diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,22244 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 4.998499565663745,
+  "eval_steps": 500,
+  "global_step": 15825,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0003158809128958383,
+      "grad_norm": 6.212856065130571,
+      "learning_rate": 6.317119393556538e-07,
+      "loss": 8.5382,
+      "step": 1
+    },
+    {
+      "epoch": 0.0015794045644791914,
+      "grad_norm": 5.441923520175301,
+      "learning_rate": 3.158559696778269e-06,
+      "loss": 8.5222,
+      "step": 5
+    },
+    {
+      "epoch": 0.003158809128958383,
+      "grad_norm": 6.224499627000259,
+      "learning_rate": 6.317119393556538e-06,
+      "loss": 8.5644,
+      "step": 10
+    },
+    {
+      "epoch": 0.004738213693437574,
+      "grad_norm": 4.55491292930646,
+      "learning_rate": 9.475679090334807e-06,
+      "loss": 8.4695,
+      "step": 15
+    },
+    {
+      "epoch": 0.006317618257916766,
+      "grad_norm": 3.7500686255294613,
+      "learning_rate": 1.2634238787113077e-05,
+      "loss": 8.384,
+      "step": 20
+    },
+    {
+      "epoch": 0.007897022822395957,
+      "grad_norm": 1.0170620331774343,
+      "learning_rate": 1.5792798483891346e-05,
+      "loss": 8.2985,
+      "step": 25
+    },
+    {
+      "epoch": 0.009476427386875147,
+      "grad_norm": 1.398708533003651,
+      "learning_rate": 1.8951358180669615e-05,
+      "loss": 8.2458,
+      "step": 30
+    },
+    {
+      "epoch": 0.01105583195135434,
+      "grad_norm": 1.445046203995498,
+      "learning_rate": 2.2109917877447884e-05,
+      "loss": 8.1732,
+      "step": 35
+    },
+    {
+      "epoch": 0.012635236515833531,
+      "grad_norm": 0.584059432072112,
+      "learning_rate": 2.5268477574226153e-05,
+      "loss": 8.1083,
+      "step": 40
+    },
+    {
+      "epoch": 0.014214641080312722,
+      "grad_norm": 0.8079302274986953,
+      "learning_rate": 2.8427037271004422e-05,
+      "loss": 8.0705,
+      "step": 45
+    },
+    {
+      "epoch": 0.015794045644791914,
+      "grad_norm": 0.49011932786257634,
+      "learning_rate": 3.158559696778269e-05,
+      "loss": 8.0232,
+      "step": 50
+    },
+    {
+      "epoch": 0.017373450209271106,
+      "grad_norm": 0.5457661408112227,
+      "learning_rate": 3.474415666456096e-05,
+      "loss": 7.9518,
+      "step": 55
+    },
+    {
+      "epoch": 0.018952854773750295,
+      "grad_norm": 0.4024225943327694,
+      "learning_rate": 3.790271636133923e-05,
+      "loss": 7.8717,
+      "step": 60
+    },
+    {
+      "epoch": 0.020532259338229487,
+      "grad_norm": 0.49433086213837385,
+      "learning_rate": 4.10612760581175e-05,
+      "loss": 7.8674,
+      "step": 65
+    },
+    {
+      "epoch": 0.02211166390270868,
+      "grad_norm": 0.34359026561624734,
+      "learning_rate": 4.421983575489577e-05,
+      "loss": 7.8532,
+      "step": 70
+    },
+    {
+      "epoch": 0.02369106846718787,
+      "grad_norm": 0.6741133299066872,
+      "learning_rate": 4.737839545167404e-05,
+      "loss": 7.7934,
+      "step": 75
+    },
+    {
+      "epoch": 0.025270473031667063,
+      "grad_norm": 0.3050937286241217,
+      "learning_rate": 5.0536955148452307e-05,
+      "loss": 7.7468,
+      "step": 80
+    },
+    {
+      "epoch": 0.02684987759614625,
+      "grad_norm": 0.2614960429830844,
+      "learning_rate": 5.3695514845230576e-05,
+      "loss": 7.7666,
+      "step": 85
+    },
+    {
+      "epoch": 0.028429282160625444,
+      "grad_norm": 0.2567785117518754,
+      "learning_rate": 5.6854074542008845e-05,
+      "loss": 7.7146,
+      "step": 90
+    },
+    {
+      "epoch": 0.030008686725104636,
+      "grad_norm": 0.6926673337604969,
+      "learning_rate": 6.0012634238787114e-05,
+      "loss": 7.6734,
+      "step": 95
+    },
+    {
+      "epoch": 0.03158809128958383,
+      "grad_norm": 0.23320925067150183,
+      "learning_rate": 6.317119393556538e-05,
+      "loss": 7.6521,
+      "step": 100
+    },
+    {
+      "epoch": 0.03316749585406302,
+      "grad_norm": 0.25069824569313287,
+      "learning_rate": 6.632975363234366e-05,
+      "loss": 7.6358,
+      "step": 105
+    },
+    {
+      "epoch": 0.03474690041854221,
+      "grad_norm": 0.1942576204740953,
+      "learning_rate": 6.948831332912192e-05,
+      "loss": 7.5837,
+      "step": 110
+    },
+    {
+      "epoch": 0.036326304983021404,
+      "grad_norm": 0.38198932971350547,
+      "learning_rate": 7.264687302590018e-05,
+      "loss": 7.509,
+      "step": 115
+    },
+    {
+      "epoch": 0.03790570954750059,
+      "grad_norm": 0.5463580858340262,
+      "learning_rate": 7.580543272267846e-05,
+      "loss": 7.5754,
+      "step": 120
+    },
+    {
+      "epoch": 0.03948511411197978,
+      "grad_norm": 0.48873012656695597,
+      "learning_rate": 7.896399241945674e-05,
+      "loss": 7.5584,
+      "step": 125
+    },
+    {
+      "epoch": 0.04106451867645897,
+      "grad_norm": 0.43371845010250426,
+      "learning_rate": 8.2122552116235e-05,
+      "loss": 7.4969,
+      "step": 130
+    },
+    {
+      "epoch": 0.042643923240938165,
+      "grad_norm": 0.14855751116650812,
+      "learning_rate": 8.528111181301326e-05,
+      "loss": 7.4691,
+      "step": 135
+    },
+    {
+      "epoch": 0.04422332780541736,
+      "grad_norm": 0.4431555849752563,
+      "learning_rate": 8.843967150979154e-05,
+      "loss": 7.4897,
+      "step": 140
+    },
+    {
+      "epoch": 0.04580273236989655,
+      "grad_norm": 0.7149119994284892,
+      "learning_rate": 9.159823120656981e-05,
+      "loss": 7.4245,
+      "step": 145
+    },
+    {
+      "epoch": 0.04738213693437574,
+      "grad_norm": 0.5337474765697023,
+      "learning_rate": 9.475679090334807e-05,
+      "loss": 7.4496,
+      "step": 150
+    },
+    {
+      "epoch": 0.048961541498854934,
+      "grad_norm": 0.41578257630377913,
+      "learning_rate": 9.791535060012634e-05,
+      "loss": 7.4453,
+      "step": 155
+    },
+    {
+      "epoch": 0.050540946063334126,
+      "grad_norm": 0.33166265263489864,
+      "learning_rate": 0.00010107391029690461,
+      "loss": 7.3195,
+      "step": 160
+    },
+    {
+      "epoch": 0.05212035062781331,
+      "grad_norm": 0.37257521426585277,
+      "learning_rate": 0.00010423246999368289,
+      "loss": 7.3363,
+      "step": 165
+    },
+    {
+      "epoch": 0.0536997551922925,
+      "grad_norm": 0.8419585693918789,
+      "learning_rate": 0.00010739102969046115,
+      "loss": 7.3107,
+      "step": 170
+    },
+    {
+      "epoch": 0.055279159756771695,
+      "grad_norm": 0.9683416314959721,
+      "learning_rate": 0.00011054958938723943,
+      "loss": 7.2638,
+      "step": 175
+    },
+    {
+      "epoch": 0.05685856432125089,
+      "grad_norm": 0.8880701070544468,
+      "learning_rate": 0.00011370814908401769,
+      "loss": 7.2934,
+      "step": 180
+    },
+    {
+      "epoch": 0.05843796888573008,
+      "grad_norm": 0.5841725028771209,
+      "learning_rate": 0.00011686670878079595,
+      "loss": 7.34,
+      "step": 185
+    },
+    {
+      "epoch": 0.06001737345020927,
+      "grad_norm": 0.3513560839888165,
+      "learning_rate": 0.00012002526847757423,
+      "loss": 7.3608,
+      "step": 190
+    },
+    {
+      "epoch": 0.061596778014688464,
+      "grad_norm": 0.447712478864365,
+      "learning_rate": 0.0001231838281743525,
+      "loss": 7.3063,
+      "step": 195
+    },
+    {
+      "epoch": 0.06317618257916766,
+      "grad_norm": 1.006450646835058,
+      "learning_rate": 0.00012634238787113077,
+      "loss": 7.243,
+      "step": 200
+    },
+    {
+      "epoch": 0.06475558714364685,
+      "grad_norm": 0.7173850731067234,
+      "learning_rate": 0.00012950094756790904,
+      "loss": 7.2677,
+      "step": 205
+    },
+    {
+      "epoch": 0.06633499170812604,
+      "grad_norm": 0.9216993218698499,
+      "learning_rate": 0.00013265950726468732,
+      "loss": 7.2272,
+      "step": 210
+    },
+    {
+      "epoch": 0.06791439627260523,
+      "grad_norm": 1.0389416519219288,
+      "learning_rate": 0.00013581806696146557,
+      "loss": 7.227,
+      "step": 215
+    },
+    {
+      "epoch": 0.06949380083708442,
+      "grad_norm": 0.8340948519918115,
+      "learning_rate": 0.00013897662665824384,
+      "loss": 7.1824,
+      "step": 220
+    },
+    {
+      "epoch": 0.07107320540156362,
+      "grad_norm": 0.3870955261336493,
+      "learning_rate": 0.00014213518635502212,
+      "loss": 7.1977,
+      "step": 225
+    },
+    {
+      "epoch": 0.07265260996604281,
+      "grad_norm": 0.3683202325041625,
+      "learning_rate": 0.00014529374605180037,
+      "loss": 7.2583,
+      "step": 230
+    },
+    {
+      "epoch": 0.07423201453052199,
+      "grad_norm": 0.3596763082283532,
+      "learning_rate": 0.00014845230574857864,
+      "loss": 7.1471,
+      "step": 235
+    },
+    {
+      "epoch": 0.07581141909500118,
+      "grad_norm": 0.42532969548337346,
+      "learning_rate": 0.00015161086544535692,
+      "loss": 7.1602,
+      "step": 240
+    },
+    {
+      "epoch": 0.07739082365948037,
+      "grad_norm": 0.4682925421768155,
+      "learning_rate": 0.0001547694251421352,
+      "loss": 7.1785,
+      "step": 245
+    },
+    {
+      "epoch": 0.07897022822395956,
+      "grad_norm": 0.2593792457091068,
+      "learning_rate": 0.00015792798483891347,
+      "loss": 7.1123,
+      "step": 250
+    },
+    {
+      "epoch": 0.08054963278843875,
+      "grad_norm": 0.3548563172798498,
+      "learning_rate": 0.00016108654453569172,
+      "loss": 7.0681,
+      "step": 255
+    },
+    {
+      "epoch": 0.08212903735291795,
+      "grad_norm": 0.23746523684222884,
+      "learning_rate": 0.00016424510423247,
+      "loss": 7.1175,
+      "step": 260
+    },
+    {
+      "epoch": 0.08370844191739714,
+      "grad_norm": 0.2931552754311947,
+      "learning_rate": 0.00016740366392924827,
+      "loss": 7.1107,
+      "step": 265
+    },
+    {
+      "epoch": 0.08528784648187633,
+      "grad_norm": 0.45247713625301256,
+      "learning_rate": 0.00017056222362602652,
+      "loss": 7.1118,
+      "step": 270
+    },
+    {
+      "epoch": 0.08686725104635552,
+      "grad_norm": 0.4946172916591009,
+      "learning_rate": 0.0001737207833228048,
+      "loss": 7.1613,
+      "step": 275
+    },
+    {
+      "epoch": 0.08844665561083472,
+      "grad_norm": 0.4604549535906028,
+      "learning_rate": 0.00017687934301958307,
+      "loss": 7.1548,
+      "step": 280
+    },
+    {
+      "epoch": 0.09002606017531391,
+      "grad_norm": 0.5301197511893415,
+      "learning_rate": 0.00018003790271636132,
+      "loss": 7.1146,
+      "step": 285
+    },
+    {
+      "epoch": 0.0916054647397931,
+      "grad_norm": 0.1660195259037277,
+      "learning_rate": 0.00018319646241313962,
+      "loss": 7.0903,
+      "step": 290
+    },
+    {
+      "epoch": 0.09318486930427229,
+      "grad_norm": 0.8970188429319172,
+      "learning_rate": 0.0001863550221099179,
+      "loss": 7.0519,
+      "step": 295
+    },
+    {
+      "epoch": 0.09476427386875148,
+      "grad_norm": 0.35501590333638516,
+      "learning_rate": 0.00018951358180669615,
+      "loss": 7.1412,
+      "step": 300
+    },
+    {
+      "epoch": 0.09634367843323068,
+      "grad_norm": 0.5402743924835812,
+      "learning_rate": 0.00019267214150347443,
+      "loss": 7.08,
+      "step": 305
+    },
+    {
+      "epoch": 0.09792308299770987,
+      "grad_norm": 0.5222405693494848,
+      "learning_rate": 0.00019583070120025267,
+      "loss": 7.0659,
+      "step": 310
+    },
+    {
+      "epoch": 0.09950248756218906,
+      "grad_norm": 0.7653128151615268,
+      "learning_rate": 0.00019898926089703095,
+      "loss": 7.0446,
+      "step": 315
+    },
+    {
+      "epoch": 0.10108189212666825,
+      "grad_norm": 0.41020927031585835,
+      "learning_rate": 0.00020214782059380923,
+      "loss": 6.9886,
+      "step": 320
+    },
+    {
+      "epoch": 0.10266129669114744,
+      "grad_norm": 0.3465446565452738,
+      "learning_rate": 0.00020530638029058748,
+      "loss": 7.0641,
+      "step": 325
+    },
+    {
+      "epoch": 0.10424070125562662,
+      "grad_norm": 0.5451369780514456,
+      "learning_rate": 0.00020846493998736578,
+      "loss": 7.0725,
+      "step": 330
+    },
+    {
+      "epoch": 0.10582010582010581,
+      "grad_norm": 0.3149566875379701,
+      "learning_rate": 0.00021162349968414405,
+      "loss": 7.0405,
+      "step": 335
+    },
+    {
+      "epoch": 0.107399510384585,
+      "grad_norm": 1.0500906357634088,
+      "learning_rate": 0.0002147820593809223,
+      "loss": 7.0727,
+      "step": 340
+    },
+    {
+      "epoch": 0.1089789149490642,
+      "grad_norm": 0.708025196449839,
+      "learning_rate": 0.00021794061907770058,
+      "loss": 7.0846,
+      "step": 345
+    },
+    {
+      "epoch": 0.11055831951354339,
+      "grad_norm": 0.40365285259163436,
+      "learning_rate": 0.00022109917877447885,
+      "loss": 7.0724,
+      "step": 350
+    },
+    {
+      "epoch": 0.11213772407802258,
+      "grad_norm": 1.009853284364277,
+      "learning_rate": 0.0002242577384712571,
+      "loss": 7.1044,
+      "step": 355
+    },
+    {
+      "epoch": 0.11371712864250177,
+      "grad_norm": 0.30563348116782707,
+      "learning_rate": 0.00022741629816803538,
+      "loss": 7.0489,
+      "step": 360
+    },
+    {
+      "epoch": 0.11529653320698097,
+      "grad_norm": 0.7687726362295777,
+      "learning_rate": 0.00023057485786481363,
+      "loss": 7.0521,
+      "step": 365
+    },
+    {
+      "epoch": 0.11687593777146016,
+      "grad_norm": 0.5391670058221071,
+      "learning_rate": 0.0002337334175615919,
+      "loss": 7.0408,
+      "step": 370
+    },
+    {
+      "epoch": 0.11845534233593935,
+      "grad_norm": 1.5486041455114379,
+      "learning_rate": 0.0002368919772583702,
+      "loss": 7.0371,
+      "step": 375
+    },
+    {
+      "epoch": 0.12003474690041854,
+      "grad_norm": 0.9903096372991533,
+      "learning_rate": 0.00024005053695514846,
+      "loss": 7.0633,
+      "step": 380
+    },
+    {
+      "epoch": 0.12161415146489774,
+      "grad_norm": 0.8877067690259648,
+      "learning_rate": 0.00024320909665192673,
+      "loss": 7.0858,
+      "step": 385
+    },
+    {
+      "epoch": 0.12319355602937693,
+      "grad_norm": 0.49172389967013586,
+      "learning_rate": 0.000246367656348705,
+      "loss": 7.0088,
+      "step": 390
+    },
+    {
+      "epoch": 0.12477296059385612,
+      "grad_norm": 0.33449461681545467,
+      "learning_rate": 0.00024952621604548323,
+      "loss": 7.041,
+      "step": 395
+    },
+    {
+      "epoch": 0.1263523651583353,
+      "grad_norm": 0.1510023218564581,
+      "learning_rate": 0.00025268477574226153,
+      "loss": 7.0501,
+      "step": 400
+    },
+    {
+      "epoch": 0.1279317697228145,
+      "grad_norm": 0.1667817614114935,
+      "learning_rate": 0.00025584333543903984,
+      "loss": 7.0478,
+      "step": 405
+    },
+    {
+      "epoch": 0.1295111742872937,
+      "grad_norm": 0.31763527428712873,
+      "learning_rate": 0.0002590018951358181,
+      "loss": 7.0516,
+      "step": 410
+    },
+    {
+      "epoch": 0.1310905788517729,
+      "grad_norm": 0.8340640684806893,
+      "learning_rate": 0.00026216045483259633,
+      "loss": 7.0509,
+      "step": 415
+    },
+    {
+      "epoch": 0.13266998341625208,
+      "grad_norm": 1.2175814470982558,
+      "learning_rate": 0.00026531901452937464,
+      "loss": 7.0111,
+      "step": 420
+    },
+    {
+      "epoch": 0.13424938798073127,
+      "grad_norm": 0.2600000626861157,
+      "learning_rate": 0.0002684775742261529,
+      "loss": 7.038,
+      "step": 425
+    },
+    {
+      "epoch": 0.13582879254521046,
+      "grad_norm": 0.3670628581593165,
+      "learning_rate": 0.00027163613392293113,
+      "loss": 6.9391,
+      "step": 430
+    },
+    {
+      "epoch": 0.13740819710968966,
+      "grad_norm": 1.6405436514269163,
+      "learning_rate": 0.00027479469361970944,
+      "loss": 6.9568,
+      "step": 435
+    },
+    {
+      "epoch": 0.13898760167416885,
+      "grad_norm": 1.2445712669582063,
+      "learning_rate": 0.0002779532533164877,
+      "loss": 6.9098,
+      "step": 440
+    },
+    {
+      "epoch": 0.14056700623864804,
+      "grad_norm": 0.370130351586787,
+      "learning_rate": 0.00028111181301326594,
+      "loss": 6.9357,
+      "step": 445
+    },
+    {
+      "epoch": 0.14214641080312723,
+      "grad_norm": 1.0843722903361295,
+      "learning_rate": 0.00028427037271004424,
+      "loss": 6.9319,
+      "step": 450
+    },
+    {
+      "epoch": 0.14372581536760642,
+      "grad_norm": 0.6683778983623138,
+      "learning_rate": 0.0002874289324068225,
+      "loss": 6.9744,
+      "step": 455
+    },
+    {
+      "epoch": 0.14530521993208562,
+      "grad_norm": 1.0477861738856202,
+      "learning_rate": 0.00029058749210360074,
+      "loss": 6.9189,
+      "step": 460
+    },
+    {
+      "epoch": 0.1468846244965648,
+      "grad_norm": 0.25574892908976704,
+      "learning_rate": 0.00029374605180037904,
+      "loss": 6.9273,
+      "step": 465
+    },
+    {
+      "epoch": 0.14846402906104397,
+      "grad_norm": 0.9522550943300727,
+      "learning_rate": 0.0002969046114971573,
+      "loss": 6.8819,
+      "step": 470
+    },
+    {
+      "epoch": 0.15004343362552316,
+      "grad_norm": 2.2990064607561798,
+      "learning_rate": 0.00030006317119393554,
+      "loss": 6.9029,
+      "step": 475
+    },
+    {
+      "epoch": 0.15162283819000236,
+      "grad_norm": 1.1494013293928298,
+      "learning_rate": 0.00030322173089071384,
+      "loss": 6.9855,
+      "step": 480
+    },
+    {
+      "epoch": 0.15320224275448155,
+      "grad_norm": 0.9865873913996686,
+      "learning_rate": 0.00030638029058749214,
+      "loss": 6.9323,
+      "step": 485
+    },
+    {
+      "epoch": 0.15478164731896074,
+      "grad_norm": 0.29333155830925645,
+      "learning_rate": 0.0003095388502842704,
+      "loss": 6.8918,
+      "step": 490
+    },
+    {
+      "epoch": 0.15636105188343993,
+      "grad_norm": 1.180574183859259,
+      "learning_rate": 0.0003126974099810487,
+      "loss": 6.9231,
+      "step": 495
+    },
+    {
+      "epoch": 0.15794045644791913,
+      "grad_norm": 0.3676278970131492,
+      "learning_rate": 0.00031585596967782694,
+      "loss": 6.824,
+      "step": 500
+    },
+    {
+      "epoch": 0.15951986101239832,
+      "grad_norm": 0.7386923557000004,
+      "learning_rate": 0.0003190145293746052,
+      "loss": 6.8202,
+      "step": 505
+    },
+    {
+      "epoch": 0.1610992655768775,
+      "grad_norm": 0.49014816122240945,
+      "learning_rate": 0.00032217308907138344,
+      "loss": 6.8693,
+      "step": 510
+    },
+    {
+      "epoch": 0.1626786701413567,
+      "grad_norm": 0.6786399651886812,
+      "learning_rate": 0.00032533164876816174,
+      "loss": 6.8113,
+      "step": 515
+    },
+    {
+      "epoch": 0.1642580747058359,
+      "grad_norm": 0.2186297422888697,
+      "learning_rate": 0.00032849020846494,
+      "loss": 6.8103,
+      "step": 520
+    },
+    {
+      "epoch": 0.16583747927031509,
+      "grad_norm": 0.2993551100105818,
+      "learning_rate": 0.00033164876816171824,
+      "loss": 6.7853,
+      "step": 525
+    },
+    {
+      "epoch": 0.16741688383479428,
+      "grad_norm": 1.0992601864877571,
+      "learning_rate": 0.00033480732785849654,
+      "loss": 6.7248,
+      "step": 530
+    },
+    {
+      "epoch": 0.16899628839927347,
+      "grad_norm": 1.7439864127638862,
+      "learning_rate": 0.0003379658875552748,
+      "loss": 6.761,
+      "step": 535
+    },
+    {
+      "epoch": 0.17057569296375266,
+      "grad_norm": 1.1409198091974713,
+      "learning_rate": 0.00034112444725205304,
+      "loss": 6.7303,
+      "step": 540
+    },
+    {
+      "epoch": 0.17215509752823185,
+      "grad_norm": 1.0606879735807513,
+      "learning_rate": 0.00034428300694883135,
+      "loss": 6.7154,
+      "step": 545
+    },
+    {
+      "epoch": 0.17373450209271105,
+      "grad_norm": 1.4442637810068535,
+      "learning_rate": 0.0003474415666456096,
+      "loss": 6.7215,
+      "step": 550
+    },
+    {
+      "epoch": 0.17531390665719024,
+      "grad_norm": 1.006395303818788,
+      "learning_rate": 0.00035060012634238784,
+      "loss": 6.6503,
+      "step": 555
+    },
+    {
+      "epoch": 0.17689331122166943,
+      "grad_norm": 1.0589832445448473,
+      "learning_rate": 0.00035375868603916615,
+      "loss": 6.7135,
+      "step": 560
+    },
+    {
+      "epoch": 0.17847271578614862,
+      "grad_norm": 0.8497334144261705,
+      "learning_rate": 0.0003569172457359444,
+      "loss": 6.6027,
+      "step": 565
+    },
+    {
+      "epoch": 0.18005212035062781,
+      "grad_norm": 1.1077562449599352,
+      "learning_rate": 0.00036007580543272264,
+      "loss": 6.5622,
+      "step": 570
+    },
+    {
+      "epoch": 0.181631524915107,
+      "grad_norm": 1.7098080815292351,
+      "learning_rate": 0.000363234365129501,
+      "loss": 6.5887,
+      "step": 575
+    },
+    {
+      "epoch": 0.1832109294795862,
+      "grad_norm": 0.9248105424600663,
+      "learning_rate": 0.00036639292482627925,
+      "loss": 6.52,
+      "step": 580
+    },
+    {
+      "epoch": 0.1847903340440654,
+      "grad_norm": 0.896852304222008,
+      "learning_rate": 0.0003695514845230575,
+      "loss": 6.4583,
+      "step": 585
+    },
+    {
+      "epoch": 0.18636973860854458,
+      "grad_norm": 0.46310282774613815,
+      "learning_rate": 0.0003727100442198358,
+      "loss": 6.4367,
+      "step": 590
+    },
+    {
+      "epoch": 0.18794914317302377,
+      "grad_norm": 0.6968747602312554,
+      "learning_rate": 0.00037586860391661405,
+      "loss": 6.3631,
+      "step": 595
+    },
+    {
+      "epoch": 0.18952854773750297,
+      "grad_norm": 0.7109858278819291,
+      "learning_rate": 0.0003790271636133923,
+      "loss": 6.3846,
+      "step": 600
+    },
+    {
+      "epoch": 0.19110795230198216,
+      "grad_norm": 1.2857212115780978,
+      "learning_rate": 0.0003821857233101706,
+      "loss": 6.3662,
+      "step": 605
+    },
+    {
+      "epoch": 0.19268735686646135,
+      "grad_norm": 0.7639201898318926,
+      "learning_rate": 0.00038534428300694885,
+      "loss": 6.3526,
+      "step": 610
+    },
+    {
+      "epoch": 0.19426676143094054,
+      "grad_norm": 0.509150106992322,
+      "learning_rate": 0.0003885028427037271,
+      "loss": 6.2548,
+      "step": 615
+    },
+    {
+      "epoch": 0.19584616599541974,
+      "grad_norm": 1.0321104885060626,
+      "learning_rate": 0.00039166140240050535,
+      "loss": 6.1297,
+      "step": 620
+    },
+    {
+      "epoch": 0.19742557055989893,
+      "grad_norm": 0.7877386488974744,
+      "learning_rate": 0.00039481996209728365,
+      "loss": 6.2514,
+      "step": 625
+    },
+    {
+      "epoch": 0.19900497512437812,
+      "grad_norm": 0.8329046476850496,
+      "learning_rate": 0.0003979785217940619,
+      "loss": 6.2613,
+      "step": 630
+    },
+    {
+      "epoch": 0.2005843796888573,
+      "grad_norm": 0.8775569920638122,
+      "learning_rate": 0.00040113708149084015,
+      "loss": 6.1495,
+      "step": 635
+    },
+    {
+      "epoch": 0.2021637842533365,
+      "grad_norm": 0.7269233689233559,
+      "learning_rate": 0.00040429564118761845,
+      "loss": 6.1335,
+      "step": 640
+    },
+    {
+      "epoch": 0.2037431888178157,
+      "grad_norm": 0.6496721449107232,
+      "learning_rate": 0.0004074542008843967,
+      "loss": 6.0668,
+      "step": 645
+    },
+    {
+      "epoch": 0.2053225933822949,
+      "grad_norm": 0.5122479752510881,
+      "learning_rate": 0.00041061276058117495,
+      "loss": 6.1253,
+      "step": 650
+    },
+    {
+      "epoch": 0.20690199794677408,
+      "grad_norm": 0.5050744799240616,
+      "learning_rate": 0.00041377132027795325,
+      "loss": 6.0586,
+      "step": 655
+    },
+    {
+      "epoch": 0.20848140251125324,
+      "grad_norm": 0.3479184996986057,
+      "learning_rate": 0.00041692987997473156,
+      "loss": 6.0856,
+      "step": 660
+    },
+    {
+      "epoch": 0.21006080707573244,
+      "grad_norm": 0.3068938111653456,
+      "learning_rate": 0.0004200884396715098,
+      "loss": 5.9654,
+      "step": 665
+    },
+    {
+      "epoch": 0.21164021164021163,
+      "grad_norm": 0.39495459621524287,
+      "learning_rate": 0.0004232469993682881,
+      "loss": 6.095,
+      "step": 670
+    },
+    {
+      "epoch": 0.21321961620469082,
+      "grad_norm": 0.9660105686777036,
+      "learning_rate": 0.00042640555906506636,
+      "loss": 5.9261,
+      "step": 675
+    },
+    {
+      "epoch": 0.21479902076917,
+      "grad_norm": 0.9919331988900892,
+      "learning_rate": 0.0004295641187618446,
+      "loss": 5.886,
+      "step": 680
+    },
+    {
+      "epoch": 0.2163784253336492,
+      "grad_norm": 0.7160578065834907,
+      "learning_rate": 0.0004327226784586229,
+      "loss": 5.8574,
+      "step": 685
+    },
+    {
+      "epoch": 0.2179578298981284,
+      "grad_norm": 0.639006532826775,
+      "learning_rate": 0.00043588123815540116,
+      "loss": 5.8255,
+      "step": 690
+    },
+    {
+      "epoch": 0.2195372344626076,
+      "grad_norm": 0.3634595486992699,
+      "learning_rate": 0.0004390397978521794,
+      "loss": 5.8781,
+      "step": 695
+    },
+    {
+      "epoch": 0.22111663902708678,
+      "grad_norm": 0.2936598155781062,
+      "learning_rate": 0.0004421983575489577,
+      "loss": 5.811,
+      "step": 700
+    },
+    {
+      "epoch": 0.22269604359156597,
+      "grad_norm": 0.325525269860369,
+      "learning_rate": 0.00044535691724573596,
+      "loss": 5.771,
+      "step": 705
+    },
+    {
+      "epoch": 0.22427544815604517,
+      "grad_norm": 0.312506499301807,
+      "learning_rate": 0.0004485154769425142,
+      "loss": 5.6925,
+      "step": 710
+    },
+    {
+      "epoch": 0.22585485272052436,
+      "grad_norm": 0.9098263319413956,
+      "learning_rate": 0.0004516740366392925,
+      "loss": 5.6828,
+      "step": 715
+    },
+    {
+      "epoch": 0.22743425728500355,
+      "grad_norm": 0.38653013169140144,
+      "learning_rate": 0.00045483259633607076,
+      "loss": 5.7282,
+      "step": 720
+    },
+    {
+      "epoch": 0.22901366184948274,
+      "grad_norm": 0.8225548383300585,
+      "learning_rate": 0.000457991156032849,
+      "loss": 5.6139,
+      "step": 725
+    },
+    {
+      "epoch": 0.23059306641396193,
+      "grad_norm": 0.674246406645935,
+      "learning_rate": 0.00046114971572962726,
+      "loss": 5.751,
+      "step": 730
+    },
+    {
+      "epoch": 0.23217247097844113,
+      "grad_norm": 0.410688929283078,
+      "learning_rate": 0.00046430827542640556,
+      "loss": 5.6573,
+      "step": 735
+    },
+    {
+      "epoch": 0.23375187554292032,
+      "grad_norm": 0.5258450644644558,
+      "learning_rate": 0.0004674668351231838,
+      "loss": 5.6017,
+      "step": 740
+    },
+    {
+      "epoch": 0.2353312801073995,
+      "grad_norm": 0.5455996191043986,
+      "learning_rate": 0.00047062539481996206,
+      "loss": 5.5352,
+      "step": 745
+    },
+    {
+      "epoch": 0.2369106846718787,
+      "grad_norm": 0.6327547779360982,
+      "learning_rate": 0.0004737839545167404,
+      "loss": 5.5942,
+      "step": 750
+    },
+    {
+      "epoch": 0.2384900892363579,
+      "grad_norm": 0.4249793063677703,
+      "learning_rate": 0.00047694251421351866,
+      "loss": 5.5499,
+      "step": 755
+    },
+    {
+      "epoch": 0.24006949380083709,
+      "grad_norm": 0.4846805701090246,
+      "learning_rate": 0.0004801010739102969,
+      "loss": 5.5201,
+      "step": 760
+    },
+    {
+      "epoch": 0.24164889836531628,
+      "grad_norm": 0.40961968063799054,
+      "learning_rate": 0.0004832596336070752,
+      "loss": 5.5148,
+      "step": 765
+    },
+    {
+      "epoch": 0.24322830292979547,
+      "grad_norm": 0.43260679031245164,
+      "learning_rate": 0.00048641819330385346,
+      "loss": 5.5397,
+      "step": 770
+    },
+    {
+      "epoch": 0.24480770749427466,
+      "grad_norm": 0.580322281525859,
+      "learning_rate": 0.0004895767530006317,
+      "loss": 5.5303,
+      "step": 775
+    },
+    {
+      "epoch": 0.24638711205875385,
+      "grad_norm": 0.5838175482889031,
+      "learning_rate": 0.00049273531269741,
+      "loss": 5.4732,
+      "step": 780
+    },
+    {
+      "epoch": 0.24796651662323305,
+      "grad_norm": 0.5355101569599774,
+      "learning_rate": 0.0004958938723941882,
+      "loss": 5.4693,
+      "step": 785
+    },
+    {
+      "epoch": 0.24954592118771224,
+      "grad_norm": 0.2647729684431385,
+      "learning_rate": 0.0004990524320909665,
+      "loss": 5.4473,
+      "step": 790
+    },
+    {
+      "epoch": 0.25112532575219143,
+      "grad_norm": 0.37971671833738974,
+      "learning_rate": 0.0005022109917877448,
+      "loss": 5.4755,
+      "step": 795
+    },
+    {
+      "epoch": 0.2527047303166706,
+      "grad_norm": 0.5315182380875813,
+      "learning_rate": 0.0005053695514845231,
+      "loss": 5.3669,
+      "step": 800
+    },
+    {
+      "epoch": 0.2542841348811498,
+      "grad_norm": 0.3985349063314641,
+      "learning_rate": 0.0005085281111813014,
+      "loss": 5.3347,
+      "step": 805
+    },
+    {
+      "epoch": 0.255863539445629,
+      "grad_norm": 0.39453329040038015,
+      "learning_rate": 0.0005116866708780797,
+      "loss": 5.34,
+      "step": 810
+    },
+    {
+      "epoch": 0.2574429440101082,
+      "grad_norm": 0.44887016233351545,
+      "learning_rate": 0.0005148452305748579,
+      "loss": 5.2464,
+      "step": 815
+    },
+    {
+      "epoch": 0.2590223485745874,
+      "grad_norm": 0.4158770980372098,
+      "learning_rate": 0.0005180037902716362,
+      "loss": 5.3033,
+      "step": 820
+    },
+    {
+      "epoch": 0.2606017531390666,
+      "grad_norm": 0.4822698046717804,
+      "learning_rate": 0.0005211623499684144,
+      "loss": 5.3215,
+      "step": 825
+    },
+    {
+      "epoch": 0.2621811577035458,
+      "grad_norm": 0.4263622236375534,
+      "learning_rate": 0.0005243209096651927,
+      "loss": 5.258,
+      "step": 830
+    },
+    {
+      "epoch": 0.26376056226802497,
+      "grad_norm": 0.416073734999235,
+      "learning_rate": 0.000527479469361971,
+      "loss": 5.1997,
+      "step": 835
+    },
+    {
+      "epoch": 0.26533996683250416,
+      "grad_norm": 0.31131344489569407,
+      "learning_rate": 0.0005306380290587493,
+      "loss": 5.2658,
+      "step": 840
+    },
+    {
+      "epoch": 0.26691937139698335,
+      "grad_norm": 0.6203655842680998,
+      "learning_rate": 0.0005337965887555275,
+      "loss": 5.4155,
+      "step": 845
+    },
+    {
+      "epoch": 0.26849877596146254,
+      "grad_norm": 0.4874508807387519,
+      "learning_rate": 0.0005369551484523058,
+      "loss": 5.2538,
+      "step": 850
+    },
+    {
+      "epoch": 0.27007818052594174,
+      "grad_norm": 0.36434499694692957,
+      "learning_rate": 0.000540113708149084,
+      "loss": 5.1956,
+      "step": 855
+    },
+    {
+      "epoch": 0.2716575850904209,
+      "grad_norm": 0.35454696093453264,
+      "learning_rate": 0.0005432722678458623,
+      "loss": 5.2159,
+      "step": 860
+    },
+    {
+      "epoch": 0.2732369896549001,
+      "grad_norm": 0.3688009439063136,
+      "learning_rate": 0.0005464308275426406,
+      "loss": 5.2339,
+      "step": 865
+    },
+    {
+      "epoch": 0.2748163942193793,
+      "grad_norm": 0.42921225196120893,
+      "learning_rate": 0.0005495893872394189,
+      "loss": 5.2704,
+      "step": 870
+    },
+    {
+      "epoch": 0.2763957987838585,
+      "grad_norm": 0.4500237273968581,
+      "learning_rate": 0.0005527479469361971,
+      "loss": 5.2321,
+      "step": 875
+    },
+    {
+      "epoch": 0.2779752033483377,
+      "grad_norm": 0.6393991884989825,
+      "learning_rate": 0.0005559065066329754,
+      "loss": 5.2822,
+      "step": 880
+    },
+    {
+      "epoch": 0.2795546079128169,
+      "grad_norm": 0.4561042638940091,
+      "learning_rate": 0.0005590650663297536,
+      "loss": 5.1967,
+      "step": 885
+    },
+    {
+      "epoch": 0.2811340124772961,
+      "grad_norm": 0.4341056142543411,
+      "learning_rate": 0.0005622236260265319,
+      "loss": 5.0808,
+      "step": 890
+    },
+    {
+      "epoch": 0.2827134170417753,
+      "grad_norm": 0.33406793605522406,
+      "learning_rate": 0.0005653821857233101,
+      "loss": 5.1819,
+      "step": 895
+    },
+    {
+      "epoch": 0.28429282160625446,
+      "grad_norm": 0.2965674356367386,
+      "learning_rate": 0.0005685407454200885,
+      "loss": 5.0906,
+      "step": 900
+    },
+    {
+      "epoch": 0.28587222617073366,
+      "grad_norm": 0.2846613701718218,
+      "learning_rate": 0.0005716993051168667,
+      "loss": 5.0733,
+      "step": 905
+    },
+    {
+      "epoch": 0.28745163073521285,
+      "grad_norm": 0.3126642707513588,
+      "learning_rate": 0.000574857864813645,
+      "loss": 5.1798,
+      "step": 910
+    },
+    {
+      "epoch": 0.28903103529969204,
+      "grad_norm": 0.500617977929322,
+      "learning_rate": 0.0005780164245104232,
+      "loss": 5.1134,
+      "step": 915
+    },
+    {
+      "epoch": 0.29061043986417123,
+      "grad_norm": 0.5064572519101513,
+      "learning_rate": 0.0005811749842072015,
+      "loss": 5.0257,
+      "step": 920
+    },
+    {
+      "epoch": 0.2921898444286504,
+      "grad_norm": 0.3277593134763086,
+      "learning_rate": 0.0005843335439039797,
+      "loss": 5.1563,
+      "step": 925
+    },
+    {
+      "epoch": 0.2937692489931296,
+      "grad_norm": 0.7447972198091461,
+      "learning_rate": 0.0005874921036007581,
+      "loss": 5.0243,
+      "step": 930
+    },
+    {
+      "epoch": 0.2953486535576088,
+      "grad_norm": 0.48383747311962916,
+      "learning_rate": 0.0005906506632975363,
+      "loss": 5.0432,
+      "step": 935
+    },
+    {
+      "epoch": 0.29692805812208795,
+      "grad_norm": 0.4219932484068117,
+      "learning_rate": 0.0005938092229943146,
+      "loss": 5.0502,
+      "step": 940
+    },
+    {
+      "epoch": 0.29850746268656714,
+      "grad_norm": 0.45534278486226426,
+      "learning_rate": 0.0005969677826910928,
+      "loss": 5.0904,
+      "step": 945
+    },
+    {
+      "epoch": 0.30008686725104633,
+      "grad_norm": 0.8801601148284834,
+      "learning_rate": 0.0006001263423878711,
+      "loss": 5.1848,
+      "step": 950
+    },
+    {
+      "epoch": 0.3016662718155255,
+      "grad_norm": 0.4182931559063342,
+      "learning_rate": 0.0006032849020846493,
+      "loss": 5.0242,
+      "step": 955
+    },
+    {
+      "epoch": 0.3032456763800047,
+      "grad_norm": 0.5575755512746062,
+      "learning_rate": 0.0006064434617814277,
+      "loss": 4.9951,
+      "step": 960
+    },
+    {
+      "epoch": 0.3048250809444839,
+      "grad_norm": 0.5825632007479437,
+      "learning_rate": 0.0006096020214782059,
+      "loss": 5.0358,
+      "step": 965
+    },
+    {
+      "epoch": 0.3064044855089631,
+      "grad_norm": 0.3190473695748764,
+      "learning_rate": 0.0006127605811749843,
+      "loss": 4.9772,
+      "step": 970
+    },
+    {
+      "epoch": 0.3079838900734423,
+      "grad_norm": 0.5708845146854377,
+      "learning_rate": 0.0006159191408717625,
+      "loss": 4.9756,
+      "step": 975
+    },
+    {
+      "epoch": 0.3095632946379215,
+      "grad_norm": 0.3903158506235631,
+      "learning_rate": 0.0006190777005685408,
+      "loss": 4.9497,
+      "step": 980
+    },
+    {
+      "epoch": 0.3111426992024007,
+      "grad_norm": 0.43408173547097856,
+      "learning_rate": 0.000622236260265319,
+      "loss": 4.9968,
+      "step": 985
+    },
+    {
+      "epoch": 0.31272210376687987,
+      "grad_norm": 0.2740192171117319,
+      "learning_rate": 0.0006253948199620974,
+      "loss": 4.8619,
+      "step": 990
+    },
+    {
+      "epoch": 0.31430150833135906,
+      "grad_norm": 0.4258379345491042,
+      "learning_rate": 0.0006285533796588756,
+      "loss": 5.0052,
+      "step": 995
+    },
+    {
+      "epoch": 0.31588091289583825,
+      "grad_norm": 0.32210281861219764,
+      "learning_rate": 0.0006317119393556539,
+      "loss": 4.9013,
+      "step": 1000
+    },
+    {
+      "epoch": 0.31746031746031744,
+      "grad_norm": 0.340983266789433,
+      "learning_rate": 0.0006348704990524321,
+      "loss": 4.9806,
+      "step": 1005
+    },
+    {
+      "epoch": 0.31903972202479663,
+      "grad_norm": 0.4434195266796685,
+      "learning_rate": 0.0006380290587492104,
+      "loss": 4.8858,
+      "step": 1010
+    },
+    {
+      "epoch": 0.3206191265892758,
+      "grad_norm": 0.4466646390477218,
+      "learning_rate": 0.0006411876184459886,
+      "loss": 4.9415,
+      "step": 1015
+    },
+    {
+      "epoch": 0.322198531153755,
+      "grad_norm": 0.4677912287920951,
+      "learning_rate": 0.0006443461781427669,
+      "loss": 4.7934,
+      "step": 1020
+    },
+    {
+      "epoch": 0.3237779357182342,
+      "grad_norm": 0.33926521745253735,
+      "learning_rate": 0.0006475047378395452,
+      "loss": 4.8824,
+      "step": 1025
+    },
+    {
+      "epoch": 0.3253573402827134,
+      "grad_norm": 0.3345925861311437,
+      "learning_rate": 0.0006506632975363235,
+      "loss": 4.9302,
+      "step": 1030
+    },
+    {
+      "epoch": 0.3269367448471926,
+      "grad_norm": 0.3515104548017844,
+      "learning_rate": 0.0006538218572331017,
+      "loss": 4.9207,
+      "step": 1035
+    },
+    {
+      "epoch": 0.3285161494116718,
+      "grad_norm": 0.412734062451674,
+      "learning_rate": 0.00065698041692988,
+      "loss": 4.8604,
+      "step": 1040
+    },
+    {
+      "epoch": 0.330095553976151,
+      "grad_norm": 0.40404802367588816,
+      "learning_rate": 0.0006601389766266582,
+      "loss": 4.8051,
+      "step": 1045
+    },
+    {
+      "epoch": 0.33167495854063017,
+      "grad_norm": 0.2687637260302526,
+      "learning_rate": 0.0006632975363234365,
+      "loss": 4.7848,
+      "step": 1050
+    },
+    {
+      "epoch": 0.33325436310510936,
+      "grad_norm": 0.43853265327336977,
+      "learning_rate": 0.0006664560960202148,
+      "loss": 4.9433,
+      "step": 1055
+    },
+    {
+      "epoch": 0.33483376766958856,
+      "grad_norm": 0.5133882529724854,
+      "learning_rate": 0.0006696146557169931,
+      "loss": 4.7974,
+      "step": 1060
+    },
+    {
+      "epoch": 0.33641317223406775,
+      "grad_norm": 0.45108000204120774,
+      "learning_rate": 0.0006727732154137713,
+      "loss": 4.8565,
+      "step": 1065
+    },
+    {
+      "epoch": 0.33799257679854694,
+      "grad_norm": 0.520840065835754,
+      "learning_rate": 0.0006759317751105496,
+      "loss": 4.7931,
+      "step": 1070
+    },
+    {
+      "epoch": 0.33957198136302613,
+      "grad_norm": 0.37929072418664284,
+      "learning_rate": 0.0006790903348073278,
+      "loss": 4.7981,
+      "step": 1075
+    },
+    {
+      "epoch": 0.3411513859275053,
+      "grad_norm": 0.37545606188093755,
+      "learning_rate": 0.0006822488945041061,
+      "loss": 4.8073,
+      "step": 1080
+    },
+    {
+      "epoch": 0.3427307904919845,
+      "grad_norm": 0.6292169690326377,
+      "learning_rate": 0.0006854074542008844,
+      "loss": 4.7367,
+      "step": 1085
+    },
+    {
+      "epoch": 0.3443101950564637,
+      "grad_norm": 0.33769916351911256,
+      "learning_rate": 0.0006885660138976627,
+      "loss": 4.7695,
+      "step": 1090
+    },
+    {
+      "epoch": 0.3458895996209429,
+      "grad_norm": 0.4320591815932741,
+      "learning_rate": 0.0006917245735944409,
+      "loss": 4.6927,
+      "step": 1095
+    },
+    {
+      "epoch": 0.3474690041854221,
+      "grad_norm": 0.4091589279145143,
+      "learning_rate": 0.0006948831332912192,
+      "loss": 4.8544,
+      "step": 1100
+    },
+    {
+      "epoch": 0.3490484087499013,
+      "grad_norm": 0.470757839112991,
+      "learning_rate": 0.0006980416929879974,
+      "loss": 4.7676,
+      "step": 1105
+    },
+    {
+      "epoch": 0.3506278133143805,
+      "grad_norm": 0.4198823601800511,
+      "learning_rate": 0.0007012002526847757,
+      "loss": 4.721,
+      "step": 1110
+    },
+    {
+      "epoch": 0.35220721787885967,
+      "grad_norm": 0.34137569065032386,
+      "learning_rate": 0.0007043588123815539,
+      "loss": 4.7626,
+      "step": 1115
+    },
+    {
+      "epoch": 0.35378662244333886,
+      "grad_norm": 0.49971050328089983,
+      "learning_rate": 0.0007075173720783323,
+      "loss": 4.8122,
+      "step": 1120
+    },
+    {
+      "epoch": 0.35536602700781805,
+      "grad_norm": 0.3953000041331016,
+      "learning_rate": 0.0007106759317751105,
+      "loss": 4.744,
+      "step": 1125
+    },
+    {
+      "epoch": 0.35694543157229724,
+      "grad_norm": 0.3817554183346107,
+      "learning_rate": 0.0007138344914718888,
+      "loss": 4.7326,
+      "step": 1130
+    },
+    {
+      "epoch": 0.35852483613677644,
+      "grad_norm": 0.3065329226094244,
+      "learning_rate": 0.000716993051168667,
+      "loss": 4.6795,
+      "step": 1135
+    },
+    {
+      "epoch": 0.36010424070125563,
+      "grad_norm": 0.2771623794292736,
+      "learning_rate": 0.0007201516108654453,
+      "loss": 4.6518,
+      "step": 1140
+    },
+    {
+      "epoch": 0.3616836452657348,
+      "grad_norm": 0.35400731064131585,
+      "learning_rate": 0.0007233101705622235,
+      "loss": 4.7473,
+      "step": 1145
+    },
+    {
+      "epoch": 0.363263049830214,
+      "grad_norm": 0.3484555026524429,
+      "learning_rate": 0.000726468730259002,
+      "loss": 4.6603,
+      "step": 1150
+    },
+    {
+      "epoch": 0.3648424543946932,
+      "grad_norm": 0.2450169475092362,
+      "learning_rate": 0.0007296272899557803,
+      "loss": 4.6341,
+      "step": 1155
+    },
+    {
+      "epoch": 0.3664218589591724,
+      "grad_norm": 0.4006637364871361,
+      "learning_rate": 0.0007327858496525585,
+      "loss": 4.6863,
+      "step": 1160
+    },
+    {
+      "epoch": 0.3680012635236516,
+      "grad_norm": 0.45669146482081835,
+      "learning_rate": 0.0007359444093493367,
+      "loss": 4.6163,
+      "step": 1165
+    },
+    {
+      "epoch": 0.3695806680881308,
+      "grad_norm": 0.28184437239622734,
+      "learning_rate": 0.000739102969046115,
+      "loss": 4.6583,
+      "step": 1170
+    },
+    {
+      "epoch": 0.37116007265261,
+      "grad_norm": 0.36542190468548746,
+      "learning_rate": 0.0007422615287428932,
+      "loss": 4.626,
+      "step": 1175
+    },
+    {
+      "epoch": 0.37273947721708917,
+      "grad_norm": 0.3211192511657023,
+      "learning_rate": 0.0007454200884396716,
+      "loss": 4.7379,
+      "step": 1180
+    },
+    {
+      "epoch": 0.37431888178156836,
+      "grad_norm": 0.20304048450743567,
+      "learning_rate": 0.0007485786481364499,
+      "loss": 4.6317,
+      "step": 1185
+    },
+    {
+      "epoch": 0.37589828634604755,
+      "grad_norm": 0.3540717270875546,
+      "learning_rate": 0.0007517372078332281,
+      "loss": 4.5685,
+      "step": 1190
+    },
+    {
+      "epoch": 0.37747769091052674,
+      "grad_norm": 0.2615691483415603,
+      "learning_rate": 0.0007548957675300064,
+      "loss": 4.6439,
+      "step": 1195
+    },
+    {
+      "epoch": 0.37905709547500593,
+      "grad_norm": 0.3503676341696831,
+      "learning_rate": 0.0007580543272267846,
+      "loss": 4.6472,
+      "step": 1200
+    },
+    {
+      "epoch": 0.3806365000394851,
+      "grad_norm": 0.381747407284883,
+      "learning_rate": 0.0007612128869235628,
+      "loss": 4.5915,
+      "step": 1205
+    },
+    {
+      "epoch": 0.3822159046039643,
+      "grad_norm": 0.5555220296942412,
+      "learning_rate": 0.0007643714466203412,
+      "loss": 4.5812,
+      "step": 1210
+    },
+    {
+      "epoch": 0.3837953091684435,
+      "grad_norm": 0.5472892354048585,
+      "learning_rate": 0.0007675300063171195,
+      "loss": 4.6288,
+      "step": 1215
+    },
+    {
+      "epoch": 0.3853747137329227,
+      "grad_norm": 0.41350514353566703,
+      "learning_rate": 0.0007706885660138977,
+      "loss": 4.511,
+      "step": 1220
+    },
+    {
+      "epoch": 0.3869541182974019,
+      "grad_norm": 0.30935412837157134,
+      "learning_rate": 0.000773847125710676,
+      "loss": 4.5887,
+      "step": 1225
+    },
+    {
+      "epoch": 0.3885335228618811,
+      "grad_norm": 0.45469414600545827,
+      "learning_rate": 0.0007770056854074542,
+      "loss": 4.545,
+      "step": 1230
+    },
+    {
+      "epoch": 0.3901129274263603,
+      "grad_norm": 0.45367406395912707,
+      "learning_rate": 0.0007801642451042324,
+      "loss": 4.6878,
+      "step": 1235
+    },
+    {
+      "epoch": 0.39169233199083947,
+      "grad_norm": 0.47051545366899983,
+      "learning_rate": 0.0007833228048010107,
+      "loss": 4.5098,
+      "step": 1240
+    },
+    {
+      "epoch": 0.39327173655531866,
+      "grad_norm": 0.356870891214942,
+      "learning_rate": 0.0007864813644977891,
+      "loss": 4.6192,
+      "step": 1245
+    },
+    {
+      "epoch": 0.39485114111979785,
+      "grad_norm": 0.3398669537171447,
+      "learning_rate": 0.0007896399241945673,
+      "loss": 4.553,
+      "step": 1250
+    },
+    {
+      "epoch": 0.39643054568427705,
+      "grad_norm": 0.38876482712395577,
+      "learning_rate": 0.0007927984838913456,
+      "loss": 4.4853,
+      "step": 1255
+    },
+    {
+      "epoch": 0.39800995024875624,
+      "grad_norm": 0.4094421799432949,
+      "learning_rate": 0.0007959570435881238,
+      "loss": 4.4592,
+      "step": 1260
+    },
+    {
+      "epoch": 0.39958935481323543,
+      "grad_norm": 0.39385781540672343,
+      "learning_rate": 0.000799115603284902,
+      "loss": 4.5032,
+      "step": 1265
+    },
+    {
+      "epoch": 0.4011687593777146,
+      "grad_norm": 0.5817554845495431,
+      "learning_rate": 0.0008022741629816803,
+      "loss": 4.649,
+      "step": 1270
+    },
+    {
+      "epoch": 0.4027481639421938,
+      "grad_norm": 0.4778375510654764,
+      "learning_rate": 0.0008054327226784587,
+      "loss": 4.5617,
+      "step": 1275
+    },
+    {
+      "epoch": 0.404327568506673,
+      "grad_norm": 0.3747112096903426,
+      "learning_rate": 0.0008085912823752369,
+      "loss": 4.4785,
+      "step": 1280
+    },
+    {
+      "epoch": 0.4059069730711522,
+      "grad_norm": 0.2862809855465098,
+      "learning_rate": 0.0008117498420720152,
+      "loss": 4.5621,
+      "step": 1285
+    },
+    {
+      "epoch": 0.4074863776356314,
+      "grad_norm": 0.2670050072937441,
+      "learning_rate": 0.0008149084017687934,
+      "loss": 4.5072,
+      "step": 1290
+    },
+    {
+      "epoch": 0.4090657822001106,
+      "grad_norm": 0.27279766937650357,
+      "learning_rate": 0.0008180669614655717,
+      "loss": 4.473,
+      "step": 1295
+    },
+    {
+      "epoch": 0.4106451867645898,
+      "grad_norm": 0.3543901390752811,
+      "learning_rate": 0.0008212255211623499,
+      "loss": 4.4827,
+      "step": 1300
+    },
+    {
+      "epoch": 0.41222459132906897,
+      "grad_norm": 0.4597168775174795,
+      "learning_rate": 0.0008243840808591283,
+      "loss": 4.4519,
+      "step": 1305
+    },
+    {
+      "epoch": 0.41380399589354816,
+      "grad_norm": 0.3161074895361928,
+      "learning_rate": 0.0008275426405559065,
+      "loss": 4.3894,
+      "step": 1310
+    },
+    {
+      "epoch": 0.4153834004580273,
+      "grad_norm": 0.3031787080639516,
+      "learning_rate": 0.0008307012002526848,
+      "loss": 4.3778,
+      "step": 1315
+    },
+    {
+      "epoch": 0.4169628050225065,
+      "grad_norm": 0.4046902940993039,
+      "learning_rate": 0.0008338597599494631,
+      "loss": 4.5039,
+      "step": 1320
+    },
+    {
+      "epoch": 0.4185422095869857,
+      "grad_norm": 0.28233846625561443,
+      "learning_rate": 0.0008370183196462414,
+      "loss": 4.4458,
+      "step": 1325
+    },
+    {
+      "epoch": 0.4201216141514649,
+      "grad_norm": 0.5878023927754359,
+      "learning_rate": 0.0008401768793430196,
+      "loss": 4.4569,
+      "step": 1330
+    },
+    {
+      "epoch": 0.42170101871594406,
+      "grad_norm": 0.28491144301264926,
+      "learning_rate": 0.000843335439039798,
+      "loss": 4.4164,
+      "step": 1335
+    },
+    {
+      "epoch": 0.42328042328042326,
+      "grad_norm": 0.339216984189955,
+      "learning_rate": 0.0008464939987365762,
+      "loss": 4.4052,
+      "step": 1340
+    },
+    {
+      "epoch": 0.42485982784490245,
+      "grad_norm": 0.3287668012766888,
+      "learning_rate": 0.0008496525584333545,
+      "loss": 4.3853,
+      "step": 1345
+    },
+    {
+      "epoch": 0.42643923240938164,
+      "grad_norm": 0.3421773901777664,
+      "learning_rate": 0.0008528111181301327,
+      "loss": 4.442,
+      "step": 1350
+    },
+    {
+      "epoch": 0.42801863697386083,
+      "grad_norm": 0.3292206362755166,
+      "learning_rate": 0.000855969677826911,
+      "loss": 4.4226,
+      "step": 1355
+    },
+    {
+      "epoch": 0.42959804153834,
+      "grad_norm": 0.3371728323598187,
+      "learning_rate": 0.0008591282375236892,
+      "loss": 4.4446,
+      "step": 1360
+    },
+    {
+      "epoch": 0.4311774461028192,
+      "grad_norm": 0.27079521890457803,
+      "learning_rate": 0.0008622867972204675,
+      "loss": 4.3547,
+      "step": 1365
+    },
+    {
+      "epoch": 0.4327568506672984,
+      "grad_norm": 0.2622769562992865,
+      "learning_rate": 0.0008654453569172458,
+      "loss": 4.3783,
+      "step": 1370
+    },
+    {
+      "epoch": 0.4343362552317776,
+      "grad_norm": 0.28526124910101036,
+      "learning_rate": 0.0008686039166140241,
+      "loss": 4.4134,
+      "step": 1375
+    },
+    {
+      "epoch": 0.4359156597962568,
+      "grad_norm": 0.4703111974864806,
+      "learning_rate": 0.0008717624763108023,
+      "loss": 4.3905,
+      "step": 1380
+    },
+    {
+      "epoch": 0.437495064360736,
+      "grad_norm": 0.42877065456487523,
+      "learning_rate": 0.0008749210360075806,
+      "loss": 4.3255,
+      "step": 1385
+    },
+    {
+      "epoch": 0.4390744689252152,
+      "grad_norm": 0.2818515041639332,
+      "learning_rate": 0.0008780795957043588,
+      "loss": 4.4326,
+      "step": 1390
+    },
+    {
+      "epoch": 0.44065387348969437,
+      "grad_norm": 0.3386738826947052,
+      "learning_rate": 0.0008812381554011371,
+      "loss": 4.3857,
+      "step": 1395
+    },
+    {
+      "epoch": 0.44223327805417356,
+      "grad_norm": 0.38221324112499194,
+      "learning_rate": 0.0008843967150979154,
+      "loss": 4.4072,
+      "step": 1400
+    },
+    {
+      "epoch": 0.44381268261865275,
+      "grad_norm": 0.4550909203693113,
+      "learning_rate": 0.0008875552747946937,
+      "loss": 4.3394,
+      "step": 1405
+    },
+    {
+      "epoch": 0.44539208718313195,
+      "grad_norm": 0.5188511582204778,
+      "learning_rate": 0.0008907138344914719,
+      "loss": 4.3843,
+      "step": 1410
+    },
+    {
+      "epoch": 0.44697149174761114,
+      "grad_norm": 0.37563874540589287,
+      "learning_rate": 0.0008938723941882502,
+      "loss": 4.3411,
+      "step": 1415
+    },
+    {
+      "epoch": 0.44855089631209033,
+      "grad_norm": 0.5066382494462849,
+      "learning_rate": 0.0008970309538850284,
+      "loss": 4.4086,
+      "step": 1420
+    },
+    {
+      "epoch": 0.4501303008765695,
+      "grad_norm": 0.354382674134275,
+      "learning_rate": 0.0009001895135818067,
+      "loss": 4.3791,
+      "step": 1425
+    },
+    {
+      "epoch": 0.4517097054410487,
+      "grad_norm": 0.3153630402026596,
+      "learning_rate": 0.000903348073278585,
+      "loss": 4.3098,
+      "step": 1430
+    },
+    {
+      "epoch": 0.4532891100055279,
+      "grad_norm": 0.2247411189874934,
+      "learning_rate": 0.0009065066329753633,
+      "loss": 4.3073,
+      "step": 1435
+    },
+    {
+      "epoch": 0.4548685145700071,
+      "grad_norm": 0.3084869625887127,
+      "learning_rate": 0.0009096651926721415,
+      "loss": 4.2301,
+      "step": 1440
+    },
+    {
+      "epoch": 0.4564479191344863,
+      "grad_norm": 0.2683775947502175,
+      "learning_rate": 0.0009128237523689198,
+      "loss": 4.3069,
+      "step": 1445
+    },
+    {
+      "epoch": 0.4580273236989655,
+      "grad_norm": 0.19696762482110167,
+      "learning_rate": 0.000915982312065698,
+      "loss": 4.3117,
+      "step": 1450
+    },
+    {
+      "epoch": 0.4596067282634447,
+      "grad_norm": 0.31780938882297877,
+      "learning_rate": 0.0009191408717624763,
+      "loss": 4.3639,
+      "step": 1455
+    },
+    {
+      "epoch": 0.46118613282792387,
+      "grad_norm": 0.34985785227208677,
+      "learning_rate": 0.0009222994314592545,
+      "loss": 4.2762,
+      "step": 1460
+    },
+    {
+      "epoch": 0.46276553739240306,
+      "grad_norm": 0.44099480740170816,
+      "learning_rate": 0.0009254579911560329,
+      "loss": 4.3993,
+      "step": 1465
+    },
+    {
+      "epoch": 0.46434494195688225,
+      "grad_norm": 0.3209951859089135,
+      "learning_rate": 0.0009286165508528111,
+      "loss": 4.3052,
+      "step": 1470
+    },
+    {
+      "epoch": 0.46592434652136144,
+      "grad_norm": 0.37335084714596994,
+      "learning_rate": 0.0009317751105495894,
+      "loss": 4.2903,
+      "step": 1475
+    },
+    {
+      "epoch": 0.46750375108584064,
+      "grad_norm": 0.2842543897263021,
+      "learning_rate": 0.0009349336702463676,
+      "loss": 4.2876,
+      "step": 1480
+    },
+    {
+      "epoch": 0.4690831556503198,
+      "grad_norm": 0.30335571616435514,
+      "learning_rate": 0.0009380922299431459,
+      "loss": 4.235,
+      "step": 1485
+    },
+    {
+      "epoch": 0.470662560214799,
+      "grad_norm": 0.266764361136666,
+      "learning_rate": 0.0009412507896399241,
+      "loss": 4.2846,
+      "step": 1490
+    },
+    {
+      "epoch": 0.4722419647792782,
+      "grad_norm": 0.3519813365858386,
+      "learning_rate": 0.0009444093493367026,
+      "loss": 4.3133,
+      "step": 1495
+    },
+    {
+      "epoch": 0.4738213693437574,
+      "grad_norm": 0.2742010924034352,
+      "learning_rate": 0.0009475679090334808,
+      "loss": 4.2552,
+      "step": 1500
+    },
+    {
+      "epoch": 0.4754007739082366,
+      "grad_norm": 0.28659499947688005,
+      "learning_rate": 0.0009507264687302591,
+      "loss": 4.276,
+      "step": 1505
+    },
+    {
+      "epoch": 0.4769801784727158,
+      "grad_norm": 0.26463115258969583,
+      "learning_rate": 0.0009538850284270373,
+      "loss": 4.2585,
+      "step": 1510
+    },
+    {
+      "epoch": 0.478559583037195,
+      "grad_norm": 0.4491072115265183,
+      "learning_rate": 0.0009570435881238156,
+      "loss": 4.2574,
+      "step": 1515
+    },
+    {
+      "epoch": 0.48013898760167417,
+      "grad_norm": 0.3236579306871126,
+      "learning_rate": 0.0009602021478205938,
+      "loss": 4.324,
+      "step": 1520
+    },
+    {
+      "epoch": 0.48171839216615336,
+      "grad_norm": 0.28007607030549353,
+      "learning_rate": 0.0009633607075173722,
+      "loss": 4.3264,
+      "step": 1525
+    },
+    {
+      "epoch": 0.48329779673063256,
+      "grad_norm": 0.4518966964830614,
+      "learning_rate": 0.0009665192672141504,
+      "loss": 4.213,
+      "step": 1530
+    },
+    {
+      "epoch": 0.48487720129511175,
+      "grad_norm": 0.29253505122413304,
+      "learning_rate": 0.0009696778269109287,
+      "loss": 4.3187,
+      "step": 1535
+    },
+    {
+      "epoch": 0.48645660585959094,
+      "grad_norm": 0.3760216806664558,
+      "learning_rate": 0.0009728363866077069,
+      "loss": 4.2086,
+      "step": 1540
+    },
+    {
+      "epoch": 0.48803601042407013,
+      "grad_norm": 0.3462163229846517,
+      "learning_rate": 0.0009759949463044852,
+      "loss": 4.2029,
+      "step": 1545
+    },
+    {
+      "epoch": 0.4896154149885493,
+      "grad_norm": 0.4678230022826408,
+      "learning_rate": 0.0009791535060012634,
+      "loss": 4.2381,
+      "step": 1550
+    },
+    {
+      "epoch": 0.4911948195530285,
+      "grad_norm": 0.281033090745962,
+      "learning_rate": 0.0009823120656980418,
+      "loss": 4.1577,
+      "step": 1555
+    },
+    {
+      "epoch": 0.4927742241175077,
+      "grad_norm": 0.25715655179867125,
+      "learning_rate": 0.00098547062539482,
+      "loss": 4.3205,
+      "step": 1560
+    },
+    {
+      "epoch": 0.4943536286819869,
+      "grad_norm": 0.2377083515339439,
+      "learning_rate": 0.0009886291850915983,
+      "loss": 4.1485,
+      "step": 1565
+    },
+    {
+      "epoch": 0.4959330332464661,
+      "grad_norm": 0.34877478211376156,
+      "learning_rate": 0.0009917877447883764,
+      "loss": 4.2703,
+      "step": 1570
+    },
+    {
+      "epoch": 0.4975124378109453,
+      "grad_norm": 0.3612121180047406,
+      "learning_rate": 0.0009949463044851548,
+      "loss": 4.2981,
+      "step": 1575
+    },
+    {
+      "epoch": 0.4990918423754245,
+      "grad_norm": 0.32745245414352575,
+      "learning_rate": 0.000998104864181933,
+      "loss": 4.188,
+      "step": 1580
+    },
+    {
+      "epoch": 0.5006712469399036,
+      "grad_norm": 0.36513716463433904,
+      "learning_rate": 0.0009999999513416054,
+      "loss": 4.2169,
+      "step": 1585
+    },
+    {
+      "epoch": 0.5022506515043829,
+      "grad_norm": 0.48696761580584264,
+      "learning_rate": 0.0009999994039347757,
+      "loss": 4.1734,
+      "step": 1590
+    },
+    {
+      "epoch": 0.503830056068862,
+      "grad_norm": 0.28706121610340524,
+      "learning_rate": 0.000999998248298791,
+      "loss": 4.2644,
+      "step": 1595
+    },
+    {
+      "epoch": 0.5054094606333412,
+      "grad_norm": 0.28463581068533467,
+      "learning_rate": 0.0009999964844350573,
+      "loss": 4.1033,
+      "step": 1600
+    },
+    {
+      "epoch": 0.5069888651978204,
+      "grad_norm": 0.281432611299984,
+      "learning_rate": 0.0009999941123457203,
+      "loss": 4.2137,
+      "step": 1605
+    },
+    {
+      "epoch": 0.5085682697622996,
+      "grad_norm": 0.2780914746609548,
+      "learning_rate": 0.0009999911320336655,
+      "loss": 4.1156,
+      "step": 1610
+    },
+    {
+      "epoch": 0.5101476743267788,
+      "grad_norm": 0.24775313426165027,
+      "learning_rate": 0.000999987543502518,
+      "loss": 4.1227,
+      "step": 1615
+    },
+    {
+      "epoch": 0.511727078891258,
+      "grad_norm": 0.36278436207383286,
+      "learning_rate": 0.0009999833467566437,
+      "loss": 4.178,
+      "step": 1620
+    },
+    {
+      "epoch": 0.5133064834557372,
+      "grad_norm": 0.1932778698252279,
+      "learning_rate": 0.0009999785418011472,
+      "loss": 4.1866,
+      "step": 1625
+    },
+    {
+      "epoch": 0.5148858880202164,
+      "grad_norm": 0.39153360957933375,
+      "learning_rate": 0.000999973128641874,
+      "loss": 4.2506,
+      "step": 1630
+    },
+    {
+      "epoch": 0.5164652925846955,
+      "grad_norm": 0.37554437506264593,
+      "learning_rate": 0.000999967107285409,
+      "loss": 4.117,
+      "step": 1635
+    },
+    {
+      "epoch": 0.5180446971491748,
+      "grad_norm": 0.28958701143166327,
+      "learning_rate": 0.0009999604777390762,
+      "loss": 4.2216,
+      "step": 1640
+    },
+    {
+      "epoch": 0.5196241017136539,
+      "grad_norm": 0.30198070915798164,
+      "learning_rate": 0.0009999532400109413,
+      "loss": 4.2161,
+      "step": 1645
+    },
+    {
+      "epoch": 0.5212035062781332,
+      "grad_norm": 0.3428198600005329,
+      "learning_rate": 0.0009999453941098076,
+      "loss": 4.0479,
+      "step": 1650
+    },
+    {
+      "epoch": 0.5227829108426123,
+      "grad_norm": 0.299713880431016,
+      "learning_rate": 0.0009999369400452201,
+      "loss": 4.1104,
+      "step": 1655
+    },
+    {
+      "epoch": 0.5243623154070916,
+      "grad_norm": 0.28532084858377904,
+      "learning_rate": 0.0009999278778274625,
+      "loss": 4.17,
+      "step": 1660
+    },
+    {
+      "epoch": 0.5259417199715707,
+      "grad_norm": 0.2984801615394102,
+      "learning_rate": 0.0009999182074675588,
+      "loss": 4.0912,
+      "step": 1665
+    },
+    {
+      "epoch": 0.5275211245360499,
+      "grad_norm": 0.25676781344144184,
+      "learning_rate": 0.0009999079289772722,
+      "loss": 4.0482,
+      "step": 1670
+    },
+    {
+      "epoch": 0.5291005291005291,
+      "grad_norm": 0.26759080611381303,
+      "learning_rate": 0.0009998970423691067,
+      "loss": 4.0194,
+      "step": 1675
+    },
+    {
+      "epoch": 0.5306799336650083,
+      "grad_norm": 0.2764566522384514,
+      "learning_rate": 0.0009998855476563051,
+      "loss": 4.1851,
+      "step": 1680
+    },
+    {
+      "epoch": 0.5322593382294875,
+      "grad_norm": 0.2904154144075542,
+      "learning_rate": 0.00099987344485285,
+      "loss": 4.0868,
+      "step": 1685
+    },
+    {
+      "epoch": 0.5338387427939667,
+      "grad_norm": 0.3681853600736386,
+      "learning_rate": 0.0009998607339734642,
+      "loss": 4.1386,
+      "step": 1690
+    },
+    {
+      "epoch": 0.5354181473584458,
+      "grad_norm": 0.4240926663878961,
+      "learning_rate": 0.0009998474150336102,
+      "loss": 4.0986,
+      "step": 1695
+    },
+    {
+      "epoch": 0.5369975519229251,
+      "grad_norm": 0.3089726259137687,
+      "learning_rate": 0.0009998334880494896,
+      "loss": 4.1597,
+      "step": 1700
+    },
+    {
+      "epoch": 0.5385769564874042,
+      "grad_norm": 0.30656590447426424,
+      "learning_rate": 0.000999818953038044,
+      "loss": 3.9891,
+      "step": 1705
+    },
+    {
+      "epoch": 0.5401563610518835,
+      "grad_norm": 0.32114946179568205,
+      "learning_rate": 0.0009998038100169553,
+      "loss": 4.1324,
+      "step": 1710
+    },
+    {
+      "epoch": 0.5417357656163626,
+      "grad_norm": 0.32282723078574843,
+      "learning_rate": 0.0009997880590046436,
+      "loss": 4.0553,
+      "step": 1715
+    },
+    {
+      "epoch": 0.5433151701808419,
+      "grad_norm": 0.21075665879048797,
+      "learning_rate": 0.0009997717000202696,
+      "loss": 4.0876,
+      "step": 1720
+    },
+    {
+      "epoch": 0.544894574745321,
+      "grad_norm": 0.2642191066111579,
+      "learning_rate": 0.0009997547330837335,
+      "loss": 3.9771,
+      "step": 1725
+    },
+    {
+      "epoch": 0.5464739793098002,
+      "grad_norm": 0.2241664738813061,
+      "learning_rate": 0.0009997371582156746,
+      "loss": 4.1472,
+      "step": 1730
+    },
+    {
+      "epoch": 0.5480533838742794,
+      "grad_norm": 0.21223821355830494,
+      "learning_rate": 0.0009997189754374725,
+      "loss": 4.1219,
+      "step": 1735
+    },
+    {
+      "epoch": 0.5496327884387586,
+      "grad_norm": 0.30437709086967074,
+      "learning_rate": 0.0009997001847712455,
+      "loss": 4.0623,
+      "step": 1740
+    },
+    {
+      "epoch": 0.5512121930032378,
+      "grad_norm": 0.2612037043109827,
+      "learning_rate": 0.0009996807862398516,
+      "loss": 4.1017,
+      "step": 1745
+    },
+    {
+      "epoch": 0.552791597567717,
+      "grad_norm": 0.21637469222681385,
+      "learning_rate": 0.0009996607798668886,
+      "loss": 4.0922,
+      "step": 1750
+    },
+    {
+      "epoch": 0.5543710021321961,
+      "grad_norm": 0.1865321522736776,
+      "learning_rate": 0.0009996401656766933,
+      "loss": 4.1306,
+      "step": 1755
+    },
+    {
+      "epoch": 0.5559504066966754,
+      "grad_norm": 0.27200198732458336,
+      "learning_rate": 0.000999618943694342,
+      "loss": 4.0081,
+      "step": 1760
+    },
+    {
+      "epoch": 0.5575298112611545,
+      "grad_norm": 0.3202280293025325,
+      "learning_rate": 0.0009995971139456503,
+      "loss": 4.0094,
+      "step": 1765
+    },
+    {
+      "epoch": 0.5591092158256338,
+      "grad_norm": 0.3869093315077745,
+      "learning_rate": 0.0009995746764571735,
+      "loss": 4.0755,
+      "step": 1770
+    },
+    {
+      "epoch": 0.5606886203901129,
+      "grad_norm": 0.38187337349101136,
+      "learning_rate": 0.0009995516312562057,
+      "loss": 4.1702,
+      "step": 1775
+    },
+    {
+      "epoch": 0.5622680249545922,
+      "grad_norm": 0.31455911111783563,
+      "learning_rate": 0.0009995279783707805,
+      "loss": 3.979,
+      "step": 1780
+    },
+    {
+      "epoch": 0.5638474295190713,
+      "grad_norm": 0.29574697329192573,
+      "learning_rate": 0.0009995037178296708,
+      "loss": 4.154,
+      "step": 1785
+    },
+    {
+      "epoch": 0.5654268340835505,
+      "grad_norm": 0.40103311872700226,
+      "learning_rate": 0.0009994788496623882,
+      "loss": 4.0433,
+      "step": 1790
+    },
+    {
+      "epoch": 0.5670062386480297,
+      "grad_norm": 0.3657271711780575,
+      "learning_rate": 0.0009994533738991844,
+      "loss": 4.0775,
+      "step": 1795
+    },
+    {
+      "epoch": 0.5685856432125089,
+      "grad_norm": 0.34299715107279133,
+      "learning_rate": 0.000999427290571049,
+      "loss": 4.1602,
+      "step": 1800
+    },
+    {
+      "epoch": 0.5701650477769881,
+      "grad_norm": 0.31471328128215614,
+      "learning_rate": 0.000999400599709712,
+      "loss": 4.1397,
+      "step": 1805
+    },
+    {
+      "epoch": 0.5717444523414673,
+      "grad_norm": 0.3249407191646648,
+      "learning_rate": 0.000999373301347641,
+      "loss": 4.0178,
+      "step": 1810
+    },
+    {
+      "epoch": 0.5733238569059464,
+      "grad_norm": 0.3627998575778931,
+      "learning_rate": 0.000999345395518044,
+      "loss": 4.073,
+      "step": 1815
+    },
+    {
+      "epoch": 0.5749032614704257,
+      "grad_norm": 0.3008083379342741,
+      "learning_rate": 0.0009993168822548671,
+      "loss": 4.0813,
+      "step": 1820
+    },
+    {
+      "epoch": 0.5764826660349048,
+      "grad_norm": 0.2922414382664129,
+      "learning_rate": 0.0009992877615927955,
+      "loss": 3.9915,
+      "step": 1825
+    },
+    {
+      "epoch": 0.5780620705993841,
+      "grad_norm": 0.281765440169215,
+      "learning_rate": 0.0009992580335672534,
+      "loss": 4.0479,
+      "step": 1830
+    },
+    {
+      "epoch": 0.5796414751638632,
+      "grad_norm": 0.2199768873265984,
+      "learning_rate": 0.0009992276982144035,
+      "loss": 3.9814,
+      "step": 1835
+    },
+    {
+      "epoch": 0.5812208797283425,
+      "grad_norm": 0.26746177527158255,
+      "learning_rate": 0.000999196755571148,
+      "loss": 4.0101,
+      "step": 1840
+    },
+    {
+      "epoch": 0.5828002842928216,
+      "grad_norm": 0.27475241843335857,
+      "learning_rate": 0.0009991652056751269,
+      "loss": 4.0526,
+      "step": 1845
+    },
+    {
+      "epoch": 0.5843796888573008,
+      "grad_norm": 0.20515852596205633,
+      "learning_rate": 0.0009991330485647194,
+      "loss": 4.0069,
+      "step": 1850
+    },
+    {
+      "epoch": 0.58595909342178,
+      "grad_norm": 0.23220844607256083,
+      "learning_rate": 0.0009991002842790438,
+      "loss": 4.1066,
+      "step": 1855
+    },
+    {
+      "epoch": 0.5875384979862592,
+      "grad_norm": 0.24900266206103544,
+      "learning_rate": 0.000999066912857956,
+      "loss": 4.0527,
+      "step": 1860
+    },
+    {
+      "epoch": 0.5891179025507384,
+      "grad_norm": 0.22772261845450373,
+      "learning_rate": 0.0009990329343420514,
+      "loss": 4.0552,
+      "step": 1865
+    },
+    {
+      "epoch": 0.5906973071152176,
+      "grad_norm": 0.34846084321759413,
+      "learning_rate": 0.0009989983487726632,
+      "loss": 4.0956,
+      "step": 1870
+    },
+    {
+      "epoch": 0.5922767116796968,
+      "grad_norm": 0.20306318495273795,
+      "learning_rate": 0.0009989631561918635,
+      "loss": 3.9788,
+      "step": 1875
+    },
+    {
+      "epoch": 0.5938561162441759,
+      "grad_norm": 0.3054746048484631,
+      "learning_rate": 0.0009989273566424629,
+      "loss": 4.0408,
+      "step": 1880
+    },
+    {
+      "epoch": 0.5954355208086551,
+      "grad_norm": 0.2689676990909977,
+      "learning_rate": 0.0009988909501680095,
+      "loss": 4.055,
+      "step": 1885
+    },
+    {
+      "epoch": 0.5970149253731343,
+      "grad_norm": 0.2879295565501649,
+      "learning_rate": 0.0009988539368127908,
+      "loss": 3.9719,
+      "step": 1890
+    },
+    {
+      "epoch": 0.5985943299376135,
+      "grad_norm": 0.25177642807914086,
+      "learning_rate": 0.000998816316621832,
+      "loss": 4.0066,
+      "step": 1895
+    },
+    {
+      "epoch": 0.6001737345020927,
+      "grad_norm": 0.32762881060791094,
+      "learning_rate": 0.0009987780896408963,
+      "loss": 3.8824,
+      "step": 1900
+    },
+    {
+      "epoch": 0.6017531390665719,
+      "grad_norm": 0.26143752591314573,
+      "learning_rate": 0.0009987392559164857,
+      "loss": 4.1152,
+      "step": 1905
+    },
+    {
+      "epoch": 0.603332543631051,
+      "grad_norm": 0.24110745226470215,
+      "learning_rate": 0.0009986998154958395,
+      "loss": 3.9775,
+      "step": 1910
+    },
+    {
+      "epoch": 0.6049119481955303,
+      "grad_norm": 0.24195042625692204,
+      "learning_rate": 0.0009986597684269354,
+      "loss": 3.9366,
+      "step": 1915
+    },
+    {
+      "epoch": 0.6064913527600094,
+      "grad_norm": 0.21330938341685474,
+      "learning_rate": 0.0009986191147584892,
+      "loss": 3.9408,
+      "step": 1920
+    },
+    {
+      "epoch": 0.6080707573244887,
+      "grad_norm": 0.23539885796681012,
+      "learning_rate": 0.0009985778545399545,
+      "loss": 4.035,
+      "step": 1925
+    },
+    {
+      "epoch": 0.6096501618889678,
+      "grad_norm": 0.2838008993926215,
+      "learning_rate": 0.0009985359878215223,
+      "loss": 4.0331,
+      "step": 1930
+    },
+    {
+      "epoch": 0.6112295664534471,
+      "grad_norm": 0.28796522858528995,
+      "learning_rate": 0.0009984935146541223,
+      "loss": 3.9946,
+      "step": 1935
+    },
+    {
+      "epoch": 0.6128089710179262,
+      "grad_norm": 0.25284741162200397,
+      "learning_rate": 0.0009984504350894212,
+      "loss": 4.1193,
+      "step": 1940
+    },
+    {
+      "epoch": 0.6143883755824054,
+      "grad_norm": 0.1883166471785072,
+      "learning_rate": 0.0009984067491798235,
+      "loss": 3.9263,
+      "step": 1945
+    },
+    {
+      "epoch": 0.6159677801468846,
+      "grad_norm": 0.31448053784092006,
+      "learning_rate": 0.0009983624569784714,
+      "loss": 3.8846,
+      "step": 1950
+    },
+    {
+      "epoch": 0.6175471847113638,
+      "grad_norm": 0.24984288586441813,
+      "learning_rate": 0.0009983175585392445,
+      "loss": 3.8951,
+      "step": 1955
+    },
+    {
+      "epoch": 0.619126589275843,
+      "grad_norm": 0.2834739736737687,
+      "learning_rate": 0.00099827205391676,
+      "loss": 3.8461,
+      "step": 1960
+    },
+    {
+      "epoch": 0.6207059938403222,
+      "grad_norm": 0.3185907483566412,
+      "learning_rate": 0.0009982259431663724,
+      "loss": 3.9755,
+      "step": 1965
+    },
+    {
+      "epoch": 0.6222853984048013,
+      "grad_norm": 0.29327583353771697,
+      "learning_rate": 0.0009981792263441737,
+      "loss": 3.9467,
+      "step": 1970
+    },
+    {
+      "epoch": 0.6238648029692806,
+      "grad_norm": 0.29352858594845627,
+      "learning_rate": 0.0009981319035069932,
+      "loss": 4.0014,
+      "step": 1975
+    },
+    {
+      "epoch": 0.6254442075337597,
+      "grad_norm": 0.25867595966057094,
+      "learning_rate": 0.0009980839747123966,
+      "loss": 3.9775,
+      "step": 1980
