{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.0648508430609598,
  "eval_steps": 500,
  "global_step": 2500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00025940337224383917,
      "grad_norm": 13.49365234375,
      "learning_rate": 1.6000000000000003e-05,
      "loss": 1.5584,
      "step": 10
    },
    {
      "epoch": 0.0005188067444876783,
      "grad_norm": 2.6006040573120117,
      "learning_rate": 3.2000000000000005e-05,
      "loss": 0.7111,
      "step": 20
    },
    {
      "epoch": 0.0007782101167315176,
      "grad_norm": 4.435553073883057,
      "learning_rate": 4.8e-05,
      "loss": 0.3076,
      "step": 30
    },
    {
      "epoch": 0.0010376134889753567,
      "grad_norm": 1.1146177053451538,
      "learning_rate": 6.400000000000001e-05,
      "loss": 0.1826,
      "step": 40
    },
    {
      "epoch": 0.0012970168612191958,
      "grad_norm": 0.8707008957862854,
      "learning_rate": 8e-05,
      "loss": 0.1341,
      "step": 50
    },
    {
      "epoch": 0.0015564202334630351,
      "grad_norm": 1.7411352396011353,
      "learning_rate": 9.6e-05,
      "loss": 0.1144,
      "step": 60
    },
    {
      "epoch": 0.0018158236057068742,
      "grad_norm": 0.9691908955574036,
      "learning_rate": 0.00011200000000000001,
      "loss": 0.0831,
      "step": 70
    },
    {
      "epoch": 0.0020752269779507134,
      "grad_norm": 1.0398532152175903,
      "learning_rate": 0.00012800000000000002,
      "loss": 0.0801,
      "step": 80
    },
    {
      "epoch": 0.0023346303501945525,
      "grad_norm": 0.5252112746238708,
      "learning_rate": 0.000144,
      "loss": 0.081,
      "step": 90
    },
    {
      "epoch": 0.0025940337224383916,
      "grad_norm": 0.6552278995513916,
      "learning_rate": 0.00016,
      "loss": 0.0821,
      "step": 100
    },
    {
      "epoch": 0.0028534370946822307,
      "grad_norm": 0.6302046775817871,
      "learning_rate": 0.00017600000000000002,
      "loss": 0.0749,
      "step": 110
    },
    {
      "epoch": 0.0031128404669260703,
      "grad_norm": 0.7439587712287903,
      "learning_rate": 0.000192,
      "loss": 0.0647,
      "step": 120
    },
    {
      "epoch": 0.0033722438391699094,
      "grad_norm": 6.8584771156311035,
      "learning_rate": 0.0001999978128380225,
      "loss": 0.0726,
      "step": 130
    },
    {
      "epoch": 0.0036316472114137485,
      "grad_norm": 2.014613389968872,
      "learning_rate": 0.0001999803161162393,
      "loss": 0.0936,
      "step": 140
    },
    {
      "epoch": 0.0038910505836575876,
      "grad_norm": 1.3758050203323364,
      "learning_rate": 0.00019994532573409262,
      "loss": 0.1089,
      "step": 150
    },
    {
      "epoch": 0.004150453955901427,
      "grad_norm": 1.3027769327163696,
      "learning_rate": 0.00019989284781388617,
      "loss": 0.0772,
      "step": 160
    },
    {
      "epoch": 0.004409857328145266,
      "grad_norm": 1.021938443183899,
      "learning_rate": 0.00019982289153773646,
      "loss": 0.0695,
      "step": 170
    },
    {
      "epoch": 0.004669260700389105,
      "grad_norm": 0.8069638609886169,
      "learning_rate": 0.00019973546914596623,
      "loss": 0.0624,
      "step": 180
    },
    {
      "epoch": 0.004928664072632944,
      "grad_norm": 0.5198134779930115,
      "learning_rate": 0.00019963059593496268,
      "loss": 0.0481,
      "step": 190
    },
    {
      "epoch": 0.005188067444876783,
      "grad_norm": 0.39965012669563293,
      "learning_rate": 0.00019950829025450114,
      "loss": 0.0493,
      "step": 200
    },
    {
      "epoch": 0.005447470817120622,
      "grad_norm": 0.8152652978897095,
      "learning_rate": 0.0001993685735045343,
      "loss": 0.0477,
      "step": 210
    },
    {
      "epoch": 0.005706874189364461,
      "grad_norm": 0.7083427309989929,
      "learning_rate": 0.0001992114701314478,
      "loss": 0.0501,
      "step": 220
    },
    {
      "epoch": 0.0059662775616083005,
      "grad_norm": 0.690096914768219,
      "learning_rate": 0.000199037007623783,
      "loss": 0.0509,
      "step": 230
    },
    {
      "epoch": 0.0062256809338521405,
      "grad_norm": 0.871699869632721,
      "learning_rate": 0.00019884521650742715,
      "loss": 0.0464,
      "step": 240
    },
    {
      "epoch": 0.00648508430609598,
      "grad_norm": 0.6445639133453369,
      "learning_rate": 0.00019863613034027224,
      "loss": 0.0462,
      "step": 250
    },
    {
      "epoch": 0.006744487678339819,
      "grad_norm": 0.4920782148838043,
      "learning_rate": 0.0001984097857063434,
      "loss": 0.0425,
      "step": 260
    },
    {
      "epoch": 0.007003891050583658,
      "grad_norm": 0.8624788522720337,
      "learning_rate": 0.0001981662222093976,
      "loss": 0.0471,
      "step": 270
    },
    {
      "epoch": 0.007263294422827497,
      "grad_norm": 0.5704030990600586,
      "learning_rate": 0.00019790548246599447,
      "loss": 0.0571,
      "step": 280
    },
    {
      "epoch": 0.007522697795071336,
      "grad_norm": 0.569247841835022,
      "learning_rate": 0.00019762761209803927,
      "loss": 0.0531,
      "step": 290
    },
    {
      "epoch": 0.007782101167315175,
      "grad_norm": 0.7414583563804626,
      "learning_rate": 0.0001973326597248006,
      "loss": 0.0455,
      "step": 300
    },
    {
      "epoch": 0.008041504539559013,
      "grad_norm": 0.42905348539352417,
      "learning_rate": 0.00019702067695440332,
      "loss": 0.0542,
      "step": 310
    },
    {
      "epoch": 0.008300907911802853,
