diff --git "a/trainer_state.json" "b/trainer_state.json"
--- "a/trainer_state.json"
+++ "b/trainer_state.json"
@@ -1,55 +1,8910 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 1.7142857142857144,
+  "epoch": 1.9995261786306564,
   "eval_steps": 500,
-  "global_step": 6,
+  "global_step": 6330,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.2857142857142857,
-      "grad_norm": 153.4065134934994,
-      "learning_rate": 0.001,
-      "loss": 21.8182,
+      "epoch": 0.0003158809128958383,
+      "grad_norm": 0.2789897719448856,
+      "learning_rate": 1.5797788309636651e-06,
+      "loss": 1.5618,
       "step": 1
     },
     {
-      "epoch": 0.8571428571428571,
-      "eval_loss": 17.570310592651367,
-      "eval_runtime": 0.3894,
-      "eval_samples_per_second": 5.136,
-      "eval_steps_per_second": 2.568,
-      "step": 3
+      "epoch": 0.0015794045644791914,
+      "grad_norm": 0.2892532097035974,
+      "learning_rate": 7.898894154818326e-06,
+      "loss": 1.491,
+      "step": 5
+    },
+    {
+      "epoch": 0.003158809128958383,
+      "grad_norm": 0.2677271963629727,
+      "learning_rate": 1.579778830963665e-05,
+      "loss": 1.5398,
+      "step": 10
+    },
+    {
+      "epoch": 0.004738213693437574,
+      "grad_norm": 0.17664334604863555,
+      "learning_rate": 2.3696682464454976e-05,
+      "loss": 1.4923,
+      "step": 15
+    },
+    {
+      "epoch": 0.006317618257916766,
+      "grad_norm": 0.1837074587639523,
+      "learning_rate": 3.15955766192733e-05,
+      "loss": 1.4707,
+      "step": 20
+    },
+    {
+      "epoch": 0.007897022822395957,
+      "grad_norm": 0.10674506928787326,
+      "learning_rate": 3.949447077409163e-05,
+      "loss": 1.4078,
+      "step": 25
+    },
+    {
+      "epoch": 0.009476427386875147,
+      "grad_norm": 0.09158554885809744,
+      "learning_rate": 4.739336492890995e-05,
+      "loss": 1.3651,
+      "step": 30
+    },
+    {
+      "epoch": 0.01105583195135434,
+      "grad_norm": 0.08454382107096206,
+      "learning_rate": 5.529225908372828e-05,
+      "loss": 1.3675,
+      "step": 35
+    },
+    {
+      "epoch": 0.012635236515833531,
+      "grad_norm": 0.06493854926352545,
+      "learning_rate": 6.31911532385466e-05,
+      "loss": 1.3226,
+      "step": 40
+    },
+    {
+      "epoch": 0.014214641080312722,
+      "grad_norm": 0.05673072207914187,
+      "learning_rate": 7.109004739336493e-05,
+      "loss": 1.3366,
+      "step": 45
+    },
+    {
+      "epoch": 0.015794045644791914,
+      "grad_norm": 0.05491667593602107,
+      "learning_rate": 7.898894154818326e-05,
+      "loss": 1.3201,
+      "step": 50
+    },
+    {
+      "epoch": 0.017373450209271106,
+      "grad_norm": 0.04964918512060703,
+      "learning_rate": 8.688783570300159e-05,
+      "loss": 1.2821,
+      "step": 55
+    },
+    {
+      "epoch": 0.018952854773750295,
+      "grad_norm": 0.05508256002416552,
+      "learning_rate": 9.47867298578199e-05,
+      "loss": 1.3296,
+      "step": 60
+    },
+    {
+      "epoch": 0.020532259338229487,
+      "grad_norm": 0.05145117328690555,
+      "learning_rate": 0.00010268562401263823,
+      "loss": 1.2242,
+      "step": 65
+    },
+    {
+      "epoch": 0.02211166390270868,
+      "grad_norm": 0.04856402203442387,
+      "learning_rate": 0.00011058451816745656,
+      "loss": 1.3047,
+      "step": 70
+    },
+    {
+      "epoch": 0.02369106846718787,
+      "grad_norm": 0.04791827380722405,
+      "learning_rate": 0.00011848341232227489,
+      "loss": 1.2547,
+      "step": 75
+    },
+    {
+      "epoch": 0.025270473031667063,
+      "grad_norm": 0.04950986336105823,
+      "learning_rate": 0.0001263823064770932,
+      "loss": 1.2478,
+      "step": 80
+    },
+    {
+      "epoch": 0.02684987759614625,
+      "grad_norm": 0.05115323866433944,
+      "learning_rate": 0.00013428120063191154,
+      "loss": 1.252,
+      "step": 85
+    },
+    {
+      "epoch": 0.028429282160625444,
+      "grad_norm": 0.050403331262978614,
+      "learning_rate": 0.00014218009478672987,
+      "loss": 1.2436,
+      "step": 90
+    },
+    {
+      "epoch": 0.030008686725104636,
+      "grad_norm": 0.047802558286327325,
+      "learning_rate": 0.0001500789889415482,
+      "loss": 1.2154,
+      "step": 95
+    },
+    {
+      "epoch": 0.03158809128958383,
+      "grad_norm": 0.058289368434715695,
+      "learning_rate": 0.00015797788309636652,
+      "loss": 1.2263,
+      "step": 100
+    },
+    {
+      "epoch": 0.03316749585406302,
+      "grad_norm": 0.052893240483671625,
+      "learning_rate": 0.00016587677725118482,
+      "loss": 1.2554,
+      "step": 105
+    },
+    {
+      "epoch": 0.03474690041854221,
+      "grad_norm": 0.05755557590352308,
+      "learning_rate": 0.00017377567140600318,
+      "loss": 1.2039,
+      "step": 110
+    },
+    {
+      "epoch": 0.036326304983021404,
+      "grad_norm": 0.05223463762568313,
+      "learning_rate": 0.00018167456556082148,
+      "loss": 1.1944,
+      "step": 115
+    },
+    {
+      "epoch": 0.03790570954750059,
+      "grad_norm": 0.05491452017858886,
+      "learning_rate": 0.0001895734597156398,
+      "loss": 1.2906,
+      "step": 120
+    },
+    {
+      "epoch": 0.03948511411197978,
+      "grad_norm": 0.05543633621414441,
+      "learning_rate": 0.00019747235387045813,
+      "loss": 1.2448,
+      "step": 125
+    },
+    {
+      "epoch": 0.04106451867645897,
+      "grad_norm": 0.05358848718368862,
+      "learning_rate": 0.00020537124802527646,
+      "loss": 1.2237,
+      "step": 130
+    },
+    {
+      "epoch": 0.042643923240938165,
+      "grad_norm": 0.06179885136266294,
+      "learning_rate": 0.0002132701421800948,
+      "loss": 1.2079,
+      "step": 135
+    },
+    {
+      "epoch": 0.04422332780541736,
+      "grad_norm": 0.05100931620495029,
+      "learning_rate": 0.00022116903633491312,
+      "loss": 1.3063,
+      "step": 140
+    },
+    {
+      "epoch": 0.04580273236989655,
+      "grad_norm": 0.0530018665651464,
+      "learning_rate": 0.00022906793048973144,
+      "loss": 1.1266,
+      "step": 145
+    },
+    {
+      "epoch": 0.04738213693437574,
+      "grad_norm": 0.06146301369787031,
+      "learning_rate": 0.00023696682464454977,
+      "loss": 1.2175,
+      "step": 150
+    },
+    {
+      "epoch": 0.048961541498854934,
+      "grad_norm": 0.058591714657941774,
+      "learning_rate": 0.0002448657187993681,
+      "loss": 1.201,
+      "step": 155
+    },
+    {
+      "epoch": 0.050540946063334126,
+      "grad_norm": 0.05380313755008262,
+      "learning_rate": 0.0002527646129541864,
+      "loss": 1.1665,
+      "step": 160
+    },
+    {
+      "epoch": 0.05212035062781331,
+      "grad_norm": 0.058558871257419466,
+      "learning_rate": 0.00026066350710900475,
+      "loss": 1.1877,
+      "step": 165
+    },
+    {
+      "epoch": 0.0536997551922925,
+      "grad_norm": 0.06380402461857115,
+      "learning_rate": 0.0002685624012638231,
+      "loss": 1.1482,
+      "step": 170
+    },
+    {
+      "epoch": 0.055279159756771695,
+      "grad_norm": 0.05479882567291893,
+      "learning_rate": 0.0002764612954186414,
+      "loss": 1.1207,
+      "step": 175
+    },
+    {
+      "epoch": 0.05685856432125089,
+      "grad_norm": 0.05855467690650077,
+      "learning_rate": 0.00028436018957345974,
+      "loss": 1.1685,
+      "step": 180
+    },
+    {
+      "epoch": 0.05843796888573008,
+      "grad_norm": 0.05564570468690808,
+      "learning_rate": 0.000292259083728278,
+      "loss": 1.1683,
+      "step": 185
+    },
+    {
+      "epoch": 0.06001737345020927,
+      "grad_norm": 0.0494359899772851,
+      "learning_rate": 0.0003001579778830964,
+      "loss": 1.2459,
+      "step": 190
+    },
+    {
+      "epoch": 0.061596778014688464,
+      "grad_norm": 0.05502377946893621,
+      "learning_rate": 0.0003080568720379147,
+      "loss": 1.1759,
+      "step": 195
+    },
+    {
+      "epoch": 0.06317618257916766,
+      "grad_norm": 0.04826733876181895,
+      "learning_rate": 0.00031595576619273305,
+      "loss": 1.2097,
+      "step": 200
+    },
+    {
+      "epoch": 0.06475558714364685,
+      "grad_norm": 0.06130531868163813,
+      "learning_rate": 0.0003238546603475513,
+      "loss": 1.2117,
+      "step": 205
+    },
+    {
+      "epoch": 0.06633499170812604,
+      "grad_norm": 0.05579359766498949,
+      "learning_rate": 0.00033175355450236965,
+      "loss": 1.1871,
+      "step": 210
+    },
+    {
+      "epoch": 0.06791439627260523,
+      "grad_norm": 0.056126237444499225,
+      "learning_rate": 0.000339652448657188,
+      "loss": 1.2226,
+      "step": 215
+    },
+    {
+      "epoch": 0.06949380083708442,
+      "grad_norm": 0.057607712153461374,
+      "learning_rate": 0.00034755134281200636,
+      "loss": 1.1417,
+      "step": 220
+    },
+    {
+      "epoch": 0.07107320540156362,
+      "grad_norm": 0.059680902683061025,
+      "learning_rate": 0.0003554502369668247,
+      "loss": 1.2532,
+      "step": 225
+    },
+    {
+      "epoch": 0.07265260996604281,
+      "grad_norm": 0.059660465306138444,
+      "learning_rate": 0.00036334913112164296,
+      "loss": 1.2279,
+      "step": 230
+    },
+    {
+      "epoch": 0.07423201453052199,
+      "grad_norm": 0.07025494414796972,
+      "learning_rate": 0.0003712480252764613,
+      "loss": 1.1505,
+      "step": 235
+    },
+    {
+      "epoch": 0.07581141909500118,
+      "grad_norm": 0.0700230881662308,
+      "learning_rate": 0.0003791469194312796,
+      "loss": 1.1622,
+      "step": 240
+    },
+    {
+      "epoch": 0.07739082365948037,
+      "grad_norm": 0.06259280855025934,
+      "learning_rate": 0.000387045813586098,
+      "loss": 1.1703,
+      "step": 245
+    },
+    {
+      "epoch": 0.07897022822395956,
+      "grad_norm": 0.05846912427798551,
+      "learning_rate": 0.00039494470774091627,
+      "loss": 1.1557,
+      "step": 250
+    },
+    {
+      "epoch": 0.08054963278843875,
+      "grad_norm": 0.059126444692307836,
+      "learning_rate": 0.0004028436018957346,
+      "loss": 1.1398,
+      "step": 255
+    },
+    {
+      "epoch": 0.08212903735291795,
+      "grad_norm": 0.0636582968368329,
+      "learning_rate": 0.0004107424960505529,
+      "loss": 1.1896,
+      "step": 260
+    },
+    {
+      "epoch": 0.08370844191739714,
+      "grad_norm": 0.054771718653728006,
+      "learning_rate": 0.00041864139020537125,
+      "loss": 1.1395,
+      "step": 265
+    },
+    {
+      "epoch": 0.08528784648187633,
+      "grad_norm": 0.052046726233918324,
+      "learning_rate": 0.0004265402843601896,
+      "loss": 1.1524,
+      "step": 270
+    },
+    {
+      "epoch": 0.08686725104635552,
+      "grad_norm": 0.05505440033261828,
+      "learning_rate": 0.0004344391785150079,
+      "loss": 1.2445,
+      "step": 275
+    },
+    {
+      "epoch": 0.08844665561083472,
+      "grad_norm": 0.05229126324293584,
+      "learning_rate": 0.00044233807266982623,
+      "loss": 1.1923,
+      "step": 280
+    },
+    {
+      "epoch": 0.09002606017531391,
+      "grad_norm": 0.056018163896207424,
+      "learning_rate": 0.00045023696682464456,
+      "loss": 1.2435,
+      "step": 285
+    },
+    {
+      "epoch": 0.0916054647397931,
+      "grad_norm": 0.05939226225728438,
+      "learning_rate": 0.0004581358609794629,
+      "loss": 1.1854,
+      "step": 290
+    },
+    {
+      "epoch": 0.09318486930427229,
+      "grad_norm": 0.060756291168264476,
+      "learning_rate": 0.0004660347551342812,
+      "loss": 1.1893,
+      "step": 295
+    },
+    {
+      "epoch": 0.09476427386875148,
+      "grad_norm": 0.05904003385364645,
+      "learning_rate": 0.00047393364928909954,
+      "loss": 1.2337,
+      "step": 300
+    },
+    {
+      "epoch": 0.09634367843323068,
+      "grad_norm": 0.056832912998354725,
+      "learning_rate": 0.00048183254344391787,
+      "loss": 1.184,
+      "step": 305
+    },
+    {
+      "epoch": 0.09792308299770987,
+      "grad_norm": 0.06534146054108758,
+      "learning_rate": 0.0004897314375987362,
+      "loss": 1.2271,
+      "step": 310
+    },
+    {
+      "epoch": 0.09950248756218906,
+      "grad_norm": 0.049751901789561674,
+      "learning_rate": 0.0004976303317535545,
+      "loss": 1.1458,
+      "step": 315
+    },
+    {
+      "epoch": 0.10108189212666825,
+      "grad_norm": 0.04997659125094788,
+      "learning_rate": 0.0005055292259083729,
+      "loss": 1.109,
+      "step": 320
+    },
+    {
+      "epoch": 0.10266129669114744,
+      "grad_norm": 0.061972227867431645,
+      "learning_rate": 0.0005134281200631912,
+      "loss": 1.1755,
+      "step": 325
+    },
+    {
+      "epoch": 0.10424070125562662,
+      "grad_norm": 0.07995162193950708,
+      "learning_rate": 0.0005213270142180095,
+      "loss": 1.1901,
+      "step": 330
+    },
+    {
+      "epoch": 0.10582010582010581,
+      "grad_norm": 0.05838498126795504,
+      "learning_rate": 0.0005292259083728278,
+      "loss": 1.1629,
+      "step": 335
+    },
+    {
+      "epoch": 0.107399510384585,
+      "grad_norm": 0.12200420477732127,
+      "learning_rate": 0.0005371248025276462,
+      "loss": 1.231,
+      "step": 340
+    },
+    {
+      "epoch": 0.1089789149490642,
+      "grad_norm": 0.05707978807468486,
+      "learning_rate": 0.0005450236966824644,
+      "loss": 1.1957,
+      "step": 345
+    },
+    {
+      "epoch": 0.11055831951354339,
+      "grad_norm": 0.06161334754095597,
+      "learning_rate": 0.0005529225908372828,
+      "loss": 1.2422,
+      "step": 350
+    },
+    {
+      "epoch": 0.11213772407802258,
+      "grad_norm": 0.05656475357013885,
+      "learning_rate": 0.0005608214849921011,
+      "loss": 1.2008,
+      "step": 355
+    },
+    {
+      "epoch": 0.11371712864250177,
+      "grad_norm": 0.056239595168653025,
+      "learning_rate": 0.0005687203791469195,
+      "loss": 1.2125,
+      "step": 360
+    },
+    {
+      "epoch": 0.11529653320698097,
+      "grad_norm": 0.058071575217863304,
+      "learning_rate": 0.0005766192733017378,
+      "loss": 1.2465,
+      "step": 365
+    },
+    {
+      "epoch": 0.11687593777146016,
+      "grad_norm": 0.06278972665091191,
+      "learning_rate": 0.000584518167456556,
+      "loss": 1.2111,
+      "step": 370
+    },
+    {
+      "epoch": 0.11845534233593935,
+      "grad_norm": 0.05703645077122024,
+      "learning_rate": 0.0005924170616113745,
+      "loss": 1.1096,
+      "step": 375
+    },
+    {
+      "epoch": 0.12003474690041854,
+      "grad_norm": 0.06994936365804813,
+      "learning_rate": 0.0006003159557661928,
+      "loss": 1.1155,
+      "step": 380
+    },
+    {
+      "epoch": 0.12161415146489774,
+      "grad_norm": 0.052372697841875995,
+      "learning_rate": 0.0006082148499210111,
+      "loss": 1.2188,
+      "step": 385
+    },
+    {
+      "epoch": 0.12319355602937693,
+      "grad_norm": 0.06645079677500605,
+      "learning_rate": 0.0006161137440758294,
+      "loss": 1.1516,
+      "step": 390
+    },
+    {
+      "epoch": 0.12477296059385612,
+      "grad_norm": 0.06386365975552465,
+      "learning_rate": 0.0006240126382306477,
+      "loss": 1.1409,
+      "step": 395
+    },
+    {
+      "epoch": 0.1263523651583353,
+      "grad_norm": 0.051198724511214086,
+      "learning_rate": 0.0006319115323854661,
+      "loss": 1.1699,
+      "step": 400
+    },
+    {
+      "epoch": 0.1279317697228145,
+      "grad_norm": 0.047789704621754345,
+      "learning_rate": 0.0006398104265402843,
+      "loss": 1.2111,
+      "step": 405
+    },
+    {
+      "epoch": 0.1295111742872937,
+      "grad_norm": 0.08292897928400877,
+      "learning_rate": 0.0006477093206951026,
+      "loss": 1.2084,
+      "step": 410
+    },
+    {
+      "epoch": 0.1310905788517729,
+      "grad_norm": 0.053440634299855995,
+      "learning_rate": 0.0006556082148499211,
+      "loss": 1.1919,
+      "step": 415
+    },
+    {
+      "epoch": 0.13266998341625208,
+      "grad_norm": 0.05318482784273273,
+      "learning_rate": 0.0006635071090047393,
+      "loss": 1.1451,
+      "step": 420
+    },
+    {
+      "epoch": 0.13424938798073127,
+      "grad_norm": 0.05665917212538103,
+      "learning_rate": 0.0006714060031595577,
+      "loss": 1.1411,
+      "step": 425
+    },
+    {
+      "epoch": 0.13582879254521046,
+      "grad_norm": 0.05574161677632433,
+      "learning_rate": 0.000679304897314376,
+      "loss": 1.1071,
+      "step": 430
+    },
+    {
+      "epoch": 0.13740819710968966,
+      "grad_norm": 0.05821755155831752,
+      "learning_rate": 0.0006872037914691943,
+      "loss": 1.1786,
+      "step": 435
+    },
+    {
+      "epoch": 0.13898760167416885,
+      "grad_norm": 0.06093075040564606,
+      "learning_rate": 0.0006951026856240127,
+      "loss": 1.092,
+      "step": 440
+    },
+    {
+      "epoch": 0.14056700623864804,
+      "grad_norm": 0.05825340962494296,
+      "learning_rate": 0.0007030015797788309,
+      "loss": 1.1626,
+      "step": 445
+    },
+    {
+      "epoch": 0.14214641080312723,
+      "grad_norm": 0.05020938068401297,
+      "learning_rate": 0.0007109004739336494,
+      "loss": 1.1822,
+      "step": 450
+    },
+    {
+      "epoch": 0.14372581536760642,
+      "grad_norm": 0.0557961806261948,
+      "learning_rate": 0.0007187993680884676,
+      "loss": 1.2147,
+      "step": 455
+    },
+    {
+      "epoch": 0.14530521993208562,
+      "grad_norm": 0.05437488653301554,
+      "learning_rate": 0.0007266982622432859,
+      "loss": 1.1479,
+      "step": 460
+    },
+    {
+      "epoch": 0.1468846244965648,
+      "grad_norm": 0.05268833934110884,
+      "learning_rate": 0.0007345971563981043,
+      "loss": 1.1473,
+      "step": 465
+    },
+    {
+      "epoch": 0.14846402906104397,
+      "grad_norm": 0.04493291882173135,
+      "learning_rate": 0.0007424960505529226,
+      "loss": 1.1499,
+      "step": 470
+    },
+    {
+      "epoch": 0.15004343362552316,
+      "grad_norm": 0.04452601287960959,
+      "learning_rate": 0.0007503949447077409,
+      "loss": 1.1688,
+      "step": 475
+    },
+    {
+      "epoch": 0.15162283819000236,
+      "grad_norm": 0.050748020091030625,
+      "learning_rate": 0.0007582938388625592,
+      "loss": 1.2168,
+      "step": 480
+    },
+    {
+      "epoch": 0.15320224275448155,
+      "grad_norm": 0.16891299036710675,
+      "learning_rate": 0.0007661927330173775,
+      "loss": 1.1953,
+      "step": 485
+    },
+    {
+      "epoch": 0.15478164731896074,
+      "grad_norm": 0.0683474674929215,
+      "learning_rate": 0.000774091627172196,
+      "loss": 1.2254,
+      "step": 490
+    },
+    {
+      "epoch": 0.15636105188343993,
+      "grad_norm": 0.06256971833638193,
+      "learning_rate": 0.0007819905213270142,
+      "loss": 1.2087,
+      "step": 495
+    },
+    {
+      "epoch": 0.15794045644791913,
+      "grad_norm": 0.05436563080376856,
+      "learning_rate": 0.0007898894154818325,
+      "loss": 1.1562,
+      "step": 500
+    },
+    {
+      "epoch": 0.15951986101239832,
+      "grad_norm": 0.055684137284571354,
+      "learning_rate": 0.0007977883096366509,
+      "loss": 1.1949,
+      "step": 505
+    },
+    {
+      "epoch": 0.1610992655768775,
+      "grad_norm": 0.050335288455669695,
+      "learning_rate": 0.0008056872037914692,
+      "loss": 1.1596,
+      "step": 510
+    },
+    {
+      "epoch": 0.1626786701413567,
+      "grad_norm": 0.05206965203978437,
+      "learning_rate": 0.0008135860979462876,
+      "loss": 1.2096,
+      "step": 515
+    },
+    {
+      "epoch": 0.1642580747058359,
+      "grad_norm": 0.06839415660825547,
+      "learning_rate": 0.0008214849921011058,
+      "loss": 1.2405,
+      "step": 520
+    },
+    {
+      "epoch": 0.16583747927031509,
+      "grad_norm": 0.05934969682125384,
+      "learning_rate": 0.0008293838862559242,
+      "loss": 1.1575,
+      "step": 525
+    },
+    {
+      "epoch": 0.16741688383479428,
+      "grad_norm": 0.048248270161178164,
+      "learning_rate": 0.0008372827804107425,
+      "loss": 1.1297,
+      "step": 530
+    },
+    {
+      "epoch": 0.16899628839927347,
+      "grad_norm": 0.04841154135491026,
+      "learning_rate": 0.0008451816745655608,
+      "loss": 1.1778,
+      "step": 535
+    },
+    {
+      "epoch": 0.17057569296375266,
+      "grad_norm": 0.04839285015392914,
+      "learning_rate": 0.0008530805687203792,
+      "loss": 1.1367,
+      "step": 540
+    },
+    {
+      "epoch": 0.17215509752823185,
+      "grad_norm": 0.05085171971926083,
+      "learning_rate": 0.0008609794628751975,
+      "loss": 1.1527,
+      "step": 545
+    },
+    {
+      "epoch": 0.17373450209271105,
+      "grad_norm": 0.048045360550468874,
+      "learning_rate": 0.0008688783570300158,
+      "loss": 1.2203,
+      "step": 550
+    },
+    {
+      "epoch": 0.17531390665719024,
+      "grad_norm": 0.05360566007829095,
+      "learning_rate": 0.0008767772511848341,
+      "loss": 1.191,
+      "step": 555
+    },
+    {
+      "epoch": 0.17689331122166943,
+      "grad_norm": 0.05153608252672655,
+      "learning_rate": 0.0008846761453396525,
+      "loss": 1.2965,
+      "step": 560
+    },
+    {
+      "epoch": 0.17847271578614862,
+      "grad_norm": 0.0537492984979501,
+      "learning_rate": 0.0008925750394944708,
+      "loss": 1.253,
+      "step": 565
+    },
+    {
+      "epoch": 0.18005212035062781,
+      "grad_norm": 0.059570203946716534,
+      "learning_rate": 0.0009004739336492891,
+      "loss": 1.1733,
+      "step": 570
+    },
+    {
+      "epoch": 0.181631524915107,
+      "grad_norm": 0.07310330867652265,
+      "learning_rate": 0.0009083728278041074,
+      "loss": 1.2127,
+      "step": 575
+    },
+    {
+      "epoch": 0.1832109294795862,
+      "grad_norm": 0.048429065249010646,
+      "learning_rate": 0.0009162717219589258,
+      "loss": 1.24,
+      "step": 580
+    },
+    {
+      "epoch": 0.1847903340440654,
+      "grad_norm": 0.05274469740017941,
+      "learning_rate": 0.0009241706161137441,
+      "loss": 1.1546,
+      "step": 585
+    },
+    {
+      "epoch": 0.18636973860854458,
+      "grad_norm": 0.04378705257118556,
+      "learning_rate": 0.0009320695102685624,
+      "loss": 1.1437,
+      "step": 590
+    },
+    {
+      "epoch": 0.18794914317302377,
+      "grad_norm": 0.05384819336054559,
+      "learning_rate": 0.0009399684044233808,
+      "loss": 1.1763,
+      "step": 595
+    },
+    {
+      "epoch": 0.18952854773750297,
+      "grad_norm": 0.07965381715501382,
+      "learning_rate": 0.0009478672985781991,
+      "loss": 1.2411,
+      "step": 600
+    },
+    {
+      "epoch": 0.19110795230198216,
+      "grad_norm": 0.059004614843427866,
+      "learning_rate": 0.0009557661927330173,
+      "loss": 1.2445,
+      "step": 605
+    },
+    {
+      "epoch": 0.19268735686646135,
+      "grad_norm": 0.05280984994326182,
+      "learning_rate": 0.0009636650868878357,
+      "loss": 1.3118,
+      "step": 610
+    },
+    {
+      "epoch": 0.19426676143094054,
+      "grad_norm": 0.04254342942038169,
+      "learning_rate": 0.0009715639810426541,
+      "loss": 1.306,
+      "step": 615
+    },
+    {
+      "epoch": 0.19584616599541974,
+      "grad_norm": 0.056142331071331146,
+      "learning_rate": 0.0009794628751974724,
+      "loss": 1.1764,
+      "step": 620
+    },
+    {
+      "epoch": 0.19742557055989893,
+      "grad_norm": 0.055000939660971034,
+      "learning_rate": 0.0009873617693522906,
+      "loss": 1.2869,
+      "step": 625
+    },
+    {
+      "epoch": 0.19900497512437812,
+      "grad_norm": 0.05848265110756856,
+      "learning_rate": 0.000995260663507109,
+      "loss": 1.2242,
+      "step": 630
+    },
+    {
+      "epoch": 0.2005843796888573,
+      "grad_norm": 0.1289533179840951,
+      "learning_rate": 0.0009999996959064125,
+      "loss": 1.2347,
+      "step": 635
+    },
+    {
+      "epoch": 0.2021637842533365,
+      "grad_norm": 0.0634015519760411,
+      "learning_rate": 0.0009999962748577986,
+      "loss": 1.296,
+      "step": 640
+    },
+    {
+      "epoch": 0.2037431888178157,
+      "grad_norm": 0.07254139591414524,
+      "learning_rate": 0.0009999890526696813,
+      "loss": 1.1969,
+      "step": 645
+    },
+    {
+      "epoch": 0.2053225933822949,
+      "grad_norm": 0.06309255152166017,
+      "learning_rate": 0.0009999780293969657,
+      "loss": 1.2262,
+      "step": 650
+    },
+    {
+      "epoch": 0.20690199794677408,
+      "grad_norm": 0.06898019039350595,
+      "learning_rate": 0.0009999632051234547,
+      "loss": 1.2251,
+      "step": 655
+    },
+    {
+      "epoch": 0.20848140251125324,
+      "grad_norm": 0.06259386096143452,
+      "learning_rate": 0.000999944579961847,
+      "loss": 1.3119,
+      "step": 660
+    },
+    {
+      "epoch": 0.21006080707573244,
+      "grad_norm": 0.06872950124265154,
+      "learning_rate": 0.0009999221540537377,
+      "loss": 1.1715,
+      "step": 665
+    },
+    {
+      "epoch": 0.21164021164021163,
+      "grad_norm": 0.9937019497079367,
+      "learning_rate": 0.000999895927569616,
+      "loss": 1.3005,
+      "step": 670
+    },
+    {
+      "epoch": 0.21321961620469082,
+      "grad_norm": 0.1354849534169185,
+      "learning_rate": 0.0009998659007088642,
+      "loss": 1.2387,
+      "step": 675
+    },
+    {
+      "epoch": 0.21479902076917,
+      "grad_norm": 0.07498188430417127,
+      "learning_rate": 0.0009998320736997568,
+      "loss": 1.2051,
+      "step": 680
+    },
+    {
+      "epoch": 0.2163784253336492,
+      "grad_norm": 2.499051766047346,
+      "learning_rate": 0.0009997944467994581,
+      "loss": 1.259,
+      "step": 685
+    },
+    {
+      "epoch": 0.2179578298981284,
+      "grad_norm": 2.048237623936503,
+      "learning_rate": 0.0009997530202940205,
+      "loss": 1.4247,
+      "step": 690
+    },
+    {
+      "epoch": 0.2195372344626076,
+      "grad_norm": 0.9568708158453652,
+      "learning_rate": 0.0009997077944983819,
+      "loss": 1.6289,
+      "step": 695
+    },
+    {
+      "epoch": 0.22111663902708678,
+      "grad_norm": 0.17185653456476527,
+      "learning_rate": 0.0009996587697563642,
+      "loss": 1.5628,
+      "step": 700
+    },
+    {
+      "epoch": 0.22269604359156597,
+      "grad_norm": 0.17039026137566396,
+      "learning_rate": 0.00099960594644067,
+      "loss": 1.4037,
+      "step": 705
+    },
+    {
+      "epoch": 0.22427544815604517,
+      "grad_norm": 5.215297156818115,
+      "learning_rate": 0.0009995493249528795,
+      "loss": 1.4015,
+      "step": 710
+    },
+    {
+      "epoch": 0.22585485272052436,
+      "grad_norm": 0.8050682876390816,
+      "learning_rate": 0.0009994889057234487,
+      "loss": 1.4124,
+      "step": 715
+    },
+    {
+      "epoch": 0.22743425728500355,
+      "grad_norm": 0.16011890487013522,
+      "learning_rate": 0.0009994246892117045,
+      "loss": 1.6748,
+      "step": 720
+    },
+    {
+      "epoch": 0.22901366184948274,
+      "grad_norm": 0.11806743445365343,
+      "learning_rate": 0.0009993566759058429,
+      "loss": 1.6867,
+      "step": 725
+    },
+    {
+      "epoch": 0.23059306641396193,
+      "grad_norm": 0.1422914641138834,
+      "learning_rate": 0.0009992848663229231,
+      "loss": 1.5684,
+      "step": 730
+    },
+    {
+      "epoch": 0.23217247097844113,
+      "grad_norm": 0.10039336502710226,
+      "learning_rate": 0.0009992092610088662,
+      "loss": 1.3594,
+      "step": 735
+    },
+    {
+      "epoch": 0.23375187554292032,
+      "grad_norm": 0.08353358714541645,
+      "learning_rate": 0.0009991298605384492,
+      "loss": 1.3553,
+      "step": 740
+    },
+    {
+      "epoch": 0.2353312801073995,
+      "grad_norm": 0.06028998203574107,
+      "learning_rate": 0.000999046665515301,
+      "loss": 1.2517,
+      "step": 745
+    },
+    {
+      "epoch": 0.2369106846718787,
+      "grad_norm": 0.1069757369931618,
+      "learning_rate": 0.0009989596765718981,
+      "loss": 1.3012,
+      "step": 750
+    },
+    {
+      "epoch": 0.2384900892363579,
+      "grad_norm": 0.057695063797185045,
+      "learning_rate": 0.0009988688943695595,
+      "loss": 1.3095,
+      "step": 755
+    },
+    {
+      "epoch": 0.24006949380083709,
+      "grad_norm": 0.05321345064994074,
+      "learning_rate": 0.000998774319598442,
+      "loss": 1.2603,
+      "step": 760
+    },
+    {
+      "epoch": 0.24164889836531628,
+      "grad_norm": 0.06181569765694586,
+      "learning_rate": 0.0009986759529775349,
+      "loss": 1.2347,
+      "step": 765
+    },
+    {
+      "epoch": 0.24322830292979547,
+      "grad_norm": 0.04799581558822299,
+      "learning_rate": 0.0009985737952546542,
+      "loss": 1.2514,
+      "step": 770
+    },
+    {
+      "epoch": 0.24480770749427466,
+      "grad_norm": 0.05659151468826247,
+      "learning_rate": 0.0009984678472064374,
+      "loss": 1.2609,
+      "step": 775
+    },
+    {
+      "epoch": 0.24638711205875385,
+      "grad_norm": 0.05127796227154854,
