diff --git "a/trainer_state.json" "b/trainer_state.json"
--- "a/trainer_state.json"
+++ "b/trainer_state.json"
@@ -1,9 +1,9 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 0.3849855630413859,
+  "epoch": 0.7699711260827719,
   "eval_steps": 500,
-  "global_step": 10000,
+  "global_step": 20000,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -14007,6 +14007,14006 @@
       "learning_rate": 0.00018229640445610988,
       "loss": 1.1324,
       "step": 10000
+    },
+    {
+      "epoch": 0.38517805582290665,
+      "grad_norm": 1.413021445274353,
+      "learning_rate": 0.00018227922177125984,
+      "loss": 1.0402,
+      "step": 10005
+    },
+    {
+      "epoch": 0.3853705486044273,
+      "grad_norm": 1.125299334526062,
+      "learning_rate": 0.00018226203156252005,
+      "loss": 1.271,
+      "step": 10010
+    },
+    {
+      "epoch": 0.38556304138594805,
+      "grad_norm": 1.2611075639724731,
+      "learning_rate": 0.00018224483383146237,
+      "loss": 1.2228,
+      "step": 10015
+    },
+    {
+      "epoch": 0.3857555341674687,
+      "grad_norm": 1.0332306623458862,
+      "learning_rate": 0.00018222762857965944,
+      "loss": 1.2059,
+      "step": 10020
+    },
+    {
+      "epoch": 0.3859480269489894,
+      "grad_norm": 1.965288758277893,
+      "learning_rate": 0.00018221041580868464,
+      "loss": 1.217,
+      "step": 10025
+    },
+    {
+      "epoch": 0.3861405197305101,
+      "grad_norm": 0.8059799075126648,
+      "learning_rate": 0.00018219319552011186,
+      "loss": 1.2039,
+      "step": 10030
+    },
+    {
+      "epoch": 0.3863330125120308,
+      "grad_norm": 1.4955195188522339,
+      "learning_rate": 0.00018217596771551584,
+      "loss": 1.2206,
+      "step": 10035
+    },
+    {
+      "epoch": 0.3865255052935515,
+      "grad_norm": 0.987479567527771,
+      "learning_rate": 0.00018215873239647197,
+      "loss": 1.3134,
+      "step": 10040
+    },
+    {
+      "epoch": 0.3867179980750722,
+      "grad_norm": 1.7247464656829834,
+      "learning_rate": 0.00018214148956455627,
+      "loss": 1.1786,
+      "step": 10045
+    },
+    {
+      "epoch": 0.3869104908565929,
+      "grad_norm": 0.9822973608970642,
+      "learning_rate": 0.00018212423922134546,
+      "loss": 1.0866,
+      "step": 10050
+    },
+    {
+      "epoch": 0.3871029836381136,
+      "grad_norm": 1.1217613220214844,
+      "learning_rate": 0.000182106981368417,
+      "loss": 1.3292,
+      "step": 10055
+    },
+    {
+      "epoch": 0.38729547641963424,
+      "grad_norm": 1.2722941637039185,
+      "learning_rate": 0.000182089716007349,
+      "loss": 1.0294,
+      "step": 10060
+    },
+    {
+      "epoch": 0.38748796920115497,
+      "grad_norm": 1.6616365909576416,
+      "learning_rate": 0.00018207244313972026,
+      "loss": 1.3691,
+      "step": 10065
+    },
+    {
+      "epoch": 0.38768046198267564,
+      "grad_norm": 4.093936443328857,
+      "learning_rate": 0.0001820551627671103,
+      "loss": 1.1916,
+      "step": 10070
+    },
+    {
+      "epoch": 0.38787295476419636,
+      "grad_norm": 1.9061866998672485,
+      "learning_rate": 0.00018203787489109926,
+      "loss": 1.3733,
+      "step": 10075
+    },
+    {
+      "epoch": 0.38806544754571703,
+      "grad_norm": 1.6439005136489868,
+      "learning_rate": 0.00018202057951326804,
+      "loss": 1.3533,
+      "step": 10080
+    },
+    {
+      "epoch": 0.3882579403272377,
+      "grad_norm": 1.535980224609375,
+      "learning_rate": 0.0001820032766351981,
+      "loss": 1.3916,
+      "step": 10085
+    },
+    {
+      "epoch": 0.38845043310875843,
+      "grad_norm": 1.6342761516571045,
+      "learning_rate": 0.00018198596625847177,
+      "loss": 1.335,
+      "step": 10090
+    },
+    {
+      "epoch": 0.3886429258902791,
+      "grad_norm": 2.2760815620422363,
+      "learning_rate": 0.00018196864838467192,
+      "loss": 1.0399,
+      "step": 10095
+    },
+    {
+      "epoch": 0.3888354186717998,
+      "grad_norm": 1.173302412033081,
+      "learning_rate": 0.0001819513230153822,
+      "loss": 1.3414,
+      "step": 10100
+    },
+    {
+      "epoch": 0.3890279114533205,
+      "grad_norm": 1.7409497499465942,
+      "learning_rate": 0.00018193399015218684,
+      "loss": 1.3377,
+      "step": 10105
+    },
+    {
+      "epoch": 0.38922040423484117,
+      "grad_norm": 1.3547555208206177,
+      "learning_rate": 0.00018191664979667085,
+      "loss": 1.2576,
+      "step": 10110
+    },
+    {
+      "epoch": 0.3894128970163619,
+      "grad_norm": 2.2421867847442627,
+      "learning_rate": 0.0001818993019504199,
+      "loss": 1.1624,
+      "step": 10115
+    },
+    {
+      "epoch": 0.38960538979788256,
+      "grad_norm": 1.5812993049621582,
+      "learning_rate": 0.00018188194661502029,
+      "loss": 1.2319,
+      "step": 10120
+    },
+    {
+      "epoch": 0.3897978825794033,
+      "grad_norm": 1.8024287223815918,
+      "learning_rate": 0.00018186458379205908,
+      "loss": 1.4016,
+      "step": 10125
+    },
+    {
+      "epoch": 0.38999037536092396,
+      "grad_norm": 0.9069392681121826,
+      "learning_rate": 0.000181847213483124,
+      "loss": 1.1683,
+      "step": 10130
+    },
+    {
+      "epoch": 0.3901828681424447,
+      "grad_norm": 1.6808935403823853,
+      "learning_rate": 0.00018182983568980346,
+      "loss": 1.3519,
+      "step": 10135
+    },
+    {
+      "epoch": 0.39037536092396535,
+      "grad_norm": 2.584958553314209,
+      "learning_rate": 0.0001818124504136865,
+      "loss": 1.3804,
+      "step": 10140
+    },
+    {
+      "epoch": 0.390567853705486,
+      "grad_norm": 1.4569361209869385,
+      "learning_rate": 0.00018179505765636287,
+      "loss": 1.2862,
+      "step": 10145
+    },
+    {
+      "epoch": 0.39076034648700675,
+      "grad_norm": 2.0809457302093506,
+      "learning_rate": 0.0001817776574194231,
+      "loss": 1.1108,
+      "step": 10150
+    },
+    {
+      "epoch": 0.3909528392685274,
+      "grad_norm": 1.7902493476867676,
+      "learning_rate": 0.00018176024970445828,
+      "loss": 1.0611,
+      "step": 10155
+    },
+    {
+      "epoch": 0.39114533205004814,
+      "grad_norm": 0.9953207969665527,
+      "learning_rate": 0.00018174283451306025,
+      "loss": 1.1883,
+      "step": 10160
+    },
+    {
+      "epoch": 0.3913378248315688,
+      "grad_norm": 1.0629642009735107,
+      "learning_rate": 0.00018172541184682147,
+      "loss": 1.3,
+      "step": 10165
+    },
+    {
+      "epoch": 0.3915303176130895,
+      "grad_norm": 1.546132206916809,
+      "learning_rate": 0.0001817079817073352,
+      "loss": 1.2446,
+      "step": 10170
+    },
+    {
+      "epoch": 0.3917228103946102,
+      "grad_norm": 1.379883050918579,
+      "learning_rate": 0.0001816905440961952,
+      "loss": 1.2964,
+      "step": 10175
+    },
+    {
+      "epoch": 0.3919153031761309,
+      "grad_norm": 1.132592797279358,
+      "learning_rate": 0.00018167309901499613,
+      "loss": 1.3951,
+      "step": 10180
+    },
+    {
+      "epoch": 0.3921077959576516,
+      "grad_norm": 1.4765934944152832,
+      "learning_rate": 0.00018165564646533322,
+      "loss": 1.2278,
+      "step": 10185
+    },
+    {
+      "epoch": 0.3923002887391723,
+      "grad_norm": 1.5826079845428467,
+      "learning_rate": 0.00018163818644880233,
+      "loss": 1.2615,
+      "step": 10190
+    },
+    {
+      "epoch": 0.392492781520693,
+      "grad_norm": 1.5647984743118286,
+      "learning_rate": 0.00018162071896700007,
+      "loss": 1.4696,
+      "step": 10195
+    },
+    {
+      "epoch": 0.39268527430221367,
+      "grad_norm": 1.0377607345581055,
+      "learning_rate": 0.0001816032440215238,
+      "loss": 1.1309,
+      "step": 10200
+    },
+    {
+      "epoch": 0.39287776708373434,
+      "grad_norm": 1.1878221035003662,
+      "learning_rate": 0.0001815857616139714,
+      "loss": 1.1442,
+      "step": 10205
+    },
+    {
+      "epoch": 0.39307025986525507,
+      "grad_norm": 1.5119047164916992,
+      "learning_rate": 0.00018156827174594157,
+      "loss": 1.2436,
+      "step": 10210
+    },
+    {
+      "epoch": 0.39326275264677574,
+      "grad_norm": 1.6624690294265747,
+      "learning_rate": 0.00018155077441903364,
+      "loss": 1.1726,
+      "step": 10215
+    },
+    {
+      "epoch": 0.39345524542829646,
+      "grad_norm": 1.2995012998580933,
+      "learning_rate": 0.0001815332696348476,
+      "loss": 1.3053,
+      "step": 10220
+    },
+    {
+      "epoch": 0.39364773820981713,
+      "grad_norm": 1.3727355003356934,
+      "learning_rate": 0.00018151575739498417,
+      "loss": 1.4224,
+      "step": 10225
+    },
+    {
+      "epoch": 0.3938402309913378,
+      "grad_norm": 1.1980619430541992,
+      "learning_rate": 0.0001814982377010447,
+      "loss": 1.0973,
+      "step": 10230
+    },
+    {
+      "epoch": 0.39403272377285853,
+      "grad_norm": 1.4235668182373047,
+      "learning_rate": 0.00018148071055463128,
+      "loss": 1.1659,
+      "step": 10235
+    },
+    {
+      "epoch": 0.3942252165543792,
+      "grad_norm": 1.1501004695892334,
+      "learning_rate": 0.00018146317595734663,
+      "loss": 1.2738,
+      "step": 10240
+    },
+    {
+      "epoch": 0.3944177093358999,
+      "grad_norm": 1.1686300039291382,
+      "learning_rate": 0.00018144563391079419,
+      "loss": 1.1691,
+      "step": 10245
+    },
+    {
+      "epoch": 0.3946102021174206,
+      "grad_norm": 1.3350188732147217,
+      "learning_rate": 0.00018142808441657806,
+      "loss": 1.2344,
+      "step": 10250
+    },
+    {
+      "epoch": 0.39480269489894126,
+      "grad_norm": 1.0583946704864502,
+      "learning_rate": 0.00018141052747630302,
+      "loss": 1.1358,
+      "step": 10255
+    },
+    {
+      "epoch": 0.394995187680462,
+      "grad_norm": 1.0637165307998657,
+      "learning_rate": 0.00018139296309157454,
+      "loss": 1.2589,
+      "step": 10260
+    },
+    {
+      "epoch": 0.39518768046198266,
+      "grad_norm": 1.971304178237915,
+      "learning_rate": 0.00018137539126399874,
+      "loss": 1.1413,
+      "step": 10265
+    },
+    {
+      "epoch": 0.3953801732435034,
+      "grad_norm": 1.1685267686843872,
+      "learning_rate": 0.0001813578119951825,
+      "loss": 1.1702,
+      "step": 10270
+    },
+    {
+      "epoch": 0.39557266602502406,
+      "grad_norm": 1.620936393737793,
+      "learning_rate": 0.0001813402252867333,
+      "loss": 1.2636,
+      "step": 10275
+    },
+    {
+      "epoch": 0.3957651588065448,
+      "grad_norm": 1.553240180015564,
+      "learning_rate": 0.00018132263114025934,
+      "loss": 1.4167,
+      "step": 10280
+    },
+    {
+      "epoch": 0.39595765158806545,
+      "grad_norm": 1.260498285293579,
+      "learning_rate": 0.00018130502955736942,
+      "loss": 1.2984,
+      "step": 10285
+    },
+    {
+      "epoch": 0.3961501443695861,
+      "grad_norm": 1.7073127031326294,
+      "learning_rate": 0.0001812874205396732,
+      "loss": 1.29,
+      "step": 10290
+    },
+    {
+      "epoch": 0.39634263715110685,
+      "grad_norm": 0.900610625743866,
+      "learning_rate": 0.00018126980408878082,
+      "loss": 1.0423,
+      "step": 10295
+    },
+    {
+      "epoch": 0.3965351299326275,
+      "grad_norm": 1.359563946723938,
+      "learning_rate": 0.00018125218020630324,
+      "loss": 1.1576,
+      "step": 10300
+    },
+    {
+      "epoch": 0.39672762271414824,
+      "grad_norm": 0.9399506449699402,
+      "learning_rate": 0.000181234548893852,
+      "loss": 1.1481,
+      "step": 10305
+    },
+    {
+      "epoch": 0.3969201154956689,
+      "grad_norm": 1.4632538557052612,
+      "learning_rate": 0.00018121691015303944,
+      "loss": 1.1404,
+      "step": 10310
+    },
+    {
+      "epoch": 0.3971126082771896,
+      "grad_norm": 1.644718050956726,
+      "learning_rate": 0.00018119926398547839,
+      "loss": 1.1783,
+      "step": 10315
+    },
+    {
+      "epoch": 0.3973051010587103,
+      "grad_norm": 1.299018144607544,
+      "learning_rate": 0.00018118161039278258,
+      "loss": 1.2076,
+      "step": 10320
+    },
+    {
+      "epoch": 0.397497593840231,
+      "grad_norm": 1.5833697319030762,
+      "learning_rate": 0.00018116394937656632,
+      "loss": 1.0825,
+      "step": 10325
+    },
+    {
+      "epoch": 0.3976900866217517,
+      "grad_norm": 1.4813597202301025,
+      "learning_rate": 0.0001811462809384445,
+      "loss": 1.263,
+      "step": 10330
+    },
+    {
+      "epoch": 0.3978825794032724,
+      "grad_norm": 1.8714033365249634,
+      "learning_rate": 0.00018112860508003284,
+      "loss": 1.2425,
+      "step": 10335
+    },
+    {
+      "epoch": 0.39807507218479304,
+      "grad_norm": 1.5847947597503662,
+      "learning_rate": 0.0001811109218029477,
+      "loss": 1.0863,
+      "step": 10340
+    },
+    {
+      "epoch": 0.39826756496631377,
+      "grad_norm": 1.339046597480774,
+      "learning_rate": 0.00018109323110880604,
+      "loss": 1.3871,
+      "step": 10345
+    },
+    {
+      "epoch": 0.39846005774783444,
+      "grad_norm": 2.370396375656128,
+      "learning_rate": 0.0001810755329992256,
+      "loss": 1.2629,
+      "step": 10350
+    },
+    {
+      "epoch": 0.39865255052935517,
+      "grad_norm": 1.2930303812026978,
+      "learning_rate": 0.00018105782747582474,
+      "loss": 1.1281,
+      "step": 10355
+    },
+    {
+      "epoch": 0.39884504331087584,
+      "grad_norm": 1.2590947151184082,
+      "learning_rate": 0.0001810401145402225,
+      "loss": 1.2229,
+      "step": 10360
+    },
+    {
+      "epoch": 0.39903753609239656,
+      "grad_norm": 0.8280492424964905,
+      "learning_rate": 0.00018102239419403866,
+      "loss": 1.2601,
+      "step": 10365
+    },
+    {
+      "epoch": 0.39923002887391723,
+      "grad_norm": 1.6567853689193726,
+      "learning_rate": 0.0001810046664388936,
+      "loss": 1.1296,
+      "step": 10370
+    },
+    {
+      "epoch": 0.3994225216554379,
+      "grad_norm": 1.2103195190429688,
+      "learning_rate": 0.00018098693127640834,
+      "loss": 1.1524,
+      "step": 10375
+    },
+    {
+      "epoch": 0.3996150144369586,
+      "grad_norm": 1.4716650247573853,
+      "learning_rate": 0.00018096918870820475,
+      "loss": 1.1805,
+      "step": 10380
+    },
+    {
+      "epoch": 0.3998075072184793,
+      "grad_norm": 1.291873574256897,
+      "learning_rate": 0.00018095143873590524,
+      "loss": 1.2877,
+      "step": 10385
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8508723974227905,
+      "learning_rate": 0.0001809336813611329,
+      "loss": 0.9215,
+      "step": 10390
+    },
+    {
+      "epoch": 0.4001924927815207,
+      "grad_norm": 1.1256935596466064,
+      "learning_rate": 0.00018091591658551154,
+      "loss": 1.3286,
+      "step": 10395
+    },
+    {
+      "epoch": 0.40038498556304136,
+      "grad_norm": 1.1910960674285889,
+      "learning_rate": 0.0001808981444106656,
+      "loss": 1.1078,
+      "step": 10400
+    },
+    {
+      "epoch": 0.4005774783445621,
+      "grad_norm": 2.188884735107422,
+      "learning_rate": 0.00018088036483822028,
+      "loss": 1.2762,
+      "step": 10405
+    },
+    {
+      "epoch": 0.40076997112608276,
+      "grad_norm": 0.9240724444389343,
+      "learning_rate": 0.00018086257786980136,
+      "loss": 1.1288,
+      "step": 10410
+    },
+    {
+      "epoch": 0.4009624639076035,
+      "grad_norm": 1.961204171180725,
+      "learning_rate": 0.00018084478350703537,
+      "loss": 1.1863,
+      "step": 10415
+    },
+    {
+      "epoch": 0.40115495668912415,
+      "grad_norm": 1.5713763236999512,
+      "learning_rate": 0.00018082698175154947,
+      "loss": 1.2157,
+      "step": 10420
+    },
+    {
+      "epoch": 0.4013474494706448,
+      "grad_norm": 2.006776809692383,
+      "learning_rate": 0.00018080917260497153,
+      "loss": 1.1671,
+      "step": 10425
+    },
+    {
+      "epoch": 0.40153994225216555,
+      "grad_norm": 1.511513352394104,
+      "learning_rate": 0.00018079135606893006,
+      "loss": 1.2428,
+      "step": 10430
+    },
+    {
+      "epoch": 0.4017324350336862,
+      "grad_norm": 1.5270637273788452,
+      "learning_rate": 0.00018077353214505427,
+      "loss": 1.2887,
+      "step": 10435
+    },
+    {
+      "epoch": 0.40192492781520694,
+      "grad_norm": 1.470389723777771,
+      "learning_rate": 0.00018075570083497407,
+      "loss": 1.2739,
+      "step": 10440
+    },
+    {
+      "epoch": 0.4021174205967276,
+      "grad_norm": 1.224330186843872,
+      "learning_rate": 0.00018073786214031992,
+      "loss": 1.0882,
+      "step": 10445
+    },
+    {
+      "epoch": 0.40230991337824834,
+      "grad_norm": 2.0693979263305664,
+      "learning_rate": 0.00018072001606272316,
+      "loss": 1.5091,
+      "step": 10450
+    },
+    {
+      "epoch": 0.402502406159769,
+      "grad_norm": 1.418346643447876,
+      "learning_rate": 0.00018070216260381567,
+      "loss": 1.2886,
+      "step": 10455
+    },
+    {
+      "epoch": 0.4026948989412897,
+      "grad_norm": 1.8632601499557495,
+      "learning_rate": 0.00018068430176522998,
+      "loss": 1.1809,
+      "step": 10460
+    },
+    {
+      "epoch": 0.4028873917228104,
+      "grad_norm": 1.6064730882644653,
+      "learning_rate": 0.00018066643354859937,
+      "loss": 1.2394,
+      "step": 10465
+    },
+    {
+      "epoch": 0.4030798845043311,
+      "grad_norm": 1.2319833040237427,
+      "learning_rate": 0.0001806485579555578,
+      "loss": 1.1979,
+      "step": 10470
+    },
+    {
+      "epoch": 0.4032723772858518,
+      "grad_norm": 1.5506865978240967,
+      "learning_rate": 0.00018063067498773987,
+      "loss": 1.1899,
+      "step": 10475
+    },
+    {
+      "epoch": 0.40346487006737247,
+      "grad_norm": 1.360120415687561,
+      "learning_rate": 0.00018061278464678082,
+      "loss": 1.0995,
+      "step": 10480
+    },
+    {
+      "epoch": 0.40365736284889314,
+      "grad_norm": 1.133346438407898,
+      "learning_rate": 0.00018059488693431664,
+      "loss": 1.1972,
+      "step": 10485
+    },
+    {
+      "epoch": 0.40384985563041387,
+      "grad_norm": 1.6961482763290405,
+      "learning_rate": 0.00018057698185198394,
+      "loss": 1.0823,
+      "step": 10490
+    },
+    {
+      "epoch": 0.40404234841193454,
+      "grad_norm": 1.0126832723617554,
+      "learning_rate": 0.00018055906940142,
+      "loss": 1.3294,
+      "step": 10495
+    },
+    {
+      "epoch": 0.40423484119345526,
+      "grad_norm": 1.378825068473816,
+      "learning_rate": 0.00018054114958426283,
+      "loss": 1.3188,
+      "step": 10500
+    },
+    {
+      "epoch": 0.40442733397497593,
+      "grad_norm": 1.1392402648925781,
+      "learning_rate": 0.00018052322240215104,
+      "loss": 1.2428,
+      "step": 10505
+    },
+    {
+      "epoch": 0.40461982675649666,
+      "grad_norm": 1.0441240072250366,
+      "learning_rate": 0.00018050528785672402,
+      "loss": 1.2997,
+      "step": 10510
+    },
+    {
+      "epoch": 0.40481231953801733,
+      "grad_norm": 1.3564190864562988,
+      "learning_rate": 0.00018048734594962171,
+      "loss": 1.3018,
+      "step": 10515
+    },
+    {
+      "epoch": 0.405004812319538,
+      "grad_norm": 1.3429349660873413,
+      "learning_rate": 0.0001804693966824848,
+      "loss": 1.0567,
+      "step": 10520
+    },
+    {
+      "epoch": 0.4051973051010587,
+      "grad_norm": 0.920313835144043,
+      "learning_rate": 0.00018045144005695462,
+      "loss": 1.1386,
+      "step": 10525
+    },
+    {
+      "epoch": 0.4053897978825794,
+      "grad_norm": 2.402700662612915,
+      "learning_rate": 0.00018043347607467317,
+      "loss": 1.2837,
+      "step": 10530
+    },
+    {
+      "epoch": 0.4055822906641001,
+      "grad_norm": 1.7154083251953125,
+      "learning_rate": 0.00018041550473728318,
+      "loss": 1.3188,
+      "step": 10535
+    },
+    {
+      "epoch": 0.4057747834456208,
+      "grad_norm": 0.8770251274108887,
+      "learning_rate": 0.000180397526046428,
+      "loss": 1.1641,
+      "step": 10540
+    },
+    {
+      "epoch": 0.40596727622714146,
+      "grad_norm": 0.9887571334838867,
+      "learning_rate": 0.0001803795400037516,
+      "loss": 1.0042,
+      "step": 10545
+    },
+    {
+      "epoch": 0.4061597690086622,
+      "grad_norm": 2.665354013442993,
+      "learning_rate": 0.00018036154661089877,
+      "loss": 1.2579,
+      "step": 10550
+    },
+    {
+      "epoch": 0.40635226179018286,
+      "grad_norm": 2.6088809967041016,
+      "learning_rate": 0.00018034354586951486,
+      "loss": 1.1098,
+      "step": 10555
+    },
+    {
+      "epoch": 0.4065447545717036,
+      "grad_norm": 1.4641830921173096,
+      "learning_rate": 0.00018032553778124586,
+      "loss": 1.1108,
+      "step": 10560
+    },
+    {
+      "epoch": 0.40673724735322425,
+      "grad_norm": 1.0744770765304565,
+      "learning_rate": 0.00018030752234773854,
+      "loss": 1.1234,
+      "step": 10565
+    },
+    {
+      "epoch": 0.4069297401347449,
+      "grad_norm": 1.2617886066436768,
+      "learning_rate": 0.00018028949957064034,
+      "loss": 1.1753,
+      "step": 10570
+    },
+    {
+      "epoch": 0.40712223291626565,
+      "grad_norm": 1.4641857147216797,
+      "learning_rate": 0.00018027146945159923,
+      "loss": 1.2671,
+      "step": 10575
+    },
+    {
+      "epoch": 0.4073147256977863,
+      "grad_norm": 1.4347914457321167,
+      "learning_rate": 0.00018025343199226402,
+      "loss": 1.1348,
+      "step": 10580
+    },
+    {
+      "epoch": 0.40750721847930704,
+      "grad_norm": 1.434019923210144,
+      "learning_rate": 0.00018023538719428407,
+      "loss": 1.2439,
+      "step": 10585
+    },
+    {
+      "epoch": 0.4076997112608277,
+      "grad_norm": 1.1034338474273682,
+      "learning_rate": 0.00018021733505930944,
+      "loss": 1.0502,
+      "step": 10590
+    },
+    {
+      "epoch": 0.40789220404234844,
+      "grad_norm": 1.591850996017456,
+      "learning_rate": 0.00018019927558899097,
+      "loss": 1.178,
+      "step": 10595
+    },
+    {
+      "epoch": 0.4080846968238691,
+      "grad_norm": 1.672735333442688,
+      "learning_rate": 0.00018018120878498,
+      "loss": 1.2363,
+      "step": 10600
+    },
+    {
+      "epoch": 0.4082771896053898,
+      "grad_norm": 1.8779442310333252,
+      "learning_rate": 0.00018016313464892862,
+      "loss": 1.2537,
+      "step": 10605
+    },
+    {
+      "epoch": 0.4084696823869105,
+      "grad_norm": 1.075453281402588,
+      "learning_rate": 0.00018014505318248963,
+      "loss": 1.081,
+      "step": 10610
+    },
+    {
+      "epoch": 0.4086621751684312,
+      "grad_norm": 1.350914478302002,
+      "learning_rate": 0.0001801269643873164,
+      "loss": 1.3958,
+      "step": 10615
+    },
+    {
+      "epoch": 0.4088546679499519,
+      "grad_norm": 1.6566729545593262,
+      "learning_rate": 0.0001801088682650631,
+      "loss": 1.3208,
+      "step": 10620
+    },
+    {
+      "epoch": 0.40904716073147257,
+      "grad_norm": 1.243171215057373,
+      "learning_rate": 0.00018009076481738446,
+      "loss": 1.17,
+      "step": 10625
+    },
+    {
+      "epoch": 0.40923965351299324,
+      "grad_norm": 1.110456109046936,
+      "learning_rate": 0.00018007265404593593,
+      "loss": 1.1311,
+      "step": 10630
+    },
+    {
+      "epoch": 0.40943214629451397,
+      "grad_norm": 2.485719919204712,
+      "learning_rate": 0.00018005453595237362,
+      "loss": 1.3703,
+      "step": 10635
+    },
+    {
+      "epoch": 0.40962463907603464,
+      "grad_norm": 1.3115043640136719,
+      "learning_rate": 0.00018003641053835435,
+      "loss": 1.2551,
+      "step": 10640
+    },
+    {
+      "epoch": 0.40981713185755536,
+      "grad_norm": 1.530535340309143,
+      "learning_rate": 0.0001800182778055355,
+      "loss": 1.3577,
+      "step": 10645
+    },
+    {
+      "epoch": 0.41000962463907603,
+      "grad_norm": 1.2339287996292114,
+      "learning_rate": 0.00018000013775557521,
+      "loss": 1.2539,
+      "step": 10650
+    },
+    {
+      "epoch": 0.4102021174205967,
+      "grad_norm": 1.579942226409912,
+      "learning_rate": 0.00017998199039013225,
+      "loss": 1.3568,
+      "step": 10655
+    },
+    {
+      "epoch": 0.4103946102021174,
+      "grad_norm": 1.831764817237854,
+      "learning_rate": 0.00017996383571086612,
+      "loss": 1.3662,
+      "step": 10660
+    },
+    {
+      "epoch": 0.4105871029836381,
+      "grad_norm": 2.1747963428497314,
+      "learning_rate": 0.00017994567371943697,
+      "loss": 1.1333,
+      "step": 10665
+    },
+    {
+      "epoch": 0.4107795957651588,
+      "grad_norm": 1.5603039264678955,
+      "learning_rate": 0.00017992750441750549,
+      "loss": 1.2327,
+      "step": 10670
+    },
+    {
+      "epoch": 0.4109720885466795,
+      "grad_norm": 1.7836112976074219,
+      "learning_rate": 0.00017990932780673324,
+      "loss": 1.0281,
+      "step": 10675
+    },
+    {
+      "epoch": 0.4111645813282002,
+      "grad_norm": 1.5049426555633545,
+      "learning_rate": 0.0001798911438887823,
+      "loss": 1.3338,
+      "step": 10680
+    },
+    {
+      "epoch": 0.4113570741097209,
+      "grad_norm": 1.6236990690231323,
+      "learning_rate": 0.00017987295266531548,
+      "loss": 1.3937,
+      "step": 10685
+    },
+    {
+      "epoch": 0.41154956689124156,
+      "grad_norm": 1.2450697422027588,
+      "learning_rate": 0.00017985475413799623,
+      "loss": 1.3456,
+      "step": 10690
+    },
+    {
+      "epoch": 0.4117420596727623,
+      "grad_norm": 1.031137228012085,
+      "learning_rate": 0.00017983654830848873,
+      "loss": 1.0254,
+      "step": 10695
+    },
+    {
+      "epoch": 0.41193455245428295,
+      "grad_norm": 1.588884949684143,
+      "learning_rate": 0.00017981833517845773,
+      "loss": 1.0554,
+      "step": 10700
+    },
+    {
+      "epoch": 0.4121270452358037,
+      "grad_norm": 1.2405824661254883,
+      "learning_rate": 0.00017980011474956874,
+      "loss": 1.4561,
+      "step": 10705
+    },
+    {
+      "epoch": 0.41231953801732435,
+      "grad_norm": 2.03009295463562,
+      "learning_rate": 0.00017978188702348792,
+      "loss": 1.2479,
+      "step": 10710
+    },
+    {
+      "epoch": 0.412512030798845,
+      "grad_norm": 0.9755954146385193,
+      "learning_rate": 0.00017976365200188198,
+      "loss": 1.1632,
+      "step": 10715
+    },
+    {
+      "epoch": 0.41270452358036575,
+      "grad_norm": 1.3121798038482666,
+      "learning_rate": 0.00017974540968641848,
+      "loss": 1.2069,
+      "step": 10720
+    },
+    {
+      "epoch": 0.4128970163618864,
+      "grad_norm": 1.880199909210205,
+      "learning_rate": 0.00017972716007876556,
+      "loss": 1.32,
+      "step": 10725
+    },
+    {
+      "epoch": 0.41308950914340714,
+      "grad_norm": 2.1090636253356934,
+      "learning_rate": 0.00017970890318059194,
+      "loss": 1.2943,
+      "step": 10730
+    },
+    {
+      "epoch": 0.4132820019249278,
+      "grad_norm": 1.2155611515045166,
+      "learning_rate": 0.00017969063899356716,
+      "loss": 1.3022,
+      "step": 10735
+    },
+    {
+      "epoch": 0.4134744947064485,
+      "grad_norm": 1.191871166229248,
+      "learning_rate": 0.00017967236751936135,
+      "loss": 1.2699,
+      "step": 10740
+    },
+    {
+      "epoch": 0.4136669874879692,
+      "grad_norm": 1.4702094793319702,
+      "learning_rate": 0.00017965408875964534,
+      "loss": 1.3936,
+      "step": 10745
+    },
+    {
+      "epoch": 0.4138594802694899,
+      "grad_norm": 1.7658724784851074,
+      "learning_rate": 0.00017963580271609052,
+      "loss": 1.1633,
+      "step": 10750
+    },
+    {
+      "epoch": 0.4140519730510106,
+      "grad_norm": 1.5030126571655273,
+      "learning_rate": 0.00017961750939036913,
+      "loss": 1.4213,
+      "step": 10755
+    },
+    {
+      "epoch": 0.4142444658325313,
+      "grad_norm": 1.5616711378097534,
+      "learning_rate": 0.0001795992087841539,
+      "loss": 1.3342,
+      "step": 10760
+    },
+    {
+      "epoch": 0.414436958614052,
+      "grad_norm": 1.2506111860275269,
+      "learning_rate": 0.0001795809008991183,
+      "loss": 1.0034,
+      "step": 10765
+    },
+    {
+      "epoch": 0.41462945139557267,
+      "grad_norm": 1.1011154651641846,
+      "learning_rate": 0.00017956258573693657,
+      "loss": 1.0936,
+      "step": 10770
+    },
+    {
+      "epoch": 0.41482194417709334,
+      "grad_norm": 1.2040156126022339,
+      "learning_rate": 0.00017954426329928335,
+      "loss": 1.1974,
+      "step": 10775
+    },
+    {
+      "epoch": 0.41501443695861406,
+      "grad_norm": 1.5271620750427246,
+      "learning_rate": 0.0001795259335878342,
+      "loss": 1.2563,
+      "step": 10780
+    },
+    {
+      "epoch": 0.41520692974013473,
+      "grad_norm": 1.342129111289978,
+      "learning_rate": 0.00017950759660426523,
+      "loss": 1.2319,
+      "step": 10785
+    },
+    {
+      "epoch": 0.41539942252165546,
+      "grad_norm": 0.9986871480941772,
+      "learning_rate": 0.00017948925235025326,
+      "loss": 1.0781,
+      "step": 10790
+    },
+    {
+      "epoch": 0.41559191530317613,
+      "grad_norm": 1.107088327407837,
+      "learning_rate": 0.00017947090082747573,
+      "loss": 1.1499,
+      "step": 10795
+    },
+    {
+      "epoch": 0.4157844080846968,
+      "grad_norm": 1.5566056966781616,
+      "learning_rate": 0.00017945254203761076,
+      "loss": 1.0997,
+      "step": 10800
+    },
+    {
+      "epoch": 0.4159769008662175,
+      "grad_norm": 1.4681777954101562,
+      "learning_rate": 0.00017943417598233715,
+      "loss": 1.3307,
+      "step": 10805
+    },
+    {
+      "epoch": 0.4161693936477382,
+      "grad_norm": 1.4198453426361084,
+      "learning_rate": 0.00017941580266333433,
+      "loss": 0.9664,
+      "step": 10810
+    },
+    {
+      "epoch": 0.4163618864292589,
+      "grad_norm": 1.1474230289459229,
+      "learning_rate": 0.00017939742208228246,
+      "loss": 1.2454,
+      "step": 10815
+    },
+    {
+      "epoch": 0.4165543792107796,
+      "grad_norm": 1.186672568321228,
+      "learning_rate": 0.00017937903424086228,
+      "loss": 1.3311,
+      "step": 10820
+    },
+    {
+      "epoch": 0.4167468719923003,
+      "grad_norm": 1.4548507928848267,
+      "learning_rate": 0.00017936063914075526,
+      "loss": 1.2508,
+      "step": 10825
+    },
+    {
+      "epoch": 0.416939364773821,
+      "grad_norm": 1.0224876403808594,
+      "learning_rate": 0.00017934223678364353,
+      "loss": 0.9364,
+      "step": 10830
+    },
+    {
+      "epoch": 0.41713185755534166,
+      "grad_norm": 1.5561485290527344,
+      "learning_rate": 0.00017932382717120984,
+      "loss": 1.1686,
+      "step": 10835
+    },
+    {
+      "epoch": 0.4173243503368624,
+      "grad_norm": 1.9549082517623901,
+      "learning_rate": 0.00017930541030513762,
+      "loss": 1.2678,
+      "step": 10840
+    },
+    {
+      "epoch": 0.41751684311838305,
+      "grad_norm": 1.2266019582748413,
+      "learning_rate": 0.00017928698618711094,
+      "loss": 1.2963,
+      "step": 10845
+    },
+    {
+      "epoch": 0.4177093358999038,
+      "grad_norm": 0.6992445588111877,
+      "learning_rate": 0.00017926855481881465,
+      "loss": 1.1042,
+      "step": 10850
+    },
+    {
+      "epoch": 0.41790182868142445,
+      "grad_norm": 1.515512466430664,
+      "learning_rate": 0.00017925011620193408,
+      "loss": 1.0718,
+      "step": 10855
+    },
+    {
+      "epoch": 0.4180943214629451,
+      "grad_norm": 1.5123271942138672,
+      "learning_rate": 0.0001792316703381554,
+      "loss": 1.1307,
+      "step": 10860
+    },
+    {
+      "epoch": 0.41828681424446584,
+      "grad_norm": 1.3709865808486938,
+      "learning_rate": 0.00017921321722916535,
+      "loss": 1.3652,
+      "step": 10865
+    },
+    {
+      "epoch": 0.4184793070259865,
+      "grad_norm": 1.3327142000198364,
+      "learning_rate": 0.0001791947568766513,
+      "loss": 1.2644,
+      "step": 10870
+    },
+    {
+      "epoch": 0.41867179980750724,
+      "grad_norm": 1.460595726966858,
+      "learning_rate": 0.00017917628928230134,
+      "loss": 1.2783,
+      "step": 10875
+    },
+    {
+      "epoch": 0.4188642925890279,
+      "grad_norm": 1.1008737087249756,
+      "learning_rate": 0.00017915781444780425,
+      "loss": 1.2889,
+      "step": 10880
+    },
+    {
+      "epoch": 0.4190567853705486,
+      "grad_norm": 1.8467929363250732,
+      "learning_rate": 0.00017913933237484936,
+      "loss": 1.1897,
+      "step": 10885
+    },
+    {
+      "epoch": 0.4192492781520693,
+      "grad_norm": 1.286544680595398,
+      "learning_rate": 0.00017912084306512683,
+      "loss": 1.1239,
+      "step": 10890
+    },
+    {
+      "epoch": 0.41944177093359,
+      "grad_norm": 1.8240995407104492,
+      "learning_rate": 0.00017910234652032726,
+      "loss": 1.3085,
+      "step": 10895
+    },
+    {
+      "epoch": 0.4196342637151107,
+      "grad_norm": 1.1262156963348389,
+      "learning_rate": 0.00017908384274214215,
+      "loss": 1.3779,
+      "step": 10900
+    },
+    {
+      "epoch": 0.41982675649663137,
+      "grad_norm": 1.2274012565612793,
+      "learning_rate": 0.0001790653317322635,
+      "loss": 1.3361,
+      "step": 10905
+    },
+    {
+      "epoch": 0.4200192492781521,
+      "grad_norm": 2.0522284507751465,
+      "learning_rate": 0.000179046813492384,
+      "loss": 1.2329,
+      "step": 10910
+    },
+    {
+      "epoch": 0.42021174205967277,
+      "grad_norm": 1.927666187286377,
+      "learning_rate": 0.0001790282880241971,
+      "loss": 1.2217,
+      "step": 10915
+    },
+    {
+      "epoch": 0.42040423484119344,
+      "grad_norm": 2.254720687866211,
+      "learning_rate": 0.0001790097553293967,
+      "loss": 1.2867,
+      "step": 10920
+    },
+    {
+      "epoch": 0.42059672762271416,
+      "grad_norm": 1.9560370445251465,
+      "learning_rate": 0.0001789912154096776,
+      "loss": 1.2959,
+      "step": 10925
+    },
+    {
+      "epoch": 0.42078922040423483,
+      "grad_norm": 1.109393835067749,
+      "learning_rate": 0.00017897266826673517,
+      "loss": 1.2397,
+      "step": 10930
+    },
+    {
+      "epoch": 0.42098171318575556,
+      "grad_norm": 1.1880956888198853,
+      "learning_rate": 0.00017895411390226527,
+      "loss": 1.192,
+      "step": 10935
+    },
+    {
+      "epoch": 0.4211742059672762,
+      "grad_norm": 1.851517677307129,
+      "learning_rate": 0.00017893555231796477,
+      "loss": 1.1866,
+      "step": 10940
+    },
+    {
+      "epoch": 0.4213666987487969,
+      "grad_norm": 1.1871724128723145,
+      "learning_rate": 0.0001789169835155309,
+      "loss": 1.1627,
+      "step": 10945
+    },
+    {
+      "epoch": 0.4215591915303176,
+      "grad_norm": 0.9478880167007446,
+      "learning_rate": 0.0001788984074966616,
+      "loss": 1.198,
+      "step": 10950
+    },
+    {
+      "epoch": 0.4217516843118383,
+      "grad_norm": 1.753989577293396,
+      "learning_rate": 0.00017887982426305566,
+      "loss": 1.2923,
+      "step": 10955
+    },
+    {
+      "epoch": 0.421944177093359,
+      "grad_norm": 2.161820650100708,
+      "learning_rate": 0.00017886123381641227,
+      "loss": 1.2651,
+      "step": 10960
+    },
+    {
+      "epoch": 0.4221366698748797,
+      "grad_norm": 1.203307867050171,
+      "learning_rate": 0.00017884263615843145,
+      "loss": 1.1854,
+      "step": 10965
+    },
+    {
+      "epoch": 0.42232916265640036,
+      "grad_norm": 1.6671913862228394,
+      "learning_rate": 0.0001788240312908139,
+      "loss": 1.2466,
+      "step": 10970
+    },
+    {
+      "epoch": 0.4225216554379211,
+      "grad_norm": 1.643796443939209,
+      "learning_rate": 0.0001788054192152608,
+      "loss": 1.202,
+      "step": 10975
+    },
+    {
+      "epoch": 0.42271414821944175,
+      "grad_norm": 1.024296522140503,
+      "learning_rate": 0.00017878679993347415,
+      "loss": 1.2392,
+      "step": 10980
+    },
+    {
+      "epoch": 0.4229066410009625,
+      "grad_norm": 1.363425612449646,
+      "learning_rate": 0.0001787681734471566,
+      "loss": 1.3577,
+      "step": 10985
+    },
+    {
+      "epoch": 0.42309913378248315,
+      "grad_norm": 1.7815190553665161,
+      "learning_rate": 0.00017874953975801134,
+      "loss": 0.9826,
+      "step": 10990
+    },
+    {
+      "epoch": 0.4232916265640039,
+      "grad_norm": 1.6736468076705933,
+      "learning_rate": 0.00017873089886774236,
+      "loss": 1.168,
+      "step": 10995
+    },
+    {
+      "epoch": 0.42348411934552455,
+      "grad_norm": 1.3047553300857544,
+      "learning_rate": 0.0001787122507780542,
+      "loss": 1.1839,
+      "step": 11000
+    },
+    {
+      "epoch": 0.4236766121270452,
+      "grad_norm": 1.5737935304641724,
+      "learning_rate": 0.00017869359549065216,
+      "loss": 1.0693,
+      "step": 11005
+    },
+    {
+      "epoch": 0.42386910490856594,
+      "grad_norm": 0.9130328893661499,
+      "learning_rate": 0.00017867493300724208,
+      "loss": 1.1609,
+      "step": 11010
+    },
+    {
+      "epoch": 0.4240615976900866,
+      "grad_norm": 2.444490432739258,
+      "learning_rate": 0.00017865626332953056,
+      "loss": 1.2422,
+      "step": 11015
+    },
+    {
+      "epoch": 0.42425409047160734,
+      "grad_norm": 1.4214091300964355,
+      "learning_rate": 0.00017863758645922481,
+      "loss": 1.2028,
+      "step": 11020
+    },
+    {
+      "epoch": 0.424446583253128,
+      "grad_norm": 1.3986276388168335,
+      "learning_rate": 0.0001786189023980327,
+      "loss": 0.9271,
+      "step": 11025
+    },
+    {
+      "epoch": 0.4246390760346487,
+      "grad_norm": 1.6309832334518433,
+      "learning_rate": 0.00017860021114766275,
+      "loss": 1.1242,
+      "step": 11030
+    },
+    {
+      "epoch": 0.4248315688161694,
+      "grad_norm": 1.0703374147415161,
+      "learning_rate": 0.00017858151270982423,
+      "loss": 1.1688,
+      "step": 11035
+    },
+    {
+      "epoch": 0.4250240615976901,
+      "grad_norm": 0.9345492720603943,
+      "learning_rate": 0.00017856280708622687,
+      "loss": 1.0759,
+      "step": 11040
+    },
+    {
+      "epoch": 0.4252165543792108,
+      "grad_norm": 1.1012792587280273,
+      "learning_rate": 0.00017854409427858124,
+      "loss": 1.3299,
+      "step": 11045
+    },
+    {
+      "epoch": 0.42540904716073147,
+      "grad_norm": 1.087344765663147,
+      "learning_rate": 0.00017852537428859853,
+      "loss": 1.1188,
+      "step": 11050
+    },
+    {
+      "epoch": 0.42560153994225214,
+      "grad_norm": 1.0374698638916016,
+      "learning_rate": 0.0001785066471179905,
+      "loss": 1.2403,
+      "step": 11055
+    },
+    {
+      "epoch": 0.42579403272377286,
+      "grad_norm": 1.2250018119812012,
+      "learning_rate": 0.00017848791276846963,
+      "loss": 1.1217,
+      "step": 11060
+    },
+    {
+      "epoch": 0.42598652550529353,
+      "grad_norm": 1.9863545894622803,
+      "learning_rate": 0.0001784691712417491,
+      "loss": 1.0159,
+      "step": 11065
+    },
+    {
+      "epoch": 0.42617901828681426,
+      "grad_norm": 1.3587582111358643,
+      "learning_rate": 0.0001784504225395427,
+      "loss": 1.1266,
+      "step": 11070
+    },
+    {
+      "epoch": 0.42637151106833493,
+      "grad_norm": 1.3274664878845215,
+      "learning_rate": 0.0001784316666635648,
+      "loss": 1.2295,
+      "step": 11075
+    },
+    {
+      "epoch": 0.42656400384985566,
+      "grad_norm": 1.594498872756958,
+      "learning_rate": 0.00017841290361553057,
+      "loss": 1.2942,
+      "step": 11080
+    },
+    {
+      "epoch": 0.4267564966313763,
+      "grad_norm": 2.5940325260162354,
+      "learning_rate": 0.00017839413339715572,
+      "loss": 1.3333,
+      "step": 11085
+    },
+    {
+      "epoch": 0.426948989412897,
+      "grad_norm": 1.5368024110794067,
+      "learning_rate": 0.0001783753560101567,
+      "loss": 1.2738,
+      "step": 11090
+    },
+    {
+      "epoch": 0.4271414821944177,
+      "grad_norm": 1.8095320463180542,
+      "learning_rate": 0.00017835657145625055,
+      "loss": 1.3245,
+      "step": 11095
+    },
+    {
+      "epoch": 0.4273339749759384,
+      "grad_norm": 1.4597771167755127,
+      "learning_rate": 0.000178337779737155,
+      "loss": 1.3837,
+      "step": 11100
+    },
+    {
+      "epoch": 0.4275264677574591,
+      "grad_norm": 1.052746057510376,
+      "learning_rate": 0.00017831898085458842,
+      "loss": 1.1603,
+      "step": 11105
+    },
+    {
+      "epoch": 0.4277189605389798,
+      "grad_norm": 1.547523856163025,
+      "learning_rate": 0.0001783001748102699,
+      "loss": 1.2277,
+      "step": 11110
+    },
+    {
+      "epoch": 0.42791145332050046,
+      "grad_norm": 2.109560012817383,
+      "learning_rate": 0.00017828136160591906,
+      "loss": 1.1299,
+      "step": 11115
+    },
+    {
+      "epoch": 0.4281039461020212,
+      "grad_norm": 0.9221099019050598,
+      "learning_rate": 0.00017826254124325626,
+      "loss": 1.1447,
+      "step": 11120
+    },
+    {
+      "epoch": 0.42829643888354185,
+      "grad_norm": 1.1257829666137695,
+      "learning_rate": 0.00017824371372400255,
+      "loss": 1.0844,
+      "step": 11125
+    },
+    {
+      "epoch": 0.4284889316650626,
+      "grad_norm": 1.9643393754959106,
+      "learning_rate": 0.00017822487904987948,
+      "loss": 1.1511,
+      "step": 11130
+    },
+    {
+      "epoch": 0.42868142444658325,
+      "grad_norm": 1.2279611825942993,
+      "learning_rate": 0.00017820603722260944,
+      "loss": 1.3039,
+      "step": 11135
+    },
+    {
+      "epoch": 0.428873917228104,
+      "grad_norm": 1.8037766218185425,
+      "learning_rate": 0.00017818718824391536,
+      "loss": 1.2338,
+      "step": 11140
+    },
+    {
+      "epoch": 0.42906641000962464,
+      "grad_norm": 2.1256327629089355,
+      "learning_rate": 0.00017816833211552085,
+      "loss": 1.2502,
+      "step": 11145
+    },
+    {
+      "epoch": 0.4292589027911453,
+      "grad_norm": 1.1520932912826538,
+      "learning_rate": 0.0001781494688391502,
+      "loss": 1.121,
+      "step": 11150
+    },
+    {
+      "epoch": 0.42945139557266604,
+      "grad_norm": 1.1287842988967896,
+      "learning_rate": 0.00017813059841652833,
+      "loss": 1.2012,
+      "step": 11155
+    },
+    {
+      "epoch": 0.4296438883541867,
+      "grad_norm": 1.2584294080734253,
+      "learning_rate": 0.00017811172084938076,
+      "loss": 1.3221,
+      "step": 11160
+    },
+    {
+      "epoch": 0.42983638113570743,
+      "grad_norm": 1.901994228363037,
+      "learning_rate": 0.0001780928361394338,
+      "loss": 1.1184,
+      "step": 11165
+    },
+    {
+      "epoch": 0.4300288739172281,
+      "grad_norm": 1.564501166343689,
+      "learning_rate": 0.00017807394428841428,
+      "loss": 1.12,
+      "step": 11170
+    },
+    {
+      "epoch": 0.4302213666987488,
+      "grad_norm": 2.138155221939087,
+      "learning_rate": 0.00017805504529804975,
+      "loss": 1.1928,
+      "step": 11175
+    },
+    {
+      "epoch": 0.4304138594802695,
+      "grad_norm": 1.3132466077804565,
+      "learning_rate": 0.00017803613917006841,
+      "loss": 1.2674,
+      "step": 11180
+    },
+    {
+      "epoch": 0.43060635226179017,
+      "grad_norm": 1.1847275495529175,
+      "learning_rate": 0.00017801722590619903,
+      "loss": 1.1457,
+      "step": 11185
+    },
+    {
+      "epoch": 0.4307988450433109,
+      "grad_norm": 1.6100077629089355,
+      "learning_rate": 0.00017799830550817124,
+      "loss": 1.3779,
+      "step": 11190
+    },
+    {
+      "epoch": 0.43099133782483157,
+      "grad_norm": 2.1193013191223145,
+      "learning_rate": 0.00017797937797771503,
+      "loss": 1.0515,
+      "step": 11195
+    },
+    {
+      "epoch": 0.43118383060635224,
+      "grad_norm": 1.6185005903244019,
+      "learning_rate": 0.0001779604433165613,
+      "loss": 1.078,
+      "step": 11200
+    },
+    {
+      "epoch": 0.43137632338787296,
+      "grad_norm": 1.275046467781067,
+      "learning_rate": 0.00017794150152644148,
+      "loss": 2.2652,
+      "step": 11205
+    },
+    {
+      "epoch": 0.43156881616939363,
+      "grad_norm": 1.4507300853729248,
+      "learning_rate": 0.00017792255260908765,
+      "loss": 1.3556,
+      "step": 11210
+    },
+    {
+      "epoch": 0.43176130895091436,
+      "grad_norm": 1.5722453594207764,
+      "learning_rate": 0.00017790359656623256,
+      "loss": 1.1115,
+      "step": 11215
+    },
+    {
+      "epoch": 0.431953801732435,
+      "grad_norm": 1.802585244178772,
+      "learning_rate": 0.00017788463339960962,
+      "loss": 1.1885,
+      "step": 11220
+    },
+    {
+      "epoch": 0.43214629451395575,
+      "grad_norm": 1.0945521593093872,
+      "learning_rate": 0.00017786566311095295,
+      "loss": 1.2419,
+      "step": 11225
+    },
+    {
+      "epoch": 0.4323387872954764,
+      "grad_norm": 1.6798467636108398,
+      "learning_rate": 0.00017784668570199714,
+      "loss": 1.0404,
+      "step": 11230
+    },
+    {
+      "epoch": 0.4325312800769971,
+      "grad_norm": 1.9263988733291626,
+      "learning_rate": 0.00017782770117447764,
+      "loss": 1.2925,
+      "step": 11235
+    },
+    {
+      "epoch": 0.4327237728585178,
+      "grad_norm": 1.3327709436416626,