+    },
+    {
+      "epoch": 0.627023612098239,
+      "grad_norm": 0.25568674878218955,
+      "learning_rate": 0.000998035440018688,
+      "loss": 3.9609,
+      "step": 1985
+    },
+    {
+      "epoch": 0.6286030166627181,
+      "grad_norm": 0.3211767242169351,
+      "learning_rate": 0.0009979862994849073,
+      "loss": 3.9838,
+      "step": 1990
+    },
+    {
+      "epoch": 0.6301824212271974,
+      "grad_norm": 0.22478636571958996,
+      "learning_rate": 0.0009979365531708325,
+      "loss": 3.9142,
+      "step": 1995
+    },
+    {
+      "epoch": 0.6317618257916765,
+      "grad_norm": 0.2892618804449103,
+      "learning_rate": 0.0009978862011369779,
+      "loss": 3.8985,
+      "step": 2000
+    },
+    {
+      "epoch": 0.6333412303561557,
+      "grad_norm": 0.24435044534248943,
+      "learning_rate": 0.0009978352434445944,
+      "loss": 3.8645,
+      "step": 2005
+    },
+    {
+      "epoch": 0.6349206349206349,
+      "grad_norm": 0.21458490976951505,
+      "learning_rate": 0.0009977836801556704,
+      "loss": 3.8908,
+      "step": 2010
+    },
+    {
+      "epoch": 0.6365000394851141,
+      "grad_norm": 0.238296404339169,
+      "learning_rate": 0.0009977315113329304,
+      "loss": 3.8564,
+      "step": 2015
+    },
+    {
+      "epoch": 0.6380794440495933,
+      "grad_norm": 0.24198303496426915,
+      "learning_rate": 0.0009976787370398355,
+      "loss": 3.9388,
+      "step": 2020
+    },
+    {
+      "epoch": 0.6396588486140725,
+      "grad_norm": 0.22491929755865114,
+      "learning_rate": 0.0009976253573405838,
+      "loss": 3.9074,
+      "step": 2025
+    },
+    {
+      "epoch": 0.6412382531785517,
+      "grad_norm": 0.20157800604718384,
+      "learning_rate": 0.0009975713723001094,
+      "loss": 3.9252,
+      "step": 2030
+    },
+    {
+      "epoch": 0.6428176577430309,
+      "grad_norm": 0.24707058750275446,
+      "learning_rate": 0.0009975167819840827,
+      "loss": 3.8766,
+      "step": 2035
+    },
+    {
+      "epoch": 0.64439706230751,
+      "grad_norm": 0.1930165825425789,
+      "learning_rate": 0.0009974615864589112,
+      "loss": 3.9563,
+      "step": 2040
+    },
+    {
+      "epoch": 0.6459764668719893,
+      "grad_norm": 0.23592774025549407,
+      "learning_rate": 0.0009974057857917373,
+      "loss": 4.0158,
+      "step": 2045
+    },
+    {
+      "epoch": 0.6475558714364684,
+      "grad_norm": 0.2744523541395304,
+      "learning_rate": 0.000997349380050441,
+      "loss": 4.015,
+      "step": 2050
+    },
+    {
+      "epoch": 0.6491352760009477,
+      "grad_norm": 0.19965782390005965,
+      "learning_rate": 0.0009972923693036373,
+      "loss": 3.8356,
+      "step": 2055
+    },
+    {
+      "epoch": 0.6507146805654268,
+      "grad_norm": 0.2102237724441891,
+      "learning_rate": 0.0009972347536206772,
+      "loss": 4.0032,
+      "step": 2060
+    },
+    {
+      "epoch": 0.652294085129906,
+      "grad_norm": 0.32443690591437163,
+      "learning_rate": 0.0009971765330716482,
+      "loss": 3.8331,
+      "step": 2065
+    },
+    {
+      "epoch": 0.6538734896943852,
+      "grad_norm": 0.2680421026750314,
+      "learning_rate": 0.0009971177077273733,
+      "loss": 3.9182,
+      "step": 2070
+    },
+    {
+      "epoch": 0.6554528942588644,
+      "grad_norm": 0.2608811339656057,
+      "learning_rate": 0.000997058277659411,
+      "loss": 3.9257,
+      "step": 2075
+    },
+    {
+      "epoch": 0.6570322988233436,
+      "grad_norm": 0.25161834765330515,
+      "learning_rate": 0.0009969982429400555,
+      "loss": 3.8755,
+      "step": 2080
+    },
+    {
+      "epoch": 0.6586117033878228,
+      "grad_norm": 0.21108033568964235,
+      "learning_rate": 0.0009969376036423367,
+      "loss": 3.8266,
+      "step": 2085
+    },
+    {
+      "epoch": 0.660191107952302,
+      "grad_norm": 0.2708969117194078,
+      "learning_rate": 0.00099687635984002,
+      "loss": 4.005,
+      "step": 2090
+    },
+    {
+      "epoch": 0.6617705125167812,
+      "grad_norm": 0.234071036138296,
+      "learning_rate": 0.0009968145116076063,
+      "loss": 3.9937,
+      "step": 2095
+    },
+    {
+      "epoch": 0.6633499170812603,
+      "grad_norm": 0.35319179407248313,
+      "learning_rate": 0.0009967520590203306,
+      "loss": 3.8658,
+      "step": 2100
+    },
+    {
+      "epoch": 0.6649293216457396,
+      "grad_norm": 0.22517639994589708,
+      "learning_rate": 0.0009966890021541647,
+      "loss": 3.9244,
+      "step": 2105
+    },
+    {
+      "epoch": 0.6665087262102187,
+      "grad_norm": 0.24835059176588564,
+      "learning_rate": 0.0009966253410858144,
+      "loss": 3.9248,
+      "step": 2110
+    },
+    {
+      "epoch": 0.668088130774698,
+      "grad_norm": 0.22461996251898433,
+      "learning_rate": 0.000996561075892721,
+      "loss": 3.8047,
+      "step": 2115
+    },
+    {
+      "epoch": 0.6696675353391771,
+      "grad_norm": 0.3486531372669808,
+      "learning_rate": 0.0009964962066530605,
+      "loss": 4.0675,
+      "step": 2120
+    },
+    {
+      "epoch": 0.6712469399036564,
+      "grad_norm": 0.2901746826514593,
+      "learning_rate": 0.0009964307334457436,
+      "loss": 3.9818,
+      "step": 2125
+    },
+    {
+      "epoch": 0.6728263444681355,
+      "grad_norm": 0.2507760755352352,
+      "learning_rate": 0.0009963646563504159,
+      "loss": 3.9974,
+      "step": 2130
+    },
+    {
+      "epoch": 0.6744057490326147,
+      "grad_norm": 0.21192458654999655,
+      "learning_rate": 0.0009962979754474576,
+      "loss": 3.8887,
+      "step": 2135
+    },
+    {
+      "epoch": 0.6759851535970939,
+      "grad_norm": 0.19157148879425212,
+      "learning_rate": 0.0009962306908179832,
+      "loss": 3.8514,
+      "step": 2140
+    },
+    {
+      "epoch": 0.6775645581615731,
+      "grad_norm": 0.19679272132932518,
+      "learning_rate": 0.0009961628025438418,
+      "loss": 3.7598,
+      "step": 2145
+    },
+    {
+      "epoch": 0.6791439627260523,
+      "grad_norm": 0.26637100711605305,
+      "learning_rate": 0.0009960943107076169,
+      "loss": 3.9017,
+      "step": 2150
+    },
+    {
+      "epoch": 0.6807233672905315,
+      "grad_norm": 0.2847253716037483,
+      "learning_rate": 0.0009960252153926258,
+      "loss": 3.7984,
+      "step": 2155
+    },
+    {
+      "epoch": 0.6823027718550106,
+      "grad_norm": 0.22262377486617854,
+      "learning_rate": 0.0009959555166829204,
+      "loss": 3.8743,
+      "step": 2160
+    },
+    {
+      "epoch": 0.6838821764194899,
+      "grad_norm": 0.18462576402162742,
+      "learning_rate": 0.0009958852146632862,
+      "loss": 3.8773,
+      "step": 2165
+    },
+    {
+      "epoch": 0.685461580983969,
+      "grad_norm": 0.2823849265980843,
+      "learning_rate": 0.0009958143094192429,
+      "loss": 3.8371,
+      "step": 2170
+    },
+    {
+      "epoch": 0.6870409855484483,
+      "grad_norm": 0.19356104342333613,
+      "learning_rate": 0.000995742801037044,
+      "loss": 3.8393,
+      "step": 2175
+    },
+    {
+      "epoch": 0.6886203901129274,
+      "grad_norm": 0.24691680125149096,
+      "learning_rate": 0.000995670689603676,
+      "loss": 3.7618,
+      "step": 2180
+    },
+    {
+      "epoch": 0.6901997946774067,
+      "grad_norm": 0.22870840714473814,
+      "learning_rate": 0.0009955979752068603,
+      "loss": 3.8878,
+      "step": 2185
+    },
+    {
+      "epoch": 0.6917791992418858,
+      "grad_norm": 0.25006435731509863,
+      "learning_rate": 0.0009955246579350506,
+      "loss": 3.8762,
+      "step": 2190
+    },
+    {
+      "epoch": 0.693358603806365,
+      "grad_norm": 0.19895321523054352,
+      "learning_rate": 0.0009954507378774344,
+      "loss": 3.8146,
+      "step": 2195
+    },
+    {
+      "epoch": 0.6949380083708442,
+      "grad_norm": 0.26802335283542106,
+      "learning_rate": 0.0009953762151239326,
+      "loss": 3.8611,
+      "step": 2200
+    },
+    {
+      "epoch": 0.6965174129353234,
+      "grad_norm": 0.2875206879898967,
+      "learning_rate": 0.0009953010897651993,
+      "loss": 3.8369,
+      "step": 2205
+    },
+    {
+      "epoch": 0.6980968174998026,
+      "grad_norm": 0.2856661439521279,
+      "learning_rate": 0.000995225361892621,
+      "loss": 4.0,
+      "step": 2210
+    },
+    {
+      "epoch": 0.6996762220642818,
+      "grad_norm": 0.30979961397328754,
+      "learning_rate": 0.000995149031598318,
+      "loss": 3.84,
+      "step": 2215
+    },
+    {
+      "epoch": 0.701255626628761,
+      "grad_norm": 0.30130512379471247,
+      "learning_rate": 0.000995072098975143,
+      "loss": 3.8299,
+      "step": 2220
+    },
+    {
+      "epoch": 0.7028350311932402,
+      "grad_norm": 0.2221796910671295,
+      "learning_rate": 0.0009949945641166812,
+      "loss": 3.8664,
+      "step": 2225
+    },
+    {
+      "epoch": 0.7044144357577193,
+      "grad_norm": 0.23762852053695657,
+      "learning_rate": 0.000994916427117251,
+      "loss": 3.7376,
+      "step": 2230
+    },
+    {
+      "epoch": 0.7059938403221986,
+      "grad_norm": 0.18218641939581295,
+      "learning_rate": 0.0009948376880719028,
+      "loss": 3.7075,
+      "step": 2235
+    },
+    {
+      "epoch": 0.7075732448866777,
+      "grad_norm": 0.1870445334301083,
+      "learning_rate": 0.0009947583470764191,
+      "loss": 3.7572,
+      "step": 2240
+    },
+    {
+      "epoch": 0.709152649451157,
+      "grad_norm": 0.1834689185397494,
+      "learning_rate": 0.0009946784042273154,
+      "loss": 3.7828,
+      "step": 2245
+    },
+    {
+      "epoch": 0.7107320540156361,
+      "grad_norm": 0.185023861033481,
+      "learning_rate": 0.000994597859621839,
+      "loss": 3.8348,
+      "step": 2250
+    },
+    {
+      "epoch": 0.7123114585801152,
+      "grad_norm": 0.19030090837858005,
+      "learning_rate": 0.000994516713357969,
+      "loss": 3.7744,
+      "step": 2255
+    },
+    {
+      "epoch": 0.7138908631445945,
+      "grad_norm": 0.21590872376404172,
+      "learning_rate": 0.0009944349655344167,
+      "loss": 3.8309,
+      "step": 2260
+    },
+    {
+      "epoch": 0.7154702677090736,
+      "grad_norm": 0.28478239021445106,
+      "learning_rate": 0.000994352616250625,
+      "loss": 3.8515,
+      "step": 2265
+    },
+    {
+      "epoch": 0.7170496722735529,
+      "grad_norm": 0.21089138394462176,
+      "learning_rate": 0.0009942696656067682,
+      "loss": 3.8179,
+      "step": 2270
+    },
+    {
+      "epoch": 0.718629076838032,
+      "grad_norm": 0.31075061877113347,
+      "learning_rate": 0.000994186113703753,
+      "loss": 3.8159,
+      "step": 2275
+    },
+    {
+      "epoch": 0.7202084814025113,
+      "grad_norm": 0.28659229497528427,
+      "learning_rate": 0.0009941019606432163,
+      "loss": 3.8684,
+      "step": 2280
+    },
+    {
+      "epoch": 0.7217878859669904,
+      "grad_norm": 0.2412094613864912,
+      "learning_rate": 0.0009940172065275273,
+      "loss": 3.8988,
+      "step": 2285
+    },
+    {
+      "epoch": 0.7233672905314696,
+      "grad_norm": 0.21405775059275908,
+      "learning_rate": 0.000993931851459786,
+      "loss": 3.7733,
+      "step": 2290
+    },
+    {
+      "epoch": 0.7249466950959488,
+      "grad_norm": 0.12139107360781308,
+      "learning_rate": 0.000993845895543823,
+      "loss": 3.648,
+      "step": 2295
+    },
+    {
+      "epoch": 0.726526099660428,
+      "grad_norm": 0.23868396704616063,
+      "learning_rate": 0.0009937593388842007,
+      "loss": 3.7886,
+      "step": 2300
+    },
+    {
+      "epoch": 0.7281055042249072,
+      "grad_norm": 0.19195542843046692,
+      "learning_rate": 0.0009936721815862117,
+      "loss": 3.7889,
+      "step": 2305
+    },
+    {
+      "epoch": 0.7296849087893864,
+      "grad_norm": 0.2257860153230892,
+      "learning_rate": 0.0009935844237558792,
+      "loss": 3.7564,
+      "step": 2310
+    },
+    {
+      "epoch": 0.7312643133538655,
+      "grad_norm": 0.17831367547050161,
+      "learning_rate": 0.000993496065499957,
+      "loss": 3.8003,
+      "step": 2315
+    },
+    {
+      "epoch": 0.7328437179183448,
+      "grad_norm": 0.2590453195673119,
+      "learning_rate": 0.0009934071069259295,
+      "loss": 3.8104,
+      "step": 2320
+    },
+    {
+      "epoch": 0.7344231224828239,
+      "grad_norm": 0.2169619591401621,
+      "learning_rate": 0.0009933175481420112,
+      "loss": 3.8378,
+      "step": 2325
+    },
+    {
+      "epoch": 0.7360025270473032,
+      "grad_norm": 0.27623130925256867,
+      "learning_rate": 0.0009932273892571467,
+      "loss": 3.7137,
+      "step": 2330
+    },
+    {
+      "epoch": 0.7375819316117823,
+      "grad_norm": 0.2294100407856401,
+      "learning_rate": 0.0009931366303810108,
+      "loss": 3.8197,
+      "step": 2335
+    },
+    {
+      "epoch": 0.7391613361762616,
+      "grad_norm": 0.23406768014525023,
+      "learning_rate": 0.0009930452716240077,
+      "loss": 3.8387,
+      "step": 2340
+    },
+    {
+      "epoch": 0.7407407407407407,
+      "grad_norm": 0.20537121017839016,
+      "learning_rate": 0.000992953313097272,
+      "loss": 3.8095,
+      "step": 2345
+    },
+    {
+      "epoch": 0.74232014530522,
+      "grad_norm": 0.22643879626344338,
+      "learning_rate": 0.0009928607549126677,
+      "loss": 3.7317,
+      "step": 2350
+    },
+    {
+      "epoch": 0.7438995498696991,
+      "grad_norm": 0.2257293267348555,
+      "learning_rate": 0.0009927675971827875,
+      "loss": 3.9312,
+      "step": 2355
+    },
+    {
+      "epoch": 0.7454789544341783,
+      "grad_norm": 0.23366862269634528,
+      "learning_rate": 0.0009926738400209546,
+      "loss": 3.76,
+      "step": 2360
+    },
+    {
+      "epoch": 0.7470583589986575,
+      "grad_norm": 0.20353001859992098,
+      "learning_rate": 0.0009925794835412205,
+      "loss": 3.6586,
+      "step": 2365
+    },
+    {
+      "epoch": 0.7486377635631367,
+      "grad_norm": 0.18424469126412626,
+      "learning_rate": 0.000992484527858366,
+      "loss": 3.9304,
+      "step": 2370
+    },
+    {
+      "epoch": 0.7502171681276159,
+      "grad_norm": 0.24559128183468742,
+      "learning_rate": 0.0009923889730879011,
+      "loss": 3.7245,
+      "step": 2375
+    },
+    {
+      "epoch": 0.7517965726920951,
+      "grad_norm": 0.3062722293190668,
+      "learning_rate": 0.0009922928193460644,
+      "loss": 3.8891,
+      "step": 2380
+    },
+    {
+      "epoch": 0.7533759772565742,
+      "grad_norm": 0.27045048766747526,
+      "learning_rate": 0.0009921960667498226,
+      "loss": 3.8022,
+      "step": 2385
+    },
+    {
+      "epoch": 0.7549553818210535,
+      "grad_norm": 0.2562345403894404,
+      "learning_rate": 0.0009920987154168719,
+      "loss": 3.7404,
+      "step": 2390
+    },
+    {
+      "epoch": 0.7565347863855326,
+      "grad_norm": 0.21991589014582286,
+      "learning_rate": 0.0009920007654656358,
+      "loss": 3.755,
+      "step": 2395
+    },
+    {
+      "epoch": 0.7581141909500119,
+      "grad_norm": 0.2304615828900788,
+      "learning_rate": 0.0009919022170152667,
+      "loss": 3.8066,
+      "step": 2400
+    },
+    {
+      "epoch": 0.759693595514491,
+      "grad_norm": 0.19738272793982797,
+      "learning_rate": 0.000991803070185645,
+      "loss": 3.6949,
+      "step": 2405
+    },
+    {
+      "epoch": 0.7612730000789703,
+      "grad_norm": 0.2030280499023611,
+      "learning_rate": 0.0009917033250973785,
+      "loss": 3.7832,
+      "step": 2410
+    },
+    {
+      "epoch": 0.7628524046434494,
+      "grad_norm": 0.22494496151508525,
+      "learning_rate": 0.0009916029818718033,
+      "loss": 3.7324,
+      "step": 2415
+    },
+    {
+      "epoch": 0.7644318092079286,
+      "grad_norm": 0.22932816027271852,
+      "learning_rate": 0.0009915020406309827,
+      "loss": 3.7687,
+      "step": 2420
+    },
+    {
+      "epoch": 0.7660112137724078,
+      "grad_norm": 0.27604691578739265,
+      "learning_rate": 0.000991400501497708,
+      "loss": 3.7379,
+      "step": 2425
+    },
+    {
+      "epoch": 0.767590618336887,
+      "grad_norm": 0.24657062639557809,
+      "learning_rate": 0.000991298364595497,
+      "loss": 3.6804,
+      "step": 2430
+    },
+    {
+      "epoch": 0.7691700229013662,
+      "grad_norm": 0.17412852873504417,
+      "learning_rate": 0.0009911956300485956,
+      "loss": 3.7576,
+      "step": 2435
+    },
+    {
+      "epoch": 0.7707494274658454,
+      "grad_norm": 0.19379318906955947,
+      "learning_rate": 0.0009910922979819762,
+      "loss": 3.8152,
+      "step": 2440
+    },
+    {
+      "epoch": 0.7723288320303245,
+      "grad_norm": 0.21644414786245522,
+      "learning_rate": 0.0009909883685213375,
+      "loss": 3.7226,
+      "step": 2445
+    },
+    {
+      "epoch": 0.7739082365948038,
+      "grad_norm": 0.2787930687738909,
+      "learning_rate": 0.000990883841793106,
+      "loss": 3.7699,
+      "step": 2450
+    },
+    {
+      "epoch": 0.7754876411592829,
+      "grad_norm": 0.24870103981372055,
+      "learning_rate": 0.0009907787179244344,
+      "loss": 3.749,
+      "step": 2455
+    },
+    {
+      "epoch": 0.7770670457237622,
+      "grad_norm": 0.19321694165856332,
+      "learning_rate": 0.0009906729970432014,
+      "loss": 3.7828,
+      "step": 2460
+    },
+    {
+      "epoch": 0.7786464502882413,
+      "grad_norm": 0.18949642569362274,
+      "learning_rate": 0.0009905666792780121,
+      "loss": 3.7458,
+      "step": 2465
+    },
+    {
+      "epoch": 0.7802258548527206,
+      "grad_norm": 0.2668235859016389,
+      "learning_rate": 0.0009904597647581981,
+      "loss": 3.6928,
+      "step": 2470
+    },
+    {
+      "epoch": 0.7818052594171997,
+      "grad_norm": 0.26892923882002495,
+      "learning_rate": 0.0009903522536138165,
+      "loss": 3.7568,
+      "step": 2475
+    },
+    {
+      "epoch": 0.7833846639816789,
+      "grad_norm": 0.23583860928869732,
+      "learning_rate": 0.00099024414597565,
+      "loss": 3.7143,
+      "step": 2480
+    },
+    {
+      "epoch": 0.7849640685461581,
+      "grad_norm": 0.24434878517883898,
+      "learning_rate": 0.0009901354419752076,
+      "loss": 3.7339,
+      "step": 2485
+    },
+    {
+      "epoch": 0.7865434731106373,
+      "grad_norm": 0.27267916282442095,
+      "learning_rate": 0.000990026141744723,
+      "loss": 3.8102,
+      "step": 2490
+    },
+    {
+      "epoch": 0.7881228776751165,
+      "grad_norm": 0.30307916098945237,
+      "learning_rate": 0.0009899162454171558,
+      "loss": 3.736,
+      "step": 2495
+    },
+    {
+      "epoch": 0.7897022822395957,
+      "grad_norm": 0.22514273495342624,
+      "learning_rate": 0.0009898057531261902,
+      "loss": 3.74,
+      "step": 2500
+    },
+    {
+      "epoch": 0.7912816868040748,
+      "grad_norm": 0.21910656895816347,
+      "learning_rate": 0.000989694665006236,
+      "loss": 3.7559,
+      "step": 2505
+    },
+    {
+      "epoch": 0.7928610913685541,
+      "grad_norm": 0.21413778361238997,
+      "learning_rate": 0.0009895829811924271,
+      "loss": 3.7509,
+      "step": 2510
+    },
+    {
+      "epoch": 0.7944404959330332,
+      "grad_norm": 0.2100958485249077,
+      "learning_rate": 0.0009894707018206223,
+      "loss": 3.6985,
+      "step": 2515
+    },
+    {
+      "epoch": 0.7960199004975125,
+      "grad_norm": 0.22357525768208342,
+      "learning_rate": 0.0009893578270274053,
+      "loss": 3.7324,
+      "step": 2520
+    },
+    {
+      "epoch": 0.7975993050619916,
+      "grad_norm": 0.1964921217332807,
+      "learning_rate": 0.0009892443569500834,
+      "loss": 3.7291,
+      "step": 2525
+    },
+    {
+      "epoch": 0.7991787096264709,
+      "grad_norm": 0.20271019779435234,
+      "learning_rate": 0.0009891302917266888,
+      "loss": 3.8733,
+      "step": 2530
+    },
+    {
+      "epoch": 0.80075811419095,
+      "grad_norm": 0.23478169827263035,
+      "learning_rate": 0.0009890156314959768,
+      "loss": 3.8878,
+      "step": 2535
+    },
+    {
+      "epoch": 0.8023375187554292,
+      "grad_norm": 0.2313386457109853,
+      "learning_rate": 0.0009889003763974271,
+      "loss": 3.6695,
+      "step": 2540
+    },
+    {
+      "epoch": 0.8039169233199084,
+      "grad_norm": 0.2618485113502287,
+      "learning_rate": 0.000988784526571243,
+      "loss": 3.7112,
+      "step": 2545
+    },
+    {
+      "epoch": 0.8054963278843876,
+      "grad_norm": 0.2634351727376466,
+      "learning_rate": 0.0009886680821583511,
+      "loss": 3.7644,
+      "step": 2550
+    },
+    {
+      "epoch": 0.8070757324488668,
+      "grad_norm": 0.2595276380762902,
+      "learning_rate": 0.0009885510433004013,
+      "loss": 3.7772,
+      "step": 2555
+    },
+    {
+      "epoch": 0.808655137013346,
+      "grad_norm": 0.2587043663729113,
+      "learning_rate": 0.0009884334101397667,
+      "loss": 3.7313,
+      "step": 2560
+    },
+    {
+      "epoch": 0.8102345415778252,
+      "grad_norm": 0.22133887180757533,
+      "learning_rate": 0.0009883151828195432,
+      "loss": 3.6144,
+      "step": 2565
+    },
+    {
+      "epoch": 0.8118139461423044,
+      "grad_norm": 0.1816423792220125,
+      "learning_rate": 0.00098819636148355,
+      "loss": 3.7003,
+      "step": 2570
+    },
+    {
+      "epoch": 0.8133933507067835,
+      "grad_norm": 0.20401144650693834,
+      "learning_rate": 0.0009880769462763278,
+      "loss": 3.7263,
+      "step": 2575
+    },
+    {
+      "epoch": 0.8149727552712628,
+      "grad_norm": 0.16695314570111003,
+      "learning_rate": 0.0009879569373431407,
+      "loss": 3.759,
+      "step": 2580
+    },
+    {
+      "epoch": 0.8165521598357419,
+      "grad_norm": 0.20536838024238005,
+      "learning_rate": 0.000987836334829975,
+      "loss": 3.6269,
+      "step": 2585
+    },
+    {
+      "epoch": 0.8181315644002212,
+      "grad_norm": 0.1991843009792855,
+      "learning_rate": 0.0009877151388835385,
+      "loss": 3.9088,
+      "step": 2590
+    },
+    {
+      "epoch": 0.8197109689647003,
+      "grad_norm": 0.20686166646618734,
+      "learning_rate": 0.0009875933496512612,
+      "loss": 3.7823,
+      "step": 2595
+    },
+    {
+      "epoch": 0.8212903735291796,
+      "grad_norm": 0.2114497635258743,
+      "learning_rate": 0.0009874709672812948,
+      "loss": 3.734,
+      "step": 2600
+    },
+    {
+      "epoch": 0.8228697780936587,
+      "grad_norm": 0.21392111329380673,
+      "learning_rate": 0.0009873479919225128,
+      "loss": 3.6851,
+      "step": 2605
+    },
+    {
+      "epoch": 0.8244491826581379,
+      "grad_norm": 0.16240566726163633,
+      "learning_rate": 0.0009872244237245096,
+      "loss": 3.7172,
+      "step": 2610
+    },
+    {
+      "epoch": 0.8260285872226171,
+      "grad_norm": 0.16766924838905117,
+      "learning_rate": 0.000987100262837601,
+      "loss": 3.6609,
+      "step": 2615
+    },
+    {
+      "epoch": 0.8276079917870963,
+      "grad_norm": 0.25885756039168034,
+      "learning_rate": 0.0009869755094128233,
+      "loss": 3.6892,
+      "step": 2620
+    },
+    {
+      "epoch": 0.8291873963515755,
+      "grad_norm": 0.20299536884768418,
+      "learning_rate": 0.0009868501636019346,
+      "loss": 3.7749,
+      "step": 2625
+    },
+    {
+      "epoch": 0.8307668009160546,
+      "grad_norm": 0.21278111733231522,
+      "learning_rate": 0.0009867242255574126,
+      "loss": 3.6753,
+      "step": 2630
+    },
+    {
+      "epoch": 0.8323462054805338,
+      "grad_norm": 0.23192355698362774,
+      "learning_rate": 0.0009865976954324563,
+      "loss": 3.689,
+      "step": 2635
+    },
+    {
+      "epoch": 0.833925610045013,
+      "grad_norm": 0.26848443384418835,
+      "learning_rate": 0.0009864705733809843,
+      "loss": 3.6186,
+      "step": 2640
+    },
+    {
+      "epoch": 0.8355050146094922,
+      "grad_norm": 0.1906325871851695,
+      "learning_rate": 0.0009863428595576352,
+      "loss": 3.6238,
+      "step": 2645
+    },
+    {
+      "epoch": 0.8370844191739714,
+      "grad_norm": 0.1641940539362934,
+      "learning_rate": 0.000986214554117768,
+      "loss": 3.7163,
+      "step": 2650
+    },
+    {
+      "epoch": 0.8386638237384506,
+      "grad_norm": 0.18932792499119566,
+      "learning_rate": 0.000986085657217461,
+      "loss": 3.75,
+      "step": 2655
+    },
+    {
+      "epoch": 0.8402432283029297,
+      "grad_norm": 0.19028118211453493,
+      "learning_rate": 0.0009859561690135125,
+      "loss": 3.5875,
+      "step": 2660
+    },
+    {
+      "epoch": 0.841822632867409,
+      "grad_norm": 0.20793049944875938,
+      "learning_rate": 0.000985826089663439,
+      "loss": 3.6361,
+      "step": 2665
+    },
+    {
+      "epoch": 0.8434020374318881,
+      "grad_norm": 0.24167945997106643,
+      "learning_rate": 0.0009856954193254773,
+      "loss": 3.7011,
+      "step": 2670
+    },
+    {
+      "epoch": 0.8449814419963674,
+      "grad_norm": 0.19134822410708058,
+      "learning_rate": 0.0009855641581585823,
+      "loss": 3.6072,
+      "step": 2675
+    },
+    {
+      "epoch": 0.8465608465608465,
+      "grad_norm": 0.30594212306849533,
+      "learning_rate": 0.0009854323063224282,
+      "loss": 3.7363,
+      "step": 2680
+    },
+    {
+      "epoch": 0.8481402511253258,
+      "grad_norm": 0.3554004161494336,
+      "learning_rate": 0.0009852998639774072,
+      "loss": 3.7074,
+      "step": 2685
+    },
+    {
+      "epoch": 0.8497196556898049,
+      "grad_norm": 0.2158252194984231,
+      "learning_rate": 0.0009851668312846303,
+      "loss": 3.719,
+      "step": 2690
+    },
+    {
+      "epoch": 0.8512990602542841,
+      "grad_norm": 0.2094551560100172,
+      "learning_rate": 0.0009850332084059262,
+      "loss": 3.6551,
+      "step": 2695
+    },
+    {
+      "epoch": 0.8528784648187633,
+      "grad_norm": 0.18976941526589675,
+      "learning_rate": 0.000984898995503842,
+      "loss": 3.659,
+      "step": 2700
+    },
+    {
+      "epoch": 0.8544578693832425,
+      "grad_norm": 0.19847958962221077,
+      "learning_rate": 0.0009847641927416423,
+      "loss": 3.6902,
+      "step": 2705
+    },
+    {
+      "epoch": 0.8560372739477217,
+      "grad_norm": 0.2354635589066857,
+      "learning_rate": 0.0009846288002833088,
+      "loss": 3.6636,
+      "step": 2710
+    },
+    {
+      "epoch": 0.8576166785122009,
+      "grad_norm": 0.20562961378170208,
+      "learning_rate": 0.0009844928182935414,
+      "loss": 3.7092,
+      "step": 2715
+    },
+    {
+      "epoch": 0.85919608307668,
+      "grad_norm": 0.24859698816861014,
+      "learning_rate": 0.0009843562469377567,
+      "loss": 3.713,
+      "step": 2720
+    },
+    {
+      "epoch": 0.8607754876411593,
+      "grad_norm": 0.2399986307557472,
+      "learning_rate": 0.000984219086382088,
+      "loss": 3.756,
+      "step": 2725
+    },
+    {
+      "epoch": 0.8623548922056384,
+      "grad_norm": 0.27013518984869855,
+      "learning_rate": 0.0009840813367933859,
+      "loss": 3.7906,
+      "step": 2730
+    },
+    {
+      "epoch": 0.8639342967701177,
+      "grad_norm": 0.23047495735441087,
+      "learning_rate": 0.000983942998339217,
+      "loss": 3.7627,
+      "step": 2735
+    },
+    {
+      "epoch": 0.8655137013345968,
+      "grad_norm": 0.20430948001029364,
+      "learning_rate": 0.0009838040711878646,
+      "loss": 3.603,
+      "step": 2740
+    },
+    {
+      "epoch": 0.8670931058990761,
+      "grad_norm": 0.18081743933907277,
+      "learning_rate": 0.0009836645555083281,
+      "loss": 3.5728,
+      "step": 2745
+    },
+    {
+      "epoch": 0.8686725104635552,
+      "grad_norm": 0.22174597927758466,
+      "learning_rate": 0.0009835244514703222,
+      "loss": 3.6436,
+      "step": 2750
+    },
+    {
+      "epoch": 0.8702519150280345,
+      "grad_norm": 0.1953983191119668,
+      "learning_rate": 0.0009833837592442786,
+      "loss": 3.6422,
+      "step": 2755
+    },
+    {
+      "epoch": 0.8718313195925136,
+      "grad_norm": 0.1829661116296364,
+      "learning_rate": 0.000983242479001343,
+      "loss": 3.6262,
+      "step": 2760
+    },
+    {
+      "epoch": 0.8734107241569928,
+      "grad_norm": 0.22235033352924125,
+      "learning_rate": 0.0009831006109133776,
+      "loss": 3.7124,
+      "step": 2765
+    },
+    {
+      "epoch": 0.874990128721472,
+      "grad_norm": 0.266676779913622,
+      "learning_rate": 0.000982958155152959,
+      "loss": 3.6524,
+      "step": 2770
+    },
+    {
+      "epoch": 0.8765695332859512,
+      "grad_norm": 0.20839375394643436,
+      "learning_rate": 0.000982815111893379,
+      "loss": 3.6547,
+      "step": 2775
+    },
+    {
+      "epoch": 0.8781489378504304,
+      "grad_norm": 0.21055619207323192,
+      "learning_rate": 0.0009826714813086438,
+      "loss": 3.7226,
+      "step": 2780
+    },
+    {
+      "epoch": 0.8797283424149096,
+      "grad_norm": 0.22345474054173048,
+      "learning_rate": 0.0009825272635734746,
+      "loss": 3.6297,
+      "step": 2785
+    },
+    {
+      "epoch": 0.8813077469793887,
+      "grad_norm": 0.20804544740257655,
+      "learning_rate": 0.0009823824588633058,
+      "loss": 3.648,
+      "step": 2790
+    },
+    {
+      "epoch": 0.882887151543868,
+      "grad_norm": 0.18295164550994766,
+      "learning_rate": 0.0009822370673542872,
+      "loss": 3.6555,
+      "step": 2795
+    },
+    {
+      "epoch": 0.8844665561083471,
+      "grad_norm": 0.19425225687908296,
+      "learning_rate": 0.0009820910892232816,
+      "loss": 3.634,
+      "step": 2800
+    },
+    {
+      "epoch": 0.8860459606728264,
+      "grad_norm": 0.19191672702600987,
+      "learning_rate": 0.0009819445246478653,
+      "loss": 3.6317,
+      "step": 2805
+    },
+    {
+      "epoch": 0.8876253652373055,
+      "grad_norm": 0.21038674810948635,
+      "learning_rate": 0.000981797373806328,
+      "loss": 3.6634,
+      "step": 2810
+    },
+    {
+      "epoch": 0.8892047698017848,
+      "grad_norm": 0.2197613989072648,
+      "learning_rate": 0.0009816496368776734,
+      "loss": 3.5895,
+      "step": 2815
+    },
+    {
+      "epoch": 0.8907841743662639,
+      "grad_norm": 0.21178720188607164,
+      "learning_rate": 0.000981501314041617,
+      "loss": 3.6104,
+      "step": 2820
+    },
+    {
+      "epoch": 0.8923635789307431,
+      "grad_norm": 0.17082253963144226,
+      "learning_rate": 0.0009813524054785878,
+      "loss": 3.7365,
+      "step": 2825
+    },
+    {
+      "epoch": 0.8939429834952223,
+      "grad_norm": 0.1725318619775281,
+      "learning_rate": 0.0009812029113697271,
+      "loss": 3.6157,
+      "step": 2830
+    },
+    {
+      "epoch": 0.8955223880597015,
+      "grad_norm": 0.2352061797282261,
+      "learning_rate": 0.0009810528318968882,
+      "loss": 3.6178,
+      "step": 2835
+    },
+    {
+      "epoch": 0.8971017926241807,
+      "grad_norm": 0.21716211764222246,
+      "learning_rate": 0.0009809021672426371,
+      "loss": 3.6164,
+      "step": 2840
+    },
+    {
+      "epoch": 0.8986811971886599,
+      "grad_norm": 0.1760557168883546,
+      "learning_rate": 0.000980750917590251,
+      "loss": 3.6029,
+      "step": 2845
+    },
+    {
+      "epoch": 0.900260601753139,
+      "grad_norm": 0.14698275622175808,
+      "learning_rate": 0.0009805990831237192,
+      "loss": 3.6432,
+      "step": 2850
+    },
+    {
+      "epoch": 0.9018400063176183,
+      "grad_norm": 0.16077926077590798,
+      "learning_rate": 0.0009804466640277421,
+      "loss": 3.5867,
+      "step": 2855
+    },
+    {
+      "epoch": 0.9034194108820974,
+      "grad_norm": 0.16707694137928564,
+      "learning_rate": 0.0009802936604877317,
+      "loss": 3.5945,
+      "step": 2860
+    },
+    {
+      "epoch": 0.9049988154465767,
+      "grad_norm": 0.21633216870374575,
+      "learning_rate": 0.0009801400726898101,
+      "loss": 3.6673,
+      "step": 2865
+    },
+    {
+      "epoch": 0.9065782200110558,
+      "grad_norm": 0.1662224581417447,
+      "learning_rate": 0.000979985900820811,
+      "loss": 3.5377,
+      "step": 2870
+    },
+    {
+      "epoch": 0.9081576245755351,
+      "grad_norm": 0.17310136270402474,
+      "learning_rate": 0.0009798311450682784,
+      "loss": 3.6498,
+      "step": 2875
+    },
+    {
+      "epoch": 0.9097370291400142,
+      "grad_norm": 0.22943147677681835,
+      "learning_rate": 0.0009796758056204661,
+      "loss": 3.5972,
+      "step": 2880
+    },
+    {
+      "epoch": 0.9113164337044934,
+      "grad_norm": 0.2643736873561172,
+      "learning_rate": 0.0009795198826663388,
+      "loss": 3.6368,
+      "step": 2885
+    },
+    {
+      "epoch": 0.9128958382689726,
+      "grad_norm": 0.25044361405077453,
+      "learning_rate": 0.00097936337639557,
+      "loss": 3.5429,
+      "step": 2890
+    },
+    {
+      "epoch": 0.9144752428334518,
+      "grad_norm": 0.23839389612516823,
+      "learning_rate": 0.0009792062869985435,
+      "loss": 3.5907,
+      "step": 2895
+    },
+    {
+      "epoch": 0.916054647397931,
+      "grad_norm": 0.2363741993442048,
+      "learning_rate": 0.000979048614666352,
+      "loss": 3.6616,
+      "step": 2900
+    },
+    {
+      "epoch": 0.9176340519624102,
+      "grad_norm": 0.2780080804073825,
+      "learning_rate": 0.000978890359590798,
+      "loss": 3.8078,
+      "step": 2905
+    },
+    {
+      "epoch": 0.9192134565268893,
+      "grad_norm": 0.15362194641567975,
+      "learning_rate": 0.000978731521964392,
+      "loss": 3.5856,
+      "step": 2910
+    },
+    {
+      "epoch": 0.9207928610913686,
+      "grad_norm": 0.3174075119221956,
+      "learning_rate": 0.0009785721019803539,
+      "loss": 3.6724,
+      "step": 2915
+    },
+    {
+      "epoch": 0.9223722656558477,
+      "grad_norm": 0.18801093915486303,
+      "learning_rate": 0.0009784120998326113,
+      "loss": 3.6253,
+      "step": 2920
+    },
+    {
+      "epoch": 0.923951670220327,
+      "grad_norm": 0.2572007567570295,
+      "learning_rate": 0.0009782515157158009,
+      "loss": 3.6182,
+      "step": 2925
+    },
+    {
+      "epoch": 0.9255310747848061,
+      "grad_norm": 0.2199486414718636,
+      "learning_rate": 0.0009780903498252664,
+      "loss": 3.5813,
+      "step": 2930
+    },
+    {
+      "epoch": 0.9271104793492854,
+      "grad_norm": 0.1805834748540574,
+      "learning_rate": 0.00097792860235706,
+      "loss": 3.5573,
+      "step": 2935
+    },
+    {
+      "epoch": 0.9286898839137645,
+      "grad_norm": 0.1807851236451506,
+      "learning_rate": 0.0009777662735079404,
+      "loss": 3.6422,
+      "step": 2940
+    },
+    {
+      "epoch": 0.9302692884782437,
+      "grad_norm": 0.19801039649431965,
+      "learning_rate": 0.0009776033634753746,
+      "loss": 3.6314,
+      "step": 2945
+    },
+    {
+      "epoch": 0.9318486930427229,
+      "grad_norm": 0.20233526907990002,
+      "learning_rate": 0.000977439872457536,
+      "loss": 3.6458,
+      "step": 2950
+    },
+    {
+      "epoch": 0.9334280976072021,
+      "grad_norm": 0.1861241478177403,
+      "learning_rate": 0.0009772758006533046,
+      "loss": 3.5384,
+      "step": 2955
+    },
+    {
+      "epoch": 0.9350075021716813,
+      "grad_norm": 0.1534334003544395,
+      "learning_rate": 0.0009771111482622676,
+      "loss": 3.61,
+      "step": 2960
+    },
+    {
+      "epoch": 0.9365869067361605,
+      "grad_norm": 0.19550599832198995,
+      "learning_rate": 0.0009769459154847177,
+      "loss": 3.6829,
+      "step": 2965
+    },
+    {
+      "epoch": 0.9381663113006397,
+      "grad_norm": 0.253267013653418,
+      "learning_rate": 0.000976780102521654,
+      "loss": 3.6187,
+      "step": 2970
+    },
+    {
+      "epoch": 0.9397457158651189,
+      "grad_norm": 0.1919114006187933,
+      "learning_rate": 0.0009766137095747812,
+      "loss": 3.5702,
+      "step": 2975
+    },
+    {
+      "epoch": 0.941325120429598,
+      "grad_norm": 0.17264392678432056,
+      "learning_rate": 0.0009764467368465098,
+      "loss": 3.7729,
+      "step": 2980
+    },
+    {
+      "epoch": 0.9429045249940773,
+      "grad_norm": 0.16643373381157617,
+      "learning_rate": 0.0009762791845399552,
+      "loss": 3.5039,
+      "step": 2985
+    },
+    {
+      "epoch": 0.9444839295585564,
+      "grad_norm": 0.18808067976092,
+      "learning_rate": 0.0009761110528589381,
+      "loss": 3.6013,
+      "step": 2990
+    },
+    {
+      "epoch": 0.9460633341230357,
+      "grad_norm": 0.25366070833950194,
+      "learning_rate": 0.000975942342007984,
+      "loss": 3.6747,
+      "step": 2995
+    },
+    {
+      "epoch": 0.9476427386875148,
+      "grad_norm": 0.23586423093199588,
+      "learning_rate": 0.000975773052192323,
+      "loss": 3.5451,
+      "step": 3000
+    },
+    {
+      "epoch": 0.9492221432519939,
+      "grad_norm": 0.25746326552700427,
+      "learning_rate": 0.0009756031836178891,
+      "loss": 3.7641,
+      "step": 3005
+    },
+    {
+      "epoch": 0.9508015478164732,
+      "grad_norm": 0.2565384006011303,
+      "learning_rate": 0.0009754327364913207,
+      "loss": 3.6605,
+      "step": 3010
+    },
+    {
+      "epoch": 0.9523809523809523,
+      "grad_norm": 0.20372628096265127,
+      "learning_rate": 0.0009752617110199598,
+      "loss": 3.6505,
+      "step": 3015
+    },
+    {
+      "epoch": 0.9539603569454316,
+      "grad_norm": 0.1948134588293486,
+      "learning_rate": 0.0009750901074118518,
+      "loss": 3.6699,
+      "step": 3020
+    },
+    {
+      "epoch": 0.9555397615099107,
+      "grad_norm": 0.180120677297602,
+      "learning_rate": 0.0009749179258757462,
+      "loss": 3.6799,
+      "step": 3025
+    },
+    {
+      "epoch": 0.95711916607439,
+      "grad_norm": 0.17287631701404943,
+      "learning_rate": 0.0009747451666210945,
+      "loss": 3.6146,
+      "step": 3030
+    },
+    {
+      "epoch": 0.9586985706388691,
+      "grad_norm": 0.19503517477380783,
+      "learning_rate": 0.0009745718298580512,
+      "loss": 3.6317,
+      "step": 3035
+    },
+    {
+      "epoch": 0.9602779752033483,
+      "grad_norm": 0.22902928630954272,
+      "learning_rate": 0.0009743979157974739,
+      "loss": 3.6152,
+      "step": 3040
+    },
+    {
+      "epoch": 0.9618573797678275,
+      "grad_norm": 0.17970255406094174,
+      "learning_rate": 0.0009742234246509217,
+      "loss": 3.6966,
+      "step": 3045
+    },
+    {
+      "epoch": 0.9634367843323067,
+      "grad_norm": 0.18801251804658714,
+      "learning_rate": 0.0009740483566306564,
+      "loss": 3.6747,
+      "step": 3050
+    },
+    {
+      "epoch": 0.9650161888967859,
+      "grad_norm": 0.16843563110513118,
+      "learning_rate": 0.0009738727119496409,
+      "loss": 3.5028,
+      "step": 3055
+    },
+    {
+      "epoch": 0.9665955934612651,
+      "grad_norm": 0.19447085681688397,
+      "learning_rate": 0.0009736964908215401,
+      "loss": 3.6577,
+      "step": 3060
+    },
+    {
+      "epoch": 0.9681749980257442,
+      "grad_norm": 0.22009121363697623,
+      "learning_rate": 0.0009735196934607198,
+      "loss": 3.5345,
+      "step": 3065
+    },
+    {
+      "epoch": 0.9697544025902235,
+      "grad_norm": 0.22448750337401296,
+      "learning_rate": 0.0009733423200822469,
+      "loss": 3.5515,
+      "step": 3070
+    },
+    {
+      "epoch": 0.9713338071547026,
+      "grad_norm": 0.20771129618077283,
+      "learning_rate": 0.0009731643709018891,
+      "loss": 3.5718,
+      "step": 3075
+    },
+    {
+      "epoch": 0.9729132117191819,
+      "grad_norm": 0.20373000120459916,
+      "learning_rate": 0.0009729858461361141,
+      "loss": 3.6764,
+      "step": 3080
+    },
+    {
+      "epoch": 0.974492616283661,
+      "grad_norm": 0.1797561361208036,
+      "learning_rate": 0.0009728067460020904,
+      "loss": 3.6827,
+      "step": 3085
+    },
+    {
+      "epoch": 0.9760720208481403,
+      "grad_norm": 0.2147655801584539,
+      "learning_rate": 0.0009726270707176858,
+      "loss": 3.5844,
+      "step": 3090
+    },
+    {
+      "epoch": 0.9776514254126194,
+      "grad_norm": 0.1960578897451964,
+      "learning_rate": 0.0009724468205014685,
+      "loss": 3.7833,
+      "step": 3095
+    },
+    {
+      "epoch": 0.9792308299770986,
+      "grad_norm": 0.15597674681769813,
+      "learning_rate": 0.0009722659955727054,
+      "loss": 3.656,
+      "step": 3100
+    },
+    {
+      "epoch": 0.9808102345415778,
+      "grad_norm": 0.21261634725346287,
+      "learning_rate": 0.0009720845961513627,
+      "loss": 3.55,
+      "step": 3105
+    },
+    {
+      "epoch": 0.982389639106057,
+      "grad_norm": 0.2435056831114221,
+      "learning_rate": 0.0009719026224581053,
+      "loss": 3.6242,
+      "step": 3110
+    },
+    {
+      "epoch": 0.9839690436705362,
+      "grad_norm": 0.16406935937973846,
+      "learning_rate": 0.0009717200747142974,
+      "loss": 3.5754,
+      "step": 3115
+    },
+    {
+      "epoch": 0.9855484482350154,
+      "grad_norm": 0.18968492314951244,
+      "learning_rate": 0.0009715369531420006,
+      "loss": 3.6399,
+      "step": 3120
+    },
+    {
+      "epoch": 0.9871278527994946,
+      "grad_norm": 0.16702625054301215,
+      "learning_rate": 0.000971353257963975,
+      "loss": 3.5255,
+      "step": 3125
+    },
+    {
+      "epoch": 0.9887072573639738,
+      "grad_norm": 0.2301624768016454,
+      "learning_rate": 0.0009711689894036784,
+      "loss": 3.5684,
+      "step": 3130
+    },
+    {
+      "epoch": 0.9902866619284529,
+      "grad_norm": 0.19561138426372737,
+      "learning_rate": 0.0009709841476852661,
+      "loss": 3.5207,
+      "step": 3135
+    },
+    {
+      "epoch": 0.9918660664929322,
+      "grad_norm": 0.2027504555908024,
+      "learning_rate": 0.0009707987330335906,
+      "loss": 3.5694,
+      "step": 3140
+    },
+    {
+      "epoch": 0.9934454710574113,
+      "grad_norm": 0.16179807583256492,
+      "learning_rate": 0.0009706127456742014,
+      "loss": 3.4524,
+      "step": 3145
+    },
+    {
+      "epoch": 0.9950248756218906,
+      "grad_norm": 0.18013411208473787,
+      "learning_rate": 0.0009704261858333445,
+      "loss": 3.5072,
+      "step": 3150
+    },
+    {
+      "epoch": 0.9966042801863697,
+      "grad_norm": 0.17933880588432946,
+      "learning_rate": 0.0009702390537379627,
+      "loss": 3.7013,
+      "step": 3155
+    },
+    {
+      "epoch": 0.998183684750849,
+      "grad_norm": 0.2131080756575249,
+      "learning_rate": 0.0009700513496156945,
+      "loss": 3.6046,
+      "step": 3160
+    },
+    {
+      "epoch": 0.9997630893153281,
+      "grad_norm": 0.29187684680182674,
+      "learning_rate": 0.0009698630736948744,
+      "loss": 3.6084,
+      "step": 3165
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 3.5630176067352295,
+      "eval_runtime": 118.6591,
+      "eval_samples_per_second": 22.324,
+      "eval_steps_per_second": 5.587,
+      "step": 3166
+    },
+    {
+      "epoch": 1.0012635236515834,
+      "grad_norm": 0.19522578739733992,
+      "learning_rate": 0.0009696742262045323,
+      "loss": 3.5774,
+      "step": 3170
+    },
+    {
+      "epoch": 1.0028429282160625,
+      "grad_norm": 0.19544019298572032,
+      "learning_rate": 0.0009694848073743941,
+      "loss": 3.5418,
+      "step": 3175
+    },
+    {
+      "epoch": 1.0044223327805417,
+      "grad_norm": 0.18717051155662734,
+      "learning_rate": 0.0009692948174348797,
+      "loss": 3.4698,
+      "step": 3180
+    },
+    {
+      "epoch": 1.006001737345021,
+      "grad_norm": 0.26296600436862455,
+      "learning_rate": 0.0009691042566171044,
+      "loss": 3.4994,
+      "step": 3185
+    },
+    {
+      "epoch": 1.0075811419095002,
+      "grad_norm": 0.1674558011162135,
+      "learning_rate": 0.0009689131251528778,
+      "loss": 3.5352,
+      "step": 3190
+    },
+    {
+      "epoch": 1.0091605464739792,
+      "grad_norm": 0.19024232705973332,
+      "learning_rate": 0.0009687214232747035,
+      "loss": 3.4942,
+      "step": 3195
+    },
+    {
+      "epoch": 1.0107399510384585,
+      "grad_norm": 0.1847075755737659,
+      "learning_rate": 0.0009685291512157792,
+      "loss": 3.5486,
+      "step": 3200
+    },
+    {
+      "epoch": 1.0123193556029377,
+      "grad_norm": 0.1769802703112888,
+      "learning_rate": 0.0009683363092099961,
+      "loss": 3.5394,
+      "step": 3205
+    },
+    {
+      "epoch": 1.013898760167417,
+      "grad_norm": 0.1405972821619923,
+      "learning_rate": 0.0009681428974919388,
+      "loss": 3.5536,
+      "step": 3210
+    },
+    {
+      "epoch": 1.015478164731896,
+      "grad_norm": 0.18248409434912527,
+      "learning_rate": 0.0009679489162968849,
+      "loss": 3.641,
+      "step": 3215
+    },
+    {
+      "epoch": 1.0170575692963753,
+      "grad_norm": 0.3581630591943084,
+      "learning_rate": 0.0009677543658608046,
+      "loss": 3.595,
+      "step": 3220
+    },
+    {
+      "epoch": 1.0186369738608545,
+      "grad_norm": 0.29219008650027906,
+      "learning_rate": 0.0009675592464203609,
+      "loss": 3.5952,
+      "step": 3225
+    },
+    {
+      "epoch": 1.0202163784253337,
+      "grad_norm": 0.29655977440988046,
+      "learning_rate": 0.0009673635582129084,
+      "loss": 3.6436,
+      "step": 3230
+    },
+    {
+      "epoch": 1.0217957829898128,
+      "grad_norm": 0.1914125404242104,
+      "learning_rate": 0.0009671673014764942,
+      "loss": 3.5391,
+      "step": 3235
+    },
+    {
+      "epoch": 1.023375187554292,
+      "grad_norm": 0.21280629483900723,
+      "learning_rate": 0.0009669704764498564,
+      "loss": 3.6341,
+      "step": 3240
+    },
+    {
+      "epoch": 1.0249545921187713,
+      "grad_norm": 0.19389573313403266,
+      "learning_rate": 0.000966773083372425,
+      "loss": 3.5238,
+      "step": 3245
+    },
+    {
+      "epoch": 1.0265339966832505,
+      "grad_norm": 0.18443862860215415,
+      "learning_rate": 0.0009665751224843209,
+      "loss": 3.6114,
+      "step": 3250
+    },
+    {
+      "epoch": 1.0281134012477295,
+      "grad_norm": 0.24078883923384495,
+      "learning_rate": 0.0009663765940263554,
+      "loss": 3.5654,
+      "step": 3255
+    },
+    {
+      "epoch": 1.0296928058122088,
+      "grad_norm": 0.16865601842164352,
+      "learning_rate": 0.0009661774982400301,
+      "loss": 3.5134,
+      "step": 3260
+    },
+    {
+      "epoch": 1.031272210376688,
+      "grad_norm": 0.18953742112811325,
+      "learning_rate": 0.0009659778353675372,
+      "loss": 3.5744,
+      "step": 3265
+    },
+    {
+      "epoch": 1.0328516149411673,
+      "grad_norm": 0.19219203532463633,
+      "learning_rate": 0.0009657776056517589,
+      "loss": 3.6115,
+      "step": 3270
+    },
+    {
+      "epoch": 1.0344310195056463,
+      "grad_norm": 0.12784162972971713,
+      "learning_rate": 0.000965576809336266,
+      "loss": 3.4902,
+      "step": 3275
+    },
+    {
+      "epoch": 1.0360104240701256,
+      "grad_norm": 0.17808282679415993,
+      "learning_rate": 0.0009653754466653195,
+      "loss": 3.5061,
+      "step": 3280
+    },
+    {
+      "epoch": 1.0375898286346048,
+      "grad_norm": 0.12403067381712111,
+      "learning_rate": 0.000965173517883869,
+      "loss": 3.5288,
+      "step": 3285
+    },
+    {
+      "epoch": 1.039169233199084,
+      "grad_norm": 0.19963383337449706,
+      "learning_rate": 0.0009649710232375525,
+      "loss": 3.5505,
+      "step": 3290
+    },
+    {
+      "epoch": 1.040748637763563,
+      "grad_norm": 0.15232981796001893,
+      "learning_rate": 0.0009647679629726968,
+      "loss": 3.4659,
+      "step": 3295
+    },
+    {
+      "epoch": 1.0423280423280423,
+      "grad_norm": 0.1315179822589232,
+      "learning_rate": 0.0009645643373363164,
+      "loss": 3.5122,
+      "step": 3300
+    },
+    {
+      "epoch": 1.0439074468925216,
+      "grad_norm": 0.1644442465246829,
+      "learning_rate": 0.0009643601465761138,
+      "loss": 3.4649,
+      "step": 3305
+    },
+    {
+      "epoch": 1.0454868514570008,
+      "grad_norm": 0.18604550370191514,
+      "learning_rate": 0.0009641553909404788,
+      "loss": 3.5517,
+      "step": 3310
+    },
+    {
+      "epoch": 1.0470662560214798,
+      "grad_norm": 0.20518387776871452,
+      "learning_rate": 0.0009639500706784885,
+      "loss": 3.527,
+      "step": 3315
+    },
+    {
+      "epoch": 1.048645660585959,
+      "grad_norm": 0.1803425231097602,
+      "learning_rate": 0.0009637441860399066,
+      "loss": 3.5415,
+      "step": 3320
+    },
+    {
+      "epoch": 1.0502250651504383,
+      "grad_norm": 0.16054121711714328,
+      "learning_rate": 0.0009635377372751835,
+      "loss": 3.6723,
+      "step": 3325
+    },
+    {
+      "epoch": 1.0518044697149174,
+      "grad_norm": 0.21733786981905223,
+      "learning_rate": 0.0009633307246354558,
+      "loss": 3.5382,
+      "step": 3330
+    },
+    {
+      "epoch": 1.0533838742793966,
+      "grad_norm": 0.1613842724324584,
+      "learning_rate": 0.000963123148372546,
+      "loss": 3.5145,
+      "step": 3335
+    },
+    {
+      "epoch": 1.0549632788438759,
+      "grad_norm": 0.14661098326997973,
+      "learning_rate": 0.0009629150087389624,
+      "loss": 3.4844,
+      "step": 3340
+    },
+    {
+      "epoch": 1.056542683408355,
+      "grad_norm": 0.18355128495159284,
+      "learning_rate": 0.0009627063059878986,
+      "loss": 3.482,
+      "step": 3345
+    },
+    {
+      "epoch": 1.0581220879728341,
+      "grad_norm": 0.213067842797003,
+      "learning_rate": 0.0009624970403732327,
+      "loss": 3.618,
+      "step": 3350
+    },
+    {
+      "epoch": 1.0597014925373134,
+      "grad_norm": 0.19809498590991662,
+      "learning_rate": 0.0009622872121495283,
+      "loss": 3.5787,
+      "step": 3355
+    },
+    {
+      "epoch": 1.0612808971017926,
+      "grad_norm": 0.14272374854343148,
+      "learning_rate": 0.0009620768215720327,
+      "loss": 3.535,
+      "step": 3360
+    },
+    {
+      "epoch": 1.0628603016662719,
+      "grad_norm": 0.15106405340295168,
+      "learning_rate": 0.0009618658688966777,
+      "loss": 3.4411,
+      "step": 3365
+    },
+    {
+      "epoch": 1.064439706230751,
+      "grad_norm": 0.18741137454692364,
+      "learning_rate": 0.0009616543543800788,
+      "loss": 3.4986,
+      "step": 3370
+    },
+    {
+      "epoch": 1.0660191107952302,
+      "grad_norm": 0.1613689259330486,
+      "learning_rate": 0.0009614422782795348,
+      "loss": 3.5067,
+      "step": 3375
+    },
+    {
+      "epoch": 1.0675985153597094,
+      "grad_norm": 0.1257460506657428,
+      "learning_rate": 0.0009612296408530277,
+      "loss": 3.5044,
+      "step": 3380
+    },
+    {
+      "epoch": 1.0691779199241886,
+      "grad_norm": 0.1754167619697149,
+      "learning_rate": 0.0009610164423592227,
+      "loss": 3.5902,
+      "step": 3385
+    },
+    {
+      "epoch": 1.0707573244886677,
+      "grad_norm": 0.17571886278255155,
+      "learning_rate": 0.0009608026830574667,
+      "loss": 3.4963,
+      "step": 3390
+    },
+    {
+      "epoch": 1.072336729053147,
+      "grad_norm": 0.18849027726038495,
+      "learning_rate": 0.0009605883632077896,
+      "loss": 3.5786,
+      "step": 3395
+    },
+    {
+      "epoch": 1.0739161336176262,
+      "grad_norm": 0.21060076571187067,
+      "learning_rate": 0.0009603734830709028,
+      "loss": 3.5587,
+      "step": 3400
+    },
+    {
+      "epoch": 1.0754955381821054,
+      "grad_norm": 0.18052163861941398,
+      "learning_rate": 0.0009601580429081993,
+      "loss": 3.593,
+      "step": 3405
+    },
+    {
+      "epoch": 1.0770749427465844,
+      "grad_norm": 0.19603356848954506,
+      "learning_rate": 0.0009599420429817533,
+      "loss": 3.4352,
+      "step": 3410
+    },
+    {
+      "epoch": 1.0786543473110637,
+      "grad_norm": 0.1893666912236303,
+      "learning_rate": 0.0009597254835543204,
+      "loss": 3.5038,
+      "step": 3415
+    },
+    {
+      "epoch": 1.080233751875543,
+      "grad_norm": 0.19888125574740098,
+      "learning_rate": 0.0009595083648893362,
+      "loss": 3.5156,
+      "step": 3420
+    },
+    {
+      "epoch": 1.0818131564400222,
+      "grad_norm": 0.16899935054600765,
+      "learning_rate": 0.0009592906872509167,
+      "loss": 3.5259,
+      "step": 3425
+    },
+    {
+      "epoch": 1.0833925610045012,
+      "grad_norm": 0.14162144798225573,
+      "learning_rate": 0.0009590724509038579,
+      "loss": 3.6076,
+      "step": 3430
+    },
+    {
+      "epoch": 1.0849719655689805,
+      "grad_norm": 0.1370291892548981,
+      "learning_rate": 0.0009588536561136358,
+      "loss": 3.4767,
+      "step": 3435
+    },
+    {
+      "epoch": 1.0865513701334597,
+      "grad_norm": 0.11325953846961405,
+      "learning_rate": 0.0009586343031464055,
+      "loss": 3.4623,
+      "step": 3440
+    },
+    {
+      "epoch": 1.088130774697939,
+      "grad_norm": 0.17410040294918944,
+      "learning_rate": 0.0009584143922690008,
+      "loss": 3.5858,
+      "step": 3445
+    },
+    {
+      "epoch": 1.089710179262418,
+      "grad_norm": 0.20972770154954912,
+      "learning_rate": 0.0009581939237489346,
+      "loss": 3.7063,
+      "step": 3450
+    },
+    {
+      "epoch": 1.0912895838268972,
+      "grad_norm": 0.1923951132039094,
+      "learning_rate": 0.000957972897854398,
+      "loss": 3.5805,
+      "step": 3455
+    },
+    {
+      "epoch": 1.0928689883913765,
+      "grad_norm": 0.16654208813122776,
+      "learning_rate": 0.00095775131485426,
+      "loss": 3.4444,
+      "step": 3460
+    },
+    {
+      "epoch": 1.0944483929558557,
+      "grad_norm": 0.14994148376566602,
+      "learning_rate": 0.0009575291750180675,
+      "loss": 3.5015,
+      "step": 3465
+    },
+    {
+      "epoch": 1.0960277975203347,
+      "grad_norm": 0.1640550694396785,
+      "learning_rate": 0.0009573064786160446,
+      "loss": 3.5014,
+      "step": 3470
+    },
+    {
+      "epoch": 1.097607202084814,
+      "grad_norm": 0.24701794107392486,
+      "learning_rate": 0.0009570832259190927,
+      "loss": 3.4355,
+      "step": 3475
+    },
+    {
+      "epoch": 1.0991866066492932,
+      "grad_norm": 0.21545079608972195,
+      "learning_rate": 0.0009568594171987893,
+      "loss": 3.6293,
+      "step": 3480
+    },
+    {
+      "epoch": 1.1007660112137725,
+      "grad_norm": 0.17616944220718544,
+      "learning_rate": 0.000956635052727389,
+      "loss": 3.5283,
+      "step": 3485
+    },
+    {
+      "epoch": 1.1023454157782515,
+      "grad_norm": 0.1880146713181073,
+      "learning_rate": 0.0009564101327778223,
+      "loss": 3.6321,
+      "step": 3490
+    },
+    {
+      "epoch": 1.1039248203427308,
+      "grad_norm": 0.17587965507269773,
+      "learning_rate": 0.000956184657623695,
+      "loss": 3.5467,
+      "step": 3495
+    },
+    {
+      "epoch": 1.10550422490721,
+      "grad_norm": 0.15875400779821083,
+      "learning_rate": 0.0009559586275392887,
+      "loss": 3.5661,
+      "step": 3500
+    },
+    {
+      "epoch": 1.1070836294716893,
+      "grad_norm": 0.1526252926690391,
+      "learning_rate": 0.0009557320427995596,
+      "loss": 3.4174,
+      "step": 3505
+    },
+    {
+      "epoch": 1.1086630340361683,
+      "grad_norm": 0.16836938588204559,
+      "learning_rate": 0.0009555049036801393,
+      "loss": 3.5962,
+      "step": 3510
+    },
+    {
+      "epoch": 1.1102424386006475,
+      "grad_norm": 0.14694322307110463,
+      "learning_rate": 0.0009552772104573332,
+      "loss": 3.5284,
+      "step": 3515
+    },
+    {
+      "epoch": 1.1118218431651268,
+      "grad_norm": 0.14691445386163327,
+      "learning_rate": 0.0009550489634081212,
+      "loss": 3.5363,
+      "step": 3520
+    },
+    {
+      "epoch": 1.113401247729606,
+      "grad_norm": 0.12514595695150219,
+      "learning_rate": 0.0009548201628101563,
+      "loss": 3.4698,
+      "step": 3525
+    },
+    {
+      "epoch": 1.114980652294085,
+      "grad_norm": 0.1466478847716101,
+      "learning_rate": 0.0009545908089417654,
+      "loss": 3.6368,
+      "step": 3530
+    },
+    {
+      "epoch": 1.1165600568585643,
+      "grad_norm": 0.18722690247559393,
+      "learning_rate": 0.0009543609020819481,
+      "loss": 3.4909,
+      "step": 3535
+    },
+    {
+      "epoch": 1.1181394614230435,
+      "grad_norm": 0.196720740346297,
+      "learning_rate": 0.0009541304425103772,
+      "loss": 3.513,
+      "step": 3540
+    },
+    {
+      "epoch": 1.1197188659875228,
+      "grad_norm": 0.15597027154323442,
+      "learning_rate": 0.000953899430507397,
+      "loss": 3.3843,
+      "step": 3545
+    },
+    {
+      "epoch": 1.1212982705520018,
+      "grad_norm": 0.14693305434429949,
+      "learning_rate": 0.0009536678663540247,
+      "loss": 3.3456,
+      "step": 3550
+    },
+    {
+      "epoch": 1.122877675116481,
+      "grad_norm": 0.17155391191624927,
+      "learning_rate": 0.0009534357503319486,
+      "loss": 3.5331,
+      "step": 3555
+    },
+    {
+      "epoch": 1.1244570796809603,
+      "grad_norm": 0.15181214586540706,
+      "learning_rate": 0.0009532030827235286,
+      "loss": 3.5211,
+      "step": 3560
+    },
+    {
+      "epoch": 1.1260364842454396,
+      "grad_norm": 0.14866555288064323,
+      "learning_rate": 0.0009529698638117954,
+      "loss": 3.4759,
+      "step": 3565
+    },
+    {
+      "epoch": 1.1276158888099186,
+      "grad_norm": 0.1619355292370477,
+      "learning_rate": 0.0009527360938804503,
+      "loss": 3.5829,
+      "step": 3570
+    },
+    {
+      "epoch": 1.1291952933743978,
+      "grad_norm": 0.17080930945098738,
+      "learning_rate": 0.0009525017732138654,
+      "loss": 3.4134,
+      "step": 3575
+    },
+    {
+      "epoch": 1.130774697938877,
+      "grad_norm": 0.1791332389027379,
+      "learning_rate": 0.000952266902097082,
+      "loss": 3.5257,
+      "step": 3580
+    },
+    {
+      "epoch": 1.1323541025033563,
+      "grad_norm": 0.16287889926792246,
+      "learning_rate": 0.0009520314808158115,
+      "loss": 3.527,
+      "step": 3585
+    },
+    {
+      "epoch": 1.1339335070678354,
+      "grad_norm": 0.22831329282252832,
+      "learning_rate": 0.0009517955096564343,
+      "loss": 3.5467,
+      "step": 3590
+    },
+    {
+      "epoch": 1.1355129116323146,
+      "grad_norm": 0.21587039636593486,
+      "learning_rate": 0.000951558988906,
+      "loss": 3.4928,
+      "step": 3595
+    },
+    {
+      "epoch": 1.1370923161967939,
+      "grad_norm": 0.15515302065487302,
+      "learning_rate": 0.0009513219188522265,
+      "loss": 3.3949,
+      "step": 3600
+    },
+    {
+      "epoch": 1.1386717207612729,
+      "grad_norm": 0.19531249498739786,
+      "learning_rate": 0.0009510842997834999,
+      "loss": 3.4776,
+      "step": 3605
+    },
+    {
+      "epoch": 1.1402511253257521,
+      "grad_norm": 0.17309648961127838,
+      "learning_rate": 0.0009508461319888743,
+      "loss": 3.4294,
+      "step": 3610
+    },
+    {
+      "epoch": 1.1418305298902314,
+      "grad_norm": 0.20374478205073432,
+      "learning_rate": 0.0009506074157580715,
+      "loss": 3.5764,
+      "step": 3615
+    },
+    {
+      "epoch": 1.1434099344547106,
+      "grad_norm": 0.1674614562290747,
+      "learning_rate": 0.0009503681513814797,
+      "loss": 3.4229,
+      "step": 3620
+    },
+    {
+      "epoch": 1.1449893390191899,
+      "grad_norm": 0.24676865618415278,
+      "learning_rate": 0.0009501283391501547,
+      "loss": 3.5104,
+      "step": 3625
+    },
+    {
+      "epoch": 1.146568743583669,
+      "grad_norm": 0.15749402214670652,
+      "learning_rate": 0.0009498879793558184,
+      "loss": 3.3673,
+      "step": 3630
+    },
+    {
+      "epoch": 1.1481481481481481,
+      "grad_norm": 0.1876087417308556,
+      "learning_rate": 0.0009496470722908586,
+      "loss": 3.554,
+      "step": 3635
+    },
+    {
+      "epoch": 1.1497275527126274,
+      "grad_norm": 0.18506446437441917,
+      "learning_rate": 0.0009494056182483292,
+      "loss": 3.4664,
+      "step": 3640
+    },
+    {
+      "epoch": 1.1513069572771064,
+      "grad_norm": 0.1650474707934297,
+      "learning_rate": 0.0009491636175219495,
+      "loss": 3.5146,
+      "step": 3645
+    },
+    {
+      "epoch": 1.1528863618415857,
+      "grad_norm": 0.1649216280629778,
+      "learning_rate": 0.0009489210704061035,
+      "loss": 3.4681,
+      "step": 3650
+    },
+    {
+      "epoch": 1.154465766406065,
+      "grad_norm": 0.18817357619334923,
+      "learning_rate": 0.0009486779771958401,
+      "loss": 3.5334,
+      "step": 3655
+    },
+    {
+      "epoch": 1.1560451709705442,
+      "grad_norm": 0.2004719054160758,
+      "learning_rate": 0.0009484343381868721,
+      "loss": 3.5297,
+      "step": 3660
+    },
+    {
+      "epoch": 1.1576245755350234,
+      "grad_norm": 0.22442527289938496,
+      "learning_rate": 0.0009481901536755768,
+      "loss": 3.4078,
+      "step": 3665
+    },
+    {
+      "epoch": 1.1592039800995024,
+      "grad_norm": 0.18039908625370335,
+      "learning_rate": 0.0009479454239589947,
+      "loss": 3.475,
+      "step": 3670
+    },
+    {
+      "epoch": 1.1607833846639817,
+      "grad_norm": 0.1968176863518416,
+      "learning_rate": 0.00094770014933483,
+      "loss": 3.479,
+      "step": 3675
+    },
+    {
+      "epoch": 1.162362789228461,
+      "grad_norm": 0.20274756267717062,
+      "learning_rate": 0.0009474543301014489,
+      "loss": 3.4736,
+      "step": 3680
+    },
+    {
+      "epoch": 1.16394219379294,
+      "grad_norm": 0.19078790909078822,
+      "learning_rate": 0.000947207966557881,
+      "loss": 3.5295,
+      "step": 3685
+    },
+    {
+      "epoch": 1.1655215983574192,
+      "grad_norm": 0.14814168748866086,
+      "learning_rate": 0.0009469610590038173,
+      "loss": 3.361,
+      "step": 3690
+    },
+    {
+      "epoch": 1.1671010029218984,
+      "grad_norm": 0.16425480114557606,
+      "learning_rate": 0.0009467136077396113,
+      "loss": 3.6061,
+      "step": 3695
+    },
+    {
+      "epoch": 1.1686804074863777,
+      "grad_norm": 0.20199057519130498,
+      "learning_rate": 0.0009464656130662773,
+      "loss": 3.4519,
+      "step": 3700
+    },
+    {
+      "epoch": 1.170259812050857,
+      "grad_norm": 0.16626861962685233,
+      "learning_rate": 0.0009462170752854908,
+      "loss": 3.4247,
+      "step": 3705
+    },
+    {
+      "epoch": 1.171839216615336,
+      "grad_norm": 0.16817046067836208,
+      "learning_rate": 0.000945967994699588,
+      "loss": 3.5423,
+      "step": 3710
+    },
+    {
+      "epoch": 1.1734186211798152,
+      "grad_norm": 0.19718785456241722,
+      "learning_rate": 0.0009457183716115655,
+      "loss": 3.3503,
+      "step": 3715
+    },
+    {
+      "epoch": 1.1749980257442945,
+      "grad_norm": 0.13969712920238997,
+      "learning_rate": 0.0009454682063250797,
+      "loss": 3.4451,
+      "step": 3720
+    },
+    {
+      "epoch": 1.1765774303087735,
+      "grad_norm": 0.16584531547624867,
+      "learning_rate": 0.0009452174991444466,
+      "loss": 3.4738,
+      "step": 3725
+    },
+    {
+      "epoch": 1.1781568348732527,
+      "grad_norm": 0.21098180520274426,
+      "learning_rate": 0.0009449662503746415,
+      "loss": 3.5065,
+      "step": 3730
+    },
+    {
+      "epoch": 1.179736239437732,
+      "grad_norm": 0.19560991681344314,
+      "learning_rate": 0.0009447144603212983,
+      "loss": 3.476,
+      "step": 3735
+    },
+    {
+      "epoch": 1.1813156440022112,
+      "grad_norm": 0.19571878846306276,
+      "learning_rate": 0.0009444621292907094,
+      "loss": 3.5014,
+      "step": 3740
+    },
+    {
+      "epoch": 1.1828950485666905,
+      "grad_norm": 0.23022472553882056,
+      "learning_rate": 0.0009442092575898253,
+      "loss": 3.3936,
+      "step": 3745
+    },
+    {
+      "epoch": 1.1844744531311695,
+      "grad_norm": 0.16129418870589246,
+      "learning_rate": 0.0009439558455262547,
+      "loss": 3.438,
+      "step": 3750
+    },
+    {
+      "epoch": 1.1860538576956487,
+      "grad_norm": 0.1420454736987748,
+      "learning_rate": 0.0009437018934082626,
+      "loss": 3.5496,
+      "step": 3755
+    },
+    {
+      "epoch": 1.187633262260128,
+      "grad_norm": 0.15833224972021434,
+      "learning_rate": 0.0009434474015447721,
+      "loss": 3.5196,
+      "step": 3760
+    },
+    {
+      "epoch": 1.189212666824607,
+      "grad_norm": 0.159555391379725,
+      "learning_rate": 0.0009431923702453617,
+      "loss": 3.4208,
+      "step": 3765
+    },
+    {
+      "epoch": 1.1907920713890863,
+      "grad_norm": 0.18211420211116194,
+      "learning_rate": 0.0009429367998202671,
+      "loss": 3.3405,
+      "step": 3770
+    },
+    {
+      "epoch": 1.1923714759535655,
+      "grad_norm": 0.15914256531789245,
+      "learning_rate": 0.0009426806905803795,
+      "loss": 3.5496,
+      "step": 3775
+    },
+    {
+      "epoch": 1.1939508805180448,
+      "grad_norm": 0.18026779470969273,
+      "learning_rate": 0.0009424240428372452,
+      "loss": 3.4273,
+      "step": 3780
+    },
+    {
+      "epoch": 1.1955302850825238,
+      "grad_norm": 0.29969190627355774,
+      "learning_rate": 0.000942166856903066,
+      "loss": 3.4439,
+      "step": 3785
+    },
+    {
+      "epoch": 1.197109689647003,
+      "grad_norm": 0.17566538847915972,
+      "learning_rate": 0.0009419091330906984,
+      "loss": 3.3116,
+      "step": 3790
+    },
+    {
+      "epoch": 1.1986890942114823,
+      "grad_norm": 0.161322829058277,
+      "learning_rate": 0.0009416508717136527,
+      "loss": 3.5267,
+      "step": 3795
+    },
+    {
+      "epoch": 1.2002684987759615,
+      "grad_norm": 0.1361218966368206,
+      "learning_rate": 0.0009413920730860937,
+      "loss": 3.4379,
+      "step": 3800
+    },
+    {
+      "epoch": 1.2018479033404406,
+      "grad_norm": 0.21398086639722805,
+      "learning_rate": 0.0009411327375228394,
+      "loss": 3.4715,
+      "step": 3805
+    },
+    {
+      "epoch": 1.2034273079049198,
+      "grad_norm": 0.21550538792847884,
+      "learning_rate": 0.0009408728653393612,
+      "loss": 3.3665,
+      "step": 3810
+    },
+    {
+      "epoch": 1.205006712469399,
+      "grad_norm": 0.21460867662063146,
+      "learning_rate": 0.0009406124568517831,
+      "loss": 3.5338,
+      "step": 3815
+    },
+    {
+      "epoch": 1.2065861170338783,
+      "grad_norm": 0.19790984454451685,
+      "learning_rate": 0.0009403515123768816,
+      "loss": 3.3732,
+      "step": 3820
+    },
+    {
+      "epoch": 1.2081655215983573,
+      "grad_norm": 0.16044676782894002,
+      "learning_rate": 0.0009400900322320851,
+      "loss": 3.2932,
+      "step": 3825
+    },
+    {
+      "epoch": 1.2097449261628366,
+      "grad_norm": 0.1782455072532651,
+      "learning_rate": 0.0009398280167354735,
+      "loss": 3.3615,
+      "step": 3830
+    },
+    {
+      "epoch": 1.2113243307273158,
+      "grad_norm": 0.16056461369257277,
+      "learning_rate": 0.0009395654662057786,
+      "loss": 3.394,
+      "step": 3835
+    },
+    {
+      "epoch": 1.212903735291795,
+      "grad_norm": 0.1617823275360994,
+      "learning_rate": 0.000939302380962382,
+      "loss": 3.409,
+      "step": 3840
+    },
+    {
+      "epoch": 1.214483139856274,
+      "grad_norm": 0.20344574605731353,
+      "learning_rate": 0.0009390387613253166,
+      "loss": 3.5088,
+      "step": 3845
+    },
+    {
+      "epoch": 1.2160625444207533,
+      "grad_norm": 0.19760384614127924,
+      "learning_rate": 0.000938774607615265,
+      "loss": 3.4202,
+      "step": 3850
+    },
+    {
+      "epoch": 1.2176419489852326,
+      "grad_norm": 0.17019002490371693,
+      "learning_rate": 0.0009385099201535596,
+      "loss": 3.5585,
+      "step": 3855
+    },
+    {
+      "epoch": 1.2192213535497118,
+      "grad_norm": 0.17409540531216763,
+      "learning_rate": 0.000938244699262182,
+      "loss": 3.4943,
+      "step": 3860
+    },
+    {
+      "epoch": 1.2208007581141909,
+      "grad_norm": 0.15846100520357334,
+      "learning_rate": 0.0009379789452637629,
+      "loss": 3.3515,
+      "step": 3865
+    },
+    {
+      "epoch": 1.2223801626786701,
+      "grad_norm": 0.21408541248812699,
+      "learning_rate": 0.0009377126584815812,
+      "loss": 3.4941,
+      "step": 3870
+    },
+    {
+      "epoch": 1.2239595672431494,
+      "grad_norm": 0.16695132569529617,
+      "learning_rate": 0.000937445839239564,
+      "loss": 3.4852,
+      "step": 3875
+    },
+    {
+      "epoch": 1.2255389718076286,
+      "grad_norm": 0.16318138945603894,
+      "learning_rate": 0.0009371784878622863,
+      "loss": 3.4315,
+      "step": 3880
+    },
+    {
+      "epoch": 1.2271183763721076,
+      "grad_norm": 0.19738953018555389,
+      "learning_rate": 0.0009369106046749703,
+      "loss": 3.5117,
+      "step": 3885
+    },
+    {
+      "epoch": 1.2286977809365869,
+      "grad_norm": 0.14204527279882048,
+      "learning_rate": 0.0009366421900034849,
+      "loss": 3.2596,
+      "step": 3890
+    },
+    {
+      "epoch": 1.2302771855010661,
+      "grad_norm": 0.14898950129104696,
+      "learning_rate": 0.0009363732441743459,
+      "loss": 3.2987,
+      "step": 3895
+    },
+    {
+      "epoch": 1.2318565900655454,
+      "grad_norm": 0.1683677844282376,
+      "learning_rate": 0.0009361037675147152,
+      "loss": 3.4681,
+      "step": 3900
+    },
+    {
+      "epoch": 1.2334359946300244,
+      "grad_norm": 0.17218443452504015,
+      "learning_rate": 0.0009358337603524001,
+      "loss": 3.4559,
+      "step": 3905
+    },
+    {
+      "epoch": 1.2350153991945036,
+      "grad_norm": 0.16230059619775924,
+      "learning_rate": 0.0009355632230158537,
+      "loss": 3.4437,
+      "step": 3910
+    },
+    {
+      "epoch": 1.236594803758983,
+      "grad_norm": 0.1637629344964814,
+      "learning_rate": 0.0009352921558341734,
+      "loss": 3.3447,
+      "step": 3915
+    },
+    {
+      "epoch": 1.2381742083234621,
+      "grad_norm": 0.18387059418034044,
+      "learning_rate": 0.0009350205591371019,
+      "loss": 3.3898,
+      "step": 3920
+    },
+    {
+      "epoch": 1.2397536128879412,
+      "grad_norm": 0.12842797710859524,
+      "learning_rate": 0.0009347484332550255,
+      "loss": 3.3435,
+      "step": 3925
+    },
+    {
+      "epoch": 1.2413330174524204,
+      "grad_norm": 0.1423647750556377,
+      "learning_rate": 0.0009344757785189744,
+      "loss": 3.3623,
+      "step": 3930
+    },
+    {
+      "epoch": 1.2429124220168997,
+      "grad_norm": 0.17711276858177502,
+      "learning_rate": 0.0009342025952606219,
+      "loss": 3.4316,
+      "step": 3935
+    },
+    {
+      "epoch": 1.244491826581379,
+      "grad_norm": 0.1647045199172262,
+      "learning_rate": 0.0009339288838122849,
+      "loss": 3.4343,
+      "step": 3940
+    },
+    {
+      "epoch": 1.246071231145858,
+      "grad_norm": 0.19616295235626366,
+      "learning_rate": 0.0009336546445069218,
+      "loss": 3.4433,
+      "step": 3945
+    },
+    {
+      "epoch": 1.2476506357103372,
+      "grad_norm": 0.1338067737830234,
+      "learning_rate": 0.0009333798776781343,
+      "loss": 3.4647,
+      "step": 3950
+    },
+    {
+      "epoch": 1.2492300402748164,
+      "grad_norm": 0.1674958325821782,
+      "learning_rate": 0.0009331045836601646,
+      "loss": 3.4153,
+      "step": 3955
+    },
+    {
+      "epoch": 1.2508094448392955,
+      "grad_norm": 0.164883261301351,
+      "learning_rate": 0.0009328287627878973,
+      "loss": 3.3946,
+      "step": 3960
+    },
+    {
+      "epoch": 1.2523888494037747,
+      "grad_norm": 0.1285324062812115,
+      "learning_rate": 0.000932552415396857,
+      "loss": 3.2798,
+      "step": 3965
