{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 9480,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0010548523206751054,
      "grad_norm": 1.281753659248352,
      "learning_rate": 0.00015789473684210527,
      "loss": 7.5122,
      "step": 10
    },
    {
      "epoch": 0.002109704641350211,
      "grad_norm": 1.1707550287246704,
      "learning_rate": 0.00031578947368421053,
      "loss": 6.9093,
      "step": 20
    },
    {
      "epoch": 0.0031645569620253164,
      "grad_norm": 0.8732145428657532,
      "learning_rate": 0.00047368421052631577,
      "loss": 6.2582,
      "step": 30
    },
    {
      "epoch": 0.004219409282700422,
      "grad_norm": 0.911007285118103,
      "learning_rate": 0.0006315789473684211,
      "loss": 5.7652,
      "step": 40
    },
    {
      "epoch": 0.005274261603375527,
      "grad_norm": 1.2406620979309082,
      "learning_rate": 0.0007894736842105263,
      "loss": 5.3073,
      "step": 50
    },
    {
      "epoch": 0.006329113924050633,
      "grad_norm": 1.3634661436080933,
      "learning_rate": 0.0009473684210526315,
      "loss": 4.7877,
      "step": 60
    },
    {
      "epoch": 0.007383966244725738,
      "grad_norm": 1.0261237621307373,
      "learning_rate": 0.0011052631578947368,
      "loss": 4.385,
      "step": 70
    },
    {
      "epoch": 0.008438818565400843,
      "grad_norm": 1.5167491436004639,
      "learning_rate": 0.0012631578947368421,
      "loss": 4.1392,
      "step": 80
    },
    {
      "epoch": 0.00949367088607595,
      "grad_norm": 0.849207878112793,
      "learning_rate": 0.0014210526315789472,
      "loss": 3.9325,
      "step": 90
    },
    {
      "epoch": 0.010548523206751054,
      "grad_norm": 0.8754621148109436,
      "learning_rate": 0.0014999989494847376,
      "loss": 3.7779,
      "step": 100
    },
    {
      "epoch": 0.011603375527426161,
      "grad_norm": 0.58613520860672,
      "learning_rate": 0.0014999905453802946,
      "loss": 3.6232,
      "step": 110
    },
    {
      "epoch": 0.012658227848101266,
      "grad_norm": 0.9189424514770508,
      "learning_rate": 0.0014999737372655805,
      "loss": 3.5039,
      "step": 120
    },
    {
      "epoch": 0.013713080168776372,
      "grad_norm": 0.7585026621818542,
      "learning_rate": 0.0014999485253289388,
      "loss": 3.399,
      "step": 130
    },
    {
      "epoch": 0.014767932489451477,
      "grad_norm": 0.8590853214263916,
      "learning_rate": 0.0014999149098528814,
      "loss": 3.301,
      "step": 140
    },
    {
      "epoch": 0.015822784810126583,
      "grad_norm": 0.5862516760826111,
      "learning_rate": 0.0014998728912140862,
      "loss": 3.2138,
      "step": 150
    },
    {
      "epoch": 0.016877637130801686,
      "grad_norm": 0.8057186007499695,
      "learning_rate": 0.0014998224698833922,
      "loss": 3.1519,
      "step": 160
    },
    {
      "epoch": 0.017932489451476793,
      "grad_norm": 0.8671597242355347,
      "learning_rate": 0.0014997636464257956,
      "loss": 3.0799,
      "step": 170
    },
    {
      "epoch": 0.0189873417721519,
      "grad_norm": 1.039334774017334,
      "learning_rate": 0.0014996964215004416,
      "loss": 3.035,
      "step": 180
    },
    {
      "epoch": 0.020042194092827006,
      "grad_norm": 0.824468731880188,
      "learning_rate": 0.0014996207958606182,
      "loss": 2.9722,
      "step": 190
    },
    {
      "epoch": 0.02109704641350211,
      "grad_norm": 0.8207864761352539,
      "learning_rate": 0.001499536770353748,
      "loss": 2.9257,
      "step": 200
    },
    {
      "epoch": 0.022151898734177215,
      "grad_norm": 0.8991343975067139,
      "learning_rate": 0.0014994443459213774,
      "loss": 2.8889,
      "step": 210
    },
    {
      "epoch": 0.023206751054852322,
      "grad_norm": 0.8417520523071289,
      "learning_rate": 0.001499343523599168,
      "loss": 2.8514,
      "step": 220
    },
    {
      "epoch": 0.024261603375527425,
      "grad_norm": 0.6757763028144836,
      "learning_rate": 0.0014992343045168823,
      "loss": 2.8067,
      "step": 230
    },
    {
      "epoch": 0.02531645569620253,
      "grad_norm": 0.7877993583679199,
      "learning_rate": 0.0014991166898983739,
      "loss": 2.7601,
      "step": 240
    },
    {
      "epoch": 0.026371308016877638,
      "grad_norm": 0.9585819840431213,
      "learning_rate": 0.001498990681061572,
      "loss": 2.7274,
      "step": 250
    },
    {
      "epoch": 0.027426160337552744,
      "grad_norm": 1.006851077079773,
      "learning_rate": 0.001498856279418467,
      "loss": 2.6915,
      "step": 260
    },
    {
      "epoch": 0.028481012658227847,
      "grad_norm": 0.8644171357154846,
      "learning_rate": 0.0014987134864750948,
      "loss": 2.6607,
      "step": 270
    },
    {
      "epoch": 0.029535864978902954,
      "grad_norm": 0.782991349697113,
      "learning_rate": 0.0014985623038315206,
      "loss": 2.6331,
      "step": 280
    },
    {
      "epoch": 0.03059071729957806,
      "grad_norm": 0.9489585757255554,
      "learning_rate": 0.0014984027331818193,
      "loss": 2.6008,
      "step": 290
    },
    {
      "epoch": 0.03164556962025317,
      "grad_norm": 1.01542329788208,
      "learning_rate": 0.0014982347763140584,
      "loss": 2.5784,
      "step": 300
    },
    {
      "epoch": 0.03270042194092827,
      "grad_norm": 0.7982136607170105,
      "learning_rate": 0.0014980584351102762,
      "loss": 2.551,
      "step": 310
    },
    {
      "epoch": 0.03375527426160337,
      "grad_norm": 1.389794111251831,
      "learning_rate": 0.001497873711546462,
      "loss": 2.5257,
      "step": 320
    },
    {
      "epoch": 0.03481012658227848,
      "grad_norm": 0.7754802107810974,
      "learning_rate": 0.0014976806076925334,
      "loss": 2.5162,
      "step": 330
    },
    {
      "epoch": 0.035864978902953586,
      "grad_norm": 0.9034753441810608,
      "learning_rate": 0.0014974791257123137,
      "loss": 2.4903,
      "step": 340
    },
    {
      "epoch": 0.03691983122362869,
      "grad_norm": 0.763302743434906,
      "learning_rate": 0.001497269267863507,
      "loss": 2.4517,
      "step": 350
    },
    {
      "epoch": 0.0379746835443038,
      "grad_norm": 1.0804508924484253,
      "learning_rate": 0.0014970510364976724,
      "loss": 2.4307,
      "step": 360
    },
    {
      "epoch": 0.039029535864978905,
      "grad_norm": 0.8051794767379761,
      "learning_rate": 0.0014968244340601996,
      "loss": 2.4207,
      "step": 370
    },
    {
      "epoch": 0.04008438818565401,
      "grad_norm": 1.0223069190979004,
      "learning_rate": 0.001496589463090279,
      "loss": 2.4081,
      "step": 380
    },
    {
      "epoch": 0.04113924050632911,
      "grad_norm": 1.0545941591262817,
      "learning_rate": 0.001496346126220875,
      "loss": 2.3881,
      "step": 390
    },
    {
      "epoch": 0.04219409282700422,
      "grad_norm": 0.7329314947128296,
      "learning_rate": 0.0014960944261786966,
      "loss": 2.3588,
      "step": 400
    },
    {
      "epoch": 0.043248945147679324,
      "grad_norm": 1.5836018323898315,
      "learning_rate": 0.0014958343657841655,
      "loss": 2.3532,
      "step": 410
    },
    {
      "epoch": 0.04430379746835443,
      "grad_norm": 1.0970488786697388,
      "learning_rate": 0.001495565947951385,
      "loss": 2.3399,
      "step": 420
    },
    {
      "epoch": 0.04535864978902954,
      "grad_norm": 0.8173145651817322,
      "learning_rate": 0.0014952891756881085,
      "loss": 2.3135,
      "step": 430
    },
    {
      "epoch": 0.046413502109704644,
      "grad_norm": 1.2800233364105225,
      "learning_rate": 0.0014950040520957037,
      "loss": 2.2934,
      "step": 440
    },
    {
      "epoch": 0.04746835443037975,
      "grad_norm": 0.8970997333526611,
      "learning_rate": 0.0014947105803691204,
      "loss": 2.2981,
      "step": 450
    },
    {
      "epoch": 0.04852320675105485,
      "grad_norm": 0.7999330163002014,
      "learning_rate": 0.0014944087637968522,
      "loss": 2.2757,
      "step": 460
    },
    {
      "epoch": 0.049578059071729956,
      "grad_norm": 0.7487330436706543,
      "learning_rate": 0.0014940986057609012,
      "loss": 2.2548,
      "step": 470
    },
    {
      "epoch": 0.05063291139240506,
      "grad_norm": 0.8741482496261597,
      "learning_rate": 0.0014937801097367396,
      "loss": 2.2454,
      "step": 480
    },
    {
      "epoch": 0.05168776371308017,
      "grad_norm": 0.8719668388366699,
      "learning_rate": 0.001493453279293271,
      "loss": 2.2234,
      "step": 490
    },
    {
      "epoch": 0.052742616033755275,
      "grad_norm": 0.8162099123001099,
      "learning_rate": 0.0014931181180927902,
      "loss": 2.2204,
      "step": 500
    },
    {
      "epoch": 0.05379746835443038,
      "grad_norm": 0.839779257774353,
      "learning_rate": 0.001492774629890942,
      "loss": 2.2083,
      "step": 510
    },
    {
      "epoch": 0.05485232067510549,
      "grad_norm": 0.7697337865829468,
      "learning_rate": 0.001492422818536679,
      "loss": 2.1983,
      "step": 520
    },
    {
      "epoch": 0.05590717299578059,
      "grad_norm": 0.7342979907989502,
      "learning_rate": 0.00149206268797222,
      "loss": 2.1783,
      "step": 530
    },
    {
      "epoch": 0.056962025316455694,
      "grad_norm": 0.9562809467315674,
      "learning_rate": 0.0014916942422330032,
      "loss": 2.163,
      "step": 540
    },
    {
      "epoch": 0.0580168776371308,
      "grad_norm": 0.6906911134719849,
      "learning_rate": 0.001491317485447643,
      "loss": 2.1658,
      "step": 550
    },
    {
      "epoch": 0.05907172995780591,
      "grad_norm": 0.8576987981796265,
      "learning_rate": 0.0014909324218378838,
      "loss": 2.1279,
      "step": 560
    },
    {
      "epoch": 0.060126582278481014,
      "grad_norm": 0.7869489789009094,
      "learning_rate": 0.0014905390557185508,
      "loss": 2.134,
      "step": 570
    },
    {
      "epoch": 0.06118143459915612,
      "grad_norm": 1.1268903017044067,
      "learning_rate": 0.0014901373914975036,
      "loss": 2.1339,
      "step": 580
    },
    {
      "epoch": 0.06223628691983123,
      "grad_norm": 0.8695049285888672,
      "learning_rate": 0.0014897274336755856,
      "loss": 2.1158,
      "step": 590
    },
    {
      "epoch": 0.06329113924050633,
      "grad_norm": 1.2872540950775146,
      "learning_rate": 0.001489309186846575,
      "loss": 2.0942,
      "step": 600
    },
    {
      "epoch": 0.06434599156118144,
      "grad_norm": 0.8001000285148621,
      "learning_rate": 0.0014888826556971313,
      "loss": 2.0952,
      "step": 610
    },
    {
      "epoch": 0.06540084388185655,
      "grad_norm": 0.8871486186981201,
      "learning_rate": 0.0014884478450067444,
      "loss": 2.0855,
      "step": 620
    },
    {
      "epoch": 0.06645569620253164,
      "grad_norm": 1.0699702501296997,
      "learning_rate": 0.0014880047596476807,
      "loss": 2.0688,
      "step": 630
    },
    {
      "epoch": 0.06751054852320675,
      "grad_norm": 0.7940484285354614,
      "learning_rate": 0.0014875534045849274,
      "loss": 2.0818,
      "step": 640
    },
    {
      "epoch": 0.06856540084388185,
      "grad_norm": 0.9331493973731995,
      "learning_rate": 0.0014870937848761388,
      "loss": 2.0704,
      "step": 650
    },
    {
      "epoch": 0.06962025316455696,
      "grad_norm": 0.9877519011497498,
      "learning_rate": 0.001486625905671578,
      "loss": 2.0657,
      "step": 660
    },
    {
      "epoch": 0.07067510548523206,
      "grad_norm": 0.8880543112754822,
      "learning_rate": 0.00148614977221406,
      "loss": 2.0417,
      "step": 670
    },
    {
      "epoch": 0.07172995780590717,
      "grad_norm": 0.865791380405426,
      "learning_rate": 0.0014856653898388927,
      "loss": 2.0386,
      "step": 680
    },
    {
      "epoch": 0.07278481012658228,
      "grad_norm": 0.7400647401809692,
      "learning_rate": 0.001485172763973817,
      "loss": 2.0399,
      "step": 690
    },
    {
      "epoch": 0.07383966244725738,
      "grad_norm": 0.8002055883407593,
      "learning_rate": 0.0014846719001389466,
      "loss": 2.0251,
      "step": 700
    },
    {
      "epoch": 0.07489451476793249,
      "grad_norm": 0.7134732604026794,
      "learning_rate": 0.001484162803946705,
      "loss": 2.0148,
      "step": 710
    },
    {
      "epoch": 0.0759493670886076,
      "grad_norm": 0.8022155165672302,
      "learning_rate": 0.0014836454811017635,
      "loss": 1.9978,
      "step": 720
    },
    {
      "epoch": 0.0770042194092827,
      "grad_norm": 0.9345313906669617,
      "learning_rate": 0.0014831199374009778,
      "loss": 1.9957,
      "step": 730
    },
    {
      "epoch": 0.07805907172995781,
      "grad_norm": 0.737798273563385,
      "learning_rate": 0.0014825861787333208,
      "loss": 2.0038,
      "step": 740
    },
    {
      "epoch": 0.07911392405063292,
      "grad_norm": 0.8962474465370178,
      "learning_rate": 0.0014820442110798197,
      "loss": 1.9859,
      "step": 750
    },
    {
      "epoch": 0.08016877637130802,
      "grad_norm": 0.7600175142288208,
      "learning_rate": 0.0014814940405134865,
      "loss": 1.9878,
      "step": 760
    },
    {
      "epoch": 0.08122362869198312,
      "grad_norm": 0.9256311655044556,
      "learning_rate": 0.001480935673199251,
      "loss": 1.9724,
      "step": 770
    },
    {
      "epoch": 0.08227848101265822,
      "grad_norm": 0.8020193576812744,
      "learning_rate": 0.0014803691153938915,
      "loss": 1.968,
      "step": 780
    },
    {
      "epoch": 0.08333333333333333,
      "grad_norm": 0.924941897392273,
      "learning_rate": 0.0014797943734459653,
      "loss": 1.9725,
      "step": 790
    },
    {
      "epoch": 0.08438818565400844,
      "grad_norm": 0.7024965286254883,
      "learning_rate": 0.001479211453795736,
      "loss": 1.9482,
      "step": 800
    },
    {
      "epoch": 0.08544303797468354,
      "grad_norm": 0.813156247138977,
      "learning_rate": 0.0014786203629751033,
      "loss": 1.9374,
      "step": 810
    },
    {
      "epoch": 0.08649789029535865,
      "grad_norm": 0.9243624210357666,
      "learning_rate": 0.0014780211076075279,
      "loss": 1.9442,
      "step": 820
    },
    {
      "epoch": 0.08755274261603375,
      "grad_norm": 1.046899437904358,
      "learning_rate": 0.0014774136944079594,
      "loss": 1.9462,
      "step": 830
    },
    {
      "epoch": 0.08860759493670886,
      "grad_norm": 0.8379732966423035,
      "learning_rate": 0.0014767981301827592,
      "loss": 1.9239,
      "step": 840
    },
    {
      "epoch": 0.08966244725738397,
      "grad_norm": 0.7675923109054565,
      "learning_rate": 0.0014761744218296249,
      "loss": 1.927,
      "step": 850
    },
    {
      "epoch": 0.09071729957805907,
      "grad_norm": 0.6716737151145935,
      "learning_rate": 0.001475542576337513,
      "loss": 1.9253,
      "step": 860
    },
    {
      "epoch": 0.09177215189873418,
      "grad_norm": 0.9615976214408875,
      "learning_rate": 0.001474902600786561,
      "loss": 1.9201,
      "step": 870
    },
    {
      "epoch": 0.09282700421940929,
      "grad_norm": 1.0707359313964844,
      "learning_rate": 0.0014742545023480075,
      "loss": 1.9155,
      "step": 880
    },
    {
      "epoch": 0.0938818565400844,
      "grad_norm": 0.9218859672546387,
      "learning_rate": 0.0014735982882841117,
      "loss": 1.9089,
      "step": 890
    },
    {
      "epoch": 0.0949367088607595,
      "grad_norm": 1.312366247177124,
      "learning_rate": 0.0014729339659480727,
      "loss": 1.8998,
      "step": 900
    },
    {
      "epoch": 0.09599156118143459,
      "grad_norm": 0.8077645301818848,
      "learning_rate": 0.0014722615427839468,
      "loss": 1.9071,
      "step": 910
    },
    {
      "epoch": 0.0970464135021097,
      "grad_norm": 0.9507492184638977,
      "learning_rate": 0.0014715810263265633,
      "loss": 1.9039,
      "step": 920
    },
    {
      "epoch": 0.0981012658227848,
      "grad_norm": 1.5656588077545166,
      "learning_rate": 0.0014708924242014423,
      "loss": 1.894,
      "step": 930
    },
    {
      "epoch": 0.09915611814345991,
      "grad_norm": 0.9367612600326538,
      "learning_rate": 0.0014701957441247064,
      "loss": 1.8873,
      "step": 940
    },
    {
      "epoch": 0.10021097046413502,
      "grad_norm": 0.8315409421920776,
      "learning_rate": 0.0014694909939029959,
      "loss": 1.8625,
      "step": 950
    },
    {
      "epoch": 0.10126582278481013,
      "grad_norm": 0.7968072891235352,
      "learning_rate": 0.0014687781814333814,
      "loss": 1.8767,
      "step": 960
    },
    {
      "epoch": 0.10232067510548523,
      "grad_norm": 0.877154529094696,
      "learning_rate": 0.0014680573147032746,
      "loss": 1.8799,
      "step": 970
    },
    {
      "epoch": 0.10337552742616034,
      "grad_norm": 0.7661100029945374,
      "learning_rate": 0.0014673284017903392,
      "loss": 1.8588,
      "step": 980
    },
    {
      "epoch": 0.10443037974683544,
      "grad_norm": 0.8402513861656189,
      "learning_rate": 0.0014665914508624,
      "loss": 1.8606,
      "step": 990
    },
    {
      "epoch": 0.10548523206751055,
      "grad_norm": 0.992628276348114,
      "learning_rate": 0.0014658464701773526,
      "loss": 1.8721,
      "step": 1000
    },
    {
      "epoch": 0.10654008438818566,
      "grad_norm": 1.1602342128753662,
      "learning_rate": 0.0014650934680830688,
      "loss": 1.8614,
      "step": 1010
    },
    {
      "epoch": 0.10759493670886076,
      "grad_norm": 1.093902349472046,
      "learning_rate": 0.0014643324530173051,
      "loss": 1.8541,
      "step": 1020
    },
    {
      "epoch": 0.10864978902953587,
      "grad_norm": 0.8267302513122559,
      "learning_rate": 0.0014635634335076067,
      "loss": 1.8529,
      "step": 1030
    },
    {
      "epoch": 0.10970464135021098,
      "grad_norm": 0.741106390953064,
      "learning_rate": 0.001462786418171213,
      "loss": 1.8433,
      "step": 1040
    },
    {
      "epoch": 0.11075949367088607,
      "grad_norm": 1.0178207159042358,
      "learning_rate": 0.0014620014157149597,
      "loss": 1.8409,
      "step": 1050
    },
    {
      "epoch": 0.11181434599156118,
      "grad_norm": 0.9372745156288147,
      "learning_rate": 0.001461208434935183,
      "loss": 1.8329,
      "step": 1060
    },
    {
      "epoch": 0.11286919831223628,
      "grad_norm": 0.7344574332237244,
      "learning_rate": 0.0014604074847176197,
      "loss": 1.8226,
      "step": 1070
    },
    {
      "epoch": 0.11392405063291139,
      "grad_norm": 0.8814179301261902,
      "learning_rate": 0.0014595985740373082,
      "loss": 1.8243,
      "step": 1080
    },
    {
      "epoch": 0.1149789029535865,
      "grad_norm": 0.9900213479995728,
      "learning_rate": 0.0014587817119584873,
      "loss": 1.8354,
      "step": 1090
    },
    {
      "epoch": 0.1160337552742616,
      "grad_norm": 0.9492560625076294,
      "learning_rate": 0.001457956907634496,
      "loss": 1.8209,
      "step": 1100
    },
    {
      "epoch": 0.11708860759493671,
      "grad_norm": 0.8075652718544006,
      "learning_rate": 0.0014571241703076692,
      "loss": 1.824,
      "step": 1110
    },
    {
      "epoch": 0.11814345991561181,
      "grad_norm": 0.694759726524353,
      "learning_rate": 0.0014562835093092348,
      "loss": 1.8105,
      "step": 1120
    },
    {
      "epoch": 0.11919831223628692,
      "grad_norm": 1.068515658378601,
      "learning_rate": 0.0014554349340592104,
      "loss": 1.7969,
      "step": 1130
    },
    {
      "epoch": 0.12025316455696203,
      "grad_norm": 0.9704002737998962,
      "learning_rate": 0.001454578454066296,
      "loss": 1.8132,
      "step": 1140
    },
    {
      "epoch": 0.12130801687763713,
      "grad_norm": 1.1903709173202515,
      "learning_rate": 0.0014537140789277678,
      "loss": 1.7983,
      "step": 1150
    },
    {
      "epoch": 0.12236286919831224,
      "grad_norm": 0.8845016360282898,
      "learning_rate": 0.0014528418183293716,
      "loss": 1.8129,
      "step": 1160
    },
    {
      "epoch": 0.12341772151898735,
      "grad_norm": 1.1104865074157715,
      "learning_rate": 0.001451961682045213,
      "loss": 1.8035,
      "step": 1170
    },
    {
      "epoch": 0.12447257383966245,
      "grad_norm": 0.8706414699554443,
      "learning_rate": 0.001451073679937649,
      "loss": 1.7847,
      "step": 1180
    },
    {
      "epoch": 0.12552742616033755,