      "grad_norm": 0.6132651567459106,
      "learning_rate": 0.00019669171837479873,
      "loss": 0.0439,
      "step": 320
    },
    {
      "epoch": 0.008560311284046693,
      "grad_norm": 0.44733384251594543,
      "learning_rate": 0.00019634584154421317,
      "loss": 0.0429,
      "step": 330
    },
    {
      "epoch": 0.008819714656290532,
      "grad_norm": 0.516659677028656,
      "learning_rate": 0.00019598310698107702,
      "loss": 0.0407,
      "step": 340
    },
    {
      "epoch": 0.009079118028534372,
      "grad_norm": 0.2340136468410492,
      "learning_rate": 0.00019560357815343577,
      "loss": 0.036,
      "step": 350
    },
    {
      "epoch": 0.00933852140077821,
      "grad_norm": 0.6629723906517029,
      "learning_rate": 0.00019520732146784491,
      "loss": 0.041,
      "step": 360
    },
    {
      "epoch": 0.00959792477302205,
      "grad_norm": 0.3777199387550354,
      "learning_rate": 0.0001947944062577507,
      "loss": 0.0414,
      "step": 370
    },
    {
      "epoch": 0.009857328145265888,
      "grad_norm": 0.44298338890075684,
      "learning_rate": 0.00019436490477135878,
      "loss": 0.0371,
      "step": 380
    },
    {
      "epoch": 0.010116731517509728,
      "grad_norm": 0.4530090093612671,
      "learning_rate": 0.00019391889215899299,
      "loss": 0.037,
      "step": 390
    },
    {
      "epoch": 0.010376134889753566,
      "grad_norm": 0.36142367124557495,
      "learning_rate": 0.0001934564464599461,
      "loss": 0.0387,
      "step": 400
    },
    {
      "epoch": 0.010635538261997406,
      "grad_norm": 0.36559927463531494,
      "learning_rate": 0.00019297764858882514,
      "loss": 0.0386,
      "step": 410
    },
    {
      "epoch": 0.010894941634241245,
      "grad_norm": 0.47660204768180847,
      "learning_rate": 0.00019248258232139388,
      "loss": 0.0316,
      "step": 420
    },
    {
      "epoch": 0.011154345006485085,
      "grad_norm": 0.2614981234073639,
      "learning_rate": 0.00019197133427991436,
      "loss": 0.0417,
      "step": 430
    },
    {
      "epoch": 0.011413748378728923,
      "grad_norm": 0.44811561703681946,
      "learning_rate": 0.00019144399391799043,
      "loss": 0.036,
      "step": 440
    },
    {
      "epoch": 0.011673151750972763,
      "grad_norm": 0.4213508367538452,
      "learning_rate": 0.00019090065350491626,
      "loss": 0.0399,
      "step": 450
    },
    {
      "epoch": 0.011932555123216601,
      "grad_norm": 0.29146406054496765,
      "learning_rate": 0.0001903414081095315,
      "loss": 0.0376,
      "step": 460
    },
    {
      "epoch": 0.012191958495460441,
      "grad_norm": 0.2714293897151947,
      "learning_rate": 0.00018976635558358722,
      "loss": 0.0362,
      "step": 470
    },
    {
      "epoch": 0.012451361867704281,
      "grad_norm": 0.3767367899417877,
      "learning_rate": 0.00018917559654462474,
      "loss": 0.0364,
      "step": 480
    },
    {
      "epoch": 0.01271076523994812,
      "grad_norm": 0.36578992009162903,
      "learning_rate": 0.00018856923435837022,
      "loss": 0.0327,
      "step": 490
    },
    {
      "epoch": 0.01297016861219196,
      "grad_norm": 0.32976096868515015,
      "learning_rate": 0.0001879473751206489,
      "loss": 0.04,
      "step": 500
    },
    {
      "epoch": 0.013229571984435798,
      "grad_norm": 0.4037308990955353,
      "learning_rate": 0.00018731012763882133,
      "loss": 0.0365,
      "step": 510
    },
    {
      "epoch": 0.013488975356679637,
      "grad_norm": 0.3599701225757599,
      "learning_rate": 0.00018665760341274505,
      "loss": 0.0337,
      "step": 520
    },
    {
      "epoch": 0.013748378728923476,
      "grad_norm": 0.6167535185813904,
      "learning_rate": 0.00018598991661526572,
      "loss": 0.0384,
      "step": 530
    },
    {
      "epoch": 0.014007782101167316,
      "grad_norm": 0.44012153148651123,
      "learning_rate": 0.00018530718407223974,
      "loss": 0.0291,
      "step": 540
    },
    {
      "epoch": 0.014267185473411154,
      "grad_norm": 0.5593993663787842,
      "learning_rate": 0.00018460952524209355,
      "loss": 0.0328,
      "step": 550
    },
    {
      "epoch": 0.014526588845654994,
      "grad_norm": 0.3970232307910919,
      "learning_rate": 0.00018389706219492147,
      "loss": 0.0282,
      "step": 560
    },
    {
      "epoch": 0.014785992217898832,
      "grad_norm": 0.2354394942522049,
      "learning_rate": 0.00018316991959112716,
      "loss": 0.0317,
      "step": 570
    },
    {
      "epoch": 0.015045395590142672,
      "grad_norm": 0.35121485590934753,
      "learning_rate": 0.00018242822465961176,
      "loss": 0.0328,
      "step": 580
    },
    {
      "epoch": 0.01530479896238651,
      "grad_norm": 0.4026910066604614,
      "learning_rate": 0.00018167210717551224,
      "loss": 0.0346,
      "step": 590
    },
    {
      "epoch": 0.01556420233463035,
      "grad_norm": 0.36058488488197327,
      "learning_rate": 0.00018090169943749476,
      "loss": 0.0356,
      "step": 600
    },
    {
      "epoch": 0.01582360570687419,
      "grad_norm": 0.43329063057899475,
      "learning_rate": 0.00018011713624460608,
      "loss": 0.039,
      "step": 610
    },
    {
      "epoch": 0.016083009079118027,
      "grad_norm": 0.5267623662948608,
      "learning_rate": 0.00017931855487268782,
      "loss": 0.034,
      "step": 620
    },
    {
      "epoch": 0.01634241245136187,
      "grad_norm": 0.4230927526950836,
      "learning_rate": 0.0001785060950503568,
      "loss": 0.031,
      "step": 630