+    },
+    {
+      "epoch": 1.253968253968254,
+      "grad_norm": 0.1708840737137744,
+      "learning_rate": 0.0009322755418232094,
+      "loss": 3.4297,
+      "step": 3970
+    },
+    {
+      "epoch": 1.2555476585327332,
+      "grad_norm": 0.143100846348656,
+      "learning_rate": 0.00093199814240376,
+      "loss": 3.524,
+      "step": 3975
+    },
+    {
+      "epoch": 1.2571270630972124,
+      "grad_norm": 0.1870541648656571,
+      "learning_rate": 0.000931720217475954,
+      "loss": 3.5105,
+      "step": 3980
+    },
+    {
+      "epoch": 1.2587064676616915,
+      "grad_norm": 0.1494970239923711,
+      "learning_rate": 0.000931441767377876,
+      "loss": 3.3396,
+      "step": 3985
+    },
+    {
+      "epoch": 1.2602858722261707,
+      "grad_norm": 0.13931866315641525,
+      "learning_rate": 0.0009311627924482493,
+      "loss": 3.3782,
+      "step": 3990
+    },
+    {
+      "epoch": 1.26186527679065,
+      "grad_norm": 0.1415432981936499,
+      "learning_rate": 0.0009308832930264354,
+      "loss": 3.3677,
+      "step": 3995
+    },
+    {
+      "epoch": 1.263444681355129,
+      "grad_norm": 0.16776120238681047,
+      "learning_rate": 0.0009306032694524345,
+      "loss": 3.3866,
+      "step": 4000
+    },
+    {
+      "epoch": 1.2650240859196082,
+      "grad_norm": 0.1786871644484267,
+      "learning_rate": 0.000930322722066884,
+      "loss": 3.3221,
+      "step": 4005
+    },
+    {
+      "epoch": 1.2666034904840875,
+      "grad_norm": 0.16315647151552112,
+      "learning_rate": 0.0009300416512110581,
+      "loss": 3.4461,
+      "step": 4010
+    },
+    {
+      "epoch": 1.2681828950485667,
+      "grad_norm": 0.1552733417787517,
+      "learning_rate": 0.0009297600572268685,
+      "loss": 3.3432,
+      "step": 4015
+    },
+    {
+      "epoch": 1.269762299613046,
+      "grad_norm": 0.1522952178479541,
+      "learning_rate": 0.0009294779404568629,
+      "loss": 3.385,
+      "step": 4020
+    },
+    {
+      "epoch": 1.271341704177525,
+      "grad_norm": 0.1662187283663952,
+      "learning_rate": 0.000929195301244225,
+      "loss": 3.4102,
+      "step": 4025
+    },
+    {
+      "epoch": 1.2729211087420043,
+      "grad_norm": 0.14647477444085574,
+      "learning_rate": 0.000928912139932774,
+      "loss": 3.3386,
+      "step": 4030
+    },
+    {
+      "epoch": 1.2745005133064835,
+      "grad_norm": 0.16711729761359687,
+      "learning_rate": 0.0009286284568669643,
+      "loss": 3.4209,
+      "step": 4035
+    },
+    {
+      "epoch": 1.2760799178709625,
+      "grad_norm": 0.14743592937515496,
+      "learning_rate": 0.0009283442523918848,
+      "loss": 3.4822,
+      "step": 4040
+    },
+    {
+      "epoch": 1.2776593224354418,
+      "grad_norm": 0.1366017208398044,
+      "learning_rate": 0.000928059526853259,
+      "loss": 3.4677,
+      "step": 4045
+    },
+    {
+      "epoch": 1.279238726999921,
+      "grad_norm": 0.16704429698198514,
+      "learning_rate": 0.000927774280597444,
+      "loss": 3.4226,
+      "step": 4050
+    },
+    {
+      "epoch": 1.2808181315644003,
+      "grad_norm": 0.15643121589785522,
+      "learning_rate": 0.0009274885139714302,
+      "loss": 3.3408,
+      "step": 4055
+    },
+    {
+      "epoch": 1.2823975361288795,
+      "grad_norm": 0.1741639877841723,
+      "learning_rate": 0.0009272022273228414,
+      "loss": 3.4501,
+      "step": 4060
+    },
+    {
+      "epoch": 1.2839769406933585,
+      "grad_norm": 0.11217281345430413,
+      "learning_rate": 0.0009269154209999338,
+      "loss": 3.296,
+      "step": 4065
+    },
+    {
+      "epoch": 1.2855563452578378,
+      "grad_norm": 0.1617728847897573,
+      "learning_rate": 0.0009266280953515956,
+      "loss": 3.3026,
+      "step": 4070
+    },
+    {
+      "epoch": 1.287135749822317,
+      "grad_norm": 0.14427201373975365,
+      "learning_rate": 0.0009263402507273471,
+      "loss": 3.4153,
+      "step": 4075
+    },
+    {
+      "epoch": 1.288715154386796,
+      "grad_norm": 0.17014303360059407,
+      "learning_rate": 0.0009260518874773394,
+      "loss": 3.3237,
+      "step": 4080
+    },
+    {
+      "epoch": 1.2902945589512753,
+      "grad_norm": 0.14224068939581733,
+      "learning_rate": 0.0009257630059523552,
+      "loss": 3.4282,
+      "step": 4085
+    },
+    {
+      "epoch": 1.2918739635157546,
+      "grad_norm": 0.17799994513807274,
+      "learning_rate": 0.0009254736065038068,
+      "loss": 3.339,
+      "step": 4090
+    },
+    {
+      "epoch": 1.2934533680802338,
+      "grad_norm": 0.1364276503857556,
+      "learning_rate": 0.0009251836894837374,
+      "loss": 3.3944,
+      "step": 4095
+    },
+    {
+      "epoch": 1.295032772644713,
+      "grad_norm": 0.14641284207636293,
+      "learning_rate": 0.0009248932552448191,
+      "loss": 3.3047,
+      "step": 4100
+    },
+    {
+      "epoch": 1.296612177209192,
+      "grad_norm": 0.16244756151602185,
+      "learning_rate": 0.0009246023041403535,
+      "loss": 3.3813,
+      "step": 4105
+    },
+    {
+      "epoch": 1.2981915817736713,
+      "grad_norm": 0.12556159589538443,
+      "learning_rate": 0.0009243108365242711,
+      "loss": 3.3201,
+      "step": 4110
+    },
+    {
+      "epoch": 1.2997709863381506,
+      "grad_norm": 0.1905706972967778,
+      "learning_rate": 0.0009240188527511303,
+      "loss": 3.3823,
+      "step": 4115
+    },
+    {
+      "epoch": 1.3013503909026296,
+      "grad_norm": 0.20641498529248553,
+      "learning_rate": 0.0009237263531761177,
+      "loss": 3.4022,
+      "step": 4120
+    },
+    {
+      "epoch": 1.3029297954671089,
+      "grad_norm": 0.1545220182472424,
+      "learning_rate": 0.0009234333381550472,
+      "loss": 3.2973,
+      "step": 4125
+    },
+    {
+      "epoch": 1.304509200031588,
+      "grad_norm": 0.18168937803347265,
+      "learning_rate": 0.0009231398080443601,
+      "loss": 3.4562,
+      "step": 4130
+    },
+    {
+      "epoch": 1.3060886045960673,
+      "grad_norm": 0.23826734915746944,
+      "learning_rate": 0.0009228457632011235,
+      "loss": 3.3164,
+      "step": 4135
+    },
+    {
+      "epoch": 1.3076680091605466,
+      "grad_norm": 0.16581129822228094,
+      "learning_rate": 0.0009225512039830315,
+      "loss": 3.3584,
+      "step": 4140
+    },
+    {
+      "epoch": 1.3092474137250256,
+      "grad_norm": 0.16210329879726387,
+      "learning_rate": 0.0009222561307484032,
+      "loss": 3.4051,
+      "step": 4145
+    },
+    {
+      "epoch": 1.3108268182895049,
+      "grad_norm": 0.15506171661463133,
+      "learning_rate": 0.0009219605438561836,
+      "loss": 3.3442,
+      "step": 4150
+    },
+    {
+      "epoch": 1.3124062228539841,
+      "grad_norm": 0.15725462551651762,
+      "learning_rate": 0.0009216644436659422,
+      "loss": 3.3336,
+      "step": 4155
+    },
+    {
+      "epoch": 1.3139856274184631,
+      "grad_norm": 0.1401664797264424,
+      "learning_rate": 0.0009213678305378727,
+      "loss": 3.35,
+      "step": 4160
+    },
+    {
+      "epoch": 1.3155650319829424,
+      "grad_norm": 0.13193412044498434,
+      "learning_rate": 0.0009210707048327935,
+      "loss": 3.3959,
+      "step": 4165
+    },
+    {
+      "epoch": 1.3171444365474216,
+      "grad_norm": 0.1643713503348713,
+      "learning_rate": 0.0009207730669121457,
+      "loss": 3.4629,
+      "step": 4170
+    },
+    {
+      "epoch": 1.3187238411119009,
+      "grad_norm": 0.18797562194535233,
+      "learning_rate": 0.000920474917137994,
+      "loss": 3.3898,
+      "step": 4175
+    },
+    {
+      "epoch": 1.3203032456763801,
+      "grad_norm": 0.1439925271657658,
+      "learning_rate": 0.0009201762558730255,
+      "loss": 3.427,
+      "step": 4180
+    },
+    {
+      "epoch": 1.3218826502408592,
+      "grad_norm": 0.17414529000788978,
+      "learning_rate": 0.0009198770834805498,
+      "loss": 3.3501,
+      "step": 4185
+    },
+    {
+      "epoch": 1.3234620548053384,
+      "grad_norm": 0.14722148228092627,
+      "learning_rate": 0.0009195774003244979,
+      "loss": 3.355,
+      "step": 4190
+    },
+    {
+      "epoch": 1.3250414593698177,
+      "grad_norm": 0.14702051757999493,
+      "learning_rate": 0.0009192772067694223,
+      "loss": 3.3385,
+      "step": 4195
+    },
+    {
+      "epoch": 1.3266208639342967,
+      "grad_norm": 0.1373432579980681,
+      "learning_rate": 0.0009189765031804965,
+      "loss": 3.3479,
+      "step": 4200
+    },
+    {
+      "epoch": 1.328200268498776,
+      "grad_norm": 0.13838263702623113,
+      "learning_rate": 0.0009186752899235142,
+      "loss": 3.3044,
+      "step": 4205
+    },
+    {
+      "epoch": 1.3297796730632552,
+      "grad_norm": 0.11732374721073062,
+      "learning_rate": 0.0009183735673648893,
+      "loss": 3.3196,
+      "step": 4210
+    },
+    {
+      "epoch": 1.3313590776277344,
+      "grad_norm": 0.13745564723029,
+      "learning_rate": 0.000918071335871655,
+      "loss": 3.3737,
+      "step": 4215
+    },
+    {
+      "epoch": 1.3329384821922137,
+      "grad_norm": 0.14220836178084498,
+      "learning_rate": 0.0009177685958114641,
+      "loss": 3.3583,
+      "step": 4220
+    },
+    {
+      "epoch": 1.3345178867566927,
+      "grad_norm": 0.16452014627189307,
+      "learning_rate": 0.0009174653475525874,
+      "loss": 3.2648,
+      "step": 4225
+    },
+    {
+      "epoch": 1.336097291321172,
+      "grad_norm": 0.14610729955363427,
+      "learning_rate": 0.0009171615914639142,
+      "loss": 3.4392,
+      "step": 4230
+    },
+    {
+      "epoch": 1.3376766958856512,
+      "grad_norm": 0.18117209766134101,
+      "learning_rate": 0.0009168573279149515,
+      "loss": 3.3022,
+      "step": 4235
+    },
+    {
+      "epoch": 1.3392561004501302,
+      "grad_norm": 0.19782231311409362,
+      "learning_rate": 0.000916552557275824,
+      "loss": 3.4022,
+      "step": 4240
+    },
+    {
+      "epoch": 1.3408355050146095,
+      "grad_norm": 0.17241089526198844,
+      "learning_rate": 0.0009162472799172725,
+      "loss": 3.3588,
+      "step": 4245
+    },
+    {
+      "epoch": 1.3424149095790887,
+      "grad_norm": 0.16410686082023807,
+      "learning_rate": 0.000915941496210655,
+      "loss": 3.3589,
+      "step": 4250
+    },
+    {
+      "epoch": 1.343994314143568,
+      "grad_norm": 0.2388905353194743,
+      "learning_rate": 0.0009156352065279448,
+      "loss": 3.3672,
+      "step": 4255
+    },
+    {
+      "epoch": 1.345573718708047,
+      "grad_norm": 0.1527648247113203,
+      "learning_rate": 0.0009153284112417313,
+      "loss": 3.2461,
+      "step": 4260
+    },
+    {
+      "epoch": 1.3471531232725262,
+      "grad_norm": 0.153856870172395,
+      "learning_rate": 0.0009150211107252181,
+      "loss": 3.3544,
+      "step": 4265
+    },
+    {
+      "epoch": 1.3487325278370055,
+      "grad_norm": 0.1539199272236105,
+      "learning_rate": 0.0009147133053522243,
+      "loss": 3.2858,
+      "step": 4270
+    },
+    {
+      "epoch": 1.3503119324014845,
+      "grad_norm": 0.1609833682589832,
+      "learning_rate": 0.0009144049954971827,
+      "loss": 3.309,
+      "step": 4275
+    },
+    {
+      "epoch": 1.3518913369659638,
+      "grad_norm": 0.19124078172755124,
+      "learning_rate": 0.0009140961815351399,
+      "loss": 3.3246,
+      "step": 4280
+    },
+    {
+      "epoch": 1.353470741530443,
+      "grad_norm": 0.16190296279226263,
+      "learning_rate": 0.0009137868638417555,
+      "loss": 3.3528,
+      "step": 4285
+    },
+    {
+      "epoch": 1.3550501460949222,
+      "grad_norm": 0.1568394556153337,
+      "learning_rate": 0.0009134770427933019,
+      "loss": 3.324,
+      "step": 4290
+    },
+    {
+      "epoch": 1.3566295506594015,
+      "grad_norm": 0.1418128016487037,
+      "learning_rate": 0.0009131667187666642,
+      "loss": 3.3073,
+      "step": 4295
+    },
+    {
+      "epoch": 1.3582089552238805,
+      "grad_norm": 0.14972736307745682,
+      "learning_rate": 0.000912855892139339,
+      "loss": 3.249,
+      "step": 4300
+    },
+    {
+      "epoch": 1.3597883597883598,
+      "grad_norm": 0.11564072656402402,
+      "learning_rate": 0.0009125445632894345,
+      "loss": 3.2773,
+      "step": 4305
+    },
+    {
+      "epoch": 1.361367764352839,
+      "grad_norm": 0.1457945817000172,
+      "learning_rate": 0.0009122327325956696,
+      "loss": 3.3964,
+      "step": 4310
+    },
+    {
+      "epoch": 1.362947168917318,
+      "grad_norm": 0.15166900783465612,
+      "learning_rate": 0.0009119204004373738,
+      "loss": 3.3749,
+      "step": 4315
+    },
+    {
+      "epoch": 1.3645265734817973,
+      "grad_norm": 0.16788847097200998,
+      "learning_rate": 0.0009116075671944864,
+      "loss": 3.2626,
+      "step": 4320
+    },
+    {
+      "epoch": 1.3661059780462765,
+      "grad_norm": 0.18345038602626523,
+      "learning_rate": 0.0009112942332475569,
+      "loss": 3.4544,
+      "step": 4325
+    },
+    {
+      "epoch": 1.3676853826107558,
+      "grad_norm": 0.16980153530949338,
+      "learning_rate": 0.0009109803989777431,
+      "loss": 3.4096,
+      "step": 4330
+    },
+    {
+      "epoch": 1.369264787175235,
+      "grad_norm": 0.1901152276047917,
+      "learning_rate": 0.0009106660647668118,
+      "loss": 3.375,
+      "step": 4335
+    },
+    {
+      "epoch": 1.370844191739714,
+      "grad_norm": 0.13319747933268186,
+      "learning_rate": 0.000910351230997138,
+      "loss": 3.2241,
+      "step": 4340
+    },
+    {
+      "epoch": 1.3724235963041933,
+      "grad_norm": 0.19034250534823186,
+      "learning_rate": 0.0009100358980517043,
+      "loss": 3.2717,
+      "step": 4345
+    },
+    {
+      "epoch": 1.3740030008686726,
+      "grad_norm": 0.16377491037494543,
+      "learning_rate": 0.0009097200663141005,
+      "loss": 3.3572,
+      "step": 4350
+    },
+    {
+      "epoch": 1.3755824054331516,
+      "grad_norm": 0.15838101408425387,
+      "learning_rate": 0.0009094037361685232,
+      "loss": 3.339,
+      "step": 4355
+    },
+    {
+      "epoch": 1.3771618099976308,
+      "grad_norm": 0.15030123154713862,
+      "learning_rate": 0.0009090869079997754,
+      "loss": 3.2,
+      "step": 4360
+    },
+    {
+      "epoch": 1.37874121456211,
+      "grad_norm": 0.19112961814499516,
+      "learning_rate": 0.0009087695821932657,
+      "loss": 3.3443,
+      "step": 4365
+    },
+    {
+      "epoch": 1.3803206191265893,
+      "grad_norm": 0.17058851459547694,
+      "learning_rate": 0.0009084517591350083,
+      "loss": 3.2637,
+      "step": 4370
+    },
+    {
+      "epoch": 1.3819000236910686,
+      "grad_norm": 0.17372701953450628,
+      "learning_rate": 0.0009081334392116218,
+      "loss": 3.3439,
+      "step": 4375
+    },
+    {
+      "epoch": 1.3834794282555476,
+      "grad_norm": 0.16146165100584847,
+      "learning_rate": 0.0009078146228103301,
+      "loss": 3.3357,
+      "step": 4380
+    },
+    {
+      "epoch": 1.3850588328200268,
+      "grad_norm": 0.1959721032697362,
+      "learning_rate": 0.0009074953103189602,
+      "loss": 3.3979,
+      "step": 4385
+    },
+    {
+      "epoch": 1.386638237384506,
+      "grad_norm": 0.13458053922970173,
+      "learning_rate": 0.0009071755021259429,
+      "loss": 3.3486,
+      "step": 4390
+    },
+    {
+      "epoch": 1.3882176419489851,
+      "grad_norm": 0.17061529633912267,
+      "learning_rate": 0.0009068551986203122,
+      "loss": 3.4066,
+      "step": 4395
+    },
+    {
+      "epoch": 1.3897970465134644,
+      "grad_norm": 0.20409384331794853,
+      "learning_rate": 0.0009065344001917042,
+      "loss": 3.2958,
+      "step": 4400
+    },
+    {
+      "epoch": 1.3913764510779436,
+      "grad_norm": 0.16679294495837618,
+      "learning_rate": 0.0009062131072303572,
+      "loss": 3.3132,
+      "step": 4405
+    },
+    {
+      "epoch": 1.3929558556424229,
+      "grad_norm": 0.1952635203259422,
+      "learning_rate": 0.0009058913201271116,
+      "loss": 3.2498,
+      "step": 4410
+    },
+    {
+      "epoch": 1.394535260206902,
+      "grad_norm": 0.15385051619032689,
+      "learning_rate": 0.000905569039273408,
+      "loss": 3.3008,
+      "step": 4415
+    },
+    {
+      "epoch": 1.3961146647713811,
+      "grad_norm": 0.2486791036593673,
+      "learning_rate": 0.0009052462650612885,
+      "loss": 3.3204,
+      "step": 4420
+    },
+    {
+      "epoch": 1.3976940693358604,
+      "grad_norm": 0.18045094172214962,
+      "learning_rate": 0.0009049229978833945,
+      "loss": 3.2637,
+      "step": 4425
+    },
+    {
+      "epoch": 1.3992734739003396,
+      "grad_norm": 0.151424008067985,
+      "learning_rate": 0.0009045992381329678,
+      "loss": 3.2764,
+      "step": 4430
+    },
+    {
+      "epoch": 1.4008528784648187,
+      "grad_norm": 0.2132183057940956,
+      "learning_rate": 0.0009042749862038491,
+      "loss": 3.2443,
+      "step": 4435
+    },
+    {
+      "epoch": 1.402432283029298,
+      "grad_norm": 0.16270294354698586,
+      "learning_rate": 0.0009039502424904777,
+      "loss": 3.2845,
+      "step": 4440
+    },
+    {
+      "epoch": 1.4040116875937771,
+      "grad_norm": 0.18019334099532144,
+      "learning_rate": 0.0009036250073878913,
+      "loss": 3.2478,
+      "step": 4445
+    },
+    {
+      "epoch": 1.4055910921582564,
+      "grad_norm": 0.1393773438924934,
+      "learning_rate": 0.0009032992812917254,
+      "loss": 3.2306,
+      "step": 4450
+    },
+    {
+      "epoch": 1.4071704967227356,
+      "grad_norm": 0.15273142716044932,
+      "learning_rate": 0.0009029730645982126,
+      "loss": 3.2648,
+      "step": 4455
+    },
+    {
+      "epoch": 1.4087499012872147,
+      "grad_norm": 0.1740693335929456,
+      "learning_rate": 0.0009026463577041824,
+      "loss": 3.3307,
+      "step": 4460
+    },
+    {
+      "epoch": 1.410329305851694,
+      "grad_norm": 0.19665063654880002,
+      "learning_rate": 0.0009023191610070607,
+      "loss": 3.3359,
+      "step": 4465
+    },
+    {
+      "epoch": 1.4119087104161732,
+      "grad_norm": 0.12561620916919017,
+      "learning_rate": 0.0009019914749048689,
+      "loss": 3.2632,
+      "step": 4470
+    },
+    {
+      "epoch": 1.4134881149806522,
+      "grad_norm": 0.13644396705083753,
+      "learning_rate": 0.0009016632997962241,
+      "loss": 3.3341,
+      "step": 4475
+    },
+    {
+      "epoch": 1.4150675195451314,
+      "grad_norm": 0.1665415155549509,
+      "learning_rate": 0.0009013346360803381,
+      "loss": 3.4047,
+      "step": 4480
+    },
+    {
+      "epoch": 1.4166469241096107,
+      "grad_norm": 0.15624730882339172,
+      "learning_rate": 0.000901005484157017,
+      "loss": 3.2707,
+      "step": 4485
+    },
+    {
+      "epoch": 1.41822632867409,
+      "grad_norm": 0.16793430979941182,
+      "learning_rate": 0.000900675844426661,
+      "loss": 3.4758,
+      "step": 4490
+    },
+    {
+      "epoch": 1.4198057332385692,
+      "grad_norm": 0.15984543570755744,
+      "learning_rate": 0.0009003457172902636,
+      "loss": 3.3969,
+      "step": 4495
+    },
+    {
+      "epoch": 1.4213851378030482,
+      "grad_norm": 0.13697976713292376,
+      "learning_rate": 0.0009000151031494109,
+      "loss": 3.3826,
+      "step": 4500
+    },
+    {
+      "epoch": 1.4229645423675275,
+      "grad_norm": 0.19809550193039518,
+      "learning_rate": 0.000899684002406282,
+      "loss": 3.2361,
+      "step": 4505
+    },
+    {
+      "epoch": 1.4245439469320067,
+      "grad_norm": 0.16229709695027106,
+      "learning_rate": 0.0008993524154636474,
+      "loss": 3.3063,
+      "step": 4510
+    },
+    {
+      "epoch": 1.4261233514964857,
+      "grad_norm": 0.12608660845067726,
+      "learning_rate": 0.0008990203427248696,
+      "loss": 3.2894,
+      "step": 4515
+    },
+    {
+      "epoch": 1.427702756060965,
+      "grad_norm": 0.1797964516783041,
+      "learning_rate": 0.0008986877845939014,
+      "loss": 3.3271,
+      "step": 4520
+    },
+    {
+      "epoch": 1.4292821606254442,
+      "grad_norm": 0.15675599346145377,
+      "learning_rate": 0.0008983547414752864,
+      "loss": 3.3151,
+      "step": 4525
+    },
+    {
+      "epoch": 1.4308615651899235,
+      "grad_norm": 0.1892068686504464,
+      "learning_rate": 0.0008980212137741584,
+      "loss": 3.2629,
+      "step": 4530
+    },
+    {
+      "epoch": 1.4324409697544027,
+      "grad_norm": 0.18537794590349257,
+      "learning_rate": 0.0008976872018962401,
+      "loss": 3.2736,
+      "step": 4535
+    },
+    {
+      "epoch": 1.4340203743188817,
+      "grad_norm": 0.2593286392625844,
+      "learning_rate": 0.0008973527062478438,
+      "loss": 3.3368,
+      "step": 4540
+    },
+    {
+      "epoch": 1.435599778883361,
+      "grad_norm": 0.17720765577244244,
+      "learning_rate": 0.0008970177272358698,
+      "loss": 3.222,
+      "step": 4545
+    },
+    {
+      "epoch": 1.4371791834478402,
+      "grad_norm": 0.14256311689707032,
+      "learning_rate": 0.0008966822652678067,
+      "loss": 3.2242,
+      "step": 4550
+    },
+    {
+      "epoch": 1.4387585880123193,
+      "grad_norm": 0.16384644938661744,
+      "learning_rate": 0.0008963463207517304,
+      "loss": 3.2097,
+      "step": 4555
+    },
+    {
+      "epoch": 1.4403379925767985,
+      "grad_norm": 0.1877920058936674,
+      "learning_rate": 0.0008960098940963041,
+      "loss": 3.3619,
+      "step": 4560
+    },
+    {
+      "epoch": 1.4419173971412778,
+      "grad_norm": 0.12584487034089764,
+      "learning_rate": 0.000895672985710777,
+      "loss": 3.2728,
+      "step": 4565
+    },
+    {
+      "epoch": 1.443496801705757,
+      "grad_norm": 0.13746308694505036,
+      "learning_rate": 0.0008953355960049847,
+      "loss": 3.4137,
+      "step": 4570
+    },
+    {
+      "epoch": 1.4450762062702363,
+      "grad_norm": 0.1578748118489456,
+      "learning_rate": 0.0008949977253893483,
+      "loss": 3.1916,
+      "step": 4575
+    },
+    {
+      "epoch": 1.4466556108347153,
+      "grad_norm": 0.1673051556573036,
+      "learning_rate": 0.0008946593742748737,
+      "loss": 3.2818,
+      "step": 4580
+    },
+    {
+      "epoch": 1.4482350153991945,
+      "grad_norm": 0.1328414390545332,
+      "learning_rate": 0.0008943205430731514,
+      "loss": 3.3013,
+      "step": 4585
+    },
+    {
+      "epoch": 1.4498144199636738,
+      "grad_norm": 0.14435362259403745,
+      "learning_rate": 0.0008939812321963561,
+      "loss": 3.2534,
+      "step": 4590
+    },
+    {
+      "epoch": 1.4513938245281528,
+      "grad_norm": 0.14208374901699425,
+      "learning_rate": 0.0008936414420572457,
+      "loss": 3.2995,
+      "step": 4595
+    },
+    {
+      "epoch": 1.452973229092632,
+      "grad_norm": 0.14376283024706615,
+      "learning_rate": 0.0008933011730691609,
+      "loss": 3.2659,
+      "step": 4600
+    },
+    {
+      "epoch": 1.4545526336571113,
+      "grad_norm": 0.1599547546108757,
+      "learning_rate": 0.0008929604256460258,
+      "loss": 3.2932,
+      "step": 4605
+    },
+    {
+      "epoch": 1.4561320382215905,
+      "grad_norm": 0.16535842899706282,
+      "learning_rate": 0.0008926192002023457,
+      "loss": 3.2339,
+      "step": 4610
+    },
+    {
+      "epoch": 1.4577114427860698,
+      "grad_norm": 0.12552135639645665,
+      "learning_rate": 0.0008922774971532076,
+      "loss": 3.1256,
+      "step": 4615
+    },
+    {
+      "epoch": 1.4592908473505488,
+      "grad_norm": 0.14820947272345442,
+      "learning_rate": 0.0008919353169142794,
+      "loss": 3.2025,
+      "step": 4620
+    },
+    {
+      "epoch": 1.460870251915028,
+      "grad_norm": 0.1376996788111554,
+      "learning_rate": 0.0008915926599018098,
+      "loss": 3.2946,
+      "step": 4625
+    },
+    {
+      "epoch": 1.462449656479507,
+      "grad_norm": 0.15470840871930625,
+      "learning_rate": 0.0008912495265326273,
+      "loss": 3.2855,
+      "step": 4630
+    },
+    {
+      "epoch": 1.4640290610439863,
+      "grad_norm": 0.14143468530300238,
+      "learning_rate": 0.0008909059172241395,
+      "loss": 3.1603,
+      "step": 4635
+    },
+    {
+      "epoch": 1.4656084656084656,
+      "grad_norm": 0.11621783222623613,
+      "learning_rate": 0.0008905618323943337,
+      "loss": 3.263,
+      "step": 4640
+    },
+    {
+      "epoch": 1.4671878701729448,
+      "grad_norm": 0.14653375693300938,
+      "learning_rate": 0.0008902172724617747,
+      "loss": 3.2858,
+      "step": 4645
+    },
+    {
+      "epoch": 1.468767274737424,
+      "grad_norm": 0.12686687887850526,
+      "learning_rate": 0.0008898722378456066,
+      "loss": 3.3301,
+      "step": 4650
+    },
+    {
+      "epoch": 1.470346679301903,
+      "grad_norm": 0.1725879758751388,
+      "learning_rate": 0.0008895267289655493,
+      "loss": 3.3617,
+      "step": 4655
+    },
+    {
+      "epoch": 1.4719260838663824,
+      "grad_norm": 0.16206213990180582,
+      "learning_rate": 0.000889180746241901,
+      "loss": 3.211,
+      "step": 4660
+    },
+    {
+      "epoch": 1.4735054884308616,
+      "grad_norm": 0.17730217086064257,
+      "learning_rate": 0.0008888342900955355,
+      "loss": 3.2401,
+      "step": 4665
+    },
+    {
+      "epoch": 1.4750848929953406,
+      "grad_norm": 0.15775239883109177,
+      "learning_rate": 0.000888487360947903,
+      "loss": 3.2833,
+      "step": 4670
+    },
+    {
+      "epoch": 1.4766642975598199,
+      "grad_norm": 0.17147232993014616,
+      "learning_rate": 0.0008881399592210286,
+      "loss": 3.2309,
+      "step": 4675
+    },
+    {
+      "epoch": 1.4782437021242991,
+      "grad_norm": 0.1582103899231393,
+      "learning_rate": 0.0008877920853375125,
+      "loss": 3.3207,
+      "step": 4680
+    },
+    {
+      "epoch": 1.4798231066887784,
+      "grad_norm": 0.16464760205681195,
+      "learning_rate": 0.0008874437397205295,
+      "loss": 3.2625,
+      "step": 4685
+    },
+    {
+      "epoch": 1.4814025112532576,
+      "grad_norm": 0.1573618442182188,
+      "learning_rate": 0.000887094922793828,
+      "loss": 3.2645,
+      "step": 4690
+    },
+    {
+      "epoch": 1.4829819158177366,
+      "grad_norm": 0.14542077447641327,
+      "learning_rate": 0.0008867456349817295,
+      "loss": 3.2381,
+      "step": 4695
+    },
+    {
+      "epoch": 1.4845613203822159,
+      "grad_norm": 0.1500640169422343,
+      "learning_rate": 0.0008863958767091289,
+      "loss": 3.2328,
+      "step": 4700
+    },
+    {
+      "epoch": 1.4861407249466951,
+      "grad_norm": 0.13869290140414445,
+      "learning_rate": 0.0008860456484014929,
+      "loss": 3.211,
+      "step": 4705
+    },
+    {
+      "epoch": 1.4877201295111742,
+      "grad_norm": 0.14829857540063193,
+      "learning_rate": 0.0008856949504848601,
+      "loss": 3.1714,
+      "step": 4710
+    },
+    {
+      "epoch": 1.4892995340756534,
+      "grad_norm": 0.16569262341577065,
+      "learning_rate": 0.0008853437833858404,
+      "loss": 3.2243,
+      "step": 4715
+    },
+    {
+      "epoch": 1.4908789386401327,
+      "grad_norm": 0.14695177504506926,
+      "learning_rate": 0.0008849921475316147,
+      "loss": 3.3036,
+      "step": 4720
+    },
+    {
+      "epoch": 1.492458343204612,
+      "grad_norm": 0.14586867054532876,
+      "learning_rate": 0.0008846400433499335,
+      "loss": 3.2349,
+      "step": 4725
+    },
+    {
+      "epoch": 1.4940377477690912,
+      "grad_norm": 0.2110658984856936,
+      "learning_rate": 0.0008842874712691175,
+      "loss": 3.2127,
+      "step": 4730
+    },
+    {
+      "epoch": 1.4956171523335702,
+      "grad_norm": 0.13573431744675368,
+      "learning_rate": 0.0008839344317180564,
+      "loss": 3.2067,
+      "step": 4735
+    },
+    {
+      "epoch": 1.4971965568980494,
+      "grad_norm": 0.14027398665568022,
+      "learning_rate": 0.0008835809251262091,
+      "loss": 3.2705,
+      "step": 4740
+    },
+    {
+      "epoch": 1.4987759614625287,
+      "grad_norm": 0.11286622399778137,
+      "learning_rate": 0.0008832269519236013,
+      "loss": 3.2677,
+      "step": 4745
+    },
+    {
+      "epoch": 1.5003553660270077,
+      "grad_norm": 0.1518608133814783,
+      "learning_rate": 0.0008828725125408276,
+      "loss": 3.2682,
+      "step": 4750
+    },
+    {
+      "epoch": 1.501934770591487,
+      "grad_norm": 0.160259811179564,
+      "learning_rate": 0.0008825176074090495,
+      "loss": 3.2156,
+      "step": 4755
+    },
+    {
+      "epoch": 1.5035141751559662,
+      "grad_norm": 0.18128847691858616,
+      "learning_rate": 0.0008821622369599944,
+      "loss": 3.2728,
+      "step": 4760
+    },
+    {
+      "epoch": 1.5050935797204454,
+      "grad_norm": 0.15979405833636123,
+      "learning_rate": 0.0008818064016259564,
+      "loss": 3.2132,
+      "step": 4765
+    },
+    {
+      "epoch": 1.5066729842849247,
+      "grad_norm": 0.16782521069305592,
+      "learning_rate": 0.0008814501018397947,
+      "loss": 3.1988,
+      "step": 4770
+    },
+    {
+      "epoch": 1.508252388849404,
+      "grad_norm": 0.16173887298085518,
+      "learning_rate": 0.0008810933380349337,
+      "loss": 3.2345,
+      "step": 4775
+    },
+    {
+      "epoch": 1.509831793413883,
+      "grad_norm": 0.1710658798394178,
+      "learning_rate": 0.0008807361106453622,
+      "loss": 3.2787,
+      "step": 4780
+    },
+    {
+      "epoch": 1.5114111979783622,
+      "grad_norm": 0.17277825281048725,
+      "learning_rate": 0.000880378420105633,
+      "loss": 3.1865,
+      "step": 4785
+    },
+    {
+      "epoch": 1.5129906025428412,
+      "grad_norm": 0.1818631892983199,
+      "learning_rate": 0.0008800202668508624,
+      "loss": 3.2306,
+      "step": 4790
+    },
+    {
+      "epoch": 1.5145700071073205,
+      "grad_norm": 0.17850343920387227,
+      "learning_rate": 0.0008796616513167291,
+      "loss": 3.118,
+      "step": 4795
+    },
+    {
+      "epoch": 1.5161494116717997,
+      "grad_norm": 0.17071356410691527,
+      "learning_rate": 0.0008793025739394745,
+      "loss": 3.1208,
+      "step": 4800
+    },
+    {
+      "epoch": 1.517728816236279,
+      "grad_norm": 0.15920996937791715,
+      "learning_rate": 0.000878943035155902,
+      "loss": 3.2293,
+      "step": 4805
+    },
+    {
+      "epoch": 1.5193082208007582,
+      "grad_norm": 0.15999249674057323,
+      "learning_rate": 0.0008785830354033759,
+      "loss": 3.189,
+      "step": 4810
+    },
+    {
+      "epoch": 1.5208876253652373,
+      "grad_norm": 0.12776982261731593,
+      "learning_rate": 0.0008782225751198216,
+      "loss": 3.2508,
+      "step": 4815
+    },
+    {
+      "epoch": 1.5224670299297165,
+      "grad_norm": 0.13987589557521737,
+      "learning_rate": 0.0008778616547437244,
+      "loss": 3.2344,
+      "step": 4820
+    },
+    {
+      "epoch": 1.5240464344941955,
+      "grad_norm": 0.13431219986869516,
+      "learning_rate": 0.0008775002747141292,
+      "loss": 3.2349,
+      "step": 4825
+    },
+    {
+      "epoch": 1.5256258390586748,
+      "grad_norm": 0.1192257573291684,
+      "learning_rate": 0.0008771384354706406,
+      "loss": 3.318,
+      "step": 4830
+    },
+    {
+      "epoch": 1.527205243623154,
+      "grad_norm": 0.13409814119711105,
+      "learning_rate": 0.0008767761374534215,
+      "loss": 3.1402,
+      "step": 4835
+    },
+    {
+      "epoch": 1.5287846481876333,
+      "grad_norm": 0.15609606836500248,
+      "learning_rate": 0.0008764133811031924,
+      "loss": 3.2244,
+      "step": 4840
+    },
+    {
+      "epoch": 1.5303640527521125,
+      "grad_norm": 0.18418778089852983,
+      "learning_rate": 0.0008760501668612324,
+      "loss": 3.1931,
+      "step": 4845
+    },
+    {
+      "epoch": 1.5319434573165918,
+      "grad_norm": 0.13804937377372667,
+      "learning_rate": 0.0008756864951693766,
+      "loss": 3.2765,
+      "step": 4850
+    },
+    {
+      "epoch": 1.5335228618810708,
+      "grad_norm": 0.12574847976657577,
+      "learning_rate": 0.0008753223664700171,
+      "loss": 3.2378,
+      "step": 4855
+    },
+    {
+      "epoch": 1.53510226644555,
+      "grad_norm": 0.1689148634919543,
+      "learning_rate": 0.0008749577812061019,
+      "loss": 3.3042,
+      "step": 4860
+    },
+    {
+      "epoch": 1.536681671010029,
+      "grad_norm": 0.2613685637963012,
+      "learning_rate": 0.0008745927398211339,
+      "loss": 3.2394,
+      "step": 4865
+    },
+    {
+      "epoch": 1.5382610755745083,
+      "grad_norm": 0.22204800684640813,
+      "learning_rate": 0.0008742272427591719,
+      "loss": 3.2851,
+      "step": 4870
+    },
+    {
+      "epoch": 1.5398404801389876,
+      "grad_norm": 0.17845446024685363,
+      "learning_rate": 0.0008738612904648279,
+      "loss": 3.2891,
+      "step": 4875
+    },
+    {
+      "epoch": 1.5414198847034668,
+      "grad_norm": 0.19052223542574034,
+      "learning_rate": 0.0008734948833832683,
+      "loss": 3.1479,
+      "step": 4880
+    },
+    {
+      "epoch": 1.542999289267946,
+      "grad_norm": 0.18442609056657755,
+      "learning_rate": 0.0008731280219602127,
+      "loss": 3.2407,
+      "step": 4885
+    },
+    {
+      "epoch": 1.5445786938324253,
+      "grad_norm": 0.16536340121177104,
+      "learning_rate": 0.000872760706641933,
+      "loss": 3.188,
+      "step": 4890
+    },
+    {
+      "epoch": 1.5461580983969043,
+      "grad_norm": 0.17738221997727127,
+      "learning_rate": 0.0008723929378752535,
+      "loss": 3.2681,
+      "step": 4895
+    },
+    {
+      "epoch": 1.5477375029613836,
+      "grad_norm": 0.14300527517916653,
+      "learning_rate": 0.0008720247161075503,
+      "loss": 3.2121,
+      "step": 4900
+    },
+    {
+      "epoch": 1.5493169075258626,
+      "grad_norm": 0.12252161093257756,
+      "learning_rate": 0.0008716560417867503,
+      "loss": 3.1576,
+      "step": 4905
+    },
+    {
+      "epoch": 1.5508963120903418,
+      "grad_norm": 0.13646202138242786,
+      "learning_rate": 0.000871286915361331,
+      "loss": 3.1993,
+      "step": 4910
+    },
+    {
+      "epoch": 1.552475716654821,
+      "grad_norm": 0.1450313805287437,
+      "learning_rate": 0.0008709173372803197,
+      "loss": 3.324,
+      "step": 4915
+    },
+    {
+      "epoch": 1.5540551212193003,
+      "grad_norm": 0.20542431408610692,
+      "learning_rate": 0.0008705473079932935,
+      "loss": 3.3023,
+      "step": 4920
+    },
+    {
+      "epoch": 1.5556345257837796,
+      "grad_norm": 0.1663832250588302,
+      "learning_rate": 0.0008701768279503779,
+      "loss": 3.2081,
+      "step": 4925
+    },
+    {
+      "epoch": 1.5572139303482588,
+      "grad_norm": 0.15828209904907997,
+      "learning_rate": 0.0008698058976022472,
+      "loss": 3.1474,
+      "step": 4930
+    },
+    {
+      "epoch": 1.5587933349127379,
+      "grad_norm": 0.13656502814277918,
+      "learning_rate": 0.0008694345174001228,
+      "loss": 3.2959,
+      "step": 4935
+    },
+    {
+      "epoch": 1.560372739477217,
+      "grad_norm": 0.16652595943649606,
+      "learning_rate": 0.0008690626877957743,
+      "loss": 3.2222,
+      "step": 4940
+    },
+    {
+      "epoch": 1.5619521440416961,
+      "grad_norm": 0.15536769823691185,
+      "learning_rate": 0.0008686904092415173,
+      "loss": 3.3182,
+      "step": 4945
+    },
+    {
+      "epoch": 1.5635315486061754,
+      "grad_norm": 0.12893633861479137,
+      "learning_rate": 0.0008683176821902135,
+      "loss": 3.1887,
+      "step": 4950
+    },
+    {
+      "epoch": 1.5651109531706546,
+      "grad_norm": 0.12321457114332846,
+      "learning_rate": 0.0008679445070952706,
+      "loss": 3.1422,
+      "step": 4955
+    },
+    {
+      "epoch": 1.5666903577351339,
+      "grad_norm": 0.1252717364208615,
+      "learning_rate": 0.0008675708844106407,
+      "loss": 3.2986,
+      "step": 4960
+    },
+    {
+      "epoch": 1.5682697622996131,
+      "grad_norm": 0.136179857864261,
+      "learning_rate": 0.0008671968145908211,
+      "loss": 3.1559,
+      "step": 4965
+    },
+    {
+      "epoch": 1.5698491668640924,
+      "grad_norm": 0.1366579774814646,
+      "learning_rate": 0.0008668222980908526,
+      "loss": 3.2452,
+      "step": 4970
+    },
+    {
+      "epoch": 1.5714285714285714,
+      "grad_norm": 0.15866997341356215,
+      "learning_rate": 0.0008664473353663196,
+      "loss": 3.1779,
+      "step": 4975
+    },
+    {
+      "epoch": 1.5730079759930506,
+      "grad_norm": 0.1570624780624178,
+      "learning_rate": 0.0008660719268733491,
+      "loss": 3.1761,
+      "step": 4980
+    },
+    {
+      "epoch": 1.5745873805575297,
+      "grad_norm": 0.16262475628462064,
+      "learning_rate": 0.0008656960730686101,
+      "loss": 3.2598,
+      "step": 4985
+    },
+    {
+      "epoch": 1.576166785122009,
+      "grad_norm": 0.16607449640434582,
+      "learning_rate": 0.0008653197744093139,
+      "loss": 3.2792,
+      "step": 4990
+    },
+    {
+      "epoch": 1.5777461896864882,
+      "grad_norm": 0.16636890691152242,
+      "learning_rate": 0.0008649430313532127,
+      "loss": 3.2285,
+      "step": 4995
+    },
+    {
+      "epoch": 1.5793255942509674,
+      "grad_norm": 0.162552103740573,
+      "learning_rate": 0.0008645658443585992,
+      "loss": 3.2509,
+      "step": 5000
+    },
+    {
+      "epoch": 1.5809049988154467,
+      "grad_norm": 0.19398036376248662,
+      "learning_rate": 0.0008641882138843062,
+      "loss": 3.1406,
+      "step": 5005
+    },
+    {
+      "epoch": 1.582484403379926,
+      "grad_norm": 0.1549550740040521,
+      "learning_rate": 0.0008638101403897061,
+      "loss": 3.22,
+      "step": 5010
+    },
+    {
+      "epoch": 1.584063807944405,
+      "grad_norm": 0.148872550051611,
+      "learning_rate": 0.0008634316243347099,
+      "loss": 3.2558,
+      "step": 5015
+    },
+    {
+      "epoch": 1.5856432125088842,
+      "grad_norm": 0.1304659991294797,
+      "learning_rate": 0.0008630526661797673,
+      "loss": 3.1085,
+      "step": 5020
+    },
+    {
+      "epoch": 1.5872226170733632,
+      "grad_norm": 0.1071200561590807,
+      "learning_rate": 0.0008626732663858655,
+      "loss": 3.2129,
+      "step": 5025
+    },
+    {
+      "epoch": 1.5888020216378425,
+      "grad_norm": 0.1321051614605033,
+      "learning_rate": 0.0008622934254145291,
+      "loss": 3.2189,
+      "step": 5030
+    },
+    {
+      "epoch": 1.5903814262023217,
+      "grad_norm": 0.15059770855107626,
+      "learning_rate": 0.0008619131437278196,
+      "loss": 3.2072,
+      "step": 5035
+    },
+    {
+      "epoch": 1.591960830766801,
+      "grad_norm": 0.15385580828759735,
+      "learning_rate": 0.0008615324217883341,
+      "loss": 3.1408,
+      "step": 5040
+    },
+    {
+      "epoch": 1.5935402353312802,
+      "grad_norm": 0.18895304159790371,
+      "learning_rate": 0.0008611512600592057,
+      "loss": 3.2543,
+      "step": 5045
+    },
+    {
+      "epoch": 1.5951196398957594,
+      "grad_norm": 0.14833800477922005,
+      "learning_rate": 0.0008607696590041021,
+      "loss": 3.2304,
+      "step": 5050
+    },
+    {
+      "epoch": 1.5966990444602385,
+      "grad_norm": 0.1829558575858106,
+      "learning_rate": 0.0008603876190872257,
+      "loss": 3.3536,
+      "step": 5055
+    },
+    {
+      "epoch": 1.5982784490247177,
+      "grad_norm": 0.13784885033258049,
+      "learning_rate": 0.000860005140773313,
+      "loss": 3.0908,
+      "step": 5060
+    },
+    {
+      "epoch": 1.5998578535891967,
+      "grad_norm": 0.13757531043144822,
+      "learning_rate": 0.0008596222245276329,
+      "loss": 3.1926,
+      "step": 5065
+    },
+    {
+      "epoch": 1.601437258153676,
+      "grad_norm": 0.15982659305698876,
+      "learning_rate": 0.000859238870815988,
+      "loss": 3.1252,
+      "step": 5070
+    },
+    {
+      "epoch": 1.6030166627181552,
+      "grad_norm": 0.15781292432563,
+      "learning_rate": 0.0008588550801047127,
+      "loss": 3.2267,
+      "step": 5075
+    },
+    {
+      "epoch": 1.6045960672826345,
+      "grad_norm": 0.13440615536611872,
+      "learning_rate": 0.0008584708528606728,
+      "loss": 3.1973,
+      "step": 5080
+    },
+    {
+      "epoch": 1.6061754718471137,
+      "grad_norm": 0.13180202622515164,
+      "learning_rate": 0.0008580861895512652,
+      "loss": 3.2315,
+      "step": 5085
+    },
+    {
+      "epoch": 1.607754876411593,
+      "grad_norm": 0.13594821623804945,
+      "learning_rate": 0.0008577010906444174,
+      "loss": 3.1541,
+      "step": 5090
+    },
+    {
+      "epoch": 1.609334280976072,
+      "grad_norm": 0.14242110458172097,
+      "learning_rate": 0.0008573155566085868,
+      "loss": 3.19,
+      "step": 5095
+    },
+    {
+      "epoch": 1.6109136855405513,
+      "grad_norm": 0.1402901958946774,
+      "learning_rate": 0.0008569295879127602,
+      "loss": 3.2011,
+      "step": 5100
+    },
+    {
+      "epoch": 1.6124930901050303,
+      "grad_norm": 0.1345845928466279,
+      "learning_rate": 0.0008565431850264527,
+      "loss": 3.153,
+      "step": 5105
+    },
+    {
+      "epoch": 1.6140724946695095,
+      "grad_norm": 0.14299194878908414,
+      "learning_rate": 0.0008561563484197079,
+      "loss": 3.1064,
+      "step": 5110
+    },
+    {
+      "epoch": 1.6156518992339888,
+      "grad_norm": 0.15337008180865508,
+      "learning_rate": 0.000855769078563097,
+      "loss": 3.2301,
+      "step": 5115
+    },
+    {
+      "epoch": 1.617231303798468,
+      "grad_norm": 0.14761041933323335,
+      "learning_rate": 0.0008553813759277184,
+      "loss": 3.303,
+      "step": 5120
+    },
+    {
+      "epoch": 1.6188107083629473,
+      "grad_norm": 0.16842837456466211,
+      "learning_rate": 0.0008549932409851965,
+      "loss": 3.0947,
+      "step": 5125
+    },
+    {
+      "epoch": 1.6203901129274265,
+      "grad_norm": 0.13929666990210726,
+      "learning_rate": 0.0008546046742076819,
+      "loss": 3.2187,
+      "step": 5130
+    },
+    {
+      "epoch": 1.6219695174919055,
+      "grad_norm": 0.14019470888113164,
+      "learning_rate": 0.0008542156760678504,
+      "loss": 3.2503,
+      "step": 5135
+    },
+    {
+      "epoch": 1.6235489220563848,
+      "grad_norm": 0.18987242622039088,
+      "learning_rate": 0.0008538262470389028,
+      "loss": 3.2645,
+      "step": 5140
+    },
+    {
+      "epoch": 1.6251283266208638,
+      "grad_norm": 0.16193318942084498,
+      "learning_rate": 0.0008534363875945637,
+      "loss": 3.1828,
+      "step": 5145
+    },
+    {
+      "epoch": 1.626707731185343,
+      "grad_norm": 0.1619889439692038,
+      "learning_rate": 0.0008530460982090812,
+      "loss": 3.2047,
+      "step": 5150
+    },
+    {
+      "epoch": 1.6282871357498223,
+      "grad_norm": 0.15342965563746738,
+      "learning_rate": 0.000852655379357227,
+      "loss": 3.2319,
+      "step": 5155
+    },
+    {
+      "epoch": 1.6298665403143016,
+      "grad_norm": 0.15132360226394195,
+      "learning_rate": 0.0008522642315142948,
+      "loss": 3.2558,
+      "step": 5160
+    },
+    {
+      "epoch": 1.6314459448787808,
+      "grad_norm": 0.18282985704516308,
+      "learning_rate": 0.0008518726551560999,
+      "loss": 3.189,
+      "step": 5165
+    },
+    {
+      "epoch": 1.63302534944326,
+      "grad_norm": 0.15024580263521556,
+      "learning_rate": 0.0008514806507589796,
+      "loss": 3.1789,
+      "step": 5170
+    },
+    {
+      "epoch": 1.634604754007739,
+      "grad_norm": 0.1669828597076097,
+      "learning_rate": 0.0008510882187997913,
+      "loss": 3.204,
+      "step": 5175
+    },
+    {
+      "epoch": 1.6361841585722183,
+      "grad_norm": 0.12997826381256986,
+      "learning_rate": 0.0008506953597559124,
+      "loss": 3.1704,
+      "step": 5180
+    },
+    {
+      "epoch": 1.6377635631366974,
+      "grad_norm": 0.19046220176512926,
+      "learning_rate": 0.0008503020741052407,
+      "loss": 3.1976,
+      "step": 5185
+    },
+    {
+      "epoch": 1.6393429677011766,
+      "grad_norm": 0.16953626947212094,
+      "learning_rate": 0.0008499083623261919,
+      "loss": 3.1796,
+      "step": 5190
+    },
+    {
+      "epoch": 1.6409223722656558,
+      "grad_norm": 0.15870142335511733,
+      "learning_rate": 0.0008495142248977007,
+      "loss": 3.1347,
+      "step": 5195
+    },
+    {
+      "epoch": 1.642501776830135,
+      "grad_norm": 0.12085159283664512,
+      "learning_rate": 0.0008491196622992194,
+      "loss": 3.1646,
+      "step": 5200
+    },
+    {
+      "epoch": 1.6440811813946143,
+      "grad_norm": 0.1356652151636163,
+      "learning_rate": 0.0008487246750107176,
+      "loss": 3.0949,
+      "step": 5205
+    },
+    {
+      "epoch": 1.6456605859590934,
+      "grad_norm": 0.18844483345578247,
+      "learning_rate": 0.0008483292635126814,
+      "loss": 3.2829,
+      "step": 5210
+    },
+    {
+      "epoch": 1.6472399905235726,
+      "grad_norm": 0.1435244567839457,
+      "learning_rate": 0.0008479334282861129,
+      "loss": 3.1473,
+      "step": 5215
+    },
+    {
+      "epoch": 1.6488193950880516,
+      "grad_norm": 0.18232492906003894,
+      "learning_rate": 0.0008475371698125297,
+      "loss": 3.1508,
+      "step": 5220
+    },
+    {
+      "epoch": 1.650398799652531,
+      "grad_norm": 0.13659700220175816,
+      "learning_rate": 0.0008471404885739644,
+      "loss": 3.0982,
+      "step": 5225
+    },
+    {
+      "epoch": 1.6519782042170101,
+      "grad_norm": 0.15726236702510096,
+      "learning_rate": 0.0008467433850529639,
+      "loss": 3.1438,
+      "step": 5230
+    },
+    {
+      "epoch": 1.6535576087814894,
+      "grad_norm": 0.16050432120825808,
+      "learning_rate": 0.0008463458597325884,
+      "loss": 3.2055,
+      "step": 5235
+    },
+    {
+      "epoch": 1.6551370133459686,
+      "grad_norm": 0.18108242926877352,
+      "learning_rate": 0.0008459479130964114,
+      "loss": 3.1179,
+      "step": 5240
+    },
+    {
+      "epoch": 1.6567164179104479,
+      "grad_norm": 0.13495676091229308,
+      "learning_rate": 0.0008455495456285193,
+      "loss": 3.1713,
+      "step": 5245
+    },
+    {
+      "epoch": 1.658295822474927,
+      "grad_norm": 0.16137588648755863,
+      "learning_rate": 0.0008451507578135098,
+      "loss": 3.0861,
+      "step": 5250
+    },
+    {
+      "epoch": 1.6598752270394062,
+      "grad_norm": 0.14343876409963272,
+      "learning_rate": 0.0008447515501364924,
+      "loss": 3.1167,
+      "step": 5255
+    },
+    {
+      "epoch": 1.6614546316038852,
+      "grad_norm": 0.13626674823775772,
+      "learning_rate": 0.0008443519230830871,
+      "loss": 3.1368,
+      "step": 5260
+    },
+    {
+      "epoch": 1.6630340361683644,
+      "grad_norm": 0.11641914134487123,
+      "learning_rate": 0.0008439518771394241,
+      "loss": 3.0619,
+      "step": 5265
+    },
+    {
+      "epoch": 1.6646134407328437,
+      "grad_norm": 0.1189405426468082,
+      "learning_rate": 0.0008435514127921431,
+      "loss": 3.1442,
+      "step": 5270
+    },
+    {
+      "epoch": 1.666192845297323,
+      "grad_norm": 0.1225580458586459,
+      "learning_rate": 0.0008431505305283933,
+      "loss": 3.1163,
+      "step": 5275
+    },
+    {
+      "epoch": 1.6677722498618022,
+      "grad_norm": 0.12802593143458157,
+      "learning_rate": 0.0008427492308358313,
+      "loss": 3.1695,
+      "step": 5280
+    },
+    {
+      "epoch": 1.6693516544262814,
+      "grad_norm": 0.18123050394278015,
+      "learning_rate": 0.0008423475142026223,
+      "loss": 3.0716,
+      "step": 5285
+    },
+    {
+      "epoch": 1.6709310589907604,
+      "grad_norm": 0.13790304915784055,
+      "learning_rate": 0.0008419453811174385,
+      "loss": 3.123,
+      "step": 5290
+    },
+    {
+      "epoch": 1.6725104635552397,
+      "grad_norm": 0.15913703714357955,
+      "learning_rate": 0.0008415428320694584,
+      "loss": 3.0708,
+      "step": 5295
+    },
+    {
+      "epoch": 1.6740898681197187,
+      "grad_norm": 0.16174700715573553,
+      "learning_rate": 0.0008411398675483668,
+      "loss": 3.0799,
+      "step": 5300
+    },
+    {
+      "epoch": 1.675669272684198,
+      "grad_norm": 0.17404012367261848,
+      "learning_rate": 0.0008407364880443539,
+      "loss": 3.1949,
+      "step": 5305
+    },
+    {
+      "epoch": 1.6772486772486772,
+      "grad_norm": 0.20852493040494388,
+      "learning_rate": 0.0008403326940481146,
+      "loss": 3.257,
+      "step": 5310
+    },
+    {
+      "epoch": 1.6788280818131565,
+      "grad_norm": 0.16777003267594165,
+      "learning_rate": 0.000839928486050848,
+      "loss": 3.2414,
+      "step": 5315
+    },
+    {
+      "epoch": 1.6804074863776357,
+      "grad_norm": 0.12454972106798985,
+      "learning_rate": 0.0008395238645442569,
+      "loss": 3.1313,
+      "step": 5320
+    },
+    {
+      "epoch": 1.681986890942115,
+      "grad_norm": 0.1330984558987137,
+      "learning_rate": 0.000839118830020547,
+      "loss": 3.145,
+      "step": 5325
+    },
+    {
+      "epoch": 1.683566295506594,
+      "grad_norm": 0.3087417482749041,
+      "learning_rate": 0.0008387133829724266,
+      "loss": 3.1466,
+      "step": 5330
+    },
+    {
+      "epoch": 1.6851457000710732,
+      "grad_norm": 0.1763043327279286,
+      "learning_rate": 0.0008383075238931057,
+      "loss": 3.1494,
+      "step": 5335
+    },
+    {
+      "epoch": 1.6867251046355523,
+      "grad_norm": 0.16430100545063897,
+      "learning_rate": 0.0008379012532762955,
+      "loss": 3.1457,
+      "step": 5340
+    },
+    {
+      "epoch": 1.6883045092000315,
+      "grad_norm": 0.13424044681308792,
+      "learning_rate": 0.0008374945716162079,
+      "loss": 3.1974,
+      "step": 5345
+    },
+    {
+      "epoch": 1.6898839137645107,
+      "grad_norm": 0.132656566575011,
+      "learning_rate": 0.0008370874794075548,
+      "loss": 3.1854,
+      "step": 5350
+    },
+    {
+      "epoch": 1.69146331832899,
+      "grad_norm": 0.1046842574599167,
+      "learning_rate": 0.0008366799771455474,
+      "loss": 3.0958,
+      "step": 5355
+    },
+    {
+      "epoch": 1.6930427228934692,
+      "grad_norm": 0.14992551766326231,
+      "learning_rate": 0.0008362720653258959,
+      "loss": 3.1479,
+      "step": 5360
+    },
+    {
+      "epoch": 1.6946221274579485,
+      "grad_norm": 0.13712636288570967,
+      "learning_rate": 0.0008358637444448085,
+      "loss": 3.1558,
+      "step": 5365
+    },
+    {
+      "epoch": 1.6962015320224275,
+      "grad_norm": 0.10402371936092365,
+      "learning_rate": 0.0008354550149989912,
+      "loss": 3.089,
+      "step": 5370
+    },
+    {
+      "epoch": 1.6977809365869068,
+      "grad_norm": 0.11701920328653621,
+      "learning_rate": 0.0008350458774856469,
+      "loss": 3.0974,
+      "step": 5375
+    },
+    {
+      "epoch": 1.6993603411513858,
+      "grad_norm": 0.13248230505359318,
+      "learning_rate": 0.000834636332402475,
+      "loss": 3.1612,
+      "step": 5380
+    },
+    {
+      "epoch": 1.700939745715865,
+      "grad_norm": 0.12477491619072674,
+      "learning_rate": 0.0008342263802476706,
+      "loss": 3.075,
+      "step": 5385
+    },
+    {
+      "epoch": 1.7025191502803443,
+      "grad_norm": 0.12147037604704496,
+      "learning_rate": 0.0008338160215199239,
+      "loss": 3.0468,
+      "step": 5390
+    },
+    {
+      "epoch": 1.7040985548448235,
+      "grad_norm": 0.13270474543121774,
+      "learning_rate": 0.0008334052567184198,
+      "loss": 3.1062,
+      "step": 5395
+    },
+    {
+      "epoch": 1.7056779594093028,
+      "grad_norm": 0.10998238774886308,
+      "learning_rate": 0.0008329940863428372,
+      "loss": 3.1386,
+      "step": 5400
+    },
+    {
+      "epoch": 1.707257363973782,
+      "grad_norm": 0.1515915492300537,
+      "learning_rate": 0.0008325825108933481,
+      "loss": 3.0909,
+      "step": 5405
+    },
+    {
+      "epoch": 1.708836768538261,
+      "grad_norm": 0.15837330357706372,
+      "learning_rate": 0.0008321705308706178,
+      "loss": 3.1,
+      "step": 5410
+    },
+    {
+      "epoch": 1.7104161731027403,
+      "grad_norm": 0.15132099851686953,
+      "learning_rate": 0.0008317581467758033,
+      "loss": 3.2848,
+      "step": 5415
+    },
+    {
+      "epoch": 1.7119955776672193,
+      "grad_norm": 0.12926189843665217,
+      "learning_rate": 0.0008313453591105533,
+      "loss": 3.0679,
+      "step": 5420
+    },
+    {
+      "epoch": 1.7135749822316986,
+      "grad_norm": 0.10577937715798899,
+      "learning_rate": 0.0008309321683770073,
+      "loss": 3.0536,
+      "step": 5425
+    },
+    {
+      "epoch": 1.7151543867961778,
+      "grad_norm": 0.12605393863349898,
+      "learning_rate": 0.0008305185750777952,
+      "loss": 3.0875,
+      "step": 5430
+    },
+    {
+      "epoch": 1.716733791360657,
+      "grad_norm": 0.14760806225530054,
+      "learning_rate": 0.0008301045797160365,
+      "loss": 3.1276,
+      "step": 5435
+    },
+    {
+      "epoch": 1.7183131959251363,
+      "grad_norm": 0.12269117451194803,
+      "learning_rate": 0.0008296901827953403,
+      "loss": 3.0501,
+      "step": 5440
+    },
+    {
+      "epoch": 1.7198926004896156,
+      "grad_norm": 0.22148513141090342,
+      "learning_rate": 0.0008292753848198034,
+      "loss": 3.0366,
+      "step": 5445
+    },
+    {
+      "epoch": 1.7214720050540946,
+      "grad_norm": 0.17485051885453007,
+      "learning_rate": 0.0008288601862940109,
+      "loss": 3.1606,
+      "step": 5450
+    },
+    {
+      "epoch": 1.7230514096185738,
+      "grad_norm": 0.16453403604485145,
+      "learning_rate": 0.0008284445877230351,
+      "loss": 3.2031,
+      "step": 5455
+    },
+    {
+      "epoch": 1.7246308141830529,
+      "grad_norm": 0.11388373614963429,
+      "learning_rate": 0.000828028589612435,
+      "loss": 3.0225,
+      "step": 5460
+    },
+    {
+      "epoch": 1.726210218747532,
+      "grad_norm": 0.15268472172554937,
+      "learning_rate": 0.0008276121924682556,
+      "loss": 3.1488,
+      "step": 5465
+    },
+    {
+      "epoch": 1.7277896233120114,
+      "grad_norm": 0.16123097535937514,
+      "learning_rate": 0.0008271953967970273,
+      "loss": 3.0955,
+      "step": 5470
+    },
+    {
+      "epoch": 1.7293690278764906,
+      "grad_norm": 0.13991029115174697,
+      "learning_rate": 0.0008267782031057651,
+      "loss": 3.0341,
+      "step": 5475
+    },
+    {
+      "epoch": 1.7309484324409699,
+      "grad_norm": 0.12181705811040357,
+      "learning_rate": 0.0008263606119019684,
+      "loss": 3.0104,
+      "step": 5480
+    },
+    {
+      "epoch": 1.732527837005449,
+      "grad_norm": 0.14873853455315397,
+      "learning_rate": 0.0008259426236936203,
+      "loss": 3.125,
+      "step": 5485
+    },
+    {
+      "epoch": 1.7341072415699281,
+      "grad_norm": 0.1491259625850415,
+      "learning_rate": 0.0008255242389891862,
+      "loss": 3.2172,
+      "step": 5490
+    },
+    {
+      "epoch": 1.7356866461344074,
+      "grad_norm": 0.14895531183393362,
+      "learning_rate": 0.0008251054582976146,
+      "loss": 3.1077,
+      "step": 5495
+    },
+    {
+      "epoch": 1.7372660506988864,
+      "grad_norm": 0.12030961427581022,
+      "learning_rate": 0.0008246862821283353,
+      "loss": 3.0649,
+      "step": 5500
+    },
+    {
+      "epoch": 1.7388454552633656,
+      "grad_norm": 0.15564231945017964,
+      "learning_rate": 0.0008242667109912592,
+      "loss": 3.1639,
+      "step": 5505
+    },
+    {
+      "epoch": 1.740424859827845,
+      "grad_norm": 0.1304107603769251,
+      "learning_rate": 0.0008238467453967778,
+      "loss": 3.1414,
+      "step": 5510
+    },
+    {
+      "epoch": 1.7420042643923241,
+      "grad_norm": 0.13419680595114025,
+      "learning_rate": 0.0008234263858557621,
+      "loss": 3.0805,
+      "step": 5515
+    },
+    {
+      "epoch": 1.7435836689568034,
+      "grad_norm": 0.1298847058824958,
+      "learning_rate": 0.0008230056328795629,
+      "loss": 3.0763,
+      "step": 5520
+    },
+    {
+      "epoch": 1.7451630735212826,
+      "grad_norm": 0.15677569802750654,
+      "learning_rate": 0.000822584486980009,
+      "loss": 3.0403,
+      "step": 5525
+    },
+    {
+      "epoch": 1.7467424780857617,
+      "grad_norm": 0.14205010542182986,
+      "learning_rate": 0.0008221629486694075,
+      "loss": 3.0167,
+      "step": 5530
+    },
+    {
+      "epoch": 1.748321882650241,
+      "grad_norm": 0.11323305619202069,
+      "learning_rate": 0.000821741018460543,
+      "loss": 3.0397,
+      "step": 5535
+    },
+    {
+      "epoch": 1.74990128721472,
+      "grad_norm": 0.12933460841485173,
+      "learning_rate": 0.0008213186968666762,
+      "loss": 3.0373,
+      "step": 5540
+    },
+    {
+      "epoch": 1.7514806917791992,
+      "grad_norm": 0.14348811976289397,
+      "learning_rate": 0.0008208959844015446,
+      "loss": 3.0593,
+      "step": 5545
+    },
+    {
+      "epoch": 1.7530600963436784,
+      "grad_norm": 0.13721521690367944,
+      "learning_rate": 0.000820472881579361,
+      "loss": 3.0771,
+      "step": 5550
+    },
+    {
+      "epoch": 1.7546395009081577,
+      "grad_norm": 0.1113321743985448,
+      "learning_rate": 0.0008200493889148129,
+      "loss": 3.0933,
+      "step": 5555
+    },
+    {
+      "epoch": 1.756218905472637,
+      "grad_norm": 0.13039510447204097,
+      "learning_rate": 0.0008196255069230618,
+      "loss": 3.1005,
+      "step": 5560
+    },
+    {
+      "epoch": 1.757798310037116,
+      "grad_norm": 0.1227027362155738,
+      "learning_rate": 0.0008192012361197434,
+      "loss": 3.0374,
+      "step": 5565
+    },
+    {
+      "epoch": 1.7593777146015952,
+      "grad_norm": 0.16124398321310454,
+      "learning_rate": 0.0008187765770209661,
+      "loss": 3.1171,
+      "step": 5570
+    },
+    {
+      "epoch": 1.7609571191660742,
+      "grad_norm": 0.13815148766215551,
+      "learning_rate": 0.0008183515301433104,
+      "loss": 3.1015,
+      "step": 5575
+    },
+    {
+      "epoch": 1.7625365237305535,
+      "grad_norm": 0.12576873501411856,
+      "learning_rate": 0.0008179260960038287,
+      "loss": 3.0402,
+      "step": 5580
+    },
+    {
+      "epoch": 1.7641159282950327,
+      "grad_norm": 0.15638893813279928,
+      "learning_rate": 0.0008175002751200447,
+      "loss": 3.1039,
+      "step": 5585
+    },
+    {
+      "epoch": 1.765695332859512,
+      "grad_norm": 0.18607696690530798,
+      "learning_rate": 0.0008170740680099519,
+      "loss": 3.1122,
+      "step": 5590
+    },
+    {
+      "epoch": 1.7672747374239912,
+      "grad_norm": 0.18841383981828605,
+      "learning_rate": 0.000816647475192015,
+      "loss": 3.0368,
+      "step": 5595
+    },
+    {
+      "epoch": 1.7688541419884705,
+      "grad_norm": 0.14018054519543763,
+      "learning_rate": 0.0008162204971851662,
+      "loss": 3.1069,
+      "step": 5600
+    },
+    {
+      "epoch": 1.7704335465529495,
+      "grad_norm": 0.1281256845789643,
+      "learning_rate": 0.0008157931345088074,
+      "loss": 3.1347,
+      "step": 5605
+    },
+    {
+      "epoch": 1.7720129511174287,
+      "grad_norm": 0.16806814294310318,
+      "learning_rate": 0.000815365387682808,
+      "loss": 3.0791,
+      "step": 5610
+    },
+    {
+      "epoch": 1.7735923556819078,
+      "grad_norm": 0.17461079002782293,
+      "learning_rate": 0.0008149372572275049,
+      "loss": 3.0854,
+      "step": 5615
+    },
+    {
+      "epoch": 1.775171760246387,
+      "grad_norm": 0.14352450743252834,
+      "learning_rate": 0.0008145087436637013,
+      "loss": 3.0062,
+      "step": 5620
+    },
+    {
+      "epoch": 1.7767511648108663,
+      "grad_norm": 0.1467818298017096,
+      "learning_rate": 0.0008140798475126671,
+      "loss": 3.0404,
+      "step": 5625
+    },
+    {
+      "epoch": 1.7783305693753455,
+      "grad_norm": 0.12627519312560972,
+      "learning_rate": 0.000813650569296137,
+      "loss": 3.0417,
+      "step": 5630
+    },
+    {
+      "epoch": 1.7799099739398248,
+      "grad_norm": 0.12415805779148288,
+      "learning_rate": 0.0008132209095363107,
+      "loss": 3.1058,
+      "step": 5635
+    },
+    {
+      "epoch": 1.781489378504304,
+      "grad_norm": 0.12978630067431146,
+      "learning_rate": 0.000812790868755852,
+      "loss": 3.0556,
+      "step": 5640
+    },
+    {
+      "epoch": 1.783068783068783,
+      "grad_norm": 0.12464317543260613,
+      "learning_rate": 0.0008123604474778881,
+      "loss": 3.0791,
+      "step": 5645
+    },
+    {
+      "epoch": 1.7846481876332623,
+      "grad_norm": 0.12344509884685934,
+      "learning_rate": 0.0008119296462260093,
+      "loss": 3.1187,
+      "step": 5650
+    },
+    {
+      "epoch": 1.7862275921977413,
+      "grad_norm": 0.11867693121313405,
+      "learning_rate": 0.0008114984655242681,
+      "loss": 3.0343,
+      "step": 5655
+    },
+    {
+      "epoch": 1.7878069967622205,
+      "grad_norm": 0.10448685844422977,
+      "learning_rate": 0.0008110669058971783,
+      "loss": 3.0101,
+      "step": 5660
+    },
+    {
+      "epoch": 1.7893864013266998,
+      "grad_norm": 0.13700981456911288,
+      "learning_rate": 0.0008106349678697147,
+      "loss": 3.1185,
+      "step": 5665
+    },
+    {
+      "epoch": 1.790965805891179,
+      "grad_norm": 0.1424263583015784,
+      "learning_rate": 0.0008102026519673127,
+      "loss": 3.0573,
+      "step": 5670
+    },
+    {
+      "epoch": 1.7925452104556583,
+      "grad_norm": 0.13955412711666954,
+      "learning_rate": 0.0008097699587158673,
+      "loss": 3.0688,
+      "step": 5675
+    },
+    {
+      "epoch": 1.7941246150201375,
+      "grad_norm": 0.14063694471408714,
+      "learning_rate": 0.0008093368886417323,
+      "loss": 3.1464,
+      "step": 5680
+    },
+    {
+      "epoch": 1.7957040195846166,
+      "grad_norm": 0.14664202038812976,
+      "learning_rate": 0.0008089034422717199,
+      "loss": 3.0739,
+      "step": 5685
+    },
+    {
+      "epoch": 1.7972834241490958,
+      "grad_norm": 0.2132974645825116,
+      "learning_rate": 0.0008084696201331004,
+      "loss": 3.1642,
+      "step": 5690
+    },
+    {
+      "epoch": 1.7988628287135748,
+      "grad_norm": 0.1616522602460956,
+      "learning_rate": 0.0008080354227536008,
+      "loss": 3.1808,
+      "step": 5695
+    },
+    {
+      "epoch": 1.800442233278054,
+      "grad_norm": 0.15325109276749146,
+      "learning_rate": 0.000807600850661405,
+      "loss": 3.1289,
+      "step": 5700
+    },
+    {
+      "epoch": 1.8020216378425333,
+      "grad_norm": 0.10407676475142526,
+      "learning_rate": 0.000807165904385152,
+      "loss": 3.1228,
+      "step": 5705
+    },
+    {
+      "epoch": 1.8036010424070126,
+      "grad_norm": 0.13943636056233583,
+      "learning_rate": 0.0008067305844539369,
+      "loss": 3.0581,
+      "step": 5710
+    },
+    {
+      "epoch": 1.8051804469714918,
+      "grad_norm": 0.13380630428432708,
+      "learning_rate": 0.0008062948913973087,
+      "loss": 3.0714,
+      "step": 5715
+    },
+    {
+      "epoch": 1.806759851535971,
+      "grad_norm": 0.14289546179063847,
+      "learning_rate": 0.0008058588257452703,
+      "loss": 3.1077,
+      "step": 5720
+    },
+    {
+      "epoch": 1.80833925610045,
+      "grad_norm": 0.1576835149035532,
+      "learning_rate": 0.0008054223880282783,
+      "loss": 3.0461,
+      "step": 5725
+    },
+    {
+      "epoch": 1.8099186606649293,
+      "grad_norm": 0.13290375552698708,
+      "learning_rate": 0.0008049855787772416,
+      "loss": 3.0308,
+      "step": 5730
+    },
+    {
+      "epoch": 1.8114980652294084,
+      "grad_norm": 0.11451184645959576,
+      "learning_rate": 0.0008045483985235207,
+      "loss": 3.1401,
+      "step": 5735
+    },
+    {
+      "epoch": 1.8130774697938876,
+      "grad_norm": 0.13278889627957352,
+      "learning_rate": 0.0008041108477989283,
+      "loss": 3.0229,
+      "step": 5740
+    },
+    {
+      "epoch": 1.8146568743583669,
+      "grad_norm": 0.09803195897353165,
+      "learning_rate": 0.0008036729271357269,
+      "loss": 3.0012,
+      "step": 5745
+    },
+    {
+      "epoch": 1.8162362789228461,
+      "grad_norm": 0.20567040452045257,
+      "learning_rate": 0.0008032346370666296,
+      "loss": 3.137,
+      "step": 5750
+    },
+    {
+      "epoch": 1.8178156834873254,
+      "grad_norm": 0.1337600870387908,
+      "learning_rate": 0.0008027959781247984,
+      "loss": 3.1482,
+      "step": 5755
+    },
+    {
+      "epoch": 1.8193950880518046,
+      "grad_norm": 0.16189524285580692,
+      "learning_rate": 0.0008023569508438444,
+      "loss": 3.0882,
+      "step": 5760
+    },
+    {
+      "epoch": 1.8209744926162836,
+      "grad_norm": 0.1358494981547762,
+      "learning_rate": 0.0008019175557578267,
+      "loss": 3.1672,
+      "step": 5765
+    },
+    {
+      "epoch": 1.8225538971807629,
+      "grad_norm": 0.09775298145324521,
+      "learning_rate": 0.0008014777934012514,
+      "loss": 3.0299,
+      "step": 5770
+    },
+    {
+      "epoch": 1.824133301745242,
+      "grad_norm": 0.1512229736528465,
+      "learning_rate": 0.0008010376643090719,
+      "loss": 3.0377,
+      "step": 5775
+    },
+    {
+      "epoch": 1.8257127063097212,
+      "grad_norm": 0.1728295251761587,
+      "learning_rate": 0.0008005971690166879,
+      "loss": 3.1011,
+      "step": 5780
+    },
+    {
+      "epoch": 1.8272921108742004,
+      "grad_norm": 0.1298951885600335,
+      "learning_rate": 0.0008001563080599437,
+      "loss": 2.9859,
+      "step": 5785
+    },
+    {
+      "epoch": 1.8288715154386797,
+      "grad_norm": 0.1209864599778657,
+      "learning_rate": 0.0007997150819751289,
+      "loss": 3.0874,
+      "step": 5790
+    },
+    {
+      "epoch": 1.830450920003159,
+      "grad_norm": 0.11706317164181328,
+      "learning_rate": 0.0007992734912989776,
+      "loss": 3.0135,
+      "step": 5795
+    },
+    {
+      "epoch": 1.8320303245676381,
+      "grad_norm": 0.16201933525966714,
+      "learning_rate": 0.0007988315365686671,
+      "loss": 3.1384,
+      "step": 5800
+    },
+    {
+      "epoch": 1.8336097291321172,
+      "grad_norm": 0.14368527580876594,
+      "learning_rate": 0.0007983892183218173,
+      "loss": 3.1312,
+      "step": 5805
+    },
+    {
+      "epoch": 1.8351891336965964,
+      "grad_norm": 0.15725717882474435,
+      "learning_rate": 0.0007979465370964904,
+      "loss": 3.0372,
+      "step": 5810
+    },
+    {
+      "epoch": 1.8367685382610754,
+      "grad_norm": 0.11697847046530074,
+      "learning_rate": 0.0007975034934311907,
+      "loss": 3.0479,
+      "step": 5815
+    },
+    {
+      "epoch": 1.8383479428255547,
+      "grad_norm": 0.15221307169759646,
+      "learning_rate": 0.000797060087864863,
+      "loss": 3.1091,
+      "step": 5820
+    },
+    {
+      "epoch": 1.839927347390034,
+      "grad_norm": 0.13501337394145765,
+      "learning_rate": 0.0007966163209368919,
+      "loss": 3.0373,
+      "step": 5825
+    },
+    {
+      "epoch": 1.8415067519545132,
+      "grad_norm": 0.1121320850988699,
+      "learning_rate": 0.0007961721931871023,
+      "loss": 3.05,
+      "step": 5830
+    },
+    {
+      "epoch": 1.8430861565189924,
+      "grad_norm": 0.14611915116235857,
+      "learning_rate": 0.0007957277051557577,
+      "loss": 3.035,
+      "step": 5835
+    },
+    {
+      "epoch": 1.8446655610834717,
+      "grad_norm": 0.1219728366948523,
+      "learning_rate": 0.00079528285738356,
+      "loss": 3.036,
+      "step": 5840
+    },
+    {
+      "epoch": 1.8462449656479507,
+      "grad_norm": 0.13072936964709259,
+      "learning_rate": 0.0007948376504116485,
+      "loss": 2.98,
+      "step": 5845
+    },
+    {
+      "epoch": 1.84782437021243,
+      "grad_norm": 0.16005339225680779,
+      "learning_rate": 0.0007943920847815995,
+      "loss": 3.1241,
+      "step": 5850
+    },
+    {
+      "epoch": 1.849403774776909,
+      "grad_norm": 0.11945810348155937,
+      "learning_rate": 0.0007939461610354258,
+      "loss": 2.9956,
+      "step": 5855
+    },
+    {
+      "epoch": 1.8509831793413882,
+      "grad_norm": 0.1697625468945632,
+      "learning_rate": 0.0007934998797155756,
+      "loss": 3.0508,
+      "step": 5860
+    },
+    {
+      "epoch": 1.8525625839058675,
+      "grad_norm": 0.13930841411775904,
+      "learning_rate": 0.0007930532413649323,
+      "loss": 3.0427,
+      "step": 5865
+    },
+    {
+      "epoch": 1.8541419884703467,
+      "grad_norm": 0.13016541216083566,
+      "learning_rate": 0.0007926062465268133,
+      "loss": 3.0849,
+      "step": 5870
+    },
+    {
+      "epoch": 1.855721393034826,
+      "grad_norm": 0.1485512809269021,
+      "learning_rate": 0.0007921588957449699,
+      "loss": 3.1645,
+      "step": 5875
+    },
+    {
+      "epoch": 1.8573007975993052,
+      "grad_norm": 0.11193581480631004,
+      "learning_rate": 0.0007917111895635864,
+      "loss": 3.1589,
+      "step": 5880
+    },
+    {
+      "epoch": 1.8588802021637842,
+      "grad_norm": 0.12341500299150605,
+      "learning_rate": 0.0007912631285272793,
+      "loss": 3.0646,
+      "step": 5885
+    },
+    {
+      "epoch": 1.8604596067282635,
+      "grad_norm": 0.13807686009862188,
+      "learning_rate": 0.0007908147131810967,
+      "loss": 2.9419,
+      "step": 5890
+    },
+    {
+      "epoch": 1.8620390112927425,
+      "grad_norm": 0.14018792900335816,
+      "learning_rate": 0.000790365944070518,
+      "loss": 3.1076,
+      "step": 5895
+    },
+    {
+      "epoch": 1.8636184158572218,
+      "grad_norm": 0.11960987802209332,
+      "learning_rate": 0.0007899168217414527,
+      "loss": 3.0158,
+      "step": 5900
+    },
+    {
+      "epoch": 1.865197820421701,
+      "grad_norm": 0.12847048660999708,
+      "learning_rate": 0.00078946734674024,
+      "loss": 3.0331,
+      "step": 5905
+    },
+    {
+      "epoch": 1.8667772249861803,
+      "grad_norm": 0.12510206818033662,
+      "learning_rate": 0.0007890175196136483,
+      "loss": 3.0068,
+      "step": 5910
+    },
+    {
+      "epoch": 1.8683566295506595,
+      "grad_norm": 0.18285088594823792,
+      "learning_rate": 0.000788567340908874,
+      "loss": 3.1459,
+      "step": 5915
+    },
+    {
+      "epoch": 1.8699360341151388,
+      "grad_norm": 0.1587901264563902,
+      "learning_rate": 0.0007881168111735416,
+      "loss": 3.0112,
+      "step": 5920
+    },
+    {
+      "epoch": 1.8715154386796178,
+      "grad_norm": 0.15136208752104693,
+      "learning_rate": 0.0007876659309557022,
+      "loss": 3.1193,
+      "step": 5925
+    },
+    {
+      "epoch": 1.873094843244097,
+      "grad_norm": 0.11287113938962186,
+      "learning_rate": 0.0007872147008038335,
+      "loss": 2.9893,
+      "step": 5930
+    },
+    {
+      "epoch": 1.874674247808576,
+      "grad_norm": 0.12637222768937448,
+      "learning_rate": 0.0007867631212668389,
+      "loss": 2.9688,
+      "step": 5935
+    },
+    {
+      "epoch": 1.8762536523730553,
+      "grad_norm": 0.09236990218264922,
+      "learning_rate": 0.0007863111928940465,
+      "loss": 2.9844,
+      "step": 5940
+    },
+    {
+      "epoch": 1.8778330569375346,
+      "grad_norm": 0.15341474154776769,
+      "learning_rate": 0.0007858589162352092,
+      "loss": 3.0107,
+      "step": 5945
+    },
+    {
+      "epoch": 1.8794124615020138,
+      "grad_norm": 0.12664771964931754,
+      "learning_rate": 0.0007854062918405034,
+      "loss": 3.1368,
+      "step": 5950
+    },
+    {
+      "epoch": 1.880991866066493,
+      "grad_norm": 0.12353028127660556,
+      "learning_rate": 0.0007849533202605284,
+      "loss": 3.0018,
+      "step": 5955
