diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,20833 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9999663084127893,
+  "eval_steps": 500,
+  "global_step": 14840,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 6.7383174421347e-05,
+      "grad_norm": 32.09206302403183,
+      "learning_rate": 6.738544474393531e-10,
+      "loss": 3.2042,
+      "step": 1
+    },
+    {
+      "epoch": 0.00033691587210673494,
+      "grad_norm": 29.85124796764343,
+      "learning_rate": 3.3692722371967655e-09,
+      "loss": 3.0738,
+      "step": 5
+    },
+    {
+      "epoch": 0.0006738317442134699,
+      "grad_norm": 35.348710336485944,
+      "learning_rate": 6.738544474393531e-09,
+      "loss": 3.0573,
+      "step": 10
+    },
+    {
+      "epoch": 0.001010747616320205,
+      "grad_norm": 35.40963118986941,
+      "learning_rate": 1.0107816711590296e-08,
+      "loss": 3.0608,
+      "step": 15
+    },
+    {
+      "epoch": 0.0013476634884269398,
+      "grad_norm": 32.39762203020303,
+      "learning_rate": 1.3477088948787062e-08,
+      "loss": 3.0198,
+      "step": 20
+    },
+    {
+      "epoch": 0.0016845793605336748,
+      "grad_norm": 30.251139866969535,
+      "learning_rate": 1.6846361185983825e-08,
+      "loss": 3.0206,
+      "step": 25
+    },
+    {
+      "epoch": 0.00202149523264041,
+      "grad_norm": 26.501991719428602,
+      "learning_rate": 2.021563342318059e-08,
+      "loss": 2.9709,
+      "step": 30
+    },
+    {
+      "epoch": 0.0023584111047471445,
+      "grad_norm": 31.481478748265243,
+      "learning_rate": 2.3584905660377358e-08,
+      "loss": 3.0565,
+      "step": 35
+    },
+    {
+      "epoch": 0.0026953269768538795,
+      "grad_norm": 34.70890491380774,
+      "learning_rate": 2.6954177897574124e-08,
+      "loss": 3.0505,
+      "step": 40
+    },
+    {
+      "epoch": 0.0030322428489606146,
+      "grad_norm": 31.691755532137382,
+      "learning_rate": 3.032345013477089e-08,
+      "loss": 3.0692,
+      "step": 45
+    },
+    {
+      "epoch": 0.0033691587210673496,
+      "grad_norm": 36.9671149567999,
+      "learning_rate": 3.369272237196765e-08,
+      "loss": 3.1505,
+      "step": 50
+    },
+    {
+      "epoch": 0.0037060745931740842,
+      "grad_norm": 30.246848965440783,
+      "learning_rate": 3.706199460916442e-08,
+      "loss": 3.08,
+      "step": 55
+    },
+    {
+      "epoch": 0.00404299046528082,
+      "grad_norm": 33.313935844298754,
+      "learning_rate": 4.043126684636118e-08,
+      "loss": 2.9876,
+      "step": 60
+    },
+    {
+      "epoch": 0.004379906337387554,
+      "grad_norm": 29.15727651429879,
+      "learning_rate": 4.380053908355795e-08,
+      "loss": 3.0608,
+      "step": 65
+    },
+    {
+      "epoch": 0.004716822209494289,
+      "grad_norm": 32.676729128831575,
+      "learning_rate": 4.7169811320754715e-08,
+      "loss": 3.1247,
+      "step": 70
+    },
+    {
+      "epoch": 0.005053738081601024,
+      "grad_norm": 31.46286104321294,
+      "learning_rate": 5.053908355795148e-08,
+      "loss": 3.0343,
+      "step": 75
+    },
+    {
+      "epoch": 0.005390653953707759,
+      "grad_norm": 25.222824338343894,
+      "learning_rate": 5.390835579514825e-08,
+      "loss": 2.9254,
+      "step": 80
+    },
+    {
+      "epoch": 0.005727569825814494,
+      "grad_norm": 22.637008002000595,
+      "learning_rate": 5.727762803234501e-08,
+      "loss": 2.7987,
+      "step": 85
+    },
+    {
+      "epoch": 0.006064485697921229,
+      "grad_norm": 27.769504016534167,
+      "learning_rate": 6.064690026954177e-08,
+      "loss": 2.9506,
+      "step": 90
+    },
+    {
+      "epoch": 0.006401401570027964,
+      "grad_norm": 26.957750943106987,
+      "learning_rate": 6.401617250673854e-08,
+      "loss": 2.9574,
+      "step": 95
+    },
+    {
+      "epoch": 0.006738317442134699,
+      "grad_norm": 26.113167121031843,
+      "learning_rate": 6.73854447439353e-08,
+      "loss": 2.9626,
+      "step": 100
+    },
+    {
+      "epoch": 0.007075233314241434,
+      "grad_norm": 27.596533306623297,
+      "learning_rate": 7.075471698113207e-08,
+      "loss": 3.0049,
+      "step": 105
+    },
+    {
+      "epoch": 0.0074121491863481685,
+      "grad_norm": 21.936343647967842,
+      "learning_rate": 7.412398921832884e-08,
+      "loss": 2.9212,
+      "step": 110
+    },
+    {
+      "epoch": 0.0077490650584549035,
+      "grad_norm": 17.587509277494334,
+      "learning_rate": 7.749326145552561e-08,
+      "loss": 2.8243,
+      "step": 115
+    },
+    {
+      "epoch": 0.00808598093056164,
+      "grad_norm": 17.039393337653188,
+      "learning_rate": 8.086253369272237e-08,
+      "loss": 2.7713,
+      "step": 120
+    },
+    {
+      "epoch": 0.008422896802668374,
+      "grad_norm": 17.11446090656002,
+      "learning_rate": 8.423180592991913e-08,
+      "loss": 2.7353,
+      "step": 125
+    },
+    {
+      "epoch": 0.008759812674775108,
+      "grad_norm": 17.005432651209553,
+      "learning_rate": 8.76010781671159e-08,
+      "loss": 2.83,
+      "step": 130
+    },
+    {
+      "epoch": 0.009096728546881844,
+      "grad_norm": 16.994628911009993,
+      "learning_rate": 9.097035040431267e-08,
+      "loss": 2.8259,
+      "step": 135
+    },
+    {
+      "epoch": 0.009433644418988578,
+      "grad_norm": 14.381055662834557,
+      "learning_rate": 9.433962264150943e-08,
+      "loss": 2.6585,
+      "step": 140
+    },
+    {
+      "epoch": 0.009770560291095314,
+      "grad_norm": 12.855855412057881,
+      "learning_rate": 9.770889487870619e-08,
+      "loss": 2.7127,
+      "step": 145
+    },
+    {
+      "epoch": 0.010107476163202048,
+      "grad_norm": 9.33136232394542,
+      "learning_rate": 1.0107816711590296e-07,
+      "loss": 2.6163,
+      "step": 150
+    },
+    {
+      "epoch": 0.010444392035308784,
+      "grad_norm": 8.974119069707546,
+      "learning_rate": 1.0444743935309973e-07,
+      "loss": 2.6374,
+      "step": 155
+    },
+    {
+      "epoch": 0.010781307907415518,
+      "grad_norm": 9.295477060485023,
+      "learning_rate": 1.078167115902965e-07,
+      "loss": 2.606,
+      "step": 160
+    },
+    {
+      "epoch": 0.011118223779522254,
+      "grad_norm": 9.296285754353445,
+      "learning_rate": 1.1118598382749325e-07,
+      "loss": 2.5663,
+      "step": 165
+    },
+    {
+      "epoch": 0.011455139651628988,
+      "grad_norm": 8.107059371834286,
+      "learning_rate": 1.1455525606469002e-07,
+      "loss": 2.5504,
+      "step": 170
+    },
+    {
+      "epoch": 0.011792055523735722,
+      "grad_norm": 8.843539027040743,
+      "learning_rate": 1.1792452830188679e-07,
+      "loss": 2.557,
+      "step": 175
+    },
+    {
+      "epoch": 0.012128971395842458,
+      "grad_norm": 7.057056300901147,
+      "learning_rate": 1.2129380053908355e-07,
+      "loss": 2.5615,
+      "step": 180
+    },
+    {
+      "epoch": 0.012465887267949192,
+      "grad_norm": 7.38030846809641,
+      "learning_rate": 1.2466307277628032e-07,
+      "loss": 2.5312,
+      "step": 185
+    },
+    {
+      "epoch": 0.012802803140055928,
+      "grad_norm": 7.8962233248696805,
+      "learning_rate": 1.280323450134771e-07,
+      "loss": 2.4867,
+      "step": 190
+    },
+    {
+      "epoch": 0.013139719012162663,
+      "grad_norm": 6.554681755642436,
+      "learning_rate": 1.3140161725067383e-07,
+      "loss": 2.5143,
+      "step": 195
+    },
+    {
+      "epoch": 0.013476634884269399,
+      "grad_norm": 7.140796549890275,
+      "learning_rate": 1.347708894878706e-07,
+      "loss": 2.4714,
+      "step": 200
+    },
+    {
+      "epoch": 0.013813550756376133,
+      "grad_norm": 6.487040702144441,
+      "learning_rate": 1.3814016172506737e-07,
+      "loss": 2.5227,
+      "step": 205
+    },
+    {
+      "epoch": 0.014150466628482869,
+      "grad_norm": 6.8916036053852645,
+      "learning_rate": 1.4150943396226414e-07,
+      "loss": 2.5232,
+      "step": 210
+    },
+    {
+      "epoch": 0.014487382500589603,
+      "grad_norm": 6.419262649425003,
+      "learning_rate": 1.448787061994609e-07,
+      "loss": 2.4131,
+      "step": 215
+    },
+    {
+      "epoch": 0.014824298372696337,
+      "grad_norm": 7.861046646197374,
+      "learning_rate": 1.4824797843665768e-07,
+      "loss": 2.4394,
+      "step": 220
+    },
+    {
+      "epoch": 0.015161214244803073,
+      "grad_norm": 7.562450211806467,
+      "learning_rate": 1.5161725067385445e-07,
+      "loss": 2.4057,
+      "step": 225
+    },
+    {
+      "epoch": 0.015498130116909807,
+      "grad_norm": 6.938949312004328,
+      "learning_rate": 1.5498652291105122e-07,
+      "loss": 2.4074,
+      "step": 230
+    },
+    {
+      "epoch": 0.01583504598901654,
+      "grad_norm": 6.496838158861888,
+      "learning_rate": 1.58355795148248e-07,
+      "loss": 2.3519,
+      "step": 235
+    },
+    {
+      "epoch": 0.01617196186112328,
+      "grad_norm": 5.7314057513508185,
+      "learning_rate": 1.6172506738544473e-07,
+      "loss": 2.367,
+      "step": 240
+    },
+    {
+      "epoch": 0.016508877733230013,
+      "grad_norm": 6.951340639666037,
+      "learning_rate": 1.650943396226415e-07,
+      "loss": 2.3327,
+      "step": 245
+    },
+    {
+      "epoch": 0.016845793605336747,
+      "grad_norm": 7.233019401736712,
+      "learning_rate": 1.6846361185983827e-07,
+      "loss": 2.3259,
+      "step": 250
+    },
+    {
+      "epoch": 0.01718270947744348,
+      "grad_norm": 6.120666820602053,
+      "learning_rate": 1.7183288409703504e-07,
+      "loss": 2.3448,
+      "step": 255
+    },
+    {
+      "epoch": 0.017519625349550216,
+      "grad_norm": 5.822657164591844,
+      "learning_rate": 1.752021563342318e-07,
+      "loss": 2.2525,
+      "step": 260
+    },
+    {
+      "epoch": 0.017856541221656953,
+      "grad_norm": 5.645590925567188,
+      "learning_rate": 1.7857142857142858e-07,
+      "loss": 2.3429,
+      "step": 265
+    },
+    {
+      "epoch": 0.018193457093763687,
+      "grad_norm": 5.542805631769041,
+      "learning_rate": 1.8194070080862535e-07,
+      "loss": 2.2838,
+      "step": 270
+    },
+    {
+      "epoch": 0.01853037296587042,
+      "grad_norm": 5.856964348688246,
+      "learning_rate": 1.853099730458221e-07,
+      "loss": 2.3266,
+      "step": 275
+    },
+    {
+      "epoch": 0.018867288837977156,
+      "grad_norm": 5.216722479418778,
+      "learning_rate": 1.8867924528301886e-07,
+      "loss": 2.2765,
+      "step": 280
+    },
+    {
+      "epoch": 0.019204204710083893,
+      "grad_norm": 5.608251688722653,
+      "learning_rate": 1.920485175202156e-07,
+      "loss": 2.2346,
+      "step": 285
+    },
+    {
+      "epoch": 0.019541120582190628,
+      "grad_norm": 6.214595981569079,
+      "learning_rate": 1.9541778975741237e-07,
+      "loss": 2.2486,
+      "step": 290
+    },
+    {
+      "epoch": 0.019878036454297362,
+      "grad_norm": 25.172949782744762,
+      "learning_rate": 1.9878706199460914e-07,
+      "loss": 2.2545,
+      "step": 295
+    },
+    {
+      "epoch": 0.020214952326404096,
+      "grad_norm": 6.03234667659201,
+      "learning_rate": 2.021563342318059e-07,
+      "loss": 2.3099,
+      "step": 300
+    },
+    {
+      "epoch": 0.02055186819851083,
+      "grad_norm": 5.5276661755222545,
+      "learning_rate": 2.0552560646900268e-07,
+      "loss": 2.1518,
+      "step": 305
+    },
+    {
+      "epoch": 0.020888784070617568,
+      "grad_norm": 6.148883486161354,
+      "learning_rate": 2.0889487870619945e-07,
+      "loss": 2.2243,
+      "step": 310
+    },
+    {
+      "epoch": 0.021225699942724302,
+      "grad_norm": 7.031330506249546,
+      "learning_rate": 2.1226415094339622e-07,
+      "loss": 2.2653,
+      "step": 315
+    },
+    {
+      "epoch": 0.021562615814831036,
+      "grad_norm": 5.720437592526712,
+      "learning_rate": 2.15633423180593e-07,
+      "loss": 2.2255,
+      "step": 320
+    },
+    {
+      "epoch": 0.02189953168693777,
+      "grad_norm": 6.232004013022435,
+      "learning_rate": 2.1900269541778973e-07,
+      "loss": 2.2326,
+      "step": 325
+    },
+    {
+      "epoch": 0.022236447559044508,
+      "grad_norm": 5.427884766447202,
+      "learning_rate": 2.223719676549865e-07,
+      "loss": 2.2255,
+      "step": 330
+    },
+    {
+      "epoch": 0.022573363431151242,
+      "grad_norm": 5.5418646369024875,
+      "learning_rate": 2.2574123989218327e-07,
+      "loss": 2.1378,
+      "step": 335
+    },
+    {
+      "epoch": 0.022910279303257976,
+      "grad_norm": 6.0425076069387895,
+      "learning_rate": 2.2911051212938004e-07,
+      "loss": 2.2097,
+      "step": 340
+    },
+    {
+      "epoch": 0.02324719517536471,
+      "grad_norm": 5.400950802947024,
+      "learning_rate": 2.324797843665768e-07,
+      "loss": 2.2412,
+      "step": 345
+    },
+    {
+      "epoch": 0.023584111047471445,
+      "grad_norm": 5.173515089054146,
+      "learning_rate": 2.3584905660377358e-07,
+      "loss": 2.1899,
+      "step": 350
+    },
+    {
+      "epoch": 0.023921026919578182,
+      "grad_norm": 5.240519930808722,
+      "learning_rate": 2.392183288409703e-07,
+      "loss": 2.2218,
+      "step": 355
+    },
+    {
+      "epoch": 0.024257942791684917,
+      "grad_norm": 5.732864030593667,
+      "learning_rate": 2.425876010781671e-07,
+      "loss": 2.1959,
+      "step": 360
+    },
+    {
+      "epoch": 0.02459485866379165,
+      "grad_norm": 5.986885229622732,
+      "learning_rate": 2.4595687331536387e-07,
+      "loss": 2.179,
+      "step": 365
+    },
+    {
+      "epoch": 0.024931774535898385,
+      "grad_norm": 6.367446128253974,
+      "learning_rate": 2.4932614555256063e-07,
+      "loss": 2.1408,
+      "step": 370
+    },
+    {
+      "epoch": 0.025268690408005123,
+      "grad_norm": 5.370594569811163,
+      "learning_rate": 2.526954177897574e-07,
+      "loss": 2.1142,
+      "step": 375
+    },
+    {
+      "epoch": 0.025605606280111857,
+      "grad_norm": 5.307650833779364,
+      "learning_rate": 2.560646900269542e-07,
+      "loss": 2.1089,
+      "step": 380
+    },
+    {
+      "epoch": 0.02594252215221859,
+      "grad_norm": 5.931005595781113,
+      "learning_rate": 2.5943396226415094e-07,
+      "loss": 2.0792,
+      "step": 385
+    },
+    {
+      "epoch": 0.026279438024325325,
+      "grad_norm": 5.636620139141842,
+      "learning_rate": 2.6280323450134766e-07,
+      "loss": 2.1345,
+      "step": 390
+    },
+    {
+      "epoch": 0.02661635389643206,
+      "grad_norm": 6.285950145006834,
+      "learning_rate": 2.661725067385445e-07,
+      "loss": 2.2002,
+      "step": 395
+    },
+    {
+      "epoch": 0.026953269768538797,
+      "grad_norm": 5.496040660546177,
+      "learning_rate": 2.695417789757412e-07,
+      "loss": 2.0634,
+      "step": 400
+    },
+    {
+      "epoch": 0.02729018564064553,
+      "grad_norm": 5.355777932542913,
+      "learning_rate": 2.72911051212938e-07,
+      "loss": 2.1038,
+      "step": 405
+    },
+    {
+      "epoch": 0.027627101512752265,
+      "grad_norm": 5.407560676614071,
+      "learning_rate": 2.7628032345013474e-07,
+      "loss": 2.116,
+      "step": 410
+    },
+    {
+      "epoch": 0.027964017384859,
+      "grad_norm": 5.510824187092772,
+      "learning_rate": 2.7964959568733156e-07,
+      "loss": 2.099,
+      "step": 415
+    },
+    {
+      "epoch": 0.028300933256965737,
+      "grad_norm": 5.6923669778496375,
+      "learning_rate": 2.830188679245283e-07,
+      "loss": 2.1419,
+      "step": 420
+    },
+    {
+      "epoch": 0.02863784912907247,
+      "grad_norm": 5.232069094231692,
+      "learning_rate": 2.863881401617251e-07,
+      "loss": 2.1309,
+      "step": 425
+    },
+    {
+      "epoch": 0.028974765001179206,
+      "grad_norm": 5.12385346701768,
+      "learning_rate": 2.897574123989218e-07,
+      "loss": 2.0957,
+      "step": 430
+    },
+    {
+      "epoch": 0.02931168087328594,
+      "grad_norm": 5.043874118840126,
+      "learning_rate": 2.9312668463611853e-07,
+      "loss": 2.0322,
+      "step": 435
+    },
+    {
+      "epoch": 0.029648596745392674,
+      "grad_norm": 5.5658668729227045,
+      "learning_rate": 2.9649595687331536e-07,
+      "loss": 2.136,
+      "step": 440
+    },
+    {
+      "epoch": 0.02998551261749941,
+      "grad_norm": 10.65312954737502,
+      "learning_rate": 2.9986522911051207e-07,
+      "loss": 2.0633,
+      "step": 445
+    },
+    {
+      "epoch": 0.030322428489606146,
+      "grad_norm": 5.397097363208325,
+      "learning_rate": 3.032345013477089e-07,
+      "loss": 2.0639,
+      "step": 450
+    },
+    {
+      "epoch": 0.03065934436171288,
+      "grad_norm": 5.639079239529771,
+      "learning_rate": 3.066037735849056e-07,
+      "loss": 2.1089,
+      "step": 455
+    },
+    {
+      "epoch": 0.030996260233819614,
+      "grad_norm": 6.098761612302749,
+      "learning_rate": 3.0997304582210244e-07,
+      "loss": 2.0253,
+      "step": 460
+    },
+    {
+      "epoch": 0.03133317610592635,
+      "grad_norm": 5.075909432856418,
+      "learning_rate": 3.1334231805929915e-07,
+      "loss": 2.107,
+      "step": 465
+    },
+    {
+      "epoch": 0.03167009197803308,
+      "grad_norm": 5.387955375678346,
+      "learning_rate": 3.16711590296496e-07,
+      "loss": 2.0536,
+      "step": 470
+    },
+    {
+      "epoch": 0.03200700785013982,
+      "grad_norm": 5.522928174932781,
+      "learning_rate": 3.200808625336927e-07,
+      "loss": 2.1081,
+      "step": 475
+    },
+    {
+      "epoch": 0.03234392372224656,
+      "grad_norm": 5.024688066890421,
+      "learning_rate": 3.2345013477088946e-07,
+      "loss": 2.0879,
+      "step": 480
+    },
+    {
+      "epoch": 0.03268083959435329,
+      "grad_norm": 5.2368353353349155,
+      "learning_rate": 3.2681940700808623e-07,
+      "loss": 2.0918,
+      "step": 485
+    },
+    {
+      "epoch": 0.033017755466460026,
+      "grad_norm": 5.118887611173772,
+      "learning_rate": 3.30188679245283e-07,
+      "loss": 2.025,
+      "step": 490
+    },
+    {
+      "epoch": 0.03335467133856676,
+      "grad_norm": 5.343662616511924,
+      "learning_rate": 3.3355795148247977e-07,
+      "loss": 2.1206,
+      "step": 495
+    },
+    {
+      "epoch": 0.033691587210673495,
+      "grad_norm": 5.28143943704075,
+      "learning_rate": 3.3692722371967654e-07,
+      "loss": 2.0416,
+      "step": 500
+    },
+    {
+      "epoch": 0.03402850308278023,
+      "grad_norm": 5.740167633758656,
+      "learning_rate": 3.402964959568733e-07,
+      "loss": 2.0,
+      "step": 505
+    },
+    {
+      "epoch": 0.03436541895488696,
+      "grad_norm": 6.130295525717783,
+      "learning_rate": 3.436657681940701e-07,
+      "loss": 2.044,
+      "step": 510
+    },
+    {
+      "epoch": 0.0347023348269937,
+      "grad_norm": 6.034524610430405,
+      "learning_rate": 3.4703504043126685e-07,
+      "loss": 2.0904,
+      "step": 515
+    },
+    {
+      "epoch": 0.03503925069910043,
+      "grad_norm": 5.618706136600561,
+      "learning_rate": 3.504043126684636e-07,
+      "loss": 2.1239,
+      "step": 520
+    },
+    {
+      "epoch": 0.03537616657120717,
+      "grad_norm": 5.5845268552377485,
+      "learning_rate": 3.5377358490566033e-07,
+      "loss": 2.0664,
+      "step": 525
+    },
+    {
+      "epoch": 0.03571308244331391,
+      "grad_norm": 4.924388794395622,
+      "learning_rate": 3.5714285714285716e-07,
+      "loss": 2.0291,
+      "step": 530
+    },
+    {
+      "epoch": 0.03604999831542064,
+      "grad_norm": 5.698445960129494,
+      "learning_rate": 3.605121293800539e-07,
+      "loss": 2.075,
+      "step": 535
+    },
+    {
+      "epoch": 0.036386914187527375,
+      "grad_norm": 5.7867177958187,
+      "learning_rate": 3.638814016172507e-07,
+      "loss": 2.0166,
+      "step": 540
+    },
+    {
+      "epoch": 0.03672383005963411,
+      "grad_norm": 5.377350472058835,
+      "learning_rate": 3.672506738544474e-07,
+      "loss": 2.056,
+      "step": 545
+    },
+    {
+      "epoch": 0.03706074593174084,
+      "grad_norm": 5.119390903437156,
+      "learning_rate": 3.706199460916442e-07,
+      "loss": 2.0499,
+      "step": 550
+    },
+    {
+      "epoch": 0.03739766180384758,
+      "grad_norm": 4.942708782421948,
+      "learning_rate": 3.7398921832884095e-07,
+      "loss": 2.0276,
+      "step": 555
+    },
+    {
+      "epoch": 0.03773457767595431,
+      "grad_norm": 5.248745016083541,
+      "learning_rate": 3.773584905660377e-07,
+      "loss": 2.0444,
+      "step": 560
+    },
+    {
+      "epoch": 0.038071493548061046,
+      "grad_norm": 5.055769934271325,
+      "learning_rate": 3.807277628032345e-07,
+      "loss": 1.9998,
+      "step": 565
+    },
+    {
+      "epoch": 0.03840840942016779,
+      "grad_norm": 5.432243157493703,
+      "learning_rate": 3.840970350404312e-07,
+      "loss": 2.0655,
+      "step": 570
+    },
+    {
+      "epoch": 0.03874532529227452,
+      "grad_norm": 4.9675516385155225,
+      "learning_rate": 3.8746630727762803e-07,
+      "loss": 1.9969,
+      "step": 575
+    },
+    {
+      "epoch": 0.039082241164381255,
+      "grad_norm": 5.378207728638795,
+      "learning_rate": 3.9083557951482475e-07,
+      "loss": 2.0253,
+      "step": 580
+    },
+    {
+      "epoch": 0.03941915703648799,
+      "grad_norm": 5.294707319370417,
+      "learning_rate": 3.9420485175202157e-07,
+      "loss": 2.0218,
+      "step": 585
+    },
+    {
+      "epoch": 0.039756072908594724,
+      "grad_norm": 5.715084163910885,
+      "learning_rate": 3.975741239892183e-07,
+      "loss": 2.0568,
+      "step": 590
+    },
+    {
+      "epoch": 0.04009298878070146,
+      "grad_norm": 6.898250059855762,
+      "learning_rate": 4.009433962264151e-07,
+      "loss": 2.0793,
+      "step": 595
+    },
+    {
+      "epoch": 0.04042990465280819,
+      "grad_norm": 5.468487143259513,
+      "learning_rate": 4.043126684636118e-07,
+      "loss": 2.0028,
+      "step": 600
+    },
+    {
+      "epoch": 0.040766820524914926,
+      "grad_norm": 5.730269710778106,
+      "learning_rate": 4.076819407008086e-07,
+      "loss": 2.0607,
+      "step": 605
+    },
+    {
+      "epoch": 0.04110373639702166,
+      "grad_norm": 5.79169857641335,
+      "learning_rate": 4.1105121293800537e-07,
+      "loss": 2.0516,
+      "step": 610
+    },
+    {
+      "epoch": 0.0414406522691284,
+      "grad_norm": 5.635207608283501,
+      "learning_rate": 4.1442048517520213e-07,
+      "loss": 2.0199,
+      "step": 615
+    },
+    {
+      "epoch": 0.041777568141235136,
+      "grad_norm": 5.521975548236661,
+      "learning_rate": 4.177897574123989e-07,
+      "loss": 2.0305,
+      "step": 620
+    },
+    {
+      "epoch": 0.04211448401334187,
+      "grad_norm": 5.083944440185206,
+      "learning_rate": 4.211590296495957e-07,
+      "loss": 2.0284,
+      "step": 625
+    },
+    {
+      "epoch": 0.042451399885448604,
+      "grad_norm": 5.377736048836437,
+      "learning_rate": 4.2452830188679244e-07,
+      "loss": 1.9927,
+      "step": 630
+    },
+    {
+      "epoch": 0.04278831575755534,
+      "grad_norm": 5.232868294304897,
+      "learning_rate": 4.278975741239892e-07,
+      "loss": 2.0419,
+      "step": 635
+    },
+    {
+      "epoch": 0.04312523162966207,
+      "grad_norm": 6.182464484906801,
+      "learning_rate": 4.31266846361186e-07,
+      "loss": 2.0166,
+      "step": 640
+    },
+    {
+      "epoch": 0.04346214750176881,
+      "grad_norm": 5.429049038299917,
+      "learning_rate": 4.3463611859838275e-07,
+      "loss": 1.9505,
+      "step": 645
+    },
+    {
+      "epoch": 0.04379906337387554,
+      "grad_norm": 5.350503007594097,
+      "learning_rate": 4.3800539083557947e-07,
+      "loss": 1.9925,
+      "step": 650
+    },
+    {
+      "epoch": 0.044135979245982275,
+      "grad_norm": 5.150651978378095,
+      "learning_rate": 4.413746630727763e-07,
+      "loss": 2.0446,
+      "step": 655
+    },
+    {
+      "epoch": 0.044472895118089016,
+      "grad_norm": 5.521163370130274,
+      "learning_rate": 4.44743935309973e-07,
+      "loss": 1.9818,
+      "step": 660
+    },
+    {
+      "epoch": 0.04480981099019575,
+      "grad_norm": 5.694227421727922,
+      "learning_rate": 4.481132075471698e-07,
+      "loss": 2.0312,
+      "step": 665
+    },
+    {
+      "epoch": 0.045146726862302484,
+      "grad_norm": 5.069392194962212,
+      "learning_rate": 4.5148247978436655e-07,
+      "loss": 2.0187,
+      "step": 670
+    },
+    {
+      "epoch": 0.04548364273440922,
+      "grad_norm": 5.147985295207899,
+      "learning_rate": 4.548517520215633e-07,
+      "loss": 2.0533,
+      "step": 675
+    },
+    {
+      "epoch": 0.04582055860651595,
+      "grad_norm": 5.647364286105832,
+      "learning_rate": 4.582210242587601e-07,
+      "loss": 2.0218,
+      "step": 680
+    },
+    {
+      "epoch": 0.04615747447862269,
+      "grad_norm": 5.296105092900961,
+      "learning_rate": 4.6159029649595686e-07,
+      "loss": 2.0438,
+      "step": 685
+    },
+    {
+      "epoch": 0.04649439035072942,
+      "grad_norm": 5.431070724140859,
+      "learning_rate": 4.649595687331536e-07,
+      "loss": 1.9468,
+      "step": 690
+    },
+    {
+      "epoch": 0.046831306222836155,
+      "grad_norm": 5.338564248947676,
+      "learning_rate": 4.6832884097035034e-07,
+      "loss": 1.9921,
+      "step": 695
+    },
+    {
+      "epoch": 0.04716822209494289,
+      "grad_norm": 5.467178934661103,
+      "learning_rate": 4.7169811320754717e-07,
+      "loss": 1.9582,
+      "step": 700
+    },
+    {
+      "epoch": 0.04750513796704963,
+      "grad_norm": 5.209811996274768,
+      "learning_rate": 4.750673854447439e-07,
+      "loss": 1.9878,
+      "step": 705
+    },
+    {
+      "epoch": 0.047842053839156365,
+      "grad_norm": 5.325725411751291,
+      "learning_rate": 4.784366576819407e-07,
+      "loss": 2.0557,
+      "step": 710
+    },
+    {
+      "epoch": 0.0481789697112631,
+      "grad_norm": 5.612384744469257,
+      "learning_rate": 4.818059299191375e-07,
+      "loss": 1.9881,
+      "step": 715
+    },
+    {
+      "epoch": 0.04851588558336983,
+      "grad_norm": 4.98603677284012,
+      "learning_rate": 4.851752021563342e-07,
+      "loss": 1.9988,
+      "step": 720
+    },
+    {
+      "epoch": 0.04885280145547657,
+      "grad_norm": 5.533953690217046,
+      "learning_rate": 4.88544474393531e-07,
+      "loss": 1.9702,
+      "step": 725
+    },
+    {
+      "epoch": 0.0491897173275833,
+      "grad_norm": 5.279439473276112,
+      "learning_rate": 4.919137466307277e-07,
+      "loss": 2.0248,
+      "step": 730
+    },
+    {
+      "epoch": 0.049526633199690036,
+      "grad_norm": 5.485075105507988,
+      "learning_rate": 4.952830188679246e-07,
+      "loss": 1.9812,
+      "step": 735
+    },
+    {
+      "epoch": 0.04986354907179677,
+      "grad_norm": 5.761664060674013,
+      "learning_rate": 4.986522911051213e-07,
+      "loss": 1.9241,
+      "step": 740
+    },
+    {
+      "epoch": 0.050200464943903504,
+      "grad_norm": 5.314097467258596,
+      "learning_rate": 5.020215633423181e-07,
+      "loss": 2.0161,
+      "step": 745
+    },
+    {
+      "epoch": 0.050537380816010245,
+      "grad_norm": 5.378664311376735,
+      "learning_rate": 5.053908355795148e-07,
+      "loss": 1.9518,
+      "step": 750
+    },
+    {
+      "epoch": 0.05087429668811698,
+      "grad_norm": 4.691452253547841,
+      "learning_rate": 5.087601078167115e-07,
+      "loss": 2.0194,
+      "step": 755
+    },
+    {
+      "epoch": 0.051211212560223714,
+      "grad_norm": 5.3831522238040845,
+      "learning_rate": 5.121293800539083e-07,
+      "loss": 1.999,
+      "step": 760
+    },
+    {
+      "epoch": 0.05154812843233045,
+      "grad_norm": 4.976811429217899,
+      "learning_rate": 5.154986522911052e-07,
+      "loss": 2.0192,
+      "step": 765
+    },
+    {
+      "epoch": 0.05188504430443718,
+      "grad_norm": 5.581599453910726,
+      "learning_rate": 5.188679245283019e-07,
+      "loss": 2.026,
+      "step": 770
+    },
+    {
+      "epoch": 0.052221960176543916,
+      "grad_norm": 5.243566709461788,
+      "learning_rate": 5.222371967654986e-07,
+      "loss": 1.9774,
+      "step": 775
+    },
+    {
+      "epoch": 0.05255887604865065,
+      "grad_norm": 21.240219094200093,
+      "learning_rate": 5.256064690026953e-07,
+      "loss": 1.9668,
+      "step": 780
+    },
+    {
+      "epoch": 0.052895791920757385,
+      "grad_norm": 5.554100631042519,
+      "learning_rate": 5.289757412398921e-07,
+      "loss": 2.0255,
+      "step": 785
+    },
+    {
+      "epoch": 0.05323270779286412,
+      "grad_norm": 5.40970243183288,
+      "learning_rate": 5.32345013477089e-07,
+      "loss": 1.9893,
+      "step": 790
+    },
+    {
+      "epoch": 0.05356962366497086,
+      "grad_norm": 5.0017195942526245,
+      "learning_rate": 5.357142857142857e-07,
+      "loss": 1.9773,
+      "step": 795
+    },
+    {
+      "epoch": 0.053906539537077594,
+      "grad_norm": 4.526154717874087,
+      "learning_rate": 5.390835579514824e-07,
+      "loss": 1.9305,
+      "step": 800
+    },
+    {
+      "epoch": 0.05424345540918433,
+      "grad_norm": 5.4425678388497865,
+      "learning_rate": 5.424528301886792e-07,
+      "loss": 1.9608,
+      "step": 805
+    },
+    {
+      "epoch": 0.05458037128129106,
+      "grad_norm": 4.623913713740688,
+      "learning_rate": 5.45822102425876e-07,
+      "loss": 2.0188,
+      "step": 810
+    },
+    {
+      "epoch": 0.0549172871533978,
+      "grad_norm": 5.881864605300756,
+      "learning_rate": 5.491913746630728e-07,
+      "loss": 1.9012,
+      "step": 815
+    },
+    {
+      "epoch": 0.05525420302550453,
+      "grad_norm": 5.011736843819804,
+      "learning_rate": 5.525606469002695e-07,
+      "loss": 2.019,
+      "step": 820
+    },
+    {
+      "epoch": 0.055591118897611265,
+      "grad_norm": 5.409773765863313,
+      "learning_rate": 5.559299191374662e-07,
+      "loss": 1.9922,
+      "step": 825
+    },
+    {
+      "epoch": 0.055928034769718,
+      "grad_norm": 4.969421140342336,
+      "learning_rate": 5.592991913746631e-07,
+      "loss": 1.9715,
+      "step": 830
+    },
+    {
+      "epoch": 0.05626495064182473,
+      "grad_norm": 5.4543763327476364,
+      "learning_rate": 5.626684636118598e-07,
+      "loss": 1.9817,
+      "step": 835
+    },
+    {
+      "epoch": 0.056601866513931474,
+      "grad_norm": 5.077339487080564,
+      "learning_rate": 5.660377358490566e-07,
+      "loss": 1.957,
+      "step": 840
+    },
+    {
+      "epoch": 0.05693878238603821,
+      "grad_norm": 5.078353916883788,
+      "learning_rate": 5.694070080862533e-07,
+      "loss": 1.9778,
+      "step": 845
+    },
+    {
+      "epoch": 0.05727569825814494,
+      "grad_norm": 5.007243744968121,
+      "learning_rate": 5.727762803234502e-07,
+      "loss": 2.014,
+      "step": 850
+    },
+    {
+      "epoch": 0.05761261413025168,
+      "grad_norm": 4.9317975602111925,
+      "learning_rate": 5.761455525606469e-07,
+      "loss": 2.0239,
+      "step": 855
+    },
+    {
+      "epoch": 0.05794953000235841,
+      "grad_norm": 4.761656162239346,
+      "learning_rate": 5.795148247978436e-07,
+      "loss": 2.0533,
+      "step": 860
+    },
+    {
+      "epoch": 0.058286445874465145,
+      "grad_norm": 5.385370063131065,
+      "learning_rate": 5.828840970350404e-07,
+      "loss": 1.9105,
+      "step": 865
+    },
+    {
+      "epoch": 0.05862336174657188,
+      "grad_norm": 5.30113234532175,
+      "learning_rate": 5.862533692722371e-07,
+      "loss": 1.9529,
+      "step": 870
+    },
+    {
+      "epoch": 0.058960277618678614,
+      "grad_norm": 6.9033659840878325,
+      "learning_rate": 5.89622641509434e-07,
+      "loss": 1.9739,
+      "step": 875
+    },
+    {
+      "epoch": 0.05929719349078535,
+      "grad_norm": 5.429285070930471,
+      "learning_rate": 5.929919137466307e-07,
+      "loss": 1.9436,
+      "step": 880
+    },
+    {
+      "epoch": 0.05963410936289209,
+      "grad_norm": 6.389343970685152,
+      "learning_rate": 5.963611859838274e-07,
+      "loss": 1.9751,
+      "step": 885
+    },
+    {
+      "epoch": 0.05997102523499882,
+      "grad_norm": 5.154716173144622,
+      "learning_rate": 5.997304582210241e-07,
+      "loss": 1.9629,
+      "step": 890
+    },
+    {
+      "epoch": 0.06030794110710556,
+      "grad_norm": 5.347558988304943,
+      "learning_rate": 6.030997304582211e-07,
+      "loss": 1.9282,
+      "step": 895
+    },
+    {
+      "epoch": 0.06064485697921229,
+      "grad_norm": 5.41923416719515,
+      "learning_rate": 6.064690026954178e-07,
+      "loss": 1.9376,
+      "step": 900
+    },
+    {
+      "epoch": 0.060981772851319026,
+      "grad_norm": 5.044466204349174,
+      "learning_rate": 6.098382749326145e-07,
+      "loss": 1.939,
+      "step": 905
+    },
+    {
+      "epoch": 0.06131868872342576,
+      "grad_norm": 5.186923659997595,
+      "learning_rate": 6.132075471698112e-07,
+      "loss": 1.9401,
+      "step": 910
+    },
+    {
+      "epoch": 0.061655604595532494,
+      "grad_norm": 6.170059318329579,
+      "learning_rate": 6.16576819407008e-07,
+      "loss": 1.9413,
+      "step": 915
+    },
+    {
+      "epoch": 0.06199252046763923,
+      "grad_norm": 5.228653101504448,
+      "learning_rate": 6.199460916442049e-07,
+      "loss": 1.9087,
+      "step": 920
+    },
+    {
+      "epoch": 0.06232943633974596,
+      "grad_norm": 4.741572431296043,
+      "learning_rate": 6.233153638814016e-07,
+      "loss": 1.94,
+      "step": 925
+    },
+    {
+      "epoch": 0.0626663522118527,
+      "grad_norm": 5.557072325156039,
+      "learning_rate": 6.266846361185983e-07,
+      "loss": 1.9267,
+      "step": 930
+    },
+    {
+      "epoch": 0.06300326808395944,
+      "grad_norm": 4.869727347438024,
+      "learning_rate": 6.300539083557951e-07,
+      "loss": 1.9688,
+      "step": 935
+    },
+    {
+      "epoch": 0.06334018395606617,
+      "grad_norm": 5.385355577314501,
+      "learning_rate": 6.33423180592992e-07,
+      "loss": 1.9268,
+      "step": 940
+    },
+    {
+      "epoch": 0.0636770998281729,
+      "grad_norm": 5.058388758820075,
+      "learning_rate": 6.367924528301887e-07,
+      "loss": 1.9199,
+      "step": 945
+    },
+    {
+      "epoch": 0.06401401570027963,
+      "grad_norm": 4.919811186559829,
+      "learning_rate": 6.401617250673854e-07,
+      "loss": 1.9054,
+      "step": 950
+    },
+    {
+      "epoch": 0.06435093157238637,
+      "grad_norm": 5.114782560307836,
+      "learning_rate": 6.435309973045822e-07,
+      "loss": 2.0087,
+      "step": 955
+    },
+    {
+      "epoch": 0.06468784744449312,
+      "grad_norm": 5.266029184598451,
+      "learning_rate": 6.469002695417789e-07,
+      "loss": 1.9631,
+      "step": 960
+    },
+    {
+      "epoch": 0.06502476331659984,
+      "grad_norm": 5.272345169294871,
+      "learning_rate": 6.502695417789757e-07,
+      "loss": 1.9649,
+      "step": 965
+    },
+    {
+      "epoch": 0.06536167918870658,
+      "grad_norm": 5.531522696661456,
+      "learning_rate": 6.536388140161725e-07,
+      "loss": 1.8831,
+      "step": 970
+    },
+    {
+      "epoch": 0.06569859506081331,
+      "grad_norm": 5.243060472631993,
+      "learning_rate": 6.570080862533693e-07,
+      "loss": 1.8948,
+      "step": 975
+    },
+    {
+      "epoch": 0.06603551093292005,
+      "grad_norm": 4.973394162552281,
+      "learning_rate": 6.60377358490566e-07,
+      "loss": 1.8546,
+      "step": 980
+    },
+    {
+      "epoch": 0.06637242680502678,
+      "grad_norm": 4.949589283745308,
+      "learning_rate": 6.637466307277628e-07,
+      "loss": 1.9344,
+      "step": 985
+    },
+    {
+      "epoch": 0.06670934267713352,
+      "grad_norm": 4.897598836158783,
+      "learning_rate": 6.671159029649595e-07,
+      "loss": 1.8681,
+      "step": 990
+    },
+    {
+      "epoch": 0.06704625854924025,
+      "grad_norm": 4.953043075542081,
+      "learning_rate": 6.704851752021563e-07,
+      "loss": 1.934,
+      "step": 995
+    },
+    {
+      "epoch": 0.06738317442134699,
+      "grad_norm": 5.051343081890513,
+      "learning_rate": 6.738544474393531e-07,
+      "loss": 1.9747,
+      "step": 1000
+    },
+    {
+      "epoch": 0.06772009029345373,
+      "grad_norm": 5.519843073235885,
+      "learning_rate": 6.772237196765498e-07,
+      "loss": 1.9198,
+      "step": 1005
+    },
+    {
+      "epoch": 0.06805700616556046,
+      "grad_norm": 6.92952990394133,
+      "learning_rate": 6.805929919137466e-07,
+      "loss": 1.9353,
+      "step": 1010
+    },
+    {
+      "epoch": 0.0683939220376672,
+      "grad_norm": 5.844500928830597,
+      "learning_rate": 6.839622641509433e-07,
+      "loss": 1.9603,
+      "step": 1015
+    },
+    {
+      "epoch": 0.06873083790977393,
+      "grad_norm": 4.680109015678994,
+      "learning_rate": 6.873315363881402e-07,
+      "loss": 1.9097,
+      "step": 1020
+    },
+    {
+      "epoch": 0.06906775378188067,
+      "grad_norm": 5.276507888101029,
+      "learning_rate": 6.907008086253369e-07,
+      "loss": 1.9169,
+      "step": 1025
+    },
+    {
+      "epoch": 0.0694046696539874,
+      "grad_norm": 4.597017974421258,
+      "learning_rate": 6.940700808625337e-07,
+      "loss": 1.866,
+      "step": 1030
+    },
+    {
+      "epoch": 0.06974158552609414,
+      "grad_norm": 5.248577301108423,
+      "learning_rate": 6.974393530997304e-07,
+      "loss": 1.9707,
+      "step": 1035
+    },
+    {
+      "epoch": 0.07007850139820086,
+      "grad_norm": 4.561722496535128,
+      "learning_rate": 7.008086253369272e-07,
+      "loss": 1.9717,
+      "step": 1040
+    },
+    {
+      "epoch": 0.0704154172703076,
+      "grad_norm": 5.822968995110593,
+      "learning_rate": 7.04177897574124e-07,
+      "loss": 1.9476,
+      "step": 1045
+    },
+    {
+      "epoch": 0.07075233314241434,
+      "grad_norm": 5.567490532077557,
+      "learning_rate": 7.075471698113207e-07,
+      "loss": 1.9454,
+      "step": 1050
+    },
+    {
+      "epoch": 0.07108924901452107,
+      "grad_norm": 5.073451663570337,
+      "learning_rate": 7.109164420485175e-07,
+      "loss": 1.9323,
+      "step": 1055
+    },
+    {
+      "epoch": 0.07142616488662781,
+      "grad_norm": 5.146228422818076,
+      "learning_rate": 7.142857142857143e-07,
+      "loss": 1.9075,
+      "step": 1060
+    },
+    {
+      "epoch": 0.07176308075873454,
+      "grad_norm": 4.631295463779331,
+      "learning_rate": 7.17654986522911e-07,
+      "loss": 1.8698,
+      "step": 1065
+    },
+    {
+      "epoch": 0.07209999663084128,
+      "grad_norm": 5.1593056457079225,
+      "learning_rate": 7.210242587601077e-07,
+      "loss": 1.9247,
+      "step": 1070
+    },
+    {
+      "epoch": 0.07243691250294801,
+      "grad_norm": 5.48417366428753,
+      "learning_rate": 7.243935309973046e-07,
+      "loss": 1.941,
+      "step": 1075
+    },
+    {
+      "epoch": 0.07277382837505475,
+      "grad_norm": 5.631796760221222,
+      "learning_rate": 7.277628032345014e-07,
+      "loss": 1.9156,
+      "step": 1080
+    },
+    {
+      "epoch": 0.07311074424716148,
+      "grad_norm": 4.777529076813078,
+      "learning_rate": 7.311320754716981e-07,
+      "loss": 1.927,
+      "step": 1085
+    },
+    {
+      "epoch": 0.07344766011926822,
+      "grad_norm": 4.90129939841515,
+      "learning_rate": 7.345013477088948e-07,
+      "loss": 1.9781,
+      "step": 1090
+    },
+    {
+      "epoch": 0.07378457599137496,
+      "grad_norm": 5.231366709132192,
+      "learning_rate": 7.378706199460915e-07,
+      "loss": 1.9127,
+      "step": 1095
+    },
+    {
+      "epoch": 0.07412149186348169,
+      "grad_norm": 4.723481493128406,
+      "learning_rate": 7.412398921832884e-07,
+      "loss": 1.918,
+      "step": 1100
+    },
+    {
+      "epoch": 0.07445840773558843,
+      "grad_norm": 5.201673325926327,
+      "learning_rate": 7.446091644204852e-07,
+      "loss": 1.9476,
+      "step": 1105
+    },
+    {
+      "epoch": 0.07479532360769515,
+      "grad_norm": 5.461855674554181,
+      "learning_rate": 7.479784366576819e-07,
+      "loss": 1.9227,
+      "step": 1110
+    },
+    {
+      "epoch": 0.0751322394798019,
+      "grad_norm": 5.122573096291021,
+      "learning_rate": 7.513477088948786e-07,
+      "loss": 1.9069,
+      "step": 1115
+    },
+    {
+      "epoch": 0.07546915535190862,
+      "grad_norm": 5.134161925716819,
+      "learning_rate": 7.547169811320754e-07,
+      "loss": 1.8723,
+      "step": 1120
+    },
+    {
+      "epoch": 0.07580607122401536,
+      "grad_norm": 4.912594707407192,
+      "learning_rate": 7.580862533692723e-07,
+      "loss": 1.925,
+      "step": 1125
+    },
+    {
+      "epoch": 0.07614298709612209,
+      "grad_norm": 4.833642336502445,
+      "learning_rate": 7.61455525606469e-07,
+      "loss": 1.9418,
+      "step": 1130
+    },
+    {
+      "epoch": 0.07647990296822883,
+      "grad_norm": 5.942285402534889,
+      "learning_rate": 7.648247978436657e-07,
+      "loss": 1.891,
+      "step": 1135
+    },
+    {
+      "epoch": 0.07681681884033557,
+      "grad_norm": 4.780486374896709,
+      "learning_rate": 7.681940700808624e-07,
+      "loss": 1.9281,
+      "step": 1140
+    },
+    {
+      "epoch": 0.0771537347124423,
+      "grad_norm": 5.0599976612815425,
+      "learning_rate": 7.715633423180593e-07,
+      "loss": 1.8856,
+      "step": 1145
+    },
+    {
+      "epoch": 0.07749065058454904,
+      "grad_norm": 5.447714555152886,
+      "learning_rate": 7.749326145552561e-07,
+      "loss": 1.9193,
+      "step": 1150
+    },
+    {
+      "epoch": 0.07782756645665577,
+      "grad_norm": 5.559443100359282,
+      "learning_rate": 7.783018867924528e-07,
+      "loss": 1.8755,
+      "step": 1155
+    },
+    {
+      "epoch": 0.07816448232876251,
+      "grad_norm": 9.485730862885632,
+      "learning_rate": 7.816711590296495e-07,
+      "loss": 1.8853,
+      "step": 1160
+    },
+    {
+      "epoch": 0.07850139820086924,
+      "grad_norm": 5.074665439939694,
+      "learning_rate": 7.850404312668463e-07,
+      "loss": 1.9718,
+      "step": 1165
+    },
+    {
+      "epoch": 0.07883831407297598,
+      "grad_norm": 5.35289626298854,
+      "learning_rate": 7.884097035040431e-07,
+      "loss": 1.9016,
+      "step": 1170
+    },
+    {
+      "epoch": 0.0791752299450827,
+      "grad_norm": 4.849839407058772,
+      "learning_rate": 7.917789757412399e-07,
+      "loss": 1.9096,
+      "step": 1175
+    },
+    {
+      "epoch": 0.07951214581718945,
+      "grad_norm": 5.292874310019159,
+      "learning_rate": 7.951482479784366e-07,
+      "loss": 1.8688,
+      "step": 1180
+    },
+    {
+      "epoch": 0.07984906168929619,
+      "grad_norm": 4.69518465130648,
+      "learning_rate": 7.985175202156334e-07,
+      "loss": 1.8804,
+      "step": 1185
+    },
+    {
+      "epoch": 0.08018597756140292,
+      "grad_norm": 5.330973093669295,
+      "learning_rate": 8.018867924528302e-07,
+      "loss": 1.9425,
+      "step": 1190
+    },
+    {
+      "epoch": 0.08052289343350966,
+      "grad_norm": 5.2166635094871,
+      "learning_rate": 8.052560646900269e-07,
+      "loss": 1.873,
+      "step": 1195
+    },
+    {
+      "epoch": 0.08085980930561638,
+      "grad_norm": 5.036327896074128,
+      "learning_rate": 8.086253369272237e-07,
+      "loss": 1.8749,
+      "step": 1200
+    },
+    {
+      "epoch": 0.08119672517772313,
+      "grad_norm": 4.569279981257993,
+      "learning_rate": 8.119946091644204e-07,
+      "loss": 1.908,
+      "step": 1205
+    },
+    {
+      "epoch": 0.08153364104982985,
+      "grad_norm": 5.277033519733169,
+      "learning_rate": 8.153638814016172e-07,
+      "loss": 1.9629,
+      "step": 1210
+    },
+    {
+      "epoch": 0.0818705569219366,
+      "grad_norm": 4.646480429400201,
+      "learning_rate": 8.18733153638814e-07,
+      "loss": 1.9139,
+      "step": 1215
+    },
+    {
+      "epoch": 0.08220747279404332,
+      "grad_norm": 5.355634775770506,
+      "learning_rate": 8.221024258760107e-07,
+      "loss": 1.8631,
+      "step": 1220
+    },
+    {
+      "epoch": 0.08254438866615006,
+      "grad_norm": 7.811459226424328,
+      "learning_rate": 8.254716981132074e-07,
+      "loss": 1.8857,
+      "step": 1225
+    },
+    {
+      "epoch": 0.0828813045382568,
+      "grad_norm": 10.013535553366285,
+      "learning_rate": 8.288409703504043e-07,
+      "loss": 1.9353,
+      "step": 1230
+    },
+    {
+      "epoch": 0.08321822041036353,
+      "grad_norm": 4.547839224400193,
+      "learning_rate": 8.322102425876011e-07,
+      "loss": 1.8719,
+      "step": 1235
+    },
+    {
+      "epoch": 0.08355513628247027,
+      "grad_norm": 4.994456574147338,
+      "learning_rate": 8.355795148247978e-07,
+      "loss": 1.9409,
+      "step": 1240
+    },
+    {
+      "epoch": 0.083892052154577,
+      "grad_norm": 4.320607866056009,
+      "learning_rate": 8.389487870619945e-07,
+      "loss": 1.9094,
+      "step": 1245
+    },
+    {
+      "epoch": 0.08422896802668374,
+      "grad_norm": 5.1285978622421675,
+      "learning_rate": 8.423180592991913e-07,
+      "loss": 1.8423,
+      "step": 1250
+    },
+    {
+      "epoch": 0.08456588389879047,
+      "grad_norm": 5.207593792225735,
+      "learning_rate": 8.456873315363881e-07,
+      "loss": 1.9206,
+      "step": 1255
+    },
+    {
+      "epoch": 0.08490279977089721,
+      "grad_norm": 5.16067220295262,
+      "learning_rate": 8.490566037735849e-07,
+      "loss": 1.8542,
+      "step": 1260
+    },
+    {
+      "epoch": 0.08523971564300394,
+      "grad_norm": 4.5869805182954275,
+      "learning_rate": 8.524258760107816e-07,
+      "loss": 1.8442,
+      "step": 1265
+    },
+    {
+      "epoch": 0.08557663151511068,
+      "grad_norm": 5.905122236250617,
+      "learning_rate": 8.557951482479784e-07,
+      "loss": 1.9592,
+      "step": 1270
+    },
+    {
+      "epoch": 0.08591354738721742,
+      "grad_norm": 4.850519676865327,
+      "learning_rate": 8.591644204851751e-07,
+      "loss": 1.8988,
+      "step": 1275
+    },
+    {
+      "epoch": 0.08625046325932414,
+      "grad_norm": 5.246664878234509,
+      "learning_rate": 8.62533692722372e-07,
+      "loss": 1.9069,
+      "step": 1280
+    },
+    {
+      "epoch": 0.08658737913143089,
+      "grad_norm": 4.686851588740122,
+      "learning_rate": 8.659029649595687e-07,
+      "loss": 1.9093,
+      "step": 1285
+    },
+    {
+      "epoch": 0.08692429500353761,
+      "grad_norm": 4.864848050723445,
+      "learning_rate": 8.692722371967655e-07,
+      "loss": 1.9621,
+      "step": 1290
+    },
+    {
+      "epoch": 0.08726121087564435,
+      "grad_norm": 5.245150673768001,
+      "learning_rate": 8.726415094339622e-07,
+      "loss": 1.9268,
+      "step": 1295
+    },
+    {
+      "epoch": 0.08759812674775108,
+      "grad_norm": 5.586645232579203,
+      "learning_rate": 8.760107816711589e-07,
+      "loss": 1.9637,
+      "step": 1300
+    },
+    {
+      "epoch": 0.08793504261985782,
+      "grad_norm": 5.129811614944534,
+      "learning_rate": 8.793800539083558e-07,
+      "loss": 1.8807,
+      "step": 1305
+    },
+    {
+      "epoch": 0.08827195849196455,
+      "grad_norm": 5.207910430535945,
+      "learning_rate": 8.827493261455526e-07,
+      "loss": 1.9112,
+      "step": 1310
+    },
+    {
+      "epoch": 0.08860887436407129,
+      "grad_norm": 5.148253867866013,
+      "learning_rate": 8.861185983827493e-07,
+      "loss": 1.8747,
+      "step": 1315
+    },
+    {
+      "epoch": 0.08894579023617803,
+      "grad_norm": 4.959406714040152,
+      "learning_rate": 8.89487870619946e-07,
+      "loss": 1.8511,
+      "step": 1320
+    },
+    {
+      "epoch": 0.08928270610828476,
+      "grad_norm": 5.311015482505079,
+      "learning_rate": 8.928571428571428e-07,
+      "loss": 1.9621,
+      "step": 1325
+    },
+    {
+      "epoch": 0.0896196219803915,
+      "grad_norm": 5.529873429359067,
+      "learning_rate": 8.962264150943396e-07,
+      "loss": 1.8751,
+      "step": 1330
+    },
+    {
+      "epoch": 0.08995653785249823,
+      "grad_norm": 5.019885851770948,
+      "learning_rate": 8.995956873315364e-07,
+      "loss": 1.8673,
+      "step": 1335
+    },
+    {
+      "epoch": 0.09029345372460497,
+      "grad_norm": 5.255259881875886,
+      "learning_rate": 9.029649595687331e-07,
+      "loss": 1.8594,
+      "step": 1340
+    },
+    {
+      "epoch": 0.0906303695967117,
+      "grad_norm": 4.864013243574321,
+      "learning_rate": 9.063342318059298e-07,
+      "loss": 1.8578,
+      "step": 1345
+    },
+    {
+      "epoch": 0.09096728546881844,
+      "grad_norm": 4.783979712270612,
+      "learning_rate": 9.097035040431266e-07,
+      "loss": 1.8668,
+      "step": 1350
+    },
+    {
+      "epoch": 0.09130420134092516,
+      "grad_norm": 5.288947009799633,
+      "learning_rate": 9.130727762803235e-07,
+      "loss": 1.8831,
+      "step": 1355
+    },
+    {
+      "epoch": 0.0916411172130319,
+      "grad_norm": 4.563876070584222,
+      "learning_rate": 9.164420485175202e-07,
+      "loss": 1.9362,
+      "step": 1360
+    },
+    {
+      "epoch": 0.09197803308513865,
+      "grad_norm": 4.368741024793251,
+      "learning_rate": 9.198113207547169e-07,
+      "loss": 1.8827,
+      "step": 1365
+    },
+    {
+      "epoch": 0.09231494895724537,
+      "grad_norm": 4.69647349179084,
+      "learning_rate": 9.231805929919137e-07,
+      "loss": 1.8673,
+      "step": 1370
+    },
+    {
+      "epoch": 0.09265186482935212,
+      "grad_norm": 5.410173364100209,
+      "learning_rate": 9.265498652291105e-07,
+      "loss": 1.9147,
+      "step": 1375
+    },
+    {
+      "epoch": 0.09298878070145884,
+      "grad_norm": 5.026718419465974,
+      "learning_rate": 9.299191374663073e-07,
+      "loss": 1.8939,
+      "step": 1380
+    },
+    {
+      "epoch": 0.09332569657356558,
+      "grad_norm": 5.226125492850762,
+      "learning_rate": 9.33288409703504e-07,
+      "loss": 1.8999,
+      "step": 1385
+    },
+    {
+      "epoch": 0.09366261244567231,
+      "grad_norm": 4.851233426364406,
+      "learning_rate": 9.366576819407007e-07,
+      "loss": 1.8784,
+      "step": 1390
+    },
+    {
+      "epoch": 0.09399952831777905,
+      "grad_norm": 4.31752533985363,
+      "learning_rate": 9.400269541778976e-07,
+      "loss": 1.825,
+      "step": 1395
+    },
+    {
+      "epoch": 0.09433644418988578,
+      "grad_norm": 5.805848435786427,
+      "learning_rate": 9.433962264150943e-07,
+      "loss": 1.8816,
+      "step": 1400
+    },
+    {
+      "epoch": 0.09467336006199252,
+      "grad_norm": 5.604277013042164,
+      "learning_rate": 9.46765498652291e-07,
+      "loss": 1.9501,
+      "step": 1405
+    },
+    {
+      "epoch": 0.09501027593409926,
+      "grad_norm": 4.640923447032217,
+      "learning_rate": 9.501347708894878e-07,
+      "loss": 1.8088,
+      "step": 1410
+    },
+    {
+      "epoch": 0.09534719180620599,
+      "grad_norm": 4.800198454517414,
+      "learning_rate": 9.535040431266847e-07,
+      "loss": 1.9026,
+      "step": 1415
+    },
+    {
+      "epoch": 0.09568410767831273,
+      "grad_norm": 4.778232798159787,
+      "learning_rate": 9.568733153638813e-07,
+      "loss": 1.8543,
+      "step": 1420
+    },
+    {
+      "epoch": 0.09602102355041946,
+      "grad_norm": 5.920465260156262,
+      "learning_rate": 9.60242587601078e-07,
+      "loss": 1.8315,
+      "step": 1425
+    },
+    {
+      "epoch": 0.0963579394225262,
+      "grad_norm": 5.5208201239955885,
+      "learning_rate": 9.63611859838275e-07,
+      "loss": 1.9189,
+      "step": 1430
+    },
+    {
+      "epoch": 0.09669485529463293,
+      "grad_norm": 4.465241133486137,
+      "learning_rate": 9.669811320754717e-07,
+      "loss": 1.8755,
+      "step": 1435
+    },
+    {
+      "epoch": 0.09703177116673967,
+      "grad_norm": 4.861772605861353,
+      "learning_rate": 9.703504043126684e-07,
+      "loss": 1.8424,
+      "step": 1440
+    },
+    {
+      "epoch": 0.0973686870388464,
+      "grad_norm": 4.932766715467049,
+      "learning_rate": 9.73719676549865e-07,
+      "loss": 1.8656,
+      "step": 1445
+    },
+    {
+      "epoch": 0.09770560291095313,
+      "grad_norm": 4.441378456407006,
+      "learning_rate": 9.77088948787062e-07,
+      "loss": 1.8494,
+      "step": 1450
+    },
+    {
+      "epoch": 0.09804251878305988,
+      "grad_norm": 5.275736089281324,
+      "learning_rate": 9.804582210242587e-07,
+      "loss": 1.8647,
+      "step": 1455
+    },
+    {
+      "epoch": 0.0983794346551666,
+      "grad_norm": 5.12890600254249,
+      "learning_rate": 9.838274932614555e-07,
+      "loss": 1.9004,
+      "step": 1460
+    },
+    {
+      "epoch": 0.09871635052727334,
+      "grad_norm": 5.705766103530472,
+      "learning_rate": 9.871967654986522e-07,
+      "loss": 1.9016,
+      "step": 1465
+    },
+    {
+      "epoch": 0.09905326639938007,
+      "grad_norm": 5.462295643224533,
+      "learning_rate": 9.90566037735849e-07,
+      "loss": 1.8279,
+      "step": 1470
+    },
+    {
+      "epoch": 0.09939018227148681,
+      "grad_norm": 4.847596040039102,
+      "learning_rate": 9.939353099730458e-07,
+      "loss": 1.8291,
+      "step": 1475
+    },
+    {
+      "epoch": 0.09972709814359354,
+      "grad_norm": 5.3575704163339655,
+      "learning_rate": 9.973045822102425e-07,
+      "loss": 1.8753,
+      "step": 1480
+    },
+    {
+      "epoch": 0.10006401401570028,
+      "grad_norm": 5.085814023260118,
+      "learning_rate": 9.999999861679377e-07,
+      "loss": 1.7741,
+      "step": 1485
+    },
+    {
+      "epoch": 0.10040092988780701,
+      "grad_norm": 5.539515142922222,
+      "learning_rate": 9.999995020458434e-07,
+      "loss": 1.8115,
+      "step": 1490
+    },
+    {
+      "epoch": 0.10073784575991375,
+      "grad_norm": 5.104632657913346,
+      "learning_rate": 9.99998326321407e-07,
+      "loss": 1.8537,
+      "step": 1495
+    },
+    {
+      "epoch": 0.10107476163202049,
+      "grad_norm": 5.037682554763403,
+      "learning_rate": 9.999964589962556e-07,
+      "loss": 1.8051,
+      "step": 1500
+    },
+    {
+      "epoch": 0.10141167750412722,
+      "grad_norm": 5.444043728664739,
+      "learning_rate": 9.999939000729715e-07,
+      "loss": 1.8778,
+      "step": 1505
+    },
+    {
+      "epoch": 0.10174859337623396,
+      "grad_norm": 4.447541781178455,
+      "learning_rate": 9.999906495550946e-07,
+      "loss": 1.8523,
+      "step": 1510
+    },
+    {
+      "epoch": 0.10208550924834069,
+      "grad_norm": 4.975360878532206,
+      "learning_rate": 9.999867074471207e-07,
+      "loss": 1.8719,
+      "step": 1515
+    },
+    {
+      "epoch": 0.10242242512044743,
+      "grad_norm": 4.71237216626977,
+      "learning_rate": 9.99982073754503e-07,
+      "loss": 1.8011,
+      "step": 1520
+    },
+    {
+      "epoch": 0.10275934099255415,
+      "grad_norm": 4.565133365936921,
+      "learning_rate": 9.999767484836502e-07,
+      "loss": 1.8379,
+      "step": 1525
+    },
+    {
+      "epoch": 0.1030962568646609,
+      "grad_norm": 5.110474397007126,
+      "learning_rate": 9.999707316419288e-07,
+      "loss": 1.7875,
+      "step": 1530
+    },
+    {
+      "epoch": 0.10343317273676762,
+      "grad_norm": 4.720086807903922,
+      "learning_rate": 9.99964023237661e-07,
+      "loss": 1.8251,
+      "step": 1535
+    },
+    {
+      "epoch": 0.10377008860887436,
+      "grad_norm": 5.341137351778559,
+      "learning_rate": 9.999566232801261e-07,
+      "loss": 1.8335,
+      "step": 1540
+    },
+    {
+      "epoch": 0.1041070044809811,
+      "grad_norm": 4.869144151159134,
+      "learning_rate": 9.9994853177956e-07,
+      "loss": 1.836,
+      "step": 1545
+    },
+    {
+      "epoch": 0.10444392035308783,
+      "grad_norm": 5.223975721103111,
+      "learning_rate": 9.999397487471543e-07,
+      "loss": 1.9321,
+      "step": 1550
+    },
+    {
+      "epoch": 0.10478083622519457,
+      "grad_norm": 4.685138611882665,
+      "learning_rate": 9.999302741950582e-07,
+      "loss": 1.8628,
+      "step": 1555
+    },
+    {
+      "epoch": 0.1051177520973013,
+      "grad_norm": 5.270431937376266,
+      "learning_rate": 9.999201081363768e-07,
+      "loss": 1.866,
+      "step": 1560
+    },
+    {
+      "epoch": 0.10545466796940804,
+      "grad_norm": 4.9359908760483995,
+      "learning_rate": 9.99909250585172e-07,
+      "loss": 1.8685,
+      "step": 1565
+    },
+    {
+      "epoch": 0.10579158384151477,
+      "grad_norm": 4.523129173891283,
+      "learning_rate": 9.998977015564617e-07,
+      "loss": 1.9224,
+      "step": 1570
+    },
+    {
+      "epoch": 0.10612849971362151,
+      "grad_norm": 4.72577503916574,
+      "learning_rate": 9.998854610662209e-07,
+      "loss": 1.8685,
+      "step": 1575
+    },
+    {
+      "epoch": 0.10646541558572824,
+      "grad_norm": 4.875038189097672,
+      "learning_rate": 9.998725291313805e-07,
+      "loss": 1.8772,
+      "step": 1580
+    },
+    {
+      "epoch": 0.10680233145783498,
+      "grad_norm": 8.560428339730194,
+      "learning_rate": 9.998589057698283e-07,
+      "loss": 1.8745,
+      "step": 1585
+    },
+    {
+      "epoch": 0.10713924732994172,
+      "grad_norm": 4.9717356721459565,
+      "learning_rate": 9.99844591000408e-07,
+      "loss": 1.7989,
+      "step": 1590
+    },
+    {
+      "epoch": 0.10747616320204845,
+      "grad_norm": 5.700211307942382,
+      "learning_rate": 9.9982958484292e-07,
+      "loss": 1.8694,
+      "step": 1595
+    },
+    {
+      "epoch": 0.10781307907415519,
+      "grad_norm": 4.9766939545781,
+      "learning_rate": 9.99813887318121e-07,
+      "loss": 1.8043,
+      "step": 1600
+    },
+    {
+      "epoch": 0.10814999494626192,
+      "grad_norm": 5.334106550163921,
+      "learning_rate": 9.997974984477236e-07,
+      "loss": 1.8789,
+      "step": 1605
+    },
+    {
+      "epoch": 0.10848691081836866,
+      "grad_norm": 4.762969343541741,
+      "learning_rate": 9.99780418254397e-07,
+      "loss": 1.8346,
+      "step": 1610
+    },
+    {
+      "epoch": 0.10882382669047538,
+      "grad_norm": 4.805895966179308,
+      "learning_rate": 9.99762646761767e-07,
+      "loss": 1.899,
+      "step": 1615
+    },
+    {
+      "epoch": 0.10916074256258212,
+      "grad_norm": 5.009948765551529,
+      "learning_rate": 9.99744183994415e-07,
+      "loss": 1.8119,
+      "step": 1620
+    },
+    {
+      "epoch": 0.10949765843468885,
+      "grad_norm": 4.932087499469652,
+      "learning_rate": 9.997250299778788e-07,
+      "loss": 1.8756,
+      "step": 1625
+    },
+    {
+      "epoch": 0.1098345743067956,
+      "grad_norm": 5.068638682364739,
+      "learning_rate": 9.997051847386524e-07,
+      "loss": 1.8727,
+      "step": 1630
+    },
+    {
+      "epoch": 0.11017149017890233,
+      "grad_norm": 4.592133002785248,
+      "learning_rate": 9.996846483041858e-07,
+      "loss": 1.9214,
+      "step": 1635
+    },
+    {
+      "epoch": 0.11050840605100906,
+      "grad_norm": 4.843565219513783,
+      "learning_rate": 9.99663420702885e-07,
+      "loss": 1.9096,
+      "step": 1640
+    },
+    {
+      "epoch": 0.1108453219231158,
+      "grad_norm": 4.7607329350093455,
+      "learning_rate": 9.996415019641124e-07,
+      "loss": 1.8273,
+      "step": 1645
+    },
+    {
+      "epoch": 0.11118223779522253,
+      "grad_norm": 4.760017744961109,
+      "learning_rate": 9.996188921181861e-07,
+      "loss": 1.8256,
+      "step": 1650
+    },
+    {
+      "epoch": 0.11151915366732927,
+      "grad_norm": 5.063047728653858,
+      "learning_rate": 9.9959559119638e-07,
+      "loss": 1.8168,
+      "step": 1655
+    },
+    {
+      "epoch": 0.111856069539436,
+      "grad_norm": 4.936422005296017,
+      "learning_rate": 9.995715992309244e-07,
+      "loss": 1.8618,
+      "step": 1660
+    },
+    {
+      "epoch": 0.11219298541154274,
+      "grad_norm": 5.094953809987295,
+      "learning_rate": 9.995469162550048e-07,
+      "loss": 1.8063,
+      "step": 1665
+    },
+    {
+      "epoch": 0.11252990128364947,
+      "grad_norm": 5.182573498494034,
+      "learning_rate": 9.99521542302763e-07,
+      "loss": 1.8227,
+      "step": 1670
+    },
+    {
+      "epoch": 0.11286681715575621,
+      "grad_norm": 4.743404258711493,
+      "learning_rate": 9.994954774092962e-07,
+      "loss": 1.7907,
+      "step": 1675
+    },
+    {
+      "epoch": 0.11320373302786295,
+      "grad_norm": 5.202730641786091,
+      "learning_rate": 9.994687216106579e-07,
+      "loss": 1.8331,
+      "step": 1680
+    },
+    {
+      "epoch": 0.11354064889996968,
+      "grad_norm": 5.090137020982806,
+      "learning_rate": 9.994412749438564e-07,
+      "loss": 1.7829,
+      "step": 1685
+    },
+    {
+      "epoch": 0.11387756477207642,
+      "grad_norm": 4.9893038014961295,
+      "learning_rate": 9.994131374468565e-07,
+      "loss": 1.7986,
+      "step": 1690
+    },
+    {
+      "epoch": 0.11421448064418314,
+      "grad_norm": 4.492198865372942,
+      "learning_rate": 9.993843091585782e-07,
+      "loss": 1.8536,
+      "step": 1695
+    },
+    {
+      "epoch": 0.11455139651628989,
+      "grad_norm": 8.98221193041383,
+      "learning_rate": 9.993547901188966e-07,
+      "loss": 1.8143,
+      "step": 1700
+    },
+    {
+      "epoch": 0.11488831238839661,
+      "grad_norm": 4.818273163330136,
+      "learning_rate": 9.993245803686426e-07,
+      "loss": 1.7691,
+      "step": 1705
+    },
+    {
+      "epoch": 0.11522522826050335,
+      "grad_norm": 5.20970940555951,
+      "learning_rate": 9.992936799496029e-07,
+      "loss": 1.849,
+      "step": 1710
+    },
+    {
+      "epoch": 0.11556214413261008,
+      "grad_norm": 4.787586490712673,
+      "learning_rate": 9.99262088904519e-07,
+      "loss": 1.8,
+      "step": 1715
+    },
+    {
+      "epoch": 0.11589906000471682,
+      "grad_norm": 4.991841549682482,
+      "learning_rate": 9.992298072770877e-07,
+      "loss": 1.8867,
+      "step": 1720
+    },
+    {
+      "epoch": 0.11623597587682356,
+      "grad_norm": 4.725283739157587,
+      "learning_rate": 9.991968351119612e-07,
+      "loss": 1.8404,
+      "step": 1725
+    },
+    {
+      "epoch": 0.11657289174893029,
+      "grad_norm": 5.151767048741731,
+      "learning_rate": 9.991631724547467e-07,
+      "loss": 1.8317,
+      "step": 1730
+    },
+    {
+      "epoch": 0.11690980762103703,
+      "grad_norm": 4.9360054197776115,
+      "learning_rate": 9.99128819352007e-07,
+      "loss": 1.8911,
+      "step": 1735
+    },
+    {
+      "epoch": 0.11724672349314376,
+      "grad_norm": 4.885881464702479,
+      "learning_rate": 9.99093775851259e-07,
+      "loss": 1.8769,
+      "step": 1740
+    },
+    {
+      "epoch": 0.1175836393652505,
+      "grad_norm": 4.760952548326224,
+      "learning_rate": 9.990580420009755e-07,
+      "loss": 1.8642,
+      "step": 1745
+    },
+    {
+      "epoch": 0.11792055523735723,
+      "grad_norm": 4.8133902321301045,
+      "learning_rate": 9.990216178505835e-07,
+      "loss": 1.8558,
+      "step": 1750
+    },
+    {
+      "epoch": 0.11825747110946397,
+      "grad_norm": 5.18502712536443,
+      "learning_rate": 9.989845034504651e-07,
+      "loss": 1.7965,
+      "step": 1755
+    },
+    {
+      "epoch": 0.1185943869815707,
+      "grad_norm": 5.582199965208816,
+      "learning_rate": 9.989466988519572e-07,
+      "loss": 1.8008,
+      "step": 1760
+    },
+    {
+      "epoch": 0.11893130285367744,
+      "grad_norm": 5.063498967253465,
+      "learning_rate": 9.989082041073517e-07,
+      "loss": 1.7832,
+      "step": 1765
+    },
+    {
+      "epoch": 0.11926821872578418,
+      "grad_norm": 4.983703694041557,
+      "learning_rate": 9.988690192698944e-07,
+      "loss": 1.8452,
+      "step": 1770
+    },
+    {
+      "epoch": 0.1196051345978909,
+      "grad_norm": 4.984092304046178,
+      "learning_rate": 9.988291443937857e-07,
+      "loss": 1.86,
+      "step": 1775
+    },
+    {
+      "epoch": 0.11994205046999765,
+      "grad_norm": 4.616659709966719,
+      "learning_rate": 9.987885795341816e-07,
+      "loss": 1.8214,
+      "step": 1780
+    },
+    {
+      "epoch": 0.12027896634210437,
+      "grad_norm": 4.824774887957237,
+      "learning_rate": 9.987473247471908e-07,
+      "loss": 1.8151,
+      "step": 1785
+    },
+    {
+      "epoch": 0.12061588221421111,
+      "grad_norm": 5.701725337107119,
+      "learning_rate": 9.98705380089878e-07,
+      "loss": 1.8337,
+      "step": 1790
+    },
+    {
+      "epoch": 0.12095279808631784,
+      "grad_norm": 4.981571796147867,
+      "learning_rate": 9.986627456202608e-07,
+      "loss": 1.8823,
+      "step": 1795
+    },
+    {
+      "epoch": 0.12128971395842458,
+      "grad_norm": 5.360549733357877,
+      "learning_rate": 9.986194213973113e-07,
+      "loss": 1.8113,
+      "step": 1800
+    },
+    {
+      "epoch": 0.12162662983053131,
+      "grad_norm": 4.673029000872237,
+      "learning_rate": 9.985754074809562e-07,
+      "loss": 1.774,
+      "step": 1805
+    },
+    {
+      "epoch": 0.12196354570263805,
+      "grad_norm": 4.636511070093299,
+      "learning_rate": 9.985307039320756e-07,
+      "loss": 1.8114,
+      "step": 1810
+    },
+    {
+      "epoch": 0.12230046157474479,
+      "grad_norm": 4.821261166128235,
+      "learning_rate": 9.98485310812504e-07,
+      "loss": 1.8265,
+      "step": 1815
+    },
+    {
+      "epoch": 0.12263737744685152,
+      "grad_norm": 4.475912019826286,
+      "learning_rate": 9.98439228185029e-07,
+      "loss": 1.8268,
+      "step": 1820
+    },
+    {
+      "epoch": 0.12297429331895826,
+      "grad_norm": 5.24713691341664,
+      "learning_rate": 9.983924561133927e-07,
+      "loss": 1.751,
+      "step": 1825
+    },
+    {
+      "epoch": 0.12331120919106499,
+      "grad_norm": 5.333336877544052,
+      "learning_rate": 9.983449946622906e-07,
+      "loss": 1.7927,
+      "step": 1830
+    },
+    {
+      "epoch": 0.12364812506317173,
+      "grad_norm": 4.630894115286722,
+      "learning_rate": 9.982968438973714e-07,
+      "loss": 1.8225,
+      "step": 1835
+    },
+    {
+      "epoch": 0.12398504093527846,
+      "grad_norm": 4.971319322552391,
+      "learning_rate": 9.982480038852375e-07,
+      "loss": 1.8832,
+      "step": 1840
+    },
+    {
+      "epoch": 0.1243219568073852,
+      "grad_norm": 5.041019026607787,
+      "learning_rate": 9.98198474693445e-07,
+      "loss": 1.8324,
+      "step": 1845
+    },
+    {
+      "epoch": 0.12465887267949192,
+      "grad_norm": 5.096166963207779,