+      "learning_rate": 0.0009983581096383368,
+      "loss": 1.2511,
+      "step": 780
+    },
+    {
+      "epoch": 0.24796651662323305,
+      "grad_norm": 0.049337974506391276,
+      "learning_rate": 0.0009982445833846146,
+      "loss": 1.2434,
+      "step": 785
+    },
+    {
+      "epoch": 0.24954592118771224,
+      "grad_norm": 0.06282817628594659,
+      "learning_rate": 0.0009981272693083349,
+      "loss": 1.2281,
+      "step": 790
+    },
+    {
+      "epoch": 0.25112532575219143,
+      "grad_norm": 0.05360363073011154,
+      "learning_rate": 0.0009980061683013592,
+      "loss": 1.261,
+      "step": 795
+    },
+    {
+      "epoch": 0.2527047303166706,
+      "grad_norm": 0.0435497395031424,
+      "learning_rate": 0.0009978812812843378,
+      "loss": 1.2305,
+      "step": 800
+    },
+    {
+      "epoch": 0.2542841348811498,
+      "grad_norm": 0.0701758491964102,
+      "learning_rate": 0.0009977526092067037,
+      "loss": 1.2389,
+      "step": 805
+    },
+    {
+      "epoch": 0.255863539445629,
+      "grad_norm": 0.04528535565880187,
+      "learning_rate": 0.0009976201530466655,
+      "loss": 1.1771,
+      "step": 810
+    },
+    {
+      "epoch": 0.2574429440101082,
+      "grad_norm": 0.04304696595003719,
+      "learning_rate": 0.0009974839138111988,
+      "loss": 1.1608,
+      "step": 815
+    },
+    {
+      "epoch": 0.2590223485745874,
+      "grad_norm": 0.05501271003472879,
+      "learning_rate": 0.0009973438925360407,
+      "loss": 1.2104,
+      "step": 820
+    },
+    {
+      "epoch": 0.2606017531390666,
+      "grad_norm": 0.06777008626565485,
+      "learning_rate": 0.0009972000902856795,
+      "loss": 1.2131,
+      "step": 825
+    },
+    {
+      "epoch": 0.2621811577035458,
+      "grad_norm": 0.03968699801552576,
+      "learning_rate": 0.0009970525081533482,
+      "loss": 1.1724,
+      "step": 830
+    },
+    {
+      "epoch": 0.26376056226802497,
+      "grad_norm": 0.07383457025829056,
+      "learning_rate": 0.0009969011472610158,
+      "loss": 1.1637,
+      "step": 835
+    },
+    {
+      "epoch": 0.26533996683250416,
+      "grad_norm": 0.05318282155905444,
+      "learning_rate": 0.0009967460087593786,
+      "loss": 1.1844,
+      "step": 840
+    },
+    {
+      "epoch": 0.26691937139698335,
+      "grad_norm": 0.0438038229881027,
+      "learning_rate": 0.0009965870938278517,
+      "loss": 1.3114,
+      "step": 845
+    },
+    {
+      "epoch": 0.26849877596146254,
+      "grad_norm": 0.044682976599825244,
+      "learning_rate": 0.0009964244036745594,
+      "loss": 1.1801,
+      "step": 850
+    },
+    {
+      "epoch": 0.27007818052594174,
+      "grad_norm": 0.04931452257775692,
+      "learning_rate": 0.000996257939536327,
+      "loss": 1.2087,
+      "step": 855
+    },
+    {
+      "epoch": 0.2716575850904209,
+      "grad_norm": 0.08794796118059663,
+      "learning_rate": 0.0009960877026786708,
+      "loss": 1.2521,
+      "step": 860
+    },
+    {
+      "epoch": 0.2732369896549001,
+      "grad_norm": 0.06090026057096657,
+      "learning_rate": 0.0009959136943957887,
+      "loss": 1.2126,
+      "step": 865
+    },
+    {
+      "epoch": 0.2748163942193793,
+      "grad_norm": 0.046466185921548456,
+      "learning_rate": 0.0009957359160105497,
+      "loss": 1.2087,
+      "step": 870
+    },
+    {
+      "epoch": 0.2763957987838585,
+      "grad_norm": 0.22759376527429676,
+      "learning_rate": 0.000995554368874485,
+      "loss": 1.2451,
+      "step": 875
+    },
+    {
+      "epoch": 0.2779752033483377,
+      "grad_norm": 0.046606113089797695,
+      "learning_rate": 0.0009953690543677768,
+      "loss": 1.2507,
+      "step": 880
+    },
+    {
+      "epoch": 0.2795546079128169,
+      "grad_norm": 0.05471794231886829,
+      "learning_rate": 0.0009951799738992485,
+      "loss": 1.2073,
+      "step": 885
+    },
+    {
+      "epoch": 0.2811340124772961,
+      "grad_norm": 2.3469505748015114,
+      "learning_rate": 0.0009949871289063525,
+      "loss": 1.1663,
+      "step": 890
+    },
+    {
+      "epoch": 0.2827134170417753,
+      "grad_norm": 0.04849558180772164,
+      "learning_rate": 0.000994790520855162,
+      "loss": 1.2141,
+      "step": 895
+    },
+    {
+      "epoch": 0.28429282160625446,
+      "grad_norm": 0.06306917133411176,
+      "learning_rate": 0.0009945901512403569,
+      "loss": 1.1744,
+      "step": 900
+    },
+    {
+      "epoch": 0.28587222617073366,
+      "grad_norm": 0.16182143067555266,
+      "learning_rate": 0.0009943860215852144,
+      "loss": 1.1608,
+      "step": 905
+    },
+    {
+      "epoch": 0.28745163073521285,
+      "grad_norm": 0.0875751596694063,
+      "learning_rate": 0.0009941781334415966,
+      "loss": 1.2676,
+      "step": 910
+    },
+    {
+      "epoch": 0.28903103529969204,
+      "grad_norm": 0.05190083485865879,
+      "learning_rate": 0.0009939664883899394,
+      "loss": 1.1948,
+      "step": 915
+    },
+    {
+      "epoch": 0.29061043986417123,
+      "grad_norm": 0.0446450044232843,
+      "learning_rate": 0.0009937510880392386,
+      "loss": 1.1432,
+      "step": 920
+    },
+    {
+      "epoch": 0.2921898444286504,
+      "grad_norm": 0.03769202980158123,
+      "learning_rate": 0.0009935319340270408,
+      "loss": 1.2313,
+      "step": 925
+    },
+    {
+      "epoch": 0.2937692489931296,
+      "grad_norm": 0.06718904210860335,
+      "learning_rate": 0.0009933090280194279,
+      "loss": 1.1912,
+      "step": 930
+    },
+    {
+      "epoch": 0.2953486535576088,
+      "grad_norm": 0.045744113522662125,
+      "learning_rate": 0.0009930823717110065,
+      "loss": 1.1365,
+      "step": 935
+    },
+    {
+      "epoch": 0.29692805812208795,
+      "grad_norm": 0.043786588552808724,
+      "learning_rate": 0.0009928519668248937,
+      "loss": 1.1888,
+      "step": 940
+    },
+    {
+      "epoch": 0.29850746268656714,
+      "grad_norm": 0.039005039119245764,
+      "learning_rate": 0.0009926178151127049,
+      "loss": 1.2473,
+      "step": 945
+    },
+    {
+      "epoch": 0.30008686725104633,
+      "grad_norm": 0.050876769890127016,
+      "learning_rate": 0.0009923799183545398,
+      "loss": 1.1889,
+      "step": 950
+    },
+    {
+      "epoch": 0.3016662718155255,
+      "grad_norm": 0.04575948024145222,
+      "learning_rate": 0.0009921382783589696,
+      "loss": 1.171,
+      "step": 955
+    },
+    {
+      "epoch": 0.3032456763800047,
+      "grad_norm": 0.05017713187941563,
+      "learning_rate": 0.0009918928969630228,
+      "loss": 1.1756,
+      "step": 960
+    },
+    {
+      "epoch": 0.3048250809444839,
+      "grad_norm": 0.04345272263747149,
+      "learning_rate": 0.0009916437760321708,
+      "loss": 1.1517,
+      "step": 965
+    },
+    {
+      "epoch": 0.3064044855089631,
+      "grad_norm": 0.05131866420472819,
+      "learning_rate": 0.0009913909174603147,
+      "loss": 1.2095,
+      "step": 970
+    },
+    {
+      "epoch": 0.3079838900734423,
+      "grad_norm": 0.05715628036510762,
+      "learning_rate": 0.0009911343231697703,
+      "loss": 1.2204,
+      "step": 975
+    },
+    {
+      "epoch": 0.3095632946379215,
+      "grad_norm": 0.04788259122222203,
+      "learning_rate": 0.0009908739951112534,
+      "loss": 1.2256,
+      "step": 980
+    },
+    {
+      "epoch": 0.3111426992024007,
+      "grad_norm": 0.04502265853173252,
+      "learning_rate": 0.0009906099352638652,
+      "loss": 1.2243,
+      "step": 985
+    },
+    {
+      "epoch": 0.31272210376687987,
+      "grad_norm": 0.040922281342619146,
+      "learning_rate": 0.0009903421456350775,
+      "loss": 1.1412,
+      "step": 990
+    },
+    {
+      "epoch": 0.31430150833135906,
+      "grad_norm": 0.03777821636386108,
+      "learning_rate": 0.000990070628260717,
+      "loss": 1.1867,
+      "step": 995
+    },
+    {
+      "epoch": 0.31588091289583825,
+      "grad_norm": 0.041838269964072416,
+      "learning_rate": 0.0009897953852049494,
+      "loss": 1.1813,
+      "step": 1000
+    },
+    {
+      "epoch": 0.31746031746031744,
+      "grad_norm": 0.09066613740501203,
+      "learning_rate": 0.0009895164185602654,
+      "loss": 1.2916,
+      "step": 1005
+    },
+    {
+      "epoch": 0.31903972202479663,
+      "grad_norm": 0.05545106093802797,
+      "learning_rate": 0.0009892337304474629,
+      "loss": 1.2022,
+      "step": 1010
+    },
+    {
+      "epoch": 0.3206191265892758,
+      "grad_norm": 0.05823267457176772,
+      "learning_rate": 0.0009889473230156316,
+      "loss": 1.1614,
+      "step": 1015
+    },
+    {
+      "epoch": 0.322198531153755,
+      "grad_norm": 0.045599148101737805,
+      "learning_rate": 0.000988657198442137,
+      "loss": 1.1038,
+      "step": 1020
+    },
+    {
+      "epoch": 0.3237779357182342,
+      "grad_norm": 0.0438141101705124,
+      "learning_rate": 0.0009883633589326038,
+      "loss": 1.122,
+      "step": 1025
+    },
+    {
+      "epoch": 0.3253573402827134,
+      "grad_norm": 0.036910474407238554,
+      "learning_rate": 0.000988065806720898,
+      "loss": 1.2285,
+      "step": 1030
+    },
+    {
+      "epoch": 0.3269367448471926,
+      "grad_norm": 0.04459568322437855,
+      "learning_rate": 0.0009877645440691122,
+      "loss": 1.2004,
+      "step": 1035
+    },
+    {
+      "epoch": 0.3285161494116718,
+      "grad_norm": 0.06088171196269682,
+      "learning_rate": 0.0009874595732675454,
+      "loss": 1.2027,
+      "step": 1040
+    },
+    {
+      "epoch": 0.330095553976151,
+      "grad_norm": 0.1301305299184432,
+      "learning_rate": 0.0009871508966346882,
+      "loss": 1.1417,
+      "step": 1045
+    },
+    {
+      "epoch": 0.33167495854063017,
+      "grad_norm": 0.0650018665905558,
+      "learning_rate": 0.0009868385165172043,
+      "loss": 1.1807,
+      "step": 1050
+    },
+    {
+      "epoch": 0.33325436310510936,
+      "grad_norm": 0.05039289090482615,
+      "learning_rate": 0.0009865224352899118,
+      "loss": 1.2863,
+      "step": 1055
+    },
+    {
+      "epoch": 0.33483376766958856,
+      "grad_norm": 0.04252576093135151,
+      "learning_rate": 0.0009862026553557669,
+      "loss": 1.2244,
+      "step": 1060
+    },
+    {
+      "epoch": 0.33641317223406775,
+      "grad_norm": 0.040040635989773796,
+      "learning_rate": 0.000985879179145843,
+      "loss": 1.1958,
+      "step": 1065
+    },
+    {
+      "epoch": 0.33799257679854694,
+      "grad_norm": 0.05510903071317426,
+      "learning_rate": 0.0009855520091193158,
+      "loss": 1.1589,
+      "step": 1070
+    },
+    {
+      "epoch": 0.33957198136302613,
+      "grad_norm": 0.06148111145590916,
+      "learning_rate": 0.000985221147763441,
+      "loss": 1.199,
+      "step": 1075
+    },
+    {
+      "epoch": 0.3411513859275053,
+      "grad_norm": 0.06385524089857669,
+      "learning_rate": 0.000984886597593538,
+      "loss": 1.1694,
+      "step": 1080
+    },
+    {
+      "epoch": 0.3427307904919845,
+      "grad_norm": 0.06039272181936096,
+      "learning_rate": 0.0009845483611529693,
+      "loss": 1.153,
+      "step": 1085
+    },
+    {
+      "epoch": 0.3443101950564637,
+      "grad_norm": 0.057409878329639244,
+      "learning_rate": 0.0009842064410131221,
+      "loss": 1.2322,
+      "step": 1090
+    },
+    {
+      "epoch": 0.3458895996209429,
+      "grad_norm": 0.0385176344260921,
+      "learning_rate": 0.000983860839773388,
+      "loss": 1.1711,
+      "step": 1095
+    },
+    {
+      "epoch": 0.3474690041854221,
+      "grad_norm": 0.046177122204221564,
+      "learning_rate": 0.0009835115600611434,
+      "loss": 1.1927,
+      "step": 1100
+    },
+    {
+      "epoch": 0.3490484087499013,
+      "grad_norm": 0.03923161192034386,
+      "learning_rate": 0.00098315860453173,
+      "loss": 1.1267,
+      "step": 1105
+    },
+    {
+      "epoch": 0.3506278133143805,
+      "grad_norm": 0.0711060541707956,
+      "learning_rate": 0.0009828019758684342,
+      "loss": 1.188,
+      "step": 1110
+    },
+    {
+      "epoch": 0.35220721787885967,
+      "grad_norm": 0.043558270735732964,
+      "learning_rate": 0.000982441676782467,
+      "loss": 1.1585,
+      "step": 1115
+    },
+    {
+      "epoch": 0.35378662244333886,
+      "grad_norm": 0.050395482724650706,
+      "learning_rate": 0.0009820777100129428,
+      "loss": 1.1803,
+      "step": 1120
+    },
+    {
+      "epoch": 0.35536602700781805,
+      "grad_norm": 0.04592194649018605,
+      "learning_rate": 0.0009817100783268591,
+      "loss": 1.1806,
+      "step": 1125
+    },
+    {
+      "epoch": 0.35694543157229724,
+      "grad_norm": 0.03797561593571615,
+      "learning_rate": 0.0009813387845190756,
+      "loss": 1.1736,
+      "step": 1130
+    },
+    {
+      "epoch": 0.35852483613677644,
+      "grad_norm": 0.05497891905537719,
+      "learning_rate": 0.0009809638314122922,
+      "loss": 1.1786,
+      "step": 1135
+    },
+    {
+      "epoch": 0.36010424070125563,
+      "grad_norm": 0.06429227680603777,
+      "learning_rate": 0.0009805852218570284,
+      "loss": 1.1582,
+      "step": 1140
+    },
+    {
+      "epoch": 0.3616836452657348,
+      "grad_norm": 0.07553160414558219,
+      "learning_rate": 0.000980202958731601,
+      "loss": 1.1873,
+      "step": 1145
+    },
+    {
+      "epoch": 0.363263049830214,
+      "grad_norm": 0.044060112680621384,
+      "learning_rate": 0.0009798170449421028,
+      "loss": 1.0936,
+      "step": 1150
+    },
+    {
+      "epoch": 0.3648424543946932,
+      "grad_norm": 0.03854689953399917,
+      "learning_rate": 0.0009794274834223798,
+      "loss": 1.1089,
+      "step": 1155
+    },
+    {
+      "epoch": 0.3664218589591724,
+      "grad_norm": 0.057582233846948015,
+      "learning_rate": 0.0009790342771340095,
+      "loss": 1.2381,
+      "step": 1160
+    },
+    {
+      "epoch": 0.3680012635236516,
+      "grad_norm": 0.042972058986148115,
+      "learning_rate": 0.000978637429066278,
+      "loss": 1.1241,
+      "step": 1165
+    },
+    {
+      "epoch": 0.3695806680881308,
+      "grad_norm": 0.04384959742998597,
+      "learning_rate": 0.0009782369422361575,
+      "loss": 1.1868,
+      "step": 1170
+    },
+    {
+      "epoch": 0.37116007265261,
+      "grad_norm": 0.03743363050153524,
+      "learning_rate": 0.0009778328196882835,
+      "loss": 1.1248,
+      "step": 1175
+    },
+    {
+      "epoch": 0.37273947721708917,
+      "grad_norm": 0.05852710309772355,
+      "learning_rate": 0.000977425064494931,
+      "loss": 1.2375,
+      "step": 1180
+    },
+    {
+      "epoch": 0.37431888178156836,
+      "grad_norm": 0.04176855553749399,
+      "learning_rate": 0.000977013679755992,
+      "loss": 1.1759,
+      "step": 1185
+    },
+    {
+      "epoch": 0.37589828634604755,
+      "grad_norm": 0.03504530283243545,
+      "learning_rate": 0.0009765986685989513,
+      "loss": 1.1413,
+      "step": 1190
+    },
+    {
+      "epoch": 0.37747769091052674,
+      "grad_norm": 0.044377355127390584,
+      "learning_rate": 0.0009761800341788632,
+      "loss": 1.1988,
+      "step": 1195
+    },
+    {
+      "epoch": 0.37905709547500593,
+      "grad_norm": 0.07278323758231763,
+      "learning_rate": 0.0009757577796783267,
+      "loss": 1.2166,
+      "step": 1200
+    },
+    {
+      "epoch": 0.3806365000394851,
+      "grad_norm": 0.03741504859286431,
+      "learning_rate": 0.0009753319083074625,
+      "loss": 1.16,
+      "step": 1205
+    },
+    {
+      "epoch": 0.3822159046039643,
+      "grad_norm": 0.043790677416908085,
+      "learning_rate": 0.0009749024233038876,
+      "loss": 1.1204,
+      "step": 1210
+    },
+    {
+      "epoch": 0.3837953091684435,
+      "grad_norm": 0.04965181579324521,
+      "learning_rate": 0.0009744693279326914,
+      "loss": 1.2009,
+      "step": 1215
+    },
+    {
+      "epoch": 0.3853747137329227,
+      "grad_norm": 0.03844420659862734,
+      "learning_rate": 0.00097403262548641,
+      "loss": 1.1129,
+      "step": 1220
+    },
+    {
+      "epoch": 0.3869541182974019,
+      "grad_norm": 0.04020927397828586,
+      "learning_rate": 0.000973592319285002,
+      "loss": 1.2122,
+      "step": 1225
+    },
+    {
+      "epoch": 0.3885335228618811,
+      "grad_norm": 0.03880487766984948,
+      "learning_rate": 0.0009731484126758229,
+      "loss": 1.1671,
+      "step": 1230
+    },
+    {
+      "epoch": 0.3901129274263603,
+      "grad_norm": 12.282611028819801,
+      "learning_rate": 0.0009727009090336001,
+      "loss": 1.2351,
+      "step": 1235
+    },
+    {
+      "epoch": 0.39169233199083947,
+      "grad_norm": 0.09915021744614969,
+      "learning_rate": 0.000972249811760406,
+      "loss": 1.1555,
+      "step": 1240
+    },
+    {
+      "epoch": 0.39327173655531866,
+      "grad_norm": 0.0788287042616388,
+      "learning_rate": 0.0009717951242856338,
+      "loss": 1.1723,
+      "step": 1245
+    },
+    {
+      "epoch": 0.39485114111979785,
+      "grad_norm": 0.05548474879341339,
+      "learning_rate": 0.00097133685006597,
+      "loss": 1.2439,
+      "step": 1250
+    },
+    {
+      "epoch": 0.39643054568427705,
+      "grad_norm": 0.05120713365037505,
+      "learning_rate": 0.0009708749925853695,
+      "loss": 1.1581,
+      "step": 1255
+    },
+    {
+      "epoch": 0.39800995024875624,
+      "grad_norm": 0.06404293801939767,
+      "learning_rate": 0.0009704095553550276,
+      "loss": 1.1447,
+      "step": 1260
+    },
+    {
+      "epoch": 0.39958935481323543,
+      "grad_norm": 0.06312319298754332,
+      "learning_rate": 0.0009699405419133542,
+      "loss": 1.1659,
+      "step": 1265
+    },
+    {
+      "epoch": 0.4011687593777146,
+      "grad_norm": 0.13502437347863908,
+      "learning_rate": 0.0009694679558259472,
+      "loss": 1.2976,
+      "step": 1270
+    },
+    {
+      "epoch": 0.4027481639421938,
+      "grad_norm": 0.05679455947202985,
+      "learning_rate": 0.0009689918006855645,
+      "loss": 1.1992,
+      "step": 1275
+    },
+    {
+      "epoch": 0.404327568506673,
+      "grad_norm": 0.042285556377311476,
+      "learning_rate": 0.0009685120801120974,
+      "loss": 1.1559,
+      "step": 1280
+    },
+    {
+      "epoch": 0.4059069730711522,
+      "grad_norm": 0.039616964902070693,
+      "learning_rate": 0.0009680287977525426,
+      "loss": 1.2237,
+      "step": 1285
+    },
+    {
+      "epoch": 0.4074863776356314,
+      "grad_norm": 0.04579132630925091,
+      "learning_rate": 0.0009675419572809748,
+      "loss": 1.1573,
+      "step": 1290
+    },
+    {
+      "epoch": 0.4090657822001106,
+      "grad_norm": 0.04404265464861563,
+      "learning_rate": 0.0009670515623985187,
+      "loss": 1.139,
+      "step": 1295
+    },
+    {
+      "epoch": 0.4106451867645898,
+      "grad_norm": 0.04072321352559835,
+      "learning_rate": 0.000966557616833321,
+      "loss": 1.1837,
+      "step": 1300
+    },
+    {
+      "epoch": 0.41222459132906897,
+      "grad_norm": 0.050235769680070874,
+      "learning_rate": 0.0009660601243405214,
+      "loss": 1.1782,
+      "step": 1305
+    },
+    {
+      "epoch": 0.41380399589354816,
+      "grad_norm": 0.033559753256380775,
+      "learning_rate": 0.000965559088702225,
+      "loss": 1.1306,
+      "step": 1310
+    },
+    {
+      "epoch": 0.4153834004580273,
+      "grad_norm": 0.04229103700252006,
+      "learning_rate": 0.0009650545137274727,
+      "loss": 1.1629,
+      "step": 1315
+    },
+    {
+      "epoch": 0.4169628050225065,
+      "grad_norm": 0.035322329411059644,
+      "learning_rate": 0.000964546403252213,
+      "loss": 1.1241,
+      "step": 1320
+    },
+    {
+      "epoch": 0.4185422095869857,
+      "grad_norm": 0.03465845928810666,
+      "learning_rate": 0.0009640347611392722,
+      "loss": 1.1461,
+      "step": 1325
+    },
+    {
+      "epoch": 0.4201216141514649,
+      "grad_norm": 0.04057442742798969,
+      "learning_rate": 0.0009635195912783254,
+      "loss": 1.1491,
+      "step": 1330
+    },
+    {
+      "epoch": 0.42170101871594406,
+      "grad_norm": 0.03684991337983024,
+      "learning_rate": 0.0009630008975858666,
+      "loss": 1.1161,
+      "step": 1335
+    },
+    {
+      "epoch": 0.42328042328042326,
+      "grad_norm": 0.03575593243336909,
+      "learning_rate": 0.0009624786840051798,
+      "loss": 1.1021,
+      "step": 1340
+    },
+    {
+      "epoch": 0.42485982784490245,
+      "grad_norm": 0.038199337694147205,
+      "learning_rate": 0.0009619529545063075,
+      "loss": 1.1164,
+      "step": 1345
+    },
+    {
+      "epoch": 0.42643923240938164,
+      "grad_norm": 0.041815107832999215,
+      "learning_rate": 0.000961423713086022,
+      "loss": 1.194,
+      "step": 1350
+    },
+    {
+      "epoch": 0.42801863697386083,
+      "grad_norm": 0.04498779180491189,
+      "learning_rate": 0.000960890963767794,
+      "loss": 1.1468,
+      "step": 1355
+    },
+    {
+      "epoch": 0.42959804153834,
+      "grad_norm": 0.036156965395200406,
+      "learning_rate": 0.0009603547106017629,
+      "loss": 1.2044,
+      "step": 1360
+    },
+    {
+      "epoch": 0.4311774461028192,
+      "grad_norm": 0.0432898856802058,
+      "learning_rate": 0.0009598149576647053,
+      "loss": 1.1368,
+      "step": 1365
+    },
+    {
+      "epoch": 0.4327568506672984,
+      "grad_norm": 0.034197240048107265,
+      "learning_rate": 0.0009592717090600039,
+      "loss": 1.1465,
+      "step": 1370
+    },
+    {
+      "epoch": 0.4343362552317776,
+      "grad_norm": 0.06043967223719665,
+      "learning_rate": 0.0009587249689176171,
+      "loss": 1.2052,
+      "step": 1375
+    },
+    {
+      "epoch": 0.4359156597962568,
+      "grad_norm": 0.06952056078208815,
+      "learning_rate": 0.0009581747413940472,
+      "loss": 1.1896,
+      "step": 1380
+    },
+    {
+      "epoch": 0.437495064360736,
+      "grad_norm": 0.05316147215022715,
+      "learning_rate": 0.000957621030672308,
+      "loss": 1.1591,
+      "step": 1385
+    },
+    {
+      "epoch": 0.4390744689252152,
+      "grad_norm": 0.04617490936576988,
+      "learning_rate": 0.0009570638409618946,
+      "loss": 1.18,
+      "step": 1390
+    },
+    {
+      "epoch": 0.44065387348969437,
+      "grad_norm": 0.04953043812653844,
+      "learning_rate": 0.0009565031764987502,
+      "loss": 1.1999,
+      "step": 1395
+    },
+    {
+      "epoch": 0.44223327805417356,
+      "grad_norm": 0.06773718629645895,
+      "learning_rate": 0.000955939041545234,
+      "loss": 1.1981,
+      "step": 1400
+    },
+    {
+      "epoch": 0.44381268261865275,
+      "grad_norm": 0.05915998407509928,
+      "learning_rate": 0.0009553714403900897,
+      "loss": 1.1665,
+      "step": 1405
+    },
+    {
+      "epoch": 0.44539208718313195,
+      "grad_norm": 0.05814230542199274,
+      "learning_rate": 0.0009548003773484114,
+      "loss": 1.1592,
+      "step": 1410
+    },
+    {
+      "epoch": 0.44697149174761114,
+      "grad_norm": 0.050530877930300165,
+      "learning_rate": 0.0009542258567616122,
+      "loss": 1.1392,
+      "step": 1415
+    },
+    {
+      "epoch": 0.44855089631209033,
+      "grad_norm": 0.0347879234970523,
+      "learning_rate": 0.0009536478829973902,
+      "loss": 1.1932,
+      "step": 1420
+    },
+    {
+      "epoch": 0.4501303008765695,
+      "grad_norm": 0.03401528298444466,
+      "learning_rate": 0.0009530664604496964,
+      "loss": 1.2068,
+      "step": 1425
+    },
+    {
+      "epoch": 0.4517097054410487,
+      "grad_norm": 0.03592298599245248,
+      "learning_rate": 0.0009524815935386997,
+      "loss": 1.1095,
+      "step": 1430
+    },
+    {
+      "epoch": 0.4532891100055279,
+      "grad_norm": 0.03974587205442548,
+      "learning_rate": 0.0009518932867107551,
+      "loss": 1.1696,
+      "step": 1435
+    },
+    {
+      "epoch": 0.4548685145700071,
+      "grad_norm": 0.03632280838981324,
+      "learning_rate": 0.0009513015444383682,
+      "loss": 1.0858,
+      "step": 1440
+    },
+    {
+      "epoch": 0.4564479191344863,
+      "grad_norm": 0.06649418475772793,
+      "learning_rate": 0.0009507063712201623,
+      "loss": 1.2024,
+      "step": 1445
+    },
+    {
+      "epoch": 0.4580273236989655,
+      "grad_norm": 0.03796203835081221,
+      "learning_rate": 0.0009501077715808444,
+      "loss": 1.1621,
+      "step": 1450
+    },
+    {
+      "epoch": 0.4596067282634447,
+      "grad_norm": 0.03697987143090938,
+      "learning_rate": 0.0009495057500711697,
+      "loss": 1.1507,
+      "step": 1455
+    },
+    {
+      "epoch": 0.46118613282792387,
+      "grad_norm": 0.060201011824411864,
+      "learning_rate": 0.0009489003112679075,
+      "loss": 1.1274,
+      "step": 1460
+    },
+    {
+      "epoch": 0.46276553739240306,
+      "grad_norm": 0.06116106558440598,
+      "learning_rate": 0.0009482914597738072,
+      "loss": 1.1842,
+      "step": 1465
+    },
+    {
+      "epoch": 0.46434494195688225,
+      "grad_norm": 0.036769642166062266,
+      "learning_rate": 0.000947679200217562,
+      "loss": 1.2194,
+      "step": 1470
+    },
+    {
+      "epoch": 0.46592434652136144,
+      "grad_norm": 0.04080679654037661,
+      "learning_rate": 0.0009470635372537748,
+      "loss": 1.1075,
+      "step": 1475
+    },
+    {
+      "epoch": 0.46750375108584064,
+      "grad_norm": 0.04134620986690867,
+      "learning_rate": 0.0009464444755629216,
+      "loss": 1.1599,
+      "step": 1480
+    },
+    {
+      "epoch": 0.4690831556503198,
+      "grad_norm": 0.04951151349439519,
+      "learning_rate": 0.0009458220198513177,
+      "loss": 1.0983,
+      "step": 1485
+    },
+    {
+      "epoch": 0.470662560214799,
+      "grad_norm": 0.036583714485988426,
+      "learning_rate": 0.00094519617485108,
+      "loss": 1.1358,
+      "step": 1490
+    },
+    {
+      "epoch": 0.4722419647792782,
+      "grad_norm": 0.03826169128595341,
+      "learning_rate": 0.0009445669453200923,
+      "loss": 1.1244,
+      "step": 1495
+    },
+    {
+      "epoch": 0.4738213693437574,
+      "grad_norm": 0.038232781887304446,
+      "learning_rate": 0.0009439343360419688,
+      "loss": 1.1603,
+      "step": 1500
+    },
+    {
+      "epoch": 0.4754007739082366,
+      "grad_norm": 0.039105164779814766,
+      "learning_rate": 0.0009432983518260174,
+      "loss": 1.1746,
+      "step": 1505
+    },
+    {
+      "epoch": 0.4769801784727158,
+      "grad_norm": 0.042248369457699976,
+      "learning_rate": 0.0009426589975072039,
+      "loss": 1.0923,
+      "step": 1510
+    },
+    {
+      "epoch": 0.478559583037195,
+      "grad_norm": 0.034606195108217934,
+      "learning_rate": 0.0009420162779461141,
+      "loss": 1.1434,
+      "step": 1515
+    },
+    {
+      "epoch": 0.48013898760167417,
+      "grad_norm": 0.05210811886077802,
+      "learning_rate": 0.000941370198028918,
+      "loss": 1.2034,
+      "step": 1520
+    },
+    {
+      "epoch": 0.48171839216615336,
+      "grad_norm": 0.042216154872195315,
+      "learning_rate": 0.0009407207626673319,
+      "loss": 1.2095,
+      "step": 1525
+    },
+    {
+      "epoch": 0.48329779673063256,
+      "grad_norm": 0.04064731158462736,
+      "learning_rate": 0.0009400679767985813,
+      "loss": 1.1016,
+      "step": 1530
+    },
+    {
+      "epoch": 0.48487720129511175,
+      "grad_norm": 0.0600499327895473,
+      "learning_rate": 0.000939411845385364,
+      "loss": 1.158,
+      "step": 1535
+    },
+    {
+      "epoch": 0.48645660585959094,
+      "grad_norm": 0.03423119648920456,
+      "learning_rate": 0.0009387523734158106,
+      "loss": 1.1099,
+      "step": 1540
+    },
+    {
+      "epoch": 0.48803601042407013,
+      "grad_norm": 0.050478652855600346,
+      "learning_rate": 0.0009380895659034485,
+      "loss": 1.1357,
+      "step": 1545
+    },
+    {
+      "epoch": 0.4896154149885493,
+      "grad_norm": 0.04697965402996422,
+      "learning_rate": 0.0009374234278871631,
+      "loss": 1.1607,
+      "step": 1550
+    },
+    {
+      "epoch": 0.4911948195530285,
+      "grad_norm": 0.03542010437099392,
+      "learning_rate": 0.0009367539644311591,
+      "loss": 1.0903,
+      "step": 1555
+    },
+    {
+      "epoch": 0.4927742241175077,
+      "grad_norm": 0.03747261262586468,
+      "learning_rate": 0.0009360811806249223,
+      "loss": 1.1855,
+      "step": 1560
+    },
+    {
+      "epoch": 0.4943536286819869,
+      "grad_norm": 0.0342672802591948,
+      "learning_rate": 0.0009354050815831811,
+      "loss": 1.1237,
+      "step": 1565
+    },
+    {
+      "epoch": 0.4959330332464661,
+      "grad_norm": 0.04341908110834359,
+      "learning_rate": 0.0009347256724458674,
+      "loss": 1.1838,
+      "step": 1570
+    },
+    {
+      "epoch": 0.4975124378109453,
+      "grad_norm": 0.035326570663710144,
+      "learning_rate": 0.0009340429583780774,
+      "loss": 1.1988,
+      "step": 1575
+    },
+    {
+      "epoch": 0.4990918423754245,
+      "grad_norm": 0.039931525021498104,
+      "learning_rate": 0.0009333569445700326,