      "grad_norm": 0.6799960136413574,
      "learning_rate": 0.0014501778219571766,
      "loss": 1.7706,
      "step": 1190
    },
    {
      "epoch": 0.12658227848101267,
      "grad_norm": 0.8295177221298218,
      "learning_rate": 0.0014492741181423225,
      "loss": 1.7895,
      "step": 1200
    },
    {
      "epoch": 0.12763713080168776,
      "grad_norm": 1.107081413269043,
      "learning_rate": 0.0014483625786195285,
      "loss": 1.7914,
      "step": 1210
    },
    {
      "epoch": 0.12869198312236288,
      "grad_norm": 0.7618156671524048,
      "learning_rate": 0.0014474432136030405,
      "loss": 1.7717,
      "step": 1220
    },
    {
      "epoch": 0.12974683544303797,
      "grad_norm": 0.8248320817947388,
      "learning_rate": 0.0014465160333947923,
      "loss": 1.7667,
      "step": 1230
    },
    {
      "epoch": 0.1308016877637131,
      "grad_norm": 0.7723872661590576,
      "learning_rate": 0.0014455810483842908,
      "loss": 1.7821,
      "step": 1240
    },
    {
      "epoch": 0.13185654008438819,
      "grad_norm": 0.7738557457923889,
      "learning_rate": 0.0014446382690484997,
      "loss": 1.7826,
      "step": 1250
    },
    {
      "epoch": 0.13291139240506328,
      "grad_norm": 0.8339896202087402,
      "learning_rate": 0.0014436877059517215,
      "loss": 1.7709,
      "step": 1260
    },
    {
      "epoch": 0.1339662447257384,
      "grad_norm": 0.7648717164993286,
      "learning_rate": 0.0014427293697454803,
      "loss": 1.7792,
      "step": 1270
    },
    {
      "epoch": 0.1350210970464135,
      "grad_norm": 0.7579119205474854,
      "learning_rate": 0.001441763271168401,
      "loss": 1.7691,
      "step": 1280
    },
    {
      "epoch": 0.1360759493670886,
      "grad_norm": 0.884143590927124,
      "learning_rate": 0.00144078942104609,
      "loss": 1.7589,
      "step": 1290
    },
    {
      "epoch": 0.1371308016877637,
      "grad_norm": 0.7581146359443665,
      "learning_rate": 0.001439807830291013,
      "loss": 1.7529,
      "step": 1300
    },
    {
      "epoch": 0.13818565400843882,
      "grad_norm": 0.7806832194328308,
      "learning_rate": 0.0014388185099023744,
      "loss": 1.7567,
      "step": 1310
    },
    {
      "epoch": 0.13924050632911392,
      "grad_norm": 0.9394419193267822,
      "learning_rate": 0.0014378214709659916,
      "loss": 1.7606,
      "step": 1320
    },
    {
      "epoch": 0.14029535864978904,
      "grad_norm": 1.15559983253479,
      "learning_rate": 0.0014368167246541733,
      "loss": 1.7549,
      "step": 1330
    },
    {
      "epoch": 0.14135021097046413,
      "grad_norm": 2.064676523208618,
      "learning_rate": 0.0014358042822255918,
      "loss": 1.7576,
      "step": 1340
    },
    {
      "epoch": 0.14240506329113925,
      "grad_norm": 0.9762903451919556,
      "learning_rate": 0.0014347841550251597,
      "loss": 1.7615,
      "step": 1350
    },
    {
      "epoch": 0.14345991561181434,
      "grad_norm": 0.7241702079772949,
      "learning_rate": 0.0014337563544838997,
      "loss": 1.7407,
      "step": 1360
    },
    {
      "epoch": 0.14451476793248946,
      "grad_norm": 0.7886254191398621,
      "learning_rate": 0.001432720892118819,
      "loss": 1.7411,
      "step": 1370
    },
    {
      "epoch": 0.14556962025316456,
      "grad_norm": 0.986682653427124,
      "learning_rate": 0.0014316777795327794,
      "loss": 1.7334,
      "step": 1380
    },
    {
      "epoch": 0.14662447257383968,
      "grad_norm": 0.9229147434234619,
      "learning_rate": 0.001430627028414366,
      "loss": 1.7454,
      "step": 1390
    },
    {
      "epoch": 0.14767932489451477,
      "grad_norm": 0.8327695727348328,
      "learning_rate": 0.0014295686505377586,
      "loss": 1.7256,
      "step": 1400
    },
    {
      "epoch": 0.14873417721518986,
      "grad_norm": 0.6828871369361877,
      "learning_rate": 0.0014285026577625982,
      "loss": 1.728,
      "step": 1410
    },
    {
      "epoch": 0.14978902953586498,
      "grad_norm": 0.7544017434120178,
      "learning_rate": 0.0014274290620338542,
      "loss": 1.739,
      "step": 1420
    },
    {
      "epoch": 0.15084388185654007,
      "grad_norm": 0.8737987875938416,
      "learning_rate": 0.0014263478753816906,
      "loss": 1.7309,
      "step": 1430
    },
    {
      "epoch": 0.1518987341772152,
      "grad_norm": 0.815551221370697,
      "learning_rate": 0.0014252591099213326,
      "loss": 1.723,
      "step": 1440
    },
    {
      "epoch": 0.1529535864978903,
      "grad_norm": 0.7114374041557312,
      "learning_rate": 0.001424162777852928,
      "loss": 1.7257,
      "step": 1450
    },
    {
      "epoch": 0.1540084388185654,
      "grad_norm": 0.7523857355117798,
      "learning_rate": 0.0014230588914614134,
      "loss": 1.7216,
      "step": 1460
    },
    {
      "epoch": 0.1550632911392405,
      "grad_norm": 0.9350714087486267,
      "learning_rate": 0.0014219474631163745,
      "loss": 1.7179,
      "step": 1470
    },
    {
      "epoch": 0.15611814345991562,
      "grad_norm": 0.7759507894515991,
      "learning_rate": 0.001420828505271909,
      "loss": 1.7189,
      "step": 1480
    },
    {
      "epoch": 0.1571729957805907,
      "grad_norm": 0.6648463606834412,
      "learning_rate": 0.0014197020304664856,
      "loss": 1.7155,
      "step": 1490
    },
    {
      "epoch": 0.15822784810126583,
      "grad_norm": 0.722392737865448,
      "learning_rate": 0.0014185680513228048,
      "loss": 1.7202,
      "step": 1500
    },
    {
      "epoch": 0.15928270042194093,
      "grad_norm": 0.9407737255096436,
      "learning_rate": 0.0014174265805476564,
      "loss": 1.7161,
      "step": 1510
    },
    {
      "epoch": 0.16033755274261605,
      "grad_norm": 0.7787057757377625,
      "learning_rate": 0.0014162776309317778,
      "loss": 1.7159,
      "step": 1520
    },
    {
      "epoch": 0.16139240506329114,
      "grad_norm": 0.7414025068283081,
      "learning_rate": 0.0014151212153497108,
      "loss": 1.6982,
      "step": 1530
    },
    {
      "epoch": 0.16244725738396623,
      "grad_norm": 0.7560591697692871,
      "learning_rate": 0.0014139573467596561,
      "loss": 1.6896,
      "step": 1540
    },
    {
      "epoch": 0.16350210970464135,
      "grad_norm": 0.6860440969467163,
      "learning_rate": 0.00141278603820333,
      "loss": 1.6907,
      "step": 1550
    },
    {
      "epoch": 0.16455696202531644,
      "grad_norm": 1.0000585317611694,
      "learning_rate": 0.0014116073028058165,
      "loss": 1.6918,
      "step": 1560
    },
    {
      "epoch": 0.16561181434599156,
      "grad_norm": 0.7621354460716248,
      "learning_rate": 0.0014104211537754217,
      "loss": 1.6871,
      "step": 1570
    },
    {
      "epoch": 0.16666666666666666,
      "grad_norm": 0.786876916885376,
      "learning_rate": 0.001409227604403524,
      "loss": 1.6967,
      "step": 1580
    },
    {
      "epoch": 0.16772151898734178,
      "grad_norm": 1.2952214479446411,
      "learning_rate": 0.0014080266680644277,
      "loss": 1.6904,
      "step": 1590
    },
    {
      "epoch": 0.16877637130801687,
      "grad_norm": 1.0267448425292969,
      "learning_rate": 0.0014068183582152103,
      "loss": 1.6962,
      "step": 1600
    },
    {
      "epoch": 0.169831223628692,
      "grad_norm": 0.9297143816947937,
      "learning_rate": 0.001405602688395574,
      "loss": 1.6934,
      "step": 1610
    },
    {
      "epoch": 0.17088607594936708,
      "grad_norm": 1.0625101327896118,
      "learning_rate": 0.0014043796722276924,
      "loss": 1.6704,
      "step": 1620
    },
    {
      "epoch": 0.1719409282700422,
      "grad_norm": 1.2425276041030884,
      "learning_rate": 0.0014031493234160591,
      "loss": 1.6779,
      "step": 1630
    },
    {
      "epoch": 0.1729957805907173,
      "grad_norm": 1.4398293495178223,
      "learning_rate": 0.0014019116557473332,
      "loss": 1.6881,
      "step": 1640
    },
    {
      "epoch": 0.17405063291139242,
      "grad_norm": 0.7769531607627869,
      "learning_rate": 0.0014006666830901854,
      "loss": 1.664,
      "step": 1650
    },
    {
      "epoch": 0.1751054852320675,
      "grad_norm": 0.8062126040458679,
      "learning_rate": 0.001399414419395142,
      "loss": 1.6585,
      "step": 1660
    },
    {
      "epoch": 0.17616033755274263,
      "grad_norm": 0.9165257811546326,
      "learning_rate": 0.0013981548786944293,
      "loss": 1.6764,
      "step": 1670
    },
    {
      "epoch": 0.17721518987341772,
      "grad_norm": 1.0061345100402832,
      "learning_rate": 0.0013968880751018158,
      "loss": 1.6655,
      "step": 1680
    },
    {
      "epoch": 0.17827004219409281,
      "grad_norm": 1.026166558265686,
      "learning_rate": 0.0013956140228124545,
      "loss": 1.6588,
      "step": 1690
    },
    {
      "epoch": 0.17932489451476794,
      "grad_norm": 0.7095503211021423,
      "learning_rate": 0.0013943327361027231,
      "loss": 1.6777,
      "step": 1700
    },
    {
      "epoch": 0.18037974683544303,
      "grad_norm": 0.7568669319152832,
      "learning_rate": 0.0013930442293300649,
      "loss": 1.6609,
      "step": 1710
    },
    {
      "epoch": 0.18143459915611815,
      "grad_norm": 0.7210203409194946,
      "learning_rate": 0.0013917485169328279,
      "loss": 1.6526,
      "step": 1720
    },
    {
      "epoch": 0.18248945147679324,
      "grad_norm": 0.763261616230011,
      "learning_rate": 0.0013904456134301016,
      "loss": 1.6611,
      "step": 1730
    },
    {
      "epoch": 0.18354430379746836,
      "grad_norm": 0.8423966765403748,
      "learning_rate": 0.0013891355334215562,
      "loss": 1.6602,
      "step": 1740
    },
    {
      "epoch": 0.18459915611814345,
      "grad_norm": 0.7480831146240234,
      "learning_rate": 0.0013878182915872776,
      "loss": 1.6824,
      "step": 1750
    },
    {
      "epoch": 0.18565400843881857,
      "grad_norm": 0.715969979763031,
      "learning_rate": 0.001386493902687604,
      "loss": 1.66,
      "step": 1760
    },
    {
      "epoch": 0.18670886075949367,
      "grad_norm": 0.6824060678482056,
      "learning_rate": 0.00138516238156296,
      "loss": 1.6617,
      "step": 1770
    },
    {
      "epoch": 0.1877637130801688,
      "grad_norm": 0.9273524880409241,
      "learning_rate": 0.0013838237431336895,
      "loss": 1.6661,
      "step": 1780
    },
    {
      "epoch": 0.18881856540084388,
      "grad_norm": 2.018726348876953,
      "learning_rate": 0.0013824780023998899,
      "loss": 1.6741,
      "step": 1790
    },
    {
      "epoch": 0.189873417721519,
      "grad_norm": 1.502572774887085,
      "learning_rate": 0.0013811251744412431,
      "loss": 1.6691,
      "step": 1800
    },
    {
      "epoch": 0.1909282700421941,
      "grad_norm": 0.90468829870224,
      "learning_rate": 0.0013797652744168473,
      "loss": 1.6386,
      "step": 1810
    },
    {
      "epoch": 0.19198312236286919,
      "grad_norm": 0.8139953017234802,
      "learning_rate": 0.0013783983175650457,
      "loss": 1.634,
      "step": 1820
    },
    {
      "epoch": 0.1930379746835443,
      "grad_norm": 0.7573320865631104,
      "learning_rate": 0.0013770243192032581,
      "loss": 1.6393,
      "step": 1830
    },
    {
      "epoch": 0.1940928270042194,
      "grad_norm": 0.803521454334259,
      "learning_rate": 0.0013756432947278064,
      "loss": 1.6417,
      "step": 1840
    },
    {
      "epoch": 0.19514767932489452,
      "grad_norm": 0.676461398601532,
      "learning_rate": 0.0013742552596137444,
      "loss": 1.641,
      "step": 1850
    },
    {
      "epoch": 0.1962025316455696,
      "grad_norm": 0.7045630216598511,
      "learning_rate": 0.0013728602294146833,
      "loss": 1.6496,
      "step": 1860
    },
    {
      "epoch": 0.19725738396624473,
      "grad_norm": 0.7004404067993164,
      "learning_rate": 0.0013714582197626175,
      "loss": 1.6488,
      "step": 1870
    },
    {
      "epoch": 0.19831223628691982,
      "grad_norm": 0.6805158853530884,
      "learning_rate": 0.0013700492463677501,
      "loss": 1.635,
      "step": 1880
    },
    {
      "epoch": 0.19936708860759494,
      "grad_norm": 0.6834223866462708,
      "learning_rate": 0.0013686333250183154,
      "loss": 1.6435,
      "step": 1890
    },
    {
      "epoch": 0.20042194092827004,
      "grad_norm": 1.061880350112915,
      "learning_rate": 0.001367210471580404,
      "loss": 1.6397,
      "step": 1900
    },
    {
      "epoch": 0.20147679324894516,
      "grad_norm": 1.1606338024139404,
      "learning_rate": 0.0013657807019977835,
      "loss": 1.6374,
      "step": 1910
    },
    {
      "epoch": 0.20253164556962025,
      "grad_norm": 0.7113993167877197,
      "learning_rate": 0.0013643440322917198,
      "loss": 1.6342,
      "step": 1920
    },
    {
      "epoch": 0.20358649789029537,
      "grad_norm": 1.2351900339126587,
      "learning_rate": 0.0013629004785607989,
      "loss": 1.6323,
      "step": 1930
    },
    {
      "epoch": 0.20464135021097046,
      "grad_norm": 0.7484331130981445,
      "learning_rate": 0.0013614500569807445,
      "loss": 1.627,
      "step": 1940
    },
    {
      "epoch": 0.20569620253164558,
      "grad_norm": 0.739666223526001,
      "learning_rate": 0.0013599927838042394,
      "loss": 1.6288,
      "step": 1950
    },
    {
      "epoch": 0.20675105485232068,
      "grad_norm": 1.1977354288101196,
      "learning_rate": 0.0013585286753607408,
      "loss": 1.6293,
      "step": 1960
    },
    {
      "epoch": 0.20780590717299577,
      "grad_norm": 0.6730018258094788,
      "learning_rate": 0.0013570577480562986,
      "loss": 1.6401,
      "step": 1970
    },
    {
      "epoch": 0.2088607594936709,
      "grad_norm": 1.1211644411087036,
      "learning_rate": 0.0013555800183733717,
      "loss": 1.6225,
      "step": 1980
    },
    {
      "epoch": 0.20991561181434598,
      "grad_norm": 0.6605643033981323,
      "learning_rate": 0.0013540955028706425,
      "loss": 1.6175,
      "step": 1990
    },
    {
      "epoch": 0.2109704641350211,
      "grad_norm": 1.1141338348388672,
      "learning_rate": 0.0013526042181828324,
      "loss": 1.6187,
      "step": 2000
    },
    {
      "epoch": 0.2120253164556962,
      "grad_norm": 0.654744029045105,
      "learning_rate": 0.0013511061810205143,
      "loss": 1.6255,
      "step": 2010
    },
    {
      "epoch": 0.21308016877637131,
      "grad_norm": 0.7747483849525452,
      "learning_rate": 0.001349601408169926,
      "loss": 1.6329,
      "step": 2020
    },
    {
      "epoch": 0.2141350210970464,
      "grad_norm": 0.8426884412765503,
      "learning_rate": 0.0013480899164927823,
      "loss": 1.6217,
      "step": 2030
    },
    {
      "epoch": 0.21518987341772153,
      "grad_norm": 0.8079628348350525,
      "learning_rate": 0.0013465717229260853,
      "loss": 1.6215,
      "step": 2040
    },
    {
      "epoch": 0.21624472573839662,
      "grad_norm": 0.70868980884552,
      "learning_rate": 0.001345046844481935,
      "loss": 1.6181,
      "step": 2050
    },
    {
      "epoch": 0.21729957805907174,
      "grad_norm": 1.0628615617752075,
      "learning_rate": 0.0013435152982473396,
      "loss": 1.609,
      "step": 2060
    },
    {
      "epoch": 0.21835443037974683,
      "grad_norm": 1.1153923273086548,
      "learning_rate": 0.0013419771013840217,
      "loss": 1.612,
      "step": 2070
    },
    {
      "epoch": 0.21940928270042195,
      "grad_norm": 1.4523441791534424,
      "learning_rate": 0.001340432271128229,
      "loss": 1.6253,
      "step": 2080
    },
    {
      "epoch": 0.22046413502109705,
      "grad_norm": 0.673022449016571,
      "learning_rate": 0.0013388808247905381,
      "loss": 1.6009,
      "step": 2090
    },
    {
      "epoch": 0.22151898734177214,
      "grad_norm": 0.8973956108093262,
      "learning_rate": 0.0013373227797556634,
      "loss": 1.6101,
      "step": 2100
    },
    {
      "epoch": 0.22257383966244726,
      "grad_norm": 0.8235937356948853,
      "learning_rate": 0.00133575815348226,
      "loss": 1.6025,
      "step": 2110
    },
    {
      "epoch": 0.22362869198312235,
      "grad_norm": 0.7105578780174255,
      "learning_rate": 0.0013341869635027292,
      "loss": 1.5997,
      "step": 2120
    },
    {
      "epoch": 0.22468354430379747,
      "grad_norm": 0.9764186143875122,
      "learning_rate": 0.001332609227423022,
      "loss": 1.6051,
      "step": 2130
    },
    {
      "epoch": 0.22573839662447256,
      "grad_norm": 0.6583696007728577,
      "learning_rate": 0.0013310249629224417,
      "loss": 1.598,
      "step": 2140
    },
    {
      "epoch": 0.22679324894514769,
      "grad_norm": 0.9083722233772278,
      "learning_rate": 0.0013294341877534454,
      "loss": 1.6055,
      "step": 2150
    },
    {
      "epoch": 0.22784810126582278,
      "grad_norm": 1.9433847665786743,
      "learning_rate": 0.0013278369197414458,
      "loss": 1.6248,
      "step": 2160
    },
    {
      "epoch": 0.2289029535864979,
      "grad_norm": 1.0756428241729736,
      "learning_rate": 0.0013262331767846104,
      "loss": 1.6139,
      "step": 2170
    },
    {
      "epoch": 0.229957805907173,
      "grad_norm": 0.8095933794975281,
      "learning_rate": 0.0013246229768536628,
      "loss": 1.5827,
      "step": 2180
    },
    {
      "epoch": 0.2310126582278481,
      "grad_norm": 0.7876030802726746,
      "learning_rate": 0.001323006337991679,
      "loss": 1.6071,
      "step": 2190
    },
    {
      "epoch": 0.2320675105485232,
      "grad_norm": 0.7395517230033875,
      "learning_rate": 0.0013213832783138873,
      "loss": 1.5944,
      "step": 2200
    },
    {
      "epoch": 0.23312236286919832,
      "grad_norm": 0.8268523812294006,
      "learning_rate": 0.0013197538160074633,
      "loss": 1.5838,
      "step": 2210
    },
    {
      "epoch": 0.23417721518987342,
      "grad_norm": 0.762881875038147,
      "learning_rate": 0.0013181179693313283,
      "loss": 1.601,
      "step": 2220
    },
    {
      "epoch": 0.23523206751054854,
      "grad_norm": 1.0932894945144653,
      "learning_rate": 0.0013164757566159428,
      "loss": 1.5951,
      "step": 2230
    },
    {
      "epoch": 0.23628691983122363,
      "grad_norm": 0.775640606880188,
      "learning_rate": 0.001314827196263102,
      "loss": 1.5927,
      "step": 2240
    },
    {
      "epoch": 0.23734177215189872,
      "grad_norm": 1.1541540622711182,
      "learning_rate": 0.0013131723067457302,
      "loss": 1.5914,
      "step": 2250
    },
    {
      "epoch": 0.23839662447257384,
      "grad_norm": 0.923848569393158,
      "learning_rate": 0.0013115111066076721,
      "loss": 1.5814,
      "step": 2260
    },
    {
      "epoch": 0.23945147679324894,
      "grad_norm": 1.0156478881835938,
      "learning_rate": 0.0013098436144634862,
      "loss": 1.612,
      "step": 2270
    },
    {
      "epoch": 0.24050632911392406,
      "grad_norm": 0.8369004726409912,
      "learning_rate": 0.0013081698489982364,
      "loss": 1.5998,
      "step": 2280
    },
    {
      "epoch": 0.24156118143459915,
      "grad_norm": 0.6826422214508057,
      "learning_rate": 0.001306489828967282,
      "loss": 1.5826,
      "step": 2290
    },
    {
      "epoch": 0.24261603375527427,
      "grad_norm": 0.924490213394165,
      "learning_rate": 0.0013048035731960679,
      "loss": 1.5806,
      "step": 2300
    },
    {
      "epoch": 0.24367088607594936,
      "grad_norm": 0.7319629788398743,
      "learning_rate": 0.0013031111005799133,
      "loss": 1.586,
      "step": 2310
    },
    {
      "epoch": 0.24472573839662448,
      "grad_norm": 0.6873080134391785,
      "learning_rate": 0.0013014124300838004,
      "loss": 1.596,
      "step": 2320
    },
    {
      "epoch": 0.24578059071729957,
      "grad_norm": 0.8748582005500793,
      "learning_rate": 0.0012997075807421612,
      "loss": 1.5774,
      "step": 2330
    },
    {
      "epoch": 0.2468354430379747,
      "grad_norm": 0.9027491211891174,
      "learning_rate": 0.0012979965716586653,
      "loss": 1.5717,
      "step": 2340
    },
    {
      "epoch": 0.2478902953586498,
      "grad_norm": 0.7842824459075928,
      "learning_rate": 0.0012962794220060048,
      "loss": 1.5804,
      "step": 2350
    },
    {
      "epoch": 0.2489451476793249,
      "grad_norm": 1.0558085441589355,
      "learning_rate": 0.0012945561510256801,
      "loss": 1.5817,
      "step": 2360
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.5054733753204346,
      "learning_rate": 0.001292826778027784,
      "loss": 1.5809,
      "step": 2370
    },
    {