+      "learning_rate": 9.981482563905025e-07,
+      "loss": 1.8775,
+      "step": 1850
+    },
+    {
+      "epoch": 0.12499578855159867,
+      "grad_norm": 4.644464767921582,
+      "learning_rate": 9.980973490458728e-07,
+      "loss": 1.8348,
+      "step": 1855
+    },
+    {
+      "epoch": 0.1253327044237054,
+      "grad_norm": 4.689654175446751,
+      "learning_rate": 9.980457527299708e-07,
+      "loss": 1.8309,
+      "step": 1860
+    },
+    {
+      "epoch": 0.12566962029581213,
+      "grad_norm": 5.007707026705937,
+      "learning_rate": 9.979934675141652e-07,
+      "loss": 1.8026,
+      "step": 1865
+    },
+    {
+      "epoch": 0.12600653616791888,
+      "grad_norm": 4.935975744739416,
+      "learning_rate": 9.979404934707771e-07,
+      "loss": 1.7989,
+      "step": 1870
+    },
+    {
+      "epoch": 0.12634345204002562,
+      "grad_norm": 5.295148949180079,
+      "learning_rate": 9.978868306730804e-07,
+      "loss": 1.814,
+      "step": 1875
+    },
+    {
+      "epoch": 0.12668036791213233,
+      "grad_norm": 6.425946944455151,
+      "learning_rate": 9.978324791953018e-07,
+      "loss": 1.808,
+      "step": 1880
+    },
+    {
+      "epoch": 0.12701728378423907,
+      "grad_norm": 4.63375202539753,
+      "learning_rate": 9.97777439112621e-07,
+      "loss": 1.8116,
+      "step": 1885
+    },
+    {
+      "epoch": 0.1273541996563458,
+      "grad_norm": 5.567189408814499,
+      "learning_rate": 9.977217105011693e-07,
+      "loss": 1.7951,
+      "step": 1890
+    },
+    {
+      "epoch": 0.12769111552845255,
+      "grad_norm": 4.967271192367276,
+      "learning_rate": 9.97665293438031e-07,
+      "loss": 1.8449,
+      "step": 1895
+    },
+    {
+      "epoch": 0.12802803140055927,
+      "grad_norm": 5.083573907507721,
+      "learning_rate": 9.976081880012426e-07,
+      "loss": 1.7821,
+      "step": 1900
+    },
+    {
+      "epoch": 0.128364947272666,
+      "grad_norm": 5.263763064210331,
+      "learning_rate": 9.975503942697925e-07,
+      "loss": 1.8511,
+      "step": 1905
+    },
+    {
+      "epoch": 0.12870186314477275,
+      "grad_norm": 4.805567854667613,
+      "learning_rate": 9.974919123236217e-07,
+      "loss": 1.818,
+      "step": 1910
+    },
+    {
+      "epoch": 0.1290387790168795,
+      "grad_norm": 5.3266090858155755,
+      "learning_rate": 9.974327422436223e-07,
+      "loss": 1.7796,
+      "step": 1915
+    },
+    {
+      "epoch": 0.12937569488898623,
+      "grad_norm": 4.317364534885699,
+      "learning_rate": 9.97372884111639e-07,
+      "loss": 1.8047,
+      "step": 1920
+    },
+    {
+      "epoch": 0.12971261076109294,
+      "grad_norm": 4.877735058995506,
+      "learning_rate": 9.97312338010468e-07,
+      "loss": 1.8578,
+      "step": 1925
+    },
+    {
+      "epoch": 0.13004952663319969,
+      "grad_norm": 4.802457765477097,
+      "learning_rate": 9.97251104023857e-07,
+      "loss": 1.8649,
+      "step": 1930
+    },
+    {
+      "epoch": 0.13038644250530643,
+      "grad_norm": 4.55290352417241,
+      "learning_rate": 9.971891822365048e-07,
+      "loss": 1.8144,
+      "step": 1935
+    },
+    {
+      "epoch": 0.13072335837741317,
+      "grad_norm": 4.808796442273753,
+      "learning_rate": 9.971265727340627e-07,
+      "loss": 1.8112,
+      "step": 1940
+    },
+    {
+      "epoch": 0.13106027424951988,
+      "grad_norm": 5.075751866145093,
+      "learning_rate": 9.970632756031322e-07,
+      "loss": 1.8588,
+      "step": 1945
+    },
+    {
+      "epoch": 0.13139719012162662,
+      "grad_norm": 5.368095057405194,
+      "learning_rate": 9.969992909312658e-07,
+      "loss": 1.8349,
+      "step": 1950
+    },
+    {
+      "epoch": 0.13173410599373336,
+      "grad_norm": 4.543717796981582,
+      "learning_rate": 9.969346188069684e-07,
+      "loss": 1.8516,
+      "step": 1955
+    },
+    {
+      "epoch": 0.1320710218658401,
+      "grad_norm": 4.810992798528759,
+      "learning_rate": 9.968692593196943e-07,
+      "loss": 1.7291,
+      "step": 1960
+    },
+    {
+      "epoch": 0.13240793773794685,
+      "grad_norm": 4.7057325097189935,
+      "learning_rate": 9.968032125598493e-07,
+      "loss": 1.9157,
+      "step": 1965
+    },
+    {
+      "epoch": 0.13274485361005356,
+      "grad_norm": 4.6279852022273635,
+      "learning_rate": 9.967364786187894e-07,
+      "loss": 1.804,
+      "step": 1970
+    },
+    {
+      "epoch": 0.1330817694821603,
+      "grad_norm": 4.797725869731827,
+      "learning_rate": 9.96669057588822e-07,
+      "loss": 1.827,
+      "step": 1975
+    },
+    {
+      "epoch": 0.13341868535426704,
+      "grad_norm": 4.63352318189097,
+      "learning_rate": 9.966009495632037e-07,
+      "loss": 1.7916,
+      "step": 1980
+    },
+    {
+      "epoch": 0.13375560122637378,
+      "grad_norm": 4.63007800091646,
+      "learning_rate": 9.965321546361421e-07,
+      "loss": 1.7955,
+      "step": 1985
+    },
+    {
+      "epoch": 0.1340925170984805,
+      "grad_norm": 4.592355530933701,
+      "learning_rate": 9.964626729027948e-07,
+      "loss": 1.8173,
+      "step": 1990
+    },
+    {
+      "epoch": 0.13442943297058724,
+      "grad_norm": 4.754656508753518,
+      "learning_rate": 9.963925044592695e-07,
+      "loss": 1.8508,
+      "step": 1995
+    },
+    {
+      "epoch": 0.13476634884269398,
+      "grad_norm": 5.036001769976464,
+      "learning_rate": 9.963216494026235e-07,
+      "loss": 1.7708,
+      "step": 2000
+    },
+    {
+      "epoch": 0.13510326471480072,
+      "grad_norm": 4.680634436660986,
+      "learning_rate": 9.962501078308636e-07,
+      "loss": 1.8545,
+      "step": 2005
+    },
+    {
+      "epoch": 0.13544018058690746,
+      "grad_norm": 5.019081719432215,
+      "learning_rate": 9.96177879842947e-07,
+      "loss": 1.8008,
+      "step": 2010
+    },
+    {
+      "epoch": 0.13577709645901417,
+      "grad_norm": 5.238373802263778,
+      "learning_rate": 9.961049655387799e-07,
+      "loss": 1.8496,
+      "step": 2015
+    },
+    {
+      "epoch": 0.13611401233112091,
+      "grad_norm": 4.796232137071534,
+      "learning_rate": 9.960313650192175e-07,
+      "loss": 1.8012,
+      "step": 2020
+    },
+    {
+      "epoch": 0.13645092820322766,
+      "grad_norm": 4.971202397273539,
+      "learning_rate": 9.959570783860647e-07,
+      "loss": 1.7943,
+      "step": 2025
+    },
+    {
+      "epoch": 0.1367878440753344,
+      "grad_norm": 4.74948378121568,
+      "learning_rate": 9.958821057420752e-07,
+      "loss": 1.84,
+      "step": 2030
+    },
+    {
+      "epoch": 0.1371247599474411,
+      "grad_norm": 4.845185928928487,
+      "learning_rate": 9.958064471909513e-07,
+      "loss": 1.7563,
+      "step": 2035
+    },
+    {
+      "epoch": 0.13746167581954785,
+      "grad_norm": 4.822222510818922,
+      "learning_rate": 9.95730102837345e-07,
+      "loss": 1.7998,
+      "step": 2040
+    },
+    {
+      "epoch": 0.1377985916916546,
+      "grad_norm": 4.219454587091505,
+      "learning_rate": 9.956530727868558e-07,
+      "loss": 1.9019,
+      "step": 2045
+    },
+    {
+      "epoch": 0.13813550756376133,
+      "grad_norm": 4.624461076983736,
+      "learning_rate": 9.955753571460322e-07,
+      "loss": 1.8316,
+      "step": 2050
+    },
+    {
+      "epoch": 0.13847242343586808,
+      "grad_norm": 4.895182906588043,
+      "learning_rate": 9.95496956022371e-07,
+      "loss": 1.822,
+      "step": 2055
+    },
+    {
+      "epoch": 0.1388093393079748,
+      "grad_norm": 5.1361871293028045,
+      "learning_rate": 9.95417869524317e-07,
+      "loss": 1.8116,
+      "step": 2060
+    },
+    {
+      "epoch": 0.13914625518008153,
+      "grad_norm": 4.598925274456651,
+      "learning_rate": 9.953380977612633e-07,
+      "loss": 1.8424,
+      "step": 2065
+    },
+    {
+      "epoch": 0.13948317105218827,
+      "grad_norm": 4.804272735140545,
+      "learning_rate": 9.952576408435505e-07,
+      "loss": 1.8828,
+      "step": 2070
+    },
+    {
+      "epoch": 0.139820086924295,
+      "grad_norm": 4.6975159711246075,
+      "learning_rate": 9.951764988824674e-07,
+      "loss": 1.8063,
+      "step": 2075
+    },
+    {
+      "epoch": 0.14015700279640173,
+      "grad_norm": 4.544302853074937,
+      "learning_rate": 9.950946719902498e-07,
+      "loss": 1.8352,
+      "step": 2080
+    },
+    {
+      "epoch": 0.14049391866850847,
+      "grad_norm": 4.3131351605485975,
+      "learning_rate": 9.950121602800813e-07,
+      "loss": 1.7853,
+      "step": 2085
+    },
+    {
+      "epoch": 0.1408308345406152,
+      "grad_norm": 4.95121376201706,
+      "learning_rate": 9.949289638660922e-07,
+      "loss": 1.7655,
+      "step": 2090
+    },
+    {
+      "epoch": 0.14116775041272195,
+      "grad_norm": 4.776858845008987,
+      "learning_rate": 9.948450828633608e-07,
+      "loss": 1.799,
+      "step": 2095
+    },
+    {
+      "epoch": 0.1415046662848287,
+      "grad_norm": 4.620614138437872,
+      "learning_rate": 9.947605173879115e-07,
+      "loss": 1.789,
+      "step": 2100
+    },
+    {
+      "epoch": 0.1418415821569354,
+      "grad_norm": 4.734440582941355,
+      "learning_rate": 9.94675267556716e-07,
+      "loss": 1.7768,
+      "step": 2105
+    },
+    {
+      "epoch": 0.14217849802904214,
+      "grad_norm": 5.165465783061854,
+      "learning_rate": 9.94589333487692e-07,
+      "loss": 1.79,
+      "step": 2110
+    },
+    {
+      "epoch": 0.14251541390114889,
+      "grad_norm": 4.910963698898701,
+      "learning_rate": 9.945027152997046e-07,
+      "loss": 1.8761,
+      "step": 2115
+    },
+    {
+      "epoch": 0.14285232977325563,
+      "grad_norm": 4.8125377943903445,
+      "learning_rate": 9.944154131125642e-07,
+      "loss": 1.8496,
+      "step": 2120
+    },
+    {
+      "epoch": 0.14318924564536234,
+      "grad_norm": 4.980214347335999,
+      "learning_rate": 9.94327427047028e-07,
+      "loss": 1.8448,
+      "step": 2125
+    },
+    {
+      "epoch": 0.14352616151746908,
+      "grad_norm": 5.2676984518019845,
+      "learning_rate": 9.942387572247983e-07,
+      "loss": 1.8144,
+      "step": 2130
+    },
+    {
+      "epoch": 0.14386307738957582,
+      "grad_norm": 4.578702371456273,
+      "learning_rate": 9.941494037685243e-07,
+      "loss": 1.8304,
+      "step": 2135
+    },
+    {
+      "epoch": 0.14419999326168256,
+      "grad_norm": 4.560416512561316,
+      "learning_rate": 9.940593668017998e-07,
+      "loss": 1.8057,
+      "step": 2140
+    },
+    {
+      "epoch": 0.1445369091337893,
+      "grad_norm": 5.543083573534103,
+      "learning_rate": 9.93968646449165e-07,
+      "loss": 1.8248,
+      "step": 2145
+    },
+    {
+      "epoch": 0.14487382500589602,
+      "grad_norm": 4.966033102521214,
+      "learning_rate": 9.938772428361045e-07,
+      "loss": 1.8475,
+      "step": 2150
+    },
+    {
+      "epoch": 0.14521074087800276,
+      "grad_norm": 5.1807178939148395,
+      "learning_rate": 9.937851560890484e-07,
+      "loss": 1.7746,
+      "step": 2155
+    },
+    {
+      "epoch": 0.1455476567501095,
+      "grad_norm": 4.75754428630411,
+      "learning_rate": 9.936923863353717e-07,
+      "loss": 1.8487,
+      "step": 2160
+    },
+    {
+      "epoch": 0.14588457262221624,
+      "grad_norm": 4.574863274091443,
+      "learning_rate": 9.935989337033939e-07,
+      "loss": 1.8253,
+      "step": 2165
+    },
+    {
+      "epoch": 0.14622148849432295,
+      "grad_norm": 4.544911337212059,
+      "learning_rate": 9.935047983223794e-07,
+      "loss": 1.7665,
+      "step": 2170
+    },
+    {
+      "epoch": 0.1465584043664297,
+      "grad_norm": 4.536591505508547,
+      "learning_rate": 9.934099803225367e-07,
+      "loss": 1.7767,
+      "step": 2175
+    },
+    {
+      "epoch": 0.14689532023853644,
+      "grad_norm": 4.588954312124332,
+      "learning_rate": 9.933144798350188e-07,
+      "loss": 1.8268,
+      "step": 2180
+    },
+    {
+      "epoch": 0.14723223611064318,
+      "grad_norm": 4.893042613752849,
+      "learning_rate": 9.932182969919228e-07,
+      "loss": 1.8961,
+      "step": 2185
+    },
+    {
+      "epoch": 0.14756915198274992,
+      "grad_norm": 4.131484985175047,
+      "learning_rate": 9.931214319262885e-07,
+      "loss": 1.7747,
+      "step": 2190
+    },
+    {
+      "epoch": 0.14790606785485663,
+      "grad_norm": 4.78203381923773,
+      "learning_rate": 9.930238847721013e-07,
+      "loss": 1.7877,
+      "step": 2195
+    },
+    {
+      "epoch": 0.14824298372696337,
+      "grad_norm": 9.076085543417564,
+      "learning_rate": 9.929256556642884e-07,
+      "loss": 1.7307,
+      "step": 2200
+    },
+    {
+      "epoch": 0.14857989959907011,
+      "grad_norm": 4.873351760425082,
+      "learning_rate": 9.92826744738721e-07,
+      "loss": 1.865,
+      "step": 2205
+    },
+    {
+      "epoch": 0.14891681547117686,
+      "grad_norm": 11.613145342066806,
+      "learning_rate": 9.927271521322134e-07,
+      "loss": 1.8322,
+      "step": 2210
+    },
+    {
+      "epoch": 0.14925373134328357,
+      "grad_norm": 4.743206734367761,
+      "learning_rate": 9.926268779825224e-07,
+      "loss": 1.8232,
+      "step": 2215
+    },
+    {
+      "epoch": 0.1495906472153903,
+      "grad_norm": 4.449486056308384,
+      "learning_rate": 9.925259224283484e-07,
+      "loss": 1.7962,
+      "step": 2220
+    },
+    {
+      "epoch": 0.14992756308749705,
+      "grad_norm": 4.712926106412736,
+      "learning_rate": 9.924242856093332e-07,
+      "loss": 1.7963,
+      "step": 2225
+    },
+    {
+      "epoch": 0.1502644789596038,
+      "grad_norm": 4.312335008664891,
+      "learning_rate": 9.923219676660614e-07,
+      "loss": 1.7713,
+      "step": 2230
+    },
+    {
+      "epoch": 0.15060139483171053,
+      "grad_norm": 4.519951384285313,
+      "learning_rate": 9.922189687400603e-07,
+      "loss": 1.7679,
+      "step": 2235
+    },
+    {
+      "epoch": 0.15093831070381725,
+      "grad_norm": 5.135990711467957,
+      "learning_rate": 9.921152889737984e-07,
+      "loss": 1.8021,
+      "step": 2240
+    },
+    {
+      "epoch": 0.151275226575924,
+      "grad_norm": 4.969576681103405,
+      "learning_rate": 9.92010928510686e-07,
+      "loss": 1.7999,
+      "step": 2245
+    },
+    {
+      "epoch": 0.15161214244803073,
+      "grad_norm": 5.093841787426275,
+      "learning_rate": 9.919058874950754e-07,
+      "loss": 1.8234,
+      "step": 2250
+    },
+    {
+      "epoch": 0.15194905832013747,
+      "grad_norm": 4.306254185807312,
+      "learning_rate": 9.9180016607226e-07,
+      "loss": 1.7889,
+      "step": 2255
+    },
+    {
+      "epoch": 0.15228597419224418,
+      "grad_norm": 5.155846839902419,
+      "learning_rate": 9.916937643884737e-07,
+      "loss": 1.8429,
+      "step": 2260
+    },
+    {
+      "epoch": 0.15262289006435092,
+      "grad_norm": 5.240572597887352,
+      "learning_rate": 9.915866825908927e-07,
+      "loss": 1.8422,
+      "step": 2265
+    },
+    {
+      "epoch": 0.15295980593645767,
+      "grad_norm": 4.27444987327846,
+      "learning_rate": 9.914789208276329e-07,
+      "loss": 1.8744,
+      "step": 2270
+    },
+    {
+      "epoch": 0.1532967218085644,
+      "grad_norm": 5.0907807471676465,
+      "learning_rate": 9.913704792477511e-07,
+      "loss": 1.8303,
+      "step": 2275
+    },
+    {
+      "epoch": 0.15363363768067115,
+      "grad_norm": 4.14235267626192,
+      "learning_rate": 9.91261358001244e-07,
+      "loss": 1.7255,
+      "step": 2280
+    },
+    {
+      "epoch": 0.15397055355277786,
+      "grad_norm": 4.74663474434792,
+      "learning_rate": 9.911515572390495e-07,
+      "loss": 1.809,
+      "step": 2285
+    },
+    {
+      "epoch": 0.1543074694248846,
+      "grad_norm": 4.984496588449255,
+      "learning_rate": 9.91041077113044e-07,
+      "loss": 1.8025,
+      "step": 2290
+    },
+    {
+      "epoch": 0.15464438529699134,
+      "grad_norm": 4.850549839234142,
+      "learning_rate": 9.909299177760445e-07,
+      "loss": 1.8208,
+      "step": 2295
+    },
+    {
+      "epoch": 0.15498130116909808,
+      "grad_norm": 4.816762254065827,
+      "learning_rate": 9.90818079381807e-07,
+      "loss": 1.8963,
+      "step": 2300
+    },
+    {
+      "epoch": 0.1553182170412048,
+      "grad_norm": 4.930040222816709,
+      "learning_rate": 9.907055620850277e-07,
+      "loss": 1.7785,
+      "step": 2305
+    },
+    {
+      "epoch": 0.15565513291331154,
+      "grad_norm": 6.8194169990152576,
+      "learning_rate": 9.905923660413409e-07,
+      "loss": 1.7386,
+      "step": 2310
+    },
+    {
+      "epoch": 0.15599204878541828,
+      "grad_norm": 5.119532349582884,
+      "learning_rate": 9.904784914073196e-07,
+      "loss": 1.8177,
+      "step": 2315
+    },
+    {
+      "epoch": 0.15632896465752502,
+      "grad_norm": 4.884587026760344,
+      "learning_rate": 9.903639383404765e-07,
+      "loss": 1.7836,
+      "step": 2320
+    },
+    {
+      "epoch": 0.15666588052963176,
+      "grad_norm": 4.790810189179543,
+      "learning_rate": 9.902487069992618e-07,
+      "loss": 1.7596,
+      "step": 2325
+    },
+    {
+      "epoch": 0.15700279640173848,
+      "grad_norm": 5.134335268782031,
+      "learning_rate": 9.901327975430645e-07,
+      "loss": 1.8548,
+      "step": 2330
+    },
+    {
+      "epoch": 0.15733971227384522,
+      "grad_norm": 5.08823307216309,
+      "learning_rate": 9.900162101322106e-07,
+      "loss": 1.7812,
+      "step": 2335
+    },
+    {
+      "epoch": 0.15767662814595196,
+      "grad_norm": 5.059545413463581,
+      "learning_rate": 9.898989449279653e-07,
+      "loss": 1.7619,
+      "step": 2340
+    },
+    {
+      "epoch": 0.1580135440180587,
+      "grad_norm": 4.680401030004896,
+      "learning_rate": 9.8978100209253e-07,
+      "loss": 1.8315,
+      "step": 2345
+    },
+    {
+      "epoch": 0.1583504598901654,
+      "grad_norm": 4.600218824615641,
+      "learning_rate": 9.89662381789044e-07,
+      "loss": 1.8484,
+      "step": 2350
+    },
+    {
+      "epoch": 0.15868737576227215,
+      "grad_norm": 4.937748124614686,
+      "learning_rate": 9.89543084181584e-07,
+      "loss": 1.7603,
+      "step": 2355
+    },
+    {
+      "epoch": 0.1590242916343789,
+      "grad_norm": 5.031865797237023,
+      "learning_rate": 9.894231094351628e-07,
+      "loss": 1.8257,
+      "step": 2360
+    },
+    {
+      "epoch": 0.15936120750648564,
+      "grad_norm": 4.489253013228351,
+      "learning_rate": 9.893024577157303e-07,
+      "loss": 1.8541,
+      "step": 2365
+    },
+    {
+      "epoch": 0.15969812337859238,
+      "grad_norm": 4.759951564369919,
+      "learning_rate": 9.891811291901727e-07,
+      "loss": 1.7888,
+      "step": 2370
+    },
+    {
+      "epoch": 0.1600350392506991,
+      "grad_norm": 4.803084737310507,
+      "learning_rate": 9.890591240263124e-07,
+      "loss": 1.7885,
+      "step": 2375
+    },
+    {
+      "epoch": 0.16037195512280583,
+      "grad_norm": 4.777856099384717,
+      "learning_rate": 9.889364423929075e-07,
+      "loss": 1.7622,
+      "step": 2380
+    },
+    {
+      "epoch": 0.16070887099491257,
+      "grad_norm": 4.658918440571349,
+      "learning_rate": 9.888130844596524e-07,
+      "loss": 1.7666,
+      "step": 2385
+    },
+    {
+      "epoch": 0.1610457868670193,
+      "grad_norm": 4.643205378034103,
+      "learning_rate": 9.88689050397176e-07,
+      "loss": 1.8317,
+      "step": 2390
+    },
+    {
+      "epoch": 0.16138270273912603,
+      "grad_norm": 4.758968390841792,
+      "learning_rate": 9.885643403770431e-07,
+      "loss": 1.7614,
+      "step": 2395
+    },
+    {
+      "epoch": 0.16171961861123277,
+      "grad_norm": 5.076582216217152,
+      "learning_rate": 9.884389545717538e-07,
+      "loss": 1.7926,
+      "step": 2400
+    },
+    {
+      "epoch": 0.1620565344833395,
+      "grad_norm": 5.28772251665507,
+      "learning_rate": 9.88312893154742e-07,
+      "loss": 1.7311,
+      "step": 2405
+    },
+    {
+      "epoch": 0.16239345035544625,
+      "grad_norm": 4.759748331327468,
+      "learning_rate": 9.881861563003766e-07,
+      "loss": 1.8694,
+      "step": 2410
+    },
+    {
+      "epoch": 0.162730366227553,
+      "grad_norm": 4.5348602623665535,
+      "learning_rate": 9.880587441839613e-07,
+      "loss": 1.783,
+      "step": 2415
+    },
+    {
+      "epoch": 0.1630672820996597,
+      "grad_norm": 4.92603836017639,
+      "learning_rate": 9.87930656981733e-07,
+      "loss": 1.8056,
+      "step": 2420
+    },
+    {
+      "epoch": 0.16340419797176645,
+      "grad_norm": 5.15426511329714,
+      "learning_rate": 9.878018948708625e-07,
+      "loss": 1.6898,
+      "step": 2425
+    },
+    {
+      "epoch": 0.1637411138438732,
+      "grad_norm": 4.970467692997654,
+      "learning_rate": 9.876724580294546e-07,
+      "loss": 1.8129,
+      "step": 2430
+    },
+    {
+      "epoch": 0.16407802971597993,
+      "grad_norm": 4.584293618459513,
+      "learning_rate": 9.875423466365471e-07,
+      "loss": 1.8001,
+      "step": 2435
+    },
+    {
+      "epoch": 0.16441494558808664,
+      "grad_norm": 4.617563282597616,
+      "learning_rate": 9.874115608721107e-07,
+      "loss": 1.7551,
+      "step": 2440
+    },
+    {
+      "epoch": 0.16475186146019338,
+      "grad_norm": 4.756934248759054,
+      "learning_rate": 9.872801009170492e-07,
+      "loss": 1.8694,
+      "step": 2445
+    },
+    {
+      "epoch": 0.16508877733230012,
+      "grad_norm": 4.474706782061384,
+      "learning_rate": 9.871479669531988e-07,
+      "loss": 1.7236,
+      "step": 2450
+    },
+    {
+      "epoch": 0.16542569320440687,
+      "grad_norm": 5.216098770623675,
+      "learning_rate": 9.87015159163328e-07,
+      "loss": 1.7731,
+      "step": 2455
+    },
+    {
+      "epoch": 0.1657626090765136,
+      "grad_norm": 5.3566900141122105,
+      "learning_rate": 9.868816777311372e-07,
+      "loss": 1.8435,
+      "step": 2460
+    },
+    {
+      "epoch": 0.16609952494862032,
+      "grad_norm": 4.724843591610029,
+      "learning_rate": 9.867475228412592e-07,
+      "loss": 1.7726,
+      "step": 2465
+    },
+    {
+      "epoch": 0.16643644082072706,
+      "grad_norm": 5.500471088365371,
+      "learning_rate": 9.866126946792572e-07,
+      "loss": 1.7542,
+      "step": 2470
+    },
+    {
+      "epoch": 0.1667733566928338,
+      "grad_norm": 4.7605627099736125,
+      "learning_rate": 9.864771934316268e-07,
+      "loss": 1.7708,
+      "step": 2475
+    },
+    {
+      "epoch": 0.16711027256494054,
+      "grad_norm": 4.725097378571309,
+      "learning_rate": 9.863410192857938e-07,
+      "loss": 1.8176,
+      "step": 2480
+    },
+    {
+      "epoch": 0.16744718843704726,
+      "grad_norm": 4.5019201902072234,
+      "learning_rate": 9.862041724301154e-07,
+      "loss": 1.8011,
+      "step": 2485
+    },
+    {
+      "epoch": 0.167784104309154,
+      "grad_norm": 4.563439436344494,
+      "learning_rate": 9.860666530538787e-07,
+      "loss": 1.8648,
+      "step": 2490
+    },
+    {
+      "epoch": 0.16812102018126074,
+      "grad_norm": 4.245541125514505,
+      "learning_rate": 9.859284613473017e-07,
+      "loss": 1.7836,
+      "step": 2495
+    },
+    {
+      "epoch": 0.16845793605336748,
+      "grad_norm": 4.289654542861537,
+      "learning_rate": 9.857895975015318e-07,
+      "loss": 1.8215,
+      "step": 2500
+    },
+    {
+      "epoch": 0.16879485192547422,
+      "grad_norm": 4.680020276493947,
+      "learning_rate": 9.856500617086463e-07,
+      "loss": 1.7658,
+      "step": 2505
+    },
+    {
+      "epoch": 0.16913176779758093,
+      "grad_norm": 4.79697808623628,
+      "learning_rate": 9.85509854161652e-07,
+      "loss": 1.7638,
+      "step": 2510
+    },
+    {
+      "epoch": 0.16946868366968768,
+      "grad_norm": 4.688992073069253,
+      "learning_rate": 9.853689750544849e-07,
+      "loss": 1.8517,
+      "step": 2515
+    },
+    {
+      "epoch": 0.16980559954179442,
+      "grad_norm": 9.766438264833477,
+      "learning_rate": 9.852274245820095e-07,
+      "loss": 1.7899,
+      "step": 2520
+    },
+    {
+      "epoch": 0.17014251541390116,
+      "grad_norm": 5.374427197648961,
+      "learning_rate": 9.850852029400198e-07,
+      "loss": 1.8161,
+      "step": 2525
+    },
+    {
+      "epoch": 0.17047943128600787,
+      "grad_norm": 4.740432344703743,
+      "learning_rate": 9.849423103252374e-07,
+      "loss": 1.767,
+      "step": 2530
+    },
+    {
+      "epoch": 0.1708163471581146,
+      "grad_norm": 4.848915892043279,
+      "learning_rate": 9.84798746935312e-07,
+      "loss": 1.7695,
+      "step": 2535
+    },
+    {
+      "epoch": 0.17115326303022135,
+      "grad_norm": 4.546388184572521,
+      "learning_rate": 9.846545129688217e-07,
+      "loss": 1.731,
+      "step": 2540
+    },
+    {
+      "epoch": 0.1714901789023281,
+      "grad_norm": 4.920312702284114,
+      "learning_rate": 9.845096086252716e-07,
+      "loss": 1.7672,
+      "step": 2545
+    },
+    {
+      "epoch": 0.17182709477443484,
+      "grad_norm": 4.6089934584516135,
+      "learning_rate": 9.843640341050944e-07,
+      "loss": 1.846,
+      "step": 2550
+    },
+    {
+      "epoch": 0.17216401064654155,
+      "grad_norm": 4.56883920848349,
+      "learning_rate": 9.842177896096493e-07,
+      "loss": 1.8548,
+      "step": 2555
+    },
+    {
+      "epoch": 0.1725009265186483,
+      "grad_norm": 4.704343360713175,
+      "learning_rate": 9.84070875341223e-07,
+      "loss": 1.7832,
+      "step": 2560
+    },
+    {
+      "epoch": 0.17283784239075503,
+      "grad_norm": 4.560170584618913,
+      "learning_rate": 9.83923291503028e-07,
+      "loss": 1.8017,
+      "step": 2565
+    },
+    {
+      "epoch": 0.17317475826286177,
+      "grad_norm": 4.648787556226848,
+      "learning_rate": 9.837750382992033e-07,
+      "loss": 1.775,
+      "step": 2570
+    },
+    {
+      "epoch": 0.17351167413496849,
+      "grad_norm": 4.863410571412938,
+      "learning_rate": 9.836261159348135e-07,
+      "loss": 1.7932,
+      "step": 2575
+    },
+    {
+      "epoch": 0.17384859000707523,
+      "grad_norm": 4.952071220515145,
+      "learning_rate": 9.834765246158488e-07,
+      "loss": 1.8207,
+      "step": 2580
+    },
+    {
+      "epoch": 0.17418550587918197,
+      "grad_norm": 4.8410219655558615,
+      "learning_rate": 9.83326264549225e-07,
+      "loss": 1.818,
+      "step": 2585
+    },
+    {
+      "epoch": 0.1745224217512887,
+      "grad_norm": 4.5337820540456795,
+      "learning_rate": 9.83175335942783e-07,
+      "loss": 1.7326,
+      "step": 2590
+    },
+    {
+      "epoch": 0.17485933762339545,
+      "grad_norm": 4.7911244830213215,
+      "learning_rate": 9.830237390052876e-07,
+      "loss": 1.7431,
+      "step": 2595
+    },
+    {
+      "epoch": 0.17519625349550216,
+      "grad_norm": 5.11291379581688,
+      "learning_rate": 9.82871473946429e-07,
+      "loss": 1.8238,
+      "step": 2600
+    },
+    {
+      "epoch": 0.1755331693676089,
+      "grad_norm": 4.8347628505716145,
+      "learning_rate": 9.82718540976821e-07,
+      "loss": 1.781,
+      "step": 2605
+    },
+    {
+      "epoch": 0.17587008523971565,
+      "grad_norm": 5.1804157814006455,
+      "learning_rate": 9.825649403080015e-07,
+      "loss": 1.781,
+      "step": 2610
+    },
+    {
+      "epoch": 0.1762070011118224,
+      "grad_norm": 4.94004659492207,
+      "learning_rate": 9.824106721524317e-07,
+      "loss": 1.737,
+      "step": 2615
+    },
+    {
+      "epoch": 0.1765439169839291,
+      "grad_norm": 4.583871412666966,
+      "learning_rate": 9.822557367234962e-07,
+      "loss": 1.7941,
+      "step": 2620
+    },
+    {
+      "epoch": 0.17688083285603584,
+      "grad_norm": 4.996180207831546,
+      "learning_rate": 9.82100134235503e-07,
+      "loss": 1.8262,
+      "step": 2625
+    },
+    {
+      "epoch": 0.17721774872814258,
+      "grad_norm": 4.799695846367883,
+      "learning_rate": 9.819438649036823e-07,
+      "loss": 1.8264,
+      "step": 2630
+    },
+    {
+      "epoch": 0.17755466460024932,
+      "grad_norm": 4.91429312170365,
+      "learning_rate": 9.817869289441864e-07,
+      "loss": 1.7259,
+      "step": 2635
+    },
+    {
+      "epoch": 0.17789158047235606,
+      "grad_norm": 4.8369882287978365,
+      "learning_rate": 9.816293265740907e-07,
+      "loss": 1.8009,
+      "step": 2640
+    },
+    {
+      "epoch": 0.17822849634446278,
+      "grad_norm": 7.820380148630697,
+      "learning_rate": 9.81471058011391e-07,
+      "loss": 1.812,
+      "step": 2645
+    },
+    {
+      "epoch": 0.17856541221656952,
+      "grad_norm": 4.591956442861727,
+      "learning_rate": 9.81312123475006e-07,
+      "loss": 1.7916,
+      "step": 2650
+    },
+    {
+      "epoch": 0.17890232808867626,
+      "grad_norm": 5.029746480437414,
+      "learning_rate": 9.811525231847746e-07,
+      "loss": 1.8452,
+      "step": 2655
+    },
+    {
+      "epoch": 0.179239243960783,
+      "grad_norm": 4.811826857148872,
+      "learning_rate": 9.809922573614569e-07,
+      "loss": 1.7681,
+      "step": 2660
+    },
+    {
+      "epoch": 0.17957615983288971,
+      "grad_norm": 4.900869827689216,
+      "learning_rate": 9.808313262267337e-07,
+      "loss": 1.7657,
+      "step": 2665
+    },
+    {
+      "epoch": 0.17991307570499646,
+      "grad_norm": 4.717354952119474,
+      "learning_rate": 9.806697300032057e-07,
+      "loss": 1.7752,
+      "step": 2670
+    },
+    {
+      "epoch": 0.1802499915771032,
+      "grad_norm": 5.128437055986362,
+      "learning_rate": 9.805074689143938e-07,
+      "loss": 1.7887,
+      "step": 2675
+    },
+    {
+      "epoch": 0.18058690744920994,
+      "grad_norm": 5.150193518417055,
+      "learning_rate": 9.803445431847388e-07,
+      "loss": 1.7366,
+      "step": 2680
+    },
+    {
+      "epoch": 0.18092382332131668,
+      "grad_norm": 4.763856294704906,
+      "learning_rate": 9.801809530396003e-07,
+      "loss": 1.8641,
+      "step": 2685
+    },
+    {
+      "epoch": 0.1812607391934234,
+      "grad_norm": 4.425971270262882,
+      "learning_rate": 9.800166987052572e-07,
+      "loss": 1.7468,
+      "step": 2690
+    },
+    {
+      "epoch": 0.18159765506553013,
+      "grad_norm": 4.638375651468754,
+      "learning_rate": 9.798517804089072e-07,
+      "loss": 1.9051,
+      "step": 2695
+    },
+    {
+      "epoch": 0.18193457093763687,
+      "grad_norm": 4.2910226154650335,
+      "learning_rate": 9.796861983786661e-07,
+      "loss": 1.729,
+      "step": 2700
+    },
+    {
+      "epoch": 0.18227148680974362,
+      "grad_norm": 4.529098176187787,
+      "learning_rate": 9.795199528435682e-07,
+      "loss": 1.7729,
+      "step": 2705
+    },
+    {
+      "epoch": 0.18260840268185033,
+      "grad_norm": 4.832329865493561,
+      "learning_rate": 9.793530440335654e-07,
+      "loss": 1.7572,
+      "step": 2710
+    },
+    {
+      "epoch": 0.18294531855395707,
+      "grad_norm": 4.810104590212989,
+      "learning_rate": 9.791854721795264e-07,
+      "loss": 1.7605,
+      "step": 2715
+    },
+    {
+      "epoch": 0.1832822344260638,
+      "grad_norm": 4.820817246781472,
+      "learning_rate": 9.790172375132385e-07,
+      "loss": 1.7599,
+      "step": 2720
+    },
+    {
+      "epoch": 0.18361915029817055,
+      "grad_norm": 4.605366486448338,
+      "learning_rate": 9.788483402674041e-07,
+      "loss": 1.8545,
+      "step": 2725
+    },
+    {
+      "epoch": 0.1839560661702773,
+      "grad_norm": 4.748751803024309,
+      "learning_rate": 9.786787806756434e-07,
+      "loss": 1.7956,
+      "step": 2730
+    },
+    {
+      "epoch": 0.184292982042384,
+      "grad_norm": 5.117426579850719,
+      "learning_rate": 9.78508558972492e-07,
+      "loss": 1.8082,
+      "step": 2735
+    },
+    {
+      "epoch": 0.18462989791449075,
+      "grad_norm": 4.569683293912544,
+      "learning_rate": 9.783376753934015e-07,
+      "loss": 1.8217,
+      "step": 2740
+    },
+    {
+      "epoch": 0.1849668137865975,
+      "grad_norm": 4.834824323691271,
+      "learning_rate": 9.781661301747393e-07,
+      "loss": 1.7359,
+      "step": 2745
+    },
+    {
+      "epoch": 0.18530372965870423,
+      "grad_norm": 4.728916185918625,
+      "learning_rate": 9.779939235537879e-07,
+      "loss": 1.8533,
+      "step": 2750
+    },
+    {
+      "epoch": 0.18564064553081094,
+      "grad_norm": 4.883042265853466,
+      "learning_rate": 9.778210557687443e-07,
+      "loss": 1.8023,
+      "step": 2755
+    },
+    {
+      "epoch": 0.18597756140291768,
+      "grad_norm": 4.652974648533559,
+      "learning_rate": 9.776475270587205e-07,
+      "loss": 1.7946,
+      "step": 2760
+    },
+    {
+      "epoch": 0.18631447727502443,
+      "grad_norm": 4.739243517296054,
+      "learning_rate": 9.774733376637421e-07,
+      "loss": 1.7767,
+      "step": 2765
+    },
+    {
+      "epoch": 0.18665139314713117,
+      "grad_norm": 5.11851445798632,
+      "learning_rate": 9.772984878247493e-07,
+      "loss": 1.7976,
+      "step": 2770
+    },
+    {
+      "epoch": 0.1869883090192379,
+      "grad_norm": 5.074930861110047,
+      "learning_rate": 9.771229777835952e-07,
+      "loss": 1.7576,
+      "step": 2775
+    },
+    {
+      "epoch": 0.18732522489134462,
+      "grad_norm": 4.520407367212375,
+      "learning_rate": 9.769468077830466e-07,
+      "loss": 1.7678,
+      "step": 2780
+    },
+    {
+      "epoch": 0.18766214076345136,
+      "grad_norm": 4.81402917555932,
+      "learning_rate": 9.767699780667827e-07,
+      "loss": 1.8164,
+      "step": 2785
+    },
+    {
+      "epoch": 0.1879990566355581,
+      "grad_norm": 4.964921504879331,
+      "learning_rate": 9.765924888793955e-07,
+      "loss": 1.7382,
+      "step": 2790
+    },
+    {
+      "epoch": 0.18833597250766484,
+      "grad_norm": 4.347692369389646,
+      "learning_rate": 9.76414340466389e-07,
+      "loss": 1.8065,
+      "step": 2795
+    },
+    {
+      "epoch": 0.18867288837977156,
+      "grad_norm": 5.298426455276457,
+      "learning_rate": 9.762355330741794e-07,
+      "loss": 1.8219,
+      "step": 2800
+    },
+    {
+      "epoch": 0.1890098042518783,
+      "grad_norm": 4.5164283817941975,
+      "learning_rate": 9.760560669500941e-07,
+      "loss": 1.7944,
+      "step": 2805
+    },
+    {
+      "epoch": 0.18934672012398504,
+      "grad_norm": 4.715610201482097,
+      "learning_rate": 9.758759423423716e-07,
+      "loss": 1.7729,
+      "step": 2810
+    },
+    {
+      "epoch": 0.18968363599609178,
+      "grad_norm": 4.716260025180655,
+      "learning_rate": 9.756951595001617e-07,
+      "loss": 1.7253,
+      "step": 2815
+    },
+    {
+      "epoch": 0.19002055186819852,
+      "grad_norm": 5.3440331043854075,
+      "learning_rate": 9.755137186735238e-07,
+      "loss": 1.7707,
+      "step": 2820
+    },
+    {
+      "epoch": 0.19035746774030524,
+      "grad_norm": 5.02041013091292,
+      "learning_rate": 9.753316201134282e-07,
+      "loss": 1.7937,
+      "step": 2825
+    },
+    {
+      "epoch": 0.19069438361241198,
+      "grad_norm": 4.555395708579455,
+      "learning_rate": 9.75148864071755e-07,
+      "loss": 1.7525,
+      "step": 2830
+    },
+    {
+      "epoch": 0.19103129948451872,
+      "grad_norm": 5.174506656565936,
+      "learning_rate": 9.74965450801293e-07,
+      "loss": 1.7621,
+      "step": 2835
+    },
+    {
+      "epoch": 0.19136821535662546,
+      "grad_norm": 4.931304293894075,
+      "learning_rate": 9.747813805557408e-07,
+      "loss": 1.7584,
+      "step": 2840
+    },
+    {
+      "epoch": 0.19170513122873217,
+      "grad_norm": 4.7073358538658185,
+      "learning_rate": 9.745966535897054e-07,
+      "loss": 1.7606,
+      "step": 2845
+    },
+    {
+      "epoch": 0.19204204710083891,
+      "grad_norm": 4.582053551568521,
+      "learning_rate": 9.744112701587024e-07,
+      "loss": 1.7712,
+      "step": 2850
+    },
+    {
+      "epoch": 0.19237896297294566,
+      "grad_norm": 4.767250149002942,
+      "learning_rate": 9.742252305191551e-07,
+      "loss": 1.7563,
+      "step": 2855
+    },
+    {
+      "epoch": 0.1927158788450524,
+      "grad_norm": 4.709027679259578,
+      "learning_rate": 9.740385349283946e-07,
+      "loss": 1.793,
+      "step": 2860
+    },
+    {
+      "epoch": 0.19305279471715914,
+      "grad_norm": 4.956158981373378,
+      "learning_rate": 9.738511836446596e-07,
+      "loss": 1.8203,
+      "step": 2865
+    },
+    {
+      "epoch": 0.19338971058926585,
+      "grad_norm": 4.474867003420559,
+      "learning_rate": 9.736631769270957e-07,
+      "loss": 1.7836,
+      "step": 2870
+    },
+    {
+      "epoch": 0.1937266264613726,
+      "grad_norm": 4.843134161039688,
+      "learning_rate": 9.734745150357544e-07,
+      "loss": 1.7853,
+      "step": 2875
+    },
+    {
+      "epoch": 0.19406354233347933,
+      "grad_norm": 4.655571844872947,
+      "learning_rate": 9.732851982315944e-07,
+      "loss": 1.8458,
+      "step": 2880
+    },
+    {
+      "epoch": 0.19440045820558607,
+      "grad_norm": 4.856240480203855,
+      "learning_rate": 9.730952267764796e-07,
+      "loss": 1.761,
+      "step": 2885
+    },
+    {
+      "epoch": 0.1947373740776928,
+      "grad_norm": 4.826308380318007,
+      "learning_rate": 9.729046009331798e-07,
+      "loss": 1.7442,
+      "step": 2890
+    },
+    {
+      "epoch": 0.19507428994979953,
+      "grad_norm": 4.632076537884298,
+      "learning_rate": 9.727133209653696e-07,
+      "loss": 1.7747,
+      "step": 2895
+    },
+    {
+      "epoch": 0.19541120582190627,
+      "grad_norm": 4.493737956408813,
+      "learning_rate": 9.72521387137629e-07,
+      "loss": 1.8317,
+      "step": 2900
+    },
+    {
+      "epoch": 0.195748121694013,
+      "grad_norm": 4.802050440898538,
+      "learning_rate": 9.723287997154419e-07,
+      "loss": 1.8332,
+      "step": 2905
+    },
+    {
+      "epoch": 0.19608503756611975,
+      "grad_norm": 4.480297291602395,
+      "learning_rate": 9.72135558965196e-07,
+      "loss": 1.7894,
+      "step": 2910
+    },
+    {
+      "epoch": 0.19642195343822647,
+      "grad_norm": 4.563719923514326,
+      "learning_rate": 9.719416651541837e-07,
+      "loss": 1.7317,
+      "step": 2915
+    },
+    {
+      "epoch": 0.1967588693103332,
+      "grad_norm": 4.315766072130242,
+      "learning_rate": 9.717471185505996e-07,
+      "loss": 1.7465,
+      "step": 2920
+    },
+    {
+      "epoch": 0.19709578518243995,
+      "grad_norm": 4.769139822472816,
+      "learning_rate": 9.715519194235422e-07,
+      "loss": 1.7184,
+      "step": 2925
+    },
+    {
+      "epoch": 0.1974327010545467,
+      "grad_norm": 4.814639381264535,
+      "learning_rate": 9.713560680430117e-07,
+      "loss": 1.801,
+      "step": 2930
+    },
+    {
+      "epoch": 0.1977696169266534,
+      "grad_norm": 4.894385656327445,
+      "learning_rate": 9.71159564679911e-07,
+      "loss": 1.7801,
+      "step": 2935
+    },
+    {
+      "epoch": 0.19810653279876014,
+      "grad_norm": 4.96835121805525,
+      "learning_rate": 9.709624096060449e-07,
+      "loss": 1.7787,
+      "step": 2940
+    },
+    {
+      "epoch": 0.19844344867086688,
+      "grad_norm": 4.395255870875639,
+      "learning_rate": 9.707646030941192e-07,
+      "loss": 1.7774,
+      "step": 2945
+    },
+    {
+      "epoch": 0.19878036454297363,
+      "grad_norm": 4.596421154913451,
+      "learning_rate": 9.705661454177416e-07,
+      "loss": 1.7741,
+      "step": 2950
+    },
+    {
+      "epoch": 0.19911728041508037,
+      "grad_norm": 4.904942759236581,
+      "learning_rate": 9.703670368514192e-07,
+      "loss": 1.7855,
+      "step": 2955
+    },
+    {
+      "epoch": 0.19945419628718708,
+      "grad_norm": 4.6289143433439035,
+      "learning_rate": 9.701672776705609e-07,
+      "loss": 1.7948,
+      "step": 2960
+    },
+    {
+      "epoch": 0.19979111215929382,
+      "grad_norm": 4.703689873640086,
+      "learning_rate": 9.699668681514746e-07,
+      "loss": 1.8122,
+      "step": 2965
+    },
+    {
+      "epoch": 0.20012802803140056,
+      "grad_norm": 4.322341146884426,
+      "learning_rate": 9.697658085713676e-07,
+      "loss": 1.8207,
+      "step": 2970
+    },
+    {
+      "epoch": 0.2004649439035073,
+      "grad_norm": 5.107208490116794,
+      "learning_rate": 9.695640992083471e-07,
+      "loss": 1.732,
+      "step": 2975
+    },
+    {
+      "epoch": 0.20080185977561402,
+      "grad_norm": 4.618863119770153,
+      "learning_rate": 9.693617403414188e-07,
+      "loss": 1.7793,
+      "step": 2980
+    },
+    {
+      "epoch": 0.20113877564772076,
+      "grad_norm": 5.063764438355028,
+      "learning_rate": 9.691587322504865e-07,
+      "loss": 1.7599,
+      "step": 2985
+    },
+    {
+      "epoch": 0.2014756915198275,
+      "grad_norm": 4.919214411010044,
+      "learning_rate": 9.68955075216352e-07,
+      "loss": 1.844,
+      "step": 2990
+    },
+    {
+      "epoch": 0.20181260739193424,
+      "grad_norm": 7.731079171027298,
+      "learning_rate": 9.687507695207154e-07,
+      "loss": 1.7809,
+      "step": 2995
+    },
+    {
+      "epoch": 0.20214952326404098,
+      "grad_norm": 4.88502046387205,
+      "learning_rate": 9.685458154461731e-07,
+      "loss": 1.7439,
+      "step": 3000
+    },
+    {
+      "epoch": 0.2024864391361477,
+      "grad_norm": 4.325672374975772,
+      "learning_rate": 9.683402132762193e-07,
+      "loss": 1.7976,
+      "step": 3005
+    },
+    {
+      "epoch": 0.20282335500825444,
+      "grad_norm": 4.696787912223931,
+      "learning_rate": 9.68133963295244e-07,
+      "loss": 1.8553,
+      "step": 3010
+    },
+    {
+      "epoch": 0.20316027088036118,
+      "grad_norm": 4.807613388751499,
+      "learning_rate": 9.679270657885334e-07,
+      "loss": 1.6998,
+      "step": 3015
+    },
+    {
+      "epoch": 0.20349718675246792,
+      "grad_norm": 4.563573042960349,
+      "learning_rate": 9.677195210422693e-07,
+      "loss": 1.8247,
+      "step": 3020
+    },
+    {
+      "epoch": 0.20383410262457463,
+      "grad_norm": 4.6931253642053115,
+      "learning_rate": 9.675113293435288e-07,
+      "loss": 1.7735,
+      "step": 3025
+    },
+    {
+      "epoch": 0.20417101849668137,
+      "grad_norm": 4.478719320401698,
+      "learning_rate": 9.673024909802841e-07,
+      "loss": 1.745,
+      "step": 3030
+    },
+    {
+      "epoch": 0.2045079343687881,
+      "grad_norm": 4.447484104577244,
+      "learning_rate": 9.670930062414017e-07,
+      "loss": 1.7063,
+      "step": 3035
+    },
+    {
+      "epoch": 0.20484485024089485,
+      "grad_norm": 4.8927883667051955,
+      "learning_rate": 9.66882875416642e-07,
+      "loss": 1.8208,
+      "step": 3040
+    },
+    {
+      "epoch": 0.2051817661130016,
+      "grad_norm": 4.682209855532116,
+      "learning_rate": 9.666720987966595e-07,
+      "loss": 1.7681,
+      "step": 3045
+    },
+    {
+      "epoch": 0.2055186819851083,
+      "grad_norm": 4.916006159522313,
+      "learning_rate": 9.664606766730012e-07,
+      "loss": 1.7754,
+      "step": 3050
+    },
+    {
+      "epoch": 0.20585559785721505,
+      "grad_norm": 4.684215933574209,
+      "learning_rate": 9.662486093381082e-07,
+      "loss": 1.7737,
+      "step": 3055
+    },
+    {
+      "epoch": 0.2061925137293218,
+      "grad_norm": 4.813741584691905,
+      "learning_rate": 9.660358970853126e-07,
+      "loss": 1.7012,
+      "step": 3060
+    },
+    {
+      "epoch": 0.20652942960142853,
+      "grad_norm": 4.382040182047615,
+      "learning_rate": 9.658225402088395e-07,
+      "loss": 1.7321,
+      "step": 3065
+    },
+    {
+      "epoch": 0.20686634547353525,
+      "grad_norm": 4.458843911617041,
+      "learning_rate": 9.656085390038058e-07,
+      "loss": 1.7993,
+      "step": 3070
+    },
+    {
+      "epoch": 0.207203261345642,
+      "grad_norm": 4.55896931783871,
+      "learning_rate": 9.653938937662187e-07,
+      "loss": 1.7798,
+      "step": 3075
+    },
+    {
+      "epoch": 0.20754017721774873,
+      "grad_norm": 4.292995630768965,
+      "learning_rate": 9.651786047929772e-07,
+      "loss": 1.768,
+      "step": 3080
+    },
+    {
+      "epoch": 0.20787709308985547,
+      "grad_norm": 4.741061046681705,
+      "learning_rate": 9.649626723818702e-07,
+      "loss": 1.7983,
+      "step": 3085
+    },
+    {
+      "epoch": 0.2082140089619622,
+      "grad_norm": 4.914730551973815,
+      "learning_rate": 9.647460968315767e-07,
+      "loss": 1.8451,
+      "step": 3090
+    },
+    {
+      "epoch": 0.20855092483406892,
+      "grad_norm": 4.748415205041859,
+      "learning_rate": 9.645288784416652e-07,
+      "loss": 1.7624,
+      "step": 3095
+    },
+    {
+      "epoch": 0.20888784070617566,
+      "grad_norm": 4.615780661397535,
+      "learning_rate": 9.643110175125935e-07,
+      "loss": 1.8054,
+      "step": 3100
+    },
+    {
+      "epoch": 0.2092247565782824,
+      "grad_norm": 4.533202597147934,
+      "learning_rate": 9.640925143457084e-07,
+      "loss": 1.7423,
+      "step": 3105
+    },
+    {
+      "epoch": 0.20956167245038915,
+      "grad_norm": 5.095719194842791,
+      "learning_rate": 9.638733692432448e-07,
+      "loss": 1.7796,
+      "step": 3110
+    },
+    {
+      "epoch": 0.20989858832249586,
+      "grad_norm": 4.785132671207265,
+      "learning_rate": 9.636535825083252e-07,
+      "loss": 1.8073,
+      "step": 3115
+    },
+    {
+      "epoch": 0.2102355041946026,
+      "grad_norm": 4.462607949552112,
+      "learning_rate": 9.634331544449601e-07,
+      "loss": 1.8014,
+      "step": 3120
+    },
+    {
+      "epoch": 0.21057242006670934,
+      "grad_norm": 4.62396497723698,
+      "learning_rate": 9.632120853580472e-07,
+      "loss": 1.8046,
+      "step": 3125
+    },
+    {
+      "epoch": 0.21090933593881608,
+      "grad_norm": 4.440303646153727,
+      "learning_rate": 9.6299037555337e-07,
+      "loss": 1.73,
+      "step": 3130
+    },
+    {
+      "epoch": 0.21124625181092282,
+      "grad_norm": 4.684820392427264,
+      "learning_rate": 9.627680253375997e-07,
+      "loss": 1.7287,
+      "step": 3135
+    },
+    {
+      "epoch": 0.21158316768302954,
+      "grad_norm": 4.699281477495129,
+      "learning_rate": 9.625450350182918e-07,
+      "loss": 1.7721,
+      "step": 3140
+    },
+    {
+      "epoch": 0.21192008355513628,
+      "grad_norm": 4.404129361919743,
+      "learning_rate": 9.62321404903888e-07,
+      "loss": 1.7183,
+      "step": 3145
+    },
+    {
+      "epoch": 0.21225699942724302,
+      "grad_norm": 4.46828072383975,
+      "learning_rate": 9.620971353037148e-07,
+      "loss": 1.7568,
+      "step": 3150
+    },
+    {
+      "epoch": 0.21259391529934976,
+      "grad_norm": 5.0456383160573015,
+      "learning_rate": 9.618722265279835e-07,
+      "loss": 1.7834,
+      "step": 3155
+    },
+    {
+      "epoch": 0.21293083117145647,
+      "grad_norm": 5.447392712626318,
+      "learning_rate": 9.61646678887789e-07,
+      "loss": 1.7902,
+      "step": 3160
+    },
+    {
+      "epoch": 0.21326774704356322,
+      "grad_norm": 5.16918774828708,
+      "learning_rate": 9.614204926951102e-07,
+      "loss": 1.8112,
+      "step": 3165
+    },
+    {
+      "epoch": 0.21360466291566996,
+      "grad_norm": 4.4757864244714005,
+      "learning_rate": 9.611936682628095e-07,
+      "loss": 1.7717,
+      "step": 3170
+    },
+    {
+      "epoch": 0.2139415787877767,
+      "grad_norm": 4.598126050187275,
+      "learning_rate": 9.609662059046315e-07,
+      "loss": 1.7461,
+      "step": 3175
+    },
+    {
+      "epoch": 0.21427849465988344,
+      "grad_norm": 4.604780314975367,
+      "learning_rate": 9.607381059352038e-07,
+      "loss": 1.7711,
+      "step": 3180
+    },
+    {
+      "epoch": 0.21461541053199015,
+      "grad_norm": 4.64864748054051,
+      "learning_rate": 9.605093686700353e-07,
+      "loss": 1.8126,
+      "step": 3185
+    },
+    {
+      "epoch": 0.2149523264040969,
+      "grad_norm": 4.781745831526942,
+      "learning_rate": 9.602799944255172e-07,
+      "loss": 1.7298,
+      "step": 3190
+    },
+    {
+      "epoch": 0.21528924227620364,
+      "grad_norm": 4.743722422243644,
+      "learning_rate": 9.60049983518921e-07,
+      "loss": 1.8324,
+      "step": 3195
+    },
+    {
+      "epoch": 0.21562615814831038,
+      "grad_norm": 4.440157266990944,
+      "learning_rate": 9.598193362683995e-07,
+      "loss": 1.7269,
+      "step": 3200
+    },
+    {
+      "epoch": 0.2159630740204171,
+      "grad_norm": 5.245985893785636,
+      "learning_rate": 9.59588052992985e-07,
+      "loss": 1.7371,
+      "step": 3205
+    },
+    {
+      "epoch": 0.21629998989252383,
+      "grad_norm": 4.593649997820267,
+      "learning_rate": 9.5935613401259e-07,
+      "loss": 1.8086,
+      "step": 3210
+    },
+    {
+      "epoch": 0.21663690576463057,
+      "grad_norm": 4.490308962417356,
+      "learning_rate": 9.591235796480064e-07,
+      "loss": 1.7968,
+      "step": 3215
+    },
+    {
+      "epoch": 0.2169738216367373,
+      "grad_norm": 4.803010238652498,
+      "learning_rate": 9.588903902209048e-07,
+      "loss": 1.7773,
+      "step": 3220
+    },
+    {
+      "epoch": 0.21731073750884405,
+      "grad_norm": 4.646133140777154,
+      "learning_rate": 9.586565660538343e-07,
+      "loss": 1.7521,
+      "step": 3225
+    },
+    {
+      "epoch": 0.21764765338095077,
+      "grad_norm": 4.473035021116014,
+      "learning_rate": 9.584221074702217e-07,
+      "loss": 1.856,
+      "step": 3230
+    },
+    {
+      "epoch": 0.2179845692530575,
+      "grad_norm": 5.055380502337204,
+      "learning_rate": 9.581870147943715e-07,
+      "loss": 1.776,
+      "step": 3235
+    },
+    {
+      "epoch": 0.21832148512516425,
+      "grad_norm": 4.594212929510007,
+      "learning_rate": 9.579512883514656e-07,
+      "loss": 1.8087,
+      "step": 3240
+    },
+    {
+      "epoch": 0.218658400997271,
+      "grad_norm": 4.858762880181832,
+      "learning_rate": 9.577149284675619e-07,
+      "loss": 1.791,
+      "step": 3245
+    },
+    {
+      "epoch": 0.2189953168693777,
+      "grad_norm": 4.3746877376150595,
+      "learning_rate": 9.574779354695951e-07,
+      "loss": 1.7996,
+      "step": 3250
+    },
+    {
+      "epoch": 0.21933223274148445,
+      "grad_norm": 4.603856398111249,
+      "learning_rate": 9.572403096853754e-07,
+      "loss": 1.7624,
+      "step": 3255
+    },
+    {
+      "epoch": 0.2196691486135912,
+      "grad_norm": 4.5083144954047825,
+      "learning_rate": 9.570020514435878e-07,
+      "loss": 1.785,
+      "step": 3260
+    },
+    {
+      "epoch": 0.22000606448569793,
+      "grad_norm": 5.316421851324029,
+      "learning_rate": 9.567631610737929e-07,
+      "loss": 1.7005,
+      "step": 3265
+    },
+    {
+      "epoch": 0.22034298035780467,
+      "grad_norm": 4.881523181211674,
+      "learning_rate": 9.565236389064255e-07,
+      "loss": 1.8059,
+      "step": 3270
+    },
+    {
+      "epoch": 0.22067989622991138,
+      "grad_norm": 4.474918888361089,
+      "learning_rate": 9.562834852727935e-07,
+      "loss": 1.7975,
+      "step": 3275
+    },
+    {
+      "epoch": 0.22101681210201812,
+      "grad_norm": 4.423307719931098,
+      "learning_rate": 9.560427005050793e-07,
+      "loss": 1.7793,
+      "step": 3280
+    },
+    {
+      "epoch": 0.22135372797412486,
+      "grad_norm": 4.413118229643694,
+      "learning_rate": 9.55801284936338e-07,
+      "loss": 1.7529,
+      "step": 3285
+    },
+    {
+      "epoch": 0.2216906438462316,
+      "grad_norm": 4.933721416291058,
+      "learning_rate": 9.555592389004966e-07,
+      "loss": 1.7266,
+      "step": 3290
+    },
+    {
+      "epoch": 0.22202755971833832,
+      "grad_norm": 4.731100188336562,
+      "learning_rate": 9.553165627323548e-07,
+      "loss": 1.7748,
+      "step": 3295
+    },
+    {
+      "epoch": 0.22236447559044506,
+      "grad_norm": 4.994696255584328,
+      "learning_rate": 9.55073256767584e-07,
+      "loss": 1.7628,
+      "step": 3300
+    },
+    {
+      "epoch": 0.2227013914625518,
+      "grad_norm": 4.733796536914988,
+      "learning_rate": 9.548293213427262e-07,
+      "loss": 1.7621,
+      "step": 3305
+    },
+    {
+      "epoch": 0.22303830733465854,
+      "grad_norm": 4.6191701476158835,
+      "learning_rate": 9.545847567951944e-07,
+      "loss": 1.7833,
+      "step": 3310
+    },
+    {
+      "epoch": 0.22337522320676528,
+      "grad_norm": 4.655742693958854,
+      "learning_rate": 9.543395634632721e-07,
+      "loss": 1.793,
+      "step": 3315
+    },
+    {
+      "epoch": 0.223712139078872,
+      "grad_norm": 4.447754347858813,
+      "learning_rate": 9.540937416861117e-07,
+      "loss": 1.7705,
+      "step": 3320
+    },
+    {
+      "epoch": 0.22404905495097874,
+      "grad_norm": 5.29641811601241,
+      "learning_rate": 9.538472918037356e-07,
+      "loss": 1.8197,
+      "step": 3325
+    },
+    {
+      "epoch": 0.22438597082308548,
+      "grad_norm": 5.925090397904477,
+      "learning_rate": 9.536002141570348e-07,
+      "loss": 1.8162,
+      "step": 3330
+    },
+    {
+      "epoch": 0.22472288669519222,
+      "grad_norm": 4.732129210837873,
+      "learning_rate": 9.533525090877688e-07,
+      "loss": 1.8002,
+      "step": 3335
+    },
+    {
+      "epoch": 0.22505980256729893,
+      "grad_norm": 4.827993433570109,
+      "learning_rate": 9.531041769385641e-07,
+      "loss": 1.7571,
+      "step": 3340
+    },
+    {
+      "epoch": 0.22539671843940567,
+      "grad_norm": 4.967863123114332,
+      "learning_rate": 9.528552180529161e-07,
+      "loss": 1.8661,
+      "step": 3345
+    },
+    {
+      "epoch": 0.22573363431151242,
+      "grad_norm": 6.94281139383591,
+      "learning_rate": 9.526056327751856e-07,
+      "loss": 1.7828,
+      "step": 3350
+    },
+    {
+      "epoch": 0.22607055018361916,
+      "grad_norm": 4.9852835606633406,
+      "learning_rate": 9.523554214506006e-07,
+      "loss": 1.8015,
+      "step": 3355
+    },
+    {
+      "epoch": 0.2264074660557259,
+      "grad_norm": 5.415149724311492,
+      "learning_rate": 9.521045844252551e-07,
+      "loss": 1.8005,
+      "step": 3360
+    },
+    {
+      "epoch": 0.2267443819278326,
+      "grad_norm": 4.803260723539399,
+      "learning_rate": 9.518531220461084e-07,
+      "loss": 1.7508,
+      "step": 3365
+    },
+    {
+      "epoch": 0.22708129779993935,
+      "grad_norm": 4.0781328070330956,
+      "learning_rate": 9.516010346609845e-07,
+      "loss": 1.7467,
+      "step": 3370
+    },
+    {
+      "epoch": 0.2274182136720461,
+      "grad_norm": 4.638943670268183,
+      "learning_rate": 9.513483226185723e-07,
+      "loss": 1.8606,
+      "step": 3375
+    },
+    {
+      "epoch": 0.22775512954415283,
+      "grad_norm": 4.7751639719223435,
+      "learning_rate": 9.510949862684248e-07,
+      "loss": 1.7434,
+      "step": 3380
+    },
+    {
+      "epoch": 0.22809204541625955,
+      "grad_norm": 4.485643625780828,
+      "learning_rate": 9.508410259609583e-07,
+      "loss": 1.7891,
+      "step": 3385
+    },
+    {
+      "epoch": 0.2284289612883663,
+      "grad_norm": 4.787103456250361,
+      "learning_rate": 9.505864420474522e-07,
+      "loss": 1.7787,
+      "step": 3390
+    },
+    {
+      "epoch": 0.22876587716047303,
+      "grad_norm": 4.746721577697795,
+      "learning_rate": 9.503312348800485e-07,
+      "loss": 1.737,
+      "step": 3395
+    },
+    {
+      "epoch": 0.22910279303257977,
+      "grad_norm": 4.484589801607281,
+      "learning_rate": 9.500754048117514e-07,
+      "loss": 1.7541,
+      "step": 3400
+    },
+    {
+      "epoch": 0.2294397089046865,
+      "grad_norm": 4.502077592044502,
+      "learning_rate": 9.498189521964263e-07,
+      "loss": 1.7034,
+      "step": 3405
+    },
+    {
+      "epoch": 0.22977662477679323,
+      "grad_norm": 4.690815487498,
+      "learning_rate": 9.495618773888006e-07,
+      "loss": 1.6797,
+      "step": 3410
+    },
+    {
+      "epoch": 0.23011354064889997,
+      "grad_norm": 4.762323506148509,
+      "learning_rate": 9.49304180744461e-07,
+      "loss": 1.7822,
+      "step": 3415
+    },
+    {
+      "epoch": 0.2304504565210067,
+      "grad_norm": 4.387162226702979,
+      "learning_rate": 9.490458626198556e-07,
+      "loss": 1.7564,
+      "step": 3420
+    },
+    {
+      "epoch": 0.23078737239311345,
+      "grad_norm": 5.108928623435626,
+      "learning_rate": 9.487869233722915e-07,
+      "loss": 1.752,
+      "step": 3425
+    },
+    {
+      "epoch": 0.23112428826522016,
+      "grad_norm": 4.696698057501602,
+      "learning_rate": 9.485273633599348e-07,
+      "loss": 1.754,
+      "step": 3430
+    },
+    {
+      "epoch": 0.2314612041373269,
+      "grad_norm": 4.664626924419995,
+      "learning_rate": 9.482671829418107e-07,
+      "loss": 1.7531,
+      "step": 3435
+    },
+    {
+      "epoch": 0.23179812000943364,
+      "grad_norm": 4.5031239498515,
+      "learning_rate": 9.480063824778024e-07,
+      "loss": 1.7475,
+      "step": 3440
+    },
+    {
+      "epoch": 0.23213503588154039,
+      "grad_norm": 6.209457625066587,
+      "learning_rate": 9.477449623286505e-07,
+      "loss": 1.6885,
+      "step": 3445
+    },
+    {
+      "epoch": 0.23247195175364713,
+      "grad_norm": 4.361227999396086,
+      "learning_rate": 9.474829228559529e-07,
+      "loss": 1.8086,
+      "step": 3450
+    },
+    {
+      "epoch": 0.23280886762575384,
+      "grad_norm": 4.969357599319759,
+      "learning_rate": 9.472202644221643e-07,
+      "loss": 1.839,
+      "step": 3455
+    },
+    {
+      "epoch": 0.23314578349786058,
+      "grad_norm": 4.872284734828359,
+      "learning_rate": 9.469569873905955e-07,
+      "loss": 1.7324,
+      "step": 3460
+    },
+    {
+      "epoch": 0.23348269936996732,
+      "grad_norm": 4.601976719361907,
+      "learning_rate": 9.466930921254128e-07,
+      "loss": 1.7301,
+      "step": 3465
+    },
+    {
+      "epoch": 0.23381961524207406,
+      "grad_norm": 4.904552543549206,
+      "learning_rate": 9.464285789916376e-07,
+      "loss": 1.777,
+      "step": 3470
+    },
+    {
+      "epoch": 0.23415653111418078,
+      "grad_norm": 4.83831726667101,
+      "learning_rate": 9.461634483551464e-07,
+      "loss": 1.7213,
+      "step": 3475
+    },
+    {
+      "epoch": 0.23449344698628752,
+      "grad_norm": 4.658287483564059,
+      "learning_rate": 9.458977005826691e-07,
+      "loss": 1.733,
+      "step": 3480
+    },
+    {
+      "epoch": 0.23483036285839426,
+      "grad_norm": 4.591347678384667,
+      "learning_rate": 9.456313360417899e-07,
+      "loss": 1.7999,
+      "step": 3485
+    },
+    {
+      "epoch": 0.235167278730501,
+      "grad_norm": 4.411330890134835,
+      "learning_rate": 9.453643551009459e-07,
+      "loss": 1.7489,
+      "step": 3490
+    },
+    {
+      "epoch": 0.23550419460260774,
+      "grad_norm": 4.8367941401303645,
+      "learning_rate": 9.450967581294265e-07,
+      "loss": 1.8119,
+      "step": 3495
+    },
+    {
+      "epoch": 0.23584111047471445,
+      "grad_norm": 4.344443850260619,
+      "learning_rate": 9.448285454973737e-07,
+      "loss": 1.7566,
+      "step": 3500
+    },
+    {
+      "epoch": 0.2361780263468212,
+      "grad_norm": 5.118104466239838,
+      "learning_rate": 9.445597175757806e-07,
+      "loss": 1.7038,
+      "step": 3505
+    },
+    {
+      "epoch": 0.23651494221892794,
+      "grad_norm": 4.453125272027257,
+      "learning_rate": 9.442902747364918e-07,
+      "loss": 1.7057,
+      "step": 3510
+    },
+    {
+      "epoch": 0.23685185809103468,
+      "grad_norm": 4.610512909539744,
+      "learning_rate": 9.440202173522022e-07,
+      "loss": 1.7718,
+      "step": 3515
+    },
+    {
+      "epoch": 0.2371887739631414,
+      "grad_norm": 4.661575193646813,
+      "learning_rate": 9.437495457964568e-07,
+      "loss": 1.762,
+      "step": 3520
+    },
+    {
+      "epoch": 0.23752568983524813,
+      "grad_norm": 4.4907095983362595,
+      "learning_rate": 9.434782604436502e-07,
+      "loss": 1.7386,
+      "step": 3525
+    },
+    {
+      "epoch": 0.23786260570735487,
+      "grad_norm": 4.405418469834188,
+      "learning_rate": 9.432063616690258e-07,
+      "loss": 1.7547,
+      "step": 3530
+    },
+    {
+      "epoch": 0.23819952157946161,
+      "grad_norm": 5.532433958681073,
+      "learning_rate": 9.429338498486758e-07,
+      "loss": 1.7206,
+      "step": 3535
+    },
+    {
+      "epoch": 0.23853643745156836,
+      "grad_norm": 4.468416192322411,
+      "learning_rate": 9.426607253595402e-07,
+      "loss": 1.7891,
+      "step": 3540
+    },
+    {
+      "epoch": 0.23887335332367507,
+      "grad_norm": 4.6161099223279525,
+      "learning_rate": 9.423869885794063e-07,
+      "loss": 1.7656,
+      "step": 3545
+    },
+    {
+      "epoch": 0.2392102691957818,
+      "grad_norm": 5.047487557295027,
+      "learning_rate": 9.421126398869086e-07,
+      "loss": 1.8102,
+      "step": 3550
+    },
+    {
+      "epoch": 0.23954718506788855,
+      "grad_norm": 4.922422180793085,
+      "learning_rate": 9.418376796615279e-07,
+      "loss": 1.7425,
+      "step": 3555
+    },
+    {
+      "epoch": 0.2398841009399953,
+      "grad_norm": 4.660146502713987,
+      "learning_rate": 9.415621082835908e-07,
+      "loss": 1.7607,
+      "step": 3560
+    },
+    {
+      "epoch": 0.240221016812102,
+      "grad_norm": 4.433734686963888,
+      "learning_rate": 9.412859261342691e-07,
+      "loss": 1.7146,
+      "step": 3565
+    },
+    {
+      "epoch": 0.24055793268420875,
+      "grad_norm": 4.960946433754072,
+      "learning_rate": 9.410091335955798e-07,
+      "loss": 1.7406,
+      "step": 3570
+    },
+    {
+      "epoch": 0.2408948485563155,
+      "grad_norm": 5.0094940207061125,
+      "learning_rate": 9.407317310503841e-07,
+      "loss": 1.7632,
+      "step": 3575
+    },
+    {
+      "epoch": 0.24123176442842223,
+      "grad_norm": 5.031730983676167,
+      "learning_rate": 9.404537188823869e-07,
+      "loss": 1.813,
+      "step": 3580
+    },
+    {
+      "epoch": 0.24156868030052897,
+      "grad_norm": 4.494391366636777,
+      "learning_rate": 9.40175097476136e-07,
+      "loss": 1.7598,
+      "step": 3585
+    },
+    {
+      "epoch": 0.24190559617263568,
+      "grad_norm": 5.229951556761748,
+      "learning_rate": 9.398958672170225e-07,
+      "loss": 1.7348,
+      "step": 3590
+    },
+    {
+      "epoch": 0.24224251204474243,
+      "grad_norm": 4.599923341117953,
+      "learning_rate": 9.396160284912795e-07,
+      "loss": 1.7758,
+      "step": 3595
+    },
+    {
+      "epoch": 0.24257942791684917,
+      "grad_norm": 4.664723816998234,
+      "learning_rate": 9.393355816859813e-07,
+      "loss": 1.7659,
+      "step": 3600
+    },
+    {
+      "epoch": 0.2429163437889559,
+      "grad_norm": 5.234803010614216,
+      "learning_rate": 9.390545271890437e-07,
+      "loss": 1.7718,
+      "step": 3605
+    },
+    {
+      "epoch": 0.24325325966106262,
+      "grad_norm": 4.8706966068665825,
+      "learning_rate": 9.387728653892233e-07,
+      "loss": 1.7528,
+      "step": 3610
+    },
+    {
+      "epoch": 0.24359017553316936,
+      "grad_norm": 4.705019456003629,
+      "learning_rate": 9.384905966761159e-07,
+      "loss": 1.7693,
+      "step": 3615
+    },
+    {
+      "epoch": 0.2439270914052761,
+      "grad_norm": 5.494231676301332,
+      "learning_rate": 9.382077214401576e-07,
+      "loss": 1.7571,
+      "step": 3620
+    },
+    {
+      "epoch": 0.24426400727738284,
+      "grad_norm": 5.008941519368937,
+      "learning_rate": 9.379242400726232e-07,
+      "loss": 1.772,
+      "step": 3625
+    },
+    {
+      "epoch": 0.24460092314948959,
+      "grad_norm": 5.222723460942831,
+      "learning_rate": 9.376401529656257e-07,
+      "loss": 1.7858,
+      "step": 3630
+    },
+    {
+      "epoch": 0.2449378390215963,
+      "grad_norm": 4.47696532920893,
+      "learning_rate": 9.373554605121161e-07,
+      "loss": 1.6848,
+      "step": 3635
+    },
+    {
+      "epoch": 0.24527475489370304,
+      "grad_norm": 4.877984037043299,
+      "learning_rate": 9.370701631058828e-07,
+      "loss": 1.7502,
+      "step": 3640
+    },
+    {
+      "epoch": 0.24561167076580978,
+      "grad_norm": 5.048132174905781,
+      "learning_rate": 9.367842611415508e-07,
+      "loss": 1.7651,
+      "step": 3645
+    },
+    {
+      "epoch": 0.24594858663791652,
+      "grad_norm": 4.855385590365053,
+      "learning_rate": 9.364977550145816e-07,
+      "loss": 1.7338,
+      "step": 3650
+    },
+    {
+      "epoch": 0.24628550251002324,
+      "grad_norm": 4.5235322260702535,
+      "learning_rate": 9.362106451212721e-07,
+      "loss": 1.8073,
+      "step": 3655
+    },
+    {
+      "epoch": 0.24662241838212998,
+      "grad_norm": 5.083006810411546,
+      "learning_rate": 9.359229318587545e-07,
+      "loss": 1.7719,
+      "step": 3660
+    },
+    {
+      "epoch": 0.24695933425423672,
+      "grad_norm": 4.72527339625108,
+      "learning_rate": 9.356346156249954e-07,
+      "loss": 1.7813,
+      "step": 3665
+    },
+    {
+      "epoch": 0.24729625012634346,
+      "grad_norm": 4.6472501747489625,
+      "learning_rate": 9.353456968187958e-07,
+      "loss": 1.799,
+      "step": 3670
+    },
+    {
+      "epoch": 0.2476331659984502,
+      "grad_norm": 4.272567937938117,
+      "learning_rate": 9.350561758397897e-07,
+      "loss": 1.769,
+      "step": 3675
+    },
+    {
+      "epoch": 0.2479700818705569,
+      "grad_norm": 5.471616163784983,
+      "learning_rate": 9.347660530884442e-07,
+      "loss": 1.7788,
+      "step": 3680
+    },
+    {
+      "epoch": 0.24830699774266365,
+      "grad_norm": 4.558443147025116,
+      "learning_rate": 9.344753289660592e-07,
+      "loss": 1.7638,
+      "step": 3685
+    },
+    {
+      "epoch": 0.2486439136147704,
+      "grad_norm": 5.069595483982285,
+      "learning_rate": 9.34184003874766e-07,
+      "loss": 1.7975,
+      "step": 3690
+    },
+    {
+      "epoch": 0.24898082948687714,
+      "grad_norm": 4.760449032055557,
+      "learning_rate": 9.338920782175269e-07,
+      "loss": 1.7857,
+      "step": 3695
+    },
+    {
+      "epoch": 0.24931774535898385,
+      "grad_norm": 5.0621587885567205,
+      "learning_rate": 9.335995523981355e-07,
+      "loss": 1.7693,
+      "step": 3700
+    },
+    {
+      "epoch": 0.2496546612310906,
+      "grad_norm": 5.071621445026028,
+      "learning_rate": 9.333064268212153e-07,
+      "loss": 1.699,
+      "step": 3705
+    },
+    {
+      "epoch": 0.24999157710319733,
+      "grad_norm": 5.508852897124171,