+    },
+    {
+      "epoch": 1.882571270630972,
+      "grad_norm": 0.11672989079025853,
+      "learning_rate": 0.0007845000020463058,
+      "loss": 2.9899,
+      "step": 5960
+    },
+    {
+      "epoch": 1.8841506751954513,
+      "grad_norm": 0.1319299033565842,
+      "learning_rate": 0.0007840463377492789,
+      "loss": 3.0766,
+      "step": 5965
+    },
+    {
+      "epoch": 1.8857300797599303,
+      "grad_norm": 0.15575784095941145,
+      "learning_rate": 0.0007835923279213124,
+      "loss": 3.0619,
+      "step": 5970
+    },
+    {
+      "epoch": 1.8873094843244096,
+      "grad_norm": 0.1655087702809232,
+      "learning_rate": 0.0007831379731146907,
+      "loss": 3.0099,
+      "step": 5975
+    },
+    {
+      "epoch": 1.8888888888888888,
+      "grad_norm": 0.13155693770891339,
+      "learning_rate": 0.0007826832738821181,
+      "loss": 3.0417,
+      "step": 5980
+    },
+    {
+      "epoch": 1.890468293453368,
+      "grad_norm": 0.13713154280184725,
+      "learning_rate": 0.0007822282307767182,
+      "loss": 3.0705,
+      "step": 5985
+    },
+    {
+      "epoch": 1.8920476980178473,
+      "grad_norm": 0.15739920600311663,
+      "learning_rate": 0.0007817728443520323,
+      "loss": 3.0458,
+      "step": 5990
+    },
+    {
+      "epoch": 1.8936271025823266,
+      "grad_norm": 0.17767884822437002,
+      "learning_rate": 0.00078131711516202,
+      "loss": 2.9771,
+      "step": 5995
+    },
+    {
+      "epoch": 1.8952065071468056,
+      "grad_norm": 0.16179754186795273,
+      "learning_rate": 0.0007808610437610573,
+      "loss": 3.0042,
+      "step": 6000
+    },
+    {
+      "epoch": 1.8967859117112849,
+      "grad_norm": 0.1209774401210003,
+      "learning_rate": 0.0007804046307039367,
+      "loss": 3.0069,
+      "step": 6005
+    },
+    {
+      "epoch": 1.8983653162757639,
+      "grad_norm": 0.14899272955577583,
+      "learning_rate": 0.0007799478765458664,
+      "loss": 3.0918,
+      "step": 6010
+    },
+    {
+      "epoch": 1.8999447208402431,
+      "grad_norm": 0.11215130241642793,
+      "learning_rate": 0.0007794907818424694,
+      "loss": 3.0635,
+      "step": 6015
+    },
+    {
+      "epoch": 1.9015241254047224,
+      "grad_norm": 0.11337944435985729,
+      "learning_rate": 0.000779033347149783,
+      "loss": 3.0359,
+      "step": 6020
+    },
+    {
+      "epoch": 1.9031035299692016,
+      "grad_norm": 0.1501609200848775,
+      "learning_rate": 0.0007785755730242584,
+      "loss": 3.1624,
+      "step": 6025
+    },
+    {
+      "epoch": 1.9046829345336809,
+      "grad_norm": 0.1805145194862832,
+      "learning_rate": 0.0007781174600227588,
+      "loss": 3.0417,
+      "step": 6030
+    },
+    {
+      "epoch": 1.9062623390981601,
+      "grad_norm": 0.1290406672184516,
+      "learning_rate": 0.0007776590087025608,
+      "loss": 3.0568,
+      "step": 6035
+    },
+    {
+      "epoch": 1.9078417436626391,
+      "grad_norm": 0.19345941569155078,
+      "learning_rate": 0.0007772002196213516,
+      "loss": 3.1087,
+      "step": 6040
+    },
+    {
+      "epoch": 1.9094211482271184,
+      "grad_norm": 0.14578283353871807,
+      "learning_rate": 0.0007767410933372297,
+      "loss": 3.0578,
+      "step": 6045
+    },
+    {
+      "epoch": 1.9110005527915974,
+      "grad_norm": 0.1832711194459425,
+      "learning_rate": 0.0007762816304087042,
+      "loss": 3.0918,
+      "step": 6050
+    },
+    {
+      "epoch": 1.9125799573560767,
+      "grad_norm": 0.1243198393896939,
+      "learning_rate": 0.0007758218313946926,
+      "loss": 3.055,
+      "step": 6055
+    },
+    {
+      "epoch": 1.914159361920556,
+      "grad_norm": 0.13805415222462827,
+      "learning_rate": 0.0007753616968545222,
+      "loss": 3.1015,
+      "step": 6060
+    },
+    {
+      "epoch": 1.9157387664850352,
+      "grad_norm": 0.12651242359343307,
+      "learning_rate": 0.0007749012273479286,
+      "loss": 3.1397,
+      "step": 6065
+    },
+    {
+      "epoch": 1.9173181710495144,
+      "grad_norm": 0.12844673711773036,
+      "learning_rate": 0.0007744404234350535,
+      "loss": 3.0258,
+      "step": 6070
+    },
+    {
+      "epoch": 1.9188975756139937,
+      "grad_norm": 0.11747370412271375,
+      "learning_rate": 0.0007739792856764472,
+      "loss": 2.9644,
+      "step": 6075
+    },
+    {
+      "epoch": 1.9204769801784727,
+      "grad_norm": 0.1188768226516703,
+      "learning_rate": 0.0007735178146330646,
+      "loss": 3.115,
+      "step": 6080
+    },
+    {
+      "epoch": 1.922056384742952,
+      "grad_norm": 0.11962149432528865,
+      "learning_rate": 0.000773056010866267,
+      "loss": 3.0715,
+      "step": 6085
+    },
+    {
+      "epoch": 1.923635789307431,
+      "grad_norm": 0.10468046220688224,
+      "learning_rate": 0.0007725938749378198,
+      "loss": 2.9093,
+      "step": 6090
+    },
+    {
+      "epoch": 1.9252151938719102,
+      "grad_norm": 0.13615012693421102,
+      "learning_rate": 0.0007721314074098932,
+      "loss": 3.0535,
+      "step": 6095
+    },
+    {
+      "epoch": 1.9267945984363894,
+      "grad_norm": 0.12332476083489902,
+      "learning_rate": 0.00077166860884506,
+      "loss": 3.0133,
+      "step": 6100
+    },
+    {
+      "epoch": 1.9283740030008687,
+      "grad_norm": 0.1554886230963916,
+      "learning_rate": 0.0007712054798062961,
+      "loss": 3.0415,
+      "step": 6105
+    },
+    {
+      "epoch": 1.929953407565348,
+      "grad_norm": 0.12618747274736528,
+      "learning_rate": 0.0007707420208569793,
+      "loss": 2.9687,
+      "step": 6110
+    },
+    {
+      "epoch": 1.9315328121298272,
+      "grad_norm": 0.11465855503094369,
+      "learning_rate": 0.0007702782325608891,
+      "loss": 3.0845,
+      "step": 6115
+    },
+    {
+      "epoch": 1.9331122166943062,
+      "grad_norm": 0.12760381851757738,
+      "learning_rate": 0.0007698141154822047,
+      "loss": 2.9836,
+      "step": 6120
+    },
+    {
+      "epoch": 1.9346916212587855,
+      "grad_norm": 0.12260016682163877,
+      "learning_rate": 0.0007693496701855063,
+      "loss": 3.0311,
+      "step": 6125
+    },
+    {
+      "epoch": 1.9362710258232645,
+      "grad_norm": 0.12577581561632112,
+      "learning_rate": 0.0007688848972357729,
+      "loss": 2.961,
+      "step": 6130
+    },
+    {
+      "epoch": 1.9378504303877437,
+      "grad_norm": 0.13281076636301217,
+      "learning_rate": 0.0007684197971983817,
+      "loss": 2.9881,
+      "step": 6135
+    },
+    {
+      "epoch": 1.939429834952223,
+      "grad_norm": 0.12485075629114431,
+      "learning_rate": 0.0007679543706391088,
+      "loss": 3.1032,
+      "step": 6140
+    },
+    {
+      "epoch": 1.9410092395167022,
+      "grad_norm": 0.13376906867906233,
+      "learning_rate": 0.0007674886181241262,
+      "loss": 3.0929,
+      "step": 6145
+    },
+    {
+      "epoch": 1.9425886440811815,
+      "grad_norm": 0.15326922072492113,
+      "learning_rate": 0.0007670225402200037,
+      "loss": 3.0059,
+      "step": 6150
+    },
+    {
+      "epoch": 1.9441680486456607,
+      "grad_norm": 0.11350105947209056,
+      "learning_rate": 0.0007665561374937059,
+      "loss": 3.0914,
+      "step": 6155
+    },
+    {
+      "epoch": 1.9457474532101398,
+      "grad_norm": 0.1277185457457496,
+      "learning_rate": 0.0007660894105125931,
+      "loss": 3.0533,
+      "step": 6160
+    },
+    {
+      "epoch": 1.947326857774619,
+      "grad_norm": 0.1185686838397195,
+      "learning_rate": 0.0007656223598444199,
+      "loss": 3.0439,
+      "step": 6165
+    },
+    {
+      "epoch": 1.948906262339098,
+      "grad_norm": 0.14586636312671342,
+      "learning_rate": 0.0007651549860573346,
+      "loss": 3.0462,
+      "step": 6170
+    },
+    {
+      "epoch": 1.9504856669035773,
+      "grad_norm": 0.11716530266799677,
+      "learning_rate": 0.0007646872897198786,
+      "loss": 3.0508,
+      "step": 6175
+    },
+    {
+      "epoch": 1.9520650714680565,
+      "grad_norm": 0.1494750093316875,
+      "learning_rate": 0.000764219271400986,
+      "loss": 3.031,
+      "step": 6180
+    },
+    {
+      "epoch": 1.9536444760325358,
+      "grad_norm": 0.15341195455485246,
+      "learning_rate": 0.0007637509316699816,
+      "loss": 3.0286,
+      "step": 6185
+    },
+    {
+      "epoch": 1.955223880597015,
+      "grad_norm": 0.1659584685319081,
+      "learning_rate": 0.0007632822710965826,
+      "loss": 2.9513,
+      "step": 6190
+    },
+    {
+      "epoch": 1.9568032851614943,
+      "grad_norm": 0.16952192795742607,
+      "learning_rate": 0.0007628132902508948,
+      "loss": 3.0618,
+      "step": 6195
+    },
+    {
+      "epoch": 1.9583826897259733,
+      "grad_norm": 0.14244345268202113,
+      "learning_rate": 0.0007623439897034154,
+      "loss": 2.96,
+      "step": 6200
+    },
+    {
+      "epoch": 1.9599620942904525,
+      "grad_norm": 0.13742088890521345,
+      "learning_rate": 0.0007618743700250292,
+      "loss": 3.0885,
+      "step": 6205
+    },
+    {
+      "epoch": 1.9615414988549316,
+      "grad_norm": 0.14591090487732308,
+      "learning_rate": 0.0007614044317870099,
+      "loss": 3.0204,
+      "step": 6210
+    },
+    {
+      "epoch": 1.9631209034194108,
+      "grad_norm": 0.1826312750944958,
+      "learning_rate": 0.0007609341755610181,
+      "loss": 3.0638,
+      "step": 6215
+    },
+    {
+      "epoch": 1.96470030798389,
+      "grad_norm": 0.13793312355059578,
+      "learning_rate": 0.0007604636019191018,
+      "loss": 2.9864,
+      "step": 6220
+    },
+    {
+      "epoch": 1.9662797125483693,
+      "grad_norm": 0.1202600309716602,
+      "learning_rate": 0.0007599927114336947,
+      "loss": 2.9795,
+      "step": 6225
+    },
+    {
+      "epoch": 1.9678591171128486,
+      "grad_norm": 0.10381821439492225,
+      "learning_rate": 0.0007595215046776165,
+      "loss": 3.0202,
+      "step": 6230
+    },
+    {
+      "epoch": 1.9694385216773278,
+      "grad_norm": 0.14311736230412642,
+      "learning_rate": 0.0007590499822240709,
+      "loss": 3.0741,
+      "step": 6235
+    },
+    {
+      "epoch": 1.9710179262418068,
+      "grad_norm": 0.13117212913237875,
+      "learning_rate": 0.0007585781446466463,
+      "loss": 2.9525,
+      "step": 6240
+    },
+    {
+      "epoch": 1.972597330806286,
+      "grad_norm": 0.09759008353033644,
+      "learning_rate": 0.0007581059925193139,
+      "loss": 2.9701,
+      "step": 6245
+    },
+    {
+      "epoch": 1.974176735370765,
+      "grad_norm": 0.14359747362895034,
+      "learning_rate": 0.0007576335264164278,
+      "loss": 3.0283,
+      "step": 6250
+    },
+    {
+      "epoch": 1.9757561399352443,
+      "grad_norm": 0.12444813025537488,
+      "learning_rate": 0.0007571607469127239,
+      "loss": 3.0474,
+      "step": 6255
+    },
+    {
+      "epoch": 1.9773355444997236,
+      "grad_norm": 0.13595472912428896,
+      "learning_rate": 0.0007566876545833197,
+      "loss": 2.9796,
+      "step": 6260
+    },
+    {
+      "epoch": 1.9789149490642028,
+      "grad_norm": 0.16565772286811356,
+      "learning_rate": 0.0007562142500037128,
+      "loss": 3.0639,
+      "step": 6265
+    },
+    {
+      "epoch": 1.980494353628682,
+      "grad_norm": 0.1418512246656067,
+      "learning_rate": 0.0007557405337497809,
+      "loss": 3.0299,
+      "step": 6270
+    },
+    {
+      "epoch": 1.9820737581931613,
+      "grad_norm": 0.15195621317984573,
+      "learning_rate": 0.0007552665063977806,
+      "loss": 3.0469,
+      "step": 6275
+    },
+    {
+      "epoch": 1.9836531627576404,
+      "grad_norm": 0.13748368645509432,
+      "learning_rate": 0.0007547921685243475,
+      "loss": 3.1035,
+      "step": 6280
+    },
+    {
+      "epoch": 1.9852325673221196,
+      "grad_norm": 0.1639705916815782,
+      "learning_rate": 0.0007543175207064941,
+      "loss": 2.9557,
+      "step": 6285
+    },
+    {
+      "epoch": 1.9868119718865986,
+      "grad_norm": 0.12493567540744403,
+      "learning_rate": 0.0007538425635216104,
+      "loss": 3.048,
+      "step": 6290
+    },
+    {
+      "epoch": 1.9883913764510779,
+      "grad_norm": 0.11329383002467514,
+      "learning_rate": 0.000753367297547463,
+      "loss": 3.0361,
+      "step": 6295
+    },
+    {
+      "epoch": 1.9899707810155571,
+      "grad_norm": 0.10547167696430386,
+      "learning_rate": 0.0007528917233621937,
+      "loss": 2.9355,
+      "step": 6300
+    },
+    {
+      "epoch": 1.9915501855800364,
+      "grad_norm": 0.10672825007534288,
+      "learning_rate": 0.0007524158415443192,
+      "loss": 3.0004,
+      "step": 6305
+    },
+    {
+      "epoch": 1.9931295901445156,
+      "grad_norm": 0.08754943202048324,
+      "learning_rate": 0.0007519396526727309,
+      "loss": 2.9292,
+      "step": 6310
+    },
+    {
+      "epoch": 1.9947089947089947,
+      "grad_norm": 0.10665404162506625,
+      "learning_rate": 0.000751463157326693,
+      "loss": 2.9146,
+      "step": 6315
+    },
+    {
+      "epoch": 1.996288399273474,
+      "grad_norm": 0.12577526351275167,
+      "learning_rate": 0.0007509863560858432,
+      "loss": 3.0294,
+      "step": 6320
+    },
+    {
+      "epoch": 1.997867803837953,
+      "grad_norm": 0.15679239066245074,
+      "learning_rate": 0.0007505092495301911,
+      "loss": 3.0487,
+      "step": 6325
+    },
+    {
+      "epoch": 1.9994472084024322,
+      "grad_norm": 0.14437984382007163,
+      "learning_rate": 0.0007500318382401173,
+      "loss": 3.0537,
+      "step": 6330
+    },
+    {
+      "epoch": 2.0,
+      "eval_loss": 3.002992630004883,
+      "eval_runtime": 118.632,
+      "eval_samples_per_second": 22.33,
+      "eval_steps_per_second": 5.589,
+      "step": 6332
+    },
+    {
+      "epoch": 2.0009476427386876,
+      "grad_norm": 0.14353458130909943,
+      "learning_rate": 0.0007495541227963736,
+      "loss": 3.0413,
+      "step": 6335
+    },
+    {
+      "epoch": 2.002527047303167,
+      "grad_norm": 0.1156761857836109,
+      "learning_rate": 0.0007490761037800815,
+      "loss": 2.9703,
+      "step": 6340
+    },
+    {
+      "epoch": 2.0041064518676457,
+      "grad_norm": 0.12923385153054337,
+      "learning_rate": 0.0007485977817727322,
+      "loss": 3.0492,
+      "step": 6345
+    },
+    {
+      "epoch": 2.005685856432125,
+      "grad_norm": 0.15716399707192955,
+      "learning_rate": 0.0007481191573561849,
+      "loss": 2.9633,
+      "step": 6350
+    },
+    {
+      "epoch": 2.007265260996604,
+      "grad_norm": 0.11284991969173797,
+      "learning_rate": 0.0007476402311126672,
+      "loss": 2.8979,
+      "step": 6355
+    },
+    {
+      "epoch": 2.0088446655610834,
+      "grad_norm": 0.10724149749864116,
+      "learning_rate": 0.0007471610036247732,
+      "loss": 2.9954,
+      "step": 6360
+    },
+    {
+      "epoch": 2.0104240701255627,
+      "grad_norm": 0.12801366254983276,
+      "learning_rate": 0.0007466814754754642,
+      "loss": 2.9294,
+      "step": 6365
+    },
+    {
+      "epoch": 2.012003474690042,
+      "grad_norm": 0.10576933472118467,
+      "learning_rate": 0.0007462016472480667,
+      "loss": 2.958,
+      "step": 6370
+    },
+    {
+      "epoch": 2.013582879254521,
+      "grad_norm": 0.13332911256730795,
+      "learning_rate": 0.0007457215195262726,
+      "loss": 2.9382,
+      "step": 6375
+    },
+    {
+      "epoch": 2.0151622838190004,
+      "grad_norm": 0.13607445289009712,
+      "learning_rate": 0.0007452410928941378,
+      "loss": 3.1146,
+      "step": 6380
+    },
+    {
+      "epoch": 2.0167416883834792,
+      "grad_norm": 0.11119726843716685,
+      "learning_rate": 0.000744760367936082,
+      "loss": 3.0029,
+      "step": 6385
+    },
+    {
+      "epoch": 2.0183210929479585,
+      "grad_norm": 0.12559115194392256,
+      "learning_rate": 0.0007442793452368879,
+      "loss": 3.0314,
+      "step": 6390
+    },
+    {
+      "epoch": 2.0199004975124377,
+      "grad_norm": 0.15043200233343293,
+      "learning_rate": 0.0007437980253817003,
+      "loss": 3.1284,
+      "step": 6395
+    },
+    {
+      "epoch": 2.021479902076917,
+      "grad_norm": 0.1408936832559694,
+      "learning_rate": 0.000743316408956025,
+      "loss": 3.0146,
+      "step": 6400
+    },
+    {
+      "epoch": 2.023059306641396,
+      "grad_norm": 0.1522614342957081,
+      "learning_rate": 0.0007428344965457294,
+      "loss": 2.9533,
+      "step": 6405
+    },
+    {
+      "epoch": 2.0246387112058755,
+      "grad_norm": 0.13815257927745242,
+      "learning_rate": 0.0007423522887370404,
+      "loss": 3.0147,
+      "step": 6410
+    },
+    {
+      "epoch": 2.0262181157703547,
+      "grad_norm": 0.13658614002521371,
+      "learning_rate": 0.0007418697861165444,
+      "loss": 3.0088,
+      "step": 6415
+    },
+    {
+      "epoch": 2.027797520334834,
+      "grad_norm": 0.1132770103349406,
+      "learning_rate": 0.0007413869892711867,
+      "loss": 3.0153,
+      "step": 6420
+    },
+    {
+      "epoch": 2.0293769248993128,
+      "grad_norm": 0.11257282557810316,
+      "learning_rate": 0.0007409038987882697,
+      "loss": 2.9768,
+      "step": 6425
+    },
+    {
+      "epoch": 2.030956329463792,
+      "grad_norm": 0.1387413305440576,
+      "learning_rate": 0.000740420515255454,
+      "loss": 2.9256,
+      "step": 6430
+    },
+    {
+      "epoch": 2.0325357340282713,
+      "grad_norm": 0.12733121878072734,
+      "learning_rate": 0.0007399368392607561,
+      "loss": 3.0091,
+      "step": 6435
+    },
+    {
+      "epoch": 2.0341151385927505,
+      "grad_norm": 0.15164433152738804,
+      "learning_rate": 0.0007394528713925481,
+      "loss": 3.0224,
+      "step": 6440
+    },
+    {
+      "epoch": 2.0356945431572298,
+      "grad_norm": 0.17255392989092136,
+      "learning_rate": 0.0007389686122395579,
+      "loss": 3.0004,
+      "step": 6445
+    },
+    {
+      "epoch": 2.037273947721709,
+      "grad_norm": 0.1494347145988662,
+      "learning_rate": 0.000738484062390867,
+      "loss": 3.0202,
+      "step": 6450
+    },
+    {
+      "epoch": 2.0388533522861882,
+      "grad_norm": 0.1286118513427532,
+      "learning_rate": 0.0007379992224359108,
+      "loss": 2.9945,
+      "step": 6455
+    },
+    {
+      "epoch": 2.0404327568506675,
+      "grad_norm": 0.1330159530967352,
+      "learning_rate": 0.0007375140929644776,
+      "loss": 2.9584,
+      "step": 6460
+    },
+    {
+      "epoch": 2.0420121614151463,
+      "grad_norm": 0.10649752052665366,
+      "learning_rate": 0.000737028674566708,
+      "loss": 2.9748,
+      "step": 6465
+    },
+    {
+      "epoch": 2.0435915659796255,
+      "grad_norm": 0.11051724224394212,
+      "learning_rate": 0.0007365429678330937,
+      "loss": 2.9452,
+      "step": 6470
+    },
+    {
+      "epoch": 2.045170970544105,
+      "grad_norm": 0.13499758720764418,
+      "learning_rate": 0.0007360569733544778,
+      "loss": 2.9613,
+      "step": 6475
+    },
+    {
+      "epoch": 2.046750375108584,
+      "grad_norm": 0.1243465553360808,
+      "learning_rate": 0.0007355706917220524,
+      "loss": 3.0741,
+      "step": 6480
+    },
+    {
+      "epoch": 2.0483297796730633,
+      "grad_norm": 0.08486693691641276,
+      "learning_rate": 0.0007350841235273602,
+      "loss": 2.9575,
+      "step": 6485
+    },
+    {
+      "epoch": 2.0499091842375425,
+      "grad_norm": 0.11758078015740667,
+      "learning_rate": 0.0007345972693622915,
+      "loss": 3.0356,
+      "step": 6490
+    },
+    {
+      "epoch": 2.051488588802022,
+      "grad_norm": 0.11659997718476248,
+      "learning_rate": 0.0007341101298190849,
+      "loss": 2.9863,
+      "step": 6495
+    },
+    {
+      "epoch": 2.053067993366501,
+      "grad_norm": 0.09389891185481668,
+      "learning_rate": 0.0007336227054903258,
+      "loss": 3.1458,
+      "step": 6500
+    },
+    {
+      "epoch": 2.05464739793098,
+      "grad_norm": 0.13480016807876233,
+      "learning_rate": 0.0007331349969689467,
+      "loss": 3.0026,
+      "step": 6505
+    },
+    {
+      "epoch": 2.056226802495459,
+      "grad_norm": 0.12055090302454477,
+      "learning_rate": 0.000732647004848225,
+      "loss": 2.9464,
+      "step": 6510
+    },
+    {
+      "epoch": 2.0578062070599383,
+      "grad_norm": 0.1148060412272345,
+      "learning_rate": 0.0007321587297217838,
+      "loss": 3.0115,
+      "step": 6515
+    },
+    {
+      "epoch": 2.0593856116244176,
+      "grad_norm": 0.1215434220029464,
+      "learning_rate": 0.0007316701721835899,
+      "loss": 2.9983,
+      "step": 6520
+    },
+    {
+      "epoch": 2.060965016188897,
+      "grad_norm": 0.14364596094061724,
+      "learning_rate": 0.000731181332827954,
+      "loss": 2.9262,
+      "step": 6525
+    },
+    {
+      "epoch": 2.062544420753376,
+      "grad_norm": 0.10659520229926726,
+      "learning_rate": 0.0007306922122495295,
+      "loss": 3.0127,
+      "step": 6530
+    },
+    {
+      "epoch": 2.0641238253178553,
+      "grad_norm": 0.13323935904303538,
+      "learning_rate": 0.0007302028110433118,
+      "loss": 3.1008,
+      "step": 6535
+    },
+    {
+      "epoch": 2.0657032298823346,
+      "grad_norm": 0.13533198826623102,
+      "learning_rate": 0.000729713129804638,
+      "loss": 3.081,
+      "step": 6540
+    },
+    {
+      "epoch": 2.0672826344468134,
+      "grad_norm": 0.142589431829919,
+      "learning_rate": 0.0007292231691291854,
+      "loss": 3.036,
+      "step": 6545
+    },
+    {
+      "epoch": 2.0688620390112926,
+      "grad_norm": 0.13569810390729722,
+      "learning_rate": 0.0007287329296129715,
+      "loss": 2.9285,
+      "step": 6550
+    },
+    {
+      "epoch": 2.070441443575772,
+      "grad_norm": 0.1155039408631233,
+      "learning_rate": 0.0007282424118523531,
+      "loss": 2.94,
+      "step": 6555
+    },
+    {
+      "epoch": 2.072020848140251,
+      "grad_norm": 0.12295926038581785,
+      "learning_rate": 0.000727751616444025,
+      "loss": 3.0605,
+      "step": 6560
+    },
+    {
+      "epoch": 2.0736002527047304,
+      "grad_norm": 0.12305628092828373,
+      "learning_rate": 0.0007272605439850205,
+      "loss": 3.023,
+      "step": 6565
+    },
+    {
+      "epoch": 2.0751796572692096,
+      "grad_norm": 0.12215737076018068,
+      "learning_rate": 0.0007267691950727089,
+      "loss": 3.0601,
+      "step": 6570
+    },
+    {
+      "epoch": 2.076759061833689,
+      "grad_norm": 0.1276869872243518,
+      "learning_rate": 0.000726277570304797,
+      "loss": 2.9532,
+      "step": 6575
+    },
+    {
+      "epoch": 2.078338466398168,
+      "grad_norm": 0.12334843350229036,
+      "learning_rate": 0.0007257856702793261,
+      "loss": 3.0194,
+      "step": 6580
+    },
+    {
+      "epoch": 2.079917870962647,
+      "grad_norm": 0.13940263746621434,
+      "learning_rate": 0.0007252934955946732,
+      "loss": 3.0748,
+      "step": 6585
+    },
+    {
+      "epoch": 2.081497275527126,
+      "grad_norm": 0.11565673251438623,
+      "learning_rate": 0.0007248010468495485,
+      "loss": 2.954,
+      "step": 6590
+    },
+    {
+      "epoch": 2.0830766800916054,
+      "grad_norm": 0.10806739091445611,
+      "learning_rate": 0.0007243083246429964,
+      "loss": 3.0082,
+      "step": 6595
+    },
+    {
+      "epoch": 2.0846560846560847,
+      "grad_norm": 0.12970993194796246,
+      "learning_rate": 0.0007238153295743935,
+      "loss": 2.8832,
+      "step": 6600
+    },
+    {
+      "epoch": 2.086235489220564,
+      "grad_norm": 0.1147358808877119,
+      "learning_rate": 0.0007233220622434488,
+      "loss": 2.9476,
+      "step": 6605
+    },
+    {
+      "epoch": 2.087814893785043,
+      "grad_norm": 0.1210669420794267,
+      "learning_rate": 0.0007228285232502015,
+      "loss": 2.9101,
+      "step": 6610
+    },
+    {
+      "epoch": 2.0893942983495224,
+      "grad_norm": 0.1606733718489849,
+      "learning_rate": 0.0007223347131950226,
+      "loss": 2.9651,
+      "step": 6615
+    },
+    {
+      "epoch": 2.0909737029140016,
+      "grad_norm": 0.13731530763006436,
+      "learning_rate": 0.0007218406326786119,
+      "loss": 3.0062,
+      "step": 6620
+    },
+    {
+      "epoch": 2.0925531074784804,
+      "grad_norm": 0.10877478662149911,
+      "learning_rate": 0.0007213462823019983,
+      "loss": 2.9584,
+      "step": 6625
+    },
+    {
+      "epoch": 2.0941325120429597,
+      "grad_norm": 0.13160049485588202,
+      "learning_rate": 0.0007208516626665394,
+      "loss": 2.9477,
+      "step": 6630
+    },
+    {
+      "epoch": 2.095711916607439,
+      "grad_norm": 0.1520902274619921,
+      "learning_rate": 0.0007203567743739198,
+      "loss": 3.0736,
+      "step": 6635
+    },
+    {
+      "epoch": 2.097291321171918,
+      "grad_norm": 0.14340603399708876,
+      "learning_rate": 0.0007198616180261514,
+      "loss": 2.9981,
+      "step": 6640
+    },
+    {
+      "epoch": 2.0988707257363974,
+      "grad_norm": 0.13195937834853005,
+      "learning_rate": 0.0007193661942255722,
+      "loss": 2.9252,
+      "step": 6645
+    },
+    {
+      "epoch": 2.1004501303008767,
+      "grad_norm": 0.1421720750596484,
+      "learning_rate": 0.0007188705035748446,
+      "loss": 2.9677,
+      "step": 6650
+    },
+    {
+      "epoch": 2.102029534865356,
+      "grad_norm": 0.11834592868666237,
+      "learning_rate": 0.0007183745466769572,
+      "loss": 2.99,
+      "step": 6655
+    },
+    {
+      "epoch": 2.1036089394298347,
+      "grad_norm": 0.12182266431595128,
+      "learning_rate": 0.000717878324135221,
+      "loss": 3.0528,
+      "step": 6660
+    },
+    {
+      "epoch": 2.105188343994314,
+      "grad_norm": 0.1022121029896845,
+      "learning_rate": 0.0007173818365532709,
+      "loss": 3.0244,
+      "step": 6665
+    },
+    {
+      "epoch": 2.1067677485587932,
+      "grad_norm": 0.1301026443398586,
+      "learning_rate": 0.0007168850845350642,
+      "loss": 3.0274,
+      "step": 6670
+    },
+    {
+      "epoch": 2.1083471531232725,
+      "grad_norm": 0.11957679595275474,
+      "learning_rate": 0.0007163880686848796,
+      "loss": 2.9305,
+      "step": 6675
+    },
+    {
+      "epoch": 2.1099265576877517,
+      "grad_norm": 0.13180237905943235,
+      "learning_rate": 0.000715890789607317,
+      "loss": 3.0343,
+      "step": 6680
+    },
+    {
+      "epoch": 2.111505962252231,
+      "grad_norm": 0.13843354550897705,
+      "learning_rate": 0.0007153932479072963,
+      "loss": 2.9075,
+      "step": 6685
+    },
+    {
+      "epoch": 2.11308536681671,
+      "grad_norm": 0.13166142357617788,
+      "learning_rate": 0.0007148954441900568,
+      "loss": 2.9276,
+      "step": 6690
+    },
+    {
+      "epoch": 2.1146647713811895,
+      "grad_norm": 0.1286711963390611,
+      "learning_rate": 0.0007143973790611571,
+      "loss": 2.9668,
+      "step": 6695
+    },
+    {
+      "epoch": 2.1162441759456683,
+      "grad_norm": 0.10836636344134955,
+      "learning_rate": 0.000713899053126473,
+      "loss": 3.054,
+      "step": 6700
+    },
+    {
+      "epoch": 2.1178235805101475,
+      "grad_norm": 0.10852062206377763,
+      "learning_rate": 0.0007134004669921983,
+      "loss": 2.9124,
+      "step": 6705
+    },
+    {
+      "epoch": 2.1194029850746268,
+      "grad_norm": 0.10533876658435133,
+      "learning_rate": 0.0007129016212648425,
+      "loss": 2.9217,
+      "step": 6710
+    },
+    {
+      "epoch": 2.120982389639106,
+      "grad_norm": 0.14748853715817997,
+      "learning_rate": 0.0007124025165512318,
+      "loss": 2.9049,
+      "step": 6715
+    },
+    {
+      "epoch": 2.1225617942035853,
+      "grad_norm": 0.10497458796252031,
+      "learning_rate": 0.0007119031534585068,
+      "loss": 2.9351,
+      "step": 6720
+    },
+    {
+      "epoch": 2.1241411987680645,
+      "grad_norm": 0.10538082248904294,
+      "learning_rate": 0.0007114035325941226,
+      "loss": 3.0758,
+      "step": 6725
+    },
+    {
+      "epoch": 2.1257206033325438,
+      "grad_norm": 0.12439664097265367,
+      "learning_rate": 0.0007109036545658478,
+      "loss": 2.9336,
+      "step": 6730
+    },
+    {
+      "epoch": 2.127300007897023,
+      "grad_norm": 0.11718790665980608,
+      "learning_rate": 0.0007104035199817642,
+      "loss": 2.9641,
+      "step": 6735
+    },
+    {
+      "epoch": 2.128879412461502,
+      "grad_norm": 0.12660875592728832,
+      "learning_rate": 0.0007099031294502651,
+      "loss": 2.987,
+      "step": 6740
+    },
+    {
+      "epoch": 2.130458817025981,
+      "grad_norm": 0.15718244268313566,
+      "learning_rate": 0.0007094024835800557,
+      "loss": 2.9108,
+      "step": 6745
+    },
+    {
+      "epoch": 2.1320382215904603,
+      "grad_norm": 0.1553720198584762,
+      "learning_rate": 0.0007089015829801513,
+      "loss": 3.0578,
+      "step": 6750
+    },
+    {
+      "epoch": 2.1336176261549396,
+      "grad_norm": 0.11568572734140488,
+      "learning_rate": 0.0007084004282598774,
+      "loss": 2.9781,
+      "step": 6755
+    },
+    {
+      "epoch": 2.135197030719419,
+      "grad_norm": 0.11782283702381052,
+      "learning_rate": 0.0007078990200288685,
+      "loss": 2.9762,
+      "step": 6760
+    },
+    {
+      "epoch": 2.136776435283898,
+      "grad_norm": 0.14707229231637933,
+      "learning_rate": 0.0007073973588970678,
+      "loss": 2.9615,
+      "step": 6765
+    },
+    {
+      "epoch": 2.1383558398483773,
+      "grad_norm": 0.1214209064411245,
+      "learning_rate": 0.0007068954454747256,
+      "loss": 2.9987,
+      "step": 6770
+    },
+    {
+      "epoch": 2.1399352444128565,
+      "grad_norm": 0.12998961493455885,
+      "learning_rate": 0.0007063932803723996,
+      "loss": 3.0439,
+      "step": 6775
+    },
+    {
+      "epoch": 2.1415146489773353,
+      "grad_norm": 0.14708814785711635,
+      "learning_rate": 0.0007058908642009531,
+      "loss": 2.8839,
+      "step": 6780
+    },
+    {
+      "epoch": 2.1430940535418146,
+      "grad_norm": 0.12748564529746748,
+      "learning_rate": 0.0007053881975715557,
+      "loss": 2.9489,
+      "step": 6785
+    },
+    {
+      "epoch": 2.144673458106294,
+      "grad_norm": 0.1321189798945764,
+      "learning_rate": 0.0007048852810956805,
+      "loss": 2.9256,
+      "step": 6790
+    },
+    {
+      "epoch": 2.146252862670773,
+      "grad_norm": 0.1304762482298313,
+      "learning_rate": 0.0007043821153851057,
+      "loss": 2.9561,
+      "step": 6795
+    },
+    {
+      "epoch": 2.1478322672352523,
+      "grad_norm": 0.11486472715401491,
+      "learning_rate": 0.0007038787010519116,
+      "loss": 2.9833,
+      "step": 6800
+    },
+    {
+      "epoch": 2.1494116717997316,
+      "grad_norm": 0.12315872643569079,
+      "learning_rate": 0.0007033750387084821,
+      "loss": 2.8684,
+      "step": 6805
+    },
+    {
+      "epoch": 2.150991076364211,
+      "grad_norm": 0.12142457266007015,
+      "learning_rate": 0.0007028711289675016,
+      "loss": 3.0012,
+      "step": 6810
+    },
+    {
+      "epoch": 2.15257048092869,
+      "grad_norm": 0.12104114201457632,
+      "learning_rate": 0.0007023669724419563,
+      "loss": 2.9969,
+      "step": 6815
+    },
+    {
+      "epoch": 2.154149885493169,
+      "grad_norm": 0.11986057914479814,
+      "learning_rate": 0.0007018625697451326,
+      "loss": 2.9424,
+      "step": 6820
+    },
+    {
+      "epoch": 2.155729290057648,
+      "grad_norm": 0.11076847146953693,
+      "learning_rate": 0.0007013579214906155,
+      "loss": 2.9856,
+      "step": 6825
+    },
+    {
+      "epoch": 2.1573086946221274,
+      "grad_norm": 0.10961247367316544,
+      "learning_rate": 0.0007008530282922896,
+      "loss": 2.9276,
+      "step": 6830
+    },
+    {
+      "epoch": 2.1588880991866066,
+      "grad_norm": 0.13011644057314592,
+      "learning_rate": 0.0007003478907643372,
+      "loss": 2.9269,
+      "step": 6835
+    },
+    {
+      "epoch": 2.160467503751086,
+      "grad_norm": 0.1166701938575446,
+      "learning_rate": 0.0006998425095212377,
+      "loss": 2.9917,
+      "step": 6840
+    },
+    {
+      "epoch": 2.162046908315565,
+      "grad_norm": 0.1456926577037329,
+      "learning_rate": 0.0006993368851777671,
+      "loss": 2.9533,
+      "step": 6845
+    },
+    {
+      "epoch": 2.1636263128800444,
+      "grad_norm": 0.1146375820891622,
+      "learning_rate": 0.0006988310183489968,
+      "loss": 2.8984,
+      "step": 6850
+    },
+    {
+      "epoch": 2.1652057174445236,
+      "grad_norm": 0.11696640157874294,
+      "learning_rate": 0.000698324909650294,
+      "loss": 2.9509,
+      "step": 6855
+    },
+    {
+      "epoch": 2.1667851220090024,
+      "grad_norm": 0.11263288404211443,
+      "learning_rate": 0.0006978185596973191,
+      "loss": 2.9958,
+      "step": 6860
+    },
+    {
+      "epoch": 2.1683645265734817,
+      "grad_norm": 0.10937990573667906,
+      "learning_rate": 0.0006973119691060266,
+      "loss": 3.0493,
+      "step": 6865
+    },
+    {
+      "epoch": 2.169943931137961,
+      "grad_norm": 0.12060921543059229,
+      "learning_rate": 0.0006968051384926634,
+      "loss": 3.0304,
+      "step": 6870
+    },
+    {
+      "epoch": 2.17152333570244,
+      "grad_norm": 0.13499039916380462,
+      "learning_rate": 0.0006962980684737688,
+      "loss": 2.9113,
+      "step": 6875
+    },
+    {
+      "epoch": 2.1731027402669194,
+      "grad_norm": 0.12796690809536898,
+      "learning_rate": 0.0006957907596661729,
+      "loss": 3.056,
+      "step": 6880
+    },
+    {
+      "epoch": 2.1746821448313987,
+      "grad_norm": 0.11823051242354714,
+      "learning_rate": 0.0006952832126869966,
+      "loss": 3.0357,
+      "step": 6885
+    },
+    {
+      "epoch": 2.176261549395878,
+      "grad_norm": 0.15676859394646797,
+      "learning_rate": 0.0006947754281536502,
+      "loss": 2.9343,
+      "step": 6890
+    },
+    {
+      "epoch": 2.177840953960357,
+      "grad_norm": 0.12703771956860213,
+      "learning_rate": 0.0006942674066838332,
+      "loss": 2.9846,
+      "step": 6895
+    },
+    {
+      "epoch": 2.179420358524836,
+      "grad_norm": 0.09810456905197461,
+      "learning_rate": 0.0006937591488955334,
+      "loss": 2.9459,
+      "step": 6900
+    },
+    {
+      "epoch": 2.180999763089315,
+      "grad_norm": 0.13643249438798305,
+      "learning_rate": 0.0006932506554070259,
+      "loss": 3.0622,
+      "step": 6905
+    },
+    {
+      "epoch": 2.1825791676537944,
+      "grad_norm": 0.15175111503072894,
+      "learning_rate": 0.0006927419268368726,
+      "loss": 2.9585,
+      "step": 6910
+    },
+    {
+      "epoch": 2.1841585722182737,
+      "grad_norm": 0.16395519307966172,
+      "learning_rate": 0.0006922329638039211,
+      "loss": 2.9912,
+      "step": 6915
+    },
+    {
+      "epoch": 2.185737976782753,
+      "grad_norm": 0.16750759783597607,
+      "learning_rate": 0.0006917237669273046,
+      "loss": 2.9087,
+      "step": 6920
+    },
+    {
+      "epoch": 2.187317381347232,
+      "grad_norm": 0.12715403500537467,
+      "learning_rate": 0.0006912143368264408,
+      "loss": 2.8529,
+      "step": 6925
+    },
+    {
+      "epoch": 2.1888967859117114,
+      "grad_norm": 0.10608221046373563,
+      "learning_rate": 0.0006907046741210308,
+      "loss": 2.8785,
+      "step": 6930
+    },
+    {
+      "epoch": 2.1904761904761907,
+      "grad_norm": 0.1208868677965757,
+      "learning_rate": 0.0006901947794310583,
+      "loss": 2.9761,
+      "step": 6935
+    },
+    {
+      "epoch": 2.1920555950406695,
+      "grad_norm": 0.12530056311098328,
+      "learning_rate": 0.0006896846533767906,
+      "loss": 2.9773,
+      "step": 6940
+    },
+    {
+      "epoch": 2.1936349996051487,
+      "grad_norm": 0.1154836339364008,
+      "learning_rate": 0.0006891742965787746,
+      "loss": 3.018,
+      "step": 6945
+    },
+    {
+      "epoch": 2.195214404169628,
+      "grad_norm": 0.10701635685168409,
+      "learning_rate": 0.0006886637096578394,
+      "loss": 3.0047,
+      "step": 6950
+    },
+    {
+      "epoch": 2.1967938087341072,
+      "grad_norm": 0.13623213124777278,
+      "learning_rate": 0.0006881528932350932,
+      "loss": 2.9208,
+      "step": 6955
+    },
+    {
+      "epoch": 2.1983732132985865,
+      "grad_norm": 0.10290165801688178,
+      "learning_rate": 0.0006876418479319238,
+      "loss": 2.932,
+      "step": 6960
+    },
+    {
+      "epoch": 2.1999526178630657,
+      "grad_norm": 0.10699922828371627,
+      "learning_rate": 0.000687130574369997,
+      "loss": 2.9233,
+      "step": 6965
+    },
+    {
+      "epoch": 2.201532022427545,
+      "grad_norm": 0.1695041583210203,
+      "learning_rate": 0.0006866190731712566,
+      "loss": 2.939,
+      "step": 6970
+    },
+    {
+      "epoch": 2.203111426992024,
+      "grad_norm": 0.13247946534689092,
+      "learning_rate": 0.0006861073449579233,
+      "loss": 3.0769,
+      "step": 6975
+    },
+    {
+      "epoch": 2.204690831556503,
+      "grad_norm": 0.12083281516035982,
+      "learning_rate": 0.0006855953903524939,
+      "loss": 2.8871,
+      "step": 6980
+    },
+    {
+      "epoch": 2.2062702361209823,
+      "grad_norm": 0.14866175804793844,
+      "learning_rate": 0.0006850832099777404,
+      "loss": 3.1146,
+      "step": 6985
+    },
+    {
+      "epoch": 2.2078496406854615,
+      "grad_norm": 0.1497492816535961,
+      "learning_rate": 0.0006845708044567099,
+      "loss": 2.9031,
+      "step": 6990
+    },
+    {
+      "epoch": 2.2094290452499408,
+      "grad_norm": 0.1421570127106526,
+      "learning_rate": 0.0006840581744127227,
+      "loss": 3.0816,
+      "step": 6995
+    },
+    {
+      "epoch": 2.21100844981442,
+      "grad_norm": 0.13829763776583967,
+      "learning_rate": 0.0006835453204693732,
+      "loss": 2.8834,
+      "step": 7000
+    },
+    {
+      "epoch": 2.2125878543788993,
+      "grad_norm": 0.12307842482923538,
+      "learning_rate": 0.000683032243250527,
+      "loss": 3.1174,
+      "step": 7005
+    },
+    {
+      "epoch": 2.2141672589433785,
+      "grad_norm": 0.16581956371172818,
+      "learning_rate": 0.0006825189433803222,
+      "loss": 2.9609,
+      "step": 7010
+    },
+    {
+      "epoch": 2.2157466635078578,
+      "grad_norm": 0.10833612483320888,
+      "learning_rate": 0.0006820054214831673,
+      "loss": 2.8756,
+      "step": 7015
+    },
+    {
+      "epoch": 2.2173260680723366,
+      "grad_norm": 0.11521930594607378,
+      "learning_rate": 0.0006814916781837413,
+      "loss": 2.9566,
+      "step": 7020
+    },
+    {
+      "epoch": 2.218905472636816,
+      "grad_norm": 0.11063510835048242,
+      "learning_rate": 0.0006809777141069917,
+      "loss": 2.8834,
+      "step": 7025
+    },
+    {
+      "epoch": 2.220484877201295,
+      "grad_norm": 0.13148973707112352,
+      "learning_rate": 0.0006804635298781358,
+      "loss": 2.9195,
+      "step": 7030
+    },
+    {
+      "epoch": 2.2220642817657743,
+      "grad_norm": 0.11620455696252242,
+      "learning_rate": 0.0006799491261226574,
+      "loss": 2.9167,
+      "step": 7035
+    },
+    {
+      "epoch": 2.2236436863302536,
+      "grad_norm": 0.1396608974442779,
+      "learning_rate": 0.0006794345034663084,
+      "loss": 2.9995,
+      "step": 7040
+    },
+    {
+      "epoch": 2.225223090894733,
+      "grad_norm": 0.10532649173801359,
+      "learning_rate": 0.0006789196625351064,
+      "loss": 2.9799,
+      "step": 7045
+    },
+    {
+      "epoch": 2.226802495459212,
+      "grad_norm": 0.1322420670483448,
+      "learning_rate": 0.0006784046039553346,
+      "loss": 3.0018,
+      "step": 7050
+    },
+    {
+      "epoch": 2.228381900023691,
+      "grad_norm": 0.13136086274387426,
+      "learning_rate": 0.0006778893283535411,
+      "loss": 2.9256,
+      "step": 7055
+    },
+    {
+      "epoch": 2.22996130458817,
+      "grad_norm": 0.10244743453725035,
+      "learning_rate": 0.0006773738363565381,
+      "loss": 2.967,
+      "step": 7060
+    },
+    {
+      "epoch": 2.2315407091526493,
+      "grad_norm": 0.13231108581539203,
+      "learning_rate": 0.0006768581285914006,
+      "loss": 2.8923,
+      "step": 7065
+    },
+    {
+      "epoch": 2.2331201137171286,
+      "grad_norm": 0.10492552753211626,
+      "learning_rate": 0.0006763422056854665,
+      "loss": 2.9883,
+      "step": 7070
+    },
+    {
+      "epoch": 2.234699518281608,
+      "grad_norm": 0.12023814021868094,
+      "learning_rate": 0.0006758260682663351,
+      "loss": 3.0296,
+      "step": 7075
+    },
+    {
+      "epoch": 2.236278922846087,
+      "grad_norm": 0.11611685386311155,
+      "learning_rate": 0.0006753097169618672,
+      "loss": 3.0089,
+      "step": 7080
+    },
+    {
+      "epoch": 2.2378583274105663,
+      "grad_norm": 0.10790971144102764,
+      "learning_rate": 0.0006747931524001829,
+      "loss": 3.0234,
+      "step": 7085
+    },
+    {
+      "epoch": 2.2394377319750456,
+      "grad_norm": 0.13591696362025332,
+      "learning_rate": 0.0006742763752096624,
+      "loss": 2.9541,
+      "step": 7090
+    },
+    {
+      "epoch": 2.241017136539525,
+      "grad_norm": 0.170771413477413,
+      "learning_rate": 0.0006737593860189444,
+      "loss": 2.9583,
+      "step": 7095
+    },
+    {
+      "epoch": 2.2425965411040036,
+      "grad_norm": 0.1171310887058597,
+      "learning_rate": 0.0006732421854569254,
+      "loss": 2.9543,
+      "step": 7100
+    },
+    {
+      "epoch": 2.244175945668483,
+      "grad_norm": 0.11847176351347768,
+      "learning_rate": 0.0006727247741527591,
+      "loss": 2.9623,
+      "step": 7105
+    },
+    {
+      "epoch": 2.245755350232962,
+      "grad_norm": 0.118122754884454,
+      "learning_rate": 0.0006722071527358556,
+      "loss": 2.9223,
+      "step": 7110
+    },
+    {
+      "epoch": 2.2473347547974414,
+      "grad_norm": 0.17184169265760627,
+      "learning_rate": 0.0006716893218358803,
+      "loss": 2.95,
+      "step": 7115
+    },
+    {
+      "epoch": 2.2489141593619206,
+      "grad_norm": 0.12136410495299987,
+      "learning_rate": 0.0006711712820827538,
+      "loss": 2.9878,
+      "step": 7120
+    },
+    {
+      "epoch": 2.2504935639264,
+      "grad_norm": 0.15124997481005967,
+      "learning_rate": 0.0006706530341066506,
+      "loss": 3.0159,
+      "step": 7125
+    },
+    {
+      "epoch": 2.252072968490879,
+      "grad_norm": 0.11705380254761119,
+      "learning_rate": 0.0006701345785379986,
+      "loss": 2.9411,
+      "step": 7130
+    },
+    {
+      "epoch": 2.253652373055358,
+      "grad_norm": 0.13077135432868156,
+      "learning_rate": 0.0006696159160074779,
+      "loss": 2.9931,
+      "step": 7135
+    },
+    {
+      "epoch": 2.255231777619837,
+      "grad_norm": 0.10159403802430324,
+      "learning_rate": 0.0006690970471460209,
+      "loss": 2.9222,
+      "step": 7140
+    },
+    {
+      "epoch": 2.2568111821843164,
+      "grad_norm": 0.11091520440334587,
+      "learning_rate": 0.0006685779725848105,
+      "loss": 2.9502,
+      "step": 7145
+    },
+    {
+      "epoch": 2.2583905867487957,
+      "grad_norm": 0.10768357345688996,
+      "learning_rate": 0.00066805869295528,
+      "loss": 2.9359,
+      "step": 7150
+    },
+    {
+      "epoch": 2.259969991313275,
+      "grad_norm": 0.11998163379469538,
+      "learning_rate": 0.0006675392088891123,
+      "loss": 2.8578,
+      "step": 7155
+    },
+    {
+      "epoch": 2.261549395877754,
+      "grad_norm": 0.13277969388364058,
+      "learning_rate": 0.0006670195210182388,
+      "loss": 2.9115,
+      "step": 7160
+    },
+    {
+      "epoch": 2.2631288004422334,
+      "grad_norm": 0.1755182491420534,
+      "learning_rate": 0.0006664996299748387,
+      "loss": 2.9665,
+      "step": 7165
+    },
+    {
+      "epoch": 2.2647082050067127,
+      "grad_norm": 0.120479369705775,
+      "learning_rate": 0.0006659795363913389,
+      "loss": 2.8196,
+      "step": 7170
+    },
+    {
+      "epoch": 2.266287609571192,
+      "grad_norm": 0.1472740095762117,
+      "learning_rate": 0.000665459240900412,
+      "loss": 2.9059,
+      "step": 7175
+    },
+    {
+      "epoch": 2.2678670141356707,
+      "grad_norm": 0.1403086377292554,
+      "learning_rate": 0.0006649387441349766,
+      "loss": 2.9567,
+      "step": 7180
+    },
+    {
+      "epoch": 2.26944641870015,
+      "grad_norm": 0.11904198908085319,
+      "learning_rate": 0.0006644180467281962,
+      "loss": 2.9466,
+      "step": 7185
+    },
+    {
+      "epoch": 2.271025823264629,
+      "grad_norm": 0.0970900038761183,
+      "learning_rate": 0.0006638971493134782,
+      "loss": 2.8502,
+      "step": 7190
+    },
+    {
+      "epoch": 2.2726052278291085,
+      "grad_norm": 0.11915325203464326,
+      "learning_rate": 0.0006633760525244733,
+      "loss": 2.9136,
+      "step": 7195
+    },
+    {
+      "epoch": 2.2741846323935877,
+      "grad_norm": 0.15596003187860796,
+      "learning_rate": 0.0006628547569950749,
+      "loss": 2.9413,
+      "step": 7200
+    },
+    {
+      "epoch": 2.275764036958067,
+      "grad_norm": 0.14361350912840504,
+      "learning_rate": 0.0006623332633594176,
+      "loss": 2.9854,
+      "step": 7205
+    },
+    {
+      "epoch": 2.2773434415225458,
+      "grad_norm": 0.1387390104010181,
+      "learning_rate": 0.0006618115722518779,
+      "loss": 2.9165,
+      "step": 7210
+    },
+    {
+      "epoch": 2.278922846087025,
+      "grad_norm": 0.12164058631463727,
+      "learning_rate": 0.000661289684307072,
+      "loss": 2.9664,
+      "step": 7215
+    },
+    {
+      "epoch": 2.2805022506515042,
+      "grad_norm": 0.1327686193020361,
+      "learning_rate": 0.0006607676001598552,
+      "loss": 2.8931,
+      "step": 7220
+    },
+    {
+      "epoch": 2.2820816552159835,
+      "grad_norm": 0.09739735935094503,
+      "learning_rate": 0.0006602453204453222,
+      "loss": 2.8827,
+      "step": 7225
+    },
+    {
+      "epoch": 2.2836610597804627,
+      "grad_norm": 0.112259704220055,
+      "learning_rate": 0.0006597228457988053,
+      "loss": 2.9629,
+      "step": 7230
+    },
+    {
+      "epoch": 2.285240464344942,
+      "grad_norm": 0.11511371392998275,
+      "learning_rate": 0.0006592001768558737,
+      "loss": 2.9948,
+      "step": 7235
+    },
+    {
+      "epoch": 2.2868198689094212,
+      "grad_norm": 0.12295746827521585,
+      "learning_rate": 0.000658677314252333,
+      "loss": 2.9594,
+      "step": 7240
+    },
+    {
+      "epoch": 2.2883992734739005,
+      "grad_norm": 0.10003248918300281,
+      "learning_rate": 0.0006581542586242251,
+      "loss": 3.0445,
+      "step": 7245
+    },
+    {
+      "epoch": 2.2899786780383797,
+      "grad_norm": 0.13571954683629917,
+      "learning_rate": 0.0006576310106078255,
+      "loss": 2.9247,
+      "step": 7250
+    },
+    {
+      "epoch": 2.2915580826028585,
+      "grad_norm": 0.10098824916493351,
+      "learning_rate": 0.0006571075708396445,
+      "loss": 2.895,
+      "step": 7255
+    },
+    {
+      "epoch": 2.293137487167338,
+      "grad_norm": 0.11393288835907761,
+      "learning_rate": 0.0006565839399564257,
+      "loss": 2.9283,
+      "step": 7260
+    },
+    {
+      "epoch": 2.294716891731817,
+      "grad_norm": 0.1335459803003351,
+      "learning_rate": 0.000656060118595145,
+      "loss": 2.9307,
+      "step": 7265
+    },
+    {
+      "epoch": 2.2962962962962963,
+      "grad_norm": 0.10505068778995834,
+      "learning_rate": 0.0006555361073930097,
+      "loss": 2.957,
+      "step": 7270
+    },
+    {
+      "epoch": 2.2978757008607755,
+      "grad_norm": 0.12431999416466642,
+      "learning_rate": 0.0006550119069874587,
+      "loss": 3.0146,
+      "step": 7275
+    },
+    {
+      "epoch": 2.2994551054252548,
+      "grad_norm": 0.13026304770658859,
+      "learning_rate": 0.0006544875180161605,
+      "loss": 2.8594,
+      "step": 7280
+    },
+    {
+      "epoch": 2.301034509989734,
+      "grad_norm": 0.16656542361563256,
+      "learning_rate": 0.0006539629411170133,
+      "loss": 2.8679,
+      "step": 7285
+    },
+    {
+      "epoch": 2.302613914554213,
+      "grad_norm": 0.12166231365670094,
+      "learning_rate": 0.0006534381769281437,
+      "loss": 2.9285,
+      "step": 7290
+    },
+    {
+      "epoch": 2.304193319118692,
+      "grad_norm": 0.14190949851348844,
+      "learning_rate": 0.0006529132260879062,
+      "loss": 2.8372,
+      "step": 7295
+    },
+    {
+      "epoch": 2.3057727236831713,
+      "grad_norm": 0.11518608507196358,
+      "learning_rate": 0.0006523880892348823,
+      "loss": 2.9607,
+      "step": 7300
+    },
+    {
+      "epoch": 2.3073521282476506,
+      "grad_norm": 0.15679921245514394,
+      "learning_rate": 0.0006518627670078802,
+      "loss": 2.9361,
+      "step": 7305
+    },
+    {
+      "epoch": 2.30893153281213,
+      "grad_norm": 0.12548306813713037,
+      "learning_rate": 0.0006513372600459329,
+      "loss": 3.0122,
+      "step": 7310
+    },
+    {
+      "epoch": 2.310510937376609,
+      "grad_norm": 0.1211400233474286,
+      "learning_rate": 0.0006508115689882985,
+      "loss": 2.8933,
+      "step": 7315
+    },
+    {
+      "epoch": 2.3120903419410883,
+      "grad_norm": 0.15006322078050777,
+      "learning_rate": 0.0006502856944744593,
+      "loss": 2.878,
+      "step": 7320
+    },
+    {
+      "epoch": 2.3136697465055676,
+      "grad_norm": 0.1319367220228824,
+      "learning_rate": 0.0006497596371441202,
+      "loss": 3.0529,
+      "step": 7325
+    },
+    {
+      "epoch": 2.315249151070047,
+      "grad_norm": 0.13369116216281307,
+      "learning_rate": 0.0006492333976372089,
+      "loss": 2.8999,
+      "step": 7330
+    },
+    {
+      "epoch": 2.3168285556345256,
+      "grad_norm": 0.10719606492938008,
+      "learning_rate": 0.0006487069765938744,
+      "loss": 2.9576,
+      "step": 7335
+    },
+    {
+      "epoch": 2.318407960199005,
+      "grad_norm": 0.1366395020043834,
+      "learning_rate": 0.000648180374654487,
+      "loss": 2.8916,
+      "step": 7340
+    },
+    {
+      "epoch": 2.319987364763484,
+      "grad_norm": 0.16736555472396292,
+      "learning_rate": 0.0006476535924596365,
+      "loss": 2.9229,
+      "step": 7345
+    },
+    {
+      "epoch": 2.3215667693279634,
+      "grad_norm": 0.10854961163919274,
+      "learning_rate": 0.0006471266306501324,
+      "loss": 2.9666,
+      "step": 7350
+    },
+    {
+      "epoch": 2.3231461738924426,
+      "grad_norm": 0.11617884785772647,
+      "learning_rate": 0.0006465994898670027,
+      "loss": 2.8805,
+      "step": 7355
+    },
+    {
+      "epoch": 2.324725578456922,
+      "grad_norm": 0.11730131395930449,
+      "learning_rate": 0.0006460721707514926,
+      "loss": 2.9296,
+      "step": 7360
+    },
+    {
+      "epoch": 2.326304983021401,
+      "grad_norm": 0.1218065285616196,
+      "learning_rate": 0.0006455446739450648,
+      "loss": 2.9287,
+      "step": 7365
+    },
+    {
+      "epoch": 2.32788438758588,
+      "grad_norm": 0.15276746108572395,
+      "learning_rate": 0.0006450170000893978,
+      "loss": 3.0416,
+      "step": 7370
+    },
+    {
+      "epoch": 2.329463792150359,
+      "grad_norm": 0.10988728730323323,
+      "learning_rate": 0.0006444891498263858,
+      "loss": 2.8179,
+      "step": 7375
+    },
+    {
+      "epoch": 2.3310431967148384,
+      "grad_norm": 0.1375312448469524,
+      "learning_rate": 0.0006439611237981373,
+      "loss": 2.8473,
+      "step": 7380
+    },
+    {
+      "epoch": 2.3326226012793176,
+      "grad_norm": 0.12200442362075764,
+      "learning_rate": 0.0006434329226469747,
+      "loss": 2.9678,
+      "step": 7385
+    },
+    {
+      "epoch": 2.334202005843797,
+      "grad_norm": 0.14015923236286695,
+      "learning_rate": 0.0006429045470154333,
+      "loss": 2.824,
+      "step": 7390
+    },
+    {
+      "epoch": 2.335781410408276,
+      "grad_norm": 0.13578554709350246,
+      "learning_rate": 0.0006423759975462611,
+      "loss": 2.8641,
+      "step": 7395
+    },
+    {
+      "epoch": 2.3373608149727554,
+      "grad_norm": 0.127381218523229,
+      "learning_rate": 0.0006418472748824171,
+      "loss": 2.8984,
+      "step": 7400
+    },
+    {
+      "epoch": 2.3389402195372346,
+      "grad_norm": 0.14646884176277938,
+      "learning_rate": 0.0006413183796670713,
+      "loss": 2.91,
+      "step": 7405
+    },
+    {
+      "epoch": 2.340519624101714,
+      "grad_norm": 0.13366509332281643,
+      "learning_rate": 0.0006407893125436031,
+      "loss": 2.9062,
+      "step": 7410
+    },
+    {
+      "epoch": 2.3420990286661927,
+      "grad_norm": 0.10031643403058177,
+      "learning_rate": 0.0006402600741556017,
+      "loss": 2.8832,
+      "step": 7415
+    },
+    {
+      "epoch": 2.343678433230672,
+      "grad_norm": 0.11108377238819567,
+      "learning_rate": 0.000639730665146864,
+      "loss": 2.9613,
+      "step": 7420
+    },
+    {
+      "epoch": 2.345257837795151,
+      "grad_norm": 0.12177336184748958,
+      "learning_rate": 0.0006392010861613951,
+      "loss": 3.1079,
+      "step": 7425
+    },
+    {
+      "epoch": 2.3468372423596304,
+      "grad_norm": 0.12803330627164533,
+      "learning_rate": 0.0006386713378434064,
+      "loss": 2.9418,
+      "step": 7430
+    },
+    {
+      "epoch": 2.3484166469241097,
+      "grad_norm": 0.09879197837614655,
+      "learning_rate": 0.0006381414208373151,
+      "loss": 2.9569,
+      "step": 7435
+    },
+    {
+      "epoch": 2.349996051488589,
+      "grad_norm": 0.09770351588664293,
+      "learning_rate": 0.0006376113357877445,
+      "loss": 2.9496,
+      "step": 7440
+    },
+    {
+      "epoch": 2.351575456053068,
+      "grad_norm": 0.11020473572258073,
+      "learning_rate": 0.0006370810833395213,
+      "loss": 2.8809,
+      "step": 7445
+    },
+    {
+      "epoch": 2.353154860617547,
+      "grad_norm": 0.10817793980080924,
+      "learning_rate": 0.0006365506641376761,
+      "loss": 2.9405,
+      "step": 7450
+    },
+    {
+      "epoch": 2.354734265182026,
+      "grad_norm": 0.11829638710718517,
+      "learning_rate": 0.0006360200788274433,
+      "loss": 2.8903,
+      "step": 7455
+    },
+    {
+      "epoch": 2.3563136697465055,
+      "grad_norm": 0.12122531438161264,
+      "learning_rate": 0.0006354893280542576,
+      "loss": 2.8649,
+      "step": 7460
+    },
+    {
+      "epoch": 2.3578930743109847,
+      "grad_norm": 0.1423364379940213,
+      "learning_rate": 0.0006349584124637568,
+      "loss": 2.8839,
+      "step": 7465
+    },
+    {
+      "epoch": 2.359472478875464,
+      "grad_norm": 0.1271453181188999,
+      "learning_rate": 0.0006344273327017778,
+      "loss": 2.9494,
+      "step": 7470
+    },
+    {
+      "epoch": 2.361051883439943,
+      "grad_norm": 0.11010006128669661,
+      "learning_rate": 0.000633896089414358,
+      "loss": 2.8252,
+      "step": 7475
+    },
+    {
+      "epoch": 2.3626312880044225,
+      "grad_norm": 0.12626868323608306,
+      "learning_rate": 0.0006333646832477333,
+      "loss": 2.9374,
+      "step": 7480
+    },
+    {
+      "epoch": 2.3642106925689017,
+      "grad_norm": 0.098916551615853,
+      "learning_rate": 0.0006328331148483382,
+      "loss": 2.8684,
+      "step": 7485
+    },
+    {
+      "epoch": 2.365790097133381,
+      "grad_norm": 0.1135466487223654,
+      "learning_rate": 0.000632301384862804,
+      "loss": 2.8282,
+      "step": 7490
+    },
+    {
+      "epoch": 2.3673695016978598,
+      "grad_norm": 0.11200927537885891,
+      "learning_rate": 0.000631769493937959,
+      "loss": 2.8443,
+      "step": 7495
+    },
+    {
+      "epoch": 2.368948906262339,
+      "grad_norm": 0.0964859433872292,
+      "learning_rate": 0.000631237442720827,
+      "loss": 2.8649,
+      "step": 7500
+    },
+    {
+      "epoch": 2.3705283108268183,
+      "grad_norm": 0.11815232292417185,
+      "learning_rate": 0.0006307052318586271,
+      "loss": 2.8965,
+      "step": 7505
+    },
+    {
+      "epoch": 2.3721077153912975,
+      "grad_norm": 0.10005956623057873,
+      "learning_rate": 0.0006301728619987721,
+      "loss": 2.9636,
+      "step": 7510
+    },
+    {
+      "epoch": 2.3736871199557767,
+      "grad_norm": 0.13347167160379853,
+      "learning_rate": 0.0006296403337888686,
+      "loss": 2.904,
+      "step": 7515
+    },
+    {
+      "epoch": 2.375266524520256,
+      "grad_norm": 0.12765953630479895,
+      "learning_rate": 0.0006291076478767159,
+      "loss": 2.9464,
+      "step": 7520
+    },
+    {
+      "epoch": 2.3768459290847352,
+      "grad_norm": 0.13646462853539004,
+      "learning_rate": 0.0006285748049103049,
+      "loss": 2.9153,
+      "step": 7525
+    },
+    {
+      "epoch": 2.378425333649214,
+      "grad_norm": 0.13430146770002008,
+      "learning_rate": 0.0006280418055378174,
+      "loss": 2.8988,
+      "step": 7530
+    },
+    {
+      "epoch": 2.3800047382136933,
+      "grad_norm": 0.11154960124006869,
+      "learning_rate": 0.0006275086504076261,
+      "loss": 2.8967,
+      "step": 7535
+    },
+    {
+      "epoch": 2.3815841427781725,
+      "grad_norm": 0.11396262290888928,
+      "learning_rate": 0.0006269753401682924,
+      "loss": 2.9453,
+      "step": 7540
+    },
+    {
+      "epoch": 2.383163547342652,
+      "grad_norm": 0.11939458362440651,
+      "learning_rate": 0.000626441875468567,
+      "loss": 2.9251,
+      "step": 7545
+    },
+    {
+      "epoch": 2.384742951907131,
+      "grad_norm": 0.13204138406120308,
+      "learning_rate": 0.000625908256957388,
+      "loss": 2.8991,
+      "step": 7550
+    },
+    {
+      "epoch": 2.3863223564716103,
+      "grad_norm": 0.09940860549838453,
+      "learning_rate": 0.000625374485283881,
+      "loss": 2.9121,
+      "step": 7555
+    },
+    {
+      "epoch": 2.3879017610360895,
+      "grad_norm": 0.12371146173838096,
+      "learning_rate": 0.0006248405610973579,
+      "loss": 2.8934,
+      "step": 7560
+    },
+    {
+      "epoch": 2.389481165600569,
+      "grad_norm": 0.10512694700874209,
+      "learning_rate": 0.0006243064850473157,
+      "loss": 2.9774,
+      "step": 7565
+    },
+    {
+      "epoch": 2.3910605701650476,
+      "grad_norm": 0.117644926009178,
+      "learning_rate": 0.0006237722577834366,
+      "loss": 3.0284,
+      "step": 7570
+    },
+    {
+      "epoch": 2.392639974729527,
+      "grad_norm": 0.14153952210007853,
+      "learning_rate": 0.0006232378799555866,
+      "loss": 2.967,
+      "step": 7575
+    },
+    {
+      "epoch": 2.394219379294006,
+      "grad_norm": 0.13043671637605017,
+      "learning_rate": 0.0006227033522138145,
+      "loss": 2.9157,
+      "step": 7580
+    },
+    {
+      "epoch": 2.3957987838584853,
+      "grad_norm": 0.13895208613939825,
+      "learning_rate": 0.0006221686752083522,
+      "loss": 3.0146,
+      "step": 7585
+    },
+    {
+      "epoch": 2.3973781884229646,
+      "grad_norm": 0.11330838502160254,
+      "learning_rate": 0.0006216338495896124,
+      "loss": 2.9736,
+      "step": 7590
+    },
+    {
+      "epoch": 2.398957592987444,
+      "grad_norm": 0.11589094914294655,
+      "learning_rate": 0.0006210988760081894,
+      "loss": 2.9057,
+      "step": 7595
+    },
+    {
+      "epoch": 2.400536997551923,
+      "grad_norm": 0.15157965170773122,
+      "learning_rate": 0.0006205637551148567,
+      "loss": 2.9991,
+      "step": 7600
+    },
+    {
+      "epoch": 2.402116402116402,
+      "grad_norm": 0.11320896179503025,
+      "learning_rate": 0.0006200284875605673,
+      "loss": 2.8798,
+      "step": 7605
+    },
+    {
+      "epoch": 2.403695806680881,
+      "grad_norm": 0.13873940669004436,
+      "learning_rate": 0.0006194930739964528,
+      "loss": 2.9625,
+      "step": 7610
+    },
+    {
+      "epoch": 2.4052752112453604,
+      "grad_norm": 0.092705040359977,
+      "learning_rate": 0.0006189575150738223,
+      "loss": 2.913,
+      "step": 7615
+    },
+    {
+      "epoch": 2.4068546158098396,
+      "grad_norm": 0.11139593937524005,
+      "learning_rate": 0.0006184218114441614,
+      "loss": 2.8262,
+      "step": 7620
+    },
+    {
+      "epoch": 2.408434020374319,
+      "grad_norm": 0.1273406414196765,
+      "learning_rate": 0.0006178859637591324,
+      "loss": 2.9821,
+      "step": 7625
+    },
+    {
+      "epoch": 2.410013424938798,
+      "grad_norm": 0.12243860879491518,
+      "learning_rate": 0.000617349972670572,
+      "loss": 2.9659,
+      "step": 7630
+    },
+    {
+      "epoch": 2.4115928295032774,
+      "grad_norm": 0.13916555381820644,
+      "learning_rate": 0.0006168138388304923,
+      "loss": 2.9555,
+      "step": 7635
+    },
+    {
+      "epoch": 2.4131722340677566,
+      "grad_norm": 0.09817635445256855,
+      "learning_rate": 0.0006162775628910781,
+      "loss": 2.8617,
+      "step": 7640
+    },
+    {
+      "epoch": 2.414751638632236,
+      "grad_norm": 0.11641800397697878,
+      "learning_rate": 0.0006157411455046877,
+      "loss": 2.8724,
+      "step": 7645
+    },
+    {
+      "epoch": 2.4163310431967147,
+      "grad_norm": 0.09597429501811984,
+      "learning_rate": 0.0006152045873238512,
+      "loss": 2.8691,
+      "step": 7650
+    },
+    {
+      "epoch": 2.417910447761194,
+      "grad_norm": 0.14373380417854187,
+      "learning_rate": 0.00061466788900127,
+      "loss": 2.865,
+      "step": 7655
+    },
+    {
+      "epoch": 2.419489852325673,
+      "grad_norm": 0.11877441608164054,
+      "learning_rate": 0.0006141310511898161,
+      "loss": 2.9486,
+      "step": 7660
+    },
+    {
+      "epoch": 2.4210692568901524,
+      "grad_norm": 0.1098346159971009,
+      "learning_rate": 0.0006135940745425314,
+      "loss": 2.8958,
+      "step": 7665
+    },
+    {
+      "epoch": 2.4226486614546316,
+      "grad_norm": 0.10845425850178957,
+      "learning_rate": 0.0006130569597126256,
+      "loss": 2.902,
+      "step": 7670
+    },
+    {
+      "epoch": 2.424228066019111,
+      "grad_norm": 0.12665847893533033,
+      "learning_rate": 0.000612519707353478,
+      "loss": 2.9425,
+      "step": 7675
+    },
+    {
+      "epoch": 2.42580747058359,
+      "grad_norm": 0.14520238937055907,
+      "learning_rate": 0.0006119823181186341,
+      "loss": 2.8898,
+      "step": 7680
+    },
+    {
+      "epoch": 2.427386875148069,
+      "grad_norm": 0.1383904256282845,
+      "learning_rate": 0.0006114447926618066,
+      "loss": 2.972,
+      "step": 7685
+    },
+    {
+      "epoch": 2.428966279712548,
+      "grad_norm": 0.12584733089244884,
+      "learning_rate": 0.0006109071316368732,
+      "loss": 2.8755,
+      "step": 7690
+    },
+    {
+      "epoch": 2.4305456842770274,
+      "grad_norm": 0.11062458148204891,
+      "learning_rate": 0.000610369335697877,
+      "loss": 2.9636,
+      "step": 7695
+    },
+    {
+      "epoch": 2.4321250888415067,
+      "grad_norm": 0.14991156013078924,
+      "learning_rate": 0.0006098314054990253,
+      "loss": 2.9908,
+      "step": 7700
+    },
+    {
+      "epoch": 2.433704493405986,
+      "grad_norm": 0.12984481610473728,
+      "learning_rate": 0.0006092933416946885,
+      "loss": 2.9305,
+      "step": 7705
+    },
+    {
+      "epoch": 2.435283897970465,
+      "grad_norm": 0.1174711868276579,
+      "learning_rate": 0.0006087551449393996,
+      "loss": 2.7334,
+      "step": 7710
+    },
+    {
+      "epoch": 2.4368633025349444,
+      "grad_norm": 0.13370455104533868,
+      "learning_rate": 0.000608216815887853,
+      "loss": 2.8389,
+      "step": 7715
+    },
+    {
+      "epoch": 2.4384427070994237,
+      "grad_norm": 0.12216399754359457,
+      "learning_rate": 0.0006076783551949046,
+      "loss": 2.8688,
+      "step": 7720
+    },
+    {
+      "epoch": 2.440022111663903,
+      "grad_norm": 0.1004581598228754,
+      "learning_rate": 0.0006071397635155701,
+      "loss": 2.8522,
+      "step": 7725
+    },
+    {
+      "epoch": 2.4416015162283817,
+      "grad_norm": 0.12995507408416646,
+      "learning_rate": 0.0006066010415050246,
+      "loss": 2.8688,
+      "step": 7730
+    },
+    {
+      "epoch": 2.443180920792861,
+      "grad_norm": 0.12200058647648904,
+      "learning_rate": 0.0006060621898186017,
+      "loss": 2.834,
+      "step": 7735
+    },
+    {
+      "epoch": 2.4447603253573402,
+      "grad_norm": 0.12845547160286597,
+      "learning_rate": 0.0006055232091117929,
+      "loss": 2.8708,
+      "step": 7740
+    },
+    {
+      "epoch": 2.4463397299218195,
+      "grad_norm": 0.13473489215683873,
+      "learning_rate": 0.0006049841000402465,
+      "loss": 2.8791,
+      "step": 7745
+    },
+    {
+      "epoch": 2.4479191344862987,
+      "grad_norm": 0.10406798504227764,
+      "learning_rate": 0.000604444863259767,
+      "loss": 2.8615,
+      "step": 7750
+    },
+    {
+      "epoch": 2.449498539050778,
+      "grad_norm": 0.10339964858007016,
+      "learning_rate": 0.0006039054994263142,
+      "loss": 2.8993,
+      "step": 7755
+    },
+    {
+      "epoch": 2.451077943615257,
+      "grad_norm": 0.09560148890671935,
+      "learning_rate": 0.0006033660091960025,
+      "loss": 2.9023,
+      "step": 7760
+    },
+    {
+      "epoch": 2.452657348179736,
+      "grad_norm": 0.1564609666925628,
+      "learning_rate": 0.0006028263932251,
+      "loss": 2.9065,
+      "step": 7765
+    },
+    {
+      "epoch": 2.4542367527442153,
+      "grad_norm": 0.15595448053044553,
+      "learning_rate": 0.000602286652170028,
+      "loss": 2.9556,
+      "step": 7770
+    },
+    {
+      "epoch": 2.4558161573086945,
+      "grad_norm": 0.13267150580512918,
+      "learning_rate": 0.0006017467866873596,
+      "loss": 2.9264,
+      "step": 7775
+    },
+    {
+      "epoch": 2.4573955618731738,
+      "grad_norm": 0.1389279559900566,
+      "learning_rate": 0.0006012067974338195,
+      "loss": 2.932,
+      "step": 7780
+    },
+    {
+      "epoch": 2.458974966437653,
+      "grad_norm": 0.11281678295707062,
+      "learning_rate": 0.0006006666850662828,
+      "loss": 2.8462,
+      "step": 7785
+    },
+    {
+      "epoch": 2.4605543710021323,
+      "grad_norm": 0.12294466833567588,
+      "learning_rate": 0.0006001264502417748,
+      "loss": 2.8788,
+      "step": 7790
+    },
+    {
+      "epoch": 2.4621337755666115,
+      "grad_norm": 0.11528135968840203,
+      "learning_rate": 0.000599586093617469,
+      "loss": 2.8603,
+      "step": 7795
+    },
+    {
+      "epoch": 2.4637131801310908,
+      "grad_norm": 0.11600416755310684,
+      "learning_rate": 0.0005990456158506878,
+      "loss": 2.8759,
+      "step": 7800
+    },
+    {
+      "epoch": 2.46529258469557,
+      "grad_norm": 0.12584827200322835,
+      "learning_rate": 0.0005985050175989005,
+      "loss": 2.8788,
+      "step": 7805
+    },
+    {
+      "epoch": 2.466871989260049,
+      "grad_norm": 0.10852126172666732,
+      "learning_rate": 0.0005979642995197231,
+      "loss": 2.8374,
+      "step": 7810
+    },
+    {
+      "epoch": 2.468451393824528,
+      "grad_norm": 0.12536981419861587,
+      "learning_rate": 0.0005974234622709173,
+      "loss": 2.8386,
+      "step": 7815
+    },
+    {
+      "epoch": 2.4700307983890073,
+      "grad_norm": 0.10695751249785046,
+      "learning_rate": 0.0005968825065103903,
+      "loss": 2.7978,
+      "step": 7820
+    },
+    {
+      "epoch": 2.4716102029534865,
+      "grad_norm": 0.09098132344627358,
+      "learning_rate": 0.0005963414328961923,
+      "loss": 2.9162,
+      "step": 7825
+    },
+    {
+      "epoch": 2.473189607517966,
+      "grad_norm": 0.1376328674091665,
+      "learning_rate": 0.0005958002420865184,
+      "loss": 2.869,
+      "step": 7830
+    },
+    {
+      "epoch": 2.474769012082445,
+      "grad_norm": 0.1246826926351846,
+      "learning_rate": 0.0005952589347397047,
+      "loss": 2.8681,
+      "step": 7835
+    },
+    {
+      "epoch": 2.4763484166469243,
+      "grad_norm": 0.11223635962115318,
+      "learning_rate": 0.0005947175115142303,
+      "loss": 2.9597,
+      "step": 7840
+    },
+    {
+      "epoch": 2.477927821211403,
+      "grad_norm": 0.09874294074136647,
+      "learning_rate": 0.0005941759730687145,
+      "loss": 2.8881,
+      "step": 7845
+    },
+    {
+      "epoch": 2.4795072257758823,
+      "grad_norm": 0.10702970527144451,
+      "learning_rate": 0.0005936343200619171,
+      "loss": 2.9165,
+      "step": 7850
+    },
+    {
+      "epoch": 2.4810866303403616,
+      "grad_norm": 0.11516417270268137,
+      "learning_rate": 0.0005930925531527373,
+      "loss": 2.9585,
+      "step": 7855
+    },
+    {
+      "epoch": 2.482666034904841,
+      "grad_norm": 0.1021710274042089,
+      "learning_rate": 0.0005925506730002125,
+      "loss": 2.8733,
+      "step": 7860
+    },
+    {
+      "epoch": 2.48424543946932,
+      "grad_norm": 0.10802543141002224,
+      "learning_rate": 0.0005920086802635182,
+      "loss": 2.9297,
+      "step": 7865
+    },
+    {
+      "epoch": 2.4858248440337993,
+      "grad_norm": 0.10653180022851777,
+      "learning_rate": 0.0005914665756019672,
+      "loss": 2.8791,
+      "step": 7870
+    },
+    {
+      "epoch": 2.4874042485982786,
+      "grad_norm": 0.0997210691058446,
+      "learning_rate": 0.0005909243596750072,
+      "loss": 2.9041,
+      "step": 7875
+    },
+    {
+      "epoch": 2.488983653162758,
+      "grad_norm": 0.11878837772321785,
+      "learning_rate": 0.0005903820331422228,
+      "loss": 2.8788,
+      "step": 7880
+    },
+    {
+      "epoch": 2.490563057727237,
+      "grad_norm": 0.11694516600327845,
+      "learning_rate": 0.0005898395966633317,
+      "loss": 2.9453,
+      "step": 7885
+    },
+    {
+      "epoch": 2.492142462291716,
+      "grad_norm": 0.11743602215417963,
+      "learning_rate": 0.0005892970508981866,
+      "loss": 2.9386,
+      "step": 7890
+    },
+    {
+      "epoch": 2.493721866856195,
+      "grad_norm": 0.12746286655196695,
+      "learning_rate": 0.0005887543965067724,
+      "loss": 2.9203,
+      "step": 7895
+    },
+    {
+      "epoch": 2.4953012714206744,
+      "grad_norm": 0.10382797812621399,
+      "learning_rate": 0.0005882116341492063,
+      "loss": 2.9383,
+      "step": 7900
+    },
+    {
+      "epoch": 2.4968806759851536,
+      "grad_norm": 0.1246978749562224,
+      "learning_rate": 0.000587668764485737,
+      "loss": 2.9262,
+      "step": 7905
+    },
+    {
+      "epoch": 2.498460080549633,
+      "grad_norm": 0.09763564074166992,
+      "learning_rate": 0.0005871257881767436,
+      "loss": 2.9287,
+      "step": 7910
+    },
+    {
+      "epoch": 2.500039485114112,
+      "grad_norm": 0.10614550079603412,
+      "learning_rate": 0.0005865827058827344,
+      "loss": 2.8489,
+      "step": 7915
+    },
+    {
+      "epoch": 2.501618889678591,
+      "grad_norm": 0.10538161842564463,
+      "learning_rate": 0.0005860395182643481,
+      "loss": 2.8041,
+      "step": 7920
+    },
+    {
+      "epoch": 2.50319829424307,
+      "grad_norm": 0.1143759722183686,
+      "learning_rate": 0.0005854962259823497,
+      "loss": 2.9817,
+      "step": 7925
+    },
+    {
+      "epoch": 2.5047776988075494,
+      "grad_norm": 0.09584559266623355,
+      "learning_rate": 0.000584952829697633,
+      "loss": 2.8908,
+      "step": 7930
+    },
+    {
+      "epoch": 2.5063571033720287,
+      "grad_norm": 0.11717848072514851,
+      "learning_rate": 0.0005844093300712175,
+      "loss": 2.8875,
+      "step": 7935
+    },
+    {
+      "epoch": 2.507936507936508,
+      "grad_norm": 0.10266982997024632,
+      "learning_rate": 0.0005838657277642484,
+      "loss": 2.8684,
+      "step": 7940
+    },
+    {
+      "epoch": 2.509515912500987,
