diff --git "a/trainer_state.json" "b/trainer_state.json"
--- "a/trainer_state.json"
+++ "b/trainer_state.json"
@@ -1,9 +1,9 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 0.44441391904394445,
+  "epoch": 0.5413769559262596,
   "eval_steps": 500,
-  "global_step": 110000,
+  "global_step": 134000,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -77007,6 +77007,16806 @@
       "learning_rate": 3.436516483539781e-05,
       "loss": 35.0977,
       "step": 110000
+    },
+    {
+      "epoch": 0.4444543203093121,
+      "grad_norm": 787.899658203125,
+      "learning_rate": 3.4361928261912254e-05,
+      "loss": 56.4767,
+      "step": 110010
+    },
+    {
+      "epoch": 0.44449472157467973,
+      "grad_norm": 0.0,
+      "learning_rate": 3.4358691505911104e-05,
+      "loss": 44.5502,
+      "step": 110020
+    },
+    {
+      "epoch": 0.44453512284004737,
+      "grad_norm": 385.0424499511719,
+      "learning_rate": 3.4355454567457445e-05,
+      "loss": 70.914,
+      "step": 110030
+    },
+    {
+      "epoch": 0.44457552410541495,
+      "grad_norm": 967.9429931640625,
+      "learning_rate": 3.435221744661438e-05,
+      "loss": 62.6088,
+      "step": 110040
+    },
+    {
+      "epoch": 0.4446159253707826,
+      "grad_norm": 389.7696533203125,
+      "learning_rate": 3.434898014344501e-05,
+      "loss": 48.413,
+      "step": 110050
+    },
+    {
+      "epoch": 0.44465632663615023,
+      "grad_norm": 1513.93017578125,
+      "learning_rate": 3.434574265801247e-05,
+      "loss": 50.9386,
+      "step": 110060
+    },
+    {
+      "epoch": 0.4446967279015179,
+      "grad_norm": 461.18096923828125,
+      "learning_rate": 3.4342504990379866e-05,
+      "loss": 66.7544,
+      "step": 110070
+    },
+    {
+      "epoch": 0.4447371291668855,
+      "grad_norm": 593.7630615234375,
+      "learning_rate": 3.433926714061032e-05,
+      "loss": 60.6121,
+      "step": 110080
+    },
+    {
+      "epoch": 0.44477753043225315,
+      "grad_norm": 1564.3096923828125,
+      "learning_rate": 3.433602910876694e-05,
+      "loss": 70.2792,
+      "step": 110090
+    },
+    {
+      "epoch": 0.4448179316976208,
+      "grad_norm": 1443.179931640625,
+      "learning_rate": 3.433279089491288e-05,
+      "loss": 36.5048,
+      "step": 110100
+    },
+    {
+      "epoch": 0.4448583329629884,
+      "grad_norm": 822.4376220703125,
+      "learning_rate": 3.432955249911125e-05,
+      "loss": 78.1726,
+      "step": 110110
+    },
+    {
+      "epoch": 0.444898734228356,
+      "grad_norm": 1508.205078125,
+      "learning_rate": 3.432631392142519e-05,
+      "loss": 66.1732,
+      "step": 110120
+    },
+    {
+      "epoch": 0.44493913549372366,
+      "grad_norm": 251.31320190429688,
+      "learning_rate": 3.432307516191783e-05,
+      "loss": 34.4932,
+      "step": 110130
+    },
+    {
+      "epoch": 0.4449795367590913,
+      "grad_norm": 1120.1636962890625,
+      "learning_rate": 3.4319836220652335e-05,
+      "loss": 38.0345,
+      "step": 110140
+    },
+    {
+      "epoch": 0.44501993802445894,
+      "grad_norm": 439.1837158203125,
+      "learning_rate": 3.431659709769183e-05,
+      "loss": 56.9619,
+      "step": 110150
+    },
+    {
+      "epoch": 0.4450603392898266,
+      "grad_norm": 969.8883666992188,
+      "learning_rate": 3.431335779309947e-05,
+      "loss": 37.3155,
+      "step": 110160
+    },
+    {
+      "epoch": 0.44510074055519416,
+      "grad_norm": 813.2791137695312,
+      "learning_rate": 3.43101183069384e-05,
+      "loss": 82.1769,
+      "step": 110170
+    },
+    {
+      "epoch": 0.4451411418205618,
+      "grad_norm": 712.302001953125,
+      "learning_rate": 3.430687863927178e-05,
+      "loss": 53.5758,
+      "step": 110180
+    },
+    {
+      "epoch": 0.44518154308592944,
+      "grad_norm": 730.143310546875,
+      "learning_rate": 3.4303638790162774e-05,
+      "loss": 61.2881,
+      "step": 110190
+    },
+    {
+      "epoch": 0.4452219443512971,
+      "grad_norm": 945.42041015625,
+      "learning_rate": 3.430039875967454e-05,
+      "loss": 62.6643,
+      "step": 110200
+    },
+    {
+      "epoch": 0.4452623456166647,
+      "grad_norm": 887.8884887695312,
+      "learning_rate": 3.429715854787024e-05,
+      "loss": 52.4225,
+      "step": 110210
+    },
+    {
+      "epoch": 0.44530274688203236,
+      "grad_norm": 778.4572143554688,
+      "learning_rate": 3.429391815481305e-05,
+      "loss": 80.6638,
+      "step": 110220
+    },
+    {
+      "epoch": 0.4453431481474,
+      "grad_norm": 701.6489868164062,
+      "learning_rate": 3.429067758056613e-05,
+      "loss": 70.6435,
+      "step": 110230
+    },
+    {
+      "epoch": 0.4453835494127676,
+      "grad_norm": 1139.0123291015625,
+      "learning_rate": 3.428743682519269e-05,
+      "loss": 78.8899,
+      "step": 110240
+    },
+    {
+      "epoch": 0.4454239506781352,
+      "grad_norm": 0.0,
+      "learning_rate": 3.428419588875588e-05,
+      "loss": 51.6693,
+      "step": 110250
+    },
+    {
+      "epoch": 0.44546435194350287,
+      "grad_norm": 919.1795654296875,
+      "learning_rate": 3.428095477131888e-05,
+      "loss": 56.4408,
+      "step": 110260
+    },
+    {
+      "epoch": 0.4455047532088705,
+      "grad_norm": 953.95263671875,
+      "learning_rate": 3.427771347294489e-05,
+      "loss": 72.902,
+      "step": 110270
+    },
+    {
+      "epoch": 0.44554515447423815,
+      "grad_norm": 1921.18359375,
+      "learning_rate": 3.427447199369711e-05,
+      "loss": 43.3139,
+      "step": 110280
+    },
+    {
+      "epoch": 0.4455855557396058,
+      "grad_norm": 790.10791015625,
+      "learning_rate": 3.4271230333638716e-05,
+      "loss": 71.2949,
+      "step": 110290
+    },
+    {
+      "epoch": 0.44562595700497337,
+      "grad_norm": 1008.7927856445312,
+      "learning_rate": 3.426798849283291e-05,
+      "loss": 60.3911,
+      "step": 110300
+    },
+    {
+      "epoch": 0.445666358270341,
+      "grad_norm": 749.1685791015625,
+      "learning_rate": 3.4264746471342905e-05,
+      "loss": 78.0302,
+      "step": 110310
+    },
+    {
+      "epoch": 0.44570675953570865,
+      "grad_norm": 967.6669921875,
+      "learning_rate": 3.4261504269231904e-05,
+      "loss": 49.5635,
+      "step": 110320
+    },
+    {
+      "epoch": 0.4457471608010763,
+      "grad_norm": 1514.6182861328125,
+      "learning_rate": 3.4258261886563104e-05,
+      "loss": 76.0732,
+      "step": 110330
+    },
+    {
+      "epoch": 0.44578756206644393,
+      "grad_norm": 1716.936279296875,
+      "learning_rate": 3.425501932339971e-05,
+      "loss": 133.2402,
+      "step": 110340
+    },
+    {
+      "epoch": 0.44582796333181157,
+      "grad_norm": 959.5086669921875,
+      "learning_rate": 3.425177657980496e-05,
+      "loss": 71.294,
+      "step": 110350
+    },
+    {
+      "epoch": 0.44586836459717916,
+      "grad_norm": 812.0950317382812,
+      "learning_rate": 3.4248533655842066e-05,
+      "loss": 57.06,
+      "step": 110360
+    },
+    {
+      "epoch": 0.4459087658625468,
+      "grad_norm": 267.0820617675781,
+      "learning_rate": 3.4245290551574237e-05,
+      "loss": 44.4798,
+      "step": 110370
+    },
+    {
+      "epoch": 0.44594916712791444,
+      "grad_norm": 1544.18310546875,
+      "learning_rate": 3.4242047267064715e-05,
+      "loss": 57.8316,
+      "step": 110380
+    },
+    {
+      "epoch": 0.4459895683932821,
+      "grad_norm": 695.9027709960938,
+      "learning_rate": 3.4238803802376716e-05,
+      "loss": 56.366,
+      "step": 110390
+    },
+    {
+      "epoch": 0.4460299696586497,
+      "grad_norm": 685.2529296875,
+      "learning_rate": 3.423556015757349e-05,
+      "loss": 57.191,
+      "step": 110400
+    },
+    {
+      "epoch": 0.44607037092401736,
+      "grad_norm": 653.3543090820312,
+      "learning_rate": 3.423231633271826e-05,
+      "loss": 61.4118,
+      "step": 110410
+    },
+    {
+      "epoch": 0.446110772189385,
+      "grad_norm": 1396.4661865234375,
+      "learning_rate": 3.4229072327874274e-05,
+      "loss": 53.9518,
+      "step": 110420
+    },
+    {
+      "epoch": 0.4461511734547526,
+      "grad_norm": 742.2490234375,
+      "learning_rate": 3.422582814310476e-05,
+      "loss": 64.2797,
+      "step": 110430
+    },
+    {
+      "epoch": 0.4461915747201202,
+      "grad_norm": 811.0308837890625,
+      "learning_rate": 3.4222583778472996e-05,
+      "loss": 34.3383,
+      "step": 110440
+    },
+    {
+      "epoch": 0.44623197598548786,
+      "grad_norm": 784.7843017578125,
+      "learning_rate": 3.421933923404219e-05,
+      "loss": 83.5336,
+      "step": 110450
+    },
+    {
+      "epoch": 0.4462723772508555,
+      "grad_norm": 368.3476257324219,
+      "learning_rate": 3.421609450987563e-05,
+      "loss": 43.9047,
+      "step": 110460
+    },
+    {
+      "epoch": 0.44631277851622314,
+      "grad_norm": 507.54205322265625,
+      "learning_rate": 3.421284960603657e-05,
+      "loss": 40.9963,
+      "step": 110470
+    },
+    {
+      "epoch": 0.4463531797815908,
+      "grad_norm": 1140.799560546875,
+      "learning_rate": 3.4209604522588255e-05,
+      "loss": 46.2735,
+      "step": 110480
+    },
+    {
+      "epoch": 0.44639358104695837,
+      "grad_norm": 1338.29541015625,
+      "learning_rate": 3.4206359259593954e-05,
+      "loss": 37.1206,
+      "step": 110490
+    },
+    {
+      "epoch": 0.446433982312326,
+      "grad_norm": 665.2498168945312,
+      "learning_rate": 3.4203113817116957e-05,
+      "loss": 57.0931,
+      "step": 110500
+    },
+    {
+      "epoch": 0.44647438357769365,
+      "grad_norm": 734.5198974609375,
+      "learning_rate": 3.4199868195220505e-05,
+      "loss": 71.1633,
+      "step": 110510
+    },
+    {
+      "epoch": 0.4465147848430613,
+      "grad_norm": 679.3670043945312,
+      "learning_rate": 3.419662239396789e-05,
+      "loss": 52.7156,
+      "step": 110520
+    },
+    {
+      "epoch": 0.4465551861084289,
+      "grad_norm": 538.1387939453125,
+      "learning_rate": 3.419337641342239e-05,
+      "loss": 87.353,
+      "step": 110530
+    },
+    {
+      "epoch": 0.44659558737379657,
+      "grad_norm": 625.4715576171875,
+      "learning_rate": 3.419013025364727e-05,
+      "loss": 52.2098,
+      "step": 110540
+    },
+    {
+      "epoch": 0.4466359886391642,
+      "grad_norm": 1380.083740234375,
+      "learning_rate": 3.4186883914705835e-05,
+      "loss": 110.8045,
+      "step": 110550
+    },
+    {
+      "epoch": 0.4466763899045318,
+      "grad_norm": 664.3757934570312,
+      "learning_rate": 3.418363739666137e-05,
+      "loss": 33.5992,
+      "step": 110560
+    },
+    {
+      "epoch": 0.44671679116989943,
+      "grad_norm": 352.8798522949219,
+      "learning_rate": 3.418039069957717e-05,
+      "loss": 54.5655,
+      "step": 110570
+    },
+    {
+      "epoch": 0.44675719243526707,
+      "grad_norm": 196.41831970214844,
+      "learning_rate": 3.417714382351652e-05,
+      "loss": 46.0049,
+      "step": 110580
+    },
+    {
+      "epoch": 0.4467975937006347,
+      "grad_norm": 1419.017822265625,
+      "learning_rate": 3.417389676854274e-05,
+      "loss": 87.4692,
+      "step": 110590
+    },
+    {
+      "epoch": 0.44683799496600235,
+      "grad_norm": 515.550537109375,
+      "learning_rate": 3.417064953471911e-05,
+      "loss": 46.8774,
+      "step": 110600
+    },
+    {
+      "epoch": 0.44687839623137,
+      "grad_norm": 729.7348022460938,
+      "learning_rate": 3.416740212210894e-05,
+      "loss": 47.5663,
+      "step": 110610
+    },
+    {
+      "epoch": 0.4469187974967376,
+      "grad_norm": 618.8557739257812,
+      "learning_rate": 3.416415453077555e-05,
+      "loss": 56.8461,
+      "step": 110620
+    },
+    {
+      "epoch": 0.4469591987621052,
+      "grad_norm": 1495.1834716796875,
+      "learning_rate": 3.416090676078225e-05,
+      "loss": 71.6594,
+      "step": 110630
+    },
+    {
+      "epoch": 0.44699960002747285,
+      "grad_norm": 537.9067993164062,
+      "learning_rate": 3.415765881219236e-05,
+      "loss": 75.8224,
+      "step": 110640
+    },
+    {
+      "epoch": 0.4470400012928405,
+      "grad_norm": 1021.9192504882812,
+      "learning_rate": 3.4154410685069196e-05,
+      "loss": 71.1588,
+      "step": 110650
+    },
+    {
+      "epoch": 0.44708040255820813,
+      "grad_norm": 1147.8583984375,
+      "learning_rate": 3.4151162379476075e-05,
+      "loss": 99.0359,
+      "step": 110660
+    },
+    {
+      "epoch": 0.4471208038235758,
+      "grad_norm": 951.2720947265625,
+      "learning_rate": 3.414791389547635e-05,
+      "loss": 62.2096,
+      "step": 110670
+    },
+    {
+      "epoch": 0.44716120508894336,
+      "grad_norm": 292.7242736816406,
+      "learning_rate": 3.414466523313332e-05,
+      "loss": 40.3493,
+      "step": 110680
+    },
+    {
+      "epoch": 0.447201606354311,
+      "grad_norm": 1284.24755859375,
+      "learning_rate": 3.414141639251033e-05,
+      "loss": 49.9673,
+      "step": 110690
+    },
+    {
+      "epoch": 0.44724200761967864,
+      "grad_norm": 1174.4100341796875,
+      "learning_rate": 3.413816737367073e-05,
+      "loss": 76.3328,
+      "step": 110700
+    },
+    {
+      "epoch": 0.4472824088850463,
+      "grad_norm": 1334.654052734375,
+      "learning_rate": 3.4134918176677846e-05,
+      "loss": 68.6847,
+      "step": 110710
+    },
+    {
+      "epoch": 0.4473228101504139,
+      "grad_norm": 1329.33203125,
+      "learning_rate": 3.4131668801595027e-05,
+      "loss": 60.1202,
+      "step": 110720
+    },
+    {
+      "epoch": 0.44736321141578156,
+      "grad_norm": 472.8536071777344,
+      "learning_rate": 3.4128419248485635e-05,
+      "loss": 55.691,
+      "step": 110730
+    },
+    {
+      "epoch": 0.4474036126811492,
+      "grad_norm": 1049.84619140625,
+      "learning_rate": 3.4125169517413e-05,
+      "loss": 70.7531,
+      "step": 110740
+    },
+    {
+      "epoch": 0.4474440139465168,
+      "grad_norm": 527.7225952148438,
+      "learning_rate": 3.412191960844049e-05,
+      "loss": 48.9243,
+      "step": 110750
+    },
+    {
+      "epoch": 0.4474844152118844,
+      "grad_norm": 415.2084045410156,
+      "learning_rate": 3.411866952163146e-05,
+      "loss": 52.9608,
+      "step": 110760
+    },
+    {
+      "epoch": 0.44752481647725206,
+      "grad_norm": 1201.3553466796875,
+      "learning_rate": 3.4115419257049286e-05,
+      "loss": 71.9567,
+      "step": 110770
+    },
+    {
+      "epoch": 0.4475652177426197,
+      "grad_norm": 245.35345458984375,
+      "learning_rate": 3.4112168814757307e-05,
+      "loss": 69.7229,
+      "step": 110780
+    },
+    {
+      "epoch": 0.44760561900798734,
+      "grad_norm": 1845.09033203125,
+      "learning_rate": 3.41089181948189e-05,
+      "loss": 68.3285,
+      "step": 110790
+    },
+    {
+      "epoch": 0.447646020273355,
+      "grad_norm": 874.4844360351562,
+      "learning_rate": 3.410566739729746e-05,
+      "loss": 42.1089,
+      "step": 110800
+    },
+    {
+      "epoch": 0.44768642153872257,
+      "grad_norm": 466.1338806152344,
+      "learning_rate": 3.410241642225633e-05,
+      "loss": 53.4753,
+      "step": 110810
+    },
+    {
+      "epoch": 0.4477268228040902,
+      "grad_norm": 391.43475341796875,
+      "learning_rate": 3.409916526975892e-05,
+      "loss": 56.1511,
+      "step": 110820
+    },
+    {
+      "epoch": 0.44776722406945785,
+      "grad_norm": 1110.3125,
+      "learning_rate": 3.409591393986859e-05,
+      "loss": 56.4363,
+      "step": 110830
+    },
+    {
+      "epoch": 0.4478076253348255,
+      "grad_norm": 441.3172302246094,
+      "learning_rate": 3.409266243264874e-05,
+      "loss": 54.6422,
+      "step": 110840
+    },
+    {
+      "epoch": 0.4478480266001931,
+      "grad_norm": 868.8545532226562,
+      "learning_rate": 3.408941074816275e-05,
+      "loss": 56.9401,
+      "step": 110850
+    },
+    {
+      "epoch": 0.44788842786556077,
+      "grad_norm": 988.2617797851562,
+      "learning_rate": 3.408615888647402e-05,
+      "loss": 58.8403,
+      "step": 110860
+    },
+    {
+      "epoch": 0.4479288291309284,
+      "grad_norm": 1164.9776611328125,
+      "learning_rate": 3.408290684764594e-05,
+      "loss": 63.0026,
+      "step": 110870
+    },
+    {
+      "epoch": 0.447969230396296,
+      "grad_norm": 592.5801391601562,
+      "learning_rate": 3.407965463174192e-05,
+      "loss": 52.5244,
+      "step": 110880
+    },
+    {
+      "epoch": 0.44800963166166363,
+      "grad_norm": 647.7515258789062,
+      "learning_rate": 3.407640223882536e-05,
+      "loss": 29.65,
+      "step": 110890
+    },
+    {
+      "epoch": 0.44805003292703127,
+      "grad_norm": 954.7613525390625,
+      "learning_rate": 3.407314966895966e-05,
+      "loss": 56.6861,
+      "step": 110900
+    },
+    {
+      "epoch": 0.4480904341923989,
+      "grad_norm": 997.4801025390625,
+      "learning_rate": 3.406989692220824e-05,
+      "loss": 72.3521,
+      "step": 110910
+    },
+    {
+      "epoch": 0.44813083545776655,
+      "grad_norm": 308.1066589355469,
+      "learning_rate": 3.4066643998634505e-05,
+      "loss": 79.9528,
+      "step": 110920
+    },
+    {
+      "epoch": 0.4481712367231342,
+      "grad_norm": 1413.3211669921875,
+      "learning_rate": 3.406339089830188e-05,
+      "loss": 90.6204,
+      "step": 110930
+    },
+    {
+      "epoch": 0.4482116379885018,
+      "grad_norm": 761.3001098632812,
+      "learning_rate": 3.406013762127379e-05,
+      "loss": 59.6376,
+      "step": 110940
+    },
+    {
+      "epoch": 0.4482520392538694,
+      "grad_norm": 550.8245239257812,
+      "learning_rate": 3.405688416761364e-05,
+      "loss": 61.1043,
+      "step": 110950
+    },
+    {
+      "epoch": 0.44829244051923706,
+      "grad_norm": 1125.8817138671875,
+      "learning_rate": 3.4053630537384885e-05,
+      "loss": 54.3687,
+      "step": 110960
+    },
+    {
+      "epoch": 0.4483328417846047,
+      "grad_norm": 1198.2652587890625,
+      "learning_rate": 3.4050376730650935e-05,
+      "loss": 52.4602,
+      "step": 110970
+    },
+    {
+      "epoch": 0.44837324304997234,
+      "grad_norm": 293.5941162109375,
+      "learning_rate": 3.4047122747475224e-05,
+      "loss": 48.189,
+      "step": 110980
+    },
+    {
+      "epoch": 0.44841364431534,
+      "grad_norm": 1793.2451171875,
+      "learning_rate": 3.40438685879212e-05,
+      "loss": 58.882,
+      "step": 110990
+    },
+    {
+      "epoch": 0.44845404558070756,
+      "grad_norm": 405.7001037597656,
+      "learning_rate": 3.4040614252052305e-05,
+      "loss": 87.0919,
+      "step": 111000
+    },
+    {
+      "epoch": 0.4484944468460752,
+      "grad_norm": 609.4334716796875,
+      "learning_rate": 3.403735973993198e-05,
+      "loss": 88.5648,
+      "step": 111010
+    },
+    {
+      "epoch": 0.44853484811144284,
+      "grad_norm": 476.433837890625,
+      "learning_rate": 3.403410505162369e-05,
+      "loss": 72.3557,
+      "step": 111020
+    },
+    {
+      "epoch": 0.4485752493768105,
+      "grad_norm": 547.7504272460938,
+      "learning_rate": 3.403085018719085e-05,
+      "loss": 71.4732,
+      "step": 111030
+    },
+    {
+      "epoch": 0.4486156506421781,
+      "grad_norm": 618.4702758789062,
+      "learning_rate": 3.402759514669694e-05,
+      "loss": 44.4472,
+      "step": 111040
+    },
+    {
+      "epoch": 0.44865605190754576,
+      "grad_norm": 731.4344482421875,
+      "learning_rate": 3.4024339930205415e-05,
+      "loss": 53.5866,
+      "step": 111050
+    },
+    {
+      "epoch": 0.4486964531729134,
+      "grad_norm": 457.8855895996094,
+      "learning_rate": 3.402108453777974e-05,
+      "loss": 58.6352,
+      "step": 111060
+    },
+    {
+      "epoch": 0.448736854438281,
+      "grad_norm": 612.9066772460938,
+      "learning_rate": 3.401782896948338e-05,
+      "loss": 80.7433,
+      "step": 111070
+    },
+    {
+      "epoch": 0.4487772557036486,
+      "grad_norm": 590.0531616210938,
+      "learning_rate": 3.401457322537979e-05,
+      "loss": 49.937,
+      "step": 111080
+    },
+    {
+      "epoch": 0.44881765696901627,
+      "grad_norm": 579.5253295898438,
+      "learning_rate": 3.401131730553247e-05,
+      "loss": 53.8527,
+      "step": 111090
+    },
+    {
+      "epoch": 0.4488580582343839,
+      "grad_norm": 761.0218505859375,
+      "learning_rate": 3.400806121000487e-05,
+      "loss": 57.2909,
+      "step": 111100
+    },
+    {
+      "epoch": 0.44889845949975155,
+      "grad_norm": 938.6339721679688,
+      "learning_rate": 3.400480493886048e-05,
+      "loss": 54.5442,
+      "step": 111110
+    },
+    {
+      "epoch": 0.4489388607651192,
+      "grad_norm": 800.48486328125,
+      "learning_rate": 3.400154849216278e-05,
+      "loss": 50.155,
+      "step": 111120
+    },
+    {
+      "epoch": 0.44897926203048677,
+      "grad_norm": 1541.491943359375,
+      "learning_rate": 3.3998291869975266e-05,
+      "loss": 47.9149,
+      "step": 111130
+    },
+    {
+      "epoch": 0.4490196632958544,
+      "grad_norm": 631.569580078125,
+      "learning_rate": 3.399503507236141e-05,
+      "loss": 60.8179,
+      "step": 111140
+    },
+    {
+      "epoch": 0.44906006456122205,
+      "grad_norm": 233.70372009277344,
+      "learning_rate": 3.399177809938472e-05,
+      "loss": 33.1188,
+      "step": 111150
+    },
+    {
+      "epoch": 0.4491004658265897,
+      "grad_norm": 492.7761535644531,
+      "learning_rate": 3.398852095110868e-05,
+      "loss": 62.9832,
+      "step": 111160
+    },
+    {
+      "epoch": 0.44914086709195733,
+      "grad_norm": 691.3985595703125,
+      "learning_rate": 3.398526362759681e-05,
+      "loss": 58.8685,
+      "step": 111170
+    },
+    {
+      "epoch": 0.44918126835732497,
+      "grad_norm": 763.817626953125,
+      "learning_rate": 3.3982006128912584e-05,
+      "loss": 34.1555,
+      "step": 111180
+    },
+    {
+      "epoch": 0.4492216696226926,
+      "grad_norm": 794.317626953125,
+      "learning_rate": 3.3978748455119536e-05,
+      "loss": 71.4623,
+      "step": 111190
+    },
+    {
+      "epoch": 0.4492620708880602,
+      "grad_norm": 423.894287109375,
+      "learning_rate": 3.397549060628116e-05,
+      "loss": 56.8908,
+      "step": 111200
+    },
+    {
+      "epoch": 0.44930247215342783,
+      "grad_norm": 2169.064697265625,
+      "learning_rate": 3.3972232582460974e-05,
+      "loss": 52.3652,
+      "step": 111210
+    },
+    {
+      "epoch": 0.4493428734187955,
+      "grad_norm": 363.28289794921875,
+      "learning_rate": 3.3968974383722495e-05,
+      "loss": 40.4274,
+      "step": 111220
+    },
+    {
+      "epoch": 0.4493832746841631,
+      "grad_norm": 973.9365234375,
+      "learning_rate": 3.3965716010129236e-05,
+      "loss": 106.0774,
+      "step": 111230
+    },
+    {
+      "epoch": 0.44942367594953075,
+      "grad_norm": 3008.888671875,
+      "learning_rate": 3.396245746174473e-05,
+      "loss": 44.22,
+      "step": 111240
+    },
+    {
+      "epoch": 0.4494640772148984,
+      "grad_norm": 926.1411743164062,
+      "learning_rate": 3.39591987386325e-05,
+      "loss": 46.0268,
+      "step": 111250
+    },
+    {
+      "epoch": 0.449504478480266,
+      "grad_norm": 177.50221252441406,
+      "learning_rate": 3.3955939840856096e-05,
+      "loss": 44.916,
+      "step": 111260
+    },
+    {
+      "epoch": 0.4495448797456336,
+      "grad_norm": 779.60888671875,
+      "learning_rate": 3.395268076847902e-05,
+      "loss": 56.0308,
+      "step": 111270
+    },
+    {
+      "epoch": 0.44958528101100126,
+      "grad_norm": 323.57818603515625,
+      "learning_rate": 3.394942152156482e-05,
+      "loss": 40.499,
+      "step": 111280
+    },
+    {
+      "epoch": 0.4496256822763689,
+      "grad_norm": 513.7103881835938,
+      "learning_rate": 3.394616210017705e-05,
+      "loss": 44.3521,
+      "step": 111290
+    },
+    {
+      "epoch": 0.44966608354173654,
+      "grad_norm": 657.2388305664062,
+      "learning_rate": 3.3942902504379235e-05,
+      "loss": 72.6372,
+      "step": 111300
+    },
+    {
+      "epoch": 0.4497064848071042,
+      "grad_norm": 513.4599609375,
+      "learning_rate": 3.3939642734234936e-05,
+      "loss": 118.4347,
+      "step": 111310
+    },
+    {
+      "epoch": 0.44974688607247176,
+      "grad_norm": 719.9033203125,
+      "learning_rate": 3.39363827898077e-05,
+      "loss": 85.1415,
+      "step": 111320
+    },
+    {
+      "epoch": 0.4497872873378394,
+      "grad_norm": 1146.169677734375,
+      "learning_rate": 3.393312267116107e-05,
+      "loss": 60.6316,
+      "step": 111330
+    },
+    {
+      "epoch": 0.44982768860320704,
+      "grad_norm": 633.3748168945312,
+      "learning_rate": 3.392986237835863e-05,
+      "loss": 60.0308,
+      "step": 111340
+    },
+    {
+      "epoch": 0.4498680898685747,
+      "grad_norm": 436.7900390625,
+      "learning_rate": 3.3926601911463915e-05,
+      "loss": 54.0903,
+      "step": 111350
+    },
+    {
+      "epoch": 0.4499084911339423,
+      "grad_norm": 425.1293029785156,
+      "learning_rate": 3.392334127054051e-05,
+      "loss": 59.7452,
+      "step": 111360
+    },
+    {
+      "epoch": 0.44994889239930996,
+      "grad_norm": 1201.4161376953125,
+      "learning_rate": 3.392008045565197e-05,
+      "loss": 56.7192,
+      "step": 111370
+    },
+    {
+      "epoch": 0.4499892936646776,
+      "grad_norm": 287.7637939453125,
+      "learning_rate": 3.391681946686186e-05,
+      "loss": 67.3147,
+      "step": 111380
+    },
+    {
+      "epoch": 0.4500296949300452,
+      "grad_norm": 818.6224975585938,
+      "learning_rate": 3.3913558304233776e-05,
+      "loss": 53.3463,
+      "step": 111390
+    },
+    {
+      "epoch": 0.4500700961954128,
+      "grad_norm": 385.80987548828125,
+      "learning_rate": 3.3910296967831266e-05,
+      "loss": 66.1127,
+      "step": 111400
+    },
+    {
+      "epoch": 0.45011049746078047,
+      "grad_norm": 1808.7755126953125,
+      "learning_rate": 3.3907035457717944e-05,
+      "loss": 67.1233,
+      "step": 111410
+    },
+    {
+      "epoch": 0.4501508987261481,
+      "grad_norm": 512.4915771484375,
+      "learning_rate": 3.390377377395738e-05,
+      "loss": 52.8904,
+      "step": 111420
+    },
+    {
+      "epoch": 0.45019129999151575,
+      "grad_norm": 1636.7056884765625,
+      "learning_rate": 3.3900511916613155e-05,
+      "loss": 101.467,
+      "step": 111430
+    },
+    {
+      "epoch": 0.4502317012568834,
+      "grad_norm": 682.1113891601562,
+      "learning_rate": 3.389724988574887e-05,
+      "loss": 47.2666,
+      "step": 111440
+    },
+    {
+      "epoch": 0.45027210252225097,
+      "grad_norm": 401.15771484375,
+      "learning_rate": 3.389398768142812e-05,
+      "loss": 40.7815,
+      "step": 111450
+    },
+    {
+      "epoch": 0.4503125037876186,
+      "grad_norm": 1251.0804443359375,
+      "learning_rate": 3.389072530371451e-05,
+      "loss": 49.2346,
+      "step": 111460
+    },
+    {
+      "epoch": 0.45035290505298625,
+      "grad_norm": 642.7127075195312,
+      "learning_rate": 3.388746275267162e-05,
+      "loss": 63.2051,
+      "step": 111470
+    },
+    {
+      "epoch": 0.4503933063183539,
+      "grad_norm": 605.9111938476562,
+      "learning_rate": 3.388420002836307e-05,
+      "loss": 52.5846,
+      "step": 111480
+    },
+    {
+      "epoch": 0.45043370758372153,
+      "grad_norm": 1575.560302734375,
+      "learning_rate": 3.3880937130852466e-05,
+      "loss": 35.7747,
+      "step": 111490
+    },
+    {
+      "epoch": 0.45047410884908917,
+      "grad_norm": 645.1834716796875,
+      "learning_rate": 3.387767406020343e-05,
+      "loss": 55.3005,
+      "step": 111500
+    },
+    {
+      "epoch": 0.4505145101144568,
+      "grad_norm": 666.5519409179688,
+      "learning_rate": 3.3874410816479564e-05,
+      "loss": 48.5718,
+      "step": 111510
+    },
+    {
+      "epoch": 0.4505549113798244,
+      "grad_norm": 654.6434326171875,
+      "learning_rate": 3.387114739974448e-05,
+      "loss": 81.415,
+      "step": 111520
+    },
+    {
+      "epoch": 0.45059531264519204,
+      "grad_norm": 2661.026123046875,
+      "learning_rate": 3.3867883810061824e-05,
+      "loss": 84.8381,
+      "step": 111530
+    },
+    {
+      "epoch": 0.4506357139105597,
+      "grad_norm": 2447.82470703125,
+      "learning_rate": 3.38646200474952e-05,
+      "loss": 48.1841,
+      "step": 111540
+    },
+    {
+      "epoch": 0.4506761151759273,
+      "grad_norm": 708.4982299804688,
+      "learning_rate": 3.3861356112108247e-05,
+      "loss": 56.9643,
+      "step": 111550
+    },
+    {
+      "epoch": 0.45071651644129496,
+      "grad_norm": 489.48846435546875,
+      "learning_rate": 3.3858092003964594e-05,
+      "loss": 57.1001,
+      "step": 111560
+    },
+    {
+      "epoch": 0.4507569177066626,
+      "grad_norm": 538.01806640625,
+      "learning_rate": 3.385482772312787e-05,
+      "loss": 43.918,
+      "step": 111570
+    },
+    {
+      "epoch": 0.4507973189720302,
+      "grad_norm": 624.0175170898438,
+      "learning_rate": 3.3851563269661726e-05,
+      "loss": 95.1447,
+      "step": 111580
+    },
+    {
+      "epoch": 0.4508377202373978,
+      "grad_norm": 565.3558959960938,
+      "learning_rate": 3.38482986436298e-05,
+      "loss": 44.5088,
+      "step": 111590
+    },
+    {
+      "epoch": 0.45087812150276546,
+      "grad_norm": 746.841796875,
+      "learning_rate": 3.384503384509574e-05,
+      "loss": 44.0926,
+      "step": 111600
+    },
+    {
+      "epoch": 0.4509185227681331,
+      "grad_norm": 969.9041748046875,
+      "learning_rate": 3.384176887412318e-05,
+      "loss": 38.6924,
+      "step": 111610
+    },
+    {
+      "epoch": 0.45095892403350074,
+      "grad_norm": 971.7702026367188,
+      "learning_rate": 3.38385037307758e-05,
+      "loss": 56.2464,
+      "step": 111620
+    },
+    {
+      "epoch": 0.4509993252988684,
+      "grad_norm": 1089.8248291015625,
+      "learning_rate": 3.383523841511723e-05,
+      "loss": 89.415,
+      "step": 111630
+    },
+    {
+      "epoch": 0.45103972656423597,
+      "grad_norm": 535.305419921875,
+      "learning_rate": 3.3831972927211135e-05,
+      "loss": 41.9581,
+      "step": 111640
+    },
+    {
+      "epoch": 0.4510801278296036,
+      "grad_norm": 1719.795166015625,
+      "learning_rate": 3.382870726712119e-05,
+      "loss": 74.1624,
+      "step": 111650
+    },
+    {
+      "epoch": 0.45112052909497125,
+      "grad_norm": 1085.8828125,
+      "learning_rate": 3.382544143491104e-05,
+      "loss": 81.0191,
+      "step": 111660
+    },
+    {
+      "epoch": 0.4511609303603389,
+      "grad_norm": 409.8812561035156,
+      "learning_rate": 3.382217543064438e-05,
+      "loss": 40.052,
+      "step": 111670
+    },
+    {
+      "epoch": 0.4512013316257065,
+      "grad_norm": 1443.2547607421875,
+      "learning_rate": 3.381890925438486e-05,
+      "loss": 37.1698,
+      "step": 111680
+    },
+    {
+      "epoch": 0.45124173289107417,
+      "grad_norm": 1432.521728515625,
+      "learning_rate": 3.3815642906196156e-05,
+      "loss": 75.6335,
+      "step": 111690
+    },
+    {
+      "epoch": 0.4512821341564418,
+      "grad_norm": 1234.38720703125,
+      "learning_rate": 3.381237638614196e-05,
+      "loss": 48.0366,
+      "step": 111700
+    },
+    {
+      "epoch": 0.4513225354218094,
+      "grad_norm": 840.4757690429688,
+      "learning_rate": 3.380910969428596e-05,
+      "loss": 58.3024,
+      "step": 111710
+    },
+    {
+      "epoch": 0.45136293668717703,
+      "grad_norm": 591.4329833984375,
+      "learning_rate": 3.380584283069183e-05,
+      "loss": 42.8441,
+      "step": 111720
+    },
+    {
+      "epoch": 0.45140333795254467,
+      "grad_norm": 579.3867797851562,
+      "learning_rate": 3.380257579542325e-05,
+      "loss": 100.9832,
+      "step": 111730
+    },
+    {
+      "epoch": 0.4514437392179123,
+      "grad_norm": 1018.6928100585938,
+      "learning_rate": 3.379930858854392e-05,
+      "loss": 40.5557,
+      "step": 111740
+    },
+    {
+      "epoch": 0.45148414048327995,
+      "grad_norm": 475.0141906738281,
+      "learning_rate": 3.3796041210117546e-05,
+      "loss": 47.8023,
+      "step": 111750
+    },
+    {
+      "epoch": 0.4515245417486476,
+      "grad_norm": 1127.599853515625,
+      "learning_rate": 3.379277366020782e-05,
+      "loss": 56.3643,
+      "step": 111760
+    },
+    {
+      "epoch": 0.4515649430140152,
+      "grad_norm": 632.768310546875,
+      "learning_rate": 3.3789505938878443e-05,
+      "loss": 42.618,
+      "step": 111770
+    },
+    {
+      "epoch": 0.4516053442793828,
+      "grad_norm": 558.4849853515625,
+      "learning_rate": 3.378623804619313e-05,
+      "loss": 38.2455,
+      "step": 111780
+    },
+    {
+      "epoch": 0.45164574554475045,
+      "grad_norm": 1455.6197509765625,
+      "learning_rate": 3.378296998221557e-05,
+      "loss": 54.4919,
+      "step": 111790
+    },
+    {
+      "epoch": 0.4516861468101181,
+      "grad_norm": 650.8120727539062,
+      "learning_rate": 3.3779701747009504e-05,
+      "loss": 68.0886,
+      "step": 111800
+    },
+    {
+      "epoch": 0.45172654807548573,
+      "grad_norm": 613.3826904296875,
+      "learning_rate": 3.377643334063862e-05,
+      "loss": 47.3737,
+      "step": 111810
+    },
+    {
+      "epoch": 0.4517669493408534,
+      "grad_norm": 814.562744140625,
+      "learning_rate": 3.3773164763166655e-05,
+      "loss": 70.561,
+      "step": 111820
+    },
+    {
+      "epoch": 0.451807350606221,
+      "grad_norm": 785.9385986328125,
+      "learning_rate": 3.376989601465733e-05,
+      "loss": 38.8111,
+      "step": 111830
+    },
+    {
+      "epoch": 0.4518477518715886,
+      "grad_norm": 640.478759765625,
+      "learning_rate": 3.376662709517435e-05,
+      "loss": 55.49,
+      "step": 111840
+    },
+    {
+      "epoch": 0.45188815313695624,
+      "grad_norm": 590.5025024414062,
+      "learning_rate": 3.3763358004781475e-05,
+      "loss": 59.1937,
+      "step": 111850
+    },
+    {
+      "epoch": 0.4519285544023239,
+      "grad_norm": 681.9274291992188,
+      "learning_rate": 3.3760088743542424e-05,
+      "loss": 121.1419,
+      "step": 111860
+    },
+    {
+      "epoch": 0.4519689556676915,
+      "grad_norm": 1086.4017333984375,
+      "learning_rate": 3.375681931152093e-05,
+      "loss": 57.0334,
+      "step": 111870
+    },
+    {
+      "epoch": 0.45200935693305916,
+      "grad_norm": 614.7588500976562,
+      "learning_rate": 3.375354970878073e-05,
+      "loss": 51.712,
+      "step": 111880
+    },
+    {
+      "epoch": 0.4520497581984268,
+      "grad_norm": 1066.3997802734375,
+      "learning_rate": 3.375027993538559e-05,
+      "loss": 62.087,
+      "step": 111890
+    },
+    {
+      "epoch": 0.4520901594637944,
+      "grad_norm": 644.5421142578125,
+      "learning_rate": 3.374700999139923e-05,
+      "loss": 58.5303,
+      "step": 111900
+    },
+    {
+      "epoch": 0.452130560729162,
+      "grad_norm": 840.0023193359375,
+      "learning_rate": 3.37437398768854e-05,
+      "loss": 71.8506,
+      "step": 111910
+    },
+    {
+      "epoch": 0.45217096199452966,
+      "grad_norm": 1441.012451171875,
+      "learning_rate": 3.374046959190786e-05,
+      "loss": 92.4958,
+      "step": 111920
+    },
+    {
+      "epoch": 0.4522113632598973,
+      "grad_norm": 1386.08349609375,
+      "learning_rate": 3.3737199136530364e-05,
+      "loss": 51.7875,
+      "step": 111930
+    },
+    {
+      "epoch": 0.45225176452526494,
+      "grad_norm": 1856.324951171875,
+      "learning_rate": 3.373392851081668e-05,
+      "loss": 74.6996,
+      "step": 111940
+    },
+    {
+      "epoch": 0.4522921657906326,
+      "grad_norm": 371.86566162109375,
+      "learning_rate": 3.373065771483056e-05,
+      "loss": 53.6041,
+      "step": 111950
+    },
+    {
+      "epoch": 0.45233256705600017,
+      "grad_norm": 862.8280639648438,
+      "learning_rate": 3.372738674863577e-05,
+      "loss": 51.361,
+      "step": 111960
+    },
+    {
+      "epoch": 0.4523729683213678,
+      "grad_norm": 923.8464965820312,
+      "learning_rate": 3.372411561229609e-05,
+      "loss": 43.7514,
+      "step": 111970
+    },
+    {
+      "epoch": 0.45241336958673545,
+      "grad_norm": 716.6856079101562,
+      "learning_rate": 3.372084430587528e-05,
+      "loss": 39.6764,
+      "step": 111980
+    },
+    {
+      "epoch": 0.4524537708521031,
+      "grad_norm": 593.4553833007812,
+      "learning_rate": 3.371757282943712e-05,
+      "loss": 39.9165,
+      "step": 111990
+    },
+    {
+      "epoch": 0.4524941721174707,
+      "grad_norm": 496.314208984375,
+      "learning_rate": 3.3714301183045385e-05,
+      "loss": 78.6274,
+      "step": 112000
+    },
+    {
+      "epoch": 0.45253457338283837,
+      "grad_norm": 811.6598510742188,
+      "learning_rate": 3.3711029366763866e-05,
+      "loss": 34.8752,
+      "step": 112010
+    },
+    {
+      "epoch": 0.452574974648206,
+      "grad_norm": 1419.599609375,
+      "learning_rate": 3.370775738065634e-05,
+      "loss": 81.8719,
+      "step": 112020
+    },
+    {
+      "epoch": 0.4526153759135736,
+      "grad_norm": 1418.275146484375,
+      "learning_rate": 3.370448522478661e-05,
+      "loss": 65.6506,
+      "step": 112030
+    },
+    {
+      "epoch": 0.45265577717894123,
+      "grad_norm": 8803.1748046875,
+      "learning_rate": 3.370121289921845e-05,
+      "loss": 112.1481,
+      "step": 112040
+    },
+    {
+      "epoch": 0.45269617844430887,
+      "grad_norm": 328.6112060546875,
+      "learning_rate": 3.369794040401567e-05,
+      "loss": 38.9246,
+      "step": 112050
+    },
+    {
+      "epoch": 0.4527365797096765,
+      "grad_norm": 576.2307739257812,
+      "learning_rate": 3.3694667739242066e-05,
+      "loss": 54.7131,
+      "step": 112060
+    },
+    {
+      "epoch": 0.45277698097504415,
+      "grad_norm": 1210.724365234375,
+      "learning_rate": 3.369139490496144e-05,
+      "loss": 60.4555,
+      "step": 112070
+    },
+    {
+      "epoch": 0.4528173822404118,
+      "grad_norm": 1678.520751953125,
+      "learning_rate": 3.368812190123759e-05,
+      "loss": 53.7549,
+      "step": 112080
+    },
+    {
+      "epoch": 0.4528577835057794,
+      "grad_norm": 754.4735717773438,
+      "learning_rate": 3.3684848728134334e-05,
+      "loss": 55.5762,
+      "step": 112090
+    },
+    {
+      "epoch": 0.452898184771147,
+      "grad_norm": 294.9254455566406,
+      "learning_rate": 3.368157538571548e-05,
+      "loss": 62.4506,
+      "step": 112100
+    },
+    {
+      "epoch": 0.45293858603651466,
+      "grad_norm": 1608.3568115234375,
+      "learning_rate": 3.367830187404484e-05,
+      "loss": 55.6337,
+      "step": 112110
+    },
+    {
+      "epoch": 0.4529789873018823,
+      "grad_norm": 680.6564331054688,
+      "learning_rate": 3.367502819318624e-05,
+      "loss": 49.9141,
+      "step": 112120
+    },
+    {
+      "epoch": 0.45301938856724994,
+      "grad_norm": 1690.9805908203125,
+      "learning_rate": 3.36717543432035e-05,
+      "loss": 58.1772,
+      "step": 112130
+    },
+    {
+      "epoch": 0.4530597898326176,
+      "grad_norm": 564.469482421875,
+      "learning_rate": 3.366848032416045e-05,
+      "loss": 59.8349,
+      "step": 112140
+    },
+    {
+      "epoch": 0.4531001910979852,
+      "grad_norm": 725.934326171875,
+      "learning_rate": 3.3665206136120906e-05,
+      "loss": 72.5532,
+      "step": 112150
+    },
+    {
+      "epoch": 0.4531405923633528,
+      "grad_norm": 949.9281616210938,
+      "learning_rate": 3.3661931779148707e-05,
+      "loss": 65.5223,
+      "step": 112160
+    },
+    {
+      "epoch": 0.45318099362872044,
+      "grad_norm": 1123.207763671875,
+      "learning_rate": 3.365865725330769e-05,
+      "loss": 52.4678,
+      "step": 112170
+    },
+    {
+      "epoch": 0.4532213948940881,
+      "grad_norm": 264.70001220703125,
+      "learning_rate": 3.3655382558661685e-05,
+      "loss": 64.133,
+      "step": 112180
+    },
+    {
+      "epoch": 0.4532617961594557,
+      "grad_norm": 1888.4464111328125,
+      "learning_rate": 3.3652107695274555e-05,
+      "loss": 45.2613,
+      "step": 112190
+    },
+    {
+      "epoch": 0.45330219742482336,
+      "grad_norm": 539.1635131835938,
+      "learning_rate": 3.3648832663210124e-05,
+      "loss": 95.8625,
+      "step": 112200
+    },
+    {
+      "epoch": 0.453342598690191,
+      "grad_norm": 510.55816650390625,
+      "learning_rate": 3.3645557462532245e-05,
+      "loss": 39.8649,
+      "step": 112210
+    },
+    {
+      "epoch": 0.4533829999555586,
+      "grad_norm": 1129.6290283203125,
+      "learning_rate": 3.364228209330477e-05,
+      "loss": 54.8464,
+      "step": 112220
+    },
+    {
+      "epoch": 0.4534234012209262,
+      "grad_norm": 718.903564453125,
+      "learning_rate": 3.363900655559157e-05,
+      "loss": 70.0421,
+      "step": 112230
+    },
+    {
+      "epoch": 0.45346380248629387,
+      "grad_norm": 765.8641967773438,
+      "learning_rate": 3.363573084945648e-05,
+      "loss": 43.2255,
+      "step": 112240
+    },
+    {
+      "epoch": 0.4535042037516615,
+      "grad_norm": 305.63763427734375,
+      "learning_rate": 3.363245497496337e-05,
+      "loss": 55.8746,
+      "step": 112250
+    },
+    {
+      "epoch": 0.45354460501702915,
+      "grad_norm": 359.51715087890625,
+      "learning_rate": 3.362917893217611e-05,
+      "loss": 73.9199,
+      "step": 112260
+    },
+    {
+      "epoch": 0.4535850062823968,
+      "grad_norm": 516.9877319335938,
+      "learning_rate": 3.362590272115855e-05,
+      "loss": 67.4331,
+      "step": 112270
+    },
+    {
+      "epoch": 0.45362540754776437,
+      "grad_norm": 862.2160034179688,
+      "learning_rate": 3.3622626341974594e-05,
+      "loss": 57.4161,
+      "step": 112280
+    },
+    {
+      "epoch": 0.453665808813132,
+      "grad_norm": 1069.54150390625,
+      "learning_rate": 3.361934979468809e-05,
+      "loss": 57.9138,
+      "step": 112290
+    },
+    {
+      "epoch": 0.45370621007849965,
+      "grad_norm": 375.27825927734375,
+      "learning_rate": 3.3616073079362926e-05,
+      "loss": 51.6367,
+      "step": 112300
+    },
+    {
+      "epoch": 0.4537466113438673,
+      "grad_norm": 450.5061950683594,
+      "learning_rate": 3.361279619606299e-05,
+      "loss": 61.73,
+      "step": 112310
+    },
+    {
+      "epoch": 0.45378701260923493,
+      "grad_norm": 1464.5079345703125,
+      "learning_rate": 3.360951914485215e-05,
+      "loss": 52.9956,
+      "step": 112320
+    },
+    {
+      "epoch": 0.45382741387460257,
+      "grad_norm": 324.18194580078125,
+      "learning_rate": 3.3606241925794295e-05,
+      "loss": 47.1093,
+      "step": 112330
+    },
+    {
+      "epoch": 0.4538678151399702,
+      "grad_norm": 836.4578857421875,
+      "learning_rate": 3.360296453895333e-05,
+      "loss": 56.2767,
+      "step": 112340
+    },
+    {
+      "epoch": 0.4539082164053378,
+      "grad_norm": 1074.7415771484375,
+      "learning_rate": 3.3599686984393134e-05,
+      "loss": 50.7534,
+      "step": 112350
+    },
+    {
+      "epoch": 0.45394861767070543,
+      "grad_norm": 225.61444091796875,
+      "learning_rate": 3.359640926217763e-05,
+      "loss": 40.0397,
+      "step": 112360
+    },
+    {
+      "epoch": 0.4539890189360731,
+      "grad_norm": 1213.6234130859375,
+      "learning_rate": 3.359313137237069e-05,
+      "loss": 47.421,
+      "step": 112370
+    },
+    {
+      "epoch": 0.4540294202014407,
+      "grad_norm": 640.165283203125,
+      "learning_rate": 3.3589853315036225e-05,
+      "loss": 60.4271,
+      "step": 112380
+    },
+    {
+      "epoch": 0.45406982146680835,
+      "grad_norm": 354.8425598144531,
+      "learning_rate": 3.358657509023815e-05,
+      "loss": 58.7236,
+      "step": 112390
+    },
+    {
+      "epoch": 0.454110222732176,
+      "grad_norm": 1010.51123046875,
+      "learning_rate": 3.3583296698040384e-05,
+      "loss": 67.7665,
+      "step": 112400
+    },
+    {
+      "epoch": 0.4541506239975436,
+      "grad_norm": 313.4957580566406,
+      "learning_rate": 3.3580018138506824e-05,
+      "loss": 73.5613,
+      "step": 112410
+    },
+    {
+      "epoch": 0.4541910252629112,
+      "grad_norm": 623.6607055664062,
+      "learning_rate": 3.3576739411701394e-05,
+      "loss": 58.9651,
+      "step": 112420
+    },
+    {
+      "epoch": 0.45423142652827886,
+      "grad_norm": 870.5923461914062,
+      "learning_rate": 3.357346051768801e-05,
+      "loss": 64.04,
+      "step": 112430
+    },
+    {
+      "epoch": 0.4542718277936465,
+      "grad_norm": 499.6788330078125,
+      "learning_rate": 3.35701814565306e-05,
+      "loss": 59.2575,
+      "step": 112440
+    },
+    {
+      "epoch": 0.45431222905901414,
+      "grad_norm": 902.0311889648438,
+      "learning_rate": 3.356690222829309e-05,
+      "loss": 37.0818,
+      "step": 112450
+    },
+    {
+      "epoch": 0.4543526303243818,
+      "grad_norm": 713.7149047851562,
+      "learning_rate": 3.356362283303941e-05,
+      "loss": 52.5913,
+      "step": 112460
+    },
+    {
+      "epoch": 0.45439303158974936,
+      "grad_norm": 671.7296142578125,
+      "learning_rate": 3.3560343270833495e-05,
+      "loss": 53.5628,
+      "step": 112470
+    },
+    {
+      "epoch": 0.454433432855117,
+      "grad_norm": 1461.3150634765625,
+      "learning_rate": 3.355706354173928e-05,
+      "loss": 78.7695,
+      "step": 112480
+    },
+    {
+      "epoch": 0.45447383412048464,
+      "grad_norm": 1066.4896240234375,
+      "learning_rate": 3.3553783645820715e-05,
+      "loss": 74.929,
+      "step": 112490
+    },
+    {
+      "epoch": 0.4545142353858523,
+      "grad_norm": 257.38531494140625,
+      "learning_rate": 3.355050358314172e-05,
+      "loss": 39.4283,
+      "step": 112500
+    },
+    {
+      "epoch": 0.4545546366512199,
+      "grad_norm": 685.919677734375,
+      "learning_rate": 3.354722335376626e-05,
+      "loss": 51.0791,
+      "step": 112510
+    },
+    {
+      "epoch": 0.45459503791658756,
+      "grad_norm": 1068.9114990234375,
+      "learning_rate": 3.354394295775829e-05,
+      "loss": 44.3589,
+      "step": 112520
+    },
+    {
+      "epoch": 0.4546354391819552,
+      "grad_norm": 604.07568359375,
+      "learning_rate": 3.354066239518174e-05,
+      "loss": 65.293,
+      "step": 112530
+    },
+    {
+      "epoch": 0.4546758404473228,
+      "grad_norm": 1334.7886962890625,
+      "learning_rate": 3.353738166610058e-05,
+      "loss": 52.1476,
+      "step": 112540
+    },
+    {
+      "epoch": 0.4547162417126904,
+      "grad_norm": 517.1060791015625,
+      "learning_rate": 3.353410077057877e-05,
+      "loss": 80.8657,
+      "step": 112550
+    },
+    {
+      "epoch": 0.45475664297805807,
+      "grad_norm": 479.4462890625,
+      "learning_rate": 3.3530819708680286e-05,
+      "loss": 54.3628,
+      "step": 112560
+    },
+    {
+      "epoch": 0.4547970442434257,
+      "grad_norm": 579.92041015625,
+      "learning_rate": 3.352753848046907e-05,
+      "loss": 41.2108,
+      "step": 112570
+    },
+    {
+      "epoch": 0.45483744550879335,
+      "grad_norm": 557.235107421875,
+      "learning_rate": 3.3524257086009104e-05,
+      "loss": 33.0313,
+      "step": 112580
+    },
+    {
+      "epoch": 0.454877846774161,
+      "grad_norm": 929.7172241210938,
+      "learning_rate": 3.352097552536435e-05,
+      "loss": 40.3618,
+      "step": 112590
+    },
+    {
+      "epoch": 0.45491824803952857,
+      "grad_norm": 1180.1580810546875,
+      "learning_rate": 3.35176937985988e-05,
+      "loss": 73.8001,
+      "step": 112600
+    },
+    {
+      "epoch": 0.4549586493048962,
+      "grad_norm": 931.85009765625,
+      "learning_rate": 3.351441190577642e-05,
+      "loss": 45.0227,
+      "step": 112610
+    },
+    {
+      "epoch": 0.45499905057026385,
+      "grad_norm": 911.3134155273438,
+      "learning_rate": 3.3511129846961184e-05,
+      "loss": 59.2229,
+      "step": 112620
+    },
+    {
+      "epoch": 0.4550394518356315,
+      "grad_norm": 675.170654296875,
+      "learning_rate": 3.35078476222171e-05,
+      "loss": 80.4119,
+      "step": 112630
+    },
+    {
+      "epoch": 0.45507985310099913,
+      "grad_norm": 1176.3988037109375,
+      "learning_rate": 3.350456523160815e-05,
+      "loss": 47.3479,
+      "step": 112640
+    },
+    {
+      "epoch": 0.45512025436636677,
+      "grad_norm": 969.1943359375,
+      "learning_rate": 3.350128267519832e-05,
+      "loss": 81.494,
+      "step": 112650
+    },
+    {
+      "epoch": 0.4551606556317344,
+      "grad_norm": 291.8066711425781,
+      "learning_rate": 3.349799995305162e-05,
+      "loss": 41.662,
+      "step": 112660
+    },
+    {
+      "epoch": 0.455201056897102,
+      "grad_norm": 658.59765625,
+      "learning_rate": 3.3494717065232016e-05,
+      "loss": 61.2986,
+      "step": 112670
+    },
+    {
+      "epoch": 0.45524145816246964,
+      "grad_norm": 917.7549438476562,
+      "learning_rate": 3.349143401180354e-05,
+      "loss": 63.2453,
+      "step": 112680
+    },
+    {
+      "epoch": 0.4552818594278373,
+      "grad_norm": 453.6599426269531,
+      "learning_rate": 3.348815079283018e-05,
+      "loss": 59.7746,
+      "step": 112690
+    },
+    {
+      "epoch": 0.4553222606932049,
+      "grad_norm": 850.1251220703125,
+      "learning_rate": 3.3484867408375954e-05,
+      "loss": 57.4509,
+      "step": 112700
+    },
+    {
+      "epoch": 0.45536266195857256,
+      "grad_norm": 701.5401000976562,
+      "learning_rate": 3.348158385850487e-05,
+      "loss": 41.1296,
+      "step": 112710
+    },
+    {
+      "epoch": 0.4554030632239402,
+      "grad_norm": 819.9852294921875,
+      "learning_rate": 3.347830014328094e-05,
+      "loss": 69.8444,
+      "step": 112720
+    },
+    {
+      "epoch": 0.4554434644893078,
+      "grad_norm": 720.2810668945312,
+      "learning_rate": 3.347501626276819e-05,
+      "loss": 75.9919,
+      "step": 112730
+    },
+    {
+      "epoch": 0.4554838657546754,
+      "grad_norm": 495.089599609375,
+      "learning_rate": 3.3471732217030625e-05,
+      "loss": 40.4414,
+      "step": 112740
+    },
+    {
+      "epoch": 0.45552426702004306,
+      "grad_norm": 573.2144775390625,
+      "learning_rate": 3.346844800613229e-05,
+      "loss": 43.111,
+      "step": 112750
+    },
+    {
+      "epoch": 0.4555646682854107,
+      "grad_norm": 1056.5557861328125,
+      "learning_rate": 3.346516363013719e-05,
+      "loss": 92.2572,
+      "step": 112760
+    },
+    {
+      "epoch": 0.45560506955077834,
+      "grad_norm": 487.2084655761719,
+      "learning_rate": 3.346187908910938e-05,
+      "loss": 36.6314,
+      "step": 112770
+    },
+    {
+      "epoch": 0.455645470816146,
+      "grad_norm": 501.5868225097656,
+      "learning_rate": 3.345859438311287e-05,
+      "loss": 56.4939,
+      "step": 112780
+    },
+    {
+      "epoch": 0.45568587208151357,
+      "grad_norm": 792.0189819335938,
+      "learning_rate": 3.345530951221171e-05,
+      "loss": 94.0912,
+      "step": 112790
+    },
+    {
+      "epoch": 0.4557262733468812,
+      "grad_norm": 778.9989624023438,
+      "learning_rate": 3.3452024476469934e-05,
+      "loss": 40.7863,
+      "step": 112800
+    },
+    {
+      "epoch": 0.45576667461224885,
+      "grad_norm": 1507.0234375,
+      "learning_rate": 3.3448739275951595e-05,
+      "loss": 61.6816,
+      "step": 112810
+    },
+    {
+      "epoch": 0.4558070758776165,
+      "grad_norm": 432.25244140625,
+      "learning_rate": 3.344545391072073e-05,
+      "loss": 44.9511,
+      "step": 112820
+    },
+    {
+      "epoch": 0.4558474771429841,
+      "grad_norm": 655.8712158203125,
+      "learning_rate": 3.34421683808414e-05,
+      "loss": 48.9905,
+      "step": 112830
+    },
+    {
+      "epoch": 0.45588787840835177,
+      "grad_norm": 940.5873413085938,
+      "learning_rate": 3.343888268637765e-05,
+      "loss": 59.8554,
+      "step": 112840
+    },
+    {
+      "epoch": 0.4559282796737194,
+      "grad_norm": 1201.72412109375,
+      "learning_rate": 3.343559682739353e-05,
+      "loss": 85.8798,
+      "step": 112850
+    },
+    {
+      "epoch": 0.455968680939087,
+      "grad_norm": 822.5703735351562,
+      "learning_rate": 3.343231080395312e-05,
+      "loss": 54.6579,
+      "step": 112860
+    },
+    {
+      "epoch": 0.45600908220445463,
+      "grad_norm": 476.291748046875,
+      "learning_rate": 3.342902461612045e-05,
+      "loss": 40.1478,
+      "step": 112870
+    },
+    {
+      "epoch": 0.45604948346982227,
+      "grad_norm": 1140.7523193359375,
+      "learning_rate": 3.3425738263959615e-05,
+      "loss": 36.0184,
+      "step": 112880
+    },
+    {
+      "epoch": 0.4560898847351899,
+      "grad_norm": 627.5087280273438,
+      "learning_rate": 3.3422451747534684e-05,
+      "loss": 67.8078,
+      "step": 112890
+    },
+    {
+      "epoch": 0.45613028600055755,
+      "grad_norm": 525.3790283203125,
+      "learning_rate": 3.3419165066909705e-05,
+      "loss": 47.1049,
+      "step": 112900
+    },
+    {
+      "epoch": 0.4561706872659252,
+      "grad_norm": 361.6174011230469,
+      "learning_rate": 3.3415878222148776e-05,
+      "loss": 66.291,
+      "step": 112910
+    },
+    {
+      "epoch": 0.4562110885312928,
+      "grad_norm": 1267.40966796875,
+      "learning_rate": 3.341259121331597e-05,
+      "loss": 39.7781,
+      "step": 112920
+    },
+    {
+      "epoch": 0.4562514897966604,
+      "grad_norm": 643.3392333984375,
+      "learning_rate": 3.340930404047537e-05,
+      "loss": 37.3639,
+      "step": 112930
+    },
+    {
+      "epoch": 0.45629189106202805,
+      "grad_norm": 481.6955871582031,
+      "learning_rate": 3.3406016703691055e-05,
+      "loss": 55.8988,
+      "step": 112940
+    },
+    {
+      "epoch": 0.4563322923273957,
+      "grad_norm": 650.4725952148438,
+      "learning_rate": 3.340272920302711e-05,
+      "loss": 91.6434,
+      "step": 112950
+    },
+    {
+      "epoch": 0.45637269359276333,
+      "grad_norm": 674.4364013671875,
+      "learning_rate": 3.339944153854764e-05,
+      "loss": 46.4506,
+      "step": 112960
+    },
+    {
+      "epoch": 0.456413094858131,
+      "grad_norm": 542.6135864257812,
+      "learning_rate": 3.3396153710316736e-05,
+      "loss": 54.1574,
+      "step": 112970
+    },
+    {
+      "epoch": 0.4564534961234986,
+      "grad_norm": 1714.6644287109375,