    },
    {
      "epoch": 0.016601815823605707,
      "grad_norm": 0.3721458613872528,
      "learning_rate": 0.00017767989893455698,
      "loss": 0.0306,
      "step": 640
    },
    {
      "epoch": 0.016861219195849545,
      "grad_norm": 0.4158923029899597,
      "learning_rate": 0.00017684011108568592,
      "loss": 0.0343,
      "step": 650
    },
    {
      "epoch": 0.017120622568093387,
      "grad_norm": 0.42805808782577515,
      "learning_rate": 0.00017598687844230088,
      "loss": 0.0348,
      "step": 660
    },
    {
      "epoch": 0.017380025940337225,
      "grad_norm": 0.2532889246940613,
      "learning_rate": 0.00017512035029540885,
      "loss": 0.0348,
      "step": 670
    },
    {
      "epoch": 0.017639429312581063,
      "grad_norm": 0.33485907316207886,
      "learning_rate": 0.000174240678262345,
      "loss": 0.0382,
      "step": 680
    },
    {
      "epoch": 0.0178988326848249,
      "grad_norm": 0.33811262249946594,
      "learning_rate": 0.000173348016260244,
      "loss": 0.0321,
      "step": 690
    },
    {
      "epoch": 0.018158236057068743,
      "grad_norm": 0.5319349765777588,
      "learning_rate": 0.00017244252047910892,
      "loss": 0.0369,
      "step": 700
    },
    {
      "epoch": 0.01841763942931258,
      "grad_norm": 0.1967012882232666,
      "learning_rate": 0.00017152434935448256,
      "loss": 0.0392,
      "step": 710
    },
    {
      "epoch": 0.01867704280155642,
      "grad_norm": 0.491212397813797,
      "learning_rate": 0.0001705936635397259,
      "loss": 0.034,
      "step": 720
    },
    {
      "epoch": 0.018936446173800258,
      "grad_norm": 0.3200204074382782,
      "learning_rate": 0.00016965062587790823,
      "loss": 0.0349,
      "step": 730
    },
    {
      "epoch": 0.0191958495460441,
      "grad_norm": 0.29090484976768494,
      "learning_rate": 0.00016869540137331445,
      "loss": 0.0302,
      "step": 740
    },
    {
      "epoch": 0.019455252918287938,
      "grad_norm": 0.4633236825466156,
      "learning_rate": 0.00016772815716257412,
      "loss": 0.0305,
      "step": 750
    },
    {
      "epoch": 0.019714656290531776,
      "grad_norm": 0.6030534505844116,
      "learning_rate": 0.00016674906248541726,
      "loss": 0.0331,
      "step": 760
    },
    {
      "epoch": 0.019974059662775615,
      "grad_norm": 0.28692400455474854,
      "learning_rate": 0.00016575828865506245,
      "loss": 0.0319,
      "step": 770
    },
    {
      "epoch": 0.020233463035019456,
      "grad_norm": 0.3799605667591095,
      "learning_rate": 0.0001647560090282419,
      "loss": 0.0316,
      "step": 780
    },
    {
      "epoch": 0.020492866407263294,
      "grad_norm": 0.529679536819458,
      "learning_rate": 0.000163742398974869,
      "loss": 0.0317,
      "step": 790
    },
    {
      "epoch": 0.020752269779507133,
      "grad_norm": 0.328056275844574,
      "learning_rate": 0.0001627176358473537,
      "loss": 0.0332,
      "step": 800
    },
    {
      "epoch": 0.021011673151750974,
      "grad_norm": 0.3303374648094177,
      "learning_rate": 0.0001616818989495711,
      "loss": 0.0323,
      "step": 810
    },
    {
      "epoch": 0.021271076523994813,
      "grad_norm": 0.23507635295391083,
      "learning_rate": 0.00016063536950548826,
      "loss": 0.0338,
      "step": 820
    },
    {
      "epoch": 0.02153047989623865,
      "grad_norm": 0.3870477080345154,
      "learning_rate": 0.0001595782306274553,
      "loss": 0.0326,
      "step": 830
    },
    {
      "epoch": 0.02178988326848249,
      "grad_norm": 0.37890198826789856,
      "learning_rate": 0.00015851066728416618,
      "loss": 0.0309,
      "step": 840
    },
    {
      "epoch": 0.02204928664072633,
      "grad_norm": 0.48460447788238525,
      "learning_rate": 0.00015743286626829437,
      "loss": 0.0351,
      "step": 850
    },
    {
      "epoch": 0.02230869001297017,
      "grad_norm": 0.38216733932495117,
      "learning_rate": 0.00015634501616380967,
      "loss": 0.0331,
      "step": 860
    },
    {
      "epoch": 0.022568093385214007,
      "grad_norm": 0.31668373942375183,
      "learning_rate": 0.00015524730731298134,
      "loss": 0.0306,
      "step": 870
    },
    {
      "epoch": 0.022827496757457846,
      "grad_norm": 0.5256275534629822,
      "learning_rate": 0.0001541399317830738,
      "loss": 0.0307,
      "step": 880
    },
    {
      "epoch": 0.023086900129701687,
      "grad_norm": 0.2295175939798355,
      "learning_rate": 0.0001530230833327405,
      "loss": 0.0297,
      "step": 890
    },
    {
      "epoch": 0.023346303501945526,
      "grad_norm": 0.35704338550567627,
      "learning_rate": 0.00015189695737812152,
      "loss": 0.0358,
      "step": 900
    },
    {
      "epoch": 0.023605706874189364,
      "grad_norm": 0.31458932161331177,
      "learning_rate": 0.0001507617509586517,
      "loss": 0.0278,
      "step": 910
    },
    {
      "epoch": 0.023865110246433202,
      "grad_norm": 0.47554951906204224,
      "learning_rate": 0.00014961766270258422,
      "loss": 0.0332,
      "step": 920
    },
    {
      "epoch": 0.024124513618677044,
      "grad_norm": 0.29959654808044434,
      "learning_rate": 0.00014846489279223652,
      "loss": 0.0268,
      "step": 930
    },
    {
      "epoch": 0.024383916990920882,
      "grad_norm": 0.31856444478034973,
      "learning_rate": 0.0001473036429289641,
      "loss": 0.028,
      "step": 940
    },
    {
      "epoch": 0.02464332036316472,
      "grad_norm": 0.23568253219127655,
      "learning_rate": 0.0001461341162978688,