+      "learning_rate": 9.330127018922193e-07,
+      "loss": 1.7654,
+      "step": 3710
+    },
+    {
+      "epoch": 0.2503284929753041,
+      "grad_norm": 4.415117796587092,
+      "learning_rate": 9.327183780174296e-07,
+      "loss": 1.7226,
+      "step": 3715
+    },
+    {
+      "epoch": 0.2506654088474108,
+      "grad_norm": 4.6970538069049095,
+      "learning_rate": 9.324234556039567e-07,
+      "loss": 1.7406,
+      "step": 3720
+    },
+    {
+      "epoch": 0.25100232471951756,
+      "grad_norm": 4.582266018979021,
+      "learning_rate": 9.321279350597393e-07,
+      "loss": 1.7339,
+      "step": 3725
+    },
+    {
+      "epoch": 0.25133924059162427,
+      "grad_norm": 4.240943943779937,
+      "learning_rate": 9.31831816793543e-07,
+      "loss": 1.7072,
+      "step": 3730
+    },
+    {
+      "epoch": 0.251676156463731,
+      "grad_norm": 4.502592636166306,
+      "learning_rate": 9.315351012149605e-07,
+      "loss": 1.7358,
+      "step": 3735
+    },
+    {
+      "epoch": 0.25201307233583775,
+      "grad_norm": 4.506192024283769,
+      "learning_rate": 9.312377887344105e-07,
+      "loss": 1.6949,
+      "step": 3740
+    },
+    {
+      "epoch": 0.25234998820794446,
+      "grad_norm": 4.595924942826764,
+      "learning_rate": 9.309398797631374e-07,
+      "loss": 1.7713,
+      "step": 3745
+    },
+    {
+      "epoch": 0.25268690408005123,
+      "grad_norm": 4.646056480790715,
+      "learning_rate": 9.306413747132108e-07,
+      "loss": 1.7086,
+      "step": 3750
+    },
+    {
+      "epoch": 0.25302381995215795,
+      "grad_norm": 5.667337214028473,
+      "learning_rate": 9.303422739975246e-07,
+      "loss": 1.8206,
+      "step": 3755
+    },
+    {
+      "epoch": 0.25336073582426466,
+      "grad_norm": 4.852279815852281,
+      "learning_rate": 9.300425780297968e-07,
+      "loss": 1.7577,
+      "step": 3760
+    },
+    {
+      "epoch": 0.25369765169637143,
+      "grad_norm": 4.908151755388344,
+      "learning_rate": 9.297422872245686e-07,
+      "loss": 1.7385,
+      "step": 3765
+    },
+    {
+      "epoch": 0.25403456756847814,
+      "grad_norm": 4.764781340681072,
+      "learning_rate": 9.294414019972043e-07,
+      "loss": 1.7649,
+      "step": 3770
+    },
+    {
+      "epoch": 0.2543714834405849,
+      "grad_norm": 5.077139413996835,
+      "learning_rate": 9.291399227638898e-07,
+      "loss": 1.7791,
+      "step": 3775
+    },
+    {
+      "epoch": 0.2547083993126916,
+      "grad_norm": 4.822220559468089,
+      "learning_rate": 9.288378499416332e-07,
+      "loss": 1.7173,
+      "step": 3780
+    },
+    {
+      "epoch": 0.25504531518479834,
+      "grad_norm": 4.829738862421561,
+      "learning_rate": 9.285351839482634e-07,
+      "loss": 1.7591,
+      "step": 3785
+    },
+    {
+      "epoch": 0.2553822310569051,
+      "grad_norm": 4.7186128991399094,
+      "learning_rate": 9.2823192520243e-07,
+      "loss": 1.7636,
+      "step": 3790
+    },
+    {
+      "epoch": 0.2557191469290118,
+      "grad_norm": 5.155182474504667,
+      "learning_rate": 9.27928074123602e-07,
+      "loss": 1.7268,
+      "step": 3795
+    },
+    {
+      "epoch": 0.25605606280111853,
+      "grad_norm": 4.627076461643508,
+      "learning_rate": 9.276236311320684e-07,
+      "loss": 1.7572,
+      "step": 3800
+    },
+    {
+      "epoch": 0.2563929786732253,
+      "grad_norm": 4.7149669408466,
+      "learning_rate": 9.273185966489365e-07,
+      "loss": 1.7925,
+      "step": 3805
+    },
+    {
+      "epoch": 0.256729894545332,
+      "grad_norm": 4.654856525421342,
+      "learning_rate": 9.270129710961318e-07,
+      "loss": 1.6885,
+      "step": 3810
+    },
+    {
+      "epoch": 0.2570668104174388,
+      "grad_norm": 4.272444207688499,
+      "learning_rate": 9.267067548963974e-07,
+      "loss": 1.7254,
+      "step": 3815
+    },
+    {
+      "epoch": 0.2574037262895455,
+      "grad_norm": 5.00860276979963,
+      "learning_rate": 9.263999484732934e-07,
+      "loss": 1.7835,
+      "step": 3820
+    },
+    {
+      "epoch": 0.2577406421616522,
+      "grad_norm": 4.624510433408824,
+      "learning_rate": 9.260925522511962e-07,
+      "loss": 1.7666,
+      "step": 3825
+    },
+    {
+      "epoch": 0.258077558033759,
+      "grad_norm": 4.466190636554243,
+      "learning_rate": 9.257845666552984e-07,
+      "loss": 1.7831,
+      "step": 3830
+    },
+    {
+      "epoch": 0.2584144739058657,
+      "grad_norm": 4.650752529976314,
+      "learning_rate": 9.254759921116073e-07,
+      "loss": 1.7767,
+      "step": 3835
+    },
+    {
+      "epoch": 0.25875138977797246,
+      "grad_norm": 4.947586647901737,
+      "learning_rate": 9.251668290469452e-07,
+      "loss": 1.7696,
+      "step": 3840
+    },
+    {
+      "epoch": 0.2590883056500792,
+      "grad_norm": 4.3973160545600365,
+      "learning_rate": 9.248570778889484e-07,
+      "loss": 1.7347,
+      "step": 3845
+    },
+    {
+      "epoch": 0.2594252215221859,
+      "grad_norm": 4.971082227222947,
+      "learning_rate": 9.245467390660664e-07,
+      "loss": 1.6983,
+      "step": 3850
+    },
+    {
+      "epoch": 0.25976213739429266,
+      "grad_norm": 4.770499620243425,
+      "learning_rate": 9.242358130075618e-07,
+      "loss": 1.7691,
+      "step": 3855
+    },
+    {
+      "epoch": 0.26009905326639937,
+      "grad_norm": 4.5826918853514735,
+      "learning_rate": 9.239243001435093e-07,
+      "loss": 1.709,
+      "step": 3860
+    },
+    {
+      "epoch": 0.26043596913850614,
+      "grad_norm": 4.714242048087625,
+      "learning_rate": 9.236122009047957e-07,
+      "loss": 1.6842,
+      "step": 3865
+    },
+    {
+      "epoch": 0.26077288501061285,
+      "grad_norm": 5.01574436470716,
+      "learning_rate": 9.232995157231182e-07,
+      "loss": 1.7489,
+      "step": 3870
+    },
+    {
+      "epoch": 0.26110980088271957,
+      "grad_norm": 4.443135356323211,
+      "learning_rate": 9.229862450309851e-07,
+      "loss": 1.7254,
+      "step": 3875
+    },
+    {
+      "epoch": 0.26144671675482634,
+      "grad_norm": 4.552093105227505,
+      "learning_rate": 9.226723892617141e-07,
+      "loss": 1.7634,
+      "step": 3880
+    },
+    {
+      "epoch": 0.26178363262693305,
+      "grad_norm": 3.9443617773456405,
+      "learning_rate": 9.223579488494327e-07,
+      "loss": 1.7155,
+      "step": 3885
+    },
+    {
+      "epoch": 0.26212054849903976,
+      "grad_norm": 4.5681734596339965,
+      "learning_rate": 9.220429242290763e-07,
+      "loss": 1.7267,
+      "step": 3890
+    },
+    {
+      "epoch": 0.26245746437114653,
+      "grad_norm": 4.716056869546119,
+      "learning_rate": 9.217273158363894e-07,
+      "loss": 1.6855,
+      "step": 3895
+    },
+    {
+      "epoch": 0.26279438024325324,
+      "grad_norm": 4.24968952182465,
+      "learning_rate": 9.214111241079232e-07,
+      "loss": 1.7645,
+      "step": 3900
+    },
+    {
+      "epoch": 0.26313129611536,
+      "grad_norm": 4.424195580222474,
+      "learning_rate": 9.21094349481036e-07,
+      "loss": 1.7845,
+      "step": 3905
+    },
+    {
+      "epoch": 0.2634682119874667,
+      "grad_norm": 4.7637558094248265,
+      "learning_rate": 9.207769923938924e-07,
+      "loss": 1.7245,
+      "step": 3910
+    },
+    {
+      "epoch": 0.26380512785957344,
+      "grad_norm": 5.284391995626364,
+      "learning_rate": 9.204590532854627e-07,
+      "loss": 1.662,
+      "step": 3915
+    },
+    {
+      "epoch": 0.2641420437316802,
+      "grad_norm": 4.7254299136588855,
+      "learning_rate": 9.20140532595522e-07,
+      "loss": 1.8177,
+      "step": 3920
+    },
+    {
+      "epoch": 0.2644789596037869,
+      "grad_norm": 4.555412680215885,
+      "learning_rate": 9.198214307646504e-07,
+      "loss": 1.7283,
+      "step": 3925
+    },
+    {
+      "epoch": 0.2648158754758937,
+      "grad_norm": 4.676744050611532,
+      "learning_rate": 9.195017482342313e-07,
+      "loss": 1.7688,
+      "step": 3930
+    },
+    {
+      "epoch": 0.2651527913480004,
+      "grad_norm": 4.923776597293947,
+      "learning_rate": 9.191814854464514e-07,
+      "loss": 1.743,
+      "step": 3935
+    },
+    {
+      "epoch": 0.2654897072201071,
+      "grad_norm": 4.624858906489115,
+      "learning_rate": 9.188606428443002e-07,
+      "loss": 1.727,
+      "step": 3940
+    },
+    {
+      "epoch": 0.2658266230922139,
+      "grad_norm": 4.50535143830333,
+      "learning_rate": 9.185392208715692e-07,
+      "loss": 1.6601,
+      "step": 3945
+    },
+    {
+      "epoch": 0.2661635389643206,
+      "grad_norm": 4.715817481252642,
+      "learning_rate": 9.182172199728513e-07,
+      "loss": 1.7566,
+      "step": 3950
+    },
+    {
+      "epoch": 0.26650045483642737,
+      "grad_norm": 5.076272945810852,
+      "learning_rate": 9.178946405935398e-07,
+      "loss": 1.7813,
+      "step": 3955
+    },
+    {
+      "epoch": 0.2668373707085341,
+      "grad_norm": 4.903244847025362,
+      "learning_rate": 9.175714831798287e-07,
+      "loss": 1.764,
+      "step": 3960
+    },
+    {
+      "epoch": 0.2671742865806408,
+      "grad_norm": 4.819745696387984,
+      "learning_rate": 9.172477481787113e-07,
+      "loss": 1.6826,
+      "step": 3965
+    },
+    {
+      "epoch": 0.26751120245274757,
+      "grad_norm": 4.686633159763879,
+      "learning_rate": 9.169234360379796e-07,
+      "loss": 1.7017,
+      "step": 3970
+    },
+    {
+      "epoch": 0.2678481183248543,
+      "grad_norm": 5.155877126897368,
+      "learning_rate": 9.165985472062244e-07,
+      "loss": 1.7647,
+      "step": 3975
+    },
+    {
+      "epoch": 0.268185034196961,
+      "grad_norm": 4.385164684257074,
+      "learning_rate": 9.162730821328337e-07,
+      "loss": 1.7524,
+      "step": 3980
+    },
+    {
+      "epoch": 0.26852195006906776,
+      "grad_norm": 4.881753774963838,
+      "learning_rate": 9.159470412679928e-07,
+      "loss": 1.7793,
+      "step": 3985
+    },
+    {
+      "epoch": 0.2688588659411745,
+      "grad_norm": 5.161223622188654,
+      "learning_rate": 9.156204250626836e-07,
+      "loss": 1.7525,
+      "step": 3990
+    },
+    {
+      "epoch": 0.26919578181328124,
+      "grad_norm": 5.334488964309807,
+      "learning_rate": 9.152932339686833e-07,
+      "loss": 1.784,
+      "step": 3995
+    },
+    {
+      "epoch": 0.26953269768538796,
+      "grad_norm": 4.5325612354681155,
+      "learning_rate": 9.149654684385647e-07,
+      "loss": 1.7371,
+      "step": 4000
+    },
+    {
+      "epoch": 0.26986961355749467,
+      "grad_norm": 5.240484693296526,
+      "learning_rate": 9.146371289256952e-07,
+      "loss": 1.6598,
+      "step": 4005
+    },
+    {
+      "epoch": 0.27020652942960144,
+      "grad_norm": 4.377746387816449,
+      "learning_rate": 9.143082158842359e-07,
+      "loss": 1.7966,
+      "step": 4010
+    },
+    {
+      "epoch": 0.27054344530170815,
+      "grad_norm": 4.593458441322993,
+      "learning_rate": 9.139787297691413e-07,
+      "loss": 1.7707,
+      "step": 4015
+    },
+    {
+      "epoch": 0.2708803611738149,
+      "grad_norm": 4.6933419598528445,
+      "learning_rate": 9.136486710361586e-07,
+      "loss": 1.8005,
+      "step": 4020
+    },
+    {
+      "epoch": 0.27121727704592163,
+      "grad_norm": 4.423648259837012,
+      "learning_rate": 9.13318040141827e-07,
+      "loss": 1.7397,
+      "step": 4025
+    },
+    {
+      "epoch": 0.27155419291802835,
+      "grad_norm": 4.58381209779022,
+      "learning_rate": 9.129868375434774e-07,
+      "loss": 1.7167,
+      "step": 4030
+    },
+    {
+      "epoch": 0.2718911087901351,
+      "grad_norm": 5.591361058814464,
+      "learning_rate": 9.12655063699231e-07,
+      "loss": 1.7489,
+      "step": 4035
+    },
+    {
+      "epoch": 0.27222802466224183,
+      "grad_norm": 4.826560920094557,
+      "learning_rate": 9.123227190679994e-07,
+      "loss": 1.6668,
+      "step": 4040
+    },
+    {
+      "epoch": 0.2725649405343486,
+      "grad_norm": 4.91009253142952,
+      "learning_rate": 9.119898041094838e-07,
+      "loss": 1.7787,
+      "step": 4045
+    },
+    {
+      "epoch": 0.2729018564064553,
+      "grad_norm": 4.752985829366886,
+      "learning_rate": 9.116563192841741e-07,
+      "loss": 1.7492,
+      "step": 4050
+    },
+    {
+      "epoch": 0.273238772278562,
+      "grad_norm": 4.934952686039615,
+      "learning_rate": 9.113222650533486e-07,
+      "loss": 1.7972,
+      "step": 4055
+    },
+    {
+      "epoch": 0.2735756881506688,
+      "grad_norm": 4.940692386574961,
+      "learning_rate": 9.109876418790731e-07,
+      "loss": 1.7277,
+      "step": 4060
+    },
+    {
+      "epoch": 0.2739126040227755,
+      "grad_norm": 4.917215340655875,
+      "learning_rate": 9.106524502242004e-07,
+      "loss": 1.7261,
+      "step": 4065
+    },
+    {
+      "epoch": 0.2742495198948822,
+      "grad_norm": 4.369817949456342,
+      "learning_rate": 9.103166905523699e-07,
+      "loss": 1.7005,
+      "step": 4070
+    },
+    {
+      "epoch": 0.274586435766989,
+      "grad_norm": 4.654979197581099,
+      "learning_rate": 9.099803633280059e-07,
+      "loss": 1.8421,
+      "step": 4075
+    },
+    {
+      "epoch": 0.2749233516390957,
+      "grad_norm": 4.987317284899288,
+      "learning_rate": 9.096434690163184e-07,
+      "loss": 1.7533,
+      "step": 4080
+    },
+    {
+      "epoch": 0.27526026751120247,
+      "grad_norm": 4.679060945617795,
+      "learning_rate": 9.093060080833019e-07,
+      "loss": 1.7128,
+      "step": 4085
+    },
+    {
+      "epoch": 0.2755971833833092,
+      "grad_norm": 4.934341742568025,
+      "learning_rate": 9.089679809957343e-07,
+      "loss": 1.7116,
+      "step": 4090
+    },
+    {
+      "epoch": 0.2759340992554159,
+      "grad_norm": 5.535564494067952,
+      "learning_rate": 9.086293882211768e-07,
+      "loss": 1.795,
+      "step": 4095
+    },
+    {
+      "epoch": 0.27627101512752267,
+      "grad_norm": 4.59416034924243,
+      "learning_rate": 9.082902302279726e-07,
+      "loss": 1.7278,
+      "step": 4100
+    },
+    {
+      "epoch": 0.2766079309996294,
+      "grad_norm": 4.81347731163381,
+      "learning_rate": 9.079505074852476e-07,
+      "loss": 1.7593,
+      "step": 4105
+    },
+    {
+      "epoch": 0.27694484687173615,
+      "grad_norm": 4.9285174913634915,
+      "learning_rate": 9.076102204629082e-07,
+      "loss": 1.7159,
+      "step": 4110
+    },
+    {
+      "epoch": 0.27728176274384286,
+      "grad_norm": 4.3498627752306716,
+      "learning_rate": 9.072693696316411e-07,
+      "loss": 1.7344,
+      "step": 4115
+    },
+    {
+      "epoch": 0.2776186786159496,
+      "grad_norm": 4.656514571023645,
+      "learning_rate": 9.069279554629137e-07,
+      "loss": 1.7313,
+      "step": 4120
+    },
+    {
+      "epoch": 0.27795559448805635,
+      "grad_norm": 4.755094094103996,
+      "learning_rate": 9.06585978428972e-07,
+      "loss": 1.7245,
+      "step": 4125
+    },
+    {
+      "epoch": 0.27829251036016306,
+      "grad_norm": 4.70651566158845,
+      "learning_rate": 9.062434390028407e-07,
+      "loss": 1.7576,
+      "step": 4130
+    },
+    {
+      "epoch": 0.27862942623226983,
+      "grad_norm": 4.5157833239482486,
+      "learning_rate": 9.059003376583223e-07,
+      "loss": 1.7557,
+      "step": 4135
+    },
+    {
+      "epoch": 0.27896634210437654,
+      "grad_norm": 4.3165534493294455,
+      "learning_rate": 9.055566748699968e-07,
+      "loss": 1.7643,
+      "step": 4140
+    },
+    {
+      "epoch": 0.27930325797648325,
+      "grad_norm": 4.511559044415423,
+      "learning_rate": 9.052124511132204e-07,
+      "loss": 1.7322,
+      "step": 4145
+    },
+    {
+      "epoch": 0.27964017384859,
+      "grad_norm": 4.491252006627144,
+      "learning_rate": 9.04867666864126e-07,
+      "loss": 1.7048,
+      "step": 4150
+    },
+    {
+      "epoch": 0.27997708972069674,
+      "grad_norm": 5.162440504545584,
+      "learning_rate": 9.045223225996207e-07,
+      "loss": 1.7349,
+      "step": 4155
+    },
+    {
+      "epoch": 0.28031400559280345,
+      "grad_norm": 4.837794801527783,
+      "learning_rate": 9.041764187973871e-07,
+      "loss": 1.7243,
+      "step": 4160
+    },
+    {
+      "epoch": 0.2806509214649102,
+      "grad_norm": 4.86864277556133,
+      "learning_rate": 9.038299559358815e-07,
+      "loss": 1.7357,
+      "step": 4165
+    },
+    {
+      "epoch": 0.28098783733701693,
+      "grad_norm": 4.675098805367262,
+      "learning_rate": 9.034829344943331e-07,
+      "loss": 1.7346,
+      "step": 4170
+    },
+    {
+      "epoch": 0.2813247532091237,
+      "grad_norm": 4.675719749121135,
+      "learning_rate": 9.031353549527444e-07,
+      "loss": 1.755,
+      "step": 4175
+    },
+    {
+      "epoch": 0.2816616690812304,
+      "grad_norm": 4.656544713004674,
+      "learning_rate": 9.027872177918894e-07,
+      "loss": 1.6781,
+      "step": 4180
+    },
+    {
+      "epoch": 0.28199858495333713,
+      "grad_norm": 4.093237425850474,
+      "learning_rate": 9.024385234933134e-07,
+      "loss": 1.7954,
+      "step": 4185
+    },
+    {
+      "epoch": 0.2823355008254439,
+      "grad_norm": 4.392579552620834,
+      "learning_rate": 9.020892725393326e-07,
+      "loss": 1.6816,
+      "step": 4190
+    },
+    {
+      "epoch": 0.2826724166975506,
+      "grad_norm": 4.79325848165355,
+      "learning_rate": 9.017394654130332e-07,
+      "loss": 1.7271,
+      "step": 4195
+    },
+    {
+      "epoch": 0.2830093325696574,
+      "grad_norm": 4.486480700182664,
+      "learning_rate": 9.013891025982703e-07,
+      "loss": 1.7621,
+      "step": 4200
+    },
+    {
+      "epoch": 0.2833462484417641,
+      "grad_norm": 4.742598461547793,
+      "learning_rate": 9.010381845796677e-07,
+      "loss": 1.7021,
+      "step": 4205
+    },
+    {
+      "epoch": 0.2836831643138708,
+      "grad_norm": 4.910310236706784,
+      "learning_rate": 9.006867118426178e-07,
+      "loss": 1.7157,
+      "step": 4210
+    },
+    {
+      "epoch": 0.2840200801859776,
+      "grad_norm": 5.0356189419192425,
+      "learning_rate": 9.003346848732793e-07,
+      "loss": 1.7219,
+      "step": 4215
+    },
+    {
+      "epoch": 0.2843569960580843,
+      "grad_norm": 4.1734770119817,
+      "learning_rate": 8.999821041585787e-07,
+      "loss": 1.6756,
+      "step": 4220
+    },
+    {
+      "epoch": 0.28469391193019106,
+      "grad_norm": 4.596840625940221,
+      "learning_rate": 8.996289701862072e-07,
+      "loss": 1.8047,
+      "step": 4225
+    },
+    {
+      "epoch": 0.28503082780229777,
+      "grad_norm": 4.6224423547224385,
+      "learning_rate": 8.99275283444622e-07,
+      "loss": 1.7589,
+      "step": 4230
+    },
+    {
+      "epoch": 0.2853677436744045,
+      "grad_norm": 4.491183430135914,
+      "learning_rate": 8.989210444230449e-07,
+      "loss": 1.7477,
+      "step": 4235
+    },
+    {
+      "epoch": 0.28570465954651125,
+      "grad_norm": 4.685179009285273,
+      "learning_rate": 8.985662536114612e-07,
+      "loss": 1.7634,
+      "step": 4240
+    },
+    {
+      "epoch": 0.28604157541861797,
+      "grad_norm": 4.750370079829212,
+      "learning_rate": 8.9821091150062e-07,
+      "loss": 1.7168,
+      "step": 4245
+    },
+    {
+      "epoch": 0.2863784912907247,
+      "grad_norm": 5.005679815546969,
+      "learning_rate": 8.978550185820323e-07,
+      "loss": 1.7179,
+      "step": 4250
+    },
+    {
+      "epoch": 0.28671540716283145,
+      "grad_norm": 4.483443011591814,
+      "learning_rate": 8.974985753479718e-07,
+      "loss": 1.7561,
+      "step": 4255
+    },
+    {
+      "epoch": 0.28705232303493816,
+      "grad_norm": 4.514312535623764,
+      "learning_rate": 8.971415822914726e-07,
+      "loss": 1.744,
+      "step": 4260
+    },
+    {
+      "epoch": 0.28738923890704493,
+      "grad_norm": 4.70665221972189,
+      "learning_rate": 8.967840399063298e-07,
+      "loss": 1.7066,
+      "step": 4265
+    },
+    {
+      "epoch": 0.28772615477915164,
+      "grad_norm": 5.139115959292068,
+      "learning_rate": 8.964259486870982e-07,
+      "loss": 1.7072,
+      "step": 4270
+    },
+    {
+      "epoch": 0.28806307065125836,
+      "grad_norm": 4.664997783932267,
+      "learning_rate": 8.960673091290916e-07,
+      "loss": 1.7119,
+      "step": 4275
+    },
+    {
+      "epoch": 0.2883999865233651,
+      "grad_norm": 4.892565476825081,
+      "learning_rate": 8.957081217283825e-07,
+      "loss": 1.7438,
+      "step": 4280
+    },
+    {
+      "epoch": 0.28873690239547184,
+      "grad_norm": 4.559694716957321,
+      "learning_rate": 8.953483869818013e-07,
+      "loss": 1.7316,
+      "step": 4285
+    },
+    {
+      "epoch": 0.2890738182675786,
+      "grad_norm": 4.698321844576039,
+      "learning_rate": 8.949881053869348e-07,
+      "loss": 1.7336,
+      "step": 4290
+    },
+    {
+      "epoch": 0.2894107341396853,
+      "grad_norm": 4.765465874545801,
+      "learning_rate": 8.946272774421271e-07,
+      "loss": 1.7119,
+      "step": 4295
+    },
+    {
+      "epoch": 0.28974765001179204,
+      "grad_norm": 4.753263661353172,
+      "learning_rate": 8.942659036464775e-07,
+      "loss": 1.74,
+      "step": 4300
+    },
+    {
+      "epoch": 0.2900845658838988,
+      "grad_norm": 4.495261380126989,
+      "learning_rate": 8.939039844998403e-07,
+      "loss": 1.7512,
+      "step": 4305
+    },
+    {
+      "epoch": 0.2904214817560055,
+      "grad_norm": 4.557559595948686,
+      "learning_rate": 8.935415205028243e-07,
+      "loss": 1.7104,
+      "step": 4310
+    },
+    {
+      "epoch": 0.2907583976281123,
+      "grad_norm": 4.202594698597534,
+      "learning_rate": 8.931785121567921e-07,
+      "loss": 1.6826,
+      "step": 4315
+    },
+    {
+      "epoch": 0.291095313500219,
+      "grad_norm": 4.729706790893457,
+      "learning_rate": 8.928149599638588e-07,
+      "loss": 1.7567,
+      "step": 4320
+    },
+    {
+      "epoch": 0.2914322293723257,
+      "grad_norm": 4.812500510407962,
+      "learning_rate": 8.924508644268921e-07,
+      "loss": 1.8317,
+      "step": 4325
+    },
+    {
+      "epoch": 0.2917691452444325,
+      "grad_norm": 4.5072102814438395,
+      "learning_rate": 8.920862260495111e-07,
+      "loss": 1.7463,
+      "step": 4330
+    },
+    {
+      "epoch": 0.2921060611165392,
+      "grad_norm": 4.815210343847766,
+      "learning_rate": 8.917210453360859e-07,
+      "loss": 1.7456,
+      "step": 4335
+    },
+    {
+      "epoch": 0.2924429769886459,
+      "grad_norm": 5.3841241146256,
+      "learning_rate": 8.913553227917365e-07,
+      "loss": 1.679,
+      "step": 4340
+    },
+    {
+      "epoch": 0.2927798928607527,
+      "grad_norm": 4.844902712349313,
+      "learning_rate": 8.909890589223329e-07,
+      "loss": 1.729,
+      "step": 4345
+    },
+    {
+      "epoch": 0.2931168087328594,
+      "grad_norm": 4.5901739139881075,
+      "learning_rate": 8.906222542344932e-07,
+      "loss": 1.7552,
+      "step": 4350
+    },
+    {
+      "epoch": 0.29345372460496616,
+      "grad_norm": 4.538804261257026,
+      "learning_rate": 8.902549092355839e-07,
+      "loss": 1.7488,
+      "step": 4355
+    },
+    {
+      "epoch": 0.2937906404770729,
+      "grad_norm": 5.2062038276879825,
+      "learning_rate": 8.898870244337189e-07,
+      "loss": 1.8083,
+      "step": 4360
+    },
+    {
+      "epoch": 0.2941275563491796,
+      "grad_norm": 4.726134277199526,
+      "learning_rate": 8.895186003377586e-07,
+      "loss": 1.7307,
+      "step": 4365
+    },
+    {
+      "epoch": 0.29446447222128636,
+      "grad_norm": 5.516228137928749,
+      "learning_rate": 8.891496374573095e-07,
+      "loss": 1.8018,
+      "step": 4370
+    },
+    {
+      "epoch": 0.29480138809339307,
+      "grad_norm": 4.799889103520739,
+      "learning_rate": 8.887801363027233e-07,
+      "loss": 1.7538,
+      "step": 4375
+    },
+    {
+      "epoch": 0.29513830396549984,
+      "grad_norm": 4.379041171560821,
+      "learning_rate": 8.884100973850962e-07,
+      "loss": 1.7804,
+      "step": 4380
+    },
+    {
+      "epoch": 0.29547521983760655,
+      "grad_norm": 4.357662889414455,
+      "learning_rate": 8.880395212162684e-07,
+      "loss": 1.7696,
+      "step": 4385
+    },
+    {
+      "epoch": 0.29581213570971326,
+      "grad_norm": 4.917230947405535,
+      "learning_rate": 8.87668408308823e-07,
+      "loss": 1.6723,
+      "step": 4390
+    },
+    {
+      "epoch": 0.29614905158182003,
+      "grad_norm": 4.543552014294589,
+      "learning_rate": 8.872967591760856e-07,
+      "loss": 1.7145,
+      "step": 4395
+    },
+    {
+      "epoch": 0.29648596745392675,
+      "grad_norm": 4.472875036743257,
+      "learning_rate": 8.869245743321234e-07,
+      "loss": 1.7572,
+      "step": 4400
+    },
+    {
+      "epoch": 0.2968228833260335,
+      "grad_norm": 5.049587485702916,
+      "learning_rate": 8.865518542917452e-07,
+      "loss": 1.6626,
+      "step": 4405
+    },
+    {
+      "epoch": 0.29715979919814023,
+      "grad_norm": 4.672040595743891,
+      "learning_rate": 8.861785995704991e-07,
+      "loss": 1.764,
+      "step": 4410
+    },
+    {
+      "epoch": 0.29749671507024694,
+      "grad_norm": 4.495170230013994,
+      "learning_rate": 8.858048106846735e-07,
+      "loss": 1.7532,
+      "step": 4415
+    },
+    {
+      "epoch": 0.2978336309423537,
+      "grad_norm": 4.823875109397055,
+      "learning_rate": 8.854304881512955e-07,
+      "loss": 1.6923,
+      "step": 4420
+    },
+    {
+      "epoch": 0.2981705468144604,
+      "grad_norm": 4.821873035905844,
+      "learning_rate": 8.850556324881302e-07,
+      "loss": 1.7809,
+      "step": 4425
+    },
+    {
+      "epoch": 0.29850746268656714,
+      "grad_norm": 4.410412924960029,
+      "learning_rate": 8.846802442136804e-07,
+      "loss": 1.7013,
+      "step": 4430
+    },
+    {
+      "epoch": 0.2988443785586739,
+      "grad_norm": 4.390679279180808,
+      "learning_rate": 8.843043238471853e-07,
+      "loss": 1.7589,
+      "step": 4435
+    },
+    {
+      "epoch": 0.2991812944307806,
+      "grad_norm": 4.734392692475669,
+      "learning_rate": 8.839278719086201e-07,
+      "loss": 1.7504,
+      "step": 4440
+    },
+    {
+      "epoch": 0.2995182103028874,
+      "grad_norm": 4.403725418684815,
+      "learning_rate": 8.835508889186956e-07,
+      "loss": 1.7657,
+      "step": 4445
+    },
+    {
+      "epoch": 0.2998551261749941,
+      "grad_norm": 5.210710942210437,
+      "learning_rate": 8.83173375398857e-07,
+      "loss": 1.7567,
+      "step": 4450
+    },
+    {
+      "epoch": 0.3001920420471008,
+      "grad_norm": 4.66405783513022,
+      "learning_rate": 8.827953318712831e-07,
+      "loss": 1.7397,
+      "step": 4455
+    },
+    {
+      "epoch": 0.3005289579192076,
+      "grad_norm": 5.118422608455292,
+      "learning_rate": 8.824167588588861e-07,
+      "loss": 1.7122,
+      "step": 4460
+    },
+    {
+      "epoch": 0.3008658737913143,
+      "grad_norm": 4.680432440576183,
+      "learning_rate": 8.820376568853105e-07,
+      "loss": 1.6974,
+      "step": 4465
+    },
+    {
+      "epoch": 0.30120278966342107,
+      "grad_norm": 4.800049981676269,
+      "learning_rate": 8.816580264749325e-07,
+      "loss": 1.735,
+      "step": 4470
+    },
+    {
+      "epoch": 0.3015397055355278,
+      "grad_norm": 4.143241744072462,
+      "learning_rate": 8.81277868152859e-07,
+      "loss": 1.7497,
+      "step": 4475
+    },
+    {
+      "epoch": 0.3018766214076345,
+      "grad_norm": 4.739632425074766,
+      "learning_rate": 8.808971824449274e-07,
+      "loss": 1.7366,
+      "step": 4480
+    },
+    {
+      "epoch": 0.30221353727974126,
+      "grad_norm": 4.389277440452169,
+      "learning_rate": 8.805159698777045e-07,
+      "loss": 1.7156,
+      "step": 4485
+    },
+    {
+      "epoch": 0.302550453151848,
+      "grad_norm": 4.837990330467177,
+      "learning_rate": 8.801342309784858e-07,
+      "loss": 1.7185,
+      "step": 4490
+    },
+    {
+      "epoch": 0.30288736902395474,
+      "grad_norm": 4.185487779722894,
+      "learning_rate": 8.79751966275295e-07,
+      "loss": 1.6727,
+      "step": 4495
+    },
+    {
+      "epoch": 0.30322428489606146,
+      "grad_norm": 5.036912125810241,
+      "learning_rate": 8.793691762968827e-07,
+      "loss": 1.7088,
+      "step": 4500
+    },
+    {
+      "epoch": 0.30356120076816817,
+      "grad_norm": 4.863760552977038,
+      "learning_rate": 8.789858615727264e-07,
+      "loss": 1.7243,
+      "step": 4505
+    },
+    {
+      "epoch": 0.30389811664027494,
+      "grad_norm": 4.464862420363283,
+      "learning_rate": 8.786020226330295e-07,
+      "loss": 1.7092,
+      "step": 4510
+    },
+    {
+      "epoch": 0.30423503251238165,
+      "grad_norm": 4.631321680323668,
+      "learning_rate": 8.782176600087203e-07,
+      "loss": 1.6525,
+      "step": 4515
+    },
+    {
+      "epoch": 0.30457194838448837,
+      "grad_norm": 4.493210645127753,
+      "learning_rate": 8.778327742314513e-07,
+      "loss": 1.7311,
+      "step": 4520
+    },
+    {
+      "epoch": 0.30490886425659514,
+      "grad_norm": 4.721685250690692,
+      "learning_rate": 8.77447365833599e-07,
+      "loss": 1.7066,
+      "step": 4525
+    },
+    {
+      "epoch": 0.30524578012870185,
+      "grad_norm": 4.731921578899134,
+      "learning_rate": 8.770614353482628e-07,
+      "loss": 1.7536,
+      "step": 4530
+    },
+    {
+      "epoch": 0.3055826960008086,
+      "grad_norm": 4.423861680399267,
+      "learning_rate": 8.766749833092638e-07,
+      "loss": 1.7985,
+      "step": 4535
+    },
+    {
+      "epoch": 0.30591961187291533,
+      "grad_norm": 4.2791878376307295,
+      "learning_rate": 8.76288010251145e-07,
+      "loss": 1.7224,
+      "step": 4540
+    },
+    {
+      "epoch": 0.30625652774502204,
+      "grad_norm": 4.903125873267787,
+      "learning_rate": 8.759005167091697e-07,
+      "loss": 1.7735,
+      "step": 4545
+    },
+    {
+      "epoch": 0.3065934436171288,
+      "grad_norm": 4.8237777891217375,
+      "learning_rate": 8.755125032193214e-07,
+      "loss": 1.656,
+      "step": 4550
+    },
+    {
+      "epoch": 0.3069303594892355,
+      "grad_norm": 4.823939629101453,
+      "learning_rate": 8.751239703183026e-07,
+      "loss": 1.7152,
+      "step": 4555
+    },
+    {
+      "epoch": 0.3072672753613423,
+      "grad_norm": 4.566940546243774,
+      "learning_rate": 8.747349185435348e-07,
+      "loss": 1.716,
+      "step": 4560
+    },
+    {
+      "epoch": 0.307604191233449,
+      "grad_norm": 4.732548527043846,
+      "learning_rate": 8.743453484331562e-07,
+      "loss": 1.7865,
+      "step": 4565
+    },
+    {
+      "epoch": 0.3079411071055557,
+      "grad_norm": 4.623294109058839,
+      "learning_rate": 8.73955260526023e-07,
+      "loss": 1.7017,
+      "step": 4570
+    },
+    {
+      "epoch": 0.3082780229776625,
+      "grad_norm": 4.485493004343804,
+      "learning_rate": 8.735646553617069e-07,
+      "loss": 1.8001,
+      "step": 4575
+    },
+    {
+      "epoch": 0.3086149388497692,
+      "grad_norm": 4.492019215877914,
+      "learning_rate": 8.731735334804953e-07,
+      "loss": 1.7515,
+      "step": 4580
+    },
+    {
+      "epoch": 0.308951854721876,
+      "grad_norm": 4.7049240303578275,
+      "learning_rate": 8.727818954233904e-07,
+      "loss": 1.7318,
+      "step": 4585
+    },
+    {
+      "epoch": 0.3092887705939827,
+      "grad_norm": 4.938502374647742,
+      "learning_rate": 8.723897417321084e-07,
+      "loss": 1.725,
+      "step": 4590
+    },
+    {
+      "epoch": 0.3096256864660894,
+      "grad_norm": 5.022532343934432,
+      "learning_rate": 8.719970729490788e-07,
+      "loss": 1.7443,
+      "step": 4595
+    },
+    {
+      "epoch": 0.30996260233819617,
+      "grad_norm": 4.738767482257275,
+      "learning_rate": 8.716038896174432e-07,
+      "loss": 1.755,
+      "step": 4600
+    },
+    {
+      "epoch": 0.3102995182103029,
+      "grad_norm": 4.7114300519860475,
+      "learning_rate": 8.712101922810551e-07,
+      "loss": 1.75,
+      "step": 4605
+    },
+    {
+      "epoch": 0.3106364340824096,
+      "grad_norm": 4.571041536354113,
+      "learning_rate": 8.708159814844793e-07,
+      "loss": 1.7098,
+      "step": 4610
+    },
+    {
+      "epoch": 0.31097334995451636,
+      "grad_norm": 4.407045547143283,
+      "learning_rate": 8.704212577729905e-07,
+      "loss": 1.7793,
+      "step": 4615
+    },
+    {
+      "epoch": 0.3113102658266231,
+      "grad_norm": 4.728634153355158,
+      "learning_rate": 8.700260216925728e-07,
+      "loss": 1.6889,
+      "step": 4620
+    },
+    {
+      "epoch": 0.31164718169872985,
+      "grad_norm": 4.563219061062255,
+      "learning_rate": 8.696302737899192e-07,
+      "loss": 1.7611,
+      "step": 4625
+    },
+    {
+      "epoch": 0.31198409757083656,
+      "grad_norm": 4.783988230321884,
+      "learning_rate": 8.692340146124308e-07,
+      "loss": 1.7104,
+      "step": 4630
+    },
+    {
+      "epoch": 0.3123210134429433,
+      "grad_norm": 4.847737749788932,
+      "learning_rate": 8.688372447082153e-07,
+      "loss": 1.7511,
+      "step": 4635
+    },
+    {
+      "epoch": 0.31265792931505004,
+      "grad_norm": 4.907400664885099,
+      "learning_rate": 8.684399646260876e-07,
+      "loss": 1.7158,
+      "step": 4640
+    },
+    {
+      "epoch": 0.31299484518715676,
+      "grad_norm": 4.876331767645437,
+      "learning_rate": 8.680421749155677e-07,
+      "loss": 1.6995,
+      "step": 4645
+    },
+    {
+      "epoch": 0.3133317610592635,
+      "grad_norm": 4.894058583174596,
+      "learning_rate": 8.676438761268808e-07,
+      "loss": 1.7764,
+      "step": 4650
+    },
+    {
+      "epoch": 0.31366867693137024,
+      "grad_norm": 4.672221745908909,
+      "learning_rate": 8.672450688109563e-07,
+      "loss": 1.7844,
+      "step": 4655
+    },
+    {
+      "epoch": 0.31400559280347695,
+      "grad_norm": 5.147054583280883,
+      "learning_rate": 8.668457535194267e-07,
+      "loss": 1.8186,
+      "step": 4660
+    },
+    {
+      "epoch": 0.3143425086755837,
+      "grad_norm": 4.51560485470913,
+      "learning_rate": 8.664459308046274e-07,
+      "loss": 1.7676,
+      "step": 4665
+    },
+    {
+      "epoch": 0.31467942454769043,
+      "grad_norm": 4.767998872271937,
+      "learning_rate": 8.660456012195957e-07,
+      "loss": 1.7302,
+      "step": 4670
+    },
+    {
+      "epoch": 0.3150163404197972,
+      "grad_norm": 4.765982682989915,
+      "learning_rate": 8.656447653180699e-07,
+      "loss": 1.7017,
+      "step": 4675
+    },
+    {
+      "epoch": 0.3153532562919039,
+      "grad_norm": 4.6837995867112285,
+      "learning_rate": 8.652434236544886e-07,
+      "loss": 1.7546,
+      "step": 4680
+    },
+    {
+      "epoch": 0.31569017216401063,
+      "grad_norm": 4.823408524687895,
+      "learning_rate": 8.648415767839899e-07,
+      "loss": 1.7132,
+      "step": 4685
+    },
+    {
+      "epoch": 0.3160270880361174,
+      "grad_norm": 4.730875315239921,
+      "learning_rate": 8.644392252624108e-07,
+      "loss": 1.6749,
+      "step": 4690
+    },
+    {
+      "epoch": 0.3163640039082241,
+      "grad_norm": 4.282912686826423,
+      "learning_rate": 8.640363696462869e-07,
+      "loss": 1.7665,
+      "step": 4695
+    },
+    {
+      "epoch": 0.3167009197803308,
+      "grad_norm": 4.436758751754393,
+      "learning_rate": 8.636330104928499e-07,
+      "loss": 1.7438,
+      "step": 4700
+    },
+    {
+      "epoch": 0.3170378356524376,
+      "grad_norm": 4.964885054397722,
+      "learning_rate": 8.632291483600289e-07,
+      "loss": 1.7099,
+      "step": 4705
+    },
+    {
+      "epoch": 0.3173747515245443,
+      "grad_norm": 4.675853565648876,
+      "learning_rate": 8.628247838064485e-07,
+      "loss": 1.7111,
+      "step": 4710
+    },
+    {
+      "epoch": 0.3177116673966511,
+      "grad_norm": 4.6934367206975365,
+      "learning_rate": 8.624199173914279e-07,
+      "loss": 1.7971,
+      "step": 4715
+    },
+    {
+      "epoch": 0.3180485832687578,
+      "grad_norm": 4.512080853138636,
+      "learning_rate": 8.620145496749811e-07,
+      "loss": 1.7319,
+      "step": 4720
+    },
+    {
+      "epoch": 0.3183854991408645,
+      "grad_norm": 5.079728958055047,
+      "learning_rate": 8.616086812178151e-07,
+      "loss": 1.737,
+      "step": 4725
+    },
+    {
+      "epoch": 0.31872241501297127,
+      "grad_norm": 4.640827385755181,
+      "learning_rate": 8.612023125813296e-07,
+      "loss": 1.7648,
+      "step": 4730
+    },
+    {
+      "epoch": 0.319059330885078,
+      "grad_norm": 4.276778816417824,
+      "learning_rate": 8.607954443276162e-07,
+      "loss": 1.7802,
+      "step": 4735
+    },
+    {
+      "epoch": 0.31939624675718475,
+      "grad_norm": 4.481763322120392,
+      "learning_rate": 8.603880770194574e-07,
+      "loss": 1.7305,
+      "step": 4740
+    },
+    {
+      "epoch": 0.31973316262929147,
+      "grad_norm": 4.615513878941768,
+      "learning_rate": 8.59980211220326e-07,
+      "loss": 1.6987,
+      "step": 4745
+    },
+    {
+      "epoch": 0.3200700785013982,
+      "grad_norm": 4.7455902906252465,
+      "learning_rate": 8.595718474943849e-07,
+      "loss": 1.7027,
+      "step": 4750
+    },
+    {
+      "epoch": 0.32040699437350495,
+      "grad_norm": 4.68398292422727,
+      "learning_rate": 8.591629864064851e-07,
+      "loss": 1.7455,
+      "step": 4755
+    },
+    {
+      "epoch": 0.32074391024561166,
+      "grad_norm": 4.9222045458553225,
+      "learning_rate": 8.587536285221655e-07,
+      "loss": 1.8028,
+      "step": 4760
+    },
+    {
+      "epoch": 0.32108082611771843,
+      "grad_norm": 4.502320783272846,
+      "learning_rate": 8.583437744076527e-07,
+      "loss": 1.7402,
+      "step": 4765
+    },
+    {
+      "epoch": 0.32141774198982515,
+      "grad_norm": 5.144779588736784,
+      "learning_rate": 8.579334246298592e-07,
+      "loss": 1.7628,
+      "step": 4770
+    },
+    {
+      "epoch": 0.32175465786193186,
+      "grad_norm": 4.504761407192705,
+      "learning_rate": 8.575225797563834e-07,
+      "loss": 1.7915,
+      "step": 4775
+    },
+    {
+      "epoch": 0.3220915737340386,
+      "grad_norm": 5.735894915412811,
+      "learning_rate": 8.571112403555083e-07,
+      "loss": 1.7218,
+      "step": 4780
+    },
+    {
+      "epoch": 0.32242848960614534,
+      "grad_norm": 4.7323785258509785,
+      "learning_rate": 8.566994069962012e-07,
+      "loss": 1.7429,
+      "step": 4785
+    },
+    {
+      "epoch": 0.32276540547825205,
+      "grad_norm": 4.487702277209751,
+      "learning_rate": 8.562870802481126e-07,
+      "loss": 1.7489,
+      "step": 4790
+    },
+    {
+      "epoch": 0.3231023213503588,
+      "grad_norm": 4.928490583624491,
+      "learning_rate": 8.55874260681575e-07,
+      "loss": 1.7347,
+      "step": 4795
+    },
+    {
+      "epoch": 0.32343923722246554,
+      "grad_norm": 4.877761123390889,
+      "learning_rate": 8.554609488676032e-07,
+      "loss": 1.7209,
+      "step": 4800
+    },
+    {
+      "epoch": 0.3237761530945723,
+      "grad_norm": 4.764011730790988,
+      "learning_rate": 8.550471453778925e-07,
+      "loss": 1.717,
+      "step": 4805
+    },
+    {
+      "epoch": 0.324113068966679,
+      "grad_norm": 4.799601103958323,
+      "learning_rate": 8.546328507848184e-07,
+      "loss": 1.7163,
+      "step": 4810
+    },
+    {
+      "epoch": 0.32444998483878573,
+      "grad_norm": 4.504239537326746,
+      "learning_rate": 8.542180656614358e-07,
+      "loss": 1.6741,
+      "step": 4815
+    },
+    {
+      "epoch": 0.3247869007108925,
+      "grad_norm": 4.770834350267517,
+      "learning_rate": 8.538027905814778e-07,
+      "loss": 1.6956,
+      "step": 4820
+    },
+    {
+      "epoch": 0.3251238165829992,
+      "grad_norm": 7.096648611390167,
+      "learning_rate": 8.533870261193556e-07,
+      "loss": 1.6588,
+      "step": 4825
+    },
+    {
+      "epoch": 0.325460732455106,
+      "grad_norm": 4.864466677890027,
+      "learning_rate": 8.529707728501571e-07,
+      "loss": 1.6915,
+      "step": 4830
+    },
+    {
+      "epoch": 0.3257976483272127,
+      "grad_norm": 4.744648508382125,
+      "learning_rate": 8.525540313496462e-07,
+      "loss": 1.7164,
+      "step": 4835
+    },
+    {
+      "epoch": 0.3261345641993194,
+      "grad_norm": 4.529035511663301,
+      "learning_rate": 8.521368021942623e-07,
+      "loss": 1.7921,
+      "step": 4840
+    },
+    {
+      "epoch": 0.3264714800714262,
+      "grad_norm": 4.9316185678104025,
+      "learning_rate": 8.517190859611195e-07,
+      "loss": 1.6849,
+      "step": 4845
+    },
+    {
+      "epoch": 0.3268083959435329,
+      "grad_norm": 5.14026374289455,
+      "learning_rate": 8.513008832280053e-07,
+      "loss": 1.8128,
+      "step": 4850
+    },
+    {
+      "epoch": 0.32714531181563966,
+      "grad_norm": 4.985258213485664,
+      "learning_rate": 8.508821945733802e-07,
+      "loss": 1.7436,
+      "step": 4855
+    },
+    {
+      "epoch": 0.3274822276877464,
+      "grad_norm": 4.869434253185966,
+      "learning_rate": 8.504630205763768e-07,
+      "loss": 1.7338,
+      "step": 4860
+    },
+    {
+      "epoch": 0.3278191435598531,
+      "grad_norm": 4.569578875489374,
+      "learning_rate": 8.500433618167992e-07,
+      "loss": 1.7813,
+      "step": 4865
+    },
+    {
+      "epoch": 0.32815605943195986,
+      "grad_norm": 4.72390686940452,
+      "learning_rate": 8.496232188751222e-07,
+      "loss": 1.7272,
+      "step": 4870
+    },
+    {
+      "epoch": 0.32849297530406657,
+      "grad_norm": 4.337728617203676,
+      "learning_rate": 8.492025923324897e-07,
+      "loss": 1.711,
+      "step": 4875
+    },
+    {
+      "epoch": 0.3288298911761733,
+      "grad_norm": 4.665189493399873,
+      "learning_rate": 8.487814827707152e-07,
+      "loss": 1.7424,
+      "step": 4880
+    },
+    {
+      "epoch": 0.32916680704828005,
+      "grad_norm": 4.764469314965838,
+      "learning_rate": 8.483598907722795e-07,
+      "loss": 1.7619,
+      "step": 4885
+    },
+    {
+      "epoch": 0.32950372292038677,
+      "grad_norm": 4.415938675590576,
+      "learning_rate": 8.479378169203317e-07,
+      "loss": 1.6712,
+      "step": 4890
+    },
+    {
+      "epoch": 0.32984063879249353,
+      "grad_norm": 4.798044411903519,
+      "learning_rate": 8.475152617986869e-07,
+      "loss": 1.698,
+      "step": 4895
+    },
+    {
+      "epoch": 0.33017755466460025,
+      "grad_norm": 4.675109222425787,
+      "learning_rate": 8.470922259918254e-07,
+      "loss": 1.6894,
+      "step": 4900
+    },
+    {
+      "epoch": 0.33051447053670696,
+      "grad_norm": 4.481553305526358,
+      "learning_rate": 8.466687100848935e-07,
+      "loss": 1.7655,
+      "step": 4905
+    },
+    {
+      "epoch": 0.33085138640881373,
+      "grad_norm": 4.93030004046697,
+      "learning_rate": 8.462447146637006e-07,
+      "loss": 1.7206,
+      "step": 4910
+    },
+    {
+      "epoch": 0.33118830228092044,
+      "grad_norm": 6.135512951857502,
+      "learning_rate": 8.458202403147199e-07,
+      "loss": 1.725,
+      "step": 4915
+    },
+    {
+      "epoch": 0.3315252181530272,
+      "grad_norm": 4.855174964493657,
+      "learning_rate": 8.453952876250867e-07,
+      "loss": 1.6884,
+      "step": 4920
+    },
+    {
+      "epoch": 0.3318621340251339,
+      "grad_norm": 4.830245687846346,
+      "learning_rate": 8.449698571825984e-07,
+      "loss": 1.777,
+      "step": 4925
+    },
+    {
+      "epoch": 0.33219904989724064,
+      "grad_norm": 4.81643428731798,
+      "learning_rate": 8.445439495757127e-07,
+      "loss": 1.7166,
+      "step": 4930
+    },
+    {
+      "epoch": 0.3325359657693474,
+      "grad_norm": 4.766650179919701,
+      "learning_rate": 8.44117565393548e-07,
+      "loss": 1.6928,
+      "step": 4935
+    },
+    {
+      "epoch": 0.3328728816414541,
+      "grad_norm": 4.504805907009001,
+      "learning_rate": 8.436907052258808e-07,
+      "loss": 1.7757,
+      "step": 4940
+    },
+    {
+      "epoch": 0.3332097975135609,
+      "grad_norm": 4.871702033132882,
+      "learning_rate": 8.432633696631473e-07,
+      "loss": 1.8047,
+      "step": 4945
+    },
+    {
+      "epoch": 0.3335467133856676,
+      "grad_norm": 4.360999298002896,
+      "learning_rate": 8.428355592964405e-07,
+      "loss": 1.7514,
+      "step": 4950
+    },
+    {
+      "epoch": 0.3338836292577743,
+      "grad_norm": 4.40738849972985,
+      "learning_rate": 8.424072747175102e-07,
+      "loss": 1.7638,
+      "step": 4955
+    },
+    {
+      "epoch": 0.3342205451298811,
+      "grad_norm": 4.402646732849628,
+      "learning_rate": 8.419785165187621e-07,
+      "loss": 1.6341,
+      "step": 4960
+    },
+    {
+      "epoch": 0.3345574610019878,
+      "grad_norm": 4.641381273642867,
+      "learning_rate": 8.415492852932573e-07,
+      "loss": 1.6812,
+      "step": 4965
+    },
+    {
+      "epoch": 0.3348943768740945,
+      "grad_norm": 4.92094782072377,
+      "learning_rate": 8.41119581634711e-07,
+      "loss": 1.8236,
+      "step": 4970
+    },
+    {
+      "epoch": 0.3352312927462013,
+      "grad_norm": 4.592892836304687,
+      "learning_rate": 8.406894061374918e-07,
+      "loss": 1.8073,
+      "step": 4975
+    },
+    {
+      "epoch": 0.335568208618308,
+      "grad_norm": 4.612862661847665,
+      "learning_rate": 8.402587593966213e-07,
+      "loss": 1.6964,
+      "step": 4980
+    },
+    {
+      "epoch": 0.33590512449041476,
+      "grad_norm": 4.537376816334139,
+      "learning_rate": 8.398276420077726e-07,
+      "loss": 1.7642,
+      "step": 4985
+    },
+    {
+      "epoch": 0.3362420403625215,
+      "grad_norm": 4.712056499426122,
+      "learning_rate": 8.393960545672698e-07,
+      "loss": 1.7134,
+      "step": 4990
+    },
+    {
+      "epoch": 0.3365789562346282,
+      "grad_norm": 4.564828506764209,
+      "learning_rate": 8.389639976720873e-07,
+      "loss": 1.7204,
+      "step": 4995
+    },
+    {
+      "epoch": 0.33691587210673496,
+      "grad_norm": 4.914182336061254,
+      "learning_rate": 8.385314719198487e-07,
+      "loss": 1.7386,
+      "step": 5000
+    },
+    {
+      "epoch": 0.3372527879788417,
+      "grad_norm": 4.484842012048959,
+      "learning_rate": 8.380984779088264e-07,
+      "loss": 1.7416,
+      "step": 5005
+    },
+    {
+      "epoch": 0.33758970385094844,
+      "grad_norm": 5.139741044975532,
+      "learning_rate": 8.376650162379404e-07,
+      "loss": 1.7514,
+      "step": 5010
+    },
+    {
+      "epoch": 0.33792661972305515,
+      "grad_norm": 4.47276838902141,
+      "learning_rate": 8.372310875067572e-07,
+      "loss": 1.7606,
+      "step": 5015
+    },
+    {
+      "epoch": 0.33826353559516187,
+      "grad_norm": 4.531427897815532,
+      "learning_rate": 8.367966923154899e-07,
+      "loss": 1.6455,
+      "step": 5020
+    },
+    {
+      "epoch": 0.33860045146726864,
+      "grad_norm": 4.661435301070422,
+      "learning_rate": 8.363618312649967e-07,
+      "loss": 1.7284,
+      "step": 5025
+    },
+    {
+      "epoch": 0.33893736733937535,
+      "grad_norm": 4.678787625067841,
+      "learning_rate": 8.359265049567796e-07,
+      "loss": 1.719,
+      "step": 5030
+    },
+    {
+      "epoch": 0.3392742832114821,
+      "grad_norm": 4.110099175210853,
+      "learning_rate": 8.35490713992985e-07,
+      "loss": 1.7321,
+      "step": 5035
+    },
+    {
+      "epoch": 0.33961119908358883,
+      "grad_norm": 4.7168890532127135,
+      "learning_rate": 8.350544589764015e-07,
+      "loss": 1.7626,
+      "step": 5040
+    },
+    {
+      "epoch": 0.33994811495569555,
+      "grad_norm": 4.5142169100739205,
+      "learning_rate": 8.346177405104595e-07,
+      "loss": 1.7402,
+      "step": 5045
+    },
+    {
+      "epoch": 0.3402850308278023,
+      "grad_norm": 4.844467413578708,
+      "learning_rate": 8.341805591992308e-07,
+      "loss": 1.7593,
+      "step": 5050
+    },
+    {
+      "epoch": 0.34062194669990903,
+      "grad_norm": 4.517706615153855,
+      "learning_rate": 8.337429156474272e-07,
+      "loss": 1.6679,
+      "step": 5055
+    },
+    {
+      "epoch": 0.34095886257201574,
+      "grad_norm": 4.377813395323879,
+      "learning_rate": 8.333048104603999e-07,
+      "loss": 1.6646,
+      "step": 5060
+    },
+    {
+      "epoch": 0.3412957784441225,
+      "grad_norm": 4.482064436846207,
+      "learning_rate": 8.328662442441388e-07,
+      "loss": 1.6389,
+      "step": 5065
+    },
+    {
+      "epoch": 0.3416326943162292,
+      "grad_norm": 4.400411146015806,
+      "learning_rate": 8.32427217605271e-07,
+      "loss": 1.7307,
+      "step": 5070
+    },
+    {
+      "epoch": 0.341969610188336,
+      "grad_norm": 4.528585841227934,
+      "learning_rate": 8.319877311510612e-07,
+      "loss": 1.7653,
+      "step": 5075
+    },
+    {
+      "epoch": 0.3423065260604427,
+      "grad_norm": 4.697407386040308,
+      "learning_rate": 8.315477854894095e-07,
+      "loss": 1.7074,
+      "step": 5080
+    },
+    {
+      "epoch": 0.3426434419325494,
+      "grad_norm": 4.59364070633189,
+      "learning_rate": 8.311073812288513e-07,
+      "loss": 1.7516,
+      "step": 5085
+    },
+    {
+      "epoch": 0.3429803578046562,
+      "grad_norm": 4.618861246582405,
+      "learning_rate": 8.306665189785567e-07,
+      "loss": 1.723,
+      "step": 5090
+    },
+    {
+      "epoch": 0.3433172736767629,
+      "grad_norm": 4.3958818021585655,
+      "learning_rate": 8.302251993483289e-07,
+      "loss": 1.7322,
+      "step": 5095
+    },
+    {
+      "epoch": 0.34365418954886967,
+      "grad_norm": 5.09198950014467,
+      "learning_rate": 8.297834229486039e-07,
+      "loss": 1.7217,
+      "step": 5100
+    },
+    {
+      "epoch": 0.3439911054209764,
+      "grad_norm": 4.498521150001597,
+      "learning_rate": 8.293411903904496e-07,
+      "loss": 1.6854,
+      "step": 5105
+    },
+    {
+      "epoch": 0.3443280212930831,
+      "grad_norm": 4.456639889055303,
+      "learning_rate": 8.288985022855645e-07,
+      "loss": 1.7743,
+      "step": 5110
+    },
+    {
+      "epoch": 0.34466493716518987,
+      "grad_norm": 4.612852747434726,
+      "learning_rate": 8.284553592462778e-07,
+      "loss": 1.7581,
+      "step": 5115
+    },
+    {
+      "epoch": 0.3450018530372966,
+      "grad_norm": 4.313561479912604,
+      "learning_rate": 8.280117618855475e-07,
+      "loss": 1.75,
+      "step": 5120
+    },
+    {
+      "epoch": 0.34533876890940335,
+      "grad_norm": 4.774021466238921,
+      "learning_rate": 8.2756771081696e-07,
+      "loss": 1.7794,
+      "step": 5125
+    },
+    {
+      "epoch": 0.34567568478151006,
+      "grad_norm": 4.532488619269362,
+      "learning_rate": 8.271232066547296e-07,
+      "loss": 1.7032,
+      "step": 5130
+    },
+    {
+      "epoch": 0.3460126006536168,
+      "grad_norm": 4.210061067829154,
+      "learning_rate": 8.266782500136971e-07,
+      "loss": 1.7748,
+      "step": 5135
+    },
+    {
+      "epoch": 0.34634951652572354,
+      "grad_norm": 4.882640898960227,
+      "learning_rate": 8.262328415093293e-07,
+      "loss": 1.7544,
+      "step": 5140
+    },
+    {
+      "epoch": 0.34668643239783026,
+      "grad_norm": 4.676380957510316,
+      "learning_rate": 8.257869817577179e-07,
+      "loss": 1.7428,
+      "step": 5145
+    },
+    {
+      "epoch": 0.34702334826993697,
+      "grad_norm": 4.494144685289505,
+      "learning_rate": 8.253406713755786e-07,
+      "loss": 1.7142,
+      "step": 5150
+    },
+    {
+      "epoch": 0.34736026414204374,
+      "grad_norm": 4.81974015891061,
+      "learning_rate": 8.24893910980251e-07,
+      "loss": 1.6675,
+      "step": 5155
+    },
+    {
+      "epoch": 0.34769718001415045,
+      "grad_norm": 4.66644833407102,
+      "learning_rate": 8.244467011896965e-07,
+      "loss": 1.66,
+      "step": 5160
+    },
+    {
+      "epoch": 0.3480340958862572,
+      "grad_norm": 4.265031234744358,
+      "learning_rate": 8.239990426224986e-07,
+      "loss": 1.7472,
+      "step": 5165
+    },
+    {
+      "epoch": 0.34837101175836394,
+      "grad_norm": 4.46755983851988,
+      "learning_rate": 8.235509358978611e-07,
+      "loss": 1.7599,
+      "step": 5170
+    },
+    {
+      "epoch": 0.34870792763047065,
+      "grad_norm": 4.846201035405282,
+      "learning_rate": 8.231023816356081e-07,
+      "loss": 1.6988,
+      "step": 5175
+    },
+    {
+      "epoch": 0.3490448435025774,
+      "grad_norm": 5.068632343934939,
+      "learning_rate": 8.226533804561826e-07,
+      "loss": 1.7098,
+      "step": 5180
+    },
+    {
+      "epoch": 0.34938175937468413,
+      "grad_norm": 4.780165677711516,
+      "learning_rate": 8.222039329806456e-07,
+      "loss": 1.7011,
+      "step": 5185
+    },
+    {
+      "epoch": 0.3497186752467909,
+      "grad_norm": 4.621335846200827,
+      "learning_rate": 8.217540398306757e-07,
+      "loss": 1.6735,
+      "step": 5190
+    },
+    {
+      "epoch": 0.3500555911188976,
+      "grad_norm": 4.513458917850163,
+      "learning_rate": 8.213037016285679e-07,
+      "loss": 1.7434,
+      "step": 5195
+    },
+    {
+      "epoch": 0.3503925069910043,
+      "grad_norm": 4.9286192435302185,
+      "learning_rate": 8.208529189972325e-07,
+      "loss": 1.7284,
+      "step": 5200
+    },
+    {
+      "epoch": 0.3507294228631111,
+      "grad_norm": 4.934007063420536,
+      "learning_rate": 8.204016925601951e-07,
+      "loss": 1.7539,
+      "step": 5205
+    },
+    {
+      "epoch": 0.3510663387352178,
+      "grad_norm": 4.855237741263649,
+      "learning_rate": 8.199500229415945e-07,
+      "loss": 1.7085,
+      "step": 5210
+    },
+    {
+      "epoch": 0.3514032546073246,
+      "grad_norm": 4.615249141184611,
+      "learning_rate": 8.19497910766183e-07,
+      "loss": 1.7298,
+      "step": 5215
+    },
+    {
+      "epoch": 0.3517401704794313,
+      "grad_norm": 4.495205667452235,
+      "learning_rate": 8.19045356659325e-07,
+      "loss": 1.7362,
+      "step": 5220
+    },
+    {
+      "epoch": 0.352077086351538,
+      "grad_norm": 4.829057942023865,
+      "learning_rate": 8.185923612469958e-07,
+      "loss": 1.6869,
+      "step": 5225
+    },
+    {
+      "epoch": 0.3524140022236448,
+      "grad_norm": 4.725434418643222,
+      "learning_rate": 8.181389251557817e-07,
+      "loss": 1.7194,
+      "step": 5230
+    },
+    {
+      "epoch": 0.3527509180957515,
+      "grad_norm": 4.58503870736107,
+      "learning_rate": 8.176850490128782e-07,
+      "loss": 1.7013,
+      "step": 5235
+    },
+    {
+      "epoch": 0.3530878339678582,
+      "grad_norm": 4.782838738330386,
+      "learning_rate": 8.172307334460892e-07,
+      "loss": 1.7747,
+      "step": 5240
+    },
+    {
+      "epoch": 0.35342474983996497,
+      "grad_norm": 5.451500093580164,
+      "learning_rate": 8.167759790838273e-07,
+      "loss": 1.7207,
+      "step": 5245
+    },
+    {
+      "epoch": 0.3537616657120717,
+      "grad_norm": 4.695952472643797,
+      "learning_rate": 8.163207865551111e-07,
+      "loss": 1.6691,
+      "step": 5250
+    },
+    {
+      "epoch": 0.35409858158417845,
+      "grad_norm": 4.953931315209139,
+      "learning_rate": 8.158651564895657e-07,
+      "loss": 1.68,
+      "step": 5255
+    },
+    {
+      "epoch": 0.35443549745628516,
+      "grad_norm": 4.731523917919795,
+      "learning_rate": 8.154090895174215e-07,
+      "loss": 1.7114,
+      "step": 5260
+    },
+    {
+      "epoch": 0.3547724133283919,
+      "grad_norm": 4.665813772982818,
+      "learning_rate": 8.149525862695131e-07,
+      "loss": 1.7682,
+      "step": 5265
+    },
+    {
+      "epoch": 0.35510932920049865,
+      "grad_norm": 4.2650023278644005,
+      "learning_rate": 8.144956473772784e-07,
+      "loss": 1.7071,
+      "step": 5270
+    },
+    {
+      "epoch": 0.35544624507260536,
+      "grad_norm": 4.583863138977418,
+      "learning_rate": 8.140382734727581e-07,
+      "loss": 1.7393,
+      "step": 5275
+    },
+    {
+      "epoch": 0.35578316094471213,
+      "grad_norm": 4.401047996458969,
+      "learning_rate": 8.135804651885946e-07,
+      "loss": 1.7556,
+      "step": 5280
+    },
+    {
+      "epoch": 0.35612007681681884,
+      "grad_norm": 4.948765436864181,
+      "learning_rate": 8.131222231580313e-07,
+      "loss": 1.6948,
+      "step": 5285
+    },
+    {
+      "epoch": 0.35645699268892556,
+      "grad_norm": 4.647264244397854,
+      "learning_rate": 8.126635480149107e-07,
+      "loss": 1.7475,
+      "step": 5290
+    },
+    {
+      "epoch": 0.3567939085610323,
+      "grad_norm": 4.6376077560112545,
+      "learning_rate": 8.122044403936759e-07,
+      "loss": 1.7301,
+      "step": 5295
+    },
+    {
+      "epoch": 0.35713082443313904,
+      "grad_norm": 4.792132621017928,
+      "learning_rate": 8.117449009293668e-07,
+      "loss": 1.7136,
+      "step": 5300
+    },
+    {
+      "epoch": 0.3574677403052458,
+      "grad_norm": 4.6050017365222535,
+      "learning_rate": 8.112849302576212e-07,
+      "loss": 1.7232,
+      "step": 5305
+    },
+    {
+      "epoch": 0.3578046561773525,
+      "grad_norm": 4.333389497999406,
+      "learning_rate": 8.108245290146735e-07,
+      "loss": 1.6384,
+      "step": 5310
+    },
+    {
+      "epoch": 0.35814157204945923,
+      "grad_norm": 4.78399376828893,
+      "learning_rate": 8.103636978373534e-07,
+      "loss": 1.6869,
+      "step": 5315
+    },
+    {
+      "epoch": 0.358478487921566,
+      "grad_norm": 4.895830217254766,
+      "learning_rate": 8.099024373630854e-07,
+      "loss": 1.67,
+      "step": 5320
+    },
+    {
+      "epoch": 0.3588154037936727,
+      "grad_norm": 4.518244253442206,
+      "learning_rate": 8.094407482298877e-07,
+      "loss": 1.7102,
+      "step": 5325
+    },
+    {
+      "epoch": 0.35915231966577943,
+      "grad_norm": 4.402545137279259,
+      "learning_rate": 8.089786310763716e-07,
+      "loss": 1.7257,
+      "step": 5330
+    },
+    {
+      "epoch": 0.3594892355378862,
+      "grad_norm": 4.404324471348795,
+      "learning_rate": 8.085160865417403e-07,
+      "loss": 1.6728,
+      "step": 5335
+    },
+    {
+      "epoch": 0.3598261514099929,
+      "grad_norm": 4.619273776898724,
+      "learning_rate": 8.080531152657884e-07,
+      "loss": 1.6772,
+      "step": 5340
+    },
+    {
+      "epoch": 0.3601630672820997,
+      "grad_norm": 5.085968829325327,
+      "learning_rate": 8.075897178889002e-07,
+      "loss": 1.7058,
+      "step": 5345
+    },
+    {
+      "epoch": 0.3604999831542064,
+      "grad_norm": 4.681702471986685,
+      "learning_rate": 8.071258950520501e-07,
+      "loss": 1.7616,
+      "step": 5350
+    },
+    {
+      "epoch": 0.3608368990263131,
+      "grad_norm": 4.542808993087876,
+      "learning_rate": 8.066616473968005e-07,
+      "loss": 1.7601,
+      "step": 5355
+    },
+    {
+      "epoch": 0.3611738148984199,
+      "grad_norm": 4.7748260170083245,
+      "learning_rate": 8.061969755653013e-07,
+      "loss": 1.7815,
+      "step": 5360
+    },
+    {
+      "epoch": 0.3615107307705266,
+      "grad_norm": 4.829390136039957,
+      "learning_rate": 8.0573188020029e-07,
+      "loss": 1.7522,
+      "step": 5365
+    },
+    {
+      "epoch": 0.36184764664263336,
+      "grad_norm": 4.538492090991936,
+      "learning_rate": 8.052663619450889e-07,
+      "loss": 1.6557,
+      "step": 5370
+    },
+    {
+      "epoch": 0.36218456251474007,
+      "grad_norm": 4.7861584328042674,
+      "learning_rate": 8.048004214436058e-07,
+      "loss": 1.7294,
+      "step": 5375
+    },
+    {
+      "epoch": 0.3625214783868468,
+      "grad_norm": 4.833761542360912,
+      "learning_rate": 8.043340593403325e-07,
+      "loss": 1.8014,
+      "step": 5380
+    },
+    {
+      "epoch": 0.36285839425895355,
+      "grad_norm": 4.986294508483786,
+      "learning_rate": 8.038672762803437e-07,
+      "loss": 1.7386,
+      "step": 5385
+    },
+    {
+      "epoch": 0.36319531013106027,
+      "grad_norm": 4.369718973973582,
+      "learning_rate": 8.034000729092967e-07,
+      "loss": 1.7509,
+      "step": 5390
+    },
+    {
+      "epoch": 0.36353222600316704,
+      "grad_norm": 5.256644811129503,
+      "learning_rate": 8.029324498734299e-07,
+      "loss": 1.6701,
+      "step": 5395
+    },
+    {
+      "epoch": 0.36386914187527375,
+      "grad_norm": 4.6274278308145105,
+      "learning_rate": 8.024644078195625e-07,
+      "loss": 1.7283,
+      "step": 5400
+    },
+    {
+      "epoch": 0.36420605774738046,
+      "grad_norm": 4.652898587715458,
+      "learning_rate": 8.01995947395093e-07,
+      "loss": 1.7084,
+      "step": 5405
+    },
+    {
+      "epoch": 0.36454297361948723,
+      "grad_norm": 4.469834686326115,
+      "learning_rate": 8.015270692479988e-07,
+      "loss": 1.7923,
+      "step": 5410
+    },
+    {
+      "epoch": 0.36487988949159395,
+      "grad_norm": 4.904048624023098,
+      "learning_rate": 8.010577740268347e-07,
+      "loss": 1.7368,
+      "step": 5415
+    },
+    {
+      "epoch": 0.36521680536370066,
+      "grad_norm": 4.77922639432497,
+      "learning_rate": 8.005880623807331e-07,
+      "loss": 1.636,
+      "step": 5420
+    },
+    {
+      "epoch": 0.3655537212358074,
+      "grad_norm": 5.2469243350698065,
+      "learning_rate": 8.001179349594016e-07,
+      "loss": 1.78,
+      "step": 5425
+    },
+    {
+      "epoch": 0.36589063710791414,
+      "grad_norm": 4.696982829348919,
+      "learning_rate": 7.996473924131236e-07,
+      "loss": 1.7995,
+      "step": 5430
+    },
+    {
+      "epoch": 0.3662275529800209,
+      "grad_norm": 4.754900374319797,
+      "learning_rate": 7.991764353927562e-07,
+      "loss": 1.7253,
+      "step": 5435
+    },
+    {
+      "epoch": 0.3665644688521276,
+      "grad_norm": 4.766043957422087,
+      "learning_rate": 7.987050645497302e-07,
+      "loss": 1.7251,
+      "step": 5440
+    },
+    {
+      "epoch": 0.36690138472423434,
+      "grad_norm": 5.495055814974811,
+      "learning_rate": 7.982332805360486e-07,
+      "loss": 1.721,
+      "step": 5445
+    },
+    {
+      "epoch": 0.3672383005963411,
+      "grad_norm": 5.185117084314556,
+      "learning_rate": 7.977610840042856e-07,
+      "loss": 1.7162,
+      "step": 5450
+    },
+    {
+      "epoch": 0.3675752164684478,
+      "grad_norm": 4.326014002095558,
+      "learning_rate": 7.972884756075867e-07,
+      "loss": 1.8196,
+      "step": 5455
+    },
+    {
+      "epoch": 0.3679121323405546,
+      "grad_norm": 4.957005167673788,
+      "learning_rate": 7.968154559996665e-07,
+      "loss": 1.7502,
+      "step": 5460
+    },
+    {
+      "epoch": 0.3682490482126613,
+      "grad_norm": 4.551368177373224,
+      "learning_rate": 7.963420258348086e-07,
+      "loss": 1.6959,
+      "step": 5465
+    },
+    {
+      "epoch": 0.368585964084768,
+      "grad_norm": 4.234515423236089,
+      "learning_rate": 7.958681857678645e-07,
+      "loss": 1.7148,
+      "step": 5470
+    },
+    {
+      "epoch": 0.3689228799568748,
+      "grad_norm": 4.230681426026155,
+      "learning_rate": 7.953939364542523e-07,
+      "loss": 1.6616,
+      "step": 5475
+    },
+    {
+      "epoch": 0.3692597958289815,
+      "grad_norm": 4.696843792805101,
+      "learning_rate": 7.949192785499573e-07,
+      "loss": 1.6891,
+      "step": 5480
+    },
+    {
+      "epoch": 0.36959671170108827,
+      "grad_norm": 4.732075643282494,
+      "learning_rate": 7.944442127115285e-07,
+      "loss": 1.6966,
+      "step": 5485
+    },
+    {
+      "epoch": 0.369933627573195,
+      "grad_norm": 4.746063635976779,
+      "learning_rate": 7.939687395960802e-07,
+      "loss": 1.6941,
+      "step": 5490
+    },
+    {
+      "epoch": 0.3702705434453017,
+      "grad_norm": 4.850456587535771,
+      "learning_rate": 7.934928598612895e-07,
+      "loss": 1.6475,
+      "step": 5495
+    },
+    {
+      "epoch": 0.37060745931740846,
+      "grad_norm": 4.415645353505913,
+      "learning_rate": 7.930165741653964e-07,
+      "loss": 1.7229,
+      "step": 5500
+    },
+    {
+      "epoch": 0.3709443751895152,
+      "grad_norm": 4.669235129577563,
+      "learning_rate": 7.925398831672018e-07,
+      "loss": 1.6973,
+      "step": 5505
+    },
+    {
+      "epoch": 0.3712812910616219,
+      "grad_norm": 4.404915667914949,
+      "learning_rate": 7.920627875260679e-07,
+      "loss": 1.8221,
+      "step": 5510
+    },
+    {
+      "epoch": 0.37161820693372866,
+      "grad_norm": 4.560927372059081,
+      "learning_rate": 7.91585287901916e-07,
+      "loss": 1.7294,
+      "step": 5515
+    },
+    {
+      "epoch": 0.37195512280583537,
+      "grad_norm": 4.552193802277202,
+      "learning_rate": 7.911073849552267e-07,
+      "loss": 1.7327,
+      "step": 5520
+    },
+    {
+      "epoch": 0.37229203867794214,
+      "grad_norm": 4.6558433797374255,
+      "learning_rate": 7.906290793470382e-07,
+      "loss": 1.7607,
+      "step": 5525
+    },
+    {
+      "epoch": 0.37262895455004885,
+      "grad_norm": 4.894533129056544,
+      "learning_rate": 7.901503717389458e-07,
+      "loss": 1.7314,
+      "step": 5530
+    },
+    {
+      "epoch": 0.37296587042215557,
+      "grad_norm": 4.8296195430988975,
+      "learning_rate": 7.896712627931004e-07,
+      "loss": 1.7676,
+      "step": 5535
+    },
+    {
+      "epoch": 0.37330278629426233,
+      "grad_norm": 4.68120650909417,
+      "learning_rate": 7.891917531722087e-07,
+      "loss": 1.728,
+      "step": 5540
+    },
+    {
+      "epoch": 0.37363970216636905,
+      "grad_norm": 4.377869202077806,
+      "learning_rate": 7.887118435395314e-07,
+      "loss": 1.7314,
+      "step": 5545
+    },
+    {
+      "epoch": 0.3739766180384758,
+      "grad_norm": 4.522942860706774,
+      "learning_rate": 7.882315345588823e-07,
+      "loss": 1.722,
+      "step": 5550
+    },
+    {
+      "epoch": 0.37431353391058253,
+      "grad_norm": 4.296050621560291,
+      "learning_rate": 7.877508268946275e-07,
+      "loss": 1.6951,
+      "step": 5555
+    },
+    {
+      "epoch": 0.37465044978268924,
+      "grad_norm": 4.295345819522486,
+      "learning_rate": 7.87269721211685e-07,
+      "loss": 1.7105,
+      "step": 5560
+    },
+    {
+      "epoch": 0.374987365654796,
+      "grad_norm": 4.613573892628392,
+      "learning_rate": 7.86788218175523e-07,
+      "loss": 1.7668,
+      "step": 5565
+    },
+    {
+      "epoch": 0.3753242815269027,
+      "grad_norm": 4.4369340983936665,
+      "learning_rate": 7.863063184521595e-07,
+      "loss": 1.6742,
+      "step": 5570
+    },