+      "loss": 1.1413,
+      "step": 1580
+    },
+    {
+      "epoch": 0.5006712469399036,
+      "grad_norm": 0.04648089686611222,
+      "learning_rate": 0.0009326676362370404,
+      "loss": 1.1775,
+      "step": 1585
+    },
+    {
+      "epoch": 0.5022506515043829,
+      "grad_norm": 0.03358345656864849,
+      "learning_rate": 0.0009319750386194537,
+      "loss": 1.0934,
+      "step": 1590
+    },
+    {
+      "epoch": 0.503830056068862,
+      "grad_norm": 0.04937018263420283,
+      "learning_rate": 0.0009312791569826324,
+      "loss": 1.2194,
+      "step": 1595
+    },
+    {
+      "epoch": 0.5054094606333412,
+      "grad_norm": 0.03677684834011018,
+      "learning_rate": 0.0009305799966169022,
+      "loss": 1.0755,
+      "step": 1600
+    },
+    {
+      "epoch": 0.5069888651978204,
+      "grad_norm": 0.0485094805968303,
+      "learning_rate": 0.000929877562837515,
+      "loss": 1.158,
+      "step": 1605
+    },
+    {
+      "epoch": 0.5085682697622996,
+      "grad_norm": 0.0444791474985611,
+      "learning_rate": 0.0009291718609846081,
+      "loss": 1.1302,
+      "step": 1610
+    },
+    {
+      "epoch": 0.5101476743267788,
+      "grad_norm": 0.04117121605197338,
+      "learning_rate": 0.0009284628964231635,
+      "loss": 1.1024,
+      "step": 1615
+    },
+    {
+      "epoch": 0.511727078891258,
+      "grad_norm": 0.04900515775546535,
+      "learning_rate": 0.0009277506745429682,
+      "loss": 1.1087,
+      "step": 1620
+    },
+    {
+      "epoch": 0.5133064834557372,
+      "grad_norm": 0.05633245648361274,
+      "learning_rate": 0.0009270352007585719,
+      "loss": 1.2387,
+      "step": 1625
+    },
+    {
+      "epoch": 0.5148858880202164,
+      "grad_norm": 0.059059357799405465,
+      "learning_rate": 0.000926316480509246,
+      "loss": 1.223,
+      "step": 1630
+    },
+    {
+      "epoch": 0.5164652925846955,
+      "grad_norm": 0.042565468524679115,
+      "learning_rate": 0.0009255945192589439,
+      "loss": 1.1179,
+      "step": 1635
+    },
+    {
+      "epoch": 0.5180446971491748,
+      "grad_norm": 0.041144265259162496,
+      "learning_rate": 0.0009248693224962567,
+      "loss": 1.239,
+      "step": 1640
+    },
+    {
+      "epoch": 0.5196241017136539,
+      "grad_norm": 0.03525706812621692,
+      "learning_rate": 0.0009241408957343739,
+      "loss": 1.2348,
+      "step": 1645
+    },
+    {
+      "epoch": 0.5212035062781332,
+      "grad_norm": 0.046457052327021985,
+      "learning_rate": 0.00092340924451104,
+      "loss": 1.1048,
+      "step": 1650
+    },
+    {
+      "epoch": 0.5227829108426123,
+      "grad_norm": 0.047403099964377,
+      "learning_rate": 0.0009226743743885134,
+      "loss": 1.1228,
+      "step": 1655
+    },
+    {
+      "epoch": 0.5243623154070916,
+      "grad_norm": 0.03485309471044975,
+      "learning_rate": 0.0009219362909535234,
+      "loss": 1.1687,
+      "step": 1660
+    },
+    {
+      "epoch": 0.5259417199715707,
+      "grad_norm": 0.03368534111902494,
+      "learning_rate": 0.0009211949998172279,
+      "loss": 1.1499,
+      "step": 1665
+    },
+    {
+      "epoch": 0.5275211245360499,
+      "grad_norm": 0.03876304390267727,
+      "learning_rate": 0.0009204505066151709,
+      "loss": 1.1442,
+      "step": 1670
+    },
+    {
+      "epoch": 0.5291005291005291,
+      "grad_norm": 0.03718587036595309,
+      "learning_rate": 0.0009197028170072397,
+      "loss": 1.0995,
+      "step": 1675
+    },
+    {
+      "epoch": 0.5306799336650083,
+      "grad_norm": 0.04248209908652153,
+      "learning_rate": 0.0009189519366776217,
+      "loss": 1.218,
+      "step": 1680
+    },
+    {
+      "epoch": 0.5322593382294875,
+      "grad_norm": 0.040832723625780394,
+      "learning_rate": 0.0009181978713347613,
+      "loss": 1.1446,
+      "step": 1685
+    },
+    {
+      "epoch": 0.5338387427939667,
+      "grad_norm": 0.04706059530054778,
+      "learning_rate": 0.000917440626711316,
+      "loss": 1.1481,
+      "step": 1690
+    },
+    {
+      "epoch": 0.5354181473584458,
+      "grad_norm": 0.039309572380869684,
+      "learning_rate": 0.0009166802085641139,
+      "loss": 1.1471,
+      "step": 1695
+    },
+    {
+      "epoch": 0.5369975519229251,
+      "grad_norm": 0.03645680498086478,
+      "learning_rate": 0.0009159166226741088,
+      "loss": 1.2631,
+      "step": 1700
+    },
+    {
+      "epoch": 0.5385769564874042,
+      "grad_norm": 0.03650813201144362,
+      "learning_rate": 0.000915149874846337,
+      "loss": 1.0914,
+      "step": 1705
+    },
+    {
+      "epoch": 0.5401563610518835,
+      "grad_norm": 0.1694935189698132,
+      "learning_rate": 0.0009143799709098728,
+      "loss": 1.161,
+      "step": 1710
+    },
+    {
+      "epoch": 0.5417357656163626,
+      "grad_norm": 0.033172052345546904,
+      "learning_rate": 0.0009136069167177844,
+      "loss": 1.0754,
+      "step": 1715
+    },
+    {
+      "epoch": 0.5433151701808419,
+      "grad_norm": 0.04336773903840129,
+      "learning_rate": 0.0009128307181470893,
+      "loss": 1.1658,
+      "step": 1720
+    },
+    {
+      "epoch": 0.544894574745321,
+      "grad_norm": 0.04377461369596686,
+      "learning_rate": 0.0009120513810987094,
+      "loss": 1.1052,
+      "step": 1725
+    },
+    {
+      "epoch": 0.5464739793098002,
+      "grad_norm": 0.041066160659134626,
+      "learning_rate": 0.0009112689114974266,
+      "loss": 1.2059,
+      "step": 1730
+    },
+    {
+      "epoch": 0.5480533838742794,
+      "grad_norm": 0.04178208265398977,
+      "learning_rate": 0.0009104833152918375,
+      "loss": 1.1699,
+      "step": 1735
+    },
+    {
+      "epoch": 0.5496327884387586,
+      "grad_norm": 0.11328835205617309,
+      "learning_rate": 0.0009096945984543081,
+      "loss": 1.1136,
+      "step": 1740
+    },
+    {
+      "epoch": 0.5512121930032378,
+      "grad_norm": 0.19182622324811044,
+      "learning_rate": 0.0009089027669809285,
+      "loss": 1.2703,
+      "step": 1745
+    },
+    {
+      "epoch": 0.552791597567717,
+      "grad_norm": 0.059773779341214854,
+      "learning_rate": 0.0009081078268914673,
+      "loss": 1.2062,
+      "step": 1750
+    },
+    {
+      "epoch": 0.5543710021321961,
+      "grad_norm": 0.15234917380906907,
+      "learning_rate": 0.000907309784229326,
+      "loss": 1.2287,
+      "step": 1755
+    },
+    {
+      "epoch": 0.5559504066966754,
+      "grad_norm": 0.06586639991112524,
+      "learning_rate": 0.0009065086450614928,
+      "loss": 1.1379,
+      "step": 1760
+    },
+    {
+      "epoch": 0.5575298112611545,
+      "grad_norm": 0.17904872602554633,
+      "learning_rate": 0.0009057044154784963,
+      "loss": 1.1576,
+      "step": 1765
+    },
+    {
+      "epoch": 0.5591092158256338,
+      "grad_norm": 0.03758826384872291,
+      "learning_rate": 0.0009048971015943599,
+      "loss": 1.1779,
+      "step": 1770
+    },
+    {
+      "epoch": 0.5606886203901129,
+      "grad_norm": 0.08338270784298804,
+      "learning_rate": 0.0009040867095465548,
+      "loss": 1.2273,
+      "step": 1775
+    },
+    {
+      "epoch": 0.5622680249545922,
+      "grad_norm": 0.03954640652450269,
+      "learning_rate": 0.0009032732454959533,
+      "loss": 1.1036,
+      "step": 1780
+    },
+    {
+      "epoch": 0.5638474295190713,
+      "grad_norm": 0.0438072209614571,
+      "learning_rate": 0.000902456715626782,
+      "loss": 1.2786,
+      "step": 1785
+    },
+    {
+      "epoch": 0.5654268340835505,
+      "grad_norm": 0.047019840479581396,
+      "learning_rate": 0.0009016371261465752,
+      "loss": 1.1721,
+      "step": 1790
+    },
+    {
+      "epoch": 0.5670062386480297,
+      "grad_norm": 0.06010554299082062,
+      "learning_rate": 0.0009008144832861272,
+      "loss": 1.1543,
+      "step": 1795
+    },
+    {
+      "epoch": 0.5685856432125089,
+      "grad_norm": 0.053420243571307535,
+      "learning_rate": 0.000899988793299445,
+      "loss": 1.1894,
+      "step": 1800
+    },
+    {
+      "epoch": 0.5701650477769881,
+      "grad_norm": 0.03523402052330801,
+      "learning_rate": 0.0008991600624637013,
+      "loss": 1.2492,
+      "step": 1805
+    },
+    {
+      "epoch": 0.5717444523414673,
+      "grad_norm": 0.03576456253696137,
+      "learning_rate": 0.0008983282970791858,
+      "loss": 1.2006,
+      "step": 1810
+    },
+    {
+      "epoch": 0.5733238569059464,
+      "grad_norm": 0.04271916038016165,
+      "learning_rate": 0.0008974935034692583,
+      "loss": 1.1345,
+      "step": 1815
+    },
+    {
+      "epoch": 0.5749032614704257,
+      "grad_norm": 0.06186038459636685,
+      "learning_rate": 0.0008966556879802998,
+      "loss": 1.1715,
+      "step": 1820
+    },
+    {
+      "epoch": 0.5764826660349048,
+      "grad_norm": 0.04007885821089377,
+      "learning_rate": 0.0008958148569816652,
+      "loss": 1.1471,
+      "step": 1825
+    },
+    {
+      "epoch": 0.5780620705993841,
+      "grad_norm": 0.06690675689465636,
+      "learning_rate": 0.0008949710168656337,
+      "loss": 1.1694,
+      "step": 1830
+    },
+    {
+      "epoch": 0.5796414751638632,
+      "grad_norm": 0.054483680948141516,
+      "learning_rate": 0.0008941241740473612,
+      "loss": 1.1302,
+      "step": 1835
+    },
+    {
+      "epoch": 0.5812208797283425,
+      "grad_norm": 0.05283051263253464,
+      "learning_rate": 0.0008932743349648312,
+      "loss": 1.1547,
+      "step": 1840
+    },
+    {
+      "epoch": 0.5828002842928216,
+      "grad_norm": 0.0351874645524128,
+      "learning_rate": 0.0008924215060788051,
+      "loss": 1.1842,
+      "step": 1845
+    },
+    {
+      "epoch": 0.5843796888573008,
+      "grad_norm": 0.03531292908646924,
+      "learning_rate": 0.000891565693872775,
+      "loss": 1.1414,
+      "step": 1850
+    },
+    {
+      "epoch": 0.58595909342178,
+      "grad_norm": 0.03252007276812236,
+      "learning_rate": 0.0008907069048529122,
+      "loss": 1.1998,
+      "step": 1855
+    },
+    {
+      "epoch": 0.5875384979862592,
+      "grad_norm": 0.048953905031448496,
+      "learning_rate": 0.000889845145548019,
+      "loss": 1.1882,
+      "step": 1860
+    },
+    {
+      "epoch": 0.5891179025507384,
+      "grad_norm": 0.03679662565463424,
+      "learning_rate": 0.000888980422509479,
+      "loss": 1.2016,
+      "step": 1865
+    },
+    {
+      "epoch": 0.5906973071152176,
+      "grad_norm": 0.0429312944873228,
+      "learning_rate": 0.0008881127423112072,
+      "loss": 1.231,
+      "step": 1870
+    },
+    {
+      "epoch": 0.5922767116796968,
+      "grad_norm": 0.05085200300369583,
+      "learning_rate": 0.0008872421115495995,
+      "loss": 1.1525,
+      "step": 1875
+    },
+    {
+      "epoch": 0.5938561162441759,
+      "grad_norm": 0.04143984691504309,
+      "learning_rate": 0.0008863685368434831,
+      "loss": 1.1573,
+      "step": 1880
+    },
+    {
+      "epoch": 0.5954355208086551,
+      "grad_norm": 0.054741188359976926,
+      "learning_rate": 0.0008854920248340662,
+      "loss": 1.1595,
+      "step": 1885
+    },
+    {
+      "epoch": 0.5970149253731343,
+      "grad_norm": 0.04157592670579889,
+      "learning_rate": 0.0008846125821848873,
+      "loss": 1.1485,
+      "step": 1890
+    },
+    {
+      "epoch": 0.5985943299376135,
+      "grad_norm": 0.05597913806097826,
+      "learning_rate": 0.0008837302155817647,
+      "loss": 1.1551,
+      "step": 1895
+    },
+    {
+      "epoch": 0.6001737345020927,
+      "grad_norm": 0.04003622147381468,
+      "learning_rate": 0.0008828449317327452,
+      "loss": 1.114,
+      "step": 1900
+    },
+    {
+      "epoch": 0.6017531390665719,
+      "grad_norm": 0.04430881613109523,
+      "learning_rate": 0.0008819567373680541,
+      "loss": 1.2302,
+      "step": 1905
+    },
+    {
+      "epoch": 0.603332543631051,
+      "grad_norm": 0.044159161743787353,
+      "learning_rate": 0.000881065639240043,
+      "loss": 1.1455,
+      "step": 1910
+    },
+    {
+      "epoch": 0.6049119481955303,
+      "grad_norm": 0.040333438068730916,
+      "learning_rate": 0.0008801716441231386,
+      "loss": 1.1751,
+      "step": 1915
+    },
+    {
+      "epoch": 0.6064913527600094,
+      "grad_norm": 0.037659445323191335,
+      "learning_rate": 0.0008792747588137924,
+      "loss": 1.1294,
+      "step": 1920
+    },
+    {
+      "epoch": 0.6080707573244887,
+      "grad_norm": 0.057367139158678554,
+      "learning_rate": 0.0008783749901304271,
+      "loss": 1.185,
+      "step": 1925
+    },
+    {
+      "epoch": 0.6096501618889678,
+      "grad_norm": 0.039046189159429465,
+      "learning_rate": 0.0008774723449133866,
+      "loss": 1.1998,
+      "step": 1930
+    },
+    {
+      "epoch": 0.6112295664534471,
+      "grad_norm": 0.038520766106106066,
+      "learning_rate": 0.0008765668300248823,
+      "loss": 1.1858,
+      "step": 1935
+    },
+    {
+      "epoch": 0.6128089710179262,
+      "grad_norm": 0.06371115248464544,
+      "learning_rate": 0.000875658452348943,
+      "loss": 1.2866,
+      "step": 1940
+    },
+    {
+      "epoch": 0.6143883755824054,
+      "grad_norm": 0.05805583054296386,
+      "learning_rate": 0.0008747472187913603,
+      "loss": 1.1246,
+      "step": 1945
+    },
+    {
+      "epoch": 0.6159677801468846,
+      "grad_norm": 0.0882727755464156,
+      "learning_rate": 0.0008738331362796375,
+      "loss": 1.1861,
+      "step": 1950
+    },
+    {
+      "epoch": 0.6175471847113638,
+      "grad_norm": 0.1291074294139289,
+      "learning_rate": 0.0008729162117629368,
+      "loss": 1.172,
+      "step": 1955
+    },
+    {
+      "epoch": 0.619126589275843,
+      "grad_norm": 0.06752168123783803,
+      "learning_rate": 0.0008719964522120261,
+      "loss": 1.1289,
+      "step": 1960
+    },
+    {
+      "epoch": 0.6207059938403222,
+      "grad_norm": 0.05230307801436827,
+      "learning_rate": 0.0008710738646192262,
+      "loss": 1.1345,
+      "step": 1965
+    },
+    {
+      "epoch": 0.6222853984048013,
+      "grad_norm": 0.05307584281041897,
+      "learning_rate": 0.0008701484559983577,
+      "loss": 1.1963,
+      "step": 1970
+    },
+    {
+      "epoch": 0.6238648029692806,
+      "grad_norm": 0.062313428830658914,
+      "learning_rate": 0.0008692202333846875,
+      "loss": 1.1784,
+      "step": 1975
+    },
+    {
+      "epoch": 0.6254442075337597,
+      "grad_norm": 0.07318456946963307,
+      "learning_rate": 0.0008682892038348756,
+      "loss": 1.1756,
+      "step": 1980
+    },
+    {
+      "epoch": 0.627023612098239,
+      "grad_norm": 0.05035910214366681,
+      "learning_rate": 0.0008673553744269207,
+      "loss": 1.1644,
+      "step": 1985
+    },
+    {
+      "epoch": 0.6286030166627181,
+      "grad_norm": 0.03930070288487179,
+      "learning_rate": 0.0008664187522601079,
+      "loss": 1.1885,
+      "step": 1990
+    },
+    {
+      "epoch": 0.6301824212271974,
+      "grad_norm": 0.04028703538445593,
+      "learning_rate": 0.0008654793444549531,
+      "loss": 1.1463,
+      "step": 1995
+    },
+    {
+      "epoch": 0.6317618257916765,
+      "grad_norm": 0.04340467125837745,
+      "learning_rate": 0.0008645371581531497,
+      "loss": 1.1571,
+      "step": 2000
+    },
+    {
+      "epoch": 0.6333412303561557,
+      "grad_norm": 0.03579329388775005,
+      "learning_rate": 0.0008635922005175143,
+      "loss": 1.1222,
+      "step": 2005
+    },
+    {
+      "epoch": 0.6349206349206349,
+      "grad_norm": 0.039633418005999524,
+      "learning_rate": 0.0008626444787319319,
+      "loss": 1.126,
+      "step": 2010
+    },
+    {
+      "epoch": 0.6365000394851141,
+      "grad_norm": 0.04656715388051395,
+      "learning_rate": 0.0008616940000013016,
+      "loss": 1.1351,
+      "step": 2015
+    },
+    {
+      "epoch": 0.6380794440495933,
+      "grad_norm": 0.04241567297447217,
+      "learning_rate": 0.0008607407715514819,
+      "loss": 1.1736,
+      "step": 2020
+    },
+    {
+      "epoch": 0.6396588486140725,
+      "grad_norm": 0.04272664455329046,
+      "learning_rate": 0.0008597848006292354,
+      "loss": 1.16,
+      "step": 2025
+    },
+    {
+      "epoch": 0.6412382531785517,
+      "grad_norm": 0.050634173141446026,
+      "learning_rate": 0.0008588260945021737,
+      "loss": 1.1535,
+      "step": 2030
+    },
+    {
+      "epoch": 0.6428176577430309,
+      "grad_norm": 0.043629941793955164,
+      "learning_rate": 0.0008578646604587028,
+      "loss": 1.1172,
+      "step": 2035
+    },
+    {
+      "epoch": 0.64439706230751,
+      "grad_norm": 0.037913467140448166,
+      "learning_rate": 0.0008569005058079671,
+      "loss": 1.1938,
+      "step": 2040
+    },
+    {
+      "epoch": 0.6459764668719893,
+      "grad_norm": 0.07347441864594863,
+      "learning_rate": 0.0008559336378797935,
+      "loss": 1.2555,
+      "step": 2045
+    },
+    {
+      "epoch": 0.6475558714364684,
+      "grad_norm": 0.04452704637473851,
+      "learning_rate": 0.0008549640640246367,
+      "loss": 1.2069,
+      "step": 2050
+    },
+    {
+      "epoch": 0.6491352760009477,
+      "grad_norm": 0.04051185517620845,
+      "learning_rate": 0.0008539917916135227,
+      "loss": 1.0992,
+      "step": 2055
+    },
+    {
+      "epoch": 0.6507146805654268,
+      "grad_norm": 0.03344259315550966,
+      "learning_rate": 0.0008530168280379924,
+      "loss": 1.2363,
+      "step": 2060
+    },
+    {
+      "epoch": 0.652294085129906,
+      "grad_norm": 0.051496549798025794,
+      "learning_rate": 0.0008520391807100465,
+      "loss": 1.1066,
+      "step": 2065
+    },
+    {
+      "epoch": 0.6538734896943852,
+      "grad_norm": 0.049313008139448104,
+      "learning_rate": 0.0008510588570620879,
+      "loss": 1.1435,
+      "step": 2070
+    },
+    {
+      "epoch": 0.6554528942588644,
+      "grad_norm": 0.036890992849123165,
+      "learning_rate": 0.0008500758645468661,
+      "loss": 1.178,
+      "step": 2075
+    },
+    {
+      "epoch": 0.6570322988233436,
+      "grad_norm": 0.05000485789172623,
+      "learning_rate": 0.0008490902106374202,
+      "loss": 1.1358,
+      "step": 2080
+    },
+    {
+      "epoch": 0.6586117033878228,
+      "grad_norm": 0.058476338194899814,
+      "learning_rate": 0.0008481019028270219,
+      "loss": 1.1535,
+      "step": 2085
+    },
+    {
+      "epoch": 0.660191107952302,
+      "grad_norm": 0.17384762993524389,
+      "learning_rate": 0.000847110948629119,
+      "loss": 1.1709,
+      "step": 2090
+    },
+    {
+      "epoch": 0.6617705125167812,
+      "grad_norm": 0.05756662582337642,
+      "learning_rate": 0.0008461173555772779,
+      "loss": 1.2323,
+      "step": 2095
+    },
+    {
+      "epoch": 0.6633499170812603,
+      "grad_norm": 0.04628423823864862,
+      "learning_rate": 0.0008451211312251266,
+      "loss": 1.1244,
+      "step": 2100
+    },
+    {
+      "epoch": 0.6649293216457396,
+      "grad_norm": 0.05753687558314109,
+      "learning_rate": 0.0008441222831462967,
+      "loss": 1.185,
+      "step": 2105
+    },
+    {
+      "epoch": 0.6665087262102187,
+      "grad_norm": 0.03541434435667187,
+      "learning_rate": 0.0008431208189343669,
+      "loss": 1.1652,
+      "step": 2110
+    },
+    {
+      "epoch": 0.668088130774698,
+      "grad_norm": 0.06365027743645021,
+      "learning_rate": 0.0008421167462028039,
+      "loss": 1.0968,
+      "step": 2115
+    },
+    {
+      "epoch": 0.6696675353391771,
+      "grad_norm": 0.15796614326365377,
+      "learning_rate": 0.000841110072584906,
+      "loss": 1.2558,
+      "step": 2120
+    },
+    {
+      "epoch": 0.6712469399036564,
+      "grad_norm": 0.04426269547365523,
+      "learning_rate": 0.0008401008057337437,
+      "loss": 1.236,
+      "step": 2125
+    },
+    {
+      "epoch": 0.6728263444681355,
+      "grad_norm": 0.0481957665679776,
+      "learning_rate": 0.0008390889533221025,
+      "loss": 1.221,
+      "step": 2130
+    },
+    {
+      "epoch": 0.6744057490326147,
+      "grad_norm": 0.0742711968724926,
+      "learning_rate": 0.0008380745230424238,
+      "loss": 1.1596,
+      "step": 2135
+    },
+    {
+      "epoch": 0.6759851535970939,
+      "grad_norm": 0.10209227348253318,
+      "learning_rate": 0.0008370575226067474,
+      "loss": 1.1409,
+      "step": 2140
+    },
+    {
+      "epoch": 0.6775645581615731,
+      "grad_norm": 0.03233639963159502,
+      "learning_rate": 0.0008360379597466518,
+      "loss": 1.0974,
+      "step": 2145
+    },
+    {
+      "epoch": 0.6791439627260523,
+      "grad_norm": 0.04112561163083512,
+      "learning_rate": 0.0008350158422131961,
+      "loss": 1.1423,
+      "step": 2150
+    },
+    {
+      "epoch": 0.6807233672905315,
+      "grad_norm": 0.04219217327869605,
+      "learning_rate": 0.0008339911777768609,
+      "loss": 1.0994,
+      "step": 2155
+    },
+    {
+      "epoch": 0.6823027718550106,
+      "grad_norm": 0.037513381823338576,
+      "learning_rate": 0.0008329639742274892,
+      "loss": 1.1342,
+      "step": 2160
+    },
+    {
+      "epoch": 0.6838821764194899,
+      "grad_norm": 0.03657974348037566,
+      "learning_rate": 0.0008319342393742268,
+      "loss": 1.1531,
+      "step": 2165
+    },
+    {
+      "epoch": 0.685461580983969,
+      "grad_norm": 0.05047343123602675,
+      "learning_rate": 0.0008309019810454643,
+      "loss": 1.0958,
+      "step": 2170
+    },
+    {
+      "epoch": 0.6870409855484483,
+      "grad_norm": 0.03243744614986179,
+      "learning_rate": 0.0008298672070887754,
+      "loss": 1.1256,
+      "step": 2175
+    },
+    {
+      "epoch": 0.6886203901129274,
+      "grad_norm": 0.044261034687072306,
+      "learning_rate": 0.0008288299253708595,
+      "loss": 1.0999,
+      "step": 2180
+    },
+    {
+      "epoch": 0.6901997946774067,
+      "grad_norm": 0.03619490329735276,
+      "learning_rate": 0.00082779014377748,
+      "loss": 1.1721,
+      "step": 2185
+    },
+    {
+      "epoch": 0.6917791992418858,
+      "grad_norm": 0.04146064866463831,
+      "learning_rate": 0.000826747870213406,
+      "loss": 1.1727,
+      "step": 2190
+    },
+    {
+      "epoch": 0.693358603806365,
+      "grad_norm": 0.04849711422712515,
+      "learning_rate": 0.0008257031126023504,
+      "loss": 1.1482,
+      "step": 2195
+    },
+    {
+      "epoch": 0.6949380083708442,
+      "grad_norm": 0.040296226435196425,
+      "learning_rate": 0.0008246558788869116,
+      "loss": 1.1527,
+      "step": 2200
+    },
+    {
+      "epoch": 0.6965174129353234,
+      "grad_norm": 0.03293749678341955,
+      "learning_rate": 0.0008236061770285119,
+      "loss": 1.1148,
+      "step": 2205
+    },
+    {
+      "epoch": 0.6980968174998026,
+      "grad_norm": 0.042233291320546755,
+      "learning_rate": 0.0008225540150073371,
+      "loss": 1.2206,
+      "step": 2210
+    },
+    {
+      "epoch": 0.6996762220642818,
+      "grad_norm": 0.03827298591812442,
+      "learning_rate": 0.0008214994008222758,
+      "loss": 1.1384,
+      "step": 2215
+    },
+    {
+      "epoch": 0.701255626628761,
+      "grad_norm": 0.045374923911993686,
+      "learning_rate": 0.0008204423424908591,
+      "loss": 1.1324,
+      "step": 2220
+    },
+    {
+      "epoch": 0.7028350311932402,
+      "grad_norm": 0.03238064122889688,
+      "learning_rate": 0.0008193828480491995,
+      "loss": 1.1544,
+      "step": 2225
+    },
+    {
+      "epoch": 0.7044144357577193,
+      "grad_norm": 0.04380634028979496,
+      "learning_rate": 0.000818320925551929,
+      "loss": 1.0635,
+      "step": 2230
+    },
+    {
+      "epoch": 0.7059938403221986,
+      "grad_norm": 0.0370632104302981,
+      "learning_rate": 0.000817256583072139,
+      "loss": 1.0649,
+      "step": 2235
+    },
+    {
+      "epoch": 0.7075732448866777,
+      "grad_norm": 0.031093840596031517,
+      "learning_rate": 0.0008161898287013184,
+      "loss": 1.1048,
+      "step": 2240
+    },
+    {
+      "epoch": 0.709152649451157,
+      "grad_norm": 0.03207987681813833,
+      "learning_rate": 0.0008151206705492919,
+      "loss": 1.1286,
+      "step": 2245
+    },
+    {
+      "epoch": 0.7107320540156361,
+      "grad_norm": 0.04164109172225732,
+      "learning_rate": 0.0008140491167441584,
+      "loss": 1.1099,
+      "step": 2250
+    },
+    {
+      "epoch": 0.7123114585801152,
+      "grad_norm": 0.03400416322772637,
+      "learning_rate": 0.0008129751754322299,
+      "loss": 1.1361,
+      "step": 2255
+    },
+    {
+      "epoch": 0.7138908631445945,
+      "grad_norm": 0.04592151794322222,
+      "learning_rate": 0.0008118988547779687,
+      "loss": 1.1323,
+      "step": 2260
+    },
+    {
+      "epoch": 0.7154702677090736,
+      "grad_norm": 0.036229621957410515,
+      "learning_rate": 0.0008108201629639256,
+      "loss": 1.1368,
+      "step": 2265
+    },
+    {
+      "epoch": 0.7170496722735529,
+      "grad_norm": 0.04151662660096531,
+      "learning_rate": 0.0008097391081906777,
+      "loss": 1.1346,
+      "step": 2270
+    },
+    {
+      "epoch": 0.718629076838032,
+      "grad_norm": 0.04873978029799004,
+      "learning_rate": 0.0008086556986767663,
+      "loss": 1.134,
+      "step": 2275
+    },
+    {
+      "epoch": 0.7202084814025113,
+      "grad_norm": 0.03333780638415451,
+      "learning_rate": 0.0008075699426586344,
+      "loss": 1.1413,
+      "step": 2280
+    },
+    {
+      "epoch": 0.7217878859669904,
+      "grad_norm": 0.034640199910440705,
+      "learning_rate": 0.0008064818483905634,
+      "loss": 1.1857,
+      "step": 2285
+    },
+    {
+      "epoch": 0.7233672905314696,
+      "grad_norm": 0.04818114443150759,
+      "learning_rate": 0.0008053914241446112,
+      "loss": 1.1447,
+      "step": 2290
+    },
+    {
+      "epoch": 0.7249466950959488,
+      "grad_norm": 0.036283227351149444,
+      "learning_rate": 0.000804298678210549,
+      "loss": 1.0211,
+      "step": 2295
+    },
+    {
+      "epoch": 0.726526099660428,
+      "grad_norm": 0.03357453215492695,
+      "learning_rate": 0.0008032036188957982,
+      "loss": 1.1105,
+      "step": 2300
+    },
+    {
+      "epoch": 0.7281055042249072,
+      "grad_norm": 0.034911998236058436,
+      "learning_rate": 0.0008021062545253672,
+      "loss": 1.1321,
+      "step": 2305
+    },
+    {
+      "epoch": 0.7296849087893864,
+      "grad_norm": 0.03072089247830198,
+      "learning_rate": 0.0008010065934417881,
+      "loss": 1.118,
+      "step": 2310
+    },
+    {
+      "epoch": 0.7312643133538655,
+      "grad_norm": 0.033945552237615126,
+      "learning_rate": 0.0007999046440050538,
+      "loss": 1.1327,
+      "step": 2315
+    },
+    {
+      "epoch": 0.7328437179183448,
+      "grad_norm": 0.036623240485144815,
+      "learning_rate": 0.0007988004145925538,
+      "loss": 1.1336,
+      "step": 2320
+    },
+    {
+      "epoch": 0.7344231224828239,
+      "grad_norm": 0.034765981851321515,
+      "learning_rate": 0.0007976939135990106,
+      "loss": 1.109,
+      "step": 2325
+    },
+    {
+      "epoch": 0.7360025270473032,
+      "grad_norm": 0.03112055116207592,
+      "learning_rate": 0.000796585149436416,
+      "loss": 1.0639,
+      "step": 2330
+    },
+    {
+      "epoch": 0.7375819316117823,
+      "grad_norm": 0.03787672321570991,
+      "learning_rate": 0.0007954741305339676,
+      "loss": 1.1191,
+      "step": 2335
+    },
+    {
+      "epoch": 0.7391613361762616,
+      "grad_norm": 0.05596768906351211,
+      "learning_rate": 0.000794360865338004,
+      "loss": 1.1129,
+      "step": 2340
+    },
+    {
+      "epoch": 0.7407407407407407,
+      "grad_norm": 0.04646968935035901,
+      "learning_rate": 0.0007932453623119407,
+      "loss": 1.1535,
+      "step": 2345
+    },
+    {
+      "epoch": 0.74232014530522,
+      "grad_norm": 0.03287872636040462,
+      "learning_rate": 0.0007921276299362062,
+      "loss": 1.0685,
+      "step": 2350
+    },
+    {
+      "epoch": 0.7438995498696991,
+      "grad_norm": 0.03761813831687904,
+      "learning_rate": 0.0007910076767081772,
+      "loss": 1.1851,
+      "step": 2355
+    },
+    {
+      "epoch": 0.7454789544341783,
+      "grad_norm": 0.04154317590568981,
+      "learning_rate": 0.0007898855111421139,
+      "loss": 1.0641,
+      "step": 2360
+    },
+    {
+      "epoch": 0.7470583589986575,
+      "grad_norm": 0.03225872527047046,
+      "learning_rate": 0.0007887611417690958,
+      "loss": 1.0459,
+      "step": 2365
+    },
+    {
+      "epoch": 0.7486377635631367,
+      "grad_norm": 0.07034127382923797,
+      "learning_rate": 0.0007876345771369564,
+      "loss": 1.301,
+      "step": 2370
+    },
+    {
+      "epoch": 0.7502171681276159,
+      "grad_norm": 0.04449126769031474,
+      "learning_rate": 0.0007865058258102177,
+      "loss": 1.1004,
+      "step": 2375
+    },
+    {
+      "epoch": 0.7517965726920951,
+      "grad_norm": 0.0350145043034834,
+      "learning_rate": 0.0007853748963700264,
+      "loss": 1.1672,
+      "step": 2380
+    },