      "loss": 0.0339,
      "step": 950
    },
    {
      "epoch": 0.024902723735408562,
      "grad_norm": 0.4032520055770874,
      "learning_rate": 0.00014495651753224705,
      "loss": 0.0291,
      "step": 960
    },
    {
      "epoch": 0.0251621271076524,
      "grad_norm": 0.36155375838279724,
      "learning_rate": 0.00014377105267778518,
      "loss": 0.0261,
      "step": 970
    },
    {
      "epoch": 0.02542153047989624,
      "grad_norm": 0.2985226511955261,
      "learning_rate": 0.00014257792915650728,
      "loss": 0.03,
      "step": 980
    },
    {
      "epoch": 0.025680933852140077,
      "grad_norm": 0.3420771658420563,
      "learning_rate": 0.00014137735573048233,
      "loss": 0.0318,
      "step": 990
    },
    {
      "epoch": 0.02594033722438392,
      "grad_norm": 0.19557547569274902,
      "learning_rate": 0.00014016954246529696,
      "loss": 0.0287,
      "step": 1000
    },
    {
      "epoch": 0.026199740596627757,
      "grad_norm": 0.2249382734298706,
      "learning_rate": 0.00013895470069330004,
      "loss": 0.0258,
      "step": 1010
    },
    {
      "epoch": 0.026459143968871595,
      "grad_norm": 0.5154671669006348,
      "learning_rate": 0.00013773304297662559,
      "loss": 0.0339,
      "step": 1020
    },
    {
      "epoch": 0.026718547341115433,
      "grad_norm": 0.3469725251197815,
      "learning_rate": 0.00013650478307000057,
      "loss": 0.032,
      "step": 1030
    },
    {
      "epoch": 0.026977950713359275,
      "grad_norm": 0.34428727626800537,
      "learning_rate": 0.00013527013588334415,
      "loss": 0.0317,
      "step": 1040
    },
    {
      "epoch": 0.027237354085603113,
      "grad_norm": 0.2862760126590729,
      "learning_rate": 0.00013402931744416433,
      "loss": 0.03,
      "step": 1050
    },
    {
      "epoch": 0.02749675745784695,
      "grad_norm": 0.5235270261764526,
      "learning_rate": 0.00013278254485975976,
      "loss": 0.0368,
      "step": 1060
    },
    {
      "epoch": 0.02775616083009079,
      "grad_norm": 0.2167377918958664,
      "learning_rate": 0.00013153003627923218,
      "loss": 0.0306,
      "step": 1070
    },
    {
      "epoch": 0.02801556420233463,
      "grad_norm": 0.29817643761634827,
      "learning_rate": 0.00013027201085531634,
      "loss": 0.0301,
      "step": 1080
    },
    {
      "epoch": 0.02827496757457847,
      "grad_norm": 0.29515424370765686,
      "learning_rate": 0.00012900868870603503,
      "loss": 0.0249,
      "step": 1090
    },
    {
      "epoch": 0.028534370946822308,
      "grad_norm": 0.3012550473213196,
      "learning_rate": 0.00012774029087618446,
      "loss": 0.0264,
      "step": 1100
    },
    {
      "epoch": 0.028793774319066146,
      "grad_norm": 0.31644681096076965,
      "learning_rate": 0.00012646703929865817,
      "loss": 0.0316,
      "step": 1110
    },
    {
      "epoch": 0.029053177691309988,
      "grad_norm": 0.37952420115470886,
      "learning_rate": 0.00012518915675561483,
      "loss": 0.0296,
      "step": 1120
    },
    {
      "epoch": 0.029312581063553826,
      "grad_norm": 0.2594399154186249,
      "learning_rate": 0.00012390686683949798,
      "loss": 0.0321,
      "step": 1130
    },
    {
      "epoch": 0.029571984435797664,
      "grad_norm": 0.34381750226020813,
      "learning_rate": 0.00012262039391391404,
      "loss": 0.0305,
      "step": 1140
    },
    {
      "epoch": 0.029831387808041506,
      "grad_norm": 0.2528727948665619,
      "learning_rate": 0.0001213299630743747,
      "loss": 0.0303,
      "step": 1150
    },
    {
      "epoch": 0.030090791180285344,
      "grad_norm": 0.32746660709381104,
      "learning_rate": 0.00012003580010891213,
      "loss": 0.0309,
      "step": 1160
    },
    {
      "epoch": 0.030350194552529183,
      "grad_norm": 0.46516871452331543,
      "learning_rate": 0.00011873813145857249,
      "loss": 0.0288,
      "step": 1170
    },
    {
      "epoch": 0.03060959792477302,
      "grad_norm": 0.21779906749725342,
      "learning_rate": 0.00011743718417779517,
      "loss": 0.0268,
      "step": 1180
    },
    {
      "epoch": 0.030869001297016863,
      "grad_norm": 0.2870500087738037,
      "learning_rate": 0.00011613318589468511,
      "loss": 0.0284,
      "step": 1190
    },
    {
      "epoch": 0.0311284046692607,
      "grad_norm": 0.223075732588768,
      "learning_rate": 0.0001148263647711842,
      "loss": 0.0283,
      "step": 1200
    },
    {
      "epoch": 0.03138780804150454,
      "grad_norm": 0.3081727921962738,
      "learning_rate": 0.0001135169494631497,
      "loss": 0.0319,
      "step": 1210
    },
    {
      "epoch": 0.03164721141374838,
      "grad_norm": 0.48067420721054077,
      "learning_rate": 0.00011220516908034601,
      "loss": 0.0286,
      "step": 1220
    },
    {
      "epoch": 0.031906614785992216,
      "grad_norm": 0.34588056802749634,
      "learning_rate": 0.00011089125314635726,
      "loss": 0.0287,
      "step": 1230
    },
    {
      "epoch": 0.032166018158236054,
      "grad_norm": 0.3428352475166321,
      "learning_rate": 0.00010957543155842702,
      "loss": 0.025,
      "step": 1240
    },
    {
      "epoch": 0.0324254215304799,
      "grad_norm": 0.4223664700984955,
      "learning_rate": 0.00010825793454723325,
      "loss": 0.0304,
      "step": 1250
    },
    {
      "epoch": 0.03268482490272374,
      "grad_norm": 0.21455545723438263,
      "learning_rate": 0.00010693899263660441,
      "loss": 0.0265,
      "step": 1260
    },
    {