+      "learning_rate": 3.339286571839848e-05,
+      "loss": 96.3534,
+      "step": 112980
+    },
+    {
+      "epoch": 0.4564938973888662,
+      "grad_norm": 525.607177734375,
+      "learning_rate": 3.338957756285699e-05,
+      "loss": 67.7952,
+      "step": 112990
+    },
+    {
+      "epoch": 0.45653429865423384,
+      "grad_norm": 1665.643798828125,
+      "learning_rate": 3.338628924375638e-05,
+      "loss": 69.8561,
+      "step": 113000
+    },
+    {
+      "epoch": 0.4565746999196015,
+      "grad_norm": 1218.552490234375,
+      "learning_rate": 3.338300076116073e-05,
+      "loss": 91.2745,
+      "step": 113010
+    },
+    {
+      "epoch": 0.4566151011849691,
+      "grad_norm": 1521.2366943359375,
+      "learning_rate": 3.337971211513417e-05,
+      "loss": 59.6287,
+      "step": 113020
+    },
+    {
+      "epoch": 0.45665550245033676,
+      "grad_norm": 898.8818969726562,
+      "learning_rate": 3.337642330574081e-05,
+      "loss": 54.526,
+      "step": 113030
+    },
+    {
+      "epoch": 0.4566959037157044,
+      "grad_norm": 390.90435791015625,
+      "learning_rate": 3.3373134333044756e-05,
+      "loss": 35.1512,
+      "step": 113040
+    },
+    {
+      "epoch": 0.456736304981072,
+      "grad_norm": 788.0700073242188,
+      "learning_rate": 3.336984519711015e-05,
+      "loss": 46.5823,
+      "step": 113050
+    },
+    {
+      "epoch": 0.4567767062464396,
+      "grad_norm": 979.811767578125,
+      "learning_rate": 3.336655589800109e-05,
+      "loss": 58.1711,
+      "step": 113060
+    },
+    {
+      "epoch": 0.45681710751180726,
+      "grad_norm": 1041.4178466796875,
+      "learning_rate": 3.336326643578172e-05,
+      "loss": 55.3556,
+      "step": 113070
+    },
+    {
+      "epoch": 0.4568575087771749,
+      "grad_norm": 594.2887573242188,
+      "learning_rate": 3.3359976810516164e-05,
+      "loss": 52.3926,
+      "step": 113080
+    },
+    {
+      "epoch": 0.45689791004254254,
+      "grad_norm": 731.5540771484375,
+      "learning_rate": 3.335668702226856e-05,
+      "loss": 107.6374,
+      "step": 113090
+    },
+    {
+      "epoch": 0.4569383113079102,
+      "grad_norm": 608.2393798828125,
+      "learning_rate": 3.3353397071103046e-05,
+      "loss": 57.6193,
+      "step": 113100
+    },
+    {
+      "epoch": 0.45697871257327777,
+      "grad_norm": 617.5512084960938,
+      "learning_rate": 3.3350106957083744e-05,
+      "loss": 72.52,
+      "step": 113110
+    },
+    {
+      "epoch": 0.4570191138386454,
+      "grad_norm": 658.0664672851562,
+      "learning_rate": 3.334681668027481e-05,
+      "loss": 94.5809,
+      "step": 113120
+    },
+    {
+      "epoch": 0.45705951510401305,
+      "grad_norm": 543.9719848632812,
+      "learning_rate": 3.334352624074039e-05,
+      "loss": 41.8438,
+      "step": 113130
+    },
+    {
+      "epoch": 0.4570999163693807,
+      "grad_norm": 696.774658203125,
+      "learning_rate": 3.334023563854463e-05,
+      "loss": 76.3532,
+      "step": 113140
+    },
+    {
+      "epoch": 0.4571403176347483,
+      "grad_norm": 2398.177734375,
+      "learning_rate": 3.333694487375168e-05,
+      "loss": 38.9571,
+      "step": 113150
+    },
+    {
+      "epoch": 0.45718071890011597,
+      "grad_norm": 455.6088562011719,
+      "learning_rate": 3.33336539464257e-05,
+      "loss": 32.256,
+      "step": 113160
+    },
+    {
+      "epoch": 0.4572211201654836,
+      "grad_norm": 1092.000732421875,
+      "learning_rate": 3.3330362856630845e-05,
+      "loss": 59.1251,
+      "step": 113170
+    },
+    {
+      "epoch": 0.4572615214308512,
+      "grad_norm": 1621.383544921875,
+      "learning_rate": 3.3327071604431275e-05,
+      "loss": 43.2729,
+      "step": 113180
+    },
+    {
+      "epoch": 0.45730192269621883,
+      "grad_norm": 680.7568969726562,
+      "learning_rate": 3.3323780189891166e-05,
+      "loss": 49.1451,
+      "step": 113190
+    },
+    {
+      "epoch": 0.45734232396158647,
+      "grad_norm": 280.8765563964844,
+      "learning_rate": 3.332048861307467e-05,
+      "loss": 54.7998,
+      "step": 113200
+    },
+    {
+      "epoch": 0.4573827252269541,
+      "grad_norm": 457.403564453125,
+      "learning_rate": 3.331719687404597e-05,
+      "loss": 91.7487,
+      "step": 113210
+    },
+    {
+      "epoch": 0.45742312649232175,
+      "grad_norm": 777.8742065429688,
+      "learning_rate": 3.331390497286922e-05,
+      "loss": 44.5807,
+      "step": 113220
+    },
+    {
+      "epoch": 0.4574635277576894,
+      "grad_norm": 1098.077880859375,
+      "learning_rate": 3.331061290960863e-05,
+      "loss": 90.1487,
+      "step": 113230
+    },
+    {
+      "epoch": 0.457503929023057,
+      "grad_norm": 882.6914672851562,
+      "learning_rate": 3.3307320684328354e-05,
+      "loss": 68.7481,
+      "step": 113240
+    },
+    {
+      "epoch": 0.4575443302884246,
+      "grad_norm": 1128.127197265625,
+      "learning_rate": 3.330402829709258e-05,
+      "loss": 65.0538,
+      "step": 113250
+    },
+    {
+      "epoch": 0.45758473155379226,
+      "grad_norm": 937.3806762695312,
+      "learning_rate": 3.3300735747965505e-05,
+      "loss": 57.1227,
+      "step": 113260
+    },
+    {
+      "epoch": 0.4576251328191599,
+      "grad_norm": 911.514404296875,
+      "learning_rate": 3.329744303701132e-05,
+      "loss": 85.5697,
+      "step": 113270
+    },
+    {
+      "epoch": 0.45766553408452754,
+      "grad_norm": 532.0978393554688,
+      "learning_rate": 3.3294150164294204e-05,
+      "loss": 41.7875,
+      "step": 113280
+    },
+    {
+      "epoch": 0.4577059353498952,
+      "grad_norm": 2276.76904296875,
+      "learning_rate": 3.329085712987836e-05,
+      "loss": 58.2938,
+      "step": 113290
+    },
+    {
+      "epoch": 0.4577463366152628,
+      "grad_norm": 633.0645751953125,
+      "learning_rate": 3.3287563933827995e-05,
+      "loss": 45.8421,
+      "step": 113300
+    },
+    {
+      "epoch": 0.4577867378806304,
+      "grad_norm": 884.1083984375,
+      "learning_rate": 3.328427057620729e-05,
+      "loss": 69.57,
+      "step": 113310
+    },
+    {
+      "epoch": 0.45782713914599804,
+      "grad_norm": 832.331298828125,
+      "learning_rate": 3.328097705708047e-05,
+      "loss": 36.6988,
+      "step": 113320
+    },
+    {
+      "epoch": 0.4578675404113657,
+      "grad_norm": 1298.6260986328125,
+      "learning_rate": 3.3277683376511744e-05,
+      "loss": 67.1316,
+      "step": 113330
+    },
+    {
+      "epoch": 0.4579079416767333,
+      "grad_norm": 690.5106811523438,
+      "learning_rate": 3.327438953456532e-05,
+      "loss": 43.1363,
+      "step": 113340
+    },
+    {
+      "epoch": 0.45794834294210096,
+      "grad_norm": 638.0509643554688,
+      "learning_rate": 3.327109553130541e-05,
+      "loss": 64.9828,
+      "step": 113350
+    },
+    {
+      "epoch": 0.4579887442074686,
+      "grad_norm": 712.4091186523438,
+      "learning_rate": 3.326780136679623e-05,
+      "loss": 51.8144,
+      "step": 113360
+    },
+    {
+      "epoch": 0.4580291454728362,
+      "grad_norm": 768.2869873046875,
+      "learning_rate": 3.326450704110201e-05,
+      "loss": 43.4623,
+      "step": 113370
+    },
+    {
+      "epoch": 0.4580695467382038,
+      "grad_norm": 667.932373046875,
+      "learning_rate": 3.3261212554286975e-05,
+      "loss": 60.1954,
+      "step": 113380
+    },
+    {
+      "epoch": 0.45810994800357147,
+      "grad_norm": 797.756591796875,
+      "learning_rate": 3.3257917906415336e-05,
+      "loss": 64.3577,
+      "step": 113390
+    },
+    {
+      "epoch": 0.4581503492689391,
+      "grad_norm": 417.9449462890625,
+      "learning_rate": 3.325462309755134e-05,
+      "loss": 52.2687,
+      "step": 113400
+    },
+    {
+      "epoch": 0.45819075053430675,
+      "grad_norm": 131.72840881347656,
+      "learning_rate": 3.325132812775922e-05,
+      "loss": 53.6158,
+      "step": 113410
+    },
+    {
+      "epoch": 0.4582311517996744,
+      "grad_norm": 1277.050537109375,
+      "learning_rate": 3.324803299710321e-05,
+      "loss": 79.7133,
+      "step": 113420
+    },
+    {
+      "epoch": 0.45827155306504197,
+      "grad_norm": 247.889404296875,
+      "learning_rate": 3.3244737705647554e-05,
+      "loss": 79.97,
+      "step": 113430
+    },
+    {
+      "epoch": 0.4583119543304096,
+      "grad_norm": 1504.002685546875,
+      "learning_rate": 3.324144225345649e-05,
+      "loss": 57.3291,
+      "step": 113440
+    },
+    {
+      "epoch": 0.45835235559577725,
+      "grad_norm": 3024.2744140625,
+      "learning_rate": 3.3238146640594256e-05,
+      "loss": 55.5882,
+      "step": 113450
+    },
+    {
+      "epoch": 0.4583927568611449,
+      "grad_norm": 1827.9388427734375,
+      "learning_rate": 3.323485086712513e-05,
+      "loss": 68.6746,
+      "step": 113460
+    },
+    {
+      "epoch": 0.45843315812651253,
+      "grad_norm": 1997.11279296875,
+      "learning_rate": 3.323155493311334e-05,
+      "loss": 60.3449,
+      "step": 113470
+    },
+    {
+      "epoch": 0.45847355939188017,
+      "grad_norm": 1310.3616943359375,
+      "learning_rate": 3.322825883862314e-05,
+      "loss": 58.0651,
+      "step": 113480
+    },
+    {
+      "epoch": 0.4585139606572478,
+      "grad_norm": 542.6761474609375,
+      "learning_rate": 3.32249625837188e-05,
+      "loss": 45.5163,
+      "step": 113490
+    },
+    {
+      "epoch": 0.4585543619226154,
+      "grad_norm": 109.23764038085938,
+      "learning_rate": 3.322166616846458e-05,
+      "loss": 46.8832,
+      "step": 113500
+    },
+    {
+      "epoch": 0.45859476318798303,
+      "grad_norm": 770.7872314453125,
+      "learning_rate": 3.321836959292475e-05,
+      "loss": 54.6829,
+      "step": 113510
+    },
+    {
+      "epoch": 0.4586351644533507,
+      "grad_norm": 1391.15185546875,
+      "learning_rate": 3.321507285716357e-05,
+      "loss": 76.5766,
+      "step": 113520
+    },
+    {
+      "epoch": 0.4586755657187183,
+      "grad_norm": 1156.2738037109375,
+      "learning_rate": 3.321177596124532e-05,
+      "loss": 44.4677,
+      "step": 113530
+    },
+    {
+      "epoch": 0.45871596698408595,
+      "grad_norm": 882.244140625,
+      "learning_rate": 3.3208478905234274e-05,
+      "loss": 64.2383,
+      "step": 113540
+    },
+    {
+      "epoch": 0.4587563682494536,
+      "grad_norm": 1060.12841796875,
+      "learning_rate": 3.32051816891947e-05,
+      "loss": 49.1175,
+      "step": 113550
+    },
+    {
+      "epoch": 0.4587967695148212,
+      "grad_norm": 1234.741455078125,
+      "learning_rate": 3.320188431319088e-05,
+      "loss": 41.2987,
+      "step": 113560
+    },
+    {
+      "epoch": 0.4588371707801888,
+      "grad_norm": 437.5050048828125,
+      "learning_rate": 3.31985867772871e-05,
+      "loss": 63.9297,
+      "step": 113570
+    },
+    {
+      "epoch": 0.45887757204555646,
+      "grad_norm": 839.0947875976562,
+      "learning_rate": 3.319528908154766e-05,
+      "loss": 55.2399,
+      "step": 113580
+    },
+    {
+      "epoch": 0.4589179733109241,
+      "grad_norm": 1059.20654296875,
+      "learning_rate": 3.319199122603683e-05,
+      "loss": 35.0855,
+      "step": 113590
+    },
+    {
+      "epoch": 0.45895837457629174,
+      "grad_norm": 1006.2760009765625,
+      "learning_rate": 3.318869321081892e-05,
+      "loss": 75.6239,
+      "step": 113600
+    },
+    {
+      "epoch": 0.4589987758416594,
+      "grad_norm": 1427.097412109375,
+      "learning_rate": 3.3185395035958224e-05,
+      "loss": 53.2945,
+      "step": 113610
+    },
+    {
+      "epoch": 0.459039177107027,
+      "grad_norm": 960.8377075195312,
+      "learning_rate": 3.318209670151904e-05,
+      "loss": 74.3849,
+      "step": 113620
+    },
+    {
+      "epoch": 0.4590795783723946,
+      "grad_norm": 780.0336303710938,
+      "learning_rate": 3.317879820756566e-05,
+      "loss": 59.6619,
+      "step": 113630
+    },
+    {
+      "epoch": 0.45911997963776224,
+      "grad_norm": 913.7399291992188,
+      "learning_rate": 3.31754995541624e-05,
+      "loss": 54.3696,
+      "step": 113640
+    },
+    {
+      "epoch": 0.4591603809031299,
+      "grad_norm": 884.738037109375,
+      "learning_rate": 3.3172200741373563e-05,
+      "loss": 87.1222,
+      "step": 113650
+    },
+    {
+      "epoch": 0.4592007821684975,
+      "grad_norm": 495.6034851074219,
+      "learning_rate": 3.3168901769263474e-05,
+      "loss": 52.4804,
+      "step": 113660
+    },
+    {
+      "epoch": 0.45924118343386516,
+      "grad_norm": 1088.6365966796875,
+      "learning_rate": 3.316560263789643e-05,
+      "loss": 46.7347,
+      "step": 113670
+    },
+    {
+      "epoch": 0.4592815846992328,
+      "grad_norm": 724.1251831054688,
+      "learning_rate": 3.3162303347336764e-05,
+      "loss": 53.7025,
+      "step": 113680
+    },
+    {
+      "epoch": 0.4593219859646004,
+      "grad_norm": 982.661376953125,
+      "learning_rate": 3.315900389764879e-05,
+      "loss": 47.1129,
+      "step": 113690
+    },
+    {
+      "epoch": 0.459362387229968,
+      "grad_norm": 662.5863037109375,
+      "learning_rate": 3.315570428889684e-05,
+      "loss": 35.4883,
+      "step": 113700
+    },
+    {
+      "epoch": 0.45940278849533567,
+      "grad_norm": 1490.281005859375,
+      "learning_rate": 3.315240452114523e-05,
+      "loss": 52.5399,
+      "step": 113710
+    },
+    {
+      "epoch": 0.4594431897607033,
+      "grad_norm": 1237.041748046875,
+      "learning_rate": 3.31491045944583e-05,
+      "loss": 59.784,
+      "step": 113720
+    },
+    {
+      "epoch": 0.45948359102607095,
+      "grad_norm": 700.3053588867188,
+      "learning_rate": 3.314580450890038e-05,
+      "loss": 70.0428,
+      "step": 113730
+    },
+    {
+      "epoch": 0.4595239922914386,
+      "grad_norm": 851.7976684570312,
+      "learning_rate": 3.3142504264535804e-05,
+      "loss": 50.4964,
+      "step": 113740
+    },
+    {
+      "epoch": 0.45956439355680617,
+      "grad_norm": 983.1741943359375,
+      "learning_rate": 3.313920386142892e-05,
+      "loss": 59.0461,
+      "step": 113750
+    },
+    {
+      "epoch": 0.4596047948221738,
+      "grad_norm": 1348.2669677734375,
+      "learning_rate": 3.313590329964406e-05,
+      "loss": 59.1032,
+      "step": 113760
+    },
+    {
+      "epoch": 0.45964519608754145,
+      "grad_norm": 487.41680908203125,
+      "learning_rate": 3.313260257924558e-05,
+      "loss": 87.3492,
+      "step": 113770
+    },
+    {
+      "epoch": 0.4596855973529091,
+      "grad_norm": 345.0389099121094,
+      "learning_rate": 3.312930170029783e-05,
+      "loss": 52.9883,
+      "step": 113780
+    },
+    {
+      "epoch": 0.45972599861827673,
+      "grad_norm": 438.7192077636719,
+      "learning_rate": 3.3126000662865156e-05,
+      "loss": 49.1873,
+      "step": 113790
+    },
+    {
+      "epoch": 0.45976639988364437,
+      "grad_norm": 493.15130615234375,
+      "learning_rate": 3.312269946701191e-05,
+      "loss": 50.4528,
+      "step": 113800
+    },
+    {
+      "epoch": 0.459806801149012,
+      "grad_norm": 1520.73974609375,
+      "learning_rate": 3.311939811280246e-05,
+      "loss": 42.8416,
+      "step": 113810
+    },
+    {
+      "epoch": 0.4598472024143796,
+      "grad_norm": 146.88551330566406,
+      "learning_rate": 3.311609660030117e-05,
+      "loss": 52.2549,
+      "step": 113820
+    },
+    {
+      "epoch": 0.45988760367974724,
+      "grad_norm": 610.1900634765625,
+      "learning_rate": 3.311279492957239e-05,
+      "loss": 81.4292,
+      "step": 113830
+    },
+    {
+      "epoch": 0.4599280049451149,
+      "grad_norm": 478.4176330566406,
+      "learning_rate": 3.31094931006805e-05,
+      "loss": 59.0037,
+      "step": 113840
+    },
+    {
+      "epoch": 0.4599684062104825,
+      "grad_norm": 698.8412475585938,
+      "learning_rate": 3.310619111368986e-05,
+      "loss": 60.8137,
+      "step": 113850
+    },
+    {
+      "epoch": 0.46000880747585016,
+      "grad_norm": 719.0433349609375,
+      "learning_rate": 3.310288896866486e-05,
+      "loss": 40.1776,
+      "step": 113860
+    },
+    {
+      "epoch": 0.4600492087412178,
+      "grad_norm": 807.5015258789062,
+      "learning_rate": 3.309958666566986e-05,
+      "loss": 87.1863,
+      "step": 113870
+    },
+    {
+      "epoch": 0.4600896100065854,
+      "grad_norm": 773.2158813476562,
+      "learning_rate": 3.309628420476926e-05,
+      "loss": 65.442,
+      "step": 113880
+    },
+    {
+      "epoch": 0.460130011271953,
+      "grad_norm": 1120.7901611328125,
+      "learning_rate": 3.309298158602742e-05,
+      "loss": 46.1566,
+      "step": 113890
+    },
+    {
+      "epoch": 0.46017041253732066,
+      "grad_norm": 2974.2939453125,
+      "learning_rate": 3.308967880950874e-05,
+      "loss": 86.9134,
+      "step": 113900
+    },
+    {
+      "epoch": 0.4602108138026883,
+      "grad_norm": 1336.4696044921875,
+      "learning_rate": 3.308637587527761e-05,
+      "loss": 79.6506,
+      "step": 113910
+    },
+    {
+      "epoch": 0.46025121506805594,
+      "grad_norm": 1928.9310302734375,
+      "learning_rate": 3.3083072783398416e-05,
+      "loss": 96.6911,
+      "step": 113920
+    },
+    {
+      "epoch": 0.4602916163334236,
+      "grad_norm": 538.5294799804688,
+      "learning_rate": 3.3079769533935556e-05,
+      "loss": 49.8037,
+      "step": 113930
+    },
+    {
+      "epoch": 0.4603320175987912,
+      "grad_norm": 809.7352294921875,
+      "learning_rate": 3.307646612695343e-05,
+      "loss": 61.5456,
+      "step": 113940
+    },
+    {
+      "epoch": 0.4603724188641588,
+      "grad_norm": 550.5420532226562,
+      "learning_rate": 3.307316256251644e-05,
+      "loss": 49.7282,
+      "step": 113950
+    },
+    {
+      "epoch": 0.46041282012952645,
+      "grad_norm": 1422.148193359375,
+      "learning_rate": 3.3069858840688994e-05,
+      "loss": 71.9855,
+      "step": 113960
+    },
+    {
+      "epoch": 0.4604532213948941,
+      "grad_norm": 426.9280090332031,
+      "learning_rate": 3.3066554961535485e-05,
+      "loss": 46.9851,
+      "step": 113970
+    },
+    {
+      "epoch": 0.4604936226602617,
+      "grad_norm": 844.1886596679688,
+      "learning_rate": 3.3063250925120334e-05,
+      "loss": 56.5634,
+      "step": 113980
+    },
+    {
+      "epoch": 0.46053402392562937,
+      "grad_norm": 590.6248779296875,
+      "learning_rate": 3.305994673150797e-05,
+      "loss": 58.149,
+      "step": 113990
+    },
+    {
+      "epoch": 0.460574425190997,
+      "grad_norm": 722.6728515625,
+      "learning_rate": 3.305664238076278e-05,
+      "loss": 48.5719,
+      "step": 114000
+    },
+    {
+      "epoch": 0.4606148264563646,
+      "grad_norm": 761.7833251953125,
+      "learning_rate": 3.30533378729492e-05,
+      "loss": 67.1143,
+      "step": 114010
+    },
+    {
+      "epoch": 0.46065522772173223,
+      "grad_norm": 258.24639892578125,
+      "learning_rate": 3.3050033208131656e-05,
+      "loss": 50.417,
+      "step": 114020
+    },
+    {
+      "epoch": 0.46069562898709987,
+      "grad_norm": 391.4142150878906,
+      "learning_rate": 3.304672838637457e-05,
+      "loss": 65.8437,
+      "step": 114030
+    },
+    {
+      "epoch": 0.4607360302524675,
+      "grad_norm": 432.7169189453125,
+      "learning_rate": 3.3043423407742375e-05,
+      "loss": 60.9631,
+      "step": 114040
+    },
+    {
+      "epoch": 0.46077643151783515,
+      "grad_norm": 893.4138793945312,
+      "learning_rate": 3.3040118272299495e-05,
+      "loss": 131.9281,
+      "step": 114050
+    },
+    {
+      "epoch": 0.4608168327832028,
+      "grad_norm": 1183.3709716796875,
+      "learning_rate": 3.303681298011037e-05,
+      "loss": 74.6201,
+      "step": 114060
+    },
+    {
+      "epoch": 0.4608572340485704,
+      "grad_norm": 327.8617248535156,
+      "learning_rate": 3.303350753123944e-05,
+      "loss": 38.2665,
+      "step": 114070
+    },
+    {
+      "epoch": 0.460897635313938,
+      "grad_norm": 642.1771850585938,
+      "learning_rate": 3.3030201925751145e-05,
+      "loss": 50.6443,
+      "step": 114080
+    },
+    {
+      "epoch": 0.46093803657930565,
+      "grad_norm": 1546.3465576171875,
+      "learning_rate": 3.302689616370993e-05,
+      "loss": 37.8828,
+      "step": 114090
+    },
+    {
+      "epoch": 0.4609784378446733,
+      "grad_norm": 634.73876953125,
+      "learning_rate": 3.302359024518024e-05,
+      "loss": 29.7485,
+      "step": 114100
+    },
+    {
+      "epoch": 0.46101883911004093,
+      "grad_norm": 658.0567626953125,
+      "learning_rate": 3.302028417022653e-05,
+      "loss": 45.2375,
+      "step": 114110
+    },
+    {
+      "epoch": 0.4610592403754086,
+      "grad_norm": 633.6270751953125,
+      "learning_rate": 3.301697793891324e-05,
+      "loss": 59.6125,
+      "step": 114120
+    },
+    {
+      "epoch": 0.4610996416407762,
+      "grad_norm": 1438.4256591796875,
+      "learning_rate": 3.301367155130485e-05,
+      "loss": 89.7295,
+      "step": 114130
+    },
+    {
+      "epoch": 0.4611400429061438,
+      "grad_norm": 1383.4329833984375,
+      "learning_rate": 3.3010365007465805e-05,
+      "loss": 48.6195,
+      "step": 114140
+    },
+    {
+      "epoch": 0.46118044417151144,
+      "grad_norm": 530.7550659179688,
+      "learning_rate": 3.300705830746057e-05,
+      "loss": 46.5861,
+      "step": 114150
+    },
+    {
+      "epoch": 0.4612208454368791,
+      "grad_norm": 633.077880859375,
+      "learning_rate": 3.300375145135361e-05,
+      "loss": 60.9828,
+      "step": 114160
+    },
+    {
+      "epoch": 0.4612612467022467,
+      "grad_norm": 376.91864013671875,
+      "learning_rate": 3.3000444439209396e-05,
+      "loss": 62.8593,
+      "step": 114170
+    },
+    {
+      "epoch": 0.46130164796761436,
+      "grad_norm": 721.732666015625,
+      "learning_rate": 3.299713727109239e-05,
+      "loss": 84.59,
+      "step": 114180
+    },
+    {
+      "epoch": 0.461342049232982,
+      "grad_norm": 473.7373352050781,
+      "learning_rate": 3.299382994706709e-05,
+      "loss": 94.0822,
+      "step": 114190
+    },
+    {
+      "epoch": 0.4613824504983496,
+      "grad_norm": 1528.2044677734375,
+      "learning_rate": 3.299052246719795e-05,
+      "loss": 81.7651,
+      "step": 114200
+    },
+    {
+      "epoch": 0.4614228517637172,
+      "grad_norm": 890.8794555664062,
+      "learning_rate": 3.298721483154946e-05,
+      "loss": 40.4416,
+      "step": 114210
+    },
+    {
+      "epoch": 0.46146325302908486,
+      "grad_norm": 541.0910034179688,
+      "learning_rate": 3.298390704018611e-05,
+      "loss": 43.575,
+      "step": 114220
+    },
+    {
+      "epoch": 0.4615036542944525,
+      "grad_norm": 915.8447875976562,
+      "learning_rate": 3.298059909317239e-05,
+      "loss": 70.035,
+      "step": 114230
+    },
+    {
+      "epoch": 0.46154405555982014,
+      "grad_norm": 1335.07275390625,
+      "learning_rate": 3.297729099057277e-05,
+      "loss": 58.3776,
+      "step": 114240
+    },
+    {
+      "epoch": 0.4615844568251878,
+      "grad_norm": 1571.07373046875,
+      "learning_rate": 3.2973982732451755e-05,
+      "loss": 72.0075,
+      "step": 114250
+    },
+    {
+      "epoch": 0.4616248580905554,
+      "grad_norm": 1263.757080078125,
+      "learning_rate": 3.297067431887384e-05,
+      "loss": 53.4101,
+      "step": 114260
+    },
+    {
+      "epoch": 0.461665259355923,
+      "grad_norm": 916.6146850585938,
+      "learning_rate": 3.296736574990353e-05,
+      "loss": 78.4652,
+      "step": 114270
+    },
+    {
+      "epoch": 0.46170566062129065,
+      "grad_norm": 365.46142578125,
+      "learning_rate": 3.296405702560532e-05,
+      "loss": 80.6376,
+      "step": 114280
+    },
+    {
+      "epoch": 0.4617460618866583,
+      "grad_norm": 944.3591918945312,
+      "learning_rate": 3.2960748146043716e-05,
+      "loss": 90.1059,
+      "step": 114290
+    },
+    {
+      "epoch": 0.46178646315202593,
+      "grad_norm": 679.8099365234375,
+      "learning_rate": 3.295743911128324e-05,
+      "loss": 48.7843,
+      "step": 114300
+    },
+    {
+      "epoch": 0.46182686441739357,
+      "grad_norm": 1261.843017578125,
+      "learning_rate": 3.295412992138838e-05,
+      "loss": 78.7052,
+      "step": 114310
+    },
+    {
+      "epoch": 0.4618672656827612,
+      "grad_norm": 481.5779724121094,
+      "learning_rate": 3.295082057642367e-05,
+      "loss": 41.2433,
+      "step": 114320
+    },
+    {
+      "epoch": 0.4619076669481288,
+      "grad_norm": 903.3895874023438,
+      "learning_rate": 3.294751107645361e-05,
+      "loss": 48.583,
+      "step": 114330
+    },
+    {
+      "epoch": 0.46194806821349643,
+      "grad_norm": 730.4287109375,
+      "learning_rate": 3.294420142154274e-05,
+      "loss": 78.3227,
+      "step": 114340
+    },
+    {
+      "epoch": 0.46198846947886407,
+      "grad_norm": 563.7612915039062,
+      "learning_rate": 3.2940891611755564e-05,
+      "loss": 53.6732,
+      "step": 114350
+    },
+    {
+      "epoch": 0.4620288707442317,
+      "grad_norm": 1336.9444580078125,
+      "learning_rate": 3.293758164715663e-05,
+      "loss": 55.2327,
+      "step": 114360
+    },
+    {
+      "epoch": 0.46206927200959935,
+      "grad_norm": 1316.6781005859375,
+      "learning_rate": 3.293427152781044e-05,
+      "loss": 39.9807,
+      "step": 114370
+    },
+    {
+      "epoch": 0.462109673274967,
+      "grad_norm": 1014.697265625,
+      "learning_rate": 3.2930961253781554e-05,
+      "loss": 81.932,
+      "step": 114380
+    },
+    {
+      "epoch": 0.4621500745403346,
+      "grad_norm": 2144.41064453125,
+      "learning_rate": 3.292765082513449e-05,
+      "loss": 80.4216,
+      "step": 114390
+    },
+    {
+      "epoch": 0.4621904758057022,
+      "grad_norm": 509.5421447753906,
+      "learning_rate": 3.29243402419338e-05,
+      "loss": 40.3778,
+      "step": 114400
+    },
+    {
+      "epoch": 0.46223087707106986,
+      "grad_norm": 432.00933837890625,
+      "learning_rate": 3.2921029504244004e-05,
+      "loss": 76.2071,
+      "step": 114410
+    },
+    {
+      "epoch": 0.4622712783364375,
+      "grad_norm": 722.176025390625,
+      "learning_rate": 3.2917718612129665e-05,
+      "loss": 41.5058,
+      "step": 114420
+    },
+    {
+      "epoch": 0.46231167960180514,
+      "grad_norm": 725.9029541015625,
+      "learning_rate": 3.291440756565533e-05,
+      "loss": 53.8693,
+      "step": 114430
+    },
+    {
+      "epoch": 0.4623520808671728,
+      "grad_norm": 281.3984375,
+      "learning_rate": 3.2911096364885544e-05,
+      "loss": 80.6305,
+      "step": 114440
+    },
+    {
+      "epoch": 0.4623924821325404,
+      "grad_norm": 811.3935546875,
+      "learning_rate": 3.290778500988485e-05,
+      "loss": 55.3654,
+      "step": 114450
+    },
+    {
+      "epoch": 0.462432883397908,
+      "grad_norm": 472.8229064941406,
+      "learning_rate": 3.2904473500717824e-05,
+      "loss": 59.1575,
+      "step": 114460
+    },
+    {
+      "epoch": 0.46247328466327564,
+      "grad_norm": 713.5052490234375,
+      "learning_rate": 3.290116183744902e-05,
+      "loss": 50.5053,
+      "step": 114470
+    },
+    {
+      "epoch": 0.4625136859286433,
+      "grad_norm": 1182.7664794921875,
+      "learning_rate": 3.2897850020143005e-05,
+      "loss": 59.1632,
+      "step": 114480
+    },
+    {
+      "epoch": 0.4625540871940109,
+      "grad_norm": 845.945556640625,
+      "learning_rate": 3.289453804886433e-05,
+      "loss": 85.3159,
+      "step": 114490
+    },
+    {
+      "epoch": 0.46259448845937856,
+      "grad_norm": 406.4460754394531,
+      "learning_rate": 3.289122592367757e-05,
+      "loss": 39.6031,
+      "step": 114500
+    },
+    {
+      "epoch": 0.4626348897247462,
+      "grad_norm": 883.4130249023438,
+      "learning_rate": 3.288791364464729e-05,
+      "loss": 44.7248,
+      "step": 114510
+    },
+    {
+      "epoch": 0.4626752909901138,
+      "grad_norm": 447.2448425292969,
+      "learning_rate": 3.2884601211838085e-05,
+      "loss": 62.0368,
+      "step": 114520
+    },
+    {
+      "epoch": 0.4627156922554814,
+      "grad_norm": 640.0363159179688,
+      "learning_rate": 3.288128862531452e-05,
+      "loss": 56.4858,
+      "step": 114530
+    },
+    {
+      "epoch": 0.46275609352084907,
+      "grad_norm": 891.1563720703125,
+      "learning_rate": 3.287797588514117e-05,
+      "loss": 70.3198,
+      "step": 114540
+    },
+    {
+      "epoch": 0.4627964947862167,
+      "grad_norm": 628.9306030273438,
+      "learning_rate": 3.287466299138262e-05,
+      "loss": 53.9219,
+      "step": 114550
+    },
+    {
+      "epoch": 0.46283689605158435,
+      "grad_norm": 382.36456298828125,
+      "learning_rate": 3.287134994410347e-05,
+      "loss": 42.8895,
+      "step": 114560
+    },
+    {
+      "epoch": 0.462877297316952,
+      "grad_norm": 1126.9713134765625,
+      "learning_rate": 3.28680367433683e-05,
+      "loss": 38.392,
+      "step": 114570
+    },
+    {
+      "epoch": 0.4629176985823196,
+      "grad_norm": 2824.01025390625,
+      "learning_rate": 3.28647233892417e-05,
+      "loss": 77.3813,
+      "step": 114580
+    },
+    {
+      "epoch": 0.4629580998476872,
+      "grad_norm": 413.0010986328125,
+      "learning_rate": 3.286140988178826e-05,
+      "loss": 50.3714,
+      "step": 114590
+    },
+    {
+      "epoch": 0.46299850111305485,
+      "grad_norm": 952.1806030273438,
+      "learning_rate": 3.28580962210726e-05,
+      "loss": 55.7656,
+      "step": 114600
+    },
+    {
+      "epoch": 0.4630389023784225,
+      "grad_norm": 491.023193359375,
+      "learning_rate": 3.2854782407159305e-05,
+      "loss": 47.8107,
+      "step": 114610
+    },
+    {
+      "epoch": 0.46307930364379013,
+      "grad_norm": 413.3840026855469,
+      "learning_rate": 3.285146844011298e-05,
+      "loss": 57.6984,
+      "step": 114620
+    },
+    {
+      "epoch": 0.46311970490915777,
+      "grad_norm": 593.4551391601562,
+      "learning_rate": 3.2848154319998235e-05,
+      "loss": 47.3308,
+      "step": 114630
+    },
+    {
+      "epoch": 0.4631601061745254,
+      "grad_norm": 543.7266845703125,
+      "learning_rate": 3.2844840046879686e-05,
+      "loss": 46.1537,
+      "step": 114640
+    },
+    {
+      "epoch": 0.463200507439893,
+      "grad_norm": 876.9476318359375,
+      "learning_rate": 3.2841525620821945e-05,
+      "loss": 39.367,
+      "step": 114650
+    },
+    {
+      "epoch": 0.46324090870526063,
+      "grad_norm": 470.560791015625,
+      "learning_rate": 3.2838211041889625e-05,
+      "loss": 56.22,
+      "step": 114660
+    },
+    {
+      "epoch": 0.4632813099706283,
+      "grad_norm": 692.4049072265625,
+      "learning_rate": 3.2834896310147336e-05,
+      "loss": 46.9886,
+      "step": 114670
+    },
+    {
+      "epoch": 0.4633217112359959,
+      "grad_norm": 1012.37890625,
+      "learning_rate": 3.283158142565971e-05,
+      "loss": 77.0895,
+      "step": 114680
+    },
+    {
+      "epoch": 0.46336211250136355,
+      "grad_norm": 1028.7650146484375,
+      "learning_rate": 3.282826638849138e-05,
+      "loss": 60.6332,
+      "step": 114690
+    },
+    {
+      "epoch": 0.4634025137667312,
+      "grad_norm": 1001.2539672851562,
+      "learning_rate": 3.2824951198706954e-05,
+      "loss": 55.1329,
+      "step": 114700
+    },
+    {
+      "epoch": 0.4634429150320988,
+      "grad_norm": 707.6488647460938,
+      "learning_rate": 3.2821635856371086e-05,
+      "loss": 68.9933,
+      "step": 114710
+    },
+    {
+      "epoch": 0.4634833162974664,
+      "grad_norm": 640.0159912109375,
+      "learning_rate": 3.28183203615484e-05,
+      "loss": 56.8567,
+      "step": 114720
+    },
+    {
+      "epoch": 0.46352371756283406,
+      "grad_norm": 1112.1865234375,
+      "learning_rate": 3.281500471430353e-05,
+      "loss": 63.7654,
+      "step": 114730
+    },
+    {
+      "epoch": 0.4635641188282017,
+      "grad_norm": 871.5289306640625,
+      "learning_rate": 3.281168891470112e-05,
+      "loss": 64.3242,
+      "step": 114740
+    },
+    {
+      "epoch": 0.46360452009356934,
+      "grad_norm": 232.17013549804688,
+      "learning_rate": 3.2808372962805816e-05,
+      "loss": 51.8217,
+      "step": 114750
+    },
+    {
+      "epoch": 0.463644921358937,
+      "grad_norm": 653.4492797851562,
+      "learning_rate": 3.280505685868226e-05,
+      "loss": 32.7491,
+      "step": 114760
+    },
+    {
+      "epoch": 0.4636853226243046,
+      "grad_norm": 806.7274780273438,
+      "learning_rate": 3.2801740602395105e-05,
+      "loss": 60.581,
+      "step": 114770
+    },
+    {
+      "epoch": 0.4637257238896722,
+      "grad_norm": 698.8681640625,
+      "learning_rate": 3.279842419400899e-05,
+      "loss": 81.199,
+      "step": 114780
+    },
+    {
+      "epoch": 0.46376612515503984,
+      "grad_norm": 1318.1488037109375,
+      "learning_rate": 3.2795107633588586e-05,
+      "loss": 61.5727,
+      "step": 114790
+    },
+    {
+      "epoch": 0.4638065264204075,
+      "grad_norm": 2012.5032958984375,
+      "learning_rate": 3.279179092119855e-05,
+      "loss": 63.0752,
+      "step": 114800
+    },
+    {
+      "epoch": 0.4638469276857751,
+      "grad_norm": 946.7776489257812,
+      "learning_rate": 3.278847405690353e-05,
+      "loss": 104.6616,
+      "step": 114810
+    },
+    {
+      "epoch": 0.46388732895114276,
+      "grad_norm": 332.1543273925781,
+      "learning_rate": 3.278515704076821e-05,
+      "loss": 57.2958,
+      "step": 114820
+    },
+    {
+      "epoch": 0.4639277302165104,
+      "grad_norm": 628.6360473632812,
+      "learning_rate": 3.278183987285724e-05,
+      "loss": 49.5415,
+      "step": 114830
+    },
+    {
+      "epoch": 0.463968131481878,
+      "grad_norm": 407.3933410644531,
+      "learning_rate": 3.277852255323529e-05,
+      "loss": 45.7567,
+      "step": 114840
+    },
+    {
+      "epoch": 0.46400853274724563,
+      "grad_norm": 1091.437255859375,
+      "learning_rate": 3.277520508196705e-05,
+      "loss": 76.9394,
+      "step": 114850
+    },
+    {
+      "epoch": 0.46404893401261327,
+      "grad_norm": 540.4053344726562,
+      "learning_rate": 3.277188745911717e-05,
+      "loss": 42.3653,
+      "step": 114860
+    },
+    {
+      "epoch": 0.4640893352779809,
+      "grad_norm": 1648.4383544921875,
+      "learning_rate": 3.276856968475035e-05,
+      "loss": 78.4631,
+      "step": 114870
+    },
+    {
+      "epoch": 0.46412973654334855,
+      "grad_norm": 469.8823547363281,
+      "learning_rate": 3.276525175893126e-05,
+      "loss": 59.8111,
+      "step": 114880
+    },
+    {
+      "epoch": 0.4641701378087162,
+      "grad_norm": 925.3463745117188,
+      "learning_rate": 3.27619336817246e-05,
+      "loss": 49.3538,
+      "step": 114890
+    },
+    {
+      "epoch": 0.46421053907408383,
+      "grad_norm": 1113.8001708984375,
+      "learning_rate": 3.2758615453195034e-05,
+      "loss": 64.5263,
+      "step": 114900
+    },
+    {
+      "epoch": 0.4642509403394514,
+      "grad_norm": 591.8302612304688,
+      "learning_rate": 3.275529707340728e-05,
+      "loss": 71.7116,
+      "step": 114910
+    },
+    {
+      "epoch": 0.46429134160481905,
+      "grad_norm": 564.6202392578125,
+      "learning_rate": 3.2751978542425995e-05,
+      "loss": 58.8196,
+      "step": 114920
+    },
+    {
+      "epoch": 0.4643317428701867,
+      "grad_norm": 1191.913330078125,
+      "learning_rate": 3.2748659860315916e-05,
+      "loss": 42.0783,
+      "step": 114930
+    },
+    {
+      "epoch": 0.46437214413555433,
+      "grad_norm": 599.6019287109375,
+      "learning_rate": 3.274534102714172e-05,
+      "loss": 50.7316,
+      "step": 114940
+    },
+    {
+      "epoch": 0.46441254540092197,
+      "grad_norm": 527.0614013671875,
+      "learning_rate": 3.2742022042968104e-05,
+      "loss": 51.8205,
+      "step": 114950
+    },
+    {
+      "epoch": 0.4644529466662896,
+      "grad_norm": 665.9529418945312,
+      "learning_rate": 3.273870290785979e-05,
+      "loss": 68.3194,
+      "step": 114960
+    },
+    {
+      "epoch": 0.4644933479316572,
+      "grad_norm": 515.5640258789062,
+      "learning_rate": 3.2735383621881485e-05,
+      "loss": 37.5466,
+      "step": 114970
+    },
+    {
+      "epoch": 0.46453374919702484,
+      "grad_norm": 822.676513671875,
+      "learning_rate": 3.273206418509788e-05,
+      "loss": 69.2502,
+      "step": 114980
+    },
+    {
+      "epoch": 0.4645741504623925,
+      "grad_norm": 1226.153076171875,
+      "learning_rate": 3.272874459757371e-05,
+      "loss": 50.1505,
+      "step": 114990
+    },
+    {
+      "epoch": 0.4646145517277601,
+      "grad_norm": 1813.3848876953125,
+      "learning_rate": 3.272542485937369e-05,
+      "loss": 77.2155,
+      "step": 115000
+    },
+    {
+      "epoch": 0.46465495299312776,
+      "grad_norm": 579.1246337890625,
+      "learning_rate": 3.2722104970562525e-05,
+      "loss": 93.8051,
+      "step": 115010
+    },
+    {
+      "epoch": 0.4646953542584954,
+      "grad_norm": 0.0,
+      "learning_rate": 3.271878493120496e-05,
+      "loss": 37.53,
+      "step": 115020
+    },
+    {
+      "epoch": 0.464735755523863,
+      "grad_norm": 285.748779296875,
+      "learning_rate": 3.27154647413657e-05,
+      "loss": 51.3141,
+      "step": 115030
+    },
+    {
+      "epoch": 0.4647761567892306,
+      "grad_norm": 585.9220581054688,
+      "learning_rate": 3.271214440110948e-05,
+      "loss": 66.3542,
+      "step": 115040
+    },
+    {
+      "epoch": 0.46481655805459826,
+      "grad_norm": 385.7046813964844,
+      "learning_rate": 3.270882391050104e-05,
+      "loss": 43.7319,
+      "step": 115050
+    },
+    {
+      "epoch": 0.4648569593199659,
+      "grad_norm": 799.7272338867188,
+      "learning_rate": 3.270550326960511e-05,
+      "loss": 73.2882,
+      "step": 115060
+    },
+    {
+      "epoch": 0.46489736058533354,
+      "grad_norm": 1642.924560546875,
+      "learning_rate": 3.270218247848642e-05,
+      "loss": 90.7406,
+      "step": 115070
+    },
+    {
+      "epoch": 0.4649377618507012,
+      "grad_norm": 890.0045776367188,
+      "learning_rate": 3.269886153720972e-05,
+      "loss": 65.4913,
+      "step": 115080
+    },
+    {
+      "epoch": 0.4649781631160688,
+      "grad_norm": 1087.22802734375,
+      "learning_rate": 3.2695540445839764e-05,
+      "loss": 66.0933,
+      "step": 115090
+    },
+    {
+      "epoch": 0.4650185643814364,
+      "grad_norm": 436.0012512207031,
+      "learning_rate": 3.269221920444127e-05,
+      "loss": 48.0393,
+      "step": 115100
+    },
+    {
+      "epoch": 0.46505896564680405,
+      "grad_norm": 821.9993286132812,
+      "learning_rate": 3.2688897813079005e-05,
+      "loss": 48.355,
+      "step": 115110
+    },
+    {
+      "epoch": 0.4650993669121717,
+      "grad_norm": 434.6961364746094,
+      "learning_rate": 3.2685576271817716e-05,
+      "loss": 54.5126,
+      "step": 115120
+    },
+    {
+      "epoch": 0.4651397681775393,
+      "grad_norm": 897.4439697265625,
+      "learning_rate": 3.268225458072217e-05,
+      "loss": 62.3177,
+      "step": 115130
+    },
+    {
+      "epoch": 0.46518016944290697,
+      "grad_norm": 2384.05224609375,
+      "learning_rate": 3.267893273985711e-05,
+      "loss": 72.2372,
+      "step": 115140
+    },
+    {
+      "epoch": 0.4652205707082746,
+      "grad_norm": 724.9418334960938,
+      "learning_rate": 3.26756107492873e-05,
+      "loss": 56.3772,
+      "step": 115150
+    },
+    {
+      "epoch": 0.4652609719736422,
+      "grad_norm": 666.811767578125,
+      "learning_rate": 3.267228860907751e-05,
+      "loss": 45.4823,
+      "step": 115160
+    },
+    {
+      "epoch": 0.46530137323900983,
+      "grad_norm": 846.5064086914062,
+      "learning_rate": 3.266896631929251e-05,
+      "loss": 49.0491,
+      "step": 115170
+    },
+    {
+      "epoch": 0.46534177450437747,
+      "grad_norm": 372.02447509765625,
+      "learning_rate": 3.2665643879997056e-05,
+      "loss": 78.3602,
+      "step": 115180
+    },
+    {
+      "epoch": 0.4653821757697451,
+      "grad_norm": 745.8118896484375,
+      "learning_rate": 3.266232129125593e-05,
+      "loss": 39.577,
+      "step": 115190
+    },
+    {
+      "epoch": 0.46542257703511275,
+      "grad_norm": 936.6083984375,
+      "learning_rate": 3.2658998553133895e-05,
+      "loss": 43.7622,
+      "step": 115200
+    },
+    {
+      "epoch": 0.4654629783004804,
+      "grad_norm": 458.98114013671875,
+      "learning_rate": 3.2655675665695754e-05,
+      "loss": 49.8332,
+      "step": 115210
+    },
+    {
+      "epoch": 0.46550337956584803,
+      "grad_norm": 636.2042846679688,
+      "learning_rate": 3.2652352629006275e-05,
+      "loss": 63.9273,
+      "step": 115220
+    },
+    {
+      "epoch": 0.4655437808312156,
+      "grad_norm": 556.1316528320312,
+      "learning_rate": 3.264902944313023e-05,
+      "loss": 46.3902,
+      "step": 115230
+    },
+    {
+      "epoch": 0.46558418209658325,
+      "grad_norm": 724.4549560546875,
+      "learning_rate": 3.2645706108132424e-05,
+      "loss": 54.3436,
+      "step": 115240
+    },
+    {
+      "epoch": 0.4656245833619509,
+      "grad_norm": 546.86767578125,
+      "learning_rate": 3.264238262407764e-05,
+      "loss": 50.2506,
+      "step": 115250
+    },
+    {
+      "epoch": 0.46566498462731853,
+      "grad_norm": 784.9192504882812,
+      "learning_rate": 3.263905899103068e-05,
+      "loss": 40.6821,
+      "step": 115260
+    },
+    {
+      "epoch": 0.4657053858926862,
+      "grad_norm": 686.7406616210938,
+      "learning_rate": 3.263573520905633e-05,
+      "loss": 52.2234,
+      "step": 115270
+    },
+    {
+      "epoch": 0.4657457871580538,
+      "grad_norm": 709.7639770507812,
+      "learning_rate": 3.263241127821938e-05,
+      "loss": 57.2129,
+      "step": 115280
+    },
+    {
+      "epoch": 0.4657861884234214,
+      "grad_norm": 704.1000366210938,
+      "learning_rate": 3.262908719858466e-05,
+      "loss": 44.5228,
+      "step": 115290
+    },
+    {
+      "epoch": 0.46582658968878904,
+      "grad_norm": 444.74029541015625,
+      "learning_rate": 3.262576297021695e-05,
+      "loss": 49.4663,
+      "step": 115300
+    },
+    {
+      "epoch": 0.4658669909541567,
+      "grad_norm": 951.1953125,
+      "learning_rate": 3.262243859318105e-05,
+      "loss": 60.8233,
+      "step": 115310
+    },
+    {
+      "epoch": 0.4659073922195243,
+      "grad_norm": 466.8185119628906,
+      "learning_rate": 3.2619114067541796e-05,
+      "loss": 92.2595,
+      "step": 115320
+    },
+    {
+      "epoch": 0.46594779348489196,
+      "grad_norm": 581.5831298828125,
+      "learning_rate": 3.2615789393363995e-05,
+      "loss": 53.4414,
+      "step": 115330
+    },
+    {
+      "epoch": 0.4659881947502596,
+      "grad_norm": 1562.436767578125,
+      "learning_rate": 3.261246457071245e-05,
+      "loss": 61.0114,
+      "step": 115340
+    },
+    {
+      "epoch": 0.4660285960156272,
+      "grad_norm": 1172.59130859375,
+      "learning_rate": 3.260913959965201e-05,
+      "loss": 71.2536,
+      "step": 115350
+    },
+    {
+      "epoch": 0.4660689972809948,
+      "grad_norm": 751.3859252929688,
+      "learning_rate": 3.260581448024745e-05,
+      "loss": 53.4582,
+      "step": 115360
+    },
+    {
+      "epoch": 0.46610939854636246,
+      "grad_norm": 479.7427673339844,
+      "learning_rate": 3.260248921256364e-05,
+      "loss": 38.3002,
+      "step": 115370
+    },
+    {
+      "epoch": 0.4661497998117301,
+      "grad_norm": 562.0839233398438,
+      "learning_rate": 3.2599163796665376e-05,
+      "loss": 74.0789,
+      "step": 115380
+    },
+    {
+      "epoch": 0.46619020107709774,
+      "grad_norm": 1441.49853515625,
+      "learning_rate": 3.25958382326175e-05,
+      "loss": 46.7784,
+      "step": 115390
+    },
+    {
+      "epoch": 0.4662306023424654,
+      "grad_norm": 586.0812377929688,
+      "learning_rate": 3.2592512520484856e-05,
+      "loss": 54.3181,
+      "step": 115400
+    },
+    {
+      "epoch": 0.466271003607833,
+      "grad_norm": 782.4759521484375,
+      "learning_rate": 3.2589186660332274e-05,
+      "loss": 59.9024,
+      "step": 115410
+    },
+    {
+      "epoch": 0.4663114048732006,
+      "grad_norm": 623.7604370117188,
+      "learning_rate": 3.2585860652224585e-05,
+      "loss": 52.6946,
+      "step": 115420
+    },
+    {
+      "epoch": 0.46635180613856825,
+      "grad_norm": 958.4971313476562,
+      "learning_rate": 3.2582534496226644e-05,
+      "loss": 56.1301,
+      "step": 115430
+    },
+    {
+      "epoch": 0.4663922074039359,
+      "grad_norm": 803.4427490234375,
+      "learning_rate": 3.257920819240328e-05,
+      "loss": 56.7052,
+      "step": 115440
+    },
+    {
+      "epoch": 0.46643260866930353,
+      "grad_norm": 481.8475646972656,
+      "learning_rate": 3.2575881740819355e-05,
+      "loss": 53.755,
+      "step": 115450
+    },
+    {
+      "epoch": 0.46647300993467117,
+      "grad_norm": 542.654052734375,
+      "learning_rate": 3.257255514153971e-05,
+      "loss": 66.2002,
+      "step": 115460
+    },
+    {
+      "epoch": 0.4665134112000388,
+      "grad_norm": 468.3689270019531,
+      "learning_rate": 3.256922839462921e-05,
+      "loss": 60.2367,
+      "step": 115470
+    },
+    {
+      "epoch": 0.4665538124654064,
+      "grad_norm": 1403.767333984375,
+      "learning_rate": 3.25659015001527e-05,
+      "loss": 108.456,
+      "step": 115480
+    },
+    {
+      "epoch": 0.46659421373077403,
+      "grad_norm": 708.3703002929688,
+      "learning_rate": 3.2562574458175044e-05,
+      "loss": 46.7288,
+      "step": 115490
+    },
+    {
+      "epoch": 0.4666346149961417,
+      "grad_norm": 1207.593017578125,
+      "learning_rate": 3.2559247268761115e-05,
+      "loss": 63.8019,
+      "step": 115500
+    },
+    {
+      "epoch": 0.4666750162615093,
+      "grad_norm": 919.4540405273438,
+      "learning_rate": 3.2555919931975766e-05,
+      "loss": 71.2822,
+      "step": 115510
+    },
+    {
+      "epoch": 0.46671541752687695,
+      "grad_norm": 2360.607421875,
+      "learning_rate": 3.2552592447883865e-05,
+      "loss": 63.8303,
+      "step": 115520
+    },
+    {
+      "epoch": 0.4667558187922446,
+      "grad_norm": 1612.3497314453125,
+      "learning_rate": 3.254926481655028e-05,
+      "loss": 45.8642,
+      "step": 115530
+    },
+    {
+      "epoch": 0.4667962200576122,
+      "grad_norm": 1008.3920288085938,
+      "learning_rate": 3.25459370380399e-05,
+      "loss": 49.0776,
+      "step": 115540
+    },
+    {
+      "epoch": 0.4668366213229798,
+      "grad_norm": 2217.866943359375,
+      "learning_rate": 3.254260911241759e-05,
+      "loss": 72.4225,
+      "step": 115550
+    },
+    {
+      "epoch": 0.46687702258834746,
+      "grad_norm": 413.4029235839844,
+      "learning_rate": 3.253928103974823e-05,
+      "loss": 66.9536,
+      "step": 115560
+    },
+    {
+      "epoch": 0.4669174238537151,
+      "grad_norm": 389.1944580078125,
+      "learning_rate": 3.253595282009671e-05,
+      "loss": 32.475,
+      "step": 115570
+    },
+    {
+      "epoch": 0.46695782511908274,
+      "grad_norm": 0.0,
+      "learning_rate": 3.253262445352791e-05,
+      "loss": 60.644,
+      "step": 115580
+    },
+    {
+      "epoch": 0.4669982263844504,
+      "grad_norm": 366.5119934082031,
+      "learning_rate": 3.252929594010671e-05,
+      "loss": 59.2907,
+      "step": 115590
+    },
+    {
+      "epoch": 0.467038627649818,
+      "grad_norm": 882.8370971679688,
+      "learning_rate": 3.2525967279898015e-05,
+      "loss": 70.1194,
+      "step": 115600
+    },
+    {
+      "epoch": 0.4670790289151856,
+      "grad_norm": 955.0043334960938,
+      "learning_rate": 3.252263847296671e-05,
+      "loss": 62.6897,
+      "step": 115610
+    },
+    {
+      "epoch": 0.46711943018055324,
+      "grad_norm": 815.323486328125,
+      "learning_rate": 3.25193095193777e-05,
+      "loss": 30.4579,
+      "step": 115620
+    },
+    {
+      "epoch": 0.4671598314459209,
+      "grad_norm": 721.734130859375,
+      "learning_rate": 3.251598041919587e-05,
+      "loss": 66.6569,
+      "step": 115630
+    },
+    {
+      "epoch": 0.4672002327112885,
+      "grad_norm": 870.7584838867188,
+      "learning_rate": 3.251265117248614e-05,
+      "loss": 52.2032,
+      "step": 115640
+    },
+    {
+      "epoch": 0.46724063397665616,
+      "grad_norm": 4967.96728515625,
+      "learning_rate": 3.25093217793134e-05,
+      "loss": 56.7999,
+      "step": 115650
+    },
+    {
+      "epoch": 0.4672810352420238,
+      "grad_norm": 805.152099609375,
+      "learning_rate": 3.250599223974258e-05,
+      "loss": 52.8073,
+      "step": 115660
+    },
+    {
+      "epoch": 0.4673214365073914,
+      "grad_norm": 614.6864624023438,
+      "learning_rate": 3.250266255383857e-05,
+      "loss": 54.9746,
+      "step": 115670
+    },
+    {
+      "epoch": 0.467361837772759,
+      "grad_norm": 1058.3001708984375,
+      "learning_rate": 3.249933272166629e-05,
+      "loss": 92.6986,
+      "step": 115680
+    },
+    {
+      "epoch": 0.46740223903812667,
+      "grad_norm": 1258.4139404296875,
+      "learning_rate": 3.249600274329066e-05,
+      "loss": 54.6076,
+      "step": 115690
+    },
+    {
+      "epoch": 0.4674426403034943,
+      "grad_norm": 712.8331909179688,
+      "learning_rate": 3.24926726187766e-05,
+      "loss": 65.3554,
+      "step": 115700
+    },
+    {
+      "epoch": 0.46748304156886195,
+      "grad_norm": 826.5805053710938,
+      "learning_rate": 3.248934234818902e-05,
+      "loss": 82.2117,
+      "step": 115710
+    },
+    {
+      "epoch": 0.4675234428342296,
+      "grad_norm": 770.5543212890625,
+      "learning_rate": 3.248601193159287e-05,
+      "loss": 73.6685,
+      "step": 115720
+    },
+    {
+      "epoch": 0.4675638440995972,
+      "grad_norm": 780.5651245117188,
+      "learning_rate": 3.248268136905304e-05,
+      "loss": 44.8846,
+      "step": 115730
+    },
+    {
+      "epoch": 0.4676042453649648,
+      "grad_norm": 1183.1986083984375,
+      "learning_rate": 3.247935066063451e-05,
+      "loss": 57.1673,
+      "step": 115740
+    },
+    {
+      "epoch": 0.46764464663033245,
+      "grad_norm": 461.81573486328125,
+      "learning_rate": 3.247601980640217e-05,
+      "loss": 45.7133,
+      "step": 115750
+    },
+    {
+      "epoch": 0.4676850478957001,
+      "grad_norm": 720.4097290039062,
+      "learning_rate": 3.247268880642098e-05,
+      "loss": 44.3927,
+      "step": 115760
+    },
+    {
+      "epoch": 0.46772544916106773,
+      "grad_norm": 1013.3065795898438,
+      "learning_rate": 3.246935766075588e-05,
+      "loss": 52.5788,
+      "step": 115770
+    },
+    {
+      "epoch": 0.46776585042643537,
+      "grad_norm": 947.147216796875,
+      "learning_rate": 3.24660263694718e-05,
+      "loss": 93.3442,
+      "step": 115780
+    },
+    {
+      "epoch": 0.467806251691803,
+      "grad_norm": 676.0074462890625,
+      "learning_rate": 3.24626949326337e-05,
+      "loss": 48.2604,
+      "step": 115790
+    },
+    {
+      "epoch": 0.4678466529571706,
+      "grad_norm": 433.5167541503906,
+      "learning_rate": 3.245936335030651e-05,
+      "loss": 68.8754,
+      "step": 115800
+    },
+    {
+      "epoch": 0.46788705422253823,
+      "grad_norm": 759.6268920898438,
+      "learning_rate": 3.2456031622555197e-05,
+      "loss": 42.3743,
+      "step": 115810
+    },
+    {
+      "epoch": 0.4679274554879059,
+      "grad_norm": 650.4202880859375,
+      "learning_rate": 3.245269974944471e-05,
+      "loss": 83.6153,
+      "step": 115820
+    },
+    {
+      "epoch": 0.4679678567532735,
+      "grad_norm": 963.0789794921875,
+      "learning_rate": 3.2449367731039996e-05,
+      "loss": 60.6028,
+      "step": 115830
+    },
+    {
+      "epoch": 0.46800825801864115,
+      "grad_norm": 2292.56005859375,
+      "learning_rate": 3.244603556740603e-05,
+      "loss": 74.8116,
+      "step": 115840
+    },
+    {
+      "epoch": 0.4680486592840088,
+      "grad_norm": 457.2271423339844,
+      "learning_rate": 3.2442703258607766e-05,
+      "loss": 37.3418,
+      "step": 115850
+    },
+    {
+      "epoch": 0.4680890605493764,
+      "grad_norm": 474.98126220703125,
+      "learning_rate": 3.243937080471017e-05,
+      "loss": 61.6,
+      "step": 115860
+    },
+    {
+      "epoch": 0.468129461814744,
+      "grad_norm": 528.2426147460938,
+      "learning_rate": 3.243603820577822e-05,
+      "loss": 37.5654,
+      "step": 115870
+    },
+    {
+      "epoch": 0.46816986308011166,
+      "grad_norm": 885.8614501953125,
+      "learning_rate": 3.243270546187687e-05,
+      "loss": 39.1532,
+      "step": 115880
+    },
+    {
+      "epoch": 0.4682102643454793,
+      "grad_norm": 2204.4306640625,
+      "learning_rate": 3.242937257307109e-05,
+      "loss": 67.4763,
+      "step": 115890
+    },
+    {
+      "epoch": 0.46825066561084694,
+      "grad_norm": 1517.2777099609375,
+      "learning_rate": 3.2426039539425876e-05,
+      "loss": 53.94,
+      "step": 115900
+    },
+    {
+      "epoch": 0.4682910668762146,
+      "grad_norm": 2055.982666015625,
+      "learning_rate": 3.2422706361006194e-05,
+      "loss": 58.1359,
+      "step": 115910
+    },
+    {
+      "epoch": 0.4683314681415822,
+      "grad_norm": 607.13330078125,
+      "learning_rate": 3.241937303787703e-05,
+      "loss": 49.6527,
+      "step": 115920
+    },
+    {
+      "epoch": 0.4683718694069498,
+      "grad_norm": 899.364990234375,
+      "learning_rate": 3.2416039570103375e-05,
+      "loss": 57.977,
+      "step": 115930
+    },
+    {
+      "epoch": 0.46841227067231744,
+      "grad_norm": 654.2115478515625,
+      "learning_rate": 3.241270595775021e-05,
+      "loss": 73.5976,
+      "step": 115940
+    },
+    {
+      "epoch": 0.4684526719376851,
+      "grad_norm": 786.1331176757812,
+      "learning_rate": 3.240937220088253e-05,
+      "loss": 35.1766,
+      "step": 115950
+    },
+    {
+      "epoch": 0.4684930732030527,
+      "grad_norm": 1251.7491455078125,
+      "learning_rate": 3.240603829956531e-05,
+      "loss": 62.6786,
+      "step": 115960
+    },
+    {
+      "epoch": 0.46853347446842036,
+      "grad_norm": 1378.3673095703125,
+      "learning_rate": 3.240270425386357e-05,
+      "loss": 75.6629,
+      "step": 115970
+    },
+    {
+      "epoch": 0.468573875733788,
+      "grad_norm": 344.47747802734375,
+      "learning_rate": 3.2399370063842294e-05,