+    {
+      "epoch": 0.7533759772565742,
+      "grad_norm": 0.03748553636786378,
+      "learning_rate": 0.0007842417974140879,
+      "loss": 1.1193,
+      "step": 2385
+    },
+    {
+      "epoch": 0.7549553818210535,
+      "grad_norm": 0.030937914052851834,
+      "learning_rate": 0.0007831065375566004,
+      "loss": 1.0449,
+      "step": 2390
+    },
+    {
+      "epoch": 0.7565347863855326,
+      "grad_norm": 0.034083331472644675,
+      "learning_rate": 0.0007819691254281905,
+      "loss": 1.0966,
+      "step": 2395
+    },
+    {
+      "epoch": 0.7581141909500119,
+      "grad_norm": 0.03047192172455573,
+      "learning_rate": 0.0007808295696758472,
+      "loss": 1.0874,
+      "step": 2400
+    },
+    {
+      "epoch": 0.759693595514491,
+      "grad_norm": 0.03554221631207444,
+      "learning_rate": 0.0007796878789628555,
+      "loss": 1.0696,
+      "step": 2405
+    },
+    {
+      "epoch": 0.7612730000789703,
+      "grad_norm": 0.03538164542716929,
+      "learning_rate": 0.0007785440619687316,
+      "loss": 1.1272,
+      "step": 2410
+    },
+    {
+      "epoch": 0.7628524046434494,
+      "grad_norm": 0.036853427979659625,
+      "learning_rate": 0.0007773981273891562,
+      "loss": 1.1094,
+      "step": 2415
+    },
+    {
+      "epoch": 0.7644318092079286,
+      "grad_norm": 0.04257817482887054,
+      "learning_rate": 0.0007762500839359084,
+      "loss": 1.1402,
+      "step": 2420
+    },
+    {
+      "epoch": 0.7660112137724078,
+      "grad_norm": 0.044417776153948235,
+      "learning_rate": 0.0007750999403368001,
+      "loss": 1.1064,
+      "step": 2425
+    },
+    {
+      "epoch": 0.767590618336887,
+      "grad_norm": 0.04767857151182902,
+      "learning_rate": 0.000773947705335609,
+      "loss": 1.0984,
+      "step": 2430
+    },
+    {
+      "epoch": 0.7691700229013662,
+      "grad_norm": 0.04717634798528141,
+      "learning_rate": 0.0007727933876920121,
+      "loss": 1.1129,
+      "step": 2435
+    },
+    {
+      "epoch": 0.7707494274658454,
+      "grad_norm": 0.03626272630855877,
+      "learning_rate": 0.0007716369961815199,
+      "loss": 1.1507,
+      "step": 2440
+    },
+    {
+      "epoch": 0.7723288320303245,
+      "grad_norm": 0.03522532598234683,
+      "learning_rate": 0.0007704785395954085,
+      "loss": 1.0645,
+      "step": 2445
+    },
+    {
+      "epoch": 0.7739082365948038,
+      "grad_norm": 0.03559706129851639,
+      "learning_rate": 0.0007693180267406539,
+      "loss": 1.1182,
+      "step": 2450
+    },
+    {
+      "epoch": 0.7754876411592829,
+      "grad_norm": 0.06678349957404967,
+      "learning_rate": 0.000768155466439864,
+      "loss": 1.1167,
+      "step": 2455
+    },
+    {
+      "epoch": 0.7770670457237622,
+      "grad_norm": 0.04029787951031313,
+      "learning_rate": 0.0007669908675312128,
+      "loss": 1.1288,
+      "step": 2460
+    },
+    {
+      "epoch": 0.7786464502882413,
+      "grad_norm": 0.060731887918175714,
+      "learning_rate": 0.000765824238868372,
+      "loss": 1.1212,
+      "step": 2465
+    },
+    {
+      "epoch": 0.7802258548527206,
+      "grad_norm": 0.029437468762583706,
+      "learning_rate": 0.0007646555893204442,
+      "loss": 1.0918,
+      "step": 2470
+    },
+    {
+      "epoch": 0.7818052594171997,
+      "grad_norm": 0.03426412759745619,
+      "learning_rate": 0.0007634849277718956,
+      "loss": 1.1042,
+      "step": 2475
+    },
+    {
+      "epoch": 0.7833846639816789,
+      "grad_norm": 0.043626338906299925,
+      "learning_rate": 0.0007623122631224881,
+      "loss": 1.0942,
+      "step": 2480
+    },
+    {
+      "epoch": 0.7849640685461581,
+      "grad_norm": 0.06053739130525586,
+      "learning_rate": 0.0007611376042872121,
+      "loss": 1.1401,
+      "step": 2485
+    },
+    {
+      "epoch": 0.7865434731106373,
+      "grad_norm": 0.04934003230834547,
+      "learning_rate": 0.0007599609601962183,
+      "loss": 1.1509,
+      "step": 2490
+    },
+    {
+      "epoch": 0.7881228776751165,
+      "grad_norm": 0.032994806646679455,
+      "learning_rate": 0.00075878233979475,
+      "loss": 1.1306,
+      "step": 2495
+    },
+    {
+      "epoch": 0.7897022822395957,
+      "grad_norm": 0.057874354577367164,
+      "learning_rate": 0.000757601752043075,
+      "loss": 1.1152,
+      "step": 2500
+    },
+    {
+      "epoch": 0.7912816868040748,
+      "grad_norm": 0.036916846930438035,
+      "learning_rate": 0.0007564192059164176,
+      "loss": 1.1254,
+      "step": 2505
+    },
+    {
+      "epoch": 0.7928610913685541,
+      "grad_norm": 0.03622601094242121,
+      "learning_rate": 0.0007552347104048908,
+      "loss": 1.1797,
+      "step": 2510
+    },
+    {
+      "epoch": 0.7944404959330332,
+      "grad_norm": 0.04034014908393007,
+      "learning_rate": 0.0007540482745134266,
+      "loss": 1.0887,
+      "step": 2515
+    },
+    {
+      "epoch": 0.7960199004975125,
+      "grad_norm": 0.04086451098985416,
+      "learning_rate": 0.000752859907261709,
+      "loss": 1.1606,
+      "step": 2520
+    },
+    {
+      "epoch": 0.7975993050619916,
+      "grad_norm": 0.04230873404823367,
+      "learning_rate": 0.0007516696176841048,
+      "loss": 1.1413,
+      "step": 2525
+    },
+    {
+      "epoch": 0.7991787096264709,
+      "grad_norm": 0.053041815848766695,
+      "learning_rate": 0.000750477414829595,
+      "loss": 1.2506,
+      "step": 2530
+    },
+    {
+      "epoch": 0.80075811419095,
+      "grad_norm": 0.038500571689324685,
+      "learning_rate": 0.0007492833077617053,
+      "loss": 1.2014,
+      "step": 2535
+    },
+    {
+      "epoch": 0.8023375187554292,
+      "grad_norm": 0.04095622602441375,
+      "learning_rate": 0.0007480873055584392,
+      "loss": 1.1081,
+      "step": 2540
+    },
+    {
+      "epoch": 0.8039169233199084,
+      "grad_norm": 0.06242247746111767,
+      "learning_rate": 0.0007468894173122063,
+      "loss": 1.1474,
+      "step": 2545
+    },
+    {
+      "epoch": 0.8054963278843876,
+      "grad_norm": 0.052696786458390654,
+      "learning_rate": 0.0007456896521297554,
+      "loss": 1.1867,
+      "step": 2550
+    },
+    {
+      "epoch": 0.8070757324488668,
+      "grad_norm": 0.03907948162129907,
+      "learning_rate": 0.000744488019132104,
+      "loss": 1.1588,
+      "step": 2555
+    },
+    {
+      "epoch": 0.808655137013346,
+      "grad_norm": 0.03537195704084729,
+      "learning_rate": 0.0007432845274544695,
+      "loss": 1.0816,
+      "step": 2560
+    },
+    {
+      "epoch": 0.8102345415778252,
+      "grad_norm": 0.0492539507126839,
+      "learning_rate": 0.0007420791862461997,
+      "loss": 1.0672,
+      "step": 2565
+    },
+    {
+      "epoch": 0.8118139461423044,
+      "grad_norm": 0.041180842720502576,
+      "learning_rate": 0.0007408720046707027,
+      "loss": 1.0945,
+      "step": 2570
+    },
+    {
+      "epoch": 0.8133933507067835,
+      "grad_norm": 0.03595581716482617,
+      "learning_rate": 0.0007396629919053785,
+      "loss": 1.1398,
+      "step": 2575
+    },
+    {
+      "epoch": 0.8149727552712628,
+      "grad_norm": 0.06860020808618335,
+      "learning_rate": 0.0007384521571415475,
+      "loss": 1.1856,
+      "step": 2580
+    },
+    {
+      "epoch": 0.8165521598357419,
+      "grad_norm": 0.034288016394964434,
+      "learning_rate": 0.0007372395095843823,
+      "loss": 1.0849,
+      "step": 2585
+    },
+    {
+      "epoch": 0.8181315644002212,
+      "grad_norm": 0.049671673626027694,
+      "learning_rate": 0.0007360250584528363,
+      "loss": 1.2389,
+      "step": 2590
+    },
+    {
+      "epoch": 0.8197109689647003,
+      "grad_norm": 0.04438153524293192,
+      "learning_rate": 0.000734808812979575,
+      "loss": 1.1411,
+      "step": 2595
+    },
+    {
+      "epoch": 0.8212903735291796,
+      "grad_norm": 0.030735268017338425,
+      "learning_rate": 0.0007335907824109046,
+      "loss": 1.1131,
+      "step": 2600
+    },
+    {
+      "epoch": 0.8228697780936587,
+      "grad_norm": 0.03405757193751862,
+      "learning_rate": 0.0007323709760067023,
+      "loss": 1.0823,
+      "step": 2605
+    },
+    {
+      "epoch": 0.8244491826581379,
+      "grad_norm": 0.030663575416957994,
+      "learning_rate": 0.0007311494030403458,
+      "loss": 1.1149,
+      "step": 2610
+    },
+    {
+      "epoch": 0.8260285872226171,
+      "grad_norm": 0.03208737020004201,
+      "learning_rate": 0.0007299260727986428,
+      "loss": 1.1078,
+      "step": 2615
+    },
+    {
+      "epoch": 0.8276079917870963,
+      "grad_norm": 0.040596884547164494,
+      "learning_rate": 0.0007287009945817605,
+      "loss": 1.072,
+      "step": 2620
+    },
+    {
+      "epoch": 0.8291873963515755,
+      "grad_norm": 0.037091725728933614,
+      "learning_rate": 0.0007274741777031544,
+      "loss": 1.1415,
+      "step": 2625
+    },
+    {
+      "epoch": 0.8307668009160546,
+      "grad_norm": 0.036465623744331946,
+      "learning_rate": 0.0007262456314894985,
+      "loss": 1.0951,
+      "step": 2630
+    },
+    {
+      "epoch": 0.8323462054805338,
+      "grad_norm": 0.0342176930658027,
+      "learning_rate": 0.0007250153652806133,
+      "loss": 1.119,
+      "step": 2635
+    },
+    {
+      "epoch": 0.833925610045013,
+      "grad_norm": 0.03140323676971682,
+      "learning_rate": 0.0007237833884293955,
+      "loss": 1.0438,
+      "step": 2640
+    },
+    {
+      "epoch": 0.8355050146094922,
+      "grad_norm": 0.03890983036796517,
+      "learning_rate": 0.0007225497103017467,
+      "loss": 1.0783,
+      "step": 2645
+    },
+    {
+      "epoch": 0.8370844191739714,
+      "grad_norm": 0.04169345402666985,
+      "learning_rate": 0.0007213143402765021,
+      "loss": 1.1149,
+      "step": 2650
+    },
+    {
+      "epoch": 0.8386638237384506,
+      "grad_norm": 0.05149900937684697,
+      "learning_rate": 0.0007200772877453593,
+      "loss": 1.1519,
+      "step": 2655
+    },
+    {
+      "epoch": 0.8402432283029297,
+      "grad_norm": 0.030798303288645443,
+      "learning_rate": 0.0007188385621128067,
+      "loss": 1.0422,
+      "step": 2660
+    },
+    {
+      "epoch": 0.841822632867409,
+      "grad_norm": 0.029135350330774433,
+      "learning_rate": 0.0007175981727960526,
+      "loss": 1.039,
+      "step": 2665
+    },
+    {
+      "epoch": 0.8434020374318881,
+      "grad_norm": 0.03493548525906682,
+      "learning_rate": 0.0007163561292249525,
+      "loss": 1.1766,
+      "step": 2670
+    },
+    {
+      "epoch": 0.8449814419963674,
+      "grad_norm": 0.03229186110696721,
+      "learning_rate": 0.0007151124408419389,
+      "loss": 1.0602,
+      "step": 2675
+    },
+    {
+      "epoch": 0.8465608465608465,
+      "grad_norm": 0.0763692511304746,
+      "learning_rate": 0.0007138671171019481,
+      "loss": 1.1464,
+      "step": 2680
+    },
+    {
+      "epoch": 0.8481402511253258,
+      "grad_norm": 0.046494324492253956,
+      "learning_rate": 0.0007126201674723492,
+      "loss": 1.1064,
+      "step": 2685
+    },
+    {
+      "epoch": 0.8497196556898049,
+      "grad_norm": 0.04244692069264259,
+      "learning_rate": 0.000711371601432872,
+      "loss": 1.1527,
+      "step": 2690
+    },
+    {
+      "epoch": 0.8512990602542841,
+      "grad_norm": 0.05873271313913313,
+      "learning_rate": 0.0007101214284755344,
+      "loss": 1.1121,
+      "step": 2695
+    },
+    {
+      "epoch": 0.8528784648187633,
+      "grad_norm": 0.038375326877799966,
+      "learning_rate": 0.000708869658104571,
+      "loss": 1.1012,
+      "step": 2700
+    },
+    {
+      "epoch": 0.8544578693832425,
+      "grad_norm": 0.04665919943954757,
+      "learning_rate": 0.0007076162998363603,
+      "loss": 1.0935,
+      "step": 2705
+    },
+    {
+      "epoch": 0.8560372739477217,
+      "grad_norm": 0.03946285449664685,
+      "learning_rate": 0.0007063613631993523,
+      "loss": 1.1081,
+      "step": 2710
+    },
+    {
+      "epoch": 0.8576166785122009,
+      "grad_norm": 0.04828063508924996,
+      "learning_rate": 0.0007051048577339968,
+      "loss": 1.1164,
+      "step": 2715
+    },
+    {
+      "epoch": 0.85919608307668,
+      "grad_norm": 0.06428297700542004,
+      "learning_rate": 0.00070384679299267,
+      "loss": 1.0972,
+      "step": 2720
+    },
+    {
+      "epoch": 0.8607754876411593,
+      "grad_norm": 0.03538152451128835,
+      "learning_rate": 0.0007025871785396023,
+      "loss": 1.1695,
+      "step": 2725
+    },
+    {
+      "epoch": 0.8623548922056384,
+      "grad_norm": 0.03802128698750148,
+      "learning_rate": 0.0007013260239508055,
+      "loss": 1.1753,
+      "step": 2730
+    },
+    {
+      "epoch": 0.8639342967701177,
+      "grad_norm": 0.048080492082236435,
+      "learning_rate": 0.0007000633388140002,
+      "loss": 1.177,
+      "step": 2735
+    },
+    {
+      "epoch": 0.8655137013345968,
+      "grad_norm": 0.028063478917697556,
+      "learning_rate": 0.0006987991327285425,
+      "loss": 1.0636,
+      "step": 2740
+    },
+    {
+      "epoch": 0.8670931058990761,
+      "grad_norm": 0.03558860357027592,
+      "learning_rate": 0.0006975334153053517,
+      "loss": 1.065,
+      "step": 2745
+    },
+    {
+      "epoch": 0.8686725104635552,
+      "grad_norm": 0.03215642075882598,
+      "learning_rate": 0.0006962661961668362,
+      "loss": 1.0859,
+      "step": 2750
+    },
+    {
+      "epoch": 0.8702519150280345,
+      "grad_norm": 0.032309742975668077,
+      "learning_rate": 0.0006949974849468212,
+      "loss": 1.1089,
+      "step": 2755
+    },
+    {
+      "epoch": 0.8718313195925136,
+      "grad_norm": 0.02951782677048253,
+      "learning_rate": 0.0006937272912904755,
+      "loss": 1.0983,
+      "step": 2760
+    },
+    {
+      "epoch": 0.8734107241569928,
+      "grad_norm": 0.036426668611823296,
+      "learning_rate": 0.0006924556248542373,
+      "loss": 1.129,
+      "step": 2765
+    },
+    {
+      "epoch": 0.874990128721472,
+      "grad_norm": 0.03957474298496717,
+      "learning_rate": 0.0006911824953057419,
+      "loss": 1.0902,
+      "step": 2770
+    },
+    {
+      "epoch": 0.8765695332859512,
+      "grad_norm": 0.03354407883805116,
+      "learning_rate": 0.0006899079123237473,
+      "loss": 1.124,
+      "step": 2775
+    },
+    {
+      "epoch": 0.8781489378504304,
+      "grad_norm": 0.041213643772183596,
+      "learning_rate": 0.0006886318855980611,
+      "loss": 1.1636,
+      "step": 2780
+    },
+    {
+      "epoch": 0.8797283424149096,
+      "grad_norm": 0.030287657014638987,
+      "learning_rate": 0.0006873544248294671,
+      "loss": 1.0797,
+      "step": 2785
+    },
+    {
+      "epoch": 0.8813077469793887,
+      "grad_norm": 0.03247509895879481,
+      "learning_rate": 0.0006860755397296505,
+      "loss": 1.0849,
+      "step": 2790
+    },
+    {
+      "epoch": 0.882887151543868,
+      "grad_norm": 0.04023725812499706,
+      "learning_rate": 0.0006847952400211252,
+      "loss": 1.1162,
+      "step": 2795
+    },
+    {
+      "epoch": 0.8844665561083471,
+      "grad_norm": 0.03890541651074166,
+      "learning_rate": 0.0006835135354371593,
+      "loss": 1.1291,
+      "step": 2800
+    },
+    {
+      "epoch": 0.8860459606728264,
+      "grad_norm": 0.03597174755417176,
+      "learning_rate": 0.0006822304357217013,
+      "loss": 1.1038,
+      "step": 2805
+    },
+    {
+      "epoch": 0.8876253652373055,
+      "grad_norm": 0.044296988823722266,
+      "learning_rate": 0.0006809459506293057,
+      "loss": 1.1106,
+      "step": 2810
+    },
+    {
+      "epoch": 0.8892047698017848,
+      "grad_norm": 0.031888241864467884,
+      "learning_rate": 0.0006796600899250596,
+      "loss": 1.0465,
+      "step": 2815
+    },
+    {
+      "epoch": 0.8907841743662639,
+      "grad_norm": 0.03375688432946613,
+      "learning_rate": 0.0006783728633845076,
+      "loss": 1.1087,
+      "step": 2820
+    },
+    {
+      "epoch": 0.8923635789307431,
+      "grad_norm": 0.03141469193813306,
+      "learning_rate": 0.0006770842807935777,
+      "loss": 1.111,
+      "step": 2825
+    },
+    {
+      "epoch": 0.8939429834952223,
+      "grad_norm": 0.029725912599760738,
+      "learning_rate": 0.0006757943519485075,
+      "loss": 1.1024,
+      "step": 2830
+    },
+    {
+      "epoch": 0.8955223880597015,
+      "grad_norm": 0.06555190731683326,
+      "learning_rate": 0.0006745030866557691,
+      "loss": 1.092,
+      "step": 2835
+    },
+    {
+      "epoch": 0.8971017926241807,
+      "grad_norm": 0.04010032166975462,
+      "learning_rate": 0.0006732104947319942,
+      "loss": 1.0763,
+      "step": 2840
+    },
+    {
+      "epoch": 0.8986811971886599,
+      "grad_norm": 0.04456482558579905,
+      "learning_rate": 0.0006719165860039009,
+      "loss": 1.121,
+      "step": 2845
+    },
+    {
+      "epoch": 0.900260601753139,
+      "grad_norm": 0.03629358875250488,
+      "learning_rate": 0.0006706213703082176,
+      "loss": 1.1179,
+      "step": 2850
+    },
+    {
+      "epoch": 0.9018400063176183,
+      "grad_norm": 0.03263092062169253,
+      "learning_rate": 0.0006693248574916086,
+      "loss": 1.0806,
+      "step": 2855
+    },
+    {
+      "epoch": 0.9034194108820974,
+      "grad_norm": 0.040114231565220586,
+      "learning_rate": 0.0006680270574105997,
+      "loss": 1.1045,
+      "step": 2860
+    },
+    {
+      "epoch": 0.9049988154465767,
+      "grad_norm": 0.032824962630205234,
+      "learning_rate": 0.0006667279799315025,
+      "loss": 1.0758,
+      "step": 2865
+    },
+    {
+      "epoch": 0.9065782200110558,
+      "grad_norm": 0.04457230563249612,
+      "learning_rate": 0.0006654276349303402,
+      "loss": 1.0575,
+      "step": 2870
+    },
+    {
+      "epoch": 0.9081576245755351,
+      "grad_norm": 0.07752050734868508,
+      "learning_rate": 0.0006641260322927718,
+      "loss": 1.0644,
+      "step": 2875
+    },
+    {
+      "epoch": 0.9097370291400142,
+      "grad_norm": 0.035445999376309226,
+      "learning_rate": 0.0006628231819140175,
+      "loss": 1.1192,
+      "step": 2880
+    },
+    {
+      "epoch": 0.9113164337044934,
+      "grad_norm": 0.05929887171190351,
+      "learning_rate": 0.0006615190936987833,
+      "loss": 1.0944,
+      "step": 2885
+    },
+    {
+      "epoch": 0.9128958382689726,
+      "grad_norm": 0.03267363437553562,
+      "learning_rate": 0.0006602137775611853,
+      "loss": 1.0551,
+      "step": 2890
+    },
+    {
+      "epoch": 0.9144752428334518,
+      "grad_norm": 0.029073524110190204,
+      "learning_rate": 0.000658907243424675,
+      "loss": 1.0458,
+      "step": 2895
+    },
+    {
+      "epoch": 0.916054647397931,
+      "grad_norm": 0.046992289036764744,
+      "learning_rate": 0.0006575995012219636,
+      "loss": 1.1098,
+      "step": 2900
+    },
+    {
+      "epoch": 0.9176340519624102,
+      "grad_norm": 0.037215881233934045,
+      "learning_rate": 0.000656290560894946,
+      "loss": 1.1957,
+      "step": 2905
+    },
+    {
+      "epoch": 0.9192134565268893,
+      "grad_norm": 0.037314969757336794,
+      "learning_rate": 0.0006549804323946261,
+      "loss": 1.067,
+      "step": 2910
+    },
+    {
+      "epoch": 0.9207928610913686,
+      "grad_norm": 0.05935064237729512,
+      "learning_rate": 0.0006536691256810404,
+      "loss": 1.0819,
+      "step": 2915
+    },
+    {
+      "epoch": 0.9223722656558477,
+      "grad_norm": 0.039382890006385236,
+      "learning_rate": 0.0006523566507231827,
+      "loss": 1.0803,
+      "step": 2920
+    },
+    {
+      "epoch": 0.923951670220327,
+      "grad_norm": 0.032551625760060814,
+      "learning_rate": 0.0006510430174989281,
+      "loss": 1.0521,
+      "step": 2925
+    },
+    {
+      "epoch": 0.9255310747848061,
+      "grad_norm": 0.03084938699937599,
+      "learning_rate": 0.0006497282359949574,
+      "loss": 1.0464,
+      "step": 2930
+    },
+    {
+      "epoch": 0.9271104793492854,
+      "grad_norm": 0.03540679422108104,
+      "learning_rate": 0.000648412316206681,
+      "loss": 1.0569,
+      "step": 2935
+    },
+    {
+      "epoch": 0.9286898839137645,
+      "grad_norm": 0.02845818505979114,
+      "learning_rate": 0.0006470952681381626,
+      "loss": 1.093,
+      "step": 2940
+    },
+    {
+      "epoch": 0.9302692884782437,
+      "grad_norm": 0.04412753782252594,
+      "learning_rate": 0.0006457771018020435,
+      "loss": 1.0981,
+      "step": 2945
+    },
+    {
+      "epoch": 0.9318486930427229,
+      "grad_norm": 0.032380564517949934,
+      "learning_rate": 0.0006444578272194672,
+      "loss": 1.1071,
+      "step": 2950
+    },
+    {
+      "epoch": 0.9334280976072021,
+      "grad_norm": 0.02960305717253739,
+      "learning_rate": 0.0006431374544200013,
+      "loss": 1.0815,
+      "step": 2955
+    },
+    {
+      "epoch": 0.9350075021716813,
+      "grad_norm": 0.0385572133108294,
+      "learning_rate": 0.0006418159934415634,
+      "loss": 1.0636,
+      "step": 2960
+    },
+    {
+      "epoch": 0.9365869067361605,
+      "grad_norm": 0.03940359113486163,
+      "learning_rate": 0.0006404934543303431,
+      "loss": 1.115,
+      "step": 2965
+    },
+    {
+      "epoch": 0.9381663113006397,
+      "grad_norm": 0.047952221587856184,
+      "learning_rate": 0.0006391698471407269,
+      "loss": 1.0677,
+      "step": 2970
+    },
+    {
+      "epoch": 0.9397457158651189,
+      "grad_norm": 0.042843188079932856,
+      "learning_rate": 0.0006378451819352206,
+      "loss": 1.1122,
+      "step": 2975
+    },
+    {
+      "epoch": 0.941325120429598,
+      "grad_norm": 0.03449156963379797,
+      "learning_rate": 0.0006365194687843743,
+      "loss": 1.1886,
+      "step": 2980
+    },
+    {
+      "epoch": 0.9429045249940773,
+      "grad_norm": 0.028173240496093405,
+      "learning_rate": 0.0006351927177667036,
+      "loss": 1.0572,
+      "step": 2985
+    },
+    {
+      "epoch": 0.9444839295585564,
+      "grad_norm": 0.030807817303202586,
+      "learning_rate": 0.0006338649389686157,
+      "loss": 1.1408,
+      "step": 2990
+    },
+    {
+      "epoch": 0.9460633341230357,
+      "grad_norm": 0.03588108022175335,
+      "learning_rate": 0.0006325361424843304,
+      "loss": 1.1679,
+      "step": 2995
+    },
+    {
+      "epoch": 0.9476427386875148,
+      "grad_norm": 0.030002262330937383,
+      "learning_rate": 0.0006312063384158043,
+      "loss": 1.067,
+      "step": 3000
+    },
+    {
+      "epoch": 0.9492221432519939,
+      "grad_norm": 0.04075260536403215,
+      "learning_rate": 0.0006298755368726548,
+      "loss": 1.1673,
+      "step": 3005
+    },
+    {
+      "epoch": 0.9508015478164732,
+      "grad_norm": 0.04006476615687794,
+      "learning_rate": 0.0006285437479720817,
+      "loss": 1.1876,
+      "step": 3010
+    },
+    {
+      "epoch": 0.9523809523809523,
+      "grad_norm": 0.02917383969273514,
+      "learning_rate": 0.0006272109818387909,
+      "loss": 1.1357,
+      "step": 3015
+    },
+    {
+      "epoch": 0.9539603569454316,
+      "grad_norm": 0.03406796068036105,
+      "learning_rate": 0.0006258772486049185,
+      "loss": 1.1703,
+      "step": 3020
+    },
+    {
+      "epoch": 0.9555397615099107,
+      "grad_norm": 0.05189254290140413,
+      "learning_rate": 0.0006245425584099518,
+      "loss": 1.1679,
+      "step": 3025
+    },
+    {
+      "epoch": 0.95711916607439,
+      "grad_norm": 0.03255022955779678,
+      "learning_rate": 0.0006232069214006536,
+      "loss": 1.0853,
+      "step": 3030
+    },
+    {
+      "epoch": 0.9586985706388691,
+      "grad_norm": 0.032704936361272335,
+      "learning_rate": 0.000621870347730985,
+      "loss": 1.116,
+      "step": 3035
+    },
+    {
+      "epoch": 0.9602779752033483,
+      "grad_norm": 0.03606712436906053,
+      "learning_rate": 0.0006205328475620275,
+      "loss": 1.1576,
+      "step": 3040
+    },
+    {
+      "epoch": 0.9618573797678275,
+      "grad_norm": 0.03527900701015813,
+      "learning_rate": 0.0006191944310619065,
+      "loss": 1.1359,
+      "step": 3045
+    },
+    {
+      "epoch": 0.9634367843323067,
+      "grad_norm": 0.04489452824337388,
+      "learning_rate": 0.0006178551084057134,
+      "loss": 1.1915,
+      "step": 3050
+    },
+    {
+      "epoch": 0.9650161888967859,
+      "grad_norm": 0.029752863180517432,
+      "learning_rate": 0.0006165148897754282,
+      "loss": 1.0507,
+      "step": 3055
+    },
+    {
+      "epoch": 0.9665955934612651,
+      "grad_norm": 0.03416207058808192,
+      "learning_rate": 0.0006151737853598432,
+      "loss": 1.1158,
+      "step": 3060
+    },
+    {
+      "epoch": 0.9681749980257442,
+      "grad_norm": 0.029662235175984018,
+      "learning_rate": 0.0006138318053544842,
+      "loss": 1.0316,
+      "step": 3065
+    },
+    {
+      "epoch": 0.9697544025902235,
+      "grad_norm": 0.05275392483944112,
+      "learning_rate": 0.0006124889599615336,
+      "loss": 1.0265,
+      "step": 3070
+    },
+    {
+      "epoch": 0.9713338071547026,
+      "grad_norm": 0.03356895671978958,
+      "learning_rate": 0.0006111452593897526,
+      "loss": 1.038,
+      "step": 3075
+    },
+    {
+      "epoch": 0.9729132117191819,
+      "grad_norm": 0.04726474414038403,
+      "learning_rate": 0.0006098007138544044,
+      "loss": 1.1369,
+      "step": 3080
+    },
+    {
+      "epoch": 0.974492616283661,
+      "grad_norm": 0.112571769030443,
+      "learning_rate": 0.0006084553335771749,
+      "loss": 1.0957,
+      "step": 3085
+    },
+    {
+      "epoch": 0.9760720208481403,
+      "grad_norm": 0.05682154704218681,
+      "learning_rate": 0.0006071091287860972,
+      "loss": 1.0644,
+      "step": 3090
+    },
+    {
+      "epoch": 0.9776514254126194,
+      "grad_norm": 0.045062530416325526,
+      "learning_rate": 0.0006057621097154715,
+      "loss": 1.1487,
+      "step": 3095
+    },
+    {
+      "epoch": 0.9792308299770986,
+      "grad_norm": 0.04500300458485399,
+      "learning_rate": 0.000604414286605789,
+      "loss": 1.148,
+      "step": 3100
+    },
+    {
+      "epoch": 0.9808102345415778,
+      "grad_norm": 0.03222115450318305,
+      "learning_rate": 0.0006030656697036534,
+      "loss": 1.0628,
+      "step": 3105
+    },
+    {
+      "epoch": 0.982389639106057,
+      "grad_norm": 0.03405499215516374,
+      "learning_rate": 0.0006017162692617031,
+      "loss": 1.0925,
+      "step": 3110
+    },
+    {
+      "epoch": 0.9839690436705362,
+      "grad_norm": 0.028050905956532823,
+      "learning_rate": 0.0006003660955385331,
+      "loss": 1.1066,
+      "step": 3115
+    },
+    {
+      "epoch": 0.9855484482350154,
+      "grad_norm": 0.029921219590913206,
+      "learning_rate": 0.0005990151587986171,
+      "loss": 1.1475,
+      "step": 3120
+    },
+    {
+      "epoch": 0.9871278527994946,
+      "grad_norm": 0.035043774118588174,
+      "learning_rate": 0.0005976634693122298,
+      "loss": 1.0828,
+      "step": 3125
+    },
+    {
+      "epoch": 0.9887072573639738,
+      "grad_norm": 0.03437726161120859,
+      "learning_rate": 0.0005963110373553686,
+      "loss": 1.1099,
+      "step": 3130
+    },
+    {
+      "epoch": 0.9902866619284529,
+      "grad_norm": 0.03221755496940865,
+      "learning_rate": 0.0005949578732096746,
+      "loss": 1.0652,
+      "step": 3135
+    },
+    {
+      "epoch": 0.9918660664929322,
+      "grad_norm": 0.036914998622386334,
+      "learning_rate": 0.0005936039871623563,
+      "loss": 1.0891,
+      "step": 3140
+    },
+    {
+      "epoch": 0.9934454710574113,
+      "grad_norm": 0.03178618200706658,
+      "learning_rate": 0.0005922493895061098,
+      "loss": 1.0234,
+      "step": 3145
+    },
+    {
+      "epoch": 0.9950248756218906,
+      "grad_norm": 0.028528350246715833,
+      "learning_rate": 0.0005908940905390408,
+      "loss": 1.0395,
+      "step": 3150
+    },
+    {
+      "epoch": 0.9966042801863697,
+      "grad_norm": 0.04501200819230332,
+      "learning_rate": 0.0005895381005645874,
+      "loss": 1.1871,
+      "step": 3155
+    },
+    {
+      "epoch": 0.998183684750849,
+      "grad_norm": 0.03144475794822077,
+      "learning_rate": 0.0005881814298914402,
+      "loss": 1.094,
+      "step": 3160
+    },
+    {
+      "epoch": 0.9997630893153281,
+      "grad_norm": 0.04082120452582063,
+      "learning_rate": 0.0005868240888334653,
+      "loss": 1.0823,
+      "step": 3165
+    },
+    {
+      "epoch": 0.9997630893153281,
+      "eval_loss": 1.0791449546813965,
+      "eval_runtime": 510.0567,
+      "eval_samples_per_second": 5.194,
+      "eval_steps_per_second": 1.3,
+      "step": 3165
+    },
+    {
+      "epoch": 1.0013424938798072,
+      "grad_norm": 0.07150112023496676,
+      "learning_rate": 0.0005854660877096246,
+      "loss": 1.13,
+      "step": 3170
+    },
+    {
+      "epoch": 1.0029218984442865,
+      "grad_norm": 0.0798487285566982,
+      "learning_rate": 0.000584107436843899,
+      "loss": 1.0078,
+      "step": 3175