      "epoch": 0.032944228274967575,
      "grad_norm": 0.2936132252216339,
      "learning_rate": 0.00010561883660318455,
      "loss": 0.0298,
      "step": 1270
    },
    {
      "epoch": 0.033203631647211414,
      "grad_norm": 0.3222610056400299,
      "learning_rate": 0.00010429769743605407,
      "loss": 0.0333,
      "step": 1280
    },
    {
      "epoch": 0.03346303501945525,
      "grad_norm": 0.42949262261390686,
      "learning_rate": 0.00010297580629631325,
      "loss": 0.0336,
      "step": 1290
    },
    {
      "epoch": 0.03372243839169909,
      "grad_norm": 0.3103349208831787,
      "learning_rate": 0.00010165339447663587,
      "loss": 0.0327,
      "step": 1300
    },
    {
      "epoch": 0.03398184176394293,
      "grad_norm": 0.3449970483779907,
      "learning_rate": 0.00010033069336079952,
      "loss": 0.0293,
      "step": 1310
    },
    {
      "epoch": 0.034241245136186774,
      "grad_norm": 0.3841015100479126,
      "learning_rate": 9.900793438320037e-05,
      "loss": 0.0305,
      "step": 1320
    },
    {
      "epoch": 0.03450064850843061,
      "grad_norm": 0.3848475515842438,
      "learning_rate": 9.768534898835862e-05,
      "loss": 0.026,
      "step": 1330
    },
    {
      "epoch": 0.03476005188067445,
      "grad_norm": 0.43429145216941833,
      "learning_rate": 9.636316859042259e-05,
      "loss": 0.0294,
      "step": 1340
    },
    {
      "epoch": 0.03501945525291829,
      "grad_norm": 0.2181561291217804,
      "learning_rate": 9.504162453267777e-05,
      "loss": 0.0275,
      "step": 1350
    },
    {
      "epoch": 0.03527885862516213,
      "grad_norm": 0.4230787754058838,
      "learning_rate": 9.372094804706867e-05,
      "loss": 0.0273,
      "step": 1360
    },
    {
      "epoch": 0.035538261997405965,
      "grad_norm": 0.273333340883255,
      "learning_rate": 9.24013702137397e-05,
      "loss": 0.0246,
      "step": 1370
    },
    {
      "epoch": 0.0357976653696498,
      "grad_norm": 0.21888947486877441,
      "learning_rate": 9.108312192060298e-05,
      "loss": 0.0223,
      "step": 1380
    },
    {
      "epoch": 0.03605706874189364,
      "grad_norm": 0.22308865189552307,
      "learning_rate": 8.97664338229395e-05,
      "loss": 0.0249,
      "step": 1390
    },
    {
      "epoch": 0.03631647211413749,
      "grad_norm": 0.2270708829164505,
      "learning_rate": 8.845153630304139e-05,
      "loss": 0.0272,
      "step": 1400
    },
    {
      "epoch": 0.036575875486381325,
      "grad_norm": 0.3046322166919708,
      "learning_rate": 8.713865942990141e-05,
      "loss": 0.0225,
      "step": 1410
    },
    {
      "epoch": 0.03683527885862516,
      "grad_norm": 0.31552374362945557,
      "learning_rate": 8.582803291895758e-05,
      "loss": 0.0302,
      "step": 1420
    },
    {
      "epoch": 0.037094682230869,
      "grad_norm": 0.46200478076934814,
      "learning_rate": 8.451988609189987e-05,
      "loss": 0.0272,
      "step": 1430
    },
    {
      "epoch": 0.03735408560311284,
      "grad_norm": 0.22868406772613525,
      "learning_rate": 8.321444783654524e-05,
      "loss": 0.0267,
      "step": 1440
    },
    {
      "epoch": 0.03761348897535668,
      "grad_norm": 0.23030489683151245,
      "learning_rate": 8.191194656678904e-05,
      "loss": 0.0287,
      "step": 1450
    },
    {
      "epoch": 0.037872892347600516,
      "grad_norm": 0.19943729043006897,
      "learning_rate": 8.061261018263919e-05,
      "loss": 0.0249,
      "step": 1460
    },
    {
      "epoch": 0.03813229571984436,
      "grad_norm": 0.3236585557460785,
      "learning_rate": 7.931666603034033e-05,
      "loss": 0.0261,
      "step": 1470
    },
    {
      "epoch": 0.0383916990920882,
      "grad_norm": 0.27057945728302,
      "learning_rate": 7.80243408625947e-05,
      "loss": 0.0313,
      "step": 1480
    },
    {
      "epoch": 0.03865110246433204,
      "grad_norm": 0.40263959765434265,
      "learning_rate": 7.673586079888698e-05,
      "loss": 0.0255,
      "step": 1490
    },
    {
      "epoch": 0.038910505836575876,
      "grad_norm": 0.36453911662101746,
      "learning_rate": 7.54514512859201e-05,
      "loss": 0.0306,
      "step": 1500
    },
    {
      "epoch": 0.039169909208819714,
      "grad_norm": 0.2801981270313263,
      "learning_rate": 7.417133705816837e-05,
      "loss": 0.0272,
      "step": 1510
    },
    {
      "epoch": 0.03942931258106355,
      "grad_norm": 0.2771183252334595,
      "learning_rate": 7.289574209855559e-05,
      "loss": 0.0252,
      "step": 1520
    },
    {
      "epoch": 0.03968871595330739,
      "grad_norm": 0.1658083200454712,
      "learning_rate": 7.16248895992645e-05,
      "loss": 0.0278,
      "step": 1530
    },
    {
      "epoch": 0.03994811932555123,
      "grad_norm": 0.19812481105327606,
      "learning_rate": 7.035900192268464e-05,
      "loss": 0.028,
      "step": 1540
    },
    {
      "epoch": 0.040207522697795074,
      "grad_norm": 0.2751558721065521,
      "learning_rate": 6.909830056250527e-05,
      "loss": 0.0293,
      "step": 1550
    },
    {
      "epoch": 0.04046692607003891,
      "grad_norm": 0.24725764989852905,
      "learning_rate": 6.784300610496048e-05,
      "loss": 0.0245,
      "step": 1560
    },
    {
      "epoch": 0.04072632944228275,
      "grad_norm": 0.6854105591773987,
      "learning_rate": 6.65933381902329e-05,
      "loss": 0.0278,
      "step": 1570
    },
    {
      "epoch": 0.04098573281452659,
      "grad_norm": 0.19634288549423218,
      "learning_rate": 6.534951547402322e-05,