+    {
+      "epoch": 0.3756611973990095,
+      "grad_norm": 4.378969798272695,
+      "learning_rate": 7.858240227081611e-07,
+      "loss": 1.6726,
+      "step": 5575
+    },
+    {
+      "epoch": 0.3759981132711162,
+      "grad_norm": 4.696000087330842,
+      "learning_rate": 7.85341331610642e-07,
+      "loss": 1.7498,
+      "step": 5580
+    },
+    {
+      "epoch": 0.3763350291432229,
+      "grad_norm": 4.399452146909051,
+      "learning_rate": 7.848582458272637e-07,
+      "loss": 1.7782,
+      "step": 5585
+    },
+    {
+      "epoch": 0.3766719450153297,
+      "grad_norm": 5.4100168587899455,
+      "learning_rate": 7.843747660262333e-07,
+      "loss": 1.6216,
+      "step": 5590
+    },
+    {
+      "epoch": 0.3770088608874364,
+      "grad_norm": 4.542837235889031,
+      "learning_rate": 7.838908928763028e-07,
+      "loss": 1.711,
+      "step": 5595
+    },
+    {
+      "epoch": 0.3773457767595431,
+      "grad_norm": 4.628730063837452,
+      "learning_rate": 7.834066270467689e-07,
+      "loss": 1.7258,
+      "step": 5600
+    },
+    {
+      "epoch": 0.3776826926316499,
+      "grad_norm": 4.599442184002633,
+      "learning_rate": 7.829219692074707e-07,
+      "loss": 1.7791,
+      "step": 5605
+    },
+    {
+      "epoch": 0.3780196085037566,
+      "grad_norm": 4.616484916394916,
+      "learning_rate": 7.824369200287899e-07,
+      "loss": 1.7359,
+      "step": 5610
+    },
+    {
+      "epoch": 0.37835652437586337,
+      "grad_norm": 4.447771634117519,
+      "learning_rate": 7.819514801816496e-07,
+      "loss": 1.7386,
+      "step": 5615
+    },
+    {
+      "epoch": 0.3786934402479701,
+      "grad_norm": 4.521321589840539,
+      "learning_rate": 7.814656503375128e-07,
+      "loss": 1.6908,
+      "step": 5620
+    },
+    {
+      "epoch": 0.3790303561200768,
+      "grad_norm": 4.903527067915301,
+      "learning_rate": 7.809794311683828e-07,
+      "loss": 1.7085,
+      "step": 5625
+    },
+    {
+      "epoch": 0.37936727199218356,
+      "grad_norm": 4.183363887244186,
+      "learning_rate": 7.804928233468006e-07,
+      "loss": 1.7351,
+      "step": 5630
+    },
+    {
+      "epoch": 0.3797041878642903,
+      "grad_norm": 4.413163244248408,
+      "learning_rate": 7.80005827545845e-07,
+      "loss": 1.7733,
+      "step": 5635
+    },
+    {
+      "epoch": 0.38004110373639705,
+      "grad_norm": 4.907578526330726,
+      "learning_rate": 7.795184444391318e-07,
+      "loss": 1.7099,
+      "step": 5640
+    },
+    {
+      "epoch": 0.38037801960850376,
+      "grad_norm": 4.351667924698213,
+      "learning_rate": 7.790306747008119e-07,
+      "loss": 1.6799,
+      "step": 5645
+    },
+    {
+      "epoch": 0.3807149354806105,
+      "grad_norm": 4.370233455711544,
+      "learning_rate": 7.785425190055719e-07,
+      "loss": 1.7144,
+      "step": 5650
+    },
+    {
+      "epoch": 0.38105185135271724,
+      "grad_norm": 5.18211571994226,
+      "learning_rate": 7.780539780286312e-07,
+      "loss": 1.6515,
+      "step": 5655
+    },
+    {
+      "epoch": 0.38138876722482395,
+      "grad_norm": 4.682021085687706,
+      "learning_rate": 7.77565052445743e-07,
+      "loss": 1.7568,
+      "step": 5660
+    },
+    {
+      "epoch": 0.3817256830969307,
+      "grad_norm": 4.383186975516879,
+      "learning_rate": 7.770757429331919e-07,
+      "loss": 1.7622,
+      "step": 5665
+    },
+    {
+      "epoch": 0.38206259896903744,
+      "grad_norm": 4.8086361127286255,
+      "learning_rate": 7.765860501677939e-07,
+      "loss": 1.7293,
+      "step": 5670
+    },
+    {
+      "epoch": 0.38239951484114415,
+      "grad_norm": 4.268414299928615,
+      "learning_rate": 7.760959748268949e-07,
+      "loss": 1.668,
+      "step": 5675
+    },
+    {
+      "epoch": 0.3827364307132509,
+      "grad_norm": 4.499269064286495,
+      "learning_rate": 7.756055175883701e-07,
+      "loss": 1.7434,
+      "step": 5680
+    },
+    {
+      "epoch": 0.38307334658535763,
+      "grad_norm": 4.662888573636476,
+      "learning_rate": 7.751146791306231e-07,
+      "loss": 1.7432,
+      "step": 5685
+    },
+    {
+      "epoch": 0.38341026245746435,
+      "grad_norm": 4.369518419175261,
+      "learning_rate": 7.746234601325843e-07,
+      "loss": 1.6939,
+      "step": 5690
+    },
+    {
+      "epoch": 0.3837471783295711,
+      "grad_norm": 4.659617437659864,
+      "learning_rate": 7.741318612737111e-07,
+      "loss": 1.6964,
+      "step": 5695
+    },
+    {
+      "epoch": 0.38408409420167783,
+      "grad_norm": 4.743139224013069,
+      "learning_rate": 7.73639883233986e-07,
+      "loss": 1.6675,
+      "step": 5700
+    },
+    {
+      "epoch": 0.3844210100737846,
+      "grad_norm": 4.438211753386986,
+      "learning_rate": 7.731475266939158e-07,
+      "loss": 1.6696,
+      "step": 5705
+    },
+    {
+      "epoch": 0.3847579259458913,
+      "grad_norm": 6.7396570413557235,
+      "learning_rate": 7.726547923345313e-07,
+      "loss": 1.7069,
+      "step": 5710
+    },
+    {
+      "epoch": 0.385094841817998,
+      "grad_norm": 4.752177328606303,
+      "learning_rate": 7.721616808373855e-07,
+      "loss": 1.7465,
+      "step": 5715
+    },
+    {
+      "epoch": 0.3854317576901048,
+      "grad_norm": 4.5518253954744035,
+      "learning_rate": 7.716681928845532e-07,
+      "loss": 1.6894,
+      "step": 5720
+    },
+    {
+      "epoch": 0.3857686735622115,
+      "grad_norm": 4.567439842382045,
+      "learning_rate": 7.711743291586298e-07,
+      "loss": 1.7141,
+      "step": 5725
+    },
+    {
+      "epoch": 0.3861055894343183,
+      "grad_norm": 4.459533696503045,
+      "learning_rate": 7.706800903427309e-07,
+      "loss": 1.6393,
+      "step": 5730
+    },
+    {
+      "epoch": 0.386442505306425,
+      "grad_norm": 4.585967312052329,
+      "learning_rate": 7.701854771204905e-07,
+      "loss": 1.7233,
+      "step": 5735
+    },
+    {
+      "epoch": 0.3867794211785317,
+      "grad_norm": 4.572181355645383,
+      "learning_rate": 7.696904901760606e-07,
+      "loss": 1.688,
+      "step": 5740
+    },
+    {
+      "epoch": 0.38711633705063847,
+      "grad_norm": 4.78962813492679,
+      "learning_rate": 7.691951301941102e-07,
+      "loss": 1.7155,
+      "step": 5745
+    },
+    {
+      "epoch": 0.3874532529227452,
+      "grad_norm": 5.055611062311816,
+      "learning_rate": 7.68699397859824e-07,
+      "loss": 1.7816,
+      "step": 5750
+    },
+    {
+      "epoch": 0.38779016879485195,
+      "grad_norm": 4.4705802156633,
+      "learning_rate": 7.682032938589023e-07,
+      "loss": 1.7335,
+      "step": 5755
+    },
+    {
+      "epoch": 0.38812708466695867,
+      "grad_norm": 4.920709221588469,
+      "learning_rate": 7.677068188775589e-07,
+      "loss": 1.6501,
+      "step": 5760
+    },
+    {
+      "epoch": 0.3884640005390654,
+      "grad_norm": 4.36634733581777,
+      "learning_rate": 7.67209973602521e-07,
+      "loss": 1.7318,
+      "step": 5765
+    },
+    {
+      "epoch": 0.38880091641117215,
+      "grad_norm": 4.501042191329174,
+      "learning_rate": 7.667127587210282e-07,
+      "loss": 1.7573,
+      "step": 5770
+    },
+    {
+      "epoch": 0.38913783228327886,
+      "grad_norm": 4.750610976620214,
+      "learning_rate": 7.66215174920831e-07,
+      "loss": 1.735,
+      "step": 5775
+    },
+    {
+      "epoch": 0.3894747481553856,
+      "grad_norm": 4.518360253175234,
+      "learning_rate": 7.657172228901905e-07,
+      "loss": 1.6926,
+      "step": 5780
+    },
+    {
+      "epoch": 0.38981166402749234,
+      "grad_norm": 4.656074322809511,
+      "learning_rate": 7.652189033178766e-07,
+      "loss": 1.7086,
+      "step": 5785
+    },
+    {
+      "epoch": 0.39014857989959906,
+      "grad_norm": 4.808862642475306,
+      "learning_rate": 7.647202168931683e-07,
+      "loss": 1.6493,
+      "step": 5790
+    },
+    {
+      "epoch": 0.3904854957717058,
+      "grad_norm": 4.652819924014255,
+      "learning_rate": 7.642211643058516e-07,
+      "loss": 1.6836,
+      "step": 5795
+    },
+    {
+      "epoch": 0.39082241164381254,
+      "grad_norm": 4.461568567691955,
+      "learning_rate": 7.637217462462189e-07,
+      "loss": 1.7185,
+      "step": 5800
+    },
+    {
+      "epoch": 0.39115932751591925,
+      "grad_norm": 4.605998238250465,
+      "learning_rate": 7.632219634050685e-07,
+      "loss": 1.6989,
+      "step": 5805
+    },
+    {
+      "epoch": 0.391496243388026,
+      "grad_norm": 4.592693407010997,
+      "learning_rate": 7.62721816473703e-07,
+      "loss": 1.7713,
+      "step": 5810
+    },
+    {
+      "epoch": 0.39183315926013274,
+      "grad_norm": 5.078801838131806,
+      "learning_rate": 7.622213061439287e-07,
+      "loss": 1.6707,
+      "step": 5815
+    },
+    {
+      "epoch": 0.3921700751322395,
+      "grad_norm": 5.172560564048931,
+      "learning_rate": 7.617204331080544e-07,
+      "loss": 1.6874,
+      "step": 5820
+    },
+    {
+      "epoch": 0.3925069910043462,
+      "grad_norm": 4.616475259743771,
+      "learning_rate": 7.612191980588907e-07,
+      "loss": 1.6954,
+      "step": 5825
+    },
+    {
+      "epoch": 0.39284390687645293,
+      "grad_norm": 4.547897998439806,
+      "learning_rate": 7.60717601689749e-07,
+      "loss": 1.7337,
+      "step": 5830
+    },
+    {
+      "epoch": 0.3931808227485597,
+      "grad_norm": 4.60176135143651,
+      "learning_rate": 7.602156446944405e-07,
+      "loss": 1.7093,
+      "step": 5835
+    },
+    {
+      "epoch": 0.3935177386206664,
+      "grad_norm": 4.695085433755888,
+      "learning_rate": 7.597133277672751e-07,
+      "loss": 1.6814,
+      "step": 5840
+    },
+    {
+      "epoch": 0.3938546544927732,
+      "grad_norm": 5.077084145839541,
+      "learning_rate": 7.592106516030607e-07,
+      "loss": 1.6684,
+      "step": 5845
+    },
+    {
+      "epoch": 0.3941915703648799,
+      "grad_norm": 4.969131663429761,
+      "learning_rate": 7.587076168971022e-07,
+      "loss": 1.6603,
+      "step": 5850
+    },
+    {
+      "epoch": 0.3945284862369866,
+      "grad_norm": 4.578673498032634,
+      "learning_rate": 7.582042243451998e-07,
+      "loss": 1.7228,
+      "step": 5855
+    },
+    {
+      "epoch": 0.3948654021090934,
+      "grad_norm": 4.866217603578348,
+      "learning_rate": 7.577004746436494e-07,
+      "loss": 1.7278,
+      "step": 5860
+    },
+    {
+      "epoch": 0.3952023179812001,
+      "grad_norm": 4.706464482205939,
+      "learning_rate": 7.571963684892404e-07,
+      "loss": 1.7142,
+      "step": 5865
+    },
+    {
+      "epoch": 0.3955392338533068,
+      "grad_norm": 4.308379972805312,
+      "learning_rate": 7.566919065792558e-07,
+      "loss": 1.7033,
+      "step": 5870
+    },
+    {
+      "epoch": 0.3958761497254136,
+      "grad_norm": 4.853147236613105,
+      "learning_rate": 7.561870896114704e-07,
+      "loss": 1.6539,
+      "step": 5875
+    },
+    {
+      "epoch": 0.3962130655975203,
+      "grad_norm": 4.690507382976398,
+      "learning_rate": 7.556819182841498e-07,
+      "loss": 1.6581,
+      "step": 5880
+    },
+    {
+      "epoch": 0.39654998146962706,
+      "grad_norm": 4.519726029042387,
+      "learning_rate": 7.551763932960502e-07,
+      "loss": 1.7522,
+      "step": 5885
+    },
+    {
+      "epoch": 0.39688689734173377,
+      "grad_norm": 4.533553164902254,
+      "learning_rate": 7.546705153464168e-07,
+      "loss": 1.7346,
+      "step": 5890
+    },
+    {
+      "epoch": 0.3972238132138405,
+      "grad_norm": 4.610001390974794,
+      "learning_rate": 7.54164285134983e-07,
+      "loss": 1.6907,
+      "step": 5895
+    },
+    {
+      "epoch": 0.39756072908594725,
+      "grad_norm": 4.760381077129122,
+      "learning_rate": 7.536577033619696e-07,
+      "loss": 1.7315,
+      "step": 5900
+    },
+    {
+      "epoch": 0.39789764495805396,
+      "grad_norm": 4.629241226901361,
+      "learning_rate": 7.531507707280836e-07,
+      "loss": 1.6959,
+      "step": 5905
+    },
+    {
+      "epoch": 0.39823456083016073,
+      "grad_norm": 4.446115728404233,
+      "learning_rate": 7.526434879345171e-07,
+      "loss": 1.7361,
+      "step": 5910
+    },
+    {
+      "epoch": 0.39857147670226745,
+      "grad_norm": 4.633205448295902,
+      "learning_rate": 7.521358556829469e-07,
+      "loss": 1.7537,
+      "step": 5915
+    },
+    {
+      "epoch": 0.39890839257437416,
+      "grad_norm": 4.670072836193434,
+      "learning_rate": 7.51627874675533e-07,
+      "loss": 1.769,
+      "step": 5920
+    },
+    {
+      "epoch": 0.39924530844648093,
+      "grad_norm": 4.631995984236844,
+      "learning_rate": 7.511195456149177e-07,
+      "loss": 1.7361,
+      "step": 5925
+    },
+    {
+      "epoch": 0.39958222431858764,
+      "grad_norm": 4.396825301474403,
+      "learning_rate": 7.50610869204225e-07,
+      "loss": 1.7077,
+      "step": 5930
+    },
+    {
+      "epoch": 0.3999191401906944,
+      "grad_norm": 4.668120397333451,
+      "learning_rate": 7.50101846147059e-07,
+      "loss": 1.8077,
+      "step": 5935
+    },
+    {
+      "epoch": 0.4002560560628011,
+      "grad_norm": 4.610399362003563,
+      "learning_rate": 7.495924771475037e-07,
+      "loss": 1.7378,
+      "step": 5940
+    },
+    {
+      "epoch": 0.40059297193490784,
+      "grad_norm": 4.662838108941898,
+      "learning_rate": 7.490827629101211e-07,
+      "loss": 1.7173,
+      "step": 5945
+    },
+    {
+      "epoch": 0.4009298878070146,
+      "grad_norm": 4.490648340857639,
+      "learning_rate": 7.485727041399513e-07,
+      "loss": 1.671,
+      "step": 5950
+    },
+    {
+      "epoch": 0.4012668036791213,
+      "grad_norm": 4.506223348196042,
+      "learning_rate": 7.480623015425105e-07,
+      "loss": 1.6699,
+      "step": 5955
+    },
+    {
+      "epoch": 0.40160371955122803,
+      "grad_norm": 4.2928272510154635,
+      "learning_rate": 7.475515558237909e-07,
+      "loss": 1.7237,
+      "step": 5960
+    },
+    {
+      "epoch": 0.4019406354233348,
+      "grad_norm": 4.6387104283175855,
+      "learning_rate": 7.470404676902587e-07,
+      "loss": 1.7306,
+      "step": 5965
+    },
+    {
+      "epoch": 0.4022775512954415,
+      "grad_norm": 4.4761023884421745,
+      "learning_rate": 7.465290378488544e-07,
+      "loss": 1.7759,
+      "step": 5970
+    },
+    {
+      "epoch": 0.4026144671675483,
+      "grad_norm": 4.732318729645672,
+      "learning_rate": 7.460172670069909e-07,
+      "loss": 1.739,
+      "step": 5975
+    },
+    {
+      "epoch": 0.402951383039655,
+      "grad_norm": 4.457319075535821,
+      "learning_rate": 7.455051558725524e-07,
+      "loss": 1.7418,
+      "step": 5980
+    },
+    {
+      "epoch": 0.4032882989117617,
+      "grad_norm": 4.478093757793555,
+      "learning_rate": 7.449927051538944e-07,
+      "loss": 1.6646,
+      "step": 5985
+    },
+    {
+      "epoch": 0.4036252147838685,
+      "grad_norm": 4.71073854533818,
+      "learning_rate": 7.444799155598419e-07,
+      "loss": 1.7251,
+      "step": 5990
+    },
+    {
+      "epoch": 0.4039621306559752,
+      "grad_norm": 4.612275010632581,
+      "learning_rate": 7.439667877996884e-07,
+      "loss": 1.6614,
+      "step": 5995
+    },
+    {
+      "epoch": 0.40429904652808196,
+      "grad_norm": 4.166735332378841,
+      "learning_rate": 7.434533225831951e-07,
+      "loss": 1.7487,
+      "step": 6000
+    },
+    {
+      "epoch": 0.4046359624001887,
+      "grad_norm": 4.437607194642227,
+      "learning_rate": 7.429395206205908e-07,
+      "loss": 1.6909,
+      "step": 6005
+    },
+    {
+      "epoch": 0.4049728782722954,
+      "grad_norm": 4.928632105168086,
+      "learning_rate": 7.424253826225689e-07,
+      "loss": 1.6897,
+      "step": 6010
+    },
+    {
+      "epoch": 0.40530979414440216,
+      "grad_norm": 4.623826941528856,
+      "learning_rate": 7.419109093002887e-07,
+      "loss": 1.6674,
+      "step": 6015
+    },
+    {
+      "epoch": 0.40564671001650887,
+      "grad_norm": 4.754454500032497,
+      "learning_rate": 7.413961013653725e-07,
+      "loss": 1.7288,
+      "step": 6020
+    },
+    {
+      "epoch": 0.40598362588861564,
+      "grad_norm": 4.848985166114699,
+      "learning_rate": 7.408809595299057e-07,
+      "loss": 1.6862,
+      "step": 6025
+    },
+    {
+      "epoch": 0.40632054176072235,
+      "grad_norm": 4.441849353174314,
+      "learning_rate": 7.403654845064358e-07,
+      "loss": 1.7385,
+      "step": 6030
+    },
+    {
+      "epoch": 0.40665745763282907,
+      "grad_norm": 5.058126999436806,
+      "learning_rate": 7.398496770079709e-07,
+      "loss": 1.7187,
+      "step": 6035
+    },
+    {
+      "epoch": 0.40699437350493584,
+      "grad_norm": 4.42595695672934,
+      "learning_rate": 7.393335377479792e-07,
+      "loss": 1.7185,
+      "step": 6040
+    },
+    {
+      "epoch": 0.40733128937704255,
+      "grad_norm": 4.560697874187131,
+      "learning_rate": 7.388170674403872e-07,
+      "loss": 1.6642,
+      "step": 6045
+    },
+    {
+      "epoch": 0.40766820524914926,
+      "grad_norm": 4.931680925309431,
+      "learning_rate": 7.383002667995804e-07,
+      "loss": 1.7109,
+      "step": 6050
+    },
+    {
+      "epoch": 0.40800512112125603,
+      "grad_norm": 4.604546109504864,
+      "learning_rate": 7.377831365404001e-07,
+      "loss": 1.6834,
+      "step": 6055
+    },
+    {
+      "epoch": 0.40834203699336274,
+      "grad_norm": 4.576988553550718,
+      "learning_rate": 7.372656773781442e-07,
+      "loss": 1.7549,
+      "step": 6060
+    },
+    {
+      "epoch": 0.4086789528654695,
+      "grad_norm": 4.46737050763205,
+      "learning_rate": 7.367478900285654e-07,
+      "loss": 1.7219,
+      "step": 6065
+    },
+    {
+      "epoch": 0.4090158687375762,
+      "grad_norm": 4.664858766858329,
+      "learning_rate": 7.362297752078702e-07,
+      "loss": 1.6568,
+      "step": 6070
+    },
+    {
+      "epoch": 0.40935278460968294,
+      "grad_norm": 4.710657927303359,
+      "learning_rate": 7.357113336327181e-07,
+      "loss": 1.7278,
+      "step": 6075
+    },
+    {
+      "epoch": 0.4096897004817897,
+      "grad_norm": 4.597622314315028,
+      "learning_rate": 7.351925660202207e-07,
+      "loss": 1.6781,
+      "step": 6080
+    },
+    {
+      "epoch": 0.4100266163538964,
+      "grad_norm": 4.731937253448,
+      "learning_rate": 7.346734730879407e-07,
+      "loss": 1.7386,
+      "step": 6085
+    },
+    {
+      "epoch": 0.4103635322260032,
+      "grad_norm": 4.532128690172484,
+      "learning_rate": 7.341540555538902e-07,
+      "loss": 1.7474,
+      "step": 6090
+    },
+    {
+      "epoch": 0.4107004480981099,
+      "grad_norm": 5.063100649227562,
+      "learning_rate": 7.33634314136531e-07,
+      "loss": 1.7513,
+      "step": 6095
+    },
+    {
+      "epoch": 0.4110373639702166,
+      "grad_norm": 4.861672431801468,
+      "learning_rate": 7.331142495547724e-07,
+      "loss": 1.7346,
+      "step": 6100
+    },
+    {
+      "epoch": 0.4113742798423234,
+      "grad_norm": 4.760654858558883,
+      "learning_rate": 7.325938625279709e-07,
+      "loss": 1.6565,
+      "step": 6105
+    },
+    {
+      "epoch": 0.4117111957144301,
+      "grad_norm": 4.591183402492795,
+      "learning_rate": 7.320731537759293e-07,
+      "loss": 1.7999,
+      "step": 6110
+    },
+    {
+      "epoch": 0.41204811158653687,
+      "grad_norm": 4.8148792163856,
+      "learning_rate": 7.315521240188944e-07,
+      "loss": 1.7689,
+      "step": 6115
+    },
+    {
+      "epoch": 0.4123850274586436,
+      "grad_norm": 5.225593664936883,
+      "learning_rate": 7.310307739775585e-07,
+      "loss": 1.7211,
+      "step": 6120
+    },
+    {
+      "epoch": 0.4127219433307503,
+      "grad_norm": 4.873654095668805,
+      "learning_rate": 7.305091043730557e-07,
+      "loss": 1.761,
+      "step": 6125
+    },
+    {
+      "epoch": 0.41305885920285706,
+      "grad_norm": 4.790346021865599,
+      "learning_rate": 7.299871159269626e-07,
+      "loss": 1.7007,
+      "step": 6130
+    },
+    {
+      "epoch": 0.4133957750749638,
+      "grad_norm": 4.52525638545263,
+      "learning_rate": 7.294648093612968e-07,
+      "loss": 1.7322,
+      "step": 6135
+    },
+    {
+      "epoch": 0.4137326909470705,
+      "grad_norm": 4.465403456484435,
+      "learning_rate": 7.28942185398516e-07,
+      "loss": 1.6736,
+      "step": 6140
+    },
+    {
+      "epoch": 0.41406960681917726,
+      "grad_norm": 4.68397753694206,
+      "learning_rate": 7.284192447615168e-07,
+      "loss": 1.689,
+      "step": 6145
+    },
+    {
+      "epoch": 0.414406522691284,
+      "grad_norm": 4.5810644765337365,
+      "learning_rate": 7.278959881736338e-07,
+      "loss": 1.7383,
+      "step": 6150
+    },
+    {
+      "epoch": 0.41474343856339074,
+      "grad_norm": 5.0164765638110635,
+      "learning_rate": 7.273724163586387e-07,
+      "loss": 1.7011,
+      "step": 6155
+    },
+    {
+      "epoch": 0.41508035443549746,
+      "grad_norm": 5.086309499155851,
+      "learning_rate": 7.268485300407394e-07,
+      "loss": 1.7185,
+      "step": 6160
+    },
+    {
+      "epoch": 0.41541727030760417,
+      "grad_norm": 3.905456024402334,
+      "learning_rate": 7.263243299445783e-07,
+      "loss": 1.7467,
+      "step": 6165
+    },
+    {
+      "epoch": 0.41575418617971094,
+      "grad_norm": 4.934234757320097,
+      "learning_rate": 7.257998167952322e-07,
+      "loss": 1.6398,
+      "step": 6170
+    },
+    {
+      "epoch": 0.41609110205181765,
+      "grad_norm": 4.438570362494904,
+      "learning_rate": 7.25274991318211e-07,
+      "loss": 1.6563,
+      "step": 6175
+    },
+    {
+      "epoch": 0.4164280179239244,
+      "grad_norm": 4.752967431338472,
+      "learning_rate": 7.247498542394566e-07,
+      "loss": 1.7653,
+      "step": 6180
+    },
+    {
+      "epoch": 0.41676493379603113,
+      "grad_norm": 4.518082197921104,
+      "learning_rate": 7.242244062853416e-07,
+      "loss": 1.7131,
+      "step": 6185
+    },
+    {
+      "epoch": 0.41710184966813785,
+      "grad_norm": 4.581616402189418,
+      "learning_rate": 7.236986481826688e-07,
+      "loss": 1.7834,
+      "step": 6190
+    },
+    {
+      "epoch": 0.4174387655402446,
+      "grad_norm": 4.220211283032085,
+      "learning_rate": 7.231725806586699e-07,
+      "loss": 1.6772,
+      "step": 6195
+    },
+    {
+      "epoch": 0.41777568141235133,
+      "grad_norm": 5.213439810668329,
+      "learning_rate": 7.22646204441005e-07,
+      "loss": 1.7462,
+      "step": 6200
+    },
+    {
+      "epoch": 0.4181125972844581,
+      "grad_norm": 4.329098312504522,
+      "learning_rate": 7.221195202577606e-07,
+      "loss": 1.713,
+      "step": 6205
+    },
+    {
+      "epoch": 0.4184495131565648,
+      "grad_norm": 4.80682214677469,
+      "learning_rate": 7.215925288374496e-07,
+      "loss": 1.7204,
+      "step": 6210
+    },
+    {
+      "epoch": 0.4187864290286715,
+      "grad_norm": 4.914051282059273,
+      "learning_rate": 7.210652309090098e-07,
+      "loss": 1.7357,
+      "step": 6215
+    },
+    {
+      "epoch": 0.4191233449007783,
+      "grad_norm": 4.845584202761493,
+      "learning_rate": 7.205376272018025e-07,
+      "loss": 1.7637,
+      "step": 6220
+    },
+    {
+      "epoch": 0.419460260772885,
+      "grad_norm": 4.643682497538772,
+      "learning_rate": 7.200097184456128e-07,
+      "loss": 1.7463,
+      "step": 6225
+    },
+    {
+      "epoch": 0.4197971766449917,
+      "grad_norm": 4.54581096465623,
+      "learning_rate": 7.19481505370647e-07,
+      "loss": 1.7802,
+      "step": 6230
+    },
+    {
+      "epoch": 0.4201340925170985,
+      "grad_norm": 4.745316630456448,
+      "learning_rate": 7.189529887075327e-07,
+      "loss": 1.7233,
+      "step": 6235
+    },
+    {
+      "epoch": 0.4204710083892052,
+      "grad_norm": 4.574264194845761,
+      "learning_rate": 7.184241691873174e-07,
+      "loss": 1.7399,
+      "step": 6240
+    },
+    {
+      "epoch": 0.42080792426131197,
+      "grad_norm": 4.473298847625997,
+      "learning_rate": 7.178950475414675e-07,
+      "loss": 1.7663,
+      "step": 6245
+    },
+    {
+      "epoch": 0.4211448401334187,
+      "grad_norm": 4.791887801785815,
+      "learning_rate": 7.173656245018671e-07,
+      "loss": 1.7203,
+      "step": 6250
+    },
+    {
+      "epoch": 0.4214817560055254,
+      "grad_norm": 4.4061237952087415,
+      "learning_rate": 7.168359008008177e-07,
+      "loss": 1.7423,
+      "step": 6255
+    },
+    {
+      "epoch": 0.42181867187763217,
+      "grad_norm": 4.862344084483968,
+      "learning_rate": 7.163058771710358e-07,
+      "loss": 1.6863,
+      "step": 6260
+    },
+    {
+      "epoch": 0.4221555877497389,
+      "grad_norm": 4.4568155120703326,
+      "learning_rate": 7.157755543456539e-07,
+      "loss": 1.7336,
+      "step": 6265
+    },
+    {
+      "epoch": 0.42249250362184565,
+      "grad_norm": 4.681291018047077,
+      "learning_rate": 7.152449330582173e-07,
+      "loss": 1.7215,
+      "step": 6270
+    },
+    {
+      "epoch": 0.42282941949395236,
+      "grad_norm": 4.665256916999848,
+      "learning_rate": 7.147140140426848e-07,
+      "loss": 1.7557,
+      "step": 6275
+    },
+    {
+      "epoch": 0.4231663353660591,
+      "grad_norm": 4.961134710940176,
+      "learning_rate": 7.141827980334265e-07,
+      "loss": 1.6411,
+      "step": 6280
+    },
+    {
+      "epoch": 0.42350325123816585,
+      "grad_norm": 4.371043870195574,
+      "learning_rate": 7.136512857652239e-07,
+      "loss": 1.7018,
+      "step": 6285
+    },
+    {
+      "epoch": 0.42384016711027256,
+      "grad_norm": 4.8217954723863174,
+      "learning_rate": 7.131194779732681e-07,
+      "loss": 1.7602,
+      "step": 6290
+    },
+    {
+      "epoch": 0.4241770829823793,
+      "grad_norm": 4.613353819208612,
+      "learning_rate": 7.125873753931586e-07,
+      "loss": 1.6274,
+      "step": 6295
+    },
+    {
+      "epoch": 0.42451399885448604,
+      "grad_norm": 4.945775372506359,
+      "learning_rate": 7.120549787609029e-07,
+      "loss": 1.6926,
+      "step": 6300
+    },
+    {
+      "epoch": 0.42485091472659275,
+      "grad_norm": 4.535099056814416,
+      "learning_rate": 7.115222888129156e-07,
+      "loss": 1.7193,
+      "step": 6305
+    },
+    {
+      "epoch": 0.4251878305986995,
+      "grad_norm": 4.593782020802655,
+      "learning_rate": 7.109893062860161e-07,
+      "loss": 1.6644,
+      "step": 6310
+    },
+    {
+      "epoch": 0.42552474647080624,
+      "grad_norm": 4.674544728983331,
+      "learning_rate": 7.104560319174296e-07,
+      "loss": 1.7831,
+      "step": 6315
+    },
+    {
+      "epoch": 0.42586166234291295,
+      "grad_norm": 4.335763702766309,
+      "learning_rate": 7.099224664447841e-07,
+      "loss": 1.7124,
+      "step": 6320
+    },
+    {
+      "epoch": 0.4261985782150197,
+      "grad_norm": 5.061683568780474,
+      "learning_rate": 7.093886106061106e-07,
+      "loss": 1.7773,
+      "step": 6325
+    },
+    {
+      "epoch": 0.42653549408712643,
+      "grad_norm": 4.67914727019353,
+      "learning_rate": 7.088544651398421e-07,
+      "loss": 1.7213,
+      "step": 6330
+    },
+    {
+      "epoch": 0.4268724099592332,
+      "grad_norm": 4.722921833002874,
+      "learning_rate": 7.083200307848115e-07,
+      "loss": 1.7383,
+      "step": 6335
+    },
+    {
+      "epoch": 0.4272093258313399,
+      "grad_norm": 4.585454080536829,
+      "learning_rate": 7.077853082802516e-07,
+      "loss": 1.755,
+      "step": 6340
+    },
+    {
+      "epoch": 0.42754624170344663,
+      "grad_norm": 4.68463042835788,
+      "learning_rate": 7.072502983657939e-07,
+      "loss": 1.6551,
+      "step": 6345
+    },
+    {
+      "epoch": 0.4278831575755534,
+      "grad_norm": 4.6391044127251995,
+      "learning_rate": 7.067150017814676e-07,
+      "loss": 1.7024,
+      "step": 6350
+    },
+    {
+      "epoch": 0.4282200734476601,
+      "grad_norm": 4.824270442615483,
+      "learning_rate": 7.061794192676979e-07,
+      "loss": 1.7594,
+      "step": 6355
+    },
+    {
+      "epoch": 0.4285569893197669,
+      "grad_norm": 4.745912639857761,
+      "learning_rate": 7.056435515653058e-07,
+      "loss": 1.7587,
+      "step": 6360
+    },
+    {
+      "epoch": 0.4288939051918736,
+      "grad_norm": 4.5638877301431,
+      "learning_rate": 7.051073994155068e-07,
+      "loss": 1.7336,
+      "step": 6365
+    },
+    {
+      "epoch": 0.4292308210639803,
+      "grad_norm": 4.842666081654743,
+      "learning_rate": 7.045709635599098e-07,
+      "loss": 1.7183,
+      "step": 6370
+    },
+    {
+      "epoch": 0.4295677369360871,
+      "grad_norm": 5.268210461664882,
+      "learning_rate": 7.040342447405161e-07,
+      "loss": 1.6839,
+      "step": 6375
+    },
+    {
+      "epoch": 0.4299046528081938,
+      "grad_norm": 4.663730097112814,
+      "learning_rate": 7.034972436997184e-07,
+      "loss": 1.6709,
+      "step": 6380
+    },
+    {
+      "epoch": 0.4302415686803005,
+      "grad_norm": 4.389744444342654,
+      "learning_rate": 7.029599611803e-07,
+      "loss": 1.7155,
+      "step": 6385
+    },
+    {
+      "epoch": 0.43057848455240727,
+      "grad_norm": 4.698100060875378,
+      "learning_rate": 7.024223979254331e-07,
+      "loss": 1.7587,
+      "step": 6390
+    },
+    {
+      "epoch": 0.430915400424514,
+      "grad_norm": 4.846149276102287,
+      "learning_rate": 7.018845546786787e-07,
+      "loss": 1.6894,
+      "step": 6395
+    },
+    {
+      "epoch": 0.43125231629662075,
+      "grad_norm": 4.697432621029227,
+      "learning_rate": 7.013464321839845e-07,
+      "loss": 1.6051,
+      "step": 6400
+    },
+    {
+      "epoch": 0.43158923216872747,
+      "grad_norm": 4.405268539181798,
+      "learning_rate": 7.00808031185685e-07,
+      "loss": 1.6693,
+      "step": 6405
+    },
+    {
+      "epoch": 0.4319261480408342,
+      "grad_norm": 4.647272804848751,
+      "learning_rate": 7.002693524284997e-07,
+      "loss": 1.6855,
+      "step": 6410
+    },
+    {
+      "epoch": 0.43226306391294095,
+      "grad_norm": 4.751884151380426,
+      "learning_rate": 6.997303966575322e-07,
+      "loss": 1.7463,
+      "step": 6415
+    },
+    {
+      "epoch": 0.43259997978504766,
+      "grad_norm": 4.829801676870609,
+      "learning_rate": 6.991911646182696e-07,
+      "loss": 1.7059,
+      "step": 6420
+    },
+    {
+      "epoch": 0.43293689565715443,
+      "grad_norm": 5.165031597531048,
+      "learning_rate": 6.986516570565809e-07,
+      "loss": 1.8163,
+      "step": 6425
+    },
+    {
+      "epoch": 0.43327381152926114,
+      "grad_norm": 4.400940338706048,
+      "learning_rate": 6.981118747187163e-07,
+      "loss": 1.6662,
+      "step": 6430
+    },
+    {
+      "epoch": 0.43361072740136786,
+      "grad_norm": 4.232144300389157,
+      "learning_rate": 6.975718183513056e-07,
+      "loss": 1.6838,
+      "step": 6435
+    },
+    {
+      "epoch": 0.4339476432734746,
+      "grad_norm": 4.88653823297036,
+      "learning_rate": 6.970314887013585e-07,
+      "loss": 1.6594,
+      "step": 6440
+    },
+    {
+      "epoch": 0.43428455914558134,
+      "grad_norm": 4.673060912673942,
+      "learning_rate": 6.964908865162617e-07,
+      "loss": 1.7143,
+      "step": 6445
+    },
+    {
+      "epoch": 0.4346214750176881,
+      "grad_norm": 5.075954008879407,
+      "learning_rate": 6.959500125437801e-07,
+      "loss": 1.7682,
+      "step": 6450
+    },
+    {
+      "epoch": 0.4349583908897948,
+      "grad_norm": 4.652966290889153,
+      "learning_rate": 6.954088675320534e-07,
+      "loss": 1.6783,
+      "step": 6455
+    },
+    {
+      "epoch": 0.43529530676190153,
+      "grad_norm": 4.460818120238903,
+      "learning_rate": 6.948674522295969e-07,
+      "loss": 1.7432,
+      "step": 6460
+    },
+    {
+      "epoch": 0.4356322226340083,
+      "grad_norm": 4.5582353433289695,
+      "learning_rate": 6.943257673852993e-07,
+      "loss": 1.6726,
+      "step": 6465
+    },
+    {
+      "epoch": 0.435969138506115,
+      "grad_norm": 5.079974355852745,
+      "learning_rate": 6.937838137484225e-07,
+      "loss": 1.7775,
+      "step": 6470
+    },
+    {
+      "epoch": 0.43630605437822173,
+      "grad_norm": 5.003592553951793,
+      "learning_rate": 6.932415920686001e-07,
+      "loss": 1.6158,
+      "step": 6475
+    },
+    {
+      "epoch": 0.4366429702503285,
+      "grad_norm": 4.383664043122529,
+      "learning_rate": 6.926991030958362e-07,
+      "loss": 1.7012,
+      "step": 6480
+    },
+    {
+      "epoch": 0.4369798861224352,
+      "grad_norm": 4.937261191482905,
+      "learning_rate": 6.921563475805051e-07,
+      "loss": 1.6558,
+      "step": 6485
+    },
+    {
+      "epoch": 0.437316801994542,
+      "grad_norm": 4.544148828349001,
+      "learning_rate": 6.916133262733493e-07,
+      "loss": 1.6468,
+      "step": 6490
+    },
+    {
+      "epoch": 0.4376537178666487,
+      "grad_norm": 4.984524384019845,
+      "learning_rate": 6.910700399254793e-07,
+      "loss": 1.7185,
+      "step": 6495
+    },
+    {
+      "epoch": 0.4379906337387554,
+      "grad_norm": 4.292729528464248,
+      "learning_rate": 6.905264892883721e-07,
+      "loss": 1.6787,
+      "step": 6500
+    },
+    {
+      "epoch": 0.4383275496108622,
+      "grad_norm": 4.11769249289178,
+      "learning_rate": 6.899826751138701e-07,
+      "loss": 1.6883,
+      "step": 6505
+    },
+    {
+      "epoch": 0.4386644654829689,
+      "grad_norm": 5.018149594762948,
+      "learning_rate": 6.894385981541804e-07,
+      "loss": 1.7168,
+      "step": 6510
+    },
+    {
+      "epoch": 0.43900138135507566,
+      "grad_norm": 4.676645625070917,
+      "learning_rate": 6.888942591618736e-07,
+      "loss": 1.6919,
+      "step": 6515
+    },
+    {
+      "epoch": 0.4393382972271824,
+      "grad_norm": 4.921357277560915,
+      "learning_rate": 6.883496588898827e-07,
+      "loss": 1.6816,
+      "step": 6520
+    },
+    {
+      "epoch": 0.4396752130992891,
+      "grad_norm": 4.9134861344297605,
+      "learning_rate": 6.87804798091502e-07,
+      "loss": 1.7108,
+      "step": 6525
+    },
+    {
+      "epoch": 0.44001212897139586,
+      "grad_norm": 4.708196322941507,
+      "learning_rate": 6.872596775203864e-07,
+      "loss": 1.7465,
+      "step": 6530
+    },
+    {
+      "epoch": 0.44034904484350257,
+      "grad_norm": 4.510797543770323,
+      "learning_rate": 6.867142979305498e-07,
+      "loss": 1.6689,
+      "step": 6535
+    },
+    {
+      "epoch": 0.44068596071560934,
+      "grad_norm": 4.522970178745911,
+      "learning_rate": 6.861686600763648e-07,
+      "loss": 1.7117,
+      "step": 6540
+    },
+    {
+      "epoch": 0.44102287658771605,
+      "grad_norm": 4.812495045573041,
+      "learning_rate": 6.856227647125607e-07,
+      "loss": 1.7273,
+      "step": 6545
+    },
+    {
+      "epoch": 0.44135979245982276,
+      "grad_norm": 4.846794241054585,
+      "learning_rate": 6.850766125942235e-07,
+      "loss": 1.7438,
+      "step": 6550
+    },
+    {
+      "epoch": 0.44169670833192953,
+      "grad_norm": 4.8551636572445265,
+      "learning_rate": 6.84530204476794e-07,
+      "loss": 1.7021,
+      "step": 6555
+    },
+    {
+      "epoch": 0.44203362420403625,
+      "grad_norm": 4.651962161849585,
+      "learning_rate": 6.839835411160673e-07,
+      "loss": 1.743,
+      "step": 6560
+    },
+    {
+      "epoch": 0.44237054007614296,
+      "grad_norm": 4.39048228251673,
+      "learning_rate": 6.834366232681915e-07,
+      "loss": 1.6973,
+      "step": 6565
+    },
+    {
+      "epoch": 0.44270745594824973,
+      "grad_norm": 4.530452505592518,
+      "learning_rate": 6.828894516896664e-07,
+      "loss": 1.6399,
+      "step": 6570
+    },
+    {
+      "epoch": 0.44304437182035644,
+      "grad_norm": 4.824517867034706,
+      "learning_rate": 6.823420271373433e-07,
+      "loss": 1.7254,
+      "step": 6575
+    },
+    {
+      "epoch": 0.4433812876924632,
+      "grad_norm": 5.0237910952503775,
+      "learning_rate": 6.817943503684232e-07,
+      "loss": 1.7502,
+      "step": 6580
+    },
+    {
+      "epoch": 0.4437182035645699,
+      "grad_norm": 4.262745340509699,
+      "learning_rate": 6.812464221404558e-07,
+      "loss": 1.6935,
+      "step": 6585
+    },
+    {
+      "epoch": 0.44405511943667664,
+      "grad_norm": 4.321290314072325,
+      "learning_rate": 6.806982432113388e-07,
+      "loss": 1.7191,
+      "step": 6590
+    },
+    {
+      "epoch": 0.4443920353087834,
+      "grad_norm": 4.424079719502189,
+      "learning_rate": 6.801498143393168e-07,
+      "loss": 1.719,
+      "step": 6595
+    },
+    {
+      "epoch": 0.4447289511808901,
+      "grad_norm": 4.7796436067671575,
+      "learning_rate": 6.796011362829794e-07,
+      "loss": 1.7639,
+      "step": 6600
+    },
+    {
+      "epoch": 0.4450658670529969,
+      "grad_norm": 4.834184053342531,
+      "learning_rate": 6.790522098012621e-07,
+      "loss": 1.6925,
+      "step": 6605
+    },
+    {
+      "epoch": 0.4454027829251036,
+      "grad_norm": 4.8753621292969305,
+      "learning_rate": 6.785030356534428e-07,
+      "loss": 1.7009,
+      "step": 6610
+    },
+    {
+      "epoch": 0.4457396987972103,
+      "grad_norm": 5.00324934803504,
+      "learning_rate": 6.779536145991427e-07,
+      "loss": 1.7248,
+      "step": 6615
+    },
+    {
+      "epoch": 0.4460766146693171,
+      "grad_norm": 4.888331632690152,
+      "learning_rate": 6.774039473983243e-07,
+      "loss": 1.6402,
+      "step": 6620
+    },
+    {
+      "epoch": 0.4464135305414238,
+      "grad_norm": 4.648432475886737,
+      "learning_rate": 6.768540348112906e-07,
+      "loss": 1.7037,
+      "step": 6625
+    },
+    {
+      "epoch": 0.44675044641353057,
+      "grad_norm": 4.413897273595553,
+      "learning_rate": 6.763038775986842e-07,
+      "loss": 1.7358,
+      "step": 6630
+    },
+    {
+      "epoch": 0.4470873622856373,
+      "grad_norm": 4.429203314804429,
+      "learning_rate": 6.757534765214858e-07,
+      "loss": 1.7404,
+      "step": 6635
+    },
+    {
+      "epoch": 0.447424278157744,
+      "grad_norm": 4.931142537284251,
+      "learning_rate": 6.752028323410134e-07,
+      "loss": 1.652,
+      "step": 6640
+    },
+    {
+      "epoch": 0.44776119402985076,
+      "grad_norm": 4.500035301333365,
+      "learning_rate": 6.746519458189214e-07,
+      "loss": 1.7007,
+      "step": 6645
+    },
+    {
+      "epoch": 0.4480981099019575,
+      "grad_norm": 4.260996915055294,
+      "learning_rate": 6.741008177171993e-07,
+      "loss": 1.7038,
+      "step": 6650
+    },
+    {
+      "epoch": 0.4484350257740642,
+      "grad_norm": 4.390040709591359,
+      "learning_rate": 6.735494487981711e-07,
+      "loss": 1.6483,
+      "step": 6655
+    },
+    {
+      "epoch": 0.44877194164617096,
+      "grad_norm": 4.197183115180016,
+      "learning_rate": 6.729978398244935e-07,
+      "loss": 1.6631,
+      "step": 6660
+    },
+    {
+      "epoch": 0.44910885751827767,
+      "grad_norm": 4.554047787891987,
+      "learning_rate": 6.724459915591551e-07,
+      "loss": 1.7313,
+      "step": 6665
+    },
+    {
+      "epoch": 0.44944577339038444,
+      "grad_norm": 4.359809841144607,
+      "learning_rate": 6.718939047654763e-07,
+      "loss": 1.7046,
+      "step": 6670
+    },
+    {
+      "epoch": 0.44978268926249115,
+      "grad_norm": 4.529598264112517,
+      "learning_rate": 6.713415802071064e-07,
+      "loss": 1.7339,
+      "step": 6675
+    },
+    {
+      "epoch": 0.45011960513459787,
+      "grad_norm": 4.422320100652781,
+      "learning_rate": 6.707890186480244e-07,
+      "loss": 1.7212,
+      "step": 6680
+    },
+    {
+      "epoch": 0.45045652100670464,
+      "grad_norm": 4.461507269286107,
+      "learning_rate": 6.702362208525366e-07,
+      "loss": 1.6548,
+      "step": 6685
+    },
+    {
+      "epoch": 0.45079343687881135,
+      "grad_norm": 4.240905367444598,
+      "learning_rate": 6.696831875852763e-07,
+      "loss": 1.6188,
+      "step": 6690
+    },
+    {
+      "epoch": 0.4511303527509181,
+      "grad_norm": 4.2509022444509545,
+      "learning_rate": 6.691299196112025e-07,
+      "loss": 1.7538,
+      "step": 6695
+    },
+    {
+      "epoch": 0.45146726862302483,
+      "grad_norm": 4.572894242799344,
+      "learning_rate": 6.685764176955991e-07,
+      "loss": 1.6576,
+      "step": 6700
+    },
+    {
+      "epoch": 0.45180418449513154,
+      "grad_norm": 4.6272307826667,
+      "learning_rate": 6.680226826040727e-07,
+      "loss": 1.7617,
+      "step": 6705
+    },
+    {
+      "epoch": 0.4521411003672383,
+      "grad_norm": 4.766793229835689,
+      "learning_rate": 6.674687151025535e-07,
+      "loss": 1.6255,
+      "step": 6710
+    },
+    {
+      "epoch": 0.452478016239345,
+      "grad_norm": 4.144650326322135,
+      "learning_rate": 6.669145159572924e-07,
+      "loss": 1.699,
+      "step": 6715
+    },
+    {
+      "epoch": 0.4528149321114518,
+      "grad_norm": 4.15141194465061,
+      "learning_rate": 6.663600859348615e-07,
+      "loss": 1.6959,
+      "step": 6720
+    },
+    {
+      "epoch": 0.4531518479835585,
+      "grad_norm": 4.311310984212423,
+      "learning_rate": 6.658054258021513e-07,
+      "loss": 1.7403,
+      "step": 6725
+    },
+    {
+      "epoch": 0.4534887638556652,
+      "grad_norm": 4.10292672099467,
+      "learning_rate": 6.652505363263712e-07,
+      "loss": 1.7093,
+      "step": 6730
+    },
+    {
+      "epoch": 0.453825679727772,
+      "grad_norm": 4.669724380678364,
+      "learning_rate": 6.646954182750478e-07,
+      "loss": 1.6632,
+      "step": 6735
+    },
+    {
+      "epoch": 0.4541625955998787,
+      "grad_norm": 4.460778912261257,
+      "learning_rate": 6.641400724160234e-07,
+      "loss": 1.6495,
+      "step": 6740
+    },
+    {
+      "epoch": 0.4544995114719854,
+      "grad_norm": 4.2960984447404025,
+      "learning_rate": 6.635844995174561e-07,
+      "loss": 1.7223,
+      "step": 6745
+    },
+    {
+      "epoch": 0.4548364273440922,
+      "grad_norm": 4.740934736927874,
+      "learning_rate": 6.630287003478176e-07,
+      "loss": 1.7362,
+      "step": 6750
+    },
+    {
+      "epoch": 0.4551733432161989,
+      "grad_norm": 4.5375663036291,
+      "learning_rate": 6.624726756758927e-07,
+      "loss": 1.625,
+      "step": 6755
+    },
+    {
+      "epoch": 0.45551025908830567,
+      "grad_norm": 4.837729915756388,
+      "learning_rate": 6.619164262707782e-07,
+      "loss": 1.7317,
+      "step": 6760
+    },
+    {
+      "epoch": 0.4558471749604124,
+      "grad_norm": 4.47076219383477,
+      "learning_rate": 6.613599529018815e-07,
+      "loss": 1.6642,
+      "step": 6765
+    },
+    {
+      "epoch": 0.4561840908325191,
+      "grad_norm": 4.5442838582563025,
+      "learning_rate": 6.608032563389198e-07,
+      "loss": 1.6893,
+      "step": 6770
+    },
+    {
+      "epoch": 0.45652100670462586,
+      "grad_norm": 4.480466187001192,
+      "learning_rate": 6.602463373519196e-07,
+      "loss": 1.6984,
+      "step": 6775
+    },
+    {
+      "epoch": 0.4568579225767326,
+      "grad_norm": 4.401722232385289,
+      "learning_rate": 6.596891967112143e-07,
+      "loss": 1.67,
+      "step": 6780
+    },
+    {
+      "epoch": 0.45719483844883935,
+      "grad_norm": 4.122189683257488,
+      "learning_rate": 6.59131835187444e-07,
+      "loss": 1.6179,
+      "step": 6785
+    },
+    {
+      "epoch": 0.45753175432094606,
+      "grad_norm": 4.893513098192096,
+      "learning_rate": 6.58574253551555e-07,
+      "loss": 1.7089,
+      "step": 6790
+    },
+    {
+      "epoch": 0.4578686701930528,
+      "grad_norm": 5.100913785880151,
+      "learning_rate": 6.580164525747973e-07,
+      "loss": 1.6715,
+      "step": 6795
+    },
+    {
+      "epoch": 0.45820558606515954,
+      "grad_norm": 4.756979979355757,
+      "learning_rate": 6.574584330287247e-07,
+      "loss": 1.7278,
+      "step": 6800
+    },
+    {
+      "epoch": 0.45854250193726626,
+      "grad_norm": 4.3372744318168115,
+      "learning_rate": 6.569001956851932e-07,
+      "loss": 1.734,
+      "step": 6805
+    },
+    {
+      "epoch": 0.458879417809373,
+      "grad_norm": 4.170241900109172,
+      "learning_rate": 6.563417413163601e-07,
+      "loss": 1.7156,
+      "step": 6810
+    },
+    {
+      "epoch": 0.45921633368147974,
+      "grad_norm": 4.411221588815247,
+      "learning_rate": 6.55783070694683e-07,
+      "loss": 1.7191,
+      "step": 6815
+    },
+    {
+      "epoch": 0.45955324955358645,
+      "grad_norm": 4.373715138284619,
+      "learning_rate": 6.55224184592918e-07,
+      "loss": 1.7069,
+      "step": 6820
+    },
+    {
+      "epoch": 0.4598901654256932,
+      "grad_norm": 4.464547670300906,
+      "learning_rate": 6.546650837841203e-07,
+      "loss": 1.7035,
+      "step": 6825
+    },
+    {
+      "epoch": 0.46022708129779993,
+      "grad_norm": 4.717468251452631,
+      "learning_rate": 6.541057690416414e-07,
+      "loss": 1.6196,
+      "step": 6830
+    },
+    {
+      "epoch": 0.46056399716990665,
+      "grad_norm": 4.841886356211968,
+      "learning_rate": 6.535462411391284e-07,
+      "loss": 1.6812,
+      "step": 6835
+    },
+    {
+      "epoch": 0.4609009130420134,
+      "grad_norm": 4.34564642283158,
+      "learning_rate": 6.529865008505244e-07,
+      "loss": 1.6883,
+      "step": 6840
+    },
+    {
+      "epoch": 0.46123782891412013,
+      "grad_norm": 5.414010936464202,
+      "learning_rate": 6.524265489500651e-07,
+      "loss": 1.7524,
+      "step": 6845
+    },
+    {
+      "epoch": 0.4615747447862269,
+      "grad_norm": 4.5914727683812995,
+      "learning_rate": 6.518663862122794e-07,
+      "loss": 1.6202,
+      "step": 6850
+    },
+    {
+      "epoch": 0.4619116606583336,
+      "grad_norm": 4.375868168460969,
+      "learning_rate": 6.513060134119878e-07,
+      "loss": 1.6169,
+      "step": 6855
+    },
+    {
+      "epoch": 0.4622485765304403,
+      "grad_norm": 4.441564082721185,
+      "learning_rate": 6.507454313243015e-07,
+      "loss": 1.6837,
+      "step": 6860
+    },
+    {
+      "epoch": 0.4625854924025471,
+      "grad_norm": 4.472804618626952,
+      "learning_rate": 6.50184640724621e-07,
+      "loss": 1.7191,
+      "step": 6865
+    },
+    {
+      "epoch": 0.4629224082746538,
+      "grad_norm": 4.819571378275609,
+      "learning_rate": 6.496236423886351e-07,
+      "loss": 1.7436,
+      "step": 6870
+    },
+    {
+      "epoch": 0.4632593241467606,
+      "grad_norm": 4.511185185898374,
+      "learning_rate": 6.490624370923201e-07,
+      "loss": 1.6813,
+      "step": 6875
+    },
+    {
+      "epoch": 0.4635962400188673,
+      "grad_norm": 4.474284380549234,
+      "learning_rate": 6.485010256119388e-07,
+      "loss": 1.6681,
+      "step": 6880
+    },
+    {
+      "epoch": 0.463933155890974,
+      "grad_norm": 4.46092530350263,
+      "learning_rate": 6.479394087240389e-07,
+      "loss": 1.7347,
+      "step": 6885
+    },
+    {
+      "epoch": 0.46427007176308077,
+      "grad_norm": 5.22492163790687,
+      "learning_rate": 6.473775872054521e-07,
+      "loss": 1.763,
+      "step": 6890
+    },
+    {
+      "epoch": 0.4646069876351875,
+      "grad_norm": 4.797720542641283,
+      "learning_rate": 6.468155618332936e-07,
+      "loss": 1.6729,
+      "step": 6895
+    },
+    {
+      "epoch": 0.46494390350729425,
+      "grad_norm": 4.562185869700717,
+      "learning_rate": 6.462533333849599e-07,
+      "loss": 1.7049,
+      "step": 6900
+    },
+    {
+      "epoch": 0.46528081937940097,
+      "grad_norm": 4.29159762801429,
+      "learning_rate": 6.456909026381292e-07,
+      "loss": 1.668,
+      "step": 6905
+    },
+    {
+      "epoch": 0.4656177352515077,
+      "grad_norm": 4.92003769196269,
+      "learning_rate": 6.451282703707591e-07,
+      "loss": 1.6669,
+      "step": 6910
+    },
+    {
+      "epoch": 0.46595465112361445,
+      "grad_norm": 5.059190111464898,
+      "learning_rate": 6.445654373610854e-07,
+      "loss": 1.7021,
+      "step": 6915
+    },
+    {
+      "epoch": 0.46629156699572116,
+      "grad_norm": 4.653309079648875,
+      "learning_rate": 6.440024043876229e-07,
+      "loss": 1.7259,
+      "step": 6920
+    },
+    {
+      "epoch": 0.4666284828678279,
+      "grad_norm": 4.274946723216935,
+      "learning_rate": 6.434391722291618e-07,
+      "loss": 1.7479,
+      "step": 6925
+    },
+    {
+      "epoch": 0.46696539873993465,
+      "grad_norm": 4.808800818933682,
+      "learning_rate": 6.428757416647683e-07,
+      "loss": 1.6954,
+      "step": 6930
+    },
+    {
+      "epoch": 0.46730231461204136,
+      "grad_norm": 5.013223950151504,
+      "learning_rate": 6.42312113473783e-07,
+      "loss": 1.6911,
+      "step": 6935
+    },
+    {
+      "epoch": 0.4676392304841481,
+      "grad_norm": 4.476533094171933,
+      "learning_rate": 6.417482884358196e-07,
+      "loss": 1.6955,
+      "step": 6940
+    },
+    {
+      "epoch": 0.46797614635625484,
+      "grad_norm": 4.28394035440652,
+      "learning_rate": 6.411842673307648e-07,
+      "loss": 1.7042,
+      "step": 6945
+    },
+    {
+      "epoch": 0.46831306222836155,
+      "grad_norm": 4.698352832884206,
+      "learning_rate": 6.406200509387756e-07,
+      "loss": 1.6375,
+      "step": 6950
+    },
+    {
+      "epoch": 0.4686499781004683,
+      "grad_norm": 4.794921943207182,
+      "learning_rate": 6.400556400402796e-07,
+      "loss": 1.7587,
+      "step": 6955
+    },
+    {
+      "epoch": 0.46898689397257504,
+      "grad_norm": 4.662930706736691,
+      "learning_rate": 6.394910354159736e-07,
+      "loss": 1.6615,
+      "step": 6960
+    },
+    {
+      "epoch": 0.4693238098446818,
+      "grad_norm": 4.693125552744149,
+      "learning_rate": 6.389262378468219e-07,
+      "loss": 1.7346,
+      "step": 6965
+    },
+    {
+      "epoch": 0.4696607257167885,
+      "grad_norm": 4.3614045657280744,
+      "learning_rate": 6.38361248114056e-07,
+      "loss": 1.6847,
+      "step": 6970
+    },
+    {
+      "epoch": 0.46999764158889523,
+      "grad_norm": 4.471600545769862,
+      "learning_rate": 6.377960669991733e-07,
+      "loss": 1.8017,
+      "step": 6975
+    },
+    {
+      "epoch": 0.470334557461002,
+      "grad_norm": 4.964624731566448,
+      "learning_rate": 6.372306952839353e-07,
+      "loss": 1.7063,
+      "step": 6980
+    },
+    {
+      "epoch": 0.4706714733331087,
+      "grad_norm": 4.616530701520827,
+      "learning_rate": 6.36665133750368e-07,
+      "loss": 1.7691,
+      "step": 6985
+    },
+    {
+      "epoch": 0.4710083892052155,
+      "grad_norm": 4.712769110974502,
+      "learning_rate": 6.360993831807593e-07,
+      "loss": 1.6286,
+      "step": 6990
+    },
+    {
+      "epoch": 0.4713453050773222,
+      "grad_norm": 4.601223725894037,
+      "learning_rate": 6.355334443576589e-07,
+      "loss": 1.7151,
+      "step": 6995
+    },
+    {
+      "epoch": 0.4716822209494289,
+      "grad_norm": 4.451414543137239,
+      "learning_rate": 6.349673180638769e-07,
+      "loss": 1.6681,
+      "step": 7000
+    },
+    {
+      "epoch": 0.4720191368215357,
+      "grad_norm": 5.152496315584951,
+      "learning_rate": 6.344010050824824e-07,
+      "loss": 1.7217,
+      "step": 7005
+    },
+    {
+      "epoch": 0.4723560526936424,
+      "grad_norm": 4.3986744071785004,
+      "learning_rate": 6.338345061968032e-07,
+      "loss": 1.7014,
+      "step": 7010
+    },
+    {
+      "epoch": 0.4726929685657491,
+      "grad_norm": 4.3217833852883345,
+      "learning_rate": 6.33267822190424e-07,
+      "loss": 1.6387,
+      "step": 7015
+    },
+    {
+      "epoch": 0.4730298844378559,
+      "grad_norm": 4.400063987546612,
+      "learning_rate": 6.327009538471853e-07,
+      "loss": 1.7177,
+      "step": 7020
+    },
+    {
+      "epoch": 0.4733668003099626,
+      "grad_norm": 4.616172815467751,
+      "learning_rate": 6.321339019511828e-07,
+      "loss": 1.6774,
+      "step": 7025
+    },
+    {
+      "epoch": 0.47370371618206936,
+      "grad_norm": 4.541432828621094,
+      "learning_rate": 6.315666672867664e-07,
+      "loss": 1.726,
+      "step": 7030
+    },
+    {
+      "epoch": 0.47404063205417607,
+      "grad_norm": 4.688208120332176,
+      "learning_rate": 6.309992506385385e-07,
+      "loss": 1.6291,
+      "step": 7035
+    },
+    {
+      "epoch": 0.4743775479262828,
+      "grad_norm": 4.6366612851099,
+      "learning_rate": 6.304316527913531e-07,
+      "loss": 1.6624,
+      "step": 7040
+    },
+    {
+      "epoch": 0.47471446379838955,
+      "grad_norm": 4.691085653627631,
+      "learning_rate": 6.29863874530315e-07,
+      "loss": 1.6532,
+      "step": 7045
+    },
+    {
+      "epoch": 0.47505137967049627,
+      "grad_norm": 4.7136046478678715,
+      "learning_rate": 6.292959166407785e-07,
+      "loss": 1.7466,
+      "step": 7050
+    },
+    {
+      "epoch": 0.47538829554260303,
+      "grad_norm": 4.9212482928274,
+      "learning_rate": 6.287277799083466e-07,
+      "loss": 1.687,
+      "step": 7055
+    },
+    {
+      "epoch": 0.47572521141470975,
+      "grad_norm": 4.526217309277262,
+      "learning_rate": 6.281594651188693e-07,
+      "loss": 1.6252,
+      "step": 7060
+    },
+    {
+      "epoch": 0.47606212728681646,
+      "grad_norm": 4.396351782230909,
+      "learning_rate": 6.275909730584431e-07,
+      "loss": 1.6776,
+      "step": 7065
+    },
+    {
+      "epoch": 0.47639904315892323,
+      "grad_norm": 4.51842155556115,
+      "learning_rate": 6.270223045134095e-07,
+      "loss": 1.7123,
+      "step": 7070
+    },
+    {
+      "epoch": 0.47673595903102994,
+      "grad_norm": 4.759606117812422,
+      "learning_rate": 6.264534602703546e-07,
+      "loss": 1.6613,
+      "step": 7075
+    },
+    {
+      "epoch": 0.4770728749031367,
+      "grad_norm": 4.707001993687859,
+      "learning_rate": 6.25884441116107e-07,
+      "loss": 1.733,
+      "step": 7080
+    },
+    {
+      "epoch": 0.4774097907752434,
+      "grad_norm": 4.467447574118067,
+      "learning_rate": 6.253152478377375e-07,
+      "loss": 1.6931,
+      "step": 7085
+    },
+    {
+      "epoch": 0.47774670664735014,
+      "grad_norm": 4.611033578404069,
+      "learning_rate": 6.247458812225576e-07,
+      "loss": 1.7302,
+      "step": 7090
+    },
+    {
+      "epoch": 0.4780836225194569,
+      "grad_norm": 4.4569214323453155,
+      "learning_rate": 6.241763420581188e-07,
+      "loss": 1.7401,
+      "step": 7095
+    },
+    {
+      "epoch": 0.4784205383915636,
+      "grad_norm": 4.982762182500195,
+      "learning_rate": 6.23606631132211e-07,
+      "loss": 1.6746,
+      "step": 7100
+    },
+    {
+      "epoch": 0.47875745426367033,
+      "grad_norm": 4.607941073155509,
+      "learning_rate": 6.23036749232862e-07,
+      "loss": 1.7207,
+      "step": 7105
+    },
+    {
+      "epoch": 0.4790943701357771,
+      "grad_norm": 4.887796210429232,
+      "learning_rate": 6.224666971483355e-07,
+      "loss": 1.6298,
+      "step": 7110
+    },
+    {
+      "epoch": 0.4794312860078838,
+      "grad_norm": 4.712956172096251,
+      "learning_rate": 6.218964756671315e-07,
+      "loss": 1.6563,
+      "step": 7115
+    },
+    {
+      "epoch": 0.4797682018799906,
+      "grad_norm": 4.778565383140019,
+      "learning_rate": 6.213260855779834e-07,
+      "loss": 1.6974,
+      "step": 7120
+    },
+    {
+      "epoch": 0.4801051177520973,
+      "grad_norm": 4.40756947233122,
+      "learning_rate": 6.207555276698584e-07,
+      "loss": 1.7371,
+      "step": 7125
+    },
+    {
+      "epoch": 0.480442033624204,
+      "grad_norm": 4.631757813103534,
+      "learning_rate": 6.201848027319556e-07,
+      "loss": 1.7037,
+      "step": 7130
+    },
+    {
+      "epoch": 0.4807789494963108,
+      "grad_norm": 4.520106006973044,
+      "learning_rate": 6.196139115537054e-07,
+      "loss": 1.6721,
+      "step": 7135
+    },
+    {
+      "epoch": 0.4811158653684175,
+      "grad_norm": 4.778967002753012,
+      "learning_rate": 6.190428549247677e-07,
+      "loss": 1.7327,
+      "step": 7140
+    },
+    {
+      "epoch": 0.48145278124052426,
+      "grad_norm": 4.366124855672215,
+      "learning_rate": 6.184716336350316e-07,
+      "loss": 1.7767,
+      "step": 7145
+    },
+    {
+      "epoch": 0.481789697112631,
+      "grad_norm": 4.938934702989457,
+      "learning_rate": 6.179002484746137e-07,
+      "loss": 1.7204,
+      "step": 7150
+    },
+    {
+      "epoch": 0.4821266129847377,
+      "grad_norm": 4.570886894075659,
+      "learning_rate": 6.173287002338577e-07,
+      "loss": 1.6922,
+      "step": 7155
+    },
+    {
+      "epoch": 0.48246352885684446,
+      "grad_norm": 4.584354407905234,
+      "learning_rate": 6.167569897033322e-07,
+      "loss": 1.6977,
+      "step": 7160
+    },
+    {
+      "epoch": 0.4828004447289512,
+      "grad_norm": 4.478905810314622,
+      "learning_rate": 6.16185117673831e-07,
+      "loss": 1.7657,
+      "step": 7165
+    },
+    {
+      "epoch": 0.48313736060105794,
+      "grad_norm": 4.468921511502001,
+      "learning_rate": 6.15613084936371e-07,
+      "loss": 1.7682,
+      "step": 7170
+    },
+    {
+      "epoch": 0.48347427647316465,
+      "grad_norm": 4.40177007449443,
+      "learning_rate": 6.150408922821911e-07,
+      "loss": 1.7354,
+      "step": 7175
+    },
+    {
+      "epoch": 0.48381119234527137,
+      "grad_norm": 4.6505765708076625,
+      "learning_rate": 6.144685405027518e-07,
+      "loss": 1.6936,
+      "step": 7180
+    },
+    {
+      "epoch": 0.48414810821737814,
+      "grad_norm": 4.939569186419913,
+      "learning_rate": 6.138960303897335e-07,
+      "loss": 1.7379,
+      "step": 7185
+    },
+    {
+      "epoch": 0.48448502408948485,
+      "grad_norm": 4.679413425948785,
+      "learning_rate": 6.133233627350355e-07,
+      "loss": 1.6535,
+      "step": 7190
+    },
+    {
+      "epoch": 0.48482193996159156,
+      "grad_norm": 4.325726315028096,
+      "learning_rate": 6.127505383307754e-07,
+      "loss": 1.6835,
+      "step": 7195
+    },
+    {
+      "epoch": 0.48515885583369833,
+      "grad_norm": 4.42673458969308,
+      "learning_rate": 6.121775579692873e-07,
+      "loss": 1.7601,
+      "step": 7200
+    },
+    {
+      "epoch": 0.48549577170580505,
+      "grad_norm": 4.845073129236577,
+      "learning_rate": 6.116044224431212e-07,
+      "loss": 1.7168,
+      "step": 7205
+    },
+    {
+      "epoch": 0.4858326875779118,
+      "grad_norm": 4.546143580258222,
+      "learning_rate": 6.110311325450416e-07,
+      "loss": 1.6114,
+      "step": 7210
+    },
+    {
+      "epoch": 0.48616960345001853,
+      "grad_norm": 4.77554325019675,
+      "learning_rate": 6.104576890680263e-07,
+      "loss": 1.69,
+      "step": 7215
+    },
+    {
+      "epoch": 0.48650651932212524,
+      "grad_norm": 4.71098928853645,
+      "learning_rate": 6.098840928052663e-07,
+      "loss": 1.7593,
+      "step": 7220
+    },
+    {
+      "epoch": 0.486843435194232,
+      "grad_norm": 4.750660306122936,
+      "learning_rate": 6.093103445501629e-07,
+      "loss": 1.6479,
+      "step": 7225
+    },
+    {
+      "epoch": 0.4871803510663387,
+      "grad_norm": 4.739154005718471,
+      "learning_rate": 6.087364450963286e-07,
+      "loss": 1.6892,
+      "step": 7230
+    },
+    {
+      "epoch": 0.4875172669384455,
+      "grad_norm": 4.530538648921853,
+      "learning_rate": 6.081623952375843e-07,
+      "loss": 1.7401,
+      "step": 7235
+    },
+    {
+      "epoch": 0.4878541828105522,
+      "grad_norm": 4.649623744583918,
+      "learning_rate": 6.075881957679593e-07,
+      "loss": 1.704,
+      "step": 7240
+    },
+    {
+      "epoch": 0.4881910986826589,
+      "grad_norm": 5.577179816173861,
+      "learning_rate": 6.0701384748169e-07,
+      "loss": 1.7216,
+      "step": 7245
+    },
+    {
+      "epoch": 0.4885280145547657,
+      "grad_norm": 4.912479991065959,
+      "learning_rate": 6.064393511732181e-07,
+      "loss": 1.7308,
+      "step": 7250
+    },
+    {
+      "epoch": 0.4888649304268724,
+      "grad_norm": 4.539142069414185,
+      "learning_rate": 6.058647076371906e-07,
+      "loss": 1.7037,
+      "step": 7255
+    },
+    {
+      "epoch": 0.48920184629897917,
+      "grad_norm": 4.557851134005118,
+      "learning_rate": 6.052899176684579e-07,
+      "loss": 1.6553,
+      "step": 7260
+    },
+    {
+      "epoch": 0.4895387621710859,
+      "grad_norm": 4.739125290233805,
+      "learning_rate": 6.047149820620729e-07,
+      "loss": 1.6802,
+      "step": 7265
+    },
+    {
+      "epoch": 0.4898756780431926,
+      "grad_norm": 4.791864662336917,
+      "learning_rate": 6.0413990161329e-07,
+      "loss": 1.6821,
+      "step": 7270
+    },
+    {
+      "epoch": 0.49021259391529937,
+      "grad_norm": 4.4891631106081284,
+      "learning_rate": 6.035646771175642e-07,
+      "loss": 1.7102,
+      "step": 7275
+    },
+    {
+      "epoch": 0.4905495097874061,
+      "grad_norm": 4.7430657098261495,
+      "learning_rate": 6.029893093705491e-07,
+      "loss": 1.6943,
+      "step": 7280
+    },
+    {
+      "epoch": 0.4908864256595128,
+      "grad_norm": 4.156109932566331,
+      "learning_rate": 6.024137991680973e-07,
+      "loss": 1.6247,
+      "step": 7285
+    },
+    {
+      "epoch": 0.49122334153161956,
+      "grad_norm": 4.328377079553324,
+      "learning_rate": 6.018381473062575e-07,
+      "loss": 1.7405,
+      "step": 7290
+    },
+    {
+      "epoch": 0.4915602574037263,
+      "grad_norm": 4.822065127779098,
+      "learning_rate": 6.012623545812754e-07,
+      "loss": 1.6312,
+      "step": 7295
+    },
+    {
+      "epoch": 0.49189717327583304,
+      "grad_norm": 4.6088445308051265,
+      "learning_rate": 6.006864217895906e-07,
+      "loss": 1.7427,
+      "step": 7300
+    },
+    {
+      "epoch": 0.49223408914793976,
+      "grad_norm": 4.379384111050929,
+      "learning_rate": 6.001103497278369e-07,
+      "loss": 1.7152,
+      "step": 7305
+    },
+    {
+      "epoch": 0.49257100502004647,
+      "grad_norm": 5.148479140033514,
+      "learning_rate": 5.995341391928408e-07,
+      "loss": 1.7061,
+      "step": 7310
+    },
+    {
+      "epoch": 0.49290792089215324,
+      "grad_norm": 4.437063736650041,
+      "learning_rate": 5.9895779098162e-07,
+      "loss": 1.7114,
+      "step": 7315
+    },
+    {
+      "epoch": 0.49324483676425995,
+      "grad_norm": 4.3579861769268335,
+      "learning_rate": 5.983813058913829e-07,
+      "loss": 1.6559,
+      "step": 7320
+    },
+    {
+      "epoch": 0.4935817526363667,
+      "grad_norm": 4.415788857014313,
+      "learning_rate": 5.978046847195272e-07,
+      "loss": 1.732,
+      "step": 7325
+    },
+    {
+      "epoch": 0.49391866850847344,
+      "grad_norm": 4.855952082704319,
+      "learning_rate": 5.97227928263639e-07,
+      "loss": 1.7554,
+      "step": 7330
+    },
+    {
+      "epoch": 0.49425558438058015,
+      "grad_norm": 4.370984522374044,
+      "learning_rate": 5.96651037321491e-07,
+      "loss": 1.7363,
+      "step": 7335
+    },
+    {
+      "epoch": 0.4945925002526869,
+      "grad_norm": 4.38389535228623,
+      "learning_rate": 5.960740126910425e-07,
+      "loss": 1.6725,
+      "step": 7340
+    },
+    {
+      "epoch": 0.49492941612479363,
+      "grad_norm": 4.6681190283608265,
+      "learning_rate": 5.954968551704373e-07,
+      "loss": 1.7207,
+      "step": 7345
+    },
+    {
+      "epoch": 0.4952663319969004,
+      "grad_norm": 4.697238368356332,
+      "learning_rate": 5.949195655580032e-07,
+      "loss": 1.6947,
+      "step": 7350
+    },
+    {
+      "epoch": 0.4956032478690071,
+      "grad_norm": 4.868515081677355,
+      "learning_rate": 5.943421446522509e-07,
+      "loss": 1.7416,
+      "step": 7355
+    },
+    {
+      "epoch": 0.4959401637411138,
+      "grad_norm": 4.611727186234703,
+      "learning_rate": 5.93764593251872e-07,
+      "loss": 1.6861,
+      "step": 7360
+    },
+    {
+      "epoch": 0.4962770796132206,
+      "grad_norm": 5.139350862969239,
+      "learning_rate": 5.931869121557397e-07,
+      "loss": 1.7155,
+      "step": 7365
+    },
+    {
+      "epoch": 0.4966139954853273,
+      "grad_norm": 4.446635040666518,
+      "learning_rate": 5.926091021629055e-07,
+      "loss": 1.6639,
+      "step": 7370
+    },
+    {
+      "epoch": 0.496950911357434,
+      "grad_norm": 4.559284920143756,
+      "learning_rate": 5.920311640726e-07,
+      "loss": 1.6769,
+      "step": 7375
+    },
+    {
+      "epoch": 0.4972878272295408,
+      "grad_norm": 4.80113478282041,
+      "learning_rate": 5.914530986842307e-07,
+      "loss": 1.7342,
+      "step": 7380
+    },
+    {
+      "epoch": 0.4976247431016475,
+      "grad_norm": 4.784159804294562,
+      "learning_rate": 5.908749067973809e-07,
+      "loss": 1.7198,
+      "step": 7385
+    },
+    {
+      "epoch": 0.4979616589737543,
+      "grad_norm": 4.570174722848811,
+      "learning_rate": 5.902965892118093e-07,
+      "loss": 1.7289,
+      "step": 7390
+    },
+    {
+      "epoch": 0.498298574845861,
+      "grad_norm": 4.374276853118066,
+      "learning_rate": 5.89718146727448e-07,
+      "loss": 1.7007,
+      "step": 7395
+    },
+    {
+      "epoch": 0.4986354907179677,
+      "grad_norm": 4.580820249984145,
+      "learning_rate": 5.891395801444026e-07,
+      "loss": 1.6943,
+      "step": 7400
+    },
+    {
+      "epoch": 0.49897240659007447,
+      "grad_norm": 4.883342042123659,
+      "learning_rate": 5.885608902629496e-07,
+      "loss": 1.6594,
+      "step": 7405
+    },
+    {
+      "epoch": 0.4993093224621812,
+      "grad_norm": 4.621394877055568,
+      "learning_rate": 5.879820778835364e-07,
+      "loss": 1.7269,
+      "step": 7410
+    },
+    {
+      "epoch": 0.49964623833428795,
+      "grad_norm": 5.160480894377575,
+      "learning_rate": 5.874031438067799e-07,
+      "loss": 1.6898,
+      "step": 7415
+    },
+    {
+      "epoch": 0.49998315420639466,
+      "grad_norm": 4.6208075920844305,
+      "learning_rate": 5.868240888334652e-07,
+      "loss": 1.694,
+      "step": 7420
+    },
+    {
+      "epoch": 0.5003200700785014,
+      "grad_norm": 4.736793560219323,
+      "learning_rate": 5.862449137645444e-07,
+      "loss": 1.7042,
+      "step": 7425
+    },
+    {
+      "epoch": 0.5006569859506081,
+      "grad_norm": 4.597792580937797,
+      "learning_rate": 5.856656194011365e-07,
+      "loss": 1.6511,
+      "step": 7430
+    },
+    {
+      "epoch": 0.5009939018227149,
+      "grad_norm": 4.866240037242486,
+      "learning_rate": 5.850862065445243e-07,