+      "grad_norm": 0.09002512148417816,
+      "learning_rate": 0.0005833220234379964,
+      "loss": 2.8304,
+      "step": 7945
+    },
+    {
+      "epoch": 2.5110953170654664,
+      "grad_norm": 0.11490604007160053,
+      "learning_rate": 0.0005827782177538558,
+      "loss": 2.9379,
+      "step": 7950
+    },
+    {
+      "epoch": 2.5126747216299457,
+      "grad_norm": 0.13998452649568366,
+      "learning_rate": 0.0005822343113733442,
+      "loss": 2.836,
+      "step": 7955
+    },
+    {
+      "epoch": 2.514254126194425,
+      "grad_norm": 0.1216994535102055,
+      "learning_rate": 0.000581690304958102,
+      "loss": 2.8637,
+      "step": 7960
+    },
+    {
+      "epoch": 2.515833530758904,
+      "grad_norm": 0.09588585536729172,
+      "learning_rate": 0.000581146199169891,
+      "loss": 2.9294,
+      "step": 7965
+    },
+    {
+      "epoch": 2.517412935323383,
+      "grad_norm": 0.0954709796689048,
+      "learning_rate": 0.0005806019946705942,
+      "loss": 2.9174,
+      "step": 7970
+    },
+    {
+      "epoch": 2.518992339887862,
+      "grad_norm": 0.12252355359878117,
+      "learning_rate": 0.0005800576921222142,
+      "loss": 2.8069,
+      "step": 7975
+    },
+    {
+      "epoch": 2.5205717444523414,
+      "grad_norm": 0.10226881145420655,
+      "learning_rate": 0.0005795132921868733,
+      "loss": 3.0369,
+      "step": 7980
+    },
+    {
+      "epoch": 2.5221511490168207,
+      "grad_norm": 0.09563660983490654,
+      "learning_rate": 0.0005789687955268119,
+      "loss": 2.8136,
+      "step": 7985
+    },
+    {
+      "epoch": 2.5237305535813,
+      "grad_norm": 0.11454572894817108,
+      "learning_rate": 0.0005784242028043885,
+      "loss": 2.8711,
+      "step": 7990
+    },
+    {
+      "epoch": 2.525309958145779,
+      "grad_norm": 0.1513588739526145,
+      "learning_rate": 0.0005778795146820783,
+      "loss": 2.9436,
+      "step": 7995
+    },
+    {
+      "epoch": 2.526889362710258,
+      "grad_norm": 0.10466678075616238,
+      "learning_rate": 0.0005773347318224725,
+      "loss": 2.8491,
+      "step": 8000
+    },
+    {
+      "epoch": 2.5284687672747372,
+      "grad_norm": 0.09832765179182253,
+      "learning_rate": 0.0005767898548882772,
+      "loss": 2.8794,
+      "step": 8005
+    },
+    {
+      "epoch": 2.5300481718392165,
+      "grad_norm": 0.09654152645996128,
+      "learning_rate": 0.0005762448845423136,
+      "loss": 2.7809,
+      "step": 8010
+    },
+    {
+      "epoch": 2.5316275764036957,
+      "grad_norm": 0.10270121322864592,
+      "learning_rate": 0.0005756998214475158,
+      "loss": 2.8804,
+      "step": 8015
+    },
+    {
+      "epoch": 2.533206980968175,
+      "grad_norm": 0.12129380452991413,
+      "learning_rate": 0.0005751546662669318,
+      "loss": 2.881,
+      "step": 8020
+    },
+    {
+      "epoch": 2.5347863855326542,
+      "grad_norm": 0.1137491271183352,
+      "learning_rate": 0.0005746094196637202,
+      "loss": 2.9944,
+      "step": 8025
+    },
+    {
+      "epoch": 2.5363657900971335,
+      "grad_norm": 0.1178310752747603,
+      "learning_rate": 0.0005740640823011519,
+      "loss": 2.9511,
+      "step": 8030
+    },
+    {
+      "epoch": 2.5379451946616127,
+      "grad_norm": 0.1057157588808938,
+      "learning_rate": 0.000573518654842608,
+      "loss": 2.8831,
+      "step": 8035
+    },
+    {
+      "epoch": 2.539524599226092,
+      "grad_norm": 0.1319056734423167,
+      "learning_rate": 0.0005729731379515787,
+      "loss": 2.9043,
+      "step": 8040
+    },
+    {
+      "epoch": 2.541104003790571,
+      "grad_norm": 0.14686616590523782,
+      "learning_rate": 0.0005724275322916636,
+      "loss": 2.8638,
+      "step": 8045
+    },
+    {
+      "epoch": 2.54268340835505,
+      "grad_norm": 0.11543070841885159,
+      "learning_rate": 0.00057188183852657,
+      "loss": 2.8652,
+      "step": 8050
+    },
+    {
+      "epoch": 2.5442628129195293,
+      "grad_norm": 0.12972477613439057,
+      "learning_rate": 0.0005713360573201123,
+      "loss": 2.8791,
+      "step": 8055
+    },
+    {
+      "epoch": 2.5458422174840085,
+      "grad_norm": 0.09660035219259314,
+      "learning_rate": 0.0005707901893362115,
+      "loss": 2.9346,
+      "step": 8060
+    },
+    {
+      "epoch": 2.5474216220484878,
+      "grad_norm": 0.10651356816686414,
+      "learning_rate": 0.000570244235238894,
+      "loss": 2.8865,
+      "step": 8065
+    },
+    {
+      "epoch": 2.549001026612967,
+      "grad_norm": 0.09675862774522141,
+      "learning_rate": 0.0005696981956922908,
+      "loss": 2.8773,
+      "step": 8070
+    },
+    {
+      "epoch": 2.5505804311774463,
+      "grad_norm": 0.15617858327168496,
+      "learning_rate": 0.0005691520713606374,
+      "loss": 2.9433,
+      "step": 8075
+    },
+    {
+      "epoch": 2.552159835741925,
+      "grad_norm": 0.11763305081741604,
+      "learning_rate": 0.0005686058629082718,
+      "loss": 2.8832,
+      "step": 8080
+    },
+    {
+      "epoch": 2.5537392403064043,
+      "grad_norm": 0.10593544492397061,
+      "learning_rate": 0.0005680595709996347,
+      "loss": 2.7942,
+      "step": 8085
+    },
+    {
+      "epoch": 2.5553186448708836,
+      "grad_norm": 0.13126937029678293,
+      "learning_rate": 0.0005675131962992684,
+      "loss": 2.986,
+      "step": 8090
+    },
+    {
+      "epoch": 2.556898049435363,
+      "grad_norm": 0.10980128811196688,
+      "learning_rate": 0.0005669667394718151,
+      "loss": 2.8083,
+      "step": 8095
+    },
+    {
+      "epoch": 2.558477453999842,
+      "grad_norm": 0.11819107446791322,
+      "learning_rate": 0.0005664202011820183,
+      "loss": 2.8245,
+      "step": 8100
+    },
+    {
+      "epoch": 2.5600568585643213,
+      "grad_norm": 0.09428546346280126,
+      "learning_rate": 0.0005658735820947195,
+      "loss": 2.7917,
+      "step": 8105
+    },
+    {
+      "epoch": 2.5616362631288006,
+      "grad_norm": 0.09680849514587951,
+      "learning_rate": 0.0005653268828748588,
+      "loss": 2.9514,
+      "step": 8110
+    },
+    {
+      "epoch": 2.56321566769328,
+      "grad_norm": 0.09532956412408176,
+      "learning_rate": 0.0005647801041874738,
+      "loss": 2.8979,
+      "step": 8115
+    },
+    {
+      "epoch": 2.564795072257759,
+      "grad_norm": 0.09313444023713757,
+      "learning_rate": 0.0005642332466976989,
+      "loss": 2.8895,
+      "step": 8120
+    },
+    {
+      "epoch": 2.5663744768222383,
+      "grad_norm": 0.09595409710264817,
+      "learning_rate": 0.000563686311070764,
+      "loss": 2.964,
+      "step": 8125
+    },
+    {
+      "epoch": 2.567953881386717,
+      "grad_norm": 0.10605804176493229,
+      "learning_rate": 0.0005631392979719945,
+      "loss": 2.9001,
+      "step": 8130
+    },
+    {
+      "epoch": 2.5695332859511963,
+      "grad_norm": 0.12266104311606131,
+      "learning_rate": 0.0005625922080668098,
+      "loss": 2.851,
+      "step": 8135
+    },
+    {
+      "epoch": 2.5711126905156756,
+      "grad_norm": 0.1197387065941255,
+      "learning_rate": 0.0005620450420207227,
+      "loss": 2.8976,
+      "step": 8140
+    },
+    {
+      "epoch": 2.572692095080155,
+      "grad_norm": 0.1025070737203573,
+      "learning_rate": 0.0005614978004993388,
+      "loss": 2.9582,
+      "step": 8145
+    },
+    {
+      "epoch": 2.574271499644634,
+      "grad_norm": 0.10279544554870783,
+      "learning_rate": 0.000560950484168355,
+      "loss": 2.9097,
+      "step": 8150
+    },
+    {
+      "epoch": 2.575850904209113,
+      "grad_norm": 0.08338297908944227,
+      "learning_rate": 0.00056040309369356,
+      "loss": 2.932,
+      "step": 8155
+    },
+    {
+      "epoch": 2.577430308773592,
+      "grad_norm": 0.10803053370715439,
+      "learning_rate": 0.0005598556297408321,
+      "loss": 2.8548,
+      "step": 8160
+    },
+    {
+      "epoch": 2.5790097133380714,
+      "grad_norm": 0.11588132304181535,
+      "learning_rate": 0.0005593080929761393,
+      "loss": 2.8645,
+      "step": 8165
+    },
+    {
+      "epoch": 2.5805891179025506,
+      "grad_norm": 0.11089896406663106,
+      "learning_rate": 0.0005587604840655379,
+      "loss": 2.8353,
+      "step": 8170
+    },
+    {
+      "epoch": 2.58216852246703,
+      "grad_norm": 0.11508150368674885,
+      "learning_rate": 0.0005582128036751724,
+      "loss": 2.9319,
+      "step": 8175
+    },
+    {
+      "epoch": 2.583747927031509,
+      "grad_norm": 0.10502399787573738,
+      "learning_rate": 0.0005576650524712735,
+      "loss": 2.9075,
+      "step": 8180
+    },
+    {
+      "epoch": 2.5853273315959884,
+      "grad_norm": 0.13345339841003134,
+      "learning_rate": 0.0005571172311201587,
+      "loss": 2.9161,
+      "step": 8185
+    },
+    {
+      "epoch": 2.5869067361604676,
+      "grad_norm": 0.12456752360732082,
+      "learning_rate": 0.0005565693402882306,
+      "loss": 2.8788,
+      "step": 8190
+    },
+    {
+      "epoch": 2.588486140724947,
+      "grad_norm": 0.14576940938930275,
+      "learning_rate": 0.0005560213806419765,
+      "loss": 2.8721,
+      "step": 8195
+    },
+    {
+      "epoch": 2.590065545289426,
+      "grad_norm": 0.11367083641648577,
+      "learning_rate": 0.0005554733528479672,
+      "loss": 2.8128,
+      "step": 8200
+    },
+    {
+      "epoch": 2.591644949853905,
+      "grad_norm": 0.08885902092672537,
+      "learning_rate": 0.0005549252575728563,
+      "loss": 2.8223,
+      "step": 8205
+    },
+    {
+      "epoch": 2.593224354418384,
+      "grad_norm": 0.08416863087959697,
+      "learning_rate": 0.0005543770954833798,
+      "loss": 2.8385,
+      "step": 8210
+    },
+    {
+      "epoch": 2.5948037589828634,
+      "grad_norm": 0.09346197267458926,
+      "learning_rate": 0.0005538288672463549,
+      "loss": 2.8702,
+      "step": 8215
+    },
+    {
+      "epoch": 2.5963831635473427,
+      "grad_norm": 0.09908019783659593,
+      "learning_rate": 0.000553280573528679,
+      "loss": 2.859,
+      "step": 8220
+    },
+    {
+      "epoch": 2.597962568111822,
+      "grad_norm": 0.11997269336385176,
+      "learning_rate": 0.0005527322149973294,
+      "loss": 2.9347,
+      "step": 8225
+    },
+    {
+      "epoch": 2.599541972676301,
+      "grad_norm": 0.11159273684240642,
+      "learning_rate": 0.0005521837923193621,
+      "loss": 2.8241,
+      "step": 8230
+    },
+    {
+      "epoch": 2.60112137724078,
+      "grad_norm": 0.11072042989007429,
+      "learning_rate": 0.0005516353061619114,
+      "loss": 2.8891,
+      "step": 8235
+    },
+    {
+      "epoch": 2.602700781805259,
+      "grad_norm": 0.10201560853653248,
+      "learning_rate": 0.0005510867571921887,
+      "loss": 2.7999,
+      "step": 8240
+    },
+    {
+      "epoch": 2.6042801863697385,
+      "grad_norm": 0.10667059215170911,
+      "learning_rate": 0.0005505381460774815,
+      "loss": 2.8858,
+      "step": 8245
+    },
+    {
+      "epoch": 2.6058595909342177,
+      "grad_norm": 0.1149422916403453,
+      "learning_rate": 0.0005499894734851533,
+      "loss": 2.924,
+      "step": 8250
+    },
+    {
+      "epoch": 2.607438995498697,
+      "grad_norm": 0.114877786687331,
+      "learning_rate": 0.0005494407400826422,
+      "loss": 2.8901,
+      "step": 8255
+    },
+    {
+      "epoch": 2.609018400063176,
+      "grad_norm": 0.12644713385260586,
+      "learning_rate": 0.0005488919465374601,
+      "loss": 2.9277,
+      "step": 8260
+    },
+    {
+      "epoch": 2.6105978046276554,
+      "grad_norm": 0.09971616604983528,
+      "learning_rate": 0.0005483430935171927,
+      "loss": 2.8212,
+      "step": 8265
+    },
+    {
+      "epoch": 2.6121772091921347,
+      "grad_norm": 0.12772171916187006,
+      "learning_rate": 0.0005477941816894972,
+      "loss": 2.8943,
+      "step": 8270
+    },
+    {
+      "epoch": 2.613756613756614,
+      "grad_norm": 0.1212709812386515,
+      "learning_rate": 0.0005472452117221031,
+      "loss": 2.8858,
+      "step": 8275
+    },
+    {
+      "epoch": 2.615336018321093,
+      "grad_norm": 0.13669330547386427,
+      "learning_rate": 0.00054669618428281,
+      "loss": 2.9107,
+      "step": 8280
+    },
+    {
+      "epoch": 2.616915422885572,
+      "grad_norm": 0.12423297098662968,
+      "learning_rate": 0.0005461471000394877,
+      "loss": 2.855,
+      "step": 8285
+    },
+    {
+      "epoch": 2.6184948274500512,
+      "grad_norm": 0.11719194897621003,
+      "learning_rate": 0.0005455979596600752,
+      "loss": 2.8654,
+      "step": 8290
+    },
+    {
+      "epoch": 2.6200742320145305,
+      "grad_norm": 0.10983667694557864,
+      "learning_rate": 0.0005450487638125798,
+      "loss": 2.9057,
+      "step": 8295
+    },
+    {
+      "epoch": 2.6216536365790097,
+      "grad_norm": 0.08788174740826656,
+      "learning_rate": 0.0005444995131650757,
+      "loss": 2.8627,
+      "step": 8300
+    },
+    {
+      "epoch": 2.623233041143489,
+      "grad_norm": 0.09902778517837978,
+      "learning_rate": 0.0005439502083857048,
+      "loss": 2.7816,
+      "step": 8305
+    },
+    {
+      "epoch": 2.6248124457079682,
+      "grad_norm": 0.11796542187832296,
+      "learning_rate": 0.0005434008501426738,
+      "loss": 2.8523,
+      "step": 8310
+    },
+    {
+      "epoch": 2.626391850272447,
+      "grad_norm": 0.10807870128539734,
+      "learning_rate": 0.0005428514391042551,
+      "loss": 2.8327,
+      "step": 8315
+    },
+    {
+      "epoch": 2.6279712548369263,
+      "grad_norm": 0.12970676311330875,
+      "learning_rate": 0.0005423019759387851,
+      "loss": 2.6965,
+      "step": 8320
+    },
+    {
+      "epoch": 2.6295506594014055,
+      "grad_norm": 0.11144602141340477,
+      "learning_rate": 0.0005417524613146636,
+      "loss": 2.9372,
+      "step": 8325
+    },
+    {
+      "epoch": 2.631130063965885,
+      "grad_norm": 0.14431448741906194,
+      "learning_rate": 0.000541202895900353,
+      "loss": 2.9028,
+      "step": 8330
+    },
+    {
+      "epoch": 2.632709468530364,
+      "grad_norm": 0.13035400654594645,
+      "learning_rate": 0.0005406532803643776,
+      "loss": 2.8818,
+      "step": 8335
+    },
+    {
+      "epoch": 2.6342888730948433,
+      "grad_norm": 0.17885692469148046,
+      "learning_rate": 0.0005401036153753224,
+      "loss": 2.8909,
+      "step": 8340
+    },
+    {
+      "epoch": 2.6358682776593225,
+      "grad_norm": 0.11894763362080127,
+      "learning_rate": 0.000539553901601833,
+      "loss": 2.7765,
+      "step": 8345
+    },
+    {
+      "epoch": 2.6374476822238018,
+      "grad_norm": 0.16610992816103776,
+      "learning_rate": 0.0005390041397126138,
+      "loss": 2.8501,
+      "step": 8350
+    },
+    {
+      "epoch": 2.639027086788281,
+      "grad_norm": 0.10533526116352965,
+      "learning_rate": 0.0005384543303764284,
+      "loss": 2.9377,
+      "step": 8355
+    },
+    {
+      "epoch": 2.6406064913527603,
+      "grad_norm": 0.1175918949534745,
+      "learning_rate": 0.0005379044742620974,
+      "loss": 2.7272,
+      "step": 8360
+    },
+    {
+      "epoch": 2.642185895917239,
+      "grad_norm": 0.10360703818981164,
+      "learning_rate": 0.0005373545720384988,
+      "loss": 2.8597,
+      "step": 8365
+    },
+    {
+      "epoch": 2.6437653004817183,
+      "grad_norm": 0.11568127032482835,
+      "learning_rate": 0.0005368046243745664,
+      "loss": 2.8109,
+      "step": 8370
+    },
+    {
+      "epoch": 2.6453447050461976,
+      "grad_norm": 0.0966862538256532,
+      "learning_rate": 0.0005362546319392895,
+      "loss": 2.8321,
+      "step": 8375
+    },
+    {
+      "epoch": 2.646924109610677,
+      "grad_norm": 0.1036770768690196,
+      "learning_rate": 0.0005357045954017117,
+      "loss": 2.8292,
+      "step": 8380
+    },
+    {
+      "epoch": 2.648503514175156,
+      "grad_norm": 0.08797383785771244,
+      "learning_rate": 0.0005351545154309304,
+      "loss": 2.9117,
+      "step": 8385
+    },
+    {
+      "epoch": 2.6500829187396353,
+      "grad_norm": 0.1140945899524084,
+      "learning_rate": 0.0005346043926960955,
+      "loss": 2.8117,
+      "step": 8390
+    },
+    {
+      "epoch": 2.651662323304114,
+      "grad_norm": 0.10659030191716873,
+      "learning_rate": 0.0005340542278664097,
+      "loss": 2.7991,
+      "step": 8395
+    },
+    {
+      "epoch": 2.6532417278685934,
+      "grad_norm": 0.1333560632282921,
+      "learning_rate": 0.0005335040216111258,
+      "loss": 2.9049,
+      "step": 8400
+    },
+    {
+      "epoch": 2.6548211324330726,
+      "grad_norm": 0.1084321450390365,
+      "learning_rate": 0.000532953774599548,
+      "loss": 2.907,
+      "step": 8405
+    },
+    {
+      "epoch": 2.656400536997552,
+      "grad_norm": 0.10121787566646603,
+      "learning_rate": 0.0005324034875010293,
+      "loss": 2.8406,
+      "step": 8410
+    },
+    {
+      "epoch": 2.657979941562031,
+      "grad_norm": 0.10903625541156818,
+      "learning_rate": 0.0005318531609849721,
+      "loss": 2.8009,
+      "step": 8415
+    },
+    {
+      "epoch": 2.6595593461265103,
+      "grad_norm": 0.09814944875787951,
+      "learning_rate": 0.0005313027957208261,
+      "loss": 2.9122,
+      "step": 8420
+    },
+    {
+      "epoch": 2.6611387506909896,
+      "grad_norm": 0.09769488231656247,
+      "learning_rate": 0.0005307523923780891,
+      "loss": 2.8387,
+      "step": 8425
+    },
+    {
+      "epoch": 2.662718155255469,
+      "grad_norm": 0.08832213539268997,
+      "learning_rate": 0.0005302019516263039,
+      "loss": 2.892,
+      "step": 8430
+    },
+    {
+      "epoch": 2.664297559819948,
+      "grad_norm": 0.10314825876543694,
+      "learning_rate": 0.0005296514741350602,
+      "loss": 2.7589,
+      "step": 8435
+    },
+    {
+      "epoch": 2.6658769643844273,
+      "grad_norm": 0.09340651096278425,
+      "learning_rate": 0.0005291009605739911,
+      "loss": 2.833,
+      "step": 8440
+    },
+    {
+      "epoch": 2.667456368948906,
+      "grad_norm": 0.0934945519190645,
+      "learning_rate": 0.0005285504116127747,
+      "loss": 2.792,
+      "step": 8445
+    },
+    {
+      "epoch": 2.6690357735133854,
+      "grad_norm": 0.08884602620064723,
+      "learning_rate": 0.0005279998279211314,
+      "loss": 2.82,
+      "step": 8450
+    },
+    {
+      "epoch": 2.6706151780778646,
+      "grad_norm": 0.10975324493893444,
+      "learning_rate": 0.0005274492101688241,
+      "loss": 2.8457,
+      "step": 8455
+    },
+    {
+      "epoch": 2.672194582642344,
+      "grad_norm": 0.13695454509385244,
+      "learning_rate": 0.0005268985590256572,
+      "loss": 2.8768,
+      "step": 8460
+    },
+    {
+      "epoch": 2.673773987206823,
+      "grad_norm": 0.12483951207448517,
+      "learning_rate": 0.0005263478751614758,
+      "loss": 2.8488,
+      "step": 8465
+    },
+    {
+      "epoch": 2.6753533917713024,
+      "grad_norm": 0.10388040882976078,
+      "learning_rate": 0.0005257971592461643,
+      "loss": 2.8538,
+      "step": 8470
+    },
+    {
+      "epoch": 2.676932796335781,
+      "grad_norm": 0.11298417140151591,
+      "learning_rate": 0.0005252464119496467,
+      "loss": 2.8867,
+      "step": 8475
+    },
+    {
+      "epoch": 2.6785122009002604,
+      "grad_norm": 0.11629704584959613,
+      "learning_rate": 0.0005246956339418848,
+      "loss": 2.8664,
+      "step": 8480
+    },
+    {
+      "epoch": 2.6800916054647397,
+      "grad_norm": 0.10522974921360309,
+      "learning_rate": 0.0005241448258928781,
+      "loss": 3.015,
+      "step": 8485
+    },
+    {
+      "epoch": 2.681671010029219,
+      "grad_norm": 0.12477955177393571,
+      "learning_rate": 0.0005235939884726624,
+      "loss": 2.8484,
+      "step": 8490
+    },
+    {
+      "epoch": 2.683250414593698,
+      "grad_norm": 0.10411492618738177,
+      "learning_rate": 0.0005230431223513089,
+      "loss": 2.9295,
+      "step": 8495
+    },
+    {
+      "epoch": 2.6848298191581774,
+      "grad_norm": 0.11828492623305749,
+      "learning_rate": 0.0005224922281989245,
+      "loss": 2.8334,
+      "step": 8500
+    },
+    {
+      "epoch": 2.6864092237226567,
+      "grad_norm": 0.11029185894654155,
+      "learning_rate": 0.0005219413066856495,
+      "loss": 2.8465,
+      "step": 8505
+    },
+    {
+      "epoch": 2.687988628287136,
+      "grad_norm": 0.10182149815224686,
+      "learning_rate": 0.0005213903584816578,
+      "loss": 2.8629,
+      "step": 8510
+    },
+    {
+      "epoch": 2.689568032851615,
+      "grad_norm": 0.1128048849382887,
+      "learning_rate": 0.000520839384257156,
+      "loss": 2.9735,
+      "step": 8515
+    },
+    {
+      "epoch": 2.691147437416094,
+      "grad_norm": 0.13170744426484057,
+      "learning_rate": 0.0005202883846823815,
+      "loss": 2.9398,
+      "step": 8520
+    },
+    {
+      "epoch": 2.692726841980573,
+      "grad_norm": 0.09150773720833716,
+      "learning_rate": 0.0005197373604276037,
+      "loss": 2.8829,
+      "step": 8525
+    },
+    {
+      "epoch": 2.6943062465450525,
+      "grad_norm": 0.09975598268082873,
+      "learning_rate": 0.0005191863121631209,
+      "loss": 2.8492,
+      "step": 8530
+    },
+    {
+      "epoch": 2.6958856511095317,
+      "grad_norm": 0.12422940940268365,
+      "learning_rate": 0.0005186352405592617,
+      "loss": 2.9577,
+      "step": 8535
+    },
+    {
+      "epoch": 2.697465055674011,
+      "grad_norm": 0.09818507336107357,
+      "learning_rate": 0.000518084146286382,
+      "loss": 2.9513,
+      "step": 8540
+    },
+    {
+      "epoch": 2.69904446023849,
+      "grad_norm": 0.10155910637986638,
+      "learning_rate": 0.0005175330300148662,
+      "loss": 2.7786,
+      "step": 8545
+    },
+    {
+      "epoch": 2.700623864802969,
+      "grad_norm": 0.11182062536399359,
+      "learning_rate": 0.000516981892415125,
+      "loss": 2.8224,
+      "step": 8550
+    },
+    {
+      "epoch": 2.7022032693674483,
+      "grad_norm": 0.09072312596104798,
+      "learning_rate": 0.0005164307341575949,
+      "loss": 2.9156,
+      "step": 8555
+    },
+    {
+      "epoch": 2.7037826739319275,
+      "grad_norm": 0.09717267367905622,
+      "learning_rate": 0.0005158795559127378,
+      "loss": 2.882,
+      "step": 8560
+    },
+    {
+      "epoch": 2.7053620784964068,
+      "grad_norm": 0.11522288125191604,
+      "learning_rate": 0.00051532835835104,
+      "loss": 2.8246,
+      "step": 8565
+    },
+    {
+      "epoch": 2.706941483060886,
+      "grad_norm": 0.13152596778141204,
+      "learning_rate": 0.0005147771421430112,
+      "loss": 2.9617,
+      "step": 8570
+    },
+    {
+      "epoch": 2.7085208876253652,
+      "grad_norm": 0.11065143019575623,
+      "learning_rate": 0.0005142259079591834,
+      "loss": 2.801,
+      "step": 8575
+    },
+    {
+      "epoch": 2.7101002921898445,
+      "grad_norm": 0.11801107468526441,
+      "learning_rate": 0.0005136746564701112,
+      "loss": 2.862,
+      "step": 8580
+    },
+    {
+      "epoch": 2.7116796967543237,
+      "grad_norm": 0.09717402661556791,
+      "learning_rate": 0.0005131233883463696,
+      "loss": 2.9415,
+      "step": 8585
+    },
+    {
+      "epoch": 2.713259101318803,
+      "grad_norm": 0.10700954717030448,
+      "learning_rate": 0.0005125721042585541,
+      "loss": 2.79,
+      "step": 8590
+    },
+    {
+      "epoch": 2.7148385058832822,
+      "grad_norm": 0.09709107788126108,
+      "learning_rate": 0.0005120208048772799,
+      "loss": 2.8614,
+      "step": 8595
+    },
+    {
+      "epoch": 2.716417910447761,
+      "grad_norm": 0.10416112756350643,
+      "learning_rate": 0.0005114694908731801,
+      "loss": 2.9702,
+      "step": 8600
+    },
+    {
+      "epoch": 2.7179973150122403,
+      "grad_norm": 0.12877731620703553,
+      "learning_rate": 0.0005109181629169063,
+      "loss": 2.8428,
+      "step": 8605
+    },
+    {
+      "epoch": 2.7195767195767195,
+      "grad_norm": 0.12569692543460473,
+      "learning_rate": 0.0005103668216791265,
+      "loss": 2.8394,
+      "step": 8610
+    },
+    {
+      "epoch": 2.721156124141199,
+      "grad_norm": 0.101443859632466,
+      "learning_rate": 0.0005098154678305253,
+      "loss": 2.8178,
+      "step": 8615
+    },
+    {
+      "epoch": 2.722735528705678,
+      "grad_norm": 0.11723589724681906,
+      "learning_rate": 0.0005092641020418026,
+      "loss": 2.8043,
+      "step": 8620
+    },
+    {
+      "epoch": 2.7243149332701573,
+      "grad_norm": 0.09726137109712447,
+      "learning_rate": 0.0005087127249836725,
+      "loss": 2.8191,
+      "step": 8625
+    },
+    {
+      "epoch": 2.725894337834636,
+      "grad_norm": 0.10264792480172079,
+      "learning_rate": 0.000508161337326863,
+      "loss": 2.8913,
+      "step": 8630
+    },
+    {
+      "epoch": 2.7274737423991153,
+      "grad_norm": 0.10925116213376099,
+      "learning_rate": 0.0005076099397421151,
+      "loss": 2.7857,
+      "step": 8635
+    },
+    {
+      "epoch": 2.7290531469635946,
+      "grad_norm": 0.10665161203445127,
+      "learning_rate": 0.0005070585329001819,
+      "loss": 2.8374,
+      "step": 8640
+    },
+    {
+      "epoch": 2.730632551528074,
+      "grad_norm": 0.09831191236696475,
+      "learning_rate": 0.0005065071174718272,
+      "loss": 2.8332,
+      "step": 8645
+    },
+    {
+      "epoch": 2.732211956092553,
+      "grad_norm": 0.09194761295937094,
+      "learning_rate": 0.0005059556941278261,
+      "loss": 2.8834,
+      "step": 8650
+    },
+    {
+      "epoch": 2.7337913606570323,
+      "grad_norm": 0.10405004093466642,
+      "learning_rate": 0.0005054042635389627,
+      "loss": 2.7754,
+      "step": 8655
+    },
+    {
+      "epoch": 2.7353707652215116,
+      "grad_norm": 0.14537068141239967,
+      "learning_rate": 0.0005048528263760301,
+      "loss": 2.7946,
+      "step": 8660
+    },
+    {
+      "epoch": 2.736950169785991,
+      "grad_norm": 0.1065988827208742,
+      "learning_rate": 0.0005043013833098296,
+      "loss": 2.8628,
+      "step": 8665
+    },
+    {
+      "epoch": 2.73852957435047,
+      "grad_norm": 0.1195366408617911,
+      "learning_rate": 0.0005037499350111692,
+      "loss": 2.9437,
+      "step": 8670
+    },
+    {
+      "epoch": 2.7401089789149493,
+      "grad_norm": 0.11459702577743501,
+      "learning_rate": 0.000503198482150864,
+      "loss": 2.7711,
+      "step": 8675
+    },
+    {
+      "epoch": 2.741688383479428,
+      "grad_norm": 0.12358540930509837,
+      "learning_rate": 0.0005026470253997339,
+      "loss": 2.9472,
+      "step": 8680
+    },
+    {
+      "epoch": 2.7432677880439074,
+      "grad_norm": 0.11603284868223569,
+      "learning_rate": 0.0005020955654286038,
+      "loss": 2.8495,
+      "step": 8685
+    },
+    {
+      "epoch": 2.7448471926083866,
+      "grad_norm": 0.08951728738117785,
+      "learning_rate": 0.0005015441029083029,
+      "loss": 2.9477,
+      "step": 8690
+    },
+    {
+      "epoch": 2.746426597172866,
+      "grad_norm": 0.10244013458323355,
+      "learning_rate": 0.0005009926385096627,
+      "loss": 2.8319,
+      "step": 8695
+    },
+    {
+      "epoch": 2.748006001737345,
+      "grad_norm": 0.12482940143306988,
+      "learning_rate": 0.0005004411729035179,
+      "loss": 2.9043,
+      "step": 8700
+    },
+    {
+      "epoch": 2.7495854063018244,
+      "grad_norm": 0.10452890378293654,
+      "learning_rate": 0.0004998897067607039,
+      "loss": 2.8477,
+      "step": 8705
+    },
+    {
+      "epoch": 2.751164810866303,
+      "grad_norm": 0.12090847404143575,
+      "learning_rate": 0.0004993382407520573,
+      "loss": 2.7886,
+      "step": 8710
+    },
+    {
+      "epoch": 2.7527442154307824,
+      "grad_norm": 0.08178133630175746,
+      "learning_rate": 0.0004987867755484141,
+      "loss": 2.8267,
+      "step": 8715
+    },
+    {
+      "epoch": 2.7543236199952617,
+      "grad_norm": 0.10062336680780973,
+      "learning_rate": 0.0004982353118206095,
+      "loss": 2.7693,
+      "step": 8720
+    },
+    {
+      "epoch": 2.755903024559741,
+      "grad_norm": 0.0943113197343758,
+      "learning_rate": 0.0004976838502394772,
+      "loss": 2.8277,
+      "step": 8725
+    },
+    {
+      "epoch": 2.75748242912422,
+      "grad_norm": 0.11023713286267932,
+      "learning_rate": 0.0004971323914758479,
+      "loss": 2.7993,
+      "step": 8730
+    },
+    {
+      "epoch": 2.7590618336886994,
+      "grad_norm": 0.1073939345323795,
+      "learning_rate": 0.000496580936200549,
+      "loss": 2.8793,
+      "step": 8735
+    },
+    {
+      "epoch": 2.7606412382531786,
+      "grad_norm": 0.11471023136135335,
+      "learning_rate": 0.0004960294850844036,
+      "loss": 2.7634,
+      "step": 8740
+    },
+    {
+      "epoch": 2.762220642817658,
+      "grad_norm": 0.12262767443184312,
+      "learning_rate": 0.0004954780387982296,
+      "loss": 2.8035,
+      "step": 8745
+    },
+    {
+      "epoch": 2.763800047382137,
+      "grad_norm": 0.131867203858333,
+      "learning_rate": 0.0004949265980128398,
+      "loss": 2.8725,
+      "step": 8750
+    },
+    {
+      "epoch": 2.7653794519466164,
+      "grad_norm": 0.10633571661017398,
+      "learning_rate": 0.0004943751633990392,
+      "loss": 2.9279,
+      "step": 8755
+    },
+    {
+      "epoch": 2.766958856511095,
+      "grad_norm": 0.10209093832053376,
+      "learning_rate": 0.0004938237356276261,
+      "loss": 2.8331,
+      "step": 8760
+    },
+    {
+      "epoch": 2.7685382610755744,
+      "grad_norm": 0.12140728220481119,
+      "learning_rate": 0.0004932723153693899,
+      "loss": 2.8959,
+      "step": 8765
+    },
+    {
+      "epoch": 2.7701176656400537,
+      "grad_norm": 0.11568062669311602,
+      "learning_rate": 0.0004927209032951113,
+      "loss": 2.8911,
+      "step": 8770
+    },
+    {
+      "epoch": 2.771697070204533,
+      "grad_norm": 0.09419916374966063,
+      "learning_rate": 0.000492169500075561,
+      "loss": 2.9308,
+      "step": 8775
+    },
+    {
+      "epoch": 2.773276474769012,
+      "grad_norm": 0.10783677082464539,
+      "learning_rate": 0.0004916181063814989,
+      "loss": 2.862,
+      "step": 8780
+    },
+    {
+      "epoch": 2.7748558793334914,
+      "grad_norm": 0.09993678976761282,
+      "learning_rate": 0.0004910667228836729,
+      "loss": 2.9124,
+      "step": 8785
+    },
+    {
+      "epoch": 2.7764352838979702,
+      "grad_norm": 0.09810924138432092,
+      "learning_rate": 0.0004905153502528192,
+      "loss": 2.8774,
+      "step": 8790
+    },
+    {
+      "epoch": 2.7780146884624495,
+      "grad_norm": 0.0882822450050144,
+      "learning_rate": 0.00048996398915966,
+      "loss": 2.8996,
+      "step": 8795
+    },
+    {
+      "epoch": 2.7795940930269287,
+      "grad_norm": 0.11509448805364654,
+      "learning_rate": 0.0004894126402749044,
+      "loss": 2.8937,
+      "step": 8800
+    },
+    {
+      "epoch": 2.781173497591408,
+      "grad_norm": 0.12405467965417506,
+      "learning_rate": 0.0004888613042692457,
+      "loss": 2.8048,
+      "step": 8805
+    },
+    {
+      "epoch": 2.782752902155887,
+      "grad_norm": 0.09425134572556984,
+      "learning_rate": 0.0004883099818133624,
+      "loss": 2.8566,
+      "step": 8810
+    },
+    {
+      "epoch": 2.7843323067203665,
+      "grad_norm": 0.09041726145112262,
+      "learning_rate": 0.0004877586735779156,
+      "loss": 2.821,
+      "step": 8815
+    },
+    {
+      "epoch": 2.7859117112848457,
+      "grad_norm": 0.08762216365221542,
+      "learning_rate": 0.00048720738023354986,
+      "loss": 2.9121,
+      "step": 8820
+    },
+    {
+      "epoch": 2.787491115849325,
+      "grad_norm": 0.0976114655616353,
+      "learning_rate": 0.00048665610245089106,
+      "loss": 2.8665,
+      "step": 8825
+    },
+    {
+      "epoch": 2.789070520413804,
+      "grad_norm": 0.11039288980681734,
+      "learning_rate": 0.00048610484090054695,
+      "loss": 2.856,
+      "step": 8830
+    },
+    {
+      "epoch": 2.7906499249782835,
+      "grad_norm": 0.10708824253884887,
+      "learning_rate": 0.0004855535962531046,
+      "loss": 2.8531,
+      "step": 8835
+    },
+    {
+      "epoch": 2.7922293295427623,
+      "grad_norm": 0.1195136121465101,
+      "learning_rate": 0.0004850023691791313,
+      "loss": 2.8933,
+      "step": 8840
+    },
+    {
+      "epoch": 2.7938087341072415,
+      "grad_norm": 0.11014538299717265,
+      "learning_rate": 0.0004844511603491722,
+      "loss": 2.9101,
+      "step": 8845
+    },
+    {
+      "epoch": 2.7953881386717208,
+      "grad_norm": 0.10473197842297582,
+      "learning_rate": 0.0004838999704337507,
+      "loss": 2.7816,
+      "step": 8850
+    },
+    {
+      "epoch": 2.7969675432362,
+      "grad_norm": 0.0894809130323671,
+      "learning_rate": 0.00048334880010336744,
+      "loss": 2.8218,
+      "step": 8855
+    },
+    {
+      "epoch": 2.7985469478006793,
+      "grad_norm": 0.09784437855539298,
+      "learning_rate": 0.00048279765002849894,
+      "loss": 2.8752,
+      "step": 8860
+    },
+    {
+      "epoch": 2.8001263523651585,
+      "grad_norm": 0.13067758045382105,
+      "learning_rate": 0.00048224652087959686,
+      "loss": 2.8194,
+      "step": 8865
+    },
+    {
+      "epoch": 2.8017057569296373,
+      "grad_norm": 0.0986468551876891,
+      "learning_rate": 0.0004816954133270878,
+      "loss": 2.963,
+      "step": 8870
+    },
+    {
+      "epoch": 2.8032851614941166,
+      "grad_norm": 0.12143975478222904,
+      "learning_rate": 0.0004811443280413716,
+      "loss": 2.7702,
+      "step": 8875
+    },
+    {
+      "epoch": 2.804864566058596,
+      "grad_norm": 0.09920981393173242,
+      "learning_rate": 0.0004805932656928218,
+      "loss": 2.893,
+      "step": 8880
+    },
+    {
+      "epoch": 2.806443970623075,
+      "grad_norm": 0.1074725681810456,
+      "learning_rate": 0.0004800422269517833,
+      "loss": 2.9134,
+      "step": 8885
+    },
+    {
+      "epoch": 2.8080233751875543,
+      "grad_norm": 0.10204281091380536,
+      "learning_rate": 0.00047949121248857277,
+      "loss": 2.8876,
+      "step": 8890
+    },
+    {
+      "epoch": 2.8096027797520335,
+      "grad_norm": 0.09270434755456333,
+      "learning_rate": 0.0004789402229734768,
+      "loss": 2.8664,
+      "step": 8895
+    },
+    {
+      "epoch": 2.811182184316513,
+      "grad_norm": 0.10166396255180883,
+      "learning_rate": 0.000478389259076752,
+      "loss": 2.8878,
+      "step": 8900
+    },
+    {
+      "epoch": 2.812761588880992,
+      "grad_norm": 0.1208537911429681,
+      "learning_rate": 0.00047783832146862403,
+      "loss": 2.8745,
+      "step": 8905
+    },
+    {
+      "epoch": 2.8143409934454713,
+      "grad_norm": 0.14123581518424533,
+      "learning_rate": 0.0004772874108192863,
+      "loss": 2.8131,
+      "step": 8910
+    },
+    {
+      "epoch": 2.81592039800995,
+      "grad_norm": 0.1263482735241249,
+      "learning_rate": 0.0004767365277988993,
+      "loss": 2.8347,
+      "step": 8915
+    },
+    {
+      "epoch": 2.8174998025744293,
+      "grad_norm": 0.10619042341997327,
+      "learning_rate": 0.0004761856730775902,
+      "loss": 2.7617,
+      "step": 8920
+    },
+    {
+      "epoch": 2.8190792071389086,
+      "grad_norm": 0.10005343105322269,
+      "learning_rate": 0.0004756348473254513,
+      "loss": 2.7529,
+      "step": 8925
+    },
+    {
+      "epoch": 2.820658611703388,
+      "grad_norm": 0.09257602942214689,
+      "learning_rate": 0.0004750840512125403,
+      "loss": 2.9872,
+      "step": 8930
+    },
+    {
+      "epoch": 2.822238016267867,
+      "grad_norm": 0.11013323338667516,
+      "learning_rate": 0.0004745332854088783,
+      "loss": 2.8541,
+      "step": 8935
+    },
+    {
+      "epoch": 2.8238174208323463,
+      "grad_norm": 0.11218466959923673,
+      "learning_rate": 0.00047398255058444996,
+      "loss": 2.8551,
+      "step": 8940
+    },
+    {
+      "epoch": 2.825396825396825,
+      "grad_norm": 0.10032747092494256,
+      "learning_rate": 0.0004734318474092018,
+      "loss": 2.839,
+      "step": 8945
+    },
+    {
+      "epoch": 2.8269762299613044,
+      "grad_norm": 0.09300804561700793,
+      "learning_rate": 0.000472881176553042,
+      "loss": 2.8955,
+      "step": 8950
+    },
+    {
+      "epoch": 2.8285556345257836,
+      "grad_norm": 0.11698869284218402,
+      "learning_rate": 0.0004723305386858399,
+      "loss": 2.764,
+      "step": 8955
+    },
+    {
+      "epoch": 2.830135039090263,
+      "grad_norm": 0.12002117606412845,
+      "learning_rate": 0.000471779934477424,
+      "loss": 2.8536,
+      "step": 8960
+    },
+    {
+      "epoch": 2.831714443654742,
+      "grad_norm": 0.09323043603377293,
+      "learning_rate": 0.00047122936459758225,
+      "loss": 2.7864,
+      "step": 8965
+    },
+    {
+      "epoch": 2.8332938482192214,
+      "grad_norm": 0.1002912090989643,
+      "learning_rate": 0.0004706788297160608,
+      "loss": 2.7198,
+      "step": 8970
+    },
+    {
+      "epoch": 2.8348732527837006,
+      "grad_norm": 0.1475023105666348,
+      "learning_rate": 0.00047012833050256287,
+      "loss": 2.8075,
+      "step": 8975
+    },
+    {
+      "epoch": 2.83645265734818,
+      "grad_norm": 0.11276908138577017,
+      "learning_rate": 0.000469577867626749,
+      "loss": 2.8639,
+      "step": 8980
+    },
+    {
+      "epoch": 2.838032061912659,
+      "grad_norm": 0.21161327588210527,
+      "learning_rate": 0.0004690274417582349,
+      "loss": 2.7783,
+      "step": 8985
+    },
+    {
+      "epoch": 2.8396114664771384,
+      "grad_norm": 0.10800754485859902,
+      "learning_rate": 0.0004684770535665917,
+      "loss": 2.8841,
+      "step": 8990
+    },
+    {
+      "epoch": 2.841190871041617,
+      "grad_norm": 0.10269668428994914,
+      "learning_rate": 0.0004679267037213443,
+      "loss": 2.818,
+      "step": 8995
+    },
+    {
+      "epoch": 2.8427702756060964,
+      "grad_norm": 0.08507380551575273,
+      "learning_rate": 0.0004673763928919712,
+      "loss": 2.8349,
+      "step": 9000
+    },
+    {
+      "epoch": 2.8443496801705757,
+      "grad_norm": 0.08736533693565444,
+      "learning_rate": 0.0004668261217479032,
+      "loss": 2.7935,
+      "step": 9005
+    },
+    {
+      "epoch": 2.845929084735055,
+      "grad_norm": 0.1114869278836615,
+      "learning_rate": 0.0004662758909585233,
+      "loss": 2.8563,
+      "step": 9010
+    },
+    {
+      "epoch": 2.847508489299534,
+      "grad_norm": 0.10751009609851755,
+      "learning_rate": 0.00046572570119316495,
+      "loss": 2.8114,
+      "step": 9015
+    },
+    {
+      "epoch": 2.8490878938640134,
+      "grad_norm": 0.08412230037944082,
+      "learning_rate": 0.000465175553121112,
+      "loss": 2.7261,
+      "step": 9020
+    },
+    {
+      "epoch": 2.850667298428492,
+      "grad_norm": 0.08604848886254374,
+      "learning_rate": 0.0004646254474115973,
+      "loss": 2.825,
+      "step": 9025
+    },
+    {
+      "epoch": 2.8522467029929714,
+      "grad_norm": 0.09052552835236849,
+      "learning_rate": 0.00046407538473380215,
+      "loss": 2.909,
+      "step": 9030
+    },
+    {
+      "epoch": 2.8538261075574507,
+      "grad_norm": 0.08449371690829786,
+      "learning_rate": 0.0004635253657568561,
+      "loss": 2.9414,
+      "step": 9035
+    },
+    {
+      "epoch": 2.85540551212193,
+      "grad_norm": 0.09900171582846604,
+      "learning_rate": 0.00046297539114983476,
+      "loss": 2.8186,
+      "step": 9040
+    },
+    {
+      "epoch": 2.856984916686409,
+      "grad_norm": 0.11936665449629383,
+      "learning_rate": 0.00046242546158176026,
+      "loss": 2.8067,
+      "step": 9045
+    },
+    {
+      "epoch": 2.8585643212508884,
+      "grad_norm": 0.10878964794862939,
+      "learning_rate": 0.0004618755777215998,
+      "loss": 2.8312,
+      "step": 9050
+    },
+    {
+      "epoch": 2.8601437258153677,
+      "grad_norm": 0.10594367710493519,
+      "learning_rate": 0.0004613257402382647,
+      "loss": 2.7821,
+      "step": 9055
+    },
+    {
+      "epoch": 2.861723130379847,
+      "grad_norm": 0.10774059810851142,
+      "learning_rate": 0.0004607759498006104,
+      "loss": 2.9061,
+      "step": 9060
+    },
+    {
+      "epoch": 2.863302534944326,
+      "grad_norm": 0.10285755907368245,
+      "learning_rate": 0.000460226207077435,
+      "loss": 2.8355,
+      "step": 9065
+    },
+    {
+      "epoch": 2.8648819395088054,
+      "grad_norm": 0.10351095774021554,
+      "learning_rate": 0.000459676512737478,
+      "loss": 2.8312,
+      "step": 9070
+    },
+    {
+      "epoch": 2.8664613440732842,
+      "grad_norm": 0.09887969502613489,
+      "learning_rate": 0.0004591268674494207,
+      "loss": 2.7871,
+      "step": 9075
+    },
+    {
+      "epoch": 2.8680407486377635,
+      "grad_norm": 0.10901193731190399,
+      "learning_rate": 0.00045857727188188425,
+      "loss": 2.8475,
+      "step": 9080
+    },
+    {
+      "epoch": 2.8696201532022427,
+      "grad_norm": 0.10770970500201059,
+      "learning_rate": 0.0004580277267034299,
+      "loss": 2.8679,
+      "step": 9085
+    },
+    {
+      "epoch": 2.871199557766722,
+      "grad_norm": 0.10326053968677797,
+      "learning_rate": 0.000457478232582557,
+      "loss": 2.7711,
+      "step": 9090
+    },
+    {
+      "epoch": 2.8727789623312012,
+      "grad_norm": 0.09136407583935778,
+      "learning_rate": 0.000456928790187703,
+      "loss": 2.8492,
+      "step": 9095
+    },
+    {
+      "epoch": 2.8743583668956805,
+      "grad_norm": 0.08441720449514221,
+      "learning_rate": 0.00045637940018724275,
+      "loss": 2.9713,
+      "step": 9100
+    },
+    {
+      "epoch": 2.8759377714601593,
+      "grad_norm": 0.09899362460145938,
+      "learning_rate": 0.00045583006324948654,
+      "loss": 2.8082,
+      "step": 9105
+    },
+    {
+      "epoch": 2.8775171760246385,
+      "grad_norm": 0.10434826649888167,
+      "learning_rate": 0.0004552807800426812,
+      "loss": 2.7625,
+      "step": 9110
+    },
+    {
+      "epoch": 2.8790965805891178,
+      "grad_norm": 0.09849217908206793,
+      "learning_rate": 0.0004547315512350075,
+      "loss": 2.7703,
+      "step": 9115
+    },
+    {
+      "epoch": 2.880675985153597,
+      "grad_norm": 0.0899109070075441,
+      "learning_rate": 0.00045418237749457994,
+      "loss": 2.7332,
+      "step": 9120
+    },
+    {
+      "epoch": 2.8822553897180763,
+      "grad_norm": 0.09411693700074429,
+      "learning_rate": 0.0004536332594894466,
+      "loss": 2.8315,
+      "step": 9125
+    },
+    {
+      "epoch": 2.8838347942825555,
+      "grad_norm": 0.11307082588068913,
+      "learning_rate": 0.00045308419788758704,
+      "loss": 2.8794,
+      "step": 9130
+    },
+    {
+      "epoch": 2.8854141988470348,
+      "grad_norm": 0.12481609564961511,
+      "learning_rate": 0.00045253519335691306,
+      "loss": 2.8332,
+      "step": 9135
+    },
+    {
+      "epoch": 2.886993603411514,
+      "grad_norm": 0.11526490471981808,
+      "learning_rate": 0.00045198624656526634,
+      "loss": 2.8797,
+      "step": 9140
+    },
+    {
+      "epoch": 2.8885730079759933,
+      "grad_norm": 0.08616360945328765,
+      "learning_rate": 0.0004514373581804187,
+      "loss": 2.8677,
+      "step": 9145
+    },
+    {
+      "epoch": 2.8901524125404725,
+      "grad_norm": 0.09639886901063495,
+      "learning_rate": 0.00045088852887007055,
+      "loss": 2.8371,
+      "step": 9150
+    },
+    {
+      "epoch": 2.8917318171049513,
+      "grad_norm": 0.09426024023747775,
+      "learning_rate": 0.00045033975930185074,
+      "loss": 2.8256,
+      "step": 9155
+    },
+    {
+      "epoch": 2.8933112216694306,
+      "grad_norm": 0.09823088057499815,
+      "learning_rate": 0.0004497910501433153,
+      "loss": 2.7695,
+      "step": 9160
+    },
+    {
+      "epoch": 2.89489062623391,
+      "grad_norm": 0.09651129158000497,
+      "learning_rate": 0.0004492424020619472,
+      "loss": 2.8388,
+      "step": 9165
+    },
+    {
+      "epoch": 2.896470030798389,
+      "grad_norm": 0.11108150616205786,
+      "learning_rate": 0.0004486938157251543,
+      "loss": 2.824,
+      "step": 9170
+    },
+    {
+      "epoch": 2.8980494353628683,
+      "grad_norm": 0.10056751400490413,
+      "learning_rate": 0.00044814529180027025,
+      "loss": 2.7821,
+      "step": 9175
+    },
+    {
+      "epoch": 2.8996288399273475,
+      "grad_norm": 0.10387820407099511,
+      "learning_rate": 0.0004475968309545519,
+      "loss": 2.8305,
+      "step": 9180
+    },
+    {
+      "epoch": 2.9012082444918263,
+      "grad_norm": 0.10673816323056329,
+      "learning_rate": 0.0004470484338551799,
+      "loss": 2.7528,
+      "step": 9185
+    },
+    {
+      "epoch": 2.9027876490563056,
+      "grad_norm": 0.10360115384761291,
+      "learning_rate": 0.00044650010116925744,
+      "loss": 2.8434,
+      "step": 9190
+    },
+    {
+      "epoch": 2.904367053620785,
+      "grad_norm": 0.11180438845853563,
+      "learning_rate": 0.00044595183356380916,
+      "loss": 2.8672,
+      "step": 9195
+    },
+    {
+      "epoch": 2.905946458185264,
+      "grad_norm": 0.0994512434438589,
+      "learning_rate": 0.0004454036317057803,
+      "loss": 2.7964,
+      "step": 9200
+    },
+    {
+      "epoch": 2.9075258627497433,
+      "grad_norm": 0.15167604489166178,
+      "learning_rate": 0.00044485549626203653,
+      "loss": 2.8185,
+      "step": 9205
+    },
+    {
+      "epoch": 2.9091052673142226,
+      "grad_norm": 0.09398806231071076,
+      "learning_rate": 0.00044430742789936244,
+      "loss": 2.8173,
+      "step": 9210
+    },
+    {
+      "epoch": 2.910684671878702,
+      "grad_norm": 0.10992730250826928,
+      "learning_rate": 0.00044375942728446145,
+      "loss": 2.8127,
+      "step": 9215
+    },
+    {
+      "epoch": 2.912264076443181,
+      "grad_norm": 0.13652749666712283,
+      "learning_rate": 0.0004432114950839539,
+      "loss": 2.891,
+      "step": 9220
+    },
+    {
+      "epoch": 2.9138434810076603,
+      "grad_norm": 0.10436724895098912,
+      "learning_rate": 0.00044266363196437757,
+      "loss": 2.8738,
+      "step": 9225
+    },
+    {
+      "epoch": 2.9154228855721396,
+      "grad_norm": 0.11944492929699857,
+      "learning_rate": 0.0004421158385921856,
+      "loss": 2.9314,
+      "step": 9230
+    },
+    {
+      "epoch": 2.9170022901366184,
+      "grad_norm": 0.11991906724849831,
+      "learning_rate": 0.0004415681156337466,
+      "loss": 2.8207,
+      "step": 9235
+    },
+    {
+      "epoch": 2.9185816947010976,
+      "grad_norm": 0.10846283055092165,
+      "learning_rate": 0.0004410204637553437,
+      "loss": 2.8969,
+      "step": 9240
+    },
+    {
+      "epoch": 2.920161099265577,
+      "grad_norm": 0.12047046083644518,
+      "learning_rate": 0.00044047288362317346,
+      "loss": 2.9509,
+      "step": 9245
+    },
+    {
+      "epoch": 2.921740503830056,
+      "grad_norm": 0.0980365989292603,
+      "learning_rate": 0.00043992537590334483,
+      "loss": 2.8516,
+      "step": 9250
+    },
+    {
+      "epoch": 2.9233199083945354,
+      "grad_norm": 0.11669989052911783,
+      "learning_rate": 0.000439377941261879,
+      "loss": 3.0115,
+      "step": 9255
+    },
+    {
+      "epoch": 2.924899312959014,
+      "grad_norm": 0.11700862846287295,
+      "learning_rate": 0.0004388305803647079,
+      "loss": 2.8834,
+      "step": 9260
+    },
+    {
+      "epoch": 2.9264787175234934,
+      "grad_norm": 0.11928964488487814,
+      "learning_rate": 0.0004382832938776747,
+      "loss": 2.9026,
+      "step": 9265
+    },
+    {
+      "epoch": 2.9280581220879727,
+      "grad_norm": 0.09061663401562066,
+      "learning_rate": 0.0004377360824665309,
+      "loss": 2.8714,
+      "step": 9270
+    },
+    {
+      "epoch": 2.929637526652452,
+      "grad_norm": 0.1143067185517572,
+      "learning_rate": 0.0004371889467969373,
+      "loss": 2.8862,
+      "step": 9275
+    },
+    {
+      "epoch": 2.931216931216931,
+      "grad_norm": 0.11881429925573886,
+      "learning_rate": 0.00043664188753446236,
+      "loss": 2.9086,
+      "step": 9280
+    },
+    {
+      "epoch": 2.9327963357814104,
+      "grad_norm": 0.10121970992950821,
+      "learning_rate": 0.0004360949053445816,
+      "loss": 2.8306,
+      "step": 9285
+    },
+    {
+      "epoch": 2.9343757403458897,
+      "grad_norm": 0.10582419788273659,
+      "learning_rate": 0.000435548000892677,
+      "loss": 2.7933,
+      "step": 9290
+    },
+    {
+      "epoch": 2.935955144910369,
+      "grad_norm": 0.08778865885627898,
+      "learning_rate": 0.00043500117484403586,
+      "loss": 2.7393,
+      "step": 9295
+    },
+    {
+      "epoch": 2.937534549474848,
+      "grad_norm": 0.09920218371219021,
+      "learning_rate": 0.00043445442786384984,
+      "loss": 2.9221,
+      "step": 9300
+    },
+    {
+      "epoch": 2.9391139540393274,
+      "grad_norm": 0.09361706009393304,
+      "learning_rate": 0.0004339077606172149,
+      "loss": 2.892,
+      "step": 9305
+    },
+    {
+      "epoch": 2.940693358603806,
+      "grad_norm": 0.09808388861515303,
+      "learning_rate": 0.0004333611737691295,
+      "loss": 2.7733,
+      "step": 9310
+    },
+    {
+      "epoch": 2.9422727631682855,
+      "grad_norm": 0.08103140627641786,
+      "learning_rate": 0.00043281466798449455,
+      "loss": 2.8325,
+      "step": 9315
+    },
+    {
+      "epoch": 2.9438521677327647,
+      "grad_norm": 0.10336914158105583,
+      "learning_rate": 0.00043226824392811255,
+      "loss": 2.9077,
+      "step": 9320
+    },
+    {
+      "epoch": 2.945431572297244,
+      "grad_norm": 0.10776845094554324,
+      "learning_rate": 0.0004317219022646864,
+      "loss": 2.8252,
+      "step": 9325
+    },
+    {
+      "epoch": 2.947010976861723,
+      "grad_norm": 0.10347109603302487,
+      "learning_rate": 0.00043117564365881847,
+      "loss": 2.9906,
+      "step": 9330
+    },
+    {
+      "epoch": 2.9485903814262024,
+      "grad_norm": 0.10262762014013083,
+      "learning_rate": 0.0004306294687750107,
+      "loss": 2.8306,
+      "step": 9335
+    },
+    {
+      "epoch": 2.9501697859906812,
+      "grad_norm": 0.11348081425096533,
+      "learning_rate": 0.0004300833782776623,
+      "loss": 2.7378,
+      "step": 9340
+    },
+    {
+      "epoch": 2.9517491905551605,
+      "grad_norm": 0.08643982582717907,
+      "learning_rate": 0.00042953737283107116,
+      "loss": 2.8455,
+      "step": 9345
+    },
+    {
+      "epoch": 2.9533285951196397,
+      "grad_norm": 0.08220527655592914,
+      "learning_rate": 0.0004289914530994303,
+      "loss": 2.7756,
+      "step": 9350
+    },
+    {
+      "epoch": 2.954907999684119,
+      "grad_norm": 0.18118766564372518,
+      "learning_rate": 0.0004284456197468296,
+      "loss": 2.8494,
+      "step": 9355
+    },
+    {
+      "epoch": 2.9564874042485982,
+      "grad_norm": 0.13143823135376625,
+      "learning_rate": 0.000427899873437253,
+      "loss": 2.7331,
+      "step": 9360
+    },
+    {
+      "epoch": 2.9580668088130775,
+      "grad_norm": 0.10058361363559926,
+      "learning_rate": 0.00042735421483457885,
+      "loss": 2.8321,
+      "step": 9365
+    },
+    {
+      "epoch": 2.9596462133775567,
+      "grad_norm": 0.12211319602843974,
+      "learning_rate": 0.0004268086446025793,
+      "loss": 2.8373,
+      "step": 9370
+    },
+    {
+      "epoch": 2.961225617942036,
+      "grad_norm": 0.11504672751453296,
+      "learning_rate": 0.00042626316340491836,
+      "loss": 2.8539,
+      "step": 9375
+    },
+    {
+      "epoch": 2.9628050225065152,
+      "grad_norm": 0.1103739918482253,
+      "learning_rate": 0.00042571777190515193,
+      "loss": 2.8314,
+      "step": 9380
+    },
+    {
+      "epoch": 2.9643844270709945,
+      "grad_norm": 0.11447689290182686,
+      "learning_rate": 0.00042517247076672695,
+      "loss": 2.7418,
+      "step": 9385
+    },
+    {
+      "epoch": 2.9659638316354733,
+      "grad_norm": 0.09751401147293268,
+      "learning_rate": 0.00042462726065297995,
+      "loss": 2.7837,
+      "step": 9390
+    },
+    {
+      "epoch": 2.9675432361999525,
+      "grad_norm": 0.1048744986956253,
+      "learning_rate": 0.00042408214222713745,
+      "loss": 2.7562,
+      "step": 9395
+    },
+    {
+      "epoch": 2.9691226407644318,
+      "grad_norm": 0.09325057594748949,
+      "learning_rate": 0.00042353711615231404,
+      "loss": 2.8928,
+      "step": 9400
+    },
+    {
+      "epoch": 2.970702045328911,
+      "grad_norm": 0.1316897961539587,
+      "learning_rate": 0.0004229921830915121,
+      "loss": 2.8762,
+      "step": 9405
+    },
+    {
+      "epoch": 2.9722814498933903,
+      "grad_norm": 0.11093338724315298,
+      "learning_rate": 0.00042244734370762036,
+      "loss": 2.8263,
+      "step": 9410
+    },
+    {
+      "epoch": 2.9738608544578695,
+      "grad_norm": 0.10257766149567363,
+      "learning_rate": 0.0004219025986634143,
+      "loss": 2.756,
+      "step": 9415
+    },
+    {
+      "epoch": 2.9754402590223483,
+      "grad_norm": 0.0990967327529078,
+      "learning_rate": 0.00042135794862155454,
+      "loss": 2.8062,
+      "step": 9420
+    },
+    {
+      "epoch": 2.9770196635868276,
+      "grad_norm": 0.10690558250376438,
+      "learning_rate": 0.0004208133942445855,
+      "loss": 2.7662,
+      "step": 9425
+    },
+    {
+      "epoch": 2.978599068151307,
+      "grad_norm": 0.11197117438239503,
+      "learning_rate": 0.00042026893619493593,
+      "loss": 2.789,
+      "step": 9430
+    },
+    {
+      "epoch": 2.980178472715786,
+      "grad_norm": 0.10816001438345199,
+      "learning_rate": 0.00041972457513491724,
+      "loss": 2.8505,
+      "step": 9435
+    },
+    {
+      "epoch": 2.9817578772802653,
+      "grad_norm": 0.1365836125794437,
+      "learning_rate": 0.00041918031172672235,
+      "loss": 2.8172,
+      "step": 9440
+    },
+    {
+      "epoch": 2.9833372818447446,
+      "grad_norm": 0.11740081577821014,
+      "learning_rate": 0.00041863614663242615,
+      "loss": 2.8843,
+      "step": 9445
+    },
+    {
+      "epoch": 2.984916686409224,
+      "grad_norm": 0.10315196514950942,
+      "learning_rate": 0.0004180920805139835,
+      "loss": 2.8653,
+      "step": 9450
+    },
+    {
+      "epoch": 2.986496090973703,
+      "grad_norm": 0.09234739238344229,
+      "learning_rate": 0.000417548114033229,
+      "loss": 2.7875,
+      "step": 9455
+    },
+    {
+      "epoch": 2.9880754955381823,
+      "grad_norm": 0.10192427331468423,
+      "learning_rate": 0.00041700424785187586,
+      "loss": 2.9056,
+      "step": 9460
+    },
+    {
+      "epoch": 2.9896549001026615,
+      "grad_norm": 0.09815002857062627,
+      "learning_rate": 0.0004164604826315155,
+      "loss": 2.8655,
+      "step": 9465
+    },
+    {
+      "epoch": 2.9912343046671404,
+      "grad_norm": 0.09440318060463548,
+      "learning_rate": 0.00041591681903361616,
+      "loss": 2.8273,
+      "step": 9470
+    },
+    {
+      "epoch": 2.9928137092316196,
+      "grad_norm": 0.08523804957990835,
+      "learning_rate": 0.00041537325771952305,
+      "loss": 2.8855,
+      "step": 9475
+    },
+    {
+      "epoch": 2.994393113796099,
+      "grad_norm": 0.10960347492187537,
+      "learning_rate": 0.00041482979935045656,
+      "loss": 2.8851,
+      "step": 9480
+    },
+    {
+      "epoch": 2.995972518360578,
+      "grad_norm": 0.11222549891471653,
+      "learning_rate": 0.000414286444587512,
+      "loss": 2.8748,
+      "step": 9485
+    },
+    {
+      "epoch": 2.9975519229250573,
+      "grad_norm": 0.11218516292236376,
+      "learning_rate": 0.0004137431940916584,
+      "loss": 2.8498,
+      "step": 9490
+    },
+    {
+      "epoch": 2.9991313274895366,
+      "grad_norm": 0.10958045434515254,
+      "learning_rate": 0.00041320004852373805,
+      "loss": 2.7871,
+      "step": 9495
+    },
+    {
+      "epoch": 3.0,
+      "eval_loss": 2.820634365081787,
+      "eval_runtime": 118.7878,
+      "eval_samples_per_second": 22.3,
+      "eval_steps_per_second": 5.581,
+      "step": 9498
+    },
+    {
+      "epoch": 3.0006317618257916,
+      "grad_norm": 0.11137564271390861,
+      "learning_rate": 0.00041265700854446605,
+      "loss": 2.7761,
+      "step": 9500
+    },
+    {
+      "epoch": 3.002211166390271,
+      "grad_norm": 0.10824032817067379,
+      "learning_rate": 0.0004121140748144283,
+      "loss": 2.9545,
+      "step": 9505
+    },
+    {
+      "epoch": 3.00379057095475,
+      "grad_norm": 0.12297364510513963,
+      "learning_rate": 0.0004115712479940821,
+      "loss": 2.8387,
+      "step": 9510
+    },
+    {
+      "epoch": 3.0053699755192294,
+      "grad_norm": 0.12154359734086224,
+      "learning_rate": 0.00041102852874375437,
+      "loss": 2.7854,
+      "step": 9515
+    },
+    {
+      "epoch": 3.0069493800837086,
+      "grad_norm": 0.0961953283591355,
+      "learning_rate": 0.000410485917723641,
+      "loss": 2.7951,
+      "step": 9520
+    },
+    {
+      "epoch": 3.008528784648188,
+      "grad_norm": 0.10474249931035579,
+      "learning_rate": 0.0004099434155938068,
+      "loss": 2.8063,
+      "step": 9525
+    },
+    {
+      "epoch": 3.0101081892126667,
+      "grad_norm": 0.10054444997655539,
+      "learning_rate": 0.0004094010230141837,
+      "loss": 2.7327,
+      "step": 9530
+    },
+    {
+      "epoch": 3.011687593777146,
+      "grad_norm": 0.09983385399225517,
+      "learning_rate": 0.0004088587406445703,
+      "loss": 2.7407,
+      "step": 9535
+    },
+    {
+      "epoch": 3.013266998341625,
+      "grad_norm": 0.08884849973502548,
+      "learning_rate": 0.0004083165691446313,
+      "loss": 2.8095,
+      "step": 9540
+    },
+    {
+      "epoch": 3.0148464029061044,
+      "grad_norm": 0.09585206257667299,
+      "learning_rate": 0.0004077745091738966,
+      "loss": 2.7738,
+      "step": 9545
+    },
+    {
+      "epoch": 3.0164258074705836,
+      "grad_norm": 0.10351425998687791,
+      "learning_rate": 0.00040723256139176044,
+      "loss": 2.8037,
+      "step": 9550
+    },
+    {
+      "epoch": 3.018005212035063,
+      "grad_norm": 0.10581996363930153,
+      "learning_rate": 0.0004066907264574803,
+      "loss": 2.8743,
+      "step": 9555
+    },
+    {
+      "epoch": 3.019584616599542,
+      "grad_norm": 0.10572898237852973,
+      "learning_rate": 0.00040614900503017665,
+      "loss": 2.8234,
+      "step": 9560
+    },
+    {
+      "epoch": 3.0211640211640214,
+      "grad_norm": 0.09120253306098203,
+      "learning_rate": 0.0004056073977688319,
+      "loss": 2.781,
+      "step": 9565
+    },
+    {
+      "epoch": 3.0227434257285,
+      "grad_norm": 0.09007962326320887,
+      "learning_rate": 0.0004050659053322892,
+      "loss": 2.8946,
+      "step": 9570
+    },
+    {
+      "epoch": 3.0243228302929794,
+      "grad_norm": 0.09230225995362885,
+      "learning_rate": 0.0004045245283792526,
+      "loss": 2.7777,
+      "step": 9575
+    },
+    {
+      "epoch": 3.0259022348574587,
+      "grad_norm": 0.14730275783137586,
+      "learning_rate": 0.0004039832675682854,
+      "loss": 2.7663,
+      "step": 9580
+    },
+    {
+      "epoch": 3.027481639421938,
+      "grad_norm": 0.11353377462122083,
+      "learning_rate": 0.0004034421235578093,
+      "loss": 2.8319,
+      "step": 9585
+    },
+    {
+      "epoch": 3.029061043986417,
+      "grad_norm": 0.1082096076103628,
+      "learning_rate": 0.0004029010970061044,
+      "loss": 2.8172,
+      "step": 9590
+    },
+    {
+      "epoch": 3.0306404485508964,
+      "grad_norm": 0.1252526709821488,
+      "learning_rate": 0.00040236018857130776,
+      "loss": 2.7878,
+      "step": 9595
+    },
+    {
+      "epoch": 3.0322198531153757,
+      "grad_norm": 0.10508346750056435,
+      "learning_rate": 0.00040181939891141273,
+      "loss": 2.805,
+      "step": 9600
+    },
+    {
+      "epoch": 3.0337992576798545,
+      "grad_norm": 0.11012507855534207,
+      "learning_rate": 0.00040127872868426807,
+      "loss": 2.8378,
+      "step": 9605
+    },
+    {
+      "epoch": 3.0353786622443337,
+      "grad_norm": 0.0905549160336993,
+      "learning_rate": 0.00040073817854757753,
+      "loss": 2.8586,
+      "step": 9610
+    },
+    {
+      "epoch": 3.036958066808813,
+      "grad_norm": 0.10658082029039984,
+      "learning_rate": 0.0004001977491588984,
+      "loss": 2.7928,
+      "step": 9615
+    },
+    {
+      "epoch": 3.038537471373292,
+      "grad_norm": 0.08473814628075636,
+      "learning_rate": 0.0003996574411756412,
+      "loss": 2.8769,
+      "step": 9620
+    },
+    {
+      "epoch": 3.0401168759377715,
+      "grad_norm": 0.10456114140621427,
+      "learning_rate": 0.00039911725525506914,
+      "loss": 2.8539,
+      "step": 9625
+    },
+    {
+      "epoch": 3.0416962805022507,
+      "grad_norm": 0.0988988689051054,
+      "learning_rate": 0.0003985771920542967,
+      "loss": 2.806,
+      "step": 9630
+    },
+    {
+      "epoch": 3.04327568506673,
+      "grad_norm": 0.09450173016461,
+      "learning_rate": 0.00039803725223028864,
+      "loss": 2.851,
+      "step": 9635
+    },
+    {
+      "epoch": 3.044855089631209,
+      "grad_norm": 0.12539248662855523,
+      "learning_rate": 0.00039749743643986035,
+      "loss": 2.8271,
+      "step": 9640
+    },
+    {
+      "epoch": 3.046434494195688,
+      "grad_norm": 0.10030288651138776,
+      "learning_rate": 0.00039695774533967586,
+      "loss": 2.7322,
+      "step": 9645
+    },
+    {
+      "epoch": 3.0480138987601673,
+      "grad_norm": 0.09442194681297006,
+      "learning_rate": 0.0003964181795862476,
+      "loss": 2.8108,
+      "step": 9650
+    },
+    {
+      "epoch": 3.0495933033246465,
+      "grad_norm": 0.09793258439010631,
+      "learning_rate": 0.00039587873983593585,
+      "loss": 2.8668,
+      "step": 9655
+    },
+    {
+      "epoch": 3.0511727078891258,
+      "grad_norm": 0.09533279772919158,
+      "learning_rate": 0.00039533942674494735,
+      "loss": 2.7553,
+      "step": 9660
+    },
+    {
+      "epoch": 3.052752112453605,
+      "grad_norm": 0.09165559609359519,
+      "learning_rate": 0.00039480024096933455,
+      "loss": 2.7224,
+      "step": 9665
+    },
+    {
+      "epoch": 3.0543315170180843,
+      "grad_norm": 0.08899617957042029,
+      "learning_rate": 0.0003942611831649953,
+      "loss": 2.8608,
+      "step": 9670
+    },
+    {
+      "epoch": 3.0559109215825635,
+      "grad_norm": 0.12462258209023344,
+      "learning_rate": 0.00039372225398767176,
+      "loss": 2.799,
+      "step": 9675
+    },
+    {
+      "epoch": 3.0574903261470427,
+      "grad_norm": 0.09919996000954215,
+      "learning_rate": 0.0003931834540929498,
+      "loss": 2.7263,
+      "step": 9680
+    },
+    {
+      "epoch": 3.0590697307115216,
+      "grad_norm": 0.10144490387767188,
+      "learning_rate": 0.0003926447841362575,
+      "loss": 2.7332,
+      "step": 9685
+    },
+    {
+      "epoch": 3.060649135276001,
+      "grad_norm": 0.13350840711628595,
+      "learning_rate": 0.0003921062447728654,
+      "loss": 2.9546,
+      "step": 9690
+    },
+    {
+      "epoch": 3.06222853984048,
+      "grad_norm": 0.11599137513418321,
+      "learning_rate": 0.0003915678366578848,
+      "loss": 2.7852,
+      "step": 9695
+    },
+    {
+      "epoch": 3.0638079444049593,
+      "grad_norm": 0.09011280977038039,
+      "learning_rate": 0.00039102956044626745,
+      "loss": 2.7473,
+      "step": 9700
+    },
+    {
+      "epoch": 3.0653873489694385,
+      "grad_norm": 0.10744348562991023,
+      "learning_rate": 0.000390491416792805,
+      "loss": 2.8039,
+      "step": 9705
+    },
+    {
+      "epoch": 3.066966753533918,
+      "grad_norm": 0.1076047456045747,
+      "learning_rate": 0.00038995340635212747,
+      "loss": 2.8721,
+      "step": 9710
+    },
+    {
+      "epoch": 3.068546158098397,
+      "grad_norm": 0.11090394334029989,
+      "learning_rate": 0.0003894155297787027,
+      "loss": 2.9012,
+      "step": 9715
+    },
+    {
+      "epoch": 3.0701255626628763,
+      "grad_norm": 0.09807974395489591,
+      "learning_rate": 0.00038887778772683605,
+      "loss": 2.7818,
+      "step": 9720
+    },
+    {
+      "epoch": 3.071704967227355,
+      "grad_norm": 0.10706239708253429,
+      "learning_rate": 0.0003883401808506688,
+      "loss": 2.7635,
+      "step": 9725
+    },
+    {
+      "epoch": 3.0732843717918343,
+      "grad_norm": 0.10143064130984526,
+      "learning_rate": 0.0003878027098041786,
+      "loss": 2.7579,
+      "step": 9730
+    },
+    {
+      "epoch": 3.0748637763563136,
+      "grad_norm": 0.10400907430117945,
+      "learning_rate": 0.00038726537524117713,
+      "loss": 2.8069,
+      "step": 9735
+    },
+    {
+      "epoch": 3.076443180920793,
+      "grad_norm": 0.09316922670618225,
+      "learning_rate": 0.00038672817781531025,
+      "loss": 2.8639,
+      "step": 9740
+    },
+    {
+      "epoch": 3.078022585485272,
+      "grad_norm": 0.09349735335320371,
+      "learning_rate": 0.0003861911181800568,
+      "loss": 2.7537,
+      "step": 9745
+    },
+    {
+      "epoch": 3.0796019900497513,
+      "grad_norm": 0.08828836960714911,
+      "learning_rate": 0.00038565419698872837,
+      "loss": 2.774,
+      "step": 9750
+    },
+    {
+      "epoch": 3.0811813946142306,
+      "grad_norm": 0.10118948073176658,
+      "learning_rate": 0.0003851174148944681,
+      "loss": 2.7001,
+      "step": 9755
+    },
+    {
+      "epoch": 3.08276079917871,
+      "grad_norm": 0.09821378293596399,
+      "learning_rate": 0.0003845807725502499,
+      "loss": 2.7579,
+      "step": 9760
+    },
+    {
+      "epoch": 3.0843402037431886,
+      "grad_norm": 0.10329086950182034,
+      "learning_rate": 0.0003840442706088772,
+      "loss": 2.829,
+      "step": 9765
+    },
+    {
+      "epoch": 3.085919608307668,
+      "grad_norm": 0.11605258739575236,
+      "learning_rate": 0.00038350790972298336,
+      "loss": 2.8862,
+      "step": 9770
+    },
+    {
+      "epoch": 3.087499012872147,
+      "grad_norm": 0.09100937689590838,
+      "learning_rate": 0.0003829716905450294,
+      "loss": 2.8117,
+      "step": 9775
+    },
+    {
+      "epoch": 3.0890784174366264,
+      "grad_norm": 0.155556621997077,
+      "learning_rate": 0.00038243561372730496,
+      "loss": 2.9684,
+      "step": 9780
+    },
+    {
+      "epoch": 3.0906578220011056,
+      "grad_norm": 0.09450981898431092,
+      "learning_rate": 0.00038189967992192545,
+      "loss": 2.7816,
+      "step": 9785
+    },
+    {
+      "epoch": 3.092237226565585,
+      "grad_norm": 0.09959380196851375,
+      "learning_rate": 0.0003813638897808331,
+      "loss": 2.8895,
+      "step": 9790
+    },
+    {
+      "epoch": 3.093816631130064,
+      "grad_norm": 0.10169324490712747,
+      "learning_rate": 0.0003808282439557948,
+      "loss": 2.8379,
+      "step": 9795
+    },
+    {
+      "epoch": 3.0953960356945434,
+      "grad_norm": 0.10058561730050815,
+      "learning_rate": 0.0003802927430984023,
+      "loss": 2.7332,
+      "step": 9800
+    },
+    {
+      "epoch": 3.096975440259022,
+      "grad_norm": 0.11816001771566845,
+      "learning_rate": 0.00037975738786007055,
+      "loss": 2.7943,
+      "step": 9805
+    },
+    {
+      "epoch": 3.0985548448235014,
+      "grad_norm": 0.08826920345868178,
+      "learning_rate": 0.00037922217889203814,
+      "loss": 2.7637,
+      "step": 9810
+    },
+    {
+      "epoch": 3.1001342493879807,
+      "grad_norm": 0.12705683278388166,
+      "learning_rate": 0.0003786871168453649,
+      "loss": 2.8163,
+      "step": 9815
+    },
+    {
+      "epoch": 3.10171365395246,
+      "grad_norm": 0.11495920571191703,
+      "learning_rate": 0.00037815220237093244,
+      "loss": 2.836,
+      "step": 9820
+    },
+    {
+      "epoch": 3.103293058516939,
+      "grad_norm": 0.12432376343554186,
+      "learning_rate": 0.00037761743611944255,
+      "loss": 2.8075,
+      "step": 9825
+    },
+    {
+      "epoch": 3.1048724630814184,
+      "grad_norm": 0.09656694722527305,
+      "learning_rate": 0.0003770828187414169,
+      "loss": 2.8801,
+      "step": 9830
+    },
+    {
+      "epoch": 3.1064518676458976,
+      "grad_norm": 0.11029553326905037,
+      "learning_rate": 0.000376548350887196,
+      "loss": 2.7515,
+      "step": 9835
+    },
+    {
+      "epoch": 3.108031272210377,
+      "grad_norm": 0.0830397231016251,
+      "learning_rate": 0.0003760140332069387,
+      "loss": 2.7623,
+      "step": 9840
+    },
+    {
+      "epoch": 3.1096106767748557,
+      "grad_norm": 0.12136488671310046,
+      "learning_rate": 0.00037547986635062076,
+      "loss": 2.7761,
+      "step": 9845
+    },
+    {
+      "epoch": 3.111190081339335,
+      "grad_norm": 0.12969324968762255,
+      "learning_rate": 0.00037494585096803476,
+      "loss": 2.8342,
+      "step": 9850
+    },
+    {
+      "epoch": 3.112769485903814,
+      "grad_norm": 0.12352241296333281,
+      "learning_rate": 0.00037441198770878857,
+      "loss": 2.8738,
+      "step": 9855
+    },
+    {
+      "epoch": 3.1143488904682934,
+      "grad_norm": 0.14236331604294866,
+      "learning_rate": 0.0003738782772223059,
+      "loss": 2.7812,
+      "step": 9860
+    },
+    {
+      "epoch": 3.1159282950327727,
+      "grad_norm": 0.10979819536457726,
+      "learning_rate": 0.00037334472015782374,
+      "loss": 2.7447,
+      "step": 9865
+    },
+    {
+      "epoch": 3.117507699597252,
+      "grad_norm": 0.09192732565456906,
+      "learning_rate": 0.00037281131716439297,
+      "loss": 2.8563,
+      "step": 9870
+    },
+    {
+      "epoch": 3.119087104161731,
+      "grad_norm": 0.1005101546494545,
+      "learning_rate": 0.00037227806889087676,
+      "loss": 2.8625,
+      "step": 9875
+    },
+    {
+      "epoch": 3.1206665087262104,
+      "grad_norm": 0.10713773883542309,
+      "learning_rate": 0.00037174497598595,
+      "loss": 2.8055,
+      "step": 9880
+    },
+    {
+      "epoch": 3.1222459132906892,
+      "grad_norm": 0.09367041726000498,
+      "learning_rate": 0.0003712120390980992,
+      "loss": 2.7897,
+      "step": 9885
+    },
+    {
+      "epoch": 3.1238253178551685,
+      "grad_norm": 0.111252472750593,
+      "learning_rate": 0.00037067925887562033,
+      "loss": 2.8767,
+      "step": 9890
+    },
+    {
+      "epoch": 3.1254047224196477,
+      "grad_norm": 0.10653969198593435,
+      "learning_rate": 0.0003701466359666191,
+      "loss": 2.8137,
+      "step": 9895
+    },
+    {
+      "epoch": 3.126984126984127,
+      "grad_norm": 0.10360624517617704,
+      "learning_rate": 0.00036961417101901003,
+      "loss": 2.8624,
+      "step": 9900
+    },
+    {
+      "epoch": 3.1285635315486062,
+      "grad_norm": 0.08869417865804269,
+      "learning_rate": 0.00036908186468051496,
+      "loss": 2.7433,
+      "step": 9905
+    },
+    {
+      "epoch": 3.1301429361130855,
+      "grad_norm": 0.08963134101472572,
+      "learning_rate": 0.0003685497175986634,
+      "loss": 2.7508,
+      "step": 9910
+    },
+    {
+      "epoch": 3.1317223406775647,
+      "grad_norm": 0.08337069015473321,
+      "learning_rate": 0.00036801773042079085,
+      "loss": 2.749,
+      "step": 9915
+    },
+    {
+      "epoch": 3.1333017452420435,
+      "grad_norm": 0.11348349319289626,
+      "learning_rate": 0.00036748590379403833,
+      "loss": 2.8682,
+      "step": 9920
+    },
+    {
+      "epoch": 3.1348811498065228,
+      "grad_norm": 0.09468176150451076,
+      "learning_rate": 0.0003669542383653514,