      "loss": 0.0251,
      "step": 1580
    },
    {
      "epoch": 0.04124513618677043,
      "grad_norm": 0.21054495871067047,
      "learning_rate": 6.411175558929152e-05,
      "loss": 0.0263,
      "step": 1590
    },
    {
      "epoch": 0.041504539559014265,
      "grad_norm": 0.2615554630756378,
      "learning_rate": 6.28802751081779e-05,
      "loss": 0.0268,
      "step": 1600
    },
    {
      "epoch": 0.041763942931258104,
      "grad_norm": 0.23431521654129028,
      "learning_rate": 6.165528950410884e-05,
      "loss": 0.0272,
      "step": 1610
    },
    {
      "epoch": 0.04202334630350195,
      "grad_norm": 0.14299079775810242,
      "learning_rate": 6.0437013114095195e-05,
      "loss": 0.0206,
      "step": 1620
    },
    {
      "epoch": 0.04228274967574579,
      "grad_norm": 0.32186359167099,
      "learning_rate": 5.922565910122967e-05,
      "loss": 0.0257,
      "step": 1630
    },
    {
      "epoch": 0.042542153047989625,
      "grad_norm": 0.3568614423274994,
      "learning_rate": 5.8021439417389444e-05,
      "loss": 0.0263,
      "step": 1640
    },
    {
      "epoch": 0.042801556420233464,
      "grad_norm": 0.2789517045021057,
      "learning_rate": 5.6824564766150726e-05,
      "loss": 0.0273,
      "step": 1650
    },
    {
      "epoch": 0.0430609597924773,
      "grad_norm": 0.23182226717472076,
      "learning_rate": 5.563524456592163e-05,
      "loss": 0.0216,
      "step": 1660
    },
    {
      "epoch": 0.04332036316472114,
      "grad_norm": 0.20413507521152496,
      "learning_rate": 5.4453686913300074e-05,
      "loss": 0.0281,
      "step": 1670
    },
    {
      "epoch": 0.04357976653696498,
      "grad_norm": 0.21678327023983002,
      "learning_rate": 5.328009854666303e-05,
      "loss": 0.0258,
      "step": 1680
    },
    {
      "epoch": 0.04383916990920882,
      "grad_norm": 0.25063127279281616,
      "learning_rate": 5.2114684809993044e-05,
      "loss": 0.0265,
      "step": 1690
    },
    {
      "epoch": 0.04409857328145266,
      "grad_norm": 0.17453815042972565,
      "learning_rate": 5.095764961694922e-05,
      "loss": 0.0256,
      "step": 1700
    },
    {
      "epoch": 0.0443579766536965,
      "grad_norm": 0.22323985397815704,
      "learning_rate": 4.980919541518796e-05,
      "loss": 0.025,
      "step": 1710
    },
    {
      "epoch": 0.04461738002594034,
      "grad_norm": 0.2994442880153656,
      "learning_rate": 4.866952315094088e-05,
      "loss": 0.0266,
      "step": 1720
    },
    {
      "epoch": 0.04487678339818418,
      "grad_norm": 0.27734553813934326,
      "learning_rate": 4.753883223385467e-05,
      "loss": 0.0247,
      "step": 1730
    },
    {
      "epoch": 0.045136186770428015,
      "grad_norm": 0.18703316152095795,
      "learning_rate": 4.6417320502100316e-05,
      "loss": 0.0232,
      "step": 1740
    },
    {
      "epoch": 0.04539559014267185,
      "grad_norm": 0.271628201007843,
      "learning_rate": 4.530518418775733e-05,
      "loss": 0.0275,
      "step": 1750
    },
    {
      "epoch": 0.04565499351491569,
      "grad_norm": 0.26739704608917236,
      "learning_rate": 4.4202617882478405e-05,
      "loss": 0.0273,
      "step": 1760
    },
    {
      "epoch": 0.045914396887159536,
      "grad_norm": 0.2327081710100174,
      "learning_rate": 4.310981450344189e-05,
      "loss": 0.0244,
      "step": 1770
    },
    {
      "epoch": 0.046173800259403375,
      "grad_norm": 0.23359940946102142,
      "learning_rate": 4.2026965259596666e-05,
      "loss": 0.0266,
      "step": 1780
    },
    {
      "epoch": 0.04643320363164721,
      "grad_norm": 0.1780378818511963,
      "learning_rate": 4.0954259618206295e-05,
      "loss": 0.0275,
      "step": 1790
    },
    {
      "epoch": 0.04669260700389105,
      "grad_norm": 0.14094606041908264,
      "learning_rate": 3.9891885271697496e-05,
      "loss": 0.0224,
      "step": 1800
    },
    {
      "epoch": 0.04695201037613489,
      "grad_norm": 0.21397313475608826,
      "learning_rate": 3.884002810481958e-05,
      "loss": 0.0262,
      "step": 1810
    },
    {
      "epoch": 0.04721141374837873,
      "grad_norm": 0.18339146673679352,
      "learning_rate": 3.779887216211995e-05,
      "loss": 0.0227,
      "step": 1820
    },
    {
      "epoch": 0.047470817120622566,
      "grad_norm": 0.3625165522098541,
      "learning_rate": 3.676859961574162e-05,
      "loss": 0.0261,
      "step": 1830
    },
    {
      "epoch": 0.047730220492866404,
      "grad_norm": 0.24323268234729767,
      "learning_rate": 3.574939073354838e-05,
      "loss": 0.0263,
      "step": 1840
    },
    {
      "epoch": 0.04798962386511025,
      "grad_norm": 0.16706861555576324,
      "learning_rate": 3.4741423847583134e-05,
      "loss": 0.0236,
      "step": 1850
    },
    {
      "epoch": 0.04824902723735409,
      "grad_norm": 0.25477734208106995,
      "learning_rate": 3.3744875322865034e-05,
      "loss": 0.0265,
      "step": 1860
    },
    {
      "epoch": 0.048508430609597926,
      "grad_norm": 0.23444534838199615,
      "learning_rate": 3.275991952653054e-05,
      "loss": 0.0255,
      "step": 1870
    },
    {
      "epoch": 0.048767833981841764,
      "grad_norm": 0.13936041295528412,
      "learning_rate": 3.178672879732435e-05,
      "loss": 0.0247,
      "step": 1880
    },
    {
      "epoch": 0.0490272373540856,
      "grad_norm": 0.27730414271354675,
      "learning_rate": 3.0825473415445074e-05,
      "loss": 0.0211,
      "step": 1890
    },
    {
      "epoch": 0.04928664072632944,