+      "learning_rate": 0.0001778087095301304,
+      "loss": 1.2621,
+      "step": 11240
+    },
+    {
+      "epoch": 0.4329162656400385,
+      "grad_norm": 1.540216088294983,
+      "learning_rate": 0.00017778971077069214,
+      "loss": 1.2733,
+      "step": 11245
+    },
+    {
+      "epoch": 0.4331087584215592,
+      "grad_norm": 0.8980332612991333,
+      "learning_rate": 0.00017777070489790014,
+      "loss": 1.1849,
+      "step": 11250
+    },
+    {
+      "epoch": 0.4333012512030799,
+      "grad_norm": 1.1286743879318237,
+      "learning_rate": 0.00017775169191349238,
+      "loss": 1.0491,
+      "step": 11255
+    },
+    {
+      "epoch": 0.43349374398460055,
+      "grad_norm": 1.5880367755889893,
+      "learning_rate": 0.0001777326718192074,
+      "loss": 1.1371,
+      "step": 11260
+    },
+    {
+      "epoch": 0.4336862367661213,
+      "grad_norm": 1.8634532690048218,
+      "learning_rate": 0.00017771364461678454,
+      "loss": 1.3491,
+      "step": 11265
+    },
+    {
+      "epoch": 0.43387872954764195,
+      "grad_norm": 1.13876473903656,
+      "learning_rate": 0.0001776946103079637,
+      "loss": 1.1284,
+      "step": 11270
+    },
+    {
+      "epoch": 0.4340712223291627,
+      "grad_norm": 1.1511520147323608,
+      "learning_rate": 0.0001776755688944854,
+      "loss": 0.9705,
+      "step": 11275
+    },
+    {
+      "epoch": 0.43426371511068335,
+      "grad_norm": 2.0832314491271973,
+      "learning_rate": 0.00017765652037809087,
+      "loss": 1.1134,
+      "step": 11280
+    },
+    {
+      "epoch": 0.434456207892204,
+      "grad_norm": 1.3219777345657349,
+      "learning_rate": 0.000177637464760522,
+      "loss": 1.1519,
+      "step": 11285
+    },
+    {
+      "epoch": 0.43464870067372474,
+      "grad_norm": 1.2205532789230347,
+      "learning_rate": 0.0001776184020435213,
+      "loss": 1.1526,
+      "step": 11290
+    },
+    {
+      "epoch": 0.4348411934552454,
+      "grad_norm": 1.1612414121627808,
+      "learning_rate": 0.00017759933222883187,
+      "loss": 1.2236,
+      "step": 11295
+    },
+    {
+      "epoch": 0.43503368623676614,
+      "grad_norm": 2.214245319366455,
+      "learning_rate": 0.00017758025531819756,
+      "loss": 1.1962,
+      "step": 11300
+    },
+    {
+      "epoch": 0.4352261790182868,
+      "grad_norm": 1.1582585573196411,
+      "learning_rate": 0.00017756117131336284,
+      "loss": 1.1488,
+      "step": 11305
+    },
+    {
+      "epoch": 0.43541867179980753,
+      "grad_norm": 1.6610682010650635,
+      "learning_rate": 0.0001775420802160728,
+      "loss": 1.2349,
+      "step": 11310
+    },
+    {
+      "epoch": 0.4356111645813282,
+      "grad_norm": 1.2163527011871338,
+      "learning_rate": 0.00017752298202807317,
+      "loss": 1.0914,
+      "step": 11315
+    },
+    {
+      "epoch": 0.4358036573628489,
+      "grad_norm": 1.3684804439544678,
+      "learning_rate": 0.00017750387675111043,
+      "loss": 1.1035,
+      "step": 11320
+    },
+    {
+      "epoch": 0.4359961501443696,
+      "grad_norm": 2.0042598247528076,
+      "learning_rate": 0.00017748476438693151,
+      "loss": 1.1783,
+      "step": 11325
+    },
+    {
+      "epoch": 0.43618864292589027,
+      "grad_norm": 1.4552195072174072,
+      "learning_rate": 0.00017746564493728424,
+      "loss": 1.1373,
+      "step": 11330
+    },
+    {
+      "epoch": 0.436381135707411,
+      "grad_norm": 1.1513317823410034,
+      "learning_rate": 0.00017744651840391685,
+      "loss": 1.122,
+      "step": 11335
+    },
+    {
+      "epoch": 0.43657362848893166,
+      "grad_norm": 1.1842467784881592,
+      "learning_rate": 0.0001774273847885784,
+      "loss": 1.085,
+      "step": 11340
+    },
+    {
+      "epoch": 0.43676612127045233,
+      "grad_norm": 1.5492455959320068,
+      "learning_rate": 0.00017740824409301852,
+      "loss": 1.1355,
+      "step": 11345
+    },
+    {
+      "epoch": 0.43695861405197306,
+      "grad_norm": 1.6276592016220093,
+      "learning_rate": 0.00017738909631898753,
+      "loss": 1.3922,
+      "step": 11350
+    },
+    {
+      "epoch": 0.43715110683349373,
+      "grad_norm": 1.5947320461273193,
+      "learning_rate": 0.0001773699414682363,
+      "loss": 1.1952,
+      "step": 11355
+    },
+    {
+      "epoch": 0.43734359961501446,
+      "grad_norm": 1.0628368854522705,
+      "learning_rate": 0.00017735077954251648,
+      "loss": 1.3908,
+      "step": 11360
+    },
+    {
+      "epoch": 0.4375360923965351,
+      "grad_norm": 1.6347852945327759,
+      "learning_rate": 0.00017733161054358027,
+      "loss": 1.3614,
+      "step": 11365
+    },
+    {
+      "epoch": 0.4377285851780558,
+      "grad_norm": 0.98406583070755,
+      "learning_rate": 0.00017731243447318055,
+      "loss": 1.0818,
+      "step": 11370
+    },
+    {
+      "epoch": 0.4379210779595765,
+      "grad_norm": 2.522155284881592,
+      "learning_rate": 0.0001772932513330708,
+      "loss": 1.1043,
+      "step": 11375
+    },
+    {
+      "epoch": 0.4381135707410972,
+      "grad_norm": 1.3053642511367798,
+      "learning_rate": 0.0001772740611250053,
+      "loss": 1.1731,
+      "step": 11380
+    },
+    {
+      "epoch": 0.4383060635226179,
+      "grad_norm": 1.540334701538086,
+      "learning_rate": 0.0001772548638507388,
+      "loss": 1.0897,
+      "step": 11385
+    },
+    {
+      "epoch": 0.4384985563041386,
+      "grad_norm": 1.756795048713684,
+      "learning_rate": 0.00017723565951202673,
+      "loss": 1.26,
+      "step": 11390
+    },
+    {
+      "epoch": 0.4386910490856593,
+      "grad_norm": 2.263253688812256,
+      "learning_rate": 0.00017721644811062524,
+      "loss": 1.2498,
+      "step": 11395
+    },
+    {
+      "epoch": 0.43888354186718,
+      "grad_norm": 1.2686541080474854,
+      "learning_rate": 0.0001771972296482911,
+      "loss": 1.2002,
+      "step": 11400
+    },
+    {
+      "epoch": 0.43907603464870065,
+      "grad_norm": 1.7692358493804932,
+      "learning_rate": 0.00017717800412678168,
+      "loss": 1.2989,
+      "step": 11405
+    },
+    {
+      "epoch": 0.4392685274302214,
+      "grad_norm": 0.9414786100387573,
+      "learning_rate": 0.00017715877154785505,
+      "loss": 1.0743,
+      "step": 11410
+    },
+    {
+      "epoch": 0.43946102021174205,
+      "grad_norm": 1.6488560438156128,
+      "learning_rate": 0.0001771395319132699,
+      "loss": 1.3217,
+      "step": 11415
+    },
+    {
+      "epoch": 0.4396535129932628,
+      "grad_norm": 0.9546147584915161,
+      "learning_rate": 0.00017712028522478556,
+      "loss": 1.1849,
+      "step": 11420
+    },
+    {
+      "epoch": 0.43984600577478344,
+      "grad_norm": 1.9460307359695435,
+      "learning_rate": 0.000177101031484162,
+      "loss": 1.3702,
+      "step": 11425
+    },
+    {
+      "epoch": 0.4400384985563041,
+      "grad_norm": 0.8990427255630493,
+      "learning_rate": 0.00017708177069315987,
+      "loss": 1.2009,
+      "step": 11430
+    },
+    {
+      "epoch": 0.44023099133782484,
+      "grad_norm": 1.3581219911575317,
+      "learning_rate": 0.0001770625028535404,
+      "loss": 1.1846,
+      "step": 11435
+    },
+    {
+      "epoch": 0.4404234841193455,
+      "grad_norm": 1.259728193283081,
+      "learning_rate": 0.00017704322796706557,
+      "loss": 1.2683,
+      "step": 11440
+    },
+    {
+      "epoch": 0.44061597690086624,
+      "grad_norm": 1.1262446641921997,
+      "learning_rate": 0.00017702394603549788,
+      "loss": 1.0015,
+      "step": 11445
+    },
+    {
+      "epoch": 0.4408084696823869,
+      "grad_norm": 2.5833356380462646,
+      "learning_rate": 0.0001770046570606006,
+      "loss": 1.1348,
+      "step": 11450
+    },
+    {
+      "epoch": 0.44100096246390763,
+      "grad_norm": 0.9725410342216492,
+      "learning_rate": 0.00017698536104413749,
+      "loss": 1.2214,
+      "step": 11455
+    },
+    {
+      "epoch": 0.4411934552454283,
+      "grad_norm": 1.0890756845474243,
+      "learning_rate": 0.00017696605798787313,
+      "loss": 1.1178,
+      "step": 11460
+    },
+    {
+      "epoch": 0.44138594802694897,
+      "grad_norm": 1.3130367994308472,
+      "learning_rate": 0.0001769467478935726,
+      "loss": 1.1582,
+      "step": 11465
+    },
+    {
+      "epoch": 0.4415784408084697,
+      "grad_norm": 2.401630163192749,
+      "learning_rate": 0.00017692743076300172,
+      "loss": 1.3043,
+      "step": 11470
+    },
+    {
+      "epoch": 0.44177093358999037,
+      "grad_norm": 1.2928016185760498,
+      "learning_rate": 0.00017690810659792686,
+      "loss": 1.1354,
+      "step": 11475
+    },
+    {
+      "epoch": 0.4419634263715111,
+      "grad_norm": 1.6433988809585571,
+      "learning_rate": 0.00017688877540011517,
+      "loss": 1.2385,
+      "step": 11480
+    },
+    {
+      "epoch": 0.44215591915303176,
+      "grad_norm": 1.5450482368469238,
+      "learning_rate": 0.00017686943717133428,
+      "loss": 1.3096,
+      "step": 11485
+    },
+    {
+      "epoch": 0.44234841193455243,
+      "grad_norm": 2.0231974124908447,
+      "learning_rate": 0.00017685009191335257,
+      "loss": 1.2129,
+      "step": 11490
+    },
+    {
+      "epoch": 0.44254090471607316,
+      "grad_norm": 1.2831270694732666,
+      "learning_rate": 0.00017683073962793908,
+      "loss": 1.2213,
+      "step": 11495
+    },
+    {
+      "epoch": 0.44273339749759383,
+      "grad_norm": 1.035520076751709,
+      "learning_rate": 0.00017681138031686337,
+      "loss": 1.248,
+      "step": 11500
+    },
+    {
+      "epoch": 0.44292589027911455,
+      "grad_norm": 1.113934874534607,
+      "learning_rate": 0.00017679201398189577,
+      "loss": 1.0017,
+      "step": 11505
+    },
+    {
+      "epoch": 0.4431183830606352,
+      "grad_norm": 1.202412724494934,
+      "learning_rate": 0.0001767726406248072,
+      "loss": 1.0873,
+      "step": 11510
+    },
+    {
+      "epoch": 0.4433108758421559,
+      "grad_norm": 1.2946287393569946,
+      "learning_rate": 0.0001767532602473692,
+      "loss": 1.3873,
+      "step": 11515
+    },
+    {
+      "epoch": 0.4435033686236766,
+      "grad_norm": 1.2840358018875122,
+      "learning_rate": 0.00017673387285135398,
+      "loss": 1.2559,
+      "step": 11520
+    },
+    {
+      "epoch": 0.4436958614051973,
+      "grad_norm": 0.9422056078910828,
+      "learning_rate": 0.00017671447843853444,
+      "loss": 1.1179,
+      "step": 11525
+    },
+    {
+      "epoch": 0.443888354186718,
+      "grad_norm": 1.9112647771835327,
+      "learning_rate": 0.000176695077010684,
+      "loss": 1.1519,
+      "step": 11530
+    },
+    {
+      "epoch": 0.4440808469682387,
+      "grad_norm": 0.9463594555854797,
+      "learning_rate": 0.00017667566856957687,
+      "loss": 1.1175,
+      "step": 11535
+    },
+    {
+      "epoch": 0.4442733397497594,
+      "grad_norm": 2.1585206985473633,
+      "learning_rate": 0.00017665625311698776,
+      "loss": 1.1535,
+      "step": 11540
+    },
+    {
+      "epoch": 0.4444658325312801,
+      "grad_norm": 1.038095474243164,
+      "learning_rate": 0.0001766368306546921,
+      "loss": 1.1633,
+      "step": 11545
+    },
+    {
+      "epoch": 0.44465832531280075,
+      "grad_norm": 1.4679070711135864,
+      "learning_rate": 0.00017661740118446594,
+      "loss": 1.3792,
+      "step": 11550
+    },
+    {
+      "epoch": 0.4448508180943215,
+      "grad_norm": 1.3058511018753052,
+      "learning_rate": 0.00017659796470808597,
+      "loss": 1.2802,
+      "step": 11555
+    },
+    {
+      "epoch": 0.44504331087584215,
+      "grad_norm": 1.0330942869186401,
+      "learning_rate": 0.0001765785212273296,
+      "loss": 1.1621,
+      "step": 11560
+    },
+    {
+      "epoch": 0.44523580365736287,
+      "grad_norm": 1.6481776237487793,
+      "learning_rate": 0.0001765590707439747,
+      "loss": 1.1098,
+      "step": 11565
+    },
+    {
+      "epoch": 0.44542829643888354,
+      "grad_norm": 1.3850781917572021,
+      "learning_rate": 0.00017653961325979998,
+      "loss": 1.3687,
+      "step": 11570
+    },
+    {
+      "epoch": 0.4456207892204042,
+      "grad_norm": 1.6551322937011719,
+      "learning_rate": 0.0001765201487765846,
+      "loss": 1.3436,
+      "step": 11575
+    },
+    {
+      "epoch": 0.44581328200192494,
+      "grad_norm": 1.0752167701721191,
+      "learning_rate": 0.00017650067729610856,
+      "loss": 1.1667,
+      "step": 11580
+    },
+    {
+      "epoch": 0.4460057747834456,
+      "grad_norm": 1.4762775897979736,
+      "learning_rate": 0.00017648119882015232,
+      "loss": 1.0119,
+      "step": 11585
+    },
+    {
+      "epoch": 0.44619826756496633,
+      "grad_norm": 0.7833762764930725,
+      "learning_rate": 0.0001764617133504971,
+      "loss": 1.1047,
+      "step": 11590
+    },
+    {
+      "epoch": 0.446390760346487,
+      "grad_norm": 1.1666022539138794,
+      "learning_rate": 0.00017644222088892473,
+      "loss": 1.2339,
+      "step": 11595
+    },
+    {
+      "epoch": 0.4465832531280077,
+      "grad_norm": 1.7897813320159912,
+      "learning_rate": 0.0001764227214372176,
+      "loss": 1.205,
+      "step": 11600
+    },
+    {
+      "epoch": 0.4467757459095284,
+      "grad_norm": 1.2021222114562988,
+      "learning_rate": 0.00017640321499715888,
+      "loss": 1.2518,
+      "step": 11605
+    },
+    {
+      "epoch": 0.44696823869104907,
+      "grad_norm": 2.9843320846557617,
+      "learning_rate": 0.00017638370157053228,
+      "loss": 0.9705,
+      "step": 11610
+    },
+    {
+      "epoch": 0.4471607314725698,
+      "grad_norm": 1.2910903692245483,
+      "learning_rate": 0.00017636418115912213,
+      "loss": 1.4018,
+      "step": 11615
+    },
+    {
+      "epoch": 0.44735322425409046,
+      "grad_norm": 1.0188699960708618,
+      "learning_rate": 0.00017634855980214943,
+      "loss": 1.6206,
+      "step": 11620
+    },
+    {
+      "epoch": 0.4475457170356112,
+      "grad_norm": 1.0419138669967651,
+      "learning_rate": 0.00017632902682262764,
+      "loss": 1.2483,
+      "step": 11625
+    },
+    {
+      "epoch": 0.44773820981713186,
+      "grad_norm": 1.665586233139038,
+      "learning_rate": 0.000176309486863322,
+      "loss": 1.1838,
+      "step": 11630
+    },
+    {
+      "epoch": 0.44793070259865253,
+      "grad_norm": 2.3444008827209473,
+      "learning_rate": 0.00017628993992601925,
+      "loss": 1.291,
+      "step": 11635
+    },
+    {
+      "epoch": 0.44812319538017326,
+      "grad_norm": 1.960339069366455,
+      "learning_rate": 0.00017627038601250686,
+      "loss": 1.2312,
+      "step": 11640
+    },
+    {
+      "epoch": 0.4483156881616939,
+      "grad_norm": 1.5672719478607178,
+      "learning_rate": 0.00017625082512457297,
+      "loss": 1.2281,
+      "step": 11645
+    },
+    {
+      "epoch": 0.44850818094321465,
+      "grad_norm": 1.5053352117538452,
+      "learning_rate": 0.00017623125726400621,
+      "loss": 1.1688,
+      "step": 11650
+    },
+    {
+      "epoch": 0.4487006737247353,
+      "grad_norm": 1.841610312461853,
+      "learning_rate": 0.00017621168243259596,
+      "loss": 1.1607,
+      "step": 11655
+    },
+    {
+      "epoch": 0.448893166506256,
+      "grad_norm": 1.1526665687561035,
+      "learning_rate": 0.0001761921006321322,
+      "loss": 1.1788,
+      "step": 11660
+    },
+    {
+      "epoch": 0.4490856592877767,
+      "grad_norm": 1.4064139127731323,
+      "learning_rate": 0.00017617251186440556,
+      "loss": 1.0825,
+      "step": 11665
+    },
+    {
+      "epoch": 0.4492781520692974,
+      "grad_norm": 1.1119096279144287,
+      "learning_rate": 0.00017615291613120736,
+      "loss": 1.2768,
+      "step": 11670
+    },
+    {
+      "epoch": 0.4494706448508181,
+      "grad_norm": 1.2367806434631348,
+      "learning_rate": 0.00017613331343432938,
+      "loss": 1.2612,
+      "step": 11675
+    },
+    {
+      "epoch": 0.4496631376323388,
+      "grad_norm": 1.093410611152649,
+      "learning_rate": 0.00017611370377556423,
+      "loss": 1.3075,
+      "step": 11680
+    },
+    {
+      "epoch": 0.44985563041385945,
+      "grad_norm": 1.0085220336914062,
+      "learning_rate": 0.00017609408715670512,
+      "loss": 1.2391,
+      "step": 11685
+    },
+    {
+      "epoch": 0.4500481231953802,
+      "grad_norm": 1.4346550703048706,
+      "learning_rate": 0.0001760744635795458,
+      "loss": 1.2241,
+      "step": 11690
+    },
+    {
+      "epoch": 0.45024061597690085,
+      "grad_norm": 1.483905553817749,
+      "learning_rate": 0.0001760548330458807,
+      "loss": 1.2696,
+      "step": 11695
+    },
+    {
+      "epoch": 0.4504331087584216,
+      "grad_norm": 1.6455215215682983,
+      "learning_rate": 0.00017603519555750498,
+      "loss": 1.2113,
+      "step": 11700
+    },
+    {
+      "epoch": 0.45062560153994224,
+      "grad_norm": 1.7613027095794678,
+      "learning_rate": 0.00017601555111621428,
+      "loss": 1.1581,
+      "step": 11705
+    },
+    {
+      "epoch": 0.45081809432146297,
+      "grad_norm": 1.5872759819030762,
+      "learning_rate": 0.000175995899723805,
+      "loss": 1.0977,
+      "step": 11710
+    },
+    {
+      "epoch": 0.45101058710298364,
+      "grad_norm": 1.5521520376205444,
+      "learning_rate": 0.00017597624138207413,
+      "loss": 1.3003,
+      "step": 11715
+    },
+    {
+      "epoch": 0.4512030798845043,
+      "grad_norm": 2.1746668815612793,
+      "learning_rate": 0.0001759565760928193,
+      "loss": 1.1861,
+      "step": 11720
+    },
+    {
+      "epoch": 0.45139557266602504,
+      "grad_norm": 1.73439359664917,
+      "learning_rate": 0.00017593690385783866,
+      "loss": 1.242,
+      "step": 11725
+    },
+    {
+      "epoch": 0.4515880654475457,
+      "grad_norm": 1.6027134656906128,
+      "learning_rate": 0.0001759172246789313,
+      "loss": 1.2936,
+      "step": 11730
+    },
+    {
+      "epoch": 0.45178055822906643,
+      "grad_norm": 1.62489652633667,
+      "learning_rate": 0.0001758975385578966,
+      "loss": 1.3521,
+      "step": 11735
+    },
+    {
+      "epoch": 0.4519730510105871,
+      "grad_norm": 1.3407773971557617,
+      "learning_rate": 0.00017587784549653477,
+      "loss": 1.1653,
+      "step": 11740
+    },
+    {
+      "epoch": 0.45216554379210777,
+      "grad_norm": 2.064875364303589,
+      "learning_rate": 0.00017585814549664664,
+      "loss": 1.2321,
+      "step": 11745
+    },
+    {
+      "epoch": 0.4523580365736285,
+      "grad_norm": 1.115850806236267,
+      "learning_rate": 0.0001758384385600336,
+      "loss": 1.0289,
+      "step": 11750
+    },
+    {
+      "epoch": 0.45255052935514917,
+      "grad_norm": 1.3943949937820435,
+      "learning_rate": 0.00017581872468849777,
+      "loss": 1.2846,
+      "step": 11755
+    },
+    {
+      "epoch": 0.4527430221366699,
+      "grad_norm": 1.0405654907226562,
+      "learning_rate": 0.0001757990038838418,
+      "loss": 1.1209,
+      "step": 11760
+    },
+    {
+      "epoch": 0.45293551491819056,
+      "grad_norm": 1.0115854740142822,
+      "learning_rate": 0.00017577927614786902,
+      "loss": 1.0178,
+      "step": 11765
+    },
+    {
+      "epoch": 0.4531280076997113,
+      "grad_norm": 2.48100209236145,
+      "learning_rate": 0.00017575954148238345,
+      "loss": 1.2485,
+      "step": 11770
+    },
+    {
+      "epoch": 0.45332050048123196,
+      "grad_norm": 1.5187568664550781,
+      "learning_rate": 0.00017573979988918967,
+      "loss": 1.3345,
+      "step": 11775
+    },
+    {
+      "epoch": 0.45351299326275263,
+      "grad_norm": 1.2286217212677002,
+      "learning_rate": 0.00017572005137009292,
+      "loss": 1.1079,
+      "step": 11780
+    },
+    {
+      "epoch": 0.45370548604427335,
+      "grad_norm": 1.5858092308044434,
+      "learning_rate": 0.00017570029592689908,
+      "loss": 1.4054,
+      "step": 11785
+    },
+    {
+      "epoch": 0.453897978825794,
+      "grad_norm": 2.0436697006225586,
+      "learning_rate": 0.00017568053356141464,
+      "loss": 1.3221,
+      "step": 11790
+    },
+    {
+      "epoch": 0.45409047160731475,
+      "grad_norm": 1.6980565786361694,
+      "learning_rate": 0.00017566076427544673,
+      "loss": 1.2384,
+      "step": 11795
+    },
+    {
+      "epoch": 0.4542829643888354,
+      "grad_norm": 1.3811545372009277,
+      "learning_rate": 0.00017564098807080315,
+      "loss": 1.171,
+      "step": 11800
+    },
+    {
+      "epoch": 0.4544754571703561,
+      "grad_norm": 1.2215286493301392,
+      "learning_rate": 0.00017562120494929228,
+      "loss": 1.1781,
+      "step": 11805
+    },
+    {
+      "epoch": 0.4546679499518768,
+      "grad_norm": 1.1313782930374146,
+      "learning_rate": 0.00017560141491272319,
+      "loss": 1.2166,
+      "step": 11810
+    },
+    {
+      "epoch": 0.4548604427333975,
+      "grad_norm": 1.2630988359451294,
+      "learning_rate": 0.0001755816179629055,
+      "loss": 1.2652,
+      "step": 11815
+    },
+    {
+      "epoch": 0.4550529355149182,
+      "grad_norm": 1.0977842807769775,
+      "learning_rate": 0.0001755618141016495,
+      "loss": 1.3057,
+      "step": 11820
+    },
+    {
+      "epoch": 0.4552454282964389,
+      "grad_norm": 0.8517459034919739,
+      "learning_rate": 0.0001755420033307662,
+      "loss": 1.1769,
+      "step": 11825
+    },
+    {
+      "epoch": 0.45543792107795955,
+      "grad_norm": 0.7195164561271667,
+      "learning_rate": 0.00017552218565206707,
+      "loss": 0.9777,
+      "step": 11830
+    },
+    {
+      "epoch": 0.4556304138594803,
+      "grad_norm": 1.125056266784668,
+      "learning_rate": 0.00017550236106736436,
+      "loss": 1.1008,
+      "step": 11835
+    },
+    {
+      "epoch": 0.45582290664100095,
+      "grad_norm": 0.8211593627929688,
+      "learning_rate": 0.00017548252957847092,
+      "loss": 1.1539,
+      "step": 11840
+    },
+    {
+      "epoch": 0.4560153994225217,
+      "grad_norm": 1.8936784267425537,
+      "learning_rate": 0.00017546269118720015,
+      "loss": 1.143,
+      "step": 11845
+    },
+    {
+      "epoch": 0.45620789220404234,
+      "grad_norm": 1.5479308366775513,
+      "learning_rate": 0.00017544284589536617,
+      "loss": 1.1481,
+      "step": 11850
+    },
+    {
+      "epoch": 0.45640038498556307,
+      "grad_norm": 1.4597593545913696,
+      "learning_rate": 0.00017542299370478372,
+      "loss": 1.2907,
+      "step": 11855
+    },
+    {
+      "epoch": 0.45659287776708374,
+      "grad_norm": 1.4036239385604858,
+      "learning_rate": 0.0001754031346172681,
+      "loss": 1.2927,
+      "step": 11860
+    },
+    {
+      "epoch": 0.4567853705486044,
+      "grad_norm": 0.9842814207077026,
+      "learning_rate": 0.00017538326863463533,
+      "loss": 1.0571,
+      "step": 11865
+    },
+    {
+      "epoch": 0.45697786333012513,
+      "grad_norm": 2.478254556655884,
+      "learning_rate": 0.000175363395758702,
+      "loss": 1.3115,
+      "step": 11870
+    },
+    {
+      "epoch": 0.4571703561116458,
+      "grad_norm": 1.000182032585144,
+      "learning_rate": 0.00017534351599128538,
+      "loss": 1.3071,
+      "step": 11875
+    },
+    {
+      "epoch": 0.45736284889316653,
+      "grad_norm": 1.8669004440307617,
+      "learning_rate": 0.0001753236293342033,
+      "loss": 1.2386,
+      "step": 11880
+    },
+    {
+      "epoch": 0.4575553416746872,
+      "grad_norm": 1.6287200450897217,
+      "learning_rate": 0.00017530373578927432,
+      "loss": 1.2196,
+      "step": 11885
+    },
+    {
+      "epoch": 0.45774783445620787,
+      "grad_norm": 2.1733322143554688,
+      "learning_rate": 0.00017528383535831755,
+      "loss": 1.6165,
+      "step": 11890
+    },
+    {
+      "epoch": 0.4579403272377286,
+      "grad_norm": 1.0370094776153564,
+      "learning_rate": 0.00017526392804315273,
+      "loss": 1.1799,
+      "step": 11895
+    },
+    {
+      "epoch": 0.45813282001924927,
+      "grad_norm": 1.3969937562942505,
+      "learning_rate": 0.00017524401384560025,
+      "loss": 1.2224,
+      "step": 11900
+    },
+    {
+      "epoch": 0.45832531280077,
+      "grad_norm": 1.3850924968719482,
+      "learning_rate": 0.00017522409276748117,
+      "loss": 1.4161,
+      "step": 11905
+    },
+    {
+      "epoch": 0.45851780558229066,
+      "grad_norm": 1.4318947792053223,
+      "learning_rate": 0.00017520416481061712,
+      "loss": 1.4166,
+      "step": 11910
+    },
+    {
+      "epoch": 0.45871029836381133,
+      "grad_norm": 1.525709629058838,
+      "learning_rate": 0.00017518422997683038,
+      "loss": 1.255,
+      "step": 11915
+    },
+    {
+      "epoch": 0.45890279114533206,
+      "grad_norm": 0.9193233847618103,
+      "learning_rate": 0.00017516428826794384,
+      "loss": 1.2299,
+      "step": 11920
+    },
+    {
+      "epoch": 0.4590952839268527,
+      "grad_norm": 1.8636525869369507,
+      "learning_rate": 0.00017514433968578107,
+      "loss": 1.1992,
+      "step": 11925
+    },
+    {
+      "epoch": 0.45928777670837345,
+      "grad_norm": 1.3876943588256836,
+      "learning_rate": 0.00017512438423216624,
+      "loss": 1.2022,
+      "step": 11930
+    },
+    {
+      "epoch": 0.4594802694898941,
+      "grad_norm": 1.5370129346847534,
+      "learning_rate": 0.00017510442190892412,
+      "loss": 1.319,
+      "step": 11935
+    },
+    {
+      "epoch": 0.45967276227141485,
+      "grad_norm": 1.8562203645706177,
+      "learning_rate": 0.00017508445271788013,
+      "loss": 1.0784,
+      "step": 11940
+    },
+    {
+      "epoch": 0.4598652550529355,
+      "grad_norm": 1.1265978813171387,
+      "learning_rate": 0.0001750644766608603,
+      "loss": 1.1591,
+      "step": 11945
+    },
+    {
+      "epoch": 0.4600577478344562,
+      "grad_norm": 1.3049321174621582,
+      "learning_rate": 0.00017504449373969137,
+      "loss": 1.2567,
+      "step": 11950
+    },
+    {
+      "epoch": 0.4602502406159769,
+      "grad_norm": 1.4252487421035767,
+      "learning_rate": 0.0001750245039562006,
+      "loss": 1.0848,
+      "step": 11955
+    },
+    {
+      "epoch": 0.4604427333974976,
+      "grad_norm": 1.888185977935791,
+      "learning_rate": 0.00017500450731221592,
+      "loss": 1.1976,
+      "step": 11960
+    },
+    {
+      "epoch": 0.4606352261790183,
+      "grad_norm": 1.4043982028961182,
+      "learning_rate": 0.00017498450380956594,
+      "loss": 1.2038,
+      "step": 11965
+    },
+    {
+      "epoch": 0.460827718960539,
+      "grad_norm": 1.182576060295105,
+      "learning_rate": 0.00017496449345007982,
+      "loss": 1.3408,
+      "step": 11970
+    },
+    {
+      "epoch": 0.46102021174205965,
+      "grad_norm": 2.084197521209717,
+      "learning_rate": 0.00017494447623558733,
+      "loss": 1.269,
+      "step": 11975
+    },
+    {
+      "epoch": 0.4612127045235804,
+      "grad_norm": 1.709518551826477,
+      "learning_rate": 0.00017492445216791896,
+      "loss": 1.3736,
+      "step": 11980
+    },
+    {
+      "epoch": 0.46140519730510104,
+      "grad_norm": 1.1446977853775024,
+      "learning_rate": 0.00017490442124890577,
+      "loss": 1.2449,
+      "step": 11985
+    },
+    {
+      "epoch": 0.46159769008662177,
+      "grad_norm": 1.9139240980148315,
+      "learning_rate": 0.00017488438348037946,
+      "loss": 1.0845,
+      "step": 11990
+    },
+    {
+      "epoch": 0.46179018286814244,
+      "grad_norm": 1.6536133289337158,
+      "learning_rate": 0.00017486433886417234,
+      "loss": 1.4398,
+      "step": 11995
+    },
+    {
+      "epoch": 0.4619826756496631,
+      "grad_norm": 1.0629438161849976,
+      "learning_rate": 0.00017484428740211736,
+      "loss": 1.2631,
+      "step": 12000
+    },
+    {
+      "epoch": 0.46217516843118384,
+      "grad_norm": 1.1966623067855835,
+      "learning_rate": 0.00017482422909604809,
+      "loss": 1.088,
+      "step": 12005
+    },
+    {
+      "epoch": 0.4623676612127045,
+      "grad_norm": 1.1087130308151245,
+      "learning_rate": 0.00017480416394779878,
+      "loss": 1.2133,
+      "step": 12010
+    },
+    {
+      "epoch": 0.46256015399422523,
+      "grad_norm": 1.9408375024795532,
+      "learning_rate": 0.00017478409195920413,
+      "loss": 1.1677,
+      "step": 12015
+    },
+    {
+      "epoch": 0.4627526467757459,
+      "grad_norm": 1.2703943252563477,
+      "learning_rate": 0.00017476401313209973,
+      "loss": 1.232,
+      "step": 12020
+    },
+    {
+      "epoch": 0.4629451395572666,
+      "grad_norm": 1.7841099500656128,
+      "learning_rate": 0.0001747439274683216,
+      "loss": 1.1688,
+      "step": 12025
+    },
+    {
+      "epoch": 0.4631376323387873,
+      "grad_norm": 1.9395395517349243,
+      "learning_rate": 0.0001747238349697064,
+      "loss": 1.2336,
+      "step": 12030
+    },
+    {
+      "epoch": 0.46333012512030797,
+      "grad_norm": 1.5011239051818848,
+      "learning_rate": 0.0001747037356380915,
+      "loss": 1.1849,
+      "step": 12035
+    },
+    {
+      "epoch": 0.4635226179018287,
+      "grad_norm": 1.6130584478378296,
+      "learning_rate": 0.00017468362947531486,
+      "loss": 1.3113,
+      "step": 12040
+    },
+    {
+      "epoch": 0.46371511068334936,
+      "grad_norm": 1.5666422843933105,
+      "learning_rate": 0.000174663516483215,
+      "loss": 1.2178,
+      "step": 12045
+    },
+    {
+      "epoch": 0.4639076034648701,
+      "grad_norm": 1.490662932395935,
+      "learning_rate": 0.0001746433966636312,
+      "loss": 1.3034,
+      "step": 12050
+    },
+    {
+      "epoch": 0.46410009624639076,
+      "grad_norm": 1.1972042322158813,
+      "learning_rate": 0.00017462327001840322,
+      "loss": 1.1732,
+      "step": 12055
+    },
+    {
+      "epoch": 0.46429258902791143,
+      "grad_norm": 1.5201470851898193,
+      "learning_rate": 0.00017460313654937154,
+      "loss": 1.1545,
+      "step": 12060
+    },
+    {
+      "epoch": 0.46448508180943215,
+      "grad_norm": 0.8927121758460999,
+      "learning_rate": 0.00017458299625837723,
+      "loss": 1.1516,
+      "step": 12065
+    },
+    {
+      "epoch": 0.4646775745909528,
+      "grad_norm": 1.394187092781067,
+      "learning_rate": 0.00017456284914726196,
+      "loss": 1.2791,
+      "step": 12070
+    },
+    {
+      "epoch": 0.46487006737247355,
+      "grad_norm": 1.8900322914123535,
+      "learning_rate": 0.00017454269521786808,
+      "loss": 1.244,
+      "step": 12075
+    },
+    {
+      "epoch": 0.4650625601539942,
+      "grad_norm": 2.20624041557312,
+      "learning_rate": 0.00017452253447203852,
+      "loss": 1.2526,
+      "step": 12080
+    },
+    {
+      "epoch": 0.46525505293551495,
+      "grad_norm": 1.404261827468872,
+      "learning_rate": 0.00017450236691161686,
+      "loss": 1.1711,
+      "step": 12085
+    },
+    {
+      "epoch": 0.4654475457170356,
+      "grad_norm": 1.6828880310058594,
+      "learning_rate": 0.00017448219253844726,
+      "loss": 1.3007,
+      "step": 12090
+    },
+    {
+      "epoch": 0.4656400384985563,
+      "grad_norm": 1.0239325761795044,
+      "learning_rate": 0.00017446201135437456,
+      "loss": 1.1359,
+      "step": 12095
+    },
+    {
+      "epoch": 0.465832531280077,
+      "grad_norm": 0.9242125749588013,
+      "learning_rate": 0.0001744418233612442,
+      "loss": 1.1848,
+      "step": 12100
+    },
+    {
+      "epoch": 0.4660250240615977,
+      "grad_norm": 2.9907031059265137,
+      "learning_rate": 0.0001744216285609022,
+      "loss": 1.155,
+      "step": 12105
+    },
+    {
+      "epoch": 0.4662175168431184,
+      "grad_norm": 0.9708018898963928,
+      "learning_rate": 0.0001744014269551953,
+      "loss": 1.4752,
+      "step": 12110
+    },
+    {
+      "epoch": 0.4664100096246391,
+      "grad_norm": 1.1917387247085571,
+      "learning_rate": 0.00017438121854597075,
+      "loss": 1.1197,
+      "step": 12115
+    },
+    {
+      "epoch": 0.46660250240615975,
+      "grad_norm": 1.5464357137680054,
+      "learning_rate": 0.00017436100333507648,
+      "loss": 1.1908,
+      "step": 12120
+    },
+    {
+      "epoch": 0.4667949951876805,
+      "grad_norm": 1.9502155780792236,
+      "learning_rate": 0.00017434078132436107,
+      "loss": 1.1888,
+      "step": 12125
+    },
+    {
+      "epoch": 0.46698748796920114,
+      "grad_norm": 2.054029941558838,
+      "learning_rate": 0.00017432055251567365,
+      "loss": 1.2771,
+      "step": 12130
+    },
+    {
+      "epoch": 0.46717998075072187,
+      "grad_norm": 1.276356816291809,
+      "learning_rate": 0.00017430031691086407,
+      "loss": 1.2392,
+      "step": 12135
+    },
+    {
+      "epoch": 0.46737247353224254,
+      "grad_norm": 1.4474079608917236,
+      "learning_rate": 0.00017428007451178267,
+      "loss": 1.2111,
+      "step": 12140
+    },
+    {
+      "epoch": 0.4675649663137632,
+      "grad_norm": 1.389797568321228,
+      "learning_rate": 0.00017425982532028053,
+      "loss": 1.2094,
+      "step": 12145
+    },
+    {
+      "epoch": 0.46775745909528393,
+      "grad_norm": 1.2491530179977417,
+      "learning_rate": 0.00017423956933820928,
+      "loss": 1.2374,
+      "step": 12150
+    },
+    {
+      "epoch": 0.4679499518768046,
+      "grad_norm": 1.0517950057983398,
+      "learning_rate": 0.00017421930656742122,
+      "loss": 1.1003,
+      "step": 12155
+    },
+    {
+      "epoch": 0.46814244465832533,
+      "grad_norm": 1.410630226135254,
+      "learning_rate": 0.00017419903700976924,
+      "loss": 1.2722,
+      "step": 12160
+    },
+    {
+      "epoch": 0.468334937439846,
+      "grad_norm": 1.5544359683990479,
+      "learning_rate": 0.00017417876066710682,
+      "loss": 1.1961,
+      "step": 12165
+    },
+    {
+      "epoch": 0.4685274302213667,
+      "grad_norm": 1.3200881481170654,
+      "learning_rate": 0.00017415847754128817,
+      "loss": 1.0058,
+      "step": 12170
+    },
+    {
+      "epoch": 0.4687199230028874,
+      "grad_norm": 1.998949646949768,
+      "learning_rate": 0.00017413818763416795,
+      "loss": 1.1513,
+      "step": 12175
+    },
+    {
+      "epoch": 0.46891241578440807,
+      "grad_norm": 1.4105117321014404,
+      "learning_rate": 0.0001741178909476016,
+      "loss": 1.2993,
+      "step": 12180
+    },
+    {
+      "epoch": 0.4691049085659288,
+      "grad_norm": 1.4521151781082153,
+      "learning_rate": 0.00017409758748344515,
+      "loss": 1.1659,
+      "step": 12185
+    },
+    {
+      "epoch": 0.46929740134744946,
+      "grad_norm": 1.3822886943817139,
+      "learning_rate": 0.00017407727724355515,
+      "loss": 1.3419,
+      "step": 12190
+    },
+    {
+      "epoch": 0.4694898941289702,
+      "grad_norm": 1.5602283477783203,
+      "learning_rate": 0.00017405696022978885,
+      "loss": 1.1506,
+      "step": 12195
+    },
+    {
+      "epoch": 0.46968238691049086,
+      "grad_norm": 1.2674669027328491,
+      "learning_rate": 0.00017403663644400413,
+      "loss": 1.2992,
+      "step": 12200
+    },
+    {
+      "epoch": 0.4698748796920115,
+      "grad_norm": 1.6091759204864502,
+      "learning_rate": 0.00017401630588805947,
+      "loss": 1.1105,
+      "step": 12205
+    },
+    {
+      "epoch": 0.47006737247353225,
+      "grad_norm": 1.591635823249817,
+      "learning_rate": 0.00017399596856381395,
+      "loss": 1.2884,
+      "step": 12210
+    },
+    {
+      "epoch": 0.4702598652550529,
+      "grad_norm": 1.5781102180480957,
+      "learning_rate": 0.00017397562447312725,
+      "loss": 1.1476,
+      "step": 12215
+    },
+    {
+      "epoch": 0.47045235803657365,
+      "grad_norm": 1.4029310941696167,
+      "learning_rate": 0.00017395527361785976,
+      "loss": 1.4271,
+      "step": 12220
+    },
+    {
+      "epoch": 0.4706448508180943,
+      "grad_norm": 1.8287990093231201,
+      "learning_rate": 0.0001739349159998724,
+      "loss": 1.2079,
+      "step": 12225
+    },
+    {
+      "epoch": 0.470837343599615,
+      "grad_norm": 0.9693268537521362,
+      "learning_rate": 0.00017391455162102677,
+      "loss": 1.2341,
+      "step": 12230
+    },
+    {
+      "epoch": 0.4710298363811357,
+      "grad_norm": 1.4181095361709595,
+      "learning_rate": 0.00017389418048318502,
+      "loss": 1.2796,
+      "step": 12235
+    },
+    {
+      "epoch": 0.4712223291626564,
+      "grad_norm": 1.9247058629989624,
+      "learning_rate": 0.00017387380258820993,
+      "loss": 1.1858,
+      "step": 12240
+    },
+    {
+      "epoch": 0.4714148219441771,
+      "grad_norm": 1.0236104726791382,
+      "learning_rate": 0.00017385341793796502,
+      "loss": 1.1713,
+      "step": 12245
+    },
+    {
+      "epoch": 0.4716073147256978,
+      "grad_norm": 1.0250846147537231,
+      "learning_rate": 0.00017383302653431427,
+      "loss": 1.3036,
+      "step": 12250
+    },
+    {
+      "epoch": 0.4717998075072185,
+      "grad_norm": 1.1760774850845337,
+      "learning_rate": 0.00017381262837912228,
+      "loss": 1.1779,
+      "step": 12255
+    },
+    {
+      "epoch": 0.4719923002887392,
+      "grad_norm": 1.6482713222503662,
+      "learning_rate": 0.00017379222347425446,
+      "loss": 1.151,
+      "step": 12260
+    },
+    {
+      "epoch": 0.47218479307025985,
+      "grad_norm": 1.3430352210998535,
+      "learning_rate": 0.00017377181182157657,
+      "loss": 1.3512,
+      "step": 12265
+    },
+    {
+      "epoch": 0.47237728585178057,
+      "grad_norm": 2.4042775630950928,
+      "learning_rate": 0.00017375139342295522,
+      "loss": 1.3002,
+      "step": 12270
+    },
+    {
+      "epoch": 0.47256977863330124,
+      "grad_norm": 0.967472493648529,
+      "learning_rate": 0.00017373096828025752,
+      "loss": 1.0813,
+      "step": 12275
+    },
+    {
+      "epoch": 0.47276227141482197,
+      "grad_norm": 1.9774664640426636,
+      "learning_rate": 0.00017371053639535117,
+      "loss": 1.2232,
+      "step": 12280
+    },
+    {
+      "epoch": 0.47295476419634264,
+      "grad_norm": 1.2525962591171265,
+      "learning_rate": 0.00017369009777010454,
+      "loss": 1.3974,
+      "step": 12285
+    },
+    {
+      "epoch": 0.4731472569778633,
+      "grad_norm": 2.268892765045166,
+      "learning_rate": 0.00017366965240638664,
+      "loss": 1.2812,
+      "step": 12290
+    },
+    {
+      "epoch": 0.47333974975938403,
+      "grad_norm": 1.143028974533081,
+      "learning_rate": 0.000173649200306067,
+      "loss": 1.2017,
+      "step": 12295
+    },
+    {
+      "epoch": 0.4735322425409047,
+      "grad_norm": 0.9833802580833435,
+      "learning_rate": 0.00017362874147101596,
+      "loss": 1.1669,
+      "step": 12300
+    },
+    {
+      "epoch": 0.4737247353224254,
+      "grad_norm": 1.5986253023147583,
+      "learning_rate": 0.0001736082759031042,
+      "loss": 1.2275,
+      "step": 12305
+    },
+    {
+      "epoch": 0.4739172281039461,
+      "grad_norm": 1.8394620418548584,
+      "learning_rate": 0.0001735878036042032,
+      "loss": 1.1077,
+      "step": 12310
+    },
+    {
+      "epoch": 0.47410972088546677,
+      "grad_norm": 2.2321078777313232,
+      "learning_rate": 0.00017356732457618506,
+      "loss": 1.1385,
+      "step": 12315
+    },
+    {
+      "epoch": 0.4743022136669875,
+      "grad_norm": 1.2479119300842285,
+      "learning_rate": 0.00017354683882092245,
+      "loss": 1.2189,
+      "step": 12320
+    },
+    {
+      "epoch": 0.47449470644850816,
+      "grad_norm": 1.6812646389007568,
+      "learning_rate": 0.0001735263463402886,
+      "loss": 1.3836,
+      "step": 12325
+    },
+    {
+      "epoch": 0.4746871992300289,
+      "grad_norm": 1.4916552305221558,
+      "learning_rate": 0.00017350584713615746,
+      "loss": 1.1306,
+      "step": 12330
+    },
+    {
+      "epoch": 0.47487969201154956,
+      "grad_norm": 1.7067712545394897,
+      "learning_rate": 0.00017348534121040354,
+      "loss": 1.3352,
+      "step": 12335
+    },
+    {
+      "epoch": 0.4750721847930703,
+      "grad_norm": 1.1849184036254883,
+      "learning_rate": 0.00017346482856490196,
+      "loss": 0.9746,
+      "step": 12340
+    },
+    {
+      "epoch": 0.47526467757459095,
+      "grad_norm": 1.700038194656372,
+      "learning_rate": 0.00017344430920152845,
+      "loss": 1.3462,
+      "step": 12345
+    },
+    {
+      "epoch": 0.4754571703561116,
+      "grad_norm": 1.4579262733459473,
+      "learning_rate": 0.0001734237831221594,
+      "loss": 1.2296,
+      "step": 12350
+    },
+    {
+      "epoch": 0.47564966313763235,
+      "grad_norm": 1.230469822883606,
+      "learning_rate": 0.00017340325032867178,
+      "loss": 1.1615,
+      "step": 12355
+    },
+    {
+      "epoch": 0.475842155919153,
+      "grad_norm": 1.4839364290237427,
+      "learning_rate": 0.00017338271082294315,
+      "loss": 1.2143,
+      "step": 12360
+    },
+    {
+      "epoch": 0.47603464870067375,
+      "grad_norm": 0.8386423587799072,
+      "learning_rate": 0.00017336216460685173,
+      "loss": 1.1173,
+      "step": 12365
+    },
+    {
+      "epoch": 0.4762271414821944,
+      "grad_norm": 1.9203957319259644,
+      "learning_rate": 0.00017334161168227634,
+      "loss": 1.2371,
+      "step": 12370
+    },
+    {
+      "epoch": 0.4764196342637151,
+      "grad_norm": 1.752314567565918,
+      "learning_rate": 0.00017332105205109641,
+      "loss": 1.1022,
+      "step": 12375
+    },
+    {
+      "epoch": 0.4766121270452358,
+      "grad_norm": 1.2998472452163696,
+      "learning_rate": 0.00017330048571519198,
+      "loss": 1.3008,
+      "step": 12380
+    },
+    {
+      "epoch": 0.4768046198267565,
+      "grad_norm": 1.8506637811660767,
+      "learning_rate": 0.0001732799126764437,
+      "loss": 1.0814,
+      "step": 12385
+    },
+    {
+      "epoch": 0.4769971126082772,
+      "grad_norm": 1.4652866125106812,
+      "learning_rate": 0.00017325933293673283,
+      "loss": 1.3528,
+      "step": 12390
+    },
+    {
+      "epoch": 0.4771896053897979,
+      "grad_norm": 1.0838465690612793,
+      "learning_rate": 0.00017323874649794127,
+      "loss": 1.1435,
+      "step": 12395
+    },
+    {
+      "epoch": 0.4773820981713186,
+      "grad_norm": 1.1437288522720337,
+      "learning_rate": 0.0001732181533619515,
+      "loss": 1.2403,
+      "step": 12400
+    },
+    {
+      "epoch": 0.4775745909528393,
+      "grad_norm": 1.5026469230651855,
+      "learning_rate": 0.00017319755353064665,
+      "loss": 1.3211,
+      "step": 12405
+    },
+    {
+      "epoch": 0.47776708373435994,
+      "grad_norm": 1.477759838104248,
+      "learning_rate": 0.00017317694700591041,
+      "loss": 1.062,
+      "step": 12410
+    },
+    {
+      "epoch": 0.47795957651588067,
+      "grad_norm": 1.6099724769592285,
+      "learning_rate": 0.00017315633378962712,
+      "loss": 1.322,
+      "step": 12415
+    },
+    {
+      "epoch": 0.47815206929740134,
+      "grad_norm": 1.2413129806518555,
+      "learning_rate": 0.00017313571388368173,
+      "loss": 1.3106,
+      "step": 12420
+    },
+    {
+      "epoch": 0.47834456207892206,
+      "grad_norm": 1.2218198776245117,
+      "learning_rate": 0.00017311508728995976,
+      "loss": 1.3899,
+      "step": 12425
+    },
+    {
+      "epoch": 0.47853705486044273,
+      "grad_norm": 1.337332844734192,
+      "learning_rate": 0.0001730944540103474,
+      "loss": 1.2143,
+      "step": 12430
+    },
+    {
+      "epoch": 0.4787295476419634,
+      "grad_norm": 1.132523775100708,
+      "learning_rate": 0.00017307381404673143,
+      "loss": 1.2243,
+      "step": 12435
+    },
+    {
+      "epoch": 0.47892204042348413,
+      "grad_norm": 1.481467366218567,
+      "learning_rate": 0.00017305316740099928,
+      "loss": 1.1563,
+      "step": 12440
+    },
+    {
+      "epoch": 0.4791145332050048,
+      "grad_norm": 1.3262776136398315,
+      "learning_rate": 0.00017303251407503885,
+      "loss": 1.1995,
+      "step": 12445
+    },
+    {
+      "epoch": 0.4793070259865255,
+      "grad_norm": 1.8554911613464355,
+      "learning_rate": 0.00017301185407073884,
+      "loss": 1.2679,
+      "step": 12450
+    },
+    {
+      "epoch": 0.4794995187680462,
+      "grad_norm": 1.5468156337738037,
+      "learning_rate": 0.0001729911873899884,
+      "loss": 1.1839,
+      "step": 12455
+    },
+    {
+      "epoch": 0.47969201154956687,
+      "grad_norm": 1.1333458423614502,
+      "learning_rate": 0.00017297051403467745,
+      "loss": 1.2927,
+      "step": 12460
+    },
+    {
+      "epoch": 0.4798845043310876,
+      "grad_norm": 1.5019558668136597,
+      "learning_rate": 0.00017294983400669632,
+      "loss": 1.1634,
+      "step": 12465
+    },
+    {
+      "epoch": 0.48007699711260826,
+      "grad_norm": 0.9122928977012634,
+      "learning_rate": 0.00017292914730793614,
+      "loss": 1.0842,
+      "step": 12470
+    },
+    {
+      "epoch": 0.480269489894129,
+      "grad_norm": 3.623866558074951,
+      "learning_rate": 0.00017290845394028853,
+      "loss": 1.2303,
+      "step": 12475
+    },
+    {
+      "epoch": 0.48046198267564966,
+      "grad_norm": 1.1163458824157715,
+      "learning_rate": 0.0001728877539056458,
+      "loss": 1.1877,
+      "step": 12480
+    },
+    {
+      "epoch": 0.4806544754571704,
+      "grad_norm": 1.1737778186798096,
+      "learning_rate": 0.00017286704720590083,
+      "loss": 1.2183,
+      "step": 12485
+    },
+    {
+      "epoch": 0.48084696823869105,
+      "grad_norm": 1.0381931066513062,