      "epoch": 0.2510548523206751,
      "grad_norm": 0.9746042490005493,
      "learning_rate": 0.0012910913223907856,
      "loss": 1.5805,
      "step": 2380
    },
    {
      "epoch": 0.2521097046413502,
      "grad_norm": 1.0360198020935059,
      "learning_rate": 0.0012893498035613123,
      "loss": 1.5692,
      "step": 2390
    },
    {
      "epoch": 0.25316455696202533,
      "grad_norm": 0.8236413598060608,
      "learning_rate": 0.001287602241053933,
      "loss": 1.5699,
      "step": 2400
    },
    {
      "epoch": 0.2542194092827004,
      "grad_norm": 0.68341064453125,
      "learning_rate": 0.0012858486544509392,
      "loss": 1.5771,
      "step": 2410
    },
    {
      "epoch": 0.2552742616033755,
      "grad_norm": 0.8383939862251282,
      "learning_rate": 0.0012840890634021249,
      "loss": 1.573,
      "step": 2420
    },
    {
      "epoch": 0.2563291139240506,
      "grad_norm": 0.9827631711959839,
      "learning_rate": 0.0012823234876245667,
      "loss": 1.5759,
      "step": 2430
    },
    {
      "epoch": 0.25738396624472576,
      "grad_norm": 0.8817957639694214,
      "learning_rate": 0.0012805519469024035,
      "loss": 1.5824,
      "step": 2440
    },
    {
      "epoch": 0.25843881856540085,
      "grad_norm": 0.8089924454689026,
      "learning_rate": 0.0012787744610866143,
      "loss": 1.5595,
      "step": 2450
    },
    {
      "epoch": 0.25949367088607594,
      "grad_norm": 0.7037054896354675,
      "learning_rate": 0.0012769910500947954,
      "loss": 1.5769,
      "step": 2460
    },
    {
      "epoch": 0.26054852320675104,
      "grad_norm": 1.0453506708145142,
      "learning_rate": 0.0012752017339109376,
      "loss": 1.5702,
      "step": 2470
    },
    {
      "epoch": 0.2616033755274262,
      "grad_norm": 0.8949272036552429,
      "learning_rate": 0.0012734065325852029,
      "loss": 1.5754,
      "step": 2480
    },
    {
      "epoch": 0.2626582278481013,
      "grad_norm": 0.9170915484428406,
      "learning_rate": 0.0012716054662336987,
      "loss": 1.564,
      "step": 2490
    },
    {
      "epoch": 0.26371308016877637,
      "grad_norm": 0.8970065116882324,
      "learning_rate": 0.001269798555038252,
      "loss": 1.558,
      "step": 2500
    },
    {
      "epoch": 0.26476793248945146,
      "grad_norm": 0.8219038844108582,
      "learning_rate": 0.0012679858192461864,
      "loss": 1.564,
      "step": 2510
    },
    {
      "epoch": 0.26582278481012656,
      "grad_norm": 0.6997198462486267,
      "learning_rate": 0.0012661672791700906,
      "loss": 1.5662,
      "step": 2520
    },
    {
      "epoch": 0.2668776371308017,
      "grad_norm": 0.8577184677124023,
      "learning_rate": 0.0012643429551875945,
      "loss": 1.5627,
      "step": 2530
    },
    {
      "epoch": 0.2679324894514768,
      "grad_norm": 1.7808401584625244,
      "learning_rate": 0.0012625128677411388,
      "loss": 1.5664,
      "step": 2540
    },
    {
      "epoch": 0.2689873417721519,
      "grad_norm": 0.9855268597602844,
      "learning_rate": 0.0012606770373377475,
      "loss": 1.5681,
      "step": 2550
    },
    {
      "epoch": 0.270042194092827,
      "grad_norm": 0.8617801070213318,
      "learning_rate": 0.0012588354845487959,
      "loss": 1.5625,
      "step": 2560
    },
    {
      "epoch": 0.27109704641350213,
      "grad_norm": 0.7381689548492432,
      "learning_rate": 0.001256988230009783,
      "loss": 1.553,
      "step": 2570
    },
    {
      "epoch": 0.2721518987341772,
      "grad_norm": 0.7527607679367065,
      "learning_rate": 0.0012551352944200976,
      "loss": 1.5575,
      "step": 2580
    },
    {
      "epoch": 0.2732067510548523,
      "grad_norm": 0.6910273432731628,
      "learning_rate": 0.0012532766985427874,
      "loss": 1.565,
      "step": 2590
    },
    {
      "epoch": 0.2742616033755274,
      "grad_norm": 0.7376472353935242,
      "learning_rate": 0.0012514124632043272,
      "loss": 1.5655,
      "step": 2600
    },
    {
      "epoch": 0.27531645569620256,
      "grad_norm": 0.7212229371070862,
      "learning_rate": 0.0012495426092943842,
      "loss": 1.562,
      "step": 2610
    },
    {
      "epoch": 0.27637130801687765,
      "grad_norm": 0.8666029572486877,
      "learning_rate": 0.0012476671577655845,
      "loss": 1.5607,
      "step": 2620
    },
    {
      "epoch": 0.27742616033755274,
      "grad_norm": 0.9047456979751587,
      "learning_rate": 0.0012457861296332774,
      "loss": 1.5511,
      "step": 2630
    },
    {
      "epoch": 0.27848101265822783,
      "grad_norm": 0.7443682551383972,
      "learning_rate": 0.001243899545975303,
      "loss": 1.5621,
      "step": 2640
    },
    {
      "epoch": 0.2795358649789029,
      "grad_norm": 0.8028952479362488,
      "learning_rate": 0.0012420074279317515,
      "loss": 1.5524,
      "step": 2650
    },
    {
      "epoch": 0.2805907172995781,
      "grad_norm": 0.9384350180625916,
      "learning_rate": 0.0012401097967047298,
      "loss": 1.5476,
      "step": 2660
    },
    {
      "epoch": 0.28164556962025317,
      "grad_norm": 0.6824517250061035,
      "learning_rate": 0.001238206673558122,
      "loss": 1.5498,
      "step": 2670
    },
    {
      "epoch": 0.28270042194092826,
      "grad_norm": 0.6348094344139099,
      "learning_rate": 0.0012362980798173526,
      "loss": 1.5482,
      "step": 2680
    },
    {
      "epoch": 0.28375527426160335,
      "grad_norm": 0.7970465421676636,
      "learning_rate": 0.0012343840368691462,
      "loss": 1.5505,
      "step": 2690
    },
    {
      "epoch": 0.2848101265822785,
      "grad_norm": 0.6667698621749878,
      "learning_rate": 0.0012324645661612886,
      "loss": 1.5453,
      "step": 2700
    },
    {
      "epoch": 0.2858649789029536,
      "grad_norm": 0.7847800254821777,
      "learning_rate": 0.0012305396892023867,
      "loss": 1.5516,
      "step": 2710
    },
    {
      "epoch": 0.2869198312236287,
      "grad_norm": 0.6802813410758972,
      "learning_rate": 0.0012286094275616264,
      "loss": 1.5545,
      "step": 2720
    },
    {
      "epoch": 0.2879746835443038,
      "grad_norm": 0.7703807950019836,
      "learning_rate": 0.0012266738028685318,
      "loss": 1.5326,
      "step": 2730
    },
    {
      "epoch": 0.2890295358649789,
      "grad_norm": 0.6709527969360352,
      "learning_rate": 0.001224732836812723,
      "loss": 1.5442,
      "step": 2740
    },
    {
      "epoch": 0.290084388185654,
      "grad_norm": 0.696071445941925,
      "learning_rate": 0.0012227865511436724,
      "loss": 1.5486,
      "step": 2750
    },
    {
      "epoch": 0.2911392405063291,
      "grad_norm": 0.8581094145774841,
      "learning_rate": 0.001220834967670461,
      "loss": 1.5557,
      "step": 2760
    },
    {
      "epoch": 0.2921940928270042,
      "grad_norm": 0.6772289872169495,
      "learning_rate": 0.0012188781082615346,
      "loss": 1.5507,
      "step": 2770
    },
    {
      "epoch": 0.29324894514767935,
      "grad_norm": 0.6636837124824524,
      "learning_rate": 0.0012169159948444588,
      "loss": 1.5481,
      "step": 2780
    },
    {
      "epoch": 0.29430379746835444,
      "grad_norm": 0.6257638931274414,
      "learning_rate": 0.001214948649405672,
      "loss": 1.545,
      "step": 2790
    },
    {
      "epoch": 0.29535864978902954,
      "grad_norm": 0.6627100706100464,
      "learning_rate": 0.0012129760939902407,
      "loss": 1.5439,
      "step": 2800
    },
    {
      "epoch": 0.29641350210970463,
      "grad_norm": 0.7219129204750061,
      "learning_rate": 0.0012109983507016114,
      "loss": 1.5453,
      "step": 2810
    },
    {
      "epoch": 0.2974683544303797,
      "grad_norm": 0.7363254427909851,
      "learning_rate": 0.0012090154417013636,
      "loss": 1.5503,
      "step": 2820
    },
    {
      "epoch": 0.29852320675105487,
      "grad_norm": 0.719441294670105,
      "learning_rate": 0.0012070273892089605,
      "loss": 1.5257,
      "step": 2830
    },
    {
      "epoch": 0.29957805907172996,
      "grad_norm": 0.745360791683197,
      "learning_rate": 0.0012050342155015012,
      "loss": 1.5273,
      "step": 2840
    },
    {
      "epoch": 0.30063291139240506,
      "grad_norm": 0.78196781873703,
      "learning_rate": 0.0012030359429134707,
      "loss": 1.5305,
      "step": 2850
    },
    {
      "epoch": 0.30168776371308015,
      "grad_norm": 0.7005848288536072,
      "learning_rate": 0.0012010325938364883,
      "loss": 1.5393,
      "step": 2860
    },
    {
      "epoch": 0.3027426160337553,
      "grad_norm": 1.3260812759399414,
      "learning_rate": 0.0011990241907190592,
      "loss": 1.5386,
      "step": 2870
    },
    {
      "epoch": 0.3037974683544304,
      "grad_norm": 0.7265012264251709,
      "learning_rate": 0.001197010756066321,
      "loss": 1.5249,
      "step": 2880
    },
    {
      "epoch": 0.3048523206751055,
      "grad_norm": 0.7650523781776428,
      "learning_rate": 0.0011949923124397917,
      "loss": 1.5272,
      "step": 2890
    },
    {
      "epoch": 0.3059071729957806,
      "grad_norm": 0.6694566607475281,
      "learning_rate": 0.001192968882457118,
      "loss": 1.5309,
      "step": 2900
    },
    {
      "epoch": 0.3069620253164557,
      "grad_norm": 0.67763751745224,
      "learning_rate": 0.001190940488791821,
      "loss": 1.5355,
      "step": 2910
    },
    {
      "epoch": 0.3080168776371308,
      "grad_norm": 1.0468145608901978,
      "learning_rate": 0.0011889071541730419,
      "loss": 1.5286,
      "step": 2920
    },
    {
      "epoch": 0.3090717299578059,
      "grad_norm": 0.7940829396247864,
      "learning_rate": 0.001186868901385288,
      "loss": 1.524,
      "step": 2930
    },
    {
      "epoch": 0.310126582278481,
      "grad_norm": 0.6903820037841797,
      "learning_rate": 0.001184825753268177,
      "loss": 1.5272,
      "step": 2940
    },
    {
      "epoch": 0.3111814345991561,
      "grad_norm": 0.6803721189498901,
      "learning_rate": 0.0011827777327161814,
      "loss": 1.54,
      "step": 2950
    },
    {
      "epoch": 0.31223628691983124,
      "grad_norm": 1.6378977298736572,
      "learning_rate": 0.0011807248626783714,
      "loss": 1.5238,
      "step": 2960
    },
    {
      "epoch": 0.31329113924050633,
      "grad_norm": 1.0977939367294312,
      "learning_rate": 0.0011786671661581584,
      "loss": 1.5177,
      "step": 2970
    },
    {
      "epoch": 0.3143459915611814,
      "grad_norm": 0.9927310347557068,
      "learning_rate": 0.001176604666213036,
      "loss": 1.5128,
      "step": 2980
    },
    {
      "epoch": 0.3154008438818565,
      "grad_norm": 0.6859022378921509,
      "learning_rate": 0.0011745373859543236,
      "loss": 1.5323,
      "step": 2990
    },
    {
      "epoch": 0.31645569620253167,
      "grad_norm": 0.8309093713760376,
      "learning_rate": 0.0011724653485469063,
      "loss": 1.5122,
      "step": 3000
    },
    {
      "epoch": 0.31751054852320676,
      "grad_norm": 0.6543201208114624,
      "learning_rate": 0.0011703885772089743,
      "loss": 1.5226,
      "step": 3010
    },
    {
      "epoch": 0.31856540084388185,
      "grad_norm": 0.8993769288063049,
      "learning_rate": 0.0011683070952117646,
      "loss": 1.5217,
      "step": 3020
    },
    {
      "epoch": 0.31962025316455694,
      "grad_norm": 0.6465635299682617,
      "learning_rate": 0.0011662209258792998,
      "loss": 1.5143,
      "step": 3030
    },
    {
      "epoch": 0.3206751054852321,
      "grad_norm": 0.8522810339927673,
      "learning_rate": 0.0011641300925881257,
      "loss": 1.5204,
      "step": 3040
    },
    {
      "epoch": 0.3217299578059072,
      "grad_norm": 0.7852684855461121,
      "learning_rate": 0.0011620346187670501,
      "loss": 1.5136,
      "step": 3050
    },
    {
      "epoch": 0.3227848101265823,
      "grad_norm": 0.6950536966323853,
      "learning_rate": 0.0011599345278968806,
      "loss": 1.5259,
      "step": 3060
    },
    {
      "epoch": 0.32383966244725737,
      "grad_norm": 0.8415475487709045,
      "learning_rate": 0.0011578298435101604,
      "loss": 1.5176,
      "step": 3070
    },
    {
      "epoch": 0.32489451476793246,
      "grad_norm": 0.8313452005386353,
      "learning_rate": 0.0011557205891909062,
      "loss": 1.5081,
      "step": 3080
    },
    {
      "epoch": 0.3259493670886076,
      "grad_norm": 0.613431453704834,
      "learning_rate": 0.0011536067885743423,
      "loss": 1.5224,
      "step": 3090
    },
    {
      "epoch": 0.3270042194092827,
      "grad_norm": 0.8133174180984497,
      "learning_rate": 0.001151488465346637,
      "loss": 1.5094,
      "step": 3100
    },
    {
      "epoch": 0.3280590717299578,
      "grad_norm": 0.7211087346076965,
      "learning_rate": 0.0011493656432446362,
      "loss": 1.5166,
      "step": 3110
    },
    {
      "epoch": 0.3291139240506329,
      "grad_norm": 0.682667076587677,
      "learning_rate": 0.0011472383460555983,
      "loss": 1.5166,
      "step": 3120
    },
    {
      "epoch": 0.33016877637130804,
      "grad_norm": 0.5974841117858887,
      "learning_rate": 0.001145106597616927,
      "loss": 1.5292,
      "step": 3130
    },
    {
      "epoch": 0.33122362869198313,
      "grad_norm": 0.6949234008789062,
      "learning_rate": 0.001142970421815904,
      "loss": 1.5038,
      "step": 3140
    },
    {
      "epoch": 0.3322784810126582,
      "grad_norm": 0.6972474455833435,
      "learning_rate": 0.0011408298425894226,
      "loss": 1.5118,
      "step": 3150
    },
    {
      "epoch": 0.3333333333333333,
      "grad_norm": 0.7973185181617737,
      "learning_rate": 0.0011386848839237186,
      "loss": 1.514,
      "step": 3160
    },
    {
      "epoch": 0.33438818565400846,
      "grad_norm": 0.7194811105728149,
      "learning_rate": 0.0011365355698541005,
      "loss": 1.5149,
      "step": 3170
    },
    {
      "epoch": 0.33544303797468356,
      "grad_norm": 0.6512601375579834,
      "learning_rate": 0.0011343819244646824,
      "loss": 1.5091,
      "step": 3180
    },
    {
      "epoch": 0.33649789029535865,
      "grad_norm": 0.6934634447097778,
      "learning_rate": 0.001132223971888112,
      "loss": 1.5124,
      "step": 3190
    },
    {
      "epoch": 0.33755274261603374,
      "grad_norm": 0.709214448928833,
      "learning_rate": 0.0011300617363053024,
      "loss": 1.5057,
      "step": 3200
    },
    {
      "epoch": 0.33860759493670883,
      "grad_norm": 0.8465114831924438,
      "learning_rate": 0.0011278952419451586,
      "loss": 1.5166,
      "step": 3210
    },
    {
      "epoch": 0.339662447257384,
      "grad_norm": 0.6625987887382507,
      "learning_rate": 0.0011257245130843077,
      "loss": 1.5097,
      "step": 3220
    },
    {
      "epoch": 0.3407172995780591,
      "grad_norm": 0.6848903298377991,
      "learning_rate": 0.0011235495740468265,
      "loss": 1.5039,
      "step": 3230
    },
    {
      "epoch": 0.34177215189873417,
      "grad_norm": 0.7153477668762207,
      "learning_rate": 0.0011213704492039694,
      "loss": 1.4827,
      "step": 3240
    },
    {
      "epoch": 0.34282700421940926,
      "grad_norm": 0.7239137291908264,
      "learning_rate": 0.001119187162973894,
      "loss": 1.5116,
      "step": 3250
    },
    {
      "epoch": 0.3438818565400844,
      "grad_norm": 0.6427382826805115,
      "learning_rate": 0.001116999739821388,
      "loss": 1.4981,
      "step": 3260
    },
    {
      "epoch": 0.3449367088607595,
      "grad_norm": 0.7354733943939209,
      "learning_rate": 0.0011148082042575968,
      "loss": 1.5166,
      "step": 3270
    },
    {
      "epoch": 0.3459915611814346,
      "grad_norm": 0.8644187450408936,
      "learning_rate": 0.0011126125808397461,
      "loss": 1.5141,
      "step": 3280
    },
    {
      "epoch": 0.3470464135021097,
      "grad_norm": 1.0546091794967651,
      "learning_rate": 0.0011104128941708683,
      "loss": 1.4937,
      "step": 3290
    },
    {
      "epoch": 0.34810126582278483,
      "grad_norm": 0.654690146446228,
      "learning_rate": 0.001108209168899527,
      "loss": 1.5011,
      "step": 3300
    },
    {
      "epoch": 0.3491561181434599,
      "grad_norm": 0.6327159404754639,
      "learning_rate": 0.0011060014297195396,
      "loss": 1.5042,
      "step": 3310
    },
    {
      "epoch": 0.350210970464135,
      "grad_norm": 0.6495179533958435,
      "learning_rate": 0.0011037897013697015,
      "loss": 1.5084,
      "step": 3320
    },
    {
      "epoch": 0.3512658227848101,
      "grad_norm": 0.7837378978729248,
      "learning_rate": 0.0011015740086335092,
      "loss": 1.4999,
      "step": 3330
    },
    {
      "epoch": 0.35232067510548526,
      "grad_norm": 0.7139352560043335,
      "learning_rate": 0.0010993543763388814,
      "loss": 1.5021,
      "step": 3340
    },
    {
      "epoch": 0.35337552742616035,
      "grad_norm": 0.6452745795249939,
      "learning_rate": 0.0010971308293578814,
      "loss": 1.4912,
      "step": 3350
    },
    {
      "epoch": 0.35443037974683544,
      "grad_norm": 0.677642285823822,
      "learning_rate": 0.0010949033926064397,
      "loss": 1.4913,
      "step": 3360
    },
    {
      "epoch": 0.35548523206751054,
      "grad_norm": 0.9014589786529541,
      "learning_rate": 0.0010926720910440725,
      "loss": 1.4995,
      "step": 3370
    },
    {
      "epoch": 0.35654008438818563,
      "grad_norm": 0.7984604239463806,
      "learning_rate": 0.001090436949673603,
      "loss": 1.4969,
      "step": 3380
    },
    {
      "epoch": 0.3575949367088608,
      "grad_norm": 0.6424825191497803,
      "learning_rate": 0.0010881979935408815,
      "loss": 1.4907,
      "step": 3390
    },
    {
      "epoch": 0.35864978902953587,
      "grad_norm": 0.74113929271698,
      "learning_rate": 0.0010859552477345052,
      "loss": 1.498,
      "step": 3400
    },
    {
      "epoch": 0.35970464135021096,
      "grad_norm": 0.6416418552398682,
      "learning_rate": 0.001083708737385536,
      "loss": 1.4946,
      "step": 3410
    },
    {
      "epoch": 0.36075949367088606,
      "grad_norm": 1.0173593759536743,
      "learning_rate": 0.0010814584876672187,
      "loss": 1.4752,
      "step": 3420
    },
    {
      "epoch": 0.3618143459915612,
      "grad_norm": 0.6335579752922058,
      "learning_rate": 0.0010792045237947008,
      "loss": 1.4919,
      "step": 3430
    },
    {
      "epoch": 0.3628691983122363,
      "grad_norm": 0.7825088500976562,
      "learning_rate": 0.0010769468710247478,
      "loss": 1.4876,
      "step": 3440
    },
    {
      "epoch": 0.3639240506329114,
      "grad_norm": 0.6434103846549988,
      "learning_rate": 0.0010746855546554612,
      "loss": 1.4863,
      "step": 3450
    },
    {
      "epoch": 0.3649789029535865,
      "grad_norm": 0.6860970854759216,
      "learning_rate": 0.0010724206000259954,
      "loss": 1.4779,
      "step": 3460
    },
    {
      "epoch": 0.36603375527426163,
      "grad_norm": 0.8527423739433289,
      "learning_rate": 0.0010701520325162727,
      "loss": 1.4952,
      "step": 3470
    },
    {
      "epoch": 0.3670886075949367,
      "grad_norm": 0.7300753593444824,
      "learning_rate": 0.0010678798775467001,
      "loss": 1.4993,
      "step": 3480
    },
    {
      "epoch": 0.3681434599156118,
      "grad_norm": 0.7711049318313599,
      "learning_rate": 0.0010656041605778832,
      "loss": 1.4929,
      "step": 3490
    },
    {
      "epoch": 0.3691983122362869,
      "grad_norm": 0.6592317223548889,
      "learning_rate": 0.001063324907110342,
      "loss": 1.4744,
      "step": 3500
    },
    {
      "epoch": 0.370253164556962,
      "grad_norm": 0.6633989214897156,
      "learning_rate": 0.0010610421426842241,
      "loss": 1.4945,
      "step": 3510
    },
    {
      "epoch": 0.37130801687763715,
      "grad_norm": 0.7151485681533813,
      "learning_rate": 0.00105875589287902,
      "loss": 1.4913,
      "step": 3520
    },
    {
      "epoch": 0.37236286919831224,
      "grad_norm": 0.7675790190696716,
      "learning_rate": 0.0010564661833132752,
      "loss": 1.4998,
      "step": 3530
    },
    {
      "epoch": 0.37341772151898733,
      "grad_norm": 1.261635184288025,
      "learning_rate": 0.001054173039644303,
      "loss": 1.4957,
      "step": 3540
    },
    {
      "epoch": 0.3744725738396624,
      "grad_norm": 0.6637124419212341,
      "learning_rate": 0.0010518764875678981,
      "loss": 1.4842,
      "step": 3550
    },
    {
      "epoch": 0.3755274261603376,
      "grad_norm": 0.7653613686561584,
      "learning_rate": 0.001049576552818048,