+      "loss": 1.6591,
+      "step": 7435
+    },
+    {
+      "epoch": 0.5013308176948216,
+      "grad_norm": 4.610706051955656,
+      "learning_rate": 5.845066759961557e-07,
+      "loss": 1.7021,
+      "step": 7440
+    },
+    {
+      "epoch": 0.5016677335669283,
+      "grad_norm": 4.425839858456235,
+      "learning_rate": 5.839270285576407e-07,
+      "loss": 1.7601,
+      "step": 7445
+    },
+    {
+      "epoch": 0.5020046494390351,
+      "grad_norm": 4.322417058164186,
+      "learning_rate": 5.833472650307509e-07,
+      "loss": 1.6423,
+      "step": 7450
+    },
+    {
+      "epoch": 0.5023415653111418,
+      "grad_norm": 4.505490220698428,
+      "learning_rate": 5.827673862174192e-07,
+      "loss": 1.7577,
+      "step": 7455
+    },
+    {
+      "epoch": 0.5026784811832485,
+      "grad_norm": 4.496510986517032,
+      "learning_rate": 5.821873929197371e-07,
+      "loss": 1.7129,
+      "step": 7460
+    },
+    {
+      "epoch": 0.5030153970553553,
+      "grad_norm": 5.187889955890472,
+      "learning_rate": 5.81607285939955e-07,
+      "loss": 1.6744,
+      "step": 7465
+    },
+    {
+      "epoch": 0.503352312927462,
+      "grad_norm": 5.006933034613527,
+      "learning_rate": 5.810270660804805e-07,
+      "loss": 1.7044,
+      "step": 7470
+    },
+    {
+      "epoch": 0.5036892287995688,
+      "grad_norm": 4.355856888149377,
+      "learning_rate": 5.80446734143877e-07,
+      "loss": 1.68,
+      "step": 7475
+    },
+    {
+      "epoch": 0.5040261446716755,
+      "grad_norm": 4.923953800176147,
+      "learning_rate": 5.798662909328633e-07,
+      "loss": 1.67,
+      "step": 7480
+    },
+    {
+      "epoch": 0.5043630605437822,
+      "grad_norm": 4.69561635565534,
+      "learning_rate": 5.792857372503119e-07,
+      "loss": 1.6919,
+      "step": 7485
+    },
+    {
+      "epoch": 0.5046999764158889,
+      "grad_norm": 4.77170903782858,
+      "learning_rate": 5.787050738992481e-07,
+      "loss": 1.6912,
+      "step": 7490
+    },
+    {
+      "epoch": 0.5050368922879956,
+      "grad_norm": 4.869923711817952,
+      "learning_rate": 5.781243016828492e-07,
+      "loss": 1.7032,
+      "step": 7495
+    },
+    {
+      "epoch": 0.5053738081601025,
+      "grad_norm": 4.585049776668319,
+      "learning_rate": 5.775434214044427e-07,
+      "loss": 1.7285,
+      "step": 7500
+    },
+    {
+      "epoch": 0.5057107240322092,
+      "grad_norm": 4.874001250553335,
+      "learning_rate": 5.769624338675057e-07,
+      "loss": 1.7505,
+      "step": 7505
+    },
+    {
+      "epoch": 0.5060476399043159,
+      "grad_norm": 4.590180315940176,
+      "learning_rate": 5.763813398756637e-07,
+      "loss": 1.7429,
+      "step": 7510
+    },
+    {
+      "epoch": 0.5063845557764226,
+      "grad_norm": 4.4868477944766925,
+      "learning_rate": 5.758001402326895e-07,
+      "loss": 1.7306,
+      "step": 7515
+    },
+    {
+      "epoch": 0.5067214716485293,
+      "grad_norm": 4.81873671171168,
+      "learning_rate": 5.752188357425019e-07,
+      "loss": 1.7365,
+      "step": 7520
+    },
+    {
+      "epoch": 0.5070583875206361,
+      "grad_norm": 4.484325181883257,
+      "learning_rate": 5.746374272091648e-07,
+      "loss": 1.7848,
+      "step": 7525
+    },
+    {
+      "epoch": 0.5073953033927429,
+      "grad_norm": 4.58168391806788,
+      "learning_rate": 5.74055915436886e-07,
+      "loss": 1.7356,
+      "step": 7530
+    },
+    {
+      "epoch": 0.5077322192648496,
+      "grad_norm": 4.6016598603627274,
+      "learning_rate": 5.734743012300162e-07,
+      "loss": 1.658,
+      "step": 7535
+    },
+    {
+      "epoch": 0.5080691351369563,
+      "grad_norm": 4.663029396984945,
+      "learning_rate": 5.728925853930475e-07,
+      "loss": 1.6654,
+      "step": 7540
+    },
+    {
+      "epoch": 0.508406051009063,
+      "grad_norm": 4.708893231070334,
+      "learning_rate": 5.72310768730613e-07,
+      "loss": 1.694,
+      "step": 7545
+    },
+    {
+      "epoch": 0.5087429668811698,
+      "grad_norm": 4.928385739239798,
+      "learning_rate": 5.717288520474849e-07,
+      "loss": 1.7709,
+      "step": 7550
+    },
+    {
+      "epoch": 0.5090798827532765,
+      "grad_norm": 4.420335047530358,
+      "learning_rate": 5.711468361485739e-07,
+      "loss": 1.687,
+      "step": 7555
+    },
+    {
+      "epoch": 0.5094167986253832,
+      "grad_norm": 4.471384860102386,
+      "learning_rate": 5.70564721838928e-07,
+      "loss": 1.7015,
+      "step": 7560
+    },
+    {
+      "epoch": 0.50975371449749,
+      "grad_norm": 4.865562364958896,
+      "learning_rate": 5.69982509923731e-07,
+      "loss": 1.7155,
+      "step": 7565
+    },
+    {
+      "epoch": 0.5100906303695967,
+      "grad_norm": 4.66852339522664,
+      "learning_rate": 5.694002012083022e-07,
+      "loss": 1.7676,
+      "step": 7570
+    },
+    {
+      "epoch": 0.5104275462417035,
+      "grad_norm": 4.242499891955819,
+      "learning_rate": 5.688177964980946e-07,
+      "loss": 1.6495,
+      "step": 7575
+    },
+    {
+      "epoch": 0.5107644621138102,
+      "grad_norm": 4.528198372867546,
+      "learning_rate": 5.682352965986935e-07,
+      "loss": 1.6835,
+      "step": 7580
+    },
+    {
+      "epoch": 0.5111013779859169,
+      "grad_norm": 5.2865806513696665,
+      "learning_rate": 5.676527023158169e-07,
+      "loss": 1.7656,
+      "step": 7585
+    },
+    {
+      "epoch": 0.5114382938580236,
+      "grad_norm": 5.006350707748134,
+      "learning_rate": 5.670700144553122e-07,
+      "loss": 1.7122,
+      "step": 7590
+    },
+    {
+      "epoch": 0.5117752097301304,
+      "grad_norm": 4.890176475676627,
+      "learning_rate": 5.664872338231571e-07,
+      "loss": 1.718,
+      "step": 7595
+    },
+    {
+      "epoch": 0.5121121256022371,
+      "grad_norm": 4.283085242905683,
+      "learning_rate": 5.659043612254573e-07,
+      "loss": 1.7049,
+      "step": 7600
+    },
+    {
+      "epoch": 0.5124490414743439,
+      "grad_norm": 4.671644584097648,
+      "learning_rate": 5.653213974684455e-07,
+      "loss": 1.6345,
+      "step": 7605
+    },
+    {
+      "epoch": 0.5127859573464506,
+      "grad_norm": 4.497378025154527,
+      "learning_rate": 5.647383433584807e-07,
+      "loss": 1.6532,
+      "step": 7610
+    },
+    {
+      "epoch": 0.5131228732185573,
+      "grad_norm": 4.770390373926907,
+      "learning_rate": 5.641551997020472e-07,
+      "loss": 1.7167,
+      "step": 7615
+    },
+    {
+      "epoch": 0.513459789090664,
+      "grad_norm": 4.644610367955537,
+      "learning_rate": 5.635719673057524e-07,
+      "loss": 1.7123,
+      "step": 7620
+    },
+    {
+      "epoch": 0.5137967049627707,
+      "grad_norm": 4.453064547486986,
+      "learning_rate": 5.629886469763273e-07,
+      "loss": 1.6907,
+      "step": 7625
+    },
+    {
+      "epoch": 0.5141336208348776,
+      "grad_norm": 4.341259433438874,
+      "learning_rate": 5.624052395206239e-07,
+      "loss": 1.7277,
+      "step": 7630
+    },
+    {
+      "epoch": 0.5144705367069843,
+      "grad_norm": 4.713591617595884,
+      "learning_rate": 5.618217457456151e-07,
+      "loss": 1.675,
+      "step": 7635
+    },
+    {
+      "epoch": 0.514807452579091,
+      "grad_norm": 4.5922560977258975,
+      "learning_rate": 5.612381664583928e-07,
+      "loss": 1.7212,
+      "step": 7640
+    },
+    {
+      "epoch": 0.5151443684511977,
+      "grad_norm": 4.508581513278687,
+      "learning_rate": 5.606545024661674e-07,
+      "loss": 1.7216,
+      "step": 7645
+    },
+    {
+      "epoch": 0.5154812843233044,
+      "grad_norm": 4.291503221118531,
+      "learning_rate": 5.600707545762667e-07,
+      "loss": 1.6839,
+      "step": 7650
+    },
+    {
+      "epoch": 0.5158182001954112,
+      "grad_norm": 4.549053737765561,
+      "learning_rate": 5.594869235961342e-07,
+      "loss": 1.7465,
+      "step": 7655
+    },
+    {
+      "epoch": 0.516155116067518,
+      "grad_norm": 4.281551741122212,
+      "learning_rate": 5.589030103333282e-07,
+      "loss": 1.7556,
+      "step": 7660
+    },
+    {
+      "epoch": 0.5164920319396247,
+      "grad_norm": 4.6183154550647165,
+      "learning_rate": 5.583190155955215e-07,
+      "loss": 1.6429,
+      "step": 7665
+    },
+    {
+      "epoch": 0.5168289478117314,
+      "grad_norm": 4.713779318598783,
+      "learning_rate": 5.57734940190499e-07,
+      "loss": 1.6639,
+      "step": 7670
+    },
+    {
+      "epoch": 0.5171658636838381,
+      "grad_norm": 4.568884925715355,
+      "learning_rate": 5.571507849261572e-07,
+      "loss": 1.6695,
+      "step": 7675
+    },
+    {
+      "epoch": 0.5175027795559449,
+      "grad_norm": 4.726799714559216,
+      "learning_rate": 5.565665506105035e-07,
+      "loss": 1.6608,
+      "step": 7680
+    },
+    {
+      "epoch": 0.5178396954280516,
+      "grad_norm": 4.747440214485794,
+      "learning_rate": 5.559822380516539e-07,
+      "loss": 1.7234,
+      "step": 7685
+    },
+    {
+      "epoch": 0.5181766113001584,
+      "grad_norm": 5.041416882602425,
+      "learning_rate": 5.553978480578335e-07,
+      "loss": 1.6882,
+      "step": 7690
+    },
+    {
+      "epoch": 0.5185135271722651,
+      "grad_norm": 4.270518135256821,
+      "learning_rate": 5.548133814373738e-07,
+      "loss": 1.6844,
+      "step": 7695
+    },
+    {
+      "epoch": 0.5188504430443718,
+      "grad_norm": 4.713569188055113,
+      "learning_rate": 5.542288389987128e-07,
+      "loss": 1.7434,
+      "step": 7700
+    },
+    {
+      "epoch": 0.5191873589164786,
+      "grad_norm": 4.424402536805608,
+      "learning_rate": 5.536442215503929e-07,
+      "loss": 1.7149,
+      "step": 7705
+    },
+    {
+      "epoch": 0.5195242747885853,
+      "grad_norm": 4.691497277737614,
+      "learning_rate": 5.530595299010606e-07,
+      "loss": 1.6705,
+      "step": 7710
+    },
+    {
+      "epoch": 0.519861190660692,
+      "grad_norm": 4.968253526258946,
+      "learning_rate": 5.524747648594651e-07,
+      "loss": 1.7873,
+      "step": 7715
+    },
+    {
+      "epoch": 0.5201981065327987,
+      "grad_norm": 4.504586292746603,
+      "learning_rate": 5.518899272344568e-07,
+      "loss": 1.7035,
+      "step": 7720
+    },
+    {
+      "epoch": 0.5205350224049055,
+      "grad_norm": 4.391879900734021,
+      "learning_rate": 5.513050178349866e-07,
+      "loss": 1.7332,
+      "step": 7725
+    },
+    {
+      "epoch": 0.5208719382770123,
+      "grad_norm": 4.961582186100007,
+      "learning_rate": 5.507200374701048e-07,
+      "loss": 1.7472,
+      "step": 7730
+    },
+    {
+      "epoch": 0.521208854149119,
+      "grad_norm": 4.274600688997273,
+      "learning_rate": 5.501349869489596e-07,
+      "loss": 1.7823,
+      "step": 7735
+    },
+    {
+      "epoch": 0.5215457700212257,
+      "grad_norm": 4.608506610910215,
+      "learning_rate": 5.495498670807967e-07,
+      "loss": 1.69,
+      "step": 7740
+    },
+    {
+      "epoch": 0.5218826858933324,
+      "grad_norm": 4.657102507065289,
+      "learning_rate": 5.489646786749574e-07,
+      "loss": 1.6813,
+      "step": 7745
+    },
+    {
+      "epoch": 0.5222196017654391,
+      "grad_norm": 4.6290567932911735,
+      "learning_rate": 5.483794225408777e-07,
+      "loss": 1.7001,
+      "step": 7750
+    },
+    {
+      "epoch": 0.522556517637546,
+      "grad_norm": 4.582264959213173,
+      "learning_rate": 5.477940994880877e-07,
+      "loss": 1.5716,
+      "step": 7755
+    },
+    {
+      "epoch": 0.5228934335096527,
+      "grad_norm": 4.457243577063691,
+      "learning_rate": 5.472087103262094e-07,
+      "loss": 1.7249,
+      "step": 7760
+    },
+    {
+      "epoch": 0.5232303493817594,
+      "grad_norm": 4.391587991132564,
+      "learning_rate": 5.46623255864957e-07,
+      "loss": 1.7002,
+      "step": 7765
+    },
+    {
+      "epoch": 0.5235672652538661,
+      "grad_norm": 4.251022723910024,
+      "learning_rate": 5.460377369141345e-07,
+      "loss": 1.7482,
+      "step": 7770
+    },
+    {
+      "epoch": 0.5239041811259728,
+      "grad_norm": 4.796986297299885,
+      "learning_rate": 5.454521542836351e-07,
+      "loss": 1.6758,
+      "step": 7775
+    },
+    {
+      "epoch": 0.5242410969980795,
+      "grad_norm": 4.9104458316986355,
+      "learning_rate": 5.448665087834405e-07,
+      "loss": 1.6565,
+      "step": 7780
+    },
+    {
+      "epoch": 0.5245780128701863,
+      "grad_norm": 4.873279146811863,
+      "learning_rate": 5.442808012236192e-07,
+      "loss": 1.6767,
+      "step": 7785
+    },
+    {
+      "epoch": 0.5249149287422931,
+      "grad_norm": 4.731997781914198,
+      "learning_rate": 5.436950324143251e-07,
+      "loss": 1.7282,
+      "step": 7790
+    },
+    {
+      "epoch": 0.5252518446143998,
+      "grad_norm": 4.5914824351135435,
+      "learning_rate": 5.431092031657973e-07,
+      "loss": 1.6881,
+      "step": 7795
+    },
+    {
+      "epoch": 0.5255887604865065,
+      "grad_norm": 4.370668534964779,
+      "learning_rate": 5.425233142883585e-07,
+      "loss": 1.7491,
+      "step": 7800
+    },
+    {
+      "epoch": 0.5259256763586132,
+      "grad_norm": 4.46181656752716,
+      "learning_rate": 5.419373665924136e-07,
+      "loss": 1.6698,
+      "step": 7805
+    },
+    {
+      "epoch": 0.52626259223072,
+      "grad_norm": 4.164236490813964,
+      "learning_rate": 5.413513608884491e-07,
+      "loss": 1.728,
+      "step": 7810
+    },
+    {
+      "epoch": 0.5265995081028267,
+      "grad_norm": 4.507873534552062,
+      "learning_rate": 5.407652979870315e-07,
+      "loss": 1.6963,
+      "step": 7815
+    },
+    {
+      "epoch": 0.5269364239749335,
+      "grad_norm": 4.677386986638094,
+      "learning_rate": 5.401791786988068e-07,
+      "loss": 1.7004,
+      "step": 7820
+    },
+    {
+      "epoch": 0.5272733398470402,
+      "grad_norm": 4.284763421389914,
+      "learning_rate": 5.395930038344986e-07,
+      "loss": 1.7088,
+      "step": 7825
+    },
+    {
+      "epoch": 0.5276102557191469,
+      "grad_norm": 4.470418550681959,
+      "learning_rate": 5.390067742049073e-07,
+      "loss": 1.6565,
+      "step": 7830
+    },
+    {
+      "epoch": 0.5279471715912537,
+      "grad_norm": 4.89005652035299,
+      "learning_rate": 5.384204906209097e-07,
+      "loss": 1.633,
+      "step": 7835
+    },
+    {
+      "epoch": 0.5282840874633604,
+      "grad_norm": 4.277901423918713,
+      "learning_rate": 5.378341538934566e-07,
+      "loss": 1.6048,
+      "step": 7840
+    },
+    {
+      "epoch": 0.5286210033354671,
+      "grad_norm": 4.935713002066088,
+      "learning_rate": 5.372477648335725e-07,
+      "loss": 1.7168,
+      "step": 7845
+    },
+    {
+      "epoch": 0.5289579192075738,
+      "grad_norm": 4.594470771134848,
+      "learning_rate": 5.366613242523544e-07,
+      "loss": 1.6782,
+      "step": 7850
+    },
+    {
+      "epoch": 0.5292948350796806,
+      "grad_norm": 4.470432848589744,
+      "learning_rate": 5.360748329609702e-07,
+      "loss": 1.7082,
+      "step": 7855
+    },
+    {
+      "epoch": 0.5296317509517874,
+      "grad_norm": 4.58838789888362,
+      "learning_rate": 5.354882917706586e-07,
+      "loss": 1.74,
+      "step": 7860
+    },
+    {
+      "epoch": 0.5299686668238941,
+      "grad_norm": 4.931385031949068,
+      "learning_rate": 5.349017014927267e-07,
+      "loss": 1.76,
+      "step": 7865
+    },
+    {
+      "epoch": 0.5303055826960008,
+      "grad_norm": 4.316385158985576,
+      "learning_rate": 5.343150629385496e-07,
+      "loss": 1.6984,
+      "step": 7870
+    },
+    {
+      "epoch": 0.5306424985681075,
+      "grad_norm": 4.95603801804168,
+      "learning_rate": 5.337283769195696e-07,
+      "loss": 1.5816,
+      "step": 7875
+    },
+    {
+      "epoch": 0.5309794144402142,
+      "grad_norm": 4.63138953237527,
+      "learning_rate": 5.331416442472941e-07,
+      "loss": 1.6088,
+      "step": 7880
+    },
+    {
+      "epoch": 0.5313163303123211,
+      "grad_norm": 5.136600945477254,
+      "learning_rate": 5.325548657332956e-07,
+      "loss": 1.6416,
+      "step": 7885
+    },
+    {
+      "epoch": 0.5316532461844278,
+      "grad_norm": 4.597913163170623,
+      "learning_rate": 5.319680421892095e-07,
+      "loss": 1.629,
+      "step": 7890
+    },
+    {
+      "epoch": 0.5319901620565345,
+      "grad_norm": 4.912203001671273,
+      "learning_rate": 5.313811744267336e-07,
+      "loss": 1.6955,
+      "step": 7895
+    },
+    {
+      "epoch": 0.5323270779286412,
+      "grad_norm": 4.898778418458837,
+      "learning_rate": 5.30794263257627e-07,
+      "loss": 1.6931,
+      "step": 7900
+    },
+    {
+      "epoch": 0.5326639938007479,
+      "grad_norm": 4.712869973378353,
+      "learning_rate": 5.302073094937089e-07,
+      "loss": 1.6538,
+      "step": 7905
+    },
+    {
+      "epoch": 0.5330009096728547,
+      "grad_norm": 4.631259210071004,
+      "learning_rate": 5.296203139468571e-07,
+      "loss": 1.6861,
+      "step": 7910
+    },
+    {
+      "epoch": 0.5333378255449615,
+      "grad_norm": 4.547471371920815,
+      "learning_rate": 5.290332774290077e-07,
+      "loss": 1.6822,
+      "step": 7915
+    },
+    {
+      "epoch": 0.5336747414170682,
+      "grad_norm": 4.628705770790463,
+      "learning_rate": 5.284462007521528e-07,
+      "loss": 1.7231,
+      "step": 7920
+    },
+    {
+      "epoch": 0.5340116572891749,
+      "grad_norm": 4.378755980314644,
+      "learning_rate": 5.278590847283407e-07,
+      "loss": 1.6724,
+      "step": 7925
+    },
+    {
+      "epoch": 0.5343485731612816,
+      "grad_norm": 4.620965174090904,
+      "learning_rate": 5.27271930169674e-07,
+      "loss": 1.6702,
+      "step": 7930
+    },
+    {
+      "epoch": 0.5346854890333884,
+      "grad_norm": 4.666805516823848,
+      "learning_rate": 5.266847378883079e-07,
+      "loss": 1.6935,
+      "step": 7935
+    },
+    {
+      "epoch": 0.5350224049054951,
+      "grad_norm": 4.800429109529865,
+      "learning_rate": 5.260975086964507e-07,
+      "loss": 1.7228,
+      "step": 7940
+    },
+    {
+      "epoch": 0.5353593207776018,
+      "grad_norm": 4.732114405330755,
+      "learning_rate": 5.255102434063612e-07,
+      "loss": 1.6925,
+      "step": 7945
+    },
+    {
+      "epoch": 0.5356962366497086,
+      "grad_norm": 4.934511610082521,
+      "learning_rate": 5.249229428303486e-07,
+      "loss": 1.7302,
+      "step": 7950
+    },
+    {
+      "epoch": 0.5360331525218153,
+      "grad_norm": 4.691119337420431,
+      "learning_rate": 5.243356077807704e-07,
+      "loss": 1.7188,
+      "step": 7955
+    },
+    {
+      "epoch": 0.536370068393922,
+      "grad_norm": 4.210217220737946,
+      "learning_rate": 5.237482390700319e-07,
+      "loss": 1.7517,
+      "step": 7960
+    },
+    {
+      "epoch": 0.5367069842660288,
+      "grad_norm": 4.463884209035557,
+      "learning_rate": 5.231608375105852e-07,
+      "loss": 1.6262,
+      "step": 7965
+    },
+    {
+      "epoch": 0.5370439001381355,
+      "grad_norm": 4.718179447099045,
+      "learning_rate": 5.225734039149277e-07,
+      "loss": 1.6842,
+      "step": 7970
+    },
+    {
+      "epoch": 0.5373808160102422,
+      "grad_norm": 4.269434927425726,
+      "learning_rate": 5.219859390956012e-07,
+      "loss": 1.6166,
+      "step": 7975
+    },
+    {
+      "epoch": 0.537717731882349,
+      "grad_norm": 4.675892807792239,
+      "learning_rate": 5.213984438651904e-07,
+      "loss": 1.671,
+      "step": 7980
+    },
+    {
+      "epoch": 0.5380546477544557,
+      "grad_norm": 5.157607246722674,
+      "learning_rate": 5.208109190363222e-07,
+      "loss": 1.7264,
+      "step": 7985
+    },
+    {
+      "epoch": 0.5383915636265625,
+      "grad_norm": 4.74183668799578,
+      "learning_rate": 5.202233654216649e-07,
+      "loss": 1.6967,
+      "step": 7990
+    },
+    {
+      "epoch": 0.5387284794986692,
+      "grad_norm": 5.059820409448004,
+      "learning_rate": 5.196357838339259e-07,
+      "loss": 1.6226,
+      "step": 7995
+    },
+    {
+      "epoch": 0.5390653953707759,
+      "grad_norm": 4.353630566773812,
+      "learning_rate": 5.190481750858516e-07,
+      "loss": 1.6884,
+      "step": 8000
+    },
+    {
+      "epoch": 0.5394023112428826,
+      "grad_norm": 4.820195518964003,
+      "learning_rate": 5.184605399902262e-07,
+      "loss": 1.6845,
+      "step": 8005
+    },
+    {
+      "epoch": 0.5397392271149893,
+      "grad_norm": 4.450052549010921,
+      "learning_rate": 5.178728793598699e-07,
+      "loss": 1.7251,
+      "step": 8010
+    },
+    {
+      "epoch": 0.5400761429870962,
+      "grad_norm": 4.7654340761853575,
+      "learning_rate": 5.172851940076387e-07,
+      "loss": 1.6985,
+      "step": 8015
+    },
+    {
+      "epoch": 0.5404130588592029,
+      "grad_norm": 4.683470373788527,
+      "learning_rate": 5.166974847464223e-07,
+      "loss": 1.7706,
+      "step": 8020
+    },
+    {
+      "epoch": 0.5407499747313096,
+      "grad_norm": 4.824119687454772,
+      "learning_rate": 5.161097523891437e-07,
+      "loss": 1.7028,
+      "step": 8025
+    },
+    {
+      "epoch": 0.5410868906034163,
+      "grad_norm": 4.516681596644095,
+      "learning_rate": 5.15521997748758e-07,
+      "loss": 1.7156,
+      "step": 8030
+    },
+    {
+      "epoch": 0.541423806475523,
+      "grad_norm": 4.349696412018392,
+      "learning_rate": 5.149342216382511e-07,
+      "loss": 1.6188,
+      "step": 8035
+    },
+    {
+      "epoch": 0.5417607223476298,
+      "grad_norm": 4.442839169291913,
+      "learning_rate": 5.143464248706381e-07,
+      "loss": 1.6777,
+      "step": 8040
+    },
+    {
+      "epoch": 0.5420976382197366,
+      "grad_norm": 4.842887030499734,
+      "learning_rate": 5.137586082589633e-07,
+      "loss": 1.6884,
+      "step": 8045
+    },
+    {
+      "epoch": 0.5424345540918433,
+      "grad_norm": 5.260125717485399,
+      "learning_rate": 5.131707726162983e-07,
+      "loss": 1.7046,
+      "step": 8050
+    },
+    {
+      "epoch": 0.54277146996395,
+      "grad_norm": 4.4292959246006305,
+      "learning_rate": 5.125829187557406e-07,
+      "loss": 1.6895,
+      "step": 8055
+    },
+    {
+      "epoch": 0.5431083858360567,
+      "grad_norm": 4.398986244582595,
+      "learning_rate": 5.119950474904137e-07,
+      "loss": 1.7533,
+      "step": 8060
+    },
+    {
+      "epoch": 0.5434453017081635,
+      "grad_norm": 4.294749638256764,
+      "learning_rate": 5.114071596334642e-07,
+      "loss": 1.6337,
+      "step": 8065
+    },
+    {
+      "epoch": 0.5437822175802702,
+      "grad_norm": 4.678822056409349,
+      "learning_rate": 5.108192559980623e-07,
+      "loss": 1.7385,
+      "step": 8070
+    },
+    {
+      "epoch": 0.544119133452377,
+      "grad_norm": 4.549614415148617,
+      "learning_rate": 5.102313373974e-07,
+      "loss": 1.7601,
+      "step": 8075
+    },
+    {
+      "epoch": 0.5444560493244837,
+      "grad_norm": 4.64005770931557,
+      "learning_rate": 5.096434046446898e-07,
+      "loss": 1.6658,
+      "step": 8080
+    },
+    {
+      "epoch": 0.5447929651965904,
+      "grad_norm": 4.0286388311888,
+      "learning_rate": 5.090554585531639e-07,
+      "loss": 1.6677,
+      "step": 8085
+    },
+    {
+      "epoch": 0.5451298810686972,
+      "grad_norm": 4.634030770095132,
+      "learning_rate": 5.084674999360729e-07,
+      "loss": 1.6842,
+      "step": 8090
+    },
+    {
+      "epoch": 0.5454667969408039,
+      "grad_norm": 4.507452689503554,
+      "learning_rate": 5.078795296066846e-07,
+      "loss": 1.7786,
+      "step": 8095
+    },
+    {
+      "epoch": 0.5458037128129106,
+      "grad_norm": 4.50967203132743,
+      "learning_rate": 5.072915483782833e-07,
+      "loss": 1.5791,
+      "step": 8100
+    },
+    {
+      "epoch": 0.5461406286850173,
+      "grad_norm": 4.401152112417684,
+      "learning_rate": 5.067035570641678e-07,
+      "loss": 1.7083,
+      "step": 8105
+    },
+    {
+      "epoch": 0.546477544557124,
+      "grad_norm": 4.680901428735062,
+      "learning_rate": 5.061155564776517e-07,
+      "loss": 1.6894,
+      "step": 8110
+    },
+    {
+      "epoch": 0.5468144604292309,
+      "grad_norm": 4.297623735657957,
+      "learning_rate": 5.055275474320609e-07,
+      "loss": 1.7282,
+      "step": 8115
+    },
+    {
+      "epoch": 0.5471513763013376,
+      "grad_norm": 4.652741505309037,
+      "learning_rate": 5.049395307407328e-07,
+      "loss": 1.7205,
+      "step": 8120
+    },
+    {
+      "epoch": 0.5474882921734443,
+      "grad_norm": 4.396575176637696,
+      "learning_rate": 5.04351507217016e-07,
+      "loss": 1.6343,
+      "step": 8125
+    },
+    {
+      "epoch": 0.547825208045551,
+      "grad_norm": 5.067335330992528,
+      "learning_rate": 5.03763477674268e-07,
+      "loss": 1.66,
+      "step": 8130
+    },
+    {
+      "epoch": 0.5481621239176577,
+      "grad_norm": 4.764149280006885,
+      "learning_rate": 5.031754429258549e-07,
+      "loss": 1.7649,
+      "step": 8135
+    },
+    {
+      "epoch": 0.5484990397897644,
+      "grad_norm": 4.689365821442324,
+      "learning_rate": 5.025874037851499e-07,
+      "loss": 1.6931,
+      "step": 8140
+    },
+    {
+      "epoch": 0.5488359556618713,
+      "grad_norm": 4.6085486128563184,
+      "learning_rate": 5.019993610655322e-07,
+      "loss": 1.6921,
+      "step": 8145
+    },
+    {
+      "epoch": 0.549172871533978,
+      "grad_norm": 4.419580714168477,
+      "learning_rate": 5.014113155803863e-07,
+      "loss": 1.7069,
+      "step": 8150
+    },
+    {
+      "epoch": 0.5495097874060847,
+      "grad_norm": 4.479763083233425,
+      "learning_rate": 5.008232681430999e-07,
+      "loss": 1.7082,
+      "step": 8155
+    },
+    {
+      "epoch": 0.5498467032781914,
+      "grad_norm": 4.729296163432992,
+      "learning_rate": 5.002352195670643e-07,
+      "loss": 1.5793,
+      "step": 8160
+    },
+    {
+      "epoch": 0.5501836191502981,
+      "grad_norm": 4.920600762682583,
+      "learning_rate": 4.996471706656715e-07,
+      "loss": 1.6859,
+      "step": 8165
+    },
+    {
+      "epoch": 0.5505205350224049,
+      "grad_norm": 4.949329604420274,
+      "learning_rate": 4.990591222523142e-07,
+      "loss": 1.7081,
+      "step": 8170
+    },
+    {
+      "epoch": 0.5508574508945117,
+      "grad_norm": 4.1980818218131555,
+      "learning_rate": 4.984710751403849e-07,
+      "loss": 1.6864,
+      "step": 8175
+    },
+    {
+      "epoch": 0.5511943667666184,
+      "grad_norm": 4.418385051349536,
+      "learning_rate": 4.978830301432738e-07,
+      "loss": 1.7467,
+      "step": 8180
+    },
+    {
+      "epoch": 0.5515312826387251,
+      "grad_norm": 4.73152418477475,
+      "learning_rate": 4.97294988074368e-07,
+      "loss": 1.7329,
+      "step": 8185
+    },
+    {
+      "epoch": 0.5518681985108318,
+      "grad_norm": 4.71362442505742,
+      "learning_rate": 4.96706949747051e-07,
+      "loss": 1.6406,
+      "step": 8190
+    },
+    {
+      "epoch": 0.5522051143829386,
+      "grad_norm": 4.487306736267114,
+      "learning_rate": 4.961189159747015e-07,
+      "loss": 1.6963,
+      "step": 8195
+    },
+    {
+      "epoch": 0.5525420302550453,
+      "grad_norm": 4.8689064749222375,
+      "learning_rate": 4.955308875706905e-07,
+      "loss": 1.7162,
+      "step": 8200
+    },
+    {
+      "epoch": 0.552878946127152,
+      "grad_norm": 4.954032629608422,
+      "learning_rate": 4.94942865348383e-07,
+      "loss": 1.6799,
+      "step": 8205
+    },
+    {
+      "epoch": 0.5532158619992588,
+      "grad_norm": 4.609550126338583,
+      "learning_rate": 4.943548501211351e-07,
+      "loss": 1.6984,
+      "step": 8210
+    },
+    {
+      "epoch": 0.5535527778713655,
+      "grad_norm": 4.592731793929496,
+      "learning_rate": 4.937668427022924e-07,
+      "loss": 1.6618,
+      "step": 8215
+    },
+    {
+      "epoch": 0.5538896937434723,
+      "grad_norm": 4.612954517966647,
+      "learning_rate": 4.931788439051909e-07,
+      "loss": 1.6494,
+      "step": 8220
+    },
+    {
+      "epoch": 0.554226609615579,
+      "grad_norm": 4.781552723452898,
+      "learning_rate": 4.925908545431537e-07,
+      "loss": 1.6531,
+      "step": 8225
+    },
+    {
+      "epoch": 0.5545635254876857,
+      "grad_norm": 4.96982515502249,
+      "learning_rate": 4.920028754294915e-07,
+      "loss": 1.6884,
+      "step": 8230
+    },
+    {
+      "epoch": 0.5549004413597924,
+      "grad_norm": 4.909103884057576,
+      "learning_rate": 4.914149073775003e-07,
+      "loss": 1.7128,
+      "step": 8235
+    },
+    {
+      "epoch": 0.5552373572318992,
+      "grad_norm": 4.559122268469247,
+      "learning_rate": 4.908269512004613e-07,
+      "loss": 1.6685,
+      "step": 8240
+    },
+    {
+      "epoch": 0.555574273104006,
+      "grad_norm": 4.7935399767154925,
+      "learning_rate": 4.902390077116392e-07,
+      "loss": 1.6834,
+      "step": 8245
+    },
+    {
+      "epoch": 0.5559111889761127,
+      "grad_norm": 4.675577266111019,
+      "learning_rate": 4.896510777242805e-07,
+      "loss": 1.6316,
+      "step": 8250
+    },
+    {
+      "epoch": 0.5562481048482194,
+      "grad_norm": 4.75778670778229,
+      "learning_rate": 4.890631620516141e-07,
+      "loss": 1.6933,
+      "step": 8255
+    },
+    {
+      "epoch": 0.5565850207203261,
+      "grad_norm": 4.88551208195117,
+      "learning_rate": 4.88475261506848e-07,
+      "loss": 1.6561,
+      "step": 8260
+    },
+    {
+      "epoch": 0.5569219365924328,
+      "grad_norm": 4.661130950299994,
+      "learning_rate": 4.878873769031702e-07,
+      "loss": 1.7364,
+      "step": 8265
+    },
+    {
+      "epoch": 0.5572588524645397,
+      "grad_norm": 4.422533661936322,
+      "learning_rate": 4.872995090537459e-07,
+      "loss": 1.6691,
+      "step": 8270
+    },
+    {
+      "epoch": 0.5575957683366464,
+      "grad_norm": 4.661440808148939,
+      "learning_rate": 4.867116587717179e-07,
+      "loss": 1.7266,
+      "step": 8275
+    },
+    {
+      "epoch": 0.5579326842087531,
+      "grad_norm": 4.239501455891196,
+      "learning_rate": 4.861238268702039e-07,
+      "loss": 1.6937,
+      "step": 8280
+    },
+    {
+      "epoch": 0.5582696000808598,
+      "grad_norm": 5.038883363121014,
+      "learning_rate": 4.855360141622965e-07,
+      "loss": 1.723,
+      "step": 8285
+    },
+    {
+      "epoch": 0.5586065159529665,
+      "grad_norm": 4.262685657903031,
+      "learning_rate": 4.849482214610623e-07,
+      "loss": 1.6939,
+      "step": 8290
+    },
+    {
+      "epoch": 0.5589434318250733,
+      "grad_norm": 4.399744779227197,
+      "learning_rate": 4.843604495795392e-07,
+      "loss": 1.7151,
+      "step": 8295
+    },
+    {
+      "epoch": 0.55928034769718,
+      "grad_norm": 4.703118750408812,
+      "learning_rate": 4.83772699330737e-07,
+      "loss": 1.6589,
+      "step": 8300
+    },
+    {
+      "epoch": 0.5596172635692868,
+      "grad_norm": 4.650840967246696,
+      "learning_rate": 4.831849715276355e-07,
+      "loss": 1.6736,
+      "step": 8305
+    },
+    {
+      "epoch": 0.5599541794413935,
+      "grad_norm": 4.992067739030684,
+      "learning_rate": 4.825972669831834e-07,
+      "loss": 1.6443,
+      "step": 8310
+    },
+    {
+      "epoch": 0.5602910953135002,
+      "grad_norm": 4.656959268996734,
+      "learning_rate": 4.82009586510297e-07,
+      "loss": 1.6643,
+      "step": 8315
+    },
+    {
+      "epoch": 0.5606280111856069,
+      "grad_norm": 4.581554977226479,
+      "learning_rate": 4.814219309218594e-07,
+      "loss": 1.6773,
+      "step": 8320
+    },
+    {
+      "epoch": 0.5609649270577137,
+      "grad_norm": 4.290945719182593,
+      "learning_rate": 4.808343010307199e-07,
+      "loss": 1.7589,
+      "step": 8325
+    },
+    {
+      "epoch": 0.5613018429298204,
+      "grad_norm": 4.7021818808486255,
+      "learning_rate": 4.802466976496911e-07,
+      "loss": 1.7013,
+      "step": 8330
+    },
+    {
+      "epoch": 0.5616387588019272,
+      "grad_norm": 4.853582184853659,
+      "learning_rate": 4.796591215915498e-07,
+      "loss": 1.7118,
+      "step": 8335
+    },
+    {
+      "epoch": 0.5619756746740339,
+      "grad_norm": 4.591796496584383,
+      "learning_rate": 4.79071573669035e-07,
+      "loss": 1.7245,
+      "step": 8340
+    },
+    {
+      "epoch": 0.5623125905461406,
+      "grad_norm": 4.312408889428106,
+      "learning_rate": 4.784840546948463e-07,
+      "loss": 1.6718,
+      "step": 8345
+    },
+    {
+      "epoch": 0.5626495064182474,
+      "grad_norm": 4.678194520576363,
+      "learning_rate": 4.778965654816435e-07,
+      "loss": 1.6318,
+      "step": 8350
+    },
+    {
+      "epoch": 0.5629864222903541,
+      "grad_norm": 4.630467160745489,
+      "learning_rate": 4.773091068420455e-07,
+      "loss": 1.7436,
+      "step": 8355
+    },
+    {
+      "epoch": 0.5633233381624608,
+      "grad_norm": 4.541776089598689,
+      "learning_rate": 4.767216795886281e-07,
+      "loss": 1.7184,
+      "step": 8360
+    },
+    {
+      "epoch": 0.5636602540345675,
+      "grad_norm": 4.467363162187823,
+      "learning_rate": 4.761342845339246e-07,
+      "loss": 1.6953,
+      "step": 8365
+    },
+    {
+      "epoch": 0.5639971699066743,
+      "grad_norm": 5.089902203731988,
+      "learning_rate": 4.7554692249042345e-07,
+      "loss": 1.7457,
+      "step": 8370
+    },
+    {
+      "epoch": 0.5643340857787811,
+      "grad_norm": 4.488762450890132,
+      "learning_rate": 4.7495959427056754e-07,
+      "loss": 1.7263,
+      "step": 8375
+    },
+    {
+      "epoch": 0.5646710016508878,
+      "grad_norm": 4.267196572613947,
+      "learning_rate": 4.743723006867523e-07,
+      "loss": 1.6892,
+      "step": 8380
+    },
+    {
+      "epoch": 0.5650079175229945,
+      "grad_norm": 4.471832534053066,
+      "learning_rate": 4.737850425513263e-07,
+      "loss": 1.699,
+      "step": 8385
+    },
+    {
+      "epoch": 0.5653448333951012,
+      "grad_norm": 4.32195057372782,
+      "learning_rate": 4.731978206765884e-07,
+      "loss": 1.69,
+      "step": 8390
+    },
+    {
+      "epoch": 0.5656817492672079,
+      "grad_norm": 4.772982759500117,
+      "learning_rate": 4.726106358747871e-07,
+      "loss": 1.7185,
+      "step": 8395
+    },
+    {
+      "epoch": 0.5660186651393148,
+      "grad_norm": 4.506780870023458,
+      "learning_rate": 4.720234889581203e-07,
+      "loss": 1.7407,
+      "step": 8400
+    },
+    {
+      "epoch": 0.5663555810114215,
+      "grad_norm": 4.775965979267713,
+      "learning_rate": 4.714363807387333e-07,
+      "loss": 1.7065,
+      "step": 8405
+    },
+    {
+      "epoch": 0.5666924968835282,
+      "grad_norm": 4.543574977359184,
+      "learning_rate": 4.708493120287175e-07,
+      "loss": 1.715,
+      "step": 8410
+    },
+    {
+      "epoch": 0.5670294127556349,
+      "grad_norm": 4.216614068049089,
+      "learning_rate": 4.7026228364010984e-07,
+      "loss": 1.6877,
+      "step": 8415
+    },
+    {
+      "epoch": 0.5673663286277416,
+      "grad_norm": 4.571120024348289,
+      "learning_rate": 4.69675296384892e-07,
+      "loss": 1.7409,
+      "step": 8420
+    },
+    {
+      "epoch": 0.5677032444998484,
+      "grad_norm": 4.452168437386814,
+      "learning_rate": 4.6908835107498775e-07,
+      "loss": 1.7086,
+      "step": 8425
+    },
+    {
+      "epoch": 0.5680401603719551,
+      "grad_norm": 4.496686032131901,
+      "learning_rate": 4.685014485222637e-07,
+      "loss": 1.6897,
+      "step": 8430
+    },
+    {
+      "epoch": 0.5683770762440619,
+      "grad_norm": 4.513536089655021,
+      "learning_rate": 4.679145895385269e-07,
+      "loss": 1.6941,
+      "step": 8435
+    },
+    {
+      "epoch": 0.5687139921161686,
+      "grad_norm": 4.919282730735273,
+      "learning_rate": 4.673277749355245e-07,
+      "loss": 1.6058,
+      "step": 8440
+    },
+    {
+      "epoch": 0.5690509079882753,
+      "grad_norm": 5.0241144585906685,
+      "learning_rate": 4.667410055249417e-07,
+      "loss": 1.6192,
+      "step": 8445
+    },
+    {
+      "epoch": 0.5693878238603821,
+      "grad_norm": 4.825190139086496,
+      "learning_rate": 4.6615428211840154e-07,
+      "loss": 1.6666,
+      "step": 8450
+    },
+    {
+      "epoch": 0.5697247397324888,
+      "grad_norm": 4.567783140676474,
+      "learning_rate": 4.655676055274637e-07,
+      "loss": 1.6983,
+      "step": 8455
+    },
+    {
+      "epoch": 0.5700616556045955,
+      "grad_norm": 4.82863687523337,
+      "learning_rate": 4.6498097656362247e-07,
+      "loss": 1.7003,
+      "step": 8460
+    },
+    {
+      "epoch": 0.5703985714767023,
+      "grad_norm": 4.7561387266728,
+      "learning_rate": 4.643943960383067e-07,
+      "loss": 1.5966,
+      "step": 8465
+    },
+    {
+      "epoch": 0.570735487348809,
+      "grad_norm": 4.680484457748234,
+      "learning_rate": 4.638078647628782e-07,
+      "loss": 1.802,
+      "step": 8470
+    },
+    {
+      "epoch": 0.5710724032209158,
+      "grad_norm": 4.6946401693960125,
+      "learning_rate": 4.632213835486305e-07,
+      "loss": 1.7394,
+      "step": 8475
+    },
+    {
+      "epoch": 0.5714093190930225,
+      "grad_norm": 4.540704185339018,
+      "learning_rate": 4.626349532067879e-07,
+      "loss": 1.7039,
+      "step": 8480
+    },
+    {
+      "epoch": 0.5717462349651292,
+      "grad_norm": 4.396382603367253,
+      "learning_rate": 4.620485745485046e-07,
+      "loss": 1.6342,
+      "step": 8485
+    },
+    {
+      "epoch": 0.5720831508372359,
+      "grad_norm": 4.650468869736771,
+      "learning_rate": 4.6146224838486287e-07,
+      "loss": 1.7561,
+      "step": 8490
+    },
+    {
+      "epoch": 0.5724200667093426,
+      "grad_norm": 4.3725778710452845,
+      "learning_rate": 4.6087597552687275e-07,
+      "loss": 1.7229,
+      "step": 8495
+    },
+    {
+      "epoch": 0.5727569825814494,
+      "grad_norm": 7.127703602208947,
+      "learning_rate": 4.602897567854705e-07,
+      "loss": 1.6767,
+      "step": 8500
+    },
+    {
+      "epoch": 0.5730938984535562,
+      "grad_norm": 4.749927884410711,
+      "learning_rate": 4.5970359297151733e-07,
+      "loss": 1.6543,
+      "step": 8505
+    },
+    {
+      "epoch": 0.5734308143256629,
+      "grad_norm": 4.559570945610894,
+      "learning_rate": 4.591174848957986e-07,
+      "loss": 1.7265,
+      "step": 8510
+    },
+    {
+      "epoch": 0.5737677301977696,
+      "grad_norm": 4.5939906488538895,
+      "learning_rate": 4.585314333690224e-07,
+      "loss": 1.7119,
+      "step": 8515
+    },
+    {
+      "epoch": 0.5741046460698763,
+      "grad_norm": 4.249421774835835,
+      "learning_rate": 4.579454392018192e-07,
+      "loss": 1.6535,
+      "step": 8520
+    },
+    {
+      "epoch": 0.574441561941983,
+      "grad_norm": 4.109291213056605,
+      "learning_rate": 4.5735950320473915e-07,
+      "loss": 1.6659,
+      "step": 8525
+    },
+    {
+      "epoch": 0.5747784778140899,
+      "grad_norm": 4.618222010138749,
+      "learning_rate": 4.5677362618825265e-07,
+      "loss": 1.6468,
+      "step": 8530
+    },
+    {
+      "epoch": 0.5751153936861966,
+      "grad_norm": 4.612200159477145,
+      "learning_rate": 4.5618780896274866e-07,
+      "loss": 1.6851,
+      "step": 8535
+    },
+    {
+      "epoch": 0.5754523095583033,
+      "grad_norm": 5.003053852009229,
+      "learning_rate": 4.556020523385326e-07,
+      "loss": 1.6119,
+      "step": 8540
+    },
+    {
+      "epoch": 0.57578922543041,
+      "grad_norm": 4.6683042791197575,
+      "learning_rate": 4.55016357125827e-07,
+      "loss": 1.6521,
+      "step": 8545
+    },
+    {
+      "epoch": 0.5761261413025167,
+      "grad_norm": 4.494709104565061,
+      "learning_rate": 4.5443072413476877e-07,
+      "loss": 1.6743,
+      "step": 8550
+    },
+    {
+      "epoch": 0.5764630571746235,
+      "grad_norm": 4.874792839672999,
+      "learning_rate": 4.5384515417540914e-07,
+      "loss": 1.6915,
+      "step": 8555
+    },
+    {
+      "epoch": 0.5767999730467303,
+      "grad_norm": 4.481805781018169,
+      "learning_rate": 4.5325964805771187e-07,
+      "loss": 1.685,
+      "step": 8560
+    },
+    {
+      "epoch": 0.577136888918837,
+      "grad_norm": 4.5394444499066715,
+      "learning_rate": 4.526742065915528e-07,
+      "loss": 1.6167,
+      "step": 8565
+    },
+    {
+      "epoch": 0.5774738047909437,
+      "grad_norm": 4.3024721598181594,
+      "learning_rate": 4.520888305867181e-07,
+      "loss": 1.647,
+      "step": 8570
+    },
+    {
+      "epoch": 0.5778107206630504,
+      "grad_norm": 4.93852787215156,
+      "learning_rate": 4.5150352085290315e-07,
+      "loss": 1.7097,
+      "step": 8575
+    },
+    {
+      "epoch": 0.5781476365351572,
+      "grad_norm": 4.604154090081095,
+      "learning_rate": 4.5091827819971207e-07,
+      "loss": 1.7154,
+      "step": 8580
+    },
+    {
+      "epoch": 0.5784845524072639,
+      "grad_norm": 4.635206348927524,
+      "learning_rate": 4.503331034366563e-07,
+      "loss": 1.6914,
+      "step": 8585
+    },
+    {
+      "epoch": 0.5788214682793706,
+      "grad_norm": 4.462169939610559,
+      "learning_rate": 4.4974799737315274e-07,
+      "loss": 1.6473,
+      "step": 8590
+    },
+    {
+      "epoch": 0.5791583841514774,
+      "grad_norm": 4.404248007791692,
+      "learning_rate": 4.491629608185237e-07,
+      "loss": 1.6687,
+      "step": 8595
+    },
+    {
+      "epoch": 0.5794953000235841,
+      "grad_norm": 4.417246419975575,
+      "learning_rate": 4.485779945819956e-07,
+      "loss": 1.6734,
+      "step": 8600
+    },
+    {
+      "epoch": 0.5798322158956909,
+      "grad_norm": 4.587835613701123,
+      "learning_rate": 4.479930994726968e-07,
+      "loss": 1.6319,
+      "step": 8605
+    },
+    {
+      "epoch": 0.5801691317677976,
+      "grad_norm": 4.543488439399577,
+      "learning_rate": 4.474082762996581e-07,
+      "loss": 1.7323,
+      "step": 8610
+    },
+    {
+      "epoch": 0.5805060476399043,
+      "grad_norm": 4.543906159067551,
+      "learning_rate": 4.468235258718105e-07,
+      "loss": 1.7344,
+      "step": 8615
+    },
+    {
+      "epoch": 0.580842963512011,
+      "grad_norm": 4.396940012204263,
+      "learning_rate": 4.4623884899798397e-07,
+      "loss": 1.6825,
+      "step": 8620
+    },
+    {
+      "epoch": 0.5811798793841177,
+      "grad_norm": 5.062533344351063,
+      "learning_rate": 4.4565424648690743e-07,
+      "loss": 1.6386,
+      "step": 8625
+    },
+    {
+      "epoch": 0.5815167952562246,
+      "grad_norm": 4.736125482927736,
+      "learning_rate": 4.450697191472067e-07,
+      "loss": 1.6547,
+      "step": 8630
+    },
+    {
+      "epoch": 0.5818537111283313,
+      "grad_norm": 4.425455347464461,
+      "learning_rate": 4.4448526778740327e-07,
+      "loss": 1.6442,
+      "step": 8635
+    },
+    {
+      "epoch": 0.582190627000438,
+      "grad_norm": 4.84707850623792,
+      "learning_rate": 4.439008932159138e-07,
+      "loss": 1.7129,
+      "step": 8640
+    },
+    {
+      "epoch": 0.5825275428725447,
+      "grad_norm": 4.67296322499853,
+      "learning_rate": 4.4331659624104876e-07,
+      "loss": 1.7726,
+      "step": 8645
+    },
+    {
+      "epoch": 0.5828644587446514,
+      "grad_norm": 4.778465340510698,
+      "learning_rate": 4.427323776710117e-07,
+      "loss": 1.7425,
+      "step": 8650
+    },
+    {
+      "epoch": 0.5832013746167583,
+      "grad_norm": 4.711550493290235,
+      "learning_rate": 4.4214823831389663e-07,
+      "loss": 1.6981,
+      "step": 8655
+    },
+    {
+      "epoch": 0.583538290488865,
+      "grad_norm": 4.512779690254907,
+      "learning_rate": 4.41564178977689e-07,
+      "loss": 1.6096,
+      "step": 8660
+    },
+    {
+      "epoch": 0.5838752063609717,
+      "grad_norm": 4.6872911357761184,
+      "learning_rate": 4.4098020047026343e-07,
+      "loss": 1.6494,
+      "step": 8665
+    },
+    {
+      "epoch": 0.5842121222330784,
+      "grad_norm": 4.547380845970744,
+      "learning_rate": 4.4039630359938194e-07,
+      "loss": 1.6356,
+      "step": 8670
+    },
+    {
+      "epoch": 0.5845490381051851,
+      "grad_norm": 4.4410239787703265,
+      "learning_rate": 4.3981248917269477e-07,
+      "loss": 1.7121,
+      "step": 8675
+    },
+    {
+      "epoch": 0.5848859539772918,
+      "grad_norm": 4.447517319316789,
+      "learning_rate": 4.3922875799773735e-07,
+      "loss": 1.6612,
+      "step": 8680
+    },
+    {
+      "epoch": 0.5852228698493986,
+      "grad_norm": 4.4385798637549,
+      "learning_rate": 4.386451108819302e-07,
+      "loss": 1.6545,
+      "step": 8685
+    },
+    {
+      "epoch": 0.5855597857215054,
+      "grad_norm": 4.926861478661365,
+      "learning_rate": 4.380615486325774e-07,
+      "loss": 1.6946,
+      "step": 8690
+    },
+    {
+      "epoch": 0.5858967015936121,
+      "grad_norm": 4.119226148156626,
+      "learning_rate": 4.3747807205686616e-07,
+      "loss": 1.6452,
+      "step": 8695
+    },
+    {
+      "epoch": 0.5862336174657188,
+      "grad_norm": 3.9868492954091796,
+      "learning_rate": 4.3689468196186433e-07,
+      "loss": 1.6384,
+      "step": 8700
+    },
+    {
+      "epoch": 0.5865705333378255,
+      "grad_norm": 4.26822813610702,
+      "learning_rate": 4.36311379154521e-07,
+      "loss": 1.7479,
+      "step": 8705
+    },
+    {
+      "epoch": 0.5869074492099323,
+      "grad_norm": 4.618293683972388,
+      "learning_rate": 4.3572816444166406e-07,
+      "loss": 1.7083,
+      "step": 8710
+    },
+    {
+      "epoch": 0.587244365082039,
+      "grad_norm": 5.183493055108485,
+      "learning_rate": 4.351450386299996e-07,
+      "loss": 1.6114,
+      "step": 8715
+    },
+    {
+      "epoch": 0.5875812809541457,
+      "grad_norm": 4.6830635864687755,
+      "learning_rate": 4.3456200252611075e-07,
+      "loss": 1.6875,
+      "step": 8720
+    },
+    {
+      "epoch": 0.5879181968262525,
+      "grad_norm": 4.48502779266541,
+      "learning_rate": 4.3397905693645653e-07,
+      "loss": 1.721,
+      "step": 8725
+    },
+    {
+      "epoch": 0.5882551126983592,
+      "grad_norm": 4.91326546399509,
+      "learning_rate": 4.3339620266737116e-07,
+      "loss": 1.6608,
+      "step": 8730
+    },
+    {
+      "epoch": 0.588592028570466,
+      "grad_norm": 4.653039821741647,
+      "learning_rate": 4.328134405250617e-07,
+      "loss": 1.6222,
+      "step": 8735
+    },
+    {
+      "epoch": 0.5889289444425727,
+      "grad_norm": 4.919960637616408,
+      "learning_rate": 4.322307713156085e-07,
+      "loss": 1.6197,
+      "step": 8740
+    },
+    {
+      "epoch": 0.5892658603146794,
+      "grad_norm": 4.813487307948052,
+      "learning_rate": 4.316481958449634e-07,
+      "loss": 1.6323,
+      "step": 8745
+    },
+    {
+      "epoch": 0.5896027761867861,
+      "grad_norm": 4.390267109347444,
+      "learning_rate": 4.310657149189478e-07,
+      "loss": 1.7268,
+      "step": 8750
+    },
+    {
+      "epoch": 0.5899396920588929,
+      "grad_norm": 4.6590548821479665,
+      "learning_rate": 4.3048332934325325e-07,
+      "loss": 1.7155,
+      "step": 8755
+    },
+    {
+      "epoch": 0.5902766079309997,
+      "grad_norm": 5.0419072706091645,
+      "learning_rate": 4.2990103992343893e-07,
+      "loss": 1.6464,
+      "step": 8760
+    },
+    {
+      "epoch": 0.5906135238031064,
+      "grad_norm": 4.331131220822677,
+      "learning_rate": 4.2931884746493107e-07,
+      "loss": 1.7478,
+      "step": 8765
+    },
+    {
+      "epoch": 0.5909504396752131,
+      "grad_norm": 4.689736914803806,
+      "learning_rate": 4.287367527730216e-07,
+      "loss": 1.6688,
+      "step": 8770
+    },
+    {
+      "epoch": 0.5912873555473198,
+      "grad_norm": 4.618094275967223,
+      "learning_rate": 4.2815475665286766e-07,
+      "loss": 1.7149,
+      "step": 8775
+    },
+    {
+      "epoch": 0.5916242714194265,
+      "grad_norm": 5.835306696527706,
+      "learning_rate": 4.2757285990948993e-07,
+      "loss": 1.6927,
+      "step": 8780
+    },
+    {
+      "epoch": 0.5919611872915334,
+      "grad_norm": 4.7646671532063465,
+      "learning_rate": 4.269910633477711e-07,
+      "loss": 1.6413,
+      "step": 8785
+    },
+    {
+      "epoch": 0.5922981031636401,
+      "grad_norm": 4.451157033474741,
+      "learning_rate": 4.264093677724561e-07,
+      "loss": 1.6888,
+      "step": 8790
+    },
+    {
+      "epoch": 0.5926350190357468,
+      "grad_norm": 4.365302012587232,
+      "learning_rate": 4.2582777398814966e-07,
+      "loss": 1.6304,
+      "step": 8795
+    },
+    {
+      "epoch": 0.5929719349078535,
+      "grad_norm": 4.360400169057322,
+      "learning_rate": 4.252462827993158e-07,
+      "loss": 1.6725,
+      "step": 8800
+    },
+    {
+      "epoch": 0.5933088507799602,
+      "grad_norm": 4.847639710618478,
+      "learning_rate": 4.246648950102765e-07,
+      "loss": 1.7565,
+      "step": 8805
+    },
+    {
+      "epoch": 0.593645766652067,
+      "grad_norm": 4.531584210029071,
+      "learning_rate": 4.240836114252112e-07,
+      "loss": 1.6825,
+      "step": 8810
+    },
+    {
+      "epoch": 0.5939826825241737,
+      "grad_norm": 4.446728074727535,
+      "learning_rate": 4.2350243284815445e-07,
+      "loss": 1.6699,
+      "step": 8815
+    },
+    {
+      "epoch": 0.5943195983962805,
+      "grad_norm": 4.557072399740796,
+      "learning_rate": 4.229213600829963e-07,
+      "loss": 1.7023,
+      "step": 8820
+    },
+    {
+      "epoch": 0.5946565142683872,
+      "grad_norm": 4.320685633324107,
+      "learning_rate": 4.223403939334802e-07,
+      "loss": 1.7022,
+      "step": 8825
+    },
+    {
+      "epoch": 0.5949934301404939,
+      "grad_norm": 5.091185808385919,
+      "learning_rate": 4.217595352032017e-07,
+      "loss": 1.717,
+      "step": 8830
+    },
+    {
+      "epoch": 0.5953303460126007,
+      "grad_norm": 4.689383125941451,
+      "learning_rate": 4.2117878469560834e-07,
+      "loss": 1.6521,
+      "step": 8835
+    },
+    {
+      "epoch": 0.5956672618847074,
+      "grad_norm": 4.2042503348182585,
+      "learning_rate": 4.205981432139978e-07,
+      "loss": 1.6621,
+      "step": 8840
+    },
+    {
+      "epoch": 0.5960041777568141,
+      "grad_norm": 4.5052957339457835,
+      "learning_rate": 4.200176115615169e-07,
+      "loss": 1.6988,
+      "step": 8845
+    },
+    {
+      "epoch": 0.5963410936289208,
+      "grad_norm": 4.6778001572913075,
+      "learning_rate": 4.1943719054116027e-07,
+      "loss": 1.647,
+      "step": 8850
+    },
+    {
+      "epoch": 0.5966780095010276,
+      "grad_norm": 4.868220790218576,
+      "learning_rate": 4.1885688095577e-07,
+      "loss": 1.6256,
+      "step": 8855
+    },
+    {
+      "epoch": 0.5970149253731343,
+      "grad_norm": 4.329463664628172,
+      "learning_rate": 4.182766836080339e-07,
+      "loss": 1.6238,
+      "step": 8860
+    },
+    {
+      "epoch": 0.5973518412452411,
+      "grad_norm": 4.70703333345465,
+      "learning_rate": 4.176965993004842e-07,
+      "loss": 1.6875,
+      "step": 8865
+    },
+    {
+      "epoch": 0.5976887571173478,
+      "grad_norm": 4.904360263585603,
+      "learning_rate": 4.171166288354971e-07,
+      "loss": 1.6849,
+      "step": 8870
+    },
+    {
+      "epoch": 0.5980256729894545,
+      "grad_norm": 4.32012699596512,
+      "learning_rate": 4.165367730152917e-07,
+      "loss": 1.6848,
+      "step": 8875
+    },
+    {
+      "epoch": 0.5983625888615612,
+      "grad_norm": 4.7755314086339,
+      "learning_rate": 4.1595703264192737e-07,
+      "loss": 1.6969,
+      "step": 8880
+    },
+    {
+      "epoch": 0.598699504733668,
+      "grad_norm": 4.404946565600434,
+      "learning_rate": 4.15377408517305e-07,
+      "loss": 1.6712,
+      "step": 8885
+    },
+    {
+      "epoch": 0.5990364206057748,
+      "grad_norm": 4.62456766911183,
+      "learning_rate": 4.147979014431642e-07,
+      "loss": 1.6782,
+      "step": 8890
+    },
+    {
+      "epoch": 0.5993733364778815,
+      "grad_norm": 5.036178958688704,
+      "learning_rate": 4.142185122210823e-07,
+      "loss": 1.6734,
+      "step": 8895
+    },
+    {
+      "epoch": 0.5997102523499882,
+      "grad_norm": 4.478025050021537,
+      "learning_rate": 4.136392416524742e-07,
+      "loss": 1.6853,
+      "step": 8900
+    },
+    {
+      "epoch": 0.6000471682220949,
+      "grad_norm": 4.802397490056363,
+      "learning_rate": 4.1306009053859043e-07,
+      "loss": 1.6727,
+      "step": 8905
+    },
+    {
+      "epoch": 0.6003840840942016,
+      "grad_norm": 4.824037986443692,
+      "learning_rate": 4.124810596805166e-07,
+      "loss": 1.6883,
+      "step": 8910
+    },
+    {
+      "epoch": 0.6007209999663085,
+      "grad_norm": 4.49654239932616,
+      "learning_rate": 4.119021498791712e-07,
+      "loss": 1.7514,
+      "step": 8915
+    },
+    {
+      "epoch": 0.6010579158384152,
+      "grad_norm": 4.118394632761788,
+      "learning_rate": 4.113233619353062e-07,
+      "loss": 1.6592,
+      "step": 8920
+    },
+    {
+      "epoch": 0.6013948317105219,
+      "grad_norm": 4.8909611172651415,
+      "learning_rate": 4.107446966495044e-07,
+      "loss": 1.7011,
+      "step": 8925
+    },
+    {
+      "epoch": 0.6017317475826286,
+      "grad_norm": 4.801420603074635,
+      "learning_rate": 4.101661548221792e-07,
+      "loss": 1.7113,
+      "step": 8930
+    },
+    {
+      "epoch": 0.6020686634547353,
+      "grad_norm": 4.893714227439976,
+      "learning_rate": 4.0958773725357297e-07,
+      "loss": 1.6746,
+      "step": 8935
+    },
+    {
+      "epoch": 0.6024055793268421,
+      "grad_norm": 4.6529537641686085,
+      "learning_rate": 4.0900944474375674e-07,
+      "loss": 1.6973,
+      "step": 8940
+    },
+    {
+      "epoch": 0.6027424951989488,
+      "grad_norm": 4.672275429926195,
+      "learning_rate": 4.084312780926279e-07,
+      "loss": 1.6491,
+      "step": 8945
+    },
+    {
+      "epoch": 0.6030794110710556,
+      "grad_norm": 4.679615356530205,
+      "learning_rate": 4.0785323809991006e-07,
+      "loss": 1.7096,
+      "step": 8950
+    },
+    {
+      "epoch": 0.6034163269431623,
+      "grad_norm": 4.81123110598509,
+      "learning_rate": 4.072753255651521e-07,
+      "loss": 1.6589,
+      "step": 8955
+    },
+    {
+      "epoch": 0.603753242815269,
+      "grad_norm": 4.916986795492,
+      "learning_rate": 4.066975412877255e-07,
+      "loss": 1.6249,
+      "step": 8960
+    },
+    {
+      "epoch": 0.6040901586873758,
+      "grad_norm": 5.411200068581318,
+      "learning_rate": 4.0611988606682544e-07,
+      "loss": 1.6395,
+      "step": 8965
+    },
+    {
+      "epoch": 0.6044270745594825,
+      "grad_norm": 4.884173846637528,
+      "learning_rate": 4.0554236070146785e-07,
+      "loss": 1.698,
+      "step": 8970
+    },
+    {
+      "epoch": 0.6047639904315892,
+      "grad_norm": 4.846861187549292,
+      "learning_rate": 4.0496496599048963e-07,
+      "loss": 1.6821,
+      "step": 8975
+    },
+    {
+      "epoch": 0.605100906303696,
+      "grad_norm": 4.879563415748026,
+      "learning_rate": 4.0438770273254624e-07,
+      "loss": 1.6807,
+      "step": 8980
+    },
+    {
+      "epoch": 0.6054378221758027,
+      "grad_norm": 4.725484457298363,
+      "learning_rate": 4.038105717261119e-07,
+      "loss": 1.6332,
+      "step": 8985
+    },
+    {
+      "epoch": 0.6057747380479095,
+      "grad_norm": 4.649479102245512,
+      "learning_rate": 4.03233573769478e-07,
+      "loss": 1.71,
+      "step": 8990
+    },
+    {
+      "epoch": 0.6061116539200162,
+      "grad_norm": 4.945730805267605,
+      "learning_rate": 4.026567096607511e-07,
+      "loss": 1.6461,
+      "step": 8995
+    },
+    {
+      "epoch": 0.6064485697921229,
+      "grad_norm": 4.501885990649764,
+      "learning_rate": 4.020799801978535e-07,
+      "loss": 1.7412,
+      "step": 9000
+    },
+    {
+      "epoch": 0.6067854856642296,
+      "grad_norm": 4.707450309124685,
+      "learning_rate": 4.015033861785208e-07,
+      "loss": 1.6886,
+      "step": 9005
+    },
+    {
+      "epoch": 0.6071224015363363,
+      "grad_norm": 4.549425438204147,
+      "learning_rate": 4.0092692840030126e-07,
+      "loss": 1.7204,
+      "step": 9010
+    },
+    {
+      "epoch": 0.6074593174084432,
+      "grad_norm": 4.412648886343521,
+      "learning_rate": 4.003506076605547e-07,
+      "loss": 1.6506,
+      "step": 9015
+    },
+    {
+      "epoch": 0.6077962332805499,
+      "grad_norm": 4.728839488949062,
+      "learning_rate": 3.997744247564519e-07,
+      "loss": 1.5739,
+      "step": 9020
+    },
+    {
+      "epoch": 0.6081331491526566,
+      "grad_norm": 4.469095502288587,
+      "learning_rate": 3.9919838048497197e-07,
+      "loss": 1.7551,
+      "step": 9025
+    },
+    {
+      "epoch": 0.6084700650247633,
+      "grad_norm": 4.439886130777974,
+      "learning_rate": 3.98622475642903e-07,
+      "loss": 1.6805,
+      "step": 9030
+    },
+    {
+      "epoch": 0.60880698089687,
+      "grad_norm": 5.252294964963045,
+      "learning_rate": 3.980467110268405e-07,
+      "loss": 1.6273,
+      "step": 9035
+    },
+    {
+      "epoch": 0.6091438967689767,
+      "grad_norm": 4.822617039300253,
+      "learning_rate": 3.9747108743318493e-07,
+      "loss": 1.7041,
+      "step": 9040
+    },
+    {
+      "epoch": 0.6094808126410836,
+      "grad_norm": 4.619438478538796,
+      "learning_rate": 3.968956056581428e-07,
+      "loss": 1.6624,
+      "step": 9045
+    },
+    {
+      "epoch": 0.6098177285131903,
+      "grad_norm": 4.6117011288419505,
+      "learning_rate": 3.9632026649772366e-07,
+      "loss": 1.6139,
+      "step": 9050
+    },
+    {
+      "epoch": 0.610154644385297,
+      "grad_norm": 4.911659054403481,
+      "learning_rate": 3.9574507074774056e-07,
+      "loss": 1.6501,
+      "step": 9055
+    },
+    {
+      "epoch": 0.6104915602574037,
+      "grad_norm": 4.6589618898907,
+      "learning_rate": 3.951700192038072e-07,
+      "loss": 1.6172,
+      "step": 9060
+    },
+    {
+      "epoch": 0.6108284761295104,
+      "grad_norm": 4.311424914105757,
+      "learning_rate": 3.945951126613387e-07,
+      "loss": 1.7492,
+      "step": 9065
+    },
+    {
+      "epoch": 0.6111653920016172,
+      "grad_norm": 5.116936063115301,
+      "learning_rate": 3.9402035191554937e-07,
+      "loss": 1.6774,
+      "step": 9070
+    },
+    {
+      "epoch": 0.611502307873724,
+      "grad_norm": 4.611695226643144,
+      "learning_rate": 3.934457377614514e-07,
+      "loss": 1.6253,
+      "step": 9075
+    },
+    {
+      "epoch": 0.6118392237458307,
+      "grad_norm": 4.345447725618097,
+      "learning_rate": 3.9287127099385483e-07,
+      "loss": 1.7269,
+      "step": 9080
+    },
+    {
+      "epoch": 0.6121761396179374,
+      "grad_norm": 5.182706851476699,
+      "learning_rate": 3.9229695240736567e-07,
+      "loss": 1.6901,
+      "step": 9085
+    },
+    {
+      "epoch": 0.6125130554900441,
+      "grad_norm": 4.751801305761913,
+      "learning_rate": 3.917227827963846e-07,
+      "loss": 1.662,
+      "step": 9090
+    },
+    {
+      "epoch": 0.6128499713621509,
+      "grad_norm": 5.101440856558443,
+      "learning_rate": 3.9114876295510653e-07,
+      "loss": 1.6569,
+      "step": 9095
+    },
+    {
+      "epoch": 0.6131868872342576,
+      "grad_norm": 4.7541345176185095,
+      "learning_rate": 3.9057489367751947e-07,
+      "loss": 1.6445,
+      "step": 9100
+    },
+    {
+      "epoch": 0.6135238031063643,
+      "grad_norm": 4.459634190814962,
+      "learning_rate": 3.900011757574024e-07,
+      "loss": 1.7521,
+      "step": 9105
+    },
+    {
+      "epoch": 0.613860718978471,
+      "grad_norm": 4.852675298666461,
+      "learning_rate": 3.894276099883258e-07,
+      "loss": 1.6383,
+      "step": 9110
+    },
+    {
+      "epoch": 0.6141976348505778,
+      "grad_norm": 4.8681803338120435,
+      "learning_rate": 3.888541971636492e-07,
+      "loss": 1.6305,
+      "step": 9115
+    },
+    {
+      "epoch": 0.6145345507226846,
+      "grad_norm": 5.104377471505262,
+      "learning_rate": 3.8828093807652095e-07,
+      "loss": 1.7823,
+      "step": 9120
+    },
+    {
+      "epoch": 0.6148714665947913,
+      "grad_norm": 4.711200384568286,
+      "learning_rate": 3.8770783351987605e-07,
+      "loss": 1.6415,
+      "step": 9125
+    },
+    {
+      "epoch": 0.615208382466898,
+      "grad_norm": 4.986953684029128,
+      "learning_rate": 3.8713488428643656e-07,
+      "loss": 1.6783,
+      "step": 9130
+    },
+    {
+      "epoch": 0.6155452983390047,
+      "grad_norm": 4.464817870891317,
+      "learning_rate": 3.8656209116870906e-07,
+      "loss": 1.6513,
+      "step": 9135
+    },
+    {
+      "epoch": 0.6158822142111114,
+      "grad_norm": 4.602491329149551,
+      "learning_rate": 3.859894549589847e-07,
+      "loss": 1.6982,
+      "step": 9140
+    },
+    {
+      "epoch": 0.6162191300832183,
+      "grad_norm": 5.174034990707735,
+      "learning_rate": 3.854169764493371e-07,
+      "loss": 1.6953,
+      "step": 9145
+    },
+    {
+      "epoch": 0.616556045955325,
+      "grad_norm": 4.567308775862828,
+      "learning_rate": 3.848446564316223e-07,
+      "loss": 1.6572,
+      "step": 9150
+    },
+    {
+      "epoch": 0.6168929618274317,
+      "grad_norm": 4.619582257540075,
+      "learning_rate": 3.8427249569747656e-07,
+      "loss": 1.6788,
+      "step": 9155
+    },
+    {
+      "epoch": 0.6172298776995384,
+      "grad_norm": 4.615649431114516,
+      "learning_rate": 3.8370049503831614e-07,
+      "loss": 1.6735,
+      "step": 9160
+    },
+    {
+      "epoch": 0.6175667935716451,
+      "grad_norm": 4.421585714223809,
+      "learning_rate": 3.8312865524533606e-07,
+      "loss": 1.6711,
+      "step": 9165
+    },
+    {
+      "epoch": 0.617903709443752,
+      "grad_norm": 4.826993098533326,
+      "learning_rate": 3.825569771095082e-07,
+      "loss": 1.7108,
+      "step": 9170
+    },
+    {
+      "epoch": 0.6182406253158587,
+      "grad_norm": 4.613669265520506,
+      "learning_rate": 3.819854614215814e-07,
+      "loss": 1.7083,
+      "step": 9175
+    },
+    {
+      "epoch": 0.6185775411879654,
+      "grad_norm": 4.94162415452775,
+      "learning_rate": 3.814141089720796e-07,
+      "loss": 1.7266,
+      "step": 9180
+    },
+    {
+      "epoch": 0.6189144570600721,
+      "grad_norm": 4.8380036054530215,
+      "learning_rate": 3.8084292055130126e-07,
+      "loss": 1.736,
+      "step": 9185
+    },
+    {
+      "epoch": 0.6192513729321788,
+      "grad_norm": 4.846548864779536,
+      "learning_rate": 3.8027189694931715e-07,
+      "loss": 1.6704,
+      "step": 9190
+    },
+    {
+      "epoch": 0.6195882888042856,
+      "grad_norm": 4.51327248172737,
+      "learning_rate": 3.797010389559708e-07,
+      "loss": 1.6785,
+      "step": 9195
+    },
+    {
+      "epoch": 0.6199252046763923,
+      "grad_norm": 5.070239294852287,
+      "learning_rate": 3.7913034736087677e-07,
+      "loss": 1.7457,
+      "step": 9200
+    },
+    {
+      "epoch": 0.620262120548499,
+      "grad_norm": 4.755063973244972,
+      "learning_rate": 3.785598229534186e-07,
+      "loss": 1.6858,
+      "step": 9205
+    },
+    {
+      "epoch": 0.6205990364206058,
+      "grad_norm": 4.561460179584289,
+      "learning_rate": 3.7798946652274943e-07,
+      "loss": 1.6691,
+      "step": 9210
+    },
+    {
+      "epoch": 0.6209359522927125,
+      "grad_norm": 4.412726574679065,
+      "learning_rate": 3.7741927885778966e-07,
+      "loss": 1.5899,
+      "step": 9215
+    },
+    {
+      "epoch": 0.6212728681648192,
+      "grad_norm": 4.3476674942315,
+      "learning_rate": 3.768492607472263e-07,
+      "loss": 1.7019,
+      "step": 9220
+    },
+    {
+      "epoch": 0.621609784036926,
+      "grad_norm": 4.390399887814367,
+      "learning_rate": 3.7627941297951183e-07,
+      "loss": 1.7051,
+      "step": 9225
+    },
+    {
+      "epoch": 0.6219466999090327,
+      "grad_norm": 4.35192116787193,
+      "learning_rate": 3.7570973634286334e-07,
+      "loss": 1.6842,
+      "step": 9230
+    },
+    {
+      "epoch": 0.6222836157811394,
+      "grad_norm": 4.431659041359632,
+      "learning_rate": 3.7514023162526066e-07,
+      "loss": 1.6323,
+      "step": 9235
+    },
+    {
+      "epoch": 0.6226205316532462,
+      "grad_norm": 4.346065772652617,
+      "learning_rate": 3.745708996144463e-07,
+      "loss": 1.7121,
+      "step": 9240
+    },
+    {
+      "epoch": 0.6229574475253529,
+      "grad_norm": 4.983399406107984,
+      "learning_rate": 3.740017410979239e-07,
+      "loss": 1.6523,
+      "step": 9245
+    },
+    {
+      "epoch": 0.6232943633974597,
+      "grad_norm": 4.70323233287576,
+      "learning_rate": 3.734327568629569e-07,
+      "loss": 1.6911,
+      "step": 9250
+    },
+    {
+      "epoch": 0.6236312792695664,
+      "grad_norm": 4.170760438250213,
+      "learning_rate": 3.728639476965678e-07,
+      "loss": 1.6334,
+      "step": 9255
+    },
+    {
+      "epoch": 0.6239681951416731,
+      "grad_norm": 5.020288150663386,
+      "learning_rate": 3.7229531438553664e-07,
+      "loss": 1.6802,
+      "step": 9260
+    },
+    {
+      "epoch": 0.6243051110137798,
+      "grad_norm": 4.723953618912066,
+      "learning_rate": 3.7172685771640076e-07,
+      "loss": 1.7039,
+      "step": 9265
+    },
+    {
+      "epoch": 0.6246420268858865,
+      "grad_norm": 4.824681211683872,
+      "learning_rate": 3.7115857847545264e-07,
+      "loss": 1.6453,
+      "step": 9270
+    },
+    {
+      "epoch": 0.6249789427579934,
+      "grad_norm": 4.396461053658457,
+      "learning_rate": 3.7059047744873955e-07,
+      "loss": 1.6911,
+      "step": 9275
+    },
+    {
+      "epoch": 0.6253158586301001,
+      "grad_norm": 4.131448856801017,
+      "learning_rate": 3.700225554220626e-07,
+      "loss": 1.6903,
+      "step": 9280
+    },
+    {
+      "epoch": 0.6256527745022068,
+      "grad_norm": 4.375253047858914,
+      "learning_rate": 3.694548131809747e-07,
+      "loss": 1.6685,
+      "step": 9285
+    },
+    {
+      "epoch": 0.6259896903743135,
+      "grad_norm": 4.418083233974953,
+      "learning_rate": 3.6888725151078024e-07,
+      "loss": 1.6877,
+      "step": 9290
+    },
+    {
+      "epoch": 0.6263266062464202,
+      "grad_norm": 4.3329316853506565,
+      "learning_rate": 3.683198711965345e-07,
+      "loss": 1.6335,
+      "step": 9295
+    },
+    {
+      "epoch": 0.626663522118527,
+      "grad_norm": 4.504174343408066,
+      "learning_rate": 3.677526730230408e-07,