+    },
+    {
+      "epoch": 1.0045013030087657,
+      "grad_norm": 0.06433145982933967,
+      "learning_rate": 0.0005827481465652079,
+      "loss": 0.9826,
+      "step": 3180
+    },
+    {
+      "epoch": 1.006080707573245,
+      "grad_norm": 0.04494583173932524,
+      "learning_rate": 0.0005813882272073325,
+      "loss": 0.9551,
+      "step": 3185
+    },
+    {
+      "epoch": 1.007660112137724,
+      "grad_norm": 0.03549329472007132,
+      "learning_rate": 0.0005800276891088362,
+      "loss": 0.9463,
+      "step": 3190
+    },
+    {
+      "epoch": 1.0092395167022032,
+      "grad_norm": 0.034186176588898505,
+      "learning_rate": 0.0005786665426129862,
+      "loss": 0.977,
+      "step": 3195
+    },
+    {
+      "epoch": 1.0108189212666825,
+      "grad_norm": 0.029817268432022398,
+      "learning_rate": 0.000577304798067675,
+      "loss": 0.9908,
+      "step": 3200
+    },
+    {
+      "epoch": 1.0123983258311617,
+      "grad_norm": 0.03394331617147919,
+      "learning_rate": 0.0005759424658253418,
+      "loss": 1.0151,
+      "step": 3205
+    },
+    {
+      "epoch": 1.0139777303956408,
+      "grad_norm": 0.13271163042849446,
+      "learning_rate": 0.0005745795562428936,
+      "loss": 1.0172,
+      "step": 3210
+    },
+    {
+      "epoch": 1.01555713496012,
+      "grad_norm": 0.04312287259535113,
+      "learning_rate": 0.0005732160796816266,
+      "loss": 0.968,
+      "step": 3215
+    },
+    {
+      "epoch": 1.0171365395245993,
+      "grad_norm": 0.06020683361964271,
+      "learning_rate": 0.000571852046507147,
+      "loss": 1.0422,
+      "step": 3220
+    },
+    {
+      "epoch": 1.0187159440890785,
+      "grad_norm": 0.030095876466864336,
+      "learning_rate": 0.0005704874670892929,
+      "loss": 1.0082,
+      "step": 3225
+    },
+    {
+      "epoch": 1.0202953486535575,
+      "grad_norm": 0.041989546074888764,
+      "learning_rate": 0.000569122351802055,
+      "loss": 1.0718,
+      "step": 3230
+    },
+    {
+      "epoch": 1.0218747532180368,
+      "grad_norm": 0.03248769015541394,
+      "learning_rate": 0.000567756711023498,
+      "loss": 1.0164,
+      "step": 3235
+    },
+    {
+      "epoch": 1.023454157782516,
+      "grad_norm": 0.033241056105033265,
+      "learning_rate": 0.0005663905551356816,
+      "loss": 1.0325,
+      "step": 3240
+    },
+    {
+      "epoch": 1.0250335623469953,
+      "grad_norm": 0.04113820390530121,
+      "learning_rate": 0.0005650238945245811,
+      "loss": 0.9916,
+      "step": 3245
+    },
+    {
+      "epoch": 1.0266129669114743,
+      "grad_norm": 0.047457711642047797,
+      "learning_rate": 0.0005636567395800095,
+      "loss": 1.0278,
+      "step": 3250
+    },
+    {
+      "epoch": 1.0281923714759535,
+      "grad_norm": 0.03425503116269514,
+      "learning_rate": 0.0005622891006955374,
+      "loss": 1.0648,
+      "step": 3255
+    },
+    {
+      "epoch": 1.0297717760404328,
+      "grad_norm": 0.032910706194540146,
+      "learning_rate": 0.0005609209882684147,
+      "loss": 0.9476,
+      "step": 3260
+    },
+    {
+      "epoch": 1.031351180604912,
+      "grad_norm": 0.031202607233272173,
+      "learning_rate": 0.0005595524126994912,
+      "loss": 1.0214,
+      "step": 3265
+    },
+    {
+      "epoch": 1.032930585169391,
+      "grad_norm": 0.034687369210111244,
+      "learning_rate": 0.0005581833843931377,
+      "loss": 0.9938,
+      "step": 3270
+    },
+    {
+      "epoch": 1.0345099897338703,
+      "grad_norm": 0.13437002311376092,
+      "learning_rate": 0.0005568139137571671,
+      "loss": 0.9909,
+      "step": 3275
+    },
+    {
+      "epoch": 1.0360893942983496,
+      "grad_norm": 0.029955677741687923,
+      "learning_rate": 0.0005554440112027546,
+      "loss": 0.9762,
+      "step": 3280
+    },
+    {
+      "epoch": 1.0376687988628288,
+      "grad_norm": 0.034097784104351415,
+      "learning_rate": 0.0005540736871443595,
+      "loss": 0.9952,
+      "step": 3285
+    },
+    {
+      "epoch": 1.0392482034273078,
+      "grad_norm": 0.03790878076303209,
+      "learning_rate": 0.0005527029519996448,
+      "loss": 1.0071,
+      "step": 3290
+    },
+    {
+      "epoch": 1.040827607991787,
+      "grad_norm": 0.02948940417484095,
+      "learning_rate": 0.0005513318161893996,
+      "loss": 0.9836,
+      "step": 3295
+    },
+    {
+      "epoch": 1.0424070125562663,
+      "grad_norm": 0.03107381970615086,
+      "learning_rate": 0.0005499602901374582,
+      "loss": 0.9885,
+      "step": 3300
+    },
+    {
+      "epoch": 1.0439864171207456,
+      "grad_norm": 0.03660851842140771,
+      "learning_rate": 0.0005485883842706224,
+      "loss": 0.9412,
+      "step": 3305
+    },
+    {
+      "epoch": 1.0455658216852246,
+      "grad_norm": 0.03214082310663574,
+      "learning_rate": 0.0005472161090185806,
+      "loss": 1.0371,
+      "step": 3310
+    },
+    {
+      "epoch": 1.0471452262497039,
+      "grad_norm": 0.03340724348849699,
+      "learning_rate": 0.0005458434748138302,
+      "loss": 0.9617,
+      "step": 3315
+    },
+    {
+      "epoch": 1.048724630814183,
+      "grad_norm": 0.056978614456297746,
+      "learning_rate": 0.0005444704920915971,
+      "loss": 1.0408,
+      "step": 3320
+    },
+    {
+      "epoch": 1.0503040353786623,
+      "grad_norm": 0.032025722101213884,
+      "learning_rate": 0.0005430971712897566,
+      "loss": 1.0889,
+      "step": 3325
+    },
+    {
+      "epoch": 1.0518834399431414,
+      "grad_norm": 0.03384344024540512,
+      "learning_rate": 0.0005417235228487546,
+      "loss": 1.0263,
+      "step": 3330
+    },
+    {
+      "epoch": 1.0534628445076206,
+      "grad_norm": 0.04349779981432125,
+      "learning_rate": 0.0005403495572115275,
+      "loss": 0.9939,
+      "step": 3335
+    },
+    {
+      "epoch": 1.0550422490720999,
+      "grad_norm": 0.0316311317750681,
+      "learning_rate": 0.0005389752848234234,
+      "loss": 0.9824,
+      "step": 3340
+    },
+    {
+      "epoch": 1.0566216536365791,
+      "grad_norm": 0.032659388007359264,
+      "learning_rate": 0.000537600716132122,
+      "loss": 0.9473,
+      "step": 3345
+    },
+    {
+      "epoch": 1.0582010582010581,
+      "grad_norm": 0.031611742536276584,
+      "learning_rate": 0.0005362258615875562,
+      "loss": 1.0733,
+      "step": 3350
+    },
+    {
+      "epoch": 1.0597804627655374,
+      "grad_norm": 0.03702335622451369,
+      "learning_rate": 0.0005348507316418313,
+      "loss": 1.0077,
+      "step": 3355
+    },
+    {
+      "epoch": 1.0613598673300166,
+      "grad_norm": 0.03582357424394027,
+      "learning_rate": 0.000533475336749147,
+      "loss": 0.9825,
+      "step": 3360
+    },
+    {
+      "epoch": 1.0629392718944959,
+      "grad_norm": 0.029384256667206336,
+      "learning_rate": 0.0005320996873657167,
+      "loss": 0.9317,
+      "step": 3365
+    },
+    {
+      "epoch": 1.064518676458975,
+      "grad_norm": 0.02909246543706195,
+      "learning_rate": 0.000530723793949689,
+      "loss": 1.0118,
+      "step": 3370
+    },
+    {
+      "epoch": 1.0660980810234542,
+      "grad_norm": 0.03610373021969205,
+      "learning_rate": 0.0005293476669610673,
+      "loss": 0.9777,
+      "step": 3375
+    },
+    {
+      "epoch": 1.0676774855879334,
+      "grad_norm": 0.036352202064281774,
+      "learning_rate": 0.0005279713168616309,
+      "loss": 0.9223,
+      "step": 3380
+    },
+    {
+      "epoch": 1.0692568901524124,
+      "grad_norm": 0.04168876104566738,
+      "learning_rate": 0.0005265947541148553,
+      "loss": 0.9932,
+      "step": 3385
+    },
+    {
+      "epoch": 1.0708362947168917,
+      "grad_norm": 0.03730727894381184,
+      "learning_rate": 0.0005252179891858326,
+      "loss": 0.955,
+      "step": 3390
+    },
+    {
+      "epoch": 1.072415699281371,
+      "grad_norm": 0.0313568608389757,
+      "learning_rate": 0.0005238410325411917,
+      "loss": 1.0055,
+      "step": 3395
+    },
+    {
+      "epoch": 1.0739951038458502,
+      "grad_norm": 0.0377620352643827,
+      "learning_rate": 0.0005224638946490191,
+      "loss": 0.9968,
+      "step": 3400
+    },
+    {
+      "epoch": 1.0755745084103294,
+      "grad_norm": 0.034285990273277014,
+      "learning_rate": 0.0005210865859787794,
+      "loss": 1.05,
+      "step": 3405
+    },
+    {
+      "epoch": 1.0771539129748084,
+      "grad_norm": 0.033053408056772364,
+      "learning_rate": 0.0005197091170012356,
+      "loss": 0.9527,
+      "step": 3410
+    },
+    {
+      "epoch": 1.0787333175392877,
+      "grad_norm": 0.03391521396574238,
+      "learning_rate": 0.000518331498188369,
+      "loss": 1.031,
+      "step": 3415
+    },
+    {
+      "epoch": 1.080312722103767,
+      "grad_norm": 0.03474538664274266,
+      "learning_rate": 0.0005169537400133002,
+      "loss": 1.0192,
+      "step": 3420
+    },
+    {
+      "epoch": 1.081892126668246,
+      "grad_norm": 0.030783132539478026,
+      "learning_rate": 0.0005155758529502095,
+      "loss": 1.0036,
+      "step": 3425
+    },
+    {
+      "epoch": 1.0834715312327252,
+      "grad_norm": 0.03057246653231956,
+      "learning_rate": 0.0005141978474742566,
+      "loss": 1.0117,
+      "step": 3430
+    },
+    {
+      "epoch": 1.0850509357972045,
+      "grad_norm": 0.02968806381266762,
+      "learning_rate": 0.0005128197340615018,
+      "loss": 0.9851,
+      "step": 3435
+    },
+    {
+      "epoch": 1.0866303403616837,
+      "grad_norm": 0.03520879529669726,
+      "learning_rate": 0.0005114415231888257,
+      "loss": 0.9676,
+      "step": 3440
+    },
+    {
+      "epoch": 1.0882097449261627,
+      "grad_norm": 0.03173846889255153,
+      "learning_rate": 0.0005100632253338499,
+      "loss": 1.0356,
+      "step": 3445
+    },
+    {
+      "epoch": 1.089789149490642,
+      "grad_norm": 0.031706825455682444,
+      "learning_rate": 0.0005086848509748577,
+      "loss": 1.1036,
+      "step": 3450
+    },
+    {
+      "epoch": 1.0913685540551212,
+      "grad_norm": 0.03537078439015193,
+      "learning_rate": 0.000507306410590713,
+      "loss": 1.0021,
+      "step": 3455
+    },
+    {
+      "epoch": 1.0929479586196005,
+      "grad_norm": 0.030282214060137556,
+      "learning_rate": 0.0005059279146607829,
+      "loss": 1.03,
+      "step": 3460
+    },
+    {
+      "epoch": 1.0945273631840795,
+      "grad_norm": 0.034400453962651394,
+      "learning_rate": 0.0005045493736648556,
+      "loss": 0.9934,
+      "step": 3465
+    },
+    {
+      "epoch": 1.0961067677485588,
+      "grad_norm": 10.606479686832095,
+      "learning_rate": 0.0005031707980830629,
+      "loss": 1.0075,
+      "step": 3470
+    },
+    {
+      "epoch": 1.097686172313038,
+      "grad_norm": 0.04574495818963683,
+      "learning_rate": 0.000501792198395799,
+      "loss": 0.9429,
+      "step": 3475
+    },
+    {
+      "epoch": 1.0992655768775172,
+      "grad_norm": 0.041434716901667105,
+      "learning_rate": 0.0005004135850836412,
+      "loss": 0.9915,
+      "step": 3480
+    },
+    {
+      "epoch": 1.1008449814419963,
+      "grad_norm": 0.036560297041780127,
+      "learning_rate": 0.0004990349686272709,
+      "loss": 1.0123,
+      "step": 3485
+    },
+    {
+      "epoch": 1.1024243860064755,
+      "grad_norm": 0.03740903677106789,
+      "learning_rate": 0.0004976563595073929,
+      "loss": 1.0287,
+      "step": 3490
+    },
+    {
+      "epoch": 1.1040037905709548,
+      "grad_norm": 0.03868760581227752,
+      "learning_rate": 0.0004962777682046565,
+      "loss": 0.9872,
+      "step": 3495
+    },
+    {
+      "epoch": 1.105583195135434,
+      "grad_norm": 0.04634060161494692,
+      "learning_rate": 0.0004948992051995756,
+      "loss": 1.0251,
+      "step": 3500
+    },
+    {
+      "epoch": 1.107162599699913,
+      "grad_norm": 0.02993297136622178,
+      "learning_rate": 0.0004935206809724488,
+      "loss": 0.9731,
+      "step": 3505
+    },
+    {
+      "epoch": 1.1087420042643923,
+      "grad_norm": 0.04079476364273126,
+      "learning_rate": 0.0004921422060032801,
+      "loss": 1.0498,
+      "step": 3510
+    },
+    {
+      "epoch": 1.1103214088288715,
+      "grad_norm": 0.04673875982677158,
+      "learning_rate": 0.0004907637907716987,
+      "loss": 1.0366,
+      "step": 3515
+    },
+    {
+      "epoch": 1.1119008133933508,
+      "grad_norm": 0.03821099301219827,
+      "learning_rate": 0.0004893854457568801,
+      "loss": 1.0495,
+      "step": 3520
+    },
+    {
+      "epoch": 1.1134802179578298,
+      "grad_norm": 0.03919150081513503,
+      "learning_rate": 0.0004880071814374656,
+      "loss": 0.9499,
+      "step": 3525
+    },
+    {
+      "epoch": 1.115059622522309,
+      "grad_norm": 0.03890374315742622,
+      "learning_rate": 0.0004866290082914831,
+      "loss": 1.0682,
+      "step": 3530
+    },
+    {
+      "epoch": 1.1166390270867883,
+      "grad_norm": 0.03362373378100303,
+      "learning_rate": 0.00048525093679626746,
+      "loss": 1.0049,
+      "step": 3535
+    },
+    {
+      "epoch": 1.1182184316512676,
+      "grad_norm": 0.0363331476945115,
+      "learning_rate": 0.00048387297742838085,
+      "loss": 1.0039,
+      "step": 3540
+    },
+    {
+      "epoch": 1.1197978362157466,
+      "grad_norm": 0.030750365373425098,
+      "learning_rate": 0.00048249514066353274,
+      "loss": 0.9417,
+      "step": 3545
+    },
+    {
+      "epoch": 1.1213772407802258,
+      "grad_norm": 0.038839666921607556,
+      "learning_rate": 0.0004811174369765008,
+      "loss": 0.9521,
+      "step": 3550
+    },
+    {
+      "epoch": 1.122956645344705,
+      "grad_norm": 0.03068903343661717,
+      "learning_rate": 0.0004797398768410509,
+      "loss": 1.0124,
+      "step": 3555
+    },
+    {
+      "epoch": 1.1245360499091843,
+      "grad_norm": 0.02862806963588536,
+      "learning_rate": 0.0004783624707298574,
+      "loss": 1.0724,
+      "step": 3560
+    },
+    {
+      "epoch": 1.1261154544736633,
+      "grad_norm": 0.03115930123636453,
+      "learning_rate": 0.00047698522911442397,
+      "loss": 1.0118,
+      "step": 3565
+    },
+    {
+      "epoch": 1.1276948590381426,
+      "grad_norm": 0.03220348520336603,
+      "learning_rate": 0.0004756081624650037,
+      "loss": 1.0294,
+      "step": 3570
+    },
+    {
+      "epoch": 1.1292742636026218,
+      "grad_norm": 0.029980214681574024,
+      "learning_rate": 0.0004742312812505194,
+      "loss": 0.9844,
+      "step": 3575
+    },
+    {
+      "epoch": 1.130853668167101,
+      "grad_norm": 0.03910674749748984,
+      "learning_rate": 0.00047285459593848425,
+      "loss": 1.0132,
+      "step": 3580
+    },
+    {
+      "epoch": 1.1324330727315801,
+      "grad_norm": 0.03391440952365017,
+      "learning_rate": 0.00047147811699492227,
+      "loss": 1.0222,
+      "step": 3585
+    },
+    {
+      "epoch": 1.1340124772960594,
+      "grad_norm": 0.05878162258843076,
+      "learning_rate": 0.00047010185488428793,
+      "loss": 1.0384,
+      "step": 3590
+    },
+    {
+      "epoch": 1.1355918818605386,
+      "grad_norm": 0.0327519856427672,
+      "learning_rate": 0.00046872582006938796,
+      "loss": 0.9858,
+      "step": 3595
+    },
+    {
+      "epoch": 1.1371712864250179,
+      "grad_norm": 0.03680010655471814,
+      "learning_rate": 0.00046735002301130093,
+      "loss": 0.9303,
+      "step": 3600
+    },
+    {
+      "epoch": 1.1387506909894969,
+      "grad_norm": 0.031895585784546855,
+      "learning_rate": 0.00046597447416929776,
+      "loss": 1.0007,
+      "step": 3605
+    },
+    {
+      "epoch": 1.1403300955539761,
+      "grad_norm": 0.03199689960710086,
+      "learning_rate": 0.0004645991840007627,
+      "loss": 1.004,
+      "step": 3610
+    },
+    {
+      "epoch": 1.1419095001184554,
+      "grad_norm": 0.03009576478041764,
+      "learning_rate": 0.00046322416296111296,
+      "loss": 1.039,
+      "step": 3615
+    },
+    {
+      "epoch": 1.1434889046829346,
+      "grad_norm": 0.032847682758630106,
+      "learning_rate": 0.00046184942150372007,
+      "loss": 1.0054,
+      "step": 3620
+    },
+    {
+      "epoch": 1.1450683092474137,
+      "grad_norm": 0.06698387106228133,
+      "learning_rate": 0.00046047497007983,
+      "loss": 1.0097,
+      "step": 3625
+    },
+    {
+      "epoch": 1.146647713811893,
+      "grad_norm": 0.035296251330652055,
+      "learning_rate": 0.0004591008191384838,
+      "loss": 0.9193,
+      "step": 3630
+    },
+    {
+      "epoch": 1.1482271183763721,
+      "grad_norm": 0.03264524280421541,
+      "learning_rate": 0.0004577269791264383,
+      "loss": 1.025,
+      "step": 3635
+    },
+    {
+      "epoch": 1.1498065229408514,
+      "grad_norm": 0.038864179284595034,
+      "learning_rate": 0.00045635346048808625,
+      "loss": 0.9855,
+      "step": 3640
+    },
+    {
+      "epoch": 1.1513859275053304,
+      "grad_norm": 0.03172597744457259,
+      "learning_rate": 0.0004549802736653775,
+      "loss": 1.0073,
+      "step": 3645
+    },
+    {
+      "epoch": 1.1529653320698097,
+      "grad_norm": 0.03272896511499134,
+      "learning_rate": 0.00045360742909773886,
+      "loss": 1.0158,
+      "step": 3650
+    },
+    {
+      "epoch": 1.154544736634289,
+      "grad_norm": 0.03181076711304137,
+      "learning_rate": 0.0004522349372219959,
+      "loss": 1.049,
+      "step": 3655
+    },
+    {
+      "epoch": 1.156124141198768,
+      "grad_norm": 0.029570420554063247,
+      "learning_rate": 0.0004508628084722923,
+      "loss": 0.9824,
+      "step": 3660
+    },
+    {
+      "epoch": 1.1577035457632472,
+      "grad_norm": 0.031518736372667645,
+      "learning_rate": 0.0004494910532800115,
+      "loss": 0.965,
+      "step": 3665
+    },
+    {
+      "epoch": 1.1592829503277264,
+      "grad_norm": 0.03120462181627199,
+      "learning_rate": 0.00044811968207369675,
+      "loss": 1.0007,
+      "step": 3670
+    },
+    {
+      "epoch": 1.1608623548922057,
+      "grad_norm": 0.030615282637634297,
+      "learning_rate": 0.0004467487052789724,
+      "loss": 0.9676,
+      "step": 3675
+    },
+    {
+      "epoch": 1.162441759456685,
+      "grad_norm": 0.03218785070440956,
+      "learning_rate": 0.00044537813331846414,
+      "loss": 1.0053,
+      "step": 3680
+    },
+    {
+      "epoch": 1.164021164021164,
+      "grad_norm": 0.037644209561606065,
+      "learning_rate": 0.00044400797661172016,
+      "loss": 1.0194,
+      "step": 3685
+    },
+    {
+      "epoch": 1.1656005685856432,
+      "grad_norm": 0.036131652266939686,
+      "learning_rate": 0.00044263824557513144,
+      "loss": 0.9678,
+      "step": 3690
+    },
+    {
+      "epoch": 1.1671799731501225,
+      "grad_norm": 0.0345485607793738,
+      "learning_rate": 0.00044126895062185324,
+      "loss": 1.0913,
+      "step": 3695
+    },
+    {
+      "epoch": 1.1687593777146015,
+      "grad_norm": 0.10534071784088957,
+      "learning_rate": 0.00043990010216172533,
+      "loss": 0.9556,
+      "step": 3700
+    },
+    {
+      "epoch": 1.1703387822790807,
+      "grad_norm": 0.031917678600092154,
+      "learning_rate": 0.000438531710601193,
+      "loss": 0.9676,
+      "step": 3705
+    },
+    {
+      "epoch": 1.17191818684356,
+      "grad_norm": 0.03528790633782125,
+      "learning_rate": 0.00043716378634322834,
+      "loss": 1.0752,
+      "step": 3710
+    },
+    {
+      "epoch": 1.1734975914080392,
+      "grad_norm": 0.06247356214049855,
+      "learning_rate": 0.00043579633978725065,
+      "loss": 0.986,
+      "step": 3715
+    },
+    {
+      "epoch": 1.1750769959725185,
+      "grad_norm": 0.03594855583679661,
+      "learning_rate": 0.00043442938132904767,
+      "loss": 0.9764,
+      "step": 3720
+    },
+    {
+      "epoch": 1.1766564005369975,
+      "grad_norm": 0.03314206038497171,
+      "learning_rate": 0.00043306292136069646,
+      "loss": 1.0098,
+      "step": 3725
+    },
+    {
+      "epoch": 1.1782358051014767,
+      "grad_norm": 0.031324091896376624,
+      "learning_rate": 0.0004316969702704842,
+      "loss": 0.9796,
+      "step": 3730
+    },
+    {
+      "epoch": 1.179815209665956,
+      "grad_norm": 0.035907780461936295,
+      "learning_rate": 0.0004303315384428298,
+      "loss": 1.0284,
+      "step": 3735
+    },
+    {
+      "epoch": 1.181394614230435,
+      "grad_norm": 0.04235651655577545,
+      "learning_rate": 0.0004289666362582041,
+      "loss": 1.0072,
+      "step": 3740
+    },
+    {
+      "epoch": 1.1829740187949143,
+      "grad_norm": 0.02825475220475871,
+      "learning_rate": 0.00042760227409305166,
+      "loss": 0.954,
+      "step": 3745
+    },
+    {
+      "epoch": 1.1845534233593935,
+      "grad_norm": 0.03422905250774198,
+      "learning_rate": 0.0004262384623197116,
+      "loss": 0.9824,
+      "step": 3750
+    },
+    {
+      "epoch": 1.1861328279238728,
+      "grad_norm": 0.03097530803468814,
+      "learning_rate": 0.0004248752113063388,
+      "loss": 1.0269,
+      "step": 3755
+    },
+    {
+      "epoch": 1.187712232488352,
+      "grad_norm": 0.03167783403621823,
+      "learning_rate": 0.0004235125314168251,
+      "loss": 1.04,
+      "step": 3760
+    },
+    {
+      "epoch": 1.189291637052831,
+      "grad_norm": 0.03228159018489407,
+      "learning_rate": 0.00042215043301072037,
+      "loss": 0.9895,
+      "step": 3765
+    },
+    {
+      "epoch": 1.1908710416173103,
+      "grad_norm": 0.028456877325231795,
+      "learning_rate": 0.00042078892644315387,
+      "loss": 0.9235,
+      "step": 3770
+    },
+    {
+      "epoch": 1.1924504461817895,
+      "grad_norm": 0.04042505241231098,
+      "learning_rate": 0.0004194280220647556,
+      "loss": 1.0244,
+      "step": 3775
+    },
+    {
+      "epoch": 1.1940298507462686,
+      "grad_norm": 0.03214462728655419,
+      "learning_rate": 0.00041806773022157716,
+      "loss": 1.0068,
+      "step": 3780
+    },
+    {
+      "epoch": 1.1956092553107478,
+      "grad_norm": 0.028459469654203682,
+      "learning_rate": 0.00041670806125501393,
+      "loss": 1.0081,
+      "step": 3785
+    },
+    {
+      "epoch": 1.197188659875227,
+      "grad_norm": 0.03278427499854533,
+      "learning_rate": 0.0004153490255017257,
+      "loss": 0.9328,
+      "step": 3790
+    },
+    {
+      "epoch": 1.1987680644397063,
+      "grad_norm": 0.03807036409466499,
+      "learning_rate": 0.00041399063329355853,
+      "loss": 1.0445,
+      "step": 3795
+    },
+    {
+      "epoch": 1.2003474690041855,
+      "grad_norm": 0.04133519873285916,
+      "learning_rate": 0.00041263289495746574,
+      "loss": 1.0262,
+      "step": 3800
+    },
+    {
+      "epoch": 1.2019268735686646,
+      "grad_norm": 0.039450935145532044,
+      "learning_rate": 0.00041127582081543,
+      "loss": 1.1415,
+      "step": 3805
+    },
+    {
+      "epoch": 1.2035062781331438,
+      "grad_norm": 0.03637001529973416,
+      "learning_rate": 0.0004099194211843847,
+      "loss": 0.9529,
+      "step": 3810
+    },
+    {
+      "epoch": 1.205085682697623,
+      "grad_norm": 0.034497184596985475,
+      "learning_rate": 0.0004085637063761346,
+      "loss": 1.0775,
+      "step": 3815
+    },
+    {
+      "epoch": 1.206665087262102,
+      "grad_norm": 0.02959202086144998,
+      "learning_rate": 0.0004072086866972789,
+      "loss": 1.0161,
+      "step": 3820
+    },
+    {
+      "epoch": 1.2082444918265813,
+      "grad_norm": 0.0311951928316788,
+      "learning_rate": 0.00040585437244913217,
+      "loss": 0.9326,
+      "step": 3825
+    },
+    {
+      "epoch": 1.2098238963910606,
+      "grad_norm": 0.03179771229016227,
+      "learning_rate": 0.0004045007739276456,
+      "loss": 0.9844,
+      "step": 3830
+    },
+    {
+      "epoch": 1.2114033009555398,
+      "grad_norm": 0.031644743069673686,
+      "learning_rate": 0.0004031479014233297,
+      "loss": 0.9812,
+      "step": 3835
+    },
+    {
+      "epoch": 1.212982705520019,
+      "grad_norm": 0.03323986149903773,
+      "learning_rate": 0.0004017957652211753,
+      "loss": 0.9487,
+      "step": 3840
+    },
+    {
+      "epoch": 1.214562110084498,
+      "grad_norm": 0.05569080062496465,
+      "learning_rate": 0.00040044437560057567,
+      "loss": 1.0618,
+      "step": 3845
+    },
+    {
+      "epoch": 1.2161415146489774,
+      "grad_norm": 0.04109910399076241,
+      "learning_rate": 0.0003990937428352482,
+      "loss": 1.0304,
+      "step": 3850
+    },
+    {
+      "epoch": 1.2177209192134566,
+      "grad_norm": 0.03505379674466537,
+      "learning_rate": 0.00039774387719315664,
+      "loss": 1.0664,
+      "step": 3855
+    },
+    {
+      "epoch": 1.2193003237779356,
+      "grad_norm": 0.03418242604516485,
+      "learning_rate": 0.00039639478893643257,
+      "loss": 1.018,
+      "step": 3860
+    },
+    {
+      "epoch": 1.2208797283424149,
+      "grad_norm": 0.05819456896356674,
+      "learning_rate": 0.00039504648832129787,
+      "loss": 0.9534,
+      "step": 3865
+    },
+    {
+      "epoch": 1.2224591329068941,
+      "grad_norm": 0.03845387810065682,
+      "learning_rate": 0.00039369898559798614,
+      "loss": 1.0085,
+      "step": 3870
+    },
+    {
+      "epoch": 1.2240385374713734,
+      "grad_norm": 0.03002188707299774,
+      "learning_rate": 0.0003923522910106656,
+      "loss": 1.0685,
+      "step": 3875
+    },
+    {
+      "epoch": 1.2256179420358524,
+      "grad_norm": 0.03350032348038046,
+      "learning_rate": 0.0003910064147973603,
+      "loss": 1.0133,
+      "step": 3880
+    },
+    {
+      "epoch": 1.2271973466003316,
+      "grad_norm": 0.032399311229717676,
+      "learning_rate": 0.0003896613671898732,
+      "loss": 1.011,
+      "step": 3885
+    },
+    {
+      "epoch": 1.2287767511648109,
+      "grad_norm": 0.03717438638700494,
+      "learning_rate": 0.00038831715841370745,
+      "loss": 0.9171,
+      "step": 3890
+    },
+    {
+      "epoch": 1.2303561557292901,
+      "grad_norm": 0.03669686520409071,
+      "learning_rate": 0.0003869737986879895,
+      "loss": 0.9085,
+      "step": 3895
+    },
+    {
+      "epoch": 1.2319355602937692,
+      "grad_norm": 0.028568004417021613,
+      "learning_rate": 0.0003856312982253909,
+      "loss": 0.98,
+      "step": 3900
+    },
+    {
+      "epoch": 1.2335149648582484,
+      "grad_norm": 0.0337167611533833,
+      "learning_rate": 0.0003842896672320506,
+      "loss": 1.0053,
+      "step": 3905
+    },
+    {
+      "epoch": 1.2350943694227277,
+      "grad_norm": 0.033327802875302,
+      "learning_rate": 0.00038294891590749783,
+      "loss": 1.0156,
+      "step": 3910
+    },
+    {
+      "epoch": 1.236673773987207,
+      "grad_norm": 0.040464892884495215,
+      "learning_rate": 0.0003816090544445741,
+      "loss": 0.9311,
+      "step": 3915
+    },
+    {
+      "epoch": 1.238253178551686,
+      "grad_norm": 0.031169717492686492,
+      "learning_rate": 0.0003802700930293563,
+      "loss": 0.9942,
+      "step": 3920
+    },
+    {
+      "epoch": 1.2398325831161652,
+      "grad_norm": 0.03608779188581459,
+      "learning_rate": 0.00037893204184107803,
+      "loss": 0.9781,
+      "step": 3925
+    },
+    {
+      "epoch": 1.2414119876806444,
+      "grad_norm": 0.034353318606698664,
+      "learning_rate": 0.0003775949110520538,
+      "loss": 0.9602,
+      "step": 3930
+    },
+    {
+      "epoch": 1.2429913922451237,
+      "grad_norm": 0.032459644273819636,
+      "learning_rate": 0.00037625871082760064,
+      "loss": 1.0103,
+      "step": 3935
+    },
+    {
+      "epoch": 1.2445707968096027,
+      "grad_norm": 0.03287385312411447,
+      "learning_rate": 0.00037492345132596113,
+      "loss": 1.0343,
+      "step": 3940
+    },
+    {
+      "epoch": 1.246150201374082,
+      "grad_norm": 0.09624463706888049,
+      "learning_rate": 0.0003735891426982262,
+      "loss": 1.001,
+      "step": 3945
+    },
+    {
+      "epoch": 1.2477296059385612,
+      "grad_norm": 0.039854815726468845,
+      "learning_rate": 0.00037225579508825803,
+      "loss": 0.9973,
+      "step": 3950
+    },
+    {
+      "epoch": 1.2493090105030404,
+      "grad_norm": 0.03282797081588262,
+      "learning_rate": 0.0003709234186326124,
+      "loss": 1.007,
+      "step": 3955
+    },
+    {
+      "epoch": 1.2508884150675195,
+      "grad_norm": 0.04323405227732425,
+      "learning_rate": 0.0003695920234604625,
+      "loss": 0.9519,
+      "step": 3960
+    },
+    {
+      "epoch": 1.2524678196319987,
+      "grad_norm": 0.030804057889002943,
+      "learning_rate": 0.00036826161969352137,
+      "loss": 0.9538,
+      "step": 3965
+    },
+    {
+      "epoch": 1.254047224196478,
+      "grad_norm": 0.033900286482032485,
+      "learning_rate": 0.00036693221744596476,
+      "loss": 0.9878,
+      "step": 3970
+    },
+    {
+      "epoch": 1.255626628760957,
+      "grad_norm": 0.029308009262548004,
+      "learning_rate": 0.000365603826824355,
+      "loss": 1.0579,
+      "step": 3975
+    },