      "loss": 1.4823,
      "step": 3560
    },
    {
      "epoch": 0.37658227848101267,
      "grad_norm": 1.5657800436019897,
      "learning_rate": 0.0010472732611666448,
      "loss": 1.4773,
      "step": 3570
    },
    {
      "epoch": 0.37763713080168776,
      "grad_norm": 1.3445134162902832,
      "learning_rate": 0.0010449666384231954,
      "loss": 1.4834,
      "step": 3580
    },
    {
      "epoch": 0.37869198312236285,
      "grad_norm": 0.716289758682251,
      "learning_rate": 0.0010426567104345346,
      "loss": 1.4885,
      "step": 3590
    },
    {
      "epoch": 0.379746835443038,
      "grad_norm": 0.7706812620162964,
      "learning_rate": 0.0010403435030845332,
      "loss": 1.4737,
      "step": 3600
    },
    {
      "epoch": 0.3808016877637131,
      "grad_norm": 1.0079050064086914,
      "learning_rate": 0.0010380270422938093,
      "loss": 1.4758,
      "step": 3610
    },
    {
      "epoch": 0.3818565400843882,
      "grad_norm": 0.6356129050254822,
      "learning_rate": 0.0010357073540194362,
      "loss": 1.4772,
      "step": 3620
    },
    {
      "epoch": 0.3829113924050633,
      "grad_norm": 1.1435695886611938,
      "learning_rate": 0.001033384464254655,
      "loss": 1.4769,
      "step": 3630
    },
    {
      "epoch": 0.38396624472573837,
      "grad_norm": 1.018843650817871,
      "learning_rate": 0.001031058399028579,
      "loss": 1.4768,
      "step": 3640
    },
    {
      "epoch": 0.3850210970464135,
      "grad_norm": 0.6845547556877136,
      "learning_rate": 0.001028729184405905,
      "loss": 1.4873,
      "step": 3650
    },
    {
      "epoch": 0.3860759493670886,
      "grad_norm": 0.7335551977157593,
      "learning_rate": 0.0010263968464866201,
      "loss": 1.4784,
      "step": 3660
    },
    {
      "epoch": 0.3871308016877637,
      "grad_norm": 0.7382904887199402,
      "learning_rate": 0.0010240614114057098,
      "loss": 1.4723,
      "step": 3670
    },
    {
      "epoch": 0.3881856540084388,
      "grad_norm": 0.9586758017539978,
      "learning_rate": 0.001021722905332864,
      "loss": 1.4744,
      "step": 3680
    },
    {
      "epoch": 0.38924050632911394,
      "grad_norm": 0.6617874503135681,
      "learning_rate": 0.0010193813544721855,
      "loss": 1.4862,
      "step": 3690
    },
    {
      "epoch": 0.39029535864978904,
      "grad_norm": 0.8000274300575256,
      "learning_rate": 0.001017036785061895,
      "loss": 1.4847,
      "step": 3700
    },
    {
      "epoch": 0.39135021097046413,
      "grad_norm": 0.6757653951644897,
      "learning_rate": 0.0010146892233740376,
      "loss": 1.4724,
      "step": 3710
    },
    {
      "epoch": 0.3924050632911392,
      "grad_norm": 0.7536963820457458,
      "learning_rate": 0.0010123386957141883,
      "loss": 1.4555,
      "step": 3720
    },
    {
      "epoch": 0.39345991561181437,
      "grad_norm": 0.6162813901901245,
      "learning_rate": 0.0010099852284211573,
      "loss": 1.4724,
      "step": 3730
    },
    {
      "epoch": 0.39451476793248946,
      "grad_norm": 0.723198413848877,
      "learning_rate": 0.0010076288478666944,
      "loss": 1.4736,
      "step": 3740
    },
    {
      "epoch": 0.39556962025316456,
      "grad_norm": 0.8122689723968506,
      "learning_rate": 0.0010052695804551946,
      "loss": 1.479,
      "step": 3750
    },
    {
      "epoch": 0.39662447257383965,
      "grad_norm": 1.0385117530822754,
      "learning_rate": 0.0010029074526234014,
      "loss": 1.4691,
      "step": 3760
    },
    {
      "epoch": 0.39767932489451474,
      "grad_norm": 0.6956025958061218,
      "learning_rate": 0.0010005424908401104,
      "loss": 1.4829,
      "step": 3770
    },
    {
      "epoch": 0.3987341772151899,
      "grad_norm": 0.6772257685661316,
      "learning_rate": 0.0009981747216058728,
      "loss": 1.4693,
      "step": 3780
    },
    {
      "epoch": 0.399789029535865,
      "grad_norm": 1.2388560771942139,
      "learning_rate": 0.0009958041714526998,
      "loss": 1.473,
      "step": 3790
    },
    {
      "epoch": 0.4008438818565401,
      "grad_norm": 1.170211672782898,
      "learning_rate": 0.0009934308669437627,
      "loss": 1.4799,
      "step": 3800
    },
    {
      "epoch": 0.40189873417721517,
      "grad_norm": 0.7374141216278076,
      "learning_rate": 0.0009910548346730972,
      "loss": 1.4595,
      "step": 3810
    },
    {
      "epoch": 0.4029535864978903,
      "grad_norm": 0.7336722016334534,
      "learning_rate": 0.0009886761012653062,
      "loss": 1.4554,
      "step": 3820
    },
    {
      "epoch": 0.4040084388185654,
      "grad_norm": 0.6598965525627136,
      "learning_rate": 0.000986294693375258,
      "loss": 1.4577,
      "step": 3830
    },
    {
      "epoch": 0.4050632911392405,
      "grad_norm": 0.6743797659873962,
      "learning_rate": 0.000983910637687791,
      "loss": 1.4775,
      "step": 3840
    },
    {
      "epoch": 0.4061181434599156,
      "grad_norm": 0.849048912525177,
      "learning_rate": 0.0009815239609174138,
      "loss": 1.4615,
      "step": 3850
    },
    {
      "epoch": 0.40717299578059074,
      "grad_norm": 0.9981573820114136,
      "learning_rate": 0.0009791346898080043,
      "loss": 1.4725,
      "step": 3860
    },
    {
      "epoch": 0.40822784810126583,
      "grad_norm": 0.691103994846344,
      "learning_rate": 0.0009767428511325122,
      "loss": 1.4629,
      "step": 3870
    },
    {
      "epoch": 0.4092827004219409,
      "grad_norm": 0.7077181935310364,
      "learning_rate": 0.0009743484716926576,
      "loss": 1.4553,
      "step": 3880
    },
    {
      "epoch": 0.410337552742616,
      "grad_norm": 0.8576418161392212,
      "learning_rate": 0.0009719515783186319,
      "loss": 1.4533,
      "step": 3890
    },
    {
      "epoch": 0.41139240506329117,
      "grad_norm": 1.1091513633728027,
      "learning_rate": 0.0009695521978687951,
      "loss": 1.4538,
      "step": 3900
    },
    {
      "epoch": 0.41244725738396626,
      "grad_norm": 0.6998710632324219,
      "learning_rate": 0.0009671503572293767,
      "loss": 1.4591,
      "step": 3910
    },
    {
      "epoch": 0.41350210970464135,
      "grad_norm": 0.7016639113426208,
      "learning_rate": 0.0009647460833141742,
      "loss": 1.4521,
      "step": 3920
    },
    {
      "epoch": 0.41455696202531644,
      "grad_norm": 0.7067856788635254,
      "learning_rate": 0.0009623394030642507,
      "loss": 1.4588,
      "step": 3930
    },
    {
      "epoch": 0.41561181434599154,
      "grad_norm": 0.8337339758872986,
      "learning_rate": 0.0009599303434476334,
      "loss": 1.4576,
      "step": 3940
    },
    {
      "epoch": 0.4166666666666667,
      "grad_norm": 0.9599086046218872,
      "learning_rate": 0.0009575189314590118,
      "loss": 1.4629,
      "step": 3950
    },
    {
      "epoch": 0.4177215189873418,
      "grad_norm": 0.6289507150650024,
      "learning_rate": 0.0009551051941194346,
      "loss": 1.4618,
      "step": 3960
    },
    {
      "epoch": 0.41877637130801687,
      "grad_norm": 0.7182209491729736,
      "learning_rate": 0.0009526891584760071,
      "loss": 1.45,
      "step": 3970
    },
    {
      "epoch": 0.41983122362869196,
      "grad_norm": 0.6944434642791748,
      "learning_rate": 0.0009502708516015889,
      "loss": 1.4609,
      "step": 3980
    },
    {
      "epoch": 0.4208860759493671,
      "grad_norm": 0.6946083307266235,
      "learning_rate": 0.0009478503005944888,
      "loss": 1.4519,
      "step": 3990
    },
    {
      "epoch": 0.4219409282700422,
      "grad_norm": 0.7730834484100342,
      "learning_rate": 0.0009454275325781632,
      "loss": 1.4645,
      "step": 4000
    },
    {
      "epoch": 0.4229957805907173,
      "grad_norm": 1.5695055723190308,
      "learning_rate": 0.0009432451682581424,
      "loss": 1.4538,
      "step": 4010
    },
    {
      "epoch": 0.4240506329113924,
      "grad_norm": 1.139216423034668,
      "learning_rate": 0.000940818262738138,
      "loss": 1.4573,
      "step": 4020
    },
    {
      "epoch": 0.42510548523206754,
      "grad_norm": 0.9682896137237549,
      "learning_rate": 0.0009383892190063256,
      "loss": 1.4422,
      "step": 4030
    },
    {
      "epoch": 0.42616033755274263,
      "grad_norm": 0.7299572229385376,
      "learning_rate": 0.0009359580642813265,
      "loss": 1.4539,
      "step": 4040
    },
    {
      "epoch": 0.4272151898734177,
      "grad_norm": 0.6831508874893188,
      "learning_rate": 0.0009335248258054162,
      "loss": 1.4539,
      "step": 4050
    },
    {
      "epoch": 0.4282700421940928,
      "grad_norm": 0.9515681266784668,
      "learning_rate": 0.0009310895308442202,
      "loss": 1.4586,
      "step": 4060
    },
    {
      "epoch": 0.4293248945147679,
      "grad_norm": 1.240949034690857,
      "learning_rate": 0.0009286522066864078,
      "loss": 1.4574,
      "step": 4070
    },
    {
      "epoch": 0.43037974683544306,
      "grad_norm": 0.6996131539344788,
      "learning_rate": 0.0009262128806433858,
      "loss": 1.4604,
      "step": 4080
    },
    {
      "epoch": 0.43143459915611815,
      "grad_norm": 0.6759021282196045,
      "learning_rate": 0.0009237715800489942,
      "loss": 1.4508,
      "step": 4090
    },
    {
      "epoch": 0.43248945147679324,
      "grad_norm": 0.5993544459342957,
      "learning_rate": 0.0009213283322591977,
      "loss": 1.4518,
      "step": 4100
    },
    {
      "epoch": 0.43354430379746833,
      "grad_norm": 0.9326764345169067,
      "learning_rate": 0.000918883164651781,
      "loss": 1.445,
      "step": 4110
    },
    {
      "epoch": 0.4345991561181435,
      "grad_norm": 0.6843815445899963,
      "learning_rate": 0.0009164361046260412,
      "loss": 1.4526,
      "step": 4120
    },
    {
      "epoch": 0.4356540084388186,
      "grad_norm": 0.6950252056121826,
      "learning_rate": 0.0009139871796024807,
      "loss": 1.4402,
      "step": 4130
    },
    {
      "epoch": 0.43670886075949367,
      "grad_norm": 0.6593072414398193,
      "learning_rate": 0.0009115364170225,
      "loss": 1.4447,
      "step": 4140
    },
    {
      "epoch": 0.43776371308016876,
      "grad_norm": 0.7160829305648804,
      "learning_rate": 0.0009090838443480903,
      "loss": 1.4504,
      "step": 4150
    },
    {
      "epoch": 0.4388185654008439,
      "grad_norm": 0.7197774052619934,
      "learning_rate": 0.0009066294890615266,
      "loss": 1.444,
      "step": 4160
    },
    {
      "epoch": 0.439873417721519,
      "grad_norm": 0.7182207107543945,
      "learning_rate": 0.0009041733786650578,
      "loss": 1.4448,
      "step": 4170
    },
    {
      "epoch": 0.4409282700421941,
      "grad_norm": 0.6120582222938538,
      "learning_rate": 0.0009017155406806006,
      "loss": 1.4405,
      "step": 4180
    },
    {
      "epoch": 0.4419831223628692,
      "grad_norm": 0.9504909515380859,
      "learning_rate": 0.0008992560026494294,
      "loss": 1.4477,
      "step": 4190
    },
    {
      "epoch": 0.4430379746835443,
      "grad_norm": 0.89927738904953,
      "learning_rate": 0.0008967947921318689,
      "loss": 1.4499,
      "step": 4200
    },
    {
      "epoch": 0.4440928270042194,
      "grad_norm": 0.8201597332954407,
      "learning_rate": 0.0008943319367069844,
      "loss": 1.4393,
      "step": 4210
    },
    {
      "epoch": 0.4451476793248945,
      "grad_norm": 0.6967815160751343,
      "learning_rate": 0.0008918674639722742,
      "loss": 1.4446,
      "step": 4220
    },
    {
      "epoch": 0.4462025316455696,
      "grad_norm": 1.0743404626846313,
      "learning_rate": 0.0008894014015433582,
      "loss": 1.4429,
      "step": 4230
    },
    {
      "epoch": 0.4472573839662447,
      "grad_norm": 0.7688655853271484,
      "learning_rate": 0.0008869337770536699,
      "loss": 1.4389,
      "step": 4240
    },
    {
      "epoch": 0.44831223628691985,
      "grad_norm": 1.0831153392791748,
      "learning_rate": 0.0008844646181541472,
      "loss": 1.4447,
      "step": 4250
    },
    {
      "epoch": 0.44936708860759494,
      "grad_norm": 1.177405834197998,
      "learning_rate": 0.0008819939525129207,
      "loss": 1.4495,
      "step": 4260
    },
    {
      "epoch": 0.45042194092827004,
      "grad_norm": 0.6696770191192627,
      "learning_rate": 0.0008795218078150056,
      "loss": 1.4398,
      "step": 4270
    },
    {
      "epoch": 0.45147679324894513,
      "grad_norm": 0.6478503346443176,
      "learning_rate": 0.0008770482117619901,
      "loss": 1.4325,
      "step": 4280
    },
    {
      "epoch": 0.4525316455696203,
      "grad_norm": 0.6330786943435669,
      "learning_rate": 0.000874573192071726,
      "loss": 1.4464,
      "step": 4290
    },
    {
      "epoch": 0.45358649789029537,
      "grad_norm": 0.7457678318023682,
      "learning_rate": 0.0008720967764780173,
      "loss": 1.4452,
      "step": 4300
    },
    {
      "epoch": 0.45464135021097046,
      "grad_norm": 0.7768924832344055,
      "learning_rate": 0.0008696189927303101,
      "loss": 1.4425,
      "step": 4310
    },
    {
      "epoch": 0.45569620253164556,
      "grad_norm": 0.7091420888900757,
      "learning_rate": 0.0008671398685933811,
      "loss": 1.4335,
      "step": 4320
    },
    {
      "epoch": 0.45675105485232065,
      "grad_norm": 0.7445207238197327,
      "learning_rate": 0.0008646594318470268,
      "loss": 1.4433,
      "step": 4330
    },
    {
      "epoch": 0.4578059071729958,
      "grad_norm": 0.9108366966247559,
      "learning_rate": 0.000862177710285752,
      "loss": 1.4334,
      "step": 4340
    },
    {
      "epoch": 0.4588607594936709,
      "grad_norm": 0.6861720085144043,
      "learning_rate": 0.0008596947317184585,
      "loss": 1.4361,
      "step": 4350
    },
    {
      "epoch": 0.459915611814346,
      "grad_norm": 0.9899851679801941,
      "learning_rate": 0.0008572105239681338,
      "loss": 1.4511,
      "step": 4360
    },
    {
      "epoch": 0.4609704641350211,
      "grad_norm": 1.09987473487854,
      "learning_rate": 0.0008547251148715386,
      "loss": 1.4302,
      "step": 4370
    },
    {
      "epoch": 0.4620253164556962,
      "grad_norm": 0.7392022013664246,
      "learning_rate": 0.0008522385322788955,
      "loss": 1.4349,
      "step": 4380
    },
    {
      "epoch": 0.4630801687763713,
      "grad_norm": 0.6557256579399109,
      "learning_rate": 0.0008497508040535766,
      "loss": 1.4266,
      "step": 4390
    },
    {
      "epoch": 0.4641350210970464,
      "grad_norm": 0.7369914054870605,
      "learning_rate": 0.0008472619580717914,
      "loss": 1.4321,
      "step": 4400
    },
    {
      "epoch": 0.4651898734177215,
      "grad_norm": 0.6846641898155212,
      "learning_rate": 0.000844772022222274,
      "loss": 1.4468,
      "step": 4410
    },
    {
      "epoch": 0.46624472573839665,
      "grad_norm": 0.6874028444290161,
      "learning_rate": 0.0008422810244059721,
      "loss": 1.4302,
      "step": 4420
    },
    {
      "epoch": 0.46729957805907174,
      "grad_norm": 0.7561067938804626,
      "learning_rate": 0.000839788992535732,
      "loss": 1.4343,
      "step": 4430
    },
    {
      "epoch": 0.46835443037974683,
      "grad_norm": 0.7646554112434387,
      "learning_rate": 0.000837295954535988,
      "loss": 1.4283,
      "step": 4440
    },
    {
      "epoch": 0.4694092827004219,
      "grad_norm": 0.6366352438926697,
      "learning_rate": 0.0008348019383424479,
      "loss": 1.4405,
      "step": 4450
    },
    {
      "epoch": 0.4704641350210971,
      "grad_norm": 0.6604663729667664,
      "learning_rate": 0.0008323069719017812,
      "loss": 1.4242,
      "step": 4460
    },
    {
      "epoch": 0.47151898734177217,
      "grad_norm": 0.7483847737312317,
      "learning_rate": 0.0008298110831713047,
      "loss": 1.4365,
      "step": 4470
    },
    {
      "epoch": 0.47257383966244726,
      "grad_norm": 0.6634123921394348,
      "learning_rate": 0.0008273143001186709,
      "loss": 1.4404,
      "step": 4480
    },
    {
      "epoch": 0.47362869198312235,
      "grad_norm": 0.6505631804466248,
      "learning_rate": 0.0008248166507215526,
      "loss": 1.4286,
      "step": 4490
    },
    {
      "epoch": 0.47468354430379744,
      "grad_norm": 0.8349068760871887,
      "learning_rate": 0.0008223181629673312,
      "loss": 1.4277,
      "step": 4500
    },
    {
      "epoch": 0.4757383966244726,
      "grad_norm": 0.8086479902267456,
      "learning_rate": 0.0008198188648527818,
      "loss": 1.4327,
      "step": 4510
    },
    {
      "epoch": 0.4767932489451477,
      "grad_norm": 0.7039761543273926,
      "learning_rate": 0.00081731878438376,
      "loss": 1.4107,
      "step": 4520
    },
    {
      "epoch": 0.4778481012658228,
      "grad_norm": 1.3149323463439941,
      "learning_rate": 0.0008148179495748885,
      "loss": 1.4359,
      "step": 4530
    },
    {
      "epoch": 0.47890295358649787,
      "grad_norm": 0.6844233870506287,
      "learning_rate": 0.0008123163884492422,
      "loss": 1.4307,
      "step": 4540
    },
    {
      "epoch": 0.479957805907173,
      "grad_norm": 1.050879716873169,
      "learning_rate": 0.0008098141290380353,
      "loss": 1.4317,
      "step": 4550
    },
    {
      "epoch": 0.4810126582278481,
      "grad_norm": 0.8016988039016724,
      "learning_rate": 0.000807311199380306,
      "loss": 1.4202,
      "step": 4560
    },
    {
      "epoch": 0.4820675105485232,
      "grad_norm": 1.0670498609542847,
      "learning_rate": 0.0008048076275226032,
      "loss": 1.4028,
      "step": 4570
    },
    {
      "epoch": 0.4831223628691983,
      "grad_norm": 1.0285097360610962,
      "learning_rate": 0.0008023034415186725,
      "loss": 1.4285,
      "step": 4580
    },
    {
      "epoch": 0.48417721518987344,
      "grad_norm": 0.8433698415756226,
      "learning_rate": 0.0007997986694291404,
      "loss": 1.4113,
      "step": 4590
    },
    {
      "epoch": 0.48523206751054854,
      "grad_norm": 0.7636685967445374,
      "learning_rate": 0.0007972933393212012,
      "loss": 1.4283,
      "step": 4600
    },
    {
      "epoch": 0.48628691983122363,
      "grad_norm": 0.9311510324478149,
      "learning_rate": 0.0007947874792683025,
      "loss": 1.4323,
      "step": 4610
    },
    {
      "epoch": 0.4873417721518987,
      "grad_norm": 0.7102547287940979,
      "learning_rate": 0.0007922811173498293,
      "loss": 1.4064,
      "step": 4620
    },
    {
      "epoch": 0.4883966244725738,
      "grad_norm": 0.7735443711280823,
      "learning_rate": 0.000789774281650791,
      "loss": 1.4264,
      "step": 4630
    },
    {
      "epoch": 0.48945147679324896,
      "grad_norm": 0.8344414234161377,
      "learning_rate": 0.0007872670002615056,
      "loss": 1.4293,
      "step": 4640
    },
    {
      "epoch": 0.49050632911392406,
      "grad_norm": 0.8580491542816162,
      "learning_rate": 0.0007847593012772852,
      "loss": 1.4285,
      "step": 4650
    },
    {
      "epoch": 0.49156118143459915,
      "grad_norm": 0.666509747505188,
      "learning_rate": 0.0007822512127981218,
      "loss": 1.4189,
      "step": 4660
    },
    {
      "epoch": 0.49261603375527424,
      "grad_norm": 0.6427246928215027,
      "learning_rate": 0.0007797427629283708,
      "loss": 1.4183,
      "step": 4670
    },
    {
      "epoch": 0.4936708860759494,
      "grad_norm": 0.6431294083595276,
      "learning_rate": 0.0007772339797764385,
      "loss": 1.413,
      "step": 4680
    },
    {
      "epoch": 0.4947257383966245,
      "grad_norm": 0.9001253843307495,
      "learning_rate": 0.0007747248914544646,
      "loss": 1.4123,
      "step": 4690
    },
    {
      "epoch": 0.4957805907172996,
      "grad_norm": 0.8244867920875549,
      "learning_rate": 0.0007722155260780093,
      "loss": 1.4209,
      "step": 4700
    },
    {
      "epoch": 0.49683544303797467,
      "grad_norm": 0.6998911499977112,
      "learning_rate": 0.0007697059117657368,
      "loss": 1.4235,
      "step": 4710
    },
    {
      "epoch": 0.4978902953586498,
      "grad_norm": 1.1831077337265015,
      "learning_rate": 0.0007671960766391008,
      "loss": 1.4207,
      "step": 4720
    },
    {
      "epoch": 0.4989451476793249,
      "grad_norm": 0.8262091875076294,
      "learning_rate": 0.0007646860488220293,
      "loss": 1.4153,
      "step": 4730
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.6739415526390076,
      "learning_rate": 0.00076217585644061,
      "loss": 1.4132,
      "step": 4740
    },
    {
      "epoch": 0.5010548523206751,
      "grad_norm": 0.6746724247932434,