+      "loss": 1.6475,
+      "step": 9300
+    },
+    {
+      "epoch": 0.6270004379906338,
+      "grad_norm": 5.38608186396199,
+      "learning_rate": 3.671856577748512e-07,
+      "loss": 1.6343,
+      "step": 9305
+    },
+    {
+      "epoch": 0.6273373538627405,
+      "grad_norm": 5.39440651463409,
+      "learning_rate": 3.666188262362648e-07,
+      "loss": 1.6907,
+      "step": 9310
+    },
+    {
+      "epoch": 0.6276742697348472,
+      "grad_norm": 4.676904103854555,
+      "learning_rate": 3.660521791913265e-07,
+      "loss": 1.6992,
+      "step": 9315
+    },
+    {
+      "epoch": 0.6280111856069539,
+      "grad_norm": 4.559801782962305,
+      "learning_rate": 3.654857174238256e-07,
+      "loss": 1.6791,
+      "step": 9320
+    },
+    {
+      "epoch": 0.6283481014790607,
+      "grad_norm": 4.136109394326187,
+      "learning_rate": 3.649194417172957e-07,
+      "loss": 1.7617,
+      "step": 9325
+    },
+    {
+      "epoch": 0.6286850173511674,
+      "grad_norm": 4.590955654084948,
+      "learning_rate": 3.6435335285501283e-07,
+      "loss": 1.5974,
+      "step": 9330
+    },
+    {
+      "epoch": 0.6290219332232742,
+      "grad_norm": 4.253775067725555,
+      "learning_rate": 3.6378745161999426e-07,
+      "loss": 1.6966,
+      "step": 9335
+    },
+    {
+      "epoch": 0.6293588490953809,
+      "grad_norm": 4.671874023298497,
+      "learning_rate": 3.632217387949983e-07,
+      "loss": 1.6913,
+      "step": 9340
+    },
+    {
+      "epoch": 0.6296957649674876,
+      "grad_norm": 4.672636100483996,
+      "learning_rate": 3.626562151625223e-07,
+      "loss": 1.6015,
+      "step": 9345
+    },
+    {
+      "epoch": 0.6300326808395944,
+      "grad_norm": 4.572731068631272,
+      "learning_rate": 3.6209088150480173e-07,
+      "loss": 1.626,
+      "step": 9350
+    },
+    {
+      "epoch": 0.6303695967117011,
+      "grad_norm": 5.048291419272467,
+      "learning_rate": 3.6152573860380964e-07,
+      "loss": 1.6399,
+      "step": 9355
+    },
+    {
+      "epoch": 0.6307065125838078,
+      "grad_norm": 4.66081177092855,
+      "learning_rate": 3.6096078724125544e-07,
+      "loss": 1.7308,
+      "step": 9360
+    },
+    {
+      "epoch": 0.6310434284559145,
+      "grad_norm": 4.917944557473536,
+      "learning_rate": 3.603960281985828e-07,
+      "loss": 1.7118,
+      "step": 9365
+    },
+    {
+      "epoch": 0.6313803443280213,
+      "grad_norm": 4.639561844571507,
+      "learning_rate": 3.5983146225697007e-07,
+      "loss": 1.6598,
+      "step": 9370
+    },
+    {
+      "epoch": 0.6317172602001281,
+      "grad_norm": 4.809402432690832,
+      "learning_rate": 3.5926709019732855e-07,
+      "loss": 1.7416,
+      "step": 9375
+    },
+    {
+      "epoch": 0.6320541760722348,
+      "grad_norm": 4.481316598432365,
+      "learning_rate": 3.587029128003006e-07,
+      "loss": 1.6618,
+      "step": 9380
+    },
+    {
+      "epoch": 0.6323910919443415,
+      "grad_norm": 4.878186037519498,
+      "learning_rate": 3.581389308462601e-07,
+      "loss": 1.6539,
+      "step": 9385
+    },
+    {
+      "epoch": 0.6327280078164482,
+      "grad_norm": 4.691055116278031,
+      "learning_rate": 3.5757514511531016e-07,
+      "loss": 1.7012,
+      "step": 9390
+    },
+    {
+      "epoch": 0.6330649236885549,
+      "grad_norm": 4.596865110399032,
+      "learning_rate": 3.5701155638728297e-07,
+      "loss": 1.6794,
+      "step": 9395
+    },
+    {
+      "epoch": 0.6334018395606617,
+      "grad_norm": 4.358380598019165,
+      "learning_rate": 3.564481654417374e-07,
+      "loss": 1.661,
+      "step": 9400
+    },
+    {
+      "epoch": 0.6337387554327685,
+      "grad_norm": 4.412280666871581,
+      "learning_rate": 3.558849730579594e-07,
+      "loss": 1.6904,
+      "step": 9405
+    },
+    {
+      "epoch": 0.6340756713048752,
+      "grad_norm": 4.595888399957703,
+      "learning_rate": 3.553219800149603e-07,
+      "loss": 1.6599,
+      "step": 9410
+    },
+    {
+      "epoch": 0.6344125871769819,
+      "grad_norm": 4.504509215702589,
+      "learning_rate": 3.547591870914752e-07,
+      "loss": 1.6241,
+      "step": 9415
+    },
+    {
+      "epoch": 0.6347495030490886,
+      "grad_norm": 4.5321538447474286,
+      "learning_rate": 3.5419659506596287e-07,
+      "loss": 1.6667,
+      "step": 9420
+    },
+    {
+      "epoch": 0.6350864189211953,
+      "grad_norm": 4.770870718979196,
+      "learning_rate": 3.536342047166039e-07,
+      "loss": 1.6589,
+      "step": 9425
+    },
+    {
+      "epoch": 0.6354233347933022,
+      "grad_norm": 4.599599013949839,
+      "learning_rate": 3.530720168213001e-07,
+      "loss": 1.6545,
+      "step": 9430
+    },
+    {
+      "epoch": 0.6357602506654089,
+      "grad_norm": 4.39991418260652,
+      "learning_rate": 3.5251003215767305e-07,
+      "loss": 1.6523,
+      "step": 9435
+    },
+    {
+      "epoch": 0.6360971665375156,
+      "grad_norm": 4.722938780137947,
+      "learning_rate": 3.519482515030636e-07,
+      "loss": 1.6727,
+      "step": 9440
+    },
+    {
+      "epoch": 0.6364340824096223,
+      "grad_norm": 4.298009738528461,
+      "learning_rate": 3.5138667563452983e-07,
+      "loss": 1.7054,
+      "step": 9445
+    },
+    {
+      "epoch": 0.636770998281729,
+      "grad_norm": 5.373001197803553,
+      "learning_rate": 3.5082530532884703e-07,
+      "loss": 1.6143,
+      "step": 9450
+    },
+    {
+      "epoch": 0.6371079141538358,
+      "grad_norm": 4.362506764538609,
+      "learning_rate": 3.5026414136250607e-07,
+      "loss": 1.6585,
+      "step": 9455
+    },
+    {
+      "epoch": 0.6374448300259425,
+      "grad_norm": 3.956197636663764,
+      "learning_rate": 3.497031845117124e-07,
+      "loss": 1.7065,
+      "step": 9460
+    },
+    {
+      "epoch": 0.6377817458980493,
+      "grad_norm": 4.968824829024163,
+      "learning_rate": 3.4914243555238476e-07,
+      "loss": 1.7093,
+      "step": 9465
+    },
+    {
+      "epoch": 0.638118661770156,
+      "grad_norm": 4.654729665807823,
+      "learning_rate": 3.4858189526015453e-07,
+      "loss": 1.6439,
+      "step": 9470
+    },
+    {
+      "epoch": 0.6384555776422627,
+      "grad_norm": 4.7396445196150605,
+      "learning_rate": 3.4802156441036467e-07,
+      "loss": 1.6802,
+      "step": 9475
+    },
+    {
+      "epoch": 0.6387924935143695,
+      "grad_norm": 4.324066565036445,
+      "learning_rate": 3.4746144377806785e-07,
+      "loss": 1.6634,
+      "step": 9480
+    },
+    {
+      "epoch": 0.6391294093864762,
+      "grad_norm": 4.746429983809976,
+      "learning_rate": 3.4690153413802653e-07,
+      "loss": 1.7065,
+      "step": 9485
+    },
+    {
+      "epoch": 0.6394663252585829,
+      "grad_norm": 4.714466290466101,
+      "learning_rate": 3.4634183626471125e-07,
+      "loss": 1.7503,
+      "step": 9490
+    },
+    {
+      "epoch": 0.6398032411306896,
+      "grad_norm": 4.702082264829062,
+      "learning_rate": 3.457823509322992e-07,
+      "loss": 1.744,
+      "step": 9495
+    },
+    {
+      "epoch": 0.6401401570027964,
+      "grad_norm": 4.871189714313392,
+      "learning_rate": 3.452230789146741e-07,
+      "loss": 1.7177,
+      "step": 9500
+    },
+    {
+      "epoch": 0.6404770728749032,
+      "grad_norm": 4.972530681728673,
+      "learning_rate": 3.4466402098542435e-07,
+      "loss": 1.6523,
+      "step": 9505
+    },
+    {
+      "epoch": 0.6408139887470099,
+      "grad_norm": 4.753078175132621,
+      "learning_rate": 3.441051779178422e-07,
+      "loss": 1.6754,
+      "step": 9510
+    },
+    {
+      "epoch": 0.6411509046191166,
+      "grad_norm": 4.491634183568621,
+      "learning_rate": 3.4354655048492277e-07,
+      "loss": 1.6013,
+      "step": 9515
+    },
+    {
+      "epoch": 0.6414878204912233,
+      "grad_norm": 4.225886018919661,
+      "learning_rate": 3.429881394593629e-07,
+      "loss": 1.6376,
+      "step": 9520
+    },
+    {
+      "epoch": 0.64182473636333,
+      "grad_norm": 4.679624121585962,
+      "learning_rate": 3.4242994561356043e-07,
+      "loss": 1.6383,
+      "step": 9525
+    },
+    {
+      "epoch": 0.6421616522354369,
+      "grad_norm": 4.679224868456952,
+      "learning_rate": 3.4187196971961185e-07,
+      "loss": 1.7326,
+      "step": 9530
+    },
+    {
+      "epoch": 0.6424985681075436,
+      "grad_norm": 4.809163806682364,
+      "learning_rate": 3.4131421254931326e-07,
+      "loss": 1.704,
+      "step": 9535
+    },
+    {
+      "epoch": 0.6428354839796503,
+      "grad_norm": 4.6325336560221615,
+      "learning_rate": 3.4075667487415785e-07,
+      "loss": 1.7466,
+      "step": 9540
+    },
+    {
+      "epoch": 0.643172399851757,
+      "grad_norm": 4.436009749711743,
+      "learning_rate": 3.4019935746533474e-07,
+      "loss": 1.6869,
+      "step": 9545
+    },
+    {
+      "epoch": 0.6435093157238637,
+      "grad_norm": 4.864438953089616,
+      "learning_rate": 3.3964226109372884e-07,
+      "loss": 1.77,
+      "step": 9550
+    },
+    {
+      "epoch": 0.6438462315959704,
+      "grad_norm": 4.522632462926502,
+      "learning_rate": 3.390853865299195e-07,
+      "loss": 1.6717,
+      "step": 9555
+    },
+    {
+      "epoch": 0.6441831474680773,
+      "grad_norm": 4.744488425863059,
+      "learning_rate": 3.385287345441786e-07,
+      "loss": 1.7493,
+      "step": 9560
+    },
+    {
+      "epoch": 0.644520063340184,
+      "grad_norm": 4.833057581829567,
+      "learning_rate": 3.3797230590647073e-07,
+      "loss": 1.6675,
+      "step": 9565
+    },
+    {
+      "epoch": 0.6448569792122907,
+      "grad_norm": 4.492473060812444,
+      "learning_rate": 3.374161013864515e-07,
+      "loss": 1.6674,
+      "step": 9570
+    },
+    {
+      "epoch": 0.6451938950843974,
+      "grad_norm": 4.5854657066579785,
+      "learning_rate": 3.368601217534661e-07,
+      "loss": 1.7302,
+      "step": 9575
+    },
+    {
+      "epoch": 0.6455308109565041,
+      "grad_norm": 4.794003300008669,
+      "learning_rate": 3.3630436777654903e-07,
+      "loss": 1.7257,
+      "step": 9580
+    },
+    {
+      "epoch": 0.6458677268286109,
+      "grad_norm": 4.582700632423808,
+      "learning_rate": 3.357488402244227e-07,
+      "loss": 1.7046,
+      "step": 9585
+    },
+    {
+      "epoch": 0.6462046427007176,
+      "grad_norm": 5.099103738821147,
+      "learning_rate": 3.3519353986549604e-07,
+      "loss": 1.5777,
+      "step": 9590
+    },
+    {
+      "epoch": 0.6465415585728244,
+      "grad_norm": 4.861722997661979,
+      "learning_rate": 3.346384674678639e-07,
+      "loss": 1.7682,
+      "step": 9595
+    },
+    {
+      "epoch": 0.6468784744449311,
+      "grad_norm": 4.8580449100967,
+      "learning_rate": 3.3408362379930576e-07,
+      "loss": 1.6197,
+      "step": 9600
+    },
+    {
+      "epoch": 0.6472153903170378,
+      "grad_norm": 4.314387420963666,
+      "learning_rate": 3.335290096272849e-07,
+      "loss": 1.6476,
+      "step": 9605
+    },
+    {
+      "epoch": 0.6475523061891446,
+      "grad_norm": 4.8963693523258485,
+      "learning_rate": 3.3297462571894673e-07,
+      "loss": 1.6491,
+      "step": 9610
+    },
+    {
+      "epoch": 0.6478892220612513,
+      "grad_norm": 4.406712709507465,
+      "learning_rate": 3.3242047284111857e-07,
+      "loss": 1.6501,
+      "step": 9615
+    },
+    {
+      "epoch": 0.648226137933358,
+      "grad_norm": 4.610325002229822,
+      "learning_rate": 3.3186655176030826e-07,
+      "loss": 1.733,
+      "step": 9620
+    },
+    {
+      "epoch": 0.6485630538054648,
+      "grad_norm": 4.313952984941844,
+      "learning_rate": 3.3131286324270234e-07,
+      "loss": 1.5754,
+      "step": 9625
+    },
+    {
+      "epoch": 0.6488999696775715,
+      "grad_norm": 4.2904274217804845,
+      "learning_rate": 3.3075940805416654e-07,
+      "loss": 1.6616,
+      "step": 9630
+    },
+    {
+      "epoch": 0.6492368855496783,
+      "grad_norm": 4.33447125046999,
+      "learning_rate": 3.3020618696024316e-07,
+      "loss": 1.707,
+      "step": 9635
+    },
+    {
+      "epoch": 0.649573801421785,
+      "grad_norm": 4.51287714996691,
+      "learning_rate": 3.2965320072615113e-07,
+      "loss": 1.722,
+      "step": 9640
+    },
+    {
+      "epoch": 0.6499107172938917,
+      "grad_norm": 4.615478632755968,
+      "learning_rate": 3.2910045011678424e-07,
+      "loss": 1.6432,
+      "step": 9645
+    },
+    {
+      "epoch": 0.6502476331659984,
+      "grad_norm": 4.413389662630831,
+      "learning_rate": 3.2854793589671046e-07,
+      "loss": 1.739,
+      "step": 9650
+    },
+    {
+      "epoch": 0.6505845490381051,
+      "grad_norm": 4.421718857382131,
+      "learning_rate": 3.279956588301712e-07,
+      "loss": 1.665,
+      "step": 9655
+    },
+    {
+      "epoch": 0.650921464910212,
+      "grad_norm": 4.566051335669866,
+      "learning_rate": 3.274436196810789e-07,
+      "loss": 1.6526,
+      "step": 9660
+    },
+    {
+      "epoch": 0.6512583807823187,
+      "grad_norm": 4.772530503841975,
+      "learning_rate": 3.268918192130178e-07,
+      "loss": 1.6811,
+      "step": 9665
+    },
+    {
+      "epoch": 0.6515952966544254,
+      "grad_norm": 4.8099020061260624,
+      "learning_rate": 3.263402581892415e-07,
+      "loss": 1.628,
+      "step": 9670
+    },
+    {
+      "epoch": 0.6519322125265321,
+      "grad_norm": 4.617221914497451,
+      "learning_rate": 3.257889373726726e-07,
+      "loss": 1.7204,
+      "step": 9675
+    },
+    {
+      "epoch": 0.6522691283986388,
+      "grad_norm": 4.488349346785449,
+      "learning_rate": 3.252378575259013e-07,
+      "loss": 1.6725,
+      "step": 9680
+    },
+    {
+      "epoch": 0.6526060442707456,
+      "grad_norm": 4.567181671797062,
+      "learning_rate": 3.246870194111849e-07,
+      "loss": 1.68,
+      "step": 9685
+    },
+    {
+      "epoch": 0.6529429601428524,
+      "grad_norm": 4.893953362104648,
+      "learning_rate": 3.2413642379044557e-07,
+      "loss": 1.656,
+      "step": 9690
+    },
+    {
+      "epoch": 0.6532798760149591,
+      "grad_norm": 4.642415265648212,
+      "learning_rate": 3.235860714252708e-07,
+      "loss": 1.707,
+      "step": 9695
+    },
+    {
+      "epoch": 0.6536167918870658,
+      "grad_norm": 4.619809492061879,
+      "learning_rate": 3.2303596307691137e-07,
+      "loss": 1.6804,
+      "step": 9700
+    },
+    {
+      "epoch": 0.6539537077591725,
+      "grad_norm": 4.518938554948811,
+      "learning_rate": 3.2248609950628023e-07,
+      "loss": 1.7232,
+      "step": 9705
+    },
+    {
+      "epoch": 0.6542906236312793,
+      "grad_norm": 5.0305921574382175,
+      "learning_rate": 3.219364814739522e-07,
+      "loss": 1.6637,
+      "step": 9710
+    },
+    {
+      "epoch": 0.654627539503386,
+      "grad_norm": 4.958018906140687,
+      "learning_rate": 3.2138710974016226e-07,
+      "loss": 1.6516,
+      "step": 9715
+    },
+    {
+      "epoch": 0.6549644553754927,
+      "grad_norm": 4.411380116115451,
+      "learning_rate": 3.208379850648046e-07,
+      "loss": 1.7346,
+      "step": 9720
+    },
+    {
+      "epoch": 0.6553013712475995,
+      "grad_norm": 4.41765731158713,
+      "learning_rate": 3.202891082074318e-07,
+      "loss": 1.6891,
+      "step": 9725
+    },
+    {
+      "epoch": 0.6556382871197062,
+      "grad_norm": 4.41496004100469,
+      "learning_rate": 3.197404799272537e-07,
+      "loss": 1.69,
+      "step": 9730
+    },
+    {
+      "epoch": 0.6559752029918129,
+      "grad_norm": 5.0894961933664495,
+      "learning_rate": 3.191921009831365e-07,
+      "loss": 1.6848,
+      "step": 9735
+    },
+    {
+      "epoch": 0.6563121188639197,
+      "grad_norm": 4.773661036376162,
+      "learning_rate": 3.1864397213360093e-07,
+      "loss": 1.6932,
+      "step": 9740
+    },
+    {
+      "epoch": 0.6566490347360264,
+      "grad_norm": 4.564536291163815,
+      "learning_rate": 3.180960941368223e-07,
+      "loss": 1.6904,
+      "step": 9745
+    },
+    {
+      "epoch": 0.6569859506081331,
+      "grad_norm": 4.480751994541764,
+      "learning_rate": 3.175484677506288e-07,
+      "loss": 1.6445,
+      "step": 9750
+    },
+    {
+      "epoch": 0.6573228664802399,
+      "grad_norm": 4.814133616042183,
+      "learning_rate": 3.1700109373250063e-07,
+      "loss": 1.6457,
+      "step": 9755
+    },
+    {
+      "epoch": 0.6576597823523466,
+      "grad_norm": 4.939708091385734,
+      "learning_rate": 3.1645397283956843e-07,
+      "loss": 1.6636,
+      "step": 9760
+    },
+    {
+      "epoch": 0.6579966982244534,
+      "grad_norm": 4.437001055650553,
+      "learning_rate": 3.159071058286138e-07,
+      "loss": 1.6724,
+      "step": 9765
+    },
+    {
+      "epoch": 0.6583336140965601,
+      "grad_norm": 4.375092204411314,
+      "learning_rate": 3.1536049345606586e-07,
+      "loss": 1.7786,
+      "step": 9770
+    },
+    {
+      "epoch": 0.6586705299686668,
+      "grad_norm": 4.495874315937704,
+      "learning_rate": 3.1481413647800247e-07,
+      "loss": 1.5484,
+      "step": 9775
+    },
+    {
+      "epoch": 0.6590074458407735,
+      "grad_norm": 4.545858285177803,
+      "learning_rate": 3.14268035650148e-07,
+      "loss": 1.7075,
+      "step": 9780
+    },
+    {
+      "epoch": 0.6593443617128802,
+      "grad_norm": 4.455537213297519,
+      "learning_rate": 3.137221917278723e-07,
+      "loss": 1.7664,
+      "step": 9785
+    },
+    {
+      "epoch": 0.6596812775849871,
+      "grad_norm": 4.599843359938901,
+      "learning_rate": 3.1317660546618986e-07,
+      "loss": 1.5995,
+      "step": 9790
+    },
+    {
+      "epoch": 0.6600181934570938,
+      "grad_norm": 4.651667753009877,
+      "learning_rate": 3.1263127761975917e-07,
+      "loss": 1.6857,
+      "step": 9795
+    },
+    {
+      "epoch": 0.6603551093292005,
+      "grad_norm": 4.935378017068128,
+      "learning_rate": 3.12086208942881e-07,
+      "loss": 1.722,
+      "step": 9800
+    },
+    {
+      "epoch": 0.6606920252013072,
+      "grad_norm": 4.434594377115127,
+      "learning_rate": 3.1154140018949736e-07,
+      "loss": 1.6791,
+      "step": 9805
+    },
+    {
+      "epoch": 0.6610289410734139,
+      "grad_norm": 4.898749340528809,
+      "learning_rate": 3.1099685211319116e-07,
+      "loss": 1.6559,
+      "step": 9810
+    },
+    {
+      "epoch": 0.6613658569455207,
+      "grad_norm": 4.364968079409384,
+      "learning_rate": 3.104525654671849e-07,
+      "loss": 1.656,
+      "step": 9815
+    },
+    {
+      "epoch": 0.6617027728176275,
+      "grad_norm": 4.756136283646096,
+      "learning_rate": 3.099085410043386e-07,
+      "loss": 1.6067,
+      "step": 9820
+    },
+    {
+      "epoch": 0.6620396886897342,
+      "grad_norm": 4.232110056496074,
+      "learning_rate": 3.0936477947715064e-07,
+      "loss": 1.6345,
+      "step": 9825
+    },
+    {
+      "epoch": 0.6623766045618409,
+      "grad_norm": 4.5660844839178525,
+      "learning_rate": 3.088212816377552e-07,
+      "loss": 1.6549,
+      "step": 9830
+    },
+    {
+      "epoch": 0.6627135204339476,
+      "grad_norm": 4.580686875748608,
+      "learning_rate": 3.0827804823792157e-07,
+      "loss": 1.6828,
+      "step": 9835
+    },
+    {
+      "epoch": 0.6630504363060544,
+      "grad_norm": 4.541892340707929,
+      "learning_rate": 3.077350800290537e-07,
+      "loss": 1.6968,
+      "step": 9840
+    },
+    {
+      "epoch": 0.6633873521781611,
+      "grad_norm": 4.383080748763885,
+      "learning_rate": 3.071923777621885e-07,
+      "loss": 1.6321,
+      "step": 9845
+    },
+    {
+      "epoch": 0.6637242680502679,
+      "grad_norm": 4.6197233210843045,
+      "learning_rate": 3.066499421879948e-07,
+      "loss": 1.7752,
+      "step": 9850
+    },
+    {
+      "epoch": 0.6640611839223746,
+      "grad_norm": 4.571457917714042,
+      "learning_rate": 3.0610777405677286e-07,
+      "loss": 1.683,
+      "step": 9855
+    },
+    {
+      "epoch": 0.6643980997944813,
+      "grad_norm": 4.209426918741808,
+      "learning_rate": 3.05565874118453e-07,
+      "loss": 1.6486,
+      "step": 9860
+    },
+    {
+      "epoch": 0.6647350156665881,
+      "grad_norm": 4.81882087804017,
+      "learning_rate": 3.050242431225948e-07,
+      "loss": 1.6834,
+      "step": 9865
+    },
+    {
+      "epoch": 0.6650719315386948,
+      "grad_norm": 4.7481008632017065,
+      "learning_rate": 3.0448288181838487e-07,
+      "loss": 1.6682,
+      "step": 9870
+    },
+    {
+      "epoch": 0.6654088474108015,
+      "grad_norm": 4.476031844564653,
+      "learning_rate": 3.0394179095463804e-07,
+      "loss": 1.7116,
+      "step": 9875
+    },
+    {
+      "epoch": 0.6657457632829082,
+      "grad_norm": 4.791042749509363,
+      "learning_rate": 3.0340097127979426e-07,
+      "loss": 1.7137,
+      "step": 9880
+    },
+    {
+      "epoch": 0.666082679155015,
+      "grad_norm": 4.725157551146566,
+      "learning_rate": 3.0286042354191844e-07,
+      "loss": 1.6872,
+      "step": 9885
+    },
+    {
+      "epoch": 0.6664195950271218,
+      "grad_norm": 4.493288785483739,
+      "learning_rate": 3.0232014848869955e-07,
+      "loss": 1.6942,
+      "step": 9890
+    },
+    {
+      "epoch": 0.6667565108992285,
+      "grad_norm": 4.855990739567336,
+      "learning_rate": 3.0178014686744966e-07,
+      "loss": 1.6451,
+      "step": 9895
+    },
+    {
+      "epoch": 0.6670934267713352,
+      "grad_norm": 4.034583816812956,
+      "learning_rate": 3.0124041942510175e-07,
+      "loss": 1.689,
+      "step": 9900
+    },
+    {
+      "epoch": 0.6674303426434419,
+      "grad_norm": 5.4785906083317455,
+      "learning_rate": 3.007009669082103e-07,
+      "loss": 1.6689,
+      "step": 9905
+    },
+    {
+      "epoch": 0.6677672585155486,
+      "grad_norm": 4.546350851107137,
+      "learning_rate": 3.001617900629496e-07,
+      "loss": 1.7015,
+      "step": 9910
+    },
+    {
+      "epoch": 0.6681041743876553,
+      "grad_norm": 4.615618260513504,
+      "learning_rate": 2.996228896351119e-07,
+      "loss": 1.6844,
+      "step": 9915
+    },
+    {
+      "epoch": 0.6684410902597622,
+      "grad_norm": 4.372487354691214,
+      "learning_rate": 2.9908426637010773e-07,
+      "loss": 1.7324,
+      "step": 9920
+    },
+    {
+      "epoch": 0.6687780061318689,
+      "grad_norm": 4.693262050453,
+      "learning_rate": 2.98545921012964e-07,
+      "loss": 1.6777,
+      "step": 9925
+    },
+    {
+      "epoch": 0.6691149220039756,
+      "grad_norm": 4.827171097205222,
+      "learning_rate": 2.9800785430832354e-07,
+      "loss": 1.6361,
+      "step": 9930
+    },
+    {
+      "epoch": 0.6694518378760823,
+      "grad_norm": 4.4256333493258255,
+      "learning_rate": 2.9747006700044295e-07,
+      "loss": 1.6837,
+      "step": 9935
+    },
+    {
+      "epoch": 0.669788753748189,
+      "grad_norm": 4.330977095836329,
+      "learning_rate": 2.969325598331932e-07,
+      "loss": 1.6356,
+      "step": 9940
+    },
+    {
+      "epoch": 0.6701256696202958,
+      "grad_norm": 4.530930969038274,
+      "learning_rate": 2.9639533355005773e-07,
+      "loss": 1.666,
+      "step": 9945
+    },
+    {
+      "epoch": 0.6704625854924026,
+      "grad_norm": 4.304911554637418,
+      "learning_rate": 2.958583888941306e-07,
+      "loss": 1.6104,
+      "step": 9950
+    },
+    {
+      "epoch": 0.6707995013645093,
+      "grad_norm": 4.810772596336117,
+      "learning_rate": 2.9532172660811745e-07,
+      "loss": 1.6974,
+      "step": 9955
+    },
+    {
+      "epoch": 0.671136417236616,
+      "grad_norm": 4.560408079183627,
+      "learning_rate": 2.9478534743433247e-07,
+      "loss": 1.6618,
+      "step": 9960
+    },
+    {
+      "epoch": 0.6714733331087227,
+      "grad_norm": 3.9849964262716044,
+      "learning_rate": 2.9424925211469876e-07,
+      "loss": 1.658,
+      "step": 9965
+    },
+    {
+      "epoch": 0.6718102489808295,
+      "grad_norm": 4.9512408667329355,
+      "learning_rate": 2.9371344139074645e-07,
+      "loss": 1.6443,
+      "step": 9970
+    },
+    {
+      "epoch": 0.6721471648529362,
+      "grad_norm": 4.808192781712395,
+      "learning_rate": 2.9317791600361243e-07,
+      "loss": 1.5463,
+      "step": 9975
+    },
+    {
+      "epoch": 0.672484080725043,
+      "grad_norm": 4.265183883192022,
+      "learning_rate": 2.9264267669403833e-07,
+      "loss": 1.6875,
+      "step": 9980
+    },
+    {
+      "epoch": 0.6728209965971497,
+      "grad_norm": 4.724613844375058,
+      "learning_rate": 2.921077242023706e-07,
+      "loss": 1.6789,
+      "step": 9985
+    },
+    {
+      "epoch": 0.6731579124692564,
+      "grad_norm": 4.932628162130658,
+      "learning_rate": 2.9157305926855893e-07,
+      "loss": 1.6627,
+      "step": 9990
+    },
+    {
+      "epoch": 0.6734948283413632,
+      "grad_norm": 4.5177985804674545,
+      "learning_rate": 2.910386826321549e-07,
+      "loss": 1.6965,
+      "step": 9995
+    },
+    {
+      "epoch": 0.6738317442134699,
+      "grad_norm": 4.774452767288286,
+      "learning_rate": 2.905045950323114e-07,
+      "loss": 1.6736,
+      "step": 10000
+    },
+    {
+      "epoch": 0.6741686600855766,
+      "grad_norm": 4.831788795618858,
+      "learning_rate": 2.899707972077817e-07,
+      "loss": 1.6555,
+      "step": 10005
+    },
+    {
+      "epoch": 0.6745055759576833,
+      "grad_norm": 5.235512873128096,
+      "learning_rate": 2.8943728989691857e-07,
+      "loss": 1.7077,
+      "step": 10010
+    },
+    {
+      "epoch": 0.6748424918297901,
+      "grad_norm": 4.913331508484025,
+      "learning_rate": 2.88904073837672e-07,
+      "loss": 1.6784,
+      "step": 10015
+    },
+    {
+      "epoch": 0.6751794077018969,
+      "grad_norm": 4.516721095762328,
+      "learning_rate": 2.883711497675899e-07,
+      "loss": 1.6134,
+      "step": 10020
+    },
+    {
+      "epoch": 0.6755163235740036,
+      "grad_norm": 4.337341745892736,
+      "learning_rate": 2.878385184238163e-07,
+      "loss": 1.6352,
+      "step": 10025
+    },
+    {
+      "epoch": 0.6758532394461103,
+      "grad_norm": 4.56965178511602,
+      "learning_rate": 2.8730618054308964e-07,
+      "loss": 1.608,
+      "step": 10030
+    },
+    {
+      "epoch": 0.676190155318217,
+      "grad_norm": 4.322189613617754,
+      "learning_rate": 2.8677413686174325e-07,
+      "loss": 1.7271,
+      "step": 10035
+    },
+    {
+      "epoch": 0.6765270711903237,
+      "grad_norm": 5.232513125407505,
+      "learning_rate": 2.8624238811570325e-07,
+      "loss": 1.6729,
+      "step": 10040
+    },
+    {
+      "epoch": 0.6768639870624306,
+      "grad_norm": 4.4509227801927205,
+      "learning_rate": 2.8571093504048737e-07,
+      "loss": 1.6776,
+      "step": 10045
+    },
+    {
+      "epoch": 0.6772009029345373,
+      "grad_norm": 4.394021508706804,
+      "learning_rate": 2.851797783712049e-07,
+      "loss": 1.5934,
+      "step": 10050
+    },
+    {
+      "epoch": 0.677537818806644,
+      "grad_norm": 4.60912000955251,
+      "learning_rate": 2.8464891884255515e-07,
+      "loss": 1.7002,
+      "step": 10055
+    },
+    {
+      "epoch": 0.6778747346787507,
+      "grad_norm": 4.764954353498366,
+      "learning_rate": 2.8411835718882593e-07,
+      "loss": 1.6836,
+      "step": 10060
+    },
+    {
+      "epoch": 0.6782116505508574,
+      "grad_norm": 4.414848267445215,
+      "learning_rate": 2.835880941438936e-07,
+      "loss": 1.6317,
+      "step": 10065
+    },
+    {
+      "epoch": 0.6785485664229642,
+      "grad_norm": 4.601263605609601,
+      "learning_rate": 2.8305813044122093e-07,
+      "loss": 1.6786,
+      "step": 10070
+    },
+    {
+      "epoch": 0.678885482295071,
+      "grad_norm": 5.073682142084875,
+      "learning_rate": 2.8252846681385734e-07,
+      "loss": 1.755,
+      "step": 10075
+    },
+    {
+      "epoch": 0.6792223981671777,
+      "grad_norm": 4.657590732042178,
+      "learning_rate": 2.8199910399443625e-07,
+      "loss": 1.706,
+      "step": 10080
+    },
+    {
+      "epoch": 0.6795593140392844,
+      "grad_norm": 4.806472860391699,
+      "learning_rate": 2.8147004271517584e-07,
+      "loss": 1.6204,
+      "step": 10085
+    },
+    {
+      "epoch": 0.6798962299113911,
+      "grad_norm": 4.6660240490616705,
+      "learning_rate": 2.8094128370787694e-07,
+      "loss": 1.7356,
+      "step": 10090
+    },
+    {
+      "epoch": 0.6802331457834978,
+      "grad_norm": 5.228044220619245,
+      "learning_rate": 2.8041282770392196e-07,
+      "loss": 1.6482,
+      "step": 10095
+    },
+    {
+      "epoch": 0.6805700616556046,
+      "grad_norm": 4.918634141984381,
+      "learning_rate": 2.7988467543427454e-07,
+      "loss": 1.6087,
+      "step": 10100
+    },
+    {
+      "epoch": 0.6809069775277113,
+      "grad_norm": 5.041088919289125,
+      "learning_rate": 2.7935682762947837e-07,
+      "loss": 1.7085,
+      "step": 10105
+    },
+    {
+      "epoch": 0.6812438933998181,
+      "grad_norm": 4.557815685478516,
+      "learning_rate": 2.788292850196553e-07,
+      "loss": 1.6837,
+      "step": 10110
+    },
+    {
+      "epoch": 0.6815808092719248,
+      "grad_norm": 4.49296018717802,
+      "learning_rate": 2.783020483345057e-07,
+      "loss": 1.6294,
+      "step": 10115
+    },
+    {
+      "epoch": 0.6819177251440315,
+      "grad_norm": 4.344023180239478,
+      "learning_rate": 2.777751183033067e-07,
+      "loss": 1.7425,
+      "step": 10120
+    },
+    {
+      "epoch": 0.6822546410161383,
+      "grad_norm": 4.352887650061733,
+      "learning_rate": 2.772484956549107e-07,
+      "loss": 1.6764,
+      "step": 10125
+    },
+    {
+      "epoch": 0.682591556888245,
+      "grad_norm": 4.701595075307773,
+      "learning_rate": 2.7672218111774566e-07,
+      "loss": 1.7387,
+      "step": 10130
+    },
+    {
+      "epoch": 0.6829284727603517,
+      "grad_norm": 4.220584191142634,
+      "learning_rate": 2.7619617541981287e-07,
+      "loss": 1.6475,
+      "step": 10135
+    },
+    {
+      "epoch": 0.6832653886324584,
+      "grad_norm": 4.72955412832327,
+      "learning_rate": 2.756704792886869e-07,
+      "loss": 1.6484,
+      "step": 10140
+    },
+    {
+      "epoch": 0.6836023045045652,
+      "grad_norm": 4.6550694748926515,
+      "learning_rate": 2.7514509345151347e-07,
+      "loss": 1.6406,
+      "step": 10145
+    },
+    {
+      "epoch": 0.683939220376672,
+      "grad_norm": 4.782860902895655,
+      "learning_rate": 2.746200186350097e-07,
+      "loss": 1.7218,
+      "step": 10150
+    },
+    {
+      "epoch": 0.6842761362487787,
+      "grad_norm": 4.376748815020733,
+      "learning_rate": 2.740952555654622e-07,
+      "loss": 1.6845,
+      "step": 10155
+    },
+    {
+      "epoch": 0.6846130521208854,
+      "grad_norm": 4.783809039044787,
+      "learning_rate": 2.735708049687262e-07,
+      "loss": 1.6067,
+      "step": 10160
+    },
+    {
+      "epoch": 0.6849499679929921,
+      "grad_norm": 4.511631761949158,
+      "learning_rate": 2.730466675702251e-07,
+      "loss": 1.7025,
+      "step": 10165
+    },
+    {
+      "epoch": 0.6852868838650988,
+      "grad_norm": 5.087600336235812,
+      "learning_rate": 2.7252284409494906e-07,
+      "loss": 1.6463,
+      "step": 10170
+    },
+    {
+      "epoch": 0.6856237997372057,
+      "grad_norm": 4.2007227960927835,
+      "learning_rate": 2.7199933526745364e-07,
+      "loss": 1.7036,
+      "step": 10175
+    },
+    {
+      "epoch": 0.6859607156093124,
+      "grad_norm": 5.117839466222247,
+      "learning_rate": 2.714761418118596e-07,
+      "loss": 1.6092,
+      "step": 10180
+    },
+    {
+      "epoch": 0.6862976314814191,
+      "grad_norm": 5.356330190705277,
+      "learning_rate": 2.7095326445185143e-07,
+      "loss": 1.7043,
+      "step": 10185
+    },
+    {
+      "epoch": 0.6866345473535258,
+      "grad_norm": 4.564267086174341,
+      "learning_rate": 2.704307039106759e-07,
+      "loss": 1.6383,
+      "step": 10190
+    },
+    {
+      "epoch": 0.6869714632256325,
+      "grad_norm": 4.164845663181458,
+      "learning_rate": 2.6990846091114205e-07,
+      "loss": 1.7422,
+      "step": 10195
+    },
+    {
+      "epoch": 0.6873083790977393,
+      "grad_norm": 4.131763760270544,
+      "learning_rate": 2.6938653617561967e-07,
+      "loss": 1.7038,
+      "step": 10200
+    },
+    {
+      "epoch": 0.687645294969846,
+      "grad_norm": 4.312564061995232,
+      "learning_rate": 2.688649304260383e-07,
+      "loss": 1.6819,
+      "step": 10205
+    },
+    {
+      "epoch": 0.6879822108419528,
+      "grad_norm": 4.61153702729294,
+      "learning_rate": 2.683436443838859e-07,
+      "loss": 1.6896,
+      "step": 10210
+    },
+    {
+      "epoch": 0.6883191267140595,
+      "grad_norm": 4.703111722307184,
+      "learning_rate": 2.678226787702086e-07,
+      "loss": 1.7162,
+      "step": 10215
+    },
+    {
+      "epoch": 0.6886560425861662,
+      "grad_norm": 4.635701429196508,
+      "learning_rate": 2.673020343056094e-07,
+      "loss": 1.7633,
+      "step": 10220
+    },
+    {
+      "epoch": 0.688992958458273,
+      "grad_norm": 4.7344644151009465,
+      "learning_rate": 2.6678171171024657e-07,
+      "loss": 1.7299,
+      "step": 10225
+    },
+    {
+      "epoch": 0.6893298743303797,
+      "grad_norm": 4.486499868121771,
+      "learning_rate": 2.6626171170383373e-07,
+      "loss": 1.6064,
+      "step": 10230
+    },
+    {
+      "epoch": 0.6896667902024864,
+      "grad_norm": 4.534823258749324,
+      "learning_rate": 2.6574203500563776e-07,
+      "loss": 1.6902,
+      "step": 10235
+    },
+    {
+      "epoch": 0.6900037060745932,
+      "grad_norm": 4.439413468251779,
+      "learning_rate": 2.6522268233447894e-07,
+      "loss": 1.7037,
+      "step": 10240
+    },
+    {
+      "epoch": 0.6903406219466999,
+      "grad_norm": 4.552475395505495,
+      "learning_rate": 2.6470365440872866e-07,
+      "loss": 1.6477,
+      "step": 10245
+    },
+    {
+      "epoch": 0.6906775378188067,
+      "grad_norm": 5.036103615948694,
+      "learning_rate": 2.641849519463099e-07,
+      "loss": 1.7112,
+      "step": 10250
+    },
+    {
+      "epoch": 0.6910144536909134,
+      "grad_norm": 4.445684686793016,
+      "learning_rate": 2.6366657566469465e-07,
+      "loss": 1.6505,
+      "step": 10255
+    },
+    {
+      "epoch": 0.6913513695630201,
+      "grad_norm": 4.4173485045453,
+      "learning_rate": 2.631485262809043e-07,
+      "loss": 1.6391,
+      "step": 10260
+    },
+    {
+      "epoch": 0.6916882854351268,
+      "grad_norm": 4.317080090850698,
+      "learning_rate": 2.6263080451150797e-07,
+      "loss": 1.6771,
+      "step": 10265
+    },
+    {
+      "epoch": 0.6920252013072336,
+      "grad_norm": 4.64941590077001,
+      "learning_rate": 2.621134110726217e-07,
+      "loss": 1.7569,
+      "step": 10270
+    },
+    {
+      "epoch": 0.6923621171793403,
+      "grad_norm": 4.590236812787074,
+      "learning_rate": 2.6159634667990683e-07,
+      "loss": 1.7185,
+      "step": 10275
+    },
+    {
+      "epoch": 0.6926990330514471,
+      "grad_norm": 4.9812699111640795,
+      "learning_rate": 2.610796120485701e-07,
+      "loss": 1.6939,
+      "step": 10280
+    },
+    {
+      "epoch": 0.6930359489235538,
+      "grad_norm": 4.375038058888252,
+      "learning_rate": 2.605632078933623e-07,
+      "loss": 1.6136,
+      "step": 10285
+    },
+    {
+      "epoch": 0.6933728647956605,
+      "grad_norm": 4.897729962824886,
+      "learning_rate": 2.600471349285763e-07,
+      "loss": 1.7073,
+      "step": 10290
+    },
+    {
+      "epoch": 0.6937097806677672,
+      "grad_norm": 4.565936668798104,
+      "learning_rate": 2.5953139386804764e-07,
+      "loss": 1.679,
+      "step": 10295
+    },
+    {
+      "epoch": 0.6940466965398739,
+      "grad_norm": 5.169593685108845,
+      "learning_rate": 2.5901598542515256e-07,
+      "loss": 1.6327,
+      "step": 10300
+    },
+    {
+      "epoch": 0.6943836124119808,
+      "grad_norm": 4.373417565376957,
+      "learning_rate": 2.5850091031280684e-07,
+      "loss": 1.6841,
+      "step": 10305
+    },
+    {
+      "epoch": 0.6947205282840875,
+      "grad_norm": 4.559109121588478,
+      "learning_rate": 2.579861692434658e-07,
+      "loss": 1.6803,
+      "step": 10310
+    },
+    {
+      "epoch": 0.6950574441561942,
+      "grad_norm": 4.571150904547719,
+      "learning_rate": 2.574717629291222e-07,
+      "loss": 1.6552,
+      "step": 10315
+    },
+    {
+      "epoch": 0.6953943600283009,
+      "grad_norm": 4.4202180568282,
+      "learning_rate": 2.5695769208130615e-07,
+      "loss": 1.6534,
+      "step": 10320
+    },
+    {
+      "epoch": 0.6957312759004076,
+      "grad_norm": 4.343547857486007,
+      "learning_rate": 2.564439574110833e-07,
+      "loss": 1.6725,
+      "step": 10325
+    },
+    {
+      "epoch": 0.6960681917725144,
+      "grad_norm": 4.568973039945001,
+      "learning_rate": 2.559305596290547e-07,
+      "loss": 1.7006,
+      "step": 10330
+    },
+    {
+      "epoch": 0.6964051076446212,
+      "grad_norm": 4.772607107022066,
+      "learning_rate": 2.554174994453555e-07,
+      "loss": 1.6212,
+      "step": 10335
+    },
+    {
+      "epoch": 0.6967420235167279,
+      "grad_norm": 4.558184718510563,
+      "learning_rate": 2.549047775696532e-07,
+      "loss": 1.65,
+      "step": 10340
+    },
+    {
+      "epoch": 0.6970789393888346,
+      "grad_norm": 4.684970264407261,
+      "learning_rate": 2.543923947111481e-07,
+      "loss": 1.6709,
+      "step": 10345
+    },
+    {
+      "epoch": 0.6974158552609413,
+      "grad_norm": 4.531575026071747,
+      "learning_rate": 2.538803515785714e-07,
+      "loss": 1.7056,
+      "step": 10350
+    },
+    {
+      "epoch": 0.6977527711330481,
+      "grad_norm": 4.527121588986311,
+      "learning_rate": 2.5336864888018393e-07,
+      "loss": 1.641,
+      "step": 10355
+    },
+    {
+      "epoch": 0.6980896870051548,
+      "grad_norm": 4.644098459060704,
+      "learning_rate": 2.528572873237761e-07,
+      "loss": 1.6763,
+      "step": 10360
+    },
+    {
+      "epoch": 0.6984266028772615,
+      "grad_norm": 4.5149636521464425,
+      "learning_rate": 2.5234626761666647e-07,
+      "loss": 1.7769,
+      "step": 10365
+    },
+    {
+      "epoch": 0.6987635187493683,
+      "grad_norm": 4.273525129862513,
+      "learning_rate": 2.5183559046570036e-07,
+      "loss": 1.7503,
+      "step": 10370
+    },
+    {
+      "epoch": 0.699100434621475,
+      "grad_norm": 4.438322061742733,
+      "learning_rate": 2.513252565772496e-07,
+      "loss": 1.6898,
+      "step": 10375
+    },
+    {
+      "epoch": 0.6994373504935818,
+      "grad_norm": 4.534576711817759,
+      "learning_rate": 2.5081526665721133e-07,
+      "loss": 1.6594,
+      "step": 10380
+    },
+    {
+      "epoch": 0.6997742663656885,
+      "grad_norm": 4.441597881770876,
+      "learning_rate": 2.503056214110062e-07,
+      "loss": 1.6447,
+      "step": 10385
+    },
+    {
+      "epoch": 0.7001111822377952,
+      "grad_norm": 4.525682977168778,
+      "learning_rate": 2.497963215435789e-07,
+      "loss": 1.6098,
+      "step": 10390
+    },
+    {
+      "epoch": 0.7004480981099019,
+      "grad_norm": 4.983403352698664,
+      "learning_rate": 2.492873677593964e-07,
+      "loss": 1.6142,
+      "step": 10395
+    },
+    {
+      "epoch": 0.7007850139820087,
+      "grad_norm": 4.6617257994767884,
+      "learning_rate": 2.4877876076244626e-07,
+      "loss": 1.7664,
+      "step": 10400
+    },
+    {
+      "epoch": 0.7011219298541155,
+      "grad_norm": 4.9910938176979105,
+      "learning_rate": 2.482705012562367e-07,
+      "loss": 1.6845,
+      "step": 10405
+    },
+    {
+      "epoch": 0.7014588457262222,
+      "grad_norm": 4.5161646653470795,
+      "learning_rate": 2.4776258994379546e-07,
+      "loss": 1.6812,
+      "step": 10410
+    },
+    {
+      "epoch": 0.7017957615983289,
+      "grad_norm": 4.645809092572502,
+      "learning_rate": 2.4725502752766883e-07,
+      "loss": 1.6761,
+      "step": 10415
+    },
+    {
+      "epoch": 0.7021326774704356,
+      "grad_norm": 4.762960103801375,
+      "learning_rate": 2.4674781470991967e-07,
+      "loss": 1.5992,
+      "step": 10420
+    },
+    {
+      "epoch": 0.7024695933425423,
+      "grad_norm": 4.868038772852577,
+      "learning_rate": 2.462409521921282e-07,
+      "loss": 1.6409,
+      "step": 10425
+    },
+    {
+      "epoch": 0.7028065092146492,
+      "grad_norm": 4.489504746004885,
+      "learning_rate": 2.4573444067538985e-07,
+      "loss": 1.7111,
+      "step": 10430
+    },
+    {
+      "epoch": 0.7031434250867559,
+      "grad_norm": 4.853823090419402,
+      "learning_rate": 2.4522828086031404e-07,
+      "loss": 1.752,
+      "step": 10435
+    },
+    {
+      "epoch": 0.7034803409588626,
+      "grad_norm": 4.867281068587823,
+      "learning_rate": 2.4472247344702424e-07,
+      "loss": 1.6618,
+      "step": 10440
+    },
+    {
+      "epoch": 0.7038172568309693,
+      "grad_norm": 4.832483270849909,
+      "learning_rate": 2.442170191351566e-07,
+      "loss": 1.7637,
+      "step": 10445
+    },
+    {
+      "epoch": 0.704154172703076,
+      "grad_norm": 4.850938617706512,
+      "learning_rate": 2.4371191862385816e-07,
+      "loss": 1.6092,
+      "step": 10450
+    },
+    {
+      "epoch": 0.7044910885751827,
+      "grad_norm": 4.641938087798896,
+      "learning_rate": 2.4320717261178715e-07,
+      "loss": 1.6935,
+      "step": 10455
+    },
+    {
+      "epoch": 0.7048280044472895,
+      "grad_norm": 4.740015041009832,
+      "learning_rate": 2.4270278179711163e-07,
+      "loss": 1.655,
+      "step": 10460
+    },
+    {
+      "epoch": 0.7051649203193963,
+      "grad_norm": 4.939551223908603,
+      "learning_rate": 2.4219874687750754e-07,
+      "loss": 1.6852,
+      "step": 10465
+    },
+    {
+      "epoch": 0.705501836191503,
+      "grad_norm": 4.711384090241523,
+      "learning_rate": 2.4169506855015923e-07,
+      "loss": 1.741,
+      "step": 10470
+    },
+    {
+      "epoch": 0.7058387520636097,
+      "grad_norm": 4.5166802497839535,
+      "learning_rate": 2.4119174751175787e-07,
+      "loss": 1.6763,
+      "step": 10475
+    },
+    {
+      "epoch": 0.7061756679357164,
+      "grad_norm": 4.663782473551683,
+      "learning_rate": 2.406887844584998e-07,
+      "loss": 1.6727,
+      "step": 10480
+    },
+    {
+      "epoch": 0.7065125838078232,
+      "grad_norm": 4.6119383856125875,
+      "learning_rate": 2.401861800860868e-07,
+      "loss": 1.651,
+      "step": 10485
+    },
+    {
+      "epoch": 0.7068494996799299,
+      "grad_norm": 4.251793594934676,
+      "learning_rate": 2.396839350897241e-07,
+      "loss": 1.7073,
+      "step": 10490
+    },
+    {
+      "epoch": 0.7071864155520367,
+      "grad_norm": 4.832759698580428,
+      "learning_rate": 2.391820501641203e-07,
+      "loss": 1.7365,
+      "step": 10495
+    },
+    {
+      "epoch": 0.7075233314241434,
+      "grad_norm": 4.917880501129237,
+      "learning_rate": 2.3868052600348524e-07,
+      "loss": 1.6904,
+      "step": 10500
+    },
+    {
+      "epoch": 0.7078602472962501,
+      "grad_norm": 4.330584540056101,
+      "learning_rate": 2.381793633015305e-07,
+      "loss": 1.6593,
+      "step": 10505
+    },
+    {
+      "epoch": 0.7081971631683569,
+      "grad_norm": 5.002212666859194,
+      "learning_rate": 2.3767856275146748e-07,
+      "loss": 1.6624,
+      "step": 10510
+    },
+    {
+      "epoch": 0.7085340790404636,
+      "grad_norm": 4.736950520168997,
+      "learning_rate": 2.3717812504600616e-07,
+      "loss": 1.6459,
+      "step": 10515
+    },
+    {
+      "epoch": 0.7088709949125703,
+      "grad_norm": 4.418489255619871,
+      "learning_rate": 2.3667805087735516e-07,
+      "loss": 1.6726,
+      "step": 10520
+    },
+    {
+      "epoch": 0.709207910784677,
+      "grad_norm": 4.6673695105380135,
+      "learning_rate": 2.3617834093722033e-07,
+      "loss": 1.637,
+      "step": 10525
+    },
+    {
+      "epoch": 0.7095448266567838,
+      "grad_norm": 4.7641420953866005,
+      "learning_rate": 2.3567899591680317e-07,
+      "loss": 1.679,
+      "step": 10530
+    },
+    {
+      "epoch": 0.7098817425288906,
+      "grad_norm": 4.505140740962068,
+      "learning_rate": 2.351800165068008e-07,
+      "loss": 1.6526,
+      "step": 10535
+    },
+    {
+      "epoch": 0.7102186584009973,
+      "grad_norm": 4.365411249482726,
+      "learning_rate": 2.346814033974047e-07,
+      "loss": 1.686,
+      "step": 10540
+    },
+    {
+      "epoch": 0.710555574273104,
+      "grad_norm": 4.693059176177268,
+      "learning_rate": 2.3418315727829962e-07,
+      "loss": 1.723,
+      "step": 10545
+    },
+    {
+      "epoch": 0.7108924901452107,
+      "grad_norm": 4.619080120396149,
+      "learning_rate": 2.336852788386623e-07,
+      "loss": 1.7133,
+      "step": 10550
+    },
+    {
+      "epoch": 0.7112294060173174,
+      "grad_norm": 4.939014950379131,
+      "learning_rate": 2.331877687671614e-07,
+      "loss": 1.6821,
+      "step": 10555
+    },
+    {
+      "epoch": 0.7115663218894243,
+      "grad_norm": 4.685813885840334,
+      "learning_rate": 2.3269062775195596e-07,
+      "loss": 1.6593,
+      "step": 10560
+    },
+    {
+      "epoch": 0.711903237761531,
+      "grad_norm": 4.524468772976415,
+      "learning_rate": 2.321938564806944e-07,
+      "loss": 1.7253,
+      "step": 10565
+    },
+    {
+      "epoch": 0.7122401536336377,
+      "grad_norm": 4.892348621641314,
+      "learning_rate": 2.3169745564051353e-07,
+      "loss": 1.6513,
+      "step": 10570
+    },
+    {
+      "epoch": 0.7125770695057444,
+      "grad_norm": 4.6502041358305,
+      "learning_rate": 2.3120142591803825e-07,
+      "loss": 1.7138,
+      "step": 10575
+    },
+    {
+      "epoch": 0.7129139853778511,
+      "grad_norm": 4.838257165305869,
+      "learning_rate": 2.307057679993797e-07,
+      "loss": 1.6281,
+      "step": 10580
+    },
+    {
+      "epoch": 0.7132509012499579,
+      "grad_norm": 4.64288900947958,
+      "learning_rate": 2.30210482570135e-07,
+      "loss": 1.7001,
+      "step": 10585
+    },
+    {
+      "epoch": 0.7135878171220646,
+      "grad_norm": 4.518161574505775,
+      "learning_rate": 2.2971557031538607e-07,
+      "loss": 1.6696,
+      "step": 10590
+    },
+    {
+      "epoch": 0.7139247329941714,
+      "grad_norm": 4.390328603188015,
+      "learning_rate": 2.2922103191969828e-07,
+      "loss": 1.7425,
+      "step": 10595
+    },
+    {
+      "epoch": 0.7142616488662781,
+      "grad_norm": 4.573838466451132,
+      "learning_rate": 2.2872686806712032e-07,
+      "loss": 1.6373,
+      "step": 10600
+    },
+    {
+      "epoch": 0.7145985647383848,
+      "grad_norm": 4.732632510738885,
+      "learning_rate": 2.2823307944118254e-07,
+      "loss": 1.6169,
+      "step": 10605
+    },
+    {
+      "epoch": 0.7149354806104916,
+      "grad_norm": 4.817100325987074,
+      "learning_rate": 2.2773966672489665e-07,
+      "loss": 1.7377,
+      "step": 10610
+    },
+    {
+      "epoch": 0.7152723964825983,
+      "grad_norm": 4.597601777541554,
+      "learning_rate": 2.2724663060075368e-07,
+      "loss": 1.644,
+      "step": 10615
+    },
+    {
+      "epoch": 0.715609312354705,
+      "grad_norm": 4.521845052152514,
+      "learning_rate": 2.2675397175072437e-07,
+      "loss": 1.6741,
+      "step": 10620
+    },
+    {
+      "epoch": 0.7159462282268118,
+      "grad_norm": 4.103897229849213,
+      "learning_rate": 2.2626169085625762e-07,
+      "loss": 1.6386,
+      "step": 10625
+    },
+    {
+      "epoch": 0.7162831440989185,
+      "grad_norm": 4.501031185020447,
+      "learning_rate": 2.25769788598279e-07,
+      "loss": 1.667,
+      "step": 10630
+    },
+    {
+      "epoch": 0.7166200599710252,
+      "grad_norm": 4.635506503161774,
+      "learning_rate": 2.2527826565719084e-07,
+      "loss": 1.6141,
+      "step": 10635
+    },
+    {
+      "epoch": 0.716956975843132,
+      "grad_norm": 4.404740981961707,
+      "learning_rate": 2.2478712271287087e-07,
+      "loss": 1.7058,
+      "step": 10640
+    },
+    {
+      "epoch": 0.7172938917152387,
+      "grad_norm": 4.688543744546726,
+      "learning_rate": 2.2429636044467059e-07,
+      "loss": 1.6766,
+      "step": 10645
+    },
+    {
+      "epoch": 0.7176308075873454,
+      "grad_norm": 4.211449995644894,
+      "learning_rate": 2.2380597953141573e-07,
+      "loss": 1.7425,
+      "step": 10650
+    },
+    {
+      "epoch": 0.7179677234594521,
+      "grad_norm": 4.504845812085762,
+      "learning_rate": 2.2331598065140396e-07,
+      "loss": 1.6595,
+      "step": 10655
+    },
+    {
+      "epoch": 0.7183046393315589,
+      "grad_norm": 4.676702954559052,
+      "learning_rate": 2.228263644824045e-07,
+      "loss": 1.7615,
+      "step": 10660
+    },
+    {
+      "epoch": 0.7186415552036657,
+      "grad_norm": 4.186233739778813,
+      "learning_rate": 2.2233713170165757e-07,
+      "loss": 1.6644,
+      "step": 10665
+    },
+    {
+      "epoch": 0.7189784710757724,
+      "grad_norm": 4.680317760117085,
+      "learning_rate": 2.2184828298587298e-07,
+      "loss": 1.6968,
+      "step": 10670
+    },
+    {
+      "epoch": 0.7193153869478791,
+      "grad_norm": 4.355114775428761,
+      "learning_rate": 2.213598190112294e-07,
+      "loss": 1.6111,
+      "step": 10675
+    },
+    {
+      "epoch": 0.7196523028199858,
+      "grad_norm": 4.722909090510962,
+      "learning_rate": 2.2087174045337275e-07,
+      "loss": 1.6572,
+      "step": 10680
+    },
+    {
+      "epoch": 0.7199892186920925,
+      "grad_norm": 4.268160931804694,
+      "learning_rate": 2.2038404798741644e-07,
+      "loss": 1.678,
+      "step": 10685
+    },
+    {
+      "epoch": 0.7203261345641994,
+      "grad_norm": 4.3897851171954,
+      "learning_rate": 2.1989674228793987e-07,
+      "loss": 1.6375,
+      "step": 10690
+    },
+    {
+      "epoch": 0.7206630504363061,
+      "grad_norm": 4.946929325535152,
+      "learning_rate": 2.1940982402898684e-07,
+      "loss": 1.6924,
+      "step": 10695
+    },
+    {
+      "epoch": 0.7209999663084128,
+      "grad_norm": 4.429108850083181,
+      "learning_rate": 2.1892329388406582e-07,
+      "loss": 1.6301,
+      "step": 10700
+    },
+    {
+      "epoch": 0.7213368821805195,
+      "grad_norm": 4.576282749266952,
+      "learning_rate": 2.1843715252614847e-07,
+      "loss": 1.6423,
+      "step": 10705
+    },
+    {
+      "epoch": 0.7216737980526262,
+      "grad_norm": 4.7374662176046956,
+      "learning_rate": 2.179514006276681e-07,
+      "loss": 1.694,
+      "step": 10710
+    },
+    {
+      "epoch": 0.722010713924733,
+      "grad_norm": 4.108799578698973,
+      "learning_rate": 2.1746603886051978e-07,
+      "loss": 1.7134,
+      "step": 10715
+    },
+    {
+      "epoch": 0.7223476297968398,
+      "grad_norm": 4.562017959059164,
+      "learning_rate": 2.169810678960591e-07,
+      "loss": 1.6636,
+      "step": 10720
+    },
+    {
+      "epoch": 0.7226845456689465,
+      "grad_norm": 4.1553309206914655,
+      "learning_rate": 2.1649648840510047e-07,
+      "loss": 1.6234,
+      "step": 10725
+    },
+    {
+      "epoch": 0.7230214615410532,
+      "grad_norm": 5.012536579623761,
+      "learning_rate": 2.1601230105791751e-07,
+      "loss": 1.709,
+      "step": 10730
+    },
+    {
+      "epoch": 0.7233583774131599,
+      "grad_norm": 4.8962071217764125,
+      "learning_rate": 2.1552850652424077e-07,
+      "loss": 1.7263,
+      "step": 10735
+    },
+    {
+      "epoch": 0.7236952932852667,
+      "grad_norm": 4.720798604388035,
+      "learning_rate": 2.150451054732581e-07,
+      "loss": 1.6333,
+      "step": 10740
+    },
+    {
+      "epoch": 0.7240322091573734,
+      "grad_norm": 4.7913493095136825,
+      "learning_rate": 2.1456209857361246e-07,
+      "loss": 1.659,
+      "step": 10745
+    },
+    {
+      "epoch": 0.7243691250294801,
+      "grad_norm": 4.498833574428302,
+      "learning_rate": 2.1407948649340208e-07,
+      "loss": 1.6841,
+      "step": 10750
+    },
+    {
+      "epoch": 0.7247060409015869,
+      "grad_norm": 5.1590058288463645,
+      "learning_rate": 2.1359726990017908e-07,
+      "loss": 1.6044,
+      "step": 10755
+    },
+    {
+      "epoch": 0.7250429567736936,
+      "grad_norm": 4.936858137916012,
+      "learning_rate": 2.13115449460948e-07,
+      "loss": 1.6876,
+      "step": 10760
+    },
+    {
+      "epoch": 0.7253798726458004,
+      "grad_norm": 4.799281359022529,
+      "learning_rate": 2.12634025842166e-07,
+      "loss": 1.6138,
+      "step": 10765
+    },
+    {
+      "epoch": 0.7257167885179071,
+      "grad_norm": 5.074248882474168,
+      "learning_rate": 2.1215299970974132e-07,
+      "loss": 1.6955,
+      "step": 10770
+    },
+    {
+      "epoch": 0.7260537043900138,
+      "grad_norm": 4.311575992161255,
+      "learning_rate": 2.116723717290318e-07,
+      "loss": 1.7084,
+      "step": 10775
+    },
+    {
+      "epoch": 0.7263906202621205,
+      "grad_norm": 5.082312637120859,
+      "learning_rate": 2.111921425648453e-07,
+      "loss": 1.7333,
+      "step": 10780
+    },
+    {
+      "epoch": 0.7267275361342272,
+      "grad_norm": 4.6329981092644195,
+      "learning_rate": 2.1071231288143777e-07,
+      "loss": 1.7053,
+      "step": 10785
+    },
+    {
+      "epoch": 0.7270644520063341,
+      "grad_norm": 4.775432945237623,
+      "learning_rate": 2.1023288334251222e-07,
+      "loss": 1.6273,
+      "step": 10790
+    },
+    {
+      "epoch": 0.7274013678784408,
+      "grad_norm": 4.185322290611864,
+      "learning_rate": 2.0975385461121864e-07,
+      "loss": 1.7184,
+      "step": 10795
+    },
+    {
+      "epoch": 0.7277382837505475,
+      "grad_norm": 4.7166590882906,
+      "learning_rate": 2.0927522735015268e-07,
+      "loss": 1.5728,
+      "step": 10800
+    },
+    {
+      "epoch": 0.7280751996226542,
+      "grad_norm": 4.4664703136065596,
+      "learning_rate": 2.0879700222135416e-07,
+      "loss": 1.6866,
+      "step": 10805
+    },
+    {
+      "epoch": 0.7284121154947609,
+      "grad_norm": 4.844312254152635,
+      "learning_rate": 2.083191798863072e-07,
+      "loss": 1.6721,
+      "step": 10810
+    },
+    {
+      "epoch": 0.7287490313668676,
+      "grad_norm": 4.6986867193190065,
+      "learning_rate": 2.0784176100593836e-07,
+      "loss": 1.6653,
+      "step": 10815
+    },
+    {
+      "epoch": 0.7290859472389745,
+      "grad_norm": 4.896429792837326,
+      "learning_rate": 2.0736474624061655e-07,
+      "loss": 1.7465,
+      "step": 10820
+    },
+    {
+      "epoch": 0.7294228631110812,
+      "grad_norm": 4.492589280114266,
+      "learning_rate": 2.0688813625015123e-07,
+      "loss": 1.6479,
+      "step": 10825
+    },
+    {
+      "epoch": 0.7297597789831879,
+      "grad_norm": 4.778220852460387,
+      "learning_rate": 2.064119316937923e-07,
+      "loss": 1.7129,
+      "step": 10830
+    },
+    {
+      "epoch": 0.7300966948552946,
+      "grad_norm": 4.643533713787365,
+      "learning_rate": 2.0593613323022907e-07,
+      "loss": 1.6287,
+      "step": 10835
+    },
+    {
+      "epoch": 0.7304336107274013,
+      "grad_norm": 5.0151875683234834,
+      "learning_rate": 2.054607415175884e-07,
+      "loss": 1.6879,
+      "step": 10840
+    },
+    {
+      "epoch": 0.7307705265995081,
+      "grad_norm": 4.846287699529274,
+      "learning_rate": 2.0498575721343525e-07,
+      "loss": 1.69,
+      "step": 10845
+    },
+    {
+      "epoch": 0.7311074424716149,
+      "grad_norm": 4.665676909970296,
+      "learning_rate": 2.0451118097477093e-07,
+      "loss": 1.6734,
+      "step": 10850
+    },
+    {
+      "epoch": 0.7314443583437216,
+      "grad_norm": 5.059613017774967,
+      "learning_rate": 2.0403701345803186e-07,
+      "loss": 1.6899,
+      "step": 10855
+    },
+    {
+      "epoch": 0.7317812742158283,
+      "grad_norm": 4.647770298843087,
+      "learning_rate": 2.0356325531908952e-07,
+      "loss": 1.6748,
+      "step": 10860
+    },
+    {
+      "epoch": 0.732118190087935,
+      "grad_norm": 4.3547322930879435,
+      "learning_rate": 2.0308990721324926e-07,
+      "loss": 1.6818,
+      "step": 10865
+    },
+    {
+      "epoch": 0.7324551059600418,
+      "grad_norm": 4.3281885006679,
+      "learning_rate": 2.0261696979524873e-07,
+      "loss": 1.6646,
+      "step": 10870
+    },
+    {
+      "epoch": 0.7327920218321485,
+      "grad_norm": 4.585176262897073,
+      "learning_rate": 2.0214444371925793e-07,
+      "loss": 1.5786,
+      "step": 10875
+    },
+    {
+      "epoch": 0.7331289377042552,
+      "grad_norm": 4.47905166124732,
+      "learning_rate": 2.0167232963887787e-07,
+      "loss": 1.6532,
+      "step": 10880
+    },
+    {
+      "epoch": 0.733465853576362,
+      "grad_norm": 4.852975997183129,
+      "learning_rate": 2.0120062820713974e-07,
+      "loss": 1.6787,
+      "step": 10885
+    },
+    {
+      "epoch": 0.7338027694484687,
+      "grad_norm": 4.907158016902662,
+      "learning_rate": 2.0072934007650345e-07,
+      "loss": 1.6677,
+      "step": 10890
+    },
+    {
+      "epoch": 0.7341396853205755,
+      "grad_norm": 4.440284305274986,
+      "learning_rate": 2.0025846589885798e-07,
+      "loss": 1.6863,
+      "step": 10895
+    },
+    {
+      "epoch": 0.7344766011926822,
+      "grad_norm": 4.427979545085985,
+      "learning_rate": 1.99788006325519e-07,
+      "loss": 1.5922,
+      "step": 10900
+    },
+    {
+      "epoch": 0.7348135170647889,
+      "grad_norm": 4.416115150299774,
+      "learning_rate": 1.9931796200722943e-07,
+      "loss": 1.6723,
+      "step": 10905
+    },
+    {
+      "epoch": 0.7351504329368956,
+      "grad_norm": 4.736296900364095,
+      "learning_rate": 1.9884833359415698e-07,
+      "loss": 1.6714,
+      "step": 10910
+    },
+    {
+      "epoch": 0.7354873488090024,
+      "grad_norm": 4.682669507953387,
+      "learning_rate": 1.9837912173589494e-07,
+      "loss": 1.7085,
+      "step": 10915
+    },
+    {
+      "epoch": 0.7358242646811092,
+      "grad_norm": 4.710098455479335,
+      "learning_rate": 1.979103270814596e-07,
+      "loss": 1.7463,
+      "step": 10920
+    },
+    {
+      "epoch": 0.7361611805532159,
+      "grad_norm": 4.56398031316537,
+      "learning_rate": 1.9744195027929072e-07,
+      "loss": 1.691,
+      "step": 10925
+    },
+    {
+      "epoch": 0.7364980964253226,
+      "grad_norm": 4.883794149339468,
+      "learning_rate": 1.9697399197725023e-07,
+      "loss": 1.7055,
+      "step": 10930
+    },
+    {
+      "epoch": 0.7368350122974293,
+      "grad_norm": 4.420876529292157,
+      "learning_rate": 1.965064528226204e-07,
+      "loss": 1.7008,
+      "step": 10935
+    },
+    {
+      "epoch": 0.737171928169536,
+      "grad_norm": 4.765157524613675,
+      "learning_rate": 1.9603933346210445e-07,
+      "loss": 1.7264,
+      "step": 10940
+    },
+    {
+      "epoch": 0.7375088440416429,
+      "grad_norm": 4.578533250992113,
+      "learning_rate": 1.9557263454182476e-07,
+      "loss": 1.6727,
+      "step": 10945
+    },
+    {
+      "epoch": 0.7378457599137496,
+      "grad_norm": 4.4105063470620625,
+      "learning_rate": 1.9510635670732216e-07,
+      "loss": 1.669,
+      "step": 10950
+    },
+    {
+      "epoch": 0.7381826757858563,
+      "grad_norm": 4.762507051489227,
+      "learning_rate": 1.946405006035548e-07,
+      "loss": 1.6808,
+      "step": 10955
+    },
+    {
+      "epoch": 0.738519591657963,
+      "grad_norm": 4.59971833063233,
+      "learning_rate": 1.9417506687489772e-07,
+      "loss": 1.6845,
+      "step": 10960
+    },
+    {
+      "epoch": 0.7388565075300697,
+      "grad_norm": 4.442562019236419,
+      "learning_rate": 1.937100561651418e-07,
+      "loss": 1.6556,
+      "step": 10965
+    },
+    {
+      "epoch": 0.7391934234021765,
+      "grad_norm": 4.809988398340547,
+      "learning_rate": 1.9324546911749246e-07,
+      "loss": 1.6909,
+      "step": 10970
+    },
+    {
+      "epoch": 0.7395303392742832,
+      "grad_norm": 4.294425250731902,
+      "learning_rate": 1.9278130637456957e-07,
+      "loss": 1.5945,
+      "step": 10975
+    },
+    {
+      "epoch": 0.73986725514639,
+      "grad_norm": 5.316130213272725,
+      "learning_rate": 1.923175685784056e-07,
+      "loss": 1.6772,
+      "step": 10980
+    },
+    {
+      "epoch": 0.7402041710184967,
+      "grad_norm": 4.942738826736831,
+      "learning_rate": 1.9185425637044567e-07,
+      "loss": 1.6838,
+      "step": 10985
+    },
+    {
+      "epoch": 0.7405410868906034,
+      "grad_norm": 4.547439281413283,
+      "learning_rate": 1.9139137039154584e-07,
+      "loss": 1.7118,
+      "step": 10990
+    },
+    {
+      "epoch": 0.7408780027627101,
+      "grad_norm": 4.497209974141303,
+      "learning_rate": 1.9092891128197308e-07,
+      "loss": 1.6566,
+      "step": 10995
+    },
+    {
+      "epoch": 0.7412149186348169,
+      "grad_norm": 4.723607597807449,
+      "learning_rate": 1.904668796814033e-07,
+      "loss": 1.782,
+      "step": 11000
+    },
+    {
+      "epoch": 0.7415518345069236,
+      "grad_norm": 5.126840547555821,
+      "learning_rate": 1.9000527622892154e-07,
+      "loss": 1.6948,
+      "step": 11005
+    },
+    {
+      "epoch": 0.7418887503790303,
+      "grad_norm": 4.564915672077015,
+      "learning_rate": 1.895441015630206e-07,
+      "loss": 1.6858,
+      "step": 11010
+    },
+    {
+      "epoch": 0.7422256662511371,
+      "grad_norm": 4.565825346690561,
+      "learning_rate": 1.8908335632160011e-07,
+      "loss": 1.7515,
+      "step": 11015
+    },
+    {
+      "epoch": 0.7425625821232438,
+      "grad_norm": 4.607141520570874,
+      "learning_rate": 1.8862304114196542e-07,
+      "loss": 1.6714,
+      "step": 11020
+    },
+    {
+      "epoch": 0.7428994979953506,
+      "grad_norm": 4.878415824316645,
+      "learning_rate": 1.8816315666082744e-07,
+      "loss": 1.6519,
+      "step": 11025
+    },
+    {
+      "epoch": 0.7432364138674573,
+      "grad_norm": 4.502110040030405,
+      "learning_rate": 1.877037035143013e-07,
+      "loss": 1.696,
+      "step": 11030
+    },
+    {
+      "epoch": 0.743573329739564,
+      "grad_norm": 4.419294474942364,
+      "learning_rate": 1.8724468233790512e-07,
+      "loss": 1.6732,
+      "step": 11035
+    },
+    {
+      "epoch": 0.7439102456116707,
+      "grad_norm": 4.957901599135894,
+      "learning_rate": 1.867860937665599e-07,
+      "loss": 1.7591,
+      "step": 11040
+    },
+    {
+      "epoch": 0.7442471614837775,
+      "grad_norm": 4.307693778434295,
+      "learning_rate": 1.8632793843458827e-07,
+      "loss": 1.7046,
+      "step": 11045
+    },
+    {
+      "epoch": 0.7445840773558843,
+      "grad_norm": 5.0797146696117155,
+      "learning_rate": 1.8587021697571313e-07,
+      "loss": 1.668,
+      "step": 11050
+    },
+    {
+      "epoch": 0.744920993227991,
+      "grad_norm": 4.733020859471026,
+      "learning_rate": 1.854129300230578e-07,
+      "loss": 1.6791,
+      "step": 11055
+    },
+    {
+      "epoch": 0.7452579091000977,
+      "grad_norm": 4.7002276751867385,
+      "learning_rate": 1.849560782091445e-07,
+      "loss": 1.6457,
+      "step": 11060
+    },
+    {
+      "epoch": 0.7455948249722044,
+      "grad_norm": 4.77322186606534,
+      "learning_rate": 1.8449966216589319e-07,
+      "loss": 1.6891,
+      "step": 11065
+    },
+    {
+      "epoch": 0.7459317408443111,
+      "grad_norm": 4.959855386815174,
+      "learning_rate": 1.8404368252462128e-07,
+      "loss": 1.6506,
+      "step": 11070
+    },
+    {
+      "epoch": 0.746268656716418,
+      "grad_norm": 4.726346375566695,
+      "learning_rate": 1.8358813991604262e-07,
+      "loss": 1.7413,
+      "step": 11075
+    },
+    {
+      "epoch": 0.7466055725885247,
+      "grad_norm": 4.487389952604869,
+      "learning_rate": 1.8313303497026673e-07,
+      "loss": 1.6747,
+      "step": 11080
+    },
+    {
+      "epoch": 0.7469424884606314,
+      "grad_norm": 5.067852653243485,
+      "learning_rate": 1.8267836831679718e-07,
+      "loss": 1.7386,
+      "step": 11085
+    },
+    {
+      "epoch": 0.7472794043327381,
+      "grad_norm": 4.796721785210641,
+      "learning_rate": 1.8222414058453183e-07,
+      "loss": 1.7148,
+      "step": 11090
+    },
+    {
+      "epoch": 0.7476163202048448,
+      "grad_norm": 4.357543678187077,
+      "learning_rate": 1.8177035240176136e-07,
+      "loss": 1.7123,
+      "step": 11095
+    },
+    {
+      "epoch": 0.7479532360769516,
+      "grad_norm": 5.122671623146191,
+      "learning_rate": 1.8131700439616803e-07,
+      "loss": 1.7409,
+      "step": 11100
+    },
+    {
+      "epoch": 0.7482901519490583,
+      "grad_norm": 4.448290814257407,
+      "learning_rate": 1.8086409719482576e-07,
+      "loss": 1.6186,
+      "step": 11105
+    },
+    {
+      "epoch": 0.7486270678211651,
+      "grad_norm": 4.511243100329989,
+      "learning_rate": 1.8041163142419857e-07,
+      "loss": 1.7047,
+      "step": 11110
+    },
+    {
+      "epoch": 0.7489639836932718,
+      "grad_norm": 4.772657575056115,
+      "learning_rate": 1.7995960771013962e-07,
+      "loss": 1.6571,
+      "step": 11115
+    },
+    {
+      "epoch": 0.7493008995653785,
+      "grad_norm": 5.016379241843775,
+      "learning_rate": 1.7950802667789107e-07,
+      "loss": 1.7197,
+      "step": 11120
+    },
+    {
+      "epoch": 0.7496378154374853,
+      "grad_norm": 4.31718338088045,
+      "learning_rate": 1.7905688895208259e-07,
+      "loss": 1.6686,
+      "step": 11125
+    },
+    {
+      "epoch": 0.749974731309592,
+      "grad_norm": 4.460748644736323,
+      "learning_rate": 1.7860619515673032e-07,
+      "loss": 1.6984,
+      "step": 11130
+    },
+    {
+      "epoch": 0.7503116471816987,
+      "grad_norm": 4.940196253932972,
+      "learning_rate": 1.7815594591523687e-07,
+      "loss": 1.7274,
+      "step": 11135
+    },
+    {
+      "epoch": 0.7506485630538055,
+      "grad_norm": 4.619600331904886,
+      "learning_rate": 1.777061418503898e-07,
+      "loss": 1.6933,
+      "step": 11140
+    },
+    {
+      "epoch": 0.7509854789259122,
+      "grad_norm": 4.643905236837146,
+      "learning_rate": 1.7725678358436053e-07,
+      "loss": 1.671,
+      "step": 11145
+    },
+    {
+      "epoch": 0.751322394798019,
+      "grad_norm": 4.49307652200486,
+      "learning_rate": 1.7680787173870454e-07,
+      "loss": 1.6081,
+      "step": 11150
+    },
+    {
+      "epoch": 0.7516593106701257,
+      "grad_norm": 4.390677025866737,
