step": 8920
    },
    {
      "epoch": 27.30886850152905,
      "grad_norm": 1.5716568231582642,
      "learning_rate": 9.167533818938606e-06,
      "loss": 0.1342,
      "step": 8930
    },
    {
      "epoch": 27.339449541284402,
      "grad_norm": 2.928072690963745,
      "learning_rate": 9.063475546305931e-06,
      "loss": 0.0834,
      "step": 8940
    },
    {
      "epoch": 27.370030581039757,
      "grad_norm": 2.3197057247161865,
      "learning_rate": 8.959417273673256e-06,
      "loss": 0.1149,
      "step": 8950
    },
    {
      "epoch": 27.400611620795107,
      "grad_norm": 7.332393646240234,
      "learning_rate": 8.855359001040583e-06,
      "loss": 0.1016,
      "step": 8960
    },
    {
      "epoch": 27.431192660550458,
      "grad_norm": 4.15636682510376,
      "learning_rate": 8.75130072840791e-06,
      "loss": 0.1038,
      "step": 8970
    },
    {
      "epoch": 27.46177370030581,
      "grad_norm": 1.6034491062164307,
      "learning_rate": 8.647242455775235e-06,
      "loss": 0.1807,
      "step": 8980
    },
    {
      "epoch": 27.492354740061163,
      "grad_norm": 4.26979398727417,
      "learning_rate": 8.543184183142561e-06,
      "loss": 0.1017,
      "step": 8990
    },
    {
      "epoch": 27.522935779816514,
      "grad_norm": 2.539623975753784,
      "learning_rate": 8.439125910509886e-06,
      "loss": 0.0951,
      "step": 9000
    },
    {
      "epoch": 27.522935779816514,
      "eval_loss": 0.4583536684513092,
      "eval_runtime": 17.3431,
      "eval_samples_per_second": 16.779,
      "eval_steps_per_second": 8.418,
      "step": 9000
    },
    {
      "epoch": 27.553516819571865,
      "grad_norm": 7.396114349365234,
      "learning_rate": 8.335067637877211e-06,
      "loss": 0.1293,
      "step": 9010
    },
    {
      "epoch": 27.584097859327215,
      "grad_norm": 3.8735439777374268,
      "learning_rate": 8.231009365244538e-06,
      "loss": 0.0991,
      "step": 9020
    },
    {
      "epoch": 27.61467889908257,
      "grad_norm": 2.752223491668701,
      "learning_rate": 8.126951092611863e-06,
      "loss": 0.1317,
      "step": 9030
    },
    {
      "epoch": 27.64525993883792,
      "grad_norm": 3.082855224609375,
      "learning_rate": 8.022892819979188e-06,
      "loss": 0.0865,
      "step": 9040
    },
    {
      "epoch": 27.67584097859327,
      "grad_norm": 3.818155527114868,
      "learning_rate": 7.918834547346513e-06,
      "loss": 0.0893,
      "step": 9050
    },
    {
      "epoch": 27.706422018348626,
      "grad_norm": 4.239770412445068,
      "learning_rate": 7.81477627471384e-06,
      "loss": 0.1284,
      "step": 9060
    },
    {
      "epoch": 27.737003058103976,
      "grad_norm": 2.538876533508301,
      "learning_rate": 7.710718002081167e-06,
      "loss": 0.1326,
      "step": 9070
    },
    {
      "epoch": 27.767584097859327,
      "grad_norm": 2.477917432785034,
      "learning_rate": 7.6066597294484915e-06,
      "loss": 0.1679,
      "step": 9080
    },
    {
      "epoch": 27.798165137614678,
      "grad_norm": 6.487174987792969,
      "learning_rate": 7.502601456815818e-06,
      "loss": 0.0955,
      "step": 9090
    },
    {
      "epoch": 27.828746177370032,
      "grad_norm": 3.586022138595581,
      "learning_rate": 7.398543184183143e-06,
      "loss": 0.1289,
      "step": 9100
    },
    {
      "epoch": 27.859327217125383,
      "grad_norm": 4.1030049324035645,
      "learning_rate": 7.294484911550468e-06,
      "loss": 0.1064,
      "step": 9110
    },
    {
      "epoch": 27.889908256880734,
      "grad_norm": 2.473414421081543,
      "learning_rate": 7.190426638917794e-06,
      "loss": 0.1148,
      "step": 9120
    },
    {
      "epoch": 27.920489296636084,
      "grad_norm": 2.4675068855285645,
      "learning_rate": 7.086368366285121e-06,
      "loss": 0.1222,
      "step": 9130
    },
    {
      "epoch": 27.95107033639144,
      "grad_norm": 2.7934279441833496,
      "learning_rate": 6.982310093652446e-06,
      "loss": 0.0951,