+      "learning_rate": 0.0001728463338429471,
+      "loss": 1.103,
+      "step": 12490
+    },
+    {
+      "epoch": 0.4810394610202117,
+      "grad_norm": 1.4400640726089478,
+      "learning_rate": 0.00017282561381867865,
+      "loss": 1.0941,
+      "step": 12495
+    },
+    {
+      "epoch": 0.48123195380173245,
+      "grad_norm": 1.774886131286621,
+      "learning_rate": 0.00017280488713499029,
+      "loss": 1.2161,
+      "step": 12500
+    },
+    {
+      "epoch": 0.4814244465832531,
+      "grad_norm": 1.680755376815796,
+      "learning_rate": 0.00017278415379377724,
+      "loss": 1.2248,
+      "step": 12505
+    },
+    {
+      "epoch": 0.48161693936477384,
+      "grad_norm": 1.0955753326416016,
+      "learning_rate": 0.00017276341379693553,
+      "loss": 1.1558,
+      "step": 12510
+    },
+    {
+      "epoch": 0.4818094321462945,
+      "grad_norm": 1.9817503690719604,
+      "learning_rate": 0.00017274266714636163,
+      "loss": 0.9682,
+      "step": 12515
+    },
+    {
+      "epoch": 0.4820019249278152,
+      "grad_norm": 1.2484976053237915,
+      "learning_rate": 0.00017272191384395266,
+      "loss": 1.1304,
+      "step": 12520
+    },
+    {
+      "epoch": 0.4821944177093359,
+      "grad_norm": 1.263295292854309,
+      "learning_rate": 0.00017270115389160645,
+      "loss": 1.1851,
+      "step": 12525
+    },
+    {
+      "epoch": 0.4823869104908566,
+      "grad_norm": 1.749971628189087,
+      "learning_rate": 0.00017268038729122126,
+      "loss": 1.2665,
+      "step": 12530
+    },
+    {
+      "epoch": 0.4825794032723773,
+      "grad_norm": 2.1695139408111572,
+      "learning_rate": 0.0001726596140446962,
+      "loss": 1.2351,
+      "step": 12535
+    },
+    {
+      "epoch": 0.482771896053898,
+      "grad_norm": 1.8199032545089722,
+      "learning_rate": 0.00017263883415393069,
+      "loss": 1.0922,
+      "step": 12540
+    },
+    {
+      "epoch": 0.48296438883541865,
+      "grad_norm": 1.5746350288391113,
+      "learning_rate": 0.00017261804762082501,
+      "loss": 1.2759,
+      "step": 12545
+    },
+    {
+      "epoch": 0.48315688161693937,
+      "grad_norm": 1.1286424398422241,
+      "learning_rate": 0.00017259725444727995,
+      "loss": 1.2286,
+      "step": 12550
+    },
+    {
+      "epoch": 0.48334937439846004,
+      "grad_norm": 1.1256860494613647,
+      "learning_rate": 0.00017257645463519686,
+      "loss": 1.1408,
+      "step": 12555
+    },
+    {
+      "epoch": 0.48354186717998077,
+      "grad_norm": 0.907913863658905,
+      "learning_rate": 0.00017255564818647776,
+      "loss": 1.1889,
+      "step": 12560
+    },
+    {
+      "epoch": 0.48373435996150144,
+      "grad_norm": 1.0480241775512695,
+      "learning_rate": 0.0001725348351030253,
+      "loss": 1.0954,
+      "step": 12565
+    },
+    {
+      "epoch": 0.48392685274302216,
+      "grad_norm": 1.4278559684753418,
+      "learning_rate": 0.0001725140153867426,
+      "loss": 1.1367,
+      "step": 12570
+    },
+    {
+      "epoch": 0.48411934552454283,
+      "grad_norm": 0.9501696825027466,
+      "learning_rate": 0.00017249318903953364,
+      "loss": 1.2135,
+      "step": 12575
+    },
+    {
+      "epoch": 0.4843118383060635,
+      "grad_norm": 1.8626338243484497,
+      "learning_rate": 0.00017247235606330271,
+      "loss": 1.2106,
+      "step": 12580
+    },
+    {
+      "epoch": 0.48450433108758423,
+      "grad_norm": 1.3876720666885376,
+      "learning_rate": 0.00017245151645995494,
+      "loss": 1.3711,
+      "step": 12585
+    },
+    {
+      "epoch": 0.4846968238691049,
+      "grad_norm": 1.6943193674087524,
+      "learning_rate": 0.0001724306702313959,
+      "loss": 1.117,
+      "step": 12590
+    },
+    {
+      "epoch": 0.4848893166506256,
+      "grad_norm": 0.9052426218986511,
+      "learning_rate": 0.00017240981737953192,
+      "loss": 1.2578,
+      "step": 12595
+    },
+    {
+      "epoch": 0.4850818094321463,
+      "grad_norm": 0.8325613141059875,
+      "learning_rate": 0.00017238895790626976,
+      "loss": 1.1599,
+      "step": 12600
+    },
+    {
+      "epoch": 0.48527430221366696,
+      "grad_norm": 1.2736178636550903,
+      "learning_rate": 0.00017236809181351697,
+      "loss": 1.266,
+      "step": 12605
+    },
+    {
+      "epoch": 0.4854667949951877,
+      "grad_norm": 1.8093243837356567,
+      "learning_rate": 0.00017234721910318158,
+      "loss": 1.2076,
+      "step": 12610
+    },
+    {
+      "epoch": 0.48565928777670836,
+      "grad_norm": 1.7740625143051147,
+      "learning_rate": 0.00017232633977717226,
+      "loss": 1.2431,
+      "step": 12615
+    },
+    {
+      "epoch": 0.4858517805582291,
+      "grad_norm": 0.83774334192276,
+      "learning_rate": 0.00017230545383739826,
+      "loss": 1.1987,
+      "step": 12620
+    },
+    {
+      "epoch": 0.48604427333974976,
+      "grad_norm": 0.987789511680603,
+      "learning_rate": 0.0001722845612857695,
+      "loss": 1.0,
+      "step": 12625
+    },
+    {
+      "epoch": 0.4862367661212704,
+      "grad_norm": 1.1741127967834473,
+      "learning_rate": 0.00017226366212419645,
+      "loss": 1.1961,
+      "step": 12630
+    },
+    {
+      "epoch": 0.48642925890279115,
+      "grad_norm": 1.9914991855621338,
+      "learning_rate": 0.00017224275635459023,
+      "loss": 1.3657,
+      "step": 12635
+    },
+    {
+      "epoch": 0.4866217516843118,
+      "grad_norm": 1.187045693397522,
+      "learning_rate": 0.00017222184397886245,
+      "loss": 1.1048,
+      "step": 12640
+    },
+    {
+      "epoch": 0.48681424446583255,
+      "grad_norm": 1.1656368970870972,
+      "learning_rate": 0.0001722009249989255,
+      "loss": 1.1226,
+      "step": 12645
+    },
+    {
+      "epoch": 0.4870067372473532,
+      "grad_norm": 1.050398349761963,
+      "learning_rate": 0.00017217999941669226,
+      "loss": 1.2018,
+      "step": 12650
+    },
+    {
+      "epoch": 0.48719923002887394,
+      "grad_norm": 1.4217538833618164,
+      "learning_rate": 0.00017215906723407618,
+      "loss": 1.0039,
+      "step": 12655
+    },
+    {
+      "epoch": 0.4873917228103946,
+      "grad_norm": 1.1657346487045288,
+      "learning_rate": 0.00017213812845299145,
+      "loss": 1.3153,
+      "step": 12660
+    },
+    {
+      "epoch": 0.4875842155919153,
+      "grad_norm": 1.0622743368148804,
+      "learning_rate": 0.0001721171830753527,
+      "loss": 1.1889,
+      "step": 12665
+    },
+    {
+      "epoch": 0.487776708373436,
+      "grad_norm": 1.3200461864471436,
+      "learning_rate": 0.00017209623110307534,
+      "loss": 1.2643,
+      "step": 12670
+    },
+    {
+      "epoch": 0.4879692011549567,
+      "grad_norm": 1.3201463222503662,
+      "learning_rate": 0.0001720752725380752,
+      "loss": 1.1021,
+      "step": 12675
+    },
+    {
+      "epoch": 0.4881616939364774,
+      "grad_norm": 1.3110108375549316,
+      "learning_rate": 0.00017205430738226885,
+      "loss": 1.297,
+      "step": 12680
+    },
+    {
+      "epoch": 0.4883541867179981,
+      "grad_norm": 1.9574589729309082,
+      "learning_rate": 0.00017203333563757344,
+      "loss": 1.0589,
+      "step": 12685
+    },
+    {
+      "epoch": 0.48854667949951874,
+      "grad_norm": 2.1387152671813965,
+      "learning_rate": 0.00017201235730590663,
+      "loss": 1.1929,
+      "step": 12690
+    },
+    {
+      "epoch": 0.48873917228103947,
+      "grad_norm": 1.007409691810608,
+      "learning_rate": 0.0001719913723891868,
+      "loss": 1.1213,
+      "step": 12695
+    },
+    {
+      "epoch": 0.48893166506256014,
+      "grad_norm": 0.9808946847915649,
+      "learning_rate": 0.00017197038088933285,
+      "loss": 1.2553,
+      "step": 12700
+    },
+    {
+      "epoch": 0.48912415784408086,
+      "grad_norm": 1.276231050491333,
+      "learning_rate": 0.00017194938280826433,
+      "loss": 1.2783,
+      "step": 12705
+    },
+    {
+      "epoch": 0.48931665062560153,
+      "grad_norm": 2.288778305053711,
+      "learning_rate": 0.0001719283781479014,
+      "loss": 1.2736,
+      "step": 12710
+    },
+    {
+      "epoch": 0.48950914340712226,
+      "grad_norm": 1.3838545083999634,
+      "learning_rate": 0.00017190736691016475,
+      "loss": 1.1101,
+      "step": 12715
+    },
+    {
+      "epoch": 0.48970163618864293,
+      "grad_norm": 1.5680103302001953,
+      "learning_rate": 0.00017188634909697572,
+      "loss": 1.3685,
+      "step": 12720
+    },
+    {
+      "epoch": 0.4898941289701636,
+      "grad_norm": 0.9690240621566772,
+      "learning_rate": 0.00017186532471025626,
+      "loss": 1.1677,
+      "step": 12725
+    },
+    {
+      "epoch": 0.4900866217516843,
+      "grad_norm": 1.617100477218628,
+      "learning_rate": 0.00017184429375192894,
+      "loss": 1.3773,
+      "step": 12730
+    },
+    {
+      "epoch": 0.490279114533205,
+      "grad_norm": 1.751895785331726,
+      "learning_rate": 0.0001718232562239169,
+      "loss": 1.1795,
+      "step": 12735
+    },
+    {
+      "epoch": 0.4904716073147257,
+      "grad_norm": 1.1766438484191895,
+      "learning_rate": 0.00017180221212814386,
+      "loss": 1.0915,
+      "step": 12740
+    },
+    {
+      "epoch": 0.4906641000962464,
+      "grad_norm": 1.384392499923706,
+      "learning_rate": 0.00017178116146653415,
+      "loss": 1.2823,
+      "step": 12745
+    },
+    {
+      "epoch": 0.49085659287776706,
+      "grad_norm": 1.2379616498947144,
+      "learning_rate": 0.00017176010424101274,
+      "loss": 1.1182,
+      "step": 12750
+    },
+    {
+      "epoch": 0.4910490856592878,
+      "grad_norm": 1.0620834827423096,
+      "learning_rate": 0.00017173904045350515,
+      "loss": 1.0414,
+      "step": 12755
+    },
+    {
+      "epoch": 0.49124157844080846,
+      "grad_norm": 1.2448549270629883,
+      "learning_rate": 0.00017171797010593755,
+      "loss": 1.2487,
+      "step": 12760
+    },
+    {
+      "epoch": 0.4914340712223292,
+      "grad_norm": 1.1862956285476685,
+      "learning_rate": 0.00017169689320023666,
+      "loss": 1.2117,
+      "step": 12765
+    },
+    {
+      "epoch": 0.49162656400384985,
+      "grad_norm": 0.8380292057991028,
+      "learning_rate": 0.00017167580973832984,
+      "loss": 1.1396,
+      "step": 12770
+    },
+    {
+      "epoch": 0.4918190567853705,
+      "grad_norm": 1.807305097579956,
+      "learning_rate": 0.00017165471972214506,
+      "loss": 1.2316,
+      "step": 12775
+    },
+    {
+      "epoch": 0.49201154956689125,
+      "grad_norm": 1.5265247821807861,
+      "learning_rate": 0.00017163362315361077,
+      "loss": 1.1191,
+      "step": 12780
+    },
+    {
+      "epoch": 0.4922040423484119,
+      "grad_norm": 1.4569288492202759,
+      "learning_rate": 0.00017161252003465626,
+      "loss": 1.3273,
+      "step": 12785
+    },
+    {
+      "epoch": 0.49239653512993264,
+      "grad_norm": 1.0343568325042725,
+      "learning_rate": 0.00017159141036721112,
+      "loss": 1.3457,
+      "step": 12790
+    },
+    {
+      "epoch": 0.4925890279114533,
+      "grad_norm": 1.4867749214172363,
+      "learning_rate": 0.00017157029415320577,
+      "loss": 1.1783,
+      "step": 12795
+    },
+    {
+      "epoch": 0.49278152069297404,
+      "grad_norm": 1.0775165557861328,
+      "learning_rate": 0.0001715491713945711,
+      "loss": 1.2355,
+      "step": 12800
+    },
+    {
+      "epoch": 0.4929740134744947,
+      "grad_norm": 1.1424553394317627,
+      "learning_rate": 0.0001715280420932387,
+      "loss": 1.1305,
+      "step": 12805
+    },
+    {
+      "epoch": 0.4931665062560154,
+      "grad_norm": 2.403656482696533,
+      "learning_rate": 0.00017150690625114065,
+      "loss": 1.2118,
+      "step": 12810
+    },
+    {
+      "epoch": 0.4933589990375361,
+      "grad_norm": 1.2673543691635132,
+      "learning_rate": 0.00017148576387020976,
+      "loss": 1.163,
+      "step": 12815
+    },
+    {
+      "epoch": 0.4935514918190568,
+      "grad_norm": 1.4545459747314453,
+      "learning_rate": 0.0001714646149523793,
+      "loss": 1.1661,
+      "step": 12820
+    },
+    {
+      "epoch": 0.4937439846005775,
+      "grad_norm": 1.8716140985488892,
+      "learning_rate": 0.0001714434594995832,
+      "loss": 1.0518,
+      "step": 12825
+    },
+    {
+      "epoch": 0.49393647738209817,
+      "grad_norm": 1.4617652893066406,
+      "learning_rate": 0.000171422297513756,
+      "loss": 0.9978,
+      "step": 12830
+    },
+    {
+      "epoch": 0.49412897016361884,
+      "grad_norm": 1.8650813102722168,
+      "learning_rate": 0.00017140112899683284,
+      "loss": 1.0077,
+      "step": 12835
+    },
+    {
+      "epoch": 0.49432146294513957,
+      "grad_norm": 1.4080079793930054,
+      "learning_rate": 0.00017137995395074938,
+      "loss": 1.2906,
+      "step": 12840
+    },
+    {
+      "epoch": 0.49451395572666024,
+      "grad_norm": 1.2144896984100342,
+      "learning_rate": 0.000171358772377442,
+      "loss": 1.1323,
+      "step": 12845
+    },
+    {
+      "epoch": 0.49470644850818096,
+      "grad_norm": 1.3294404745101929,
+      "learning_rate": 0.0001713375842788476,
+      "loss": 1.1987,
+      "step": 12850
+    },
+    {
+      "epoch": 0.49489894128970163,
+      "grad_norm": 1.3663264513015747,
+      "learning_rate": 0.00017131638965690372,
+      "loss": 1.224,
+      "step": 12855
+    },
+    {
+      "epoch": 0.4950914340712223,
+      "grad_norm": 2.325491428375244,
+      "learning_rate": 0.0001712951885135484,
+      "loss": 1.2449,
+      "step": 12860
+    },
+    {
+      "epoch": 0.49528392685274303,
+      "grad_norm": 1.3226628303527832,
+      "learning_rate": 0.00017127398085072039,
+      "loss": 1.2321,
+      "step": 12865
+    },
+    {
+      "epoch": 0.4954764196342637,
+      "grad_norm": 1.309049367904663,
+      "learning_rate": 0.00017125276667035895,
+      "loss": 1.1242,
+      "step": 12870
+    },
+    {
+      "epoch": 0.4956689124157844,
+      "grad_norm": 2.189549446105957,
+      "learning_rate": 0.00017123154597440402,
+      "loss": 1.1479,
+      "step": 12875
+    },
+    {
+      "epoch": 0.4958614051973051,
+      "grad_norm": 1.258787989616394,
+      "learning_rate": 0.00017121031876479606,
+      "loss": 1.3375,
+      "step": 12880
+    },
+    {
+      "epoch": 0.4960538979788258,
+      "grad_norm": 1.4555957317352295,
+      "learning_rate": 0.00017118908504347623,
+      "loss": 1.29,
+      "step": 12885
+    },
+    {
+      "epoch": 0.4962463907603465,
+      "grad_norm": 1.4220309257507324,
+      "learning_rate": 0.00017116784481238613,
+      "loss": 1.2438,
+      "step": 12890
+    },
+    {
+      "epoch": 0.49643888354186716,
+      "grad_norm": 1.1141269207000732,
+      "learning_rate": 0.00017114659807346803,
+      "loss": 1.1286,
+      "step": 12895
+    },
+    {
+      "epoch": 0.4966313763233879,
+      "grad_norm": 2.7541897296905518,
+      "learning_rate": 0.00017112534482866486,
+      "loss": 1.2506,
+      "step": 12900
+    },
+    {
+      "epoch": 0.49682386910490856,
+      "grad_norm": 1.8478270769119263,
+      "learning_rate": 0.00017110408507992007,
+      "loss": 1.2337,
+      "step": 12905
+    },
+    {
+      "epoch": 0.4970163618864293,
+      "grad_norm": 1.3013496398925781,
+      "learning_rate": 0.00017108281882917775,
+      "loss": 1.109,
+      "step": 12910
+    },
+    {
+      "epoch": 0.49720885466794995,
+      "grad_norm": 1.6363266706466675,
+      "learning_rate": 0.00017106154607838249,
+      "loss": 1.2546,
+      "step": 12915
+    },
+    {
+      "epoch": 0.4974013474494706,
+      "grad_norm": 1.4996516704559326,
+      "learning_rate": 0.0001710402668294796,
+      "loss": 1.3066,
+      "step": 12920
+    },
+    {
+      "epoch": 0.49759384023099135,
+      "grad_norm": 1.2411632537841797,
+      "learning_rate": 0.0001710189810844149,
+      "loss": 1.2678,
+      "step": 12925
+    },
+    {
+      "epoch": 0.497786333012512,
+      "grad_norm": 1.197771430015564,
+      "learning_rate": 0.00017099768884513484,
+      "loss": 1.1069,
+      "step": 12930
+    },
+    {
+      "epoch": 0.49797882579403274,
+      "grad_norm": 1.1568015813827515,
+      "learning_rate": 0.00017097639011358644,
+      "loss": 1.1863,
+      "step": 12935
+    },
+    {
+      "epoch": 0.4981713185755534,
+      "grad_norm": 1.8848886489868164,
+      "learning_rate": 0.00017095508489171736,
+      "loss": 1.3294,
+      "step": 12940
+    },
+    {
+      "epoch": 0.4983638113570741,
+      "grad_norm": 1.4993650913238525,
+      "learning_rate": 0.00017093377318147578,
+      "loss": 1.1768,
+      "step": 12945
+    },
+    {
+      "epoch": 0.4985563041385948,
+      "grad_norm": 1.1212975978851318,
+      "learning_rate": 0.00017091245498481055,
+      "loss": 1.2018,
+      "step": 12950
+    },
+    {
+      "epoch": 0.4987487969201155,
+      "grad_norm": 1.086147665977478,
+      "learning_rate": 0.00017089113030367107,
+      "loss": 1.3022,
+      "step": 12955
+    },
+    {
+      "epoch": 0.4989412897016362,
+      "grad_norm": 2.185974359512329,
+      "learning_rate": 0.00017086979914000732,
+      "loss": 1.239,
+      "step": 12960
+    },
+    {
+      "epoch": 0.4991337824831569,
+      "grad_norm": 1.0493237972259521,
+      "learning_rate": 0.00017084846149576993,
+      "loss": 1.1614,
+      "step": 12965
+    },
+    {
+      "epoch": 0.4993262752646776,
+      "grad_norm": 2.034449815750122,
+      "learning_rate": 0.00017082711737291005,
+      "loss": 1.2636,
+      "step": 12970
+    },
+    {
+      "epoch": 0.49951876804619827,
+      "grad_norm": 1.6736334562301636,
+      "learning_rate": 0.00017080576677337944,
+      "loss": 1.1721,
+      "step": 12975
+    },
+    {
+      "epoch": 0.49971126082771894,
+      "grad_norm": 1.8189449310302734,
+      "learning_rate": 0.00017078440969913055,
+      "loss": 1.3739,
+      "step": 12980
+    },
+    {
+      "epoch": 0.49990375360923966,
+      "grad_norm": 1.500243067741394,
+      "learning_rate": 0.00017076304615211627,
+      "loss": 1.0446,
+      "step": 12985
+    },
+    {
+      "epoch": 0.5000962463907603,
+      "grad_norm": 1.2671639919281006,
+      "learning_rate": 0.0001707416761342902,
+      "loss": 1.1841,
+      "step": 12990
+    },
+    {
+      "epoch": 0.500288739172281,
+      "grad_norm": 1.3602843284606934,
+      "learning_rate": 0.00017072029964760644,
+      "loss": 1.3311,
+      "step": 12995
+    },
+    {
+      "epoch": 0.5004812319538018,
+      "grad_norm": 1.106224775314331,
+      "learning_rate": 0.00017069891669401982,
+      "loss": 1.1083,
+      "step": 13000
+    },
+    {
+      "epoch": 0.5006737247353225,
+      "grad_norm": 1.3517072200775146,
+      "learning_rate": 0.00017067752727548555,
+      "loss": 1.2598,
+      "step": 13005
+    },
+    {
+      "epoch": 0.5008662175168431,
+      "grad_norm": 1.1175580024719238,
+      "learning_rate": 0.00017065613139395958,
+      "loss": 1.0876,
+      "step": 13010
+    },
+    {
+      "epoch": 0.5010587102983638,
+      "grad_norm": 1.8314218521118164,
+      "learning_rate": 0.00017063472905139854,
+      "loss": 1.3194,
+      "step": 13015
+    },
+    {
+      "epoch": 0.5012512030798845,
+      "grad_norm": 0.7892528176307678,
+      "learning_rate": 0.00017061332024975937,
+      "loss": 1.0172,
+      "step": 13020
+    },
+    {
+      "epoch": 0.5014436958614052,
+      "grad_norm": 0.8713880777359009,
+      "learning_rate": 0.00017059190499099986,
+      "loss": 1.0775,
+      "step": 13025
+    },
+    {
+      "epoch": 0.5016361886429259,
+      "grad_norm": 2.4726779460906982,
+      "learning_rate": 0.00017057048327707826,
+      "loss": 1.3351,
+      "step": 13030
+    },
+    {
+      "epoch": 0.5018286814244466,
+      "grad_norm": 1.1800824403762817,
+      "learning_rate": 0.00017054905510995342,
+      "loss": 1.3102,
+      "step": 13035
+    },
+    {
+      "epoch": 0.5020211742059673,
+      "grad_norm": 1.583617091178894,
+      "learning_rate": 0.0001705276204915849,
+      "loss": 1.3633,
+      "step": 13040
+    },
+    {
+      "epoch": 0.5022136669874879,
+      "grad_norm": 2.0497727394104004,
+      "learning_rate": 0.00017050617942393264,
+      "loss": 1.2055,
+      "step": 13045
+    },
+    {
+      "epoch": 0.5024061597690087,
+      "grad_norm": 1.775793433189392,
+      "learning_rate": 0.00017048473190895734,
+      "loss": 1.1747,
+      "step": 13050
+    },
+    {
+      "epoch": 0.5025986525505294,
+      "grad_norm": 1.8494744300842285,
+      "learning_rate": 0.00017046327794862024,
+      "loss": 1.2159,
+      "step": 13055
+    },
+    {
+      "epoch": 0.50279114533205,
+      "grad_norm": 1.6188912391662598,
+      "learning_rate": 0.00017044181754488315,
+      "loss": 1.156,
+      "step": 13060
+    },
+    {
+      "epoch": 0.5029836381135707,
+      "grad_norm": 1.575838565826416,
+      "learning_rate": 0.00017042035069970846,
+      "loss": 1.2103,
+      "step": 13065
+    },
+    {
+      "epoch": 0.5031761308950914,
+      "grad_norm": 1.7594157457351685,
+      "learning_rate": 0.0001703988774150592,
+      "loss": 1.4271,
+      "step": 13070
+    },
+    {
+      "epoch": 0.5033686236766122,
+      "grad_norm": 2.066418170928955,
+      "learning_rate": 0.00017037739769289894,
+      "loss": 1.215,
+      "step": 13075
+    },
+    {
+      "epoch": 0.5035611164581328,
+      "grad_norm": 1.8343390226364136,
+      "learning_rate": 0.0001703559115351919,
+      "loss": 1.3318,
+      "step": 13080
+    },
+    {
+      "epoch": 0.5037536092396535,
+      "grad_norm": 1.177186369895935,
+      "learning_rate": 0.00017033441894390278,
+      "loss": 1.2845,
+      "step": 13085
+    },
+    {
+      "epoch": 0.5039461020211742,
+      "grad_norm": 1.7650407552719116,
+      "learning_rate": 0.000170312919920997,
+      "loss": 1.2133,
+      "step": 13090
+    },
+    {
+      "epoch": 0.5041385948026949,
+      "grad_norm": 1.0483806133270264,
+      "learning_rate": 0.00017029141446844043,
+      "loss": 1.2309,
+      "step": 13095
+    },
+    {
+      "epoch": 0.5043310875842156,
+      "grad_norm": 1.1729573011398315,
+      "learning_rate": 0.00017026990258819968,
+      "loss": 1.2975,
+      "step": 13100
+    },
+    {
+      "epoch": 0.5045235803657363,
+      "grad_norm": 1.8557977676391602,
+      "learning_rate": 0.00017024838428224184,
+      "loss": 1.1332,
+      "step": 13105
+    },
+    {
+      "epoch": 0.504716073147257,
+      "grad_norm": 2.2607064247131348,
+      "learning_rate": 0.00017022685955253458,
+      "loss": 1.186,
+      "step": 13110
+    },
+    {
+      "epoch": 0.5049085659287776,
+      "grad_norm": 1.0992141962051392,
+      "learning_rate": 0.00017020532840104625,
+      "loss": 1.0708,
+      "step": 13115
+    },
+    {
+      "epoch": 0.5051010587102983,
+      "grad_norm": 0.9550696611404419,
+      "learning_rate": 0.0001701837908297457,
+      "loss": 1.2069,
+      "step": 13120
+    },
+    {
+      "epoch": 0.5052935514918191,
+      "grad_norm": 1.9301509857177734,
+      "learning_rate": 0.00017016224684060242,
+      "loss": 1.1152,
+      "step": 13125
+    },
+    {
+      "epoch": 0.5054860442733398,
+      "grad_norm": 1.2657769918441772,
+      "learning_rate": 0.0001701406964355864,
+      "loss": 1.1028,
+      "step": 13130
+    },
+    {
+      "epoch": 0.5056785370548604,
+      "grad_norm": 1.549902081489563,
+      "learning_rate": 0.00017011913961666837,
+      "loss": 1.1677,
+      "step": 13135
+    },
+    {
+      "epoch": 0.5058710298363811,
+      "grad_norm": 1.5015727281570435,
+      "learning_rate": 0.00017009757638581952,
+      "loss": 1.1799,
+      "step": 13140
+    },
+    {
+      "epoch": 0.5060635226179019,
+      "grad_norm": 1.206145167350769,
+      "learning_rate": 0.00017007600674501166,
+      "loss": 1.131,
+      "step": 13145
+    },
+    {
+      "epoch": 0.5062560153994226,
+      "grad_norm": 1.1034317016601562,
+      "learning_rate": 0.00017005443069621716,
+      "loss": 1.0504,
+      "step": 13150
+    },
+    {
+      "epoch": 0.5064485081809432,
+      "grad_norm": 1.0830001831054688,
+      "learning_rate": 0.00017003284824140908,
+      "loss": 1.1653,
+      "step": 13155
+    },
+    {
+      "epoch": 0.5066410009624639,
+      "grad_norm": 1.8119686841964722,
+      "learning_rate": 0.00017001125938256094,
+      "loss": 1.1646,
+      "step": 13160
+    },
+    {
+      "epoch": 0.5068334937439846,
+      "grad_norm": 1.1078890562057495,
+      "learning_rate": 0.00016998966412164692,
+      "loss": 1.3281,
+      "step": 13165
+    },
+    {
+      "epoch": 0.5070259865255053,
+      "grad_norm": 2.1965198516845703,
+      "learning_rate": 0.00016996806246064174,
+      "loss": 1.1042,
+      "step": 13170
+    },
+    {
+      "epoch": 0.507218479307026,
+      "grad_norm": 1.3997282981872559,
+      "learning_rate": 0.00016994645440152075,
+      "loss": 1.1662,
+      "step": 13175
+    },
+    {
+      "epoch": 0.5074109720885467,
+      "grad_norm": 1.2493458986282349,
+      "learning_rate": 0.00016992483994625985,
+      "loss": 1.1594,
+      "step": 13180
+    },
+    {
+      "epoch": 0.5076034648700674,
+      "grad_norm": 0.8307852745056152,
+      "learning_rate": 0.00016990321909683557,
+      "loss": 1.1701,
+      "step": 13185
+    },
+    {
+      "epoch": 0.507795957651588,
+      "grad_norm": 1.1798492670059204,
+      "learning_rate": 0.00016988159185522497,
+      "loss": 1.2217,
+      "step": 13190
+    },
+    {
+      "epoch": 0.5079884504331088,
+      "grad_norm": 2.131786823272705,
+      "learning_rate": 0.00016985995822340567,
+      "loss": 1.277,
+      "step": 13195
+    },
+    {
+      "epoch": 0.5081809432146295,
+      "grad_norm": 1.0258443355560303,
+      "learning_rate": 0.00016983831820335603,
+      "loss": 1.0954,
+      "step": 13200
+    },
+    {
+      "epoch": 0.5083734359961501,
+      "grad_norm": 2.4005777835845947,
+      "learning_rate": 0.0001698166717970548,
+      "loss": 1.3484,
+      "step": 13205
+    },
+    {
+      "epoch": 0.5085659287776708,
+      "grad_norm": 1.3329745531082153,
+      "learning_rate": 0.00016979501900648143,
+      "loss": 1.2548,
+      "step": 13210
+    },
+    {
+      "epoch": 0.5087584215591915,
+      "grad_norm": 1.2792582511901855,
+      "learning_rate": 0.00016977335983361594,
+      "loss": 1.1056,
+      "step": 13215
+    },
+    {
+      "epoch": 0.5089509143407123,
+      "grad_norm": 1.1146180629730225,
+      "learning_rate": 0.00016975169428043888,
+      "loss": 1.1728,
+      "step": 13220
+    },
+    {
+      "epoch": 0.5091434071222329,
+      "grad_norm": 1.155003309249878,
+      "learning_rate": 0.0001697300223489314,
+      "loss": 1.2469,
+      "step": 13225
+    },
+    {
+      "epoch": 0.5093358999037536,
+      "grad_norm": 1.8456053733825684,
+      "learning_rate": 0.00016970834404107535,
+      "loss": 1.1515,
+      "step": 13230
+    },
+    {
+      "epoch": 0.5095283926852743,
+      "grad_norm": 1.3863856792449951,
+      "learning_rate": 0.000169686659358853,
+      "loss": 1.2561,
+      "step": 13235
+    },
+    {
+      "epoch": 0.509720885466795,
+      "grad_norm": 2.099985361099243,
+      "learning_rate": 0.00016966496830424728,
+      "loss": 1.2639,
+      "step": 13240
+    },
+    {
+      "epoch": 0.5099133782483157,
+      "grad_norm": 1.4132083654403687,
+      "learning_rate": 0.0001696432708792417,
+      "loss": 0.9859,
+      "step": 13245
+    },
+    {
+      "epoch": 0.5101058710298364,
+      "grad_norm": 1.0421473979949951,
+      "learning_rate": 0.00016962156708582037,
+      "loss": 1.1239,
+      "step": 13250
+    },
+    {
+      "epoch": 0.5102983638113571,
+      "grad_norm": 1.4971591234207153,
+      "learning_rate": 0.0001695998569259679,
+      "loss": 1.183,
+      "step": 13255
+    },
+    {
+      "epoch": 0.5104908565928777,
+      "grad_norm": 1.7850632667541504,
+      "learning_rate": 0.00016957814040166955,
+      "loss": 1.2342,
+      "step": 13260
+    },
+    {
+      "epoch": 0.5106833493743984,
+      "grad_norm": 2.817624092102051,
+      "learning_rate": 0.0001695564175149112,
+      "loss": 1.1795,
+      "step": 13265
+    },
+    {
+      "epoch": 0.5108758421559192,
+      "grad_norm": 1.4107112884521484,
+      "learning_rate": 0.00016953468826767925,
+      "loss": 1.2791,
+      "step": 13270
+    },
+    {
+      "epoch": 0.5110683349374399,
+      "grad_norm": 1.4817914962768555,
+      "learning_rate": 0.00016951295266196063,
+      "loss": 1.0816,
+      "step": 13275
+    },
+    {
+      "epoch": 0.5112608277189605,
+      "grad_norm": 1.054870367050171,
+      "learning_rate": 0.00016949121069974302,
+      "loss": 1.2114,
+      "step": 13280
+    },
+    {
+      "epoch": 0.5114533205004812,
+      "grad_norm": 1.2629690170288086,
+      "learning_rate": 0.00016946946238301453,
+      "loss": 1.1014,
+      "step": 13285
+    },
+    {
+      "epoch": 0.5116458132820019,
+      "grad_norm": 0.9189853668212891,
+      "learning_rate": 0.00016944770771376387,
+      "loss": 1.319,
+      "step": 13290
+    },
+    {
+      "epoch": 0.5118383060635227,
+      "grad_norm": 1.4326847791671753,
+      "learning_rate": 0.0001694259466939804,
+      "loss": 1.299,
+      "step": 13295
+    },
+    {
+      "epoch": 0.5120307988450433,
+      "grad_norm": 1.1833186149597168,
+      "learning_rate": 0.00016940417932565402,
+      "loss": 1.2863,
+      "step": 13300
+    },
+    {
+      "epoch": 0.512223291626564,
+      "grad_norm": 1.1329289674758911,
+      "learning_rate": 0.0001693824056107752,
+      "loss": 1.2228,
+      "step": 13305
+    },
+    {
+      "epoch": 0.5124157844080847,
+      "grad_norm": 1.2103817462921143,
+      "learning_rate": 0.000169360625551335,
+      "loss": 1.042,
+      "step": 13310
+    },
+    {
+      "epoch": 0.5126082771896054,
+      "grad_norm": 1.2664172649383545,
+      "learning_rate": 0.00016933883914932506,
+      "loss": 1.299,
+      "step": 13315
+    },
+    {
+      "epoch": 0.5128007699711261,
+      "grad_norm": 1.8509985208511353,
+      "learning_rate": 0.0001693170464067376,
+      "loss": 1.135,
+      "step": 13320
+    },
+    {
+      "epoch": 0.5129932627526468,
+      "grad_norm": 2.1004250049591064,
+      "learning_rate": 0.00016929524732556546,
+      "loss": 1.1321,
+      "step": 13325
+    },
+    {
+      "epoch": 0.5131857555341675,
+      "grad_norm": 1.1648815870285034,
+      "learning_rate": 0.00016927344190780197,
+      "loss": 1.094,
+      "step": 13330
+    },
+    {
+      "epoch": 0.5133782483156881,
+      "grad_norm": 0.9492617249488831,
+      "learning_rate": 0.0001692516301554411,
+      "loss": 1.2055,
+      "step": 13335
+    },
+    {
+      "epoch": 0.5135707410972089,
+      "grad_norm": 1.7911789417266846,
+      "learning_rate": 0.00016922981207047742,
+      "loss": 1.1726,
+      "step": 13340
+    },
+    {
+      "epoch": 0.5137632338787296,
+      "grad_norm": 1.2055487632751465,
+      "learning_rate": 0.00016920798765490601,
+      "loss": 1.158,
+      "step": 13345
+    },
+    {
+      "epoch": 0.5139557266602502,
+      "grad_norm": 1.1120411157608032,
+      "learning_rate": 0.0001691861569107226,
+      "loss": 1.2508,
+      "step": 13350
+    },
+    {
+      "epoch": 0.5141482194417709,
+      "grad_norm": 1.1816275119781494,
+      "learning_rate": 0.0001691643198399235,
+      "loss": 1.0294,
+      "step": 13355
+    },
+    {
+      "epoch": 0.5143407122232916,
+      "grad_norm": 1.1714962720870972,
+      "learning_rate": 0.00016914247644450546,
+      "loss": 1.2843,
+      "step": 13360
+    },
+    {
+      "epoch": 0.5145332050048124,
+      "grad_norm": 2.0812292098999023,
+      "learning_rate": 0.000169120626726466,
+      "loss": 1.284,
+      "step": 13365
+    },
+    {
+      "epoch": 0.514725697786333,
+      "grad_norm": 1.7628620862960815,
+      "learning_rate": 0.00016909877068780314,
+      "loss": 1.1104,
+      "step": 13370
+    },
+    {
+      "epoch": 0.5149181905678537,
+      "grad_norm": 1.7429643869400024,
+      "learning_rate": 0.0001690769083305154,
+      "loss": 1.1387,
+      "step": 13375
+    },
+    {
+      "epoch": 0.5151106833493744,
+      "grad_norm": 2.087916612625122,
+      "learning_rate": 0.00016905503965660196,
+      "loss": 1.2737,
+      "step": 13380
+    },
+    {
+      "epoch": 0.515303176130895,
+      "grad_norm": 1.5689221620559692,
+      "learning_rate": 0.00016903316466806265,
+      "loss": 1.145,
+      "step": 13385
+    },
+    {
+      "epoch": 0.5154956689124158,
+      "grad_norm": 1.0740375518798828,
+      "learning_rate": 0.0001690112833668977,
+      "loss": 0.9748,
+      "step": 13390
+    },
+    {
+      "epoch": 0.5156881616939365,
+      "grad_norm": 1.4595876932144165,
+      "learning_rate": 0.00016898939575510805,
+      "loss": 1.1378,
+      "step": 13395
+    },
+    {
+      "epoch": 0.5158806544754572,
+      "grad_norm": 1.9210182428359985,
+      "learning_rate": 0.00016896750183469517,
+      "loss": 1.2581,
+      "step": 13400
+    },
+    {
+      "epoch": 0.5160731472569778,
+      "grad_norm": 1.0922927856445312,
+      "learning_rate": 0.00016894560160766117,
+      "loss": 1.1601,
+      "step": 13405
+    },
+    {
+      "epoch": 0.5162656400384985,
+      "grad_norm": 2.037611246109009,
+      "learning_rate": 0.00016892369507600855,
+      "loss": 1.2394,
+      "step": 13410
+    },
+    {
+      "epoch": 0.5164581328200193,
+      "grad_norm": 1.1577821969985962,
+      "learning_rate": 0.0001689017822417406,
+      "loss": 1.2428,
+      "step": 13415
+    },
+    {
+      "epoch": 0.51665062560154,
+      "grad_norm": 1.1762430667877197,
+      "learning_rate": 0.00016887986310686114,
+      "loss": 1.1648,
+      "step": 13420
+    },
+    {
+      "epoch": 0.5168431183830606,
+      "grad_norm": 1.8631316423416138,
+      "learning_rate": 0.00016885793767337445,
+      "loss": 1.2288,
+      "step": 13425
+    },
+    {
+      "epoch": 0.5170356111645813,
+      "grad_norm": 1.191747784614563,
+      "learning_rate": 0.0001688360059432855,
+      "loss": 1.1287,
+      "step": 13430
+    },
+    {
+      "epoch": 0.517228103946102,
+      "grad_norm": 1.092367172241211,
+      "learning_rate": 0.00016881406791859985,
+      "loss": 1.2073,
+      "step": 13435
+    },
+    {
+      "epoch": 0.5174205967276228,
+      "grad_norm": 0.9805938601493835,
+      "learning_rate": 0.00016879212360132345,
+      "loss": 1.3199,
+      "step": 13440
+    },
+    {
+      "epoch": 0.5176130895091434,
+      "grad_norm": 1.0042074918746948,
+      "learning_rate": 0.00016877017299346314,
+      "loss": 1.1389,
+      "step": 13445
+    },
+    {
+      "epoch": 0.5178055822906641,
+      "grad_norm": 1.3087821006774902,
+      "learning_rate": 0.00016874821609702605,
+      "loss": 1.2112,
+      "step": 13450
+    },
+    {
+      "epoch": 0.5179980750721848,
+      "grad_norm": 1.4208637475967407,
+      "learning_rate": 0.00016872625291401998,
+      "loss": 1.119,
+      "step": 13455
+    },
+    {
+      "epoch": 0.5181905678537055,
+      "grad_norm": 0.9211226105690002,
+      "learning_rate": 0.0001687042834464534,
+      "loss": 1.1458,
+      "step": 13460
+    },
+    {
+      "epoch": 0.5183830606352262,
+      "grad_norm": 1.1774996519088745,
+      "learning_rate": 0.00016868230769633518,
+      "loss": 1.195,
+      "step": 13465
+    },
+    {
+      "epoch": 0.5185755534167469,
+      "grad_norm": 1.395883321762085,
+      "learning_rate": 0.0001686603256656749,
+      "loss": 1.1633,
+      "step": 13470
+    },
+    {
+      "epoch": 0.5187680461982676,
+      "grad_norm": 2.2554938793182373,
+      "learning_rate": 0.00016863833735648268,
+      "loss": 1.1345,
+      "step": 13475
+    },
+    {
+      "epoch": 0.5189605389797882,
+      "grad_norm": 1.2396293878555298,
+      "learning_rate": 0.00016861634277076922,
+      "loss": 1.1109,
+      "step": 13480
+    },
+    {
+      "epoch": 0.519153031761309,
+      "grad_norm": 1.2292909622192383,
+      "learning_rate": 0.00016859434191054574,
+      "loss": 1.1029,
+      "step": 13485
+    },
+    {
+      "epoch": 0.5193455245428297,
+      "grad_norm": 1.145571231842041,
+      "learning_rate": 0.00016857233477782409,
+      "loss": 1.1734,
+      "step": 13490
+    },
+    {
+      "epoch": 0.5195380173243503,
+      "grad_norm": 1.7307795286178589,
+      "learning_rate": 0.00016855032137461667,
+      "loss": 1.1476,
+      "step": 13495
+    },
+    {
+      "epoch": 0.519730510105871,
+      "grad_norm": 1.611140489578247,
+      "learning_rate": 0.0001685283017029365,
+      "loss": 1.1304,
+      "step": 13500
+    },
+    {
+      "epoch": 0.5199230028873917,
+      "grad_norm": 1.3966014385223389,
+      "learning_rate": 0.00016850627576479705,
+      "loss": 1.1231,
+      "step": 13505
+    },
+    {
+      "epoch": 0.5201154956689125,
+      "grad_norm": 1.505765676498413,
+      "learning_rate": 0.0001684842435622125,
+      "loss": 1.0741,
+      "step": 13510
+    },
+    {
+      "epoch": 0.5203079884504331,
+      "grad_norm": 1.791595220565796,
+      "learning_rate": 0.00016846220509719755,
+      "loss": 1.1928,
+      "step": 13515
+    },
+    {
+      "epoch": 0.5205004812319538,
+      "grad_norm": 1.2992479801177979,
+      "learning_rate": 0.00016844016037176744,
+      "loss": 1.1523,
+      "step": 13520
+    },
+    {
+      "epoch": 0.5206929740134745,
+      "grad_norm": 1.8747221231460571,
+      "learning_rate": 0.00016841810938793807,
+      "loss": 1.0704,
+      "step": 13525
+    },
+    {
+      "epoch": 0.5208854667949951,
+      "grad_norm": 1.3441274166107178,
+      "learning_rate": 0.00016839605214772583,
+      "loss": 1.1979,
+      "step": 13530
+    },
+    {
+      "epoch": 0.5210779595765159,
+      "grad_norm": 0.8640159964561462,
+      "learning_rate": 0.0001683739886531477,
+      "loss": 1.1577,
+      "step": 13535
+    },
+    {
+      "epoch": 0.5212704523580366,
+      "grad_norm": 1.7198442220687866,
+      "learning_rate": 0.00016835191890622123,
+      "loss": 1.2623,
+      "step": 13540
+    },
+    {
+      "epoch": 0.5214629451395573,
+      "grad_norm": 1.2651041746139526,
+      "learning_rate": 0.0001683298429089646,
+      "loss": 1.2428,
+      "step": 13545
+    },
+    {
+      "epoch": 0.5216554379210779,
+      "grad_norm": 1.9191710948944092,
+      "learning_rate": 0.00016830776066339642,
+      "loss": 1.2872,
+      "step": 13550
+    },
+    {
+      "epoch": 0.5218479307025986,
+      "grad_norm": 1.8098481893539429,
+      "learning_rate": 0.00016828567217153605,
+      "loss": 1.2838,
+      "step": 13555
+    },
+    {
+      "epoch": 0.5220404234841194,
+      "grad_norm": 1.732160210609436,
+      "learning_rate": 0.00016826357743540332,
+      "loss": 1.0766,
+      "step": 13560
+    },
+    {
+      "epoch": 0.5222329162656401,
+      "grad_norm": 1.4580518007278442,
+      "learning_rate": 0.00016824147645701863,
+      "loss": 1.2825,
+      "step": 13565
+    },
+    {
+      "epoch": 0.5224254090471607,
+      "grad_norm": 1.5836480855941772,
+      "learning_rate": 0.000168219369238403,
+      "loss": 1.1772,
+      "step": 13570
+    },
+    {
+      "epoch": 0.5226179018286814,
+      "grad_norm": 1.5529143810272217,
+      "learning_rate": 0.00016819725578157794,
+      "loss": 1.2795,
+      "step": 13575
+    },
+    {
+      "epoch": 0.5228103946102021,
+      "grad_norm": 1.1405484676361084,
+      "learning_rate": 0.0001681751360885656,
+      "loss": 1.2133,
+      "step": 13580
+    },
+    {
+      "epoch": 0.5230028873917228,
+      "grad_norm": 1.0912057161331177,
+      "learning_rate": 0.00016815301016138873,
+      "loss": 1.0493,
+      "step": 13585
+    },
+    {
+      "epoch": 0.5231953801732435,
+      "grad_norm": 0.9384201169013977,
+      "learning_rate": 0.0001681308780020705,
+      "loss": 1.1638,
+      "step": 13590
+    },
+    {
+      "epoch": 0.5233878729547642,
+      "grad_norm": 1.3467286825180054,
+      "learning_rate": 0.0001681087396126348,
+      "loss": 1.1927,
+      "step": 13595
+    },
+    {
+      "epoch": 0.5235803657362849,
+      "grad_norm": 0.9008259773254395,
+      "learning_rate": 0.00016808659499510607,
+      "loss": 1.2158,
+      "step": 13600
+    },
+    {
+      "epoch": 0.5237728585178055,
+      "grad_norm": 1.1013727188110352,
+      "learning_rate": 0.00016806444415150927,
+      "loss": 1.1575,
+      "step": 13605
+    },
+    {
+      "epoch": 0.5239653512993263,
+      "grad_norm": 1.160654902458191,
+      "learning_rate": 0.00016804228708386992,
+      "loss": 1.1662,
+      "step": 13610
+    },
+    {
+      "epoch": 0.524157844080847,
+      "grad_norm": 1.5752032995224,
+      "learning_rate": 0.00016802012379421414,
+      "loss": 1.1596,
+      "step": 13615
+    },
+    {
+      "epoch": 0.5243503368623676,
+      "grad_norm": 1.1819881200790405,
+      "learning_rate": 0.00016799795428456865,
+      "loss": 1.1686,
+      "step": 13620
+    },
+    {
+      "epoch": 0.5245428296438883,
+      "grad_norm": 0.9841921329498291,
+      "learning_rate": 0.00016797577855696069,
+      "loss": 1.0872,
+      "step": 13625
+    },
+    {
+      "epoch": 0.5247353224254091,
+      "grad_norm": 1.2292228937149048,
+      "learning_rate": 0.00016795359661341808,
+      "loss": 1.3943,
+      "step": 13630
+    },
+    {
+      "epoch": 0.5249278152069298,
+      "grad_norm": 1.2674068212509155,
+      "learning_rate": 0.0001679314084559692,
+      "loss": 1.232,
+      "step": 13635
+    },
+    {
+      "epoch": 0.5251203079884504,
+      "grad_norm": 1.1942312717437744,
+      "learning_rate": 0.00016790921408664302,
+      "loss": 1.2223,
+      "step": 13640
+    },
+    {
+      "epoch": 0.5253128007699711,
+      "grad_norm": 1.5753337144851685,
+      "learning_rate": 0.00016788701350746907,
+      "loss": 1.2936,
+      "step": 13645
+    },
+    {
+      "epoch": 0.5255052935514918,
+      "grad_norm": 1.1031461954116821,
+      "learning_rate": 0.00016786480672047744,
+      "loss": 1.2651,
+      "step": 13650
+    },
+    {
+      "epoch": 0.5256977863330126,
+      "grad_norm": 3.8325674533843994,
+      "learning_rate": 0.00016784259372769884,
+      "loss": 1.1693,
+      "step": 13655
+    },
+    {
+      "epoch": 0.5258902791145332,
+      "grad_norm": 1.6535909175872803,
+      "learning_rate": 0.0001678203745311644,
+      "loss": 1.1606,
+      "step": 13660
+    },
+    {
+      "epoch": 0.5260827718960539,
+      "grad_norm": 1.6406097412109375,
+      "learning_rate": 0.000167798149132906,
+      "loss": 1.3232,
+      "step": 13665
+    },
+    {
+      "epoch": 0.5262752646775746,
+      "grad_norm": 1.6994904279708862,
+      "learning_rate": 0.000167775917534956,
+      "loss": 1.0998,
+      "step": 13670
+    },
+    {
+      "epoch": 0.5264677574590952,
+      "grad_norm": 1.6446374654769897,
+      "learning_rate": 0.0001677536797393473,
+      "loss": 1.2997,
+      "step": 13675
+    },
+    {
+      "epoch": 0.526660250240616,
+      "grad_norm": 1.6050851345062256,
+      "learning_rate": 0.0001677314357481134,
+      "loss": 1.19,
+      "step": 13680
+    },
+    {
+      "epoch": 0.5268527430221367,
+      "grad_norm": 1.473940134048462,
+      "learning_rate": 0.00016770918556328844,
+      "loss": 1.2007,
+      "step": 13685
+    },
+    {
+      "epoch": 0.5270452358036574,
+      "grad_norm": 1.1209567785263062,
+      "learning_rate": 0.00016768692918690695,
+      "loss": 1.2956,
+      "step": 13690
+    },
+    {
+      "epoch": 0.527237728585178,
+      "grad_norm": 1.4143558740615845,
+      "learning_rate": 0.00016766466662100415,
+      "loss": 1.1734,
+      "step": 13695
+    },
+    {
+      "epoch": 0.5274302213666987,
+      "grad_norm": 1.138107180595398,
+      "learning_rate": 0.00016764239786761585,
+      "loss": 1.1318,
+      "step": 13700
+    },
+    {
+      "epoch": 0.5276227141482195,
+      "grad_norm": 1.5194774866104126,
+      "learning_rate": 0.00016762012292877835,
+      "loss": 1.1525,
+      "step": 13705
+    },
+    {
+      "epoch": 0.5278152069297402,
+      "grad_norm": 1.136946439743042,
+      "learning_rate": 0.00016759784180652858,
+      "loss": 1.1289,
+      "step": 13710
+    },
+    {
+      "epoch": 0.5280076997112608,
+      "grad_norm": 1.5263949632644653,
+      "learning_rate": 0.00016757555450290396,
+      "loss": 1.2811,
+      "step": 13715
+    },
+    {
+      "epoch": 0.5282001924927815,
+      "grad_norm": 2.306833505630493,
+      "learning_rate": 0.00016755326101994248,
+      "loss": 1.0326,
+      "step": 13720
+    },
+    {
+      "epoch": 0.5283926852743022,
+      "grad_norm": 1.4330452680587769,
+      "learning_rate": 0.0001675309613596828,
+      "loss": 1.0126,
+      "step": 13725
+    },
+    {
+      "epoch": 0.528585178055823,
+      "grad_norm": 0.8746087551116943,
+      "learning_rate": 0.00016750865552416408,
+      "loss": 1.224,
+      "step": 13730
+    },
+    {
+      "epoch": 0.5287776708373436,
+      "grad_norm": 2.576612949371338,
+      "learning_rate": 0.000167486343515426,
+      "loss": 1.178,
+      "step": 13735
+    },
+    {
+      "epoch": 0.5289701636188643,
+      "grad_norm": 1.3074976205825806,
+      "learning_rate": 0.00016746402533550887,
+      "loss": 1.2453,
+      "step": 13740