+      "learning_rate": 1.763594069343589e-07,
+      "loss": 1.6297,
+      "step": 11155
+    },
+    {
+      "epoch": 0.7519962265422324,
+      "grad_norm": 5.09476617875323,
+      "learning_rate": 1.7591138979164337e-07,
+      "loss": 1.6135,
+      "step": 11160
+    },
+    {
+      "epoch": 0.7523331424143391,
+      "grad_norm": 4.56890945684151,
+      "learning_rate": 1.7546382093025758e-07,
+      "loss": 1.6907,
+      "step": 11165
+    },
+    {
+      "epoch": 0.7526700582864458,
+      "grad_norm": 4.894976797363558,
+      "learning_rate": 1.7501670096928162e-07,
+      "loss": 1.7158,
+      "step": 11170
+    },
+    {
+      "epoch": 0.7530069741585526,
+      "grad_norm": 4.160433485102773,
+      "learning_rate": 1.7457003052717473e-07,
+      "loss": 1.7441,
+      "step": 11175
+    },
+    {
+      "epoch": 0.7533438900306594,
+      "grad_norm": 5.015497260674881,
+      "learning_rate": 1.741238102217738e-07,
+      "loss": 1.6565,
+      "step": 11180
+    },
+    {
+      "epoch": 0.7536808059027661,
+      "grad_norm": 4.777756060009474,
+      "learning_rate": 1.736780406702937e-07,
+      "loss": 1.6735,
+      "step": 11185
+    },
+    {
+      "epoch": 0.7540177217748728,
+      "grad_norm": 4.831225099167053,
+      "learning_rate": 1.7323272248932564e-07,
+      "loss": 1.6308,
+      "step": 11190
+    },
+    {
+      "epoch": 0.7543546376469795,
+      "grad_norm": 4.364088966880527,
+      "learning_rate": 1.727878562948362e-07,
+      "loss": 1.6505,
+      "step": 11195
+    },
+    {
+      "epoch": 0.7546915535190862,
+      "grad_norm": 5.26461802860794,
+      "learning_rate": 1.723434427021671e-07,
+      "loss": 1.6145,
+      "step": 11200
+    },
+    {
+      "epoch": 0.7550284693911931,
+      "grad_norm": 4.703754105055933,
+      "learning_rate": 1.7189948232603412e-07,
+      "loss": 1.6986,
+      "step": 11205
+    },
+    {
+      "epoch": 0.7553653852632998,
+      "grad_norm": 4.1370611882167685,
+      "learning_rate": 1.7145597578052557e-07,
+      "loss": 1.7309,
+      "step": 11210
+    },
+    {
+      "epoch": 0.7557023011354065,
+      "grad_norm": 4.56949563785717,
+      "learning_rate": 1.7101292367910259e-07,
+      "loss": 1.7101,
+      "step": 11215
+    },
+    {
+      "epoch": 0.7560392170075132,
+      "grad_norm": 4.734655354891645,
+      "learning_rate": 1.7057032663459768e-07,
+      "loss": 1.7221,
+      "step": 11220
+    },
+    {
+      "epoch": 0.7563761328796199,
+      "grad_norm": 5.206838932444794,
+      "learning_rate": 1.701281852592134e-07,
+      "loss": 1.6888,
+      "step": 11225
+    },
+    {
+      "epoch": 0.7567130487517267,
+      "grad_norm": 4.6074064994191835,
+      "learning_rate": 1.696865001645228e-07,
+      "loss": 1.6462,
+      "step": 11230
+    },
+    {
+      "epoch": 0.7570499646238334,
+      "grad_norm": 4.177604176716723,
+      "learning_rate": 1.6924527196146692e-07,
+      "loss": 1.738,
+      "step": 11235
+    },
+    {
+      "epoch": 0.7573868804959402,
+      "grad_norm": 4.560519854232525,
+      "learning_rate": 1.6880450126035572e-07,
+      "loss": 1.6563,
+      "step": 11240
+    },
+    {
+      "epoch": 0.7577237963680469,
+      "grad_norm": 4.882181848246992,
+      "learning_rate": 1.683641886708655e-07,
+      "loss": 1.6493,
+      "step": 11245
+    },
+    {
+      "epoch": 0.7580607122401536,
+      "grad_norm": 4.570694014559624,
+      "learning_rate": 1.6792433480203955e-07,
+      "loss": 1.6609,
+      "step": 11250
+    },
+    {
+      "epoch": 0.7583976281122604,
+      "grad_norm": 4.615121303405667,
+      "learning_rate": 1.674849402622865e-07,
+      "loss": 1.6867,
+      "step": 11255
+    },
+    {
+      "epoch": 0.7587345439843671,
+      "grad_norm": 4.570503569438364,
+      "learning_rate": 1.6704600565937927e-07,
+      "loss": 1.7028,
+      "step": 11260
+    },
+    {
+      "epoch": 0.7590714598564738,
+      "grad_norm": 4.702666604621073,
+      "learning_rate": 1.6660753160045498e-07,
+      "loss": 1.6276,
+      "step": 11265
+    },
+    {
+      "epoch": 0.7594083757285806,
+      "grad_norm": 4.62630415804712,
+      "learning_rate": 1.6616951869201378e-07,
+      "loss": 1.6527,
+      "step": 11270
+    },
+    {
+      "epoch": 0.7597452916006873,
+      "grad_norm": 4.096803295319819,
+      "learning_rate": 1.6573196753991747e-07,
+      "loss": 1.7448,
+      "step": 11275
+    },
+    {
+      "epoch": 0.7600822074727941,
+      "grad_norm": 4.56230206217443,
+      "learning_rate": 1.652948787493896e-07,
+      "loss": 1.6791,
+      "step": 11280
+    },
+    {
+      "epoch": 0.7604191233449008,
+      "grad_norm": 4.79645465297604,
+      "learning_rate": 1.64858252925014e-07,
+      "loss": 1.7032,
+      "step": 11285
+    },
+    {
+      "epoch": 0.7607560392170075,
+      "grad_norm": 4.617093579090467,
+      "learning_rate": 1.6442209067073442e-07,
+      "loss": 1.6793,
+      "step": 11290
+    },
+    {
+      "epoch": 0.7610929550891142,
+      "grad_norm": 4.69948875641977,
+      "learning_rate": 1.639863925898527e-07,
+      "loss": 1.6961,
+      "step": 11295
+    },
+    {
+      "epoch": 0.761429870961221,
+      "grad_norm": 4.811597345530875,
+      "learning_rate": 1.6355115928502934e-07,
+      "loss": 1.7011,
+      "step": 11300
+    },
+    {
+      "epoch": 0.7617667868333278,
+      "grad_norm": 4.718286454240784,
+      "learning_rate": 1.6311639135828176e-07,
+      "loss": 1.7075,
+      "step": 11305
+    },
+    {
+      "epoch": 0.7621037027054345,
+      "grad_norm": 5.058155884325266,
+      "learning_rate": 1.6268208941098344e-07,
+      "loss": 1.6541,
+      "step": 11310
+    },
+    {
+      "epoch": 0.7624406185775412,
+      "grad_norm": 4.513026537436967,
+      "learning_rate": 1.6224825404386326e-07,
+      "loss": 1.6382,
+      "step": 11315
+    },
+    {
+      "epoch": 0.7627775344496479,
+      "grad_norm": 4.346149115041241,
+      "learning_rate": 1.6181488585700541e-07,
+      "loss": 1.5756,
+      "step": 11320
+    },
+    {
+      "epoch": 0.7631144503217546,
+      "grad_norm": 5.096384285195435,
+      "learning_rate": 1.6138198544984692e-07,
+      "loss": 1.7126,
+      "step": 11325
+    },
+    {
+      "epoch": 0.7634513661938614,
+      "grad_norm": 4.680091495711113,
+      "learning_rate": 1.609495534211785e-07,
+      "loss": 1.6963,
+      "step": 11330
+    },
+    {
+      "epoch": 0.7637882820659682,
+      "grad_norm": 4.456805907972039,
+      "learning_rate": 1.6051759036914286e-07,
+      "loss": 1.6394,
+      "step": 11335
+    },
+    {
+      "epoch": 0.7641251979380749,
+      "grad_norm": 4.667742373252477,
+      "learning_rate": 1.6008609689123364e-07,
+      "loss": 1.7233,
+      "step": 11340
+    },
+    {
+      "epoch": 0.7644621138101816,
+      "grad_norm": 4.881823308634071,
+      "learning_rate": 1.596550735842953e-07,
+      "loss": 1.7119,
+      "step": 11345
+    },
+    {
+      "epoch": 0.7647990296822883,
+      "grad_norm": 4.588255263137127,
+      "learning_rate": 1.5922452104452204e-07,
+      "loss": 1.6889,
+      "step": 11350
+    },
+    {
+      "epoch": 0.765135945554395,
+      "grad_norm": 4.937138907995441,
+      "learning_rate": 1.5879443986745678e-07,
+      "loss": 1.709,
+      "step": 11355
+    },
+    {
+      "epoch": 0.7654728614265018,
+      "grad_norm": 4.623024164988955,
+      "learning_rate": 1.583648306479901e-07,
+      "loss": 1.6697,
+      "step": 11360
+    },
+    {
+      "epoch": 0.7658097772986086,
+      "grad_norm": 4.8931561862953865,
+      "learning_rate": 1.5793569398036032e-07,
+      "loss": 1.7047,
+      "step": 11365
+    },
+    {
+      "epoch": 0.7661466931707153,
+      "grad_norm": 4.33678060112718,
+      "learning_rate": 1.57507030458152e-07,
+      "loss": 1.6264,
+      "step": 11370
+    },
+    {
+      "epoch": 0.766483609042822,
+      "grad_norm": 4.519741492496316,
+      "learning_rate": 1.5707884067429471e-07,
+      "loss": 1.7615,
+      "step": 11375
+    },
+    {
+      "epoch": 0.7668205249149287,
+      "grad_norm": 4.4632645139705085,
+      "learning_rate": 1.566511252210635e-07,
+      "loss": 1.645,
+      "step": 11380
+    },
+    {
+      "epoch": 0.7671574407870355,
+      "grad_norm": 4.230664086216158,
+      "learning_rate": 1.5622388469007696e-07,
+      "loss": 1.6724,
+      "step": 11385
+    },
+    {
+      "epoch": 0.7674943566591422,
+      "grad_norm": 4.568825730112045,
+      "learning_rate": 1.5579711967229652e-07,
+      "loss": 1.6245,
+      "step": 11390
+    },
+    {
+      "epoch": 0.7678312725312489,
+      "grad_norm": 4.781519886710452,
+      "learning_rate": 1.5537083075802647e-07,
+      "loss": 1.7008,
+      "step": 11395
+    },
+    {
+      "epoch": 0.7681681884033557,
+      "grad_norm": 4.4330956754887705,
+      "learning_rate": 1.5494501853691195e-07,
+      "loss": 1.6979,
+      "step": 11400
+    },
+    {
+      "epoch": 0.7685051042754624,
+      "grad_norm": 5.104008891311832,
+      "learning_rate": 1.5451968359793927e-07,
+      "loss": 1.6358,
+      "step": 11405
+    },
+    {
+      "epoch": 0.7688420201475692,
+      "grad_norm": 4.526648061418248,
+      "learning_rate": 1.5409482652943396e-07,
+      "loss": 1.641,
+      "step": 11410
+    },
+    {
+      "epoch": 0.7691789360196759,
+      "grad_norm": 4.78227432726269,
+      "learning_rate": 1.536704479190611e-07,
+      "loss": 1.6142,
+      "step": 11415
+    },
+    {
+      "epoch": 0.7695158518917826,
+      "grad_norm": 4.3575853373097155,
+      "learning_rate": 1.5324654835382384e-07,
+      "loss": 1.7033,
+      "step": 11420
+    },
+    {
+      "epoch": 0.7698527677638893,
+      "grad_norm": 4.574266836550961,
+      "learning_rate": 1.5282312842006238e-07,
+      "loss": 1.6819,
+      "step": 11425
+    },
+    {
+      "epoch": 0.770189683635996,
+      "grad_norm": 4.453632511820118,
+      "learning_rate": 1.5240018870345388e-07,
+      "loss": 1.6517,
+      "step": 11430
+    },
+    {
+      "epoch": 0.7705265995081029,
+      "grad_norm": 4.8945122291555005,
+      "learning_rate": 1.519777297890113e-07,
+      "loss": 1.6195,
+      "step": 11435
+    },
+    {
+      "epoch": 0.7708635153802096,
+      "grad_norm": 5.369730796017658,
+      "learning_rate": 1.5155575226108198e-07,
+      "loss": 1.6486,
+      "step": 11440
+    },
+    {
+      "epoch": 0.7712004312523163,
+      "grad_norm": 4.538548054640214,
+      "learning_rate": 1.51134256703348e-07,
+      "loss": 1.7399,
+      "step": 11445
+    },
+    {
+      "epoch": 0.771537347124423,
+      "grad_norm": 4.460016295353051,
+      "learning_rate": 1.5071324369882478e-07,
+      "loss": 1.6284,
+      "step": 11450
+    },
+    {
+      "epoch": 0.7718742629965297,
+      "grad_norm": 4.738163460247311,
+      "learning_rate": 1.5029271382985964e-07,
+      "loss": 1.683,
+      "step": 11455
+    },
+    {
+      "epoch": 0.7722111788686365,
+      "grad_norm": 4.641725415440229,
+      "learning_rate": 1.498726676781323e-07,
+      "loss": 1.7271,
+      "step": 11460
+    },
+    {
+      "epoch": 0.7725480947407433,
+      "grad_norm": 4.921504012646379,
+      "learning_rate": 1.4945310582465327e-07,
+      "loss": 1.7093,
+      "step": 11465
+    },
+    {
+      "epoch": 0.77288501061285,
+      "grad_norm": 4.554110504574998,
+      "learning_rate": 1.4903402884976262e-07,
+      "loss": 1.7071,
+      "step": 11470
+    },
+    {
+      "epoch": 0.7732219264849567,
+      "grad_norm": 5.112272763733612,
+      "learning_rate": 1.4861543733313065e-07,
+      "loss": 1.6968,
+      "step": 11475
+    },
+    {
+      "epoch": 0.7735588423570634,
+      "grad_norm": 4.556784459285417,
+      "learning_rate": 1.4819733185375531e-07,
+      "loss": 1.7668,
+      "step": 11480
+    },
+    {
+      "epoch": 0.7738957582291702,
+      "grad_norm": 4.508391558534176,
+      "learning_rate": 1.4777971298996288e-07,
+      "loss": 1.6464,
+      "step": 11485
+    },
+    {
+      "epoch": 0.7742326741012769,
+      "grad_norm": 4.411711682730823,
+      "learning_rate": 1.4736258131940605e-07,
+      "loss": 1.6264,
+      "step": 11490
+    },
+    {
+      "epoch": 0.7745695899733837,
+      "grad_norm": 5.015932123347625,
+      "learning_rate": 1.4694593741906403e-07,
+      "loss": 1.7101,
+      "step": 11495
+    },
+    {
+      "epoch": 0.7749065058454904,
+      "grad_norm": 4.787863644165874,
+      "learning_rate": 1.4652978186524135e-07,
+      "loss": 1.6964,
+      "step": 11500
+    },
+    {
+      "epoch": 0.7752434217175971,
+      "grad_norm": 4.2832899134958815,
+      "learning_rate": 1.4611411523356653e-07,
+      "loss": 1.6985,
+      "step": 11505
+    },
+    {
+      "epoch": 0.7755803375897039,
+      "grad_norm": 4.406066609439304,
+      "learning_rate": 1.4569893809899242e-07,
+      "loss": 1.7463,
+      "step": 11510
+    },
+    {
+      "epoch": 0.7759172534618106,
+      "grad_norm": 4.635127474132326,
+      "learning_rate": 1.452842510357946e-07,
+      "loss": 1.6584,
+      "step": 11515
+    },
+    {
+      "epoch": 0.7762541693339173,
+      "grad_norm": 4.433167172796796,
+      "learning_rate": 1.4487005461757051e-07,
+      "loss": 1.6375,
+      "step": 11520
+    },
+    {
+      "epoch": 0.776591085206024,
+      "grad_norm": 4.447787680123461,
+      "learning_rate": 1.4445634941723927e-07,
+      "loss": 1.6763,
+      "step": 11525
+    },
+    {
+      "epoch": 0.7769280010781308,
+      "grad_norm": 4.5157465552949505,
+      "learning_rate": 1.4404313600704054e-07,
+      "loss": 1.7647,
+      "step": 11530
+    },
+    {
+      "epoch": 0.7772649169502375,
+      "grad_norm": 4.968898885521456,
+      "learning_rate": 1.4363041495853334e-07,
+      "loss": 1.6746,
+      "step": 11535
+    },
+    {
+      "epoch": 0.7776018328223443,
+      "grad_norm": 4.778289663537236,
+      "learning_rate": 1.4321818684259607e-07,
+      "loss": 1.7011,
+      "step": 11540
+    },
+    {
+      "epoch": 0.777938748694451,
+      "grad_norm": 4.433981527727124,
+      "learning_rate": 1.4280645222942535e-07,
+      "loss": 1.6845,
+      "step": 11545
+    },
+    {
+      "epoch": 0.7782756645665577,
+      "grad_norm": 4.404414874555339,
+      "learning_rate": 1.4239521168853458e-07,
+      "loss": 1.6378,
+      "step": 11550
+    },
+    {
+      "epoch": 0.7786125804386644,
+      "grad_norm": 4.4047141653845046,
+      "learning_rate": 1.4198446578875444e-07,
+      "loss": 1.6676,
+      "step": 11555
+    },
+    {
+      "epoch": 0.7789494963107712,
+      "grad_norm": 4.5657660248855505,
+      "learning_rate": 1.4157421509823119e-07,
+      "loss": 1.6269,
+      "step": 11560
+    },
+    {
+      "epoch": 0.779286412182878,
+      "grad_norm": 4.45762595000863,
+      "learning_rate": 1.4116446018442608e-07,
+      "loss": 1.5861,
+      "step": 11565
+    },
+    {
+      "epoch": 0.7796233280549847,
+      "grad_norm": 5.183066557912235,
+      "learning_rate": 1.4075520161411425e-07,
+      "loss": 1.6269,
+      "step": 11570
+    },
+    {
+      "epoch": 0.7799602439270914,
+      "grad_norm": 4.763250474607571,
+      "learning_rate": 1.403464399533849e-07,
+      "loss": 1.6923,
+      "step": 11575
+    },
+    {
+      "epoch": 0.7802971597991981,
+      "grad_norm": 4.606123422875513,
+      "learning_rate": 1.3993817576763983e-07,
+      "loss": 1.7652,
+      "step": 11580
+    },
+    {
+      "epoch": 0.7806340756713048,
+      "grad_norm": 4.8621835506838496,
+      "learning_rate": 1.3953040962159207e-07,
+      "loss": 1.686,
+      "step": 11585
+    },
+    {
+      "epoch": 0.7809709915434117,
+      "grad_norm": 4.839325519832752,
+      "learning_rate": 1.3912314207926657e-07,
+      "loss": 1.6517,
+      "step": 11590
+    },
+    {
+      "epoch": 0.7813079074155184,
+      "grad_norm": 4.441109434606644,
+      "learning_rate": 1.3871637370399824e-07,
+      "loss": 1.6852,
+      "step": 11595
+    },
+    {
+      "epoch": 0.7816448232876251,
+      "grad_norm": 4.643129780812165,
+      "learning_rate": 1.3831010505843139e-07,
+      "loss": 1.6516,
+      "step": 11600
+    },
+    {
+      "epoch": 0.7819817391597318,
+      "grad_norm": 4.509617088988801,
+      "learning_rate": 1.3790433670451927e-07,
+      "loss": 1.6263,
+      "step": 11605
+    },
+    {
+      "epoch": 0.7823186550318385,
+      "grad_norm": 4.4981983354573245,
+      "learning_rate": 1.374990692035235e-07,
+      "loss": 1.6663,
+      "step": 11610
+    },
+    {
+      "epoch": 0.7826555709039453,
+      "grad_norm": 4.709987900224131,
+      "learning_rate": 1.3709430311601205e-07,
+      "loss": 1.6481,
+      "step": 11615
+    },
+    {
+      "epoch": 0.782992486776052,
+      "grad_norm": 4.519779682868838,
+      "learning_rate": 1.366900390018601e-07,
+      "loss": 1.5717,
+      "step": 11620
+    },
+    {
+      "epoch": 0.7833294026481588,
+      "grad_norm": 4.5215699973254395,
+      "learning_rate": 1.3628627742024812e-07,
+      "loss": 1.738,
+      "step": 11625
+    },
+    {
+      "epoch": 0.7836663185202655,
+      "grad_norm": 5.250318064412388,
+      "learning_rate": 1.3588301892966182e-07,
+      "loss": 1.5477,
+      "step": 11630
+    },
+    {
+      "epoch": 0.7840032343923722,
+      "grad_norm": 4.88331049164947,
+      "learning_rate": 1.3548026408789044e-07,
+      "loss": 1.7003,
+      "step": 11635
+    },
+    {
+      "epoch": 0.784340150264479,
+      "grad_norm": 4.931461361213741,
+      "learning_rate": 1.350780134520272e-07,
+      "loss": 1.7331,
+      "step": 11640
+    },
+    {
+      "epoch": 0.7846770661365857,
+      "grad_norm": 4.53955275256855,
+      "learning_rate": 1.3467626757846733e-07,
+      "loss": 1.6507,
+      "step": 11645
+    },
+    {
+      "epoch": 0.7850139820086924,
+      "grad_norm": 4.583660293606964,
+      "learning_rate": 1.342750270229085e-07,
+      "loss": 1.6317,
+      "step": 11650
+    },
+    {
+      "epoch": 0.7853508978807991,
+      "grad_norm": 4.975257250805664,
+      "learning_rate": 1.338742923403487e-07,
+      "loss": 1.5883,
+      "step": 11655
+    },
+    {
+      "epoch": 0.7856878137529059,
+      "grad_norm": 5.910548010361466,
+      "learning_rate": 1.3347406408508694e-07,
+      "loss": 1.6157,
+      "step": 11660
+    },
+    {
+      "epoch": 0.7860247296250127,
+      "grad_norm": 4.495450387975911,
+      "learning_rate": 1.3307434281072106e-07,
+      "loss": 1.7168,
+      "step": 11665
+    },
+    {
+      "epoch": 0.7863616454971194,
+      "grad_norm": 4.871421983989498,
+      "learning_rate": 1.326751290701481e-07,
+      "loss": 1.6807,
+      "step": 11670
+    },
+    {
+      "epoch": 0.7866985613692261,
+      "grad_norm": 4.3982658790692,
+      "learning_rate": 1.3227642341556306e-07,
+      "loss": 1.703,
+      "step": 11675
+    },
+    {
+      "epoch": 0.7870354772413328,
+      "grad_norm": 4.295004600481432,
+      "learning_rate": 1.318782263984577e-07,
+      "loss": 1.6987,
+      "step": 11680
+    },
+    {
+      "epoch": 0.7873723931134395,
+      "grad_norm": 4.58100374784792,
+      "learning_rate": 1.314805385696207e-07,
+      "loss": 1.6905,
+      "step": 11685
+    },
+    {
+      "epoch": 0.7877093089855464,
+      "grad_norm": 4.834714800486215,
+      "learning_rate": 1.3108336047913633e-07,
+      "loss": 1.7305,
+      "step": 11690
+    },
+    {
+      "epoch": 0.7880462248576531,
+      "grad_norm": 4.328656303402874,
+      "learning_rate": 1.3068669267638377e-07,
+      "loss": 1.6603,
+      "step": 11695
+    },
+    {
+      "epoch": 0.7883831407297598,
+      "grad_norm": 4.512299890876141,
+      "learning_rate": 1.3029053571003619e-07,
+      "loss": 1.6928,
+      "step": 11700
+    },
+    {
+      "epoch": 0.7887200566018665,
+      "grad_norm": 4.760951108743499,
+      "learning_rate": 1.2989489012806033e-07,
+      "loss": 1.7068,
+      "step": 11705
+    },
+    {
+      "epoch": 0.7890569724739732,
+      "grad_norm": 5.381231127812247,
+      "learning_rate": 1.294997564777157e-07,
+      "loss": 1.6708,
+      "step": 11710
+    },
+    {
+      "epoch": 0.7893938883460799,
+      "grad_norm": 4.90579764808831,
+      "learning_rate": 1.291051353055534e-07,
+      "loss": 1.7308,
+      "step": 11715
+    },
+    {
+      "epoch": 0.7897308042181868,
+      "grad_norm": 4.641833152824364,
+      "learning_rate": 1.28711027157416e-07,
+      "loss": 1.6489,
+      "step": 11720
+    },
+    {
+      "epoch": 0.7900677200902935,
+      "grad_norm": 4.5697377870925235,
+      "learning_rate": 1.2831743257843597e-07,
+      "loss": 1.6771,
+      "step": 11725
+    },
+    {
+      "epoch": 0.7904046359624002,
+      "grad_norm": 4.638576200215656,
+      "learning_rate": 1.279243521130361e-07,
+      "loss": 1.6553,
+      "step": 11730
+    },
+    {
+      "epoch": 0.7907415518345069,
+      "grad_norm": 4.898438222833477,
+      "learning_rate": 1.2753178630492733e-07,
+      "loss": 1.7493,
+      "step": 11735
+    },
+    {
+      "epoch": 0.7910784677066136,
+      "grad_norm": 4.8068008257623696,
+      "learning_rate": 1.271397356971094e-07,
+      "loss": 1.6344,
+      "step": 11740
+    },
+    {
+      "epoch": 0.7914153835787204,
+      "grad_norm": 5.032163072622715,
+      "learning_rate": 1.267482008318687e-07,
+      "loss": 1.6641,
+      "step": 11745
+    },
+    {
+      "epoch": 0.7917522994508271,
+      "grad_norm": 4.739291643331048,
+      "learning_rate": 1.2635718225077884e-07,
+      "loss": 1.7942,
+      "step": 11750
+    },
+    {
+      "epoch": 0.7920892153229339,
+      "grad_norm": 4.620255847392547,
+      "learning_rate": 1.259666804946991e-07,
+      "loss": 1.6376,
+      "step": 11755
+    },
+    {
+      "epoch": 0.7924261311950406,
+      "grad_norm": 4.826991904490677,
+      "learning_rate": 1.2557669610377397e-07,
+      "loss": 1.6789,
+      "step": 11760
+    },
+    {
+      "epoch": 0.7927630470671473,
+      "grad_norm": 4.653677114654714,
+      "learning_rate": 1.25187229617432e-07,
+      "loss": 1.7308,
+      "step": 11765
+    },
+    {
+      "epoch": 0.7930999629392541,
+      "grad_norm": 4.354166048350967,
+      "learning_rate": 1.247982815743857e-07,
+      "loss": 1.6842,
+      "step": 11770
+    },
+    {
+      "epoch": 0.7934368788113608,
+      "grad_norm": 4.791557277404768,
+      "learning_rate": 1.2440985251263054e-07,
+      "loss": 1.6012,
+      "step": 11775
+    },
+    {
+      "epoch": 0.7937737946834675,
+      "grad_norm": 4.593305862093821,
+      "learning_rate": 1.2402194296944363e-07,
+      "loss": 1.6843,
+      "step": 11780
+    },
+    {
+      "epoch": 0.7941107105555743,
+      "grad_norm": 4.673979764626909,
+      "learning_rate": 1.236345534813839e-07,
+      "loss": 1.6214,
+      "step": 11785
+    },
+    {
+      "epoch": 0.794447626427681,
+      "grad_norm": 4.479878569305674,
+      "learning_rate": 1.2324768458429107e-07,
+      "loss": 1.6624,
+      "step": 11790
+    },
+    {
+      "epoch": 0.7947845422997878,
+      "grad_norm": 4.945547685723316,
+      "learning_rate": 1.228613368132842e-07,
+      "loss": 1.7297,
+      "step": 11795
+    },
+    {
+      "epoch": 0.7951214581718945,
+      "grad_norm": 4.36357357134437,
+      "learning_rate": 1.2247551070276207e-07,
+      "loss": 1.5405,
+      "step": 11800
+    },
+    {
+      "epoch": 0.7954583740440012,
+      "grad_norm": 4.782501621716016,
+      "learning_rate": 1.2209020678640176e-07,
+      "loss": 1.743,
+      "step": 11805
+    },
+    {
+      "epoch": 0.7957952899161079,
+      "grad_norm": 4.780308924663592,
+      "learning_rate": 1.2170542559715775e-07,
+      "loss": 1.6265,
+      "step": 11810
+    },
+    {
+      "epoch": 0.7961322057882146,
+      "grad_norm": 4.7301440980568215,
+      "learning_rate": 1.2132116766726196e-07,
+      "loss": 1.6937,
+      "step": 11815
+    },
+    {
+      "epoch": 0.7964691216603215,
+      "grad_norm": 4.7183597873504795,
+      "learning_rate": 1.2093743352822206e-07,
+      "loss": 1.6882,
+      "step": 11820
+    },
+    {
+      "epoch": 0.7968060375324282,
+      "grad_norm": 5.4563281880583165,
+      "learning_rate": 1.2055422371082168e-07,
+      "loss": 1.6439,
+      "step": 11825
+    },
+    {
+      "epoch": 0.7971429534045349,
+      "grad_norm": 4.558590412599557,
+      "learning_rate": 1.2017153874511865e-07,
+      "loss": 1.6541,
+      "step": 11830
+    },
+    {
+      "epoch": 0.7974798692766416,
+      "grad_norm": 4.769716161730735,
+      "learning_rate": 1.1978937916044534e-07,
+      "loss": 1.6999,
+      "step": 11835
+    },
+    {
+      "epoch": 0.7978167851487483,
+      "grad_norm": 4.766641399567675,
+      "learning_rate": 1.1940774548540733e-07,
+      "loss": 1.6257,
+      "step": 11840
+    },
+    {
+      "epoch": 0.7981537010208551,
+      "grad_norm": 4.4988420622667755,
+      "learning_rate": 1.1902663824788233e-07,
+      "loss": 1.6595,
+      "step": 11845
+    },
+    {
+      "epoch": 0.7984906168929619,
+      "grad_norm": 4.384373031174729,
+      "learning_rate": 1.1864605797502031e-07,
+      "loss": 1.6879,
+      "step": 11850
+    },
+    {
+      "epoch": 0.7988275327650686,
+      "grad_norm": 4.172777385405688,
+      "learning_rate": 1.1826600519324237e-07,
+      "loss": 1.7245,
+      "step": 11855
+    },
+    {
+      "epoch": 0.7991644486371753,
+      "grad_norm": 4.817536572367274,
+      "learning_rate": 1.1788648042823956e-07,
+      "loss": 1.6383,
+      "step": 11860
+    },
+    {
+      "epoch": 0.799501364509282,
+      "grad_norm": 4.74080247539222,
+      "learning_rate": 1.1750748420497298e-07,
+      "loss": 1.635,
+      "step": 11865
+    },
+    {
+      "epoch": 0.7998382803813888,
+      "grad_norm": 4.966353394973762,
+      "learning_rate": 1.1712901704767253e-07,
+      "loss": 1.649,
+      "step": 11870
+    },
+    {
+      "epoch": 0.8001751962534955,
+      "grad_norm": 4.347264494303227,
+      "learning_rate": 1.1675107947983615e-07,
+      "loss": 1.698,
+      "step": 11875
+    },
+    {
+      "epoch": 0.8005121121256022,
+      "grad_norm": 4.573477117456977,
+      "learning_rate": 1.1637367202422943e-07,
+      "loss": 1.7489,
+      "step": 11880
+    },
+    {
+      "epoch": 0.800849027997709,
+      "grad_norm": 4.351379168903883,
+      "learning_rate": 1.159967952028848e-07,
+      "loss": 1.7147,
+      "step": 11885
+    },
+    {
+      "epoch": 0.8011859438698157,
+      "grad_norm": 4.5742306257265755,
+      "learning_rate": 1.1562044953710032e-07,
+      "loss": 1.5991,
+      "step": 11890
+    },
+    {
+      "epoch": 0.8015228597419224,
+      "grad_norm": 4.24646806235061,
+      "learning_rate": 1.152446355474398e-07,
+      "loss": 1.6763,
+      "step": 11895
+    },
+    {
+      "epoch": 0.8018597756140292,
+      "grad_norm": 5.2105463873461035,
+      "learning_rate": 1.1486935375373124e-07,
+      "loss": 1.7155,
+      "step": 11900
+    },
+    {
+      "epoch": 0.8021966914861359,
+      "grad_norm": 5.47357024870505,
+      "learning_rate": 1.1449460467506689e-07,
+      "loss": 1.6485,
+      "step": 11905
+    },
+    {
+      "epoch": 0.8025336073582426,
+      "grad_norm": 4.295231435421802,
+      "learning_rate": 1.1412038882980174e-07,
+      "loss": 1.6926,
+      "step": 11910
+    },
+    {
+      "epoch": 0.8028705232303494,
+      "grad_norm": 4.339085780781735,
+      "learning_rate": 1.1374670673555348e-07,
+      "loss": 1.6742,
+      "step": 11915
+    },
+    {
+      "epoch": 0.8032074391024561,
+      "grad_norm": 4.371106160876894,
+      "learning_rate": 1.1337355890920169e-07,
+      "loss": 1.7035,
+      "step": 11920
+    },
+    {
+      "epoch": 0.8035443549745629,
+      "grad_norm": 4.112654956639733,
+      "learning_rate": 1.130009458668863e-07,
+      "loss": 1.6723,
+      "step": 11925
+    },
+    {
+      "epoch": 0.8038812708466696,
+      "grad_norm": 4.500666264030187,
+      "learning_rate": 1.1262886812400813e-07,
+      "loss": 1.6475,
+      "step": 11930
+    },
+    {
+      "epoch": 0.8042181867187763,
+      "grad_norm": 4.561129079706414,
+      "learning_rate": 1.1225732619522754e-07,
+      "loss": 1.6282,
+      "step": 11935
+    },
+    {
+      "epoch": 0.804555102590883,
+      "grad_norm": 4.683667511801432,
+      "learning_rate": 1.118863205944633e-07,
+      "loss": 1.6257,
+      "step": 11940
+    },
+    {
+      "epoch": 0.8048920184629897,
+      "grad_norm": 5.150745324206002,
+      "learning_rate": 1.1151585183489266e-07,
+      "loss": 1.6967,
+      "step": 11945
+    },
+    {
+      "epoch": 0.8052289343350966,
+      "grad_norm": 4.535603700059998,
+      "learning_rate": 1.1114592042895044e-07,
+      "loss": 1.6512,
+      "step": 11950
+    },
+    {
+      "epoch": 0.8055658502072033,
+      "grad_norm": 4.697947520091324,
+      "learning_rate": 1.1077652688832772e-07,
+      "loss": 1.672,
+      "step": 11955
+    },
+    {
+      "epoch": 0.80590276607931,
+      "grad_norm": 4.58900673943732,
+      "learning_rate": 1.1040767172397209e-07,
+      "loss": 1.6287,
+      "step": 11960
+    },
+    {
+      "epoch": 0.8062396819514167,
+      "grad_norm": 4.697800139739896,
+      "learning_rate": 1.1003935544608612e-07,
+      "loss": 1.6606,
+      "step": 11965
+    },
+    {
+      "epoch": 0.8065765978235234,
+      "grad_norm": 4.517313865394391,
+      "learning_rate": 1.0967157856412739e-07,
+      "loss": 1.6497,
+      "step": 11970
+    },
+    {
+      "epoch": 0.8069135136956302,
+      "grad_norm": 4.488096600110159,
+      "learning_rate": 1.093043415868069e-07,
+      "loss": 1.7794,
+      "step": 11975
+    },
+    {
+      "epoch": 0.807250429567737,
+      "grad_norm": 4.441004794884173,
+      "learning_rate": 1.0893764502208891e-07,
+      "loss": 1.7114,
+      "step": 11980
+    },
+    {
+      "epoch": 0.8075873454398437,
+      "grad_norm": 4.795206946774002,
+      "learning_rate": 1.0857148937719063e-07,
+      "loss": 1.6427,
+      "step": 11985
+    },
+    {
+      "epoch": 0.8079242613119504,
+      "grad_norm": 4.659472879632121,
+      "learning_rate": 1.0820587515858054e-07,
+      "loss": 1.6943,
+      "step": 11990
+    },
+    {
+      "epoch": 0.8082611771840571,
+      "grad_norm": 4.606554652971352,
+      "learning_rate": 1.078408028719785e-07,
+      "loss": 1.6448,
+      "step": 11995
+    },
+    {
+      "epoch": 0.8085980930561639,
+      "grad_norm": 4.291101869310125,
+      "learning_rate": 1.0747627302235491e-07,
+      "loss": 1.7005,
+      "step": 12000
+    },
+    {
+      "epoch": 0.8089350089282706,
+      "grad_norm": 4.841403602767737,
+      "learning_rate": 1.0711228611392936e-07,
+      "loss": 1.7057,
+      "step": 12005
+    },
+    {
+      "epoch": 0.8092719248003774,
+      "grad_norm": 4.756702586221024,
+      "learning_rate": 1.0674884265017086e-07,
+      "loss": 1.6921,
+      "step": 12010
+    },
+    {
+      "epoch": 0.8096088406724841,
+      "grad_norm": 4.727776169377255,
+      "learning_rate": 1.0638594313379678e-07,
+      "loss": 1.5935,
+      "step": 12015
+    },
+    {
+      "epoch": 0.8099457565445908,
+      "grad_norm": 4.685242195126687,
+      "learning_rate": 1.060235880667717e-07,
+      "loss": 1.7255,
+      "step": 12020
+    },
+    {
+      "epoch": 0.8102826724166976,
+      "grad_norm": 4.314878206791833,
+      "learning_rate": 1.056617779503074e-07,
+      "loss": 1.6571,
+      "step": 12025
+    },
+    {
+      "epoch": 0.8106195882888043,
+      "grad_norm": 5.029269653182732,
+      "learning_rate": 1.053005132848619e-07,
+      "loss": 1.675,
+      "step": 12030
+    },
+    {
+      "epoch": 0.810956504160911,
+      "grad_norm": 4.342726447389877,
+      "learning_rate": 1.0493979457013874e-07,
+      "loss": 1.7437,
+      "step": 12035
+    },
+    {
+      "epoch": 0.8112934200330177,
+      "grad_norm": 4.235853993662135,
+      "learning_rate": 1.0457962230508599e-07,
+      "loss": 1.6567,
+      "step": 12040
+    },
+    {
+      "epoch": 0.8116303359051245,
+      "grad_norm": 4.687902595868502,
+      "learning_rate": 1.042199969878963e-07,
+      "loss": 1.6397,
+      "step": 12045
+    },
+    {
+      "epoch": 0.8119672517772313,
+      "grad_norm": 4.520022707269915,
+      "learning_rate": 1.0386091911600564e-07,
+      "loss": 1.6126,
+      "step": 12050
+    },
+    {
+      "epoch": 0.812304167649338,
+      "grad_norm": 4.744073377652037,
+      "learning_rate": 1.0350238918609244e-07,
+      "loss": 1.6284,
+      "step": 12055
+    },
+    {
+      "epoch": 0.8126410835214447,
+      "grad_norm": 4.328349619953094,
+      "learning_rate": 1.0314440769407784e-07,
+      "loss": 1.6412,
+      "step": 12060
+    },
+    {
+      "epoch": 0.8129779993935514,
+      "grad_norm": 4.867324986966181,
+      "learning_rate": 1.0278697513512375e-07,
+      "loss": 1.5896,
+      "step": 12065
+    },
+    {
+      "epoch": 0.8133149152656581,
+      "grad_norm": 4.7885057818111365,
+      "learning_rate": 1.0243009200363295e-07,
+      "loss": 1.5919,
+      "step": 12070
+    },
+    {
+      "epoch": 0.8136518311377648,
+      "grad_norm": 5.090797412053734,
+      "learning_rate": 1.0207375879324853e-07,
+      "loss": 1.6436,
+      "step": 12075
+    },
+    {
+      "epoch": 0.8139887470098717,
+      "grad_norm": 4.920044634686522,
+      "learning_rate": 1.0171797599685284e-07,
+      "loss": 1.6884,
+      "step": 12080
+    },
+    {
+      "epoch": 0.8143256628819784,
+      "grad_norm": 4.217021254878892,
+      "learning_rate": 1.0136274410656653e-07,
+      "loss": 1.689,
+      "step": 12085
+    },
+    {
+      "epoch": 0.8146625787540851,
+      "grad_norm": 4.706327392597494,
+      "learning_rate": 1.010080636137487e-07,
+      "loss": 1.6614,
+      "step": 12090
+    },
+    {
+      "epoch": 0.8149994946261918,
+      "grad_norm": 4.917225692982013,
+      "learning_rate": 1.0065393500899549e-07,
+      "loss": 1.5856,
+      "step": 12095
+    },
+    {
+      "epoch": 0.8153364104982985,
+      "grad_norm": 4.909986128400832,
+      "learning_rate": 1.0030035878213988e-07,
+      "loss": 1.6692,
+      "step": 12100
+    },
+    {
+      "epoch": 0.8156733263704053,
+      "grad_norm": 4.634205918538887,
+      "learning_rate": 9.994733542225037e-08,
+      "loss": 1.6299,
+      "step": 12105
+    },
+    {
+      "epoch": 0.8160102422425121,
+      "grad_norm": 4.313804535342684,
+      "learning_rate": 9.959486541763118e-08,
+      "loss": 1.6404,
+      "step": 12110
+    },
+    {
+      "epoch": 0.8163471581146188,
+      "grad_norm": 4.680390901578063,
+      "learning_rate": 9.924294925582105e-08,
+      "loss": 1.6987,
+      "step": 12115
+    },
+    {
+      "epoch": 0.8166840739867255,
+      "grad_norm": 4.753692254136698,
+      "learning_rate": 9.889158742359233e-08,
+      "loss": 1.6577,
+      "step": 12120
+    },
+    {
+      "epoch": 0.8170209898588322,
+      "grad_norm": 4.5693819665720525,
+      "learning_rate": 9.854078040695096e-08,
+      "loss": 1.6935,
+      "step": 12125
+    },
+    {
+      "epoch": 0.817357905730939,
+      "grad_norm": 4.549210623708017,
+      "learning_rate": 9.819052869113543e-08,
+      "loss": 1.6723,
+      "step": 12130
+    },
+    {
+      "epoch": 0.8176948216030457,
+      "grad_norm": 4.8659961093999975,
+      "learning_rate": 9.784083276061578e-08,
+      "loss": 1.6669,
+      "step": 12135
+    },
+    {
+      "epoch": 0.8180317374751525,
+      "grad_norm": 5.086215136308865,
+      "learning_rate": 9.749169309909382e-08,
+      "loss": 1.6106,
+      "step": 12140
+    },
+    {
+      "epoch": 0.8183686533472592,
+      "grad_norm": 4.46335613876711,
+      "learning_rate": 9.714311018950139e-08,
+      "loss": 1.6771,
+      "step": 12145
+    },
+    {
+      "epoch": 0.8187055692193659,
+      "grad_norm": 4.577487028769546,
+      "learning_rate": 9.67950845140007e-08,
+      "loss": 1.7008,
+      "step": 12150
+    },
+    {
+      "epoch": 0.8190424850914727,
+      "grad_norm": 4.742273478717242,
+      "learning_rate": 9.644761655398276e-08,
+      "loss": 1.6126,
+      "step": 12155
+    },
+    {
+      "epoch": 0.8193794009635794,
+      "grad_norm": 5.175051265752663,
+      "learning_rate": 9.61007067900675e-08,
+      "loss": 1.7169,
+      "step": 12160
+    },
+    {
+      "epoch": 0.8197163168356861,
+      "grad_norm": 4.305291595947235,
+      "learning_rate": 9.575435570210266e-08,
+      "loss": 1.6949,
+      "step": 12165
+    },
+    {
+      "epoch": 0.8200532327077928,
+      "grad_norm": 4.412018556271386,
+      "learning_rate": 9.5408563769163e-08,
+      "loss": 1.6687,
+      "step": 12170
+    },
+    {
+      "epoch": 0.8203901485798996,
+      "grad_norm": 4.170303077983749,
+      "learning_rate": 9.506333146955009e-08,
+      "loss": 1.5886,
+      "step": 12175
+    },
+    {
+      "epoch": 0.8207270644520064,
+      "grad_norm": 4.287215524208355,
+      "learning_rate": 9.471865928079148e-08,
+      "loss": 1.6817,
+      "step": 12180
+    },
+    {
+      "epoch": 0.8210639803241131,
+      "grad_norm": 4.1863139310236805,
+      "learning_rate": 9.437454767963954e-08,
+      "loss": 1.6514,
+      "step": 12185
+    },
+    {
+      "epoch": 0.8214008961962198,
+      "grad_norm": 4.782584978988166,
+      "learning_rate": 9.403099714207174e-08,
+      "loss": 1.6714,
+      "step": 12190
+    },
+    {
+      "epoch": 0.8217378120683265,
+      "grad_norm": 4.61780885498575,
+      "learning_rate": 9.368800814328931e-08,
+      "loss": 1.6771,
+      "step": 12195
+    },
+    {
+      "epoch": 0.8220747279404332,
+      "grad_norm": 4.920092623434809,
+      "learning_rate": 9.334558115771646e-08,
+      "loss": 1.6746,
+      "step": 12200
+    },
+    {
+      "epoch": 0.8224116438125401,
+      "grad_norm": 4.308366345769101,
+      "learning_rate": 9.300371665900048e-08,
+      "loss": 1.6434,
+      "step": 12205
+    },
+    {
+      "epoch": 0.8227485596846468,
+      "grad_norm": 4.8831688275939,
+      "learning_rate": 9.266241512001044e-08,
+      "loss": 1.5913,
+      "step": 12210
+    },
+    {
+      "epoch": 0.8230854755567535,
+      "grad_norm": 4.686820772021366,
+      "learning_rate": 9.23216770128365e-08,
+      "loss": 1.6709,
+      "step": 12215
+    },
+    {
+      "epoch": 0.8234223914288602,
+      "grad_norm": 4.719416017221441,
+      "learning_rate": 9.19815028087898e-08,
+      "loss": 1.6748,
+      "step": 12220
+    },
+    {
+      "epoch": 0.8237593073009669,
+      "grad_norm": 4.340410251877525,
+      "learning_rate": 9.164189297840147e-08,
+      "loss": 1.6961,
+      "step": 12225
+    },
+    {
+      "epoch": 0.8240962231730737,
+      "grad_norm": 5.11434879075709,
+      "learning_rate": 9.130284799142179e-08,
+      "loss": 1.6409,
+      "step": 12230
+    },
+    {
+      "epoch": 0.8244331390451805,
+      "grad_norm": 4.862919162784708,
+      "learning_rate": 9.09643683168197e-08,
+      "loss": 1.667,
+      "step": 12235
+    },
+    {
+      "epoch": 0.8247700549172872,
+      "grad_norm": 4.439123911206303,
+      "learning_rate": 9.062645442278244e-08,
+      "loss": 1.706,
+      "step": 12240
+    },
+    {
+      "epoch": 0.8251069707893939,
+      "grad_norm": 4.410873781311947,
+      "learning_rate": 9.028910677671469e-08,
+      "loss": 1.7492,
+      "step": 12245
+    },
+    {
+      "epoch": 0.8254438866615006,
+      "grad_norm": 4.661431721051673,
+      "learning_rate": 8.995232584523754e-08,
+      "loss": 1.6371,
+      "step": 12250
+    },
+    {
+      "epoch": 0.8257808025336073,
+      "grad_norm": 4.885583165906465,
+      "learning_rate": 8.961611209418851e-08,
+      "loss": 1.6724,
+      "step": 12255
+    },
+    {
+      "epoch": 0.8261177184057141,
+      "grad_norm": 4.583887928593409,
+      "learning_rate": 8.928046598862065e-08,
+      "loss": 1.6665,
+      "step": 12260
+    },
+    {
+      "epoch": 0.8264546342778208,
+      "grad_norm": 4.62236702480345,
+      "learning_rate": 8.894538799280138e-08,
+      "loss": 1.732,
+      "step": 12265
+    },
+    {
+      "epoch": 0.8267915501499276,
+      "grad_norm": 4.196005535999056,
+      "learning_rate": 8.861087857021282e-08,
+      "loss": 1.6653,
+      "step": 12270
+    },
+    {
+      "epoch": 0.8271284660220343,
+      "grad_norm": 5.14421976931363,
+      "learning_rate": 8.827693818355048e-08,
+      "loss": 1.5841,
+      "step": 12275
+    },
+    {
+      "epoch": 0.827465381894141,
+      "grad_norm": 4.836820671207669,
+      "learning_rate": 8.794356729472252e-08,
+      "loss": 1.6485,
+      "step": 12280
+    },
+    {
+      "epoch": 0.8278022977662478,
+      "grad_norm": 4.303365938631198,
+      "learning_rate": 8.76107663648497e-08,
+      "loss": 1.6251,
+      "step": 12285
+    },
+    {
+      "epoch": 0.8281392136383545,
+      "grad_norm": 4.4132607946935885,
+      "learning_rate": 8.727853585426436e-08,
+      "loss": 1.7011,
+      "step": 12290
+    },
+    {
+      "epoch": 0.8284761295104612,
+      "grad_norm": 4.489080068730756,
+      "learning_rate": 8.694687622250963e-08,
+      "loss": 1.6709,
+      "step": 12295
+    },
+    {
+      "epoch": 0.828813045382568,
+      "grad_norm": 4.431555011379005,
+      "learning_rate": 8.661578792833907e-08,
+      "loss": 1.6917,
+      "step": 12300
+    },
+    {
+      "epoch": 0.8291499612546747,
+      "grad_norm": 4.606480976685833,
+      "learning_rate": 8.628527142971632e-08,
+      "loss": 1.5667,
+      "step": 12305
+    },
+    {
+      "epoch": 0.8294868771267815,
+      "grad_norm": 4.415920989112211,
+      "learning_rate": 8.595532718381338e-08,
+      "loss": 1.6725,
+      "step": 12310
+    },
+    {
+      "epoch": 0.8298237929988882,
+      "grad_norm": 4.577414386363807,
+      "learning_rate": 8.562595564701153e-08,
+      "loss": 1.7033,
+      "step": 12315
+    },
+    {
+      "epoch": 0.8301607088709949,
+      "grad_norm": 4.807819936598261,
+      "learning_rate": 8.529715727489912e-08,
+      "loss": 1.7281,
+      "step": 12320
+    },
+    {
+      "epoch": 0.8304976247431016,
+      "grad_norm": 4.591824431615779,
+      "learning_rate": 8.496893252227238e-08,
+      "loss": 1.6812,
+      "step": 12325
+    },
+    {
+      "epoch": 0.8308345406152083,
+      "grad_norm": 4.232434383612212,
+      "learning_rate": 8.464128184313346e-08,
+      "loss": 1.6564,
+      "step": 12330
+    },
+    {
+      "epoch": 0.8311714564873152,
+      "grad_norm": 4.660216015221876,
+      "learning_rate": 8.431420569069093e-08,
+      "loss": 1.6009,
+      "step": 12335
+    },
+    {
+      "epoch": 0.8315083723594219,
+      "grad_norm": 4.84665822926301,
+      "learning_rate": 8.398770451735865e-08,
+      "loss": 1.6645,
+      "step": 12340
+    },
+    {
+      "epoch": 0.8318452882315286,
+      "grad_norm": 4.5925610308319165,
+      "learning_rate": 8.366177877475473e-08,
+      "loss": 1.6727,
+      "step": 12345
+    },
+    {
+      "epoch": 0.8321822041036353,
+      "grad_norm": 4.273499635590044,
+      "learning_rate": 8.333642891370174e-08,
+      "loss": 1.7604,
+      "step": 12350
+    },
+    {
+      "epoch": 0.832519119975742,
+      "grad_norm": 4.289284690779844,
+      "learning_rate": 8.301165538422577e-08,
+      "loss": 1.6998,
+      "step": 12355
+    },
+    {
+      "epoch": 0.8328560358478488,
+      "grad_norm": 4.708588816998743,
+      "learning_rate": 8.268745863555521e-08,
+      "loss": 1.5891,
+      "step": 12360
+    },
+    {
+      "epoch": 0.8331929517199556,
+      "grad_norm": 4.731731935159316,
+      "learning_rate": 8.236383911612116e-08,
+      "loss": 1.6399,
+      "step": 12365
+    },
+    {
+      "epoch": 0.8335298675920623,
+      "grad_norm": 4.368128769303862,
+      "learning_rate": 8.204079727355611e-08,
+      "loss": 1.6887,
+      "step": 12370
+    },
+    {
+      "epoch": 0.833866783464169,
+      "grad_norm": 5.031103795653992,
+      "learning_rate": 8.171833355469354e-08,
+      "loss": 1.7402,
+      "step": 12375
+    },
+    {
+      "epoch": 0.8342036993362757,
+      "grad_norm": 5.039376453564224,
+      "learning_rate": 8.139644840556703e-08,
+      "loss": 1.6775,
+      "step": 12380
+    },
+    {
+      "epoch": 0.8345406152083825,
+      "grad_norm": 4.652131385190657,
+      "learning_rate": 8.107514227141032e-08,
+      "loss": 1.6316,
+      "step": 12385
+    },
+    {
+      "epoch": 0.8348775310804892,
+      "grad_norm": 4.482192191227387,
+      "learning_rate": 8.075441559665569e-08,
+      "loss": 1.6908,
+      "step": 12390
+    },
+    {
+      "epoch": 0.835214446952596,
+      "grad_norm": 4.178251657515203,
+      "learning_rate": 8.04342688249346e-08,
+      "loss": 1.7173,
+      "step": 12395
+    },
+    {
+      "epoch": 0.8355513628247027,
+      "grad_norm": 4.644346393479015,
+      "learning_rate": 8.011470239907558e-08,
+      "loss": 1.6686,
+      "step": 12400
+    },
+    {
+      "epoch": 0.8358882786968094,
+      "grad_norm": 4.762354253384477,
+      "learning_rate": 7.979571676110525e-08,
+      "loss": 1.6626,
+      "step": 12405
+    },
+    {
+      "epoch": 0.8362251945689162,
+      "grad_norm": 5.008823526360671,
+      "learning_rate": 7.947731235224614e-08,
+      "loss": 1.7438,
+      "step": 12410
+    },
+    {
+      "epoch": 0.8365621104410229,
+      "grad_norm": 4.502314779203351,
+      "learning_rate": 7.915948961291729e-08,
+      "loss": 1.6587,
+      "step": 12415
+    },
+    {
+      "epoch": 0.8368990263131296,
+      "grad_norm": 4.793427388427419,
+      "learning_rate": 7.884224898273322e-08,
+      "loss": 1.6446,
+      "step": 12420
+    },
+    {
+      "epoch": 0.8372359421852363,
+      "grad_norm": 4.939647629786324,
+      "learning_rate": 7.852559090050276e-08,
+      "loss": 1.6378,
+      "step": 12425
+    },
+    {
+      "epoch": 0.837572858057343,
+      "grad_norm": 4.808437181984224,
+      "learning_rate": 7.820951580422952e-08,
+      "loss": 1.6157,
+      "step": 12430
+    },
+    {
+      "epoch": 0.8379097739294498,
+      "grad_norm": 4.546950694277577,
+      "learning_rate": 7.789402413111041e-08,
+      "loss": 1.6062,
+      "step": 12435
+    },
+    {
+      "epoch": 0.8382466898015566,
+      "grad_norm": 4.4368217224106505,
+      "learning_rate": 7.757911631753556e-08,
+      "loss": 1.6451,
+      "step": 12440
+    },
+    {
+      "epoch": 0.8385836056736633,
+      "grad_norm": 4.455951392656074,
+      "learning_rate": 7.72647927990871e-08,
+      "loss": 1.7056,
+      "step": 12445
+    },
+    {
+      "epoch": 0.83892052154577,
+      "grad_norm": 4.4739168194216195,
+      "learning_rate": 7.695105401053942e-08,
+      "loss": 1.7121,
+      "step": 12450
+    },
+    {
+      "epoch": 0.8392574374178767,
+      "grad_norm": 4.440868666208389,
+      "learning_rate": 7.663790038585794e-08,
+      "loss": 1.7122,
+      "step": 12455
+    },
+    {
+      "epoch": 0.8395943532899834,
+      "grad_norm": 4.547711961346844,
+      "learning_rate": 7.63253323581985e-08,
+      "loss": 1.6672,
+      "step": 12460
+    },
+    {
+      "epoch": 0.8399312691620903,
+      "grad_norm": 4.768378256765404,
+      "learning_rate": 7.601335035990714e-08,
+      "loss": 1.7202,
+      "step": 12465
+    },
+    {
+      "epoch": 0.840268185034197,
+      "grad_norm": 4.852415223059402,
+      "learning_rate": 7.57019548225194e-08,
+      "loss": 1.6276,
+      "step": 12470
+    },
+    {
+      "epoch": 0.8406051009063037,
+      "grad_norm": 4.900333256099145,
+      "learning_rate": 7.539114617675941e-08,
+      "loss": 1.6976,
+      "step": 12475
+    },
+    {
+      "epoch": 0.8409420167784104,
+      "grad_norm": 4.267078803636009,
+      "learning_rate": 7.508092485253936e-08,
+      "loss": 1.6408,
+      "step": 12480
+    },
+    {
+      "epoch": 0.8412789326505171,
+      "grad_norm": 4.717401305057401,
+      "learning_rate": 7.477129127895954e-08,
+      "loss": 1.6725,
+      "step": 12485
+    },
+    {
+      "epoch": 0.8416158485226239,
+      "grad_norm": 4.46002079290251,
+      "learning_rate": 7.446224588430678e-08,
+      "loss": 1.6679,
+      "step": 12490
+    },
+    {
+      "epoch": 0.8419527643947307,
+      "grad_norm": 4.269028617491822,
+      "learning_rate": 7.415378909605457e-08,
+      "loss": 1.75,
+      "step": 12495
+    },
+    {
+      "epoch": 0.8422896802668374,
+      "grad_norm": 4.578804117489516,
+      "learning_rate": 7.384592134086231e-08,
+      "loss": 1.6489,
+      "step": 12500
+    },
+    {
+      "epoch": 0.8426265961389441,
+      "grad_norm": 4.81371661431306,
+      "learning_rate": 7.353864304457463e-08,
+      "loss": 1.687,
+      "step": 12505
+    },
+    {
+      "epoch": 0.8429635120110508,
+      "grad_norm": 4.980069333005033,
+      "learning_rate": 7.323195463222054e-08,
+      "loss": 1.6767,
+      "step": 12510
+    },
+    {
+      "epoch": 0.8433004278831576,
+      "grad_norm": 4.524134363057272,
+      "learning_rate": 7.292585652801331e-08,
+      "loss": 1.6008,
+      "step": 12515
+    },
+    {
+      "epoch": 0.8436373437552643,
+      "grad_norm": 4.4534725851542,
+      "learning_rate": 7.262034915534993e-08,
+      "loss": 1.6912,
+      "step": 12520
+    },
+    {
+      "epoch": 0.843974259627371,
+      "grad_norm": 4.747417878677229,
+      "learning_rate": 7.231543293680969e-08,
+      "loss": 1.6438,
+      "step": 12525
+    },
+    {
+      "epoch": 0.8443111754994778,
+      "grad_norm": 5.1122874036227755,
+      "learning_rate": 7.20111082941548e-08,
+      "loss": 1.7049,
+      "step": 12530
+    },
+    {
+      "epoch": 0.8446480913715845,
+      "grad_norm": 4.536409694031237,
+      "learning_rate": 7.170737564832902e-08,
+      "loss": 1.7071,
+      "step": 12535
+    },
+    {
+      "epoch": 0.8449850072436913,
+      "grad_norm": 4.587192664311009,
+      "learning_rate": 7.14042354194569e-08,
+      "loss": 1.6938,
+      "step": 12540
+    },
+    {
+      "epoch": 0.845321923115798,
+      "grad_norm": 4.385212609700637,
+      "learning_rate": 7.110168802684408e-08,
+      "loss": 1.589,
+      "step": 12545
+    },
+    {
+      "epoch": 0.8456588389879047,
+      "grad_norm": 4.395202562503766,
+      "learning_rate": 7.079973388897592e-08,
+      "loss": 1.6925,
+      "step": 12550
+    },
+    {
+      "epoch": 0.8459957548600114,
+      "grad_norm": 4.560065549888014,
+      "learning_rate": 7.049837342351706e-08,
+      "loss": 1.6441,
+      "step": 12555
+    },
+    {
+      "epoch": 0.8463326707321182,
+      "grad_norm": 5.114443613657291,
+      "learning_rate": 7.019760704731131e-08,
+      "loss": 1.716,
+      "step": 12560
+    },
+    {
+      "epoch": 0.846669586604225,
+      "grad_norm": 4.429698035578612,
+      "learning_rate": 6.989743517638053e-08,
+      "loss": 1.6688,
+      "step": 12565
+    },
+    {
+      "epoch": 0.8470065024763317,
+      "grad_norm": 4.620772449380527,
+      "learning_rate": 6.959785822592402e-08,
+      "loss": 1.7159,
+      "step": 12570
+    },
+    {
+      "epoch": 0.8473434183484384,
+      "grad_norm": 4.888061963996254,
+      "learning_rate": 6.929887661031864e-08,
+      "loss": 1.6162,
+      "step": 12575
+    },
+    {
+      "epoch": 0.8476803342205451,
+      "grad_norm": 4.2420582924615795,
+      "learning_rate": 6.900049074311753e-08,
+      "loss": 1.651,
+      "step": 12580
+    },
+    {
+      "epoch": 0.8480172500926518,
+      "grad_norm": 4.839083686871792,
+      "learning_rate": 6.870270103705e-08,
+      "loss": 1.6384,
+      "step": 12585
+    },
+    {
+      "epoch": 0.8483541659647587,
+      "grad_norm": 4.6168931337714465,
+      "learning_rate": 6.840550790402027e-08,
+      "loss": 1.69,
+      "step": 12590
+    },
+    {
+      "epoch": 0.8486910818368654,
+      "grad_norm": 4.487334521520262,
+      "learning_rate": 6.810891175510792e-08,
+      "loss": 1.7088,
+      "step": 12595
+    },
+    {
+      "epoch": 0.8490279977089721,
+      "grad_norm": 4.6057406298661,
+      "learning_rate": 6.781291300056647e-08,
+      "loss": 1.6772,
+      "step": 12600
+    },
+    {
+      "epoch": 0.8493649135810788,
+      "grad_norm": 4.6230805899692236,
+      "learning_rate": 6.751751204982309e-08,
+      "loss": 1.5621,
+      "step": 12605
+    },
+    {
+      "epoch": 0.8497018294531855,
+      "grad_norm": 4.645235329625154,
+      "learning_rate": 6.722270931147827e-08,
+      "loss": 1.6998,
+      "step": 12610
+    },
+    {
+      "epoch": 0.8500387453252922,
+      "grad_norm": 4.556960627434323,
+      "learning_rate": 6.692850519330506e-08,
+      "loss": 1.6397,
+      "step": 12615
+    },
+    {
+      "epoch": 0.850375661197399,
+      "grad_norm": 4.723727212397595,
+      "learning_rate": 6.66349001022481e-08,
+      "loss": 1.5809,
+      "step": 12620
+    },
+    {
+      "epoch": 0.8507125770695058,
+      "grad_norm": 4.276618460688099,
+      "learning_rate": 6.634189444442389e-08,
+      "loss": 1.7389,
+      "step": 12625
+    },
+    {
+      "epoch": 0.8510494929416125,
+      "grad_norm": 4.8765777916342845,
+      "learning_rate": 6.604948862511977e-08,
+      "loss": 1.7438,
+      "step": 12630
+    },
+    {
+      "epoch": 0.8513864088137192,
+      "grad_norm": 4.6285342572963195,
+      "learning_rate": 6.575768304879292e-08,
+      "loss": 1.6929,
+      "step": 12635
+    },
+    {
+      "epoch": 0.8517233246858259,
+      "grad_norm": 4.637050698862129,
+      "learning_rate": 6.546647811907091e-08,
+      "loss": 1.5862,
+      "step": 12640
+    },
+    {
+      "epoch": 0.8520602405579327,
+      "grad_norm": 4.592917589678165,
+      "learning_rate": 6.517587423874988e-08,
+      "loss": 1.6897,
+      "step": 12645
+    },
+    {
+      "epoch": 0.8523971564300394,
+      "grad_norm": 4.345754092042132,
+      "learning_rate": 6.48858718097951e-08,
+      "loss": 1.5901,
+      "step": 12650
+    },
+    {
+      "epoch": 0.8527340723021462,
+      "grad_norm": 4.297560031338897,
+      "learning_rate": 6.459647123333956e-08,
+      "loss": 1.5688,
+      "step": 12655
+    },
+    {
+      "epoch": 0.8530709881742529,
+      "grad_norm": 4.712073899028917,
+      "learning_rate": 6.430767290968387e-08,
+      "loss": 1.6449,
+      "step": 12660
+    },
+    {
+      "epoch": 0.8534079040463596,
+      "grad_norm": 4.744460867221172,
+      "learning_rate": 6.401947723829576e-08,
+      "loss": 1.6518,
+      "step": 12665
+    },
+    {
+      "epoch": 0.8537448199184664,
+      "grad_norm": 4.4916813376687355,
+      "learning_rate": 6.373188461780904e-08,
+      "loss": 1.6605,
+      "step": 12670
+    },
+    {
+      "epoch": 0.8540817357905731,
+      "grad_norm": 4.551996624487793,
+      "learning_rate": 6.344489544602371e-08,
+      "loss": 1.7003,
+      "step": 12675
+    },
+    {
+      "epoch": 0.8544186516626798,
+      "grad_norm": 4.699348939364616,
+      "learning_rate": 6.315851011990498e-08,
+      "loss": 1.7217,
+      "step": 12680
+    },
+    {
+      "epoch": 0.8547555675347865,
+      "grad_norm": 4.314103210677744,
+      "learning_rate": 6.28727290355826e-08,
+      "loss": 1.6255,
+      "step": 12685
+    },
+    {
+      "epoch": 0.8550924834068933,
+      "grad_norm": 4.616664419421673,
+      "learning_rate": 6.258755258835075e-08,
+      "loss": 1.674,
+      "step": 12690
+    },
+    {
+      "epoch": 0.8554293992790001,
+      "grad_norm": 4.769742677903213,
+      "learning_rate": 6.230298117266736e-08,
+      "loss": 1.6362,
+      "step": 12695
+    },
+    {
+      "epoch": 0.8557663151511068,
+      "grad_norm": 4.750891827945278,
+      "learning_rate": 6.201901518215313e-08,
+      "loss": 1.6539,
+      "step": 12700
+    },
+    {
+      "epoch": 0.8561032310232135,
+      "grad_norm": 5.300031969588006,
+      "learning_rate": 6.173565500959165e-08,
+      "loss": 1.7062,
+      "step": 12705
+    },
+    {
+      "epoch": 0.8564401468953202,
+      "grad_norm": 4.9877195879588925,
+      "learning_rate": 6.14529010469284e-08,
+      "loss": 1.7033,
+      "step": 12710
+    },
+    {
+      "epoch": 0.8567770627674269,
+      "grad_norm": 4.465620946765142,
+      "learning_rate": 6.117075368527053e-08,
+      "loss": 1.5669,
+      "step": 12715
+    },
+    {
+      "epoch": 0.8571139786395338,
+      "grad_norm": 4.567565940501714,
+      "learning_rate": 6.088921331488566e-08,
+      "loss": 1.6474,
+      "step": 12720
+    },
+    {
+      "epoch": 0.8574508945116405,
+      "grad_norm": 4.5006741327829465,
+      "learning_rate": 6.060828032520249e-08,
+      "loss": 1.662,
+      "step": 12725
+    },
+    {
+      "epoch": 0.8577878103837472,
+      "grad_norm": 4.7060071164822155,
+      "learning_rate": 6.032795510480904e-08,
+      "loss": 1.6264,
+      "step": 12730
+    },
+    {
+      "epoch": 0.8581247262558539,
+      "grad_norm": 4.648153265111918,
+      "learning_rate": 6.004823804145276e-08,
+      "loss": 1.7249,
+      "step": 12735
+    },
+    {
+      "epoch": 0.8584616421279606,
+      "grad_norm": 4.47213474728508,
+      "learning_rate": 5.976912952204016e-08,
+      "loss": 1.6887,
+      "step": 12740
+    },
+    {
+      "epoch": 0.8587985580000674,
+      "grad_norm": 4.524244747562429,
+      "learning_rate": 5.9490629932635815e-08,
+      "loss": 1.7353,
+      "step": 12745
+    },
+    {
+      "epoch": 0.8591354738721741,
+      "grad_norm": 5.297426837911389,
+      "learning_rate": 5.921273965846191e-08,
+      "loss": 1.648,
+      "step": 12750
+    },
+    {
+      "epoch": 0.8594723897442809,
+      "grad_norm": 4.965235141374218,
+      "learning_rate": 5.893545908389807e-08,
+      "loss": 1.5881,
+      "step": 12755
+    },
+    {
+      "epoch": 0.8598093056163876,
+      "grad_norm": 4.652217619220733,
+      "learning_rate": 5.865878859248058e-08,
+      "loss": 1.5889,
+      "step": 12760
+    },
+    {
+      "epoch": 0.8601462214884943,
+      "grad_norm": 4.837036525431273,
+      "learning_rate": 5.838272856690146e-08,
+      "loss": 1.596,
+      "step": 12765
+    },
+    {
+      "epoch": 0.860483137360601,
+      "grad_norm": 4.317476977979007,
+      "learning_rate": 5.810727938900878e-08,
+      "loss": 1.6771,
+      "step": 12770
+    },
+    {
+      "epoch": 0.8608200532327078,
+      "grad_norm": 4.485083808259918,
+      "learning_rate": 5.7832441439805536e-08,
+      "loss": 1.6996,
+      "step": 12775
+    },
+    {
+      "epoch": 0.8611569691048145,
+      "grad_norm": 4.602719696368838,
+      "learning_rate": 5.755821509944925e-08,
+      "loss": 1.6995,
+      "step": 12780
+    },
+    {
+      "epoch": 0.8614938849769213,
+      "grad_norm": 4.82873736404331,
+      "learning_rate": 5.728460074725133e-08,
+      "loss": 1.6445,
+      "step": 12785
+    },
+    {
+      "epoch": 0.861830800849028,
+      "grad_norm": 4.613031757102517,
+      "learning_rate": 5.701159876167688e-08,
+      "loss": 1.6395,
+      "step": 12790
+    },
+    {
+      "epoch": 0.8621677167211347,
+      "grad_norm": 4.739452889007518,
+      "learning_rate": 5.673920952034406e-08,
+      "loss": 1.6522,
+      "step": 12795
+    },
+    {
+      "epoch": 0.8625046325932415,
+      "grad_norm": 4.825662079465411,
+      "learning_rate": 5.646743340002302e-08,
+      "loss": 1.6343,
+      "step": 12800
+    },
+    {
+      "epoch": 0.8628415484653482,
+      "grad_norm": 5.3045173960429635,
+      "learning_rate": 5.619627077663636e-08,
+      "loss": 1.7305,
+      "step": 12805
+    },
+    {
+      "epoch": 0.8631784643374549,
+      "grad_norm": 4.556268810510644,
+      "learning_rate": 5.5925722025257746e-08,
+      "loss": 1.735,
+      "step": 12810
+    },
+    {
+      "epoch": 0.8635153802095616,
+      "grad_norm": 4.584380506792433,
+      "learning_rate": 5.5655787520111966e-08,
+      "loss": 1.7302,
+      "step": 12815
+    },
+    {
+      "epoch": 0.8638522960816684,
+      "grad_norm": 4.495296345448851,
+      "learning_rate": 5.538646763457389e-08,
+      "loss": 1.6434,
+      "step": 12820
+    },
+    {
+      "epoch": 0.8641892119537752,
+      "grad_norm": 4.587139082186452,
+      "learning_rate": 5.511776274116864e-08,
+      "loss": 1.6176,
+      "step": 12825
+    },
+    {
+      "epoch": 0.8645261278258819,
+      "grad_norm": 4.861128343283088,
+      "learning_rate": 5.484967321157019e-08,
+      "loss": 1.6692,
+      "step": 12830
+    },
+    {
+      "epoch": 0.8648630436979886,
+      "grad_norm": 4.753311652307588,
+      "learning_rate": 5.4582199416601746e-08,
+      "loss": 1.6115,
+      "step": 12835
+    },
+    {
+      "epoch": 0.8651999595700953,
+      "grad_norm": 4.775924852820231,
+      "learning_rate": 5.43153417262347e-08,
+      "loss": 1.7149,
+      "step": 12840
+    },
+    {
+      "epoch": 0.865536875442202,
+      "grad_norm": 5.105625607161886,
+      "learning_rate": 5.404910050958833e-08,
+      "loss": 1.6891,
+      "step": 12845
+    },
+    {
+      "epoch": 0.8658737913143089,
+      "grad_norm": 4.469873946974137,
+      "learning_rate": 5.378347613492884e-08,
+      "loss": 1.6519,
+      "step": 12850
+    },
+    {
+      "epoch": 0.8662107071864156,
+      "grad_norm": 4.364768077012948,
+      "learning_rate": 5.351846896966966e-08,
+      "loss": 1.6981,
+      "step": 12855
+    },
+    {
+      "epoch": 0.8665476230585223,
+      "grad_norm": 4.9119501639318415,
+      "learning_rate": 5.32540793803703e-08,
+      "loss": 1.6897,
+      "step": 12860
+    },
+    {
+      "epoch": 0.866884538930629,
+      "grad_norm": 4.43964878962444,
+      "learning_rate": 5.299030773273594e-08,
+      "loss": 1.6619,
+      "step": 12865
+    },
+    {
+      "epoch": 0.8672214548027357,
+      "grad_norm": 4.672406178991766,
+      "learning_rate": 5.272715439161718e-08,
+      "loss": 1.6509,
+      "step": 12870
+    },
+    {
+      "epoch": 0.8675583706748425,
+      "grad_norm": 4.772545341332354,
+      "learning_rate": 5.246461972100941e-08,
+      "loss": 1.6213,
+      "step": 12875
+    },
+    {
+      "epoch": 0.8678952865469493,
+      "grad_norm": 4.286308937980272,
+      "learning_rate": 5.220270408405197e-08,
+      "loss": 1.6349,
+      "step": 12880
+    },
+    {
+      "epoch": 0.868232202419056,
+      "grad_norm": 4.634791844153955,
+      "learning_rate": 5.194140784302836e-08,
+      "loss": 1.6495,
+      "step": 12885
+    },
+    {
+      "epoch": 0.8685691182911627,
+      "grad_norm": 3.8557299998167927,
+      "learning_rate": 5.168073135936496e-08,
+      "loss": 1.6539,
+      "step": 12890
+    },
+    {
+      "epoch": 0.8689060341632694,
+      "grad_norm": 4.746078778573034,
+      "learning_rate": 5.1420674993631285e-08,
+      "loss": 1.6101,
+      "step": 12895
+    },
+    {
+      "epoch": 0.8692429500353762,
+      "grad_norm": 4.596974751954337,
+      "learning_rate": 5.116123910553854e-08,
+      "loss": 1.6959,
+      "step": 12900
+    },
+    {
+      "epoch": 0.8695798659074829,
+      "grad_norm": 4.183227031111133,
+      "learning_rate": 5.0902424053940406e-08,
+      "loss": 1.7306,
+      "step": 12905
+    },
+    {
+      "epoch": 0.8699167817795896,
+      "grad_norm": 4.706195131417817,
+      "learning_rate": 5.064423019683106e-08,
+      "loss": 1.6512,
+      "step": 12910
+    },
+    {
+      "epoch": 0.8702536976516964,
+      "grad_norm": 4.883634643878311,
+      "learning_rate": 5.0386657891346e-08,
+      "loss": 1.6405,
+      "step": 12915
+    },
+    {
+      "epoch": 0.8705906135238031,
+      "grad_norm": 4.121854755010217,
+      "learning_rate": 5.012970749376083e-08,
+      "loss": 1.6358,
+      "step": 12920
+    },
+    {
+      "epoch": 0.8709275293959099,
+      "grad_norm": 4.713311198178696,
+      "learning_rate": 4.987337935949087e-08,
+      "loss": 1.6824,
+      "step": 12925
+    },
+    {
+      "epoch": 0.8712644452680166,
+      "grad_norm": 4.535731428384727,
+      "learning_rate": 4.961767384309068e-08,
+      "loss": 1.7024,
+      "step": 12930
+    },
+    {
+      "epoch": 0.8716013611401233,
+      "grad_norm": 4.190502023346841,
+      "learning_rate": 4.936259129825376e-08,
+      "loss": 1.662,
+      "step": 12935
+    },
+    {
+      "epoch": 0.87193827701223,
+      "grad_norm": 4.541026079557271,
+      "learning_rate": 4.9108132077811836e-08,
+      "loss": 1.6413,
+      "step": 12940
+    },
+    {
+      "epoch": 0.8722751928843367,
+      "grad_norm": 4.692232599207379,
+      "learning_rate": 4.885429653373435e-08,
+      "loss": 1.6707,
+      "step": 12945
+    },
+    {
+      "epoch": 0.8726121087564435,
+      "grad_norm": 5.017348687830675,
+      "learning_rate": 4.860108501712823e-08,
+      "loss": 1.6969,
+      "step": 12950
+    },
+    {
+      "epoch": 0.8729490246285503,
+      "grad_norm": 4.679428429265382,
+      "learning_rate": 4.834849787823725e-08,
+      "loss": 1.6445,
+      "step": 12955
+    },
+    {
+      "epoch": 0.873285940500657,
+      "grad_norm": 5.00770849393764,
+      "learning_rate": 4.809653546644132e-08,
+      "loss": 1.7464,
+      "step": 12960
+    },
+    {
+      "epoch": 0.8736228563727637,
+      "grad_norm": 4.780319317876716,
+      "learning_rate": 4.7845198130256395e-08,
+      "loss": 1.7075,
+      "step": 12965
+    },
+    {
+      "epoch": 0.8739597722448704,
+      "grad_norm": 4.568955739105629,
+      "learning_rate": 4.759448621733403e-08,
+      "loss": 1.6402,
+      "step": 12970
+    },
+    {
+      "epoch": 0.8742966881169771,
+      "grad_norm": 4.670815008886996,
+      "learning_rate": 4.7344400074460276e-08,
+      "loss": 1.6548,
+      "step": 12975
+    },
+    {
+      "epoch": 0.874633603989084,
+      "grad_norm": 4.6192757150973565,
+      "learning_rate": 4.709494004755571e-08,
+      "loss": 1.6799,
+      "step": 12980
+    },
+    {
+      "epoch": 0.8749705198611907,
+      "grad_norm": 4.607392535169383,
+      "learning_rate": 4.684610648167503e-08,
+      "loss": 1.6268,
+      "step": 12985
+    },
+    {
+      "epoch": 0.8753074357332974,
+      "grad_norm": 4.975556056688224,
+      "learning_rate": 4.659789972100647e-08,
+      "loss": 1.6801,
+      "step": 12990
+    },
+    {
+      "epoch": 0.8756443516054041,
+      "grad_norm": 4.383102938364268,
+      "learning_rate": 4.635032010887097e-08,
+      "loss": 1.7203,
+      "step": 12995
+    },
+    {
+      "epoch": 0.8759812674775108,
+      "grad_norm": 4.367463559610928,
+      "learning_rate": 4.610336798772213e-08,
+      "loss": 1.6703,
+      "step": 13000
+    },
+    {
+      "epoch": 0.8763181833496176,
+      "grad_norm": 4.442038336994325,
+      "learning_rate": 4.5857043699145834e-08,
+      "loss": 1.6961,
+      "step": 13005
+    },
+    {