+      "loss": 46.953,
+      "step": 115980
+    },
+    {
+      "epoch": 0.4686142769991556,
+      "grad_norm": 383.6866760253906,
+      "learning_rate": 3.23960357295665e-05,
+      "loss": 47.923,
+      "step": 115990
+    },
+    {
+      "epoch": 0.46865467826452323,
+      "grad_norm": 471.9056091308594,
+      "learning_rate": 3.239270125110117e-05,
+      "loss": 77.3322,
+      "step": 116000
+    },
+    {
+      "epoch": 0.46869507952989087,
+      "grad_norm": 879.043701171875,
+      "learning_rate": 3.238936662851133e-05,
+      "loss": 48.9826,
+      "step": 116010
+    },
+    {
+      "epoch": 0.4687354807952585,
+      "grad_norm": 632.9555053710938,
+      "learning_rate": 3.2386031861861976e-05,
+      "loss": 31.8865,
+      "step": 116020
+    },
+    {
+      "epoch": 0.46877588206062615,
+      "grad_norm": 1467.3824462890625,
+      "learning_rate": 3.2382696951218135e-05,
+      "loss": 95.1815,
+      "step": 116030
+    },
+    {
+      "epoch": 0.4688162833259938,
+      "grad_norm": 336.1209411621094,
+      "learning_rate": 3.2379361896644816e-05,
+      "loss": 54.7575,
+      "step": 116040
+    },
+    {
+      "epoch": 0.46885668459136143,
+      "grad_norm": 873.735595703125,
+      "learning_rate": 3.237602669820704e-05,
+      "loss": 63.7571,
+      "step": 116050
+    },
+    {
+      "epoch": 0.468897085856729,
+      "grad_norm": 537.8552856445312,
+      "learning_rate": 3.2372691355969816e-05,
+      "loss": 55.9651,
+      "step": 116060
+    },
+    {
+      "epoch": 0.46893748712209665,
+      "grad_norm": 353.98614501953125,
+      "learning_rate": 3.2369355869998185e-05,
+      "loss": 39.4995,
+      "step": 116070
+    },
+    {
+      "epoch": 0.4689778883874643,
+      "grad_norm": 195.72950744628906,
+      "learning_rate": 3.236602024035716e-05,
+      "loss": 53.4221,
+      "step": 116080
+    },
+    {
+      "epoch": 0.46901828965283193,
+      "grad_norm": 657.9056396484375,
+      "learning_rate": 3.236268446711179e-05,
+      "loss": 66.8073,
+      "step": 116090
+    },
+    {
+      "epoch": 0.4690586909181996,
+      "grad_norm": 483.935546875,
+      "learning_rate": 3.235934855032709e-05,
+      "loss": 51.066,
+      "step": 116100
+    },
+    {
+      "epoch": 0.4690990921835672,
+      "grad_norm": 1542.11865234375,
+      "learning_rate": 3.23560124900681e-05,
+      "loss": 89.0544,
+      "step": 116110
+    },
+    {
+      "epoch": 0.4691394934489348,
+      "grad_norm": 838.8958740234375,
+      "learning_rate": 3.235267628639987e-05,
+      "loss": 60.8755,
+      "step": 116120
+    },
+    {
+      "epoch": 0.46917989471430244,
+      "grad_norm": 861.8223266601562,
+      "learning_rate": 3.234933993938742e-05,
+      "loss": 47.4594,
+      "step": 116130
+    },
+    {
+      "epoch": 0.4692202959796701,
+      "grad_norm": 550.157958984375,
+      "learning_rate": 3.2346003449095805e-05,
+      "loss": 53.5828,
+      "step": 116140
+    },
+    {
+      "epoch": 0.4692606972450377,
+      "grad_norm": 503.9146728515625,
+      "learning_rate": 3.234266681559007e-05,
+      "loss": 78.4864,
+      "step": 116150
+    },
+    {
+      "epoch": 0.46930109851040536,
+      "grad_norm": 333.4444885253906,
+      "learning_rate": 3.2339330038935265e-05,
+      "loss": 54.1187,
+      "step": 116160
+    },
+    {
+      "epoch": 0.469341499775773,
+      "grad_norm": 666.8385620117188,
+      "learning_rate": 3.233599311919644e-05,
+      "loss": 60.4023,
+      "step": 116170
+    },
+    {
+      "epoch": 0.4693819010411406,
+      "grad_norm": 641.484619140625,
+      "learning_rate": 3.233265605643866e-05,
+      "loss": 63.9357,
+      "step": 116180
+    },
+    {
+      "epoch": 0.4694223023065082,
+      "grad_norm": 1227.126953125,
+      "learning_rate": 3.232931885072697e-05,
+      "loss": 61.4396,
+      "step": 116190
+    },
+    {
+      "epoch": 0.46946270357187586,
+      "grad_norm": 1222.384765625,
+      "learning_rate": 3.2325981502126433e-05,
+      "loss": 61.0376,
+      "step": 116200
+    },
+    {
+      "epoch": 0.4695031048372435,
+      "grad_norm": 706.1589965820312,
+      "learning_rate": 3.232264401070213e-05,
+      "loss": 58.2779,
+      "step": 116210
+    },
+    {
+      "epoch": 0.46954350610261114,
+      "grad_norm": 558.5364990234375,
+      "learning_rate": 3.231930637651909e-05,
+      "loss": 52.5935,
+      "step": 116220
+    },
+    {
+      "epoch": 0.4695839073679788,
+      "grad_norm": 928.905029296875,
+      "learning_rate": 3.231596859964242e-05,
+      "loss": 61.5201,
+      "step": 116230
+    },
+    {
+      "epoch": 0.4696243086333464,
+      "grad_norm": 630.4097290039062,
+      "learning_rate": 3.2312630680137175e-05,
+      "loss": 37.5664,
+      "step": 116240
+    },
+    {
+      "epoch": 0.469664709898714,
+      "grad_norm": 1384.9384765625,
+      "learning_rate": 3.230929261806842e-05,
+      "loss": 53.7496,
+      "step": 116250
+    },
+    {
+      "epoch": 0.46970511116408165,
+      "grad_norm": 147.7491912841797,
+      "learning_rate": 3.230595441350125e-05,
+      "loss": 55.0247,
+      "step": 116260
+    },
+    {
+      "epoch": 0.4697455124294493,
+      "grad_norm": 334.8374938964844,
+      "learning_rate": 3.2302616066500735e-05,
+      "loss": 46.3881,
+      "step": 116270
+    },
+    {
+      "epoch": 0.4697859136948169,
+      "grad_norm": 789.24755859375,
+      "learning_rate": 3.229927757713196e-05,
+      "loss": 70.6806,
+      "step": 116280
+    },
+    {
+      "epoch": 0.46982631496018457,
+      "grad_norm": 1363.5418701171875,
+      "learning_rate": 3.229593894546001e-05,
+      "loss": 58.1901,
+      "step": 116290
+    },
+    {
+      "epoch": 0.4698667162255522,
+      "grad_norm": 609.9053344726562,
+      "learning_rate": 3.229260017154997e-05,
+      "loss": 98.0226,
+      "step": 116300
+    },
+    {
+      "epoch": 0.4699071174909198,
+      "grad_norm": 865.0936279296875,
+      "learning_rate": 3.228926125546695e-05,
+      "loss": 61.3911,
+      "step": 116310
+    },
+    {
+      "epoch": 0.46994751875628743,
+      "grad_norm": 590.865234375,
+      "learning_rate": 3.228592219727602e-05,
+      "loss": 86.893,
+      "step": 116320
+    },
+    {
+      "epoch": 0.46998792002165507,
+      "grad_norm": 789.493408203125,
+      "learning_rate": 3.2282582997042285e-05,
+      "loss": 49.3227,
+      "step": 116330
+    },
+    {
+      "epoch": 0.4700283212870227,
+      "grad_norm": 707.34814453125,
+      "learning_rate": 3.2279243654830836e-05,
+      "loss": 69.9071,
+      "step": 116340
+    },
+    {
+      "epoch": 0.47006872255239035,
+      "grad_norm": 1062.26123046875,
+      "learning_rate": 3.2275904170706797e-05,
+      "loss": 55.3705,
+      "step": 116350
+    },
+    {
+      "epoch": 0.470109123817758,
+      "grad_norm": 967.4932861328125,
+      "learning_rate": 3.227256454473526e-05,
+      "loss": 42.835,
+      "step": 116360
+    },
+    {
+      "epoch": 0.47014952508312563,
+      "grad_norm": 645.4476318359375,
+      "learning_rate": 3.226922477698133e-05,
+      "loss": 57.326,
+      "step": 116370
+    },
+    {
+      "epoch": 0.4701899263484932,
+      "grad_norm": 884.6676025390625,
+      "learning_rate": 3.226588486751012e-05,
+      "loss": 66.296,
+      "step": 116380
+    },
+    {
+      "epoch": 0.47023032761386085,
+      "grad_norm": 3511.986083984375,
+      "learning_rate": 3.2262544816386745e-05,
+      "loss": 49.9027,
+      "step": 116390
+    },
+    {
+      "epoch": 0.4702707288792285,
+      "grad_norm": 476.053955078125,
+      "learning_rate": 3.225920462367632e-05,
+      "loss": 68.0334,
+      "step": 116400
+    },
+    {
+      "epoch": 0.47031113014459613,
+      "grad_norm": 367.8993225097656,
+      "learning_rate": 3.225586428944396e-05,
+      "loss": 41.9805,
+      "step": 116410
+    },
+    {
+      "epoch": 0.4703515314099638,
+      "grad_norm": 500.8946533203125,
+      "learning_rate": 3.225252381375479e-05,
+      "loss": 65.6448,
+      "step": 116420
+    },
+    {
+      "epoch": 0.4703919326753314,
+      "grad_norm": 1217.2655029296875,
+      "learning_rate": 3.224918319667394e-05,
+      "loss": 83.4919,
+      "step": 116430
+    },
+    {
+      "epoch": 0.470432333940699,
+      "grad_norm": 662.4497680664062,
+      "learning_rate": 3.2245842438266526e-05,
+      "loss": 65.4877,
+      "step": 116440
+    },
+    {
+      "epoch": 0.47047273520606664,
+      "grad_norm": 725.7619018554688,
+      "learning_rate": 3.224250153859769e-05,
+      "loss": 55.515,
+      "step": 116450
+    },
+    {
+      "epoch": 0.4705131364714343,
+      "grad_norm": 910.7822265625,
+      "learning_rate": 3.223916049773256e-05,
+      "loss": 43.4893,
+      "step": 116460
+    },
+    {
+      "epoch": 0.4705535377368019,
+      "grad_norm": 2010.1138916015625,
+      "learning_rate": 3.223581931573625e-05,
+      "loss": 58.6502,
+      "step": 116470
+    },
+    {
+      "epoch": 0.47059393900216956,
+      "grad_norm": 715.5496826171875,
+      "learning_rate": 3.223247799267394e-05,
+      "loss": 84.7735,
+      "step": 116480
+    },
+    {
+      "epoch": 0.4706343402675372,
+      "grad_norm": 737.7570190429688,
+      "learning_rate": 3.2229136528610736e-05,
+      "loss": 39.7872,
+      "step": 116490
+    },
+    {
+      "epoch": 0.4706747415329048,
+      "grad_norm": 532.3171997070312,
+      "learning_rate": 3.222579492361179e-05,
+      "loss": 67.2663,
+      "step": 116500
+    },
+    {
+      "epoch": 0.4707151427982724,
+      "grad_norm": 1556.9482421875,
+      "learning_rate": 3.222245317774226e-05,
+      "loss": 57.7589,
+      "step": 116510
+    },
+    {
+      "epoch": 0.47075554406364006,
+      "grad_norm": 1564.6119384765625,
+      "learning_rate": 3.221911129106728e-05,
+      "loss": 41.327,
+      "step": 116520
+    },
+    {
+      "epoch": 0.4707959453290077,
+      "grad_norm": 668.7022094726562,
+      "learning_rate": 3.221576926365202e-05,
+      "loss": 53.1316,
+      "step": 116530
+    },
+    {
+      "epoch": 0.47083634659437534,
+      "grad_norm": 940.4009399414062,
+      "learning_rate": 3.221242709556161e-05,
+      "loss": 52.2775,
+      "step": 116540
+    },
+    {
+      "epoch": 0.470876747859743,
+      "grad_norm": 929.6431884765625,
+      "learning_rate": 3.220908478686123e-05,
+      "loss": 63.8318,
+      "step": 116550
+    },
+    {
+      "epoch": 0.4709171491251106,
+      "grad_norm": 770.2941284179688,
+      "learning_rate": 3.220574233761603e-05,
+      "loss": 50.6914,
+      "step": 116560
+    },
+    {
+      "epoch": 0.4709575503904782,
+      "grad_norm": 2690.2001953125,
+      "learning_rate": 3.220239974789117e-05,
+      "loss": 103.2043,
+      "step": 116570
+    },
+    {
+      "epoch": 0.47099795165584585,
+      "grad_norm": 744.5656127929688,
+      "learning_rate": 3.219905701775182e-05,
+      "loss": 40.3893,
+      "step": 116580
+    },
+    {
+      "epoch": 0.4710383529212135,
+      "grad_norm": 2249.509765625,
+      "learning_rate": 3.219571414726315e-05,
+      "loss": 55.5555,
+      "step": 116590
+    },
+    {
+      "epoch": 0.47107875418658113,
+      "grad_norm": 392.30657958984375,
+      "learning_rate": 3.219237113649032e-05,
+      "loss": 63.3421,
+      "step": 116600
+    },
+    {
+      "epoch": 0.47111915545194877,
+      "grad_norm": 0.0,
+      "learning_rate": 3.2189027985498514e-05,
+      "loss": 82.4716,
+      "step": 116610
+    },
+    {
+      "epoch": 0.4711595567173164,
+      "grad_norm": 626.5792846679688,
+      "learning_rate": 3.2185684694352916e-05,
+      "loss": 38.248,
+      "step": 116620
+    },
+    {
+      "epoch": 0.471199957982684,
+      "grad_norm": 0.0,
+      "learning_rate": 3.218234126311869e-05,
+      "loss": 39.9608,
+      "step": 116630
+    },
+    {
+      "epoch": 0.47124035924805163,
+      "grad_norm": 959.5233154296875,
+      "learning_rate": 3.2178997691861014e-05,
+      "loss": 56.8771,
+      "step": 116640
+    },
+    {
+      "epoch": 0.4712807605134193,
+      "grad_norm": 368.611083984375,
+      "learning_rate": 3.217565398064509e-05,
+      "loss": 30.2148,
+      "step": 116650
+    },
+    {
+      "epoch": 0.4713211617787869,
+      "grad_norm": 2147.029296875,
+      "learning_rate": 3.2172310129536096e-05,
+      "loss": 56.8126,
+      "step": 116660
+    },
+    {
+      "epoch": 0.47136156304415455,
+      "grad_norm": 528.3900146484375,
+      "learning_rate": 3.2168966138599225e-05,
+      "loss": 93.9726,
+      "step": 116670
+    },
+    {
+      "epoch": 0.4714019643095222,
+      "grad_norm": 1705.246337890625,
+      "learning_rate": 3.2165622007899676e-05,
+      "loss": 42.9779,
+      "step": 116680
+    },
+    {
+      "epoch": 0.47144236557488983,
+      "grad_norm": 1022.4556884765625,
+      "learning_rate": 3.216227773750262e-05,
+      "loss": 40.3776,
+      "step": 116690
+    },
+    {
+      "epoch": 0.4714827668402574,
+      "grad_norm": 836.2701416015625,
+      "learning_rate": 3.215893332747328e-05,
+      "loss": 75.0597,
+      "step": 116700
+    },
+    {
+      "epoch": 0.47152316810562506,
+      "grad_norm": 493.19110107421875,
+      "learning_rate": 3.2155588777876856e-05,
+      "loss": 45.9303,
+      "step": 116710
+    },
+    {
+      "epoch": 0.4715635693709927,
+      "grad_norm": 800.6790771484375,
+      "learning_rate": 3.215224408877854e-05,
+      "loss": 52.3173,
+      "step": 116720
+    },
+    {
+      "epoch": 0.47160397063636034,
+      "grad_norm": 1031.7352294921875,
+      "learning_rate": 3.2148899260243545e-05,
+      "loss": 41.5077,
+      "step": 116730
+    },
+    {
+      "epoch": 0.471644371901728,
+      "grad_norm": 524.3075561523438,
+      "learning_rate": 3.214555429233707e-05,
+      "loss": 43.6509,
+      "step": 116740
+    },
+    {
+      "epoch": 0.4716847731670956,
+      "grad_norm": 2053.65576171875,
+      "learning_rate": 3.214220918512434e-05,
+      "loss": 52.429,
+      "step": 116750
+    },
+    {
+      "epoch": 0.4717251744324632,
+      "grad_norm": 660.8265380859375,
+      "learning_rate": 3.213886393867057e-05,
+      "loss": 59.9484,
+      "step": 116760
+    },
+    {
+      "epoch": 0.47176557569783084,
+      "grad_norm": 539.3421630859375,
+      "learning_rate": 3.2135518553040964e-05,
+      "loss": 56.5854,
+      "step": 116770
+    },
+    {
+      "epoch": 0.4718059769631985,
+      "grad_norm": 461.3349304199219,
+      "learning_rate": 3.2132173028300756e-05,
+      "loss": 56.3017,
+      "step": 116780
+    },
+    {
+      "epoch": 0.4718463782285661,
+      "grad_norm": 1058.8426513671875,
+      "learning_rate": 3.212882736451516e-05,
+      "loss": 47.9099,
+      "step": 116790
+    },
+    {
+      "epoch": 0.47188677949393376,
+      "grad_norm": 521.5872802734375,
+      "learning_rate": 3.21254815617494e-05,
+      "loss": 48.9531,
+      "step": 116800
+    },
+    {
+      "epoch": 0.4719271807593014,
+      "grad_norm": 789.1767578125,
+      "learning_rate": 3.212213562006872e-05,
+      "loss": 77.1051,
+      "step": 116810
+    },
+    {
+      "epoch": 0.471967582024669,
+      "grad_norm": 710.3841552734375,
+      "learning_rate": 3.2118789539538335e-05,
+      "loss": 69.6694,
+      "step": 116820
+    },
+    {
+      "epoch": 0.4720079832900366,
+      "grad_norm": 1419.693115234375,
+      "learning_rate": 3.211544332022348e-05,
+      "loss": 65.4239,
+      "step": 116830
+    },
+    {
+      "epoch": 0.47204838455540427,
+      "grad_norm": 518.5504760742188,
+      "learning_rate": 3.21120969621894e-05,
+      "loss": 60.0513,
+      "step": 116840
+    },
+    {
+      "epoch": 0.4720887858207719,
+      "grad_norm": 385.5600280761719,
+      "learning_rate": 3.210875046550132e-05,
+      "loss": 39.933,
+      "step": 116850
+    },
+    {
+      "epoch": 0.47212918708613955,
+      "grad_norm": 850.2272338867188,
+      "learning_rate": 3.210540383022449e-05,
+      "loss": 81.0095,
+      "step": 116860
+    },
+    {
+      "epoch": 0.4721695883515072,
+      "grad_norm": 443.19189453125,
+      "learning_rate": 3.210205705642416e-05,
+      "loss": 81.4084,
+      "step": 116870
+    },
+    {
+      "epoch": 0.4722099896168748,
+      "grad_norm": 490.591796875,
+      "learning_rate": 3.209871014416557e-05,
+      "loss": 44.2058,
+      "step": 116880
+    },
+    {
+      "epoch": 0.4722503908822424,
+      "grad_norm": 1418.0242919921875,
+      "learning_rate": 3.209536309351397e-05,
+      "loss": 80.6199,
+      "step": 116890
+    },
+    {
+      "epoch": 0.47229079214761005,
+      "grad_norm": 1384.6348876953125,
+      "learning_rate": 3.209201590453461e-05,
+      "loss": 90.8569,
+      "step": 116900
+    },
+    {
+      "epoch": 0.4723311934129777,
+      "grad_norm": 595.4918823242188,
+      "learning_rate": 3.208866857729276e-05,
+      "loss": 63.6417,
+      "step": 116910
+    },
+    {
+      "epoch": 0.47237159467834533,
+      "grad_norm": 689.3577880859375,
+      "learning_rate": 3.208532111185365e-05,
+      "loss": 57.5565,
+      "step": 116920
+    },
+    {
+      "epoch": 0.47241199594371297,
+      "grad_norm": 611.20751953125,
+      "learning_rate": 3.208197350828257e-05,
+      "loss": 80.1349,
+      "step": 116930
+    },
+    {
+      "epoch": 0.4724523972090806,
+      "grad_norm": 793.7510375976562,
+      "learning_rate": 3.207862576664477e-05,
+      "loss": 73.2373,
+      "step": 116940
+    },
+    {
+      "epoch": 0.4724927984744482,
+      "grad_norm": 777.3289184570312,
+      "learning_rate": 3.207527788700551e-05,
+      "loss": 74.5804,
+      "step": 116950
+    },
+    {
+      "epoch": 0.47253319973981583,
+      "grad_norm": 1934.6875,
+      "learning_rate": 3.207192986943006e-05,
+      "loss": 78.4664,
+      "step": 116960
+    },
+    {
+      "epoch": 0.4725736010051835,
+      "grad_norm": 573.094482421875,
+      "learning_rate": 3.206858171398371e-05,
+      "loss": 77.9668,
+      "step": 116970
+    },
+    {
+      "epoch": 0.4726140022705511,
+      "grad_norm": 290.15185546875,
+      "learning_rate": 3.206523342073172e-05,
+      "loss": 56.8567,
+      "step": 116980
+    },
+    {
+      "epoch": 0.47265440353591875,
+      "grad_norm": 603.3993530273438,
+      "learning_rate": 3.206188498973935e-05,
+      "loss": 62.5748,
+      "step": 116990
+    },
+    {
+      "epoch": 0.4726948048012864,
+      "grad_norm": 459.7473449707031,
+      "learning_rate": 3.205853642107192e-05,
+      "loss": 55.5894,
+      "step": 117000
+    },
+    {
+      "epoch": 0.47273520606665403,
+      "grad_norm": 2034.221923828125,
+      "learning_rate": 3.2055187714794674e-05,
+      "loss": 94.7735,
+      "step": 117010
+    },
+    {
+      "epoch": 0.4727756073320216,
+      "grad_norm": 556.3843383789062,
+      "learning_rate": 3.205183887097291e-05,
+      "loss": 49.9171,
+      "step": 117020
+    },
+    {
+      "epoch": 0.47281600859738926,
+      "grad_norm": 618.3563232421875,
+      "learning_rate": 3.2048489889671915e-05,
+      "loss": 64.9501,
+      "step": 117030
+    },
+    {
+      "epoch": 0.4728564098627569,
+      "grad_norm": 1050.610107421875,
+      "learning_rate": 3.204514077095699e-05,
+      "loss": 45.0024,
+      "step": 117040
+    },
+    {
+      "epoch": 0.47289681112812454,
+      "grad_norm": 676.1912841796875,
+      "learning_rate": 3.2041791514893416e-05,
+      "loss": 71.348,
+      "step": 117050
+    },
+    {
+      "epoch": 0.4729372123934922,
+      "grad_norm": 338.9596252441406,
+      "learning_rate": 3.2038442121546487e-05,
+      "loss": 60.3322,
+      "step": 117060
+    },
+    {
+      "epoch": 0.4729776136588598,
+      "grad_norm": 1104.4334716796875,
+      "learning_rate": 3.2035092590981514e-05,
+      "loss": 65.3458,
+      "step": 117070
+    },
+    {
+      "epoch": 0.4730180149242274,
+      "grad_norm": 977.1143188476562,
+      "learning_rate": 3.203174292326378e-05,
+      "loss": 65.3839,
+      "step": 117080
+    },
+    {
+      "epoch": 0.47305841618959504,
+      "grad_norm": 552.0875854492188,
+      "learning_rate": 3.20283931184586e-05,
+      "loss": 52.0979,
+      "step": 117090
+    },
+    {
+      "epoch": 0.4730988174549627,
+      "grad_norm": 615.712646484375,
+      "learning_rate": 3.202504317663128e-05,
+      "loss": 32.2021,
+      "step": 117100
+    },
+    {
+      "epoch": 0.4731392187203303,
+      "grad_norm": 550.6535034179688,
+      "learning_rate": 3.2021693097847125e-05,
+      "loss": 70.6825,
+      "step": 117110
+    },
+    {
+      "epoch": 0.47317961998569796,
+      "grad_norm": 610.3817749023438,
+      "learning_rate": 3.2018342882171445e-05,
+      "loss": 36.3148,
+      "step": 117120
+    },
+    {
+      "epoch": 0.4732200212510656,
+      "grad_norm": 755.9873657226562,
+      "learning_rate": 3.2014992529669566e-05,
+      "loss": 44.023,
+      "step": 117130
+    },
+    {
+      "epoch": 0.4732604225164332,
+      "grad_norm": 847.390625,
+      "learning_rate": 3.2011642040406784e-05,
+      "loss": 49.9786,
+      "step": 117140
+    },
+    {
+      "epoch": 0.47330082378180083,
+      "grad_norm": 644.954345703125,
+      "learning_rate": 3.200829141444844e-05,
+      "loss": 93.2963,
+      "step": 117150
+    },
+    {
+      "epoch": 0.47334122504716847,
+      "grad_norm": 646.0951538085938,
+      "learning_rate": 3.2004940651859844e-05,
+      "loss": 55.8603,
+      "step": 117160
+    },
+    {
+      "epoch": 0.4733816263125361,
+      "grad_norm": 1527.83447265625,
+      "learning_rate": 3.200158975270633e-05,
+      "loss": 69.3903,
+      "step": 117170
+    },
+    {
+      "epoch": 0.47342202757790375,
+      "grad_norm": 1161.49658203125,
+      "learning_rate": 3.1998238717053206e-05,
+      "loss": 63.7593,
+      "step": 117180
+    },
+    {
+      "epoch": 0.4734624288432714,
+      "grad_norm": 1293.619140625,
+      "learning_rate": 3.199488754496582e-05,
+      "loss": 63.1005,
+      "step": 117190
+    },
+    {
+      "epoch": 0.47350283010863903,
+      "grad_norm": 505.3819580078125,
+      "learning_rate": 3.19915362365095e-05,
+      "loss": 50.2402,
+      "step": 117200
+    },
+    {
+      "epoch": 0.4735432313740066,
+      "grad_norm": 648.815185546875,
+      "learning_rate": 3.198818479174959e-05,
+      "loss": 53.0908,
+      "step": 117210
+    },
+    {
+      "epoch": 0.47358363263937425,
+      "grad_norm": 601.2451171875,
+      "learning_rate": 3.198483321075141e-05,
+      "loss": 56.0472,
+      "step": 117220
+    },
+    {
+      "epoch": 0.4736240339047419,
+      "grad_norm": 401.19842529296875,
+      "learning_rate": 3.198148149358031e-05,
+      "loss": 58.7673,
+      "step": 117230
+    },
+    {
+      "epoch": 0.47366443517010953,
+      "grad_norm": 942.248046875,
+      "learning_rate": 3.197812964030164e-05,
+      "loss": 42.9004,
+      "step": 117240
+    },
+    {
+      "epoch": 0.4737048364354772,
+      "grad_norm": 1621.0218505859375,
+      "learning_rate": 3.1974777650980735e-05,
+      "loss": 59.7592,
+      "step": 117250
+    },
+    {
+      "epoch": 0.4737452377008448,
+      "grad_norm": 1157.42431640625,
+      "learning_rate": 3.197142552568295e-05,
+      "loss": 53.8315,
+      "step": 117260
+    },
+    {
+      "epoch": 0.4737856389662124,
+      "grad_norm": 2135.51611328125,
+      "learning_rate": 3.196807326447363e-05,
+      "loss": 74.3025,
+      "step": 117270
+    },
+    {
+      "epoch": 0.47382604023158004,
+      "grad_norm": 982.652587890625,
+      "learning_rate": 3.196472086741815e-05,
+      "loss": 56.9442,
+      "step": 117280
+    },
+    {
+      "epoch": 0.4738664414969477,
+      "grad_norm": 303.63623046875,
+      "learning_rate": 3.1961368334581844e-05,
+      "loss": 44.213,
+      "step": 117290
+    },
+    {
+      "epoch": 0.4739068427623153,
+      "grad_norm": 555.4284057617188,
+      "learning_rate": 3.195801566603007e-05,
+      "loss": 53.6509,
+      "step": 117300
+    },
+    {
+      "epoch": 0.47394724402768296,
+      "grad_norm": 946.455078125,
+      "learning_rate": 3.1954662861828204e-05,
+      "loss": 82.5156,
+      "step": 117310
+    },
+    {
+      "epoch": 0.4739876452930506,
+      "grad_norm": 1169.52685546875,
+      "learning_rate": 3.195130992204161e-05,
+      "loss": 57.4077,
+      "step": 117320
+    },
+    {
+      "epoch": 0.47402804655841824,
+      "grad_norm": 595.98291015625,
+      "learning_rate": 3.1947956846735645e-05,
+      "loss": 76.41,
+      "step": 117330
+    },
+    {
+      "epoch": 0.4740684478237858,
+      "grad_norm": 463.3590393066406,
+      "learning_rate": 3.194460363597569e-05,
+      "loss": 44.9645,
+      "step": 117340
+    },
+    {
+      "epoch": 0.47410884908915346,
+      "grad_norm": 443.7630615234375,
+      "learning_rate": 3.1941250289827104e-05,
+      "loss": 49.6183,
+      "step": 117350
+    },
+    {
+      "epoch": 0.4741492503545211,
+      "grad_norm": 2856.91259765625,
+      "learning_rate": 3.193789680835527e-05,
+      "loss": 59.4347,
+      "step": 117360
+    },
+    {
+      "epoch": 0.47418965161988874,
+      "grad_norm": 502.811279296875,
+      "learning_rate": 3.193454319162557e-05,
+      "loss": 52.4119,
+      "step": 117370
+    },
+    {
+      "epoch": 0.4742300528852564,
+      "grad_norm": 867.6959228515625,
+      "learning_rate": 3.193118943970338e-05,
+      "loss": 90.0065,
+      "step": 117380
+    },
+    {
+      "epoch": 0.474270454150624,
+      "grad_norm": 1995.4403076171875,
+      "learning_rate": 3.192783555265408e-05,
+      "loss": 68.9316,
+      "step": 117390
+    },
+    {
+      "epoch": 0.4743108554159916,
+      "grad_norm": 522.066162109375,
+      "learning_rate": 3.192448153054306e-05,
+      "loss": 41.8321,
+      "step": 117400
+    },
+    {
+      "epoch": 0.47435125668135925,
+      "grad_norm": 835.8606567382812,
+      "learning_rate": 3.1921127373435714e-05,
+      "loss": 67.2701,
+      "step": 117410
+    },
+    {
+      "epoch": 0.4743916579467269,
+      "grad_norm": 628.2836303710938,
+      "learning_rate": 3.191777308139742e-05,
+      "loss": 57.3309,
+      "step": 117420
+    },
+    {
+      "epoch": 0.4744320592120945,
+      "grad_norm": 582.8585205078125,
+      "learning_rate": 3.1914418654493586e-05,
+      "loss": 79.4639,
+      "step": 117430
+    },
+    {
+      "epoch": 0.47447246047746217,
+      "grad_norm": 452.00830078125,
+      "learning_rate": 3.191106409278959e-05,
+      "loss": 57.4894,
+      "step": 117440
+    },
+    {
+      "epoch": 0.4745128617428298,
+      "grad_norm": 867.4202270507812,
+      "learning_rate": 3.1907709396350844e-05,
+      "loss": 54.852,
+      "step": 117450
+    },
+    {
+      "epoch": 0.4745532630081974,
+      "grad_norm": 375.7869567871094,
+      "learning_rate": 3.190435456524275e-05,
+      "loss": 42.05,
+      "step": 117460
+    },
+    {
+      "epoch": 0.47459366427356503,
+      "grad_norm": 453.1134033203125,
+      "learning_rate": 3.190099959953071e-05,
+      "loss": 34.9451,
+      "step": 117470
+    },
+    {
+      "epoch": 0.47463406553893267,
+      "grad_norm": 614.6144409179688,
+      "learning_rate": 3.189764449928012e-05,
+      "loss": 86.9492,
+      "step": 117480
+    },
+    {
+      "epoch": 0.4746744668043003,
+      "grad_norm": 673.782470703125,
+      "learning_rate": 3.1894289264556417e-05,
+      "loss": 69.0601,
+      "step": 117490
+    },
+    {
+      "epoch": 0.47471486806966795,
+      "grad_norm": 541.9735107421875,
+      "learning_rate": 3.1890933895424976e-05,
+      "loss": 46.3959,
+      "step": 117500
+    },
+    {
+      "epoch": 0.4747552693350356,
+      "grad_norm": 379.82891845703125,
+      "learning_rate": 3.188757839195125e-05,
+      "loss": 44.2336,
+      "step": 117510
+    },
+    {
+      "epoch": 0.47479567060040323,
+      "grad_norm": 680.2620849609375,
+      "learning_rate": 3.1884222754200625e-05,
+      "loss": 54.9478,
+      "step": 117520
+    },
+    {
+      "epoch": 0.4748360718657708,
+      "grad_norm": 1045.8995361328125,
+      "learning_rate": 3.188086698223853e-05,
+      "loss": 62.2926,
+      "step": 117530
+    },
+    {
+      "epoch": 0.47487647313113845,
+      "grad_norm": 1066.854248046875,
+      "learning_rate": 3.1877511076130404e-05,
+      "loss": 52.8988,
+      "step": 117540
+    },
+    {
+      "epoch": 0.4749168743965061,
+      "grad_norm": 1133.3275146484375,
+      "learning_rate": 3.187415503594166e-05,
+      "loss": 50.2685,
+      "step": 117550
+    },
+    {
+      "epoch": 0.47495727566187373,
+      "grad_norm": 561.2869873046875,
+      "learning_rate": 3.1870798861737705e-05,
+      "loss": 71.2999,
+      "step": 117560
+    },
+    {
+      "epoch": 0.4749976769272414,
+      "grad_norm": 905.6627197265625,
+      "learning_rate": 3.1867442553584e-05,
+      "loss": 35.9215,
+      "step": 117570
+    },
+    {
+      "epoch": 0.475038078192609,
+      "grad_norm": 2765.510009765625,
+      "learning_rate": 3.186408611154597e-05,
+      "loss": 62.86,
+      "step": 117580
+    },
+    {
+      "epoch": 0.4750784794579766,
+      "grad_norm": 3249.31298828125,
+      "learning_rate": 3.186072953568905e-05,
+      "loss": 89.4793,
+      "step": 117590
+    },
+    {
+      "epoch": 0.47511888072334424,
+      "grad_norm": 1029.83642578125,
+      "learning_rate": 3.185737282607867e-05,
+      "loss": 52.9404,
+      "step": 117600
+    },
+    {
+      "epoch": 0.4751592819887119,
+      "grad_norm": 638.6604614257812,
+      "learning_rate": 3.1854015982780275e-05,
+      "loss": 70.1836,
+      "step": 117610
+    },
+    {
+      "epoch": 0.4751996832540795,
+      "grad_norm": 866.2382202148438,
+      "learning_rate": 3.185065900585931e-05,
+      "loss": 60.6042,
+      "step": 117620
+    },
+    {
+      "epoch": 0.47524008451944716,
+      "grad_norm": 967.8104858398438,
+      "learning_rate": 3.184730189538122e-05,
+      "loss": 67.1052,
+      "step": 117630
+    },
+    {
+      "epoch": 0.4752804857848148,
+      "grad_norm": 990.4212646484375,
+      "learning_rate": 3.1843944651411456e-05,
+      "loss": 44.4866,
+      "step": 117640
+    },
+    {
+      "epoch": 0.47532088705018244,
+      "grad_norm": 528.0682983398438,
+      "learning_rate": 3.184058727401546e-05,
+      "loss": 51.6501,
+      "step": 117650
+    },
+    {
+      "epoch": 0.47536128831555,
+      "grad_norm": 936.557373046875,
+      "learning_rate": 3.1837229763258705e-05,
+      "loss": 97.2483,
+      "step": 117660
+    },
+    {
+      "epoch": 0.47540168958091766,
+      "grad_norm": 870.8411865234375,
+      "learning_rate": 3.183387211920663e-05,
+      "loss": 81.8832,
+      "step": 117670
+    },
+    {
+      "epoch": 0.4754420908462853,
+      "grad_norm": 628.0894165039062,
+      "learning_rate": 3.183051434192471e-05,
+      "loss": 44.0012,
+      "step": 117680
+    },
+    {
+      "epoch": 0.47548249211165294,
+      "grad_norm": 423.3321228027344,
+      "learning_rate": 3.1827156431478386e-05,
+      "loss": 77.5057,
+      "step": 117690
+    },
+    {
+      "epoch": 0.4755228933770206,
+      "grad_norm": 483.94091796875,
+      "learning_rate": 3.1823798387933134e-05,
+      "loss": 39.6376,
+      "step": 117700
+    },
+    {
+      "epoch": 0.4755632946423882,
+      "grad_norm": 791.345703125,
+      "learning_rate": 3.182044021135442e-05,
+      "loss": 55.5527,
+      "step": 117710
+    },
+    {
+      "epoch": 0.4756036959077558,
+      "grad_norm": 498.8603210449219,
+      "learning_rate": 3.181708190180771e-05,
+      "loss": 41.6617,
+      "step": 117720
+    },
+    {
+      "epoch": 0.47564409717312345,
+      "grad_norm": 771.0953979492188,
+      "learning_rate": 3.181372345935848e-05,
+      "loss": 48.7883,
+      "step": 117730
+    },
+    {
+      "epoch": 0.4756844984384911,
+      "grad_norm": 661.4583740234375,
+      "learning_rate": 3.1810364884072205e-05,
+      "loss": 72.9009,
+      "step": 117740
+    },
+    {
+      "epoch": 0.47572489970385873,
+      "grad_norm": 632.1604614257812,
+      "learning_rate": 3.180700617601436e-05,
+      "loss": 59.3145,
+      "step": 117750
+    },
+    {
+      "epoch": 0.47576530096922637,
+      "grad_norm": 806.9335327148438,
+      "learning_rate": 3.180364733525043e-05,
+      "loss": 41.7695,
+      "step": 117760
+    },
+    {
+      "epoch": 0.475805702234594,
+      "grad_norm": 776.6650390625,
+      "learning_rate": 3.1800288361845883e-05,
+      "loss": 62.7163,
+      "step": 117770
+    },
+    {
+      "epoch": 0.4758461034999616,
+      "grad_norm": 601.4182739257812,
+      "learning_rate": 3.179692925586622e-05,
+      "loss": 54.4891,
+      "step": 117780
+    },
+    {
+      "epoch": 0.47588650476532923,
+      "grad_norm": 907.4159545898438,
+      "learning_rate": 3.179357001737692e-05,
+      "loss": 54.7433,
+      "step": 117790
+    },
+    {
+      "epoch": 0.4759269060306969,
+      "grad_norm": 792.1842651367188,
+      "learning_rate": 3.179021064644347e-05,
+      "loss": 73.5924,
+      "step": 117800
+    },
+    {
+      "epoch": 0.4759673072960645,
+      "grad_norm": 641.330810546875,
+      "learning_rate": 3.178685114313137e-05,
+      "loss": 35.0172,
+      "step": 117810
+    },
+    {
+      "epoch": 0.47600770856143215,
+      "grad_norm": 272.5650939941406,
+      "learning_rate": 3.178349150750612e-05,
+      "loss": 77.2498,
+      "step": 117820
+    },
+    {
+      "epoch": 0.4760481098267998,
+      "grad_norm": 917.3323974609375,
+      "learning_rate": 3.1780131739633204e-05,
+      "loss": 66.6369,
+      "step": 117830
+    },
+    {
+      "epoch": 0.47608851109216743,
+      "grad_norm": 924.8410034179688,
+      "learning_rate": 3.177677183957813e-05,
+      "loss": 50.9081,
+      "step": 117840
+    },
+    {
+      "epoch": 0.476128912357535,
+      "grad_norm": 539.0849609375,
+      "learning_rate": 3.17734118074064e-05,
+      "loss": 43.3925,
+      "step": 117850
+    },
+    {
+      "epoch": 0.47616931362290266,
+      "grad_norm": 1520.1893310546875,
+      "learning_rate": 3.177005164318353e-05,
+      "loss": 52.2887,
+      "step": 117860
+    },
+    {
+      "epoch": 0.4762097148882703,
+      "grad_norm": 0.0,
+      "learning_rate": 3.1766691346974996e-05,
+      "loss": 31.6033,
+      "step": 117870
+    },
+    {
+      "epoch": 0.47625011615363794,
+      "grad_norm": 0.0,
+      "learning_rate": 3.176333091884635e-05,
+      "loss": 59.7049,
+      "step": 117880
+    },
+    {
+      "epoch": 0.4762905174190056,
+      "grad_norm": 1278.6094970703125,
+      "learning_rate": 3.175997035886307e-05,
+      "loss": 56.4383,
+      "step": 117890
+    },
+    {
+      "epoch": 0.4763309186843732,
+      "grad_norm": 1077.4521484375,
+      "learning_rate": 3.1756609667090696e-05,
+      "loss": 70.2687,
+      "step": 117900
+    },
+    {
+      "epoch": 0.4763713199497408,
+      "grad_norm": 842.562255859375,
+      "learning_rate": 3.175324884359474e-05,
+      "loss": 112.2075,
+      "step": 117910
+    },
+    {
+      "epoch": 0.47641172121510844,
+      "grad_norm": 310.8381042480469,
+      "learning_rate": 3.174988788844072e-05,
+      "loss": 48.3912,
+      "step": 117920
+    },
+    {
+      "epoch": 0.4764521224804761,
+      "grad_norm": 1312.25244140625,
+      "learning_rate": 3.1746526801694156e-05,
+      "loss": 52.8015,
+      "step": 117930
+    },
+    {
+      "epoch": 0.4764925237458437,
+      "grad_norm": 885.5872802734375,
+      "learning_rate": 3.174316558342059e-05,
+      "loss": 78.4678,
+      "step": 117940
+    },
+    {
+      "epoch": 0.47653292501121136,
+      "grad_norm": 629.9498291015625,
+      "learning_rate": 3.173980423368553e-05,
+      "loss": 57.0286,
+      "step": 117950
+    },
+    {
+      "epoch": 0.476573326276579,
+      "grad_norm": 622.9977416992188,
+      "learning_rate": 3.173644275255451e-05,
+      "loss": 61.0304,
+      "step": 117960
+    },
+    {
+      "epoch": 0.47661372754194664,
+      "grad_norm": 772.329833984375,
+      "learning_rate": 3.173308114009308e-05,
+      "loss": 34.8157,
+      "step": 117970
+    },
+    {
+      "epoch": 0.4766541288073142,
+      "grad_norm": 884.4105834960938,
+      "learning_rate": 3.1729719396366765e-05,
+      "loss": 44.0258,
+      "step": 117980
+    },
+    {
+      "epoch": 0.47669453007268187,
+      "grad_norm": 1267.8201904296875,
+      "learning_rate": 3.172635752144111e-05,
+      "loss": 53.9944,
+      "step": 117990
+    },
+    {
+      "epoch": 0.4767349313380495,
+      "grad_norm": 1093.1011962890625,
+      "learning_rate": 3.172299551538164e-05,
+      "loss": 49.2085,
+      "step": 118000
+    },
+    {
+      "epoch": 0.47677533260341715,
+      "grad_norm": 345.0997009277344,
+      "learning_rate": 3.1719633378253924e-05,
+      "loss": 48.0018,
+      "step": 118010
+    },
+    {
+      "epoch": 0.4768157338687848,
+      "grad_norm": 610.406494140625,
+      "learning_rate": 3.171627111012349e-05,
+      "loss": 52.8992,
+      "step": 118020
+    },
+    {
+      "epoch": 0.4768561351341524,
+      "grad_norm": 1354.5128173828125,
+      "learning_rate": 3.1712908711055897e-05,
+      "loss": 72.9081,
+      "step": 118030
+    },
+    {
+      "epoch": 0.47689653639952,
+      "grad_norm": 446.1787109375,
+      "learning_rate": 3.170954618111669e-05,
+      "loss": 63.6848,
+      "step": 118040
+    },
+    {
+      "epoch": 0.47693693766488765,
+      "grad_norm": 704.4627075195312,
+      "learning_rate": 3.170618352037142e-05,
+      "loss": 58.9005,
+      "step": 118050
+    },
+    {
+      "epoch": 0.4769773389302553,
+      "grad_norm": 481.2560119628906,
+      "learning_rate": 3.170282072888566e-05,
+      "loss": 46.0265,
+      "step": 118060
+    },
+    {
+      "epoch": 0.47701774019562293,
+      "grad_norm": 560.6388549804688,
+      "learning_rate": 3.169945780672495e-05,
+      "loss": 42.9609,
+      "step": 118070
+    },
+    {
+      "epoch": 0.47705814146099057,
+      "grad_norm": 2451.45556640625,
+      "learning_rate": 3.169609475395486e-05,
+      "loss": 84.6672,
+      "step": 118080
+    },
+    {
+      "epoch": 0.4770985427263582,
+      "grad_norm": 458.1382141113281,
+      "learning_rate": 3.169273157064097e-05,
+      "loss": 59.9135,
+      "step": 118090
+    },
+    {
+      "epoch": 0.4771389439917258,
+      "grad_norm": 736.26416015625,
+      "learning_rate": 3.168936825684882e-05,
+      "loss": 42.6383,
+      "step": 118100
+    },
+    {
+      "epoch": 0.47717934525709343,
+      "grad_norm": 3540.630126953125,
+      "learning_rate": 3.1686004812644e-05,
+      "loss": 58.3743,
+      "step": 118110
+    },
+    {
+      "epoch": 0.4772197465224611,
+      "grad_norm": 609.9805908203125,
+      "learning_rate": 3.1682641238092064e-05,
+      "loss": 44.9807,
+      "step": 118120
+    },
+    {
+      "epoch": 0.4772601477878287,
+      "grad_norm": 864.6240234375,
+      "learning_rate": 3.16792775332586e-05,
+      "loss": 91.4741,
+      "step": 118130
+    },
+    {
+      "epoch": 0.47730054905319635,
+      "grad_norm": 804.3956298828125,
+      "learning_rate": 3.167591369820918e-05,
+      "loss": 64.7296,
+      "step": 118140
+    },
+    {
+      "epoch": 0.477340950318564,
+      "grad_norm": 365.2998962402344,
+      "learning_rate": 3.1672549733009396e-05,
+      "loss": 50.6288,
+      "step": 118150
+    },
+    {
+      "epoch": 0.47738135158393163,
+      "grad_norm": 402.1658630371094,
+      "learning_rate": 3.166918563772481e-05,
+      "loss": 53.5217,
+      "step": 118160
+    },
+    {
+      "epoch": 0.4774217528492992,
+      "grad_norm": 668.7144165039062,
+      "learning_rate": 3.1665821412421015e-05,
+      "loss": 54.7511,
+      "step": 118170
+    },
+    {
+      "epoch": 0.47746215411466686,
+      "grad_norm": 919.4798583984375,
+      "learning_rate": 3.1662457057163604e-05,
+      "loss": 47.6506,
+      "step": 118180
+    },
+    {
+      "epoch": 0.4775025553800345,
+      "grad_norm": 1031.98193359375,
+      "learning_rate": 3.165909257201816e-05,
+      "loss": 71.545,
+      "step": 118190
+    },
+    {
+      "epoch": 0.47754295664540214,
+      "grad_norm": 390.0589294433594,
+      "learning_rate": 3.1655727957050285e-05,
+      "loss": 49.304,
+      "step": 118200
+    },
+    {
+      "epoch": 0.4775833579107698,
+      "grad_norm": 1422.1484375,
+      "learning_rate": 3.165236321232557e-05,
+      "loss": 52.9677,
+      "step": 118210
+    },
+    {
+      "epoch": 0.4776237591761374,
+      "grad_norm": 846.0404052734375,
+      "learning_rate": 3.1648998337909594e-05,
+      "loss": 62.2265,
+      "step": 118220
+    },
+    {
+      "epoch": 0.477664160441505,
+      "grad_norm": 408.1695556640625,
+      "learning_rate": 3.164563333386798e-05,
+      "loss": 67.1796,
+      "step": 118230
+    },
+    {
+      "epoch": 0.47770456170687264,
+      "grad_norm": 256.5215759277344,
+      "learning_rate": 3.1642268200266317e-05,
+      "loss": 40.5587,
+      "step": 118240
+    },
+    {
+      "epoch": 0.4777449629722403,
+      "grad_norm": 1751.73486328125,
+      "learning_rate": 3.163890293717022e-05,
+      "loss": 52.8397,
+      "step": 118250
+    },
+    {
+      "epoch": 0.4777853642376079,
+      "grad_norm": 564.2115478515625,
+      "learning_rate": 3.1635537544645296e-05,
+      "loss": 67.6077,
+      "step": 118260
+    },
+    {
+      "epoch": 0.47782576550297556,
+      "grad_norm": 1232.5323486328125,
+      "learning_rate": 3.163217202275715e-05,
+      "loss": 49.1063,
+      "step": 118270
+    },
+    {
+      "epoch": 0.4778661667683432,
+      "grad_norm": 3210.296142578125,
+      "learning_rate": 3.162880637157139e-05,
+      "loss": 69.347,
+      "step": 118280
+    },
+    {
+      "epoch": 0.47790656803371084,
+      "grad_norm": 674.1715087890625,
+      "learning_rate": 3.1625440591153645e-05,
+      "loss": 63.0254,
+      "step": 118290
+    },
+    {
+      "epoch": 0.47794696929907843,
+      "grad_norm": 369.24017333984375,
+      "learning_rate": 3.162207468156952e-05,
+      "loss": 44.4211,
+      "step": 118300
+    },
+    {
+      "epoch": 0.47798737056444607,
+      "grad_norm": 634.0050659179688,
+      "learning_rate": 3.161870864288464e-05,
+      "loss": 58.934,
+      "step": 118310
+    },
+    {
+      "epoch": 0.4780277718298137,
+      "grad_norm": 1877.2578125,
+      "learning_rate": 3.1615342475164636e-05,
+      "loss": 62.1043,
+      "step": 118320
+    },
+    {
+      "epoch": 0.47806817309518135,
+      "grad_norm": 550.0879516601562,
+      "learning_rate": 3.161197617847511e-05,
+      "loss": 41.5093,
+      "step": 118330
+    },
+    {
+      "epoch": 0.478108574360549,
+      "grad_norm": 544.1702270507812,
+      "learning_rate": 3.160860975288171e-05,
+      "loss": 59.2241,
+      "step": 118340
+    },
+    {
+      "epoch": 0.47814897562591663,
+      "grad_norm": 316.4408264160156,
+      "learning_rate": 3.1605243198450066e-05,
+      "loss": 60.8573,
+      "step": 118350
+    },
+    {
+      "epoch": 0.4781893768912842,
+      "grad_norm": 412.7470703125,
+      "learning_rate": 3.16018765152458e-05,
+      "loss": 47.1499,
+      "step": 118360
+    },
+    {
+      "epoch": 0.47822977815665185,
+      "grad_norm": 998.5579833984375,
+      "learning_rate": 3.159850970333456e-05,
+      "loss": 62.5579,
+      "step": 118370
+    },
+    {
+      "epoch": 0.4782701794220195,
+      "grad_norm": 425.9685974121094,
+      "learning_rate": 3.159514276278197e-05,
+      "loss": 55.9611,
+      "step": 118380
+    },
+    {
+      "epoch": 0.47831058068738713,
+      "grad_norm": 393.94635009765625,
+      "learning_rate": 3.1591775693653674e-05,
+      "loss": 50.0992,
+      "step": 118390
+    },
+    {
+      "epoch": 0.4783509819527548,
+      "grad_norm": 1880.3455810546875,
+      "learning_rate": 3.158840849601532e-05,
+      "loss": 65.1443,
+      "step": 118400
+    },
+    {
+      "epoch": 0.4783913832181224,
+      "grad_norm": 503.0107421875,
+      "learning_rate": 3.1585041169932545e-05,
+      "loss": 44.3258,
+      "step": 118410
+    },
+    {
+      "epoch": 0.47843178448349,
+      "grad_norm": 0.0,
+      "learning_rate": 3.1581673715471006e-05,
+      "loss": 53.8743,
+      "step": 118420
+    },
+    {
+      "epoch": 0.47847218574885764,
+      "grad_norm": 608.0613403320312,
+      "learning_rate": 3.157830613269635e-05,
+      "loss": 54.61,
+      "step": 118430
+    },
+    {
+      "epoch": 0.4785125870142253,
+      "grad_norm": 427.8965148925781,
+      "learning_rate": 3.157493842167423e-05,
+      "loss": 63.0965,
+      "step": 118440
+    },
+    {
+      "epoch": 0.4785529882795929,
+      "grad_norm": 507.14190673828125,
+      "learning_rate": 3.15715705824703e-05,
+      "loss": 50.7432,
+      "step": 118450
+    },
+    {
+      "epoch": 0.47859338954496056,
+      "grad_norm": 983.0687255859375,
+      "learning_rate": 3.156820261515022e-05,
+      "loss": 87.0616,
+      "step": 118460
+    },
+    {
+      "epoch": 0.4786337908103282,
+      "grad_norm": 1234.0418701171875,
+      "learning_rate": 3.1564834519779647e-05,
+      "loss": 48.0925,
+      "step": 118470
+    },
+    {
+      "epoch": 0.47867419207569584,
+      "grad_norm": 935.4703369140625,
+      "learning_rate": 3.156146629642425e-05,
+      "loss": 53.1847,
+      "step": 118480
+    },
+    {
+      "epoch": 0.4787145933410634,
+      "grad_norm": 820.0416259765625,
+      "learning_rate": 3.155809794514968e-05,
+      "loss": 77.3875,
+      "step": 118490
+    },
+    {
+      "epoch": 0.47875499460643106,
+      "grad_norm": 877.6055297851562,
+      "learning_rate": 3.155472946602162e-05,
+      "loss": 59.0853,
+      "step": 118500
+    },
+    {
+      "epoch": 0.4787953958717987,
+      "grad_norm": 979.28662109375,
+      "learning_rate": 3.155136085910573e-05,
+      "loss": 60.1543,
+      "step": 118510
+    },
+    {
+      "epoch": 0.47883579713716634,
+      "grad_norm": 702.0006713867188,
+      "learning_rate": 3.15479921244677e-05,
+      "loss": 44.0273,
+      "step": 118520
+    },
+    {
+      "epoch": 0.478876198402534,
+      "grad_norm": 1405.5360107421875,
+      "learning_rate": 3.1544623262173176e-05,
+      "loss": 63.5543,
+      "step": 118530
+    },
+    {
+      "epoch": 0.4789165996679016,
+      "grad_norm": 971.0574340820312,
+      "learning_rate": 3.1541254272287865e-05,
+      "loss": 55.8479,
+      "step": 118540
+    },
+    {
+      "epoch": 0.4789570009332692,
+      "grad_norm": 1137.547119140625,
+      "learning_rate": 3.153788515487742e-05,
+      "loss": 66.1714,
+      "step": 118550
+    },
+    {
+      "epoch": 0.47899740219863685,
+      "grad_norm": 813.9231567382812,
+      "learning_rate": 3.153451591000756e-05,
+      "loss": 102.5478,
+      "step": 118560
+    },
+    {
+      "epoch": 0.4790378034640045,
+      "grad_norm": 870.7437133789062,
+      "learning_rate": 3.153114653774393e-05,
+      "loss": 81.2061,
+      "step": 118570
+    },
+    {
+      "epoch": 0.4790782047293721,
+      "grad_norm": 984.029296875,
+      "learning_rate": 3.152777703815223e-05,
+      "loss": 64.9325,
+      "step": 118580
+    },
+    {
+      "epoch": 0.47911860599473977,
+      "grad_norm": 853.7675170898438,
+      "learning_rate": 3.152440741129817e-05,
+      "loss": 69.4327,
+      "step": 118590
+    },
+    {
+      "epoch": 0.4791590072601074,
+      "grad_norm": 1823.9547119140625,
+      "learning_rate": 3.152103765724743e-05,
+      "loss": 51.6635,
+      "step": 118600
+    },
+    {
+      "epoch": 0.479199408525475,
+      "grad_norm": 1002.6013793945312,
+      "learning_rate": 3.1517667776065696e-05,
+      "loss": 53.4132,
+      "step": 118610
+    },
+    {
+      "epoch": 0.47923980979084263,
+      "grad_norm": 1199.445556640625,
+      "learning_rate": 3.151429776781868e-05,
+      "loss": 57.5089,
+      "step": 118620
+    },
+    {
+      "epoch": 0.47928021105621027,
+      "grad_norm": 545.897705078125,
+      "learning_rate": 3.151092763257206e-05,
+      "loss": 66.9297,
+      "step": 118630
+    },
+    {
+      "epoch": 0.4793206123215779,
+      "grad_norm": 1726.234375,
+      "learning_rate": 3.150755737039157e-05,
+      "loss": 53.6566,
+      "step": 118640
+    },
+    {
+      "epoch": 0.47936101358694555,
+      "grad_norm": 719.9454956054688,
+      "learning_rate": 3.150418698134289e-05,
+      "loss": 48.0167,
+      "step": 118650
+    },
+    {
+      "epoch": 0.4794014148523132,
+      "grad_norm": 864.4921264648438,
+      "learning_rate": 3.150081646549174e-05,
+      "loss": 84.6051,
+      "step": 118660
+    },
+    {
+      "epoch": 0.47944181611768083,
+      "grad_norm": 824.414306640625,
+      "learning_rate": 3.149744582290383e-05,
+      "loss": 53.0983,
+      "step": 118670
+    },
+    {
+      "epoch": 0.4794822173830484,
+      "grad_norm": 471.1961975097656,
+      "learning_rate": 3.149407505364486e-05,
+      "loss": 69.1806,
+      "step": 118680
+    },
+    {
+      "epoch": 0.47952261864841605,
+      "grad_norm": 684.9292602539062,
+      "learning_rate": 3.149070415778056e-05,
+      "loss": 43.8025,
+      "step": 118690
+    },
+    {
+      "epoch": 0.4795630199137837,
+      "grad_norm": 507.1294860839844,
+      "learning_rate": 3.148733313537664e-05,
+      "loss": 53.6068,
+      "step": 118700
+    },
+    {
+      "epoch": 0.47960342117915133,
+      "grad_norm": 1452.51953125,
+      "learning_rate": 3.148396198649882e-05,
+      "loss": 62.2304,
+      "step": 118710
+    },
+    {
+      "epoch": 0.479643822444519,
+      "grad_norm": 568.4696655273438,
+      "learning_rate": 3.148059071121282e-05,
+      "loss": 69.3875,
+      "step": 118720
+    },
+    {
+      "epoch": 0.4796842237098866,
+      "grad_norm": 527.501953125,
+      "learning_rate": 3.147721930958437e-05,
+      "loss": 53.4141,
+      "step": 118730
+    },
+    {
+      "epoch": 0.4797246249752542,
+      "grad_norm": 1219.31640625,
+      "learning_rate": 3.14738477816792e-05,
+      "loss": 58.4269,
+      "step": 118740
+    },
+    {
+      "epoch": 0.47976502624062184,
+      "grad_norm": 1039.8255615234375,
+      "learning_rate": 3.147047612756302e-05,
+      "loss": 39.0244,
+      "step": 118750
+    },
+    {
+      "epoch": 0.4798054275059895,
+      "grad_norm": 1056.8941650390625,
+      "learning_rate": 3.146710434730159e-05,
+      "loss": 54.2339,
+      "step": 118760
+    },
+    {
+      "epoch": 0.4798458287713571,
+      "grad_norm": 756.7150268554688,
+      "learning_rate": 3.1463732440960625e-05,
+      "loss": 53.7804,
+      "step": 118770
+    },
+    {
+      "epoch": 0.47988623003672476,
+      "grad_norm": 990.2318725585938,
+      "learning_rate": 3.1460360408605866e-05,
+      "loss": 45.0402,
+      "step": 118780
+    },
+    {
+      "epoch": 0.4799266313020924,
+      "grad_norm": 782.805908203125,
+      "learning_rate": 3.145698825030307e-05,
+      "loss": 52.3285,
+      "step": 118790
+    },
+    {
+      "epoch": 0.47996703256746004,
+      "grad_norm": 781.7633666992188,
+      "learning_rate": 3.145361596611795e-05,
+      "loss": 53.3584,
+      "step": 118800
+    },
+    {
+      "epoch": 0.4800074338328276,
+      "grad_norm": 429.9420166015625,
+      "learning_rate": 3.1450243556116266e-05,
+      "loss": 61.5989,
+      "step": 118810
+    },
+    {
+      "epoch": 0.48004783509819526,
+      "grad_norm": 1086.17822265625,
+      "learning_rate": 3.144687102036376e-05,
+      "loss": 44.8307,
+      "step": 118820
+    },
+    {
+      "epoch": 0.4800882363635629,
+      "grad_norm": 858.8822631835938,
+      "learning_rate": 3.1443498358926186e-05,
+      "loss": 63.265,
+      "step": 118830
+    },
+    {
+      "epoch": 0.48012863762893054,
+      "grad_norm": 530.9268188476562,
+      "learning_rate": 3.1440125571869306e-05,
+      "loss": 45.8561,
+      "step": 118840
+    },
+    {
+      "epoch": 0.4801690388942982,
+      "grad_norm": 426.1288146972656,
+      "learning_rate": 3.143675265925885e-05,
+      "loss": 52.3246,
+      "step": 118850
+    },
+    {
+      "epoch": 0.4802094401596658,
+      "grad_norm": 374.76849365234375,
+      "learning_rate": 3.1433379621160586e-05,
+      "loss": 44.3146,
+      "step": 118860
+    },
+    {
+      "epoch": 0.4802498414250334,
+      "grad_norm": 727.5894165039062,
+      "learning_rate": 3.143000645764028e-05,
+      "loss": 45.2602,
+      "step": 118870
+    },
+    {
+      "epoch": 0.48029024269040105,
+      "grad_norm": 914.193603515625,
+      "learning_rate": 3.142663316876368e-05,
+      "loss": 67.587,
+      "step": 118880
+    },
+    {
+      "epoch": 0.4803306439557687,
+      "grad_norm": 119.14258575439453,
+      "learning_rate": 3.1423259754596576e-05,
+      "loss": 48.5501,
+      "step": 118890
+    },
+    {
+      "epoch": 0.48037104522113633,
+      "grad_norm": 947.4161376953125,
+      "learning_rate": 3.1419886215204694e-05,
+      "loss": 56.8426,
+      "step": 118900
+    },
+    {
+      "epoch": 0.48041144648650397,
+      "grad_norm": 593.0452270507812,
+      "learning_rate": 3.1416512550653835e-05,
+      "loss": 70.3968,
+      "step": 118910
+    },
+    {
+      "epoch": 0.4804518477518716,
+      "grad_norm": 757.6504516601562,
+      "learning_rate": 3.141313876100976e-05,
+      "loss": 59.1005,
+      "step": 118920
+    },
+    {
+      "epoch": 0.4804922490172392,
+      "grad_norm": 3180.365966796875,
+      "learning_rate": 3.1409764846338245e-05,
+      "loss": 64.8498,
+      "step": 118930
+    },
+    {
+      "epoch": 0.48053265028260683,
+      "grad_norm": 566.63671875,
+      "learning_rate": 3.140639080670507e-05,
+      "loss": 64.9276,
+      "step": 118940
+    },
+    {
+      "epoch": 0.4805730515479745,
+      "grad_norm": 714.4104614257812,
+      "learning_rate": 3.140301664217599e-05,
+      "loss": 59.2069,
+      "step": 118950
+    },
+    {
+      "epoch": 0.4806134528133421,
+      "grad_norm": 575.2021484375,
+      "learning_rate": 3.139964235281682e-05,
+      "loss": 43.4193,
+      "step": 118960
+    },
+    {
+      "epoch": 0.48065385407870975,
+      "grad_norm": 1377.7861328125,
+      "learning_rate": 3.1396267938693316e-05,
+      "loss": 63.2737,
+      "step": 118970
+    },
+    {
+      "epoch": 0.4806942553440774,
+      "grad_norm": 661.6658935546875,
+      "learning_rate": 3.1392893399871295e-05,
+      "loss": 60.5852,
+      "step": 118980
+    },
+    {
+      "epoch": 0.48073465660944503,
+      "grad_norm": 513.0794067382812,