      "step": 9140
    },
    {
      "epoch": 27.98165137614679,
      "grad_norm": 7.207830429077148,
      "learning_rate": 6.878251821019771e-06,
      "loss": 0.1336,
      "step": 9150
    },
    {
      "epoch": 28.01223241590214,
      "grad_norm": 2.382178544998169,
      "learning_rate": 6.774193548387098e-06,
      "loss": 0.0944,
      "step": 9160
    },
    {
      "epoch": 28.04281345565749,
      "grad_norm": 3.0280263423919678,
      "learning_rate": 6.670135275754423e-06,
      "loss": 0.0803,
      "step": 9170
    },
    {
      "epoch": 28.073394495412845,
      "grad_norm": 2.1945552825927734,
      "learning_rate": 6.5660770031217485e-06,
      "loss": 0.0828,
      "step": 9180
    },
    {
      "epoch": 28.103975535168196,
      "grad_norm": 4.57396936416626,
      "learning_rate": 6.4620187304890735e-06,
      "loss": 0.124,
      "step": 9190
    },
    {
      "epoch": 28.134556574923547,
      "grad_norm": 9.008644104003906,
      "learning_rate": 6.3579604578564e-06,
      "loss": 0.1269,
      "step": 9200
    },
    {
      "epoch": 28.165137614678898,
      "grad_norm": 2.750105857849121,
      "learning_rate": 6.253902185223725e-06,
      "loss": 0.1416,
      "step": 9210
    },
    {
      "epoch": 28.195718654434252,
      "grad_norm": 5.317841529846191,
      "learning_rate": 6.149843912591052e-06,
      "loss": 0.1053,
      "step": 9220
    },
    {
      "epoch": 28.226299694189603,
      "grad_norm": 2.429410219192505,
      "learning_rate": 6.045785639958377e-06,
      "loss": 0.1105,
      "step": 9230
    },
    {
      "epoch": 28.256880733944953,
      "grad_norm": 4.838024139404297,
      "learning_rate": 5.941727367325703e-06,
      "loss": 0.1138,
      "step": 9240
    },
    {
      "epoch": 28.287461773700304,
      "grad_norm": 12.658930778503418,
      "learning_rate": 5.837669094693028e-06,
      "loss": 0.13,
      "step": 9250
    },
    {
      "epoch": 28.31804281345566,
      "grad_norm": 1.7371057271957397,
      "learning_rate": 5.733610822060355e-06,
      "loss": 0.1251,
      "step": 9260
    },
    {
      "epoch": 28.34862385321101,
      "grad_norm": 2.2298595905303955,
      "learning_rate": 5.62955254942768e-06,
      "loss": 0.0757,
      "step": 9270
    },
    {
      "epoch": 28.37920489296636,
      "grad_norm": 4.572708606719971,
      "learning_rate": 5.5254942767950055e-06,
      "loss": 0.1127,
      "step": 9280
    },
    {
      "epoch": 28.40978593272171,
      "grad_norm": 3.1214189529418945,
      "learning_rate": 5.421436004162331e-06,
      "loss": 0.1395,
      "step": 9290
    },
    {
      "epoch": 28.440366972477065,
      "grad_norm": 3.5031135082244873,
      "learning_rate": 5.317377731529656e-06,
      "loss": 0.1255,
      "step": 9300
    },
    {
      "epoch": 28.470948012232416,
      "grad_norm": 2.455829620361328,
      "learning_rate": 5.213319458896983e-06,
      "loss": 0.0986,
      "step": 9310
    },
    {
      "epoch": 28.501529051987767,
      "grad_norm": 2.4105734825134277,
      "learning_rate": 5.109261186264308e-06,
      "loss": 0.1033,
      "step": 9320
    },
    {
      "epoch": 28.53211009174312,
      "grad_norm": 5.415170669555664,
      "learning_rate": 5.005202913631634e-06,
      "loss": 0.106,
      "step": 9330
    },
    {
      "epoch": 28.56269113149847,
      "grad_norm": 1.5614656209945679,
      "learning_rate": 4.901144640998959e-06,
      "loss": 0.1119,
      "step": 9340
    },
    {
      "epoch": 28.593272171253822,
      "grad_norm": 11.242796897888184,
      "learning_rate": 4.797086368366286e-06,
      "loss": 0.1031,
      "step": 9350
    },
    {
      "epoch": 28.623853211009173,
      "grad_norm": 1.98354172706604,
      "learning_rate": 4.693028095733612e-06,
      "loss": 0.1192,
      "step": 9360
    },
    {
      "epoch": 28.654434250764528,
      "grad_norm": 3.658066749572754,
      "learning_rate": 4.588969823100937e-06,
      "loss": 0.08,
      "step": 9370
    },
    {
      "epoch": 28.68501529051988,