+      "epoch": 0.8766550992217244,
+      "grad_norm": 4.793599970051047,
+      "learning_rate": 4.5611347583859095e-08,
+      "loss": 1.6731,
+      "step": 13010
+    },
+    {
+      "epoch": 0.8769920150938311,
+      "grad_norm": 4.9522389045914625,
+      "learning_rate": 4.536627998171033e-08,
+      "loss": 1.6996,
+      "step": 13015
+    },
+    {
+      "epoch": 0.8773289309659378,
+      "grad_norm": 4.980252194482352,
+      "learning_rate": 4.512184123167867e-08,
+      "loss": 1.6263,
+      "step": 13020
+    },
+    {
+      "epoch": 0.8776658468380445,
+      "grad_norm": 4.9646395476564384,
+      "learning_rate": 4.487803167187304e-08,
+      "loss": 1.6801,
+      "step": 13025
+    },
+    {
+      "epoch": 0.8780027627101513,
+      "grad_norm": 4.778315521986718,
+      "learning_rate": 4.463485163953246e-08,
+      "loss": 1.7304,
+      "step": 13030
+    },
+    {
+      "epoch": 0.878339678582258,
+      "grad_norm": 4.36506282947457,
+      "learning_rate": 4.4392301471025074e-08,
+      "loss": 1.7238,
+      "step": 13035
+    },
+    {
+      "epoch": 0.8786765944543647,
+      "grad_norm": 4.4129541893069835,
+      "learning_rate": 4.415038150184758e-08,
+      "loss": 1.6556,
+      "step": 13040
+    },
+    {
+      "epoch": 0.8790135103264715,
+      "grad_norm": 4.600992013387407,
+      "learning_rate": 4.3909092066625245e-08,
+      "loss": 1.6354,
+      "step": 13045
+    },
+    {
+      "epoch": 0.8793504261985782,
+      "grad_norm": 4.501353404766527,
+      "learning_rate": 4.366843349911109e-08,
+      "loss": 1.6193,
+      "step": 13050
+    },
+    {
+      "epoch": 0.879687342070685,
+      "grad_norm": 4.793533490720773,
+      "learning_rate": 4.342840613218546e-08,
+      "loss": 1.6213,
+      "step": 13055
+    },
+    {
+      "epoch": 0.8800242579427917,
+      "grad_norm": 4.873768264726616,
+      "learning_rate": 4.318901029785571e-08,
+      "loss": 1.6917,
+      "step": 13060
+    },
+    {
+      "epoch": 0.8803611738148984,
+      "grad_norm": 4.514341122154152,
+      "learning_rate": 4.2950246327255523e-08,
+      "loss": 1.6181,
+      "step": 13065
+    },
+    {
+      "epoch": 0.8806980896870051,
+      "grad_norm": 4.267152040052711,
+      "learning_rate": 4.271211455064483e-08,
+      "loss": 1.7006,
+      "step": 13070
+    },
+    {
+      "epoch": 0.8810350055591119,
+      "grad_norm": 4.785818717093623,
+      "learning_rate": 4.2474615297408754e-08,
+      "loss": 1.647,
+      "step": 13075
+    },
+    {
+      "epoch": 0.8813719214312187,
+      "grad_norm": 4.30173252242199,
+      "learning_rate": 4.223774889605775e-08,
+      "loss": 1.6541,
+      "step": 13080
+    },
+    {
+      "epoch": 0.8817088373033254,
+      "grad_norm": 4.565940042679433,
+      "learning_rate": 4.200151567422699e-08,
+      "loss": 1.6905,
+      "step": 13085
+    },
+    {
+      "epoch": 0.8820457531754321,
+      "grad_norm": 4.69539783865438,
+      "learning_rate": 4.176591595867557e-08,
+      "loss": 1.6944,
+      "step": 13090
+    },
+    {
+      "epoch": 0.8823826690475388,
+      "grad_norm": 4.380967972336634,
+      "learning_rate": 4.153095007528645e-08,
+      "loss": 1.6728,
+      "step": 13095
+    },
+    {
+      "epoch": 0.8827195849196455,
+      "grad_norm": 4.516035300863843,
+      "learning_rate": 4.1296618349066e-08,
+      "loss": 1.603,
+      "step": 13100
+    },
+    {
+      "epoch": 0.8830565007917524,
+      "grad_norm": 4.296821049635356,
+      "learning_rate": 4.106292110414311e-08,
+      "loss": 1.7034,
+      "step": 13105
+    },
+    {
+      "epoch": 0.8833934166638591,
+      "grad_norm": 4.437609089662178,
+      "learning_rate": 4.082985866376926e-08,
+      "loss": 1.7042,
+      "step": 13110
+    },
+    {
+      "epoch": 0.8837303325359658,
+      "grad_norm": 5.020502142421371,
+      "learning_rate": 4.05974313503179e-08,
+      "loss": 1.7177,
+      "step": 13115
+    },
+    {
+      "epoch": 0.8840672484080725,
+      "grad_norm": 4.797384079762569,
+      "learning_rate": 4.036563948528393e-08,
+      "loss": 1.6754,
+      "step": 13120
+    },
+    {
+      "epoch": 0.8844041642801792,
+      "grad_norm": 4.420504657788444,
+      "learning_rate": 4.01344833892831e-08,
+      "loss": 1.7012,
+      "step": 13125
+    },
+    {
+      "epoch": 0.8847410801522859,
+      "grad_norm": 5.010665239478789,
+      "learning_rate": 3.990396338205204e-08,
+      "loss": 1.6727,
+      "step": 13130
+    },
+    {
+      "epoch": 0.8850779960243927,
+      "grad_norm": 4.875792248941043,
+      "learning_rate": 3.967407978244747e-08,
+      "loss": 1.5873,
+      "step": 13135
+    },
+    {
+      "epoch": 0.8854149118964995,
+      "grad_norm": 4.920319954446455,
+      "learning_rate": 3.944483290844575e-08,
+      "loss": 1.6536,
+      "step": 13140
+    },
+    {
+      "epoch": 0.8857518277686062,
+      "grad_norm": 4.702279123087858,
+      "learning_rate": 3.9216223077142394e-08,
+      "loss": 1.7247,
+      "step": 13145
+    },
+    {
+      "epoch": 0.8860887436407129,
+      "grad_norm": 4.690033811483646,
+      "learning_rate": 3.8988250604752135e-08,
+      "loss": 1.6406,
+      "step": 13150
+    },
+    {
+      "epoch": 0.8864256595128196,
+      "grad_norm": 5.082844262732006,
+      "learning_rate": 3.876091580660762e-08,
+      "loss": 1.7613,
+      "step": 13155
+    },
+    {
+      "epoch": 0.8867625753849264,
+      "grad_norm": 4.662418270415757,
+      "learning_rate": 3.853421899715992e-08,
+      "loss": 1.638,
+      "step": 13160
+    },
+    {
+      "epoch": 0.8870994912570331,
+      "grad_norm": 5.224077723143157,
+      "learning_rate": 3.8308160489977424e-08,
+      "loss": 1.709,
+      "step": 13165
+    },
+    {
+      "epoch": 0.8874364071291398,
+      "grad_norm": 4.947169326251201,
+      "learning_rate": 3.808274059774552e-08,
+      "loss": 1.7174,
+      "step": 13170
+    },
+    {
+      "epoch": 0.8877733230012466,
+      "grad_norm": 4.600213672685725,
+      "learning_rate": 3.785795963226646e-08,
+      "loss": 1.6272,
+      "step": 13175
+    },
+    {
+      "epoch": 0.8881102388733533,
+      "grad_norm": 4.644106545994671,
+      "learning_rate": 3.7633817904458574e-08,
+      "loss": 1.709,
+      "step": 13180
+    },
+    {
+      "epoch": 0.8884471547454601,
+      "grad_norm": 4.881787365902224,
+      "learning_rate": 3.741031572435615e-08,
+      "loss": 1.6412,
+      "step": 13185
+    },
+    {
+      "epoch": 0.8887840706175668,
+      "grad_norm": 4.601119029827488,
+      "learning_rate": 3.718745340110868e-08,
+      "loss": 1.6599,
+      "step": 13190
+    },
+    {
+      "epoch": 0.8891209864896735,
+      "grad_norm": 4.876499663076449,
+      "learning_rate": 3.6965231242980624e-08,
+      "loss": 1.6504,
+      "step": 13195
+    },
+    {
+      "epoch": 0.8894579023617802,
+      "grad_norm": 4.614845675235782,
+      "learning_rate": 3.6743649557351265e-08,
+      "loss": 1.6961,
+      "step": 13200
+    },
+    {
+      "epoch": 0.889794818233887,
+      "grad_norm": 4.2887236933033055,
+      "learning_rate": 3.652270865071344e-08,
+      "loss": 1.6644,
+      "step": 13205
+    },
+    {
+      "epoch": 0.8901317341059938,
+      "grad_norm": 4.013386340686639,
+      "learning_rate": 3.630240882867408e-08,
+      "loss": 1.609,
+      "step": 13210
+    },
+    {
+      "epoch": 0.8904686499781005,
+      "grad_norm": 4.589800677924639,
+      "learning_rate": 3.608275039595332e-08,
+      "loss": 1.6885,
+      "step": 13215
+    },
+    {
+      "epoch": 0.8908055658502072,
+      "grad_norm": 4.996418734298018,
+      "learning_rate": 3.5863733656383844e-08,
+      "loss": 1.6432,
+      "step": 13220
+    },
+    {
+      "epoch": 0.8911424817223139,
+      "grad_norm": 4.561616629992068,
+      "learning_rate": 3.564535891291115e-08,
+      "loss": 1.6334,
+      "step": 13225
+    },
+    {
+      "epoch": 0.8914793975944206,
+      "grad_norm": 4.5190475576262195,
+      "learning_rate": 3.542762646759234e-08,
+      "loss": 1.6786,
+      "step": 13230
+    },
+    {
+      "epoch": 0.8918163134665275,
+      "grad_norm": 4.885770968192226,
+      "learning_rate": 3.521053662159629e-08,
+      "loss": 1.6563,
+      "step": 13235
+    },
+    {
+      "epoch": 0.8921532293386342,
+      "grad_norm": 4.311300465911783,
+      "learning_rate": 3.499408967520295e-08,
+      "loss": 1.6554,
+      "step": 13240
+    },
+    {
+      "epoch": 0.8924901452107409,
+      "grad_norm": 4.661435819756512,
+      "learning_rate": 3.477828592780319e-08,
+      "loss": 1.6824,
+      "step": 13245
+    },
+    {
+      "epoch": 0.8928270610828476,
+      "grad_norm": 4.507106001053341,
+      "learning_rate": 3.456312567789793e-08,
+      "loss": 1.7095,
+      "step": 13250
+    },
+    {
+      "epoch": 0.8931639769549543,
+      "grad_norm": 4.3042004781365675,
+      "learning_rate": 3.4348609223098125e-08,
+      "loss": 1.6764,
+      "step": 13255
+    },
+    {
+      "epoch": 0.8935008928270611,
+      "grad_norm": 4.459285002947242,
+      "learning_rate": 3.41347368601243e-08,
+      "loss": 1.608,
+      "step": 13260
+    },
+    {
+      "epoch": 0.8938378086991678,
+      "grad_norm": 5.012943421245698,
+      "learning_rate": 3.39215088848061e-08,
+      "loss": 1.7396,
+      "step": 13265
+    },
+    {
+      "epoch": 0.8941747245712746,
+      "grad_norm": 5.035475395776995,
+      "learning_rate": 3.370892559208155e-08,
+      "loss": 1.6454,
+      "step": 13270
+    },
+    {
+      "epoch": 0.8945116404433813,
+      "grad_norm": 4.46377384535415,
+      "learning_rate": 3.34969872759972e-08,
+      "loss": 1.6899,
+      "step": 13275
+    },
+    {
+      "epoch": 0.894848556315488,
+      "grad_norm": 4.47489904509951,
+      "learning_rate": 3.328569422970762e-08,
+      "loss": 1.7486,
+      "step": 13280
+    },
+    {
+      "epoch": 0.8951854721875948,
+      "grad_norm": 4.39226374900135,
+      "learning_rate": 3.307504674547429e-08,
+      "loss": 1.6666,
+      "step": 13285
+    },
+    {
+      "epoch": 0.8955223880597015,
+      "grad_norm": 6.3409003089939935,
+      "learning_rate": 3.286504511466631e-08,
+      "loss": 1.7211,
+      "step": 13290
+    },
+    {
+      "epoch": 0.8958593039318082,
+      "grad_norm": 4.92187136039575,
+      "learning_rate": 3.265568962775927e-08,
+      "loss": 1.672,
+      "step": 13295
+    },
+    {
+      "epoch": 0.896196219803915,
+      "grad_norm": 4.862516911103943,
+      "learning_rate": 3.2446980574334706e-08,
+      "loss": 1.6782,
+      "step": 13300
+    },
+    {
+      "epoch": 0.8965331356760217,
+      "grad_norm": 4.954131476977874,
+      "learning_rate": 3.2238918243080505e-08,
+      "loss": 1.6374,
+      "step": 13305
+    },
+    {
+      "epoch": 0.8968700515481284,
+      "grad_norm": 5.185311614713857,
+      "learning_rate": 3.203150292178952e-08,
+      "loss": 1.6918,
+      "step": 13310
+    },
+    {
+      "epoch": 0.8972069674202352,
+      "grad_norm": 4.859702788623765,
+      "learning_rate": 3.182473489736004e-08,
+      "loss": 1.6278,
+      "step": 13315
+    },
+    {
+      "epoch": 0.8975438832923419,
+      "grad_norm": 5.444427223124619,
+      "learning_rate": 3.161861445579478e-08,
+      "loss": 1.6935,
+      "step": 13320
+    },
+    {
+      "epoch": 0.8978807991644486,
+      "grad_norm": 4.6618984522024345,
+      "learning_rate": 3.1413141882200736e-08,
+      "loss": 1.6618,
+      "step": 13325
+    },
+    {
+      "epoch": 0.8982177150365553,
+      "grad_norm": 4.597572632532803,
+      "learning_rate": 3.120831746078895e-08,
+      "loss": 1.6797,
+      "step": 13330
+    },
+    {
+      "epoch": 0.898554630908662,
+      "grad_norm": 4.66671806352106,
+      "learning_rate": 3.100414147487368e-08,
+      "loss": 1.7287,
+      "step": 13335
+    },
+    {
+      "epoch": 0.8988915467807689,
+      "grad_norm": 4.432222168219578,
+      "learning_rate": 3.0800614206872413e-08,
+      "loss": 1.6574,
+      "step": 13340
+    },
+    {
+      "epoch": 0.8992284626528756,
+      "grad_norm": 5.389463648441865,
+      "learning_rate": 3.059773593830539e-08,
+      "loss": 1.6732,
+      "step": 13345
+    },
+    {
+      "epoch": 0.8995653785249823,
+      "grad_norm": 4.51707048609424,
+      "learning_rate": 3.039550694979492e-08,
+      "loss": 1.6256,
+      "step": 13350
+    },
+    {
+      "epoch": 0.899902294397089,
+      "grad_norm": 4.596592919058153,
+      "learning_rate": 3.019392752106548e-08,
+      "loss": 1.6426,
+      "step": 13355
+    },
+    {
+      "epoch": 0.9002392102691957,
+      "grad_norm": 4.534597115346769,
+      "learning_rate": 2.9992997930942954e-08,
+      "loss": 1.6079,
+      "step": 13360
+    },
+    {
+      "epoch": 0.9005761261413026,
+      "grad_norm": 4.501842736749262,
+      "learning_rate": 2.979271845735426e-08,
+      "loss": 1.5888,
+      "step": 13365
+    },
+    {
+      "epoch": 0.9009130420134093,
+      "grad_norm": 4.543584289239795,
+      "learning_rate": 2.9593089377327242e-08,
+      "loss": 1.6151,
+      "step": 13370
+    },
+    {
+      "epoch": 0.901249957885516,
+      "grad_norm": 4.656857934626395,
+      "learning_rate": 2.9394110966990184e-08,
+      "loss": 1.6873,
+      "step": 13375
+    },
+    {
+      "epoch": 0.9015868737576227,
+      "grad_norm": 4.501681745676623,
+      "learning_rate": 2.9195783501570982e-08,
+      "loss": 1.6992,
+      "step": 13380
+    },
+    {
+      "epoch": 0.9019237896297294,
+      "grad_norm": 4.907087522799292,
+      "learning_rate": 2.8998107255397643e-08,
+      "loss": 1.6739,
+      "step": 13385
+    },
+    {
+      "epoch": 0.9022607055018362,
+      "grad_norm": 4.490445324603956,
+      "learning_rate": 2.880108250189689e-08,
+      "loss": 1.6372,
+      "step": 13390
+    },
+    {
+      "epoch": 0.902597621373943,
+      "grad_norm": 4.762399631606492,
+      "learning_rate": 2.860470951359478e-08,
+      "loss": 1.649,
+      "step": 13395
+    },
+    {
+      "epoch": 0.9029345372460497,
+      "grad_norm": 4.965958597373103,
+      "learning_rate": 2.8408988562115488e-08,
+      "loss": 1.678,
+      "step": 13400
+    },
+    {
+      "epoch": 0.9032714531181564,
+      "grad_norm": 4.873670679901971,
+      "learning_rate": 2.8213919918181393e-08,
+      "loss": 1.7255,
+      "step": 13405
+    },
+    {
+      "epoch": 0.9036083689902631,
+      "grad_norm": 4.637630650167888,
+      "learning_rate": 2.8019503851612837e-08,
+      "loss": 1.6833,
+      "step": 13410
+    },
+    {
+      "epoch": 0.9039452848623699,
+      "grad_norm": 4.614907766408516,
+      "learning_rate": 2.782574063132703e-08,
+      "loss": 1.7035,
+      "step": 13415
+    },
+    {
+      "epoch": 0.9042822007344766,
+      "grad_norm": 4.698127813148663,
+      "learning_rate": 2.7632630525338597e-08,
+      "loss": 1.6726,
+      "step": 13420
+    },
+    {
+      "epoch": 0.9046191166065833,
+      "grad_norm": 4.908958225004308,
+      "learning_rate": 2.7440173800758583e-08,
+      "loss": 1.693,
+      "step": 13425
+    },
+    {
+      "epoch": 0.90495603247869,
+      "grad_norm": 4.553215982139727,
+      "learning_rate": 2.7248370723794268e-08,
+      "loss": 1.6548,
+      "step": 13430
+    },
+    {
+      "epoch": 0.9052929483507968,
+      "grad_norm": 4.873564430369931,
+      "learning_rate": 2.7057221559748822e-08,
+      "loss": 1.6762,
+      "step": 13435
+    },
+    {
+      "epoch": 0.9056298642229036,
+      "grad_norm": 4.485392880009042,
+      "learning_rate": 2.6866726573021025e-08,
+      "loss": 1.6518,
+      "step": 13440
+    },
+    {
+      "epoch": 0.9059667800950103,
+      "grad_norm": 4.932314532115502,
+      "learning_rate": 2.667688602710455e-08,
+      "loss": 1.6006,
+      "step": 13445
+    },
+    {
+      "epoch": 0.906303695967117,
+      "grad_norm": 4.503122097316531,
+      "learning_rate": 2.648770018458807e-08,
+      "loss": 1.6812,
+      "step": 13450
+    },
+    {
+      "epoch": 0.9066406118392237,
+      "grad_norm": 4.574449543803867,
+      "learning_rate": 2.6299169307154535e-08,
+      "loss": 1.6163,
+      "step": 13455
+    },
+    {
+      "epoch": 0.9069775277113304,
+      "grad_norm": 4.834936908101332,
+      "learning_rate": 2.611129365558118e-08,
+      "loss": 1.6345,
+      "step": 13460
+    },
+    {
+      "epoch": 0.9073144435834373,
+      "grad_norm": 4.628745448605451,
+      "learning_rate": 2.592407348973852e-08,
+      "loss": 1.6528,
+      "step": 13465
+    },
+    {
+      "epoch": 0.907651359455544,
+      "grad_norm": 4.335332294524596,
+      "learning_rate": 2.573750906859079e-08,
+      "loss": 1.6507,
+      "step": 13470
+    },
+    {
+      "epoch": 0.9079882753276507,
+      "grad_norm": 4.6604992266607095,
+      "learning_rate": 2.5551600650194906e-08,
+      "loss": 1.6526,
+      "step": 13475
+    },
+    {
+      "epoch": 0.9083251911997574,
+      "grad_norm": 4.4333865162999055,
+      "learning_rate": 2.536634849170055e-08,
+      "loss": 1.6162,
+      "step": 13480
+    },
+    {
+      "epoch": 0.9086621070718641,
+      "grad_norm": 4.85890159742819,
+      "learning_rate": 2.5181752849349593e-08,
+      "loss": 1.7159,
+      "step": 13485
+    },
+    {
+      "epoch": 0.9089990229439708,
+      "grad_norm": 4.729622761774583,
+      "learning_rate": 2.4997813978476e-08,
+      "loss": 1.629,
+      "step": 13490
+    },
+    {
+      "epoch": 0.9093359388160777,
+      "grad_norm": 4.445209264708288,
+      "learning_rate": 2.481453213350493e-08,
+      "loss": 1.6073,
+      "step": 13495
+    },
+    {
+      "epoch": 0.9096728546881844,
+      "grad_norm": 4.331937548330955,
+      "learning_rate": 2.463190756795308e-08,
+      "loss": 1.7122,
+      "step": 13500
+    },
+    {
+      "epoch": 0.9100097705602911,
+      "grad_norm": 4.210089541483122,
+      "learning_rate": 2.4449940534427836e-08,
+      "loss": 1.6794,
+      "step": 13505
+    },
+    {
+      "epoch": 0.9103466864323978,
+      "grad_norm": 4.522200369117284,
+      "learning_rate": 2.4268631284627027e-08,
+      "loss": 1.6555,
+      "step": 13510
+    },
+    {
+      "epoch": 0.9106836023045045,
+      "grad_norm": 4.620662367302598,
+      "learning_rate": 2.408798006933882e-08,
+      "loss": 1.6215,
+      "step": 13515
+    },
+    {
+      "epoch": 0.9110205181766113,
+      "grad_norm": 4.935550615365647,
+      "learning_rate": 2.3907987138440945e-08,
+      "loss": 1.745,
+      "step": 13520
+    },
+    {
+      "epoch": 0.911357434048718,
+      "grad_norm": 4.809365117645508,
+      "learning_rate": 2.3728652740900856e-08,
+      "loss": 1.7111,
+      "step": 13525
+    },
+    {
+      "epoch": 0.9116943499208248,
+      "grad_norm": 4.958322472785031,
+      "learning_rate": 2.3549977124774857e-08,
+      "loss": 1.6901,
+      "step": 13530
+    },
+    {
+      "epoch": 0.9120312657929315,
+      "grad_norm": 4.935261215153186,
+      "learning_rate": 2.337196053720819e-08,
+      "loss": 1.73,
+      "step": 13535
+    },
+    {
+      "epoch": 0.9123681816650382,
+      "grad_norm": 4.511277252542813,
+      "learning_rate": 2.319460322443456e-08,
+      "loss": 1.7142,
+      "step": 13540
+    },
+    {
+      "epoch": 0.912705097537145,
+      "grad_norm": 4.723475456942302,
+      "learning_rate": 2.301790543177551e-08,
+      "loss": 1.7186,
+      "step": 13545
+    },
+    {
+      "epoch": 0.9130420134092517,
+      "grad_norm": 4.584142787010846,
+      "learning_rate": 2.284186740364069e-08,
+      "loss": 1.6963,
+      "step": 13550
+    },
+    {
+      "epoch": 0.9133789292813584,
+      "grad_norm": 4.415909599002564,
+      "learning_rate": 2.266648938352672e-08,
+      "loss": 1.6398,
+      "step": 13555
+    },
+    {
+      "epoch": 0.9137158451534652,
+      "grad_norm": 4.569776155689798,
+      "learning_rate": 2.249177161401783e-08,
+      "loss": 1.633,
+      "step": 13560
+    },
+    {
+      "epoch": 0.9140527610255719,
+      "grad_norm": 4.694072322744868,
+      "learning_rate": 2.2317714336784422e-08,
+      "loss": 1.6676,
+      "step": 13565
+    },
+    {
+      "epoch": 0.9143896768976787,
+      "grad_norm": 4.6866620357865605,
+      "learning_rate": 2.21443177925838e-08,
+      "loss": 1.6787,
+      "step": 13570
+    },
+    {
+      "epoch": 0.9147265927697854,
+      "grad_norm": 4.860394188870198,
+      "learning_rate": 2.1971582221258944e-08,
+      "loss": 1.5489,
+      "step": 13575
+    },
+    {
+      "epoch": 0.9150635086418921,
+      "grad_norm": 4.196003767451703,
+      "learning_rate": 2.1799507861738788e-08,
+      "loss": 1.6717,
+      "step": 13580
+    },
+    {
+      "epoch": 0.9154004245139988,
+      "grad_norm": 4.565008110985724,
+      "learning_rate": 2.1628094952037713e-08,
+      "loss": 1.657,
+      "step": 13585
+    },
+    {
+      "epoch": 0.9157373403861055,
+      "grad_norm": 4.4052643514904695,
+      "learning_rate": 2.1457343729255062e-08,
+      "loss": 1.6563,
+      "step": 13590
+    },
+    {
+      "epoch": 0.9160742562582124,
+      "grad_norm": 4.565101687356525,
+      "learning_rate": 2.128725442957491e-08,
+      "loss": 1.6745,
+      "step": 13595
+    },
+    {
+      "epoch": 0.9164111721303191,
+      "grad_norm": 4.6695200252580955,
+      "learning_rate": 2.111782728826583e-08,
+      "loss": 1.6678,
+      "step": 13600
+    },
+    {
+      "epoch": 0.9167480880024258,
+      "grad_norm": 4.512246052954719,
+      "learning_rate": 2.0949062539680486e-08,
+      "loss": 1.7268,
+      "step": 13605
+    },
+    {
+      "epoch": 0.9170850038745325,
+      "grad_norm": 4.864232850938187,
+      "learning_rate": 2.07809604172553e-08,
+      "loss": 1.644,
+      "step": 13610
+    },
+    {
+      "epoch": 0.9174219197466392,
+      "grad_norm": 4.5261117115863545,
+      "learning_rate": 2.0613521153510115e-08,
+      "loss": 1.6487,
+      "step": 13615
+    },
+    {
+      "epoch": 0.917758835618746,
+      "grad_norm": 4.720489538078507,
+      "learning_rate": 2.0446744980048002e-08,
+      "loss": 1.6534,
+      "step": 13620
+    },
+    {
+      "epoch": 0.9180957514908528,
+      "grad_norm": 4.8300960188069055,
+      "learning_rate": 2.0280632127554708e-08,
+      "loss": 1.7123,
+      "step": 13625
+    },
+    {
+      "epoch": 0.9184326673629595,
+      "grad_norm": 4.36414883570687,
+      "learning_rate": 2.011518282579855e-08,
+      "loss": 1.6826,
+      "step": 13630
+    },
+    {
+      "epoch": 0.9187695832350662,
+      "grad_norm": 4.419503790016664,
+      "learning_rate": 1.9950397303630075e-08,
+      "loss": 1.7687,
+      "step": 13635
+    },
+    {
+      "epoch": 0.9191064991071729,
+      "grad_norm": 4.517493564314203,
+      "learning_rate": 1.9786275788981565e-08,
+      "loss": 1.6415,
+      "step": 13640
+    },
+    {
+      "epoch": 0.9194434149792797,
+      "grad_norm": 4.63746474526432,
+      "learning_rate": 1.9622818508866823e-08,
+      "loss": 1.6636,
+      "step": 13645
+    },
+    {
+      "epoch": 0.9197803308513864,
+      "grad_norm": 4.791877863432056,
+      "learning_rate": 1.9460025689381043e-08,
+      "loss": 1.7187,
+      "step": 13650
+    },
+    {
+      "epoch": 0.9201172467234932,
+      "grad_norm": 4.074508619807229,
+      "learning_rate": 1.9297897555700216e-08,
+      "loss": 1.6615,
+      "step": 13655
+    },
+    {
+      "epoch": 0.9204541625955999,
+      "grad_norm": 4.488988677170429,
+      "learning_rate": 1.9136434332080898e-08,
+      "loss": 1.6658,
+      "step": 13660
+    },
+    {
+      "epoch": 0.9207910784677066,
+      "grad_norm": 4.572599413440709,
+      "learning_rate": 1.8975636241860048e-08,
+      "loss": 1.6673,
+      "step": 13665
+    },
+    {
+      "epoch": 0.9211279943398133,
+      "grad_norm": 4.591900564378601,
+      "learning_rate": 1.8815503507454644e-08,
+      "loss": 1.5942,
+      "step": 13670
+    },
+    {
+      "epoch": 0.9214649102119201,
+      "grad_norm": 4.651051241662432,
+      "learning_rate": 1.8656036350361117e-08,
+      "loss": 1.6395,
+      "step": 13675
+    },
+    {
+      "epoch": 0.9218018260840268,
+      "grad_norm": 4.819658152787535,
+      "learning_rate": 1.8497234991155463e-08,
+      "loss": 1.6572,
+      "step": 13680
+    },
+    {
+      "epoch": 0.9221387419561335,
+      "grad_norm": 4.300806686835021,
+      "learning_rate": 1.8339099649492762e-08,
+      "loss": 1.6533,
+      "step": 13685
+    },
+    {
+      "epoch": 0.9224756578282403,
+      "grad_norm": 4.702436296805395,
+      "learning_rate": 1.8181630544106653e-08,
+      "loss": 1.6403,
+      "step": 13690
+    },
+    {
+      "epoch": 0.922812573700347,
+      "grad_norm": 5.236256063034201,
+      "learning_rate": 1.8024827892809346e-08,
+      "loss": 1.7093,
+      "step": 13695
+    },
+    {
+      "epoch": 0.9231494895724538,
+      "grad_norm": 4.781425278603808,
+      "learning_rate": 1.7868691912491352e-08,
+      "loss": 1.6634,
+      "step": 13700
+    },
+    {
+      "epoch": 0.9234864054445605,
+      "grad_norm": 5.000408465760196,
+      "learning_rate": 1.77132228191208e-08,
+      "loss": 1.7284,
+      "step": 13705
+    },
+    {
+      "epoch": 0.9238233213166672,
+      "grad_norm": 4.511488885466128,
+      "learning_rate": 1.7558420827743505e-08,
+      "loss": 1.6045,
+      "step": 13710
+    },
+    {
+      "epoch": 0.9241602371887739,
+      "grad_norm": 4.4336520051897415,
+      "learning_rate": 1.7404286152482573e-08,
+      "loss": 1.7259,
+      "step": 13715
+    },
+    {
+      "epoch": 0.9244971530608806,
+      "grad_norm": 4.913660389081854,
+      "learning_rate": 1.725081900653791e-08,
+      "loss": 1.6808,
+      "step": 13720
+    },
+    {
+      "epoch": 0.9248340689329875,
+      "grad_norm": 4.697631827271392,
+      "learning_rate": 1.7098019602186376e-08,
+      "loss": 1.6687,
+      "step": 13725
+    },
+    {
+      "epoch": 0.9251709848050942,
+      "grad_norm": 4.553841575367267,
+      "learning_rate": 1.6945888150780797e-08,
+      "loss": 1.6873,
+      "step": 13730
+    },
+    {
+      "epoch": 0.9255079006772009,
+      "grad_norm": 4.4992528881221485,
+      "learning_rate": 1.6794424862750568e-08,
+      "loss": 1.6747,
+      "step": 13735
+    },
+    {
+      "epoch": 0.9258448165493076,
+      "grad_norm": 5.104827419984812,
+      "learning_rate": 1.664362994760038e-08,
+      "loss": 1.6637,
+      "step": 13740
+    },
+    {
+      "epoch": 0.9261817324214143,
+      "grad_norm": 4.680722518422123,
+      "learning_rate": 1.649350361391083e-08,
+      "loss": 1.6732,
+      "step": 13745
+    },
+    {
+      "epoch": 0.9265186482935212,
+      "grad_norm": 5.31436095897836,
+      "learning_rate": 1.6344046069337646e-08,
+      "loss": 1.6788,
+      "step": 13750
+    },
+    {
+      "epoch": 0.9268555641656279,
+      "grad_norm": 4.234586773332621,
+      "learning_rate": 1.6195257520611182e-08,
+      "loss": 1.6974,
+      "step": 13755
+    },
+    {
+      "epoch": 0.9271924800377346,
+      "grad_norm": 4.272670471430088,
+      "learning_rate": 1.604713817353681e-08,
+      "loss": 1.6658,
+      "step": 13760
+    },
+    {
+      "epoch": 0.9275293959098413,
+      "grad_norm": 4.6885220919191966,
+      "learning_rate": 1.5899688232994147e-08,
+      "loss": 1.6807,
+      "step": 13765
+    },
+    {
+      "epoch": 0.927866311781948,
+      "grad_norm": 4.662232899644492,
+      "learning_rate": 1.5752907902936707e-08,
+      "loss": 1.7172,
+      "step": 13770
+    },
+    {
+      "epoch": 0.9282032276540548,
+      "grad_norm": 4.6833045439786405,
+      "learning_rate": 1.560679738639198e-08,
+      "loss": 1.68,
+      "step": 13775
+    },
+    {
+      "epoch": 0.9285401435261615,
+      "grad_norm": 4.756429528552997,
+      "learning_rate": 1.5461356885461075e-08,
+      "loss": 1.6828,
+      "step": 13780
+    },
+    {
+      "epoch": 0.9288770593982683,
+      "grad_norm": 4.710974308535877,
+      "learning_rate": 1.5316586601317905e-08,
+      "loss": 1.6583,
+      "step": 13785
+    },
+    {
+      "epoch": 0.929213975270375,
+      "grad_norm": 4.489974244824489,
+      "learning_rate": 1.5172486734209788e-08,
+      "loss": 1.6617,
+      "step": 13790
+    },
+    {
+      "epoch": 0.9295508911424817,
+      "grad_norm": 4.54926827654004,
+      "learning_rate": 1.502905748345651e-08,
+      "loss": 1.712,
+      "step": 13795
+    },
+    {
+      "epoch": 0.9298878070145885,
+      "grad_norm": 4.55393602811608,
+      "learning_rate": 1.4886299047450257e-08,
+      "loss": 1.7058,
+      "step": 13800
+    },
+    {
+      "epoch": 0.9302247228866952,
+      "grad_norm": 4.473944643605008,
+      "learning_rate": 1.4744211623655356e-08,
+      "loss": 1.6269,
+      "step": 13805
+    },
+    {
+      "epoch": 0.9305616387588019,
+      "grad_norm": 4.874670195538181,
+      "learning_rate": 1.4602795408607982e-08,
+      "loss": 1.739,
+      "step": 13810
+    },
+    {
+      "epoch": 0.9308985546309086,
+      "grad_norm": 4.354158570656502,
+      "learning_rate": 1.4462050597915942e-08,
+      "loss": 1.6832,
+      "step": 13815
+    },
+    {
+      "epoch": 0.9312354705030154,
+      "grad_norm": 4.510572112208238,
+      "learning_rate": 1.4321977386258289e-08,
+      "loss": 1.6461,
+      "step": 13820
+    },
+    {
+      "epoch": 0.9315723863751222,
+      "grad_norm": 4.497444842424594,
+      "learning_rate": 1.4182575967385092e-08,
+      "loss": 1.683,
+      "step": 13825
+    },
+    {
+      "epoch": 0.9319093022472289,
+      "grad_norm": 4.651578766108472,
+      "learning_rate": 1.4043846534117331e-08,
+      "loss": 1.7365,
+      "step": 13830
+    },
+    {
+      "epoch": 0.9322462181193356,
+      "grad_norm": 4.604232959356164,
+      "learning_rate": 1.3905789278346347e-08,
+      "loss": 1.7186,
+      "step": 13835
+    },
+    {
+      "epoch": 0.9325831339914423,
+      "grad_norm": 4.8235544459128885,
+      "learning_rate": 1.3768404391033717e-08,
+      "loss": 1.6283,
+      "step": 13840
+    },
+    {
+      "epoch": 0.932920049863549,
+      "grad_norm": 4.535664336555867,
+      "learning_rate": 1.3631692062211209e-08,
+      "loss": 1.674,
+      "step": 13845
+    },
+    {
+      "epoch": 0.9332569657356558,
+      "grad_norm": 4.220888842317781,
+      "learning_rate": 1.3495652480979947e-08,
+      "loss": 1.593,
+      "step": 13850
+    },
+    {
+      "epoch": 0.9335938816077626,
+      "grad_norm": 4.7901986238588234,
+      "learning_rate": 1.3360285835510854e-08,
+      "loss": 1.6707,
+      "step": 13855
+    },
+    {
+      "epoch": 0.9339307974798693,
+      "grad_norm": 4.325902736144492,
+      "learning_rate": 1.322559231304382e-08,
+      "loss": 1.6401,
+      "step": 13860
+    },
+    {
+      "epoch": 0.934267713351976,
+      "grad_norm": 4.222275752497884,
+      "learning_rate": 1.3091572099887816e-08,
+      "loss": 1.6086,
+      "step": 13865
+    },
+    {
+      "epoch": 0.9346046292240827,
+      "grad_norm": 4.971828618203986,
+      "learning_rate": 1.2958225381420329e-08,
+      "loss": 1.7208,
+      "step": 13870
+    },
+    {
+      "epoch": 0.9349415450961894,
+      "grad_norm": 4.478315582636149,
+      "learning_rate": 1.282555234208732e-08,
+      "loss": 1.6177,
+      "step": 13875
+    },
+    {
+      "epoch": 0.9352784609682963,
+      "grad_norm": 4.480192189419776,
+      "learning_rate": 1.2693553165403104e-08,
+      "loss": 1.6847,
+      "step": 13880
+    },
+    {
+      "epoch": 0.935615376840403,
+      "grad_norm": 4.280287532922316,
+      "learning_rate": 1.2562228033949628e-08,
+      "loss": 1.6754,
+      "step": 13885
+    },
+    {
+      "epoch": 0.9359522927125097,
+      "grad_norm": 4.6862475175220135,
+      "learning_rate": 1.243157712937659e-08,
+      "loss": 1.673,
+      "step": 13890
+    },
+    {
+      "epoch": 0.9362892085846164,
+      "grad_norm": 4.444473930426543,
+      "learning_rate": 1.230160063240121e-08,
+      "loss": 1.675,
+      "step": 13895
+    },
+    {
+      "epoch": 0.9366261244567231,
+      "grad_norm": 4.765538180006135,
+      "learning_rate": 1.2172298722807617e-08,
+      "loss": 1.6323,
+      "step": 13900
+    },
+    {
+      "epoch": 0.9369630403288299,
+      "grad_norm": 5.033579348083475,
+      "learning_rate": 1.204367157944708e-08,
+      "loss": 1.6684,
+      "step": 13905
+    },
+    {
+      "epoch": 0.9372999562009366,
+      "grad_norm": 4.739347042185302,
+      "learning_rate": 1.19157193802375e-08,
+      "loss": 1.6221,
+      "step": 13910
+    },
+    {
+      "epoch": 0.9376368720730434,
+      "grad_norm": 4.498216776239286,
+      "learning_rate": 1.1788442302163026e-08,
+      "loss": 1.5921,
+      "step": 13915
+    },
+    {
+      "epoch": 0.9379737879451501,
+      "grad_norm": 4.109009748201763,
+      "learning_rate": 1.1661840521274168e-08,
+      "loss": 1.6288,
+      "step": 13920
+    },
+    {
+      "epoch": 0.9383107038172568,
+      "grad_norm": 4.419909793832432,
+      "learning_rate": 1.1535914212687237e-08,
+      "loss": 1.6096,
+      "step": 13925
+    },
+    {
+      "epoch": 0.9386476196893636,
+      "grad_norm": 4.7263234586750125,
+      "learning_rate": 1.1410663550584287e-08,
+      "loss": 1.6956,
+      "step": 13930
+    },
+    {
+      "epoch": 0.9389845355614703,
+      "grad_norm": 4.617355329685817,
+      "learning_rate": 1.1286088708212793e-08,
+      "loss": 1.6558,
+      "step": 13935
+    },
+    {
+      "epoch": 0.939321451433577,
+      "grad_norm": 4.721202684803363,
+      "learning_rate": 1.1162189857885362e-08,
+      "loss": 1.5686,
+      "step": 13940
+    },
+    {
+      "epoch": 0.9396583673056838,
+      "grad_norm": 4.7742840434126474,
+      "learning_rate": 1.1038967170979741e-08,
+      "loss": 1.718,
+      "step": 13945
+    },
+    {
+      "epoch": 0.9399952831777905,
+      "grad_norm": 4.78729301438576,
+      "learning_rate": 1.0916420817938254e-08,
+      "loss": 1.6548,
+      "step": 13950
+    },
+    {
+      "epoch": 0.9403321990498973,
+      "grad_norm": 4.736016915905797,
+      "learning_rate": 1.0794550968267701e-08,
+      "loss": 1.6462,
+      "step": 13955
+    },
+    {
+      "epoch": 0.940669114922004,
+      "grad_norm": 3.9679066975931954,
+      "learning_rate": 1.0673357790539294e-08,
+      "loss": 1.6619,
+      "step": 13960
+    },
+    {
+      "epoch": 0.9410060307941107,
+      "grad_norm": 4.608423358301694,
+      "learning_rate": 1.0552841452388105e-08,
+      "loss": 1.6532,
+      "step": 13965
+    },
+    {
+      "epoch": 0.9413429466662174,
+      "grad_norm": 4.90480375150542,
+      "learning_rate": 1.0433002120513123e-08,
+      "loss": 1.7195,
+      "step": 13970
+    },
+    {
+      "epoch": 0.9416798625383241,
+      "grad_norm": 4.2680828511473665,
+      "learning_rate": 1.0313839960676751e-08,
+      "loss": 1.6184,
+      "step": 13975
+    },
+    {
+      "epoch": 0.942016778410431,
+      "grad_norm": 4.180504127514293,
+      "learning_rate": 1.019535513770492e-08,
+      "loss": 1.5969,
+      "step": 13980
+    },
+    {
+      "epoch": 0.9423536942825377,
+      "grad_norm": 4.751749615006871,
+      "learning_rate": 1.0077547815486476e-08,
+      "loss": 1.7189,
+      "step": 13985
+    },
+    {
+      "epoch": 0.9426906101546444,
+      "grad_norm": 4.557131971025592,
+      "learning_rate": 9.960418156973238e-09,
+      "loss": 1.7,
+      "step": 13990
+    },
+    {
+      "epoch": 0.9430275260267511,
+      "grad_norm": 4.8323487176650355,
+      "learning_rate": 9.843966324179609e-09,
+      "loss": 1.668,
+      "step": 13995
+    },
+    {
+      "epoch": 0.9433644418988578,
+      "grad_norm": 4.241230044588303,
+      "learning_rate": 9.728192478182573e-09,
+      "loss": 1.6605,
+      "step": 14000
+    },
+    {
+      "epoch": 0.9437013577709646,
+      "grad_norm": 4.681284350687982,
+      "learning_rate": 9.613096779121089e-09,
+      "loss": 1.6712,
+      "step": 14005
+    },
+    {
+      "epoch": 0.9440382736430714,
+      "grad_norm": 4.862508336979253,
+      "learning_rate": 9.498679386196417e-09,
+      "loss": 1.5407,
+      "step": 14010
+    },
+    {
+      "epoch": 0.9443751895151781,
+      "grad_norm": 4.877091231184954,
+      "learning_rate": 9.384940457671186e-09,
+      "loss": 1.6501,
+      "step": 14015
+    },
+    {
+      "epoch": 0.9447121053872848,
+      "grad_norm": 4.841069261109426,
+      "learning_rate": 9.271880150869882e-09,
+      "loss": 1.7064,
+      "step": 14020
+    },
+    {
+      "epoch": 0.9450490212593915,
+      "grad_norm": 4.725458488667362,
+      "learning_rate": 9.15949862217824e-09,
+      "loss": 1.684,
+      "step": 14025
+    },
+    {
+      "epoch": 0.9453859371314982,
+      "grad_norm": 4.8010599888096595,
+      "learning_rate": 9.04779602704292e-09,
+      "loss": 1.7179,
+      "step": 14030
+    },
+    {
+      "epoch": 0.945722853003605,
+      "grad_norm": 4.666711180892035,
+      "learning_rate": 8.936772519971769e-09,
+      "loss": 1.6452,
+      "step": 14035
+    },
+    {
+      "epoch": 0.9460597688757117,
+      "grad_norm": 4.563602390830038,
+      "learning_rate": 8.826428254533169e-09,
+      "loss": 1.709,
+      "step": 14040
+    },
+    {
+      "epoch": 0.9463966847478185,
+      "grad_norm": 4.238437561918158,
+      "learning_rate": 8.716763383355862e-09,
+      "loss": 1.6716,
+      "step": 14045
+    },
+    {
+      "epoch": 0.9467336006199252,
+      "grad_norm": 4.777398386226606,
+      "learning_rate": 8.607778058129122e-09,
+      "loss": 1.6786,
+      "step": 14050
+    },
+    {
+      "epoch": 0.9470705164920319,
+      "grad_norm": 4.558655591881119,
+      "learning_rate": 8.499472429601972e-09,
+      "loss": 1.618,
+      "step": 14055
+    },
+    {
+      "epoch": 0.9474074323641387,
+      "grad_norm": 4.747135549338924,
+      "learning_rate": 8.391846647583468e-09,
+      "loss": 1.692,
+      "step": 14060
+    },
+    {
+      "epoch": 0.9477443482362454,
+      "grad_norm": 4.881503970705126,
+      "learning_rate": 8.284900860942246e-09,
+      "loss": 1.7147,
+      "step": 14065
+    },
+    {
+      "epoch": 0.9480812641083521,
+      "grad_norm": 4.694090559014302,
+      "learning_rate": 8.178635217606367e-09,
+      "loss": 1.6388,
+      "step": 14070
+    },
+    {
+      "epoch": 0.9484181799804589,
+      "grad_norm": 4.462356406892563,
+      "learning_rate": 8.073049864563142e-09,
+      "loss": 1.6103,
+      "step": 14075
+    },
+    {
+      "epoch": 0.9487550958525656,
+      "grad_norm": 4.268419903555753,
+      "learning_rate": 7.968144947858801e-09,
+      "loss": 1.6273,
+      "step": 14080
+    },
+    {
+      "epoch": 0.9490920117246724,
+      "grad_norm": 4.72808438390043,
+      "learning_rate": 7.863920612598496e-09,
+      "loss": 1.6714,
+      "step": 14085
+    },
+    {
+      "epoch": 0.9494289275967791,
+      "grad_norm": 4.456322087598728,
+      "learning_rate": 7.760377002945961e-09,
+      "loss": 1.677,
+      "step": 14090
+    },
+    {
+      "epoch": 0.9497658434688858,
+      "grad_norm": 4.381592471889056,
+      "learning_rate": 7.657514262123354e-09,
+      "loss": 1.6303,
+      "step": 14095
+    },
+    {
+      "epoch": 0.9501027593409925,
+      "grad_norm": 5.084398199395626,
+      "learning_rate": 7.55533253241103e-09,
+      "loss": 1.6657,
+      "step": 14100
+    },
+    {
+      "epoch": 0.9504396752130992,
+      "grad_norm": 4.506764254750484,
+      "learning_rate": 7.453831955147428e-09,
+      "loss": 1.67,
+      "step": 14105
+    },
+    {
+      "epoch": 0.9507765910852061,
+      "grad_norm": 4.1594856114480265,
+      "learning_rate": 7.353012670728631e-09,
+      "loss": 1.6293,
+      "step": 14110
+    },
+    {
+      "epoch": 0.9511135069573128,
+      "grad_norm": 4.2947296679424225,
+      "learning_rate": 7.252874818608645e-09,
+      "loss": 1.6333,
+      "step": 14115
+    },
+    {
+      "epoch": 0.9514504228294195,
+      "grad_norm": 4.54393570232016,
+      "learning_rate": 7.153418537298617e-09,
+      "loss": 1.7011,
+      "step": 14120
+    },
+    {
+      "epoch": 0.9517873387015262,
+      "grad_norm": 4.728057239456153,
+      "learning_rate": 7.0546439643671685e-09,
+      "loss": 1.6314,
+      "step": 14125
+    },
+    {
+      "epoch": 0.9521242545736329,
+      "grad_norm": 4.668701870639253,
+      "learning_rate": 6.9565512364398445e-09,
+      "loss": 1.6822,
+      "step": 14130
+    },
+    {
+      "epoch": 0.9524611704457397,
+      "grad_norm": 4.381883653333877,
+      "learning_rate": 6.859140489199167e-09,
+      "loss": 1.6452,
+      "step": 14135
+    },
+    {
+      "epoch": 0.9527980863178465,
+      "grad_norm": 4.4705819028917535,
+      "learning_rate": 6.762411857384187e-09,
+      "loss": 1.71,
+      "step": 14140
+    },
+    {
+      "epoch": 0.9531350021899532,
+      "grad_norm": 5.099898183629039,
+      "learning_rate": 6.666365474790492e-09,
+      "loss": 1.6914,
+      "step": 14145
+    },
+    {
+      "epoch": 0.9534719180620599,
+      "grad_norm": 5.008579652213383,
+      "learning_rate": 6.571001474270144e-09,
+      "loss": 1.6008,
+      "step": 14150
+    },
+    {
+      "epoch": 0.9538088339341666,
+      "grad_norm": 4.4435992523929215,
+      "learning_rate": 6.4763199877311825e-09,
+      "loss": 1.6745,
+      "step": 14155
+    },
+    {
+      "epoch": 0.9541457498062734,
+      "grad_norm": 4.416720324518652,
+      "learning_rate": 6.382321146137571e-09,
+      "loss": 1.7335,
+      "step": 14160
+    },
+    {
+      "epoch": 0.9544826656783801,
+      "grad_norm": 4.6157251630017155,
+      "learning_rate": 6.28900507950908e-09,
+      "loss": 1.6982,
+      "step": 14165
+    },
+    {
+      "epoch": 0.9548195815504869,
+      "grad_norm": 4.576436969845522,
+      "learning_rate": 6.196371916921073e-09,
+      "loss": 1.6821,
+      "step": 14170
+    },
+    {
+      "epoch": 0.9551564974225936,
+      "grad_norm": 4.8030806252738385,
+      "learning_rate": 6.1044217865043325e-09,
+      "loss": 1.6766,
+      "step": 14175
+    },
+    {
+      "epoch": 0.9554934132947003,
+      "grad_norm": 4.788114104485619,
+      "learning_rate": 6.013154815444732e-09,
+      "loss": 1.7202,
+      "step": 14180
+    },
+    {
+      "epoch": 0.9558303291668071,
+      "grad_norm": 4.497660090547918,
+      "learning_rate": 5.922571129983456e-09,
+      "loss": 1.716,
+      "step": 14185
+    },
+    {
+      "epoch": 0.9561672450389138,
+      "grad_norm": 4.977854172865926,
+      "learning_rate": 5.832670855416277e-09,
+      "loss": 1.6901,
+      "step": 14190
+    },
+    {
+      "epoch": 0.9565041609110205,
+      "grad_norm": 4.665062252009661,
+      "learning_rate": 5.7434541160938375e-09,
+      "loss": 1.6698,
+      "step": 14195
+    },
+    {
+      "epoch": 0.9568410767831272,
+      "grad_norm": 4.503462544044519,
+      "learning_rate": 5.6549210354212565e-09,
+      "loss": 1.7074,
+      "step": 14200
+    },
+    {
+      "epoch": 0.957177992655234,
+      "grad_norm": 4.828820266872503,
+      "learning_rate": 5.567071735858131e-09,
+      "loss": 1.6173,
+      "step": 14205
+    },
+    {
+      "epoch": 0.9575149085273407,
+      "grad_norm": 4.577331498261228,
+      "learning_rate": 5.4799063389179834e-09,
+      "loss": 1.6618,
+      "step": 14210
+    },
+    {
+      "epoch": 0.9578518243994475,
+      "grad_norm": 4.70378738630458,
+      "learning_rate": 5.393424965168702e-09,
+      "loss": 1.7063,
+      "step": 14215
+    },
+    {
+      "epoch": 0.9581887402715542,
+      "grad_norm": 5.166081051847853,
+      "learning_rate": 5.307627734231657e-09,
+      "loss": 1.7223,
+      "step": 14220
+    },
+    {
+      "epoch": 0.9585256561436609,
+      "grad_norm": 4.51665984078035,
+      "learning_rate": 5.222514764782193e-09,
+      "loss": 1.735,
+      "step": 14225
+    },
+    {
+      "epoch": 0.9588625720157676,
+      "grad_norm": 4.452215666971605,
+      "learning_rate": 5.138086174549083e-09,
+      "loss": 1.6995,
+      "step": 14230
+    },
+    {
+      "epoch": 0.9591994878878743,
+      "grad_norm": 4.794617784858695,
+      "learning_rate": 5.054342080314522e-09,
+      "loss": 1.6653,
+      "step": 14235
+    },
+    {
+      "epoch": 0.9595364037599812,
+      "grad_norm": 4.384539843873983,
+      "learning_rate": 4.97128259791374e-09,
+      "loss": 1.5975,
+      "step": 14240
+    },
+    {
+      "epoch": 0.9598733196320879,
+      "grad_norm": 4.7112314437798775,
+      "learning_rate": 4.888907842235113e-09,
+      "loss": 1.6114,
+      "step": 14245
+    },
+    {
+      "epoch": 0.9602102355041946,
+      "grad_norm": 4.718198652581357,
+      "learning_rate": 4.807217927220053e-09,
+      "loss": 1.6625,
+      "step": 14250
+    },
+    {
+      "epoch": 0.9605471513763013,
+      "grad_norm": 5.051593564690264,
+      "learning_rate": 4.726212965862342e-09,
+      "loss": 1.6603,
+      "step": 14255
+    },
+    {
+      "epoch": 0.960884067248408,
+      "grad_norm": 4.688799213508077,
+      "learning_rate": 4.645893070208684e-09,
+      "loss": 1.6349,
+      "step": 14260
+    },
+    {
+      "epoch": 0.9612209831205148,
+      "grad_norm": 4.922175546380499,
+      "learning_rate": 4.566258351357988e-09,
+      "loss": 1.6075,
+      "step": 14265
+    },
+    {
+      "epoch": 0.9615578989926216,
+      "grad_norm": 4.917762419900809,
+      "learning_rate": 4.48730891946153e-09,
+      "loss": 1.5986,
+      "step": 14270
+    },
+    {
+      "epoch": 0.9618948148647283,
+      "grad_norm": 4.598810064957941,
+      "learning_rate": 4.409044883722568e-09,
+      "loss": 1.6256,
+      "step": 14275
+    },
+    {
+      "epoch": 0.962231730736835,
+      "grad_norm": 4.356705697302099,
+      "learning_rate": 4.331466352396396e-09,
+      "loss": 1.7101,
+      "step": 14280
+    },
+    {
+      "epoch": 0.9625686466089417,
+      "grad_norm": 4.328389990749196,
+      "learning_rate": 4.2545734327902315e-09,
+      "loss": 1.6263,
+      "step": 14285
+    },
+    {
+      "epoch": 0.9629055624810485,
+      "grad_norm": 4.995883922084056,
+      "learning_rate": 4.178366231262665e-09,
+      "loss": 1.6551,
+      "step": 14290
+    },
+    {
+      "epoch": 0.9632424783531552,
+      "grad_norm": 4.475910128620982,
+      "learning_rate": 4.102844853224041e-09,
+      "loss": 1.6777,
+      "step": 14295
+    },
+    {
+      "epoch": 0.963579394225262,
+      "grad_norm": 4.658822634083728,
+      "learning_rate": 4.028009403135968e-09,
+      "loss": 1.6598,
+      "step": 14300
+    },
+    {
+      "epoch": 0.9639163100973687,
+      "grad_norm": 4.7826733156830255,
+      "learning_rate": 3.95385998451131e-09,
+      "loss": 1.7016,
+      "step": 14305
+    },
+    {
+      "epoch": 0.9642532259694754,
+      "grad_norm": 4.233830555284409,
+      "learning_rate": 3.880396699913968e-09,
+      "loss": 1.6825,
+      "step": 14310
+    },
+    {
+      "epoch": 0.9645901418415822,
+      "grad_norm": 4.995866302785352,
+      "learning_rate": 3.807619650958827e-09,
+      "loss": 1.6545,
+      "step": 14315
+    },
+    {
+      "epoch": 0.9649270577136889,
+      "grad_norm": 4.873719835419959,
+      "learning_rate": 3.7355289383115276e-09,
+      "loss": 1.6973,
+      "step": 14320
+    },
+    {
+      "epoch": 0.9652639735857956,
+      "grad_norm": 4.569288481946704,
+      "learning_rate": 3.664124661688417e-09,
+      "loss": 1.7109,
+      "step": 14325
+    },
+    {
+      "epoch": 0.9656008894579023,
+      "grad_norm": 5.142916963140143,
+      "learning_rate": 3.5934069198562677e-09,
+      "loss": 1.5927,
+      "step": 14330
+    },
+    {
+      "epoch": 0.9659378053300091,
+      "grad_norm": 4.503412551781982,
+      "learning_rate": 3.5233758106322787e-09,
+      "loss": 1.6719,
+      "step": 14335
+    },
+    {
+      "epoch": 0.9662747212021159,
+      "grad_norm": 4.78116076892331,
+      "learning_rate": 3.4540314308839635e-09,
+      "loss": 1.6598,
+      "step": 14340
+    },
+    {
+      "epoch": 0.9666116370742226,
+      "grad_norm": 4.503686920939599,
+      "learning_rate": 3.385373876528874e-09,
+      "loss": 1.7066,
+      "step": 14345
+    },
+    {
+      "epoch": 0.9669485529463293,
+      "grad_norm": 4.86363545716925,
+      "learning_rate": 3.3174032425345444e-09,
+      "loss": 1.6455,
+      "step": 14350
+    },
+    {
+      "epoch": 0.967285468818436,
+      "grad_norm": 4.69142574421157,
+      "learning_rate": 3.250119622918379e-09,
+      "loss": 1.6145,
+      "step": 14355
+    },
+    {
+      "epoch": 0.9676223846905427,
+      "grad_norm": 4.598668097010771,
+      "learning_rate": 3.1835231107474323e-09,
+      "loss": 1.6796,
+      "step": 14360
+    },
+    {
+      "epoch": 0.9679593005626496,
+      "grad_norm": 4.704583569768571,
+      "learning_rate": 3.1176137981385185e-09,
+      "loss": 1.6821,
+      "step": 14365
+    },
+    {
+      "epoch": 0.9682962164347563,
+      "grad_norm": 4.451333372128447,
+      "learning_rate": 3.0523917762576568e-09,
+      "loss": 1.6355,
+      "step": 14370
+    },
+    {
+      "epoch": 0.968633132306863,
+      "grad_norm": 4.419630961259664,
+      "learning_rate": 2.9878571353204595e-09,
+      "loss": 1.7504,
+      "step": 14375
+    },
+    {
+      "epoch": 0.9689700481789697,
+      "grad_norm": 4.668303542736468,
+      "learning_rate": 2.924009964591578e-09,
+      "loss": 1.6296,
+      "step": 14380
+    },
+    {
+      "epoch": 0.9693069640510764,
+      "grad_norm": 4.481187207413503,
+      "learning_rate": 2.8608503523848136e-09,
+      "loss": 1.7121,
+      "step": 14385
+    },
+    {
+      "epoch": 0.9696438799231831,
+      "grad_norm": 5.098540734967322,
+      "learning_rate": 2.7983783860629496e-09,
+      "loss": 1.5944,
+      "step": 14390
+    },
+    {
+      "epoch": 0.96998079579529,
+      "grad_norm": 4.44133751167953,
+      "learning_rate": 2.7365941520375303e-09,
+      "loss": 1.6733,
+      "step": 14395
+    },
+    {
+      "epoch": 0.9703177116673967,
+      "grad_norm": 4.46888227527462,
+      "learning_rate": 2.6754977357689724e-09,
+      "loss": 1.6552,
+      "step": 14400
+    },
+    {
+      "epoch": 0.9706546275395034,
+      "grad_norm": 4.348433001446189,
+      "learning_rate": 2.6150892217660647e-09,
+      "loss": 1.5949,
+      "step": 14405
+    },
+    {
+      "epoch": 0.9709915434116101,
+      "grad_norm": 4.788880812069154,
+      "learning_rate": 2.5553686935864126e-09,
+      "loss": 1.7321,
+      "step": 14410
+    },
+    {
+      "epoch": 0.9713284592837168,
+      "grad_norm": 4.64048628521751,
+      "learning_rate": 2.496336233835661e-09,
+      "loss": 1.6318,
+      "step": 14415
+    },
+    {
+      "epoch": 0.9716653751558236,
+      "grad_norm": 5.386824440878254,
+      "learning_rate": 2.437991924167937e-09,
+      "loss": 1.72,
+      "step": 14420
+    },
+    {
+      "epoch": 0.9720022910279303,
+      "grad_norm": 4.345030594331005,
+      "learning_rate": 2.380335845285464e-09,
+      "loss": 1.6684,
+      "step": 14425
+    },
+    {
+      "epoch": 0.9723392069000371,
+      "grad_norm": 5.205111442516552,
+      "learning_rate": 2.323368076938448e-09,
+      "loss": 1.6683,
+      "step": 14430
+    },
+    {
+      "epoch": 0.9726761227721438,
+      "grad_norm": 4.539297425674057,
+      "learning_rate": 2.2670886979250235e-09,
+      "loss": 1.6197,
+      "step": 14435
+    },
+    {
+      "epoch": 0.9730130386442505,
+      "grad_norm": 5.3078680962201386,
+      "learning_rate": 2.211497786091143e-09,
+      "loss": 1.6613,
+      "step": 14440
+    },
+    {
+      "epoch": 0.9733499545163573,
+      "grad_norm": 4.093183911517651,
+      "learning_rate": 2.1565954183306313e-09,
+      "loss": 1.6449,
+      "step": 14445
+    },
+    {
+      "epoch": 0.973686870388464,
+      "grad_norm": 4.956998797141432,
+      "learning_rate": 2.1023816705846853e-09,
+      "loss": 1.677,
+      "step": 14450
+    },
+    {
+      "epoch": 0.9740237862605707,
+      "grad_norm": 4.401344494084036,
+      "learning_rate": 2.048856617842043e-09,
+      "loss": 1.5762,
+      "step": 14455
+    },
+    {
+      "epoch": 0.9743607021326774,
+      "grad_norm": 4.460249220034049,
+      "learning_rate": 1.9960203341389813e-09,
+      "loss": 1.6107,
+      "step": 14460
+    },
+    {
+      "epoch": 0.9746976180047842,
+      "grad_norm": 4.684091839068217,
+      "learning_rate": 1.943872892558929e-09,
+      "loss": 1.693,
+      "step": 14465
+    },
+    {
+      "epoch": 0.975034533876891,
+      "grad_norm": 4.627584897372627,
+      "learning_rate": 1.8924143652325196e-09,
+      "loss": 1.6675,
+      "step": 14470
+    },
+    {
+      "epoch": 0.9753714497489977,
+      "grad_norm": 4.5804086807255855,
+      "learning_rate": 1.8416448233374848e-09,
+      "loss": 1.6034,
+      "step": 14475
+    },
+    {
+      "epoch": 0.9757083656211044,
+      "grad_norm": 4.784239841374212,
+      "learning_rate": 1.79156433709865e-09,
+      "loss": 1.698,
+      "step": 14480
+    },
+    {
+      "epoch": 0.9760452814932111,
+      "grad_norm": 4.50530685834436,
+      "learning_rate": 1.742172975787548e-09,
+      "loss": 1.6361,
+      "step": 14485
+    },
+    {
+      "epoch": 0.9763821973653178,
+      "grad_norm": 4.68302514179738,
+      "learning_rate": 1.6934708077226411e-09,
+      "loss": 1.6463,
+      "step": 14490
+    },
+    {
+      "epoch": 0.9767191132374247,
+      "grad_norm": 4.602230205378361,
+      "learning_rate": 1.6454579002690982e-09,
+      "loss": 1.6715,
+      "step": 14495
+    },
+    {
+      "epoch": 0.9770560291095314,
+      "grad_norm": 4.913828834206417,
+      "learning_rate": 1.5981343198386288e-09,
+      "loss": 1.6651,
+      "step": 14500
+    },
+    {
+      "epoch": 0.9773929449816381,
+      "grad_norm": 4.734186118385281,
+      "learning_rate": 1.5515001318895382e-09,
+      "loss": 1.6073,
+      "step": 14505
+    },
+    {
+      "epoch": 0.9777298608537448,
+      "grad_norm": 4.4545084698284505,
+      "learning_rate": 1.5055554009264505e-09,
+      "loss": 1.6929,
+      "step": 14510
+    },
+    {
+      "epoch": 0.9780667767258515,
+      "grad_norm": 4.785564771151706,
+      "learning_rate": 1.4603001905004187e-09,
+      "loss": 1.6346,
+      "step": 14515
+    },
+    {
+      "epoch": 0.9784036925979583,
+      "grad_norm": 4.645996933378816,
+      "learning_rate": 1.4157345632087592e-09,
+      "loss": 1.5743,
+      "step": 14520
+    },
+    {
+      "epoch": 0.978740608470065,
+      "grad_norm": 4.391419944898517,
+      "learning_rate": 1.37185858069494e-09,
+      "loss": 1.6349,
+      "step": 14525
+    },
+    {
+      "epoch": 0.9790775243421718,
+      "grad_norm": 4.301204477830003,
+      "learning_rate": 1.328672303648415e-09,
+      "loss": 1.6671,
+      "step": 14530
+    },
+    {
+      "epoch": 0.9794144402142785,
+      "grad_norm": 4.65876078456801,
+      "learning_rate": 1.2861757918046778e-09,
+      "loss": 1.6325,
+      "step": 14535
+    },
+    {
+      "epoch": 0.9797513560863852,
+      "grad_norm": 4.656387193046978,
+      "learning_rate": 1.2443691039452642e-09,
+      "loss": 1.6799,
+      "step": 14540
+    },
+    {
+      "epoch": 0.980088271958492,
+      "grad_norm": 4.856871488888764,
+      "learning_rate": 1.203252297897417e-09,
+      "loss": 1.6712,
+      "step": 14545
+    },
+    {
+      "epoch": 0.9804251878305987,
+      "grad_norm": 4.4246714925954755,
+      "learning_rate": 1.1628254305340869e-09,
+      "loss": 1.6516,
+      "step": 14550
+    },
+    {
+      "epoch": 0.9807621037027054,
+      "grad_norm": 4.861361000481014,
+      "learning_rate": 1.123088557773988e-09,
+      "loss": 1.6693,
+      "step": 14555
+    },
+    {
+      "epoch": 0.9810990195748122,
+      "grad_norm": 4.5445239138117595,
+      "learning_rate": 1.0840417345814312e-09,
+      "loss": 1.6838,
+      "step": 14560
+    },
+    {
+      "epoch": 0.9814359354469189,
+      "grad_norm": 4.296831382491651,
+      "learning_rate": 1.0456850149662134e-09,
+      "loss": 1.6672,
+      "step": 14565
+    },
+    {
+      "epoch": 0.9817728513190256,
+      "grad_norm": 4.360980348229525,
+      "learning_rate": 1.0080184519835056e-09,
+      "loss": 1.6291,
+      "step": 14570
+    },
+    {
+      "epoch": 0.9821097671911324,
+      "grad_norm": 4.6410736774559584,
+      "learning_rate": 9.71042097734076e-10,
+      "loss": 1.5869,
+      "step": 14575
+    },
+    {
+      "epoch": 0.9824466830632391,
+      "grad_norm": 4.672315989477686,
+      "learning_rate": 9.347560033637347e-10,
+      "loss": 1.687,
+      "step": 14580
+    },
+    {
+      "epoch": 0.9827835989353458,
+      "grad_norm": 4.760298790473192,
+      "learning_rate": 8.991602190636105e-10,
+      "loss": 1.6779,
+      "step": 14585
+    },
+    {
+      "epoch": 0.9831205148074526,
+      "grad_norm": 4.5956833044108105,
+      "learning_rate": 8.642547940700961e-10,
+      "loss": 1.6209,
+      "step": 14590
+    },
+    {
+      "epoch": 0.9834574306795593,
+      "grad_norm": 4.852699110524061,
+      "learning_rate": 8.300397766644595e-10,
+      "loss": 1.66,
+      "step": 14595
+    },
+    {
+      "epoch": 0.9837943465516661,
+      "grad_norm": 4.573099918969528,
+      "learning_rate": 7.965152141732878e-10,
+      "loss": 1.6092,
+      "step": 14600
+    },
+    {
+      "epoch": 0.9841312624237728,
+      "grad_norm": 5.041789334977715,
+      "learning_rate": 7.636811529678211e-10,
+      "loss": 1.7897,
+      "step": 14605
+    },
+    {
+      "epoch": 0.9844681782958795,
+      "grad_norm": 4.865882608674358,
+      "learning_rate": 7.315376384643968e-10,
+      "loss": 1.6522,
+      "step": 14610
+    },
+    {
+      "epoch": 0.9848050941679862,
+      "grad_norm": 4.754816598507756,
+      "learning_rate": 7.000847151240608e-10,
+      "loss": 1.6973,
+      "step": 14615
+    },
+    {
+      "epoch": 0.9851420100400929,
+      "grad_norm": 4.544176767785006,
+      "learning_rate": 6.693224264527897e-10,
+      "loss": 1.6629,
+      "step": 14620
+    },
+    {
+      "epoch": 0.9854789259121998,
+      "grad_norm": 4.345974794927777,
+      "learning_rate": 6.392508150011023e-10,
+      "loss": 1.6632,
+      "step": 14625
+    },
+    {
+      "epoch": 0.9858158417843065,
+      "grad_norm": 4.473605639509804,
+      "learning_rate": 6.098699223641701e-10,
+      "loss": 1.6382,
+      "step": 14630
+    },
+    {
+      "epoch": 0.9861527576564132,
+      "grad_norm": 4.2474880352741256,
+      "learning_rate": 5.811797891819847e-10,
+      "loss": 1.6289,
+      "step": 14635
+    },
+    {
+      "epoch": 0.9864896735285199,
+      "grad_norm": 4.706147394472757,
+      "learning_rate": 5.531804551387464e-10,
+      "loss": 1.6884,
+      "step": 14640
+    },
+    {
+      "epoch": 0.9868265894006266,
+      "grad_norm": 4.6198804722109506,
+      "learning_rate": 5.258719589634198e-10,
+      "loss": 1.5749,
+      "step": 14645
+    },
+    {
+      "epoch": 0.9871635052727334,
+      "grad_norm": 5.169200607824497,
+      "learning_rate": 4.992543384291781e-10,
+      "loss": 1.6929,
+      "step": 14650
+    },
+    {
+      "epoch": 0.9875004211448402,
+      "grad_norm": 5.093086280987769,
+      "learning_rate": 4.733276303537925e-10,
+      "loss": 1.7123,
+      "step": 14655
+    },
+    {
+      "epoch": 0.9878373370169469,
+      "grad_norm": 4.469256470389324,
+      "learning_rate": 4.480918705991321e-10,
+      "loss": 1.6728,
+      "step": 14660
+    },
+    {
+      "epoch": 0.9881742528890536,
+      "grad_norm": 4.937986515455233,
+      "learning_rate": 4.235470940715524e-10,
+      "loss": 1.6923,
+      "step": 14665
+    },
+    {
+      "epoch": 0.9885111687611603,
+      "grad_norm": 4.774095402654197,
+      "learning_rate": 3.99693334721507e-10,
+      "loss": 1.6331,
+      "step": 14670
+    },
+    {
+      "epoch": 0.9888480846332671,
+      "grad_norm": 4.895959891617084,
+      "learning_rate": 3.765306255436029e-10,
+      "loss": 1.6954,
+      "step": 14675
+    },
+    {
+      "epoch": 0.9891850005053738,
+      "grad_norm": 4.588526481995063,
+      "learning_rate": 3.540589985766562e-10,
+      "loss": 1.6825,
+      "step": 14680
+    },
+    {
+      "epoch": 0.9895219163774805,
+      "grad_norm": 4.632597593371431,
+      "learning_rate": 3.322784849036364e-10,
+      "loss": 1.6749,
+      "step": 14685
+    },
+    {
+      "epoch": 0.9898588322495873,
+      "grad_norm": 4.7885886471647,
+      "learning_rate": 3.11189114651389e-10,
+      "loss": 1.5306,
+      "step": 14690
+    },
+    {
+      "epoch": 0.990195748121694,
+      "grad_norm": 5.4431857664056285,
+      "learning_rate": 2.9079091699091287e-10,
+      "loss": 1.6503,
+      "step": 14695
+    },
+    {
+      "epoch": 0.9905326639938008,
+      "grad_norm": 4.560136830848719,
+      "learning_rate": 2.710839201370829e-10,
+      "loss": 1.7277,
+      "step": 14700
+    },
+    {
+      "epoch": 0.9908695798659075,
+      "grad_norm": 4.534976189027352,
+      "learning_rate": 2.5206815134881655e-10,
+      "loss": 1.6648,
+      "step": 14705
+    },
+    {
+      "epoch": 0.9912064957380142,
+      "grad_norm": 4.549435008199678,
+      "learning_rate": 2.337436369287404e-10,
+      "loss": 1.6607,
+      "step": 14710
+    },
+    {
+      "epoch": 0.9915434116101209,
+      "grad_norm": 4.429287754065413,
+      "learning_rate": 2.1611040222346833e-10,
+      "loss": 1.6488,
+      "step": 14715
+    },
+    {
+      "epoch": 0.9918803274822277,
+      "grad_norm": 4.450792157939868,
+      "learning_rate": 1.9916847162343432e-10,
+      "loss": 1.6863,
+      "step": 14720
+    },
+    {
+      "epoch": 0.9922172433543345,
+      "grad_norm": 4.456301955946831,
+      "learning_rate": 1.829178685627264e-10,
+      "loss": 1.6766,
+      "step": 14725
+    },
+    {
+      "epoch": 0.9925541592264412,
+      "grad_norm": 4.536387361240234,
+      "learning_rate": 1.6735861551936402e-10,
+      "loss": 1.6072,
+      "step": 14730
+    },
+    {
+      "epoch": 0.9928910750985479,
+      "grad_norm": 4.5301760503399064,
+      "learning_rate": 1.5249073401502055e-10,
+      "loss": 1.6165,
+      "step": 14735
+    },
+    {
+      "epoch": 0.9932279909706546,
+      "grad_norm": 4.626033879980004,
+      "learning_rate": 1.3831424461496766e-10,
+      "loss": 1.6704,
+      "step": 14740
+    },
+    {
+      "epoch": 0.9935649068427613,
+      "grad_norm": 4.638098471878353,
+      "learning_rate": 1.2482916692824197e-10,
+      "loss": 1.6164,
+      "step": 14745
+    },
+    {
+      "epoch": 0.993901822714868,
+      "grad_norm": 4.640388344532539,
+      "learning_rate": 1.1203551960742297e-10,
+      "loss": 1.6507,
+      "step": 14750
+    },
+    {
+      "epoch": 0.9942387385869749,
+      "grad_norm": 4.703240878023018,
+      "learning_rate": 9.993332034891056e-11,
+      "loss": 1.6233,
+      "step": 14755
+    },
+    {
+      "epoch": 0.9945756544590816,
+      "grad_norm": 4.59940282508,
+      "learning_rate": 8.852258589236994e-11,
+      "loss": 1.6481,
+      "step": 14760
+    },
+    {
+      "epoch": 0.9949125703311883,
+      "grad_norm": 4.95979390658074,
+      "learning_rate": 7.780333202134226e-11,
+      "loss": 1.6457,
+      "step": 14765
+    },
+    {
+      "epoch": 0.995249486203295,
+      "grad_norm": 4.709459134710041,
+      "learning_rate": 6.777557356263397e-11,
+      "loss": 1.6773,
+      "step": 14770
+    },
+    {
+      "epoch": 0.9955864020754017,
+      "grad_norm": 4.6127703869383545,
+      "learning_rate": 5.843932438681643e-11,
+      "loss": 1.7132,
+      "step": 14775
+    },
+    {
+      "epoch": 0.9959233179475085,
+      "grad_norm": 4.708135222754572,
+      "learning_rate": 4.97945974077818e-11,
+      "loss": 1.7054,
+      "step": 14780
+    },
+    {
+      "epoch": 0.9962602338196153,
+      "grad_norm": 5.013474030961767,
+      "learning_rate": 4.1841404582965143e-11,
+      "loss": 1.6342,
+      "step": 14785
+    },
+    {
+      "epoch": 0.996597149691722,
+      "grad_norm": 4.300498430663266,
+      "learning_rate": 3.457975691334436e-11,
+      "loss": 1.5682,
+      "step": 14790
+    },
+    {
+      "epoch": 0.9969340655638287,
+      "grad_norm": 4.445836794215863,
+      "learning_rate": 2.800966444316266e-11,
+      "loss": 1.6325,
+      "step": 14795
+    },
+    {
+      "epoch": 0.9972709814359354,
+      "grad_norm": 4.491154818955257,
+      "learning_rate": 2.213113626026164e-11,
+      "loss": 1.6928,
+      "step": 14800
+    },
+    {
+      "epoch": 0.9976078973080422,
+      "grad_norm": 4.732056867538516,
+      "learning_rate": 1.6944180495914728e-11,
+      "loss": 1.5832,
+      "step": 14805
+    },
+    {
+      "epoch": 0.9979448131801489,
+      "grad_norm": 4.736707269460872,
+      "learning_rate": 1.2448804324660667e-11,
+      "loss": 1.7032,
+      "step": 14810
+    },
+    {
+      "epoch": 0.9982817290522557,
+      "grad_norm": 4.805801710749307,
+      "learning_rate": 8.645013964581061e-12,
+      "loss": 1.7129,
+      "step": 14815
+    },
+    {
+      "epoch": 0.9986186449243624,
+      "grad_norm": 4.748295330431418,
+      "learning_rate": 5.532814677133846e-12,
+      "loss": 1.684,
+      "step": 14820
+    },
+    {
+      "epoch": 0.9989555607964691,
+      "grad_norm": 4.76094035726937,
+      "learning_rate": 3.112210767042267e-12,
+      "loss": 1.692,
+      "step": 14825
+    },
+    {
+      "epoch": 0.9992924766685759,
+      "grad_norm": 4.743959538402305,
+      "learning_rate": 1.383205582516922e-12,
+      "loss": 1.5827,
+      "step": 14830
+    },
+    {
+      "epoch": 0.9996293925406826,
+      "grad_norm": 4.50407060815357,
+      "learning_rate": 3.4580151520025024e-13,
+      "loss": 1.7556,
+      "step": 14835
+    },
+    {
+      "epoch": 0.9999663084127893,
+      "grad_norm": 4.787374440275167,
+      "learning_rate": 0.0,
+      "loss": 1.6964,
+      "step": 14840
+    },
+    {
+      "epoch": 0.9999663084127893,
+      "eval_loss": NaN,
+      "eval_runtime": 130.4563,
+      "eval_samples_per_second": 36.779,
+      "eval_steps_per_second": 4.599,
+      "step": 14840
+    },
+    {
+      "epoch": 0.9999663084127893,
+      "step": 14840,
+      "total_flos": 1.8023025589165752e+18,
+      "train_loss": 1.7489710538856424,
+      "train_runtime": 43730.9212,
+      "train_samples_per_second": 10.859,
+      "train_steps_per_second": 0.339
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 14840,
+  "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": 1.8023025589165752e+18,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}