+    },
+    {
+      "epoch": 0.529162656400385,
+      "grad_norm": 1.0941317081451416,
+      "learning_rate": 0.00016744170098645353,
+      "loss": 0.9341,
+      "step": 13745
+    },
+    {
+      "epoch": 0.5293551491819056,
+      "grad_norm": 1.6738418340682983,
+      "learning_rate": 0.00016741937047030139,
+      "loss": 1.1423,
+      "step": 13750
+    },
+    {
+      "epoch": 0.5295476419634264,
+      "grad_norm": 1.9735844135284424,
+      "learning_rate": 0.00016739703378909444,
+      "loss": 1.0691,
+      "step": 13755
+    },
+    {
+      "epoch": 0.5297401347449471,
+      "grad_norm": 1.0063233375549316,
+      "learning_rate": 0.00016737469094487518,
+      "loss": 1.0096,
+      "step": 13760
+    },
+    {
+      "epoch": 0.5299326275264677,
+      "grad_norm": 1.2500115633010864,
+      "learning_rate": 0.00016735234193968678,
+      "loss": 1.1627,
+      "step": 13765
+    },
+    {
+      "epoch": 0.5301251203079884,
+      "grad_norm": 1.0908536911010742,
+      "learning_rate": 0.00016732998677557287,
+      "loss": 1.0477,
+      "step": 13770
+    },
+    {
+      "epoch": 0.5303176130895092,
+      "grad_norm": 1.609208106994629,
+      "learning_rate": 0.0001673076254545777,
+      "loss": 0.964,
+      "step": 13775
+    },
+    {
+      "epoch": 0.5305101058710299,
+      "grad_norm": 1.0210634469985962,
+      "learning_rate": 0.00016728525797874607,
+      "loss": 1.2982,
+      "step": 13780
+    },
+    {
+      "epoch": 0.5307025986525505,
+      "grad_norm": 2.0595545768737793,
+      "learning_rate": 0.0001672628843501233,
+      "loss": 1.1969,
+      "step": 13785
+    },
+    {
+      "epoch": 0.5308950914340712,
+      "grad_norm": 1.7514983415603638,
+      "learning_rate": 0.00016724050457075533,
+      "loss": 1.2918,
+      "step": 13790
+    },
+    {
+      "epoch": 0.5310875842155919,
+      "grad_norm": 1.483798861503601,
+      "learning_rate": 0.00016721811864268865,
+      "loss": 1.0163,
+      "step": 13795
+    },
+    {
+      "epoch": 0.5312800769971127,
+      "grad_norm": 1.4174484014511108,
+      "learning_rate": 0.0001671957265679703,
+      "loss": 1.1936,
+      "step": 13800
+    },
+    {
+      "epoch": 0.5314725697786333,
+      "grad_norm": 1.4664232730865479,
+      "learning_rate": 0.00016717332834864787,
+      "loss": 1.2553,
+      "step": 13805
+    },
+    {
+      "epoch": 0.531665062560154,
+      "grad_norm": 0.6863868832588196,
+      "learning_rate": 0.00016715092398676958,
+      "loss": 0.8998,
+      "step": 13810
+    },
+    {
+      "epoch": 0.5318575553416747,
+      "grad_norm": 2.3511574268341064,
+      "learning_rate": 0.00016712851348438408,
+      "loss": 1.4484,
+      "step": 13815
+    },
+    {
+      "epoch": 0.5320500481231953,
+      "grad_norm": 1.418361783027649,
+      "learning_rate": 0.00016710609684354074,
+      "loss": 1.1139,
+      "step": 13820
+    },
+    {
+      "epoch": 0.5322425409047161,
+      "grad_norm": 1.5918070077896118,
+      "learning_rate": 0.00016708367406628938,
+      "loss": 1.1045,
+      "step": 13825
+    },
+    {
+      "epoch": 0.5324350336862368,
+      "grad_norm": 1.1937044858932495,
+      "learning_rate": 0.00016706124515468042,
+      "loss": 1.2665,
+      "step": 13830
+    },
+    {
+      "epoch": 0.5326275264677575,
+      "grad_norm": 1.60366952419281,
+      "learning_rate": 0.00016703881011076482,
+      "loss": 1.3277,
+      "step": 13835
+    },
+    {
+      "epoch": 0.5328200192492781,
+      "grad_norm": 1.2769535779953003,
+      "learning_rate": 0.00016701636893659414,
+      "loss": 1.2517,
+      "step": 13840
+    },
+    {
+      "epoch": 0.5330125120307988,
+      "grad_norm": 1.3906430006027222,
+      "learning_rate": 0.00016699392163422043,
+      "loss": 1.3485,
+      "step": 13845
+    },
+    {
+      "epoch": 0.5332050048123196,
+      "grad_norm": 1.461391568183899,
+      "learning_rate": 0.0001669714682056964,
+      "loss": 1.1297,
+      "step": 13850
+    },
+    {
+      "epoch": 0.5333974975938403,
+      "grad_norm": 1.3566093444824219,
+      "learning_rate": 0.00016694900865307525,
+      "loss": 1.2833,
+      "step": 13855
+    },
+    {
+      "epoch": 0.5335899903753609,
+      "grad_norm": 1.4480105638504028,
+      "learning_rate": 0.00016692654297841076,
+      "loss": 1.0877,
+      "step": 13860
+    },
+    {
+      "epoch": 0.5337824831568816,
+      "grad_norm": 1.0896391868591309,
+      "learning_rate": 0.00016690407118375724,
+      "loss": 1.1286,
+      "step": 13865
+    },
+    {
+      "epoch": 0.5339749759384023,
+      "grad_norm": 1.101636290550232,
+      "learning_rate": 0.00016688159327116962,
+      "loss": 1.0802,
+      "step": 13870
+    },
+    {
+      "epoch": 0.534167468719923,
+      "grad_norm": 1.1488208770751953,
+      "learning_rate": 0.00016685910924270337,
+      "loss": 1.144,
+      "step": 13875
+    },
+    {
+      "epoch": 0.5343599615014437,
+      "grad_norm": 1.4691115617752075,
+      "learning_rate": 0.00016683661910041445,
+      "loss": 1.2133,
+      "step": 13880
+    },
+    {
+      "epoch": 0.5345524542829644,
+      "grad_norm": 0.9920752048492432,
+      "learning_rate": 0.0001668141228463595,
+      "loss": 1.1326,
+      "step": 13885
+    },
+    {
+      "epoch": 0.534744947064485,
+      "grad_norm": 1.2828654050827026,
+      "learning_rate": 0.00016679162048259557,
+      "loss": 1.2162,
+      "step": 13890
+    },
+    {
+      "epoch": 0.5349374398460057,
+      "grad_norm": 1.3294516801834106,
+      "learning_rate": 0.00016676911201118043,
+      "loss": 1.1797,
+      "step": 13895
+    },
+    {
+      "epoch": 0.5351299326275265,
+      "grad_norm": 1.5326685905456543,
+      "learning_rate": 0.00016674659743417232,
+      "loss": 1.1147,
+      "step": 13900
+    },
+    {
+      "epoch": 0.5353224254090472,
+      "grad_norm": 1.9222960472106934,
+      "learning_rate": 0.00016672407675363,
+      "loss": 1.1615,
+      "step": 13905
+    },
+    {
+      "epoch": 0.5355149181905678,
+      "grad_norm": 1.412458062171936,
+      "learning_rate": 0.00016670154997161288,
+      "loss": 1.1556,
+      "step": 13910
+    },
+    {
+      "epoch": 0.5357074109720885,
+      "grad_norm": 1.230669617652893,
+      "learning_rate": 0.00016667901709018087,
+      "loss": 1.062,
+      "step": 13915
+    },
+    {
+      "epoch": 0.5358999037536092,
+      "grad_norm": 1.431746006011963,
+      "learning_rate": 0.00016665647811139444,
+      "loss": 1.0561,
+      "step": 13920
+    },
+    {
+      "epoch": 0.53609239653513,
+      "grad_norm": 1.6623647212982178,
+      "learning_rate": 0.00016663393303731466,
+      "loss": 1.1495,
+      "step": 13925
+    },
+    {
+      "epoch": 0.5362848893166506,
+      "grad_norm": 1.5261880159378052,
+      "learning_rate": 0.00016661138187000312,
+      "loss": 1.3093,
+      "step": 13930
+    },
+    {
+      "epoch": 0.5364773820981713,
+      "grad_norm": 1.5623407363891602,
+      "learning_rate": 0.00016658882461152195,
+      "loss": 1.0859,
+      "step": 13935
+    },
+    {
+      "epoch": 0.536669874879692,
+      "grad_norm": 1.2155213356018066,
+      "learning_rate": 0.0001665662612639339,
+      "loss": 1.2502,
+      "step": 13940
+    },
+    {
+      "epoch": 0.5368623676612128,
+      "grad_norm": 0.7948794364929199,
+      "learning_rate": 0.0001665436918293022,
+      "loss": 1.1741,
+      "step": 13945
+    },
+    {
+      "epoch": 0.5370548604427334,
+      "grad_norm": 1.370322585105896,
+      "learning_rate": 0.0001665211163096907,
+      "loss": 1.2727,
+      "step": 13950
+    },
+    {
+      "epoch": 0.5372473532242541,
+      "grad_norm": 1.146519660949707,
+      "learning_rate": 0.00016649853470716378,
+      "loss": 1.2603,
+      "step": 13955
+    },
+    {
+      "epoch": 0.5374398460057748,
+      "grad_norm": 1.1492048501968384,
+      "learning_rate": 0.00016647594702378637,
+      "loss": 1.1772,
+      "step": 13960
+    },
+    {
+      "epoch": 0.5376323387872954,
+      "grad_norm": 2.4730112552642822,
+      "learning_rate": 0.00016645335326162397,
+      "loss": 1.4024,
+      "step": 13965
+    },
+    {
+      "epoch": 0.5378248315688162,
+      "grad_norm": 1.411889910697937,
+      "learning_rate": 0.00016643075342274264,
+      "loss": 1.1121,
+      "step": 13970
+    },
+    {
+      "epoch": 0.5380173243503369,
+      "grad_norm": 1.0818617343902588,
+      "learning_rate": 0.00016640814750920895,
+      "loss": 1.2139,
+      "step": 13975
+    },
+    {
+      "epoch": 0.5382098171318576,
+      "grad_norm": 1.1196002960205078,
+      "learning_rate": 0.0001663855355230901,
+      "loss": 1.0877,
+      "step": 13980
+    },
+    {
+      "epoch": 0.5384023099133782,
+      "grad_norm": 1.5476993322372437,
+      "learning_rate": 0.00016636291746645378,
+      "loss": 1.1055,
+      "step": 13985
+    },
+    {
+      "epoch": 0.5385948026948989,
+      "grad_norm": 0.924186646938324,
+      "learning_rate": 0.00016634029334136827,
+      "loss": 1.0307,
+      "step": 13990
+    },
+    {
+      "epoch": 0.5387872954764197,
+      "grad_norm": 1.157355546951294,
+      "learning_rate": 0.0001663176631499024,
+      "loss": 1.0783,
+      "step": 13995
+    },
+    {
+      "epoch": 0.5389797882579404,
+      "grad_norm": 1.1704423427581787,
+      "learning_rate": 0.00016629502689412555,
+      "loss": 1.3452,
+      "step": 14000
+    },
+    {
+      "epoch": 0.539172281039461,
+      "grad_norm": 2.0251457691192627,
+      "learning_rate": 0.00016627238457610766,
+      "loss": 1.3611,
+      "step": 14005
+    },
+    {
+      "epoch": 0.5393647738209817,
+      "grad_norm": 1.018612265586853,
+      "learning_rate": 0.0001662497361979192,
+      "loss": 1.115,
+      "step": 14010
+    },
+    {
+      "epoch": 0.5395572666025024,
+      "grad_norm": 1.2389349937438965,
+      "learning_rate": 0.00016622708176163126,
+      "loss": 1.2055,
+      "step": 14015
+    },
+    {
+      "epoch": 0.5397497593840231,
+      "grad_norm": 2.2555086612701416,
+      "learning_rate": 0.0001662044212693154,
+      "loss": 1.0512,
+      "step": 14020
+    },
+    {
+      "epoch": 0.5399422521655438,
+      "grad_norm": 1.059856653213501,
+      "learning_rate": 0.00016618175472304375,
+      "loss": 1.2114,
+      "step": 14025
+    },
+    {
+      "epoch": 0.5401347449470645,
+      "grad_norm": 1.484417200088501,
+      "learning_rate": 0.00016615908212488906,
+      "loss": 1.1872,
+      "step": 14030
+    },
+    {
+      "epoch": 0.5403272377285852,
+      "grad_norm": 1.4816780090332031,
+      "learning_rate": 0.00016613640347692458,
+      "loss": 1.1261,
+      "step": 14035
+    },
+    {
+      "epoch": 0.5405197305101058,
+      "grad_norm": 1.6735597848892212,
+      "learning_rate": 0.00016611371878122412,
+      "loss": 1.2311,
+      "step": 14040
+    },
+    {
+      "epoch": 0.5407122232916266,
+      "grad_norm": 1.8882919549942017,
+      "learning_rate": 0.00016609102803986204,
+      "loss": 1.3099,
+      "step": 14045
+    },
+    {
+      "epoch": 0.5409047160731473,
+      "grad_norm": 1.4272384643554688,
+      "learning_rate": 0.00016606833125491327,
+      "loss": 1.2343,
+      "step": 14050
+    },
+    {
+      "epoch": 0.5410972088546679,
+      "grad_norm": 1.2361105680465698,
+      "learning_rate": 0.0001660456284284532,
+      "loss": 1.155,
+      "step": 14055
+    },
+    {
+      "epoch": 0.5412897016361886,
+      "grad_norm": 1.294826626777649,
+      "learning_rate": 0.000166022919562558,
+      "loss": 1.0691,
+      "step": 14060
+    },
+    {
+      "epoch": 0.5414821944177093,
+      "grad_norm": 2.163748264312744,
+      "learning_rate": 0.00016600020465930415,
+      "loss": 1.4603,
+      "step": 14065
+    },
+    {
+      "epoch": 0.5416746871992301,
+      "grad_norm": 2.8181777000427246,
+      "learning_rate": 0.00016597748372076878,
+      "loss": 1.1513,
+      "step": 14070
+    },
+    {
+      "epoch": 0.5418671799807507,
+      "grad_norm": 1.558497667312622,
+      "learning_rate": 0.00016595475674902957,
+      "loss": 1.1758,
+      "step": 14075
+    },
+    {
+      "epoch": 0.5420596727622714,
+      "grad_norm": 1.5868738889694214,
+      "learning_rate": 0.0001659320237461648,
+      "loss": 1.1867,
+      "step": 14080
+    },
+    {
+      "epoch": 0.5422521655437921,
+      "grad_norm": 0.850387692451477,
+      "learning_rate": 0.0001659092847142532,
+      "loss": 0.8849,
+      "step": 14085
+    },
+    {
+      "epoch": 0.5424446583253129,
+      "grad_norm": 1.334726095199585,
+      "learning_rate": 0.00016588653965537412,
+      "loss": 1.252,
+      "step": 14090
+    },
+    {
+      "epoch": 0.5426371511068335,
+      "grad_norm": 1.1548973321914673,
+      "learning_rate": 0.00016586378857160743,
+      "loss": 1.2255,
+      "step": 14095
+    },
+    {
+      "epoch": 0.5428296438883542,
+      "grad_norm": 1.3282769918441772,
+      "learning_rate": 0.00016584103146503364,
+      "loss": 1.0991,
+      "step": 14100
+    },
+    {
+      "epoch": 0.5430221366698749,
+      "grad_norm": 1.635657548904419,
+      "learning_rate": 0.00016581826833773363,
+      "loss": 1.1963,
+      "step": 14105
+    },
+    {
+      "epoch": 0.5432146294513955,
+      "grad_norm": 1.7892380952835083,
+      "learning_rate": 0.00016579549919178903,
+      "loss": 1.0593,
+      "step": 14110
+    },
+    {
+      "epoch": 0.5434071222329163,
+      "grad_norm": 2.381394147872925,
+      "learning_rate": 0.00016577272402928183,
+      "loss": 1.2743,
+      "step": 14115
+    },
+    {
+      "epoch": 0.543599615014437,
+      "grad_norm": 1.1770328283309937,
+      "learning_rate": 0.00016574994285229478,
+      "loss": 1.1433,
+      "step": 14120
+    },
+    {
+      "epoch": 0.5437921077959577,
+      "grad_norm": 1.9077178239822388,
+      "learning_rate": 0.00016572715566291098,
+      "loss": 1.2422,
+      "step": 14125
+    },
+    {
+      "epoch": 0.5439846005774783,
+      "grad_norm": 1.2600334882736206,
+      "learning_rate": 0.00016570436246321417,
+      "loss": 1.1479,
+      "step": 14130
+    },
+    {
+      "epoch": 0.544177093358999,
+      "grad_norm": 1.0997780561447144,
+      "learning_rate": 0.0001656815632552887,
+      "loss": 1.2516,
+      "step": 14135
+    },
+    {
+      "epoch": 0.5443695861405198,
+      "grad_norm": 1.1767383813858032,
+      "learning_rate": 0.00016565875804121935,
+      "loss": 1.1713,
+      "step": 14140
+    },
+    {
+      "epoch": 0.5445620789220404,
+      "grad_norm": 1.62860906124115,
+      "learning_rate": 0.00016563594682309152,
+      "loss": 1.2017,
+      "step": 14145
+    },
+    {
+      "epoch": 0.5447545717035611,
+      "grad_norm": 1.6149252653121948,
+      "learning_rate": 0.0001656131296029912,
+      "loss": 1.2104,
+      "step": 14150
+    },
+    {
+      "epoch": 0.5449470644850818,
+      "grad_norm": 1.0693351030349731,
+      "learning_rate": 0.0001655903063830048,
+      "loss": 1.2656,
+      "step": 14155
+    },
+    {
+      "epoch": 0.5451395572666025,
+      "grad_norm": 1.4624438285827637,
+      "learning_rate": 0.00016556747716521937,
+      "loss": 1.2323,
+      "step": 14160
+    },
+    {
+      "epoch": 0.5453320500481232,
+      "grad_norm": 1.8848096132278442,
+      "learning_rate": 0.0001655446419517225,
+      "loss": 1.2104,
+      "step": 14165
+    },
+    {
+      "epoch": 0.5455245428296439,
+      "grad_norm": 1.076907753944397,
+      "learning_rate": 0.00016552180074460231,
+      "loss": 1.2503,
+      "step": 14170
+    },
+    {
+      "epoch": 0.5457170356111646,
+      "grad_norm": 2.496718645095825,
+      "learning_rate": 0.00016549895354594748,
+      "loss": 1.146,
+      "step": 14175
+    },
+    {
+      "epoch": 0.5459095283926853,
+      "grad_norm": 1.8133556842803955,
+      "learning_rate": 0.00016547610035784724,
+      "loss": 1.274,
+      "step": 14180
+    },
+    {
+      "epoch": 0.5461020211742059,
+      "grad_norm": 1.1353720426559448,
+      "learning_rate": 0.0001654532411823914,
+      "loss": 1.3842,
+      "step": 14185
+    },
+    {
+      "epoch": 0.5462945139557267,
+      "grad_norm": 2.368894577026367,
+      "learning_rate": 0.00016543037602167017,
+      "loss": 1.3566,
+      "step": 14190
+    },
+    {
+      "epoch": 0.5464870067372474,
+      "grad_norm": 1.884104609489441,
+      "learning_rate": 0.00016540750487777455,
+      "loss": 1.15,
+      "step": 14195
+    },
+    {
+      "epoch": 0.546679499518768,
+      "grad_norm": 1.1348326206207275,
+      "learning_rate": 0.00016538462775279587,
+      "loss": 1.1782,
+      "step": 14200
+    },
+    {
+      "epoch": 0.5468719923002887,
+      "grad_norm": 1.2342017889022827,
+      "learning_rate": 0.00016536174464882613,
+      "loss": 1.1361,
+      "step": 14205
+    },
+    {
+      "epoch": 0.5470644850818094,
+      "grad_norm": 1.0037345886230469,
+      "learning_rate": 0.0001653388555679578,
+      "loss": 1.1282,
+      "step": 14210
+    },
+    {
+      "epoch": 0.5472569778633302,
+      "grad_norm": 2.8669965267181396,
+      "learning_rate": 0.000165315960512284,
+      "loss": 1.184,
+      "step": 14215
+    },
+    {
+      "epoch": 0.5474494706448508,
+      "grad_norm": 1.0212280750274658,
+      "learning_rate": 0.00016529305948389825,
+      "loss": 1.1422,
+      "step": 14220
+    },
+    {
+      "epoch": 0.5476419634263715,
+      "grad_norm": 1.1197772026062012,
+      "learning_rate": 0.00016527015248489474,
+      "loss": 1.077,
+      "step": 14225
+    },
+    {
+      "epoch": 0.5478344562078922,
+      "grad_norm": 1.4821882247924805,
+      "learning_rate": 0.0001652472395173682,
+      "loss": 1.3187,
+      "step": 14230
+    },
+    {
+      "epoch": 0.5480269489894128,
+      "grad_norm": 1.1993844509124756,
+      "learning_rate": 0.00016522432058341377,
+      "loss": 1.1834,
+      "step": 14235
+    },
+    {
+      "epoch": 0.5482194417709336,
+      "grad_norm": 1.9386481046676636,
+      "learning_rate": 0.00016520139568512734,
+      "loss": 1.1461,
+      "step": 14240
+    },
+    {
+      "epoch": 0.5484119345524543,
+      "grad_norm": 0.8914703130722046,
+      "learning_rate": 0.00016517846482460517,
+      "loss": 1.4175,
+      "step": 14245
+    },
+    {
+      "epoch": 0.548604427333975,
+      "grad_norm": 1.8703666925430298,
+      "learning_rate": 0.00016515552800394417,
+      "loss": 1.2483,
+      "step": 14250
+    },
+    {
+      "epoch": 0.5487969201154956,
+      "grad_norm": 1.1656851768493652,
+      "learning_rate": 0.00016513258522524177,
+      "loss": 1.2293,
+      "step": 14255
+    },
+    {
+      "epoch": 0.5489894128970164,
+      "grad_norm": 1.402370810508728,
+      "learning_rate": 0.0001651096364905959,
+      "loss": 1.3326,
+      "step": 14260
+    },
+    {
+      "epoch": 0.5491819056785371,
+      "grad_norm": 1.8804208040237427,
+      "learning_rate": 0.00016508668180210506,
+      "loss": 1.1033,
+      "step": 14265
+    },
+    {
+      "epoch": 0.5493743984600578,
+      "grad_norm": 1.0970590114593506,
+      "learning_rate": 0.00016506372116186836,
+      "loss": 1.1887,
+      "step": 14270
+    },
+    {
+      "epoch": 0.5495668912415784,
+      "grad_norm": 1.3364982604980469,
+      "learning_rate": 0.00016504075457198533,
+      "loss": 1.1183,
+      "step": 14275
+    },
+    {
+      "epoch": 0.5497593840230991,
+      "grad_norm": 1.4718800783157349,
+      "learning_rate": 0.0001650177820345562,
+      "loss": 1.1305,
+      "step": 14280
+    },
+    {
+      "epoch": 0.5499518768046199,
+      "grad_norm": 1.3023836612701416,
+      "learning_rate": 0.00016499480355168156,
+      "loss": 1.2267,
+      "step": 14285
+    },
+    {
+      "epoch": 0.5501443695861405,
+      "grad_norm": 1.809346079826355,
+      "learning_rate": 0.0001649718191254627,
+      "loss": 1.081,
+      "step": 14290
+    },
+    {
+      "epoch": 0.5503368623676612,
+      "grad_norm": 1.3828262090682983,
+      "learning_rate": 0.0001649488287580014,
+      "loss": 1.209,
+      "step": 14295
+    },
+    {
+      "epoch": 0.5505293551491819,
+      "grad_norm": 1.4741365909576416,
+      "learning_rate": 0.00016492583245139995,
+      "loss": 1.0607,
+      "step": 14300
+    },
+    {
+      "epoch": 0.5507218479307026,
+      "grad_norm": 1.057210922241211,
+      "learning_rate": 0.0001649028302077612,
+      "loss": 1.281,
+      "step": 14305
+    },
+    {
+      "epoch": 0.5509143407122233,
+      "grad_norm": 2.588911294937134,
+      "learning_rate": 0.00016487982202918858,
+      "loss": 1.382,
+      "step": 14310
+    },
+    {
+      "epoch": 0.551106833493744,
+      "grad_norm": 2.2811248302459717,
+      "learning_rate": 0.00016485680791778604,
+      "loss": 1.3173,
+      "step": 14315
+    },
+    {
+      "epoch": 0.5512993262752647,
+      "grad_norm": 1.675776481628418,
+      "learning_rate": 0.00016483378787565802,
+      "loss": 1.1948,
+      "step": 14320
+    },
+    {
+      "epoch": 0.5514918190567853,
+      "grad_norm": 1.1149309873580933,
+      "learning_rate": 0.0001648107619049096,
+      "loss": 1.1406,
+      "step": 14325
+    },
+    {
+      "epoch": 0.551684311838306,
+      "grad_norm": 1.0165066719055176,
+      "learning_rate": 0.00016478773000764635,
+      "loss": 1.1491,
+      "step": 14330
+    },
+    {
+      "epoch": 0.5518768046198268,
+      "grad_norm": 1.8692020177841187,
+      "learning_rate": 0.00016476469218597433,
+      "loss": 1.0848,
+      "step": 14335
+    },
+    {
+      "epoch": 0.5520692974013475,
+      "grad_norm": 0.9627811908721924,
+      "learning_rate": 0.0001647416484420003,
+      "loss": 1.2159,
+      "step": 14340
+    },
+    {
+      "epoch": 0.5522617901828681,
+      "grad_norm": 2.1085097789764404,
+      "learning_rate": 0.00016471859877783133,
+      "loss": 1.1551,
+      "step": 14345
+    },
+    {
+      "epoch": 0.5524542829643888,
+      "grad_norm": 2.2478790283203125,
+      "learning_rate": 0.00016469554319557527,
+      "loss": 1.3081,
+      "step": 14350
+    },
+    {
+      "epoch": 0.5526467757459095,
+      "grad_norm": 1.0580302476882935,
+      "learning_rate": 0.00016467248169734037,
+      "loss": 1.3293,
+      "step": 14355
+    },
+    {
+      "epoch": 0.5528392685274303,
+      "grad_norm": 1.3953101634979248,
+      "learning_rate": 0.00016464941428523538,
+      "loss": 1.1256,
+      "step": 14360
+    },
+    {
+      "epoch": 0.5530317613089509,
+      "grad_norm": 0.9302542209625244,
+      "learning_rate": 0.0001646263409613697,
+      "loss": 0.9645,
+      "step": 14365
+    },
+    {
+      "epoch": 0.5532242540904716,
+      "grad_norm": 1.9415937662124634,
+      "learning_rate": 0.00016460326172785332,
+      "loss": 1.3428,
+      "step": 14370
+    },
+    {
+      "epoch": 0.5534167468719923,
+      "grad_norm": 0.9449756145477295,
+      "learning_rate": 0.00016458017658679656,
+      "loss": 1.3183,
+      "step": 14375
+    },
+    {
+      "epoch": 0.5536092396535129,
+      "grad_norm": 1.2944326400756836,
+      "learning_rate": 0.00016455708554031047,
+      "loss": 1.1277,
+      "step": 14380
+    },
+    {
+      "epoch": 0.5538017324350337,
+      "grad_norm": 1.3632171154022217,
+      "learning_rate": 0.00016453398859050657,
+      "loss": 1.3262,
+      "step": 14385
+    },
+    {
+      "epoch": 0.5539942252165544,
+      "grad_norm": 1.119086503982544,
+      "learning_rate": 0.00016451088573949692,
+      "loss": 1.2639,
+      "step": 14390
+    },
+    {
+      "epoch": 0.5541867179980751,
+      "grad_norm": 1.3261640071868896,
+      "learning_rate": 0.00016448777698939407,
+      "loss": 1.0911,
+      "step": 14395
+    },
+    {
+      "epoch": 0.5543792107795957,
+      "grad_norm": 1.6098653078079224,
+      "learning_rate": 0.00016446466234231125,
+      "loss": 1.2942,
+      "step": 14400
+    },
+    {
+      "epoch": 0.5545717035611165,
+      "grad_norm": 2.1425249576568604,
+      "learning_rate": 0.0001644415418003621,
+      "loss": 1.1566,
+      "step": 14405
+    },
+    {
+      "epoch": 0.5547641963426372,
+      "grad_norm": 1.0087484121322632,
+      "learning_rate": 0.0001644184153656608,
+      "loss": 1.1515,
+      "step": 14410
+    },
+    {
+      "epoch": 0.5549566891241579,
+      "grad_norm": 1.3792825937271118,
+      "learning_rate": 0.00016439528304032218,
+      "loss": 1.3815,
+      "step": 14415
+    },
+    {
+      "epoch": 0.5551491819056785,
+      "grad_norm": 1.0076264142990112,
+      "learning_rate": 0.0001643721448264615,
+      "loss": 1.1996,
+      "step": 14420
+    },
+    {
+      "epoch": 0.5553416746871992,
+      "grad_norm": 1.6108455657958984,
+      "learning_rate": 0.0001643490007261946,
+      "loss": 1.2801,
+      "step": 14425
+    },
+    {
+      "epoch": 0.55553416746872,
+      "grad_norm": 1.4850428104400635,
+      "learning_rate": 0.00016432585074163783,
+      "loss": 1.1272,
+      "step": 14430
+    },
+    {
+      "epoch": 0.5557266602502406,
+      "grad_norm": 1.482926607131958,
+      "learning_rate": 0.0001643026948749082,
+      "loss": 0.9271,
+      "step": 14435
+    },
+    {
+      "epoch": 0.5559191530317613,
+      "grad_norm": 1.404266119003296,
+      "learning_rate": 0.000164279533128123,
+      "loss": 1.2476,
+      "step": 14440
+    },
+    {
+      "epoch": 0.556111645813282,
+      "grad_norm": 1.5951578617095947,
+      "learning_rate": 0.00016425636550340035,
+      "loss": 1.036,
+      "step": 14445
+    },
+    {
+      "epoch": 0.5563041385948027,
+      "grad_norm": 1.11802339553833,
+      "learning_rate": 0.00016423319200285877,
+      "loss": 1.2595,
+      "step": 14450
+    },
+    {
+      "epoch": 0.5564966313763234,
+      "grad_norm": 0.9702684283256531,
+      "learning_rate": 0.00016421001262861723,
+      "loss": 1.1478,
+      "step": 14455
+    },
+    {
+      "epoch": 0.5566891241578441,
+      "grad_norm": 1.4077606201171875,
+      "learning_rate": 0.00016418682738279542,
+      "loss": 1.2807,
+      "step": 14460
+    },
+    {
+      "epoch": 0.5568816169393648,
+      "grad_norm": 1.5000783205032349,
+      "learning_rate": 0.00016416363626751344,
+      "loss": 1.3231,
+      "step": 14465
+    },
+    {
+      "epoch": 0.5570741097208854,
+      "grad_norm": 1.0804152488708496,
+      "learning_rate": 0.00016414043928489195,
+      "loss": 1.2609,
+      "step": 14470
+    },
+    {
+      "epoch": 0.5572666025024061,
+      "grad_norm": 2.0902814865112305,
+      "learning_rate": 0.0001641172364370522,
+      "loss": 1.1007,
+      "step": 14475
+    },
+    {
+      "epoch": 0.5574590952839269,
+      "grad_norm": 0.9129114151000977,
+      "learning_rate": 0.0001640940277261159,
+      "loss": 1.2385,
+      "step": 14480
+    },
+    {
+      "epoch": 0.5576515880654476,
+      "grad_norm": 1.5251227617263794,
+      "learning_rate": 0.0001640708131542054,
+      "loss": 1.3068,
+      "step": 14485
+    },
+    {
+      "epoch": 0.5578440808469682,
+      "grad_norm": 1.7822771072387695,
+      "learning_rate": 0.00016404759272344342,
+      "loss": 1.2942,
+      "step": 14490
+    },
+    {
+      "epoch": 0.5580365736284889,
+      "grad_norm": 1.7675615549087524,
+      "learning_rate": 0.00016402436643595336,
+      "loss": 0.9753,
+      "step": 14495
+    },
+    {
+      "epoch": 0.5582290664100096,
+      "grad_norm": 1.4113742113113403,
+      "learning_rate": 0.0001640011342938591,
+      "loss": 1.2727,
+      "step": 14500
+    },
+    {
+      "epoch": 0.5584215591915304,
+      "grad_norm": 2.213724136352539,
+      "learning_rate": 0.0001639778962992851,
+      "loss": 1.1275,
+      "step": 14505
+    },
+    {
+      "epoch": 0.558614051973051,
+      "grad_norm": 0.777229368686676,
+      "learning_rate": 0.0001639546524543563,
+      "loss": 1.0434,
+      "step": 14510
+    },
+    {
+      "epoch": 0.5588065447545717,
+      "grad_norm": 1.0420740842819214,
+      "learning_rate": 0.00016393140276119817,
+      "loss": 1.1202,
+      "step": 14515
+    },
+    {
+      "epoch": 0.5589990375360924,
+      "grad_norm": 1.4241138696670532,
+      "learning_rate": 0.00016390814722193678,
+      "loss": 1.0245,
+      "step": 14520
+    },
+    {
+      "epoch": 0.559191530317613,
+      "grad_norm": 1.1826037168502808,
+      "learning_rate": 0.00016388488583869872,
+      "loss": 1.1894,
+      "step": 14525
+    },
+    {
+      "epoch": 0.5593840230991338,
+      "grad_norm": 1.136072039604187,
+      "learning_rate": 0.000163861618613611,
+      "loss": 1.2093,
+      "step": 14530
+    },
+    {
+      "epoch": 0.5595765158806545,
+      "grad_norm": 1.0932581424713135,
+      "learning_rate": 0.0001638383455488013,
+      "loss": 1.26,
+      "step": 14535
+    },
+    {
+      "epoch": 0.5597690086621752,
+      "grad_norm": 1.4892606735229492,
+      "learning_rate": 0.00016381506664639784,
+      "loss": 1.0244,
+      "step": 14540
+    },
+    {
+      "epoch": 0.5599615014436958,
+      "grad_norm": 1.4259272813796997,
+      "learning_rate": 0.0001637917819085292,
+      "loss": 0.9896,
+      "step": 14545
+    },
+    {
+      "epoch": 0.5601539942252165,
+      "grad_norm": 1.0615971088409424,
+      "learning_rate": 0.00016376849133732473,
+      "loss": 1.1619,
+      "step": 14550
+    },
+    {
+      "epoch": 0.5603464870067373,
+      "grad_norm": 0.8815811276435852,
+      "learning_rate": 0.00016374519493491413,
+      "loss": 1.1123,
+      "step": 14555
+    },
+    {
+      "epoch": 0.560538979788258,
+      "grad_norm": 1.2956461906433105,
+      "learning_rate": 0.00016372189270342778,
+      "loss": 1.1978,
+      "step": 14560
+    },
+    {
+      "epoch": 0.5607314725697786,
+      "grad_norm": 1.8797427415847778,
+      "learning_rate": 0.00016369858464499641,
+      "loss": 1.4186,
+      "step": 14565
+    },
+    {
+      "epoch": 0.5609239653512993,
+      "grad_norm": 1.6631108522415161,
+      "learning_rate": 0.00016367527076175143,
+      "loss": 0.9839,
+      "step": 14570
+    },
+    {
+      "epoch": 0.5611164581328201,
+      "grad_norm": 1.8200160264968872,
+      "learning_rate": 0.0001636519510558248,
+      "loss": 1.1272,
+      "step": 14575
+    },
+    {
+      "epoch": 0.5613089509143407,
+      "grad_norm": 1.884712815284729,
+      "learning_rate": 0.00016362862552934886,
+      "loss": 1.15,
+      "step": 14580
+    },
+    {
+      "epoch": 0.5615014436958614,
+      "grad_norm": 0.7094476222991943,
+      "learning_rate": 0.00016360529418445662,
+      "loss": 0.9581,
+      "step": 14585
+    },
+    {
+      "epoch": 0.5616939364773821,
+      "grad_norm": 0.9652591347694397,
+      "learning_rate": 0.00016358195702328158,
+      "loss": 1.0858,
+      "step": 14590
+    },
+    {
+      "epoch": 0.5618864292589028,
+      "grad_norm": 1.3010308742523193,
+      "learning_rate": 0.00016355861404795778,
+      "loss": 1.2491,
+      "step": 14595
+    },
+    {
+      "epoch": 0.5620789220404235,
+      "grad_norm": 1.459953784942627,
+      "learning_rate": 0.00016353526526061973,
+      "loss": 1.1194,
+      "step": 14600
+    },
+    {
+      "epoch": 0.5622714148219442,
+      "grad_norm": 1.0818215608596802,
+      "learning_rate": 0.0001635119106634026,
+      "loss": 1.202,
+      "step": 14605
+    },
+    {
+      "epoch": 0.5624639076034649,
+      "grad_norm": 1.0625619888305664,
+      "learning_rate": 0.0001634885502584419,
+      "loss": 1.3284,
+      "step": 14610
+    },
+    {
+      "epoch": 0.5626564003849855,
+      "grad_norm": 1.5708478689193726,
+      "learning_rate": 0.0001634651840478739,
+      "loss": 1.036,
+      "step": 14615
+    },
+    {
+      "epoch": 0.5628488931665062,
+      "grad_norm": 1.2847293615341187,
+      "learning_rate": 0.00016344181203383523,
+      "loss": 1.0858,
+      "step": 14620
+    },
+    {
+      "epoch": 0.563041385948027,
+      "grad_norm": 1.082689881324768,
+      "learning_rate": 0.00016341843421846313,
+      "loss": 1.3457,
+      "step": 14625
+    },
+    {
+      "epoch": 0.5632338787295477,
+      "grad_norm": 1.9000965356826782,
+      "learning_rate": 0.0001633950506038953,
+      "loss": 1.3901,
+      "step": 14630
+    },
+    {
+      "epoch": 0.5634263715110683,
+      "grad_norm": 1.4664018154144287,
+      "learning_rate": 0.0001633716611922701,
+      "loss": 1.0836,
+      "step": 14635
+    },
+    {
+      "epoch": 0.563618864292589,
+      "grad_norm": 1.6126337051391602,
+      "learning_rate": 0.0001633482659857262,
+      "loss": 1.0794,
+      "step": 14640
+    },
+    {
+      "epoch": 0.5638113570741097,
+      "grad_norm": 1.865504503250122,
+      "learning_rate": 0.00016332486498640307,
+      "loss": 0.9427,
+      "step": 14645
+    },
+    {
+      "epoch": 0.5640038498556305,
+      "grad_norm": 1.4346791505813599,
+      "learning_rate": 0.0001633014581964405,
+      "loss": 1.1952,
+      "step": 14650
+    },
+    {
+      "epoch": 0.5641963426371511,
+      "grad_norm": 1.3558484315872192,
+      "learning_rate": 0.00016327804561797895,
+      "loss": 1.1679,
+      "step": 14655
+    },
+    {
+      "epoch": 0.5643888354186718,
+      "grad_norm": 1.3297834396362305,
+      "learning_rate": 0.00016325462725315926,
+      "loss": 1.2225,
+      "step": 14660
+    },
+    {
+      "epoch": 0.5645813282001925,
+      "grad_norm": 2.106694221496582,
+      "learning_rate": 0.00016323120310412297,
+      "loss": 1.072,
+      "step": 14665
+    },
+    {
+      "epoch": 0.5647738209817131,
+      "grad_norm": 1.284629225730896,
+      "learning_rate": 0.00016320777317301198,
+      "loss": 1.0004,
+      "step": 14670
+    },
+    {
+      "epoch": 0.5649663137632339,
+      "grad_norm": 1.4289201498031616,
+      "learning_rate": 0.0001631843374619689,
+      "loss": 1.1239,
+      "step": 14675
+    },
+    {
+      "epoch": 0.5651588065447546,
+      "grad_norm": 1.9027820825576782,
+      "learning_rate": 0.0001631608959731367,
+      "loss": 1.2137,
+      "step": 14680
+    },
+    {
+      "epoch": 0.5653512993262753,
+      "grad_norm": 1.878009557723999,
+      "learning_rate": 0.00016313744870865895,
+      "loss": 1.247,
+      "step": 14685
+    },
+    {
+      "epoch": 0.5655437921077959,
+      "grad_norm": 1.4919451475143433,
+      "learning_rate": 0.00016311399567067974,
+      "loss": 1.3506,
+      "step": 14690
+    },
+    {
+      "epoch": 0.5657362848893166,
+      "grad_norm": 2.0583205223083496,
+      "learning_rate": 0.00016309053686134378,
+      "loss": 1.1191,
+      "step": 14695
+    },
+    {
+      "epoch": 0.5659287776708374,
+      "grad_norm": 1.1545616388320923,
+      "learning_rate": 0.00016306707228279615,
+      "loss": 1.2105,
+      "step": 14700
+    },
+    {
+      "epoch": 0.566121270452358,
+      "grad_norm": 0.8714199662208557,
+      "learning_rate": 0.0001630436019371825,
+      "loss": 1.0834,
+      "step": 14705
+    },
+    {
+      "epoch": 0.5663137632338787,
+      "grad_norm": 2.1866228580474854,
+      "learning_rate": 0.0001630201258266491,
+      "loss": 1.3334,
+      "step": 14710
+    },
+    {
+      "epoch": 0.5665062560153994,
+      "grad_norm": 1.4117622375488281,
+      "learning_rate": 0.00016299664395334266,
+      "loss": 1.1353,
+      "step": 14715
+    },
+    {
+      "epoch": 0.5666987487969202,
+      "grad_norm": 1.5454515218734741,
+      "learning_rate": 0.00016297315631941045,
+      "loss": 1.096,
+      "step": 14720
+    },
+    {
+      "epoch": 0.5668912415784408,
+      "grad_norm": 1.1799986362457275,
+      "learning_rate": 0.00016294966292700026,
+      "loss": 1.214,
+      "step": 14725
+    },
+    {
+      "epoch": 0.5670837343599615,
+      "grad_norm": 1.2906007766723633,
+      "learning_rate": 0.00016292616377826038,
+      "loss": 1.2613,
+      "step": 14730
+    },
+    {
+      "epoch": 0.5672762271414822,
+      "grad_norm": 2.8731329441070557,
+      "learning_rate": 0.00016290265887533968,
+      "loss": 1.3257,
+      "step": 14735
+    },
+    {
+      "epoch": 0.5674687199230029,
+      "grad_norm": 1.0078117847442627,
+      "learning_rate": 0.0001628791482203875,
+      "loss": 1.2053,
+      "step": 14740
+    },
+    {
+      "epoch": 0.5676612127045236,
+      "grad_norm": 1.05767023563385,
+      "learning_rate": 0.0001628556318155538,
+      "loss": 0.9775,
+      "step": 14745
+    },
+    {
+      "epoch": 0.5678537054860443,
+      "grad_norm": 2.118110418319702,
+      "learning_rate": 0.0001628321096629889,
+      "loss": 1.1801,
+      "step": 14750
+    },
+    {
+      "epoch": 0.568046198267565,
+      "grad_norm": 1.1577699184417725,
+      "learning_rate": 0.00016280858176484384,
+      "loss": 1.2156,
+      "step": 14755
+    },
+    {
+      "epoch": 0.5682386910490856,
+      "grad_norm": 1.5565030574798584,
+      "learning_rate": 0.00016278504812327002,
+      "loss": 1.0586,
+      "step": 14760
+    },
+    {
+      "epoch": 0.5684311838306063,
+      "grad_norm": 1.5205986499786377,
+      "learning_rate": 0.00016276150874041946,
+      "loss": 1.3679,
+      "step": 14765
+    },
+    {
+      "epoch": 0.5686236766121271,
+      "grad_norm": 0.9402291774749756,
+      "learning_rate": 0.00016273796361844468,
+      "loss": 1.0996,
+      "step": 14770
+    },
+    {
+      "epoch": 0.5688161693936478,
+      "grad_norm": 1.3806294202804565,
+      "learning_rate": 0.00016271441275949875,
+      "loss": 1.1815,
+      "step": 14775
+    },
+    {
+      "epoch": 0.5690086621751684,
+      "grad_norm": 2.0714609622955322,
+      "learning_rate": 0.0001626908561657352,
+      "loss": 1.1945,
+      "step": 14780
+    },
+    {
+      "epoch": 0.5692011549566891,
+      "grad_norm": 0.9732249975204468,
+      "learning_rate": 0.00016266729383930816,
+      "loss": 1.0233,
+      "step": 14785
+    },
+    {
+      "epoch": 0.5693936477382098,
+      "grad_norm": 1.3748955726623535,
+      "learning_rate": 0.0001626437257823722,
+      "loss": 1.25,
+      "step": 14790
+    },
+    {
+      "epoch": 0.5695861405197306,
+      "grad_norm": 1.9781707525253296,
+      "learning_rate": 0.00016262015199708252,
+      "loss": 1.1745,
+      "step": 14795
+    },
+    {
+      "epoch": 0.5697786333012512,
+      "grad_norm": 1.5062282085418701,
+      "learning_rate": 0.00016259657248559475,
+      "loss": 1.098,
+      "step": 14800
+    },
+    {
+      "epoch": 0.5699711260827719,
+      "grad_norm": 1.7073885202407837,
+      "learning_rate": 0.0001625729872500651,
+      "loss": 1.1191,
+      "step": 14805
+    },
+    {
+      "epoch": 0.5701636188642926,
+      "grad_norm": 2.0891575813293457,
+      "learning_rate": 0.00016254939629265026,
+      "loss": 1.2533,
+      "step": 14810
+    },
+    {
+      "epoch": 0.5703561116458132,
+      "grad_norm": 1.6380434036254883,
+      "learning_rate": 0.0001625257996155075,
+      "loss": 1.2756,
+      "step": 14815
+    },
+    {
+      "epoch": 0.570548604427334,
+      "grad_norm": 1.1182715892791748,
+      "learning_rate": 0.00016250219722079452,
+      "loss": 1.3084,
+      "step": 14820
+    },
+    {
+      "epoch": 0.5707410972088547,
+      "grad_norm": 1.2113651037216187,
+      "learning_rate": 0.0001624785891106697,
+      "loss": 1.287,
+      "step": 14825
+    },
+    {
+      "epoch": 0.5709335899903754,
+      "grad_norm": 1.1726208925247192,
+      "learning_rate": 0.00016245497528729174,
+      "loss": 1.1491,
+      "step": 14830
+    },
+    {
+      "epoch": 0.571126082771896,
+      "grad_norm": 1.0203557014465332,
+      "learning_rate": 0.00016243135575282004,
+      "loss": 1.0809,
+      "step": 14835
+    },
+    {
+      "epoch": 0.5713185755534167,
+      "grad_norm": 1.2878923416137695,
+      "learning_rate": 0.00016240773050941443,
+      "loss": 1.1848,
+      "step": 14840
+    },
+    {
+      "epoch": 0.5715110683349375,
+      "grad_norm": 1.5805665254592896,
+      "learning_rate": 0.00016238409955923527,
+      "loss": 1.1191,
+      "step": 14845
+    },
+    {
+      "epoch": 0.5717035611164581,
+      "grad_norm": 1.089296579360962,
+      "learning_rate": 0.00016236046290444347,
+      "loss": 1.066,
+      "step": 14850
+    },
+    {
+      "epoch": 0.5718960538979788,
+      "grad_norm": 1.1492708921432495,
+      "learning_rate": 0.0001623368205472004,
+      "loss": 1.174,
+      "step": 14855
+    },
+    {
+      "epoch": 0.5720885466794995,
+      "grad_norm": 1.9744573831558228,
+      "learning_rate": 0.00016231317248966809,
+      "loss": 1.248,
+      "step": 14860
+    },
+    {
+      "epoch": 0.5722810394610202,
+      "grad_norm": 2.2061898708343506,
+      "learning_rate": 0.0001622895187340089,
+      "loss": 1.2028,
+      "step": 14865
+    },
+    {
+      "epoch": 0.5724735322425409,
+      "grad_norm": 1.0993640422821045,
+      "learning_rate": 0.0001622658592823859,
+      "loss": 1.059,
+      "step": 14870
+    },
+    {
+      "epoch": 0.5726660250240616,
+      "grad_norm": 1.5680936574935913,
+      "learning_rate": 0.00016224219413696252,
+      "loss": 1.4181,
+      "step": 14875
+    },
+    {
+      "epoch": 0.5728585178055823,
+      "grad_norm": 1.3295773267745972,
+      "learning_rate": 0.00016221852329990276,
+      "loss": 1.214,
+      "step": 14880
+    },
+    {
+      "epoch": 0.573051010587103,
+      "grad_norm": 1.2004729509353638,
+      "learning_rate": 0.00016219484677337126,
+      "loss": 1.1474,
+      "step": 14885
+    },
+    {
+      "epoch": 0.5732435033686237,
+      "grad_norm": 1.3868520259857178,
+      "learning_rate": 0.000162171164559533,
+      "loss": 1.2373,
+      "step": 14890
+    },
+    {
+      "epoch": 0.5734359961501444,
+      "grad_norm": 1.2218377590179443,
+      "learning_rate": 0.00016214747666055358,
+      "loss": 1.1009,
+      "step": 14895
+    },
+    {
+      "epoch": 0.5736284889316651,
+      "grad_norm": 1.1113415956497192,
+      "learning_rate": 0.00016212378307859914,
+      "loss": 1.2191,
+      "step": 14900
+    },
+    {
+      "epoch": 0.5738209817131857,
+      "grad_norm": 1.099223256111145,
+      "learning_rate": 0.00016210008381583623,
+      "loss": 1.2024,
+      "step": 14905
+    },
+    {
+      "epoch": 0.5740134744947064,
+      "grad_norm": 1.3597705364227295,
+      "learning_rate": 0.00016207637887443208,
+      "loss": 1.1785,
+      "step": 14910
+    },
+    {
+      "epoch": 0.5742059672762272,
+      "grad_norm": 1.675276279449463,
+      "learning_rate": 0.00016205266825655427,
+      "loss": 1.1492,
+      "step": 14915
+    },
+    {
+      "epoch": 0.5743984600577479,
+      "grad_norm": 1.5977553129196167,
+      "learning_rate": 0.000162028951964371,
+      "loss": 1.1355,
+      "step": 14920
+    },
+    {
+      "epoch": 0.5745909528392685,
+      "grad_norm": 2.0862395763397217,
+      "learning_rate": 0.000162005230000051,
+      "loss": 1.1747,
+      "step": 14925
+    },
+    {
+      "epoch": 0.5747834456207892,
+      "grad_norm": 0.8812354803085327,
+      "learning_rate": 0.00016198150236576347,
+      "loss": 1.2876,
+      "step": 14930
+    },
+    {
+      "epoch": 0.5749759384023099,
+      "grad_norm": 1.3878661394119263,
+      "learning_rate": 0.0001619577690636781,
+      "loss": 1.2269,
+      "step": 14935
+    },
+    {
+      "epoch": 0.5751684311838307,
+      "grad_norm": 1.0739976167678833,
+      "learning_rate": 0.0001619340300959652,
+      "loss": 1.2467,
+      "step": 14940
+    },
+    {
+      "epoch": 0.5753609239653513,
+      "grad_norm": 0.766392707824707,
+      "learning_rate": 0.0001619102854647955,
+      "loss": 1.0829,
+      "step": 14945
+    },
+    {
+      "epoch": 0.575553416746872,
+      "grad_norm": 1.2837680578231812,
+      "learning_rate": 0.00016188653517234036,
+      "loss": 1.2027,
+      "step": 14950
+    },
+    {
+      "epoch": 0.5757459095283927,
+      "grad_norm": 2.2257256507873535,
+      "learning_rate": 0.00016186277922077152,
+      "loss": 1.1181,
+      "step": 14955
+    },
+    {
+      "epoch": 0.5759384023099133,
+      "grad_norm": 1.257380723953247,
+      "learning_rate": 0.00016183901761226133,
+      "loss": 1.1899,
+      "step": 14960
+    },
+    {
+      "epoch": 0.5761308950914341,
+      "grad_norm": 1.1324365139007568,
+      "learning_rate": 0.00016181525034898261,
+      "loss": 1.0823,
+      "step": 14965
+    },
+    {
+      "epoch": 0.5763233878729548,
+      "grad_norm": 0.9696788787841797,
+      "learning_rate": 0.00016179147743310872,
+      "loss": 1.176,
+      "step": 14970
+    },
+    {
+      "epoch": 0.5765158806544755,
+      "grad_norm": 1.8557454347610474,
+      "learning_rate": 0.00016176769886681357,
+      "loss": 1.1396,
+      "step": 14975
+    },
+    {
+      "epoch": 0.5767083734359961,
+      "grad_norm": 1.2395600080490112,
+      "learning_rate": 0.00016174391465227154,
+      "loss": 1.0799,
+      "step": 14980
+    },
+    {
+      "epoch": 0.5769008662175168,
+      "grad_norm": 1.8957431316375732,
+      "learning_rate": 0.00016172012479165752,
+      "loss": 1.146,
+      "step": 14985
+    },
+    {
+      "epoch": 0.5770933589990376,
+      "grad_norm": 1.191486120223999,
+      "learning_rate": 0.00016169632928714697,
+      "loss": 1.0166,
+      "step": 14990
+    },
+    {
+      "epoch": 0.5772858517805582,
+      "grad_norm": 1.7964496612548828,
+      "learning_rate": 0.0001616725281409158,