+      "learning_rate": 3.1389518736416507e-05,
+      "loss": 50.5535,
+      "step": 118990
+    },
+    {
+      "epoch": 0.4807750578748126,
+      "grad_norm": 1068.8321533203125,
+      "learning_rate": 3.138614394839476e-05,
+      "loss": 56.0899,
+      "step": 119000
+    },
+    {
+      "epoch": 0.48081545914018026,
+      "grad_norm": 346.4224548339844,
+      "learning_rate": 3.138276903587186e-05,
+      "loss": 62.7875,
+      "step": 119010
+    },
+    {
+      "epoch": 0.4808558604055479,
+      "grad_norm": 591.9736328125,
+      "learning_rate": 3.137939399891359e-05,
+      "loss": 50.5829,
+      "step": 119020
+    },
+    {
+      "epoch": 0.48089626167091554,
+      "grad_norm": 762.9810791015625,
+      "learning_rate": 3.1376018837585747e-05,
+      "loss": 66.904,
+      "step": 119030
+    },
+    {
+      "epoch": 0.4809366629362832,
+      "grad_norm": 700.6834106445312,
+      "learning_rate": 3.137264355195413e-05,
+      "loss": 66.1691,
+      "step": 119040
+    },
+    {
+      "epoch": 0.4809770642016508,
+      "grad_norm": 315.9108581542969,
+      "learning_rate": 3.1369268142084556e-05,
+      "loss": 49.7968,
+      "step": 119050
+    },
+    {
+      "epoch": 0.4810174654670184,
+      "grad_norm": 0.0,
+      "learning_rate": 3.136589260804282e-05,
+      "loss": 47.6848,
+      "step": 119060
+    },
+    {
+      "epoch": 0.48105786673238604,
+      "grad_norm": 395.9146423339844,
+      "learning_rate": 3.1362516949894725e-05,
+      "loss": 57.0576,
+      "step": 119070
+    },
+    {
+      "epoch": 0.4810982679977537,
+      "grad_norm": 476.3524475097656,
+      "learning_rate": 3.135914116770609e-05,
+      "loss": 58.1471,
+      "step": 119080
+    },
+    {
+      "epoch": 0.4811386692631213,
+      "grad_norm": 626.7835083007812,
+      "learning_rate": 3.135576526154272e-05,
+      "loss": 39.7096,
+      "step": 119090
+    },
+    {
+      "epoch": 0.48117907052848896,
+      "grad_norm": 1119.859619140625,
+      "learning_rate": 3.135238923147043e-05,
+      "loss": 87.1097,
+      "step": 119100
+    },
+    {
+      "epoch": 0.4812194717938566,
+      "grad_norm": 804.6915283203125,
+      "learning_rate": 3.1349013077555045e-05,
+      "loss": 36.7657,
+      "step": 119110
+    },
+    {
+      "epoch": 0.48125987305922424,
+      "grad_norm": 769.036865234375,
+      "learning_rate": 3.134563679986238e-05,
+      "loss": 40.638,
+      "step": 119120
+    },
+    {
+      "epoch": 0.4813002743245918,
+      "grad_norm": 191.26966857910156,
+      "learning_rate": 3.134226039845827e-05,
+      "loss": 50.2752,
+      "step": 119130
+    },
+    {
+      "epoch": 0.48134067558995947,
+      "grad_norm": 778.8662109375,
+      "learning_rate": 3.1338883873408516e-05,
+      "loss": 58.7929,
+      "step": 119140
+    },
+    {
+      "epoch": 0.4813810768553271,
+      "grad_norm": 514.7074584960938,
+      "learning_rate": 3.133550722477896e-05,
+      "loss": 66.8902,
+      "step": 119150
+    },
+    {
+      "epoch": 0.48142147812069475,
+      "grad_norm": 355.43096923828125,
+      "learning_rate": 3.133213045263543e-05,
+      "loss": 58.968,
+      "step": 119160
+    },
+    {
+      "epoch": 0.4814618793860624,
+      "grad_norm": 3042.458984375,
+      "learning_rate": 3.132875355704376e-05,
+      "loss": 88.0314,
+      "step": 119170
+    },
+    {
+      "epoch": 0.48150228065143,
+      "grad_norm": 656.7515869140625,
+      "learning_rate": 3.1325376538069776e-05,
+      "loss": 68.5067,
+      "step": 119180
+    },
+    {
+      "epoch": 0.4815426819167976,
+      "grad_norm": 495.6517028808594,
+      "learning_rate": 3.132199939577932e-05,
+      "loss": 73.5789,
+      "step": 119190
+    },
+    {
+      "epoch": 0.48158308318216525,
+      "grad_norm": 1507.7938232421875,
+      "learning_rate": 3.1318622130238236e-05,
+      "loss": 55.6776,
+      "step": 119200
+    },
+    {
+      "epoch": 0.4816234844475329,
+      "grad_norm": 802.6456298828125,
+      "learning_rate": 3.1315244741512356e-05,
+      "loss": 49.8764,
+      "step": 119210
+    },
+    {
+      "epoch": 0.48166388571290053,
+      "grad_norm": 814.6461181640625,
+      "learning_rate": 3.131186722966753e-05,
+      "loss": 51.0948,
+      "step": 119220
+    },
+    {
+      "epoch": 0.48170428697826817,
+      "grad_norm": 826.6341552734375,
+      "learning_rate": 3.1308489594769605e-05,
+      "loss": 54.8559,
+      "step": 119230
+    },
+    {
+      "epoch": 0.4817446882436358,
+      "grad_norm": 536.5158081054688,
+      "learning_rate": 3.1305111836884425e-05,
+      "loss": 49.7287,
+      "step": 119240
+    },
+    {
+      "epoch": 0.4817850895090034,
+      "grad_norm": 1080.1739501953125,
+      "learning_rate": 3.130173395607785e-05,
+      "loss": 96.4672,
+      "step": 119250
+    },
+    {
+      "epoch": 0.48182549077437103,
+      "grad_norm": 875.5479125976562,
+      "learning_rate": 3.129835595241571e-05,
+      "loss": 41.3818,
+      "step": 119260
+    },
+    {
+      "epoch": 0.4818658920397387,
+      "grad_norm": 977.9671020507812,
+      "learning_rate": 3.129497782596389e-05,
+      "loss": 44.1907,
+      "step": 119270
+    },
+    {
+      "epoch": 0.4819062933051063,
+      "grad_norm": 469.0617370605469,
+      "learning_rate": 3.129159957678824e-05,
+      "loss": 50.5765,
+      "step": 119280
+    },
+    {
+      "epoch": 0.48194669457047395,
+      "grad_norm": 851.140869140625,
+      "learning_rate": 3.128822120495462e-05,
+      "loss": 50.6099,
+      "step": 119290
+    },
+    {
+      "epoch": 0.4819870958358416,
+      "grad_norm": 608.3344116210938,
+      "learning_rate": 3.1284842710528876e-05,
+      "loss": 72.0118,
+      "step": 119300
+    },
+    {
+      "epoch": 0.48202749710120923,
+      "grad_norm": 380.56195068359375,
+      "learning_rate": 3.128146409357689e-05,
+      "loss": 83.0865,
+      "step": 119310
+    },
+    {
+      "epoch": 0.4820678983665768,
+      "grad_norm": 601.8255615234375,
+      "learning_rate": 3.127808535416454e-05,
+      "loss": 48.3086,
+      "step": 119320
+    },
+    {
+      "epoch": 0.48210829963194446,
+      "grad_norm": 949.0869140625,
+      "learning_rate": 3.127470649235768e-05,
+      "loss": 94.7472,
+      "step": 119330
+    },
+    {
+      "epoch": 0.4821487008973121,
+      "grad_norm": 398.5122985839844,
+      "learning_rate": 3.1271327508222174e-05,
+      "loss": 73.9752,
+      "step": 119340
+    },
+    {
+      "epoch": 0.48218910216267974,
+      "grad_norm": 683.5873413085938,
+      "learning_rate": 3.126794840182392e-05,
+      "loss": 65.0503,
+      "step": 119350
+    },
+    {
+      "epoch": 0.4822295034280474,
+      "grad_norm": 574.1619873046875,
+      "learning_rate": 3.126456917322878e-05,
+      "loss": 55.5617,
+      "step": 119360
+    },
+    {
+      "epoch": 0.482269904693415,
+      "grad_norm": 918.1862182617188,
+      "learning_rate": 3.1261189822502644e-05,
+      "loss": 48.1121,
+      "step": 119370
+    },
+    {
+      "epoch": 0.4823103059587826,
+      "grad_norm": 1102.67919921875,
+      "learning_rate": 3.125781034971139e-05,
+      "loss": 79.4511,
+      "step": 119380
+    },
+    {
+      "epoch": 0.48235070722415024,
+      "grad_norm": 865.8079223632812,
+      "learning_rate": 3.125443075492089e-05,
+      "loss": 58.5068,
+      "step": 119390
+    },
+    {
+      "epoch": 0.4823911084895179,
+      "grad_norm": 644.5775756835938,
+      "learning_rate": 3.1251051038197055e-05,
+      "loss": 53.9188,
+      "step": 119400
+    },
+    {
+      "epoch": 0.4824315097548855,
+      "grad_norm": 472.7004699707031,
+      "learning_rate": 3.124767119960576e-05,
+      "loss": 46.8185,
+      "step": 119410
+    },
+    {
+      "epoch": 0.48247191102025316,
+      "grad_norm": 988.4273681640625,
+      "learning_rate": 3.1244291239212896e-05,
+      "loss": 64.0862,
+      "step": 119420
+    },
+    {
+      "epoch": 0.4825123122856208,
+      "grad_norm": 320.8127746582031,
+      "learning_rate": 3.124091115708436e-05,
+      "loss": 66.8379,
+      "step": 119430
+    },
+    {
+      "epoch": 0.48255271355098844,
+      "grad_norm": 703.6538696289062,
+      "learning_rate": 3.123753095328604e-05,
+      "loss": 74.7884,
+      "step": 119440
+    },
+    {
+      "epoch": 0.48259311481635603,
+      "grad_norm": 741.8955078125,
+      "learning_rate": 3.123415062788385e-05,
+      "loss": 47.4477,
+      "step": 119450
+    },
+    {
+      "epoch": 0.48263351608172367,
+      "grad_norm": 764.6326904296875,
+      "learning_rate": 3.123077018094369e-05,
+      "loss": 48.5175,
+      "step": 119460
+    },
+    {
+      "epoch": 0.4826739173470913,
+      "grad_norm": 436.60516357421875,
+      "learning_rate": 3.122738961253145e-05,
+      "loss": 66.414,
+      "step": 119470
+    },
+    {
+      "epoch": 0.48271431861245895,
+      "grad_norm": 369.33343505859375,
+      "learning_rate": 3.1224008922713044e-05,
+      "loss": 71.549,
+      "step": 119480
+    },
+    {
+      "epoch": 0.4827547198778266,
+      "grad_norm": 921.5698852539062,
+      "learning_rate": 3.122062811155438e-05,
+      "loss": 50.5309,
+      "step": 119490
+    },
+    {
+      "epoch": 0.48279512114319423,
+      "grad_norm": 597.4305419921875,
+      "learning_rate": 3.121724717912138e-05,
+      "loss": 50.5436,
+      "step": 119500
+    },
+    {
+      "epoch": 0.4828355224085618,
+      "grad_norm": 1035.259765625,
+      "learning_rate": 3.121386612547993e-05,
+      "loss": 31.3353,
+      "step": 119510
+    },
+    {
+      "epoch": 0.48287592367392945,
+      "grad_norm": 1129.0030517578125,
+      "learning_rate": 3.121048495069596e-05,
+      "loss": 74.6693,
+      "step": 119520
+    },
+    {
+      "epoch": 0.4829163249392971,
+      "grad_norm": 1025.2454833984375,
+      "learning_rate": 3.1207103654835394e-05,
+      "loss": 102.4706,
+      "step": 119530
+    },
+    {
+      "epoch": 0.48295672620466473,
+      "grad_norm": 379.8052062988281,
+      "learning_rate": 3.120372223796415e-05,
+      "loss": 49.5203,
+      "step": 119540
+    },
+    {
+      "epoch": 0.4829971274700324,
+      "grad_norm": 607.2550048828125,
+      "learning_rate": 3.120034070014814e-05,
+      "loss": 70.4186,
+      "step": 119550
+    },
+    {
+      "epoch": 0.4830375287354,
+      "grad_norm": 690.1021728515625,
+      "learning_rate": 3.11969590414533e-05,
+      "loss": 43.4509,
+      "step": 119560
+    },
+    {
+      "epoch": 0.4830779300007676,
+      "grad_norm": 1215.8643798828125,
+      "learning_rate": 3.119357726194556e-05,
+      "loss": 42.1141,
+      "step": 119570
+    },
+    {
+      "epoch": 0.48311833126613524,
+      "grad_norm": 259.37872314453125,
+      "learning_rate": 3.119019536169083e-05,
+      "loss": 58.9376,
+      "step": 119580
+    },
+    {
+      "epoch": 0.4831587325315029,
+      "grad_norm": 755.9671020507812,
+      "learning_rate": 3.118681334075506e-05,
+      "loss": 45.3444,
+      "step": 119590
+    },
+    {
+      "epoch": 0.4831991337968705,
+      "grad_norm": 1142.39892578125,
+      "learning_rate": 3.118343119920418e-05,
+      "loss": 58.759,
+      "step": 119600
+    },
+    {
+      "epoch": 0.48323953506223816,
+      "grad_norm": 651.2678833007812,
+      "learning_rate": 3.1180048937104114e-05,
+      "loss": 51.0291,
+      "step": 119610
+    },
+    {
+      "epoch": 0.4832799363276058,
+      "grad_norm": 555.2761840820312,
+      "learning_rate": 3.117666655452083e-05,
+      "loss": 48.9106,
+      "step": 119620
+    },
+    {
+      "epoch": 0.48332033759297344,
+      "grad_norm": 1292.247802734375,
+      "learning_rate": 3.117328405152024e-05,
+      "loss": 67.4017,
+      "step": 119630
+    },
+    {
+      "epoch": 0.483360738858341,
+      "grad_norm": 579.1093139648438,
+      "learning_rate": 3.11699014281683e-05,
+      "loss": 56.0868,
+      "step": 119640
+    },
+    {
+      "epoch": 0.48340114012370866,
+      "grad_norm": 571.940673828125,
+      "learning_rate": 3.116651868453097e-05,
+      "loss": 57.993,
+      "step": 119650
+    },
+    {
+      "epoch": 0.4834415413890763,
+      "grad_norm": 592.5040893554688,
+      "learning_rate": 3.116313582067416e-05,
+      "loss": 51.3045,
+      "step": 119660
+    },
+    {
+      "epoch": 0.48348194265444394,
+      "grad_norm": 820.21337890625,
+      "learning_rate": 3.115975283666386e-05,
+      "loss": 50.0175,
+      "step": 119670
+    },
+    {
+      "epoch": 0.4835223439198116,
+      "grad_norm": 1067.8109130859375,
+      "learning_rate": 3.1156369732566006e-05,
+      "loss": 31.4268,
+      "step": 119680
+    },
+    {
+      "epoch": 0.4835627451851792,
+      "grad_norm": 589.9420776367188,
+      "learning_rate": 3.115298650844655e-05,
+      "loss": 85.4477,
+      "step": 119690
+    },
+    {
+      "epoch": 0.4836031464505468,
+      "grad_norm": 485.7039489746094,
+      "learning_rate": 3.114960316437145e-05,
+      "loss": 53.188,
+      "step": 119700
+    },
+    {
+      "epoch": 0.48364354771591445,
+      "grad_norm": 809.940673828125,
+      "learning_rate": 3.1146219700406674e-05,
+      "loss": 73.8831,
+      "step": 119710
+    },
+    {
+      "epoch": 0.4836839489812821,
+      "grad_norm": 1266.80419921875,
+      "learning_rate": 3.114283611661818e-05,
+      "loss": 65.2104,
+      "step": 119720
+    },
+    {
+      "epoch": 0.4837243502466497,
+      "grad_norm": 1203.2113037109375,
+      "learning_rate": 3.113945241307194e-05,
+      "loss": 64.743,
+      "step": 119730
+    },
+    {
+      "epoch": 0.48376475151201737,
+      "grad_norm": 655.8514404296875,
+      "learning_rate": 3.1136068589833914e-05,
+      "loss": 70.6579,
+      "step": 119740
+    },
+    {
+      "epoch": 0.483805152777385,
+      "grad_norm": 842.0756225585938,
+      "learning_rate": 3.1132684646970064e-05,
+      "loss": 41.7421,
+      "step": 119750
+    },
+    {
+      "epoch": 0.48384555404275265,
+      "grad_norm": 859.2930297851562,
+      "learning_rate": 3.1129300584546375e-05,
+      "loss": 61.2274,
+      "step": 119760
+    },
+    {
+      "epoch": 0.48388595530812023,
+      "grad_norm": 972.3688354492188,
+      "learning_rate": 3.1125916402628814e-05,
+      "loss": 58.9962,
+      "step": 119770
+    },
+    {
+      "epoch": 0.48392635657348787,
+      "grad_norm": 983.5337524414062,
+      "learning_rate": 3.112253210128336e-05,
+      "loss": 57.9547,
+      "step": 119780
+    },
+    {
+      "epoch": 0.4839667578388555,
+      "grad_norm": 997.2120361328125,
+      "learning_rate": 3.111914768057599e-05,
+      "loss": 47.62,
+      "step": 119790
+    },
+    {
+      "epoch": 0.48400715910422315,
+      "grad_norm": 1432.3392333984375,
+      "learning_rate": 3.111576314057268e-05,
+      "loss": 48.3912,
+      "step": 119800
+    },
+    {
+      "epoch": 0.4840475603695908,
+      "grad_norm": 502.978759765625,
+      "learning_rate": 3.1112378481339425e-05,
+      "loss": 60.9147,
+      "step": 119810
+    },
+    {
+      "epoch": 0.48408796163495843,
+      "grad_norm": 881.69580078125,
+      "learning_rate": 3.1108993702942205e-05,
+      "loss": 63.0135,
+      "step": 119820
+    },
+    {
+      "epoch": 0.484128362900326,
+      "grad_norm": 619.3832397460938,
+      "learning_rate": 3.110560880544701e-05,
+      "loss": 55.6975,
+      "step": 119830
+    },
+    {
+      "epoch": 0.48416876416569365,
+      "grad_norm": 933.4210815429688,
+      "learning_rate": 3.1102223788919824e-05,
+      "loss": 45.4596,
+      "step": 119840
+    },
+    {
+      "epoch": 0.4842091654310613,
+      "grad_norm": 1533.591552734375,
+      "learning_rate": 3.1098838653426645e-05,
+      "loss": 62.3054,
+      "step": 119850
+    },
+    {
+      "epoch": 0.48424956669642893,
+      "grad_norm": 815.6768798828125,
+      "learning_rate": 3.1095453399033466e-05,
+      "loss": 53.5172,
+      "step": 119860
+    },
+    {
+      "epoch": 0.4842899679617966,
+      "grad_norm": 1024.9600830078125,
+      "learning_rate": 3.109206802580629e-05,
+      "loss": 63.8872,
+      "step": 119870
+    },
+    {
+      "epoch": 0.4843303692271642,
+      "grad_norm": 0.0,
+      "learning_rate": 3.10886825338111e-05,
+      "loss": 48.7491,
+      "step": 119880
+    },
+    {
+      "epoch": 0.4843707704925318,
+      "grad_norm": 610.7237548828125,
+      "learning_rate": 3.108529692311391e-05,
+      "loss": 59.5073,
+      "step": 119890
+    },
+    {
+      "epoch": 0.48441117175789944,
+      "grad_norm": 999.9380493164062,
+      "learning_rate": 3.108191119378073e-05,
+      "loss": 54.3617,
+      "step": 119900
+    },
+    {
+      "epoch": 0.4844515730232671,
+      "grad_norm": 589.252685546875,
+      "learning_rate": 3.107852534587756e-05,
+      "loss": 70.6535,
+      "step": 119910
+    },
+    {
+      "epoch": 0.4844919742886347,
+      "grad_norm": 1168.3255615234375,
+      "learning_rate": 3.107513937947041e-05,
+      "loss": 65.7924,
+      "step": 119920
+    },
+    {
+      "epoch": 0.48453237555400236,
+      "grad_norm": 452.32415771484375,
+      "learning_rate": 3.107175329462529e-05,
+      "loss": 33.2393,
+      "step": 119930
+    },
+    {
+      "epoch": 0.48457277681937,
+      "grad_norm": 365.3132629394531,
+      "learning_rate": 3.106836709140821e-05,
+      "loss": 61.387,
+      "step": 119940
+    },
+    {
+      "epoch": 0.48461317808473764,
+      "grad_norm": 431.0849914550781,
+      "learning_rate": 3.1064980769885187e-05,
+      "loss": 49.9083,
+      "step": 119950
+    },
+    {
+      "epoch": 0.4846535793501052,
+      "grad_norm": 706.0770874023438,
+      "learning_rate": 3.1061594330122246e-05,
+      "loss": 41.8204,
+      "step": 119960
+    },
+    {
+      "epoch": 0.48469398061547286,
+      "grad_norm": 519.536865234375,
+      "learning_rate": 3.10582077721854e-05,
+      "loss": 39.0427,
+      "step": 119970
+    },
+    {
+      "epoch": 0.4847343818808405,
+      "grad_norm": 1396.991455078125,
+      "learning_rate": 3.1054821096140676e-05,
+      "loss": 70.4659,
+      "step": 119980
+    },
+    {
+      "epoch": 0.48477478314620814,
+      "grad_norm": 1468.1710205078125,
+      "learning_rate": 3.10514343020541e-05,
+      "loss": 91.4469,
+      "step": 119990
+    },
+    {
+      "epoch": 0.4848151844115758,
+      "grad_norm": 483.8679504394531,
+      "learning_rate": 3.104804738999169e-05,
+      "loss": 31.8263,
+      "step": 120000
+    },
+    {
+      "epoch": 0.4848555856769434,
+      "grad_norm": 1103.7783203125,
+      "learning_rate": 3.10446603600195e-05,
+      "loss": 48.717,
+      "step": 120010
+    },
+    {
+      "epoch": 0.484895986942311,
+      "grad_norm": 0.0,
+      "learning_rate": 3.104127321220353e-05,
+      "loss": 44.1308,
+      "step": 120020
+    },
+    {
+      "epoch": 0.48493638820767865,
+      "grad_norm": 2047.26025390625,
+      "learning_rate": 3.1037885946609824e-05,
+      "loss": 61.4208,
+      "step": 120030
+    },
+    {
+      "epoch": 0.4849767894730463,
+      "grad_norm": 2711.773681640625,
+      "learning_rate": 3.103449856330443e-05,
+      "loss": 70.1431,
+      "step": 120040
+    },
+    {
+      "epoch": 0.48501719073841393,
+      "grad_norm": 1064.7301025390625,
+      "learning_rate": 3.1031111062353373e-05,
+      "loss": 59.401,
+      "step": 120050
+    },
+    {
+      "epoch": 0.48505759200378157,
+      "grad_norm": 1243.82763671875,
+      "learning_rate": 3.102772344382271e-05,
+      "loss": 78.9579,
+      "step": 120060
+    },
+    {
+      "epoch": 0.4850979932691492,
+      "grad_norm": 935.9784545898438,
+      "learning_rate": 3.102433570777847e-05,
+      "loss": 56.5204,
+      "step": 120070
+    },
+    {
+      "epoch": 0.48513839453451685,
+      "grad_norm": 583.158935546875,
+      "learning_rate": 3.102094785428671e-05,
+      "loss": 63.3833,
+      "step": 120080
+    },
+    {
+      "epoch": 0.48517879579988443,
+      "grad_norm": 749.0221557617188,
+      "learning_rate": 3.101755988341347e-05,
+      "loss": 48.2709,
+      "step": 120090
+    },
+    {
+      "epoch": 0.4852191970652521,
+      "grad_norm": 876.2115478515625,
+      "learning_rate": 3.101417179522479e-05,
+      "loss": 53.4147,
+      "step": 120100
+    },
+    {
+      "epoch": 0.4852595983306197,
+      "grad_norm": 1196.973388671875,
+      "learning_rate": 3.101078358978675e-05,
+      "loss": 56.0996,
+      "step": 120110
+    },
+    {
+      "epoch": 0.48529999959598735,
+      "grad_norm": 580.7915649414062,
+      "learning_rate": 3.100739526716538e-05,
+      "loss": 58.37,
+      "step": 120120
+    },
+    {
+      "epoch": 0.485340400861355,
+      "grad_norm": 942.7879028320312,
+      "learning_rate": 3.100400682742675e-05,
+      "loss": 64.2288,
+      "step": 120130
+    },
+    {
+      "epoch": 0.48538080212672263,
+      "grad_norm": 627.4921264648438,
+      "learning_rate": 3.100061827063692e-05,
+      "loss": 58.1287,
+      "step": 120140
+    },
+    {
+      "epoch": 0.4854212033920902,
+      "grad_norm": 1478.2410888671875,
+      "learning_rate": 3.0997229596861944e-05,
+      "loss": 91.4508,
+      "step": 120150
+    },
+    {
+      "epoch": 0.48546160465745786,
+      "grad_norm": 1782.7314453125,
+      "learning_rate": 3.099384080616789e-05,
+      "loss": 56.9644,
+      "step": 120160
+    },
+    {
+      "epoch": 0.4855020059228255,
+      "grad_norm": 359.5166320800781,
+      "learning_rate": 3.099045189862081e-05,
+      "loss": 43.2427,
+      "step": 120170
+    },
+    {
+      "epoch": 0.48554240718819314,
+      "grad_norm": 604.357177734375,
+      "learning_rate": 3.0987062874286804e-05,
+      "loss": 43.9985,
+      "step": 120180
+    },
+    {
+      "epoch": 0.4855828084535608,
+      "grad_norm": 536.3892211914062,
+      "learning_rate": 3.098367373323192e-05,
+      "loss": 60.2531,
+      "step": 120190
+    },
+    {
+      "epoch": 0.4856232097189284,
+      "grad_norm": 762.7628784179688,
+      "learning_rate": 3.098028447552224e-05,
+      "loss": 45.4865,
+      "step": 120200
+    },
+    {
+      "epoch": 0.485663610984296,
+      "grad_norm": 621.0496215820312,
+      "learning_rate": 3.097689510122382e-05,
+      "loss": 49.9217,
+      "step": 120210
+    },
+    {
+      "epoch": 0.48570401224966364,
+      "grad_norm": 1028.444580078125,
+      "learning_rate": 3.0973505610402765e-05,
+      "loss": 54.5276,
+      "step": 120220
+    },
+    {
+      "epoch": 0.4857444135150313,
+      "grad_norm": 639.4829711914062,
+      "learning_rate": 3.0970116003125146e-05,
+      "loss": 46.6422,
+      "step": 120230
+    },
+    {
+      "epoch": 0.4857848147803989,
+      "grad_norm": 955.115234375,
+      "learning_rate": 3.0966726279457034e-05,
+      "loss": 59.0818,
+      "step": 120240
+    },
+    {
+      "epoch": 0.48582521604576656,
+      "grad_norm": 1233.369873046875,
+      "learning_rate": 3.0963336439464526e-05,
+      "loss": 56.1891,
+      "step": 120250
+    },
+    {
+      "epoch": 0.4858656173111342,
+      "grad_norm": 3286.7587890625,
+      "learning_rate": 3.09599464832137e-05,
+      "loss": 70.8362,
+      "step": 120260
+    },
+    {
+      "epoch": 0.48590601857650184,
+      "grad_norm": 749.9498291015625,
+      "learning_rate": 3.0956556410770655e-05,
+      "loss": 71.6872,
+      "step": 120270
+    },
+    {
+      "epoch": 0.4859464198418694,
+      "grad_norm": 337.7810974121094,
+      "learning_rate": 3.0953166222201476e-05,
+      "loss": 34.0317,
+      "step": 120280
+    },
+    {
+      "epoch": 0.48598682110723707,
+      "grad_norm": 969.7774047851562,
+      "learning_rate": 3.094977591757224e-05,
+      "loss": 39.0394,
+      "step": 120290
+    },
+    {
+      "epoch": 0.4860272223726047,
+      "grad_norm": 600.6663208007812,
+      "learning_rate": 3.094638549694908e-05,
+      "loss": 72.5898,
+      "step": 120300
+    },
+    {
+      "epoch": 0.48606762363797235,
+      "grad_norm": 388.6318664550781,
+      "learning_rate": 3.0942994960398064e-05,
+      "loss": 58.7893,
+      "step": 120310
+    },
+    {
+      "epoch": 0.48610802490334,
+      "grad_norm": 1114.3638916015625,
+      "learning_rate": 3.09396043079853e-05,
+      "loss": 67.1577,
+      "step": 120320
+    },
+    {
+      "epoch": 0.4861484261687076,
+      "grad_norm": 723.9693603515625,
+      "learning_rate": 3.0936213539776895e-05,
+      "loss": 66.5562,
+      "step": 120330
+    },
+    {
+      "epoch": 0.4861888274340752,
+      "grad_norm": 608.6270141601562,
+      "learning_rate": 3.093282265583895e-05,
+      "loss": 61.2591,
+      "step": 120340
+    },
+    {
+      "epoch": 0.48622922869944285,
+      "grad_norm": 455.3746643066406,
+      "learning_rate": 3.092943165623758e-05,
+      "loss": 57.6604,
+      "step": 120350
+    },
+    {
+      "epoch": 0.4862696299648105,
+      "grad_norm": 494.3013610839844,
+      "learning_rate": 3.092604054103888e-05,
+      "loss": 42.1755,
+      "step": 120360
+    },
+    {
+      "epoch": 0.48631003123017813,
+      "grad_norm": 2547.077392578125,
+      "learning_rate": 3.092264931030897e-05,
+      "loss": 86.2226,
+      "step": 120370
+    },
+    {
+      "epoch": 0.48635043249554577,
+      "grad_norm": 671.0562133789062,
+      "learning_rate": 3.0919257964113964e-05,
+      "loss": 60.703,
+      "step": 120380
+    },
+    {
+      "epoch": 0.4863908337609134,
+      "grad_norm": 471.1807556152344,
+      "learning_rate": 3.0915866502519975e-05,
+      "loss": 55.5357,
+      "step": 120390
+    },
+    {
+      "epoch": 0.48643123502628105,
+      "grad_norm": 656.9327392578125,
+      "learning_rate": 3.091247492559312e-05,
+      "loss": 66.8196,
+      "step": 120400
+    },
+    {
+      "epoch": 0.48647163629164863,
+      "grad_norm": 2174.91064453125,
+      "learning_rate": 3.090908323339952e-05,
+      "loss": 62.3699,
+      "step": 120410
+    },
+    {
+      "epoch": 0.4865120375570163,
+      "grad_norm": 663.234619140625,
+      "learning_rate": 3.090569142600531e-05,
+      "loss": 58.3517,
+      "step": 120420
+    },
+    {
+      "epoch": 0.4865524388223839,
+      "grad_norm": 1044.9910888671875,
+      "learning_rate": 3.09022995034766e-05,
+      "loss": 45.2753,
+      "step": 120430
+    },
+    {
+      "epoch": 0.48659284008775155,
+      "grad_norm": 1259.57666015625,
+      "learning_rate": 3.089890746587953e-05,
+      "loss": 37.8244,
+      "step": 120440
+    },
+    {
+      "epoch": 0.4866332413531192,
+      "grad_norm": 619.0317993164062,
+      "learning_rate": 3.089551531328021e-05,
+      "loss": 52.6852,
+      "step": 120450
+    },
+    {
+      "epoch": 0.48667364261848683,
+      "grad_norm": 547.9066772460938,
+      "learning_rate": 3.0892123045744785e-05,
+      "loss": 35.2198,
+      "step": 120460
+    },
+    {
+      "epoch": 0.4867140438838544,
+      "grad_norm": 538.3446044921875,
+      "learning_rate": 3.08887306633394e-05,
+      "loss": 49.3895,
+      "step": 120470
+    },
+    {
+      "epoch": 0.48675444514922206,
+      "grad_norm": 1102.380859375,
+      "learning_rate": 3.088533816613017e-05,
+      "loss": 78.057,
+      "step": 120480
+    },
+    {
+      "epoch": 0.4867948464145897,
+      "grad_norm": 836.9898071289062,
+      "learning_rate": 3.0881945554183235e-05,
+      "loss": 73.6896,
+      "step": 120490
+    },
+    {
+      "epoch": 0.48683524767995734,
+      "grad_norm": 657.3259887695312,
+      "learning_rate": 3.087855282756475e-05,
+      "loss": 45.1531,
+      "step": 120500
+    },
+    {
+      "epoch": 0.486875648945325,
+      "grad_norm": 440.7289733886719,
+      "learning_rate": 3.087515998634085e-05,
+      "loss": 57.9621,
+      "step": 120510
+    },
+    {
+      "epoch": 0.4869160502106926,
+      "grad_norm": 955.15966796875,
+      "learning_rate": 3.087176703057769e-05,
+      "loss": 56.2972,
+      "step": 120520
+    },
+    {
+      "epoch": 0.4869564514760602,
+      "grad_norm": 293.07012939453125,
+      "learning_rate": 3.08683739603414e-05,
+      "loss": 42.8358,
+      "step": 120530
+    },
+    {
+      "epoch": 0.48699685274142784,
+      "grad_norm": 516.2140502929688,
+      "learning_rate": 3.0864980775698145e-05,
+      "loss": 92.3896,
+      "step": 120540
+    },
+    {
+      "epoch": 0.4870372540067955,
+      "grad_norm": 425.08343505859375,
+      "learning_rate": 3.086158747671406e-05,
+      "loss": 90.2946,
+      "step": 120550
+    },
+    {
+      "epoch": 0.4870776552721631,
+      "grad_norm": 302.85626220703125,
+      "learning_rate": 3.085819406345532e-05,
+      "loss": 73.1443,
+      "step": 120560
+    },
+    {
+      "epoch": 0.48711805653753076,
+      "grad_norm": 910.6437377929688,
+      "learning_rate": 3.0854800535988064e-05,
+      "loss": 54.2145,
+      "step": 120570
+    },
+    {
+      "epoch": 0.4871584578028984,
+      "grad_norm": 1036.3970947265625,
+      "learning_rate": 3.085140689437846e-05,
+      "loss": 49.4283,
+      "step": 120580
+    },
+    {
+      "epoch": 0.48719885906826604,
+      "grad_norm": 1020.8902587890625,
+      "learning_rate": 3.084801313869266e-05,
+      "loss": 50.4306,
+      "step": 120590
+    },
+    {
+      "epoch": 0.48723926033363363,
+      "grad_norm": 683.2181396484375,
+      "learning_rate": 3.0844619268996845e-05,
+      "loss": 52.7355,
+      "step": 120600
+    },
+    {
+      "epoch": 0.48727966159900127,
+      "grad_norm": 712.2384033203125,
+      "learning_rate": 3.084122528535717e-05,
+      "loss": 76.377,
+      "step": 120610
+    },
+    {
+      "epoch": 0.4873200628643689,
+      "grad_norm": 1040.790771484375,
+      "learning_rate": 3.0837831187839784e-05,
+      "loss": 66.5839,
+      "step": 120620
+    },
+    {
+      "epoch": 0.48736046412973655,
+      "grad_norm": 826.3641967773438,
+      "learning_rate": 3.083443697651088e-05,
+      "loss": 81.0511,
+      "step": 120630
+    },
+    {
+      "epoch": 0.4874008653951042,
+      "grad_norm": 1687.8861083984375,
+      "learning_rate": 3.083104265143663e-05,
+      "loss": 73.7149,
+      "step": 120640
+    },
+    {
+      "epoch": 0.48744126666047183,
+      "grad_norm": 700.1195678710938,
+      "learning_rate": 3.08276482126832e-05,
+      "loss": 49.3427,
+      "step": 120650
+    },
+    {
+      "epoch": 0.4874816679258394,
+      "grad_norm": 661.125244140625,
+      "learning_rate": 3.082425366031676e-05,
+      "loss": 37.9126,
+      "step": 120660
+    },
+    {
+      "epoch": 0.48752206919120705,
+      "grad_norm": 773.9281005859375,
+      "learning_rate": 3.08208589944035e-05,
+      "loss": 47.0344,
+      "step": 120670
+    },
+    {
+      "epoch": 0.4875624704565747,
+      "grad_norm": 1276.460205078125,
+      "learning_rate": 3.08174642150096e-05,
+      "loss": 53.2406,
+      "step": 120680
+    },
+    {
+      "epoch": 0.48760287172194233,
+      "grad_norm": 669.7227172851562,
+      "learning_rate": 3.081406932220123e-05,
+      "loss": 39.3701,
+      "step": 120690
+    },
+    {
+      "epoch": 0.48764327298731,
+      "grad_norm": 2131.654541015625,
+      "learning_rate": 3.08106743160446e-05,
+      "loss": 64.2498,
+      "step": 120700
+    },
+    {
+      "epoch": 0.4876836742526776,
+      "grad_norm": 621.3330078125,
+      "learning_rate": 3.0807279196605876e-05,
+      "loss": 64.5872,
+      "step": 120710
+    },
+    {
+      "epoch": 0.48772407551804525,
+      "grad_norm": 565.8031616210938,
+      "learning_rate": 3.0803883963951255e-05,
+      "loss": 59.7642,
+      "step": 120720
+    },
+    {
+      "epoch": 0.48776447678341284,
+      "grad_norm": 936.9571533203125,
+      "learning_rate": 3.080048861814693e-05,
+      "loss": 34.2556,
+      "step": 120730
+    },
+    {
+      "epoch": 0.4878048780487805,
+      "grad_norm": 361.34063720703125,
+      "learning_rate": 3.0797093159259085e-05,
+      "loss": 52.7916,
+      "step": 120740
+    },
+    {
+      "epoch": 0.4878452793141481,
+      "grad_norm": 325.8556213378906,
+      "learning_rate": 3.079369758735393e-05,
+      "loss": 41.6967,
+      "step": 120750
+    },
+    {
+      "epoch": 0.48788568057951576,
+      "grad_norm": 522.2239990234375,
+      "learning_rate": 3.0790301902497666e-05,
+      "loss": 44.1217,
+      "step": 120760
+    },
+    {
+      "epoch": 0.4879260818448834,
+      "grad_norm": 1932.948486328125,
+      "learning_rate": 3.078690610475647e-05,
+      "loss": 70.4253,
+      "step": 120770
+    },
+    {
+      "epoch": 0.48796648311025104,
+      "grad_norm": 1131.1263427734375,
+      "learning_rate": 3.0783510194196576e-05,
+      "loss": 101.964,
+      "step": 120780
+    },
+    {
+      "epoch": 0.4880068843756186,
+      "grad_norm": 1616.0198974609375,
+      "learning_rate": 3.078011417088416e-05,
+      "loss": 55.159,
+      "step": 120790
+    },
+    {
+      "epoch": 0.48804728564098626,
+      "grad_norm": 942.608642578125,
+      "learning_rate": 3.0776718034885454e-05,
+      "loss": 47.1413,
+      "step": 120800
+    },
+    {
+      "epoch": 0.4880876869063539,
+      "grad_norm": 503.5368347167969,
+      "learning_rate": 3.0773321786266644e-05,
+      "loss": 36.5475,
+      "step": 120810
+    },
+    {
+      "epoch": 0.48812808817172154,
+      "grad_norm": 844.5652465820312,
+      "learning_rate": 3.076992542509396e-05,
+      "loss": 52.4938,
+      "step": 120820
+    },
+    {
+      "epoch": 0.4881684894370892,
+      "grad_norm": 856.3523559570312,
+      "learning_rate": 3.07665289514336e-05,
+      "loss": 83.7883,
+      "step": 120830
+    },
+    {
+      "epoch": 0.4882088907024568,
+      "grad_norm": 400.933349609375,
+      "learning_rate": 3.07631323653518e-05,
+      "loss": 55.8697,
+      "step": 120840
+    },
+    {
+      "epoch": 0.4882492919678244,
+      "grad_norm": 604.0474853515625,
+      "learning_rate": 3.075973566691477e-05,
+      "loss": 72.4179,
+      "step": 120850
+    },
+    {
+      "epoch": 0.48828969323319205,
+      "grad_norm": 584.9152221679688,
+      "learning_rate": 3.0756338856188716e-05,
+      "loss": 59.9413,
+      "step": 120860
+    },
+    {
+      "epoch": 0.4883300944985597,
+      "grad_norm": 734.94287109375,
+      "learning_rate": 3.075294193323988e-05,
+      "loss": 81.9922,
+      "step": 120870
+    },
+    {
+      "epoch": 0.4883704957639273,
+      "grad_norm": 775.8441162109375,
+      "learning_rate": 3.074954489813449e-05,
+      "loss": 59.0604,
+      "step": 120880
+    },
+    {
+      "epoch": 0.48841089702929497,
+      "grad_norm": 464.788330078125,
+      "learning_rate": 3.074614775093874e-05,
+      "loss": 77.2058,
+      "step": 120890
+    },
+    {
+      "epoch": 0.4884512982946626,
+      "grad_norm": 250.41958618164062,
+      "learning_rate": 3.074275049171889e-05,
+      "loss": 40.2773,
+      "step": 120900
+    },
+    {
+      "epoch": 0.48849169956003025,
+      "grad_norm": 608.3348388671875,
+      "learning_rate": 3.073935312054117e-05,
+      "loss": 34.6437,
+      "step": 120910
+    },
+    {
+      "epoch": 0.48853210082539783,
+      "grad_norm": 1844.59033203125,
+      "learning_rate": 3.0735955637471794e-05,
+      "loss": 71.1501,
+      "step": 120920
+    },
+    {
+      "epoch": 0.48857250209076547,
+      "grad_norm": 385.8973388671875,
+      "learning_rate": 3.073255804257702e-05,
+      "loss": 56.5991,
+      "step": 120930
+    },
+    {
+      "epoch": 0.4886129033561331,
+      "grad_norm": 503.91558837890625,
+      "learning_rate": 3.072916033592307e-05,
+      "loss": 62.6247,
+      "step": 120940
+    },
+    {
+      "epoch": 0.48865330462150075,
+      "grad_norm": 590.6072387695312,
+      "learning_rate": 3.0725762517576195e-05,
+      "loss": 39.7041,
+      "step": 120950
+    },
+    {
+      "epoch": 0.4886937058868684,
+      "grad_norm": 369.4613952636719,
+      "learning_rate": 3.072236458760262e-05,
+      "loss": 37.5118,
+      "step": 120960
+    },
+    {
+      "epoch": 0.48873410715223603,
+      "grad_norm": 743.9142456054688,
+      "learning_rate": 3.071896654606862e-05,
+      "loss": 48.3398,
+      "step": 120970
+    },
+    {
+      "epoch": 0.4887745084176036,
+      "grad_norm": 611.325439453125,
+      "learning_rate": 3.0715568393040405e-05,
+      "loss": 82.7277,
+      "step": 120980
+    },
+    {
+      "epoch": 0.48881490968297125,
+      "grad_norm": 1614.0037841796875,
+      "learning_rate": 3.071217012858425e-05,
+      "loss": 102.5452,
+      "step": 120990
+    },
+    {
+      "epoch": 0.4888553109483389,
+      "grad_norm": 984.4487915039062,
+      "learning_rate": 3.0708771752766394e-05,
+      "loss": 56.9584,
+      "step": 121000
+    },
+    {
+      "epoch": 0.48889571221370653,
+      "grad_norm": 1377.5753173828125,
+      "learning_rate": 3.07053732656531e-05,
+      "loss": 75.7293,
+      "step": 121010
+    },
+    {
+      "epoch": 0.4889361134790742,
+      "grad_norm": 438.0140686035156,
+      "learning_rate": 3.070197466731061e-05,
+      "loss": 43.8571,
+      "step": 121020
+    },
+    {
+      "epoch": 0.4889765147444418,
+      "grad_norm": 710.8450317382812,
+      "learning_rate": 3.069857595780519e-05,
+      "loss": 66.9329,
+      "step": 121030
+    },
+    {
+      "epoch": 0.48901691600980945,
+      "grad_norm": 762.4956665039062,
+      "learning_rate": 3.06951771372031e-05,
+      "loss": 52.6469,
+      "step": 121040
+    },
+    {
+      "epoch": 0.48905731727517704,
+      "grad_norm": 732.955322265625,
+      "learning_rate": 3.06917782055706e-05,
+      "loss": 84.5691,
+      "step": 121050
+    },
+    {
+      "epoch": 0.4890977185405447,
+      "grad_norm": 899.1510009765625,
+      "learning_rate": 3.0688379162973955e-05,
+      "loss": 78.2393,
+      "step": 121060
+    },
+    {
+      "epoch": 0.4891381198059123,
+      "grad_norm": 798.7470092773438,
+      "learning_rate": 3.0684980009479424e-05,
+      "loss": 58.7748,
+      "step": 121070
+    },
+    {
+      "epoch": 0.48917852107127996,
+      "grad_norm": 897.6027221679688,
+      "learning_rate": 3.068158074515328e-05,
+      "loss": 52.4273,
+      "step": 121080
+    },
+    {
+      "epoch": 0.4892189223366476,
+      "grad_norm": 503.83392333984375,
+      "learning_rate": 3.0678181370061805e-05,
+      "loss": 59.4657,
+      "step": 121090
+    },
+    {
+      "epoch": 0.48925932360201524,
+      "grad_norm": 1317.728759765625,
+      "learning_rate": 3.0674781884271254e-05,
+      "loss": 93.741,
+      "step": 121100
+    },
+    {
+      "epoch": 0.4892997248673828,
+      "grad_norm": 1355.2291259765625,
+      "learning_rate": 3.067138228784791e-05,
+      "loss": 84.8171,
+      "step": 121110
+    },
+    {
+      "epoch": 0.48934012613275046,
+      "grad_norm": 1263.6253662109375,
+      "learning_rate": 3.0667982580858044e-05,
+      "loss": 65.6653,
+      "step": 121120
+    },
+    {
+      "epoch": 0.4893805273981181,
+      "grad_norm": 630.547607421875,
+      "learning_rate": 3.066458276336794e-05,
+      "loss": 62.2357,
+      "step": 121130
+    },
+    {
+      "epoch": 0.48942092866348574,
+      "grad_norm": 1021.1478271484375,
+      "learning_rate": 3.0661182835443884e-05,
+      "loss": 57.1426,
+      "step": 121140
+    },
+    {
+      "epoch": 0.4894613299288534,
+      "grad_norm": 1333.4888916015625,
+      "learning_rate": 3.065778279715215e-05,
+      "loss": 49.1796,
+      "step": 121150
+    },
+    {
+      "epoch": 0.489501731194221,
+      "grad_norm": 481.13165283203125,
+      "learning_rate": 3.0654382648559026e-05,
+      "loss": 65.8724,
+      "step": 121160
+    },
+    {
+      "epoch": 0.4895421324595886,
+      "grad_norm": 359.9427795410156,
+      "learning_rate": 3.065098238973081e-05,
+      "loss": 77.4488,
+      "step": 121170
+    },
+    {
+      "epoch": 0.48958253372495625,
+      "grad_norm": 274.55328369140625,
+      "learning_rate": 3.064758202073377e-05,
+      "loss": 54.833,
+      "step": 121180
+    },
+    {
+      "epoch": 0.4896229349903239,
+      "grad_norm": 489.7810974121094,
+      "learning_rate": 3.064418154163422e-05,
+      "loss": 54.8473,
+      "step": 121190
+    },
+    {
+      "epoch": 0.48966333625569153,
+      "grad_norm": 674.3745727539062,
+      "learning_rate": 3.064078095249844e-05,
+      "loss": 51.562,
+      "step": 121200
+    },
+    {
+      "epoch": 0.48970373752105917,
+      "grad_norm": 443.5577087402344,
+      "learning_rate": 3.0637380253392736e-05,
+      "loss": 68.5722,
+      "step": 121210
+    },
+    {
+      "epoch": 0.4897441387864268,
+      "grad_norm": 824.37158203125,
+      "learning_rate": 3.06339794443834e-05,
+      "loss": 46.6327,
+      "step": 121220
+    },
+    {
+      "epoch": 0.48978454005179445,
+      "grad_norm": 275.5827331542969,
+      "learning_rate": 3.063057852553674e-05,
+      "loss": 32.1468,
+      "step": 121230
+    },
+    {
+      "epoch": 0.48982494131716203,
+      "grad_norm": 573.8633422851562,
+      "learning_rate": 3.062717749691904e-05,
+      "loss": 58.9486,
+      "step": 121240
+    },
+    {
+      "epoch": 0.4898653425825297,
+      "grad_norm": 383.3703918457031,
+      "learning_rate": 3.062377635859663e-05,
+      "loss": 39.3934,
+      "step": 121250
+    },
+    {
+      "epoch": 0.4899057438478973,
+      "grad_norm": 942.3863525390625,
+      "learning_rate": 3.06203751106358e-05,
+      "loss": 49.6613,
+      "step": 121260
+    },
+    {
+      "epoch": 0.48994614511326495,
+      "grad_norm": 0.0,
+      "learning_rate": 3.0616973753102856e-05,
+      "loss": 37.8872,
+      "step": 121270
+    },
+    {
+      "epoch": 0.4899865463786326,
+      "grad_norm": 1657.1973876953125,
+      "learning_rate": 3.0613572286064125e-05,
+      "loss": 68.007,
+      "step": 121280
+    },
+    {
+      "epoch": 0.49002694764400023,
+      "grad_norm": 647.4043579101562,
+      "learning_rate": 3.061017070958591e-05,
+      "loss": 92.0626,
+      "step": 121290
+    },
+    {
+      "epoch": 0.4900673489093678,
+      "grad_norm": 885.7959594726562,
+      "learning_rate": 3.0606769023734536e-05,
+      "loss": 30.4313,
+      "step": 121300
+    },
+    {
+      "epoch": 0.49010775017473546,
+      "grad_norm": 1316.2650146484375,
+      "learning_rate": 3.060336722857631e-05,
+      "loss": 83.0374,
+      "step": 121310
+    },
+    {
+      "epoch": 0.4901481514401031,
+      "grad_norm": 1012.3235473632812,
+      "learning_rate": 3.059996532417754e-05,
+      "loss": 41.8788,
+      "step": 121320
+    },
+    {
+      "epoch": 0.49018855270547074,
+      "grad_norm": 449.5831604003906,
+      "learning_rate": 3.059656331060458e-05,
+      "loss": 55.4948,
+      "step": 121330
+    },
+    {
+      "epoch": 0.4902289539708384,
+      "grad_norm": 723.0786743164062,
+      "learning_rate": 3.0593161187923736e-05,
+      "loss": 68.319,
+      "step": 121340
+    },
+    {
+      "epoch": 0.490269355236206,
+      "grad_norm": 569.5224609375,
+      "learning_rate": 3.0589758956201327e-05,
+      "loss": 62.8122,
+      "step": 121350
+    },
+    {
+      "epoch": 0.49030975650157366,
+      "grad_norm": 572.9276733398438,
+      "learning_rate": 3.058635661550369e-05,
+      "loss": 60.2772,
+      "step": 121360
+    },
+    {
+      "epoch": 0.49035015776694124,
+      "grad_norm": 379.9477844238281,
+      "learning_rate": 3.058295416589716e-05,
+      "loss": 69.8202,
+      "step": 121370
+    },
+    {
+      "epoch": 0.4903905590323089,
+      "grad_norm": 534.4418334960938,
+      "learning_rate": 3.0579551607448066e-05,
+      "loss": 51.7835,
+      "step": 121380
+    },
+    {
+      "epoch": 0.4904309602976765,
+      "grad_norm": 803.1372680664062,
+      "learning_rate": 3.057614894022274e-05,
+      "loss": 43.7992,
+      "step": 121390
+    },
+    {
+      "epoch": 0.49047136156304416,
+      "grad_norm": 2386.29736328125,
+      "learning_rate": 3.0572746164287514e-05,
+      "loss": 54.0955,
+      "step": 121400
+    },
+    {
+      "epoch": 0.4905117628284118,
+      "grad_norm": 997.156005859375,
+      "learning_rate": 3.0569343279708734e-05,
+      "loss": 65.676,
+      "step": 121410
+    },
+    {
+      "epoch": 0.49055216409377944,
+      "grad_norm": 525.251953125,
+      "learning_rate": 3.056594028655274e-05,
+      "loss": 52.9306,
+      "step": 121420
+    },
+    {
+      "epoch": 0.490592565359147,
+      "grad_norm": 1752.92236328125,
+      "learning_rate": 3.056253718488588e-05,
+      "loss": 56.5857,
+      "step": 121430
+    },
+    {
+      "epoch": 0.49063296662451467,
+      "grad_norm": 704.6538696289062,
+      "learning_rate": 3.055913397477448e-05,
+      "loss": 68.621,
+      "step": 121440
+    },
+    {
+      "epoch": 0.4906733678898823,
+      "grad_norm": 615.273681640625,
+      "learning_rate": 3.0555730656284914e-05,
+      "loss": 54.718,
+      "step": 121450
+    },
+    {
+      "epoch": 0.49071376915524995,
+      "grad_norm": 676.6206665039062,
+      "learning_rate": 3.0552327229483515e-05,
+      "loss": 60.8578,
+      "step": 121460
+    },
+    {
+      "epoch": 0.4907541704206176,
+      "grad_norm": 512.4281005859375,
+      "learning_rate": 3.054892369443663e-05,
+      "loss": 48.8229,
+      "step": 121470
+    },
+    {
+      "epoch": 0.4907945716859852,
+      "grad_norm": 885.903076171875,
+      "learning_rate": 3.054552005121064e-05,
+      "loss": 83.9689,
+      "step": 121480
+    },
+    {
+      "epoch": 0.4908349729513528,
+      "grad_norm": 405.67919921875,
+      "learning_rate": 3.054211629987187e-05,
+      "loss": 50.5087,
+      "step": 121490
+    },
+    {
+      "epoch": 0.49087537421672045,
+      "grad_norm": 430.7709045410156,
+      "learning_rate": 3.053871244048669e-05,
+      "loss": 54.2436,
+      "step": 121500
+    },
+    {
+      "epoch": 0.4909157754820881,
+      "grad_norm": 394.2481689453125,
+      "learning_rate": 3.0535308473121455e-05,
+      "loss": 37.6645,
+      "step": 121510
+    },
+    {
+      "epoch": 0.49095617674745573,
+      "grad_norm": 2119.940673828125,
+      "learning_rate": 3.053190439784253e-05,
+      "loss": 85.2859,
+      "step": 121520
+    },
+    {
+      "epoch": 0.49099657801282337,
+      "grad_norm": 509.3744201660156,
+      "learning_rate": 3.052850021471629e-05,
+      "loss": 74.4669,
+      "step": 121530
+    },
+    {
+      "epoch": 0.491036979278191,
+      "grad_norm": 0.0,
+      "learning_rate": 3.052509592380909e-05,
+      "loss": 51.207,
+      "step": 121540
+    },
+    {
+      "epoch": 0.49107738054355865,
+      "grad_norm": 615.7112426757812,
+      "learning_rate": 3.052169152518729e-05,
+      "loss": 37.4327,
+      "step": 121550
+    },
+    {
+      "epoch": 0.49111778180892623,
+      "grad_norm": 624.2453002929688,
+      "learning_rate": 3.051828701891729e-05,
+      "loss": 52.5991,
+      "step": 121560
+    },
+    {
+      "epoch": 0.4911581830742939,
+      "grad_norm": 558.4168090820312,
+      "learning_rate": 3.0514882405065432e-05,
+      "loss": 63.3106,
+      "step": 121570
+    },
+    {
+      "epoch": 0.4911985843396615,
+      "grad_norm": 509.6629333496094,
+      "learning_rate": 3.0511477683698108e-05,
+      "loss": 63.4842,
+      "step": 121580
+    },
+    {
+      "epoch": 0.49123898560502915,
+      "grad_norm": 806.631103515625,
+      "learning_rate": 3.050807285488168e-05,
+      "loss": 62.4659,
+      "step": 121590
+    },
+    {
+      "epoch": 0.4912793868703968,
+      "grad_norm": 814.9642944335938,
+      "learning_rate": 3.050466791868254e-05,
+      "loss": 66.1909,
+      "step": 121600
+    },
+    {
+      "epoch": 0.49131978813576443,
+      "grad_norm": 712.5091552734375,
+      "learning_rate": 3.0501262875167063e-05,
+      "loss": 74.5207,
+      "step": 121610
+    },
+    {
+      "epoch": 0.491360189401132,
+      "grad_norm": 615.8709106445312,
+      "learning_rate": 3.0497857724401642e-05,
+      "loss": 79.785,
+      "step": 121620
+    },
+    {
+      "epoch": 0.49140059066649966,
+      "grad_norm": 729.620361328125,
+      "learning_rate": 3.0494452466452644e-05,
+      "loss": 75.2466,
+      "step": 121630
+    },
+    {
+      "epoch": 0.4914409919318673,
+      "grad_norm": 232.82257080078125,
+      "learning_rate": 3.049104710138647e-05,
+      "loss": 40.4371,
+      "step": 121640
+    },
+    {
+      "epoch": 0.49148139319723494,
+      "grad_norm": 528.4829711914062,
+      "learning_rate": 3.0487641629269516e-05,
+      "loss": 49.6708,
+      "step": 121650
+    },
+    {
+      "epoch": 0.4915217944626026,
+      "grad_norm": 2002.76416015625,
+      "learning_rate": 3.0484236050168153e-05,
+      "loss": 52.0956,
+      "step": 121660
+    },
+    {
+      "epoch": 0.4915621957279702,
+      "grad_norm": 575.5763549804688,
+      "learning_rate": 3.048083036414878e-05,
+      "loss": 39.8568,
+      "step": 121670
+    },
+    {
+      "epoch": 0.4916025969933378,
+      "grad_norm": 0.0,
+      "learning_rate": 3.0477424571277807e-05,
+      "loss": 39.302,
+      "step": 121680
+    },
+    {
+      "epoch": 0.49164299825870544,
+      "grad_norm": 412.2666015625,
+      "learning_rate": 3.047401867162162e-05,
+      "loss": 46.3011,
+      "step": 121690
+    },
+    {
+      "epoch": 0.4916833995240731,
+      "grad_norm": 749.7333374023438,
+      "learning_rate": 3.0470612665246618e-05,
+      "loss": 63.5483,
+      "step": 121700
+    },
+    {
+      "epoch": 0.4917238007894407,
+      "grad_norm": 686.0115966796875,
+      "learning_rate": 3.0467206552219208e-05,
+      "loss": 37.7942,
+      "step": 121710
+    },
+    {
+      "epoch": 0.49176420205480836,
+      "grad_norm": 1040.0323486328125,
+      "learning_rate": 3.0463800332605784e-05,
+      "loss": 69.1257,
+      "step": 121720
+    },
+    {
+      "epoch": 0.491804603320176,
+      "grad_norm": 547.2194213867188,
+      "learning_rate": 3.046039400647277e-05,
+      "loss": 28.7517,
+      "step": 121730
+    },
+    {
+      "epoch": 0.49184500458554364,
+      "grad_norm": 1607.9515380859375,
+      "learning_rate": 3.0456987573886564e-05,
+      "loss": 67.1605,
+      "step": 121740
+    },
+    {
+      "epoch": 0.49188540585091123,
+      "grad_norm": 1343.8623046875,
+      "learning_rate": 3.045358103491357e-05,
+      "loss": 61.9395,
+      "step": 121750
+    },
+    {
+      "epoch": 0.49192580711627887,
+      "grad_norm": 360.9034729003906,
+      "learning_rate": 3.0450174389620205e-05,
+      "loss": 60.9107,
+      "step": 121760
+    },
+    {
+      "epoch": 0.4919662083816465,
+      "grad_norm": 658.2755126953125,
+      "learning_rate": 3.044676763807288e-05,
+      "loss": 44.2099,
+      "step": 121770
+    },
+    {
+      "epoch": 0.49200660964701415,
+      "grad_norm": 463.0184326171875,
+      "learning_rate": 3.044336078033803e-05,
+      "loss": 50.352,
+      "step": 121780
+    },
+    {
+      "epoch": 0.4920470109123818,
+      "grad_norm": 1313.499755859375,
+      "learning_rate": 3.043995381648205e-05,
+      "loss": 65.2679,
+      "step": 121790
+    },
+    {
+      "epoch": 0.49208741217774943,
+      "grad_norm": 662.1671142578125,
+      "learning_rate": 3.0436546746571372e-05,
+      "loss": 47.1219,
+      "step": 121800
+    },
+    {
+      "epoch": 0.492127813443117,
+      "grad_norm": 3354.996826171875,
+      "learning_rate": 3.0433139570672407e-05,
+      "loss": 87.668,
+      "step": 121810
+    },
+    {
+      "epoch": 0.49216821470848465,
+      "grad_norm": 931.5249633789062,
+      "learning_rate": 3.0429732288851603e-05,
+      "loss": 44.4428,
+      "step": 121820
+    },
+    {
+      "epoch": 0.4922086159738523,
+      "grad_norm": 798.1958618164062,
+      "learning_rate": 3.0426324901175374e-05,
+      "loss": 51.0042,
+      "step": 121830
+    },
+    {
+      "epoch": 0.49224901723921993,
+      "grad_norm": 674.7509765625,
+      "learning_rate": 3.0422917407710137e-05,
+      "loss": 33.4516,
+      "step": 121840
+    },
+    {
+      "epoch": 0.4922894185045876,
+      "grad_norm": 734.49169921875,
+      "learning_rate": 3.0419509808522334e-05,
+      "loss": 77.9193,
+      "step": 121850
+    },
+    {
+      "epoch": 0.4923298197699552,
+      "grad_norm": 561.5974731445312,
+      "learning_rate": 3.0416102103678402e-05,
+      "loss": 52.3013,
+      "step": 121860
+    },
+    {
+      "epoch": 0.49237022103532285,
+      "grad_norm": 292.3045959472656,
+      "learning_rate": 3.041269429324477e-05,
+      "loss": 31.1926,
+      "step": 121870
+    },
+    {
+      "epoch": 0.49241062230069044,
+      "grad_norm": 613.70361328125,
+      "learning_rate": 3.040928637728787e-05,
+      "loss": 56.5007,
+      "step": 121880
+    },
+    {
+      "epoch": 0.4924510235660581,
+      "grad_norm": 1232.8988037109375,
+      "learning_rate": 3.040587835587415e-05,
+      "loss": 73.1134,
+      "step": 121890
+    },
+    {
+      "epoch": 0.4924914248314257,
+      "grad_norm": 831.9037475585938,
+      "learning_rate": 3.0402470229070056e-05,
+      "loss": 46.9302,
+      "step": 121900
+    },
+    {
+      "epoch": 0.49253182609679336,
+      "grad_norm": 409.0111389160156,
+      "learning_rate": 3.039906199694202e-05,
+      "loss": 39.5951,
+      "step": 121910
+    },
+    {
+      "epoch": 0.492572227362161,
+      "grad_norm": 646.8606567382812,
+      "learning_rate": 3.0395653659556488e-05,
+      "loss": 47.5818,
+      "step": 121920
+    },
+    {
+      "epoch": 0.49261262862752864,
+      "grad_norm": 363.55609130859375,
+      "learning_rate": 3.039224521697991e-05,
+      "loss": 72.8517,
+      "step": 121930
+    },
+    {
+      "epoch": 0.4926530298928962,
+      "grad_norm": 498.0539855957031,
+      "learning_rate": 3.0388836669278738e-05,
+      "loss": 53.5023,
+      "step": 121940
+    },
+    {
+      "epoch": 0.49269343115826386,
+      "grad_norm": 943.4515991210938,
+      "learning_rate": 3.038542801651941e-05,
+      "loss": 59.0278,
+      "step": 121950
+    },
+    {
+      "epoch": 0.4927338324236315,
+      "grad_norm": 3113.823486328125,
+      "learning_rate": 3.0382019258768403e-05,
+      "loss": 72.7263,
+      "step": 121960
+    },
+    {
+      "epoch": 0.49277423368899914,
+      "grad_norm": 509.8355407714844,
+      "learning_rate": 3.0378610396092154e-05,
+      "loss": 51.3322,
+      "step": 121970
+    },
+    {
+      "epoch": 0.4928146349543668,
+      "grad_norm": 791.736572265625,
+      "learning_rate": 3.0375201428557132e-05,
+      "loss": 48.9846,
+      "step": 121980
+    },
+    {
+      "epoch": 0.4928550362197344,
+      "grad_norm": 532.8504638671875,
+      "learning_rate": 3.0371792356229783e-05,