      "learning_rate": 0.0007596655276227739,
      "loss": 1.4319,
      "step": 4750
    },
    {
      "epoch": 0.5021097046413502,
      "grad_norm": 0.7905839085578918,
      "learning_rate": 0.0007571550904979812,
      "loss": 1.4155,
      "step": 4760
    },
    {
      "epoch": 0.5031645569620253,
      "grad_norm": 0.8470945358276367,
      "learning_rate": 0.0007546445731969056,
      "loss": 1.4148,
      "step": 4770
    },
    {
      "epoch": 0.5042194092827004,
      "grad_norm": 0.7767723798751831,
      "learning_rate": 0.0007521340038511196,
      "loss": 1.4253,
      "step": 4780
    },
    {
      "epoch": 0.5052742616033755,
      "grad_norm": 0.7497154474258423,
      "learning_rate": 0.0007496234105927785,
      "loss": 1.407,
      "step": 4790
    },
    {
      "epoch": 0.5063291139240507,
      "grad_norm": 0.7352545261383057,
      "learning_rate": 0.0007471128215543056,
      "loss": 1.4122,
      "step": 4800
    },
    {
      "epoch": 0.5073839662447257,
      "grad_norm": 0.677992582321167,
      "learning_rate": 0.0007446022648680768,
      "loss": 1.4133,
      "step": 4810
    },
    {
      "epoch": 0.5084388185654009,
      "grad_norm": 0.8154480457305908,
      "learning_rate": 0.0007420917686661055,
      "loss": 1.4318,
      "step": 4820
    },
    {
      "epoch": 0.509493670886076,
      "grad_norm": 0.6944006681442261,
      "learning_rate": 0.0007395813610797283,
      "loss": 1.4139,
      "step": 4830
    },
    {
      "epoch": 0.510548523206751,
      "grad_norm": 1.21970796585083,
      "learning_rate": 0.0007370710702392873,
      "loss": 1.4091,
      "step": 4840
    },
    {
      "epoch": 0.5116033755274262,
      "grad_norm": 0.9442689418792725,
      "learning_rate": 0.0007345609242738173,
      "loss": 1.4194,
      "step": 4850
    },
    {
      "epoch": 0.5126582278481012,
      "grad_norm": 0.6289509534835815,
      "learning_rate": 0.0007320509513107296,
      "loss": 1.4063,
      "step": 4860
    },
    {
      "epoch": 0.5137130801687764,
      "grad_norm": 0.8363620638847351,
      "learning_rate": 0.0007295411794754967,
      "loss": 1.4006,
      "step": 4870
    },
    {
      "epoch": 0.5147679324894515,
      "grad_norm": 0.7297316193580627,
      "learning_rate": 0.0007270316368913374,
      "loss": 1.4184,
      "step": 4880
    },
    {
      "epoch": 0.5158227848101266,
      "grad_norm": 1.0297095775604248,
      "learning_rate": 0.0007245223516789019,
      "loss": 1.4117,
      "step": 4890
    },
    {
      "epoch": 0.5168776371308017,
      "grad_norm": 0.7083756327629089,
      "learning_rate": 0.0007220133519559563,
      "loss": 1.4193,
      "step": 4900
    },
    {
      "epoch": 0.5179324894514767,
      "grad_norm": 0.8673659563064575,
      "learning_rate": 0.0007195046658370675,
      "loss": 1.4143,
      "step": 4910
    },
    {
      "epoch": 0.5189873417721519,
      "grad_norm": 0.9152645468711853,
      "learning_rate": 0.0007169963214332885,
      "loss": 1.4022,
      "step": 4920
    },
    {
      "epoch": 0.520042194092827,
      "grad_norm": 0.8333408832550049,
      "learning_rate": 0.000714488346851843,
      "loss": 1.4126,
      "step": 4930
    },
    {
      "epoch": 0.5210970464135021,
      "grad_norm": 0.7255474328994751,
      "learning_rate": 0.000711980770195811,
      "loss": 1.4098,
      "step": 4940
    },
    {
      "epoch": 0.5221518987341772,
      "grad_norm": 0.7678618431091309,
      "learning_rate": 0.0007094736195638128,
      "loss": 1.4179,
      "step": 4950
    },
    {
      "epoch": 0.5232067510548524,
      "grad_norm": 0.7069323658943176,
      "learning_rate": 0.0007069669230496961,
      "loss": 1.4,
      "step": 4960
    },
    {
      "epoch": 0.5242616033755274,
      "grad_norm": 1.0740123987197876,
      "learning_rate": 0.0007044607087422191,
      "loss": 1.4057,
      "step": 4970
    },
    {
      "epoch": 0.5253164556962026,
      "grad_norm": 0.7819149494171143,
      "learning_rate": 0.000701955004724737,
      "loss": 1.4024,
      "step": 4980
    },
    {
      "epoch": 0.5263713080168776,
      "grad_norm": 0.646610677242279,
      "learning_rate": 0.0006994498390748865,
      "loss": 1.4019,
      "step": 4990
    },
    {
      "epoch": 0.5274261603375527,
      "grad_norm": 0.7318787574768066,
      "learning_rate": 0.0006969452398642721,
      "loss": 1.4126,
      "step": 5000
    },
    {
      "epoch": 0.5284810126582279,
      "grad_norm": 0.6314866542816162,
      "learning_rate": 0.0006944412351581506,
      "loss": 1.4061,
      "step": 5010
    },
    {
      "epoch": 0.5295358649789029,
      "grad_norm": 0.834693431854248,
      "learning_rate": 0.0006919378530151182,
      "loss": 1.4016,
      "step": 5020
    },
    {
      "epoch": 0.5305907172995781,
      "grad_norm": 0.6236812472343445,
      "learning_rate": 0.0006894351214867937,
      "loss": 1.3846,
      "step": 5030
    },
    {
      "epoch": 0.5316455696202531,
      "grad_norm": 0.664553165435791,
      "learning_rate": 0.0006869330686175058,
      "loss": 1.4118,
      "step": 5040
    },
    {
      "epoch": 0.5327004219409283,
      "grad_norm": 0.6874618530273438,
      "learning_rate": 0.0006844317224439788,
      "loss": 1.3986,
      "step": 5050
    },
    {
      "epoch": 0.5337552742616034,
      "grad_norm": 0.6190400123596191,
      "learning_rate": 0.0006819311109950177,
      "loss": 1.4111,
      "step": 5060
    },
    {
      "epoch": 0.5348101265822784,
      "grad_norm": 1.047164797782898,
      "learning_rate": 0.0006794312622911953,
      "loss": 1.4003,
      "step": 5070
    },
    {
      "epoch": 0.5358649789029536,
      "grad_norm": 0.7667319774627686,
      "learning_rate": 0.0006769322043445363,
      "loss": 1.4141,
      "step": 5080
    },
    {
      "epoch": 0.5369198312236287,
      "grad_norm": 0.9032647013664246,
      "learning_rate": 0.0006744339651582059,
      "loss": 1.4043,
      "step": 5090
    },
    {
      "epoch": 0.5379746835443038,
      "grad_norm": 0.7803893089294434,
      "learning_rate": 0.0006719365727261935,
      "loss": 1.3884,
      "step": 5100
    },
    {
      "epoch": 0.5390295358649789,
      "grad_norm": 0.714446485042572,
      "learning_rate": 0.0006694400550330013,
      "loss": 1.4024,
      "step": 5110
    },
    {
      "epoch": 0.540084388185654,
      "grad_norm": 0.6937583684921265,
      "learning_rate": 0.0006669444400533286,
      "loss": 1.3993,
      "step": 5120
    },
    {
      "epoch": 0.5411392405063291,
      "grad_norm": 0.8323858976364136,
      "learning_rate": 0.0006644497557517599,
      "loss": 1.3971,
      "step": 5130
    },
    {
      "epoch": 0.5421940928270043,
      "grad_norm": 0.8532124757766724,
      "learning_rate": 0.0006619560300824507,
      "loss": 1.4123,
      "step": 5140
    },
    {
      "epoch": 0.5432489451476793,
      "grad_norm": 0.7675754427909851,
      "learning_rate": 0.0006594632909888154,
      "loss": 1.4055,
      "step": 5150
    },
    {
      "epoch": 0.5443037974683544,
      "grad_norm": 0.8182488679885864,
      "learning_rate": 0.0006569715664032124,
      "loss": 1.401,
      "step": 5160
    },
    {
      "epoch": 0.5453586497890295,
      "grad_norm": 0.7972416281700134,
      "learning_rate": 0.0006544808842466324,
      "loss": 1.3999,
      "step": 5170
    },
    {
      "epoch": 0.5464135021097046,
      "grad_norm": 0.8812068104743958,
      "learning_rate": 0.0006519912724283851,
      "loss": 1.41,
      "step": 5180
    },
    {
      "epoch": 0.5474683544303798,
      "grad_norm": 0.7234752774238586,
      "learning_rate": 0.0006495027588457864,
      "loss": 1.3926,
      "step": 5190
    },
    {
      "epoch": 0.5485232067510548,
      "grad_norm": 0.9225847721099854,
      "learning_rate": 0.0006470153713838463,
      "loss": 1.4054,
      "step": 5200
    },
    {
      "epoch": 0.54957805907173,
      "grad_norm": 0.7419421672821045,
      "learning_rate": 0.0006445291379149556,
      "loss": 1.3891,
      "step": 5210
    },
    {
      "epoch": 0.5506329113924051,
      "grad_norm": 0.6472369432449341,
      "learning_rate": 0.0006420440862985748,
      "loss": 1.3972,
      "step": 5220
    },
    {
      "epoch": 0.5516877637130801,
      "grad_norm": 1.104505181312561,
      "learning_rate": 0.0006395602443809203,
      "loss": 1.3924,
      "step": 5230
    },
    {
      "epoch": 0.5527426160337553,
      "grad_norm": 0.7523930072784424,
      "learning_rate": 0.0006370776399946536,
      "loss": 1.3994,
      "step": 5240
    },
    {
      "epoch": 0.5537974683544303,
      "grad_norm": 1.0366647243499756,
      "learning_rate": 0.0006345963009585694,
      "loss": 1.391,
      "step": 5250
    },
    {
      "epoch": 0.5548523206751055,
      "grad_norm": 0.8382375836372375,
      "learning_rate": 0.0006321162550772829,
      "loss": 1.3939,
      "step": 5260
    },
    {
      "epoch": 0.5559071729957806,
      "grad_norm": 0.6724934577941895,
      "learning_rate": 0.0006296375301409187,
      "loss": 1.4023,
      "step": 5270
    },
    {
      "epoch": 0.5569620253164557,
      "grad_norm": 0.8834611773490906,
      "learning_rate": 0.0006271601539248012,
      "loss": 1.3906,
      "step": 5280
    },
    {
      "epoch": 0.5580168776371308,
      "grad_norm": 0.8029678463935852,
      "learning_rate": 0.0006246841541891399,
      "loss": 1.3942,
      "step": 5290
    },
    {
      "epoch": 0.5590717299578059,
      "grad_norm": 0.7332834005355835,
      "learning_rate": 0.0006222095586787208,
      "loss": 1.396,
      "step": 5300
    },
    {
      "epoch": 0.560126582278481,
      "grad_norm": 0.8929516077041626,
      "learning_rate": 0.0006197363951225951,
      "loss": 1.386,
      "step": 5310
    },
    {
      "epoch": 0.5611814345991561,
      "grad_norm": 0.8867559432983398,
      "learning_rate": 0.0006172646912337678,
      "loss": 1.3861,
      "step": 5320
    },
    {
      "epoch": 0.5622362869198312,
      "grad_norm": 0.6337891817092896,
      "learning_rate": 0.0006147944747088881,
      "loss": 1.3954,
      "step": 5330
    },
    {
      "epoch": 0.5632911392405063,
      "grad_norm": 0.7554860711097717,
      "learning_rate": 0.000612325773227938,
      "loss": 1.3878,
      "step": 5340
    },
    {
      "epoch": 0.5643459915611815,
      "grad_norm": 0.8300672173500061,
      "learning_rate": 0.0006098586144539235,
      "loss": 1.3822,
      "step": 5350
    },
    {
      "epoch": 0.5654008438818565,
      "grad_norm": 0.6472622156143188,
      "learning_rate": 0.0006073930260325632,
      "loss": 1.3894,
      "step": 5360
    },
    {
      "epoch": 0.5664556962025317,
      "grad_norm": 0.678819477558136,
      "learning_rate": 0.0006049290355919792,
      "loss": 1.3886,
      "step": 5370
    },
    {
      "epoch": 0.5675105485232067,
      "grad_norm": 0.7347920536994934,
      "learning_rate": 0.0006024666707423875,
      "loss": 1.3831,
      "step": 5380
    },
    {
      "epoch": 0.5685654008438819,
      "grad_norm": 0.7904503345489502,
      "learning_rate": 0.0006000059590757886,
      "loss": 1.3858,
      "step": 5390
    },
    {
      "epoch": 0.569620253164557,
      "grad_norm": 0.802675187587738,
      "learning_rate": 0.0005975469281656581,
      "loss": 1.3842,
      "step": 5400
    },
    {
      "epoch": 0.570675105485232,
      "grad_norm": 0.6231290698051453,
      "learning_rate": 0.0005950896055666384,
      "loss": 1.3863,
      "step": 5410
    },
    {
      "epoch": 0.5717299578059072,
      "grad_norm": 0.6504110097885132,
      "learning_rate": 0.0005926340188142289,
      "loss": 1.3792,
      "step": 5420
    },
    {
      "epoch": 0.5727848101265823,
      "grad_norm": 0.6663057804107666,
      "learning_rate": 0.0005901801954244782,
      "loss": 1.3779,
      "step": 5430
    },
    {
      "epoch": 0.5738396624472574,
      "grad_norm": 0.68030846118927,
      "learning_rate": 0.0005877281628936756,
      "loss": 1.3756,
      "step": 5440
    },
    {
      "epoch": 0.5748945147679325,
      "grad_norm": 0.688315749168396,
      "learning_rate": 0.0005852779486980427,
      "loss": 1.3829,
      "step": 5450
    },
    {
      "epoch": 0.5759493670886076,
      "grad_norm": 0.6646932363510132,
      "learning_rate": 0.0005828295802934263,
      "loss": 1.3842,
      "step": 5460
    },
    {
      "epoch": 0.5770042194092827,
      "grad_norm": 0.7625579237937927,
      "learning_rate": 0.0005803830851149892,
      "loss": 1.3835,
      "step": 5470
    },
    {
      "epoch": 0.5780590717299579,
      "grad_norm": 0.6739574074745178,
      "learning_rate": 0.0005779384905769053,
      "loss": 1.3878,
      "step": 5480
    },
    {
      "epoch": 0.5791139240506329,
      "grad_norm": 0.6768545508384705,
      "learning_rate": 0.0005754958240720498,
      "loss": 1.3907,
      "step": 5490
    },
    {
      "epoch": 0.580168776371308,
      "grad_norm": 0.7745399475097656,
      "learning_rate": 0.0005730551129716936,
      "loss": 1.3809,
      "step": 5500
    },
    {
      "epoch": 0.5812236286919831,
      "grad_norm": 1.0100916624069214,
      "learning_rate": 0.0005706163846251961,
      "loss": 1.3883,
      "step": 5510
    },
    {
      "epoch": 0.5822784810126582,
      "grad_norm": 0.6042978167533875,
      "learning_rate": 0.0005681796663596996,
      "loss": 1.3784,
      "step": 5520
    },
    {
      "epoch": 0.5833333333333334,
      "grad_norm": 0.8127723932266235,
      "learning_rate": 0.0005657449854798216,
      "loss": 1.3695,
      "step": 5530
    },
    {
      "epoch": 0.5843881856540084,
      "grad_norm": 0.6895067691802979,
      "learning_rate": 0.0005633123692673503,
      "loss": 1.3701,
      "step": 5540
    },
    {
      "epoch": 0.5854430379746836,
      "grad_norm": 0.703092098236084,
      "learning_rate": 0.0005608818449809387,
      "loss": 1.3871,
      "step": 5550
    },
    {
      "epoch": 0.5864978902953587,
      "grad_norm": 0.7611879706382751,
      "learning_rate": 0.0005584534398557977,
      "loss": 1.3944,
      "step": 5560
    },
    {
      "epoch": 0.5875527426160337,
      "grad_norm": 0.644504725933075,
      "learning_rate": 0.0005560271811033928,
      "loss": 1.3787,
      "step": 5570
    },
    {
      "epoch": 0.5886075949367089,
      "grad_norm": 0.952104389667511,
      "learning_rate": 0.0005536030959111377,
      "loss": 1.3756,
      "step": 5580
    },
    {
      "epoch": 0.5896624472573839,
      "grad_norm": 0.7188992500305176,
      "learning_rate": 0.0005511812114420908,
      "loss": 1.3782,
      "step": 5590
    },
    {
      "epoch": 0.5907172995780591,
      "grad_norm": 0.7364240884780884,
      "learning_rate": 0.0005487615548346502,
      "loss": 1.3728,
      "step": 5600
    },
    {
      "epoch": 0.5917721518987342,
      "grad_norm": 0.6614511013031006,
      "learning_rate": 0.0005463441532022495,
      "loss": 1.3784,
      "step": 5610
    },
    {
      "epoch": 0.5928270042194093,
      "grad_norm": 0.666408360004425,
      "learning_rate": 0.0005439290336330545,
      "loss": 1.3886,
      "step": 5620
    },
    {
      "epoch": 0.5938818565400844,
      "grad_norm": 0.6617103219032288,
      "learning_rate": 0.0005415162231896593,
      "loss": 1.3792,
      "step": 5630
    },
    {
      "epoch": 0.5949367088607594,
      "grad_norm": 0.9930887222290039,
      "learning_rate": 0.000539105748908783,
      "loss": 1.3881,
      "step": 5640
    },
    {
      "epoch": 0.5959915611814346,
      "grad_norm": 0.8630357384681702,
      "learning_rate": 0.0005366976378009668,
      "loss": 1.382,
      "step": 5650
    },
    {
      "epoch": 0.5970464135021097,
      "grad_norm": 0.8617135882377625,
      "learning_rate": 0.0005342919168502717,
      "loss": 1.3762,
      "step": 5660
    },
    {
      "epoch": 0.5981012658227848,
      "grad_norm": 0.7107936143875122,
      "learning_rate": 0.0005318886130139753,
      "loss": 1.3852,
      "step": 5670
    },
    {
      "epoch": 0.5991561181434599,
      "grad_norm": 0.6863983869552612,
      "learning_rate": 0.0005294877532222709,
      "loss": 1.3864,
      "step": 5680
    },
    {
      "epoch": 0.6002109704641351,
      "grad_norm": 0.9210748076438904,
      "learning_rate": 0.0005270893643779649,
      "loss": 1.3644,
      "step": 5690
    },
    {
      "epoch": 0.6012658227848101,
      "grad_norm": 0.8707157373428345,
      "learning_rate": 0.0005246934733561751,
      "loss": 1.3717,
      "step": 5700
    },
    {
      "epoch": 0.6023206751054853,
      "grad_norm": 0.6799771785736084,
      "learning_rate": 0.0005223001070040305,
      "loss": 1.3735,
      "step": 5710
    },
    {
      "epoch": 0.6033755274261603,
      "grad_norm": 0.8240166902542114,
      "learning_rate": 0.0005199092921403696,
      "loss": 1.3775,
      "step": 5720
    },
    {
      "epoch": 0.6044303797468354,
      "grad_norm": 0.7255812883377075,
      "learning_rate": 0.00051752105555544,
      "loss": 1.3746,
      "step": 5730
    },
    {
      "epoch": 0.6054852320675106,
      "grad_norm": 0.7232790589332581,
      "learning_rate": 0.0005151354240105994,
      "loss": 1.3724,
      "step": 5740
    },
    {
      "epoch": 0.6065400843881856,
      "grad_norm": 0.7431049346923828,
      "learning_rate": 0.0005127524242380139,
      "loss": 1.3494,
      "step": 5750
    },
    {
      "epoch": 0.6075949367088608,
      "grad_norm": 0.7193424105644226,
      "learning_rate": 0.0005103720829403594,
      "loss": 1.3568,
      "step": 5760
    },
    {
      "epoch": 0.6086497890295358,
      "grad_norm": 0.9045636057853699,
      "learning_rate": 0.0005079944267905226,
      "loss": 1.36,
      "step": 5770
    },
    {
      "epoch": 0.609704641350211,
      "grad_norm": 0.7302725911140442,
      "learning_rate": 0.0005056194824313015,
      "loss": 1.3678,
      "step": 5780
    },
    {
      "epoch": 0.6107594936708861,
      "grad_norm": 0.6335333585739136,
      "learning_rate": 0.0005032472764751074,
      "loss": 1.3645,
      "step": 5790
    },
    {
      "epoch": 0.6118143459915611,
      "grad_norm": 0.8211780786514282,
      "learning_rate": 0.000500877835503666,
      "loss": 1.3655,
      "step": 5800
    },
    {
      "epoch": 0.6128691983122363,
      "grad_norm": 0.7296932339668274,
      "learning_rate": 0.000498511186067721,
      "loss": 1.3731,
      "step": 5810
    },
    {
      "epoch": 0.6139240506329114,
      "grad_norm": 0.7274990081787109,
      "learning_rate": 0.0004961473546867346,
      "loss": 1.384,
      "step": 5820
    },
    {
      "epoch": 0.6149789029535865,
      "grad_norm": 0.9230065941810608,
      "learning_rate": 0.0004937863678485915,
      "loss": 1.3689,
      "step": 5830
    },
    {
      "epoch": 0.6160337552742616,
      "grad_norm": 0.7609516978263855,
      "learning_rate": 0.0004914282520093023,
      "loss": 1.3798,
      "step": 5840
    },
    {
      "epoch": 0.6170886075949367,
      "grad_norm": 0.7594123482704163,
      "learning_rate": 0.0004890730335927063,
      "loss": 1.3821,
      "step": 5850
    },
    {
      "epoch": 0.6181434599156118,
      "grad_norm": 0.7904174327850342,
      "learning_rate": 0.00048672073899017564,
      "loss": 1.3682,
      "step": 5860
    },
    {
      "epoch": 0.619198312236287,
      "grad_norm": 0.6791795492172241,
      "learning_rate": 0.0004843713945603205,
      "loss": 1.3622,
      "step": 5870
    },
    {
      "epoch": 0.620253164556962,
      "grad_norm": 0.6898028254508972,
      "learning_rate": 0.00048202502662869195,
      "loss": 1.3737,
      "step": 5880
    },
    {
      "epoch": 0.6213080168776371,
      "grad_norm": 0.9083935618400574,
      "learning_rate": 0.0004796816614874885,
      "loss": 1.363,
      "step": 5890
    },
    {
      "epoch": 0.6223628691983122,
      "grad_norm": 0.8479238152503967,
      "learning_rate": 0.00047734132539526086,
      "loss": 1.3658,
      "step": 5900
    },
    {
      "epoch": 0.6234177215189873,
      "grad_norm": 1.0418318510055542,
      "learning_rate": 0.00047500404457661747,
      "loss": 1.3602,
      "step": 5910
    },
    {
      "epoch": 0.6244725738396625,
      "grad_norm": 0.9775272607803345,
      "learning_rate": 0.00047266984522193134,
      "loss": 1.357,
      "step": 5920
    },
    {
      "epoch": 0.6255274261603375,
      "grad_norm": 0.631068766117096,
      "learning_rate": 0.00047033875348704576,
      "loss": 1.3635,
      "step": 5930
    },
    {
      "epoch": 0.6265822784810127,
      "grad_norm": 0.7417823076248169,