+      "loss": 1.2131,
+      "step": 14995
+    },
+    {
+      "epoch": 0.5774783445620789,
+      "grad_norm": 1.4722768068313599,
+      "learning_rate": 0.00016164872135514044,
+      "loss": 1.0148,
+      "step": 15000
+    },
+    {
+      "epoch": 0.5776708373435996,
+      "grad_norm": 1.265663981437683,
+      "learning_rate": 0.00016162490893199791,
+      "loss": 1.1166,
+      "step": 15005
+    },
+    {
+      "epoch": 0.5778633301251203,
+      "grad_norm": 1.2796491384506226,
+      "learning_rate": 0.0001616010908736657,
+      "loss": 1.2911,
+      "step": 15010
+    },
+    {
+      "epoch": 0.578055822906641,
+      "grad_norm": 1.025158166885376,
+      "learning_rate": 0.00016157726718232177,
+      "loss": 1.0723,
+      "step": 15015
+    },
+    {
+      "epoch": 0.5782483156881617,
+      "grad_norm": 1.5206444263458252,
+      "learning_rate": 0.00016155343786014472,
+      "loss": 1.0406,
+      "step": 15020
+    },
+    {
+      "epoch": 0.5784408084696824,
+      "grad_norm": 1.5212637186050415,
+      "learning_rate": 0.0001615296029093135,
+      "loss": 1.0445,
+      "step": 15025
+    },
+    {
+      "epoch": 0.578633301251203,
+      "grad_norm": 1.3746932744979858,
+      "learning_rate": 0.0001615057623320077,
+      "loss": 1.1385,
+      "step": 15030
+    },
+    {
+      "epoch": 0.5788257940327238,
+      "grad_norm": 1.8660439252853394,
+      "learning_rate": 0.00016148191613040734,
+      "loss": 1.0786,
+      "step": 15035
+    },
+    {
+      "epoch": 0.5790182868142445,
+      "grad_norm": 2.343719720840454,
+      "learning_rate": 0.0001614580643066931,
+      "loss": 1.276,
+      "step": 15040
+    },
+    {
+      "epoch": 0.5792107795957652,
+      "grad_norm": 1.1358321905136108,
+      "learning_rate": 0.00016143420686304594,
+      "loss": 1.3055,
+      "step": 15045
+    },
+    {
+      "epoch": 0.5794032723772858,
+      "grad_norm": 1.6678638458251953,
+      "learning_rate": 0.00016141034380164754,
+      "loss": 1.0694,
+      "step": 15050
+    },
+    {
+      "epoch": 0.5795957651588065,
+      "grad_norm": 1.6096512079238892,
+      "learning_rate": 0.00016138647512468004,
+      "loss": 1.4079,
+      "step": 15055
+    },
+    {
+      "epoch": 0.5797882579403273,
+      "grad_norm": 2.3922042846679688,
+      "learning_rate": 0.000161362600834326,
+      "loss": 1.2739,
+      "step": 15060
+    },
+    {
+      "epoch": 0.579980750721848,
+      "grad_norm": 1.167476773262024,
+      "learning_rate": 0.0001613387209327686,
+      "loss": 1.2336,
+      "step": 15065
+    },
+    {
+      "epoch": 0.5801732435033686,
+      "grad_norm": 0.9550272226333618,
+      "learning_rate": 0.00016131483542219152,
+      "loss": 1.1557,
+      "step": 15070
+    },
+    {
+      "epoch": 0.5803657362848893,
+      "grad_norm": 1.1105631589889526,
+      "learning_rate": 0.00016129094430477893,
+      "loss": 1.0289,
+      "step": 15075
+    },
+    {
+      "epoch": 0.58055822906641,
+      "grad_norm": 1.3411318063735962,
+      "learning_rate": 0.00016126704758271548,
+      "loss": 1.2454,
+      "step": 15080
+    },
+    {
+      "epoch": 0.5807507218479308,
+      "grad_norm": 1.2867335081100464,
+      "learning_rate": 0.00016124314525818635,
+      "loss": 1.2983,
+      "step": 15085
+    },
+    {
+      "epoch": 0.5809432146294514,
+      "grad_norm": 1.6035441160202026,
+      "learning_rate": 0.00016121923733337736,
+      "loss": 1.2227,
+      "step": 15090
+    },
+    {
+      "epoch": 0.5811357074109721,
+      "grad_norm": 1.6657713651657104,
+      "learning_rate": 0.0001611953238104746,
+      "loss": 1.1144,
+      "step": 15095
+    },
+    {
+      "epoch": 0.5813282001924928,
+      "grad_norm": 1.8781518936157227,
+      "learning_rate": 0.00016117140469166486,
+      "loss": 1.1393,
+      "step": 15100
+    },
+    {
+      "epoch": 0.5815206929740134,
+      "grad_norm": 1.542438268661499,
+      "learning_rate": 0.00016114747997913542,
+      "loss": 1.188,
+      "step": 15105
+    },
+    {
+      "epoch": 0.5817131857555342,
+      "grad_norm": 2.148175001144409,
+      "learning_rate": 0.00016112354967507398,
+      "loss": 1.1323,
+      "step": 15110
+    },
+    {
+      "epoch": 0.5819056785370549,
+      "grad_norm": 1.3092713356018066,
+      "learning_rate": 0.0001610996137816688,
+      "loss": 1.1799,
+      "step": 15115
+    },
+    {
+      "epoch": 0.5820981713185756,
+      "grad_norm": 1.4203580617904663,
+      "learning_rate": 0.00016107567230110874,
+      "loss": 1.0916,
+      "step": 15120
+    },
+    {
+      "epoch": 0.5822906641000962,
+      "grad_norm": 1.2932054996490479,
+      "learning_rate": 0.00016105172523558301,
+      "loss": 1.13,
+      "step": 15125
+    },
+    {
+      "epoch": 0.5824831568816169,
+      "grad_norm": 2.218705654144287,
+      "learning_rate": 0.00016102777258728142,
+      "loss": 1.153,
+      "step": 15130
+    },
+    {
+      "epoch": 0.5826756496631377,
+      "grad_norm": 1.180166244506836,
+      "learning_rate": 0.00016100381435839433,
+      "loss": 1.2611,
+      "step": 15135
+    },
+    {
+      "epoch": 0.5828681424446583,
+      "grad_norm": 2.007887125015259,
+      "learning_rate": 0.00016097985055111256,
+      "loss": 1.2046,
+      "step": 15140
+    },
+    {
+      "epoch": 0.583060635226179,
+      "grad_norm": 1.20327889919281,
+      "learning_rate": 0.00016095588116762734,
+      "loss": 1.3217,
+      "step": 15145
+    },
+    {
+      "epoch": 0.5832531280076997,
+      "grad_norm": 1.2758903503417969,
+      "learning_rate": 0.00016093190621013063,
+      "loss": 1.1277,
+      "step": 15150
+    },
+    {
+      "epoch": 0.5834456207892204,
+      "grad_norm": 1.851881980895996,
+      "learning_rate": 0.00016090792568081473,
+      "loss": 1.1701,
+      "step": 15155
+    },
+    {
+      "epoch": 0.5836381135707411,
+      "grad_norm": 1.6895406246185303,
+      "learning_rate": 0.00016088393958187247,
+      "loss": 1.3331,
+      "step": 15160
+    },
+    {
+      "epoch": 0.5838306063522618,
+      "grad_norm": 1.4138762950897217,
+      "learning_rate": 0.0001608599479154973,
+      "loss": 1.3016,
+      "step": 15165
+    },
+    {
+      "epoch": 0.5840230991337825,
+      "grad_norm": 1.3571628332138062,
+      "learning_rate": 0.00016083595068388303,
+      "loss": 1.1407,
+      "step": 15170
+    },
+    {
+      "epoch": 0.5842155919153031,
+      "grad_norm": 1.3217098712921143,
+      "learning_rate": 0.00016081194788922405,
+      "loss": 1.0032,
+      "step": 15175
+    },
+    {
+      "epoch": 0.5844080846968238,
+      "grad_norm": 0.9765079617500305,
+      "learning_rate": 0.00016078793953371533,
+      "loss": 1.1543,
+      "step": 15180
+    },
+    {
+      "epoch": 0.5846005774783446,
+      "grad_norm": 1.0757596492767334,
+      "learning_rate": 0.0001607639256195522,
+      "loss": 1.0828,
+      "step": 15185
+    },
+    {
+      "epoch": 0.5847930702598653,
+      "grad_norm": 1.2296372652053833,
+      "learning_rate": 0.00016073990614893057,
+      "loss": 1.2089,
+      "step": 15190
+    },
+    {
+      "epoch": 0.5849855630413859,
+      "grad_norm": 1.8743308782577515,
+      "learning_rate": 0.00016071588112404693,
+      "loss": 1.2195,
+      "step": 15195
+    },
+    {
+      "epoch": 0.5851780558229066,
+      "grad_norm": 1.347332239151001,
+      "learning_rate": 0.00016069185054709814,
+      "loss": 1.1664,
+      "step": 15200
+    },
+    {
+      "epoch": 0.5853705486044274,
+      "grad_norm": 1.629981279373169,
+      "learning_rate": 0.00016066781442028165,
+      "loss": 1.1888,
+      "step": 15205
+    },
+    {
+      "epoch": 0.5855630413859481,
+      "grad_norm": 1.353702425956726,
+      "learning_rate": 0.00016064377274579544,
+      "loss": 1.265,
+      "step": 15210
+    },
+    {
+      "epoch": 0.5857555341674687,
+      "grad_norm": 0.9861169457435608,
+      "learning_rate": 0.00016061972552583795,
+      "loss": 1.0908,
+      "step": 15215
+    },
+    {
+      "epoch": 0.5859480269489894,
+      "grad_norm": 1.1305365562438965,
+      "learning_rate": 0.00016059567276260813,
+      "loss": 1.0076,
+      "step": 15220
+    },
+    {
+      "epoch": 0.5861405197305101,
+      "grad_norm": 1.4098013639450073,
+      "learning_rate": 0.00016057161445830542,
+      "loss": 1.1882,
+      "step": 15225
+    },
+    {
+      "epoch": 0.5863330125120308,
+      "grad_norm": 1.1900111436843872,
+      "learning_rate": 0.00016054755061512986,
+      "loss": 1.1961,
+      "step": 15230
+    },
+    {
+      "epoch": 0.5865255052935515,
+      "grad_norm": 1.0856738090515137,
+      "learning_rate": 0.00016052348123528183,
+      "loss": 1.2169,
+      "step": 15235
+    },
+    {
+      "epoch": 0.5867179980750722,
+      "grad_norm": 1.109937071800232,
+      "learning_rate": 0.0001604994063209624,
+      "loss": 1.0818,
+      "step": 15240
+    },
+    {
+      "epoch": 0.5869104908565929,
+      "grad_norm": 1.9059746265411377,
+      "learning_rate": 0.00016047532587437304,
+      "loss": 1.3035,
+      "step": 15245
+    },
+    {
+      "epoch": 0.5871029836381135,
+      "grad_norm": 1.089796781539917,
+      "learning_rate": 0.00016045123989771575,
+      "loss": 1.0872,
+      "step": 15250
+    },
+    {
+      "epoch": 0.5872954764196343,
+      "grad_norm": 1.3014196157455444,
+      "learning_rate": 0.00016042714839319298,
+      "loss": 1.1809,
+      "step": 15255
+    },
+    {
+      "epoch": 0.587487969201155,
+      "grad_norm": 1.5097154378890991,
+      "learning_rate": 0.00016040305136300783,
+      "loss": 1.1026,
+      "step": 15260
+    },
+    {
+      "epoch": 0.5876804619826757,
+      "grad_norm": 1.9508148431777954,
+      "learning_rate": 0.00016037894880936376,
+      "loss": 1.0489,
+      "step": 15265
+    },
+    {
+      "epoch": 0.5878729547641963,
+      "grad_norm": 1.2007025480270386,
+      "learning_rate": 0.0001603548407344648,
+      "loss": 1.2376,
+      "step": 15270
+    },
+    {
+      "epoch": 0.588065447545717,
+      "grad_norm": 4.035842418670654,
+      "learning_rate": 0.00016033072714051545,
+      "loss": 1.2894,
+      "step": 15275
+    },
+    {
+      "epoch": 0.5882579403272378,
+      "grad_norm": 1.2279680967330933,
+      "learning_rate": 0.00016030660802972074,
+      "loss": 1.1945,
+      "step": 15280
+    },
+    {
+      "epoch": 0.5884504331087584,
+      "grad_norm": 1.0882714986801147,
+      "learning_rate": 0.00016028248340428625,
+      "loss": 1.0842,
+      "step": 15285
+    },
+    {
+      "epoch": 0.5886429258902791,
+      "grad_norm": 1.3169769048690796,
+      "learning_rate": 0.00016025835326641797,
+      "loss": 1.0085,
+      "step": 15290
+    },
+    {
+      "epoch": 0.5888354186717998,
+      "grad_norm": 1.3032643795013428,
+      "learning_rate": 0.00016023421761832246,
+      "loss": 1.1994,
+      "step": 15295
+    },
+    {
+      "epoch": 0.5890279114533205,
+      "grad_norm": 1.053415060043335,
+      "learning_rate": 0.00016021007646220678,
+      "loss": 1.0983,
+      "step": 15300
+    },
+    {
+      "epoch": 0.5892204042348412,
+      "grad_norm": 1.483736515045166,
+      "learning_rate": 0.00016018592980027846,
+      "loss": 1.1709,
+      "step": 15305
+    },
+    {
+      "epoch": 0.5894128970163619,
+      "grad_norm": 1.4688469171524048,
+      "learning_rate": 0.00016016177763474555,
+      "loss": 1.0505,
+      "step": 15310
+    },
+    {
+      "epoch": 0.5896053897978826,
+      "grad_norm": 1.7809165716171265,
+      "learning_rate": 0.00016013761996781661,
+      "loss": 1.1585,
+      "step": 15315
+    },
+    {
+      "epoch": 0.5897978825794032,
+      "grad_norm": 1.5344901084899902,
+      "learning_rate": 0.00016011345680170072,
+      "loss": 1.1269,
+      "step": 15320
+    },
+    {
+      "epoch": 0.5899903753609239,
+      "grad_norm": 1.298094630241394,
+      "learning_rate": 0.0001600892881386074,
+      "loss": 1.1804,
+      "step": 15325
+    },
+    {
+      "epoch": 0.5901828681424447,
+      "grad_norm": 1.8283668756484985,
+      "learning_rate": 0.0001600651139807467,
+      "loss": 1.2059,
+      "step": 15330
+    },
+    {
+      "epoch": 0.5903753609239654,
+      "grad_norm": 1.3290801048278809,
+      "learning_rate": 0.00016004093433032924,
+      "loss": 1.2334,
+      "step": 15335
+    },
+    {
+      "epoch": 0.590567853705486,
+      "grad_norm": 1.461422324180603,
+      "learning_rate": 0.00016001674918956612,
+      "loss": 1.2987,
+      "step": 15340
+    },
+    {
+      "epoch": 0.5907603464870067,
+      "grad_norm": 1.6681803464889526,
+      "learning_rate": 0.00015999255856066885,
+      "loss": 1.0221,
+      "step": 15345
+    },
+    {
+      "epoch": 0.5909528392685275,
+      "grad_norm": 1.1714918613433838,
+      "learning_rate": 0.00015996836244584948,
+      "loss": 1.0144,
+      "step": 15350
+    },
+    {
+      "epoch": 0.5911453320500482,
+      "grad_norm": 0.9316911697387695,
+      "learning_rate": 0.00015994416084732062,
+      "loss": 1.3241,
+      "step": 15355
+    },
+    {
+      "epoch": 0.5913378248315688,
+      "grad_norm": 2.429568290710449,
+      "learning_rate": 0.00015991995376729535,
+      "loss": 1.3155,
+      "step": 15360
+    },
+    {
+      "epoch": 0.5915303176130895,
+      "grad_norm": 1.3793234825134277,
+      "learning_rate": 0.00015989574120798725,
+      "loss": 1.2822,
+      "step": 15365
+    },
+    {
+      "epoch": 0.5917228103946102,
+      "grad_norm": 1.1756724119186401,
+      "learning_rate": 0.0001598715231716104,
+      "loss": 1.0682,
+      "step": 15370
+    },
+    {
+      "epoch": 0.591915303176131,
+      "grad_norm": 1.9872701168060303,
+      "learning_rate": 0.00015984729966037934,
+      "loss": 1.2034,
+      "step": 15375
+    },
+    {
+      "epoch": 0.5921077959576516,
+      "grad_norm": 1.5333032608032227,
+      "learning_rate": 0.00015982307067650918,
+      "loss": 1.3922,
+      "step": 15380
+    },
+    {
+      "epoch": 0.5923002887391723,
+      "grad_norm": 1.1813582181930542,
+      "learning_rate": 0.00015979883622221555,
+      "loss": 1.1811,
+      "step": 15385
+    },
+    {
+      "epoch": 0.592492781520693,
+      "grad_norm": 1.632565975189209,
+      "learning_rate": 0.00015977459629971442,
+      "loss": 1.0877,
+      "step": 15390
+    },
+    {
+      "epoch": 0.5926852743022136,
+      "grad_norm": 1.0945332050323486,
+      "learning_rate": 0.00015975035091122245,
+      "loss": 1.0836,
+      "step": 15395
+    },
+    {
+      "epoch": 0.5928777670837344,
+      "grad_norm": 0.8069517016410828,
+      "learning_rate": 0.0001597261000589567,
+      "loss": 1.1574,
+      "step": 15400
+    },
+    {
+      "epoch": 0.5930702598652551,
+      "grad_norm": 1.8364413976669312,
+      "learning_rate": 0.00015970184374513476,
+      "loss": 1.1935,
+      "step": 15405
+    },
+    {
+      "epoch": 0.5932627526467757,
+      "grad_norm": 1.5146484375,
+      "learning_rate": 0.00015967758197197468,
+      "loss": 1.06,
+      "step": 15410
+    },
+    {
+      "epoch": 0.5934552454282964,
+      "grad_norm": 1.5792328119277954,
+      "learning_rate": 0.00015965331474169508,
+      "loss": 1.1464,
+      "step": 15415
+    },
+    {
+      "epoch": 0.5936477382098171,
+      "grad_norm": 1.887292742729187,
+      "learning_rate": 0.00015962904205651495,
+      "loss": 1.2039,
+      "step": 15420
+    },
+    {
+      "epoch": 0.5938402309913379,
+      "grad_norm": 1.8241037130355835,
+      "learning_rate": 0.000159604763918654,
+      "loss": 1.2011,
+      "step": 15425
+    },
+    {
+      "epoch": 0.5940327237728585,
+      "grad_norm": 1.2130569219589233,
+      "learning_rate": 0.0001595804803303322,
+      "loss": 1.2401,
+      "step": 15430
+    },
+    {
+      "epoch": 0.5942252165543792,
+      "grad_norm": 1.1083897352218628,
+      "learning_rate": 0.00015955619129377017,
+      "loss": 1.2919,
+      "step": 15435
+    },
+    {
+      "epoch": 0.5944177093358999,
+      "grad_norm": 1.8266736268997192,
+      "learning_rate": 0.00015953189681118895,
+      "loss": 1.1609,
+      "step": 15440
+    },
+    {
+      "epoch": 0.5946102021174205,
+      "grad_norm": 1.5710999965667725,
+      "learning_rate": 0.0001595075968848102,
+      "loss": 1.2178,
+      "step": 15445
+    },
+    {
+      "epoch": 0.5948026948989413,
+      "grad_norm": 2.023061752319336,
+      "learning_rate": 0.00015948329151685583,
+      "loss": 1.2577,
+      "step": 15450
+    },
+    {
+      "epoch": 0.594995187680462,
+      "grad_norm": 1.3245149850845337,
+      "learning_rate": 0.00015945898070954853,
+      "loss": 0.9832,
+      "step": 15455
+    },
+    {
+      "epoch": 0.5951876804619827,
+      "grad_norm": 1.7696577310562134,
+      "learning_rate": 0.00015943466446511132,
+      "loss": 1.1991,
+      "step": 15460
+    },
+    {
+      "epoch": 0.5953801732435033,
+      "grad_norm": 1.0893733501434326,
+      "learning_rate": 0.00015941034278576775,
+      "loss": 1.3321,
+      "step": 15465
+    },
+    {
+      "epoch": 0.595572666025024,
+      "grad_norm": 1.294731616973877,
+      "learning_rate": 0.0001593860156737419,
+      "loss": 1.0485,
+      "step": 15470
+    },
+    {
+      "epoch": 0.5957651588065448,
+      "grad_norm": 1.1282588243484497,
+      "learning_rate": 0.00015936168313125833,
+      "loss": 1.0585,
+      "step": 15475
+    },
+    {
+      "epoch": 0.5959576515880655,
+      "grad_norm": 0.9207860231399536,
+      "learning_rate": 0.00015933734516054203,
+      "loss": 1.1343,
+      "step": 15480
+    },
+    {
+      "epoch": 0.5961501443695861,
+      "grad_norm": 2.2860140800476074,
+      "learning_rate": 0.00015931300176381865,
+      "loss": 1.3317,
+      "step": 15485
+    },
+    {
+      "epoch": 0.5963426371511068,
+      "grad_norm": 1.2698768377304077,
+      "learning_rate": 0.00015928865294331413,
+      "loss": 1.175,
+      "step": 15490
+    },
+    {
+      "epoch": 0.5965351299326275,
+      "grad_norm": 1.0986465215682983,
+      "learning_rate": 0.00015926429870125505,
+      "loss": 1.1309,
+      "step": 15495
+    },
+    {
+      "epoch": 0.5967276227141483,
+      "grad_norm": 1.5664902925491333,
+      "learning_rate": 0.00015923993903986844,
+      "loss": 1.0117,
+      "step": 15500
+    },
+    {
+      "epoch": 0.5969201154956689,
+      "grad_norm": 1.3162322044372559,
+      "learning_rate": 0.00015921557396138188,
+      "loss": 1.1964,
+      "step": 15505
+    },
+    {
+      "epoch": 0.5971126082771896,
+      "grad_norm": 0.8635309934616089,
+      "learning_rate": 0.0001591912034680233,
+      "loss": 1.1119,
+      "step": 15510
+    },
+    {
+      "epoch": 0.5973051010587103,
+      "grad_norm": 1.3118690252304077,
+      "learning_rate": 0.00015916682756202127,
+      "loss": 1.0618,
+      "step": 15515
+    },
+    {
+      "epoch": 0.597497593840231,
+      "grad_norm": 1.0313913822174072,
+      "learning_rate": 0.00015914244624560481,
+      "loss": 1.0686,
+      "step": 15520
+    },
+    {
+      "epoch": 0.5976900866217517,
+      "grad_norm": 1.3414394855499268,
+      "learning_rate": 0.00015911805952100347,
+      "loss": 1.2013,
+      "step": 15525
+    },
+    {
+      "epoch": 0.5978825794032724,
+      "grad_norm": 1.2710504531860352,
+      "learning_rate": 0.00015909366739044715,
+      "loss": 1.3748,
+      "step": 15530
+    },
+    {
+      "epoch": 0.598075072184793,
+      "grad_norm": 1.6694974899291992,
+      "learning_rate": 0.0001590692698561664,
+      "loss": 0.9833,
+      "step": 15535
+    },
+    {
+      "epoch": 0.5982675649663137,
+      "grad_norm": 1.5924476385116577,
+      "learning_rate": 0.00015904486692039227,
+      "loss": 1.2046,
+      "step": 15540
+    },
+    {
+      "epoch": 0.5984600577478345,
+      "grad_norm": 2.3105616569519043,
+      "learning_rate": 0.00015902045858535616,
+      "loss": 1.233,
+      "step": 15545
+    },
+    {
+      "epoch": 0.5986525505293552,
+      "grad_norm": 1.3003478050231934,
+      "learning_rate": 0.00015899604485329012,
+      "loss": 1.1891,
+      "step": 15550
+    },
+    {
+      "epoch": 0.5988450433108758,
+      "grad_norm": 1.2988343238830566,
+      "learning_rate": 0.00015897162572642656,
+      "loss": 1.0767,
+      "step": 15555
+    },
+    {
+      "epoch": 0.5990375360923965,
+      "grad_norm": 1.0845260620117188,
+      "learning_rate": 0.00015894720120699849,
+      "loss": 1.2702,
+      "step": 15560
+    },
+    {
+      "epoch": 0.5992300288739172,
+      "grad_norm": 1.0050013065338135,
+      "learning_rate": 0.00015892277129723935,
+      "loss": 1.2267,
+      "step": 15565
+    },
+    {
+      "epoch": 0.599422521655438,
+      "grad_norm": 1.3145102262496948,
+      "learning_rate": 0.0001588983359993831,
+      "loss": 1.1086,
+      "step": 15570
+    },
+    {
+      "epoch": 0.5996150144369586,
+      "grad_norm": 1.817396640777588,
+      "learning_rate": 0.00015887389531566424,
+      "loss": 1.0999,
+      "step": 15575
+    },
+    {
+      "epoch": 0.5998075072184793,
+      "grad_norm": 1.4001067876815796,
+      "learning_rate": 0.0001588494492483176,
+      "loss": 1.2802,
+      "step": 15580
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1305971145629883,
+      "learning_rate": 0.00015882499779957868,
+      "loss": 1.2481,
+      "step": 15585
+    },
+    {
+      "epoch": 0.6001924927815206,
+      "grad_norm": 1.5675426721572876,
+      "learning_rate": 0.00015880054097168337,
+      "loss": 1.2555,
+      "step": 15590
+    },
+    {
+      "epoch": 0.6003849855630414,
+      "grad_norm": 1.3107160329818726,
+      "learning_rate": 0.00015877607876686815,
+      "loss": 1.273,
+      "step": 15595
+    },
+    {
+      "epoch": 0.6005774783445621,
+      "grad_norm": 0.5348256230354309,
+      "learning_rate": 0.00015875161118736986,
+      "loss": 0.9708,
+      "step": 15600
+    },
+    {
+      "epoch": 0.6007699711260828,
+      "grad_norm": 1.0877107381820679,
+      "learning_rate": 0.00015872713823542593,
+      "loss": 1.1419,
+      "step": 15605
+    },
+    {
+      "epoch": 0.6009624639076034,
+      "grad_norm": 1.0563950538635254,
+      "learning_rate": 0.00015870265991327424,
+      "loss": 1.0216,
+      "step": 15610
+    },
+    {
+      "epoch": 0.6011549566891241,
+      "grad_norm": 1.0346797704696655,
+      "learning_rate": 0.00015867817622315316,
+      "loss": 1.205,
+      "step": 15615
+    },
+    {
+      "epoch": 0.6013474494706449,
+      "grad_norm": 1.67006254196167,
+      "learning_rate": 0.00015865368716730158,
+      "loss": 1.2875,
+      "step": 15620
+    },
+    {
+      "epoch": 0.6015399422521656,
+      "grad_norm": 1.8183788061141968,
+      "learning_rate": 0.00015862919274795884,
+      "loss": 1.1703,
+      "step": 15625
+    },
+    {
+      "epoch": 0.6017324350336862,
+      "grad_norm": 1.1460903882980347,
+      "learning_rate": 0.00015860469296736482,
+      "loss": 1.1998,
+      "step": 15630
+    },
+    {
+      "epoch": 0.6019249278152069,
+      "grad_norm": 1.5365129709243774,
+      "learning_rate": 0.00015858018782775985,
+      "loss": 1.054,
+      "step": 15635
+    },
+    {
+      "epoch": 0.6021174205967276,
+      "grad_norm": 1.4886486530303955,
+      "learning_rate": 0.00015855567733138478,
+      "loss": 1.2914,
+      "step": 15640
+    },
+    {
+      "epoch": 0.6023099133782484,
+      "grad_norm": 1.519114375114441,
+      "learning_rate": 0.00015853116148048087,
+      "loss": 1.0586,
+      "step": 15645
+    },
+    {
+      "epoch": 0.602502406159769,
+      "grad_norm": 1.2735627889633179,
+      "learning_rate": 0.00015850664027729,
+      "loss": 1.1287,
+      "step": 15650
+    },
+    {
+      "epoch": 0.6026948989412897,
+      "grad_norm": 2.464672327041626,
+      "learning_rate": 0.00015848211372405444,
+      "loss": 1.0616,
+      "step": 15655
+    },
+    {
+      "epoch": 0.6028873917228104,
+      "grad_norm": 0.9507278800010681,
+      "learning_rate": 0.000158457581823017,
+      "loss": 1.0118,
+      "step": 15660
+    },
+    {
+      "epoch": 0.6030798845043311,
+      "grad_norm": 1.155150294303894,
+      "learning_rate": 0.00015843304457642093,
+      "loss": 1.0563,
+      "step": 15665
+    },
+    {
+      "epoch": 0.6032723772858518,
+      "grad_norm": 2.669029474258423,
+      "learning_rate": 0.00015840850198651002,
+      "loss": 1.1918,
+      "step": 15670
+    },
+    {
+      "epoch": 0.6034648700673725,
+      "grad_norm": 1.4008570909500122,
+      "learning_rate": 0.00015838395405552854,
+      "loss": 1.2122,
+      "step": 15675
+    },
+    {
+      "epoch": 0.6036573628488932,
+      "grad_norm": 1.4199731349945068,
+      "learning_rate": 0.0001583594007857212,
+      "loss": 1.2546,
+      "step": 15680
+    },
+    {
+      "epoch": 0.6038498556304138,
+      "grad_norm": 2.2346031665802,
+      "learning_rate": 0.0001583348421793333,
+      "loss": 1.1184,
+      "step": 15685
+    },
+    {
+      "epoch": 0.6040423484119346,
+      "grad_norm": 1.1559759378433228,
+      "learning_rate": 0.00015831027823861048,
+      "loss": 1.157,
+      "step": 15690
+    },
+    {
+      "epoch": 0.6042348411934553,
+      "grad_norm": 1.9930438995361328,
+      "learning_rate": 0.00015828570896579897,
+      "loss": 1.1095,
+      "step": 15695
+    },
+    {
+      "epoch": 0.6044273339749759,
+      "grad_norm": 1.040358304977417,
+      "learning_rate": 0.00015826113436314548,
+      "loss": 1.062,
+      "step": 15700
+    },
+    {
+      "epoch": 0.6046198267564966,
+      "grad_norm": 0.8409137725830078,
+      "learning_rate": 0.00015823655443289724,
+      "loss": 1.0204,
+      "step": 15705
+    },
+    {
+      "epoch": 0.6048123195380173,
+      "grad_norm": 1.477950930595398,
+      "learning_rate": 0.00015821196917730184,
+      "loss": 1.2479,
+      "step": 15710
+    },
+    {
+      "epoch": 0.6050048123195381,
+      "grad_norm": 1.5752694606781006,
+      "learning_rate": 0.00015818737859860752,
+      "loss": 1.343,
+      "step": 15715
+    },
+    {
+      "epoch": 0.6051973051010587,
+      "grad_norm": 1.505356788635254,
+      "learning_rate": 0.00015816278269906284,
+      "loss": 1.0742,
+      "step": 15720
+    },
+    {
+      "epoch": 0.6053897978825794,
+      "grad_norm": 1.165273904800415,
+      "learning_rate": 0.000158138181480917,
+      "loss": 1.1023,
+      "step": 15725
+    },
+    {
+      "epoch": 0.6055822906641001,
+      "grad_norm": 1.7088487148284912,
+      "learning_rate": 0.00015811357494641958,
+      "loss": 1.2899,
+      "step": 15730
+    },
+    {
+      "epoch": 0.6057747834456207,
+      "grad_norm": 1.6200921535491943,
+      "learning_rate": 0.0001580889630978207,
+      "loss": 1.0353,
+      "step": 15735
+    },
+    {
+      "epoch": 0.6059672762271415,
+      "grad_norm": 1.1059575080871582,
+      "learning_rate": 0.00015806434593737095,
+      "loss": 0.8117,
+      "step": 15740
+    },
+    {
+      "epoch": 0.6061597690086622,
+      "grad_norm": 1.3026262521743774,
+      "learning_rate": 0.00015803972346732143,
+      "loss": 1.1648,
+      "step": 15745
+    },
+    {
+      "epoch": 0.6063522617901829,
+      "grad_norm": 1.316931128501892,
+      "learning_rate": 0.00015801509568992366,
+      "loss": 1.0999,
+      "step": 15750
+    },
+    {
+      "epoch": 0.6065447545717035,
+      "grad_norm": 0.9396672248840332,
+      "learning_rate": 0.00015799046260742968,
+      "loss": 1.0374,
+      "step": 15755
+    },
+    {
+      "epoch": 0.6067372473532242,
+      "grad_norm": 1.1851413249969482,
+      "learning_rate": 0.00015796582422209206,
+      "loss": 1.1861,
+      "step": 15760
+    },
+    {
+      "epoch": 0.606929740134745,
+      "grad_norm": 2.0202128887176514,
+      "learning_rate": 0.00015794118053616383,
+      "loss": 1.139,
+      "step": 15765
+    },
+    {
+      "epoch": 0.6071222329162657,
+      "grad_norm": 1.642561912536621,
+      "learning_rate": 0.00015791653155189841,
+      "loss": 1.1811,
+      "step": 15770
+    },
+    {
+      "epoch": 0.6073147256977863,
+      "grad_norm": 1.4148608446121216,
+      "learning_rate": 0.0001578918772715499,
+      "loss": 1.3261,
+      "step": 15775
+    },
+    {
+      "epoch": 0.607507218479307,
+      "grad_norm": 1.160662293434143,
+      "learning_rate": 0.0001578672176973727,
+      "loss": 1.2117,
+      "step": 15780
+    },
+    {
+      "epoch": 0.6076997112608277,
+      "grad_norm": 1.4699779748916626,
+      "learning_rate": 0.00015784255283162176,
+      "loss": 1.1937,
+      "step": 15785
+    },
+    {
+      "epoch": 0.6078922040423484,
+      "grad_norm": 1.579142451286316,
+      "learning_rate": 0.00015781788267655252,
+      "loss": 1.2722,
+      "step": 15790
+    },
+    {
+      "epoch": 0.6080846968238691,
+      "grad_norm": 1.3598978519439697,
+      "learning_rate": 0.00015779320723442096,
+      "loss": 1.0829,
+      "step": 15795
+    },
+    {
+      "epoch": 0.6082771896053898,
+      "grad_norm": 1.1840283870697021,
+      "learning_rate": 0.0001577734630755471,
+      "loss": 1.3599,
+      "step": 15800
+    },
+    {
+      "epoch": 0.6084696823869105,
+      "grad_norm": 1.1004847288131714,
+      "learning_rate": 0.00015774877812238972,
+      "loss": 1.3756,
+      "step": 15805
+    },
+    {
+      "epoch": 0.6086621751684311,
+      "grad_norm": 1.6455458402633667,
+      "learning_rate": 0.00015772408788848914,
+      "loss": 1.1067,
+      "step": 15810
+    },
+    {
+      "epoch": 0.6088546679499519,
+      "grad_norm": 1.6274205446243286,
+      "learning_rate": 0.00015769939237610312,
+      "loss": 1.3339,
+      "step": 15815
+    },
+    {
+      "epoch": 0.6090471607314726,
+      "grad_norm": 1.2150076627731323,
+      "learning_rate": 0.00015767469158748987,
+      "loss": 0.9821,
+      "step": 15820
+    },
+    {
+      "epoch": 0.6092396535129933,
+      "grad_norm": 1.2452518939971924,
+      "learning_rate": 0.00015764998552490815,
+      "loss": 0.9994,
+      "step": 15825
+    },
+    {
+      "epoch": 0.6094321462945139,
+      "grad_norm": 1.4766079187393188,
+      "learning_rate": 0.00015762527419061715,
+      "loss": 1.0281,
+      "step": 15830
+    },
+    {
+      "epoch": 0.6096246390760347,
+      "grad_norm": 1.1288725137710571,
+      "learning_rate": 0.00015760055758687655,
+      "loss": 1.082,
+      "step": 15835
+    },
+    {
+      "epoch": 0.6098171318575554,
+      "grad_norm": 1.181159496307373,
+      "learning_rate": 0.00015757583571594653,
+      "loss": 1.1515,
+      "step": 15840
+    },
+    {
+      "epoch": 0.610009624639076,
+      "grad_norm": 1.2939519882202148,
+      "learning_rate": 0.00015755110858008773,
+      "loss": 0.9892,
+      "step": 15845
+    },
+    {
+      "epoch": 0.6102021174205967,
+      "grad_norm": 2.3088269233703613,
+      "learning_rate": 0.0001575263761815613,
+      "loss": 0.9798,
+      "step": 15850
+    },
+    {
+      "epoch": 0.6103946102021174,
+      "grad_norm": 1.4175939559936523,
+      "learning_rate": 0.00015750163852262886,
+      "loss": 1.1408,
+      "step": 15855
+    },
+    {
+      "epoch": 0.6105871029836382,
+      "grad_norm": 1.0206336975097656,
+      "learning_rate": 0.00015747689560555248,
+      "loss": 1.2078,
+      "step": 15860
+    },
+    {
+      "epoch": 0.6107795957651588,
+      "grad_norm": 0.9995696544647217,
+      "learning_rate": 0.0001574521474325948,
+      "loss": 1.1922,
+      "step": 15865
+    },
+    {
+      "epoch": 0.6109720885466795,
+      "grad_norm": 1.6652652025222778,
+      "learning_rate": 0.00015742739400601872,
+      "loss": 1.1039,
+      "step": 15870
+    },
+    {
+      "epoch": 0.6111645813282002,
+      "grad_norm": 1.3411548137664795,
+      "learning_rate": 0.00015740263532808792,
+      "loss": 1.1592,
+      "step": 15875
+    },
+    {
+      "epoch": 0.6113570741097208,
+      "grad_norm": 0.9215561151504517,
+      "learning_rate": 0.0001573778714010664,
+      "loss": 1.1379,
+      "step": 15880
+    },
+    {
+      "epoch": 0.6115495668912416,
+      "grad_norm": 1.269482970237732,
+      "learning_rate": 0.00015735310222721863,
+      "loss": 1.2042,
+      "step": 15885
+    },
+    {
+      "epoch": 0.6117420596727623,
+      "grad_norm": 1.316909909248352,
+      "learning_rate": 0.00015732832780880957,
+      "loss": 1.2702,
+      "step": 15890
+    },
+    {
+      "epoch": 0.611934552454283,
+      "grad_norm": 1.2689425945281982,
+      "learning_rate": 0.0001573035481481047,
+      "loss": 1.0755,
+      "step": 15895
+    },
+    {
+      "epoch": 0.6121270452358036,
+      "grad_norm": 1.0369685888290405,
+      "learning_rate": 0.00015727876324736996,
+      "loss": 1.0574,
+      "step": 15900
+    },
+    {
+      "epoch": 0.6123195380173243,
+      "grad_norm": 1.0056127309799194,
+      "learning_rate": 0.00015725397310887174,
+      "loss": 1.2219,
+      "step": 15905
+    },
+    {
+      "epoch": 0.6125120307988451,
+      "grad_norm": 1.3123587369918823,
+      "learning_rate": 0.00015722917773487702,
+      "loss": 1.2203,
+      "step": 15910
+    },
+    {
+      "epoch": 0.6127045235803658,
+      "grad_norm": 1.0959875583648682,
+      "learning_rate": 0.00015720437712765306,
+      "loss": 1.2516,
+      "step": 15915
+    },
+    {
+      "epoch": 0.6128970163618864,
+      "grad_norm": 2.0152196884155273,
+      "learning_rate": 0.00015717957128946774,
+      "loss": 1.2099,
+      "step": 15920
+    },
+    {
+      "epoch": 0.6130895091434071,
+      "grad_norm": 2.816568374633789,
+      "learning_rate": 0.00015715476022258942,
+      "loss": 1.1093,
+      "step": 15925
+    },
+    {
+      "epoch": 0.6132820019249278,
+      "grad_norm": 1.8223321437835693,
+      "learning_rate": 0.00015712994392928689,
+      "loss": 1.0474,
+      "step": 15930
+    },
+    {
+      "epoch": 0.6134744947064485,
+      "grad_norm": 1.2718263864517212,
+      "learning_rate": 0.00015710512241182945,
+      "loss": 1.1405,
+      "step": 15935
+    },
+    {
+      "epoch": 0.6136669874879692,
+      "grad_norm": 1.2518097162246704,
+      "learning_rate": 0.00015708029567248683,
+      "loss": 1.13,
+      "step": 15940
+    },
+    {
+      "epoch": 0.6138594802694899,
+      "grad_norm": 0.8542113900184631,
+      "learning_rate": 0.0001570554637135293,
+      "loss": 1.0871,
+      "step": 15945
+    },
+    {
+      "epoch": 0.6140519730510106,
+      "grad_norm": 1.0798470973968506,
+      "learning_rate": 0.00015703062653722757,
+      "loss": 1.1563,
+      "step": 15950
+    },
+    {
+      "epoch": 0.6142444658325312,
+      "grad_norm": 1.123974084854126,
+      "learning_rate": 0.00015700578414585284,
+      "loss": 1.1253,
+      "step": 15955
+    },
+    {
+      "epoch": 0.614436958614052,
+      "grad_norm": 1.2129628658294678,
+      "learning_rate": 0.0001569809365416768,
+      "loss": 1.1123,
+      "step": 15960
+    },
+    {
+      "epoch": 0.6146294513955727,
+      "grad_norm": 1.4137890338897705,
+      "learning_rate": 0.00015695608372697154,
+      "loss": 1.115,
+      "step": 15965
+    },
+    {
+      "epoch": 0.6148219441770933,
+      "grad_norm": 1.2815289497375488,
+      "learning_rate": 0.00015693122570400975,
+      "loss": 1.0876,
+      "step": 15970
+    },
+    {
+      "epoch": 0.615014436958614,
+      "grad_norm": 0.9300668835639954,
+      "learning_rate": 0.00015690636247506448,
+      "loss": 1.1442,
+      "step": 15975
+    },
+    {
+      "epoch": 0.6152069297401348,
+      "grad_norm": 0.9866906404495239,
+      "learning_rate": 0.00015688149404240938,
+      "loss": 1.0664,
+      "step": 15980
+    },
+    {
+      "epoch": 0.6153994225216555,
+      "grad_norm": 1.1951825618743896,
+      "learning_rate": 0.0001568566204083184,
+      "loss": 1.0933,
+      "step": 15985
+    },
+    {
+      "epoch": 0.6155919153031761,
+      "grad_norm": 1.4439541101455688,
+      "learning_rate": 0.00015683174157506616,
+      "loss": 1.1618,
+      "step": 15990
+    },
+    {
+      "epoch": 0.6157844080846968,
+      "grad_norm": 1.242619276046753,
+      "learning_rate": 0.00015680685754492762,
+      "loss": 1.0794,
+      "step": 15995
+    },
+    {
+      "epoch": 0.6159769008662175,
+      "grad_norm": 1.9631248712539673,
+      "learning_rate": 0.00015678196832017823,
+      "loss": 1.1082,
+      "step": 16000
+    },
+    {
+      "epoch": 0.6161693936477383,
+      "grad_norm": 1.056715488433838,
+      "learning_rate": 0.00015675707390309403,
+      "loss": 1.0893,
+      "step": 16005
+    },
+    {
+      "epoch": 0.6163618864292589,
+      "grad_norm": 2.3864753246307373,
+      "learning_rate": 0.00015673217429595143,
+      "loss": 1.3378,
+      "step": 16010
+    },
+    {
+      "epoch": 0.6165543792107796,
+      "grad_norm": 1.3226178884506226,
+      "learning_rate": 0.00015670726950102725,
+      "loss": 1.1959,
+      "step": 16015
+    },
+    {
+      "epoch": 0.6167468719923003,
+      "grad_norm": 2.254422426223755,
+      "learning_rate": 0.00015668235952059892,
+      "loss": 1.1495,
+      "step": 16020
+    },
+    {
+      "epoch": 0.6169393647738209,
+      "grad_norm": 1.6376910209655762,
+      "learning_rate": 0.00015665744435694435,
+      "loss": 1.0027,
+      "step": 16025
+    },
+    {
+      "epoch": 0.6171318575553417,
+      "grad_norm": 1.190169334411621,
+      "learning_rate": 0.00015663252401234177,
+      "loss": 1.0419,
+      "step": 16030
+    },
+    {
+      "epoch": 0.6173243503368624,
+      "grad_norm": 1.6388911008834839,
+      "learning_rate": 0.00015660759848907008,
+      "loss": 1.3868,
+      "step": 16035
+    },
+    {
+      "epoch": 0.6175168431183831,
+      "grad_norm": 0.9445647597312927,
+      "learning_rate": 0.00015658266778940843,
+      "loss": 1.1382,
+      "step": 16040
+    },
+    {
+      "epoch": 0.6177093358999037,
+      "grad_norm": 0.9717797636985779,
+      "learning_rate": 0.00015655773191563664,
+      "loss": 1.3385,
+      "step": 16045
+    },
+    {
+      "epoch": 0.6179018286814244,
+      "grad_norm": 1.7297828197479248,
+      "learning_rate": 0.000156532790870035,
+      "loss": 1.2607,
+      "step": 16050
+    },
+    {
+      "epoch": 0.6180943214629452,
+      "grad_norm": 1.3885836601257324,
+      "learning_rate": 0.00015650784465488405,
+      "loss": 1.2271,
+      "step": 16055
+    },
+    {
+      "epoch": 0.6182868142444659,
+      "grad_norm": 1.3968501091003418,
+      "learning_rate": 0.00015648289327246508,
+      "loss": 1.255,
+      "step": 16060
+    },
+    {
+      "epoch": 0.6184793070259865,
+      "grad_norm": 1.7532678842544556,
+      "learning_rate": 0.00015645793672505967,
+      "loss": 1.2088,
+      "step": 16065
+    },
+    {
+      "epoch": 0.6186717998075072,
+      "grad_norm": 1.4146851301193237,
+      "learning_rate": 0.00015643297501494999,
+      "loss": 0.9797,
+      "step": 16070
+    },
+    {
+      "epoch": 0.6188642925890279,
+      "grad_norm": 1.4249024391174316,
+      "learning_rate": 0.00015640800814441851,
+      "loss": 1.1446,
+      "step": 16075
+    },
+    {
+      "epoch": 0.6190567853705486,
+      "grad_norm": 1.3387399911880493,
+      "learning_rate": 0.0001563830361157484,
+      "loss": 1.2204,
+      "step": 16080
+    },
+    {
+      "epoch": 0.6192492781520693,
+      "grad_norm": 1.137149691581726,
+      "learning_rate": 0.00015635805893122312,
+      "loss": 1.1626,
+      "step": 16085
+    },
+    {
+      "epoch": 0.61944177093359,
+      "grad_norm": 1.8353437185287476,
+      "learning_rate": 0.0001563330765931267,
+      "loss": 1.2476,
+      "step": 16090
+    },
+    {
+      "epoch": 0.6196342637151107,
+      "grad_norm": 0.969289243221283,
+      "learning_rate": 0.00015630808910374358,
+      "loss": 1.026,
+      "step": 16095
+    },
+    {
+      "epoch": 0.6198267564966313,
+      "grad_norm": 1.0529965162277222,
+      "learning_rate": 0.0001562830964653587,
+      "loss": 1.15,
+      "step": 16100
+    },
+    {
+      "epoch": 0.6200192492781521,
+      "grad_norm": 1.2508490085601807,
+      "learning_rate": 0.00015625809868025756,
+      "loss": 1.08,
+      "step": 16105
+    },
+    {
+      "epoch": 0.6202117420596728,
+      "grad_norm": 1.1188933849334717,
+      "learning_rate": 0.0001562330957507259,
+      "loss": 1.2273,
+      "step": 16110
+    },
+    {
+      "epoch": 0.6204042348411934,
+      "grad_norm": 1.9137325286865234,
+      "learning_rate": 0.00015620808767905018,
+      "loss": 1.1073,
+      "step": 16115
+    },
+    {
+      "epoch": 0.6205967276227141,
+      "grad_norm": 1.146921157836914,
+      "learning_rate": 0.0001561830744675172,
+      "loss": 1.0292,
+      "step": 16120
+    },
+    {
+      "epoch": 0.6207892204042348,
+      "grad_norm": 1.6574608087539673,
+      "learning_rate": 0.00015615805611841424,
+      "loss": 1.2067,
+      "step": 16125
+    },
+    {
+      "epoch": 0.6209817131857556,
+      "grad_norm": 1.599156379699707,
+      "learning_rate": 0.00015613303263402903,
+      "loss": 1.4416,
+      "step": 16130
+    },
+    {
+      "epoch": 0.6211742059672762,
+      "grad_norm": 1.9472912549972534,
+      "learning_rate": 0.00015610800401664988,
+      "loss": 0.9591,
+      "step": 16135
+    },
+    {
+      "epoch": 0.6213666987487969,
+      "grad_norm": 1.2037914991378784,
+      "learning_rate": 0.00015608297026856538,
+      "loss": 1.0899,
+      "step": 16140
+    },
+    {
+      "epoch": 0.6215591915303176,
+      "grad_norm": 1.0116618871688843,
+      "learning_rate": 0.0001560579313920648,
+      "loss": 1.2294,
+      "step": 16145
+    },
+    {
+      "epoch": 0.6217516843118384,
+      "grad_norm": 1.6344687938690186,
+      "learning_rate": 0.00015603288738943774,
+      "loss": 1.1918,
+      "step": 16150
+    },
+    {
+      "epoch": 0.621944177093359,
+      "grad_norm": 1.3862853050231934,
+      "learning_rate": 0.0001560078382629743,
+      "loss": 1.157,
+      "step": 16155
+    },
+    {
+      "epoch": 0.6221366698748797,
+      "grad_norm": 0.9576367139816284,
+      "learning_rate": 0.00015598278401496508,
+      "loss": 1.0759,
+      "step": 16160
+    },
+    {
+      "epoch": 0.6223291626564004,
+      "grad_norm": 1.2092609405517578,
+      "learning_rate": 0.0001559577246477011,
+      "loss": 1.1928,
+      "step": 16165
+    },
+    {
+      "epoch": 0.622521655437921,
+      "grad_norm": 1.594510793685913,
+      "learning_rate": 0.0001559326601634739,
+      "loss": 1.2336,
+      "step": 16170
+    },
+    {
+      "epoch": 0.6227141482194418,
+      "grad_norm": 0.851620078086853,
+      "learning_rate": 0.00015590759056457546,
+      "loss": 1.1646,
+      "step": 16175
+    },
+    {
+      "epoch": 0.6229066410009625,
+      "grad_norm": 1.1468600034713745,
+      "learning_rate": 0.0001558825158532982,
+      "loss": 1.1879,
+      "step": 16180
+    },
+    {
+      "epoch": 0.6230991337824832,
+      "grad_norm": 1.934251308441162,
+      "learning_rate": 0.00015585743603193505,
+      "loss": 1.1207,
+      "step": 16185
+    },
+    {
+      "epoch": 0.6232916265640038,
+      "grad_norm": 0.9963223934173584,
+      "learning_rate": 0.00015583235110277943,
+      "loss": 1.068,
+      "step": 16190
+    },
+    {
+      "epoch": 0.6234841193455245,
+      "grad_norm": 0.8857359290122986,
+      "learning_rate": 0.00015580726106812512,
+      "loss": 1.1148,
+      "step": 16195
+    },
+    {
+      "epoch": 0.6236766121270453,
+      "grad_norm": 1.2589722871780396,
+      "learning_rate": 0.00015578216593026647,
+      "loss": 1.0485,
+      "step": 16200
+    },
+    {
+      "epoch": 0.623869104908566,
+      "grad_norm": 1.0346484184265137,
+      "learning_rate": 0.0001557570656914983,
+      "loss": 1.1276,
+      "step": 16205
+    },
+    {
+      "epoch": 0.6240615976900866,
+      "grad_norm": 0.8794786334037781,
+      "learning_rate": 0.0001557319603541158,
+      "loss": 1.2591,
+      "step": 16210
+    },
+    {
+      "epoch": 0.6242540904716073,
+      "grad_norm": 1.0909137725830078,
+      "learning_rate": 0.00015570684992041473,
+      "loss": 1.1197,
+      "step": 16215
+    },
+    {
+      "epoch": 0.624446583253128,
+      "grad_norm": 1.3499592542648315,
+      "learning_rate": 0.0001556817343926913,
+      "loss": 1.0165,
+      "step": 16220
+    },
+    {
+      "epoch": 0.6246390760346487,
+      "grad_norm": 1.5356526374816895,
+      "learning_rate": 0.00015565661377324203,
+      "loss": 1.0144,
+      "step": 16225
+    },
+    {
+      "epoch": 0.6248315688161694,
+      "grad_norm": 1.849442958831787,
+      "learning_rate": 0.0001556314880643642,
+      "loss": 1.2191,
+      "step": 16230
+    },
+    {
+      "epoch": 0.6250240615976901,
+      "grad_norm": 1.1928755044937134,
+      "learning_rate": 0.00015560635726835525,
+      "loss": 1.2685,
+      "step": 16235
+    },
+    {
+      "epoch": 0.6252165543792108,
+      "grad_norm": 1.1445300579071045,
+      "learning_rate": 0.00015558122138751332,
+      "loss": 1.445,
+      "step": 16240
+    },
+    {
+      "epoch": 0.6254090471607314,
+      "grad_norm": 1.7465559244155884,
+      "learning_rate": 0.00015555608042413689,
+      "loss": 1.2479,
+      "step": 16245
+    },
+    {
+      "epoch": 0.6256015399422522,
+      "grad_norm": 1.1695505380630493,