+      "loss": 35.4836,
+      "step": 121990
+    },
+    {
+      "epoch": 0.492895437485102,
+      "grad_norm": 419.5154113769531,
+      "learning_rate": 3.0368383179176585e-05,
+      "loss": 69.5128,
+      "step": 122000
+    },
+    {
+      "epoch": 0.49293583875046965,
+      "grad_norm": 3489.129150390625,
+      "learning_rate": 3.036497389746399e-05,
+      "loss": 68.8055,
+      "step": 122010
+    },
+    {
+      "epoch": 0.4929762400158373,
+      "grad_norm": 1044.6541748046875,
+      "learning_rate": 3.0361564511158457e-05,
+      "loss": 62.8193,
+      "step": 122020
+    },
+    {
+      "epoch": 0.4930166412812049,
+      "grad_norm": 1858.55615234375,
+      "learning_rate": 3.0358155020326477e-05,
+      "loss": 77.7083,
+      "step": 122030
+    },
+    {
+      "epoch": 0.49305704254657257,
+      "grad_norm": 682.7332153320312,
+      "learning_rate": 3.0354745425034498e-05,
+      "loss": 64.1814,
+      "step": 122040
+    },
+    {
+      "epoch": 0.4930974438119402,
+      "grad_norm": 653.3331298828125,
+      "learning_rate": 3.0351335725349e-05,
+      "loss": 49.824,
+      "step": 122050
+    },
+    {
+      "epoch": 0.49313784507730785,
+      "grad_norm": 859.71337890625,
+      "learning_rate": 3.0347925921336463e-05,
+      "loss": 54.182,
+      "step": 122060
+    },
+    {
+      "epoch": 0.49317824634267543,
+      "grad_norm": 1686.3477783203125,
+      "learning_rate": 3.0344516013063357e-05,
+      "loss": 43.8826,
+      "step": 122070
+    },
+    {
+      "epoch": 0.49321864760804307,
+      "grad_norm": 1858.2529296875,
+      "learning_rate": 3.034110600059616e-05,
+      "loss": 42.0115,
+      "step": 122080
+    },
+    {
+      "epoch": 0.4932590488734107,
+      "grad_norm": 857.298828125,
+      "learning_rate": 3.0337695884001343e-05,
+      "loss": 44.892,
+      "step": 122090
+    },
+    {
+      "epoch": 0.49329945013877835,
+      "grad_norm": 3229.34765625,
+      "learning_rate": 3.0334285663345404e-05,
+      "loss": 49.8066,
+      "step": 122100
+    },
+    {
+      "epoch": 0.493339851404146,
+      "grad_norm": 782.3743896484375,
+      "learning_rate": 3.033087533869482e-05,
+      "loss": 46.6393,
+      "step": 122110
+    },
+    {
+      "epoch": 0.49338025266951363,
+      "grad_norm": 1177.9285888671875,
+      "learning_rate": 3.032746491011607e-05,
+      "loss": 54.9906,
+      "step": 122120
+    },
+    {
+      "epoch": 0.4934206539348812,
+      "grad_norm": 997.4578247070312,
+      "learning_rate": 3.0324054377675654e-05,
+      "loss": 47.8525,
+      "step": 122130
+    },
+    {
+      "epoch": 0.49346105520024885,
+      "grad_norm": 1065.852294921875,
+      "learning_rate": 3.032064374144005e-05,
+      "loss": 58.0819,
+      "step": 122140
+    },
+    {
+      "epoch": 0.4935014564656165,
+      "grad_norm": 2079.648681640625,
+      "learning_rate": 3.031723300147577e-05,
+      "loss": 95.1813,
+      "step": 122150
+    },
+    {
+      "epoch": 0.49354185773098413,
+      "grad_norm": 759.1595458984375,
+      "learning_rate": 3.0313822157849287e-05,
+      "loss": 54.4963,
+      "step": 122160
+    },
+    {
+      "epoch": 0.4935822589963518,
+      "grad_norm": 2073.466064453125,
+      "learning_rate": 3.031041121062711e-05,
+      "loss": 70.8082,
+      "step": 122170
+    },
+    {
+      "epoch": 0.4936226602617194,
+      "grad_norm": 1743.0040283203125,
+      "learning_rate": 3.030700015987573e-05,
+      "loss": 57.2004,
+      "step": 122180
+    },
+    {
+      "epoch": 0.49366306152708705,
+      "grad_norm": 714.07275390625,
+      "learning_rate": 3.030358900566165e-05,
+      "loss": 60.7911,
+      "step": 122190
+    },
+    {
+      "epoch": 0.49370346279245464,
+      "grad_norm": 1210.780029296875,
+      "learning_rate": 3.0300177748051373e-05,
+      "loss": 57.1338,
+      "step": 122200
+    },
+    {
+      "epoch": 0.4937438640578223,
+      "grad_norm": 815.665771484375,
+      "learning_rate": 3.02967663871114e-05,
+      "loss": 74.5411,
+      "step": 122210
+    },
+    {
+      "epoch": 0.4937842653231899,
+      "grad_norm": 1104.837646484375,
+      "learning_rate": 3.0293354922908235e-05,
+      "loss": 69.4025,
+      "step": 122220
+    },
+    {
+      "epoch": 0.49382466658855756,
+      "grad_norm": 693.127197265625,
+      "learning_rate": 3.0289943355508392e-05,
+      "loss": 63.0513,
+      "step": 122230
+    },
+    {
+      "epoch": 0.4938650678539252,
+      "grad_norm": 683.1317749023438,
+      "learning_rate": 3.028653168497838e-05,
+      "loss": 46.7121,
+      "step": 122240
+    },
+    {
+      "epoch": 0.49390546911929284,
+      "grad_norm": 317.39532470703125,
+      "learning_rate": 3.028311991138472e-05,
+      "loss": 63.5555,
+      "step": 122250
+    },
+    {
+      "epoch": 0.4939458703846604,
+      "grad_norm": 466.8677062988281,
+      "learning_rate": 3.0279708034793907e-05,
+      "loss": 45.8454,
+      "step": 122260
+    },
+    {
+      "epoch": 0.49398627165002806,
+      "grad_norm": 1304.4661865234375,
+      "learning_rate": 3.027629605527248e-05,
+      "loss": 67.0937,
+      "step": 122270
+    },
+    {
+      "epoch": 0.4940266729153957,
+      "grad_norm": 341.1806335449219,
+      "learning_rate": 3.0272883972886935e-05,
+      "loss": 32.4989,
+      "step": 122280
+    },
+    {
+      "epoch": 0.49406707418076334,
+      "grad_norm": 668.6641235351562,
+      "learning_rate": 3.02694717877038e-05,
+      "loss": 58.9114,
+      "step": 122290
+    },
+    {
+      "epoch": 0.494107475446131,
+      "grad_norm": 1312.1033935546875,
+      "learning_rate": 3.02660594997896e-05,
+      "loss": 52.9595,
+      "step": 122300
+    },
+    {
+      "epoch": 0.4941478767114986,
+      "grad_norm": 206.2220916748047,
+      "learning_rate": 3.0262647109210867e-05,
+      "loss": 42.2671,
+      "step": 122310
+    },
+    {
+      "epoch": 0.4941882779768662,
+      "grad_norm": 533.5427856445312,
+      "learning_rate": 3.0259234616034116e-05,
+      "loss": 54.6035,
+      "step": 122320
+    },
+    {
+      "epoch": 0.49422867924223385,
+      "grad_norm": 389.654296875,
+      "learning_rate": 3.0255822020325873e-05,
+      "loss": 46.6373,
+      "step": 122330
+    },
+    {
+      "epoch": 0.4942690805076015,
+      "grad_norm": 509.89276123046875,
+      "learning_rate": 3.025240932215268e-05,
+      "loss": 65.848,
+      "step": 122340
+    },
+    {
+      "epoch": 0.49430948177296913,
+      "grad_norm": 738.1375732421875,
+      "learning_rate": 3.024899652158107e-05,
+      "loss": 46.7456,
+      "step": 122350
+    },
+    {
+      "epoch": 0.49434988303833677,
+      "grad_norm": 272.38751220703125,
+      "learning_rate": 3.0245583618677558e-05,
+      "loss": 58.7914,
+      "step": 122360
+    },
+    {
+      "epoch": 0.4943902843037044,
+      "grad_norm": 599.438232421875,
+      "learning_rate": 3.0242170613508692e-05,
+      "loss": 57.4107,
+      "step": 122370
+    },
+    {
+      "epoch": 0.49443068556907205,
+      "grad_norm": 323.33062744140625,
+      "learning_rate": 3.0238757506141012e-05,
+      "loss": 48.1619,
+      "step": 122380
+    },
+    {
+      "epoch": 0.49447108683443963,
+      "grad_norm": 1464.4964599609375,
+      "learning_rate": 3.0235344296641067e-05,
+      "loss": 58.9648,
+      "step": 122390
+    },
+    {
+      "epoch": 0.4945114880998073,
+      "grad_norm": 2316.94873046875,
+      "learning_rate": 3.023193098507538e-05,
+      "loss": 80.6663,
+      "step": 122400
+    },
+    {
+      "epoch": 0.4945518893651749,
+      "grad_norm": 777.03173828125,
+      "learning_rate": 3.0228517571510507e-05,
+      "loss": 74.6564,
+      "step": 122410
+    },
+    {
+      "epoch": 0.49459229063054255,
+      "grad_norm": 832.2376098632812,
+      "learning_rate": 3.0225104056013e-05,
+      "loss": 50.8063,
+      "step": 122420
+    },
+    {
+      "epoch": 0.4946326918959102,
+      "grad_norm": 1374.7584228515625,
+      "learning_rate": 3.0221690438649386e-05,
+      "loss": 84.333,
+      "step": 122430
+    },
+    {
+      "epoch": 0.49467309316127783,
+      "grad_norm": 586.2657470703125,
+      "learning_rate": 3.0218276719486244e-05,
+      "loss": 54.8292,
+      "step": 122440
+    },
+    {
+      "epoch": 0.4947134944266454,
+      "grad_norm": 362.1689147949219,
+      "learning_rate": 3.0214862898590095e-05,
+      "loss": 45.3425,
+      "step": 122450
+    },
+    {
+      "epoch": 0.49475389569201306,
+      "grad_norm": 740.5028076171875,
+      "learning_rate": 3.021144897602752e-05,
+      "loss": 65.865,
+      "step": 122460
+    },
+    {
+      "epoch": 0.4947942969573807,
+      "grad_norm": 776.755615234375,
+      "learning_rate": 3.020803495186506e-05,
+      "loss": 45.3244,
+      "step": 122470
+    },
+    {
+      "epoch": 0.49483469822274834,
+      "grad_norm": 1503.520263671875,
+      "learning_rate": 3.020462082616928e-05,
+      "loss": 44.012,
+      "step": 122480
+    },
+    {
+      "epoch": 0.494875099488116,
+      "grad_norm": 772.3239135742188,
+      "learning_rate": 3.0201206599006733e-05,
+      "loss": 50.5993,
+      "step": 122490
+    },
+    {
+      "epoch": 0.4949155007534836,
+      "grad_norm": 343.9445495605469,
+      "learning_rate": 3.0197792270443982e-05,
+      "loss": 43.6323,
+      "step": 122500
+    },
+    {
+      "epoch": 0.49495590201885126,
+      "grad_norm": 765.3672485351562,
+      "learning_rate": 3.0194377840547606e-05,
+      "loss": 48.1465,
+      "step": 122510
+    },
+    {
+      "epoch": 0.49499630328421884,
+      "grad_norm": 936.2794799804688,
+      "learning_rate": 3.0190963309384156e-05,
+      "loss": 58.2882,
+      "step": 122520
+    },
+    {
+      "epoch": 0.4950367045495865,
+      "grad_norm": 770.3196411132812,
+      "learning_rate": 3.01875486770202e-05,
+      "loss": 43.9133,
+      "step": 122530
+    },
+    {
+      "epoch": 0.4950771058149541,
+      "grad_norm": 411.43255615234375,
+      "learning_rate": 3.0184133943522314e-05,
+      "loss": 50.147,
+      "step": 122540
+    },
+    {
+      "epoch": 0.49511750708032176,
+      "grad_norm": 589.0484619140625,
+      "learning_rate": 3.0180719108957063e-05,
+      "loss": 48.991,
+      "step": 122550
+    },
+    {
+      "epoch": 0.4951579083456894,
+      "grad_norm": 317.48046875,
+      "learning_rate": 3.0177304173391037e-05,
+      "loss": 47.1381,
+      "step": 122560
+    },
+    {
+      "epoch": 0.49519830961105704,
+      "grad_norm": 2199.10888671875,
+      "learning_rate": 3.0173889136890786e-05,
+      "loss": 60.0664,
+      "step": 122570
+    },
+    {
+      "epoch": 0.4952387108764246,
+      "grad_norm": 2371.26025390625,
+      "learning_rate": 3.0170473999522915e-05,
+      "loss": 53.2513,
+      "step": 122580
+    },
+    {
+      "epoch": 0.49527911214179227,
+      "grad_norm": 2204.5810546875,
+      "learning_rate": 3.016705876135399e-05,
+      "loss": 75.0456,
+      "step": 122590
+    },
+    {
+      "epoch": 0.4953195134071599,
+      "grad_norm": 1063.3651123046875,
+      "learning_rate": 3.016364342245059e-05,
+      "loss": 85.0364,
+      "step": 122600
+    },
+    {
+      "epoch": 0.49535991467252755,
+      "grad_norm": 1718.4310302734375,
+      "learning_rate": 3.016022798287931e-05,
+      "loss": 66.0194,
+      "step": 122610
+    },
+    {
+      "epoch": 0.4954003159378952,
+      "grad_norm": 953.0413208007812,
+      "learning_rate": 3.0156812442706715e-05,
+      "loss": 54.6552,
+      "step": 122620
+    },
+    {
+      "epoch": 0.4954407172032628,
+      "grad_norm": 702.347900390625,
+      "learning_rate": 3.015339680199941e-05,
+      "loss": 39.405,
+      "step": 122630
+    },
+    {
+      "epoch": 0.4954811184686304,
+      "grad_norm": 477.78302001953125,
+      "learning_rate": 3.0149981060823995e-05,
+      "loss": 39.2307,
+      "step": 122640
+    },
+    {
+      "epoch": 0.49552151973399805,
+      "grad_norm": 893.2791137695312,
+      "learning_rate": 3.0146565219247036e-05,
+      "loss": 70.0523,
+      "step": 122650
+    },
+    {
+      "epoch": 0.4955619209993657,
+      "grad_norm": 493.2530517578125,
+      "learning_rate": 3.0143149277335138e-05,
+      "loss": 59.8175,
+      "step": 122660
+    },
+    {
+      "epoch": 0.49560232226473333,
+      "grad_norm": 2252.927734375,
+      "learning_rate": 3.01397332351549e-05,
+      "loss": 79.7985,
+      "step": 122670
+    },
+    {
+      "epoch": 0.49564272353010097,
+      "grad_norm": 639.860595703125,
+      "learning_rate": 3.013631709277292e-05,
+      "loss": 80.3618,
+      "step": 122680
+    },
+    {
+      "epoch": 0.4956831247954686,
+      "grad_norm": 542.6490478515625,
+      "learning_rate": 3.013290085025579e-05,
+      "loss": 64.4751,
+      "step": 122690
+    },
+    {
+      "epoch": 0.49572352606083625,
+      "grad_norm": 933.4337158203125,
+      "learning_rate": 3.0129484507670115e-05,
+      "loss": 57.9422,
+      "step": 122700
+    },
+    {
+      "epoch": 0.49576392732620383,
+      "grad_norm": 1414.308837890625,
+      "learning_rate": 3.0126068065082504e-05,
+      "loss": 50.3865,
+      "step": 122710
+    },
+    {
+      "epoch": 0.4958043285915715,
+      "grad_norm": 900.0178833007812,
+      "learning_rate": 3.0122651522559553e-05,
+      "loss": 54.273,
+      "step": 122720
+    },
+    {
+      "epoch": 0.4958447298569391,
+      "grad_norm": 1027.331787109375,
+      "learning_rate": 3.0119234880167867e-05,
+      "loss": 46.3332,
+      "step": 122730
+    },
+    {
+      "epoch": 0.49588513112230675,
+      "grad_norm": 527.3615112304688,
+      "learning_rate": 3.0115818137974067e-05,
+      "loss": 98.7442,
+      "step": 122740
+    },
+    {
+      "epoch": 0.4959255323876744,
+      "grad_norm": 546.8230590820312,
+      "learning_rate": 3.0112401296044757e-05,
+      "loss": 68.4313,
+      "step": 122750
+    },
+    {
+      "epoch": 0.49596593365304203,
+      "grad_norm": 835.651611328125,
+      "learning_rate": 3.0108984354446556e-05,
+      "loss": 45.3258,
+      "step": 122760
+    },
+    {
+      "epoch": 0.4960063349184096,
+      "grad_norm": 459.29144287109375,
+      "learning_rate": 3.0105567313246074e-05,
+      "loss": 70.6443,
+      "step": 122770
+    },
+    {
+      "epoch": 0.49604673618377726,
+      "grad_norm": 1120.5982666015625,
+      "learning_rate": 3.010215017250993e-05,
+      "loss": 73.4087,
+      "step": 122780
+    },
+    {
+      "epoch": 0.4960871374491449,
+      "grad_norm": 815.575927734375,
+      "learning_rate": 3.0098732932304734e-05,
+      "loss": 51.7506,
+      "step": 122790
+    },
+    {
+      "epoch": 0.49612753871451254,
+      "grad_norm": 529.2800903320312,
+      "learning_rate": 3.0095315592697126e-05,
+      "loss": 43.3278,
+      "step": 122800
+    },
+    {
+      "epoch": 0.4961679399798802,
+      "grad_norm": 352.13128662109375,
+      "learning_rate": 3.0091898153753705e-05,
+      "loss": 52.9101,
+      "step": 122810
+    },
+    {
+      "epoch": 0.4962083412452478,
+      "grad_norm": 2164.66064453125,
+      "learning_rate": 3.0088480615541113e-05,
+      "loss": 69.647,
+      "step": 122820
+    },
+    {
+      "epoch": 0.49624874251061546,
+      "grad_norm": 468.87158203125,
+      "learning_rate": 3.0085062978125967e-05,
+      "loss": 65.0103,
+      "step": 122830
+    },
+    {
+      "epoch": 0.49628914377598304,
+      "grad_norm": 771.53857421875,
+      "learning_rate": 3.008164524157491e-05,
+      "loss": 47.0131,
+      "step": 122840
+    },
+    {
+      "epoch": 0.4963295450413507,
+      "grad_norm": 630.9580078125,
+      "learning_rate": 3.0078227405954557e-05,
+      "loss": 36.1967,
+      "step": 122850
+    },
+    {
+      "epoch": 0.4963699463067183,
+      "grad_norm": 632.0338745117188,
+      "learning_rate": 3.007480947133155e-05,
+      "loss": 78.9644,
+      "step": 122860
+    },
+    {
+      "epoch": 0.49641034757208596,
+      "grad_norm": 748.582763671875,
+      "learning_rate": 3.0071391437772516e-05,
+      "loss": 52.5742,
+      "step": 122870
+    },
+    {
+      "epoch": 0.4964507488374536,
+      "grad_norm": 402.97882080078125,
+      "learning_rate": 3.00679733053441e-05,
+      "loss": 49.2446,
+      "step": 122880
+    },
+    {
+      "epoch": 0.49649115010282124,
+      "grad_norm": 814.5822143554688,
+      "learning_rate": 3.0064555074112927e-05,
+      "loss": 49.1269,
+      "step": 122890
+    },
+    {
+      "epoch": 0.49653155136818883,
+      "grad_norm": 484.45751953125,
+      "learning_rate": 3.0061136744145652e-05,
+      "loss": 46.5377,
+      "step": 122900
+    },
+    {
+      "epoch": 0.49657195263355647,
+      "grad_norm": 1381.5684814453125,
+      "learning_rate": 3.0057718315508905e-05,
+      "loss": 56.9945,
+      "step": 122910
+    },
+    {
+      "epoch": 0.4966123538989241,
+      "grad_norm": 1052.2681884765625,
+      "learning_rate": 3.005429978826934e-05,
+      "loss": 54.9941,
+      "step": 122920
+    },
+    {
+      "epoch": 0.49665275516429175,
+      "grad_norm": 1034.386474609375,
+      "learning_rate": 3.0050881162493593e-05,
+      "loss": 46.4848,
+      "step": 122930
+    },
+    {
+      "epoch": 0.4966931564296594,
+      "grad_norm": 919.5922241210938,
+      "learning_rate": 3.004746243824833e-05,
+      "loss": 56.9942,
+      "step": 122940
+    },
+    {
+      "epoch": 0.49673355769502703,
+      "grad_norm": 787.20947265625,
+      "learning_rate": 3.0044043615600175e-05,
+      "loss": 65.2977,
+      "step": 122950
+    },
+    {
+      "epoch": 0.4967739589603946,
+      "grad_norm": 0.0,
+      "learning_rate": 3.0040624694615803e-05,
+      "loss": 41.2134,
+      "step": 122960
+    },
+    {
+      "epoch": 0.49681436022576225,
+      "grad_norm": 683.5494995117188,
+      "learning_rate": 3.003720567536185e-05,
+      "loss": 72.0073,
+      "step": 122970
+    },
+    {
+      "epoch": 0.4968547614911299,
+      "grad_norm": 575.966796875,
+      "learning_rate": 3.003378655790498e-05,
+      "loss": 65.4787,
+      "step": 122980
+    },
+    {
+      "epoch": 0.49689516275649753,
+      "grad_norm": 1353.0836181640625,
+      "learning_rate": 3.0030367342311848e-05,
+      "loss": 67.0195,
+      "step": 122990
+    },
+    {
+      "epoch": 0.4969355640218652,
+      "grad_norm": 590.2861328125,
+      "learning_rate": 3.002694802864912e-05,
+      "loss": 53.1483,
+      "step": 123000
+    },
+    {
+      "epoch": 0.4969759652872328,
+      "grad_norm": 1127.41259765625,
+      "learning_rate": 3.002352861698345e-05,
+      "loss": 88.6338,
+      "step": 123010
+    },
+    {
+      "epoch": 0.49701636655260045,
+      "grad_norm": 902.4027099609375,
+      "learning_rate": 3.00201091073815e-05,
+      "loss": 39.8632,
+      "step": 123020
+    },
+    {
+      "epoch": 0.49705676781796804,
+      "grad_norm": 760.8033447265625,
+      "learning_rate": 3.0016689499909945e-05,
+      "loss": 37.9242,
+      "step": 123030
+    },
+    {
+      "epoch": 0.4970971690833357,
+      "grad_norm": 612.1544189453125,
+      "learning_rate": 3.0013269794635446e-05,
+      "loss": 56.6241,
+      "step": 123040
+    },
+    {
+      "epoch": 0.4971375703487033,
+      "grad_norm": 777.5595092773438,
+      "learning_rate": 3.0009849991624662e-05,
+      "loss": 82.5385,
+      "step": 123050
+    },
+    {
+      "epoch": 0.49717797161407096,
+      "grad_norm": 1702.484619140625,
+      "learning_rate": 3.0006430090944277e-05,
+      "loss": 59.1968,
+      "step": 123060
+    },
+    {
+      "epoch": 0.4972183728794386,
+      "grad_norm": 537.38134765625,
+      "learning_rate": 3.000301009266096e-05,
+      "loss": 46.0222,
+      "step": 123070
+    },
+    {
+      "epoch": 0.49725877414480624,
+      "grad_norm": 305.5016784667969,
+      "learning_rate": 2.9999589996841386e-05,
+      "loss": 57.2696,
+      "step": 123080
+    },
+    {
+      "epoch": 0.4972991754101738,
+      "grad_norm": 844.5637817382812,
+      "learning_rate": 2.9996169803552233e-05,
+      "loss": 56.0423,
+      "step": 123090
+    },
+    {
+      "epoch": 0.49733957667554146,
+      "grad_norm": 883.1146850585938,
+      "learning_rate": 2.9992749512860173e-05,
+      "loss": 32.5165,
+      "step": 123100
+    },
+    {
+      "epoch": 0.4973799779409091,
+      "grad_norm": 1908.025634765625,
+      "learning_rate": 2.99893291248319e-05,
+      "loss": 110.2845,
+      "step": 123110
+    },
+    {
+      "epoch": 0.49742037920627674,
+      "grad_norm": 1058.1068115234375,
+      "learning_rate": 2.9985908639534075e-05,
+      "loss": 67.6201,
+      "step": 123120
+    },
+    {
+      "epoch": 0.4974607804716444,
+      "grad_norm": 1846.614013671875,
+      "learning_rate": 2.998248805703341e-05,
+      "loss": 49.1556,
+      "step": 123130
+    },
+    {
+      "epoch": 0.497501181737012,
+      "grad_norm": 1011.847412109375,
+      "learning_rate": 2.9979067377396565e-05,
+      "loss": 25.5965,
+      "step": 123140
+    },
+    {
+      "epoch": 0.49754158300237966,
+      "grad_norm": 627.6660766601562,
+      "learning_rate": 2.9975646600690234e-05,
+      "loss": 34.3208,
+      "step": 123150
+    },
+    {
+      "epoch": 0.49758198426774725,
+      "grad_norm": 770.2060546875,
+      "learning_rate": 2.9972225726981113e-05,
+      "loss": 52.1218,
+      "step": 123160
+    },
+    {
+      "epoch": 0.4976223855331149,
+      "grad_norm": 398.234130859375,
+      "learning_rate": 2.99688047563359e-05,
+      "loss": 60.0683,
+      "step": 123170
+    },
+    {
+      "epoch": 0.4976627867984825,
+      "grad_norm": 1440.9393310546875,
+      "learning_rate": 2.996538368882127e-05,
+      "loss": 94.9321,
+      "step": 123180
+    },
+    {
+      "epoch": 0.49770318806385017,
+      "grad_norm": 580.39892578125,
+      "learning_rate": 2.9961962524503927e-05,
+      "loss": 46.6695,
+      "step": 123190
+    },
+    {
+      "epoch": 0.4977435893292178,
+      "grad_norm": 402.80419921875,
+      "learning_rate": 2.9958541263450584e-05,
+      "loss": 58.4956,
+      "step": 123200
+    },
+    {
+      "epoch": 0.49778399059458545,
+      "grad_norm": 709.5952758789062,
+      "learning_rate": 2.9955119905727925e-05,
+      "loss": 54.5307,
+      "step": 123210
+    },
+    {
+      "epoch": 0.49782439185995303,
+      "grad_norm": 418.1485290527344,
+      "learning_rate": 2.995169845140264e-05,
+      "loss": 33.9692,
+      "step": 123220
+    },
+    {
+      "epoch": 0.49786479312532067,
+      "grad_norm": 555.658203125,
+      "learning_rate": 2.994827690054145e-05,
+      "loss": 57.7686,
+      "step": 123230
+    },
+    {
+      "epoch": 0.4979051943906883,
+      "grad_norm": 1826.1614990234375,
+      "learning_rate": 2.9944855253211052e-05,
+      "loss": 59.2031,
+      "step": 123240
+    },
+    {
+      "epoch": 0.49794559565605595,
+      "grad_norm": 728.795166015625,
+      "learning_rate": 2.9941433509478156e-05,
+      "loss": 64.7671,
+      "step": 123250
+    },
+    {
+      "epoch": 0.4979859969214236,
+      "grad_norm": 562.8650512695312,
+      "learning_rate": 2.993801166940947e-05,
+      "loss": 30.4642,
+      "step": 123260
+    },
+    {
+      "epoch": 0.49802639818679123,
+      "grad_norm": 718.48681640625,
+      "learning_rate": 2.9934589733071704e-05,
+      "loss": 43.2636,
+      "step": 123270
+    },
+    {
+      "epoch": 0.4980667994521588,
+      "grad_norm": 775.666748046875,
+      "learning_rate": 2.9931167700531578e-05,
+      "loss": 64.5946,
+      "step": 123280
+    },
+    {
+      "epoch": 0.49810720071752645,
+      "grad_norm": 941.3685302734375,
+      "learning_rate": 2.9927745571855786e-05,
+      "loss": 68.1941,
+      "step": 123290
+    },
+    {
+      "epoch": 0.4981476019828941,
+      "grad_norm": 806.8750610351562,
+      "learning_rate": 2.9924323347111073e-05,
+      "loss": 75.4404,
+      "step": 123300
+    },
+    {
+      "epoch": 0.49818800324826173,
+      "grad_norm": 1879.86279296875,
+      "learning_rate": 2.992090102636413e-05,
+      "loss": 82.7253,
+      "step": 123310
+    },
+    {
+      "epoch": 0.4982284045136294,
+      "grad_norm": 500.6525573730469,
+      "learning_rate": 2.991747860968168e-05,
+      "loss": 59.617,
+      "step": 123320
+    },
+    {
+      "epoch": 0.498268805778997,
+      "grad_norm": 576.9788818359375,
+      "learning_rate": 2.9914056097130473e-05,
+      "loss": 37.7997,
+      "step": 123330
+    },
+    {
+      "epoch": 0.49830920704436465,
+      "grad_norm": 1132.8011474609375,
+      "learning_rate": 2.9910633488777196e-05,
+      "loss": 60.6151,
+      "step": 123340
+    },
+    {
+      "epoch": 0.49834960830973224,
+      "grad_norm": 1793.87109375,
+      "learning_rate": 2.99072107846886e-05,
+      "loss": 49.606,
+      "step": 123350
+    },
+    {
+      "epoch": 0.4983900095750999,
+      "grad_norm": 226.36434936523438,
+      "learning_rate": 2.9903787984931396e-05,
+      "loss": 29.6053,
+      "step": 123360
+    },
+    {
+      "epoch": 0.4984304108404675,
+      "grad_norm": 257.8052673339844,
+      "learning_rate": 2.9900365089572328e-05,
+      "loss": 42.242,
+      "step": 123370
+    },
+    {
+      "epoch": 0.49847081210583516,
+      "grad_norm": 719.3456420898438,
+      "learning_rate": 2.9896942098678122e-05,
+      "loss": 46.6997,
+      "step": 123380
+    },
+    {
+      "epoch": 0.4985112133712028,
+      "grad_norm": 641.3980102539062,
+      "learning_rate": 2.9893519012315503e-05,
+      "loss": 67.5891,
+      "step": 123390
+    },
+    {
+      "epoch": 0.49855161463657044,
+      "grad_norm": 685.2548828125,
+      "learning_rate": 2.9890095830551207e-05,
+      "loss": 60.4898,
+      "step": 123400
+    },
+    {
+      "epoch": 0.498592015901938,
+      "grad_norm": 632.763671875,
+      "learning_rate": 2.9886672553451985e-05,
+      "loss": 61.0664,
+      "step": 123410
+    },
+    {
+      "epoch": 0.49863241716730566,
+      "grad_norm": 0.0,
+      "learning_rate": 2.988324918108456e-05,
+      "loss": 56.5434,
+      "step": 123420
+    },
+    {
+      "epoch": 0.4986728184326733,
+      "grad_norm": 1752.1431884765625,
+      "learning_rate": 2.9879825713515676e-05,
+      "loss": 105.4006,
+      "step": 123430
+    },
+    {
+      "epoch": 0.49871321969804094,
+      "grad_norm": 575.599853515625,
+      "learning_rate": 2.9876402150812078e-05,
+      "loss": 52.491,
+      "step": 123440
+    },
+    {
+      "epoch": 0.4987536209634086,
+      "grad_norm": 620.3259887695312,
+      "learning_rate": 2.9872978493040514e-05,
+      "loss": 48.9634,
+      "step": 123450
+    },
+    {
+      "epoch": 0.4987940222287762,
+      "grad_norm": 837.0609741210938,
+      "learning_rate": 2.9869554740267724e-05,
+      "loss": 46.9197,
+      "step": 123460
+    },
+    {
+      "epoch": 0.49883442349414386,
+      "grad_norm": 231.35792541503906,
+      "learning_rate": 2.986613089256046e-05,
+      "loss": 62.4719,
+      "step": 123470
+    },
+    {
+      "epoch": 0.49887482475951145,
+      "grad_norm": 848.2174072265625,
+      "learning_rate": 2.9862706949985463e-05,
+      "loss": 63.5692,
+      "step": 123480
+    },
+    {
+      "epoch": 0.4989152260248791,
+      "grad_norm": 482.8079833984375,
+      "learning_rate": 2.9859282912609497e-05,
+      "loss": 48.4904,
+      "step": 123490
+    },
+    {
+      "epoch": 0.49895562729024673,
+      "grad_norm": 786.4293212890625,
+      "learning_rate": 2.98558587804993e-05,
+      "loss": 51.905,
+      "step": 123500
+    },
+    {
+      "epoch": 0.49899602855561437,
+      "grad_norm": 967.361083984375,
+      "learning_rate": 2.9852434553721642e-05,
+      "loss": 55.1196,
+      "step": 123510
+    },
+    {
+      "epoch": 0.499036429820982,
+      "grad_norm": 691.3804321289062,
+      "learning_rate": 2.984901023234327e-05,
+      "loss": 63.2347,
+      "step": 123520
+    },
+    {
+      "epoch": 0.49907683108634965,
+      "grad_norm": 2461.232421875,
+      "learning_rate": 2.9845585816430955e-05,
+      "loss": 86.1027,
+      "step": 123530
+    },
+    {
+      "epoch": 0.49911723235171723,
+      "grad_norm": 625.51220703125,
+      "learning_rate": 2.9842161306051446e-05,
+      "loss": 59.4591,
+      "step": 123540
+    },
+    {
+      "epoch": 0.4991576336170849,
+      "grad_norm": 937.5235595703125,
+      "learning_rate": 2.9838736701271514e-05,
+      "loss": 86.4743,
+      "step": 123550
+    },
+    {
+      "epoch": 0.4991980348824525,
+      "grad_norm": 719.2523803710938,
+      "learning_rate": 2.9835312002157913e-05,
+      "loss": 57.3074,
+      "step": 123560
+    },
+    {
+      "epoch": 0.49923843614782015,
+      "grad_norm": 1146.9417724609375,
+      "learning_rate": 2.983188720877741e-05,
+      "loss": 59.0976,
+      "step": 123570
+    },
+    {
+      "epoch": 0.4992788374131878,
+      "grad_norm": 643.1884155273438,
+      "learning_rate": 2.9828462321196788e-05,
+      "loss": 59.2565,
+      "step": 123580
+    },
+    {
+      "epoch": 0.49931923867855543,
+      "grad_norm": 717.5197143554688,
+      "learning_rate": 2.9825037339482804e-05,
+      "loss": 62.1937,
+      "step": 123590
+    },
+    {
+      "epoch": 0.499359639943923,
+      "grad_norm": 911.0812377929688,
+      "learning_rate": 2.9821612263702226e-05,
+      "loss": 53.7806,
+      "step": 123600
+    },
+    {
+      "epoch": 0.49940004120929066,
+      "grad_norm": 655.2685546875,
+      "learning_rate": 2.981818709392184e-05,
+      "loss": 76.8916,
+      "step": 123610
+    },
+    {
+      "epoch": 0.4994404424746583,
+      "grad_norm": 694.6813354492188,
+      "learning_rate": 2.981476183020842e-05,
+      "loss": 54.2385,
+      "step": 123620
+    },
+    {
+      "epoch": 0.49948084374002594,
+      "grad_norm": 277.7393798828125,
+      "learning_rate": 2.9811336472628737e-05,
+      "loss": 56.6968,
+      "step": 123630
+    },
+    {
+      "epoch": 0.4995212450053936,
+      "grad_norm": 2142.0654296875,
+      "learning_rate": 2.9807911021249573e-05,
+      "loss": 71.4143,
+      "step": 123640
+    },
+    {
+      "epoch": 0.4995616462707612,
+      "grad_norm": 685.448974609375,
+      "learning_rate": 2.9804485476137706e-05,
+      "loss": 41.1674,
+      "step": 123650
+    },
+    {
+      "epoch": 0.49960204753612886,
+      "grad_norm": 0.0,
+      "learning_rate": 2.9801059837359925e-05,
+      "loss": 67.2189,
+      "step": 123660
+    },
+    {
+      "epoch": 0.49964244880149644,
+      "grad_norm": 563.3724975585938,
+      "learning_rate": 2.979763410498301e-05,
+      "loss": 42.097,
+      "step": 123670
+    },
+    {
+      "epoch": 0.4996828500668641,
+      "grad_norm": 823.16259765625,
+      "learning_rate": 2.9794208279073743e-05,
+      "loss": 46.5123,
+      "step": 123680
+    },
+    {
+      "epoch": 0.4997232513322317,
+      "grad_norm": 396.4298095703125,
+      "learning_rate": 2.9790782359698914e-05,
+      "loss": 38.3301,
+      "step": 123690
+    },
+    {
+      "epoch": 0.49976365259759936,
+      "grad_norm": 487.17205810546875,
+      "learning_rate": 2.9787356346925327e-05,
+      "loss": 54.4319,
+      "step": 123700
+    },
+    {
+      "epoch": 0.499804053862967,
+      "grad_norm": 1401.6353759765625,
+      "learning_rate": 2.9783930240819758e-05,
+      "loss": 89.3769,
+      "step": 123710
+    },
+    {
+      "epoch": 0.49984445512833464,
+      "grad_norm": 401.6158447265625,
+      "learning_rate": 2.978050404144901e-05,
+      "loss": 39.3288,
+      "step": 123720
+    },
+    {
+      "epoch": 0.4998848563937022,
+      "grad_norm": 1638.397216796875,
+      "learning_rate": 2.977707774887987e-05,
+      "loss": 85.3148,
+      "step": 123730
+    },
+    {
+      "epoch": 0.49992525765906987,
+      "grad_norm": 525.0819702148438,
+      "learning_rate": 2.9773651363179144e-05,
+      "loss": 62.3121,
+      "step": 123740
+    },
+    {
+      "epoch": 0.4999656589244375,
+      "grad_norm": 576.019775390625,
+      "learning_rate": 2.9770224884413623e-05,
+      "loss": 34.6271,
+      "step": 123750
+    },
+    {
+      "epoch": 0.5000060601898051,
+      "grad_norm": 596.5224609375,
+      "learning_rate": 2.9766798312650112e-05,
+      "loss": 45.5462,
+      "step": 123760
+    },
+    {
+      "epoch": 0.5000464614551727,
+      "grad_norm": 1210.31005859375,
+      "learning_rate": 2.976337164795541e-05,
+      "loss": 43.6167,
+      "step": 123770
+    },
+    {
+      "epoch": 0.5000868627205404,
+      "grad_norm": 1921.7060546875,
+      "learning_rate": 2.975994489039634e-05,
+      "loss": 52.5438,
+      "step": 123780
+    },
+    {
+      "epoch": 0.500127263985908,
+      "grad_norm": 906.3182373046875,
+      "learning_rate": 2.9756518040039682e-05,
+      "loss": 64.2235,
+      "step": 123790
+    },
+    {
+      "epoch": 0.5001676652512757,
+      "grad_norm": 724.0418701171875,
+      "learning_rate": 2.9753091096952255e-05,
+      "loss": 63.5248,
+      "step": 123800
+    },
+    {
+      "epoch": 0.5002080665166433,
+      "grad_norm": 1490.1466064453125,
+      "learning_rate": 2.9749664061200877e-05,
+      "loss": 38.7024,
+      "step": 123810
+    },
+    {
+      "epoch": 0.5002484677820109,
+      "grad_norm": 262.531494140625,
+      "learning_rate": 2.9746236932852355e-05,
+      "loss": 66.6208,
+      "step": 123820
+    },
+    {
+      "epoch": 0.5002888690473786,
+      "grad_norm": 283.681640625,
+      "learning_rate": 2.974280971197349e-05,
+      "loss": 45.1853,
+      "step": 123830
+    },
+    {
+      "epoch": 0.5003292703127462,
+      "grad_norm": 636.9791259765625,
+      "learning_rate": 2.973938239863111e-05,
+      "loss": 54.2482,
+      "step": 123840
+    },
+    {
+      "epoch": 0.5003696715781139,
+      "grad_norm": 1471.0166015625,
+      "learning_rate": 2.9735954992892033e-05,
+      "loss": 73.4811,
+      "step": 123850
+    },
+    {
+      "epoch": 0.5004100728434815,
+      "grad_norm": 870.6261596679688,
+      "learning_rate": 2.9732527494823083e-05,
+      "loss": 69.1818,
+      "step": 123860
+    },
+    {
+      "epoch": 0.5004504741088491,
+      "grad_norm": 1015.466064453125,
+      "learning_rate": 2.9729099904491058e-05,
+      "loss": 65.9109,
+      "step": 123870
+    },
+    {
+      "epoch": 0.5004908753742168,
+      "grad_norm": 769.5751953125,
+      "learning_rate": 2.97256722219628e-05,
+      "loss": 41.2782,
+      "step": 123880
+    },
+    {
+      "epoch": 0.5005312766395843,
+      "grad_norm": 814.8590087890625,
+      "learning_rate": 2.9722244447305135e-05,
+      "loss": 84.8106,
+      "step": 123890
+    },
+    {
+      "epoch": 0.5005716779049519,
+      "grad_norm": 476.1192932128906,
+      "learning_rate": 2.9718816580584884e-05,
+      "loss": 43.7513,
+      "step": 123900
+    },
+    {
+      "epoch": 0.5006120791703196,
+      "grad_norm": 729.9635620117188,
+      "learning_rate": 2.9715388621868873e-05,
+      "loss": 44.9804,
+      "step": 123910
+    },
+    {
+      "epoch": 0.5006524804356872,
+      "grad_norm": 657.6481323242188,
+      "learning_rate": 2.971196057122393e-05,
+      "loss": 68.3173,
+      "step": 123920
+    },
+    {
+      "epoch": 0.5006928817010549,
+      "grad_norm": 620.4969482421875,
+      "learning_rate": 2.9708532428716883e-05,
+      "loss": 50.9084,
+      "step": 123930
+    },
+    {
+      "epoch": 0.5007332829664225,
+      "grad_norm": 840.12890625,
+      "learning_rate": 2.9705104194414586e-05,
+      "loss": 63.0937,
+      "step": 123940
+    },
+    {
+      "epoch": 0.5007736842317901,
+      "grad_norm": 1180.94970703125,
+      "learning_rate": 2.9701675868383848e-05,
+      "loss": 69.3689,
+      "step": 123950
+    },
+    {
+      "epoch": 0.5008140854971578,
+      "grad_norm": 630.738037109375,
+      "learning_rate": 2.9698247450691525e-05,
+      "loss": 56.1486,
+      "step": 123960
+    },
+    {
+      "epoch": 0.5008544867625254,
+      "grad_norm": 1723.3427734375,
+      "learning_rate": 2.9694818941404444e-05,
+      "loss": 52.5067,
+      "step": 123970
+    },
+    {
+      "epoch": 0.5008948880278931,
+      "grad_norm": 1080.7830810546875,
+      "learning_rate": 2.9691390340589466e-05,
+      "loss": 63.3121,
+      "step": 123980
+    },
+    {
+      "epoch": 0.5009352892932607,
+      "grad_norm": 428.3757629394531,
+      "learning_rate": 2.9687961648313405e-05,
+      "loss": 45.4763,
+      "step": 123990
+    },
+    {
+      "epoch": 0.5009756905586283,
+      "grad_norm": 585.084228515625,
+      "learning_rate": 2.9684532864643122e-05,
+      "loss": 47.8538,
+      "step": 124000
+    },
+    {
+      "epoch": 0.501016091823996,
+      "grad_norm": 951.522705078125,
+      "learning_rate": 2.9681103989645453e-05,
+      "loss": 64.0775,
+      "step": 124010
+    },
+    {
+      "epoch": 0.5010564930893635,
+      "grad_norm": 414.710693359375,
+      "learning_rate": 2.9677675023387258e-05,
+      "loss": 35.3562,
+      "step": 124020
+    },
+    {
+      "epoch": 0.5010968943547311,
+      "grad_norm": 1510.4676513671875,
+      "learning_rate": 2.9674245965935378e-05,
+      "loss": 85.7635,
+      "step": 124030
+    },
+    {
+      "epoch": 0.5011372956200988,
+      "grad_norm": 555.8351440429688,
+      "learning_rate": 2.9670816817356668e-05,
+      "loss": 39.8303,
+      "step": 124040
+    },
+    {
+      "epoch": 0.5011776968854664,
+      "grad_norm": 976.0780639648438,
+      "learning_rate": 2.9667387577717976e-05,
+      "loss": 61.9712,
+      "step": 124050
+    },
+    {
+      "epoch": 0.5012180981508341,
+      "grad_norm": 1387.890869140625,
+      "learning_rate": 2.9663958247086166e-05,
+      "loss": 58.7585,
+      "step": 124060
+    },
+    {
+      "epoch": 0.5012584994162017,
+      "grad_norm": 1036.8333740234375,
+      "learning_rate": 2.966052882552809e-05,
+      "loss": 44.4313,
+      "step": 124070
+    },
+    {
+      "epoch": 0.5012989006815693,
+      "grad_norm": 1022.4268798828125,
+      "learning_rate": 2.9657099313110593e-05,
+      "loss": 57.0276,
+      "step": 124080
+    },
+    {
+      "epoch": 0.501339301946937,
+      "grad_norm": 1398.85546875,
+      "learning_rate": 2.9653669709900555e-05,
+      "loss": 58.4346,
+      "step": 124090
+    },
+    {
+      "epoch": 0.5013797032123046,
+      "grad_norm": 496.8158264160156,
+      "learning_rate": 2.9650240015964825e-05,
+      "loss": 62.5083,
+      "step": 124100
+    },
+    {
+      "epoch": 0.5014201044776723,
+      "grad_norm": 911.7360229492188,
+      "learning_rate": 2.964681023137028e-05,
+      "loss": 52.9003,
+      "step": 124110
+    },
+    {
+      "epoch": 0.5014605057430399,
+      "grad_norm": 932.399658203125,
+      "learning_rate": 2.9643380356183775e-05,
+      "loss": 56.112,
+      "step": 124120
+    },
+    {
+      "epoch": 0.5015009070084075,
+      "grad_norm": 0.0,
+      "learning_rate": 2.9639950390472177e-05,
+      "loss": 45.8005,
+      "step": 124130
+    },
+    {
+      "epoch": 0.5015413082737752,
+      "grad_norm": 789.1591796875,
+      "learning_rate": 2.9636520334302354e-05,
+      "loss": 73.2515,
+      "step": 124140
+    },
+    {
+      "epoch": 0.5015817095391427,
+      "grad_norm": 1213.167236328125,
+      "learning_rate": 2.9633090187741185e-05,
+      "loss": 45.1217,
+      "step": 124150
+    },
+    {
+      "epoch": 0.5016221108045104,
+      "grad_norm": 1115.8275146484375,
+      "learning_rate": 2.9629659950855544e-05,
+      "loss": 48.8063,
+      "step": 124160
+    },
+    {
+      "epoch": 0.501662512069878,
+      "grad_norm": 1056.0667724609375,
+      "learning_rate": 2.9626229623712288e-05,
+      "loss": 59.1212,
+      "step": 124170
+    },
+    {
+      "epoch": 0.5017029133352456,
+      "grad_norm": 269.9258728027344,
+      "learning_rate": 2.9622799206378305e-05,
+      "loss": 83.2247,
+      "step": 124180
+    },
+    {
+      "epoch": 0.5017433146006133,
+      "grad_norm": 539.0936889648438,
+      "learning_rate": 2.961936869892048e-05,
+      "loss": 53.4077,
+      "step": 124190
+    },
+    {
+      "epoch": 0.5017837158659809,
+      "grad_norm": 3030.404052734375,
+      "learning_rate": 2.9615938101405676e-05,
+      "loss": 76.7048,
+      "step": 124200
+    },
+    {
+      "epoch": 0.5018241171313486,
+      "grad_norm": 1794.911376953125,
+      "learning_rate": 2.961250741390078e-05,
+      "loss": 60.769,
+      "step": 124210
+    },
+    {
+      "epoch": 0.5018645183967162,
+      "grad_norm": 1027.421875,
+      "learning_rate": 2.960907663647268e-05,
+      "loss": 43.8524,
+      "step": 124220
+    },
+    {
+      "epoch": 0.5019049196620838,
+      "grad_norm": 726.89697265625,
+      "learning_rate": 2.9605645769188268e-05,
+      "loss": 57.9974,
+      "step": 124230
+    },
+    {
+      "epoch": 0.5019453209274515,
+      "grad_norm": 741.7726440429688,
+      "learning_rate": 2.9602214812114415e-05,
+      "loss": 45.0123,
+      "step": 124240
+    },
+    {
+      "epoch": 0.5019857221928191,
+      "grad_norm": 572.0324096679688,
+      "learning_rate": 2.9598783765318007e-05,
+      "loss": 41.7237,
+      "step": 124250
+    },
+    {
+      "epoch": 0.5020261234581868,
+      "grad_norm": 905.0762329101562,
+      "learning_rate": 2.9595352628865947e-05,
+      "loss": 60.8909,
+      "step": 124260
+    },
+    {
+      "epoch": 0.5020665247235544,
+      "grad_norm": 418.3033447265625,
+      "learning_rate": 2.9591921402825123e-05,
+      "loss": 56.9024,
+      "step": 124270
+    },
+    {
+      "epoch": 0.5021069259889219,
+      "grad_norm": 642.4514770507812,
+      "learning_rate": 2.958849008726242e-05,
+      "loss": 56.9975,
+      "step": 124280
+    },
+    {
+      "epoch": 0.5021473272542896,
+      "grad_norm": 545.2382202148438,
+      "learning_rate": 2.9585058682244748e-05,
+      "loss": 54.7912,
+      "step": 124290
+    },
+    {
+      "epoch": 0.5021877285196572,
+      "grad_norm": 1271.816162109375,
+      "learning_rate": 2.9581627187838994e-05,
+      "loss": 67.879,
+      "step": 124300
+    },
+    {
+      "epoch": 0.5022281297850248,
+      "grad_norm": 404.8785095214844,
+      "learning_rate": 2.9578195604112064e-05,
+      "loss": 68.5172,
+      "step": 124310
+    },
+    {
+      "epoch": 0.5022685310503925,
+      "grad_norm": 1688.64599609375,
+      "learning_rate": 2.9574763931130843e-05,
+      "loss": 53.8717,
+      "step": 124320
+    },
+    {
+      "epoch": 0.5023089323157601,
+      "grad_norm": 574.3252563476562,
+      "learning_rate": 2.9571332168962256e-05,
+      "loss": 66.5893,
+      "step": 124330
+    },
+    {
+      "epoch": 0.5023493335811278,
+      "grad_norm": 418.29864501953125,
+      "learning_rate": 2.956790031767319e-05,
+      "loss": 47.6842,
+      "step": 124340
+    },
+    {
+      "epoch": 0.5023897348464954,
+      "grad_norm": 1550.780029296875,
+      "learning_rate": 2.9564468377330556e-05,
+      "loss": 66.8126,
+      "step": 124350
+    },
+    {
+      "epoch": 0.502430136111863,
+      "grad_norm": 5627.90478515625,
+      "learning_rate": 2.956103634800126e-05,
+      "loss": 86.4269,
+      "step": 124360
+    },
+    {
+      "epoch": 0.5024705373772307,
+      "grad_norm": 2004.539306640625,
+      "learning_rate": 2.9557604229752212e-05,
+      "loss": 57.9298,
+      "step": 124370
+    },
+    {
+      "epoch": 0.5025109386425983,
+      "grad_norm": 531.3135986328125,
+      "learning_rate": 2.9554172022650317e-05,
+      "loss": 53.1231,
+      "step": 124380
+    },
+    {
+      "epoch": 0.502551339907966,
+      "grad_norm": 1106.7730712890625,
+      "learning_rate": 2.9550739726762507e-05,
+      "loss": 62.664,
+      "step": 124390
+    },
+    {
+      "epoch": 0.5025917411733335,
+      "grad_norm": 1226.38818359375,
+      "learning_rate": 2.9547307342155673e-05,
+      "loss": 94.7886,
+      "step": 124400
+    },
+    {
+      "epoch": 0.5026321424387011,
+      "grad_norm": 608.1491088867188,
+      "learning_rate": 2.9543874868896747e-05,
+      "loss": 71.896,
+      "step": 124410
+    },
+    {
+      "epoch": 0.5026725437040688,
+      "grad_norm": 240.57858276367188,
+      "learning_rate": 2.954044230705264e-05,
+      "loss": 96.0706,
+      "step": 124420
+    },
+    {
+      "epoch": 0.5027129449694364,
+      "grad_norm": 596.6292114257812,
+      "learning_rate": 2.9537009656690275e-05,
+      "loss": 50.2395,
+      "step": 124430
+    },
+    {
+      "epoch": 0.502753346234804,
+      "grad_norm": 742.9195556640625,
+      "learning_rate": 2.953357691787656e-05,
+      "loss": 39.4977,
+      "step": 124440
+    },
+    {
+      "epoch": 0.5027937475001717,
+      "grad_norm": 715.0127563476562,
+      "learning_rate": 2.9530144090678435e-05,
+      "loss": 40.7906,
+      "step": 124450
+    },
+    {
+      "epoch": 0.5028341487655393,
+      "grad_norm": 494.5076599121094,
+      "learning_rate": 2.952671117516282e-05,
+      "loss": 57.1695,
+      "step": 124460
+    },
+    {
+      "epoch": 0.502874550030907,
+      "grad_norm": 766.6427001953125,
+      "learning_rate": 2.952327817139664e-05,
+      "loss": 62.9475,
+      "step": 124470
+    },
+    {
+      "epoch": 0.5029149512962746,
+      "grad_norm": 670.2885131835938,
+      "learning_rate": 2.9519845079446823e-05,
+      "loss": 47.9745,
+      "step": 124480
+    },
+    {
+      "epoch": 0.5029553525616423,
+      "grad_norm": 636.1236572265625,
+      "learning_rate": 2.9516411899380296e-05,
+      "loss": 57.113,
+      "step": 124490
+    },
+    {
+      "epoch": 0.5029957538270099,
+      "grad_norm": 434.45477294921875,
+      "learning_rate": 2.9512978631264006e-05,
+      "loss": 44.9944,
+      "step": 124500
+    },
+    {
+      "epoch": 0.5030361550923775,
+      "grad_norm": 687.5653686523438,
+      "learning_rate": 2.950954527516487e-05,
+      "loss": 50.9721,
+      "step": 124510
+    },
+    {
+      "epoch": 0.5030765563577452,
+      "grad_norm": 451.1505126953125,
+      "learning_rate": 2.9506111831149818e-05,
+      "loss": 44.7546,
+      "step": 124520
+    },
+    {
+      "epoch": 0.5031169576231127,
+      "grad_norm": 1212.3389892578125,
+      "learning_rate": 2.9502678299285798e-05,
+      "loss": 44.7764,
+      "step": 124530
+    },
+    {
+      "epoch": 0.5031573588884803,
+      "grad_norm": 572.8779907226562,
+      "learning_rate": 2.949924467963975e-05,
+      "loss": 48.557,
+      "step": 124540
+    },
+    {
+      "epoch": 0.503197760153848,
+      "grad_norm": 520.7388916015625,
+      "learning_rate": 2.949581097227861e-05,
+      "loss": 80.5346,
+      "step": 124550
+    },
+    {
+      "epoch": 0.5032381614192156,
+      "grad_norm": 713.9329833984375,
+      "learning_rate": 2.9492377177269315e-05,
+      "loss": 32.9715,
+      "step": 124560
+    },
+    {
+      "epoch": 0.5032785626845833,
+      "grad_norm": 978.047119140625,
+      "learning_rate": 2.9488943294678818e-05,
+      "loss": 52.3525,
+      "step": 124570
+    },
+    {
+      "epoch": 0.5033189639499509,
+      "grad_norm": 1501.214599609375,
+      "learning_rate": 2.948550932457407e-05,
+      "loss": 55.6765,
+      "step": 124580
+    },
+    {
+      "epoch": 0.5033593652153185,
+      "grad_norm": 625.777587890625,
+      "learning_rate": 2.9482075267021995e-05,
+      "loss": 46.8336,
+      "step": 124590
+    },
+    {
+      "epoch": 0.5033997664806862,
+      "grad_norm": 601.8237915039062,
+      "learning_rate": 2.9478641122089562e-05,
+      "loss": 51.4297,
+      "step": 124600
+    },
+    {
+      "epoch": 0.5034401677460538,
+      "grad_norm": 924.4441528320312,
+      "learning_rate": 2.947520688984371e-05,
+      "loss": 60.7378,
+      "step": 124610
+    },
+    {
+      "epoch": 0.5034805690114215,
+      "grad_norm": 516.689453125,
+      "learning_rate": 2.9471772570351398e-05,
+      "loss": 31.4356,
+      "step": 124620
+    },
+    {
+      "epoch": 0.5035209702767891,
+      "grad_norm": 1193.8798828125,
+      "learning_rate": 2.9468338163679577e-05,
+      "loss": 60.812,
+      "step": 124630
+    },
+    {
+      "epoch": 0.5035613715421567,
+      "grad_norm": 663.6973266601562,
+      "learning_rate": 2.9464903669895205e-05,
+      "loss": 35.3694,
+      "step": 124640
+    },
+    {
+      "epoch": 0.5036017728075244,
+      "grad_norm": 1353.3345947265625,
+      "learning_rate": 2.9461469089065234e-05,
+      "loss": 51.9646,
+      "step": 124650
+    },
+    {
+      "epoch": 0.5036421740728919,
+      "grad_norm": 529.1851806640625,
+      "learning_rate": 2.945803442125663e-05,
+      "loss": 48.2301,
+      "step": 124660
+    },
+    {
+      "epoch": 0.5036825753382596,
+      "grad_norm": 331.9324035644531,
+      "learning_rate": 2.9454599666536347e-05,
+      "loss": 54.444,
+      "step": 124670
+    },
+    {
+      "epoch": 0.5037229766036272,
+      "grad_norm": 620.619384765625,
+      "learning_rate": 2.9451164824971356e-05,
+      "loss": 92.8183,
+      "step": 124680
+    },
+    {
+      "epoch": 0.5037633778689948,
+      "grad_norm": 441.77734375,
+      "learning_rate": 2.9447729896628612e-05,
+      "loss": 47.082,
+      "step": 124690
+    },
+    {
+      "epoch": 0.5038037791343625,
+      "grad_norm": 891.2379150390625,
+      "learning_rate": 2.944429488157508e-05,
+      "loss": 70.2383,
+      "step": 124700
+    },
+    {
+      "epoch": 0.5038441803997301,
+      "grad_norm": 1449.03955078125,
+      "learning_rate": 2.9440859779877728e-05,
+      "loss": 57.3963,
+      "step": 124710
+    },
+    {
+      "epoch": 0.5038845816650978,
+      "grad_norm": 1024.277587890625,
+      "learning_rate": 2.943742459160354e-05,
+      "loss": 48.08,
+      "step": 124720
+    },
+    {
+      "epoch": 0.5039249829304654,
+      "grad_norm": 272.8819885253906,
+      "learning_rate": 2.9433989316819467e-05,
+      "loss": 65.6629,
+      "step": 124730
+    },
+    {
+      "epoch": 0.503965384195833,
+      "grad_norm": 665.0719604492188,
+      "learning_rate": 2.943055395559249e-05,
+      "loss": 60.4535,
+      "step": 124740
+    },
+    {
+      "epoch": 0.5040057854612007,
+      "grad_norm": 1616.3466796875,
+      "learning_rate": 2.9427118507989586e-05,
+      "loss": 88.4401,
+      "step": 124750
+    },
+    {
+      "epoch": 0.5040461867265683,
+      "grad_norm": 1536.4730224609375,
+      "learning_rate": 2.942368297407772e-05,
+      "loss": 67.7257,
+      "step": 124760
+    },
+    {
+      "epoch": 0.504086587991936,
+      "grad_norm": 258.504638671875,
+      "learning_rate": 2.942024735392389e-05,
+      "loss": 45.7606,
+      "step": 124770
+    },
+    {
+      "epoch": 0.5041269892573036,
+      "grad_norm": 1484.1357421875,
+      "learning_rate": 2.9416811647595048e-05,
+      "loss": 65.3528,
+      "step": 124780
+    },
+    {
+      "epoch": 0.5041673905226711,
+      "grad_norm": 933.3783569335938,
+      "learning_rate": 2.94133758551582e-05,
+      "loss": 30.6998,
+      "step": 124790
+    },
+    {
+      "epoch": 0.5042077917880388,
+      "grad_norm": 437.1009216308594,
+      "learning_rate": 2.9409939976680313e-05,
+      "loss": 73.0693,
+      "step": 124800
+    },
+    {
+      "epoch": 0.5042481930534064,
+      "grad_norm": 888.6316528320312,
+      "learning_rate": 2.9406504012228375e-05,
+      "loss": 38.9805,
+      "step": 124810
+    },
+    {
+      "epoch": 0.504288594318774,
+      "grad_norm": 454.36627197265625,
+      "learning_rate": 2.9403067961869367e-05,
+      "loss": 32.8442,
+      "step": 124820
+    },
+    {
+      "epoch": 0.5043289955841417,
+      "grad_norm": 648.9022827148438,
+      "learning_rate": 2.9399631825670292e-05,
+      "loss": 40.4288,
+      "step": 124830
+    },
+    {
+      "epoch": 0.5043693968495093,
+      "grad_norm": 967.1868286132812,
+      "learning_rate": 2.939619560369813e-05,
+      "loss": 63.7735,
+      "step": 124840
+    },
+    {
+      "epoch": 0.504409798114877,
+      "grad_norm": 1663.5084228515625,
+      "learning_rate": 2.9392759296019867e-05,
+      "loss": 57.5468,
+      "step": 124850
+    },
+    {
+      "epoch": 0.5044501993802446,
+      "grad_norm": 1367.6448974609375,
+      "learning_rate": 2.9389322902702497e-05,
+      "loss": 61.3237,
+      "step": 124860
+    },
+    {
+      "epoch": 0.5044906006456122,
+      "grad_norm": 756.0995483398438,
+      "learning_rate": 2.9385886423813024e-05,
+      "loss": 65.3766,
+      "step": 124870
+    },
+    {
+      "epoch": 0.5045310019109799,
+      "grad_norm": 630.2203369140625,
+      "learning_rate": 2.938244985941844e-05,
+      "loss": 42.1302,
+      "step": 124880
+    },
+    {
+      "epoch": 0.5045714031763475,
+      "grad_norm": 451.697998046875,
+      "learning_rate": 2.9379013209585726e-05,
+      "loss": 47.7857,
+      "step": 124890
+    },
+    {
+      "epoch": 0.5046118044417152,
+      "grad_norm": 890.1043701171875,
+      "learning_rate": 2.9375576474381905e-05,
+      "loss": 51.9966,
+      "step": 124900
+    },
+    {
+      "epoch": 0.5046522057070827,
+      "grad_norm": 901.4338989257812,
+      "learning_rate": 2.9372139653873958e-05,
+      "loss": 65.1977,
+      "step": 124910
+    },
+    {
+      "epoch": 0.5046926069724503,
+      "grad_norm": 1069.64208984375,
+      "learning_rate": 2.9368702748128912e-05,
+      "loss": 55.9412,
+      "step": 124920
+    },
+    {
+      "epoch": 0.504733008237818,
+      "grad_norm": 1394.1937255859375,
+      "learning_rate": 2.9365265757213745e-05,
+      "loss": 66.5137,
+      "step": 124930
+    },
+    {
+      "epoch": 0.5047734095031856,
+      "grad_norm": 260.0003967285156,
+      "learning_rate": 2.9361828681195484e-05,
+      "loss": 46.3195,
+      "step": 124940
+    },
+    {
+      "epoch": 0.5048138107685533,
+      "grad_norm": 521.545654296875,
+      "learning_rate": 2.9358391520141122e-05,
+      "loss": 54.6765,
+      "step": 124950
+    },
+    {
+      "epoch": 0.5048542120339209,
+      "grad_norm": 676.0771484375,
+      "learning_rate": 2.935495427411768e-05,
+      "loss": 27.8798,
+      "step": 124960
+    },
+    {
+      "epoch": 0.5048946132992885,
+      "grad_norm": 554.6319580078125,
+      "learning_rate": 2.9351516943192155e-05,
+      "loss": 39.7234,
+      "step": 124970
+    },
+    {
+      "epoch": 0.5049350145646562,
+      "grad_norm": 1820.2799072265625,
+      "learning_rate": 2.9348079527431567e-05,
+      "loss": 47.2992,
+      "step": 124980
+    },
+    {
+      "epoch": 0.5049754158300238,
+      "grad_norm": 730.9981689453125,
+      "learning_rate": 2.9344642026902924e-05,
+      "loss": 58.4385,