      "learning_rate": 0.00046801079549298224,
      "loss": 1.3669,
      "step": 5940
    },
    {
      "epoch": 0.6276371308016878,
      "grad_norm": 0.7994781136512756,
      "learning_rate": 0.0004656859973256466,
      "loss": 1.3595,
      "step": 5950
    },
    {
      "epoch": 0.6286919831223629,
      "grad_norm": 0.6689051389694214,
      "learning_rate": 0.00046336438503553754,
      "loss": 1.3655,
      "step": 5960
    },
    {
      "epoch": 0.629746835443038,
      "grad_norm": 0.6540760397911072,
      "learning_rate": 0.00046104598463745424,
      "loss": 1.3641,
      "step": 5970
    },
    {
      "epoch": 0.630801687763713,
      "grad_norm": 0.8898299336433411,
      "learning_rate": 0.0004587308221102053,
      "loss": 1.3472,
      "step": 5980
    },
    {
      "epoch": 0.6318565400843882,
      "grad_norm": 0.7407407164573669,
      "learning_rate": 0.00045641892339631703,
      "loss": 1.3652,
      "step": 5990
    },
    {
      "epoch": 0.6329113924050633,
      "grad_norm": 0.7216312289237976,
      "learning_rate": 0.000454110314401744,
      "loss": 1.3737,
      "step": 6000
    },
    {
      "epoch": 0.6339662447257384,
      "grad_norm": 0.7877267003059387,
      "learning_rate": 0.00045180502099557686,
      "loss": 1.3612,
      "step": 6010
    },
    {
      "epoch": 0.6350210970464135,
      "grad_norm": 0.6999595761299133,
      "learning_rate": 0.00044950306900975377,
      "loss": 1.3642,
      "step": 6020
    },
    {
      "epoch": 0.6360759493670886,
      "grad_norm": 0.732111930847168,
      "learning_rate": 0.00044720448423877113,
      "loss": 1.3457,
      "step": 6030
    },
    {
      "epoch": 0.6371308016877637,
      "grad_norm": 0.9228755235671997,
      "learning_rate": 0.0004449092924393933,
      "loss": 1.3651,
      "step": 6040
    },
    {
      "epoch": 0.6381856540084389,
      "grad_norm": 0.7102705836296082,
      "learning_rate": 0.00044261751933036525,
      "loss": 1.3477,
      "step": 6050
    },
    {
      "epoch": 0.6392405063291139,
      "grad_norm": 0.9746001958847046,
      "learning_rate": 0.0004403291905921233,
      "loss": 1.3675,
      "step": 6060
    },
    {
      "epoch": 0.640295358649789,
      "grad_norm": 0.7171458601951599,
      "learning_rate": 0.00043804433186650916,
      "loss": 1.3559,
      "step": 6070
    },
    {
      "epoch": 0.6413502109704642,
      "grad_norm": 0.690568208694458,
      "learning_rate": 0.00043576296875647984,
      "loss": 1.3561,
      "step": 6080
    },
    {
      "epoch": 0.6424050632911392,
      "grad_norm": 0.7487748265266418,
      "learning_rate": 0.0004334851268258234,
      "loss": 1.3543,
      "step": 6090
    },
    {
      "epoch": 0.6434599156118144,
      "grad_norm": 0.8148131966590881,
      "learning_rate": 0.00043121083159887056,
      "loss": 1.3482,
      "step": 6100
    },
    {
      "epoch": 0.6445147679324894,
      "grad_norm": 0.7849488258361816,
      "learning_rate": 0.00042894010856020997,
      "loss": 1.3628,
      "step": 6110
    },
    {
      "epoch": 0.6455696202531646,
      "grad_norm": 0.7116225361824036,
      "learning_rate": 0.0004266729831544017,
      "loss": 1.3564,
      "step": 6120
    },
    {
      "epoch": 0.6466244725738397,
      "grad_norm": 0.6714718341827393,
      "learning_rate": 0.0004244094807856936,
      "loss": 1.3548,
      "step": 6130
    },
    {
      "epoch": 0.6476793248945147,
      "grad_norm": 0.7513552904129028,
      "learning_rate": 0.00042214962681773457,
      "loss": 1.354,
      "step": 6140
    },
    {
      "epoch": 0.6487341772151899,
      "grad_norm": 0.8191859722137451,
      "learning_rate": 0.00041989344657329187,
      "loss": 1.3662,
      "step": 6150
    },
    {
      "epoch": 0.6497890295358649,
      "grad_norm": 0.7799637317657471,
      "learning_rate": 0.00041764096533396667,
      "loss": 1.3558,
      "step": 6160
    },
    {
      "epoch": 0.6508438818565401,
      "grad_norm": 0.6763932704925537,
      "learning_rate": 0.00041539220833991124,
      "loss": 1.3351,
      "step": 6170
    },
    {
      "epoch": 0.6518987341772152,
      "grad_norm": 0.6493298411369324,
      "learning_rate": 0.0004131472007895457,
      "loss": 1.3591,
      "step": 6180
    },
    {
      "epoch": 0.6529535864978903,
      "grad_norm": 0.6854337453842163,
      "learning_rate": 0.00041090596783927583,
      "loss": 1.361,
      "step": 6190
    },
    {
      "epoch": 0.6540084388185654,
      "grad_norm": 0.6839235424995422,
      "learning_rate": 0.0004086685346032111,
      "loss": 1.3539,
      "step": 6200
    },
    {
      "epoch": 0.6550632911392406,
      "grad_norm": 0.6952365040779114,
      "learning_rate": 0.00040643492615288367,
      "loss": 1.3494,
      "step": 6210
    },
    {
      "epoch": 0.6561181434599156,
      "grad_norm": 0.6574273109436035,
      "learning_rate": 0.00040420516751696664,
      "loss": 1.3568,
      "step": 6220
    },
    {
      "epoch": 0.6571729957805907,
      "grad_norm": 0.6705141067504883,
      "learning_rate": 0.00040197928368099445,
      "loss": 1.3452,
      "step": 6230
    },
    {
      "epoch": 0.6582278481012658,
      "grad_norm": 0.8345258235931396,
      "learning_rate": 0.00039975729958708223,
      "loss": 1.3511,
      "step": 6240
    },
    {
      "epoch": 0.6592827004219409,
      "grad_norm": 0.8301558494567871,
      "learning_rate": 0.0003975392401336468,
      "loss": 1.3596,
      "step": 6250
    },
    {
      "epoch": 0.6603375527426161,
      "grad_norm": 0.6865615844726562,
      "learning_rate": 0.00039532513017512694,
      "loss": 1.3479,
      "step": 6260
    },
    {
      "epoch": 0.6613924050632911,
      "grad_norm": 0.6713852286338806,
      "learning_rate": 0.00039311499452170665,
      "loss": 1.3568,
      "step": 6270
    },
    {
      "epoch": 0.6624472573839663,
      "grad_norm": 0.660412073135376,
      "learning_rate": 0.0003909088579390347,
      "loss": 1.3555,
      "step": 6280
    },
    {
      "epoch": 0.6635021097046413,
      "grad_norm": 0.7791770696640015,
      "learning_rate": 0.00038870674514794877,
      "loss": 1.3545,
      "step": 6290
    },
    {
      "epoch": 0.6645569620253164,
      "grad_norm": 0.7100890874862671,
      "learning_rate": 0.0003865086808241979,
      "loss": 1.3466,
      "step": 6300
    },
    {
      "epoch": 0.6656118143459916,
      "grad_norm": 0.6848272681236267,
      "learning_rate": 0.0003843146895981661,
      "loss": 1.3469,
      "step": 6310
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 0.7608482837677002,
      "learning_rate": 0.00038212479605459617,
      "loss": 1.341,
      "step": 6320
    },
    {
      "epoch": 0.6677215189873418,
      "grad_norm": 0.8545164465904236,
      "learning_rate": 0.000379939024732315,
      "loss": 1.3504,
      "step": 6330
    },
    {
      "epoch": 0.6687763713080169,
      "grad_norm": 0.6385216116905212,
      "learning_rate": 0.0003777574001239573,
      "loss": 1.3445,
      "step": 6340
    },
    {
      "epoch": 0.669831223628692,
      "grad_norm": 0.6825957298278809,
      "learning_rate": 0.00037557994667569217,
      "loss": 1.3509,
      "step": 6350
    },
    {
      "epoch": 0.6708860759493671,
      "grad_norm": 0.8143519759178162,
      "learning_rate": 0.0003734066887869485,
      "loss": 1.3481,
      "step": 6360
    },
    {
      "epoch": 0.6719409282700421,
      "grad_norm": 0.6440000534057617,
      "learning_rate": 0.0003712376508101424,
      "loss": 1.3307,
      "step": 6370
    },
    {
      "epoch": 0.6729957805907173,
      "grad_norm": 0.6486859321594238,
      "learning_rate": 0.0003690728570504032,
      "loss": 1.3346,
      "step": 6380
    },
    {
      "epoch": 0.6740506329113924,
      "grad_norm": 0.6770305633544922,
      "learning_rate": 0.00036691233176530197,
      "loss": 1.3519,
      "step": 6390
    },
    {
      "epoch": 0.6751054852320675,
      "grad_norm": 0.941678524017334,
      "learning_rate": 0.00036475609916457996,
      "loss": 1.3512,
      "step": 6400
    },
    {
      "epoch": 0.6761603375527426,
      "grad_norm": 0.8228393793106079,
      "learning_rate": 0.000362604183409876,
      "loss": 1.3447,
      "step": 6410
    },
    {
      "epoch": 0.6772151898734177,
      "grad_norm": 0.684606671333313,
      "learning_rate": 0.00036045660861445684,
      "loss": 1.3373,
      "step": 6420
    },
    {
      "epoch": 0.6782700421940928,
      "grad_norm": 0.7011838555335999,
      "learning_rate": 0.0003583133988429468,
      "loss": 1.3429,
      "step": 6430
    },
    {
      "epoch": 0.679324894514768,
      "grad_norm": 0.9510440826416016,
      "learning_rate": 0.0003561745781110579,
      "loss": 1.3289,
      "step": 6440
    },
    {
      "epoch": 0.680379746835443,
      "grad_norm": 0.7210097908973694,
      "learning_rate": 0.00035404017038532045,
      "loss": 1.3333,
      "step": 6450
    },
    {
      "epoch": 0.6814345991561181,
      "grad_norm": 0.769101619720459,
      "learning_rate": 0.00035191019958281575,
      "loss": 1.3452,
      "step": 6460
    },
    {
      "epoch": 0.6824894514767933,
      "grad_norm": 0.740253746509552,
      "learning_rate": 0.00034978468957090635,
      "loss": 1.3414,
      "step": 6470
    },
    {
      "epoch": 0.6835443037974683,
      "grad_norm": 1.2986252307891846,
      "learning_rate": 0.0003476636641669699,
      "loss": 1.35,
      "step": 6480
    },
    {
      "epoch": 0.6845991561181435,
      "grad_norm": 0.7250733375549316,
      "learning_rate": 0.0003455471471381318,
      "loss": 1.3444,
      "step": 6490
    },
    {
      "epoch": 0.6856540084388185,
      "grad_norm": 0.7665097117424011,
      "learning_rate": 0.0003434351622009985,
      "loss": 1.3568,
      "step": 6500
    },
    {
      "epoch": 0.6867088607594937,
      "grad_norm": 0.7032374143600464,
      "learning_rate": 0.0003413277330213928,
      "loss": 1.3595,
      "step": 6510
    },
    {
      "epoch": 0.6877637130801688,
      "grad_norm": 0.6787393689155579,
      "learning_rate": 0.0003392248832140876,
      "loss": 1.3418,
      "step": 6520
    },
    {
      "epoch": 0.6888185654008439,
      "grad_norm": 0.745445728302002,
      "learning_rate": 0.00033712663634254163,
      "loss": 1.3354,
      "step": 6530
    },
    {
      "epoch": 0.689873417721519,
      "grad_norm": 0.6580391526222229,
      "learning_rate": 0.00033503301591863586,
      "loss": 1.3427,
      "step": 6540
    },
    {
      "epoch": 0.6909282700421941,
      "grad_norm": 0.6638602018356323,
      "learning_rate": 0.0003329440454024092,
      "loss": 1.3398,
      "step": 6550
    },
    {
      "epoch": 0.6919831223628692,
      "grad_norm": 1.0527986288070679,
      "learning_rate": 0.0003308597482017965,
      "loss": 1.3418,
      "step": 6560
    },
    {
      "epoch": 0.6930379746835443,
      "grad_norm": 0.9146166443824768,
      "learning_rate": 0.0003287801476723656,
      "loss": 1.3368,
      "step": 6570
    },
    {
      "epoch": 0.6940928270042194,
      "grad_norm": 0.6148539781570435,
      "learning_rate": 0.00032670526711705536,
      "loss": 1.3349,
      "step": 6580
    },
    {
      "epoch": 0.6951476793248945,
      "grad_norm": 0.6986439228057861,
      "learning_rate": 0.0003246351297859164,
      "loss": 1.3378,
      "step": 6590
    },
    {
      "epoch": 0.6962025316455697,
      "grad_norm": 0.6541133522987366,
      "learning_rate": 0.00032256975887584783,
      "loss": 1.3441,
      "step": 6600
    },
    {
      "epoch": 0.6972573839662447,
      "grad_norm": 0.7345884442329407,
      "learning_rate": 0.00032050917753033935,
      "loss": 1.3303,
      "step": 6610
    },
    {
      "epoch": 0.6983122362869199,
      "grad_norm": 0.6537117958068848,
      "learning_rate": 0.000318453408839211,
      "loss": 1.3396,
      "step": 6620
    },
    {
      "epoch": 0.6993670886075949,
      "grad_norm": 0.6956437230110168,
      "learning_rate": 0.0003164024758383548,
      "loss": 1.3324,
      "step": 6630
    },
    {
      "epoch": 0.70042194092827,
      "grad_norm": 0.7272790670394897,
      "learning_rate": 0.00031435640150947645,
      "loss": 1.3369,
      "step": 6640
    },
    {
      "epoch": 0.7014767932489452,
      "grad_norm": 0.6868227124214172,
      "learning_rate": 0.0003123152087798376,
      "loss": 1.3319,
      "step": 6650
    },
    {
      "epoch": 0.7025316455696202,
      "grad_norm": 0.6691982746124268,
      "learning_rate": 0.00031027892052200003,
      "loss": 1.3325,
      "step": 6660
    },
    {
      "epoch": 0.7035864978902954,
      "grad_norm": 0.6686353087425232,
      "learning_rate": 0.0003082475595535677,
      "loss": 1.3339,
      "step": 6670
    },
    {
      "epoch": 0.7046413502109705,
      "grad_norm": 0.6458573341369629,
      "learning_rate": 0.00030622114863693205,
      "loss": 1.3375,
      "step": 6680
    },
    {
      "epoch": 0.7056962025316456,
      "grad_norm": 0.6652058959007263,
      "learning_rate": 0.00030419971047901704,
      "loss": 1.3294,
      "step": 6690
    },
    {
      "epoch": 0.7067510548523207,
      "grad_norm": 0.6503669619560242,
      "learning_rate": 0.00030218326773102407,
      "loss": 1.3315,
      "step": 6700
    },
    {
      "epoch": 0.7078059071729957,
      "grad_norm": 0.6382567286491394,
      "learning_rate": 0.00030017184298817873,
      "loss": 1.3344,
      "step": 6710
    },
    {
      "epoch": 0.7088607594936709,
      "grad_norm": 0.8746625185012817,
      "learning_rate": 0.00029816545878947763,
      "loss": 1.3251,
      "step": 6720
    },
    {
      "epoch": 0.709915611814346,
      "grad_norm": 0.7243242263793945,
      "learning_rate": 0.00029616413761743537,
      "loss": 1.3336,
      "step": 6730
    },
    {
      "epoch": 0.7109704641350211,
      "grad_norm": 0.8179724812507629,
      "learning_rate": 0.00029416790189783286,
      "loss": 1.3369,
      "step": 6740
    },
    {
      "epoch": 0.7120253164556962,
      "grad_norm": 0.6592196226119995,
      "learning_rate": 0.000292176773999466,
      "loss": 1.3368,
      "step": 6750
    },
    {
      "epoch": 0.7130801687763713,
      "grad_norm": 0.8134679794311523,
      "learning_rate": 0.0002901907762338952,
      "loss": 1.3321,
      "step": 6760
    },
    {
      "epoch": 0.7141350210970464,
      "grad_norm": 0.65492182970047,
      "learning_rate": 0.0002882099308551951,
      "loss": 1.335,
      "step": 6770
    },
    {
      "epoch": 0.7151898734177216,
      "grad_norm": 0.7491779923439026,
      "learning_rate": 0.00028623426005970517,
      "loss": 1.3328,
      "step": 6780
    },
    {
      "epoch": 0.7162447257383966,
      "grad_norm": 0.6567303538322449,
      "learning_rate": 0.00028426378598578187,
      "loss": 1.3366,
      "step": 6790
    },
    {
      "epoch": 0.7172995780590717,
      "grad_norm": 0.6415104866027832,
      "learning_rate": 0.0002822985307135491,
      "loss": 1.3228,
      "step": 6800
    },
    {
      "epoch": 0.7183544303797469,
      "grad_norm": 1.0002161264419556,
      "learning_rate": 0.0002803385162646518,
      "loss": 1.3253,
      "step": 6810
    },
    {
      "epoch": 0.7194092827004219,
      "grad_norm": 0.6753124594688416,
      "learning_rate": 0.0002783837646020089,
      "loss": 1.3301,
      "step": 6820
    },
    {
      "epoch": 0.7204641350210971,
      "grad_norm": 0.7003511786460876,
      "learning_rate": 0.0002764342976295673,
      "loss": 1.3277,
      "step": 6830
    },
    {
      "epoch": 0.7215189873417721,
      "grad_norm": 0.7631194591522217,
      "learning_rate": 0.00027449013719205623,
      "loss": 1.3168,
      "step": 6840
    },
    {
      "epoch": 0.7225738396624473,
      "grad_norm": 0.6557474136352539,
      "learning_rate": 0.00027255130507474276,
      "loss": 1.3255,
      "step": 6850
    },
    {
      "epoch": 0.7236286919831224,
      "grad_norm": 0.8119695782661438,
      "learning_rate": 0.00027061782300318726,
      "loss": 1.321,
      "step": 6860
    },
    {
      "epoch": 0.7246835443037974,
      "grad_norm": 0.6647341251373291,
      "learning_rate": 0.0002686897126430009,
      "loss": 1.3201,
      "step": 6870
    },
    {
      "epoch": 0.7257383966244726,
      "grad_norm": 0.6715971827507019,
      "learning_rate": 0.00026676699559960145,
      "loss": 1.3282,
      "step": 6880
    },
    {
      "epoch": 0.7267932489451476,
      "grad_norm": 0.7210767269134521,
      "learning_rate": 0.00026484969341797224,
      "loss": 1.3172,
      "step": 6890
    },
    {
      "epoch": 0.7278481012658228,
      "grad_norm": 0.6769729852676392,
      "learning_rate": 0.0002629378275824204,
      "loss": 1.3189,
      "step": 6900
    },
    {
      "epoch": 0.7289029535864979,
      "grad_norm": 0.728450357913971,
      "learning_rate": 0.00026103141951633617,
      "loss": 1.3447,
      "step": 6910
    },
    {
      "epoch": 0.729957805907173,
      "grad_norm": 0.8838924765586853,
      "learning_rate": 0.00025913049058195277,
      "loss": 1.3227,
      "step": 6920
    },
    {
      "epoch": 0.7310126582278481,
      "grad_norm": 0.6649219393730164,
      "learning_rate": 0.0002572350620801072,
      "loss": 1.3299,
      "step": 6930
    },
    {
      "epoch": 0.7320675105485233,
      "grad_norm": 0.691104531288147,
      "learning_rate": 0.0002553451552500012,
      "loss": 1.3322,
      "step": 6940
    },
    {
      "epoch": 0.7331223628691983,
      "grad_norm": 0.6900149583816528,
      "learning_rate": 0.0002534607912689637,
      "loss": 1.3296,
      "step": 6950
    },
    {
      "epoch": 0.7341772151898734,
      "grad_norm": 0.6626471281051636,
      "learning_rate": 0.00025158199125221325,
      "loss": 1.3195,
      "step": 6960
    },
    {
      "epoch": 0.7352320675105485,
      "grad_norm": 0.6535481810569763,
      "learning_rate": 0.0002497087762526211,
      "loss": 1.3159,
      "step": 6970
    },
    {
      "epoch": 0.7362869198312236,
      "grad_norm": 0.6971670985221863,
      "learning_rate": 0.0002478411672604766,
      "loss": 1.3218,
      "step": 6980
    },
    {
      "epoch": 0.7373417721518988,
      "grad_norm": 0.6511154174804688,
      "learning_rate": 0.00024597918520324994,
      "loss": 1.3203,
      "step": 6990
    },
    {
      "epoch": 0.7383966244725738,
      "grad_norm": 0.7733587622642517,
      "learning_rate": 0.00024412285094535952,
      "loss": 1.331,
      "step": 7000
    },
    {
      "epoch": 0.739451476793249,
      "grad_norm": 0.685684859752655,
      "learning_rate": 0.00024227218528793696,
      "loss": 1.3041,
      "step": 7010
    },
    {
      "epoch": 0.740506329113924,
      "grad_norm": 0.7792125344276428,
      "learning_rate": 0.00024042720896859471,
      "loss": 1.318,
      "step": 7020
    },
    {
      "epoch": 0.7415611814345991,
      "grad_norm": 0.7213562726974487,
      "learning_rate": 0.00023858794266119323,
      "loss": 1.3249,
      "step": 7030
    },
    {
      "epoch": 0.7426160337552743,
      "grad_norm": 0.9945384860038757,
      "learning_rate": 0.00023675440697560943,
      "loss": 1.3295,
      "step": 7040
    },
    {
      "epoch": 0.7436708860759493,
      "grad_norm": 0.6952188014984131,
      "learning_rate": 0.0002349266224575063,
      "loss": 1.3156,
      "step": 7050
    },
    {
      "epoch": 0.7447257383966245,
      "grad_norm": 0.6873391270637512,
      "learning_rate": 0.0002331046095881017,
      "loss": 1.3196,
      "step": 7060
    },
    {
      "epoch": 0.7457805907172996,
      "grad_norm": 0.7940738797187805,
      "learning_rate": 0.00023128838878393946,
      "loss": 1.3278,
      "step": 7070
    },
    {
      "epoch": 0.7468354430379747,
      "grad_norm": 0.6671113967895508,
      "learning_rate": 0.00022947798039666051,
      "loss": 1.3225,
      "step": 7080
    },
    {
      "epoch": 0.7478902953586498,
      "grad_norm": 0.6724119782447815,
      "learning_rate": 0.00022767340471277492,
      "loss": 1.322,
      "step": 7090
    },
    {
      "epoch": 0.7489451476793249,
      "grad_norm": 0.6412546038627625,
      "learning_rate": 0.00022587468195343436,
      "loss": 1.318,
      "step": 7100
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.6673798561096191,
      "learning_rate": 0.00022408183227420528,
      "loss": 1.327,
      "step": 7110
    },
    {
      "epoch": 0.7510548523206751,
      "grad_norm": 0.7734135985374451,
      "learning_rate": 0.0002222948757648443,
      "loss": 1.3212,
      "step": 7120
    },
    {