      "grad_norm": 0.44268760085105896,
      "learning_rate": 2.9876321572751144e-05,
      "loss": 0.0293,
      "step": 1900
    },
    {
      "epoch": 0.04954604409857328,
      "grad_norm": 0.2908082902431488,
      "learning_rate": 2.8939439343332086e-05,
      "loss": 0.0283,
      "step": 1910
    },
    {
      "epoch": 0.049805447470817124,
      "grad_norm": 0.20642030239105225,
      "learning_rate": 2.8014990654450325e-05,
      "loss": 0.0237,
      "step": 1920
    },
    {
      "epoch": 0.05006485084306096,
      "grad_norm": 0.23467542231082916,
      "learning_rate": 2.7103137257858868e-05,
      "loss": 0.0257,
      "step": 1930
    },
    {
      "epoch": 0.0503242542153048,
      "grad_norm": 0.30216318368911743,
      "learning_rate": 2.6204038701499056e-05,
      "loss": 0.023,
      "step": 1940
    },
    {
      "epoch": 0.05058365758754864,
      "grad_norm": 0.2494220733642578,
      "learning_rate": 2.5317852301584643e-05,
      "loss": 0.0248,
      "step": 1950
    },
    {
      "epoch": 0.05084306095979248,
      "grad_norm": 0.3801492154598236,
      "learning_rate": 2.4444733115075823e-05,
      "loss": 0.0208,
      "step": 1960
    },
    {
      "epoch": 0.051102464332036315,
      "grad_norm": 0.18340712785720825,
      "learning_rate": 2.3584833912548888e-05,
      "loss": 0.0265,
      "step": 1970
    },
    {
      "epoch": 0.051361867704280154,
      "grad_norm": 0.20898978412151337,
      "learning_rate": 2.2738305151465645e-05,
      "loss": 0.0242,
      "step": 1980
    },
    {
      "epoch": 0.05162127107652399,
      "grad_norm": 0.2393597811460495,
      "learning_rate": 2.190529494984782e-05,
      "loss": 0.0223,
      "step": 1990
    },
    {
      "epoch": 0.05188067444876784,
      "grad_norm": 0.16330353915691376,
      "learning_rate": 2.1085949060360654e-05,
      "loss": 0.0207,
      "step": 2000
    },
    {
      "epoch": 0.052140077821011675,
      "grad_norm": 0.2321455329656601,
      "learning_rate": 2.0280410844810428e-05,
      "loss": 0.0208,
      "step": 2010
    },
    {
      "epoch": 0.052399481193255514,
      "grad_norm": 0.1408013105392456,
      "learning_rate": 1.9488821249060297e-05,
      "loss": 0.0248,
      "step": 2020
    },
    {
      "epoch": 0.05265888456549935,
      "grad_norm": 0.38080984354019165,
      "learning_rate": 1.871131877836879e-05,
      "loss": 0.0286,
      "step": 2030
    },
    {
      "epoch": 0.05291828793774319,
      "grad_norm": 0.17722366750240326,
      "learning_rate": 1.7948039473155554e-05,
      "loss": 0.0283,
      "step": 2040
    },
    {
      "epoch": 0.05317769130998703,
      "grad_norm": 0.2016279697418213,
      "learning_rate": 1.7199116885197995e-05,
      "loss": 0.0258,
      "step": 2050
    },
    {
      "epoch": 0.053437094682230867,
      "grad_norm": 0.24826838076114655,
      "learning_rate": 1.646468205426377e-05,
      "loss": 0.0233,
      "step": 2060
    },
    {
      "epoch": 0.053696498054474705,
      "grad_norm": 0.21984995901584625,
      "learning_rate": 1.5744863485182537e-05,
      "loss": 0.0234,
      "step": 2070
    },
    {
      "epoch": 0.05395590142671855,
      "grad_norm": 0.16736479103565216,
      "learning_rate": 1.5039787125361326e-05,
      "loss": 0.0229,
      "step": 2080
    },
    {
      "epoch": 0.05421530479896239,
      "grad_norm": 0.1707315593957901,
      "learning_rate": 1.4349576342747462e-05,
      "loss": 0.02,
      "step": 2090
    },
    {
      "epoch": 0.054474708171206226,
      "grad_norm": 0.20376580953598022,
      "learning_rate": 1.3674351904242611e-05,
      "loss": 0.0193,
      "step": 2100
    },
    {
      "epoch": 0.054734111543450065,
      "grad_norm": 0.18196424841880798,
      "learning_rate": 1.3014231954572287e-05,
      "loss": 0.0201,
      "step": 2110
    },
    {
      "epoch": 0.0549935149156939,
      "grad_norm": 0.2857169806957245,
      "learning_rate": 1.2369331995613665e-05,
      "loss": 0.0222,
      "step": 2120
    },
    {
      "epoch": 0.05525291828793774,
      "grad_norm": 0.23490458726882935,
      "learning_rate": 1.173976486618631e-05,
      "loss": 0.0224,
      "step": 2130
    },
    {
      "epoch": 0.05551232166018158,
      "grad_norm": 0.2817946672439575,
      "learning_rate": 1.1125640722308628e-05,
      "loss": 0.0227,
      "step": 2140
    },
    {
      "epoch": 0.055771725032425425,
      "grad_norm": 0.2497781217098236,
      "learning_rate": 1.0527067017923654e-05,
      "loss": 0.0216,
      "step": 2150
    },
    {
      "epoch": 0.05603112840466926,
      "grad_norm": 0.2638106048107147,
      "learning_rate": 9.944148486097793e-06,
      "loss": 0.0236,
      "step": 2160
    },
    {
      "epoch": 0.0562905317769131,
      "grad_norm": 0.11962006986141205,
      "learning_rate": 9.376987120695545e-06,
      "loss": 0.0187,
      "step": 2170
    },
    {
      "epoch": 0.05654993514915694,
      "grad_norm": 0.3043888211250305,
      "learning_rate": 8.825682158533554e-06,
      "loss": 0.0254,
      "step": 2180
    },
    {
      "epoch": 0.05680933852140078,
      "grad_norm": 0.18844349682331085,
      "learning_rate": 8.290330062017016e-06,
      "loss": 0.0214,
      "step": 2190
    },
    {
      "epoch": 0.057068741893644616,
      "grad_norm": 0.320204496383667,
      "learning_rate": 7.771024502261526e-06,
      "loss": 0.0239,
      "step": 2200
    },
    {
      "epoch": 0.057328145265888454,
      "grad_norm": 0.16253778338432312,
      "learning_rate": 7.267856342703461e-06,
      "loss": 0.0207,