+      "loss": 2.8021,
+      "step": 9925
+    },
+    {
+      "epoch": 3.136460554371002,
+      "grad_norm": 0.10888557336854489,
+      "learning_rate": 0.00036642273478147957,
+      "loss": 2.8114,
+      "step": 9930
+    },
+    {
+      "epoch": 3.1380399589354813,
+      "grad_norm": 0.09812788122332593,
+      "learning_rate": 0.000365891393688976,
+      "loss": 2.8673,
+      "step": 9935
+    },
+    {
+      "epoch": 3.1396193634999605,
+      "grad_norm": 0.10373767188692451,
+      "learning_rate": 0.0003653602157341953,
+      "loss": 2.8692,
+      "step": 9940
+    },
+    {
+      "epoch": 3.1411987680644398,
+      "grad_norm": 0.10425056849464936,
+      "learning_rate": 0.0003648292015632942,
+      "loss": 2.7598,
+      "step": 9945
+    },
+    {
+      "epoch": 3.142778172628919,
+      "grad_norm": 0.10437038930991278,
+      "learning_rate": 0.00036429835182223024,
+      "loss": 2.8128,
+      "step": 9950
+    },
+    {
+      "epoch": 3.1443575771933983,
+      "grad_norm": 0.08773464643806514,
+      "learning_rate": 0.00036376766715676053,
+      "loss": 2.8262,
+      "step": 9955
+    },
+    {
+      "epoch": 3.1459369817578775,
+      "grad_norm": 0.10175662607669828,
+      "learning_rate": 0.0003632371482124416,
+      "loss": 2.7323,
+      "step": 9960
+    },
+    {
+      "epoch": 3.1475163863223563,
+      "grad_norm": 0.08826994901790985,
+      "learning_rate": 0.00036270679563462873,
+      "loss": 2.8165,
+      "step": 9965
+    },
+    {
+      "epoch": 3.1490957908868356,
+      "grad_norm": 0.10089045643559466,
+      "learning_rate": 0.00036217661006847417,
+      "loss": 2.7692,
+      "step": 9970
+    },
+    {
+      "epoch": 3.150675195451315,
+      "grad_norm": 0.0963958566936885,
+      "learning_rate": 0.0003616465921589275,
+      "loss": 2.8025,
+      "step": 9975
+    },
+    {
+      "epoch": 3.152254600015794,
+      "grad_norm": 0.12276808862628645,
+      "learning_rate": 0.00036111674255073415,
+      "loss": 2.8106,
+      "step": 9980
+    },
+    {
+      "epoch": 3.1538340045802733,
+      "grad_norm": 0.11509772915867757,
+      "learning_rate": 0.0003605870618884345,
+      "loss": 2.8525,
+      "step": 9985
+    },
+    {
+      "epoch": 3.1554134091447525,
+      "grad_norm": 0.0983241286227083,
+      "learning_rate": 0.00036005755081636425,
+      "loss": 2.8276,
+      "step": 9990
+    },
+    {
+      "epoch": 3.156992813709232,
+      "grad_norm": 0.11038222745065804,
+      "learning_rate": 0.0003595282099786523,
+      "loss": 2.7956,
+      "step": 9995
+    },
+    {
+      "epoch": 3.1585722182737106,
+      "grad_norm": 0.10526613695833549,
+      "learning_rate": 0.0003589990400192201,
+      "loss": 2.8081,
+      "step": 10000
+    },
+    {
+      "epoch": 3.16015162283819,
+      "grad_norm": 0.09230450379133098,
+      "learning_rate": 0.00035847004158178186,
+      "loss": 2.7626,
+      "step": 10005
+    },
+    {
+      "epoch": 3.161731027402669,
+      "grad_norm": 0.12200782727310795,
+      "learning_rate": 0.0003579412153098428,
+      "loss": 2.8417,
+      "step": 10010
+    },
+    {
+      "epoch": 3.1633104319671483,
+      "grad_norm": 0.10083668653829685,
+      "learning_rate": 0.00035741256184669903,
+      "loss": 2.8797,
+      "step": 10015
+    },
+    {
+      "epoch": 3.1648898365316276,
+      "grad_norm": 0.11019845863132575,
+      "learning_rate": 0.00035688408183543586,
+      "loss": 2.7975,
+      "step": 10020
+    },
+    {
+      "epoch": 3.166469241096107,
+      "grad_norm": 0.11159164688843977,
+      "learning_rate": 0.0003563557759189282,
+      "loss": 2.7171,
+      "step": 10025
+    },
+    {
+      "epoch": 3.168048645660586,
+      "grad_norm": 0.10537361417941192,
+      "learning_rate": 0.00035582764473983896,
+      "loss": 2.8344,
+      "step": 10030
+    },
+    {
+      "epoch": 3.1696280502250653,
+      "grad_norm": 0.09542451280482253,
+      "learning_rate": 0.00035529968894061815,
+      "loss": 2.7607,
+      "step": 10035
+    },
+    {
+      "epoch": 3.1712074547895446,
+      "grad_norm": 0.08830766235266202,
+      "learning_rate": 0.0003547719091635031,
+      "loss": 2.7306,
+      "step": 10040
+    },
+    {
+      "epoch": 3.1727868593540234,
+      "grad_norm": 0.09387085756674177,
+      "learning_rate": 0.0003542443060505167,
+      "loss": 2.8492,
+      "step": 10045
+    },
+    {
+      "epoch": 3.1743662639185026,
+      "grad_norm": 0.10198816922678111,
+      "learning_rate": 0.0003537168802434666,
+      "loss": 2.8642,
+      "step": 10050
+    },
+    {
+      "epoch": 3.175945668482982,
+      "grad_norm": 0.13322243689409521,
+      "learning_rate": 0.00035318963238394524,
+      "loss": 2.6974,
+      "step": 10055
+    },
+    {
+      "epoch": 3.177525073047461,
+      "grad_norm": 0.1138497993736802,
+      "learning_rate": 0.0003526625631133283,
+      "loss": 2.7591,
+      "step": 10060
+    },
+    {
+      "epoch": 3.1791044776119404,
+      "grad_norm": 0.12308413441898473,
+      "learning_rate": 0.0003521356730727747,
+      "loss": 2.8083,
+      "step": 10065
+    },
+    {
+      "epoch": 3.1806838821764196,
+      "grad_norm": 0.10174180898082953,
+      "learning_rate": 0.00035160896290322466,
+      "loss": 2.7881,
+      "step": 10070
+    },
+    {
+      "epoch": 3.182263286740899,
+      "grad_norm": 0.09672870359748681,
+      "learning_rate": 0.0003510824332454,
+      "loss": 2.734,
+      "step": 10075
+    },
+    {
+      "epoch": 3.1838426913053777,
+      "grad_norm": 0.08547351924749266,
+      "learning_rate": 0.0003505560847398027,
+      "loss": 2.7196,
+      "step": 10080
+    },
+    {
+      "epoch": 3.185422095869857,
+      "grad_norm": 0.08723168197605567,
+      "learning_rate": 0.0003500299180267146,
+      "loss": 2.796,
+      "step": 10085
+    },
+    {
+      "epoch": 3.187001500434336,
+      "grad_norm": 0.08406353772095086,
+      "learning_rate": 0.0003495039337461966,
+      "loss": 2.7767,
+      "step": 10090
+    },
+    {
+      "epoch": 3.1885809049988154,
+      "grad_norm": 0.09271636093914204,
+      "learning_rate": 0.00034897813253808717,
+      "loss": 2.8579,
+      "step": 10095
+    },
+    {
+      "epoch": 3.1901603095632947,
+      "grad_norm": 0.09685899156814314,
+      "learning_rate": 0.0003484525150420024,
+      "loss": 2.802,
+      "step": 10100
+    },
+    {
+      "epoch": 3.191739714127774,
+      "grad_norm": 0.08296004406299497,
+      "learning_rate": 0.00034792708189733477,
+      "loss": 2.7294,
+      "step": 10105
+    },
+    {
+      "epoch": 3.193319118692253,
+      "grad_norm": 0.09573421531424595,
+      "learning_rate": 0.00034740183374325255,
+      "loss": 2.7771,
+      "step": 10110
+    },
+    {
+      "epoch": 3.1948985232567324,
+      "grad_norm": 0.1026502934765465,
+      "learning_rate": 0.00034687677121869885,
+      "loss": 2.7712,
+      "step": 10115
+    },
+    {
+      "epoch": 3.196477927821211,
+      "grad_norm": 0.1197013920449785,
+      "learning_rate": 0.0003463518949623914,
+      "loss": 2.7689,
+      "step": 10120
+    },
+    {
+      "epoch": 3.1980573323856905,
+      "grad_norm": 0.10892579808211512,
+      "learning_rate": 0.0003458272056128211,
+      "loss": 2.8117,
+      "step": 10125
+    },
+    {
+      "epoch": 3.1996367369501697,
+      "grad_norm": 0.11654615110479964,
+      "learning_rate": 0.000345302703808251,
+      "loss": 2.7712,
+      "step": 10130
+    },
+    {
+      "epoch": 3.201216141514649,
+      "grad_norm": 0.08037702352022942,
+      "learning_rate": 0.00034477839018671677,
+      "loss": 2.7656,
+      "step": 10135
+    },
+    {
+      "epoch": 3.202795546079128,
+      "grad_norm": 0.09728981313291721,
+      "learning_rate": 0.00034425426538602457,
+      "loss": 2.7511,
+      "step": 10140
+    },
+    {
+      "epoch": 3.2043749506436074,
+      "grad_norm": 0.09791391597565863,
+      "learning_rate": 0.00034373033004375154,
+      "loss": 2.7805,
+      "step": 10145
+    },
+    {
+      "epoch": 3.2059543552080867,
+      "grad_norm": 0.10599170335724191,
+      "learning_rate": 0.00034320658479724354,
+      "loss": 2.8416,
+      "step": 10150
+    },
+    {
+      "epoch": 3.207533759772566,
+      "grad_norm": 0.0871599731028459,
+      "learning_rate": 0.00034268303028361593,
+      "loss": 2.7097,
+      "step": 10155
+    },
+    {
+      "epoch": 3.2091131643370447,
+      "grad_norm": 0.0850412540153791,
+      "learning_rate": 0.00034215966713975135,
+      "loss": 2.8742,
+      "step": 10160
+    },
+    {
+      "epoch": 3.210692568901524,
+      "grad_norm": 0.08043909999541177,
+      "learning_rate": 0.0003416364960023001,
+      "loss": 2.7685,
+      "step": 10165
+    },
+    {
+      "epoch": 3.2122719734660032,
+      "grad_norm": 0.08821836004911278,
+      "learning_rate": 0.000341113517507679,
+      "loss": 2.7397,
+      "step": 10170
+    },
+    {
+      "epoch": 3.2138513780304825,
+      "grad_norm": 0.08649994686945457,
+      "learning_rate": 0.00034059073229207034,
+      "loss": 2.8747,
+      "step": 10175
+    },
+    {
+      "epoch": 3.2154307825949617,
+      "grad_norm": 0.10390318980319042,
+      "learning_rate": 0.0003400681409914211,
+      "loss": 2.9359,
+      "step": 10180
+    },
+    {
+      "epoch": 3.217010187159441,
+      "grad_norm": 0.09047296521943848,
+      "learning_rate": 0.00033954574424144274,
+      "loss": 2.8406,
+      "step": 10185
+    },
+    {
+      "epoch": 3.2185895917239202,
+      "grad_norm": 0.09267422370405444,
+      "learning_rate": 0.0003390235426776095,
+      "loss": 2.7719,
+      "step": 10190
+    },
+    {
+      "epoch": 3.2201689962883995,
+      "grad_norm": 0.09331054979286894,
+      "learning_rate": 0.00033850153693515915,
+      "loss": 2.8203,
+      "step": 10195
+    },
+    {
+      "epoch": 3.2217484008528783,
+      "grad_norm": 0.09305039470541963,
+      "learning_rate": 0.0003379797276490904,
+      "loss": 2.8131,
+      "step": 10200
+    },
+    {
+      "epoch": 3.2233278054173575,
+      "grad_norm": 0.08844182082942555,
+      "learning_rate": 0.00033745811545416327,
+      "loss": 2.7837,
+      "step": 10205
+    },
+    {
+      "epoch": 3.2249072099818368,
+      "grad_norm": 0.10344603201593157,
+      "learning_rate": 0.0003369367009848979,
+      "loss": 2.8531,
+      "step": 10210
+    },
+    {
+      "epoch": 3.226486614546316,
+      "grad_norm": 0.08988469720539088,
+      "learning_rate": 0.00033641548487557406,
+      "loss": 2.8106,
+      "step": 10215
+    },
+    {
+      "epoch": 3.2280660191107953,
+      "grad_norm": 0.09553604500726276,
+      "learning_rate": 0.00033589446776023023,
+      "loss": 2.9042,
+      "step": 10220
+    },
+    {
+      "epoch": 3.2296454236752745,
+      "grad_norm": 0.09403774521688882,
+      "learning_rate": 0.00033537365027266284,
+      "loss": 2.8888,
+      "step": 10225
+    },
+    {
+      "epoch": 3.2312248282397538,
+      "grad_norm": 0.09265964429985306,
+      "learning_rate": 0.0003348530330464252,
+      "loss": 2.7721,
+      "step": 10230
+    },
+    {
+      "epoch": 3.2328042328042326,
+      "grad_norm": 0.09851990312142413,
+      "learning_rate": 0.0003343326167148275,
+      "loss": 2.8712,
+      "step": 10235
+    },
+    {
+      "epoch": 3.234383637368712,
+      "grad_norm": 0.08719383377135549,
+      "learning_rate": 0.00033381240191093477,
+      "loss": 2.8202,
+      "step": 10240
+    },
+    {
+      "epoch": 3.235963041933191,
+      "grad_norm": 0.09013388573499846,
+      "learning_rate": 0.0003332923892675679,
+      "loss": 2.7795,
+      "step": 10245
+    },
+    {
+      "epoch": 3.2375424464976703,
+      "grad_norm": 0.08995633861369724,
+      "learning_rate": 0.0003327725794173011,
+      "loss": 2.8209,
+      "step": 10250
+    },
+    {
+      "epoch": 3.2391218510621496,
+      "grad_norm": 0.10856360963772668,
+      "learning_rate": 0.00033225297299246227,
+      "loss": 2.7667,
+      "step": 10255
+    },
+    {
+      "epoch": 3.240701255626629,
+      "grad_norm": 0.1051886244149574,
+      "learning_rate": 0.00033173357062513153,
+      "loss": 2.7641,
+      "step": 10260
+    },
+    {
+      "epoch": 3.242280660191108,
+      "grad_norm": 0.10921839089379684,
+      "learning_rate": 0.00033121437294714103,
+      "loss": 2.8636,
+      "step": 10265
+    },
+    {
+      "epoch": 3.2438600647555873,
+      "grad_norm": 0.10352380755657233,
+      "learning_rate": 0.000330695380590074,
+      "loss": 2.8334,
+      "step": 10270
+    },
+    {
+      "epoch": 3.2454394693200666,
+      "grad_norm": 0.1100045225541539,
+      "learning_rate": 0.00033017659418526366,
+      "loss": 2.8854,
+      "step": 10275
+    },
+    {
+      "epoch": 3.2470188738845454,
+      "grad_norm": 0.09270834220250747,
+      "learning_rate": 0.0003296580143637927,
+      "loss": 2.6529,
+      "step": 10280
+    },
+    {
+      "epoch": 3.2485982784490246,
+      "grad_norm": 0.08939893829290244,
+      "learning_rate": 0.0003291396417564927,
+      "loss": 2.8705,
+      "step": 10285
+    },
+    {
+      "epoch": 3.250177683013504,
+      "grad_norm": 0.10116051723993615,
+      "learning_rate": 0.00032862147699394307,
+      "loss": 2.7462,
+      "step": 10290
+    },
+    {
+      "epoch": 3.251757087577983,
+      "grad_norm": 0.08254665948382854,
+      "learning_rate": 0.0003281035207064702,
+      "loss": 2.7523,
+      "step": 10295
+    },
+    {
+      "epoch": 3.2533364921424623,
+      "grad_norm": 0.0927012310467256,
+      "learning_rate": 0.00032758577352414743,
+      "loss": 2.8168,
+      "step": 10300
+    },
+    {
+      "epoch": 3.2549158967069416,
+      "grad_norm": 0.0727240409285992,
+      "learning_rate": 0.0003270682360767933,
+      "loss": 2.7105,
+      "step": 10305
+    },
+    {
+      "epoch": 3.256495301271421,
+      "grad_norm": 0.07641937991051459,
+      "learning_rate": 0.000326550908993971,
+      "loss": 2.6676,
+      "step": 10310
+    },
+    {
+      "epoch": 3.2580747058358996,
+      "grad_norm": 0.09083967725825964,
+      "learning_rate": 0.00032603379290498845,
+      "loss": 2.8156,
+      "step": 10315
+    },
+    {
+      "epoch": 3.259654110400379,
+      "grad_norm": 0.10311120598563625,
+      "learning_rate": 0.0003255168884388962,
+      "loss": 2.7448,
+      "step": 10320
+    },
+    {
+      "epoch": 3.261233514964858,
+      "grad_norm": 0.12285696927565312,
+      "learning_rate": 0.0003250001962244881,
+      "loss": 2.7653,
+      "step": 10325
+    },
+    {
+      "epoch": 3.2628129195293374,
+      "grad_norm": 0.10189192939942499,
+      "learning_rate": 0.00032448371689029916,
+      "loss": 2.74,
+      "step": 10330
+    },
+    {
+      "epoch": 3.2643923240938166,
+      "grad_norm": 0.08569907776229424,
+      "learning_rate": 0.000323967451064606,
+      "loss": 2.8235,
+      "step": 10335
+    },
+    {
+      "epoch": 3.265971728658296,
+      "grad_norm": 0.10915203379452394,
+      "learning_rate": 0.0003234513993754249,
+      "loss": 2.8322,
+      "step": 10340
+    },
+    {
+      "epoch": 3.267551133222775,
+      "grad_norm": 0.10559515915941756,
+      "learning_rate": 0.00032293556245051205,
+      "loss": 2.911,
+      "step": 10345
+    },
+    {
+      "epoch": 3.2691305377872544,
+      "grad_norm": 0.09274098866921747,
+      "learning_rate": 0.0003224199409173626,
+      "loss": 2.7084,
+      "step": 10350
+    },
+    {
+      "epoch": 3.2707099423517336,
+      "grad_norm": 0.10169340022249343,
+      "learning_rate": 0.00032190453540320905,
+      "loss": 2.7828,
+      "step": 10355
+    },
+    {
+      "epoch": 3.2722893469162124,
+      "grad_norm": 0.10538902352745173,
+      "learning_rate": 0.00032138934653502154,
+      "loss": 2.7102,
+      "step": 10360
+    },
+    {
+      "epoch": 3.2738687514806917,
+      "grad_norm": 0.13638379454753577,
+      "learning_rate": 0.0003208743749395068,
+      "loss": 2.8244,
+      "step": 10365
+    },
+    {
+      "epoch": 3.275448156045171,
+      "grad_norm": 0.14513930133062752,
+      "learning_rate": 0.00032035962124310675,
+      "loss": 2.8425,
+      "step": 10370
+    },
+    {
+      "epoch": 3.27702756060965,
+      "grad_norm": 0.08349472002147552,
+      "learning_rate": 0.00031984508607199873,
+      "loss": 2.7892,
+      "step": 10375
+    },
+    {
+      "epoch": 3.2786069651741294,
+      "grad_norm": 0.08104299402779074,
+      "learning_rate": 0.0003193307700520941,
+      "loss": 2.7423,
+      "step": 10380
+    },
+    {
+      "epoch": 3.2801863697386087,
+      "grad_norm": 0.08236614473434854,
+      "learning_rate": 0.0003188166738090377,
+      "loss": 2.8255,
+      "step": 10385
+    },
+    {
+      "epoch": 3.281765774303088,
+      "grad_norm": 0.08969459608903811,
+      "learning_rate": 0.00031830279796820656,
+      "loss": 2.7319,
+      "step": 10390
+    },
+    {
+      "epoch": 3.2833451788675667,
+      "grad_norm": 0.11484517090134234,
+      "learning_rate": 0.0003177891431547101,
+      "loss": 2.9348,
+      "step": 10395
+    },
+    {
+      "epoch": 3.284924583432046,
+      "grad_norm": 0.11750554407948001,
+      "learning_rate": 0.000317275709993389,
+      "loss": 2.7242,
+      "step": 10400
+    },
+    {
+      "epoch": 3.286503987996525,
+      "grad_norm": 0.12448605915336501,
+      "learning_rate": 0.00031676249910881374,
+      "loss": 2.8041,
+      "step": 10405
+    },
+    {
+      "epoch": 3.2880833925610045,
+      "grad_norm": 0.0826535725225018,
+      "learning_rate": 0.00031624951112528484,
+      "loss": 2.7142,
+      "step": 10410
+    },
+    {
+      "epoch": 3.2896627971254837,
+      "grad_norm": 0.09599474247091126,
+      "learning_rate": 0.0003157367466668316,
+      "loss": 2.7681,
+      "step": 10415
+    },
+    {
+      "epoch": 3.291242201689963,
+      "grad_norm": 0.08587535416157806,
+      "learning_rate": 0.00031522420635721107,
+      "loss": 2.7437,
+      "step": 10420
+    },
+    {
+      "epoch": 3.292821606254442,
+      "grad_norm": 0.08780026752528526,
+      "learning_rate": 0.00031471189081990814,
+      "loss": 2.7516,
+      "step": 10425
+    },
+    {
+      "epoch": 3.2944010108189214,
+      "grad_norm": 0.08833883706658188,
+      "learning_rate": 0.0003141998006781341,
+      "loss": 2.745,
+      "step": 10430
+    },
+    {
+      "epoch": 3.2959804153834007,
+      "grad_norm": 0.09358958670004401,
+      "learning_rate": 0.000313687936554826,
+      "loss": 2.81,
+      "step": 10435
+    },
+    {
+      "epoch": 3.2975598199478795,
+      "grad_norm": 0.10450158707026738,
+      "learning_rate": 0.0003131762990726457,
+      "loss": 2.85,
+      "step": 10440
+    },
+    {
+      "epoch": 3.2991392245123587,
+      "grad_norm": 0.09643583499425244,
+      "learning_rate": 0.0003126648888539798,
+      "loss": 2.8408,
+      "step": 10445
+    },
+    {
+      "epoch": 3.300718629076838,
+      "grad_norm": 0.08745485523804206,
+      "learning_rate": 0.0003121537065209382,
+      "loss": 2.7717,
+      "step": 10450
+    },
+    {
+      "epoch": 3.3022980336413172,
+      "grad_norm": 0.08974871708382999,
+      "learning_rate": 0.0003116427526953536,
+      "loss": 2.799,
+      "step": 10455
+    },
+    {
+      "epoch": 3.3038774382057965,
+      "grad_norm": 0.11404985610167258,
+      "learning_rate": 0.000311132027998781,
+      "loss": 2.8022,
+      "step": 10460
+    },
+    {
+      "epoch": 3.3054568427702757,
+      "grad_norm": 0.12015100369305097,
+      "learning_rate": 0.0003106215330524962,
+      "loss": 2.7478,
+      "step": 10465
+    },
+    {
+      "epoch": 3.307036247334755,
+      "grad_norm": 0.10342486455572253,
+      "learning_rate": 0.00031011126847749573,
+      "loss": 2.7389,
+      "step": 10470
+    },
+    {
+      "epoch": 3.308615651899234,
+      "grad_norm": 0.14353090615973468,
+      "learning_rate": 0.000309601234894496,
+      "loss": 2.8025,
+      "step": 10475
+    },
+    {
+      "epoch": 3.310195056463713,
+      "grad_norm": 0.12943337260732343,
+      "learning_rate": 0.0003090914329239325,
+      "loss": 2.785,
+      "step": 10480
+    },
+    {
+      "epoch": 3.3117744610281923,
+      "grad_norm": 0.11866623983880317,
+      "learning_rate": 0.0003085818631859585,
+      "loss": 2.9396,
+      "step": 10485
+    },
+    {
+      "epoch": 3.3133538655926715,
+      "grad_norm": 0.11061480562524836,
+      "learning_rate": 0.00030807252630044534,
+      "loss": 2.8291,
+      "step": 10490
+    },
+    {
+      "epoch": 3.314933270157151,
+      "grad_norm": 0.11813030425267225,
+      "learning_rate": 0.0003075634228869808,
+      "loss": 2.7673,
+      "step": 10495
+    },
+    {
+      "epoch": 3.31651267472163,
+      "grad_norm": 0.11844596787273207,
+      "learning_rate": 0.00030705455356486844,
+      "loss": 2.7361,
+      "step": 10500
+    },
+    {
+      "epoch": 3.3180920792861093,
+      "grad_norm": 0.1072983792160873,
+      "learning_rate": 0.00030654591895312765,
+      "loss": 2.8479,
+      "step": 10505
+    },
+    {
+      "epoch": 3.3196714838505885,
+      "grad_norm": 0.09770505021791172,
+      "learning_rate": 0.0003060375196704919,
+      "loss": 2.7568,
+      "step": 10510
+    },
+    {
+      "epoch": 3.3212508884150673,
+      "grad_norm": 0.08337777054489283,
+      "learning_rate": 0.00030552935633540836,
+      "loss": 2.7652,
+      "step": 10515
+    },
+    {
+      "epoch": 3.3228302929795466,
+      "grad_norm": 0.09925597058384632,
+      "learning_rate": 0.0003050214295660373,
+      "loss": 2.7135,
+      "step": 10520
+    },
+    {
+      "epoch": 3.324409697544026,
+      "grad_norm": 0.11284231096345444,
+      "learning_rate": 0.00030451373998025103,
+      "loss": 2.7402,
+      "step": 10525
+    },
+    {
+      "epoch": 3.325989102108505,
+      "grad_norm": 0.1161877790093571,
+      "learning_rate": 0.0003040062881956339,
+      "loss": 2.8164,
+      "step": 10530
+    },
+    {
+      "epoch": 3.3275685066729843,
+      "grad_norm": 0.11312346073070796,
+      "learning_rate": 0.00030349907482948033,
+      "loss": 2.8635,
+      "step": 10535
+    },
+    {
+      "epoch": 3.3291479112374636,
+      "grad_norm": 0.10143003846192548,
+      "learning_rate": 0.000302992100498795,
+      "loss": 2.8035,
+      "step": 10540
+    },
+    {
+      "epoch": 3.330727315801943,
+      "grad_norm": 0.11416579286384988,
+      "learning_rate": 0.00030248536582029177,
+      "loss": 2.6941,
+      "step": 10545
+    },
+    {
+      "epoch": 3.3323067203664216,
+      "grad_norm": 0.09756681992157866,
+      "learning_rate": 0.00030197887141039296,
+      "loss": 2.7982,
+      "step": 10550
+    },
+    {
+      "epoch": 3.333886124930901,
+      "grad_norm": 0.09618299258170292,
+      "learning_rate": 0.0003014726178852286,
+      "loss": 2.8402,
+      "step": 10555
+    },
+    {
+      "epoch": 3.33546552949538,
+      "grad_norm": 0.08986382600295917,
+      "learning_rate": 0.0003009666058606361,
+      "loss": 2.8109,
+      "step": 10560
+    },
+    {
+      "epoch": 3.3370449340598594,
+      "grad_norm": 0.10525015223717464,
+      "learning_rate": 0.00030046083595215825,
+      "loss": 2.7715,
+      "step": 10565
+    },
+    {
+      "epoch": 3.3386243386243386,
+      "grad_norm": 0.0871796093885151,
+      "learning_rate": 0.0002999553087750441,
+      "loss": 2.821,
+      "step": 10570
+    },
+    {
+      "epoch": 3.340203743188818,
+      "grad_norm": 0.08747846767578411,
+      "learning_rate": 0.0002994500249442467,
+      "loss": 2.824,
+      "step": 10575
+    },
+    {
+      "epoch": 3.341783147753297,
+      "grad_norm": 0.08686435971083976,
+      "learning_rate": 0.00029894498507442404,
+      "loss": 2.7703,
+      "step": 10580
+    },
+    {
+      "epoch": 3.3433625523177763,
+      "grad_norm": 0.0948554219898756,
+      "learning_rate": 0.00029844018977993647,
+      "loss": 2.8411,
+      "step": 10585
+    },
+    {
+      "epoch": 3.3449419568822556,
+      "grad_norm": 0.09792856835877507,
+      "learning_rate": 0.00029793563967484737,
+      "loss": 2.8168,
+      "step": 10590
+    },
+    {
+      "epoch": 3.3465213614467344,
+      "grad_norm": 0.09260715681034247,
+      "learning_rate": 0.00029743133537292146,
+      "loss": 2.7013,
+      "step": 10595
+    },
+    {
+      "epoch": 3.3481007660112136,
+      "grad_norm": 0.09671818934996361,
+      "learning_rate": 0.0002969272774876246,
+      "loss": 2.7433,
+      "step": 10600
+    },
+    {
+      "epoch": 3.349680170575693,
+      "grad_norm": 0.07068673760594434,
+      "learning_rate": 0.0002964234666321229,
+      "loss": 2.8983,
+      "step": 10605
+    },
+    {
+      "epoch": 3.351259575140172,
+      "grad_norm": 0.08911111410190489,
+      "learning_rate": 0.0002959199034192823,
+      "loss": 2.7506,
+      "step": 10610
+    },
+    {
+      "epoch": 3.3528389797046514,
+      "grad_norm": 0.08770121377828956,
+      "learning_rate": 0.0002954165884616669,
+      "loss": 2.6839,
+      "step": 10615
+    },
+    {
+      "epoch": 3.3544183842691306,
+      "grad_norm": 0.10305670333029622,
+      "learning_rate": 0.00029491352237153925,
+      "loss": 2.8885,
+      "step": 10620
+    },
+    {
+      "epoch": 3.35599778883361,
+      "grad_norm": 0.08706783547374539,
+      "learning_rate": 0.0002944107057608588,
+      "loss": 2.77,
+      "step": 10625
+    },
+    {
+      "epoch": 3.3575771933980887,
+      "grad_norm": 0.08626509141126583,
+      "learning_rate": 0.00029390813924128187,
+      "loss": 2.863,
+      "step": 10630
+    },
+    {
+      "epoch": 3.359156597962568,
+      "grad_norm": 0.07681826210713953,
+      "learning_rate": 0.0002934058234241604,
+      "loss": 2.8356,
+      "step": 10635
+    },
+    {
+      "epoch": 3.360736002527047,
+      "grad_norm": 0.08735499339197443,
+      "learning_rate": 0.00029290375892054144,
+      "loss": 2.8735,
+      "step": 10640
+    },
+    {
+      "epoch": 3.3623154070915264,
+      "grad_norm": 0.08461473487663433,
+      "learning_rate": 0.00029240194634116615,
+      "loss": 2.7668,
+      "step": 10645
+    },
+    {
+      "epoch": 3.3638948116560057,
+      "grad_norm": 0.08954118905206135,
+      "learning_rate": 0.00029190038629646925,
+      "loss": 2.7989,
+      "step": 10650
+    },
+    {
+      "epoch": 3.365474216220485,
+      "grad_norm": 0.08580660015308204,
+      "learning_rate": 0.0002913990793965785,
+      "loss": 2.8045,
+      "step": 10655
+    },
+    {
+      "epoch": 3.367053620784964,
+      "grad_norm": 0.09208676256006787,
+      "learning_rate": 0.00029089802625131356,
+      "loss": 2.7887,
+      "step": 10660
+    },
+    {
+      "epoch": 3.3686330253494434,
+      "grad_norm": 0.10581784088945766,
+      "learning_rate": 0.0002903972274701854,
+      "loss": 2.8631,
+      "step": 10665
+    },
+    {
+      "epoch": 3.3702124299139227,
+      "grad_norm": 0.08496701975959248,
+      "learning_rate": 0.00028989668366239557,
+      "loss": 2.9046,
+      "step": 10670
+    },
+    {
+      "epoch": 3.3717918344784015,
+      "grad_norm": 0.09830241330937724,
+      "learning_rate": 0.0002893963954368357,
+      "loss": 2.6857,
+      "step": 10675
+    },
+    {
+      "epoch": 3.3733712390428807,
+      "grad_norm": 0.10063256563142661,
+      "learning_rate": 0.00028889636340208557,
+      "loss": 2.796,
+      "step": 10680
+    },
+    {
+      "epoch": 3.37495064360736,
+      "grad_norm": 0.14218562279325908,
+      "learning_rate": 0.00028839658816641483,
+      "loss": 2.8761,
+      "step": 10685
+    },
+    {
+      "epoch": 3.376530048171839,
+      "grad_norm": 0.08569913365755431,
+      "learning_rate": 0.00028789707033777956,
+      "loss": 2.7671,
+      "step": 10690
+    },
+    {
+      "epoch": 3.3781094527363185,
+      "grad_norm": 0.08385131104192822,
+      "learning_rate": 0.0002873978105238234,
+      "loss": 2.7571,
+      "step": 10695
+    },
+    {
+      "epoch": 3.3796888573007977,
+      "grad_norm": 0.09983927441112955,
+      "learning_rate": 0.00028689880933187545,
+      "loss": 2.7044,
+      "step": 10700
+    },
+    {
+      "epoch": 3.381268261865277,
+      "grad_norm": 0.10194084669294304,
+      "learning_rate": 0.00028640006736895045,
+      "loss": 2.8349,
+      "step": 10705
+    },
+    {
+      "epoch": 3.3828476664297558,
+      "grad_norm": 0.0826488277966761,
+      "learning_rate": 0.0002859015852417485,
+      "loss": 2.7198,
+      "step": 10710
+    },
+    {
+      "epoch": 3.384427070994235,
+      "grad_norm": 0.08591632570056654,
+      "learning_rate": 0.00028540336355665287,
+      "loss": 2.7896,
+      "step": 10715
+    },
+    {
+      "epoch": 3.3860064755587143,
+      "grad_norm": 0.10755343923794718,
+      "learning_rate": 0.00028490540291972987,
+      "loss": 2.6768,
+      "step": 10720
+    },
+    {
+      "epoch": 3.3875858801231935,
+      "grad_norm": 0.0992873652173549,
+      "learning_rate": 0.00028440770393672876,
+      "loss": 2.7535,
+      "step": 10725
+    },
+    {
+      "epoch": 3.3891652846876728,
+      "grad_norm": 0.0907066773586846,
+      "learning_rate": 0.00028391026721308045,
+      "loss": 2.6961,
+      "step": 10730
+    },
+    {
+      "epoch": 3.390744689252152,
+      "grad_norm": 0.08815581129035133,
+      "learning_rate": 0.0002834130933538965,
+      "loss": 2.9166,
+      "step": 10735
+    },
+    {
+      "epoch": 3.3923240938166312,
+      "grad_norm": 0.09925993594935574,
+      "learning_rate": 0.00028291618296396903,
+      "loss": 2.8347,
+      "step": 10740
+    },
+    {
+      "epoch": 3.3939034983811105,
+      "grad_norm": 0.08854222364072202,
+      "learning_rate": 0.00028241953664776947,
+      "loss": 2.7158,
+      "step": 10745
+    },
+    {
+      "epoch": 3.3954829029455897,
+      "grad_norm": 0.10593514231411359,
+      "learning_rate": 0.00028192315500944815,
+      "loss": 2.75,
+      "step": 10750
+    },
+    {
+      "epoch": 3.3970623075100685,
+      "grad_norm": 0.08871145686865252,
+      "learning_rate": 0.0002814270386528335,
+      "loss": 2.8973,
+      "step": 10755
+    },
+    {
+      "epoch": 3.398641712074548,
+      "grad_norm": 0.10109032402250467,
+      "learning_rate": 0.00028093118818143056,
+      "loss": 2.8363,
+      "step": 10760
+    },
+    {
+      "epoch": 3.400221116639027,
+      "grad_norm": 0.08939857135079345,
+      "learning_rate": 0.000280435604198422,
+      "loss": 2.7487,
+      "step": 10765
+    },
+    {
+      "epoch": 3.4018005212035063,
+      "grad_norm": 0.07345794157408284,
+      "learning_rate": 0.00027994028730666566,
+      "loss": 2.7973,
+      "step": 10770
+    },
+    {
+      "epoch": 3.4033799257679855,
+      "grad_norm": 0.0976729103298637,
+      "learning_rate": 0.0002794452381086947,
+      "loss": 2.6475,
+      "step": 10775
+    },
+    {
+      "epoch": 3.404959330332465,
+      "grad_norm": 0.10542577295220248,
+      "learning_rate": 0.0002789504572067163,
+      "loss": 2.8145,
+      "step": 10780
+    },
+    {
+      "epoch": 3.406538734896944,
+      "grad_norm": 0.08879000218903442,
+      "learning_rate": 0.00027845594520261143,
+      "loss": 2.8503,
+      "step": 10785
+    },
+    {
+      "epoch": 3.408118139461423,
+      "grad_norm": 0.08992673674464459,
+      "learning_rate": 0.00027796170269793447,
+      "loss": 2.774,
+      "step": 10790
+    },
+    {
+      "epoch": 3.409697544025902,
+      "grad_norm": 0.08304893286281426,
+      "learning_rate": 0.0002774677302939115,
+      "loss": 2.8154,
+      "step": 10795
+    },
+    {
+      "epoch": 3.4112769485903813,
+      "grad_norm": 0.09054083293396371,
+      "learning_rate": 0.00027697402859143974,
+      "loss": 2.7216,
+      "step": 10800
+    },
+    {
+      "epoch": 3.4128563531548606,
+      "grad_norm": 0.11485679109368063,
+      "learning_rate": 0.0002764805981910875,
+      "loss": 2.779,
+      "step": 10805
+    },
+    {
+      "epoch": 3.41443575771934,
+      "grad_norm": 0.09986355742999838,
+      "learning_rate": 0.0002759874396930932,
+      "loss": 2.777,
+      "step": 10810
+    },
+    {
+      "epoch": 3.416015162283819,
+      "grad_norm": 0.08634040176509505,
+      "learning_rate": 0.0002754945536973642,
+      "loss": 2.7817,
+      "step": 10815
+    },
+    {
+      "epoch": 3.4175945668482983,
+      "grad_norm": 0.1387185273790178,
+      "learning_rate": 0.0002750019408034765,
+      "loss": 2.8803,
+      "step": 10820
+    },
+    {
+      "epoch": 3.4191739714127776,
+      "grad_norm": 0.09169017218900806,
+      "learning_rate": 0.00027450960161067386,
+      "loss": 2.8645,
+      "step": 10825
+    },
+    {
+      "epoch": 3.420753375977257,
+      "grad_norm": 0.09312706836528975,
+      "learning_rate": 0.00027401753671786713,
+      "loss": 2.7441,
+      "step": 10830
+    },
+    {
+      "epoch": 3.4223327805417356,
+      "grad_norm": 0.07638921895269593,
+      "learning_rate": 0.0002735257467236333,
+      "loss": 2.725,
+      "step": 10835
+    },
+    {
+      "epoch": 3.423912185106215,
+      "grad_norm": 0.09152532966595302,
+      "learning_rate": 0.0002730342322262153,
+      "loss": 2.8329,
+      "step": 10840
+    },
+    {
+      "epoch": 3.425491589670694,
+      "grad_norm": 0.10319959902796268,
+      "learning_rate": 0.0002725429938235207,
+      "loss": 2.7832,
+      "step": 10845
+    },
+    {
+      "epoch": 3.4270709942351734,
+      "grad_norm": 0.085158962330912,
+      "learning_rate": 0.00027205203211312114,
+      "loss": 2.7716,
+      "step": 10850
+    },
+    {
+      "epoch": 3.4286503987996526,
+      "grad_norm": 0.09396859809862895,
+      "learning_rate": 0.00027156134769225213,
+      "loss": 2.678,
+      "step": 10855
+    },
+    {
+      "epoch": 3.430229803364132,
+      "grad_norm": 0.09498066575529822,
+      "learning_rate": 0.0002710709411578108,
+      "loss": 2.7378,
+      "step": 10860
+    },
+    {
+      "epoch": 3.431809207928611,
+      "grad_norm": 0.10037601950402124,
+      "learning_rate": 0.0002705808131063576,
+      "loss": 2.8111,
+      "step": 10865
+    },
+    {
+      "epoch": 3.43338861249309,
+      "grad_norm": 0.09284252009302106,
+      "learning_rate": 0.0002700909641341136,
+      "loss": 2.7683,
+      "step": 10870
+    },
+    {
+      "epoch": 3.434968017057569,
+      "grad_norm": 0.09157793859610316,
+      "learning_rate": 0.00026960139483696,
+      "loss": 2.7061,
+      "step": 10875
+    },
+    {
+      "epoch": 3.4365474216220484,
+      "grad_norm": 0.10394576925136403,
+      "learning_rate": 0.00026911210581043827,
+      "loss": 2.7692,
+      "step": 10880
+    },
+    {
+      "epoch": 3.4381268261865277,
+      "grad_norm": 0.08246710987942577,
+      "learning_rate": 0.0002686230976497487,
+      "loss": 2.7873,
+      "step": 10885
+    },
+    {
+      "epoch": 3.439706230751007,
+      "grad_norm": 0.08815358578578188,
+      "learning_rate": 0.0002681343709497506,
+      "loss": 2.7813,
+      "step": 10890
+    },
+    {
+      "epoch": 3.441285635315486,
+      "grad_norm": 0.15415479096951565,
+      "learning_rate": 0.00026764592630495966,
+      "loss": 2.8441,
+      "step": 10895
+    },
+    {
+      "epoch": 3.4428650398799654,
+      "grad_norm": 0.0924510602430328,
+      "learning_rate": 0.0002671577643095495,
+      "loss": 2.7931,
+      "step": 10900
+    },
+    {
+      "epoch": 3.4444444444444446,
+      "grad_norm": 0.13299265963339613,
+      "learning_rate": 0.0002666698855573494,
+      "loss": 2.7371,
+      "step": 10905
+    },
+    {
+      "epoch": 3.4460238490089234,
+      "grad_norm": 0.0908679979361624,
+      "learning_rate": 0.0002661822906418443,
+      "loss": 2.8087,
+      "step": 10910
+    },
+    {
+      "epoch": 3.4476032535734027,
+      "grad_norm": 0.08558429963702503,
+      "learning_rate": 0.00026569498015617375,
+      "loss": 2.8402,
+      "step": 10915
+    },
+    {
+      "epoch": 3.449182658137882,
+      "grad_norm": 0.11112704951636004,
+      "learning_rate": 0.0002652079546931314,
+      "loss": 2.7451,
+      "step": 10920
+    },
+    {
+      "epoch": 3.450762062702361,
+      "grad_norm": 0.10937437672718385,
+      "learning_rate": 0.0002647212148451641,
+      "loss": 2.7759,
+      "step": 10925
+    },
+    {
+      "epoch": 3.4523414672668404,
+      "grad_norm": 0.09702527426316783,
+      "learning_rate": 0.0002642347612043713,
+      "loss": 2.7748,
+      "step": 10930
+    },
+    {
+      "epoch": 3.4539208718313197,
+      "grad_norm": 0.10383828392331206,
+      "learning_rate": 0.00026374859436250443,
+      "loss": 2.7683,
+      "step": 10935
+    },
+    {
+      "epoch": 3.455500276395799,
+      "grad_norm": 0.07690619190890755,
+      "learning_rate": 0.00026326271491096533,
+      "loss": 2.6818,
+      "step": 10940
+    },
+    {
+      "epoch": 3.4570796809602777,
+      "grad_norm": 0.08124921569841642,
+      "learning_rate": 0.00026277712344080744,
+      "loss": 2.8036,
+      "step": 10945
+    },
+    {
+      "epoch": 3.458659085524757,
+      "grad_norm": 0.08092124730352737,
+      "learning_rate": 0.0002622918205427332,
+      "loss": 2.7491,
+      "step": 10950
+    },
+    {
+      "epoch": 3.4602384900892362,
+      "grad_norm": 0.08633152504961933,
+      "learning_rate": 0.0002618068068070937,
+      "loss": 2.7884,
+      "step": 10955
+    },
+    {
+      "epoch": 3.4618178946537155,
+      "grad_norm": 0.08300162993971244,
+      "learning_rate": 0.0002613220828238887,
+      "loss": 2.7962,
+      "step": 10960
+    },
+    {
+      "epoch": 3.4633972992181947,
+      "grad_norm": 0.14525904455931402,
+      "learning_rate": 0.0002608376491827653,
+      "loss": 2.7566,
+      "step": 10965
+    },
+    {
+      "epoch": 3.464976703782674,
+      "grad_norm": 0.11516134927374742,
+      "learning_rate": 0.00026035350647301826,
+      "loss": 2.7814,
+      "step": 10970
+    },
+    {
+      "epoch": 3.466556108347153,
+      "grad_norm": 0.08151651771920508,
+      "learning_rate": 0.00025986965528358686,
+      "loss": 2.8004,
+      "step": 10975
+    },
+    {
+      "epoch": 3.4681355129116325,
+      "grad_norm": 0.09385265602259785,
+      "learning_rate": 0.000259386096203057,
+      "loss": 2.7664,
+      "step": 10980
+    },
+    {
+      "epoch": 3.4697149174761117,
+      "grad_norm": 0.08359061385939004,
+      "learning_rate": 0.0002589028298196587,
+      "loss": 2.8729,
+      "step": 10985
+    },
+    {
+      "epoch": 3.4712943220405905,
+      "grad_norm": 0.09326773515852947,
+      "learning_rate": 0.00025841985672126627,
+      "loss": 2.8564,
+      "step": 10990
+    },
+    {
+      "epoch": 3.4728737266050698,
+      "grad_norm": 0.09217841367489349,
+      "learning_rate": 0.0002579371774953969,
+      "loss": 2.8335,
+      "step": 10995
+    },
+    {
+      "epoch": 3.474453131169549,
+      "grad_norm": 0.09981611640743568,
+      "learning_rate": 0.00025745479272921035,
+      "loss": 2.8001,
+      "step": 11000
+    },
+    {
+      "epoch": 3.4760325357340283,
+      "grad_norm": 0.08179164642245464,
+      "learning_rate": 0.00025697270300950847,
+      "loss": 2.9299,
+      "step": 11005
+    },
+    {
+      "epoch": 3.4776119402985075,
+      "grad_norm": 0.091949551376435,
+      "learning_rate": 0.0002564909089227339,
+      "loss": 2.6954,
+      "step": 11010
+    },
+    {
+      "epoch": 3.4791913448629868,
+      "grad_norm": 0.09166900983954186,
+      "learning_rate": 0.00025600941105496976,
+      "loss": 2.8015,
+      "step": 11015
+    },
+    {
+      "epoch": 3.480770749427466,
+      "grad_norm": 0.0795310556658113,
+      "learning_rate": 0.0002555282099919389,
+      "loss": 2.7736,
+      "step": 11020
+    },
+    {
+      "epoch": 3.482350153991945,
+      "grad_norm": 0.0900562553354306,
+      "learning_rate": 0.0002550473063190031,
+      "loss": 2.7664,
+      "step": 11025
+    },
+    {
+      "epoch": 3.483929558556424,
+      "grad_norm": 0.09407254281721275,
+      "learning_rate": 0.0002545667006211623,
+      "loss": 2.8369,
+      "step": 11030
+    },
+    {
+      "epoch": 3.4855089631209033,
+      "grad_norm": 0.07723378262992991,
+      "learning_rate": 0.00025408639348305375,
+      "loss": 2.832,
+      "step": 11035
+    },
+    {
+      "epoch": 3.4870883676853826,
+      "grad_norm": 0.07979269610322459,
+      "learning_rate": 0.00025360638548895177,
+      "loss": 2.7741,
+      "step": 11040
+    },
+    {
+      "epoch": 3.488667772249862,
+      "grad_norm": 0.09323028206334563,
+      "learning_rate": 0.00025312667722276707,
+      "loss": 2.7649,
+      "step": 11045
+    },
+    {
+      "epoch": 3.490247176814341,
+      "grad_norm": 0.12105776258792499,
+      "learning_rate": 0.00025264726926804546,
+      "loss": 2.8477,
+      "step": 11050
+    },
+    {
+      "epoch": 3.4918265813788203,
+      "grad_norm": 0.1105682808381471,
+      "learning_rate": 0.0002521681622079672,
+      "loss": 2.7613,
+      "step": 11055
+    },
+    {
+      "epoch": 3.4934059859432995,
+      "grad_norm": 0.11059096312067039,
+      "learning_rate": 0.00025168935662534675,
+      "loss": 2.7119,
+      "step": 11060
+    },
+    {
+      "epoch": 3.494985390507779,
+      "grad_norm": 0.09524605317480227,
+      "learning_rate": 0.0002512108531026318,
+      "loss": 2.7114,
+      "step": 11065
+    },
+    {
+      "epoch": 3.4965647950722576,
+      "grad_norm": 0.0990559059014925,
+      "learning_rate": 0.00025073265222190304,
+      "loss": 2.8923,
+      "step": 11070
+    },
+    {
+      "epoch": 3.498144199636737,
+      "grad_norm": 0.1099256285770096,
+      "learning_rate": 0.00025025475456487217,
+      "loss": 2.7998,
+      "step": 11075
+    },
+    {
+      "epoch": 3.499723604201216,
+      "grad_norm": 0.10100376221497276,
+      "learning_rate": 0.0002497771607128826,
+      "loss": 2.8201,
+      "step": 11080
+    },
+    {
+      "epoch": 3.5013030087656953,
+      "grad_norm": 0.1052486162156182,
+      "learning_rate": 0.0002492998712469079,
+      "loss": 2.7215,
+      "step": 11085
+    },
+    {
+      "epoch": 3.5028824133301746,
+      "grad_norm": 0.09928927738606914,
+      "learning_rate": 0.00024882288674755196,
+      "loss": 2.7733,
+      "step": 11090
+    },
+    {
+      "epoch": 3.504461817894654,
+      "grad_norm": 0.09818223447896932,
+      "learning_rate": 0.0002483462077950464,
+      "loss": 2.7213,
+      "step": 11095
+    },
+    {
+      "epoch": 3.5060412224591326,
+      "grad_norm": 0.08596510791164912,
+      "learning_rate": 0.0002478698349692527,
+      "loss": 2.809,
+      "step": 11100
+    },
+    {
+      "epoch": 3.507620627023612,
+      "grad_norm": 0.0846114049791052,
+      "learning_rate": 0.000247393768849659,
+      "loss": 2.746,
+      "step": 11105
+    },
+    {
+      "epoch": 3.509200031588091,
+      "grad_norm": 0.09427065622660098,
+      "learning_rate": 0.00024691801001538083,
+      "loss": 2.7923,
+      "step": 11110
+    },
+    {
+      "epoch": 3.5107794361525704,
+      "grad_norm": 0.10380969457280094,
+      "learning_rate": 0.00024644255904515916,
+      "loss": 2.7766,
+      "step": 11115
+    },
+    {
+      "epoch": 3.5123588407170496,
+      "grad_norm": 0.09213310919614595,
+      "learning_rate": 0.0002459674165173611,
+      "loss": 2.8597,
+      "step": 11120
+    },
+    {
+      "epoch": 3.513938245281529,
+      "grad_norm": 0.09392699835767081,
+      "learning_rate": 0.00024549258300997866,
+      "loss": 2.8485,
+      "step": 11125
+    },
+    {
+      "epoch": 3.515517649846008,
+      "grad_norm": 0.08174692469312347,
+      "learning_rate": 0.0002450180591006278,
+      "loss": 2.7071,
+      "step": 11130
+    },
+    {
+      "epoch": 3.5170970544104874,
+      "grad_norm": 0.09132574355106149,
+      "learning_rate": 0.00024454384536654733,
+      "loss": 2.8601,
+      "step": 11135
+    },
+    {
+      "epoch": 3.5186764589749666,
+      "grad_norm": 0.09142503988901095,
+      "learning_rate": 0.0002440699423845994,
+      "loss": 2.8217,
+      "step": 11140
+    },
+    {
+      "epoch": 3.520255863539446,
+      "grad_norm": 0.10370066500341804,
+      "learning_rate": 0.00024359635073126768,
+      "loss": 2.7277,
+      "step": 11145
+    },
+    {
+      "epoch": 3.5218352681039247,
+      "grad_norm": 0.12108542661119506,
+      "learning_rate": 0.00024312307098265802,
+      "loss": 2.7941,
+      "step": 11150
+    },
+    {
+      "epoch": 3.523414672668404,
+      "grad_norm": 0.09631209388704579,
+      "learning_rate": 0.00024265010371449548,
+      "loss": 2.8293,
+      "step": 11155
+    },
+    {
+      "epoch": 3.524994077232883,
+      "grad_norm": 0.09490360968591363,
+      "learning_rate": 0.000242177449502126,
+      "loss": 2.7607,
+      "step": 11160
+    },
+    {
+      "epoch": 3.5265734817973624,
+      "grad_norm": 0.10513561014615604,
+      "learning_rate": 0.0002417051089205144,
+      "loss": 2.7314,
+      "step": 11165
+    },
+    {
+      "epoch": 3.5281528863618417,
+      "grad_norm": 0.09288770685274528,
+      "learning_rate": 0.00024123308254424397,
+      "loss": 2.668,
+      "step": 11170
+    },
+    {
+      "epoch": 3.529732290926321,
+      "grad_norm": 0.09114131498496017,
+      "learning_rate": 0.00024076137094751582,
+      "loss": 2.7992,
+      "step": 11175
+    },
+    {
+      "epoch": 3.5313116954907997,
+      "grad_norm": 0.11313615103787365,
+      "learning_rate": 0.0002402899747041481,
+      "loss": 2.7992,
+      "step": 11180
+    },
+    {
+      "epoch": 3.532891100055279,
+      "grad_norm": 0.1036181606441657,
+      "learning_rate": 0.00023981889438757538,
+      "loss": 2.7268,
+      "step": 11185
+    },
+    {
+      "epoch": 3.534470504619758,
+      "grad_norm": 0.12020951307713118,
+      "learning_rate": 0.0002393481305708481,
+      "loss": 2.8105,
+      "step": 11190
+    },
+    {
+      "epoch": 3.5360499091842374,
+      "grad_norm": 0.09703809632260463,
+      "learning_rate": 0.00023887768382663095,
+      "loss": 2.7822,
+      "step": 11195
+    },
+    {
+      "epoch": 3.5376293137487167,
+      "grad_norm": 0.09432814464265707,
+      "learning_rate": 0.000238407554727204,
+      "loss": 2.7567,
+      "step": 11200
+    },
+    {
+      "epoch": 3.539208718313196,
+      "grad_norm": 0.11908118014259843,
+      "learning_rate": 0.0002379377438444602,
+      "loss": 2.8585,
+      "step": 11205
+    },
+    {
+      "epoch": 3.540788122877675,
+      "grad_norm": 0.07999832467414614,
+      "learning_rate": 0.00023746825174990582,
+      "loss": 2.7772,
+      "step": 11210
+    },
+    {
+      "epoch": 3.5423675274421544,
+      "grad_norm": 0.0829297522285289,
+      "learning_rate": 0.0002369990790146586,
+      "loss": 2.813,
+      "step": 11215
+    },
+    {
+      "epoch": 3.5439469320066337,
+      "grad_norm": 0.08156986811427414,
+      "learning_rate": 0.00023653022620944848,
+      "loss": 2.7536,
+      "step": 11220
+    },
+    {
+      "epoch": 3.545526336571113,
+      "grad_norm": 0.09170687949810832,
+      "learning_rate": 0.00023606169390461647,
+      "loss": 2.749,
+      "step": 11225
+    },
+    {
+      "epoch": 3.5471057411355917,
+      "grad_norm": 0.09254289254155507,
+      "learning_rate": 0.00023559348267011265,
+      "loss": 2.8507,
+      "step": 11230
+    },
+    {
+      "epoch": 3.548685145700071,
+      "grad_norm": 0.08877591653132519,
+      "learning_rate": 0.00023512559307549747,
+      "loss": 2.7886,
+      "step": 11235
+    },
+    {
+      "epoch": 3.5502645502645502,
+      "grad_norm": 0.07578799768096475,
+      "learning_rate": 0.0002346580256899397,
+      "loss": 2.7956,
+      "step": 11240
+    },
+    {
+      "epoch": 3.5518439548290295,
+      "grad_norm": 0.10027186839404545,
+      "learning_rate": 0.0002341907810822163,
+      "loss": 2.7303,
+      "step": 11245
+    },
+    {
+      "epoch": 3.5534233593935087,
+      "grad_norm": 0.08779188132077755,
+      "learning_rate": 0.00023372385982071154,
+      "loss": 2.8202,
+      "step": 11250
+    },
+    {
+      "epoch": 3.555002763957988,
+      "grad_norm": 0.10738559283983715,
+      "learning_rate": 0.00023325726247341627,
+      "loss": 2.889,
+      "step": 11255
+    },
+    {
+      "epoch": 3.556582168522467,
+      "grad_norm": 0.11921392227900801,
+      "learning_rate": 0.00023279098960792745,
+      "loss": 2.7683,
+      "step": 11260
+    },
+    {
+      "epoch": 3.558161573086946,
+      "grad_norm": 0.11346109853767701,
+      "learning_rate": 0.00023232504179144725,
+      "loss": 2.8041,
+      "step": 11265
+    },
+    {
+      "epoch": 3.5597409776514253,
+      "grad_norm": 0.09307444252039754,
+      "learning_rate": 0.00023185941959078261,
+      "loss": 2.7866,
+      "step": 11270
+    },
+    {
+      "epoch": 3.5613203822159045,
+      "grad_norm": 0.08626558856259384,
+      "learning_rate": 0.00023139412357234368,
+      "loss": 2.7719,
+      "step": 11275
+    },
+    {
+      "epoch": 3.5628997867803838,
+      "grad_norm": 0.1078184947468944,
+      "learning_rate": 0.00023092915430214484,
+      "loss": 2.7441,
+      "step": 11280
+    },
+    {
+      "epoch": 3.564479191344863,
+      "grad_norm": 0.10113108741500824,
+      "learning_rate": 0.00023046451234580233,
+      "loss": 2.7064,
+      "step": 11285
+    },
+    {
+      "epoch": 3.5660585959093423,
+      "grad_norm": 0.1157179687956109,
+      "learning_rate": 0.00023000019826853463,
+      "loss": 2.7445,
+      "step": 11290
+    },
+    {
+      "epoch": 3.5676380004738215,
+      "grad_norm": 0.0927942211817932,
+      "learning_rate": 0.00022953621263516072,
+      "loss": 2.7825,
+      "step": 11295
+    },
+    {
+      "epoch": 3.5692174050383008,
+      "grad_norm": 0.10375826834603429,
+      "learning_rate": 0.00022907255601010046,
+      "loss": 2.8141,
+      "step": 11300
+    },
+    {
+      "epoch": 3.57079680960278,
+      "grad_norm": 0.09543513718767459,
+      "learning_rate": 0.000228609228957374,
+      "loss": 2.7531,
+      "step": 11305
+    },
+    {
+      "epoch": 3.572376214167259,
+      "grad_norm": 0.0890887111388023,
+      "learning_rate": 0.00022814623204059952,
+      "loss": 2.779,
+      "step": 11310
+    },
+    {
+      "epoch": 3.573955618731738,
+      "grad_norm": 0.08073292664590946,
+      "learning_rate": 0.00022768356582299432,
+      "loss": 2.6918,
+      "step": 11315
+    },
+    {
+      "epoch": 3.5755350232962173,
+      "grad_norm": 0.08676561072352555,
+      "learning_rate": 0.00022722123086737329,
+      "loss": 2.6887,
+      "step": 11320
+    },
+    {
+      "epoch": 3.5771144278606966,
+      "grad_norm": 0.09426925946308844,
+      "learning_rate": 0.0002267592277361482,
+      "loss": 2.7954,
+      "step": 11325
+    },
+    {
+      "epoch": 3.578693832425176,
+      "grad_norm": 0.09229311484095422,
+      "learning_rate": 0.00022629755699132736,
+      "loss": 2.7214,
+      "step": 11330
+    },
+    {
+      "epoch": 3.580273236989655,
+      "grad_norm": 0.08140271955266823,
+      "learning_rate": 0.00022583621919451463,
+      "loss": 2.8671,
+      "step": 11335
+    },
+    {
+      "epoch": 3.581852641554134,
+      "grad_norm": 0.09556558459606468,
+      "learning_rate": 0.00022537521490690883,
+      "loss": 2.7257,
+      "step": 11340
+    },
+    {
+      "epoch": 3.583432046118613,
+      "grad_norm": 0.10077992058409145,
+      "learning_rate": 0.00022491454468930318,
+      "loss": 2.7685,
+      "step": 11345
+    },
+    {
+      "epoch": 3.5850114506830923,
+      "grad_norm": 0.087570674890972,
+      "learning_rate": 0.0002244542091020844,
+      "loss": 2.756,
+      "step": 11350
+    },
+    {
+      "epoch": 3.5865908552475716,
+      "grad_norm": 0.10849227424982827,
+      "learning_rate": 0.0002239942087052323,
+      "loss": 2.7807,
+      "step": 11355
+    },
+    {
+      "epoch": 3.588170259812051,
+      "grad_norm": 0.08597460501625169,
+      "learning_rate": 0.00022353454405831875,
+      "loss": 2.8133,
+      "step": 11360
+    },
+    {
+      "epoch": 3.58974966437653,
+      "grad_norm": 0.0998748647754923,
+      "learning_rate": 0.00022307521572050736,
+      "loss": 2.8042,
+      "step": 11365
+    },
+    {
+      "epoch": 3.5913290689410093,
+      "grad_norm": 0.11126051646572314,
+      "learning_rate": 0.00022261622425055272,
+      "loss": 2.7533,
+      "step": 11370
+    },
+    {
+      "epoch": 3.5929084735054886,
+      "grad_norm": 0.09055924524444105,
+      "learning_rate": 0.000222157570206799,
+      "loss": 2.8279,
+      "step": 11375
+    },
+    {
+      "epoch": 3.594487878069968,
+      "grad_norm": 0.08984142006571276,
+      "learning_rate": 0.00022169925414718085,
+      "loss": 2.7877,
+      "step": 11380
+    },
+    {
+      "epoch": 3.596067282634447,
+      "grad_norm": 0.10052235488529815,
+      "learning_rate": 0.00022124127662922132,
+      "loss": 2.8896,
+      "step": 11385
+    },
+    {
+      "epoch": 3.597646687198926,
+      "grad_norm": 0.08811739473430545,
+      "learning_rate": 0.0002207836382100314,
+      "loss": 2.7709,
+      "step": 11390
+    },
+    {
+      "epoch": 3.599226091763405,
+      "grad_norm": 0.08353319058341495,
+      "learning_rate": 0.0002203263394463098,
+      "loss": 2.8369,
+      "step": 11395
+    },
+    {
+      "epoch": 3.6008054963278844,
+      "grad_norm": 0.09449205017726552,
+      "learning_rate": 0.00021986938089434217,
+      "loss": 2.8294,
+      "step": 11400
+    },
+    {
+      "epoch": 3.6023849008923636,
+      "grad_norm": 0.08941806037298423,
+      "learning_rate": 0.00021941276311000026,
+      "loss": 2.8409,
+      "step": 11405
+    },
+    {
+      "epoch": 3.603964305456843,
+      "grad_norm": 0.09097010972981263,
+      "learning_rate": 0.00021895648664874108,
+      "loss": 2.7626,
+      "step": 11410
+    },
+    {
+      "epoch": 3.605543710021322,
+      "grad_norm": 0.0952754837032473,
+      "learning_rate": 0.00021850055206560666,
+      "loss": 2.7939,
+      "step": 11415
+    },
+    {
+      "epoch": 3.607123114585801,
+      "grad_norm": 0.08485773446474305,
+      "learning_rate": 0.0002180449599152231,
+      "loss": 2.7381,
+      "step": 11420
+    },
+    {
+      "epoch": 3.60870251915028,
+      "grad_norm": 0.07453026280015991,
+      "learning_rate": 0.00021758971075179988,
+      "loss": 2.7899,
+      "step": 11425
+    },
+    {
+      "epoch": 3.6102819237147594,
+      "grad_norm": 0.09183300642945505,
+      "learning_rate": 0.0002171348051291293,
+      "loss": 2.7926,
+      "step": 11430
+    },
+    {
+      "epoch": 3.6118613282792387,
+      "grad_norm": 0.0839019053264377,
+      "learning_rate": 0.00021668024360058574,
+      "loss": 2.8769,
+      "step": 11435
+    },
+    {
+      "epoch": 3.613440732843718,
+      "grad_norm": 0.10134168322160107,
+      "learning_rate": 0.00021622602671912507,
+      "loss": 2.8372,
+      "step": 11440
+    },
+    {
+      "epoch": 3.615020137408197,
+      "grad_norm": 0.09412058984971264,
+      "learning_rate": 0.00021577215503728393,
+      "loss": 2.811,
+      "step": 11445
+    },
+    {
+      "epoch": 3.6165995419726764,
+      "grad_norm": 0.11903314838257359,
+      "learning_rate": 0.00021531862910717864,
+      "loss": 2.7716,
+      "step": 11450
+    },
+    {
+      "epoch": 3.6181789465371557,
+      "grad_norm": 0.12363190718408543,
+      "learning_rate": 0.00021486544948050524,
+      "loss": 2.8333,
+      "step": 11455
+    },
+    {
+      "epoch": 3.619758351101635,
+      "grad_norm": 0.08863366856434014,
+      "learning_rate": 0.00021441261670853884,
+      "loss": 2.6796,
+      "step": 11460
+    },
+    {
+      "epoch": 3.6213377556661137,
+      "grad_norm": 0.13499636006682605,
+      "learning_rate": 0.0002139601313421324,
+      "loss": 2.7957,
+      "step": 11465
+    },
+    {
+      "epoch": 3.622917160230593,
+      "grad_norm": 0.09507917881349866,
+      "learning_rate": 0.00021350799393171567,
+      "loss": 2.7729,
+      "step": 11470
+    },
+    {
+      "epoch": 3.624496564795072,
+      "grad_norm": 0.06858739471914442,
+      "learning_rate": 0.00021305620502729583,
+      "loss": 2.7611,
+      "step": 11475
+    },
+    {
+      "epoch": 3.6260759693595515,
+      "grad_norm": 0.09214856126188252,
+      "learning_rate": 0.00021260476517845573,
+      "loss": 2.7689,
+      "step": 11480
+    },
+    {
+      "epoch": 3.6276553739240307,
+      "grad_norm": 0.09131495394135357,
+      "learning_rate": 0.0002121536749343544,
+      "loss": 2.8118,
+      "step": 11485
+    },
+    {
+      "epoch": 3.62923477848851,
+      "grad_norm": 0.07816970729114443,
+      "learning_rate": 0.00021170293484372427,
+      "loss": 2.7645,
+      "step": 11490
+    },
+    {
+      "epoch": 3.6308141830529888,
+      "grad_norm": 0.10929095968764148,
+      "learning_rate": 0.00021125254545487283,
+      "loss": 2.7951,
+      "step": 11495
+    },
+    {
+      "epoch": 3.632393587617468,
+      "grad_norm": 0.10301738196087953,
+      "learning_rate": 0.00021080250731568057,
+      "loss": 2.867,
+      "step": 11500
+    },
+    {
+      "epoch": 3.6339729921819472,
+      "grad_norm": 0.08960227429034275,
+      "learning_rate": 0.00021035282097360086,
+      "loss": 2.8183,
+      "step": 11505
+    },
+    {
+      "epoch": 3.6355523967464265,
+      "grad_norm": 0.08737624858384084,
+      "learning_rate": 0.00020990348697565896,
+      "loss": 2.8485,
+      "step": 11510
+    },
+    {
+      "epoch": 3.6371318013109057,
+      "grad_norm": 0.07791929158219078,
+      "learning_rate": 0.00020945450586845165,
+      "loss": 2.8107,
+      "step": 11515
+    },
+    {
+      "epoch": 3.638711205875385,
+      "grad_norm": 0.08913956949374928,
+      "learning_rate": 0.00020900587819814637,
+      "loss": 2.7753,
+      "step": 11520
+    },
+    {
+      "epoch": 3.6402906104398642,
+      "grad_norm": 0.087704377877188,
+      "learning_rate": 0.0002085576045104808,
+      "loss": 2.8193,
+      "step": 11525
+    },
+    {
+      "epoch": 3.6418700150043435,
+      "grad_norm": 0.09306829106179089,
+      "learning_rate": 0.00020810968535076125,
+      "loss": 2.8075,
+      "step": 11530
+    },
+    {
+      "epoch": 3.6434494195688227,
+      "grad_norm": 0.08484563394410759,
+      "learning_rate": 0.00020766212126386397,
+      "loss": 2.6954,
+      "step": 11535
+    },
+    {
+      "epoch": 3.645028824133302,
+      "grad_norm": 0.0798912479090602,
+      "learning_rate": 0.00020721491279423242,
+      "loss": 2.7521,
+      "step": 11540
+    },
+    {
+      "epoch": 3.646608228697781,
+      "grad_norm": 0.08750716646371902,
+      "learning_rate": 0.000206768060485878,
+      "loss": 2.8699,
+      "step": 11545
+    },
+    {
+      "epoch": 3.64818763326226,
+      "grad_norm": 0.09061907295127078,
+      "learning_rate": 0.00020632156488237808,
+      "loss": 2.8274,
+      "step": 11550
+    },
+    {
+      "epoch": 3.6497670378267393,
+      "grad_norm": 0.09301703574210217,
+      "learning_rate": 0.00020587542652687662,
+      "loss": 2.7568,
+      "step": 11555
+    },
+    {
+      "epoch": 3.6513464423912185,
+      "grad_norm": 0.09156106394288291,
+      "learning_rate": 0.0002054296459620834,
+      "loss": 2.8077,
+      "step": 11560
+    },
+    {
+      "epoch": 3.6529258469556978,
+      "grad_norm": 0.08949852492673643,
+      "learning_rate": 0.00020498422373027247,
+      "loss": 2.7622,
+      "step": 11565
+    },
+    {
+      "epoch": 3.654505251520177,
+      "grad_norm": 0.10457491176962785,
+      "learning_rate": 0.00020453916037328174,
+      "loss": 2.7507,
+      "step": 11570
+    },
+    {
+      "epoch": 3.656084656084656,
+      "grad_norm": 0.08779017195219393,
+      "learning_rate": 0.000204094456432513,
+      "loss": 2.8668,
+      "step": 11575
+    },
+    {
+      "epoch": 3.657664060649135,
+      "grad_norm": 0.09069030608376441,
+      "learning_rate": 0.00020365011244893077,
+      "loss": 2.8472,
+      "step": 11580
+    },
+    {
+      "epoch": 3.6592434652136143,
+      "grad_norm": 0.08884167819693801,
+      "learning_rate": 0.00020320612896306158,
+      "loss": 2.7427,
+      "step": 11585
+    },
+    {
+      "epoch": 3.6608228697780936,
+      "grad_norm": 0.08189253953823854,
+      "learning_rate": 0.00020276250651499346,
+      "loss": 2.7255,
+      "step": 11590
+    },
+    {
+      "epoch": 3.662402274342573,
+      "grad_norm": 0.06780471843155822,
+      "learning_rate": 0.00020231924564437527,
+      "loss": 2.8178,
+      "step": 11595
+    },
+    {
+      "epoch": 3.663981678907052,
+      "grad_norm": 0.08192279195750907,
+      "learning_rate": 0.000201876346890416,
+      "loss": 2.6784,
+      "step": 11600
+    },
+    {
+      "epoch": 3.6655610834715313,
+      "grad_norm": 0.08128488744507092,
+      "learning_rate": 0.00020143381079188444,
+      "loss": 2.799,
+      "step": 11605
+    },
+    {
+      "epoch": 3.6671404880360106,
+      "grad_norm": 0.09287388991092453,
+      "learning_rate": 0.0002009916378871074,
+      "loss": 2.8106,
+      "step": 11610
+    },
+    {
+      "epoch": 3.66871989260049,
+      "grad_norm": 0.06955081594717988,
+      "learning_rate": 0.00020054982871397083,
+      "loss": 2.6967,
+      "step": 11615
+    },
+    {
+      "epoch": 3.670299297164969,
+      "grad_norm": 0.07732653629557824,
+      "learning_rate": 0.00020010838380991774,
+      "loss": 2.7274,
+      "step": 11620
+    },
+    {
+      "epoch": 3.671878701729448,
+      "grad_norm": 0.08847447417470254,
+      "learning_rate": 0.00019966730371194825,
+      "loss": 2.8347,
+      "step": 11625
+    },
+    {
+      "epoch": 3.673458106293927,
+      "grad_norm": 0.08323413983065359,
+      "learning_rate": 0.00019922658895661817,
+      "loss": 2.8466,
+      "step": 11630
+    },
+    {
+      "epoch": 3.6750375108584064,
+      "grad_norm": 0.07423365005518999,
+      "learning_rate": 0.00019878624008003927,
+      "loss": 2.7574,
+      "step": 11635
+    },
+    {
+      "epoch": 3.6766169154228856,
+      "grad_norm": 0.08596050904098117,
+      "learning_rate": 0.00019834625761787862,
+      "loss": 2.7952,
+      "step": 11640
+    },
+    {
+      "epoch": 3.678196319987365,
+      "grad_norm": 0.08447539960506971,
+      "learning_rate": 0.00019790664210535714,
+      "loss": 2.8661,
+      "step": 11645
+    },
+    {
+      "epoch": 3.679775724551844,
+      "grad_norm": 0.08848325804957861,
+      "learning_rate": 0.00019746739407724913,
+      "loss": 2.8027,
+      "step": 11650
+    },
+    {
+      "epoch": 3.681355129116323,
+      "grad_norm": 0.09719722365226477,
+      "learning_rate": 0.00019702851406788225,
+      "loss": 2.7186,
+      "step": 11655
+    },
+    {
+      "epoch": 3.682934533680802,
+      "grad_norm": 0.10118128059502188,
+      "learning_rate": 0.00019659000261113642,
+      "loss": 2.8854,
+      "step": 11660
+    },
+    {
+      "epoch": 3.6845139382452814,
+      "grad_norm": 0.07497857561322374,
+      "learning_rate": 0.00019615186024044313,
+      "loss": 2.8331,
+      "step": 11665
+    },
+    {
+      "epoch": 3.6860933428097606,
+      "grad_norm": 0.10780605274986786,
+      "learning_rate": 0.00019571408748878495,
+      "loss": 2.72,
+      "step": 11670
+    },
+    {
+      "epoch": 3.68767274737424,
+      "grad_norm": 0.09273671983890912,
+      "learning_rate": 0.00019527668488869484,
+      "loss": 2.7444,
+      "step": 11675
+    },
+    {
+      "epoch": 3.689252151938719,
+      "grad_norm": 0.09267158414129262,
+      "learning_rate": 0.00019483965297225543,
+      "loss": 2.7272,
+      "step": 11680
+    },
+    {
+      "epoch": 3.6908315565031984,
+      "grad_norm": 0.09858600391860631,
+      "learning_rate": 0.00019440299227109853,
+      "loss": 2.679,
+      "step": 11685
+    },
+    {
+      "epoch": 3.6924109610676776,
+      "grad_norm": 0.08707185882765187,
+      "learning_rate": 0.00019396670331640425,
+      "loss": 2.7163,
+      "step": 11690
+    },
+    {
+      "epoch": 3.693990365632157,
+      "grad_norm": 0.08386738183816067,
+      "learning_rate": 0.00019353078663890056,
+      "loss": 2.7726,
+      "step": 11695
+    },
+    {
+      "epoch": 3.695569770196636,
+      "grad_norm": 0.09285610818152679,
+      "learning_rate": 0.00019309524276886258,
+      "loss": 2.7636,
+      "step": 11700
+    },
+    {
+      "epoch": 3.697149174761115,
+      "grad_norm": 0.08276309730190091,
+      "learning_rate": 0.00019266007223611205,
+      "loss": 2.6908,
+      "step": 11705
+    },
+    {
+      "epoch": 3.698728579325594,
+      "grad_norm": 0.09344772692451785,
+      "learning_rate": 0.00019222527557001583,
+      "loss": 2.7216,
+      "step": 11710
+    },
+    {
+      "epoch": 3.7003079838900734,
+      "grad_norm": 0.08138562850581343,
+      "learning_rate": 0.00019179085329948725,
+      "loss": 2.7707,
+      "step": 11715
+    },
+    {
+      "epoch": 3.7018873884545527,
+      "grad_norm": 0.09182017905548755,
+      "learning_rate": 0.00019135680595298315,
+      "loss": 2.7875,
+      "step": 11720
+    },
+    {
+      "epoch": 3.703466793019032,
+      "grad_norm": 0.08299814767009533,
+      "learning_rate": 0.00019092313405850503,
+      "loss": 2.7206,
+      "step": 11725
+    },
+    {
+      "epoch": 3.705046197583511,
+      "grad_norm": 0.1004891033089223,
+      "learning_rate": 0.00019048983814359683,
+      "loss": 2.7871,
+      "step": 11730
+    },
+    {
+      "epoch": 3.70662560214799,
+      "grad_norm": 0.08497364732744732,
+      "learning_rate": 0.0001900569187353458,
+      "loss": 2.7905,
+      "step": 11735
+    },
+    {
+      "epoch": 3.708205006712469,
+      "grad_norm": 0.10208046299317208,
+      "learning_rate": 0.00018962437636038093,
+      "loss": 2.7942,
+      "step": 11740
+    },
+    {
+      "epoch": 3.7097844112769485,
+      "grad_norm": 0.08901693082989817,
+      "learning_rate": 0.0001891922115448727,
+      "loss": 2.8117,
+      "step": 11745
+    },
+    {
+      "epoch": 3.7113638158414277,
+      "grad_norm": 0.07914149877510862,
+      "learning_rate": 0.00018876042481453221,
+      "loss": 2.8137,
+      "step": 11750
+    },
+    {
+      "epoch": 3.712943220405907,
+      "grad_norm": 0.09657293929383366,
+      "learning_rate": 0.00018832901669461056,
+      "loss": 2.8863,
+      "step": 11755
+    },
+    {
+      "epoch": 3.714522624970386,
+      "grad_norm": 0.08524187526118726,
+      "learning_rate": 0.0001878979877098984,
+      "loss": 2.7986,
+      "step": 11760
+    },
+    {
+      "epoch": 3.7161020295348655,
+      "grad_norm": 0.0893121969017454,
+      "learning_rate": 0.0001874673383847252,
+      "loss": 2.8444,
+      "step": 11765
+    },
+    {
+      "epoch": 3.7176814340993447,
+      "grad_norm": 0.07737661971954736,
+      "learning_rate": 0.00018703706924295849,
+      "loss": 2.8618,
+      "step": 11770
+    },
+    {
+      "epoch": 3.719260838663824,
+      "grad_norm": 0.11339322775943667,
+      "learning_rate": 0.00018660718080800337,
+      "loss": 2.6354,
+      "step": 11775
+    },
+    {
+      "epoch": 3.720840243228303,
+      "grad_norm": 0.09029538393369561,
+      "learning_rate": 0.0001861776736028018,
+      "loss": 2.7916,
+      "step": 11780
+    },
+    {
+      "epoch": 3.722419647792782,
+      "grad_norm": 0.09222139162095129,
+      "learning_rate": 0.00018574854814983228,
+      "loss": 2.7123,
+      "step": 11785
+    },
+    {
+      "epoch": 3.7239990523572613,
+      "grad_norm": 0.09272711747658678,
+      "learning_rate": 0.000185319804971108,
+      "loss": 2.7622,
+      "step": 11790
+    },
+    {
+      "epoch": 3.7255784569217405,
+      "grad_norm": 0.09311251342299524,
+      "learning_rate": 0.0001848914445881784,
+      "loss": 2.6836,
+      "step": 11795
+    },
+    {
+      "epoch": 3.7271578614862197,
+      "grad_norm": 0.07043590075001643,
+      "learning_rate": 0.00018446346752212662,
+      "loss": 2.757,
+      "step": 11800
+    },
+    {
+      "epoch": 3.728737266050699,
+      "grad_norm": 0.07975371796499488,
+      "learning_rate": 0.00018403587429356916,
+      "loss": 2.7193,
+      "step": 11805
+    },
+    {
+      "epoch": 3.7303166706151782,
+      "grad_norm": 0.08351856594277221,
+      "learning_rate": 0.00018360866542265625,
+      "loss": 2.8014,
+      "step": 11810
+    },
+    {
+      "epoch": 3.731896075179657,
+      "grad_norm": 0.08923111864867268,
+      "learning_rate": 0.00018318184142907,
+      "loss": 2.7806,
+      "step": 11815
+    },
+    {
+      "epoch": 3.7334754797441363,
+      "grad_norm": 0.0868170822441492,
+      "learning_rate": 0.0001827554028320252,
+      "loss": 2.7466,
+      "step": 11820
+    },
+    {
+      "epoch": 3.7350548843086155,
+      "grad_norm": 0.09740873126793242,
+      "learning_rate": 0.0001823293501502667,
+      "loss": 2.6942,
+      "step": 11825
+    },
+    {
+      "epoch": 3.736634288873095,
+      "grad_norm": 0.09994819493713986,
+      "learning_rate": 0.00018190368390207063,
+      "loss": 2.8106,
+      "step": 11830
+    },
+    {
+      "epoch": 3.738213693437574,
+      "grad_norm": 0.11005739113864033,
+      "learning_rate": 0.0001814784046052429,
+      "loss": 2.8167,
+      "step": 11835
+    },
+    {
+      "epoch": 3.7397930980020533,
+      "grad_norm": 0.09021459683105348,
+      "learning_rate": 0.00018105351277711857,
+      "loss": 2.8033,
+      "step": 11840
+    },
+    {
+      "epoch": 3.7413725025665325,
+      "grad_norm": 0.10913132440175603,
+      "learning_rate": 0.00018062900893456147,
+      "loss": 2.8071,
+      "step": 11845
+    },
+    {
+      "epoch": 3.742951907131012,
+      "grad_norm": 0.1025572689490513,
+      "learning_rate": 0.00018020489359396353,
+      "loss": 2.7539,
+      "step": 11850
+    },
+    {
+      "epoch": 3.744531311695491,
+      "grad_norm": 0.11027023905298508,
+      "learning_rate": 0.00017978116727124387,
+      "loss": 2.829,
+      "step": 11855
+    },
+    {
+      "epoch": 3.74611071625997,
+      "grad_norm": 0.10401771895295076,
+      "learning_rate": 0.00017935783048184868,
+      "loss": 2.7328,
+      "step": 11860
+    },
+    {
+      "epoch": 3.747690120824449,
+      "grad_norm": 0.07543245605329015,
+      "learning_rate": 0.0001789348837407499,
+      "loss": 2.723,
+      "step": 11865
+    },
+    {
+      "epoch": 3.7492695253889283,
+      "grad_norm": 0.08070299948779706,
+      "learning_rate": 0.0001785123275624454,
+      "loss": 2.7909,
+      "step": 11870
+    },
+    {
+      "epoch": 3.7508489299534076,
+      "grad_norm": 0.08267456058643989,
+      "learning_rate": 0.00017809016246095772,
+      "loss": 2.7677,
+      "step": 11875
+    },
+    {
+      "epoch": 3.752428334517887,
+      "grad_norm": 0.07934566141525058,
+      "learning_rate": 0.0001776683889498339,
+      "loss": 2.7921,
+      "step": 11880
+    },
+    {
+      "epoch": 3.754007739082366,
+      "grad_norm": 0.07808235021780177,
+      "learning_rate": 0.00017724700754214403,
+      "loss": 2.8413,
+      "step": 11885
+    },
+    {
+      "epoch": 3.755587143646845,
+      "grad_norm": 0.0842968243503666,
+      "learning_rate": 0.00017682601875048187,
+      "loss": 2.8465,
+      "step": 11890
+    },
+    {
+      "epoch": 3.757166548211324,
+      "grad_norm": 0.08020341407065772,
+      "learning_rate": 0.00017640542308696317,
+      "loss": 2.7862,
+      "step": 11895
+    },
+    {
+      "epoch": 3.7587459527758034,
+      "grad_norm": 0.08540408558560887,