+    {
+      "epoch": 1.2572060333254362,
+      "grad_norm": 0.03242275017266416,
+      "learning_rate": 0.00036427645792756335,
+      "loss": 1.0618,
+      "step": 3980
+    },
+    {
+      "epoch": 1.2587854378899155,
+      "grad_norm": 0.0302414222139075,
+      "learning_rate": 0.0003629501208466938,
+      "loss": 0.9297,
+      "step": 3985
+    },
+    {
+      "epoch": 1.2603648424543947,
+      "grad_norm": 0.03279689599577794,
+      "learning_rate": 0.000361624825665006,
+      "loss": 0.9962,
+      "step": 3990
+    },
+    {
+      "epoch": 1.261944247018874,
+      "grad_norm": 0.033006660420700024,
+      "learning_rate": 0.0003603005824578386,
+      "loss": 0.93,
+      "step": 3995
+    },
+    {
+      "epoch": 1.263523651583353,
+      "grad_norm": 0.03197012821917078,
+      "learning_rate": 0.00035897740129253296,
+      "loss": 0.9839,
+      "step": 4000
+    },
+    {
+      "epoch": 1.2651030561478323,
+      "grad_norm": 0.03230977381844998,
+      "learning_rate": 0.00035765529222835666,
+      "loss": 0.9787,
+      "step": 4005
+    },
+    {
+      "epoch": 1.2666824607123115,
+      "grad_norm": 0.031821233663209894,
+      "learning_rate": 0.00035633426531642625,
+      "loss": 0.9809,
+      "step": 4010
+    },
+    {
+      "epoch": 1.2682618652767905,
+      "grad_norm": 0.03646680985111504,
+      "learning_rate": 0.00035501433059963194,
+      "loss": 0.9671,
+      "step": 4015
+    },
+    {
+      "epoch": 1.2698412698412698,
+      "grad_norm": 0.034147950422965034,
+      "learning_rate": 0.00035369549811256043,
+      "loss": 0.9884,
+      "step": 4020
+    },
+    {
+      "epoch": 1.271420674405749,
+      "grad_norm": 0.04178973681830019,
+      "learning_rate": 0.00035237777788141896,
+      "loss": 1.019,
+      "step": 4025
+    },
+    {
+      "epoch": 1.2730000789702283,
+      "grad_norm": 0.03610221769429688,
+      "learning_rate": 0.00035106117992395893,
+      "loss": 0.9549,
+      "step": 4030
+    },
+    {
+      "epoch": 1.2745794835347075,
+      "grad_norm": 0.032643633732576,
+      "learning_rate": 0.00034974571424940007,
+      "loss": 1.0115,
+      "step": 4035
+    },
+    {
+      "epoch": 1.2761588880991865,
+      "grad_norm": 0.03551475992576843,
+      "learning_rate": 0.0003484313908583538,
+      "loss": 1.0712,
+      "step": 4040
+    },
+    {
+      "epoch": 1.2777382926636658,
+      "grad_norm": 0.03168699509132295,
+      "learning_rate": 0.0003471182197427477,
+      "loss": 0.9868,
+      "step": 4045
+    },
+    {
+      "epoch": 1.279317697228145,
+      "grad_norm": 0.04010531950002586,
+      "learning_rate": 0.00034580621088574944,
+      "loss": 1.0084,
+      "step": 4050
+    },
+    {
+      "epoch": 1.280897101792624,
+      "grad_norm": 0.034498613645441746,
+      "learning_rate": 0.00034449537426169065,
+      "loss": 0.9823,
+      "step": 4055
+    },
+    {
+      "epoch": 1.2824765063571033,
+      "grad_norm": 0.03821759329478121,
+      "learning_rate": 0.00034318571983599146,
+      "loss": 1.0177,
+      "step": 4060
+    },
+    {
+      "epoch": 1.2840559109215826,
+      "grad_norm": 0.03673721054940819,
+      "learning_rate": 0.00034187725756508426,
+      "loss": 0.9367,
+      "step": 4065
+    },
+    {
+      "epoch": 1.2856353154860618,
+      "grad_norm": 0.03384213798714474,
+      "learning_rate": 0.0003405699973963384,
+      "loss": 0.9227,
+      "step": 4070
+    },
+    {
+      "epoch": 1.287214720050541,
+      "grad_norm": 0.034090758745436614,
+      "learning_rate": 0.0003392639492679846,
+      "loss": 1.0423,
+      "step": 4075
+    },
+    {
+      "epoch": 1.28879412461502,
+      "grad_norm": 0.03245882007994995,
+      "learning_rate": 0.0003379591231090391,
+      "loss": 0.9629,
+      "step": 4080
+    },
+    {
+      "epoch": 1.2903735291794993,
+      "grad_norm": 0.03496114506115646,
+      "learning_rate": 0.00033665552883922815,
+      "loss": 1.0429,
+      "step": 4085
+    },
+    {
+      "epoch": 1.2919529337439786,
+      "grad_norm": 0.03294034716543751,
+      "learning_rate": 0.00033535317636891306,
+      "loss": 0.9318,
+      "step": 4090
+    },
+    {
+      "epoch": 1.2935323383084576,
+      "grad_norm": 0.03295135988949754,
+      "learning_rate": 0.0003340520755990144,
+      "loss": 0.9867,
+      "step": 4095
+    },
+    {
+      "epoch": 1.2951117428729368,
+      "grad_norm": 0.03690628663937476,
+      "learning_rate": 0.0003327522364209369,
+      "loss": 0.9245,
+      "step": 4100
+    },
+    {
+      "epoch": 1.296691147437416,
+      "grad_norm": 0.029297125596864402,
+      "learning_rate": 0.0003314536687164944,
+      "loss": 1.0122,
+      "step": 4105
+    },
+    {
+      "epoch": 1.2982705520018953,
+      "grad_norm": 0.029269281575792814,
+      "learning_rate": 0.0003301563823578343,
+      "loss": 0.9307,
+      "step": 4110
+    },
+    {
+      "epoch": 1.2998499565663746,
+      "grad_norm": 0.034022614599823864,
+      "learning_rate": 0.0003288603872073631,
+      "loss": 0.9801,
+      "step": 4115
+    },
+    {
+      "epoch": 1.3014293611308536,
+      "grad_norm": 0.08295291494782484,
+      "learning_rate": 0.00032756569311767083,
+      "loss": 1.0192,
+      "step": 4120
+    },
+    {
+      "epoch": 1.3030087656953329,
+      "grad_norm": 0.05198063709991113,
+      "learning_rate": 0.00032627230993145643,
+      "loss": 0.9641,
+      "step": 4125
+    },
+    {
+      "epoch": 1.304588170259812,
+      "grad_norm": 0.0386293310510888,
+      "learning_rate": 0.0003249802474814532,
+      "loss": 1.0402,
+      "step": 4130
+    },
+    {
+      "epoch": 1.3061675748242911,
+      "grad_norm": 0.03262356907346096,
+      "learning_rate": 0.0003236895155903533,
+      "loss": 0.9112,
+      "step": 4135
+    },
+    {
+      "epoch": 1.3077469793887704,
+      "grad_norm": 0.036464664779840325,
+      "learning_rate": 0.000322400124070734,
+      "loss": 1.0295,
+      "step": 4140
+    },
+    {
+      "epoch": 1.3093263839532496,
+      "grad_norm": 0.0498083025164631,
+      "learning_rate": 0.0003211120827249827,
+      "loss": 0.9529,
+      "step": 4145
+    },
+    {
+      "epoch": 1.3109057885177289,
+      "grad_norm": 0.03277934651515331,
+      "learning_rate": 0.0003198254013452214,
+      "loss": 0.9878,
+      "step": 4150
+    },
+    {
+      "epoch": 1.3124851930822081,
+      "grad_norm": 0.03799182246479796,
+      "learning_rate": 0.0003185400897132341,
+      "loss": 0.9981,
+      "step": 4155
+    },
+    {
+      "epoch": 1.3140645976466871,
+      "grad_norm": 0.030928525858774297,
+      "learning_rate": 0.0003172561576003913,
+      "loss": 0.9849,
+      "step": 4160
+    },
+    {
+      "epoch": 1.3156440022111664,
+      "grad_norm": 0.032931441479663676,
+      "learning_rate": 0.00031597361476757587,
+      "loss": 0.9871,
+      "step": 4165
+    },
+    {
+      "epoch": 1.3172234067756456,
+      "grad_norm": 0.03860824643776728,
+      "learning_rate": 0.0003146924709651089,
+      "loss": 1.0179,
+      "step": 4170
+    },
+    {
+      "epoch": 1.3188028113401247,
+      "grad_norm": 0.036134302819682045,
+      "learning_rate": 0.0003134127359326755,
+      "loss": 0.9913,
+      "step": 4175
+    },
+    {
+      "epoch": 1.320382215904604,
+      "grad_norm": 0.04505322900858715,
+      "learning_rate": 0.000312134419399251,
+      "loss": 0.9884,
+      "step": 4180
+    },
+    {
+      "epoch": 1.3219616204690832,
+      "grad_norm": 0.03583021484325502,
+      "learning_rate": 0.0003108575310830266,
+      "loss": 1.0335,
+      "step": 4185
+    },
+    {
+      "epoch": 1.3235410250335624,
+      "grad_norm": 0.028584860073887837,
+      "learning_rate": 0.00030958208069133613,
+      "loss": 0.9472,
+      "step": 4190
+    },
+    {
+      "epoch": 1.3251204295980417,
+      "grad_norm": 0.032560726753026195,
+      "learning_rate": 0.00030830807792058137,
+      "loss": 0.9358,
+      "step": 4195
+    },
+    {
+      "epoch": 1.3266998341625207,
+      "grad_norm": 0.03201200510999583,
+      "learning_rate": 0.0003070355324561591,
+      "loss": 0.9581,
+      "step": 4200
+    },
+    {
+      "epoch": 1.328279238727,
+      "grad_norm": 0.02863524059045198,
+      "learning_rate": 0.0003057644539723871,
+      "loss": 0.9479,
+      "step": 4205
+    },
+    {
+      "epoch": 1.3298586432914792,
+      "grad_norm": 0.02861412563443391,
+      "learning_rate": 0.00030449485213243047,
+      "loss": 0.9869,
+      "step": 4210
+    },
+    {
+      "epoch": 1.3314380478559582,
+      "grad_norm": 0.0297763674050006,
+      "learning_rate": 0.00030322673658822864,
+      "loss": 0.9998,
+      "step": 4215
+    },
+    {
+      "epoch": 1.3330174524204375,
+      "grad_norm": 0.03234585505718821,
+      "learning_rate": 0.0003019601169804216,
+      "loss": 0.9828,
+      "step": 4220
+    },
+    {
+      "epoch": 1.3345968569849167,
+      "grad_norm": 0.03260293388067336,
+      "learning_rate": 0.00030069500293827676,
+      "loss": 0.9266,
+      "step": 4225
+    },
+    {
+      "epoch": 1.336176261549396,
+      "grad_norm": 0.032776920157258686,
+      "learning_rate": 0.00029943140407961565,
+      "loss": 1.0039,
+      "step": 4230
+    },
+    {
+      "epoch": 1.3377556661138752,
+      "grad_norm": 0.08314838417227068,
+      "learning_rate": 0.000298169330010741,
+      "loss": 0.9715,
+      "step": 4235
+    },
+    {
+      "epoch": 1.3393350706783542,
+      "grad_norm": 0.028657099663928653,
+      "learning_rate": 0.0002969087903263635,
+      "loss": 0.9944,
+      "step": 4240
+    },
+    {
+      "epoch": 1.3409144752428335,
+      "grad_norm": 0.03255072028797001,
+      "learning_rate": 0.0002956497946095289,
+      "loss": 0.9984,
+      "step": 4245
+    },
+    {
+      "epoch": 1.3424938798073127,
+      "grad_norm": 0.029853697845549056,
+      "learning_rate": 0.0002943923524315451,
+      "loss": 0.9864,
+      "step": 4250
+    },
+    {
+      "epoch": 1.3440732843717917,
+      "grad_norm": 0.02927271301894455,
+      "learning_rate": 0.00029313647335190975,
+      "loss": 0.9946,
+      "step": 4255
+    },
+    {
+      "epoch": 1.345652688936271,
+      "grad_norm": 0.03457055693857293,
+      "learning_rate": 0.0002918821669182372,
+      "loss": 0.9119,
+      "step": 4260
+    },
+    {
+      "epoch": 1.3472320935007502,
+      "grad_norm": 0.030506666861336504,
+      "learning_rate": 0.00029062944266618565,
+      "loss": 0.95,
+      "step": 4265
+    },
+    {
+      "epoch": 1.3488114980652295,
+      "grad_norm": 0.02777303737819793,
+      "learning_rate": 0.00028937831011938565,
+      "loss": 0.9663,
+      "step": 4270
+    },
+    {
+      "epoch": 1.3503909026297087,
+      "grad_norm": 0.028472861399310736,
+      "learning_rate": 0.0002881287787893666,
+      "loss": 0.9641,
+      "step": 4275
+    },
+    {
+      "epoch": 1.3519703071941878,
+      "grad_norm": 0.03335275141344005,
+      "learning_rate": 0.00028688085817548504,
+      "loss": 0.9274,
+      "step": 4280
+    },
+    {
+      "epoch": 1.353549711758667,
+      "grad_norm": 0.03476091838113158,
+      "learning_rate": 0.0002856345577648526,
+      "loss": 1.0074,
+      "step": 4285
+    },
+    {
+      "epoch": 1.3551291163231463,
+      "grad_norm": 0.029396912738058084,
+      "learning_rate": 0.00028438988703226287,
+      "loss": 0.95,
+      "step": 4290
+    },
+    {
+      "epoch": 1.3567085208876253,
+      "grad_norm": 0.03292824454164419,
+      "learning_rate": 0.000283146855440121,
+      "loss": 1.0241,
+      "step": 4295
+    },
+    {
+      "epoch": 1.3582879254521045,
+      "grad_norm": 0.02933063711821805,
+      "learning_rate": 0.00028190547243836994,
+      "loss": 0.9273,
+      "step": 4300
+    },
+    {
+      "epoch": 1.3598673300165838,
+      "grad_norm": 0.0487532524969538,
+      "learning_rate": 0.0002806657474644204,
+      "loss": 0.9284,
+      "step": 4305
+    },
+    {
+      "epoch": 1.361446734581063,
+      "grad_norm": 0.04641854021034858,
+      "learning_rate": 0.00027942768994307734,
+      "loss": 0.9597,
+      "step": 4310
+    },
+    {
+      "epoch": 1.363026139145542,
+      "grad_norm": 0.03899217449204588,
+      "learning_rate": 0.0002781913092864699,
+      "loss": 1.0238,
+      "step": 4315
+    },
+    {
+      "epoch": 1.3646055437100213,
+      "grad_norm": 0.41874580707406633,
+      "learning_rate": 0.0002769566148939787,
+      "loss": 0.9863,
+      "step": 4320
+    },
+    {
+      "epoch": 1.3661849482745005,
+      "grad_norm": 0.03254665656229802,
+      "learning_rate": 0.0002757236161521647,
+      "loss": 1.0182,
+      "step": 4325
+    },
+    {
+      "epoch": 1.3677643528389796,
+      "grad_norm": 0.031051834601741202,
+      "learning_rate": 0.00027449232243469856,
+      "loss": 1.0195,
+      "step": 4330
+    },
+    {
+      "epoch": 1.3693437574034588,
+      "grad_norm": 0.03681112953763141,
+      "learning_rate": 0.00027326274310228806,
+      "loss": 1.0055,
+      "step": 4335
+    },
+    {
+      "epoch": 1.370923161967938,
+      "grad_norm": 0.029500637011015705,
+      "learning_rate": 0.0002720348875026083,
+      "loss": 0.9259,
+      "step": 4340
+    },
+    {
+      "epoch": 1.3725025665324173,
+      "grad_norm": 0.029781036898435677,
+      "learning_rate": 0.0002708087649702294,
+      "loss": 0.9307,
+      "step": 4345
+    },
+    {
+      "epoch": 1.3740819710968966,
+      "grad_norm": 0.033142957599914515,
+      "learning_rate": 0.00026958438482654667,
+      "loss": 0.9918,
+      "step": 4350
+    },
+    {
+      "epoch": 1.3756613756613756,
+      "grad_norm": 0.028352927356558807,
+      "learning_rate": 0.0002683617563797088,
+      "loss": 0.982,
+      "step": 4355
+    },
+    {
+      "epoch": 1.3772407802258548,
+      "grad_norm": 0.03140033057258207,
+      "learning_rate": 0.0002671408889245475,
+      "loss": 0.9288,
+      "step": 4360
+    },
+    {
+      "epoch": 1.378820184790334,
+      "grad_norm": 0.035233655109260735,
+      "learning_rate": 0.0002659217917425071,
+      "loss": 0.9746,
+      "step": 4365
+    },
+    {
+      "epoch": 1.380399589354813,
+      "grad_norm": 0.0310879621744834,
+      "learning_rate": 0.00026470447410157353,
+      "loss": 0.9245,
+      "step": 4370
+    },
+    {
+      "epoch": 1.3819789939192924,
+      "grad_norm": 0.03545557030251083,
+      "learning_rate": 0.0002634889452562041,
+      "loss": 0.9832,
+      "step": 4375
+    },
+    {
+      "epoch": 1.3835583984837716,
+      "grad_norm": 0.03221098890038632,
+      "learning_rate": 0.00026227521444725685,
+      "loss": 0.9617,
+      "step": 4380
+    },
+    {
+      "epoch": 1.3851378030482508,
+      "grad_norm": 0.03575604783785744,
+      "learning_rate": 0.0002610632909019211,
+      "loss": 0.9869,
+      "step": 4385
+    },
+    {
+      "epoch": 1.38671720761273,
+      "grad_norm": 0.03000837213143947,
+      "learning_rate": 0.0002598531838336461,
+      "loss": 0.9722,
+      "step": 4390
+    },
+    {
+      "epoch": 1.3882966121772091,
+      "grad_norm": 0.0313471817306669,
+      "learning_rate": 0.0002586449024420724,
+      "loss": 1.0356,
+      "step": 4395
+    },
+    {
+      "epoch": 1.3898760167416884,
+      "grad_norm": 0.03130166860322959,
+      "learning_rate": 0.0002574384559129602,
+      "loss": 0.9598,
+      "step": 4400
+    },
+    {
+      "epoch": 1.3914554213061676,
+      "grad_norm": 0.08264592100818809,
+      "learning_rate": 0.00025623385341812135,
+      "loss": 0.9846,
+      "step": 4405
+    },
+    {
+      "epoch": 1.3930348258706466,
+      "grad_norm": 0.030114349700177814,
+      "learning_rate": 0.0002550311041153482,
+      "loss": 0.9582,
+      "step": 4410
+    },
+    {
+      "epoch": 1.394614230435126,
+      "grad_norm": 0.03521345245155562,
+      "learning_rate": 0.0002538302171483444,
+      "loss": 0.9928,
+      "step": 4415
+    },
+    {
+      "epoch": 1.3961936349996051,
+      "grad_norm": 0.03229933078643899,
+      "learning_rate": 0.0002526312016466562,
+      "loss": 0.9755,
+      "step": 4420
+    },
+    {
+      "epoch": 1.3977730395640844,
+      "grad_norm": 0.032182885993634974,
+      "learning_rate": 0.0002514340667256014,
+      "loss": 0.9613,
+      "step": 4425
+    },
+    {
+      "epoch": 1.3993524441285636,
+      "grad_norm": 0.030631941234258123,
+      "learning_rate": 0.00025023882148620205,
+      "loss": 0.9776,
+      "step": 4430
+    },
+    {
+      "epoch": 1.4009318486930427,
+      "grad_norm": 0.02987505008410226,
+      "learning_rate": 0.00024904547501511306,
+      "loss": 0.913,
+      "step": 4435
+    },
+    {
+      "epoch": 1.402511253257522,
+      "grad_norm": 0.028145259876542112,
+      "learning_rate": 0.00024785403638455535,
+      "loss": 0.934,
+      "step": 4440
+    },
+    {
+      "epoch": 1.4040906578220012,
+      "grad_norm": 0.028978602256717955,
+      "learning_rate": 0.000246664514652245,
+      "loss": 0.9603,
+      "step": 4445
+    },
+    {
+      "epoch": 1.4056700623864802,
+      "grad_norm": 0.028847643444511347,
+      "learning_rate": 0.0002454769188613254,
+      "loss": 0.8939,
+      "step": 4450
+    },
+    {
+      "epoch": 1.4072494669509594,
+      "grad_norm": 0.029573447000154355,
+      "learning_rate": 0.00024429125804029865,
+      "loss": 0.9665,
+      "step": 4455
+    },
+    {
+      "epoch": 1.4088288715154387,
+      "grad_norm": 0.028470803895962127,
+      "learning_rate": 0.00024310754120295596,
+      "loss": 0.9971,
+      "step": 4460
+    },
+    {
+      "epoch": 1.410408276079918,
+      "grad_norm": 0.02826169613197225,
+      "learning_rate": 0.00024192577734831046,
+      "loss": 0.969,
+      "step": 4465
+    },
+    {
+      "epoch": 1.4119876806443972,
+      "grad_norm": 0.031375362200801706,
+      "learning_rate": 0.00024074597546052713,
+      "loss": 0.9437,
+      "step": 4470
+    },
+    {
+      "epoch": 1.4135670852088762,
+      "grad_norm": 0.030043580137543383,
+      "learning_rate": 0.00023956814450885633,
+      "loss": 0.9972,
+      "step": 4475
+    },
+    {
+      "epoch": 1.4151464897733554,
+      "grad_norm": 0.043291478405645165,
+      "learning_rate": 0.00023839229344756418,
+      "loss": 1.0168,
+      "step": 4480
+    },
+    {
+      "epoch": 1.4167258943378347,
+      "grad_norm": 0.029231571062768233,
+      "learning_rate": 0.00023721843121586505,
+      "loss": 0.9198,
+      "step": 4485
+    },
+    {
+      "epoch": 1.4183052989023137,
+      "grad_norm": 0.030135526164777927,
+      "learning_rate": 0.0002360465667378534,
+      "loss": 1.1433,
+      "step": 4490
+    },
+    {
+      "epoch": 1.419884703466793,
+      "grad_norm": 0.034022565203319415,
+      "learning_rate": 0.00023487670892243683,
+      "loss": 1.0713,
+      "step": 4495
+    },
+    {
+      "epoch": 1.4214641080312722,
+      "grad_norm": 0.029417036409679694,
+      "learning_rate": 0.0002337088666632668,
+      "loss": 0.9955,
+      "step": 4500
+    },
+    {
+      "epoch": 1.4230435125957515,
+      "grad_norm": 0.03610772984932751,
+      "learning_rate": 0.00023254304883867205,
+      "loss": 0.9593,
+      "step": 4505
+    },
+    {
+      "epoch": 1.4246229171602307,
+      "grad_norm": 0.031236680351112527,
+      "learning_rate": 0.00023137926431159129,
+      "loss": 0.9978,
+      "step": 4510
+    },
+    {
+      "epoch": 1.4262023217247097,
+      "grad_norm": 0.030707645526992056,
+      "learning_rate": 0.00023021752192950472,
+      "loss": 0.9331,
+      "step": 4515
+    },
+    {
+      "epoch": 1.427781726289189,
+      "grad_norm": 0.03726122575915596,
+      "learning_rate": 0.00022905783052436834,
+      "loss": 1.0044,
+      "step": 4520
+    },
+    {
+      "epoch": 1.4293611308536682,
+      "grad_norm": 0.03361264973926007,
+      "learning_rate": 0.00022790019891254506,
+      "loss": 1.0059,
+      "step": 4525
+    },
+    {
+      "epoch": 1.4309405354181473,
+      "grad_norm": 0.027948864001244987,
+      "learning_rate": 0.00022674463589473926,
+      "loss": 0.906,
+      "step": 4530
+    },
+    {
+      "epoch": 1.4325199399826265,
+      "grad_norm": 0.028728985899873962,
+      "learning_rate": 0.0002255911502559287,
+      "loss": 0.966,
+      "step": 4535
+    },
+    {
+      "epoch": 1.4340993445471057,
+      "grad_norm": 0.02915994123656055,
+      "learning_rate": 0.0002244397507652982,
+      "loss": 0.9845,
+      "step": 4540
+    },
+    {
+      "epoch": 1.435678749111585,
+      "grad_norm": 0.027532167934338617,
+      "learning_rate": 0.00022329044617617355,
+      "loss": 0.9283,
+      "step": 4545
+    },
+    {
+      "epoch": 1.4372581536760642,
+      "grad_norm": 0.02787896461830844,
+      "learning_rate": 0.0002221432452259536,
+      "loss": 0.9426,
+      "step": 4550
+    },
+    {
+      "epoch": 1.4388375582405433,
+      "grad_norm": 0.02915052050382484,
+      "learning_rate": 0.00022099815663604533,
+      "loss": 0.8726,
+      "step": 4555
+    },
+    {
+      "epoch": 1.4404169628050225,
+      "grad_norm": 0.031834082078436965,
+      "learning_rate": 0.00021985518911179624,
+      "loss": 1.0137,
+      "step": 4560
+    },
+    {
+      "epoch": 1.4419963673695018,
+      "grad_norm": 0.03190501236818658,
+      "learning_rate": 0.0002187143513424295,
+      "loss": 0.9506,
+      "step": 4565
+    },
+    {
+      "epoch": 1.4435757719339808,
+      "grad_norm": 0.033915132392950936,
+      "learning_rate": 0.0002175756520009765,
+      "loss": 1.0225,
+      "step": 4570
+    },
+    {
+      "epoch": 1.44515517649846,
+      "grad_norm": 0.027089734040549717,
+      "learning_rate": 0.00021643909974421166,
+      "loss": 0.92,
+      "step": 4575
+    },
+    {
+      "epoch": 1.4467345810629393,
+      "grad_norm": 0.03244887285384843,
+      "learning_rate": 0.0002153047032125871,
+      "loss": 0.9505,
+      "step": 4580
+    },
+    {
+      "epoch": 1.4483139856274185,
+      "grad_norm": 0.034126797735564554,
+      "learning_rate": 0.00021417247103016563,
+      "loss": 0.9887,
+      "step": 4585
+    },
+    {
+      "epoch": 1.4498933901918978,
+      "grad_norm": 0.03479014165453759,
+      "learning_rate": 0.00021304241180455675,
+      "loss": 0.9509,
+      "step": 4590
+    },
+    {
+      "epoch": 1.4514727947563768,
+      "grad_norm": 0.03606094910601667,
+      "learning_rate": 0.0002119145341268497,
+      "loss": 0.9849,
+      "step": 4595
+    },
+    {
+      "epoch": 1.453052199320856,
+      "grad_norm": 0.033649259298392964,
+      "learning_rate": 0.00021078884657154922,
+      "loss": 0.9333,
+      "step": 4600
+    },
+    {
+      "epoch": 1.4546316038853353,
+      "grad_norm": 0.033304936568712216,
+      "learning_rate": 0.0002096653576965098,
+      "loss": 1.0053,
+      "step": 4605
+    },
+    {
+      "epoch": 1.4562110084498143,
+      "grad_norm": 0.03169005737288277,
+      "learning_rate": 0.00020854407604287123,
+      "loss": 0.9485,
+      "step": 4610
+    },
+    {
+      "epoch": 1.4577904130142936,
+      "grad_norm": 0.03260405386554439,
+      "learning_rate": 0.0002074250101349927,
+      "loss": 0.8876,
+      "step": 4615
+    },
+    {
+      "epoch": 1.4593698175787728,
+      "grad_norm": 0.02873933047741036,
+      "learning_rate": 0.0002063081684803892,
+      "loss": 0.9522,
+      "step": 4620
+    },
+    {
+      "epoch": 1.460949222143252,
+      "grad_norm": 0.038484761951163585,
+      "learning_rate": 0.00020519355956966567,
+      "loss": 1.0253,
+      "step": 4625
+    },
+    {
+      "epoch": 1.4625286267077313,
+      "grad_norm": 0.02998983242693224,
+      "learning_rate": 0.000204081191876453,
+      "loss": 0.9557,
+      "step": 4630
+    },
+    {
+      "epoch": 1.4641080312722103,
+      "grad_norm": 0.0343455817655151,
+      "learning_rate": 0.0002029710738573441,
+      "loss": 0.9206,
+      "step": 4635
+    },
+    {
+      "epoch": 1.4656874358366896,
+      "grad_norm": 0.030228779530593563,
+      "learning_rate": 0.00020186321395182838,
+      "loss": 0.9486,
+      "step": 4640
+    },
+    {
+      "epoch": 1.4672668404011688,
+      "grad_norm": 0.030423290706468993,
+      "learning_rate": 0.00020075762058222914,
+      "loss": 0.9493,
+      "step": 4645
+    },
+    {
+      "epoch": 1.4688462449656479,
+      "grad_norm": 0.029878637950206986,
+      "learning_rate": 0.00019965430215363779,
+      "loss": 1.0168,
+      "step": 4650
+    },
+    {
+      "epoch": 1.470425649530127,
+      "grad_norm": 0.03965356236244884,
+      "learning_rate": 0.00019855326705385174,
+      "loss": 1.0224,
+      "step": 4655
+    },
+    {
+      "epoch": 1.4720050540946064,
+      "grad_norm": 0.03325347822092596,
+      "learning_rate": 0.00019745452365330923,
+      "loss": 0.9515,
+      "step": 4660
+    },
+    {
+      "epoch": 1.4735844586590856,
+      "grad_norm": 0.032050231462300495,
+      "learning_rate": 0.00019635808030502616,
+      "loss": 0.9337,
+      "step": 4665
+    },
+    {
+      "epoch": 1.4751638632235649,
+      "grad_norm": 0.03942197951048198,
+      "learning_rate": 0.00019526394534453328,
+      "loss": 0.9647,
+      "step": 4670
+    },
+    {
+      "epoch": 1.4767432677880439,
+      "grad_norm": 0.03706566932643626,
+      "learning_rate": 0.00019417212708981146,
+      "loss": 0.9601,
+      "step": 4675
+    },
+    {
+      "epoch": 1.4783226723525231,
+      "grad_norm": 0.031943909303215,
+      "learning_rate": 0.00019308263384122987,
+      "loss": 0.9851,
+      "step": 4680
+    },
+    {
+      "epoch": 1.4799020769170022,
+      "grad_norm": 0.030359878975780616,
+      "learning_rate": 0.00019199547388148148,
+      "loss": 0.9544,
+      "step": 4685
+    },
+    {
+      "epoch": 1.4814814814814814,
+      "grad_norm": 0.03122343055875806,
+      "learning_rate": 0.0001909106554755216,
+      "loss": 0.983,
+      "step": 4690
+    },
+    {
+      "epoch": 1.4830608860459606,
+      "grad_norm": 0.03029868467794547,
+      "learning_rate": 0.00018982818687050368,
+      "loss": 0.9617,
+      "step": 4695
+    },
+    {
+      "epoch": 1.48464029061044,
+      "grad_norm": 0.029666257350296574,
+      "learning_rate": 0.00018874807629571722,
+      "loss": 0.9558,
+      "step": 4700
+    },
+    {
+      "epoch": 1.4862196951749191,
+      "grad_norm": 0.03656648047904474,
+      "learning_rate": 0.0001876703319625257,
+      "loss": 0.9576,
+      "step": 4705
+    },
+    {
+      "epoch": 1.4877990997393982,
+      "grad_norm": 0.028585469241267418,
+      "learning_rate": 0.00018659496206430303,
+      "loss": 0.9123,
+      "step": 4710
+    },
+    {
+      "epoch": 1.4893785043038774,
+      "grad_norm": 0.03493954328106872,
+      "learning_rate": 0.0001855219747763723,
+      "loss": 0.9263,
+      "step": 4715
+    },
+    {
+      "epoch": 1.4909579088683567,
+      "grad_norm": 0.030950110520780533,
+      "learning_rate": 0.0001844513782559426,
+      "loss": 0.9996,
+      "step": 4720
+    },
+    {
+      "epoch": 1.4925373134328357,
+      "grad_norm": 0.029563002599243136,
+      "learning_rate": 0.00018338318064204856,
+      "loss": 0.9425,
+      "step": 4725
+    },
+    {
+      "epoch": 1.494116717997315,
+      "grad_norm": 0.02886888207876382,
+      "learning_rate": 0.00018231739005548654,
+      "loss": 0.9078,
+      "step": 4730
+    },
+    {
+      "epoch": 1.4956961225617942,
+      "grad_norm": 0.030011112415565738,
+      "learning_rate": 0.00018125401459875474,
+      "loss": 0.9273,
+      "step": 4735
+    },
+    {
+      "epoch": 1.4972755271262734,
+      "grad_norm": 0.027887955051654714,
+      "learning_rate": 0.00018019306235598983,
+      "loss": 0.9781,
+      "step": 4740
+    },
+    {
+      "epoch": 1.4988549316907527,
+      "grad_norm": 0.03007756668275834,
+      "learning_rate": 0.0001791345413929073,
+      "loss": 1.0021,
+      "step": 4745
+    },
+    {
+      "epoch": 1.500434336255232,
+      "grad_norm": 0.03740847640057817,
+      "learning_rate": 0.0001780784597567386,
+      "loss": 0.9867,
+      "step": 4750
+    },
+    {
+      "epoch": 1.502013740819711,
+      "grad_norm": 0.02990805205321503,
+      "learning_rate": 0.00017702482547617067,
+      "loss": 0.9218,
+      "step": 4755
+    },
+    {
+      "epoch": 1.5035931453841902,
+      "grad_norm": 0.03897827731571156,
+      "learning_rate": 0.00017597364656128517,
+      "loss": 1.0062,
+      "step": 4760
+    },
+    {
+      "epoch": 1.5051725499486692,
+      "grad_norm": 0.027694122754996994,
+      "learning_rate": 0.0001749249310034969,
+      "loss": 0.9356,
+      "step": 4765
+    },
+    {
+      "epoch": 1.5067519545131485,
+      "grad_norm": 0.033195432920009905,
+      "learning_rate": 0.00017387868677549368,
+      "loss": 0.9328,
+      "step": 4770
+    },
+    {
+      "epoch": 1.5083313590776277,
+      "grad_norm": 0.031499993107616385,
+      "learning_rate": 0.0001728349218311751,