+      "step": 124990
+    },
+    {
+      "epoch": 0.5050158170953915,
+      "grad_norm": 232.58773803710938,
+      "learning_rate": 2.9341204441673266e-05,
+      "loss": 51.1445,
+      "step": 125000
+    },
+    {
+      "epoch": 0.5050562183607591,
+      "grad_norm": 744.6976318359375,
+      "learning_rate": 2.9337766771809577e-05,
+      "loss": 56.7142,
+      "step": 125010
+    },
+    {
+      "epoch": 0.5050966196261267,
+      "grad_norm": 807.66748046875,
+      "learning_rate": 2.9334329017378898e-05,
+      "loss": 66.2732,
+      "step": 125020
+    },
+    {
+      "epoch": 0.5051370208914944,
+      "grad_norm": 1156.4183349609375,
+      "learning_rate": 2.933089117844824e-05,
+      "loss": 53.764,
+      "step": 125030
+    },
+    {
+      "epoch": 0.5051774221568619,
+      "grad_norm": 1409.17626953125,
+      "learning_rate": 2.9327453255084638e-05,
+      "loss": 51.1774,
+      "step": 125040
+    },
+    {
+      "epoch": 0.5052178234222295,
+      "grad_norm": 1206.735107421875,
+      "learning_rate": 2.9324015247355098e-05,
+      "loss": 50.6902,
+      "step": 125050
+    },
+    {
+      "epoch": 0.5052582246875972,
+      "grad_norm": 1590.0701904296875,
+      "learning_rate": 2.932057715532665e-05,
+      "loss": 61.7914,
+      "step": 125060
+    },
+    {
+      "epoch": 0.5052986259529648,
+      "grad_norm": 905.3733520507812,
+      "learning_rate": 2.9317138979066327e-05,
+      "loss": 44.9613,
+      "step": 125070
+    },
+    {
+      "epoch": 0.5053390272183325,
+      "grad_norm": 747.811279296875,
+      "learning_rate": 2.9313700718641167e-05,
+      "loss": 65.6674,
+      "step": 125080
+    },
+    {
+      "epoch": 0.5053794284837001,
+      "grad_norm": 0.0,
+      "learning_rate": 2.9310262374118185e-05,
+      "loss": 45.0609,
+      "step": 125090
+    },
+    {
+      "epoch": 0.5054198297490677,
+      "grad_norm": 528.493408203125,
+      "learning_rate": 2.9306823945564422e-05,
+      "loss": 58.0582,
+      "step": 125100
+    },
+    {
+      "epoch": 0.5054602310144354,
+      "grad_norm": 680.7048950195312,
+      "learning_rate": 2.9303385433046902e-05,
+      "loss": 36.582,
+      "step": 125110
+    },
+    {
+      "epoch": 0.505500632279803,
+      "grad_norm": 767.6009521484375,
+      "learning_rate": 2.9299946836632673e-05,
+      "loss": 40.1557,
+      "step": 125120
+    },
+    {
+      "epoch": 0.5055410335451707,
+      "grad_norm": 1023.16015625,
+      "learning_rate": 2.929650815638877e-05,
+      "loss": 47.9998,
+      "step": 125130
+    },
+    {
+      "epoch": 0.5055814348105383,
+      "grad_norm": 588.2609252929688,
+      "learning_rate": 2.9293069392382224e-05,
+      "loss": 50.6491,
+      "step": 125140
+    },
+    {
+      "epoch": 0.5056218360759059,
+      "grad_norm": 831.1854858398438,
+      "learning_rate": 2.9289630544680075e-05,
+      "loss": 68.4844,
+      "step": 125150
+    },
+    {
+      "epoch": 0.5056622373412736,
+      "grad_norm": 1181.1595458984375,
+      "learning_rate": 2.9286191613349374e-05,
+      "loss": 84.0284,
+      "step": 125160
+    },
+    {
+      "epoch": 0.5057026386066411,
+      "grad_norm": 435.8255920410156,
+      "learning_rate": 2.9282752598457165e-05,
+      "loss": 58.2437,
+      "step": 125170
+    },
+    {
+      "epoch": 0.5057430398720087,
+      "grad_norm": 727.278564453125,
+      "learning_rate": 2.9279313500070483e-05,
+      "loss": 66.3914,
+      "step": 125180
+    },
+    {
+      "epoch": 0.5057834411373764,
+      "grad_norm": 302.980224609375,
+      "learning_rate": 2.927587431825639e-05,
+      "loss": 44.221,
+      "step": 125190
+    },
+    {
+      "epoch": 0.505823842402744,
+      "grad_norm": 758.2953491210938,
+      "learning_rate": 2.9272435053081922e-05,
+      "loss": 65.575,
+      "step": 125200
+    },
+    {
+      "epoch": 0.5058642436681117,
+      "grad_norm": 181.88172912597656,
+      "learning_rate": 2.9268995704614132e-05,
+      "loss": 40.2345,
+      "step": 125210
+    },
+    {
+      "epoch": 0.5059046449334793,
+      "grad_norm": 920.4985961914062,
+      "learning_rate": 2.926555627292007e-05,
+      "loss": 51.6828,
+      "step": 125220
+    },
+    {
+      "epoch": 0.505945046198847,
+      "grad_norm": 1029.88330078125,
+      "learning_rate": 2.9262116758066793e-05,
+      "loss": 64.6292,
+      "step": 125230
+    },
+    {
+      "epoch": 0.5059854474642146,
+      "grad_norm": 733.74951171875,
+      "learning_rate": 2.9258677160121352e-05,
+      "loss": 53.2531,
+      "step": 125240
+    },
+    {
+      "epoch": 0.5060258487295822,
+      "grad_norm": 2031.1961669921875,
+      "learning_rate": 2.9255237479150816e-05,
+      "loss": 62.1065,
+      "step": 125250
+    },
+    {
+      "epoch": 0.5060662499949499,
+      "grad_norm": 900.5405883789062,
+      "learning_rate": 2.925179771522223e-05,
+      "loss": 52.0671,
+      "step": 125260
+    },
+    {
+      "epoch": 0.5061066512603175,
+      "grad_norm": 554.5584106445312,
+      "learning_rate": 2.924835786840266e-05,
+      "loss": 38.7164,
+      "step": 125270
+    },
+    {
+      "epoch": 0.5061470525256851,
+      "grad_norm": 386.526123046875,
+      "learning_rate": 2.9244917938759163e-05,
+      "loss": 54.9804,
+      "step": 125280
+    },
+    {
+      "epoch": 0.5061874537910528,
+      "grad_norm": 616.0713500976562,
+      "learning_rate": 2.9241477926358818e-05,
+      "loss": 56.0183,
+      "step": 125290
+    },
+    {
+      "epoch": 0.5062278550564203,
+      "grad_norm": 831.8729858398438,
+      "learning_rate": 2.923803783126866e-05,
+      "loss": 46.269,
+      "step": 125300
+    },
+    {
+      "epoch": 0.506268256321788,
+      "grad_norm": 506.3004455566406,
+      "learning_rate": 2.923459765355578e-05,
+      "loss": 44.9364,
+      "step": 125310
+    },
+    {
+      "epoch": 0.5063086575871556,
+      "grad_norm": 662.8009033203125,
+      "learning_rate": 2.9231157393287234e-05,
+      "loss": 44.294,
+      "step": 125320
+    },
+    {
+      "epoch": 0.5063490588525232,
+      "grad_norm": 1044.3468017578125,
+      "learning_rate": 2.9227717050530107e-05,
+      "loss": 59.7929,
+      "step": 125330
+    },
+    {
+      "epoch": 0.5063894601178909,
+      "grad_norm": 836.830322265625,
+      "learning_rate": 2.922427662535145e-05,
+      "loss": 50.6092,
+      "step": 125340
+    },
+    {
+      "epoch": 0.5064298613832585,
+      "grad_norm": 848.728759765625,
+      "learning_rate": 2.9220836117818344e-05,
+      "loss": 46.637,
+      "step": 125350
+    },
+    {
+      "epoch": 0.5064702626486262,
+      "grad_norm": 456.876953125,
+      "learning_rate": 2.9217395527997875e-05,
+      "loss": 45.2742,
+      "step": 125360
+    },
+    {
+      "epoch": 0.5065106639139938,
+      "grad_norm": 780.4397583007812,
+      "learning_rate": 2.921395485595711e-05,
+      "loss": 47.9569,
+      "step": 125370
+    },
+    {
+      "epoch": 0.5065510651793614,
+      "grad_norm": 851.5180053710938,
+      "learning_rate": 2.9210514101763113e-05,
+      "loss": 74.7248,
+      "step": 125380
+    },
+    {
+      "epoch": 0.5065914664447291,
+      "grad_norm": 776.1827392578125,
+      "learning_rate": 2.9207073265482982e-05,
+      "loss": 49.9978,
+      "step": 125390
+    },
+    {
+      "epoch": 0.5066318677100967,
+      "grad_norm": 528.63330078125,
+      "learning_rate": 2.920363234718379e-05,
+      "loss": 73.2618,
+      "step": 125400
+    },
+    {
+      "epoch": 0.5066722689754644,
+      "grad_norm": 1279.852294921875,
+      "learning_rate": 2.9200191346932627e-05,
+      "loss": 48.9025,
+      "step": 125410
+    },
+    {
+      "epoch": 0.506712670240832,
+      "grad_norm": 1353.2442626953125,
+      "learning_rate": 2.919675026479656e-05,
+      "loss": 51.6571,
+      "step": 125420
+    },
+    {
+      "epoch": 0.5067530715061995,
+      "grad_norm": 1191.01416015625,
+      "learning_rate": 2.9193309100842693e-05,
+      "loss": 46.1766,
+      "step": 125430
+    },
+    {
+      "epoch": 0.5067934727715672,
+      "grad_norm": 1029.6173095703125,
+      "learning_rate": 2.9189867855138103e-05,
+      "loss": 58.42,
+      "step": 125440
+    },
+    {
+      "epoch": 0.5068338740369348,
+      "grad_norm": 575.3421630859375,
+      "learning_rate": 2.918642652774989e-05,
+      "loss": 49.3698,
+      "step": 125450
+    },
+    {
+      "epoch": 0.5068742753023024,
+      "grad_norm": 710.7742919921875,
+      "learning_rate": 2.9182985118745136e-05,
+      "loss": 46.0317,
+      "step": 125460
+    },
+    {
+      "epoch": 0.5069146765676701,
+      "grad_norm": 794.8711547851562,
+      "learning_rate": 2.9179543628190925e-05,
+      "loss": 61.0076,
+      "step": 125470
+    },
+    {
+      "epoch": 0.5069550778330377,
+      "grad_norm": 621.3440551757812,
+      "learning_rate": 2.9176102056154363e-05,
+      "loss": 58.3854,
+      "step": 125480
+    },
+    {
+      "epoch": 0.5069954790984054,
+      "grad_norm": 617.9365844726562,
+      "learning_rate": 2.9172660402702546e-05,
+      "loss": 47.3791,
+      "step": 125490
+    },
+    {
+      "epoch": 0.507035880363773,
+      "grad_norm": 902.1868286132812,
+      "learning_rate": 2.916921866790256e-05,
+      "loss": 49.7716,
+      "step": 125500
+    },
+    {
+      "epoch": 0.5070762816291406,
+      "grad_norm": 882.5235595703125,
+      "learning_rate": 2.9165776851821508e-05,
+      "loss": 44.4464,
+      "step": 125510
+    },
+    {
+      "epoch": 0.5071166828945083,
+      "grad_norm": 893.1301879882812,
+      "learning_rate": 2.9162334954526493e-05,
+      "loss": 40.7801,
+      "step": 125520
+    },
+    {
+      "epoch": 0.5071570841598759,
+      "grad_norm": 457.0363464355469,
+      "learning_rate": 2.915889297608462e-05,
+      "loss": 43.9219,
+      "step": 125530
+    },
+    {
+      "epoch": 0.5071974854252436,
+      "grad_norm": 504.5679626464844,
+      "learning_rate": 2.9155450916562994e-05,
+      "loss": 33.9357,
+      "step": 125540
+    },
+    {
+      "epoch": 0.5072378866906111,
+      "grad_norm": 866.373046875,
+      "learning_rate": 2.91520087760287e-05,
+      "loss": 72.3561,
+      "step": 125550
+    },
+    {
+      "epoch": 0.5072782879559787,
+      "grad_norm": 599.7020874023438,
+      "learning_rate": 2.9148566554548857e-05,
+      "loss": 53.9173,
+      "step": 125560
+    },
+    {
+      "epoch": 0.5073186892213464,
+      "grad_norm": 567.0419921875,
+      "learning_rate": 2.914512425219058e-05,
+      "loss": 71.651,
+      "step": 125570
+    },
+    {
+      "epoch": 0.507359090486714,
+      "grad_norm": 628.0565795898438,
+      "learning_rate": 2.914168186902097e-05,
+      "loss": 56.5678,
+      "step": 125580
+    },
+    {
+      "epoch": 0.5073994917520817,
+      "grad_norm": 528.4154663085938,
+      "learning_rate": 2.9138239405107136e-05,
+      "loss": 42.4691,
+      "step": 125590
+    },
+    {
+      "epoch": 0.5074398930174493,
+      "grad_norm": 829.2301635742188,
+      "learning_rate": 2.9134796860516194e-05,
+      "loss": 49.0979,
+      "step": 125600
+    },
+    {
+      "epoch": 0.5074802942828169,
+      "grad_norm": 1135.746826171875,
+      "learning_rate": 2.9131354235315268e-05,
+      "loss": 60.1448,
+      "step": 125610
+    },
+    {
+      "epoch": 0.5075206955481846,
+      "grad_norm": 402.7625427246094,
+      "learning_rate": 2.912791152957145e-05,
+      "loss": 69.6794,
+      "step": 125620
+    },
+    {
+      "epoch": 0.5075610968135522,
+      "grad_norm": 654.6475219726562,
+      "learning_rate": 2.9124468743351884e-05,
+      "loss": 58.2971,
+      "step": 125630
+    },
+    {
+      "epoch": 0.5076014980789199,
+      "grad_norm": 552.7247924804688,
+      "learning_rate": 2.9121025876723674e-05,
+      "loss": 66.4654,
+      "step": 125640
+    },
+    {
+      "epoch": 0.5076418993442875,
+      "grad_norm": 887.902099609375,
+      "learning_rate": 2.9117582929753932e-05,
+      "loss": 76.861,
+      "step": 125650
+    },
+    {
+      "epoch": 0.5076823006096551,
+      "grad_norm": 966.7318725585938,
+      "learning_rate": 2.9114139902509807e-05,
+      "loss": 57.0813,
+      "step": 125660
+    },
+    {
+      "epoch": 0.5077227018750228,
+      "grad_norm": 1011.2301635742188,
+      "learning_rate": 2.9110696795058394e-05,
+      "loss": 45.7711,
+      "step": 125670
+    },
+    {
+      "epoch": 0.5077631031403903,
+      "grad_norm": 557.7918090820312,
+      "learning_rate": 2.9107253607466832e-05,
+      "loss": 55.3871,
+      "step": 125680
+    },
+    {
+      "epoch": 0.5078035044057579,
+      "grad_norm": 3118.763916015625,
+      "learning_rate": 2.910381033980225e-05,
+      "loss": 50.6585,
+      "step": 125690
+    },
+    {
+      "epoch": 0.5078439056711256,
+      "grad_norm": 863.99658203125,
+      "learning_rate": 2.910036699213178e-05,
+      "loss": 68.7242,
+      "step": 125700
+    },
+    {
+      "epoch": 0.5078843069364932,
+      "grad_norm": 1471.3134765625,
+      "learning_rate": 2.909692356452254e-05,
+      "loss": 52.9448,
+      "step": 125710
+    },
+    {
+      "epoch": 0.5079247082018609,
+      "grad_norm": 553.6021118164062,
+      "learning_rate": 2.9093480057041662e-05,
+      "loss": 55.443,
+      "step": 125720
+    },
+    {
+      "epoch": 0.5079651094672285,
+      "grad_norm": 329.5192565917969,
+      "learning_rate": 2.9090036469756276e-05,
+      "loss": 67.455,
+      "step": 125730
+    },
+    {
+      "epoch": 0.5080055107325961,
+      "grad_norm": 1218.230712890625,
+      "learning_rate": 2.9086592802733536e-05,
+      "loss": 84.1644,
+      "step": 125740
+    },
+    {
+      "epoch": 0.5080459119979638,
+      "grad_norm": 140.51890563964844,
+      "learning_rate": 2.908314905604056e-05,
+      "loss": 58.1322,
+      "step": 125750
+    },
+    {
+      "epoch": 0.5080863132633314,
+      "grad_norm": 557.7669677734375,
+      "learning_rate": 2.9079705229744493e-05,
+      "loss": 58.6665,
+      "step": 125760
+    },
+    {
+      "epoch": 0.5081267145286991,
+      "grad_norm": 357.09234619140625,
+      "learning_rate": 2.907626132391246e-05,
+      "loss": 33.4991,
+      "step": 125770
+    },
+    {
+      "epoch": 0.5081671157940667,
+      "grad_norm": 404.6373596191406,
+      "learning_rate": 2.9072817338611636e-05,
+      "loss": 89.8573,
+      "step": 125780
+    },
+    {
+      "epoch": 0.5082075170594343,
+      "grad_norm": 606.2252807617188,
+      "learning_rate": 2.9069373273909123e-05,
+      "loss": 61.2002,
+      "step": 125790
+    },
+    {
+      "epoch": 0.508247918324802,
+      "grad_norm": 458.4631652832031,
+      "learning_rate": 2.9065929129872094e-05,
+      "loss": 46.1727,
+      "step": 125800
+    },
+    {
+      "epoch": 0.5082883195901695,
+      "grad_norm": 901.8555908203125,
+      "learning_rate": 2.906248490656768e-05,
+      "loss": 48.2973,
+      "step": 125810
+    },
+    {
+      "epoch": 0.5083287208555372,
+      "grad_norm": 676.2325439453125,
+      "learning_rate": 2.905904060406303e-05,
+      "loss": 57.2065,
+      "step": 125820
+    },
+    {
+      "epoch": 0.5083691221209048,
+      "grad_norm": 1410.12158203125,
+      "learning_rate": 2.905559622242529e-05,
+      "loss": 45.1312,
+      "step": 125830
+    },
+    {
+      "epoch": 0.5084095233862724,
+      "grad_norm": 386.5525207519531,
+      "learning_rate": 2.9052151761721617e-05,
+      "loss": 73.7721,
+      "step": 125840
+    },
+    {
+      "epoch": 0.5084499246516401,
+      "grad_norm": 1108.298583984375,
+      "learning_rate": 2.9048707222019154e-05,
+      "loss": 66.7324,
+      "step": 125850
+    },
+    {
+      "epoch": 0.5084903259170077,
+      "grad_norm": 752.5873413085938,
+      "learning_rate": 2.904526260338507e-05,
+      "loss": 52.5618,
+      "step": 125860
+    },
+    {
+      "epoch": 0.5085307271823754,
+      "grad_norm": 393.9751281738281,
+      "learning_rate": 2.9041817905886504e-05,
+      "loss": 55.0743,
+      "step": 125870
+    },
+    {
+      "epoch": 0.508571128447743,
+      "grad_norm": 509.97735595703125,
+      "learning_rate": 2.9038373129590622e-05,
+      "loss": 51.6777,
+      "step": 125880
+    },
+    {
+      "epoch": 0.5086115297131106,
+      "grad_norm": 1160.9713134765625,
+      "learning_rate": 2.903492827456457e-05,
+      "loss": 59.5145,
+      "step": 125890
+    },
+    {
+      "epoch": 0.5086519309784783,
+      "grad_norm": 510.33514404296875,
+      "learning_rate": 2.903148334087552e-05,
+      "loss": 34.4883,
+      "step": 125900
+    },
+    {
+      "epoch": 0.5086923322438459,
+      "grad_norm": 778.451171875,
+      "learning_rate": 2.9028038328590617e-05,
+      "loss": 57.813,
+      "step": 125910
+    },
+    {
+      "epoch": 0.5087327335092136,
+      "grad_norm": 604.14990234375,
+      "learning_rate": 2.9024593237777037e-05,
+      "loss": 44.8038,
+      "step": 125920
+    },
+    {
+      "epoch": 0.5087731347745812,
+      "grad_norm": 1083.8634033203125,
+      "learning_rate": 2.902114806850194e-05,
+      "loss": 66.7281,
+      "step": 125930
+    },
+    {
+      "epoch": 0.5088135360399487,
+      "grad_norm": 626.0941772460938,
+      "learning_rate": 2.9017702820832498e-05,
+      "loss": 64.0597,
+      "step": 125940
+    },
+    {
+      "epoch": 0.5088539373053164,
+      "grad_norm": 2549.89453125,
+      "learning_rate": 2.9014257494835862e-05,
+      "loss": 83.0733,
+      "step": 125950
+    },
+    {
+      "epoch": 0.508894338570684,
+      "grad_norm": 271.8528137207031,
+      "learning_rate": 2.901081209057921e-05,
+      "loss": 72.4616,
+      "step": 125960
+    },
+    {
+      "epoch": 0.5089347398360516,
+      "grad_norm": 418.9285583496094,
+      "learning_rate": 2.900736660812972e-05,
+      "loss": 32.0242,
+      "step": 125970
+    },
+    {
+      "epoch": 0.5089751411014193,
+      "grad_norm": 748.1174926757812,
+      "learning_rate": 2.900392104755455e-05,
+      "loss": 62.5196,
+      "step": 125980
+    },
+    {
+      "epoch": 0.5090155423667869,
+      "grad_norm": 1057.87060546875,
+      "learning_rate": 2.900047540892088e-05,
+      "loss": 66.0844,
+      "step": 125990
+    },
+    {
+      "epoch": 0.5090559436321546,
+      "grad_norm": 1053.06689453125,
+      "learning_rate": 2.8997029692295874e-05,
+      "loss": 59.8133,
+      "step": 126000
+    },
+    {
+      "epoch": 0.5090963448975222,
+      "grad_norm": 1008.4521484375,
+      "learning_rate": 2.8993583897746717e-05,
+      "loss": 67.719,
+      "step": 126010
+    },
+    {
+      "epoch": 0.5091367461628898,
+      "grad_norm": 400.8494567871094,
+      "learning_rate": 2.8990138025340596e-05,
+      "loss": 55.9787,
+      "step": 126020
+    },
+    {
+      "epoch": 0.5091771474282575,
+      "grad_norm": 776.619140625,
+      "learning_rate": 2.8986692075144673e-05,
+      "loss": 53.733,
+      "step": 126030
+    },
+    {
+      "epoch": 0.5092175486936251,
+      "grad_norm": 344.3357238769531,
+      "learning_rate": 2.8983246047226135e-05,
+      "loss": 49.7655,
+      "step": 126040
+    },
+    {
+      "epoch": 0.5092579499589928,
+      "grad_norm": 461.5896911621094,
+      "learning_rate": 2.897979994165217e-05,
+      "loss": 66.4076,
+      "step": 126050
+    },
+    {
+      "epoch": 0.5092983512243604,
+      "grad_norm": 0.0,
+      "learning_rate": 2.8976353758489955e-05,
+      "loss": 44.4132,
+      "step": 126060
+    },
+    {
+      "epoch": 0.5093387524897279,
+      "grad_norm": 446.7651062011719,
+      "learning_rate": 2.897290749780667e-05,
+      "loss": 48.4824,
+      "step": 126070
+    },
+    {
+      "epoch": 0.5093791537550956,
+      "grad_norm": 1131.740966796875,
+      "learning_rate": 2.8969461159669513e-05,
+      "loss": 45.0098,
+      "step": 126080
+    },
+    {
+      "epoch": 0.5094195550204632,
+      "grad_norm": 1675.897705078125,
+      "learning_rate": 2.8966014744145663e-05,
+      "loss": 93.9438,
+      "step": 126090
+    },
+    {
+      "epoch": 0.5094599562858309,
+      "grad_norm": 665.18994140625,
+      "learning_rate": 2.8962568251302324e-05,
+      "loss": 52.6255,
+      "step": 126100
+    },
+    {
+      "epoch": 0.5095003575511985,
+      "grad_norm": 404.408203125,
+      "learning_rate": 2.895912168120667e-05,
+      "loss": 57.1605,
+      "step": 126110
+    },
+    {
+      "epoch": 0.5095407588165661,
+      "grad_norm": 690.7555541992188,
+      "learning_rate": 2.8955675033925895e-05,
+      "loss": 63.0719,
+      "step": 126120
+    },
+    {
+      "epoch": 0.5095811600819338,
+      "grad_norm": 937.3909912109375,
+      "learning_rate": 2.89522283095272e-05,
+      "loss": 61.4134,
+      "step": 126130
+    },
+    {
+      "epoch": 0.5096215613473014,
+      "grad_norm": 1157.58642578125,
+      "learning_rate": 2.8948781508077786e-05,
+      "loss": 43.2224,
+      "step": 126140
+    },
+    {
+      "epoch": 0.509661962612669,
+      "grad_norm": 766.99609375,
+      "learning_rate": 2.894533462964485e-05,
+      "loss": 45.9795,
+      "step": 126150
+    },
+    {
+      "epoch": 0.5097023638780367,
+      "grad_norm": 1816.978515625,
+      "learning_rate": 2.894188767429557e-05,
+      "loss": 53.0026,
+      "step": 126160
+    },
+    {
+      "epoch": 0.5097427651434043,
+      "grad_norm": 813.2989501953125,
+      "learning_rate": 2.8938440642097164e-05,
+      "loss": 37.4644,
+      "step": 126170
+    },
+    {
+      "epoch": 0.509783166408772,
+      "grad_norm": 691.785888671875,
+      "learning_rate": 2.893499353311683e-05,
+      "loss": 51.4019,
+      "step": 126180
+    },
+    {
+      "epoch": 0.5098235676741395,
+      "grad_norm": 1490.9388427734375,
+      "learning_rate": 2.8931546347421773e-05,
+      "loss": 45.6698,
+      "step": 126190
+    },
+    {
+      "epoch": 0.5098639689395071,
+      "grad_norm": 548.1686401367188,
+      "learning_rate": 2.8928099085079197e-05,
+      "loss": 41.4219,
+      "step": 126200
+    },
+    {
+      "epoch": 0.5099043702048748,
+      "grad_norm": 1110.064697265625,
+      "learning_rate": 2.89246517461563e-05,
+      "loss": 71.6463,
+      "step": 126210
+    },
+    {
+      "epoch": 0.5099447714702424,
+      "grad_norm": 549.9130859375,
+      "learning_rate": 2.892120433072031e-05,
+      "loss": 45.8994,
+      "step": 126220
+    },
+    {
+      "epoch": 0.5099851727356101,
+      "grad_norm": 4628.6025390625,
+      "learning_rate": 2.8917756838838418e-05,
+      "loss": 75.2251,
+      "step": 126230
+    },
+    {
+      "epoch": 0.5100255740009777,
+      "grad_norm": 518.2777099609375,
+      "learning_rate": 2.8914309270577834e-05,
+      "loss": 48.5844,
+      "step": 126240
+    },
+    {
+      "epoch": 0.5100659752663453,
+      "grad_norm": 677.185302734375,
+      "learning_rate": 2.8910861626005776e-05,
+      "loss": 61.1778,
+      "step": 126250
+    },
+    {
+      "epoch": 0.510106376531713,
+      "grad_norm": 1070.945556640625,
+      "learning_rate": 2.8907413905189456e-05,
+      "loss": 43.0625,
+      "step": 126260
+    },
+    {
+      "epoch": 0.5101467777970806,
+      "grad_norm": 1414.0343017578125,
+      "learning_rate": 2.8903966108196096e-05,
+      "loss": 69.5534,
+      "step": 126270
+    },
+    {
+      "epoch": 0.5101871790624483,
+      "grad_norm": 1134.17041015625,
+      "learning_rate": 2.8900518235092905e-05,
+      "loss": 73.4212,
+      "step": 126280
+    },
+    {
+      "epoch": 0.5102275803278159,
+      "grad_norm": 970.3851318359375,
+      "learning_rate": 2.8897070285947098e-05,
+      "loss": 58.0343,
+      "step": 126290
+    },
+    {
+      "epoch": 0.5102679815931835,
+      "grad_norm": 340.81231689453125,
+      "learning_rate": 2.8893622260825904e-05,
+      "loss": 76.9128,
+      "step": 126300
+    },
+    {
+      "epoch": 0.5103083828585512,
+      "grad_norm": 637.6676025390625,
+      "learning_rate": 2.889017415979654e-05,
+      "loss": 68.674,
+      "step": 126310
+    },
+    {
+      "epoch": 0.5103487841239187,
+      "grad_norm": 643.8228759765625,
+      "learning_rate": 2.8886725982926232e-05,
+      "loss": 81.6261,
+      "step": 126320
+    },
+    {
+      "epoch": 0.5103891853892863,
+      "grad_norm": 378.8822021484375,
+      "learning_rate": 2.8883277730282194e-05,
+      "loss": 46.2597,
+      "step": 126330
+    },
+    {
+      "epoch": 0.510429586654654,
+      "grad_norm": 1168.7510986328125,
+      "learning_rate": 2.8879829401931652e-05,
+      "loss": 71.7926,
+      "step": 126340
+    },
+    {
+      "epoch": 0.5104699879200216,
+      "grad_norm": 522.0606689453125,
+      "learning_rate": 2.8876380997941847e-05,
+      "loss": 54.6862,
+      "step": 126350
+    },
+    {
+      "epoch": 0.5105103891853893,
+      "grad_norm": 396.95263671875,
+      "learning_rate": 2.8872932518379997e-05,
+      "loss": 44.9638,
+      "step": 126360
+    },
+    {
+      "epoch": 0.5105507904507569,
+      "grad_norm": 794.6138305664062,
+      "learning_rate": 2.886948396331333e-05,
+      "loss": 76.7646,
+      "step": 126370
+    },
+    {
+      "epoch": 0.5105911917161245,
+      "grad_norm": 1046.785888671875,
+      "learning_rate": 2.8866035332809084e-05,
+      "loss": 53.4936,
+      "step": 126380
+    },
+    {
+      "epoch": 0.5106315929814922,
+      "grad_norm": 869.3177490234375,
+      "learning_rate": 2.886258662693449e-05,
+      "loss": 43.9115,
+      "step": 126390
+    },
+    {
+      "epoch": 0.5106719942468598,
+      "grad_norm": 724.6814575195312,
+      "learning_rate": 2.8859137845756784e-05,
+      "loss": 45.5164,
+      "step": 126400
+    },
+    {
+      "epoch": 0.5107123955122275,
+      "grad_norm": 411.1709289550781,
+      "learning_rate": 2.8855688989343193e-05,
+      "loss": 57.5156,
+      "step": 126410
+    },
+    {
+      "epoch": 0.5107527967775951,
+      "grad_norm": 326.8053283691406,
+      "learning_rate": 2.885224005776096e-05,
+      "loss": 51.5223,
+      "step": 126420
+    },
+    {
+      "epoch": 0.5107931980429627,
+      "grad_norm": 843.7996215820312,
+      "learning_rate": 2.884879105107733e-05,
+      "loss": 44.6353,
+      "step": 126430
+    },
+    {
+      "epoch": 0.5108335993083304,
+      "grad_norm": 1049.7962646484375,
+      "learning_rate": 2.884534196935953e-05,
+      "loss": 58.139,
+      "step": 126440
+    },
+    {
+      "epoch": 0.5108740005736979,
+      "grad_norm": 2668.488037109375,
+      "learning_rate": 2.8841892812674808e-05,
+      "loss": 37.7585,
+      "step": 126450
+    },
+    {
+      "epoch": 0.5109144018390656,
+      "grad_norm": 1897.6781005859375,
+      "learning_rate": 2.8838443581090412e-05,
+      "loss": 50.773,
+      "step": 126460
+    },
+    {
+      "epoch": 0.5109548031044332,
+      "grad_norm": 556.2504272460938,
+      "learning_rate": 2.8834994274673582e-05,
+      "loss": 56.0956,
+      "step": 126470
+    },
+    {
+      "epoch": 0.5109952043698008,
+      "grad_norm": 292.6050109863281,
+      "learning_rate": 2.8831544893491563e-05,
+      "loss": 49.7071,
+      "step": 126480
+    },
+    {
+      "epoch": 0.5110356056351685,
+      "grad_norm": 442.0137939453125,
+      "learning_rate": 2.882809543761161e-05,
+      "loss": 51.5451,
+      "step": 126490
+    },
+    {
+      "epoch": 0.5110760069005361,
+      "grad_norm": 810.0142211914062,
+      "learning_rate": 2.8824645907100954e-05,
+      "loss": 55.3643,
+      "step": 126500
+    },
+    {
+      "epoch": 0.5111164081659038,
+      "grad_norm": 1782.1546630859375,
+      "learning_rate": 2.8821196302026863e-05,
+      "loss": 68.3655,
+      "step": 126510
+    },
+    {
+      "epoch": 0.5111568094312714,
+      "grad_norm": 900.7822875976562,
+      "learning_rate": 2.881774662245658e-05,
+      "loss": 48.6357,
+      "step": 126520
+    },
+    {
+      "epoch": 0.511197210696639,
+      "grad_norm": 648.9537963867188,
+      "learning_rate": 2.8814296868457364e-05,
+      "loss": 51.299,
+      "step": 126530
+    },
+    {
+      "epoch": 0.5112376119620067,
+      "grad_norm": 949.2380981445312,
+      "learning_rate": 2.8810847040096467e-05,
+      "loss": 66.241,
+      "step": 126540
+    },
+    {
+      "epoch": 0.5112780132273743,
+      "grad_norm": 227.76019287109375,
+      "learning_rate": 2.8807397137441145e-05,
+      "loss": 32.5676,
+      "step": 126550
+    },
+    {
+      "epoch": 0.511318414492742,
+      "grad_norm": 545.3878173828125,
+      "learning_rate": 2.8803947160558652e-05,
+      "loss": 77.758,
+      "step": 126560
+    },
+    {
+      "epoch": 0.5113588157581096,
+      "grad_norm": 621.6892700195312,
+      "learning_rate": 2.8800497109516263e-05,
+      "loss": 58.1789,
+      "step": 126570
+    },
+    {
+      "epoch": 0.5113992170234771,
+      "grad_norm": 1468.47509765625,
+      "learning_rate": 2.8797046984381208e-05,
+      "loss": 72.7109,
+      "step": 126580
+    },
+    {
+      "epoch": 0.5114396182888448,
+      "grad_norm": 406.4592590332031,
+      "learning_rate": 2.8793596785220783e-05,
+      "loss": 64.9677,
+      "step": 126590
+    },
+    {
+      "epoch": 0.5114800195542124,
+      "grad_norm": 290.8427429199219,
+      "learning_rate": 2.879014651210223e-05,
+      "loss": 47.5232,
+      "step": 126600
+    },
+    {
+      "epoch": 0.51152042081958,
+      "grad_norm": 1523.126220703125,
+      "learning_rate": 2.8786696165092812e-05,
+      "loss": 42.1534,
+      "step": 126610
+    },
+    {
+      "epoch": 0.5115608220849477,
+      "grad_norm": 1009.9312133789062,
+      "learning_rate": 2.8783245744259806e-05,
+      "loss": 61.5785,
+      "step": 126620
+    },
+    {
+      "epoch": 0.5116012233503153,
+      "grad_norm": 527.927490234375,
+      "learning_rate": 2.877979524967048e-05,
+      "loss": 61.9816,
+      "step": 126630
+    },
+    {
+      "epoch": 0.511641624615683,
+      "grad_norm": 575.1068725585938,
+      "learning_rate": 2.8776344681392105e-05,
+      "loss": 60.9859,
+      "step": 126640
+    },
+    {
+      "epoch": 0.5116820258810506,
+      "grad_norm": 338.4177551269531,
+      "learning_rate": 2.877289403949194e-05,
+      "loss": 54.107,
+      "step": 126650
+    },
+    {
+      "epoch": 0.5117224271464182,
+      "grad_norm": 354.56005859375,
+      "learning_rate": 2.876944332403726e-05,
+      "loss": 54.2842,
+      "step": 126660
+    },
+    {
+      "epoch": 0.5117628284117859,
+      "grad_norm": 989.342041015625,
+      "learning_rate": 2.8765992535095345e-05,
+      "loss": 100.1021,
+      "step": 126670
+    },
+    {
+      "epoch": 0.5118032296771535,
+      "grad_norm": 547.06982421875,
+      "learning_rate": 2.8762541672733472e-05,
+      "loss": 48.5291,
+      "step": 126680
+    },
+    {
+      "epoch": 0.5118436309425212,
+      "grad_norm": 1068.9014892578125,
+      "learning_rate": 2.8759090737018902e-05,
+      "loss": 58.7899,
+      "step": 126690
+    },
+    {
+      "epoch": 0.5118840322078888,
+      "grad_norm": 1199.883544921875,
+      "learning_rate": 2.875563972801893e-05,
+      "loss": 43.4314,
+      "step": 126700
+    },
+    {
+      "epoch": 0.5119244334732563,
+      "grad_norm": 616.8649291992188,
+      "learning_rate": 2.8752188645800822e-05,
+      "loss": 46.0097,
+      "step": 126710
+    },
+    {
+      "epoch": 0.511964834738624,
+      "grad_norm": 2177.439697265625,
+      "learning_rate": 2.874873749043187e-05,
+      "loss": 33.2443,
+      "step": 126720
+    },
+    {
+      "epoch": 0.5120052360039916,
+      "grad_norm": 3227.6328125,
+      "learning_rate": 2.8745286261979348e-05,
+      "loss": 64.0699,
+      "step": 126730
+    },
+    {
+      "epoch": 0.5120456372693593,
+      "grad_norm": 754.943115234375,
+      "learning_rate": 2.874183496051055e-05,
+      "loss": 40.789,
+      "step": 126740
+    },
+    {
+      "epoch": 0.5120860385347269,
+      "grad_norm": 560.1809692382812,
+      "learning_rate": 2.8738383586092745e-05,
+      "loss": 42.9098,
+      "step": 126750
+    },
+    {
+      "epoch": 0.5121264398000945,
+      "grad_norm": 1200.2210693359375,
+      "learning_rate": 2.8734932138793225e-05,
+      "loss": 58.1281,
+      "step": 126760
+    },
+    {
+      "epoch": 0.5121668410654622,
+      "grad_norm": 569.3571166992188,
+      "learning_rate": 2.8731480618679285e-05,
+      "loss": 71.4217,
+      "step": 126770
+    },
+    {
+      "epoch": 0.5122072423308298,
+      "grad_norm": 1209.4700927734375,
+      "learning_rate": 2.8728029025818204e-05,
+      "loss": 51.1106,
+      "step": 126780
+    },
+    {
+      "epoch": 0.5122476435961975,
+      "grad_norm": 541.9495849609375,
+      "learning_rate": 2.872457736027728e-05,
+      "loss": 42.31,
+      "step": 126790
+    },
+    {
+      "epoch": 0.5122880448615651,
+      "grad_norm": 1008.5130004882812,
+      "learning_rate": 2.8721125622123806e-05,
+      "loss": 101.0686,
+      "step": 126800
+    },
+    {
+      "epoch": 0.5123284461269327,
+      "grad_norm": 794.3250732421875,
+      "learning_rate": 2.8717673811425072e-05,
+      "loss": 56.3601,
+      "step": 126810
+    },
+    {
+      "epoch": 0.5123688473923004,
+      "grad_norm": 796.3021240234375,
+      "learning_rate": 2.8714221928248368e-05,
+      "loss": 50.3433,
+      "step": 126820
+    },
+    {
+      "epoch": 0.5124092486576679,
+      "grad_norm": 536.2626953125,
+      "learning_rate": 2.8710769972661e-05,
+      "loss": 53.0116,
+      "step": 126830
+    },
+    {
+      "epoch": 0.5124496499230355,
+      "grad_norm": 816.8209228515625,
+      "learning_rate": 2.8707317944730268e-05,
+      "loss": 42.8337,
+      "step": 126840
+    },
+    {
+      "epoch": 0.5124900511884032,
+      "grad_norm": 1031.8271484375,
+      "learning_rate": 2.8703865844523452e-05,
+      "loss": 47.6963,
+      "step": 126850
+    },
+    {
+      "epoch": 0.5125304524537708,
+      "grad_norm": 1931.7418212890625,
+      "learning_rate": 2.8700413672107866e-05,
+      "loss": 52.6271,
+      "step": 126860
+    },
+    {
+      "epoch": 0.5125708537191385,
+      "grad_norm": 747.9002075195312,
+      "learning_rate": 2.869696142755081e-05,
+      "loss": 57.4997,
+      "step": 126870
+    },
+    {
+      "epoch": 0.5126112549845061,
+      "grad_norm": 460.6169128417969,
+      "learning_rate": 2.8693509110919598e-05,
+      "loss": 49.3172,
+      "step": 126880
+    },
+    {
+      "epoch": 0.5126516562498737,
+      "grad_norm": 395.9313659667969,
+      "learning_rate": 2.8690056722281513e-05,
+      "loss": 50.4048,
+      "step": 126890
+    },
+    {
+      "epoch": 0.5126920575152414,
+      "grad_norm": 673.880859375,
+      "learning_rate": 2.8686604261703875e-05,
+      "loss": 44.8503,
+      "step": 126900
+    },
+    {
+      "epoch": 0.512732458780609,
+      "grad_norm": 745.4364624023438,
+      "learning_rate": 2.8683151729253994e-05,
+      "loss": 50.778,
+      "step": 126910
+    },
+    {
+      "epoch": 0.5127728600459767,
+      "grad_norm": 887.1217041015625,
+      "learning_rate": 2.8679699124999166e-05,
+      "loss": 44.0796,
+      "step": 126920
+    },
+    {
+      "epoch": 0.5128132613113443,
+      "grad_norm": 565.948486328125,
+      "learning_rate": 2.8676246449006715e-05,
+      "loss": 79.4067,
+      "step": 126930
+    },
+    {
+      "epoch": 0.5128536625767119,
+      "grad_norm": 602.2057495117188,
+      "learning_rate": 2.8672793701343946e-05,
+      "loss": 53.1035,
+      "step": 126940
+    },
+    {
+      "epoch": 0.5128940638420796,
+      "grad_norm": 670.5474243164062,
+      "learning_rate": 2.8669340882078166e-05,
+      "loss": 60.5767,
+      "step": 126950
+    },
+    {
+      "epoch": 0.5129344651074471,
+      "grad_norm": 447.6339111328125,
+      "learning_rate": 2.866588799127671e-05,
+      "loss": 67.3426,
+      "step": 126960
+    },
+    {
+      "epoch": 0.5129748663728148,
+      "grad_norm": 728.9623413085938,
+      "learning_rate": 2.8662435029006868e-05,
+      "loss": 72.0813,
+      "step": 126970
+    },
+    {
+      "epoch": 0.5130152676381824,
+      "grad_norm": 600.4077758789062,
+      "learning_rate": 2.865898199533597e-05,
+      "loss": 54.8805,
+      "step": 126980
+    },
+    {
+      "epoch": 0.51305566890355,
+      "grad_norm": 580.8352661132812,
+      "learning_rate": 2.865552889033134e-05,
+      "loss": 55.582,
+      "step": 126990
+    },
+    {
+      "epoch": 0.5130960701689177,
+      "grad_norm": 1138.0733642578125,
+      "learning_rate": 2.8652075714060295e-05,
+      "loss": 64.3382,
+      "step": 127000
+    },
+    {
+      "epoch": 0.5131364714342853,
+      "grad_norm": 281.50640869140625,
+      "learning_rate": 2.864862246659015e-05,
+      "loss": 45.8135,
+      "step": 127010
+    },
+    {
+      "epoch": 0.513176872699653,
+      "grad_norm": 1133.2386474609375,
+      "learning_rate": 2.8645169147988226e-05,
+      "loss": 56.7795,
+      "step": 127020
+    },
+    {
+      "epoch": 0.5132172739650206,
+      "grad_norm": 846.5953979492188,
+      "learning_rate": 2.8641715758321857e-05,
+      "loss": 31.9575,
+      "step": 127030
+    },
+    {
+      "epoch": 0.5132576752303882,
+      "grad_norm": 3238.64013671875,
+      "learning_rate": 2.8638262297658368e-05,
+      "loss": 78.6456,
+      "step": 127040
+    },
+    {
+      "epoch": 0.5132980764957559,
+      "grad_norm": 1550.1759033203125,
+      "learning_rate": 2.863480876606508e-05,
+      "loss": 56.8611,
+      "step": 127050
+    },
+    {
+      "epoch": 0.5133384777611235,
+      "grad_norm": 442.9187927246094,
+      "learning_rate": 2.863135516360932e-05,
+      "loss": 41.2573,
+      "step": 127060
+    },
+    {
+      "epoch": 0.5133788790264912,
+      "grad_norm": 1137.1116943359375,
+      "learning_rate": 2.8627901490358422e-05,
+      "loss": 46.5267,
+      "step": 127070
+    },
+    {
+      "epoch": 0.5134192802918588,
+      "grad_norm": 1043.6861572265625,
+      "learning_rate": 2.8624447746379722e-05,
+      "loss": 60.2835,
+      "step": 127080
+    },
+    {
+      "epoch": 0.5134596815572263,
+      "grad_norm": 739.7015380859375,
+      "learning_rate": 2.862099393174055e-05,
+      "loss": 48.624,
+      "step": 127090
+    },
+    {
+      "epoch": 0.513500082822594,
+      "grad_norm": 945.224853515625,
+      "learning_rate": 2.861754004650823e-05,
+      "loss": 63.0741,
+      "step": 127100
+    },
+    {
+      "epoch": 0.5135404840879616,
+      "grad_norm": 425.7527160644531,
+      "learning_rate": 2.8614086090750103e-05,
+      "loss": 46.5734,
+      "step": 127110
+    },
+    {
+      "epoch": 0.5135808853533292,
+      "grad_norm": 1409.0302734375,
+      "learning_rate": 2.8610632064533517e-05,
+      "loss": 47.2388,
+      "step": 127120
+    },
+    {
+      "epoch": 0.5136212866186969,
+      "grad_norm": 581.88916015625,
+      "learning_rate": 2.8607177967925792e-05,
+      "loss": 69.2831,
+      "step": 127130
+    },
+    {
+      "epoch": 0.5136616878840645,
+      "grad_norm": 478.259765625,
+      "learning_rate": 2.8603723800994275e-05,
+      "loss": 60.1264,
+      "step": 127140
+    },
+    {
+      "epoch": 0.5137020891494322,
+      "grad_norm": 2480.73095703125,
+      "learning_rate": 2.8600269563806302e-05,
+      "loss": 68.6383,
+      "step": 127150
+    },
+    {
+      "epoch": 0.5137424904147998,
+      "grad_norm": 673.7091064453125,
+      "learning_rate": 2.859681525642923e-05,
+      "loss": 65.6581,
+      "step": 127160
+    },
+    {
+      "epoch": 0.5137828916801674,
+      "grad_norm": 675.0384521484375,
+      "learning_rate": 2.8593360878930392e-05,
+      "loss": 57.0633,
+      "step": 127170
+    },
+    {
+      "epoch": 0.5138232929455351,
+      "grad_norm": 613.3917236328125,
+      "learning_rate": 2.8589906431377134e-05,
+      "loss": 50.5051,
+      "step": 127180
+    },
+    {
+      "epoch": 0.5138636942109027,
+      "grad_norm": 358.6791687011719,
+      "learning_rate": 2.8586451913836797e-05,
+      "loss": 48.4163,
+      "step": 127190
+    },
+    {
+      "epoch": 0.5139040954762704,
+      "grad_norm": 1500.3143310546875,
+      "learning_rate": 2.858299732637674e-05,
+      "loss": 69.6514,
+      "step": 127200
+    },
+    {
+      "epoch": 0.513944496741638,
+      "grad_norm": 303.2386474609375,
+      "learning_rate": 2.8579542669064296e-05,
+      "loss": 46.2417,
+      "step": 127210
+    },
+    {
+      "epoch": 0.5139848980070055,
+      "grad_norm": 936.28662109375,
+      "learning_rate": 2.8576087941966835e-05,
+      "loss": 59.2663,
+      "step": 127220
+    },
+    {
+      "epoch": 0.5140252992723732,
+      "grad_norm": 550.0284423828125,
+      "learning_rate": 2.857263314515169e-05,
+      "loss": 46.5445,
+      "step": 127230
+    },
+    {
+      "epoch": 0.5140657005377408,
+      "grad_norm": 1154.9842529296875,
+      "learning_rate": 2.856917827868622e-05,
+      "loss": 48.4179,
+      "step": 127240
+    },
+    {
+      "epoch": 0.5141061018031085,
+      "grad_norm": 400.9862365722656,
+      "learning_rate": 2.8565723342637796e-05,
+      "loss": 84.6768,
+      "step": 127250
+    },
+    {
+      "epoch": 0.5141465030684761,
+      "grad_norm": 1014.5621948242188,
+      "learning_rate": 2.856226833707375e-05,
+      "loss": 60.3505,
+      "step": 127260
+    },
+    {
+      "epoch": 0.5141869043338437,
+      "grad_norm": 351.1904602050781,
+      "learning_rate": 2.855881326206145e-05,
+      "loss": 59.1501,
+      "step": 127270
+    },
+    {
+      "epoch": 0.5142273055992114,
+      "grad_norm": 520.452392578125,
+      "learning_rate": 2.855535811766825e-05,
+      "loss": 60.3091,
+      "step": 127280
+    },
+    {
+      "epoch": 0.514267706864579,
+      "grad_norm": 717.9920043945312,
+      "learning_rate": 2.8551902903961526e-05,
+      "loss": 40.0301,
+      "step": 127290
+    },
+    {
+      "epoch": 0.5143081081299467,
+      "grad_norm": 560.2972412109375,
+      "learning_rate": 2.854844762100861e-05,
+      "loss": 50.7043,
+      "step": 127300
+    },
+    {
+      "epoch": 0.5143485093953143,
+      "grad_norm": 1860.7342529296875,
+      "learning_rate": 2.854499226887689e-05,
+      "loss": 63.0763,
+      "step": 127310
+    },
+    {
+      "epoch": 0.5143889106606819,
+      "grad_norm": 135.43519592285156,
+      "learning_rate": 2.8541536847633717e-05,
+      "loss": 45.6506,
+      "step": 127320
+    },
+    {
+      "epoch": 0.5144293119260496,
+      "grad_norm": 766.4346313476562,
+      "learning_rate": 2.8538081357346465e-05,
+      "loss": 72.0964,
+      "step": 127330
+    },
+    {
+      "epoch": 0.5144697131914172,
+      "grad_norm": 778.751220703125,
+      "learning_rate": 2.8534625798082488e-05,
+      "loss": 65.8823,
+      "step": 127340
+    },
+    {
+      "epoch": 0.5145101144567847,
+      "grad_norm": 490.54461669921875,
+      "learning_rate": 2.853117016990917e-05,
+      "loss": 41.0561,
+      "step": 127350
+    },
+    {
+      "epoch": 0.5145505157221524,
+      "grad_norm": 450.0224914550781,
+      "learning_rate": 2.8527714472893862e-05,
+      "loss": 61.406,
+      "step": 127360
+    },
+    {
+      "epoch": 0.51459091698752,
+      "grad_norm": 298.4858093261719,
+      "learning_rate": 2.8524258707103957e-05,
+      "loss": 88.5857,
+      "step": 127370
+    },
+    {
+      "epoch": 0.5146313182528877,
+      "grad_norm": 1860.1826171875,
+      "learning_rate": 2.85208028726068e-05,
+      "loss": 108.9013,
+      "step": 127380
+    },
+    {
+      "epoch": 0.5146717195182553,
+      "grad_norm": 948.806396484375,
+      "learning_rate": 2.8517346969469782e-05,
+      "loss": 47.6132,
+      "step": 127390
+    },
+    {
+      "epoch": 0.5147121207836229,
+      "grad_norm": 1325.517578125,
+      "learning_rate": 2.8513890997760272e-05,
+      "loss": 45.2896,
+      "step": 127400
+    },
+    {
+      "epoch": 0.5147525220489906,
+      "grad_norm": 605.529052734375,
+      "learning_rate": 2.851043495754566e-05,
+      "loss": 43.9705,
+      "step": 127410
+    },
+    {
+      "epoch": 0.5147929233143582,
+      "grad_norm": 426.92193603515625,
+      "learning_rate": 2.8506978848893302e-05,
+      "loss": 39.5349,
+      "step": 127420
+    },
+    {
+      "epoch": 0.5148333245797259,
+      "grad_norm": 1157.5235595703125,
+      "learning_rate": 2.8503522671870585e-05,
+      "loss": 56.4158,
+      "step": 127430
+    },
+    {
+      "epoch": 0.5148737258450935,
+      "grad_norm": 782.9813232421875,
+      "learning_rate": 2.8500066426544896e-05,
+      "loss": 85.5858,
+      "step": 127440
+    },
+    {
+      "epoch": 0.5149141271104611,
+      "grad_norm": 1053.0633544921875,
+      "learning_rate": 2.849661011298361e-05,
+      "loss": 64.2638,
+      "step": 127450
+    },
+    {
+      "epoch": 0.5149545283758288,
+      "grad_norm": 652.576171875,
+      "learning_rate": 2.8493153731254102e-05,
+      "loss": 43.2801,
+      "step": 127460
+    },
+    {
+      "epoch": 0.5149949296411963,
+      "grad_norm": 1399.2457275390625,
+      "learning_rate": 2.8489697281423767e-05,
+      "loss": 57.9567,
+      "step": 127470
+    },
+    {
+      "epoch": 0.515035330906564,
+      "grad_norm": 1636.6573486328125,
+      "learning_rate": 2.8486240763559986e-05,
+      "loss": 87.6768,
+      "step": 127480
+    },
+    {
+      "epoch": 0.5150757321719316,
+      "grad_norm": 0.0,
+      "learning_rate": 2.848278417773015e-05,
+      "loss": 68.869,
+      "step": 127490
+    },
+    {
+      "epoch": 0.5151161334372992,
+      "grad_norm": 499.08416748046875,
+      "learning_rate": 2.8479327524001636e-05,
+      "loss": 44.9667,
+      "step": 127500
+    },
+    {
+      "epoch": 0.5151565347026669,
+      "grad_norm": 905.2075805664062,
+      "learning_rate": 2.8475870802441844e-05,
+      "loss": 51.4555,
+      "step": 127510
+    },
+    {
+      "epoch": 0.5151969359680345,
+      "grad_norm": 567.349365234375,
+      "learning_rate": 2.847241401311817e-05,
+      "loss": 37.061,
+      "step": 127520
+    },
+    {
+      "epoch": 0.5152373372334021,
+      "grad_norm": 499.4542541503906,
+      "learning_rate": 2.846895715609799e-05,
+      "loss": 46.9687,
+      "step": 127530
+    },
+    {
+      "epoch": 0.5152777384987698,
+      "grad_norm": 1374.189208984375,
+      "learning_rate": 2.8465500231448704e-05,
+      "loss": 74.0274,
+      "step": 127540
+    },
+    {
+      "epoch": 0.5153181397641374,
+      "grad_norm": 1301.90673828125,
+      "learning_rate": 2.8462043239237707e-05,
+      "loss": 51.4661,
+      "step": 127550
+    },
+    {
+      "epoch": 0.5153585410295051,
+      "grad_norm": 290.5340576171875,
+      "learning_rate": 2.845858617953239e-05,
+      "loss": 49.4092,
+      "step": 127560
+    },
+    {
+      "epoch": 0.5153989422948727,
+      "grad_norm": 617.203857421875,
+      "learning_rate": 2.8455129052400166e-05,
+      "loss": 48.6922,
+      "step": 127570
+    },
+    {
+      "epoch": 0.5154393435602403,
+      "grad_norm": 877.1514282226562,
+      "learning_rate": 2.8451671857908415e-05,
+      "loss": 57.4519,
+      "step": 127580
+    },
+    {
+      "epoch": 0.515479744825608,
+      "grad_norm": 1293.734375,
+      "learning_rate": 2.844821459612454e-05,
+      "loss": 97.9196,
+      "step": 127590
+    },
+    {
+      "epoch": 0.5155201460909755,
+      "grad_norm": 628.9718017578125,
+      "learning_rate": 2.844475726711595e-05,
+      "loss": 44.7918,
+      "step": 127600
+    },
+    {
+      "epoch": 0.5155605473563432,
+      "grad_norm": 645.7540283203125,
+      "learning_rate": 2.844129987095005e-05,
+      "loss": 42.1978,
+      "step": 127610
+    },
+    {
+      "epoch": 0.5156009486217108,
+      "grad_norm": 470.0179138183594,
+      "learning_rate": 2.8437842407694236e-05,
+      "loss": 59.6109,
+      "step": 127620
+    },
+    {
+      "epoch": 0.5156413498870784,
+      "grad_norm": 548.3775024414062,
+      "learning_rate": 2.843438487741591e-05,
+      "loss": 53.7288,
+      "step": 127630
+    },
+    {
+      "epoch": 0.5156817511524461,
+      "grad_norm": 294.67889404296875,
+      "learning_rate": 2.843092728018248e-05,
+      "loss": 49.3608,
+      "step": 127640
+    },
+    {
+      "epoch": 0.5157221524178137,
+      "grad_norm": 2212.821533203125,
+      "learning_rate": 2.8427469616061364e-05,
+      "loss": 59.0455,
+      "step": 127650
+    },
+    {
+      "epoch": 0.5157625536831814,
+      "grad_norm": 849.8643798828125,
+      "learning_rate": 2.8424011885119954e-05,
+      "loss": 44.9502,
+      "step": 127660
+    },
+    {
+      "epoch": 0.515802954948549,
+      "grad_norm": 565.5281372070312,
+      "learning_rate": 2.842055408742567e-05,
+      "loss": 53.8755,
+      "step": 127670
+    },
+    {
+      "epoch": 0.5158433562139166,
+      "grad_norm": 874.5379638671875,
+      "learning_rate": 2.8417096223045925e-05,
+      "loss": 53.4064,
+      "step": 127680
+    },
+    {
+      "epoch": 0.5158837574792843,
+      "grad_norm": 761.3839111328125,
+      "learning_rate": 2.841363829204814e-05,
+      "loss": 40.8466,
+      "step": 127690
+    },
+    {
+      "epoch": 0.5159241587446519,
+      "grad_norm": 409.4591369628906,
+      "learning_rate": 2.841018029449971e-05,
+      "loss": 40.9471,
+      "step": 127700
+    },
+    {
+      "epoch": 0.5159645600100196,
+      "grad_norm": 896.9284057617188,
+      "learning_rate": 2.8406722230468063e-05,
+      "loss": 69.344,
+      "step": 127710
+    },
+    {
+      "epoch": 0.5160049612753872,
+      "grad_norm": 248.2589569091797,
+      "learning_rate": 2.840326410002061e-05,
+      "loss": 53.1701,
+      "step": 127720
+    },
+    {
+      "epoch": 0.5160453625407547,
+      "grad_norm": 770.6051025390625,
+      "learning_rate": 2.839980590322477e-05,
+      "loss": 37.5784,
+      "step": 127730
+    },
+    {
+      "epoch": 0.5160857638061224,
+      "grad_norm": 1029.7786865234375,
+      "learning_rate": 2.8396347640147962e-05,
+      "loss": 57.3108,
+      "step": 127740
+    },
+    {
+      "epoch": 0.51612616507149,
+      "grad_norm": 450.9689636230469,
+      "learning_rate": 2.8392889310857612e-05,
+      "loss": 47.8311,
+      "step": 127750
+    },
+    {
+      "epoch": 0.5161665663368576,
+      "grad_norm": 471.0567932128906,
+      "learning_rate": 2.8389430915421132e-05,
+      "loss": 58.49,
+      "step": 127760
+    },
+    {
+      "epoch": 0.5162069676022253,
+      "grad_norm": 991.8955688476562,
+      "learning_rate": 2.8385972453905958e-05,
+      "loss": 43.4331,
+      "step": 127770
+    },
+    {
+      "epoch": 0.5162473688675929,
+      "grad_norm": 0.0,
+      "learning_rate": 2.8382513926379504e-05,
+      "loss": 66.81,
+      "step": 127780
+    },
+    {
+      "epoch": 0.5162877701329606,
+      "grad_norm": 1275.402587890625,
+      "learning_rate": 2.837905533290921e-05,
+      "loss": 33.7486,
+      "step": 127790
+    },
+    {
+      "epoch": 0.5163281713983282,
+      "grad_norm": 921.3496704101562,
+      "learning_rate": 2.8375596673562482e-05,
+      "loss": 66.0726,
+      "step": 127800
+    },
+    {
+      "epoch": 0.5163685726636958,
+      "grad_norm": 1107.1654052734375,
+      "learning_rate": 2.8372137948406762e-05,
+      "loss": 64.4679,
+      "step": 127810
+    },
+    {
+      "epoch": 0.5164089739290635,
+      "grad_norm": 978.436279296875,
+      "learning_rate": 2.8368679157509477e-05,
+      "loss": 55.1411,
+      "step": 127820
+    },
+    {
+      "epoch": 0.5164493751944311,
+      "grad_norm": 794.9774169921875,
+      "learning_rate": 2.8365220300938055e-05,
+      "loss": 44.3256,
+      "step": 127830
+    },
+    {
+      "epoch": 0.5164897764597988,
+      "grad_norm": 2009.2308349609375,
+      "learning_rate": 2.8361761378759934e-05,
+      "loss": 52.7043,
+      "step": 127840
+    },
+    {
+      "epoch": 0.5165301777251664,
+      "grad_norm": 989.152099609375,
+      "learning_rate": 2.8358302391042536e-05,
+      "loss": 37.631,
+      "step": 127850
+    },
+    {
+      "epoch": 0.5165705789905339,
+      "grad_norm": 356.483642578125,
+      "learning_rate": 2.8354843337853314e-05,
+      "loss": 77.6594,
+      "step": 127860
+    },
+    {
+      "epoch": 0.5166109802559016,
+      "grad_norm": 1363.230224609375,
+      "learning_rate": 2.835138421925969e-05,
+      "loss": 79.6859,
+      "step": 127870
+    },
+    {
+      "epoch": 0.5166513815212692,
+      "grad_norm": 811.1392211914062,
+      "learning_rate": 2.834792503532911e-05,
+      "loss": 75.7948,
+      "step": 127880
+    },
+    {
+      "epoch": 0.5166917827866369,
+      "grad_norm": 1689.7584228515625,
+      "learning_rate": 2.8344465786129e-05,
+      "loss": 79.6354,
+      "step": 127890
+    },
+    {
+      "epoch": 0.5167321840520045,
+      "grad_norm": 251.05508422851562,
+      "learning_rate": 2.8341006471726816e-05,
+      "loss": 38.7622,
+      "step": 127900
+    },
+    {
+      "epoch": 0.5167725853173721,
+      "grad_norm": 1307.02880859375,
+      "learning_rate": 2.833754709218998e-05,
+      "loss": 75.8891,
+      "step": 127910
+    },
+    {
+      "epoch": 0.5168129865827398,
+      "grad_norm": 743.4976196289062,
+      "learning_rate": 2.833408764758595e-05,
+      "loss": 37.3798,
+      "step": 127920
+    },
+    {
+      "epoch": 0.5168533878481074,
+      "grad_norm": 395.01214599609375,
+      "learning_rate": 2.833062813798216e-05,
+      "loss": 53.7051,
+      "step": 127930
+    },
+    {
+      "epoch": 0.516893789113475,
+      "grad_norm": 1006.5745849609375,
+      "learning_rate": 2.832716856344607e-05,
+      "loss": 53.2251,
+      "step": 127940
+    },
+    {
+      "epoch": 0.5169341903788427,
+      "grad_norm": 649.6722412109375,
+      "learning_rate": 2.832370892404511e-05,
+      "loss": 47.6598,
+      "step": 127950
+    },
+    {
+      "epoch": 0.5169745916442103,
+      "grad_norm": 952.0791625976562,
+      "learning_rate": 2.832024921984674e-05,
+      "loss": 46.0533,
+      "step": 127960
+    },
+    {
+      "epoch": 0.517014992909578,
+      "grad_norm": 663.2807006835938,
+      "learning_rate": 2.8316789450918396e-05,
+      "loss": 60.6055,
+      "step": 127970
+    },
+    {
+      "epoch": 0.5170553941749455,
+      "grad_norm": 1854.5562744140625,
+      "learning_rate": 2.8313329617327537e-05,
+      "loss": 104.9597,
+      "step": 127980
+    },
+    {
+      "epoch": 0.5170957954403131,
+      "grad_norm": 399.9706726074219,
+      "learning_rate": 2.8309869719141608e-05,
+      "loss": 73.0087,
+      "step": 127990
+    },
+    {
+      "epoch": 0.5171361967056808,