+      "learning_rate": 0.00017598522106322619,
+      "loss": 2.731,
+      "step": 11900
+    },
+    {
+      "epoch": 3.7603253573402826,
+      "grad_norm": 0.09374602539245858,
+      "learning_rate": 0.00017556541319042911,
+      "loss": 2.8661,
+      "step": 11905
+    },
+    {
+      "epoch": 3.761904761904762,
+      "grad_norm": 0.08328870658200717,
+      "learning_rate": 0.00017514599997925167,
+      "loss": 2.7336,
+      "step": 11910
+    },
+    {
+      "epoch": 3.763484166469241,
+      "grad_norm": 0.09328402945620633,
+      "learning_rate": 0.00017472698193989305,
+      "loss": 2.6946,
+      "step": 11915
+    },
+    {
+      "epoch": 3.7650635710337204,
+      "grad_norm": 0.08556856440346546,
+      "learning_rate": 0.00017430835958207185,
+      "loss": 2.8038,
+      "step": 11920
+    },
+    {
+      "epoch": 3.7666429755981996,
+      "grad_norm": 0.08605764458620771,
+      "learning_rate": 0.0001738901334150254,
+      "loss": 2.8046,
+      "step": 11925
+    },
+    {
+      "epoch": 3.768222380162679,
+      "grad_norm": 0.07969013072361213,
+      "learning_rate": 0.0001734723039475089,
+      "loss": 2.8382,
+      "step": 11930
+    },
+    {
+      "epoch": 3.769801784727158,
+      "grad_norm": 0.08754956682484208,
+      "learning_rate": 0.00017305487168779515,
+      "loss": 2.8156,
+      "step": 11935
+    },
+    {
+      "epoch": 3.771381189291637,
+      "grad_norm": 0.10029042879887169,
+      "learning_rate": 0.00017263783714367386,
+      "loss": 2.6554,
+      "step": 11940
+    },
+    {
+      "epoch": 3.772960593856116,
+      "grad_norm": 0.10504648793028643,
+      "learning_rate": 0.00017222120082245018,
+      "loss": 2.708,
+      "step": 11945
+    },
+    {
+      "epoch": 3.7745399984205954,
+      "grad_norm": 0.10522389032354743,
+      "learning_rate": 0.00017180496323094608,
+      "loss": 2.7088,
+      "step": 11950
+    },
+    {
+      "epoch": 3.7761194029850746,
+      "grad_norm": 0.09359727624397036,
+      "learning_rate": 0.00017138912487549756,
+      "loss": 2.8039,
+      "step": 11955
+    },
+    {
+      "epoch": 3.777698807549554,
+      "grad_norm": 0.11422346362497239,
+      "learning_rate": 0.00017097368626195548,
+      "loss": 2.7834,
+      "step": 11960
+    },
+    {
+      "epoch": 3.779278212114033,
+      "grad_norm": 0.08018043274815162,
+      "learning_rate": 0.00017055864789568376,
+      "loss": 2.6771,
+      "step": 11965
+    },
+    {
+      "epoch": 3.780857616678512,
+      "grad_norm": 0.082621279623475,
+      "learning_rate": 0.00017014401028156,
+      "loss": 2.7746,
+      "step": 11970
+    },
+    {
+      "epoch": 3.782437021242991,
+      "grad_norm": 0.08863049292682723,
+      "learning_rate": 0.00016972977392397444,
+      "loss": 2.7527,
+      "step": 11975
+    },
+    {
+      "epoch": 3.7840164258074704,
+      "grad_norm": 0.08119101989236832,
+      "learning_rate": 0.00016931593932682893,
+      "loss": 2.8148,
+      "step": 11980
+    },
+    {
+      "epoch": 3.7855958303719497,
+      "grad_norm": 0.09138427007636155,
+      "learning_rate": 0.0001689025069935363,
+      "loss": 2.6863,
+      "step": 11985
+    },
+    {
+      "epoch": 3.787175234936429,
+      "grad_norm": 0.09879479871575889,
+      "learning_rate": 0.00016848947742702046,
+      "loss": 2.782,
+      "step": 11990
+    },
+    {
+      "epoch": 3.788754639500908,
+      "grad_norm": 0.08980149895256091,
+      "learning_rate": 0.0001680768511297152,
+      "loss": 2.7806,
+      "step": 11995
+    },
+    {
+      "epoch": 3.7903340440653874,
+      "grad_norm": 0.08770742462877869,
+      "learning_rate": 0.00016766462860356423,
+      "loss": 2.8098,
+      "step": 12000
+    },
+    {
+      "epoch": 3.7919134486298667,
+      "grad_norm": 0.07695351477939083,
+      "learning_rate": 0.00016725281035001916,
+      "loss": 2.7902,
+      "step": 12005
+    },
+    {
+      "epoch": 3.793492853194346,
+      "grad_norm": 0.08662585356629579,
+      "learning_rate": 0.00016684139687004053,
+      "loss": 2.7889,
+      "step": 12010
+    },
+    {
+      "epoch": 3.795072257758825,
+      "grad_norm": 0.07423777411389267,
+      "learning_rate": 0.0001664303886640962,
+      "loss": 2.78,
+      "step": 12015
+    },
+    {
+      "epoch": 3.796651662323304,
+      "grad_norm": 0.08469752187569327,
+      "learning_rate": 0.00016601978623216124,
+      "loss": 2.7842,
+      "step": 12020
+    },
+    {
+      "epoch": 3.7982310668877832,
+      "grad_norm": 0.08826241649397007,
+      "learning_rate": 0.00016560959007371685,
+      "loss": 2.7844,
+      "step": 12025
+    },
+    {
+      "epoch": 3.7998104714522625,
+      "grad_norm": 0.06895911425297126,
+      "learning_rate": 0.00016519980068775025,
+      "loss": 2.7493,
+      "step": 12030
+    },
+    {
+      "epoch": 3.8013898760167417,
+      "grad_norm": 0.0776441798856278,
+      "learning_rate": 0.00016479041857275374,
+      "loss": 2.73,
+      "step": 12035
+    },
+    {
+      "epoch": 3.802969280581221,
+      "grad_norm": 0.08136608195262024,
+      "learning_rate": 0.0001643814442267243,
+      "loss": 2.7479,
+      "step": 12040
+    },
+    {
+      "epoch": 3.8045486851457,
+      "grad_norm": 0.08478394868745333,
+      "learning_rate": 0.00016397287814716243,
+      "loss": 2.714,
+      "step": 12045
+    },
+    {
+      "epoch": 3.806128089710179,
+      "grad_norm": 0.08316813607004452,
+      "learning_rate": 0.00016356472083107237,
+      "loss": 2.8003,
+      "step": 12050
+    },
+    {
+      "epoch": 3.8077074942746583,
+      "grad_norm": 0.0904850713575638,
+      "learning_rate": 0.00016315697277496138,
+      "loss": 2.833,
+      "step": 12055
+    },
+    {
+      "epoch": 3.8092868988391375,
+      "grad_norm": 0.07648298666227166,
+      "learning_rate": 0.00016274963447483854,
+      "loss": 2.8567,
+      "step": 12060
+    },
+    {
+      "epoch": 3.8108663034036168,
+      "grad_norm": 0.07238836488587672,
+      "learning_rate": 0.00016234270642621424,
+      "loss": 2.7621,
+      "step": 12065
+    },
+    {
+      "epoch": 3.812445707968096,
+      "grad_norm": 0.08909847817275786,
+      "learning_rate": 0.0001619361891241002,
+      "loss": 2.7587,
+      "step": 12070
+    },
+    {
+      "epoch": 3.8140251125325753,
+      "grad_norm": 0.07552419420026069,
+      "learning_rate": 0.00016153008306300814,
+      "loss": 2.8112,
+      "step": 12075
+    },
+    {
+      "epoch": 3.8156045170970545,
+      "grad_norm": 0.0876622533187228,
+      "learning_rate": 0.0001611243887369503,
+      "loss": 2.7627,
+      "step": 12080
+    },
+    {
+      "epoch": 3.8171839216615338,
+      "grad_norm": 0.08761521312444666,
+      "learning_rate": 0.000160719106639437,
+      "loss": 2.6827,
+      "step": 12085
+    },
+    {
+      "epoch": 3.818763326226013,
+      "grad_norm": 0.10525892878890712,
+      "learning_rate": 0.00016031423726347777,
+      "loss": 2.8289,
+      "step": 12090
+    },
+    {
+      "epoch": 3.8203427307904922,
+      "grad_norm": 0.09079503509259372,
+      "learning_rate": 0.0001599097811015799,
+      "loss": 2.7873,
+      "step": 12095
+    },
+    {
+      "epoch": 3.821922135354971,
+      "grad_norm": 0.07449785388655988,
+      "learning_rate": 0.00015950573864574808,
+      "loss": 2.7647,
+      "step": 12100
+    },
+    {
+      "epoch": 3.8235015399194503,
+      "grad_norm": 0.07493346474050362,
+      "learning_rate": 0.00015910211038748363,
+      "loss": 2.8051,
+      "step": 12105
+    },
+    {
+      "epoch": 3.8250809444839295,
+      "grad_norm": 0.08858856892366813,
+      "learning_rate": 0.0001586988968177841,
+      "loss": 2.6911,
+      "step": 12110
+    },
+    {
+      "epoch": 3.826660349048409,
+      "grad_norm": 0.09711661023425233,
+      "learning_rate": 0.0001582960984271426,
+      "loss": 2.9282,
+      "step": 12115
+    },
+    {
+      "epoch": 3.828239753612888,
+      "grad_norm": 0.08481939048470226,
+      "learning_rate": 0.00015789371570554728,
+      "loss": 2.7376,
+      "step": 12120
+    },
+    {
+      "epoch": 3.8298191581773673,
+      "grad_norm": 0.08332736816622259,
+      "learning_rate": 0.0001574917491424801,
+      "loss": 2.7929,
+      "step": 12125
+    },
+    {
+      "epoch": 3.831398562741846,
+      "grad_norm": 0.0742516035203001,
+      "learning_rate": 0.0001570901992269177,
+      "loss": 2.7426,
+      "step": 12130
+    },
+    {
+      "epoch": 3.8329779673063253,
+      "grad_norm": 0.07167536684934585,
+      "learning_rate": 0.00015668906644732917,
+      "loss": 2.7953,
+      "step": 12135
+    },
+    {
+      "epoch": 3.8345573718708046,
+      "grad_norm": 0.07976658207700503,
+      "learning_rate": 0.00015628835129167663,
+      "loss": 2.7906,
+      "step": 12140
+    },
+    {
+      "epoch": 3.836136776435284,
+      "grad_norm": 0.08991311149016072,
+      "learning_rate": 0.00015588805424741352,
+      "loss": 2.7448,
+      "step": 12145
+    },
+    {
+      "epoch": 3.837716180999763,
+      "grad_norm": 0.10173820928136434,
+      "learning_rate": 0.00015548817580148517,
+      "loss": 2.7555,
+      "step": 12150
+    },
+    {
+      "epoch": 3.8392955855642423,
+      "grad_norm": 0.06561034464832172,
+      "learning_rate": 0.00015508871644032807,
+      "loss": 2.7457,
+      "step": 12155
+    },
+    {
+      "epoch": 3.8408749901287216,
+      "grad_norm": 0.07733120609909291,
+      "learning_rate": 0.00015468967664986798,
+      "loss": 2.7664,
+      "step": 12160
+    },
+    {
+      "epoch": 3.842454394693201,
+      "grad_norm": 0.11128728912296254,
+      "learning_rate": 0.0001542910569155209,
+      "loss": 2.7219,
+      "step": 12165
+    },
+    {
+      "epoch": 3.84403379925768,
+      "grad_norm": 0.08670042557414782,
+      "learning_rate": 0.00015389285772219176,
+      "loss": 2.7775,
+      "step": 12170
+    },
+    {
+      "epoch": 3.8456132038221593,
+      "grad_norm": 0.08895531137890758,
+      "learning_rate": 0.00015349507955427378,
+      "loss": 2.7067,
+      "step": 12175
+    },
+    {
+      "epoch": 3.847192608386638,
+      "grad_norm": 0.08427607915004383,
+      "learning_rate": 0.00015309772289564806,
+      "loss": 2.7725,
+      "step": 12180
+    },
+    {
+      "epoch": 3.8487720129511174,
+      "grad_norm": 0.07370797804326407,
+      "learning_rate": 0.00015270078822968307,
+      "loss": 2.7484,
+      "step": 12185
+    },
+    {
+      "epoch": 3.8503514175155966,
+      "grad_norm": 0.07708525753337944,
+      "learning_rate": 0.00015230427603923387,
+      "loss": 2.7986,
+      "step": 12190
+    },
+    {
+      "epoch": 3.851930822080076,
+      "grad_norm": 0.08109018579361335,
+      "learning_rate": 0.00015190818680664147,
+      "loss": 2.7234,
+      "step": 12195
+    },
+    {
+      "epoch": 3.853510226644555,
+      "grad_norm": 0.08237238380223535,
+      "learning_rate": 0.00015151252101373264,
+      "loss": 2.8295,
+      "step": 12200
+    },
+    {
+      "epoch": 3.855089631209034,
+      "grad_norm": 0.08654806501913762,
+      "learning_rate": 0.00015111727914181877,
+      "loss": 2.7166,
+      "step": 12205
+    },
+    {
+      "epoch": 3.856669035773513,
+      "grad_norm": 0.07056141952098877,
+      "learning_rate": 0.00015072246167169574,
+      "loss": 2.8158,
+      "step": 12210
+    },
+    {
+      "epoch": 3.8582484403379924,
+      "grad_norm": 0.08098815316771699,
+      "learning_rate": 0.0001503280690836431,
+      "loss": 2.7101,
+      "step": 12215
+    },
+    {
+      "epoch": 3.8598278449024717,
+      "grad_norm": 0.08432444865308394,
+      "learning_rate": 0.00014993410185742373,
+      "loss": 2.723,
+      "step": 12220
+    },
+    {
+      "epoch": 3.861407249466951,
+      "grad_norm": 0.08230097290036,
+      "learning_rate": 0.0001495405604722826,
+      "loss": 2.7922,
+      "step": 12225
+    },
+    {
+      "epoch": 3.86298665403143,
+      "grad_norm": 0.08559451418421558,
+      "learning_rate": 0.00014914744540694698,
+      "loss": 2.7691,
+      "step": 12230
+    },
+    {
+      "epoch": 3.8645660585959094,
+      "grad_norm": 0.08263002382153399,
+      "learning_rate": 0.00014875475713962594,
+      "loss": 2.7754,
+      "step": 12235
+    },
+    {
+      "epoch": 3.8661454631603887,
+      "grad_norm": 0.08536841281141583,
+      "learning_rate": 0.00014836249614800856,
+      "loss": 2.7422,
+      "step": 12240
+    },
+    {
+      "epoch": 3.867724867724868,
+      "grad_norm": 0.07981680067080599,
+      "learning_rate": 0.00014797066290926465,
+      "loss": 2.7872,
+      "step": 12245
+    },
+    {
+      "epoch": 3.869304272289347,
+      "grad_norm": 0.0851273897360789,
+      "learning_rate": 0.0001475792579000436,
+      "loss": 2.7314,
+      "step": 12250
+    },
+    {
+      "epoch": 3.870883676853826,
+      "grad_norm": 0.08062361561989131,
+      "learning_rate": 0.00014718828159647384,
+      "loss": 2.8101,
+      "step": 12255
+    },
+    {
+      "epoch": 3.872463081418305,
+      "grad_norm": 0.0981840696392559,
+      "learning_rate": 0.0001467977344741624,
+      "loss": 2.7741,
+      "step": 12260
+    },
+    {
+      "epoch": 3.8740424859827844,
+      "grad_norm": 0.06646153017665142,
+      "learning_rate": 0.000146407617008194,
+      "loss": 2.717,
+      "step": 12265
+    },
+    {
+      "epoch": 3.8756218905472637,
+      "grad_norm": 0.08534845102315676,
+      "learning_rate": 0.00014601792967313093,
+      "loss": 2.7099,
+      "step": 12270
+    },
+    {
+      "epoch": 3.877201295111743,
+      "grad_norm": 0.07236067988844529,
+      "learning_rate": 0.00014562867294301207,
+      "loss": 2.7099,
+      "step": 12275
+    },
+    {
+      "epoch": 3.878780699676222,
+      "grad_norm": 0.07012044315380965,
+      "learning_rate": 0.0001452398472913527,
+      "loss": 3.0013,
+      "step": 12280
+    },
+    {
+      "epoch": 3.880360104240701,
+      "grad_norm": 0.07284188811108293,
+      "learning_rate": 0.00014485145319114345,
+      "loss": 2.7175,
+      "step": 12285
+    },
+    {
+      "epoch": 3.8819395088051802,
+      "grad_norm": 0.07010366764819255,
+      "learning_rate": 0.00014446349111485018,
+      "loss": 2.7621,
+      "step": 12290
+    },
+    {
+      "epoch": 3.8835189133696595,
+      "grad_norm": 0.07467745080448321,
+      "learning_rate": 0.00014407596153441328,
+      "loss": 2.7282,
+      "step": 12295
+    },
+    {
+      "epoch": 3.8850983179341387,
+      "grad_norm": 0.0770014215554429,
+      "learning_rate": 0.0001436888649212466,
+      "loss": 2.8074,
+      "step": 12300
+    },
+    {
+      "epoch": 3.886677722498618,
+      "grad_norm": 0.08929085293217041,
+      "learning_rate": 0.0001433022017462376,
+      "loss": 2.7233,
+      "step": 12305
+    },
+    {
+      "epoch": 3.8882571270630972,
+      "grad_norm": 0.07900647808172974,
+      "learning_rate": 0.00014291597247974668,
+      "loss": 2.742,
+      "step": 12310
+    },
+    {
+      "epoch": 3.8898365316275765,
+      "grad_norm": 0.08403907746645993,
+      "learning_rate": 0.00014253017759160636,
+      "loss": 2.7749,
+      "step": 12315
+    },
+    {
+      "epoch": 3.8914159361920557,
+      "grad_norm": 0.08830804078019384,
+      "learning_rate": 0.0001421448175511202,
+      "loss": 2.8338,
+      "step": 12320
+    },
+    {
+      "epoch": 3.892995340756535,
+      "grad_norm": 0.08036052320678963,
+      "learning_rate": 0.00014175989282706332,
+      "loss": 2.7507,
+      "step": 12325
+    },
+    {
+      "epoch": 3.894574745321014,
+      "grad_norm": 0.07391711139175576,
+      "learning_rate": 0.00014137540388768107,
+      "loss": 2.683,
+      "step": 12330
+    },
+    {
+      "epoch": 3.896154149885493,
+      "grad_norm": 0.06940022388169716,
+      "learning_rate": 0.00014099135120068911,
+      "loss": 2.8218,
+      "step": 12335
+    },
+    {
+      "epoch": 3.8977335544499723,
+      "grad_norm": 0.0894235128784357,
+      "learning_rate": 0.00014060773523327176,
+      "loss": 2.7571,
+      "step": 12340
+    },
+    {
+      "epoch": 3.8993129590144515,
+      "grad_norm": 0.07888755492894607,
+      "learning_rate": 0.00014022455645208248,
+      "loss": 2.7721,
+      "step": 12345
+    },
+    {
+      "epoch": 3.9008923635789308,
+      "grad_norm": 0.10058397800877593,
+      "learning_rate": 0.00013984181532324293,
+      "loss": 2.7646,
+      "step": 12350
+    },
+    {
+      "epoch": 3.90247176814341,
+      "grad_norm": 0.07999818986103556,
+      "learning_rate": 0.00013945951231234216,
+      "loss": 2.6823,
+      "step": 12355
+    },
+    {
+      "epoch": 3.9040511727078893,
+      "grad_norm": 0.07010208553521481,
+      "learning_rate": 0.00013907764788443649,
+      "loss": 2.8145,
+      "step": 12360
+    },
+    {
+      "epoch": 3.905630577272368,
+      "grad_norm": 0.08192124375429993,
+      "learning_rate": 0.00013869622250404855,
+      "loss": 2.8116,
+      "step": 12365
+    },
+    {
+      "epoch": 3.9072099818368473,
+      "grad_norm": 0.08035748821022067,
+      "learning_rate": 0.0001383152366351671,
+      "loss": 2.7304,
+      "step": 12370
+    },
+    {
+      "epoch": 3.9087893864013266,
+      "grad_norm": 0.0875957355975858,
+      "learning_rate": 0.00013793469074124614,
+      "loss": 2.807,
+      "step": 12375
+    },
+    {
+      "epoch": 3.910368790965806,
+      "grad_norm": 0.07905581991232238,
+      "learning_rate": 0.0001375545852852042,
+      "loss": 2.7509,
+      "step": 12380
+    },
+    {
+      "epoch": 3.911948195530285,
+      "grad_norm": 0.0862463921453058,
+      "learning_rate": 0.00013717492072942423,
+      "loss": 2.7711,
+      "step": 12385
+    },
+    {
+      "epoch": 3.9135276000947643,
+      "grad_norm": 0.08596933906927828,
+      "learning_rate": 0.00013679569753575322,
+      "loss": 2.7928,
+      "step": 12390
+    },
+    {
+      "epoch": 3.9151070046592436,
+      "grad_norm": 0.07644688214878918,
+      "learning_rate": 0.00013641691616550096,
+      "loss": 2.7458,
+      "step": 12395
+    },
+    {
+      "epoch": 3.916686409223723,
+      "grad_norm": 0.07795462137024338,
+      "learning_rate": 0.00013603857707943933,
+      "loss": 2.6595,
+      "step": 12400
+    },
+    {
+      "epoch": 3.918265813788202,
+      "grad_norm": 0.07544869027380872,
+      "learning_rate": 0.00013566068073780286,
+      "loss": 2.7311,
+      "step": 12405
+    },
+    {
+      "epoch": 3.9198452183526813,
+      "grad_norm": 0.07465173186397134,
+      "learning_rate": 0.00013528322760028705,
+      "loss": 2.8227,
+      "step": 12410
+    },
+    {
+      "epoch": 3.92142462291716,
+      "grad_norm": 0.07568641966275272,
+      "learning_rate": 0.00013490621812604892,
+      "loss": 2.7298,
+      "step": 12415
+    },
+    {
+      "epoch": 3.9230040274816393,
+      "grad_norm": 0.07156080882726594,
+      "learning_rate": 0.00013452965277370487,
+      "loss": 2.7488,
+      "step": 12420
+    },
+    {
+      "epoch": 3.9245834320461186,
+      "grad_norm": 0.08462040333981412,
+      "learning_rate": 0.00013415353200133163,
+      "loss": 2.748,
+      "step": 12425
+    },
+    {
+      "epoch": 3.926162836610598,
+      "grad_norm": 0.07149390183535212,
+      "learning_rate": 0.00013377785626646506,
+      "loss": 2.6944,
+      "step": 12430
+    },
+    {
+      "epoch": 3.927742241175077,
+      "grad_norm": 0.09290857724360561,
+      "learning_rate": 0.00013340262602609942,
+      "loss": 2.7834,
+      "step": 12435
+    },
+    {
+      "epoch": 3.9293216457395563,
+      "grad_norm": 0.0709730136169655,
+      "learning_rate": 0.0001330278417366873,
+      "loss": 2.7245,
+      "step": 12440
+    },
+    {
+      "epoch": 3.930901050304035,
+      "grad_norm": 0.07376630063431873,
+      "learning_rate": 0.00013265350385413871,
+      "loss": 2.8153,
+      "step": 12445
+    },
+    {
+      "epoch": 3.9324804548685144,
+      "grad_norm": 0.08242174101966185,
+      "learning_rate": 0.00013227961283382068,
+      "loss": 2.8471,
+      "step": 12450
+    },
+    {
+      "epoch": 3.9340598594329936,
+      "grad_norm": 0.0799237354023802,
+      "learning_rate": 0.00013190616913055659,
+      "loss": 2.8095,
+      "step": 12455
+    },
+    {
+      "epoch": 3.935639263997473,
+      "grad_norm": 0.06942648505961056,
+      "learning_rate": 0.00013153317319862528,
+      "loss": 2.6888,
+      "step": 12460
+    },
+    {
+      "epoch": 3.937218668561952,
+      "grad_norm": 0.07440402727491054,
+      "learning_rate": 0.00013116062549176183,
+      "loss": 2.6996,
+      "step": 12465
+    },
+    {
+      "epoch": 3.9387980731264314,
+      "grad_norm": 0.07871576677595668,
+      "learning_rate": 0.0001307885264631553,
+      "loss": 2.8121,
+      "step": 12470
+    },
+    {
+      "epoch": 3.9403774776909106,
+      "grad_norm": 0.07980672808189275,
+      "learning_rate": 0.00013041687656544938,
+      "loss": 2.711,
+      "step": 12475
+    },
+    {
+      "epoch": 3.94195688225539,
+      "grad_norm": 0.08117275324957014,
+      "learning_rate": 0.0001300456762507408,
+      "loss": 2.7112,
+      "step": 12480
+    },
+    {
+      "epoch": 3.943536286819869,
+      "grad_norm": 0.07012399895331334,
+      "learning_rate": 0.00012967492597058,
+      "loss": 2.7654,
+      "step": 12485
+    },
+    {
+      "epoch": 3.9451156913843484,
+      "grad_norm": 0.07933854900775285,
+      "learning_rate": 0.00012930462617596994,
+      "loss": 2.7743,
+      "step": 12490
+    },
+    {
+      "epoch": 3.946695095948827,
+      "grad_norm": 0.08119374219657345,
+      "learning_rate": 0.00012893477731736546,
+      "loss": 2.7857,
+      "step": 12495
+    },
+    {
+      "epoch": 3.9482745005133064,
+      "grad_norm": 0.10841667058137638,
+      "learning_rate": 0.00012856537984467248,
+      "loss": 2.7342,
+      "step": 12500
+    },
+    {
+      "epoch": 3.9498539050777857,
+      "grad_norm": 0.08529119395669767,
+      "learning_rate": 0.00012819643420724835,
+      "loss": 2.7458,
+      "step": 12505
+    },
+    {
+      "epoch": 3.951433309642265,
+      "grad_norm": 0.08069660608208816,
+      "learning_rate": 0.0001278279408539006,
+      "loss": 2.8239,
+      "step": 12510
+    },
+    {
+      "epoch": 3.953012714206744,
+      "grad_norm": 0.06719522396418853,
+      "learning_rate": 0.00012745990023288657,
+      "loss": 2.7527,
+      "step": 12515
+    },
+    {
+      "epoch": 3.9545921187712234,
+      "grad_norm": 0.09172062976577304,
+      "learning_rate": 0.0001270923127919128,
+      "loss": 2.6765,
+      "step": 12520
+    },
+    {
+      "epoch": 3.956171523335702,
+      "grad_norm": 0.0946439871431242,
+      "learning_rate": 0.00012672517897813462,
+      "loss": 2.7211,
+      "step": 12525
+    },
+    {
+      "epoch": 3.9577509279001815,
+      "grad_norm": 0.07169281189124944,
+      "learning_rate": 0.00012635849923815562,
+      "loss": 2.7036,
+      "step": 12530
+    },
+    {
+      "epoch": 3.9593303324646607,
+      "grad_norm": 0.06873997840461551,
+      "learning_rate": 0.0001259922740180271,
+      "loss": 2.7091,
+      "step": 12535
+    },
+    {
+      "epoch": 3.96090973702914,
+      "grad_norm": 0.07795319134608658,
+      "learning_rate": 0.00012562650376324674,
+      "loss": 2.7317,
+      "step": 12540
+    },
+    {
+      "epoch": 3.962489141593619,
+      "grad_norm": 0.07771723173444081,
+      "learning_rate": 0.00012526118891875991,
+      "loss": 2.8105,
+      "step": 12545
+    },
+    {
+      "epoch": 3.9640685461580984,
+      "grad_norm": 0.08365154714386186,
+      "learning_rate": 0.00012489632992895722,
+      "loss": 2.793,
+      "step": 12550
+    },
+    {
+      "epoch": 3.9656479507225777,
+      "grad_norm": 0.08829169223148622,
+      "learning_rate": 0.0001245319272376751,
+      "loss": 2.7862,
+      "step": 12555
+    },
+    {
+      "epoch": 3.967227355287057,
+      "grad_norm": 0.07959096556734646,
+      "learning_rate": 0.00012416798128819445,
+      "loss": 2.7274,
+      "step": 12560
+    },
+    {
+      "epoch": 3.968806759851536,
+      "grad_norm": 0.07992753763610859,
+      "learning_rate": 0.0001238044925232409,
+      "loss": 2.8568,
+      "step": 12565
+    },
+    {
+      "epoch": 3.9703861644160154,
+      "grad_norm": 0.08179913452703812,
+      "learning_rate": 0.00012344146138498413,
+      "loss": 2.7581,
+      "step": 12570
+    },
+    {
+      "epoch": 3.9719655689804942,
+      "grad_norm": 0.0922006090466073,
+      "learning_rate": 0.000123078888315037,
+      "loss": 2.7934,
+      "step": 12575
+    },
+    {
+      "epoch": 3.9735449735449735,
+      "grad_norm": 0.08126675516242796,
+      "learning_rate": 0.00012271677375445472,
+      "loss": 2.7128,
+      "step": 12580
+    },
+    {
+      "epoch": 3.9751243781094527,
+      "grad_norm": 0.07367105073888085,
+      "learning_rate": 0.00012235511814373524,
+      "loss": 2.6988,
+      "step": 12585
+    },
+    {
+      "epoch": 3.976703782673932,
+      "grad_norm": 0.08732065316831548,
+      "learning_rate": 0.00012199392192281806,
+      "loss": 2.7254,
+      "step": 12590
+    },
+    {
+      "epoch": 3.9782831872384112,
+      "grad_norm": 0.08797895856176832,
+      "learning_rate": 0.00012163318553108383,
+      "loss": 2.6384,
+      "step": 12595
+    },
+    {
+      "epoch": 3.97986259180289,
+      "grad_norm": 0.07485047149535816,
+      "learning_rate": 0.00012127290940735387,
+      "loss": 2.8072,
+      "step": 12600
+    },
+    {
+      "epoch": 3.9814419963673693,
+      "grad_norm": 0.0739096520837601,
+      "learning_rate": 0.00012091309398988959,
+      "loss": 2.7176,
+      "step": 12605
+    },
+    {
+      "epoch": 3.9830214009318485,
+      "grad_norm": 0.08794552730291774,
+      "learning_rate": 0.00012055373971639194,
+      "loss": 2.8281,
+      "step": 12610
+    },
+    {
+      "epoch": 3.984600805496328,
+      "grad_norm": 0.08730105494120556,
+      "learning_rate": 0.00012019484702400102,
+      "loss": 2.7105,
+      "step": 12615
+    },
+    {
+      "epoch": 3.986180210060807,
+      "grad_norm": 0.08248874844921628,
+      "learning_rate": 0.0001198364163492952,
+      "loss": 2.8115,
+      "step": 12620
+    },
+    {
+      "epoch": 3.9877596146252863,
+      "grad_norm": 0.08023159465661149,
+      "learning_rate": 0.00011947844812829112,
+      "loss": 2.7232,
+      "step": 12625
+    },
+    {
+      "epoch": 3.9893390191897655,
+      "grad_norm": 0.07409895266284665,
+      "learning_rate": 0.00011912094279644264,
+      "loss": 2.7581,
+      "step": 12630
+    },
+    {
+      "epoch": 3.9909184237542448,
+      "grad_norm": 0.07309501686116365,
+      "learning_rate": 0.00011876390078864074,
+      "loss": 2.8141,
+      "step": 12635
+    },
+    {
+      "epoch": 3.992497828318724,
+      "grad_norm": 0.0843364088296787,
+      "learning_rate": 0.00011840732253921226,
+      "loss": 2.7141,
+      "step": 12640
+    },
+    {
+      "epoch": 3.9940772328832033,
+      "grad_norm": 0.06949766629434186,
+      "learning_rate": 0.00011805120848192069,
+      "loss": 2.8486,
+      "step": 12645
+    },
+    {
+      "epoch": 3.995656637447682,
+      "grad_norm": 0.09031300657317012,
+      "learning_rate": 0.00011769555904996454,
+      "loss": 2.7412,
+      "step": 12650
+    },
+    {
+      "epoch": 3.9972360420121613,
+      "grad_norm": 0.08610290159458256,
+      "learning_rate": 0.00011734037467597663,
+      "loss": 2.7564,
+      "step": 12655
+    },
+    {
+      "epoch": 3.9988154465766406,
+      "grad_norm": 0.0704581876158133,
+      "learning_rate": 0.00011698565579202464,
+      "loss": 2.8096,
+      "step": 12660
+    },
+    {
+      "epoch": 4.0,
+      "eval_loss": 2.7555720806121826,
+      "eval_runtime": 118.9074,
+      "eval_samples_per_second": 22.278,
+      "eval_steps_per_second": 5.576,
+      "step": 12664
+    },
+    {
+      "epoch": 4.000315880912896,
+      "grad_norm": 0.07267795935226085,
+      "learning_rate": 0.00011663140282960972,
+      "loss": 2.7881,
+      "step": 12665
+    },
+    {
+      "epoch": 4.001895285477375,
+      "grad_norm": 0.11335087572849066,
+      "learning_rate": 0.00011627761621966671,
+      "loss": 2.7376,
+      "step": 12670
+    },
+    {
+      "epoch": 4.0034746900418545,
+      "grad_norm": 0.09454711336777949,
+      "learning_rate": 0.00011592429639256236,
+      "loss": 2.7298,
+      "step": 12675
+    },
+    {
+      "epoch": 4.005054094606334,
+      "grad_norm": 0.07858817668238845,
+      "learning_rate": 0.00011557144377809626,
+      "loss": 2.7434,
+      "step": 12680
+    },
+    {
+      "epoch": 4.006633499170813,
+      "grad_norm": 0.07858793747755334,
+      "learning_rate": 0.00011521905880549927,
+      "loss": 2.7363,
+      "step": 12685
+    },
+    {
+      "epoch": 4.008212903735291,
+      "grad_norm": 0.07070561973242187,
+      "learning_rate": 0.00011486714190343368,
+      "loss": 2.7429,
+      "step": 12690
+    },
+    {
+      "epoch": 4.009792308299771,
+      "grad_norm": 0.07734112618433828,
+      "learning_rate": 0.00011451569349999208,
+      "loss": 2.785,
+      "step": 12695
+    },
+    {
+      "epoch": 4.01137171286425,
+      "grad_norm": 0.0930543497197193,
+      "learning_rate": 0.00011416471402269745,
+      "loss": 2.702,
+      "step": 12700
+    },
+    {
+      "epoch": 4.012951117428729,
+      "grad_norm": 0.0870420210926015,
+      "learning_rate": 0.0001138142038985021,
+      "loss": 2.7423,
+      "step": 12705
+    },
+    {
+      "epoch": 4.014530521993208,
+      "grad_norm": 0.07013096337671941,
+      "learning_rate": 0.00011346416355378763,
+      "loss": 2.7589,
+      "step": 12710
+    },
+    {
+      "epoch": 4.016109926557688,
+      "grad_norm": 0.06466982904356627,
+      "learning_rate": 0.00011311459341436398,
+      "loss": 2.7705,
+      "step": 12715
+    },
+    {
+      "epoch": 4.017689331122167,
+      "grad_norm": 0.06498097336324272,
+      "learning_rate": 0.00011276549390546891,
+      "loss": 2.6803,
+      "step": 12720
+    },
+    {
+      "epoch": 4.019268735686646,
+      "grad_norm": 0.06619826905104946,
+      "learning_rate": 0.00011241686545176821,
+      "loss": 2.7004,
+      "step": 12725
+    },
+    {
+      "epoch": 4.020848140251125,
+      "grad_norm": 0.0737305188549197,
+      "learning_rate": 0.00011206870847735451,
+      "loss": 2.8536,
+      "step": 12730
+    },
+    {
+      "epoch": 4.022427544815605,
+      "grad_norm": 0.07863028121907374,
+      "learning_rate": 0.0001117210234057463,
+      "loss": 2.7752,
+      "step": 12735
+    },
+    {
+      "epoch": 4.024006949380084,
+      "grad_norm": 0.07744110506698226,
+      "learning_rate": 0.00011137381065988878,
+      "loss": 2.8128,
+      "step": 12740
+    },
+    {
+      "epoch": 4.025586353944563,
+      "grad_norm": 0.08781241077290056,
+      "learning_rate": 0.00011102707066215207,
+      "loss": 2.8003,
+      "step": 12745
+    },
+    {
+      "epoch": 4.027165758509042,
+      "grad_norm": 0.07560593544354333,
+      "learning_rate": 0.00011068080383433188,
+      "loss": 2.746,
+      "step": 12750
+    },
+    {
+      "epoch": 4.028745163073522,
+      "grad_norm": 0.07883754256609818,
+      "learning_rate": 0.00011033501059764739,
+      "loss": 2.6919,
+      "step": 12755
+    },
+    {
+      "epoch": 4.030324567638001,
+      "grad_norm": 0.08393257191797195,
+      "learning_rate": 0.00010998969137274233,
+      "loss": 2.6511,
+      "step": 12760
+    },
+    {
+      "epoch": 4.03190397220248,
+      "grad_norm": 0.0829523093658688,
+      "learning_rate": 0.00010964484657968366,
+      "loss": 2.7328,
+      "step": 12765
+    },
+    {
+      "epoch": 4.0334833767669585,
+      "grad_norm": 0.06908183601758058,
+      "learning_rate": 0.00010930047663796117,
+      "loss": 2.7249,
+      "step": 12770
+    },
+    {
+      "epoch": 4.035062781331438,
+      "grad_norm": 0.09022581140433972,
+      "learning_rate": 0.00010895658196648705,
+      "loss": 2.7721,
+      "step": 12775
+    },
+    {
+      "epoch": 4.036642185895917,
+      "grad_norm": 0.07657464578816578,
+      "learning_rate": 0.00010861316298359535,
+      "loss": 2.7059,
+      "step": 12780
+    },
+    {
+      "epoch": 4.038221590460396,
+      "grad_norm": 0.08410938024536432,
+      "learning_rate": 0.0001082702201070414,
+      "loss": 2.7395,
+      "step": 12785
+    },
+    {
+      "epoch": 4.039800995024875,
+      "grad_norm": 0.08752998570639688,
+      "learning_rate": 0.00010792775375400143,
+      "loss": 2.7439,
+      "step": 12790
+    },
+    {
+      "epoch": 4.041380399589355,
+      "grad_norm": 0.0790726622604146,
+      "learning_rate": 0.00010758576434107198,
+      "loss": 2.8198,
+      "step": 12795
+    },
+    {
+      "epoch": 4.042959804153834,
+      "grad_norm": 0.08010916906242367,
+      "learning_rate": 0.00010724425228426937,
+      "loss": 2.7416,
+      "step": 12800
+    },
+    {
+      "epoch": 4.044539208718313,
+      "grad_norm": 0.07123591418461958,
+      "learning_rate": 0.00010690321799902935,
+      "loss": 2.7365,
+      "step": 12805
+    },
+    {
+      "epoch": 4.046118613282792,
+      "grad_norm": 0.07012860548872378,
+      "learning_rate": 0.00010656266190020647,
+      "loss": 2.7608,
+      "step": 12810
+    },
+    {
+      "epoch": 4.047698017847272,
+      "grad_norm": 0.09592957225014759,
+      "learning_rate": 0.00010622258440207332,
+      "loss": 2.7532,
+      "step": 12815
+    },
+    {
+      "epoch": 4.049277422411751,
+      "grad_norm": 0.07239864807799998,
+      "learning_rate": 0.0001058829859183204,
+      "loss": 2.6603,
+      "step": 12820
+    },
+    {
+      "epoch": 4.05085682697623,
+      "grad_norm": 0.09866936223922237,
+      "learning_rate": 0.00010554386686205598,
+      "loss": 2.6921,
+      "step": 12825
+    },
+    {
+      "epoch": 4.052436231540709,
+      "grad_norm": 0.0992711192539062,
+      "learning_rate": 0.00010520522764580465,
+      "loss": 2.7069,
+      "step": 12830
+    },
+    {
+      "epoch": 4.054015636105189,
+      "grad_norm": 0.11154957581863764,
+      "learning_rate": 0.00010486706868150719,
+      "loss": 2.6719,
+      "step": 12835
+    },
+    {
+      "epoch": 4.055595040669668,
+      "grad_norm": 0.06616443878302948,
+      "learning_rate": 0.00010452939038052045,
+      "loss": 2.7901,
+      "step": 12840
+    },
+    {
+      "epoch": 4.057174445234147,
+      "grad_norm": 0.0742117036830686,
+      "learning_rate": 0.0001041921931536165,
+      "loss": 2.7356,
+      "step": 12845
+    },
+    {
+      "epoch": 4.0587538497986255,
+      "grad_norm": 0.08669868316456582,
+      "learning_rate": 0.00010385547741098222,
+      "loss": 2.7192,
+      "step": 12850
+    },
+    {
+      "epoch": 4.060333254363105,
+      "grad_norm": 0.07619909737865753,
+      "learning_rate": 0.00010351924356221881,
+      "loss": 2.7892,
+      "step": 12855
+    },
+    {
+      "epoch": 4.061912658927584,
+      "grad_norm": 0.07369461427803699,
+      "learning_rate": 0.00010318349201634114,
+      "loss": 2.6935,
+      "step": 12860
+    },
+    {
+      "epoch": 4.063492063492063,
+      "grad_norm": 0.07486900644344201,
+      "learning_rate": 0.00010284822318177745,
+      "loss": 2.7046,
+      "step": 12865
+    },
+    {
+      "epoch": 4.0650714680565425,
+      "grad_norm": 0.06591349144085443,
+      "learning_rate": 0.00010251343746636898,
+      "loss": 2.8372,
+      "step": 12870
+    },
+    {
+      "epoch": 4.066650872621022,
+      "grad_norm": 0.08545955955022334,
+      "learning_rate": 0.00010217913527736866,
+      "loss": 2.8099,
+      "step": 12875
+    },
+    {
+      "epoch": 4.068230277185501,
+      "grad_norm": 0.09741800113980427,
+      "learning_rate": 0.00010184531702144201,
+      "loss": 2.8099,
+      "step": 12880
+    },
+    {
+      "epoch": 4.06980968174998,
+      "grad_norm": 0.06915029241686076,
+      "learning_rate": 0.00010151198310466542,
+      "loss": 2.7801,
+      "step": 12885
+    },
+    {
+      "epoch": 4.0713890863144595,
+      "grad_norm": 0.07865162471831375,
+      "learning_rate": 0.00010117913393252631,
+      "loss": 2.7565,
+      "step": 12890
+    },
+    {
+      "epoch": 4.072968490878939,
+      "grad_norm": 0.07877350645722128,
+      "learning_rate": 0.00010084676990992198,
+      "loss": 2.7584,
+      "step": 12895
+    },
+    {
+      "epoch": 4.074547895443418,
+      "grad_norm": 0.08714047286311652,
+      "learning_rate": 0.00010051489144115999,
+      "loss": 2.7372,
+      "step": 12900
+    },
+    {
+      "epoch": 4.076127300007897,
+      "grad_norm": 0.07739346995909928,
+      "learning_rate": 0.00010018349892995737,
+      "loss": 2.7185,
+      "step": 12905
+    },
+    {
+      "epoch": 4.0777067045723765,
+      "grad_norm": 0.08920991036051683,
+      "learning_rate": 9.985259277943976e-05,
+      "loss": 2.775,
+      "step": 12910
+    },
+    {
+      "epoch": 4.079286109136856,
+      "grad_norm": 0.07472659972509982,
+      "learning_rate": 9.952217339214087e-05,
+      "loss": 2.8163,
+      "step": 12915
+    },
+    {
+      "epoch": 4.080865513701335,
+      "grad_norm": 0.07826325288038806,
+      "learning_rate": 9.91922411700028e-05,
+      "loss": 2.7713,
+      "step": 12920
+    },
+    {
+      "epoch": 4.082444918265814,
+      "grad_norm": 0.07986269543461305,
+      "learning_rate": 9.886279651437463e-05,
+      "loss": 2.7425,
+      "step": 12925
+    },
+    {
+      "epoch": 4.084024322830293,
+      "grad_norm": 0.08008789236435811,
+      "learning_rate": 9.853383982601293e-05,
+      "loss": 2.7371,
+      "step": 12930
+    },
+    {
+      "epoch": 4.085603727394772,
+      "grad_norm": 0.07425675172749298,
+      "learning_rate": 9.820537150507997e-05,
+      "loss": 2.6831,
+      "step": 12935
+    },
+    {
+      "epoch": 4.087183131959251,
+      "grad_norm": 0.073958189495471,
+      "learning_rate": 9.787739195114425e-05,
+      "loss": 2.683,
+      "step": 12940
+    },
+    {
+      "epoch": 4.08876253652373,
+      "grad_norm": 0.08580862030204887,
+      "learning_rate": 9.754990156317978e-05,
+      "loss": 2.6973,
+      "step": 12945
+    },
+    {
+      "epoch": 4.09034194108821,
+      "grad_norm": 0.06697056889956048,
+      "learning_rate": 9.722290073956536e-05,
+      "loss": 2.766,
+      "step": 12950
+    },
+    {
+      "epoch": 4.091921345652689,
+      "grad_norm": 0.07514374406514496,
+      "learning_rate": 9.689638987808441e-05,
+      "loss": 2.6808,
+      "step": 12955
+    },
+    {
+      "epoch": 4.093500750217168,
+      "grad_norm": 0.07904483649894045,
+      "learning_rate": 9.657036937592422e-05,
+      "loss": 2.8603,
+      "step": 12960
+    },
+    {
+      "epoch": 4.095080154781647,
+      "grad_norm": 0.07407896816517008,
+      "learning_rate": 9.624483962967568e-05,
+      "loss": 2.7394,
+      "step": 12965
+    },
+    {
+      "epoch": 4.096659559346127,
+      "grad_norm": 0.0813140930775043,
+      "learning_rate": 9.59198010353326e-05,
+      "loss": 2.655,
+      "step": 12970
+    },
+    {
+      "epoch": 4.098238963910606,
+      "grad_norm": 0.07840620482991092,
+      "learning_rate": 9.559525398829111e-05,
+      "loss": 2.8495,
+      "step": 12975
+    },
+    {
+      "epoch": 4.099818368475085,
+      "grad_norm": 0.0768145193026349,
+      "learning_rate": 9.527119888334995e-05,
+      "loss": 2.6923,
+      "step": 12980
+    },
+    {
+      "epoch": 4.101397773039564,
+      "grad_norm": 0.08710073076402247,
+      "learning_rate": 9.494763611470903e-05,
+      "loss": 2.7411,
+      "step": 12985
+    },
+    {
+      "epoch": 4.102977177604044,
+      "grad_norm": 0.06974392194527812,
+      "learning_rate": 9.462456607596953e-05,
+      "loss": 2.708,
+      "step": 12990
+    },
+    {
+      "epoch": 4.104556582168523,
+      "grad_norm": 0.07353572274840098,
+      "learning_rate": 9.430198916013294e-05,
+      "loss": 2.7189,
+      "step": 12995
+    },
+    {
+      "epoch": 4.106135986733002,
+      "grad_norm": 0.0660256957832875,
+      "learning_rate": 9.397990575960102e-05,
+      "loss": 2.6645,
+      "step": 13000
+    },
+    {
+      "epoch": 4.10771539129748,
+      "grad_norm": 0.06266918290859314,
+      "learning_rate": 9.365831626617555e-05,
+      "loss": 2.7111,
+      "step": 13005
+    },
+    {
+      "epoch": 4.10929479586196,
+      "grad_norm": 0.06435040279489714,
+      "learning_rate": 9.333722107105724e-05,
+      "loss": 2.7237,
+      "step": 13010
+    },
+    {
+      "epoch": 4.110874200426439,
+      "grad_norm": 0.061291168489293696,
+      "learning_rate": 9.301662056484522e-05,
+      "loss": 2.7257,
+      "step": 13015
+    },
+    {
+      "epoch": 4.112453604990918,
+      "grad_norm": 0.06917342741798833,
+      "learning_rate": 9.269651513753724e-05,
+      "loss": 2.7505,
+      "step": 13020
+    },
+    {
+      "epoch": 4.114033009555397,
+      "grad_norm": 0.0808240567172112,
+      "learning_rate": 9.237690517852859e-05,
+      "loss": 2.8223,
+      "step": 13025
+    },
+    {
+      "epoch": 4.115612414119877,
+      "grad_norm": 0.07901186635350496,
+      "learning_rate": 9.2057791076612e-05,
+      "loss": 2.8034,
+      "step": 13030
+    },
+    {
+      "epoch": 4.117191818684356,
+      "grad_norm": 0.10154864996328554,
+      "learning_rate": 9.173917321997693e-05,
+      "loss": 2.7343,
+      "step": 13035
+    },
+    {
+      "epoch": 4.118771223248835,
+      "grad_norm": 0.07225281959690563,
+      "learning_rate": 9.142105199620915e-05,
+      "loss": 2.8152,
+      "step": 13040
+    },
+    {
+      "epoch": 4.120350627813314,
+      "grad_norm": 0.06857353203537747,
+      "learning_rate": 9.11034277922903e-05,
+      "loss": 2.7683,
+      "step": 13045
+    },
+    {
+      "epoch": 4.121930032377794,
+      "grad_norm": 0.07312913741081528,
+      "learning_rate": 9.078630099459768e-05,
+      "loss": 2.7367,
+      "step": 13050
+    },
+    {
+      "epoch": 4.123509436942273,
+      "grad_norm": 0.06819091028683402,
+      "learning_rate": 9.046967198890283e-05,
+      "loss": 2.7787,
+      "step": 13055
+    },
+    {
+      "epoch": 4.125088841506752,
+      "grad_norm": 0.06621544445289144,
+      "learning_rate": 9.015354116037255e-05,
+      "loss": 2.7888,
+      "step": 13060
+    },
+    {
+      "epoch": 4.126668246071231,
+      "grad_norm": 0.06836253581250609,
+      "learning_rate": 8.983790889356714e-05,
+      "loss": 2.8769,
+      "step": 13065
+    },
+    {
+      "epoch": 4.128247650635711,
+      "grad_norm": 0.0749274451688237,
+      "learning_rate": 8.952277557244076e-05,
+      "loss": 2.8301,
+      "step": 13070
+    },
+    {
+      "epoch": 4.12982705520019,
+      "grad_norm": 0.07840530326600034,
+      "learning_rate": 8.920814158034008e-05,
+      "loss": 2.723,
+      "step": 13075
+    },
+    {
+      "epoch": 4.131406459764669,
+      "grad_norm": 0.07538609612683901,
+      "learning_rate": 8.889400730000474e-05,
+      "loss": 2.7714,
+      "step": 13080
+    },
+    {
+      "epoch": 4.1329858643291475,
+      "grad_norm": 0.07840343227079008,
+      "learning_rate": 8.858037311356676e-05,
+      "loss": 2.8266,
+      "step": 13085
+    },
+    {
+      "epoch": 4.134565268893627,
+      "grad_norm": 0.08453554879200816,
+      "learning_rate": 8.826723940254922e-05,
+      "loss": 2.8083,
+      "step": 13090
+    },
+    {
+      "epoch": 4.136144673458106,
+      "grad_norm": 0.08588294433678069,
+      "learning_rate": 8.795460654786675e-05,
+      "loss": 2.6759,
+      "step": 13095
+    },
+    {
+      "epoch": 4.137724078022585,
+      "grad_norm": 0.06968909508591559,
+      "learning_rate": 8.764247492982469e-05,
+      "loss": 2.7308,
+      "step": 13100
+    },
+    {
+      "epoch": 4.1393034825870645,
+      "grad_norm": 0.08330260139514357,
+      "learning_rate": 8.73308449281185e-05,
+      "loss": 2.8009,
+      "step": 13105
+    },
+    {
+      "epoch": 4.140882887151544,
+      "grad_norm": 0.07035989414823073,
+      "learning_rate": 8.701971692183364e-05,
+      "loss": 2.6717,
+      "step": 13110
+    },
+    {
+      "epoch": 4.142462291716023,
+      "grad_norm": 0.07396310603579366,
+      "learning_rate": 8.670909128944471e-05,
+      "loss": 2.7767,
+      "step": 13115
+    },
+    {
+      "epoch": 4.144041696280502,
+      "grad_norm": 0.06781275294051901,
+      "learning_rate": 8.639896840881533e-05,
+      "loss": 2.7749,
+      "step": 13120
+    },
+    {
+      "epoch": 4.1456211008449815,
+      "grad_norm": 0.07113564030328245,
+      "learning_rate": 8.608934865719759e-05,
+      "loss": 2.6876,
+      "step": 13125
+    },
+    {
+      "epoch": 4.147200505409461,
+      "grad_norm": 0.07591610433087836,
+      "learning_rate": 8.578023241123134e-05,
+      "loss": 2.7316,
+      "step": 13130
+    },
+    {
+      "epoch": 4.14877990997394,
+      "grad_norm": 0.0701649084899664,
+      "learning_rate": 8.547162004694408e-05,
+      "loss": 2.7393,
+      "step": 13135
+    },
+    {
+      "epoch": 4.150359314538419,
+      "grad_norm": 0.0707216056523151,
+      "learning_rate": 8.516351193975041e-05,
+      "loss": 2.7782,
+      "step": 13140
+    },
+    {
+      "epoch": 4.1519387191028985,
+      "grad_norm": 0.06585288088366432,
+      "learning_rate": 8.485590846445134e-05,
+      "loss": 2.79,
+      "step": 13145
+    },
+    {
+      "epoch": 4.153518123667378,
+      "grad_norm": 0.07924071303504657,
+      "learning_rate": 8.454880999523434e-05,
+      "loss": 2.8983,
+      "step": 13150
+    },
+    {
+      "epoch": 4.155097528231857,
+      "grad_norm": 0.07506187079168304,
+      "learning_rate": 8.424221690567185e-05,
+      "loss": 2.7565,
+      "step": 13155
+    },
+    {
+      "epoch": 4.156676932796336,
+      "grad_norm": 0.06815158920242109,
+      "learning_rate": 8.393612956872254e-05,
+      "loss": 2.8069,
+      "step": 13160
+    },
+    {
+      "epoch": 4.158256337360815,
+      "grad_norm": 0.07248305637947533,
+      "learning_rate": 8.363054835672923e-05,
+      "loss": 2.7453,
+      "step": 13165
+    },
+    {
+      "epoch": 4.159835741925294,
+      "grad_norm": 0.08402316260156116,
+      "learning_rate": 8.33254736414189e-05,
+      "loss": 2.8404,
+      "step": 13170
+    },
+    {
+      "epoch": 4.161415146489773,
+      "grad_norm": 0.07211372668448897,
+      "learning_rate": 8.302090579390292e-05,
+      "loss": 2.783,
+      "step": 13175
+    },
+    {
+      "epoch": 4.162994551054252,
+      "grad_norm": 0.06899978064304814,
+      "learning_rate": 8.27168451846757e-05,
+      "loss": 2.74,
+      "step": 13180
+    },
+    {
+      "epoch": 4.164573955618732,
+      "grad_norm": 0.07452756938770923,
+      "learning_rate": 8.241329218361481e-05,
+      "loss": 2.8303,
+      "step": 13185
+    },
+    {
+      "epoch": 4.166153360183211,
+      "grad_norm": 0.07142349650656299,
+      "learning_rate": 8.211024715998022e-05,
+      "loss": 2.7092,
+      "step": 13190
+    },
+    {
+      "epoch": 4.16773276474769,
+      "grad_norm": 0.07575208809051998,
+      "learning_rate": 8.180771048241403e-05,
+      "loss": 2.7697,
+      "step": 13195
+    },
+    {
+      "epoch": 4.169312169312169,
+      "grad_norm": 0.06957683151967588,
+      "learning_rate": 8.150568251893991e-05,
+      "loss": 2.6347,
+      "step": 13200
+    },
+    {
+      "epoch": 4.1708915738766486,
+      "grad_norm": 0.07486281848146506,
+      "learning_rate": 8.120416363696276e-05,
+      "loss": 2.7276,
+      "step": 13205
+    },
+    {
+      "epoch": 4.172470978441128,
+      "grad_norm": 0.0916192059245111,
+      "learning_rate": 8.090315420326811e-05,
+      "loss": 2.7572,
+      "step": 13210
+    },
+    {
+      "epoch": 4.174050383005607,
+      "grad_norm": 0.06858563839143138,
+      "learning_rate": 8.060265458402189e-05,
+      "loss": 2.6835,
+      "step": 13215
+    },
+    {
+      "epoch": 4.175629787570086,
+      "grad_norm": 0.07071131140325115,
+      "learning_rate": 8.030266514476975e-05,
+      "loss": 2.8856,
+      "step": 13220
+    },
+    {
+      "epoch": 4.1772091921345655,
+      "grad_norm": 0.08802252899394522,
+      "learning_rate": 8.000318625043684e-05,
+      "loss": 2.7193,
+      "step": 13225
+    },
+    {
+      "epoch": 4.178788596699045,
+      "grad_norm": 0.07430944547478276,
+      "learning_rate": 7.970421826532708e-05,
+      "loss": 2.7029,
+      "step": 13230
+    },
+    {
+      "epoch": 4.180368001263524,
+      "grad_norm": 0.08417855867691813,
+      "learning_rate": 7.940576155312291e-05,
+      "loss": 2.701,
+      "step": 13235
+    },
+    {
+      "epoch": 4.181947405828003,
+      "grad_norm": 0.07304615966559413,
+      "learning_rate": 7.910781647688514e-05,
+      "loss": 2.857,
+      "step": 13240
+    },
+    {
+      "epoch": 4.183526810392482,
+      "grad_norm": 0.06287700243132766,
+      "learning_rate": 7.8810383399052e-05,
+      "loss": 2.718,
+      "step": 13245
+    },
+    {
+      "epoch": 4.185106214956961,
+      "grad_norm": 0.06915417223034416,
+      "learning_rate": 7.851346268143861e-05,
+      "loss": 2.8032,
+      "step": 13250
+    },
+    {
+      "epoch": 4.18668561952144,
+      "grad_norm": 0.08706746214365717,
+      "learning_rate": 7.821705468523716e-05,
+      "loss": 2.7408,
+      "step": 13255
+    },
+    {
+      "epoch": 4.188265024085919,
+      "grad_norm": 0.0695987755863437,
+      "learning_rate": 7.7921159771016e-05,
+      "loss": 2.8078,
+      "step": 13260
+    },
+    {
+      "epoch": 4.189844428650399,
+      "grad_norm": 0.07315416293232646,
+      "learning_rate": 7.762577829871964e-05,
+      "loss": 2.7824,
+      "step": 13265
+    },
+    {
+      "epoch": 4.191423833214878,
+      "grad_norm": 0.07281198087456109,
+      "learning_rate": 7.73309106276675e-05,
+      "loss": 2.7185,
+      "step": 13270
+    },
+    {
+      "epoch": 4.193003237779357,
+      "grad_norm": 0.07879492288896292,
+      "learning_rate": 7.703655711655433e-05,
+      "loss": 2.6875,
+      "step": 13275
+    },
+    {
+      "epoch": 4.194582642343836,
+      "grad_norm": 0.07156009221206502,
+      "learning_rate": 7.674271812344936e-05,
+      "loss": 2.7295,
+      "step": 13280
+    },
+    {
+      "epoch": 4.196162046908316,
+      "grad_norm": 0.07588370500677988,
+      "learning_rate": 7.644939400579582e-05,
+      "loss": 2.7847,
+      "step": 13285
+    },
+    {
+      "epoch": 4.197741451472795,
+      "grad_norm": 0.06610658084447268,
+      "learning_rate": 7.615658512041068e-05,
+      "loss": 2.7541,
+      "step": 13290
+    },
+    {
+      "epoch": 4.199320856037274,
+      "grad_norm": 0.08835917770634552,
+      "learning_rate": 7.586429182348431e-05,
+      "loss": 2.831,
+      "step": 13295
+    },
+    {
+      "epoch": 4.200900260601753,
+      "grad_norm": 0.08165579997433896,
+      "learning_rate": 7.557251447057961e-05,
+      "loss": 2.7883,
+      "step": 13300
+    },
+    {
+      "epoch": 4.202479665166233,
+      "grad_norm": 0.056710901407484034,
+      "learning_rate": 7.528125341663216e-05,
+      "loss": 2.7187,
+      "step": 13305
+    },
+    {
+      "epoch": 4.204059069730712,
+      "grad_norm": 0.06606261383500421,
+      "learning_rate": 7.499050901594895e-05,
+      "loss": 2.7384,
+      "step": 13310
+    },
+    {
+      "epoch": 4.205638474295191,
+      "grad_norm": 0.08710969003358883,
+      "learning_rate": 7.470028162220921e-05,
+      "loss": 2.7903,
+      "step": 13315
+    },
+    {
+      "epoch": 4.2072178788596695,
+      "grad_norm": 0.07616166018552015,
+      "learning_rate": 7.441057158846276e-05,
+      "loss": 2.7545,
+      "step": 13320
+    },
+    {
+      "epoch": 4.208797283424149,
+      "grad_norm": 0.06864727212517878,
+      "learning_rate": 7.41213792671303e-05,
+      "loss": 2.7866,
+      "step": 13325
+    },
+    {
+      "epoch": 4.210376687988628,
+      "grad_norm": 0.07239789201367354,
+      "learning_rate": 7.383270501000244e-05,
+      "loss": 2.6703,
+      "step": 13330
+    },
+    {
+      "epoch": 4.211956092553107,
+      "grad_norm": 0.07398804471101839,
+      "learning_rate": 7.354454916823988e-05,
+      "loss": 2.6702,
+      "step": 13335
+    },
+    {
+      "epoch": 4.2135354971175865,
+      "grad_norm": 0.0713376054157643,
+      "learning_rate": 7.325691209237251e-05,
+      "loss": 2.7825,
+      "step": 13340
+    },
+    {
+      "epoch": 4.215114901682066,
+      "grad_norm": 0.1059775076324938,
+      "learning_rate": 7.296979413229965e-05,
+      "loss": 2.8092,
+      "step": 13345
+    },
+    {
+      "epoch": 4.216694306246545,
+      "grad_norm": 0.08755920739028457,
+      "learning_rate": 7.26831956372883e-05,
+      "loss": 2.7413,
+      "step": 13350
+    },
+    {
+      "epoch": 4.218273710811024,
+      "grad_norm": 0.08570092338531314,
+      "learning_rate": 7.239711695597423e-05,
+      "loss": 2.8745,
+      "step": 13355
+    },
+    {
+      "epoch": 4.2198531153755034,
+      "grad_norm": 0.08907014603792289,
+      "learning_rate": 7.211155843636058e-05,
+      "loss": 2.7157,
+      "step": 13360
+    },
+    {
+      "epoch": 4.221432519939983,
+      "grad_norm": 0.06238522496569736,
+      "learning_rate": 7.182652042581777e-05,
+      "loss": 2.6032,
+      "step": 13365
+    },
+    {
+      "epoch": 4.223011924504462,
+      "grad_norm": 0.0654451679189967,
+      "learning_rate": 7.154200327108313e-05,
+      "loss": 2.7601,
+      "step": 13370
+    },
+    {
+      "epoch": 4.224591329068941,
+      "grad_norm": 0.08077192983965602,
+      "learning_rate": 7.125800731826027e-05,
+      "loss": 2.7445,
+      "step": 13375
+    },
+    {
+      "epoch": 4.22617073363342,
+      "grad_norm": 0.0695802242339257,
+      "learning_rate": 7.097453291281886e-05,
+      "loss": 2.9043,
+      "step": 13380
+    },
+    {
+      "epoch": 4.2277501381979,
+      "grad_norm": 0.06603611733803791,
+      "learning_rate": 7.069158039959428e-05,
+      "loss": 2.6915,
+      "step": 13385
+    },
+    {
+      "epoch": 4.229329542762379,
+      "grad_norm": 0.07245836114459824,
+      "learning_rate": 7.040915012278649e-05,
+      "loss": 2.7769,
+      "step": 13390
+    },
+    {
+      "epoch": 4.230908947326858,
+      "grad_norm": 0.07601383842300888,
+      "learning_rate": 7.01272424259608e-05,
+      "loss": 2.7277,
+      "step": 13395
+    },
+    {
+      "epoch": 4.2324883518913365,
+      "grad_norm": 0.0674693714620134,
+      "learning_rate": 6.984585765204665e-05,
+      "loss": 2.7123,
+      "step": 13400
+    },
+    {
+      "epoch": 4.234067756455816,
+      "grad_norm": 0.06622712675349786,
+      "learning_rate": 6.956499614333728e-05,
+      "loss": 2.7395,
+      "step": 13405
+    },
+    {
+      "epoch": 4.235647161020295,
+      "grad_norm": 0.07659108550055024,
+      "learning_rate": 6.928465824148921e-05,
+      "loss": 2.7132,
+      "step": 13410
+    },
+    {
+      "epoch": 4.237226565584774,
+      "grad_norm": 0.06989111688990207,
+      "learning_rate": 6.900484428752229e-05,
+      "loss": 2.7415,
+      "step": 13415
+    },
+    {
+      "epoch": 4.2388059701492535,
+      "grad_norm": 0.0812313906742365,
+      "learning_rate": 6.872555462181907e-05,
+      "loss": 2.9352,
+      "step": 13420
+    },
+    {
+      "epoch": 4.240385374713733,
+      "grad_norm": 0.08118905036233727,
+      "learning_rate": 6.84467895841242e-05,
+      "loss": 2.7589,
+      "step": 13425
+    },
+    {
+      "epoch": 4.241964779278212,
+      "grad_norm": 0.0815071465384909,
+      "learning_rate": 6.816854951354395e-05,
+      "loss": 2.7103,
+      "step": 13430
+    },
+    {
+      "epoch": 4.243544183842691,
+      "grad_norm": 0.07226807620972926,
+      "learning_rate": 6.789083474854623e-05,
+      "loss": 2.7102,
+      "step": 13435
+    },
+    {
+      "epoch": 4.2451235884071705,
+      "grad_norm": 0.06507752689397496,
+      "learning_rate": 6.761364562695993e-05,
+      "loss": 2.8181,
+      "step": 13440
+    },
+    {
+      "epoch": 4.24670299297165,
+      "grad_norm": 0.06343344383579275,
+      "learning_rate": 6.733698248597442e-05,
+      "loss": 2.6587,
+      "step": 13445
+    },
+    {
+      "epoch": 4.248282397536129,
+      "grad_norm": 0.08499124262494784,
+      "learning_rate": 6.706084566213933e-05,
+      "loss": 2.7394,
+      "step": 13450
+    },
+    {
+      "epoch": 4.249861802100608,
+      "grad_norm": 0.07541637564123528,
+      "learning_rate": 6.678523549136395e-05,
+      "loss": 2.7756,
+      "step": 13455
+    },
+    {
+      "epoch": 4.2514412066650875,
+      "grad_norm": 0.07850718407085357,
+      "learning_rate": 6.651015230891694e-05,
+      "loss": 2.7917,
+      "step": 13460
+    },
+    {
+      "epoch": 4.253020611229567,
+      "grad_norm": 0.08729045027729063,
+      "learning_rate": 6.6235596449426e-05,
+      "loss": 2.6784,
+      "step": 13465
+    },
+    {
+      "epoch": 4.254600015794046,
+      "grad_norm": 0.06528484978373501,
+      "learning_rate": 6.59615682468772e-05,
+      "loss": 2.6629,
+      "step": 13470
+    },
+    {
+      "epoch": 4.256179420358524,
+      "grad_norm": 0.06004152882965481,
+      "learning_rate": 6.568806803461486e-05,
+      "loss": 2.763,
+      "step": 13475
+    },
+    {
+      "epoch": 4.257758824923004,
+      "grad_norm": 0.06942651796503685,
+      "learning_rate": 6.541509614534102e-05,
+      "loss": 2.7736,
+      "step": 13480
+    },
+    {
+      "epoch": 4.259338229487483,
+      "grad_norm": 0.08931798912696844,
+      "learning_rate": 6.514265291111505e-05,
+      "loss": 2.7304,
+      "step": 13485
+    },
+    {
+      "epoch": 4.260917634051962,
+      "grad_norm": 0.06593449167140476,
+      "learning_rate": 6.487073866335297e-05,
+      "loss": 2.7512,
+      "step": 13490
+    },
+    {
+      "epoch": 4.262497038616441,
+      "grad_norm": 0.06953667307281435,
+      "learning_rate": 6.459935373282754e-05,
+      "loss": 2.8361,
+      "step": 13495
+    },
+    {
+      "epoch": 4.264076443180921,
+      "grad_norm": 0.06738249714486153,
+      "learning_rate": 6.432849844966781e-05,
+      "loss": 2.7549,
+      "step": 13500
+    },
+    {
+      "epoch": 4.2656558477454,
+      "grad_norm": 0.06926766218445686,
+      "learning_rate": 6.405817314335838e-05,
+      "loss": 2.7942,
+      "step": 13505
+    },
+    {
+      "epoch": 4.267235252309879,
+      "grad_norm": 0.06177416969559798,
+      "learning_rate": 6.378837814273886e-05,
+      "loss": 2.7003,
+      "step": 13510
+    },
+    {
+      "epoch": 4.268814656874358,
+      "grad_norm": 0.07142728930853254,
+      "learning_rate": 6.351911377600405e-05,
+      "loss": 2.7676,
+      "step": 13515
+    },
+    {
+      "epoch": 4.270394061438838,
+      "grad_norm": 0.07247437337603856,
+      "learning_rate": 6.325038037070335e-05,
+      "loss": 2.6987,
+      "step": 13520
+    },
+    {
+      "epoch": 4.271973466003317,
+      "grad_norm": 0.06667755074853544,
+      "learning_rate": 6.298217825374003e-05,
+      "loss": 2.6546,
+      "step": 13525
+    },
+    {
+      "epoch": 4.273552870567796,
+      "grad_norm": 0.05841983559812674,
+      "learning_rate": 6.271450775137116e-05,
+      "loss": 2.6878,
+      "step": 13530
+    },
+    {
+      "epoch": 4.275132275132275,
+      "grad_norm": 0.13813052672559056,
+      "learning_rate": 6.244736918920723e-05,
+      "loss": 2.7327,
+      "step": 13535
+    },
+    {
+      "epoch": 4.276711679696755,
+      "grad_norm": 0.09279953583701123,
+      "learning_rate": 6.218076289221153e-05,
+      "loss": 2.7295,
+      "step": 13540
+    },
+    {
+      "epoch": 4.278291084261234,
+      "grad_norm": 0.08170515928486509,
+      "learning_rate": 6.191468918469983e-05,
+      "loss": 2.7195,
+      "step": 13545
+    },
+    {
+      "epoch": 4.279870488825713,
+      "grad_norm": 0.08292936633247144,
+      "learning_rate": 6.164914839034009e-05,
+      "loss": 2.8554,
+      "step": 13550
+    },
+    {
+      "epoch": 4.281449893390192,
+      "grad_norm": 0.07496049391155023,
+      "learning_rate": 6.13841408321521e-05,
+      "loss": 2.7582,
+      "step": 13555
+    },
+    {
+      "epoch": 4.283029297954671,
+      "grad_norm": 0.07430535646232635,
+      "learning_rate": 6.111966683250681e-05,
+      "loss": 2.8655,
+      "step": 13560
+    },
+    {
+      "epoch": 4.28460870251915,
+      "grad_norm": 0.06984431750702808,
+      "learning_rate": 6.085572671312628e-05,
+      "loss": 2.7455,
+      "step": 13565
+    },
+    {
+      "epoch": 4.286188107083629,
+      "grad_norm": 0.06264677255343279,
+      "learning_rate": 6.059232079508276e-05,
+      "loss": 2.7517,
+      "step": 13570
+    },
+    {
+      "epoch": 4.287767511648108,
+      "grad_norm": 0.06003493490936945,
+      "learning_rate": 6.0329449398799306e-05,
+      "loss": 2.6768,
+      "step": 13575
+    },
+    {
+      "epoch": 4.289346916212588,
+      "grad_norm": 0.06981023649376253,
+      "learning_rate": 6.006711284404837e-05,
+      "loss": 2.7726,
+      "step": 13580
+    },
+    {
+      "epoch": 4.290926320777067,
+      "grad_norm": 0.07681366787794795,
+      "learning_rate": 5.980531144995155e-05,
+      "loss": 2.7501,
+      "step": 13585
+    },
+    {
+      "epoch": 4.292505725341546,
+      "grad_norm": 0.07795095111464338,
+      "learning_rate": 5.9544045534979885e-05,
+      "loss": 2.727,
+      "step": 13590
+    },
+    {
+      "epoch": 4.294085129906025,
+      "grad_norm": 0.07648082875555658,
+      "learning_rate": 5.9283315416952696e-05,
+      "loss": 2.722,
+      "step": 13595
+    },
+    {
+      "epoch": 4.295664534470505,
+      "grad_norm": 0.06563449012297855,
+      "learning_rate": 5.9023121413038064e-05,
+      "loss": 2.7084,
+      "step": 13600
+    },
+    {
+      "epoch": 4.297243939034984,
+      "grad_norm": 0.07913454594352652,
+      "learning_rate": 5.8763463839751065e-05,
+      "loss": 2.7199,
+      "step": 13605
+    },
+    {
+      "epoch": 4.298823343599463,
+      "grad_norm": 0.08395026529050832,
+      "learning_rate": 5.850434301295493e-05,
+      "loss": 2.7976,
+      "step": 13610
+    },
+    {
+      "epoch": 4.300402748163942,
+      "grad_norm": 0.0763829408994177,
+      "learning_rate": 5.824575924785969e-05,
+      "loss": 2.7019,
+      "step": 13615
+    },
+    {
+      "epoch": 4.301982152728422,
+      "grad_norm": 0.0773512882126801,
+      "learning_rate": 5.798771285902205e-05,
+      "loss": 2.9049,
+      "step": 13620
+    },
+    {
+      "epoch": 4.303561557292901,
+      "grad_norm": 0.06828440170825315,
+      "learning_rate": 5.7730204160345135e-05,
+      "loss": 2.7715,
+      "step": 13625
+    },
+    {
+      "epoch": 4.30514096185738,
+      "grad_norm": 0.073421634901678,
+      "learning_rate": 5.7473233465077766e-05,
+      "loss": 2.7953,
+      "step": 13630
+    },
+    {
+      "epoch": 4.3067203664218585,
+      "grad_norm": 0.07285346751400801,
+      "learning_rate": 5.7216801085814616e-05,
+      "loss": 2.8742,
+      "step": 13635
+    },
+    {
+      "epoch": 4.308299770986338,
+      "grad_norm": 0.06569249021510394,
+      "learning_rate": 5.6960907334495274e-05,
+      "loss": 2.7577,
+      "step": 13640
+    },
+    {
+      "epoch": 4.309879175550817,
+      "grad_norm": 0.07456857019857352,
+      "learning_rate": 5.6705552522404226e-05,
+      "loss": 2.7515,
+      "step": 13645
+    },
+    {
+      "epoch": 4.311458580115296,
+      "grad_norm": 0.07198010974084602,
+      "learning_rate": 5.645073696017028e-05,
+      "loss": 2.7084,
+      "step": 13650
+    },
+    {
+      "epoch": 4.3130379846797755,
+      "grad_norm": 0.06138065120692361,
+      "learning_rate": 5.619646095776632e-05,
+      "loss": 2.7768,
+      "step": 13655
+    },
+    {
+      "epoch": 4.314617389244255,
+      "grad_norm": 0.06453928405053538,
+      "learning_rate": 5.5942724824509014e-05,
+      "loss": 2.7546,
+      "step": 13660
+    },
+    {
+      "epoch": 4.316196793808734,
+      "grad_norm": 0.0747783956818153,
+      "learning_rate": 5.5689528869057924e-05,
+      "loss": 2.7236,
+      "step": 13665
+    },
+    {
+      "epoch": 4.317776198373213,
+      "grad_norm": 0.06628218044947043,
+      "learning_rate": 5.5436873399415836e-05,
+      "loss": 2.7179,
+      "step": 13670
+    },
+    {
+      "epoch": 4.3193556029376925,
+      "grad_norm": 0.07117624637547765,
+      "learning_rate": 5.518475872292789e-05,
+      "loss": 2.6556,
+      "step": 13675
+    },
+    {
+      "epoch": 4.320935007502172,
+      "grad_norm": 0.06278166768111325,
+      "learning_rate": 5.4933185146281706e-05,
+      "loss": 2.7812,
+      "step": 13680
+    },
+    {
+      "epoch": 4.322514412066651,
+      "grad_norm": 0.06052245340859566,
+      "learning_rate": 5.468215297550616e-05,
+      "loss": 2.7103,
+      "step": 13685
+    },
+    {
+      "epoch": 4.32409381663113,
+      "grad_norm": 0.07885858716670277,
+      "learning_rate": 5.443166251597187e-05,
+      "loss": 2.6971,
+      "step": 13690
+    },
+    {
+      "epoch": 4.3256732211956095,
+      "grad_norm": 0.06694542098736136,
+      "learning_rate": 5.418171407239042e-05,
+      "loss": 2.8413,
+      "step": 13695
+    },
+    {
+      "epoch": 4.327252625760089,
+      "grad_norm": 0.07576900006105601,
+      "learning_rate": 5.393230794881398e-05,
+      "loss": 2.7751,
+      "step": 13700
+    },
+    {
+      "epoch": 4.328832030324568,
+      "grad_norm": 0.07256073278844476,
+      "learning_rate": 5.36834444486351e-05,
+      "loss": 2.7187,
+      "step": 13705
+    },
+    {
+      "epoch": 4.330411434889047,
+      "grad_norm": 0.08385522204414514,
+      "learning_rate": 5.343512387458621e-05,
+      "loss": 2.7229,
+      "step": 13710
+    },
+    {
+      "epoch": 4.3319908394535265,
+      "grad_norm": 0.07873581195704993,
+      "learning_rate": 5.31873465287393e-05,
+      "loss": 2.7325,
+      "step": 13715
+    },
+    {
+      "epoch": 4.333570244018005,
+      "grad_norm": 0.0750069810527468,
+      "learning_rate": 5.2940112712505485e-05,
+      "loss": 2.7698,
+      "step": 13720
+    },
+    {
+      "epoch": 4.335149648582484,
+      "grad_norm": 0.06965511574224907,
+      "learning_rate": 5.269342272663486e-05,
+      "loss": 2.7588,
+      "step": 13725
+    },
+    {
+      "epoch": 4.336729053146963,
+      "grad_norm": 0.06150813914298518,
+      "learning_rate": 5.244727687121581e-05,
+      "loss": 2.7442,
+      "step": 13730
+    },
+    {
+      "epoch": 4.338308457711443,
+      "grad_norm": 0.08021150446847683,
+      "learning_rate": 5.220167544567483e-05,
+      "loss": 2.7578,
+      "step": 13735
+    },
+    {
+      "epoch": 4.339887862275922,
+      "grad_norm": 0.06595907033478059,
+      "learning_rate": 5.195661874877633e-05,
+      "loss": 2.8048,
+      "step": 13740
+    },
+    {
+      "epoch": 4.341467266840401,
+      "grad_norm": 0.06715984794095874,
+      "learning_rate": 5.1712107078621674e-05,
+      "loss": 2.7568,
+      "step": 13745
+    },
+    {
+      "epoch": 4.34304667140488,
+      "grad_norm": 0.07371042512940375,
+      "learning_rate": 5.1468140732649495e-05,
+      "loss": 2.7925,
+      "step": 13750
+    },
+    {
+      "epoch": 4.34462607596936,
+      "grad_norm": 0.0700871588028233,
+      "learning_rate": 5.122472000763523e-05,
+      "loss": 2.7745,
+      "step": 13755
+    },
+    {
+      "epoch": 4.346205480533839,
+      "grad_norm": 0.06950713265252911,
+      "learning_rate": 5.09818451996904e-05,
+      "loss": 2.7787,
+      "step": 13760
+    },
+    {
+      "epoch": 4.347784885098318,
+      "grad_norm": 0.06535950114577534,
+      "learning_rate": 5.0739516604262234e-05,
+      "loss": 2.8471,
+      "step": 13765
+    },
+    {
+      "epoch": 4.349364289662797,
+      "grad_norm": 0.06463402111041774,
+      "learning_rate": 5.0497734516133816e-05,
+      "loss": 2.7385,
+      "step": 13770
+    },
+    {
+      "epoch": 4.350943694227277,
+      "grad_norm": 0.07359128160034697,
+      "learning_rate": 5.025649922942322e-05,
+      "loss": 2.7414,
+      "step": 13775
+    },
+    {
+      "epoch": 4.352523098791756,
+      "grad_norm": 0.07421338773029013,
+      "learning_rate": 5.001581103758374e-05,
+      "loss": 2.7344,
+      "step": 13780
+    },
+    {
+      "epoch": 4.354102503356235,
+      "grad_norm": 0.07800647904117446,
+      "learning_rate": 4.977567023340263e-05,
+      "loss": 2.7717,
+      "step": 13785
+    },
+    {
+      "epoch": 4.355681907920714,
+      "grad_norm": 0.0763861872196818,
+      "learning_rate": 4.95360771090016e-05,
+      "loss": 2.7473,
+      "step": 13790
+    },
+    {
+      "epoch": 4.357261312485193,
+      "grad_norm": 0.08171719347294958,
+      "learning_rate": 4.9297031955836014e-05,
+      "loss": 2.6722,
+      "step": 13795
+    },
+    {
+      "epoch": 4.358840717049672,
+      "grad_norm": 0.10817254137085992,
+      "learning_rate": 4.9058535064694764e-05,
+      "loss": 2.7582,
+      "step": 13800
+    },
+    {
+      "epoch": 4.360420121614151,
+      "grad_norm": 0.06503752608596988,
+      "learning_rate": 4.882058672569961e-05,
+      "loss": 2.7507,
+      "step": 13805
+    },
+    {
+      "epoch": 4.36199952617863,
+      "grad_norm": 0.06677998740072949,
+      "learning_rate": 4.858318722830518e-05,
+      "loss": 2.7941,
+      "step": 13810
+    },
+    {
+      "epoch": 4.36357893074311,
+      "grad_norm": 0.06369201479861068,
+      "learning_rate": 4.834633686129841e-05,
+      "loss": 2.7379,
+      "step": 13815
+    },
+    {
+      "epoch": 4.365158335307589,
+      "grad_norm": 0.06259694786399726,
+      "learning_rate": 4.8110035912798334e-05,
+      "loss": 2.7584,
+      "step": 13820
+    },
+    {
+      "epoch": 4.366737739872068,
+      "grad_norm": 0.07670825601864158,
+      "learning_rate": 4.7874284670255395e-05,
+      "loss": 2.8117,
+      "step": 13825
+    },
+    {
+      "epoch": 4.368317144436547,
+      "grad_norm": 0.09683683755337852,
+      "learning_rate": 4.7639083420451423e-05,
+      "loss": 2.6355,
+      "step": 13830
+    },
+    {
+      "epoch": 4.369896549001027,
+      "grad_norm": 0.06627193836192438,
+      "learning_rate": 4.740443244949949e-05,
+      "loss": 2.636,
+      "step": 13835
+    },
+    {
+      "epoch": 4.371475953565506,
+      "grad_norm": 0.07032739646481113,
+      "learning_rate": 4.71703320428431e-05,
+      "loss": 2.8651,
+      "step": 13840
+    },
+    {
+      "epoch": 4.373055358129985,
+      "grad_norm": 0.06890569235735645,
+      "learning_rate": 4.6936782485255734e-05,
+      "loss": 2.725,
+      "step": 13845
+    },
+    {
+      "epoch": 4.374634762694464,
+      "grad_norm": 0.08070104843324633,
+      "learning_rate": 4.670378406084119e-05,
+      "loss": 2.6644,
+      "step": 13850
+    },
+    {
+      "epoch": 4.376214167258944,
+      "grad_norm": 0.0688429788742678,
+      "learning_rate": 4.6471337053032466e-05,
+      "loss": 2.7465,
+      "step": 13855
+    },
+    {
+      "epoch": 4.377793571823423,
+      "grad_norm": 0.056364800051506535,
+      "learning_rate": 4.623944174459238e-05,
+      "loss": 2.8138,
+      "step": 13860
+    },
+    {
+      "epoch": 4.379372976387902,
+      "grad_norm": 0.0706255673968221,
+      "learning_rate": 4.600809841761194e-05,
+      "loss": 2.7108,
+      "step": 13865
+    },
+    {
+      "epoch": 4.380952380952381,
+      "grad_norm": 0.07042155634195384,
+      "learning_rate": 4.57773073535111e-05,
+      "loss": 2.7321,