      "step": 2210
    },
    {
      "epoch": 0.05758754863813229,
      "grad_norm": 0.1236085444688797,
      "learning_rate": 6.780913623201346e-06,
      "loss": 0.0194,
      "step": 2220
    },
    {
      "epoch": 0.05784695201037614,
      "grad_norm": 0.17225727438926697,
      "learning_rate": 6.310281544631546e-06,
      "loss": 0.0229,
      "step": 2230
    },
    {
      "epoch": 0.058106355382619976,
      "grad_norm": 0.15956400334835052,
      "learning_rate": 5.856042453980526e-06,
      "loss": 0.0213,
      "step": 2240
    },
    {
      "epoch": 0.058365758754863814,
      "grad_norm": 0.14553219079971313,
      "learning_rate": 5.418275829936537e-06,
      "loss": 0.0267,
      "step": 2250
    },
    {
      "epoch": 0.05862516212710765,
      "grad_norm": 0.1686364859342575,
      "learning_rate": 4.997058268983135e-06,
      "loss": 0.0218,
      "step": 2260
    },
    {
      "epoch": 0.05888456549935149,
      "grad_norm": 0.2441764622926712,
      "learning_rate": 4.592463471997022e-06,
      "loss": 0.0232,
      "step": 2270
    },
    {
      "epoch": 0.05914396887159533,
      "grad_norm": 0.3006925880908966,
      "learning_rate": 4.204562231352516e-06,
      "loss": 0.027,
      "step": 2280
    },
    {
      "epoch": 0.05940337224383917,
      "grad_norm": 0.23077550530433655,
      "learning_rate": 3.83342241853496e-06,
      "loss": 0.0232,
      "step": 2290
    },
    {
      "epoch": 0.05966277561608301,
      "grad_norm": 0.32399365305900574,
      "learning_rate": 3.4791089722651436e-06,
      "loss": 0.0211,
      "step": 2300
    },
    {
      "epoch": 0.05992217898832685,
      "grad_norm": 0.2478347271680832,
      "learning_rate": 3.1416838871368924e-06,
      "loss": 0.027,
      "step": 2310
    },
    {
      "epoch": 0.06018158236057069,
      "grad_norm": 0.17069502174854279,
      "learning_rate": 2.821206202769899e-06,
      "loss": 0.023,
      "step": 2320
    },
    {
      "epoch": 0.06044098573281453,
      "grad_norm": 0.1713675558567047,
      "learning_rate": 2.5177319934794e-06,
      "loss": 0.0275,
      "step": 2330
    },
    {
      "epoch": 0.060700389105058365,
      "grad_norm": 0.13431312143802643,
      "learning_rate": 2.2313143584648423e-06,
      "loss": 0.0209,
      "step": 2340
    },
    {
      "epoch": 0.0609597924773022,
      "grad_norm": 0.20045480132102966,
      "learning_rate": 1.9620034125190644e-06,
      "loss": 0.026,
      "step": 2350
    },
    {
      "epoch": 0.06121919584954604,
      "grad_norm": 0.1736753284931183,
      "learning_rate": 1.7098462772596302e-06,
      "loss": 0.0224,
      "step": 2360
    },
    {
      "epoch": 0.06147859922178988,
      "grad_norm": 0.20570042729377747,
      "learning_rate": 1.4748870728839347e-06,
      "loss": 0.0213,
      "step": 2370
    },
    {
      "epoch": 0.061738002594033725,
      "grad_norm": 0.15525294840335846,
      "learning_rate": 1.2571669104494256e-06,
      "loss": 0.0215,
      "step": 2380
    },
    {
      "epoch": 0.06199740596627756,
      "grad_norm": 0.16905346512794495,
      "learning_rate": 1.0567238846803996e-06,
      "loss": 0.0231,
      "step": 2390
    },
    {
      "epoch": 0.0622568093385214,
      "grad_norm": 0.18170209228992462,
      "learning_rate": 8.735930673024806e-07,
      "loss": 0.0233,
      "step": 2400
    },
    {
      "epoch": 0.06251621271076524,
      "grad_norm": 0.19480860233306885,
      "learning_rate": 7.078065009060941e-07,
      "loss": 0.0215,
      "step": 2410
    },
    {
      "epoch": 0.06277561608300908,
      "grad_norm": 0.1651337891817093,
      "learning_rate": 5.593931933399854e-07,
      "loss": 0.0224,
      "step": 2420
    },
    {
      "epoch": 0.06303501945525292,
      "grad_norm": 0.1658225804567337,
      "learning_rate": 4.2837911263562404e-07,
      "loss": 0.024,
      "step": 2430
    },
    {
      "epoch": 0.06329442282749675,
      "grad_norm": 0.2529785633087158,
      "learning_rate": 3.1478718246357173e-07,
      "loss": 0.0275,
      "step": 2440
    },
    {
      "epoch": 0.06355382619974059,
      "grad_norm": 0.14123444259166718,
      "learning_rate": 2.1863727812254653e-07,
      "loss": 0.0212,
      "step": 2450
    },
    {
      "epoch": 0.06381322957198443,
      "grad_norm": 0.14571131765842438,
      "learning_rate": 1.3994622306173765e-07,
      "loss": 0.0225,
      "step": 2460
    },
    {
      "epoch": 0.06407263294422827,
      "grad_norm": 0.20619624853134155,
      "learning_rate": 7.872778593728258e-08,
      "loss": 0.0214,
      "step": 2470
    },
    {
      "epoch": 0.06433203631647211,
      "grad_norm": 0.19361430406570435,
      "learning_rate": 3.499267820307184e-08,
      "loss": 0.0233,
      "step": 2480
    },
    {
      "epoch": 0.06459143968871596,
      "grad_norm": 0.20903588831424713,
      "learning_rate": 8.748552236603757e-09,
      "loss": 0.0234,
      "step": 2490
    },
    {
      "epoch": 0.0648508430609598,
      "grad_norm": 0.17228850722312927,
      "learning_rate": 0.0,
      "loss": 0.0261,
      "step": 2500
    },
    {
      "epoch": 0.0648508430609598,
      "step": 2500,
      "total_flos": 3.2864989085696e+17,
      "train_loss": 0.043296425741910935,
      "train_runtime": 1930.1512,
      "train_samples_per_second": 82.895,
      "train_steps_per_second": 1.295
    }
  ],
  "logging_steps": 10,
  "max_steps": 2500,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 3.2864989085696e+17,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}