+      "learning_rate": 0.0001555309343805249,
+      "loss": 1.2347,
+      "step": 16250
+    },
+    {
+      "epoch": 0.6257940327237729,
+      "grad_norm": 1.2655342817306519,
+      "learning_rate": 0.00015550578325897687,
+      "loss": 1.2343,
+      "step": 16255
+    },
+    {
+      "epoch": 0.6259865255052935,
+      "grad_norm": 1.569800853729248,
+      "learning_rate": 0.0001554806270617926,
+      "loss": 1.0798,
+      "step": 16260
+    },
+    {
+      "epoch": 0.6261790182868142,
+      "grad_norm": 2.0027542114257812,
+      "learning_rate": 0.00015545546579127256,
+      "loss": 1.0084,
+      "step": 16265
+    },
+    {
+      "epoch": 0.6263715110683349,
+      "grad_norm": 2.259096145629883,
+      "learning_rate": 0.0001554302994497175,
+      "loss": 1.1921,
+      "step": 16270
+    },
+    {
+      "epoch": 0.6265640038498557,
+      "grad_norm": 1.092046856880188,
+      "learning_rate": 0.00015540512803942878,
+      "loss": 1.112,
+      "step": 16275
+    },
+    {
+      "epoch": 0.6267564966313763,
+      "grad_norm": 1.67642343044281,
+      "learning_rate": 0.00015537995156270808,
+      "loss": 1.3709,
+      "step": 16280
+    },
+    {
+      "epoch": 0.626948989412897,
+      "grad_norm": 1.7039928436279297,
+      "learning_rate": 0.0001553547700218577,
+      "loss": 1.2211,
+      "step": 16285
+    },
+    {
+      "epoch": 0.6271414821944177,
+      "grad_norm": 2.0744543075561523,
+      "learning_rate": 0.00015532958341918027,
+      "loss": 1.2324,
+      "step": 16290
+    },
+    {
+      "epoch": 0.6273339749759385,
+      "grad_norm": 1.2610362768173218,
+      "learning_rate": 0.00015530439175697898,
+      "loss": 1.1924,
+      "step": 16295
+    },
+    {
+      "epoch": 0.6275264677574591,
+      "grad_norm": 1.8385295867919922,
+      "learning_rate": 0.00015527919503755742,
+      "loss": 1.2602,
+      "step": 16300
+    },
+    {
+      "epoch": 0.6277189605389798,
+      "grad_norm": 1.62607741355896,
+      "learning_rate": 0.00015525399326321966,
+      "loss": 1.2135,
+      "step": 16305
+    },
+    {
+      "epoch": 0.6279114533205005,
+      "grad_norm": 1.164507508277893,
+      "learning_rate": 0.00015522878643627023,
+      "loss": 1.07,
+      "step": 16310
+    },
+    {
+      "epoch": 0.6281039461020211,
+      "grad_norm": 0.9871059060096741,
+      "learning_rate": 0.0001552035745590142,
+      "loss": 1.1749,
+      "step": 16315
+    },
+    {
+      "epoch": 0.6282964388835419,
+      "grad_norm": 1.1414002180099487,
+      "learning_rate": 0.00015517835763375688,
+      "loss": 1.233,
+      "step": 16320
+    },
+    {
+      "epoch": 0.6284889316650626,
+      "grad_norm": 1.1266084909439087,
+      "learning_rate": 0.00015515313566280428,
+      "loss": 1.1642,
+      "step": 16325
+    },
+    {
+      "epoch": 0.6286814244465833,
+      "grad_norm": 1.8156638145446777,
+      "learning_rate": 0.00015512790864846286,
+      "loss": 1.0328,
+      "step": 16330
+    },
+    {
+      "epoch": 0.6288739172281039,
+      "grad_norm": 1.9357597827911377,
+      "learning_rate": 0.00015510267659303933,
+      "loss": 1.3325,
+      "step": 16335
+    },
+    {
+      "epoch": 0.6290664100096246,
+      "grad_norm": 1.767910122871399,
+      "learning_rate": 0.00015507743949884104,
+      "loss": 1.2381,
+      "step": 16340
+    },
+    {
+      "epoch": 0.6292589027911454,
+      "grad_norm": 2.1196887493133545,
+      "learning_rate": 0.0001550521973681758,
+      "loss": 1.2286,
+      "step": 16345
+    },
+    {
+      "epoch": 0.629451395572666,
+      "grad_norm": 1.7220022678375244,
+      "learning_rate": 0.00015502695020335177,
+      "loss": 1.1699,
+      "step": 16350
+    },
+    {
+      "epoch": 0.6296438883541867,
+      "grad_norm": 1.9612696170806885,
+      "learning_rate": 0.00015500169800667765,
+      "loss": 1.0786,
+      "step": 16355
+    },
+    {
+      "epoch": 0.6298363811357074,
+      "grad_norm": 1.666223406791687,
+      "learning_rate": 0.00015497644078046261,
+      "loss": 1.2211,
+      "step": 16360
+    },
+    {
+      "epoch": 0.6300288739172281,
+      "grad_norm": 1.7156059741973877,
+      "learning_rate": 0.00015495117852701626,
+      "loss": 1.0621,
+      "step": 16365
+    },
+    {
+      "epoch": 0.6302213666987488,
+      "grad_norm": 1.5840719938278198,
+      "learning_rate": 0.00015492591124864865,
+      "loss": 1.2364,
+      "step": 16370
+    },
+    {
+      "epoch": 0.6304138594802695,
+      "grad_norm": 1.1821776628494263,
+      "learning_rate": 0.0001549006389476703,
+      "loss": 1.1479,
+      "step": 16375
+    },
+    {
+      "epoch": 0.6306063522617902,
+      "grad_norm": 1.2549364566802979,
+      "learning_rate": 0.00015487536162639223,
+      "loss": 1.0564,
+      "step": 16380
+    },
+    {
+      "epoch": 0.6307988450433109,
+      "grad_norm": 1.5308479070663452,
+      "learning_rate": 0.0001548500792871258,
+      "loss": 1.1825,
+      "step": 16385
+    },
+    {
+      "epoch": 0.6309913378248315,
+      "grad_norm": 1.6546053886413574,
+      "learning_rate": 0.000154824791932183,
+      "loss": 1.1673,
+      "step": 16390
+    },
+    {
+      "epoch": 0.6311838306063523,
+      "grad_norm": 1.1561111211776733,
+      "learning_rate": 0.00015479949956387617,
+      "loss": 1.1014,
+      "step": 16395
+    },
+    {
+      "epoch": 0.631376323387873,
+      "grad_norm": 1.6901589632034302,
+      "learning_rate": 0.0001547742021845181,
+      "loss": 1.2377,
+      "step": 16400
+    },
+    {
+      "epoch": 0.6315688161693936,
+      "grad_norm": 1.2808809280395508,
+      "learning_rate": 0.0001547488997964221,
+      "loss": 1.2976,
+      "step": 16405
+    },
+    {
+      "epoch": 0.6317613089509143,
+      "grad_norm": 0.9793625473976135,
+      "learning_rate": 0.0001547235924019019,
+      "loss": 1.0328,
+      "step": 16410
+    },
+    {
+      "epoch": 0.631953801732435,
+      "grad_norm": 1.6001505851745605,
+      "learning_rate": 0.00015469828000327164,
+      "loss": 1.0232,
+      "step": 16415
+    },
+    {
+      "epoch": 0.6321462945139558,
+      "grad_norm": 1.3900479078292847,
+      "learning_rate": 0.00015467296260284605,
+      "loss": 1.2412,
+      "step": 16420
+    },
+    {
+      "epoch": 0.6323387872954764,
+      "grad_norm": 1.6030535697937012,
+      "learning_rate": 0.0001546476402029402,
+      "loss": 1.178,
+      "step": 16425
+    },
+    {
+      "epoch": 0.6325312800769971,
+      "grad_norm": 1.5602627992630005,
+      "learning_rate": 0.00015462231280586965,
+      "loss": 1.2834,
+      "step": 16430
+    },
+    {
+      "epoch": 0.6327237728585178,
+      "grad_norm": 1.3648455142974854,
+      "learning_rate": 0.00015459698041395045,
+      "loss": 1.1425,
+      "step": 16435
+    },
+    {
+      "epoch": 0.6329162656400384,
+      "grad_norm": 1.4346479177474976,
+      "learning_rate": 0.00015457164302949908,
+      "loss": 1.0076,
+      "step": 16440
+    },
+    {
+      "epoch": 0.6331087584215592,
+      "grad_norm": 0.9692068696022034,
+      "learning_rate": 0.00015454630065483242,
+      "loss": 1.0133,
+      "step": 16445
+    },
+    {
+      "epoch": 0.6333012512030799,
+      "grad_norm": 1.479915976524353,
+      "learning_rate": 0.0001545209532922679,
+      "loss": 1.131,
+      "step": 16450
+    },
+    {
+      "epoch": 0.6334937439846006,
+      "grad_norm": 1.0446960926055908,
+      "learning_rate": 0.00015449560094412342,
+      "loss": 1.2545,
+      "step": 16455
+    },
+    {
+      "epoch": 0.6336862367661212,
+      "grad_norm": 1.458414077758789,
+      "learning_rate": 0.00015447024361271721,
+      "loss": 1.325,
+      "step": 16460
+    },
+    {
+      "epoch": 0.633878729547642,
+      "grad_norm": 1.2071151733398438,
+      "learning_rate": 0.00015444488130036802,
+      "loss": 1.2303,
+      "step": 16465
+    },
+    {
+      "epoch": 0.6340712223291627,
+      "grad_norm": 1.9108256101608276,
+      "learning_rate": 0.00015441951400939515,
+      "loss": 1.2031,
+      "step": 16470
+    },
+    {
+      "epoch": 0.6342637151106834,
+      "grad_norm": 1.1393382549285889,
+      "learning_rate": 0.0001543941417421182,
+      "loss": 1.016,
+      "step": 16475
+    },
+    {
+      "epoch": 0.634456207892204,
+      "grad_norm": 2.0735628604888916,
+      "learning_rate": 0.00015436876450085728,
+      "loss": 1.1619,
+      "step": 16480
+    },
+    {
+      "epoch": 0.6346487006737247,
+      "grad_norm": 1.6895620822906494,
+      "learning_rate": 0.00015434338228793306,
+      "loss": 1.1621,
+      "step": 16485
+    },
+    {
+      "epoch": 0.6348411934552455,
+      "grad_norm": 1.9663159847259521,
+      "learning_rate": 0.0001543179951056665,
+      "loss": 1.2465,
+      "step": 16490
+    },
+    {
+      "epoch": 0.6350336862367661,
+      "grad_norm": 1.1372085809707642,
+      "learning_rate": 0.0001542926029563791,
+      "loss": 1.1643,
+      "step": 16495
+    },
+    {
+      "epoch": 0.6352261790182868,
+      "grad_norm": 0.5948193669319153,
+      "learning_rate": 0.00015426720584239283,
+      "loss": 0.9659,
+      "step": 16500
+    },
+    {
+      "epoch": 0.6354186717998075,
+      "grad_norm": 1.829047441482544,
+      "learning_rate": 0.00015424180376603008,
+      "loss": 1.1334,
+      "step": 16505
+    },
+    {
+      "epoch": 0.6356111645813282,
+      "grad_norm": 1.4863371849060059,
+      "learning_rate": 0.00015421639672961367,
+      "loss": 1.1206,
+      "step": 16510
+    },
+    {
+      "epoch": 0.6358036573628489,
+      "grad_norm": 1.2481038570404053,
+      "learning_rate": 0.00015419098473546696,
+      "loss": 1.1101,
+      "step": 16515
+    },
+    {
+      "epoch": 0.6359961501443696,
+      "grad_norm": 1.8721559047698975,
+      "learning_rate": 0.00015416556778591363,
+      "loss": 1.1293,
+      "step": 16520
+    },
+    {
+      "epoch": 0.6361886429258903,
+      "grad_norm": 1.5730985403060913,
+      "learning_rate": 0.000154140145883278,
+      "loss": 1.277,
+      "step": 16525
+    },
+    {
+      "epoch": 0.636381135707411,
+      "grad_norm": 1.4351321458816528,
+      "learning_rate": 0.00015411471902988463,
+      "loss": 1.2475,
+      "step": 16530
+    },
+    {
+      "epoch": 0.6365736284889316,
+      "grad_norm": 0.8733989596366882,
+      "learning_rate": 0.00015408928722805874,
+      "loss": 1.0728,
+      "step": 16535
+    },
+    {
+      "epoch": 0.6367661212704524,
+      "grad_norm": 1.454068899154663,
+      "learning_rate": 0.00015406385048012577,
+      "loss": 1.0163,
+      "step": 16540
+    },
+    {
+      "epoch": 0.6369586140519731,
+      "grad_norm": 0.9600105285644531,
+      "learning_rate": 0.00015403840878841182,
+      "loss": 1.097,
+      "step": 16545
+    },
+    {
+      "epoch": 0.6371511068334937,
+      "grad_norm": 2.419609546661377,
+      "learning_rate": 0.00015401296215524345,
+      "loss": 1.2003,
+      "step": 16550
+    },
+    {
+      "epoch": 0.6373435996150144,
+      "grad_norm": 1.313755989074707,
+      "learning_rate": 0.0001539875105829474,
+      "loss": 1.1276,
+      "step": 16555
+    },
+    {
+      "epoch": 0.6375360923965351,
+      "grad_norm": 1.6932001113891602,
+      "learning_rate": 0.00015396205407385116,
+      "loss": 1.1689,
+      "step": 16560
+    },
+    {
+      "epoch": 0.6377285851780559,
+      "grad_norm": 1.076905608177185,
+      "learning_rate": 0.00015393659263028257,
+      "loss": 1.189,
+      "step": 16565
+    },
+    {
+      "epoch": 0.6379210779595765,
+      "grad_norm": 1.2433785200119019,
+      "learning_rate": 0.00015391112625456983,
+      "loss": 0.9797,
+      "step": 16570
+    },
+    {
+      "epoch": 0.6381135707410972,
+      "grad_norm": 1.1299281120300293,
+      "learning_rate": 0.00015388565494904176,
+      "loss": 1.1399,
+      "step": 16575
+    },
+    {
+      "epoch": 0.6383060635226179,
+      "grad_norm": 1.0440160036087036,
+      "learning_rate": 0.0001538601787160275,
+      "loss": 1.2491,
+      "step": 16580
+    },
+    {
+      "epoch": 0.6384985563041385,
+      "grad_norm": 1.1874500513076782,
+      "learning_rate": 0.00015383469755785668,
+      "loss": 1.2762,
+      "step": 16585
+    },
+    {
+      "epoch": 0.6386910490856593,
+      "grad_norm": 1.2737995386123657,
+      "learning_rate": 0.0001538092114768594,
+      "loss": 1.2102,
+      "step": 16590
+    },
+    {
+      "epoch": 0.63888354186718,
+      "grad_norm": 1.8649038076400757,
+      "learning_rate": 0.0001537837204753662,
+      "loss": 1.0711,
+      "step": 16595
+    },
+    {
+      "epoch": 0.6390760346487007,
+      "grad_norm": 1.0375845432281494,
+      "learning_rate": 0.000153758224555708,
+      "loss": 1.0349,
+      "step": 16600
+    },
+    {
+      "epoch": 0.6392685274302213,
+      "grad_norm": 1.4500465393066406,
+      "learning_rate": 0.0001537327237202163,
+      "loss": 1.1501,
+      "step": 16605
+    },
+    {
+      "epoch": 0.6394610202117421,
+      "grad_norm": 1.5905102491378784,
+      "learning_rate": 0.000153707217971223,
+      "loss": 1.1946,
+      "step": 16610
+    },
+    {
+      "epoch": 0.6396535129932628,
+      "grad_norm": 1.224752426147461,
+      "learning_rate": 0.00015368170731106036,
+      "loss": 1.1101,
+      "step": 16615
+    },
+    {
+      "epoch": 0.6398460057747835,
+      "grad_norm": 2.605717182159424,
+      "learning_rate": 0.00015365619174206117,
+      "loss": 1.0483,
+      "step": 16620
+    },
+    {
+      "epoch": 0.6400384985563041,
+      "grad_norm": 1.2829294204711914,
+      "learning_rate": 0.00015363067126655873,
+      "loss": 1.2265,
+      "step": 16625
+    },
+    {
+      "epoch": 0.6402309913378248,
+      "grad_norm": 1.1748125553131104,
+      "learning_rate": 0.00015360514588688665,
+      "loss": 1.0909,
+      "step": 16630
+    },
+    {
+      "epoch": 0.6404234841193456,
+      "grad_norm": 1.0052121877670288,
+      "learning_rate": 0.00015357961560537908,
+      "loss": 1.3145,
+      "step": 16635
+    },
+    {
+      "epoch": 0.6406159769008662,
+      "grad_norm": 1.1692798137664795,
+      "learning_rate": 0.00015355408042437061,
+      "loss": 1.3134,
+      "step": 16640
+    },
+    {
+      "epoch": 0.6408084696823869,
+      "grad_norm": 1.1379728317260742,
+      "learning_rate": 0.00015352854034619622,
+      "loss": 1.0519,
+      "step": 16645
+    },
+    {
+      "epoch": 0.6410009624639076,
+      "grad_norm": 1.067920207977295,
+      "learning_rate": 0.00015350299537319147,
+      "loss": 1.17,
+      "step": 16650
+    },
+    {
+      "epoch": 0.6411934552454283,
+      "grad_norm": 1.9951469898223877,
+      "learning_rate": 0.00015347744550769216,
+      "loss": 1.0478,
+      "step": 16655
+    },
+    {
+      "epoch": 0.641385948026949,
+      "grad_norm": 1.036605715751648,
+      "learning_rate": 0.00015345189075203477,
+      "loss": 1.1288,
+      "step": 16660
+    },
+    {
+      "epoch": 0.6415784408084697,
+      "grad_norm": 0.5938658714294434,
+      "learning_rate": 0.000153426331108556,
+      "loss": 1.0589,
+      "step": 16665
+    },
+    {
+      "epoch": 0.6417709335899904,
+      "grad_norm": 1.212049961090088,
+      "learning_rate": 0.00015340076657959317,
+      "loss": 1.1104,
+      "step": 16670
+    },
+    {
+      "epoch": 0.641963426371511,
+      "grad_norm": 1.3548222780227661,
+      "learning_rate": 0.00015337519716748403,
+      "loss": 1.2639,
+      "step": 16675
+    },
+    {
+      "epoch": 0.6421559191530317,
+      "grad_norm": 1.210879921913147,
+      "learning_rate": 0.00015334962287456665,
+      "loss": 1.0576,
+      "step": 16680
+    },
+    {
+      "epoch": 0.6423484119345525,
+      "grad_norm": 2.2316668033599854,
+      "learning_rate": 0.00015332404370317965,
+      "loss": 1.2075,
+      "step": 16685
+    },
+    {
+      "epoch": 0.6425409047160732,
+      "grad_norm": 1.0065557956695557,
+      "learning_rate": 0.00015329845965566215,
+      "loss": 1.0872,
+      "step": 16690
+    },
+    {
+      "epoch": 0.6427333974975938,
+      "grad_norm": 1.36894953250885,
+      "learning_rate": 0.00015327287073435355,
+      "loss": 0.9866,
+      "step": 16695
+    },
+    {
+      "epoch": 0.6429258902791145,
+      "grad_norm": 0.9726212620735168,
+      "learning_rate": 0.0001532472769415938,
+      "loss": 1.0069,
+      "step": 16700
+    },
+    {
+      "epoch": 0.6431183830606352,
+      "grad_norm": 0.9447348117828369,
+      "learning_rate": 0.00015322167827972334,
+      "loss": 1.3184,
+      "step": 16705
+    },
+    {
+      "epoch": 0.643310875842156,
+      "grad_norm": 1.7236000299453735,
+      "learning_rate": 0.00015319607475108296,
+      "loss": 1.2547,
+      "step": 16710
+    },
+    {
+      "epoch": 0.6435033686236766,
+      "grad_norm": 2.3541550636291504,
+      "learning_rate": 0.00015317046635801392,
+      "loss": 1.2886,
+      "step": 16715
+    },
+    {
+      "epoch": 0.6436958614051973,
+      "grad_norm": 1.8849072456359863,
+      "learning_rate": 0.00015314485310285796,
+      "loss": 1.1295,
+      "step": 16720
+    },
+    {
+      "epoch": 0.643888354186718,
+      "grad_norm": 4.183611869812012,
+      "learning_rate": 0.00015311923498795724,
+      "loss": 1.1109,
+      "step": 16725
+    },
+    {
+      "epoch": 0.6440808469682386,
+      "grad_norm": 1.4037699699401855,
+      "learning_rate": 0.00015309361201565436,
+      "loss": 1.1097,
+      "step": 16730
+    },
+    {
+      "epoch": 0.6442733397497594,
+      "grad_norm": 1.626489520072937,
+      "learning_rate": 0.00015306798418829236,
+      "loss": 1.2515,
+      "step": 16735
+    },
+    {
+      "epoch": 0.6444658325312801,
+      "grad_norm": 2.0744874477386475,
+      "learning_rate": 0.00015304235150821475,
+      "loss": 1.2196,
+      "step": 16740
+    },
+    {
+      "epoch": 0.6446583253128008,
+      "grad_norm": 1.2196972370147705,
+      "learning_rate": 0.0001530167139777655,
+      "loss": 1.0935,
+      "step": 16745
+    },
+    {
+      "epoch": 0.6448508180943214,
+      "grad_norm": 2.1287968158721924,
+      "learning_rate": 0.00015299107159928897,
+      "loss": 0.9476,
+      "step": 16750
+    },
+    {
+      "epoch": 0.6450433108758421,
+      "grad_norm": 1.6050670146942139,
+      "learning_rate": 0.00015296542437512995,
+      "loss": 1.2276,
+      "step": 16755
+    },
+    {
+      "epoch": 0.6452358036573629,
+      "grad_norm": 1.316373348236084,
+      "learning_rate": 0.0001529397723076337,
+      "loss": 1.121,
+      "step": 16760
+    },
+    {
+      "epoch": 0.6454282964388836,
+      "grad_norm": 1.4219224452972412,
+      "learning_rate": 0.00015291411539914603,
+      "loss": 1.3219,
+      "step": 16765
+    },
+    {
+      "epoch": 0.6456207892204042,
+      "grad_norm": 1.3470525741577148,
+      "learning_rate": 0.00015288845365201299,
+      "loss": 1.0538,
+      "step": 16770
+    },
+    {
+      "epoch": 0.6458132820019249,
+      "grad_norm": 1.6893870830535889,
+      "learning_rate": 0.0001528627870685812,
+      "loss": 1.1907,
+      "step": 16775
+    },
+    {
+      "epoch": 0.6460057747834457,
+      "grad_norm": 1.7264561653137207,
+      "learning_rate": 0.00015283711565119775,
+      "loss": 1.167,
+      "step": 16780
+    },
+    {
+      "epoch": 0.6461982675649663,
+      "grad_norm": 1.1093302965164185,
+      "learning_rate": 0.0001528114394022101,
+      "loss": 1.1477,
+      "step": 16785
+    },
+    {
+      "epoch": 0.646390760346487,
+      "grad_norm": 1.1114470958709717,
+      "learning_rate": 0.00015278575832396613,
+      "loss": 1.1224,
+      "step": 16790
+    },
+    {
+      "epoch": 0.6465832531280077,
+      "grad_norm": 2.0239744186401367,
+      "learning_rate": 0.00015276007241881424,
+      "loss": 1.1655,
+      "step": 16795
+    },
+    {
+      "epoch": 0.6467757459095284,
+      "grad_norm": 1.0726968050003052,
+      "learning_rate": 0.00015273438168910322,
+      "loss": 0.9021,
+      "step": 16800
+    },
+    {
+      "epoch": 0.6469682386910491,
+      "grad_norm": 1.2715688943862915,
+      "learning_rate": 0.00015270868613718238,
+      "loss": 1.1776,
+      "step": 16805
+    },
+    {
+      "epoch": 0.6471607314725698,
+      "grad_norm": 1.4808478355407715,
+      "learning_rate": 0.00015268298576540129,
+      "loss": 1.1023,
+      "step": 16810
+    },
+    {
+      "epoch": 0.6473532242540905,
+      "grad_norm": 1.63973069190979,
+      "learning_rate": 0.0001526572805761102,
+      "loss": 1.0025,
+      "step": 16815
+    },
+    {
+      "epoch": 0.6475457170356111,
+      "grad_norm": 1.0935505628585815,
+      "learning_rate": 0.0001526315705716596,
+      "loss": 1.1039,
+      "step": 16820
+    },
+    {
+      "epoch": 0.6477382098171318,
+      "grad_norm": 1.0586233139038086,
+      "learning_rate": 0.00015260585575440052,
+      "loss": 1.0884,
+      "step": 16825
+    },
+    {
+      "epoch": 0.6479307025986526,
+      "grad_norm": 1.0608752965927124,
+      "learning_rate": 0.0001525801361266844,
+      "loss": 1.2997,
+      "step": 16830
+    },
+    {
+      "epoch": 0.6481231953801733,
+      "grad_norm": 1.0017322301864624,
+      "learning_rate": 0.00015255441169086318,
+      "loss": 1.3023,
+      "step": 16835
+    },
+    {
+      "epoch": 0.6483156881616939,
+      "grad_norm": 0.9409940242767334,
+      "learning_rate": 0.00015252868244928914,
+      "loss": 1.2462,
+      "step": 16840
+    },
+    {
+      "epoch": 0.6485081809432146,
+      "grad_norm": 1.646735429763794,
+      "learning_rate": 0.00015250294840431504,
+      "loss": 1.1759,
+      "step": 16845
+    },
+    {
+      "epoch": 0.6487006737247353,
+      "grad_norm": 2.878627300262451,
+      "learning_rate": 0.00015247720955829412,
+      "loss": 1.2458,
+      "step": 16850
+    },
+    {
+      "epoch": 0.6488931665062561,
+      "grad_norm": 1.6578867435455322,
+      "learning_rate": 0.00015245146591358002,
+      "loss": 1.297,
+      "step": 16855
+    },
+    {
+      "epoch": 0.6490856592877767,
+      "grad_norm": 1.9454634189605713,
+      "learning_rate": 0.00015242571747252682,
+      "loss": 1.2366,
+      "step": 16860
+    },
+    {
+      "epoch": 0.6492781520692974,
+      "grad_norm": 1.8211311101913452,
+      "learning_rate": 0.00015239996423748906,
+      "loss": 1.1163,
+      "step": 16865
+    },
+    {
+      "epoch": 0.6494706448508181,
+      "grad_norm": 1.5382091999053955,
+      "learning_rate": 0.00015237420621082163,
+      "loss": 1.0103,
+      "step": 16870
+    },
+    {
+      "epoch": 0.6496631376323387,
+      "grad_norm": 1.7348453998565674,
+      "learning_rate": 0.00015234844339488004,
+      "loss": 1.1667,
+      "step": 16875
+    },
+    {
+      "epoch": 0.6498556304138595,
+      "grad_norm": 1.0255297422409058,
+      "learning_rate": 0.0001523226757920201,
+      "loss": 1.1472,
+      "step": 16880
+    },
+    {
+      "epoch": 0.6500481231953802,
+      "grad_norm": 1.730460524559021,
+      "learning_rate": 0.00015229690340459802,
+      "loss": 1.2442,
+      "step": 16885
+    },
+    {
+      "epoch": 0.6502406159769009,
+      "grad_norm": 1.6826850175857544,
+      "learning_rate": 0.00015227112623497058,
+      "loss": 1.2426,
+      "step": 16890
+    },
+    {
+      "epoch": 0.6504331087584215,
+      "grad_norm": 1.6523195505142212,
+      "learning_rate": 0.00015224534428549488,
+      "loss": 1.1543,
+      "step": 16895
+    },
+    {
+      "epoch": 0.6506256015399422,
+      "grad_norm": 2.3335843086242676,
+      "learning_rate": 0.00015221955755852858,
+      "loss": 1.115,
+      "step": 16900
+    },
+    {
+      "epoch": 0.650818094321463,
+      "grad_norm": 1.0122956037521362,
+      "learning_rate": 0.00015219376605642962,
+      "loss": 1.2913,
+      "step": 16905
+    },
+    {
+      "epoch": 0.6510105871029837,
+      "grad_norm": 1.5100213289260864,
+      "learning_rate": 0.00015216796978155655,
+      "loss": 1.0309,
+      "step": 16910
+    },
+    {
+      "epoch": 0.6512030798845043,
+      "grad_norm": 1.1331759691238403,
+      "learning_rate": 0.0001521421687362682,
+      "loss": 1.0732,
+      "step": 16915
+    },
+    {
+      "epoch": 0.651395572666025,
+      "grad_norm": 0.9450187087059021,
+      "learning_rate": 0.00015211636292292394,
+      "loss": 1.2011,
+      "step": 16920
+    },
+    {
+      "epoch": 0.6515880654475458,
+      "grad_norm": 1.1546697616577148,
+      "learning_rate": 0.00015209055234388354,
+      "loss": 1.1368,
+      "step": 16925
+    },
+    {
+      "epoch": 0.6517805582290664,
+      "grad_norm": 1.5972734689712524,
+      "learning_rate": 0.00015206473700150717,
+      "loss": 1.0546,
+      "step": 16930
+    },
+    {
+      "epoch": 0.6519730510105871,
+      "grad_norm": 1.1828382015228271,
+      "learning_rate": 0.0001520389168981555,
+      "loss": 1.0311,
+      "step": 16935
+    },
+    {
+      "epoch": 0.6521655437921078,
+      "grad_norm": 1.0515602827072144,
+      "learning_rate": 0.00015201309203618962,
+      "loss": 1.3763,
+      "step": 16940
+    },
+    {
+      "epoch": 0.6523580365736285,
+      "grad_norm": 1.0648945569992065,
+      "learning_rate": 0.00015198726241797103,
+      "loss": 1.136,
+      "step": 16945
+    },
+    {
+      "epoch": 0.6525505293551492,
+      "grad_norm": 1.3983291387557983,
+      "learning_rate": 0.00015196142804586166,
+      "loss": 1.121,
+      "step": 16950
+    },
+    {
+      "epoch": 0.6527430221366699,
+      "grad_norm": 1.1980384588241577,
+      "learning_rate": 0.00015193558892222394,
+      "loss": 1.1442,
+      "step": 16955
+    },
+    {
+      "epoch": 0.6529355149181906,
+      "grad_norm": 0.92877596616745,
+      "learning_rate": 0.00015190974504942064,
+      "loss": 1.1025,
+      "step": 16960
+    },
+    {
+      "epoch": 0.6531280076997112,
+      "grad_norm": 1.3868606090545654,
+      "learning_rate": 0.00015188389642981502,
+      "loss": 1.0714,
+      "step": 16965
+    },
+    {
+      "epoch": 0.6533205004812319,
+      "grad_norm": 2.058389663696289,
+      "learning_rate": 0.00015185804306577075,
+      "loss": 1.3543,
+      "step": 16970
+    },
+    {
+      "epoch": 0.6535129932627527,
+      "grad_norm": 0.5963343381881714,
+      "learning_rate": 0.00015183218495965202,
+      "loss": 0.9247,
+      "step": 16975
+    },
+    {
+      "epoch": 0.6537054860442734,
+      "grad_norm": 1.6353943347930908,
+      "learning_rate": 0.0001518063221138233,
+      "loss": 1.1284,
+      "step": 16980
+    },
+    {
+      "epoch": 0.653897978825794,
+      "grad_norm": 2.303635597229004,
+      "learning_rate": 0.00015178045453064962,
+      "loss": 1.3496,
+      "step": 16985
+    },
+    {
+      "epoch": 0.6540904716073147,
+      "grad_norm": 0.9238683581352234,
+      "learning_rate": 0.00015175458221249638,
+      "loss": 1.1348,
+      "step": 16990
+    },
+    {
+      "epoch": 0.6542829643888354,
+      "grad_norm": 1.4203814268112183,
+      "learning_rate": 0.00015172870516172942,
+      "loss": 1.1032,
+      "step": 16995
+    },
+    {
+      "epoch": 0.6544754571703562,
+      "grad_norm": 1.018648386001587,
+      "learning_rate": 0.0001517028233807151,
+      "loss": 1.237,
+      "step": 17000
+    },
+    {
+      "epoch": 0.6546679499518768,
+      "grad_norm": 1.4779586791992188,
+      "learning_rate": 0.00015167693687182,
+      "loss": 1.173,
+      "step": 17005
+    },
+    {
+      "epoch": 0.6548604427333975,
+      "grad_norm": 1.7097437381744385,
+      "learning_rate": 0.0001516510456374114,
+      "loss": 1.1935,
+      "step": 17010
+    },
+    {
+      "epoch": 0.6550529355149182,
+      "grad_norm": 1.4055527448654175,
+      "learning_rate": 0.00015162514967985682,
+      "loss": 1.0832,
+      "step": 17015
+    },
+    {
+      "epoch": 0.6552454282964388,
+      "grad_norm": 1.5012494325637817,
+      "learning_rate": 0.00015159924900152432,
+      "loss": 1.3221,
+      "step": 17020
+    },
+    {
+      "epoch": 0.6554379210779596,
+      "grad_norm": 1.13307785987854,
+      "learning_rate": 0.00015157334360478228,
+      "loss": 1.2599,
+      "step": 17025
+    },
+    {
+      "epoch": 0.6556304138594803,
+      "grad_norm": 2.10911226272583,
+      "learning_rate": 0.0001515474334919996,
+      "loss": 1.1446,
+      "step": 17030
+    },
+    {
+      "epoch": 0.655822906641001,
+      "grad_norm": 1.4689563512802124,
+      "learning_rate": 0.00015152151866554563,
+      "loss": 1.3851,
+      "step": 17035
+    },
+    {
+      "epoch": 0.6560153994225216,
+      "grad_norm": 1.3363420963287354,
+      "learning_rate": 0.00015149559912779005,
+      "loss": 1.1939,
+      "step": 17040
+    },
+    {
+      "epoch": 0.6562078922040423,
+      "grad_norm": 1.665319561958313,
+      "learning_rate": 0.00015146967488110307,
+      "loss": 1.3353,
+      "step": 17045
+    },
+    {
+      "epoch": 0.6564003849855631,
+      "grad_norm": 1.03946852684021,
+      "learning_rate": 0.00015144374592785528,
+      "loss": 1.0736,
+      "step": 17050
+    },
+    {
+      "epoch": 0.6565928777670837,
+      "grad_norm": 1.941311240196228,
+      "learning_rate": 0.0001514178122704177,
+      "loss": 1.1745,
+      "step": 17055
+    },
+    {
+      "epoch": 0.6567853705486044,
+      "grad_norm": 2.091871738433838,
+      "learning_rate": 0.00015139187391116182,
+      "loss": 0.9826,
+      "step": 17060
+    },
+    {
+      "epoch": 0.6569778633301251,
+      "grad_norm": 1.3722056150436401,
+      "learning_rate": 0.0001513659308524595,
+      "loss": 1.0969,
+      "step": 17065
+    },
+    {
+      "epoch": 0.6571703561116458,
+      "grad_norm": 1.9604045152664185,
+      "learning_rate": 0.00015133998309668306,
+      "loss": 1.0726,
+      "step": 17070
+    },
+    {
+      "epoch": 0.6573628488931665,
+      "grad_norm": 1.1731983423233032,
+      "learning_rate": 0.00015131403064620527,
+      "loss": 1.0909,
+      "step": 17075
+    },
+    {
+      "epoch": 0.6575553416746872,
+      "grad_norm": 1.3418563604354858,
+      "learning_rate": 0.0001512880735033993,
+      "loss": 1.2574,
+      "step": 17080
+    },
+    {
+      "epoch": 0.6577478344562079,
+      "grad_norm": 2.054722785949707,
+      "learning_rate": 0.00015126211167063876,
+      "loss": 1.1705,
+      "step": 17085
+    },
+    {
+      "epoch": 0.6579403272377286,
+      "grad_norm": 1.1431398391723633,
+      "learning_rate": 0.00015123614515029772,
+      "loss": 1.2606,
+      "step": 17090
+    },
+    {
+      "epoch": 0.6581328200192493,
+      "grad_norm": 1.4750339984893799,
+      "learning_rate": 0.0001512101739447506,
+      "loss": 1.1471,
+      "step": 17095
+    },
+    {
+      "epoch": 0.65832531280077,
+      "grad_norm": 1.6877497434616089,
+      "learning_rate": 0.00015118419805637228,
+      "loss": 0.986,
+      "step": 17100
+    },
+    {
+      "epoch": 0.6585178055822907,
+      "grad_norm": 0.6538336873054504,
+      "learning_rate": 0.0001511582174875381,
+      "loss": 0.9426,
+      "step": 17105
+    },
+    {
+      "epoch": 0.6587102983638113,
+      "grad_norm": 1.1754498481750488,
+      "learning_rate": 0.00015113223224062384,
+      "loss": 1.0994,
+      "step": 17110
+    },
+    {
+      "epoch": 0.658902791145332,
+      "grad_norm": 2.219837188720703,
+      "learning_rate": 0.00015110624231800567,
+      "loss": 1.1205,
+      "step": 17115
+    },
+    {
+      "epoch": 0.6590952839268528,
+      "grad_norm": 1.826324701309204,
+      "learning_rate": 0.0001510802477220602,
+      "loss": 1.2335,
+      "step": 17120
+    },
+    {
+      "epoch": 0.6592877767083735,
+      "grad_norm": 1.8668159246444702,
+      "learning_rate": 0.00015105424845516445,
+      "loss": 1.2609,
+      "step": 17125
+    },
+    {
+      "epoch": 0.6594802694898941,
+      "grad_norm": 0.9887051582336426,
+      "learning_rate": 0.00015102824451969585,
+      "loss": 1.0539,
+      "step": 17130
+    },
+    {
+      "epoch": 0.6596727622714148,
+      "grad_norm": 1.2473443746566772,
+      "learning_rate": 0.00015100223591803236,
+      "loss": 1.2355,
+      "step": 17135
+    },
+    {
+      "epoch": 0.6598652550529355,
+      "grad_norm": 1.2736021280288696,
+      "learning_rate": 0.00015097622265255222,
+      "loss": 1.3073,
+      "step": 17140
+    },
+    {
+      "epoch": 0.6600577478344563,
+      "grad_norm": 1.0870583057403564,
+      "learning_rate": 0.00015095020472563424,
+      "loss": 0.8381,
+      "step": 17145
+    },
+    {
+      "epoch": 0.6602502406159769,
+      "grad_norm": 1.6099382638931274,
+      "learning_rate": 0.0001509241821396575,
+      "loss": 1.2738,
+      "step": 17150
+    },
+    {
+      "epoch": 0.6604427333974976,
+      "grad_norm": 1.3321658372879028,
+      "learning_rate": 0.0001508981548970017,
+      "loss": 1.2924,
+      "step": 17155
+    },
+    {
+      "epoch": 0.6606352261790183,
+      "grad_norm": 1.0399209260940552,
+      "learning_rate": 0.00015087212300004678,
+      "loss": 0.9254,
+      "step": 17160
+    },
+    {
+      "epoch": 0.6608277189605389,
+      "grad_norm": 0.9332255721092224,
+      "learning_rate": 0.0001508460864511732,
+      "loss": 1.2693,
+      "step": 17165
+    },
+    {
+      "epoch": 0.6610202117420597,
+      "grad_norm": 1.408109188079834,
+      "learning_rate": 0.00015082004525276185,
+      "loss": 1.0394,
+      "step": 17170
+    },
+    {
+      "epoch": 0.6612127045235804,
+      "grad_norm": 1.3958436250686646,
+      "learning_rate": 0.00015079399940719402,
+      "loss": 1.1119,
+      "step": 17175
+    },
+    {
+      "epoch": 0.6614051973051011,
+      "grad_norm": 1.3326903581619263,
+      "learning_rate": 0.00015076794891685143,
+      "loss": 1.0996,
+      "step": 17180
+    },
+    {
+      "epoch": 0.6615976900866217,
+      "grad_norm": 1.1485531330108643,
+      "learning_rate": 0.00015074189378411622,
+      "loss": 1.1617,
+      "step": 17185
+    },
+    {
+      "epoch": 0.6617901828681424,
+      "grad_norm": 1.9735444784164429,
+      "learning_rate": 0.00015071583401137092,
+      "loss": 1.1168,
+      "step": 17190
+    },
+    {
+      "epoch": 0.6619826756496632,
+      "grad_norm": 1.6123241186141968,
+      "learning_rate": 0.00015068976960099862,
+      "loss": 1.1232,
+      "step": 17195
+    },
+    {
+      "epoch": 0.6621751684311838,
+      "grad_norm": 1.3553659915924072,
+      "learning_rate": 0.0001506637005553826,
+      "loss": 0.9969,
+      "step": 17200
+    },
+    {
+      "epoch": 0.6623676612127045,
+      "grad_norm": 1.3059508800506592,
+      "learning_rate": 0.00015063762687690684,
+      "loss": 1.0852,
+      "step": 17205
+    },
+    {
+      "epoch": 0.6625601539942252,
+      "grad_norm": 0.9797844290733337,
+      "learning_rate": 0.00015061154856795553,
+      "loss": 0.8927,
+      "step": 17210
+    },
+    {
+      "epoch": 0.6627526467757459,
+      "grad_norm": 1.2405691146850586,
+      "learning_rate": 0.00015058546563091337,
+      "loss": 1.1381,
+      "step": 17215
+    },
+    {
+      "epoch": 0.6629451395572666,
+      "grad_norm": 0.7226620316505432,
+      "learning_rate": 0.00015055937806816548,
+      "loss": 0.9773,
+      "step": 17220
+    },
+    {
+      "epoch": 0.6631376323387873,
+      "grad_norm": 1.302935004234314,
+      "learning_rate": 0.0001505332858820974,
+      "loss": 1.2386,
+      "step": 17225
+    },
+    {
+      "epoch": 0.663330125120308,
+      "grad_norm": 0.8981648683547974,
+      "learning_rate": 0.00015050718907509505,
+      "loss": 1.1499,
+      "step": 17230
+    },
+    {
+      "epoch": 0.6635226179018286,
+      "grad_norm": 1.6177557706832886,
+      "learning_rate": 0.00015048108764954487,
+      "loss": 1.0118,
+      "step": 17235
+    },
+    {
+      "epoch": 0.6637151106833494,
+      "grad_norm": 1.4030743837356567,
+      "learning_rate": 0.00015045498160783362,
+      "loss": 1.2892,
+      "step": 17240
+    },
+    {
+      "epoch": 0.6639076034648701,
+      "grad_norm": 1.3468968868255615,
+      "learning_rate": 0.00015042887095234852,
+      "loss": 1.2397,
+      "step": 17245
+    },
+    {
+      "epoch": 0.6641000962463908,
+      "grad_norm": 0.9706347584724426,
+      "learning_rate": 0.00015040275568547728,
+      "loss": 1.0251,
+      "step": 17250
+    },
+    {
+      "epoch": 0.6642925890279114,
+      "grad_norm": 1.623147964477539,
+      "learning_rate": 0.00015037663580960787,
+      "loss": 1.1651,
+      "step": 17255
+    },
+    {
+      "epoch": 0.6644850818094321,
+      "grad_norm": 0.9518052935600281,
+      "learning_rate": 0.00015035051132712883,
+      "loss": 1.1605,
+      "step": 17260
+    },
+    {
+      "epoch": 0.6646775745909529,
+      "grad_norm": 1.36576509475708,
+      "learning_rate": 0.00015032438224042908,
+      "loss": 1.1485,
+      "step": 17265
+    },
+    {
+      "epoch": 0.6648700673724736,
+      "grad_norm": 1.4218300580978394,
+      "learning_rate": 0.00015029824855189797,
+      "loss": 1.0527,
+      "step": 17270
+    },
+    {
+      "epoch": 0.6650625601539942,
+      "grad_norm": 1.573996663093567,
+      "learning_rate": 0.0001502721102639252,
+      "loss": 1.1692,
+      "step": 17275
+    },
+    {
+      "epoch": 0.6652550529355149,
+      "grad_norm": 1.1809152364730835,
+      "learning_rate": 0.00015024596737890097,
+      "loss": 1.0801,
+      "step": 17280
+    },
+    {
+      "epoch": 0.6654475457170356,
+      "grad_norm": 1.043346881866455,
+      "learning_rate": 0.00015021981989921587,
+      "loss": 1.181,
+      "step": 17285
+    },
+    {
+      "epoch": 0.6656400384985564,
+      "grad_norm": 0.9252155423164368,
+      "learning_rate": 0.00015019366782726093,
+      "loss": 1.0204,
+      "step": 17290
+    },
+    {
+      "epoch": 0.665832531280077,
+      "grad_norm": 1.4319888353347778,
+      "learning_rate": 0.00015016751116542757,
+      "loss": 1.2009,
+      "step": 17295
+    },
+    {
+      "epoch": 0.6660250240615977,
+      "grad_norm": 0.6749492287635803,
+      "learning_rate": 0.00015014134991610766,
+      "loss": 1.1157,
+      "step": 17300
+    },
+    {
+      "epoch": 0.6662175168431184,
+      "grad_norm": 0.9866890907287598,
+      "learning_rate": 0.0001501151840816934,
+      "loss": 1.1943,
+      "step": 17305
+    },
+    {
+      "epoch": 0.666410009624639,
+      "grad_norm": 1.4207334518432617,
+      "learning_rate": 0.00015008901366457756,
+      "loss": 1.1103,
+      "step": 17310
+    },
+    {
+      "epoch": 0.6666025024061598,
+      "grad_norm": 1.0321522951126099,
+      "learning_rate": 0.00015006283866715326,
+      "loss": 1.079,
+      "step": 17315
+    },
+    {
+      "epoch": 0.6667949951876805,
+      "grad_norm": 1.6033141613006592,
+      "learning_rate": 0.000150036659091814,
+      "loss": 0.982,
+      "step": 17320
+    },
+    {
+      "epoch": 0.6669874879692012,
+      "grad_norm": 1.503190279006958,
+      "learning_rate": 0.00015001047494095368,
+      "loss": 1.1371,
+      "step": 17325
+    },
+    {
+      "epoch": 0.6671799807507218,
+      "grad_norm": 1.2487331628799438,
+      "learning_rate": 0.00014998428621696677,
+      "loss": 1.1328,
+      "step": 17330
+    },
+    {
+      "epoch": 0.6673724735322425,
+      "grad_norm": 1.2876261472702026,
+      "learning_rate": 0.00014995809292224797,
+      "loss": 1.2034,
+      "step": 17335
+    },
+    {
+      "epoch": 0.6675649663137633,
+      "grad_norm": 1.0377410650253296,
+      "learning_rate": 0.0001499318950591925,
+      "loss": 1.2794,
+      "step": 17340
+    },
+    {
+      "epoch": 0.667757459095284,
+      "grad_norm": 2.4566397666931152,
+      "learning_rate": 0.00014990569263019602,
+      "loss": 1.1211,
+      "step": 17345
+    },
+    {
+      "epoch": 0.6679499518768046,
+      "grad_norm": 1.3069671392440796,
+      "learning_rate": 0.00014987948563765455,
+      "loss": 1.1101,
+      "step": 17350
+    },
+    {
+      "epoch": 0.6681424446583253,
+      "grad_norm": 1.0914125442504883,
+      "learning_rate": 0.0001498532740839645,
+      "loss": 1.0383,
+      "step": 17355
+    },
+    {
+      "epoch": 0.668334937439846,
+      "grad_norm": 1.1379315853118896,
+      "learning_rate": 0.00014982705797152285,
+      "loss": 1.0903,
+      "step": 17360
+    },
+    {
+      "epoch": 0.6685274302213667,
+      "grad_norm": 0.9188007712364197,
+      "learning_rate": 0.00014980083730272675,
+      "loss": 1.0696,
+      "step": 17365
+    },
+    {
+      "epoch": 0.6687199230028874,
+      "grad_norm": 1.2434134483337402,
+      "learning_rate": 0.00014977461207997403,
+      "loss": 1.2438,
+      "step": 17370
+    },
+    {
+      "epoch": 0.6689124157844081,
+      "grad_norm": 1.1543229818344116,
+      "learning_rate": 0.00014974838230566274,
+      "loss": 1.12,
+      "step": 17375
+    },
+    {
+      "epoch": 0.6691049085659287,
+      "grad_norm": 1.4789245128631592,
+      "learning_rate": 0.00014972214798219144,
+      "loss": 1.0437,
+      "step": 17380
+    },
+    {
+      "epoch": 0.6692974013474494,
+      "grad_norm": 1.4191787242889404,
+      "learning_rate": 0.0001496959091119591,
+      "loss": 1.1827,
+      "step": 17385
+    },
+    {
+      "epoch": 0.6694898941289702,
+      "grad_norm": 1.749631404876709,
+      "learning_rate": 0.00014966966569736508,
+      "loss": 1.0353,
+      "step": 17390
+    },
+    {
+      "epoch": 0.6696823869104909,
+      "grad_norm": 1.4120956659317017,
+      "learning_rate": 0.00014964341774080912,
+      "loss": 1.2257,
+      "step": 17395
+    },
+    {
+      "epoch": 0.6698748796920115,
+      "grad_norm": 1.6030794382095337,
+      "learning_rate": 0.00014961716524469152,
+      "loss": 1.0767,
+      "step": 17400
+    },
+    {
+      "epoch": 0.6700673724735322,
+      "grad_norm": 1.4263496398925781,
+      "learning_rate": 0.00014959090821141282,
+      "loss": 1.1188,
+      "step": 17405
+    },
+    {
+      "epoch": 0.670259865255053,
+      "grad_norm": 1.1514267921447754,
+      "learning_rate": 0.00014956464664337408,
+      "loss": 1.0731,
+      "step": 17410
+    },
+    {
+      "epoch": 0.6704523580365737,
+      "grad_norm": 1.5985325574874878,
+      "learning_rate": 0.00014953838054297672,
+      "loss": 1.1342,
+      "step": 17415
+    },
+    {
+      "epoch": 0.6706448508180943,
+      "grad_norm": 2.1868584156036377,
+      "learning_rate": 0.00014951210991262262,
+      "loss": 1.1169,
+      "step": 17420
+    },
+    {
+      "epoch": 0.670837343599615,
+      "grad_norm": 1.1203131675720215,
+      "learning_rate": 0.0001494858347547141,
+      "loss": 1.051,
+      "step": 17425
+    },
+    {
+      "epoch": 0.6710298363811357,
+      "grad_norm": 1.3077278137207031,
+      "learning_rate": 0.00014945955507165377,
+      "loss": 1.19,
+      "step": 17430
+    },
+    {
+      "epoch": 0.6712223291626565,
+      "grad_norm": 1.1149485111236572,
+      "learning_rate": 0.00014943327086584476,
+      "loss": 1.3471,
+      "step": 17435
+    },
+    {
+      "epoch": 0.6714148219441771,
+      "grad_norm": 1.7210713624954224,
+      "learning_rate": 0.00014940698213969063,
+      "loss": 1.0918,
+      "step": 17440
+    },
+    {
+      "epoch": 0.6716073147256978,
+      "grad_norm": 1.265023946762085,
+      "learning_rate": 0.00014938068889559526,
+      "loss": 1.0716,
+      "step": 17445
+    },
+    {
+      "epoch": 0.6717998075072185,
+      "grad_norm": 1.37469482421875,
+      "learning_rate": 0.00014935439113596298,
+      "loss": 1.1524,
+      "step": 17450
+    },
+    {
+      "epoch": 0.6719923002887391,
+      "grad_norm": 1.189141035079956,
+      "learning_rate": 0.0001493280888631986,
+      "loss": 1.1097,
+      "step": 17455
+    },
+    {
+      "epoch": 0.6721847930702599,
+      "grad_norm": 1.5825908184051514,
+      "learning_rate": 0.00014930178207970727,
+      "loss": 1.2842,
+      "step": 17460
+    },
+    {
+      "epoch": 0.6723772858517806,
+      "grad_norm": 1.1093425750732422,
+      "learning_rate": 0.00014927547078789452,
+      "loss": 1.0679,
+      "step": 17465
+    },
+    {
+      "epoch": 0.6725697786333013,
+      "grad_norm": 1.3306807279586792,
+      "learning_rate": 0.00014924915499016646,
+      "loss": 1.2877,
+      "step": 17470
+    },
+    {
+      "epoch": 0.6727622714148219,
+      "grad_norm": 1.9391852617263794,
+      "learning_rate": 0.00014922283468892935,
+      "loss": 1.1743,
+      "step": 17475
+    },
+    {
+      "epoch": 0.6729547641963426,
+      "grad_norm": 1.5213755369186401,
+      "learning_rate": 0.0001491965098865901,
+      "loss": 1.1793,
+      "step": 17480
+    },
+    {
+      "epoch": 0.6731472569778634,
+      "grad_norm": 1.6637414693832397,
+      "learning_rate": 0.00014917018058555593,
+      "loss": 1.1441,
+      "step": 17485
+    },
+    {
+      "epoch": 0.673339749759384,
+      "grad_norm": 1.7859970331192017,
+      "learning_rate": 0.00014914384678823447,
+      "loss": 1.1376,
+      "step": 17490
+    },
+    {
+      "epoch": 0.6735322425409047,
+      "grad_norm": 0.9251899719238281,
+      "learning_rate": 0.00014911750849703378,
+      "loss": 1.0523,
+      "step": 17495
+    },
+    {
+      "epoch": 0.6737247353224254,
+      "grad_norm": 2.6382827758789062,
+      "learning_rate": 0.00014909116571436228,
+      "loss": 1.311,
+      "step": 17500
+    },
+    {
+      "epoch": 0.673917228103946,