+      "step": 13870
+    },
+    {
+      "epoch": 4.38253178551686,
+      "grad_norm": 0.06509540616282554,
+      "learning_rate": 4.554706883303783e-05,
+      "loss": 2.7258,
+      "step": 13875
+    },
+    {
+      "epoch": 4.384111190081339,
+      "grad_norm": 0.06879444588675486,
+      "learning_rate": 4.53173831362681e-05,
+      "loss": 2.7727,
+      "step": 13880
+    },
+    {
+      "epoch": 4.385690594645818,
+      "grad_norm": 0.07447573997972565,
+      "learning_rate": 4.508825054260529e-05,
+      "loss": 2.7852,
+      "step": 13885
+    },
+    {
+      "epoch": 4.3872699992102975,
+      "grad_norm": 0.08917821801221523,
+      "learning_rate": 4.485967133078001e-05,
+      "loss": 2.6625,
+      "step": 13890
+    },
+    {
+      "epoch": 4.388849403774777,
+      "grad_norm": 0.06734782982930355,
+      "learning_rate": 4.463164577884959e-05,
+      "loss": 2.7126,
+      "step": 13895
+    },
+    {
+      "epoch": 4.390428808339256,
+      "grad_norm": 0.060382294964502946,
+      "learning_rate": 4.4404174164198116e-05,
+      "loss": 2.8383,
+      "step": 13900
+    },
+    {
+      "epoch": 4.392008212903735,
+      "grad_norm": 0.07387672965341177,
+      "learning_rate": 4.41772567635354e-05,
+      "loss": 2.8482,
+      "step": 13905
+    },
+    {
+      "epoch": 4.3935876174682145,
+      "grad_norm": 0.060981720869160934,
+      "learning_rate": 4.3950893852897465e-05,
+      "loss": 2.7127,
+      "step": 13910
+    },
+    {
+      "epoch": 4.395167022032694,
+      "grad_norm": 0.07892445094347626,
+      "learning_rate": 4.372508570764572e-05,
+      "loss": 2.7342,
+      "step": 13915
+    },
+    {
+      "epoch": 4.396746426597173,
+      "grad_norm": 0.06792673838931045,
+      "learning_rate": 4.349983260246676e-05,
+      "loss": 2.6993,
+      "step": 13920
+    },
+    {
+      "epoch": 4.398325831161652,
+      "grad_norm": 0.07555246416109847,
+      "learning_rate": 4.327513481137168e-05,
+      "loss": 2.731,
+      "step": 13925
+    },
+    {
+      "epoch": 4.3999052357261315,
+      "grad_norm": 0.07844031330721625,
+      "learning_rate": 4.305099260769635e-05,
+      "loss": 2.6728,
+      "step": 13930
+    },
+    {
+      "epoch": 4.401484640290611,
+      "grad_norm": 0.06945229492805172,
+      "learning_rate": 4.2827406264100976e-05,
+      "loss": 2.754,
+      "step": 13935
+    },
+    {
+      "epoch": 4.40306404485509,
+      "grad_norm": 0.06506067535766738,
+      "learning_rate": 4.260437605256912e-05,
+      "loss": 2.6624,
+      "step": 13940
+    },
+    {
+      "epoch": 4.404643449419569,
+      "grad_norm": 0.0801259534187138,
+      "learning_rate": 4.238190224440813e-05,
+      "loss": 2.7922,
+      "step": 13945
+    },
+    {
+      "epoch": 4.406222853984048,
+      "grad_norm": 0.07372482715922321,
+      "learning_rate": 4.2159985110248435e-05,
+      "loss": 2.7701,
+      "step": 13950
+    },
+    {
+      "epoch": 4.407802258548527,
+      "grad_norm": 0.06525766264560198,
+      "learning_rate": 4.1938624920043356e-05,
+      "loss": 2.7111,
+      "step": 13955
+    },
+    {
+      "epoch": 4.409381663113006,
+      "grad_norm": 0.06493386078079086,
+      "learning_rate": 4.171782194306856e-05,
+      "loss": 2.6602,
+      "step": 13960
+    },
+    {
+      "epoch": 4.410961067677485,
+      "grad_norm": 0.06553366615343312,
+      "learning_rate": 4.149757644792207e-05,
+      "loss": 2.836,
+      "step": 13965
+    },
+    {
+      "epoch": 4.4125404722419646,
+      "grad_norm": 0.0698484117012733,
+      "learning_rate": 4.127788870252358e-05,
+      "loss": 2.7079,
+      "step": 13970
+    },
+    {
+      "epoch": 4.414119876806444,
+      "grad_norm": 0.06464569018163112,
+      "learning_rate": 4.1058758974114485e-05,
+      "loss": 2.8094,
+      "step": 13975
+    },
+    {
+      "epoch": 4.415699281370923,
+      "grad_norm": 0.06047139827966574,
+      "learning_rate": 4.0840187529257275e-05,
+      "loss": 2.7841,
+      "step": 13980
+    },
+    {
+      "epoch": 4.417278685935402,
+      "grad_norm": 0.07218113510282796,
+      "learning_rate": 4.062217463383516e-05,
+      "loss": 2.7843,
+      "step": 13985
+    },
+    {
+      "epoch": 4.4188580904998815,
+      "grad_norm": 0.06689263430779493,
+      "learning_rate": 4.0404720553052224e-05,
+      "loss": 2.7752,
+      "step": 13990
+    },
+    {
+      "epoch": 4.420437495064361,
+      "grad_norm": 0.0745169954223912,
+      "learning_rate": 4.018782555143258e-05,
+      "loss": 2.7223,
+      "step": 13995
+    },
+    {
+      "epoch": 4.42201689962884,
+      "grad_norm": 0.05396426558685231,
+      "learning_rate": 3.997148989282035e-05,
+      "loss": 2.7539,
+      "step": 14000
+    },
+    {
+      "epoch": 4.423596304193319,
+      "grad_norm": 0.06034232058814344,
+      "learning_rate": 3.9755713840378906e-05,
+      "loss": 2.6927,
+      "step": 14005
+    },
+    {
+      "epoch": 4.4251757087577985,
+      "grad_norm": 0.07655578601385024,
+      "learning_rate": 3.9540497656591234e-05,
+      "loss": 2.7493,
+      "step": 14010
+    },
+    {
+      "epoch": 4.426755113322278,
+      "grad_norm": 0.08736839497775496,
+      "learning_rate": 3.9325841603259414e-05,
+      "loss": 2.7368,
+      "step": 14015
+    },
+    {
+      "epoch": 4.428334517886757,
+      "grad_norm": 0.07381437602143522,
+      "learning_rate": 3.911174594150352e-05,
+      "loss": 2.7831,
+      "step": 14020
+    },
+    {
+      "epoch": 4.429913922451236,
+      "grad_norm": 0.05704036250688668,
+      "learning_rate": 3.889821093176255e-05,
+      "loss": 2.7748,
+      "step": 14025
+    },
+    {
+      "epoch": 4.4314933270157155,
+      "grad_norm": 0.06313743725519254,
+      "learning_rate": 3.868523683379316e-05,
+      "loss": 2.7115,
+      "step": 14030
+    },
+    {
+      "epoch": 4.433072731580194,
+      "grad_norm": 0.06809934158165958,
+      "learning_rate": 3.8472823906669784e-05,
+      "loss": 2.7187,
+      "step": 14035
+    },
+    {
+      "epoch": 4.434652136144673,
+      "grad_norm": 0.06549677567756533,
+      "learning_rate": 3.8260972408784235e-05,
+      "loss": 2.6307,
+      "step": 14040
+    },
+    {
+      "epoch": 4.436231540709152,
+      "grad_norm": 0.06997860888308473,
+      "learning_rate": 3.80496825978453e-05,
+      "loss": 2.863,
+      "step": 14045
+    },
+    {
+      "epoch": 4.437810945273632,
+      "grad_norm": 0.0651886553309917,
+      "learning_rate": 3.783895473087851e-05,
+      "loss": 2.665,
+      "step": 14050
+    },
+    {
+      "epoch": 4.439390349838111,
+      "grad_norm": 0.0665741756348438,
+      "learning_rate": 3.76287890642259e-05,
+      "loss": 2.6741,
+      "step": 14055
+    },
+    {
+      "epoch": 4.44096975440259,
+      "grad_norm": 0.06426164049312041,
+      "learning_rate": 3.741918585354548e-05,
+      "loss": 2.6937,
+      "step": 14060
+    },
+    {
+      "epoch": 4.442549158967069,
+      "grad_norm": 0.06891509753497858,
+      "learning_rate": 3.721014535381117e-05,
+      "loss": 2.7693,
+      "step": 14065
+    },
+    {
+      "epoch": 4.444128563531549,
+      "grad_norm": 0.06288296963737722,
+      "learning_rate": 3.70016678193123e-05,
+      "loss": 2.8424,
+      "step": 14070
+    },
+    {
+      "epoch": 4.445707968096028,
+      "grad_norm": 0.05996547439265329,
+      "learning_rate": 3.6793753503653385e-05,
+      "loss": 2.6971,
+      "step": 14075
+    },
+    {
+      "epoch": 4.447287372660507,
+      "grad_norm": 0.07084741729207963,
+      "learning_rate": 3.6586402659753995e-05,
+      "loss": 2.8812,
+      "step": 14080
+    },
+    {
+      "epoch": 4.448866777224986,
+      "grad_norm": 0.06084729264743166,
+      "learning_rate": 3.6379615539847756e-05,
+      "loss": 2.8096,
+      "step": 14085
+    },
+    {
+      "epoch": 4.450446181789466,
+      "grad_norm": 0.07677261064708935,
+      "learning_rate": 3.617339239548312e-05,
+      "loss": 2.7297,
+      "step": 14090
+    },
+    {
+      "epoch": 4.452025586353945,
+      "grad_norm": 0.05650388293033085,
+      "learning_rate": 3.5967733477522246e-05,
+      "loss": 2.7602,
+      "step": 14095
+    },
+    {
+      "epoch": 4.453604990918424,
+      "grad_norm": 0.062323534595875925,
+      "learning_rate": 3.576263903614085e-05,
+      "loss": 2.7709,
+      "step": 14100
+    },
+    {
+      "epoch": 4.455184395482903,
+      "grad_norm": 0.06398308241638362,
+      "learning_rate": 3.555810932082809e-05,
+      "loss": 2.8182,
+      "step": 14105
+    },
+    {
+      "epoch": 4.456763800047382,
+      "grad_norm": 0.06880813740246586,
+      "learning_rate": 3.5354144580385994e-05,
+      "loss": 2.6806,
+      "step": 14110
+    },
+    {
+      "epoch": 4.458343204611861,
+      "grad_norm": 0.0601846120350076,
+      "learning_rate": 3.515074506292981e-05,
+      "loss": 2.7093,
+      "step": 14115
+    },
+    {
+      "epoch": 4.45992260917634,
+      "grad_norm": 0.06296228083020446,
+      "learning_rate": 3.494791101588657e-05,
+      "loss": 2.7168,
+      "step": 14120
+    },
+    {
+      "epoch": 4.4615020137408195,
+      "grad_norm": 0.06191844446700565,
+      "learning_rate": 3.474564268599584e-05,
+      "loss": 2.7745,
+      "step": 14125
+    },
+    {
+      "epoch": 4.463081418305299,
+      "grad_norm": 0.06884176731460052,
+      "learning_rate": 3.454394031930885e-05,
+      "loss": 2.7455,
+      "step": 14130
+    },
+    {
+      "epoch": 4.464660822869778,
+      "grad_norm": 0.06755656670539843,
+      "learning_rate": 3.4342804161188456e-05,
+      "loss": 2.7446,
+      "step": 14135
+    },
+    {
+      "epoch": 4.466240227434257,
+      "grad_norm": 0.06319580791901808,
+      "learning_rate": 3.414223445630865e-05,
+      "loss": 2.7739,
+      "step": 14140
+    },
+    {
+      "epoch": 4.467819631998736,
+      "grad_norm": 0.08458901559524377,
+      "learning_rate": 3.3942231448654494e-05,
+      "loss": 2.7623,
+      "step": 14145
+    },
+    {
+      "epoch": 4.469399036563216,
+      "grad_norm": 0.08682005555589914,
+      "learning_rate": 3.374279538152153e-05,
+      "loss": 2.8679,
+      "step": 14150
+    },
+    {
+      "epoch": 4.470978441127695,
+      "grad_norm": 0.06654776150384711,
+      "learning_rate": 3.3543926497515806e-05,
+      "loss": 2.7028,
+      "step": 14155
+    },
+    {
+      "epoch": 4.472557845692174,
+      "grad_norm": 0.06494438763287523,
+      "learning_rate": 3.334562503855321e-05,
+      "loss": 2.8037,
+      "step": 14160
+    },
+    {
+      "epoch": 4.474137250256653,
+      "grad_norm": 0.06230766616900129,
+      "learning_rate": 3.3147891245859374e-05,
+      "loss": 2.7058,
+      "step": 14165
+    },
+    {
+      "epoch": 4.475716654821133,
+      "grad_norm": 0.0654662855004723,
+      "learning_rate": 3.2950725359969735e-05,
+      "loss": 2.6733,
+      "step": 14170
+    },
+    {
+      "epoch": 4.477296059385612,
+      "grad_norm": 0.06401045176203156,
+      "learning_rate": 3.2754127620728714e-05,
+      "loss": 2.7908,
+      "step": 14175
+    },
+    {
+      "epoch": 4.478875463950091,
+      "grad_norm": 0.07287652394147869,
+      "learning_rate": 3.2558098267289226e-05,
+      "loss": 2.7242,
+      "step": 14180
+    },
+    {
+      "epoch": 4.48045486851457,
+      "grad_norm": 0.08527784503859166,
+      "learning_rate": 3.2362637538113305e-05,
+      "loss": 2.6897,
+      "step": 14185
+    },
+    {
+      "epoch": 4.48203427307905,
+      "grad_norm": 0.06939312669332218,
+      "learning_rate": 3.216774567097097e-05,
+      "loss": 2.8094,
+      "step": 14190
+    },
+    {
+      "epoch": 4.483613677643528,
+      "grad_norm": 0.061218931687837694,
+      "learning_rate": 3.197342290294053e-05,
+      "loss": 2.7239,
+      "step": 14195
+    },
+    {
+      "epoch": 4.485193082208007,
+      "grad_norm": 0.06603079912688316,
+      "learning_rate": 3.177966947040761e-05,
+      "loss": 2.819,
+      "step": 14200
+    },
+    {
+      "epoch": 4.4867724867724865,
+      "grad_norm": 0.060929969118875785,
+      "learning_rate": 3.158648560906552e-05,
+      "loss": 2.7157,
+      "step": 14205
+    },
+    {
+      "epoch": 4.488351891336966,
+      "grad_norm": 0.06468485810349213,
+      "learning_rate": 3.139387155391465e-05,
+      "loss": 2.7658,
+      "step": 14210
+    },
+    {
+      "epoch": 4.489931295901445,
+      "grad_norm": 0.07268389043589228,
+      "learning_rate": 3.120182753926226e-05,
+      "loss": 2.8512,
+      "step": 14215
+    },
+    {
+      "epoch": 4.491510700465924,
+      "grad_norm": 0.060452501315795604,
+      "learning_rate": 3.101035379872219e-05,
+      "loss": 2.6202,
+      "step": 14220
+    },
+    {
+      "epoch": 4.4930901050304035,
+      "grad_norm": 0.06324868957888526,
+      "learning_rate": 3.081945056521451e-05,
+      "loss": 2.8334,
+      "step": 14225
+    },
+    {
+      "epoch": 4.494669509594883,
+      "grad_norm": 0.0648943320479459,
+      "learning_rate": 3.06291180709653e-05,
+      "loss": 2.7852,
+      "step": 14230
+    },
+    {
+      "epoch": 4.496248914159362,
+      "grad_norm": 0.095625503609685,
+      "learning_rate": 3.0439356547506513e-05,
+      "loss": 2.6689,
+      "step": 14235
+    },
+    {
+      "epoch": 4.497828318723841,
+      "grad_norm": 0.07565829725378756,
+      "learning_rate": 3.0250166225675114e-05,
+      "loss": 2.7775,
+      "step": 14240
+    },
+    {
+      "epoch": 4.4994077232883205,
+      "grad_norm": 0.06339715848310948,
+      "learning_rate": 3.006154733561378e-05,
+      "loss": 2.704,
+      "step": 14245
+    },
+    {
+      "epoch": 4.5009871278528,
+      "grad_norm": 0.07563055696776189,
+      "learning_rate": 2.9873500106769758e-05,
+      "loss": 2.9049,
+      "step": 14250
+    },
+    {
+      "epoch": 4.502566532417279,
+      "grad_norm": 0.05983804079064975,
+      "learning_rate": 2.968602476789495e-05,
+      "loss": 2.8519,
+      "step": 14255
+    },
+    {
+      "epoch": 4.504145936981758,
+      "grad_norm": 0.06868463639208783,
+      "learning_rate": 2.9499121547045425e-05,
+      "loss": 2.7555,
+      "step": 14260
+    },
+    {
+      "epoch": 4.505725341546237,
+      "grad_norm": 0.060978626347398604,
+      "learning_rate": 2.9312790671581434e-05,
+      "loss": 2.7537,
+      "step": 14265
+    },
+    {
+      "epoch": 4.507304746110716,
+      "grad_norm": 0.07264461101387482,
+      "learning_rate": 2.9127032368167216e-05,
+      "loss": 2.726,
+      "step": 14270
+    },
+    {
+      "epoch": 4.508884150675195,
+      "grad_norm": 0.06238319967834464,
+      "learning_rate": 2.894184686277013e-05,
+      "loss": 2.6569,
+      "step": 14275
+    },
+    {
+      "epoch": 4.510463555239674,
+      "grad_norm": 0.05858114580348566,
+      "learning_rate": 2.8757234380660858e-05,
+      "loss": 2.7033,
+      "step": 14280
+    },
+    {
+      "epoch": 4.512042959804154,
+      "grad_norm": 0.056648958840228705,
+      "learning_rate": 2.8573195146413046e-05,
+      "loss": 2.7374,
+      "step": 14285
+    },
+    {
+      "epoch": 4.513622364368633,
+      "grad_norm": 0.06109579452131664,
+      "learning_rate": 2.8389729383903107e-05,
+      "loss": 2.6694,
+      "step": 14290
+    },
+    {
+      "epoch": 4.515201768933112,
+      "grad_norm": 0.06955546274393315,
+      "learning_rate": 2.8206837316309686e-05,
+      "loss": 2.8184,
+      "step": 14295
+    },
+    {
+      "epoch": 4.516781173497591,
+      "grad_norm": 0.06389816912715901,
+      "learning_rate": 2.802451916611365e-05,
+      "loss": 2.7332,
+      "step": 14300
+    },
+    {
+      "epoch": 4.518360578062071,
+      "grad_norm": 0.07540072853690875,
+      "learning_rate": 2.7842775155097698e-05,
+      "loss": 2.7632,
+      "step": 14305
+    },
+    {
+      "epoch": 4.51993998262655,
+      "grad_norm": 0.06443257502762123,
+      "learning_rate": 2.7661605504346043e-05,
+      "loss": 2.6654,
+      "step": 14310
+    },
+    {
+      "epoch": 4.521519387191029,
+      "grad_norm": 0.06193483198132655,
+      "learning_rate": 2.7481010434244448e-05,
+      "loss": 2.752,
+      "step": 14315
+    },
+    {
+      "epoch": 4.523098791755508,
+      "grad_norm": 0.058886621958504706,
+      "learning_rate": 2.7300990164479288e-05,
+      "loss": 2.7912,
+      "step": 14320
+    },
+    {
+      "epoch": 4.524678196319988,
+      "grad_norm": 0.07651985633840072,
+      "learning_rate": 2.7121544914038178e-05,
+      "loss": 2.822,
+      "step": 14325
+    },
+    {
+      "epoch": 4.526257600884467,
+      "grad_norm": 0.05360993115806157,
+      "learning_rate": 2.6942674901208996e-05,
+      "loss": 2.7693,
+      "step": 14330
+    },
+    {
+      "epoch": 4.527837005448946,
+      "grad_norm": 0.06575026619758434,
+      "learning_rate": 2.6764380343579975e-05,
+      "loss": 2.7918,
+      "step": 14335
+    },
+    {
+      "epoch": 4.529416410013425,
+      "grad_norm": 0.07334945930308437,
+      "learning_rate": 2.6586661458039118e-05,
+      "loss": 2.8417,
+      "step": 14340
+    },
+    {
+      "epoch": 4.530995814577905,
+      "grad_norm": 0.058699707796910444,
+      "learning_rate": 2.640951846077433e-05,
+      "loss": 2.7152,
+      "step": 14345
+    },
+    {
+      "epoch": 4.532575219142384,
+      "grad_norm": 0.09061721674917018,
+      "learning_rate": 2.6232951567273012e-05,
+      "loss": 2.8641,
+      "step": 14350
+    },
+    {
+      "epoch": 4.534154623706862,
+      "grad_norm": 0.07214711896857261,
+      "learning_rate": 2.60569609923218e-05,
+      "loss": 2.6287,
+      "step": 14355
+    },
+    {
+      "epoch": 4.535734028271341,
+      "grad_norm": 0.08400853618040989,
+      "learning_rate": 2.5881546950005884e-05,
+      "loss": 2.7658,
+      "step": 14360
+    },
+    {
+      "epoch": 4.537313432835821,
+      "grad_norm": 0.07646229902988555,
+      "learning_rate": 2.5706709653709526e-05,
+      "loss": 2.7203,
+      "step": 14365
+    },
+    {
+      "epoch": 4.5388928374003,
+      "grad_norm": 0.06658085307713038,
+      "learning_rate": 2.5532449316115204e-05,
+      "loss": 2.8077,
+      "step": 14370
+    },
+    {
+      "epoch": 4.540472241964779,
+      "grad_norm": 0.06374948967721511,
+      "learning_rate": 2.5358766149203684e-05,
+      "loss": 2.8396,
+      "step": 14375
+    },
+    {
+      "epoch": 4.542051646529258,
+      "grad_norm": 0.056954466249409016,
+      "learning_rate": 2.5185660364253514e-05,
+      "loss": 2.769,
+      "step": 14380
+    },
+    {
+      "epoch": 4.543631051093738,
+      "grad_norm": 0.06443322257692594,
+      "learning_rate": 2.5013132171840925e-05,
+      "loss": 2.7951,
+      "step": 14385
+    },
+    {
+      "epoch": 4.545210455658217,
+      "grad_norm": 0.0659027665384806,
+      "learning_rate": 2.484118178183953e-05,
+      "loss": 2.7829,
+      "step": 14390
+    },
+    {
+      "epoch": 4.546789860222696,
+      "grad_norm": 0.06019612237489631,
+      "learning_rate": 2.4669809403420007e-05,
+      "loss": 2.7491,
+      "step": 14395
+    },
+    {
+      "epoch": 4.548369264787175,
+      "grad_norm": 0.0621857388979336,
+      "learning_rate": 2.4499015245049994e-05,
+      "loss": 2.7726,
+      "step": 14400
+    },
+    {
+      "epoch": 4.549948669351655,
+      "grad_norm": 0.06058665884209109,
+      "learning_rate": 2.432879951449368e-05,
+      "loss": 2.722,
+      "step": 14405
+    },
+    {
+      "epoch": 4.551528073916134,
+      "grad_norm": 0.05787388725056837,
+      "learning_rate": 2.415916241881172e-05,
+      "loss": 2.6921,
+      "step": 14410
+    },
+    {
+      "epoch": 4.553107478480613,
+      "grad_norm": 0.07026398839890141,
+      "learning_rate": 2.3990104164360872e-05,
+      "loss": 2.7933,
+      "step": 14415
+    },
+    {
+      "epoch": 4.5546868830450915,
+      "grad_norm": 0.07729955095066493,
+      "learning_rate": 2.382162495679341e-05,
+      "loss": 2.6791,
+      "step": 14420
+    },
+    {
+      "epoch": 4.556266287609571,
+      "grad_norm": 0.05699547599441058,
+      "learning_rate": 2.365372500105778e-05,
+      "loss": 2.7441,
+      "step": 14425
+    },
+    {
+      "epoch": 4.55784569217405,
+      "grad_norm": 0.05395722696644634,
+      "learning_rate": 2.3486404501397494e-05,
+      "loss": 2.6912,
+      "step": 14430
+    },
+    {
+      "epoch": 4.559425096738529,
+      "grad_norm": 0.05671082677172548,
+      "learning_rate": 2.3319663661351187e-05,
+      "loss": 2.7675,
+      "step": 14435
+    },
+    {
+      "epoch": 4.5610045013030085,
+      "grad_norm": 0.09085770144352527,
+      "learning_rate": 2.315350268375227e-05,
+      "loss": 2.6886,
+      "step": 14440
+    },
+    {
+      "epoch": 4.562583905867488,
+      "grad_norm": 0.0724620056678275,
+      "learning_rate": 2.2987921770728894e-05,
+      "loss": 2.7525,
+      "step": 14445
+    },
+    {
+      "epoch": 4.564163310431967,
+      "grad_norm": 0.060969514802193876,
+      "learning_rate": 2.282292112370382e-05,
+      "loss": 2.7198,
+      "step": 14450
+    },
+    {
+      "epoch": 4.565742714996446,
+      "grad_norm": 0.06432353698873515,
+      "learning_rate": 2.2658500943393445e-05,
+      "loss": 2.7714,
+      "step": 14455
+    },
+    {
+      "epoch": 4.5673221195609255,
+      "grad_norm": 0.062306067911699396,
+      "learning_rate": 2.2494661429808438e-05,
+      "loss": 2.8476,
+      "step": 14460
+    },
+    {
+      "epoch": 4.568901524125405,
+      "grad_norm": 0.06591504034774806,
+      "learning_rate": 2.233140278225282e-05,
+      "loss": 2.7526,
+      "step": 14465
+    },
+    {
+      "epoch": 4.570480928689884,
+      "grad_norm": 0.05234082082736486,
+      "learning_rate": 2.2168725199324337e-05,
+      "loss": 2.7742,
+      "step": 14470
+    },
+    {
+      "epoch": 4.572060333254363,
+      "grad_norm": 0.061327728160049846,
+      "learning_rate": 2.2006628878913638e-05,
+      "loss": 2.7647,
+      "step": 14475
+    },
+    {
+      "epoch": 4.5736397378188425,
+      "grad_norm": 0.06096072179000841,
+      "learning_rate": 2.184511401820438e-05,
+      "loss": 2.7635,
+      "step": 14480
+    },
+    {
+      "epoch": 4.575219142383322,
+      "grad_norm": 0.05662154341075043,
+      "learning_rate": 2.1684180813672904e-05,
+      "loss": 2.732,
+      "step": 14485
+    },
+    {
+      "epoch": 4.576798546947801,
+      "grad_norm": 0.05752992552601801,
+      "learning_rate": 2.1523829461087997e-05,
+      "loss": 2.7743,
+      "step": 14490
+    },
+    {
+      "epoch": 4.57837795151228,
+      "grad_norm": 0.06506393985290536,
+      "learning_rate": 2.1364060155510623e-05,
+      "loss": 2.7648,
+      "step": 14495
+    },
+    {
+      "epoch": 4.5799573560767595,
+      "grad_norm": 0.05979071791978911,
+      "learning_rate": 2.1204873091293598e-05,
+      "loss": 2.7522,
+      "step": 14500
+    },
+    {
+      "epoch": 4.581536760641239,
+      "grad_norm": 0.06475097488467664,
+      "learning_rate": 2.1046268462081685e-05,
+      "loss": 2.725,
+      "step": 14505
+    },
+    {
+      "epoch": 4.583116165205717,
+      "grad_norm": 0.07002462214313948,
+      "learning_rate": 2.0888246460811165e-05,
+      "loss": 2.6988,
+      "step": 14510
+    },
+    {
+      "epoch": 4.584695569770196,
+      "grad_norm": 0.06338470337206913,
+      "learning_rate": 2.0730807279709162e-05,
+      "loss": 2.8524,
+      "step": 14515
+    },
+    {
+      "epoch": 4.586274974334676,
+      "grad_norm": 0.05494800113422327,
+      "learning_rate": 2.0573951110294307e-05,
+      "loss": 2.7232,
+      "step": 14520
+    },
+    {
+      "epoch": 4.587854378899155,
+      "grad_norm": 0.05821450686335452,
+      "learning_rate": 2.041767814337564e-05,
+      "loss": 2.7765,
+      "step": 14525
+    },
+    {
+      "epoch": 4.589433783463634,
+      "grad_norm": 0.0651294804089357,
+      "learning_rate": 2.02619885690532e-05,
+      "loss": 2.8371,
+      "step": 14530
+    },
+    {
+      "epoch": 4.591013188028113,
+      "grad_norm": 0.07958280233068674,
+      "learning_rate": 2.010688257671689e-05,
+      "loss": 2.8957,
+      "step": 14535
+    },
+    {
+      "epoch": 4.592592592592593,
+      "grad_norm": 0.06409496074399801,
+      "learning_rate": 1.9952360355046938e-05,
+      "loss": 2.673,
+      "step": 14540
+    },
+    {
+      "epoch": 4.594171997157072,
+      "grad_norm": 0.06179795070704282,
+      "learning_rate": 1.9798422092013437e-05,
+      "loss": 2.7927,
+      "step": 14545
+    },
+    {
+      "epoch": 4.595751401721551,
+      "grad_norm": 0.06246801898318563,
+      "learning_rate": 1.9645067974876086e-05,
+      "loss": 2.6022,
+      "step": 14550
+    },
+    {
+      "epoch": 4.59733080628603,
+      "grad_norm": 0.057736502834117825,
+      "learning_rate": 1.9492298190184e-05,
+      "loss": 2.6849,
+      "step": 14555
+    },
+    {
+      "epoch": 4.5989102108505096,
+      "grad_norm": 0.06317115402348944,
+      "learning_rate": 1.9340112923775465e-05,
+      "loss": 2.727,
+      "step": 14560
+    },
+    {
+      "epoch": 4.600489615414989,
+      "grad_norm": 0.0760149568562783,
+      "learning_rate": 1.9188512360777733e-05,
+      "loss": 2.7165,
+      "step": 14565
+    },
+    {
+      "epoch": 4.602069019979468,
+      "grad_norm": 0.05643531205222235,
+      "learning_rate": 1.903749668560678e-05,
+      "loss": 2.7542,
+      "step": 14570
+    },
+    {
+      "epoch": 4.603648424543947,
+      "grad_norm": 0.05650550416887614,
+      "learning_rate": 1.8887066081967163e-05,
+      "loss": 2.723,
+      "step": 14575
+    },
+    {
+      "epoch": 4.605227829108426,
+      "grad_norm": 0.06232952422218689,
+      "learning_rate": 1.8737220732851556e-05,
+      "loss": 2.6474,
+      "step": 14580
+    },
+    {
+      "epoch": 4.606807233672905,
+      "grad_norm": 0.06272739678715326,
+      "learning_rate": 1.8587960820540873e-05,
+      "loss": 2.7013,
+      "step": 14585
+    },
+    {
+      "epoch": 4.608386638237384,
+      "grad_norm": 0.0596058380745117,
+      "learning_rate": 1.8439286526603815e-05,
+      "loss": 2.7787,
+      "step": 14590
+    },
+    {
+      "epoch": 4.609966042801863,
+      "grad_norm": 0.0586885996955403,
+      "learning_rate": 1.829119803189655e-05,
+      "loss": 2.6843,
+      "step": 14595
+    },
+    {
+      "epoch": 4.611545447366343,
+      "grad_norm": 0.058935178132082246,
+      "learning_rate": 1.814369551656281e-05,
+      "loss": 2.7648,
+      "step": 14600
+    },
+    {
+      "epoch": 4.613124851930822,
+      "grad_norm": 0.05902770135942328,
+      "learning_rate": 1.799677916003356e-05,
+      "loss": 2.7793,
+      "step": 14605
+    },
+    {
+      "epoch": 4.614704256495301,
+      "grad_norm": 0.05913727813921137,
+      "learning_rate": 1.7850449141026627e-05,
+      "loss": 2.7881,
+      "step": 14610
+    },
+    {
+      "epoch": 4.61628366105978,
+      "grad_norm": 0.06421761140676323,
+      "learning_rate": 1.7704705637546504e-05,
+      "loss": 2.7424,
+      "step": 14615
+    },
+    {
+      "epoch": 4.61786306562426,
+      "grad_norm": 0.061288198027929716,
+      "learning_rate": 1.755954882688432e-05,
+      "loss": 2.674,
+      "step": 14620
+    },
+    {
+      "epoch": 4.619442470188739,
+      "grad_norm": 0.058127282817275416,
+      "learning_rate": 1.741497888561755e-05,
+      "loss": 2.7704,
+      "step": 14625
+    },
+    {
+      "epoch": 4.621021874753218,
+      "grad_norm": 0.0688027425402634,
+      "learning_rate": 1.7270995989609684e-05,
+      "loss": 2.7269,
+      "step": 14630
+    },
+    {
+      "epoch": 4.622601279317697,
+      "grad_norm": 0.06690719700150118,
+      "learning_rate": 1.7127600314010118e-05,
+      "loss": 2.864,
+      "step": 14635
+    },
+    {
+      "epoch": 4.624180683882177,
+      "grad_norm": 0.055957725104073146,
+      "learning_rate": 1.698479203325387e-05,
+      "loss": 2.7936,
+      "step": 14640
+    },
+    {
+      "epoch": 4.625760088446656,
+      "grad_norm": 0.06070257519349371,
+      "learning_rate": 1.684257132106154e-05,
+      "loss": 2.8367,
+      "step": 14645
+    },
+    {
+      "epoch": 4.627339493011135,
+      "grad_norm": 0.06885399968723249,
+      "learning_rate": 1.6700938350438898e-05,
+      "loss": 2.7519,
+      "step": 14650
+    },
+    {
+      "epoch": 4.628918897575614,
+      "grad_norm": 0.06265774692629597,
+      "learning_rate": 1.6559893293676685e-05,
+      "loss": 2.7064,
+      "step": 14655
+    },
+    {
+      "epoch": 4.630498302140094,
+      "grad_norm": 0.07387155982227189,
+      "learning_rate": 1.64194363223506e-05,
+      "loss": 2.6803,
+      "step": 14660
+    },
+    {
+      "epoch": 4.632077706704573,
+      "grad_norm": 0.05610098872321232,
+      "learning_rate": 1.627956760732091e-05,
+      "loss": 2.7879,
+      "step": 14665
+    },
+    {
+      "epoch": 4.633657111269051,
+      "grad_norm": 0.0687555797648038,
+      "learning_rate": 1.6140287318732294e-05,
+      "loss": 2.7316,
+      "step": 14670
+    },
+    {
+      "epoch": 4.6352365158335305,
+      "grad_norm": 0.0652842017046115,
+      "learning_rate": 1.60015956260135e-05,
+      "loss": 2.7575,
+      "step": 14675
+    },
+    {
+      "epoch": 4.63681592039801,
+      "grad_norm": 0.061715779369038784,
+      "learning_rate": 1.58634926978774e-05,
+      "loss": 2.7488,
+      "step": 14680
+    },
+    {
+      "epoch": 4.638395324962489,
+      "grad_norm": 0.05873635883033044,
+      "learning_rate": 1.5725978702320788e-05,
+      "loss": 2.7626,
+      "step": 14685
+    },
+    {
+      "epoch": 4.639974729526968,
+      "grad_norm": 0.06498451375033829,
+      "learning_rate": 1.5589053806623842e-05,
+      "loss": 2.646,
+      "step": 14690
+    },
+    {
+      "epoch": 4.6415541340914475,
+      "grad_norm": 0.05968919227964137,
+      "learning_rate": 1.5452718177350167e-05,
+      "loss": 2.7291,
+      "step": 14695
+    },
+    {
+      "epoch": 4.643133538655927,
+      "grad_norm": 0.05939990170483963,
+      "learning_rate": 1.5316971980346595e-05,
+      "loss": 2.7875,
+      "step": 14700
+    },
+    {
+      "epoch": 4.644712943220406,
+      "grad_norm": 0.07305067144805817,
+      "learning_rate": 1.5181815380742814e-05,
+      "loss": 2.7002,
+      "step": 14705
+    },
+    {
+      "epoch": 4.646292347784885,
+      "grad_norm": 0.06599968413260995,
+      "learning_rate": 1.5047248542951585e-05,
+      "loss": 2.7398,
+      "step": 14710
+    },
+    {
+      "epoch": 4.6478717523493644,
+      "grad_norm": 0.06060760394733419,
+      "learning_rate": 1.4913271630667856e-05,
+      "loss": 2.7438,
+      "step": 14715
+    },
+    {
+      "epoch": 4.649451156913844,
+      "grad_norm": 0.06457200143485123,
+      "learning_rate": 1.4779884806869259e-05,
+      "loss": 2.7779,
+      "step": 14720
+    },
+    {
+      "epoch": 4.651030561478323,
+      "grad_norm": 0.06766896925804818,
+      "learning_rate": 1.4647088233815442e-05,
+      "loss": 2.7243,
+      "step": 14725
+    },
+    {
+      "epoch": 4.652609966042802,
+      "grad_norm": 0.053428952697001,
+      "learning_rate": 1.4514882073048185e-05,
+      "loss": 2.7115,
+      "step": 14730
+    },
+    {
+      "epoch": 4.654189370607281,
+      "grad_norm": 0.07209620805743888,
+      "learning_rate": 1.4383266485390845e-05,
+      "loss": 2.7503,
+      "step": 14735
+    },
+    {
+      "epoch": 4.65576877517176,
+      "grad_norm": 0.06486017597551075,
+      "learning_rate": 1.4252241630948514e-05,
+      "loss": 2.7112,
+      "step": 14740
+    },
+    {
+      "epoch": 4.657348179736239,
+      "grad_norm": 0.05877887994164633,
+      "learning_rate": 1.41218076691077e-05,
+      "loss": 2.7173,
+      "step": 14745
+    },
+    {
+      "epoch": 4.658927584300718,
+      "grad_norm": 0.05404621626263031,
+      "learning_rate": 1.3991964758536147e-05,
+      "loss": 2.762,
+      "step": 14750
+    },
+    {
+      "epoch": 4.6605069888651975,
+      "grad_norm": 0.05062655885257775,
+      "learning_rate": 1.3862713057182285e-05,
+      "loss": 2.7538,
+      "step": 14755
+    },
+    {
+      "epoch": 4.662086393429677,
+      "grad_norm": 0.05829852880050507,
+      "learning_rate": 1.3734052722275848e-05,
+      "loss": 2.6997,
+      "step": 14760
+    },
+    {
+      "epoch": 4.663665797994156,
+      "grad_norm": 0.060772444171327794,
+      "learning_rate": 1.3605983910326803e-05,
+      "loss": 2.6966,
+      "step": 14765
+    },
+    {
+      "epoch": 4.665245202558635,
+      "grad_norm": 0.0644898450615056,
+      "learning_rate": 1.3478506777125866e-05,
+      "loss": 2.731,
+      "step": 14770
+    },
+    {
+      "epoch": 4.6668246071231145,
+      "grad_norm": 0.05827542320067822,
+      "learning_rate": 1.3351621477743714e-05,
+      "loss": 2.7233,
+      "step": 14775
+    },
+    {
+      "epoch": 4.668404011687594,
+      "grad_norm": 0.06691029805021395,
+      "learning_rate": 1.3225328166531158e-05,
+      "loss": 2.7853,
+      "step": 14780
+    },
+    {
+      "epoch": 4.669983416252073,
+      "grad_norm": 0.06566268495273672,
+      "learning_rate": 1.3099626997119029e-05,
+      "loss": 2.7213,
+      "step": 14785
+    },
+    {
+      "epoch": 4.671562820816552,
+      "grad_norm": 0.061236687846343285,
+      "learning_rate": 1.297451812241779e-05,
+      "loss": 2.756,
+      "step": 14790
+    },
+    {
+      "epoch": 4.6731422253810315,
+      "grad_norm": 0.0536860947465952,
+      "learning_rate": 1.2850001694617253e-05,
+      "loss": 2.6802,
+      "step": 14795
+    },
+    {
+      "epoch": 4.674721629945511,
+      "grad_norm": 0.06032907143406907,
+      "learning_rate": 1.2726077865186648e-05,
+      "loss": 2.7773,
+      "step": 14800
+    },
+    {
+      "epoch": 4.67630103450999,
+      "grad_norm": 0.06119163783834023,
+      "learning_rate": 1.260274678487433e-05,
+      "loss": 2.763,
+      "step": 14805
+    },
+    {
+      "epoch": 4.677880439074469,
+      "grad_norm": 0.069980184949926,
+      "learning_rate": 1.2480008603707626e-05,
+      "loss": 2.7077,
+      "step": 14810
+    },
+    {
+      "epoch": 4.6794598436389485,
+      "grad_norm": 0.07154881039704317,
+      "learning_rate": 1.2357863470992604e-05,
+      "loss": 2.6343,
+      "step": 14815
+    },
+    {
+      "epoch": 4.681039248203428,
+      "grad_norm": 0.06226557030888638,
+      "learning_rate": 1.2236311535313849e-05,
+      "loss": 2.7263,
+      "step": 14820
+    },
+    {
+      "epoch": 4.682618652767906,
+      "grad_norm": 0.05868134529593452,
+      "learning_rate": 1.2115352944534474e-05,
+      "loss": 2.7429,
+      "step": 14825
+    },
+    {
+      "epoch": 4.684198057332385,
+      "grad_norm": 0.05644731559254908,
+      "learning_rate": 1.1994987845795724e-05,
+      "loss": 2.6599,
+      "step": 14830
+    },
+    {
+      "epoch": 4.685777461896865,
+      "grad_norm": 0.0865592401429452,
+      "learning_rate": 1.1875216385516751e-05,
+      "loss": 2.7562,
+      "step": 14835
+    },
+    {
+      "epoch": 4.687356866461344,
+      "grad_norm": 0.05669551027219937,
+      "learning_rate": 1.17560387093949e-05,
+      "loss": 2.6981,
+      "step": 14840
+    },
+    {
+      "epoch": 4.688936271025823,
+      "grad_norm": 0.061677090036888774,
+      "learning_rate": 1.1637454962404982e-05,
+      "loss": 2.7543,
+      "step": 14845
+    },
+    {
+      "epoch": 4.690515675590302,
+      "grad_norm": 0.08284595517474833,
+      "learning_rate": 1.1519465288799325e-05,
+      "loss": 2.7559,
+      "step": 14850
+    },
+    {
+      "epoch": 4.692095080154782,
+      "grad_norm": 0.05984777724798328,
+      "learning_rate": 1.140206983210762e-05,
+      "loss": 2.7345,
+      "step": 14855
+    },
+    {
+      "epoch": 4.693674484719261,
+      "grad_norm": 0.05838786521203706,
+      "learning_rate": 1.1285268735136633e-05,
+      "loss": 2.7114,
+      "step": 14860
+    },
+    {
+      "epoch": 4.69525388928374,
+      "grad_norm": 0.0543130897314997,
+      "learning_rate": 1.1169062139970376e-05,
+      "loss": 2.7835,
+      "step": 14865
+    },
+    {
+      "epoch": 4.696833293848219,
+      "grad_norm": 0.0784933235031214,
+      "learning_rate": 1.1053450187969383e-05,
+      "loss": 2.726,
+      "step": 14870
+    },
+    {
+      "epoch": 4.698412698412699,
+      "grad_norm": 0.05947895732414855,
+      "learning_rate": 1.0938433019770932e-05,
+      "loss": 2.7097,
+      "step": 14875
+    },
+    {
+      "epoch": 4.699992102977178,
+      "grad_norm": 0.06675918644773202,
+      "learning_rate": 1.0824010775288829e-05,
+      "loss": 2.7526,
+      "step": 14880
+    },
+    {
+      "epoch": 4.701571507541657,
+      "grad_norm": 0.06628644168818526,
+      "learning_rate": 1.0710183593713063e-05,
+      "loss": 2.7017,
+      "step": 14885
+    },
+    {
+      "epoch": 4.703150912106136,
+      "grad_norm": 0.06590363253954858,
+      "learning_rate": 1.059695161350993e-05,
+      "loss": 2.7512,
+      "step": 14890
+    },
+    {
+      "epoch": 4.704730316670615,
+      "grad_norm": 0.053947613534192154,
+      "learning_rate": 1.0484314972421471e-05,
+      "loss": 2.6796,
+      "step": 14895
+    },
+    {
+      "epoch": 4.706309721235094,
+      "grad_norm": 0.05509173657733855,
+      "learning_rate": 1.0372273807465638e-05,
+      "loss": 2.7717,
+      "step": 14900
+    },
+    {
+      "epoch": 4.707889125799573,
+      "grad_norm": 0.07077772894040348,
+      "learning_rate": 1.0260828254936072e-05,
+      "loss": 2.7494,
+      "step": 14905
+    },
+    {
+      "epoch": 4.709468530364052,
+      "grad_norm": 0.05472870317866765,
+      "learning_rate": 1.0149978450401776e-05,
+      "loss": 2.7635,
+      "step": 14910
+    },
+    {
+      "epoch": 4.711047934928532,
+      "grad_norm": 0.0618392835438327,
+      "learning_rate": 1.0039724528707051e-05,
+      "loss": 2.7097,
+      "step": 14915
+    },
+    {
+      "epoch": 4.712627339493011,
+      "grad_norm": 0.06141125101382459,
+      "learning_rate": 9.930066623971334e-06,
+      "loss": 2.7059,
+      "step": 14920
+    },
+    {
+      "epoch": 4.71420674405749,
+      "grad_norm": 0.06226514742487043,
+      "learning_rate": 9.82100486958909e-06,
+      "loss": 2.7471,
+      "step": 14925
+    },
+    {
+      "epoch": 4.715786148621969,
+      "grad_norm": 0.05799250583353126,
+      "learning_rate": 9.712539398229635e-06,
+      "loss": 2.843,
+      "step": 14930
+    },
+    {
+      "epoch": 4.717365553186449,
+      "grad_norm": 0.06510162586705369,
+      "learning_rate": 9.604670341836652e-06,
+      "loss": 2.7466,
+      "step": 14935
+    },
+    {
+      "epoch": 4.718944957750928,
+      "grad_norm": 0.0527705611228643,
+      "learning_rate": 9.497397831628674e-06,
+      "loss": 2.6647,
+      "step": 14940
+    },
+    {
+      "epoch": 4.720524362315407,
+      "grad_norm": 0.05916914370321485,
+      "learning_rate": 9.390721998098372e-06,
+      "loss": 2.798,
+      "step": 14945
+    },
+    {
+      "epoch": 4.722103766879886,
+      "grad_norm": 0.05273078939253314,
+      "learning_rate": 9.284642971012557e-06,
+      "loss": 2.8165,
+      "step": 14950
+    },
+    {
+      "epoch": 4.723683171444366,
+      "grad_norm": 0.07317781115947353,
+      "learning_rate": 9.179160879412063e-06,
+      "loss": 2.903,
+      "step": 14955
+    },
+    {
+      "epoch": 4.725262576008845,
+      "grad_norm": 0.06095942833581712,
+      "learning_rate": 9.074275851611691e-06,
+      "loss": 2.6464,
+      "step": 14960
+    },
+    {
+      "epoch": 4.726841980573324,
+      "grad_norm": 0.16499149627846274,
+      "learning_rate": 8.969988015199826e-06,
+      "loss": 2.7903,
+      "step": 14965
+    },
+    {
+      "epoch": 4.728421385137803,
+      "grad_norm": 0.059296679735940355,
+      "learning_rate": 8.866297497038434e-06,
+      "loss": 2.7781,
+      "step": 14970
+    },
+    {
+      "epoch": 4.730000789702283,
+      "grad_norm": 0.055793367942026134,
+      "learning_rate": 8.763204423262838e-06,
+      "loss": 2.7091,
+      "step": 14975
+    },
+    {
+      "epoch": 4.731580194266762,
+      "grad_norm": 0.060455988803085246,
+      "learning_rate": 8.660708919281612e-06,
+      "loss": 2.7383,
+      "step": 14980
+    },
+    {
+      "epoch": 4.73315959883124,
+      "grad_norm": 0.05854906963710378,
+      "learning_rate": 8.558811109776465e-06,
+      "loss": 2.7276,
+      "step": 14985
+    },
+    {
+      "epoch": 4.7347390033957195,
+      "grad_norm": 0.10055956445366193,
+      "learning_rate": 8.45751111870191e-06,
+      "loss": 2.7601,
+      "step": 14990
+    },
+    {
+      "epoch": 4.736318407960199,
+      "grad_norm": 0.0672929149842486,
+      "learning_rate": 8.35680906928532e-06,
+      "loss": 2.6883,
+      "step": 14995
+    },
+    {
+      "epoch": 4.737897812524678,
+      "grad_norm": 0.06645703766918439,
+      "learning_rate": 8.256705084026761e-06,
+      "loss": 2.7759,
+      "step": 15000
+    },
+    {
+      "epoch": 4.739477217089157,
+      "grad_norm": 0.0773292275020157,
+      "learning_rate": 8.157199284698601e-06,
+      "loss": 2.7386,
+      "step": 15005
+    },
+    {
+      "epoch": 4.7410566216536365,
+      "grad_norm": 0.06896932354488351,
+      "learning_rate": 8.05829179234574e-06,
+      "loss": 2.7304,
+      "step": 15010
+    },
+    {
+      "epoch": 4.742636026218116,
+      "grad_norm": 0.058377033841892786,
+      "learning_rate": 7.959982727285042e-06,
+      "loss": 2.7119,
+      "step": 15015
+    },
+    {
+      "epoch": 4.744215430782595,
+      "grad_norm": 0.0616384069995505,
+      "learning_rate": 7.862272209105625e-06,
+      "loss": 2.7639,
+      "step": 15020
+    },
+    {
+      "epoch": 4.745794835347074,
+      "grad_norm": 0.05627891739676093,
+      "learning_rate": 7.76516035666841e-06,
+      "loss": 2.7197,
+      "step": 15025
+    },
+    {
+      "epoch": 4.7473742399115535,
+      "grad_norm": 0.05663360905497998,
+      "learning_rate": 7.668647288106012e-06,
+      "loss": 2.7569,
+      "step": 15030
+    },
+    {
+      "epoch": 4.748953644476033,
+      "grad_norm": 0.057975134522910164,
+      "learning_rate": 7.5727331208226835e-06,
+      "loss": 2.7978,
+      "step": 15035
+    },
+    {
+      "epoch": 4.750533049040512,
+      "grad_norm": 0.05821822725552486,
+      "learning_rate": 7.4774179714941495e-06,
+      "loss": 2.7771,
+      "step": 15040
+    },
+    {
+      "epoch": 4.752112453604991,
+      "grad_norm": 0.05950304018346307,
+      "learning_rate": 7.3827019560674945e-06,
+      "loss": 2.629,
+      "step": 15045
+    },
+    {
+      "epoch": 4.7536918581694705,
+      "grad_norm": 0.06162795839189552,
+      "learning_rate": 7.288585189760943e-06,
+      "loss": 2.6555,
+      "step": 15050
+    },
+    {
+      "epoch": 4.755271262733949,
+      "grad_norm": 0.05914445886403031,
+      "learning_rate": 7.195067787063692e-06,
+      "loss": 2.7604,
+      "step": 15055
+    },
+    {
+      "epoch": 4.756850667298428,
+      "grad_norm": 0.058657165867435516,
+      "learning_rate": 7.102149861735963e-06,
+      "loss": 2.7221,
+      "step": 15060
+    },
+    {
+      "epoch": 4.758430071862907,
+      "grad_norm": 0.05915469649322643,
+      "learning_rate": 7.009831526808675e-06,
+      "loss": 2.7357,
+      "step": 15065
+    },
+    {
+      "epoch": 4.760009476427387,
+      "grad_norm": 0.06440039807456888,
+      "learning_rate": 6.918112894583328e-06,
+      "loss": 2.7004,
+      "step": 15070
+    },
+    {
+      "epoch": 4.761588880991866,
+      "grad_norm": 0.06368301718764023,
+      "learning_rate": 6.8269940766318986e-06,
+      "loss": 2.7296,
+      "step": 15075
+    },
+    {
+      "epoch": 4.763168285556345,
+      "grad_norm": 0.0678473225203925,
+      "learning_rate": 6.736475183796886e-06,
+      "loss": 2.6353,
+      "step": 15080
+    },
+    {
+      "epoch": 4.764747690120824,
+      "grad_norm": 0.052843028068742595,
+      "learning_rate": 6.646556326190822e-06,
+      "loss": 2.7018,
+      "step": 15085
+    },
+    {
+      "epoch": 4.766327094685304,
+      "grad_norm": 0.06127133807691744,
+      "learning_rate": 6.557237613196321e-06,
+      "loss": 2.7277,
+      "step": 15090
+    },
+    {
+      "epoch": 4.767906499249783,
+      "grad_norm": 0.06577587907109655,
+      "learning_rate": 6.468519153466134e-06,
+      "loss": 2.7959,
+      "step": 15095
+    },
+    {
+      "epoch": 4.769485903814262,
+      "grad_norm": 0.05874482833404608,
+      "learning_rate": 6.3804010549225465e-06,
+      "loss": 2.7262,
+      "step": 15100
+    },
+    {
+      "epoch": 4.771065308378741,
+      "grad_norm": 0.06110105448734113,
+      "learning_rate": 6.292883424757867e-06,
+      "loss": 2.6778,
+      "step": 15105
+    },
+    {
+      "epoch": 4.772644712943221,
+      "grad_norm": 0.05864471882579104,
+      "learning_rate": 6.205966369433547e-06,
+      "loss": 2.6706,
+      "step": 15110
+    },
+    {
+      "epoch": 4.7742241175077,
+      "grad_norm": 0.05639066484026297,
+      "learning_rate": 6.119649994680842e-06,
+      "loss": 2.7794,
+      "step": 15115
+    },
+    {
+      "epoch": 4.775803522072179,
+      "grad_norm": 0.07878649928148707,
+      "learning_rate": 6.033934405500041e-06,
+      "loss": 2.7969,
+      "step": 15120
+    },
+    {
+      "epoch": 4.777382926636658,
+      "grad_norm": 0.07749243563922507,
+      "learning_rate": 5.948819706160901e-06,
+      "loss": 2.8182,
+      "step": 15125
+    },
+    {
+      "epoch": 4.778962331201138,
+      "grad_norm": 0.06604600619512055,
+      "learning_rate": 5.864306000201825e-06,
+      "loss": 2.7728,
+      "step": 15130
+    },
+    {
+      "epoch": 4.780541735765617,
+      "grad_norm": 0.06483095830054528,
+      "learning_rate": 5.780393390430405e-06,
+      "loss": 2.6885,
+      "step": 15135
+    },
+    {
+      "epoch": 4.782121140330095,
+      "grad_norm": 0.054909753523317575,
+      "learning_rate": 5.697081978922935e-06,
+      "loss": 2.7436,
+      "step": 15140
+    },
+    {
+      "epoch": 4.783700544894574,
+      "grad_norm": 0.07589710384598491,
+      "learning_rate": 5.6143718670244594e-06,
+      "loss": 2.8059,
+      "step": 15145
+    },
+    {
+      "epoch": 4.785279949459054,
+      "grad_norm": 0.06750839314901379,
+      "learning_rate": 5.532263155348438e-06,
+      "loss": 2.6943,
+      "step": 15150
+    },
+    {
+      "epoch": 4.786859354023533,
+      "grad_norm": 0.0746353227848116,
+      "learning_rate": 5.450755943776864e-06,
+      "loss": 2.718,
+      "step": 15155
+    },
+    {
+      "epoch": 4.788438758588012,
+      "grad_norm": 0.0691116307433447,
+      "learning_rate": 5.369850331459925e-06,
+      "loss": 2.755,
+      "step": 15160
+    },
+    {
+      "epoch": 4.790018163152491,
+      "grad_norm": 0.05531640854987408,
+      "learning_rate": 5.289546416816116e-06,
+      "loss": 2.6978,
+      "step": 15165
+    },
+    {
+      "epoch": 4.791597567716971,
+      "grad_norm": 0.056533255556961814,
+      "learning_rate": 5.209844297531796e-06,
+      "loss": 2.7005,
+      "step": 15170
+    },
+    {
+      "epoch": 4.79317697228145,
+      "grad_norm": 0.0562531483383305,
+      "learning_rate": 5.130744070561522e-06,
+      "loss": 2.7563,
+      "step": 15175
+    },
+    {
+      "epoch": 4.794756376845929,
+      "grad_norm": 0.0625340328968042,
+      "learning_rate": 5.0522458321274335e-06,
+      "loss": 2.7568,
+      "step": 15180
+    },
+    {
+      "epoch": 4.796335781410408,
+      "grad_norm": 0.056404367673192526,
+      "learning_rate": 4.974349677719591e-06,
+      "loss": 2.7705,
+      "step": 15185
+    },
+    {
+      "epoch": 4.797915185974888,
+      "grad_norm": 0.06908653457847379,
+      "learning_rate": 4.897055702095421e-06,
+      "loss": 2.7288,
+      "step": 15190
+    },
+    {
+      "epoch": 4.799494590539367,
+      "grad_norm": 0.06513915287870246,
+      "learning_rate": 4.820363999279987e-06,
+      "loss": 2.7565,
+      "step": 15195
+    },
+    {
+      "epoch": 4.801073995103846,
+      "grad_norm": 0.057377722680944654,
+      "learning_rate": 4.7442746625656616e-06,
+      "loss": 2.7648,
+      "step": 15200
+    },
+    {
+      "epoch": 4.802653399668325,
+      "grad_norm": 0.06775039201509281,
+      "learning_rate": 4.6687877845120185e-06,
+      "loss": 2.6989,
+      "step": 15205
+    },
+    {
+      "epoch": 4.804232804232804,
+      "grad_norm": 0.06411973092923852,
+      "learning_rate": 4.59390345694588e-06,
+      "loss": 2.8046,
+      "step": 15210
+    },
+    {
+      "epoch": 4.805812208797283,
+      "grad_norm": 0.06531844572614323,
+      "learning_rate": 4.519621770960936e-06,
+      "loss": 2.8352,
+      "step": 15215
+    },
+    {
+      "epoch": 4.807391613361762,
+      "grad_norm": 0.057181390031279165,
+      "learning_rate": 4.445942816917958e-06,
+      "loss": 2.7532,
+      "step": 15220
+    },
+    {
+      "epoch": 4.8089710179262415,
+      "grad_norm": 0.05483137340649482,
+      "learning_rate": 4.37286668444431e-06,
+      "loss": 2.8256,
+      "step": 15225
+    },
+    {
+      "epoch": 4.810550422490721,
+      "grad_norm": 0.05106913732806533,
+      "learning_rate": 4.300393462434271e-06,
+      "loss": 2.7968,
+      "step": 15230
+    },
+    {
+      "epoch": 4.8121298270552,
+      "grad_norm": 0.059138307189591,
+      "learning_rate": 4.228523239048543e-06,
+      "loss": 2.7916,
+      "step": 15235
+    },
+    {
+      "epoch": 4.813709231619679,
+      "grad_norm": 0.05877993488243228,
+      "learning_rate": 4.157256101714413e-06,
+      "loss": 2.6521,
+      "step": 15240
+    },
+    {
+      "epoch": 4.8152886361841585,
+      "grad_norm": 0.055401634394682805,
+      "learning_rate": 4.0865921371254224e-06,
+      "loss": 2.7479,
+      "step": 15245
+    },
+    {
+      "epoch": 4.816868040748638,
+      "grad_norm": 0.05057956724789247,
+      "learning_rate": 4.016531431241532e-06,
+      "loss": 2.7482,
+      "step": 15250
+    },
+    {
+      "epoch": 4.818447445313117,
+      "grad_norm": 0.05726350015391884,
+      "learning_rate": 3.947074069288625e-06,
+      "loss": 2.7244,
+      "step": 15255
+    },
+    {
+      "epoch": 4.820026849877596,
+      "grad_norm": 0.06076411959971761,
+      "learning_rate": 3.878220135758948e-06,
+      "loss": 2.704,
+      "step": 15260
+    },
+    {
+      "epoch": 4.8216062544420755,
+      "grad_norm": 0.06304090033074285,
+      "learning_rate": 3.8099697144104438e-06,
+      "loss": 2.7629,
+      "step": 15265
+    },
+    {
+      "epoch": 4.823185659006555,
+      "grad_norm": 0.05814717358794076,
+      "learning_rate": 3.7423228882670358e-06,
+      "loss": 2.7172,
+      "step": 15270
+    },
+    {
+      "epoch": 4.824765063571034,
+      "grad_norm": 0.055460152393600984,
+      "learning_rate": 3.6752797396182867e-06,
+      "loss": 2.7555,
+      "step": 15275
+    },
+    {
+      "epoch": 4.826344468135513,
+      "grad_norm": 0.06435768058340195,
+      "learning_rate": 3.6088403500196267e-06,
+      "loss": 2.7931,
+      "step": 15280
+    },
+    {
+      "epoch": 4.8279238726999925,
+      "grad_norm": 0.05810767081405668,
+      "learning_rate": 3.5430048002918493e-06,
+      "loss": 2.6553,
+      "step": 15285
+    },
+    {
+      "epoch": 4.829503277264472,
+      "grad_norm": 0.10008919099175449,
+      "learning_rate": 3.4777731705211703e-06,
+      "loss": 2.7711,
+      "step": 15290
+    },
+    {
+      "epoch": 4.831082681828951,
+      "grad_norm": 0.06245798141826981,
+      "learning_rate": 3.4131455400593368e-06,
+      "loss": 2.7068,
+      "step": 15295
+    },
+    {
+      "epoch": 4.832662086393429,
+      "grad_norm": 0.055230100674052006,
+      "learning_rate": 3.3491219875232403e-06,
+      "loss": 2.6398,
+      "step": 15300
+    },
+    {
+      "epoch": 4.834241490957909,
+      "grad_norm": 0.06924589388567168,
+      "learning_rate": 3.2857025907949146e-06,
+      "loss": 2.6182,
+      "step": 15305
+    },
+    {
+      "epoch": 4.835820895522388,
+      "grad_norm": 0.05871203678959051,
+      "learning_rate": 3.222887427021537e-06,
+      "loss": 2.7068,
+      "step": 15310
+    },
+    {
+      "epoch": 4.837400300086867,
+      "grad_norm": 0.060324558691795774,
+      "learning_rate": 3.160676572615262e-06,
+      "loss": 2.6595,
+      "step": 15315
+    },
+    {
+      "epoch": 4.838979704651346,
+      "grad_norm": 0.06256345492332065,
+      "learning_rate": 3.0990701032530542e-06,
+      "loss": 2.6762,
+      "step": 15320
+    },
+    {
+      "epoch": 4.8405591092158256,
+      "grad_norm": 0.054424348737375214,
+      "learning_rate": 3.0380680938766337e-06,
+      "loss": 2.7133,
+      "step": 15325
+    },
+    {
+      "epoch": 4.842138513780305,
+      "grad_norm": 0.05087877330618913,
+      "learning_rate": 2.9776706186926407e-06,
+      "loss": 2.8461,
+      "step": 15330
+    },
+    {
+      "epoch": 4.843717918344784,
+      "grad_norm": 0.05234768882543315,
+      "learning_rate": 2.917877751172027e-06,
+      "loss": 2.7088,
+      "step": 15335
+    },
+    {
+      "epoch": 4.845297322909263,
+      "grad_norm": 0.053568382157766374,
+      "learning_rate": 2.8586895640504985e-06,
+      "loss": 2.75,
+      "step": 15340
+    },
+    {
+      "epoch": 4.8468767274737425,
+      "grad_norm": 0.055739675394195665,
+      "learning_rate": 2.800106129328128e-06,
+      "loss": 2.7616,
+      "step": 15345
+    },
+    {
+      "epoch": 4.848456132038222,
+      "grad_norm": 0.05351193031715695,
+      "learning_rate": 2.7421275182691884e-06,
+      "loss": 2.687,
+      "step": 15350
+    },
+    {
+      "epoch": 4.850035536602701,
+      "grad_norm": 0.04983056997371717,
+      "learning_rate": 2.6847538014024285e-06,
+      "loss": 2.7052,
+      "step": 15355
+    },
+    {
+      "epoch": 4.85161494116718,
+      "grad_norm": 0.06148903219444012,
+      "learning_rate": 2.6279850485206314e-06,
+      "loss": 2.794,
+      "step": 15360
+    },
+    {
+      "epoch": 4.8531943457316595,
+      "grad_norm": 0.05592778447049073,
+      "learning_rate": 2.571821328680668e-06,
+      "loss": 2.7798,
+      "step": 15365
+    },
+    {
+      "epoch": 4.854773750296138,
+      "grad_norm": 0.05962306167498793,
+      "learning_rate": 2.516262710203554e-06,
+      "loss": 2.6869,
+      "step": 15370
+    },
+    {
+      "epoch": 4.856353154860617,
+      "grad_norm": 0.053960696536328595,
+      "learning_rate": 2.4613092606739496e-06,
+      "loss": 2.6778,
+      "step": 15375
+    },
+    {
+      "epoch": 4.857932559425096,
+      "grad_norm": 0.10449612336693467,
+      "learning_rate": 2.406961046940659e-06,
+      "loss": 2.7655,
+      "step": 15380
+    },
+    {
+      "epoch": 4.859511963989576,
+      "grad_norm": 0.06596788826797334,
+      "learning_rate": 2.35321813511602e-06,
+      "loss": 2.7192,
+      "step": 15385
+    },
+    {
+      "epoch": 4.861091368554055,
+      "grad_norm": 0.05756139751792677,
+      "learning_rate": 2.3000805905761814e-06,
+      "loss": 2.7363,
+      "step": 15390
+    },
+    {
+      "epoch": 4.862670773118534,
+      "grad_norm": 0.06448921203442322,
+      "learning_rate": 2.24754847796077e-06,
+      "loss": 2.8249,
+      "step": 15395
+    },
+    {
+      "epoch": 4.864250177683013,
+      "grad_norm": 0.061459534392145035,
+      "learning_rate": 2.195621861173003e-06,
+      "loss": 2.7841,
+      "step": 15400
+    },
+    {
+      "epoch": 4.865829582247493,
+      "grad_norm": 0.04969925481491452,
+      "learning_rate": 2.1443008033795174e-06,
+      "loss": 2.7858,
+      "step": 15405
+    },
+    {
+      "epoch": 4.867408986811972,
+      "grad_norm": 0.057196089747991834,
+      "learning_rate": 2.09358536701032e-06,
+      "loss": 2.6766,
+      "step": 15410
+    },
+    {
+      "epoch": 4.868988391376451,
+      "grad_norm": 0.05714107205418257,
+      "learning_rate": 2.0434756137586717e-06,
+      "loss": 2.7223,
+      "step": 15415
+    },
+    {
+      "epoch": 4.87056779594093,
+      "grad_norm": 0.060405759017855686,
+      "learning_rate": 1.993971604581146e-06,
+      "loss": 2.7119,
+      "step": 15420
+    },
+    {
+      "epoch": 4.87214720050541,
+      "grad_norm": 0.0662054910886324,
+      "learning_rate": 1.9450733996973503e-06,
+      "loss": 2.7402,
+      "step": 15425
+    },
+    {
+      "epoch": 4.873726605069889,
+      "grad_norm": 0.06102989770270585,
+      "learning_rate": 1.8967810585898692e-06,
+      "loss": 2.6732,
+      "step": 15430
+    },
+    {
+      "epoch": 4.875306009634368,
+      "grad_norm": 0.0649814842059288,
+      "learning_rate": 1.849094640004545e-06,
+      "loss": 2.7352,
+      "step": 15435
+    },
+    {
+      "epoch": 4.876885414198847,
+      "grad_norm": 0.054831489915207365,
+      "learning_rate": 1.8020142019499752e-06,
+      "loss": 2.7626,
+      "step": 15440
+    },
+    {
+      "epoch": 4.878464818763327,
+      "grad_norm": 0.05717943145989557,
+      "learning_rate": 1.7555398016975143e-06,
+      "loss": 2.7748,
+      "step": 15445
+    },
+    {
+      "epoch": 4.880044223327806,
+      "grad_norm": 0.0683332453356377,
+      "learning_rate": 1.7096714957814952e-06,
+      "loss": 2.7078,
+      "step": 15450
+    },
+    {
+      "epoch": 4.881623627892285,
+      "grad_norm": 0.06792411946161687,
+      "learning_rate": 1.6644093399987848e-06,
+      "loss": 2.7,
+      "step": 15455
+    },
+    {
+      "epoch": 4.8832030324567635,
+      "grad_norm": 0.054876135445462194,
+      "learning_rate": 1.6197533894090622e-06,
+      "loss": 2.7288,
+      "step": 15460
+    },
+    {
+      "epoch": 4.884782437021243,
+      "grad_norm": 0.06811680379297462,
+      "learning_rate": 1.5757036983344297e-06,
+      "loss": 2.6166,
+      "step": 15465
+    },
+    {
+      "epoch": 4.886361841585722,
+      "grad_norm": 0.05210136140369539,
+      "learning_rate": 1.5322603203595797e-06,
+      "loss": 2.7213,
+      "step": 15470
+    },
+    {
+      "epoch": 4.887941246150201,
+      "grad_norm": 0.051948038479980814,
+      "learning_rate": 1.4894233083316277e-06,
+      "loss": 2.8177,
+      "step": 15475
+    },
+    {
+      "epoch": 4.8895206507146804,
+      "grad_norm": 0.052774872054946194,
+      "learning_rate": 1.4471927143601126e-06,
+      "loss": 2.762,
+      "step": 15480
+    },
+    {
+      "epoch": 4.89110005527916,
+      "grad_norm": 0.05486860863597665,
+      "learning_rate": 1.4055685898167746e-06,
+      "loss": 2.7276,
+      "step": 15485
+    },
+    {
+      "epoch": 4.892679459843639,
+      "grad_norm": 0.05238293396836016,
+      "learning_rate": 1.3645509853357774e-06,
+      "loss": 2.7149,
+      "step": 15490
+    },
+    {
+      "epoch": 4.894258864408118,
+      "grad_norm": 0.054587768861856456,
+      "learning_rate": 1.3241399508133744e-06,
+      "loss": 2.837,
+      "step": 15495
+    },
+    {
+      "epoch": 4.895838268972597,
+      "grad_norm": 0.0798321225020181,
+      "learning_rate": 1.28433553540791e-06,
+      "loss": 2.7395,
+      "step": 15500
+    },
+    {
+      "epoch": 4.897417673537077,
+      "grad_norm": 0.07792929085811533,
+      "learning_rate": 1.245137787539874e-06,
+      "loss": 2.7809,
+      "step": 15505
+    },
+    {
+      "epoch": 4.898997078101556,
+      "grad_norm": 0.06061549204972541,
+      "learning_rate": 1.2065467548917353e-06,
+      "loss": 2.7374,
+      "step": 15510
+    },
+    {
+      "epoch": 4.900576482666035,
+      "grad_norm": 0.06378125097136801,
+      "learning_rate": 1.1685624844079978e-06,
+      "loss": 2.7399,
+      "step": 15515
+    },
+    {
+      "epoch": 4.902155887230514,
+      "grad_norm": 0.06160280704408907,
+      "learning_rate": 1.1311850222949226e-06,
+      "loss": 2.6053,
+      "step": 15520
+    },
+    {
+      "epoch": 4.903735291794993,
+      "grad_norm": 0.056435727044791974,
+      "learning_rate": 1.0944144140206945e-06,
+      "loss": 2.6938,
+      "step": 15525
+    },
+    {
+      "epoch": 4.905314696359472,
+      "grad_norm": 0.06563342311549544,
+      "learning_rate": 1.0582507043153112e-06,
+      "loss": 2.744,
+      "step": 15530
+    },
+    {
+      "epoch": 4.906894100923951,
+      "grad_norm": 0.06126523878343254,
+      "learning_rate": 1.0226939371704714e-06,
+      "loss": 2.6803,
+      "step": 15535
+    },
+    {
+      "epoch": 4.9084735054884305,
+      "grad_norm": 0.062461213853841034,
+      "learning_rate": 9.877441558395762e-07,
+      "loss": 2.727,
+      "step": 15540
+    },
+    {
+      "epoch": 4.91005291005291,
+      "grad_norm": 0.05783775706720171,
+      "learning_rate": 9.534014028375615e-07,
+      "loss": 2.7089,
+      "step": 15545
+    },
+    {
+      "epoch": 4.911632314617389,
+      "grad_norm": 0.05012947137324418,
+      "learning_rate": 9.196657199410097e-07,
+      "loss": 2.667,
+      "step": 15550
+    },
+    {
+      "epoch": 4.913211719181868,
+      "grad_norm": 0.05529155175103065,
+      "learning_rate": 8.865371481880935e-07,
+      "loss": 2.7449,
+      "step": 15555
+    },
+    {
+      "epoch": 4.9147911237463475,
+      "grad_norm": 0.05983162805552894,
+      "learning_rate": 8.540157278782989e-07,
+      "loss": 2.7101,
+      "step": 15560
+    },
+    {
+      "epoch": 4.916370528310827,
+      "grad_norm": 0.05450651501679739,
+      "learning_rate": 8.221014985727027e-07,
+      "loss": 2.6899,
+      "step": 15565
+    },
+    {
+      "epoch": 4.917949932875306,
+      "grad_norm": 0.05703893233347161,
+      "learning_rate": 7.907944990936389e-07,
+      "loss": 2.6503,
+      "step": 15570
+    },
+    {
+      "epoch": 4.919529337439785,
+      "grad_norm": 0.05595284348716262,
+      "learning_rate": 7.600947675248104e-07,
+      "loss": 2.7128,
+      "step": 15575
+    },
+    {
+      "epoch": 4.9211087420042645,
+      "grad_norm": 0.05411673035204549,
+      "learning_rate": 7.300023412111778e-07,
+      "loss": 2.6972,
+      "step": 15580
+    },
+    {
+      "epoch": 4.922688146568744,
+      "grad_norm": 0.06387772424566021,
+      "learning_rate": 7.005172567590146e-07,
+      "loss": 2.6588,
+      "step": 15585
+    },
+    {
+      "epoch": 4.924267551133223,
+      "grad_norm": 0.05143834451553453,
+      "learning_rate": 6.716395500357963e-07,
+      "loss": 2.7175,
+      "step": 15590
+    },
+    {
+      "epoch": 4.925846955697702,
+      "grad_norm": 0.05708835270307971,
+      "learning_rate": 6.433692561699789e-07,
+      "loss": 2.7446,
+      "step": 15595
+    },
+    {
+      "epoch": 4.9274263602621815,
+      "grad_norm": 0.056247765959520635,
+      "learning_rate": 6.157064095512755e-07,
+      "loss": 2.7048,
+      "step": 15600
+    },
+    {
+      "epoch": 4.929005764826661,
+      "grad_norm": 0.06054402367957489,
+      "learning_rate": 5.886510438304349e-07,
+      "loss": 2.7331,
+      "step": 15605
+    },
+    {
+      "epoch": 4.93058516939114,
+      "grad_norm": 0.06558951718720128,
+      "learning_rate": 5.622031919191861e-07,
+      "loss": 2.6663,
+      "step": 15610
+    },
+    {
+      "epoch": 4.932164573955618,
+      "grad_norm": 0.05795946759771694,
+      "learning_rate": 5.363628859903491e-07,
+      "loss": 2.739,
+      "step": 15615
+    },
+    {
+      "epoch": 4.933743978520098,
+      "grad_norm": 0.05563515738066562,
+      "learning_rate": 5.111301574775573e-07,
+      "loss": 2.767,
+      "step": 15620
+    },
+    {
+      "epoch": 4.935323383084577,
+      "grad_norm": 0.1032472602937949,
+      "learning_rate": 4.865050370754242e-07,
+      "loss": 2.7165,
+      "step": 15625
+    },
+    {
+      "epoch": 4.936902787649056,
+      "grad_norm": 0.05779905262933907,
+      "learning_rate": 4.624875547394325e-07,
+      "loss": 2.7258,
+      "step": 15630
+    },
+    {
+      "epoch": 4.938482192213535,
+      "grad_norm": 0.05461029508193594,
+      "learning_rate": 4.3907773968587804e-07,
+      "loss": 2.7765,
+      "step": 15635
+    },
+    {
+      "epoch": 4.940061596778015,
+      "grad_norm": 0.050173376485073085,
+      "learning_rate": 4.16275620391815e-07,
+      "loss": 2.6581,
+      "step": 15640
+    },
+    {
+      "epoch": 4.941641001342494,
+      "grad_norm": 0.05470222003207647,
+      "learning_rate": 3.9408122459516636e-07,
+      "loss": 2.7145,
+      "step": 15645
+    },
+    {
+      "epoch": 4.943220405906973,
+      "grad_norm": 0.05798096882177209,
+      "learning_rate": 3.7249457929450225e-07,
+      "loss": 2.6783,
+      "step": 15650
+    },
+    {
+      "epoch": 4.944799810471452,
+      "grad_norm": 0.06083505621472735,
+      "learning_rate": 3.5151571074909515e-07,
+      "loss": 2.8144,
+      "step": 15655
+    },
+    {
+      "epoch": 4.946379215035932,
+      "grad_norm": 0.05196750496017329,
+      "learning_rate": 3.311446444789201e-07,
+      "loss": 2.7412,
+      "step": 15660
+    },
+    {
+      "epoch": 4.947958619600411,
+      "grad_norm": 0.056580318756435044,
+      "learning_rate": 3.113814052644881e-07,
+      "loss": 2.6301,
+      "step": 15665
+    },
+    {
+      "epoch": 4.94953802416489,
+      "grad_norm": 0.06440244683881301,
+      "learning_rate": 2.922260171470681e-07,
+      "loss": 2.7279,
+      "step": 15670
+    },
+    {
+      "epoch": 4.951117428729369,
+      "grad_norm": 0.06859737188430524,
+      "learning_rate": 2.736785034284095e-07,
+      "loss": 2.8842,
+      "step": 15675
+    },
+    {
+      "epoch": 4.952696833293849,
+      "grad_norm": 0.058608207485147536,
+      "learning_rate": 2.557388866707977e-07,
+      "loss": 2.7541,
+      "step": 15680
+    },
+    {
+      "epoch": 4.954276237858327,
+      "grad_norm": 0.05435903069168684,
+      "learning_rate": 2.3840718869699852e-07,
+      "loss": 2.7838,
+      "step": 15685
+    },
+    {
+      "epoch": 4.955855642422806,
+      "grad_norm": 0.05554975694652394,
+      "learning_rate": 2.2168343059042472e-07,
+      "loss": 2.7771,
+      "step": 15690
+    },
+    {
+      "epoch": 4.957435046987285,
+      "grad_norm": 0.06402777154756942,
+      "learning_rate": 2.0556763269480304e-07,
+      "loss": 2.7447,
+      "step": 15695
+    },
+    {
+      "epoch": 4.959014451551765,
+      "grad_norm": 0.059406494327933136,
+      "learning_rate": 1.9005981461434064e-07,
+      "loss": 2.7157,
+      "step": 15700
+    },
+    {
+      "epoch": 4.960593856116244,
+      "grad_norm": 0.058115189680075535,
+      "learning_rate": 1.7515999521366954e-07,
+      "loss": 2.7347,
+      "step": 15705
+    },
+    {
+      "epoch": 4.962173260680723,
+      "grad_norm": 0.06345300802524664,
+      "learning_rate": 1.6086819261790232e-07,
+      "loss": 2.7423,
+      "step": 15710
+    },
+    {
+      "epoch": 4.963752665245202,
+      "grad_norm": 0.049900722632364655,
+      "learning_rate": 1.4718442421235434e-07,
+      "loss": 2.683,
+      "step": 15715
+    },
+    {
+      "epoch": 4.965332069809682,
+      "grad_norm": 0.06586977817792498,
+      "learning_rate": 1.3410870664276597e-07,
+      "loss": 2.7338,
+      "step": 15720
+    },
+    {
+      "epoch": 4.966911474374161,
+      "grad_norm": 0.08799394671190368,
+      "learning_rate": 1.216410558153025e-07,
+      "loss": 2.7152,
+      "step": 15725
+    },
+    {
+      "epoch": 4.96849087893864,
+      "grad_norm": 0.07154488795314753,
+      "learning_rate": 1.0978148689633205e-07,
+      "loss": 2.7308,
+      "step": 15730
+    },
+    {
+      "epoch": 4.970070283503119,
+      "grad_norm": 0.05346095356682859,
+      "learning_rate": 9.853001431253672e-08,
+      "loss": 2.7842,
+      "step": 15735
+    },
+    {
+      "epoch": 4.971649688067599,
+      "grad_norm": 0.06197114870761926,
+      "learning_rate": 8.788665175085698e-08,
+      "loss": 2.7664,
+      "step": 15740
+    },
+    {
+      "epoch": 4.973229092632078,
+      "grad_norm": 0.058910878720595174,
+      "learning_rate": 7.785141215849167e-08,
+      "loss": 2.7765,
+      "step": 15745
+    },
+    {
+      "epoch": 4.974808497196557,
+      "grad_norm": 0.06344942449685999,
+      "learning_rate": 6.842430774300912e-08,
+      "loss": 2.7572,
+      "step": 15750
+    },
+    {
+      "epoch": 4.976387901761036,
+      "grad_norm": 0.05046781102909866,
+      "learning_rate": 5.960534997201395e-08,
+      "loss": 2.7001,
+      "step": 15755
+    },
+    {
+      "epoch": 4.977967306325516,
+      "grad_norm": 0.0647640050458623,
+      "learning_rate": 5.139454957342471e-08,
+      "loss": 2.7129,
+      "step": 15760
+    },
+    {
+      "epoch": 4.979546710889995,
+      "grad_norm": 0.0757413197898649,
+      "learning_rate": 4.379191653536285e-08,
+      "loss": 2.7075,
+      "step": 15765
+    },
+    {
+      "epoch": 4.981126115454474,
+      "grad_norm": 0.06596993656633884,
+      "learning_rate": 3.67974601061527e-08,
+      "loss": 2.7624,
+      "step": 15770
+    },
+    {
+      "epoch": 4.9827055200189525,
+      "grad_norm": 0.05902247217698239,
+      "learning_rate": 3.041118879421045e-08,
+      "loss": 2.6504,
+      "step": 15775
+    },
+    {
+      "epoch": 4.984284924583432,
+      "grad_norm": 0.05794600829523902,
+      "learning_rate": 2.463311036826621e-08,
+      "loss": 2.7527,
+      "step": 15780
+    },
+    {
+      "epoch": 4.985864329147911,
+      "grad_norm": 0.0573280053151581,
+      "learning_rate": 1.9463231857030915e-08,
+      "loss": 2.7713,
+      "step": 15785
+    },
+    {
+      "epoch": 4.98744373371239,
+      "grad_norm": 0.07010005276664519,
+      "learning_rate": 1.490155954947392e-08,
+      "loss": 2.7121,
+      "step": 15790
+    },
+    {
+      "epoch": 4.9890231382768695,
+      "grad_norm": 0.060817764954682635,
+      "learning_rate": 1.0948098994711942e-08,
+      "loss": 2.6928,
+      "step": 15795
+    },
+    {
+      "epoch": 4.990602542841349,
+      "grad_norm": 0.04809928983104812,
+      "learning_rate": 7.602855001953568e-09,
+      "loss": 2.8476,
+      "step": 15800
+    },
+    {
+      "epoch": 4.992181947405828,
+      "grad_norm": 0.06480893682473204,
+      "learning_rate": 4.865831640554763e-09,
+      "loss": 2.6281,
+      "step": 15805
+    },
+    {
+      "epoch": 4.993761351970307,
+      "grad_norm": 0.06219200033889189,
+      "learning_rate": 2.7370322400188664e-09,
+      "loss": 2.7179,
+      "step": 15810
+    },
+    {
+      "epoch": 4.9953407565347865,
+      "grad_norm": 0.05647383150360001,
+      "learning_rate": 1.2164593899410825e-09,
+      "loss": 2.7328,
+      "step": 15815
+    },
+    {
+      "epoch": 4.996920161099266,
+      "grad_norm": 0.05182288923095421,
+      "learning_rate": 3.041149399529708e-10,
+      "loss": 2.7402,
+      "step": 15820
+    },
+    {
+      "epoch": 4.998499565663745,
+      "grad_norm": 0.05430101175745804,
+      "learning_rate": 0.0,
+      "loss": 2.6927,
+      "step": 15825
+    },
+    {
+      "epoch": 4.998499565663745,
+      "eval_loss": 2.7424681186676025,
+      "eval_runtime": 118.6065,
+      "eval_samples_per_second": 22.334,
+      "eval_steps_per_second": 5.59,
+      "step": 15825
+    },
+    {
+      "epoch": 4.998499565663745,
+      "step": 15825,
+      "total_flos": 4.574102807563469e+16,
+      "train_loss": 3.298353223815725,
+      "train_runtime": 39698.8648,
+      "train_samples_per_second": 6.379,
+      "train_steps_per_second": 0.399
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 15825,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 5,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": false,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 4.574102807563469e+16,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}