      "epoch": 0.7521097046413502,
      "grad_norm": 0.6902635097503662,
      "learning_rate": 0.00022051383244907143,
      "loss": 1.3307,
      "step": 7130
    },
    {
      "epoch": 0.7531645569620253,
      "grad_norm": 0.7915105819702148,
      "learning_rate": 0.0002187387222843467,
      "loss": 1.3149,
      "step": 7140
    },
    {
      "epoch": 0.7542194092827004,
      "grad_norm": 0.6439545750617981,
      "learning_rate": 0.0002169695651616463,
      "loss": 1.3108,
      "step": 7150
    },
    {
      "epoch": 0.7552742616033755,
      "grad_norm": 0.7627606391906738,
      "learning_rate": 0.00021520638090523955,
      "loss": 1.3194,
      "step": 7160
    },
    {
      "epoch": 0.7563291139240507,
      "grad_norm": 1.0325992107391357,
      "learning_rate": 0.00021344918927246678,
      "loss": 1.3272,
      "step": 7170
    },
    {
      "epoch": 0.7573839662447257,
      "grad_norm": 0.744968593120575,
      "learning_rate": 0.00021169800995351874,
      "loss": 1.3062,
      "step": 7180
    },
    {
      "epoch": 0.7584388185654009,
      "grad_norm": 0.6517015099525452,
      "learning_rate": 0.00020995286257121453,
      "loss": 1.3097,
      "step": 7190
    },
    {
      "epoch": 0.759493670886076,
      "grad_norm": 0.6919528841972351,
      "learning_rate": 0.00020821376668078264,
      "loss": 1.314,
      "step": 7200
    },
    {
      "epoch": 0.760548523206751,
      "grad_norm": 0.6990111470222473,
      "learning_rate": 0.00020648074176964182,
      "loss": 1.3179,
      "step": 7210
    },
    {
      "epoch": 0.7616033755274262,
      "grad_norm": 0.8284183740615845,
      "learning_rate": 0.00020475380725718228,
      "loss": 1.331,
      "step": 7220
    },
    {
      "epoch": 0.7626582278481012,
      "grad_norm": 0.7366883754730225,
      "learning_rate": 0.00020303298249454857,
      "loss": 1.3269,
      "step": 7230
    },
    {
      "epoch": 0.7637130801687764,
      "grad_norm": 0.7097327709197998,
      "learning_rate": 0.00020131828676442237,
      "loss": 1.3219,
      "step": 7240
    },
    {
      "epoch": 0.7647679324894515,
      "grad_norm": 0.6799291968345642,
      "learning_rate": 0.00019960973928080666,
      "loss": 1.3186,
      "step": 7250
    },
    {
      "epoch": 0.7658227848101266,
      "grad_norm": 0.6854842901229858,
      "learning_rate": 0.0001979073591888101,
      "loss": 1.3169,
      "step": 7260
    },
    {
      "epoch": 0.7668776371308017,
      "grad_norm": 0.6777929067611694,
      "learning_rate": 0.000196211165564433,
      "loss": 1.3129,
      "step": 7270
    },
    {
      "epoch": 0.7679324894514767,
      "grad_norm": 0.9655599594116211,
      "learning_rate": 0.00019452117741435314,
      "loss": 1.3171,
      "step": 7280
    },
    {
      "epoch": 0.7689873417721519,
      "grad_norm": 0.6834695339202881,
      "learning_rate": 0.00019283741367571294,
      "loss": 1.3248,
      "step": 7290
    },
    {
      "epoch": 0.770042194092827,
      "grad_norm": 0.8187342286109924,
      "learning_rate": 0.00019115989321590694,
      "loss": 1.3301,
      "step": 7300
    },
    {
      "epoch": 0.7710970464135021,
      "grad_norm": 0.6526690125465393,
      "learning_rate": 0.00018948863483237154,
      "loss": 1.3129,
      "step": 7310
    },
    {
      "epoch": 0.7721518987341772,
      "grad_norm": 0.7523473501205444,
      "learning_rate": 0.00018782365725237272,
      "loss": 1.3089,
      "step": 7320
    },
    {
      "epoch": 0.7732067510548524,
      "grad_norm": 0.7795331478118896,
      "learning_rate": 0.00018616497913279728,
      "loss": 1.3111,
      "step": 7330
    },
    {
      "epoch": 0.7742616033755274,
      "grad_norm": 0.8736000657081604,
      "learning_rate": 0.0001845126190599434,
      "loss": 1.3104,
      "step": 7340
    },
    {
      "epoch": 0.7753164556962026,
      "grad_norm": 0.7167467474937439,
      "learning_rate": 0.00018286659554931254,
      "loss": 1.3073,
      "step": 7350
    },
    {
      "epoch": 0.7763713080168776,
      "grad_norm": 0.7300223708152771,
      "learning_rate": 0.00018122692704540194,
      "loss": 1.3142,
      "step": 7360
    },
    {
      "epoch": 0.7774261603375527,
      "grad_norm": 0.671873152256012,
      "learning_rate": 0.00017959363192149752,
      "loss": 1.3149,
      "step": 7370
    },
    {
      "epoch": 0.7784810126582279,
      "grad_norm": 0.6954656839370728,
      "learning_rate": 0.00017796672847946905,
      "loss": 1.3026,
      "step": 7380
    },
    {
      "epoch": 0.7795358649789029,
      "grad_norm": 0.7982514500617981,
      "learning_rate": 0.0001763462349495639,
      "loss": 1.3143,
      "step": 7390
    },
    {
      "epoch": 0.7805907172995781,
      "grad_norm": 0.8448479771614075,
      "learning_rate": 0.00017473216949020326,
      "loss": 1.3192,
      "step": 7400
    },
    {
      "epoch": 0.7816455696202531,
      "grad_norm": 0.7651928663253784,
      "learning_rate": 0.0001731245501877787,
      "loss": 1.3012,
      "step": 7410
    },
    {
      "epoch": 0.7827004219409283,
      "grad_norm": 0.808005690574646,
      "learning_rate": 0.00017152339505644963,
      "loss": 1.307,
      "step": 7420
    },
    {
      "epoch": 0.7837552742616034,
      "grad_norm": 0.6326072812080383,
      "learning_rate": 0.0001699287220379407,
      "loss": 1.2925,
      "step": 7430
    },
    {
      "epoch": 0.7848101265822784,
      "grad_norm": 0.641252875328064,
      "learning_rate": 0.00016834054900134228,
      "loss": 1.2964,
      "step": 7440
    },
    {
      "epoch": 0.7858649789029536,
      "grad_norm": 0.6453474760055542,
      "learning_rate": 0.00016675889374290852,
      "loss": 1.2998,
      "step": 7450
    },
    {
      "epoch": 0.7869198312236287,
      "grad_norm": 0.6745063066482544,
      "learning_rate": 0.0001651837739858589,
      "loss": 1.3118,
      "step": 7460
    },
    {
      "epoch": 0.7879746835443038,
      "grad_norm": 0.6970017552375793,
      "learning_rate": 0.00016361520738017934,
      "loss": 1.3023,
      "step": 7470
    },
    {
      "epoch": 0.7890295358649789,
      "grad_norm": 0.819610595703125,
      "learning_rate": 0.00016205321150242454,
      "loss": 1.3102,
      "step": 7480
    },
    {
      "epoch": 0.790084388185654,
      "grad_norm": 0.6704266667366028,
      "learning_rate": 0.00016049780385552113,
      "loss": 1.3147,
      "step": 7490
    },
    {
      "epoch": 0.7911392405063291,
      "grad_norm": 0.684683620929718,
      "learning_rate": 0.00015894900186857105,
      "loss": 1.3099,
      "step": 7500
    },
    {
      "epoch": 0.7921940928270043,
      "grad_norm": 0.6652682423591614,
      "learning_rate": 0.00015740682289665714,
      "loss": 1.3181,
      "step": 7510
    },
    {
      "epoch": 0.7932489451476793,
      "grad_norm": 0.65765780210495,
      "learning_rate": 0.0001558712842206477,
      "loss": 1.3013,
      "step": 7520
    },
    {
      "epoch": 0.7943037974683544,
      "grad_norm": 0.6687241196632385,
      "learning_rate": 0.00015434240304700332,
      "loss": 1.3025,
      "step": 7530
    },
    {
      "epoch": 0.7953586497890295,
      "grad_norm": 0.6935544013977051,
      "learning_rate": 0.0001528201965075841,
      "loss": 1.299,
      "step": 7540
    },
    {
      "epoch": 0.7964135021097046,
      "grad_norm": 0.7061299681663513,
      "learning_rate": 0.0001513046816594575,
      "loss": 1.2953,
      "step": 7550
    },
    {
      "epoch": 0.7974683544303798,
      "grad_norm": 0.7440963387489319,
      "learning_rate": 0.0001497958754847076,
      "loss": 1.3,
      "step": 7560
    },
    {
      "epoch": 0.7985232067510548,
      "grad_norm": 0.8268978595733643,
      "learning_rate": 0.00014829379489024415,
      "loss": 1.3121,
      "step": 7570
    },
    {
      "epoch": 0.79957805907173,
      "grad_norm": 0.7233864665031433,
      "learning_rate": 0.0001467984567076137,
      "loss": 1.3084,
      "step": 7580
    },
    {
      "epoch": 0.8006329113924051,
      "grad_norm": 0.6765027046203613,
      "learning_rate": 0.00014530987769281075,
      "loss": 1.2992,
      "step": 7590
    },
    {
      "epoch": 0.8016877637130801,
      "grad_norm": 0.7198473215103149,
      "learning_rate": 0.00014382807452609003,
      "loss": 1.3033,
      "step": 7600
    },
    {
      "epoch": 0.8027426160337553,
      "grad_norm": 0.7215249538421631,
      "learning_rate": 0.00014235306381177952,
      "loss": 1.3146,
      "step": 7610
    },
    {
      "epoch": 0.8037974683544303,
      "grad_norm": 0.6656304597854614,
      "learning_rate": 0.00014088486207809449,
      "loss": 1.2876,
      "step": 7620
    },
    {
      "epoch": 0.8048523206751055,
      "grad_norm": 0.6388023495674133,
      "learning_rate": 0.0001394234857769521,
      "loss": 1.3128,
      "step": 7630
    },
    {
      "epoch": 0.8059071729957806,
      "grad_norm": 0.6659093499183655,
      "learning_rate": 0.0001379689512837878,
      "loss": 1.3029,
      "step": 7640
    },
    {
      "epoch": 0.8069620253164557,
      "grad_norm": 0.6641107797622681,
      "learning_rate": 0.00013652127489737067,
      "loss": 1.3196,
      "step": 7650
    },
    {
      "epoch": 0.8080168776371308,
      "grad_norm": 0.6964825391769409,
      "learning_rate": 0.00013508047283962137,
      "loss": 1.2966,
      "step": 7660
    },
    {
      "epoch": 0.8090717299578059,
      "grad_norm": 0.6330327391624451,
      "learning_rate": 0.00013364656125543044,
      "loss": 1.303,
      "step": 7670
    },
    {
      "epoch": 0.810126582278481,
      "grad_norm": 0.8244850039482117,
      "learning_rate": 0.00013221955621247749,
      "loss": 1.3179,
      "step": 7680
    },
    {
      "epoch": 0.8111814345991561,
      "grad_norm": 0.7391464710235596,
      "learning_rate": 0.00013079947370105057,
      "loss": 1.2909,
      "step": 7690
    },
    {
      "epoch": 0.8122362869198312,
      "grad_norm": 0.6498596668243408,
      "learning_rate": 0.00012938632963386808,
      "loss": 1.3022,
      "step": 7700
    },
    {
      "epoch": 0.8132911392405063,
      "grad_norm": 0.9545032382011414,
      "learning_rate": 0.00012798013984589894,
      "loss": 1.2983,
      "step": 7710
    },
    {
      "epoch": 0.8143459915611815,
      "grad_norm": 0.9015225172042847,
      "learning_rate": 0.00012658092009418652,
      "loss": 1.303,
      "step": 7720
    },
    {
      "epoch": 0.8154008438818565,
      "grad_norm": 0.6953148245811462,
      "learning_rate": 0.00012518868605767118,
      "loss": 1.3097,
      "step": 7730
    },
    {
      "epoch": 0.8164556962025317,
      "grad_norm": 0.7007344365119934,
      "learning_rate": 0.0001238034533370153,
      "loss": 1.3097,
      "step": 7740
    },
    {
      "epoch": 0.8175105485232067,
      "grad_norm": 0.6556501984596252,
      "learning_rate": 0.0001224252374544278,
      "loss": 1.294,
      "step": 7750
    },
    {
      "epoch": 0.8185654008438819,
      "grad_norm": 0.7583401203155518,
      "learning_rate": 0.00012105405385349047,
      "loss": 1.2953,
      "step": 7760
    },
    {
      "epoch": 0.819620253164557,
      "grad_norm": 0.6536235213279724,
      "learning_rate": 0.00011968991789898533,
      "loss": 1.3026,
      "step": 7770
    },
    {
      "epoch": 0.820675105485232,
      "grad_norm": 0.9007168412208557,
      "learning_rate": 0.00011833284487672185,
      "loss": 1.2947,
      "step": 7780
    },
    {
      "epoch": 0.8217299578059072,
      "grad_norm": 0.822363018989563,
      "learning_rate": 0.00011698284999336578,
      "loss": 1.3124,
      "step": 7790
    },
    {
      "epoch": 0.8227848101265823,
      "grad_norm": 0.6882434487342834,
      "learning_rate": 0.00011563994837626898,
      "loss": 1.301,
      "step": 7800
    },
    {
      "epoch": 0.8238396624472574,
      "grad_norm": 0.6219406723976135,
      "learning_rate": 0.00011430415507329975,
      "loss": 1.2994,
      "step": 7810
    },
    {
      "epoch": 0.8248945147679325,
      "grad_norm": 0.7564154863357544,
      "learning_rate": 0.00011297548505267424,
      "loss": 1.3064,
      "step": 7820
    },
    {
      "epoch": 0.8259493670886076,
      "grad_norm": 0.7380446791648865,
      "learning_rate": 0.00011165395320278898,
      "loss": 1.2975,
      "step": 7830
    },
    {
      "epoch": 0.8270042194092827,
      "grad_norm": 0.6821550726890564,
      "learning_rate": 0.00011033957433205364,
      "loss": 1.2846,
      "step": 7840
    },
    {
      "epoch": 0.8280590717299579,
      "grad_norm": 0.6716055274009705,
      "learning_rate": 0.00010903236316872514,
      "loss": 1.2925,
      "step": 7850
    },
    {
      "epoch": 0.8291139240506329,
      "grad_norm": 0.6739335656166077,
      "learning_rate": 0.00010773233436074287,
      "loss": 1.2812,
      "step": 7860
    },
    {
      "epoch": 0.830168776371308,
      "grad_norm": 0.649252712726593,
      "learning_rate": 0.00010643950247556447,
      "loss": 1.2939,
      "step": 7870
    },
    {
      "epoch": 0.8312236286919831,
      "grad_norm": 0.7520459890365601,
      "learning_rate": 0.00010515388200000245,
      "loss": 1.3141,
      "step": 7880
    },
    {
      "epoch": 0.8322784810126582,
      "grad_norm": 0.6486638784408569,
      "learning_rate": 0.00010387548734006195,
      "loss": 1.3097,
      "step": 7890
    },
    {
      "epoch": 0.8333333333333334,
      "grad_norm": 0.6894299983978271,
      "learning_rate": 0.00010260433282077944,
      "loss": 1.2909,
      "step": 7900
    },
    {
      "epoch": 0.8343881856540084,
      "grad_norm": 0.7930160164833069,
      "learning_rate": 0.00010134043268606191,
      "loss": 1.2994,
      "step": 7910
    },
    {
      "epoch": 0.8354430379746836,
      "grad_norm": 0.7372665405273438,
      "learning_rate": 0.00010008380109852752,
      "loss": 1.2958,
      "step": 7920
    },
    {
      "epoch": 0.8364978902953587,
      "grad_norm": 0.7255853414535522,
      "learning_rate": 9.883445213934675e-05,
      "loss": 1.3076,
      "step": 7930
    },
    {
      "epoch": 0.8375527426160337,
      "grad_norm": 0.641669750213623,
      "learning_rate": 9.759239980808494e-05,
      "loss": 1.2994,
      "step": 7940
    },
    {
      "epoch": 0.8386075949367089,
      "grad_norm": 0.6384446024894714,
      "learning_rate": 9.635765802254482e-05,
      "loss": 1.3037,
      "step": 7950
    },
    {
      "epoch": 0.8396624472573839,
      "grad_norm": 0.752605140209198,
      "learning_rate": 9.5130240618611e-05,
      "loss": 1.2909,
      "step": 7960
    },
    {
      "epoch": 0.8407172995780591,
      "grad_norm": 0.7616172432899475,
      "learning_rate": 9.391016135009484e-05,
      "loss": 1.2905,
      "step": 7970
    },
    {
      "epoch": 0.8417721518987342,
      "grad_norm": 0.6532970666885376,
      "learning_rate": 9.269743388858019e-05,
      "loss": 1.2887,
      "step": 7980
    },
    {
      "epoch": 0.8428270042194093,
      "grad_norm": 0.6917771100997925,
      "learning_rate": 9.149207182327054e-05,
      "loss": 1.2991,
      "step": 7990
    },
    {
      "epoch": 0.8438818565400844,
      "grad_norm": 0.6677715182304382,
      "learning_rate": 9.029408866083638e-05,
      "loss": 1.2947,
      "step": 8000
    },
    {
      "epoch": 0.8449367088607594,
      "grad_norm": 0.8115660548210144,
      "learning_rate": 8.92222238732397e-05,
      "loss": 1.3028,
      "step": 8010
    },
    {
      "epoch": 0.8459915611814346,
      "grad_norm": 0.676626443862915,
      "learning_rate": 8.803829754089138e-05,
      "loss": 1.2942,
      "step": 8020
    },
    {
      "epoch": 0.8470464135021097,
      "grad_norm": 0.6849190592765808,
      "learning_rate": 8.686178881264568e-05,
      "loss": 1.2918,
      "step": 8030
    },
    {
      "epoch": 0.8481012658227848,
      "grad_norm": 0.6659732460975647,
      "learning_rate": 8.569271087185756e-05,
      "loss": 1.3049,
      "step": 8040
    },
    {
      "epoch": 0.8491561181434599,
      "grad_norm": 0.6546719074249268,
      "learning_rate": 8.453107681861616e-05,
      "loss": 1.2842,
      "step": 8050
    },
    {
      "epoch": 0.8502109704641351,
      "grad_norm": 0.6582532525062561,
      "learning_rate": 8.337689966959819e-05,
      "loss": 1.2939,
      "step": 8060
    },
    {
      "epoch": 0.8512658227848101,
      "grad_norm": 0.6461778283119202,
      "learning_rate": 8.223019235792214e-05,
      "loss": 1.2993,
      "step": 8070
    },
    {
      "epoch": 0.8523206751054853,
      "grad_norm": 0.6988905072212219,
      "learning_rate": 8.109096773300348e-05,
      "loss": 1.3008,
      "step": 8080
    },
    {
      "epoch": 0.8533755274261603,
      "grad_norm": 0.6567661166191101,
      "learning_rate": 7.995923856041013e-05,
      "loss": 1.2877,
      "step": 8090
    },
    {
      "epoch": 0.8544303797468354,
      "grad_norm": 0.8190459609031677,
      "learning_rate": 7.883501752172038e-05,
      "loss": 1.2844,
      "step": 8100
    },
    {
      "epoch": 0.8554852320675106,
      "grad_norm": 0.6638689637184143,
      "learning_rate": 7.771831721437989e-05,
      "loss": 1.3008,
      "step": 8110
    },
    {
      "epoch": 0.8565400843881856,
      "grad_norm": 0.6531414985656738,
      "learning_rate": 7.660915015156067e-05,
      "loss": 1.289,
      "step": 8120
    },
    {
      "epoch": 0.8575949367088608,
      "grad_norm": 0.7257735133171082,
      "learning_rate": 7.55075287620215e-05,
      "loss": 1.2923,
      "step": 8130
    },
    {
      "epoch": 0.8586497890295358,
      "grad_norm": 0.6310346126556396,
      "learning_rate": 7.441346538996769e-05,
      "loss": 1.291,
      "step": 8140
    },
    {
      "epoch": 0.859704641350211,
      "grad_norm": 0.6479582786560059,
      "learning_rate": 7.332697229491373e-05,
      "loss": 1.2854,
      "step": 8150
    },
    {
      "epoch": 0.8607594936708861,
      "grad_norm": 0.6325903534889221,
      "learning_rate": 7.224806165154504e-05,
      "loss": 1.2963,
      "step": 8160
    },
    {
      "epoch": 0.8618143459915611,
      "grad_norm": 0.677290141582489,
      "learning_rate": 7.117674554958253e-05,
      "loss": 1.2903,
      "step": 8170
    },
    {
      "epoch": 0.8628691983122363,
      "grad_norm": 0.707441508769989,
      "learning_rate": 7.011303599364608e-05,
      "loss": 1.2964,
      "step": 8180
    },
    {
      "epoch": 0.8639240506329114,
      "grad_norm": 0.7063124775886536,
      "learning_rate": 6.905694490312064e-05,
      "loss": 1.2892,
      "step": 8190
    },
    {
      "epoch": 0.8649789029535865,
      "grad_norm": 0.6436252593994141,
      "learning_rate": 6.80084841120226e-05,
      "loss": 1.3,
      "step": 8200
    },
    {
      "epoch": 0.8660337552742616,
      "grad_norm": 0.6381955742835999,
      "learning_rate": 6.696766536886692e-05,
      "loss": 1.2943,
      "step": 8210
    },
    {
      "epoch": 0.8670886075949367,
      "grad_norm": 0.7381530404090881,
      "learning_rate": 6.593450033653586e-05,
      "loss": 1.2994,
      "step": 8220
    },
    {
      "epoch": 0.8681434599156118,
      "grad_norm": 0.6792395710945129,
      "learning_rate": 6.490900059214836e-05,
      "loss": 1.2952,
      "step": 8230
    },
    {
      "epoch": 0.869198312236287,
      "grad_norm": 0.6321101188659668,
      "learning_rate": 6.389117762692952e-05,
      "loss": 1.2916,
      "step": 8240
    },
    {
      "epoch": 0.870253164556962,
      "grad_norm": 0.6617045998573303,
      "learning_rate": 6.288104284608284e-05,
      "loss": 1.3058,
      "step": 8250
    },
    {
      "epoch": 0.8713080168776371,
      "grad_norm": 0.7012975215911865,
      "learning_rate": 6.187860756866157e-05,
      "loss": 1.2969,
      "step": 8260
    },
    {
      "epoch": 0.8723628691983122,
      "grad_norm": 0.6854604482650757,
      "learning_rate": 6.088388302744266e-05,
      "loss": 1.2942,
      "step": 8270
    },
    {
      "epoch": 0.8734177215189873,
      "grad_norm": 0.6472890973091125,
      "learning_rate": 5.9896880368800115e-05,
      "loss": 1.2779,
      "step": 8280
    },
    {
      "epoch": 0.8744725738396625,
      "grad_norm": 0.6605722904205322,
      "learning_rate": 5.891761065258089e-05,
      "loss": 1.2792,
      "step": 8290
    },
    {
      "epoch": 0.8755274261603375,
      "grad_norm": 0.6920452117919922,
      "learning_rate": 5.794608485198008e-05,
      "loss": 1.3041,
      "step": 8300
    },
    {
      "epoch": 0.8765822784810127,
      "grad_norm": 0.7682325839996338,
      "learning_rate": 5.698231385341887e-05,