+      "grad_norm": 2.1472413539886475,
+      "learning_rate": 0.00014906481844262888,
+      "loss": 1.3515,
+      "step": 17505
+    },
+    {
+      "epoch": 0.6741097208854668,
+      "grad_norm": 1.6070085763931274,
+      "learning_rate": 0.0001490384666842429,
+      "loss": 1.121,
+      "step": 17510
+    },
+    {
+      "epoch": 0.6743022136669875,
+      "grad_norm": 1.637009620666504,
+      "learning_rate": 0.00014901211044161393,
+      "loss": 1.1249,
+      "step": 17515
+    },
+    {
+      "epoch": 0.6744947064485082,
+      "grad_norm": 1.4050389528274536,
+      "learning_rate": 0.00014898574971715218,
+      "loss": 1.1719,
+      "step": 17520
+    },
+    {
+      "epoch": 0.6746871992300288,
+      "grad_norm": 1.7863889932632446,
+      "learning_rate": 0.0001489593845132681,
+      "loss": 1.2576,
+      "step": 17525
+    },
+    {
+      "epoch": 0.6748796920115495,
+      "grad_norm": 1.149431586265564,
+      "learning_rate": 0.00014893301483237263,
+      "loss": 1.0863,
+      "step": 17530
+    },
+    {
+      "epoch": 0.6750721847930703,
+      "grad_norm": 1.4066704511642456,
+      "learning_rate": 0.0001489066406768771,
+      "loss": 1.1338,
+      "step": 17535
+    },
+    {
+      "epoch": 0.675264677574591,
+      "grad_norm": 1.2270228862762451,
+      "learning_rate": 0.00014888026204919327,
+      "loss": 1.1118,
+      "step": 17540
+    },
+    {
+      "epoch": 0.6754571703561116,
+      "grad_norm": 1.6182643175125122,
+      "learning_rate": 0.0001488538789517333,
+      "loss": 1.3269,
+      "step": 17545
+    },
+    {
+      "epoch": 0.6756496631376323,
+      "grad_norm": 2.3642048835754395,
+      "learning_rate": 0.0001488274913869097,
+      "loss": 1.439,
+      "step": 17550
+    },
+    {
+      "epoch": 0.6758421559191531,
+      "grad_norm": 1.8097171783447266,
+      "learning_rate": 0.00014880109935713548,
+      "loss": 1.093,
+      "step": 17555
+    },
+    {
+      "epoch": 0.6760346487006738,
+      "grad_norm": 0.8650147914886475,
+      "learning_rate": 0.00014877470286482397,
+      "loss": 1.0413,
+      "step": 17560
+    },
+    {
+      "epoch": 0.6762271414821944,
+      "grad_norm": 1.2217522859573364,
+      "learning_rate": 0.00014874830191238903,
+      "loss": 1.1818,
+      "step": 17565
+    },
+    {
+      "epoch": 0.6764196342637151,
+      "grad_norm": 1.1500258445739746,
+      "learning_rate": 0.00014872189650224477,
+      "loss": 1.0607,
+      "step": 17570
+    },
+    {
+      "epoch": 0.6766121270452358,
+      "grad_norm": 1.1867146492004395,
+      "learning_rate": 0.00014869548663680584,
+      "loss": 0.9716,
+      "step": 17575
+    },
+    {
+      "epoch": 0.6768046198267565,
+      "grad_norm": 1.0046483278274536,
+      "learning_rate": 0.00014866907231848723,
+      "loss": 1.1875,
+      "step": 17580
+    },
+    {
+      "epoch": 0.6769971126082772,
+      "grad_norm": 2.1072323322296143,
+      "learning_rate": 0.00014864265354970436,
+      "loss": 1.194,
+      "step": 17585
+    },
+    {
+      "epoch": 0.6771896053897979,
+      "grad_norm": 1.4290494918823242,
+      "learning_rate": 0.00014861623033287307,
+      "loss": 1.2389,
+      "step": 17590
+    },
+    {
+      "epoch": 0.6773820981713186,
+      "grad_norm": 0.8890597820281982,
+      "learning_rate": 0.00014858980267040957,
+      "loss": 0.9362,
+      "step": 17595
+    },
+    {
+      "epoch": 0.6775745909528392,
+      "grad_norm": 0.9515128135681152,
+      "learning_rate": 0.00014856337056473045,
+      "loss": 1.039,
+      "step": 17600
+    },
+    {
+      "epoch": 0.67776708373436,
+      "grad_norm": 1.540008544921875,
+      "learning_rate": 0.00014853693401825283,
+      "loss": 1.1778,
+      "step": 17605
+    },
+    {
+      "epoch": 0.6779595765158807,
+      "grad_norm": 1.0766023397445679,
+      "learning_rate": 0.00014851049303339414,
+      "loss": 0.9362,
+      "step": 17610
+    },
+    {
+      "epoch": 0.6781520692974013,
+      "grad_norm": 1.854201078414917,
+      "learning_rate": 0.00014848404761257217,
+      "loss": 0.9427,
+      "step": 17615
+    },
+    {
+      "epoch": 0.678344562078922,
+      "grad_norm": 2.292722463607788,
+      "learning_rate": 0.00014845759775820527,
+      "loss": 1.0835,
+      "step": 17620
+    },
+    {
+      "epoch": 0.6785370548604427,
+      "grad_norm": 1.768997311592102,
+      "learning_rate": 0.00014843114347271204,
+      "loss": 1.0976,
+      "step": 17625
+    },
+    {
+      "epoch": 0.6787295476419635,
+      "grad_norm": 2.223881721496582,
+      "learning_rate": 0.00014840468475851154,
+      "loss": 1.1417,
+      "step": 17630
+    },
+    {
+      "epoch": 0.6789220404234841,
+      "grad_norm": 1.1589646339416504,
+      "learning_rate": 0.0001483782216180233,
+      "loss": 1.2536,
+      "step": 17635
+    },
+    {
+      "epoch": 0.6791145332050048,
+      "grad_norm": 1.6478285789489746,
+      "learning_rate": 0.00014835175405366718,
+      "loss": 1.1534,
+      "step": 17640
+    },
+    {
+      "epoch": 0.6793070259865255,
+      "grad_norm": 1.6837091445922852,
+      "learning_rate": 0.00014832528206786344,
+      "loss": 1.3415,
+      "step": 17645
+    },
+    {
+      "epoch": 0.6794995187680462,
+      "grad_norm": 1.6697105169296265,
+      "learning_rate": 0.00014829880566303273,
+      "loss": 1.0241,
+      "step": 17650
+    },
+    {
+      "epoch": 0.6796920115495669,
+      "grad_norm": 1.08551025390625,
+      "learning_rate": 0.00014827232484159624,
+      "loss": 1.2322,
+      "step": 17655
+    },
+    {
+      "epoch": 0.6798845043310876,
+      "grad_norm": 1.9399616718292236,
+      "learning_rate": 0.00014824583960597543,
+      "loss": 1.393,
+      "step": 17660
+    },
+    {
+      "epoch": 0.6800769971126083,
+      "grad_norm": 1.0628485679626465,
+      "learning_rate": 0.00014821934995859216,
+      "loss": 1.2078,
+      "step": 17665
+    },
+    {
+      "epoch": 0.6802694898941289,
+      "grad_norm": 0.9613144397735596,
+      "learning_rate": 0.00014819285590186875,
+      "loss": 1.1234,
+      "step": 17670
+    },
+    {
+      "epoch": 0.6804619826756496,
+      "grad_norm": 0.9686816930770874,
+      "learning_rate": 0.00014816635743822795,
+      "loss": 1.1959,
+      "step": 17675
+    },
+    {
+      "epoch": 0.6806544754571704,
+      "grad_norm": 1.4415709972381592,
+      "learning_rate": 0.00014813985457009282,
+      "loss": 1.0775,
+      "step": 17680
+    },
+    {
+      "epoch": 0.6808469682386911,
+      "grad_norm": 1.5800002813339233,
+      "learning_rate": 0.00014811334729988688,
+      "loss": 1.0802,
+      "step": 17685
+    },
+    {
+      "epoch": 0.6810394610202117,
+      "grad_norm": 1.1061028242111206,
+      "learning_rate": 0.0001480868356300341,
+      "loss": 1.0415,
+      "step": 17690
+    },
+    {
+      "epoch": 0.6812319538017324,
+      "grad_norm": 2.3262946605682373,
+      "learning_rate": 0.00014806031956295868,
+      "loss": 1.2431,
+      "step": 17695
+    },
+    {
+      "epoch": 0.6814244465832531,
+      "grad_norm": 1.6517562866210938,
+      "learning_rate": 0.00014803379910108543,
+      "loss": 1.1792,
+      "step": 17700
+    },
+    {
+      "epoch": 0.6816169393647739,
+      "grad_norm": 1.3823506832122803,
+      "learning_rate": 0.00014800727424683948,
+      "loss": 1.1293,
+      "step": 17705
+    },
+    {
+      "epoch": 0.6818094321462945,
+      "grad_norm": 1.5448585748672485,
+      "learning_rate": 0.00014798074500264627,
+      "loss": 1.2126,
+      "step": 17710
+    },
+    {
+      "epoch": 0.6820019249278152,
+      "grad_norm": 1.2395973205566406,
+      "learning_rate": 0.0001479542113709318,
+      "loss": 1.3002,
+      "step": 17715
+    },
+    {
+      "epoch": 0.6821944177093359,
+      "grad_norm": 1.8366637229919434,
+      "learning_rate": 0.00014792767335412233,
+      "loss": 1.1798,
+      "step": 17720
+    },
+    {
+      "epoch": 0.6823869104908566,
+      "grad_norm": 1.3830804824829102,
+      "learning_rate": 0.00014790113095464465,
+      "loss": 1.3001,
+      "step": 17725
+    },
+    {
+      "epoch": 0.6825794032723773,
+      "grad_norm": 10.001764297485352,
+      "learning_rate": 0.0001478745841749259,
+      "loss": 1.1643,
+      "step": 17730
+    },
+    {
+      "epoch": 0.682771896053898,
+      "grad_norm": 1.0113561153411865,
+      "learning_rate": 0.00014784803301739352,
+      "loss": 1.1725,
+      "step": 17735
+    },
+    {
+      "epoch": 0.6829643888354187,
+      "grad_norm": 2.7240827083587646,
+      "learning_rate": 0.00014782147748447554,
+      "loss": 1.2348,
+      "step": 17740
+    },
+    {
+      "epoch": 0.6831568816169393,
+      "grad_norm": 1.0802150964736938,
+      "learning_rate": 0.00014779491757860015,
+      "loss": 1.3556,
+      "step": 17745
+    },
+    {
+      "epoch": 0.6833493743984601,
+      "grad_norm": 1.6339032649993896,
+      "learning_rate": 0.00014776835330219623,
+      "loss": 0.9967,
+      "step": 17750
+    },
+    {
+      "epoch": 0.6835418671799808,
+      "grad_norm": 1.6983892917633057,
+      "learning_rate": 0.0001477417846576928,
+      "loss": 1.041,
+      "step": 17755
+    },
+    {
+      "epoch": 0.6837343599615014,
+      "grad_norm": 1.6230486631393433,
+      "learning_rate": 0.00014771521164751942,
+      "loss": 1.2298,
+      "step": 17760
+    },
+    {
+      "epoch": 0.6839268527430221,
+      "grad_norm": 1.1079175472259521,
+      "learning_rate": 0.00014768863427410604,
+      "loss": 1.214,
+      "step": 17765
+    },
+    {
+      "epoch": 0.6841193455245428,
+      "grad_norm": 1.1601203680038452,
+      "learning_rate": 0.00014766205253988294,
+      "loss": 1.2399,
+      "step": 17770
+    },
+    {
+      "epoch": 0.6843118383060636,
+      "grad_norm": 2.2776849269866943,
+      "learning_rate": 0.00014763546644728088,
+      "loss": 1.0071,
+      "step": 17775
+    },
+    {
+      "epoch": 0.6845043310875842,
+      "grad_norm": 1.362021565437317,
+      "learning_rate": 0.00014760887599873094,
+      "loss": 1.1233,
+      "step": 17780
+    },
+    {
+      "epoch": 0.6846968238691049,
+      "grad_norm": 1.933518409729004,
+      "learning_rate": 0.00014758228119666472,
+      "loss": 1.0854,
+      "step": 17785
+    },
+    {
+      "epoch": 0.6848893166506256,
+      "grad_norm": 1.148533582687378,
+      "learning_rate": 0.00014755568204351407,
+      "loss": 1.0694,
+      "step": 17790
+    },
+    {
+      "epoch": 0.6850818094321462,
+      "grad_norm": 1.2880831956863403,
+      "learning_rate": 0.0001475290785417113,
+      "loss": 1.0814,
+      "step": 17795
+    },
+    {
+      "epoch": 0.685274302213667,
+      "grad_norm": 1.5790437459945679,
+      "learning_rate": 0.0001475024706936892,
+      "loss": 1.0467,
+      "step": 17800
+    },
+    {
+      "epoch": 0.6854667949951877,
+      "grad_norm": 1.636828899383545,
+      "learning_rate": 0.0001474758585018808,
+      "loss": 1.3419,
+      "step": 17805
+    },
+    {
+      "epoch": 0.6856592877767084,
+      "grad_norm": 1.0403766632080078,
+      "learning_rate": 0.00014744924196871963,
+      "loss": 1.1468,
+      "step": 17810
+    },
+    {
+      "epoch": 0.685851780558229,
+      "grad_norm": 1.1266472339630127,
+      "learning_rate": 0.0001474226210966396,
+      "loss": 1.2723,
+      "step": 17815
+    },
+    {
+      "epoch": 0.6860442733397497,
+      "grad_norm": 1.352543830871582,
+      "learning_rate": 0.00014739599588807506,
+      "loss": 1.1345,
+      "step": 17820
+    },
+    {
+      "epoch": 0.6862367661212705,
+      "grad_norm": 1.674023151397705,
+      "learning_rate": 0.00014736936634546062,
+      "loss": 1.2522,
+      "step": 17825
+    },
+    {
+      "epoch": 0.6864292589027912,
+      "grad_norm": 1.3684656620025635,
+      "learning_rate": 0.00014734273247123144,
+      "loss": 1.1169,
+      "step": 17830
+    },
+    {
+      "epoch": 0.6866217516843118,
+      "grad_norm": 1.917075514793396,
+      "learning_rate": 0.00014731609426782297,
+      "loss": 1.2523,
+      "step": 17835
+    },
+    {
+      "epoch": 0.6868142444658325,
+      "grad_norm": 1.5463966131210327,
+      "learning_rate": 0.00014728945173767116,
+      "loss": 0.9929,
+      "step": 17840
+    },
+    {
+      "epoch": 0.6870067372473532,
+      "grad_norm": 1.7427698373794556,
+      "learning_rate": 0.00014726280488321222,
+      "loss": 1.22,
+      "step": 17845
+    },
+    {
+      "epoch": 0.687199230028874,
+      "grad_norm": 1.8021422624588013,
+      "learning_rate": 0.0001472361537068829,
+      "loss": 1.0429,
+      "step": 17850
+    },
+    {
+      "epoch": 0.6873917228103946,
+      "grad_norm": 1.571053147315979,
+      "learning_rate": 0.0001472094982111202,
+      "loss": 1.17,
+      "step": 17855
+    },
+    {
+      "epoch": 0.6875842155919153,
+      "grad_norm": 1.3607596158981323,
+      "learning_rate": 0.00014718283839836166,
+      "loss": 1.0644,
+      "step": 17860
+    },
+    {
+      "epoch": 0.687776708373436,
+      "grad_norm": 0.9396845102310181,
+      "learning_rate": 0.00014715617427104504,
+      "loss": 1.0807,
+      "step": 17865
+    },
+    {
+      "epoch": 0.6879692011549567,
+      "grad_norm": 1.605432152748108,
+      "learning_rate": 0.00014712950583160872,
+      "loss": 1.0641,
+      "step": 17870
+    },
+    {
+      "epoch": 0.6881616939364774,
+      "grad_norm": 1.4847965240478516,
+      "learning_rate": 0.0001471081679769722,
+      "loss": 1.1625,
+      "step": 17875
+    },
+    {
+      "epoch": 0.6883541867179981,
+      "grad_norm": 1.930336594581604,
+      "learning_rate": 0.00014708149178186593,
+      "loss": 1.3346,
+      "step": 17880
+    },
+    {
+      "epoch": 0.6885466794995188,
+      "grad_norm": 1.7398570775985718,
+      "learning_rate": 0.00014705481128146917,
+      "loss": 1.2316,
+      "step": 17885
+    },
+    {
+      "epoch": 0.6887391722810394,
+      "grad_norm": 1.5817015171051025,
+      "learning_rate": 0.00014702812647822162,
+      "loss": 1.0292,
+      "step": 17890
+    },
+    {
+      "epoch": 0.6889316650625602,
+      "grad_norm": 3.2520430088043213,
+      "learning_rate": 0.00014700143737456342,
+      "loss": 1.088,
+      "step": 17895
+    },
+    {
+      "epoch": 0.6891241578440809,
+      "grad_norm": 2.165456533432007,
+      "learning_rate": 0.00014697474397293517,
+      "loss": 0.9452,
+      "step": 17900
+    },
+    {
+      "epoch": 0.6893166506256015,
+      "grad_norm": 0.9637191295623779,
+      "learning_rate": 0.00014694804627577771,
+      "loss": 1.266,
+      "step": 17905
+    },
+    {
+      "epoch": 0.6895091434071222,
+      "grad_norm": 1.9606934785842896,
+      "learning_rate": 0.00014692134428553248,
+      "loss": 1.0773,
+      "step": 17910
+    },
+    {
+      "epoch": 0.6897016361886429,
+      "grad_norm": 1.1911338567733765,
+      "learning_rate": 0.0001468946380046411,
+      "loss": 1.1359,
+      "step": 17915
+    },
+    {
+      "epoch": 0.6898941289701637,
+      "grad_norm": 1.3913235664367676,
+      "learning_rate": 0.00014686792743554575,
+      "loss": 1.3053,
+      "step": 17920
+    },
+    {
+      "epoch": 0.6900866217516843,
+      "grad_norm": 1.2314075231552124,
+      "learning_rate": 0.00014684121258068888,
+      "loss": 1.0624,
+      "step": 17925
+    },
+    {
+      "epoch": 0.690279114533205,
+      "grad_norm": 2.1499176025390625,
+      "learning_rate": 0.00014681449344251338,
+      "loss": 1.2147,
+      "step": 17930
+    },
+    {
+      "epoch": 0.6904716073147257,
+      "grad_norm": 1.6417664289474487,
+      "learning_rate": 0.00014678777002346264,
+      "loss": 1.1139,
+      "step": 17935
+    },
+    {
+      "epoch": 0.6906641000962463,
+      "grad_norm": 1.181154727935791,
+      "learning_rate": 0.00014676104232598026,
+      "loss": 1.0503,
+      "step": 17940
+    },
+    {
+      "epoch": 0.6908565928777671,
+      "grad_norm": 1.7786331176757812,
+      "learning_rate": 0.00014673431035251027,
+      "loss": 1.05,
+      "step": 17945
+    },
+    {
+      "epoch": 0.6910490856592878,
+      "grad_norm": 0.948625922203064,
+      "learning_rate": 0.00014670757410549724,
+      "loss": 1.0888,
+      "step": 17950
+    },
+    {
+      "epoch": 0.6912415784408085,
+      "grad_norm": 1.9812164306640625,
+      "learning_rate": 0.00014668083358738597,
+      "loss": 1.1467,
+      "step": 17955
+    },
+    {
+      "epoch": 0.6914340712223291,
+      "grad_norm": 0.9091313481330872,
+      "learning_rate": 0.0001466540888006217,
+      "loss": 1.1226,
+      "step": 17960
+    },
+    {
+      "epoch": 0.6916265640038498,
+      "grad_norm": 2.100114583969116,
+      "learning_rate": 0.00014662733974765005,
+      "loss": 1.1233,
+      "step": 17965
+    },
+    {
+      "epoch": 0.6918190567853706,
+      "grad_norm": 2.0999033451080322,
+      "learning_rate": 0.00014660058643091702,
+      "loss": 1.086,
+      "step": 17970
+    },
+    {
+      "epoch": 0.6920115495668913,
+      "grad_norm": 1.543411374092102,
+      "learning_rate": 0.0001465738288528691,
+      "loss": 1.218,
+      "step": 17975
+    },
+    {
+      "epoch": 0.6922040423484119,
+      "grad_norm": 2.6429097652435303,
+      "learning_rate": 0.00014654706701595305,
+      "loss": 1.1425,
+      "step": 17980
+    },
+    {
+      "epoch": 0.6923965351299326,
+      "grad_norm": 1.258535385131836,
+      "learning_rate": 0.00014652030092261606,
+      "loss": 1.124,
+      "step": 17985
+    },
+    {
+      "epoch": 0.6925890279114533,
+      "grad_norm": 0.9203128814697266,
+      "learning_rate": 0.00014649353057530573,
+      "loss": 1.0035,
+      "step": 17990
+    },
+    {
+      "epoch": 0.692781520692974,
+      "grad_norm": 1.7482789754867554,
+      "learning_rate": 0.00014646675597647003,
+      "loss": 1.2393,
+      "step": 17995
+    },
+    {
+      "epoch": 0.6929740134744947,
+      "grad_norm": 1.3026279211044312,
+      "learning_rate": 0.0001464399771285573,
+      "loss": 1.294,
+      "step": 18000
+    },
+    {
+      "epoch": 0.6931665062560154,
+      "grad_norm": 1.5518649816513062,
+      "learning_rate": 0.00014641319403401628,
+      "loss": 1.2397,
+      "step": 18005
+    },
+    {
+      "epoch": 0.6933589990375361,
+      "grad_norm": 1.3904852867126465,
+      "learning_rate": 0.00014638640669529615,
+      "loss": 1.079,
+      "step": 18010
+    },
+    {
+      "epoch": 0.6935514918190567,
+      "grad_norm": 0.7677931189537048,
+      "learning_rate": 0.0001463596151148464,
+      "loss": 1.1485,
+      "step": 18015
+    },
+    {
+      "epoch": 0.6937439846005775,
+      "grad_norm": 1.1935845613479614,
+      "learning_rate": 0.00014633281929511696,
+      "loss": 1.167,
+      "step": 18020
+    },
+    {
+      "epoch": 0.6939364773820982,
+      "grad_norm": 1.8612521886825562,
+      "learning_rate": 0.00014630601923855814,
+      "loss": 1.2335,
+      "step": 18025
+    },
+    {
+      "epoch": 0.6941289701636189,
+      "grad_norm": 1.9979881048202515,
+      "learning_rate": 0.00014627921494762055,
+      "loss": 1.0421,
+      "step": 18030
+    },
+    {
+      "epoch": 0.6943214629451395,
+      "grad_norm": 1.9426400661468506,
+      "learning_rate": 0.00014625240642475538,
+      "loss": 1.0918,
+      "step": 18035
+    },
+    {
+      "epoch": 0.6945139557266603,
+      "grad_norm": 0.9990954399108887,
+      "learning_rate": 0.000146225593672414,
+      "loss": 1.3456,
+      "step": 18040
+    },
+    {
+      "epoch": 0.694706448508181,
+      "grad_norm": 2.187206745147705,
+      "learning_rate": 0.00014619877669304834,
+      "loss": 1.0926,
+      "step": 18045
+    },
+    {
+      "epoch": 0.6948989412897016,
+      "grad_norm": 1.5417639017105103,
+      "learning_rate": 0.00014617195548911053,
+      "loss": 1.2796,
+      "step": 18050
+    },
+    {
+      "epoch": 0.6950914340712223,
+      "grad_norm": 1.476150631904602,
+      "learning_rate": 0.0001461451300630533,
+      "loss": 1.1623,
+      "step": 18055
+    },
+    {
+      "epoch": 0.695283926852743,
+      "grad_norm": 1.6524615287780762,
+      "learning_rate": 0.0001461183004173296,
+      "loss": 1.0976,
+      "step": 18060
+    },
+    {
+      "epoch": 0.6954764196342638,
+      "grad_norm": 1.4800169467926025,
+      "learning_rate": 0.0001460914665543928,
+      "loss": 1.2641,
+      "step": 18065
+    },
+    {
+      "epoch": 0.6956689124157844,
+      "grad_norm": 1.2046303749084473,
+      "learning_rate": 0.00014606462847669674,
+      "loss": 1.0037,
+      "step": 18070
+    },
+    {
+      "epoch": 0.6958614051973051,
+      "grad_norm": 1.3457711935043335,
+      "learning_rate": 0.00014603778618669556,
+      "loss": 1.1599,
+      "step": 18075
+    },
+    {
+      "epoch": 0.6960538979788258,
+      "grad_norm": 1.8690896034240723,
+      "learning_rate": 0.0001460109396868438,
+      "loss": 1.3016,
+      "step": 18080
+    },
+    {
+      "epoch": 0.6962463907603464,
+      "grad_norm": 0.8788353204727173,
+      "learning_rate": 0.00014598408897959639,
+      "loss": 1.0261,
+      "step": 18085
+    },
+    {
+      "epoch": 0.6964388835418672,
+      "grad_norm": 1.064239501953125,
+      "learning_rate": 0.00014595723406740868,
+      "loss": 1.1159,
+      "step": 18090
+    },
+    {
+      "epoch": 0.6966313763233879,
+      "grad_norm": 0.9102209210395813,
+      "learning_rate": 0.00014593037495273635,
+      "loss": 1.1263,
+      "step": 18095
+    },
+    {
+      "epoch": 0.6968238691049086,
+      "grad_norm": 1.4841855764389038,
+      "learning_rate": 0.00014590351163803545,
+      "loss": 1.0526,
+      "step": 18100
+    },
+    {
+      "epoch": 0.6970163618864292,
+      "grad_norm": 2.282543182373047,
+      "learning_rate": 0.00014587664412576254,
+      "loss": 1.0876,
+      "step": 18105
+    },
+    {
+      "epoch": 0.6972088546679499,
+      "grad_norm": 1.149782657623291,
+      "learning_rate": 0.0001458497724183744,
+      "loss": 1.2092,
+      "step": 18110
+    },
+    {
+      "epoch": 0.6974013474494707,
+      "grad_norm": 1.6531153917312622,
+      "learning_rate": 0.0001458228965183283,
+      "loss": 1.2421,
+      "step": 18115
+    },
+    {
+      "epoch": 0.6975938402309914,
+      "grad_norm": 2.376281976699829,
+      "learning_rate": 0.00014579601642808192,
+      "loss": 1.2179,
+      "step": 18120
+    },
+    {
+      "epoch": 0.697786333012512,
+      "grad_norm": 1.9077723026275635,
+      "learning_rate": 0.0001457691321500932,
+      "loss": 1.1962,
+      "step": 18125
+    },
+    {
+      "epoch": 0.6979788257940327,
+      "grad_norm": 1.3130842447280884,
+      "learning_rate": 0.00014574224368682048,
+      "loss": 1.3169,
+      "step": 18130
+    },
+    {
+      "epoch": 0.6981713185755534,
+      "grad_norm": 1.0211979150772095,
+      "learning_rate": 0.00014571535104072262,
+      "loss": 1.0256,
+      "step": 18135
+    },
+    {
+      "epoch": 0.6983638113570741,
+      "grad_norm": 1.7479397058486938,
+      "learning_rate": 0.00014568845421425875,
+      "loss": 1.0906,
+      "step": 18140
+    },
+    {
+      "epoch": 0.6985563041385948,
+      "grad_norm": 1.3305407762527466,
+      "learning_rate": 0.00014566155320988838,
+      "loss": 1.206,
+      "step": 18145
+    },
+    {
+      "epoch": 0.6987487969201155,
+      "grad_norm": 1.2185992002487183,
+      "learning_rate": 0.00014563464803007145,
+      "loss": 1.2765,
+      "step": 18150
+    },
+    {
+      "epoch": 0.6989412897016362,
+      "grad_norm": 1.3256112337112427,
+      "learning_rate": 0.00014560773867726827,
+      "loss": 1.0899,
+      "step": 18155
+    },
+    {
+      "epoch": 0.6991337824831568,
+      "grad_norm": 1.9090956449508667,
+      "learning_rate": 0.0001455808251539395,
+      "loss": 1.1944,
+      "step": 18160
+    },
+    {
+      "epoch": 0.6993262752646776,
+      "grad_norm": 1.078116774559021,
+      "learning_rate": 0.00014555390746254622,
+      "loss": 1.1393,
+      "step": 18165
+    },
+    {
+      "epoch": 0.6995187680461983,
+      "grad_norm": 1.21144437789917,
+      "learning_rate": 0.00014552698560554988,
+      "loss": 1.0835,
+      "step": 18170
+    },
+    {
+      "epoch": 0.699711260827719,
+      "grad_norm": 1.4013081789016724,
+      "learning_rate": 0.00014550005958541227,
+      "loss": 1.0785,
+      "step": 18175
+    },
+    {
+      "epoch": 0.6999037536092396,
+      "grad_norm": 1.102122187614441,
+      "learning_rate": 0.00014547312940459562,
+      "loss": 1.0839,
+      "step": 18180
+    },
+    {
+      "epoch": 0.7000962463907604,
+      "grad_norm": 1.602994680404663,
+      "learning_rate": 0.00014544619506556256,
+      "loss": 1.2608,
+      "step": 18185
+    },
+    {
+      "epoch": 0.7002887391722811,
+      "grad_norm": 2.8694801330566406,
+      "learning_rate": 0.000145419256570776,
+      "loss": 1.3161,
+      "step": 18190
+    },
+    {
+      "epoch": 0.7004812319538017,
+      "grad_norm": 1.5687551498413086,
+      "learning_rate": 0.00014539231392269927,
+      "loss": 1.0668,
+      "step": 18195
+    },
+    {
+      "epoch": 0.7006737247353224,
+      "grad_norm": 1.1013094186782837,
+      "learning_rate": 0.00014536536712379618,
+      "loss": 1.0829,
+      "step": 18200
+    },
+    {
+      "epoch": 0.7008662175168431,
+      "grad_norm": 1.4294344186782837,
+      "learning_rate": 0.00014533841617653075,
+      "loss": 1.0003,
+      "step": 18205
+    },
+    {
+      "epoch": 0.7010587102983639,
+      "grad_norm": 1.168997049331665,
+      "learning_rate": 0.0001453114610833675,
+      "loss": 1.2252,
+      "step": 18210
+    },
+    {
+      "epoch": 0.7012512030798845,
+      "grad_norm": 1.21929132938385,
+      "learning_rate": 0.0001452845018467713,
+      "loss": 1.102,
+      "step": 18215
+    },
+    {
+      "epoch": 0.7014436958614052,
+      "grad_norm": 1.0682016611099243,
+      "learning_rate": 0.00014525753846920738,
+      "loss": 1.0219,
+      "step": 18220
+    },
+    {
+      "epoch": 0.7016361886429259,
+      "grad_norm": 1.210161566734314,
+      "learning_rate": 0.00014523057095314142,
+      "loss": 1.0666,
+      "step": 18225
+    },
+    {
+      "epoch": 0.7018286814244465,
+      "grad_norm": 0.9966996312141418,
+      "learning_rate": 0.0001452035993010393,
+      "loss": 1.1343,
+      "step": 18230
+    },
+    {
+      "epoch": 0.7020211742059673,
+      "grad_norm": 1.2477959394454956,
+      "learning_rate": 0.00014517662351536752,
+      "loss": 1.2147,
+      "step": 18235
+    },
+    {
+      "epoch": 0.702213666987488,
+      "grad_norm": 1.8020172119140625,
+      "learning_rate": 0.00014514964359859276,
+      "loss": 1.1945,
+      "step": 18240
+    },
+    {
+      "epoch": 0.7024061597690087,
+      "grad_norm": 1.0535303354263306,
+      "learning_rate": 0.0001451226595531822,
+      "loss": 1.0792,
+      "step": 18245
+    },
+    {
+      "epoch": 0.7025986525505293,
+      "grad_norm": 1.913590431213379,
+      "learning_rate": 0.0001450956713816033,
+      "loss": 1.1344,
+      "step": 18250
+    },
+    {
+      "epoch": 0.70279114533205,
+      "grad_norm": 0.998621940612793,
+      "learning_rate": 0.00014506867908632403,
+      "loss": 1.1139,
+      "step": 18255
+    },
+    {
+      "epoch": 0.7029836381135708,
+      "grad_norm": 1.8913546800613403,
+      "learning_rate": 0.0001450416826698126,
+      "loss": 1.0621,
+      "step": 18260
+    },
+    {
+      "epoch": 0.7031761308950915,
+      "grad_norm": 1.0329716205596924,
+      "learning_rate": 0.00014501468213453763,
+      "loss": 1.2732,
+      "step": 18265
+    },
+    {
+      "epoch": 0.7033686236766121,
+      "grad_norm": 0.9243387579917908,
+      "learning_rate": 0.0001449876774829682,
+      "loss": 1.2272,
+      "step": 18270
+    },
+    {
+      "epoch": 0.7035611164581328,
+      "grad_norm": 1.6289262771606445,
+      "learning_rate": 0.0001449606687175737,
+      "loss": 1.0912,
+      "step": 18275
+    },
+    {
+      "epoch": 0.7037536092396535,
+      "grad_norm": 2.005293607711792,
+      "learning_rate": 0.00014493365584082384,
+      "loss": 1.018,
+      "step": 18280
+    },
+    {
+      "epoch": 0.7039461020211742,
+      "grad_norm": 1.2743504047393799,
+      "learning_rate": 0.00014490663885518881,
+      "loss": 1.0026,
+      "step": 18285
+    },
+    {
+      "epoch": 0.7041385948026949,
+      "grad_norm": 1.4915635585784912,
+      "learning_rate": 0.00014487961776313922,
+      "loss": 1.0489,
+      "step": 18290
+    },
+    {
+      "epoch": 0.7043310875842156,
+      "grad_norm": 0.9605044722557068,
+      "learning_rate": 0.00014485259256714577,
+      "loss": 1.1053,
+      "step": 18295
+    },
+    {
+      "epoch": 0.7045235803657363,
+      "grad_norm": 1.8121784925460815,
+      "learning_rate": 0.0001448255632696799,
+      "loss": 1.1862,
+      "step": 18300
+    },
+    {
+      "epoch": 0.7047160731472569,
+      "grad_norm": 1.2540571689605713,
+      "learning_rate": 0.00014479852987321322,
+      "loss": 1.1361,
+      "step": 18305
+    },
+    {
+      "epoch": 0.7049085659287777,
+      "grad_norm": 1.4160270690917969,
+      "learning_rate": 0.00014477149238021776,
+      "loss": 1.0917,
+      "step": 18310
+    },
+    {
+      "epoch": 0.7051010587102984,
+      "grad_norm": 1.4298075437545776,
+      "learning_rate": 0.0001447444507931659,
+      "loss": 1.1407,
+      "step": 18315
+    },
+    {
+      "epoch": 0.705293551491819,
+      "grad_norm": 1.0214334726333618,
+      "learning_rate": 0.00014471740511453037,
+      "loss": 1.0714,
+      "step": 18320
+    },
+    {
+      "epoch": 0.7054860442733397,
+      "grad_norm": 1.6246428489685059,
+      "learning_rate": 0.00014469035534678444,
+      "loss": 1.258,
+      "step": 18325
+    },
+    {
+      "epoch": 0.7056785370548604,
+      "grad_norm": 1.5467473268508911,
+      "learning_rate": 0.0001446633014924015,
+      "loss": 1.1811,
+      "step": 18330
+    },
+    {
+      "epoch": 0.7058710298363812,
+      "grad_norm": 2.038041114807129,
+      "learning_rate": 0.00014463624355385557,
+      "loss": 1.1339,
+      "step": 18335
+    },
+    {
+      "epoch": 0.7060635226179018,
+      "grad_norm": 1.5328725576400757,
+      "learning_rate": 0.0001446091815336208,
+      "loss": 1.2261,
+      "step": 18340
+    },
+    {
+      "epoch": 0.7062560153994225,
+      "grad_norm": 0.9550712704658508,
+      "learning_rate": 0.0001445821154341719,
+      "loss": 1.0973,
+      "step": 18345
+    },
+    {
+      "epoch": 0.7064485081809432,
+      "grad_norm": 1.4610974788665771,
+      "learning_rate": 0.0001445550452579839,
+      "loss": 1.2341,
+      "step": 18350
+    },
+    {
+      "epoch": 0.706641000962464,
+      "grad_norm": 1.9539941549301147,
+      "learning_rate": 0.00014452797100753212,
+      "loss": 1.1115,
+      "step": 18355
+    },
+    {
+      "epoch": 0.7068334937439846,
+      "grad_norm": 1.136670708656311,
+      "learning_rate": 0.0001445008926852924,
+      "loss": 1.1883,
+      "step": 18360
+    },
+    {
+      "epoch": 0.7070259865255053,
+      "grad_norm": 1.2136088609695435,
+      "learning_rate": 0.00014447381029374082,
+      "loss": 1.1384,
+      "step": 18365
+    },
+    {
+      "epoch": 0.707218479307026,
+      "grad_norm": 1.3836339712142944,
+      "learning_rate": 0.00014444672383535388,
+      "loss": 1.2371,
+      "step": 18370
+    },
+    {
+      "epoch": 0.7074109720885466,
+      "grad_norm": 1.6226662397384644,
+      "learning_rate": 0.00014441963331260848,
+      "loss": 1.3057,
+      "step": 18375
+    },
+    {
+      "epoch": 0.7076034648700674,
+      "grad_norm": 1.249576449394226,
+      "learning_rate": 0.0001443925387279819,
+      "loss": 1.0849,
+      "step": 18380
+    },
+    {
+      "epoch": 0.7077959576515881,
+      "grad_norm": 1.9330114126205444,
+      "learning_rate": 0.0001443654400839517,
+      "loss": 0.9933,
+      "step": 18385
+    },
+    {
+      "epoch": 0.7079884504331088,
+      "grad_norm": 1.4878582954406738,
+      "learning_rate": 0.0001443383373829959,
+      "loss": 0.8842,
+      "step": 18390
+    },
+    {
+      "epoch": 0.7081809432146294,
+      "grad_norm": 2.3553292751312256,
+      "learning_rate": 0.00014431123062759286,
+      "loss": 1.1733,
+      "step": 18395
+    },
+    {
+      "epoch": 0.7083734359961501,
+      "grad_norm": 0.8834003210067749,
+      "learning_rate": 0.00014428411982022135,
+      "loss": 1.1275,
+      "step": 18400
+    },
+    {
+      "epoch": 0.7085659287776709,
+      "grad_norm": 1.331040620803833,
+      "learning_rate": 0.00014425700496336038,
+      "loss": 1.0753,
+      "step": 18405
+    },
+    {
+      "epoch": 0.7087584215591916,
+      "grad_norm": 1.0972214937210083,
+      "learning_rate": 0.0001442298860594895,
+      "loss": 1.2045,
+      "step": 18410
+    },
+    {
+      "epoch": 0.7089509143407122,
+      "grad_norm": 1.5350794792175293,
+      "learning_rate": 0.00014420276311108857,
+      "loss": 1.0097,
+      "step": 18415
+    },
+    {
+      "epoch": 0.7091434071222329,
+      "grad_norm": 1.8360435962677002,
+      "learning_rate": 0.00014417563612063777,
+      "loss": 1.177,
+      "step": 18420
+    },
+    {
+      "epoch": 0.7093358999037536,
+      "grad_norm": 1.0898863077163696,
+      "learning_rate": 0.00014414850509061764,
+      "loss": 1.0374,
+      "step": 18425
+    },
+    {
+      "epoch": 0.7095283926852743,
+      "grad_norm": 1.2654744386672974,
+      "learning_rate": 0.00014412137002350919,
+      "loss": 1.1494,
+      "step": 18430
+    },
+    {
+      "epoch": 0.709720885466795,
+      "grad_norm": 1.8603087663650513,
+      "learning_rate": 0.00014409423092179375,
+      "loss": 1.2723,
+      "step": 18435
+    },
+    {
+      "epoch": 0.7099133782483157,
+      "grad_norm": 0.9974476099014282,
+      "learning_rate": 0.00014406708778795296,
+      "loss": 1.1139,
+      "step": 18440
+    },
+    {
+      "epoch": 0.7101058710298364,
+      "grad_norm": 0.998330295085907,
+      "learning_rate": 0.00014403994062446893,
+      "loss": 1.2881,
+      "step": 18445
+    },
+    {
+      "epoch": 0.710298363811357,
+      "grad_norm": 2.04758882522583,
+      "learning_rate": 0.00014401278943382406,
+      "loss": 1.0089,
+      "step": 18450
+    },
+    {
+      "epoch": 0.7104908565928778,
+      "grad_norm": 1.301059603691101,
+      "learning_rate": 0.0001439856342185012,
+      "loss": 1.1405,
+      "step": 18455
+    },
+    {
+      "epoch": 0.7106833493743985,
+      "grad_norm": 1.684041142463684,
+      "learning_rate": 0.00014395847498098338,
+      "loss": 1.1387,
+      "step": 18460
+    },
+    {
+      "epoch": 0.7108758421559191,
+      "grad_norm": 1.95292067527771,
+      "learning_rate": 0.0001439313117237543,
+      "loss": 1.1659,
+      "step": 18465
+    },
+    {
+      "epoch": 0.7110683349374398,
+      "grad_norm": 1.1917790174484253,
+      "learning_rate": 0.00014390414444929775,
+      "loss": 1.0497,
+      "step": 18470
+    },
+    {
+      "epoch": 0.7112608277189605,
+      "grad_norm": 1.1583658456802368,
+      "learning_rate": 0.000143876973160098,
+      "loss": 1.2276,
+      "step": 18475
+    },
+    {
+      "epoch": 0.7114533205004813,
+      "grad_norm": 1.116721749305725,
+      "learning_rate": 0.00014384979785863976,
+      "loss": 1.3688,
+      "step": 18480
+    },
+    {
+      "epoch": 0.7116458132820019,
+      "grad_norm": 1.1651076078414917,
+      "learning_rate": 0.00014382261854740795,
+      "loss": 1.3093,
+      "step": 18485
+    },
+    {
+      "epoch": 0.7118383060635226,
+      "grad_norm": 1.2162317037582397,
+      "learning_rate": 0.00014379543522888798,
+      "loss": 1.1324,
+      "step": 18490
+    },
+    {
+      "epoch": 0.7120307988450433,
+      "grad_norm": 1.5792020559310913,
+      "learning_rate": 0.0001437682479055656,
+      "loss": 1.123,
+      "step": 18495
+    },
+    {
+      "epoch": 0.7122232916265641,
+      "grad_norm": 0.9636641144752502,
+      "learning_rate": 0.00014374105657992688,
+      "loss": 1.0547,
+      "step": 18500
+    },
+    {
+      "epoch": 0.7124157844080847,
+      "grad_norm": 1.1409319639205933,
+      "learning_rate": 0.00014371386125445828,
+      "loss": 1.1277,
+      "step": 18505
+    },
+    {
+      "epoch": 0.7126082771896054,
+      "grad_norm": 1.074267029762268,
+      "learning_rate": 0.00014368666193164664,
+      "loss": 1.1041,
+      "step": 18510
+    },
+    {
+      "epoch": 0.7128007699711261,
+      "grad_norm": 1.2324203252792358,
+      "learning_rate": 0.00014365945861397918,
+      "loss": 1.1274,
+      "step": 18515
+    },
+    {
+      "epoch": 0.7129932627526467,
+      "grad_norm": 1.2441449165344238,
+      "learning_rate": 0.00014363225130394343,
+      "loss": 1.0739,
+      "step": 18520
+    },
+    {
+      "epoch": 0.7131857555341675,
+      "grad_norm": 1.0249239206314087,
+      "learning_rate": 0.00014360504000402737,
+      "loss": 1.1945,
+      "step": 18525
+    },
+    {
+      "epoch": 0.7133782483156882,
+      "grad_norm": 1.0297977924346924,
+      "learning_rate": 0.00014357782471671922,
+      "loss": 1.1694,
+      "step": 18530
+    },
+    {
+      "epoch": 0.7135707410972089,
+      "grad_norm": 1.6610252857208252,
+      "learning_rate": 0.00014355060544450767,
+      "loss": 1.2034,
+      "step": 18535
+    },
+    {
+      "epoch": 0.7137632338787295,
+      "grad_norm": 1.290869951248169,
+      "learning_rate": 0.0001435233821898818,
+      "loss": 1.1195,
+      "step": 18540
+    },
+    {
+      "epoch": 0.7139557266602502,
+      "grad_norm": 1.4730745553970337,
+      "learning_rate": 0.0001434961549553309,
+      "loss": 1.1237,
+      "step": 18545
+    },
+    {
+      "epoch": 0.714148219441771,
+      "grad_norm": 1.0857551097869873,
+      "learning_rate": 0.00014346892374334479,
+      "loss": 1.013,
+      "step": 18550
+    },
+    {
+      "epoch": 0.7143407122232917,
+      "grad_norm": 1.0761737823486328,
+      "learning_rate": 0.00014344168855641356,
+      "loss": 0.9948,
+      "step": 18555
+    },
+    {
+      "epoch": 0.7145332050048123,
+      "grad_norm": 2.012099027633667,
+      "learning_rate": 0.00014341444939702767,
+      "loss": 1.1598,
+      "step": 18560
+    },
+    {
+      "epoch": 0.714725697786333,
+      "grad_norm": 1.837538242340088,
+      "learning_rate": 0.000143387206267678,
+      "loss": 1.1389,
+      "step": 18565
+    },
+    {
+      "epoch": 0.7149181905678537,
+      "grad_norm": 1.1099295616149902,
+      "learning_rate": 0.0001433599591708557,
+      "loss": 1.0835,
+      "step": 18570
+    },
+    {
+      "epoch": 0.7151106833493744,
+      "grad_norm": 0.9746969938278198,
+      "learning_rate": 0.00014333270810905238,
+      "loss": 0.973,
+      "step": 18575
+    },
+    {
+      "epoch": 0.7153031761308951,
+      "grad_norm": 1.9786537885665894,
+      "learning_rate": 0.00014330545308475996,
+      "loss": 1.1564,
+      "step": 18580
+    },
+    {
+      "epoch": 0.7154956689124158,
+      "grad_norm": 1.020973801612854,
+      "learning_rate": 0.0001432781941004707,
+      "loss": 1.0202,
+      "step": 18585
+    },
+    {
+      "epoch": 0.7156881616939365,
+      "grad_norm": 1.2314329147338867,
+      "learning_rate": 0.0001432509311586773,
+      "loss": 1.2654,
+      "step": 18590
+    },
+    {
+      "epoch": 0.7158806544754571,
+      "grad_norm": 1.1897294521331787,
+      "learning_rate": 0.00014322366426187277,
+      "loss": 1.3241,
+      "step": 18595
+    },
+    {
+      "epoch": 0.7160731472569779,
+      "grad_norm": 1.2122468948364258,
+      "learning_rate": 0.00014319639341255048,
+      "loss": 1.0044,
+      "step": 18600
+    },
+    {
+      "epoch": 0.7162656400384986,
+      "grad_norm": 1.5471996068954468,
+      "learning_rate": 0.00014316911861320415,
+      "loss": 1.2251,
+      "step": 18605
+    },
+    {
+      "epoch": 0.7164581328200192,
+      "grad_norm": 1.4441865682601929,
+      "learning_rate": 0.00014314183986632788,
+      "loss": 1.1717,
+      "step": 18610
+    },
+    {
+      "epoch": 0.7166506256015399,
+      "grad_norm": 1.092637300491333,
+      "learning_rate": 0.00014311455717441616,
+      "loss": 1.0724,
+      "step": 18615
+    },
+    {
+      "epoch": 0.7168431183830606,
+      "grad_norm": 1.0974675416946411,
+      "learning_rate": 0.00014308727053996377,
+      "loss": 1.0623,
+      "step": 18620
+    },
+    {
+      "epoch": 0.7170356111645814,
+      "grad_norm": 1.513769507408142,
+      "learning_rate": 0.00014305997996546594,
+      "loss": 1.1027,
+      "step": 18625
+    },
+    {
+      "epoch": 0.717228103946102,
+      "grad_norm": 1.0637279748916626,
+      "learning_rate": 0.00014303268545341817,
+      "loss": 1.0313,
+      "step": 18630
+    },
+    {
+      "epoch": 0.7174205967276227,
+      "grad_norm": 1.3569130897521973,
+      "learning_rate": 0.00014300538700631643,
+      "loss": 1.0324,
+      "step": 18635
+    },
+    {
+      "epoch": 0.7176130895091434,
+      "grad_norm": 1.0008260011672974,
+      "learning_rate": 0.00014297808462665688,
+      "loss": 1.0383,
+      "step": 18640
+    },
+    {
+      "epoch": 0.717805582290664,
+      "grad_norm": 1.291493535041809,
+      "learning_rate": 0.0001429507783169362,
+      "loss": 1.2128,
+      "step": 18645
+    },
+    {
+      "epoch": 0.7179980750721848,
+      "grad_norm": 2.5597760677337646,
+      "learning_rate": 0.0001429234680796514,
+      "loss": 1.146,
+      "step": 18650
+    },
+    {
+      "epoch": 0.7181905678537055,
+      "grad_norm": 2.4308478832244873,
+      "learning_rate": 0.00014289615391729974,
+      "loss": 1.3797,
+      "step": 18655
+    },
+    {
+      "epoch": 0.7183830606352262,
+      "grad_norm": 1.1110010147094727,
+      "learning_rate": 0.00014286883583237896,
+      "loss": 1.2471,
+      "step": 18660
+    },
+    {
+      "epoch": 0.7185755534167468,
+      "grad_norm": 1.075013279914856,
+      "learning_rate": 0.00014284151382738718,
+      "loss": 1.0836,
+      "step": 18665
+    },
+    {
+      "epoch": 0.7187680461982676,
+      "grad_norm": 1.9422922134399414,
+      "learning_rate": 0.00014281418790482273,
+      "loss": 1.3271,
+      "step": 18670
+    },
+    {
+      "epoch": 0.7189605389797883,
+      "grad_norm": 1.58540678024292,
+      "learning_rate": 0.00014278685806718442,
+      "loss": 1.0762,
+      "step": 18675
+    },
+    {
+      "epoch": 0.719153031761309,
+      "grad_norm": 1.1696521043777466,
+      "learning_rate": 0.00014275952431697138,
+      "loss": 1.1783,
+      "step": 18680
+    },
+    {
+      "epoch": 0.7193455245428296,
+      "grad_norm": 1.6518898010253906,
+      "learning_rate": 0.0001427321866566831,
+      "loss": 0.9509,
+      "step": 18685
+    },
+    {
+      "epoch": 0.7195380173243503,
+      "grad_norm": 1.2448405027389526,
+      "learning_rate": 0.0001427048450888194,
+      "loss": 1.1316,
+      "step": 18690
+    },
+    {
+      "epoch": 0.7197305101058711,
+      "grad_norm": 0.9715486168861389,
+      "learning_rate": 0.00014267749961588053,
+      "loss": 1.1547,
+      "step": 18695
+    },
+    {
+      "epoch": 0.7199230028873917,
+      "grad_norm": 1.176511287689209,
+      "learning_rate": 0.00014265015024036702,
+      "loss": 1.0325,
+      "step": 18700
+    },
+    {
+      "epoch": 0.7201154956689124,
+      "grad_norm": 1.096604824066162,
+      "learning_rate": 0.0001426227969647798,
+      "loss": 0.9389,
+      "step": 18705
+    },
+    {
+      "epoch": 0.7203079884504331,
+      "grad_norm": 1.2895269393920898,
+      "learning_rate": 0.00014259543979162017,
+      "loss": 1.1157,
+      "step": 18710
+    },
+    {
+      "epoch": 0.7205004812319538,
+      "grad_norm": 1.1590831279754639,
+      "learning_rate": 0.00014256807872338974,
+      "loss": 1.0154,
+      "step": 18715
+    },
+    {
+      "epoch": 0.7206929740134745,
+      "grad_norm": 1.1659713983535767,
+      "learning_rate": 0.00014254071376259046,
+      "loss": 1.0744,
+      "step": 18720
+    },
+    {
+      "epoch": 0.7208854667949952,
+      "grad_norm": 1.3548671007156372,
+      "learning_rate": 0.00014251334491172473,
+      "loss": 0.9823,
+      "step": 18725
+    },
+    {
+      "epoch": 0.7210779595765159,
+      "grad_norm": 1.5639405250549316,
+      "learning_rate": 0.00014248597217329526,
+      "loss": 1.1793,
+      "step": 18730
+    },
+    {
+      "epoch": 0.7212704523580366,
+      "grad_norm": 1.9836759567260742,
+      "learning_rate": 0.00014245859554980504,
+      "loss": 1.1827,
+      "step": 18735
+    },
+    {
+      "epoch": 0.7214629451395572,
+      "grad_norm": 1.2241086959838867,
+      "learning_rate": 0.00014243121504375753,
+      "loss": 1.2403,
+      "step": 18740
+    },
+    {
+      "epoch": 0.721655437921078,
+      "grad_norm": 1.1298317909240723,
+      "learning_rate": 0.0001424038306576565,
+      "loss": 1.1577,
+      "step": 18745
+    },
+    {
+      "epoch": 0.7218479307025987,
+      "grad_norm": 1.325210452079773,
+      "learning_rate": 0.00014237644239400605,
+      "loss": 1.1232,
+      "step": 18750
+    },
+    {
+      "epoch": 0.7220404234841193,
+      "grad_norm": 1.613929033279419,
+      "learning_rate": 0.00014234905025531066,
+      "loss": 0.9627,