      "grad_norm": 2.7829060554504395,
      "learning_rate": 4.4849115504682625e-06,
      "loss": 0.0842,
      "step": 9380
    },
    {
      "epoch": 28.71559633027523,
      "grad_norm": 3.9607882499694824,
      "learning_rate": 4.3808532778355875e-06,
      "loss": 0.0955,
      "step": 9390
    },
    {
      "epoch": 28.74617737003058,
      "grad_norm": 5.205819606781006,
      "learning_rate": 4.276795005202914e-06,
      "loss": 0.1385,
      "step": 9400
    },
    {
      "epoch": 28.776758409785934,
      "grad_norm": 2.6670114994049072,
      "learning_rate": 4.172736732570239e-06,
      "loss": 0.1485,
      "step": 9410
    },
    {
      "epoch": 28.807339449541285,
      "grad_norm": 6.059555530548096,
      "learning_rate": 4.068678459937565e-06,
      "loss": 0.0743,
      "step": 9420
    },
    {
      "epoch": 28.837920489296636,
      "grad_norm": 2.8076841831207275,
      "learning_rate": 3.964620187304891e-06,
      "loss": 0.0958,
      "step": 9430
    },
    {
      "epoch": 28.868501529051986,
      "grad_norm": 6.3077616691589355,
      "learning_rate": 3.860561914672217e-06,
      "loss": 0.0761,
      "step": 9440
    },
    {
      "epoch": 28.89908256880734,
      "grad_norm": 7.713011264801025,
      "learning_rate": 3.7565036420395423e-06,
      "loss": 0.0894,
      "step": 9450
    },
    {
      "epoch": 28.92966360856269,
      "grad_norm": 2.3206980228424072,
      "learning_rate": 3.6524453694068677e-06,
      "loss": 0.1241,
      "step": 9460
    },
    {
      "epoch": 28.960244648318042,
      "grad_norm": 3.4398224353790283,
      "learning_rate": 3.5483870967741936e-06,
      "loss": 0.0933,
      "step": 9470
    },
    {
      "epoch": 28.990825688073393,
      "grad_norm": 7.903160572052002,
      "learning_rate": 3.44432882414152e-06,
      "loss": 0.1263,
      "step": 9480
    },
    {
      "epoch": 29.021406727828747,
      "grad_norm": 1.9907540082931519,
      "learning_rate": 3.340270551508845e-06,
      "loss": 0.0967,
      "step": 9490
    },
    {
      "epoch": 29.051987767584098,
      "grad_norm": 1.7297402620315552,
      "learning_rate": 3.236212278876171e-06,
      "loss": 0.0935,
      "step": 9500
    },
    {
      "epoch": 29.051987767584098,
      "eval_loss": 0.46150296926498413,
      "eval_runtime": 17.3886,
      "eval_samples_per_second": 16.735,
      "eval_steps_per_second": 8.396,
      "step": 9500
    },
    {
      "epoch": 29.08256880733945,
      "grad_norm": 4.011406421661377,
      "learning_rate": 3.1321540062434962e-06,
      "loss": 0.1097,
      "step": 9510
    },
    {
      "epoch": 29.1131498470948,
      "grad_norm": 3.3061842918395996,
      "learning_rate": 3.0280957336108225e-06,
      "loss": 0.1327,
      "step": 9520
    },
    {
      "epoch": 29.143730886850154,
      "grad_norm": 11.67180061340332,
      "learning_rate": 2.924037460978148e-06,
      "loss": 0.1092,
      "step": 9530
    },
    {
      "epoch": 29.174311926605505,
      "grad_norm": 4.720720291137695,
      "learning_rate": 2.819979188345474e-06,
      "loss": 0.1067,
      "step": 9540
    },
    {
      "epoch": 29.204892966360855,
      "grad_norm": 3.8615598678588867,
      "learning_rate": 2.7159209157127993e-06,
      "loss": 0.0915,
      "step": 9550
    },
    {
      "epoch": 29.235474006116206,
      "grad_norm": 4.769123554229736,
      "learning_rate": 2.6118626430801247e-06,
      "loss": 0.1029,
      "step": 9560
    },
    {
      "epoch": 29.26605504587156,
      "grad_norm": 10.71865177154541,
      "learning_rate": 2.5078043704474506e-06,
      "loss": 0.1094,
      "step": 9570
    },
    {
      "epoch": 29.29663608562691,
      "grad_norm": 4.421198844909668,
      "learning_rate": 2.4037460978147764e-06,
      "loss": 0.1504,
      "step": 9580
    },
    {
      "epoch": 29.327217125382262,
      "grad_norm": 4.271670341491699,
      "learning_rate": 2.2996878251821023e-06,
      "loss": 0.1025,
      "step": 9590
    },
    {