+      "grad_norm": 742.0582275390625,
+      "learning_rate": 2.8306409756428064e-05,
+      "loss": 63.6711,
+      "step": 128000
+    },
+    {
+      "epoch": 0.5171765979710484,
+      "grad_norm": 426.78814697265625,
+      "learning_rate": 2.8302949729254358e-05,
+      "loss": 44.5783,
+      "step": 128010
+    },
+    {
+      "epoch": 0.5172169992364161,
+      "grad_norm": 537.81494140625,
+      "learning_rate": 2.8299489637687954e-05,
+      "loss": 46.2207,
+      "step": 128020
+    },
+    {
+      "epoch": 0.5172574005017837,
+      "grad_norm": 545.3671264648438,
+      "learning_rate": 2.8296029481796292e-05,
+      "loss": 52.6649,
+      "step": 128030
+    },
+    {
+      "epoch": 0.5172978017671513,
+      "grad_norm": 1167.154541015625,
+      "learning_rate": 2.829256926164685e-05,
+      "loss": 68.5006,
+      "step": 128040
+    },
+    {
+      "epoch": 0.517338203032519,
+      "grad_norm": 743.8500366210938,
+      "learning_rate": 2.8289108977307067e-05,
+      "loss": 49.9936,
+      "step": 128050
+    },
+    {
+      "epoch": 0.5173786042978866,
+      "grad_norm": 708.7972412109375,
+      "learning_rate": 2.8285648628844413e-05,
+      "loss": 58.3343,
+      "step": 128060
+    },
+    {
+      "epoch": 0.5174190055632543,
+      "grad_norm": 952.4853515625,
+      "learning_rate": 2.8282188216326345e-05,
+      "loss": 51.3421,
+      "step": 128070
+    },
+    {
+      "epoch": 0.5174594068286219,
+      "grad_norm": 761.9249267578125,
+      "learning_rate": 2.8278727739820333e-05,
+      "loss": 47.8939,
+      "step": 128080
+    },
+    {
+      "epoch": 0.5174998080939895,
+      "grad_norm": 1164.4027099609375,
+      "learning_rate": 2.827526719939383e-05,
+      "loss": 112.3727,
+      "step": 128090
+    },
+    {
+      "epoch": 0.5175402093593572,
+      "grad_norm": 1486.0279541015625,
+      "learning_rate": 2.827180659511431e-05,
+      "loss": 54.5274,
+      "step": 128100
+    },
+    {
+      "epoch": 0.5175806106247247,
+      "grad_norm": 1370.9776611328125,
+      "learning_rate": 2.8268345927049234e-05,
+      "loss": 65.6802,
+      "step": 128110
+    },
+    {
+      "epoch": 0.5176210118900924,
+      "grad_norm": 796.0413818359375,
+      "learning_rate": 2.8264885195266065e-05,
+      "loss": 58.2697,
+      "step": 128120
+    },
+    {
+      "epoch": 0.51766141315546,
+      "grad_norm": 703.8768310546875,
+      "learning_rate": 2.8261424399832293e-05,
+      "loss": 49.1636,
+      "step": 128130
+    },
+    {
+      "epoch": 0.5177018144208276,
+      "grad_norm": 419.2386779785156,
+      "learning_rate": 2.825796354081537e-05,
+      "loss": 39.2348,
+      "step": 128140
+    },
+    {
+      "epoch": 0.5177422156861953,
+      "grad_norm": 687.80859375,
+      "learning_rate": 2.8254502618282763e-05,
+      "loss": 33.7805,
+      "step": 128150
+    },
+    {
+      "epoch": 0.5177826169515629,
+      "grad_norm": 0.0,
+      "learning_rate": 2.8251041632301957e-05,
+      "loss": 54.0625,
+      "step": 128160
+    },
+    {
+      "epoch": 0.5178230182169306,
+      "grad_norm": 458.79937744140625,
+      "learning_rate": 2.8247580582940413e-05,
+      "loss": 60.1953,
+      "step": 128170
+    },
+    {
+      "epoch": 0.5178634194822982,
+      "grad_norm": 739.254638671875,
+      "learning_rate": 2.824411947026563e-05,
+      "loss": 67.9315,
+      "step": 128180
+    },
+    {
+      "epoch": 0.5179038207476658,
+      "grad_norm": 998.814208984375,
+      "learning_rate": 2.824065829434505e-05,
+      "loss": 52.8274,
+      "step": 128190
+    },
+    {
+      "epoch": 0.5179442220130335,
+      "grad_norm": 862.947265625,
+      "learning_rate": 2.8237197055246172e-05,
+      "loss": 47.5776,
+      "step": 128200
+    },
+    {
+      "epoch": 0.5179846232784011,
+      "grad_norm": 0.0,
+      "learning_rate": 2.8233735753036484e-05,
+      "loss": 58.8695,
+      "step": 128210
+    },
+    {
+      "epoch": 0.5180250245437688,
+      "grad_norm": 1318.5653076171875,
+      "learning_rate": 2.823027438778344e-05,
+      "loss": 46.4017,
+      "step": 128220
+    },
+    {
+      "epoch": 0.5180654258091364,
+      "grad_norm": 1210.0972900390625,
+      "learning_rate": 2.8226812959554537e-05,
+      "loss": 61.901,
+      "step": 128230
+    },
+    {
+      "epoch": 0.5181058270745039,
+      "grad_norm": 756.2396850585938,
+      "learning_rate": 2.8223351468417254e-05,
+      "loss": 49.0368,
+      "step": 128240
+    },
+    {
+      "epoch": 0.5181462283398716,
+      "grad_norm": 419.0620422363281,
+      "learning_rate": 2.8219889914439074e-05,
+      "loss": 52.058,
+      "step": 128250
+    },
+    {
+      "epoch": 0.5181866296052392,
+      "grad_norm": 868.4646606445312,
+      "learning_rate": 2.821642829768748e-05,
+      "loss": 78.654,
+      "step": 128260
+    },
+    {
+      "epoch": 0.5182270308706068,
+      "grad_norm": 424.6618957519531,
+      "learning_rate": 2.8212966618229964e-05,
+      "loss": 94.6811,
+      "step": 128270
+    },
+    {
+      "epoch": 0.5182674321359745,
+      "grad_norm": 535.59228515625,
+      "learning_rate": 2.8209504876134007e-05,
+      "loss": 44.2588,
+      "step": 128280
+    },
+    {
+      "epoch": 0.5183078334013421,
+      "grad_norm": 578.2286987304688,
+      "learning_rate": 2.8206043071467102e-05,
+      "loss": 38.9842,
+      "step": 128290
+    },
+    {
+      "epoch": 0.5183482346667098,
+      "grad_norm": 948.8009033203125,
+      "learning_rate": 2.8202581204296742e-05,
+      "loss": 66.6254,
+      "step": 128300
+    },
+    {
+      "epoch": 0.5183886359320774,
+      "grad_norm": 956.4691772460938,
+      "learning_rate": 2.819911927469041e-05,
+      "loss": 83.3204,
+      "step": 128310
+    },
+    {
+      "epoch": 0.518429037197445,
+      "grad_norm": 662.3033447265625,
+      "learning_rate": 2.8195657282715594e-05,
+      "loss": 33.0041,
+      "step": 128320
+    },
+    {
+      "epoch": 0.5184694384628127,
+      "grad_norm": 841.9630737304688,
+      "learning_rate": 2.81921952284398e-05,
+      "loss": 40.0045,
+      "step": 128330
+    },
+    {
+      "epoch": 0.5185098397281803,
+      "grad_norm": 644.8792114257812,
+      "learning_rate": 2.818873311193051e-05,
+      "loss": 49.6458,
+      "step": 128340
+    },
+    {
+      "epoch": 0.518550240993548,
+      "grad_norm": 1007.34765625,
+      "learning_rate": 2.8185270933255237e-05,
+      "loss": 45.5391,
+      "step": 128350
+    },
+    {
+      "epoch": 0.5185906422589156,
+      "grad_norm": 412.73883056640625,
+      "learning_rate": 2.8181808692481453e-05,
+      "loss": 45.3399,
+      "step": 128360
+    },
+    {
+      "epoch": 0.5186310435242831,
+      "grad_norm": 775.2849731445312,
+      "learning_rate": 2.817834638967668e-05,
+      "loss": 56.8275,
+      "step": 128370
+    },
+    {
+      "epoch": 0.5186714447896508,
+      "grad_norm": 880.3590698242188,
+      "learning_rate": 2.817488402490841e-05,
+      "loss": 62.0417,
+      "step": 128380
+    },
+    {
+      "epoch": 0.5187118460550184,
+      "grad_norm": 0.0,
+      "learning_rate": 2.8171421598244134e-05,
+      "loss": 100.0851,
+      "step": 128390
+    },
+    {
+      "epoch": 0.518752247320386,
+      "grad_norm": 757.4847412109375,
+      "learning_rate": 2.816795910975137e-05,
+      "loss": 81.3571,
+      "step": 128400
+    },
+    {
+      "epoch": 0.5187926485857537,
+      "grad_norm": 986.7409057617188,
+      "learning_rate": 2.8164496559497605e-05,
+      "loss": 99.9801,
+      "step": 128410
+    },
+    {
+      "epoch": 0.5188330498511213,
+      "grad_norm": 1231.8388671875,
+      "learning_rate": 2.816103394755035e-05,
+      "loss": 36.7881,
+      "step": 128420
+    },
+    {
+      "epoch": 0.518873451116489,
+      "grad_norm": 551.5934448242188,
+      "learning_rate": 2.8157571273977117e-05,
+      "loss": 48.0908,
+      "step": 128430
+    },
+    {
+      "epoch": 0.5189138523818566,
+      "grad_norm": 957.9308471679688,
+      "learning_rate": 2.8154108538845404e-05,
+      "loss": 54.7105,
+      "step": 128440
+    },
+    {
+      "epoch": 0.5189542536472243,
+      "grad_norm": 746.921875,
+      "learning_rate": 2.8150645742222714e-05,
+      "loss": 47.9935,
+      "step": 128450
+    },
+    {
+      "epoch": 0.5189946549125919,
+      "grad_norm": 817.0953369140625,
+      "learning_rate": 2.814718288417657e-05,
+      "loss": 34.7214,
+      "step": 128460
+    },
+    {
+      "epoch": 0.5190350561779595,
+      "grad_norm": 652.5723876953125,
+      "learning_rate": 2.814371996477448e-05,
+      "loss": 51.5591,
+      "step": 128470
+    },
+    {
+      "epoch": 0.5190754574433272,
+      "grad_norm": 1426.8663330078125,
+      "learning_rate": 2.8140256984083947e-05,
+      "loss": 53.6987,
+      "step": 128480
+    },
+    {
+      "epoch": 0.5191158587086948,
+      "grad_norm": 1349.2745361328125,
+      "learning_rate": 2.8136793942172483e-05,
+      "loss": 67.8195,
+      "step": 128490
+    },
+    {
+      "epoch": 0.5191562599740623,
+      "grad_norm": 1255.9248046875,
+      "learning_rate": 2.8133330839107608e-05,
+      "loss": 89.4671,
+      "step": 128500
+    },
+    {
+      "epoch": 0.51919666123943,
+      "grad_norm": 772.8133544921875,
+      "learning_rate": 2.8129867674956838e-05,
+      "loss": 99.116,
+      "step": 128510
+    },
+    {
+      "epoch": 0.5192370625047976,
+      "grad_norm": 799.4025268554688,
+      "learning_rate": 2.8126404449787685e-05,
+      "loss": 37.4048,
+      "step": 128520
+    },
+    {
+      "epoch": 0.5192774637701653,
+      "grad_norm": 767.2107543945312,
+      "learning_rate": 2.8122941163667667e-05,
+      "loss": 53.6676,
+      "step": 128530
+    },
+    {
+      "epoch": 0.5193178650355329,
+      "grad_norm": 717.8088989257812,
+      "learning_rate": 2.8119477816664296e-05,
+      "loss": 53.9604,
+      "step": 128540
+    },
+    {
+      "epoch": 0.5193582663009005,
+      "grad_norm": 568.8512573242188,
+      "learning_rate": 2.8116014408845116e-05,
+      "loss": 52.336,
+      "step": 128550
+    },
+    {
+      "epoch": 0.5193986675662682,
+      "grad_norm": 1033.583984375,
+      "learning_rate": 2.8112550940277616e-05,
+      "loss": 60.4758,
+      "step": 128560
+    },
+    {
+      "epoch": 0.5194390688316358,
+      "grad_norm": 267.0958251953125,
+      "learning_rate": 2.810908741102934e-05,
+      "loss": 87.3894,
+      "step": 128570
+    },
+    {
+      "epoch": 0.5194794700970035,
+      "grad_norm": 966.0076293945312,
+      "learning_rate": 2.8105623821167804e-05,
+      "loss": 64.1264,
+      "step": 128580
+    },
+    {
+      "epoch": 0.5195198713623711,
+      "grad_norm": 888.2479858398438,
+      "learning_rate": 2.810216017076053e-05,
+      "loss": 37.3439,
+      "step": 128590
+    },
+    {
+      "epoch": 0.5195602726277387,
+      "grad_norm": 1010.2241821289062,
+      "learning_rate": 2.8098696459875046e-05,
+      "loss": 59.4499,
+      "step": 128600
+    },
+    {
+      "epoch": 0.5196006738931064,
+      "grad_norm": 672.6810302734375,
+      "learning_rate": 2.8095232688578883e-05,
+      "loss": 59.5572,
+      "step": 128610
+    },
+    {
+      "epoch": 0.5196410751584739,
+      "grad_norm": 1322.0794677734375,
+      "learning_rate": 2.809176885693956e-05,
+      "loss": 75.077,
+      "step": 128620
+    },
+    {
+      "epoch": 0.5196814764238415,
+      "grad_norm": 647.8616943359375,
+      "learning_rate": 2.8088304965024614e-05,
+      "loss": 54.4772,
+      "step": 128630
+    },
+    {
+      "epoch": 0.5197218776892092,
+      "grad_norm": 895.9517211914062,
+      "learning_rate": 2.8084841012901574e-05,
+      "loss": 32.6104,
+      "step": 128640
+    },
+    {
+      "epoch": 0.5197622789545768,
+      "grad_norm": 346.7025451660156,
+      "learning_rate": 2.808137700063797e-05,
+      "loss": 39.3208,
+      "step": 128650
+    },
+    {
+      "epoch": 0.5198026802199445,
+      "grad_norm": 911.7136840820312,
+      "learning_rate": 2.807791292830133e-05,
+      "loss": 58.0953,
+      "step": 128660
+    },
+    {
+      "epoch": 0.5198430814853121,
+      "grad_norm": 522.1300659179688,
+      "learning_rate": 2.8074448795959203e-05,
+      "loss": 73.4923,
+      "step": 128670
+    },
+    {
+      "epoch": 0.5198834827506797,
+      "grad_norm": 657.18505859375,
+      "learning_rate": 2.8070984603679107e-05,
+      "loss": 42.6445,
+      "step": 128680
+    },
+    {
+      "epoch": 0.5199238840160474,
+      "grad_norm": 389.8962707519531,
+      "learning_rate": 2.8067520351528587e-05,
+      "loss": 43.4139,
+      "step": 128690
+    },
+    {
+      "epoch": 0.519964285281415,
+      "grad_norm": 851.1239013671875,
+      "learning_rate": 2.806405603957517e-05,
+      "loss": 72.5371,
+      "step": 128700
+    },
+    {
+      "epoch": 0.5200046865467827,
+      "grad_norm": 3148.775634765625,
+      "learning_rate": 2.8060591667886416e-05,
+      "loss": 56.5808,
+      "step": 128710
+    },
+    {
+      "epoch": 0.5200450878121503,
+      "grad_norm": 1202.496826171875,
+      "learning_rate": 2.8057127236529844e-05,
+      "loss": 46.4309,
+      "step": 128720
+    },
+    {
+      "epoch": 0.520085489077518,
+      "grad_norm": 962.0496826171875,
+      "learning_rate": 2.805366274557301e-05,
+      "loss": 50.884,
+      "step": 128730
+    },
+    {
+      "epoch": 0.5201258903428856,
+      "grad_norm": 453.456298828125,
+      "learning_rate": 2.8050198195083444e-05,
+      "loss": 59.2252,
+      "step": 128740
+    },
+    {
+      "epoch": 0.5201662916082531,
+      "grad_norm": 239.76718139648438,
+      "learning_rate": 2.8046733585128687e-05,
+      "loss": 70.7534,
+      "step": 128750
+    },
+    {
+      "epoch": 0.5202066928736208,
+      "grad_norm": 457.1228942871094,
+      "learning_rate": 2.80432689157763e-05,
+      "loss": 52.9828,
+      "step": 128760
+    },
+    {
+      "epoch": 0.5202470941389884,
+      "grad_norm": 877.8587036132812,
+      "learning_rate": 2.8039804187093816e-05,
+      "loss": 52.8952,
+      "step": 128770
+    },
+    {
+      "epoch": 0.520287495404356,
+      "grad_norm": 1153.672119140625,
+      "learning_rate": 2.803633939914878e-05,
+      "loss": 60.0604,
+      "step": 128780
+    },
+    {
+      "epoch": 0.5203278966697237,
+      "grad_norm": 120.61048126220703,
+      "learning_rate": 2.803287455200875e-05,
+      "loss": 47.4182,
+      "step": 128790
+    },
+    {
+      "epoch": 0.5203682979350913,
+      "grad_norm": 1153.82080078125,
+      "learning_rate": 2.8029409645741267e-05,
+      "loss": 59.4981,
+      "step": 128800
+    },
+    {
+      "epoch": 0.520408699200459,
+      "grad_norm": 418.24920654296875,
+      "learning_rate": 2.8025944680413878e-05,
+      "loss": 37.4401,
+      "step": 128810
+    },
+    {
+      "epoch": 0.5204491004658266,
+      "grad_norm": 781.8113403320312,
+      "learning_rate": 2.8022479656094154e-05,
+      "loss": 48.7074,
+      "step": 128820
+    },
+    {
+      "epoch": 0.5204895017311942,
+      "grad_norm": 836.5515747070312,
+      "learning_rate": 2.801901457284962e-05,
+      "loss": 51.3323,
+      "step": 128830
+    },
+    {
+      "epoch": 0.5205299029965619,
+      "grad_norm": 1007.6831665039062,
+      "learning_rate": 2.8015549430747852e-05,
+      "loss": 63.1766,
+      "step": 128840
+    },
+    {
+      "epoch": 0.5205703042619295,
+      "grad_norm": 1059.2291259765625,
+      "learning_rate": 2.8012084229856382e-05,
+      "loss": 48.2031,
+      "step": 128850
+    },
+    {
+      "epoch": 0.5206107055272972,
+      "grad_norm": 435.08837890625,
+      "learning_rate": 2.800861897024279e-05,
+      "loss": 61.0167,
+      "step": 128860
+    },
+    {
+      "epoch": 0.5206511067926648,
+      "grad_norm": 520.9664916992188,
+      "learning_rate": 2.8005153651974614e-05,
+      "loss": 68.8364,
+      "step": 128870
+    },
+    {
+      "epoch": 0.5206915080580323,
+      "grad_norm": 1375.3497314453125,
+      "learning_rate": 2.8001688275119432e-05,
+      "loss": 73.5849,
+      "step": 128880
+    },
+    {
+      "epoch": 0.5207319093234,
+      "grad_norm": 570.526611328125,
+      "learning_rate": 2.799822283974478e-05,
+      "loss": 45.2284,
+      "step": 128890
+    },
+    {
+      "epoch": 0.5207723105887676,
+      "grad_norm": 3453.25927734375,
+      "learning_rate": 2.7994757345918244e-05,
+      "loss": 65.3527,
+      "step": 128900
+    },
+    {
+      "epoch": 0.5208127118541352,
+      "grad_norm": 807.7914428710938,
+      "learning_rate": 2.7991291793707357e-05,
+      "loss": 37.759,
+      "step": 128910
+    },
+    {
+      "epoch": 0.5208531131195029,
+      "grad_norm": 2692.131103515625,
+      "learning_rate": 2.7987826183179712e-05,
+      "loss": 70.0045,
+      "step": 128920
+    },
+    {
+      "epoch": 0.5208935143848705,
+      "grad_norm": 316.426025390625,
+      "learning_rate": 2.798436051440284e-05,
+      "loss": 50.6802,
+      "step": 128930
+    },
+    {
+      "epoch": 0.5209339156502382,
+      "grad_norm": 1145.9749755859375,
+      "learning_rate": 2.7980894787444334e-05,
+      "loss": 44.7317,
+      "step": 128940
+    },
+    {
+      "epoch": 0.5209743169156058,
+      "grad_norm": 571.5775146484375,
+      "learning_rate": 2.7977429002371747e-05,
+      "loss": 39.6738,
+      "step": 128950
+    },
+    {
+      "epoch": 0.5210147181809734,
+      "grad_norm": 625.548828125,
+      "learning_rate": 2.797396315925265e-05,
+      "loss": 34.7972,
+      "step": 128960
+    },
+    {
+      "epoch": 0.5210551194463411,
+      "grad_norm": 733.4710083007812,
+      "learning_rate": 2.7970497258154603e-05,
+      "loss": 38.4996,
+      "step": 128970
+    },
+    {
+      "epoch": 0.5210955207117087,
+      "grad_norm": 1038.670654296875,
+      "learning_rate": 2.7967031299145193e-05,
+      "loss": 49.1648,
+      "step": 128980
+    },
+    {
+      "epoch": 0.5211359219770764,
+      "grad_norm": 545.0449829101562,
+      "learning_rate": 2.7963565282291977e-05,
+      "loss": 48.0368,
+      "step": 128990
+    },
+    {
+      "epoch": 0.521176323242444,
+      "grad_norm": 428.3600769042969,
+      "learning_rate": 2.7960099207662532e-05,
+      "loss": 60.4256,
+      "step": 129000
+    },
+    {
+      "epoch": 0.5212167245078115,
+      "grad_norm": 714.7028198242188,
+      "learning_rate": 2.7956633075324424e-05,
+      "loss": 60.8093,
+      "step": 129010
+    },
+    {
+      "epoch": 0.5212571257731792,
+      "grad_norm": 1208.1380615234375,
+      "learning_rate": 2.795316688534523e-05,
+      "loss": 74.678,
+      "step": 129020
+    },
+    {
+      "epoch": 0.5212975270385468,
+      "grad_norm": 534.9729614257812,
+      "learning_rate": 2.794970063779253e-05,
+      "loss": 51.2989,
+      "step": 129030
+    },
+    {
+      "epoch": 0.5213379283039145,
+      "grad_norm": 756.8164672851562,
+      "learning_rate": 2.79462343327339e-05,
+      "loss": 66.086,
+      "step": 129040
+    },
+    {
+      "epoch": 0.5213783295692821,
+      "grad_norm": 575.9217529296875,
+      "learning_rate": 2.794276797023691e-05,
+      "loss": 49.4426,
+      "step": 129050
+    },
+    {
+      "epoch": 0.5214187308346497,
+      "grad_norm": 498.71441650390625,
+      "learning_rate": 2.7939301550369146e-05,
+      "loss": 48.8029,
+      "step": 129060
+    },
+    {
+      "epoch": 0.5214591321000174,
+      "grad_norm": 594.4951782226562,
+      "learning_rate": 2.7935835073198192e-05,
+      "loss": 34.5576,
+      "step": 129070
+    },
+    {
+      "epoch": 0.521499533365385,
+      "grad_norm": 1014.3102416992188,
+      "learning_rate": 2.793236853879161e-05,
+      "loss": 58.8379,
+      "step": 129080
+    },
+    {
+      "epoch": 0.5215399346307527,
+      "grad_norm": 560.2127685546875,
+      "learning_rate": 2.7928901947217008e-05,
+      "loss": 63.1494,
+      "step": 129090
+    },
+    {
+      "epoch": 0.5215803358961203,
+      "grad_norm": 666.1370239257812,
+      "learning_rate": 2.792543529854194e-05,
+      "loss": 53.1794,
+      "step": 129100
+    },
+    {
+      "epoch": 0.5216207371614879,
+      "grad_norm": 636.662353515625,
+      "learning_rate": 2.7921968592834006e-05,
+      "loss": 25.9828,
+      "step": 129110
+    },
+    {
+      "epoch": 0.5216611384268556,
+      "grad_norm": 835.5521850585938,
+      "learning_rate": 2.79185018301608e-05,
+      "loss": 57.4081,
+      "step": 129120
+    },
+    {
+      "epoch": 0.5217015396922232,
+      "grad_norm": 992.4476928710938,
+      "learning_rate": 2.791503501058989e-05,
+      "loss": 52.8603,
+      "step": 129130
+    },
+    {
+      "epoch": 0.5217419409575907,
+      "grad_norm": 328.57794189453125,
+      "learning_rate": 2.7911568134188875e-05,
+      "loss": 88.1504,
+      "step": 129140
+    },
+    {
+      "epoch": 0.5217823422229584,
+      "grad_norm": 1140.0035400390625,
+      "learning_rate": 2.7908101201025337e-05,
+      "loss": 48.8337,
+      "step": 129150
+    },
+    {
+      "epoch": 0.521822743488326,
+      "grad_norm": 745.549072265625,
+      "learning_rate": 2.7904634211166876e-05,
+      "loss": 55.3195,
+      "step": 129160
+    },
+    {
+      "epoch": 0.5218631447536937,
+      "grad_norm": 969.3555297851562,
+      "learning_rate": 2.7901167164681073e-05,
+      "loss": 55.3426,
+      "step": 129170
+    },
+    {
+      "epoch": 0.5219035460190613,
+      "grad_norm": 726.4353637695312,
+      "learning_rate": 2.7897700061635517e-05,
+      "loss": 49.0015,
+      "step": 129180
+    },
+    {
+      "epoch": 0.5219439472844289,
+      "grad_norm": 374.95843505859375,
+      "learning_rate": 2.7894232902097813e-05,
+      "loss": 46.3873,
+      "step": 129190
+    },
+    {
+      "epoch": 0.5219843485497966,
+      "grad_norm": 511.00885009765625,
+      "learning_rate": 2.7890765686135544e-05,
+      "loss": 47.9072,
+      "step": 129200
+    },
+    {
+      "epoch": 0.5220247498151642,
+      "grad_norm": 512.494873046875,
+      "learning_rate": 2.788729841381631e-05,
+      "loss": 29.6466,
+      "step": 129210
+    },
+    {
+      "epoch": 0.5220651510805319,
+      "grad_norm": 551.0260620117188,
+      "learning_rate": 2.7883831085207707e-05,
+      "loss": 65.4076,
+      "step": 129220
+    },
+    {
+      "epoch": 0.5221055523458995,
+      "grad_norm": 621.8154907226562,
+      "learning_rate": 2.788036370037733e-05,
+      "loss": 55.3526,
+      "step": 129230
+    },
+    {
+      "epoch": 0.5221459536112671,
+      "grad_norm": 477.7518310546875,
+      "learning_rate": 2.7876896259392788e-05,
+      "loss": 70.7479,
+      "step": 129240
+    },
+    {
+      "epoch": 0.5221863548766348,
+      "grad_norm": 228.69644165039062,
+      "learning_rate": 2.787342876232167e-05,
+      "loss": 45.7491,
+      "step": 129250
+    },
+    {
+      "epoch": 0.5222267561420023,
+      "grad_norm": 406.46026611328125,
+      "learning_rate": 2.7869961209231577e-05,
+      "loss": 40.5636,
+      "step": 129260
+    },
+    {
+      "epoch": 0.52226715740737,
+      "grad_norm": 1248.6951904296875,
+      "learning_rate": 2.7866493600190107e-05,
+      "loss": 43.6552,
+      "step": 129270
+    },
+    {
+      "epoch": 0.5223075586727376,
+      "grad_norm": 780.5540161132812,
+      "learning_rate": 2.7863025935264875e-05,
+      "loss": 42.3198,
+      "step": 129280
+    },
+    {
+      "epoch": 0.5223479599381052,
+      "grad_norm": 830.3927612304688,
+      "learning_rate": 2.785955821452348e-05,
+      "loss": 50.129,
+      "step": 129290
+    },
+    {
+      "epoch": 0.5223883612034729,
+      "grad_norm": 162.7823028564453,
+      "learning_rate": 2.7856090438033522e-05,
+      "loss": 35.8086,
+      "step": 129300
+    },
+    {
+      "epoch": 0.5224287624688405,
+      "grad_norm": 577.8681030273438,
+      "learning_rate": 2.785262260586261e-05,
+      "loss": 48.4552,
+      "step": 129310
+    },
+    {
+      "epoch": 0.5224691637342082,
+      "grad_norm": 495.5697021484375,
+      "learning_rate": 2.7849154718078346e-05,
+      "loss": 41.2047,
+      "step": 129320
+    },
+    {
+      "epoch": 0.5225095649995758,
+      "grad_norm": 650.5645141601562,
+      "learning_rate": 2.784568677474836e-05,
+      "loss": 65.8662,
+      "step": 129330
+    },
+    {
+      "epoch": 0.5225499662649434,
+      "grad_norm": 276.66241455078125,
+      "learning_rate": 2.7842218775940237e-05,
+      "loss": 33.7441,
+      "step": 129340
+    },
+    {
+      "epoch": 0.5225903675303111,
+      "grad_norm": 840.7539672851562,
+      "learning_rate": 2.783875072172159e-05,
+      "loss": 49.3532,
+      "step": 129350
+    },
+    {
+      "epoch": 0.5226307687956787,
+      "grad_norm": 315.2930603027344,
+      "learning_rate": 2.783528261216004e-05,
+      "loss": 37.2258,
+      "step": 129360
+    },
+    {
+      "epoch": 0.5226711700610464,
+      "grad_norm": 398.0323791503906,
+      "learning_rate": 2.78318144473232e-05,
+      "loss": 85.9904,
+      "step": 129370
+    },
+    {
+      "epoch": 0.522711571326414,
+      "grad_norm": 779.0153198242188,
+      "learning_rate": 2.7828346227278674e-05,
+      "loss": 53.5858,
+      "step": 129380
+    },
+    {
+      "epoch": 0.5227519725917815,
+      "grad_norm": 941.3084106445312,
+      "learning_rate": 2.782487795209408e-05,
+      "loss": 77.1186,
+      "step": 129390
+    },
+    {
+      "epoch": 0.5227923738571492,
+      "grad_norm": 1087.8790283203125,
+      "learning_rate": 2.782140962183704e-05,
+      "loss": 75.8238,
+      "step": 129400
+    },
+    {
+      "epoch": 0.5228327751225168,
+      "grad_norm": 836.9957275390625,
+      "learning_rate": 2.7817941236575173e-05,
+      "loss": 60.6787,
+      "step": 129410
+    },
+    {
+      "epoch": 0.5228731763878844,
+      "grad_norm": 542.893798828125,
+      "learning_rate": 2.781447279637608e-05,
+      "loss": 54.1359,
+      "step": 129420
+    },
+    {
+      "epoch": 0.5229135776532521,
+      "grad_norm": 706.9591064453125,
+      "learning_rate": 2.7811004301307403e-05,
+      "loss": 42.0238,
+      "step": 129430
+    },
+    {
+      "epoch": 0.5229539789186197,
+      "grad_norm": 975.5703735351562,
+      "learning_rate": 2.7807535751436738e-05,
+      "loss": 42.8671,
+      "step": 129440
+    },
+    {
+      "epoch": 0.5229943801839874,
+      "grad_norm": 1424.2943115234375,
+      "learning_rate": 2.7804067146831725e-05,
+      "loss": 69.0784,
+      "step": 129450
+    },
+    {
+      "epoch": 0.523034781449355,
+      "grad_norm": 1137.7862548828125,
+      "learning_rate": 2.7800598487559975e-05,
+      "loss": 46.7197,
+      "step": 129460
+    },
+    {
+      "epoch": 0.5230751827147226,
+      "grad_norm": 195.2952423095703,
+      "learning_rate": 2.7797129773689118e-05,
+      "loss": 55.5766,
+      "step": 129470
+    },
+    {
+      "epoch": 0.5231155839800903,
+      "grad_norm": 810.8615112304688,
+      "learning_rate": 2.7793661005286774e-05,
+      "loss": 42.1326,
+      "step": 129480
+    },
+    {
+      "epoch": 0.5231559852454579,
+      "grad_norm": 259.7433166503906,
+      "learning_rate": 2.7790192182420578e-05,
+      "loss": 40.0814,
+      "step": 129490
+    },
+    {
+      "epoch": 0.5231963865108256,
+      "grad_norm": 465.7834777832031,
+      "learning_rate": 2.7786723305158136e-05,
+      "loss": 41.8416,
+      "step": 129500
+    },
+    {
+      "epoch": 0.5232367877761932,
+      "grad_norm": 309.6441650390625,
+      "learning_rate": 2.7783254373567103e-05,
+      "loss": 92.4435,
+      "step": 129510
+    },
+    {
+      "epoch": 0.5232771890415607,
+      "grad_norm": 689.2532348632812,
+      "learning_rate": 2.7779785387715078e-05,
+      "loss": 50.8395,
+      "step": 129520
+    },
+    {
+      "epoch": 0.5233175903069284,
+      "grad_norm": 551.0718994140625,
+      "learning_rate": 2.7776316347669722e-05,
+      "loss": 53.9811,
+      "step": 129530
+    },
+    {
+      "epoch": 0.523357991572296,
+      "grad_norm": 792.376953125,
+      "learning_rate": 2.7772847253498636e-05,
+      "loss": 38.2251,
+      "step": 129540
+    },
+    {
+      "epoch": 0.5233983928376637,
+      "grad_norm": 687.2542724609375,
+      "learning_rate": 2.7769378105269467e-05,
+      "loss": 72.9509,
+      "step": 129550
+    },
+    {
+      "epoch": 0.5234387941030313,
+      "grad_norm": 856.1763916015625,
+      "learning_rate": 2.7765908903049848e-05,
+      "loss": 42.2359,
+      "step": 129560
+    },
+    {
+      "epoch": 0.5234791953683989,
+      "grad_norm": 865.5512084960938,
+      "learning_rate": 2.7762439646907417e-05,
+      "loss": 46.5808,
+      "step": 129570
+    },
+    {
+      "epoch": 0.5235195966337666,
+      "grad_norm": 218.27801513671875,
+      "learning_rate": 2.7758970336909795e-05,
+      "loss": 58.7573,
+      "step": 129580
+    },
+    {
+      "epoch": 0.5235599978991342,
+      "grad_norm": 797.5692749023438,
+      "learning_rate": 2.7755500973124625e-05,
+      "loss": 64.8211,
+      "step": 129590
+    },
+    {
+      "epoch": 0.5236003991645019,
+      "grad_norm": 1161.782470703125,
+      "learning_rate": 2.7752031555619555e-05,
+      "loss": 62.3158,
+      "step": 129600
+    },
+    {
+      "epoch": 0.5236408004298695,
+      "grad_norm": 495.8021545410156,
+      "learning_rate": 2.774856208446221e-05,
+      "loss": 34.1338,
+      "step": 129610
+    },
+    {
+      "epoch": 0.5236812016952371,
+      "grad_norm": 538.5377807617188,
+      "learning_rate": 2.7745092559720227e-05,
+      "loss": 84.3193,
+      "step": 129620
+    },
+    {
+      "epoch": 0.5237216029606048,
+      "grad_norm": 472.7118225097656,
+      "learning_rate": 2.7741622981461253e-05,
+      "loss": 88.2121,
+      "step": 129630
+    },
+    {
+      "epoch": 0.5237620042259724,
+      "grad_norm": 680.1412353515625,
+      "learning_rate": 2.773815334975292e-05,
+      "loss": 61.683,
+      "step": 129640
+    },
+    {
+      "epoch": 0.5238024054913399,
+      "grad_norm": 194.98385620117188,
+      "learning_rate": 2.7734683664662892e-05,
+      "loss": 35.9615,
+      "step": 129650
+    },
+    {
+      "epoch": 0.5238428067567076,
+      "grad_norm": 664.1810913085938,
+      "learning_rate": 2.7731213926258794e-05,
+      "loss": 76.9602,
+      "step": 129660
+    },
+    {
+      "epoch": 0.5238832080220752,
+      "grad_norm": 675.6226806640625,
+      "learning_rate": 2.7727744134608263e-05,
+      "loss": 49.6265,
+      "step": 129670
+    },
+    {
+      "epoch": 0.5239236092874429,
+      "grad_norm": 424.91912841796875,
+      "learning_rate": 2.7724274289778974e-05,
+      "loss": 49.0146,
+      "step": 129680
+    },
+    {
+      "epoch": 0.5239640105528105,
+      "grad_norm": 1368.484619140625,
+      "learning_rate": 2.7720804391838544e-05,
+      "loss": 53.2469,
+      "step": 129690
+    },
+    {
+      "epoch": 0.5240044118181781,
+      "grad_norm": 988.5595092773438,
+      "learning_rate": 2.771733444085463e-05,
+      "loss": 72.9639,
+      "step": 129700
+    },
+    {
+      "epoch": 0.5240448130835458,
+      "grad_norm": 1024.283447265625,
+      "learning_rate": 2.771386443689489e-05,
+      "loss": 35.8563,
+      "step": 129710
+    },
+    {
+      "epoch": 0.5240852143489134,
+      "grad_norm": 2501.96142578125,
+      "learning_rate": 2.7710394380026954e-05,
+      "loss": 93.8633,
+      "step": 129720
+    },
+    {
+      "epoch": 0.5241256156142811,
+      "grad_norm": 397.55413818359375,
+      "learning_rate": 2.7706924270318496e-05,
+      "loss": 51.6667,
+      "step": 129730
+    },
+    {
+      "epoch": 0.5241660168796487,
+      "grad_norm": 401.48687744140625,
+      "learning_rate": 2.770345410783715e-05,
+      "loss": 53.5041,
+      "step": 129740
+    },
+    {
+      "epoch": 0.5242064181450163,
+      "grad_norm": 561.6734619140625,
+      "learning_rate": 2.7699983892650573e-05,
+      "loss": 60.5054,
+      "step": 129750
+    },
+    {
+      "epoch": 0.524246819410384,
+      "grad_norm": 1459.1732177734375,
+      "learning_rate": 2.769651362482642e-05,
+      "loss": 80.3514,
+      "step": 129760
+    },
+    {
+      "epoch": 0.5242872206757516,
+      "grad_norm": 211.4439239501953,
+      "learning_rate": 2.7693043304432354e-05,
+      "loss": 53.1344,
+      "step": 129770
+    },
+    {
+      "epoch": 0.5243276219411191,
+      "grad_norm": 535.6337890625,
+      "learning_rate": 2.7689572931536017e-05,
+      "loss": 59.1036,
+      "step": 129780
+    },
+    {
+      "epoch": 0.5243680232064868,
+      "grad_norm": 212.34646606445312,
+      "learning_rate": 2.7686102506205068e-05,
+      "loss": 50.72,
+      "step": 129790
+    },
+    {
+      "epoch": 0.5244084244718544,
+      "grad_norm": 1038.571044921875,
+      "learning_rate": 2.7682632028507167e-05,
+      "loss": 43.0176,
+      "step": 129800
+    },
+    {
+      "epoch": 0.5244488257372221,
+      "grad_norm": 621.1820678710938,
+      "learning_rate": 2.7679161498509976e-05,
+      "loss": 25.86,
+      "step": 129810
+    },
+    {
+      "epoch": 0.5244892270025897,
+      "grad_norm": 441.5834045410156,
+      "learning_rate": 2.7675690916281156e-05,
+      "loss": 54.9453,
+      "step": 129820
+    },
+    {
+      "epoch": 0.5245296282679573,
+      "grad_norm": 317.9010925292969,
+      "learning_rate": 2.7672220281888357e-05,
+      "loss": 95.5096,
+      "step": 129830
+    },
+    {
+      "epoch": 0.524570029533325,
+      "grad_norm": 862.2459106445312,
+      "learning_rate": 2.766874959539925e-05,
+      "loss": 69.8221,
+      "step": 129840
+    },
+    {
+      "epoch": 0.5246104307986926,
+      "grad_norm": 531.95263671875,
+      "learning_rate": 2.76652788568815e-05,
+      "loss": 49.4422,
+      "step": 129850
+    },
+    {
+      "epoch": 0.5246508320640603,
+      "grad_norm": 694.7820434570312,
+      "learning_rate": 2.7661808066402767e-05,
+      "loss": 46.5116,
+      "step": 129860
+    },
+    {
+      "epoch": 0.5246912333294279,
+      "grad_norm": 1010.5025024414062,
+      "learning_rate": 2.765833722403071e-05,
+      "loss": 71.7905,
+      "step": 129870
+    },
+    {
+      "epoch": 0.5247316345947955,
+      "grad_norm": 285.8402099609375,
+      "learning_rate": 2.7654866329833002e-05,
+      "loss": 37.7266,
+      "step": 129880
+    },
+    {
+      "epoch": 0.5247720358601632,
+      "grad_norm": 556.6085205078125,
+      "learning_rate": 2.7651395383877304e-05,
+      "loss": 54.9936,
+      "step": 129890
+    },
+    {
+      "epoch": 0.5248124371255307,
+      "grad_norm": 1335.543701171875,
+      "learning_rate": 2.76479243862313e-05,
+      "loss": 71.0658,
+      "step": 129900
+    },
+    {
+      "epoch": 0.5248528383908984,
+      "grad_norm": 510.7142639160156,
+      "learning_rate": 2.7644453336962633e-05,
+      "loss": 86.5262,
+      "step": 129910
+    },
+    {
+      "epoch": 0.524893239656266,
+      "grad_norm": 629.845458984375,
+      "learning_rate": 2.7640982236138992e-05,
+      "loss": 47.5349,
+      "step": 129920
+    },
+    {
+      "epoch": 0.5249336409216336,
+      "grad_norm": 1113.2122802734375,
+      "learning_rate": 2.7637511083828043e-05,
+      "loss": 66.3521,
+      "step": 129930
+    },
+    {
+      "epoch": 0.5249740421870013,
+      "grad_norm": 365.0244140625,
+      "learning_rate": 2.763403988009746e-05,
+      "loss": 54.905,
+      "step": 129940
+    },
+    {
+      "epoch": 0.5250144434523689,
+      "grad_norm": 543.3900756835938,
+      "learning_rate": 2.7630568625014917e-05,
+      "loss": 48.9247,
+      "step": 129950
+    },
+    {
+      "epoch": 0.5250548447177366,
+      "grad_norm": 381.6798400878906,
+      "learning_rate": 2.7627097318648076e-05,
+      "loss": 39.9984,
+      "step": 129960
+    },
+    {
+      "epoch": 0.5250952459831042,
+      "grad_norm": 1278.4371337890625,
+      "learning_rate": 2.7623625961064618e-05,
+      "loss": 60.6886,
+      "step": 129970
+    },
+    {
+      "epoch": 0.5251356472484718,
+      "grad_norm": 500.2679443359375,
+      "learning_rate": 2.7620154552332232e-05,
+      "loss": 54.9953,
+      "step": 129980
+    },
+    {
+      "epoch": 0.5251760485138395,
+      "grad_norm": 482.51458740234375,
+      "learning_rate": 2.7616683092518576e-05,
+      "loss": 55.6544,
+      "step": 129990
+    },
+    {
+      "epoch": 0.5252164497792071,
+      "grad_norm": 359.9608154296875,
+      "learning_rate": 2.761321158169134e-05,
+      "loss": 64.5541,
+      "step": 130000
+    },
+    {
+      "epoch": 0.5252568510445748,
+      "grad_norm": 398.5904541015625,
+      "learning_rate": 2.7609740019918197e-05,
+      "loss": 53.5339,
+      "step": 130010
+    },
+    {
+      "epoch": 0.5252972523099424,
+      "grad_norm": 698.3560791015625,
+      "learning_rate": 2.7606268407266827e-05,
+      "loss": 61.4191,
+      "step": 130020
+    },
+    {
+      "epoch": 0.5253376535753099,
+      "grad_norm": 768.4830932617188,
+      "learning_rate": 2.7602796743804922e-05,
+      "loss": 47.3355,
+      "step": 130030
+    },
+    {
+      "epoch": 0.5253780548406776,
+      "grad_norm": 775.7195434570312,
+      "learning_rate": 2.7599325029600143e-05,
+      "loss": 25.0439,
+      "step": 130040
+    },
+    {
+      "epoch": 0.5254184561060452,
+      "grad_norm": 1195.4205322265625,
+      "learning_rate": 2.7595853264720184e-05,
+      "loss": 70.0329,
+      "step": 130050
+    },
+    {
+      "epoch": 0.5254588573714128,
+      "grad_norm": 606.2637329101562,
+      "learning_rate": 2.759238144923274e-05,
+      "loss": 53.4816,
+      "step": 130060
+    },
+    {
+      "epoch": 0.5254992586367805,
+      "grad_norm": 469.8358459472656,
+      "learning_rate": 2.7588909583205475e-05,
+      "loss": 49.5405,
+      "step": 130070
+    },
+    {
+      "epoch": 0.5255396599021481,
+      "grad_norm": 648.6143188476562,
+      "learning_rate": 2.7585437666706087e-05,
+      "loss": 69.9888,
+      "step": 130080
+    },
+    {
+      "epoch": 0.5255800611675158,
+      "grad_norm": 612.1096801757812,
+      "learning_rate": 2.758196569980226e-05,
+      "loss": 51.9348,
+      "step": 130090
+    },
+    {
+      "epoch": 0.5256204624328834,
+      "grad_norm": 1045.827392578125,
+      "learning_rate": 2.7578493682561685e-05,
+      "loss": 75.3227,
+      "step": 130100
+    },
+    {
+      "epoch": 0.525660863698251,
+      "grad_norm": 990.9442138671875,
+      "learning_rate": 2.757502161505205e-05,
+      "loss": 73.4528,
+      "step": 130110
+    },
+    {
+      "epoch": 0.5257012649636187,
+      "grad_norm": 510.2374572753906,
+      "learning_rate": 2.7571549497341042e-05,
+      "loss": 60.0831,
+      "step": 130120
+    },
+    {
+      "epoch": 0.5257416662289863,
+      "grad_norm": 1088.753662109375,
+      "learning_rate": 2.756807732949635e-05,
+      "loss": 63.8334,
+      "step": 130130
+    },
+    {
+      "epoch": 0.525782067494354,
+      "grad_norm": 1749.695068359375,
+      "learning_rate": 2.756460511158567e-05,
+      "loss": 54.9062,
+      "step": 130140
+    },
+    {
+      "epoch": 0.5258224687597216,
+      "grad_norm": 992.9537963867188,
+      "learning_rate": 2.756113284367669e-05,
+      "loss": 55.7644,
+      "step": 130150
+    },
+    {
+      "epoch": 0.5258628700250891,
+      "grad_norm": 679.5419311523438,
+      "learning_rate": 2.7557660525837108e-05,
+      "loss": 51.0159,
+      "step": 130160
+    },
+    {
+      "epoch": 0.5259032712904568,
+      "grad_norm": 444.13031005859375,
+      "learning_rate": 2.7554188158134616e-05,
+      "loss": 19.1531,
+      "step": 130170
+    },
+    {
+      "epoch": 0.5259436725558244,
+      "grad_norm": 416.7135314941406,
+      "learning_rate": 2.7550715740636917e-05,
+      "loss": 46.4561,
+      "step": 130180
+    },
+    {
+      "epoch": 0.525984073821192,
+      "grad_norm": 933.5376586914062,
+      "learning_rate": 2.7547243273411695e-05,
+      "loss": 66.3492,
+      "step": 130190
+    },
+    {
+      "epoch": 0.5260244750865597,
+      "grad_norm": 444.9652404785156,
+      "learning_rate": 2.754377075652666e-05,
+      "loss": 50.9244,
+      "step": 130200
+    },
+    {
+      "epoch": 0.5260648763519273,
+      "grad_norm": 652.8956298828125,
+      "learning_rate": 2.7540298190049503e-05,
+      "loss": 59.4898,
+      "step": 130210
+    },
+    {
+      "epoch": 0.526105277617295,
+      "grad_norm": 645.0548095703125,
+      "learning_rate": 2.7536825574047925e-05,
+      "loss": 35.8312,
+      "step": 130220
+    },
+    {
+      "epoch": 0.5261456788826626,
+      "grad_norm": 534.1641845703125,
+      "learning_rate": 2.7533352908589622e-05,
+      "loss": 40.8292,
+      "step": 130230
+    },
+    {
+      "epoch": 0.5261860801480303,
+      "grad_norm": 420.173583984375,
+      "learning_rate": 2.7529880193742297e-05,
+      "loss": 30.2268,
+      "step": 130240
+    },
+    {
+      "epoch": 0.5262264814133979,
+      "grad_norm": 465.4990234375,
+      "learning_rate": 2.7526407429573657e-05,
+      "loss": 64.3663,
+      "step": 130250
+    },
+    {
+      "epoch": 0.5262668826787655,
+      "grad_norm": 511.3827819824219,
+      "learning_rate": 2.7522934616151414e-05,
+      "loss": 56.9852,
+      "step": 130260
+    },
+    {
+      "epoch": 0.5263072839441332,
+      "grad_norm": 624.65087890625,
+      "learning_rate": 2.751946175354325e-05,
+      "loss": 63.1493,
+      "step": 130270
+    },
+    {
+      "epoch": 0.5263476852095008,
+      "grad_norm": 751.966064453125,
+      "learning_rate": 2.7515988841816887e-05,
+      "loss": 69.2688,
+      "step": 130280
+    },
+    {
+      "epoch": 0.5263880864748683,
+      "grad_norm": 907.3563842773438,
+      "learning_rate": 2.7512515881040028e-05,
+      "loss": 54.0088,
+      "step": 130290
+    },
+    {
+      "epoch": 0.526428487740236,
+      "grad_norm": 893.8932495117188,
+      "learning_rate": 2.7509042871280372e-05,
+      "loss": 54.6191,
+      "step": 130300
+    },
+    {
+      "epoch": 0.5264688890056036,
+      "grad_norm": 1132.3380126953125,
+      "learning_rate": 2.750556981260564e-05,
+      "loss": 51.4256,
+      "step": 130310
+    },
+    {
+      "epoch": 0.5265092902709713,
+      "grad_norm": 408.7311096191406,
+      "learning_rate": 2.7502096705083535e-05,
+      "loss": 44.0828,
+      "step": 130320
+    },
+    {
+      "epoch": 0.5265496915363389,
+      "grad_norm": 569.1919555664062,
+      "learning_rate": 2.749862354878176e-05,
+      "loss": 34.8458,
+      "step": 130330
+    },
+    {
+      "epoch": 0.5265900928017065,
+      "grad_norm": 759.1968383789062,
+      "learning_rate": 2.7495150343768034e-05,
+      "loss": 59.1251,
+      "step": 130340
+    },
+    {
+      "epoch": 0.5266304940670742,
+      "grad_norm": 563.669677734375,
+      "learning_rate": 2.7491677090110076e-05,
+      "loss": 48.0987,
+      "step": 130350
+    },
+    {
+      "epoch": 0.5266708953324418,
+      "grad_norm": 1401.8712158203125,
+      "learning_rate": 2.7488203787875577e-05,
+      "loss": 67.7033,
+      "step": 130360
+    },
+    {
+      "epoch": 0.5267112965978095,
+      "grad_norm": 496.8099060058594,
+      "learning_rate": 2.7484730437132278e-05,
+      "loss": 44.0139,
+      "step": 130370
+    },
+    {
+      "epoch": 0.5267516978631771,
+      "grad_norm": 533.447998046875,
+      "learning_rate": 2.7481257037947872e-05,
+      "loss": 47.0373,
+      "step": 130380
+    },
+    {
+      "epoch": 0.5267920991285447,
+      "grad_norm": 825.5570068359375,
+      "learning_rate": 2.7477783590390082e-05,
+      "loss": 63.7042,
+      "step": 130390
+    },
+    {
+      "epoch": 0.5268325003939124,
+      "grad_norm": 577.7359619140625,
+      "learning_rate": 2.747431009452663e-05,
+      "loss": 43.8473,
+      "step": 130400
+    },
+    {
+      "epoch": 0.52687290165928,
+      "grad_norm": 2243.215087890625,
+      "learning_rate": 2.747083655042522e-05,
+      "loss": 58.3573,
+      "step": 130410
+    },
+    {
+      "epoch": 0.5269133029246476,
+      "grad_norm": 897.6618041992188,
+      "learning_rate": 2.7467362958153587e-05,
+      "loss": 85.0911,
+      "step": 130420
+    },
+    {
+      "epoch": 0.5269537041900152,
+      "grad_norm": 976.666259765625,
+      "learning_rate": 2.7463889317779446e-05,
+      "loss": 55.5755,
+      "step": 130430
+    },
+    {
+      "epoch": 0.5269941054553828,
+      "grad_norm": 1133.8804931640625,
+      "learning_rate": 2.7460415629370508e-05,
+      "loss": 59.8554,
+      "step": 130440
+    },
+    {
+      "epoch": 0.5270345067207505,
+      "grad_norm": 757.4534912109375,
+      "learning_rate": 2.7456941892994497e-05,
+      "loss": 43.9233,
+      "step": 130450
+    },
+    {
+      "epoch": 0.5270749079861181,
+      "grad_norm": 962.9010009765625,
+      "learning_rate": 2.7453468108719145e-05,
+      "loss": 65.6795,
+      "step": 130460
+    },
+    {
+      "epoch": 0.5271153092514858,
+      "grad_norm": 394.7574157714844,
+      "learning_rate": 2.744999427661217e-05,
+      "loss": 69.4007,
+      "step": 130470
+    },
+    {
+      "epoch": 0.5271557105168534,
+      "grad_norm": 762.9472045898438,
+      "learning_rate": 2.744652039674129e-05,
+      "loss": 49.1858,
+      "step": 130480
+    },
+    {
+      "epoch": 0.527196111782221,
+      "grad_norm": 2199.121337890625,
+      "learning_rate": 2.7443046469174237e-05,
+      "loss": 77.0622,
+      "step": 130490
+    },
+    {
+      "epoch": 0.5272365130475887,
+      "grad_norm": 370.2516784667969,
+      "learning_rate": 2.7439572493978736e-05,
+      "loss": 51.3727,
+      "step": 130500
+    },
+    {
+      "epoch": 0.5272769143129563,
+      "grad_norm": 1177.574951171875,
+      "learning_rate": 2.7436098471222522e-05,
+      "loss": 47.053,
+      "step": 130510
+    },
+    {
+      "epoch": 0.527317315578324,
+      "grad_norm": 1359.2659912109375,
+      "learning_rate": 2.74326244009733e-05,
+      "loss": 61.0058,
+      "step": 130520
+    },
+    {
+      "epoch": 0.5273577168436916,
+      "grad_norm": 1003.060546875,
+      "learning_rate": 2.7429150283298817e-05,
+      "loss": 52.1394,
+      "step": 130530
+    },
+    {
+      "epoch": 0.5273981181090591,
+      "grad_norm": 1656.4512939453125,
+      "learning_rate": 2.7425676118266808e-05,
+      "loss": 46.5512,
+      "step": 130540
+    },
+    {
+      "epoch": 0.5274385193744268,
+      "grad_norm": 509.01220703125,
+      "learning_rate": 2.7422201905944982e-05,
+      "loss": 52.2351,
+      "step": 130550
+    },
+    {
+      "epoch": 0.5274789206397944,
+      "grad_norm": 1386.0987548828125,
+      "learning_rate": 2.7418727646401094e-05,
+      "loss": 44.8062,
+      "step": 130560
+    },
+    {
+      "epoch": 0.527519321905162,
+      "grad_norm": 708.5176391601562,
+      "learning_rate": 2.741525333970285e-05,
+      "loss": 36.5751,
+      "step": 130570
+    },
+    {
+      "epoch": 0.5275597231705297,
+      "grad_norm": 3067.193115234375,
+      "learning_rate": 2.7411778985918006e-05,
+      "loss": 98.0832,
+      "step": 130580
+    },
+    {
+      "epoch": 0.5276001244358973,
+      "grad_norm": 598.675537109375,
+      "learning_rate": 2.7408304585114298e-05,
+      "loss": 60.4569,
+      "step": 130590
+    },
+    {
+      "epoch": 0.527640525701265,
+      "grad_norm": 821.6907348632812,
+      "learning_rate": 2.7404830137359444e-05,
+      "loss": 71.1968,
+      "step": 130600
+    },
+    {
+      "epoch": 0.5276809269666326,
+      "grad_norm": 796.5547485351562,
+      "learning_rate": 2.740135564272119e-05,
+      "loss": 64.8786,
+      "step": 130610
+    },
+    {
+      "epoch": 0.5277213282320002,
+      "grad_norm": 587.248291015625,
+      "learning_rate": 2.7397881101267263e-05,
+      "loss": 61.1327,
+      "step": 130620
+    },
+    {
+      "epoch": 0.5277617294973679,
+      "grad_norm": 603.5186157226562,
+      "learning_rate": 2.7394406513065423e-05,
+      "loss": 54.6609,
+      "step": 130630
+    },
+    {
+      "epoch": 0.5278021307627355,
+      "grad_norm": 1664.0540771484375,
+      "learning_rate": 2.739093187818339e-05,
+      "loss": 69.9045,
+      "step": 130640
+    },
+    {
+      "epoch": 0.5278425320281032,
+      "grad_norm": 847.181640625,
+      "learning_rate": 2.7387457196688908e-05,
+      "loss": 60.1619,
+      "step": 130650
+    },
+    {
+      "epoch": 0.5278829332934708,
+      "grad_norm": 1435.385498046875,
+      "learning_rate": 2.7383982468649714e-05,
+      "loss": 48.167,
+      "step": 130660
+    },
+    {
+      "epoch": 0.5279233345588383,
+      "grad_norm": 884.9189453125,
+      "learning_rate": 2.738050769413357e-05,
+      "loss": 57.8182,
+      "step": 130670
+    },
+    {
+      "epoch": 0.527963735824206,
+      "grad_norm": 568.2256469726562,
+      "learning_rate": 2.7377032873208186e-05,
+      "loss": 66.162,
+      "step": 130680
+    },
+    {
+      "epoch": 0.5280041370895736,
+      "grad_norm": 744.9805908203125,
+      "learning_rate": 2.737355800594133e-05,
+      "loss": 60.4696,
+      "step": 130690
+    },
+    {
+      "epoch": 0.5280445383549413,
+      "grad_norm": 907.050537109375,
+      "learning_rate": 2.7370083092400735e-05,
+      "loss": 43.196,
+      "step": 130700
+    },
+    {
+      "epoch": 0.5280849396203089,
+      "grad_norm": 778.4954833984375,
+      "learning_rate": 2.7366608132654154e-05,
+      "loss": 51.7738,
+      "step": 130710
+    },
+    {
+      "epoch": 0.5281253408856765,
+      "grad_norm": 405.15338134765625,
+      "learning_rate": 2.7363133126769325e-05,
+      "loss": 64.0481,
+      "step": 130720
+    },
+    {
+      "epoch": 0.5281657421510442,
+      "grad_norm": 591.9571533203125,
+      "learning_rate": 2.735965807481401e-05,
+      "loss": 45.1595,
+      "step": 130730
+    },
+    {
+      "epoch": 0.5282061434164118,
+      "grad_norm": 985.9591064453125,
+      "learning_rate": 2.7356182976855934e-05,
+      "loss": 35.7408,
+      "step": 130740
+    },
+    {
+      "epoch": 0.5282465446817795,
+      "grad_norm": 460.90887451171875,
+      "learning_rate": 2.7352707832962865e-05,
+      "loss": 52.3368,
+      "step": 130750
+    },
+    {
+      "epoch": 0.5282869459471471,
+      "grad_norm": 900.5870971679688,
+      "learning_rate": 2.734923264320254e-05,
+      "loss": 31.7685,
+      "step": 130760
+    },
+    {
+      "epoch": 0.5283273472125147,
+      "grad_norm": 851.9434814453125,
+      "learning_rate": 2.7345757407642714e-05,
+      "loss": 41.7662,
+      "step": 130770
+    },
+    {
+      "epoch": 0.5283677484778824,
+      "grad_norm": 2293.946533203125,
+      "learning_rate": 2.7342282126351144e-05,
+      "loss": 96.9833,
+      "step": 130780
+    },
+    {
+      "epoch": 0.52840814974325,
+      "grad_norm": 837.9186401367188,
+      "learning_rate": 2.7338806799395577e-05,
+      "loss": 53.9366,
+      "step": 130790
+    },
+    {
+      "epoch": 0.5284485510086175,
+      "grad_norm": 314.82135009765625,
+      "learning_rate": 2.733533142684377e-05,
+      "loss": 56.3716,
+      "step": 130800
+    },
+    {
+      "epoch": 0.5284889522739852,
+      "grad_norm": 918.69287109375,
+      "learning_rate": 2.7331856008763472e-05,
+      "loss": 50.4389,
+      "step": 130810
+    },
+    {
+      "epoch": 0.5285293535393528,
+      "grad_norm": 775.286865234375,
+      "learning_rate": 2.7328380545222436e-05,
+      "loss": 64.7548,
+      "step": 130820
+    },
+    {
+      "epoch": 0.5285697548047205,
+      "grad_norm": 461.6344909667969,
+      "learning_rate": 2.732490503628843e-05,
+      "loss": 53.2063,
+      "step": 130830
+    },
+    {
+      "epoch": 0.5286101560700881,
+      "grad_norm": 415.2874755859375,
+      "learning_rate": 2.73214294820292e-05,
+      "loss": 94.1432,
+      "step": 130840
+    },
+    {
+      "epoch": 0.5286505573354557,
+      "grad_norm": 718.537109375,
+      "learning_rate": 2.7317953882512504e-05,
+      "loss": 47.0089,
+      "step": 130850
+    },
+    {
+      "epoch": 0.5286909586008234,
+      "grad_norm": 524.0968017578125,
+      "learning_rate": 2.7314478237806107e-05,
+      "loss": 50.3915,
+      "step": 130860
+    },
+    {
+      "epoch": 0.528731359866191,
+      "grad_norm": 685.7463989257812,
+      "learning_rate": 2.7311002547977766e-05,
+      "loss": 47.7177,
+      "step": 130870
+    },
+    {
+      "epoch": 0.5287717611315587,
+      "grad_norm": 658.6242065429688,
+      "learning_rate": 2.730752681309524e-05,
+      "loss": 44.9115,
+      "step": 130880
+    },
+    {
+      "epoch": 0.5288121623969263,
+      "grad_norm": 831.5072021484375,
+      "learning_rate": 2.730405103322629e-05,
+      "loss": 36.1585,
+      "step": 130890
+    },
+    {
+      "epoch": 0.5288525636622939,
+      "grad_norm": 2113.5791015625,
+      "learning_rate": 2.7300575208438683e-05,
+      "loss": 64.7953,
+      "step": 130900
+    },
+    {
+      "epoch": 0.5288929649276616,
+      "grad_norm": 1694.7403564453125,
+      "learning_rate": 2.729709933880017e-05,
+      "loss": 64.9504,
+      "step": 130910
+    },
+    {
+      "epoch": 0.5289333661930292,
+      "grad_norm": 959.0403442382812,
+      "learning_rate": 2.7293623424378535e-05,
+      "loss": 62.3716,
+      "step": 130920
+    },
+    {
+      "epoch": 0.5289737674583967,
+      "grad_norm": 2491.924072265625,
+      "learning_rate": 2.7290147465241517e-05,
+      "loss": 52.1772,
+      "step": 130930
+    },
+    {
+      "epoch": 0.5290141687237644,
+      "grad_norm": 243.4221954345703,
+      "learning_rate": 2.7286671461456897e-05,
+      "loss": 38.3009,
+      "step": 130940
+    },
+    {
+      "epoch": 0.529054569989132,
+      "grad_norm": 1660.8577880859375,
+      "learning_rate": 2.7283195413092445e-05,
+      "loss": 60.126,
+      "step": 130950
+    },
+    {
+      "epoch": 0.5290949712544997,
+      "grad_norm": 609.1519165039062,
+      "learning_rate": 2.7279719320215924e-05,
+      "loss": 59.0399,
+      "step": 130960
+    },
+    {
+      "epoch": 0.5291353725198673,
+      "grad_norm": 379.8237609863281,
+      "learning_rate": 2.7276243182895094e-05,
+      "loss": 64.3225,
+      "step": 130970
+    },
+    {
+      "epoch": 0.529175773785235,
+      "grad_norm": 535.1643676757812,
+      "learning_rate": 2.7272767001197742e-05,
+      "loss": 37.2136,
+      "step": 130980
+    },
+    {
+      "epoch": 0.5292161750506026,
+      "grad_norm": 587.9727783203125,
+      "learning_rate": 2.726929077519162e-05,
+      "loss": 41.9734,
+      "step": 130990
+    },
+    {
+      "epoch": 0.5292565763159702,
+      "grad_norm": 652.3465576171875,