+      "loss": 0.9293,
+      "step": 4775
+    },
+    {
+      "epoch": 1.509910763642107,
+      "grad_norm": 0.029061947708505044,
+      "learning_rate": 0.00017179364410559284,
+      "loss": 0.9617,
+      "step": 4780
+    },
+    {
+      "epoch": 1.5114901682065862,
+      "grad_norm": 0.028165272129141487,
+      "learning_rate": 0.00017075486151488955,
+      "loss": 0.9074,
+      "step": 4785
+    },
+    {
+      "epoch": 1.5130695727710655,
+      "grad_norm": 0.02758690423618732,
+      "learning_rate": 0.00016971858195623897,
+      "loss": 0.9602,
+      "step": 4790
+    },
+    {
+      "epoch": 1.5146489773355445,
+      "grad_norm": 0.03188469784355415,
+      "learning_rate": 0.00016868481330778646,
+      "loss": 0.9115,
+      "step": 4795
+    },
+    {
+      "epoch": 1.5162283819000237,
+      "grad_norm": 0.028648464031357315,
+      "learning_rate": 0.00016765356342858794,
+      "loss": 0.874,
+      "step": 4800
+    },
+    {
+      "epoch": 1.5178077864645028,
+      "grad_norm": 0.02934106811153413,
+      "learning_rate": 0.00016662484015855152,
+      "loss": 0.9345,
+      "step": 4805
+    },
+    {
+      "epoch": 1.519387191028982,
+      "grad_norm": 0.03524406400581884,
+      "learning_rate": 0.0001655986513183763,
+      "loss": 0.9732,
+      "step": 4810
+    },
+    {
+      "epoch": 1.5209665955934613,
+      "grad_norm": 0.030630658877105973,
+      "learning_rate": 0.00016457500470949476,
+      "loss": 0.9414,
+      "step": 4815
+    },
+    {
+      "epoch": 1.5225460001579405,
+      "grad_norm": 0.03469678327082884,
+      "learning_rate": 0.00016355390811401176,
+      "loss": 0.9199,
+      "step": 4820
+    },
+    {
+      "epoch": 1.5241254047224198,
+      "grad_norm": 0.030372740472785865,
+      "learning_rate": 0.0001625353692946464,
+      "loss": 0.9456,
+      "step": 4825
+    },
+    {
+      "epoch": 1.525704809286899,
+      "grad_norm": 0.031793028072404264,
+      "learning_rate": 0.00016151939599467246,
+      "loss": 0.9958,
+      "step": 4830
+    },
+    {
+      "epoch": 1.527284213851378,
+      "grad_norm": 0.0310066550568518,
+      "learning_rate": 0.0001605059959378603,
+      "loss": 0.9201,
+      "step": 4835
+    },
+    {
+      "epoch": 1.5288636184158573,
+      "grad_norm": 0.029379841977030172,
+      "learning_rate": 0.00015949517682841712,
+      "loss": 0.9616,
+      "step": 4840
+    },
+    {
+      "epoch": 1.5304430229803363,
+      "grad_norm": 0.033451450559209034,
+      "learning_rate": 0.00015848694635092896,
+      "loss": 0.9249,
+      "step": 4845
+    },
+    {
+      "epoch": 1.5320224275448155,
+      "grad_norm": 0.031728724996236635,
+      "learning_rate": 0.00015748131217030258,
+      "loss": 0.9942,
+      "step": 4850
+    },
+    {
+      "epoch": 1.5336018321092948,
+      "grad_norm": 0.03353294850579401,
+      "learning_rate": 0.00015647828193170632,
+      "loss": 0.9606,
+      "step": 4855
+    },
+    {
+      "epoch": 1.535181236673774,
+      "grad_norm": 0.029190786275313668,
+      "learning_rate": 0.00015547786326051293,
+      "loss": 0.9939,
+      "step": 4860
+    },
+    {
+      "epoch": 1.5367606412382533,
+      "grad_norm": 0.02937275253640311,
+      "learning_rate": 0.00015448006376224066,
+      "loss": 0.9728,
+      "step": 4865
+    },
+    {
+      "epoch": 1.5383400458027325,
+      "grad_norm": 0.02972987748663535,
+      "learning_rate": 0.00015348489102249657,
+      "loss": 0.9799,
+      "step": 4870
+    },
+    {
+      "epoch": 1.5399194503672116,
+      "grad_norm": 0.028079108417540122,
+      "learning_rate": 0.00015249235260691763,
+      "loss": 0.9692,
+      "step": 4875
+    },
+    {
+      "epoch": 1.5414988549316906,
+      "grad_norm": 0.027471739413706998,
+      "learning_rate": 0.0001515024560611139,
+      "loss": 0.9209,
+      "step": 4880
+    },
+    {
+      "epoch": 1.5430782594961698,
+      "grad_norm": 0.028941544868934965,
+      "learning_rate": 0.00015051520891061143,
+      "loss": 0.9596,
+      "step": 4885
+    },
+    {
+      "epoch": 1.544657664060649,
+      "grad_norm": 0.02723478992768709,
+      "learning_rate": 0.0001495306186607942,
+      "loss": 0.9286,
+      "step": 4890
+    },
+    {
+      "epoch": 1.5462370686251283,
+      "grad_norm": 0.028916965276510317,
+      "learning_rate": 0.00014854869279684808,
+      "loss": 0.9878,
+      "step": 4895
+    },
+    {
+      "epoch": 1.5478164731896076,
+      "grad_norm": 0.030312213406193604,
+      "learning_rate": 0.00014756943878370266,
+      "loss": 0.9618,
+      "step": 4900
+    },
+    {
+      "epoch": 1.5493958777540868,
+      "grad_norm": 0.03473525605902538,
+      "learning_rate": 0.000146592864065976,
+      "loss": 0.9115,
+      "step": 4905
+    },
+    {
+      "epoch": 1.5509752823185659,
+      "grad_norm": 0.031870739232060324,
+      "learning_rate": 0.00014561897606791673,
+      "loss": 0.9396,
+      "step": 4910
+    },
+    {
+      "epoch": 1.552554686883045,
+      "grad_norm": 0.03690047717608636,
+      "learning_rate": 0.00014464778219334812,
+      "loss": 1.0796,
+      "step": 4915
+    },
+    {
+      "epoch": 1.5541340914475241,
+      "grad_norm": 0.03234558759055333,
+      "learning_rate": 0.00014367928982561234,
+      "loss": 0.96,
+      "step": 4920
+    },
+    {
+      "epoch": 1.5557134960120034,
+      "grad_norm": 0.031191637794967527,
+      "learning_rate": 0.00014271350632751313,
+      "loss": 0.9271,
+      "step": 4925
+    },
+    {
+      "epoch": 1.5572929005764826,
+      "grad_norm": 0.03077065665554426,
+      "learning_rate": 0.00014175043904126117,
+      "loss": 0.9352,
+      "step": 4930
+    },
+    {
+      "epoch": 1.5588723051409619,
+      "grad_norm": 0.030775870100129745,
+      "learning_rate": 0.00014079009528841668,
+      "loss": 1.0264,
+      "step": 4935
+    },
+    {
+      "epoch": 1.5604517097054411,
+      "grad_norm": 0.02898000438139762,
+      "learning_rate": 0.0001398324823698357,
+      "loss": 0.9509,
+      "step": 4940
+    },
+    {
+      "epoch": 1.5620311142699204,
+      "grad_norm": 0.027784668808561445,
+      "learning_rate": 0.00013887760756561268,
+      "loss": 1.0466,
+      "step": 4945
+    },
+    {
+      "epoch": 1.5636105188343994,
+      "grad_norm": 0.03145514968781478,
+      "learning_rate": 0.00013792547813502675,
+      "loss": 0.9426,
+      "step": 4950
+    },
+    {
+      "epoch": 1.5651899233988786,
+      "grad_norm": 0.031223582023416144,
+      "learning_rate": 0.0001369761013164851,
+      "loss": 0.9717,
+      "step": 4955
+    },
+    {
+      "epoch": 1.5667693279633577,
+      "grad_norm": 0.029446106573280874,
+      "learning_rate": 0.00013602948432746916,
+      "loss": 1.0059,
+      "step": 4960
+    },
+    {
+      "epoch": 1.568348732527837,
+      "grad_norm": 0.03060047149808178,
+      "learning_rate": 0.000135085634364479,
+      "loss": 0.9386,
+      "step": 4965
+    },
+    {
+      "epoch": 1.5699281370923162,
+      "grad_norm": 0.031111881817250466,
+      "learning_rate": 0.00013414455860297865,
+      "loss": 0.969,
+      "step": 4970
+    },
+    {
+      "epoch": 1.5715075416567954,
+      "grad_norm": 0.03352967313819241,
+      "learning_rate": 0.00013320626419734217,
+      "loss": 0.9655,
+      "step": 4975
+    },
+    {
+      "epoch": 1.5730869462212747,
+      "grad_norm": 0.029547593732058695,
+      "learning_rate": 0.00013227075828079832,
+      "loss": 0.9662,
+      "step": 4980
+    },
+    {
+      "epoch": 1.574666350785754,
+      "grad_norm": 0.029508552107430478,
+      "learning_rate": 0.00013133804796537735,
+      "loss": 0.9803,
+      "step": 4985
+    },
+    {
+      "epoch": 1.576245755350233,
+      "grad_norm": 0.02975554398000049,
+      "learning_rate": 0.00013040814034185588,
+      "loss": 1.0374,
+      "step": 4990
+    },
+    {
+      "epoch": 1.5778251599147122,
+      "grad_norm": 0.031578737148424206,
+      "learning_rate": 0.000129481042479704,
+      "loss": 0.9395,
+      "step": 4995
+    },
+    {
+      "epoch": 1.5794045644791912,
+      "grad_norm": 0.029663746491364242,
+      "learning_rate": 0.00012855676142703077,
+      "loss": 0.9928,
+      "step": 5000
+    },
+    {
+      "epoch": 1.5809839690436704,
+      "grad_norm": 0.029549754079616287,
+      "learning_rate": 0.00012763530421053076,
+      "loss": 0.8866,
+      "step": 5005
+    },
+    {
+      "epoch": 1.5825633736081497,
+      "grad_norm": 0.030569386705799836,
+      "learning_rate": 0.0001267166778354314,
+      "loss": 0.9777,
+      "step": 5010
+    },
+    {
+      "epoch": 1.584142778172629,
+      "grad_norm": 0.028048496212648384,
+      "learning_rate": 0.00012580088928543836,
+      "loss": 0.962,
+      "step": 5015
+    },
+    {
+      "epoch": 1.5857221827371082,
+      "grad_norm": 0.030860169017198416,
+      "learning_rate": 0.00012488794552268395,
+      "loss": 0.9141,
+      "step": 5020
+    },
+    {
+      "epoch": 1.5873015873015874,
+      "grad_norm": 0.029711018675052436,
+      "learning_rate": 0.0001239778534876727,
+      "loss": 0.9897,
+      "step": 5025
+    },
+    {
+      "epoch": 1.5888809918660665,
+      "grad_norm": 0.02982786530992063,
+      "learning_rate": 0.00012307062009923005,
+      "loss": 0.9729,
+      "step": 5030
+    },
+    {
+      "epoch": 1.5904603964305457,
+      "grad_norm": 0.029748108496392242,
+      "learning_rate": 0.0001221662522544486,
+      "loss": 0.9373,
+      "step": 5035
+    },
+    {
+      "epoch": 1.5920398009950247,
+      "grad_norm": 0.042065215046750926,
+      "learning_rate": 0.00012126475682863608,
+      "loss": 0.9409,
+      "step": 5040
+    },
+    {
+      "epoch": 1.593619205559504,
+      "grad_norm": 0.029375244036325774,
+      "learning_rate": 0.00012036614067526364,
+      "loss": 0.9917,
+      "step": 5045
+    },
+    {
+      "epoch": 1.5951986101239832,
+      "grad_norm": 0.040792830285248355,
+      "learning_rate": 0.00011947041062591274,
+      "loss": 0.9492,
+      "step": 5050
+    },
+    {
+      "epoch": 1.5967780146884625,
+      "grad_norm": 0.02942938481672079,
+      "learning_rate": 0.0001185775734902238,
+      "loss": 1.0339,
+      "step": 5055
+    },
+    {
+      "epoch": 1.5983574192529417,
+      "grad_norm": 0.027670992950519373,
+      "learning_rate": 0.00011768763605584437,
+      "loss": 0.898,
+      "step": 5060
+    },
+    {
+      "epoch": 1.599936823817421,
+      "grad_norm": 0.029414668742697336,
+      "learning_rate": 0.0001168006050883777,
+      "loss": 0.9601,
+      "step": 5065
+    },
+    {
+      "epoch": 1.6015162283819,
+      "grad_norm": 0.02715641122003609,
+      "learning_rate": 0.0001159164873313307,
+      "loss": 0.9179,
+      "step": 5070
+    },
+    {
+      "epoch": 1.6030956329463792,
+      "grad_norm": 0.0344629215298313,
+      "learning_rate": 0.00011503528950606363,
+      "loss": 0.9857,
+      "step": 5075
+    },
+    {
+      "epoch": 1.6046750375108583,
+      "grad_norm": 0.03305054684378565,
+      "learning_rate": 0.0001141570183117378,
+      "loss": 0.9162,
+      "step": 5080
+    },
+    {
+      "epoch": 1.6062544420753375,
+      "grad_norm": 0.0302924518110162,
+      "learning_rate": 0.00011328168042526594,
+      "loss": 0.9946,
+      "step": 5085
+    },
+    {
+      "epoch": 1.6078338466398168,
+      "grad_norm": 0.0348715109293973,
+      "learning_rate": 0.00011240928250126026,
+      "loss": 0.9149,
+      "step": 5090
+    },
+    {
+      "epoch": 1.609413251204296,
+      "grad_norm": 0.028970342452736657,
+      "learning_rate": 0.00011153983117198252,
+      "loss": 0.9643,
+      "step": 5095
+    },
+    {
+      "epoch": 1.6109926557687753,
+      "grad_norm": 0.02909734940218002,
+      "learning_rate": 0.00011067333304729382,
+      "loss": 0.9972,
+      "step": 5100
+    },
+    {
+      "epoch": 1.6125720603332545,
+      "grad_norm": 0.030720347491084186,
+      "learning_rate": 0.00010980979471460339,
+      "loss": 0.9243,
+      "step": 5105
+    },
+    {
+      "epoch": 1.6141514648977335,
+      "grad_norm": 0.028974249632074664,
+      "learning_rate": 0.0001089492227388199,
+      "loss": 0.9629,
+      "step": 5110
+    },
+    {
+      "epoch": 1.6157308694622128,
+      "grad_norm": 0.03122595729483744,
+      "learning_rate": 0.00010809162366229996,
+      "loss": 1.0212,
+      "step": 5115
+    },
+    {
+      "epoch": 1.6173102740266918,
+      "grad_norm": 0.03136511591330045,
+      "learning_rate": 0.00010723700400479997,
+      "loss": 1.0273,
+      "step": 5120
+    },
+    {
+      "epoch": 1.618889678591171,
+      "grad_norm": 0.02825384699316145,
+      "learning_rate": 0.00010638537026342515,
+      "loss": 0.9214,
+      "step": 5125
+    },
+    {
+      "epoch": 1.6204690831556503,
+      "grad_norm": 0.03165649689077606,
+      "learning_rate": 0.00010553672891258104,
+      "loss": 0.9782,
+      "step": 5130
+    },
+    {
+      "epoch": 1.6220484877201296,
+      "grad_norm": 0.034210643516055385,
+      "learning_rate": 0.00010469108640392422,
+      "loss": 1.0466,
+      "step": 5135
+    },
+    {
+      "epoch": 1.6236278922846088,
+      "grad_norm": 0.026834499698806056,
+      "learning_rate": 0.00010384844916631264,
+      "loss": 0.9678,
+      "step": 5140
+    },
+    {
+      "epoch": 1.625207296849088,
+      "grad_norm": 0.028988043467771108,
+      "learning_rate": 0.00010300882360575775,
+      "loss": 0.9699,
+      "step": 5145
+    },
+    {
+      "epoch": 1.626786701413567,
+      "grad_norm": 0.030870225331222148,
+      "learning_rate": 0.00010217221610537448,
+      "loss": 0.97,
+      "step": 5150
+    },
+    {
+      "epoch": 1.6283661059780463,
+      "grad_norm": 0.0350020851497268,
+      "learning_rate": 0.0001013386330253343,
+      "loss": 1.0223,
+      "step": 5155
+    },
+    {
+      "epoch": 1.6299455105425253,
+      "grad_norm": 0.0281915476142278,
+      "learning_rate": 0.00010050808070281508,
+      "loss": 1.0168,
+      "step": 5160
+    },
+    {
+      "epoch": 1.6315249151070046,
+      "grad_norm": 0.029815543401568854,
+      "learning_rate": 9.968056545195476e-05,
+      "loss": 0.9887,
+      "step": 5165
+    },
+    {
+      "epoch": 1.6331043196714838,
+      "grad_norm": 0.03017266635366265,
+      "learning_rate": 9.88560935638017e-05,
+      "loss": 0.9811,
+      "step": 5170
+    },
+    {
+      "epoch": 1.634683724235963,
+      "grad_norm": 0.02814923543583342,
+      "learning_rate": 9.80346713062682e-05,
+      "loss": 0.9602,
+      "step": 5175
+    },
+    {
+      "epoch": 1.6362631288004423,
+      "grad_norm": 0.027320031708459765,
+      "learning_rate": 9.72163049240819e-05,
+      "loss": 0.9273,
+      "step": 5180
+    },
+    {
+      "epoch": 1.6378425333649216,
+      "grad_norm": 0.029917340873650416,
+      "learning_rate": 9.640100063873852e-05,
+      "loss": 0.9804,
+      "step": 5185
+    },
+    {
+      "epoch": 1.6394219379294006,
+      "grad_norm": 0.027497721634484554,
+      "learning_rate": 9.558876464845517e-05,
+      "loss": 0.9379,
+      "step": 5190
+    },
+    {
+      "epoch": 1.6410013424938799,
+      "grad_norm": 0.029168251943545543,
+      "learning_rate": 9.477960312812217e-05,
+      "loss": 0.9427,
+      "step": 5195
+    },
+    {
+      "epoch": 1.6425807470583589,
+      "grad_norm": 0.029310743945918462,
+      "learning_rate": 9.397352222925737e-05,
+      "loss": 0.935,
+      "step": 5200
+    },
+    {
+      "epoch": 1.6441601516228381,
+      "grad_norm": 0.03490911408333341,
+      "learning_rate": 9.317052807995797e-05,
+      "loss": 0.9037,
+      "step": 5205
+    },
+    {
+      "epoch": 1.6457395561873174,
+      "grad_norm": 0.029412651272320862,
+      "learning_rate": 9.23706267848553e-05,
+      "loss": 1.0066,
+      "step": 5210
+    },
+    {
+      "epoch": 1.6473189607517966,
+      "grad_norm": 0.03265358238275424,
+      "learning_rate": 9.157382442506734e-05,
+      "loss": 0.9727,
+      "step": 5215
+    },
+    {
+      "epoch": 1.6488983653162759,
+      "grad_norm": 0.03207810008519121,
+      "learning_rate": 9.078012705815297e-05,
+      "loss": 0.9691,
+      "step": 5220
+    },
+    {
+      "epoch": 1.6504777698807551,
+      "grad_norm": 0.033924430870274636,
+      "learning_rate": 8.998954071806625e-05,
+      "loss": 0.9386,
+      "step": 5225
+    },
+    {
+      "epoch": 1.6520571744452341,
+      "grad_norm": 0.03025695180847416,
+      "learning_rate": 8.920207141510962e-05,
+      "loss": 0.9494,
+      "step": 5230
+    },
+    {
+      "epoch": 1.6536365790097134,
+      "grad_norm": 0.04335743382466986,
+      "learning_rate": 8.841772513588919e-05,
+      "loss": 0.9915,
+      "step": 5235
+    },
+    {
+      "epoch": 1.6552159835741924,
+      "grad_norm": 0.030378840122880894,
+      "learning_rate": 8.763650784326855e-05,
+      "loss": 0.9172,
+      "step": 5240
+    },
+    {
+      "epoch": 1.6567953881386717,
+      "grad_norm": 0.028274908027186057,
+      "learning_rate": 8.685842547632395e-05,
+      "loss": 0.9718,
+      "step": 5245
+    },
+    {
+      "epoch": 1.658374792703151,
+      "grad_norm": 0.03839721528686618,
+      "learning_rate": 8.608348395029859e-05,
+      "loss": 0.9117,
+      "step": 5250
+    },
+    {
+      "epoch": 1.6599541972676302,
+      "grad_norm": 0.03365555844021039,
+      "learning_rate": 8.531168915655785e-05,
+      "loss": 0.9401,
+      "step": 5255
+    },
+    {
+      "epoch": 1.6615336018321094,
+      "grad_norm": 0.02967917392975241,
+      "learning_rate": 8.454304696254516e-05,
+      "loss": 0.9328,
+      "step": 5260
+    },
+    {
+      "epoch": 1.6631130063965884,
+      "grad_norm": 0.02915381706391397,
+      "learning_rate": 8.377756321173629e-05,
+      "loss": 0.898,
+      "step": 5265
+    },
+    {
+      "epoch": 1.6646924109610677,
+      "grad_norm": 0.02733055819538654,
+      "learning_rate": 8.30152437235957e-05,
+      "loss": 0.9382,
+      "step": 5270
+    },
+    {
+      "epoch": 1.6662718155255467,
+      "grad_norm": 0.03054724576737309,
+      "learning_rate": 8.225609429353187e-05,
+      "loss": 0.9498,
+      "step": 5275
+    },
+    {
+      "epoch": 1.667851220090026,
+      "grad_norm": 0.028099357270467477,
+      "learning_rate": 8.150012069285373e-05,
+      "loss": 0.9599,
+      "step": 5280
+    },
+    {
+      "epoch": 1.6694306246545052,
+      "grad_norm": 0.03324146580650551,
+      "learning_rate": 8.074732866872619e-05,
+      "loss": 0.9326,
+      "step": 5285
+    },
+    {
+      "epoch": 1.6710100292189844,
+      "grad_norm": 0.029211888199556814,
+      "learning_rate": 7.999772394412713e-05,
+      "loss": 0.9251,
+      "step": 5290
+    },
+    {
+      "epoch": 1.6725894337834637,
+      "grad_norm": 0.02838078725555603,
+      "learning_rate": 7.925131221780297e-05,
+      "loss": 0.9026,
+      "step": 5295
+    },
+    {
+      "epoch": 1.674168838347943,
+      "grad_norm": 0.027425274238882927,
+      "learning_rate": 7.85080991642264e-05,
+      "loss": 0.9298,
+      "step": 5300
+    },
+    {
+      "epoch": 1.675748242912422,
+      "grad_norm": 0.031163908927535056,
+      "learning_rate": 7.776809043355254e-05,
+      "loss": 0.9946,
+      "step": 5305
+    },
+    {
+      "epoch": 1.6773276474769012,
+      "grad_norm": 0.030309070448334655,
+      "learning_rate": 7.703129165157586e-05,
+      "loss": 1.0243,
+      "step": 5310
+    },
+    {
+      "epoch": 1.6789070520413802,
+      "grad_norm": 0.031389990741688696,
+      "learning_rate": 7.629770841968837e-05,
+      "loss": 1.0145,
+      "step": 5315
+    },
+    {
+      "epoch": 1.6804864566058595,
+      "grad_norm": 0.032247096608187115,
+      "learning_rate": 7.556734631483564e-05,
+      "loss": 0.9475,
+      "step": 5320
+    },
+    {
+      "epoch": 1.6820658611703387,
+      "grad_norm": 0.03219927524276064,
+      "learning_rate": 7.484021088947591e-05,
+      "loss": 0.9795,
+      "step": 5325
+    },
+    {
+      "epoch": 1.683645265734818,
+      "grad_norm": 0.028086436761979167,
+      "learning_rate": 7.411630767153643e-05,
+      "loss": 0.969,
+      "step": 5330
+    },
+    {
+      "epoch": 1.6852246702992972,
+      "grad_norm": 0.03177602647774136,
+      "learning_rate": 7.339564216437273e-05,
+      "loss": 0.9481,
+      "step": 5335
+    },
+    {
+      "epoch": 1.6868040748637765,
+      "grad_norm": 0.029274911998221066,
+      "learning_rate": 7.267821984672573e-05,
+      "loss": 0.9434,
+      "step": 5340
+    },
+    {
+      "epoch": 1.6883834794282555,
+      "grad_norm": 0.030745705722408884,
+      "learning_rate": 7.196404617268059e-05,
+      "loss": 1.0108,
+      "step": 5345
+    },
+    {
+      "epoch": 1.6899628839927348,
+      "grad_norm": 0.031067297390136,
+      "learning_rate": 7.125312657162547e-05,
+      "loss": 0.9551,
+      "step": 5350
+    },
+    {
+      "epoch": 1.6915422885572138,
+      "grad_norm": 0.02853230768950431,
+      "learning_rate": 7.054546644820964e-05,
+      "loss": 0.9126,
+      "step": 5355
+    },
+    {
+      "epoch": 1.693121693121693,
+      "grad_norm": 0.027851652070922075,
+      "learning_rate": 6.984107118230309e-05,
+      "loss": 0.9397,
+      "step": 5360
+    },
+    {
+      "epoch": 1.6947010976861723,
+      "grad_norm": 0.03409950787348448,
+      "learning_rate": 6.91399461289548e-05,
+      "loss": 0.9741,
+      "step": 5365
+    },
+    {
+      "epoch": 1.6962805022506515,
+      "grad_norm": 0.031246138770397752,
+      "learning_rate": 6.844209661835299e-05,
+      "loss": 0.91,
+      "step": 5370
+    },
+    {
+      "epoch": 1.6978599068151308,
+      "grad_norm": 0.02818147666089724,
+      "learning_rate": 6.774752795578365e-05,
+      "loss": 0.9463,
+      "step": 5375
+    },
+    {
+      "epoch": 1.69943931137961,
+      "grad_norm": 0.02768868505742041,
+      "learning_rate": 6.705624542159123e-05,
+      "loss": 0.9933,
+      "step": 5380
+    },
+    {
+      "epoch": 1.701018715944089,
+      "grad_norm": 0.02933701599491612,
+      "learning_rate": 6.636825427113718e-05,
+      "loss": 0.9009,
+      "step": 5385
+    },
+    {
+      "epoch": 1.7025981205085683,
+      "grad_norm": 0.028010286533507792,
+      "learning_rate": 6.568355973476136e-05,
+      "loss": 0.9082,
+      "step": 5390
+    },
+    {
+      "epoch": 1.7041775250730473,
+      "grad_norm": 0.02810226017668788,
+      "learning_rate": 6.500216701774147e-05,
+      "loss": 0.9339,
+      "step": 5395
+    },
+    {
+      "epoch": 1.7057569296375266,
+      "grad_norm": 0.03158388238396385,
+      "learning_rate": 6.432408130025347e-05,
+      "loss": 0.9326,
+      "step": 5400
+    },
+    {
+      "epoch": 1.7073363342020058,
+      "grad_norm": 0.028939507313212626,
+      "learning_rate": 6.36493077373328e-05,
+      "loss": 0.9599,
+      "step": 5405
+    },
+    {
+      "epoch": 1.708915738766485,
+      "grad_norm": 0.029375924859892183,
+      "learning_rate": 6.297785145883439e-05,
+      "loss": 0.9643,
+      "step": 5410
+    },
+    {
+      "epoch": 1.7104951433309643,
+      "grad_norm": 0.028567698771268964,
+      "learning_rate": 6.230971756939441e-05,
+      "loss": 1.0528,
+      "step": 5415
+    },
+    {
+      "epoch": 1.7120745478954436,
+      "grad_norm": 0.03190272267423509,
+      "learning_rate": 6.164491114839077e-05,
+      "loss": 0.9126,
+      "step": 5420
+    },
+    {
+      "epoch": 1.7136539524599226,
+      "grad_norm": 0.03278810172581454,
+      "learning_rate": 6.098343724990524e-05,
+      "loss": 0.9017,
+      "step": 5425
+    },
+    {
+      "epoch": 1.7152333570244018,
+      "grad_norm": 0.03128291196025068,
+      "learning_rate": 6.032530090268429e-05,
+      "loss": 0.9239,
+      "step": 5430
+    },
+    {
+      "epoch": 1.7168127615888809,
+      "grad_norm": 0.029785540541996565,
+      "learning_rate": 5.967050711010119e-05,
+      "loss": 0.9322,
+      "step": 5435
+    },
+    {
+      "epoch": 1.71839216615336,
+      "grad_norm": 0.030098440572351545,
+      "learning_rate": 5.9019060850118434e-05,
+      "loss": 0.9287,
+      "step": 5440
+    },
+    {
+      "epoch": 1.7199715707178393,
+      "grad_norm": 0.028159321421875484,
+      "learning_rate": 5.837096707524886e-05,
+      "loss": 0.8821,
+      "step": 5445
+    },
+    {
+      "epoch": 1.7215509752823186,
+      "grad_norm": 0.028063721609171772,
+      "learning_rate": 5.772623071251915e-05,
+      "loss": 0.9866,
+      "step": 5450
+    },
+    {
+      "epoch": 1.7231303798467978,
+      "grad_norm": 0.03500911377144775,
+      "learning_rate": 5.7084856663431216e-05,
+      "loss": 1.0065,
+      "step": 5455
+    },
+    {
+      "epoch": 1.724709784411277,
+      "grad_norm": 0.02767605696654335,
+      "learning_rate": 5.644684980392617e-05,
+      "loss": 0.8997,
+      "step": 5460
+    },
+    {
+      "epoch": 1.7262891889757561,
+      "grad_norm": 0.026826975417503198,
+      "learning_rate": 5.5812214984346074e-05,
+      "loss": 0.9658,
+      "step": 5465
+    },
+    {
+      "epoch": 1.7278685935402354,
+      "grad_norm": 0.027611668243687007,
+      "learning_rate": 5.518095702939807e-05,
+      "loss": 0.9327,
+      "step": 5470
+    },
+    {
+      "epoch": 1.7294479981047144,
+      "grad_norm": 0.02754288098514866,
+      "learning_rate": 5.4553080738116826e-05,
+      "loss": 0.9229,
+      "step": 5475
+    },
+    {
+      "epoch": 1.7310274026691936,
+      "grad_norm": 0.027193439151797328,
+      "learning_rate": 5.392859088382856e-05,
+      "loss": 0.8753,
+      "step": 5480
+    },
+    {
+      "epoch": 1.7326068072336729,
+      "grad_norm": 0.027601347646716512,
+      "learning_rate": 5.330749221411507e-05,
+      "loss": 0.9554,
+      "step": 5485
+    },
+    {
+      "epoch": 1.7341862117981521,
+      "grad_norm": 0.03515769678408284,
+      "learning_rate": 5.268978945077668e-05,
+      "loss": 1.0282,
+      "step": 5490
+    },
+    {
+      "epoch": 1.7357656163626314,
+      "grad_norm": 0.02894574402350519,
+      "learning_rate": 5.207548728979716e-05,
+      "loss": 0.9606,
+      "step": 5495
+    },
+    {
+      "epoch": 1.7373450209271106,
+      "grad_norm": 0.027702253930178652,
+      "learning_rate": 5.1464590401307684e-05,
+      "loss": 0.8957,
+      "step": 5500
+    },
+    {
+      "epoch": 1.7389244254915897,
+      "grad_norm": 0.03350890536428408,
+      "learning_rate": 5.085710342955163e-05,
+      "loss": 0.9537,
+      "step": 5505
+    },
+    {
+      "epoch": 1.740503830056069,
+      "grad_norm": 0.03142247327424322,
+      "learning_rate": 5.0253030992848616e-05,
+      "loss": 0.9703,
+      "step": 5510
+    },
+    {
+      "epoch": 1.742083234620548,
+      "grad_norm": 0.030854271335729343,
+      "learning_rate": 4.965237768356029e-05,
+      "loss": 0.8991,
+      "step": 5515
+    },
+    {
+      "epoch": 1.7436626391850272,
+      "grad_norm": 0.02972820758050759,
+      "learning_rate": 4.905514806805456e-05,
+      "loss": 0.9566,
+      "step": 5520
+    },
+    {
+      "epoch": 1.7452420437495064,
+      "grad_norm": 0.027053904669588996,
+      "learning_rate": 4.8461346686671405e-05,
+      "loss": 0.9043,
+      "step": 5525
+    },
+    {
+      "epoch": 1.7468214483139857,
+      "grad_norm": 0.02752918466003326,
+      "learning_rate": 4.787097805368839e-05,
+      "loss": 0.9152,
+      "step": 5530
+    },
+    {
+      "epoch": 1.748400852878465,
+      "grad_norm": 0.027898582750512425,
+      "learning_rate": 4.728404665728586e-05,
+      "loss": 0.8763,
+      "step": 5535
+    },
+    {
+      "epoch": 1.7499802574429442,
+      "grad_norm": 0.03174675042457659,
+      "learning_rate": 4.670055695951342e-05,
+      "loss": 0.9377,
+      "step": 5540
+    },
+    {
+      "epoch": 1.7515596620074232,
+      "grad_norm": 0.02953452769699611,
+      "learning_rate": 4.6120513396255446e-05,
+      "loss": 0.9275,
+      "step": 5545
+    },
+    {
+      "epoch": 1.7531390665719024,
+      "grad_norm": 0.028278131840278015,
+      "learning_rate": 4.554392037719801e-05,
+      "loss": 0.9219,
+      "step": 5550
+    },
+    {
+      "epoch": 1.7547184711363815,
+      "grad_norm": 0.02883058939172368,
+      "learning_rate": 4.4970782285794484e-05,
+      "loss": 0.9637,
+      "step": 5555
+    },
+    {
+      "epoch": 1.7562978757008607,
+      "grad_norm": 0.02892465549363483,
+      "learning_rate": 4.440110347923332e-05,
+      "loss": 0.9655,
+      "step": 5560
+    },
+    {
+      "epoch": 1.75787728026534,
+      "grad_norm": 0.028065875175745866,
+      "learning_rate": 4.383488828840387e-05,
+      "loss": 0.9238,
+      "step": 5565
+    },
+    {
+      "epoch": 1.7594566848298192,
+      "grad_norm": 0.029174737233365906,
+      "learning_rate": 4.327214101786397e-05,
+      "loss": 0.9855,
+      "step": 5570
+    },
+    {
+      "epoch": 1.7610360893942985,