+      "step": 18755
+    },
+    {
+      "epoch": 0.72223291626564,
+      "grad_norm": 1.6307876110076904,
+      "learning_rate": 0.00014232165424407517,
+      "loss": 1.0229,
+      "step": 18760
+    },
+    {
+      "epoch": 0.7224254090471607,
+      "grad_norm": 0.8971173167228699,
+      "learning_rate": 0.00014229425436280475,
+      "loss": 1.1371,
+      "step": 18765
+    },
+    {
+      "epoch": 0.7226179018286815,
+      "grad_norm": 1.3740814924240112,
+      "learning_rate": 0.00014226685061400496,
+      "loss": 1.1431,
+      "step": 18770
+    },
+    {
+      "epoch": 0.7228103946102021,
+      "grad_norm": 1.7433820962905884,
+      "learning_rate": 0.00014223944300018163,
+      "loss": 1.1876,
+      "step": 18775
+    },
+    {
+      "epoch": 0.7230028873917228,
+      "grad_norm": 1.1470065116882324,
+      "learning_rate": 0.0001422120315238411,
+      "loss": 1.1971,
+      "step": 18780
+    },
+    {
+      "epoch": 0.7231953801732435,
+      "grad_norm": 2.0566489696502686,
+      "learning_rate": 0.00014218461618748987,
+      "loss": 1.0274,
+      "step": 18785
+    },
+    {
+      "epoch": 0.7233878729547641,
+      "grad_norm": 1.589087724685669,
+      "learning_rate": 0.00014215719699363496,
+      "loss": 1.067,
+      "step": 18790
+    },
+    {
+      "epoch": 0.7235803657362849,
+      "grad_norm": 2.473461866378784,
+      "learning_rate": 0.00014212977394478365,
+      "loss": 1.2185,
+      "step": 18795
+    },
+    {
+      "epoch": 0.7237728585178056,
+      "grad_norm": 1.3214609622955322,
+      "learning_rate": 0.00014210234704344359,
+      "loss": 1.2501,
+      "step": 18800
+    },
+    {
+      "epoch": 0.7239653512993263,
+      "grad_norm": 2.281226873397827,
+      "learning_rate": 0.0001420749162921228,
+      "loss": 1.2262,
+      "step": 18805
+    },
+    {
+      "epoch": 0.7241578440808469,
+      "grad_norm": 1.188148021697998,
+      "learning_rate": 0.0001420474816933296,
+      "loss": 1.3338,
+      "step": 18810
+    },
+    {
+      "epoch": 0.7243503368623677,
+      "grad_norm": 2.0242867469787598,
+      "learning_rate": 0.00014202004324957279,
+      "loss": 1.1157,
+      "step": 18815
+    },
+    {
+      "epoch": 0.7245428296438884,
+      "grad_norm": 1.2399152517318726,
+      "learning_rate": 0.00014199260096336134,
+      "loss": 1.119,
+      "step": 18820
+    },
+    {
+      "epoch": 0.7247353224254091,
+      "grad_norm": 1.7323557138442993,
+      "learning_rate": 0.00014196515483720477,
+      "loss": 1.1672,
+      "step": 18825
+    },
+    {
+      "epoch": 0.7249278152069297,
+      "grad_norm": 1.4888850450515747,
+      "learning_rate": 0.00014193770487361273,
+      "loss": 0.9814,
+      "step": 18830
+    },
+    {
+      "epoch": 0.7251203079884504,
+      "grad_norm": 1.5241479873657227,
+      "learning_rate": 0.0001419102510750954,
+      "loss": 1.0614,
+      "step": 18835
+    },
+    {
+      "epoch": 0.7253128007699712,
+      "grad_norm": 1.2932441234588623,
+      "learning_rate": 0.00014188279344416323,
+      "loss": 1.1905,
+      "step": 18840
+    },
+    {
+      "epoch": 0.7255052935514918,
+      "grad_norm": 1.4357131719589233,
+      "learning_rate": 0.0001418553319833271,
+      "loss": 1.1303,
+      "step": 18845
+    },
+    {
+      "epoch": 0.7256977863330125,
+      "grad_norm": 2.1818439960479736,
+      "learning_rate": 0.00014182786669509806,
+      "loss": 1.2141,
+      "step": 18850
+    },
+    {
+      "epoch": 0.7258902791145332,
+      "grad_norm": 2.111520528793335,
+      "learning_rate": 0.00014180039758198774,
+      "loss": 1.165,
+      "step": 18855
+    },
+    {
+      "epoch": 0.7260827718960539,
+      "grad_norm": 1.3923039436340332,
+      "learning_rate": 0.00014177292464650796,
+      "loss": 1.3364,
+      "step": 18860
+    },
+    {
+      "epoch": 0.7262752646775746,
+      "grad_norm": 1.905661702156067,
+      "learning_rate": 0.0001417454478911709,
+      "loss": 1.1535,
+      "step": 18865
+    },
+    {
+      "epoch": 0.7264677574590953,
+      "grad_norm": 1.1814746856689453,
+      "learning_rate": 0.0001417179673184892,
+      "loss": 1.2141,
+      "step": 18870
+    },
+    {
+      "epoch": 0.726660250240616,
+      "grad_norm": 1.4515434503555298,
+      "learning_rate": 0.00014169048293097576,
+      "loss": 1.0955,
+      "step": 18875
+    },
+    {
+      "epoch": 0.7268527430221366,
+      "grad_norm": 1.2174112796783447,
+      "learning_rate": 0.0001416629947311438,
+      "loss": 1.0399,
+      "step": 18880
+    },
+    {
+      "epoch": 0.7270452358036573,
+      "grad_norm": 1.3769662380218506,
+      "learning_rate": 0.00014163550272150698,
+      "loss": 1.2164,
+      "step": 18885
+    },
+    {
+      "epoch": 0.7272377285851781,
+      "grad_norm": 1.3401464223861694,
+      "learning_rate": 0.00014160800690457927,
+      "loss": 1.1039,
+      "step": 18890
+    },
+    {
+      "epoch": 0.7274302213666988,
+      "grad_norm": 1.1210380792617798,
+      "learning_rate": 0.0001415805072828749,
+      "loss": 1.0771,
+      "step": 18895
+    },
+    {
+      "epoch": 0.7276227141482194,
+      "grad_norm": 1.3425636291503906,
+      "learning_rate": 0.00014155300385890863,
+      "loss": 1.1506,
+      "step": 18900
+    },
+    {
+      "epoch": 0.7278152069297401,
+      "grad_norm": 1.155220866203308,
+      "learning_rate": 0.0001415254966351954,
+      "loss": 1.1321,
+      "step": 18905
+    },
+    {
+      "epoch": 0.7280076997112608,
+      "grad_norm": 1.440024733543396,
+      "learning_rate": 0.0001414979856142506,
+      "loss": 1.2324,
+      "step": 18910
+    },
+    {
+      "epoch": 0.7282001924927816,
+      "grad_norm": 1.6521823406219482,
+      "learning_rate": 0.0001414704707985899,
+      "loss": 1.196,
+      "step": 18915
+    },
+    {
+      "epoch": 0.7283926852743022,
+      "grad_norm": 3.4958372116088867,
+      "learning_rate": 0.00014144295219072937,
+      "loss": 0.9906,
+      "step": 18920
+    },
+    {
+      "epoch": 0.7285851780558229,
+      "grad_norm": 0.9254593849182129,
+      "learning_rate": 0.00014141542979318538,
+      "loss": 1.2552,
+      "step": 18925
+    },
+    {
+      "epoch": 0.7287776708373436,
+      "grad_norm": 1.519364833831787,
+      "learning_rate": 0.00014138790360847473,
+      "loss": 1.0491,
+      "step": 18930
+    },
+    {
+      "epoch": 0.7289701636188642,
+      "grad_norm": 1.199167013168335,
+      "learning_rate": 0.0001413603736391144,
+      "loss": 0.9939,
+      "step": 18935
+    },
+    {
+      "epoch": 0.729162656400385,
+      "grad_norm": 1.0213391780853271,
+      "learning_rate": 0.00014133283988762192,
+      "loss": 1.222,
+      "step": 18940
+    },
+    {
+      "epoch": 0.7293551491819057,
+      "grad_norm": 1.27894127368927,
+      "learning_rate": 0.00014130530235651506,
+      "loss": 1.2881,
+      "step": 18945
+    },
+    {
+      "epoch": 0.7295476419634264,
+      "grad_norm": 1.1660070419311523,
+      "learning_rate": 0.0001412777610483119,
+      "loss": 1.1839,
+      "step": 18950
+    },
+    {
+      "epoch": 0.729740134744947,
+      "grad_norm": 0.9614414572715759,
+      "learning_rate": 0.00014125021596553093,
+      "loss": 1.0397,
+      "step": 18955
+    },
+    {
+      "epoch": 0.7299326275264677,
+      "grad_norm": 1.5278538465499878,
+      "learning_rate": 0.00014122266711069095,
+      "loss": 1.2835,
+      "step": 18960
+    },
+    {
+      "epoch": 0.7301251203079885,
+      "grad_norm": 1.2992238998413086,
+      "learning_rate": 0.00014119511448631118,
+      "loss": 1.2873,
+      "step": 18965
+    },
+    {
+      "epoch": 0.7303176130895092,
+      "grad_norm": 1.0794028043746948,
+      "learning_rate": 0.00014116755809491104,
+      "loss": 1.1677,
+      "step": 18970
+    },
+    {
+      "epoch": 0.7305101058710298,
+      "grad_norm": 1.672555685043335,
+      "learning_rate": 0.00014113999793901046,
+      "loss": 0.9295,
+      "step": 18975
+    },
+    {
+      "epoch": 0.7307025986525505,
+      "grad_norm": 1.630053997039795,
+      "learning_rate": 0.00014111243402112957,
+      "loss": 1.1635,
+      "step": 18980
+    },
+    {
+      "epoch": 0.7308950914340713,
+      "grad_norm": 1.3171367645263672,
+      "learning_rate": 0.00014108486634378895,
+      "loss": 1.065,
+      "step": 18985
+    },
+    {
+      "epoch": 0.731087584215592,
+      "grad_norm": 1.1997402906417847,
+      "learning_rate": 0.00014105729490950948,
+      "loss": 1.0747,
+      "step": 18990
+    },
+    {
+      "epoch": 0.7312800769971126,
+      "grad_norm": 1.6320029497146606,
+      "learning_rate": 0.00014102971972081233,
+      "loss": 1.2414,
+      "step": 18995
+    },
+    {
+      "epoch": 0.7314725697786333,
+      "grad_norm": 1.3852897882461548,
+      "learning_rate": 0.00014100214078021915,
+      "loss": 1.0307,
+      "step": 19000
+    },
+    {
+      "epoch": 0.731665062560154,
+      "grad_norm": 1.29547119140625,
+      "learning_rate": 0.00014097455809025178,
+      "loss": 1.1411,
+      "step": 19005
+    },
+    {
+      "epoch": 0.7318575553416747,
+      "grad_norm": 1.0764034986495972,
+      "learning_rate": 0.00014094697165343252,
+      "loss": 1.1789,
+      "step": 19010
+    },
+    {
+      "epoch": 0.7320500481231954,
+      "grad_norm": 1.7445317506790161,
+      "learning_rate": 0.00014091938147228395,
+      "loss": 1.2379,
+      "step": 19015
+    },
+    {
+      "epoch": 0.7322425409047161,
+      "grad_norm": 1.844789743423462,
+      "learning_rate": 0.00014089178754932898,
+      "loss": 1.0126,
+      "step": 19020
+    },
+    {
+      "epoch": 0.7324350336862367,
+      "grad_norm": 1.370970368385315,
+      "learning_rate": 0.00014086418988709095,
+      "loss": 1.3182,
+      "step": 19025
+    },
+    {
+      "epoch": 0.7326275264677574,
+      "grad_norm": 1.2565025091171265,
+      "learning_rate": 0.00014083658848809347,
+      "loss": 1.1753,
+      "step": 19030
+    },
+    {
+      "epoch": 0.7328200192492782,
+      "grad_norm": 1.7159111499786377,
+      "learning_rate": 0.00014080898335486046,
+      "loss": 1.0572,
+      "step": 19035
+    },
+    {
+      "epoch": 0.7330125120307989,
+      "grad_norm": 1.8323345184326172,
+      "learning_rate": 0.0001407813744899163,
+      "loss": 1.0822,
+      "step": 19040
+    },
+    {
+      "epoch": 0.7332050048123195,
+      "grad_norm": 1.6878646612167358,
+      "learning_rate": 0.00014075376189578553,
+      "loss": 1.1133,
+      "step": 19045
+    },
+    {
+      "epoch": 0.7333974975938402,
+      "grad_norm": 1.7448841333389282,
+      "learning_rate": 0.00014072614557499323,
+      "loss": 1.0922,
+      "step": 19050
+    },
+    {
+      "epoch": 0.7335899903753609,
+      "grad_norm": 0.9125509262084961,
+      "learning_rate": 0.00014069852553006472,
+      "loss": 1.1788,
+      "step": 19055
+    },
+    {
+      "epoch": 0.7337824831568817,
+      "grad_norm": 1.8741627931594849,
+      "learning_rate": 0.00014067090176352563,
+      "loss": 1.0538,
+      "step": 19060
+    },
+    {
+      "epoch": 0.7339749759384023,
+      "grad_norm": 3.1138720512390137,
+      "learning_rate": 0.00014064327427790201,
+      "loss": 1.256,
+      "step": 19065
+    },
+    {
+      "epoch": 0.734167468719923,
+      "grad_norm": 1.3083161115646362,
+      "learning_rate": 0.00014061564307572022,
+      "loss": 0.976,
+      "step": 19070
+    },
+    {
+      "epoch": 0.7343599615014437,
+      "grad_norm": 1.176721215248108,
+      "learning_rate": 0.00014058800815950687,
+      "loss": 1.1733,
+      "step": 19075
+    },
+    {
+      "epoch": 0.7345524542829643,
+      "grad_norm": 0.9016759395599365,
+      "learning_rate": 0.00014056036953178906,
+      "loss": 1.0671,
+      "step": 19080
+    },
+    {
+      "epoch": 0.7347449470644851,
+      "grad_norm": 1.4011337757110596,
+      "learning_rate": 0.00014053272719509417,
+      "loss": 1.1453,
+      "step": 19085
+    },
+    {
+      "epoch": 0.7349374398460058,
+      "grad_norm": 1.2671010494232178,
+      "learning_rate": 0.00014050508115194988,
+      "loss": 1.1453,
+      "step": 19090
+    },
+    {
+      "epoch": 0.7351299326275265,
+      "grad_norm": 1.3316471576690674,
+      "learning_rate": 0.00014047743140488422,
+      "loss": 0.9451,
+      "step": 19095
+    },
+    {
+      "epoch": 0.7353224254090471,
+      "grad_norm": 1.963815689086914,
+      "learning_rate": 0.0001404497779564256,
+      "loss": 0.993,
+      "step": 19100
+    },
+    {
+      "epoch": 0.7355149181905678,
+      "grad_norm": 1.4354350566864014,
+      "learning_rate": 0.00014042212080910276,
+      "loss": 1.1263,
+      "step": 19105
+    },
+    {
+      "epoch": 0.7357074109720886,
+      "grad_norm": 1.6670982837677002,
+      "learning_rate": 0.00014039445996544473,
+      "loss": 1.0964,
+      "step": 19110
+    },
+    {
+      "epoch": 0.7358999037536093,
+      "grad_norm": 0.9805311560630798,
+      "learning_rate": 0.00014036679542798092,
+      "loss": 1.056,
+      "step": 19115
+    },
+    {
+      "epoch": 0.7360923965351299,
+      "grad_norm": 1.4659690856933594,
+      "learning_rate": 0.0001403391271992411,
+      "loss": 1.0984,
+      "step": 19120
+    },
+    {
+      "epoch": 0.7362848893166506,
+      "grad_norm": 0.5292593240737915,
+      "learning_rate": 0.00014031145528175525,
+      "loss": 1.0774,
+      "step": 19125
+    },
+    {
+      "epoch": 0.7364773820981714,
+      "grad_norm": 1.9471726417541504,
+      "learning_rate": 0.00014028377967805392,
+      "loss": 1.1648,
+      "step": 19130
+    },
+    {
+      "epoch": 0.736669874879692,
+      "grad_norm": 1.2082020044326782,
+      "learning_rate": 0.0001402561003906678,
+      "loss": 1.0764,
+      "step": 19135
+    },
+    {
+      "epoch": 0.7368623676612127,
+      "grad_norm": 1.558237075805664,
+      "learning_rate": 0.00014022841742212792,
+      "loss": 1.3944,
+      "step": 19140
+    },
+    {
+      "epoch": 0.7370548604427334,
+      "grad_norm": 1.7463306188583374,
+      "learning_rate": 0.0001402007307749658,
+      "loss": 1.0599,
+      "step": 19145
+    },
+    {
+      "epoch": 0.737247353224254,
+      "grad_norm": 1.2820191383361816,
+      "learning_rate": 0.00014017304045171316,
+      "loss": 1.042,
+      "step": 19150
+    },
+    {
+      "epoch": 0.7374398460057748,
+      "grad_norm": 1.617754340171814,
+      "learning_rate": 0.00014014534645490206,
+      "loss": 1.1031,
+      "step": 19155
+    },
+    {
+      "epoch": 0.7376323387872955,
+      "grad_norm": 1.0561091899871826,
+      "learning_rate": 0.00014011764878706497,
+      "loss": 1.1711,
+      "step": 19160
+    },
+    {
+      "epoch": 0.7378248315688162,
+      "grad_norm": 1.0614964962005615,
+      "learning_rate": 0.00014008994745073468,
+      "loss": 1.0783,
+      "step": 19165
+    },
+    {
+      "epoch": 0.7380173243503368,
+      "grad_norm": 1.5456453561782837,
+      "learning_rate": 0.0001400622424484442,
+      "loss": 1.0303,
+      "step": 19170
+    },
+    {
+      "epoch": 0.7382098171318575,
+      "grad_norm": 1.4854921102523804,
+      "learning_rate": 0.00014003453378272712,
+      "loss": 1.0719,
+      "step": 19175
+    },
+    {
+      "epoch": 0.7384023099133783,
+      "grad_norm": 1.4764469861984253,
+      "learning_rate": 0.00014000682145611708,
+      "loss": 1.2755,
+      "step": 19180
+    },
+    {
+      "epoch": 0.738594802694899,
+      "grad_norm": 1.6524717807769775,
+      "learning_rate": 0.00013997910547114826,
+      "loss": 1.1086,
+      "step": 19185
+    },
+    {
+      "epoch": 0.7387872954764196,
+      "grad_norm": 1.264930248260498,
+      "learning_rate": 0.00013995138583035508,
+      "loss": 1.1087,
+      "step": 19190
+    },
+    {
+      "epoch": 0.7389797882579403,
+      "grad_norm": 1.8001179695129395,
+      "learning_rate": 0.0001399236625362723,
+      "loss": 1.1736,
+      "step": 19195
+    },
+    {
+      "epoch": 0.739172281039461,
+      "grad_norm": 1.0975139141082764,
+      "learning_rate": 0.00013989593559143507,
+      "loss": 1.1669,
+      "step": 19200
+    },
+    {
+      "epoch": 0.7393647738209818,
+      "grad_norm": 1.078940987586975,
+      "learning_rate": 0.0001398682049983788,
+      "loss": 1.1259,
+      "step": 19205
+    },
+    {
+      "epoch": 0.7395572666025024,
+      "grad_norm": 1.0370323657989502,
+      "learning_rate": 0.0001398404707596393,
+      "loss": 1.2454,
+      "step": 19210
+    },
+    {
+      "epoch": 0.7397497593840231,
+      "grad_norm": 1.8001567125320435,
+      "learning_rate": 0.00013981273287775266,
+      "loss": 1.2803,
+      "step": 19215
+    },
+    {
+      "epoch": 0.7399422521655438,
+      "grad_norm": 1.00836181640625,
+      "learning_rate": 0.00013978499135525535,
+      "loss": 1.2406,
+      "step": 19220
+    },
+    {
+      "epoch": 0.7401347449470644,
+      "grad_norm": 1.169600009918213,
+      "learning_rate": 0.00013975724619468414,
+      "loss": 1.2738,
+      "step": 19225
+    },
+    {
+      "epoch": 0.7403272377285852,
+      "grad_norm": 1.6229758262634277,
+      "learning_rate": 0.00013972949739857613,
+      "loss": 1.1428,
+      "step": 19230
+    },
+    {
+      "epoch": 0.7405197305101059,
+      "grad_norm": 1.573930263519287,
+      "learning_rate": 0.00013970174496946873,
+      "loss": 1.1467,
+      "step": 19235
+    },
+    {
+      "epoch": 0.7407122232916266,
+      "grad_norm": 1.4224984645843506,
+      "learning_rate": 0.00013967398890989979,
+      "loss": 1.1335,
+      "step": 19240
+    },
+    {
+      "epoch": 0.7409047160731472,
+      "grad_norm": 1.5381492376327515,
+      "learning_rate": 0.00013964622922240736,
+      "loss": 1.1332,
+      "step": 19245
+    },
+    {
+      "epoch": 0.7410972088546679,
+      "grad_norm": 1.7980502843856812,
+      "learning_rate": 0.0001396184659095299,
+      "loss": 1.2107,
+      "step": 19250
+    },
+    {
+      "epoch": 0.7412897016361887,
+      "grad_norm": 0.8735668063163757,
+      "learning_rate": 0.00013959069897380617,
+      "loss": 1.0948,
+      "step": 19255
+    },
+    {
+      "epoch": 0.7414821944177094,
+      "grad_norm": 1.1920636892318726,
+      "learning_rate": 0.0001395629284177753,
+      "loss": 1.1663,
+      "step": 19260
+    },
+    {
+      "epoch": 0.74167468719923,
+      "grad_norm": 1.3055362701416016,
+      "learning_rate": 0.0001395351542439767,
+      "loss": 1.1108,
+      "step": 19265
+    },
+    {
+      "epoch": 0.7418671799807507,
+      "grad_norm": 1.8382583856582642,
+      "learning_rate": 0.00013950737645495014,
+      "loss": 1.0279,
+      "step": 19270
+    },
+    {
+      "epoch": 0.7420596727622714,
+      "grad_norm": 0.865042507648468,
+      "learning_rate": 0.00013947959505323577,
+      "loss": 1.1005,
+      "step": 19275
+    },
+    {
+      "epoch": 0.7422521655437921,
+      "grad_norm": 1.182671070098877,
+      "learning_rate": 0.0001394518100413739,
+      "loss": 1.199,
+      "step": 19280
+    },
+    {
+      "epoch": 0.7424446583253128,
+      "grad_norm": 1.5641695261001587,
+      "learning_rate": 0.00013942402142190532,
+      "loss": 1.098,
+      "step": 19285
+    },
+    {
+      "epoch": 0.7426371511068335,
+      "grad_norm": 0.924503743648529,
+      "learning_rate": 0.0001393962291973712,
+      "loss": 1.2765,
+      "step": 19290
+    },
+    {
+      "epoch": 0.7428296438883542,
+      "grad_norm": 1.1100239753723145,
+      "learning_rate": 0.00013936843337031287,
+      "loss": 1.2573,
+      "step": 19295
+    },
+    {
+      "epoch": 0.7430221366698749,
+      "grad_norm": 1.2185837030410767,
+      "learning_rate": 0.0001393406339432721,
+      "loss": 1.2064,
+      "step": 19300
+    },
+    {
+      "epoch": 0.7432146294513956,
+      "grad_norm": 1.5463718175888062,
+      "learning_rate": 0.000139312830918791,
+      "loss": 1.1559,
+      "step": 19305
+    },
+    {
+      "epoch": 0.7434071222329163,
+      "grad_norm": 1.8315119743347168,
+      "learning_rate": 0.00013928502429941188,
+      "loss": 1.2292,
+      "step": 19310
+    },
+    {
+      "epoch": 0.7435996150144369,
+      "grad_norm": 1.303144097328186,
+      "learning_rate": 0.00013925721408767757,
+      "loss": 1.1463,
+      "step": 19315
+    },
+    {
+      "epoch": 0.7437921077959576,
+      "grad_norm": 1.2040412425994873,
+      "learning_rate": 0.00013922940028613106,
+      "loss": 0.9717,
+      "step": 19320
+    },
+    {
+      "epoch": 0.7439846005774784,
+      "grad_norm": 1.0547009706497192,
+      "learning_rate": 0.0001392015828973158,
+      "loss": 1.0389,
+      "step": 19325
+    },
+    {
+      "epoch": 0.7441770933589991,
+      "grad_norm": 1.597541093826294,
+      "learning_rate": 0.00013917376192377543,
+      "loss": 1.0937,
+      "step": 19330
+    },
+    {
+      "epoch": 0.7443695861405197,
+      "grad_norm": 0.9714812636375427,
+      "learning_rate": 0.00013914593736805402,
+      "loss": 1.1641,
+      "step": 19335
+    },
+    {
+      "epoch": 0.7445620789220404,
+      "grad_norm": 1.2114696502685547,
+      "learning_rate": 0.00013911810923269603,
+      "loss": 1.2167,
+      "step": 19340
+    },
+    {
+      "epoch": 0.7447545717035611,
+      "grad_norm": 1.330718755722046,
+      "learning_rate": 0.000139090277520246,
+      "loss": 1.0583,
+      "step": 19345
+    },
+    {
+      "epoch": 0.7449470644850819,
+      "grad_norm": 2.2889277935028076,
+      "learning_rate": 0.0001390624422332491,
+      "loss": 0.9747,
+      "step": 19350
+    },
+    {
+      "epoch": 0.7451395572666025,
+      "grad_norm": 1.2337167263031006,
+      "learning_rate": 0.0001390346033742506,
+      "loss": 1.0294,
+      "step": 19355
+    },
+    {
+      "epoch": 0.7453320500481232,
+      "grad_norm": 1.0727423429489136,
+      "learning_rate": 0.0001390067609457962,
+      "loss": 1.0961,
+      "step": 19360
+    },
+    {
+      "epoch": 0.7455245428296439,
+      "grad_norm": 1.7654608488082886,
+      "learning_rate": 0.0001389789149504319,
+      "loss": 1.0775,
+      "step": 19365
+    },
+    {
+      "epoch": 0.7457170356111645,
+      "grad_norm": 1.2609182596206665,
+      "learning_rate": 0.0001389510653907041,
+      "loss": 1.0706,
+      "step": 19370
+    },
+    {
+      "epoch": 0.7459095283926853,
+      "grad_norm": 1.906533122062683,
+      "learning_rate": 0.00013892321226915933,
+      "loss": 1.1035,
+      "step": 19375
+    },
+    {
+      "epoch": 0.746102021174206,
+      "grad_norm": 1.1768391132354736,
+      "learning_rate": 0.00013889535558834462,
+      "loss": 1.097,
+      "step": 19380
+    },
+    {
+      "epoch": 0.7462945139557267,
+      "grad_norm": 1.7764432430267334,
+      "learning_rate": 0.00013886749535080737,
+      "loss": 1.2136,
+      "step": 19385
+    },
+    {
+      "epoch": 0.7464870067372473,
+      "grad_norm": 2.2302021980285645,
+      "learning_rate": 0.0001388396315590951,
+      "loss": 1.1236,
+      "step": 19390
+    },
+    {
+      "epoch": 0.746679499518768,
+      "grad_norm": 1.0161263942718506,
+      "learning_rate": 0.00013881176421575583,
+      "loss": 1.3818,
+      "step": 19395
+    },
+    {
+      "epoch": 0.7468719923002888,
+      "grad_norm": 1.002767562866211,
+      "learning_rate": 0.00013878389332333784,
+      "loss": 1.1468,
+      "step": 19400
+    },
+    {
+      "epoch": 0.7470644850818094,
+      "grad_norm": 1.637412667274475,
+      "learning_rate": 0.00013875601888438968,
+      "loss": 1.2074,
+      "step": 19405
+    },
+    {
+      "epoch": 0.7472569778633301,
+      "grad_norm": 1.440507173538208,
+      "learning_rate": 0.00013872814090146036,
+      "loss": 0.9903,
+      "step": 19410
+    },
+    {
+      "epoch": 0.7474494706448508,
+      "grad_norm": 1.8063361644744873,
+      "learning_rate": 0.00013870025937709913,
+      "loss": 1.1508,
+      "step": 19415
+    },
+    {
+      "epoch": 0.7476419634263715,
+      "grad_norm": 1.0809664726257324,
+      "learning_rate": 0.0001386723743138555,
+      "loss": 1.1509,
+      "step": 19420
+    },
+    {
+      "epoch": 0.7478344562078922,
+      "grad_norm": 0.7736053466796875,
+      "learning_rate": 0.00013864448571427945,
+      "loss": 1.048,
+      "step": 19425
+    },
+    {
+      "epoch": 0.7480269489894129,
+      "grad_norm": 1.588610291481018,
+      "learning_rate": 0.00013861659358092117,
+      "loss": 1.1118,
+      "step": 19430
+    },
+    {
+      "epoch": 0.7482194417709336,
+      "grad_norm": 1.5141923427581787,
+      "learning_rate": 0.00013858869791633124,
+      "loss": 1.1847,
+      "step": 19435
+    },
+    {
+      "epoch": 0.7484119345524542,
+      "grad_norm": 1.6033471822738647,
+      "learning_rate": 0.00013856079872306046,
+      "loss": 1.1109,
+      "step": 19440
+    },
+    {
+      "epoch": 0.748604427333975,
+      "grad_norm": 1.1898064613342285,
+      "learning_rate": 0.0001385328960036601,
+      "loss": 1.1758,
+      "step": 19445
+    },
+    {
+      "epoch": 0.7487969201154957,
+      "grad_norm": 1.8286123275756836,
+      "learning_rate": 0.00013850498976068166,
+      "loss": 1.1827,
+      "step": 19450
+    },
+    {
+      "epoch": 0.7489894128970164,
+      "grad_norm": 1.6806395053863525,
+      "learning_rate": 0.00013847707999667698,
+      "loss": 0.9877,
+      "step": 19455
+    },
+    {
+      "epoch": 0.749181905678537,
+      "grad_norm": 2.1087961196899414,
+      "learning_rate": 0.00013844916671419823,
+      "loss": 1.2052,
+      "step": 19460
+    },
+    {
+      "epoch": 0.7493743984600577,
+      "grad_norm": 1.4643951654434204,
+      "learning_rate": 0.0001384212499157979,
+      "loss": 0.9652,
+      "step": 19465
+    },
+    {
+      "epoch": 0.7495668912415785,
+      "grad_norm": 1.4248661994934082,
+      "learning_rate": 0.00013839332960402872,
+      "loss": 1.1903,
+      "step": 19470
+    },
+    {
+      "epoch": 0.7497593840230992,
+      "grad_norm": 1.8101911544799805,
+      "learning_rate": 0.0001383709908267036,
+      "loss": 1.0335,
+      "step": 19475
+    },
+    {
+      "epoch": 0.7499518768046198,
+      "grad_norm": 1.9175690412521362,
+      "learning_rate": 0.00013834306419730473,
+      "loss": 1.2055,
+      "step": 19480
+    },
+    {
+      "epoch": 0.7501443695861405,
+      "grad_norm": 0.930147647857666,
+      "learning_rate": 0.00013831513406168663,
+      "loss": 1.2383,
+      "step": 19485
+    },
+    {
+      "epoch": 0.7503368623676612,
+      "grad_norm": 2.496994733810425,
+      "learning_rate": 0.00013828720042240338,
+      "loss": 1.2861,
+      "step": 19490
+    },
+    {
+      "epoch": 0.750529355149182,
+      "grad_norm": 1.9224547147750854,
+      "learning_rate": 0.00013825926328200926,
+      "loss": 1.247,
+      "step": 19495
+    },
+    {
+      "epoch": 0.7507218479307026,
+      "grad_norm": 1.7266567945480347,
+      "learning_rate": 0.00013823132264305894,
+      "loss": 1.1127,
+      "step": 19500
+    },
+    {
+      "epoch": 0.7509143407122233,
+      "grad_norm": 1.2304484844207764,
+      "learning_rate": 0.00013820337850810744,
+      "loss": 1.1432,
+      "step": 19505
+    },
+    {
+      "epoch": 0.751106833493744,
+      "grad_norm": 2.311600685119629,
+      "learning_rate": 0.00013817543087971004,
+      "loss": 1.2405,
+      "step": 19510
+    },
+    {
+      "epoch": 0.7512993262752646,
+      "grad_norm": 2.358445644378662,
+      "learning_rate": 0.0001381474797604224,
+      "loss": 1.2407,
+      "step": 19515
+    },
+    {
+      "epoch": 0.7514918190567854,
+      "grad_norm": 1.2297358512878418,
+      "learning_rate": 0.00013811952515280042,
+      "loss": 0.9701,
+      "step": 19520
+    },
+    {
+      "epoch": 0.7516843118383061,
+      "grad_norm": 1.260389804840088,
+      "learning_rate": 0.00013809156705940037,
+      "loss": 1.1711,
+      "step": 19525
+    },
+    {
+      "epoch": 0.7518768046198268,
+      "grad_norm": 0.9936877489089966,
+      "learning_rate": 0.00013806360548277886,
+      "loss": 0.9045,
+      "step": 19530
+    },
+    {
+      "epoch": 0.7520692974013474,
+      "grad_norm": 1.951540470123291,
+      "learning_rate": 0.0001380356404254928,
+      "loss": 1.0988,
+      "step": 19535
+    },
+    {
+      "epoch": 0.7522617901828681,
+      "grad_norm": 1.0897135734558105,
+      "learning_rate": 0.00013800767189009935,
+      "loss": 1.0216,
+      "step": 19540
+    },
+    {
+      "epoch": 0.7524542829643889,
+      "grad_norm": 1.3618708848953247,
+      "learning_rate": 0.00013797969987915608,
+      "loss": 1.0604,
+      "step": 19545
+    },
+    {
+      "epoch": 0.7526467757459095,
+      "grad_norm": 1.413282871246338,
+      "learning_rate": 0.00013795172439522087,
+      "loss": 1.2045,
+      "step": 19550
+    },
+    {
+      "epoch": 0.7528392685274302,
+      "grad_norm": 1.4086360931396484,
+      "learning_rate": 0.00013792374544085187,
+      "loss": 1.0724,
+      "step": 19555
+    },
+    {
+      "epoch": 0.7530317613089509,
+      "grad_norm": 1.2165982723236084,
+      "learning_rate": 0.00013789576301860757,
+      "loss": 1.1886,
+      "step": 19560
+    },
+    {
+      "epoch": 0.7532242540904716,
+      "grad_norm": 1.4711132049560547,
+      "learning_rate": 0.00013786777713104678,
+      "loss": 1.1847,
+      "step": 19565
+    },
+    {
+      "epoch": 0.7534167468719923,
+      "grad_norm": 1.0978587865829468,
+      "learning_rate": 0.00013783978778072862,
+      "loss": 1.1521,
+      "step": 19570
+    },
+    {
+      "epoch": 0.753609239653513,
+      "grad_norm": 1.1508560180664062,
+      "learning_rate": 0.00013781179497021251,
+      "loss": 1.089,
+      "step": 19575
+    },
+    {
+      "epoch": 0.7538017324350337,
+      "grad_norm": 1.3086730241775513,
+      "learning_rate": 0.00013778379870205829,
+      "loss": 1.2293,
+      "step": 19580
+    },
+    {
+      "epoch": 0.7539942252165543,
+      "grad_norm": 1.63782799243927,
+      "learning_rate": 0.0001377557989788259,
+      "loss": 1.0373,
+      "step": 19585
+    },
+    {
+      "epoch": 0.754186717998075,
+      "grad_norm": 1.4707633256912231,
+      "learning_rate": 0.00013772779580307584,
+      "loss": 1.0481,
+      "step": 19590
+    },
+    {
+      "epoch": 0.7543792107795958,
+      "grad_norm": 1.6030997037887573,
+      "learning_rate": 0.0001376997891773688,
+      "loss": 1.0953,
+      "step": 19595
+    },
+    {
+      "epoch": 0.7545717035611165,
+      "grad_norm": 1.0709367990493774,
+      "learning_rate": 0.00013767177910426574,
+      "loss": 1.1094,
+      "step": 19600
+    },
+    {
+      "epoch": 0.7547641963426371,
+      "grad_norm": 1.2302757501602173,
+      "learning_rate": 0.00013764376558632807,
+      "loss": 0.9815,
+      "step": 19605
+    },
+    {
+      "epoch": 0.7549566891241578,
+      "grad_norm": 2.4043121337890625,
+      "learning_rate": 0.00013761574862611737,
+      "loss": 1.1146,
+      "step": 19610
+    },
+    {
+      "epoch": 0.7551491819056786,
+      "grad_norm": 1.2333440780639648,
+      "learning_rate": 0.00013758772822619565,
+      "loss": 1.367,
+      "step": 19615
+    },
+    {
+      "epoch": 0.7553416746871993,
+      "grad_norm": 2.032453775405884,
+      "learning_rate": 0.0001375597043891252,
+      "loss": 1.1401,
+      "step": 19620
+    },
+    {
+      "epoch": 0.7555341674687199,
+      "grad_norm": 1.1483811140060425,
+      "learning_rate": 0.00013753167711746858,
+      "loss": 1.0757,
+      "step": 19625
+    },
+    {
+      "epoch": 0.7557266602502406,
+      "grad_norm": 2.314659833908081,
+      "learning_rate": 0.0001375036464137887,
+      "loss": 1.162,
+      "step": 19630
+    },
+    {
+      "epoch": 0.7559191530317613,
+      "grad_norm": 1.460924744606018,
+      "learning_rate": 0.0001374756122806488,
+      "loss": 1.1596,
+      "step": 19635
+    },
+    {
+      "epoch": 0.756111645813282,
+      "grad_norm": 1.628796935081482,
+      "learning_rate": 0.0001374475747206124,
+      "loss": 1.2437,
+      "step": 19640
+    },
+    {
+      "epoch": 0.7563041385948027,
+      "grad_norm": 0.9428819417953491,
+      "learning_rate": 0.0001374195337362434,
+      "loss": 1.1804,
+      "step": 19645
+    },
+    {
+      "epoch": 0.7564966313763234,
+      "grad_norm": 1.1497089862823486,
+      "learning_rate": 0.00013739148933010587,
+      "loss": 1.0776,
+      "step": 19650
+    },
+    {
+      "epoch": 0.7566891241578441,
+      "grad_norm": 1.2695974111557007,
+      "learning_rate": 0.00013736344150476435,
+      "loss": 1.2446,
+      "step": 19655
+    },
+    {
+      "epoch": 0.7568816169393647,
+      "grad_norm": 1.4802236557006836,
+      "learning_rate": 0.00013733539026278364,
+      "loss": 1.066,
+      "step": 19660
+    },
+    {
+      "epoch": 0.7570741097208855,
+      "grad_norm": 1.7089695930480957,
+      "learning_rate": 0.0001373073356067288,
+      "loss": 1.0265,
+      "step": 19665
+    },
+    {
+      "epoch": 0.7572666025024062,
+      "grad_norm": 1.4578391313552856,
+      "learning_rate": 0.00013727927753916523,
+      "loss": 1.0214,
+      "step": 19670
+    },
+    {
+      "epoch": 0.7574590952839269,
+      "grad_norm": 0.7848085761070251,
+      "learning_rate": 0.00013725121606265872,
+      "loss": 1.0254,
+      "step": 19675
+    },
+    {
+      "epoch": 0.7576515880654475,
+      "grad_norm": 1.5217918157577515,
+      "learning_rate": 0.00013722315117977525,
+      "loss": 1.2226,
+      "step": 19680
+    },
+    {
+      "epoch": 0.7578440808469682,
+      "grad_norm": 1.0195049047470093,
+      "learning_rate": 0.00013719508289308118,
+      "loss": 1.0981,
+      "step": 19685
+    },
+    {
+      "epoch": 0.758036573628489,
+      "grad_norm": 0.8907167315483093,
+      "learning_rate": 0.00013716701120514323,
+      "loss": 1.0012,
+      "step": 19690
+    },
+    {
+      "epoch": 0.7582290664100096,
+      "grad_norm": 1.6701477766036987,
+      "learning_rate": 0.00013713893611852824,
+      "loss": 1.1048,
+      "step": 19695
+    },
+    {
+      "epoch": 0.7584215591915303,
+      "grad_norm": 1.4811270236968994,
+      "learning_rate": 0.0001371108576358036,
+      "loss": 1.2534,
+      "step": 19700
+    },
+    {
+      "epoch": 0.758614051973051,
+      "grad_norm": 2.0855724811553955,
+      "learning_rate": 0.00013708277575953686,
+      "loss": 0.968,
+      "step": 19705
+    },
+    {
+      "epoch": 0.7588065447545717,
+      "grad_norm": 1.4841949939727783,
+      "learning_rate": 0.00013705469049229594,
+      "loss": 1.1039,
+      "step": 19710
+    },
+    {
+      "epoch": 0.7589990375360924,
+      "grad_norm": 1.2720580101013184,
+      "learning_rate": 0.000137026601836649,
+      "loss": 0.9966,
+      "step": 19715
+    },
+    {
+      "epoch": 0.7591915303176131,
+      "grad_norm": 2.275491714477539,
+      "learning_rate": 0.00013699850979516465,
+      "loss": 1.1684,
+      "step": 19720
+    },
+    {
+      "epoch": 0.7593840230991338,
+      "grad_norm": 1.2187795639038086,
+      "learning_rate": 0.00013697041437041167,
+      "loss": 0.9793,
+      "step": 19725
+    },
+    {
+      "epoch": 0.7595765158806544,
+      "grad_norm": 1.1858078241348267,
+      "learning_rate": 0.00013694231556495915,
+      "loss": 1.1833,
+      "step": 19730
+    },
+    {
+      "epoch": 0.7597690086621751,
+      "grad_norm": 1.2739187479019165,
+      "learning_rate": 0.00013691421338137664,
+      "loss": 1.1139,
+      "step": 19735
+    },
+    {
+      "epoch": 0.7599615014436959,
+      "grad_norm": 1.7635918855667114,
+      "learning_rate": 0.00013688610782223382,
+      "loss": 1.0935,
+      "step": 19740
+    },
+    {
+      "epoch": 0.7601539942252166,
+      "grad_norm": 1.7312551736831665,
+      "learning_rate": 0.00013685799889010084,
+      "loss": 1.036,
+      "step": 19745
+    },
+    {
+      "epoch": 0.7603464870067372,
+      "grad_norm": 1.222069501876831,
+      "learning_rate": 0.00013682988658754797,
+      "loss": 1.1653,
+      "step": 19750
+    },
+    {
+      "epoch": 0.7605389797882579,
+      "grad_norm": 2.664635181427002,
+      "learning_rate": 0.00013680177091714596,
+      "loss": 1.281,
+      "step": 19755
+    },
+    {
+      "epoch": 0.7607314725697787,
+      "grad_norm": 1.2842050790786743,
+      "learning_rate": 0.00013677365188146577,
+      "loss": 1.1799,
+      "step": 19760
+    },
+    {
+      "epoch": 0.7609239653512994,
+      "grad_norm": 1.45145583152771,
+      "learning_rate": 0.00013674552948307874,
+      "loss": 1.1625,
+      "step": 19765
+    },
+    {
+      "epoch": 0.76111645813282,
+      "grad_norm": 1.8897767066955566,
+      "learning_rate": 0.00013671740372455648,
+      "loss": 1.1714,
+      "step": 19770
+    },
+    {
+      "epoch": 0.7613089509143407,
+      "grad_norm": 1.171235203742981,
+      "learning_rate": 0.00013668927460847084,
+      "loss": 1.2752,
+      "step": 19775
+    },
+    {
+      "epoch": 0.7615014436958614,
+      "grad_norm": 0.9240397810935974,
+      "learning_rate": 0.00013666114213739408,
+      "loss": 1.0669,
+      "step": 19780
+    },
+    {
+      "epoch": 0.7616939364773821,
+      "grad_norm": 1.654099941253662,
+      "learning_rate": 0.0001366330063138988,
+      "loss": 1.1941,
+      "step": 19785
+    },
+    {
+      "epoch": 0.7618864292589028,
+      "grad_norm": 1.2961543798446655,
+      "learning_rate": 0.00013660486714055768,
+      "loss": 1.2552,
+      "step": 19790
+    },
+    {
+      "epoch": 0.7620789220404235,
+      "grad_norm": 1.2810674905776978,
+      "learning_rate": 0.00013657672461994398,
+      "loss": 1.1035,
+      "step": 19795
+    },
+    {
+      "epoch": 0.7622714148219442,
+      "grad_norm": 3.574240207672119,
+      "learning_rate": 0.00013654857875463111,
+      "loss": 1.1724,
+      "step": 19800
+    },
+    {
+      "epoch": 0.7624639076034648,
+      "grad_norm": 1.0426640510559082,
+      "learning_rate": 0.00013652042954719282,
+      "loss": 1.351,
+      "step": 19805
+    },
+    {
+      "epoch": 0.7626564003849856,
+      "grad_norm": 0.9059193730354309,
+      "learning_rate": 0.00013649227700020318,
+      "loss": 1.1989,
+      "step": 19810
+    },
+    {
+      "epoch": 0.7628488931665063,
+      "grad_norm": 2.0250661373138428,
+      "learning_rate": 0.00013646412111623657,
+      "loss": 1.5794,
+      "step": 19815
+    },
+    {
+      "epoch": 0.763041385948027,
+      "grad_norm": 1.030274748802185,
+      "learning_rate": 0.00013643596189786758,
+      "loss": 0.965,
+      "step": 19820
+    },
+    {
+      "epoch": 0.7632338787295476,
+      "grad_norm": 1.976044774055481,
+      "learning_rate": 0.0001364077993476713,
+      "loss": 1.1147,
+      "step": 19825
+    },
+    {
+      "epoch": 0.7634263715110683,
+      "grad_norm": 1.6923823356628418,
+      "learning_rate": 0.00013637963346822292,
+      "loss": 1.1618,
+      "step": 19830
+    },
+    {
+      "epoch": 0.7636188642925891,
+      "grad_norm": 1.3266521692276,
+      "learning_rate": 0.00013635146426209805,
+      "loss": 1.1782,
+      "step": 19835
+    },
+    {
+      "epoch": 0.7638113570741097,
+      "grad_norm": 1.6700036525726318,
+      "learning_rate": 0.00013632329173187256,
+      "loss": 0.9154,
+      "step": 19840
+    },
+    {
+      "epoch": 0.7640038498556304,
+      "grad_norm": 1.5041186809539795,
+      "learning_rate": 0.00013629511588012273,
+      "loss": 1.1082,
+      "step": 19845
+    },
+    {
+      "epoch": 0.7641963426371511,
+      "grad_norm": 1.4730234146118164,
+      "learning_rate": 0.0001362669367094249,
+      "loss": 1.128,
+      "step": 19850
+    },
+    {
+      "epoch": 0.7643888354186718,
+      "grad_norm": 1.415727972984314,
+      "learning_rate": 0.00013623875422235602,
+      "loss": 1.1844,
+      "step": 19855
+    },
+    {
+      "epoch": 0.7645813282001925,
+      "grad_norm": 1.9785696268081665,
+      "learning_rate": 0.00013621056842149306,
+      "loss": 1.1183,
+      "step": 19860
+    },
+    {
+      "epoch": 0.7647738209817132,
+      "grad_norm": 1.0615553855895996,
+      "learning_rate": 0.00013618237930941357,
+      "loss": 1.1257,
+      "step": 19865
+    },
+    {
+      "epoch": 0.7649663137632339,
+      "grad_norm": 1.749930500984192,
+      "learning_rate": 0.00013615418688869512,
+      "loss": 0.9933,
+      "step": 19870
+    },
+    {
+      "epoch": 0.7651588065447545,
+      "grad_norm": 1.5585590600967407,
+      "learning_rate": 0.0001361259911619158,
+      "loss": 1.1877,
+      "step": 19875
+    },
+    {
+      "epoch": 0.7653512993262752,
+      "grad_norm": 1.4237456321716309,
+      "learning_rate": 0.00013609779213165393,
+      "loss": 1.0918,
+      "step": 19880
+    },
+    {
+      "epoch": 0.765543792107796,
+      "grad_norm": 1.2364110946655273,
+      "learning_rate": 0.00013606958980048805,
+      "loss": 1.0557,
+      "step": 19885
+    },
+    {
+      "epoch": 0.7657362848893167,
+      "grad_norm": 1.0982424020767212,
+      "learning_rate": 0.00013604138417099712,
+      "loss": 1.1845,
+      "step": 19890
+    },
+    {
+      "epoch": 0.7659287776708373,
+      "grad_norm": 0.8089034557342529,
+      "learning_rate": 0.00013601317524576038,
+      "loss": 1.139,
+      "step": 19895
+    },
+    {
+      "epoch": 0.766121270452358,
+      "grad_norm": 1.0913941860198975,
+      "learning_rate": 0.0001359849630273573,
+      "loss": 1.2198,
+      "step": 19900
+    },
+    {
+      "epoch": 0.7663137632338787,
+      "grad_norm": 1.3398661613464355,
+      "learning_rate": 0.00013595674751836777,
+      "loss": 0.9453,
+      "step": 19905
+    },
+    {
+      "epoch": 0.7665062560153995,
+      "grad_norm": 2.1962482929229736,
+      "learning_rate": 0.00013592852872137186,
+      "loss": 1.3174,
+      "step": 19910
+    },
+    {
+      "epoch": 0.7666987487969201,
+      "grad_norm": 1.4308804273605347,
+      "learning_rate": 0.00013590030663895001,
+      "loss": 1.1015,
+      "step": 19915
+    },
+    {
+      "epoch": 0.7668912415784408,
+      "grad_norm": 0.915403425693512,
+      "learning_rate": 0.00013587208127368292,
+      "loss": 1.0555,
+      "step": 19920
+    },
+    {
+      "epoch": 0.7670837343599615,
+      "grad_norm": 1.0108091831207275,
+      "learning_rate": 0.00013584385262815164,
+      "loss": 1.1591,
+      "step": 19925
+    },
+    {
+      "epoch": 0.7672762271414822,
+      "grad_norm": 1.7001339197158813,
+      "learning_rate": 0.00013581562070493747,
+      "loss": 1.1671,
+      "step": 19930
+    },
+    {
+      "epoch": 0.7674687199230029,
+      "grad_norm": 0.7533661723136902,
+      "learning_rate": 0.00013578738550662207,
+      "loss": 0.9644,
+      "step": 19935
+    },
+    {
+      "epoch": 0.7676612127045236,
+      "grad_norm": 1.1101553440093994,
+      "learning_rate": 0.0001357591470357873,
+      "loss": 1.0671,
+      "step": 19940
+    },
+    {
+      "epoch": 0.7678537054860443,
+      "grad_norm": 2.112529993057251,
+      "learning_rate": 0.00013573090529501544,
+      "loss": 1.1979,
+      "step": 19945
+    },
+    {
+      "epoch": 0.7680461982675649,
+      "grad_norm": 1.2636981010437012,
+      "learning_rate": 0.000135702660286889,
+      "loss": 1.1494,
+      "step": 19950
+    },
+    {
+      "epoch": 0.7682386910490857,
+      "grad_norm": 1.5712652206420898,
+      "learning_rate": 0.00013567441201399073,
+      "loss": 1.1171,
+      "step": 19955
+    },
+    {
+      "epoch": 0.7684311838306064,
+      "grad_norm": 1.0548748970031738,
+      "learning_rate": 0.00013564616047890383,
+      "loss": 0.9806,
+      "step": 19960
+    },
+    {
+      "epoch": 0.768623676612127,
+      "grad_norm": 1.828020453453064,
+      "learning_rate": 0.00013561790568421172,
+      "loss": 1.2924,
+      "step": 19965
+    },
+    {
+      "epoch": 0.7688161693936477,
+      "grad_norm": 1.037383794784546,
+      "learning_rate": 0.00013558964763249804,
+      "loss": 1.0602,
+      "step": 19970
+    },
+    {
+      "epoch": 0.7690086621751684,
+      "grad_norm": 1.3003454208374023,
+      "learning_rate": 0.00013556138632634686,
+      "loss": 1.0049,
+      "step": 19975
+    },
+    {
+      "epoch": 0.7692011549566892,
+      "grad_norm": 1.0770816802978516,
+      "learning_rate": 0.00013553312176834247,
+      "loss": 1.2497,
+      "step": 19980
+    },
+    {
+      "epoch": 0.7693936477382098,
+      "grad_norm": 1.5512239933013916,
+      "learning_rate": 0.00013550485396106947,
+      "loss": 1.053,
+      "step": 19985
+    },
+    {
+      "epoch": 0.7695861405197305,
+      "grad_norm": 1.0486184358596802,
+      "learning_rate": 0.0001354765829071128,
+      "loss": 1.2014,
+      "step": 19990
+    },
+    {
+      "epoch": 0.7697786333012512,
+      "grad_norm": 1.2066504955291748,
+      "learning_rate": 0.00013544830860905762,
+      "loss": 1.1933,
+      "step": 19995
+    },
+    {
+      "epoch": 0.7699711260827719,
+      "grad_norm": 1.090703010559082,
+      "learning_rate": 0.00013542003106948943,
+      "loss": 1.1504,
+      "step": 20000
     }
   ],
   "logging_steps": 5,
@@ -14014,7 +28014,7 @@
   "num_input_tokens_seen": 0,
   "num_train_epochs": 2,
   "save_steps": 10000,
-  "total_flos": 3.133033729973453e+17,
+  "total_flos": 6.241026539658117e+17,
   "train_batch_size": 2,
   "trial_name": null,
   "trial_params": null