+      "learning_rate": 2.726581450494451e-05,
+      "loss": 83.3366,
+      "step": 131000
+    },
+    {
+      "epoch": 0.5292969775813379,
+      "grad_norm": 395.1925964355469,
+      "learning_rate": 2.7262338190524173e-05,
+      "loss": 70.538,
+      "step": 131010
+    },
+    {
+      "epoch": 0.5293373788467055,
+      "grad_norm": 602.4820556640625,
+      "learning_rate": 2.7258861831998388e-05,
+      "loss": 55.8564,
+      "step": 131020
+    },
+    {
+      "epoch": 0.5293777801120731,
+      "grad_norm": 2387.6884765625,
+      "learning_rate": 2.7255385429434932e-05,
+      "loss": 77.701,
+      "step": 131030
+    },
+    {
+      "epoch": 0.5294181813774408,
+      "grad_norm": 626.4151000976562,
+      "learning_rate": 2.725190898290158e-05,
+      "loss": 71.4265,
+      "step": 131040
+    },
+    {
+      "epoch": 0.5294585826428083,
+      "grad_norm": 1227.336669921875,
+      "learning_rate": 2.7248432492466096e-05,
+      "loss": 63.7168,
+      "step": 131050
+    },
+    {
+      "epoch": 0.529498983908176,
+      "grad_norm": 506.81475830078125,
+      "learning_rate": 2.7244955958196265e-05,
+      "loss": 38.2177,
+      "step": 131060
+    },
+    {
+      "epoch": 0.5295393851735436,
+      "grad_norm": 550.5950317382812,
+      "learning_rate": 2.7241479380159868e-05,
+      "loss": 69.3767,
+      "step": 131070
+    },
+    {
+      "epoch": 0.5295797864389112,
+      "grad_norm": 751.3677368164062,
+      "learning_rate": 2.723800275842468e-05,
+      "loss": 45.0376,
+      "step": 131080
+    },
+    {
+      "epoch": 0.5296201877042789,
+      "grad_norm": 459.6380310058594,
+      "learning_rate": 2.7234526093058464e-05,
+      "loss": 59.1021,
+      "step": 131090
+    },
+    {
+      "epoch": 0.5296605889696465,
+      "grad_norm": 463.3095703125,
+      "learning_rate": 2.7231049384129016e-05,
+      "loss": 37.5911,
+      "step": 131100
+    },
+    {
+      "epoch": 0.5297009902350142,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7227572631704107e-05,
+      "loss": 52.0239,
+      "step": 131110
+    },
+    {
+      "epoch": 0.5297413915003818,
+      "grad_norm": 1287.0667724609375,
+      "learning_rate": 2.7224095835851525e-05,
+      "loss": 85.9998,
+      "step": 131120
+    },
+    {
+      "epoch": 0.5297817927657494,
+      "grad_norm": 710.130615234375,
+      "learning_rate": 2.722061899663905e-05,
+      "loss": 41.3392,
+      "step": 131130
+    },
+    {
+      "epoch": 0.5298221940311171,
+      "grad_norm": 241.11277770996094,
+      "learning_rate": 2.7217142114134463e-05,
+      "loss": 75.5654,
+      "step": 131140
+    },
+    {
+      "epoch": 0.5298625952964847,
+      "grad_norm": 2309.976806640625,
+      "learning_rate": 2.7213665188405556e-05,
+      "loss": 62.305,
+      "step": 131150
+    },
+    {
+      "epoch": 0.5299029965618524,
+      "grad_norm": 592.4625244140625,
+      "learning_rate": 2.721018821952011e-05,
+      "loss": 49.6338,
+      "step": 131160
+    },
+    {
+      "epoch": 0.52994339782722,
+      "grad_norm": 3443.605224609375,
+      "learning_rate": 2.7206711207545893e-05,
+      "loss": 65.5017,
+      "step": 131170
+    },
+    {
+      "epoch": 0.5299837990925875,
+      "grad_norm": 924.45654296875,
+      "learning_rate": 2.7203234152550712e-05,
+      "loss": 52.5658,
+      "step": 131180
+    },
+    {
+      "epoch": 0.5300242003579552,
+      "grad_norm": 385.39910888671875,
+      "learning_rate": 2.719975705460234e-05,
+      "loss": 35.2366,
+      "step": 131190
+    },
+    {
+      "epoch": 0.5300646016233228,
+      "grad_norm": 709.4951171875,
+      "learning_rate": 2.7196279913768584e-05,
+      "loss": 56.7616,
+      "step": 131200
+    },
+    {
+      "epoch": 0.5301050028886904,
+      "grad_norm": 849.27294921875,
+      "learning_rate": 2.719280273011721e-05,
+      "loss": 62.0372,
+      "step": 131210
+    },
+    {
+      "epoch": 0.5301454041540581,
+      "grad_norm": 943.2574462890625,
+      "learning_rate": 2.7189325503716022e-05,
+      "loss": 81.1754,
+      "step": 131220
+    },
+    {
+      "epoch": 0.5301858054194257,
+      "grad_norm": 839.4765014648438,
+      "learning_rate": 2.7185848234632803e-05,
+      "loss": 65.2041,
+      "step": 131230
+    },
+    {
+      "epoch": 0.5302262066847934,
+      "grad_norm": 879.865966796875,
+      "learning_rate": 2.7182370922935353e-05,
+      "loss": 57.2173,
+      "step": 131240
+    },
+    {
+      "epoch": 0.530266607950161,
+      "grad_norm": 980.9358520507812,
+      "learning_rate": 2.717889356869146e-05,
+      "loss": 47.6376,
+      "step": 131250
+    },
+    {
+      "epoch": 0.5303070092155286,
+      "grad_norm": 680.2994384765625,
+      "learning_rate": 2.717541617196891e-05,
+      "loss": 35.0096,
+      "step": 131260
+    },
+    {
+      "epoch": 0.5303474104808963,
+      "grad_norm": 870.2879638671875,
+      "learning_rate": 2.71719387328355e-05,
+      "loss": 54.6478,
+      "step": 131270
+    },
+    {
+      "epoch": 0.5303878117462639,
+      "grad_norm": 456.5863952636719,
+      "learning_rate": 2.716846125135903e-05,
+      "loss": 66.5173,
+      "step": 131280
+    },
+    {
+      "epoch": 0.5304282130116316,
+      "grad_norm": 1249.5938720703125,
+      "learning_rate": 2.716498372760729e-05,
+      "loss": 44.1854,
+      "step": 131290
+    },
+    {
+      "epoch": 0.5304686142769992,
+      "grad_norm": 364.5703430175781,
+      "learning_rate": 2.7161506161648076e-05,
+      "loss": 58.9766,
+      "step": 131300
+    },
+    {
+      "epoch": 0.5305090155423667,
+      "grad_norm": 890.297119140625,
+      "learning_rate": 2.7158028553549187e-05,
+      "loss": 56.6682,
+      "step": 131310
+    },
+    {
+      "epoch": 0.5305494168077344,
+      "grad_norm": 925.9620361328125,
+      "learning_rate": 2.715455090337842e-05,
+      "loss": 43.9666,
+      "step": 131320
+    },
+    {
+      "epoch": 0.530589818073102,
+      "grad_norm": 511.55609130859375,
+      "learning_rate": 2.715107321120358e-05,
+      "loss": 51.9629,
+      "step": 131330
+    },
+    {
+      "epoch": 0.5306302193384697,
+      "grad_norm": 547.1328125,
+      "learning_rate": 2.7147595477092457e-05,
+      "loss": 43.9675,
+      "step": 131340
+    },
+    {
+      "epoch": 0.5306706206038373,
+      "grad_norm": 784.9970703125,
+      "learning_rate": 2.7144117701112846e-05,
+      "loss": 65.4148,
+      "step": 131350
+    },
+    {
+      "epoch": 0.5307110218692049,
+      "grad_norm": 653.8527221679688,
+      "learning_rate": 2.7140639883332564e-05,
+      "loss": 64.9822,
+      "step": 131360
+    },
+    {
+      "epoch": 0.5307514231345726,
+      "grad_norm": 480.8045959472656,
+      "learning_rate": 2.713716202381941e-05,
+      "loss": 53.8297,
+      "step": 131370
+    },
+    {
+      "epoch": 0.5307918243999402,
+      "grad_norm": 251.93240356445312,
+      "learning_rate": 2.713368412264118e-05,
+      "loss": 55.8464,
+      "step": 131380
+    },
+    {
+      "epoch": 0.5308322256653079,
+      "grad_norm": 609.9501953125,
+      "learning_rate": 2.713020617986567e-05,
+      "loss": 38.5014,
+      "step": 131390
+    },
+    {
+      "epoch": 0.5308726269306755,
+      "grad_norm": 921.5916748046875,
+      "learning_rate": 2.7126728195560702e-05,
+      "loss": 52.8066,
+      "step": 131400
+    },
+    {
+      "epoch": 0.5309130281960431,
+      "grad_norm": 912.7449951171875,
+      "learning_rate": 2.7123250169794075e-05,
+      "loss": 54.2252,
+      "step": 131410
+    },
+    {
+      "epoch": 0.5309534294614108,
+      "grad_norm": 718.9556274414062,
+      "learning_rate": 2.711977210263359e-05,
+      "loss": 57.7856,
+      "step": 131420
+    },
+    {
+      "epoch": 0.5309938307267784,
+      "grad_norm": 457.2162170410156,
+      "learning_rate": 2.7116293994147053e-05,
+      "loss": 64.9568,
+      "step": 131430
+    },
+    {
+      "epoch": 0.5310342319921459,
+      "grad_norm": 621.4869995117188,
+      "learning_rate": 2.711281584440228e-05,
+      "loss": 62.9255,
+      "step": 131440
+    },
+    {
+      "epoch": 0.5310746332575136,
+      "grad_norm": 792.2979125976562,
+      "learning_rate": 2.710933765346707e-05,
+      "loss": 47.9132,
+      "step": 131450
+    },
+    {
+      "epoch": 0.5311150345228812,
+      "grad_norm": 539.8186645507812,
+      "learning_rate": 2.710585942140924e-05,
+      "loss": 54.2183,
+      "step": 131460
+    },
+    {
+      "epoch": 0.5311554357882489,
+      "grad_norm": 558.9265747070312,
+      "learning_rate": 2.710238114829659e-05,
+      "loss": 49.2721,
+      "step": 131470
+    },
+    {
+      "epoch": 0.5311958370536165,
+      "grad_norm": 709.4071655273438,
+      "learning_rate": 2.7098902834196943e-05,
+      "loss": 58.9437,
+      "step": 131480
+    },
+    {
+      "epoch": 0.5312362383189841,
+      "grad_norm": 654.2186889648438,
+      "learning_rate": 2.7095424479178106e-05,
+      "loss": 36.4761,
+      "step": 131490
+    },
+    {
+      "epoch": 0.5312766395843518,
+      "grad_norm": 889.7582397460938,
+      "learning_rate": 2.7091946083307896e-05,
+      "loss": 68.6755,
+      "step": 131500
+    },
+    {
+      "epoch": 0.5313170408497194,
+      "grad_norm": 1404.284912109375,
+      "learning_rate": 2.708846764665411e-05,
+      "loss": 67.0801,
+      "step": 131510
+    },
+    {
+      "epoch": 0.5313574421150871,
+      "grad_norm": 774.0541381835938,
+      "learning_rate": 2.7084989169284568e-05,
+      "loss": 70.2885,
+      "step": 131520
+    },
+    {
+      "epoch": 0.5313978433804547,
+      "grad_norm": 676.1631469726562,
+      "learning_rate": 2.70815106512671e-05,
+      "loss": 57.3858,
+      "step": 131530
+    },
+    {
+      "epoch": 0.5314382446458223,
+      "grad_norm": 288.2684631347656,
+      "learning_rate": 2.7078032092669502e-05,
+      "loss": 57.0647,
+      "step": 131540
+    },
+    {
+      "epoch": 0.53147864591119,
+      "grad_norm": 692.8833618164062,
+      "learning_rate": 2.70745534935596e-05,
+      "loss": 52.4784,
+      "step": 131550
+    },
+    {
+      "epoch": 0.5315190471765576,
+      "grad_norm": 360.534912109375,
+      "learning_rate": 2.707107485400521e-05,
+      "loss": 76.2931,
+      "step": 131560
+    },
+    {
+      "epoch": 0.5315594484419252,
+      "grad_norm": 902.7174682617188,
+      "learning_rate": 2.7067596174074155e-05,
+      "loss": 66.0437,
+      "step": 131570
+    },
+    {
+      "epoch": 0.5315998497072928,
+      "grad_norm": 1684.62451171875,
+      "learning_rate": 2.7064117453834243e-05,
+      "loss": 48.1203,
+      "step": 131580
+    },
+    {
+      "epoch": 0.5316402509726604,
+      "grad_norm": 74.62276458740234,
+      "learning_rate": 2.70606386933533e-05,
+      "loss": 69.9313,
+      "step": 131590
+    },
+    {
+      "epoch": 0.5316806522380281,
+      "grad_norm": 347.76605224609375,
+      "learning_rate": 2.705715989269914e-05,
+      "loss": 67.9091,
+      "step": 131600
+    },
+    {
+      "epoch": 0.5317210535033957,
+      "grad_norm": 2573.411865234375,
+      "learning_rate": 2.70536810519396e-05,
+      "loss": 58.865,
+      "step": 131610
+    },
+    {
+      "epoch": 0.5317614547687634,
+      "grad_norm": 265.0232849121094,
+      "learning_rate": 2.705020217114248e-05,
+      "loss": 47.0783,
+      "step": 131620
+    },
+    {
+      "epoch": 0.531801856034131,
+      "grad_norm": 801.228515625,
+      "learning_rate": 2.7046723250375617e-05,
+      "loss": 51.892,
+      "step": 131630
+    },
+    {
+      "epoch": 0.5318422572994986,
+      "grad_norm": 833.2240600585938,
+      "learning_rate": 2.7043244289706826e-05,
+      "loss": 49.0188,
+      "step": 131640
+    },
+    {
+      "epoch": 0.5318826585648663,
+      "grad_norm": 984.8556518554688,
+      "learning_rate": 2.7039765289203946e-05,
+      "loss": 55.2805,
+      "step": 131650
+    },
+    {
+      "epoch": 0.5319230598302339,
+      "grad_norm": 779.8922119140625,
+      "learning_rate": 2.703628624893478e-05,
+      "loss": 40.7662,
+      "step": 131660
+    },
+    {
+      "epoch": 0.5319634610956016,
+      "grad_norm": 687.851318359375,
+      "learning_rate": 2.703280716896717e-05,
+      "loss": 71.506,
+      "step": 131670
+    },
+    {
+      "epoch": 0.5320038623609692,
+      "grad_norm": 561.60888671875,
+      "learning_rate": 2.702932804936894e-05,
+      "loss": 76.0556,
+      "step": 131680
+    },
+    {
+      "epoch": 0.5320442636263367,
+      "grad_norm": 597.3526611328125,
+      "learning_rate": 2.7025848890207917e-05,
+      "loss": 51.0563,
+      "step": 131690
+    },
+    {
+      "epoch": 0.5320846648917044,
+      "grad_norm": 1162.3173828125,
+      "learning_rate": 2.7022369691551917e-05,
+      "loss": 67.7418,
+      "step": 131700
+    },
+    {
+      "epoch": 0.532125066157072,
+      "grad_norm": 1007.927734375,
+      "learning_rate": 2.7018890453468788e-05,
+      "loss": 55.7545,
+      "step": 131710
+    },
+    {
+      "epoch": 0.5321654674224396,
+      "grad_norm": 583.6637573242188,
+      "learning_rate": 2.7015411176026344e-05,
+      "loss": 59.1643,
+      "step": 131720
+    },
+    {
+      "epoch": 0.5322058686878073,
+      "grad_norm": 503.95025634765625,
+      "learning_rate": 2.7011931859292427e-05,
+      "loss": 57.7244,
+      "step": 131730
+    },
+    {
+      "epoch": 0.5322462699531749,
+      "grad_norm": 545.329345703125,
+      "learning_rate": 2.7008452503334858e-05,
+      "loss": 46.0308,
+      "step": 131740
+    },
+    {
+      "epoch": 0.5322866712185426,
+      "grad_norm": 380.2509765625,
+      "learning_rate": 2.7004973108221472e-05,
+      "loss": 40.5332,
+      "step": 131750
+    },
+    {
+      "epoch": 0.5323270724839102,
+      "grad_norm": 740.1156005859375,
+      "learning_rate": 2.700149367402011e-05,
+      "loss": 60.6691,
+      "step": 131760
+    },
+    {
+      "epoch": 0.5323674737492778,
+      "grad_norm": 1089.05712890625,
+      "learning_rate": 2.69980142007986e-05,
+      "loss": 65.5629,
+      "step": 131770
+    },
+    {
+      "epoch": 0.5324078750146455,
+      "grad_norm": 537.0314331054688,
+      "learning_rate": 2.699453468862477e-05,
+      "loss": 55.9601,
+      "step": 131780
+    },
+    {
+      "epoch": 0.5324482762800131,
+      "grad_norm": 266.5636291503906,
+      "learning_rate": 2.699105513756645e-05,
+      "loss": 59.2656,
+      "step": 131790
+    },
+    {
+      "epoch": 0.5324886775453808,
+      "grad_norm": 889.6873779296875,
+      "learning_rate": 2.6987575547691497e-05,
+      "loss": 76.0885,
+      "step": 131800
+    },
+    {
+      "epoch": 0.5325290788107484,
+      "grad_norm": 1049.7486572265625,
+      "learning_rate": 2.698409591906773e-05,
+      "loss": 50.9139,
+      "step": 131810
+    },
+    {
+      "epoch": 0.5325694800761159,
+      "grad_norm": 820.5618286132812,
+      "learning_rate": 2.6980616251762997e-05,
+      "loss": 36.7931,
+      "step": 131820
+    },
+    {
+      "epoch": 0.5326098813414836,
+      "grad_norm": 643.490966796875,
+      "learning_rate": 2.6977136545845122e-05,
+      "loss": 56.8342,
+      "step": 131830
+    },
+    {
+      "epoch": 0.5326502826068512,
+      "grad_norm": 853.44970703125,
+      "learning_rate": 2.6973656801381963e-05,
+      "loss": 60.0717,
+      "step": 131840
+    },
+    {
+      "epoch": 0.5326906838722189,
+      "grad_norm": 440.1711730957031,
+      "learning_rate": 2.697017701844134e-05,
+      "loss": 45.1691,
+      "step": 131850
+    },
+    {
+      "epoch": 0.5327310851375865,
+      "grad_norm": 468.0743408203125,
+      "learning_rate": 2.6966697197091108e-05,
+      "loss": 54.0596,
+      "step": 131860
+    },
+    {
+      "epoch": 0.5327714864029541,
+      "grad_norm": 1050.264892578125,
+      "learning_rate": 2.69632173373991e-05,
+      "loss": 55.2322,
+      "step": 131870
+    },
+    {
+      "epoch": 0.5328118876683218,
+      "grad_norm": 720.6689453125,
+      "learning_rate": 2.695973743943315e-05,
+      "loss": 55.2848,
+      "step": 131880
+    },
+    {
+      "epoch": 0.5328522889336894,
+      "grad_norm": 1108.028564453125,
+      "learning_rate": 2.6956257503261116e-05,
+      "loss": 40.9585,
+      "step": 131890
+    },
+    {
+      "epoch": 0.532892690199057,
+      "grad_norm": 1092.5072021484375,
+      "learning_rate": 2.695277752895084e-05,
+      "loss": 44.4573,
+      "step": 131900
+    },
+    {
+      "epoch": 0.5329330914644247,
+      "grad_norm": 625.7965087890625,
+      "learning_rate": 2.6949297516570156e-05,
+      "loss": 28.5109,
+      "step": 131910
+    },
+    {
+      "epoch": 0.5329734927297923,
+      "grad_norm": 681.8804931640625,
+      "learning_rate": 2.6945817466186912e-05,
+      "loss": 43.6512,
+      "step": 131920
+    },
+    {
+      "epoch": 0.53301389399516,
+      "grad_norm": 756.3519897460938,
+      "learning_rate": 2.694233737786896e-05,
+      "loss": 76.8343,
+      "step": 131930
+    },
+    {
+      "epoch": 0.5330542952605276,
+      "grad_norm": 1092.2066650390625,
+      "learning_rate": 2.693885725168414e-05,
+      "loss": 46.8568,
+      "step": 131940
+    },
+    {
+      "epoch": 0.5330946965258951,
+      "grad_norm": 988.5387573242188,
+      "learning_rate": 2.6935377087700297e-05,
+      "loss": 73.0625,
+      "step": 131950
+    },
+    {
+      "epoch": 0.5331350977912628,
+      "grad_norm": 940.03076171875,
+      "learning_rate": 2.693189688598528e-05,
+      "loss": 63.5997,
+      "step": 131960
+    },
+    {
+      "epoch": 0.5331754990566304,
+      "grad_norm": 372.99151611328125,
+      "learning_rate": 2.6928416646606936e-05,
+      "loss": 57.4221,
+      "step": 131970
+    },
+    {
+      "epoch": 0.5332159003219981,
+      "grad_norm": 1288.4371337890625,
+      "learning_rate": 2.6924936369633125e-05,
+      "loss": 48.3105,
+      "step": 131980
+    },
+    {
+      "epoch": 0.5332563015873657,
+      "grad_norm": 2535.275634765625,
+      "learning_rate": 2.6921456055131683e-05,
+      "loss": 67.3763,
+      "step": 131990
+    },
+    {
+      "epoch": 0.5332967028527333,
+      "grad_norm": 3868.3837890625,
+      "learning_rate": 2.6917975703170466e-05,
+      "loss": 65.5701,
+      "step": 132000
+    },
+    {
+      "epoch": 0.533337104118101,
+      "grad_norm": 810.7564697265625,
+      "learning_rate": 2.691449531381733e-05,
+      "loss": 64.9525,
+      "step": 132010
+    },
+    {
+      "epoch": 0.5333775053834686,
+      "grad_norm": 1662.32373046875,
+      "learning_rate": 2.6911014887140122e-05,
+      "loss": 57.4445,
+      "step": 132020
+    },
+    {
+      "epoch": 0.5334179066488363,
+      "grad_norm": 528.4577026367188,
+      "learning_rate": 2.6907534423206692e-05,
+      "loss": 49.0796,
+      "step": 132030
+    },
+    {
+      "epoch": 0.5334583079142039,
+      "grad_norm": 798.1748657226562,
+      "learning_rate": 2.6904053922084895e-05,
+      "loss": 51.6413,
+      "step": 132040
+    },
+    {
+      "epoch": 0.5334987091795715,
+      "grad_norm": 589.7070922851562,
+      "learning_rate": 2.6900573383842583e-05,
+      "loss": 33.4644,
+      "step": 132050
+    },
+    {
+      "epoch": 0.5335391104449392,
+      "grad_norm": 757.95458984375,
+      "learning_rate": 2.689709280854762e-05,
+      "loss": 52.9057,
+      "step": 132060
+    },
+    {
+      "epoch": 0.5335795117103068,
+      "grad_norm": 567.9570922851562,
+      "learning_rate": 2.6893612196267853e-05,
+      "loss": 46.4571,
+      "step": 132070
+    },
+    {
+      "epoch": 0.5336199129756743,
+      "grad_norm": 732.6300048828125,
+      "learning_rate": 2.6890131547071147e-05,
+      "loss": 66.3549,
+      "step": 132080
+    },
+    {
+      "epoch": 0.533660314241042,
+      "grad_norm": 1564.1744384765625,
+      "learning_rate": 2.6886650861025343e-05,
+      "loss": 82.0941,
+      "step": 132090
+    },
+    {
+      "epoch": 0.5337007155064096,
+      "grad_norm": 6092.22314453125,
+      "learning_rate": 2.6883170138198323e-05,
+      "loss": 73.8271,
+      "step": 132100
+    },
+    {
+      "epoch": 0.5337411167717773,
+      "grad_norm": 3677.495361328125,
+      "learning_rate": 2.6879689378657923e-05,
+      "loss": 83.4729,
+      "step": 132110
+    },
+    {
+      "epoch": 0.5337815180371449,
+      "grad_norm": 830.318603515625,
+      "learning_rate": 2.6876208582472012e-05,
+      "loss": 43.1947,
+      "step": 132120
+    },
+    {
+      "epoch": 0.5338219193025125,
+      "grad_norm": 861.20849609375,
+      "learning_rate": 2.687272774970845e-05,
+      "loss": 79.1685,
+      "step": 132130
+    },
+    {
+      "epoch": 0.5338623205678802,
+      "grad_norm": 541.7510375976562,
+      "learning_rate": 2.6869246880435095e-05,
+      "loss": 50.6851,
+      "step": 132140
+    },
+    {
+      "epoch": 0.5339027218332478,
+      "grad_norm": 2233.982177734375,
+      "learning_rate": 2.686576597471981e-05,
+      "loss": 64.9713,
+      "step": 132150
+    },
+    {
+      "epoch": 0.5339431230986155,
+      "grad_norm": 801.9125366210938,
+      "learning_rate": 2.686228503263045e-05,
+      "loss": 67.6367,
+      "step": 132160
+    },
+    {
+      "epoch": 0.5339835243639831,
+      "grad_norm": 577.6140747070312,
+      "learning_rate": 2.685880405423489e-05,
+      "loss": 56.8992,
+      "step": 132170
+    },
+    {
+      "epoch": 0.5340239256293507,
+      "grad_norm": 303.9769592285156,
+      "learning_rate": 2.6855323039601e-05,
+      "loss": 55.4841,
+      "step": 132180
+    },
+    {
+      "epoch": 0.5340643268947184,
+      "grad_norm": 0.0,
+      "learning_rate": 2.685184198879662e-05,
+      "loss": 79.864,
+      "step": 132190
+    },
+    {
+      "epoch": 0.534104728160086,
+      "grad_norm": 1271.53759765625,
+      "learning_rate": 2.684836090188963e-05,
+      "loss": 53.1909,
+      "step": 132200
+    },
+    {
+      "epoch": 0.5341451294254536,
+      "grad_norm": 390.48504638671875,
+      "learning_rate": 2.6844879778947884e-05,
+      "loss": 45.1817,
+      "step": 132210
+    },
+    {
+      "epoch": 0.5341855306908212,
+      "grad_norm": 1045.8829345703125,
+      "learning_rate": 2.6841398620039273e-05,
+      "loss": 57.4035,
+      "step": 132220
+    },
+    {
+      "epoch": 0.5342259319561888,
+      "grad_norm": 211.4333038330078,
+      "learning_rate": 2.6837917425231633e-05,
+      "loss": 93.9413,
+      "step": 132230
+    },
+    {
+      "epoch": 0.5342663332215565,
+      "grad_norm": 362.2946472167969,
+      "learning_rate": 2.6834436194592853e-05,
+      "loss": 33.1701,
+      "step": 132240
+    },
+    {
+      "epoch": 0.5343067344869241,
+      "grad_norm": 875.4388427734375,
+      "learning_rate": 2.6830954928190794e-05,
+      "loss": 76.5484,
+      "step": 132250
+    },
+    {
+      "epoch": 0.5343471357522918,
+      "grad_norm": 546.877197265625,
+      "learning_rate": 2.682747362609333e-05,
+      "loss": 38.9788,
+      "step": 132260
+    },
+    {
+      "epoch": 0.5343875370176594,
+      "grad_norm": 674.6320190429688,
+      "learning_rate": 2.6823992288368322e-05,
+      "loss": 70.6211,
+      "step": 132270
+    },
+    {
+      "epoch": 0.534427938283027,
+      "grad_norm": 868.1038818359375,
+      "learning_rate": 2.6820510915083648e-05,
+      "loss": 49.9378,
+      "step": 132280
+    },
+    {
+      "epoch": 0.5344683395483947,
+      "grad_norm": 355.66741943359375,
+      "learning_rate": 2.681702950630717e-05,
+      "loss": 34.1495,
+      "step": 132290
+    },
+    {
+      "epoch": 0.5345087408137623,
+      "grad_norm": 460.1276550292969,
+      "learning_rate": 2.6813548062106775e-05,
+      "loss": 64.0979,
+      "step": 132300
+    },
+    {
+      "epoch": 0.53454914207913,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6810066582550324e-05,
+      "loss": 59.0976,
+      "step": 132310
+    },
+    {
+      "epoch": 0.5345895433444976,
+      "grad_norm": 310.49627685546875,
+      "learning_rate": 2.6806585067705692e-05,
+      "loss": 64.1751,
+      "step": 132320
+    },
+    {
+      "epoch": 0.5346299446098651,
+      "grad_norm": 862.5602416992188,
+      "learning_rate": 2.680310351764075e-05,
+      "loss": 63.3662,
+      "step": 132330
+    },
+    {
+      "epoch": 0.5346703458752328,
+      "grad_norm": 802.3466186523438,
+      "learning_rate": 2.679962193242338e-05,
+      "loss": 71.8971,
+      "step": 132340
+    },
+    {
+      "epoch": 0.5347107471406004,
+      "grad_norm": 1650.000244140625,
+      "learning_rate": 2.6796140312121458e-05,
+      "loss": 59.4168,
+      "step": 132350
+    },
+    {
+      "epoch": 0.534751148405968,
+      "grad_norm": 832.1773071289062,
+      "learning_rate": 2.6792658656802856e-05,
+      "loss": 45.569,
+      "step": 132360
+    },
+    {
+      "epoch": 0.5347915496713357,
+      "grad_norm": 917.97412109375,
+      "learning_rate": 2.6789176966535444e-05,
+      "loss": 45.4459,
+      "step": 132370
+    },
+    {
+      "epoch": 0.5348319509367033,
+      "grad_norm": 583.6998901367188,
+      "learning_rate": 2.678569524138711e-05,
+      "loss": 66.7321,
+      "step": 132380
+    },
+    {
+      "epoch": 0.534872352202071,
+      "grad_norm": 418.53680419921875,
+      "learning_rate": 2.678221348142573e-05,
+      "loss": 34.0593,
+      "step": 132390
+    },
+    {
+      "epoch": 0.5349127534674386,
+      "grad_norm": 402.7685241699219,
+      "learning_rate": 2.6778731686719178e-05,
+      "loss": 55.3928,
+      "step": 132400
+    },
+    {
+      "epoch": 0.5349531547328062,
+      "grad_norm": 372.7942199707031,
+      "learning_rate": 2.6775249857335333e-05,
+      "loss": 30.0867,
+      "step": 132410
+    },
+    {
+      "epoch": 0.5349935559981739,
+      "grad_norm": 1119.2479248046875,
+      "learning_rate": 2.677176799334208e-05,
+      "loss": 58.895,
+      "step": 132420
+    },
+    {
+      "epoch": 0.5350339572635415,
+      "grad_norm": 407.2303771972656,
+      "learning_rate": 2.6768286094807298e-05,
+      "loss": 64.0234,
+      "step": 132430
+    },
+    {
+      "epoch": 0.5350743585289092,
+      "grad_norm": 855.4847412109375,
+      "learning_rate": 2.6764804161798867e-05,
+      "loss": 89.9496,
+      "step": 132440
+    },
+    {
+      "epoch": 0.5351147597942768,
+      "grad_norm": 1015.7791748046875,
+      "learning_rate": 2.6761322194384674e-05,
+      "loss": 60.1903,
+      "step": 132450
+    },
+    {
+      "epoch": 0.5351551610596443,
+      "grad_norm": 1254.9368896484375,
+      "learning_rate": 2.6757840192632598e-05,
+      "loss": 61.4563,
+      "step": 132460
+    },
+    {
+      "epoch": 0.535195562325012,
+      "grad_norm": 299.9359436035156,
+      "learning_rate": 2.6754358156610525e-05,
+      "loss": 38.3524,
+      "step": 132470
+    },
+    {
+      "epoch": 0.5352359635903796,
+      "grad_norm": 811.1012573242188,
+      "learning_rate": 2.6750876086386328e-05,
+      "loss": 47.9699,
+      "step": 132480
+    },
+    {
+      "epoch": 0.5352763648557473,
+      "grad_norm": 893.4951171875,
+      "learning_rate": 2.6747393982027903e-05,
+      "loss": 48.4848,
+      "step": 132490
+    },
+    {
+      "epoch": 0.5353167661211149,
+      "grad_norm": 727.3240966796875,
+      "learning_rate": 2.674391184360313e-05,
+      "loss": 34.6639,
+      "step": 132500
+    },
+    {
+      "epoch": 0.5353571673864825,
+      "grad_norm": 2415.583251953125,
+      "learning_rate": 2.6740429671179907e-05,
+      "loss": 73.8259,
+      "step": 132510
+    },
+    {
+      "epoch": 0.5353975686518502,
+      "grad_norm": 443.5478515625,
+      "learning_rate": 2.6736947464826108e-05,
+      "loss": 48.6447,
+      "step": 132520
+    },
+    {
+      "epoch": 0.5354379699172178,
+      "grad_norm": 1031.6845703125,
+      "learning_rate": 2.6733465224609622e-05,
+      "loss": 54.8037,
+      "step": 132530
+    },
+    {
+      "epoch": 0.5354783711825855,
+      "grad_norm": 564.192138671875,
+      "learning_rate": 2.6729982950598338e-05,
+      "loss": 51.8397,
+      "step": 132540
+    },
+    {
+      "epoch": 0.5355187724479531,
+      "grad_norm": 472.5278625488281,
+      "learning_rate": 2.6726500642860154e-05,
+      "loss": 49.6563,
+      "step": 132550
+    },
+    {
+      "epoch": 0.5355591737133207,
+      "grad_norm": 551.7787475585938,
+      "learning_rate": 2.6723018301462937e-05,
+      "loss": 43.9799,
+      "step": 132560
+    },
+    {
+      "epoch": 0.5355995749786884,
+      "grad_norm": 1193.7701416015625,
+      "learning_rate": 2.67195359264746e-05,
+      "loss": 64.9433,
+      "step": 132570
+    },
+    {
+      "epoch": 0.535639976244056,
+      "grad_norm": 799.9347534179688,
+      "learning_rate": 2.671605351796302e-05,
+      "loss": 41.734,
+      "step": 132580
+    },
+    {
+      "epoch": 0.5356803775094235,
+      "grad_norm": 759.9945068359375,
+      "learning_rate": 2.67125710759961e-05,
+      "loss": 76.4194,
+      "step": 132590
+    },
+    {
+      "epoch": 0.5357207787747912,
+      "grad_norm": 647.0215454101562,
+      "learning_rate": 2.6709088600641717e-05,
+      "loss": 43.6254,
+      "step": 132600
+    },
+    {
+      "epoch": 0.5357611800401588,
+      "grad_norm": 720.9840698242188,
+      "learning_rate": 2.6705606091967778e-05,
+      "loss": 49.2521,
+      "step": 132610
+    },
+    {
+      "epoch": 0.5358015813055265,
+      "grad_norm": 1699.47216796875,
+      "learning_rate": 2.670212355004217e-05,
+      "loss": 48.1515,
+      "step": 132620
+    },
+    {
+      "epoch": 0.5358419825708941,
+      "grad_norm": 912.8600463867188,
+      "learning_rate": 2.6698640974932793e-05,
+      "loss": 60.5469,
+      "step": 132630
+    },
+    {
+      "epoch": 0.5358823838362617,
+      "grad_norm": 562.8499755859375,
+      "learning_rate": 2.6695158366707522e-05,
+      "loss": 53.9552,
+      "step": 132640
+    },
+    {
+      "epoch": 0.5359227851016294,
+      "grad_norm": 778.0060424804688,
+      "learning_rate": 2.6691675725434272e-05,
+      "loss": 52.4513,
+      "step": 132650
+    },
+    {
+      "epoch": 0.535963186366997,
+      "grad_norm": 773.4092407226562,
+      "learning_rate": 2.6688193051180933e-05,
+      "loss": 39.1972,
+      "step": 132660
+    },
+    {
+      "epoch": 0.5360035876323647,
+      "grad_norm": 608.3802490234375,
+      "learning_rate": 2.66847103440154e-05,
+      "loss": 37.7895,
+      "step": 132670
+    },
+    {
+      "epoch": 0.5360439888977323,
+      "grad_norm": 775.6356811523438,
+      "learning_rate": 2.6681227604005576e-05,
+      "loss": 41.3886,
+      "step": 132680
+    },
+    {
+      "epoch": 0.5360843901630999,
+      "grad_norm": 1359.913818359375,
+      "learning_rate": 2.6677744831219348e-05,
+      "loss": 69.0232,
+      "step": 132690
+    },
+    {
+      "epoch": 0.5361247914284676,
+      "grad_norm": 540.3826293945312,
+      "learning_rate": 2.6674262025724627e-05,
+      "loss": 32.8396,
+      "step": 132700
+    },
+    {
+      "epoch": 0.5361651926938352,
+      "grad_norm": 862.3062744140625,
+      "learning_rate": 2.66707791875893e-05,
+      "loss": 64.9376,
+      "step": 132710
+    },
+    {
+      "epoch": 0.5362055939592028,
+      "grad_norm": 557.792236328125,
+      "learning_rate": 2.666729631688128e-05,
+      "loss": 53.6906,
+      "step": 132720
+    },
+    {
+      "epoch": 0.5362459952245704,
+      "grad_norm": 711.6216430664062,
+      "learning_rate": 2.6663813413668455e-05,
+      "loss": 55.813,
+      "step": 132730
+    },
+    {
+      "epoch": 0.536286396489938,
+      "grad_norm": 795.2600708007812,
+      "learning_rate": 2.6660330478018726e-05,
+      "loss": 57.3366,
+      "step": 132740
+    },
+    {
+      "epoch": 0.5363267977553057,
+      "grad_norm": 375.18975830078125,
+      "learning_rate": 2.6656847510000012e-05,
+      "loss": 64.0476,
+      "step": 132750
+    },
+    {
+      "epoch": 0.5363671990206733,
+      "grad_norm": 765.218994140625,
+      "learning_rate": 2.6653364509680188e-05,
+      "loss": 52.8319,
+      "step": 132760
+    },
+    {
+      "epoch": 0.536407600286041,
+      "grad_norm": 609.1416015625,
+      "learning_rate": 2.664988147712718e-05,
+      "loss": 34.5201,
+      "step": 132770
+    },
+    {
+      "epoch": 0.5364480015514086,
+      "grad_norm": 629.1380615234375,
+      "learning_rate": 2.664639841240888e-05,
+      "loss": 30.9847,
+      "step": 132780
+    },
+    {
+      "epoch": 0.5364884028167762,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6642915315593204e-05,
+      "loss": 41.8538,
+      "step": 132790
+    },
+    {
+      "epoch": 0.5365288040821439,
+      "grad_norm": 751.7651977539062,
+      "learning_rate": 2.6639432186748043e-05,
+      "loss": 53.989,
+      "step": 132800
+    },
+    {
+      "epoch": 0.5365692053475115,
+      "grad_norm": 829.0548095703125,
+      "learning_rate": 2.6635949025941303e-05,
+      "loss": 55.0998,
+      "step": 132810
+    },
+    {
+      "epoch": 0.5366096066128792,
+      "grad_norm": 1614.1934814453125,
+      "learning_rate": 2.6632465833240893e-05,
+      "loss": 67.0182,
+      "step": 132820
+    },
+    {
+      "epoch": 0.5366500078782468,
+      "grad_norm": 1317.6927490234375,
+      "learning_rate": 2.662898260871473e-05,
+      "loss": 61.6694,
+      "step": 132830
+    },
+    {
+      "epoch": 0.5366904091436144,
+      "grad_norm": 427.5072937011719,
+      "learning_rate": 2.662549935243071e-05,
+      "loss": 50.9423,
+      "step": 132840
+    },
+    {
+      "epoch": 0.536730810408982,
+      "grad_norm": 1174.3800048828125,
+      "learning_rate": 2.6622016064456738e-05,
+      "loss": 72.6584,
+      "step": 132850
+    },
+    {
+      "epoch": 0.5367712116743496,
+      "grad_norm": 379.9467468261719,
+      "learning_rate": 2.661853274486073e-05,
+      "loss": 64.3494,
+      "step": 132860
+    },
+    {
+      "epoch": 0.5368116129397172,
+      "grad_norm": 1121.682861328125,
+      "learning_rate": 2.6615049393710596e-05,
+      "loss": 87.2399,
+      "step": 132870
+    },
+    {
+      "epoch": 0.5368520142050849,
+      "grad_norm": 1015.9825439453125,
+      "learning_rate": 2.661156601107424e-05,
+      "loss": 76.0801,
+      "step": 132880
+    },
+    {
+      "epoch": 0.5368924154704525,
+      "grad_norm": 364.2757568359375,
+      "learning_rate": 2.660808259701958e-05,
+      "loss": 44.1075,
+      "step": 132890
+    },
+    {
+      "epoch": 0.5369328167358202,
+      "grad_norm": 350.0528259277344,
+      "learning_rate": 2.6604599151614513e-05,
+      "loss": 41.6351,
+      "step": 132900
+    },
+    {
+      "epoch": 0.5369732180011878,
+      "grad_norm": 742.4920654296875,
+      "learning_rate": 2.660111567492696e-05,
+      "loss": 47.2528,
+      "step": 132910
+    },
+    {
+      "epoch": 0.5370136192665554,
+      "grad_norm": 588.64599609375,
+      "learning_rate": 2.6597632167024843e-05,
+      "loss": 49.0769,
+      "step": 132920
+    },
+    {
+      "epoch": 0.5370540205319231,
+      "grad_norm": 519.329345703125,
+      "learning_rate": 2.6594148627976056e-05,
+      "loss": 27.8517,
+      "step": 132930
+    },
+    {
+      "epoch": 0.5370944217972907,
+      "grad_norm": 620.1698608398438,
+      "learning_rate": 2.659066505784852e-05,
+      "loss": 30.6305,
+      "step": 132940
+    },
+    {
+      "epoch": 0.5371348230626584,
+      "grad_norm": 615.6215209960938,
+      "learning_rate": 2.6587181456710153e-05,
+      "loss": 70.2314,
+      "step": 132950
+    },
+    {
+      "epoch": 0.537175224328026,
+      "grad_norm": 766.043212890625,
+      "learning_rate": 2.6583697824628868e-05,
+      "loss": 59.5871,
+      "step": 132960
+    },
+    {
+      "epoch": 0.5372156255933935,
+      "grad_norm": 719.8123168945312,
+      "learning_rate": 2.6580214161672577e-05,
+      "loss": 92.2357,
+      "step": 132970
+    },
+    {
+      "epoch": 0.5372560268587612,
+      "grad_norm": 0.0,
+      "learning_rate": 2.65767304679092e-05,
+      "loss": 56.5093,
+      "step": 132980
+    },
+    {
+      "epoch": 0.5372964281241288,
+      "grad_norm": 615.7906494140625,
+      "learning_rate": 2.6573246743406643e-05,
+      "loss": 82.8496,
+      "step": 132990
+    },
+    {
+      "epoch": 0.5373368293894965,
+      "grad_norm": 377.28985595703125,
+      "learning_rate": 2.656976298823284e-05,
+      "loss": 61.0799,
+      "step": 133000
+    },
+    {
+      "epoch": 0.5373772306548641,
+      "grad_norm": 839.7172241210938,
+      "learning_rate": 2.656627920245569e-05,
+      "loss": 57.0245,
+      "step": 133010
+    },
+    {
+      "epoch": 0.5374176319202317,
+      "grad_norm": 296.6985168457031,
+      "learning_rate": 2.6562795386143126e-05,
+      "loss": 83.4302,
+      "step": 133020
+    },
+    {
+      "epoch": 0.5374580331855994,
+      "grad_norm": 615.6727294921875,
+      "learning_rate": 2.6559311539363057e-05,
+      "loss": 83.2225,
+      "step": 133030
+    },
+    {
+      "epoch": 0.537498434450967,
+      "grad_norm": 518.0780029296875,
+      "learning_rate": 2.6555827662183414e-05,
+      "loss": 61.7653,
+      "step": 133040
+    },
+    {
+      "epoch": 0.5375388357163347,
+      "grad_norm": 585.662841796875,
+      "learning_rate": 2.6552343754672103e-05,
+      "loss": 30.4691,
+      "step": 133050
+    },
+    {
+      "epoch": 0.5375792369817023,
+      "grad_norm": 763.933349609375,
+      "learning_rate": 2.654885981689706e-05,
+      "loss": 47.9403,
+      "step": 133060
+    },
+    {
+      "epoch": 0.5376196382470699,
+      "grad_norm": 4562.193359375,
+      "learning_rate": 2.654537584892619e-05,
+      "loss": 94.3141,
+      "step": 133070
+    },
+    {
+      "epoch": 0.5376600395124376,
+      "grad_norm": 2078.553466796875,
+      "learning_rate": 2.6541891850827427e-05,
+      "loss": 59.5855,
+      "step": 133080
+    },
+    {
+      "epoch": 0.5377004407778052,
+      "grad_norm": 882.0210571289062,
+      "learning_rate": 2.653840782266868e-05,
+      "loss": 48.3455,
+      "step": 133090
+    },
+    {
+      "epoch": 0.5377408420431727,
+      "grad_norm": 643.2105712890625,
+      "learning_rate": 2.653492376451789e-05,
+      "loss": 50.6237,
+      "step": 133100
+    },
+    {
+      "epoch": 0.5377812433085404,
+      "grad_norm": 468.3760986328125,
+      "learning_rate": 2.6531439676442966e-05,
+      "loss": 50.7116,
+      "step": 133110
+    },
+    {
+      "epoch": 0.537821644573908,
+      "grad_norm": 763.0037841796875,
+      "learning_rate": 2.6527955558511842e-05,
+      "loss": 78.7773,
+      "step": 133120
+    },
+    {
+      "epoch": 0.5378620458392757,
+      "grad_norm": 289.5901184082031,
+      "learning_rate": 2.652447141079243e-05,
+      "loss": 46.9934,
+      "step": 133130
+    },
+    {
+      "epoch": 0.5379024471046433,
+      "grad_norm": 776.5145874023438,
+      "learning_rate": 2.6520987233352668e-05,
+      "loss": 65.4828,
+      "step": 133140
+    },
+    {
+      "epoch": 0.5379428483700109,
+      "grad_norm": 2137.632568359375,
+      "learning_rate": 2.6517503026260477e-05,
+      "loss": 51.402,
+      "step": 133150
+    },
+    {
+      "epoch": 0.5379832496353786,
+      "grad_norm": 1670.868896484375,
+      "learning_rate": 2.6514018789583784e-05,
+      "loss": 58.1956,
+      "step": 133160
+    },
+    {
+      "epoch": 0.5380236509007462,
+      "grad_norm": 451.3445739746094,
+      "learning_rate": 2.651053452339051e-05,
+      "loss": 47.8556,
+      "step": 133170
+    },
+    {
+      "epoch": 0.5380640521661139,
+      "grad_norm": 993.9606323242188,
+      "learning_rate": 2.650705022774859e-05,
+      "loss": 53.2049,
+      "step": 133180
+    },
+    {
+      "epoch": 0.5381044534314815,
+      "grad_norm": 836.609130859375,
+      "learning_rate": 2.6503565902725945e-05,
+      "loss": 43.5089,
+      "step": 133190
+    },
+    {
+      "epoch": 0.5381448546968491,
+      "grad_norm": 999.381103515625,
+      "learning_rate": 2.650008154839052e-05,
+      "loss": 61.5361,
+      "step": 133200
+    },
+    {
+      "epoch": 0.5381852559622168,
+      "grad_norm": 806.6781616210938,
+      "learning_rate": 2.6496597164810228e-05,
+      "loss": 82.2622,
+      "step": 133210
+    },
+    {
+      "epoch": 0.5382256572275844,
+      "grad_norm": 576.73974609375,
+      "learning_rate": 2.6493112752053e-05,
+      "loss": 32.6687,
+      "step": 133220
+    },
+    {
+      "epoch": 0.538266058492952,
+      "grad_norm": 362.6260986328125,
+      "learning_rate": 2.6489628310186777e-05,
+      "loss": 27.9188,
+      "step": 133230
+    },
+    {
+      "epoch": 0.5383064597583196,
+      "grad_norm": 422.7489013671875,
+      "learning_rate": 2.6486143839279487e-05,
+      "loss": 54.0248,
+      "step": 133240
+    },
+    {
+      "epoch": 0.5383468610236872,
+      "grad_norm": 578.5451049804688,
+      "learning_rate": 2.6482659339399045e-05,
+      "loss": 60.191,
+      "step": 133250
+    },
+    {
+      "epoch": 0.5383872622890549,
+      "grad_norm": 2537.281005859375,
+      "learning_rate": 2.64791748106134e-05,
+      "loss": 58.6423,
+      "step": 133260
+    },
+    {
+      "epoch": 0.5384276635544225,
+      "grad_norm": 537.8567504882812,
+      "learning_rate": 2.647569025299048e-05,
+      "loss": 39.9818,
+      "step": 133270
+    },
+    {
+      "epoch": 0.5384680648197901,
+      "grad_norm": 415.552001953125,
+      "learning_rate": 2.647220566659822e-05,
+      "loss": 41.9125,
+      "step": 133280
+    },
+    {
+      "epoch": 0.5385084660851578,
+      "grad_norm": 603.8895874023438,
+      "learning_rate": 2.6468721051504554e-05,
+      "loss": 74.081,
+      "step": 133290
+    },
+    {
+      "epoch": 0.5385488673505254,
+      "grad_norm": 918.3076171875,
+      "learning_rate": 2.646523640777741e-05,
+      "loss": 67.8838,
+      "step": 133300
+    },
+    {
+      "epoch": 0.5385892686158931,
+      "grad_norm": 644.3486938476562,
+      "learning_rate": 2.646175173548474e-05,
+      "loss": 39.718,
+      "step": 133310
+    },
+    {
+      "epoch": 0.5386296698812607,
+      "grad_norm": 1127.198486328125,
+      "learning_rate": 2.6458267034694463e-05,
+      "loss": 50.1396,
+      "step": 133320
+    },
+    {
+      "epoch": 0.5386700711466283,
+      "grad_norm": 1032.7508544921875,
+      "learning_rate": 2.645478230547451e-05,
+      "loss": 60.5964,
+      "step": 133330
+    },
+    {
+      "epoch": 0.538710472411996,
+      "grad_norm": 2308.70654296875,
+      "learning_rate": 2.6451297547892834e-05,
+      "loss": 86.4414,
+      "step": 133340
+    },
+    {
+      "epoch": 0.5387508736773636,
+      "grad_norm": 250.09999084472656,
+      "learning_rate": 2.644781276201736e-05,
+      "loss": 38.5809,
+      "step": 133350
+    },
+    {
+      "epoch": 0.5387912749427312,
+      "grad_norm": 892.7601318359375,
+      "learning_rate": 2.6444327947916036e-05,
+      "loss": 54.9328,
+      "step": 133360
+    },
+    {
+      "epoch": 0.5388316762080988,
+      "grad_norm": 742.9862670898438,
+      "learning_rate": 2.6440843105656793e-05,
+      "loss": 53.7666,
+      "step": 133370
+    },
+    {
+      "epoch": 0.5388720774734664,
+      "grad_norm": 580.2630004882812,
+      "learning_rate": 2.6437358235307576e-05,
+      "loss": 42.9191,
+      "step": 133380
+    },
+    {
+      "epoch": 0.5389124787388341,
+      "grad_norm": 657.9290161132812,
+      "learning_rate": 2.643387333693631e-05,
+      "loss": 48.353,
+      "step": 133390
+    },
+    {
+      "epoch": 0.5389528800042017,
+      "grad_norm": 278.8210144042969,
+      "learning_rate": 2.6430388410610955e-05,
+      "loss": 28.9319,
+      "step": 133400
+    },
+    {
+      "epoch": 0.5389932812695694,
+      "grad_norm": 4367.41015625,
+      "learning_rate": 2.6426903456399442e-05,
+      "loss": 80.4687,
+      "step": 133410
+    },
+    {
+      "epoch": 0.539033682534937,
+      "grad_norm": 192.4734649658203,
+      "learning_rate": 2.6423418474369704e-05,
+      "loss": 34.9511,
+      "step": 133420
+    },
+    {
+      "epoch": 0.5390740838003046,
+      "grad_norm": 2244.14892578125,
+      "learning_rate": 2.6419933464589695e-05,
+      "loss": 88.4328,
+      "step": 133430
+    },
+    {
+      "epoch": 0.5391144850656723,
+      "grad_norm": 520.8280639648438,
+      "learning_rate": 2.641644842712735e-05,
+      "loss": 60.3455,
+      "step": 133440
+    },
+    {
+      "epoch": 0.5391548863310399,
+      "grad_norm": 356.8963623046875,
+      "learning_rate": 2.6412963362050618e-05,
+      "loss": 42.0201,
+      "step": 133450
+    },
+    {
+      "epoch": 0.5391952875964076,
+      "grad_norm": 578.19873046875,
+      "learning_rate": 2.640947826942743e-05,
+      "loss": 80.3492,
+      "step": 133460
+    },
+    {
+      "epoch": 0.5392356888617752,
+      "grad_norm": 855.69677734375,
+      "learning_rate": 2.640599314932574e-05,
+      "loss": 64.0816,
+      "step": 133470
+    },
+    {
+      "epoch": 0.5392760901271428,
+      "grad_norm": 1241.7418212890625,
+      "learning_rate": 2.6402508001813496e-05,
+      "loss": 59.2113,
+      "step": 133480
+    },
+    {
+      "epoch": 0.5393164913925104,
+      "grad_norm": 245.85995483398438,
+      "learning_rate": 2.6399022826958635e-05,
+      "loss": 40.4469,
+      "step": 133490
+    },
+    {
+      "epoch": 0.539356892657878,
+      "grad_norm": 1086.880615234375,
+      "learning_rate": 2.6395537624829096e-05,
+      "loss": 47.158,
+      "step": 133500
+    },
+    {
+      "epoch": 0.5393972939232456,
+      "grad_norm": 1283.5257568359375,
+      "learning_rate": 2.639205239549284e-05,
+      "loss": 75.8677,
+      "step": 133510
+    },
+    {
+      "epoch": 0.5394376951886133,
+      "grad_norm": 973.451904296875,
+      "learning_rate": 2.63885671390178e-05,
+      "loss": 60.6037,
+      "step": 133520
+    },
+    {
+      "epoch": 0.5394780964539809,
+      "grad_norm": 743.983642578125,
+      "learning_rate": 2.6385081855471937e-05,
+      "loss": 46.7652,
+      "step": 133530
+    },
+    {
+      "epoch": 0.5395184977193486,
+      "grad_norm": 338.3520812988281,
+      "learning_rate": 2.638159654492318e-05,
+      "loss": 84.5076,
+      "step": 133540
+    },
+    {
+      "epoch": 0.5395588989847162,
+      "grad_norm": 460.87103271484375,
+      "learning_rate": 2.6378111207439494e-05,
+      "loss": 51.0171,
+      "step": 133550
+    },
+    {
+      "epoch": 0.5395993002500838,
+      "grad_norm": 0.0,
+      "learning_rate": 2.637462584308881e-05,
+      "loss": 49.7129,
+      "step": 133560
+    },
+    {
+      "epoch": 0.5396397015154515,
+      "grad_norm": 1022.263427734375,
+      "learning_rate": 2.6371140451939103e-05,
+      "loss": 69.8698,
+      "step": 133570
+    },
+    {
+      "epoch": 0.5396801027808191,
+      "grad_norm": 1827.470947265625,
+      "learning_rate": 2.6367655034058302e-05,
+      "loss": 79.5532,
+      "step": 133580
+    },
+    {
+      "epoch": 0.5397205040461868,
+      "grad_norm": 608.9415893554688,
+      "learning_rate": 2.6364169589514358e-05,
+      "loss": 61.6891,
+      "step": 133590
+    },
+    {
+      "epoch": 0.5397609053115544,
+      "grad_norm": 1445.496826171875,
+      "learning_rate": 2.636068411837523e-05,
+      "loss": 67.8332,
+      "step": 133600
+    },
+    {
+      "epoch": 0.5398013065769219,
+      "grad_norm": 633.453369140625,
+      "learning_rate": 2.6357198620708868e-05,
+      "loss": 41.0135,
+      "step": 133610
+    },
+    {
+      "epoch": 0.5398417078422896,
+      "grad_norm": 1123.755615234375,
+      "learning_rate": 2.635371309658321e-05,
+      "loss": 65.9337,
+      "step": 133620
+    },
+    {
+      "epoch": 0.5398821091076572,
+      "grad_norm": 412.05029296875,
+      "learning_rate": 2.6350227546066218e-05,
+      "loss": 57.8581,
+      "step": 133630
+    },
+    {
+      "epoch": 0.5399225103730249,
+      "grad_norm": 794.2384643554688,
+      "learning_rate": 2.634674196922585e-05,
+      "loss": 47.7789,
+      "step": 133640
+    },
+    {
+      "epoch": 0.5399629116383925,
+      "grad_norm": 839.2406616210938,
+      "learning_rate": 2.6343256366130054e-05,
+      "loss": 49.6203,
+      "step": 133650
+    },
+    {
+      "epoch": 0.5400033129037601,
+      "grad_norm": 474.2958068847656,
+      "learning_rate": 2.633977073684679e-05,
+      "loss": 73.9534,
+      "step": 133660
+    },
+    {
+      "epoch": 0.5400437141691278,
+      "grad_norm": 611.3136596679688,
+      "learning_rate": 2.6336285081443996e-05,
+      "loss": 90.5097,
+      "step": 133670
+    },
+    {
+      "epoch": 0.5400841154344954,
+      "grad_norm": 1800.31884765625,
+      "learning_rate": 2.633279939998964e-05,
+      "loss": 66.3581,
+      "step": 133680
+    },
+    {
+      "epoch": 0.540124516699863,
+      "grad_norm": 437.4395751953125,
+      "learning_rate": 2.6329313692551672e-05,
+      "loss": 54.8113,
+      "step": 133690
+    },
+    {
+      "epoch": 0.5401649179652307,
+      "grad_norm": 891.3457641601562,
+      "learning_rate": 2.6325827959198045e-05,
+      "loss": 59.0138,
+      "step": 133700
+    },
+    {
+      "epoch": 0.5402053192305983,
+      "grad_norm": 971.3360595703125,
+      "learning_rate": 2.6322342199996726e-05,
+      "loss": 42.0446,
+      "step": 133710
+    },
+    {
+      "epoch": 0.540245720495966,
+      "grad_norm": 1470.095703125,
+      "learning_rate": 2.6318856415015664e-05,
+      "loss": 67.5125,
+      "step": 133720
+    },
+    {
+      "epoch": 0.5402861217613336,
+      "grad_norm": 553.43408203125,
+      "learning_rate": 2.631537060432282e-05,
+      "loss": 45.4572,
+      "step": 133730
+    },
+    {
+      "epoch": 0.5403265230267011,
+      "grad_norm": 1151.4344482421875,
+      "learning_rate": 2.631188476798614e-05,
+      "loss": 41.0355,
+      "step": 133740
+    },
+    {
+      "epoch": 0.5403669242920688,
+      "grad_norm": 1145.021728515625,
+      "learning_rate": 2.63083989060736e-05,
+      "loss": 48.7941,
+      "step": 133750
+    },
+    {
+      "epoch": 0.5404073255574364,
+      "grad_norm": 640.4161987304688,
+      "learning_rate": 2.6304913018653144e-05,
+      "loss": 58.756,
+      "step": 133760
+    },
+    {
+      "epoch": 0.5404477268228041,
+      "grad_norm": 887.5857543945312,
+      "learning_rate": 2.630142710579274e-05,
+      "loss": 54.0097,
+      "step": 133770
+    },
+    {
+      "epoch": 0.5404881280881717,
+      "grad_norm": 817.3486328125,
+      "learning_rate": 2.6297941167560346e-05,
+      "loss": 45.1971,
+      "step": 133780
+    },
+    {
+      "epoch": 0.5405285293535393,
+      "grad_norm": 959.50732421875,
+      "learning_rate": 2.6294455204023915e-05,
+      "loss": 43.2953,
+      "step": 133790
+    },
+    {
+      "epoch": 0.540568930618907,
+      "grad_norm": 315.968017578125,
+      "learning_rate": 2.6290969215251416e-05,
+      "loss": 76.2952,
+      "step": 133800
+    },
+    {
+      "epoch": 0.5406093318842746,
+      "grad_norm": 1908.3179931640625,
+      "learning_rate": 2.628748320131081e-05,
+      "loss": 71.9718,
+      "step": 133810
+    },
+    {
+      "epoch": 0.5406497331496423,
+      "grad_norm": 869.3490600585938,
+      "learning_rate": 2.6283997162270052e-05,
+      "loss": 48.5477,
+      "step": 133820
+    },
+    {
+      "epoch": 0.5406901344150099,
+      "grad_norm": 1390.5584716796875,
+      "learning_rate": 2.6280511098197113e-05,
+      "loss": 57.5297,
+      "step": 133830
+    },
+    {
+      "epoch": 0.5407305356803775,
+      "grad_norm": 1634.8409423828125,
+      "learning_rate": 2.627702500915995e-05,
+      "loss": 56.8467,
+      "step": 133840
+    },
+    {
+      "epoch": 0.5407709369457452,
+      "grad_norm": 796.09716796875,
+      "learning_rate": 2.6273538895226522e-05,
+      "loss": 43.6418,
+      "step": 133850
+    },
+    {
+      "epoch": 0.5408113382111128,
+      "grad_norm": 676.2175903320312,
+      "learning_rate": 2.6270052756464803e-05,
+      "loss": 61.2231,
+      "step": 133860
+    },
+    {
+      "epoch": 0.5408517394764804,
+      "grad_norm": 843.5305786132812,
+      "learning_rate": 2.626656659294275e-05,
+      "loss": 39.5511,
+      "step": 133870
+    },
+    {
+      "epoch": 0.540892140741848,
+      "grad_norm": 584.4107055664062,
+      "learning_rate": 2.6263080404728325e-05,
+      "loss": 40.8211,
+      "step": 133880
+    },
+    {
+      "epoch": 0.5409325420072156,
+      "grad_norm": 617.6050415039062,
+      "learning_rate": 2.62595941918895e-05,
+      "loss": 46.0637,
+      "step": 133890
+    },
+    {
+      "epoch": 0.5409729432725833,
+      "grad_norm": 323.1057434082031,
+      "learning_rate": 2.6256107954494242e-05,
+      "loss": 59.272,
+      "step": 133900
+    },
+    {
+      "epoch": 0.5410133445379509,
+      "grad_norm": 534.6497802734375,
+      "learning_rate": 2.6252621692610507e-05,
+      "loss": 47.1891,
+      "step": 133910
+    },
+    {
+      "epoch": 0.5410537458033186,
+      "grad_norm": 1699.9815673828125,
+      "learning_rate": 2.6249135406306273e-05,
+      "loss": 34.6852,
+      "step": 133920
+    },
+    {
+      "epoch": 0.5410941470686862,
+      "grad_norm": 595.4637451171875,
+      "learning_rate": 2.6245649095649494e-05,
+      "loss": 54.051,
+      "step": 133930
+    },
+    {
+      "epoch": 0.5411345483340538,
+      "grad_norm": 538.2552490234375,
+      "learning_rate": 2.6242162760708154e-05,
+      "loss": 35.2021,
+      "step": 133940
+    },
+    {
+      "epoch": 0.5411749495994215,
+      "grad_norm": 1009.6656494140625,
+      "learning_rate": 2.6238676401550207e-05,
+      "loss": 40.208,
+      "step": 133950
+    },
+    {
+      "epoch": 0.5412153508647891,
+      "grad_norm": 941.228515625,
+      "learning_rate": 2.623519001824362e-05,
+      "loss": 48.1016,
+      "step": 133960
+    },
+    {
+      "epoch": 0.5412557521301568,
+      "grad_norm": 884.5780029296875,
+      "learning_rate": 2.6231703610856373e-05,
+      "loss": 57.1634,
+      "step": 133970
+    },
+    {
+      "epoch": 0.5412961533955244,
+      "grad_norm": 670.623291015625,
+      "learning_rate": 2.6228217179456433e-05,
+      "loss": 56.2344,
+      "step": 133980
+    },
+    {
+      "epoch": 0.541336554660892,
+      "grad_norm": 432.82928466796875,
+      "learning_rate": 2.6224730724111758e-05,
+      "loss": 37.2939,
+      "step": 133990
+    },
+    {
+      "epoch": 0.5413769559262596,
+      "grad_norm": 505.16387939453125,
+      "learning_rate": 2.6221244244890336e-05,
+      "loss": 44.5701,
+      "step": 134000
     }
   ],
   "logging_steps": 10,