+      "grad_norm": 0.0285435239640047,
+      "learning_rate": 4.271286594580748e-05,
+      "loss": 0.9132,
+      "step": 5575
+    },
+    {
+      "epoch": 1.7626154939587777,
+      "grad_norm": 0.0288796816691122,
+      "learning_rate": 4.215706732403096e-05,
+      "loss": 0.9095,
+      "step": 5580
+    },
+    {
+      "epoch": 1.7641948985232567,
+      "grad_norm": 0.02956368548489994,
+      "learning_rate": 4.160474937790232e-05,
+      "loss": 0.9589,
+      "step": 5585
+    },
+    {
+      "epoch": 1.765774303087736,
+      "grad_norm": 0.028132040846672902,
+      "learning_rate": 4.105591630632777e-05,
+      "loss": 0.9333,
+      "step": 5590
+    },
+    {
+      "epoch": 1.767353707652215,
+      "grad_norm": 0.02957005723683004,
+      "learning_rate": 4.051057228172073e-05,
+      "loss": 0.9258,
+      "step": 5595
+    },
+    {
+      "epoch": 1.7689331122166942,
+      "grad_norm": 0.032839128835579115,
+      "learning_rate": 3.996872144996938e-05,
+      "loss": 0.9229,
+      "step": 5600
+    },
+    {
+      "epoch": 1.7705125167811735,
+      "grad_norm": 0.028340339388521443,
+      "learning_rate": 3.9430367930405666e-05,
+      "loss": 0.9532,
+      "step": 5605
+    },
+    {
+      "epoch": 1.7720919213456527,
+      "grad_norm": 0.02798928468339342,
+      "learning_rate": 3.8895515815773774e-05,
+      "loss": 0.9623,
+      "step": 5610
+    },
+    {
+      "epoch": 1.773671325910132,
+      "grad_norm": 0.027450223264395033,
+      "learning_rate": 3.836416917219881e-05,
+      "loss": 0.9309,
+      "step": 5615
+    },
+    {
+      "epoch": 1.775250730474611,
+      "grad_norm": 0.02861113029376848,
+      "learning_rate": 3.783633203915654e-05,
+      "loss": 0.9085,
+      "step": 5620
+    },
+    {
+      "epoch": 1.7768301350390903,
+      "grad_norm": 0.028077426206203647,
+      "learning_rate": 3.731200842944182e-05,
+      "loss": 0.9115,
+      "step": 5625
+    },
+    {
+      "epoch": 1.7784095396035693,
+      "grad_norm": 0.027683250225642483,
+      "learning_rate": 3.6791202329138965e-05,
+      "loss": 0.8986,
+      "step": 5630
+    },
+    {
+      "epoch": 1.7799889441680485,
+      "grad_norm": 0.028563107486100536,
+      "learning_rate": 3.6273917697590475e-05,
+      "loss": 0.9682,
+      "step": 5635
+    },
+    {
+      "epoch": 1.7815683487325278,
+      "grad_norm": 0.029066045783134917,
+      "learning_rate": 3.576015846736797e-05,
+      "loss": 0.9404,
+      "step": 5640
+    },
+    {
+      "epoch": 1.783147753297007,
+      "grad_norm": 0.031421469836084966,
+      "learning_rate": 3.524992854424147e-05,
+      "loss": 0.94,
+      "step": 5645
+    },
+    {
+      "epoch": 1.7847271578614863,
+      "grad_norm": 0.0290390057106864,
+      "learning_rate": 3.4743231807150056e-05,
+      "loss": 0.9669,
+      "step": 5650
+    },
+    {
+      "epoch": 1.7863065624259655,
+      "grad_norm": 0.027429872792924102,
+      "learning_rate": 3.4240072108172485e-05,
+      "loss": 0.9191,
+      "step": 5655
+    },
+    {
+      "epoch": 1.7878859669904446,
+      "grad_norm": 0.02743182866648419,
+      "learning_rate": 3.3740453272497585e-05,
+      "loss": 0.8912,
+      "step": 5660
+    },
+    {
+      "epoch": 1.7894653715549238,
+      "grad_norm": 0.03289260196623006,
+      "learning_rate": 3.324437909839556e-05,
+      "loss": 0.9597,
+      "step": 5665
+    },
+    {
+      "epoch": 1.7910447761194028,
+      "grad_norm": 0.031067149441916974,
+      "learning_rate": 3.275185335718861e-05,
+      "loss": 0.9613,
+      "step": 5670
+    },
+    {
+      "epoch": 1.792624180683882,
+      "grad_norm": 0.027403840944075043,
+      "learning_rate": 3.226287979322295e-05,
+      "loss": 0.9077,
+      "step": 5675
+    },
+    {
+      "epoch": 1.7942035852483613,
+      "grad_norm": 0.027638003709822466,
+      "learning_rate": 3.177746212383953e-05,
+      "loss": 1.0154,
+      "step": 5680
+    },
+    {
+      "epoch": 1.7957829898128406,
+      "grad_norm": 0.027981137344302076,
+      "learning_rate": 3.1295604039346615e-05,
+      "loss": 0.9223,
+      "step": 5685
+    },
+    {
+      "epoch": 1.7973623943773198,
+      "grad_norm": 0.02869606714386023,
+      "learning_rate": 3.0817309202990916e-05,
+      "loss": 0.9966,
+      "step": 5690
     },
     {
-      "epoch": 1.4285714285714286,
-      "grad_norm": 31.90907042465711,
-      "learning_rate": 9.549150281252633e-05,
-      "loss": 19.9897,
-      "step": 5
+      "epoch": 1.798941798941799,
+      "grad_norm": 0.029474493487004923,
+      "learning_rate": 3.0342581250930368e-05,
+      "loss": 1.0082,
+      "step": 5695
+    },
+    {
+      "epoch": 1.800521203506278,
+      "grad_norm": 0.029469302160620065,
+      "learning_rate": 2.9871423792206252e-05,
+      "loss": 0.9922,
+      "step": 5700
+    },
+    {
+      "epoch": 1.8021006080707573,
+      "grad_norm": 0.030119473606862637,
+      "learning_rate": 2.940384040871563e-05,
+      "loss": 0.9676,
+      "step": 5705
+    },
+    {
+      "epoch": 1.8036800126352364,
+      "grad_norm": 0.03392551374610534,
+      "learning_rate": 2.893983465518446e-05,
+      "loss": 0.9468,
+      "step": 5710
+    },
+    {
+      "epoch": 1.8052594171997156,
+      "grad_norm": 0.028895121987632597,
+      "learning_rate": 2.847941005914012e-05,
+      "loss": 0.9319,
+      "step": 5715
+    },
+    {
+      "epoch": 1.8068388217641949,
+      "grad_norm": 0.03129999681800566,
+      "learning_rate": 2.8022570120884937e-05,
+      "loss": 1.0065,
+      "step": 5720
+    },
+    {
+      "epoch": 1.808418226328674,
+      "grad_norm": 0.027645868323954787,
+      "learning_rate": 2.756931831346937e-05,
+      "loss": 0.9222,
+      "step": 5725
+    },
+    {
+      "epoch": 1.8099976308931534,
+      "grad_norm": 0.029121807025221366,
+      "learning_rate": 2.7119658082666034e-05,
+      "loss": 0.895,
+      "step": 5730
+    },
+    {
+      "epoch": 1.8115770354576326,
+      "grad_norm": 0.02854616322766256,
+      "learning_rate": 2.6673592846942707e-05,
+      "loss": 0.988,
+      "step": 5735
+    },
+    {
+      "epoch": 1.8131564400221116,
+      "grad_norm": 0.030400035928911648,
+      "learning_rate": 2.62311259974371e-05,
+      "loss": 0.9156,
+      "step": 5740
+    },
+    {
+      "epoch": 1.8147358445865909,
+      "grad_norm": 0.02879072839156828,
+      "learning_rate": 2.579226089793074e-05,
+      "loss": 0.9106,
+      "step": 5745
+    },
+    {
+      "epoch": 1.81631524915107,
+      "grad_norm": 0.02891714024162058,
+      "learning_rate": 2.5357000884823344e-05,
+      "loss": 0.982,
+      "step": 5750
+    },
+    {
+      "epoch": 1.8178946537155491,
+      "grad_norm": 0.027645185731986426,
+      "learning_rate": 2.4925349267107765e-05,
+      "loss": 0.9915,
+      "step": 5755
+    },
+    {
+      "epoch": 1.8194740582800284,
+      "grad_norm": 0.028868560331025948,
+      "learning_rate": 2.4497309326344364e-05,
+      "loss": 0.9583,
+      "step": 5760
+    },
+    {
+      "epoch": 1.8210534628445076,
+      "grad_norm": 0.028199422420434998,
+      "learning_rate": 2.4072884316636512e-05,
+      "loss": 0.9967,
+      "step": 5765
+    },
+    {
+      "epoch": 1.822632867408987,
+      "grad_norm": 0.026658531525045798,
+      "learning_rate": 2.3652077464605514e-05,
+      "loss": 0.9488,
+      "step": 5770
+    },
+    {
+      "epoch": 1.8242122719734661,
+      "grad_norm": 0.035756806143134,
+      "learning_rate": 2.32348919693664e-05,
+      "loss": 0.9512,
+      "step": 5775
+    },
+    {
+      "epoch": 1.8257916765379452,
+      "grad_norm": 0.027506596305714767,
+      "learning_rate": 2.2821331002503276e-05,
+      "loss": 0.9161,
+      "step": 5780
+    },
+    {
+      "epoch": 1.8273710811024244,
+      "grad_norm": 0.028991229999471242,
+      "learning_rate": 2.2411397708045346e-05,
+      "loss": 0.8853,
+      "step": 5785
+    },
+    {
+      "epoch": 1.8289504856669034,
+      "grad_norm": 0.02828101088834613,
+      "learning_rate": 2.200509520244326e-05,
+      "loss": 0.9581,
+      "step": 5790
+    },
+    {
+      "epoch": 1.8305298902313827,
+      "grad_norm": 0.08332681448836027,
+      "learning_rate": 2.1602426574544863e-05,
+      "loss": 0.916,
+      "step": 5795
+    },
+    {
+      "epoch": 1.832109294795862,
+      "grad_norm": 0.02904806118099692,
+      "learning_rate": 2.1203394885572436e-05,
+      "loss": 0.9825,
+      "step": 5800
+    },
+    {
+      "epoch": 1.8336886993603412,
+      "grad_norm": 0.027670952342160212,
+      "learning_rate": 2.0808003169098587e-05,
+      "loss": 1.0126,
+      "step": 5805
+    },
+    {
+      "epoch": 1.8352681039248204,
+      "grad_norm": 0.027656961469499166,
+      "learning_rate": 2.0416254431024073e-05,
+      "loss": 0.8975,
+      "step": 5810
+    },
+    {
+      "epoch": 1.8368475084892997,
+      "grad_norm": 0.028413585230628282,
+      "learning_rate": 2.0028151649554126e-05,
+      "loss": 0.9336,
+      "step": 5815
+    },
+    {
+      "epoch": 1.8384269130537787,
+      "grad_norm": 0.027511851975369536,
+      "learning_rate": 1.964369777517644e-05,
+      "loss": 0.9466,
+      "step": 5820
+    },
+    {
+      "epoch": 1.840006317618258,
+      "grad_norm": 0.02815057689234334,
+      "learning_rate": 1.9262895730638387e-05,
+      "loss": 0.9238,
+      "step": 5825
+    },
+    {
+      "epoch": 1.841585722182737,
+      "grad_norm": 0.031429023025303086,
+      "learning_rate": 1.8885748410924884e-05,
+      "loss": 0.9359,
+      "step": 5830
+    },
+    {
+      "epoch": 1.8431651267472162,
+      "grad_norm": 0.027798306464001606,
+      "learning_rate": 1.8512258683236525e-05,
+      "loss": 0.9302,
+      "step": 5835
+    },
+    {
+      "epoch": 1.8447445313116955,
+      "grad_norm": 0.028933369811650223,
+      "learning_rate": 1.8142429386967473e-05,
+      "loss": 0.923,
+      "step": 5840
+    },
+    {
+      "epoch": 1.8463239358761747,
+      "grad_norm": 0.027960711205781297,
+      "learning_rate": 1.7776263333684316e-05,
+      "loss": 0.9287,
+      "step": 5845
+    },
+    {
+      "epoch": 1.847903340440654,
+      "grad_norm": 0.026802432992602397,
+      "learning_rate": 1.7413763307104092e-05,
+      "loss": 0.9873,
+      "step": 5850
+    },
+    {
+      "epoch": 1.8494827450051332,
+      "grad_norm": 0.030407224531095515,
+      "learning_rate": 1.70549320630739e-05,
+      "loss": 0.9345,
+      "step": 5855
+    },
+    {
+      "epoch": 1.8510621495696122,
+      "grad_norm": 0.03016268684640237,
+      "learning_rate": 1.669977232954911e-05,
+      "loss": 0.9718,
+      "step": 5860
+    },
+    {
+      "epoch": 1.8526415541340915,
+      "grad_norm": 0.029031201241125883,
+      "learning_rate": 1.6348286806573354e-05,
+      "loss": 0.9148,
+      "step": 5865
+    },
+    {
+      "epoch": 1.8542209586985705,
+      "grad_norm": 0.026706981294734724,
+      "learning_rate": 1.6000478166257494e-05,
+      "loss": 0.9294,
+      "step": 5870
+    },
+    {
+      "epoch": 1.8558003632630498,
+      "grad_norm": 0.027765072457612058,
+      "learning_rate": 1.5656349052759533e-05,
+      "loss": 1.0392,
+      "step": 5875
+    },
+    {
+      "epoch": 1.857379767827529,
+      "grad_norm": 0.02930140524141711,
+      "learning_rate": 1.5315902082264577e-05,
+      "loss": 1.0084,
+      "step": 5880
+    },
+    {
+      "epoch": 1.8589591723920083,
+      "grad_norm": 0.02615279031598757,
+      "learning_rate": 1.4979139842964674e-05,
+      "loss": 0.9409,
+      "step": 5885
+    },
+    {
+      "epoch": 1.8605385769564875,
+      "grad_norm": 0.027246642272342872,
+      "learning_rate": 1.4646064895039502e-05,
+      "loss": 0.8993,
+      "step": 5890
+    },
+    {
+      "epoch": 1.8621179815209667,
+      "grad_norm": 0.030922173080159505,
+      "learning_rate": 1.4316679770636498e-05,
+      "loss": 0.9891,
+      "step": 5895
+    },
+    {
+      "epoch": 1.8636973860854458,
+      "grad_norm": 0.029229489681193894,
+      "learning_rate": 1.3990986973852039e-05,
+      "loss": 0.9553,
+      "step": 5900
+    },
+    {
+      "epoch": 1.865276790649925,
+      "grad_norm": 0.02719994991525794,
+      "learning_rate": 1.3668988980712005e-05,
+      "loss": 0.9284,
+      "step": 5905
+    },
+    {
+      "epoch": 1.866856195214404,
+      "grad_norm": 0.02866064080644265,
+      "learning_rate": 1.3350688239153196e-05,
+      "loss": 0.9299,
+      "step": 5910
+    },
+    {
+      "epoch": 1.8684355997788833,
+      "grad_norm": 0.06285937243773189,
+      "learning_rate": 1.303608716900484e-05,
+      "loss": 1.0569,
+      "step": 5915
+    },
+    {
+      "epoch": 1.8700150043433625,
+      "grad_norm": 0.02930390176914617,
+      "learning_rate": 1.2725188161969659e-05,
+      "loss": 0.9373,
+      "step": 5920
+    },
+    {
+      "epoch": 1.8715944089078418,
+      "grad_norm": 0.027319721436793096,
+      "learning_rate": 1.2417993581606447e-05,
+      "loss": 0.9802,
+      "step": 5925
+    },
+    {
+      "epoch": 1.873173813472321,
+      "grad_norm": 0.027699411481698355,
+      "learning_rate": 1.2114505763311356e-05,
+      "loss": 0.9254,
+      "step": 5930
+    },
+    {
+      "epoch": 1.8747532180368003,
+      "grad_norm": 0.027341193658824932,
+      "learning_rate": 1.1814727014300807e-05,
+      "loss": 0.8946,
+      "step": 5935
+    },
+    {
+      "epoch": 1.8763326226012793,
+      "grad_norm": 0.02829423795989287,
+      "learning_rate": 1.151865961359333e-05,
+      "loss": 0.9129,
+      "step": 5940
+    },
+    {
+      "epoch": 1.8779120271657586,
+      "grad_norm": 0.02724668316770682,
+      "learning_rate": 1.1226305811992743e-05,
+      "loss": 0.9241,
+      "step": 5945
+    },
+    {
+      "epoch": 1.8794914317302376,
+      "grad_norm": 0.027357269499672484,
+      "learning_rate": 1.093766783207073e-05,
+      "loss": 1.029,
+      "step": 5950
+    },
+    {
+      "epoch": 1.8810708362947168,
+      "grad_norm": 0.02737662342078075,
+      "learning_rate": 1.0652747868150125e-05,
+      "loss": 0.9155,
+      "step": 5955
+    },
+    {
+      "epoch": 1.882650240859196,
+      "grad_norm": 0.02909080740490778,
+      "learning_rate": 1.0371548086288207e-05,
+      "loss": 0.9197,
+      "step": 5960
+    },
+    {
+      "epoch": 1.8842296454236753,
+      "grad_norm": 0.02851471169073095,
+      "learning_rate": 1.0094070624259877e-05,
+      "loss": 0.9217,
+      "step": 5965
+    },
+    {
+      "epoch": 1.8858090499881546,
+      "grad_norm": 0.030988378240067956,
+      "learning_rate": 9.820317591542172e-06,
+      "loss": 0.9359,
+      "step": 5970
+    },
+    {
+      "epoch": 1.8873884545526338,
+      "grad_norm": 0.03114149770416287,
+      "learning_rate": 9.550291069297445e-06,
+      "loss": 0.9198,
+      "step": 5975
+    },
+    {
+      "epoch": 1.8889678591171128,
+      "grad_norm": 0.03499816306156768,
+      "learning_rate": 9.283993110357936e-06,
+      "loss": 0.9136,
+      "step": 5980
+    },
+    {
+      "epoch": 1.890547263681592,
+      "grad_norm": 0.028861934812947413,
+      "learning_rate": 9.021425739210054e-06,
+      "loss": 1.0039,
+      "step": 5985
+    },
+    {
+      "epoch": 1.8921266682460711,
+      "grad_norm": 0.029981362197630507,
+      "learning_rate": 8.762590951979232e-06,
+      "loss": 0.9442,
+      "step": 5990
+    },
+    {
+      "epoch": 1.8937060728105504,
+      "grad_norm": 0.027640443348192135,
+      "learning_rate": 8.507490716414269e-06,
+      "loss": 0.878,
+      "step": 5995
+    },
+    {
+      "epoch": 1.8952854773750296,
+      "grad_norm": 0.027628387408123886,
+      "learning_rate": 8.256126971872835e-06,
+      "loss": 0.9266,
+      "step": 6000
+    },
+    {
+      "epoch": 1.8968648819395089,
+      "grad_norm": 0.028137456844671317,
+      "learning_rate": 8.008501629306497e-06,
+      "loss": 0.9445,
+      "step": 6005
+    },
+    {
+      "epoch": 1.898444286503988,
+      "grad_norm": 0.030635181924957434,
+      "learning_rate": 7.764616571246218e-06,
+      "loss": 0.9492,
+      "step": 6010
+    },
+    {
+      "epoch": 1.9000236910684671,
+      "grad_norm": 0.03022552040333277,
+      "learning_rate": 7.524473651788044e-06,
+      "loss": 0.9865,
+      "step": 6015
+    },
+    {
+      "epoch": 1.9016030956329464,
+      "grad_norm": 0.027532356125552514,
+      "learning_rate": 7.288074696578995e-06,
+      "loss": 0.9205,
+      "step": 6020
+    },
+    {
+      "epoch": 1.9031825001974254,
+      "grad_norm": 0.03037252231239617,
+      "learning_rate": 7.055421502803416e-06,
+      "loss": 1.0084,
+      "step": 6025
+    },
+    {
+      "epoch": 1.9047619047619047,
+      "grad_norm": 0.029191841877876303,
+      "learning_rate": 6.826515839168934e-06,
+      "loss": 0.9652,
+      "step": 6030
+    },
+    {
+      "epoch": 1.906341309326384,
+      "grad_norm": 0.03132558846523111,
+      "learning_rate": 6.6013594458931295e-06,
+      "loss": 0.9523,
+      "step": 6035
+    },
+    {
+      "epoch": 1.9079207138908632,
+      "grad_norm": 0.02818847729664874,
+      "learning_rate": 6.379954034690605e-06,
+      "loss": 0.9847,
+      "step": 6040
+    },
+    {
+      "epoch": 1.9095001184553424,
+      "grad_norm": 0.028021954410402724,
+      "learning_rate": 6.162301288759498e-06,
+      "loss": 0.9539,
+      "step": 6045
+    },
+    {
+      "epoch": 1.9110795230198216,
+      "grad_norm": 0.029421920413576684,
+      "learning_rate": 5.9484028627692085e-06,
+      "loss": 0.9986,
+      "step": 6050
+    },
+    {
+      "epoch": 1.9126589275843007,
+      "grad_norm": 0.026853037333214733,
+      "learning_rate": 5.738260382847193e-06,
+      "loss": 0.939,
+      "step": 6055
+    },
+    {
+      "epoch": 1.91423833214878,
+      "grad_norm": 0.02934335338064871,
+      "learning_rate": 5.531875446567136e-06,
+      "loss": 0.9809,
+      "step": 6060
+    },
+    {
+      "epoch": 1.915817736713259,
+      "grad_norm": 0.02949182368243745,
+      "learning_rate": 5.3292496229366824e-06,
+      "loss": 1.0094,
+      "step": 6065
+    },
+    {
+      "epoch": 1.9173971412777382,
+      "grad_norm": 0.028757305323247218,
+      "learning_rate": 5.130384452385339e-06,
+      "loss": 0.9217,
+      "step": 6070
+    },
+    {
+      "epoch": 1.9189765458422174,
+      "grad_norm": 0.027371319131601297,
+      "learning_rate": 4.93528144675276e-06,
+      "loss": 0.8874,
+      "step": 6075
+    },
+    {
+      "epoch": 1.9205559504066967,
+      "grad_norm": 0.02981376355537962,
+      "learning_rate": 4.743942089277642e-06,
+      "loss": 1.0042,
+      "step": 6080
+    },
+    {
+      "epoch": 1.922135354971176,
+      "grad_norm": 0.027760082391345534,
+      "learning_rate": 4.556367834585961e-06,
+      "loss": 0.941,
+      "step": 6085
+    },
+    {
+      "epoch": 1.9237147595356552,
+      "grad_norm": 0.0279524329788298,
+      "learning_rate": 4.3725601086800345e-06,
+      "loss": 0.9074,
+      "step": 6090
+    },
+    {
+      "epoch": 1.9252941641001342,
+      "grad_norm": 0.02941844183259332,
+      "learning_rate": 4.192520308928083e-06,
+      "loss": 0.9607,
+      "step": 6095
+    },
+    {
+      "epoch": 1.9268735686646135,
+      "grad_norm": 0.028310654875225674,
+      "learning_rate": 4.016249804052907e-06,
+      "loss": 0.9221,
+      "step": 6100
+    },
+    {
+      "epoch": 1.9284529732290925,
+      "grad_norm": 0.02790363498934764,
+      "learning_rate": 3.843749934122231e-06,
+      "loss": 0.9637,
+      "step": 6105
+    },
+    {
+      "epoch": 1.9300323777935717,
+      "grad_norm": 0.02883951143907877,
+      "learning_rate": 3.6750220105378206e-06,
+      "loss": 0.926,
+      "step": 6110
+    },
+    {
+      "epoch": 1.931611782358051,
+      "grad_norm": 0.029616968182773224,
+      "learning_rate": 3.5100673160260442e-06,
+      "loss": 0.9938,
+      "step": 6115
+    },
+    {
+      "epoch": 1.9331911869225302,
+      "grad_norm": 0.03017986090096489,
+      "learning_rate": 3.3488871046278844e-06,
+      "loss": 0.9115,
+      "step": 6120
+    },
+    {
+      "epoch": 1.9347705914870095,
+      "grad_norm": 0.03190498090763136,
+      "learning_rate": 3.191482601689333e-06,
+      "loss": 0.9357,
+      "step": 6125
+    },
+    {
+      "epoch": 1.9363499960514887,
+      "grad_norm": 0.029102205438411393,
+      "learning_rate": 3.0378550038522855e-06,
+      "loss": 0.9068,
+      "step": 6130
+    },
+    {
+      "epoch": 1.9379294006159677,
+      "grad_norm": 0.02897397063321125,
+      "learning_rate": 2.8880054790453304e-06,
+      "loss": 0.9336,
+      "step": 6135
+    },
+    {
+      "epoch": 1.939508805180447,
+      "grad_norm": 0.029670071722107515,
+      "learning_rate": 2.741935166474807e-06,
+      "loss": 0.9588,
+      "step": 6140
+    },
+    {
+      "epoch": 1.941088209744926,
+      "grad_norm": 0.032069141806492965,
+      "learning_rate": 2.5996451766163165e-06,
+      "loss": 1.0055,
+      "step": 6145
+    },
+    {
+      "epoch": 1.9426676143094053,
+      "grad_norm": 0.028650482223296685,
+      "learning_rate": 2.4611365912061143e-06,
+      "loss": 0.9142,
+      "step": 6150
+    },
+    {
+      "epoch": 1.9442470188738845,
+      "grad_norm": 0.02895708267084703,
+      "learning_rate": 2.3264104632328974e-06,
+      "loss": 0.9812,
+      "step": 6155
+    },
+    {
+      "epoch": 1.9458264234383638,
+      "grad_norm": 0.027893607901280315,
+      "learning_rate": 2.1954678169299745e-06,
+      "loss": 0.9621,
+      "step": 6160
+    },
+    {
+      "epoch": 1.947405828002843,
+      "grad_norm": 0.025988777604333536,
+      "learning_rate": 2.0683096477672747e-06,
+      "loss": 0.9891,
+      "step": 6165
+    },
+    {
+      "epoch": 1.9489852325673223,
+      "grad_norm": 0.02763337859997564,
+      "learning_rate": 1.9449369224438517e-06,
+      "loss": 0.9667,
+      "step": 6170
+    },
+    {
+      "epoch": 1.9505646371318013,
+      "grad_norm": 0.02684017287977009,
+      "learning_rate": 1.8253505788806136e-06,
+      "loss": 0.9869,
+      "step": 6175
+    },
+    {
+      "epoch": 1.9521440416962805,
+      "grad_norm": 0.03379619360602131,
+      "learning_rate": 1.7095515262129935e-06,
+      "loss": 0.9331,
+      "step": 6180
+    },
+    {
+      "epoch": 1.9537234462607596,
+      "grad_norm": 0.02904286384558498,
+      "learning_rate": 1.597540644784401e-06,
+      "loss": 0.9409,
+      "step": 6185
+    },
+    {
+      "epoch": 1.9553028508252388,
+      "grad_norm": 0.028873506388812257,
+      "learning_rate": 1.48931878613906e-06,
+      "loss": 0.8905,
+      "step": 6190
+    },
+    {
+      "epoch": 1.956882255389718,
+      "grad_norm": 0.027643832216016774,
+      "learning_rate": 1.3848867730158476e-06,
+      "loss": 0.9781,
+      "step": 6195
+    },
+    {
+      "epoch": 1.9584616599541973,
+      "grad_norm": 0.02723841917931141,
+      "learning_rate": 1.2842453993420765e-06,
+      "loss": 0.9066,
+      "step": 6200
+    },
+    {
+      "epoch": 1.9600410645186765,
+      "grad_norm": 0.027939513974217022,
+      "learning_rate": 1.1873954302271118e-06,
+      "loss": 0.9705,
+      "step": 6205
+    },
+    {
+      "epoch": 1.9616204690831558,
+      "grad_norm": 0.027685289739808893,
+      "learning_rate": 1.0943376019570962e-06,
+      "loss": 0.9548,
+      "step": 6210
+    },
+    {
+      "epoch": 1.9631998736476348,
+      "grad_norm": 0.031984172346232166,
+      "learning_rate": 1.0050726219886785e-06,
+      "loss": 0.9961,
+      "step": 6215
+    },
+    {
+      "epoch": 1.964779278212114,
+      "grad_norm": 0.027664449474453394,
+      "learning_rate": 9.196011689444061e-07,
+      "loss": 0.9057,
+      "step": 6220
+    },
+    {
+      "epoch": 1.966358682776593,
+      "grad_norm": 0.02804811457422032,
+      "learning_rate": 8.379238926067845e-07,
+      "loss": 0.9279,
+      "step": 6225
+    },
+    {
+      "epoch": 1.9679380873410723,
+      "grad_norm": 0.028313917901337752,
+      "learning_rate": 7.600414139139477e-07,
+      "loss": 0.9599,
+      "step": 6230
+    },
+    {
+      "epoch": 1.9695174919055516,
+      "grad_norm": 0.028323922173206457,
+      "learning_rate": 6.859543249546074e-07,
+      "loss": 0.9481,
+      "step": 6235
+    },
+    {
+      "epoch": 1.9710968964700308,
+      "grad_norm": 0.02668801851595795,
+      "learning_rate": 6.156631889637776e-07,
+      "loss": 0.8966,
+      "step": 6240
+    },
+    {
+      "epoch": 1.97267630103451,
+      "grad_norm": 0.027012280802929903,
+      "learning_rate": 5.491685403181679e-07,
+      "loss": 0.9364,
+      "step": 6245
+    },
+    {
+      "epoch": 1.9742557055989893,
+      "grad_norm": 0.03458699490095528,
+      "learning_rate": 4.864708845324639e-07,
+      "loss": 0.9719,
+      "step": 6250
+    },
+    {
+      "epoch": 1.9758351101634684,
+      "grad_norm": 0.029890185354397428,
+      "learning_rate": 4.275706982552752e-07,
+      "loss": 0.9293,
+      "step": 6255
+    },
+    {
+      "epoch": 1.9774145147279476,
+      "grad_norm": 0.028008736761539147,
+      "learning_rate": 3.724684292655822e-07,
+      "loss": 0.9318,
+      "step": 6260
+    },
+    {
+      "epoch": 1.9789939192924266,
+      "grad_norm": 0.02920905193937555,
+      "learning_rate": 3.21164496469295e-07,
+      "loss": 0.9837,
+      "step": 6265
+    },
+    {
+      "epoch": 1.9805733238569059,
+      "grad_norm": 0.02677931624268966,
+      "learning_rate": 2.736592898961998e-07,
+      "loss": 0.9539,
+      "step": 6270
+    },
+    {
+      "epoch": 1.9821527284213851,
+      "grad_norm": 0.027770174529727857,
+      "learning_rate": 2.29953170696795e-07,
+      "loss": 0.9783,
+      "step": 6275
+    },
+    {
+      "epoch": 1.9837321329858644,
+      "grad_norm": 0.02942134960133055,
+      "learning_rate": 1.900464711396821e-07,
+      "loss": 0.9974,
+      "step": 6280
+    },
+    {
+      "epoch": 1.9853115375503436,
+      "grad_norm": 0.02938903540449821,
+      "learning_rate": 1.5393949460895674e-07,
+      "loss": 0.8912,
+      "step": 6285
+    },
+    {
+      "epoch": 1.9868909421148229,
+      "grad_norm": 0.02772513890848461,
+      "learning_rate": 1.2163251560198817e-07,
+      "loss": 0.96,
+      "step": 6290
+    },
+    {
+      "epoch": 1.988470346679302,
+      "grad_norm": 0.02890495825239336,
+      "learning_rate": 9.312577972725444e-08,
+      "loss": 0.9747,
+      "step": 6295
+    },
+    {
+      "epoch": 1.9900497512437811,
+      "grad_norm": 0.0291827619706324,
+      "learning_rate": 6.841950370256589e-08,
+      "loss": 0.8942,
+      "step": 6300
+    },
+    {
+      "epoch": 1.9916291558082602,
+      "grad_norm": 0.02705666292763451,
+      "learning_rate": 4.751387535328888e-08,
+      "loss": 0.9338,
+      "step": 6305
+    },
+    {
+      "epoch": 1.9932085603727394,
+      "grad_norm": 0.03065964434744364,
+      "learning_rate": 3.0409053611013534e-08,
+      "loss": 0.8768,
+      "step": 6310
+    },
+    {
+      "epoch": 1.9947879649372187,
+      "grad_norm": 0.031459567261155605,
+      "learning_rate": 1.7105168512443482e-08,
+      "loss": 0.9127,
+      "step": 6315
+    },
+    {
+      "epoch": 1.996367369501698,
+      "grad_norm": 0.02752383355078723,
+      "learning_rate": 7.602321198063589e-09,
+      "loss": 0.9482,
+      "step": 6320
+    },
+    {
+      "epoch": 1.9979467740661772,
+      "grad_norm": 0.0273474840031279,
+      "learning_rate": 1.900583911751408e-09,
+      "loss": 0.9588,
+      "step": 6325
+    },
+    {
+      "epoch": 1.9995261786306564,
+      "grad_norm": 0.06363709530898917,
+      "learning_rate": 0.0,
+      "loss": 0.9738,
+      "step": 6330
     },
     {
-      "epoch": 1.7142857142857144,
-      "eval_loss": 13.667020797729492,
-      "eval_runtime": 0.4014,
-      "eval_samples_per_second": 4.983,
-      "eval_steps_per_second": 2.491,
-      "step": 6
+      "epoch": 1.9995261786306564,
+      "eval_loss": 1.0175522565841675,
+      "eval_runtime": 509.7076,
+      "eval_samples_per_second": 5.197,
+      "eval_steps_per_second": 1.301,
+      "step": 6330
     },
     {
-      "epoch": 1.7142857142857144,
-      "step": 6,
-      "total_flos": 1113549778714624.0,
-      "train_loss": 19.26901610692342,
-      "train_runtime": 77.567,
-      "train_samples_per_second": 1.444,
-      "train_steps_per_second": 0.077
+      "epoch": 1.9995261786306564,
+      "step": 6330,
+      "total_flos": 1.174779724933628e+18,
+      "train_loss": 1.0698493373337514,
+      "train_runtime": 77836.8266,
+      "train_samples_per_second": 1.301,
+      "train_steps_per_second": 0.081
     }
   ],
   "logging_steps": 5,
-  "max_steps": 6,
+  "max_steps": 6330,
   "num_input_tokens_seen": 0,
   "num_train_epochs": 2,
   "save_steps": 500,
@@ -65,7 +8920,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 1113549778714624.0,
+  "total_flos": 1.174779724933628e+18,
   "train_batch_size": 4,
   "trial_name": null,
   "trial_params": null