      "loss": 1.2819,
      "step": 8310
    },
    {
      "epoch": 0.8776371308016878,
      "grad_norm": 0.7030491828918457,
      "learning_rate": 5.60263084564217e-05,
      "loss": 1.3004,
      "step": 8320
    },
    {
      "epoch": 0.8786919831223629,
      "grad_norm": 0.6490678787231445,
      "learning_rate": 5.507807937349604e-05,
      "loss": 1.2936,
      "step": 8330
    },
    {
      "epoch": 0.879746835443038,
      "grad_norm": 0.6462332010269165,
      "learning_rate": 5.413763723001164e-05,
      "loss": 1.2805,
      "step": 8340
    },
    {
      "epoch": 0.880801687763713,
      "grad_norm": 0.645033061504364,
      "learning_rate": 5.320499256408204e-05,
      "loss": 1.2853,
      "step": 8350
    },
    {
      "epoch": 0.8818565400843882,
      "grad_norm": 0.6679040789604187,
      "learning_rate": 5.228015582644585e-05,
      "loss": 1.2823,
      "step": 8360
    },
    {
      "epoch": 0.8829113924050633,
      "grad_norm": 0.6407747268676758,
      "learning_rate": 5.136313738035059e-05,
      "loss": 1.2916,
      "step": 8370
    },
    {
      "epoch": 0.8839662447257384,
      "grad_norm": 0.6732246279716492,
      "learning_rate": 5.045394750143567e-05,
      "loss": 1.2967,
      "step": 8380
    },
    {
      "epoch": 0.8850210970464135,
      "grad_norm": 0.7067869901657104,
      "learning_rate": 4.955259637761761e-05,
      "loss": 1.2874,
      "step": 8390
    },
    {
      "epoch": 0.8860759493670886,
      "grad_norm": 0.675907552242279,
      "learning_rate": 4.865909410897576e-05,
      "loss": 1.2868,
      "step": 8400
    },
    {
      "epoch": 0.8871308016877637,
      "grad_norm": 0.6973927021026611,
      "learning_rate": 4.7773450707639414e-05,
      "loss": 1.3052,
      "step": 8410
    },
    {
      "epoch": 0.8881856540084389,
      "grad_norm": 0.6466608643531799,
      "learning_rate": 4.6895676097675225e-05,
      "loss": 1.2947,
      "step": 8420
    },
    {
      "epoch": 0.8892405063291139,
      "grad_norm": 0.6730669736862183,
      "learning_rate": 4.6025780114976545e-05,
      "loss": 1.2911,
      "step": 8430
    },
    {
      "epoch": 0.890295358649789,
      "grad_norm": 0.6851930618286133,
      "learning_rate": 4.5163772507152425e-05,
      "loss": 1.2837,
      "step": 8440
    },
    {
      "epoch": 0.8913502109704642,
      "grad_norm": 0.7445343732833862,
      "learning_rate": 4.430966293341912e-05,
      "loss": 1.2903,
      "step": 8450
    },
    {
      "epoch": 0.8924050632911392,
      "grad_norm": 0.6899713277816772,
      "learning_rate": 4.346346096449136e-05,
      "loss": 1.2882,
      "step": 8460
    },
    {
      "epoch": 0.8934599156118144,
      "grad_norm": 0.6613866686820984,
      "learning_rate": 4.26251760824754e-05,
      "loss": 1.29,
      "step": 8470
    },
    {
      "epoch": 0.8945147679324894,
      "grad_norm": 0.6546576619148254,
      "learning_rate": 4.179481768076274e-05,
      "loss": 1.2821,
      "step": 8480
    },
    {
      "epoch": 0.8955696202531646,
      "grad_norm": 0.6866971254348755,
      "learning_rate": 4.0972395063924554e-05,
      "loss": 1.2879,
      "step": 8490
    },
    {
      "epoch": 0.8966244725738397,
      "grad_norm": 0.6406285762786865,
      "learning_rate": 4.015791744760811e-05,
      "loss": 1.2798,
      "step": 8500
    },
    {
      "epoch": 0.8976793248945147,
      "grad_norm": 0.6423125267028809,
      "learning_rate": 3.93513939584326e-05,
      "loss": 1.2911,
      "step": 8510
    },
    {
      "epoch": 0.8987341772151899,
      "grad_norm": 0.6347548961639404,
      "learning_rate": 3.855283363388762e-05,
      "loss": 1.2851,
      "step": 8520
    },
    {
      "epoch": 0.8997890295358649,
      "grad_norm": 0.6761476397514343,
      "learning_rate": 3.7762245422231476e-05,
      "loss": 1.3001,
      "step": 8530
    },
    {
      "epoch": 0.9008438818565401,
      "grad_norm": 0.636894941329956,
      "learning_rate": 3.697963818239117e-05,
      "loss": 1.2959,
      "step": 8540
    },
    {
      "epoch": 0.9018987341772152,
      "grad_norm": 0.6979252099990845,
      "learning_rate": 3.6205020683862836e-05,
      "loss": 1.2869,
      "step": 8550
    },
    {
      "epoch": 0.9029535864978903,
      "grad_norm": 0.671042263507843,
      "learning_rate": 3.543840160661396e-05,
      "loss": 1.2785,
      "step": 8560
    },
    {
      "epoch": 0.9040084388185654,
      "grad_norm": 0.7815701961517334,
      "learning_rate": 3.467978954098549e-05,
      "loss": 1.2979,
      "step": 8570
    },
    {
      "epoch": 0.9050632911392406,
      "grad_norm": 0.6872392892837524,
      "learning_rate": 3.392919298759623e-05,
      "loss": 1.2876,
      "step": 8580
    },
    {
      "epoch": 0.9061181434599156,
      "grad_norm": 0.7770042419433594,
      "learning_rate": 3.318662035724679e-05,
      "loss": 1.2938,
      "step": 8590
    },
    {
      "epoch": 0.9071729957805907,
      "grad_norm": 0.6594602465629578,
      "learning_rate": 3.2452079970826335e-05,
      "loss": 1.2852,
      "step": 8600
    },
    {
      "epoch": 0.9082278481012658,
      "grad_norm": 0.6838461756706238,
      "learning_rate": 3.172558005921841e-05,
      "loss": 1.2876,
      "step": 8610
    },
    {
      "epoch": 0.9092827004219409,
      "grad_norm": 0.7709410190582275,
      "learning_rate": 3.100712876320924e-05,
      "loss": 1.2833,
      "step": 8620
    },
    {
      "epoch": 0.9103375527426161,
      "grad_norm": 0.6450920701026917,
      "learning_rate": 3.029673413339651e-05,
      "loss": 1.2933,
      "step": 8630
    },
    {
      "epoch": 0.9113924050632911,
      "grad_norm": 0.7203512191772461,
      "learning_rate": 2.959440413009895e-05,
      "loss": 1.2753,
      "step": 8640
    },
    {
      "epoch": 0.9124472573839663,
      "grad_norm": 0.6634292006492615,
      "learning_rate": 2.890014662326701e-05,
      "loss": 1.2788,
      "step": 8650
    },
    {
      "epoch": 0.9135021097046413,
      "grad_norm": 0.7300726771354675,
      "learning_rate": 2.8213969392395233e-05,
      "loss": 1.2803,
      "step": 8660
    },
    {
      "epoch": 0.9145569620253164,
      "grad_norm": 0.6458709836006165,
      "learning_rate": 2.7535880126434433e-05,
      "loss": 1.2883,
      "step": 8670
    },
    {
      "epoch": 0.9156118143459916,
      "grad_norm": 0.6189814805984497,
      "learning_rate": 2.686588642370591e-05,
      "loss": 1.2965,
      "step": 8680
    },
    {
      "epoch": 0.9166666666666666,
      "grad_norm": 0.6375811100006104,
      "learning_rate": 2.6203995791816372e-05,
      "loss": 1.2885,
      "step": 8690
    },
    {
      "epoch": 0.9177215189873418,
      "grad_norm": 0.6480770707130432,
      "learning_rate": 2.5550215647573482e-05,
      "loss": 1.2895,
      "step": 8700
    },
    {
      "epoch": 0.9187763713080169,
      "grad_norm": 0.6383376121520996,
      "learning_rate": 2.490455331690303e-05,
      "loss": 1.2849,
      "step": 8710
    },
    {
      "epoch": 0.919831223628692,
      "grad_norm": 0.6496900320053101,
      "learning_rate": 2.4267016034766637e-05,
      "loss": 1.2814,
      "step": 8720
    },
    {
      "epoch": 0.9208860759493671,
      "grad_norm": 0.6400648355484009,
      "learning_rate": 2.363761094508085e-05,
      "loss": 1.2928,
      "step": 8730
    },
    {
      "epoch": 0.9219409282700421,
      "grad_norm": 0.6740437150001526,
      "learning_rate": 2.301634510063702e-05,
      "loss": 1.2831,
      "step": 8740
    },
    {
      "epoch": 0.9229957805907173,
      "grad_norm": 0.6760841608047485,
      "learning_rate": 2.2403225463022288e-05,
      "loss": 1.2797,
      "step": 8750
    },
    {
      "epoch": 0.9240506329113924,
      "grad_norm": 0.7134016156196594,
      "learning_rate": 2.1798258902541723e-05,
      "loss": 1.269,
      "step": 8760
    },
    {
      "epoch": 0.9251054852320675,
      "grad_norm": 0.626210629940033,
      "learning_rate": 2.120145219814082e-05,
      "loss": 1.2793,
      "step": 8770
    },
    {
      "epoch": 0.9261603375527426,
      "grad_norm": 0.674151599407196,
      "learning_rate": 2.0612812037330202e-05,
      "loss": 1.2868,
      "step": 8780
    },
    {
      "epoch": 0.9272151898734177,
      "grad_norm": 0.6493917107582092,
      "learning_rate": 2.003234501611037e-05,
      "loss": 1.2798,
      "step": 8790
    },
    {
      "epoch": 0.9282700421940928,
      "grad_norm": 0.652468740940094,
      "learning_rate": 1.9460057638897578e-05,
      "loss": 1.2799,
      "step": 8800
    },
    {
      "epoch": 0.929324894514768,
      "grad_norm": 0.6398833394050598,
      "learning_rate": 1.8895956318451398e-05,
      "loss": 1.297,
      "step": 8810
    },
    {
      "epoch": 0.930379746835443,
      "grad_norm": 0.6292800903320312,
      "learning_rate": 1.8340047375802693e-05,
      "loss": 1.2827,
      "step": 8820
    },
    {
      "epoch": 0.9314345991561181,
      "grad_norm": 0.6489097476005554,
      "learning_rate": 1.7792337040182434e-05,
      "loss": 1.2999,
      "step": 8830
    },
    {
      "epoch": 0.9324894514767933,
      "grad_norm": 0.6469489336013794,
      "learning_rate": 1.72528314489524e-05,
      "loss": 1.2852,
      "step": 8840
    },
    {
      "epoch": 0.9335443037974683,
      "grad_norm": 0.7080603837966919,
      "learning_rate": 1.6721536647536255e-05,
      "loss": 1.2844,
      "step": 8850
    },
    {
      "epoch": 0.9345991561181435,
      "grad_norm": 0.7085657119750977,
      "learning_rate": 1.6198458589351595e-05,
      "loss": 1.2844,
      "step": 8860
    },
    {
      "epoch": 0.9356540084388185,
      "grad_norm": 0.6312565207481384,
      "learning_rate": 1.568360313574349e-05,
      "loss": 1.2871,
      "step": 8870
    },
    {
      "epoch": 0.9367088607594937,
      "grad_norm": 0.6665783524513245,
      "learning_rate": 1.517697605591864e-05,
      "loss": 1.2884,
      "step": 8880
    },
    {
      "epoch": 0.9377637130801688,
      "grad_norm": 0.6456872820854187,
      "learning_rate": 1.4678583026880993e-05,
      "loss": 1.2891,
      "step": 8890
    },
    {
      "epoch": 0.9388185654008439,
      "grad_norm": 0.6229240298271179,
      "learning_rate": 1.4188429633367721e-05,
      "loss": 1.2797,
      "step": 8900
    },
    {
      "epoch": 0.939873417721519,
      "grad_norm": 0.6531126499176025,
      "learning_rate": 1.370652136778694e-05,
      "loss": 1.2781,
      "step": 8910
    },
    {
      "epoch": 0.9409282700421941,
      "grad_norm": 0.6330586671829224,
      "learning_rate": 1.3232863630156077e-05,
      "loss": 1.2872,
      "step": 8920
    },
    {
      "epoch": 0.9419831223628692,
      "grad_norm": 0.6318323016166687,
      "learning_rate": 1.2767461728041357e-05,
      "loss": 1.2819,
      "step": 8930
    },
    {
      "epoch": 0.9430379746835443,
      "grad_norm": 0.6362433433532715,
      "learning_rate": 1.2310320876498333e-05,
      "loss": 1.2853,
      "step": 8940
    },
    {
      "epoch": 0.9440928270042194,
      "grad_norm": 0.6001209616661072,
      "learning_rate": 1.186144619801352e-05,
      "loss": 1.2858,
      "step": 8950
    },
    {
      "epoch": 0.9451476793248945,
      "grad_norm": 0.6504714488983154,
      "learning_rate": 1.14208427224467e-05,
      "loss": 1.2883,
      "step": 8960
    },
    {
      "epoch": 0.9462025316455697,
      "grad_norm": 0.7938284873962402,
      "learning_rate": 1.0988515386975206e-05,
      "loss": 1.2853,
      "step": 8970
    },
    {
      "epoch": 0.9472573839662447,
      "grad_norm": 0.7599204182624817,
      "learning_rate": 1.0564469036037722e-05,
      "loss": 1.2822,
      "step": 8980
    },
    {
      "epoch": 0.9483122362869199,
      "grad_norm": 0.6392332911491394,
      "learning_rate": 1.0148708421280822e-05,
      "loss": 1.2866,
      "step": 8990
    },
    {
      "epoch": 0.9493670886075949,
      "grad_norm": 0.6156899929046631,
      "learning_rate": 9.74123820150502e-06,
      "loss": 1.2804,
      "step": 9000
    },
    {
      "epoch": 0.95042194092827,
      "grad_norm": 0.7023313641548157,
      "learning_rate": 9.342062942613222e-06,
      "loss": 1.277,
      "step": 9010
    },
    {
      "epoch": 0.9514767932489452,
      "grad_norm": 0.6554434895515442,
      "learning_rate": 8.9511871175591e-06,
      "loss": 1.2802,
      "step": 9020
    },
    {
      "epoch": 0.9525316455696202,
      "grad_norm": 0.6764358878135681,
      "learning_rate": 8.568615106297223e-06,
      "loss": 1.2984,
      "step": 9030
    },
    {
      "epoch": 0.9535864978902954,
      "grad_norm": 0.664933443069458,
      "learning_rate": 8.194351195733585e-06,
      "loss": 1.2756,
      "step": 9040
    },
    {
      "epoch": 0.9546413502109705,
      "grad_norm": 0.6336010694503784,
      "learning_rate": 7.828399579678153e-06,
      "loss": 1.2864,
      "step": 9050
    },
    {
      "epoch": 0.9556962025316456,
      "grad_norm": 0.6007170081138611,
      "learning_rate": 7.470764358797566e-06,
      "loss": 1.2877,
      "step": 9060
    },
    {
      "epoch": 0.9567510548523207,
      "grad_norm": 0.6316767334938049,
      "learning_rate": 7.121449540568842e-06,
      "loss": 1.288,
      "step": 9070
    },
    {
      "epoch": 0.9578059071729957,
      "grad_norm": 0.6550620198249817,
      "learning_rate": 6.780459039235409e-06,
      "loss": 1.2899,
      "step": 9080
    },
    {
      "epoch": 0.9588607594936709,
      "grad_norm": 0.6269824504852295,
      "learning_rate": 6.447796675762146e-06,
      "loss": 1.2903,
      "step": 9090
    },
    {
      "epoch": 0.959915611814346,
      "grad_norm": 0.6472267508506775,
      "learning_rate": 6.123466177793247e-06,
      "loss": 1.2686,
      "step": 9100
    },
    {
      "epoch": 0.9609704641350211,
      "grad_norm": 0.6144204139709473,
      "learning_rate": 5.807471179610418e-06,
      "loss": 1.2793,
      "step": 9110
    },
    {
      "epoch": 0.9620253164556962,
      "grad_norm": 0.6408663988113403,
      "learning_rate": 5.499815222091836e-06,
      "loss": 1.2911,
      "step": 9120
    },
    {
      "epoch": 0.9630801687763713,
      "grad_norm": 0.6615253686904907,
      "learning_rate": 5.200501752672754e-06,
      "loss": 1.2771,
      "step": 9130
    },
    {
      "epoch": 0.9641350210970464,
      "grad_norm": 0.6785182356834412,
      "learning_rate": 4.909534125306702e-06,
      "loss": 1.2934,
      "step": 9140
    },
    {
      "epoch": 0.9651898734177216,
      "grad_norm": 0.6627257466316223,
      "learning_rate": 4.626915600428105e-06,
      "loss": 1.2817,
      "step": 9150
    },
    {
      "epoch": 0.9662447257383966,
      "grad_norm": 0.6076435446739197,
      "learning_rate": 4.352649344915471e-06,
      "loss": 1.288,
      "step": 9160
    },
    {
      "epoch": 0.9672995780590717,
      "grad_norm": 0.6198477745056152,
      "learning_rate": 4.086738432056092e-06,
      "loss": 1.2865,
      "step": 9170
    },
    {
      "epoch": 0.9683544303797469,
      "grad_norm": 0.623679518699646,
      "learning_rate": 3.8291858415117344e-06,
      "loss": 1.2867,
      "step": 9180
    },
    {
      "epoch": 0.9694092827004219,
      "grad_norm": 0.6261500716209412,
      "learning_rate": 3.579994459284752e-06,
      "loss": 1.2793,
      "step": 9190
    },
    {
      "epoch": 0.9704641350210971,
      "grad_norm": 0.6912320256233215,
      "learning_rate": 3.339167077686278e-06,
      "loss": 1.2832,
      "step": 9200
    },
    {
      "epoch": 0.9715189873417721,
      "grad_norm": 0.6676372289657593,
      "learning_rate": 3.1067063953048313e-06,
      "loss": 1.2831,
      "step": 9210
    },
    {
      "epoch": 0.9725738396624473,
      "grad_norm": 0.6578547358512878,
      "learning_rate": 2.8826150169758425e-06,
      "loss": 1.2809,
      "step": 9220
    },
    {
      "epoch": 0.9736286919831224,
      "grad_norm": 0.6263319253921509,
      "learning_rate": 2.66689545375251e-06,
      "loss": 1.2739,
      "step": 9230
    },
    {
      "epoch": 0.9746835443037974,
      "grad_norm": 0.6633256673812866,
      "learning_rate": 2.4595501228779906e-06,
      "loss": 1.265,
      "step": 9240
    },
    {
      "epoch": 0.9757383966244726,
      "grad_norm": 0.6043865084648132,
      "learning_rate": 2.2605813477579172e-06,
      "loss": 1.2813,
      "step": 9250
    },
    {
      "epoch": 0.9767932489451476,
      "grad_norm": 0.6398628950119019,
      "learning_rate": 2.069991357934592e-06,
      "loss": 1.2751,
      "step": 9260
    },
    {
      "epoch": 0.9778481012658228,
      "grad_norm": 0.653148889541626,
      "learning_rate": 1.8877822890618346e-06,
      "loss": 1.2876,
      "step": 9270
    },
    {
      "epoch": 0.9789029535864979,
      "grad_norm": 0.6248723268508911,
      "learning_rate": 1.7139561828813377e-06,
      "loss": 1.2757,
      "step": 9280
    },
    {
      "epoch": 0.979957805907173,
      "grad_norm": 0.6223885416984558,
      "learning_rate": 1.5485149871995175e-06,
      "loss": 1.2816,
      "step": 9290
    },
    {
      "epoch": 0.9810126582278481,
      "grad_norm": 0.6285667419433594,
      "learning_rate": 1.3914605558656146e-06,
      "loss": 1.2819,
      "step": 9300
    },
    {
      "epoch": 0.9820675105485233,
      "grad_norm": 0.658174991607666,
      "learning_rate": 1.2427946487512941e-06,
      "loss": 1.2816,
      "step": 9310
    },
    {
      "epoch": 0.9831223628691983,
      "grad_norm": 0.6376784443855286,
      "learning_rate": 1.1025189317305784e-06,
      "loss": 1.2771,
      "step": 9320
    },
    {
      "epoch": 0.9841772151898734,
      "grad_norm": 0.6339856386184692,
      "learning_rate": 9.706349766615275e-07,
      "loss": 1.2896,
      "step": 9330
    },
    {
      "epoch": 0.9852320675105485,
      "grad_norm": 0.6104322671890259,
      "learning_rate": 8.47144261368088e-07,
      "loss": 1.2852,
      "step": 9340
    },
    {
      "epoch": 0.9862869198312236,
      "grad_norm": 0.6234305500984192,
      "learning_rate": 7.320481696241887e-07,
      "loss": 1.2957,
      "step": 9350
    },
    {
      "epoch": 0.9873417721518988,
      "grad_norm": 0.652967631816864,
      "learning_rate": 6.253479911375037e-07,
      "loss": 1.2872,
      "step": 9360
    },
    {
      "epoch": 0.9883966244725738,
      "grad_norm": 0.6444950699806213,
      "learning_rate": 5.270449215358797e-07,
      "loss": 1.2873,
      "step": 9370
    },
    {
      "epoch": 0.989451476793249,
      "grad_norm": 0.6644588112831116,
      "learning_rate": 4.371400623530142e-07,
      "loss": 1.2727,
      "step": 9380
    },
    {
      "epoch": 0.990506329113924,
      "grad_norm": 0.6382048726081848,
      "learning_rate": 3.5563442101696486e-07,
      "loss": 1.278,
      "step": 9390
    },
    {
      "epoch": 0.9915611814345991,
      "grad_norm": 0.6332500576972961,
      "learning_rate": 2.825289108379925e-07,
      "loss": 1.292,
      "step": 9400
    },
    {
      "epoch": 0.9926160337552743,
      "grad_norm": 0.6931483149528503,
      "learning_rate": 2.1782435099923503e-07,
      "loss": 1.2669,
      "step": 9410
    },
    {
      "epoch": 0.9936708860759493,
      "grad_norm": 0.6163036227226257,
      "learning_rate": 1.6152146654671573e-07,
      "loss": 1.2862,
      "step": 9420
    },
    {
      "epoch": 0.9947257383966245,
      "grad_norm": 0.638114333152771,
      "learning_rate": 1.1362088838193229e-07,
      "loss": 1.2858,
      "step": 9430
    },
    {
      "epoch": 0.9957805907172996,
      "grad_norm": 0.6300340294837952,
      "learning_rate": 7.412315325411312e-08,
      "loss": 1.2823,
      "step": 9440
    },
    {
      "epoch": 0.9968354430379747,
      "grad_norm": 0.6341334581375122,
      "learning_rate": 4.302870375472168e-08,
      "loss": 1.2922,
      "step": 9450
    },
    {
      "epoch": 0.9978902953586498,
      "grad_norm": 0.6342418789863586,
      "learning_rate": 2.0337888312210727e-08,
      "loss": 1.2737,
      "step": 9460
    },
    {
      "epoch": 0.9989451476793249,
      "grad_norm": 0.6190896034240723,
      "learning_rate": 6.050961188358573e-09,
      "loss": 1.2885,
      "step": 9470
    },
    {
      "epoch": 1.0,
      "grad_norm": 1.9135394096374512,
      "learning_rate": 1.6808247493838026e-10,
      "loss": 1.2788,
      "step": 9480
    }
  ],
  "logging_steps": 10,
  "max_steps": 9480,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 1000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.832308198648013e+16,
  "train_batch_size": 1024,
  "trial_name": null,
  "trial_params": null
}