      "epoch": 29.357798165137616,
      "grad_norm": 3.598496913909912,
      "learning_rate": 2.1956295525494278e-06,
      "loss": 0.1062,
      "step": 9600
    },
    {
      "epoch": 29.388379204892967,
      "grad_norm": 2.3889999389648438,
      "learning_rate": 2.0915712799167536e-06,
      "loss": 0.0964,
      "step": 9610
    },
    {
      "epoch": 29.418960244648318,
      "grad_norm": 4.520473957061768,
      "learning_rate": 1.987513007284079e-06,
      "loss": 0.1036,
      "step": 9620
    },
    {
      "epoch": 29.44954128440367,
      "grad_norm": 6.458251476287842,
      "learning_rate": 1.8834547346514047e-06,
      "loss": 0.129,
      "step": 9630
    },
    {
      "epoch": 29.480122324159023,
      "grad_norm": 2.3070294857025146,
      "learning_rate": 1.7793964620187304e-06,
      "loss": 0.1162,
      "step": 9640
    },
    {
      "epoch": 29.510703363914374,
      "grad_norm": 2.367161512374878,
      "learning_rate": 1.6753381893860565e-06,
      "loss": 0.104,
      "step": 9650
    },
    {
      "epoch": 29.541284403669724,
      "grad_norm": 6.095667839050293,
      "learning_rate": 1.5712799167533821e-06,
      "loss": 0.0903,
      "step": 9660
    },
    {
      "epoch": 29.571865443425075,
      "grad_norm": 10.448101043701172,
      "learning_rate": 1.4672216441207078e-06,
      "loss": 0.1061,
      "step": 9670
    },
    {
      "epoch": 29.60244648318043,
      "grad_norm": 3.1639466285705566,
      "learning_rate": 1.3631633714880334e-06,
      "loss": 0.1307,
      "step": 9680
    },
    {
      "epoch": 29.63302752293578,
      "grad_norm": 2.028094530105591,
      "learning_rate": 1.259105098855359e-06,
      "loss": 0.1,
      "step": 9690
    },
    {
      "epoch": 29.66360856269113,
      "grad_norm": 4.976256847381592,
      "learning_rate": 1.1550468262226847e-06,
      "loss": 0.0844,
      "step": 9700
    },
    {
      "epoch": 29.69418960244648,
      "grad_norm": 4.531723499298096,
      "learning_rate": 1.0509885535900104e-06,
      "loss": 0.1207,
      "step": 9710
    },
    {
      "epoch": 29.724770642201836,
      "grad_norm": 3.63309645652771,
      "learning_rate": 9.46930280957336e-07,
      "loss": 0.0812,
      "step": 9720
    },
    {
      "epoch": 29.755351681957187,
      "grad_norm": 14.414322853088379,
      "learning_rate": 8.428720083246619e-07,
      "loss": 0.1532,
      "step": 9730
    },
    {
      "epoch": 29.785932721712538,
      "grad_norm": 3.6692416667938232,
      "learning_rate": 7.388137356919876e-07,
      "loss": 0.1036,
      "step": 9740
    },
    {
      "epoch": 29.81651376146789,
      "grad_norm": 2.8107922077178955,
      "learning_rate": 6.347554630593132e-07,
      "loss": 0.0884,
      "step": 9750
    },
    {
      "epoch": 29.847094801223243,
      "grad_norm": 9.009915351867676,
      "learning_rate": 5.306971904266389e-07,
      "loss": 0.0975,
      "step": 9760
    },
    {
      "epoch": 29.877675840978593,
      "grad_norm": 4.969494819641113,
      "learning_rate": 4.2663891779396464e-07,
      "loss": 0.1096,
      "step": 9770
    },
    {
      "epoch": 29.908256880733944,
      "grad_norm": 3.147620916366577,
      "learning_rate": 3.2258064516129035e-07,
      "loss": 0.1088,
      "step": 9780
    },
    {
      "epoch": 29.938837920489295,
      "grad_norm": 12.04923152923584,
      "learning_rate": 2.1852237252861603e-07,
      "loss": 0.1043,
      "step": 9790
    },
    {
      "epoch": 29.96941896024465,
      "grad_norm": 4.238800048828125,
      "learning_rate": 1.1446409989594173e-07,
      "loss": 0.1298,
      "step": 9800
    },
    {
      "epoch": 30.0,
      "grad_norm": 3.123566150665283,
      "learning_rate": 1.040582726326743e-08,
      "loss": 0.099,
      "step": 9810
    }
  ],
  "logging_steps": 10,
  "max_steps": 9810,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 30,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 3.5687795117499023e+18,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}